From d9b2b2a277219d4812311d995054ce4f95067725 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Wed, 13 Feb 2008 16:56:49 -0800
Subject: [LIB]: Make PowerPC LMB code generic so sparc64 can use it too.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/lmb.h | 83 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 83 insertions(+)
 create mode 100644 include/linux/lmb.h

(limited to 'include/linux')

diff --git a/include/linux/lmb.h b/include/linux/lmb.h
new file mode 100644
index 000000000000..8b93f63407e9
--- /dev/null
+++ b/include/linux/lmb.h
@@ -0,0 +1,83 @@
+#ifndef _LINUX_LMB_H
+#define _LINUX_LMB_H
+#ifdef __KERNEL__
+
+/*
+ * Logical memory blocks.
+ *
+ * Copyright (C) 2001 Peter Bergner, IBM Corp.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/init.h>
+#include <linux/mm.h>
+
+#define MAX_LMB_REGIONS 128
+
+struct lmb_property {
+	unsigned long base;
+	unsigned long size;
+};
+
+struct lmb_region {
+	unsigned long cnt;
+	unsigned long size;
+	struct lmb_property region[MAX_LMB_REGIONS+1];
+};
+
+struct lmb {
+	unsigned long debug;
+	unsigned long rmo_size;
+	struct lmb_region memory;
+	struct lmb_region reserved;
+};
+
+extern struct lmb lmb;
+
+extern void __init lmb_init(void);
+extern void __init lmb_analyze(void);
+extern long __init lmb_add(unsigned long base, unsigned long size);
+extern long __init lmb_reserve(unsigned long base, unsigned long size);
+extern unsigned long __init lmb_alloc(unsigned long size, unsigned long align);
+extern unsigned long __init lmb_alloc_base(unsigned long size,
+		unsigned long align, unsigned long max_addr);
+extern unsigned long __init __lmb_alloc_base(unsigned long size,
+		unsigned long align, unsigned long max_addr);
+extern unsigned long __init lmb_phys_mem_size(void);
+extern unsigned long __init lmb_end_of_DRAM(void);
+extern void __init lmb_enforce_memory_limit(unsigned long memory_limit);
+extern int __init lmb_is_reserved(unsigned long addr);
+
+extern void lmb_dump_all(void);
+
+static inline unsigned long
+lmb_size_bytes(struct lmb_region *type, unsigned long region_nr)
+{
+	return type->region[region_nr].size;
+}
+static inline unsigned long
+lmb_size_pages(struct lmb_region *type, unsigned long region_nr)
+{
+	return lmb_size_bytes(type, region_nr) >> PAGE_SHIFT;
+}
+static inline unsigned long
+lmb_start_pfn(struct lmb_region *type, unsigned long region_nr)
+{
+	return type->region[region_nr].base >> PAGE_SHIFT;
+}
+static inline unsigned long
+lmb_end_pfn(struct lmb_region *type, unsigned long region_nr)
+{
+	return lmb_start_pfn(type, region_nr) +
+	       lmb_size_pages(type, region_nr);
+}
+
+#include <asm/lmb.h>
+
+#endif /* __KERNEL__ */
+
+#endif /* _LINUX_LMB_H */
-- 
cgit v1.2.3


From e5f270954364a4add74e8445b1db925ac534fcfb Mon Sep 17 00:00:00 2001
From: Becky Bruce <becky.bruce@freescale.com>
Date: Wed, 13 Feb 2008 16:58:39 -0800
Subject: [LMB]: Make lmb support large physical addressing

Convert the lmb code to use u64 instead of unsigned long for physical
addresses and sizes.  This is needed to support large amounts of RAM
on 32-bit systems that support 36-bit physical addressing.

Signed-off-by: Becky Bruce <becky.bruce@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/lmb.h | 38 +++++++++++-----------
 lib/lmb.c           | 93 ++++++++++++++++++++++++++---------------------------
 2 files changed, 65 insertions(+), 66 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/lmb.h b/include/linux/lmb.h
index 8b93f63407e9..632717c6a2ba 100644
--- a/include/linux/lmb.h
+++ b/include/linux/lmb.h
@@ -19,19 +19,19 @@
 #define MAX_LMB_REGIONS 128
 
 struct lmb_property {
-	unsigned long base;
-	unsigned long size;
+	u64 base;
+	u64 size;
 };
 
 struct lmb_region {
 	unsigned long cnt;
-	unsigned long size;
+	u64 size;
 	struct lmb_property region[MAX_LMB_REGIONS+1];
 };
 
 struct lmb {
 	unsigned long debug;
-	unsigned long rmo_size;
+	u64 rmo_size;
 	struct lmb_region memory;
 	struct lmb_region reserved;
 };
@@ -40,36 +40,36 @@ extern struct lmb lmb;
 
 extern void __init lmb_init(void);
 extern void __init lmb_analyze(void);
-extern long __init lmb_add(unsigned long base, unsigned long size);
-extern long __init lmb_reserve(unsigned long base, unsigned long size);
-extern unsigned long __init lmb_alloc(unsigned long size, unsigned long align);
-extern unsigned long __init lmb_alloc_base(unsigned long size,
-		unsigned long align, unsigned long max_addr);
-extern unsigned long __init __lmb_alloc_base(unsigned long size,
-		unsigned long align, unsigned long max_addr);
-extern unsigned long __init lmb_phys_mem_size(void);
-extern unsigned long __init lmb_end_of_DRAM(void);
-extern void __init lmb_enforce_memory_limit(unsigned long memory_limit);
-extern int __init lmb_is_reserved(unsigned long addr);
+extern long __init lmb_add(u64 base, u64 size);
+extern long __init lmb_reserve(u64 base, u64 size);
+extern u64 __init lmb_alloc(u64 size, u64 align);
+extern u64 __init lmb_alloc_base(u64 size,
+		u64, u64 max_addr);
+extern u64 __init __lmb_alloc_base(u64 size,
+		u64 align, u64 max_addr);
+extern u64 __init lmb_phys_mem_size(void);
+extern u64 __init lmb_end_of_DRAM(void);
+extern void __init lmb_enforce_memory_limit(u64 memory_limit);
+extern int __init lmb_is_reserved(u64 addr);
 
 extern void lmb_dump_all(void);
 
-static inline unsigned long
+static inline u64
 lmb_size_bytes(struct lmb_region *type, unsigned long region_nr)
 {
 	return type->region[region_nr].size;
 }
-static inline unsigned long
+static inline u64
 lmb_size_pages(struct lmb_region *type, unsigned long region_nr)
 {
 	return lmb_size_bytes(type, region_nr) >> PAGE_SHIFT;
 }
-static inline unsigned long
+static inline u64
 lmb_start_pfn(struct lmb_region *type, unsigned long region_nr)
 {
 	return type->region[region_nr].base >> PAGE_SHIFT;
 }
-static inline unsigned long
+static inline u64
 lmb_end_pfn(struct lmb_region *type, unsigned long region_nr)
 {
 	return lmb_start_pfn(type, region_nr) +
diff --git a/lib/lmb.c b/lib/lmb.c
index e34a9e586c42..e3c8dcb04b46 100644
--- a/lib/lmb.c
+++ b/lib/lmb.c
@@ -34,33 +34,34 @@ void lmb_dump_all(void)
 
 	DBG("lmb_dump_all:\n");
 	DBG("    memory.cnt		  = 0x%lx\n", lmb.memory.cnt);
-	DBG("    memory.size		  = 0x%lx\n", lmb.memory.size);
+	DBG("    memory.size		  = 0x%llx\n",
+	    (unsigned long long)lmb.memory.size);
 	for (i=0; i < lmb.memory.cnt ;i++) {
-		DBG("    memory.region[0x%x].base       = 0x%lx\n",
-			    i, lmb.memory.region[i].base);
-		DBG("		      .size     = 0x%lx\n",
-			    lmb.memory.region[i].size);
+		DBG("    memory.region[0x%x].base       = 0x%llx\n",
+		    i, (unsigned long long)lmb.memory.region[i].base);
+		DBG("		      .size     = 0x%llx\n",
+		    (unsigned long long)lmb.memory.region[i].size);
 	}
 
 	DBG("\n    reserved.cnt	  = 0x%lx\n", lmb.reserved.cnt);
 	DBG("    reserved.size	  = 0x%lx\n", lmb.reserved.size);
 	for (i=0; i < lmb.reserved.cnt ;i++) {
-		DBG("    reserved.region[0x%x].base       = 0x%lx\n",
-			    i, lmb.reserved.region[i].base);
-		DBG("		      .size     = 0x%lx\n",
-			    lmb.reserved.region[i].size);
+		DBG("    reserved.region[0x%x].base       = 0x%llx\n",
+		    i, (unsigned long long)lmb.reserved.region[i].base);
+		DBG("		      .size     = 0x%llx\n",
+		    (unsigned long long)lmb.reserved.region[i].size);
 	}
 #endif /* DEBUG */
 }
 
-static unsigned long __init lmb_addrs_overlap(unsigned long base1,
-		unsigned long size1, unsigned long base2, unsigned long size2)
+static unsigned long __init lmb_addrs_overlap(u64 base1,
+		u64 size1, u64 base2, u64 size2)
 {
 	return ((base1 < (base2+size2)) && (base2 < (base1+size1)));
 }
 
-static long __init lmb_addrs_adjacent(unsigned long base1, unsigned long size1,
-		unsigned long base2, unsigned long size2)
+static long __init lmb_addrs_adjacent(u64 base1, u64 size1,
+		u64 base2, u64 size2)
 {
 	if (base2 == base1 + size1)
 		return 1;
@@ -73,10 +74,10 @@ static long __init lmb_addrs_adjacent(unsigned long base1, unsigned long size1,
 static long __init lmb_regions_adjacent(struct lmb_region *rgn,
 		unsigned long r1, unsigned long r2)
 {
-	unsigned long base1 = rgn->region[r1].base;
-	unsigned long size1 = rgn->region[r1].size;
-	unsigned long base2 = rgn->region[r2].base;
-	unsigned long size2 = rgn->region[r2].size;
+	u64 base1 = rgn->region[r1].base;
+	u64 size1 = rgn->region[r1].size;
+	u64 base2 = rgn->region[r2].base;
+	u64 size2 = rgn->region[r2].size;
 
 	return lmb_addrs_adjacent(base1, size1, base2, size2);
 }
@@ -128,8 +129,7 @@ void __init lmb_analyze(void)
 }
 
 /* This routine called with relocation disabled. */
-static long __init lmb_add_region(struct lmb_region *rgn, unsigned long base,
-				  unsigned long size)
+static long __init lmb_add_region(struct lmb_region *rgn, u64 base, u64 size)
 {
 	unsigned long coalesced = 0;
 	long adjacent, i;
@@ -142,8 +142,8 @@ static long __init lmb_add_region(struct lmb_region *rgn, unsigned long base,
 
 	/* First try and coalesce this LMB with another. */
 	for (i=0; i < rgn->cnt; i++) {
-		unsigned long rgnbase = rgn->region[i].base;
-		unsigned long rgnsize = rgn->region[i].size;
+		u64 rgnbase = rgn->region[i].base;
+		u64 rgnsize = rgn->region[i].size;
 
 		if ((rgnbase == base) && (rgnsize == size))
 			/* Already have this region, so we're done */
@@ -190,7 +190,7 @@ static long __init lmb_add_region(struct lmb_region *rgn, unsigned long base,
 }
 
 /* This routine may be called with relocation disabled. */
-long __init lmb_add(unsigned long base, unsigned long size)
+long __init lmb_add(u64 base, u64 size)
 {
 	struct lmb_region *_rgn = &(lmb.memory);
 
@@ -202,7 +202,7 @@ long __init lmb_add(unsigned long base, unsigned long size)
 
 }
 
-long __init lmb_reserve(unsigned long base, unsigned long size)
+long __init lmb_reserve(u64 base, u64 size)
 {
 	struct lmb_region *_rgn = &(lmb.reserved);
 
@@ -211,14 +211,14 @@ long __init lmb_reserve(unsigned long base, unsigned long size)
 	return lmb_add_region(_rgn, base, size);
 }
 
-long __init lmb_overlaps_region(struct lmb_region *rgn, unsigned long base,
-				unsigned long size)
+long __init lmb_overlaps_region(struct lmb_region *rgn, u64 base,
+				u64 size)
 {
 	unsigned long i;
 
 	for (i=0; i < rgn->cnt; i++) {
-		unsigned long rgnbase = rgn->region[i].base;
-		unsigned long rgnsize = rgn->region[i].size;
+		u64 rgnbase = rgn->region[i].base;
+		u64 rgnsize = rgn->region[i].size;
 		if ( lmb_addrs_overlap(base,size,rgnbase,rgnsize) ) {
 			break;
 		}
@@ -227,40 +227,38 @@ long __init lmb_overlaps_region(struct lmb_region *rgn, unsigned long base,
 	return (i < rgn->cnt) ? i : -1;
 }
 
-unsigned long __init lmb_alloc(unsigned long size, unsigned long align)
+u64 __init lmb_alloc(u64 size, u64 align)
 {
 	return lmb_alloc_base(size, align, LMB_ALLOC_ANYWHERE);
 }
 
-unsigned long __init lmb_alloc_base(unsigned long size, unsigned long align,
-				    unsigned long max_addr)
+u64 __init lmb_alloc_base(u64 size, u64 align, u64 max_addr)
 {
-	unsigned long alloc;
+	u64 alloc;
 
 	alloc = __lmb_alloc_base(size, align, max_addr);
 
 	if (alloc == 0)
-		panic("ERROR: Failed to allocate 0x%lx bytes below 0x%lx.\n",
-				size, max_addr);
+		panic("ERROR: Failed to allocate 0x%llx bytes below 0x%llx.\n",
+		      (unsigned long long) size, (unsigned long long) max_addr);
 
 	return alloc;
 }
 
-static unsigned long lmb_align_down(unsigned long addr, unsigned long size)
+static u64 lmb_align_down(u64 addr, u64 size)
 {
 	return addr & ~(size - 1);
 }
 
-static unsigned long lmb_align_up(unsigned long addr, unsigned long size)
+static u64 lmb_align_up(u64 addr, u64 size)
 {
 	return (addr + (size - 1)) & ~(size - 1);
 }
 
-unsigned long __init __lmb_alloc_base(unsigned long size, unsigned long align,
-				    unsigned long max_addr)
+u64 __init __lmb_alloc_base(u64 size, u64 align, u64 max_addr)
 {
 	long i, j;
-	unsigned long base = 0;
+	u64 base = 0;
 
 	BUG_ON(0 == size);
 
@@ -269,8 +267,8 @@ unsigned long __init __lmb_alloc_base(unsigned long size, unsigned long align,
 		max_addr = LMB_REAL_LIMIT;
 
 	for (i = lmb.memory.cnt-1; i >= 0; i--) {
-		unsigned long lmbbase = lmb.memory.region[i].base;
-		unsigned long lmbsize = lmb.memory.region[i].size;
+		u64 lmbbase = lmb.memory.region[i].base;
+		u64 lmbsize = lmb.memory.region[i].size;
 
 		if (max_addr == LMB_ALLOC_ANYWHERE)
 			base = lmb_align_down(lmbbase + lmbsize - size, align);
@@ -299,12 +297,12 @@ unsigned long __init __lmb_alloc_base(unsigned long size, unsigned long align,
 }
 
 /* You must call lmb_analyze() before this. */
-unsigned long __init lmb_phys_mem_size(void)
+u64 __init lmb_phys_mem_size(void)
 {
 	return lmb.memory.size;
 }
 
-unsigned long __init lmb_end_of_DRAM(void)
+u64 __init lmb_end_of_DRAM(void)
 {
 	int idx = lmb.memory.cnt - 1;
 
@@ -312,9 +310,10 @@ unsigned long __init lmb_end_of_DRAM(void)
 }
 
 /* You must call lmb_analyze() after this. */
-void __init lmb_enforce_memory_limit(unsigned long memory_limit)
+void __init lmb_enforce_memory_limit(u64 memory_limit)
 {
-	unsigned long i, limit;
+	unsigned long i;
+	u64 limit;
 	struct lmb_property *p;
 
 	if (! memory_limit)
@@ -352,13 +351,13 @@ void __init lmb_enforce_memory_limit(unsigned long memory_limit)
 	}
 }
 
-int __init lmb_is_reserved(unsigned long addr)
+int __init lmb_is_reserved(u64 addr)
 {
 	int i;
 
 	for (i = 0; i < lmb.reserved.cnt; i++) {
-		unsigned long upper = lmb.reserved.region[i].base +
-				      lmb.reserved.region[i].size - 1;
+		u64 upper = lmb.reserved.region[i].base +
+			lmb.reserved.region[i].size - 1;
 		if ((addr >= lmb.reserved.region[i].base) && (addr <= upper))
 			return 1;
 	}
-- 
cgit v1.2.3


From b35905c16ad6428551eb9e49525011bd2700cf56 Mon Sep 17 00:00:00 2001
From: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Date: Mon, 25 Feb 2008 16:54:37 -0500
Subject: ext4: Fix memory and buffer head leak in callers to
 ext4_ext_find_extent()

The path variable returned via ext4_ext_find_extent is a kmalloc
variable and needs to be freeded.  It also contains a reference to
buffer_head which needs to be dropped.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/extents.c               | 6 +++---
 fs/ext4/migrate.c               | 5 +++++
 include/linux/ext4_fs_extents.h | 1 +
 3 files changed, 9 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index e856f660fc30..995ac16102a9 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -349,7 +349,7 @@ static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
 #define ext4_ext_show_leaf(inode,path)
 #endif
 
-static void ext4_ext_drop_refs(struct ext4_ext_path *path)
+void ext4_ext_drop_refs(struct ext4_ext_path *path)
 {
 	int depth = path->p_depth;
 	int i;
@@ -2200,10 +2200,10 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 		newdepth = ext_depth(inode);
 		if (newdepth != depth) {
 			depth = newdepth;
-			path = ext4_ext_find_extent(inode, iblock, NULL);
+			ext4_ext_drop_refs(path);
+			path = ext4_ext_find_extent(inode, iblock, path);
 			if (IS_ERR(path)) {
 				err = PTR_ERR(path);
-				path = NULL;
 				goto out;
 			}
 			eh = path[depth].p_hdr;
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index 8c6c685b9d22..5c1e27de7755 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -43,6 +43,7 @@ static int finish_range(handle_t *handle, struct inode *inode,
 
 	if (IS_ERR(path)) {
 		retval = PTR_ERR(path);
+		path = NULL;
 		goto err_out;
 	}
 
@@ -74,6 +75,10 @@ static int finish_range(handle_t *handle, struct inode *inode,
 	}
 	retval = ext4_ext_insert_extent(handle, inode, path, &newext);
 err_out:
+	if (path) {
+		ext4_ext_drop_refs(path);
+		kfree(path);
+	}
 	lb->first_pblock = 0;
 	return retval;
 }
diff --git a/include/linux/ext4_fs_extents.h b/include/linux/ext4_fs_extents.h
index 697da4bce6c5..1285c583b2d8 100644
--- a/include/linux/ext4_fs_extents.h
+++ b/include/linux/ext4_fs_extents.h
@@ -227,5 +227,6 @@ extern int ext4_ext_search_left(struct inode *, struct ext4_ext_path *,
 						ext4_lblk_t *, ext4_fsblk_t *);
 extern int ext4_ext_search_right(struct inode *, struct ext4_ext_path *,
 						ext4_lblk_t *, ext4_fsblk_t *);
+extern void ext4_ext_drop_refs(struct ext4_ext_path *);
 #endif /* _LINUX_EXT4_EXTENTS */
 
-- 
cgit v1.2.3


From fea5fa875eb235dc186b1f5184eb36abc63e26cc Mon Sep 17 00:00:00 2001
From: Joe Perches <joe@perches.com>
Date: Sun, 17 Feb 2008 23:35:28 -0800
Subject: [NET]: Remove MAC_FMT

MAC_FMT is no longer used

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if_ether.h | 1 -
 1 file changed, 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/if_ether.h b/include/linux/if_ether.h
index e157c1399b61..7a1e011b8a2c 100644
--- a/include/linux/if_ether.h
+++ b/include/linux/if_ether.h
@@ -130,7 +130,6 @@ extern ssize_t sysfs_format_mac(char *buf, const unsigned char *addr, int len);
  *	Display a 6 byte device address (MAC) in a readable format.
  */
 extern char *print_mac(char *buf, const unsigned char *addr);
-#define MAC_FMT "%02x:%02x:%02x:%02x:%02x:%02x"
 #define MAC_BUF_SIZE	18
 #define DECLARE_MAC_BUF(var) char var[MAC_BUF_SIZE] __maybe_unused
 
-- 
cgit v1.2.3


From 83db33a3e5c4200eaf5dec5b4883c0e73f468a3b Mon Sep 17 00:00:00 2001
From: Jiri Slaby <jirislaby@gmail.com>
Date: Wed, 13 Feb 2008 20:04:49 -0300
Subject: V4L/DVB (7198): V4L, include ioctl.h in videodev headers

Fix compilation of user processes which includes videodev*.h but
not includes linux/ioctl.h:

v4l2ext_helper.c: In function 'process_ioctl':
v4l2ext_helper.c:183: warning: implicit declaration of function '_IOWR'
v4l2ext_helper.c:183: error: expected expression before 'struct'
v4l2ext_helper.c:183: error: case label does not reduce to an integer constant

Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
---
 include/linux/videodev.h  | 1 +
 include/linux/videodev2.h | 1 +
 2 files changed, 2 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/videodev.h b/include/linux/videodev.h
index 52e3d5fd5be4..9385a566aed8 100644
--- a/include/linux/videodev.h
+++ b/include/linux/videodev.h
@@ -12,6 +12,7 @@
 #ifndef __LINUX_VIDEODEV_H
 #define __LINUX_VIDEODEV_H
 
+#include <linux/ioctl.h>
 #include <linux/videodev2.h>
 
 #if defined(CONFIG_VIDEO_V4L1_COMPAT) || !defined (__KERNEL__)
diff --git a/include/linux/videodev2.h b/include/linux/videodev2.h
index 439474f24e34..17a80177a674 100644
--- a/include/linux/videodev2.h
+++ b/include/linux/videodev2.h
@@ -62,6 +62,7 @@
 #define __user
 #include <sys/time.h>
 #endif
+#include <linux/ioctl.h>
 #include <linux/types.h>
 
 /*
-- 
cgit v1.2.3


From 84f7e451e9213d8c328752d0f39bc362519d53d2 Mon Sep 17 00:00:00 2001
From: Andrew Smith <asmith@tranquility.fsbusiness.co.uk>
Date: Tue, 19 Feb 2008 01:41:26 +0100
Subject: via82cxxx: add new PCI id for cx700

[bart: manually ported it over via82cxxx changes]

From: Andrew Smith <asmith@tranquility.fsbusiness.co.uk>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/pci/via82cxxx.c | 1 +
 include/linux/pci_ids.h     | 1 +
 2 files changed, 2 insertions(+)

(limited to 'include/linux')

diff --git a/drivers/ide/pci/via82cxxx.c b/drivers/ide/pci/via82cxxx.c
index f3f79f805813..9004e7521889 100644
--- a/drivers/ide/pci/via82cxxx.c
+++ b/drivers/ide/pci/via82cxxx.c
@@ -479,6 +479,7 @@ static int __devinit via_init_one(struct pci_dev *dev, const struct pci_device_i
 static const struct pci_device_id via_pci_tbl[] = {
 	{ PCI_VDEVICE(VIA, PCI_DEVICE_ID_VIA_82C576_1),  0 },
 	{ PCI_VDEVICE(VIA, PCI_DEVICE_ID_VIA_82C586_1),  0 },
+	{ PCI_VDEVICE(VIA, PCI_DEVICE_ID_VIA_CX700_IDE), 0 },
 	{ PCI_VDEVICE(VIA, PCI_DEVICE_ID_VIA_6410),      1 },
 	{ PCI_VDEVICE(VIA, PCI_DEVICE_ID_VIA_SATA_EIDE), 1 },
 	{ 0, },
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index df6dd79a0d3b..0519e53d5eb2 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -1351,6 +1351,7 @@
 #define PCI_DEVICE_ID_VIA_8231_4	0x8235
 #define PCI_DEVICE_ID_VIA_8365_1	0x8305
 #define PCI_DEVICE_ID_VIA_CX700		0x8324
+#define PCI_DEVICE_ID_VIA_CX700_IDE	0x0581
 #define PCI_DEVICE_ID_VIA_VX800		0x8353
 #define PCI_DEVICE_ID_VIA_8371_1	0x8391
 #define PCI_DEVICE_ID_VIA_82C598_1	0x8598
-- 
cgit v1.2.3


From f401fd55b5a9ef1a4572abaefaa6957af11ced61 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Tue, 19 Feb 2008 01:41:26 +0100
Subject: linux/hdsmart.h: fix goofups (take 2)

Fix goofups of commit 76166952bbc81dda1c8a8c14e75a2aa06f6c052c
("<linux/hdsmart.h> is not used by kernel code").

Also update include/linux/Kbuild to reflect the fact that hdsmart.h
uses __KERNEL__ ifdefs now.

Reported-by: "Robert P. J. Day" <rpjday@crashcourse.ca>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 include/linux/Kbuild    | 2 +-
 include/linux/hdsmart.h | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 5cae9b5960ea..aada32fffec2 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -70,7 +70,6 @@ header-y += fuse.h
 header-y += genetlink.h
 header-y += gen_stats.h
 header-y += gigaset_dev.h
-header-y += hdsmart.h
 header-y += hysdn_if.h
 header-y += i2o-dev.h
 header-y += i8k.h
@@ -211,6 +210,7 @@ unifdef-y += hayesesp.h
 unifdef-y += hdlcdrv.h
 unifdef-y += hdlc.h
 unifdef-y += hdreg.h
+unifdef-y += hdsmart.h
 unifdef-y += hiddev.h
 unifdef-y += hpet.h
 unifdef-y += i2c.h
diff --git a/include/linux/hdsmart.h b/include/linux/hdsmart.h
index e69192159d40..4f4faf9d4238 100644
--- a/include/linux/hdsmart.h
+++ b/include/linux/hdsmart.h
@@ -17,7 +17,7 @@
 #ifndef _LINUX_HDSMART_H
 #define _LINUX_HDSMART_H
 
-#ifndef __KERNEL
+#ifndef __KERNEL__
 #define OFFLINE_FULL_SCAN		0
 #define SHORT_SELF_TEST			1
 #define EXTEND_SELF_TEST		2
@@ -121,6 +121,6 @@ typedef struct ata_smart_selftestlog_s {
 	unsigned char			resevered[2];
 	unsigned char			chksum;
 } __attribute__ ((packed)) ata_smart_selftestlog_t;
-#endif /* __KERNEL__ *
+#endif /* __KERNEL__ */
 
 #endif	/* _LINUX_HDSMART_H */
-- 
cgit v1.2.3


From 8f789c48448aed74fe1c07af76de8f04adacec7d Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Mon, 18 Feb 2008 16:50:22 -0800
Subject: [NET]: Elminate spurious print_mac() calls.

Patrick McHardy notes that print_mac() can get invoked
even if the result it unused (f.e. as an argument to
pr_debug() when DEBUG is not defined).

Mark this function as "__pure" to eliminate this problem.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if_ether.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/if_ether.h b/include/linux/if_ether.h
index 7a1e011b8a2c..42dc6a3571ec 100644
--- a/include/linux/if_ether.h
+++ b/include/linux/if_ether.h
@@ -129,7 +129,7 @@ extern ssize_t sysfs_format_mac(char *buf, const unsigned char *addr, int len);
 /*
  *	Display a 6 byte device address (MAC) in a readable format.
  */
-extern char *print_mac(char *buf, const unsigned char *addr);
+extern __pure char *print_mac(char *buf, const unsigned char *addr);
 #define MAC_BUF_SIZE	18
 #define DECLARE_MAC_BUF(var) char var[MAC_BUF_SIZE] __maybe_unused
 
-- 
cgit v1.2.3


From 31729363418ea25b01aa9410838c38e36792e44c Mon Sep 17 00:00:00 2001
From: Randy Dunlap <randy.dunlap@oracle.com>
Date: Mon, 18 Feb 2008 20:52:13 -0800
Subject: net: fix kernel-doc warnings in header files

Add missing structure kernel-doc descriptions to sock.h & skbuff.h
to fix kernel-doc warnings.

(I think that Stephen H. sent a similar patch, but I can't find it.
I just want to kill the warnings, with either patch.)

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 2 ++
 include/net/sock.h     | 1 +
 2 files changed, 3 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 412672a79e8a..bbd8d0027e2f 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -232,6 +232,8 @@ typedef unsigned char *sk_buff_data_t;
  *	@mark: Generic packet mark
  *	@nfct: Associated connection, if any
  *	@ipvs_property: skbuff is owned by ipvs
+ *	@peeked: this packet has been seen already, so stats have been
+ *		done for it, don't do them again
  *	@nf_trace: netfilter packet trace flag
  *	@nfctinfo: Relationship of this skb to the connection
  *	@nfct_reasm: netfilter conntrack re-assembly pointer
diff --git a/include/net/sock.h b/include/net/sock.h
index 8a7889b35810..fd9876087651 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -180,6 +180,7 @@ struct sock_common {
   *	@sk_sndmsg_off: cached offset for sendmsg
   *	@sk_send_head: front of stuff to transmit
   *	@sk_security: used by security modules
+  *	@sk_mark: generic packet mark
   *	@sk_write_pending: a write to stream socket waits to start
   *	@sk_state_change: callback to indicate change in the state of the sock
   *	@sk_data_ready: callback to indicate there is data to be processed
-- 
cgit v1.2.3


From 4c54ac62dceecedd82d4a865017bba0b738e2897 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Mon, 18 Feb 2008 13:48:31 +0100
Subject: make struct def_blk_aops static

This patch makes the needlessly global struct def_blk_aops static.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <axboe@carl.home.kernel.dk>
---
 fs/block_dev.c     | 4 +++-
 include/linux/fs.h | 1 -
 2 files changed, 3 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/fs/block_dev.c b/fs/block_dev.c
index 67fe72ce6ac7..8335f0e1b0fb 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -31,6 +31,8 @@ struct bdev_inode {
 	struct inode vfs_inode;
 };
 
+static const struct address_space_operations def_blk_aops;
+
 static inline struct bdev_inode *BDEV_I(struct inode *inode)
 {
 	return container_of(inode, struct bdev_inode, vfs_inode);
@@ -1334,7 +1336,7 @@ static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg)
 	return blkdev_ioctl(file->f_mapping->host, file, cmd, arg);
 }
 
-const struct address_space_operations def_blk_aops = {
+static const struct address_space_operations def_blk_aops = {
 	.readpage	= blkdev_readpage,
 	.writepage	= blkdev_writepage,
 	.sync_page	= block_sync_page,
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 98ffb6ead434..b84b848431f2 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1590,7 +1590,6 @@ extern void bd_set_size(struct block_device *, loff_t size);
 extern void bd_forget(struct inode *inode);
 extern void bdput(struct block_device *);
 extern struct block_device *open_by_devnum(dev_t, unsigned);
-extern const struct address_space_operations def_blk_aops;
 #else
 static inline void bd_forget(struct inode *inode) {}
 #endif
-- 
cgit v1.2.3


From 86b6c7a7f78feca58d2d8615e53aee4d59ab9dc6 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Mon, 18 Feb 2008 13:48:32 +0100
Subject: fs/block_dev.c: remove #if 0'ed code

Commit b2e895dbd80c420bfc0937c3729b4afe073b3848 #if 0'ed this code stating:

<--  snip  -->

    [PATCH] revert blockdev direct io back to 2.6.19 version

    Andrew Vasquez is reporting as-iosched oopses and a 65% throughput
    slowdown due to the recent special-casing of direct-io against
    blockdevs.  We don't know why either of these things are occurring.

    The patch minimally reverts us back to the 2.6.19 code for a 2.6.20
    release.

<--  snip  -->

It has since been dead code, and unless someone wants to revive it now
it's time to remove it.

This patch also makes bio_release_pages() static again and removes the
ki_bio_count member from struct kiocb, reverting changes that had been
done for this dead code.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <axboe@carl.home.kernel.dk>
---
 fs/bio.c            |   2 +-
 fs/block_dev.c      | 197 ----------------------------------------------------
 include/linux/aio.h |   1 -
 include/linux/bio.h |   1 -
 4 files changed, 1 insertion(+), 200 deletions(-)

(limited to 'include/linux')

diff --git a/fs/bio.c b/fs/bio.c
index 242e409dab4b..3312fcc3c098 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -903,7 +903,7 @@ void bio_set_pages_dirty(struct bio *bio)
 	}
 }
 
-void bio_release_pages(struct bio *bio)
+static void bio_release_pages(struct bio *bio)
 {
 	struct bio_vec *bvec = bio->bi_io_vec;
 	int i;
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 8335f0e1b0fb..7d822fae7765 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -173,203 +173,6 @@ blkdev_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
 				iov, offset, nr_segs, blkdev_get_blocks, NULL);
 }
 
-#if 0
-static void blk_end_aio(struct bio *bio, int error)
-{
-	struct kiocb *iocb = bio->bi_private;
-	atomic_t *bio_count = &iocb->ki_bio_count;
-
-	if (bio_data_dir(bio) == READ)
-		bio_check_pages_dirty(bio);
-	else {
-		bio_release_pages(bio);
-		bio_put(bio);
-	}
-
-	/* iocb->ki_nbytes stores error code from LLDD */
-	if (error)
-		iocb->ki_nbytes = -EIO;
-
-	if (atomic_dec_and_test(bio_count)) {
-		if ((long)iocb->ki_nbytes < 0)
-			aio_complete(iocb, iocb->ki_nbytes, 0);
-		else
-			aio_complete(iocb, iocb->ki_left, 0);
-	}
-
-	return 0;
-}
-
-#define VEC_SIZE	16
-struct pvec {
-	unsigned short nr;
-	unsigned short idx;
-	struct page *page[VEC_SIZE];
-};
-
-#define PAGES_SPANNED(addr, len)	\
-	(DIV_ROUND_UP((addr) + (len), PAGE_SIZE) - (addr) / PAGE_SIZE);
-
-/*
- * get page pointer for user addr, we internally cache struct page array for
- * (addr, count) range in pvec to avoid frequent call to get_user_pages.  If
- * internal page list is exhausted, a batch count of up to VEC_SIZE is used
- * to get next set of page struct.
- */
-static struct page *blk_get_page(unsigned long addr, size_t count, int rw,
-				 struct pvec *pvec)
-{
-	int ret, nr_pages;
-	if (pvec->idx == pvec->nr) {
-		nr_pages = PAGES_SPANNED(addr, count);
-		nr_pages = min(nr_pages, VEC_SIZE);
-		down_read(&current->mm->mmap_sem);
-		ret = get_user_pages(current, current->mm, addr, nr_pages,
-				     rw == READ, 0, pvec->page, NULL);
-		up_read(&current->mm->mmap_sem);
-		if (ret < 0)
-			return ERR_PTR(ret);
-		pvec->nr = ret;
-		pvec->idx = 0;
-	}
-	return pvec->page[pvec->idx++];
-}
-
-/* return a page back to pvec array */
-static void blk_unget_page(struct page *page, struct pvec *pvec)
-{
-	pvec->page[--pvec->idx] = page;
-}
-
-static ssize_t
-blkdev_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
-		 loff_t pos, unsigned long nr_segs)
-{
-	struct inode *inode = iocb->ki_filp->f_mapping->host;
-	unsigned blkbits = blksize_bits(bdev_hardsect_size(I_BDEV(inode)));
-	unsigned blocksize_mask = (1 << blkbits) - 1;
-	unsigned long seg = 0;	/* iov segment iterator */
-	unsigned long nvec;	/* number of bio vec needed */
-	unsigned long cur_off;	/* offset into current page */
-	unsigned long cur_len;	/* I/O len of current page, up to PAGE_SIZE */
-
-	unsigned long addr;	/* user iovec address */
-	size_t count;		/* user iovec len */
-	size_t nbytes = iocb->ki_nbytes = iocb->ki_left; /* total xfer size */
-	loff_t size;		/* size of block device */
-	struct bio *bio;
-	atomic_t *bio_count = &iocb->ki_bio_count;
-	struct page *page;
-	struct pvec pvec;
-
-	pvec.nr = 0;
-	pvec.idx = 0;
-
-	if (pos & blocksize_mask)
-		return -EINVAL;
-
-	size = i_size_read(inode);
-	if (pos + nbytes > size) {
-		nbytes = size - pos;
-		iocb->ki_left = nbytes;
-	}
-
-	/*
-	 * check first non-zero iov alignment, the remaining
-	 * iov alignment is checked inside bio loop below.
-	 */
-	do {
-		addr = (unsigned long) iov[seg].iov_base;
-		count = min(iov[seg].iov_len, nbytes);
-		if (addr & blocksize_mask || count & blocksize_mask)
-			return -EINVAL;
-	} while (!count && ++seg < nr_segs);
-	atomic_set(bio_count, 1);
-
-	while (nbytes) {
-		/* roughly estimate number of bio vec needed */
-		nvec = (nbytes + PAGE_SIZE - 1) / PAGE_SIZE;
-		nvec = max(nvec, nr_segs - seg);
-		nvec = min(nvec, (unsigned long) BIO_MAX_PAGES);
-
-		/* bio_alloc should not fail with GFP_KERNEL flag */
-		bio = bio_alloc(GFP_KERNEL, nvec);
-		bio->bi_bdev = I_BDEV(inode);
-		bio->bi_end_io = blk_end_aio;
-		bio->bi_private = iocb;
-		bio->bi_sector = pos >> blkbits;
-same_bio:
-		cur_off = addr & ~PAGE_MASK;
-		cur_len = PAGE_SIZE - cur_off;
-		if (count < cur_len)
-			cur_len = count;
-
-		page = blk_get_page(addr, count, rw, &pvec);
-		if (unlikely(IS_ERR(page)))
-			goto backout;
-
-		if (bio_add_page(bio, page, cur_len, cur_off)) {
-			pos += cur_len;
-			addr += cur_len;
-			count -= cur_len;
-			nbytes -= cur_len;
-
-			if (count)
-				goto same_bio;
-			while (++seg < nr_segs) {
-				addr = (unsigned long) iov[seg].iov_base;
-				count = iov[seg].iov_len;
-				if (!count)
-					continue;
-				if (unlikely(addr & blocksize_mask ||
-					     count & blocksize_mask)) {
-					page = ERR_PTR(-EINVAL);
-					goto backout;
-				}
-				count = min(count, nbytes);
-				goto same_bio;
-			}
-		} else {
-			blk_unget_page(page, &pvec);
-		}
-
-		/* bio is ready, submit it */
-		if (rw == READ)
-			bio_set_pages_dirty(bio);
-		atomic_inc(bio_count);
-		submit_bio(rw, bio);
-	}
-
-completion:
-	iocb->ki_left -= nbytes;
-	nbytes = iocb->ki_left;
-	iocb->ki_pos += nbytes;
-
-	blk_run_address_space(inode->i_mapping);
-	if (atomic_dec_and_test(bio_count))
-		aio_complete(iocb, nbytes, 0);
-
-	return -EIOCBQUEUED;
-
-backout:
-	/*
-	 * back out nbytes count constructed so far for this bio,
-	 * we will throw away current bio.
-	 */
-	nbytes += bio->bi_size;
-	bio_release_pages(bio);
-	bio_put(bio);
-
-	/*
-	 * if no bio was submmitted, return the error code.
-	 * otherwise, proceed with pending I/O completion.
-	 */
-	if (atomic_read(bio_count) == 1)
-		return PTR_ERR(page);
-	goto completion;
-}
-#endif
-
 static int blkdev_writepage(struct page *page, struct writeback_control *wbc)
 {
 	return block_write_full_page(page, blkdev_get_block, wbc);
diff --git a/include/linux/aio.h b/include/linux/aio.h
index a9931e2e5624..0d0b7f629bd3 100644
--- a/include/linux/aio.h
+++ b/include/linux/aio.h
@@ -105,7 +105,6 @@ struct kiocb {
 	wait_queue_t		ki_wait;
 	loff_t			ki_pos;
 
-	atomic_t		ki_bio_count;	/* num bio used for this iocb */
 	void			*private;
 	/* State that we remember to be able to restart/retry  */
 	unsigned short		ki_opcode;
diff --git a/include/linux/bio.h b/include/linux/bio.h
index 4da441337d6e..4c59bdccd3ee 100644
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -326,7 +326,6 @@ extern struct bio *bio_map_kern(struct request_queue *, void *, unsigned int,
 				gfp_t);
 extern void bio_set_pages_dirty(struct bio *bio);
 extern void bio_check_pages_dirty(struct bio *bio);
-extern void bio_release_pages(struct bio *bio);
 extern struct bio *bio_copy_user(struct request_queue *, unsigned long, unsigned int, int);
 extern int bio_uncopy_user(struct bio *);
 void zero_fill_bio(struct bio *bio);
-- 
cgit v1.2.3


From ffc4e7595734cf768fa60cea8a4d545dfef8231a Mon Sep 17 00:00:00 2001
From: Jens Axboe <jens.axboe@oracle.com>
Date: Tue, 19 Feb 2008 10:02:29 +0100
Subject: cfq-iosched: add hlist for browsing parallel to the radix tree

It's cumbersome to browse a radix tree from start to finish, especially
since we modify keys when a process exits. So add a hlist for the single
purpose of browsing over all known cfq_io_contexts, used for exit,
io prio change, etc.

This fixes http://bugzilla.kernel.org/show_bug.cgi?id=9948

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 block/blk-ioc.c           | 35 +++++++++++++++--------------------
 block/cfq-iosched.c       | 38 ++++++++++++--------------------------
 include/linux/iocontext.h |  2 ++
 3 files changed, 29 insertions(+), 46 deletions(-)

(limited to 'include/linux')

diff --git a/block/blk-ioc.c b/block/blk-ioc.c
index 4ae0929c6e38..e34df7c9fc36 100644
--- a/block/blk-ioc.c
+++ b/block/blk-ioc.c
@@ -17,17 +17,13 @@ static struct kmem_cache *iocontext_cachep;
 
 static void cfq_dtor(struct io_context *ioc)
 {
-	struct cfq_io_context *cic[1];
-	int r;
+	if (!hlist_empty(&ioc->cic_list)) {
+		struct cfq_io_context *cic;
 
-	/*
-	 * We don't have a specific key to lookup with, so use the gang
-	 * lookup to just retrieve the first item stored. The cfq exit
-	 * function will iterate the full tree, so any member will do.
-	 */
-	r = radix_tree_gang_lookup(&ioc->radix_root, (void **) cic, 0, 1);
-	if (r > 0)
-		cic[0]->dtor(ioc);
+		cic = list_entry(ioc->cic_list.first, struct cfq_io_context,
+								cic_list);
+		cic->dtor(ioc);
+	}
 }
 
 /*
@@ -57,18 +53,16 @@ EXPORT_SYMBOL(put_io_context);
 
 static void cfq_exit(struct io_context *ioc)
 {
-	struct cfq_io_context *cic[1];
-	int r;
-
 	rcu_read_lock();
-	/*
-	 * See comment for cfq_dtor()
-	 */
-	r = radix_tree_gang_lookup(&ioc->radix_root, (void **) cic, 0, 1);
-	rcu_read_unlock();
 
-	if (r > 0)
-		cic[0]->exit(ioc);
+	if (!hlist_empty(&ioc->cic_list)) {
+		struct cfq_io_context *cic;
+
+		cic = list_entry(ioc->cic_list.first, struct cfq_io_context,
+								cic_list);
+		cic->exit(ioc);
+	}
+	rcu_read_unlock();
 }
 
 /* Called by the exitting task */
@@ -105,6 +99,7 @@ struct io_context *alloc_io_context(gfp_t gfp_flags, int node)
 		ret->nr_batch_requests = 0; /* because this is 0 */
 		ret->aic = NULL;
 		INIT_RADIX_TREE(&ret->radix_root, GFP_ATOMIC | __GFP_HIGH);
+		INIT_HLIST_HEAD(&ret->cic_list);
 		ret->ioc_data = NULL;
 	}
 
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index ca198e61fa65..0f962ecae91f 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -1145,38 +1145,19 @@ static void cfq_put_queue(struct cfq_queue *cfqq)
 /*
  * Call func for each cic attached to this ioc. Returns number of cic's seen.
  */
-#define CIC_GANG_NR	16
 static unsigned int
 call_for_each_cic(struct io_context *ioc,
 		  void (*func)(struct io_context *, struct cfq_io_context *))
 {
-	struct cfq_io_context *cics[CIC_GANG_NR];
-	unsigned long index = 0;
-	unsigned int called = 0;
-	int nr;
+	struct cfq_io_context *cic;
+	struct hlist_node *n;
+	int called = 0;
 
 	rcu_read_lock();
-
-	do {
-		int i;
-
-		/*
-		 * Perhaps there's a better way - this just gang lookups from
-		 * 0 to the end, restarting after each CIC_GANG_NR from the
-		 * last key + 1.
-		 */
-		nr = radix_tree_gang_lookup(&ioc->radix_root, (void **) cics,
-						index, CIC_GANG_NR);
-		if (!nr)
-			break;
-
-		called += nr;
-		index = 1 + (unsigned long) cics[nr - 1]->key;
-
-		for (i = 0; i < nr; i++)
-			func(ioc, cics[i]);
-	} while (nr == CIC_GANG_NR);
-
+	hlist_for_each_entry_rcu(cic, n, &ioc->cic_list, cic_list) {
+		func(ioc, cic);
+		called++;
+	}
 	rcu_read_unlock();
 
 	return called;
@@ -1190,6 +1171,7 @@ static void cic_free_func(struct io_context *ioc, struct cfq_io_context *cic)
 
 	spin_lock_irqsave(&ioc->lock, flags);
 	radix_tree_delete(&ioc->radix_root, cic->dead_key);
+	hlist_del_rcu(&cic->cic_list);
 	spin_unlock_irqrestore(&ioc->lock, flags);
 
 	kmem_cache_free(cfq_ioc_pool, cic);
@@ -1280,6 +1262,7 @@ cfq_alloc_io_context(struct cfq_data *cfqd, gfp_t gfp_mask)
 	if (cic) {
 		cic->last_end_request = jiffies;
 		INIT_LIST_HEAD(&cic->queue_list);
+		INIT_HLIST_NODE(&cic->cic_list);
 		cic->dtor = cfq_free_io_context;
 		cic->exit = cfq_exit_io_context;
 		elv_ioc_count_inc(ioc_count);
@@ -1501,6 +1484,7 @@ cfq_drop_dead_cic(struct cfq_data *cfqd, struct io_context *ioc,
 		rcu_assign_pointer(ioc->ioc_data, NULL);
 
 	radix_tree_delete(&ioc->radix_root, (unsigned long) cfqd);
+	hlist_del_rcu(&cic->cic_list);
 	spin_unlock_irqrestore(&ioc->lock, flags);
 
 	cfq_cic_free(cic);
@@ -1561,6 +1545,8 @@ static int cfq_cic_link(struct cfq_data *cfqd, struct io_context *ioc,
 		spin_lock_irqsave(&ioc->lock, flags);
 		ret = radix_tree_insert(&ioc->radix_root,
 						(unsigned long) cfqd, cic);
+		if (!ret)
+			hlist_add_head_rcu(&cic->cic_list, &ioc->cic_list);
 		spin_unlock_irqrestore(&ioc->lock, flags);
 
 		radix_tree_preload_end();
diff --git a/include/linux/iocontext.h b/include/linux/iocontext.h
index 593b222d9dcc..1b4ccf25b4d2 100644
--- a/include/linux/iocontext.h
+++ b/include/linux/iocontext.h
@@ -50,6 +50,7 @@ struct cfq_io_context {
 	sector_t seek_mean;
 
 	struct list_head queue_list;
+	struct hlist_node cic_list;
 
 	void (*dtor)(struct io_context *); /* destructor */
 	void (*exit)(struct io_context *); /* called on task exit */
@@ -77,6 +78,7 @@ struct io_context {
 
 	struct as_io_context *aic;
 	struct radix_tree_root radix_root;
+	struct hlist_head cic_list;
 	void *ioc_data;
 };
 
-- 
cgit v1.2.3


From 6b00769fe1502b4ad97bb327ef7ac971b208bfb5 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 19 Feb 2008 11:36:35 +0100
Subject: block: add request->raw_data_len

With padding and draining moved into it, block layer now may extend
requests as directed by queue parameters, so now a request has two
sizes - the original request size and the extended size which matches
the size of area pointed to by bios and later by sgs.  The latter size
is what lower layers are primarily interested in when allocating,
filling up DMA tables and setting up the controller.

Both padding and draining extend the data area to accomodate
controller characteristics.  As any controller which speaks SCSI can
handle underflows, feeding larger data area is safe.

So, this patch makes the primary data length field, request->data_len,
indicate the size of full data area and add a separate length field,
request->raw_data_len, for the unmodified request size.  The latter is
used to report to higher layer (userland) and where the original
request size should be fed to the controller or device.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 block/blk-core.c        | 2 ++
 block/blk-map.c         | 2 ++
 block/blk-merge.c       | 1 +
 block/bsg.c             | 8 ++++----
 block/scsi_ioctl.c      | 3 ++-
 drivers/scsi/scsi_lib.c | 8 ++++----
 include/linux/blkdev.h  | 1 +
 7 files changed, 16 insertions(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/block/blk-core.c b/block/blk-core.c
index c013ca22eb67..775c8516abf5 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -127,6 +127,7 @@ void rq_init(struct request_queue *q, struct request *rq)
 	rq->nr_hw_segments = 0;
 	rq->ioprio = 0;
 	rq->special = NULL;
+	rq->raw_data_len = 0;
 	rq->buffer = NULL;
 	rq->tag = -1;
 	rq->errors = 0;
@@ -2015,6 +2016,7 @@ void blk_rq_bio_prep(struct request_queue *q, struct request *rq,
 	rq->hard_cur_sectors = rq->current_nr_sectors;
 	rq->hard_nr_sectors = rq->nr_sectors = bio_sectors(bio);
 	rq->buffer = bio_data(bio);
+	rq->raw_data_len = bio->bi_size;
 	rq->data_len = bio->bi_size;
 
 	rq->bio = rq->biotail = bio;
diff --git a/block/blk-map.c b/block/blk-map.c
index a7cf63ccb5cf..09f7fd0bcb73 100644
--- a/block/blk-map.c
+++ b/block/blk-map.c
@@ -19,6 +19,7 @@ int blk_rq_append_bio(struct request_queue *q, struct request *rq,
 		rq->biotail->bi_next = bio;
 		rq->biotail = bio;
 
+		rq->raw_data_len += bio->bi_size;
 		rq->data_len += bio->bi_size;
 	}
 	return 0;
@@ -154,6 +155,7 @@ int blk_rq_map_user(struct request_queue *q, struct request *rq,
 
 		bio->bi_io_vec[bio->bi_vcnt - 1].bv_len += pad_len;
 		bio->bi_size += pad_len;
+		rq->data_len += pad_len;
 	}
 
 	rq->buffer = rq->data = NULL;
diff --git a/block/blk-merge.c b/block/blk-merge.c
index d3b84bbb776a..39f2e077a014 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -228,6 +228,7 @@ new_segment:
 			    ((unsigned long)q->dma_drain_buffer) &
 			    (PAGE_SIZE - 1));
 		nsegs++;
+		rq->data_len += q->dma_drain_size;
 	}
 
 	if (sg)
diff --git a/block/bsg.c b/block/bsg.c
index 8917c5174dc2..7f3c09549e4b 100644
--- a/block/bsg.c
+++ b/block/bsg.c
@@ -437,14 +437,14 @@ static int blk_complete_sgv4_hdr_rq(struct request *rq, struct sg_io_v4 *hdr,
 	}
 
 	if (rq->next_rq) {
-		hdr->dout_resid = rq->data_len;
-		hdr->din_resid = rq->next_rq->data_len;
+		hdr->dout_resid = rq->raw_data_len;
+		hdr->din_resid = rq->next_rq->raw_data_len;
 		blk_rq_unmap_user(bidi_bio);
 		blk_put_request(rq->next_rq);
 	} else if (rq_data_dir(rq) == READ)
-		hdr->din_resid = rq->data_len;
+		hdr->din_resid = rq->raw_data_len;
 	else
-		hdr->dout_resid = rq->data_len;
+		hdr->dout_resid = rq->raw_data_len;
 
 	/*
 	 * If the request generated a negative error number, return it
diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c
index 9675b34638d4..e993cac4911d 100644
--- a/block/scsi_ioctl.c
+++ b/block/scsi_ioctl.c
@@ -266,7 +266,7 @@ static int blk_complete_sghdr_rq(struct request *rq, struct sg_io_hdr *hdr,
 	hdr->info = 0;
 	if (hdr->masked_status || hdr->host_status || hdr->driver_status)
 		hdr->info |= SG_INFO_CHECK;
-	hdr->resid = rq->data_len;
+	hdr->resid = rq->raw_data_len;
 	hdr->sb_len_wr = 0;
 
 	if (rq->sense_len && hdr->sbp) {
@@ -528,6 +528,7 @@ static int __blk_send_generic(struct request_queue *q, struct gendisk *bd_disk,
 	rq = blk_get_request(q, WRITE, __GFP_WAIT);
 	rq->cmd_type = REQ_TYPE_BLOCK_PC;
 	rq->data = NULL;
+	rq->raw_data_len = 0;
 	rq->data_len = 0;
 	rq->timeout = BLK_DEFAULT_SG_TIMEOUT;
 	memset(rq->cmd, 0, sizeof(rq->cmd));
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 135c1d054701..ba21d97d1855 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1014,10 +1014,6 @@ static int scsi_init_sgtable(struct request *req, struct scsi_data_buffer *sdb,
 	}
 
 	req->buffer = NULL;
-	if (blk_pc_request(req))
-		sdb->length = req->data_len;
-	else
-		sdb->length = req->nr_sectors << 9;
 
 	/* 
 	 * Next, walk the list, and fill in the addresses and sizes of
@@ -1026,6 +1022,10 @@ static int scsi_init_sgtable(struct request *req, struct scsi_data_buffer *sdb,
 	count = blk_rq_map_sg(req->q, req, sdb->table.sgl);
 	BUG_ON(count > sdb->table.nents);
 	sdb->table.nents = count;
+	if (blk_pc_request(req))
+		sdb->length = req->data_len;
+	else
+		sdb->length = req->nr_sectors << 9;
 	return BLKPREP_OK;
 }
 
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index e1888cc5b8ae..f1fe9fbf1c0e 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -216,6 +216,7 @@ struct request {
 	unsigned int cmd_len;
 	unsigned char cmd[BLK_MAX_CDB];
 
+	unsigned int raw_data_len;
 	unsigned int data_len;
 	unsigned int sense_len;
 	void *data;
-- 
cgit v1.2.3


From 2fb98e8414c42cb14698833aac640b143b9ade4f Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 19 Feb 2008 11:36:53 +0100
Subject: block: implement request_queue->dma_drain_needed

Draining shouldn't be done for commands where overflow may indicate
data integrity issues.  Add dma_drain_needed callback to
request_queue.  Drain buffer is appened iff this function returns
non-zero.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 block/blk-merge.c      | 2 +-
 block/blk-settings.c   | 7 +++++--
 include/linux/blkdev.h | 7 +++++--
 3 files changed, 11 insertions(+), 5 deletions(-)

(limited to 'include/linux')

diff --git a/block/blk-merge.c b/block/blk-merge.c
index 39f2e077a014..bef1b4d0fc02 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -220,7 +220,7 @@ new_segment:
 		bvprv = bvec;
 	} /* segments in rq */
 
-	if (q->dma_drain_size) {
+	if (q->dma_drain_size && q->dma_drain_needed(rq)) {
 		sg->page_link &= ~0x02;
 		sg = sg_next(sg);
 		sg_set_page(sg, virt_to_page(q->dma_drain_buffer),
diff --git a/block/blk-settings.c b/block/blk-settings.c
index 13536a388d27..9a8ffdd0ce3d 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -296,6 +296,7 @@ EXPORT_SYMBOL(blk_queue_stack_limits);
  * blk_queue_dma_drain - Set up a drain buffer for excess dma.
  *
  * @q:  the request queue for the device
+ * @dma_drain_needed: fn which returns non-zero if drain is necessary
  * @buf:	physically contiguous buffer
  * @size:	size of the buffer in bytes
  *
@@ -315,14 +316,16 @@ EXPORT_SYMBOL(blk_queue_stack_limits);
  * device can support otherwise there won't be room for the drain
  * buffer.
  */
-int blk_queue_dma_drain(struct request_queue *q, void *buf,
-				unsigned int size)
+extern int blk_queue_dma_drain(struct request_queue *q,
+			       dma_drain_needed_fn *dma_drain_needed,
+			       void *buf, unsigned int size)
 {
 	if (q->max_hw_segments < 2 || q->max_phys_segments < 2)
 		return -EINVAL;
 	/* make room for appending the drain */
 	--q->max_hw_segments;
 	--q->max_phys_segments;
+	q->dma_drain_needed = dma_drain_needed;
 	q->dma_drain_buffer = buf;
 	q->dma_drain_size = size;
 
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index f1fe9fbf1c0e..6fe67d1939c2 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -259,6 +259,7 @@ struct bio_vec;
 typedef int (merge_bvec_fn) (struct request_queue *, struct bio *, struct bio_vec *);
 typedef void (prepare_flush_fn) (struct request_queue *, struct request *);
 typedef void (softirq_done_fn)(struct request *);
+typedef int (dma_drain_needed_fn)(struct request *);
 
 enum blk_queue_state {
 	Queue_down,
@@ -295,6 +296,7 @@ struct request_queue
 	merge_bvec_fn		*merge_bvec_fn;
 	prepare_flush_fn	*prepare_flush_fn;
 	softirq_done_fn		*softirq_done_fn;
+	dma_drain_needed_fn	*dma_drain_needed;
 
 	/*
 	 * Dispatch queue sorting
@@ -699,8 +701,9 @@ extern void blk_queue_max_hw_segments(struct request_queue *, unsigned short);
 extern void blk_queue_max_segment_size(struct request_queue *, unsigned int);
 extern void blk_queue_hardsect_size(struct request_queue *, unsigned short);
 extern void blk_queue_stack_limits(struct request_queue *t, struct request_queue *b);
-extern int blk_queue_dma_drain(struct request_queue *q, void *buf,
-			       unsigned int size);
+extern int blk_queue_dma_drain(struct request_queue *q,
+			       dma_drain_needed_fn *dma_drain_needed,
+			       void *buf, unsigned int size);
 extern void blk_queue_segment_boundary(struct request_queue *, unsigned long);
 extern void blk_queue_prep_rq(struct request_queue *, prep_rq_fn *pfn);
 extern void blk_queue_merge_bvec(struct request_queue *, merge_bvec_fn *);
-- 
cgit v1.2.3


From dde2020754aeb14e17052d61784dcb37f252aac2 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@HansenPartnership.com>
Date: Tue, 19 Feb 2008 11:36:56 +0100
Subject: libata: eliminate the home grown dma padding in favour of

that provided by the block layer

ATA requires that all DMA transfers begin and end on word boundaries.
Because of this, a large amount of machinery grew up in ide to adjust
scatterlists on this basis.  However, as of 2.5, the block layer has a
dma_alignment variable which ensures both the beginning and length of a
DMA transfer are aligned on the dma_alignment boundary.  Although the
block layer does adjust the beginning of the transfer to ensure this
happens, it doesn't actually adjust the length, it merely makes sure
that space is allocated for transfers beyond the declared length.  The
upshot of this is that scatterlists may be padded to any size between
the actual length and the length adjusted to the dma_alignment safely
knowing that memory is allocated in this region.

Right at the moment, SCSI takes the default dma_aligment which is on a
512 byte boundary.  Note that this aligment only applies to transfers
coming in from user space.  However, since all kernel allocations are
automatically aligned on a minimum of 32 byte boundaries, it is safe to
adjust them in this manner as well.

tj: * Adjusting sg after padding is done in block layer.  Make libata
      set queue alignment correctly for ATAPI devices and drop broken
      sg mangling from ata_sg_setup().
    * Use request->raw_data_len for ATAPI transfer chunk size.
    * Killed qc->raw_nbytes.
    * Separated out killing qc->n_iter.

Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 drivers/ata/ahci.c            |   5 --
 drivers/ata/libata-core.c     | 145 +++---------------------------------------
 drivers/ata/libata-scsi.c     |  23 ++-----
 drivers/ata/pata_icside.c     |   8 ---
 drivers/ata/sata_fsl.c        |  13 ----
 drivers/ata/sata_mv.c         |   6 +-
 drivers/ata/sata_sil24.c      |   5 --
 drivers/scsi/ipr.c            |   4 +-
 drivers/scsi/libsas/sas_ata.c |   4 +-
 include/linux/libata.h        |  28 +-------
 10 files changed, 21 insertions(+), 220 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 29e71bddd6ff..3c06e457b4dc 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1975,16 +1975,11 @@ static int ahci_port_start(struct ata_port *ap)
 	struct ahci_port_priv *pp;
 	void *mem;
 	dma_addr_t mem_dma;
-	int rc;
 
 	pp = devm_kzalloc(dev, sizeof(*pp), GFP_KERNEL);
 	if (!pp)
 		return -ENOMEM;
 
-	rc = ata_pad_alloc(ap, dev);
-	if (rc)
-		return rc;
-
 	mem = dmam_alloc_coherent(dev, AHCI_PORT_PRIV_DMA_SZ, &mem_dma,
 				  GFP_KERNEL);
 	if (!mem)
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index a109ccbda9ca..3587ac3fe3f3 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -4493,30 +4493,13 @@ void ata_sg_clean(struct ata_queued_cmd *qc)
 	struct ata_port *ap = qc->ap;
 	struct scatterlist *sg = qc->sg;
 	int dir = qc->dma_dir;
-	void *pad_buf = NULL;
 
 	WARN_ON(sg == NULL);
 
-	VPRINTK("unmapping %u sg elements\n", qc->mapped_n_elem);
+	VPRINTK("unmapping %u sg elements\n", qc->n_elem);
 
-	/* if we padded the buffer out to 32-bit bound, and data
-	 * xfer direction is from-device, we must copy from the
-	 * pad buffer back into the supplied buffer
-	 */
-	if (qc->pad_len && !(qc->tf.flags & ATA_TFLAG_WRITE))
-		pad_buf = ap->pad + (qc->tag * ATA_DMA_PAD_SZ);
-
-	if (qc->mapped_n_elem)
-		dma_unmap_sg(ap->dev, sg, qc->mapped_n_elem, dir);
-	/* restore last sg */
-	if (qc->last_sg)
-		*qc->last_sg = qc->saved_last_sg;
-	if (pad_buf) {
-		struct scatterlist *psg = &qc->extra_sg[1];
-		void *addr = kmap_atomic(sg_page(psg), KM_IRQ0);
-		memcpy(addr + psg->offset, pad_buf, qc->pad_len);
-		kunmap_atomic(addr, KM_IRQ0);
-	}
+	if (qc->n_elem)
+		dma_unmap_sg(ap->dev, sg, qc->n_elem, dir);
 
 	qc->flags &= ~ATA_QCFLAG_DMAMAP;
 	qc->sg = NULL;
@@ -4767,97 +4750,6 @@ void ata_sg_init(struct ata_queued_cmd *qc, struct scatterlist *sg,
 	qc->cursg = qc->sg;
 }
 
-static unsigned int ata_sg_setup_extra(struct ata_queued_cmd *qc,
-				       unsigned int *n_elem_extra,
-				       unsigned int *nbytes_extra)
-{
-	struct ata_port *ap = qc->ap;
-	unsigned int n_elem = qc->n_elem;
-	struct scatterlist *lsg, *copy_lsg = NULL, *tsg = NULL, *esg = NULL;
-
-	*n_elem_extra = 0;
-	*nbytes_extra = 0;
-
-	/* needs padding? */
-	qc->pad_len = qc->nbytes & 3;
-
-	if (likely(!qc->pad_len))
-		return n_elem;
-
-	/* locate last sg and save it */
-	lsg = sg_last(qc->sg, n_elem);
-	qc->last_sg = lsg;
-	qc->saved_last_sg = *lsg;
-
-	sg_init_table(qc->extra_sg, ARRAY_SIZE(qc->extra_sg));
-
-	if (qc->pad_len) {
-		struct scatterlist *psg = &qc->extra_sg[1];
-		void *pad_buf = ap->pad + (qc->tag * ATA_DMA_PAD_SZ);
-		unsigned int offset;
-
-		WARN_ON(qc->dev->class != ATA_DEV_ATAPI);
-
-		memset(pad_buf, 0, ATA_DMA_PAD_SZ);
-
-		/* psg->page/offset are used to copy to-be-written
-		 * data in this function or read data in ata_sg_clean.
-		 */
-		offset = lsg->offset + lsg->length - qc->pad_len;
-		sg_set_page(psg, nth_page(sg_page(lsg), offset >> PAGE_SHIFT),
-			    qc->pad_len, offset_in_page(offset));
-
-		if (qc->tf.flags & ATA_TFLAG_WRITE) {
-			void *addr = kmap_atomic(sg_page(psg), KM_IRQ0);
-			memcpy(pad_buf, addr + psg->offset, qc->pad_len);
-			kunmap_atomic(addr, KM_IRQ0);
-		}
-
-		sg_dma_address(psg) = ap->pad_dma + (qc->tag * ATA_DMA_PAD_SZ);
-		sg_dma_len(psg) = ATA_DMA_PAD_SZ;
-
-		/* Trim the last sg entry and chain the original and
-		 * padding sg lists.
-		 *
-		 * Because chaining consumes one sg entry, one extra
-		 * sg entry is allocated and the last sg entry is
-		 * copied to it if the length isn't zero after padded
-		 * amount is removed.
-		 *
-		 * If the last sg entry is completely replaced by
-		 * padding sg entry, the first sg entry is skipped
-		 * while chaining.
-		 */
-		lsg->length -= qc->pad_len;
-		if (lsg->length) {
-			copy_lsg = &qc->extra_sg[0];
-			tsg = &qc->extra_sg[0];
-		} else {
-			n_elem--;
-			tsg = &qc->extra_sg[1];
-		}
-
-		esg = &qc->extra_sg[1];
-
-		(*n_elem_extra)++;
-		(*nbytes_extra) += 4 - qc->pad_len;
-	}
-
-	if (copy_lsg)
-		sg_set_page(copy_lsg, sg_page(lsg), lsg->length, lsg->offset);
-
-	sg_chain(lsg, 1, tsg);
-	sg_mark_end(esg);
-
-	/* sglist can't start with chaining sg entry, fast forward */
-	if (qc->sg == lsg) {
-		qc->sg = tsg;
-		qc->cursg = tsg;
-	}
-
-	return n_elem;
-}
-
 /**
  *	ata_sg_setup - DMA-map the scatter-gather table associated with a command.
  *	@qc: Command with scatter-gather table to be mapped.
@@ -4874,26 +4766,17 @@ static unsigned int ata_sg_setup_extra(struct ata_queued_cmd *qc,
 static int ata_sg_setup(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
-	unsigned int n_elem, n_elem_extra, nbytes_extra;
+	unsigned int n_elem;
 
 	VPRINTK("ENTER, ata%u\n", ap->print_id);
 
-	n_elem = ata_sg_setup_extra(qc, &n_elem_extra, &nbytes_extra);
+	n_elem = dma_map_sg(ap->dev, qc->sg, qc->n_elem, qc->dma_dir);
+	if (n_elem < 1)
+		return -1;
 
-	if (n_elem) {
-		n_elem = dma_map_sg(ap->dev, qc->sg, n_elem, qc->dma_dir);
-		if (n_elem < 1) {
-			/* restore last sg */
-			if (qc->last_sg)
-				*qc->last_sg = qc->saved_last_sg;
-			return -1;
-		}
-		DPRINTK("%d sg elements mapped\n", n_elem);
-	}
+	DPRINTK("%d sg elements mapped\n", n_elem);
 
-	qc->n_elem = qc->mapped_n_elem = n_elem;
-	qc->n_elem += n_elem_extra;
-	qc->nbytes += nbytes_extra;
+	qc->n_elem = n_elem;
 	qc->flags |= ATA_QCFLAG_DMAMAP;
 
 	return 0;
@@ -5962,9 +5845,6 @@ void ata_qc_issue(struct ata_queued_cmd *qc)
 	 */
 	BUG_ON(ata_is_data(prot) && (!qc->sg || !qc->n_elem || !qc->nbytes));
 
-	/* ata_sg_setup() may update nbytes */
-	qc->raw_nbytes = qc->nbytes;
-
 	if (ata_is_dma(prot) || (ata_is_pio(prot) &&
 				 (ap->flags & ATA_FLAG_PIO_DMA)))
 		if (ata_sg_setup(qc))
@@ -6573,19 +6453,12 @@ void ata_host_resume(struct ata_host *host)
 int ata_port_start(struct ata_port *ap)
 {
 	struct device *dev = ap->dev;
-	int rc;
 
 	ap->prd = dmam_alloc_coherent(dev, ATA_PRD_TBL_SZ, &ap->prd_dma,
 				      GFP_KERNEL);
 	if (!ap->prd)
 		return -ENOMEM;
 
-	rc = ata_pad_alloc(ap, dev);
-	if (rc)
-		return rc;
-
-	DPRINTK("prd alloc, virt %p, dma %llx\n", ap->prd,
-		(unsigned long long)ap->prd_dma);
 	return 0;
 }
 
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 1cea18f62abc..6e15c5ddae6d 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -832,24 +832,16 @@ static void ata_scsi_dev_config(struct scsi_device *sdev,
 	/* configure max sectors */
 	blk_queue_max_sectors(sdev->request_queue, dev->max_sectors);
 
-	/* SATA DMA transfers must be multiples of 4 byte, so
-	 * we need to pad ATAPI transfers using an extra sg.
-	 * Decrement max hw segments accordingly.
-	 */
-	if (dev->class == ATA_DEV_ATAPI) {
-		struct request_queue *q = sdev->request_queue;
-		blk_queue_max_hw_segments(q, q->max_hw_segments - 1);
-
+	if (dev->class == ATA_DEV_ATAPI)
 		/* set the min alignment */
 		blk_queue_update_dma_alignment(sdev->request_queue,
 					       ATA_DMA_PAD_SZ - 1);
-	} else
+	else {
 		/* ATA devices must be sector aligned */
 		blk_queue_update_dma_alignment(sdev->request_queue,
 					       ATA_SECT_SIZE - 1);
-
-	if (dev->class == ATA_DEV_ATA)
 		sdev->manage_start_stop = 1;
+	}
 
 	if (dev->flags & ATA_DFLAG_AN)
 		set_bit(SDEV_EVT_MEDIA_CHANGE, sdev->supported_events);
@@ -2500,7 +2492,7 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc)
 	 * want to set it properly, and for DMA where it is
 	 * effectively meaningless.
 	 */
-	nbytes = min(qc->nbytes, (unsigned int)63 * 1024);
+	nbytes = min(scmd->request->raw_data_len, (unsigned int)63 * 1024);
 
 	/* Most ATAPI devices which honor transfer chunk size don't
 	 * behave according to the spec when odd chunk size which
@@ -3555,7 +3547,7 @@ EXPORT_SYMBOL_GPL(ata_sas_port_alloc);
  *	@ap: Port to initialize
  *
  *	Called just after data structures for each port are
- *	initialized.  Allocates DMA pad.
+ *	initialized.
  *
  *	May be used as the port_start() entry in ata_port_operations.
  *
@@ -3564,7 +3556,7 @@ EXPORT_SYMBOL_GPL(ata_sas_port_alloc);
  */
 int ata_sas_port_start(struct ata_port *ap)
 {
-	return ata_pad_alloc(ap, ap->dev);
+	return 0;
 }
 EXPORT_SYMBOL_GPL(ata_sas_port_start);
 
@@ -3572,8 +3564,6 @@ EXPORT_SYMBOL_GPL(ata_sas_port_start);
  *	ata_port_stop - Undo ata_sas_port_start()
  *	@ap: Port to shut down
  *
- *	Frees the DMA pad.
- *
  *	May be used as the port_stop() entry in ata_port_operations.
  *
  *	LOCKING:
@@ -3582,7 +3572,6 @@ EXPORT_SYMBOL_GPL(ata_sas_port_start);
 
 void ata_sas_port_stop(struct ata_port *ap)
 {
-	ata_pad_free(ap, ap->dev);
 }
 EXPORT_SYMBOL_GPL(ata_sas_port_stop);
 
diff --git a/drivers/ata/pata_icside.c b/drivers/ata/pata_icside.c
index 5b8586dac63b..f97068be2d79 100644
--- a/drivers/ata/pata_icside.c
+++ b/drivers/ata/pata_icside.c
@@ -304,12 +304,6 @@ static int icside_dma_init(struct pata_icside_info *info)
 }
 
 
-static int pata_icside_port_start(struct ata_port *ap)
-{
-	/* No PRD to alloc */
-	return ata_pad_alloc(ap, ap->dev);
-}
-
 static struct scsi_host_template pata_icside_sht = {
 	.module			= THIS_MODULE,
 	.name			= DRV_NAME,
@@ -389,8 +383,6 @@ static struct ata_port_operations pata_icside_port_ops = {
 	.irq_clear		= ata_dummy_noret,
 	.irq_on			= ata_irq_on,
 
-	.port_start		= pata_icside_port_start,
-
 	.bmdma_stop		= pata_icside_bmdma_stop,
 	.bmdma_status		= pata_icside_bmdma_status,
 };
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index efcb66b6ccef..9323dd0c7d8d 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -601,21 +601,9 @@ static int sata_fsl_port_start(struct ata_port *ap)
 	if (!pp)
 		return -ENOMEM;
 
-	/*
-	 * allocate per command dma alignment pad buffer, which is used
-	 * internally by libATA to ensure that all transfers ending on
-	 * unaligned boundaries are padded, to align on Dword boundaries
-	 */
-	retval = ata_pad_alloc(ap, dev);
-	if (retval) {
-		kfree(pp);
-		return retval;
-	}
-
 	mem = dma_alloc_coherent(dev, SATA_FSL_PORT_PRIV_DMA_SZ, &mem_dma,
 				 GFP_KERNEL);
 	if (!mem) {
-		ata_pad_free(ap, dev);
 		kfree(pp);
 		return -ENOMEM;
 	}
@@ -694,7 +682,6 @@ static void sata_fsl_port_stop(struct ata_port *ap)
 	dma_free_coherent(dev, SATA_FSL_PORT_PRIV_DMA_SZ,
 			  pp->cmdslot, pp->cmdslot_paddr);
 
-	ata_pad_free(ap, dev);
 	kfree(pp);
 }
 
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 2ecd44db4142..1c1fbf375d9a 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -1158,17 +1158,13 @@ static int mv_port_start(struct ata_port *ap)
 	struct mv_port_priv *pp;
 	void __iomem *port_mmio = mv_ap_base(ap);
 	unsigned long flags;
-	int tag, rc;
+	int tag;
 
 	pp = devm_kzalloc(dev, sizeof(*pp), GFP_KERNEL);
 	if (!pp)
 		return -ENOMEM;
 	ap->private_data = pp;
 
-	rc = ata_pad_alloc(ap, dev);
-	if (rc)
-		return rc;
-
 	pp->crqb = dma_pool_alloc(hpriv->crqb_pool, GFP_KERNEL, &pp->crqb_dma);
 	if (!pp->crqb)
 		return -ENOMEM;
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index b4b1f91ea693..df7988df7908 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -1234,7 +1234,6 @@ static int sil24_port_start(struct ata_port *ap)
 	union sil24_cmd_block *cb;
 	size_t cb_size = sizeof(*cb) * SIL24_MAX_CMDS;
 	dma_addr_t cb_dma;
-	int rc;
 
 	pp = devm_kzalloc(dev, sizeof(*pp), GFP_KERNEL);
 	if (!pp)
@@ -1247,10 +1246,6 @@ static int sil24_port_start(struct ata_port *ap)
 		return -ENOMEM;
 	memset(cb, 0, cb_size);
 
-	rc = ata_pad_alloc(ap, dev);
-	if (rc)
-		return rc;
-
 	pp->cmd_block = cb;
 	pp->cmd_block_dma = cb_dma;
 
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index 2074701f7e76..c72014a3e7d4 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -5140,7 +5140,7 @@ static void ipr_build_ata_ioadl(struct ipr_cmnd *ipr_cmd,
 	struct ipr_ioarcb *ioarcb = &ipr_cmd->ioarcb;
 	struct ipr_ioadl_desc *ioadl = ipr_cmd->ioadl;
 	struct ipr_ioadl_desc *last_ioadl = NULL;
-	int len = qc->nbytes + qc->pad_len;
+	int len = qc->nbytes;
 	struct scatterlist *sg;
 	unsigned int si;
 
@@ -5206,7 +5206,7 @@ static unsigned int ipr_qc_issue(struct ata_queued_cmd *qc)
 	ioarcb->cmd_pkt.request_type = IPR_RQTYPE_ATA_PASSTHRU;
 	ioarcb->cmd_pkt.flags_hi |= IPR_FLAGS_HI_NO_LINK_DESC;
 	ioarcb->cmd_pkt.flags_hi |= IPR_FLAGS_HI_NO_ULEN_CHK;
-	ipr_cmd->dma_use_sg = qc->pad_len ? qc->n_elem + 1 : qc->n_elem;
+	ipr_cmd->dma_use_sg = qc->n_elem;
 
 	ipr_build_ata_ioadl(ipr_cmd, qc);
 	regs->flags |= IPR_ATA_FLAG_STATUS_ON_GOOD_COMPLETION;
diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index 0996f866f14c..7cd05b599a12 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -178,8 +178,8 @@ static unsigned int sas_ata_qc_issue(struct ata_queued_cmd *qc)
 	task->uldd_task = qc;
 	if (ata_is_atapi(qc->tf.protocol)) {
 		memcpy(task->ata_task.atapi_packet, qc->cdb, qc->dev->cdb_len);
-		task->total_xfer_len = qc->nbytes + qc->pad_len;
-		task->num_scatter = qc->pad_len ? qc->n_elem + 1 : qc->n_elem;
+		task->total_xfer_len = qc->nbytes;
+		task->num_scatter = qc->n_elem;
 	} else {
 		for_each_sg(qc->sg, sg, qc->n_elem, si)
 			xfer += sg->length;
diff --git a/include/linux/libata.h b/include/linux/libata.h
index bc5a8d0c7090..2e098f940cec 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -278,7 +278,6 @@ enum {
 
 	/* size of buffer to pad xfers ending on unaligned boundaries */
 	ATA_DMA_PAD_SZ		= 4,
-	ATA_DMA_PAD_BUF_SZ	= ATA_DMA_PAD_SZ * ATA_MAX_QUEUE,
 
 	/* ering size */
 	ATA_ERING_SIZE		= 32,
@@ -457,24 +456,18 @@ struct ata_queued_cmd {
 	unsigned long		flags;		/* ATA_QCFLAG_xxx */
 	unsigned int		tag;
 	unsigned int		n_elem;
-	unsigned int		mapped_n_elem;
 
 	int			dma_dir;
 
-	unsigned int		pad_len;
 	unsigned int		sect_size;
 
 	unsigned int		nbytes;
-	unsigned int		raw_nbytes;
 	unsigned int		curbytes;
 
 	struct scatterlist	*cursg;
 	unsigned int		cursg_ofs;
 
-	struct scatterlist	*last_sg;
-	struct scatterlist	saved_last_sg;
 	struct scatterlist	sgent;
-	struct scatterlist	extra_sg[2];
 
 	struct scatterlist	*sg;
 
@@ -619,9 +612,6 @@ struct ata_port {
 	struct ata_prd		*prd;	 /* our SG list */
 	dma_addr_t		prd_dma; /* and its DMA mapping */
 
-	void			*pad;	/* array of DMA pad buffers */
-	dma_addr_t		pad_dma;
-
 	struct ata_ioports	ioaddr;	/* ATA cmd/ctl/dma register blocks */
 
 	u8			ctl;	/* cache of ATA control register */
@@ -1363,12 +1353,9 @@ static inline void ata_qc_reinit(struct ata_queued_cmd *qc)
 	qc->flags = 0;
 	qc->cursg = NULL;
 	qc->cursg_ofs = 0;
-	qc->nbytes = qc->raw_nbytes = qc->curbytes = 0;
+	qc->nbytes = qc->curbytes = 0;
 	qc->n_elem = 0;
-	qc->mapped_n_elem = 0;
 	qc->err_mask = 0;
-	qc->pad_len = 0;
-	qc->last_sg = NULL;
 	qc->sect_size = ATA_SECT_SIZE;
 
 	ata_tf_init(qc->dev, &qc->tf);
@@ -1423,19 +1410,6 @@ static inline unsigned int __ac_err_mask(u8 status)
 	return mask;
 }
 
-static inline int ata_pad_alloc(struct ata_port *ap, struct device *dev)
-{
-	ap->pad_dma = 0;
-	ap->pad = dmam_alloc_coherent(dev, ATA_DMA_PAD_BUF_SZ,
-				      &ap->pad_dma, GFP_KERNEL);
-	return (ap->pad == NULL) ? -ENOMEM : 0;
-}
-
-static inline void ata_pad_free(struct ata_port *ap, struct device *dev)
-{
-	dmam_free_coherent(dev, ATA_DMA_PAD_BUF_SZ, ap->pad, ap->pad_dma);
-}
-
 static inline struct ata_port *ata_shost_to_port(struct Scsi_Host *host)
 {
 	return *(struct ata_port **)&host->hostdata[0];
-- 
cgit v1.2.3


From 37c514e3dfc8f55145d9c6895e2838ac31859aa4 Mon Sep 17 00:00:00 2001
From: Sam Ravnborg <sam@ravnborg.org>
Date: Tue, 19 Feb 2008 21:00:18 +0100
Subject: Add missing init section definitions

When adding __devinitconst etc. the __initconst variant
were missed.
Add this one and proper definitions for .head.text for use
in .S files.
The naming .head.text is preferred over .text.head as the
latter will conflict for a function named head when introducing
-ffunctions-sections.

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
---
 include/asm-generic/vmlinux.lds.h | 3 +++
 include/linux/init.h              | 2 ++
 2 files changed, 5 insertions(+)

(limited to 'include/linux')

diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index f784d2f34149..f054778e916c 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -238,6 +238,9 @@
 		*(.kprobes.text)					\
 		VMLINUX_SYMBOL(__kprobes_text_end) = .;
 
+/* Section used for early init (in .S files) */
+#define HEAD_TEXT  *(.head.text)
+
 /* init and exit section handling */
 #define INIT_DATA							\
 	*(.init.data)							\
diff --git a/include/linux/init.h b/include/linux/init.h
index a404a0055dd7..fb58c0493cf2 100644
--- a/include/linux/init.h
+++ b/include/linux/init.h
@@ -42,6 +42,7 @@
    discard it in modules) */
 #define __init		__section(.init.text) __cold
 #define __initdata	__section(.init.data)
+#define __initconst	__section(.init.rodata)
 #define __exitdata	__section(.exit.data)
 #define __exit_call	__used __section(.exitcall.exit)
 
@@ -106,6 +107,7 @@
 #define __memexitconst   __section(.memexit.rodata)
 
 /* For assembly routines */
+#define __HEAD		.section	".head.text","ax"
 #define __INIT		.section	".init.text","ax"
 #define __FINIT		.previous
 
-- 
cgit v1.2.3


From 651acc9893f2cd036e9485ce782d0c4cffbc27f5 Mon Sep 17 00:00:00 2001
From: "Robert P. J. Day" <rpjday@crashcourse.ca>
Date: Tue, 19 Feb 2008 17:19:14 -0800
Subject: [NETFILTER]: Make sure xt_policy.h is unifdef'ed.

Since the header file xt_policy.h tests __KERNEL__, it should be
unifdef'ed before exporting to userspace.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/Kbuild | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/Kbuild b/include/linux/netfilter/Kbuild
index 91fef0cae42f..3aff513d12c8 100644
--- a/include/linux/netfilter/Kbuild
+++ b/include/linux/netfilter/Kbuild
@@ -30,7 +30,6 @@ header-y += xt_mark.h
 header-y += xt_multiport.h
 header-y += xt_owner.h
 header-y += xt_pkttype.h
-header-y += xt_policy.h
 header-y += xt_rateest.h
 header-y += xt_realm.h
 header-y += xt_sctp.h
@@ -47,3 +46,4 @@ unifdef-y += nfnetlink.h
 unifdef-y += nfnetlink_compat.h
 unifdef-y += x_tables.h
 unifdef-y += xt_physdev.h
+unifdef-y += xt_policy.h
-- 
cgit v1.2.3


From 7d9904c2604ce091c97c225b5e2f92f3e7cc2e3a Mon Sep 17 00:00:00 2001
From: Jan Engelhardt <jengelh@computergmbh.de>
Date: Tue, 19 Feb 2008 17:19:44 -0800
Subject: [NETFILTER]: xt_hashlimit: remove unneeded struct member

By allocating ->hinfo, we already have the needed indirection to cope
with the per-cpu xtables struct match_entry.

[Patrick: do this now before the revision 1 struct is used by userspace]

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/xt_hashlimit.h | 1 -
 net/netfilter/xt_hashlimit.c           | 3 ---
 2 files changed, 4 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/xt_hashlimit.h b/include/linux/netfilter/xt_hashlimit.h
index 58b818ee41ca..51b18d83b477 100644
--- a/include/linux/netfilter/xt_hashlimit.h
+++ b/include/linux/netfilter/xt_hashlimit.h
@@ -61,7 +61,6 @@ struct xt_hashlimit_mtinfo1 {
 
 	/* Used internally by the kernel */
 	struct xt_hashlimit_htable *hinfo __attribute__((aligned(8)));
-	struct xt_hashlimit_mtinfo1 *master __attribute__((aligned(8)));
 };
 
 #endif /*_XT_HASHLIMIT_H*/
diff --git a/net/netfilter/xt_hashlimit.c b/net/netfilter/xt_hashlimit.c
index 744c7f2ab0b1..5418ce59ac3a 100644
--- a/net/netfilter/xt_hashlimit.c
+++ b/net/netfilter/xt_hashlimit.c
@@ -774,9 +774,6 @@ hashlimit_mt_check(const char *tablename, const void *inf,
 		return false;
 	}
 	mutex_unlock(&hlimit_mutex);
-
-	/* Ugly hack: For SMP, we only want to use one set */
-	info->master = info;
 	return true;
 }
 
-- 
cgit v1.2.3


From 7b33ed22194d8f0dbcf682f5cdf5b9c2ef551e7c Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 19 Feb 2008 17:20:33 -0800
Subject: [NETFILTER]: Use __u32 in struct nf_inet_addr

As reported by David Woodhouse <dwmw2@infradead.org>, using u_int32_t
in struct nf_inet_addr breaks the busybox build. Fix by using __u32.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h
index d74e79bacd2d..b74b615492e8 100644
--- a/include/linux/netfilter.h
+++ b/include/linux/netfilter.h
@@ -51,7 +51,7 @@ enum nf_inet_hooks {
 };
 
 union nf_inet_addr {
-	u_int32_t	all[4];
+	__u32		all[4];
 	__be32		ip;
 	__be32		ip6[4];
 	struct in_addr	in;
-- 
cgit v1.2.3


From 5d44b414daa8c1d8551aed6130d86d54175db43f Mon Sep 17 00:00:00 2001
From: Harvey Harrison <harvey.harrison@gmail.com>
Date: Fri, 15 Feb 2008 13:41:32 -0800
Subject: ata: fix sparse warning in libata.h

Avoids lots of these, also is more readable.
include/linux/libata.h:1210:13: warning: potentially expensive pointer subtraction

Change the subtraction to addition on the other side of the comparison.

Thanks to Christer Weinigel for the suggestion.

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Acked-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 include/linux/libata.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/libata.h b/include/linux/libata.h
index 2e098f940cec..ce7603a73156 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1197,7 +1197,7 @@ static inline struct ata_link *ata_port_next_link(struct ata_link *link)
 		return ap->pmp_link;
 	}
 
-	if (++link - ap->pmp_link < ap->nr_pmp_links)
+	if (++link < ap->nr_pmp_links + ap->pmp_link)
 		return link;
 	return NULL;
 }
-- 
cgit v1.2.3


From 58ff70d4feae29cbb7ace410fa6585ef3afb44b6 Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Mon, 18 Feb 2008 21:44:39 +0100
Subject: ssb: Fix serial console on new bcm47xx devices

This fixes the baud settings for new devices
like the Linksys WRT350n.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/driver_chipcommon.c           | 36 ++++++++++++++++++++++++++-----
 include/linux/ssb/ssb_driver_chipcommon.h |  3 +++
 2 files changed, 34 insertions(+), 5 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ssb/driver_chipcommon.c b/drivers/ssb/driver_chipcommon.c
index 6fbf1c53b6f2..7cc03f2dd5a6 100644
--- a/drivers/ssb/driver_chipcommon.c
+++ b/drivers/ssb/driver_chipcommon.c
@@ -376,6 +376,7 @@ int ssb_chipco_serial_init(struct ssb_chipcommon *cc,
 	unsigned int irq;
 	u32 baud_base, div;
 	u32 i, n;
+	unsigned int ccrev = cc->dev->id.revision;
 
 	plltype = (cc->capabilities & SSB_CHIPCO_CAP_PLLT);
 	irq = ssb_mips_irq(cc->dev);
@@ -387,14 +388,39 @@ int ssb_chipco_serial_init(struct ssb_chipcommon *cc,
 						chipco_read32(cc, SSB_CHIPCO_CLOCK_M2));
 		div = 1;
 	} else {
-		if (cc->dev->id.revision >= 11) {
+		if (ccrev == 20) {
+			/* BCM5354 uses constant 25MHz clock */
+			baud_base = 25000000;
+			div = 48;
+			/* Set the override bit so we don't divide it */
+			chipco_write32(cc, SSB_CHIPCO_CORECTL,
+				       chipco_read32(cc, SSB_CHIPCO_CORECTL)
+				       | SSB_CHIPCO_CORECTL_UARTCLK0);
+		} else if ((ccrev >= 11) && (ccrev != 15)) {
 			/* Fixed ALP clock */
 			baud_base = 20000000;
+			if (cc->capabilities & SSB_CHIPCO_CAP_PMU) {
+				/* FIXME: baud_base is different for devices with a PMU */
+				SSB_WARN_ON(1);
+			}
 			div = 1;
+			if (ccrev >= 21) {
+				/* Turn off UART clock before switching clocksource. */
+				chipco_write32(cc, SSB_CHIPCO_CORECTL,
+					       chipco_read32(cc, SSB_CHIPCO_CORECTL)
+					       & ~SSB_CHIPCO_CORECTL_UARTCLKEN);
+			}
 			/* Set the override bit so we don't divide it */
 			chipco_write32(cc, SSB_CHIPCO_CORECTL,
-				       SSB_CHIPCO_CORECTL_UARTCLK0);
-		} else if (cc->dev->id.revision >= 3) {
+				       chipco_read32(cc, SSB_CHIPCO_CORECTL)
+				       | SSB_CHIPCO_CORECTL_UARTCLK0);
+			if (ccrev >= 21) {
+				/* Re-enable the UART clock. */
+				chipco_write32(cc, SSB_CHIPCO_CORECTL,
+					       chipco_read32(cc, SSB_CHIPCO_CORECTL)
+					       | SSB_CHIPCO_CORECTL_UARTCLKEN);
+			}
+		} else if (ccrev >= 3) {
 			/* Internal backplane clock */
 			baud_base = ssb_clockspeed(bus);
 			div = chipco_read32(cc, SSB_CHIPCO_CLKDIV)
@@ -406,7 +432,7 @@ int ssb_chipco_serial_init(struct ssb_chipcommon *cc,
 		}
 
 		/* Clock source depends on strapping if UartClkOverride is unset */
-		if ((cc->dev->id.revision > 0) &&
+		if ((ccrev > 0) &&
 		    !(chipco_read32(cc, SSB_CHIPCO_CORECTL) & SSB_CHIPCO_CORECTL_UARTCLK0)) {
 			if ((cc->capabilities & SSB_CHIPCO_CAP_UARTCLK) ==
 			    SSB_CHIPCO_CAP_UARTCLK_INT) {
@@ -428,7 +454,7 @@ int ssb_chipco_serial_init(struct ssb_chipcommon *cc,
 		cc_mmio = cc->dev->bus->mmio + (cc->dev->core_index * SSB_CORE_SIZE);
 		uart_regs = cc_mmio + SSB_CHIPCO_UART0_DATA;
 		/* Offset changed at after rev 0 */
-		if (cc->dev->id.revision == 0)
+		if (ccrev == 0)
 			uart_regs += (i * 8);
 		else
 			uart_regs += (i * 256);
diff --git a/include/linux/ssb/ssb_driver_chipcommon.h b/include/linux/ssb/ssb_driver_chipcommon.h
index 4cb995494662..35717b400cef 100644
--- a/include/linux/ssb/ssb_driver_chipcommon.h
+++ b/include/linux/ssb/ssb_driver_chipcommon.h
@@ -51,9 +51,12 @@
 #define  SSB_CHIPCO_CAP_JTAGM		0x00400000	/* JTAG master present */
 #define  SSB_CHIPCO_CAP_BROM		0x00800000	/* Internal boot ROM active */
 #define  SSB_CHIPCO_CAP_64BIT		0x08000000	/* 64-bit Backplane */
+#define  SSB_CHIPCO_CAP_PMU		0x10000000	/* PMU available (rev >= 20) */
+#define  SSB_CHIPCO_CAP_ECI		0x20000000	/* ECI available (rev >= 20) */
 #define SSB_CHIPCO_CORECTL		0x0008
 #define  SSB_CHIPCO_CORECTL_UARTCLK0	0x00000001	/* Drive UART with internal clock */
 #define	 SSB_CHIPCO_CORECTL_SE		0x00000002	/* sync clk out enable (corerev >= 3) */
+#define  SSB_CHIPCO_CORECTL_UARTCLKEN	0x00000008	/* UART clock enable (rev >= 21) */
 #define SSB_CHIPCO_BIST			0x000C
 #define SSB_CHIPCO_OTPS			0x0010		/* OTP status */
 #define	 SSB_CHIPCO_OTPS_PROGFAIL	0x80000000
-- 
cgit v1.2.3


From 42bfad4f71637c4eb4791aa8062063c4a8526522 Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Tue, 19 Feb 2008 12:41:30 +0100
Subject: ssb: Fix watchdog access for devices without a chipcommon

This fixes the SSB watchdog access for devices without a chipcommon.
These devices have the watchdog on the extif.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/Kconfig                       |  6 ++++++
 drivers/ssb/Makefile                      |  1 +
 drivers/ssb/driver_extif.c                |  6 ++++++
 drivers/ssb/embedded.c                    | 26 ++++++++++++++++++++++++++
 include/linux/ssb/ssb_driver_chipcommon.h |  5 +++++
 include/linux/ssb/ssb_driver_extif.h      |  9 +++++++++
 include/linux/ssb/ssb_embedded.h          | 10 ++++++++++
 7 files changed, 63 insertions(+)
 create mode 100644 drivers/ssb/embedded.c
 create mode 100644 include/linux/ssb/ssb_embedded.h

(limited to 'include/linux')

diff --git a/drivers/ssb/Kconfig b/drivers/ssb/Kconfig
index d976660cb7f0..78fd33125e02 100644
--- a/drivers/ssb/Kconfig
+++ b/drivers/ssb/Kconfig
@@ -105,6 +105,12 @@ config SSB_DRIVER_MIPS
 
 	  If unsure, say N
 
+# Assumption: We are on embedded, if we compile the MIPS core.
+config SSB_EMBEDDED
+	bool
+	depends on SSB_DRIVER_MIPS
+	default y
+
 config SSB_DRIVER_EXTIF
 	bool "SSB Broadcom EXTIF core driver (EXPERIMENTAL)"
 	depends on SSB_DRIVER_MIPS && EXPERIMENTAL
diff --git a/drivers/ssb/Makefile b/drivers/ssb/Makefile
index 7be397595805..e235144add7c 100644
--- a/drivers/ssb/Makefile
+++ b/drivers/ssb/Makefile
@@ -1,5 +1,6 @@
 # core
 ssb-y					+= main.o scan.o
+ssb-$(CONFIG_SSB_EMBEDDED)		+= embedded.o
 
 # host support
 ssb-$(CONFIG_SSB_PCIHOST)		+= pci.o pcihost_wrapper.o
diff --git a/drivers/ssb/driver_extif.c b/drivers/ssb/driver_extif.c
index fe55eb8b038a..b1899f422a54 100644
--- a/drivers/ssb/driver_extif.c
+++ b/drivers/ssb/driver_extif.c
@@ -110,6 +110,12 @@ void ssb_extif_get_clockcontrol(struct ssb_extif *extif,
 	*m = extif_read32(extif, SSB_EXTIF_CLOCK_SB);
 }
 
+void ssb_extif_watchdog_timer_set(struct ssb_extif *extif,
+				  u32 ticks)
+{
+	extif_write32(extif, SSB_EXTIF_WATCHDOG, ticks);
+}
+
 u32 ssb_extif_gpio_in(struct ssb_extif *extif, u32 mask)
 {
 	return extif_read32(extif, SSB_EXTIF_GPIO_IN) & mask;
diff --git a/drivers/ssb/embedded.c b/drivers/ssb/embedded.c
new file mode 100644
index 000000000000..751f58ac612c
--- /dev/null
+++ b/drivers/ssb/embedded.c
@@ -0,0 +1,26 @@
+/*
+ * Sonics Silicon Backplane
+ * Embedded systems support code
+ *
+ * Copyright 2005-2008, Broadcom Corporation
+ * Copyright 2006-2008, Michael Buesch <mb@bu3sch.de>
+ *
+ * Licensed under the GNU/GPL. See COPYING for details.
+ */
+
+#include <linux/ssb/ssb.h>
+#include <linux/ssb/ssb_embedded.h>
+
+
+int ssb_watchdog_timer_set(struct ssb_bus *bus, u32 ticks)
+{
+	if (ssb_chipco_available(&bus->chipco)) {
+		ssb_chipco_watchdog_timer_set(&bus->chipco, ticks);
+		return 0;
+	}
+	if (ssb_extif_available(&bus->extif)) {
+		ssb_extif_watchdog_timer_set(&bus->extif, ticks);
+		return 0;
+	}
+	return -ENODEV;
+}
diff --git a/include/linux/ssb/ssb_driver_chipcommon.h b/include/linux/ssb/ssb_driver_chipcommon.h
index 35717b400cef..89638153cbe1 100644
--- a/include/linux/ssb/ssb_driver_chipcommon.h
+++ b/include/linux/ssb/ssb_driver_chipcommon.h
@@ -360,6 +360,11 @@ struct ssb_chipcommon {
 	u16 fast_pwrup_delay;
 };
 
+static inline bool ssb_chipco_available(struct ssb_chipcommon *cc)
+{
+	return (cc->dev != NULL);
+}
+
 extern void ssb_chipcommon_init(struct ssb_chipcommon *cc);
 
 #include <linux/pm.h>
diff --git a/include/linux/ssb/ssb_driver_extif.h b/include/linux/ssb/ssb_driver_extif.h
index a9164357b5ae..0d7c9bfa965d 100644
--- a/include/linux/ssb/ssb_driver_extif.h
+++ b/include/linux/ssb/ssb_driver_extif.h
@@ -171,6 +171,9 @@ extern void ssb_extif_get_clockcontrol(struct ssb_extif *extif,
 extern void ssb_extif_timing_init(struct ssb_extif *extif,
 				  unsigned long ns);
 
+extern void ssb_extif_watchdog_timer_set(struct ssb_extif *extif,
+					 u32 ticks);
+
 u32 ssb_extif_gpio_in(struct ssb_extif *extif, u32 mask);
 
 void ssb_extif_gpio_out(struct ssb_extif *extif, u32 mask, u32 value);
@@ -200,5 +203,11 @@ void ssb_extif_get_clockcontrol(struct ssb_extif *extif,
 {
 }
 
+static inline
+void ssb_extif_watchdog_timer_set(struct ssb_extif *extif,
+				  u32 ticks)
+{
+}
+
 #endif /* CONFIG_SSB_DRIVER_EXTIF */
 #endif /* LINUX_SSB_EXTIFCORE_H_ */
diff --git a/include/linux/ssb/ssb_embedded.h b/include/linux/ssb/ssb_embedded.h
new file mode 100644
index 000000000000..80bd58496450
--- /dev/null
+++ b/include/linux/ssb/ssb_embedded.h
@@ -0,0 +1,10 @@
+#ifndef LINUX_SSB_EMBEDDED_H_
+#define LINUX_SSB_EMBEDDED_H_
+
+#include <linux/types.h>
+#include <linux/ssb/ssb.h>
+
+
+extern int ssb_watchdog_timer_set(struct ssb_bus *bus, u32 ticks);
+
+#endif /* LINUX_SSB_EMBEDDED_H_ */
-- 
cgit v1.2.3


From c2bcbe65fc88d61f9a806367ff6eab76c9eabb3a Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Tue, 19 Feb 2008 14:53:35 +0100
Subject: ssb: Fix the GPIO API

This fixes the GPIO API to be usable.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/driver_chipcommon.c           | 35 +++++++++++++++++++++++++------
 drivers/ssb/driver_extif.c                | 24 +++++++++++++++++----
 include/linux/ssb/ssb_driver_chipcommon.h | 10 +++++----
 include/linux/ssb/ssb_driver_extif.h      |  9 ++++----
 4 files changed, 60 insertions(+), 18 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ssb/driver_chipcommon.c b/drivers/ssb/driver_chipcommon.c
index 7cc03f2dd5a6..7ea0c0faa9ab 100644
--- a/drivers/ssb/driver_chipcommon.c
+++ b/drivers/ssb/driver_chipcommon.c
@@ -39,12 +39,14 @@ static inline void chipco_write32(struct ssb_chipcommon *cc,
 	ssb_write32(cc->dev, offset, value);
 }
 
-static inline void chipco_write32_masked(struct ssb_chipcommon *cc, u16 offset,
-					 u32 mask, u32 value)
+static inline u32 chipco_write32_masked(struct ssb_chipcommon *cc, u16 offset,
+					u32 mask, u32 value)
 {
 	value &= mask;
 	value |= chipco_read32(cc, offset) & ~mask;
 	chipco_write32(cc, offset, value);
+
+	return value;
 }
 
 void ssb_chipco_set_clockmode(struct ssb_chipcommon *cc,
@@ -355,16 +357,37 @@ u32 ssb_chipco_gpio_in(struct ssb_chipcommon *cc, u32 mask)
 {
 	return chipco_read32(cc, SSB_CHIPCO_GPIOIN) & mask;
 }
+EXPORT_SYMBOL(ssb_chipco_gpio_in);
+
+u32 ssb_chipco_gpio_out(struct ssb_chipcommon *cc, u32 mask, u32 value)
+{
+	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOOUT, mask, value);
+}
+EXPORT_SYMBOL(ssb_chipco_gpio_out);
+
+u32 ssb_chipco_gpio_outen(struct ssb_chipcommon *cc, u32 mask, u32 value)
+{
+	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOOUTEN, mask, value);
+}
+EXPORT_SYMBOL(ssb_chipco_gpio_outen);
+
+u32 ssb_chipco_gpio_control(struct ssb_chipcommon *cc, u32 mask, u32 value)
+{
+	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOCTL, mask, value);
+}
+EXPORT_SYMBOL(ssb_chipco_gpio_control);
 
-void ssb_chipco_gpio_out(struct ssb_chipcommon *cc, u32 mask, u32 value)
+u32 ssb_chipco_gpio_intmask(struct ssb_chipcommon *cc, u32 mask, u32 value)
 {
-	chipco_write32_masked(cc, SSB_CHIPCO_GPIOOUT, mask, value);
+	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOIRQ, mask, value);
 }
+EXPORT_SYMBOL(ssb_chipco_gpio_intmask);
 
-void ssb_chipco_gpio_outen(struct ssb_chipcommon *cc, u32 mask, u32 value)
+u32 ssb_chipco_gpio_polarity(struct ssb_chipcommon *cc, u32 mask, u32 value)
 {
-	chipco_write32_masked(cc, SSB_CHIPCO_GPIOOUTEN, mask, value);
+	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOPOL, mask, value);
 }
+EXPORT_SYMBOL(ssb_chipco_gpio_polarity);
 
 #ifdef CONFIG_SSB_SERIAL
 int ssb_chipco_serial_init(struct ssb_chipcommon *cc,
diff --git a/drivers/ssb/driver_extif.c b/drivers/ssb/driver_extif.c
index b1899f422a54..10c6b287f8bb 100644
--- a/drivers/ssb/driver_extif.c
+++ b/drivers/ssb/driver_extif.c
@@ -27,12 +27,14 @@ static inline void extif_write32(struct ssb_extif *extif, u16 offset, u32 value)
 	ssb_write32(extif->dev, offset, value);
 }
 
-static inline void extif_write32_masked(struct ssb_extif *extif, u16 offset,
-					u32 mask, u32 value)
+static inline u32 extif_write32_masked(struct ssb_extif *extif, u16 offset,
+				       u32 mask, u32 value)
 {
 	value &= mask;
 	value |= extif_read32(extif, offset) & ~mask;
 	extif_write32(extif, offset, value);
+
+	return value;
 }
 
 #ifdef CONFIG_SSB_SERIAL
@@ -120,16 +122,30 @@ u32 ssb_extif_gpio_in(struct ssb_extif *extif, u32 mask)
 {
 	return extif_read32(extif, SSB_EXTIF_GPIO_IN) & mask;
 }
+EXPORT_SYMBOL(ssb_extif_gpio_in);
 
-void ssb_extif_gpio_out(struct ssb_extif *extif, u32 mask, u32 value)
+u32 ssb_extif_gpio_out(struct ssb_extif *extif, u32 mask, u32 value)
 {
 	return extif_write32_masked(extif, SSB_EXTIF_GPIO_OUT(0),
 				   mask, value);
 }
+EXPORT_SYMBOL(ssb_extif_gpio_out);
 
-void ssb_extif_gpio_outen(struct ssb_extif *extif, u32 mask, u32 value)
+u32 ssb_extif_gpio_outen(struct ssb_extif *extif, u32 mask, u32 value)
 {
 	return extif_write32_masked(extif, SSB_EXTIF_GPIO_OUTEN(0),
 				   mask, value);
 }
+EXPORT_SYMBOL(ssb_extif_gpio_outen);
+
+u32 ssb_extif_gpio_polarity(struct ssb_extif *extif, u32 mask, u32 value)
+{
+	return extif_write32_masked(extif, SSB_EXTIF_GPIO_INTPOL, mask, value);
+}
+EXPORT_SYMBOL(ssb_extif_gpio_polarity);
 
+u32 ssb_extif_gpio_intmask(struct ssb_extif *extif, u32 mask, u32 value)
+{
+	return extif_write32_masked(extif, SSB_EXTIF_GPIO_INTMASK, mask, value);
+}
+EXPORT_SYMBOL(ssb_extif_gpio_intmask);
diff --git a/include/linux/ssb/ssb_driver_chipcommon.h b/include/linux/ssb/ssb_driver_chipcommon.h
index 89638153cbe1..536851b946f6 100644
--- a/include/linux/ssb/ssb_driver_chipcommon.h
+++ b/include/linux/ssb/ssb_driver_chipcommon.h
@@ -390,11 +390,13 @@ extern void ssb_chipco_set_clockmode(struct ssb_chipcommon *cc,
 extern void ssb_chipco_watchdog_timer_set(struct ssb_chipcommon *cc,
 					  u32 ticks);
 
+/* Chipcommon GPIO pin access. */
 u32 ssb_chipco_gpio_in(struct ssb_chipcommon *cc, u32 mask);
-
-void ssb_chipco_gpio_out(struct ssb_chipcommon *cc, u32 mask, u32 value);
-
-void ssb_chipco_gpio_outen(struct ssb_chipcommon *cc, u32 mask, u32 value);
+u32 ssb_chipco_gpio_out(struct ssb_chipcommon *cc, u32 mask, u32 value);
+u32 ssb_chipco_gpio_outen(struct ssb_chipcommon *cc, u32 mask, u32 value);
+u32 ssb_chipco_gpio_control(struct ssb_chipcommon *cc, u32 mask, u32 value);
+u32 ssb_chipco_gpio_intmask(struct ssb_chipcommon *cc, u32 mask, u32 value);
+u32 ssb_chipco_gpio_polarity(struct ssb_chipcommon *cc, u32 mask, u32 value);
 
 #ifdef CONFIG_SSB_SERIAL
 extern int ssb_chipco_serial_init(struct ssb_chipcommon *cc,
diff --git a/include/linux/ssb/ssb_driver_extif.h b/include/linux/ssb/ssb_driver_extif.h
index 0d7c9bfa965d..91161f0aa22b 100644
--- a/include/linux/ssb/ssb_driver_extif.h
+++ b/include/linux/ssb/ssb_driver_extif.h
@@ -174,11 +174,12 @@ extern void ssb_extif_timing_init(struct ssb_extif *extif,
 extern void ssb_extif_watchdog_timer_set(struct ssb_extif *extif,
 					 u32 ticks);
 
+/* Extif GPIO pin access */
 u32 ssb_extif_gpio_in(struct ssb_extif *extif, u32 mask);
-
-void ssb_extif_gpio_out(struct ssb_extif *extif, u32 mask, u32 value);
-
-void ssb_extif_gpio_outen(struct ssb_extif *extif, u32 mask, u32 value);
+u32 ssb_extif_gpio_out(struct ssb_extif *extif, u32 mask, u32 value);
+u32 ssb_extif_gpio_outen(struct ssb_extif *extif, u32 mask, u32 value);
+u32 ssb_extif_gpio_polarity(struct ssb_extif *extif, u32 mask, u32 value);
+u32 ssb_extif_gpio_intmask(struct ssb_extif *extif, u32 mask, u32 value);
 
 #ifdef CONFIG_SSB_SERIAL
 extern int ssb_extif_serial_init(struct ssb_extif *extif,
-- 
cgit v1.2.3


From 53521d8c90d366191b6c134f88a8ebe83de60614 Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Tue, 19 Feb 2008 16:22:50 +0100
Subject: ssb: Make the GPIO API reentrancy safe

This fixes the GPIO API to be reentrancy safe.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/driver_chipcommon.c  |   6 ---
 drivers/ssb/driver_extif.c       |   5 --
 drivers/ssb/embedded.c           | 106 +++++++++++++++++++++++++++++++++++++++
 drivers/ssb/main.c               |   3 ++
 include/linux/ssb/ssb.h          |   5 ++
 include/linux/ssb/ssb_embedded.h |   8 +++
 6 files changed, 122 insertions(+), 11 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ssb/driver_chipcommon.c b/drivers/ssb/driver_chipcommon.c
index 7ea0c0faa9ab..e586321a473a 100644
--- a/drivers/ssb/driver_chipcommon.c
+++ b/drivers/ssb/driver_chipcommon.c
@@ -357,37 +357,31 @@ u32 ssb_chipco_gpio_in(struct ssb_chipcommon *cc, u32 mask)
 {
 	return chipco_read32(cc, SSB_CHIPCO_GPIOIN) & mask;
 }
-EXPORT_SYMBOL(ssb_chipco_gpio_in);
 
 u32 ssb_chipco_gpio_out(struct ssb_chipcommon *cc, u32 mask, u32 value)
 {
 	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOOUT, mask, value);
 }
-EXPORT_SYMBOL(ssb_chipco_gpio_out);
 
 u32 ssb_chipco_gpio_outen(struct ssb_chipcommon *cc, u32 mask, u32 value)
 {
 	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOOUTEN, mask, value);
 }
-EXPORT_SYMBOL(ssb_chipco_gpio_outen);
 
 u32 ssb_chipco_gpio_control(struct ssb_chipcommon *cc, u32 mask, u32 value)
 {
 	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOCTL, mask, value);
 }
-EXPORT_SYMBOL(ssb_chipco_gpio_control);
 
 u32 ssb_chipco_gpio_intmask(struct ssb_chipcommon *cc, u32 mask, u32 value)
 {
 	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOIRQ, mask, value);
 }
-EXPORT_SYMBOL(ssb_chipco_gpio_intmask);
 
 u32 ssb_chipco_gpio_polarity(struct ssb_chipcommon *cc, u32 mask, u32 value)
 {
 	return chipco_write32_masked(cc, SSB_CHIPCO_GPIOPOL, mask, value);
 }
-EXPORT_SYMBOL(ssb_chipco_gpio_polarity);
 
 #ifdef CONFIG_SSB_SERIAL
 int ssb_chipco_serial_init(struct ssb_chipcommon *cc,
diff --git a/drivers/ssb/driver_extif.c b/drivers/ssb/driver_extif.c
index 10c6b287f8bb..c3e1d3e6d610 100644
--- a/drivers/ssb/driver_extif.c
+++ b/drivers/ssb/driver_extif.c
@@ -122,30 +122,25 @@ u32 ssb_extif_gpio_in(struct ssb_extif *extif, u32 mask)
 {
 	return extif_read32(extif, SSB_EXTIF_GPIO_IN) & mask;
 }
-EXPORT_SYMBOL(ssb_extif_gpio_in);
 
 u32 ssb_extif_gpio_out(struct ssb_extif *extif, u32 mask, u32 value)
 {
 	return extif_write32_masked(extif, SSB_EXTIF_GPIO_OUT(0),
 				   mask, value);
 }
-EXPORT_SYMBOL(ssb_extif_gpio_out);
 
 u32 ssb_extif_gpio_outen(struct ssb_extif *extif, u32 mask, u32 value)
 {
 	return extif_write32_masked(extif, SSB_EXTIF_GPIO_OUTEN(0),
 				   mask, value);
 }
-EXPORT_SYMBOL(ssb_extif_gpio_outen);
 
 u32 ssb_extif_gpio_polarity(struct ssb_extif *extif, u32 mask, u32 value)
 {
 	return extif_write32_masked(extif, SSB_EXTIF_GPIO_INTPOL, mask, value);
 }
-EXPORT_SYMBOL(ssb_extif_gpio_polarity);
 
 u32 ssb_extif_gpio_intmask(struct ssb_extif *extif, u32 mask, u32 value)
 {
 	return extif_write32_masked(extif, SSB_EXTIF_GPIO_INTMASK, mask, value);
 }
-EXPORT_SYMBOL(ssb_extif_gpio_intmask);
diff --git a/drivers/ssb/embedded.c b/drivers/ssb/embedded.c
index 751f58ac612c..d3ade821555c 100644
--- a/drivers/ssb/embedded.c
+++ b/drivers/ssb/embedded.c
@@ -11,6 +11,8 @@
 #include <linux/ssb/ssb.h>
 #include <linux/ssb/ssb_embedded.h>
 
+#include "ssb_private.h"
+
 
 int ssb_watchdog_timer_set(struct ssb_bus *bus, u32 ticks)
 {
@@ -24,3 +26,107 @@ int ssb_watchdog_timer_set(struct ssb_bus *bus, u32 ticks)
 	}
 	return -ENODEV;
 }
+
+u32 ssb_gpio_in(struct ssb_bus *bus, u32 mask)
+{
+	unsigned long flags;
+	u32 res = 0;
+
+	spin_lock_irqsave(&bus->gpio_lock, flags);
+	if (ssb_chipco_available(&bus->chipco))
+		res = ssb_chipco_gpio_in(&bus->chipco, mask);
+	else if (ssb_extif_available(&bus->extif))
+		res = ssb_extif_gpio_in(&bus->extif, mask);
+	else
+		SSB_WARN_ON(1);
+	spin_unlock_irqrestore(&bus->gpio_lock, flags);
+
+	return res;
+}
+EXPORT_SYMBOL(ssb_gpio_in);
+
+u32 ssb_gpio_out(struct ssb_bus *bus, u32 mask, u32 value)
+{
+	unsigned long flags;
+	u32 res = 0;
+
+	spin_lock_irqsave(&bus->gpio_lock, flags);
+	if (ssb_chipco_available(&bus->chipco))
+		res = ssb_chipco_gpio_out(&bus->chipco, mask, value);
+	else if (ssb_extif_available(&bus->extif))
+		res = ssb_extif_gpio_out(&bus->extif, mask, value);
+	else
+		SSB_WARN_ON(1);
+	spin_unlock_irqrestore(&bus->gpio_lock, flags);
+
+	return res;
+}
+EXPORT_SYMBOL(ssb_gpio_out);
+
+u32 ssb_gpio_outen(struct ssb_bus *bus, u32 mask, u32 value)
+{
+	unsigned long flags;
+	u32 res = 0;
+
+	spin_lock_irqsave(&bus->gpio_lock, flags);
+	if (ssb_chipco_available(&bus->chipco))
+		res = ssb_chipco_gpio_outen(&bus->chipco, mask, value);
+	else if (ssb_extif_available(&bus->extif))
+		res = ssb_extif_gpio_outen(&bus->extif, mask, value);
+	else
+		SSB_WARN_ON(1);
+	spin_unlock_irqrestore(&bus->gpio_lock, flags);
+
+	return res;
+}
+EXPORT_SYMBOL(ssb_gpio_outen);
+
+u32 ssb_gpio_control(struct ssb_bus *bus, u32 mask, u32 value)
+{
+	unsigned long flags;
+	u32 res = 0;
+
+	spin_lock_irqsave(&bus->gpio_lock, flags);
+	if (ssb_chipco_available(&bus->chipco))
+		res = ssb_chipco_gpio_control(&bus->chipco, mask, value);
+	spin_unlock_irqrestore(&bus->gpio_lock, flags);
+
+	return res;
+}
+EXPORT_SYMBOL(ssb_gpio_control);
+
+u32 ssb_gpio_intmask(struct ssb_bus *bus, u32 mask, u32 value)
+{
+	unsigned long flags;
+	u32 res = 0;
+
+	spin_lock_irqsave(&bus->gpio_lock, flags);
+	if (ssb_chipco_available(&bus->chipco))
+		res = ssb_chipco_gpio_intmask(&bus->chipco, mask, value);
+	else if (ssb_extif_available(&bus->extif))
+		res = ssb_extif_gpio_intmask(&bus->extif, mask, value);
+	else
+		SSB_WARN_ON(1);
+	spin_unlock_irqrestore(&bus->gpio_lock, flags);
+
+	return res;
+}
+EXPORT_SYMBOL(ssb_gpio_intmask);
+
+u32 ssb_gpio_polarity(struct ssb_bus *bus, u32 mask, u32 value)
+{
+	unsigned long flags;
+	u32 res = 0;
+
+	spin_lock_irqsave(&bus->gpio_lock, flags);
+	if (ssb_chipco_available(&bus->chipco))
+		res = ssb_chipco_gpio_polarity(&bus->chipco, mask, value);
+	else if (ssb_extif_available(&bus->extif))
+		res = ssb_extif_gpio_polarity(&bus->extif, mask, value);
+	else
+		SSB_WARN_ON(1);
+	spin_unlock_irqrestore(&bus->gpio_lock, flags);
+
+	return res;
+}
+EXPORT_SYMBOL(ssb_gpio_polarity);
diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index 9028ed5715a1..af07ab22708f 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -569,6 +569,9 @@ static int ssb_bus_register(struct ssb_bus *bus,
 
 	spin_lock_init(&bus->bar_lock);
 	INIT_LIST_HEAD(&bus->list);
+#ifdef CONFIG_SSB_EMBEDDED
+	spin_lock_init(&bus->gpio_lock);
+#endif
 
 	/* Powerup the bus */
 	err = ssb_pci_xtal(bus, SSB_GPIO_XTAL | SSB_GPIO_PLL, 1);
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index 9d5da8b2ccf9..d14c03685717 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -283,6 +283,11 @@ struct ssb_bus {
 	/* Contents of the SPROM. */
 	struct ssb_sprom sprom;
 
+#ifdef CONFIG_SSB_EMBEDDED
+	/* Lock for GPIO register access. */
+	spinlock_t gpio_lock;
+#endif /* EMBEDDED */
+
 	/* Internal-only stuff follows. Do not touch. */
 	struct list_head list;
 #ifdef CONFIG_SSB_DEBUG
diff --git a/include/linux/ssb/ssb_embedded.h b/include/linux/ssb/ssb_embedded.h
index 80bd58496450..8d8dedff059d 100644
--- a/include/linux/ssb/ssb_embedded.h
+++ b/include/linux/ssb/ssb_embedded.h
@@ -7,4 +7,12 @@
 
 extern int ssb_watchdog_timer_set(struct ssb_bus *bus, u32 ticks);
 
+/* Generic GPIO API */
+u32 ssb_gpio_in(struct ssb_bus *bus, u32 mask);
+u32 ssb_gpio_out(struct ssb_bus *bus, u32 mask, u32 value);
+u32 ssb_gpio_outen(struct ssb_bus *bus, u32 mask, u32 value);
+u32 ssb_gpio_control(struct ssb_bus *bus, u32 mask, u32 value);
+u32 ssb_gpio_intmask(struct ssb_bus *bus, u32 mask, u32 value);
+u32 ssb_gpio_polarity(struct ssb_bus *bus, u32 mask, u32 value);
+
 #endif /* LINUX_SSB_EMBEDDED_H_ */
-- 
cgit v1.2.3


From 7cb4461520f307a6e3fb2bb32cb8daee45aa1fae Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Tue, 19 Feb 2008 17:46:48 +0100
Subject: ssb: Fix pcicore cardbus mode

This fixes the pcicore driver to not die a horrible
crash death when inserting a cardbus card.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/driver_pcicore.c       | 28 ++++++++++++++++++++++++++--
 drivers/ssb/main.c                 |  1 +
 include/linux/ssb/ssb.h            |  7 +++++++
 include/linux/ssb/ssb_driver_pci.h |  5 +++++
 4 files changed, 39 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ssb/driver_pcicore.c b/drivers/ssb/driver_pcicore.c
index 2faaa906d5d6..059452fbb168 100644
--- a/drivers/ssb/driver_pcicore.c
+++ b/drivers/ssb/driver_pcicore.c
@@ -11,6 +11,7 @@
 #include <linux/ssb/ssb.h>
 #include <linux/pci.h>
 #include <linux/delay.h>
+#include <linux/ssb/ssb_embedded.h>
 
 #include "ssb_private.h"
 
@@ -27,6 +28,18 @@ void pcicore_write32(struct ssb_pcicore *pc, u16 offset, u32 value)
 	ssb_write32(pc->dev, offset, value);
 }
 
+static inline
+u16 pcicore_read16(struct ssb_pcicore *pc, u16 offset)
+{
+	return ssb_read16(pc->dev, offset);
+}
+
+static inline
+void pcicore_write16(struct ssb_pcicore *pc, u16 offset, u16 value)
+{
+	ssb_write16(pc->dev, offset, value);
+}
+
 /**************************************************
  * Code for hostmode operation.
  **************************************************/
@@ -117,8 +130,10 @@ static u32 get_cfgspace_addr(struct ssb_pcicore *pc,
 	u32 addr = 0;
 	u32 tmp;
 
-	if (unlikely(pc->cardbusmode && dev > 1))
+	/* We do only have one cardbus device behind the bridge. */
+	if (pc->cardbusmode && (dev >= 1))
 		goto out;
+
 	if (bus == 0) {
 		/* Type 0 transaction */
 		if (unlikely(dev >= SSB_PCI_SLOT_MAX))
@@ -318,7 +333,16 @@ static void ssb_pcicore_init_hostmode(struct ssb_pcicore *pc)
 	pcicore_write32(pc, SSB_PCICORE_ARBCTL, val);
 	udelay(1); /* Assertion time demanded by the PCI standard */
 
-	/*TODO cardbus mode */
+	if (pc->dev->bus->has_cardbus_slot) {
+		ssb_dprintk(KERN_INFO PFX "CardBus slot detected\n");
+		pc->cardbusmode = 1;
+		/* GPIO 1 resets the bridge */
+		ssb_gpio_out(pc->dev->bus, 1, 1);
+		ssb_gpio_outen(pc->dev->bus, 1, 1);
+		pcicore_write16(pc, SSB_PCICORE_SPROM(0),
+				pcicore_read16(pc, SSB_PCICORE_SPROM(0))
+				| 0x0400);
+	}
 
 	/* 64MB I/O window */
 	pcicore_write32(pc, SSB_PCICORE_SBTOPCI0,
diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index af07ab22708f..bedb2b4ee9d2 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -557,6 +557,7 @@ static int ssb_fetch_invariants(struct ssb_bus *bus,
 		goto out;
 	memcpy(&bus->boardinfo, &iv.boardinfo, sizeof(iv.boardinfo));
 	memcpy(&bus->sprom, &iv.sprom, sizeof(iv.sprom));
+	bus->has_cardbus_slot = iv.has_cardbus_slot;
 out:
 	return err;
 }
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index d14c03685717..20add65215af 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -282,6 +282,8 @@ struct ssb_bus {
 	struct ssb_boardinfo boardinfo;
 	/* Contents of the SPROM. */
 	struct ssb_sprom sprom;
+	/* If the board has a cardbus slot, this is set to true. */
+	bool has_cardbus_slot;
 
 #ifdef CONFIG_SSB_EMBEDDED
 	/* Lock for GPIO register access. */
@@ -299,8 +301,13 @@ struct ssb_bus {
 
 /* The initialization-invariants. */
 struct ssb_init_invariants {
+	/* Versioning information about the PCB. */
 	struct ssb_boardinfo boardinfo;
+	/* The SPROM information. That's either stored in an
+	 * EEPROM or NVRAM on the board. */
 	struct ssb_sprom sprom;
+	/* If the board has a cardbus slot, this is set to true. */
+	bool has_cardbus_slot;
 };
 /* Type of function to fetch the invariants. */
 typedef int (*ssb_invariants_func_t)(struct ssb_bus *bus,
diff --git a/include/linux/ssb/ssb_driver_pci.h b/include/linux/ssb/ssb_driver_pci.h
index 9cfffb7b1a27..5e25bac4ed31 100644
--- a/include/linux/ssb/ssb_driver_pci.h
+++ b/include/linux/ssb/ssb_driver_pci.h
@@ -51,6 +51,11 @@
 #define  SSB_PCICORE_SBTOPCI1_MASK	0xFC000000
 #define SSB_PCICORE_SBTOPCI2		0x0108	/* Backplane to PCI translation 2 (sbtopci2) */
 #define  SSB_PCICORE_SBTOPCI2_MASK	0xC0000000
+#define SSB_PCICORE_PCICFG0		0x0400	/* PCI config space 0 (rev >= 8) */
+#define SSB_PCICORE_PCICFG1		0x0500	/* PCI config space 1 (rev >= 8) */
+#define SSB_PCICORE_PCICFG2		0x0600	/* PCI config space 2 (rev >= 8) */
+#define SSB_PCICORE_PCICFG3		0x0700	/* PCI config space 3 (rev >= 8) */
+#define SSB_PCICORE_SPROM(wordoffset)	(0x0800 + ((wordoffset) * 2)) /* SPROM shadow area (72 bytes) */
 
 /* SBtoPCIx */
 #define SSB_PCICORE_SBTOPCI_MEM		0x00000000
-- 
cgit v1.2.3


From 8a235efad548abd2ab5ebea45a9ffa750c814375 Mon Sep 17 00:00:00 2001
From: "Rafael J. Wysocki" <rjw@sisk.pl>
Date: Wed, 20 Feb 2008 01:47:44 +0100
Subject: Hibernation: Handle DEBUG_PAGEALLOC on x86

Make hibernation work with CONFIG_DEBUG_PAGEALLOC set on x86, by
checking if the pages to be copied are marked as present in the
kernel mapping and temporarily marking them as present if that's not
the case.  No functional modifications are introduced if
CONFIG_DEBUG_PAGEALLOC is unset.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Len Brown <len.brown@intel.com>
---
 arch/x86/mm/pageattr.c  | 19 ++++++++++++++++++-
 include/linux/mm.h      |  6 ++++++
 kernel/power/snapshot.c | 42 +++++++++++++++++++++++++++++-------------
 3 files changed, 53 insertions(+), 14 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c
index e2a74ea11a53..464d8fc21ce6 100644
--- a/arch/x86/mm/pageattr.c
+++ b/arch/x86/mm/pageattr.c
@@ -899,7 +899,24 @@ void kernel_map_pages(struct page *page, int numpages, int enable)
 	 */
 	cpa_fill_pool();
 }
-#endif
+
+#ifdef CONFIG_HIBERNATION
+
+bool kernel_page_present(struct page *page)
+{
+	unsigned int level;
+	pte_t *pte;
+
+	if (PageHighMem(page))
+		return false;
+
+	pte = lookup_address((unsigned long)page_address(page), &level);
+	return (pte_val(*pte) & _PAGE_PRESENT);
+}
+
+#endif /* CONFIG_HIBERNATION */
+
+#endif /* CONFIG_DEBUG_PAGEALLOC */
 
 /*
  * The testcases use internal knowledge of the implementation that shouldn't
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 26c7124b841a..3b3e1341163f 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1171,12 +1171,18 @@ static inline void enable_debug_pagealloc(void)
 {
 	debug_pagealloc_enabled = 1;
 }
+#ifdef CONFIG_HIBERNATION
+extern bool kernel_page_present(struct page *page);
+#endif /* CONFIG_HIBERNATION */
 #else
 static inline void
 kernel_map_pages(struct page *page, int numpages, int enable) {}
 static inline void enable_debug_pagealloc(void)
 {
 }
+#ifdef CONFIG_HIBERNATION
+static inline bool kernel_page_present(struct page *page) { return true; }
+#endif /* CONFIG_HIBERNATION */
 #endif
 
 extern struct vm_area_struct *get_gate_vma(struct task_struct *tsk);
diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c
index 95250d7c8d91..72a020cabb4c 100644
--- a/kernel/power/snapshot.c
+++ b/kernel/power/snapshot.c
@@ -875,8 +875,8 @@ static inline void *saveable_highmem_page(unsigned long pfn) { return NULL; }
 #endif /* CONFIG_HIGHMEM */
 
 /**
- *	saveable - Determine whether a non-highmem page should be included in
- *	the suspend image.
+ *	saveable_page - Determine whether a non-highmem page should be included
+ *	in the suspend image.
  *
  *	We should save the page if it isn't Nosave, and is not in the range
  *	of pages statically defined as 'unsaveable', and it isn't a part of
@@ -897,7 +897,8 @@ static struct page *saveable_page(unsigned long pfn)
 	if (swsusp_page_is_forbidden(page) || swsusp_page_is_free(page))
 		return NULL;
 
-	if (PageReserved(page) && pfn_is_nosave(pfn))
+	if (PageReserved(page)
+	    && (!kernel_page_present(page) || pfn_is_nosave(pfn)))
 		return NULL;
 
 	return page;
@@ -938,6 +939,25 @@ static inline void do_copy_page(long *dst, long *src)
 		*dst++ = *src++;
 }
 
+
+/**
+ *	safe_copy_page - check if the page we are going to copy is marked as
+ *		present in the kernel page tables (this always is the case if
+ *		CONFIG_DEBUG_PAGEALLOC is not set and in that case
+ *		kernel_page_present() always returns 'true').
+ */
+static void safe_copy_page(void *dst, struct page *s_page)
+{
+	if (kernel_page_present(s_page)) {
+		do_copy_page(dst, page_address(s_page));
+	} else {
+		kernel_map_pages(s_page, 1, 1);
+		do_copy_page(dst, page_address(s_page));
+		kernel_map_pages(s_page, 1, 0);
+	}
+}
+
+
 #ifdef CONFIG_HIGHMEM
 static inline struct page *
 page_is_saveable(struct zone *zone, unsigned long pfn)
@@ -946,8 +966,7 @@ page_is_saveable(struct zone *zone, unsigned long pfn)
 			saveable_highmem_page(pfn) : saveable_page(pfn);
 }
 
-static inline void
-copy_data_page(unsigned long dst_pfn, unsigned long src_pfn)
+static void copy_data_page(unsigned long dst_pfn, unsigned long src_pfn)
 {
 	struct page *s_page, *d_page;
 	void *src, *dst;
@@ -961,29 +980,26 @@ copy_data_page(unsigned long dst_pfn, unsigned long src_pfn)
 		kunmap_atomic(src, KM_USER0);
 		kunmap_atomic(dst, KM_USER1);
 	} else {
-		src = page_address(s_page);
 		if (PageHighMem(d_page)) {
 			/* Page pointed to by src may contain some kernel
 			 * data modified by kmap_atomic()
 			 */
-			do_copy_page(buffer, src);
+			safe_copy_page(buffer, s_page);
 			dst = kmap_atomic(pfn_to_page(dst_pfn), KM_USER0);
 			memcpy(dst, buffer, PAGE_SIZE);
 			kunmap_atomic(dst, KM_USER0);
 		} else {
-			dst = page_address(d_page);
-			do_copy_page(dst, src);
+			safe_copy_page(page_address(d_page), s_page);
 		}
 	}
 }
 #else
 #define page_is_saveable(zone, pfn)	saveable_page(pfn)
 
-static inline void
-copy_data_page(unsigned long dst_pfn, unsigned long src_pfn)
+static inline void copy_data_page(unsigned long dst_pfn, unsigned long src_pfn)
 {
-	do_copy_page(page_address(pfn_to_page(dst_pfn)),
-			page_address(pfn_to_page(src_pfn)));
+	safe_copy_page(page_address(pfn_to_page(dst_pfn)),
+				pfn_to_page(src_pfn));
 }
 #endif /* CONFIG_HIGHMEM */
 
-- 
cgit v1.2.3


From eb87f65f78bb4fc6dc532e6d7082d90ba0890cb8 Mon Sep 17 00:00:00 2001
From: Jason Gaston <jason.d.gaston@intel.com>
Date: Mon, 28 Jan 2008 17:20:36 -0800
Subject: PCI: pci_ids: patch for Intel ICH10 DeviceID's

This patch adds the Intel ICH10 LPC and SMBus Controller DeviceID's.

Signed-off-by:  Jason Gaston <jason.d.gaston@intel.com>
Cc: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/pci_ids.h | 6 ++++++
 1 file changed, 6 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index 0519e53d5eb2..effdb558a588 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -2374,6 +2374,12 @@
 #define PCI_DEVICE_ID_INTEL_MCH_PC1	0x359a
 #define PCI_DEVICE_ID_INTEL_E7525_MCH	0x359e
 #define PCI_DEVICE_ID_INTEL_IOAT_CNB	0x360b
+#define PCI_DEVICE_ID_INTEL_ICH10_0	0x3a14
+#define PCI_DEVICE_ID_INTEL_ICH10_1	0x3a16
+#define PCI_DEVICE_ID_INTEL_ICH10_2	0x3a18
+#define PCI_DEVICE_ID_INTEL_ICH10_3	0x3a1a
+#define PCI_DEVICE_ID_INTEL_ICH10_4	0x3a30
+#define PCI_DEVICE_ID_INTEL_ICH10_5	0x3a60
 #define PCI_DEVICE_ID_INTEL_IOAT_SNB	0x402f
 #define PCI_DEVICE_ID_INTEL_IOAT_SCNB	0x65ff
 #define PCI_DEVICE_ID_INTEL_TOLAPAI_0	0x5031
-- 
cgit v1.2.3


From a62f735cbb1a0937ab6a9196dadf752694ef0b7f Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Thu, 21 Feb 2008 16:12:45 +0000
Subject: MM: Fix macro argument substitution in PageHead() and PageTail()

Fix macro argument substitution in PageHead() and PageTail() - 'page' should
have brackets surrounding it (commit 6d7779538f765963ced45a3fa4bed7ba8d2c277d).

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/page-flags.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
index bbad43fb8181..b5b30f1c1e59 100644
--- a/include/linux/page-flags.h
+++ b/include/linux/page-flags.h
@@ -266,7 +266,7 @@ static inline void SetPageUptodate(struct page *page)
 
 #define PG_head_tail_mask ((1L << PG_compound) | (1L << PG_reclaim))
 
-#define PageTail(page)	((page->flags & PG_head_tail_mask) \
+#define PageTail(page)	(((page)->flags & PG_head_tail_mask)	\
 				== PG_head_tail_mask)
 
 static inline void __SetPageTail(struct page *page)
@@ -279,7 +279,7 @@ static inline void __ClearPageTail(struct page *page)
 	page->flags &= ~PG_head_tail_mask;
 }
 
-#define PageHead(page)	((page->flags & PG_head_tail_mask) \
+#define PageHead(page)	(((page)->flags & PG_head_tail_mask)	\
 				== (1L << PG_compound))
 #define __SetPageHead(page)	__SetPageCompound(page)
 #define __ClearPageHead(page)	__ClearPageCompound(page)
-- 
cgit v1.2.3


From 3a2d5b700132f35401f1d9e22fe3c2cab02c2549 Mon Sep 17 00:00:00 2001
From: "Rafael J. Wysocki" <rjw@sisk.pl>
Date: Sat, 23 Feb 2008 19:13:25 +0100
Subject: PM: Introduce PM_EVENT_HIBERNATE callback state

During the last step of hibernation in the "platform" mode (with the
help of ACPI) we use the suspend code, including the devices'
->suspend() methods, to prepare the system for entering the ACPI S4
system sleep state.

But at least for some devices the operations performed by the
->suspend() callback in that case must be different from its operations
during regular suspend.

For this reason, introduce the new PM event type PM_EVENT_HIBERNATE and
pass it to the device drivers' ->suspend() methods during the last phase
of hibernation, so that they can distinguish this case and handle it as
appropriate.  Modify the drivers that handle PM_EVENT_SUSPEND in a
special way and need to handle PM_EVENT_HIBERNATE in the same way.

These changes are necessary to fix a hibernation regression related
to the i915 driver (ref. http://lkml.org/lkml/2008/2/22/488).

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Pavel Machek <pavel@ucw.cz>
Tested-by: Jeff Chua <jeff.chua.linux@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 Documentation/power/devices.txt        | 13 ++++++++-----
 drivers/ata/ahci.c                     |  2 +-
 drivers/ata/ata_piix.c                 |  2 +-
 drivers/ata/libata-core.c              |  2 +-
 drivers/ide/ppc/pmac.c                 |  4 ++--
 drivers/macintosh/mediabay.c           |  3 ++-
 drivers/pci/pci.c                      |  1 +
 drivers/scsi/aic7xxx/aic79xx_osm_pci.c |  2 +-
 drivers/scsi/aic7xxx/aic7xxx_osm_pci.c |  2 +-
 drivers/scsi/mesh.c                    |  1 +
 drivers/scsi/sd.c                      |  3 +--
 drivers/usb/host/sl811-hcd.c           |  1 +
 drivers/usb/host/u132-hcd.c            | 11 ++++++++---
 drivers/video/chipsfb.c                |  2 +-
 drivers/video/nvidia/nvidia.c          |  2 +-
 include/linux/pm.h                     |  9 ++++++++-
 kernel/power/disk.c                    |  4 ++--
 net/rfkill/rfkill.c                    |  2 +-
 18 files changed, 42 insertions(+), 24 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/power/devices.txt b/Documentation/power/devices.txt
index c53d26361919..461e4f1dbec4 100644
--- a/Documentation/power/devices.txt
+++ b/Documentation/power/devices.txt
@@ -310,9 +310,12 @@ used with suspend-to-disk:
     PM_EVENT_SUSPEND -- quiesce the driver and put hardware into a low-power
 	state.  When used with system sleep states like "suspend-to-RAM" or
 	"standby", the upcoming resume() call will often be able to rely on
-	state kept in hardware, or issue system wakeup events.  When used
-	instead with suspend-to-disk, few devices support this capability;
-	most are completely powered off.
+	state kept in hardware, or issue system wakeup events.
+
+    PM_EVENT_HIBERNATE -- Put hardware into a low-power state and enable wakeup
+	events as appropriate.  It is only used with hibernation
+	(suspend-to-disk) and few devices are able to wake up the system from
+	this state; most are completely powered off.
 
     PM_EVENT_FREEZE -- quiesce the driver, but don't necessarily change into
 	any low power mode.  A system snapshot is about to be taken, often
@@ -329,8 +332,8 @@ used with suspend-to-disk:
 	wakeup events nor DMA are allowed.
 
 To enter "standby" (ACPI S1) or "Suspend to RAM" (STR, ACPI S3) states, or
-the similarly named APM states, only PM_EVENT_SUSPEND is used; for "Suspend
-to Disk" (STD, hibernate, ACPI S4), all of those event codes are used.
+the similarly named APM states, only PM_EVENT_SUSPEND is used; the other event
+codes are used for hibernation ("Suspend to Disk", STD, ACPI S4).
 
 There's also PM_EVENT_ON, a value which never appears as a suspend event
 but is sometimes used to record the "not suspended" device state.
diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 3c06e457b4dc..6dd12f7019a0 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1932,7 +1932,7 @@ static int ahci_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 	void __iomem *mmio = host->iomap[AHCI_PCI_BAR];
 	u32 ctl;
 
-	if (mesg.event == PM_EVENT_SUSPEND) {
+	if (mesg.event & PM_EVENT_SLEEP) {
 		/* AHCI spec rev1.1 section 8.3.3:
 		 * Software must disable interrupts prior to requesting a
 		 * transition of the HBA to D3 state.
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index 752e7d2f3b2f..fae8404254c0 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -1339,7 +1339,7 @@ static int piix_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg)
 	 * cycles and power trying to do something to the sleeping
 	 * beauty.
 	 */
-	if (piix_broken_suspend() && mesg.event == PM_EVENT_SUSPEND) {
+	if (piix_broken_suspend() && (mesg.event & PM_EVENT_SLEEP)) {
 		pci_save_state(pdev);
 
 		/* mark its power state as "unknown", since we don't
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 60d1bb556973..4cf8662df99e 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -7368,7 +7368,7 @@ void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t mesg)
 	pci_save_state(pdev);
 	pci_disable_device(pdev);
 
-	if (mesg.event == PM_EVENT_SUSPEND)
+	if (mesg.event & PM_EVENT_SLEEP)
 		pci_set_power_state(pdev, PCI_D3hot);
 }
 
diff --git a/drivers/ide/ppc/pmac.c b/drivers/ide/ppc/pmac.c
index 12ac3bfb4f9a..78c9eeb85634 100644
--- a/drivers/ide/ppc/pmac.c
+++ b/drivers/ide/ppc/pmac.c
@@ -1254,7 +1254,7 @@ pmac_ide_macio_suspend(struct macio_dev *mdev, pm_message_t mesg)
 	int		rc = 0;
 
 	if (mesg.event != mdev->ofdev.dev.power.power_state.event
-			&& mesg.event == PM_EVENT_SUSPEND) {
+			&& (mesg.event & PM_EVENT_SLEEP)) {
 		rc = pmac_ide_do_suspend(hwif);
 		if (rc == 0)
 			mdev->ofdev.dev.power.power_state = mesg;
@@ -1364,7 +1364,7 @@ pmac_ide_pci_suspend(struct pci_dev *pdev, pm_message_t mesg)
 	int		rc = 0;
 	
 	if (mesg.event != pdev->dev.power.power_state.event
-			&& mesg.event == PM_EVENT_SUSPEND) {
+			&& (mesg.event & PM_EVENT_SLEEP)) {
 		rc = pmac_ide_do_suspend(hwif);
 		if (rc == 0)
 			pdev->dev.power.power_state = mesg;
diff --git a/drivers/macintosh/mediabay.c b/drivers/macintosh/mediabay.c
index 51a112815f46..bd8a1d14b45d 100644
--- a/drivers/macintosh/mediabay.c
+++ b/drivers/macintosh/mediabay.c
@@ -698,7 +698,8 @@ static int media_bay_suspend(struct macio_dev *mdev, pm_message_t state)
 {
 	struct media_bay_info	*bay = macio_get_drvdata(mdev);
 
-	if (state.event != mdev->ofdev.dev.power.power_state.event && state.event == PM_EVENT_SUSPEND) {
+	if (state.event != mdev->ofdev.dev.power.power_state.event
+	    && (state.event & PM_EVENT_SLEEP)) {
 		down(&bay->lock);
 		bay->sleeping = 1;
 		set_mb_power(bay, 0);
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index ae3df46eaabf..183fddaa38b7 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -554,6 +554,7 @@ pci_power_t pci_choose_state(struct pci_dev *dev, pm_message_t state)
 	case PM_EVENT_PRETHAW:
 		/* REVISIT both freeze and pre-thaw "should" use D0 */
 	case PM_EVENT_SUSPEND:
+	case PM_EVENT_HIBERNATE:
 		return PCI_D3hot;
 	default:
 		printk("Unrecognized suspend event %d\n", state.event);
diff --git a/drivers/scsi/aic7xxx/aic79xx_osm_pci.c b/drivers/scsi/aic7xxx/aic79xx_osm_pci.c
index 4150c8a8fdc2..dfaaae5e73ae 100644
--- a/drivers/scsi/aic7xxx/aic79xx_osm_pci.c
+++ b/drivers/scsi/aic7xxx/aic79xx_osm_pci.c
@@ -89,7 +89,7 @@ ahd_linux_pci_dev_suspend(struct pci_dev *pdev, pm_message_t mesg)
 	pci_save_state(pdev);
 	pci_disable_device(pdev);
 
-	if (mesg.event == PM_EVENT_SUSPEND)
+	if (mesg.event & PM_EVENT_SLEEP)
 		pci_set_power_state(pdev, PCI_D3hot);
 
 	return rc;
diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c b/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c
index dd6e21d6f1dd..3d3eaef65fb3 100644
--- a/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c
+++ b/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c
@@ -134,7 +134,7 @@ ahc_linux_pci_dev_suspend(struct pci_dev *pdev, pm_message_t mesg)
 	pci_save_state(pdev);
 	pci_disable_device(pdev);
 
-	if (mesg.event == PM_EVENT_SUSPEND)
+	if (mesg.event & PM_EVENT_SLEEP)
 		pci_set_power_state(pdev, PCI_D3hot);
 
 	return rc;
diff --git a/drivers/scsi/mesh.c b/drivers/scsi/mesh.c
index 651d09b08f2a..fd63b06d9ef1 100644
--- a/drivers/scsi/mesh.c
+++ b/drivers/scsi/mesh.c
@@ -1759,6 +1759,7 @@ static int mesh_suspend(struct macio_dev *mdev, pm_message_t mesg)
 
 	switch (mesg.event) {
 	case PM_EVENT_SUSPEND:
+	case PM_EVENT_HIBERNATE:
 	case PM_EVENT_FREEZE:
 		break;
 	default:
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index 37df8bbe7f46..7aee64dbfbeb 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -1835,8 +1835,7 @@ static int sd_suspend(struct device *dev, pm_message_t mesg)
 			goto done;
 	}
 
-	if (mesg.event == PM_EVENT_SUSPEND &&
-	    sdkp->device->manage_start_stop) {
+	if ((mesg.event & PM_EVENT_SLEEP) && sdkp->device->manage_start_stop) {
 		sd_printk(KERN_NOTICE, sdkp, "Stopping disk\n");
 		ret = sd_start_stop_device(sdkp, 0);
 	}
diff --git a/drivers/usb/host/sl811-hcd.c b/drivers/usb/host/sl811-hcd.c
index ba370c56172c..59be276ccd9d 100644
--- a/drivers/usb/host/sl811-hcd.c
+++ b/drivers/usb/host/sl811-hcd.c
@@ -1766,6 +1766,7 @@ sl811h_suspend(struct platform_device *dev, pm_message_t state)
 		retval = sl811h_bus_suspend(hcd);
 		break;
 	case PM_EVENT_SUSPEND:
+	case PM_EVENT_HIBERNATE:
 	case PM_EVENT_PRETHAW:		/* explicitly discard hw state */
 		port_power(sl811, 0);
 		break;
diff --git a/drivers/usb/host/u132-hcd.c b/drivers/usb/host/u132-hcd.c
index ac283b09a63f..6fca06961559 100644
--- a/drivers/usb/host/u132-hcd.c
+++ b/drivers/usb/host/u132-hcd.c
@@ -3214,14 +3214,19 @@ static int u132_suspend(struct platform_device *pdev, pm_message_t state)
                 return -ESHUTDOWN;
         } else {
                 int retval = 0;
-                if (state.event == PM_EVENT_FREEZE) {
+
+		switch (state.event) {
+		case PM_EVENT_FREEZE:
                         retval = u132_bus_suspend(hcd);
-                } else if (state.event == PM_EVENT_SUSPEND) {
+			break;
+		case PM_EVENT_SUSPEND:
+		case PM_EVENT_HIBERNATE:
                         int ports = MAX_U132_PORTS;
                         while (ports-- > 0) {
                                 port_power(u132, ports, 0);
                         }
-                }
+			break;
+		}
                 if (retval == 0)
                         pdev->dev.power.power_state = state;
                 return retval;
diff --git a/drivers/video/chipsfb.c b/drivers/video/chipsfb.c
index 6796ba62c3c6..777389c40988 100644
--- a/drivers/video/chipsfb.c
+++ b/drivers/video/chipsfb.c
@@ -459,7 +459,7 @@ static int chipsfb_pci_suspend(struct pci_dev *pdev, pm_message_t state)
 
 	if (state.event == pdev->dev.power.power_state.event)
 		return 0;
-	if (state.event != PM_EVENT_SUSPEND)
+	if (!(state.event & PM_EVENT_SLEEP))
 		goto done;
 
 	acquire_console_sem();
diff --git a/drivers/video/nvidia/nvidia.c b/drivers/video/nvidia/nvidia.c
index 74517b1b26a6..596652d2831f 100644
--- a/drivers/video/nvidia/nvidia.c
+++ b/drivers/video/nvidia/nvidia.c
@@ -1066,7 +1066,7 @@ static int nvidiafb_suspend(struct pci_dev *dev, pm_message_t mesg)
 	acquire_console_sem();
 	par->pm_state = mesg.event;
 
-	if (mesg.event == PM_EVENT_SUSPEND) {
+	if (mesg.event & PM_EVENT_SLEEP) {
 		fb_set_suspend(info, 1);
 		nvidiafb_blank(FB_BLANK_POWERDOWN, info);
 		nvidia_write_regs(par, &par->SavedReg);
diff --git a/include/linux/pm.h b/include/linux/pm.h
index eccf59ea2a77..015b735811b4 100644
--- a/include/linux/pm.h
+++ b/include/linux/pm.h
@@ -143,6 +143,9 @@ typedef struct pm_message {
  * 		the upcoming system state (such as PCI_D3hot), and enable
  * 		wakeup events as appropriate.
  *
+ * HIBERNATE	Enter a low power device state appropriate for the hibernation
+ * 		state (eg. ACPI S4) and enable wakeup events as appropriate.
+ *
  * FREEZE	Quiesce operations so that a consistent image can be saved;
  * 		but do NOT otherwise enter a low power device state, and do
  * 		NOT emit system wakeup events.
@@ -166,11 +169,15 @@ typedef struct pm_message {
 #define PM_EVENT_ON 0
 #define PM_EVENT_FREEZE 1
 #define PM_EVENT_SUSPEND 2
-#define PM_EVENT_PRETHAW 3
+#define PM_EVENT_HIBERNATE 4
+#define PM_EVENT_PRETHAW 8
+
+#define PM_EVENT_SLEEP	(PM_EVENT_SUSPEND | PM_EVENT_HIBERNATE)
 
 #define PMSG_FREEZE	((struct pm_message){ .event = PM_EVENT_FREEZE, })
 #define PMSG_PRETHAW	((struct pm_message){ .event = PM_EVENT_PRETHAW, })
 #define PMSG_SUSPEND	((struct pm_message){ .event = PM_EVENT_SUSPEND, })
+#define PMSG_HIBERNATE	((struct pm_message){ .event = PM_EVENT_HIBERNATE, })
 #define PMSG_ON		((struct pm_message){ .event = PM_EVENT_ON, })
 
 struct dev_pm_info {
diff --git a/kernel/power/disk.c b/kernel/power/disk.c
index 859a8e59773a..14a656cdc652 100644
--- a/kernel/power/disk.c
+++ b/kernel/power/disk.c
@@ -391,7 +391,7 @@ int hibernation_platform_enter(void)
 		goto Close;
 
 	suspend_console();
-	error = device_suspend(PMSG_SUSPEND);
+	error = device_suspend(PMSG_HIBERNATE);
 	if (error)
 		goto Resume_console;
 
@@ -404,7 +404,7 @@ int hibernation_platform_enter(void)
 		goto Finish;
 
 	local_irq_disable();
-	error = device_power_down(PMSG_SUSPEND);
+	error = device_power_down(PMSG_HIBERNATE);
 	if (!error) {
 		hibernation_ops->enter();
 		/* We should never get here */
diff --git a/net/rfkill/rfkill.c b/net/rfkill/rfkill.c
index 1a47f5d1be17..140a0a8c6b02 100644
--- a/net/rfkill/rfkill.c
+++ b/net/rfkill/rfkill.c
@@ -232,7 +232,7 @@ static int rfkill_suspend(struct device *dev, pm_message_t state)
 	struct rfkill *rfkill = to_rfkill(dev);
 
 	if (dev->power.power_state.event != state.event) {
-		if (state.event == PM_EVENT_SUSPEND) {
+		if (state.event & PM_EVENT_SLEEP) {
 			mutex_lock(&rfkill->mutex);
 
 			if (rfkill->state == RFKILL_STATE_ON)
-- 
cgit v1.2.3


From eaeb16883bd6aa2d6b6b61b825c0d2b0dc793f60 Mon Sep 17 00:00:00 2001
From: David Rientjes <rientjes@google.com>
Date: Sat, 23 Feb 2008 15:23:34 -0800
Subject: sparc: fix build

Fix build failure on sparc:

	In file included from include/linux/mm.h:39,
	                from include/linux/memcontrol.h:24,
	                from include/linux/swap.h:8,
	                from include/linux/suspend.h:7,
	                from init/do_mounts.c:6:
	include/asm/pgtable.h:344: warning: parameter names (without
		types) in function declaration
	include/asm/pgtable.h:345: warning: parameter names (without
		types) in function declaration
	include/asm/pgtable.h:346: error: expected '=', ',', ';', 'asm' or
		'__attribute__' before '___f___swp_entry'

viro sayeth:

  I've run allmodconfig builds on a bunch of target, FWIW (essentially the
  same patch).  Note that these includes are recent addition caused by added
  inline function that had since then become a define.  So while I agree with
  your comments in general, in _this_ case it's pretty safe.

  The commit that had done it is 3062fc67dad01b1d2a15d58c709eff946389eca4
  ("memcontrol: move mm_cgroup to header file") and the switch to #define
  is in commit 60c12b1202a60eabb1c61317e5d2678fcea9893f ("memcontrol: add
  vm_match_cgroup()") (BTW, that probably warranted mentioning in the
  changelog of the latter).

Cc: Adrian Bunk <bunk@kernel.org>
Cc: Robert Reif <reif@earthlink.net>
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/memcontrol.h | 3 ---
 1 file changed, 3 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 925d57b236aa..04075628cb9a 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -20,9 +20,6 @@
 #ifndef _LINUX_MEMCONTROL_H
 #define _LINUX_MEMCONTROL_H
 
-#include <linux/rcupdate.h>
-#include <linux/mm.h>
-
 struct mem_cgroup;
 struct page_cgroup;
 struct page;
-- 
cgit v1.2.3


From 8ca3ed87db062201e1fa15b64a9214e193fc3a8a Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Sat, 23 Feb 2008 15:23:37 -0800
Subject: NOMMU: is_vmalloc_addr() won't compile if !MMU

Make is_vmalloc_addr() contingent on CONFIG_MMU=y, as it won't compile
in !MMU mode.

[ Bug introduced in commit 9e2779fa281cfda13ac060753d674bbcaa23367e:
  "is_vmalloc_addr(): Check if an address is within the vmalloc
  boundaries" ].

Signed-off-by: David Howells <dhowells@redhat.com>
Cc: Greg Ungerer <gerg@snapgear.com>
Cc: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/mm.h | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 3b3e1341163f..3f3ccfe42de0 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -235,6 +235,7 @@ static inline int get_page_unless_zero(struct page *page)
 struct page *vmalloc_to_page(const void *addr);
 unsigned long vmalloc_to_pfn(const void *addr);
 
+#ifdef CONFIG_MMU
 /* Determine if an address is within the vmalloc range */
 static inline int is_vmalloc_addr(const void *x)
 {
@@ -242,6 +243,7 @@ static inline int is_vmalloc_addr(const void *x)
 
 	return addr >= VMALLOC_START && addr < VMALLOC_END;
 }
+#endif
 
 static inline struct page *compound_head(struct page *page)
 {
-- 
cgit v1.2.3


From 45254b4fb2aef51c94a7397df1e481c4137b4b97 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Sat, 23 Feb 2008 15:23:51 -0800
Subject: efs: move headers out of include/linux/

Merge include/linux/efs_fs{_i,_dir}.h into fs/efs/efs.h.  efs_vh.h remains
there because this is the IRIX volume header and shouldn't really be
handled by efs but by the partitioning code.  efs_sb.h remains there for
now because it's exported to userspace.  Of course this wrong and aboot
should have a copy of it's own, but I'll leave that to a separate patch to
avoid any contention.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 fs/efs/dir.c             |   2 +-
 fs/efs/efs.h             | 140 +++++++++++++++++++++++++++++++++++++++++++++++
 fs/efs/file.c            |   2 +-
 fs/efs/inode.c           |   4 +-
 fs/efs/namei.c           |   2 +-
 fs/efs/super.c           |   7 ++-
 fs/efs/symlink.c         |   2 +-
 include/linux/efs_dir.h  |  42 --------------
 include/linux/efs_fs.h   |  56 -------------------
 include/linux/efs_fs_i.h |  68 -----------------------
 10 files changed, 150 insertions(+), 175 deletions(-)
 create mode 100644 fs/efs/efs.h
 delete mode 100644 include/linux/efs_dir.h
 delete mode 100644 include/linux/efs_fs.h
 delete mode 100644 include/linux/efs_fs_i.h

(limited to 'include/linux')

diff --git a/fs/efs/dir.c b/fs/efs/dir.c
index dfb5cb400217..49308a29798a 100644
--- a/fs/efs/dir.c
+++ b/fs/efs/dir.c
@@ -5,8 +5,8 @@
  */
 
 #include <linux/buffer_head.h>
-#include <linux/efs_fs.h>
 #include <linux/smp_lock.h>
+#include "efs.h"
 
 static int efs_readdir(struct file *, void *, filldir_t);
 
diff --git a/fs/efs/efs.h b/fs/efs/efs.h
new file mode 100644
index 000000000000..d8305b582ab0
--- /dev/null
+++ b/fs/efs/efs.h
@@ -0,0 +1,140 @@
+/*
+ * Copyright (c) 1999 Al Smith
+ *
+ * Portions derived from work (c) 1995,1996 Christian Vogelgsang.
+ * Portions derived from IRIX header files (c) 1988 Silicon Graphics
+ */
+#ifndef _EFS_EFS_H_
+#define _EFS_EFS_H_
+
+#include <linux/fs.h>
+#include <asm/uaccess.h>
+
+#define EFS_VERSION "1.0a"
+
+static const char cprt[] = "EFS: "EFS_VERSION" - (c) 1999 Al Smith <Al.Smith@aeschi.ch.eu.org>";
+
+
+/* 1 block is 512 bytes */
+#define	EFS_BLOCKSIZE_BITS	9
+#define	EFS_BLOCKSIZE		(1 << EFS_BLOCKSIZE_BITS)
+
+typedef	int32_t		efs_block_t;
+typedef uint32_t	efs_ino_t;
+
+#define	EFS_DIRECTEXTENTS	12
+
+/*
+ * layout of an extent, in memory and on disk. 8 bytes exactly.
+ */
+typedef union extent_u {
+	unsigned char raw[8];
+	struct extent_s {
+		unsigned int	ex_magic:8;	/* magic # (zero) */
+		unsigned int	ex_bn:24;	/* basic block */
+		unsigned int	ex_length:8;	/* numblocks in this extent */
+		unsigned int	ex_offset:24;	/* logical offset into file */
+	} cooked;
+} efs_extent;
+
+typedef struct edevs {
+	__be16		odev;
+	__be32		ndev;
+} efs_devs;
+
+/*
+ * extent based filesystem inode as it appears on disk.  The efs inode
+ * is exactly 128 bytes long.
+ */
+struct	efs_dinode {
+	__be16		di_mode;	/* mode and type of file */
+	__be16		di_nlink;	/* number of links to file */
+	__be16		di_uid;		/* owner's user id */
+	__be16		di_gid;		/* owner's group id */
+	__be32		di_size;	/* number of bytes in file */
+	__be32		di_atime;	/* time last accessed */
+	__be32		di_mtime;	/* time last modified */
+	__be32		di_ctime;	/* time created */
+	__be32		di_gen;		/* generation number */
+	__be16		di_numextents;	/* # of extents */
+	u_char		di_version;	/* version of inode */
+	u_char		di_spare;	/* spare - used by AFS */
+	union di_addr {
+		efs_extent	di_extents[EFS_DIRECTEXTENTS];
+		efs_devs	di_dev;	/* device for IFCHR/IFBLK */
+	} di_u;
+};
+
+/* efs inode storage in memory */
+struct efs_inode_info {
+	int		numextents;
+	int		lastextent;
+
+	efs_extent	extents[EFS_DIRECTEXTENTS];
+	struct inode	vfs_inode;
+};
+
+#include <linux/efs_fs_sb.h>
+
+#define EFS_DIRBSIZE_BITS	EFS_BLOCKSIZE_BITS
+#define EFS_DIRBSIZE		(1 << EFS_DIRBSIZE_BITS)
+
+struct efs_dentry {
+	__be32		inode;
+	unsigned char	namelen;
+	char		name[3];
+};
+
+#define EFS_DENTSIZE	(sizeof(struct efs_dentry) - 3 + 1)
+#define EFS_MAXNAMELEN  ((1 << (sizeof(char) * 8)) - 1)
+
+#define EFS_DIRBLK_HEADERSIZE	4
+#define EFS_DIRBLK_MAGIC	0xbeef	/* moo */
+
+struct efs_dir {
+	__be16	magic;
+	unsigned char	firstused;
+	unsigned char	slots;
+
+	unsigned char	space[EFS_DIRBSIZE - EFS_DIRBLK_HEADERSIZE];
+};
+
+#define EFS_MAXENTS \
+	((EFS_DIRBSIZE - EFS_DIRBLK_HEADERSIZE) / \
+	 (EFS_DENTSIZE + sizeof(char)))
+
+#define EFS_SLOTAT(dir, slot) EFS_REALOFF((dir)->space[slot])
+
+#define EFS_REALOFF(offset) ((offset << 1))
+
+
+static inline struct efs_inode_info *INODE_INFO(struct inode *inode)
+{
+	return container_of(inode, struct efs_inode_info, vfs_inode);
+}
+
+static inline struct efs_sb_info *SUPER_INFO(struct super_block *sb)
+{
+	return sb->s_fs_info;
+}
+
+struct statfs;
+struct fid;
+
+extern const struct inode_operations efs_dir_inode_operations;
+extern const struct file_operations efs_dir_operations;
+extern const struct address_space_operations efs_symlink_aops;
+
+extern struct inode *efs_iget(struct super_block *, unsigned long);
+extern efs_block_t efs_map_block(struct inode *, efs_block_t);
+extern int efs_get_block(struct inode *, sector_t, struct buffer_head *, int);
+
+extern struct dentry *efs_lookup(struct inode *, struct dentry *, struct nameidata *);
+extern struct dentry *efs_fh_to_dentry(struct super_block *sb, struct fid *fid,
+		int fh_len, int fh_type);
+extern struct dentry *efs_fh_to_parent(struct super_block *sb, struct fid *fid,
+		int fh_len, int fh_type);
+extern struct dentry *efs_get_parent(struct dentry *);
+extern int efs_bmap(struct inode *, int);
+
+#endif /* _EFS_EFS_H_ */
diff --git a/fs/efs/file.c b/fs/efs/file.c
index 5db20129681e..1ccb364ffa63 100644
--- a/fs/efs/file.c
+++ b/fs/efs/file.c
@@ -7,7 +7,7 @@
  */
 
 #include <linux/buffer_head.h>
-#include <linux/efs_fs.h>
+#include "efs.h"
 
 int efs_get_block(struct inode *inode, sector_t iblock,
 		  struct buffer_head *bh_result, int create)
diff --git a/fs/efs/inode.c b/fs/efs/inode.c
index 627c3026946d..79e19e5958e1 100644
--- a/fs/efs/inode.c
+++ b/fs/efs/inode.c
@@ -7,11 +7,11 @@
  *              and from work (c) 1998 Mike Shaver.
  */
 
-#include <linux/efs_fs.h>
-#include <linux/efs_fs_sb.h>
 #include <linux/buffer_head.h>
 #include <linux/module.h>
 #include <linux/fs.h>
+#include "efs.h"
+#include <linux/efs_fs_sb.h>
 
 static int efs_readpage(struct file *file, struct page *page)
 {
diff --git a/fs/efs/namei.c b/fs/efs/namei.c
index e26704742d41..3a404e7fad53 100644
--- a/fs/efs/namei.c
+++ b/fs/efs/namei.c
@@ -8,9 +8,9 @@
 
 #include <linux/buffer_head.h>
 #include <linux/string.h>
-#include <linux/efs_fs.h>
 #include <linux/smp_lock.h>
 #include <linux/exportfs.h>
+#include "efs.h"
 
 
 static efs_ino_t efs_find_entry(struct inode *inode, const char *name, int len) {
diff --git a/fs/efs/super.c b/fs/efs/super.c
index 14082405cdd1..d733531b55e2 100644
--- a/fs/efs/super.c
+++ b/fs/efs/super.c
@@ -8,14 +8,15 @@
 
 #include <linux/init.h>
 #include <linux/module.h>
-#include <linux/efs_fs.h>
-#include <linux/efs_vh.h>
-#include <linux/efs_fs_sb.h>
 #include <linux/exportfs.h>
 #include <linux/slab.h>
 #include <linux/buffer_head.h>
 #include <linux/vfs.h>
 
+#include "efs.h"
+#include <linux/efs_vh.h>
+#include <linux/efs_fs_sb.h>
+
 static int efs_statfs(struct dentry *dentry, struct kstatfs *buf);
 static int efs_fill_super(struct super_block *s, void *d, int silent);
 
diff --git a/fs/efs/symlink.c b/fs/efs/symlink.c
index 1d30d2ff440f..41911ec83aaf 100644
--- a/fs/efs/symlink.c
+++ b/fs/efs/symlink.c
@@ -7,10 +7,10 @@
  */
 
 #include <linux/string.h>
-#include <linux/efs_fs.h>
 #include <linux/pagemap.h>
 #include <linux/buffer_head.h>
 #include <linux/smp_lock.h>
+#include "efs.h"
 
 static int efs_symlink_readpage(struct file *file, struct page *page)
 {
diff --git a/include/linux/efs_dir.h b/include/linux/efs_dir.h
deleted file mode 100644
index a09ec010569c..000000000000
--- a/include/linux/efs_dir.h
+++ /dev/null
@@ -1,42 +0,0 @@
-/*
- * efs_dir.h
- *
- * Copyright (c) 1999 Al Smith
- */
-
-#ifndef __EFS_DIR_H__
-#define __EFS_DIR_H__
-
-#define EFS_DIRBSIZE_BITS	EFS_BLOCKSIZE_BITS
-#define EFS_DIRBSIZE		(1 << EFS_DIRBSIZE_BITS)
-
-struct efs_dentry {
-	__be32		inode;
-	unsigned char	namelen;
-	char		name[3];
-};
-
-#define EFS_DENTSIZE	(sizeof(struct efs_dentry) - 3 + 1)
-#define EFS_MAXNAMELEN  ((1 << (sizeof(char) * 8)) - 1)
-
-#define EFS_DIRBLK_HEADERSIZE	4
-#define EFS_DIRBLK_MAGIC	0xbeef	/* moo */
-
-struct efs_dir {
-	__be16	magic;
-	unsigned char	firstused;
-	unsigned char	slots;
-
-	unsigned char	space[EFS_DIRBSIZE - EFS_DIRBLK_HEADERSIZE];
-};
-
-#define EFS_MAXENTS \
-	((EFS_DIRBSIZE - EFS_DIRBLK_HEADERSIZE) / \
-	 (EFS_DENTSIZE + sizeof(char)))
-
-#define EFS_SLOTAT(dir, slot) EFS_REALOFF((dir)->space[slot])
-
-#define EFS_REALOFF(offset) ((offset << 1))
-
-#endif /* __EFS_DIR_H__ */
-
diff --git a/include/linux/efs_fs.h b/include/linux/efs_fs.h
deleted file mode 100644
index a695d63a07af..000000000000
--- a/include/linux/efs_fs.h
+++ /dev/null
@@ -1,56 +0,0 @@
-/*
- * efs_fs.h
- *
- * Copyright (c) 1999 Al Smith
- *
- * Portions derived from work (c) 1995,1996 Christian Vogelgsang.
- */
-
-#ifndef __EFS_FS_H__
-#define __EFS_FS_H__
-
-#define EFS_VERSION "1.0a"
-
-static const char cprt[] = "EFS: "EFS_VERSION" - (c) 1999 Al Smith <Al.Smith@aeschi.ch.eu.org>";
-
-#include <asm/uaccess.h>
-
-/* 1 block is 512 bytes */
-#define	EFS_BLOCKSIZE_BITS	9
-#define	EFS_BLOCKSIZE		(1 << EFS_BLOCKSIZE_BITS)
-
-#include <linux/fs.h>
-#include <linux/efs_fs_i.h>
-#include <linux/efs_fs_sb.h>
-#include <linux/efs_dir.h>
-
-static inline struct efs_inode_info *INODE_INFO(struct inode *inode)
-{
-	return container_of(inode, struct efs_inode_info, vfs_inode);
-}
-
-static inline struct efs_sb_info *SUPER_INFO(struct super_block *sb)
-{
-	return sb->s_fs_info;
-}
-
-struct statfs;
-struct fid;
-
-extern const struct inode_operations efs_dir_inode_operations;
-extern const struct file_operations efs_dir_operations;
-extern const struct address_space_operations efs_symlink_aops;
-
-extern struct inode *efs_iget(struct super_block *, unsigned long);
-extern efs_block_t efs_map_block(struct inode *, efs_block_t);
-extern int efs_get_block(struct inode *, sector_t, struct buffer_head *, int);
-
-extern struct dentry *efs_lookup(struct inode *, struct dentry *, struct nameidata *);
-extern struct dentry *efs_fh_to_dentry(struct super_block *sb, struct fid *fid,
-		int fh_len, int fh_type);
-extern struct dentry *efs_fh_to_parent(struct super_block *sb, struct fid *fid,
-		int fh_len, int fh_type);
-extern struct dentry *efs_get_parent(struct dentry *);
-extern int efs_bmap(struct inode *, int);
-
-#endif /* __EFS_FS_H__ */
diff --git a/include/linux/efs_fs_i.h b/include/linux/efs_fs_i.h
deleted file mode 100644
index 617c474ca659..000000000000
--- a/include/linux/efs_fs_i.h
+++ /dev/null
@@ -1,68 +0,0 @@
-/*
- * efs_fs_i.h
- *
- * Copyright (c) 1999 Al Smith
- *
- * Portions derived from IRIX header files (c) 1988 Silicon Graphics
- */
-
-#ifndef	__EFS_FS_I_H__
-#define	__EFS_FS_I_H__
-
-typedef	int32_t		efs_block_t;
-typedef uint32_t	efs_ino_t;
-
-#define	EFS_DIRECTEXTENTS	12
-
-/*
- * layout of an extent, in memory and on disk. 8 bytes exactly.
- */
-typedef union extent_u {
-	unsigned char raw[8];
-	struct extent_s {
-		unsigned int	ex_magic:8;	/* magic # (zero) */
-		unsigned int	ex_bn:24;	/* basic block */
-		unsigned int	ex_length:8;	/* numblocks in this extent */
-		unsigned int	ex_offset:24;	/* logical offset into file */
-	} cooked;
-} efs_extent;
-
-typedef struct edevs {
-	__be16		odev;
-	__be32		ndev;
-} efs_devs;
-
-/*
- * extent based filesystem inode as it appears on disk.  The efs inode
- * is exactly 128 bytes long.
- */
-struct	efs_dinode {
-	__be16		di_mode;	/* mode and type of file */
-	__be16		di_nlink;	/* number of links to file */
-	__be16		di_uid;		/* owner's user id */
-	__be16		di_gid;		/* owner's group id */
-	__be32		di_size;	/* number of bytes in file */
-	__be32		di_atime;	/* time last accessed */
-	__be32		di_mtime;	/* time last modified */
-	__be32		di_ctime;	/* time created */
-	__be32		di_gen;		/* generation number */
-	__be16		di_numextents;	/* # of extents */
-	u_char		di_version;	/* version of inode */
-	u_char		di_spare;	/* spare - used by AFS */
-	union di_addr {
-		efs_extent	di_extents[EFS_DIRECTEXTENTS];
-		efs_devs	di_dev;	/* device for IFCHR/IFBLK */
-	} di_u;
-};
-
-/* efs inode storage in memory */
-struct efs_inode_info {
-	int		numextents;
-	int		lastextent;
-
-	efs_extent	extents[EFS_DIRECTEXTENTS];
-	struct inode	vfs_inode;
-};
-
-#endif	/* __EFS_FS_I_H__ */
-
-- 
cgit v1.2.3


From a0c1e9073ef7428a14309cba010633a6cd6719ea Mon Sep 17 00:00:00 2001
From: Thomas Gleixner <tglx@linutronix.de>
Date: Sat, 23 Feb 2008 15:23:57 -0800
Subject: futex: runtime enable pi and robust functionality

Not all architectures implement futex_atomic_cmpxchg_inatomic().  The default
implementation returns -ENOSYS, which is currently not handled inside of the
futex guts.

Futex PI calls and robust list exits with a held futex result in an endless
loop in the futex code on architectures which have no support.

Fixing up every place where futex_atomic_cmpxchg_inatomic() is called would
add a fair amount of extra if/else constructs to the already complex code.  It
is also not possible to disable the robust feature before user space tries to
register robust lists.

Compile time disabling is not a good idea either, as there are already
architectures with runtime detection of futex_atomic_cmpxchg_inatomic support.

Detect the functionality at runtime instead by calling
cmpxchg_futex_value_locked() with a NULL pointer from the futex initialization
code.  This is guaranteed to fail, but the call of
futex_atomic_cmpxchg_inatomic() happens with pagefaults disabled.

On architectures, which use the asm-generic implementation or have a runtime
CPU feature detection, a -ENOSYS return value disables the PI/robust features.

On architectures with a working implementation the call returns -EFAULT and
the PI/robust features are enabled.

The relevant syscalls return -ENOSYS and the robust list exit code is blocked,
when the detection fails.

Fixes http://lkml.org/lkml/2008/2/11/149
Originally reported by: Lennart Buytenhek

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Lennert Buytenhek <buytenh@wantstofly.org>
Cc: Riku Voipio <riku.voipio@movial.fi>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/futex.h |  1 +
 kernel/futex.c        | 38 ++++++++++++++++++++++++++++++++++----
 kernel/futex_compat.c |  9 +++++++++
 3 files changed, 44 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/futex.h b/include/linux/futex.h
index 90048fb28a38..586ab56a3ec3 100644
--- a/include/linux/futex.h
+++ b/include/linux/futex.h
@@ -167,6 +167,7 @@ union futex_key {
 #ifdef CONFIG_FUTEX
 extern void exit_robust_list(struct task_struct *curr);
 extern void exit_pi_state_list(struct task_struct *curr);
+extern int futex_cmpxchg_enabled;
 #else
 static inline void exit_robust_list(struct task_struct *curr)
 {
diff --git a/kernel/futex.c b/kernel/futex.c
index c21f667c63f6..06968cd79200 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -60,6 +60,8 @@
 
 #include "rtmutex_common.h"
 
+int __read_mostly futex_cmpxchg_enabled;
+
 #define FUTEX_HASHBITS (CONFIG_BASE_SMALL ? 4 : 8)
 
 /*
@@ -469,6 +471,8 @@ void exit_pi_state_list(struct task_struct *curr)
 	struct futex_hash_bucket *hb;
 	union futex_key key;
 
+	if (!futex_cmpxchg_enabled)
+		return;
 	/*
 	 * We are a ZOMBIE and nobody can enqueue itself on
 	 * pi_state_list anymore, but we have to be careful
@@ -1870,6 +1874,8 @@ asmlinkage long
 sys_set_robust_list(struct robust_list_head __user *head,
 		    size_t len)
 {
+	if (!futex_cmpxchg_enabled)
+		return -ENOSYS;
 	/*
 	 * The kernel knows only one size for now:
 	 */
@@ -1894,6 +1900,9 @@ sys_get_robust_list(int pid, struct robust_list_head __user * __user *head_ptr,
 	struct robust_list_head __user *head;
 	unsigned long ret;
 
+	if (!futex_cmpxchg_enabled)
+		return -ENOSYS;
+
 	if (!pid)
 		head = current->robust_list;
 	else {
@@ -1997,6 +2006,9 @@ void exit_robust_list(struct task_struct *curr)
 	unsigned long futex_offset;
 	int rc;
 
+	if (!futex_cmpxchg_enabled)
+		return;
+
 	/*
 	 * Fetch the list head (which was registered earlier, via
 	 * sys_set_robust_list()):
@@ -2051,7 +2063,7 @@ void exit_robust_list(struct task_struct *curr)
 long do_futex(u32 __user *uaddr, int op, u32 val, ktime_t *timeout,
 		u32 __user *uaddr2, u32 val2, u32 val3)
 {
-	int ret;
+	int ret = -ENOSYS;
 	int cmd = op & FUTEX_CMD_MASK;
 	struct rw_semaphore *fshared = NULL;
 
@@ -2083,13 +2095,16 @@ long do_futex(u32 __user *uaddr, int op, u32 val, ktime_t *timeout,
 		ret = futex_wake_op(uaddr, fshared, uaddr2, val, val2, val3);
 		break;
 	case FUTEX_LOCK_PI:
-		ret = futex_lock_pi(uaddr, fshared, val, timeout, 0);
+		if (futex_cmpxchg_enabled)
+			ret = futex_lock_pi(uaddr, fshared, val, timeout, 0);
 		break;
 	case FUTEX_UNLOCK_PI:
-		ret = futex_unlock_pi(uaddr, fshared);
+		if (futex_cmpxchg_enabled)
+			ret = futex_unlock_pi(uaddr, fshared);
 		break;
 	case FUTEX_TRYLOCK_PI:
-		ret = futex_lock_pi(uaddr, fshared, 0, timeout, 1);
+		if (futex_cmpxchg_enabled)
+			ret = futex_lock_pi(uaddr, fshared, 0, timeout, 1);
 		break;
 	default:
 		ret = -ENOSYS;
@@ -2145,8 +2160,23 @@ static struct file_system_type futex_fs_type = {
 
 static int __init init(void)
 {
+	u32 curval;
 	int i;
 
+	/*
+	 * This will fail and we want it. Some arch implementations do
+	 * runtime detection of the futex_atomic_cmpxchg_inatomic()
+	 * functionality. We want to know that before we call in any
+	 * of the complex code paths. Also we want to prevent
+	 * registration of robust lists in that case. NULL is
+	 * guaranteed to fault and we get -EFAULT on functional
+	 * implementation, the non functional ones will return
+	 * -ENOSYS.
+	 */
+	curval = cmpxchg_futex_value_locked(NULL, 0, 0);
+	if (curval == -EFAULT)
+		futex_cmpxchg_enabled = 1;
+
 	for (i = 0; i < ARRAY_SIZE(futex_queues); i++) {
 		plist_head_init(&futex_queues[i].chain, &futex_queues[i].lock);
 		spin_lock_init(&futex_queues[i].lock);
diff --git a/kernel/futex_compat.c b/kernel/futex_compat.c
index 7d5e4b016f39..ff90f049f8f6 100644
--- a/kernel/futex_compat.c
+++ b/kernel/futex_compat.c
@@ -54,6 +54,9 @@ void compat_exit_robust_list(struct task_struct *curr)
 	compat_long_t futex_offset;
 	int rc;
 
+	if (!futex_cmpxchg_enabled)
+		return;
+
 	/*
 	 * Fetch the list head (which was registered earlier, via
 	 * sys_set_robust_list()):
@@ -115,6 +118,9 @@ asmlinkage long
 compat_sys_set_robust_list(struct compat_robust_list_head __user *head,
 			   compat_size_t len)
 {
+	if (!futex_cmpxchg_enabled)
+		return -ENOSYS;
+
 	if (unlikely(len != sizeof(*head)))
 		return -EINVAL;
 
@@ -130,6 +136,9 @@ compat_sys_get_robust_list(int pid, compat_uptr_t __user *head_ptr,
 	struct compat_robust_list_head __user *head;
 	unsigned long ret;
 
+	if (!futex_cmpxchg_enabled)
+		return -ENOSYS;
+
 	if (!pid)
 		head = current->compat_robust_list;
 	else {
-- 
cgit v1.2.3


From a043e3b2c63445512c5592cbe3c8694f3c655e81 Mon Sep 17 00:00:00 2001
From: Li Zefan <lizf@cn.fujitsu.com>
Date: Sat, 23 Feb 2008 15:24:09 -0800
Subject: cgroup: fix comments

fix:
- comments about need_forkexit_callback
- comments about release agent
- typo and comment style, etc.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/cgroup.h |   2 +-
 kernel/cgroup.c        | 142 +++++++++++++++++++++++++++----------------------
 2 files changed, 80 insertions(+), 64 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index ff9055fc3d2a..2ebf7afedd9f 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -175,7 +175,7 @@ struct css_set {
  *
  *
  * When reading/writing to a file:
- *	- the cgroup to use in file->f_dentry->d_parent->d_fsdata
+ *	- the cgroup to use is file->f_dentry->d_parent->d_fsdata
  *	- the 'cftype' of the file is file->f_dentry->d_fsdata
  */
 
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 4766bb65e4d9..36066d8a4911 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -113,9 +113,9 @@ static int root_count;
 #define dummytop (&rootnode.top_cgroup)
 
 /* This flag indicates whether tasks in the fork and exit paths should
- * take callback_mutex and check for fork/exit handlers to call. This
- * avoids us having to do extra work in the fork/exit path if none of the
- * subsystems need to be called.
+ * check for fork/exit handlers to call. This avoids us having to do
+ * extra work in the fork/exit path if none of the subsystems need to
+ * be called.
  */
 static int need_forkexit_callback;
 
@@ -307,7 +307,6 @@ static inline void put_css_set_taskexit(struct css_set *cg)
  * template: location in which to build the desired set of subsystem
  * state objects for the new cgroup group
  */
-
 static struct css_set *find_existing_css_set(
 	struct css_set *oldcg,
 	struct cgroup *cgrp,
@@ -354,7 +353,6 @@ static struct css_set *find_existing_css_set(
  * and chains them on tmp through their cgrp_link_list fields. Returns 0 on
  * success or a negative error
  */
-
 static int allocate_cg_links(int count, struct list_head *tmp)
 {
 	struct cg_cgroup_link *link;
@@ -396,7 +394,6 @@ static void free_cg_links(struct list_head *tmp)
  * substituted into the appropriate hierarchy. Must be called with
  * cgroup_mutex held
  */
-
 static struct css_set *find_css_set(
 	struct css_set *oldcg, struct cgroup *cgrp)
 {
@@ -507,8 +504,8 @@ static struct css_set *find_css_set(
  * critical pieces of code here.  The exception occurs on cgroup_exit(),
  * when a task in a notify_on_release cgroup exits.  Then cgroup_mutex
  * is taken, and if the cgroup count is zero, a usermode call made
- * to /sbin/cgroup_release_agent with the name of the cgroup (path
- * relative to the root of cgroup file system) as the argument.
+ * to the release agent with the name of the cgroup (path relative to
+ * the root of cgroup file system) as the argument.
  *
  * A cgroup can only be deleted if both its 'count' of using tasks
  * is zero, and its list of 'children' cgroups is empty.  Since all
@@ -521,7 +518,7 @@ static struct css_set *find_css_set(
  *
  * The need for this exception arises from the action of
  * cgroup_attach_task(), which overwrites one tasks cgroup pointer with
- * another.  It does so using cgroup_mutexe, however there are
+ * another.  It does so using cgroup_mutex, however there are
  * several performance critical places that need to reference
  * task->cgroup without the expense of grabbing a system global
  * mutex.  Therefore except as noted below, when dereferencing or, as
@@ -537,7 +534,6 @@ static struct css_set *find_css_set(
  * cgroup_lock - lock out any changes to cgroup structures
  *
  */
-
 void cgroup_lock(void)
 {
 	mutex_lock(&cgroup_mutex);
@@ -548,7 +544,6 @@ void cgroup_lock(void)
  *
  * Undo the lock taken in a previous cgroup_lock() call.
  */
-
 void cgroup_unlock(void)
 {
 	mutex_unlock(&cgroup_mutex);
@@ -590,7 +585,6 @@ static struct inode *cgroup_new_inode(mode_t mode, struct super_block *sb)
  * Call subsys's pre_destroy handler.
  * This is called before css refcnt check.
  */
-
 static void cgroup_call_pre_destroy(struct cgroup *cgrp)
 {
 	struct cgroup_subsys *ss;
@@ -600,7 +594,6 @@ static void cgroup_call_pre_destroy(struct cgroup *cgrp)
 	return;
 }
 
-
 static void cgroup_diput(struct dentry *dentry, struct inode *inode)
 {
 	/* is dentry a directory ? if so, kfree() associated cgroup */
@@ -1129,8 +1122,13 @@ static inline struct cftype *__d_cft(struct dentry *dentry)
 	return dentry->d_fsdata;
 }
 
-/*
- * Called with cgroup_mutex held.  Writes path of cgroup into buf.
+/**
+ * cgroup_path - generate the path of a cgroup
+ * @cgrp: the cgroup in question
+ * @buf: the buffer to write the path into
+ * @buflen: the length of the buffer
+ *
+ * Called with cgroup_mutex held. Writes path of cgroup into buf.
  * Returns 0 on success, -errno on error.
  */
 int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen)
@@ -1188,11 +1186,13 @@ static void get_first_subsys(const struct cgroup *cgrp,
 		*subsys_id = test_ss->subsys_id;
 }
 
-/*
- * Attach task 'tsk' to cgroup 'cgrp'
+/**
+ * cgroup_attach_task - attach task 'tsk' to cgroup 'cgrp'
+ * @cgrp: the cgroup the task is attaching to
+ * @tsk: the task to be attached
  *
- * Call holding cgroup_mutex.  May take task_lock of
- * the task 'pid' during call.
+ * Call holding cgroup_mutex. May take task_lock of
+ * the task 'tsk' during call.
  */
 int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk)
 {
@@ -1293,7 +1293,6 @@ static int attach_task_by_pid(struct cgroup *cgrp, char *pidbuf)
 }
 
 /* The various types of files and directories in a cgroup file system */
-
 enum cgroup_filetype {
 	FILE_ROOT,
 	FILE_DIR,
@@ -1584,12 +1583,11 @@ static int cgroup_create_file(struct dentry *dentry, int mode,
 }
 
 /*
- *	cgroup_create_dir - create a directory for an object.
- *	cgrp:	the cgroup we create the directory for.
- *		It must have a valid ->parent field
- *		And we are going to fill its ->dentry field.
- *	dentry: dentry of the new cgroup
- *	mode:	mode to set on new directory.
+ * cgroup_create_dir - create a directory for an object.
+ * @cgrp: the cgroup we create the directory for. It must have a valid
+ *        ->parent field. And we are going to fill its ->dentry field.
+ * @dentry: dentry of the new cgroup
+ * @mode: mode to set on new directory.
  */
 static int cgroup_create_dir(struct cgroup *cgrp, struct dentry *dentry,
 				int mode)
@@ -1651,8 +1649,12 @@ int cgroup_add_files(struct cgroup *cgrp,
 	return 0;
 }
 
-/* Count the number of tasks in a cgroup. */
-
+/**
+ * cgroup_task_count - count the number of tasks in a cgroup.
+ * @cgrp: the cgroup in question
+ *
+ * Return the number of tasks in the cgroup.
+ */
 int cgroup_task_count(const struct cgroup *cgrp)
 {
 	int count = 0;
@@ -1962,12 +1964,13 @@ static int pid_array_load(pid_t *pidarray, int npids, struct cgroup *cgrp)
 }
 
 /**
- * Build and fill cgroupstats so that taskstats can export it to user
- * space.
- *
+ * cgroupstats_build - build and fill cgroupstats
  * @stats: cgroupstats to fill information into
  * @dentry: A dentry entry belonging to the cgroup for which stats have
  * been requested.
+ *
+ * Build and fill cgroupstats so that taskstats can export it to user
+ * space.
  */
 int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
 {
@@ -2199,14 +2202,13 @@ static void init_cgroup_css(struct cgroup_subsys_state *css,
 }
 
 /*
- *	cgroup_create - create a cgroup
- *	parent:	cgroup that will be parent of the new cgroup.
- *	name:		name of the new cgroup. Will be strcpy'ed.
- *	mode:		mode to set on new inode
+ * cgroup_create - create a cgroup
+ * @parent: cgroup that will be parent of the new cgroup
+ * @dentry: dentry of the new cgroup
+ * @mode: mode to set on new inode
  *
- *	Must be called with the mutex on the parent inode held
+ * Must be called with the mutex on the parent inode held
  */
-
 static long cgroup_create(struct cgroup *parent, struct dentry *dentry,
 			     int mode)
 {
@@ -2349,13 +2351,12 @@ static int cgroup_rmdir(struct inode *unused_dir, struct dentry *dentry)
 	parent = cgrp->parent;
 	root = cgrp->root;
 	sb = root->sb;
+
 	/*
-	 * Call pre_destroy handlers of subsys
+	 * Call pre_destroy handlers of subsys. Notify subsystems
+	 * that rmdir() request comes.
 	 */
 	cgroup_call_pre_destroy(cgrp);
-	/*
-	 * Notify subsyses that rmdir() request comes.
-	 */
 
 	if (cgroup_has_css_refs(cgrp)) {
 		mutex_unlock(&cgroup_mutex);
@@ -2431,8 +2432,10 @@ static void cgroup_init_subsys(struct cgroup_subsys *ss)
 }
 
 /**
- * cgroup_init_early - initialize cgroups at system boot, and
- * initialize any subsystems that request early init.
+ * cgroup_init_early - cgroup initialization at system boot
+ *
+ * Initialize cgroups at system boot, and initialize any
+ * subsystems that request early init.
  */
 int __init cgroup_init_early(void)
 {
@@ -2474,8 +2477,10 @@ int __init cgroup_init_early(void)
 }
 
 /**
- * cgroup_init - register cgroup filesystem and /proc file, and
- * initialize any subsystems that didn't request early init.
+ * cgroup_init - cgroup initialization
+ *
+ * Register cgroup filesystem and /proc file, and initialize
+ * any subsystems that didn't request early init.
  */
 int __init cgroup_init(void)
 {
@@ -2618,7 +2623,7 @@ static struct file_operations proc_cgroupstats_operations = {
 
 /**
  * cgroup_fork - attach newly forked task to its parents cgroup.
- * @tsk: pointer to task_struct of forking parent process.
+ * @child: pointer to task_struct of forking parent process.
  *
  * Description: A task inherits its parent's cgroup at fork().
  *
@@ -2642,9 +2647,12 @@ void cgroup_fork(struct task_struct *child)
 }
 
 /**
- * cgroup_fork_callbacks - called on a new task very soon before
- * adding it to the tasklist. No need to take any locks since no-one
- * can be operating on this task
+ * cgroup_fork_callbacks - run fork callbacks
+ * @child: the new task
+ *
+ * Called on a new task very soon before adding it to the
+ * tasklist. No need to take any locks since no-one can
+ * be operating on this task.
  */
 void cgroup_fork_callbacks(struct task_struct *child)
 {
@@ -2659,11 +2667,14 @@ void cgroup_fork_callbacks(struct task_struct *child)
 }
 
 /**
- * cgroup_post_fork - called on a new task after adding it to the
- * task list. Adds the task to the list running through its css_set
- * if necessary. Has to be after the task is visible on the task list
- * in case we race with the first call to cgroup_iter_start() - to
- * guarantee that the new task ends up on its list. */
+ * cgroup_post_fork - called on a new task after adding it to the task list
+ * @child: the task in question
+ *
+ * Adds the task to the list running through its css_set if necessary.
+ * Has to be after the task is visible on the task list in case we race
+ * with the first call to cgroup_iter_start() - to guarantee that the
+ * new task ends up on its list.
+ */
 void cgroup_post_fork(struct task_struct *child)
 {
 	if (use_task_css_set_links) {
@@ -2676,6 +2687,7 @@ void cgroup_post_fork(struct task_struct *child)
 /**
  * cgroup_exit - detach cgroup from exiting task
  * @tsk: pointer to task_struct of exiting process
+ * @run_callback: run exit callbacks?
  *
  * Description: Detach cgroup from @tsk and release it.
  *
@@ -2706,7 +2718,6 @@ void cgroup_post_fork(struct task_struct *child)
  *    top_cgroup isn't going away, and either task has PF_EXITING set,
  *    which wards off any cgroup_attach_task() attempts, or task is a failed
  *    fork, never visible to cgroup_attach_task.
- *
  */
 void cgroup_exit(struct task_struct *tsk, int run_callbacks)
 {
@@ -2743,9 +2754,13 @@ void cgroup_exit(struct task_struct *tsk, int run_callbacks)
 }
 
 /**
- * cgroup_clone - duplicate the current cgroup in the hierarchy
- * that the given subsystem is attached to, and move this task into
- * the new child
+ * cgroup_clone - clone the cgroup the given subsystem is attached to
+ * @tsk: the task to be moved
+ * @subsys: the given subsystem
+ *
+ * Duplicate the current cgroup in the hierarchy that the given
+ * subsystem is attached to, and move this task into the new
+ * child.
  */
 int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *subsys)
 {
@@ -2858,9 +2873,12 @@ int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *subsys)
 	return ret;
 }
 
-/*
- * See if "cgrp" is a descendant of the current task's cgroup in
- * the appropriate hierarchy
+/**
+ * cgroup_is_descendant - see if @cgrp is a descendant of current task's cgrp
+ * @cgrp: the cgroup in question
+ *
+ * See if @cgrp is a descendant of the current task's cgroup in
+ * the appropriate hierarchy.
  *
  * If we are sending in dummytop, then presumably we are creating
  * the top cgroup in the subsystem.
@@ -2939,9 +2957,7 @@ void __css_put(struct cgroup_subsys_state *css)
  * release agent task.  We don't bother to wait because the caller of
  * this routine has no use for the exit status of the release agent
  * task, so no sense holding our caller up for that.
- *
  */
-
 static void cgroup_release_agent(struct work_struct *work)
 {
 	BUG_ON(work != &release_agent_work);
-- 
cgit v1.2.3


From ffd2d883399cbbb641e55730676ce1ec4845d99d Mon Sep 17 00:00:00 2001
From: Li Zefan <lizf@cn.fujitsu.com>
Date: Sat, 23 Feb 2008 15:24:09 -0800
Subject: cgroup: clean up cgroup.h

- replace old name 'cont' with 'cgrp' (Paul Menage did this cleanup for
  cgroup.c in commit bd89aabc6761de1c35b154fe6f914a445d301510)
- remove a duplicate declaration of cgroup_path()

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/cgroup.h | 48 +++++++++++++++++++++++-------------------------
 1 file changed, 23 insertions(+), 25 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index 2ebf7afedd9f..028ba3b523b1 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -186,15 +186,15 @@ struct cftype {
 	char name[MAX_CFTYPE_NAME];
 	int private;
 	int (*open) (struct inode *inode, struct file *file);
-	ssize_t (*read) (struct cgroup *cont, struct cftype *cft,
+	ssize_t (*read) (struct cgroup *cgrp, struct cftype *cft,
 			 struct file *file,
 			 char __user *buf, size_t nbytes, loff_t *ppos);
 	/*
 	 * read_uint() is a shortcut for the common case of returning a
 	 * single integer. Use it in place of read()
 	 */
-	u64 (*read_uint) (struct cgroup *cont, struct cftype *cft);
-	ssize_t (*write) (struct cgroup *cont, struct cftype *cft,
+	u64 (*read_uint) (struct cgroup *cgrp, struct cftype *cft);
+	ssize_t (*write) (struct cgroup *cgrp, struct cftype *cft,
 			  struct file *file,
 			  const char __user *buf, size_t nbytes, loff_t *ppos);
 
@@ -203,7 +203,7 @@ struct cftype {
 	 * a single integer (as parsed by simple_strtoull) from
 	 * userspace. Use in place of write(); return 0 or error.
 	 */
-	int (*write_uint) (struct cgroup *cont, struct cftype *cft, u64 val);
+	int (*write_uint) (struct cgroup *cgrp, struct cftype *cft, u64 val);
 
 	int (*release) (struct inode *inode, struct file *file);
 };
@@ -218,41 +218,41 @@ struct cgroup_scanner {
 
 /* Add a new file to the given cgroup directory. Should only be
  * called by subsystems from within a populate() method */
-int cgroup_add_file(struct cgroup *cont, struct cgroup_subsys *subsys,
+int cgroup_add_file(struct cgroup *cgrp, struct cgroup_subsys *subsys,
 		       const struct cftype *cft);
 
 /* Add a set of new files to the given cgroup directory. Should
  * only be called by subsystems from within a populate() method */
-int cgroup_add_files(struct cgroup *cont,
+int cgroup_add_files(struct cgroup *cgrp,
 			struct cgroup_subsys *subsys,
 			const struct cftype cft[],
 			int count);
 
-int cgroup_is_removed(const struct cgroup *cont);
+int cgroup_is_removed(const struct cgroup *cgrp);
 
-int cgroup_path(const struct cgroup *cont, char *buf, int buflen);
+int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen);
 
-int cgroup_task_count(const struct cgroup *cont);
+int cgroup_task_count(const struct cgroup *cgrp);
 
 /* Return true if the cgroup is a descendant of the current cgroup */
-int cgroup_is_descendant(const struct cgroup *cont);
+int cgroup_is_descendant(const struct cgroup *cgrp);
 
 /* Control Group subsystem type. See Documentation/cgroups.txt for details */
 
 struct cgroup_subsys {
 	struct cgroup_subsys_state *(*create)(struct cgroup_subsys *ss,
-						  struct cgroup *cont);
-	void (*pre_destroy)(struct cgroup_subsys *ss, struct cgroup *cont);
-	void (*destroy)(struct cgroup_subsys *ss, struct cgroup *cont);
+						  struct cgroup *cgrp);
+	void (*pre_destroy)(struct cgroup_subsys *ss, struct cgroup *cgrp);
+	void (*destroy)(struct cgroup_subsys *ss, struct cgroup *cgrp);
 	int (*can_attach)(struct cgroup_subsys *ss,
-			  struct cgroup *cont, struct task_struct *tsk);
-	void (*attach)(struct cgroup_subsys *ss, struct cgroup *cont,
-			struct cgroup *old_cont, struct task_struct *tsk);
+			  struct cgroup *cgrp, struct task_struct *tsk);
+	void (*attach)(struct cgroup_subsys *ss, struct cgroup *cgrp,
+			struct cgroup *old_cgrp, struct task_struct *tsk);
 	void (*fork)(struct cgroup_subsys *ss, struct task_struct *task);
 	void (*exit)(struct cgroup_subsys *ss, struct task_struct *task);
 	int (*populate)(struct cgroup_subsys *ss,
-			struct cgroup *cont);
-	void (*post_clone)(struct cgroup_subsys *ss, struct cgroup *cont);
+			struct cgroup *cgrp);
+	void (*post_clone)(struct cgroup_subsys *ss, struct cgroup *cgrp);
 	void (*bind)(struct cgroup_subsys *ss, struct cgroup *root);
 	int subsys_id;
 	int active;
@@ -273,9 +273,9 @@ struct cgroup_subsys {
 #undef SUBSYS
 
 static inline struct cgroup_subsys_state *cgroup_subsys_state(
-	struct cgroup *cont, int subsys_id)
+	struct cgroup *cgrp, int subsys_id)
 {
-	return cont->subsys[subsys_id];
+	return cgrp->subsys[subsys_id];
 }
 
 static inline struct cgroup_subsys_state *task_subsys_state(
@@ -290,8 +290,6 @@ static inline struct cgroup* task_cgroup(struct task_struct *task,
 	return task_subsys_state(task, subsys_id)->cgroup;
 }
 
-int cgroup_path(const struct cgroup *cont, char *buf, int buflen);
-
 int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *ss);
 
 /* A cgroup_iter should be treated as an opaque object */
@@ -313,10 +311,10 @@ struct cgroup_iter {
  *    - cgroup_scan_tasks() holds the css_set_lock when calling the test_task()
  *      callback, but not while calling the process_task() callback.
  */
-void cgroup_iter_start(struct cgroup *cont, struct cgroup_iter *it);
-struct task_struct *cgroup_iter_next(struct cgroup *cont,
+void cgroup_iter_start(struct cgroup *cgrp, struct cgroup_iter *it);
+struct task_struct *cgroup_iter_next(struct cgroup *cgrp,
 					struct cgroup_iter *it);
-void cgroup_iter_end(struct cgroup *cont, struct cgroup_iter *it);
+void cgroup_iter_end(struct cgroup *cgrp, struct cgroup_iter *it);
 int cgroup_scan_tasks(struct cgroup_scanner *scan);
 int cgroup_attach_task(struct cgroup *, struct task_struct *);
 
-- 
cgit v1.2.3


From 55b01e8681cba392ccda4ff6184054d38968115d Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Sat, 23 Feb 2008 20:09:11 -0800
Subject: [NET]: Restore sanity wrt. print_mac().

MAC_FMT had only one user and we tried to get rid of
that, but this created more problems than it solved.

As a result, this reverts three commits:

235365f3aaaa10b7056293877c0ead50425f25c7 ("net/8021q/vlan_dev.c: Use
print_mac."), fea5fa875eb235dc186b1f5184eb36abc63e26cc ("[NET]: Remove
MAC_FMT"), and 8f789c48448aed74fe1c07af76de8f04adacec7d ("[NET]:
Elminate spurious print_mac() calls.")

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if_ether.h |  3 ++-
 net/8021q/vlan_dev.c     | 10 ++++++----
 2 files changed, 8 insertions(+), 5 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/if_ether.h b/include/linux/if_ether.h
index 42dc6a3571ec..e157c1399b61 100644
--- a/include/linux/if_ether.h
+++ b/include/linux/if_ether.h
@@ -129,7 +129,8 @@ extern ssize_t sysfs_format_mac(char *buf, const unsigned char *addr, int len);
 /*
  *	Display a 6 byte device address (MAC) in a readable format.
  */
-extern __pure char *print_mac(char *buf, const unsigned char *addr);
+extern char *print_mac(char *buf, const unsigned char *addr);
+#define MAC_FMT "%02x:%02x:%02x:%02x:%02x:%02x"
 #define MAC_BUF_SIZE	18
 #define DECLARE_MAC_BUF(var) char var[MAC_BUF_SIZE] __maybe_unused
 
diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
index fc60c6d096b9..77f04e49a1a0 100644
--- a/net/8021q/vlan_dev.c
+++ b/net/8021q/vlan_dev.c
@@ -366,8 +366,7 @@ static int vlan_dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct net_device_stats *stats = &dev->stats;
 	struct vlan_ethhdr *veth = (struct vlan_ethhdr *)(skb->data);
-	DECLARE_MAC_BUF(mac);
-	DECLARE_MAC_BUF(mac2);
+
 	/* Handle non-VLAN frames if they are sent to us, for example by DHCP.
 	 *
 	 * NOTE: THIS ASSUMES DIX ETHERNET, SPECIFICALLY NOT SUPPORTING
@@ -405,8 +404,11 @@ static int vlan_dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	pr_debug("%s: about to send skb: %p to dev: %s\n",
 		__FUNCTION__, skb, skb->dev->name);
-	pr_debug("  %s %s %4hx %4hx %4hx\n",
-		 print_mac(mac, veth->h_dest), print_mac(mac2, veth->h_source),
+	pr_debug("  " MAC_FMT " " MAC_FMT " %4hx %4hx %4hx\n",
+		 veth->h_dest[0], veth->h_dest[1], veth->h_dest[2],
+		 veth->h_dest[3], veth->h_dest[4], veth->h_dest[5],
+		 veth->h_source[0], veth->h_source[1], veth->h_source[2],
+		 veth->h_source[3], veth->h_source[4], veth->h_source[5],
 		 veth->h_vlan_proto, veth->h_vlan_TCI,
 		 veth->h_vlan_encapsulated_proto);
 
-- 
cgit v1.2.3


From 9116300634c5c76cfcd0d2af689846e04d172256 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 21 Feb 2008 13:25:50 +0900
Subject: libata: automatically use DMADIR if drive/bridge requires it

Back in 2.6.17-rc2, a libata module parameter was added for atapi_dmadir.

That's nice, but most SATA devices which need it will tell us about it
in their IDENTIFY PACKET response, as bit-15 of word-62 of the
returned data (as per ATA7, ATA8 specifications).

So for those which specify it, we should automatically use the DMADIR bit.
Otherwise, disc writing will fail by default on many SATA-ATAPI drives.

This patch adds ATA_DFLAG_DMADIR and make ata_dev_configure() set it
if atapi_dmadir is set or identify data indicates DMADIR is necessary.
atapi_xlat() is converted to check ATA_DFLAG_DMADIR before setting
DMADIR.

Original patch is from Mark Lord.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Mark Lord <mlord@pobox.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 11 +++++++++--
 drivers/ata/libata-scsi.c |  3 ++-
 include/linux/ata.h       |  5 +++++
 include/linux/libata.h    |  1 +
 4 files changed, 17 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 9812bbf05e62..1845119cc7f0 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -2396,6 +2396,7 @@ int ata_dev_configure(struct ata_device *dev)
 	else if (dev->class == ATA_DEV_ATAPI) {
 		const char *cdb_intr_string = "";
 		const char *atapi_an_string = "";
+		const char *dma_dir_string = "";
 		u32 sntf;
 
 		rc = atapi_cdb_len(id);
@@ -2436,13 +2437,19 @@ int ata_dev_configure(struct ata_device *dev)
 			cdb_intr_string = ", CDB intr";
 		}
 
+		if (atapi_dmadir || atapi_id_dmadir(dev->id)) {
+			dev->flags |= ATA_DFLAG_DMADIR;
+			dma_dir_string = ", DMADIR";
+		}
+
 		/* print device info to dmesg */
 		if (ata_msg_drv(ap) && print_info)
 			ata_dev_printk(dev, KERN_INFO,
-				       "ATAPI: %s, %s, max %s%s%s\n",
+				       "ATAPI: %s, %s, max %s%s%s%s\n",
 				       modelbuf, fwrevbuf,
 				       ata_mode_string(xfer_mask),
-				       cdb_intr_string, atapi_an_string);
+				       cdb_intr_string, atapi_an_string,
+				       dma_dir_string);
 	}
 
 	/* determine max_sectors */
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index f888babc8283..0562b0a49f3b 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -2582,7 +2582,8 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc)
 		qc->tf.protocol = ATAPI_PROT_DMA;
 		qc->tf.feature |= ATAPI_PKT_DMA;
 
-		if (atapi_dmadir && (scmd->sc_data_direction != DMA_TO_DEVICE))
+		if ((dev->flags & ATA_DFLAG_DMADIR) &&
+		    (scmd->sc_data_direction != DMA_TO_DEVICE))
 			/* some SATA bridges need us to indicate data xfer direction */
 			qc->tf.feature |= ATAPI_DMADIR;
 	}
diff --git a/include/linux/ata.h b/include/linux/ata.h
index 78bbacaed8c4..1c622e2b0504 100644
--- a/include/linux/ata.h
+++ b/include/linux/ata.h
@@ -659,6 +659,11 @@ static inline int atapi_command_packet_set(const u16 *dev_id)
 	return (dev_id[0] >> 8) & 0x1f;
 }
 
+static inline int atapi_id_dmadir(const u16 *dev_id)
+{
+	return ata_id_major_version(dev_id) >= 7 && (dev_id[62] & 0x8000);
+}
+
 static inline int is_multi_taskfile(struct ata_taskfile *tf)
 {
 	return (tf->command == ATA_CMD_READ_MULTI) ||
diff --git a/include/linux/libata.h b/include/linux/libata.h
index ce7603a73156..a05f60013642 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -138,6 +138,7 @@ enum {
 	ATA_DFLAG_AN		= (1 << 7), /* AN configured */
 	ATA_DFLAG_HIPM		= (1 << 8), /* device supports HIPM */
 	ATA_DFLAG_DIPM		= (1 << 9), /* device supports DIPM */
+	ATA_DFLAG_DMADIR	= (1 << 10), /* device requires DMADIR */
 	ATA_DFLAG_CFG_MASK	= (1 << 12) - 1,
 
 	ATA_DFLAG_PIO		= (1 << 12), /* device limited to PIO mode */
-- 
cgit v1.2.3


From 305183fc3ec8aac55179ef0fcb65dab9b97a9145 Mon Sep 17 00:00:00 2001
From: Tobias Klauser <tklauser@distanz.ch>
Date: Sun, 24 Feb 2008 20:03:42 +0100
Subject: i2c: Storage class should be before const qualifier

The C99 specification states in section 6.11.5:

The placement of a storage-class specifier other than at the
beginning of the declaration specifiers in a declaration is an
obsolescent feature.

Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
---
 drivers/i2c/busses/i2c-pmcmsp.c | 4 ++--
 include/linux/i2c.h             | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/i2c/busses/i2c-pmcmsp.c b/drivers/i2c/busses/i2c-pmcmsp.c
index be99c02ecac5..b03af5653c65 100644
--- a/drivers/i2c/busses/i2c-pmcmsp.c
+++ b/drivers/i2c/busses/i2c-pmcmsp.c
@@ -122,7 +122,7 @@ struct pmcmsptwi_data {
 };
 
 /* The default settings */
-const static struct pmcmsptwi_clockcfg pmcmsptwi_defclockcfg = {
+static const struct pmcmsptwi_clockcfg pmcmsptwi_defclockcfg = {
 	.standard = {
 		.filter	= 0x3,
 		.clock	= 0x1f,
@@ -133,7 +133,7 @@ const static struct pmcmsptwi_clockcfg pmcmsptwi_defclockcfg = {
 	},
 };
 
-const static struct pmcmsptwi_cfg pmcmsptwi_defcfg = {
+static const struct pmcmsptwi_cfg pmcmsptwi_defcfg = {
 	.arbf		= 0x03,
 	.nak		= 0x03,
 	.add10		= 0x00,
diff --git a/include/linux/i2c.h b/include/linux/i2c.h
index 76014f8f3c60..2d1c608eb5db 100644
--- a/include/linux/i2c.h
+++ b/include/linux/i2c.h
@@ -598,7 +598,7 @@ I2C_CLIENT_MODULE_PARM(probe, "List of adapter,address pairs to scan "	\
 		       "additionally");					\
 I2C_CLIENT_MODULE_PARM(ignore, "List of adapter,address pairs not to "	\
 		       "scan");						\
-const static struct i2c_client_address_data addr_data = {		\
+static const struct i2c_client_address_data addr_data = {		\
 	.normal_i2c	= normal_i2c,					\
 	.probe		= probe,					\
 	.ignore		= ignore,					\
-- 
cgit v1.2.3


From 263867631ea02741baf878ca9faaf94b1563b9d7 Mon Sep 17 00:00:00 2001
From: David Brownell <david-b@pacbell.net>
Date: Sun, 24 Feb 2008 20:03:42 +0100
Subject: i2c: Make i2c_register_board_info() a NOP when CONFIG_I2C_BOARDINFO=n

Don't require platform code to be #ifdeffed according to whether
I2C is enabled or not ... if it's not enabled, let GCC compile out
all I2C device declarations.  (Issue noted on an NSLU2 build that
didn't configure I2C.)

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
---
 include/linux/i2c.h | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/i2c.h b/include/linux/i2c.h
index 2d1c608eb5db..365e0df3646b 100644
--- a/include/linux/i2c.h
+++ b/include/linux/i2c.h
@@ -271,9 +271,16 @@ extern void i2c_unregister_device(struct i2c_client *);
  * This is done at arch_initcall time, before declaring any i2c adapters.
  * Modules for add-on boards must use other calls.
  */
+#ifdef CONFIG_I2C_BOARDINFO
 extern int
 i2c_register_board_info(int busnum, struct i2c_board_info const *info, unsigned n);
-
+#else
+static inline int
+i2c_register_board_info(int busnum, struct i2c_board_info const *info, unsigned n)
+{
+	return 0;
+}
+#endif
 
 /*
  * The following structs are for those who like to implement new bus drivers:
-- 
cgit v1.2.3


From 2d07b255c7b8a9723010e5c74778e058dc05162e Mon Sep 17 00:00:00 2001
From: Harvey Harrison <harvey.harrison@gmail.com>
Date: Fri, 15 Feb 2008 09:56:34 -0800
Subject: sched: add declaration of sched_tail to sched.h

Avoids sparse warnings:
kernel/sched.c:2170:17: warning: symbol 'schedule_tail' was not declared. Should it be static?

Avoids the need for an external declaration in arch/um/process.c

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/um/kernel/process.c | 2 --
 include/linux/sched.h    | 1 +
 2 files changed, 1 insertion(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/arch/um/kernel/process.c b/arch/um/kernel/process.c
index fc50d2f959d1..e8cb9ff183e9 100644
--- a/arch/um/kernel/process.c
+++ b/arch/um/kernel/process.c
@@ -128,8 +128,6 @@ void *get_current(void)
 	return current;
 }
 
-extern void schedule_tail(struct task_struct *prev);
-
 /*
  * This is called magically, by its address being stuffed in a jmp_buf
  * and being longjmp-d to.
diff --git a/include/linux/sched.h b/include/linux/sched.h
index e217d188a102..9c17e828d6d4 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -242,6 +242,7 @@ struct task_struct;
 
 extern void sched_init(void);
 extern void sched_init_smp(void);
+extern asmlinkage void schedule_tail(struct task_struct *prev);
 extern void init_idle(struct task_struct *idle, int cpu);
 extern void init_idle_bootup_task(struct task_struct *idle);
 
-- 
cgit v1.2.3


From bdb9441e9c325d50b5ae17f7d3205d65b8ed2e5f Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Mon, 25 Feb 2008 23:02:48 +0100
Subject: lockdep: increase MAX_LOCK_DEPTH

Some code paths exceed the current max lock depth (XFS), so increase
this limit a bit. I looked at making this a dynamic allocated array,
but we should not advocate insane lock depths, so stay with this as
long as it works...

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index e217d188a102..e3ea12437547 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1189,7 +1189,7 @@ struct task_struct {
 	int softirq_context;
 #endif
 #ifdef CONFIG_LOCKDEP
-# define MAX_LOCK_DEPTH 30UL
+# define MAX_LOCK_DEPTH 48UL
 	u64 curr_chain_key;
 	int lockdep_depth;
 	struct held_lock held_locks[MAX_LOCK_DEPTH];
-- 
cgit v1.2.3


From 24d10f0c37d301e88f6965e3dc0aa684311544e5 Mon Sep 17 00:00:00 2001
From: Adrian McMenamin <adrian@newgolddream.dyndns.info>
Date: Sat, 16 Feb 2008 23:37:33 +0000
Subject: maple: remove unused variable

Remove an unused variable from the definition of struct maple_device

Signed-off-by: Adrian McMenamin <adrian@mcmen.demon.co.uk>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
---
 include/linux/maple.h | 1 -
 1 file changed, 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/maple.h b/include/linux/maple.h
index 3f01e2bae1a1..d31e36ebb436 100644
--- a/include/linux/maple.h
+++ b/include/linux/maple.h
@@ -64,7 +64,6 @@ struct maple_driver {
 	int (*connect) (struct maple_device * dev);
 	void (*disconnect) (struct maple_device * dev);
 	struct device_driver drv;
-	int registered;
 };
 
 void maple_getcond_callback(struct maple_device *dev,
-- 
cgit v1.2.3


From 96de1a8f0275bd67f243833e7088baced518f873 Mon Sep 17 00:00:00 2001
From: Paul Mundt <lethal@linux-sh.org>
Date: Tue, 26 Feb 2008 14:52:45 +0900
Subject: serial: Move asm-sh/sci.h to linux/serial_sci.h.

This header is needed on other architectures as well (namely h8300),
which currently fails to build without this in place. Rather than
duplicating the port definition completely there, just move this to a
common location instead.

This should get h8300 working again for 2.6.25, in addition to the
changes already pushed by Sato-san in -rc2.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
---
 arch/sh/kernel/cpu/sh2/setup-sh7619.c  |  2 +-
 arch/sh/kernel/cpu/sh2a/setup-sh7203.c |  2 +-
 arch/sh/kernel/cpu/sh2a/setup-sh7206.c |  2 +-
 arch/sh/kernel/cpu/sh3/setup-sh7705.c  |  2 +-
 arch/sh/kernel/cpu/sh3/setup-sh770x.c  |  2 +-
 arch/sh/kernel/cpu/sh3/setup-sh7710.c  |  2 +-
 arch/sh/kernel/cpu/sh3/setup-sh7720.c  |  2 +-
 arch/sh/kernel/cpu/sh4/setup-sh4-202.c |  2 +-
 arch/sh/kernel/cpu/sh4/setup-sh7750.c  |  2 +-
 arch/sh/kernel/cpu/sh4/setup-sh7760.c  |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-sh7343.c |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-sh7366.c |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-sh7722.c |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-sh7763.c |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-sh7770.c |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-sh7780.c |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-sh7785.c |  2 +-
 arch/sh/kernel/cpu/sh4a/setup-shx3.c   |  2 +-
 drivers/serial/sh-sci.c                |  2 +-
 include/asm-sh/sci.h                   | 34 ----------------------------------
 include/linux/serial_sci.h             | 32 ++++++++++++++++++++++++++++++++
 21 files changed, 51 insertions(+), 53 deletions(-)
 delete mode 100644 include/asm-sh/sci.h
 create mode 100644 include/linux/serial_sci.h

(limited to 'include/linux')

diff --git a/arch/sh/kernel/cpu/sh2/setup-sh7619.c b/arch/sh/kernel/cpu/sh2/setup-sh7619.c
index b230eb278cef..cc530f4d84d6 100644
--- a/arch/sh/kernel/cpu/sh2/setup-sh7619.c
+++ b/arch/sh/kernel/cpu/sh2/setup-sh7619.c
@@ -10,7 +10,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 enum {
 	UNUSED = 0,
diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7203.c b/arch/sh/kernel/cpu/sh2a/setup-sh7203.c
index db6ef5cecde1..e98dc4450352 100644
--- a/arch/sh/kernel/cpu/sh2a/setup-sh7203.c
+++ b/arch/sh/kernel/cpu/sh2a/setup-sh7203.c
@@ -10,7 +10,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 enum {
 	UNUSED = 0,
diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7206.c b/arch/sh/kernel/cpu/sh2a/setup-sh7206.c
index a564425b905f..e6d4ec445dd8 100644
--- a/arch/sh/kernel/cpu/sh2a/setup-sh7206.c
+++ b/arch/sh/kernel/cpu/sh2a/setup-sh7206.c
@@ -10,7 +10,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 enum {
 	UNUSED = 0,
diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7705.c b/arch/sh/kernel/cpu/sh3/setup-sh7705.c
index dd0a20a685f7..f581534cb732 100644
--- a/arch/sh/kernel/cpu/sh3/setup-sh7705.c
+++ b/arch/sh/kernel/cpu/sh3/setup-sh7705.c
@@ -12,7 +12,7 @@
 #include <linux/init.h>
 #include <linux/irq.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 #include <asm/rtc.h>
 
 enum {
diff --git a/arch/sh/kernel/cpu/sh3/setup-sh770x.c b/arch/sh/kernel/cpu/sh3/setup-sh770x.c
index 9066ed78e283..d3733b13ea52 100644
--- a/arch/sh/kernel/cpu/sh3/setup-sh770x.c
+++ b/arch/sh/kernel/cpu/sh3/setup-sh770x.c
@@ -16,7 +16,7 @@
 #include <linux/irq.h>
 #include <linux/platform_device.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 enum {
 	UNUSED = 0,
diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7710.c b/arch/sh/kernel/cpu/sh3/setup-sh7710.c
index 0cc0e2bf135d..7406c9ad9259 100644
--- a/arch/sh/kernel/cpu/sh3/setup-sh7710.c
+++ b/arch/sh/kernel/cpu/sh3/setup-sh7710.c
@@ -12,7 +12,7 @@
 #include <linux/init.h>
 #include <linux/irq.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 #include <asm/rtc.h>
 
 enum {
diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7720.c b/arch/sh/kernel/cpu/sh3/setup-sh7720.c
index 3855ea4c21c8..8028082527c5 100644
--- a/arch/sh/kernel/cpu/sh3/setup-sh7720.c
+++ b/arch/sh/kernel/cpu/sh3/setup-sh7720.c
@@ -16,7 +16,7 @@
 #include <linux/init.h>
 #include <linux/serial.h>
 #include <linux/io.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 #include <asm/rtc.h>
 
 #define INTC_ICR1	0xA4140010UL
diff --git a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c
index dab193293f20..7371abf64f80 100644
--- a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c
+++ b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c
@@ -10,7 +10,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 static struct plat_sci_port sci_platform_data[] = {
 	{
diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7750.c b/arch/sh/kernel/cpu/sh4/setup-sh7750.c
index ae3603aca615..ec884039b914 100644
--- a/arch/sh/kernel/cpu/sh4/setup-sh7750.c
+++ b/arch/sh/kernel/cpu/sh4/setup-sh7750.c
@@ -12,7 +12,7 @@
 #include <linux/init.h>
 #include <linux/serial.h>
 #include <linux/io.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 static struct resource rtc_resources[] = {
 	[0] = {
diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7760.c b/arch/sh/kernel/cpu/sh4/setup-sh7760.c
index 85f81579b97e..254c5c55ab91 100644
--- a/arch/sh/kernel/cpu/sh4/setup-sh7760.c
+++ b/arch/sh/kernel/cpu/sh4/setup-sh7760.c
@@ -10,7 +10,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 enum {
 	UNUSED = 0,
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c
index c0a3f079dfdc..6d4f50cd4aaf 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c
@@ -10,7 +10,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 static struct plat_sci_port sci_platform_data[] = {
 	{
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c
index 967e8b69a2f8..f26b5cdad0d1 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c
@@ -12,7 +12,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 static struct plat_sci_port sci_platform_data[] = {
 	{
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c
index 73c778d40d13..b98b4bc93ec9 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c
@@ -10,9 +10,9 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
+#include <linux/serial_sci.h>
 #include <linux/mm.h>
 #include <asm/mmzone.h>
-#include <asm/sci.h>
 
 static struct resource usbf_resources[] = {
 	[0] = {
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7763.c b/arch/sh/kernel/cpu/sh4a/setup-sh7763.c
index eabd5386812d..07c988dc9de6 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7763.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7763.c
@@ -12,7 +12,7 @@
 #include <linux/init.h>
 #include <linux/serial.h>
 #include <linux/io.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 static struct resource rtc_resources[] = {
 	[0] = {
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7770.c b/arch/sh/kernel/cpu/sh4a/setup-sh7770.c
index 32f4f59a837b..b9cec48b1808 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7770.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7770.c
@@ -10,7 +10,7 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 static struct plat_sci_port sci_platform_data[] = {
 	{
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c
index 293004b526ff..18dbbe23fea1 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c
@@ -11,7 +11,7 @@
 #include <linux/init.h>
 #include <linux/serial.h>
 #include <linux/io.h>
-#include <asm/sci.h>
+#include <linux/serial_sci.h>
 
 static struct resource rtc_resources[] = {
 	[0] = {
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c
index 74b60e96cdf4..621e7329ec63 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c
@@ -10,10 +10,10 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
+#include <linux/serial_sci.h>
 #include <linux/io.h>
 #include <linux/mm.h>
 #include <asm/mmzone.h>
-#include <asm/sci.h>
 
 static struct plat_sci_port sci_platform_data[] = {
 	{
diff --git a/arch/sh/kernel/cpu/sh4a/setup-shx3.c b/arch/sh/kernel/cpu/sh4a/setup-shx3.c
index 4dc958b6b314..bd35f32534b9 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-shx3.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-shx3.c
@@ -10,9 +10,9 @@
 #include <linux/platform_device.h>
 #include <linux/init.h>
 #include <linux/serial.h>
+#include <linux/serial_sci.h>
 #include <linux/io.h>
 #include <asm/mmzone.h>
-#include <asm/sci.h>
 
 static struct plat_sci_port sci_platform_data[] = {
 	{
diff --git a/drivers/serial/sh-sci.c b/drivers/serial/sh-sci.c
index 9ce12cb2cebc..a8c116b80bff 100644
--- a/drivers/serial/sh-sci.c
+++ b/drivers/serial/sh-sci.c
@@ -41,6 +41,7 @@
 #include <linux/delay.h>
 #include <linux/console.h>
 #include <linux/platform_device.h>
+#include <linux/serial_sci.h>
 
 #ifdef CONFIG_CPU_FREQ
 #include <linux/notifier.h>
@@ -54,7 +55,6 @@
 #include <asm/kgdb.h>
 #endif
 
-#include <asm/sci.h>
 #include "sh-sci.h"
 
 struct sci_port {
diff --git a/include/asm-sh/sci.h b/include/asm-sh/sci.h
deleted file mode 100644
index 52e73660c129..000000000000
--- a/include/asm-sh/sci.h
+++ /dev/null
@@ -1,34 +0,0 @@
-#ifndef __ASM_SH_SCI_H
-#define __ASM_SH_SCI_H
-
-#include <linux/serial_core.h>
-
-/*
- * Generic header for SuperH SCI(F)
- *
- * Do not place SH-specific parts in here, sh64 and h8300 depend on this too.
- */
-
-/* Offsets into the sci_port->irqs array */
-enum {
-	SCIx_ERI_IRQ,
-	SCIx_RXI_IRQ,
-	SCIx_TXI_IRQ,
-	SCIx_BRI_IRQ,
-	SCIx_NR_IRQS,
-};
-
-/*
- * Platform device specific platform_data struct
- */
-struct plat_sci_port {
-	void __iomem	*membase;		/* io cookie */
-	unsigned long	mapbase;		/* resource base */
-	unsigned int	irqs[SCIx_NR_IRQS];	/* ERI, RXI, TXI, BRI */
-	unsigned int	type;			/* SCI / SCIF / IRDA */
-	upf_t		flags;			/* UPF_* flags */
-};
-
-int early_sci_setup(struct uart_port *port);
-
-#endif /* __ASM_SH_SCI_H */
diff --git a/include/linux/serial_sci.h b/include/linux/serial_sci.h
new file mode 100644
index 000000000000..893cc53486bc
--- /dev/null
+++ b/include/linux/serial_sci.h
@@ -0,0 +1,32 @@
+#ifndef __LINUX_SERIAL_SCI_H
+#define __LINUX_SERIAL_SCI_H
+
+#include <linux/serial_core.h>
+
+/*
+ * Generic header for SuperH SCI(F) (used by sh/sh64/h8300 and related parts)
+ */
+
+/* Offsets into the sci_port->irqs array */
+enum {
+	SCIx_ERI_IRQ,
+	SCIx_RXI_IRQ,
+	SCIx_TXI_IRQ,
+	SCIx_BRI_IRQ,
+	SCIx_NR_IRQS,
+};
+
+/*
+ * Platform device specific platform_data struct
+ */
+struct plat_sci_port {
+	void __iomem	*membase;		/* io cookie */
+	unsigned long	mapbase;		/* resource base */
+	unsigned int	irqs[SCIx_NR_IRQS];	/* ERI, RXI, TXI, BRI */
+	unsigned int	type;			/* SCI / SCIF / IRDA */
+	upf_t		flags;			/* UPF_* flags */
+};
+
+int early_sci_setup(struct uart_port *port);
+
+#endif /* __LINUX_SERIAL_SCI_H */
-- 
cgit v1.2.3


From 78374676efae525094aee45c0aab4bcab95ea9d1 Mon Sep 17 00:00:00 2001
From: Li Zefan <lizf@cn.fujitsu.com>
Date: Tue, 26 Feb 2008 18:25:53 -0800
Subject: CONNECTOR: make cn_already_initialized static

It is used in connector.c only, so make it static.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/connector/connector.c | 2 +-
 include/linux/connector.h     | 2 --
 2 files changed, 1 insertion(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c
index fea2d3ed9cbd..85e2ba7fcfba 100644
--- a/drivers/connector/connector.c
+++ b/drivers/connector/connector.c
@@ -47,7 +47,7 @@ static LIST_HEAD(notify_list);
 
 static struct cn_dev cdev;
 
-int cn_already_initialized = 0;
+static int cn_already_initialized;
 
 /*
  * msg->seq and msg->ack are used to determine message genealogy.
diff --git a/include/linux/connector.h b/include/linux/connector.h
index da6dd957f908..96a89d3d6727 100644
--- a/include/linux/connector.h
+++ b/include/linux/connector.h
@@ -170,7 +170,5 @@ int cn_cb_equal(struct cb_id *, struct cb_id *);
 
 void cn_queue_wrapper(struct work_struct *work);
 
-extern int cn_already_initialized;
-
 #endif				/* __KERNEL__ */
 #endif				/* __CONNECTOR_H */
-- 
cgit v1.2.3


From fbabbed8284d1526ed01754ecd4fabdb941a1ff2 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 27 Feb 2008 12:21:18 -0800
Subject: [NETFILTER]: Fix NF_QUEUE_NR() parenthesis

Properly add parens around the macro argument. This is not needed by
the kernel but the macro is exported to userspace, so it shouldn't
make any assumptions.

Also use NF_VERDICT_BITS instead of NF_VERDICT_QBTIS for the left-shift
since thats whats logically correct.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h
index b74b615492e8..f0680c2bee73 100644
--- a/include/linux/netfilter.h
+++ b/include/linux/netfilter.h
@@ -31,7 +31,7 @@
 #define NF_VERDICT_QMASK 0xffff0000
 #define NF_VERDICT_QBITS 16
 
-#define NF_QUEUE_NR(x) (((x << NF_VERDICT_QBITS) & NF_VERDICT_QMASK) | NF_QUEUE)
+#define NF_QUEUE_NR(x) ((((x) << NF_VERDICT_BITS) & NF_VERDICT_QMASK) | NF_QUEUE)
 
 /* only for userspace compatibility */
 #ifndef __KERNEL__
-- 
cgit v1.2.3


From e5b13cb10de209f924fdf9478214bcf7e4008d6d Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Thu, 28 Feb 2008 20:51:43 -0800
Subject: [NETNS]: Process devinet ioctl in the correct namespace.

Add namespace parameter to devinet_ioctl and locate device inside it for
state changes.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/inetdevice.h | 2 +-
 net/ipv4/af_inet.c         | 7 ++++---
 net/ipv4/devinet.c         | 6 +++---
 net/ipv4/ipconfig.c        | 2 +-
 4 files changed, 9 insertions(+), 8 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h
index fc4e3db649e8..da05ab47ff2f 100644
--- a/include/linux/inetdevice.h
+++ b/include/linux/inetdevice.h
@@ -129,7 +129,7 @@ extern int unregister_inetaddr_notifier(struct notifier_block *nb);
 
 extern struct net_device *ip_dev_find(struct net *net, __be32 addr);
 extern int		inet_addr_onlink(struct in_device *in_dev, __be32 a, __be32 b);
-extern int		devinet_ioctl(unsigned int cmd, void __user *);
+extern int		devinet_ioctl(struct net *net, unsigned int cmd, void __user *);
 extern void		devinet_init(void);
 extern struct in_device	*inetdev_by_index(struct net *, int);
 extern __be32		inet_select_addr(const struct net_device *dev, __be32 dst, int scope);
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 09ca5293d08f..c270080f370e 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -784,6 +784,7 @@ int inet_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
 {
 	struct sock *sk = sock->sk;
 	int err = 0;
+	struct net *net = sk->sk_net;
 
 	switch (cmd) {
 		case SIOCGSTAMP:
@@ -795,12 +796,12 @@ int inet_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
 		case SIOCADDRT:
 		case SIOCDELRT:
 		case SIOCRTMSG:
-			err = ip_rt_ioctl(sk->sk_net, cmd, (void __user *)arg);
+			err = ip_rt_ioctl(net, cmd, (void __user *)arg);
 			break;
 		case SIOCDARP:
 		case SIOCGARP:
 		case SIOCSARP:
-			err = arp_ioctl(sk->sk_net, cmd, (void __user *)arg);
+			err = arp_ioctl(net, cmd, (void __user *)arg);
 			break;
 		case SIOCGIFADDR:
 		case SIOCSIFADDR:
@@ -813,7 +814,7 @@ int inet_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
 		case SIOCSIFPFLAGS:
 		case SIOCGIFPFLAGS:
 		case SIOCSIFFLAGS:
-			err = devinet_ioctl(cmd, (void __user *)arg);
+			err = devinet_ioctl(net, cmd, (void __user *)arg);
 			break;
 		default:
 			if (sk->sk_prot->ioctl)
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 90210a74d638..af752fc5d0ab 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -595,7 +595,7 @@ static __inline__ int inet_abc_len(__be32 addr)
 }
 
 
-int devinet_ioctl(unsigned int cmd, void __user *arg)
+int devinet_ioctl(struct net *net, unsigned int cmd, void __user *arg)
 {
 	struct ifreq ifr;
 	struct sockaddr_in sin_orig;
@@ -624,7 +624,7 @@ int devinet_ioctl(unsigned int cmd, void __user *arg)
 		*colon = 0;
 
 #ifdef CONFIG_KMOD
-	dev_load(&init_net, ifr.ifr_name);
+	dev_load(net, ifr.ifr_name);
 #endif
 
 	switch (cmd) {
@@ -665,7 +665,7 @@ int devinet_ioctl(unsigned int cmd, void __user *arg)
 	rtnl_lock();
 
 	ret = -ENODEV;
-	if ((dev = __dev_get_by_name(&init_net, ifr.ifr_name)) == NULL)
+	if ((dev = __dev_get_by_name(net, ifr.ifr_name)) == NULL)
 		goto done;
 
 	if (colon)
diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c
index 10013ccee8dd..c90e75a66e81 100644
--- a/net/ipv4/ipconfig.c
+++ b/net/ipv4/ipconfig.c
@@ -291,7 +291,7 @@ static int __init ic_dev_ioctl(unsigned int cmd, struct ifreq *arg)
 
 	mm_segment_t oldfs = get_fs();
 	set_fs(get_ds());
-	res = devinet_ioctl(cmd, (struct ifreq __user *) arg);
+	res = devinet_ioctl(&init_net, cmd, (struct ifreq __user *) arg);
 	set_fs(oldfs);
 	return res;
 }
-- 
cgit v1.2.3


From 4c563f7669c10a12354b72b518c2287ffc6ebfb3 Mon Sep 17 00:00:00 2001
From: Timo Teras <timo.teras@iki.fi>
Date: Thu, 28 Feb 2008 21:31:08 -0800
Subject: [XFRM]: Speed up xfrm_policy and xfrm_state walking

Change xfrm_policy and xfrm_state walking algorithm from O(n^2) to O(n).
This is achieved adding the entries to one more list which is used
solely for walking the entries.

This also fixes some races where the dump can have duplicate or missing
entries when the SPD/SADB is modified during an ongoing dump.

Dumping SADB with 20000 entries using "time ip xfrm state" the sys
time dropped from 1.012s to 0.080s.

Signed-off-by: Timo Teras <timo.teras@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/xfrm.h   |  3 +-
 include/net/xfrm.h     | 52 +++++++++++++++++++++++++++++++--
 net/key/af_key.c       | 24 ++++++++++++---
 net/xfrm/xfrm_policy.c | 79 +++++++++++++++++++++++++++++---------------------
 net/xfrm/xfrm_state.c  | 53 ++++++++++++++++++++++-----------
 net/xfrm/xfrm_user.c   | 71 +++++++++++++++++++++++++++------------------
 6 files changed, 197 insertions(+), 85 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index e31b8c84f2c9..0c82c80b277f 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -113,7 +113,8 @@ enum
 {
 	XFRM_POLICY_TYPE_MAIN	= 0,
 	XFRM_POLICY_TYPE_SUB	= 1,
-	XFRM_POLICY_TYPE_MAX	= 2
+	XFRM_POLICY_TYPE_MAX	= 2,
+	XFRM_POLICY_TYPE_ANY	= 255
 };
 
 enum
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index eea7785cc757..9b6205665190 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -121,6 +121,7 @@ extern struct mutex xfrm_cfg_mutex;
 struct xfrm_state
 {
 	/* Note: bydst is re-used during gc */
+	struct list_head	all;
 	struct hlist_node	bydst;
 	struct hlist_node	bysrc;
 	struct hlist_node	byspi;
@@ -424,6 +425,7 @@ struct xfrm_tmpl
 struct xfrm_policy
 {
 	struct xfrm_policy	*next;
+	struct list_head	bytype;
 	struct hlist_node	bydst;
 	struct hlist_node	byidx;
 
@@ -1160,6 +1162,18 @@ struct xfrm6_tunnel {
 	int priority;
 };
 
+struct xfrm_state_walk {
+	struct xfrm_state *state;
+	int count;
+	u8 proto;
+};
+
+struct xfrm_policy_walk {
+	struct xfrm_policy *policy;
+	int count;
+	u8 type, cur_type;
+};
+
 extern void xfrm_init(void);
 extern void xfrm4_init(void);
 extern void xfrm_state_init(void);
@@ -1184,7 +1198,23 @@ static inline void xfrm6_fini(void)
 extern int xfrm_proc_init(void);
 #endif
 
-extern int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*), void *);
+static inline void xfrm_state_walk_init(struct xfrm_state_walk *walk, u8 proto)
+{
+	walk->proto = proto;
+	walk->state = NULL;
+	walk->count = 0;
+}
+
+static inline void xfrm_state_walk_done(struct xfrm_state_walk *walk)
+{
+	if (walk->state != NULL) {
+		xfrm_state_put(walk->state);
+		walk->state = NULL;
+	}
+}
+
+extern int xfrm_state_walk(struct xfrm_state_walk *walk,
+			   int (*func)(struct xfrm_state *, int, void*), void *);
 extern struct xfrm_state *xfrm_state_alloc(void);
 extern struct xfrm_state *xfrm_state_find(xfrm_address_t *daddr, xfrm_address_t *saddr, 
 					  struct flowi *fl, struct xfrm_tmpl *tmpl,
@@ -1306,7 +1336,25 @@ static inline int xfrm4_udp_encap_rcv(struct sock *sk, struct sk_buff *skb)
 #endif
 
 struct xfrm_policy *xfrm_policy_alloc(gfp_t gfp);
-extern int xfrm_policy_walk(u8 type, int (*func)(struct xfrm_policy *, int, int, void*), void *);
+
+static inline void xfrm_policy_walk_init(struct xfrm_policy_walk *walk, u8 type)
+{
+	walk->cur_type = XFRM_POLICY_TYPE_MAIN;
+	walk->type = type;
+	walk->policy = NULL;
+	walk->count = 0;
+}
+
+static inline void xfrm_policy_walk_done(struct xfrm_policy_walk *walk)
+{
+	if (walk->policy != NULL) {
+		xfrm_pol_put(walk->policy);
+		walk->policy = NULL;
+	}
+}
+
+extern int xfrm_policy_walk(struct xfrm_policy_walk *walk,
+	int (*func)(struct xfrm_policy *, int, int, void*), void *);
 int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl);
 struct xfrm_policy *xfrm_policy_bysel_ctx(u8 type, int dir,
 					  struct xfrm_selector *sel,
diff --git a/net/key/af_key.c b/net/key/af_key.c
index 8b5f486ac80f..7cb6f1213360 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -1742,12 +1742,18 @@ static int pfkey_dump(struct sock *sk, struct sk_buff *skb, struct sadb_msg *hdr
 {
 	u8 proto;
 	struct pfkey_dump_data data = { .skb = skb, .hdr = hdr, .sk = sk };
+	struct xfrm_state_walk walk;
+	int rc;
 
 	proto = pfkey_satype2proto(hdr->sadb_msg_satype);
 	if (proto == 0)
 		return -EINVAL;
 
-	return xfrm_state_walk(proto, dump_sa, &data);
+	xfrm_state_walk_init(&walk, proto);
+	rc = xfrm_state_walk(&walk, dump_sa, &data);
+	xfrm_state_walk_done(&walk);
+
+	return rc;
 }
 
 static int pfkey_promisc(struct sock *sk, struct sk_buff *skb, struct sadb_msg *hdr, void **ext_hdrs)
@@ -1780,7 +1786,9 @@ static int check_reqid(struct xfrm_policy *xp, int dir, int count, void *ptr)
 
 static u32 gen_reqid(void)
 {
+	struct xfrm_policy_walk walk;
 	u32 start;
+	int rc;
 	static u32 reqid = IPSEC_MANUAL_REQID_MAX;
 
 	start = reqid;
@@ -1788,8 +1796,10 @@ static u32 gen_reqid(void)
 		++reqid;
 		if (reqid == 0)
 			reqid = IPSEC_MANUAL_REQID_MAX+1;
-		if (xfrm_policy_walk(XFRM_POLICY_TYPE_MAIN, check_reqid,
-				     (void*)&reqid) != -EEXIST)
+		xfrm_policy_walk_init(&walk, XFRM_POLICY_TYPE_MAIN);
+		rc = xfrm_policy_walk(&walk, check_reqid, (void*)&reqid);
+		xfrm_policy_walk_done(&walk);
+		if (rc != -EEXIST)
 			return reqid;
 	} while (reqid != start);
 	return 0;
@@ -2665,8 +2675,14 @@ static int dump_sp(struct xfrm_policy *xp, int dir, int count, void *ptr)
 static int pfkey_spddump(struct sock *sk, struct sk_buff *skb, struct sadb_msg *hdr, void **ext_hdrs)
 {
 	struct pfkey_dump_data data = { .skb = skb, .hdr = hdr, .sk = sk };
+	struct xfrm_policy_walk walk;
+	int rc;
+
+	xfrm_policy_walk_init(&walk, XFRM_POLICY_TYPE_MAIN);
+	rc = xfrm_policy_walk(&walk, dump_sp, &data);
+	xfrm_policy_walk_done(&walk);
 
-	return xfrm_policy_walk(XFRM_POLICY_TYPE_MAIN, dump_sp, &data);
+	return rc;
 }
 
 static int key_notify_policy_flush(struct km_event *c)
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 9fc4c315f6cd..bae94a8031a2 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -46,6 +46,7 @@ EXPORT_SYMBOL(xfrm_cfg_mutex);
 
 static DEFINE_RWLOCK(xfrm_policy_lock);
 
+static struct list_head xfrm_policy_bytype[XFRM_POLICY_TYPE_MAX];
 unsigned int xfrm_policy_count[XFRM_POLICY_MAX*2];
 EXPORT_SYMBOL(xfrm_policy_count);
 
@@ -208,6 +209,7 @@ struct xfrm_policy *xfrm_policy_alloc(gfp_t gfp)
 	policy = kzalloc(sizeof(struct xfrm_policy), gfp);
 
 	if (policy) {
+		INIT_LIST_HEAD(&policy->bytype);
 		INIT_HLIST_NODE(&policy->bydst);
 		INIT_HLIST_NODE(&policy->byidx);
 		rwlock_init(&policy->lock);
@@ -230,6 +232,10 @@ void xfrm_policy_destroy(struct xfrm_policy *policy)
 	if (del_timer(&policy->timer))
 		BUG();
 
+	write_lock_bh(&xfrm_policy_lock);
+	list_del(&policy->bytype);
+	write_unlock_bh(&xfrm_policy_lock);
+
 	security_xfrm_policy_free(policy);
 	kfree(policy);
 }
@@ -584,6 +590,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 	policy->curlft.use_time = 0;
 	if (!mod_timer(&policy->timer, jiffies + HZ))
 		xfrm_pol_hold(policy);
+	list_add_tail(&policy->bytype, &xfrm_policy_bytype[policy->type]);
 	write_unlock_bh(&xfrm_policy_lock);
 
 	if (delpol)
@@ -822,57 +829,60 @@ out:
 }
 EXPORT_SYMBOL(xfrm_policy_flush);
 
-int xfrm_policy_walk(u8 type, int (*func)(struct xfrm_policy *, int, int, void*),
+int xfrm_policy_walk(struct xfrm_policy_walk *walk,
+		     int (*func)(struct xfrm_policy *, int, int, void*),
 		     void *data)
 {
-	struct xfrm_policy *pol, *last = NULL;
-	struct hlist_node *entry;
-	int dir, last_dir = 0, count, error;
+	struct xfrm_policy *old, *pol, *last = NULL;
+	int error = 0;
+
+	if (walk->type >= XFRM_POLICY_TYPE_MAX &&
+	    walk->type != XFRM_POLICY_TYPE_ANY)
+		return -EINVAL;
 
+	if (walk->policy == NULL && walk->count != 0)
+		return 0;
+
+	old = pol = walk->policy;
+	walk->policy = NULL;
 	read_lock_bh(&xfrm_policy_lock);
-	count = 0;
 
-	for (dir = 0; dir < 2*XFRM_POLICY_MAX; dir++) {
-		struct hlist_head *table = xfrm_policy_bydst[dir].table;
-		int i;
+	for (; walk->cur_type < XFRM_POLICY_TYPE_MAX; walk->cur_type++) {
+		if (walk->type != walk->cur_type &&
+		    walk->type != XFRM_POLICY_TYPE_ANY)
+			continue;
 
-		hlist_for_each_entry(pol, entry,
-				     &xfrm_policy_inexact[dir], bydst) {
-			if (pol->type != type)
+		if (pol == NULL) {
+			pol = list_first_entry(&xfrm_policy_bytype[walk->cur_type],
+					       struct xfrm_policy, bytype);
+		}
+		list_for_each_entry_from(pol, &xfrm_policy_bytype[walk->cur_type], bytype) {
+			if (pol->dead)
 				continue;
 			if (last) {
-				error = func(last, last_dir % XFRM_POLICY_MAX,
-					     count, data);
-				if (error)
+				error = func(last, xfrm_policy_id2dir(last->index),
+					     walk->count, data);
+				if (error) {
+					xfrm_pol_hold(last);
+					walk->policy = last;
 					goto out;
-			}
-			last = pol;
-			last_dir = dir;
-			count++;
-		}
-		for (i = xfrm_policy_bydst[dir].hmask; i >= 0; i--) {
-			hlist_for_each_entry(pol, entry, table + i, bydst) {
-				if (pol->type != type)
-					continue;
-				if (last) {
-					error = func(last, last_dir % XFRM_POLICY_MAX,
-						     count, data);
-					if (error)
-						goto out;
 				}
-				last = pol;
-				last_dir = dir;
-				count++;
 			}
+			last = pol;
+			walk->count++;
 		}
+		pol = NULL;
 	}
-	if (count == 0) {
+	if (walk->count == 0) {
 		error = -ENOENT;
 		goto out;
 	}
-	error = func(last, last_dir % XFRM_POLICY_MAX, 0, data);
+	if (last)
+		error = func(last, xfrm_policy_id2dir(last->index), 0, data);
 out:
 	read_unlock_bh(&xfrm_policy_lock);
+	if (old != NULL)
+		xfrm_pol_put(old);
 	return error;
 }
 EXPORT_SYMBOL(xfrm_policy_walk);
@@ -2365,6 +2375,9 @@ static void __init xfrm_policy_init(void)
 			panic("XFRM: failed to allocate bydst hash\n");
 	}
 
+	for (dir = 0; dir < XFRM_POLICY_TYPE_MAX; dir++)
+		INIT_LIST_HEAD(&xfrm_policy_bytype[dir]);
+
 	INIT_WORK(&xfrm_policy_gc_work, xfrm_policy_gc_task);
 	register_netdevice_notifier(&xfrm_dev_notifier);
 }
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 7ba65e82941c..9880b792e6a5 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -50,6 +50,7 @@ static DEFINE_SPINLOCK(xfrm_state_lock);
  * Main use is finding SA after policy selected tunnel or transport mode.
  * Also, it can be used by ah/esp icmp error handler to find offending SA.
  */
+static LIST_HEAD(xfrm_state_all);
 static struct hlist_head *xfrm_state_bydst __read_mostly;
 static struct hlist_head *xfrm_state_bysrc __read_mostly;
 static struct hlist_head *xfrm_state_byspi __read_mostly;
@@ -510,6 +511,7 @@ struct xfrm_state *xfrm_state_alloc(void)
 	if (x) {
 		atomic_set(&x->refcnt, 1);
 		atomic_set(&x->tunnel_users, 0);
+		INIT_LIST_HEAD(&x->all);
 		INIT_HLIST_NODE(&x->bydst);
 		INIT_HLIST_NODE(&x->bysrc);
 		INIT_HLIST_NODE(&x->byspi);
@@ -533,6 +535,10 @@ void __xfrm_state_destroy(struct xfrm_state *x)
 {
 	BUG_TRAP(x->km.state == XFRM_STATE_DEAD);
 
+	spin_lock_bh(&xfrm_state_lock);
+	list_del(&x->all);
+	spin_unlock_bh(&xfrm_state_lock);
+
 	spin_lock_bh(&xfrm_state_gc_lock);
 	hlist_add_head(&x->bydst, &xfrm_state_gc_list);
 	spin_unlock_bh(&xfrm_state_gc_lock);
@@ -909,6 +915,8 @@ static void __xfrm_state_insert(struct xfrm_state *x)
 
 	x->genid = ++xfrm_state_genid;
 
+	list_add_tail(&x->all, &xfrm_state_all);
+
 	h = xfrm_dst_hash(&x->id.daddr, &x->props.saddr,
 			  x->props.reqid, x->props.family);
 	hlist_add_head(&x->bydst, xfrm_state_bydst+h);
@@ -1518,36 +1526,47 @@ unlock:
 }
 EXPORT_SYMBOL(xfrm_alloc_spi);
 
-int xfrm_state_walk(u8 proto, int (*func)(struct xfrm_state *, int, void*),
+int xfrm_state_walk(struct xfrm_state_walk *walk,
+		    int (*func)(struct xfrm_state *, int, void*),
 		    void *data)
 {
-	int i;
-	struct xfrm_state *x, *last = NULL;
-	struct hlist_node *entry;
-	int count = 0;
+	struct xfrm_state *old, *x, *last = NULL;
 	int err = 0;
 
+	if (walk->state == NULL && walk->count != 0)
+		return 0;
+
+	old = x = walk->state;
+	walk->state = NULL;
 	spin_lock_bh(&xfrm_state_lock);
-	for (i = 0; i <= xfrm_state_hmask; i++) {
-		hlist_for_each_entry(x, entry, xfrm_state_bydst+i, bydst) {
-			if (!xfrm_id_proto_match(x->id.proto, proto))
-				continue;
-			if (last) {
-				err = func(last, count, data);
-				if (err)
-					goto out;
+	if (x == NULL)
+		x = list_first_entry(&xfrm_state_all, struct xfrm_state, all);
+	list_for_each_entry_from(x, &xfrm_state_all, all) {
+		if (x->km.state == XFRM_STATE_DEAD)
+			continue;
+		if (!xfrm_id_proto_match(x->id.proto, walk->proto))
+			continue;
+		if (last) {
+			err = func(last, walk->count, data);
+			if (err) {
+				xfrm_state_hold(last);
+				walk->state = last;
+				goto out;
 			}
-			last = x;
-			count++;
 		}
+		last = x;
+		walk->count++;
 	}
-	if (count == 0) {
+	if (walk->count == 0) {
 		err = -ENOENT;
 		goto out;
 	}
-	err = func(last, 0, data);
+	if (last)
+		err = func(last, 0, data);
 out:
 	spin_unlock_bh(&xfrm_state_lock);
+	if (old != NULL)
+		xfrm_state_put(old);
 	return err;
 }
 EXPORT_SYMBOL(xfrm_state_walk);
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index f971ca5645f8..f5fd5b3147cc 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -532,8 +532,6 @@ struct xfrm_dump_info {
 	struct sk_buff *out_skb;
 	u32 nlmsg_seq;
 	u16 nlmsg_flags;
-	int start_idx;
-	int this_idx;
 };
 
 static int copy_sec_ctx(struct xfrm_sec_ctx *s, struct sk_buff *skb)
@@ -600,9 +598,6 @@ static int dump_one_state(struct xfrm_state *x, int count, void *ptr)
 	struct nlmsghdr *nlh;
 	int err;
 
-	if (sp->this_idx < sp->start_idx)
-		goto out;
-
 	nlh = nlmsg_put(skb, NETLINK_CB(in_skb).pid, sp->nlmsg_seq,
 			XFRM_MSG_NEWSA, sizeof(*p), sp->nlmsg_flags);
 	if (nlh == NULL)
@@ -615,8 +610,6 @@ static int dump_one_state(struct xfrm_state *x, int count, void *ptr)
 		goto nla_put_failure;
 
 	nlmsg_end(skb, nlh);
-out:
-	sp->this_idx++;
 	return 0;
 
 nla_put_failure:
@@ -624,18 +617,32 @@ nla_put_failure:
 	return err;
 }
 
+static int xfrm_dump_sa_done(struct netlink_callback *cb)
+{
+	struct xfrm_state_walk *walk = (struct xfrm_state_walk *) &cb->args[1];
+	xfrm_state_walk_done(walk);
+	return 0;
+}
+
 static int xfrm_dump_sa(struct sk_buff *skb, struct netlink_callback *cb)
 {
+	struct xfrm_state_walk *walk = (struct xfrm_state_walk *) &cb->args[1];
 	struct xfrm_dump_info info;
 
+	BUILD_BUG_ON(sizeof(struct xfrm_state_walk) >
+		     sizeof(cb->args) - sizeof(cb->args[0]));
+
 	info.in_skb = cb->skb;
 	info.out_skb = skb;
 	info.nlmsg_seq = cb->nlh->nlmsg_seq;
 	info.nlmsg_flags = NLM_F_MULTI;
-	info.this_idx = 0;
-	info.start_idx = cb->args[0];
-	(void) xfrm_state_walk(0, dump_one_state, &info);
-	cb->args[0] = info.this_idx;
+
+	if (!cb->args[0]) {
+		cb->args[0] = 1;
+		xfrm_state_walk_init(walk, 0);
+	}
+
+	(void) xfrm_state_walk(walk, dump_one_state, &info);
 
 	return skb->len;
 }
@@ -654,7 +661,6 @@ static struct sk_buff *xfrm_state_netlink(struct sk_buff *in_skb,
 	info.out_skb = skb;
 	info.nlmsg_seq = seq;
 	info.nlmsg_flags = 0;
-	info.this_idx = info.start_idx = 0;
 
 	if (dump_one_state(x, 0, &info)) {
 		kfree_skb(skb);
@@ -1232,9 +1238,6 @@ static int dump_one_policy(struct xfrm_policy *xp, int dir, int count, void *ptr
 	struct sk_buff *skb = sp->out_skb;
 	struct nlmsghdr *nlh;
 
-	if (sp->this_idx < sp->start_idx)
-		goto out;
-
 	nlh = nlmsg_put(skb, NETLINK_CB(in_skb).pid, sp->nlmsg_seq,
 			XFRM_MSG_NEWPOLICY, sizeof(*p), sp->nlmsg_flags);
 	if (nlh == NULL)
@@ -1250,8 +1253,6 @@ static int dump_one_policy(struct xfrm_policy *xp, int dir, int count, void *ptr
 		goto nlmsg_failure;
 
 	nlmsg_end(skb, nlh);
-out:
-	sp->this_idx++;
 	return 0;
 
 nlmsg_failure:
@@ -1259,21 +1260,33 @@ nlmsg_failure:
 	return -EMSGSIZE;
 }
 
+static int xfrm_dump_policy_done(struct netlink_callback *cb)
+{
+	struct xfrm_policy_walk *walk = (struct xfrm_policy_walk *) &cb->args[1];
+
+	xfrm_policy_walk_done(walk);
+	return 0;
+}
+
 static int xfrm_dump_policy(struct sk_buff *skb, struct netlink_callback *cb)
 {
+	struct xfrm_policy_walk *walk = (struct xfrm_policy_walk *) &cb->args[1];
 	struct xfrm_dump_info info;
 
+	BUILD_BUG_ON(sizeof(struct xfrm_policy_walk) >
+		     sizeof(cb->args) - sizeof(cb->args[0]));
+
 	info.in_skb = cb->skb;
 	info.out_skb = skb;
 	info.nlmsg_seq = cb->nlh->nlmsg_seq;
 	info.nlmsg_flags = NLM_F_MULTI;
-	info.this_idx = 0;
-	info.start_idx = cb->args[0];
-	(void) xfrm_policy_walk(XFRM_POLICY_TYPE_MAIN, dump_one_policy, &info);
-#ifdef CONFIG_XFRM_SUB_POLICY
-	(void) xfrm_policy_walk(XFRM_POLICY_TYPE_SUB, dump_one_policy, &info);
-#endif
-	cb->args[0] = info.this_idx;
+
+	if (!cb->args[0]) {
+		cb->args[0] = 1;
+		xfrm_policy_walk_init(walk, XFRM_POLICY_TYPE_ANY);
+	}
+
+	(void) xfrm_policy_walk(walk, dump_one_policy, &info);
 
 	return skb->len;
 }
@@ -1293,7 +1306,6 @@ static struct sk_buff *xfrm_policy_netlink(struct sk_buff *in_skb,
 	info.out_skb = skb;
 	info.nlmsg_seq = seq;
 	info.nlmsg_flags = 0;
-	info.this_idx = info.start_idx = 0;
 
 	if (dump_one_policy(xp, dir, 0, &info) < 0) {
 		kfree_skb(skb);
@@ -1891,15 +1903,18 @@ static const struct nla_policy xfrma_policy[XFRMA_MAX+1] = {
 static struct xfrm_link {
 	int (*doit)(struct sk_buff *, struct nlmsghdr *, struct nlattr **);
 	int (*dump)(struct sk_buff *, struct netlink_callback *);
+	int (*done)(struct netlink_callback *);
 } xfrm_dispatch[XFRM_NR_MSGTYPES] = {
 	[XFRM_MSG_NEWSA       - XFRM_MSG_BASE] = { .doit = xfrm_add_sa        },
 	[XFRM_MSG_DELSA       - XFRM_MSG_BASE] = { .doit = xfrm_del_sa        },
 	[XFRM_MSG_GETSA       - XFRM_MSG_BASE] = { .doit = xfrm_get_sa,
-						   .dump = xfrm_dump_sa       },
+						   .dump = xfrm_dump_sa,
+						   .done = xfrm_dump_sa_done  },
 	[XFRM_MSG_NEWPOLICY   - XFRM_MSG_BASE] = { .doit = xfrm_add_policy    },
 	[XFRM_MSG_DELPOLICY   - XFRM_MSG_BASE] = { .doit = xfrm_get_policy    },
 	[XFRM_MSG_GETPOLICY   - XFRM_MSG_BASE] = { .doit = xfrm_get_policy,
-						   .dump = xfrm_dump_policy   },
+						   .dump = xfrm_dump_policy,
+						   .done = xfrm_dump_policy_done },
 	[XFRM_MSG_ALLOCSPI    - XFRM_MSG_BASE] = { .doit = xfrm_alloc_userspi },
 	[XFRM_MSG_ACQUIRE     - XFRM_MSG_BASE] = { .doit = xfrm_add_acquire   },
 	[XFRM_MSG_EXPIRE      - XFRM_MSG_BASE] = { .doit = xfrm_add_sa_expire },
@@ -1938,7 +1953,7 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 		if (link->dump == NULL)
 			return -EINVAL;
 
-		return netlink_dump_start(xfrm_nl, skb, nlh, link->dump, NULL);
+		return netlink_dump_start(xfrm_nl, skb, nlh, link->dump, link->done);
 	}
 
 	err = nlmsg_parse(nlh, xfrm_msg_min[type], attrs, XFRMA_MAX,
-- 
cgit v1.2.3


From b59931649256685f294d2d163a4f6d6286fbff05 Mon Sep 17 00:00:00 2001
From: Roland McGrath <roland@redhat.com>
Date: Tue, 26 Feb 2008 13:20:58 -0800
Subject: elfcore-compat fix uid/gid types

I overlooked the difference between __kernel_uid_t and uid_t when defining
struct compat_elf_prpsinfo.  The result is a regression in 32-bit core
dumps on x86_64, where the NT_PRPSINFO note has the wrong size and layout.
This patch fixes it.

Signed-off-by: Roland McGrath <roland@redhat.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/elfcore-compat.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/elfcore-compat.h b/include/linux/elfcore-compat.h
index 532d13adabc4..0a90e1c3a422 100644
--- a/include/linux/elfcore-compat.h
+++ b/include/linux/elfcore-compat.h
@@ -45,8 +45,8 @@ struct compat_elf_prpsinfo
 	char				pr_zomb;
 	char				pr_nice;
 	compat_ulong_t			pr_flag;
-	compat_uid_t			pr_uid;
-	compat_gid_t			pr_gid;
+	__compat_uid_t			pr_uid;
+	__compat_gid_t			pr_gid;
 	compat_pid_t			pr_pid, pr_ppid, pr_pgrp, pr_sid;
 	char				pr_fname[16];
 	char				pr_psargs[ELF_PRARGSZ];
-- 
cgit v1.2.3


From 57ce36feb4d1281247755bc445bae77728298955 Mon Sep 17 00:00:00 2001
From: Uwe Kleine-König <Uwe.Kleine-Koenig@digi.com>
Date: Mon, 25 Feb 2008 16:45:03 +0100
Subject: let __dec_zone_page_state use __dec_zone_state
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This removes code duplication and makes __dec_zone_page_state look like
__inc_zone_page_state.

Signed-off-by: Uwe Kleine-König <Uwe.Kleine-Koenig@digi.com>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/vmstat.h | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h
index 75370ec0923e..9f1b4b46151e 100644
--- a/include/linux/vmstat.h
+++ b/include/linux/vmstat.h
@@ -246,8 +246,7 @@ static inline void __dec_zone_state(struct zone *zone, enum zone_stat_item item)
 static inline void __dec_zone_page_state(struct page *page,
 			enum zone_stat_item item)
 {
-	atomic_long_dec(&page_zone(page)->vm_stat[item]);
-	atomic_long_dec(&vm_stat[item]);
+	__dec_zone_state(page_zone(page), item);
 }
 
 /*
-- 
cgit v1.2.3


From 2232c2d8e0a6a31061dec311f3d1cf7624bc14f1 Mon Sep 17 00:00:00 2001
From: Steven Rostedt <rostedt@goodmis.org>
Date: Fri, 29 Feb 2008 18:46:50 +0100
Subject: rcu: add support for dynamic ticks and preempt rcu

The PREEMPT-RCU can get stuck if a CPU goes idle and NO_HZ is set. The
idle CPU will not progress the RCU through its grace period and a
synchronize_rcu my get stuck. Without this patch I have a box that will
not boot when PREEMPT_RCU and NO_HZ are set. That same box boots fine
with this patch.

This patch comes from the -rt kernel where it has been tested for
several months.

Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/hardirq.h    |  10 ++
 include/linux/rcuclassic.h |   3 +
 include/linux/rcupreempt.h |  22 +++++
 kernel/rcupreempt.c        | 224 ++++++++++++++++++++++++++++++++++++++++++++-
 kernel/softirq.c           |   1 +
 kernel/time/tick-sched.c   |   3 +
 6 files changed, 259 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h
index 2961ec788046..49829988bfa0 100644
--- a/include/linux/hardirq.h
+++ b/include/linux/hardirq.h
@@ -109,6 +109,14 @@ static inline void account_system_vtime(struct task_struct *tsk)
 }
 #endif
 
+#if defined(CONFIG_PREEMPT_RCU) && defined(CONFIG_NO_HZ)
+extern void rcu_irq_enter(void);
+extern void rcu_irq_exit(void);
+#else
+# define rcu_irq_enter() do { } while (0)
+# define rcu_irq_exit() do { } while (0)
+#endif /* CONFIG_PREEMPT_RCU */
+
 /*
  * It is safe to do non-atomic ops on ->hardirq_context,
  * because NMI handlers may not preempt and the ops are
@@ -117,6 +125,7 @@ static inline void account_system_vtime(struct task_struct *tsk)
  */
 #define __irq_enter()					\
 	do {						\
+		rcu_irq_enter();			\
 		account_system_vtime(current);		\
 		add_preempt_count(HARDIRQ_OFFSET);	\
 		trace_hardirq_enter();			\
@@ -135,6 +144,7 @@ extern void irq_enter(void);
 		trace_hardirq_exit();			\
 		account_system_vtime(current);		\
 		sub_preempt_count(HARDIRQ_OFFSET);	\
+		rcu_irq_exit();				\
 	} while (0)
 
 /*
diff --git a/include/linux/rcuclassic.h b/include/linux/rcuclassic.h
index 4d6624260b4c..b3dccd68629e 100644
--- a/include/linux/rcuclassic.h
+++ b/include/linux/rcuclassic.h
@@ -160,5 +160,8 @@ extern void rcu_restart_cpu(int cpu);
 extern long rcu_batches_completed(void);
 extern long rcu_batches_completed_bh(void);
 
+#define rcu_enter_nohz()	do { } while (0)
+#define rcu_exit_nohz()		do { } while (0)
+
 #endif /* __KERNEL__ */
 #endif /* __LINUX_RCUCLASSIC_H */
diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h
index 60c2a033b19e..01152ed532c8 100644
--- a/include/linux/rcupreempt.h
+++ b/include/linux/rcupreempt.h
@@ -82,5 +82,27 @@ extern struct rcupreempt_trace *rcupreempt_trace_cpu(int cpu);
 
 struct softirq_action;
 
+#ifdef CONFIG_NO_HZ
+DECLARE_PER_CPU(long, dynticks_progress_counter);
+
+static inline void rcu_enter_nohz(void)
+{
+	__get_cpu_var(dynticks_progress_counter)++;
+	WARN_ON(__get_cpu_var(dynticks_progress_counter) & 0x1);
+	mb();
+}
+
+static inline void rcu_exit_nohz(void)
+{
+	mb();
+	__get_cpu_var(dynticks_progress_counter)++;
+	WARN_ON(!(__get_cpu_var(dynticks_progress_counter) & 0x1));
+}
+
+#else /* CONFIG_NO_HZ */
+#define rcu_enter_nohz()	do { } while (0)
+#define rcu_exit_nohz()		do { } while (0)
+#endif /* CONFIG_NO_HZ */
+
 #endif /* __KERNEL__ */
 #endif /* __LINUX_RCUPREEMPT_H */
diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c
index 987cfb7ade89..c7c52096df48 100644
--- a/kernel/rcupreempt.c
+++ b/kernel/rcupreempt.c
@@ -23,6 +23,10 @@
  *		to Suparna Bhattacharya for pushing me completely away
  *		from atomic instructions on the read side.
  *
+ *  - Added handling of Dynamic Ticks
+ *      Copyright 2007 - Paul E. Mckenney <paulmck@us.ibm.com>
+ *                     - Steven Rostedt <srostedt@redhat.com>
+ *
  * Papers:  http://www.rdrop.com/users/paulmck/RCU
  *
  * Design Document: http://lwn.net/Articles/253651/
@@ -409,6 +413,212 @@ static void __rcu_advance_callbacks(struct rcu_data *rdp)
 	}
 }
 
+#ifdef CONFIG_NO_HZ
+
+DEFINE_PER_CPU(long, dynticks_progress_counter) = 1;
+static DEFINE_PER_CPU(long, rcu_dyntick_snapshot);
+static DEFINE_PER_CPU(int, rcu_update_flag);
+
+/**
+ * rcu_irq_enter - Called from Hard irq handlers and NMI/SMI.
+ *
+ * If the CPU was idle with dynamic ticks active, this updates the
+ * dynticks_progress_counter to let the RCU handling know that the
+ * CPU is active.
+ */
+void rcu_irq_enter(void)
+{
+	int cpu = smp_processor_id();
+
+	if (per_cpu(rcu_update_flag, cpu))
+		per_cpu(rcu_update_flag, cpu)++;
+
+	/*
+	 * Only update if we are coming from a stopped ticks mode
+	 * (dynticks_progress_counter is even).
+	 */
+	if (!in_interrupt() &&
+	    (per_cpu(dynticks_progress_counter, cpu) & 0x1) == 0) {
+		/*
+		 * The following might seem like we could have a race
+		 * with NMI/SMIs. But this really isn't a problem.
+		 * Here we do a read/modify/write, and the race happens
+		 * when an NMI/SMI comes in after the read and before
+		 * the write. But NMI/SMIs will increment this counter
+		 * twice before returning, so the zero bit will not
+		 * be corrupted by the NMI/SMI which is the most important
+		 * part.
+		 *
+		 * The only thing is that we would bring back the counter
+		 * to a postion that it was in during the NMI/SMI.
+		 * But the zero bit would be set, so the rest of the
+		 * counter would again be ignored.
+		 *
+		 * On return from the IRQ, the counter may have the zero
+		 * bit be 0 and the counter the same as the return from
+		 * the NMI/SMI. If the state machine was so unlucky to
+		 * see that, it still doesn't matter, since all
+		 * RCU read-side critical sections on this CPU would
+		 * have already completed.
+		 */
+		per_cpu(dynticks_progress_counter, cpu)++;
+		/*
+		 * The following memory barrier ensures that any
+		 * rcu_read_lock() primitives in the irq handler
+		 * are seen by other CPUs to follow the above
+		 * increment to dynticks_progress_counter. This is
+		 * required in order for other CPUs to correctly
+		 * determine when it is safe to advance the RCU
+		 * grace-period state machine.
+		 */
+		smp_mb(); /* see above block comment. */
+		/*
+		 * Since we can't determine the dynamic tick mode from
+		 * the dynticks_progress_counter after this routine,
+		 * we use a second flag to acknowledge that we came
+		 * from an idle state with ticks stopped.
+		 */
+		per_cpu(rcu_update_flag, cpu)++;
+		/*
+		 * If we take an NMI/SMI now, they will also increment
+		 * the rcu_update_flag, and will not update the
+		 * dynticks_progress_counter on exit. That is for
+		 * this IRQ to do.
+		 */
+	}
+}
+
+/**
+ * rcu_irq_exit - Called from exiting Hard irq context.
+ *
+ * If the CPU was idle with dynamic ticks active, update the
+ * dynticks_progress_counter to put let the RCU handling be
+ * aware that the CPU is going back to idle with no ticks.
+ */
+void rcu_irq_exit(void)
+{
+	int cpu = smp_processor_id();
+
+	/*
+	 * rcu_update_flag is set if we interrupted the CPU
+	 * when it was idle with ticks stopped.
+	 * Once this occurs, we keep track of interrupt nesting
+	 * because a NMI/SMI could also come in, and we still
+	 * only want the IRQ that started the increment of the
+	 * dynticks_progress_counter to be the one that modifies
+	 * it on exit.
+	 */
+	if (per_cpu(rcu_update_flag, cpu)) {
+		if (--per_cpu(rcu_update_flag, cpu))
+			return;
+
+		/* This must match the interrupt nesting */
+		WARN_ON(in_interrupt());
+
+		/*
+		 * If an NMI/SMI happens now we are still
+		 * protected by the dynticks_progress_counter being odd.
+		 */
+
+		/*
+		 * The following memory barrier ensures that any
+		 * rcu_read_unlock() primitives in the irq handler
+		 * are seen by other CPUs to preceed the following
+		 * increment to dynticks_progress_counter. This
+		 * is required in order for other CPUs to determine
+		 * when it is safe to advance the RCU grace-period
+		 * state machine.
+		 */
+		smp_mb(); /* see above block comment. */
+		per_cpu(dynticks_progress_counter, cpu)++;
+		WARN_ON(per_cpu(dynticks_progress_counter, cpu) & 0x1);
+	}
+}
+
+static void dyntick_save_progress_counter(int cpu)
+{
+	per_cpu(rcu_dyntick_snapshot, cpu) =
+		per_cpu(dynticks_progress_counter, cpu);
+}
+
+static inline int
+rcu_try_flip_waitack_needed(int cpu)
+{
+	long curr;
+	long snap;
+
+	curr = per_cpu(dynticks_progress_counter, cpu);
+	snap = per_cpu(rcu_dyntick_snapshot, cpu);
+	smp_mb(); /* force ordering with cpu entering/leaving dynticks. */
+
+	/*
+	 * If the CPU remained in dynticks mode for the entire time
+	 * and didn't take any interrupts, NMIs, SMIs, or whatever,
+	 * then it cannot be in the middle of an rcu_read_lock(), so
+	 * the next rcu_read_lock() it executes must use the new value
+	 * of the counter.  So we can safely pretend that this CPU
+	 * already acknowledged the counter.
+	 */
+
+	if ((curr == snap) && ((curr & 0x1) == 0))
+		return 0;
+
+	/*
+	 * If the CPU passed through or entered a dynticks idle phase with
+	 * no active irq handlers, then, as above, we can safely pretend
+	 * that this CPU already acknowledged the counter.
+	 */
+
+	if ((curr - snap) > 2 || (snap & 0x1) == 0)
+		return 0;
+
+	/* We need this CPU to explicitly acknowledge the counter flip. */
+
+	return 1;
+}
+
+static inline int
+rcu_try_flip_waitmb_needed(int cpu)
+{
+	long curr;
+	long snap;
+
+	curr = per_cpu(dynticks_progress_counter, cpu);
+	snap = per_cpu(rcu_dyntick_snapshot, cpu);
+	smp_mb(); /* force ordering with cpu entering/leaving dynticks. */
+
+	/*
+	 * If the CPU remained in dynticks mode for the entire time
+	 * and didn't take any interrupts, NMIs, SMIs, or whatever,
+	 * then it cannot have executed an RCU read-side critical section
+	 * during that time, so there is no need for it to execute a
+	 * memory barrier.
+	 */
+
+	if ((curr == snap) && ((curr & 0x1) == 0))
+		return 0;
+
+	/*
+	 * If the CPU either entered or exited an outermost interrupt,
+	 * SMI, NMI, or whatever handler, then we know that it executed
+	 * a memory barrier when doing so.  So we don't need another one.
+	 */
+	if (curr != snap)
+		return 0;
+
+	/* We need the CPU to execute a memory barrier. */
+
+	return 1;
+}
+
+#else /* !CONFIG_NO_HZ */
+
+# define dyntick_save_progress_counter(cpu)	do { } while (0)
+# define rcu_try_flip_waitack_needed(cpu)	(1)
+# define rcu_try_flip_waitmb_needed(cpu)	(1)
+
+#endif /* CONFIG_NO_HZ */
+
 /*
  * Get here when RCU is idle.  Decide whether we need to
  * move out of idle state, and return non-zero if so.
@@ -447,8 +657,10 @@ rcu_try_flip_idle(void)
 
 	/* Now ask each CPU for acknowledgement of the flip. */
 
-	for_each_cpu_mask(cpu, rcu_cpu_online_map)
+	for_each_cpu_mask(cpu, rcu_cpu_online_map) {
 		per_cpu(rcu_flip_flag, cpu) = rcu_flipped;
+		dyntick_save_progress_counter(cpu);
+	}
 
 	return 1;
 }
@@ -464,7 +676,8 @@ rcu_try_flip_waitack(void)
 
 	RCU_TRACE_ME(rcupreempt_trace_try_flip_a1);
 	for_each_cpu_mask(cpu, rcu_cpu_online_map)
-		if (per_cpu(rcu_flip_flag, cpu) != rcu_flip_seen) {
+		if (rcu_try_flip_waitack_needed(cpu) &&
+		    per_cpu(rcu_flip_flag, cpu) != rcu_flip_seen) {
 			RCU_TRACE_ME(rcupreempt_trace_try_flip_ae1);
 			return 0;
 		}
@@ -509,8 +722,10 @@ rcu_try_flip_waitzero(void)
 	smp_mb();  /*  ^^^^^^^^^^^^ */
 
 	/* Call for a memory barrier from each CPU. */
-	for_each_cpu_mask(cpu, rcu_cpu_online_map)
+	for_each_cpu_mask(cpu, rcu_cpu_online_map) {
 		per_cpu(rcu_mb_flag, cpu) = rcu_mb_needed;
+		dyntick_save_progress_counter(cpu);
+	}
 
 	RCU_TRACE_ME(rcupreempt_trace_try_flip_z2);
 	return 1;
@@ -528,7 +743,8 @@ rcu_try_flip_waitmb(void)
 
 	RCU_TRACE_ME(rcupreempt_trace_try_flip_m1);
 	for_each_cpu_mask(cpu, rcu_cpu_online_map)
-		if (per_cpu(rcu_mb_flag, cpu) != rcu_mb_done) {
+		if (rcu_try_flip_waitmb_needed(cpu) &&
+		    per_cpu(rcu_mb_flag, cpu) != rcu_mb_done) {
 			RCU_TRACE_ME(rcupreempt_trace_try_flip_me1);
 			return 0;
 		}
diff --git a/kernel/softirq.c b/kernel/softirq.c
index 5b3aea5f471e..31e9f2a47928 100644
--- a/kernel/softirq.c
+++ b/kernel/softirq.c
@@ -313,6 +313,7 @@ void irq_exit(void)
 	/* Make sure that timer wheel updates are propagated */
 	if (!in_interrupt() && idle_cpu(smp_processor_id()) && !need_resched())
 		tick_nohz_stop_sched_tick();
+	rcu_irq_exit();
 #endif
 	preempt_enable_no_resched();
 }
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index fa9bb73dbdb4..2968298f8f36 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -282,6 +282,7 @@ void tick_nohz_stop_sched_tick(void)
 			ts->idle_tick = ts->sched_timer.expires;
 			ts->tick_stopped = 1;
 			ts->idle_jiffies = last_jiffies;
+			rcu_enter_nohz();
 		}
 
 		/*
@@ -375,6 +376,8 @@ void tick_nohz_restart_sched_tick(void)
 		return;
 	}
 
+	rcu_exit_nohz();
+
 	/* Update jiffies first */
 	select_nohz_load_balancer(0);
 	now = ktime_get();
-- 
cgit v1.2.3


From 9b0f976f27f00a81cf47643d90854659626795b4 Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Fri, 29 Feb 2008 11:13:15 -0800
Subject: [INET]: Remove struct net_proto_family* from _init calls.

struct net_proto_family* is not used in icmp[v6]_init, ndisc_init,
igmp_init and tcp_v4_init. Remove it.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Daniel Lezcano <dlezcano@fr.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/icmpv6.h | 2 +-
 include/net/icmp.h     | 2 +-
 include/net/ndisc.h    | 4 ++--
 include/net/tcp.h      | 2 +-
 net/ipv4/af_inet.c     | 4 ++--
 net/ipv4/icmp.c        | 2 +-
 net/ipv4/tcp_ipv4.c    | 2 +-
 net/ipv6/af_inet6.c    | 6 +++---
 net/ipv6/icmp.c        | 2 +-
 net/ipv6/mcast.c       | 2 +-
 net/ipv6/ndisc.c       | 2 +-
 11 files changed, 15 insertions(+), 15 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/icmpv6.h b/include/linux/icmpv6.h
index 7c5e9817e998..8f86d6b621c8 100644
--- a/include/linux/icmpv6.h
+++ b/include/linux/icmpv6.h
@@ -176,7 +176,7 @@ extern void				icmpv6_send(struct sk_buff *skb,
 						    __u32 info, 
 						    struct net_device *dev);
 
-extern int				icmpv6_init(struct net_proto_family *ops);
+extern int				icmpv6_init(void);
 extern int				icmpv6_err_convert(int type, int code,
 							   int *err);
 extern void				icmpv6_cleanup(void);
diff --git a/include/net/icmp.h b/include/net/icmp.h
index 9f7ef3c8baef..7bf714d9d7c7 100644
--- a/include/net/icmp.h
+++ b/include/net/icmp.h
@@ -48,7 +48,7 @@ struct sk_buff;
 extern void	icmp_send(struct sk_buff *skb_in,  int type, int code, __be32 info);
 extern int	icmp_rcv(struct sk_buff *skb);
 extern int	icmp_ioctl(struct sock *sk, int cmd, unsigned long arg);
-extern void	icmp_init(struct net_proto_family *ops);
+extern void	icmp_init(void);
 extern void	icmp_out_count(unsigned char type);
 
 /* Move into dst.h ? */
diff --git a/include/net/ndisc.h b/include/net/ndisc.h
index 59b70624b056..5aedf324de66 100644
--- a/include/net/ndisc.h
+++ b/include/net/ndisc.h
@@ -77,7 +77,7 @@ struct nd_opt_hdr {
 } __attribute__((__packed__));
 
 
-extern int			ndisc_init(struct net_proto_family *ops);
+extern int			ndisc_init(void);
 
 extern void			ndisc_cleanup(void);
 
@@ -107,7 +107,7 @@ extern int			ndisc_mc_map(struct in6_addr *addr, char *buf, struct net_device *d
 /*
  *	IGMP
  */
-extern int			igmp6_init(struct net_proto_family *ops);
+extern int			igmp6_init(void);
 
 extern void			igmp6_cleanup(void);
 
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 7de4ea3a04d9..ae9774b478f5 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1373,7 +1373,7 @@ struct tcp_request_sock_ops {
 #endif
 };
 
-extern void tcp_v4_init(struct net_proto_family *ops);
+extern void tcp_v4_init(void);
 extern void tcp_init(void);
 
 #endif	/* _TCP_H */
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index c270080f370e..a7a99ac856dc 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1415,7 +1415,7 @@ static int __init inet_init(void)
 
 	ip_init();
 
-	tcp_v4_init(&inet_family_ops);
+	tcp_v4_init();
 
 	/* Setup TCP slab cache for open requests. */
 	tcp_init();
@@ -1430,7 +1430,7 @@ static int __init inet_init(void)
 	 *	Set the ICMP layer up
 	 */
 
-	icmp_init(&inet_family_ops);
+	icmp_init();
 
 	/*
 	 *	Initialise the multicast router
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index a13c074dac09..98372db66c66 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -1139,7 +1139,7 @@ static const struct icmp_control icmp_pointers[NR_ICMP_TYPES + 1] = {
 	},
 };
 
-void __init icmp_init(struct net_proto_family *ops)
+void __init icmp_init(void)
 {
 	struct inet_sock *inet;
 	int i;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 00156bf421ca..256032a41069 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -2443,7 +2443,7 @@ struct proto tcp_prot = {
 	REF_PROTO_INUSE(tcp)
 };
 
-void __init tcp_v4_init(struct net_proto_family *ops)
+void __init tcp_v4_init(void)
 {
 	if (inet_csk_ctl_sock_create(&tcp_socket, PF_INET, SOCK_RAW,
 				     IPPROTO_TCP) < 0)
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index f0aa97738746..9869f87243cf 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -808,13 +808,13 @@ static int __init inet6_init(void)
 	if (err)
 		goto sysctl_fail;
 #endif
-	err = icmpv6_init(&inet6_family_ops);
+	err = icmpv6_init();
 	if (err)
 		goto icmp_fail;
-	err = ndisc_init(&inet6_family_ops);
+	err = ndisc_init();
 	if (err)
 		goto ndisc_fail;
-	err = igmp6_init(&inet6_family_ops);
+	err = igmp6_init();
 	if (err)
 		goto igmp_fail;
 	err = ipv6_netfilter_init();
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index 121d517bf91c..b9b13a77ba30 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -780,7 +780,7 @@ drop_no_count:
  */
 static struct lock_class_key icmpv6_socket_sk_dst_lock_key;
 
-int __init icmpv6_init(struct net_proto_family *ops)
+int __init icmpv6_init(void)
 {
 	struct sock *sk;
 	int err, i, j;
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index ab228d1ea114..8ce894d90063 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -2597,7 +2597,7 @@ static const struct file_operations igmp6_mcf_seq_fops = {
 };
 #endif
 
-int __init igmp6_init(struct net_proto_family *ops)
+int __init igmp6_init(void)
 {
 	struct ipv6_pinfo *np;
 	struct sock *sk;
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 0d33a7d32125..1fc33c8c7232 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1733,7 +1733,7 @@ static int ndisc_ifinfo_sysctl_strategy(ctl_table *ctl, int __user *name,
 
 #endif
 
-int __init ndisc_init(struct net_proto_family *ops)
+int __init ndisc_init(void)
 {
 	struct ipv6_pinfo *np;
 	struct sock *sk;
-- 
cgit v1.2.3


From ee688b000d35f413f33561ec9c7d3355be561e2f Mon Sep 17 00:00:00 2001
From: Johannes Berg <johannes@sipsolutions.net>
Date: Thu, 24 Jan 2008 19:38:39 +0100
Subject: nl80211: export hardware bitrate/channel capabilities

This makes nl80211 export the hardware bitrate/channel capabilities
as registered in a wiphy.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/linux/nl80211.h | 64 ++++++++++++++++++++++++++++++++++++++++++
 net/wireless/nl80211.c  | 74 +++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 138 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h
index 9fecf902419c..63695060db9f 100644
--- a/include/linux/nl80211.h
+++ b/include/linux/nl80211.h
@@ -161,6 +161,9 @@ enum nl80211_commands {
  *	given for %NL80211_CMD_GET_STATION, nested attribute containing
  *	info as possible, see &enum nl80211_sta_stats.
  *
+ * @NL80211_ATTR_WIPHY_BANDS: Information about an operating bands,
+ *	consisting of a nested array.
+ *
  * @NL80211_ATTR_MAX: highest attribute number currently defined
  * @__NL80211_ATTR_AFTER_LAST: internal use
  */
@@ -195,6 +198,8 @@ enum nl80211_attrs {
 	NL80211_ATTR_STA_VLAN,
 	NL80211_ATTR_STA_STATS,
 
+	NL80211_ATTR_WIPHY_BANDS,
+
 	/* add attributes here, update the policy in nl80211.c */
 
 	__NL80211_ATTR_AFTER_LAST,
@@ -280,4 +285,63 @@ enum nl80211_sta_stats {
 	NL80211_STA_STAT_MAX = __NL80211_STA_STAT_AFTER_LAST - 1
 };
 
+/**
+ * enum nl80211_band_attr - band attributes
+ * @__NL80211_BAND_ATTR_INVALID: attribute number 0 is reserved
+ * @NL80211_BAND_ATTR_FREQS: supported frequencies in this band,
+ *	an array of nested frequency attributes
+ * @NL80211_BAND_ATTR_RATES: supported bitrates in this band,
+ *	an array of nested bitrate attributes
+ */
+enum nl80211_band_attr {
+	__NL80211_BAND_ATTR_INVALID,
+	NL80211_BAND_ATTR_FREQS,
+	NL80211_BAND_ATTR_RATES,
+
+	/* keep last */
+	__NL80211_BAND_ATTR_AFTER_LAST,
+	NL80211_BAND_ATTR_MAX = __NL80211_BAND_ATTR_AFTER_LAST - 1
+};
+
+/**
+ * enum nl80211_frequency_attr - frequency attributes
+ * @NL80211_FREQUENCY_ATTR_FREQ: Frequency in MHz
+ * @NL80211_FREQUENCY_ATTR_DISABLED: Channel is disabled in current
+ *	regulatory domain.
+ * @NL80211_FREQUENCY_ATTR_PASSIVE_SCAN: Only passive scanning is
+ *	permitted on this channel in current regulatory domain.
+ * @NL80211_FREQUENCY_ATTR_NO_IBSS: IBSS networks are not permitted
+ *	on this channel in current regulatory domain.
+ * @NL80211_FREQUENCY_ATTR_RADAR: Radar detection is mandatory
+ *	on this channel in current regulatory domain.
+ */
+enum nl80211_frequency_attr {
+	__NL80211_FREQUENCY_ATTR_INVALID,
+	NL80211_FREQUENCY_ATTR_FREQ,
+	NL80211_FREQUENCY_ATTR_DISABLED,
+	NL80211_FREQUENCY_ATTR_PASSIVE_SCAN,
+	NL80211_FREQUENCY_ATTR_NO_IBSS,
+	NL80211_FREQUENCY_ATTR_RADAR,
+
+	/* keep last */
+	__NL80211_FREQUENCY_ATTR_AFTER_LAST,
+	NL80211_FREQUENCY_ATTR_MAX = __NL80211_FREQUENCY_ATTR_AFTER_LAST - 1
+};
+
+/**
+ * enum nl80211_bitrate_attr - bitrate attributes
+ * @NL80211_BITRATE_ATTR_RATE: Bitrate in units of 100 kbps
+ * @NL80211_BITRATE_ATTR_2GHZ_SHORTPREAMBLE: Short preamble supported
+ *	in 2.4 GHz band.
+ */
+enum nl80211_bitrate_attr {
+	__NL80211_BITRATE_ATTR_INVALID,
+	NL80211_BITRATE_ATTR_RATE,
+	NL80211_BITRATE_ATTR_2GHZ_SHORTPREAMBLE,
+
+	/* keep last */
+	__NL80211_BITRATE_ATTR_AFTER_LAST,
+	NL80211_BITRATE_ATTR_MAX = __NL80211_BITRATE_ATTR_AFTER_LAST - 1
+};
+
 #endif /* __LINUX_NL80211_H */
diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c
index e3a214f63f91..b123f58d3909 100644
--- a/net/wireless/nl80211.c
+++ b/net/wireless/nl80211.c
@@ -98,6 +98,13 @@ static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags,
 			      struct cfg80211_registered_device *dev)
 {
 	void *hdr;
+	struct nlattr *nl_bands, *nl_band;
+	struct nlattr *nl_freqs, *nl_freq;
+	struct nlattr *nl_rates, *nl_rate;
+	enum ieee80211_band band;
+	struct ieee80211_channel *chan;
+	struct ieee80211_rate *rate;
+	int i;
 
 	hdr = nl80211hdr_put(msg, pid, seq, flags, NL80211_CMD_NEW_WIPHY);
 	if (!hdr)
@@ -105,6 +112,73 @@ static int nl80211_send_wiphy(struct sk_buff *msg, u32 pid, u32 seq, int flags,
 
 	NLA_PUT_U32(msg, NL80211_ATTR_WIPHY, dev->idx);
 	NLA_PUT_STRING(msg, NL80211_ATTR_WIPHY_NAME, wiphy_name(&dev->wiphy));
+
+	nl_bands = nla_nest_start(msg, NL80211_ATTR_WIPHY_BANDS);
+	if (!nl_bands)
+		goto nla_put_failure;
+
+	for (band = 0; band < IEEE80211_NUM_BANDS; band++) {
+		if (!dev->wiphy.bands[band])
+			continue;
+
+		nl_band = nla_nest_start(msg, band);
+		if (!nl_band)
+			goto nla_put_failure;
+
+		/* add frequencies */
+		nl_freqs = nla_nest_start(msg, NL80211_BAND_ATTR_FREQS);
+		if (!nl_freqs)
+			goto nla_put_failure;
+
+		for (i = 0; i < dev->wiphy.bands[band]->n_channels; i++) {
+			nl_freq = nla_nest_start(msg, i);
+			if (!nl_freq)
+				goto nla_put_failure;
+
+			chan = &dev->wiphy.bands[band]->channels[i];
+			NLA_PUT_U32(msg, NL80211_FREQUENCY_ATTR_FREQ,
+				    chan->center_freq);
+
+			if (chan->flags & IEEE80211_CHAN_DISABLED)
+				NLA_PUT_FLAG(msg, NL80211_FREQUENCY_ATTR_DISABLED);
+			if (chan->flags & IEEE80211_CHAN_PASSIVE_SCAN)
+				NLA_PUT_FLAG(msg, NL80211_FREQUENCY_ATTR_PASSIVE_SCAN);
+			if (chan->flags & IEEE80211_CHAN_NO_IBSS)
+				NLA_PUT_FLAG(msg, NL80211_FREQUENCY_ATTR_NO_IBSS);
+			if (chan->flags & IEEE80211_CHAN_RADAR)
+				NLA_PUT_FLAG(msg, NL80211_FREQUENCY_ATTR_RADAR);
+
+			nla_nest_end(msg, nl_freq);
+		}
+
+		nla_nest_end(msg, nl_freqs);
+
+		/* add bitrates */
+		nl_rates = nla_nest_start(msg, NL80211_BAND_ATTR_RATES);
+		if (!nl_rates)
+			goto nla_put_failure;
+
+		for (i = 0; i < dev->wiphy.bands[band]->n_bitrates; i++) {
+			nl_rate = nla_nest_start(msg, i);
+			if (!nl_rate)
+				goto nla_put_failure;
+
+			rate = &dev->wiphy.bands[band]->bitrates[i];
+			NLA_PUT_U32(msg, NL80211_BITRATE_ATTR_RATE,
+				    rate->bitrate);
+			if (rate->flags & IEEE80211_RATE_SHORT_PREAMBLE)
+				NLA_PUT_FLAG(msg,
+					NL80211_BITRATE_ATTR_2GHZ_SHORTPREAMBLE);
+
+			nla_nest_end(msg, nl_rate);
+		}
+
+		nla_nest_end(msg, nl_rates);
+
+		nla_nest_end(msg, nl_band);
+	}
+	nla_nest_end(msg, nl_bands);
+
 	return genlmsg_end(msg, hdr);
 
  nla_put_failure:
-- 
cgit v1.2.3


From 66f7ac50ed7cc5c19a62bc97e8f6e7891004a03a Mon Sep 17 00:00:00 2001
From: Michael Wu <flamingice@sourmilk.net>
Date: Thu, 31 Jan 2008 19:48:22 +0100
Subject: nl80211: Add monitor interface configuration flags

This allows precise control over what a monitor interface shows.

Signed-off-by: Michael Wu <flamingice@sourmilk.net>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/linux/nl80211.h | 35 +++++++++++++++++++++++++++++++++++
 include/net/cfg80211.h  | 24 ++++++++++++++++++++++--
 net/mac80211/cfg.c      |  4 ++--
 net/wireless/nl80211.c  | 44 ++++++++++++++++++++++++++++++++++++++++++--
 4 files changed, 101 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h
index 63695060db9f..a9f0b93324a2 100644
--- a/include/linux/nl80211.h
+++ b/include/linux/nl80211.h
@@ -164,6 +164,9 @@ enum nl80211_commands {
  * @NL80211_ATTR_WIPHY_BANDS: Information about an operating bands,
  *	consisting of a nested array.
  *
+ * @NL80211_ATTR_MNTR_FLAGS: flags, nested element with NLA_FLAG attributes of
+ *      &enum nl80211_mntr_flags.
+ *
  * @NL80211_ATTR_MAX: highest attribute number currently defined
  * @__NL80211_ATTR_AFTER_LAST: internal use
  */
@@ -200,6 +203,8 @@ enum nl80211_attrs {
 
 	NL80211_ATTR_WIPHY_BANDS,
 
+	NL80211_ATTR_MNTR_FLAGS,
+
 	/* add attributes here, update the policy in nl80211.c */
 
 	__NL80211_ATTR_AFTER_LAST,
@@ -344,4 +349,34 @@ enum nl80211_bitrate_attr {
 	NL80211_BITRATE_ATTR_MAX = __NL80211_BITRATE_ATTR_AFTER_LAST - 1
 };
 
+/**
+ * enum nl80211_mntr_flags - monitor configuration flags
+ *
+ * Monitor configuration flags.
+ *
+ * @__NL80211_MNTR_FLAG_INVALID: reserved
+ *
+ * @NL80211_MNTR_FLAG_FCSFAIL: pass frames with bad FCS
+ * @NL80211_MNTR_FLAG_PLCPFAIL: pass frames with bad PLCP
+ * @NL80211_MNTR_FLAG_CONTROL: pass control frames
+ * @NL80211_MNTR_FLAG_OTHER_BSS: disable BSSID filtering
+ * @NL80211_MNTR_FLAG_COOK_FRAMES: report frames after processing.
+ *	overrides all other flags.
+ *
+ * @__NL80211_MNTR_FLAG_AFTER_LAST: internal use
+ * @NL80211_MNTR_FLAG_MAX: highest possible monitor flag
+ */
+enum nl80211_mntr_flags {
+	__NL80211_MNTR_FLAG_INVALID,
+	NL80211_MNTR_FLAG_FCSFAIL,
+	NL80211_MNTR_FLAG_PLCPFAIL,
+	NL80211_MNTR_FLAG_CONTROL,
+	NL80211_MNTR_FLAG_OTHER_BSS,
+	NL80211_MNTR_FLAG_COOK_FRAMES,
+
+	/* keep last */
+	__NL80211_MNTR_FLAG_AFTER_LAST,
+	NL80211_MNTR_FLAG_MAX = __NL80211_MNTR_FLAG_AFTER_LAST - 1
+};
+
 #endif /* __LINUX_NL80211_H */
diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h
index bcc480b8892a..ab4caf63954f 100644
--- a/include/net/cfg80211.h
+++ b/include/net/cfg80211.h
@@ -163,6 +163,26 @@ struct station_stats {
 	u32 tx_bytes;
 };
 
+/**
+ * enum monitor_flags - monitor flags
+ *
+ * Monitor interface configuration flags. Note that these must be the bits
+ * according to the nl80211 flags.
+ *
+ * @MONITOR_FLAG_FCSFAIL: pass frames with bad FCS
+ * @MONITOR_FLAG_PLCPFAIL: pass frames with bad PLCP
+ * @MONITOR_FLAG_CONTROL: pass control frames
+ * @MONITOR_FLAG_OTHER_BSS: disable BSSID filtering
+ * @MONITOR_FLAG_COOK_FRAMES: report frames after processing
+ */
+enum monitor_flags {
+	MONITOR_FLAG_FCSFAIL		= 1<<NL80211_MNTR_FLAG_FCSFAIL,
+	MONITOR_FLAG_PLCPFAIL		= 1<<NL80211_MNTR_FLAG_PLCPFAIL,
+	MONITOR_FLAG_CONTROL		= 1<<NL80211_MNTR_FLAG_CONTROL,
+	MONITOR_FLAG_OTHER_BSS		= 1<<NL80211_MNTR_FLAG_OTHER_BSS,
+	MONITOR_FLAG_COOK_FRAMES	= 1<<NL80211_MNTR_FLAG_COOK_FRAMES,
+};
+
 /* from net/wireless.h */
 struct wiphy;
 
@@ -213,10 +233,10 @@ struct wiphy;
  */
 struct cfg80211_ops {
 	int	(*add_virtual_intf)(struct wiphy *wiphy, char *name,
-				    enum nl80211_iftype type);
+				    enum nl80211_iftype type, u32 *flags);
 	int	(*del_virtual_intf)(struct wiphy *wiphy, int ifindex);
 	int	(*change_virtual_intf)(struct wiphy *wiphy, int ifindex,
-				       enum nl80211_iftype type);
+				       enum nl80211_iftype type, u32 *flags);
 
 	int	(*add_key)(struct wiphy *wiphy, struct net_device *netdev,
 			   u8 key_index, u8 *mac_addr,
diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c
index 15b8cf94f510..706418d477c2 100644
--- a/net/mac80211/cfg.c
+++ b/net/mac80211/cfg.c
@@ -34,7 +34,7 @@ nl80211_type_to_mac80211_type(enum nl80211_iftype type)
 }
 
 static int ieee80211_add_iface(struct wiphy *wiphy, char *name,
-			       enum nl80211_iftype type)
+			       enum nl80211_iftype type, u32 *flags)
 {
 	struct ieee80211_local *local = wiphy_priv(wiphy);
 	enum ieee80211_if_types itype;
@@ -69,7 +69,7 @@ static int ieee80211_del_iface(struct wiphy *wiphy, int ifindex)
 }
 
 static int ieee80211_change_iface(struct wiphy *wiphy, int ifindex,
-				  enum nl80211_iftype type)
+				  enum nl80211_iftype type, u32 *flags)
 {
 	struct ieee80211_local *local = wiphy_priv(wiphy);
 	struct net_device *dev;
diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c
index b123f58d3909..5b3474798b8d 100644
--- a/net/wireless/nl80211.c
+++ b/net/wireless/nl80211.c
@@ -82,6 +82,7 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = {
 	[NL80211_ATTR_STA_SUPPORTED_RATES] = { .type = NLA_BINARY,
 					       .len = NL80211_MAX_SUPP_RATES },
 	[NL80211_ATTR_STA_VLAN] = { .type = NLA_U32 },
+	[NL80211_ATTR_MNTR_FLAGS] = { .type = NLA_NESTED },
 };
 
 /* message building helper */
@@ -336,12 +337,42 @@ static int nl80211_get_interface(struct sk_buff *skb, struct genl_info *info)
 	return -ENOBUFS;
 }
 
+static const struct nla_policy mntr_flags_policy[NL80211_MNTR_FLAG_MAX + 1] = {
+	[NL80211_MNTR_FLAG_FCSFAIL] = { .type = NLA_FLAG },
+	[NL80211_MNTR_FLAG_PLCPFAIL] = { .type = NLA_FLAG },
+	[NL80211_MNTR_FLAG_CONTROL] = { .type = NLA_FLAG },
+	[NL80211_MNTR_FLAG_OTHER_BSS] = { .type = NLA_FLAG },
+	[NL80211_MNTR_FLAG_COOK_FRAMES] = { .type = NLA_FLAG },
+};
+
+static int parse_monitor_flags(struct nlattr *nla, u32 *mntrflags)
+{
+	struct nlattr *flags[NL80211_MNTR_FLAG_MAX + 1];
+	int flag;
+
+	*mntrflags = 0;
+
+	if (!nla)
+		return -EINVAL;
+
+	if (nla_parse_nested(flags, NL80211_MNTR_FLAG_MAX,
+			     nla, mntr_flags_policy))
+		return -EINVAL;
+
+	for (flag = 1; flag <= NL80211_MNTR_FLAG_MAX; flag++)
+		if (flags[flag])
+			*mntrflags |= (1<<flag);
+
+	return 0;
+}
+
 static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info)
 {
 	struct cfg80211_registered_device *drv;
 	int err, ifindex;
 	enum nl80211_iftype type;
 	struct net_device *dev;
+	u32 flags;
 
 	if (info->attrs[NL80211_ATTR_IFTYPE]) {
 		type = nla_get_u32(info->attrs[NL80211_ATTR_IFTYPE]);
@@ -362,7 +393,11 @@ static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info)
 	}
 
 	rtnl_lock();
-	err = drv->ops->change_virtual_intf(&drv->wiphy, ifindex, type);
+	err = parse_monitor_flags(type == NL80211_IFTYPE_MONITOR ?
+				  info->attrs[NL80211_ATTR_MNTR_FLAGS] : NULL,
+				  &flags);
+	err = drv->ops->change_virtual_intf(&drv->wiphy, ifindex,
+					    type, err ? NULL : &flags);
 	rtnl_unlock();
 
  unlock:
@@ -375,6 +410,7 @@ static int nl80211_new_interface(struct sk_buff *skb, struct genl_info *info)
 	struct cfg80211_registered_device *drv;
 	int err;
 	enum nl80211_iftype type = NL80211_IFTYPE_UNSPECIFIED;
+	u32 flags;
 
 	if (!info->attrs[NL80211_ATTR_IFNAME])
 		return -EINVAL;
@@ -395,8 +431,12 @@ static int nl80211_new_interface(struct sk_buff *skb, struct genl_info *info)
 	}
 
 	rtnl_lock();
+	err = parse_monitor_flags(type == NL80211_IFTYPE_MONITOR ?
+				  info->attrs[NL80211_ATTR_MNTR_FLAGS] : NULL,
+				  &flags);
 	err = drv->ops->add_virtual_intf(&drv->wiphy,
-		nla_data(info->attrs[NL80211_ATTR_IFNAME]), type);
+		nla_data(info->attrs[NL80211_ATTR_IFNAME]),
+		type, err ? NULL : &flags);
 	rtnl_unlock();
 
  unlock:
-- 
cgit v1.2.3


From ffc7689ddae5cbe12bde437ae0f2b386d568b5cd Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Wed, 20 Feb 2008 19:08:10 +0100
Subject: ssb: Add support for 8bit register access

This adds support for 8bit wide register reads/writes.
This is needed in order to support the gigabit ethernet core.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/main.c      | 18 ++++++++++++++++++
 drivers/ssb/pci.c       | 28 ++++++++++++++++++++++++++++
 drivers/ssb/pcmcia.c    | 32 ++++++++++++++++++++++++++++++++
 include/linux/ssb/ssb.h | 10 ++++++++++
 4 files changed, 88 insertions(+)

(limited to 'include/linux')

diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index bedb2b4ee9d2..8db40c4b86e9 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -505,6 +505,14 @@ error:
 	return err;
 }
 
+static u8 ssb_ssb_read8(struct ssb_device *dev, u16 offset)
+{
+	struct ssb_bus *bus = dev->bus;
+
+	offset += dev->core_index * SSB_CORE_SIZE;
+	return readb(bus->mmio + offset);
+}
+
 static u16 ssb_ssb_read16(struct ssb_device *dev, u16 offset)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -521,6 +529,14 @@ static u32 ssb_ssb_read32(struct ssb_device *dev, u16 offset)
 	return readl(bus->mmio + offset);
 }
 
+static void ssb_ssb_write8(struct ssb_device *dev, u16 offset, u8 value)
+{
+	struct ssb_bus *bus = dev->bus;
+
+	offset += dev->core_index * SSB_CORE_SIZE;
+	writeb(value, bus->mmio + offset);
+}
+
 static void ssb_ssb_write16(struct ssb_device *dev, u16 offset, u16 value)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -539,8 +555,10 @@ static void ssb_ssb_write32(struct ssb_device *dev, u16 offset, u32 value)
 
 /* Ops for the plain SSB bus without a host-device (no PCI or PCMCIA). */
 static const struct ssb_bus_ops ssb_ssb_ops = {
+	.read8		= ssb_ssb_read8,
 	.read16		= ssb_ssb_read16,
 	.read32		= ssb_ssb_read32,
+	.write8		= ssb_ssb_write8,
 	.write16	= ssb_ssb_write16,
 	.write32	= ssb_ssb_write32,
 };
diff --git a/drivers/ssb/pci.c b/drivers/ssb/pci.c
index b434df75047f..1facc7620fc8 100644
--- a/drivers/ssb/pci.c
+++ b/drivers/ssb/pci.c
@@ -572,6 +572,19 @@ static inline int ssb_pci_assert_buspower(struct ssb_bus *bus)
 }
 #endif /* DEBUG */
 
+static u8 ssb_pci_read8(struct ssb_device *dev, u16 offset)
+{
+	struct ssb_bus *bus = dev->bus;
+
+	if (unlikely(ssb_pci_assert_buspower(bus)))
+		return 0xFF;
+	if (unlikely(bus->mapped_device != dev)) {
+		if (unlikely(ssb_pci_switch_core(bus, dev)))
+			return 0xFF;
+	}
+	return ioread8(bus->mmio + offset);
+}
+
 static u16 ssb_pci_read16(struct ssb_device *dev, u16 offset)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -598,6 +611,19 @@ static u32 ssb_pci_read32(struct ssb_device *dev, u16 offset)
 	return ioread32(bus->mmio + offset);
 }
 
+static void ssb_pci_write8(struct ssb_device *dev, u16 offset, u8 value)
+{
+	struct ssb_bus *bus = dev->bus;
+
+	if (unlikely(ssb_pci_assert_buspower(bus)))
+		return;
+	if (unlikely(bus->mapped_device != dev)) {
+		if (unlikely(ssb_pci_switch_core(bus, dev)))
+			return;
+	}
+	iowrite8(value, bus->mmio + offset);
+}
+
 static void ssb_pci_write16(struct ssb_device *dev, u16 offset, u16 value)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -626,8 +652,10 @@ static void ssb_pci_write32(struct ssb_device *dev, u16 offset, u32 value)
 
 /* Not "static", as it's used in main.c */
 const struct ssb_bus_ops ssb_pci_ops = {
+	.read8		= ssb_pci_read8,
 	.read16		= ssb_pci_read16,
 	.read32		= ssb_pci_read32,
+	.write8		= ssb_pci_write8,
 	.write16	= ssb_pci_write16,
 	.write32	= ssb_pci_write32,
 };
diff --git a/drivers/ssb/pcmcia.c b/drivers/ssb/pcmcia.c
index 46816cda8b98..84b3a845a8a8 100644
--- a/drivers/ssb/pcmcia.c
+++ b/drivers/ssb/pcmcia.c
@@ -172,6 +172,22 @@ static int select_core_and_segment(struct ssb_device *dev,
 	return 0;
 }
 
+static u8 ssb_pcmcia_read8(struct ssb_device *dev, u16 offset)
+{
+	struct ssb_bus *bus = dev->bus;
+	unsigned long flags;
+	int err;
+	u8 value = 0xFF;
+
+	spin_lock_irqsave(&bus->bar_lock, flags);
+	err = select_core_and_segment(dev, &offset);
+	if (likely(!err))
+		value = readb(bus->mmio + offset);
+	spin_unlock_irqrestore(&bus->bar_lock, flags);
+
+	return value;
+}
+
 static u16 ssb_pcmcia_read16(struct ssb_device *dev, u16 offset)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -206,6 +222,20 @@ static u32 ssb_pcmcia_read32(struct ssb_device *dev, u16 offset)
 	return (lo | (hi << 16));
 }
 
+static void ssb_pcmcia_write8(struct ssb_device *dev, u16 offset, u8 value)
+{
+	struct ssb_bus *bus = dev->bus;
+	unsigned long flags;
+	int err;
+
+	spin_lock_irqsave(&bus->bar_lock, flags);
+	err = select_core_and_segment(dev, &offset);
+	if (likely(!err))
+		writeb(value, bus->mmio + offset);
+	mmiowb();
+	spin_unlock_irqrestore(&bus->bar_lock, flags);
+}
+
 static void ssb_pcmcia_write16(struct ssb_device *dev, u16 offset, u16 value)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -238,8 +268,10 @@ static void ssb_pcmcia_write32(struct ssb_device *dev, u16 offset, u32 value)
 
 /* Not "static", as it's used in main.c */
 const struct ssb_bus_ops ssb_pcmcia_ops = {
+	.read8		= ssb_pcmcia_read8,
 	.read16		= ssb_pcmcia_read16,
 	.read32		= ssb_pcmcia_read32,
+	.write8		= ssb_pcmcia_write8,
 	.write16	= ssb_pcmcia_write16,
 	.write32	= ssb_pcmcia_write32,
 };
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index 20add65215af..860d28c6d149 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -72,8 +72,10 @@ struct ssb_device;
 /* Lowlevel read/write operations on the device MMIO.
  * Internal, don't use that outside of ssb. */
 struct ssb_bus_ops {
+	u8 (*read8)(struct ssb_device *dev, u16 offset);
 	u16 (*read16)(struct ssb_device *dev, u16 offset);
 	u32 (*read32)(struct ssb_device *dev, u16 offset);
+	void (*write8)(struct ssb_device *dev, u16 offset, u8 value);
 	void (*write16)(struct ssb_device *dev, u16 offset, u16 value);
 	void (*write32)(struct ssb_device *dev, u16 offset, u32 value);
 };
@@ -344,6 +346,10 @@ void ssb_device_disable(struct ssb_device *dev, u32 core_specific_flags);
 
 
 /* Device MMIO register read/write functions. */
+static inline u8 ssb_read8(struct ssb_device *dev, u16 offset)
+{
+	return dev->ops->read8(dev, offset);
+}
 static inline u16 ssb_read16(struct ssb_device *dev, u16 offset)
 {
 	return dev->ops->read16(dev, offset);
@@ -352,6 +358,10 @@ static inline u32 ssb_read32(struct ssb_device *dev, u16 offset)
 {
 	return dev->ops->read32(dev, offset);
 }
+static inline void ssb_write8(struct ssb_device *dev, u16 offset, u8 value)
+{
+	dev->ops->write8(dev, offset, value);
+}
 static inline void ssb_write16(struct ssb_device *dev, u16 offset, u16 value)
 {
 	dev->ops->write16(dev, offset, value);
-- 
cgit v1.2.3


From 28de57d1a9eb7e67badb731297197fcbef0cc19e Mon Sep 17 00:00:00 2001
From: Aurelien Jarno <aurelien@aurel32.net>
Date: Fri, 22 Feb 2008 16:14:58 +0100
Subject: ssb: Add CHIPCO IRQ access functions

This patch adds functions to setup and read the CHIPCO IRQ.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/driver_chipcommon.c           | 10 ++++++++++
 include/linux/ssb/ssb_driver_chipcommon.h |  4 ++++
 2 files changed, 14 insertions(+)

(limited to 'include/linux')

diff --git a/drivers/ssb/driver_chipcommon.c b/drivers/ssb/driver_chipcommon.c
index e586321a473a..45b672a69003 100644
--- a/drivers/ssb/driver_chipcommon.c
+++ b/drivers/ssb/driver_chipcommon.c
@@ -353,6 +353,16 @@ void ssb_chipco_watchdog_timer_set(struct ssb_chipcommon *cc, u32 ticks)
 	chipco_write32(cc, SSB_CHIPCO_WATCHDOG, ticks);
 }
 
+void ssb_chipco_irq_mask(struct ssb_chipcommon *cc, u32 mask, u32 value)
+{
+	chipco_write32_masked(cc, SSB_CHIPCO_IRQMASK, mask, value);
+}
+
+u32 ssb_chipco_irq_status(struct ssb_chipcommon *cc, u32 mask)
+{
+	return chipco_read32(cc, SSB_CHIPCO_IRQSTAT) & mask;
+}
+
 u32 ssb_chipco_gpio_in(struct ssb_chipcommon *cc, u32 mask)
 {
 	return chipco_read32(cc, SSB_CHIPCO_GPIOIN) & mask;
diff --git a/include/linux/ssb/ssb_driver_chipcommon.h b/include/linux/ssb/ssb_driver_chipcommon.h
index 536851b946f6..b548a54ff1f5 100644
--- a/include/linux/ssb/ssb_driver_chipcommon.h
+++ b/include/linux/ssb/ssb_driver_chipcommon.h
@@ -390,6 +390,10 @@ extern void ssb_chipco_set_clockmode(struct ssb_chipcommon *cc,
 extern void ssb_chipco_watchdog_timer_set(struct ssb_chipcommon *cc,
 					  u32 ticks);
 
+void ssb_chipco_irq_mask(struct ssb_chipcommon *cc, u32 mask, u32 value);
+
+u32 ssb_chipco_irq_status(struct ssb_chipcommon *cc, u32 mask);
+
 /* Chipcommon GPIO pin access. */
 u32 ssb_chipco_gpio_in(struct ssb_chipcommon *cc, u32 mask);
 u32 ssb_chipco_gpio_out(struct ssb_chipcommon *cc, u32 mask, u32 value);
-- 
cgit v1.2.3


From 674eea0fc4d1d693250b5d3ddad42ca931c87dfd Mon Sep 17 00:00:00 2001
From: Avi Kivity <avi@qumranet.com>
Date: Mon, 11 Feb 2008 18:37:23 +0200
Subject: KVM: Make the supported cpuid list a host property rather than a vm
 property

One of the use cases for the supported cpuid list is to create a "greatest
common denominator" of cpu capabilities in a server farm.  As such, it is
useful to be able to get the list without creating a virtual machine first.

Since the code does not depend on the vm in any way, all that is needed is
to move it to the device ioctl handler.  The capability identifier is also
changed so that binaries made against -rc1 will fail gracefully.

Signed-off-by: Avi Kivity <avi@qumranet.com>
---
 arch/x86/kvm/x86.c  | 42 ++++++++++++++++++++++--------------------
 include/linux/kvm.h |  4 ++--
 2 files changed, 24 insertions(+), 22 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index ec60409299a3..a7069ec2267c 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -46,6 +46,9 @@
 #define VM_STAT(x) offsetof(struct kvm, stat.x), KVM_STAT_VM
 #define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU
 
+static int kvm_dev_ioctl_get_supported_cpuid(struct kvm_cpuid2 *cpuid,
+				    struct kvm_cpuid_entry2 __user *entries);
+
 struct kvm_x86_ops *kvm_x86_ops;
 
 struct kvm_stats_debugfs_item debugfs_entries[] = {
@@ -727,6 +730,24 @@ long kvm_arch_dev_ioctl(struct file *filp,
 		r = 0;
 		break;
 	}
+	case KVM_GET_SUPPORTED_CPUID: {
+		struct kvm_cpuid2 __user *cpuid_arg = argp;
+		struct kvm_cpuid2 cpuid;
+
+		r = -EFAULT;
+		if (copy_from_user(&cpuid, cpuid_arg, sizeof cpuid))
+			goto out;
+		r = kvm_dev_ioctl_get_supported_cpuid(&cpuid,
+			cpuid_arg->entries);
+		if (r)
+			goto out;
+
+		r = -EFAULT;
+		if (copy_to_user(cpuid_arg, &cpuid, sizeof cpuid))
+			goto out;
+		r = 0;
+		break;
+	}
 	default:
 		r = -EINVAL;
 	}
@@ -974,8 +995,7 @@ static void do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 	put_cpu();
 }
 
-static int kvm_vm_ioctl_get_supported_cpuid(struct kvm *kvm,
-				    struct kvm_cpuid2 *cpuid,
+static int kvm_dev_ioctl_get_supported_cpuid(struct kvm_cpuid2 *cpuid,
 				    struct kvm_cpuid_entry2 __user *entries)
 {
 	struct kvm_cpuid_entry2 *cpuid_entries;
@@ -1487,24 +1507,6 @@ long kvm_arch_vm_ioctl(struct file *filp,
 		r = 0;
 		break;
 	}
-	case KVM_GET_SUPPORTED_CPUID: {
-		struct kvm_cpuid2 __user *cpuid_arg = argp;
-		struct kvm_cpuid2 cpuid;
-
-		r = -EFAULT;
-		if (copy_from_user(&cpuid, cpuid_arg, sizeof cpuid))
-			goto out;
-		r = kvm_vm_ioctl_get_supported_cpuid(kvm, &cpuid,
-			cpuid_arg->entries);
-		if (r)
-			goto out;
-
-		r = -EFAULT;
-		if (copy_to_user(cpuid_arg, &cpuid, sizeof cpuid))
-			goto out;
-		r = 0;
-		break;
-	}
 	default:
 		;
 	}
diff --git a/include/linux/kvm.h b/include/linux/kvm.h
index 4de4fd2d8607..c1ec04fd000d 100644
--- a/include/linux/kvm.h
+++ b/include/linux/kvm.h
@@ -221,6 +221,7 @@ struct kvm_vapic_addr {
  * Get size for mmap(vcpu_fd)
  */
 #define KVM_GET_VCPU_MMAP_SIZE    _IO(KVMIO,   0x04) /* in bytes */
+#define KVM_GET_SUPPORTED_CPUID   _IOWR(KVMIO, 0x05, struct kvm_cpuid2)
 
 /*
  * Extension capability list.
@@ -230,8 +231,8 @@ struct kvm_vapic_addr {
 #define KVM_CAP_MMU_SHADOW_CACHE_CONTROL 2
 #define KVM_CAP_USER_MEMORY 3
 #define KVM_CAP_SET_TSS_ADDR 4
-#define KVM_CAP_EXT_CPUID 5
 #define KVM_CAP_VAPIC 6
+#define KVM_CAP_EXT_CPUID 7
 
 /*
  * ioctls for VM fds
@@ -249,7 +250,6 @@ struct kvm_vapic_addr {
 #define KVM_CREATE_VCPU           _IO(KVMIO,  0x41)
 #define KVM_GET_DIRTY_LOG         _IOW(KVMIO, 0x42, struct kvm_dirty_log)
 #define KVM_SET_MEMORY_ALIAS      _IOW(KVMIO, 0x43, struct kvm_memory_alias)
-#define KVM_GET_SUPPORTED_CPUID   _IOWR(KVMIO, 0x48, struct kvm_cpuid2)
 /* Device model IOC */
 #define KVM_CREATE_IRQCHIP	  _IO(KVMIO,  0x60)
 #define KVM_IRQ_LINE		  _IOW(KVMIO, 0x61, struct kvm_irq_level)
-- 
cgit v1.2.3


From d0bcabcd72dda5f553322a1ca92ae31c15b408b6 Mon Sep 17 00:00:00 2001
From: Randy Dunlap <randy.dunlap@oracle.com>
Date: Fri, 29 Feb 2008 22:03:07 -0800
Subject: docbook: fix usb source files

Fix docbook problems in USB source files.
These cause the generated docbook to be incorrect.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/usb/core/usb.c | 6 ++----
 include/linux/usb.h    | 9 +++------
 2 files changed, 5 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/usb/core/usb.c b/drivers/usb/core/usb.c
index 4e984060c984..f6f19908f5f0 100644
--- a/drivers/usb/core/usb.c
+++ b/drivers/usb/core/usb.c
@@ -99,8 +99,7 @@ struct usb_interface *usb_ifnum_to_if(const struct usb_device *dev,
 EXPORT_SYMBOL_GPL(usb_ifnum_to_if);
 
 /**
- * usb_altnum_to_altsetting - get the altsetting structure with a given
- *	alternate setting number.
+ * usb_altnum_to_altsetting - get the altsetting structure with a given alternate setting number.
  * @intf: the interface containing the altsetting in question
  * @altnum: the desired alternate setting number
  *
@@ -442,8 +441,7 @@ EXPORT_SYMBOL_GPL(usb_put_intf);
  */
 
 /**
- * usb_lock_device_for_reset - cautiously acquire the lock for a
- *	usb device structure
+ * usb_lock_device_for_reset - cautiously acquire the lock for a usb device structure
  * @udev: device that's being locked
  * @iface: interface bound to the driver making the request (optional)
  *
diff --git a/include/linux/usb.h b/include/linux/usb.h
index 2372e2e6b527..5bd3ae8aaaf4 100644
--- a/include/linux/usb.h
+++ b/include/linux/usb.h
@@ -781,8 +781,7 @@ static inline int usb_endpoint_is_isoc_out(
 	.idVendor = (vend), \
 	.idProduct = (prod)
 /**
- * USB_DEVICE_VER - macro used to describe a specific usb device with a
- *		version range
+ * USB_DEVICE_VER - describe a specific usb device with a version range
  * @vend: the 16 bit USB Vendor ID
  * @prod: the 16 bit USB Product ID
  * @lo: the bcdDevice_lo value
@@ -799,8 +798,7 @@ static inline int usb_endpoint_is_isoc_out(
 	.bcdDevice_hi = (hi)
 
 /**
- * USB_DEVICE_INTERFACE_PROTOCOL - macro used to describe a usb
- *		device with a specific interface protocol
+ * USB_DEVICE_INTERFACE_PROTOCOL - describe a usb device with a specific interface protocol
  * @vend: the 16 bit USB Vendor ID
  * @prod: the 16 bit USB Product ID
  * @pr: bInterfaceProtocol value
@@ -846,8 +844,7 @@ static inline int usb_endpoint_is_isoc_out(
 	.bInterfaceProtocol = (pr)
 
 /**
- * USB_DEVICE_AND_INTERFACE_INFO - macro used to describe a specific usb device
- * 		with a class of usb interfaces
+ * USB_DEVICE_AND_INTERFACE_INFO - describe a specific usb device with a class of usb interfaces
  * @vend: the 16 bit USB Vendor ID
  * @prod: the 16 bit USB Product ID
  * @cl: bInterfaceClass value
-- 
cgit v1.2.3


From 988b705077d8f922408913f4f521ae073256d4a1 Mon Sep 17 00:00:00 2001
From: Pavel Emelyanov <xemul@openvz.org>
Date: Mon, 3 Mar 2008 12:20:57 -0800
Subject: [ARP]: Introduce the arp_hdr_len helper.

There are some place, that calculate the ARP header length. These
calculations are correct, but
 a) some operate with "magic" constants,
 b) enlarge the code length (sometimes at the cost of coding style),
 c) are not informative from the first glance.

The proposal is to introduce a helper, that includes all the good
sides of these calculations.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/bonding/bond_main.c | 5 +----
 include/linux/if_arp.h          | 6 ++++++
 net/core/netpoll.c              | 6 ++----
 net/ipv4/arp.c                  | 9 +++------
 net/ipv4/ipconfig.c             | 5 +----
 net/ipv4/netfilter/arp_tables.c | 5 +----
 6 files changed, 14 insertions(+), 22 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 966643473da7..5fc9d8d58ece 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -2646,10 +2646,7 @@ static int bond_arp_rcv(struct sk_buff *skb, struct net_device *dev, struct pack
 	if (!slave || !slave_do_arp_validate(bond, slave))
 		goto out_unlock;
 
-	/* ARP header, plus 2 device addresses, plus 2 IP addresses.  */
-	if (!pskb_may_pull(skb, (sizeof(struct arphdr) +
-				 (2 * dev->addr_len) +
-				 (2 * sizeof(u32)))))
+	if (!pskb_may_pull(skb, arp_hdr_len(dev)))
 		goto out_unlock;
 
 	arp = arp_hdr(skb);
diff --git a/include/linux/if_arp.h b/include/linux/if_arp.h
index 296e8e86e91d..4d3401812e6c 100644
--- a/include/linux/if_arp.h
+++ b/include/linux/if_arp.h
@@ -156,6 +156,12 @@ static inline struct arphdr *arp_hdr(const struct sk_buff *skb)
 {
 	return (struct arphdr *)skb_network_header(skb);
 }
+
+static inline int arp_hdr_len(struct net_device *dev)
+{
+	/* ARP header, plus 2 device addresses, plus 2 IP addresses. */
+	return sizeof(struct arphdr) + (dev->addr_len + sizeof(u32)) * 2;
+}
 #endif
 
 #endif	/* _LINUX_IF_ARP_H */
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index 6faa128a4c8e..7ae98659d79d 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -384,9 +384,7 @@ static void arp_reply(struct sk_buff *skb)
 	if (skb->dev->flags & IFF_NOARP)
 		return;
 
-	if (!pskb_may_pull(skb, (sizeof(struct arphdr) +
-				 (2 * skb->dev->addr_len) +
-				 (2 * sizeof(u32)))))
+	if (!pskb_may_pull(skb, arp_hdr_len(skb->dev)))
 		return;
 
 	skb_reset_network_header(skb);
@@ -414,7 +412,7 @@ static void arp_reply(struct sk_buff *skb)
 	    ipv4_is_loopback(tip) || ipv4_is_multicast(tip))
 		return;
 
-	size = sizeof(struct arphdr) + 2 * (skb->dev->addr_len + 4);
+	size = arp_hdr_len(skb->dev);
 	send_skb = find_skb(np, size + LL_RESERVED_SPACE(np->dev),
 			    LL_RESERVED_SPACE(np->dev));
 
diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c
index 8e17f65f4002..69e80bd9774a 100644
--- a/net/ipv4/arp.c
+++ b/net/ipv4/arp.c
@@ -570,14 +570,13 @@ struct sk_buff *arp_create(int type, int ptype, __be32 dest_ip,
 	 *	Allocate a buffer
 	 */
 
-	skb = alloc_skb(sizeof(struct arphdr)+ 2*(dev->addr_len+4)
-				+ LL_RESERVED_SPACE(dev), GFP_ATOMIC);
+	skb = alloc_skb(arp_hdr_len(dev) + LL_RESERVED_SPACE(dev), GFP_ATOMIC);
 	if (skb == NULL)
 		return NULL;
 
 	skb_reserve(skb, LL_RESERVED_SPACE(dev));
 	skb_reset_network_header(skb);
-	arp = (struct arphdr *) skb_put(skb,sizeof(struct arphdr) + 2*(dev->addr_len+4));
+	arp = (struct arphdr *) skb_put(skb, arp_hdr_len(dev));
 	skb->dev = dev;
 	skb->protocol = htons(ETH_P_ARP);
 	if (src_hw == NULL)
@@ -916,9 +915,7 @@ static int arp_rcv(struct sk_buff *skb, struct net_device *dev,
 		goto freeskb;
 
 	/* ARP header, plus 2 device addresses, plus 2 IP addresses.  */
-	if (!pskb_may_pull(skb, (sizeof(struct arphdr) +
-				 (2 * dev->addr_len) +
-				 (2 * sizeof(u32)))))
+	if (!pskb_may_pull(skb, arp_hdr_len(dev)))
 		goto freeskb;
 
 	arp = arp_hdr(skb);
diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c
index c90e75a66e81..f84041d1f623 100644
--- a/net/ipv4/ipconfig.c
+++ b/net/ipv4/ipconfig.c
@@ -459,10 +459,7 @@ ic_rarp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt
 	if (rarp->ar_pro != htons(ETH_P_IP))
 		goto drop;
 
-	if (!pskb_may_pull(skb,
-			   sizeof(struct arphdr) +
-			   (2 * dev->addr_len) +
-			   (2 * 4)))
+	if (!pskb_may_pull(skb, arp_hdr_len(dev)))
 		goto drop;
 
 	/* OK, it is all there and looks valid, process... */
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index a7591ce344d2..9b5904486184 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -233,10 +233,7 @@ unsigned int arpt_do_table(struct sk_buff *skb,
 	void *table_base;
 	struct xt_table_info *private;
 
-	/* ARP header, plus 2 device addresses, plus 2 IP addresses.  */
-	if (!pskb_may_pull(skb, (sizeof(struct arphdr) +
-				 (2 * skb->dev->addr_len) +
-				 (2 * sizeof(u32)))))
+	if (!pskb_may_pull(skb, arp_hdr_len(skb->dev)))
 		return NF_DROP;
 
 	indev = in ? in->name : nulldevname;
-- 
cgit v1.2.3


From a973e9dd1e140a65bed694a2c5c8d53e9cba1a23 Mon Sep 17 00:00:00 2001
From: Christoph Lameter <clameter@sgi.com>
Date: Sat, 1 Mar 2008 13:40:44 -0800
Subject: Revert "unique end pointer" patch

This only made sense for the alternate fastpath which was reverted last week.

Mathieu is working on a new version that addresses the fastpath issues but that
new code first needs to go through mm and it is not clear if we need the
unique end pointers with his new scheme.

Reviewed-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
---
 include/linux/mm_types.h |  5 +---
 mm/slub.c                | 70 ++++++++++++++++--------------------------------
 2 files changed, 24 insertions(+), 51 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
index bfee0bd1d435..34023c65d466 100644
--- a/include/linux/mm_types.h
+++ b/include/linux/mm_types.h
@@ -64,10 +64,7 @@ struct page {
 #if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS
 	    spinlock_t ptl;
 #endif
-	    struct {
-		   struct kmem_cache *slab;	/* SLUB: Pointer to slab */
-		   void *end;			/* SLUB: end marker */
-	    };
+	    struct kmem_cache *slab;	/* SLUB: Pointer to slab */
 	    struct page *first_page;	/* Compound tail pages */
 	};
 	union {
diff --git a/mm/slub.c b/mm/slub.c
index 74c65af0a54f..a873953e5a11 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -291,32 +291,15 @@ static inline struct kmem_cache_cpu *get_cpu_slab(struct kmem_cache *s, int cpu)
 #endif
 }
 
-/*
- * The end pointer in a slab is special. It points to the first object in the
- * slab but has bit 0 set to mark it.
- *
- * Note that SLUB relies on page_mapping returning NULL for pages with bit 0
- * in the mapping set.
- */
-static inline int is_end(void *addr)
-{
-	return (unsigned long)addr & PAGE_MAPPING_ANON;
-}
-
-static void *slab_address(struct page *page)
-{
-	return page->end - PAGE_MAPPING_ANON;
-}
-
 static inline int check_valid_pointer(struct kmem_cache *s,
 				struct page *page, const void *object)
 {
 	void *base;
 
-	if (object == page->end)
+	if (!object)
 		return 1;
 
-	base = slab_address(page);
+	base = page_address(page);
 	if (object < base || object >= base + s->objects * s->size ||
 		(object - base) % s->size) {
 		return 0;
@@ -349,8 +332,7 @@ static inline void set_freepointer(struct kmem_cache *s, void *object, void *fp)
 
 /* Scan freelist */
 #define for_each_free_object(__p, __s, __free) \
-	for (__p = (__free); (__p) != page->end; __p = get_freepointer((__s),\
-		__p))
+	for (__p = (__free); __p; __p = get_freepointer((__s), __p))
 
 /* Determine object index from a given position */
 static inline int slab_index(void *p, struct kmem_cache *s, void *addr)
@@ -502,7 +484,7 @@ static void slab_fix(struct kmem_cache *s, char *fmt, ...)
 static void print_trailer(struct kmem_cache *s, struct page *page, u8 *p)
 {
 	unsigned int off;	/* Offset of last byte */
-	u8 *addr = slab_address(page);
+	u8 *addr = page_address(page);
 
 	print_tracking(s, p);
 
@@ -680,7 +662,7 @@ static int slab_pad_check(struct kmem_cache *s, struct page *page)
 	if (!(s->flags & SLAB_POISON))
 		return 1;
 
-	start = slab_address(page);
+	start = page_address(page);
 	end = start + (PAGE_SIZE << s->order);
 	length = s->objects * s->size;
 	remainder = end - (start + length);
@@ -748,7 +730,7 @@ static int check_object(struct kmem_cache *s, struct page *page,
 		 * of the free objects in this slab. May cause
 		 * another error because the object count is now wrong.
 		 */
-		set_freepointer(s, p, page->end);
+		set_freepointer(s, p, NULL);
 		return 0;
 	}
 	return 1;
@@ -782,18 +764,18 @@ static int on_freelist(struct kmem_cache *s, struct page *page, void *search)
 	void *fp = page->freelist;
 	void *object = NULL;
 
-	while (fp != page->end && nr <= s->objects) {
+	while (fp && nr <= s->objects) {
 		if (fp == search)
 			return 1;
 		if (!check_valid_pointer(s, page, fp)) {
 			if (object) {
 				object_err(s, page, object,
 					"Freechain corrupt");
-				set_freepointer(s, object, page->end);
+				set_freepointer(s, object, NULL);
 				break;
 			} else {
 				slab_err(s, page, "Freepointer corrupt");
-				page->freelist = page->end;
+				page->freelist = NULL;
 				page->inuse = s->objects;
 				slab_fix(s, "Freelist cleared");
 				return 0;
@@ -899,7 +881,7 @@ bad:
 		 */
 		slab_fix(s, "Marking all objects used");
 		page->inuse = s->objects;
-		page->freelist = page->end;
+		page->freelist = NULL;
 	}
 	return 0;
 }
@@ -939,7 +921,7 @@ static int free_debug_processing(struct kmem_cache *s, struct page *page,
 	}
 
 	/* Special debug activities for freeing objects */
-	if (!SlabFrozen(page) && page->freelist == page->end)
+	if (!SlabFrozen(page) && !page->freelist)
 		remove_full(s, page);
 	if (s->flags & SLAB_STORE_USER)
 		set_track(s, object, TRACK_FREE, addr);
@@ -1124,7 +1106,6 @@ static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node)
 		SetSlabDebug(page);
 
 	start = page_address(page);
-	page->end = start + 1;
 
 	if (unlikely(s->flags & SLAB_POISON))
 		memset(start, POISON_INUSE, PAGE_SIZE << s->order);
@@ -1136,7 +1117,7 @@ static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node)
 		last = p;
 	}
 	setup_object(s, page, last);
-	set_freepointer(s, last, page->end);
+	set_freepointer(s, last, NULL);
 
 	page->freelist = start;
 	page->inuse = 0;
@@ -1152,7 +1133,7 @@ static void __free_slab(struct kmem_cache *s, struct page *page)
 		void *p;
 
 		slab_pad_check(s, page);
-		for_each_object(p, s, slab_address(page))
+		for_each_object(p, s, page_address(page))
 			check_object(s, page, p, 0);
 		ClearSlabDebug(page);
 	}
@@ -1162,7 +1143,6 @@ static void __free_slab(struct kmem_cache *s, struct page *page)
 		NR_SLAB_RECLAIMABLE : NR_SLAB_UNRECLAIMABLE,
 		-pages);
 
-	page->mapping = NULL;
 	__free_pages(page, s->order);
 }
 
@@ -1366,7 +1346,7 @@ static void unfreeze_slab(struct kmem_cache *s, struct page *page, int tail)
 	ClearSlabFrozen(page);
 	if (page->inuse) {
 
-		if (page->freelist != page->end) {
+		if (page->freelist) {
 			add_partial(n, page, tail);
 			stat(c, tail ? DEACTIVATE_TO_TAIL : DEACTIVATE_TO_HEAD);
 		} else {
@@ -1410,12 +1390,8 @@ static void deactivate_slab(struct kmem_cache *s, struct kmem_cache_cpu *c)
 	 * Merge cpu freelist into freelist. Typically we get here
 	 * because both freelists are empty. So this is unlikely
 	 * to occur.
-	 *
-	 * We need to use _is_end here because deactivate slab may
-	 * be called for a debug slab. Then c->freelist may contain
-	 * a dummy pointer.
 	 */
-	while (unlikely(!is_end(c->freelist))) {
+	while (unlikely(c->freelist)) {
 		void **object;
 
 		tail = 0;	/* Hot objects. Put the slab first */
@@ -1517,7 +1493,7 @@ static void *__slab_alloc(struct kmem_cache *s,
 	stat(c, ALLOC_REFILL);
 load_freelist:
 	object = c->page->freelist;
-	if (unlikely(object == c->page->end))
+	if (unlikely(!object))
 		goto another_slab;
 	if (unlikely(SlabDebug(c->page)))
 		goto debug;
@@ -1525,7 +1501,7 @@ load_freelist:
 	object = c->page->freelist;
 	c->freelist = object[c->offset];
 	c->page->inuse = s->objects;
-	c->page->freelist = c->page->end;
+	c->page->freelist = NULL;
 	c->node = page_to_nid(c->page);
 unlock_out:
 	slab_unlock(c->page);
@@ -1607,7 +1583,7 @@ static __always_inline void *slab_alloc(struct kmem_cache *s,
 
 	local_irq_save(flags);
 	c = get_cpu_slab(s, smp_processor_id());
-	if (unlikely(is_end(c->freelist) || !node_match(c, node)))
+	if (unlikely(!c->freelist || !node_match(c, node)))
 
 		object = __slab_alloc(s, gfpflags, node, addr, c);
 
@@ -1677,7 +1653,7 @@ checks_ok:
 	 * was not on the partial list before
 	 * then add it.
 	 */
-	if (unlikely(prior == page->end)) {
+	if (unlikely(!prior)) {
 		add_partial(get_node(s, page_to_nid(page)), page, 1);
 		stat(c, FREE_ADD_PARTIAL);
 	}
@@ -1687,7 +1663,7 @@ out_unlock:
 	return;
 
 slab_empty:
-	if (prior != page->end) {
+	if (prior) {
 		/*
 		 * Slab still on the partial list.
 		 */
@@ -1910,7 +1886,7 @@ static void init_kmem_cache_cpu(struct kmem_cache *s,
 			struct kmem_cache_cpu *c)
 {
 	c->page = NULL;
-	c->freelist = (void *)PAGE_MAPPING_ANON;
+	c->freelist = NULL;
 	c->node = 0;
 	c->offset = s->offset / sizeof(void *);
 	c->objsize = s->objsize;
@@ -3199,7 +3175,7 @@ static int validate_slab(struct kmem_cache *s, struct page *page,
 						unsigned long *map)
 {
 	void *p;
-	void *addr = slab_address(page);
+	void *addr = page_address(page);
 
 	if (!check_slab(s, page) ||
 			!on_freelist(s, page, NULL))
@@ -3482,7 +3458,7 @@ static int add_location(struct loc_track *t, struct kmem_cache *s,
 static void process_slab(struct loc_track *t, struct kmem_cache *s,
 		struct page *page, enum track_item alloc)
 {
-	void *addr = slab_address(page);
+	void *addr = page_address(page);
 	DECLARE_BITMAP(map, s->objects);
 	void *p;
 
-- 
cgit v1.2.3


From 6446faa2ff30ca77c5b25e886bbbfb81c63f1c91 Mon Sep 17 00:00:00 2001
From: Christoph Lameter <clameter@sgi.com>
Date: Fri, 15 Feb 2008 23:45:26 -0800
Subject: slub: Fix up comments

Provide comments and fix up various spelling / style issues.

Signed-off-by: Christoph Lameter <clameter@sgi.com>
---
 include/linux/slub_def.h |  4 ++--
 mm/slub.c                | 49 +++++++++++++++++++++++++++---------------------
 2 files changed, 30 insertions(+), 23 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h
index 57deecc79d52..b00c1c73eb0a 100644
--- a/include/linux/slub_def.h
+++ b/include/linux/slub_def.h
@@ -61,7 +61,7 @@ struct kmem_cache {
 	int size;		/* The size of an object including meta data */
 	int objsize;		/* The size of an object without meta data */
 	int offset;		/* Free pointer offset. */
-	int order;
+	int order;		/* Current preferred allocation order */
 
 	/*
 	 * Avoid an extra cache line for UP, SMP and for the node local to
@@ -138,11 +138,11 @@ static __always_inline int kmalloc_index(size_t size)
 	if (size <=        512) return 9;
 	if (size <=       1024) return 10;
 	if (size <=   2 * 1024) return 11;
+	if (size <=   4 * 1024) return 12;
 /*
  * The following is only needed to support architectures with a larger page
  * size than 4k.
  */
-	if (size <=   4 * 1024) return 12;
 	if (size <=   8 * 1024) return 13;
 	if (size <=  16 * 1024) return 14;
 	if (size <=  32 * 1024) return 15;
diff --git a/mm/slub.c b/mm/slub.c
index 72f5f4ecd1d2..10d546954efa 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -291,6 +291,7 @@ static inline struct kmem_cache_cpu *get_cpu_slab(struct kmem_cache *s, int cpu)
 #endif
 }
 
+/* Verify that a pointer has an address that is valid within a slab page */
 static inline int check_valid_pointer(struct kmem_cache *s,
 				struct page *page, const void *object)
 {
@@ -619,7 +620,7 @@ static int check_bytes_and_report(struct kmem_cache *s, struct page *page,
  * 	A. Free pointer (if we cannot overwrite object on free)
  * 	B. Tracking data for SLAB_STORE_USER
  * 	C. Padding to reach required alignment boundary or at mininum
- * 		one word if debuggin is on to be able to detect writes
+ * 		one word if debugging is on to be able to detect writes
  * 		before the word boundary.
  *
  *	Padding is done using 0x5a (POISON_INUSE)
@@ -1268,7 +1269,7 @@ static struct page *get_any_partial(struct kmem_cache *s, gfp_t flags)
 	 * may return off node objects because partial slabs are obtained
 	 * from other nodes and filled up.
 	 *
-	 * If /sys/slab/xx/defrag_ratio is set to 100 (which makes
+	 * If /sys/kernel/slab/xx/defrag_ratio is set to 100 (which makes
 	 * defrag_ratio = 1000) then every (well almost) allocation will
 	 * first attempt to defrag slab caches on other nodes. This means
 	 * scanning over all nodes to look for partial slabs which may be
@@ -1343,9 +1344,11 @@ static void unfreeze_slab(struct kmem_cache *s, struct page *page, int tail)
 			 * Adding an empty slab to the partial slabs in order
 			 * to avoid page allocator overhead. This slab needs
 			 * to come after the other slabs with objects in
-			 * order to fill them up. That way the size of the
-			 * partial list stays small. kmem_cache_shrink can
-			 * reclaim empty slabs from the partial list.
+			 * so that the others get filled first. That way the
+			 * size of the partial list stays small.
+			 *
+			 * kmem_cache_shrink can reclaim any empty slabs from the
+			 * partial list.
 			 */
 			add_partial(n, page, 1);
 			slab_unlock(page);
@@ -1368,7 +1371,7 @@ static void deactivate_slab(struct kmem_cache *s, struct kmem_cache_cpu *c)
 	if (c->freelist)
 		stat(c, DEACTIVATE_REMOTE_FREES);
 	/*
-	 * Merge cpu freelist into freelist. Typically we get here
+	 * Merge cpu freelist into slab freelist. Typically we get here
 	 * because both freelists are empty. So this is unlikely
 	 * to occur.
 	 */
@@ -1399,6 +1402,7 @@ static inline void flush_slab(struct kmem_cache *s, struct kmem_cache_cpu *c)
 
 /*
  * Flush cpu slab.
+ *
  * Called from IPI handler with interrupts disabled.
  */
 static inline void __flush_cpu_slab(struct kmem_cache *s, int cpu)
@@ -1457,7 +1461,8 @@ static inline int node_match(struct kmem_cache_cpu *c, int node)
  * rest of the freelist to the lockless freelist.
  *
  * And if we were unable to get a new slab from the partial slab lists then
- * we need to allocate a new slab. This is slowest path since we may sleep.
+ * we need to allocate a new slab. This is the slowest path since it involves
+ * a call to the page allocator and the setup of a new slab.
  */
 static void *__slab_alloc(struct kmem_cache *s,
 		gfp_t gfpflags, int node, void *addr, struct kmem_cache_cpu *c)
@@ -1471,7 +1476,9 @@ static void *__slab_alloc(struct kmem_cache *s,
 	slab_lock(c->page);
 	if (unlikely(!node_match(c, node)))
 		goto another_slab;
+
 	stat(c, ALLOC_REFILL);
+
 load_freelist:
 	object = c->page->freelist;
 	if (unlikely(!object))
@@ -1616,6 +1623,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page,
 
 	if (unlikely(SlabDebug(page)))
 		goto debug;
+
 checks_ok:
 	prior = object[offset] = page->freelist;
 	page->freelist = object;
@@ -1630,8 +1638,7 @@ checks_ok:
 		goto slab_empty;
 
 	/*
-	 * Objects left in the slab. If it
-	 * was not on the partial list before
+	 * Objects left in the slab. If it was not on the partial list before
 	 * then add it.
 	 */
 	if (unlikely(!prior)) {
@@ -1845,13 +1852,11 @@ static unsigned long calculate_alignment(unsigned long flags,
 		unsigned long align, unsigned long size)
 {
 	/*
-	 * If the user wants hardware cache aligned objects then
-	 * follow that suggestion if the object is sufficiently
-	 * large.
+	 * If the user wants hardware cache aligned objects then follow that
+	 * suggestion if the object is sufficiently large.
 	 *
-	 * The hardware cache alignment cannot override the
-	 * specified alignment though. If that is greater
-	 * then use it.
+	 * The hardware cache alignment cannot override the specified
+	 * alignment though. If that is greater then use it.
 	 */
 	if ((flags & SLAB_HWCACHE_ALIGN) &&
 			size > cache_line_size() / 2)
@@ -2049,6 +2054,7 @@ static struct kmem_cache_node *early_kmem_cache_node_alloc(gfp_t gfpflags,
 #endif
 	init_kmem_cache_node(n);
 	atomic_long_inc(&n->nr_slabs);
+
 	/*
 	 * lockdep requires consistent irq usage for each lock
 	 * so even though there cannot be a race this early in
@@ -2301,7 +2307,7 @@ int kmem_ptr_validate(struct kmem_cache *s, const void *object)
 	/*
 	 * We could also check if the object is on the slabs freelist.
 	 * But this would be too expensive and it seems that the main
-	 * purpose of kmem_ptr_valid is to check if the object belongs
+	 * purpose of kmem_ptr_valid() is to check if the object belongs
 	 * to a certain slab.
 	 */
 	return 1;
@@ -2913,7 +2919,7 @@ void __init kmem_cache_init(void)
 	/*
 	 * Patch up the size_index table if we have strange large alignment
 	 * requirements for the kmalloc array. This is only the case for
-	 * mips it seems. The standard arches will not generate any code here.
+	 * MIPS it seems. The standard arches will not generate any code here.
 	 *
 	 * Largest permitted alignment is 256 bytes due to the way we
 	 * handle the index determination for the smaller caches.
@@ -2942,7 +2948,6 @@ void __init kmem_cache_init(void)
 	kmem_size = sizeof(struct kmem_cache);
 #endif
 
-
 	printk(KERN_INFO
 		"SLUB: Genslabs=%d, HWalign=%d, Order=%d-%d, MinObjects=%d,"
 		" CPUs=%d, Nodes=%d\n",
@@ -3039,12 +3044,15 @@ struct kmem_cache *kmem_cache_create(const char *name, size_t size,
 		 */
 		for_each_online_cpu(cpu)
 			get_cpu_slab(s, cpu)->objsize = s->objsize;
+
 		s->inuse = max_t(int, s->inuse, ALIGN(size, sizeof(void *)));
 		up_write(&slub_lock);
+
 		if (sysfs_slab_alias(s, name))
 			goto err;
 		return s;
 	}
+
 	s = kmalloc(kmem_size, GFP_KERNEL);
 	if (s) {
 		if (kmem_cache_open(s, GFP_KERNEL, name,
@@ -3927,7 +3935,6 @@ SLAB_ATTR(remote_node_defrag_ratio);
 #endif
 
 #ifdef CONFIG_SLUB_STATS
-
 static int show_stat(struct kmem_cache *s, char *buf, enum stat_item si)
 {
 	unsigned long sum  = 0;
@@ -4111,8 +4118,8 @@ static struct kset *slab_kset;
 #define ID_STR_LENGTH 64
 
 /* Create a unique string id for a slab cache:
- * format
- * :[flags-]size:[memory address of kmemcache]
+ *
+ * Format	:[flags-]size
  */
 static char *create_unique_id(struct kmem_cache *s)
 {
-- 
cgit v1.2.3


From e898d4db2749c6052072e9bc4448e396cbdeb06a Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Sat, 1 Mar 2008 01:06:47 +0900
Subject: [UDP]: Allow users to configure UDP-Lite.

Let's give users an option for disabling UDP-Lite (~4K).

old:
|    text	   data	    bss	    dec	    hex	filename
|  286498	  12432	   6072	 305002	  4a76a	net/ipv4/built-in.o
|  193830	   8192	   3204	 205226	  321aa	net/ipv6/ipv6.o

new (without UDP-Lite):
|    text	   data	    bss	    dec	    hex	filename
|  284086	  12136	   5432	 301654	  49a56	net/ipv4/built-in.o
|  191835	   7832	   3076	 202743	  317f7	net/ipv6/ipv6.o

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/udp.h      | 11 +++++++++++
 include/net/ipv6.h       |  5 +++++
 include/net/transp_v6.h  |  5 +++++
 include/net/udplite.h    |  9 +++++++--
 net/ipv4/Kconfig         | 10 ++++++++++
 net/ipv4/Makefile        |  3 ++-
 net/ipv4/af_inet.c       |  7 ++++++-
 net/ipv4/proc.c          |  5 ++++-
 net/ipv4/udp.c           | 24 ++++++++++++++----------
 net/ipv6/Makefile        |  3 ++-
 net/ipv6/af_inet6.c      | 14 ++++++++++++++
 net/ipv6/ipv6_sockglue.c |  6 +++++-
 net/ipv6/proc.c          |  6 ++++++
 net/ipv6/udp.c           | 16 ++++++++--------
 14 files changed, 99 insertions(+), 25 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/udp.h b/include/linux/udp.h
index 8ec703f462da..4144664d69d9 100644
--- a/include/linux/udp.h
+++ b/include/linux/udp.h
@@ -70,8 +70,10 @@ struct udp_sock {
 #define UDPLITE_BIT      0x1  		/* set by udplite proto init function */
 #define UDPLITE_SEND_CC  0x2  		/* set via udplite setsockopt         */
 #define UDPLITE_RECV_CC  0x4		/* set via udplite setsocktopt        */
+#ifdef CONFIG_IP_UDPLITE
 	__u8		 pcflag;        /* marks socket as UDP-Lite if > 0    */
 	__u8		 unused[3];
+#endif
 	/*
 	 * For encapsulation sockets.
 	 */
@@ -82,7 +84,16 @@ static inline struct udp_sock *udp_sk(const struct sock *sk)
 {
 	return (struct udp_sock *)sk;
 }
+
+#ifdef CONFIG_IP_UDPLITE
 #define IS_UDPLITE(__sk) (udp_sk(__sk)->pcflag)
+#define IS_PROTO_UDPLITE(__proto) ((__proto) == IPPROTO_UDPLITE)
+#define IS_SOL_UDPFAMILY(level) ((level) == SOL_UDP || (level) == SOL_UDPLITE)
+#else
+#define IS_UDPLITE(__sk) 0
+#define IS_PROTO_UDPLITE(__proto) 0
+#define IS_SOL_UDPFAMILY(level) ((level) == SOL_UDP)
+#endif
 
 #endif
 
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 8b05c65415cb..96b1763bfcaa 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -603,8 +603,13 @@ extern int  tcp6_proc_init(void);
 extern void tcp6_proc_exit(void);
 extern int  udp6_proc_init(void);
 extern void udp6_proc_exit(void);
+#ifdef CONFIG_IP_UDPLITE
 extern int  udplite6_proc_init(void);
 extern void udplite6_proc_exit(void);
+#else
+static inline int udplite6_proc_init(void) { return 0; }
+static inline void udplite6_proc_exit(void) { }
+#endif
 extern int  ipv6_misc_proc_init(void);
 extern void ipv6_misc_proc_exit(void);
 extern int snmp6_register_dev(struct inet6_dev *idev);
diff --git a/include/net/transp_v6.h b/include/net/transp_v6.h
index 27394e0447d8..902e6c6bc793 100644
--- a/include/net/transp_v6.h
+++ b/include/net/transp_v6.h
@@ -27,8 +27,13 @@ extern int				rawv6_init(void);
 extern void				rawv6_exit(void);
 extern int				udpv6_init(void);
 extern void				udpv6_exit(void);
+#ifdef CONFIG_IP_UDPLITE
 extern int 				udplitev6_init(void);
 extern void 				udplitev6_exit(void);
+#else
+static inline int udplitev6_init(void) { return 0; }
+static inline void udplitev6_exit(void) { }
+#endif
 extern int				tcpv6_init(void);
 extern void				tcpv6_exit(void);
 
diff --git a/include/net/udplite.h b/include/net/udplite.h
index b76b2e377af4..01ddb2c20264 100644
--- a/include/net/udplite.h
+++ b/include/net/udplite.h
@@ -25,7 +25,9 @@ static __inline__ int udplite_getfrag(void *from, char *to, int  offset,
 /* Designate sk as UDP-Lite socket */
 static inline int udplite_sk_init(struct sock *sk)
 {
+#ifdef CONFIG_IP_UDPLITE
 	udp_sk(sk)->pcflag = UDPLITE_BIT;
+#endif
 	return 0;
 }
 
@@ -69,7 +71,7 @@ static inline int udplite_checksum_init(struct sk_buff *skb, struct udphdr *uh)
 static inline int udplite_sender_cscov(struct udp_sock *up, struct udphdr *uh)
 {
 	int cscov = up->len;
-
+#ifdef CONFIG_IP_UDPLITE
 	/*
 	 * Sender has set `partial coverage' option on UDP-Lite socket
 	 */
@@ -93,13 +95,15 @@ static inline int udplite_sender_cscov(struct udp_sock *up, struct udphdr *uh)
 	 *       illegal, we fall back to the defaults here.
 	 */
 	}
+#endif
 	return cscov;
 }
 
 static inline __wsum udplite_csum_outgoing(struct sock *sk, struct sk_buff *skb)
 {
-	int cscov = udplite_sender_cscov(udp_sk(sk), udp_hdr(skb));
 	__wsum csum = 0;
+#ifdef CONFIG_IP_UDPLITE
+	int cscov = udplite_sender_cscov(udp_sk(sk), udp_hdr(skb));
 
 	skb->ip_summed = CHECKSUM_NONE;     /* no HW support for checksumming */
 
@@ -112,6 +116,7 @@ static inline __wsum udplite_csum_outgoing(struct sock *sk, struct sk_buff *skb)
 		if ((cscov -= len) <= 0)
 			break;
 	}
+#endif
 	return csum;
 }
 
diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig
index 19880b086e71..efe3832c4ad8 100644
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -632,5 +632,15 @@ config TCP_MD5SIG
 
 	  If unsure, say N.
 
+config IP_UDPLITE
+	bool "IP: UDP-Lite Protocol (RFC 3828)"
+	default n
+	---help---
+	  UDP-Lite (RFC 3828) is a UDP-like protocol with variable-length
+	  checksum.  Read <file:Documentation/networking/udplite.txt> for
+	  details.
+
+	  If unsure, say N.
+
 source "net/ipv4/ipvs/Kconfig"
 
diff --git a/net/ipv4/Makefile b/net/ipv4/Makefile
index ad40ef3f9ebc..e88cebdf3e30 100644
--- a/net/ipv4/Makefile
+++ b/net/ipv4/Makefile
@@ -8,7 +8,7 @@ obj-y     := route.o inetpeer.o protocol.o \
 	     inet_timewait_sock.o inet_connection_sock.o \
 	     tcp.o tcp_input.o tcp_output.o tcp_timer.o tcp_ipv4.o \
 	     tcp_minisocks.o tcp_cong.o \
-	     datagram.o raw.o udp.o udplite.o \
+	     datagram.o raw.o udp.o \
 	     arp.o icmp.o devinet.o af_inet.o  igmp.o \
 	     fib_frontend.o fib_semantics.o \
 	     inet_fragment.o
@@ -49,6 +49,7 @@ obj-$(CONFIG_TCP_CONG_SCALABLE) += tcp_scalable.o
 obj-$(CONFIG_TCP_CONG_LP) += tcp_lp.o
 obj-$(CONFIG_TCP_CONG_YEAH) += tcp_yeah.o
 obj-$(CONFIG_TCP_CONG_ILLINOIS) += tcp_illinois.o
+obj-$(CONFIG_IP_UDPLITE) += udplite.o
 obj-$(CONFIG_NETLABEL) += cipso_ipv4.o
 
 obj-$(CONFIG_XFRM) += xfrm4_policy.o xfrm4_state.o xfrm4_input.o \
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 4f539bd48718..67260c0eaaa8 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1317,15 +1317,18 @@ static int __init init_ipv4_mibs(void)
 	if (snmp_mib_init((void **)udp_statistics,
 			  sizeof(struct udp_mib)) < 0)
 		goto err_udp_mib;
+#ifdef CONFIG_IP_UDPLITE
 	if (snmp_mib_init((void **)udplite_statistics,
 			  sizeof(struct udp_mib)) < 0)
 		goto err_udplite_mib;
-
+#endif
 	tcp_mib_init();
 
 	return 0;
 
+#ifdef CONFIG_IP_UDPLITE
 err_udplite_mib:
+#endif
 	snmp_mib_free((void **)udp_statistics);
 err_udp_mib:
 	snmp_mib_free((void **)tcp_statistics);
@@ -1423,8 +1426,10 @@ static int __init inet_init(void)
 	/* Setup UDP memory threshold */
 	udp_init();
 
+#ifdef CONFIG_IP_UDPLITE
 	/* Add UDP-Lite (RFC 3828) */
 	udplite4_register();
+#endif
 
 	/*
 	 *	Set the ICMP layer up
diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c
index d63474c6b400..d75ddb7fa4b8 100644
--- a/net/ipv4/proc.c
+++ b/net/ipv4/proc.c
@@ -59,7 +59,9 @@ static int sockstat_seq_show(struct seq_file *seq, void *v)
 		   atomic_read(&tcp_memory_allocated));
 	seq_printf(seq, "UDP: inuse %d mem %d\n", sock_prot_inuse_get(&udp_prot),
 		   atomic_read(&udp_memory_allocated));
+#ifdef CONFIG_IP_UDPLITE
 	seq_printf(seq, "UDPLITE: inuse %d\n", sock_prot_inuse_get(&udplite_prot));
+#endif
 	seq_printf(seq, "RAW: inuse %d\n", sock_prot_inuse_get(&raw_prot));
 	seq_printf(seq,  "FRAG: inuse %d memory %d\n",
 			ip_frag_nqueues(&init_net), ip_frag_mem(&init_net));
@@ -349,6 +351,7 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 			   snmp_fold_field((void **)udp_statistics,
 					   snmp4_udp_list[i].entry));
 
+#ifdef CONFIG_IP_UDPLITE
 	/* the UDP and UDP-Lite MIBs are the same */
 	seq_puts(seq, "\nUdpLite:");
 	for (i = 0; snmp4_udp_list[i].name != NULL; i++)
@@ -359,7 +362,7 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 		seq_printf(seq, " %lu",
 			   snmp_fold_field((void **)udplite_statistics,
 					   snmp4_udp_list[i].entry));
-
+#endif
 	seq_putc(seq, '\n');
 	return 0;
 }
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 7ea1b67b6de1..acc353aa89eb 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -1127,7 +1127,7 @@ static inline int udp4_csum_init(struct sk_buff *skb, struct udphdr *uh,
 	UDP_SKB_CB(skb)->partial_cov = 0;
 	UDP_SKB_CB(skb)->cscov = skb->len;
 
-	if (proto == IPPROTO_UDPLITE) {
+	if (IS_PROTO_UDPLITE(proto)) {
 		err = udplite_checksum_init(skb, uh);
 		if (err)
 			return err;
@@ -1175,7 +1175,7 @@ int __udp4_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 	if (ulen > skb->len)
 		goto short_packet;
 
-	if (proto == IPPROTO_UDP) {
+	if (IS_PROTO_UDPLITE(proto)) {
 		/* UDP validates ulen. */
 		if (ulen < sizeof(*uh) || pskb_trim_rcsum(skb, ulen))
 			goto short_packet;
@@ -1217,7 +1217,7 @@ int __udp4_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 	if (udp_lib_checksum_complete(skb))
 		goto csum_error;
 
-	UDP_INC_STATS_BH(UDP_MIB_NOPORTS, proto == IPPROTO_UDPLITE);
+	UDP_INC_STATS_BH(UDP_MIB_NOPORTS, IS_PROTO_UDPLITE(proto));
 	icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);
 
 	/*
@@ -1229,7 +1229,7 @@ int __udp4_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 
 short_packet:
 	LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: short packet: From %u.%u.%u.%u:%u %d/%d to %u.%u.%u.%u:%u\n",
-		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
+		       IS_PROTO_UDPLITE(proto) ? "-Lite" : "",
 		       NIPQUAD(saddr),
 		       ntohs(uh->source),
 		       ulen,
@@ -1244,14 +1244,14 @@ csum_error:
 	 * the network is concerned, anyway) as per 4.1.3.4 (MUST).
 	 */
 	LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: bad checksum. From %d.%d.%d.%d:%d to %d.%d.%d.%d:%d ulen %d\n",
-		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
+		       IS_PROTO_UDPLITE(proto) ? "-Lite" : "",
 		       NIPQUAD(saddr),
 		       ntohs(uh->source),
 		       NIPQUAD(daddr),
 		       ntohs(uh->dest),
 		       ulen);
 drop:
-	UDP_INC_STATS_BH(UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
+	UDP_INC_STATS_BH(UDP_MIB_INERRORS, IS_PROTO_UDPLITE(proto));
 	kfree_skb(skb);
 	return 0;
 }
@@ -1279,7 +1279,9 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 	struct udp_sock *up = udp_sk(sk);
 	int val;
 	int err = 0;
+#ifdef CONFIG_IP_UDPLITE
 	int is_udplite = IS_UDPLITE(sk);
+#endif
 
 	if (optlen<sizeof(int))
 		return -EINVAL;
@@ -1315,6 +1317,7 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 		}
 		break;
 
+#ifdef CONFIG_IP_UDPLITE
 	/*
 	 * 	UDP-Lite's partial checksum coverage (RFC 3828).
 	 */
@@ -1340,6 +1343,7 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 		up->pcrlen = val;
 		up->pcflag |= UDPLITE_RECV_CC;
 		break;
+#endif
 
 	default:
 		err = -ENOPROTOOPT;
@@ -1352,7 +1356,7 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 int udp_setsockopt(struct sock *sk, int level, int optname,
 		   char __user *optval, int optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
 					  udp_push_pending_frames);
 	return ip_setsockopt(sk, level, optname, optval, optlen);
@@ -1362,7 +1366,7 @@ int udp_setsockopt(struct sock *sk, int level, int optname,
 int compat_udp_setsockopt(struct sock *sk, int level, int optname,
 			  char __user *optval, int optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
 					  udp_push_pending_frames);
 	return compat_ip_setsockopt(sk, level, optname, optval, optlen);
@@ -1416,7 +1420,7 @@ int udp_lib_getsockopt(struct sock *sk, int level, int optname,
 int udp_getsockopt(struct sock *sk, int level, int optname,
 		   char __user *optval, int __user *optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
 	return ip_getsockopt(sk, level, optname, optval, optlen);
 }
@@ -1425,7 +1429,7 @@ int udp_getsockopt(struct sock *sk, int level, int optname,
 int compat_udp_getsockopt(struct sock *sk, int level, int optname,
 				 char __user *optval, int __user *optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
 	return compat_ip_getsockopt(sk, level, optname, optval, optlen);
 }
diff --git a/net/ipv6/Makefile b/net/ipv6/Makefile
index ae14617e607f..81969479955f 100644
--- a/net/ipv6/Makefile
+++ b/net/ipv6/Makefile
@@ -6,7 +6,7 @@ obj-$(CONFIG_IPV6) += ipv6.o
 
 ipv6-objs :=	af_inet6.o anycast.o ip6_output.o ip6_input.o addrconf.o \
 		addrlabel.o \
-		route.o ip6_fib.o ipv6_sockglue.o ndisc.o udp.o udplite.o \
+		route.o ip6_fib.o ipv6_sockglue.o ndisc.o udp.o \
 		raw.o protocol.o icmp.o mcast.o reassembly.o tcp_ipv6.o \
 		exthdrs.o datagram.o ip6_flowlabel.o inet6_connection_sock.o
 
@@ -17,6 +17,7 @@ ipv6-$(CONFIG_NETFILTER) += netfilter.o
 ipv6-$(CONFIG_IPV6_MULTIPLE_TABLES) += fib6_rules.o
 ipv6-$(CONFIG_PROC_FS) += proc.o
 ipv6-$(CONFIG_SYN_COOKIES) += syncookies.o
+ipv6-$(CONFIG_IP_UDPLITE) += udplite.o
 
 ipv6-objs += $(ipv6-y)
 
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index 9869f87243cf..243c42a6b80d 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -691,12 +691,16 @@ static int __init init_ipv6_mibs(void)
 		goto err_icmpmsg_mib;
 	if (snmp_mib_init((void **)udp_stats_in6, sizeof (struct udp_mib)) < 0)
 		goto err_udp_mib;
+#ifdef CONFIG_IP_UDPLITE
 	if (snmp_mib_init((void **)udplite_stats_in6,
 			  sizeof (struct udp_mib)) < 0)
 		goto err_udplite_mib;
+#endif
 	return 0;
 
+#ifdef CONFIG_IP_UDPLITE
 err_udplite_mib:
+#endif
 	snmp_mib_free((void **)udp_stats_in6);
 err_udp_mib:
 	snmp_mib_free((void **)icmpv6msg_statistics);
@@ -715,7 +719,9 @@ static void cleanup_ipv6_mibs(void)
 	snmp_mib_free((void **)icmpv6_statistics);
 	snmp_mib_free((void **)icmpv6msg_statistics);
 	snmp_mib_free((void **)udp_stats_in6);
+#ifdef CONFIG_IP_UDPLITE
 	snmp_mib_free((void **)udplite_stats_in6);
+#endif
 }
 
 static int inet6_net_init(struct net *net)
@@ -760,9 +766,11 @@ static int __init inet6_init(void)
 	if (err)
 		goto out_unregister_tcp_proto;
 
+#ifdef CONFIG_IP_UDPLITE
 	err = proto_register(&udplitev6_prot, 1);
 	if (err)
 		goto out_unregister_udp_proto;
+#endif
 
 	err = proto_register(&rawv6_prot, 1);
 	if (err)
@@ -933,8 +941,10 @@ out_sock_register_fail:
 out_unregister_raw_proto:
 	proto_unregister(&rawv6_prot);
 out_unregister_udplite_proto:
+#ifdef CONFIG_IP_UDPLITE
 	proto_unregister(&udplitev6_prot);
 out_unregister_udp_proto:
+#endif
 	proto_unregister(&udpv6_prot);
 out_unregister_tcp_proto:
 	proto_unregister(&tcpv6_prot);
@@ -950,7 +960,9 @@ static void __exit inet6_exit(void)
 	rtnl_unregister_all(PF_INET6);
 
 	udpv6_exit();
+#ifdef CONFIG_IP_UDPLITE
 	udplitev6_exit();
+#endif
 	tcpv6_exit();
 
 	/* Cleanup code parts. */
@@ -982,7 +994,9 @@ static void __exit inet6_exit(void)
 	unregister_pernet_subsys(&inet6_net_ops);
 	cleanup_ipv6_mibs();
 	proto_unregister(&rawv6_prot);
+#ifdef CONFIG_IP_UDPLITE
 	proto_unregister(&udplitev6_prot);
+#endif
 	proto_unregister(&udpv6_prot);
 	proto_unregister(&tcpv6_prot);
 }
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index bf2a686aa13d..0a18fecb93d1 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -239,7 +239,9 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 			struct sk_buff *pktopt;
 
 			if (sk->sk_protocol != IPPROTO_UDP &&
+#ifdef CONFIG_IP_UDPLITE
 			    sk->sk_protocol != IPPROTO_UDPLITE &&
+#endif
 			    sk->sk_protocol != IPPROTO_TCP)
 				break;
 
@@ -279,7 +281,7 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 			} else {
 				struct proto *prot = &udp_prot;
 
-				if (sk->sk_protocol == IPPROTO_UDPLITE)
+				if (IS_PROTO_UDPLITE(sk->sk_protocol))
 					prot = &udplite_prot;
 				local_bh_disable();
 				sock_prot_inuse_add(sk->sk_prot, -1);
@@ -844,7 +846,9 @@ static int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
 	switch (optname) {
 	case IPV6_ADDRFORM:
 		if (sk->sk_protocol != IPPROTO_UDP &&
+#ifdef CONFIG_IP_UDPLITE
 		    sk->sk_protocol != IPPROTO_UDPLITE &&
+#endif
 		    sk->sk_protocol != IPPROTO_TCP)
 			return -EINVAL;
 		if (sk->sk_state != TCP_ESTABLISHED)
diff --git a/net/ipv6/proc.c b/net/ipv6/proc.c
index 199ef379e501..5ba7ae849d04 100644
--- a/net/ipv6/proc.c
+++ b/net/ipv6/proc.c
@@ -39,8 +39,10 @@ static int sockstat6_seq_show(struct seq_file *seq, void *v)
 		       sock_prot_inuse_get(&tcpv6_prot));
 	seq_printf(seq, "UDP6: inuse %d\n",
 		       sock_prot_inuse_get(&udpv6_prot));
+#ifdef CONFIG_IP_UDPLITE
 	seq_printf(seq, "UDPLITE6: inuse %d\n",
 			sock_prot_inuse_get(&udplitev6_prot));
+#endif
 	seq_printf(seq, "RAW6: inuse %d\n",
 		       sock_prot_inuse_get(&rawv6_prot));
 	seq_printf(seq, "FRAG6: inuse %d memory %d\n",
@@ -111,6 +113,7 @@ static struct snmp_mib snmp6_udp6_list[] = {
 	SNMP_MIB_SENTINEL
 };
 
+#ifdef CONFIG_IP_UDPLITE
 static struct snmp_mib snmp6_udplite6_list[] = {
 	SNMP_MIB_ITEM("UdpLite6InDatagrams", UDP_MIB_INDATAGRAMS),
 	SNMP_MIB_ITEM("UdpLite6NoPorts", UDP_MIB_NOPORTS),
@@ -118,6 +121,7 @@ static struct snmp_mib snmp6_udplite6_list[] = {
 	SNMP_MIB_ITEM("UdpLite6OutDatagrams", UDP_MIB_OUTDATAGRAMS),
 	SNMP_MIB_SENTINEL
 };
+#endif
 
 static void snmp6_seq_show_icmpv6msg(struct seq_file *seq, void **mib)
 {
@@ -176,7 +180,9 @@ static int snmp6_seq_show(struct seq_file *seq, void *v)
 		snmp6_seq_show_item(seq, (void **)icmpv6_statistics, snmp6_icmp6_list);
 		snmp6_seq_show_icmpv6msg(seq, (void **)icmpv6msg_statistics);
 		snmp6_seq_show_item(seq, (void **)udp_stats_in6, snmp6_udp6_list);
+#ifdef CONFIG_IP_UDPLITE
 		snmp6_seq_show_item(seq, (void **)udplite_stats_in6, snmp6_udplite6_list);
+#endif
 	}
 	return 0;
 }
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 53739de829db..55feac7ba717 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -400,7 +400,7 @@ static inline int udp6_csum_init(struct sk_buff *skb, struct udphdr *uh,
 	UDP_SKB_CB(skb)->partial_cov = 0;
 	UDP_SKB_CB(skb)->cscov = skb->len;
 
-	if (proto == IPPROTO_UDPLITE) {
+	if (IS_PROTO_UDPLITE(proto)) {
 		err = udplite_checksum_init(skb, uh);
 		if (err)
 			return err;
@@ -489,7 +489,7 @@ int __udp6_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 
 		if (udp_lib_checksum_complete(skb))
 			goto discard;
-		UDP6_INC_STATS_BH(UDP_MIB_NOPORTS, proto == IPPROTO_UDPLITE);
+		UDP6_INC_STATS_BH(UDP_MIB_NOPORTS, IS_PROTO_UDPLITE(proto));
 
 		icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0, dev);
 
@@ -510,11 +510,11 @@ int __udp6_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 
 short_packet:
 	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: %d/%u\n",
-		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
+		       IS_PROTO_UDPLITE(proto) ? "-Lite" : "",
 		       ulen, skb->len);
 
 discard:
-	UDP6_INC_STATS_BH(UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
+	UDP6_INC_STATS_BH(UDP_MIB_INERRORS, IS_PROTO_UDPLITE(proto));
 	kfree_skb(skb);
 	return 0;
 }
@@ -890,7 +890,7 @@ int udpv6_destroy_sock(struct sock *sk)
 int udpv6_setsockopt(struct sock *sk, int level, int optname,
 		     char __user *optval, int optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
 					  udp_v6_push_pending_frames);
 	return ipv6_setsockopt(sk, level, optname, optval, optlen);
@@ -900,7 +900,7 @@ int udpv6_setsockopt(struct sock *sk, int level, int optname,
 int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
 			    char __user *optval, int optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
 					  udp_v6_push_pending_frames);
 	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
@@ -910,7 +910,7 @@ int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
 int udpv6_getsockopt(struct sock *sk, int level, int optname,
 		     char __user *optval, int __user *optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
 	return ipv6_getsockopt(sk, level, optname, optval, optlen);
 }
@@ -919,7 +919,7 @@ int udpv6_getsockopt(struct sock *sk, int level, int optname,
 int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
 			    char __user *optval, int __user *optlen)
 {
-	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+	if (IS_SOL_UDPFAMILY(level))
 		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
 	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
 }
-- 
cgit v1.2.3


From 95e41e93e18d8e1e272ce23d96bae4f17ce11d42 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Thu, 6 Dec 2007 15:43:30 -0800
Subject: [IPV6]: Make ndisc_flow_init() common for later use.

For later use, this patch is renaming ndisc_flow_init() to
icmpv6_flow_init() and putting it in common place.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/icmpv6.h |  8 ++++++++
 net/ipv6/icmp.c        | 16 ++++++++++++++++
 net/ipv6/ndisc.c       | 23 ++++-------------------
 3 files changed, 28 insertions(+), 19 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/icmpv6.h b/include/linux/icmpv6.h
index 8f86d6b621c8..e4d4300d768f 100644
--- a/include/linux/icmpv6.h
+++ b/include/linux/icmpv6.h
@@ -182,6 +182,14 @@ extern int				icmpv6_err_convert(int type, int code,
 extern void				icmpv6_cleanup(void);
 extern void				icmpv6_param_prob(struct sk_buff *skb,
 							  int code, int pos);
+
+struct flowi;
+extern void				icmpv6_flow_init(struct sock *sk,
+							 struct flowi *fl,
+							 u8 type,
+							 const struct in6_addr *saddr,
+							 const struct in6_addr *daddr,
+							 int oif);
 #endif
 
 #endif
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index 12c0b85d6c46..cff74127ea32 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -777,6 +777,22 @@ drop_no_count:
 	return 0;
 }
 
+void icmpv6_flow_init(struct sock *sk, struct flowi *fl,
+		      u8 type,
+		      const struct in6_addr *saddr,
+		      const struct in6_addr *daddr,
+		      int oif)
+{
+	memset(fl, 0, sizeof(*fl));
+	ipv6_addr_copy(&fl->fl6_src, saddr);
+	ipv6_addr_copy(&fl->fl6_dst, daddr);
+	fl->proto	 	= IPPROTO_ICMPV6;
+	fl->fl_icmp_type	= type;
+	fl->fl_icmp_code	= 0;
+	fl->oif			= oif;
+	security_sk_classify_flow(sk, fl);
+}
+
 /*
  * Special lock-class for __icmpv6_sk:
  */
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 1fc33c8c7232..8db5f4a419aa 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -441,21 +441,6 @@ static void pndisc_destructor(struct pneigh_entry *n)
 /*
  *	Send a Neighbour Advertisement
  */
-
-static inline void ndisc_flow_init(struct flowi *fl, u8 type,
-			    struct in6_addr *saddr, struct in6_addr *daddr,
-			    int oif)
-{
-	memset(fl, 0, sizeof(*fl));
-	ipv6_addr_copy(&fl->fl6_src, saddr);
-	ipv6_addr_copy(&fl->fl6_dst, daddr);
-	fl->proto	 	= IPPROTO_ICMPV6;
-	fl->fl_icmp_type	= type;
-	fl->fl_icmp_code	= 0;
-	fl->oif			= oif;
-	security_sk_classify_flow(ndisc_socket->sk, fl);
-}
-
 static void __ndisc_send(struct net_device *dev,
 			 struct neighbour *neigh,
 			 struct in6_addr *daddr, struct in6_addr *saddr,
@@ -474,8 +459,8 @@ static void __ndisc_send(struct net_device *dev,
 
 	type = icmp6h->icmp6_type;
 
-	ndisc_flow_init(&fl, type, saddr, daddr,
-			dev->ifindex);
+	icmpv6_flow_init(ndisc_socket->sk, &fl, type,
+			 saddr, daddr, dev->ifindex);
 
 	dst = ndisc_dst_alloc(dev, neigh, daddr, ip6_output);
 	if (!dst)
@@ -1439,8 +1424,8 @@ void ndisc_send_redirect(struct sk_buff *skb, struct neighbour *neigh,
 		return;
 	}
 
-	ndisc_flow_init(&fl, NDISC_REDIRECT, &saddr_buf, &ipv6_hdr(skb)->saddr,
-			dev->ifindex);
+	icmpv6_flow_init(ndisc_socket->sk, &fl, NDISC_REDIRECT,
+			 &saddr_buf, &ipv6_hdr(skb)->saddr, dev->ifindex);
 
 	dst = ip6_route_output(NULL, &fl);
 	if (dst == NULL)
-- 
cgit v1.2.3


From 7a85f8896f4b4a4a0249563b92af9e3161a6b467 Mon Sep 17 00:00:00 2001
From: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Date: Tue, 4 Mar 2008 11:17:11 +0100
Subject: block: restore the meaning of rq->data_len to the true data length

The meaning of rq->data_len was changed to the length of an allocated
buffer from the true data length. It breaks SG_IO friends and
bsg. This patch restores the meaning of rq->data_len to the true data
length and adds rq->extra_len to store an extended length (due to
drain buffer and padding).

This patch also removes the code to update bio in blk_rq_map_user
introduced by the commit 40b01b9bbdf51ae543a04744283bf2d56c4a6afa.
The commit adjusts bio according to memory alignment
(queue_dma_alignment). However, memory alignment is NOT padding
alignment. This adjustment also breaks SG_IO friends and bsg. Padding
alignment needs to be fixed in a proper way (by a separate patch).

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <axboe@carl.home.kernel.dk>
---
 block/blk-core.c          | 3 +--
 block/blk-map.c           | 6 +-----
 block/blk-merge.c         | 2 +-
 block/bsg.c               | 8 ++++----
 block/scsi_ioctl.c        | 4 ++--
 drivers/ata/libata-scsi.c | 6 +++---
 include/linux/blkdev.h    | 2 +-
 7 files changed, 13 insertions(+), 18 deletions(-)

(limited to 'include/linux')

diff --git a/block/blk-core.c b/block/blk-core.c
index 2d7e3a2f56c4..a248cf1c98dd 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -127,7 +127,6 @@ void rq_init(struct request_queue *q, struct request *rq)
 	rq->nr_hw_segments = 0;
 	rq->ioprio = 0;
 	rq->special = NULL;
-	rq->raw_data_len = 0;
 	rq->buffer = NULL;
 	rq->tag = -1;
 	rq->errors = 0;
@@ -135,6 +134,7 @@ void rq_init(struct request_queue *q, struct request *rq)
 	rq->cmd_len = 0;
 	memset(rq->cmd, 0, sizeof(rq->cmd));
 	rq->data_len = 0;
+	rq->extra_len = 0;
 	rq->sense_len = 0;
 	rq->data = NULL;
 	rq->sense = NULL;
@@ -2018,7 +2018,6 @@ void blk_rq_bio_prep(struct request_queue *q, struct request *rq,
 	rq->hard_cur_sectors = rq->current_nr_sectors;
 	rq->hard_nr_sectors = rq->nr_sectors = bio_sectors(bio);
 	rq->buffer = bio_data(bio);
-	rq->raw_data_len = bio->bi_size;
 	rq->data_len = bio->bi_size;
 
 	rq->bio = rq->biotail = bio;
diff --git a/block/blk-map.c b/block/blk-map.c
index 09f7fd0bcb73..f5598322954d 100644
--- a/block/blk-map.c
+++ b/block/blk-map.c
@@ -19,7 +19,6 @@ int blk_rq_append_bio(struct request_queue *q, struct request *rq,
 		rq->biotail->bi_next = bio;
 		rq->biotail = bio;
 
-		rq->raw_data_len += bio->bi_size;
 		rq->data_len += bio->bi_size;
 	}
 	return 0;
@@ -151,11 +150,8 @@ int blk_rq_map_user(struct request_queue *q, struct request *rq,
 	 */
 	if (len & queue_dma_alignment(q)) {
 		unsigned int pad_len = (queue_dma_alignment(q) & ~len) + 1;
-		struct bio *bio = rq->biotail;
 
-		bio->bi_io_vec[bio->bi_vcnt - 1].bv_len += pad_len;
-		bio->bi_size += pad_len;
-		rq->data_len += pad_len;
+		rq->extra_len += pad_len;
 	}
 
 	rq->buffer = rq->data = NULL;
diff --git a/block/blk-merge.c b/block/blk-merge.c
index 7506c4fe0264..0f58616bcd7f 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -231,7 +231,7 @@ new_segment:
 			    ((unsigned long)q->dma_drain_buffer) &
 			    (PAGE_SIZE - 1));
 		nsegs++;
-		rq->data_len += q->dma_drain_size;
+		rq->extra_len += q->dma_drain_size;
 	}
 
 	if (sg)
diff --git a/block/bsg.c b/block/bsg.c
index 7f3c09549e4b..8917c5174dc2 100644
--- a/block/bsg.c
+++ b/block/bsg.c
@@ -437,14 +437,14 @@ static int blk_complete_sgv4_hdr_rq(struct request *rq, struct sg_io_v4 *hdr,
 	}
 
 	if (rq->next_rq) {
-		hdr->dout_resid = rq->raw_data_len;
-		hdr->din_resid = rq->next_rq->raw_data_len;
+		hdr->dout_resid = rq->data_len;
+		hdr->din_resid = rq->next_rq->data_len;
 		blk_rq_unmap_user(bidi_bio);
 		blk_put_request(rq->next_rq);
 	} else if (rq_data_dir(rq) == READ)
-		hdr->din_resid = rq->raw_data_len;
+		hdr->din_resid = rq->data_len;
 	else
-		hdr->dout_resid = rq->raw_data_len;
+		hdr->dout_resid = rq->data_len;
 
 	/*
 	 * If the request generated a negative error number, return it
diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c
index e993cac4911d..a2c3a936ebf9 100644
--- a/block/scsi_ioctl.c
+++ b/block/scsi_ioctl.c
@@ -266,7 +266,7 @@ static int blk_complete_sghdr_rq(struct request *rq, struct sg_io_hdr *hdr,
 	hdr->info = 0;
 	if (hdr->masked_status || hdr->host_status || hdr->driver_status)
 		hdr->info |= SG_INFO_CHECK;
-	hdr->resid = rq->raw_data_len;
+	hdr->resid = rq->data_len;
 	hdr->sb_len_wr = 0;
 
 	if (rq->sense_len && hdr->sbp) {
@@ -528,8 +528,8 @@ static int __blk_send_generic(struct request_queue *q, struct gendisk *bd_disk,
 	rq = blk_get_request(q, WRITE, __GFP_WAIT);
 	rq->cmd_type = REQ_TYPE_BLOCK_PC;
 	rq->data = NULL;
-	rq->raw_data_len = 0;
 	rq->data_len = 0;
+	rq->extra_len = 0;
 	rq->timeout = BLK_DEFAULT_SG_TIMEOUT;
 	memset(rq->cmd, 0, sizeof(rq->cmd));
 	rq->cmd[0] = cmd;
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 7b1f1ee8131d..fe47922dd69e 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -2538,7 +2538,7 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc)
 	}
 
 	qc->tf.command = ATA_CMD_PACKET;
-	qc->nbytes = scsi_bufflen(scmd);
+	qc->nbytes = scsi_bufflen(scmd) + scmd->request->extra_len;
 
 	/* check whether ATAPI DMA is safe */
 	if (!using_pio && ata_check_atapi_dma(qc))
@@ -2549,7 +2549,7 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc)
 	 * want to set it properly, and for DMA where it is
 	 * effectively meaningless.
 	 */
-	nbytes = min(scmd->request->raw_data_len, (unsigned int)63 * 1024);
+	nbytes = min(scmd->request->data_len, (unsigned int)63 * 1024);
 
 	/* Most ATAPI devices which honor transfer chunk size don't
 	 * behave according to the spec when odd chunk size which
@@ -2875,7 +2875,7 @@ static unsigned int ata_scsi_pass_thru(struct ata_queued_cmd *qc)
 	 * TODO: find out if we need to do more here to
 	 *       cover scatter/gather case.
 	 */
-	qc->nbytes = scsi_bufflen(scmd);
+	qc->nbytes = scsi_bufflen(scmd) + scmd->request->extra_len;
 
 	/* request result TF and be quiet about device error */
 	qc->flags |= ATA_QCFLAG_RESULT_TF | ATA_QCFLAG_QUIET;
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 6fe67d1939c2..b72526c13ca0 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -216,8 +216,8 @@ struct request {
 	unsigned int cmd_len;
 	unsigned char cmd[BLK_MAX_CDB];
 
-	unsigned int raw_data_len;
 	unsigned int data_len;
+	unsigned int extra_len;	/* length of alignment and padding */
 	unsigned int sense_len;
 	void *data;
 	void *sense;
-- 
cgit v1.2.3


From e3790c7d42a545e8fe8b38b513613ca96687b670 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 4 Mar 2008 11:18:17 +0100
Subject: block: separate out padding from alignment

Block layer alignment was used for two different purposes - memory
alignment and padding.  This causes problems in lower layers because
drivers which only require memory alignment ends up with adjusted
rq->data_len.  Separate out padding such that padding occurs iff
driver explicitly requests it.

Tomo: restorethe code to update bio in blk_rq_map_user
      introduced by the commit 40b01b9bbdf51ae543a04744283bf2d56c4a6afa
      according to padding alignment.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 block/blk-map.c           | 20 +++++++++++++-------
 block/blk-settings.c      | 17 +++++++++++++++++
 drivers/ata/libata-scsi.c |  3 ++-
 include/linux/blkdev.h    |  2 ++
 4 files changed, 34 insertions(+), 8 deletions(-)

(limited to 'include/linux')

diff --git a/block/blk-map.c b/block/blk-map.c
index f5598322954d..4e17dfd0035d 100644
--- a/block/blk-map.c
+++ b/block/blk-map.c
@@ -43,6 +43,7 @@ static int __blk_rq_map_user(struct request_queue *q, struct request *rq,
 			     void __user *ubuf, unsigned int len)
 {
 	unsigned long uaddr;
+	unsigned int alignment;
 	struct bio *bio, *orig_bio;
 	int reading, ret;
 
@@ -53,8 +54,8 @@ static int __blk_rq_map_user(struct request_queue *q, struct request *rq,
 	 * direct dma. else, set up kernel bounce buffers
 	 */
 	uaddr = (unsigned long) ubuf;
-	if (!(uaddr & queue_dma_alignment(q)) &&
-	    !(len & queue_dma_alignment(q)))
+	alignment = queue_dma_alignment(q) | q->dma_pad_mask;
+	if (!(uaddr & alignment) && !(len & alignment))
 		bio = bio_map_user(q, NULL, uaddr, len, reading);
 	else
 		bio = bio_copy_user(q, uaddr, len, reading);
@@ -141,15 +142,20 @@ int blk_rq_map_user(struct request_queue *q, struct request *rq,
 
 	/*
 	 * __blk_rq_map_user() copies the buffers if starting address
-	 * or length isn't aligned.  As the copied buffer is always
-	 * page aligned, we know that there's enough room for padding.
-	 * Extend the last bio and update rq->data_len accordingly.
+	 * or length isn't aligned to dma_pad_mask.  As the copied
+	 * buffer is always page aligned, we know that there's enough
+	 * room for padding.  Extend the last bio and update
+	 * rq->data_len accordingly.
 	 *
 	 * On unmap, bio_uncopy_user() will use unmodified
 	 * bio_map_data pointed to by bio->bi_private.
 	 */
-	if (len & queue_dma_alignment(q)) {
-		unsigned int pad_len = (queue_dma_alignment(q) & ~len) + 1;
+	if (len & q->dma_pad_mask) {
+		unsigned int pad_len = (q->dma_pad_mask & ~len) + 1;
+		struct bio *bio = rq->biotail;
+
+		bio->bi_io_vec[bio->bi_vcnt - 1].bv_len += pad_len;
+		bio->bi_size += pad_len;
 
 		rq->extra_len += pad_len;
 	}
diff --git a/block/blk-settings.c b/block/blk-settings.c
index da923fed1f2c..a9f37f530b15 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -292,6 +292,23 @@ void blk_queue_stack_limits(struct request_queue *t, struct request_queue *b)
 }
 EXPORT_SYMBOL(blk_queue_stack_limits);
 
+/**
+ * blk_queue_dma_pad - set pad mask
+ * @q:     the request queue for the device
+ * @mask:  pad mask
+ *
+ * Set pad mask.  Direct IO requests are padded to the mask specified.
+ *
+ * Appending pad buffer to a request modifies ->data_len such that it
+ * includes the pad buffer.  The original requested data length can be
+ * obtained using blk_rq_raw_data_len().
+ **/
+void blk_queue_dma_pad(struct request_queue *q, unsigned int mask)
+{
+	q->dma_pad_mask = mask;
+}
+EXPORT_SYMBOL(blk_queue_dma_pad);
+
 /**
  * blk_queue_dma_drain - Set up a drain buffer for excess dma.
  * @q:  the request queue for the device
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index fe47922dd69e..8f0e8f2bc628 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -862,9 +862,10 @@ static int ata_scsi_dev_config(struct scsi_device *sdev,
 		struct request_queue *q = sdev->request_queue;
 		void *buf;
 
-		/* set the min alignment */
+		/* set the min alignment and padding */
 		blk_queue_update_dma_alignment(sdev->request_queue,
 					       ATA_DMA_PAD_SZ - 1);
+		blk_queue_dma_pad(sdev->request_queue, ATA_DMA_PAD_SZ - 1);
 
 		/* configure draining */
 		buf = kmalloc(ATAPI_MAX_DRAIN, q->bounce_gfp | GFP_KERNEL);
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index b72526c13ca0..6f79d40dd3c0 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -362,6 +362,7 @@ struct request_queue
 	unsigned long		seg_boundary_mask;
 	void			*dma_drain_buffer;
 	unsigned int		dma_drain_size;
+	unsigned int		dma_pad_mask;
 	unsigned int		dma_alignment;
 
 	struct blk_queue_tag	*queue_tags;
@@ -701,6 +702,7 @@ extern void blk_queue_max_hw_segments(struct request_queue *, unsigned short);
 extern void blk_queue_max_segment_size(struct request_queue *, unsigned int);
 extern void blk_queue_hardsect_size(struct request_queue *, unsigned short);
 extern void blk_queue_stack_limits(struct request_queue *t, struct request_queue *b);
+extern void blk_queue_dma_pad(struct request_queue *, unsigned int);
 extern int blk_queue_dma_drain(struct request_queue *q,
 			       dma_drain_needed_fn *dma_drain_needed,
 			       void *buf, unsigned int size);
-- 
cgit v1.2.3


From 1826eadfc42839af7c1c5a1859510aff635d3fa1 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Tue, 4 Mar 2008 11:23:46 +0100
Subject: block/genhd.c: cleanups

This patch contains the following cleanups:
- make the needlessly global struct disk_type static
- #if 0 the unused genhd_media_change_notify()

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 block/genhd.c         | 6 +++++-
 include/linux/genhd.h | 2 --
 2 files changed, 5 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/block/genhd.c b/block/genhd.c
index abc6feddc8c6..c44527d16c52 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -24,6 +24,8 @@ static DEFINE_MUTEX(block_class_lock);
 struct kobject *block_depr;
 #endif
 
+static struct device_type disk_type;
+
 /*
  * Can be deleted altogether. Later.
  *
@@ -502,7 +504,7 @@ struct class block_class = {
 	.name		= "block",
 };
 
-struct device_type disk_type = {
+static struct device_type disk_type = {
 	.name		= "disk",
 	.groups		= disk_attr_groups,
 	.release	= disk_release,
@@ -632,12 +634,14 @@ static void media_change_notify_thread(struct work_struct *work)
 	put_device(gd->driverfs_dev);
 }
 
+#if 0
 void genhd_media_change_notify(struct gendisk *disk)
 {
 	get_device(disk->driverfs_dev);
 	schedule_work(&disk->async_notify);
 }
 EXPORT_SYMBOL_GPL(genhd_media_change_notify);
+#endif  /*  0  */
 
 dev_t blk_lookup_devt(const char *name)
 {
diff --git a/include/linux/genhd.h b/include/linux/genhd.h
index 09a3b18918c7..cd048e3cc96d 100644
--- a/include/linux/genhd.h
+++ b/include/linux/genhd.h
@@ -18,7 +18,6 @@
 #define dev_to_disk(device) container_of(device, struct gendisk, dev)
 #define dev_to_part(device) container_of(device, struct hd_struct, dev)
 
-extern struct device_type disk_type;
 extern struct device_type part_type;
 extern struct kobject *block_depr;
 extern struct class block_class;
@@ -556,7 +555,6 @@ extern struct gendisk *alloc_disk_node(int minors, int node_id);
 extern struct gendisk *alloc_disk(int minors);
 extern struct kobject *get_disk(struct gendisk *disk);
 extern void put_disk(struct gendisk *disk);
-extern void genhd_media_change_notify(struct gendisk *disk);
 extern void blk_register_region(dev_t devt, unsigned long range,
 			struct module *module,
 			struct kobject *(*probe)(dev_t, int *, void *),
-- 
cgit v1.2.3


From a0db701a6bf767320e4471bd55e70702d230f6fb Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Tue, 4 Mar 2008 11:23:50 +0100
Subject: block/genhd.c: proper externs

This patch adds proper externs for two structs in include/linux/genhd.h

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 fs/proc/proc_misc.c   | 3 +--
 include/linux/genhd.h | 3 +++
 2 files changed, 4 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/fs/proc/proc_misc.c b/fs/proc/proc_misc.c
index 468805d40e2b..2d563979cb02 100644
--- a/fs/proc/proc_misc.c
+++ b/fs/proc/proc_misc.c
@@ -32,6 +32,7 @@
 #include <linux/interrupt.h>
 #include <linux/swap.h>
 #include <linux/slab.h>
+#include <linux/genhd.h>
 #include <linux/smp.h>
 #include <linux/signal.h>
 #include <linux/module.h>
@@ -377,7 +378,6 @@ static int stram_read_proc(char *page, char **start, off_t off,
 #endif
 
 #ifdef CONFIG_BLOCK
-extern const struct seq_operations partitions_op;
 static int partitions_open(struct inode *inode, struct file *file)
 {
 	return seq_open(file, &partitions_op);
@@ -389,7 +389,6 @@ static const struct file_operations proc_partitions_operations = {
 	.release	= seq_release,
 };
 
-extern const struct seq_operations diskstats_op;
 static int diskstats_open(struct inode *inode, struct file *file)
 {
 	return seq_open(file, &diskstats_op);
diff --git a/include/linux/genhd.h b/include/linux/genhd.h
index cd048e3cc96d..32c2ac49a070 100644
--- a/include/linux/genhd.h
+++ b/include/linux/genhd.h
@@ -22,6 +22,9 @@ extern struct device_type part_type;
 extern struct kobject *block_depr;
 extern struct class block_class;
 
+extern const struct seq_operations partitions_op;
+extern const struct seq_operations diskstats_op;
+
 enum {
 /* These three have identical behaviour; use the second one if DOS FDISK gets
    confused about extended/logical partitions starting past cylinder 1023. */
-- 
cgit v1.2.3


From 2a341f5cf57dce9d89b41484a69e88adc6422f6c Mon Sep 17 00:00:00 2001
From: David Brownell <david-b@pacbell.net>
Date: Fri, 22 Feb 2008 17:23:23 -0800
Subject: atmel_tc library

Create <linux/atmel_tc.h> based on <asm-arm/arch-at91/at91-tc.h> and the
at91sam9263 and at32ap7000 datasheets.  Most AT91 and AT32 SOCs have one
or two of these TC blocks, which include three 16-bit timers that can be
interconnected in various ways.

These TC blocks can be used for external interfacing (such as PWM and
measurement), or used as somewhat quirky sixteen-bit timers.

Changes relative to the original version:
  * Drop unneeded inclusion of <linux/mutex.h>
  * Support an arbitrary number of TC blocks
  * Return a struct with information about a TC block from
    atmel_tc_alloc() instead of using a combination of return values
    and "out" parameters.
  * ioremap() the I/O registers on allocation
  * Look up clocks and irqs for all channels
  * Add "name" parameter to atmel_tc_alloc() and use this when
    requesting the iomem resource.
  * Check if the platform provided the necessary resources at probe()
    time instead of when the TCB is allocated.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
---
 drivers/misc/Kconfig       |   8 ++
 drivers/misc/Makefile      |   1 +
 drivers/misc/atmel_tclib.c | 161 +++++++++++++++++++++++++++++
 include/linux/atmel_tc.h   | 252 +++++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 422 insertions(+)
 create mode 100644 drivers/misc/atmel_tclib.c
 create mode 100644 include/linux/atmel_tc.h

(limited to 'include/linux')

diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig
index 982e27b86d10..b3ba68170b81 100644
--- a/drivers/misc/Kconfig
+++ b/drivers/misc/Kconfig
@@ -22,6 +22,14 @@ config ATMEL_PWM
 	  purposes including software controlled power-efficent backlights
 	  on LCD displays, motor control, and waveform generation.
 
+config ATMEL_TCLIB
+	bool "Atmel AT32/AT91 Timer/Counter Library"
+	depends on (AVR32 || ARCH_AT91)
+	help
+	  Select this if you want a library to allocate the Timer/Counter
+	  blocks found on many Atmel processors.  This facilitates using
+	  these blocks by different drivers despite processor differences.
+
 config IBM_ASM
 	tristate "Device driver for IBM RSA service processor"
 	depends on X86 && PCI && INPUT && EXPERIMENTAL
diff --git a/drivers/misc/Makefile b/drivers/misc/Makefile
index 3b12f5da8562..c975028f101b 100644
--- a/drivers/misc/Makefile
+++ b/drivers/misc/Makefile
@@ -10,6 +10,7 @@ obj-$(CONFIG_ACER_WMI)     += acer-wmi.o
 obj-$(CONFIG_ASUS_LAPTOP)     += asus-laptop.o
 obj-$(CONFIG_ATMEL_PWM)		+= atmel_pwm.o
 obj-$(CONFIG_ATMEL_SSC)		+= atmel-ssc.o
+obj-$(CONFIG_ATMEL_TCLIB)	+= atmel_tclib.o
 obj-$(CONFIG_TC1100_WMI)	+= tc1100-wmi.o
 obj-$(CONFIG_LKDTM)		+= lkdtm.o
 obj-$(CONFIG_TIFM_CORE)       	+= tifm_core.o
diff --git a/drivers/misc/atmel_tclib.c b/drivers/misc/atmel_tclib.c
new file mode 100644
index 000000000000..05dc8a31f280
--- /dev/null
+++ b/drivers/misc/atmel_tclib.c
@@ -0,0 +1,161 @@
+#include <linux/atmel_tc.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/ioport.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+
+/* Number of bytes to reserve for the iomem resource */
+#define ATMEL_TC_IOMEM_SIZE	256
+
+
+/*
+ * This is a thin library to solve the problem of how to portably allocate
+ * one of the TC blocks.  For simplicity, it doesn't currently expect to
+ * share individual timers between different drivers.
+ */
+
+#if defined(CONFIG_AVR32)
+/* AVR32 has these divide PBB */
+const u8 atmel_tc_divisors[5] = { 0, 4, 8, 16, 32, };
+EXPORT_SYMBOL(atmel_tc_divisors);
+
+#elif defined(CONFIG_ARCH_AT91)
+/* AT91 has these divide MCK */
+const u8 atmel_tc_divisors[5] = { 2, 8, 32, 128, 0, };
+EXPORT_SYMBOL(atmel_tc_divisors);
+
+#endif
+
+static DEFINE_SPINLOCK(tc_list_lock);
+static LIST_HEAD(tc_list);
+
+/**
+ * atmel_tc_alloc - allocate a specified TC block
+ * @block: which block to allocate
+ * @name: name to be associated with the iomem resource
+ *
+ * Caller allocates a block.  If it is available, a pointer to a
+ * pre-initialized struct atmel_tc is returned. The caller can access
+ * the registers directly through the "regs" field.
+ */
+struct atmel_tc *atmel_tc_alloc(unsigned block, const char *name)
+{
+	struct atmel_tc		*tc;
+	struct platform_device	*pdev = NULL;
+	struct resource		*r;
+
+	spin_lock(&tc_list_lock);
+	list_for_each_entry(tc, &tc_list, node) {
+		if (tc->pdev->id == block) {
+			pdev = tc->pdev;
+			break;
+		}
+	}
+
+	if (!pdev || tc->iomem)
+		goto fail;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	r = request_mem_region(r->start, ATMEL_TC_IOMEM_SIZE, name);
+	if (!r)
+		goto fail;
+
+	tc->regs = ioremap(r->start, ATMEL_TC_IOMEM_SIZE);
+	if (!tc->regs)
+		goto fail_ioremap;
+
+	tc->iomem = r;
+
+out:
+	spin_unlock(&tc_list_lock);
+	return tc;
+
+fail_ioremap:
+	release_resource(r);
+fail:
+	tc = NULL;
+	goto out;
+}
+EXPORT_SYMBOL_GPL(atmel_tc_alloc);
+
+/**
+ * atmel_tc_free - release a specified TC block
+ * @tc: Timer/counter block that was returned by atmel_tc_alloc()
+ *
+ * This reverses the effect of atmel_tc_alloc(), unmapping the I/O
+ * registers, invalidating the resource returned by that routine and
+ * making the TC available to other drivers.
+ */
+void atmel_tc_free(struct atmel_tc *tc)
+{
+	spin_lock(&tc_list_lock);
+	if (tc->regs) {
+		iounmap(tc->regs);
+		release_resource(tc->iomem);
+		tc->regs = NULL;
+		tc->iomem = NULL;
+	}
+	spin_unlock(&tc_list_lock);
+}
+EXPORT_SYMBOL_GPL(atmel_tc_free);
+
+static int __init tc_probe(struct platform_device *pdev)
+{
+	struct atmel_tc *tc;
+	struct clk	*clk;
+	int		irq;
+
+	if (!platform_get_resource(pdev, IORESOURCE_MEM, 0))
+		return -EINVAL;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return -EINVAL;
+
+	tc = kzalloc(sizeof(struct atmel_tc), GFP_KERNEL);
+	if (!tc)
+		return -ENOMEM;
+
+	tc->pdev = pdev;
+
+	clk = clk_get(&pdev->dev, "t0_clk");
+	if (IS_ERR(clk)) {
+		kfree(tc);
+		return -EINVAL;
+	}
+
+	tc->clk[0] = clk;
+	tc->clk[1] = clk_get(&pdev->dev, "t1_clk");
+	if (IS_ERR(tc->clk[1]))
+		tc->clk[1] = clk;
+	tc->clk[2] = clk_get(&pdev->dev, "t2_clk");
+	if (IS_ERR(tc->clk[2]))
+		tc->clk[2] = clk;
+
+	tc->irq[0] = irq;
+	tc->irq[1] = platform_get_irq(pdev, 1);
+	if (tc->irq[1] < 0)
+		tc->irq[1] = irq;
+	tc->irq[2] = platform_get_irq(pdev, 2);
+	if (tc->irq[2] < 0)
+		tc->irq[2] = irq;
+
+	spin_lock(&tc_list_lock);
+	list_add_tail(&tc->node, &tc_list);
+	spin_unlock(&tc_list_lock);
+
+	return 0;
+}
+
+static struct platform_driver tc_driver = {
+	.driver.name	= "atmel_tcb",
+};
+
+static int __init tc_init(void)
+{
+	return platform_driver_probe(&tc_driver, tc_probe);
+}
+arch_initcall(tc_init);
diff --git a/include/linux/atmel_tc.h b/include/linux/atmel_tc.h
new file mode 100644
index 000000000000..53ba65e30caa
--- /dev/null
+++ b/include/linux/atmel_tc.h
@@ -0,0 +1,252 @@
+/*
+ * Timer/Counter Unit (TC) registers.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#ifndef ATMEL_TC_H
+#define ATMEL_TC_H
+
+#include <linux/compiler.h>
+#include <linux/list.h>
+
+/*
+ * Many 32-bit Atmel SOCs include one or more TC blocks, each of which holds
+ * three general-purpose 16-bit timers.  These timers share one register bank.
+ * Depending on the SOC, each timer may have its own clock and IRQ, or those
+ * may be shared by the whole TC block.
+ *
+ * These TC blocks may have up to nine external pins:  TCLK0..2 signals for
+ * clocks or clock gates, and per-timer TIOA and TIOB signals used for PWM
+ * or triggering.  Those pins need to be set up for use with the TC block,
+ * else they will be used as GPIOs or for a different controller.
+ *
+ * Although we expect each TC block to have a platform_device node, those
+ * nodes are not what drivers bind to.  Instead, they ask for a specific
+ * TC block, by number ... which is a common approach on systems with many
+ * timers.  Then they use clk_get() and platform_get_irq() to get clock and
+ * IRQ resources.
+ */
+
+struct clk;
+
+/**
+ * struct atmel_tc - information about a Timer/Counter Block
+ * @pdev: physical device
+ * @iomem: resource associated with the I/O register
+ * @regs: mapping through which the I/O registers can be accessed
+ * @irq: irq for each of the three channels
+ * @clk: internal clock source for each of the three channels
+ * @node: list node, for tclib internal use
+ *
+ * On some platforms, each TC channel has its own clocks and IRQs,
+ * while on others, all TC channels share the same clock and IRQ.
+ * Drivers should clk_enable() all the clocks they need even though
+ * all the entries in @clk may point to the same physical clock.
+ * Likewise, drivers should request irqs independently for each
+ * channel, but they must use IRQF_SHARED in case some of the entries
+ * in @irq are actually the same IRQ.
+ */
+struct atmel_tc {
+	struct platform_device	*pdev;
+	struct resource		*iomem;
+	void __iomem		*regs;
+	int			irq[3];
+	struct clk		*clk[3];
+	struct list_head	node;
+};
+
+extern struct atmel_tc *atmel_tc_alloc(unsigned block, const char *name);
+extern void atmel_tc_free(struct atmel_tc *tc);
+
+/* platform-specific ATMEL_TC_TIMER_CLOCKx divisors (0 means 32KiHz) */
+extern const u8 atmel_tc_divisors[5];
+
+
+/*
+ * Two registers have block-wide controls.  These are: configuring the three
+ * "external" clocks (or event sources) used by the timer channels; and
+ * synchronizing the timers by resetting them all at once.
+ *
+ * "External" can mean "external to chip" using the TCLK0, TCLK1, or TCLK2
+ * signals.  Or, it can mean "external to timer", using the TIOA output from
+ * one of the other two timers that's being run in waveform mode.
+ */
+
+#define ATMEL_TC_BCR	0xc0		/* TC Block Control Register */
+#define     ATMEL_TC_SYNC	(1 << 0)	/* synchronize timers */
+
+#define ATMEL_TC_BMR	0xc4		/* TC Block Mode Register */
+#define     ATMEL_TC_TC0XC0S	(3 << 0)	/* external clock 0 source */
+#define        ATMEL_TC_TC0XC0S_TCLK0	(0 << 0)
+#define        ATMEL_TC_TC0XC0S_NONE	(1 << 0)
+#define        ATMEL_TC_TC0XC0S_TIOA1	(2 << 0)
+#define        ATMEL_TC_TC0XC0S_TIOA2	(3 << 0)
+#define     ATMEL_TC_TC1XC1S	(3 << 2)	/* external clock 1 source */
+#define        ATMEL_TC_TC1XC1S_TCLK1	(0 << 2)
+#define        ATMEL_TC_TC1XC1S_NONE	(1 << 2)
+#define        ATMEL_TC_TC1XC1S_TIOA0	(2 << 2)
+#define        ATMEL_TC_TC1XC1S_TIOA2	(3 << 2)
+#define     ATMEL_TC_TC2XC2S	(3 << 4)	/* external clock 2 source */
+#define        ATMEL_TC_TC2XC2S_TCLK2	(0 << 4)
+#define        ATMEL_TC_TC2XC2S_NONE	(1 << 4)
+#define        ATMEL_TC_TC2XC2S_TIOA0	(2 << 4)
+#define        ATMEL_TC_TC2XC2S_TIOA1	(3 << 4)
+
+
+/*
+ * Each TC block has three "channels", each with one counter and controls.
+ *
+ * Note that the semantics of ATMEL_TC_TIMER_CLOCKx (input clock selection
+ * when it's not "external") is silicon-specific.  AT91 platforms use one
+ * set of definitions; AVR32 platforms use a different set.  Don't hard-wire
+ * such knowledge into your code, use the global "atmel_tc_divisors" ...
+ * where index N is the divisor for clock N+1, else zero to indicate it uses
+ * the 32 KiHz clock.
+ *
+ * The timers can be chained in various ways, and operated in "waveform"
+ * generation mode (including PWM) or "capture" mode (to time events).  In
+ * both modes, behavior can be configured in many ways.
+ *
+ * Each timer has two I/O pins, TIOA and TIOB.  Waveform mode uses TIOA as a
+ * PWM output, and TIOB as either another PWM or as a trigger.  Capture mode
+ * uses them only as inputs.
+ */
+#define ATMEL_TC_CHAN(idx)	((idx)*0x40)
+#define ATMEL_TC_REG(idx, reg)	(ATMEL_TC_CHAN(idx) + ATMEL_TC_ ## reg)
+
+#define ATMEL_TC_CCR	0x00		/* Channel Control Register */
+#define     ATMEL_TC_CLKEN	(1 << 0)	/* clock enable */
+#define     ATMEL_TC_CLKDIS	(1 << 1)	/* clock disable */
+#define     ATMEL_TC_SWTRG	(1 << 2)	/* software trigger */
+
+#define ATMEL_TC_CMR	0x04		/* Channel Mode Register */
+
+/* Both modes share some CMR bits */
+#define     ATMEL_TC_TCCLKS	(7 << 0)	/* clock source */
+#define        ATMEL_TC_TIMER_CLOCK1	(0 << 0)
+#define        ATMEL_TC_TIMER_CLOCK2	(1 << 0)
+#define        ATMEL_TC_TIMER_CLOCK3	(2 << 0)
+#define        ATMEL_TC_TIMER_CLOCK4	(3 << 0)
+#define        ATMEL_TC_TIMER_CLOCK5	(4 << 0)
+#define        ATMEL_TC_XC0		(5 << 0)
+#define        ATMEL_TC_XC1		(6 << 0)
+#define        ATMEL_TC_XC2		(7 << 0)
+#define     ATMEL_TC_CLKI	(1 << 3)	/* clock invert */
+#define     ATMEL_TC_BURST	(3 << 4)	/* clock gating */
+#define        ATMEL_TC_GATE_NONE	(0 << 4)
+#define        ATMEL_TC_GATE_XC0	(1 << 4)
+#define        ATMEL_TC_GATE_XC1	(2 << 4)
+#define        ATMEL_TC_GATE_XC2	(3 << 4)
+#define     ATMEL_TC_WAVE	(1 << 15)	/* true = Waveform mode */
+
+/* CAPTURE mode CMR bits */
+#define     ATMEL_TC_LDBSTOP	(1 << 6)	/* counter stops on RB load */
+#define     ATMEL_TC_LDBDIS	(1 << 7)	/* counter disable on RB load */
+#define     ATMEL_TC_ETRGEDG	(3 << 8)	/* external trigger edge */
+#define        ATMEL_TC_ETRGEDG_NONE	(0 << 8)
+#define        ATMEL_TC_ETRGEDG_RISING	(1 << 8)
+#define        ATMEL_TC_ETRGEDG_FALLING	(2 << 8)
+#define        ATMEL_TC_ETRGEDG_BOTH	(3 << 8)
+#define     ATMEL_TC_ABETRG	(1 << 10)	/* external trigger is TIOA? */
+#define     ATMEL_TC_CPCTRG	(1 << 14)	/* RC compare trigger enable */
+#define     ATMEL_TC_LDRA	(3 << 16)	/* RA loading edge (of TIOA) */
+#define        ATMEL_TC_LDRA_NONE	(0 << 16)
+#define        ATMEL_TC_LDRA_RISING	(1 << 16)
+#define        ATMEL_TC_LDRA_FALLING	(2 << 16)
+#define        ATMEL_TC_LDRA_BOTH	(3 << 16)
+#define     ATMEL_TC_LDRB	(3 << 18)	/* RB loading edge (of TIOA) */
+#define        ATMEL_TC_LDRB_NONE	(0 << 18)
+#define        ATMEL_TC_LDRB_RISING	(1 << 18)
+#define        ATMEL_TC_LDRB_FALLING	(2 << 18)
+#define        ATMEL_TC_LDRB_BOTH	(3 << 18)
+
+/* WAVEFORM mode CMR bits */
+#define     ATMEL_TC_CPCSTOP	(1 <<  6)	/* RC compare stops counter */
+#define     ATMEL_TC_CPCDIS	(1 <<  7)	/* RC compare disables counter */
+#define     ATMEL_TC_EEVTEDG	(3 <<  8)	/* external event edge */
+#define        ATMEL_TC_EEVTEDG_NONE	(0 << 8)
+#define        ATMEL_TC_EEVTEDG_RISING	(1 << 8)
+#define        ATMEL_TC_EEVTEDG_FALLING	(2 << 8)
+#define        ATMEL_TC_EEVTEDG_BOTH	(3 << 8)
+#define     ATMEL_TC_EEVT	(3 << 10)	/* external event source */
+#define        ATMEL_TC_EEVT_TIOB	(0 << 10)
+#define        ATMEL_TC_EEVT_XC0	(1 << 10)
+#define        ATMEL_TC_EEVT_XC1	(2 << 10)
+#define        ATMEL_TC_EEVT_XC2	(3 << 10)
+#define     ATMEL_TC_ENETRG	(1 << 12)	/* external event is trigger */
+#define     ATMEL_TC_WAVESEL	(3 << 13)	/* waveform type */
+#define        ATMEL_TC_WAVESEL_UP	(0 << 13)
+#define        ATMEL_TC_WAVESEL_UPDOWN	(1 << 13)
+#define        ATMEL_TC_WAVESEL_UP_AUTO	(2 << 13)
+#define        ATMEL_TC_WAVESEL_UPDOWN_AUTO (3 << 13)
+#define     ATMEL_TC_ACPA	(3 << 16)	/* RA compare changes TIOA */
+#define        ATMEL_TC_ACPA_NONE	(0 << 16)
+#define        ATMEL_TC_ACPA_SET	(1 << 16)
+#define        ATMEL_TC_ACPA_CLEAR	(2 << 16)
+#define        ATMEL_TC_ACPA_TOGGLE	(3 << 16)
+#define     ATMEL_TC_ACPC	(3 << 18)	/* RC compare changes TIOA */
+#define        ATMEL_TC_ACPC_NONE	(0 << 18)
+#define        ATMEL_TC_ACPC_SET	(1 << 18)
+#define        ATMEL_TC_ACPC_CLEAR	(2 << 18)
+#define        ATMEL_TC_ACPC_TOGGLE	(3 << 18)
+#define     ATMEL_TC_AEEVT	(3 << 20)	/* external event changes TIOA */
+#define        ATMEL_TC_AEEVT_NONE	(0 << 20)
+#define        ATMEL_TC_AEEVT_SET	(1 << 20)
+#define        ATMEL_TC_AEEVT_CLEAR	(2 << 20)
+#define        ATMEL_TC_AEEVT_TOGGLE	(3 << 20)
+#define     ATMEL_TC_ASWTRG	(3 << 22)	/* software trigger changes TIOA */
+#define        ATMEL_TC_ASWTRG_NONE	(0 << 22)
+#define        ATMEL_TC_ASWTRG_SET	(1 << 22)
+#define        ATMEL_TC_ASWTRG_CLEAR	(2 << 22)
+#define        ATMEL_TC_ASWTRG_TOGGLE	(3 << 22)
+#define     ATMEL_TC_BCPB	(3 << 24)	/* RB compare changes TIOB */
+#define        ATMEL_TC_BCPB_NONE	(0 << 24)
+#define        ATMEL_TC_BCPB_SET	(1 << 24)
+#define        ATMEL_TC_BCPB_CLEAR	(2 << 24)
+#define        ATMEL_TC_BCPB_TOGGLE	(3 << 24)
+#define     ATMEL_TC_BCPC	(3 << 26)	/* RC compare changes TIOB */
+#define        ATMEL_TC_BCPC_NONE	(0 << 26)
+#define        ATMEL_TC_BCPC_SET	(1 << 26)
+#define        ATMEL_TC_BCPC_CLEAR	(2 << 26)
+#define        ATMEL_TC_BCPC_TOGGLE	(3 << 26)
+#define     ATMEL_TC_BEEVT	(3 << 28)	/* external event changes TIOB */
+#define        ATMEL_TC_BEEVT_NONE	(0 << 28)
+#define        ATMEL_TC_BEEVT_SET	(1 << 28)
+#define        ATMEL_TC_BEEVT_CLEAR	(2 << 28)
+#define        ATMEL_TC_BEEVT_TOGGLE	(3 << 28)
+#define     ATMEL_TC_BSWTRG	(3 << 30)	/* software trigger changes TIOB */
+#define        ATMEL_TC_BSWTRG_NONE	(0 << 30)
+#define        ATMEL_TC_BSWTRG_SET	(1 << 30)
+#define        ATMEL_TC_BSWTRG_CLEAR	(2 << 30)
+#define        ATMEL_TC_BSWTRG_TOGGLE	(3 << 30)
+
+#define ATMEL_TC_CV	0x10		/* counter Value */
+#define ATMEL_TC_RA	0x14		/* register A */
+#define ATMEL_TC_RB	0x18		/* register B */
+#define ATMEL_TC_RC	0x1c		/* register C */
+
+#define ATMEL_TC_SR	0x20		/* status (read-only) */
+/* Status-only flags */
+#define     ATMEL_TC_CLKSTA	(1 << 16)	/* clock enabled */
+#define     ATMEL_TC_MTIOA	(1 << 17)	/* TIOA mirror */
+#define     ATMEL_TC_MTIOB	(1 << 18)	/* TIOB mirror */
+
+#define ATMEL_TC_IER	0x24		/* interrupt enable (write-only) */
+#define ATMEL_TC_IDR	0x28		/* interrupt disable (write-only) */
+#define ATMEL_TC_IMR	0x2c		/* interrupt mask (read-only) */
+
+/* Status and IRQ flags */
+#define     ATMEL_TC_COVFS	(1 <<  0)	/* counter overflow */
+#define     ATMEL_TC_LOVRS	(1 <<  1)	/* load overrun */
+#define     ATMEL_TC_CPAS	(1 <<  2)	/* RA compare */
+#define     ATMEL_TC_CPBS	(1 <<  3)	/* RB compare */
+#define     ATMEL_TC_CPCS	(1 <<  4)	/* RC compare */
+#define     ATMEL_TC_LDRAS	(1 <<  5)	/* RA loading */
+#define     ATMEL_TC_LDRBS	(1 <<  6)	/* RB loading */
+#define     ATMEL_TC_ETRGS	(1 <<  7)	/* external trigger */
+
+#endif
-- 
cgit v1.2.3


From 72dc67a69690288538142df73a7e3ac66fea68dc Mon Sep 17 00:00:00 2001
From: Izik Eidus <izike@qumranet.com>
Date: Sun, 10 Feb 2008 18:04:15 +0200
Subject: KVM: remove the usage of the mmap_sem for the protection of the
 memory slots.

This patch replaces the mmap_sem lock for the memory slots with a new
kvm private lock, it is needed beacuse untill now there were cases where
kvm accesses user memory while holding the mmap semaphore.

Signed-off-by: Izik Eidus <izike@qumranet.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
---
 arch/x86/kvm/mmu.c         | 24 ++++++++++++++---
 arch/x86/kvm/paging_tmpl.h | 13 +++++++---
 arch/x86/kvm/vmx.c         |  7 +++--
 arch/x86/kvm/x86.c         | 65 ++++++++++++++++++++++++++--------------------
 include/linux/kvm_host.h   |  1 +
 virt/kvm/kvm_main.c        |  5 ++--
 6 files changed, 75 insertions(+), 40 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 8efdcdbebb03..26037106ad19 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -876,11 +876,18 @@ static void page_header_update_slot(struct kvm *kvm, void *pte, gfn_t gfn)
 
 struct page *gva_to_page(struct kvm_vcpu *vcpu, gva_t gva)
 {
+	struct page *page;
+
 	gpa_t gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, gva);
 
 	if (gpa == UNMAPPED_GVA)
 		return NULL;
-	return gfn_to_page(vcpu->kvm, gpa >> PAGE_SHIFT);
+
+	down_read(&current->mm->mmap_sem);
+	page = gfn_to_page(vcpu->kvm, gpa >> PAGE_SHIFT);
+	up_read(&current->mm->mmap_sem);
+
+	return page;
 }
 
 static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
@@ -1020,15 +1027,18 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn)
 
 	struct page *page;
 
+	down_read(&vcpu->kvm->slots_lock);
+
 	down_read(&current->mm->mmap_sem);
 	page = gfn_to_page(vcpu->kvm, gfn);
+	up_read(&current->mm->mmap_sem);
 
 	spin_lock(&vcpu->kvm->mmu_lock);
 	kvm_mmu_free_some_pages(vcpu);
 	r = __nonpaging_map(vcpu, v, write, gfn, page);
 	spin_unlock(&vcpu->kvm->mmu_lock);
 
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 
 	return r;
 }
@@ -1362,6 +1372,7 @@ static void mmu_guess_page_from_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	gfn_t gfn;
 	int r;
 	u64 gpte = 0;
+	struct page *page;
 
 	if (bytes != 4 && bytes != 8)
 		return;
@@ -1389,6 +1400,11 @@ static void mmu_guess_page_from_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	if (!is_present_pte(gpte))
 		return;
 	gfn = (gpte & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT;
+
+	down_read(&current->mm->mmap_sem);
+	page = gfn_to_page(vcpu->kvm, gfn);
+	up_read(&current->mm->mmap_sem);
+
 	vcpu->arch.update_pte.gfn = gfn;
 	vcpu->arch.update_pte.page = gfn_to_page(vcpu->kvm, gfn);
 }
@@ -1496,9 +1512,9 @@ int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva)
 	gpa_t gpa;
 	int r;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, gva);
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 
 	spin_lock(&vcpu->kvm->mmu_lock);
 	r = kvm_mmu_unprotect_page(vcpu->kvm, gpa >> PAGE_SHIFT);
diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
index 03ba8608fe0f..2009c6e9dc4d 100644
--- a/arch/x86/kvm/paging_tmpl.h
+++ b/arch/x86/kvm/paging_tmpl.h
@@ -91,7 +91,10 @@ static bool FNAME(cmpxchg_gpte)(struct kvm *kvm,
 	pt_element_t *table;
 	struct page *page;
 
+	down_read(&current->mm->mmap_sem);
 	page = gfn_to_page(kvm, table_gfn);
+	up_read(&current->mm->mmap_sem);
+
 	table = kmap_atomic(page, KM_USER0);
 
 	ret = CMPXCHG(&table[index], orig_pte, new_pte);
@@ -378,7 +381,7 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
 	if (r)
 		return r;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	/*
 	 * Look up the shadow pte for the faulting address.
 	 */
@@ -392,11 +395,13 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
 		pgprintk("%s: guest page fault\n", __FUNCTION__);
 		inject_page_fault(vcpu, addr, walker.error_code);
 		vcpu->arch.last_pt_write_count = 0; /* reset fork detector */
-		up_read(&current->mm->mmap_sem);
+		up_read(&vcpu->kvm->slots_lock);
 		return 0;
 	}
 
+	down_read(&current->mm->mmap_sem);
 	page = gfn_to_page(vcpu->kvm, walker.gfn);
+	up_read(&current->mm->mmap_sem);
 
 	spin_lock(&vcpu->kvm->mmu_lock);
 	kvm_mmu_free_some_pages(vcpu);
@@ -413,14 +418,14 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr,
 	 */
 	if (shadow_pte && is_io_pte(*shadow_pte)) {
 		spin_unlock(&vcpu->kvm->mmu_lock);
-		up_read(&current->mm->mmap_sem);
+		up_read(&vcpu->kvm->slots_lock);
 		return 1;
 	}
 
 	++vcpu->stat.pf_fixed;
 	kvm_mmu_audit(vcpu, "post page fault (fixed)");
 	spin_unlock(&vcpu->kvm->mmu_lock);
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 
 	return write_pt;
 }
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index ad36447e696e..86f5bf121838 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -1477,7 +1477,7 @@ static int alloc_apic_access_page(struct kvm *kvm)
 	struct kvm_userspace_memory_region kvm_userspace_mem;
 	int r = 0;
 
-	down_write(&current->mm->mmap_sem);
+	down_write(&kvm->slots_lock);
 	if (kvm->arch.apic_access_page)
 		goto out;
 	kvm_userspace_mem.slot = APIC_ACCESS_PAGE_PRIVATE_MEMSLOT;
@@ -1487,9 +1487,12 @@ static int alloc_apic_access_page(struct kvm *kvm)
 	r = __kvm_set_memory_region(kvm, &kvm_userspace_mem, 0);
 	if (r)
 		goto out;
+
+	down_read(&current->mm->mmap_sem);
 	kvm->arch.apic_access_page = gfn_to_page(kvm, 0xfee00);
+	up_read(&current->mm->mmap_sem);
 out:
-	up_write(&current->mm->mmap_sem);
+	up_write(&kvm->slots_lock);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 338764fa5391..6b01552bd1f1 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -184,7 +184,7 @@ int load_pdptrs(struct kvm_vcpu *vcpu, unsigned long cr3)
 	int ret;
 	u64 pdpte[ARRAY_SIZE(vcpu->arch.pdptrs)];
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	ret = kvm_read_guest_page(vcpu->kvm, pdpt_gfn, pdpte,
 				  offset * sizeof(u64), sizeof(pdpte));
 	if (ret < 0) {
@@ -201,7 +201,7 @@ int load_pdptrs(struct kvm_vcpu *vcpu, unsigned long cr3)
 
 	memcpy(vcpu->arch.pdptrs, pdpte, sizeof(vcpu->arch.pdptrs));
 out:
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 
 	return ret;
 }
@@ -215,13 +215,13 @@ static bool pdptrs_changed(struct kvm_vcpu *vcpu)
 	if (is_long_mode(vcpu) || !is_pae(vcpu))
 		return false;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	r = kvm_read_guest(vcpu->kvm, vcpu->arch.cr3 & ~31u, pdpte, sizeof(pdpte));
 	if (r < 0)
 		goto out;
 	changed = memcmp(pdpte, vcpu->arch.pdptrs, sizeof(pdpte)) != 0;
 out:
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 
 	return changed;
 }
@@ -359,7 +359,7 @@ void set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)
 		 */
 	}
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	/*
 	 * Does the new cr3 value map to physical memory? (Note, we
 	 * catch an invalid cr3 even in real-mode, because it would
@@ -375,7 +375,7 @@ void set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)
 		vcpu->arch.cr3 = cr3;
 		vcpu->arch.mmu.new_cr3(vcpu);
 	}
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 }
 EXPORT_SYMBOL_GPL(set_cr3);
 
@@ -1232,12 +1232,12 @@ static int kvm_vm_ioctl_set_nr_mmu_pages(struct kvm *kvm,
 	if (kvm_nr_mmu_pages < KVM_MIN_ALLOC_MMU_PAGES)
 		return -EINVAL;
 
-	down_write(&current->mm->mmap_sem);
+	down_write(&kvm->slots_lock);
 
 	kvm_mmu_change_mmu_pages(kvm, kvm_nr_mmu_pages);
 	kvm->arch.n_requested_mmu_pages = kvm_nr_mmu_pages;
 
-	up_write(&current->mm->mmap_sem);
+	up_write(&kvm->slots_lock);
 	return 0;
 }
 
@@ -1286,7 +1286,7 @@ static int kvm_vm_ioctl_set_memory_alias(struct kvm *kvm,
 	    < alias->target_phys_addr)
 		goto out;
 
-	down_write(&current->mm->mmap_sem);
+	down_write(&kvm->slots_lock);
 
 	p = &kvm->arch.aliases[alias->slot];
 	p->base_gfn = alias->guest_phys_addr >> PAGE_SHIFT;
@@ -1300,7 +1300,7 @@ static int kvm_vm_ioctl_set_memory_alias(struct kvm *kvm,
 
 	kvm_mmu_zap_all(kvm);
 
-	up_write(&current->mm->mmap_sem);
+	up_write(&kvm->slots_lock);
 
 	return 0;
 
@@ -1376,7 +1376,7 @@ int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
 	struct kvm_memory_slot *memslot;
 	int is_dirty = 0;
 
-	down_write(&current->mm->mmap_sem);
+	down_write(&kvm->slots_lock);
 
 	r = kvm_get_dirty_log(kvm, log, &is_dirty);
 	if (r)
@@ -1392,7 +1392,7 @@ int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
 	}
 	r = 0;
 out:
-	up_write(&current->mm->mmap_sem);
+	up_write(&kvm->slots_lock);
 	return r;
 }
 
@@ -1570,7 +1570,7 @@ int emulator_read_std(unsigned long addr,
 	void *data = val;
 	int r = X86EMUL_CONTINUE;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	while (bytes) {
 		gpa_t gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);
 		unsigned offset = addr & (PAGE_SIZE-1);
@@ -1592,7 +1592,7 @@ int emulator_read_std(unsigned long addr,
 		addr += tocopy;
 	}
 out:
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 	return r;
 }
 EXPORT_SYMBOL_GPL(emulator_read_std);
@@ -1611,9 +1611,9 @@ static int emulator_read_emulated(unsigned long addr,
 		return X86EMUL_CONTINUE;
 	}
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 
 	/* For APIC access vmexit */
 	if ((gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE)
@@ -1651,14 +1651,14 @@ static int emulator_write_phys(struct kvm_vcpu *vcpu, gpa_t gpa,
 {
 	int ret;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	ret = kvm_write_guest(vcpu->kvm, gpa, val, bytes);
 	if (ret < 0) {
-		up_read(&current->mm->mmap_sem);
+		up_read(&vcpu->kvm->slots_lock);
 		return 0;
 	}
 	kvm_mmu_pte_write(vcpu, gpa, val, bytes);
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 	return 1;
 }
 
@@ -1670,9 +1670,9 @@ static int emulator_write_emulated_onepage(unsigned long addr,
 	struct kvm_io_device *mmio_dev;
 	gpa_t                 gpa;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 
 	if (gpa == UNMAPPED_GVA) {
 		kvm_inject_page_fault(vcpu, addr, 2);
@@ -1749,7 +1749,7 @@ static int emulator_cmpxchg_emulated(unsigned long addr,
 		char *kaddr;
 		u64 val;
 
-		down_read(&current->mm->mmap_sem);
+		down_read(&vcpu->kvm->slots_lock);
 		gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);
 
 		if (gpa == UNMAPPED_GVA ||
@@ -1760,13 +1760,17 @@ static int emulator_cmpxchg_emulated(unsigned long addr,
 			goto emul_write;
 
 		val = *(u64 *)new;
+
+		down_read(&current->mm->mmap_sem);
 		page = gfn_to_page(vcpu->kvm, gpa >> PAGE_SHIFT);
+		up_read(&current->mm->mmap_sem);
+
 		kaddr = kmap_atomic(page, KM_USER0);
 		set_64bit((u64 *)(kaddr + offset_in_page(gpa)), val);
 		kunmap_atomic(kaddr, KM_USER0);
 		kvm_release_page_dirty(page);
 	emul_write:
-		up_read(&current->mm->mmap_sem);
+		up_read(&vcpu->kvm->slots_lock);
 	}
 #endif
 
@@ -2159,10 +2163,10 @@ int kvm_emulate_pio_string(struct kvm_vcpu *vcpu, struct kvm_run *run, int in,
 		kvm_x86_ops->skip_emulated_instruction(vcpu);
 
 	for (i = 0; i < nr_pages; ++i) {
-		down_read(&current->mm->mmap_sem);
+		down_read(&vcpu->kvm->slots_lock);
 		page = gva_to_page(vcpu, address + i * PAGE_SIZE);
 		vcpu->arch.pio.guest_pages[i] = page;
-		up_read(&current->mm->mmap_sem);
+		up_read(&vcpu->kvm->slots_lock);
 		if (!page) {
 			kvm_inject_gp(vcpu, 0);
 			free_pio_guest_pages(vcpu);
@@ -2485,8 +2489,9 @@ static void vapic_enter(struct kvm_vcpu *vcpu)
 
 	down_read(&current->mm->mmap_sem);
 	page = gfn_to_page(vcpu->kvm, apic->vapic_addr >> PAGE_SHIFT);
-	vcpu->arch.apic->vapic_page = page;
 	up_read(&current->mm->mmap_sem);
+
+	vcpu->arch.apic->vapic_page = page;
 }
 
 static void vapic_exit(struct kvm_vcpu *vcpu)
@@ -2959,9 +2964,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	gpa_t gpa;
 
 	vcpu_load(vcpu);
-	down_read(&current->mm->mmap_sem);
+	down_read(&vcpu->kvm->slots_lock);
 	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, vaddr);
-	up_read(&current->mm->mmap_sem);
+	up_read(&vcpu->kvm->slots_lock);
 	tr->physical_address = gpa;
 	tr->valid = gpa != UNMAPPED_GVA;
 	tr->writeable = 1;
@@ -3234,11 +3239,13 @@ int kvm_arch_set_memory_region(struct kvm *kvm,
 	 */
 	if (!user_alloc) {
 		if (npages && !old.rmap) {
+			down_write(&current->mm->mmap_sem);
 			memslot->userspace_addr = do_mmap(NULL, 0,
 						     npages * PAGE_SIZE,
 						     PROT_READ | PROT_WRITE,
 						     MAP_SHARED | MAP_ANONYMOUS,
 						     0);
+			up_write(&current->mm->mmap_sem);
 
 			if (IS_ERR((void *)memslot->userspace_addr))
 				return PTR_ERR((void *)memslot->userspace_addr);
@@ -3246,8 +3253,10 @@ int kvm_arch_set_memory_region(struct kvm *kvm,
 			if (!old.user_alloc && old.rmap) {
 				int ret;
 
+				down_write(&current->mm->mmap_sem);
 				ret = do_munmap(current->mm, old.userspace_addr,
 						old.npages * PAGE_SIZE);
+				up_write(&current->mm->mmap_sem);
 				if (ret < 0)
 					printk(KERN_WARNING
 				       "kvm_vm_ioctl_set_memory_region: "
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index ea4764b0a2f4..928b0d59e9ba 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -107,6 +107,7 @@ struct kvm_memory_slot {
 struct kvm {
 	struct mutex lock; /* protects the vcpus array and APIC accesses */
 	spinlock_t mmu_lock;
+	struct rw_semaphore slots_lock;
 	struct mm_struct *mm; /* userspace tied to this vm */
 	int nmemslots;
 	struct kvm_memory_slot memslots[KVM_MEMORY_SLOTS +
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 32fbf8006969..b2e12893e3f4 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -169,6 +169,7 @@ static struct kvm *kvm_create_vm(void)
 	kvm_io_bus_init(&kvm->pio_bus);
 	mutex_init(&kvm->lock);
 	kvm_io_bus_init(&kvm->mmio_bus);
+	init_rwsem(&kvm->slots_lock);
 	spin_lock(&kvm_lock);
 	list_add(&kvm->vm_list, &vm_list);
 	spin_unlock(&kvm_lock);
@@ -339,9 +340,9 @@ int kvm_set_memory_region(struct kvm *kvm,
 {
 	int r;
 
-	down_write(&current->mm->mmap_sem);
+	down_write(&kvm->slots_lock);
 	r = __kvm_set_memory_region(kvm, mem, user_alloc);
-	up_write(&current->mm->mmap_sem);
+	up_write(&kvm->slots_lock);
 	return r;
 }
 EXPORT_SYMBOL_GPL(kvm_set_memory_region);
-- 
cgit v1.2.3


From 62fb185130e4d420f71a30ff59d8b16b74ef5d2b Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Mon, 25 Feb 2008 17:34:02 +0100
Subject: sched: revert load_balance_monitor() changes

The following commits cause a number of regressions:

  commit 58e2d4ca581167c2a079f4ee02be2f0bc52e8729
  Author: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
  Date:   Fri Jan 25 21:08:00 2008 +0100
  sched: group scheduling, change how cpu load is calculated

  commit 6b2d7700266b9402e12824e11e0099ae6a4a6a79
  Author: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
  Date:   Fri Jan 25 21:08:00 2008 +0100
  sched: group scheduler, fix fairness of cpu bandwidth allocation for task groups

Namely:
 - very frequent wakeups on SMP, reported by PowerTop users.
 - cacheline trashing on (large) SMP
 - some latencies larger than 500ms

While there is a mergeable patch to fix the latter, the former issues
are not fixable in a manner suitable for .25 (we're at -rc3 now).

Hence we revert them and try again in v2.6.26.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
CC: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Tested-by: Alexey Zaytsev <alexey.zaytsev@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |   4 -
 kernel/sched.c        | 283 +++++++-------------------------------------------
 kernel/sched_fair.c   | 115 +++++++-------------
 kernel/sched_rt.c     |   4 -
 kernel/sysctl.c       |  18 ----
 5 files changed, 70 insertions(+), 354 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 2c9621f8bf87..9ae4030067a9 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1542,10 +1542,6 @@ extern unsigned int sysctl_sched_child_runs_first;
 extern unsigned int sysctl_sched_features;
 extern unsigned int sysctl_sched_migration_cost;
 extern unsigned int sysctl_sched_nr_migrate;
-#if defined(CONFIG_FAIR_GROUP_SCHED) && defined(CONFIG_SMP)
-extern unsigned int sysctl_sched_min_bal_int_shares;
-extern unsigned int sysctl_sched_max_bal_int_shares;
-#endif
 
 int sched_nr_latency_handler(struct ctl_table *table, int write,
 		struct file *file, void __user *buffer, size_t *length,
diff --git a/kernel/sched.c b/kernel/sched.c
index f06950c8a6ce..dcd553cc4ee8 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -174,41 +174,6 @@ struct task_group {
 	struct sched_entity **se;
 	/* runqueue "owned" by this group on each cpu */
 	struct cfs_rq **cfs_rq;
-
-	/*
-	 * shares assigned to a task group governs how much of cpu bandwidth
-	 * is allocated to the group. The more shares a group has, the more is
-	 * the cpu bandwidth allocated to it.
-	 *
-	 * For ex, lets say that there are three task groups, A, B and C which
-	 * have been assigned shares 1000, 2000 and 3000 respectively. Then,
-	 * cpu bandwidth allocated by the scheduler to task groups A, B and C
-	 * should be:
-	 *
-	 *	Bw(A) = 1000/(1000+2000+3000) * 100 = 16.66%
-	 *	Bw(B) = 2000/(1000+2000+3000) * 100 = 33.33%
-	 *	Bw(C) = 3000/(1000+2000+3000) * 100 = 50%
-	 *
-	 * The weight assigned to a task group's schedulable entities on every
-	 * cpu (task_group.se[a_cpu]->load.weight) is derived from the task
-	 * group's shares. For ex: lets say that task group A has been
-	 * assigned shares of 1000 and there are two CPUs in a system. Then,
-	 *
-	 *  tg_A->se[0]->load.weight = tg_A->se[1]->load.weight = 1000;
-	 *
-	 * Note: It's not necessary that each of a task's group schedulable
-	 *	 entity have the same weight on all CPUs. If the group
-	 *	 has 2 of its tasks on CPU0 and 1 task on CPU1, then a
-	 *	 better distribution of weight could be:
-	 *
-	 *	tg_A->se[0]->load.weight = 2/3 * 2000 = 1333
-	 *	tg_A->se[1]->load.weight = 1/2 * 2000 =  667
-	 *
-	 * rebalance_shares() is responsible for distributing the shares of a
-	 * task groups like this among the group's schedulable entities across
-	 * cpus.
-	 *
-	 */
 	unsigned long shares;
 #endif
 
@@ -250,22 +215,12 @@ static DEFINE_SPINLOCK(task_group_lock);
 static DEFINE_MUTEX(doms_cur_mutex);
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
-#ifdef CONFIG_SMP
-/* kernel thread that runs rebalance_shares() periodically */
-static struct task_struct *lb_monitor_task;
-static int load_balance_monitor(void *unused);
-#endif
-
-static void set_se_shares(struct sched_entity *se, unsigned long shares);
-
 #ifdef CONFIG_USER_SCHED
 # define INIT_TASK_GROUP_LOAD	(2*NICE_0_LOAD)
 #else
 # define INIT_TASK_GROUP_LOAD	NICE_0_LOAD
 #endif
 
-#define MIN_GROUP_SHARES	2
-
 static int init_task_group_load = INIT_TASK_GROUP_LOAD;
 #endif
 
@@ -1245,16 +1200,6 @@ static void cpuacct_charge(struct task_struct *tsk, u64 cputime);
 static inline void cpuacct_charge(struct task_struct *tsk, u64 cputime) {}
 #endif
 
-static inline void inc_cpu_load(struct rq *rq, unsigned long load)
-{
-	update_load_add(&rq->load, load);
-}
-
-static inline void dec_cpu_load(struct rq *rq, unsigned long load)
-{
-	update_load_sub(&rq->load, load);
-}
-
 #ifdef CONFIG_SMP
 static unsigned long source_load(int cpu, int type);
 static unsigned long target_load(int cpu, int type);
@@ -1272,14 +1217,26 @@ static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd);
 
 #define sched_class_highest (&rt_sched_class)
 
-static void inc_nr_running(struct rq *rq)
+static inline void inc_load(struct rq *rq, const struct task_struct *p)
+{
+	update_load_add(&rq->load, p->se.load.weight);
+}
+
+static inline void dec_load(struct rq *rq, const struct task_struct *p)
+{
+	update_load_sub(&rq->load, p->se.load.weight);
+}
+
+static void inc_nr_running(struct task_struct *p, struct rq *rq)
 {
 	rq->nr_running++;
+	inc_load(rq, p);
 }
 
-static void dec_nr_running(struct rq *rq)
+static void dec_nr_running(struct task_struct *p, struct rq *rq)
 {
 	rq->nr_running--;
+	dec_load(rq, p);
 }
 
 static void set_load_weight(struct task_struct *p)
@@ -1371,7 +1328,7 @@ static void activate_task(struct rq *rq, struct task_struct *p, int wakeup)
 		rq->nr_uninterruptible--;
 
 	enqueue_task(rq, p, wakeup);
-	inc_nr_running(rq);
+	inc_nr_running(p, rq);
 }
 
 /*
@@ -1383,7 +1340,7 @@ static void deactivate_task(struct rq *rq, struct task_struct *p, int sleep)
 		rq->nr_uninterruptible++;
 
 	dequeue_task(rq, p, sleep);
-	dec_nr_running(rq);
+	dec_nr_running(p, rq);
 }
 
 /**
@@ -2023,7 +1980,7 @@ void wake_up_new_task(struct task_struct *p, unsigned long clone_flags)
 		 * management (if any):
 		 */
 		p->sched_class->task_new(rq, p);
-		inc_nr_running(rq);
+		inc_nr_running(p, rq);
 	}
 	check_preempt_curr(rq, p);
 #ifdef CONFIG_SMP
@@ -4362,8 +4319,10 @@ void set_user_nice(struct task_struct *p, long nice)
 		goto out_unlock;
 	}
 	on_rq = p->se.on_rq;
-	if (on_rq)
+	if (on_rq) {
 		dequeue_task(rq, p, 0);
+		dec_load(rq, p);
+	}
 
 	p->static_prio = NICE_TO_PRIO(nice);
 	set_load_weight(p);
@@ -4373,6 +4332,7 @@ void set_user_nice(struct task_struct *p, long nice)
 
 	if (on_rq) {
 		enqueue_task(rq, p, 0);
+		inc_load(rq, p);
 		/*
 		 * If the task increased its priority or is running and
 		 * lowered its priority, then reschedule its CPU:
@@ -7087,21 +7047,6 @@ void __init sched_init_smp(void)
 	if (set_cpus_allowed(current, non_isolated_cpus) < 0)
 		BUG();
 	sched_init_granularity();
-
-#ifdef CONFIG_FAIR_GROUP_SCHED
-	if (nr_cpu_ids == 1)
-		return;
-
-	lb_monitor_task = kthread_create(load_balance_monitor, NULL,
-					 "group_balance");
-	if (!IS_ERR(lb_monitor_task)) {
-		lb_monitor_task->flags |= PF_NOFREEZE;
-		wake_up_process(lb_monitor_task);
-	} else {
-		printk(KERN_ERR "Could not create load balance monitor thread"
-			"(error = %ld) \n", PTR_ERR(lb_monitor_task));
-	}
-#endif
 }
 #else
 void __init sched_init_smp(void)
@@ -7424,157 +7369,6 @@ void set_curr_task(int cpu, struct task_struct *p)
 
 #ifdef CONFIG_GROUP_SCHED
 
-#if defined CONFIG_FAIR_GROUP_SCHED && defined CONFIG_SMP
-/*
- * distribute shares of all task groups among their schedulable entities,
- * to reflect load distribution across cpus.
- */
-static int rebalance_shares(struct sched_domain *sd, int this_cpu)
-{
-	struct cfs_rq *cfs_rq;
-	struct rq *rq = cpu_rq(this_cpu);
-	cpumask_t sdspan = sd->span;
-	int balanced = 1;
-
-	/* Walk thr' all the task groups that we have */
-	for_each_leaf_cfs_rq(rq, cfs_rq) {
-		int i;
-		unsigned long total_load = 0, total_shares;
-		struct task_group *tg = cfs_rq->tg;
-
-		/* Gather total task load of this group across cpus */
-		for_each_cpu_mask(i, sdspan)
-			total_load += tg->cfs_rq[i]->load.weight;
-
-		/* Nothing to do if this group has no load */
-		if (!total_load)
-			continue;
-
-		/*
-		 * tg->shares represents the number of cpu shares the task group
-		 * is eligible to hold on a single cpu. On N cpus, it is
-		 * eligible to hold (N * tg->shares) number of cpu shares.
-		 */
-		total_shares = tg->shares * cpus_weight(sdspan);
-
-		/*
-		 * redistribute total_shares across cpus as per the task load
-		 * distribution.
-		 */
-		for_each_cpu_mask(i, sdspan) {
-			unsigned long local_load, local_shares;
-
-			local_load = tg->cfs_rq[i]->load.weight;
-			local_shares = (local_load * total_shares) / total_load;
-			if (!local_shares)
-				local_shares = MIN_GROUP_SHARES;
-			if (local_shares == tg->se[i]->load.weight)
-				continue;
-
-			spin_lock_irq(&cpu_rq(i)->lock);
-			set_se_shares(tg->se[i], local_shares);
-			spin_unlock_irq(&cpu_rq(i)->lock);
-			balanced = 0;
-		}
-	}
-
-	return balanced;
-}
-
-/*
- * How frequently should we rebalance_shares() across cpus?
- *
- * The more frequently we rebalance shares, the more accurate is the fairness
- * of cpu bandwidth distribution between task groups. However higher frequency
- * also implies increased scheduling overhead.
- *
- * sysctl_sched_min_bal_int_shares represents the minimum interval between
- * consecutive calls to rebalance_shares() in the same sched domain.
- *
- * sysctl_sched_max_bal_int_shares represents the maximum interval between
- * consecutive calls to rebalance_shares() in the same sched domain.
- *
- * These settings allows for the appropriate trade-off between accuracy of
- * fairness and the associated overhead.
- *
- */
-
-/* default: 8ms, units: milliseconds */
-const_debug unsigned int sysctl_sched_min_bal_int_shares = 8;
-
-/* default: 128ms, units: milliseconds */
-const_debug unsigned int sysctl_sched_max_bal_int_shares = 128;
-
-/* kernel thread that runs rebalance_shares() periodically */
-static int load_balance_monitor(void *unused)
-{
-	unsigned int timeout = sysctl_sched_min_bal_int_shares;
-	struct sched_param schedparm;
-	int ret;
-
-	/*
-	 * We don't want this thread's execution to be limited by the shares
-	 * assigned to default group (init_task_group). Hence make it run
-	 * as a SCHED_RR RT task at the lowest priority.
-	 */
-	schedparm.sched_priority = 1;
-	ret = sched_setscheduler(current, SCHED_RR, &schedparm);
-	if (ret)
-		printk(KERN_ERR "Couldn't set SCHED_RR policy for load balance"
-				" monitor thread (error = %d) \n", ret);
-
-	while (!kthread_should_stop()) {
-		int i, cpu, balanced = 1;
-
-		/* Prevent cpus going down or coming up */
-		get_online_cpus();
-		/* lockout changes to doms_cur[] array */
-		lock_doms_cur();
-		/*
-		 * Enter a rcu read-side critical section to safely walk rq->sd
-		 * chain on various cpus and to walk task group list
-		 * (rq->leaf_cfs_rq_list) in rebalance_shares().
-		 */
-		rcu_read_lock();
-
-		for (i = 0; i < ndoms_cur; i++) {
-			cpumask_t cpumap = doms_cur[i];
-			struct sched_domain *sd = NULL, *sd_prev = NULL;
-
-			cpu = first_cpu(cpumap);
-
-			/* Find the highest domain at which to balance shares */
-			for_each_domain(cpu, sd) {
-				if (!(sd->flags & SD_LOAD_BALANCE))
-					continue;
-				sd_prev = sd;
-			}
-
-			sd = sd_prev;
-			/* sd == NULL? No load balance reqd in this domain */
-			if (!sd)
-				continue;
-
-			balanced &= rebalance_shares(sd, cpu);
-		}
-
-		rcu_read_unlock();
-
-		unlock_doms_cur();
-		put_online_cpus();
-
-		if (!balanced)
-			timeout = sysctl_sched_min_bal_int_shares;
-		else if (timeout < sysctl_sched_max_bal_int_shares)
-			timeout *= 2;
-
-		msleep_interruptible(timeout);
-	}
-
-	return 0;
-}
-#endif	/* CONFIG_SMP */
-
 #ifdef CONFIG_FAIR_GROUP_SCHED
 static void free_fair_sched_group(struct task_group *tg)
 {
@@ -7841,29 +7635,25 @@ void sched_move_task(struct task_struct *tsk)
 }
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
-/* rq->lock to be locked by caller */
 static void set_se_shares(struct sched_entity *se, unsigned long shares)
 {
 	struct cfs_rq *cfs_rq = se->cfs_rq;
 	struct rq *rq = cfs_rq->rq;
 	int on_rq;
 
-	if (!shares)
-		shares = MIN_GROUP_SHARES;
+	spin_lock_irq(&rq->lock);
 
 	on_rq = se->on_rq;
-	if (on_rq) {
+	if (on_rq)
 		dequeue_entity(cfs_rq, se, 0);
-		dec_cpu_load(rq, se->load.weight);
-	}
 
 	se->load.weight = shares;
 	se->load.inv_weight = div64_64((1ULL<<32), shares);
 
-	if (on_rq) {
+	if (on_rq)
 		enqueue_entity(cfs_rq, se, 0);
-		inc_cpu_load(rq, se->load.weight);
-	}
+
+	spin_unlock_irq(&rq->lock);
 }
 
 static DEFINE_MUTEX(shares_mutex);
@@ -7873,18 +7663,18 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares)
 	int i;
 	unsigned long flags;
 
+	/*
+	 * A weight of 0 or 1 can cause arithmetics problems.
+	 * (The default weight is 1024 - so there's no practical
+	 *  limitation from this.)
+	 */
+	if (shares < 2)
+		shares = 2;
+
 	mutex_lock(&shares_mutex);
 	if (tg->shares == shares)
 		goto done;
 
-	if (shares < MIN_GROUP_SHARES)
-		shares = MIN_GROUP_SHARES;
-
-	/*
-	 * Prevent any load balance activity (rebalance_shares,
-	 * load_balance_fair) from referring to this group first,
-	 * by taking it off the rq->leaf_cfs_rq_list on each cpu.
-	 */
 	spin_lock_irqsave(&task_group_lock, flags);
 	for_each_possible_cpu(i)
 		unregister_fair_sched_group(tg, i);
@@ -7898,11 +7688,8 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares)
 	 * w/o tripping rebalance_share or load_balance_fair.
 	 */
 	tg->shares = shares;
-	for_each_possible_cpu(i) {
-		spin_lock_irq(&cpu_rq(i)->lock);
+	for_each_possible_cpu(i)
 		set_se_shares(tg->se[i], shares);
-		spin_unlock_irq(&cpu_rq(i)->lock);
-	}
 
 	/*
 	 * Enable load balance activity on this group, by inserting it back on
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index c8e6492c5925..3df4d46994ca 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -727,8 +727,6 @@ static inline struct sched_entity *parent_entity(struct sched_entity *se)
 	return se->parent;
 }
 
-#define GROUP_IMBALANCE_PCT	20
-
 #else	/* CONFIG_FAIR_GROUP_SCHED */
 
 #define for_each_sched_entity(se) \
@@ -819,26 +817,15 @@ hrtick_start_fair(struct rq *rq, struct task_struct *p)
 static void enqueue_task_fair(struct rq *rq, struct task_struct *p, int wakeup)
 {
 	struct cfs_rq *cfs_rq;
-	struct sched_entity *se = &p->se,
-			    *topse = NULL;	/* Highest schedulable entity */
-	int incload = 1;
+	struct sched_entity *se = &p->se;
 
 	for_each_sched_entity(se) {
-		topse = se;
-		if (se->on_rq) {
-			incload = 0;
+		if (se->on_rq)
 			break;
-		}
 		cfs_rq = cfs_rq_of(se);
 		enqueue_entity(cfs_rq, se, wakeup);
 		wakeup = 1;
 	}
-	/* Increment cpu load if we just enqueued the first task of a group on
-	 * 'rq->cpu'. 'topse' represents the group to which task 'p' belongs
-	 * at the highest grouping level.
-	 */
-	if (incload)
-		inc_cpu_load(rq, topse->load.weight);
 
 	hrtick_start_fair(rq, rq->curr);
 }
@@ -851,28 +838,16 @@ static void enqueue_task_fair(struct rq *rq, struct task_struct *p, int wakeup)
 static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int sleep)
 {
 	struct cfs_rq *cfs_rq;
-	struct sched_entity *se = &p->se,
-			    *topse = NULL; 	/* Highest schedulable entity */
-	int decload = 1;
+	struct sched_entity *se = &p->se;
 
 	for_each_sched_entity(se) {
-		topse = se;
 		cfs_rq = cfs_rq_of(se);
 		dequeue_entity(cfs_rq, se, sleep);
 		/* Don't dequeue parent if it has other entities besides us */
-		if (cfs_rq->load.weight) {
-			if (parent_entity(se))
-				decload = 0;
+		if (cfs_rq->load.weight)
 			break;
-		}
 		sleep = 1;
 	}
-	/* Decrement cpu load if we just dequeued the last task of a group on
-	 * 'rq->cpu'. 'topse' represents the group to which task 'p' belongs
-	 * at the highest grouping level.
-	 */
-	if (decload)
-		dec_cpu_load(rq, topse->load.weight);
 
 	hrtick_start_fair(rq, rq->curr);
 }
@@ -1186,6 +1161,25 @@ static struct task_struct *load_balance_next_fair(void *arg)
 	return __load_balance_iterator(cfs_rq, cfs_rq->rb_load_balance_curr);
 }
 
+#ifdef CONFIG_FAIR_GROUP_SCHED
+static int cfs_rq_best_prio(struct cfs_rq *cfs_rq)
+{
+	struct sched_entity *curr;
+	struct task_struct *p;
+
+	if (!cfs_rq->nr_running || !first_fair(cfs_rq))
+		return MAX_PRIO;
+
+	curr = cfs_rq->curr;
+	if (!curr)
+		curr = __pick_next_entity(cfs_rq);
+
+	p = task_of(curr);
+
+	return p->prio;
+}
+#endif
+
 static unsigned long
 load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
 		  unsigned long max_load_move,
@@ -1195,45 +1189,28 @@ load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
 	struct cfs_rq *busy_cfs_rq;
 	long rem_load_move = max_load_move;
 	struct rq_iterator cfs_rq_iterator;
-	unsigned long load_moved;
 
 	cfs_rq_iterator.start = load_balance_start_fair;
 	cfs_rq_iterator.next = load_balance_next_fair;
 
 	for_each_leaf_cfs_rq(busiest, busy_cfs_rq) {
 #ifdef CONFIG_FAIR_GROUP_SCHED
-		struct cfs_rq *this_cfs_rq = busy_cfs_rq->tg->cfs_rq[this_cpu];
-		unsigned long maxload, task_load, group_weight;
-		unsigned long thisload, per_task_load;
-		struct sched_entity *se = busy_cfs_rq->tg->se[busiest->cpu];
-
-		task_load = busy_cfs_rq->load.weight;
-		group_weight = se->load.weight;
+		struct cfs_rq *this_cfs_rq;
+		long imbalance;
+		unsigned long maxload;
 
-		/*
-		 * 'group_weight' is contributed by tasks of total weight
-		 * 'task_load'. To move 'rem_load_move' worth of weight only,
-		 * we need to move a maximum task load of:
-		 *
-		 * 	maxload = (remload / group_weight) * task_load;
-		 */
-		maxload = (rem_load_move * task_load) / group_weight;
+		this_cfs_rq = cpu_cfs_rq(busy_cfs_rq, this_cpu);
 
-		if (!maxload || !task_load)
+		imbalance = busy_cfs_rq->load.weight - this_cfs_rq->load.weight;
+		/* Don't pull if this_cfs_rq has more load than busy_cfs_rq */
+		if (imbalance <= 0)
 			continue;
 
-		per_task_load = task_load / busy_cfs_rq->nr_running;
-		/*
-		 * balance_tasks will try to forcibly move atleast one task if
-		 * possible (because of SCHED_LOAD_SCALE_FUZZ). Avoid that if
-		 * maxload is less than GROUP_IMBALANCE_FUZZ% the per_task_load.
-		 */
-		 if (100 * maxload < GROUP_IMBALANCE_PCT * per_task_load)
-			continue;
+		/* Don't pull more than imbalance/2 */
+		imbalance /= 2;
+		maxload = min(rem_load_move, imbalance);
 
-		/* Disable priority-based load balance */
-		*this_best_prio = 0;
-		thisload = this_cfs_rq->load.weight;
+		*this_best_prio = cfs_rq_best_prio(this_cfs_rq);
 #else
 # define maxload rem_load_move
 #endif
@@ -1242,33 +1219,11 @@ load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
 		 * load_balance_[start|next]_fair iterators
 		 */
 		cfs_rq_iterator.arg = busy_cfs_rq;
-		load_moved = balance_tasks(this_rq, this_cpu, busiest,
+		rem_load_move -= balance_tasks(this_rq, this_cpu, busiest,
 					       maxload, sd, idle, all_pinned,
 					       this_best_prio,
 					       &cfs_rq_iterator);
 
-#ifdef CONFIG_FAIR_GROUP_SCHED
-		/*
-		 * load_moved holds the task load that was moved. The
-		 * effective (group) weight moved would be:
-		 * 	load_moved_eff = load_moved/task_load * group_weight;
-		 */
-		load_moved = (group_weight * load_moved) / task_load;
-
-		/* Adjust shares on both cpus to reflect load_moved */
-		group_weight -= load_moved;
-		set_se_shares(se, group_weight);
-
-		se = busy_cfs_rq->tg->se[this_cpu];
-		if (!thisload)
-			group_weight = load_moved;
-		else
-			group_weight = se->load.weight + load_moved;
-		set_se_shares(se, group_weight);
-#endif
-
-		rem_load_move -= load_moved;
-
 		if (rem_load_move <= 0)
 			break;
 	}
diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index f54792b175b2..76e828517541 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -393,8 +393,6 @@ static void enqueue_task_rt(struct rq *rq, struct task_struct *p, int wakeup)
 	 */
 	for_each_sched_rt_entity(rt_se)
 		enqueue_rt_entity(rt_se);
-
-	inc_cpu_load(rq, p->se.load.weight);
 }
 
 static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int sleep)
@@ -414,8 +412,6 @@ static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int sleep)
 		if (rt_rq && rt_rq->rt_nr_running)
 			enqueue_rt_entity(rt_se);
 	}
-
-	dec_cpu_load(rq, p->se.load.weight);
 }
 
 /*
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 8b7e95411795..b2a2d6889bab 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -311,24 +311,6 @@ static struct ctl_table kern_table[] = {
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec,
 	},
-#if defined(CONFIG_FAIR_GROUP_SCHED) && defined(CONFIG_SMP)
-	{
-		.ctl_name       = CTL_UNNUMBERED,
-		.procname       = "sched_min_bal_int_shares",
-		.data           = &sysctl_sched_min_bal_int_shares,
-		.maxlen         = sizeof(unsigned int),
-		.mode           = 0644,
-		.proc_handler   = &proc_dointvec,
-	},
-	{
-		.ctl_name       = CTL_UNNUMBERED,
-		.procname       = "sched_max_bal_int_shares",
-		.data           = &sysctl_sched_max_bal_int_shares,
-		.maxlen         = sizeof(unsigned int),
-		.mode           = 0644,
-		.proc_handler   = &proc_dointvec,
-	},
-#endif
 #endif
 	{
 		.ctl_name	= CTL_UNNUMBERED,
-- 
cgit v1.2.3


From ec8670f1f795badedaa056a3a3245b9b82201747 Mon Sep 17 00:00:00 2001
From: Dan Williams <dan.j.williams@intel.com>
Date: Sat, 1 Mar 2008 07:51:29 -0700
Subject: dmaengine: fix sparse warning

include/linux/dmaengine.h:364:2: warning: returning void-valued expression

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 include/linux/dmaengine.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index acbb364674ff..261e43a4c873 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -366,7 +366,7 @@ __dma_has_cap(enum dma_transaction_type tx_type, dma_cap_mask_t *srcp)
  */
 static inline void dma_async_issue_pending(struct dma_chan *chan)
 {
-	return chan->device->device_issue_pending(chan);
+	chan->device->device_issue_pending(chan);
 }
 
 #define dma_async_memcpy_issue_pending(chan) dma_async_issue_pending(chan)
-- 
cgit v1.2.3


From d9452e9f81e997cbd0c9bface8d2c2a4b064cc3e Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Tue, 4 Mar 2008 12:28:49 -0800
Subject: [NETPOLL]: Revert two bogus cleanups that broke netconsole.

Based upon a report by Andrew Morton and code analysis done
by Jarek Poplawski.

This reverts 33f807ba0d9259e7c75c7a2ce8bd2787e5b540c7 ("[NETPOLL]:
Kill NETPOLL_RX_DROP, set but never tested.")  and
c7b6ea24b43afb5749cb704e143df19d70e23dea ("[NETPOLL]: Don't need
rx_flags.").

The rx_flags did get tested for zero vs. non-zero and therefore we do
need those tests and that code which sets NETPOLL_RX_DROP et al.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netpoll.h |  7 ++++---
 net/core/netpoll.c      | 12 ++++++++----
 2 files changed, 12 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h
index a0525a1f4715..e3d79593fb3a 100644
--- a/include/linux/netpoll.h
+++ b/include/linux/netpoll.h
@@ -25,6 +25,7 @@ struct netpoll {
 
 struct netpoll_info {
 	atomic_t refcnt;
+	int rx_flags;
 	spinlock_t rx_lock;
 	struct netpoll *rx_np; /* netpoll that registered an rx_hook */
 	struct sk_buff_head arp_tx; /* list of arp requests to reply to */
@@ -50,12 +51,12 @@ static inline int netpoll_rx(struct sk_buff *skb)
 	unsigned long flags;
 	int ret = 0;
 
-	if (!npinfo || !npinfo->rx_np)
+	if (!npinfo || (!npinfo->rx_np && !npinfo->rx_flags))
 		return 0;
 
 	spin_lock_irqsave(&npinfo->rx_lock, flags);
-	/* check rx_np again with the lock held */
-	if (npinfo->rx_np && __netpoll_rx(skb))
+	/* check rx_flags again with the lock held */
+	if (npinfo->rx_flags && __netpoll_rx(skb))
 		ret = 1;
 	spin_unlock_irqrestore(&npinfo->rx_lock, flags);
 
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index 6faa128a4c8e..4b7e756181c9 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -39,6 +39,8 @@ static struct sk_buff_head skb_pool;
 static atomic_t trapped;
 
 #define USEC_PER_POLL	50
+#define NETPOLL_RX_ENABLED  1
+#define NETPOLL_RX_DROP     2
 
 #define MAX_SKB_SIZE \
 		(MAX_UDP_CHUNK + sizeof(struct udphdr) + \
@@ -126,11 +128,13 @@ static int poll_one_napi(struct netpoll_info *npinfo,
 	if (!test_bit(NAPI_STATE_SCHED, &napi->state))
 		return budget;
 
+	npinfo->rx_flags |= NETPOLL_RX_DROP;
 	atomic_inc(&trapped);
 
 	work = napi->poll(napi, budget);
 
 	atomic_dec(&trapped);
+	npinfo->rx_flags &= ~NETPOLL_RX_DROP;
 
 	return budget - work;
 }
@@ -472,7 +476,7 @@ int __netpoll_rx(struct sk_buff *skb)
 	if (skb->dev->type != ARPHRD_ETHER)
 		goto out;
 
-	/* if receive ARP during middle of NAPI poll, then queue */
+	/* check if netpoll clients need ARP */
 	if (skb->protocol == htons(ETH_P_ARP) &&
 	    atomic_read(&trapped)) {
 		skb_queue_tail(&npi->arp_tx, skb);
@@ -534,9 +538,6 @@ int __netpoll_rx(struct sk_buff *skb)
 	return 1;
 
 out:
-	/* If packet received while already in poll then just
-	 * silently drop.
-	 */
 	if (atomic_read(&trapped)) {
 		kfree_skb(skb);
 		return 1;
@@ -675,6 +676,7 @@ int netpoll_setup(struct netpoll *np)
 			goto release;
 		}
 
+		npinfo->rx_flags = 0;
 		npinfo->rx_np = NULL;
 
 		spin_lock_init(&npinfo->rx_lock);
@@ -756,6 +758,7 @@ int netpoll_setup(struct netpoll *np)
 
 	if (np->rx_hook) {
 		spin_lock_irqsave(&npinfo->rx_lock, flags);
+		npinfo->rx_flags |= NETPOLL_RX_ENABLED;
 		npinfo->rx_np = np;
 		spin_unlock_irqrestore(&npinfo->rx_lock, flags);
 	}
@@ -797,6 +800,7 @@ void netpoll_cleanup(struct netpoll *np)
 			if (npinfo->rx_np == np) {
 				spin_lock_irqsave(&npinfo->rx_lock, flags);
 				npinfo->rx_np = NULL;
+				npinfo->rx_flags &= ~NETPOLL_RX_ENABLED;
 				spin_unlock_irqrestore(&npinfo->rx_lock, flags);
 			}
 
-- 
cgit v1.2.3


From 3634634edd49c115da931998b9540bcc17665b05 Mon Sep 17 00:00:00 2001
From: Harvey Harrison <harvey.harrison@gmail.com>
Date: Wed, 13 Feb 2008 17:08:16 -0800
Subject: debugfs: fix sparse warnings

extern does not belong in C files, move declaration to linux/debugfs.h
fs/debugfs/file.c:42:30: warning: symbol 'debugfs_file_operations' was not declared. Should it be static?
fs/debugfs/file.c:54:31: warning: symbol 'debugfs_link_operations' was not declared. Should it be static?

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 fs/debugfs/inode.c      | 4 ----
 include/linux/debugfs.h | 5 +++++
 2 files changed, 5 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index d26e2826ba5b..e9602d85c11d 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -29,10 +29,6 @@
 
 #define DEBUGFS_MAGIC	0x64626720
 
-/* declared over in file.c */
-extern struct file_operations debugfs_file_operations;
-extern struct inode_operations debugfs_link_operations;
-
 static struct vfsmount *debugfs_mount;
 static int debugfs_mount_count;
 
diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h
index f592d6de3b97..7266124361b4 100644
--- a/include/linux/debugfs.h
+++ b/include/linux/debugfs.h
@@ -27,6 +27,11 @@ struct debugfs_blob_wrapper {
 };
 
 #if defined(CONFIG_DEBUG_FS)
+
+/* declared over in file.c */
+extern const struct file_operations debugfs_file_operations;
+extern const struct inode_operations debugfs_link_operations;
+
 struct dentry *debugfs_create_file(const char *name, mode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops);
-- 
cgit v1.2.3


From d6d914f52b15d5a8e81ad481e02d9ab30d412a29 Mon Sep 17 00:00:00 2001
From: Lei Ming <tom.leiming@gmail.com>
Date: Mon, 25 Feb 2008 18:07:28 +0800
Subject: USB: fix comment of struct usb_interface

update the comment for the removed "driver" field  and  being
out-of-order of  @cur_altsetting and @num_altsetting.

Signed-off-by: Lei Ming <tom.leiming@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/usb.h | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/usb.h b/include/linux/usb.h
index 5bd3ae8aaaf4..583e0481dfa0 100644
--- a/include/linux/usb.h
+++ b/include/linux/usb.h
@@ -94,10 +94,9 @@ enum usb_interface_condition {
  * @altsetting: array of interface structures, one for each alternate
  * 	setting that may be selected.  Each one includes a set of
  * 	endpoint configurations.  They will be in no particular order.
- * @num_altsetting: number of altsettings defined.
  * @cur_altsetting: the current altsetting.
+ * @num_altsetting: number of altsettings defined.
  * @intf_assoc: interface association descriptor
- * @driver: the USB driver that is bound to this interface.
  * @minor: the minor number assigned to this interface, if this
  *	interface is bound to a driver that uses the USB major number.
  *	If this interface does not use the USB major, this field should
-- 
cgit v1.2.3


From 90a1ba0c5e39eeea278f263c28ae02166c5911c8 Mon Sep 17 00:00:00 2001
From: Jonas Bonn <jonas@southpole.se>
Date: Fri, 22 Feb 2008 11:02:21 +0100
Subject: PCI: Add DECLARE_PCI_DEVICE_TABLE macro

The definitions of struct pci_device_id arrays should generally follow
the same pattern across the entire kernel.  This macro defines this
array as const and puts it into the __devinitconst section.

There are currently many definitions scattered about the kernel that
omit the __devinitdata modifier despite the documentation stating that
it should always be there.  These definitions really also should have
been const, which wasn't possible before but has become so with the
addition of the __devinitconst attribute.

Furthermore, there are definitions that use "const" and __devinitdata,
which is explicitly wrong but the compiler doesn't catch section
mismatches if there's only one such one case in the module (which is
often the case).

Adding the __devinitconst modifier where there was nothing before buys
us memory.  Adding the const modifier gives the compiler a chance to do
its thing.  Changing __devinitdata to __devinitconst where it was wrong
actually fixes some compiler errors in older (mid-release) kernels that
were patched over by "removing" the section attribute altogether (which
wastes memory).

This macro makes it pretty difficult to get this definition wrong in
the future...

Signed-off-by: Jonas Bonn <jonas@southpole.se>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 Documentation/pci.txt |  6 ++++--
 include/linux/pci.h   | 10 ++++++++++
 2 files changed, 14 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/pci.txt b/Documentation/pci.txt
index 72b20c639596..bb7bd27d4682 100644
--- a/Documentation/pci.txt
+++ b/Documentation/pci.txt
@@ -123,7 +123,8 @@ initialization with a pointer to a structure describing the driver
 
 
 The ID table is an array of struct pci_device_id entries ending with an
-all-zero entry.  Each entry consists of:
+all-zero entry; use of the macro DECLARE_PCI_DEVICE_TABLE is the preferred
+method of declaring the table.  Each entry consists of:
 
 	vendor,device	Vendor and device ID to match (or PCI_ANY_ID)
 
@@ -191,7 +192,8 @@ Tips on when/where to use the above attributes:
 
 	o Do not mark the struct pci_driver.
 
-	o The ID table array should be marked __devinitdata.
+	o The ID table array should be marked __devinitconst; this is done
+	  automatically if the table is declared with DECLARE_PCI_DEVICE_TABLE().
 
 	o The probe() and remove() functions should be marked __devinit
 	  and __devexit respectively.  All initialization functions
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 87195b62de52..f3165e7ac431 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -388,6 +388,16 @@ struct pci_driver {
 
 #define	to_pci_driver(drv) container_of(drv, struct pci_driver, driver)
 
+/**
+ * DECLARE_PCI_DEVICE_TABLE - macro used to describe a pci device table
+ * @_table: device table name
+ *
+ * This macro is used to create a struct pci_device_id array (a device table)
+ * in a generic manner.
+ */
+#define DECLARE_PCI_DEVICE_TABLE(_table) \
+	const struct pci_device_id _table[] __devinitconst
+
 /**
  * PCI_DEVICE - macro used to describe a specific pci device
  * @vend: the 16 bit PCI Vendor ID
-- 
cgit v1.2.3


From 7560fa60fcdcdb0da662f6a9fad9064b554ef46c Mon Sep 17 00:00:00 2001
From: David Brownell <david-b@pacbell.net>
Date: Tue, 4 Mar 2008 14:28:27 -0800
Subject: gpio: <linux/gpio.h> and "no GPIO support here" stubs

Add a <linux/gpio.h> defining fail/warn stubs for GPIO calls on platforms that
don't support the GPIO programming interface.  That includes the arch-specific
implementation glue otherwise.

This facilitates a new model for GPIO usage: drivers that can use GPIOs if
they're available, but don't require them.  One example of such a driver is
NAND driver for various FreeScale chips.  On platforms update with GPIO
support, they can be used instead of a worst-case delay to verify that the
BUSY signal is off.

(Also includes a couple minor unrelated doc updates.)

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 Documentation/gpio.txt | 16 ++++++---
 include/linux/gpio.h   | 95 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 107 insertions(+), 4 deletions(-)
 create mode 100644 include/linux/gpio.h

(limited to 'include/linux')

diff --git a/Documentation/gpio.txt b/Documentation/gpio.txt
index 8da724e2a0ff..54630095aa3c 100644
--- a/Documentation/gpio.txt
+++ b/Documentation/gpio.txt
@@ -2,6 +2,9 @@ GPIO Interfaces
 
 This provides an overview of GPIO access conventions on Linux.
 
+These calls use the gpio_* naming prefix.  No other calls should use that
+prefix, or the related __gpio_* prefix.
+
 
 What is a GPIO?
 ===============
@@ -69,11 +72,13 @@ in this document, but drivers acting as clients to the GPIO interface must
 not care how it's implemented.)
 
 That said, if the convention is supported on their platform, drivers should
-use it when possible.  Platforms should declare GENERIC_GPIO support in
-Kconfig (boolean true), which multi-platform drivers can depend on when
-using the include file:
+use it when possible.  Platforms must declare GENERIC_GPIO support in their
+Kconfig (boolean true), and provide an <asm/gpio.h> file.  Drivers that can't
+work without standard GPIO calls should have Kconfig entries which depend
+on GENERIC_GPIO.  The GPIO calls are available, either as "real code" or as
+optimized-away stubs, when drivers use the include file:
 
-	#include <asm/gpio.h>
+	#include <linux/gpio.h>
 
 If you stick to this convention then it'll be easier for other developers to
 see what your code is doing, and help maintain it.
@@ -316,6 +321,9 @@ pulldowns integrated on some platforms.  Not all platforms support them,
 or support them in the same way; and any given board might use external
 pullups (or pulldowns) so that the on-chip ones should not be used.
 (When a circuit needs 5 kOhm, on-chip 100 kOhm resistors won't do.)
+Likewise drive strength (2 mA vs 20 mA) and voltage (1.8V vs 3.3V) is a
+platform-specific issue, as are models like (not) having a one-to-one
+correspondence between configurable pins and GPIOs.
 
 There are other system-specific mechanisms that are not specified here,
 like the aforementioned options for input de-glitching and wire-OR output.
diff --git a/include/linux/gpio.h b/include/linux/gpio.h
new file mode 100644
index 000000000000..4987a84078ef
--- /dev/null
+++ b/include/linux/gpio.h
@@ -0,0 +1,95 @@
+#ifndef __LINUX_GPIO_H
+#define __LINUX_GPIO_H
+
+/* see Documentation/gpio.txt */
+
+#ifdef CONFIG_GENERIC_GPIO
+#include <asm/gpio.h>
+
+#else
+
+/*
+ * Some platforms don't support the GPIO programming interface.
+ *
+ * In case some driver uses it anyway (it should normally have
+ * depended on GENERIC_GPIO), these routines help the compiler
+ * optimize out much GPIO-related code ... or trigger a runtime
+ * warning when something is wrongly called.
+ */
+
+static inline int gpio_is_valid(int number)
+{
+	return 0;
+}
+
+static inline int gpio_request(unsigned gpio, const char *label)
+{
+	return -ENOSYS;
+}
+
+static inline void gpio_free(unsigned gpio)
+{
+	/* GPIO can never have been requested */
+	WARN_ON(1);
+}
+
+static inline int gpio_direction_input(unsigned gpio)
+{
+	return -ENOSYS;
+}
+
+static inline int gpio_direction_output(unsigned gpio, int value)
+{
+	return -ENOSYS;
+}
+
+static inline int gpio_get_value(unsigned gpio)
+{
+	/* GPIO can never have been requested or set as {in,out}put */
+	WARN_ON(1);
+	return 0;
+}
+
+static inline void gpio_set_value(unsigned gpio, int value)
+{
+	/* GPIO can never have been requested or set as output */
+	WARN_ON(1);
+}
+
+static inline int gpio_cansleep(unsigned gpio)
+{
+	/* GPIO can never have been requested or set as {in,out}put */
+	WARN_ON(1);
+	return 0;
+}
+
+static inline int gpio_get_value_cansleep(unsigned gpio)
+{
+	/* GPIO can never have been requested or set as {in,out}put */
+	WARN_ON(1);
+	return 0;
+}
+
+static inline void gpio_set_value_cansleep(unsigned gpio, int value)
+{
+	/* GPIO can never have been requested or set as output */
+	WARN_ON(1);
+}
+
+static inline int gpio_to_irq(unsigned gpio)
+{
+	/* GPIO can never have been requested or set as input */
+	WARN_ON(1);
+	return -EINVAL;
+}
+
+static inline int irq_to_gpio(unsigned irq)
+{
+	/* irq can never have been returned from gpio_to_irq() */
+	WARN_ON(1);
+	return -EINVAL;
+}
+
+#endif
+
+#endif /* __LINUX_GPIO_H */
-- 
cgit v1.2.3


From 9edddaa200df18e08fe0cf21036e8ae467b1363c Mon Sep 17 00:00:00 2001
From: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Date: Tue, 4 Mar 2008 14:28:37 -0800
Subject: Kprobes: indicate kretprobe support in Kconfig

Add CONFIG_HAVE_KRETPROBES to the arch/<arch>/Kconfig file for relevant
architectures with kprobes support.  This facilitates easy handling of
in-kernel modules (like samples/kprobes/kretprobe_example.c) that depend on
kretprobes being present in the kernel.

Thanks to Sam Ravnborg for helping make the patch more lean.

Per Mathieu's suggestion, added CONFIG_KRETPROBES and fixed up dependencies.

Signed-off-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Acked-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/Kconfig                  | 7 +++++++
 arch/arm/Kconfig              | 1 +
 arch/ia64/Kconfig             | 1 +
 arch/powerpc/Kconfig          | 1 +
 arch/s390/Kconfig             | 1 +
 arch/sparc64/Kconfig          | 1 +
 arch/x86/Kconfig              | 1 +
 include/asm-arm/kprobes.h     | 1 -
 include/asm-ia64/kprobes.h    | 1 -
 include/asm-powerpc/kprobes.h | 1 -
 include/asm-s390/kprobes.h    | 1 -
 include/asm-sparc64/kprobes.h | 2 --
 include/asm-x86/kprobes.h     | 1 -
 include/linux/kprobes.h       | 6 +++---
 kernel/kprobes.c              | 9 +++------
 15 files changed, 19 insertions(+), 16 deletions(-)

(limited to 'include/linux')

diff --git a/arch/Kconfig b/arch/Kconfig
index 3d72dc3fc8f5..694c9af520bb 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -27,5 +27,12 @@ config KPROBES
 	  for kernel debugging, non-intrusive instrumentation and testing.
 	  If in doubt, say "N".
 
+config KRETPROBES
+	def_bool y
+	depends on KPROBES && HAVE_KRETPROBES
+
 config HAVE_KPROBES
 	def_bool n
+
+config HAVE_KRETPROBES
+	def_bool n
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 16b82e1272b0..955fc53c1c01 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -12,6 +12,7 @@ config ARM
 	select SYS_SUPPORTS_APM_EMULATION
 	select HAVE_OPROFILE
 	select HAVE_KPROBES if (!XIP_KERNEL)
+	select HAVE_KRETPROBES if (HAVE_KPROBES)
 	help
 	  The ARM series is a line of low-power-consumption RISC chip designs
 	  licensed by ARM Ltd and targeted at embedded applications and
diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
index dff9edfc7465..56762d3c2a6a 100644
--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -18,6 +18,7 @@ config IA64
 	select HAVE_IDE
 	select HAVE_OPROFILE
 	select HAVE_KPROBES
+	select HAVE_KRETPROBES
 	default y
 	help
 	  The Itanium Processor Family is Intel's 64-bit successor to
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 5b8d8382b762..1189d8d6170d 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -90,6 +90,7 @@ config PPC
 	select HAVE_IDE
 	select HAVE_OPROFILE
 	select HAVE_KPROBES
+	select HAVE_KRETPROBES
 
 config EARLY_PRINTK
 	bool
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index b21444b681b6..9892827b6176 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -61,6 +61,7 @@ config S390
 	def_bool y
 	select HAVE_OPROFILE
 	select HAVE_KPROBES
+	select HAVE_KRETPROBES
 
 source "init/Kconfig"
 
diff --git a/arch/sparc64/Kconfig b/arch/sparc64/Kconfig
index 3af378ddb6ae..463d1be32c98 100644
--- a/arch/sparc64/Kconfig
+++ b/arch/sparc64/Kconfig
@@ -10,6 +10,7 @@ config SPARC
 	default y
 	select HAVE_OPROFILE
 	select HAVE_KPROBES
+	select HAVE_KRETPROBES
 
 config SPARC64
 	bool
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 53800b80a204..f41c9538ca30 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -21,6 +21,7 @@ config X86
 	select HAVE_IDE
 	select HAVE_OPROFILE
 	select HAVE_KPROBES
+	select HAVE_KRETPROBES
 	select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64)
 
 
diff --git a/include/asm-arm/kprobes.h b/include/asm-arm/kprobes.h
index 4e7bd32288ae..c042194d3ab5 100644
--- a/include/asm-arm/kprobes.h
+++ b/include/asm-arm/kprobes.h
@@ -20,7 +20,6 @@
 #include <linux/ptrace.h>
 #include <linux/percpu.h>
 
-#define ARCH_SUPPORTS_KRETPROBES
 #define __ARCH_WANT_KPROBES_INSN_SLOT
 #define MAX_INSN_SIZE			2
 #define MAX_STACK_SIZE			64	/* 32 would probably be OK */
diff --git a/include/asm-ia64/kprobes.h b/include/asm-ia64/kprobes.h
index a93ce9ef07ff..adbaba14eb0a 100644
--- a/include/asm-ia64/kprobes.h
+++ b/include/asm-ia64/kprobes.h
@@ -82,7 +82,6 @@ struct kprobe_ctlblk {
 	struct prev_kprobe prev_kprobe[ARCH_PREV_KPROBE_SZ];
 };
 
-#define ARCH_SUPPORTS_KRETPROBES
 #define kretprobe_blacklist_size 0
 
 #define SLOT0_OPCODE_SHIFT	(37)
diff --git a/include/asm-powerpc/kprobes.h b/include/asm-powerpc/kprobes.h
index afabad230dbb..d0e7701fa1f6 100644
--- a/include/asm-powerpc/kprobes.h
+++ b/include/asm-powerpc/kprobes.h
@@ -80,7 +80,6 @@ typedef unsigned int kprobe_opcode_t;
 #define is_trap(instr)	(IS_TW(instr) || IS_TWI(instr))
 #endif
 
-#define ARCH_SUPPORTS_KRETPROBES
 #define flush_insn_slot(p)	do { } while (0)
 #define kretprobe_blacklist_size 0
 
diff --git a/include/asm-s390/kprobes.h b/include/asm-s390/kprobes.h
index 948db3d0d05c..330f68caffe4 100644
--- a/include/asm-s390/kprobes.h
+++ b/include/asm-s390/kprobes.h
@@ -46,7 +46,6 @@ typedef u16 kprobe_opcode_t;
 	? (MAX_STACK_SIZE) \
 	: (((unsigned long)current_thread_info()) + THREAD_SIZE - (ADDR)))
 
-#define ARCH_SUPPORTS_KRETPROBES
 #define kretprobe_blacklist_size 0
 
 #define KPROBE_SWAP_INST	0x10
diff --git a/include/asm-sparc64/kprobes.h b/include/asm-sparc64/kprobes.h
index 7237dd87663e..5879d71afdaa 100644
--- a/include/asm-sparc64/kprobes.h
+++ b/include/asm-sparc64/kprobes.h
@@ -14,8 +14,6 @@ typedef u32 kprobe_opcode_t;
 
 #define arch_remove_kprobe(p)	do {} while (0)
 
-#define ARCH_SUPPORTS_KRETPROBES
-
 #define flush_insn_slot(p)		\
 do { 	flushi(&(p)->ainsn.insn[0]);	\
 	flushi(&(p)->ainsn.insn[1]);	\
diff --git a/include/asm-x86/kprobes.h b/include/asm-x86/kprobes.h
index 143476a3cb52..61ad7b5d142e 100644
--- a/include/asm-x86/kprobes.h
+++ b/include/asm-x86/kprobes.h
@@ -42,7 +42,6 @@ typedef u8 kprobe_opcode_t;
 	: (((unsigned long)current_thread_info()) + THREAD_SIZE \
 	   - (unsigned long)(ADDR)))
 
-#define ARCH_SUPPORTS_KRETPROBES
 #define flush_insn_slot(p)	do { } while (0)
 
 extern const int kretprobe_blacklist_size;
diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h
index 4a6ce82ba039..0f28486f6360 100644
--- a/include/linux/kprobes.h
+++ b/include/linux/kprobes.h
@@ -125,11 +125,11 @@ struct jprobe {
 DECLARE_PER_CPU(struct kprobe *, current_kprobe);
 DECLARE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
 
-#ifdef ARCH_SUPPORTS_KRETPROBES
+#ifdef CONFIG_KRETPROBES
 extern void arch_prepare_kretprobe(struct kretprobe_instance *ri,
 				   struct pt_regs *regs);
 extern int arch_trampoline_kprobe(struct kprobe *p);
-#else /* ARCH_SUPPORTS_KRETPROBES */
+#else /* CONFIG_KRETPROBES */
 static inline void arch_prepare_kretprobe(struct kretprobe *rp,
 					struct pt_regs *regs)
 {
@@ -138,7 +138,7 @@ static inline int arch_trampoline_kprobe(struct kprobe *p)
 {
 	return 0;
 }
-#endif /* ARCH_SUPPORTS_KRETPROBES */
+#endif /* CONFIG_KRETPROBES */
 /*
  * Function-return probe -
  * Note:
diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index 7a86e6432338..e6a61dcbc578 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -678,8 +678,7 @@ void __kprobes unregister_jprobe(struct jprobe *jp)
 	unregister_kprobe(&jp->kp);
 }
 
-#ifdef ARCH_SUPPORTS_KRETPROBES
-
+#ifdef CONFIG_KRETPROBES
 /*
  * This kprobe pre_handler is registered with every kretprobe. When probe
  * hits it will set up the return probe.
@@ -769,8 +768,7 @@ int __kprobes register_kretprobe(struct kretprobe *rp)
 	return ret;
 }
 
-#else /* ARCH_SUPPORTS_KRETPROBES */
-
+#else /* CONFIG_KRETPROBES */
 int __kprobes register_kretprobe(struct kretprobe *rp)
 {
 	return -ENOSYS;
@@ -781,8 +779,7 @@ static int __kprobes pre_handler_kretprobe(struct kprobe *p,
 {
 	return 0;
 }
-
-#endif /* ARCH_SUPPORTS_KRETPROBES */
+#endif /* CONFIG_KRETPROBES */
 
 void __kprobes unregister_kretprobe(struct kretprobe *rp)
 {
-- 
cgit v1.2.3


From 00f0b8259e48979c37212995d798f3fbd0374690 Mon Sep 17 00:00:00 2001
From: Balbir Singh <balbir@linux.vnet.ibm.com>
Date: Tue, 4 Mar 2008 14:28:39 -0800
Subject: Memory controller: rename to Memory Resource Controller

Rename Memory Controller to Memory Resource Controller.  Reflect the same
changes in the CONFIG definition for the Memory Resource Controller.  Group
together the config options for Resource Counters and Memory Resource
Controller.

Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 Documentation/controllers/memory.txt |  8 ++++++--
 include/linux/cgroup_subsys.h        |  2 +-
 include/linux/memcontrol.h           |  4 ++--
 include/linux/mm_types.h             |  4 ++--
 init/Kconfig                         | 30 +++++++++++++++---------------
 mm/Makefile                          |  2 +-
 mm/oom_kill.c                        |  2 +-
 mm/vmscan.c                          |  4 ++--
 8 files changed, 30 insertions(+), 26 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/controllers/memory.txt b/Documentation/controllers/memory.txt
index fba6af45225c..866b9cd9a959 100644
--- a/Documentation/controllers/memory.txt
+++ b/Documentation/controllers/memory.txt
@@ -1,4 +1,8 @@
-Memory Controller
+Memory Resource Controller
+
+NOTE: The Memory Resource Controller has been generically been referred
+to as the memory controller in this document. Do not confuse memory controller
+used here with the memory controller that is used in hardware.
 
 Salient features
 
@@ -152,7 +156,7 @@ The memory controller uses the following hierarchy
 
 a. Enable CONFIG_CGROUPS
 b. Enable CONFIG_RESOURCE_COUNTERS
-c. Enable CONFIG_CGROUP_MEM_CONT
+c. Enable CONFIG_CGROUP_MEM_RES_CTLR
 
 1. Prepare the cgroups
 # mkdir -p /cgroups
diff --git a/include/linux/cgroup_subsys.h b/include/linux/cgroup_subsys.h
index ac6aad98b607..1ddebfc52565 100644
--- a/include/linux/cgroup_subsys.h
+++ b/include/linux/cgroup_subsys.h
@@ -37,7 +37,7 @@ SUBSYS(cpuacct)
 
 /* */
 
-#ifdef CONFIG_CGROUP_MEM_CONT
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
 SUBSYS(mem_cgroup)
 #endif
 
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 04075628cb9a..a8be8073b9e6 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -25,7 +25,7 @@ struct page_cgroup;
 struct page;
 struct mm_struct;
 
-#ifdef CONFIG_CGROUP_MEM_CONT
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
 
 extern void mm_init_cgroup(struct mm_struct *mm, struct task_struct *p);
 extern void mm_free_cgroup(struct mm_struct *mm);
@@ -72,7 +72,7 @@ extern long mem_cgroup_calc_reclaim_active(struct mem_cgroup *mem,
 extern long mem_cgroup_calc_reclaim_inactive(struct mem_cgroup *mem,
 				struct zone *zone, int priority);
 
-#else /* CONFIG_CGROUP_MEM_CONT */
+#else /* CONFIG_CGROUP_MEM_RES_CTLR */
 static inline void mm_init_cgroup(struct mm_struct *mm,
 					struct task_struct *p)
 {
diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
index 34023c65d466..af190ceab971 100644
--- a/include/linux/mm_types.h
+++ b/include/linux/mm_types.h
@@ -88,7 +88,7 @@ struct page {
 	void *virtual;			/* Kernel virtual address (NULL if
 					   not kmapped, ie. highmem) */
 #endif /* WANT_PAGE_VIRTUAL */
-#ifdef CONFIG_CGROUP_MEM_CONT
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
 	unsigned long page_cgroup;
 #endif
 };
@@ -222,7 +222,7 @@ struct mm_struct {
 	/* aio bits */
 	rwlock_t		ioctx_list_lock;
 	struct kioctx		*ioctx_list;
-#ifdef CONFIG_CGROUP_MEM_CONT
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
 	struct mem_cgroup *mem_cgroup;
 #endif
 };
diff --git a/init/Kconfig b/init/Kconfig
index f698a5af5007..442850b984be 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -366,6 +366,21 @@ config RESOURCE_COUNTERS
           infrastructure that works with cgroups
 	depends on CGROUPS
 
+config CGROUP_MEM_RES_CTLR
+	bool "Memory Resource Controller for Control Groups"
+	depends on CGROUPS && RESOURCE_COUNTERS
+	help
+	  Provides a memory resource controller that manages both page cache and
+	  RSS memory.
+
+	  Note that setting this option increases fixed memory overhead
+	  associated with each page of memory in the system by 4/8 bytes
+	  and also increases cache misses because struct page on many 64bit
+	  systems will not fit into a single cache line anymore.
+
+	  Only enable when you're ok with these trade offs and really
+	  sure you need the memory resource controller.
+
 config SYSFS_DEPRECATED
 	bool "Create deprecated sysfs files"
 	depends on SYSFS
@@ -387,21 +402,6 @@ config SYSFS_DEPRECATED
 	  If you are using a distro that was released in 2006 or later,
 	  it should be safe to say N here.
 
-config CGROUP_MEM_CONT
-	bool "Memory controller for cgroups"
-	depends on CGROUPS && RESOURCE_COUNTERS
-	help
-	  Provides a memory controller that manages both page cache and
-	  RSS memory.
-
-	  Note that setting this option increases fixed memory overhead
-	  associated with each page of memory in the system by 4/8 bytes
-	  and also increases cache misses because struct page on many 64bit
-	  systems will not fit into a single cache line anymore.
-
-	  Only enable when you're ok with these trade offs and really
-	  sure you need the memory controller.
-
 config PROC_PID_CPUSET
 	bool "Include legacy /proc/<pid>/cpuset file"
 	depends on CPUSETS
diff --git a/mm/Makefile b/mm/Makefile
index 9f117bab5322..a5b0dd93427a 100644
--- a/mm/Makefile
+++ b/mm/Makefile
@@ -32,5 +32,5 @@ obj-$(CONFIG_FS_XIP) += filemap_xip.o
 obj-$(CONFIG_MIGRATION) += migrate.o
 obj-$(CONFIG_SMP) += allocpercpu.o
 obj-$(CONFIG_QUICKLIST) += quicklist.o
-obj-$(CONFIG_CGROUP_MEM_CONT) += memcontrol.o
+obj-$(CONFIG_CGROUP_MEM_RES_CTLR) += memcontrol.o
 
diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index 4194b9db0104..44b2da11bf43 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -412,7 +412,7 @@ static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order,
 	return oom_kill_task(p);
 }
 
-#ifdef CONFIG_CGROUP_MEM_CONT
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
 void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask)
 {
 	unsigned long points = 0;
diff --git a/mm/vmscan.c b/mm/vmscan.c
index a26dabd62fed..106ba10e1ac6 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -126,7 +126,7 @@ long vm_total_pages;	/* The total number of pages which the VM controls */
 static LIST_HEAD(shrinker_list);
 static DECLARE_RWSEM(shrinker_rwsem);
 
-#ifdef CONFIG_CGROUP_MEM_CONT
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
 #define scan_global_lru(sc)	(!(sc)->mem_cgroup)
 #else
 #define scan_global_lru(sc)	(1)
@@ -1427,7 +1427,7 @@ unsigned long try_to_free_pages(struct zone **zones, int order, gfp_t gfp_mask)
 	return do_try_to_free_pages(zones, gfp_mask, &sc);
 }
 
-#ifdef CONFIG_CGROUP_MEM_CONT
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR
 
 unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont,
 						gfp_t gfp_mask)
-- 
cgit v1.2.3


From 735c4fb916e9f83a9350aeb2680d77d01ea75094 Mon Sep 17 00:00:00 2001
From: Andrew Morton <akpm@linux-foundation.org>
Date: Tue, 4 Mar 2008 14:28:40 -0800
Subject: add noinline_for_stack

People are adding `noinline' in various places to prevent excess stack
consumption due to gcc inlining.  But once this is done, it is quite unobvious
why the `noinline' is present in the code.  We can comment each and every
site, or we can use noinline_for_stack.

Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/compiler.h | 6 ++++++
 1 file changed, 6 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/compiler.h b/include/linux/compiler.h
index d0e17e1657dc..dcae0c8d97e6 100644
--- a/include/linux/compiler.h
+++ b/include/linux/compiler.h
@@ -138,6 +138,12 @@ extern void __chk_io_ptr(const volatile void __iomem *);
 #define noinline
 #endif
 
+/*
+ * Rather then using noinline to prevent stack consumption, use
+ * noinline_for_stack instead.  For documentaiton reasons.
+ */
+#define noinline_for_stack noinline
+
 #ifndef __always_inline
 #define __always_inline inline
 #endif
-- 
cgit v1.2.3


From 5cba6d22e35a05adb28fdea191b232501518c455 Mon Sep 17 00:00:00 2001
From: Andrew Morton <akpm@linux-foundation.org>
Date: Tue, 4 Mar 2008 14:28:45 -0800
Subject: ndelay(): switch to C function to avoid 64-bit division

We should be able to do ndelay(some_u64), but that can cause a call to
__divdi3() to be emitted because the ndelay() macros does a divide.

Fix it by switching to static inline which will force the u64 arg to be
treated as an unsigned long.  udelay() takes an unsigned long arg.

[bunk@kernel.org: reported m68k build breakage]
Cc: Adrian Bunk <bunk@kernel.org>
Cc: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Cc: Martin Michlmayr <tbm@cyrius.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/delay.h | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/delay.h b/include/linux/delay.h
index 17ddb55430ae..54552d21296e 100644
--- a/include/linux/delay.h
+++ b/include/linux/delay.h
@@ -7,6 +7,8 @@
  * Delay routines, using a pre-computed "loops_per_jiffy" value.
  */
 
+#include <linux/kernel.h>
+
 extern unsigned long loops_per_jiffy;
 
 #include <asm/delay.h>
@@ -32,7 +34,11 @@ extern unsigned long loops_per_jiffy;
 #endif
 
 #ifndef ndelay
-#define ndelay(x)	udelay(((x)+999)/1000)
+static inline void ndelay(unsigned long x)
+{
+	udelay(DIV_ROUND_UP(x, 1000));
+}
+#define ndelay(x) ndelay(x)
 #endif
 
 void calibrate_delay(void);
-- 
cgit v1.2.3


From 3149be50d3a31df095bcc83d752293da65a37f62 Mon Sep 17 00:00:00 2001
From: Ville Syrjala <syrjala@sci.fi>
Date: Tue, 4 Mar 2008 14:28:50 -0800
Subject: sm501: add support for the SM502 programmable PLL

SM502 has a programmable PLL which can provide the panel pixel clock instead
of the 288MHz and 336MHz PLLs.

[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Ville Syrjala <syrjala@sci.fi>
Cc: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/mfd/sm501.c        | 163 +++++++++++++++++++++++++++++++++++----------
 include/linux/sm501-regs.h |   3 +
 include/linux/sm501.h      |   3 +-
 3 files changed, 133 insertions(+), 36 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/mfd/sm501.c b/drivers/mfd/sm501.c
index 4de8d467762a..13bac53db69a 100644
--- a/drivers/mfd/sm501.c
+++ b/drivers/mfd/sm501.c
@@ -48,6 +48,7 @@ struct sm501_devdata {
 	unsigned int			 pdev_id;
 	unsigned int			 irq;
 	void __iomem			*regs;
+	unsigned int			 rev;
 };
 
 #define MHZ (1000 * 1000)
@@ -417,46 +418,108 @@ struct sm501_clock {
 	unsigned long mclk;
 	int divider;
 	int shift;
+	unsigned int m, n, k;
 };
 
+/* sm501_calc_clock
+ *
+ * Calculates the nearest discrete clock frequency that
+ * can be achieved with the specified input clock.
+ *   the maximum divisor is 3 or 5
+ */
+
+static int sm501_calc_clock(unsigned long freq,
+			    struct sm501_clock *clock,
+			    int max_div,
+			    unsigned long mclk,
+			    long *best_diff)
+{
+	int ret = 0;
+	int divider;
+	int shift;
+	long diff;
+
+	/* try dividers 1 and 3 for CRT and for panel,
+	   try divider 5 for panel only.*/
+
+	for (divider = 1; divider <= max_div; divider += 2) {
+		/* try all 8 shift values.*/
+		for (shift = 0; shift < 8; shift++) {
+			/* Calculate difference to requested clock */
+			diff = sm501fb_round_div(mclk, divider << shift) - freq;
+			if (diff < 0)
+				diff = -diff;
+
+			/* If it is less than the current, use it */
+			if (diff < *best_diff) {
+				*best_diff = diff;
+
+				clock->mclk = mclk;
+				clock->divider = divider;
+				clock->shift = shift;
+				ret = 1;
+			}
+		}
+	}
+
+	return ret;
+}
+
+/* sm501_calc_pll
+ *
+ * Calculates the nearest discrete clock frequency that can be
+ * achieved using the programmable PLL.
+ *   the maximum divisor is 3 or 5
+ */
+
+static unsigned long sm501_calc_pll(unsigned long freq,
+					struct sm501_clock *clock,
+					int max_div)
+{
+	unsigned long mclk;
+	unsigned int m, n, k;
+	long best_diff = 999999999;
+
+	/*
+	 * The SM502 datasheet doesn't specify the min/max values for M and N.
+	 * N = 1 at least doesn't work in practice.
+	 */
+	for (m = 2; m <= 255; m++) {
+		for (n = 2; n <= 127; n++) {
+			for (k = 0; k <= 1; k++) {
+				mclk = (24000000UL * m / n) >> k;
+
+				if (sm501_calc_clock(freq, clock, max_div,
+						     mclk, &best_diff)) {
+					clock->m = m;
+					clock->n = n;
+					clock->k = k;
+				}
+			}
+		}
+	}
+
+	/* Return best clock. */
+	return clock->mclk / (clock->divider << clock->shift);
+}
+
 /* sm501_select_clock
  *
- * selects nearest discrete clock frequency the SM501 can achive
+ * Calculates the nearest discrete clock frequency that can be
+ * achieved using the 288MHz and 336MHz PLLs.
  *   the maximum divisor is 3 or 5
  */
+
 static unsigned long sm501_select_clock(unsigned long freq,
 					struct sm501_clock *clock,
 					int max_div)
 {
 	unsigned long mclk;
-	int divider;
-	int shift;
-	long diff;
 	long best_diff = 999999999;
 
 	/* Try 288MHz and 336MHz clocks. */
 	for (mclk = 288000000; mclk <= 336000000; mclk += 48000000) {
-		/* try dividers 1 and 3 for CRT and for panel,
-		   try divider 5 for panel only.*/
-
-		for (divider = 1; divider <= max_div; divider += 2) {
-			/* try all 8 shift values.*/
-			for (shift = 0; shift < 8; shift++) {
-				/* Calculate difference to requested clock */
-				diff = sm501fb_round_div(mclk, divider << shift) - freq;
-				if (diff < 0)
-					diff = -diff;
-
-				/* If it is less than the current, use it */
-				if (diff < best_diff) {
-					best_diff = diff;
-
-					clock->mclk = mclk;
-					clock->divider = divider;
-					clock->shift = shift;
-				}
-			}
-		}
+		sm501_calc_clock(freq, clock, max_div, mclk, &best_diff);
 	}
 
 	/* Return best clock. */
@@ -478,6 +541,7 @@ unsigned long sm501_set_clock(struct device *dev,
 	unsigned long gate = readl(sm->regs + SM501_CURRENT_GATE);
 	unsigned long clock = readl(sm->regs + SM501_CURRENT_CLOCK);
 	unsigned char reg;
+	unsigned int pll_reg = 0;
 	unsigned long sm501_freq; /* the actual frequency acheived */
 
 	struct sm501_clock to;
@@ -492,14 +556,28 @@ unsigned long sm501_set_clock(struct device *dev,
 		 * requested frequency the value must be multiplied by
 		 * 2. This clock also has an additional pre divisor */
 
-		sm501_freq = (sm501_select_clock(2 * req_freq, &to, 5) / 2);
-		reg=to.shift & 0x07;/* bottom 3 bits are shift */
-		if (to.divider == 3)
-			reg |= 0x08; /* /3 divider required */
-		else if (to.divider == 5)
-			reg |= 0x10; /* /5 divider required */
-		if (to.mclk != 288000000)
-			reg |= 0x20; /* which mclk pll is source */
+		if (sm->rev >= 0xC0) {
+			/* SM502 -> use the programmable PLL */
+			sm501_freq = (sm501_calc_pll(2 * req_freq,
+						     &to, 5) / 2);
+			reg = to.shift & 0x07;/* bottom 3 bits are shift */
+			if (to.divider == 3)
+				reg |= 0x08; /* /3 divider required */
+			else if (to.divider == 5)
+				reg |= 0x10; /* /5 divider required */
+			reg |= 0x40; /* select the programmable PLL */
+			pll_reg = 0x20000 | (to.k << 15) | (to.n << 8) | to.m;
+		} else {
+			sm501_freq = (sm501_select_clock(2 * req_freq,
+							 &to, 5) / 2);
+			reg = to.shift & 0x07;/* bottom 3 bits are shift */
+			if (to.divider == 3)
+				reg |= 0x08; /* /3 divider required */
+			else if (to.divider == 5)
+				reg |= 0x10; /* /5 divider required */
+			if (to.mclk != 288000000)
+				reg |= 0x20; /* which mclk pll is source */
+		}
 		break;
 
 	case SM501_CLOCK_V2XCLK:
@@ -560,6 +638,10 @@ unsigned long sm501_set_clock(struct device *dev,
 	}
 
 	writel(mode, sm->regs + SM501_POWER_MODE_CONTROL);
+
+	if (pll_reg)
+		writel(pll_reg, sm->regs + SM501_PROGRAMMABLE_PLL_CONTROL);
+
 	sm501_sync_regs(sm);
 
 	dev_info(sm->dev, "gate %08lx, clock %08lx, mode %08lx\n",
@@ -580,15 +662,24 @@ EXPORT_SYMBOL_GPL(sm501_set_clock);
  * finds the closest available frequency for a given clock
 */
 
-unsigned long sm501_find_clock(int clksrc,
+unsigned long sm501_find_clock(struct device *dev,
+			       int clksrc,
 			       unsigned long req_freq)
 {
+	struct sm501_devdata *sm = dev_get_drvdata(dev);
 	unsigned long sm501_freq; /* the frequency achiveable by the 501 */
 	struct sm501_clock to;
 
 	switch (clksrc) {
 	case SM501_CLOCK_P2XCLK:
-		sm501_freq = (sm501_select_clock(2 * req_freq, &to, 5) / 2);
+		if (sm->rev >= 0xC0) {
+			/* SM502 -> use the programmable PLL */
+			sm501_freq = (sm501_calc_pll(2 * req_freq,
+						     &to, 5) / 2);
+		} else {
+			sm501_freq = (sm501_select_clock(2 * req_freq,
+							 &to, 5) / 2);
+		}
 		break;
 
 	case SM501_CLOCK_V2XCLK:
@@ -895,6 +986,8 @@ static int sm501_init_dev(struct sm501_devdata *sm)
 	dev_info(sm->dev, "SM501 At %p: Version %08lx, %ld Mb, IRQ %d\n",
 		 sm->regs, devid, (unsigned long)mem_avail >> 20, sm->irq);
 
+	sm->rev = devid & SM501_DEVICEID_REVMASK;
+
 	sm501_dump_gate(sm);
 
 	ret = device_create_file(sm->dev, &dev_attr_dbg_regs);
diff --git a/include/linux/sm501-regs.h b/include/linux/sm501-regs.h
index 64236b73c724..d53642d2d899 100644
--- a/include/linux/sm501-regs.h
+++ b/include/linux/sm501-regs.h
@@ -129,11 +129,14 @@
 
 #define SM501_DEVICEID_SM501		(0x05010000)
 #define SM501_DEVICEID_IDMASK		(0xffff0000)
+#define SM501_DEVICEID_REVMASK		(0x000000ff)
 
 #define SM501_PLLCLOCK_COUNT		(0x000064)
 #define SM501_MISC_TIMING		(0x000068)
 #define SM501_CURRENT_SDRAM_CLOCK	(0x00006C)
 
+#define SM501_PROGRAMMABLE_PLL_CONTROL	(0x000074)
+
 /* GPIO base */
 #define SM501_GPIO			(0x010000)
 #define SM501_GPIO_DATA_LOW		(0x00)
diff --git a/include/linux/sm501.h b/include/linux/sm501.h
index 932a9efee8a5..bca134544700 100644
--- a/include/linux/sm501.h
+++ b/include/linux/sm501.h
@@ -24,7 +24,8 @@ extern int sm501_unit_power(struct device *dev,
 extern unsigned long sm501_set_clock(struct device *dev,
 				     int clksrc, unsigned long freq);
 
-extern unsigned long sm501_find_clock(int clksrc, unsigned long req_freq);
+extern unsigned long sm501_find_clock(struct device *dev,
+				      int clksrc, unsigned long req_freq);
 
 /* sm501_misc_control
  *
-- 
cgit v1.2.3


From 040922c04cf2c8ac70be2e88a8a9614ecdb41d2e Mon Sep 17 00:00:00 2001
From: Eric Sandeen <sandeen@redhat.com>
Date: Tue, 4 Mar 2008 14:28:53 -0800
Subject: include falloc.h in header-y

Include falloc.h in header-y; it defines a flag for the fallocate sysctl.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/Kbuild | 1 +
 1 file changed, 1 insertion(+)

(limited to 'include/linux')

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index aada32fffec2..994df3780007 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -61,6 +61,7 @@ header-y += efs_fs_sb.h
 header-y += elf-fdpic.h
 header-y += elf-em.h
 header-y += fadvise.h
+header-y += falloc.h
 header-y += fd.h
 header-y += fdreg.h
 header-y += fib_rules.h
-- 
cgit v1.2.3


From acc4988bcf38f9618886eaeb9802aeacc6978ec2 Mon Sep 17 00:00:00 2001
From: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Date: Tue, 4 Mar 2008 14:29:00 -0800
Subject: markers: add an if(0) to __mark_check_format()

Wrap __mark_check_format() into an if(0) to make sure that parameters such as

trace_mark(mm_page_alloc, "order %u pfn %lu", order, page?page_to_pfn(page):0);

(where page_to_pfn() has side-effects) won't generate code because of the
__mark_check_format().

Thanks to Jan Kiszka for reporting this.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Cc: Jan Kiszka <jan.kiszka@siemens.com>
Cc: "Frank Ch. Eigler" <fche@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/marker.h | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/marker.h b/include/linux/marker.h
index 5df879dc3776..430f6adf9762 100644
--- a/include/linux/marker.h
+++ b/include/linux/marker.h
@@ -104,10 +104,16 @@ static inline void marker_update_probe_range(struct marker *begin,
 #define MARK_NOARGS " "
 
 /* To be used for string format validity checking with gcc */
-static inline void __printf(1, 2) __mark_check_format(const char *fmt, ...)
+static inline void __printf(1, 2) ___mark_check_format(const char *fmt, ...)
 {
 }
 
+#define __mark_check_format(format, args...)				\
+	do {								\
+		if (0)							\
+			___mark_check_format(format, ## args);		\
+	} while (0)
+
 extern marker_probe_func __mark_empty_function;
 
 extern void marker_probe_cb(const struct marker *mdata,
-- 
cgit v1.2.3


From bd845e38c7a7251a95a8f2c38aa7fb87140b771d Mon Sep 17 00:00:00 2001
From: Hugh Dickins <hugh@veritas.com>
Date: Tue, 4 Mar 2008 14:29:01 -0800
Subject: memcg: mm_match_cgroup not vm_match_cgroup

vm_match_cgroup is a perverse name for a macro to match mm with cgroup: rename
it mm_match_cgroup, matching mm_init_cgroup and mm_free_cgroup.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Hirokazu Takahashi <taka@valinux.co.jp>
Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/memcontrol.h | 4 ++--
 mm/memcontrol.c            | 2 +-
 mm/rmap.c                  | 4 ++--
 3 files changed, 5 insertions(+), 5 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index a8be8073b9e6..e4247c83c1c7 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -48,7 +48,7 @@ extern int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
 					gfp_t gfp_mask);
 int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem);
 
-#define vm_match_cgroup(mm, cgroup)	\
+#define mm_match_cgroup(mm, cgroup)	\
 	((cgroup) == rcu_dereference((mm)->mem_cgroup))
 
 extern int mem_cgroup_prepare_migration(struct page *page);
@@ -118,7 +118,7 @@ static inline int mem_cgroup_cache_charge(struct page *page,
 	return 0;
 }
 
-static inline int vm_match_cgroup(struct mm_struct *mm, struct mem_cgroup *mem)
+static inline int mm_match_cgroup(struct mm_struct *mm, struct mem_cgroup *mem)
 {
 	return 1;
 }
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 631002d085d1..41041c0a6898 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -399,7 +399,7 @@ int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem)
 	int ret;
 
 	task_lock(task);
-	ret = task->mm && vm_match_cgroup(task->mm, mem);
+	ret = task->mm && mm_match_cgroup(task->mm, mem);
 	task_unlock(task);
 	return ret;
 }
diff --git a/mm/rmap.c b/mm/rmap.c
index 8fd527c4e2bf..0c9a2df06c39 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -321,7 +321,7 @@ static int page_referenced_anon(struct page *page,
 		 * counting on behalf of references from different
 		 * cgroups
 		 */
-		if (mem_cont && !vm_match_cgroup(vma->vm_mm, mem_cont))
+		if (mem_cont && !mm_match_cgroup(vma->vm_mm, mem_cont))
 			continue;
 		referenced += page_referenced_one(page, vma, &mapcount);
 		if (!mapcount)
@@ -382,7 +382,7 @@ static int page_referenced_file(struct page *page,
 		 * counting on behalf of references from different
 		 * cgroups
 		 */
-		if (mem_cont && !vm_match_cgroup(vma->vm_mm, mem_cont))
+		if (mem_cont && !mm_match_cgroup(vma->vm_mm, mem_cont))
 			continue;
 		if ((vma->vm_flags & (VM_LOCKED|VM_MAYSHARE))
 				  == (VM_LOCKED|VM_MAYSHARE)) {
-- 
cgit v1.2.3


From 427d5416f317681498337ab19218d195edea02d6 Mon Sep 17 00:00:00 2001
From: Hugh Dickins <hugh@veritas.com>
Date: Tue, 4 Mar 2008 14:29:03 -0800
Subject: memcg: move_lists on page not page_cgroup

Each caller of mem_cgroup_move_lists is having to use page_get_page_cgroup:
it's more convenient if it acts upon the page itself not the page_cgroup; and
in a later patch this becomes important to handle within memcontrol.c.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Cc: David Rientjes <rientjes@google.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Hirokazu Takahashi <taka@valinux.co.jp>
Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/memcontrol.h | 5 ++---
 mm/memcontrol.c            | 4 +++-
 mm/swap.c                  | 2 +-
 mm/vmscan.c                | 5 +++--
 4 files changed, 9 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index e4247c83c1c7..56432ff8d4e3 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -36,7 +36,7 @@ extern int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
 				gfp_t gfp_mask);
 extern void mem_cgroup_uncharge(struct page_cgroup *pc);
 extern void mem_cgroup_uncharge_page(struct page *page);
-extern void mem_cgroup_move_lists(struct page_cgroup *pc, bool active);
+extern void mem_cgroup_move_lists(struct page *page, bool active);
 extern unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan,
 					struct list_head *dst,
 					unsigned long *scanned, int order,
@@ -106,8 +106,7 @@ static inline void mem_cgroup_uncharge_page(struct page *page)
 {
 }
 
-static inline void mem_cgroup_move_lists(struct page_cgroup *pc,
-						bool active)
+static inline void mem_cgroup_move_lists(struct page *page, bool active)
 {
 }
 
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 41041c0a6898..afdd406f618a 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -407,11 +407,13 @@ int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem)
 /*
  * This routine assumes that the appropriate zone's lru lock is already held
  */
-void mem_cgroup_move_lists(struct page_cgroup *pc, bool active)
+void mem_cgroup_move_lists(struct page *page, bool active)
 {
+	struct page_cgroup *pc;
 	struct mem_cgroup_per_zone *mz;
 	unsigned long flags;
 
+	pc = page_get_page_cgroup(page);
 	if (!pc)
 		return;
 
diff --git a/mm/swap.c b/mm/swap.c
index 710a20bb9749..d4ec59aa5c46 100644
--- a/mm/swap.c
+++ b/mm/swap.c
@@ -176,7 +176,7 @@ void activate_page(struct page *page)
 		SetPageActive(page);
 		add_page_to_active_list(zone, page);
 		__count_vm_event(PGACTIVATE);
-		mem_cgroup_move_lists(page_get_page_cgroup(page), true);
+		mem_cgroup_move_lists(page, true);
 	}
 	spin_unlock_irq(&zone->lru_lock);
 }
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 106ba10e1ac6..45711585684e 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1128,7 +1128,7 @@ static void shrink_active_list(unsigned long nr_pages, struct zone *zone,
 		ClearPageActive(page);
 
 		list_move(&page->lru, &zone->inactive_list);
-		mem_cgroup_move_lists(page_get_page_cgroup(page), false);
+		mem_cgroup_move_lists(page, false);
 		pgmoved++;
 		if (!pagevec_add(&pvec, page)) {
 			__mod_zone_page_state(zone, NR_INACTIVE, pgmoved);
@@ -1156,8 +1156,9 @@ static void shrink_active_list(unsigned long nr_pages, struct zone *zone,
 		VM_BUG_ON(PageLRU(page));
 		SetPageLRU(page);
 		VM_BUG_ON(!PageActive(page));
+
 		list_move(&page->lru, &zone->active_list);
-		mem_cgroup_move_lists(page_get_page_cgroup(page), true);
+		mem_cgroup_move_lists(page, true);
 		pgmoved++;
 		if (!pagevec_add(&pvec, page)) {
 			__mod_zone_page_state(zone, NR_ACTIVE, pgmoved);
-- 
cgit v1.2.3


From 9442ec9df40d952b0de185ae5638a74970388e01 Mon Sep 17 00:00:00 2001
From: Hugh Dickins <hugh@veritas.com>
Date: Tue, 4 Mar 2008 14:29:07 -0800
Subject: memcg: bad page if page_cgroup when free

Replace free_hot_cold_page's VM_BUG_ON(page_get_page_cgroup(page)) by a "Bad
page state" and clear: most users don't have CONFIG_DEBUG_VM on, and if it
were set here, it'd likely cause corruption when the page is reused.

Don't use page_assign_page_cgroup to clear it: that should be private to
memcontrol.c, and always called with the lock taken; and memmap_init_zone
doesn't need it either - like page->mapping and other pointers throughout the
kernel, Linux assumes pointers in zeroed structures are NULL pointers.

Instead use page_reset_bad_cgroup, added to memcontrol.h for this only.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Cc: David Rientjes <rientjes@google.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Hirokazu Takahashi <taka@valinux.co.jp>
Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/memcontrol.h |  8 ++++----
 mm/memcontrol.c            | 27 ++++++++++++---------------
 mm/page_alloc.c            | 18 ++++++++++++------
 3 files changed, 28 insertions(+), 25 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 56432ff8d4e3..70789df7dab4 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -29,8 +29,9 @@ struct mm_struct;
 
 extern void mm_init_cgroup(struct mm_struct *mm, struct task_struct *p);
 extern void mm_free_cgroup(struct mm_struct *mm);
-extern void page_assign_page_cgroup(struct page *page,
-					struct page_cgroup *pc);
+
+#define page_reset_bad_cgroup(page)	((page)->page_cgroup = 0)
+
 extern struct page_cgroup *page_get_page_cgroup(struct page *page);
 extern int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
 				gfp_t gfp_mask);
@@ -82,8 +83,7 @@ static inline void mm_free_cgroup(struct mm_struct *mm)
 {
 }
 
-static inline void page_assign_page_cgroup(struct page *page,
-						struct page_cgroup *pc)
+static inline void page_reset_bad_cgroup(struct page *page)
 {
 }
 
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index afdd406f618a..9e170d3c71e5 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -140,11 +140,17 @@ struct mem_cgroup {
 
 /*
  * We use the lower bit of the page->page_cgroup pointer as a bit spin
- * lock. We need to ensure that page->page_cgroup is atleast two
- * byte aligned (based on comments from Nick Piggin)
+ * lock.  We need to ensure that page->page_cgroup is at least two
+ * byte aligned (based on comments from Nick Piggin).  But since
+ * bit_spin_lock doesn't actually set that lock bit in a non-debug
+ * uniprocessor kernel, we should avoid setting it here too.
  */
 #define PAGE_CGROUP_LOCK_BIT 	0x0
-#define PAGE_CGROUP_LOCK 		(1 << PAGE_CGROUP_LOCK_BIT)
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK)
+#define PAGE_CGROUP_LOCK 	(1 << PAGE_CGROUP_LOCK_BIT)
+#else
+#define PAGE_CGROUP_LOCK	0x0
+#endif
 
 /*
  * A page_cgroup page is associated with every page descriptor. The
@@ -271,19 +277,10 @@ static inline int page_cgroup_locked(struct page *page)
 					&page->page_cgroup);
 }
 
-void page_assign_page_cgroup(struct page *page, struct page_cgroup *pc)
+static void page_assign_page_cgroup(struct page *page, struct page_cgroup *pc)
 {
-	int locked;
-
-	/*
-	 * While resetting the page_cgroup we might not hold the
-	 * page_cgroup lock. free_hot_cold_page() is an example
-	 * of such a scenario
-	 */
-	if (pc)
-		VM_BUG_ON(!page_cgroup_locked(page));
-	locked = (page->page_cgroup & PAGE_CGROUP_LOCK);
-	page->page_cgroup = ((unsigned long)pc | locked);
+	VM_BUG_ON(!page_cgroup_locked(page));
+	page->page_cgroup = ((unsigned long)pc | PAGE_CGROUP_LOCK);
 }
 
 struct page_cgroup *page_get_page_cgroup(struct page *page)
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index e76cf94725c9..402a504f1228 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -222,13 +222,19 @@ static inline int bad_range(struct zone *zone, struct page *page)
 
 static void bad_page(struct page *page)
 {
-	printk(KERN_EMERG "Bad page state in process '%s'\n"
-		KERN_EMERG "page:%p flags:0x%0*lx mapping:%p mapcount:%d count:%d\n"
-		KERN_EMERG "Trying to fix it up, but a reboot is needed\n"
-		KERN_EMERG "Backtrace:\n",
+	void *pc = page_get_page_cgroup(page);
+
+	printk(KERN_EMERG "Bad page state in process '%s'\n" KERN_EMERG
+		"page:%p flags:0x%0*lx mapping:%p mapcount:%d count:%d\n",
 		current->comm, page, (int)(2*sizeof(unsigned long)),
 		(unsigned long)page->flags, page->mapping,
 		page_mapcount(page), page_count(page));
+	if (pc) {
+		printk(KERN_EMERG "cgroup:%p\n", pc);
+		page_reset_bad_cgroup(page);
+	}
+	printk(KERN_EMERG "Trying to fix it up, but a reboot is needed\n"
+		KERN_EMERG "Backtrace:\n");
 	dump_stack();
 	page->flags &= ~(1 << PG_lru	|
 			1 << PG_private |
@@ -454,6 +460,7 @@ static inline int free_pages_check(struct page *page)
 {
 	if (unlikely(page_mapcount(page) |
 		(page->mapping != NULL)  |
+		(page_get_page_cgroup(page) != NULL) |
 		(page_count(page) != 0)  |
 		(page->flags & (
 			1 << PG_lru	|
@@ -603,6 +610,7 @@ static int prep_new_page(struct page *page, int order, gfp_t gfp_flags)
 {
 	if (unlikely(page_mapcount(page) |
 		(page->mapping != NULL)  |
+		(page_get_page_cgroup(page) != NULL) |
 		(page_count(page) != 0)  |
 		(page->flags & (
 			1 << PG_lru	|
@@ -989,7 +997,6 @@ static void free_hot_cold_page(struct page *page, int cold)
 
 	if (!PageHighMem(page))
 		debug_check_no_locks_freed(page_address(page), PAGE_SIZE);
-	VM_BUG_ON(page_get_page_cgroup(page));
 	arch_free_page(page, 0);
 	kernel_map_pages(page, 1, 0);
 
@@ -2528,7 +2535,6 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,
 		set_page_links(page, zone, nid, pfn);
 		init_page_count(page);
 		reset_page_mapcount(page);
-		page_assign_page_cgroup(page, NULL);
 		SetPageReserved(page);
 
 		/*
-- 
cgit v1.2.3


From 8289546e573d5ff681cdf0fc7a1184cca66fdb55 Mon Sep 17 00:00:00 2001
From: Hugh Dickins <hugh@veritas.com>
Date: Tue, 4 Mar 2008 14:29:08 -0800
Subject: memcg: remove mem_cgroup_uncharge

Nothing uses mem_cgroup_uncharge apart from mem_cgroup_uncharge_page, (a
trivial wrapper around it) and mem_cgroup_end_migration (which does the same
as mem_cgroup_uncharge_page).  And it often ends up having to lock just to let
its caller unlock.  Remove it (but leave the silly locking until a later
patch).

Moved mem_cgroup_cache_charge next to mem_cgroup_charge in memcontrol.h.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Cc: David Rientjes <rientjes@google.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Hirokazu Takahashi <taka@valinux.co.jp>
Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/memcontrol.h | 20 +++++++-------------
 mm/memcontrol.c            | 23 ++++++++---------------
 2 files changed, 15 insertions(+), 28 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 70789df7dab4..8b1c4295848b 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -35,7 +35,8 @@ extern void mm_free_cgroup(struct mm_struct *mm);
 extern struct page_cgroup *page_get_page_cgroup(struct page *page);
 extern int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
 				gfp_t gfp_mask);
-extern void mem_cgroup_uncharge(struct page_cgroup *pc);
+extern int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
+					gfp_t gfp_mask);
 extern void mem_cgroup_uncharge_page(struct page *page);
 extern void mem_cgroup_move_lists(struct page *page, bool active);
 extern unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan,
@@ -45,8 +46,6 @@ extern unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan,
 					struct mem_cgroup *mem_cont,
 					int active);
 extern void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask);
-extern int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
-					gfp_t gfp_mask);
 int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem);
 
 #define mm_match_cgroup(mm, cgroup)	\
@@ -92,14 +91,16 @@ static inline struct page_cgroup *page_get_page_cgroup(struct page *page)
 	return NULL;
 }
 
-static inline int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
-					gfp_t gfp_mask)
+static inline int mem_cgroup_charge(struct page *page,
+					struct mm_struct *mm, gfp_t gfp_mask)
 {
 	return 0;
 }
 
-static inline void mem_cgroup_uncharge(struct page_cgroup *pc)
+static inline int mem_cgroup_cache_charge(struct page *page,
+					struct mm_struct *mm, gfp_t gfp_mask)
 {
+	return 0;
 }
 
 static inline void mem_cgroup_uncharge_page(struct page *page)
@@ -110,13 +111,6 @@ static inline void mem_cgroup_move_lists(struct page *page, bool active)
 {
 }
 
-static inline int mem_cgroup_cache_charge(struct page *page,
-						struct mm_struct *mm,
-						gfp_t gfp_mask)
-{
-	return 0;
-}
-
 static inline int mm_match_cgroup(struct mm_struct *mm, struct mem_cgroup *mem)
 {
 	return 1;
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 83ba13ad31e1..1333d25163bb 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -697,20 +697,22 @@ int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm,
 
 /*
  * Uncharging is always a welcome operation, we never complain, simply
- * uncharge. This routine should be called with lock_page_cgroup held
+ * uncharge.
  */
-void mem_cgroup_uncharge(struct page_cgroup *pc)
+void mem_cgroup_uncharge_page(struct page *page)
 {
+	struct page_cgroup *pc;
 	struct mem_cgroup *mem;
 	struct mem_cgroup_per_zone *mz;
-	struct page *page;
 	unsigned long flags;
 
 	/*
 	 * Check if our page_cgroup is valid
 	 */
+	lock_page_cgroup(page);
+	pc = page_get_page_cgroup(page);
 	if (!pc)
-		return;
+		goto unlock;
 
 	if (atomic_dec_and_test(&pc->ref_cnt)) {
 		page = pc->page;
@@ -731,12 +733,8 @@ void mem_cgroup_uncharge(struct page_cgroup *pc)
 		}
 		lock_page_cgroup(page);
 	}
-}
 
-void mem_cgroup_uncharge_page(struct page *page)
-{
-	lock_page_cgroup(page);
-	mem_cgroup_uncharge(page_get_page_cgroup(page));
+unlock:
 	unlock_page_cgroup(page);
 }
 
@@ -759,12 +757,7 @@ int mem_cgroup_prepare_migration(struct page *page)
 
 void mem_cgroup_end_migration(struct page *page)
 {
-	struct page_cgroup *pc;
-
-	lock_page_cgroup(page);
-	pc = page_get_page_cgroup(page);
-	mem_cgroup_uncharge(pc);
-	unlock_page_cgroup(page);
+	mem_cgroup_uncharge_page(page);
 }
 /*
  * We know both *page* and *newpage* are now not-on-LRU and Pg_locked.
-- 
cgit v1.2.3


From 3715863aa142c4f4c5208f5f3e5e9bac06006d2f Mon Sep 17 00:00:00 2001
From: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Date: Tue, 4 Mar 2008 14:29:27 -0800
Subject: iommu: export iommu_is_span_boundary helper function

iommu_is_span_boundary is used internally in the IOMMU helper
(lib/iommu-helper.c), a primitive function that judges whether a memory area
spans LLD's segment boundary or not.

It's difficult to convert some IOMMUs to use the IOMMU helper but
iommu_is_span_boundary is still useful for them.  So this patch exports it.

This is needed for the parisc iommu fixes.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Kyle McMartin <kyle@parisc-linux.org>
Cc: Matthew Wilcox <matthew@wil.cx>
Cc: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/iommu-helper.h |  3 +++
 lib/iommu-helper.c           | 10 ++++++----
 2 files changed, 9 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/iommu-helper.h b/include/linux/iommu-helper.h
index 4dd4c04ff2f4..c975caf75385 100644
--- a/include/linux/iommu-helper.h
+++ b/include/linux/iommu-helper.h
@@ -1,3 +1,6 @@
+extern int iommu_is_span_boundary(unsigned int index, unsigned int nr,
+				  unsigned long shift,
+				  unsigned long boundary_size);
 extern unsigned long iommu_area_alloc(unsigned long *map, unsigned long size,
 				      unsigned long start, unsigned int nr,
 				      unsigned long shift,
diff --git a/lib/iommu-helper.c b/lib/iommu-helper.c
index 495575a59ca6..a3b8d4c3f77a 100644
--- a/lib/iommu-helper.c
+++ b/lib/iommu-helper.c
@@ -40,10 +40,12 @@ static inline void set_bit_area(unsigned long *map, unsigned long i,
 	}
 }
 
-static inline int is_span_boundary(unsigned int index, unsigned int nr,
-				   unsigned long shift,
-				   unsigned long boundary_size)
+int iommu_is_span_boundary(unsigned int index, unsigned int nr,
+			   unsigned long shift,
+			   unsigned long boundary_size)
 {
+	BUG_ON(!is_power_of_2(boundary_size));
+
 	shift = (shift + index) & (boundary_size - 1);
 	return shift + nr > boundary_size;
 }
@@ -57,7 +59,7 @@ unsigned long iommu_area_alloc(unsigned long *map, unsigned long size,
 again:
 	index = find_next_zero_area(map, size, start, nr, align_mask);
 	if (index != -1) {
-		if (is_span_boundary(index, nr, shift, boundary_size)) {
+		if (iommu_is_span_boundary(index, nr, shift, boundary_size)) {
 			/* we could do more effectively */
 			start = index + 1;
 			goto again;
-- 
cgit v1.2.3


From 8311c29d40235062a843f4a8e8a70a44af6fe4c9 Mon Sep 17 00:00:00 2001
From: NeilBrown <neilb@suse.de>
Date: Tue, 4 Mar 2008 14:29:30 -0800
Subject: md: reduce CPU wastage on idle md array with a write-intent bitmap

On an md array with a write-intent bitmap, a thread wakes up every few seconds
and scans the bitmap looking for work to do.  If the array is idle, there will
be no work to do, but a lot of scanning is done to discover this.

So cache the fact that the bitmap is completely clean, and avoid scanning the
whole bitmap when the cache is known to be clean.

Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/md/bitmap.c         | 19 +++++++++++++++++--
 include/linux/raid/bitmap.h |  2 ++
 2 files changed, 19 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c
index 7aeceedcf7d4..831aed9c56ff 100644
--- a/drivers/md/bitmap.c
+++ b/drivers/md/bitmap.c
@@ -1047,6 +1047,11 @@ void bitmap_daemon_work(struct bitmap *bitmap)
 	if (time_before(jiffies, bitmap->daemon_lastrun + bitmap->daemon_sleep*HZ))
 		return;
 	bitmap->daemon_lastrun = jiffies;
+	if (bitmap->allclean) {
+		bitmap->mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
+		return;
+	}
+	bitmap->allclean = 1;
 
 	for (j = 0; j < bitmap->chunks; j++) {
 		bitmap_counter_t *bmc;
@@ -1068,8 +1073,10 @@ void bitmap_daemon_work(struct bitmap *bitmap)
 					clear_page_attr(bitmap, page, BITMAP_PAGE_NEEDWRITE);
 
 				spin_unlock_irqrestore(&bitmap->lock, flags);
-				if (need_write)
+				if (need_write) {
 					write_page(bitmap, page, 0);
+					bitmap->allclean = 0;
+				}
 				continue;
 			}
 
@@ -1098,6 +1105,9 @@ void bitmap_daemon_work(struct bitmap *bitmap)
 /*
   if (j < 100) printk("bitmap: j=%lu, *bmc = 0x%x\n", j, *bmc);
 */
+			if (*bmc)
+				bitmap->allclean = 0;
+
 			if (*bmc == 2) {
 				*bmc=1; /* maybe clear the bit next time */
 				set_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
@@ -1132,6 +1142,8 @@ void bitmap_daemon_work(struct bitmap *bitmap)
 		}
 	}
 
+	if (bitmap->allclean == 0)
+		bitmap->mddev->thread->timeout = bitmap->daemon_sleep * HZ;
 }
 
 static bitmap_counter_t *bitmap_get_counter(struct bitmap *bitmap,
@@ -1226,6 +1238,7 @@ int bitmap_startwrite(struct bitmap *bitmap, sector_t offset, unsigned long sect
 			sectors -= blocks;
 		else sectors = 0;
 	}
+	bitmap->allclean = 0;
 	return 0;
 }
 
@@ -1296,6 +1309,7 @@ int bitmap_start_sync(struct bitmap *bitmap, sector_t offset, int *blocks,
 		}
 	}
 	spin_unlock_irq(&bitmap->lock);
+	bitmap->allclean = 0;
 	return rv;
 }
 
@@ -1332,6 +1346,7 @@ void bitmap_end_sync(struct bitmap *bitmap, sector_t offset, int *blocks, int ab
 	}
  unlock:
 	spin_unlock_irqrestore(&bitmap->lock, flags);
+	bitmap->allclean = 0;
 }
 
 void bitmap_close_sync(struct bitmap *bitmap)
@@ -1399,7 +1414,7 @@ static void bitmap_set_memory_bits(struct bitmap *bitmap, sector_t offset, int n
 		set_page_attr(bitmap, page, BITMAP_PAGE_CLEAN);
 	}
 	spin_unlock_irq(&bitmap->lock);
-
+	bitmap->allclean = 0;
 }
 
 /* dirty the memory and file bits for bitmap chunks "s" to "e" */
diff --git a/include/linux/raid/bitmap.h b/include/linux/raid/bitmap.h
index e51b531cd0b2..47fbcba11850 100644
--- a/include/linux/raid/bitmap.h
+++ b/include/linux/raid/bitmap.h
@@ -235,6 +235,8 @@ struct bitmap {
 
 	unsigned long flags;
 
+	int allclean;
+
 	unsigned long max_write_behind; /* write-behind mode */
 	atomic_t behind_writes;
 
-- 
cgit v1.2.3


From d0fae18f1b53a1d39135a968792be034bdf7ff26 Mon Sep 17 00:00:00 2001
From: NeilBrown <neilb@suse.de>
Date: Tue, 4 Mar 2008 14:29:31 -0800
Subject: md: clean up irregularity with raid autodetect

When a raid1 array is stopped, all components currently get added to the list
for auto-detection.  However we should really only add components that were
found by autodetection in the first place.  So add a flag to record that
information, and use it.

Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/md/md.c           | 4 +++-
 include/linux/raid/md_k.h | 1 +
 2 files changed, 4 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/md/md.c b/drivers/md/md.c
index b375de5c1af2..a71241c5ae72 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -1503,7 +1503,8 @@ static void export_rdev(mdk_rdev_t * rdev)
 	free_disk_sb(rdev);
 	list_del_init(&rdev->same_set);
 #ifndef MODULE
-	md_autodetect_dev(rdev->bdev->bd_dev);
+	if (test_bit(AutoDetected, &rdev->flags))
+		md_autodetect_dev(rdev->bdev->bd_dev);
 #endif
 	unlock_rdev(rdev);
 	kobject_put(&rdev->kobj);
@@ -6025,6 +6026,7 @@ static void autostart_arrays(int part)
 			MD_BUG();
 			continue;
 		}
+		set_bit(AutoDetected, &rdev->flags);
 		list_add(&rdev->same_set, &pending_raid_disks);
 		i_passed++;
 	}
diff --git a/include/linux/raid/md_k.h b/include/linux/raid/md_k.h
index 85a068bab625..7bb6d1abf71e 100644
--- a/include/linux/raid/md_k.h
+++ b/include/linux/raid/md_k.h
@@ -83,6 +83,7 @@ struct mdk_rdev_s
 #define	BarriersNotsupp	5		/* BIO_RW_BARRIER is not supported */
 #define	AllReserved	6		/* If whole device is reserved for
 					 * one array */
+#define	AutoDetected	7		/* added by auto-detect */
 
 	int desc_nr;			/* descriptor index in the superblock */
 	int raid_disk;			/* role of device in array */
-- 
cgit v1.2.3


From e0007529893c1c064be90bd21422ca0da4a0198e Mon Sep 17 00:00:00 2001
From: Eric Paris <eparis@redhat.com>
Date: Wed, 5 Mar 2008 10:31:54 -0500
Subject: LSM/SELinux: Interfaces to allow FS to control mount options

Introduce new LSM interfaces to allow an FS to deal with their own mount
options.  This includes a new string parsing function exported from the
LSM that an FS can use to get a security data blob and a new security
data blob.  This is particularly useful for an FS which uses binary
mount data, like NFS, which does not pass strings into the vfs to be
handled by the loaded LSM.  Also fix a BUG() in both SELinux and SMACK
when dealing with binary mount data.  If the binary mount data is less
than one page the copy_page() in security_sb_copy_data() can cause an
illegal page fault and boom.  Remove all NFSisms from the SELinux code
since they were broken by past NFS changes.

Signed-off-by: Eric Paris <eparis@redhat.com>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Acked-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: James Morris <jmorris@namei.org>
---
 fs/super.c                          |   4 +-
 include/linux/security.h            |  99 ++++++++++++++------
 security/dummy.c                    |  23 ++---
 security/security.c                 |  23 +++--
 security/selinux/hooks.c            | 175 +++++++++++++++++++-----------------
 security/selinux/include/security.h |   5 ++
 security/smack/smack_lsm.c          |   9 +-
 7 files changed, 204 insertions(+), 134 deletions(-)

(limited to 'include/linux')

diff --git a/fs/super.c b/fs/super.c
index 88811f60c8de..010446d8c40a 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -870,12 +870,12 @@ vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void
 	if (!mnt)
 		goto out;
 
-	if (data) {
+	if (data && !(type->fs_flags & FS_BINARY_MOUNTDATA)) {
 		secdata = alloc_secdata();
 		if (!secdata)
 			goto out_mnt;
 
-		error = security_sb_copy_data(type, data, secdata);
+		error = security_sb_copy_data(data, secdata);
 		if (error)
 			goto out_free_secdata;
 	}
diff --git a/include/linux/security.h b/include/linux/security.h
index fe52cdeab0a6..b07357ca2137 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -34,12 +34,6 @@
 #include <linux/xfrm.h>
 #include <net/flow.h>
 
-/* only a char in selinux superblock security struct flags */
-#define FSCONTEXT_MNT		0x01
-#define CONTEXT_MNT		0x02
-#define ROOTCONTEXT_MNT		0x04
-#define DEFCONTEXT_MNT		0x08
-
 extern unsigned securebits;
 
 struct ctl_table;
@@ -114,6 +108,32 @@ struct request_sock;
 
 #ifdef CONFIG_SECURITY
 
+struct security_mnt_opts {
+	char **mnt_opts;
+	int *mnt_opts_flags;
+	int num_mnt_opts;
+};
+
+static inline void security_init_mnt_opts(struct security_mnt_opts *opts)
+{
+	opts->mnt_opts = NULL;
+	opts->mnt_opts_flags = NULL;
+	opts->num_mnt_opts = 0;
+}
+
+static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
+{
+	int i;
+	if (opts->mnt_opts)
+		for(i = 0; i < opts->num_mnt_opts; i++)
+			kfree(opts->mnt_opts[i]);
+	kfree(opts->mnt_opts);
+	opts->mnt_opts = NULL;
+	kfree(opts->mnt_opts_flags);
+	opts->mnt_opts_flags = NULL;
+	opts->num_mnt_opts = 0;
+}
+
 /**
  * struct security_operations - main security structure
  *
@@ -262,19 +282,19 @@ struct request_sock;
  * @sb_get_mnt_opts:
  *	Get the security relevant mount options used for a superblock
  *	@sb the superblock to get security mount options from
- *	@mount_options array for pointers to mount options
- *	@mount_flags array of ints specifying what each mount options is
- *	@num_opts number of options in the arrays
+ *	@opts binary data structure containing all lsm mount data
  * @sb_set_mnt_opts:
  *	Set the security relevant mount options used for a superblock
  *	@sb the superblock to set security mount options for
- *	@mount_options array for pointers to mount options
- *	@mount_flags array of ints specifying what each mount options is
- *	@num_opts number of options in the arrays
+ *	@opts binary data structure containing all lsm mount data
  * @sb_clone_mnt_opts:
  *	Copy all security options from a given superblock to another
  *	@oldsb old superblock which contain information to clone
  *	@newsb new superblock which needs filled in
+ * @sb_parse_opts_str:
+ *	Parse a string of security data filling in the opts structure
+ *	@options string containing all mount options known by the LSM
+ *	@opts binary data structure usable by the LSM
  *
  * Security hooks for inode operations.
  *
@@ -1238,8 +1258,7 @@ struct security_operations {
 
 	int (*sb_alloc_security) (struct super_block * sb);
 	void (*sb_free_security) (struct super_block * sb);
-	int (*sb_copy_data)(struct file_system_type *type,
-			    void *orig, void *copy);
+	int (*sb_copy_data)(char *orig, char *copy);
 	int (*sb_kern_mount) (struct super_block *sb, void *data);
 	int (*sb_statfs) (struct dentry *dentry);
 	int (*sb_mount) (char *dev_name, struct nameidata * nd,
@@ -1257,12 +1276,12 @@ struct security_operations {
 	void (*sb_post_pivotroot) (struct nameidata * old_nd,
 				   struct nameidata * new_nd);
 	int (*sb_get_mnt_opts) (const struct super_block *sb,
-				char ***mount_options, int **flags,
-				int *num_opts);
-	int (*sb_set_mnt_opts) (struct super_block *sb, char **mount_options,
-				int *flags, int num_opts);
+				struct security_mnt_opts *opts);
+	int (*sb_set_mnt_opts) (struct super_block *sb,
+				struct security_mnt_opts *opts);
 	void (*sb_clone_mnt_opts) (const struct super_block *oldsb,
 				   struct super_block *newsb);
+	int (*sb_parse_opts_str) (char *options, struct security_mnt_opts *opts);
 
 	int (*inode_alloc_security) (struct inode *inode);	
 	void (*inode_free_security) (struct inode *inode);
@@ -1507,7 +1526,7 @@ int security_bprm_check(struct linux_binprm *bprm);
 int security_bprm_secureexec(struct linux_binprm *bprm);
 int security_sb_alloc(struct super_block *sb);
 void security_sb_free(struct super_block *sb);
-int security_sb_copy_data(struct file_system_type *type, void *orig, void *copy);
+int security_sb_copy_data(char *orig, char *copy);
 int security_sb_kern_mount(struct super_block *sb, void *data);
 int security_sb_statfs(struct dentry *dentry);
 int security_sb_mount(char *dev_name, struct nameidata *nd,
@@ -1520,12 +1539,12 @@ void security_sb_post_remount(struct vfsmount *mnt, unsigned long flags, void *d
 void security_sb_post_addmount(struct vfsmount *mnt, struct nameidata *mountpoint_nd);
 int security_sb_pivotroot(struct nameidata *old_nd, struct nameidata *new_nd);
 void security_sb_post_pivotroot(struct nameidata *old_nd, struct nameidata *new_nd);
-int security_sb_get_mnt_opts(const struct super_block *sb, char ***mount_options,
-			     int **flags, int *num_opts);
-int security_sb_set_mnt_opts(struct super_block *sb, char **mount_options,
-			     int *flags, int num_opts);
+int security_sb_get_mnt_opts(const struct super_block *sb,
+				struct security_mnt_opts *opts);
+int security_sb_set_mnt_opts(struct super_block *sb, struct security_mnt_opts *opts);
 void security_sb_clone_mnt_opts(const struct super_block *oldsb,
 				struct super_block *newsb);
+int security_sb_parse_opts_str(char *options, struct security_mnt_opts *opts);
 
 int security_inode_alloc(struct inode *inode);
 void security_inode_free(struct inode *inode);
@@ -1635,6 +1654,16 @@ int security_secctx_to_secid(char *secdata, u32 seclen, u32 *secid);
 void security_release_secctx(char *secdata, u32 seclen);
 
 #else /* CONFIG_SECURITY */
+struct security_mnt_opts {
+};
+
+static inline void security_init_mnt_opts(struct security_mnt_opts *opts)
+{
+}
+
+static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
+{
+}
 
 /*
  * This is the default capabilities functionality.  Most of these functions
@@ -1762,8 +1791,7 @@ static inline int security_sb_alloc (struct super_block *sb)
 static inline void security_sb_free (struct super_block *sb)
 { }
 
-static inline int security_sb_copy_data (struct file_system_type *type,
-					 void *orig, void *copy)
+static inline int security_sb_copy_data (char *orig, char *copy)
 {
 	return 0;
 }
@@ -1819,6 +1847,27 @@ static inline int security_sb_pivotroot (struct nameidata *old_nd,
 static inline void security_sb_post_pivotroot (struct nameidata *old_nd,
 					       struct nameidata *new_nd)
 { }
+static inline int security_sb_get_mnt_opts(const struct super_block *sb,
+					   struct security_mnt_opts *opts)
+{
+	security_init_mnt_opts(opts);
+	return 0;
+}
+
+static inline int security_sb_set_mnt_opts(struct super_block *sb,
+					   struct security_mnt_opts *opts)
+{
+	return 0;
+}
+
+static inline void security_sb_clone_mnt_opts(const struct super_block *oldsb,
+					      struct super_block *newsb)
+{ }
+
+static inline int security_sb_parse_opts_str(char *options, struct security_mnt_opts *opts)
+{
+	return 0;
+}
 
 static inline int security_inode_alloc (struct inode *inode)
 {
diff --git a/security/dummy.c b/security/dummy.c
index 649326bf64ea..78d8f92310a4 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -181,8 +181,7 @@ static void dummy_sb_free_security (struct super_block *sb)
 	return;
 }
 
-static int dummy_sb_copy_data (struct file_system_type *type,
-			       void *orig, void *copy)
+static int dummy_sb_copy_data (char *orig, char *copy)
 {
 	return 0;
 }
@@ -245,19 +244,17 @@ static void dummy_sb_post_pivotroot (struct nameidata *old_nd, struct nameidata
 	return;
 }
 
-static int dummy_sb_get_mnt_opts(const struct super_block *sb, char ***mount_options,
-				 int **flags, int *num_opts)
+static int dummy_sb_get_mnt_opts(const struct super_block *sb,
+				 struct security_mnt_opts *opts)
 {
-	*mount_options = NULL;
-	*flags = NULL;
-	*num_opts = 0;
+	security_init_mnt_opts(opts);
 	return 0;
 }
 
-static int dummy_sb_set_mnt_opts(struct super_block *sb, char **mount_options,
-				 int *flags, int num_opts)
+static int dummy_sb_set_mnt_opts(struct super_block *sb,
+				 struct security_mnt_opts *opts)
 {
-	if (unlikely(num_opts))
+	if (unlikely(opts->num_mnt_opts))
 		return -EOPNOTSUPP;
 	return 0;
 }
@@ -268,6 +265,11 @@ static void dummy_sb_clone_mnt_opts(const struct super_block *oldsb,
 	return;
 }
 
+static int dummy_sb_parse_opts_str(char *options, struct security_mnt_opts *opts)
+{
+	return 0;
+}
+
 static int dummy_inode_alloc_security (struct inode *inode)
 {
 	return 0;
@@ -1028,6 +1030,7 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, sb_get_mnt_opts);
 	set_to_dummy_if_null(ops, sb_set_mnt_opts);
 	set_to_dummy_if_null(ops, sb_clone_mnt_opts);
+	set_to_dummy_if_null(ops, sb_parse_opts_str);
 	set_to_dummy_if_null(ops, inode_alloc_security);
 	set_to_dummy_if_null(ops, inode_free_security);
 	set_to_dummy_if_null(ops, inode_init_security);
diff --git a/security/security.c b/security/security.c
index d15e56cbaade..b1387a6b416d 100644
--- a/security/security.c
+++ b/security/security.c
@@ -244,10 +244,11 @@ void security_sb_free(struct super_block *sb)
 	security_ops->sb_free_security(sb);
 }
 
-int security_sb_copy_data(struct file_system_type *type, void *orig, void *copy)
+int security_sb_copy_data(char *orig, char *copy)
 {
-	return security_ops->sb_copy_data(type, orig, copy);
+	return security_ops->sb_copy_data(orig, copy);
 }
+EXPORT_SYMBOL(security_sb_copy_data);
 
 int security_sb_kern_mount(struct super_block *sb, void *data)
 {
@@ -306,24 +307,30 @@ void security_sb_post_pivotroot(struct nameidata *old_nd, struct nameidata *new_
 }
 
 int security_sb_get_mnt_opts(const struct super_block *sb,
-			      char ***mount_options,
-			      int **flags, int *num_opts)
+				struct security_mnt_opts *opts)
 {
-	return security_ops->sb_get_mnt_opts(sb, mount_options, flags, num_opts);
+	return security_ops->sb_get_mnt_opts(sb, opts);
 }
 
 int security_sb_set_mnt_opts(struct super_block *sb,
-			      char **mount_options,
-			      int *flags, int num_opts)
+				struct security_mnt_opts *opts)
 {
-	return security_ops->sb_set_mnt_opts(sb, mount_options, flags, num_opts);
+	return security_ops->sb_set_mnt_opts(sb, opts);
 }
+EXPORT_SYMBOL(security_sb_set_mnt_opts);
 
 void security_sb_clone_mnt_opts(const struct super_block *oldsb,
 				struct super_block *newsb)
 {
 	security_ops->sb_clone_mnt_opts(oldsb, newsb);
 }
+EXPORT_SYMBOL(security_sb_clone_mnt_opts);
+
+int security_sb_parse_opts_str(char *options, struct security_mnt_opts *opts)
+{
+	return security_ops->sb_parse_opts_str(options, opts);
+}
+EXPORT_SYMBOL(security_sb_parse_opts_str);
 
 int security_inode_alloc(struct inode *inode)
 {
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 75c2e99bfb81..4bf4807f2d44 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -443,8 +443,7 @@ out:
  * mount options, or whatever.
  */
 static int selinux_get_mnt_opts(const struct super_block *sb,
-				char ***mount_options, int **mnt_opts_flags,
-				int *num_opts)
+				struct security_mnt_opts *opts)
 {
 	int rc = 0, i;
 	struct superblock_security_struct *sbsec = sb->s_security;
@@ -452,9 +451,7 @@ static int selinux_get_mnt_opts(const struct super_block *sb,
 	u32 len;
 	char tmp;
 
-	*num_opts = 0;
-	*mount_options = NULL;
-	*mnt_opts_flags = NULL;
+	security_init_mnt_opts(opts);
 
 	if (!sbsec->initialized)
 		return -EINVAL;
@@ -470,18 +467,18 @@ static int selinux_get_mnt_opts(const struct super_block *sb,
 	/* count the number of mount options for this sb */
 	for (i = 0; i < 8; i++) {
 		if (tmp & 0x01)
-			(*num_opts)++;
+			opts->num_mnt_opts++;
 		tmp >>= 1;
 	}
 
-	*mount_options = kcalloc(*num_opts, sizeof(char *), GFP_ATOMIC);
-	if (!*mount_options) {
+	opts->mnt_opts = kcalloc(opts->num_mnt_opts, sizeof(char *), GFP_ATOMIC);
+	if (!opts->mnt_opts) {
 		rc = -ENOMEM;
 		goto out_free;
 	}
 
-	*mnt_opts_flags = kcalloc(*num_opts, sizeof(int), GFP_ATOMIC);
-	if (!*mnt_opts_flags) {
+	opts->mnt_opts_flags = kcalloc(opts->num_mnt_opts, sizeof(int), GFP_ATOMIC);
+	if (!opts->mnt_opts_flags) {
 		rc = -ENOMEM;
 		goto out_free;
 	}
@@ -491,22 +488,22 @@ static int selinux_get_mnt_opts(const struct super_block *sb,
 		rc = security_sid_to_context(sbsec->sid, &context, &len);
 		if (rc)
 			goto out_free;
-		(*mount_options)[i] = context;
-		(*mnt_opts_flags)[i++] = FSCONTEXT_MNT;
+		opts->mnt_opts[i] = context;
+		opts->mnt_opts_flags[i++] = FSCONTEXT_MNT;
 	}
 	if (sbsec->flags & CONTEXT_MNT) {
 		rc = security_sid_to_context(sbsec->mntpoint_sid, &context, &len);
 		if (rc)
 			goto out_free;
-		(*mount_options)[i] = context;
-		(*mnt_opts_flags)[i++] = CONTEXT_MNT;
+		opts->mnt_opts[i] = context;
+		opts->mnt_opts_flags[i++] = CONTEXT_MNT;
 	}
 	if (sbsec->flags & DEFCONTEXT_MNT) {
 		rc = security_sid_to_context(sbsec->def_sid, &context, &len);
 		if (rc)
 			goto out_free;
-		(*mount_options)[i] = context;
-		(*mnt_opts_flags)[i++] = DEFCONTEXT_MNT;
+		opts->mnt_opts[i] = context;
+		opts->mnt_opts_flags[i++] = DEFCONTEXT_MNT;
 	}
 	if (sbsec->flags & ROOTCONTEXT_MNT) {
 		struct inode *root = sbsec->sb->s_root->d_inode;
@@ -515,24 +512,16 @@ static int selinux_get_mnt_opts(const struct super_block *sb,
 		rc = security_sid_to_context(isec->sid, &context, &len);
 		if (rc)
 			goto out_free;
-		(*mount_options)[i] = context;
-		(*mnt_opts_flags)[i++] = ROOTCONTEXT_MNT;
+		opts->mnt_opts[i] = context;
+		opts->mnt_opts_flags[i++] = ROOTCONTEXT_MNT;
 	}
 
-	BUG_ON(i != *num_opts);
+	BUG_ON(i != opts->num_mnt_opts);
 
 	return 0;
 
 out_free:
-	/* don't leak context string if security_sid_to_context had an error */
-	if (*mount_options && i)
-		for (; i > 0; i--)
-			kfree((*mount_options)[i-1]);
-	kfree(*mount_options);
-	*mount_options = NULL;
-	kfree(*mnt_opts_flags);
-	*mnt_opts_flags = NULL;
-	*num_opts = 0;
+	security_free_mnt_opts(opts);
 	return rc;
 }
 
@@ -553,12 +542,13 @@ static int bad_option(struct superblock_security_struct *sbsec, char flag,
 			return 1;
 	return 0;
 }
+
 /*
  * Allow filesystems with binary mount data to explicitly set mount point
  * labeling information.
  */
-static int selinux_set_mnt_opts(struct super_block *sb, char **mount_options,
-				int *flags, int num_opts)
+static int selinux_set_mnt_opts(struct super_block *sb,
+				struct security_mnt_opts *opts)
 {
 	int rc = 0, i;
 	struct task_security_struct *tsec = current->security;
@@ -568,6 +558,9 @@ static int selinux_set_mnt_opts(struct super_block *sb, char **mount_options,
 	struct inode_security_struct *root_isec = inode->i_security;
 	u32 fscontext_sid = 0, context_sid = 0, rootcontext_sid = 0;
 	u32 defcontext_sid = 0;
+	char **mount_options = opts->mnt_opts;
+	int *flags = opts->mnt_opts_flags;
+	int num_opts = opts->num_mnt_opts;
 
 	mutex_lock(&sbsec->lock);
 
@@ -588,6 +581,21 @@ static int selinux_set_mnt_opts(struct super_block *sb, char **mount_options,
 		goto out;
 	}
 
+	/*
+	 * Binary mount data FS will come through this function twice.  Once
+	 * from an explicit call and once from the generic calls from the vfs.
+	 * Since the generic VFS calls will not contain any security mount data
+	 * we need to skip the double mount verification.
+	 *
+	 * This does open a hole in which we will not notice if the first
+	 * mount using this sb set explict options and a second mount using
+	 * this sb does not set any security options.  (The first options
+	 * will be used for both mounts)
+	 */
+	if (sbsec->initialized && (sb->s_type->fs_flags & FS_BINARY_MOUNTDATA)
+	    && (num_opts == 0))
+	        goto out;
+
 	/*
 	 * parse the mount options, check if they are valid sids.
 	 * also check if someone is trying to mount the same sb more
@@ -792,43 +800,14 @@ static void selinux_sb_clone_mnt_opts(const struct super_block *oldsb,
 	mutex_unlock(&newsbsec->lock);
 }
 
-/*
- * string mount options parsing and call set the sbsec
- */
-static int superblock_doinit(struct super_block *sb, void *data)
+int selinux_parse_opts_str(char *options, struct security_mnt_opts *opts)
 {
+	char *p;
 	char *context = NULL, *defcontext = NULL;
 	char *fscontext = NULL, *rootcontext = NULL;
-	int rc = 0;
-	char *p, *options = data;
-	/* selinux only know about a fixed number of mount options */
-	char *mnt_opts[NUM_SEL_MNT_OPTS];
-	int mnt_opts_flags[NUM_SEL_MNT_OPTS], num_mnt_opts = 0;
-
-	if (!data)
-		goto out;
+	int rc, num_mnt_opts = 0;
 
-	/* with the nfs patch this will become a goto out; */
-	if (sb->s_type->fs_flags & FS_BINARY_MOUNTDATA) {
-		const char *name = sb->s_type->name;
-		/* NFS we understand. */
-		if (!strcmp(name, "nfs")) {
-			struct nfs_mount_data *d = data;
-
-			if (d->version !=  NFS_MOUNT_VERSION)
-				goto out;
-
-			if (d->context[0]) {
-				context = kstrdup(d->context, GFP_KERNEL);
-				if (!context) {
-					rc = -ENOMEM;
-					goto out;
-				}
-			}
-			goto build_flags;
-		} else
-			goto out;
-	}
+	opts->num_mnt_opts = 0;
 
 	/* Standard string-based options. */
 	while ((p = strsep(&options, "|")) != NULL) {
@@ -901,26 +880,37 @@ static int superblock_doinit(struct super_block *sb, void *data)
 		}
 	}
 
-build_flags:
+	rc = -ENOMEM;
+	opts->mnt_opts = kcalloc(NUM_SEL_MNT_OPTS, sizeof(char *), GFP_ATOMIC);
+	if (!opts->mnt_opts)
+		goto out_err;
+
+	opts->mnt_opts_flags = kcalloc(NUM_SEL_MNT_OPTS, sizeof(int), GFP_ATOMIC);
+	if (!opts->mnt_opts_flags) {
+		kfree(opts->mnt_opts);
+		goto out_err;
+	}
+
 	if (fscontext) {
-		mnt_opts[num_mnt_opts] = fscontext;
-		mnt_opts_flags[num_mnt_opts++] = FSCONTEXT_MNT;
+		opts->mnt_opts[num_mnt_opts] = fscontext;
+		opts->mnt_opts_flags[num_mnt_opts++] = FSCONTEXT_MNT;
 	}
 	if (context) {
-		mnt_opts[num_mnt_opts] = context;
-		mnt_opts_flags[num_mnt_opts++] = CONTEXT_MNT;
+		opts->mnt_opts[num_mnt_opts] = context;
+		opts->mnt_opts_flags[num_mnt_opts++] = CONTEXT_MNT;
 	}
 	if (rootcontext) {
-		mnt_opts[num_mnt_opts] = rootcontext;
-		mnt_opts_flags[num_mnt_opts++] = ROOTCONTEXT_MNT;
+		opts->mnt_opts[num_mnt_opts] = rootcontext;
+		opts->mnt_opts_flags[num_mnt_opts++] = ROOTCONTEXT_MNT;
 	}
 	if (defcontext) {
-		mnt_opts[num_mnt_opts] = defcontext;
-		mnt_opts_flags[num_mnt_opts++] = DEFCONTEXT_MNT;
+		opts->mnt_opts[num_mnt_opts] = defcontext;
+		opts->mnt_opts_flags[num_mnt_opts++] = DEFCONTEXT_MNT;
 	}
 
-out:
-	rc = selinux_set_mnt_opts(sb, mnt_opts, mnt_opts_flags, num_mnt_opts);
+	opts->num_mnt_opts = num_mnt_opts;
+	return 0;
+
 out_err:
 	kfree(context);
 	kfree(defcontext);
@@ -928,6 +918,33 @@ out_err:
 	kfree(rootcontext);
 	return rc;
 }
+/*
+ * string mount options parsing and call set the sbsec
+ */
+static int superblock_doinit(struct super_block *sb, void *data)
+{
+	int rc = 0;
+	char *options = data;
+	struct security_mnt_opts opts;
+
+	security_init_mnt_opts(&opts);
+
+	if (!data)
+		goto out;
+
+	BUG_ON(sb->s_type->fs_flags & FS_BINARY_MOUNTDATA);
+
+	rc = selinux_parse_opts_str(options, &opts);
+	if (rc)
+		goto out_err;
+
+out:
+	rc = selinux_set_mnt_opts(sb, &opts);
+
+out_err:
+	security_free_mnt_opts(&opts);
+	return rc;
+}
 
 static inline u16 inode_mode_to_security_class(umode_t mode)
 {
@@ -2253,7 +2270,7 @@ static inline void take_selinux_option(char **to, char *from, int *first,
 	}
 }
 
-static int selinux_sb_copy_data(struct file_system_type *type, void *orig, void *copy)
+static int selinux_sb_copy_data(char *orig, char *copy)
 {
 	int fnosec, fsec, rc = 0;
 	char *in_save, *in_curr, *in_end;
@@ -2263,12 +2280,6 @@ static int selinux_sb_copy_data(struct file_system_type *type, void *orig, void
 	in_curr = orig;
 	sec_curr = copy;
 
-	/* Binary mount data: just copy */
-	if (type->fs_flags & FS_BINARY_MOUNTDATA) {
-		copy_page(sec_curr, in_curr);
-		goto out;
-	}
-
 	nosec = (char *)get_zeroed_page(GFP_KERNEL);
 	if (!nosec) {
 		rc = -ENOMEM;
@@ -5251,6 +5262,8 @@ static struct security_operations selinux_ops = {
 	.sb_get_mnt_opts =		selinux_get_mnt_opts,
 	.sb_set_mnt_opts =		selinux_set_mnt_opts,
 	.sb_clone_mnt_opts = 		selinux_sb_clone_mnt_opts,
+	.sb_parse_opts_str = 		selinux_parse_opts_str,
+
 
 	.inode_alloc_security =		selinux_inode_alloc_security,
 	.inode_free_security =		selinux_inode_free_security,
diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
index 837ce420d2f6..f7d2f03781f2 100644
--- a/security/selinux/include/security.h
+++ b/security/selinux/include/security.h
@@ -35,6 +35,11 @@
 #define POLICYDB_VERSION_MAX	POLICYDB_VERSION_POLCAP
 #endif
 
+#define CONTEXT_MNT	0x01
+#define FSCONTEXT_MNT	0x02
+#define ROOTCONTEXT_MNT	0x04
+#define DEFCONTEXT_MNT	0x08
+
 struct netlbl_lsm_secattr;
 
 extern int selinux_enabled;
diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
index 770eb067e165..0241fd359675 100644
--- a/security/smack/smack_lsm.c
+++ b/security/smack/smack_lsm.c
@@ -189,17 +189,10 @@ static void smack_sb_free_security(struct super_block *sb)
  * Copy the Smack specific mount options out of the mount
  * options list.
  */
-static int smack_sb_copy_data(struct file_system_type *type, void *orig,
-			      void *smackopts)
+static int smack_sb_copy_data(char *orig, char *smackopts)
 {
 	char *cp, *commap, *otheropts, *dp;
 
-	/* Binary mount data: just copy */
-	if (type->fs_flags & FS_BINARY_MOUNTDATA) {
-		copy_page(smackopts, orig);
-		return 0;
-	}
-
 	otheropts = (char *)get_zeroed_page(GFP_KERNEL);
 	if (otheropts == NULL)
 		return -ENOMEM;
-- 
cgit v1.2.3


From ee6b967301b4aa5d4a4b61e2f682f086266db9fb Mon Sep 17 00:00:00 2001
From: Eric Dumazet <dada1@cosmosbay.com>
Date: Wed, 5 Mar 2008 18:30:47 -0800
Subject: [IPV4]: Add 'rtable' field in struct sk_buff to alias 'dst' and avoid
 casts

(Anonymous) unions can help us to avoid ugly casts.

A common cast it the (struct rtable *)skb->dst one.

Defining an union like  :
union {
     struct dst_entry *dst;
     struct rtable *rtable;
};
permits to use skb->rtable in place.

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h                  |  5 ++++-
 include/net/inet_sock.h                 |  2 +-
 net/bridge/br_netfilter.c               | 14 +++++++-------
 net/dccp/ipv4.c                         |  7 +++----
 net/ipv4/arp.c                          |  4 ++--
 net/ipv4/icmp.c                         | 10 +++++-----
 net/ipv4/igmp.c                         |  2 +-
 net/ipv4/ip_forward.c                   |  2 +-
 net/ipv4/ip_gre.c                       |  4 ++--
 net/ipv4/ip_input.c                     |  2 +-
 net/ipv4/ip_options.c                   | 16 ++++++++--------
 net/ipv4/ip_output.c                    | 10 +++++-----
 net/ipv4/ip_sockglue.c                  |  2 +-
 net/ipv4/ipip.c                         |  2 +-
 net/ipv4/ipmr.c                         |  6 +++---
 net/ipv4/netfilter/ipt_MASQUERADE.c     |  2 +-
 net/ipv4/netfilter/nf_nat_helper.c      |  4 ++--
 net/ipv4/route.c                        | 22 +++++++++++-----------
 net/ipv4/tcp_ipv4.c                     |  5 ++---
 net/ipv4/udp_ipv4.c                     |  2 +-
 net/netfilter/nf_conntrack_netbios_ns.c |  2 +-
 net/sched/em_meta.c                     |  4 ++--
 net/sctp/protocol.c                     |  8 ++++----
 23 files changed, 69 insertions(+), 68 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index bbd8d0027e2f..7beb239d2ee0 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -256,7 +256,10 @@ struct sk_buff {
 	ktime_t			tstamp;
 	struct net_device	*dev;
 
-	struct  dst_entry	*dst;
+	union {
+		struct  dst_entry	*dst;
+		struct  rtable		*rtable;
+	};
 	struct	sec_path	*sp;
 
 	/*
diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h
index 89cd011edb99..8660cb0fa0dd 100644
--- a/include/net/inet_sock.h
+++ b/include/net/inet_sock.h
@@ -195,7 +195,7 @@ static inline int inet_sk_ehashfn(const struct sock *sk)
 
 static inline int inet_iif(const struct sk_buff *skb)
 {
-	return ((struct rtable *)skb->dst)->rt_iif;
+	return skb->rtable->rt_iif;
 }
 
 #endif	/* _INET_SOCK_H */
diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c
index 1c0efd8ad9f3..0278a069c6f1 100644
--- a/net/bridge/br_netfilter.c
+++ b/net/bridge/br_netfilter.c
@@ -223,8 +223,8 @@ static int br_nf_pre_routing_finish_ipv6(struct sk_buff *skb)
 	}
 	nf_bridge->mask ^= BRNF_NF_BRIDGE_PREROUTING;
 
-	skb->dst = (struct dst_entry *)&__fake_rtable;
-	dst_hold(skb->dst);
+	skb->rtable = &__fake_rtable;
+	dst_hold(&__fake_rtable.u.dst);
 
 	skb->dev = nf_bridge->physindev;
 	nf_bridge_push_encap_header(skb);
@@ -388,8 +388,8 @@ bridged_dnat:
 			skb->pkt_type = PACKET_HOST;
 		}
 	} else {
-		skb->dst = (struct dst_entry *)&__fake_rtable;
-		dst_hold(skb->dst);
+		skb->rtable = &__fake_rtable;
+		dst_hold(&__fake_rtable.u.dst);
 	}
 
 	skb->dev = nf_bridge->physindev;
@@ -608,9 +608,9 @@ static unsigned int br_nf_local_in(unsigned int hook, struct sk_buff *skb,
 				   const struct net_device *out,
 				   int (*okfn)(struct sk_buff *))
 {
-	if (skb->dst == (struct dst_entry *)&__fake_rtable) {
-		dst_release(skb->dst);
-		skb->dst = NULL;
+	if (skb->rtable == &__fake_rtable) {
+		dst_release(&__fake_rtable.u.dst);
+		skb->rtable = NULL;
 	}
 
 	return NF_ACCEPT;
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index 514a40b7fc7f..17ad69e90e48 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -450,7 +450,7 @@ static struct dst_entry* dccp_v4_route_skb(struct sock *sk,
 					   struct sk_buff *skb)
 {
 	struct rtable *rt;
-	struct flowi fl = { .oif = ((struct rtable *)skb->dst)->rt_iif,
+	struct flowi fl = { .oif = skb->rtable->rt_iif,
 			    .nl_u = { .ip4_u =
 				      { .daddr = ip_hdr(skb)->saddr,
 					.saddr = ip_hdr(skb)->daddr,
@@ -511,7 +511,7 @@ static void dccp_v4_ctl_send_reset(struct sock *sk, struct sk_buff *rxskb)
 	if (dccp_hdr(rxskb)->dccph_type == DCCP_PKT_RESET)
 		return;
 
-	if (((struct rtable *)rxskb->dst)->rt_type != RTN_LOCAL)
+	if (rxskb->rtable->rt_type != RTN_LOCAL)
 		return;
 
 	dst = dccp_v4_route_skb(dccp_v4_ctl_socket->sk, rxskb);
@@ -563,8 +563,7 @@ int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
 	struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
 
 	/* Never answer to DCCP_PKT_REQUESTs send to broadcast or multicast */
-	if (((struct rtable *)skb->dst)->rt_flags &
-	    (RTCF_BROADCAST | RTCF_MULTICAST))
+	if (skb->rtable->rt_flags & (RTCF_BROADCAST | RTCF_MULTICAST))
 		return 0;	/* discard, don't send a reset here */
 
 	if (dccp_bad_service_code(sk, service)) {
diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c
index 69e80bd9774a..efe01df8fc0e 100644
--- a/net/ipv4/arp.c
+++ b/net/ipv4/arp.c
@@ -475,7 +475,7 @@ int arp_find(unsigned char *haddr, struct sk_buff *skb)
 		return 1;
 	}
 
-	paddr = ((struct rtable*)skb->dst)->rt_gateway;
+	paddr = skb->rtable->rt_gateway;
 
 	if (arp_set_predefined(inet_addr_type(&init_net, paddr), haddr, paddr, dev))
 		return 0;
@@ -814,7 +814,7 @@ static int arp_process(struct sk_buff *skb)
 	if (arp->ar_op == htons(ARPOP_REQUEST) &&
 	    ip_route_input(skb, tip, sip, 0, dev) == 0) {
 
-		rt = (struct rtable*)skb->dst;
+		rt = skb->rtable;
 		addr_type = rt->rt_type;
 
 		if (addr_type == RTN_LOCAL) {
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index cee77d606fbe..ff9a8e643fcc 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -381,7 +381,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param,
 static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb)
 {
 	struct ipcm_cookie ipc;
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct net *net = rt->u.dst.dev->nd_net;
 	struct sock *sk = icmp_sk(net);
 	struct inet_sock *inet = inet_sk(sk);
@@ -438,7 +438,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info)
 	struct iphdr *iph;
 	int room;
 	struct icmp_bxm icmp_param;
-	struct rtable *rt = (struct rtable *)skb_in->dst;
+	struct rtable *rt = skb_in->rtable;
 	struct ipcm_cookie ipc;
 	__be32 saddr;
 	u8  tos;
@@ -616,7 +616,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info)
 					     RT_TOS(tos), rt2->u.dst.dev);
 
 			dst_release(&rt2->u.dst);
-			rt2 = (struct rtable *)skb_in->dst;
+			rt2 = skb_in->rtable;
 			skb_in->dst = odst;
 		}
 
@@ -943,7 +943,7 @@ static void icmp_address(struct sk_buff *skb)
 
 static void icmp_address_reply(struct sk_buff *skb)
 {
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct net_device *dev = skb->dev;
 	struct in_device *in_dev;
 	struct in_ifaddr *ifa;
@@ -988,7 +988,7 @@ static void icmp_discard(struct sk_buff *skb)
 int icmp_rcv(struct sk_buff *skb)
 {
 	struct icmphdr *icmph;
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 
 	if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
 		int nh;
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index d3f34a772f3b..6a4ee8da6994 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -948,7 +948,7 @@ int igmp_rcv(struct sk_buff *skb)
 	case IGMPV2_HOST_MEMBERSHIP_REPORT:
 	case IGMPV3_HOST_MEMBERSHIP_REPORT:
 		/* Is it our report looped back? */
-		if (((struct rtable*)skb->dst)->fl.iif == 0)
+		if (skb->rtable->fl.iif == 0)
 			break;
 		/* don't rely on MC router hearing unicast reports */
 		if (skb->pkt_type == PACKET_MULTICAST ||
diff --git a/net/ipv4/ip_forward.c b/net/ipv4/ip_forward.c
index 0b3b328d82db..9d6d3befd854 100644
--- a/net/ipv4/ip_forward.c
+++ b/net/ipv4/ip_forward.c
@@ -80,7 +80,7 @@ int ip_forward(struct sk_buff *skb)
 	if (!xfrm4_route_forward(skb))
 		goto drop;
 
-	rt = (struct rtable*)skb->dst;
+	rt = skb->rtable;
 
 	if (opt->is_strictroute && rt->rt_dst != rt->rt_gateway)
 		goto sr_failed;
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index e7821ba7a9a0..f9ee84420cb3 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -619,7 +619,7 @@ static int ipgre_rcv(struct sk_buff *skb)
 #ifdef CONFIG_NET_IPGRE_BROADCAST
 		if (ipv4_is_multicast(iph->daddr)) {
 			/* Looped back packet, drop it! */
-			if (((struct rtable*)skb->dst)->fl.iif == 0)
+			if (skb->rtable->fl.iif == 0)
 				goto drop;
 			tunnel->stat.multicast++;
 			skb->pkt_type = PACKET_BROADCAST;
@@ -699,7 +699,7 @@ static int ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev)
 		}
 
 		if (skb->protocol == htons(ETH_P_IP)) {
-			rt = (struct rtable*)skb->dst;
+			rt = skb->rtable;
 			if ((dst = rt->rt_gateway) == 0)
 				goto tx_error_icmp;
 		}
diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c
index 65631391d479..d36e310b314d 100644
--- a/net/ipv4/ip_input.c
+++ b/net/ipv4/ip_input.c
@@ -351,7 +351,7 @@ static int ip_rcv_finish(struct sk_buff *skb)
 	if (iph->ihl > 5 && ip_rcv_options(skb))
 		goto drop;
 
-	rt = (struct rtable*)skb->dst;
+	rt = skb->rtable;
 	if (rt->rt_type == RTN_MULTICAST)
 		IP_INC_STATS_BH(IPSTATS_MIB_INMCASTPKTS);
 	else if (rt->rt_type == RTN_BROADCAST)
diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
index baaedd9689a0..df93a9c2efda 100644
--- a/net/ipv4/ip_options.c
+++ b/net/ipv4/ip_options.c
@@ -107,7 +107,7 @@ int ip_options_echo(struct ip_options * dopt, struct sk_buff * skb)
 	sptr = skb_network_header(skb);
 	dptr = dopt->__data;
 
-	daddr = ((struct rtable*)skb->dst)->rt_spec_dst;
+	daddr = skb->rtable->rt_spec_dst;
 
 	if (sopt->rr) {
 		optlen  = sptr[sopt->rr+1];
@@ -258,7 +258,7 @@ int ip_options_compile(struct ip_options * opt, struct sk_buff * skb)
 	unsigned char * optptr;
 	int optlen;
 	unsigned char * pp_ptr = NULL;
-	struct rtable *rt = skb ? (struct rtable*)skb->dst : NULL;
+	struct rtable *rt = skb ? skb->rtable : NULL;
 
 	if (!opt) {
 		opt = &(IPCB(skb)->opt);
@@ -558,7 +558,7 @@ void ip_forward_options(struct sk_buff *skb)
 {
 	struct   ip_options * opt	= &(IPCB(skb)->opt);
 	unsigned char * optptr;
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	unsigned char *raw = skb_network_header(skb);
 
 	if (opt->rr_needaddr) {
@@ -606,7 +606,7 @@ int ip_options_rcv_srr(struct sk_buff *skb)
 	__be32 nexthop;
 	struct iphdr *iph = ip_hdr(skb);
 	unsigned char *optptr = skb_network_header(skb) + opt->srr;
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct rtable *rt2;
 	int err;
 
@@ -631,13 +631,13 @@ int ip_options_rcv_srr(struct sk_buff *skb)
 		}
 		memcpy(&nexthop, &optptr[srrptr-1], 4);
 
-		rt = (struct rtable*)skb->dst;
-		skb->dst = NULL;
+		rt = skb->rtable;
+		skb->rtable = NULL;
 		err = ip_route_input(skb, nexthop, iph->saddr, iph->tos, skb->dev);
-		rt2 = (struct rtable*)skb->dst;
+		rt2 = skb->rtable;
 		if (err || (rt2->rt_type != RTN_UNICAST && rt2->rt_type != RTN_LOCAL)) {
 			ip_rt_put(rt2);
-			skb->dst = &rt->u.dst;
+			skb->rtable = rt;
 			return -EINVAL;
 		}
 		ip_rt_put(rt);
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 341779e685d9..dc494ea594a7 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -142,7 +142,7 @@ int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk,
 			  __be32 saddr, __be32 daddr, struct ip_options *opt)
 {
 	struct inet_sock *inet = inet_sk(sk);
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct iphdr *iph;
 
 	/* Build the IP header. */
@@ -240,7 +240,7 @@ static int ip_finish_output(struct sk_buff *skb)
 int ip_mc_output(struct sk_buff *skb)
 {
 	struct sock *sk = skb->sk;
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct net_device *dev = rt->u.dst.dev;
 
 	/*
@@ -321,7 +321,7 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok)
 	/* Skip all of this if the packet is already routed,
 	 * f.e. by something like SCTP.
 	 */
-	rt = (struct rtable *) skb->dst;
+	rt = skb->rtable;
 	if (rt != NULL)
 		goto packet_routed;
 
@@ -441,7 +441,7 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff*))
 	unsigned int mtu, hlen, left, len, ll_rs, pad;
 	int offset;
 	__be16 not_last_frag;
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	int err = 0;
 
 	dev = rt->u.dst.dev;
@@ -1357,7 +1357,7 @@ void ip_send_reply(struct sock *sk, struct sk_buff *skb, struct ip_reply_arg *ar
 	} replyopts;
 	struct ipcm_cookie ipc;
 	__be32 daddr;
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 
 	if (ip_options_echo(&replyopts.opt, skb))
 		return;
diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c
index de0572c88859..e7c9e4e72327 100644
--- a/net/ipv4/ip_sockglue.c
+++ b/net/ipv4/ip_sockglue.c
@@ -57,7 +57,7 @@
 static void ip_cmsg_recv_pktinfo(struct msghdr *msg, struct sk_buff *skb)
 {
 	struct in_pktinfo info;
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 
 	info.ipi_addr.s_addr = ip_hdr(skb)->daddr;
 	if (rt) {
diff --git a/net/ipv4/ipip.c b/net/ipv4/ipip.c
index dbaed69de06a..894bce96284a 100644
--- a/net/ipv4/ipip.c
+++ b/net/ipv4/ipip.c
@@ -528,7 +528,7 @@ static int ipip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	if (!dst) {
 		/* NBMA tunnel */
-		if ((rt = (struct rtable*)skb->dst) == NULL) {
+		if ((rt = skb->rtable) == NULL) {
 			tunnel->stat.tx_fifo_errors++;
 			goto tx_error;
 		}
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index a94f52c207a7..7d63d74ef62a 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -1283,7 +1283,7 @@ static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local
 	if (vif_table[vif].dev != skb->dev) {
 		int true_vifi;
 
-		if (((struct rtable*)skb->dst)->fl.iif == 0) {
+		if (skb->rtable->fl.iif == 0) {
 			/* It is our own packet, looped back.
 			   Very complicated situation...
 
@@ -1357,7 +1357,7 @@ dont_forward:
 int ip_mr_input(struct sk_buff *skb)
 {
 	struct mfc_cache *cache;
-	int local = ((struct rtable*)skb->dst)->rt_flags&RTCF_LOCAL;
+	int local = skb->rtable->rt_flags&RTCF_LOCAL;
 
 	/* Packet is looped back after forward, it should not be
 	   forwarded second time, but still can be delivered locally.
@@ -1594,7 +1594,7 @@ int ipmr_get_route(struct sk_buff *skb, struct rtmsg *rtm, int nowait)
 {
 	int err;
 	struct mfc_cache *cache;
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 
 	read_lock(&mrt_lock);
 	cache = ipmr_cache_find(rt->rt_src, rt->rt_dst);
diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c
index 313b3fcf387e..c6817b18366a 100644
--- a/net/ipv4/netfilter/ipt_MASQUERADE.c
+++ b/net/ipv4/netfilter/ipt_MASQUERADE.c
@@ -77,7 +77,7 @@ masquerade_tg(struct sk_buff *skb, const struct net_device *in,
 		return NF_ACCEPT;
 
 	mr = targinfo;
-	rt = (struct rtable *)skb->dst;
+	rt = skb->rtable;
 	newsrc = inet_select_addr(out, rt->rt_gateway, RT_SCOPE_UNIVERSE);
 	if (!newsrc) {
 		printk("MASQUERADE: %s ate my IP address\n", out->name);
diff --git a/net/ipv4/netfilter/nf_nat_helper.c b/net/ipv4/netfilter/nf_nat_helper.c
index ca57f47bbd25..2fca727aa8ba 100644
--- a/net/ipv4/netfilter/nf_nat_helper.c
+++ b/net/ipv4/netfilter/nf_nat_helper.c
@@ -139,7 +139,7 @@ nf_nat_mangle_tcp_packet(struct sk_buff *skb,
 			 const char *rep_buffer,
 			 unsigned int rep_len)
 {
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct iphdr *iph;
 	struct tcphdr *tcph;
 	int oldlen, datalen;
@@ -217,7 +217,7 @@ nf_nat_mangle_udp_packet(struct sk_buff *skb,
 			 const char *rep_buffer,
 			 unsigned int rep_len)
 {
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct iphdr *iph;
 	struct udphdr *udph;
 	int datalen, oldlen;
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 8c3e165f0034..1051326c36b2 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -1289,7 +1289,7 @@ reject_redirect:
 
 static struct dst_entry *ipv4_negative_advice(struct dst_entry *dst)
 {
-	struct rtable *rt = (struct rtable*)dst;
+	struct rtable *rt = (struct rtable *)dst;
 	struct dst_entry *ret = dst;
 
 	if (rt) {
@@ -1330,7 +1330,7 @@ static struct dst_entry *ipv4_negative_advice(struct dst_entry *dst)
 
 void ip_rt_send_redirect(struct sk_buff *skb)
 {
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct in_device *in_dev = in_dev_get(rt->u.dst.dev);
 
 	if (!in_dev)
@@ -1379,7 +1379,7 @@ out:
 
 static int ip_error(struct sk_buff *skb)
 {
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	unsigned long now;
 	int code;
 
@@ -1548,7 +1548,7 @@ static void ipv4_link_failure(struct sk_buff *skb)
 
 	icmp_send(skb, ICMP_DEST_UNREACH, ICMP_HOST_UNREACH, 0);
 
-	rt = (struct rtable *) skb->dst;
+	rt = skb->rtable;
 	if (rt)
 		dst_set_expires(&rt->u.dst, 0);
 }
@@ -1708,7 +1708,7 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 
 	in_dev_put(in_dev);
 	hash = rt_hash(daddr, saddr, dev->ifindex);
-	return rt_intern_hash(hash, rth, (struct rtable**) &skb->dst);
+	return rt_intern_hash(hash, rth, &skb->rtable);
 
 e_nobufs:
 	in_dev_put(in_dev);
@@ -1869,7 +1869,7 @@ static inline int ip_mkroute_input(struct sk_buff *skb,
 
 	/* put it into the cache */
 	hash = rt_hash(daddr, saddr, fl->iif);
-	return rt_intern_hash(hash, rth, (struct rtable**)&skb->dst);
+	return rt_intern_hash(hash, rth, &skb->rtable);
 }
 
 /*
@@ -2025,7 +2025,7 @@ local_input:
 	}
 	rth->rt_type	= res.type;
 	hash = rt_hash(daddr, saddr, fl.iif);
-	err = rt_intern_hash(hash, rth, (struct rtable**)&skb->dst);
+	err = rt_intern_hash(hash, rth, &skb->rtable);
 	goto done;
 
 no_route:
@@ -2091,7 +2091,7 @@ int ip_route_input(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 			dst_use(&rth->u.dst, jiffies);
 			RT_CACHE_STAT_INC(in_hit);
 			rcu_read_unlock();
-			skb->dst = (struct dst_entry*)rth;
+			skb->rtable = rth;
 			return 0;
 		}
 		RT_CACHE_STAT_INC(in_hlist_search);
@@ -2598,7 +2598,7 @@ int ip_route_output_key(struct net *net, struct rtable **rp, struct flowi *flp)
 static int rt_fill_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
 			int nowait, unsigned int flags)
 {
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct rtmsg *r;
 	struct nlmsghdr *nlh;
 	long expires;
@@ -2742,7 +2742,7 @@ static int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void
 		err = ip_route_input(skb, dst, src, rtm->rtm_tos, dev);
 		local_bh_enable();
 
-		rt = (struct rtable*) skb->dst;
+		rt = skb->rtable;
 		if (err == 0 && rt->u.dst.error)
 			err = -rt->u.dst.error;
 	} else {
@@ -2762,7 +2762,7 @@ static int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void
 	if (err)
 		goto errout_free;
 
-	skb->dst = &rt->u.dst;
+	skb->rtable = rt;
 	if (rtm->rtm_flags & RTM_F_NOTIFY)
 		rt->rt_flags |= RTCF_NOTIFY;
 
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 3873c4dbeaeb..a79e324638eb 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -552,7 +552,7 @@ static void tcp_v4_send_reset(struct sock *sk, struct sk_buff *skb)
 	if (th->rst)
 		return;
 
-	if (((struct rtable *)skb->dst)->rt_type != RTN_LOCAL)
+	if (skb->rtable->rt_type != RTN_LOCAL)
 		return;
 
 	/* Swap the send and the receive. */
@@ -1262,8 +1262,7 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
 #endif
 
 	/* Never answer to SYNs send to broadcast or multicast */
-	if (((struct rtable *)skb->dst)->rt_flags &
-	    (RTCF_BROADCAST | RTCF_MULTICAST))
+	if (skb->rtable->rt_flags & (RTCF_BROADCAST | RTCF_MULTICAST))
 		goto drop;
 
 	/* TW buckets are converted to open requests without
diff --git a/net/ipv4/udp_ipv4.c b/net/ipv4/udp_ipv4.c
index 40978de7fb51..fd14c2c50ed4 100644
--- a/net/ipv4/udp_ipv4.c
+++ b/net/ipv4/udp_ipv4.c
@@ -893,7 +893,7 @@ int __udp4_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 	struct sock *sk;
 	struct udphdr *uh = udp_hdr(skb);
 	unsigned short ulen;
-	struct rtable *rt = (struct rtable*)skb->dst;
+	struct rtable *rt = skb->rtable;
 	__be32 saddr = ip_hdr(skb)->saddr;
 	__be32 daddr = ip_hdr(skb)->daddr;
 
diff --git a/net/netfilter/nf_conntrack_netbios_ns.c b/net/netfilter/nf_conntrack_netbios_ns.c
index 9810d81e2a06..60dedaded84e 100644
--- a/net/netfilter/nf_conntrack_netbios_ns.c
+++ b/net/netfilter/nf_conntrack_netbios_ns.c
@@ -47,7 +47,7 @@ static int help(struct sk_buff *skb, unsigned int protoff,
 {
 	struct nf_conntrack_expect *exp;
 	struct iphdr *iph = ip_hdr(skb);
-	struct rtable *rt = (struct rtable *)skb->dst;
+	struct rtable *rt = skb->rtable;
 	struct in_device *in_dev;
 	__be32 mask = 0;
 
diff --git a/net/sched/em_meta.c b/net/sched/em_meta.c
index 3da4129b89d1..72cf86e3c090 100644
--- a/net/sched/em_meta.c
+++ b/net/sched/em_meta.c
@@ -256,10 +256,10 @@ META_COLLECTOR(int_rtclassid)
 
 META_COLLECTOR(int_rtiif)
 {
-	if (unlikely(skb->dst == NULL))
+	if (unlikely(skb->rtable == NULL))
 		*err = -1;
 	else
-		dst->value = ((struct rtable*) skb->dst)->fl.iif;
+		dst->value = skb->rtable->fl.iif;
 }
 
 /**************************************************************************
diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c
index 8d9d929f6cea..1afef08f6c1d 100644
--- a/net/sctp/protocol.c
+++ b/net/sctp/protocol.c
@@ -363,7 +363,7 @@ static int sctp_v4_addr_valid(union sctp_addr *addr,
 		return 0;
 
 	/* Is this a broadcast address? */
-	if (skb && ((struct rtable *)skb->dst)->rt_flags & RTCF_BROADCAST)
+	if (skb && skb->rtable->rt_flags & RTCF_BROADCAST)
 		return 0;
 
 	return 1;
@@ -539,7 +539,7 @@ static void sctp_v4_get_saddr(struct sctp_association *asoc,
 /* What interface did this skb arrive on? */
 static int sctp_v4_skb_iif(const struct sk_buff *skb)
 {
-	return ((struct rtable *)skb->dst)->rt_iif;
+	return skb->rtable->rt_iif;
 }
 
 /* Was this packet marked by Explicit Congestion Notification? */
@@ -828,8 +828,8 @@ static inline int sctp_v4_xmit(struct sk_buff *skb,
 	SCTP_DEBUG_PRINTK("%s: skb:%p, len:%d, "
 			  "src:%u.%u.%u.%u, dst:%u.%u.%u.%u\n",
 			  __FUNCTION__, skb, skb->len,
-			  NIPQUAD(((struct rtable *)skb->dst)->rt_src),
-			  NIPQUAD(((struct rtable *)skb->dst)->rt_dst));
+			  NIPQUAD(skb->rtable->rt_src),
+			  NIPQUAD(skb->rtable->rt_dst));
 
 	SCTP_INC_STATS(SCTP_MIB_OUTSCTPPACKS);
 	return ip_queue_xmit(skb, ipfragok);
-- 
cgit v1.2.3


From f59d43899e279c77924a7ada4bec8c70e5aeca06 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Wed, 5 Mar 2008 20:58:10 -0800
Subject: [IPV6]: Fix powerpc allmodconfig build warnings.

Introduced by changeset 95e41e93e18d8e1e272ce23d96bae4f17ce11d42
("[IPV6]: Make ndisc_flow_init() common for later use.")

Reported by Stephen Rothwell.

In file included from net/ipv6/netfilter/ip6_tables.c:21:
include/linux/icmpv6.h:192: warning: 'struct in6_addr' declared inside parameter list
include/linux/icmpv6.h:192: warning: its scope is only this definition or declaration, which is probably not what you want

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/icmpv6.h | 1 +
 1 file changed, 1 insertion(+)

(limited to 'include/linux')

diff --git a/include/linux/icmpv6.h b/include/linux/icmpv6.h
index e4d4300d768f..03067443198a 100644
--- a/include/linux/icmpv6.h
+++ b/include/linux/icmpv6.h
@@ -184,6 +184,7 @@ extern void				icmpv6_param_prob(struct sk_buff *skb,
 							  int code, int pos);
 
 struct flowi;
+struct in6_addr;
 extern void				icmpv6_flow_init(struct sock *sk,
 							 struct flowi *fl,
 							 u8 type,
-- 
cgit v1.2.3


From 37c5798968d0ce4d479f114f1d5785551b57bfa5 Mon Sep 17 00:00:00 2001
From: Luis Carlos Cobo <luisca@cozybit.com>
Date: Sat, 23 Feb 2008 15:17:04 +0100
Subject: wireless: various definitions for mesh networking

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/linux/ieee80211.h | 35 +++++++++++++++++++++++++++++++++++
 1 file changed, 35 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/ieee80211.h b/include/linux/ieee80211.h
index f577c8f1c66d..f27d11ab418b 100644
--- a/include/linux/ieee80211.h
+++ b/include/linux/ieee80211.h
@@ -97,6 +97,7 @@
 #define IEEE80211_MAX_FRAME_LEN		2352
 
 #define IEEE80211_MAX_SSID_LEN		32
+#define IEEE80211_MAX_MESH_ID_LEN	32
 
 struct ieee80211_hdr {
 	__le16 frame_control;
@@ -109,6 +110,16 @@ struct ieee80211_hdr {
 } __attribute__ ((packed));
 
 
+struct ieee80211s_hdr {
+	u8 flags;
+	u8 ttl;
+	u8 seqnum[3];
+	u8 eaddr1[6];
+	u8 eaddr2[6];
+	u8 eaddr3[6];
+} __attribute__ ((packed));
+
+
 struct ieee80211_mgmt {
 	__le16 frame_control;
 	__le16 duration;
@@ -206,6 +217,23 @@ struct ieee80211_mgmt {
 					__le16 params;
 					__le16 reason_code;
 				} __attribute__((packed)) delba;
+				struct{
+					u8 action_code;
+					/* capab_info for open and confirm,
+					 * reason for close
+					 */
+					__le16 aux;
+					/* Followed in plink_confirm by status
+					 * code, AID and supported rates,
+					 * and directly by supported rates in
+					 * plink_open and plink_close
+					 */
+					u8 variable[0];
+				} __attribute__((packed)) plink_action;
+				struct{
+					u8 action_code;
+					u8 variable[0];
+				} __attribute__((packed)) mesh_action;
 			} u;
 		} __attribute__ ((packed)) action;
 	} u;
@@ -437,6 +465,13 @@ enum ieee80211_eid {
 	WLAN_EID_TS_DELAY = 43,
 	WLAN_EID_TCLAS_PROCESSING = 44,
 	WLAN_EID_QOS_CAPA = 46,
+	/* 802.11s */
+	WLAN_EID_MESH_CONFIG = 36,      /* Pending IEEE 802.11 ANA approval */
+	WLAN_EID_MESH_ID = 37,          /* Pending IEEE 802.11 ANA approval */
+	WLAN_EID_PEER_LINK = 40,	/* Pending IEEE 802.11 ANA approval */
+	WLAN_EID_PREQ = 53,		/* Pending IEEE 802.11 ANA approval */
+	WLAN_EID_PREP = 54,		/* Pending IEEE 802.11 ANA approval */
+	WLAN_EID_PERR = 55,		/* Pending IEEE 802.11 ANA approval */
 	/* 802.11h */
 	WLAN_EID_PWR_CONSTRAINT = 32,
 	WLAN_EID_PWR_CAPABILITY = 33,
-- 
cgit v1.2.3


From cc0672a1066829be7e1b0128a13e36a2d0a15479 Mon Sep 17 00:00:00 2001
From: Johannes Berg <johannes@sipsolutions.net>
Date: Sat, 23 Feb 2008 15:17:05 +0100
Subject: WEXT: add mesh interface type

This introduces a new WEXT type IW_MODE_MESH for mesh networks,
used for scan results.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/linux/wireless.h | 1 +
 1 file changed, 1 insertion(+)

(limited to 'include/linux')

diff --git a/include/linux/wireless.h b/include/linux/wireless.h
index 3160dfed73ca..2864b1699ecc 100644
--- a/include/linux/wireless.h
+++ b/include/linux/wireless.h
@@ -455,6 +455,7 @@
 #define IW_MODE_REPEAT	4	/* Wireless Repeater (forwarder) */
 #define IW_MODE_SECOND	5	/* Secondary master/repeater (backup) */
 #define IW_MODE_MONITOR	6	/* Passive monitor (listen only) */
+#define IW_MODE_MESH	7	/* Mesh (IEEE 802.11s) network */
 
 /* Statistics flags (bitmask in updated) */
 #define IW_QUAL_QUAL_UPDATED	0x01	/* Value was updated since last read */
-- 
cgit v1.2.3


From 2ec600d672e74488f8d1acf67a0a2baed222564c Mon Sep 17 00:00:00 2001
From: Luis Carlos Cobo <luisca@cozybit.com>
Date: Sat, 23 Feb 2008 15:17:06 +0100
Subject: nl80211/cfg80211: support for mesh, sta dumping

Added support for mesh id and mesh path operation as well as
station structure dumping.

Signed-off-by: Luis Carlos Cobo <luisca@cozybit.com>
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 include/linux/nl80211.h | 119 +++++++++++--
 include/net/cfg80211.h  | 139 +++++++++++++--
 net/mac80211/cfg.c      |  20 ++-
 net/wireless/nl80211.c  | 438 +++++++++++++++++++++++++++++++++++++++++++++---
 4 files changed, 650 insertions(+), 66 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/nl80211.h b/include/linux/nl80211.h
index a9f0b93324a2..ea6517e58b04 100644
--- a/include/linux/nl80211.h
+++ b/include/linux/nl80211.h
@@ -78,6 +78,18 @@
  *	or, if no MAC address given, all stations, on the interface identified
  *	by %NL80211_ATTR_IFINDEX.
  *
+ * @NL80211_CMD_GET_MPATH: Get mesh path attributes for mesh path to
+ * 	destination %NL80211_ATTR_MAC on the interface identified by
+ * 	%NL80211_ATTR_IFINDEX.
+ * @NL80211_CMD_SET_MPATH:  Set mesh path attributes for mesh path to
+ * 	destination %NL80211_ATTR_MAC on the interface identified by
+ * 	%NL80211_ATTR_IFINDEX.
+ * @NL80211_CMD_NEW_PATH: Add a mesh path with given attributes to the
+ *	the interface identified by %NL80211_ATTR_IFINDEX.
+ * @NL80211_CMD_DEL_PATH: Remove a mesh path identified by %NL80211_ATTR_MAC
+ *	or, if no MAC address given, all mesh paths, on the interface identified
+ *	by %NL80211_ATTR_IFINDEX.
+ *
  * @NL80211_CMD_MAX: highest used command number
  * @__NL80211_CMD_AFTER_LAST: internal use
  */
@@ -112,6 +124,11 @@ enum nl80211_commands {
 
 	/* add commands here */
 
+	NL80211_CMD_GET_MPATH,
+	NL80211_CMD_SET_MPATH,
+	NL80211_CMD_NEW_MPATH,
+	NL80211_CMD_DEL_MPATH,
+
 	/* used to define NL80211_CMD_MAX below */
 	__NL80211_CMD_AFTER_LAST,
 	NL80211_CMD_MAX = __NL80211_CMD_AFTER_LAST - 1
@@ -157,13 +174,21 @@ enum nl80211_commands {
  *	restriction (at most %NL80211_MAX_SUPP_RATES).
  * @NL80211_ATTR_STA_VLAN: interface index of VLAN interface to move station
  *	to, or the AP interface the station was originally added to to.
- * @NL80211_ATTR_STA_STATS: statistics for a station, part of station info
+ * @NL80211_ATTR_STA_INFO: information about a station, part of station info
  *	given for %NL80211_CMD_GET_STATION, nested attribute containing
- *	info as possible, see &enum nl80211_sta_stats.
+ *	info as possible, see &enum nl80211_sta_info.
  *
  * @NL80211_ATTR_WIPHY_BANDS: Information about an operating bands,
  *	consisting of a nested array.
  *
+ * @NL80211_ATTR_MESH_ID: mesh id (1-32 bytes).
+ * @NL80211_ATTR_PLINK_ACTION: action to perform on the mesh peer link.
+ * @NL80211_ATTR_MPATH_NEXT_HOP: MAC address of the next hop for a mesh path.
+ * @NL80211_ATTR_MPATH_INFO: information about a mesh_path, part of mesh path
+ * 	info given for %NL80211_CMD_GET_MPATH, nested attribute described at
+ *	&enum nl80211_mpath_info.
+ *
+ *
  * @NL80211_ATTR_MNTR_FLAGS: flags, nested element with NLA_FLAG attributes of
  *      &enum nl80211_mntr_flags.
  *
@@ -199,7 +224,7 @@ enum nl80211_attrs {
 	NL80211_ATTR_STA_LISTEN_INTERVAL,
 	NL80211_ATTR_STA_SUPPORTED_RATES,
 	NL80211_ATTR_STA_VLAN,
-	NL80211_ATTR_STA_STATS,
+	NL80211_ATTR_STA_INFO,
 
 	NL80211_ATTR_WIPHY_BANDS,
 
@@ -207,6 +232,11 @@ enum nl80211_attrs {
 
 	/* add attributes here, update the policy in nl80211.c */
 
+	NL80211_ATTR_MESH_ID,
+	NL80211_ATTR_STA_PLINK_ACTION,
+	NL80211_ATTR_MPATH_NEXT_HOP,
+	NL80211_ATTR_MPATH_INFO,
+
 	__NL80211_ATTR_AFTER_LAST,
 	NL80211_ATTR_MAX = __NL80211_ATTR_AFTER_LAST - 1
 };
@@ -223,6 +253,7 @@ enum nl80211_attrs {
  * @NL80211_IFTYPE_AP_VLAN: VLAN interface for access points
  * @NL80211_IFTYPE_WDS: wireless distribution interface
  * @NL80211_IFTYPE_MONITOR: monitor interface receiving all frames
+ * @NL80211_IFTYPE_MESH_POINT: mesh point
  * @NL80211_IFTYPE_MAX: highest interface type number currently defined
  * @__NL80211_IFTYPE_AFTER_LAST: internal use
  *
@@ -238,6 +269,7 @@ enum nl80211_iftype {
 	NL80211_IFTYPE_AP_VLAN,
 	NL80211_IFTYPE_WDS,
 	NL80211_IFTYPE_MONITOR,
+	NL80211_IFTYPE_MESH_POINT,
 
 	/* keep last */
 	__NL80211_IFTYPE_AFTER_LAST,
@@ -267,27 +299,78 @@ enum nl80211_sta_flags {
 };
 
 /**
- * enum nl80211_sta_stats - station statistics
+ * enum nl80211_sta_info - station information
  *
- * These attribute types are used with %NL80211_ATTR_STA_STATS
+ * These attribute types are used with %NL80211_ATTR_STA_INFO
  * when getting information about a station.
  *
- * @__NL80211_STA_STAT_INVALID: attribute number 0 is reserved
- * @NL80211_STA_STAT_INACTIVE_TIME: time since last activity (u32, msecs)
- * @NL80211_STA_STAT_RX_BYTES: total received bytes (u32, from this station)
- * @NL80211_STA_STAT_TX_BYTES: total transmitted bytes (u32, to this station)
- * @__NL80211_STA_STAT_AFTER_LAST: internal
- * @NL80211_STA_STAT_MAX: highest possible station stats attribute
+ * @__NL80211_STA_INFO_INVALID: attribute number 0 is reserved
+ * @NL80211_STA_INFO_INACTIVE_TIME: time since last activity (u32, msecs)
+ * @NL80211_STA_INFO_RX_BYTES: total received bytes (u32, from this station)
+ * @NL80211_STA_INFO_TX_BYTES: total transmitted bytes (u32, to this station)
+ * @__NL80211_STA_INFO_AFTER_LAST: internal
+ * @NL80211_STA_INFO_MAX: highest possible station info attribute
+ */
+enum nl80211_sta_info {
+	__NL80211_STA_INFO_INVALID,
+	NL80211_STA_INFO_INACTIVE_TIME,
+	NL80211_STA_INFO_RX_BYTES,
+	NL80211_STA_INFO_TX_BYTES,
+	NL80211_STA_INFO_LLID,
+	NL80211_STA_INFO_PLID,
+	NL80211_STA_INFO_PLINK_STATE,
+
+	/* keep last */
+	__NL80211_STA_INFO_AFTER_LAST,
+	NL80211_STA_INFO_MAX = __NL80211_STA_INFO_AFTER_LAST - 1
+};
+
+/**
+ * enum nl80211_mpath_flags - nl80211 mesh path flags
+ *
+ * @NL80211_MPATH_FLAG_ACTIVE: the mesh path is active
+ * @NL80211_MPATH_FLAG_RESOLVING: the mesh path discovery process is running
+ * @NL80211_MPATH_FLAG_DSN_VALID: the mesh path contains a valid DSN
+ * @NL80211_MPATH_FLAG_FIXED: the mesh path has been manually set
+ * @NL80211_MPATH_FLAG_RESOLVED: the mesh path discovery process succeeded
+ */
+enum nl80211_mpath_flags {
+	NL80211_MPATH_FLAG_ACTIVE =	1<<0,
+	NL80211_MPATH_FLAG_RESOLVING =	1<<1,
+	NL80211_MPATH_FLAG_DSN_VALID =	1<<2,
+	NL80211_MPATH_FLAG_FIXED =	1<<3,
+	NL80211_MPATH_FLAG_RESOLVED =	1<<4,
+};
+
+/**
+ * enum nl80211_mpath_info - mesh path information
+ *
+ * These attribute types are used with %NL80211_ATTR_MPATH_INFO when getting
+ * information about a mesh path.
+ *
+ * @__NL80211_MPATH_INFO_INVALID: attribute number 0 is reserved
+ * @NL80211_ATTR_MPATH_FRAME_QLEN: number of queued frames for this destination
+ * @NL80211_ATTR_MPATH_DSN: destination sequence number
+ * @NL80211_ATTR_MPATH_METRIC: metric (cost) of this mesh path
+ * @NL80211_ATTR_MPATH_EXPTIME: expiration time for the path, in msec from now
+ * @NL80211_ATTR_MPATH_FLAGS: mesh path flags, enumerated in
+ * 	&enum nl80211_mpath_flags;
+ * @NL80211_ATTR_MPATH_DISCOVERY_TIMEOUT: total path discovery timeout, in msec
+ * @NL80211_ATTR_MPATH_DISCOVERY_RETRIES: mesh path discovery retries
  */
-enum nl80211_sta_stats {
-	__NL80211_STA_STAT_INVALID,
-	NL80211_STA_STAT_INACTIVE_TIME,
-	NL80211_STA_STAT_RX_BYTES,
-	NL80211_STA_STAT_TX_BYTES,
+enum nl80211_mpath_info {
+	__NL80211_MPATH_INFO_INVALID,
+	NL80211_MPATH_INFO_FRAME_QLEN,
+	NL80211_MPATH_INFO_DSN,
+	NL80211_MPATH_INFO_METRIC,
+	NL80211_MPATH_INFO_EXPTIME,
+	NL80211_MPATH_INFO_FLAGS,
+	NL80211_MPATH_INFO_DISCOVERY_TIMEOUT,
+	NL80211_MPATH_INFO_DISCOVERY_RETRIES,
 
 	/* keep last */
-	__NL80211_STA_STAT_AFTER_LAST,
-	NL80211_STA_STAT_MAX = __NL80211_STA_STAT_AFTER_LAST - 1
+	__NL80211_MPATH_INFO_AFTER_LAST,
+	NL80211_MPATH_INFO_MAX = __NL80211_MPATH_INFO_AFTER_LAST - 1
 };
 
 /**
diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h
index ab4caf63954f..e00750836ba5 100644
--- a/include/net/cfg80211.h
+++ b/include/net/cfg80211.h
@@ -12,6 +12,16 @@
  * Copyright 2006, 2007	Johannes Berg <johannes@sipsolutions.net>
  */
 
+/**
+ * struct vif_params - describes virtual interface parameters
+ * @mesh_id: mesh ID to use
+ * @mesh_id_len: length of the mesh ID
+ */
+struct vif_params {
+       u8 *mesh_id;
+       int mesh_id_len;
+};
+
 /* Radiotap header iteration
  *   implemented in net/wireless/radiotap.c
  *   docs in Documentation/networking/radiotap-headers.txt
@@ -108,6 +118,19 @@ enum station_flags {
 	STATION_FLAG_WME		= 1<<NL80211_STA_FLAG_WME,
 };
 
+/**
+ * enum plink_action - actions to perform in mesh peers
+ *
+ * @PLINK_ACTION_INVALID: action 0 is reserved
+ * @PLINK_ACTION_OPEN: start mesh peer link establishment
+ * @PLINK_ACTION_BLOCL: block traffic from this mesh peer
+ */
+enum plink_actions {
+	PLINK_ACTION_INVALID,
+	PLINK_ACTION_OPEN,
+	PLINK_ACTION_BLOCK,
+};
+
 /**
  * struct station_parameters - station parameters
  *
@@ -128,39 +151,52 @@ struct station_parameters {
 	int listen_interval;
 	u16 aid;
 	u8 supported_rates_len;
+	u8 plink_action;
 };
 
 /**
- * enum station_stats_flags - station statistics flags
+ * enum station_info_flags - station information flags
  *
- * Used by the driver to indicate which info in &struct station_stats
- * it has filled in during get_station().
+ * Used by the driver to indicate which info in &struct station_info
+ * it has filled in during get_station() or dump_station().
  *
- * @STATION_STAT_INACTIVE_TIME: @inactive_time filled
- * @STATION_STAT_RX_BYTES: @rx_bytes filled
- * @STATION_STAT_TX_BYTES: @tx_bytes filled
+ * @STATION_INFO_INACTIVE_TIME: @inactive_time filled
+ * @STATION_INFO_RX_BYTES: @rx_bytes filled
+ * @STATION_INFO_TX_BYTES: @tx_bytes filled
+ * @STATION_INFO_LLID: @llid filled
+ * @STATION_INFO_PLID: @plid filled
+ * @STATION_INFO_PLINK_STATE: @plink_state filled
  */
-enum station_stats_flags {
-	STATION_STAT_INACTIVE_TIME	= 1<<0,
-	STATION_STAT_RX_BYTES		= 1<<1,
-	STATION_STAT_TX_BYTES		= 1<<2,
+enum station_info_flags {
+	STATION_INFO_INACTIVE_TIME	= 1<<0,
+	STATION_INFO_RX_BYTES		= 1<<1,
+	STATION_INFO_TX_BYTES		= 1<<2,
+	STATION_INFO_LLID		= 1<<3,
+	STATION_INFO_PLID		= 1<<4,
+	STATION_INFO_PLINK_STATE	= 1<<5,
 };
 
 /**
- * struct station_stats - station statistics
+ * struct station_info - station information
  *
- * Station information filled by driver for get_station().
+ * Station information filled by driver for get_station() and dump_station.
  *
- * @filled: bitflag of flags from &enum station_stats_flags
+ * @filled: bitflag of flags from &enum station_info_flags
  * @inactive_time: time since last station activity (tx/rx) in milliseconds
  * @rx_bytes: bytes received from this station
  * @tx_bytes: bytes transmitted to this station
+ * @llid: mesh local link id
+ * @plid: mesh peer link id
+ * @plink_state: mesh peer link state
  */
-struct station_stats {
+struct station_info {
 	u32 filled;
 	u32 inactive_time;
 	u32 rx_bytes;
 	u32 tx_bytes;
+	u16 llid;
+	u16 plid;
+	u8 plink_state;
 };
 
 /**
@@ -183,6 +219,56 @@ enum monitor_flags {
 	MONITOR_FLAG_COOK_FRAMES	= 1<<NL80211_MNTR_FLAG_COOK_FRAMES,
 };
 
+/**
+ * enum mpath_info_flags -  mesh path information flags
+ *
+ * Used by the driver to indicate which info in &struct mpath_info it has filled
+ * in during get_station() or dump_station().
+ *
+ * MPATH_INFO_FRAME_QLEN: @frame_qlen filled
+ * MPATH_INFO_DSN: @dsn filled
+ * MPATH_INFO_METRIC: @metric filled
+ * MPATH_INFO_EXPTIME: @exptime filled
+ * MPATH_INFO_DISCOVERY_TIMEOUT: @discovery_timeout filled
+ * MPATH_INFO_DISCOVERY_RETRIES: @discovery_retries filled
+ * MPATH_INFO_FLAGS: @flags filled
+ */
+enum mpath_info_flags {
+	MPATH_INFO_FRAME_QLEN		= BIT(0),
+	MPATH_INFO_DSN			= BIT(1),
+	MPATH_INFO_METRIC		= BIT(2),
+	MPATH_INFO_EXPTIME		= BIT(3),
+	MPATH_INFO_DISCOVERY_TIMEOUT	= BIT(4),
+	MPATH_INFO_DISCOVERY_RETRIES	= BIT(5),
+	MPATH_INFO_FLAGS		= BIT(6),
+};
+
+/**
+ * struct mpath_info - mesh path information
+ *
+ * Mesh path information filled by driver for get_mpath() and dump_mpath().
+ *
+ * @filled: bitfield of flags from &enum mpath_info_flags
+ * @frame_qlen: number of queued frames for this destination
+ * @dsn: destination sequence number
+ * @metric: metric (cost) of this mesh path
+ * @exptime: expiration time for the mesh path from now, in msecs
+ * @flags: mesh path flags
+ * @discovery_timeout: total mesh path discovery timeout, in msecs
+ * @discovery_retries: mesh path discovery retries
+ */
+struct mpath_info {
+	u32 filled;
+	u32 frame_qlen;
+	u32 dsn;
+	u32 metric;
+	u32 exptime;
+	u32 discovery_timeout;
+	u8 discovery_retries;
+	u8 flags;
+};
+
+
 /* from net/wireless.h */
 struct wiphy;
 
@@ -230,13 +316,17 @@ struct wiphy;
  * @del_station: Remove a station; @mac may be NULL to remove all stations.
  *
  * @change_station: Modify a given station.
+ *
+ * @set_mesh_cfg: set mesh parameters (by now, just mesh id)
  */
 struct cfg80211_ops {
 	int	(*add_virtual_intf)(struct wiphy *wiphy, char *name,
-				    enum nl80211_iftype type, u32 *flags);
+				    enum nl80211_iftype type, u32 *flags,
+				    struct vif_params *params);
 	int	(*del_virtual_intf)(struct wiphy *wiphy, int ifindex);
 	int	(*change_virtual_intf)(struct wiphy *wiphy, int ifindex,
-				       enum nl80211_iftype type, u32 *flags);
+				       enum nl80211_iftype type, u32 *flags,
+				       struct vif_params *params);
 
 	int	(*add_key)(struct wiphy *wiphy, struct net_device *netdev,
 			   u8 key_index, u8 *mac_addr,
@@ -264,7 +354,22 @@ struct cfg80211_ops {
 	int	(*change_station)(struct wiphy *wiphy, struct net_device *dev,
 				  u8 *mac, struct station_parameters *params);
 	int	(*get_station)(struct wiphy *wiphy, struct net_device *dev,
-			       u8 *mac, struct station_stats *stats);
+			       u8 *mac, struct station_info *sinfo);
+	int	(*dump_station)(struct wiphy *wiphy, struct net_device *dev,
+			       int idx, u8 *mac, struct station_info *sinfo);
+
+	int	(*add_mpath)(struct wiphy *wiphy, struct net_device *dev,
+			       u8 *dst, u8 *next_hop);
+	int	(*del_mpath)(struct wiphy *wiphy, struct net_device *dev,
+			       u8 *dst);
+	int	(*change_mpath)(struct wiphy *wiphy, struct net_device *dev,
+				  u8 *dst, u8 *next_hop);
+	int	(*get_mpath)(struct wiphy *wiphy, struct net_device *dev,
+			       u8 *dst, u8 *next_hop,
+			       struct mpath_info *pinfo);
+	int	(*dump_mpath)(struct wiphy *wiphy, struct net_device *dev,
+			       int idx, u8 *dst, u8 *next_hop,
+			       struct mpath_info *pinfo);
 };
 
 #endif /* __NET_CFG80211_H */
diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c
index e7535ffc8e1c..006da6a2e71b 100644
--- a/net/mac80211/cfg.c
+++ b/net/mac80211/cfg.c
@@ -34,7 +34,8 @@ nl80211_type_to_mac80211_type(enum nl80211_iftype type)
 }
 
 static int ieee80211_add_iface(struct wiphy *wiphy, char *name,
-			       enum nl80211_iftype type, u32 *flags)
+			       enum nl80211_iftype type, u32 *flags,
+			       struct vif_params *params)
 {
 	struct ieee80211_local *local = wiphy_priv(wiphy);
 	enum ieee80211_if_types itype;
@@ -78,7 +79,8 @@ static int ieee80211_del_iface(struct wiphy *wiphy, int ifindex)
 }
 
 static int ieee80211_change_iface(struct wiphy *wiphy, int ifindex,
-				  enum nl80211_iftype type, u32 *flags)
+				  enum nl80211_iftype type, u32 *flags,
+				  struct vif_params *params)
 {
 	struct ieee80211_local *local = wiphy_priv(wiphy);
 	struct net_device *dev;
@@ -296,7 +298,7 @@ static int ieee80211_config_default_key(struct wiphy *wiphy,
 }
 
 static int ieee80211_get_station(struct wiphy *wiphy, struct net_device *dev,
-				 u8 *mac, struct station_stats *stats)
+				 u8 *mac, struct station_info *sinfo)
 {
 	struct ieee80211_local *local = wdev_priv(dev->ieee80211_ptr);
 	struct sta_info *sta;
@@ -307,13 +309,13 @@ static int ieee80211_get_station(struct wiphy *wiphy, struct net_device *dev,
 
 	/* XXX: verify sta->dev == dev */
 
-	stats->filled = STATION_STAT_INACTIVE_TIME |
-			STATION_STAT_RX_BYTES |
-			STATION_STAT_TX_BYTES;
+	sinfo->filled = STATION_INFO_INACTIVE_TIME |
+			STATION_INFO_RX_BYTES |
+			STATION_INFO_TX_BYTES;
 
-	stats->inactive_time = jiffies_to_msecs(jiffies - sta->last_rx);
-	stats->rx_bytes = sta->rx_bytes;
-	stats->tx_bytes = sta->tx_bytes;
+	sinfo->inactive_time = jiffies_to_msecs(jiffies - sta->last_rx);
+	sinfo->rx_bytes = sta->rx_bytes;
+	sinfo->tx_bytes = sta->tx_bytes;
 
 	sta_info_put(sta);
 
diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c
index 5b3474798b8d..64a7460af734 100644
--- a/net/wireless/nl80211.c
+++ b/net/wireless/nl80211.c
@@ -81,8 +81,12 @@ static struct nla_policy nl80211_policy[NL80211_ATTR_MAX+1] __read_mostly = {
 	[NL80211_ATTR_STA_LISTEN_INTERVAL] = { .type = NLA_U16 },
 	[NL80211_ATTR_STA_SUPPORTED_RATES] = { .type = NLA_BINARY,
 					       .len = NL80211_MAX_SUPP_RATES },
+	[NL80211_ATTR_STA_PLINK_ACTION] = { .type = NLA_U8 },
 	[NL80211_ATTR_STA_VLAN] = { .type = NLA_U32 },
 	[NL80211_ATTR_MNTR_FLAGS] = { .type = NLA_NESTED },
+	[NL80211_ATTR_MESH_ID] = { .type = NLA_BINARY,
+				.len = IEEE80211_MAX_MESH_ID_LEN },
+	[NL80211_ATTR_MPATH_NEXT_HOP] = { .type = NLA_U32 },
 };
 
 /* message building helper */
@@ -369,11 +373,14 @@ static int parse_monitor_flags(struct nlattr *nla, u32 *mntrflags)
 static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info)
 {
 	struct cfg80211_registered_device *drv;
+	struct vif_params params;
 	int err, ifindex;
 	enum nl80211_iftype type;
 	struct net_device *dev;
 	u32 flags;
 
+	memset(&params, 0, sizeof(params));
+
 	if (info->attrs[NL80211_ATTR_IFTYPE]) {
 		type = nla_get_u32(info->attrs[NL80211_ATTR_IFTYPE]);
 		if (type > NL80211_IFTYPE_MAX)
@@ -392,12 +399,18 @@ static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info)
 		goto unlock;
 	}
 
+	if (type == NL80211_IFTYPE_MESH_POINT &&
+	    info->attrs[NL80211_ATTR_MESH_ID]) {
+		params.mesh_id = nla_data(info->attrs[NL80211_ATTR_MESH_ID]);
+		params.mesh_id_len = nla_len(info->attrs[NL80211_ATTR_MESH_ID]);
+	}
+
 	rtnl_lock();
 	err = parse_monitor_flags(type == NL80211_IFTYPE_MONITOR ?
 				  info->attrs[NL80211_ATTR_MNTR_FLAGS] : NULL,
 				  &flags);
 	err = drv->ops->change_virtual_intf(&drv->wiphy, ifindex,
-					    type, err ? NULL : &flags);
+					    type, err ? NULL : &flags, &params);
 	rtnl_unlock();
 
  unlock:
@@ -408,10 +421,13 @@ static int nl80211_set_interface(struct sk_buff *skb, struct genl_info *info)
 static int nl80211_new_interface(struct sk_buff *skb, struct genl_info *info)
 {
 	struct cfg80211_registered_device *drv;
+	struct vif_params params;
 	int err;
 	enum nl80211_iftype type = NL80211_IFTYPE_UNSPECIFIED;
 	u32 flags;
 
+	memset(&params, 0, sizeof(params));
+
 	if (!info->attrs[NL80211_ATTR_IFNAME])
 		return -EINVAL;
 
@@ -430,15 +446,22 @@ static int nl80211_new_interface(struct sk_buff *skb, struct genl_info *info)
 		goto unlock;
 	}
 
+	if (type == NL80211_IFTYPE_MESH_POINT &&
+	    info->attrs[NL80211_ATTR_MESH_ID]) {
+		params.mesh_id = nla_data(info->attrs[NL80211_ATTR_MESH_ID]);
+		params.mesh_id_len = nla_len(info->attrs[NL80211_ATTR_MESH_ID]);
+	}
+
 	rtnl_lock();
 	err = parse_monitor_flags(type == NL80211_IFTYPE_MONITOR ?
 				  info->attrs[NL80211_ATTR_MNTR_FLAGS] : NULL,
 				  &flags);
 	err = drv->ops->add_virtual_intf(&drv->wiphy,
 		nla_data(info->attrs[NL80211_ATTR_IFNAME]),
-		type, err ? NULL : &flags);
+		type, err ? NULL : &flags, &params);
 	rtnl_unlock();
 
+
  unlock:
 	cfg80211_put_dev(drv);
 	return err;
@@ -866,10 +889,10 @@ static int parse_station_flags(struct nlattr *nla, u32 *staflags)
 
 static int nl80211_send_station(struct sk_buff *msg, u32 pid, u32 seq,
 				int flags, struct net_device *dev,
-				u8 *mac_addr, struct station_stats *stats)
+				u8 *mac_addr, struct station_info *sinfo)
 {
 	void *hdr;
-	struct nlattr *statsattr;
+	struct nlattr *sinfoattr;
 
 	hdr = nl80211hdr_put(msg, pid, seq, flags, NL80211_CMD_NEW_STATION);
 	if (!hdr)
@@ -878,20 +901,29 @@ static int nl80211_send_station(struct sk_buff *msg, u32 pid, u32 seq,
 	NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, dev->ifindex);
 	NLA_PUT(msg, NL80211_ATTR_MAC, ETH_ALEN, mac_addr);
 
-	statsattr = nla_nest_start(msg, NL80211_ATTR_STA_STATS);
-	if (!statsattr)
+	sinfoattr = nla_nest_start(msg, NL80211_ATTR_STA_INFO);
+	if (!sinfoattr)
 		goto nla_put_failure;
-	if (stats->filled & STATION_STAT_INACTIVE_TIME)
-		NLA_PUT_U32(msg, NL80211_STA_STAT_INACTIVE_TIME,
-			    stats->inactive_time);
-	if (stats->filled & STATION_STAT_RX_BYTES)
-		NLA_PUT_U32(msg, NL80211_STA_STAT_RX_BYTES,
-			    stats->rx_bytes);
-	if (stats->filled & STATION_STAT_TX_BYTES)
-		NLA_PUT_U32(msg, NL80211_STA_STAT_TX_BYTES,
-			    stats->tx_bytes);
-
-	nla_nest_end(msg, statsattr);
+	if (sinfo->filled & STATION_INFO_INACTIVE_TIME)
+		NLA_PUT_U32(msg, NL80211_STA_INFO_INACTIVE_TIME,
+			    sinfo->inactive_time);
+	if (sinfo->filled & STATION_INFO_RX_BYTES)
+		NLA_PUT_U32(msg, NL80211_STA_INFO_RX_BYTES,
+			    sinfo->rx_bytes);
+	if (sinfo->filled & STATION_INFO_TX_BYTES)
+		NLA_PUT_U32(msg, NL80211_STA_INFO_TX_BYTES,
+			    sinfo->tx_bytes);
+	if (sinfo->filled & STATION_INFO_LLID)
+		NLA_PUT_U16(msg, NL80211_STA_INFO_LLID,
+			    sinfo->llid);
+	if (sinfo->filled & STATION_INFO_PLID)
+		NLA_PUT_U16(msg, NL80211_STA_INFO_PLID,
+			    sinfo->plid);
+	if (sinfo->filled & STATION_INFO_PLINK_STATE)
+		NLA_PUT_U8(msg, NL80211_STA_INFO_PLINK_STATE,
+			    sinfo->plink_state);
+
+	nla_nest_end(msg, sinfoattr);
 
 	return genlmsg_end(msg, hdr);
 
@@ -899,17 +931,80 @@ static int nl80211_send_station(struct sk_buff *msg, u32 pid, u32 seq,
 	return genlmsg_cancel(msg, hdr);
 }
 
+static int nl80211_dump_station(struct sk_buff *skb,
+		struct netlink_callback *cb)
+{
+	int wp_idx = 0;
+	int if_idx = 0;
+	int sta_idx = cb->args[2];
+	int wp_start = cb->args[0];
+	int if_start = cb->args[1];
+	struct station_info sinfo;
+	struct cfg80211_registered_device *dev;
+	struct wireless_dev *wdev;
+	u8 mac_addr[ETH_ALEN];
+	int err;
+	int exit = 0;
+
+	/* TODO: filter by device */
+	mutex_lock(&cfg80211_drv_mutex);
+	list_for_each_entry(dev, &cfg80211_drv_list, list) {
+		if (exit)
+			break;
+		if (++wp_idx < wp_start)
+			continue;
+		if_idx = 0;
+
+		mutex_lock(&dev->devlist_mtx);
+		list_for_each_entry(wdev, &dev->netdev_list, list) {
+			if (exit)
+				break;
+			if (++if_idx < if_start)
+				continue;
+			if (!dev->ops->dump_station)
+				continue;
+
+			for (;; ++sta_idx) {
+				rtnl_lock();
+				err = dev->ops->dump_station(&dev->wiphy,
+						wdev->netdev, sta_idx, mac_addr,
+						&sinfo);
+				rtnl_unlock();
+				if (err) {
+					sta_idx = 0;
+					break;
+				}
+				if (nl80211_send_station(skb,
+						NETLINK_CB(cb->skb).pid,
+						cb->nlh->nlmsg_seq, NLM_F_MULTI,
+						wdev->netdev, mac_addr,
+						&sinfo) < 0) {
+					exit = 1;
+					break;
+				}
+			}
+		}
+		mutex_unlock(&dev->devlist_mtx);
+	}
+	mutex_unlock(&cfg80211_drv_mutex);
+
+	cb->args[0] = wp_idx;
+	cb->args[1] = if_idx;
+	cb->args[2] = sta_idx;
+
+	return skb->len;
+}
 
 static int nl80211_get_station(struct sk_buff *skb, struct genl_info *info)
 {
 	struct cfg80211_registered_device *drv;
 	int err;
 	struct net_device *dev;
-	struct station_stats stats;
+	struct station_info sinfo;
 	struct sk_buff *msg;
 	u8 *mac_addr = NULL;
 
-	memset(&stats, 0, sizeof(stats));
+	memset(&sinfo, 0, sizeof(sinfo));
 
 	if (!info->attrs[NL80211_ATTR_MAC])
 		return -EINVAL;
@@ -926,15 +1021,18 @@ static int nl80211_get_station(struct sk_buff *skb, struct genl_info *info)
 	}
 
 	rtnl_lock();
-	err = drv->ops->get_station(&drv->wiphy, dev, mac_addr, &stats);
+	err = drv->ops->get_station(&drv->wiphy, dev, mac_addr, &sinfo);
 	rtnl_unlock();
 
+	if (err)
+		goto out;
+
 	msg = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
 	if (!msg)
 		goto out;
 
 	if (nl80211_send_station(msg, info->snd_pid, info->snd_seq, 0,
-				 dev, mac_addr, &stats) < 0)
+				 dev, mac_addr, &sinfo) < 0)
 		goto out_free;
 
 	err = genlmsg_unicast(msg, info->snd_pid);
@@ -1005,6 +1103,10 @@ static int nl80211_set_station(struct sk_buff *skb, struct genl_info *info)
 				&params.station_flags))
 		return -EINVAL;
 
+	if (info->attrs[NL80211_ATTR_STA_PLINK_ACTION])
+		params.plink_action =
+		    nla_get_u8(info->attrs[NL80211_ATTR_STA_PLINK_ACTION]);
+
 	err = get_drv_dev_by_info_ifindex(info, &drv, &dev);
 	if (err)
 		return err;
@@ -1119,6 +1221,273 @@ static int nl80211_del_station(struct sk_buff *skb, struct genl_info *info)
 	return err;
 }
 
+static int nl80211_send_mpath(struct sk_buff *msg, u32 pid, u32 seq,
+				int flags, struct net_device *dev,
+				u8 *dst, u8 *next_hop,
+				struct mpath_info *pinfo)
+{
+	void *hdr;
+	struct nlattr *pinfoattr;
+
+	hdr = nl80211hdr_put(msg, pid, seq, flags, NL80211_CMD_NEW_STATION);
+	if (!hdr)
+		return -1;
+
+	NLA_PUT_U32(msg, NL80211_ATTR_IFINDEX, dev->ifindex);
+	NLA_PUT(msg, NL80211_ATTR_MAC, ETH_ALEN, dst);
+	NLA_PUT(msg, NL80211_ATTR_MPATH_NEXT_HOP, ETH_ALEN, next_hop);
+
+	pinfoattr = nla_nest_start(msg, NL80211_ATTR_MPATH_INFO);
+	if (!pinfoattr)
+		goto nla_put_failure;
+	if (pinfo->filled & MPATH_INFO_FRAME_QLEN)
+		NLA_PUT_U32(msg, NL80211_MPATH_INFO_FRAME_QLEN,
+			    pinfo->frame_qlen);
+	if (pinfo->filled & MPATH_INFO_DSN)
+		NLA_PUT_U32(msg, NL80211_MPATH_INFO_DSN,
+			    pinfo->dsn);
+	if (pinfo->filled & MPATH_INFO_METRIC)
+		NLA_PUT_U32(msg, NL80211_MPATH_INFO_METRIC,
+			    pinfo->metric);
+	if (pinfo->filled & MPATH_INFO_EXPTIME)
+		NLA_PUT_U32(msg, NL80211_MPATH_INFO_EXPTIME,
+			    pinfo->exptime);
+	if (pinfo->filled & MPATH_INFO_FLAGS)
+		NLA_PUT_U8(msg, NL80211_MPATH_INFO_FLAGS,
+			    pinfo->flags);
+	if (pinfo->filled & MPATH_INFO_DISCOVERY_TIMEOUT)
+		NLA_PUT_U32(msg, NL80211_MPATH_INFO_DISCOVERY_TIMEOUT,
+			    pinfo->discovery_timeout);
+	if (pinfo->filled & MPATH_INFO_DISCOVERY_RETRIES)
+		NLA_PUT_U8(msg, NL80211_MPATH_INFO_DISCOVERY_RETRIES,
+			    pinfo->discovery_retries);
+
+	nla_nest_end(msg, pinfoattr);
+
+	return genlmsg_end(msg, hdr);
+
+ nla_put_failure:
+	return genlmsg_cancel(msg, hdr);
+}
+
+static int nl80211_dump_mpath(struct sk_buff *skb,
+		struct netlink_callback *cb)
+{
+	int wp_idx = 0;
+	int if_idx = 0;
+	int sta_idx = cb->args[2];
+	int wp_start = cb->args[0];
+	int if_start = cb->args[1];
+	struct mpath_info pinfo;
+	struct cfg80211_registered_device *dev;
+	struct wireless_dev *wdev;
+	u8 dst[ETH_ALEN];
+	u8 next_hop[ETH_ALEN];
+	int err;
+	int exit = 0;
+
+	/* TODO: filter by device */
+	mutex_lock(&cfg80211_drv_mutex);
+	list_for_each_entry(dev, &cfg80211_drv_list, list) {
+		if (exit)
+			break;
+		if (++wp_idx < wp_start)
+			continue;
+		if_idx = 0;
+
+		mutex_lock(&dev->devlist_mtx);
+		list_for_each_entry(wdev, &dev->netdev_list, list) {
+			if (exit)
+				break;
+			if (++if_idx < if_start)
+				continue;
+			if (!dev->ops->dump_mpath)
+				continue;
+
+			for (;; ++sta_idx) {
+				rtnl_lock();
+				err = dev->ops->dump_mpath(&dev->wiphy,
+						wdev->netdev, sta_idx, dst,
+						next_hop, &pinfo);
+				rtnl_unlock();
+				if (err) {
+					sta_idx = 0;
+					break;
+				}
+				if (nl80211_send_mpath(skb,
+						NETLINK_CB(cb->skb).pid,
+						cb->nlh->nlmsg_seq, NLM_F_MULTI,
+						wdev->netdev, dst, next_hop,
+						&pinfo) < 0) {
+					exit = 1;
+					break;
+				}
+			}
+		}
+		mutex_unlock(&dev->devlist_mtx);
+	}
+	mutex_unlock(&cfg80211_drv_mutex);
+
+	cb->args[0] = wp_idx;
+	cb->args[1] = if_idx;
+	cb->args[2] = sta_idx;
+
+	return skb->len;
+}
+
+static int nl80211_get_mpath(struct sk_buff *skb, struct genl_info *info)
+{
+	struct cfg80211_registered_device *drv;
+	int err;
+	struct net_device *dev;
+	struct mpath_info pinfo;
+	struct sk_buff *msg;
+	u8 *dst = NULL;
+	u8 next_hop[ETH_ALEN];
+
+	memset(&pinfo, 0, sizeof(pinfo));
+
+	if (!info->attrs[NL80211_ATTR_MAC])
+		return -EINVAL;
+
+	dst = nla_data(info->attrs[NL80211_ATTR_MAC]);
+
+	err = get_drv_dev_by_info_ifindex(info, &drv, &dev);
+	if (err)
+		return err;
+
+	if (!drv->ops->get_mpath) {
+		err = -EOPNOTSUPP;
+		goto out;
+	}
+
+	rtnl_lock();
+	err = drv->ops->get_mpath(&drv->wiphy, dev, dst, next_hop, &pinfo);
+	rtnl_unlock();
+
+	if (err)
+		goto out;
+
+	msg = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
+	if (!msg)
+		goto out;
+
+	if (nl80211_send_mpath(msg, info->snd_pid, info->snd_seq, 0,
+				 dev, dst, next_hop, &pinfo) < 0)
+		goto out_free;
+
+	err = genlmsg_unicast(msg, info->snd_pid);
+	goto out;
+
+ out_free:
+	nlmsg_free(msg);
+
+ out:
+	cfg80211_put_dev(drv);
+	dev_put(dev);
+	return err;
+}
+
+static int nl80211_set_mpath(struct sk_buff *skb, struct genl_info *info)
+{
+	struct cfg80211_registered_device *drv;
+	int err;
+	struct net_device *dev;
+	u8 *dst = NULL;
+	u8 *next_hop = NULL;
+
+	if (!info->attrs[NL80211_ATTR_MAC])
+		return -EINVAL;
+
+	if (!info->attrs[NL80211_ATTR_MPATH_NEXT_HOP])
+		return -EINVAL;
+
+	dst = nla_data(info->attrs[NL80211_ATTR_MAC]);
+	next_hop = nla_data(info->attrs[NL80211_ATTR_MPATH_NEXT_HOP]);
+
+	err = get_drv_dev_by_info_ifindex(info, &drv, &dev);
+	if (err)
+		return err;
+
+	if (!drv->ops->change_mpath) {
+		err = -EOPNOTSUPP;
+		goto out;
+	}
+
+	rtnl_lock();
+	err = drv->ops->change_mpath(&drv->wiphy, dev, dst, next_hop);
+	rtnl_unlock();
+
+ out:
+	cfg80211_put_dev(drv);
+	dev_put(dev);
+	return err;
+}
+static int nl80211_new_mpath(struct sk_buff *skb, struct genl_info *info)
+{
+	struct cfg80211_registered_device *drv;
+	int err;
+	struct net_device *dev;
+	u8 *dst = NULL;
+	u8 *next_hop = NULL;
+
+	if (!info->attrs[NL80211_ATTR_MAC])
+		return -EINVAL;
+
+	if (!info->attrs[NL80211_ATTR_MPATH_NEXT_HOP])
+		return -EINVAL;
+
+	dst = nla_data(info->attrs[NL80211_ATTR_MAC]);
+	next_hop = nla_data(info->attrs[NL80211_ATTR_MPATH_NEXT_HOP]);
+
+	err = get_drv_dev_by_info_ifindex(info, &drv, &dev);
+	if (err)
+		return err;
+
+	if (!drv->ops->add_mpath) {
+		err = -EOPNOTSUPP;
+		goto out;
+	}
+
+	rtnl_lock();
+	err = drv->ops->add_mpath(&drv->wiphy, dev, dst, next_hop);
+	rtnl_unlock();
+
+ out:
+	cfg80211_put_dev(drv);
+	dev_put(dev);
+	return err;
+}
+
+static int nl80211_del_mpath(struct sk_buff *skb, struct genl_info *info)
+{
+	struct cfg80211_registered_device *drv;
+	int err;
+	struct net_device *dev;
+	u8 *dst = NULL;
+
+	if (info->attrs[NL80211_ATTR_MAC])
+		dst = nla_data(info->attrs[NL80211_ATTR_MAC]);
+
+	err = get_drv_dev_by_info_ifindex(info, &drv, &dev);
+	if (err)
+		return err;
+
+	if (!drv->ops->del_mpath) {
+		err = -EOPNOTSUPP;
+		goto out;
+	}
+
+	rtnl_lock();
+	err = drv->ops->del_mpath(&drv->wiphy, dev, dst);
+	rtnl_unlock();
+
+ out:
+	cfg80211_put_dev(drv);
+	dev_put(dev);
+	return err;
+}
+
 static struct genl_ops nl80211_ops[] = {
 	{
 		.cmd = NL80211_CMD_GET_WIPHY,
@@ -1203,7 +1572,7 @@ static struct genl_ops nl80211_ops[] = {
 	{
 		.cmd = NL80211_CMD_GET_STATION,
 		.doit = nl80211_get_station,
-		/* TODO: implement dumpit */
+		.dumpit = nl80211_dump_station,
 		.policy = nl80211_policy,
 		.flags = GENL_ADMIN_PERM,
 	},
@@ -1225,6 +1594,31 @@ static struct genl_ops nl80211_ops[] = {
 		.policy = nl80211_policy,
 		.flags = GENL_ADMIN_PERM,
 	},
+	{
+		.cmd = NL80211_CMD_GET_MPATH,
+		.doit = nl80211_get_mpath,
+		.dumpit = nl80211_dump_mpath,
+		.policy = nl80211_policy,
+		.flags = GENL_ADMIN_PERM,
+	},
+	{
+		.cmd = NL80211_CMD_SET_MPATH,
+		.doit = nl80211_set_mpath,
+		.policy = nl80211_policy,
+		.flags = GENL_ADMIN_PERM,
+	},
+	{
+		.cmd = NL80211_CMD_NEW_MPATH,
+		.doit = nl80211_new_mpath,
+		.policy = nl80211_policy,
+		.flags = GENL_ADMIN_PERM,
+	},
+	{
+		.cmd = NL80211_CMD_DEL_MPATH,
+		.doit = nl80211_del_mpath,
+		.policy = nl80211_policy,
+		.flags = GENL_ADMIN_PERM,
+	},
 };
 
 /* multicast groups */
-- 
cgit v1.2.3


From aab547ce0d1493d400b6468c521a0137cd8c1edf Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Fri, 29 Feb 2008 11:36:12 +0100
Subject: ssb: Add Gigabit Ethernet driver

This adds the Gigabit Ethernet driver for the SSB
Gigabit Ethernet core. This driver actually is a frontend to
the Tigon3 driver. So the real work is done by tg3.
This device is used in the Linksys WRT350N.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/Kconfig                 |   9 ++
 drivers/ssb/Makefile                |   1 +
 drivers/ssb/driver_gige.c           | 294 ++++++++++++++++++++++++++++++++++++
 drivers/ssb/driver_mipscore.c       |   1 +
 drivers/ssb/driver_pcicore.c        | 160 +++++++++++---------
 drivers/ssb/embedded.c              |  90 +++++++++++
 drivers/ssb/main.c                  |  30 +++-
 drivers/ssb/ssb_private.h           |   2 +
 include/linux/ssb/ssb.h             |   7 +
 include/linux/ssb/ssb_driver_gige.h | 174 +++++++++++++++++++++
 include/linux/ssb/ssb_driver_pci.h  |  19 +++
 11 files changed, 715 insertions(+), 72 deletions(-)
 create mode 100644 drivers/ssb/driver_gige.c
 create mode 100644 include/linux/ssb/ssb_driver_gige.h

(limited to 'include/linux')

diff --git a/drivers/ssb/Kconfig b/drivers/ssb/Kconfig
index adea792fb675..f69ef0ba2613 100644
--- a/drivers/ssb/Kconfig
+++ b/drivers/ssb/Kconfig
@@ -125,4 +125,13 @@ config SSB_DRIVER_EXTIF
 
 	  If unsure, say N
 
+config SSB_DRIVER_GIGE
+	bool "SSB Broadcom Gigabit Ethernet driver"
+	depends on SSB_PCIHOST_POSSIBLE && SSB_EMBEDDED && MIPS
+	help
+	  Driver for the Sonics Silicon Backplane attached
+	  Broadcom Gigabit Ethernet.
+
+	  If unsure, say N
+
 endmenu
diff --git a/drivers/ssb/Makefile b/drivers/ssb/Makefile
index de94c2eb7a37..910f35e32fc9 100644
--- a/drivers/ssb/Makefile
+++ b/drivers/ssb/Makefile
@@ -11,6 +11,7 @@ ssb-y					+= driver_chipcommon.o
 ssb-$(CONFIG_SSB_DRIVER_MIPS)		+= driver_mipscore.o
 ssb-$(CONFIG_SSB_DRIVER_EXTIF)		+= driver_extif.o
 ssb-$(CONFIG_SSB_DRIVER_PCICORE)	+= driver_pcicore.o
+ssb-$(CONFIG_SSB_DRIVER_GIGE)		+= driver_gige.o
 
 # b43 pci-ssb-bridge driver
 # Not strictly a part of SSB, but kept here for convenience
diff --git a/drivers/ssb/driver_gige.c b/drivers/ssb/driver_gige.c
new file mode 100644
index 000000000000..172f90407b93
--- /dev/null
+++ b/drivers/ssb/driver_gige.c
@@ -0,0 +1,294 @@
+/*
+ * Sonics Silicon Backplane
+ * Broadcom Gigabit Ethernet core driver
+ *
+ * Copyright 2008, Broadcom Corporation
+ * Copyright 2008, Michael Buesch <mb@bu3sch.de>
+ *
+ * Licensed under the GNU/GPL. See COPYING for details.
+ */
+
+#include <linux/ssb/ssb.h>
+#include <linux/ssb/ssb_driver_gige.h>
+#include <linux/pci.h>
+#include <linux/pci_regs.h>
+
+
+/*
+MODULE_DESCRIPTION("SSB Broadcom Gigabit Ethernet driver");
+MODULE_AUTHOR("Michael Buesch");
+MODULE_LICENSE("GPL");
+*/
+
+static const struct ssb_device_id ssb_gige_tbl[] = {
+	SSB_DEVICE(SSB_VENDOR_BROADCOM, SSB_DEV_ETHERNET_GBIT, SSB_ANY_REV),
+	SSB_DEVTABLE_END
+};
+/* MODULE_DEVICE_TABLE(ssb, ssb_gige_tbl); */
+
+
+static inline u8 gige_read8(struct ssb_gige *dev, u16 offset)
+{
+	return ssb_read8(dev->dev, offset);
+}
+
+static inline u16 gige_read16(struct ssb_gige *dev, u16 offset)
+{
+	return ssb_read16(dev->dev, offset);
+}
+
+static inline u32 gige_read32(struct ssb_gige *dev, u16 offset)
+{
+	return ssb_read32(dev->dev, offset);
+}
+
+static inline void gige_write8(struct ssb_gige *dev,
+			       u16 offset, u8 value)
+{
+	ssb_write8(dev->dev, offset, value);
+}
+
+static inline void gige_write16(struct ssb_gige *dev,
+				u16 offset, u16 value)
+{
+	ssb_write16(dev->dev, offset, value);
+}
+
+static inline void gige_write32(struct ssb_gige *dev,
+				u16 offset, u32 value)
+{
+	ssb_write32(dev->dev, offset, value);
+}
+
+static inline
+u8 gige_pcicfg_read8(struct ssb_gige *dev, unsigned int offset)
+{
+	BUG_ON(offset >= 256);
+	return gige_read8(dev, SSB_GIGE_PCICFG + offset);
+}
+
+static inline
+u16 gige_pcicfg_read16(struct ssb_gige *dev, unsigned int offset)
+{
+	BUG_ON(offset >= 256);
+	return gige_read16(dev, SSB_GIGE_PCICFG + offset);
+}
+
+static inline
+u32 gige_pcicfg_read32(struct ssb_gige *dev, unsigned int offset)
+{
+	BUG_ON(offset >= 256);
+	return gige_read32(dev, SSB_GIGE_PCICFG + offset);
+}
+
+static inline
+void gige_pcicfg_write8(struct ssb_gige *dev,
+			unsigned int offset, u8 value)
+{
+	BUG_ON(offset >= 256);
+	gige_write8(dev, SSB_GIGE_PCICFG + offset, value);
+}
+
+static inline
+void gige_pcicfg_write16(struct ssb_gige *dev,
+			 unsigned int offset, u16 value)
+{
+	BUG_ON(offset >= 256);
+	gige_write16(dev, SSB_GIGE_PCICFG + offset, value);
+}
+
+static inline
+void gige_pcicfg_write32(struct ssb_gige *dev,
+			 unsigned int offset, u32 value)
+{
+	BUG_ON(offset >= 256);
+	gige_write32(dev, SSB_GIGE_PCICFG + offset, value);
+}
+
+static int ssb_gige_pci_read_config(struct pci_bus *bus, unsigned int devfn,
+				    int reg, int size, u32 *val)
+{
+	struct ssb_gige *dev = container_of(bus->ops, struct ssb_gige, pci_ops);
+	unsigned long flags;
+
+	if ((PCI_SLOT(devfn) > 0) || (PCI_FUNC(devfn) > 0))
+		return PCIBIOS_DEVICE_NOT_FOUND;
+	if (reg >= 256)
+		return PCIBIOS_DEVICE_NOT_FOUND;
+
+	spin_lock_irqsave(&dev->lock, flags);
+	switch (size) {
+	case 1:
+		*val = gige_pcicfg_read8(dev, reg);
+		break;
+	case 2:
+		*val = gige_pcicfg_read16(dev, reg);
+		break;
+	case 4:
+		*val = gige_pcicfg_read32(dev, reg);
+		break;
+	default:
+		WARN_ON(1);
+	}
+	spin_unlock_irqrestore(&dev->lock, flags);
+
+	return PCIBIOS_SUCCESSFUL;
+}
+
+static int ssb_gige_pci_write_config(struct pci_bus *bus, unsigned int devfn,
+				     int reg, int size, u32 val)
+{
+	struct ssb_gige *dev = container_of(bus->ops, struct ssb_gige, pci_ops);
+	unsigned long flags;
+
+	if ((PCI_SLOT(devfn) > 0) || (PCI_FUNC(devfn) > 0))
+		return PCIBIOS_DEVICE_NOT_FOUND;
+	if (reg >= 256)
+		return PCIBIOS_DEVICE_NOT_FOUND;
+
+	spin_lock_irqsave(&dev->lock, flags);
+	switch (size) {
+	case 1:
+		gige_pcicfg_write8(dev, reg, val);
+		break;
+	case 2:
+		gige_pcicfg_write16(dev, reg, val);
+		break;
+	case 4:
+		gige_pcicfg_write32(dev, reg, val);
+		break;
+	default:
+		WARN_ON(1);
+	}
+	spin_unlock_irqrestore(&dev->lock, flags);
+
+	return PCIBIOS_SUCCESSFUL;
+}
+
+static int ssb_gige_probe(struct ssb_device *sdev, const struct ssb_device_id *id)
+{
+	struct ssb_gige *dev;
+	u32 base, tmslow, tmshigh;
+
+	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+	if (!dev)
+		return -ENOMEM;
+	dev->dev = sdev;
+
+	spin_lock_init(&dev->lock);
+	dev->pci_controller.pci_ops = &dev->pci_ops;
+	dev->pci_controller.io_resource = &dev->io_resource;
+	dev->pci_controller.mem_resource = &dev->mem_resource;
+	dev->pci_controller.io_map_base = 0x800;
+	dev->pci_ops.read = ssb_gige_pci_read_config;
+	dev->pci_ops.write = ssb_gige_pci_write_config;
+
+	dev->io_resource.name = SSB_GIGE_IO_RES_NAME;
+	dev->io_resource.start = 0x800;
+	dev->io_resource.end = 0x8FF;
+	dev->io_resource.flags = IORESOURCE_IO | IORESOURCE_PCI_FIXED;
+
+	if (!ssb_device_is_enabled(sdev))
+		ssb_device_enable(sdev, 0);
+
+	/* Setup BAR0. This is a 64k MMIO region. */
+	base = ssb_admatch_base(ssb_read32(sdev, SSB_ADMATCH1));
+	gige_pcicfg_write32(dev, PCI_BASE_ADDRESS_0, base);
+	gige_pcicfg_write32(dev, PCI_BASE_ADDRESS_1, 0);
+
+	dev->mem_resource.name = SSB_GIGE_MEM_RES_NAME;
+	dev->mem_resource.start = base;
+	dev->mem_resource.end = base + 0x10000 - 1;
+	dev->mem_resource.flags = IORESOURCE_MEM | IORESOURCE_PCI_FIXED;
+
+	/* Enable the memory region. */
+	gige_pcicfg_write16(dev, PCI_COMMAND,
+			    gige_pcicfg_read16(dev, PCI_COMMAND)
+			    | PCI_COMMAND_MEMORY);
+
+	/* Write flushing is controlled by the Flush Status Control register.
+	 * We want to flush every register write with a timeout and we want
+	 * to disable the IRQ mask while flushing to avoid concurrency.
+	 * Note that automatic write flushing does _not_ work from
+	 * an IRQ handler. The driver must flush manually by reading a register.
+	 */
+	gige_write32(dev, SSB_GIGE_SHIM_FLUSHSTAT, 0x00000068);
+
+	/* Check if we have an RGMII or GMII PHY-bus.
+	 * On RGMII do not bypass the DLLs */
+	tmslow = ssb_read32(sdev, SSB_TMSLOW);
+	tmshigh = ssb_read32(sdev, SSB_TMSHIGH);
+	if (tmshigh & SSB_GIGE_TMSHIGH_RGMII) {
+		tmslow &= ~SSB_GIGE_TMSLOW_TXBYPASS;
+		tmslow &= ~SSB_GIGE_TMSLOW_RXBYPASS;
+		dev->has_rgmii = 1;
+	} else {
+		tmslow |= SSB_GIGE_TMSLOW_TXBYPASS;
+		tmslow |= SSB_GIGE_TMSLOW_RXBYPASS;
+		dev->has_rgmii = 0;
+	}
+	tmslow |= SSB_GIGE_TMSLOW_DLLEN;
+	ssb_write32(sdev, SSB_TMSLOW, tmslow);
+
+	ssb_set_drvdata(sdev, dev);
+	register_pci_controller(&dev->pci_controller);
+
+	return 0;
+}
+
+bool pdev_is_ssb_gige_core(struct pci_dev *pdev)
+{
+	if (!pdev->resource[0].name)
+		return 0;
+	return (strcmp(pdev->resource[0].name, SSB_GIGE_MEM_RES_NAME) == 0);
+}
+EXPORT_SYMBOL(pdev_is_ssb_gige_core);
+
+int ssb_gige_pcibios_plat_dev_init(struct ssb_device *sdev,
+				   struct pci_dev *pdev)
+{
+	struct ssb_gige *dev = ssb_get_drvdata(sdev);
+	struct resource *res;
+
+	if (pdev->bus->ops != &dev->pci_ops) {
+		/* The PCI device is not on this SSB GigE bridge device. */
+		return -ENODEV;
+	}
+
+	/* Fixup the PCI resources. */
+	res = &(pdev->resource[0]);
+	res->flags = IORESOURCE_MEM | IORESOURCE_PCI_FIXED;
+	res->name = dev->mem_resource.name;
+	res->start = dev->mem_resource.start;
+	res->end = dev->mem_resource.end;
+
+	/* Fixup interrupt lines. */
+	pdev->irq = ssb_mips_irq(sdev) + 2;
+	pci_write_config_byte(pdev, PCI_INTERRUPT_LINE, pdev->irq);
+
+	return 0;
+}
+
+int ssb_gige_map_irq(struct ssb_device *sdev,
+		     const struct pci_dev *pdev)
+{
+	struct ssb_gige *dev = ssb_get_drvdata(sdev);
+
+	if (pdev->bus->ops != &dev->pci_ops) {
+		/* The PCI device is not on this SSB GigE bridge device. */
+		return -ENODEV;
+	}
+
+	return ssb_mips_irq(sdev) + 2;
+}
+
+static struct ssb_driver ssb_gige_driver = {
+	.name		= "BCM-GigE",
+	.id_table	= ssb_gige_tbl,
+	.probe		= ssb_gige_probe,
+};
+
+int ssb_gige_init(void)
+{
+	return ssb_driver_register(&ssb_gige_driver);
+}
diff --git a/drivers/ssb/driver_mipscore.c b/drivers/ssb/driver_mipscore.c
index 3d3dd32bf3ab..e3fad3123ecb 100644
--- a/drivers/ssb/driver_mipscore.c
+++ b/drivers/ssb/driver_mipscore.c
@@ -209,6 +209,7 @@ void ssb_mipscore_init(struct ssb_mipscore *mcore)
 			/* fallthrough */
 		case SSB_DEV_PCI:
 		case SSB_DEV_ETHERNET:
+		case SSB_DEV_ETHERNET_GBIT:
 		case SSB_DEV_80211:
 		case SSB_DEV_USB20_HOST:
 			/* These devices get their own IRQ line if available, the rest goes on IRQ0 */
diff --git a/drivers/ssb/driver_pcicore.c b/drivers/ssb/driver_pcicore.c
index 74b9a8aea52b..33a7d5620474 100644
--- a/drivers/ssb/driver_pcicore.c
+++ b/drivers/ssb/driver_pcicore.c
@@ -60,77 +60,6 @@ static DEFINE_SPINLOCK(cfgspace_lock);
 /* Core to access the external PCI config space. Can only have one. */
 static struct ssb_pcicore *extpci_core;
 
-static u32 ssb_pcicore_pcibus_iobase = 0x100;
-static u32 ssb_pcicore_pcibus_membase = SSB_PCI_DMA;
-
-int pcibios_plat_dev_init(struct pci_dev *d)
-{
-	struct resource *res;
-	int pos, size;
-	u32 *base;
-
-	ssb_printk(KERN_INFO "PCI: Fixing up device %s\n",
-		   pci_name(d));
-
-	/* Fix up resource bases */
-	for (pos = 0; pos < 6; pos++) {
-		res = &d->resource[pos];
-		if (res->flags & IORESOURCE_IO)
-			base = &ssb_pcicore_pcibus_iobase;
-		else
-			base = &ssb_pcicore_pcibus_membase;
-		res->flags |= IORESOURCE_PCI_FIXED;
-		if (res->end) {
-			size = res->end - res->start + 1;
-			if (*base & (size - 1))
-				*base = (*base + size) & ~(size - 1);
-			res->start = *base;
-			res->end = res->start + size - 1;
-			*base += size;
-			pci_write_config_dword(d, PCI_BASE_ADDRESS_0 + (pos << 2), res->start);
-		}
-		/* Fix up PCI bridge BAR0 only */
-		if (d->bus->number == 0 && PCI_SLOT(d->devfn) == 0)
-			break;
-	}
-	/* Fix up interrupt lines */
-	d->irq = ssb_mips_irq(extpci_core->dev) + 2;
-	pci_write_config_byte(d, PCI_INTERRUPT_LINE, d->irq);
-
-	return 0;
-}
-
-static void __init ssb_fixup_pcibridge(struct pci_dev *dev)
-{
-	u8 lat;
-
-	if (dev->bus->number != 0 || PCI_SLOT(dev->devfn) != 0)
-		return;
-
-	ssb_printk(KERN_INFO "PCI: Fixing up bridge %s\n", pci_name(dev));
-
-	/* Enable PCI bridge bus mastering and memory space */
-	pci_set_master(dev);
-	if (pcibios_enable_device(dev, ~0) < 0) {
-		ssb_printk(KERN_ERR "PCI: SSB bridge enable failed\n");
-		return;
-	}
-
-	/* Enable PCI bridge BAR1 prefetch and burst */
-	pci_write_config_dword(dev, SSB_BAR1_CONTROL, 3);
-
-	/* Make sure our latency is high enough to handle the devices behind us */
-	lat = 168;
-	ssb_printk(KERN_INFO "PCI: Fixing latency timer of device %s to %u\n",
-		   pci_name(dev), lat);
-	pci_write_config_byte(dev, PCI_LATENCY_TIMER, lat);
-}
-DECLARE_PCI_FIXUP_EARLY(PCI_ANY_ID, PCI_ANY_ID, ssb_fixup_pcibridge);
-
-int __init pcibios_map_irq(const struct pci_dev *dev, u8 slot, u8 pin)
-{
-	return ssb_mips_irq(extpci_core->dev) + 2;
-}
 
 static u32 get_cfgspace_addr(struct ssb_pcicore *pc,
 			     unsigned int bus, unsigned int dev,
@@ -320,6 +249,95 @@ static struct pci_controller ssb_pcicore_controller = {
 	.mem_offset	= 0x24000000,
 };
 
+static u32 ssb_pcicore_pcibus_iobase = 0x100;
+static u32 ssb_pcicore_pcibus_membase = SSB_PCI_DMA;
+
+/* This function is called when doing a pci_enable_device().
+ * We must first check if the device is a device on the PCI-core bridge. */
+int ssb_pcicore_plat_dev_init(struct pci_dev *d)
+{
+	struct resource *res;
+	int pos, size;
+	u32 *base;
+
+	if (d->bus->ops != &ssb_pcicore_pciops) {
+		/* This is not a device on the PCI-core bridge. */
+		return -ENODEV;
+	}
+
+	ssb_printk(KERN_INFO "PCI: Fixing up device %s\n",
+		   pci_name(d));
+
+	/* Fix up resource bases */
+	for (pos = 0; pos < 6; pos++) {
+		res = &d->resource[pos];
+		if (res->flags & IORESOURCE_IO)
+			base = &ssb_pcicore_pcibus_iobase;
+		else
+			base = &ssb_pcicore_pcibus_membase;
+		res->flags |= IORESOURCE_PCI_FIXED;
+		if (res->end) {
+			size = res->end - res->start + 1;
+			if (*base & (size - 1))
+				*base = (*base + size) & ~(size - 1);
+			res->start = *base;
+			res->end = res->start + size - 1;
+			*base += size;
+			pci_write_config_dword(d, PCI_BASE_ADDRESS_0 + (pos << 2), res->start);
+		}
+		/* Fix up PCI bridge BAR0 only */
+		if (d->bus->number == 0 && PCI_SLOT(d->devfn) == 0)
+			break;
+	}
+	/* Fix up interrupt lines */
+	d->irq = ssb_mips_irq(extpci_core->dev) + 2;
+	pci_write_config_byte(d, PCI_INTERRUPT_LINE, d->irq);
+
+	return 0;
+}
+
+/* Early PCI fixup for a device on the PCI-core bridge. */
+static void ssb_pcicore_fixup_pcibridge(struct pci_dev *dev)
+{
+	u8 lat;
+
+	if (dev->bus->ops != &ssb_pcicore_pciops) {
+		/* This is not a device on the PCI-core bridge. */
+		return;
+	}
+	if (dev->bus->number != 0 || PCI_SLOT(dev->devfn) != 0)
+		return;
+
+	ssb_printk(KERN_INFO "PCI: Fixing up bridge %s\n", pci_name(dev));
+
+	/* Enable PCI bridge bus mastering and memory space */
+	pci_set_master(dev);
+	if (pcibios_enable_device(dev, ~0) < 0) {
+		ssb_printk(KERN_ERR "PCI: SSB bridge enable failed\n");
+		return;
+	}
+
+	/* Enable PCI bridge BAR1 prefetch and burst */
+	pci_write_config_dword(dev, SSB_BAR1_CONTROL, 3);
+
+	/* Make sure our latency is high enough to handle the devices behind us */
+	lat = 168;
+	ssb_printk(KERN_INFO "PCI: Fixing latency timer of device %s to %u\n",
+		   pci_name(dev), lat);
+	pci_write_config_byte(dev, PCI_LATENCY_TIMER, lat);
+}
+DECLARE_PCI_FIXUP_EARLY(PCI_ANY_ID, PCI_ANY_ID, ssb_pcicore_fixup_pcibridge);
+
+/* PCI device IRQ mapping. */
+int ssb_pcicore_pcibios_map_irq(const struct pci_dev *dev, u8 slot, u8 pin)
+{
+	if (dev->bus->ops != &ssb_pcicore_pciops) {
+		/* This is not a device on the PCI-core bridge. */
+		return -ENODEV;
+	}
+	return ssb_mips_irq(extpci_core->dev) + 2;
+}
+
 static void ssb_pcicore_init_hostmode(struct ssb_pcicore *pc)
 {
 	u32 val;
diff --git a/drivers/ssb/embedded.c b/drivers/ssb/embedded.c
index d3ade821555c..7dc3a6b41397 100644
--- a/drivers/ssb/embedded.c
+++ b/drivers/ssb/embedded.c
@@ -10,6 +10,9 @@
 
 #include <linux/ssb/ssb.h>
 #include <linux/ssb/ssb_embedded.h>
+#include <linux/ssb/ssb_driver_pci.h>
+#include <linux/ssb/ssb_driver_gige.h>
+#include <linux/pci.h>
 
 #include "ssb_private.h"
 
@@ -130,3 +133,90 @@ u32 ssb_gpio_polarity(struct ssb_bus *bus, u32 mask, u32 value)
 	return res;
 }
 EXPORT_SYMBOL(ssb_gpio_polarity);
+
+#ifdef CONFIG_SSB_DRIVER_GIGE
+static int gige_pci_init_callback(struct ssb_bus *bus, unsigned long data)
+{
+	struct pci_dev *pdev = (struct pci_dev *)data;
+	struct ssb_device *dev;
+	unsigned int i;
+	int res;
+
+	for (i = 0; i < bus->nr_devices; i++) {
+		dev = &(bus->devices[i]);
+		if (dev->id.coreid != SSB_DEV_ETHERNET_GBIT)
+			continue;
+		if (!dev->dev ||
+		    !dev->dev->driver ||
+		    !device_is_registered(dev->dev))
+			continue;
+		res = ssb_gige_pcibios_plat_dev_init(dev, pdev);
+		if (res >= 0)
+			return res;
+	}
+
+	return -ENODEV;
+}
+#endif /* CONFIG_SSB_DRIVER_GIGE */
+
+int ssb_pcibios_plat_dev_init(struct pci_dev *dev)
+{
+	int err;
+
+	err = ssb_pcicore_plat_dev_init(dev);
+	if (!err)
+		return 0;
+#ifdef CONFIG_SSB_DRIVER_GIGE
+	err = ssb_for_each_bus_call((unsigned long)dev, gige_pci_init_callback);
+	if (err >= 0)
+		return err;
+#endif
+	/* This is not a PCI device on any SSB device. */
+
+	return -ENODEV;
+}
+
+#ifdef CONFIG_SSB_DRIVER_GIGE
+static int gige_map_irq_callback(struct ssb_bus *bus, unsigned long data)
+{
+	const struct pci_dev *pdev = (const struct pci_dev *)data;
+	struct ssb_device *dev;
+	unsigned int i;
+	int res;
+
+	for (i = 0; i < bus->nr_devices; i++) {
+		dev = &(bus->devices[i]);
+		if (dev->id.coreid != SSB_DEV_ETHERNET_GBIT)
+			continue;
+		if (!dev->dev ||
+		    !dev->dev->driver ||
+		    !device_is_registered(dev->dev))
+			continue;
+		res = ssb_gige_map_irq(dev, pdev);
+		if (res >= 0)
+			return res;
+	}
+
+	return -ENODEV;
+}
+#endif /* CONFIG_SSB_DRIVER_GIGE */
+
+int ssb_pcibios_map_irq(const struct pci_dev *dev, u8 slot, u8 pin)
+{
+	int res;
+
+	/* Check if this PCI device is a device on a SSB bus or device
+	 * and return the IRQ number for it. */
+
+	res = ssb_pcicore_pcibios_map_irq(dev, slot, pin);
+	if (res >= 0)
+		return res;
+#ifdef CONFIG_SSB_DRIVER_GIGE
+	res = ssb_for_each_bus_call((unsigned long)dev, gige_map_irq_callback);
+	if (res >= 0)
+		return res;
+#endif
+	/* This is not a PCI device on any SSB device. */
+
+	return -ENODEV;
+}
diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index 8db40c4b86e9..49d7bbb9bea7 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -14,6 +14,7 @@
 #include <linux/io.h>
 #include <linux/ssb/ssb.h>
 #include <linux/ssb/ssb_regs.h>
+#include <linux/ssb/ssb_driver_gige.h>
 #include <linux/dma-mapping.h>
 #include <linux/pci.h>
 
@@ -68,6 +69,25 @@ found:
 }
 #endif /* CONFIG_SSB_PCIHOST */
 
+int ssb_for_each_bus_call(unsigned long data,
+			  int (*func)(struct ssb_bus *bus, unsigned long data))
+{
+	struct ssb_bus *bus;
+	int res;
+
+	ssb_buses_lock();
+	list_for_each_entry(bus, &buses, list) {
+		res = func(bus, data);
+		if (res >= 0) {
+			ssb_buses_unlock();
+			return res;
+		}
+	}
+	ssb_buses_unlock();
+
+	return -ENODEV;
+}
+
 static struct ssb_device *ssb_device_get(struct ssb_device *dev)
 {
 	if (dev)
@@ -1171,7 +1191,14 @@ static int __init ssb_modinit(void)
 	err = b43_pci_ssb_bridge_init();
 	if (err) {
 		ssb_printk(KERN_ERR "Broadcom 43xx PCI-SSB-bridge "
-			   "initialization failed");
+			   "initialization failed\n");
+		/* don't fail SSB init because of this */
+		err = 0;
+	}
+	err = ssb_gige_init();
+	if (err) {
+		ssb_printk(KERN_ERR "SSB Broadcom Gigabit Ethernet "
+			   "driver initialization failed\n");
 		/* don't fail SSB init because of this */
 		err = 0;
 	}
@@ -1185,6 +1212,7 @@ fs_initcall(ssb_modinit);
 
 static void __exit ssb_modexit(void)
 {
+	ssb_gige_exit();
 	b43_pci_ssb_bridge_exit();
 	bus_unregister(&ssb_bustype);
 }
diff --git a/drivers/ssb/ssb_private.h b/drivers/ssb/ssb_private.h
index 21eca2b5118b..d03b20983b1e 100644
--- a/drivers/ssb/ssb_private.h
+++ b/drivers/ssb/ssb_private.h
@@ -118,6 +118,8 @@ extern u32 ssb_calc_clock_rate(u32 plltype, u32 n, u32 m);
 extern int ssb_devices_freeze(struct ssb_bus *bus);
 extern int ssb_devices_thaw(struct ssb_bus *bus);
 extern struct ssb_bus *ssb_pci_dev_to_bus(struct pci_dev *pdev);
+int ssb_for_each_bus_call(unsigned long data,
+			  int (*func)(struct ssb_bus *bus, unsigned long data));
 
 /* b43_pci_bridge.c */
 #ifdef CONFIG_SSB_B43_PCI_BRIDGE
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index 860d28c6d149..b7c388972fcf 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -422,5 +422,12 @@ extern int ssb_bus_powerup(struct ssb_bus *bus, bool dynamic_pctl);
 extern u32 ssb_admatch_base(u32 adm);
 extern u32 ssb_admatch_size(u32 adm);
 
+/* PCI device mapping and fixup routines.
+ * Called from the architecture pcibios init code.
+ * These are only available on SSB_EMBEDDED configurations. */
+#ifdef CONFIG_SSB_EMBEDDED
+int ssb_pcibios_plat_dev_init(struct pci_dev *dev);
+int ssb_pcibios_map_irq(const struct pci_dev *dev, u8 slot, u8 pin);
+#endif /* CONFIG_SSB_EMBEDDED */
 
 #endif /* LINUX_SSB_H_ */
diff --git a/include/linux/ssb/ssb_driver_gige.h b/include/linux/ssb/ssb_driver_gige.h
new file mode 100644
index 000000000000..01fbdf5fef22
--- /dev/null
+++ b/include/linux/ssb/ssb_driver_gige.h
@@ -0,0 +1,174 @@
+#ifndef LINUX_SSB_DRIVER_GIGE_H_
+#define LINUX_SSB_DRIVER_GIGE_H_
+
+#include <linux/ssb/ssb.h>
+#include <linux/pci.h>
+#include <linux/spinlock.h>
+
+
+#ifdef CONFIG_SSB_DRIVER_GIGE
+
+
+#define SSB_GIGE_PCIIO			0x0000 /* PCI I/O Registers (1024 bytes) */
+#define SSB_GIGE_RESERVED		0x0400 /* Reserved (1024 bytes) */
+#define SSB_GIGE_PCICFG			0x0800 /* PCI config space (256 bytes) */
+#define SSB_GIGE_SHIM_FLUSHSTAT		0x0C00 /* PCI to OCP: Flush status control (32bit) */
+#define SSB_GIGE_SHIM_FLUSHRDA		0x0C04 /* PCI to OCP: Flush read address (32bit) */
+#define SSB_GIGE_SHIM_FLUSHTO		0x0C08 /* PCI to OCP: Flush timeout counter (32bit) */
+#define SSB_GIGE_SHIM_BARRIER		0x0C0C /* PCI to OCP: Barrier register (32bit) */
+#define SSB_GIGE_SHIM_MAOCPSI		0x0C10 /* PCI to OCP: MaocpSI Control (32bit) */
+#define SSB_GIGE_SHIM_SIOCPMA		0x0C14 /* PCI to OCP: SiocpMa Control (32bit) */
+
+/* TM Status High flags */
+#define SSB_GIGE_TMSHIGH_RGMII		0x00010000 /* Have an RGMII PHY-bus */
+/* TM Status Low flags */
+#define SSB_GIGE_TMSLOW_TXBYPASS	0x00080000 /* TX bypass (no delay) */
+#define SSB_GIGE_TMSLOW_RXBYPASS	0x00100000 /* RX bypass (no delay) */
+#define SSB_GIGE_TMSLOW_DLLEN		0x01000000 /* Enable DLL controls */
+
+/* Boardflags (low) */
+#define SSB_GIGE_BFL_ROBOSWITCH		0x0010
+
+
+#define SSB_GIGE_MEM_RES_NAME		"SSB Broadcom 47xx GigE memory"
+#define SSB_GIGE_IO_RES_NAME		"SSB Broadcom 47xx GigE I/O"
+
+struct ssb_gige {
+	struct ssb_device *dev;
+
+	spinlock_t lock;
+
+	/* True, if the device has an RGMII bus.
+	 * False, if the device has a GMII bus. */
+	bool has_rgmii;
+
+	/* The PCI controller device. */
+	struct pci_controller pci_controller;
+	struct pci_ops pci_ops;
+	struct resource mem_resource;
+	struct resource io_resource;
+};
+
+/* Check whether a PCI device is a SSB Gigabit Ethernet core. */
+extern bool pdev_is_ssb_gige_core(struct pci_dev *pdev);
+
+/* Convert a pci_dev pointer to a ssb_gige pointer. */
+static inline struct ssb_gige * pdev_to_ssb_gige(struct pci_dev *pdev)
+{
+	if (!pdev_is_ssb_gige_core(pdev))
+		return NULL;
+	return container_of(pdev->bus->ops, struct ssb_gige, pci_ops);
+}
+
+/* Returns whether the PHY is connected by an RGMII bus. */
+static inline bool ssb_gige_is_rgmii(struct pci_dev *pdev)
+{
+	struct ssb_gige *dev = pdev_to_ssb_gige(pdev);
+	return (dev ? dev->has_rgmii : 0);
+}
+
+/* Returns whether we have a Roboswitch. */
+static inline bool ssb_gige_have_roboswitch(struct pci_dev *pdev)
+{
+	struct ssb_gige *dev = pdev_to_ssb_gige(pdev);
+	if (dev)
+		return !!(dev->dev->bus->sprom.boardflags_lo &
+			  SSB_GIGE_BFL_ROBOSWITCH);
+	return 0;
+}
+
+/* Returns whether we can only do one DMA at once. */
+static inline bool ssb_gige_one_dma_at_once(struct pci_dev *pdev)
+{
+	struct ssb_gige *dev = pdev_to_ssb_gige(pdev);
+	if (dev)
+		return ((dev->dev->bus->chip_id == 0x4785) &&
+			(dev->dev->bus->chip_rev < 2));
+	return 0;
+}
+
+/* Returns whether we must flush posted writes. */
+static inline bool ssb_gige_must_flush_posted_writes(struct pci_dev *pdev)
+{
+	struct ssb_gige *dev = pdev_to_ssb_gige(pdev);
+	if (dev)
+		return (dev->dev->bus->chip_id == 0x4785);
+	return 0;
+}
+
+extern char * nvram_get(const char *name);
+/* Get the device MAC address */
+static inline void ssb_gige_get_macaddr(struct pci_dev *pdev, u8 *macaddr)
+{
+#ifdef CONFIG_BCM947XX
+	char *res = nvram_get("et0macaddr");
+	if (res)
+		memcpy(macaddr, res, 6);
+#endif
+}
+
+extern int ssb_gige_pcibios_plat_dev_init(struct ssb_device *sdev,
+					  struct pci_dev *pdev);
+extern int ssb_gige_map_irq(struct ssb_device *sdev,
+			    const struct pci_dev *pdev);
+
+/* The GigE driver is not a standalone module, because we don't have support
+ * for unregistering the driver. So we could not unload the module anyway. */
+extern int ssb_gige_init(void);
+static inline void ssb_gige_exit(void)
+{
+	/* Currently we can not unregister the GigE driver,
+	 * because we can not unregister the PCI bridge. */
+	BUG();
+}
+
+
+#else /* CONFIG_SSB_DRIVER_GIGE */
+/* Gigabit Ethernet driver disabled */
+
+
+static inline int ssb_gige_pcibios_plat_dev_init(struct ssb_device *sdev,
+						 struct pci_dev *pdev)
+{
+	return -ENOSYS;
+}
+static inline int ssb_gige_map_irq(struct ssb_device *sdev,
+				   const struct pci_dev *pdev)
+{
+	return -ENOSYS;
+}
+static inline int ssb_gige_init(void)
+{
+	return 0;
+}
+static inline void ssb_gige_exit(void)
+{
+}
+
+static inline bool pdev_is_ssb_gige_core(struct pci_dev *pdev)
+{
+	return 0;
+}
+static inline struct ssb_gige * pdev_to_ssb_gige(struct pci_dev *pdev)
+{
+	return NULL;
+}
+static inline bool ssb_gige_is_rgmii(struct pci_dev *pdev)
+{
+	return 0;
+}
+static inline bool ssb_gige_have_roboswitch(struct pci_dev *pdev)
+{
+	return 0;
+}
+static inline bool ssb_gige_one_dma_at_once(struct pci_dev *pdev)
+{
+	return 0;
+}
+static inline bool ssb_gige_must_flush_posted_writes(struct pci_dev *pdev)
+{
+	return 0;
+}
+
+#endif /* CONFIG_SSB_DRIVER_GIGE */
+#endif /* LINUX_SSB_DRIVER_GIGE_H_ */
diff --git a/include/linux/ssb/ssb_driver_pci.h b/include/linux/ssb/ssb_driver_pci.h
index 5e25bac4ed31..41e330e51c2a 100644
--- a/include/linux/ssb/ssb_driver_pci.h
+++ b/include/linux/ssb/ssb_driver_pci.h
@@ -1,6 +1,11 @@
 #ifndef LINUX_SSB_PCICORE_H_
 #define LINUX_SSB_PCICORE_H_
 
+#include <linux/types.h>
+
+struct pci_dev;
+
+
 #ifdef CONFIG_SSB_DRIVER_PCICORE
 
 /* PCI core registers. */
@@ -88,6 +93,9 @@ extern void ssb_pcicore_init(struct ssb_pcicore *pc);
 extern int ssb_pcicore_dev_irqvecs_enable(struct ssb_pcicore *pc,
 					  struct ssb_device *dev);
 
+int ssb_pcicore_plat_dev_init(struct pci_dev *d);
+int ssb_pcicore_pcibios_map_irq(const struct pci_dev *dev, u8 slot, u8 pin);
+
 
 #else /* CONFIG_SSB_DRIVER_PCICORE */
 
@@ -107,5 +115,16 @@ int ssb_pcicore_dev_irqvecs_enable(struct ssb_pcicore *pc,
 	return 0;
 }
 
+static inline
+int ssb_pcicore_plat_dev_init(struct pci_dev *d)
+{
+	return -ENODEV;
+}
+static inline
+int ssb_pcicore_pcibios_map_irq(const struct pci_dev *dev, u8 slot, u8 pin)
+{
+	return -ENODEV;
+}
+
 #endif /* CONFIG_SSB_DRIVER_PCICORE */
 #endif /* LINUX_SSB_PCICORE_H_ */
-- 
cgit v1.2.3


From 1c61fc40fc264059ff41a614ed2d899127288281 Mon Sep 17 00:00:00 2001
From: Joe Perches <joe@perches.com>
Date: Wed, 5 Mar 2008 13:58:17 -0800
Subject: slab - use angle brackets for include of kmalloc_sizes.h

Make them all use angle brackets and the directory name.

Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
---
 include/linux/slab_def.h | 4 ++--
 mm/slab.c                | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/slab_def.h b/include/linux/slab_def.h
index fcc48096ee64..39c3a5eb8ebe 100644
--- a/include/linux/slab_def.h
+++ b/include/linux/slab_def.h
@@ -41,7 +41,7 @@ static inline void *kmalloc(size_t size, gfp_t flags)
 			goto found; \
 		else \
 			i++;
-#include "kmalloc_sizes.h"
+#include <linux/kmalloc_sizes.h>
 #undef CACHE
 		{
 			extern void __you_cannot_kmalloc_that_much(void);
@@ -75,7 +75,7 @@ static inline void *kmalloc_node(size_t size, gfp_t flags, int node)
 			goto found; \
 		else \
 			i++;
-#include "kmalloc_sizes.h"
+#include <linux/kmalloc_sizes.h>
 #undef CACHE
 		{
 			extern void __you_cannot_kmalloc_that_much(void);
diff --git a/mm/slab.c b/mm/slab.c
index 5d16c8a30499..f7faff72cf56 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -333,7 +333,7 @@ static __always_inline int index_of(const size_t size)
 		return i; \
 	else \
 		i++;
-#include "linux/kmalloc_sizes.h"
+#include <linux/kmalloc_sizes.h>
 #undef CACHE
 		__bad_size();
 	} else
-- 
cgit v1.2.3


From db8dac20d5199307dcfcf4e01dac4bda5edf9e89 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Thu, 6 Mar 2008 16:22:02 -0800
Subject: [UDP]: Revert udplite and code split.

This reverts commit db1ed684f6c430c4cdad67d058688b8a1b5e607c ("[IPV6]
UDP: Rename IPv6 UDP files."), commit
8be8af8fa4405652e6c0797db5465a4be8afb998 ("[IPV4] UDP: Move
IPv4-specific bits to other file.") and commit
e898d4db2749c6052072e9bc4448e396cbdeb06a ("[UDP]: Allow users to
configure UDP-Lite.").

First, udplite is of such small cost, and it is a core protocol just
like TCP and normal UDP are.

We spent enormous amounts of effort to make udplite share as much code
with core UDP as possible.  All of that work is less valuable if we're
just going to slap a config option on udplite support.

It is also causing build failures, as reported on linux-next, showing
that the changeset was not tested very well.  In fact, this is the
second build failure resulting from the udplite change.

Finally, the config options provided was a bool, instead of a modular
option.  Meaning the udplite code does not even get build tested
by allmodconfig builds, and furthermore the user is not presented
with a reasonable modular build option which is particularly needed
by distribution vendors.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/udp.h      |   10 -
 include/net/ipv6.h       |    5 -
 include/net/transp_v6.h  |    5 -
 include/net/udplite.h    |    9 +-
 net/ipv4/Kconfig         |   10 -
 net/ipv4/Makefile        |    3 +-
 net/ipv4/af_inet.c       |    7 +-
 net/ipv4/proc.c          |    5 +-
 net/ipv4/udp.c           | 1090 +++++++++++++++++++++++++++++++++++++++++++-
 net/ipv4/udp_ipv4.c      | 1134 ----------------------------------------------
 net/ipv4/udplite.c       |  121 +++++
 net/ipv4/udplite_ipv4.c  |  121 -----
 net/ipv6/Makefile        |    3 +-
 net/ipv6/af_inet6.c      |   14 -
 net/ipv6/ipv6_sockglue.c |    6 +-
 net/ipv6/proc.c          |    6 -
 net/ipv6/udp.c           | 1065 +++++++++++++++++++++++++++++++++++++++++++
 net/ipv6/udp_ipv6.c      | 1065 -------------------------------------------
 net/ipv6/udplite.c       |  125 +++++
 net/ipv6/udplite_ipv6.c  |  125 -----
 20 files changed, 2404 insertions(+), 2525 deletions(-)
 delete mode 100644 net/ipv4/udp_ipv4.c
 create mode 100644 net/ipv4/udplite.c
 delete mode 100644 net/ipv4/udplite_ipv4.c
 create mode 100644 net/ipv6/udp.c
 delete mode 100644 net/ipv6/udp_ipv6.c
 create mode 100644 net/ipv6/udplite.c
 delete mode 100644 net/ipv6/udplite_ipv6.c

(limited to 'include/linux')

diff --git a/include/linux/udp.h b/include/linux/udp.h
index 4144664d69d9..1e7b7cb5703b 100644
--- a/include/linux/udp.h
+++ b/include/linux/udp.h
@@ -70,10 +70,8 @@ struct udp_sock {
 #define UDPLITE_BIT      0x1  		/* set by udplite proto init function */
 #define UDPLITE_SEND_CC  0x2  		/* set via udplite setsockopt         */
 #define UDPLITE_RECV_CC  0x4		/* set via udplite setsocktopt        */
-#ifdef CONFIG_IP_UDPLITE
 	__u8		 pcflag;        /* marks socket as UDP-Lite if > 0    */
 	__u8		 unused[3];
-#endif
 	/*
 	 * For encapsulation sockets.
 	 */
@@ -85,15 +83,7 @@ static inline struct udp_sock *udp_sk(const struct sock *sk)
 	return (struct udp_sock *)sk;
 }
 
-#ifdef CONFIG_IP_UDPLITE
 #define IS_UDPLITE(__sk) (udp_sk(__sk)->pcflag)
-#define IS_PROTO_UDPLITE(__proto) ((__proto) == IPPROTO_UDPLITE)
-#define IS_SOL_UDPFAMILY(level) ((level) == SOL_UDP || (level) == SOL_UDPLITE)
-#else
-#define IS_UDPLITE(__sk) 0
-#define IS_PROTO_UDPLITE(__proto) 0
-#define IS_SOL_UDPFAMILY(level) ((level) == SOL_UDP)
-#endif
 
 #endif
 
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 5f6df50a33a9..8db06af1efbb 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -599,13 +599,8 @@ extern int  tcp6_proc_init(void);
 extern void tcp6_proc_exit(void);
 extern int  udp6_proc_init(void);
 extern void udp6_proc_exit(void);
-#ifdef CONFIG_IP_UDPLITE
 extern int  udplite6_proc_init(void);
 extern void udplite6_proc_exit(void);
-#else
-static inline int udplite6_proc_init(void) { return 0; }
-static inline void udplite6_proc_exit(void) { }
-#endif
 extern int  ipv6_misc_proc_init(void);
 extern void ipv6_misc_proc_exit(void);
 extern int snmp6_register_dev(struct inet6_dev *idev);
diff --git a/include/net/transp_v6.h b/include/net/transp_v6.h
index 902e6c6bc793..27394e0447d8 100644
--- a/include/net/transp_v6.h
+++ b/include/net/transp_v6.h
@@ -27,13 +27,8 @@ extern int				rawv6_init(void);
 extern void				rawv6_exit(void);
 extern int				udpv6_init(void);
 extern void				udpv6_exit(void);
-#ifdef CONFIG_IP_UDPLITE
 extern int 				udplitev6_init(void);
 extern void 				udplitev6_exit(void);
-#else
-static inline int udplitev6_init(void) { return 0; }
-static inline void udplitev6_exit(void) { }
-#endif
 extern int				tcpv6_init(void);
 extern void				tcpv6_exit(void);
 
diff --git a/include/net/udplite.h b/include/net/udplite.h
index 01ddb2c20264..b76b2e377af4 100644
--- a/include/net/udplite.h
+++ b/include/net/udplite.h
@@ -25,9 +25,7 @@ static __inline__ int udplite_getfrag(void *from, char *to, int  offset,
 /* Designate sk as UDP-Lite socket */
 static inline int udplite_sk_init(struct sock *sk)
 {
-#ifdef CONFIG_IP_UDPLITE
 	udp_sk(sk)->pcflag = UDPLITE_BIT;
-#endif
 	return 0;
 }
 
@@ -71,7 +69,7 @@ static inline int udplite_checksum_init(struct sk_buff *skb, struct udphdr *uh)
 static inline int udplite_sender_cscov(struct udp_sock *up, struct udphdr *uh)
 {
 	int cscov = up->len;
-#ifdef CONFIG_IP_UDPLITE
+
 	/*
 	 * Sender has set `partial coverage' option on UDP-Lite socket
 	 */
@@ -95,15 +93,13 @@ static inline int udplite_sender_cscov(struct udp_sock *up, struct udphdr *uh)
 	 *       illegal, we fall back to the defaults here.
 	 */
 	}
-#endif
 	return cscov;
 }
 
 static inline __wsum udplite_csum_outgoing(struct sock *sk, struct sk_buff *skb)
 {
-	__wsum csum = 0;
-#ifdef CONFIG_IP_UDPLITE
 	int cscov = udplite_sender_cscov(udp_sk(sk), udp_hdr(skb));
+	__wsum csum = 0;
 
 	skb->ip_summed = CHECKSUM_NONE;     /* no HW support for checksumming */
 
@@ -116,7 +112,6 @@ static inline __wsum udplite_csum_outgoing(struct sock *sk, struct sk_buff *skb)
 		if ((cscov -= len) <= 0)
 			break;
 	}
-#endif
 	return csum;
 }
 
diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig
index 5098fd2ff4d0..9c7e5ffb223d 100644
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -632,15 +632,5 @@ config TCP_MD5SIG
 
 	  If unsure, say N.
 
-config IP_UDPLITE
-	bool "IP: UDP-Lite Protocol (RFC 3828)"
-	default n
-	---help---
-	  UDP-Lite (RFC 3828) is a UDP-like protocol with variable-length
-	  checksum.  Read <file:Documentation/networking/udplite.txt> for
-	  details.
-
-	  If unsure, say N.
-
 source "net/ipv4/ipvs/Kconfig"
 
diff --git a/net/ipv4/Makefile b/net/ipv4/Makefile
index d5226241d5ed..ad40ef3f9ebc 100644
--- a/net/ipv4/Makefile
+++ b/net/ipv4/Makefile
@@ -8,7 +8,7 @@ obj-y     := route.o inetpeer.o protocol.o \
 	     inet_timewait_sock.o inet_connection_sock.o \
 	     tcp.o tcp_input.o tcp_output.o tcp_timer.o tcp_ipv4.o \
 	     tcp_minisocks.o tcp_cong.o \
-	     datagram.o raw.o udp.o udp_ipv4.o \
+	     datagram.o raw.o udp.o udplite.o \
 	     arp.o icmp.o devinet.o af_inet.o  igmp.o \
 	     fib_frontend.o fib_semantics.o \
 	     inet_fragment.o
@@ -49,7 +49,6 @@ obj-$(CONFIG_TCP_CONG_SCALABLE) += tcp_scalable.o
 obj-$(CONFIG_TCP_CONG_LP) += tcp_lp.o
 obj-$(CONFIG_TCP_CONG_YEAH) += tcp_yeah.o
 obj-$(CONFIG_TCP_CONG_ILLINOIS) += tcp_illinois.o
-obj-$(CONFIG_IP_UDPLITE) += udplite_ipv4.o
 obj-$(CONFIG_NETLABEL) += cipso_ipv4.o
 
 obj-$(CONFIG_XFRM) += xfrm4_policy.o xfrm4_state.o xfrm4_input.o \
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 25871c6c7444..4cb8a1385539 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1317,18 +1317,15 @@ static int __init init_ipv4_mibs(void)
 	if (snmp_mib_init((void **)udp_statistics,
 			  sizeof(struct udp_mib)) < 0)
 		goto err_udp_mib;
-#ifdef CONFIG_IP_UDPLITE
 	if (snmp_mib_init((void **)udplite_statistics,
 			  sizeof(struct udp_mib)) < 0)
 		goto err_udplite_mib;
-#endif
+
 	tcp_mib_init();
 
 	return 0;
 
-#ifdef CONFIG_IP_UDPLITE
 err_udplite_mib:
-#endif
 	snmp_mib_free((void **)udp_statistics);
 err_udp_mib:
 	snmp_mib_free((void **)tcp_statistics);
@@ -1426,10 +1423,8 @@ static int __init inet_init(void)
 	/* Setup UDP memory threshold */
 	udp_init();
 
-#ifdef CONFIG_IP_UDPLITE
 	/* Add UDP-Lite (RFC 3828) */
 	udplite4_register();
-#endif
 
 	/*
 	 *	Set the ICMP layer up
diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c
index d75ddb7fa4b8..d63474c6b400 100644
--- a/net/ipv4/proc.c
+++ b/net/ipv4/proc.c
@@ -59,9 +59,7 @@ static int sockstat_seq_show(struct seq_file *seq, void *v)
 		   atomic_read(&tcp_memory_allocated));
 	seq_printf(seq, "UDP: inuse %d mem %d\n", sock_prot_inuse_get(&udp_prot),
 		   atomic_read(&udp_memory_allocated));
-#ifdef CONFIG_IP_UDPLITE
 	seq_printf(seq, "UDPLITE: inuse %d\n", sock_prot_inuse_get(&udplite_prot));
-#endif
 	seq_printf(seq, "RAW: inuse %d\n", sock_prot_inuse_get(&raw_prot));
 	seq_printf(seq,  "FRAG: inuse %d memory %d\n",
 			ip_frag_nqueues(&init_net), ip_frag_mem(&init_net));
@@ -351,7 +349,6 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 			   snmp_fold_field((void **)udp_statistics,
 					   snmp4_udp_list[i].entry));
 
-#ifdef CONFIG_IP_UDPLITE
 	/* the UDP and UDP-Lite MIBs are the same */
 	seq_puts(seq, "\nUdpLite:");
 	for (i = 0; snmp4_udp_list[i].name != NULL; i++)
@@ -362,7 +359,7 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 		seq_printf(seq, " %lu",
 			   snmp_fold_field((void **)udplite_statistics,
 					   snmp4_udp_list[i].entry));
-#endif
+
 	seq_putc(seq, '\n');
 	return 0;
 }
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index c53d7673b57d..7ea1b67b6de1 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -246,6 +246,553 @@ int udp_get_port(struct sock *sk, unsigned short snum,
 	return  __udp_lib_get_port(sk, snum, udp_hash, scmp);
 }
 
+int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2)
+{
+	struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2);
+
+	return 	( !ipv6_only_sock(sk2)  &&
+		  (!inet1->rcv_saddr || !inet2->rcv_saddr ||
+		   inet1->rcv_saddr == inet2->rcv_saddr      ));
+}
+
+static inline int udp_v4_get_port(struct sock *sk, unsigned short snum)
+{
+	return udp_get_port(sk, snum, ipv4_rcv_saddr_equal);
+}
+
+/* UDP is nearly always wildcards out the wazoo, it makes no sense to try
+ * harder than this. -DaveM
+ */
+static struct sock *__udp4_lib_lookup(struct net *net, __be32 saddr,
+		__be16 sport, __be32 daddr, __be16 dport,
+		int dif, struct hlist_head udptable[])
+{
+	struct sock *sk, *result = NULL;
+	struct hlist_node *node;
+	unsigned short hnum = ntohs(dport);
+	int badness = -1;
+
+	read_lock(&udp_hash_lock);
+	sk_for_each(sk, node, &udptable[hnum & (UDP_HTABLE_SIZE - 1)]) {
+		struct inet_sock *inet = inet_sk(sk);
+
+		if (sk->sk_net == net && sk->sk_hash == hnum &&
+				!ipv6_only_sock(sk)) {
+			int score = (sk->sk_family == PF_INET ? 1 : 0);
+			if (inet->rcv_saddr) {
+				if (inet->rcv_saddr != daddr)
+					continue;
+				score+=2;
+			}
+			if (inet->daddr) {
+				if (inet->daddr != saddr)
+					continue;
+				score+=2;
+			}
+			if (inet->dport) {
+				if (inet->dport != sport)
+					continue;
+				score+=2;
+			}
+			if (sk->sk_bound_dev_if) {
+				if (sk->sk_bound_dev_if != dif)
+					continue;
+				score+=2;
+			}
+			if (score == 9) {
+				result = sk;
+				break;
+			} else if (score > badness) {
+				result = sk;
+				badness = score;
+			}
+		}
+	}
+	if (result)
+		sock_hold(result);
+	read_unlock(&udp_hash_lock);
+	return result;
+}
+
+static inline struct sock *udp_v4_mcast_next(struct sock *sk,
+					     __be16 loc_port, __be32 loc_addr,
+					     __be16 rmt_port, __be32 rmt_addr,
+					     int dif)
+{
+	struct hlist_node *node;
+	struct sock *s = sk;
+	unsigned short hnum = ntohs(loc_port);
+
+	sk_for_each_from(s, node) {
+		struct inet_sock *inet = inet_sk(s);
+
+		if (s->sk_hash != hnum					||
+		    (inet->daddr && inet->daddr != rmt_addr)		||
+		    (inet->dport != rmt_port && inet->dport)		||
+		    (inet->rcv_saddr && inet->rcv_saddr != loc_addr)	||
+		    ipv6_only_sock(s)					||
+		    (s->sk_bound_dev_if && s->sk_bound_dev_if != dif))
+			continue;
+		if (!ip_mc_sf_allow(s, loc_addr, rmt_addr, dif))
+			continue;
+		goto found;
+	}
+	s = NULL;
+found:
+	return s;
+}
+
+/*
+ * This routine is called by the ICMP module when it gets some
+ * sort of error condition.  If err < 0 then the socket should
+ * be closed and the error returned to the user.  If err > 0
+ * it's just the icmp type << 8 | icmp code.
+ * Header points to the ip header of the error packet. We move
+ * on past this. Then (as it used to claim before adjustment)
+ * header points to the first 8 bytes of the udp header.  We need
+ * to find the appropriate port.
+ */
+
+void __udp4_lib_err(struct sk_buff *skb, u32 info, struct hlist_head udptable[])
+{
+	struct inet_sock *inet;
+	struct iphdr *iph = (struct iphdr*)skb->data;
+	struct udphdr *uh = (struct udphdr*)(skb->data+(iph->ihl<<2));
+	const int type = icmp_hdr(skb)->type;
+	const int code = icmp_hdr(skb)->code;
+	struct sock *sk;
+	int harderr;
+	int err;
+
+	sk = __udp4_lib_lookup(skb->dev->nd_net, iph->daddr, uh->dest,
+			iph->saddr, uh->source, skb->dev->ifindex, udptable);
+	if (sk == NULL) {
+		ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
+		return;	/* No socket for error */
+	}
+
+	err = 0;
+	harderr = 0;
+	inet = inet_sk(sk);
+
+	switch (type) {
+	default:
+	case ICMP_TIME_EXCEEDED:
+		err = EHOSTUNREACH;
+		break;
+	case ICMP_SOURCE_QUENCH:
+		goto out;
+	case ICMP_PARAMETERPROB:
+		err = EPROTO;
+		harderr = 1;
+		break;
+	case ICMP_DEST_UNREACH:
+		if (code == ICMP_FRAG_NEEDED) { /* Path MTU discovery */
+			if (inet->pmtudisc != IP_PMTUDISC_DONT) {
+				err = EMSGSIZE;
+				harderr = 1;
+				break;
+			}
+			goto out;
+		}
+		err = EHOSTUNREACH;
+		if (code <= NR_ICMP_UNREACH) {
+			harderr = icmp_err_convert[code].fatal;
+			err = icmp_err_convert[code].errno;
+		}
+		break;
+	}
+
+	/*
+	 *      RFC1122: OK.  Passes ICMP errors back to application, as per
+	 *	4.1.3.3.
+	 */
+	if (!inet->recverr) {
+		if (!harderr || sk->sk_state != TCP_ESTABLISHED)
+			goto out;
+	} else {
+		ip_icmp_error(sk, skb, err, uh->dest, info, (u8*)(uh+1));
+	}
+	sk->sk_err = err;
+	sk->sk_error_report(sk);
+out:
+	sock_put(sk);
+}
+
+void udp_err(struct sk_buff *skb, u32 info)
+{
+	__udp4_lib_err(skb, info, udp_hash);
+}
+
+/*
+ * Throw away all pending data and cancel the corking. Socket is locked.
+ */
+static void udp_flush_pending_frames(struct sock *sk)
+{
+	struct udp_sock *up = udp_sk(sk);
+
+	if (up->pending) {
+		up->len = 0;
+		up->pending = 0;
+		ip_flush_pending_frames(sk);
+	}
+}
+
+/**
+ * 	udp4_hwcsum_outgoing  -  handle outgoing HW checksumming
+ * 	@sk: 	socket we are sending on
+ * 	@skb: 	sk_buff containing the filled-in UDP header
+ * 	        (checksum field must be zeroed out)
+ */
+static void udp4_hwcsum_outgoing(struct sock *sk, struct sk_buff *skb,
+				 __be32 src, __be32 dst, int len      )
+{
+	unsigned int offset;
+	struct udphdr *uh = udp_hdr(skb);
+	__wsum csum = 0;
+
+	if (skb_queue_len(&sk->sk_write_queue) == 1) {
+		/*
+		 * Only one fragment on the socket.
+		 */
+		skb->csum_start = skb_transport_header(skb) - skb->head;
+		skb->csum_offset = offsetof(struct udphdr, check);
+		uh->check = ~csum_tcpudp_magic(src, dst, len, IPPROTO_UDP, 0);
+	} else {
+		/*
+		 * HW-checksum won't work as there are two or more
+		 * fragments on the socket so that all csums of sk_buffs
+		 * should be together
+		 */
+		offset = skb_transport_offset(skb);
+		skb->csum = skb_checksum(skb, offset, skb->len - offset, 0);
+
+		skb->ip_summed = CHECKSUM_NONE;
+
+		skb_queue_walk(&sk->sk_write_queue, skb) {
+			csum = csum_add(csum, skb->csum);
+		}
+
+		uh->check = csum_tcpudp_magic(src, dst, len, IPPROTO_UDP, csum);
+		if (uh->check == 0)
+			uh->check = CSUM_MANGLED_0;
+	}
+}
+
+/*
+ * Push out all pending data as one UDP datagram. Socket is locked.
+ */
+static int udp_push_pending_frames(struct sock *sk)
+{
+	struct udp_sock  *up = udp_sk(sk);
+	struct inet_sock *inet = inet_sk(sk);
+	struct flowi *fl = &inet->cork.fl;
+	struct sk_buff *skb;
+	struct udphdr *uh;
+	int err = 0;
+	int is_udplite = IS_UDPLITE(sk);
+	__wsum csum = 0;
+
+	/* Grab the skbuff where UDP header space exists. */
+	if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
+		goto out;
+
+	/*
+	 * Create a UDP header
+	 */
+	uh = udp_hdr(skb);
+	uh->source = fl->fl_ip_sport;
+	uh->dest = fl->fl_ip_dport;
+	uh->len = htons(up->len);
+	uh->check = 0;
+
+	if (is_udplite)  				 /*     UDP-Lite      */
+		csum  = udplite_csum_outgoing(sk, skb);
+
+	else if (sk->sk_no_check == UDP_CSUM_NOXMIT) {   /* UDP csum disabled */
+
+		skb->ip_summed = CHECKSUM_NONE;
+		goto send;
+
+	} else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
+
+		udp4_hwcsum_outgoing(sk, skb, fl->fl4_src,fl->fl4_dst, up->len);
+		goto send;
+
+	} else						 /*   `normal' UDP    */
+		csum = udp_csum_outgoing(sk, skb);
+
+	/* add protocol-dependent pseudo-header */
+	uh->check = csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst, up->len,
+				      sk->sk_protocol, csum             );
+	if (uh->check == 0)
+		uh->check = CSUM_MANGLED_0;
+
+send:
+	err = ip_push_pending_frames(sk);
+out:
+	up->len = 0;
+	up->pending = 0;
+	if (!err)
+		UDP_INC_STATS_USER(UDP_MIB_OUTDATAGRAMS, is_udplite);
+	return err;
+}
+
+int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
+		size_t len)
+{
+	struct inet_sock *inet = inet_sk(sk);
+	struct udp_sock *up = udp_sk(sk);
+	int ulen = len;
+	struct ipcm_cookie ipc;
+	struct rtable *rt = NULL;
+	int free = 0;
+	int connected = 0;
+	__be32 daddr, faddr, saddr;
+	__be16 dport;
+	u8  tos;
+	int err, is_udplite = IS_UDPLITE(sk);
+	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
+	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
+
+	if (len > 0xFFFF)
+		return -EMSGSIZE;
+
+	/*
+	 *	Check the flags.
+	 */
+
+	if (msg->msg_flags&MSG_OOB)	/* Mirror BSD error message compatibility */
+		return -EOPNOTSUPP;
+
+	ipc.opt = NULL;
+
+	if (up->pending) {
+		/*
+		 * There are pending frames.
+		 * The socket lock must be held while it's corked.
+		 */
+		lock_sock(sk);
+		if (likely(up->pending)) {
+			if (unlikely(up->pending != AF_INET)) {
+				release_sock(sk);
+				return -EINVAL;
+			}
+			goto do_append_data;
+		}
+		release_sock(sk);
+	}
+	ulen += sizeof(struct udphdr);
+
+	/*
+	 *	Get and verify the address.
+	 */
+	if (msg->msg_name) {
+		struct sockaddr_in * usin = (struct sockaddr_in*)msg->msg_name;
+		if (msg->msg_namelen < sizeof(*usin))
+			return -EINVAL;
+		if (usin->sin_family != AF_INET) {
+			if (usin->sin_family != AF_UNSPEC)
+				return -EAFNOSUPPORT;
+		}
+
+		daddr = usin->sin_addr.s_addr;
+		dport = usin->sin_port;
+		if (dport == 0)
+			return -EINVAL;
+	} else {
+		if (sk->sk_state != TCP_ESTABLISHED)
+			return -EDESTADDRREQ;
+		daddr = inet->daddr;
+		dport = inet->dport;
+		/* Open fast path for connected socket.
+		   Route will not be used, if at least one option is set.
+		 */
+		connected = 1;
+	}
+	ipc.addr = inet->saddr;
+
+	ipc.oif = sk->sk_bound_dev_if;
+	if (msg->msg_controllen) {
+		err = ip_cmsg_send(msg, &ipc);
+		if (err)
+			return err;
+		if (ipc.opt)
+			free = 1;
+		connected = 0;
+	}
+	if (!ipc.opt)
+		ipc.opt = inet->opt;
+
+	saddr = ipc.addr;
+	ipc.addr = faddr = daddr;
+
+	if (ipc.opt && ipc.opt->srr) {
+		if (!daddr)
+			return -EINVAL;
+		faddr = ipc.opt->faddr;
+		connected = 0;
+	}
+	tos = RT_TOS(inet->tos);
+	if (sock_flag(sk, SOCK_LOCALROUTE) ||
+	    (msg->msg_flags & MSG_DONTROUTE) ||
+	    (ipc.opt && ipc.opt->is_strictroute)) {
+		tos |= RTO_ONLINK;
+		connected = 0;
+	}
+
+	if (ipv4_is_multicast(daddr)) {
+		if (!ipc.oif)
+			ipc.oif = inet->mc_index;
+		if (!saddr)
+			saddr = inet->mc_addr;
+		connected = 0;
+	}
+
+	if (connected)
+		rt = (struct rtable*)sk_dst_check(sk, 0);
+
+	if (rt == NULL) {
+		struct flowi fl = { .oif = ipc.oif,
+				    .nl_u = { .ip4_u =
+					      { .daddr = faddr,
+						.saddr = saddr,
+						.tos = tos } },
+				    .proto = sk->sk_protocol,
+				    .uli_u = { .ports =
+					       { .sport = inet->sport,
+						 .dport = dport } } };
+		security_sk_classify_flow(sk, &fl);
+		err = ip_route_output_flow(&init_net, &rt, &fl, sk, 1);
+		if (err) {
+			if (err == -ENETUNREACH)
+				IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
+			goto out;
+		}
+
+		err = -EACCES;
+		if ((rt->rt_flags & RTCF_BROADCAST) &&
+		    !sock_flag(sk, SOCK_BROADCAST))
+			goto out;
+		if (connected)
+			sk_dst_set(sk, dst_clone(&rt->u.dst));
+	}
+
+	if (msg->msg_flags&MSG_CONFIRM)
+		goto do_confirm;
+back_from_confirm:
+
+	saddr = rt->rt_src;
+	if (!ipc.addr)
+		daddr = ipc.addr = rt->rt_dst;
+
+	lock_sock(sk);
+	if (unlikely(up->pending)) {
+		/* The socket is already corked while preparing it. */
+		/* ... which is an evident application bug. --ANK */
+		release_sock(sk);
+
+		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
+		err = -EINVAL;
+		goto out;
+	}
+	/*
+	 *	Now cork the socket to pend data.
+	 */
+	inet->cork.fl.fl4_dst = daddr;
+	inet->cork.fl.fl_ip_dport = dport;
+	inet->cork.fl.fl4_src = saddr;
+	inet->cork.fl.fl_ip_sport = inet->sport;
+	up->pending = AF_INET;
+
+do_append_data:
+	up->len += ulen;
+	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
+	err = ip_append_data(sk, getfrag, msg->msg_iov, ulen,
+			sizeof(struct udphdr), &ipc, rt,
+			corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags);
+	if (err)
+		udp_flush_pending_frames(sk);
+	else if (!corkreq)
+		err = udp_push_pending_frames(sk);
+	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
+		up->pending = 0;
+	release_sock(sk);
+
+out:
+	ip_rt_put(rt);
+	if (free)
+		kfree(ipc.opt);
+	if (!err)
+		return len;
+	/*
+	 * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space.  Reporting
+	 * ENOBUFS might not be good (it's not tunable per se), but otherwise
+	 * we don't have a good statistic (IpOutDiscards but it can be too many
+	 * things).  We could add another new stat but at least for now that
+	 * seems like overkill.
+	 */
+	if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
+		UDP_INC_STATS_USER(UDP_MIB_SNDBUFERRORS, is_udplite);
+	}
+	return err;
+
+do_confirm:
+	dst_confirm(&rt->u.dst);
+	if (!(msg->msg_flags&MSG_PROBE) || len)
+		goto back_from_confirm;
+	err = 0;
+	goto out;
+}
+
+int udp_sendpage(struct sock *sk, struct page *page, int offset,
+		 size_t size, int flags)
+{
+	struct udp_sock *up = udp_sk(sk);
+	int ret;
+
+	if (!up->pending) {
+		struct msghdr msg = {	.msg_flags = flags|MSG_MORE };
+
+		/* Call udp_sendmsg to specify destination address which
+		 * sendpage interface can't pass.
+		 * This will succeed only when the socket is connected.
+		 */
+		ret = udp_sendmsg(NULL, sk, &msg, 0);
+		if (ret < 0)
+			return ret;
+	}
+
+	lock_sock(sk);
+
+	if (unlikely(!up->pending)) {
+		release_sock(sk);
+
+		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 3\n");
+		return -EINVAL;
+	}
+
+	ret = ip_append_page(sk, page, offset, size, flags);
+	if (ret == -EOPNOTSUPP) {
+		release_sock(sk);
+		return sock_no_sendpage(sk->sk_socket, page, offset,
+					size, flags);
+	}
+	if (ret < 0) {
+		udp_flush_pending_frames(sk);
+		goto out;
+	}
+
+	up->len += size;
+	if (!(up->corkflag || (flags&MSG_MORE)))
+		ret = udp_push_pending_frames(sk);
+	if (!ret)
+		ret = size;
+out:
+	release_sock(sk);
+	return ret;
+}
+
 /*
  *	IOCTL requests applicable to the UDP protocol
  */
@@ -286,6 +833,107 @@ int udp_ioctl(struct sock *sk, int cmd, unsigned long arg)
 	return 0;
 }
 
+/*
+ * 	This should be easy, if there is something there we
+ * 	return it, otherwise we block.
+ */
+
+int udp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
+		size_t len, int noblock, int flags, int *addr_len)
+{
+	struct inet_sock *inet = inet_sk(sk);
+	struct sockaddr_in *sin = (struct sockaddr_in *)msg->msg_name;
+	struct sk_buff *skb;
+	unsigned int ulen, copied;
+	int peeked;
+	int err;
+	int is_udplite = IS_UDPLITE(sk);
+
+	/*
+	 *	Check any passed addresses
+	 */
+	if (addr_len)
+		*addr_len=sizeof(*sin);
+
+	if (flags & MSG_ERRQUEUE)
+		return ip_recv_error(sk, msg, len);
+
+try_again:
+	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
+				  &peeked, &err);
+	if (!skb)
+		goto out;
+
+	ulen = skb->len - sizeof(struct udphdr);
+	copied = len;
+	if (copied > ulen)
+		copied = ulen;
+	else if (copied < ulen)
+		msg->msg_flags |= MSG_TRUNC;
+
+	/*
+	 * If checksum is needed at all, try to do it while copying the
+	 * data.  If the data is truncated, or if we only want a partial
+	 * coverage checksum (UDP-Lite), do it before the copy.
+	 */
+
+	if (copied < ulen || UDP_SKB_CB(skb)->partial_cov) {
+		if (udp_lib_checksum_complete(skb))
+			goto csum_copy_err;
+	}
+
+	if (skb_csum_unnecessary(skb))
+		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
+					      msg->msg_iov, copied       );
+	else {
+		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
+
+		if (err == -EINVAL)
+			goto csum_copy_err;
+	}
+
+	if (err)
+		goto out_free;
+
+	if (!peeked)
+		UDP_INC_STATS_USER(UDP_MIB_INDATAGRAMS, is_udplite);
+
+	sock_recv_timestamp(msg, sk, skb);
+
+	/* Copy the address. */
+	if (sin)
+	{
+		sin->sin_family = AF_INET;
+		sin->sin_port = udp_hdr(skb)->source;
+		sin->sin_addr.s_addr = ip_hdr(skb)->saddr;
+		memset(sin->sin_zero, 0, sizeof(sin->sin_zero));
+	}
+	if (inet->cmsg_flags)
+		ip_cmsg_recv(msg, skb);
+
+	err = copied;
+	if (flags & MSG_TRUNC)
+		err = ulen;
+
+out_free:
+	lock_sock(sk);
+	skb_free_datagram(sk, skb);
+	release_sock(sk);
+out:
+	return err;
+
+csum_copy_err:
+	lock_sock(sk);
+	if (!skb_kill_datagram(sk, skb, flags))
+		UDP_INC_STATS_USER(UDP_MIB_INERRORS, is_udplite);
+	release_sock(sk);
+
+	if (noblock)
+		return -EAGAIN;
+	goto try_again;
+}
+
+
 int udp_disconnect(struct sock *sk, int flags)
 {
 	struct inet_sock *inet = inet_sk(sk);
@@ -308,6 +956,319 @@ int udp_disconnect(struct sock *sk, int flags)
 	return 0;
 }
 
+/* returns:
+ *  -1: error
+ *   0: success
+ *  >0: "udp encap" protocol resubmission
+ *
+ * Note that in the success and error cases, the skb is assumed to
+ * have either been requeued or freed.
+ */
+int udp_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
+{
+	struct udp_sock *up = udp_sk(sk);
+	int rc;
+	int is_udplite = IS_UDPLITE(sk);
+
+	/*
+	 *	Charge it to the socket, dropping if the queue is full.
+	 */
+	if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb))
+		goto drop;
+	nf_reset(skb);
+
+	if (up->encap_type) {
+		/*
+		 * This is an encapsulation socket so pass the skb to
+		 * the socket's udp_encap_rcv() hook. Otherwise, just
+		 * fall through and pass this up the UDP socket.
+		 * up->encap_rcv() returns the following value:
+		 * =0 if skb was successfully passed to the encap
+		 *    handler or was discarded by it.
+		 * >0 if skb should be passed on to UDP.
+		 * <0 if skb should be resubmitted as proto -N
+		 */
+
+		/* if we're overly short, let UDP handle it */
+		if (skb->len > sizeof(struct udphdr) &&
+		    up->encap_rcv != NULL) {
+			int ret;
+
+			ret = (*up->encap_rcv)(sk, skb);
+			if (ret <= 0) {
+				UDP_INC_STATS_BH(UDP_MIB_INDATAGRAMS,
+						 is_udplite);
+				return -ret;
+			}
+		}
+
+		/* FALLTHROUGH -- it's a UDP Packet */
+	}
+
+	/*
+	 * 	UDP-Lite specific tests, ignored on UDP sockets
+	 */
+	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
+
+		/*
+		 * MIB statistics other than incrementing the error count are
+		 * disabled for the following two types of errors: these depend
+		 * on the application settings, not on the functioning of the
+		 * protocol stack as such.
+		 *
+		 * RFC 3828 here recommends (sec 3.3): "There should also be a
+		 * way ... to ... at least let the receiving application block
+		 * delivery of packets with coverage values less than a value
+		 * provided by the application."
+		 */
+		if (up->pcrlen == 0) {          /* full coverage was set  */
+			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE: partial coverage "
+				"%d while full coverage %d requested\n",
+				UDP_SKB_CB(skb)->cscov, skb->len);
+			goto drop;
+		}
+		/* The next case involves violating the min. coverage requested
+		 * by the receiver. This is subtle: if receiver wants x and x is
+		 * greater than the buffersize/MTU then receiver will complain
+		 * that it wants x while sender emits packets of smaller size y.
+		 * Therefore the above ...()->partial_cov statement is essential.
+		 */
+		if (UDP_SKB_CB(skb)->cscov  <  up->pcrlen) {
+			LIMIT_NETDEBUG(KERN_WARNING
+				"UDPLITE: coverage %d too small, need min %d\n",
+				UDP_SKB_CB(skb)->cscov, up->pcrlen);
+			goto drop;
+		}
+	}
+
+	if (sk->sk_filter) {
+		if (udp_lib_checksum_complete(skb))
+			goto drop;
+	}
+
+	if ((rc = sock_queue_rcv_skb(sk,skb)) < 0) {
+		/* Note that an ENOMEM error is charged twice */
+		if (rc == -ENOMEM)
+			UDP_INC_STATS_BH(UDP_MIB_RCVBUFERRORS, is_udplite);
+		goto drop;
+	}
+
+	return 0;
+
+drop:
+	UDP_INC_STATS_BH(UDP_MIB_INERRORS, is_udplite);
+	kfree_skb(skb);
+	return -1;
+}
+
+/*
+ *	Multicasts and broadcasts go to each listener.
+ *
+ *	Note: called only from the BH handler context,
+ *	so we don't need to lock the hashes.
+ */
+static int __udp4_lib_mcast_deliver(struct sk_buff *skb,
+				    struct udphdr  *uh,
+				    __be32 saddr, __be32 daddr,
+				    struct hlist_head udptable[])
+{
+	struct sock *sk;
+	int dif;
+
+	read_lock(&udp_hash_lock);
+	sk = sk_head(&udptable[ntohs(uh->dest) & (UDP_HTABLE_SIZE - 1)]);
+	dif = skb->dev->ifindex;
+	sk = udp_v4_mcast_next(sk, uh->dest, daddr, uh->source, saddr, dif);
+	if (sk) {
+		struct sock *sknext = NULL;
+
+		do {
+			struct sk_buff *skb1 = skb;
+
+			sknext = udp_v4_mcast_next(sk_next(sk), uh->dest, daddr,
+						   uh->source, saddr, dif);
+			if (sknext)
+				skb1 = skb_clone(skb, GFP_ATOMIC);
+
+			if (skb1) {
+				int ret = 0;
+
+				bh_lock_sock_nested(sk);
+				if (!sock_owned_by_user(sk))
+					ret = udp_queue_rcv_skb(sk, skb1);
+				else
+					sk_add_backlog(sk, skb1);
+				bh_unlock_sock(sk);
+
+				if (ret > 0)
+					/* we should probably re-process instead
+					 * of dropping packets here. */
+					kfree_skb(skb1);
+			}
+			sk = sknext;
+		} while (sknext);
+	} else
+		kfree_skb(skb);
+	read_unlock(&udp_hash_lock);
+	return 0;
+}
+
+/* Initialize UDP checksum. If exited with zero value (success),
+ * CHECKSUM_UNNECESSARY means, that no more checks are required.
+ * Otherwise, csum completion requires chacksumming packet body,
+ * including udp header and folding it to skb->csum.
+ */
+static inline int udp4_csum_init(struct sk_buff *skb, struct udphdr *uh,
+				 int proto)
+{
+	const struct iphdr *iph;
+	int err;
+
+	UDP_SKB_CB(skb)->partial_cov = 0;
+	UDP_SKB_CB(skb)->cscov = skb->len;
+
+	if (proto == IPPROTO_UDPLITE) {
+		err = udplite_checksum_init(skb, uh);
+		if (err)
+			return err;
+	}
+
+	iph = ip_hdr(skb);
+	if (uh->check == 0) {
+		skb->ip_summed = CHECKSUM_UNNECESSARY;
+	} else if (skb->ip_summed == CHECKSUM_COMPLETE) {
+	       if (!csum_tcpudp_magic(iph->saddr, iph->daddr, skb->len,
+				      proto, skb->csum))
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+	}
+	if (!skb_csum_unnecessary(skb))
+		skb->csum = csum_tcpudp_nofold(iph->saddr, iph->daddr,
+					       skb->len, proto, 0);
+	/* Probably, we should checksum udp header (it should be in cache
+	 * in any case) and data in tiny packets (< rx copybreak).
+	 */
+
+	return 0;
+}
+
+/*
+ *	All we need to do is get the socket, and then do a checksum.
+ */
+
+int __udp4_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
+		   int proto)
+{
+	struct sock *sk;
+	struct udphdr *uh = udp_hdr(skb);
+	unsigned short ulen;
+	struct rtable *rt = (struct rtable*)skb->dst;
+	__be32 saddr = ip_hdr(skb)->saddr;
+	__be32 daddr = ip_hdr(skb)->daddr;
+
+	/*
+	 *  Validate the packet.
+	 */
+	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
+		goto drop;		/* No space for header. */
+
+	ulen = ntohs(uh->len);
+	if (ulen > skb->len)
+		goto short_packet;
+
+	if (proto == IPPROTO_UDP) {
+		/* UDP validates ulen. */
+		if (ulen < sizeof(*uh) || pskb_trim_rcsum(skb, ulen))
+			goto short_packet;
+		uh = udp_hdr(skb);
+	}
+
+	if (udp4_csum_init(skb, uh, proto))
+		goto csum_error;
+
+	if (rt->rt_flags & (RTCF_BROADCAST|RTCF_MULTICAST))
+		return __udp4_lib_mcast_deliver(skb, uh, saddr, daddr, udptable);
+
+	sk = __udp4_lib_lookup(skb->dev->nd_net, saddr, uh->source, daddr,
+			uh->dest, inet_iif(skb), udptable);
+
+	if (sk != NULL) {
+		int ret = 0;
+		bh_lock_sock_nested(sk);
+		if (!sock_owned_by_user(sk))
+			ret = udp_queue_rcv_skb(sk, skb);
+		else
+			sk_add_backlog(sk, skb);
+		bh_unlock_sock(sk);
+		sock_put(sk);
+
+		/* a return value > 0 means to resubmit the input, but
+		 * it wants the return to be -protocol, or 0
+		 */
+		if (ret > 0)
+			return -ret;
+		return 0;
+	}
+
+	if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb))
+		goto drop;
+	nf_reset(skb);
+
+	/* No socket. Drop packet silently, if checksum is wrong */
+	if (udp_lib_checksum_complete(skb))
+		goto csum_error;
+
+	UDP_INC_STATS_BH(UDP_MIB_NOPORTS, proto == IPPROTO_UDPLITE);
+	icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);
+
+	/*
+	 * Hmm.  We got an UDP packet to a port to which we
+	 * don't wanna listen.  Ignore it.
+	 */
+	kfree_skb(skb);
+	return 0;
+
+short_packet:
+	LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: short packet: From %u.%u.%u.%u:%u %d/%d to %u.%u.%u.%u:%u\n",
+		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
+		       NIPQUAD(saddr),
+		       ntohs(uh->source),
+		       ulen,
+		       skb->len,
+		       NIPQUAD(daddr),
+		       ntohs(uh->dest));
+	goto drop;
+
+csum_error:
+	/*
+	 * RFC1122: OK.  Discards the bad packet silently (as far as
+	 * the network is concerned, anyway) as per 4.1.3.4 (MUST).
+	 */
+	LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: bad checksum. From %d.%d.%d.%d:%d to %d.%d.%d.%d:%d ulen %d\n",
+		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
+		       NIPQUAD(saddr),
+		       ntohs(uh->source),
+		       NIPQUAD(daddr),
+		       ntohs(uh->dest),
+		       ulen);
+drop:
+	UDP_INC_STATS_BH(UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
+	kfree_skb(skb);
+	return 0;
+}
+
+int udp_rcv(struct sk_buff *skb)
+{
+	return __udp4_lib_rcv(skb, udp_hash, IPPROTO_UDP);
+}
+
+int udp_destroy_sock(struct sock *sk)
+{
+	lock_sock(sk);
+	udp_flush_pending_frames(sk);
+	release_sock(sk);
+	return 0;
+}
+
 /*
  *	Socket option code for UDP
  */
@@ -318,9 +1279,7 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 	struct udp_sock *up = udp_sk(sk);
 	int val;
 	int err = 0;
-#ifdef CONFIG_IP_UDPLITE
 	int is_udplite = IS_UDPLITE(sk);
-#endif
 
 	if (optlen<sizeof(int))
 		return -EINVAL;
@@ -356,7 +1315,6 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 		}
 		break;
 
-#ifdef CONFIG_IP_UDPLITE
 	/*
 	 * 	UDP-Lite's partial checksum coverage (RFC 3828).
 	 */
@@ -382,7 +1340,6 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 		up->pcrlen = val;
 		up->pcflag |= UDPLITE_RECV_CC;
 		break;
-#endif
 
 	default:
 		err = -ENOPROTOOPT;
@@ -392,6 +1349,26 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 	return err;
 }
 
+int udp_setsockopt(struct sock *sk, int level, int optname,
+		   char __user *optval, int optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
+					  udp_push_pending_frames);
+	return ip_setsockopt(sk, level, optname, optval, optlen);
+}
+
+#ifdef CONFIG_COMPAT
+int compat_udp_setsockopt(struct sock *sk, int level, int optname,
+			  char __user *optval, int optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
+					  udp_push_pending_frames);
+	return compat_ip_setsockopt(sk, level, optname, optval, optlen);
+}
+#endif
+
 int udp_lib_getsockopt(struct sock *sk, int level, int optname,
 		       char __user *optval, int __user *optlen)
 {
@@ -436,6 +1413,23 @@ int udp_lib_getsockopt(struct sock *sk, int level, int optname,
 	return 0;
 }
 
+int udp_getsockopt(struct sock *sk, int level, int optname,
+		   char __user *optval, int __user *optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
+	return ip_getsockopt(sk, level, optname, optval, optlen);
+}
+
+#ifdef CONFIG_COMPAT
+int compat_udp_getsockopt(struct sock *sk, int level, int optname,
+				 char __user *optval, int __user *optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
+	return compat_ip_getsockopt(sk, level, optname, optval, optlen);
+}
+#endif
 /**
  * 	udp_poll - wait for a UDP event.
  *	@file - file struct
@@ -480,6 +1474,36 @@ unsigned int udp_poll(struct file *file, struct socket *sock, poll_table *wait)
 
 }
 
+DEFINE_PROTO_INUSE(udp)
+
+struct proto udp_prot = {
+	.name		   = "UDP",
+	.owner		   = THIS_MODULE,
+	.close		   = udp_lib_close,
+	.connect	   = ip4_datagram_connect,
+	.disconnect	   = udp_disconnect,
+	.ioctl		   = udp_ioctl,
+	.destroy	   = udp_destroy_sock,
+	.setsockopt	   = udp_setsockopt,
+	.getsockopt	   = udp_getsockopt,
+	.sendmsg	   = udp_sendmsg,
+	.recvmsg	   = udp_recvmsg,
+	.sendpage	   = udp_sendpage,
+	.backlog_rcv	   = udp_queue_rcv_skb,
+	.hash		   = udp_lib_hash,
+	.unhash		   = udp_lib_unhash,
+	.get_port	   = udp_v4_get_port,
+	.memory_allocated  = &udp_memory_allocated,
+	.sysctl_mem	   = sysctl_udp_mem,
+	.sysctl_wmem	   = &sysctl_udp_wmem_min,
+	.sysctl_rmem	   = &sysctl_udp_rmem_min,
+	.obj_size	   = sizeof(struct udp_sock),
+#ifdef CONFIG_COMPAT
+	.compat_setsockopt = compat_udp_setsockopt,
+	.compat_getsockopt = compat_udp_getsockopt,
+#endif
+	REF_PROTO_INUSE(udp)
+};
 
 /* ------------------------------------------------------------------------ */
 #ifdef CONFIG_PROC_FS
@@ -612,6 +1636,62 @@ void udp_proc_unregister(struct udp_seq_afinfo *afinfo)
 	proc_net_remove(&init_net, afinfo->name);
 	memset(afinfo->seq_fops, 0, sizeof(*afinfo->seq_fops));
 }
+
+/* ------------------------------------------------------------------------ */
+static void udp4_format_sock(struct sock *sp, char *tmpbuf, int bucket)
+{
+	struct inet_sock *inet = inet_sk(sp);
+	__be32 dest = inet->daddr;
+	__be32 src  = inet->rcv_saddr;
+	__u16 destp	  = ntohs(inet->dport);
+	__u16 srcp	  = ntohs(inet->sport);
+
+	sprintf(tmpbuf, "%4d: %08X:%04X %08X:%04X"
+		" %02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p",
+		bucket, src, srcp, dest, destp, sp->sk_state,
+		atomic_read(&sp->sk_wmem_alloc),
+		atomic_read(&sp->sk_rmem_alloc),
+		0, 0L, 0, sock_i_uid(sp), 0, sock_i_ino(sp),
+		atomic_read(&sp->sk_refcnt), sp);
+}
+
+int udp4_seq_show(struct seq_file *seq, void *v)
+{
+	if (v == SEQ_START_TOKEN)
+		seq_printf(seq, "%-127s\n",
+			   "  sl  local_address rem_address   st tx_queue "
+			   "rx_queue tr tm->when retrnsmt   uid  timeout "
+			   "inode");
+	else {
+		char tmpbuf[129];
+		struct udp_iter_state *state = seq->private;
+
+		udp4_format_sock(v, tmpbuf, state->bucket);
+		seq_printf(seq, "%-127s\n", tmpbuf);
+	}
+	return 0;
+}
+
+/* ------------------------------------------------------------------------ */
+static struct file_operations udp4_seq_fops;
+static struct udp_seq_afinfo udp4_seq_afinfo = {
+	.owner		= THIS_MODULE,
+	.name		= "udp",
+	.family		= AF_INET,
+	.hashtable	= udp_hash,
+	.seq_show	= udp4_seq_show,
+	.seq_fops	= &udp4_seq_fops,
+};
+
+int __init udp4_proc_init(void)
+{
+	return udp_proc_register(&udp4_seq_afinfo);
+}
+
+void udp4_proc_exit(void)
+{
+	udp_proc_unregister(&udp4_seq_afinfo);
+}
 #endif /* CONFIG_PROC_FS */
 
 void __init udp_init(void)
@@ -638,6 +1718,8 @@ EXPORT_SYMBOL(udp_hash);
 EXPORT_SYMBOL(udp_hash_lock);
 EXPORT_SYMBOL(udp_ioctl);
 EXPORT_SYMBOL(udp_get_port);
+EXPORT_SYMBOL(udp_prot);
+EXPORT_SYMBOL(udp_sendmsg);
 EXPORT_SYMBOL(udp_lib_getsockopt);
 EXPORT_SYMBOL(udp_lib_setsockopt);
 EXPORT_SYMBOL(udp_poll);
diff --git a/net/ipv4/udp_ipv4.c b/net/ipv4/udp_ipv4.c
deleted file mode 100644
index fd14c2c50ed4..000000000000
--- a/net/ipv4/udp_ipv4.c
+++ /dev/null
@@ -1,1134 +0,0 @@
-/*
- * INET		An implementation of the TCP/IP protocol suite for the LINUX
- *		operating system.  INET is implemented using the  BSD Socket
- *		interface as the means of communication with the user level.
- *
- *		UDP for IPv4.
- *
- *		For full credits, see net/ipv4/udp.c.
- *
- *		This program is free software; you can redistribute it and/or
- *		modify it under the terms of the GNU General Public License
- *		as published by the Free Software Foundation; either version
- *		2 of the License, or (at your option) any later version.
- */
-
-#include <asm/system.h>
-#include <asm/uaccess.h>
-#include <asm/ioctls.h>
-#include <linux/bootmem.h>
-#include <linux/types.h>
-#include <linux/fcntl.h>
-#include <linux/module.h>
-#include <linux/socket.h>
-#include <linux/sockios.h>
-#include <linux/igmp.h>
-#include <linux/in.h>
-#include <linux/errno.h>
-#include <linux/timer.h>
-#include <linux/mm.h>
-#include <linux/inet.h>
-#include <linux/netdevice.h>
-#include <net/tcp_states.h>
-#include <linux/skbuff.h>
-#include <linux/proc_fs.h>
-#include <linux/seq_file.h>
-#include <net/net_namespace.h>
-#include <net/icmp.h>
-#include <net/route.h>
-#include <net/checksum.h>
-#include <net/xfrm.h>
-#include "udp_impl.h"
-
-int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2)
-{
-	struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2);
-
-	return 	( !ipv6_only_sock(sk2)  &&
-		  (!inet1->rcv_saddr || !inet2->rcv_saddr ||
-		   inet1->rcv_saddr == inet2->rcv_saddr      ));
-}
-
-static inline int udp_v4_get_port(struct sock *sk, unsigned short snum)
-{
-	return udp_get_port(sk, snum, ipv4_rcv_saddr_equal);
-}
-
-/* UDP is nearly always wildcards out the wazoo, it makes no sense to try
- * harder than this. -DaveM
- */
-static struct sock *__udp4_lib_lookup(struct net *net, __be32 saddr,
-		__be16 sport, __be32 daddr, __be16 dport,
-		int dif, struct hlist_head udptable[])
-{
-	struct sock *sk, *result = NULL;
-	struct hlist_node *node;
-	unsigned short hnum = ntohs(dport);
-	int badness = -1;
-
-	read_lock(&udp_hash_lock);
-	sk_for_each(sk, node, &udptable[hnum & (UDP_HTABLE_SIZE - 1)]) {
-		struct inet_sock *inet = inet_sk(sk);
-
-		if (sk->sk_net == net && sk->sk_hash == hnum &&
-				!ipv6_only_sock(sk)) {
-			int score = (sk->sk_family == PF_INET ? 1 : 0);
-			if (inet->rcv_saddr) {
-				if (inet->rcv_saddr != daddr)
-					continue;
-				score+=2;
-			}
-			if (inet->daddr) {
-				if (inet->daddr != saddr)
-					continue;
-				score+=2;
-			}
-			if (inet->dport) {
-				if (inet->dport != sport)
-					continue;
-				score+=2;
-			}
-			if (sk->sk_bound_dev_if) {
-				if (sk->sk_bound_dev_if != dif)
-					continue;
-				score+=2;
-			}
-			if (score == 9) {
-				result = sk;
-				break;
-			} else if (score > badness) {
-				result = sk;
-				badness = score;
-			}
-		}
-	}
-	if (result)
-		sock_hold(result);
-	read_unlock(&udp_hash_lock);
-	return result;
-}
-
-static inline struct sock *udp_v4_mcast_next(struct sock *sk,
-					     __be16 loc_port, __be32 loc_addr,
-					     __be16 rmt_port, __be32 rmt_addr,
-					     int dif)
-{
-	struct hlist_node *node;
-	struct sock *s = sk;
-	unsigned short hnum = ntohs(loc_port);
-
-	sk_for_each_from(s, node) {
-		struct inet_sock *inet = inet_sk(s);
-
-		if (s->sk_hash != hnum					||
-		    (inet->daddr && inet->daddr != rmt_addr)		||
-		    (inet->dport != rmt_port && inet->dport)		||
-		    (inet->rcv_saddr && inet->rcv_saddr != loc_addr)	||
-		    ipv6_only_sock(s)					||
-		    (s->sk_bound_dev_if && s->sk_bound_dev_if != dif))
-			continue;
-		if (!ip_mc_sf_allow(s, loc_addr, rmt_addr, dif))
-			continue;
-		goto found;
-	}
-	s = NULL;
-found:
-	return s;
-}
-
-/*
- * This routine is called by the ICMP module when it gets some
- * sort of error condition.  If err < 0 then the socket should
- * be closed and the error returned to the user.  If err > 0
- * it's just the icmp type << 8 | icmp code.
- * Header points to the ip header of the error packet. We move
- * on past this. Then (as it used to claim before adjustment)
- * header points to the first 8 bytes of the udp header.  We need
- * to find the appropriate port.
- */
-
-void __udp4_lib_err(struct sk_buff *skb, u32 info, struct hlist_head udptable[])
-{
-	struct inet_sock *inet;
-	struct iphdr *iph = (struct iphdr*)skb->data;
-	struct udphdr *uh = (struct udphdr*)(skb->data+(iph->ihl<<2));
-	const int type = icmp_hdr(skb)->type;
-	const int code = icmp_hdr(skb)->code;
-	struct sock *sk;
-	int harderr;
-	int err;
-
-	sk = __udp4_lib_lookup(skb->dev->nd_net, iph->daddr, uh->dest,
-			iph->saddr, uh->source, skb->dev->ifindex, udptable);
-	if (sk == NULL) {
-		ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
-		return;	/* No socket for error */
-	}
-
-	err = 0;
-	harderr = 0;
-	inet = inet_sk(sk);
-
-	switch (type) {
-	default:
-	case ICMP_TIME_EXCEEDED:
-		err = EHOSTUNREACH;
-		break;
-	case ICMP_SOURCE_QUENCH:
-		goto out;
-	case ICMP_PARAMETERPROB:
-		err = EPROTO;
-		harderr = 1;
-		break;
-	case ICMP_DEST_UNREACH:
-		if (code == ICMP_FRAG_NEEDED) { /* Path MTU discovery */
-			if (inet->pmtudisc != IP_PMTUDISC_DONT) {
-				err = EMSGSIZE;
-				harderr = 1;
-				break;
-			}
-			goto out;
-		}
-		err = EHOSTUNREACH;
-		if (code <= NR_ICMP_UNREACH) {
-			harderr = icmp_err_convert[code].fatal;
-			err = icmp_err_convert[code].errno;
-		}
-		break;
-	}
-
-	/*
-	 *      RFC1122: OK.  Passes ICMP errors back to application, as per
-	 *	4.1.3.3.
-	 */
-	if (!inet->recverr) {
-		if (!harderr || sk->sk_state != TCP_ESTABLISHED)
-			goto out;
-	} else {
-		ip_icmp_error(sk, skb, err, uh->dest, info, (u8*)(uh+1));
-	}
-	sk->sk_err = err;
-	sk->sk_error_report(sk);
-out:
-	sock_put(sk);
-}
-
-void udp_err(struct sk_buff *skb, u32 info)
-{
-	__udp4_lib_err(skb, info, udp_hash);
-}
-
-/*
- * Throw away all pending data and cancel the corking. Socket is locked.
- */
-static void udp_flush_pending_frames(struct sock *sk)
-{
-	struct udp_sock *up = udp_sk(sk);
-
-	if (up->pending) {
-		up->len = 0;
-		up->pending = 0;
-		ip_flush_pending_frames(sk);
-	}
-}
-
-/**
- * 	udp4_hwcsum_outgoing  -  handle outgoing HW checksumming
- * 	@sk: 	socket we are sending on
- * 	@skb: 	sk_buff containing the filled-in UDP header
- * 	        (checksum field must be zeroed out)
- */
-static void udp4_hwcsum_outgoing(struct sock *sk, struct sk_buff *skb,
-				 __be32 src, __be32 dst, int len      )
-{
-	unsigned int offset;
-	struct udphdr *uh = udp_hdr(skb);
-	__wsum csum = 0;
-
-	if (skb_queue_len(&sk->sk_write_queue) == 1) {
-		/*
-		 * Only one fragment on the socket.
-		 */
-		skb->csum_start = skb_transport_header(skb) - skb->head;
-		skb->csum_offset = offsetof(struct udphdr, check);
-		uh->check = ~csum_tcpudp_magic(src, dst, len, IPPROTO_UDP, 0);
-	} else {
-		/*
-		 * HW-checksum won't work as there are two or more
-		 * fragments on the socket so that all csums of sk_buffs
-		 * should be together
-		 */
-		offset = skb_transport_offset(skb);
-		skb->csum = skb_checksum(skb, offset, skb->len - offset, 0);
-
-		skb->ip_summed = CHECKSUM_NONE;
-
-		skb_queue_walk(&sk->sk_write_queue, skb) {
-			csum = csum_add(csum, skb->csum);
-		}
-
-		uh->check = csum_tcpudp_magic(src, dst, len, IPPROTO_UDP, csum);
-		if (uh->check == 0)
-			uh->check = CSUM_MANGLED_0;
-	}
-}
-
-/*
- * Push out all pending data as one UDP datagram. Socket is locked.
- */
-static int udp_push_pending_frames(struct sock *sk)
-{
-	struct udp_sock  *up = udp_sk(sk);
-	struct inet_sock *inet = inet_sk(sk);
-	struct flowi *fl = &inet->cork.fl;
-	struct sk_buff *skb;
-	struct udphdr *uh;
-	int err = 0;
-	int is_udplite = IS_UDPLITE(sk);
-	__wsum csum = 0;
-
-	/* Grab the skbuff where UDP header space exists. */
-	if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
-		goto out;
-
-	/*
-	 * Create a UDP header
-	 */
-	uh = udp_hdr(skb);
-	uh->source = fl->fl_ip_sport;
-	uh->dest = fl->fl_ip_dport;
-	uh->len = htons(up->len);
-	uh->check = 0;
-
-	if (is_udplite)  				 /*     UDP-Lite      */
-		csum  = udplite_csum_outgoing(sk, skb);
-
-	else if (sk->sk_no_check == UDP_CSUM_NOXMIT) {   /* UDP csum disabled */
-
-		skb->ip_summed = CHECKSUM_NONE;
-		goto send;
-
-	} else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
-
-		udp4_hwcsum_outgoing(sk, skb, fl->fl4_src,fl->fl4_dst, up->len);
-		goto send;
-
-	} else						 /*   `normal' UDP    */
-		csum = udp_csum_outgoing(sk, skb);
-
-	/* add protocol-dependent pseudo-header */
-	uh->check = csum_tcpudp_magic(fl->fl4_src, fl->fl4_dst, up->len,
-				      sk->sk_protocol, csum             );
-	if (uh->check == 0)
-		uh->check = CSUM_MANGLED_0;
-
-send:
-	err = ip_push_pending_frames(sk);
-out:
-	up->len = 0;
-	up->pending = 0;
-	if (!err)
-		UDP_INC_STATS_USER(UDP_MIB_OUTDATAGRAMS, is_udplite);
-	return err;
-}
-
-int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
-		size_t len)
-{
-	struct inet_sock *inet = inet_sk(sk);
-	struct udp_sock *up = udp_sk(sk);
-	int ulen = len;
-	struct ipcm_cookie ipc;
-	struct rtable *rt = NULL;
-	int free = 0;
-	int connected = 0;
-	__be32 daddr, faddr, saddr;
-	__be16 dport;
-	u8  tos;
-	int err, is_udplite = IS_UDPLITE(sk);
-	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
-	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
-
-	if (len > 0xFFFF)
-		return -EMSGSIZE;
-
-	/*
-	 *	Check the flags.
-	 */
-
-	if (msg->msg_flags&MSG_OOB)	/* Mirror BSD error message compatibility */
-		return -EOPNOTSUPP;
-
-	ipc.opt = NULL;
-
-	if (up->pending) {
-		/*
-		 * There are pending frames.
-		 * The socket lock must be held while it's corked.
-		 */
-		lock_sock(sk);
-		if (likely(up->pending)) {
-			if (unlikely(up->pending != AF_INET)) {
-				release_sock(sk);
-				return -EINVAL;
-			}
-			goto do_append_data;
-		}
-		release_sock(sk);
-	}
-	ulen += sizeof(struct udphdr);
-
-	/*
-	 *	Get and verify the address.
-	 */
-	if (msg->msg_name) {
-		struct sockaddr_in * usin = (struct sockaddr_in*)msg->msg_name;
-		if (msg->msg_namelen < sizeof(*usin))
-			return -EINVAL;
-		if (usin->sin_family != AF_INET) {
-			if (usin->sin_family != AF_UNSPEC)
-				return -EAFNOSUPPORT;
-		}
-
-		daddr = usin->sin_addr.s_addr;
-		dport = usin->sin_port;
-		if (dport == 0)
-			return -EINVAL;
-	} else {
-		if (sk->sk_state != TCP_ESTABLISHED)
-			return -EDESTADDRREQ;
-		daddr = inet->daddr;
-		dport = inet->dport;
-		/* Open fast path for connected socket.
-		   Route will not be used, if at least one option is set.
-		 */
-		connected = 1;
-	}
-	ipc.addr = inet->saddr;
-
-	ipc.oif = sk->sk_bound_dev_if;
-	if (msg->msg_controllen) {
-		err = ip_cmsg_send(msg, &ipc);
-		if (err)
-			return err;
-		if (ipc.opt)
-			free = 1;
-		connected = 0;
-	}
-	if (!ipc.opt)
-		ipc.opt = inet->opt;
-
-	saddr = ipc.addr;
-	ipc.addr = faddr = daddr;
-
-	if (ipc.opt && ipc.opt->srr) {
-		if (!daddr)
-			return -EINVAL;
-		faddr = ipc.opt->faddr;
-		connected = 0;
-	}
-	tos = RT_TOS(inet->tos);
-	if (sock_flag(sk, SOCK_LOCALROUTE) ||
-	    (msg->msg_flags & MSG_DONTROUTE) ||
-	    (ipc.opt && ipc.opt->is_strictroute)) {
-		tos |= RTO_ONLINK;
-		connected = 0;
-	}
-
-	if (ipv4_is_multicast(daddr)) {
-		if (!ipc.oif)
-			ipc.oif = inet->mc_index;
-		if (!saddr)
-			saddr = inet->mc_addr;
-		connected = 0;
-	}
-
-	if (connected)
-		rt = (struct rtable*)sk_dst_check(sk, 0);
-
-	if (rt == NULL) {
-		struct flowi fl = { .oif = ipc.oif,
-				    .nl_u = { .ip4_u =
-					      { .daddr = faddr,
-						.saddr = saddr,
-						.tos = tos } },
-				    .proto = sk->sk_protocol,
-				    .uli_u = { .ports =
-					       { .sport = inet->sport,
-						 .dport = dport } } };
-		security_sk_classify_flow(sk, &fl);
-		err = ip_route_output_flow(&init_net, &rt, &fl, sk, 1);
-		if (err) {
-			if (err == -ENETUNREACH)
-				IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
-			goto out;
-		}
-
-		err = -EACCES;
-		if ((rt->rt_flags & RTCF_BROADCAST) &&
-		    !sock_flag(sk, SOCK_BROADCAST))
-			goto out;
-		if (connected)
-			sk_dst_set(sk, dst_clone(&rt->u.dst));
-	}
-
-	if (msg->msg_flags&MSG_CONFIRM)
-		goto do_confirm;
-back_from_confirm:
-
-	saddr = rt->rt_src;
-	if (!ipc.addr)
-		daddr = ipc.addr = rt->rt_dst;
-
-	lock_sock(sk);
-	if (unlikely(up->pending)) {
-		/* The socket is already corked while preparing it. */
-		/* ... which is an evident application bug. --ANK */
-		release_sock(sk);
-
-		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
-		err = -EINVAL;
-		goto out;
-	}
-	/*
-	 *	Now cork the socket to pend data.
-	 */
-	inet->cork.fl.fl4_dst = daddr;
-	inet->cork.fl.fl_ip_dport = dport;
-	inet->cork.fl.fl4_src = saddr;
-	inet->cork.fl.fl_ip_sport = inet->sport;
-	up->pending = AF_INET;
-
-do_append_data:
-	up->len += ulen;
-	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
-	err = ip_append_data(sk, getfrag, msg->msg_iov, ulen,
-			sizeof(struct udphdr), &ipc, rt,
-			corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags);
-	if (err)
-		udp_flush_pending_frames(sk);
-	else if (!corkreq)
-		err = udp_push_pending_frames(sk);
-	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
-		up->pending = 0;
-	release_sock(sk);
-
-out:
-	ip_rt_put(rt);
-	if (free)
-		kfree(ipc.opt);
-	if (!err)
-		return len;
-	/*
-	 * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space.  Reporting
-	 * ENOBUFS might not be good (it's not tunable per se), but otherwise
-	 * we don't have a good statistic (IpOutDiscards but it can be too many
-	 * things).  We could add another new stat but at least for now that
-	 * seems like overkill.
-	 */
-	if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
-		UDP_INC_STATS_USER(UDP_MIB_SNDBUFERRORS, is_udplite);
-	}
-	return err;
-
-do_confirm:
-	dst_confirm(&rt->u.dst);
-	if (!(msg->msg_flags&MSG_PROBE) || len)
-		goto back_from_confirm;
-	err = 0;
-	goto out;
-}
-
-int udp_sendpage(struct sock *sk, struct page *page, int offset,
-		 size_t size, int flags)
-{
-	struct udp_sock *up = udp_sk(sk);
-	int ret;
-
-	if (!up->pending) {
-		struct msghdr msg = {	.msg_flags = flags|MSG_MORE };
-
-		/* Call udp_sendmsg to specify destination address which
-		 * sendpage interface can't pass.
-		 * This will succeed only when the socket is connected.
-		 */
-		ret = udp_sendmsg(NULL, sk, &msg, 0);
-		if (ret < 0)
-			return ret;
-	}
-
-	lock_sock(sk);
-
-	if (unlikely(!up->pending)) {
-		release_sock(sk);
-
-		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 3\n");
-		return -EINVAL;
-	}
-
-	ret = ip_append_page(sk, page, offset, size, flags);
-	if (ret == -EOPNOTSUPP) {
-		release_sock(sk);
-		return sock_no_sendpage(sk->sk_socket, page, offset,
-					size, flags);
-	}
-	if (ret < 0) {
-		udp_flush_pending_frames(sk);
-		goto out;
-	}
-
-	up->len += size;
-	if (!(up->corkflag || (flags&MSG_MORE)))
-		ret = udp_push_pending_frames(sk);
-	if (!ret)
-		ret = size;
-out:
-	release_sock(sk);
-	return ret;
-}
-
-/*
- * 	This should be easy, if there is something there we
- * 	return it, otherwise we block.
- */
-
-int udp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
-		size_t len, int noblock, int flags, int *addr_len)
-{
-	struct inet_sock *inet = inet_sk(sk);
-	struct sockaddr_in *sin = (struct sockaddr_in *)msg->msg_name;
-	struct sk_buff *skb;
-	unsigned int ulen, copied;
-	int peeked;
-	int err;
-	int is_udplite = IS_UDPLITE(sk);
-
-	/*
-	 *	Check any passed addresses
-	 */
-	if (addr_len)
-		*addr_len=sizeof(*sin);
-
-	if (flags & MSG_ERRQUEUE)
-		return ip_recv_error(sk, msg, len);
-
-try_again:
-	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
-				  &peeked, &err);
-	if (!skb)
-		goto out;
-
-	ulen = skb->len - sizeof(struct udphdr);
-	copied = len;
-	if (copied > ulen)
-		copied = ulen;
-	else if (copied < ulen)
-		msg->msg_flags |= MSG_TRUNC;
-
-	/*
-	 * If checksum is needed at all, try to do it while copying the
-	 * data.  If the data is truncated, or if we only want a partial
-	 * coverage checksum (UDP-Lite), do it before the copy.
-	 */
-
-	if (copied < ulen || UDP_SKB_CB(skb)->partial_cov) {
-		if (udp_lib_checksum_complete(skb))
-			goto csum_copy_err;
-	}
-
-	if (skb_csum_unnecessary(skb))
-		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
-					      msg->msg_iov, copied       );
-	else {
-		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
-
-		if (err == -EINVAL)
-			goto csum_copy_err;
-	}
-
-	if (err)
-		goto out_free;
-
-	if (!peeked)
-		UDP_INC_STATS_USER(UDP_MIB_INDATAGRAMS, is_udplite);
-
-	sock_recv_timestamp(msg, sk, skb);
-
-	/* Copy the address. */
-	if (sin)
-	{
-		sin->sin_family = AF_INET;
-		sin->sin_port = udp_hdr(skb)->source;
-		sin->sin_addr.s_addr = ip_hdr(skb)->saddr;
-		memset(sin->sin_zero, 0, sizeof(sin->sin_zero));
-	}
-	if (inet->cmsg_flags)
-		ip_cmsg_recv(msg, skb);
-
-	err = copied;
-	if (flags & MSG_TRUNC)
-		err = ulen;
-
-out_free:
-	lock_sock(sk);
-	skb_free_datagram(sk, skb);
-	release_sock(sk);
-out:
-	return err;
-
-csum_copy_err:
-	lock_sock(sk);
-	if (!skb_kill_datagram(sk, skb, flags))
-		UDP_INC_STATS_USER(UDP_MIB_INERRORS, is_udplite);
-	release_sock(sk);
-
-	if (noblock)
-		return -EAGAIN;
-	goto try_again;
-}
-
-
-/* returns:
- *  -1: error
- *   0: success
- *  >0: "udp encap" protocol resubmission
- *
- * Note that in the success and error cases, the skb is assumed to
- * have either been requeued or freed.
- */
-int udp_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
-{
-	struct udp_sock *up = udp_sk(sk);
-	int rc;
-	int is_udplite = IS_UDPLITE(sk);
-
-	/*
-	 *	Charge it to the socket, dropping if the queue is full.
-	 */
-	if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb))
-		goto drop;
-	nf_reset(skb);
-
-	if (up->encap_type) {
-		/*
-		 * This is an encapsulation socket so pass the skb to
-		 * the socket's udp_encap_rcv() hook. Otherwise, just
-		 * fall through and pass this up the UDP socket.
-		 * up->encap_rcv() returns the following value:
-		 * =0 if skb was successfully passed to the encap
-		 *    handler or was discarded by it.
-		 * >0 if skb should be passed on to UDP.
-		 * <0 if skb should be resubmitted as proto -N
-		 */
-
-		/* if we're overly short, let UDP handle it */
-		if (skb->len > sizeof(struct udphdr) &&
-		    up->encap_rcv != NULL) {
-			int ret;
-
-			ret = (*up->encap_rcv)(sk, skb);
-			if (ret <= 0) {
-				UDP_INC_STATS_BH(UDP_MIB_INDATAGRAMS,
-						 is_udplite);
-				return -ret;
-			}
-		}
-
-		/* FALLTHROUGH -- it's a UDP Packet */
-	}
-
-	/*
-	 * 	UDP-Lite specific tests, ignored on UDP sockets
-	 */
-	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
-
-		/*
-		 * MIB statistics other than incrementing the error count are
-		 * disabled for the following two types of errors: these depend
-		 * on the application settings, not on the functioning of the
-		 * protocol stack as such.
-		 *
-		 * RFC 3828 here recommends (sec 3.3): "There should also be a
-		 * way ... to ... at least let the receiving application block
-		 * delivery of packets with coverage values less than a value
-		 * provided by the application."
-		 */
-		if (up->pcrlen == 0) {          /* full coverage was set  */
-			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE: partial coverage "
-				"%d while full coverage %d requested\n",
-				UDP_SKB_CB(skb)->cscov, skb->len);
-			goto drop;
-		}
-		/* The next case involves violating the min. coverage requested
-		 * by the receiver. This is subtle: if receiver wants x and x is
-		 * greater than the buffersize/MTU then receiver will complain
-		 * that it wants x while sender emits packets of smaller size y.
-		 * Therefore the above ...()->partial_cov statement is essential.
-		 */
-		if (UDP_SKB_CB(skb)->cscov  <  up->pcrlen) {
-			LIMIT_NETDEBUG(KERN_WARNING
-				"UDPLITE: coverage %d too small, need min %d\n",
-				UDP_SKB_CB(skb)->cscov, up->pcrlen);
-			goto drop;
-		}
-	}
-
-	if (sk->sk_filter) {
-		if (udp_lib_checksum_complete(skb))
-			goto drop;
-	}
-
-	if ((rc = sock_queue_rcv_skb(sk,skb)) < 0) {
-		/* Note that an ENOMEM error is charged twice */
-		if (rc == -ENOMEM)
-			UDP_INC_STATS_BH(UDP_MIB_RCVBUFERRORS, is_udplite);
-		goto drop;
-	}
-
-	return 0;
-
-drop:
-	UDP_INC_STATS_BH(UDP_MIB_INERRORS, is_udplite);
-	kfree_skb(skb);
-	return -1;
-}
-
-/*
- *	Multicasts and broadcasts go to each listener.
- *
- *	Note: called only from the BH handler context,
- *	so we don't need to lock the hashes.
- */
-static int __udp4_lib_mcast_deliver(struct sk_buff *skb,
-				    struct udphdr  *uh,
-				    __be32 saddr, __be32 daddr,
-				    struct hlist_head udptable[])
-{
-	struct sock *sk;
-	int dif;
-
-	read_lock(&udp_hash_lock);
-	sk = sk_head(&udptable[ntohs(uh->dest) & (UDP_HTABLE_SIZE - 1)]);
-	dif = skb->dev->ifindex;
-	sk = udp_v4_mcast_next(sk, uh->dest, daddr, uh->source, saddr, dif);
-	if (sk) {
-		struct sock *sknext = NULL;
-
-		do {
-			struct sk_buff *skb1 = skb;
-
-			sknext = udp_v4_mcast_next(sk_next(sk), uh->dest, daddr,
-						   uh->source, saddr, dif);
-			if (sknext)
-				skb1 = skb_clone(skb, GFP_ATOMIC);
-
-			if (skb1) {
-				int ret = 0;
-
-				bh_lock_sock_nested(sk);
-				if (!sock_owned_by_user(sk))
-					ret = udp_queue_rcv_skb(sk, skb1);
-				else
-					sk_add_backlog(sk, skb1);
-				bh_unlock_sock(sk);
-
-				if (ret > 0)
-					/* we should probably re-process instead
-					 * of dropping packets here. */
-					kfree_skb(skb1);
-			}
-			sk = sknext;
-		} while (sknext);
-	} else
-		kfree_skb(skb);
-	read_unlock(&udp_hash_lock);
-	return 0;
-}
-
-/* Initialize UDP checksum. If exited with zero value (success),
- * CHECKSUM_UNNECESSARY means, that no more checks are required.
- * Otherwise, csum completion requires chacksumming packet body,
- * including udp header and folding it to skb->csum.
- */
-static inline int udp4_csum_init(struct sk_buff *skb, struct udphdr *uh,
-				 int proto)
-{
-	const struct iphdr *iph;
-	int err;
-
-	UDP_SKB_CB(skb)->partial_cov = 0;
-	UDP_SKB_CB(skb)->cscov = skb->len;
-
-	if (IS_PROTO_UDPLITE(proto)) {
-		err = udplite_checksum_init(skb, uh);
-		if (err)
-			return err;
-	}
-
-	iph = ip_hdr(skb);
-	if (uh->check == 0) {
-		skb->ip_summed = CHECKSUM_UNNECESSARY;
-	} else if (skb->ip_summed == CHECKSUM_COMPLETE) {
-	       if (!csum_tcpudp_magic(iph->saddr, iph->daddr, skb->len,
-				      proto, skb->csum))
-			skb->ip_summed = CHECKSUM_UNNECESSARY;
-	}
-	if (!skb_csum_unnecessary(skb))
-		skb->csum = csum_tcpudp_nofold(iph->saddr, iph->daddr,
-					       skb->len, proto, 0);
-	/* Probably, we should checksum udp header (it should be in cache
-	 * in any case) and data in tiny packets (< rx copybreak).
-	 */
-
-	return 0;
-}
-
-/*
- *	All we need to do is get the socket, and then do a checksum.
- */
-
-int __udp4_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
-		   int proto)
-{
-	struct sock *sk;
-	struct udphdr *uh = udp_hdr(skb);
-	unsigned short ulen;
-	struct rtable *rt = skb->rtable;
-	__be32 saddr = ip_hdr(skb)->saddr;
-	__be32 daddr = ip_hdr(skb)->daddr;
-
-	/*
-	 *  Validate the packet.
-	 */
-	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
-		goto drop;		/* No space for header. */
-
-	ulen = ntohs(uh->len);
-	if (ulen > skb->len)
-		goto short_packet;
-
-	if (IS_PROTO_UDPLITE(proto)) {
-		/* UDP validates ulen. */
-		if (ulen < sizeof(*uh) || pskb_trim_rcsum(skb, ulen))
-			goto short_packet;
-		uh = udp_hdr(skb);
-	}
-
-	if (udp4_csum_init(skb, uh, proto))
-		goto csum_error;
-
-	if (rt->rt_flags & (RTCF_BROADCAST|RTCF_MULTICAST))
-		return __udp4_lib_mcast_deliver(skb, uh, saddr, daddr, udptable);
-
-	sk = __udp4_lib_lookup(skb->dev->nd_net, saddr, uh->source, daddr,
-			uh->dest, inet_iif(skb), udptable);
-
-	if (sk != NULL) {
-		int ret = 0;
-		bh_lock_sock_nested(sk);
-		if (!sock_owned_by_user(sk))
-			ret = udp_queue_rcv_skb(sk, skb);
-		else
-			sk_add_backlog(sk, skb);
-		bh_unlock_sock(sk);
-		sock_put(sk);
-
-		/* a return value > 0 means to resubmit the input, but
-		 * it wants the return to be -protocol, or 0
-		 */
-		if (ret > 0)
-			return -ret;
-		return 0;
-	}
-
-	if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb))
-		goto drop;
-	nf_reset(skb);
-
-	/* No socket. Drop packet silently, if checksum is wrong */
-	if (udp_lib_checksum_complete(skb))
-		goto csum_error;
-
-	UDP_INC_STATS_BH(UDP_MIB_NOPORTS, IS_PROTO_UDPLITE(proto));
-	icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);
-
-	/*
-	 * Hmm.  We got an UDP packet to a port to which we
-	 * don't wanna listen.  Ignore it.
-	 */
-	kfree_skb(skb);
-	return 0;
-
-short_packet:
-	LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: short packet: From %u.%u.%u.%u:%u %d/%d to %u.%u.%u.%u:%u\n",
-		       IS_PROTO_UDPLITE(proto) ? "-Lite" : "",
-		       NIPQUAD(saddr),
-		       ntohs(uh->source),
-		       ulen,
-		       skb->len,
-		       NIPQUAD(daddr),
-		       ntohs(uh->dest));
-	goto drop;
-
-csum_error:
-	/*
-	 * RFC1122: OK.  Discards the bad packet silently (as far as
-	 * the network is concerned, anyway) as per 4.1.3.4 (MUST).
-	 */
-	LIMIT_NETDEBUG(KERN_DEBUG "UDP%s: bad checksum. From %d.%d.%d.%d:%d to %d.%d.%d.%d:%d ulen %d\n",
-		       IS_PROTO_UDPLITE(proto) ? "-Lite" : "",
-		       NIPQUAD(saddr),
-		       ntohs(uh->source),
-		       NIPQUAD(daddr),
-		       ntohs(uh->dest),
-		       ulen);
-drop:
-	UDP_INC_STATS_BH(UDP_MIB_INERRORS, IS_PROTO_UDPLITE(proto));
-	kfree_skb(skb);
-	return 0;
-}
-
-int udp_rcv(struct sk_buff *skb)
-{
-	return __udp4_lib_rcv(skb, udp_hash, IPPROTO_UDP);
-}
-
-int udp_destroy_sock(struct sock *sk)
-{
-	lock_sock(sk);
-	udp_flush_pending_frames(sk);
-	release_sock(sk);
-	return 0;
-}
-
-int udp_setsockopt(struct sock *sk, int level, int optname,
-		   char __user *optval, int optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
-					  udp_push_pending_frames);
-	return ip_setsockopt(sk, level, optname, optval, optlen);
-}
-
-#ifdef CONFIG_COMPAT
-int compat_udp_setsockopt(struct sock *sk, int level, int optname,
-			  char __user *optval, int optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
-					  udp_push_pending_frames);
-	return compat_ip_setsockopt(sk, level, optname, optval, optlen);
-}
-#endif
-
-int udp_getsockopt(struct sock *sk, int level, int optname,
-		   char __user *optval, int __user *optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
-	return ip_getsockopt(sk, level, optname, optval, optlen);
-}
-
-#ifdef CONFIG_COMPAT
-int compat_udp_getsockopt(struct sock *sk, int level, int optname,
-			  char __user *optval, int __user *optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
-	return compat_ip_getsockopt(sk, level, optname, optval, optlen);
-}
-#endif
-
-/* ------------------------------------------------------------------------ */
-DEFINE_PROTO_INUSE(udp)
-
-struct proto udp_prot = {
-	.name		   = "UDP",
-	.owner		   = THIS_MODULE,
-	.close		   = udp_lib_close,
-	.connect	   = ip4_datagram_connect,
-	.disconnect	   = udp_disconnect,
-	.ioctl		   = udp_ioctl,
-	.destroy	   = udp_destroy_sock,
-	.setsockopt	   = udp_setsockopt,
-	.getsockopt	   = udp_getsockopt,
-	.sendmsg	   = udp_sendmsg,
-	.recvmsg	   = udp_recvmsg,
-	.sendpage	   = udp_sendpage,
-	.backlog_rcv	   = udp_queue_rcv_skb,
-	.hash		   = udp_lib_hash,
-	.unhash		   = udp_lib_unhash,
-	.get_port	   = udp_v4_get_port,
-	.memory_allocated  = &udp_memory_allocated,
-	.sysctl_mem	   = sysctl_udp_mem,
-	.sysctl_wmem	   = &sysctl_udp_wmem_min,
-	.sysctl_rmem	   = &sysctl_udp_rmem_min,
-	.obj_size	   = sizeof(struct udp_sock),
-#ifdef CONFIG_COMPAT
-	.compat_setsockopt = compat_udp_setsockopt,
-	.compat_getsockopt = compat_udp_getsockopt,
-#endif
-	REF_PROTO_INUSE(udp)
-};
-
-/* ------------------------------------------------------------------------ */
-static void udp4_format_sock(struct sock *sp, char *tmpbuf, int bucket)
-{
-	struct inet_sock *inet = inet_sk(sp);
-	__be32 dest = inet->daddr;
-	__be32 src  = inet->rcv_saddr;
-	__u16 destp	  = ntohs(inet->dport);
-	__u16 srcp	  = ntohs(inet->sport);
-
-	sprintf(tmpbuf, "%4d: %08X:%04X %08X:%04X"
-		" %02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p",
-		bucket, src, srcp, dest, destp, sp->sk_state,
-		atomic_read(&sp->sk_wmem_alloc),
-		atomic_read(&sp->sk_rmem_alloc),
-		0, 0L, 0, sock_i_uid(sp), 0, sock_i_ino(sp),
-		atomic_read(&sp->sk_refcnt), sp);
-}
-
-int udp4_seq_show(struct seq_file *seq, void *v)
-{
-	if (v == SEQ_START_TOKEN)
-		seq_printf(seq, "%-127s\n",
-			   "  sl  local_address rem_address   st tx_queue "
-			   "rx_queue tr tm->when retrnsmt   uid  timeout "
-			   "inode");
-	else {
-		char tmpbuf[129];
-		struct udp_iter_state *state = seq->private;
-
-		udp4_format_sock(v, tmpbuf, state->bucket);
-		seq_printf(seq, "%-127s\n", tmpbuf);
-	}
-	return 0;
-}
-
-/* ------------------------------------------------------------------------ */
-#ifdef CONFIG_PROC_FS
-static struct file_operations udp4_seq_fops;
-static struct udp_seq_afinfo udp4_seq_afinfo = {
-	.owner		= THIS_MODULE,
-	.name		= "udp",
-	.family		= AF_INET,
-	.hashtable	= udp_hash,
-	.seq_show	= udp4_seq_show,
-	.seq_fops	= &udp4_seq_fops,
-};
-
-int __init udp4_proc_init(void)
-{
-	return udp_proc_register(&udp4_seq_afinfo);
-}
-
-void udp4_proc_exit(void)
-{
-	udp_proc_unregister(&udp4_seq_afinfo);
-}
-#endif /* CONFIG_PROC_FS */
-
-EXPORT_SYMBOL(udp_prot);
-EXPORT_SYMBOL(udp_sendmsg);
-
diff --git a/net/ipv4/udplite.c b/net/ipv4/udplite.c
new file mode 100644
index 000000000000..d49c6d68c8a9
--- /dev/null
+++ b/net/ipv4/udplite.c
@@ -0,0 +1,121 @@
+/*
+ *  UDPLITE     An implementation of the UDP-Lite protocol (RFC 3828).
+ *
+ *  Version:    $Id: udplite.c,v 1.25 2006/10/19 07:22:36 gerrit Exp $
+ *
+ *  Authors:    Gerrit Renker       <gerrit@erg.abdn.ac.uk>
+ *
+ *  Changes:
+ *  Fixes:
+ *		This program is free software; you can redistribute it and/or
+ *		modify it under the terms of the GNU General Public License
+ *		as published by the Free Software Foundation; either version
+ *		2 of the License, or (at your option) any later version.
+ */
+#include "udp_impl.h"
+DEFINE_SNMP_STAT(struct udp_mib, udplite_statistics)	__read_mostly;
+
+struct hlist_head 	udplite_hash[UDP_HTABLE_SIZE];
+
+int udplite_get_port(struct sock *sk, unsigned short p,
+		     int (*c)(const struct sock *, const struct sock *))
+{
+	return  __udp_lib_get_port(sk, p, udplite_hash, c);
+}
+
+static int udplite_v4_get_port(struct sock *sk, unsigned short snum)
+{
+	return udplite_get_port(sk, snum, ipv4_rcv_saddr_equal);
+}
+
+static int udplite_rcv(struct sk_buff *skb)
+{
+	return __udp4_lib_rcv(skb, udplite_hash, IPPROTO_UDPLITE);
+}
+
+static void udplite_err(struct sk_buff *skb, u32 info)
+{
+	__udp4_lib_err(skb, info, udplite_hash);
+}
+
+static	struct net_protocol udplite_protocol = {
+	.handler	= udplite_rcv,
+	.err_handler	= udplite_err,
+	.no_policy	= 1,
+};
+
+DEFINE_PROTO_INUSE(udplite)
+
+struct proto 	udplite_prot = {
+	.name		   = "UDP-Lite",
+	.owner		   = THIS_MODULE,
+	.close		   = udp_lib_close,
+	.connect	   = ip4_datagram_connect,
+	.disconnect	   = udp_disconnect,
+	.ioctl		   = udp_ioctl,
+	.init		   = udplite_sk_init,
+	.destroy	   = udp_destroy_sock,
+	.setsockopt	   = udp_setsockopt,
+	.getsockopt	   = udp_getsockopt,
+	.sendmsg	   = udp_sendmsg,
+	.recvmsg	   = udp_recvmsg,
+	.sendpage	   = udp_sendpage,
+	.backlog_rcv	   = udp_queue_rcv_skb,
+	.hash		   = udp_lib_hash,
+	.unhash		   = udp_lib_unhash,
+	.get_port	   = udplite_v4_get_port,
+	.obj_size	   = sizeof(struct udp_sock),
+#ifdef CONFIG_COMPAT
+	.compat_setsockopt = compat_udp_setsockopt,
+	.compat_getsockopt = compat_udp_getsockopt,
+#endif
+	REF_PROTO_INUSE(udplite)
+};
+
+static struct inet_protosw udplite4_protosw = {
+	.type		=  SOCK_DGRAM,
+	.protocol	=  IPPROTO_UDPLITE,
+	.prot		=  &udplite_prot,
+	.ops		=  &inet_dgram_ops,
+	.capability	= -1,
+	.no_check	=  0,		/* must checksum (RFC 3828) */
+	.flags		=  INET_PROTOSW_PERMANENT,
+};
+
+#ifdef CONFIG_PROC_FS
+static struct file_operations udplite4_seq_fops;
+static struct udp_seq_afinfo udplite4_seq_afinfo = {
+	.owner		= THIS_MODULE,
+	.name		= "udplite",
+	.family		= AF_INET,
+	.hashtable	= udplite_hash,
+	.seq_show	= udp4_seq_show,
+	.seq_fops	= &udplite4_seq_fops,
+};
+#endif
+
+void __init udplite4_register(void)
+{
+	if (proto_register(&udplite_prot, 1))
+		goto out_register_err;
+
+	if (inet_add_protocol(&udplite_protocol, IPPROTO_UDPLITE) < 0)
+		goto out_unregister_proto;
+
+	inet_register_protosw(&udplite4_protosw);
+
+#ifdef CONFIG_PROC_FS
+	if (udp_proc_register(&udplite4_seq_afinfo)) /* udplite4_proc_init() */
+		printk(KERN_ERR "%s: Cannot register /proc!\n", __func__);
+#endif
+	return;
+
+out_unregister_proto:
+	proto_unregister(&udplite_prot);
+out_register_err:
+	printk(KERN_CRIT "%s: Cannot add UDP-Lite protocol.\n", __func__);
+}
+
+EXPORT_SYMBOL(udplite_hash);
+EXPORT_SYMBOL(udplite_prot);
+EXPORT_SYMBOL(udplite_get_port);
diff --git a/net/ipv4/udplite_ipv4.c b/net/ipv4/udplite_ipv4.c
deleted file mode 100644
index d49c6d68c8a9..000000000000
--- a/net/ipv4/udplite_ipv4.c
+++ /dev/null
@@ -1,121 +0,0 @@
-/*
- *  UDPLITE     An implementation of the UDP-Lite protocol (RFC 3828).
- *
- *  Version:    $Id: udplite.c,v 1.25 2006/10/19 07:22:36 gerrit Exp $
- *
- *  Authors:    Gerrit Renker       <gerrit@erg.abdn.ac.uk>
- *
- *  Changes:
- *  Fixes:
- *		This program is free software; you can redistribute it and/or
- *		modify it under the terms of the GNU General Public License
- *		as published by the Free Software Foundation; either version
- *		2 of the License, or (at your option) any later version.
- */
-#include "udp_impl.h"
-DEFINE_SNMP_STAT(struct udp_mib, udplite_statistics)	__read_mostly;
-
-struct hlist_head 	udplite_hash[UDP_HTABLE_SIZE];
-
-int udplite_get_port(struct sock *sk, unsigned short p,
-		     int (*c)(const struct sock *, const struct sock *))
-{
-	return  __udp_lib_get_port(sk, p, udplite_hash, c);
-}
-
-static int udplite_v4_get_port(struct sock *sk, unsigned short snum)
-{
-	return udplite_get_port(sk, snum, ipv4_rcv_saddr_equal);
-}
-
-static int udplite_rcv(struct sk_buff *skb)
-{
-	return __udp4_lib_rcv(skb, udplite_hash, IPPROTO_UDPLITE);
-}
-
-static void udplite_err(struct sk_buff *skb, u32 info)
-{
-	__udp4_lib_err(skb, info, udplite_hash);
-}
-
-static	struct net_protocol udplite_protocol = {
-	.handler	= udplite_rcv,
-	.err_handler	= udplite_err,
-	.no_policy	= 1,
-};
-
-DEFINE_PROTO_INUSE(udplite)
-
-struct proto 	udplite_prot = {
-	.name		   = "UDP-Lite",
-	.owner		   = THIS_MODULE,
-	.close		   = udp_lib_close,
-	.connect	   = ip4_datagram_connect,
-	.disconnect	   = udp_disconnect,
-	.ioctl		   = udp_ioctl,
-	.init		   = udplite_sk_init,
-	.destroy	   = udp_destroy_sock,
-	.setsockopt	   = udp_setsockopt,
-	.getsockopt	   = udp_getsockopt,
-	.sendmsg	   = udp_sendmsg,
-	.recvmsg	   = udp_recvmsg,
-	.sendpage	   = udp_sendpage,
-	.backlog_rcv	   = udp_queue_rcv_skb,
-	.hash		   = udp_lib_hash,
-	.unhash		   = udp_lib_unhash,
-	.get_port	   = udplite_v4_get_port,
-	.obj_size	   = sizeof(struct udp_sock),
-#ifdef CONFIG_COMPAT
-	.compat_setsockopt = compat_udp_setsockopt,
-	.compat_getsockopt = compat_udp_getsockopt,
-#endif
-	REF_PROTO_INUSE(udplite)
-};
-
-static struct inet_protosw udplite4_protosw = {
-	.type		=  SOCK_DGRAM,
-	.protocol	=  IPPROTO_UDPLITE,
-	.prot		=  &udplite_prot,
-	.ops		=  &inet_dgram_ops,
-	.capability	= -1,
-	.no_check	=  0,		/* must checksum (RFC 3828) */
-	.flags		=  INET_PROTOSW_PERMANENT,
-};
-
-#ifdef CONFIG_PROC_FS
-static struct file_operations udplite4_seq_fops;
-static struct udp_seq_afinfo udplite4_seq_afinfo = {
-	.owner		= THIS_MODULE,
-	.name		= "udplite",
-	.family		= AF_INET,
-	.hashtable	= udplite_hash,
-	.seq_show	= udp4_seq_show,
-	.seq_fops	= &udplite4_seq_fops,
-};
-#endif
-
-void __init udplite4_register(void)
-{
-	if (proto_register(&udplite_prot, 1))
-		goto out_register_err;
-
-	if (inet_add_protocol(&udplite_protocol, IPPROTO_UDPLITE) < 0)
-		goto out_unregister_proto;
-
-	inet_register_protosw(&udplite4_protosw);
-
-#ifdef CONFIG_PROC_FS
-	if (udp_proc_register(&udplite4_seq_afinfo)) /* udplite4_proc_init() */
-		printk(KERN_ERR "%s: Cannot register /proc!\n", __func__);
-#endif
-	return;
-
-out_unregister_proto:
-	proto_unregister(&udplite_prot);
-out_register_err:
-	printk(KERN_CRIT "%s: Cannot add UDP-Lite protocol.\n", __func__);
-}
-
-EXPORT_SYMBOL(udplite_hash);
-EXPORT_SYMBOL(udplite_prot);
-EXPORT_SYMBOL(udplite_get_port);
diff --git a/net/ipv6/Makefile b/net/ipv6/Makefile
index 107051f7c227..ae14617e607f 100644
--- a/net/ipv6/Makefile
+++ b/net/ipv6/Makefile
@@ -6,7 +6,7 @@ obj-$(CONFIG_IPV6) += ipv6.o
 
 ipv6-objs :=	af_inet6.o anycast.o ip6_output.o ip6_input.o addrconf.o \
 		addrlabel.o \
-		route.o ip6_fib.o ipv6_sockglue.o ndisc.o udp_ipv6.o \
+		route.o ip6_fib.o ipv6_sockglue.o ndisc.o udp.o udplite.o \
 		raw.o protocol.o icmp.o mcast.o reassembly.o tcp_ipv6.o \
 		exthdrs.o datagram.o ip6_flowlabel.o inet6_connection_sock.o
 
@@ -17,7 +17,6 @@ ipv6-$(CONFIG_NETFILTER) += netfilter.o
 ipv6-$(CONFIG_IPV6_MULTIPLE_TABLES) += fib6_rules.o
 ipv6-$(CONFIG_PROC_FS) += proc.o
 ipv6-$(CONFIG_SYN_COOKIES) += syncookies.o
-ipv6-$(CONFIG_IP_UDPLITE) += udplite_ipv6.o
 
 ipv6-objs += $(ipv6-y)
 
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index afe9276d0420..730a861b8f41 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -813,16 +813,12 @@ static int __init init_ipv6_mibs(void)
 		goto err_icmpmsg_mib;
 	if (snmp_mib_init((void **)udp_stats_in6, sizeof (struct udp_mib)) < 0)
 		goto err_udp_mib;
-#ifdef CONFIG_IP_UDPLITE
 	if (snmp_mib_init((void **)udplite_stats_in6,
 			  sizeof (struct udp_mib)) < 0)
 		goto err_udplite_mib;
-#endif
 	return 0;
 
-#ifdef CONFIG_IP_UDPLITE
 err_udplite_mib:
-#endif
 	snmp_mib_free((void **)udp_stats_in6);
 err_udp_mib:
 	snmp_mib_free((void **)icmpv6msg_statistics);
@@ -841,9 +837,7 @@ static void cleanup_ipv6_mibs(void)
 	snmp_mib_free((void **)icmpv6_statistics);
 	snmp_mib_free((void **)icmpv6msg_statistics);
 	snmp_mib_free((void **)udp_stats_in6);
-#ifdef CONFIG_IP_UDPLITE
 	snmp_mib_free((void **)udplite_stats_in6);
-#endif
 }
 
 static int inet6_net_init(struct net *net)
@@ -888,11 +882,9 @@ static int __init inet6_init(void)
 	if (err)
 		goto out_unregister_tcp_proto;
 
-#ifdef CONFIG_IP_UDPLITE
 	err = proto_register(&udplitev6_prot, 1);
 	if (err)
 		goto out_unregister_udp_proto;
-#endif
 
 	err = proto_register(&rawv6_prot, 1);
 	if (err)
@@ -1063,10 +1055,8 @@ out_sock_register_fail:
 out_unregister_raw_proto:
 	proto_unregister(&rawv6_prot);
 out_unregister_udplite_proto:
-#ifdef CONFIG_IP_UDPLITE
 	proto_unregister(&udplitev6_prot);
 out_unregister_udp_proto:
-#endif
 	proto_unregister(&udpv6_prot);
 out_unregister_tcp_proto:
 	proto_unregister(&tcpv6_prot);
@@ -1085,9 +1075,7 @@ static void __exit inet6_exit(void)
 	ipv6_sysctl_unregister();
 #endif
 	udpv6_exit();
-#ifdef CONFIG_IP_UDPLITE
 	udplitev6_exit();
-#endif
 	tcpv6_exit();
 
 	/* Cleanup code parts. */
@@ -1117,9 +1105,7 @@ static void __exit inet6_exit(void)
 	unregister_pernet_subsys(&inet6_net_ops);
 	cleanup_ipv6_mibs();
 	proto_unregister(&rawv6_prot);
-#ifdef CONFIG_IP_UDPLITE
 	proto_unregister(&udplitev6_prot);
-#endif
 	proto_unregister(&udpv6_prot);
 	proto_unregister(&tcpv6_prot);
 }
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 3bbfdff698d2..5eea6fa506e5 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -127,9 +127,7 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 			struct sk_buff *pktopt;
 
 			if (sk->sk_protocol != IPPROTO_UDP &&
-#ifdef CONFIG_IP_UDPLITE
 			    sk->sk_protocol != IPPROTO_UDPLITE &&
-#endif
 			    sk->sk_protocol != IPPROTO_TCP)
 				break;
 
@@ -169,7 +167,7 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 			} else {
 				struct proto *prot = &udp_prot;
 
-				if (IS_PROTO_UDPLITE(sk->sk_protocol))
+				if (sk->sk_protocol == IPPROTO_UDPLITE)
 					prot = &udplite_prot;
 				local_bh_disable();
 				sock_prot_inuse_add(sk->sk_prot, -1);
@@ -734,9 +732,7 @@ static int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
 	switch (optname) {
 	case IPV6_ADDRFORM:
 		if (sk->sk_protocol != IPPROTO_UDP &&
-#ifdef CONFIG_IP_UDPLITE
 		    sk->sk_protocol != IPPROTO_UDPLITE &&
-#endif
 		    sk->sk_protocol != IPPROTO_TCP)
 			return -EINVAL;
 		if (sk->sk_state != TCP_ESTABLISHED)
diff --git a/net/ipv6/proc.c b/net/ipv6/proc.c
index 2453f2229ef7..8a5be290c710 100644
--- a/net/ipv6/proc.c
+++ b/net/ipv6/proc.c
@@ -39,10 +39,8 @@ static int sockstat6_seq_show(struct seq_file *seq, void *v)
 		       sock_prot_inuse_get(&tcpv6_prot));
 	seq_printf(seq, "UDP6: inuse %d\n",
 		       sock_prot_inuse_get(&udpv6_prot));
-#ifdef CONFIG_IP_UDPLITE
 	seq_printf(seq, "UDPLITE6: inuse %d\n",
 			sock_prot_inuse_get(&udplitev6_prot));
-#endif
 	seq_printf(seq, "RAW6: inuse %d\n",
 		       sock_prot_inuse_get(&rawv6_prot));
 	seq_printf(seq, "FRAG6: inuse %d memory %d\n",
@@ -113,7 +111,6 @@ static struct snmp_mib snmp6_udp6_list[] = {
 	SNMP_MIB_SENTINEL
 };
 
-#ifdef CONFIG_IP_UDPLITE
 static struct snmp_mib snmp6_udplite6_list[] = {
 	SNMP_MIB_ITEM("UdpLite6InDatagrams", UDP_MIB_INDATAGRAMS),
 	SNMP_MIB_ITEM("UdpLite6NoPorts", UDP_MIB_NOPORTS),
@@ -121,7 +118,6 @@ static struct snmp_mib snmp6_udplite6_list[] = {
 	SNMP_MIB_ITEM("UdpLite6OutDatagrams", UDP_MIB_OUTDATAGRAMS),
 	SNMP_MIB_SENTINEL
 };
-#endif
 
 static void snmp6_seq_show_icmpv6msg(struct seq_file *seq, void **mib)
 {
@@ -180,9 +176,7 @@ static int snmp6_seq_show(struct seq_file *seq, void *v)
 		snmp6_seq_show_item(seq, (void **)icmpv6_statistics, snmp6_icmp6_list);
 		snmp6_seq_show_icmpv6msg(seq, (void **)icmpv6msg_statistics);
 		snmp6_seq_show_item(seq, (void **)udp_stats_in6, snmp6_udp6_list);
-#ifdef CONFIG_IP_UDPLITE
 		snmp6_seq_show_item(seq, (void **)udplite_stats_in6, snmp6_udplite6_list);
-#endif
 	}
 	return 0;
 }
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
new file mode 100644
index 000000000000..53739de829db
--- /dev/null
+++ b/net/ipv6/udp.c
@@ -0,0 +1,1065 @@
+/*
+ *	UDP over IPv6
+ *	Linux INET6 implementation
+ *
+ *	Authors:
+ *	Pedro Roque		<roque@di.fc.ul.pt>
+ *
+ *	Based on linux/ipv4/udp.c
+ *
+ *	$Id: udp.c,v 1.65 2002/02/01 22:01:04 davem Exp $
+ *
+ *	Fixes:
+ *	Hideaki YOSHIFUJI	:	sin6_scope_id support
+ *	YOSHIFUJI Hideaki @USAGI and:	Support IPV6_V6ONLY socket option, which
+ *	Alexey Kuznetsov		allow both IPv4 and IPv6 sockets to bind
+ *					a single port at the same time.
+ *      Kazunori MIYAZAWA @USAGI:       change process style to use ip6_append_data
+ *      YOSHIFUJI Hideaki @USAGI:	convert /proc/net/udp6 to seq_file.
+ *
+ *	This program is free software; you can redistribute it and/or
+ *      modify it under the terms of the GNU General Public License
+ *      as published by the Free Software Foundation; either version
+ *      2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/socket.h>
+#include <linux/sockios.h>
+#include <linux/net.h>
+#include <linux/in6.h>
+#include <linux/netdevice.h>
+#include <linux/if_arp.h>
+#include <linux/ipv6.h>
+#include <linux/icmpv6.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <asm/uaccess.h>
+
+#include <net/ndisc.h>
+#include <net/protocol.h>
+#include <net/transp_v6.h>
+#include <net/ip6_route.h>
+#include <net/raw.h>
+#include <net/tcp_states.h>
+#include <net/ip6_checksum.h>
+#include <net/xfrm.h>
+
+#include <linux/proc_fs.h>
+#include <linux/seq_file.h>
+#include "udp_impl.h"
+
+static inline int udp_v6_get_port(struct sock *sk, unsigned short snum)
+{
+	return udp_get_port(sk, snum, ipv6_rcv_saddr_equal);
+}
+
+static struct sock *__udp6_lib_lookup(struct net *net,
+				      struct in6_addr *saddr, __be16 sport,
+				      struct in6_addr *daddr, __be16 dport,
+				      int dif, struct hlist_head udptable[])
+{
+	struct sock *sk, *result = NULL;
+	struct hlist_node *node;
+	unsigned short hnum = ntohs(dport);
+	int badness = -1;
+
+	read_lock(&udp_hash_lock);
+	sk_for_each(sk, node, &udptable[hnum & (UDP_HTABLE_SIZE - 1)]) {
+		struct inet_sock *inet = inet_sk(sk);
+
+		if (sk->sk_net == net && sk->sk_hash == hnum &&
+				sk->sk_family == PF_INET6) {
+			struct ipv6_pinfo *np = inet6_sk(sk);
+			int score = 0;
+			if (inet->dport) {
+				if (inet->dport != sport)
+					continue;
+				score++;
+			}
+			if (!ipv6_addr_any(&np->rcv_saddr)) {
+				if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
+					continue;
+				score++;
+			}
+			if (!ipv6_addr_any(&np->daddr)) {
+				if (!ipv6_addr_equal(&np->daddr, saddr))
+					continue;
+				score++;
+			}
+			if (sk->sk_bound_dev_if) {
+				if (sk->sk_bound_dev_if != dif)
+					continue;
+				score++;
+			}
+			if (score == 4) {
+				result = sk;
+				break;
+			} else if (score > badness) {
+				result = sk;
+				badness = score;
+			}
+		}
+	}
+	if (result)
+		sock_hold(result);
+	read_unlock(&udp_hash_lock);
+	return result;
+}
+
+/*
+ * 	This should be easy, if there is something there we
+ * 	return it, otherwise we block.
+ */
+
+int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
+		  struct msghdr *msg, size_t len,
+		  int noblock, int flags, int *addr_len)
+{
+	struct ipv6_pinfo *np = inet6_sk(sk);
+	struct inet_sock *inet = inet_sk(sk);
+	struct sk_buff *skb;
+	unsigned int ulen, copied;
+	int peeked;
+	int err;
+	int is_udplite = IS_UDPLITE(sk);
+
+	if (addr_len)
+		*addr_len=sizeof(struct sockaddr_in6);
+
+	if (flags & MSG_ERRQUEUE)
+		return ipv6_recv_error(sk, msg, len);
+
+try_again:
+	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
+				  &peeked, &err);
+	if (!skb)
+		goto out;
+
+	ulen = skb->len - sizeof(struct udphdr);
+	copied = len;
+	if (copied > ulen)
+		copied = ulen;
+	else if (copied < ulen)
+		msg->msg_flags |= MSG_TRUNC;
+
+	/*
+	 * If checksum is needed at all, try to do it while copying the
+	 * data.  If the data is truncated, or if we only want a partial
+	 * coverage checksum (UDP-Lite), do it before the copy.
+	 */
+
+	if (copied < ulen || UDP_SKB_CB(skb)->partial_cov) {
+		if (udp_lib_checksum_complete(skb))
+			goto csum_copy_err;
+	}
+
+	if (skb_csum_unnecessary(skb))
+		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
+					      msg->msg_iov, copied       );
+	else {
+		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
+		if (err == -EINVAL)
+			goto csum_copy_err;
+	}
+	if (err)
+		goto out_free;
+
+	if (!peeked)
+		UDP6_INC_STATS_USER(UDP_MIB_INDATAGRAMS, is_udplite);
+
+	sock_recv_timestamp(msg, sk, skb);
+
+	/* Copy the address. */
+	if (msg->msg_name) {
+		struct sockaddr_in6 *sin6;
+
+		sin6 = (struct sockaddr_in6 *) msg->msg_name;
+		sin6->sin6_family = AF_INET6;
+		sin6->sin6_port = udp_hdr(skb)->source;
+		sin6->sin6_flowinfo = 0;
+		sin6->sin6_scope_id = 0;
+
+		if (skb->protocol == htons(ETH_P_IP))
+			ipv6_addr_set(&sin6->sin6_addr, 0, 0,
+				      htonl(0xffff), ip_hdr(skb)->saddr);
+		else {
+			ipv6_addr_copy(&sin6->sin6_addr,
+				       &ipv6_hdr(skb)->saddr);
+			if (ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)
+				sin6->sin6_scope_id = IP6CB(skb)->iif;
+		}
+
+	}
+	if (skb->protocol == htons(ETH_P_IP)) {
+		if (inet->cmsg_flags)
+			ip_cmsg_recv(msg, skb);
+	} else {
+		if (np->rxopt.all)
+			datagram_recv_ctl(sk, msg, skb);
+	}
+
+	err = copied;
+	if (flags & MSG_TRUNC)
+		err = ulen;
+
+out_free:
+	lock_sock(sk);
+	skb_free_datagram(sk, skb);
+	release_sock(sk);
+out:
+	return err;
+
+csum_copy_err:
+	lock_sock(sk);
+	if (!skb_kill_datagram(sk, skb, flags))
+		UDP6_INC_STATS_USER(UDP_MIB_INERRORS, is_udplite);
+	release_sock(sk);
+
+	if (flags & MSG_DONTWAIT)
+		return -EAGAIN;
+	goto try_again;
+}
+
+void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+		    int type, int code, int offset, __be32 info,
+		    struct hlist_head udptable[]                    )
+{
+	struct ipv6_pinfo *np;
+	struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data;
+	struct in6_addr *saddr = &hdr->saddr;
+	struct in6_addr *daddr = &hdr->daddr;
+	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
+	struct sock *sk;
+	int err;
+
+	sk = __udp6_lib_lookup(skb->dev->nd_net, daddr, uh->dest,
+			       saddr, uh->source, inet6_iif(skb), udptable);
+	if (sk == NULL)
+		return;
+
+	np = inet6_sk(sk);
+
+	if (!icmpv6_err_convert(type, code, &err) && !np->recverr)
+		goto out;
+
+	if (sk->sk_state != TCP_ESTABLISHED && !np->recverr)
+		goto out;
+
+	if (np->recverr)
+		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
+
+	sk->sk_err = err;
+	sk->sk_error_report(sk);
+out:
+	sock_put(sk);
+}
+
+static __inline__ void udpv6_err(struct sk_buff *skb,
+				 struct inet6_skb_parm *opt, int type,
+				 int code, int offset, __be32 info     )
+{
+	__udp6_lib_err(skb, opt, type, code, offset, info, udp_hash);
+}
+
+int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
+{
+	struct udp_sock *up = udp_sk(sk);
+	int rc;
+	int is_udplite = IS_UDPLITE(sk);
+
+	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
+		goto drop;
+
+	/*
+	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
+	 */
+	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
+
+		if (up->pcrlen == 0) {          /* full coverage was set  */
+			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE6: partial coverage"
+				" %d while full coverage %d requested\n",
+				UDP_SKB_CB(skb)->cscov, skb->len);
+			goto drop;
+		}
+		if (UDP_SKB_CB(skb)->cscov  <  up->pcrlen) {
+			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE6: coverage %d "
+						    "too small, need min %d\n",
+				       UDP_SKB_CB(skb)->cscov, up->pcrlen);
+			goto drop;
+		}
+	}
+
+	if (sk->sk_filter) {
+		if (udp_lib_checksum_complete(skb))
+			goto drop;
+	}
+
+	if ((rc = sock_queue_rcv_skb(sk,skb)) < 0) {
+		/* Note that an ENOMEM error is charged twice */
+		if (rc == -ENOMEM)
+			UDP6_INC_STATS_BH(UDP_MIB_RCVBUFERRORS, is_udplite);
+		goto drop;
+	}
+
+	return 0;
+drop:
+	UDP6_INC_STATS_BH(UDP_MIB_INERRORS, is_udplite);
+	kfree_skb(skb);
+	return -1;
+}
+
+static struct sock *udp_v6_mcast_next(struct sock *sk,
+				      __be16 loc_port, struct in6_addr *loc_addr,
+				      __be16 rmt_port, struct in6_addr *rmt_addr,
+				      int dif)
+{
+	struct hlist_node *node;
+	struct sock *s = sk;
+	unsigned short num = ntohs(loc_port);
+
+	sk_for_each_from(s, node) {
+		struct inet_sock *inet = inet_sk(s);
+
+		if (s->sk_hash == num && s->sk_family == PF_INET6) {
+			struct ipv6_pinfo *np = inet6_sk(s);
+			if (inet->dport) {
+				if (inet->dport != rmt_port)
+					continue;
+			}
+			if (!ipv6_addr_any(&np->daddr) &&
+			    !ipv6_addr_equal(&np->daddr, rmt_addr))
+				continue;
+
+			if (s->sk_bound_dev_if && s->sk_bound_dev_if != dif)
+				continue;
+
+			if (!ipv6_addr_any(&np->rcv_saddr)) {
+				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
+					continue;
+			}
+			if (!inet6_mc_check(s, loc_addr, rmt_addr))
+				continue;
+			return s;
+		}
+	}
+	return NULL;
+}
+
+/*
+ * Note: called only from the BH handler context,
+ * so we don't need to lock the hashes.
+ */
+static int __udp6_lib_mcast_deliver(struct sk_buff *skb, struct in6_addr *saddr,
+			   struct in6_addr *daddr, struct hlist_head udptable[])
+{
+	struct sock *sk, *sk2;
+	const struct udphdr *uh = udp_hdr(skb);
+	int dif;
+
+	read_lock(&udp_hash_lock);
+	sk = sk_head(&udptable[ntohs(uh->dest) & (UDP_HTABLE_SIZE - 1)]);
+	dif = inet6_iif(skb);
+	sk = udp_v6_mcast_next(sk, uh->dest, daddr, uh->source, saddr, dif);
+	if (!sk) {
+		kfree_skb(skb);
+		goto out;
+	}
+
+	sk2 = sk;
+	while ((sk2 = udp_v6_mcast_next(sk_next(sk2), uh->dest, daddr,
+					uh->source, saddr, dif))) {
+		struct sk_buff *buff = skb_clone(skb, GFP_ATOMIC);
+		if (buff) {
+			bh_lock_sock_nested(sk2);
+			if (!sock_owned_by_user(sk2))
+				udpv6_queue_rcv_skb(sk2, buff);
+			else
+				sk_add_backlog(sk2, buff);
+			bh_unlock_sock(sk2);
+		}
+	}
+	bh_lock_sock_nested(sk);
+	if (!sock_owned_by_user(sk))
+		udpv6_queue_rcv_skb(sk, skb);
+	else
+		sk_add_backlog(sk, skb);
+	bh_unlock_sock(sk);
+out:
+	read_unlock(&udp_hash_lock);
+	return 0;
+}
+
+static inline int udp6_csum_init(struct sk_buff *skb, struct udphdr *uh,
+				 int proto)
+{
+	int err;
+
+	UDP_SKB_CB(skb)->partial_cov = 0;
+	UDP_SKB_CB(skb)->cscov = skb->len;
+
+	if (proto == IPPROTO_UDPLITE) {
+		err = udplite_checksum_init(skb, uh);
+		if (err)
+			return err;
+	}
+
+	if (uh->check == 0) {
+		/* RFC 2460 section 8.1 says that we SHOULD log
+		   this error. Well, it is reasonable.
+		 */
+		LIMIT_NETDEBUG(KERN_INFO "IPv6: udp checksum is 0\n");
+		return 1;
+	}
+	if (skb->ip_summed == CHECKSUM_COMPLETE &&
+	    !csum_ipv6_magic(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr,
+			     skb->len, proto, skb->csum))
+		skb->ip_summed = CHECKSUM_UNNECESSARY;
+
+	if (!skb_csum_unnecessary(skb))
+		skb->csum = ~csum_unfold(csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
+							 &ipv6_hdr(skb)->daddr,
+							 skb->len, proto, 0));
+
+	return 0;
+}
+
+int __udp6_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
+		   int proto)
+{
+	struct sock *sk;
+	struct udphdr *uh;
+	struct net_device *dev = skb->dev;
+	struct in6_addr *saddr, *daddr;
+	u32 ulen = 0;
+
+	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
+		goto short_packet;
+
+	saddr = &ipv6_hdr(skb)->saddr;
+	daddr = &ipv6_hdr(skb)->daddr;
+	uh = udp_hdr(skb);
+
+	ulen = ntohs(uh->len);
+	if (ulen > skb->len)
+		goto short_packet;
+
+	if (proto == IPPROTO_UDP) {
+		/* UDP validates ulen. */
+
+		/* Check for jumbo payload */
+		if (ulen == 0)
+			ulen = skb->len;
+
+		if (ulen < sizeof(*uh))
+			goto short_packet;
+
+		if (ulen < skb->len) {
+			if (pskb_trim_rcsum(skb, ulen))
+				goto short_packet;
+			saddr = &ipv6_hdr(skb)->saddr;
+			daddr = &ipv6_hdr(skb)->daddr;
+			uh = udp_hdr(skb);
+		}
+	}
+
+	if (udp6_csum_init(skb, uh, proto))
+		goto discard;
+
+	/*
+	 *	Multicast receive code
+	 */
+	if (ipv6_addr_is_multicast(daddr))
+		return __udp6_lib_mcast_deliver(skb, saddr, daddr, udptable);
+
+	/* Unicast */
+
+	/*
+	 * check socket cache ... must talk to Alan about his plans
+	 * for sock caches... i'll skip this for now.
+	 */
+	sk = __udp6_lib_lookup(skb->dev->nd_net, saddr, uh->source,
+			       daddr, uh->dest, inet6_iif(skb), udptable);
+
+	if (sk == NULL) {
+		if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
+			goto discard;
+
+		if (udp_lib_checksum_complete(skb))
+			goto discard;
+		UDP6_INC_STATS_BH(UDP_MIB_NOPORTS, proto == IPPROTO_UDPLITE);
+
+		icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0, dev);
+
+		kfree_skb(skb);
+		return 0;
+	}
+
+	/* deliver */
+
+	bh_lock_sock_nested(sk);
+	if (!sock_owned_by_user(sk))
+		udpv6_queue_rcv_skb(sk, skb);
+	else
+		sk_add_backlog(sk, skb);
+	bh_unlock_sock(sk);
+	sock_put(sk);
+	return 0;
+
+short_packet:
+	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: %d/%u\n",
+		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
+		       ulen, skb->len);
+
+discard:
+	UDP6_INC_STATS_BH(UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
+	kfree_skb(skb);
+	return 0;
+}
+
+static __inline__ int udpv6_rcv(struct sk_buff *skb)
+{
+	return __udp6_lib_rcv(skb, udp_hash, IPPROTO_UDP);
+}
+
+/*
+ * Throw away all pending data and cancel the corking. Socket is locked.
+ */
+static void udp_v6_flush_pending_frames(struct sock *sk)
+{
+	struct udp_sock *up = udp_sk(sk);
+
+	if (up->pending) {
+		up->len = 0;
+		up->pending = 0;
+		ip6_flush_pending_frames(sk);
+	}
+}
+
+/*
+ *	Sending
+ */
+
+static int udp_v6_push_pending_frames(struct sock *sk)
+{
+	struct sk_buff *skb;
+	struct udphdr *uh;
+	struct udp_sock  *up = udp_sk(sk);
+	struct inet_sock *inet = inet_sk(sk);
+	struct flowi *fl = &inet->cork.fl;
+	int err = 0;
+	int is_udplite = IS_UDPLITE(sk);
+	__wsum csum = 0;
+
+	/* Grab the skbuff where UDP header space exists. */
+	if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
+		goto out;
+
+	/*
+	 * Create a UDP header
+	 */
+	uh = udp_hdr(skb);
+	uh->source = fl->fl_ip_sport;
+	uh->dest = fl->fl_ip_dport;
+	uh->len = htons(up->len);
+	uh->check = 0;
+
+	if (is_udplite)
+		csum = udplite_csum_outgoing(sk, skb);
+	 else
+		csum = udp_csum_outgoing(sk, skb);
+
+	/* add protocol-dependent pseudo-header */
+	uh->check = csum_ipv6_magic(&fl->fl6_src, &fl->fl6_dst,
+				    up->len, fl->proto, csum   );
+	if (uh->check == 0)
+		uh->check = CSUM_MANGLED_0;
+
+	err = ip6_push_pending_frames(sk);
+out:
+	up->len = 0;
+	up->pending = 0;
+	if (!err)
+		UDP6_INC_STATS_USER(UDP_MIB_OUTDATAGRAMS, is_udplite);
+	return err;
+}
+
+int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
+		  struct msghdr *msg, size_t len)
+{
+	struct ipv6_txoptions opt_space;
+	struct udp_sock *up = udp_sk(sk);
+	struct inet_sock *inet = inet_sk(sk);
+	struct ipv6_pinfo *np = inet6_sk(sk);
+	struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) msg->msg_name;
+	struct in6_addr *daddr, *final_p = NULL, final;
+	struct ipv6_txoptions *opt = NULL;
+	struct ip6_flowlabel *flowlabel = NULL;
+	struct flowi fl;
+	struct dst_entry *dst;
+	int addr_len = msg->msg_namelen;
+	int ulen = len;
+	int hlimit = -1;
+	int tclass = -1;
+	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
+	int err;
+	int connected = 0;
+	int is_udplite = IS_UDPLITE(sk);
+	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
+
+	/* destination address check */
+	if (sin6) {
+		if (addr_len < offsetof(struct sockaddr, sa_data))
+			return -EINVAL;
+
+		switch (sin6->sin6_family) {
+		case AF_INET6:
+			if (addr_len < SIN6_LEN_RFC2133)
+				return -EINVAL;
+			daddr = &sin6->sin6_addr;
+			break;
+		case AF_INET:
+			goto do_udp_sendmsg;
+		case AF_UNSPEC:
+			msg->msg_name = sin6 = NULL;
+			msg->msg_namelen = addr_len = 0;
+			daddr = NULL;
+			break;
+		default:
+			return -EINVAL;
+		}
+	} else if (!up->pending) {
+		if (sk->sk_state != TCP_ESTABLISHED)
+			return -EDESTADDRREQ;
+		daddr = &np->daddr;
+	} else
+		daddr = NULL;
+
+	if (daddr) {
+		if (ipv6_addr_v4mapped(daddr)) {
+			struct sockaddr_in sin;
+			sin.sin_family = AF_INET;
+			sin.sin_port = sin6 ? sin6->sin6_port : inet->dport;
+			sin.sin_addr.s_addr = daddr->s6_addr32[3];
+			msg->msg_name = &sin;
+			msg->msg_namelen = sizeof(sin);
+do_udp_sendmsg:
+			if (__ipv6_only_sock(sk))
+				return -ENETUNREACH;
+			return udp_sendmsg(iocb, sk, msg, len);
+		}
+	}
+
+	if (up->pending == AF_INET)
+		return udp_sendmsg(iocb, sk, msg, len);
+
+	/* Rough check on arithmetic overflow,
+	   better check is made in ip6_append_data().
+	   */
+	if (len > INT_MAX - sizeof(struct udphdr))
+		return -EMSGSIZE;
+
+	if (up->pending) {
+		/*
+		 * There are pending frames.
+		 * The socket lock must be held while it's corked.
+		 */
+		lock_sock(sk);
+		if (likely(up->pending)) {
+			if (unlikely(up->pending != AF_INET6)) {
+				release_sock(sk);
+				return -EAFNOSUPPORT;
+			}
+			dst = NULL;
+			goto do_append_data;
+		}
+		release_sock(sk);
+	}
+	ulen += sizeof(struct udphdr);
+
+	memset(&fl, 0, sizeof(fl));
+
+	if (sin6) {
+		if (sin6->sin6_port == 0)
+			return -EINVAL;
+
+		fl.fl_ip_dport = sin6->sin6_port;
+		daddr = &sin6->sin6_addr;
+
+		if (np->sndflow) {
+			fl.fl6_flowlabel = sin6->sin6_flowinfo&IPV6_FLOWINFO_MASK;
+			if (fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) {
+				flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
+				if (flowlabel == NULL)
+					return -EINVAL;
+				daddr = &flowlabel->dst;
+			}
+		}
+
+		/*
+		 * Otherwise it will be difficult to maintain
+		 * sk->sk_dst_cache.
+		 */
+		if (sk->sk_state == TCP_ESTABLISHED &&
+		    ipv6_addr_equal(daddr, &np->daddr))
+			daddr = &np->daddr;
+
+		if (addr_len >= sizeof(struct sockaddr_in6) &&
+		    sin6->sin6_scope_id &&
+		    ipv6_addr_type(daddr)&IPV6_ADDR_LINKLOCAL)
+			fl.oif = sin6->sin6_scope_id;
+	} else {
+		if (sk->sk_state != TCP_ESTABLISHED)
+			return -EDESTADDRREQ;
+
+		fl.fl_ip_dport = inet->dport;
+		daddr = &np->daddr;
+		fl.fl6_flowlabel = np->flow_label;
+		connected = 1;
+	}
+
+	if (!fl.oif)
+		fl.oif = sk->sk_bound_dev_if;
+
+	if (msg->msg_controllen) {
+		opt = &opt_space;
+		memset(opt, 0, sizeof(struct ipv6_txoptions));
+		opt->tot_len = sizeof(*opt);
+
+		err = datagram_send_ctl(msg, &fl, opt, &hlimit, &tclass);
+		if (err < 0) {
+			fl6_sock_release(flowlabel);
+			return err;
+		}
+		if ((fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
+			flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
+			if (flowlabel == NULL)
+				return -EINVAL;
+		}
+		if (!(opt->opt_nflen|opt->opt_flen))
+			opt = NULL;
+		connected = 0;
+	}
+	if (opt == NULL)
+		opt = np->opt;
+	if (flowlabel)
+		opt = fl6_merge_options(&opt_space, flowlabel, opt);
+	opt = ipv6_fixup_options(&opt_space, opt);
+
+	fl.proto = sk->sk_protocol;
+	ipv6_addr_copy(&fl.fl6_dst, daddr);
+	if (ipv6_addr_any(&fl.fl6_src) && !ipv6_addr_any(&np->saddr))
+		ipv6_addr_copy(&fl.fl6_src, &np->saddr);
+	fl.fl_ip_sport = inet->sport;
+
+	/* merge ip6_build_xmit from ip6_output */
+	if (opt && opt->srcrt) {
+		struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt;
+		ipv6_addr_copy(&final, &fl.fl6_dst);
+		ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
+		final_p = &final;
+		connected = 0;
+	}
+
+	if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst)) {
+		fl.oif = np->mcast_oif;
+		connected = 0;
+	}
+
+	security_sk_classify_flow(sk, &fl);
+
+	err = ip6_sk_dst_lookup(sk, &dst, &fl);
+	if (err)
+		goto out;
+	if (final_p)
+		ipv6_addr_copy(&fl.fl6_dst, final_p);
+
+	if ((err = __xfrm_lookup(&dst, &fl, sk, XFRM_LOOKUP_WAIT)) < 0) {
+		if (err == -EREMOTE)
+			err = ip6_dst_blackhole(sk, &dst, &fl);
+		if (err < 0)
+			goto out;
+	}
+
+	if (hlimit < 0) {
+		if (ipv6_addr_is_multicast(&fl.fl6_dst))
+			hlimit = np->mcast_hops;
+		else
+			hlimit = np->hop_limit;
+		if (hlimit < 0)
+			hlimit = dst_metric(dst, RTAX_HOPLIMIT);
+		if (hlimit < 0)
+			hlimit = ipv6_get_hoplimit(dst->dev);
+	}
+
+	if (tclass < 0) {
+		tclass = np->tclass;
+		if (tclass < 0)
+			tclass = 0;
+	}
+
+	if (msg->msg_flags&MSG_CONFIRM)
+		goto do_confirm;
+back_from_confirm:
+
+	lock_sock(sk);
+	if (unlikely(up->pending)) {
+		/* The socket is already corked while preparing it. */
+		/* ... which is an evident application bug. --ANK */
+		release_sock(sk);
+
+		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
+		err = -EINVAL;
+		goto out;
+	}
+
+	up->pending = AF_INET6;
+
+do_append_data:
+	up->len += ulen;
+	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
+	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
+		sizeof(struct udphdr), hlimit, tclass, opt, &fl,
+		(struct rt6_info*)dst,
+		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags);
+	if (err)
+		udp_v6_flush_pending_frames(sk);
+	else if (!corkreq)
+		err = udp_v6_push_pending_frames(sk);
+	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
+		up->pending = 0;
+
+	if (dst) {
+		if (connected) {
+			ip6_dst_store(sk, dst,
+				      ipv6_addr_equal(&fl.fl6_dst, &np->daddr) ?
+				      &np->daddr : NULL,
+#ifdef CONFIG_IPV6_SUBTREES
+				      ipv6_addr_equal(&fl.fl6_src, &np->saddr) ?
+				      &np->saddr :
+#endif
+				      NULL);
+		} else {
+			dst_release(dst);
+		}
+	}
+
+	if (err > 0)
+		err = np->recverr ? net_xmit_errno(err) : 0;
+	release_sock(sk);
+out:
+	fl6_sock_release(flowlabel);
+	if (!err)
+		return len;
+	/*
+	 * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space.  Reporting
+	 * ENOBUFS might not be good (it's not tunable per se), but otherwise
+	 * we don't have a good statistic (IpOutDiscards but it can be too many
+	 * things).  We could add another new stat but at least for now that
+	 * seems like overkill.
+	 */
+	if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
+		UDP6_INC_STATS_USER(UDP_MIB_SNDBUFERRORS, is_udplite);
+	}
+	return err;
+
+do_confirm:
+	dst_confirm(dst);
+	if (!(msg->msg_flags&MSG_PROBE) || len)
+		goto back_from_confirm;
+	err = 0;
+	goto out;
+}
+
+int udpv6_destroy_sock(struct sock *sk)
+{
+	lock_sock(sk);
+	udp_v6_flush_pending_frames(sk);
+	release_sock(sk);
+
+	inet6_destroy_sock(sk);
+
+	return 0;
+}
+
+/*
+ *	Socket option code for UDP
+ */
+int udpv6_setsockopt(struct sock *sk, int level, int optname,
+		     char __user *optval, int optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
+					  udp_v6_push_pending_frames);
+	return ipv6_setsockopt(sk, level, optname, optval, optlen);
+}
+
+#ifdef CONFIG_COMPAT
+int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
+			    char __user *optval, int optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
+					  udp_v6_push_pending_frames);
+	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
+}
+#endif
+
+int udpv6_getsockopt(struct sock *sk, int level, int optname,
+		     char __user *optval, int __user *optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
+	return ipv6_getsockopt(sk, level, optname, optval, optlen);
+}
+
+#ifdef CONFIG_COMPAT
+int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
+			    char __user *optval, int __user *optlen)
+{
+	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
+		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
+	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
+}
+#endif
+
+static struct inet6_protocol udpv6_protocol = {
+	.handler	=	udpv6_rcv,
+	.err_handler	=	udpv6_err,
+	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
+};
+
+/* ------------------------------------------------------------------------ */
+#ifdef CONFIG_PROC_FS
+
+static void udp6_sock_seq_show(struct seq_file *seq, struct sock *sp, int bucket)
+{
+	struct inet_sock *inet = inet_sk(sp);
+	struct ipv6_pinfo *np = inet6_sk(sp);
+	struct in6_addr *dest, *src;
+	__u16 destp, srcp;
+
+	dest  = &np->daddr;
+	src   = &np->rcv_saddr;
+	destp = ntohs(inet->dport);
+	srcp  = ntohs(inet->sport);
+	seq_printf(seq,
+		   "%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
+		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p\n",
+		   bucket,
+		   src->s6_addr32[0], src->s6_addr32[1],
+		   src->s6_addr32[2], src->s6_addr32[3], srcp,
+		   dest->s6_addr32[0], dest->s6_addr32[1],
+		   dest->s6_addr32[2], dest->s6_addr32[3], destp,
+		   sp->sk_state,
+		   atomic_read(&sp->sk_wmem_alloc),
+		   atomic_read(&sp->sk_rmem_alloc),
+		   0, 0L, 0,
+		   sock_i_uid(sp), 0,
+		   sock_i_ino(sp),
+		   atomic_read(&sp->sk_refcnt), sp);
+}
+
+int udp6_seq_show(struct seq_file *seq, void *v)
+{
+	if (v == SEQ_START_TOKEN)
+		seq_printf(seq,
+			   "  sl  "
+			   "local_address                         "
+			   "remote_address                        "
+			   "st tx_queue rx_queue tr tm->when retrnsmt"
+			   "   uid  timeout inode\n");
+	else
+		udp6_sock_seq_show(seq, v, ((struct udp_iter_state *)seq->private)->bucket);
+	return 0;
+}
+
+static struct file_operations udp6_seq_fops;
+static struct udp_seq_afinfo udp6_seq_afinfo = {
+	.owner		= THIS_MODULE,
+	.name		= "udp6",
+	.family		= AF_INET6,
+	.hashtable	= udp_hash,
+	.seq_show	= udp6_seq_show,
+	.seq_fops	= &udp6_seq_fops,
+};
+
+int __init udp6_proc_init(void)
+{
+	return udp_proc_register(&udp6_seq_afinfo);
+}
+
+void udp6_proc_exit(void) {
+	udp_proc_unregister(&udp6_seq_afinfo);
+}
+#endif /* CONFIG_PROC_FS */
+
+/* ------------------------------------------------------------------------ */
+
+DEFINE_PROTO_INUSE(udpv6)
+
+struct proto udpv6_prot = {
+	.name		   = "UDPv6",
+	.owner		   = THIS_MODULE,
+	.close		   = udp_lib_close,
+	.connect	   = ip6_datagram_connect,
+	.disconnect	   = udp_disconnect,
+	.ioctl		   = udp_ioctl,
+	.destroy	   = udpv6_destroy_sock,
+	.setsockopt	   = udpv6_setsockopt,
+	.getsockopt	   = udpv6_getsockopt,
+	.sendmsg	   = udpv6_sendmsg,
+	.recvmsg	   = udpv6_recvmsg,
+	.backlog_rcv	   = udpv6_queue_rcv_skb,
+	.hash		   = udp_lib_hash,
+	.unhash		   = udp_lib_unhash,
+	.get_port	   = udp_v6_get_port,
+	.memory_allocated  = &udp_memory_allocated,
+	.sysctl_mem	   = sysctl_udp_mem,
+	.sysctl_wmem	   = &sysctl_udp_wmem_min,
+	.sysctl_rmem	   = &sysctl_udp_rmem_min,
+	.obj_size	   = sizeof(struct udp6_sock),
+#ifdef CONFIG_COMPAT
+	.compat_setsockopt = compat_udpv6_setsockopt,
+	.compat_getsockopt = compat_udpv6_getsockopt,
+#endif
+	REF_PROTO_INUSE(udpv6)
+};
+
+static struct inet_protosw udpv6_protosw = {
+	.type =      SOCK_DGRAM,
+	.protocol =  IPPROTO_UDP,
+	.prot =      &udpv6_prot,
+	.ops =       &inet6_dgram_ops,
+	.capability =-1,
+	.no_check =  UDP_CSUM_DEFAULT,
+	.flags =     INET_PROTOSW_PERMANENT,
+};
+
+
+int __init udpv6_init(void)
+{
+	int ret;
+
+	ret = inet6_add_protocol(&udpv6_protocol, IPPROTO_UDP);
+	if (ret)
+		goto out;
+
+	ret = inet6_register_protosw(&udpv6_protosw);
+	if (ret)
+		goto out_udpv6_protocol;
+out:
+	return ret;
+
+out_udpv6_protocol:
+	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
+	goto out;
+}
+
+void udpv6_exit(void)
+{
+	inet6_unregister_protosw(&udpv6_protosw);
+	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
+}
diff --git a/net/ipv6/udp_ipv6.c b/net/ipv6/udp_ipv6.c
deleted file mode 100644
index 55feac7ba717..000000000000
--- a/net/ipv6/udp_ipv6.c
+++ /dev/null
@@ -1,1065 +0,0 @@
-/*
- *	UDP over IPv6
- *	Linux INET6 implementation
- *
- *	Authors:
- *	Pedro Roque		<roque@di.fc.ul.pt>
- *
- *	Based on linux/ipv4/udp.c
- *
- *	$Id: udp.c,v 1.65 2002/02/01 22:01:04 davem Exp $
- *
- *	Fixes:
- *	Hideaki YOSHIFUJI	:	sin6_scope_id support
- *	YOSHIFUJI Hideaki @USAGI and:	Support IPV6_V6ONLY socket option, which
- *	Alexey Kuznetsov		allow both IPv4 and IPv6 sockets to bind
- *					a single port at the same time.
- *      Kazunori MIYAZAWA @USAGI:       change process style to use ip6_append_data
- *      YOSHIFUJI Hideaki @USAGI:	convert /proc/net/udp6 to seq_file.
- *
- *	This program is free software; you can redistribute it and/or
- *      modify it under the terms of the GNU General Public License
- *      as published by the Free Software Foundation; either version
- *      2 of the License, or (at your option) any later version.
- */
-
-#include <linux/errno.h>
-#include <linux/types.h>
-#include <linux/socket.h>
-#include <linux/sockios.h>
-#include <linux/net.h>
-#include <linux/in6.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/ipv6.h>
-#include <linux/icmpv6.h>
-#include <linux/init.h>
-#include <linux/module.h>
-#include <linux/skbuff.h>
-#include <asm/uaccess.h>
-
-#include <net/ndisc.h>
-#include <net/protocol.h>
-#include <net/transp_v6.h>
-#include <net/ip6_route.h>
-#include <net/raw.h>
-#include <net/tcp_states.h>
-#include <net/ip6_checksum.h>
-#include <net/xfrm.h>
-
-#include <linux/proc_fs.h>
-#include <linux/seq_file.h>
-#include "udp_impl.h"
-
-static inline int udp_v6_get_port(struct sock *sk, unsigned short snum)
-{
-	return udp_get_port(sk, snum, ipv6_rcv_saddr_equal);
-}
-
-static struct sock *__udp6_lib_lookup(struct net *net,
-				      struct in6_addr *saddr, __be16 sport,
-				      struct in6_addr *daddr, __be16 dport,
-				      int dif, struct hlist_head udptable[])
-{
-	struct sock *sk, *result = NULL;
-	struct hlist_node *node;
-	unsigned short hnum = ntohs(dport);
-	int badness = -1;
-
-	read_lock(&udp_hash_lock);
-	sk_for_each(sk, node, &udptable[hnum & (UDP_HTABLE_SIZE - 1)]) {
-		struct inet_sock *inet = inet_sk(sk);
-
-		if (sk->sk_net == net && sk->sk_hash == hnum &&
-				sk->sk_family == PF_INET6) {
-			struct ipv6_pinfo *np = inet6_sk(sk);
-			int score = 0;
-			if (inet->dport) {
-				if (inet->dport != sport)
-					continue;
-				score++;
-			}
-			if (!ipv6_addr_any(&np->rcv_saddr)) {
-				if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
-					continue;
-				score++;
-			}
-			if (!ipv6_addr_any(&np->daddr)) {
-				if (!ipv6_addr_equal(&np->daddr, saddr))
-					continue;
-				score++;
-			}
-			if (sk->sk_bound_dev_if) {
-				if (sk->sk_bound_dev_if != dif)
-					continue;
-				score++;
-			}
-			if (score == 4) {
-				result = sk;
-				break;
-			} else if (score > badness) {
-				result = sk;
-				badness = score;
-			}
-		}
-	}
-	if (result)
-		sock_hold(result);
-	read_unlock(&udp_hash_lock);
-	return result;
-}
-
-/*
- * 	This should be easy, if there is something there we
- * 	return it, otherwise we block.
- */
-
-int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
-		  struct msghdr *msg, size_t len,
-		  int noblock, int flags, int *addr_len)
-{
-	struct ipv6_pinfo *np = inet6_sk(sk);
-	struct inet_sock *inet = inet_sk(sk);
-	struct sk_buff *skb;
-	unsigned int ulen, copied;
-	int peeked;
-	int err;
-	int is_udplite = IS_UDPLITE(sk);
-
-	if (addr_len)
-		*addr_len=sizeof(struct sockaddr_in6);
-
-	if (flags & MSG_ERRQUEUE)
-		return ipv6_recv_error(sk, msg, len);
-
-try_again:
-	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
-				  &peeked, &err);
-	if (!skb)
-		goto out;
-
-	ulen = skb->len - sizeof(struct udphdr);
-	copied = len;
-	if (copied > ulen)
-		copied = ulen;
-	else if (copied < ulen)
-		msg->msg_flags |= MSG_TRUNC;
-
-	/*
-	 * If checksum is needed at all, try to do it while copying the
-	 * data.  If the data is truncated, or if we only want a partial
-	 * coverage checksum (UDP-Lite), do it before the copy.
-	 */
-
-	if (copied < ulen || UDP_SKB_CB(skb)->partial_cov) {
-		if (udp_lib_checksum_complete(skb))
-			goto csum_copy_err;
-	}
-
-	if (skb_csum_unnecessary(skb))
-		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
-					      msg->msg_iov, copied       );
-	else {
-		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
-		if (err == -EINVAL)
-			goto csum_copy_err;
-	}
-	if (err)
-		goto out_free;
-
-	if (!peeked)
-		UDP6_INC_STATS_USER(UDP_MIB_INDATAGRAMS, is_udplite);
-
-	sock_recv_timestamp(msg, sk, skb);
-
-	/* Copy the address. */
-	if (msg->msg_name) {
-		struct sockaddr_in6 *sin6;
-
-		sin6 = (struct sockaddr_in6 *) msg->msg_name;
-		sin6->sin6_family = AF_INET6;
-		sin6->sin6_port = udp_hdr(skb)->source;
-		sin6->sin6_flowinfo = 0;
-		sin6->sin6_scope_id = 0;
-
-		if (skb->protocol == htons(ETH_P_IP))
-			ipv6_addr_set(&sin6->sin6_addr, 0, 0,
-				      htonl(0xffff), ip_hdr(skb)->saddr);
-		else {
-			ipv6_addr_copy(&sin6->sin6_addr,
-				       &ipv6_hdr(skb)->saddr);
-			if (ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)
-				sin6->sin6_scope_id = IP6CB(skb)->iif;
-		}
-
-	}
-	if (skb->protocol == htons(ETH_P_IP)) {
-		if (inet->cmsg_flags)
-			ip_cmsg_recv(msg, skb);
-	} else {
-		if (np->rxopt.all)
-			datagram_recv_ctl(sk, msg, skb);
-	}
-
-	err = copied;
-	if (flags & MSG_TRUNC)
-		err = ulen;
-
-out_free:
-	lock_sock(sk);
-	skb_free_datagram(sk, skb);
-	release_sock(sk);
-out:
-	return err;
-
-csum_copy_err:
-	lock_sock(sk);
-	if (!skb_kill_datagram(sk, skb, flags))
-		UDP6_INC_STATS_USER(UDP_MIB_INERRORS, is_udplite);
-	release_sock(sk);
-
-	if (flags & MSG_DONTWAIT)
-		return -EAGAIN;
-	goto try_again;
-}
-
-void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
-		    int type, int code, int offset, __be32 info,
-		    struct hlist_head udptable[]                    )
-{
-	struct ipv6_pinfo *np;
-	struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data;
-	struct in6_addr *saddr = &hdr->saddr;
-	struct in6_addr *daddr = &hdr->daddr;
-	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
-	struct sock *sk;
-	int err;
-
-	sk = __udp6_lib_lookup(skb->dev->nd_net, daddr, uh->dest,
-			       saddr, uh->source, inet6_iif(skb), udptable);
-	if (sk == NULL)
-		return;
-
-	np = inet6_sk(sk);
-
-	if (!icmpv6_err_convert(type, code, &err) && !np->recverr)
-		goto out;
-
-	if (sk->sk_state != TCP_ESTABLISHED && !np->recverr)
-		goto out;
-
-	if (np->recverr)
-		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
-
-	sk->sk_err = err;
-	sk->sk_error_report(sk);
-out:
-	sock_put(sk);
-}
-
-static __inline__ void udpv6_err(struct sk_buff *skb,
-				 struct inet6_skb_parm *opt, int type,
-				 int code, int offset, __be32 info     )
-{
-	__udp6_lib_err(skb, opt, type, code, offset, info, udp_hash);
-}
-
-int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
-{
-	struct udp_sock *up = udp_sk(sk);
-	int rc;
-	int is_udplite = IS_UDPLITE(sk);
-
-	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
-		goto drop;
-
-	/*
-	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
-	 */
-	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
-
-		if (up->pcrlen == 0) {          /* full coverage was set  */
-			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE6: partial coverage"
-				" %d while full coverage %d requested\n",
-				UDP_SKB_CB(skb)->cscov, skb->len);
-			goto drop;
-		}
-		if (UDP_SKB_CB(skb)->cscov  <  up->pcrlen) {
-			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE6: coverage %d "
-						    "too small, need min %d\n",
-				       UDP_SKB_CB(skb)->cscov, up->pcrlen);
-			goto drop;
-		}
-	}
-
-	if (sk->sk_filter) {
-		if (udp_lib_checksum_complete(skb))
-			goto drop;
-	}
-
-	if ((rc = sock_queue_rcv_skb(sk,skb)) < 0) {
-		/* Note that an ENOMEM error is charged twice */
-		if (rc == -ENOMEM)
-			UDP6_INC_STATS_BH(UDP_MIB_RCVBUFERRORS, is_udplite);
-		goto drop;
-	}
-
-	return 0;
-drop:
-	UDP6_INC_STATS_BH(UDP_MIB_INERRORS, is_udplite);
-	kfree_skb(skb);
-	return -1;
-}
-
-static struct sock *udp_v6_mcast_next(struct sock *sk,
-				      __be16 loc_port, struct in6_addr *loc_addr,
-				      __be16 rmt_port, struct in6_addr *rmt_addr,
-				      int dif)
-{
-	struct hlist_node *node;
-	struct sock *s = sk;
-	unsigned short num = ntohs(loc_port);
-
-	sk_for_each_from(s, node) {
-		struct inet_sock *inet = inet_sk(s);
-
-		if (s->sk_hash == num && s->sk_family == PF_INET6) {
-			struct ipv6_pinfo *np = inet6_sk(s);
-			if (inet->dport) {
-				if (inet->dport != rmt_port)
-					continue;
-			}
-			if (!ipv6_addr_any(&np->daddr) &&
-			    !ipv6_addr_equal(&np->daddr, rmt_addr))
-				continue;
-
-			if (s->sk_bound_dev_if && s->sk_bound_dev_if != dif)
-				continue;
-
-			if (!ipv6_addr_any(&np->rcv_saddr)) {
-				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
-					continue;
-			}
-			if (!inet6_mc_check(s, loc_addr, rmt_addr))
-				continue;
-			return s;
-		}
-	}
-	return NULL;
-}
-
-/*
- * Note: called only from the BH handler context,
- * so we don't need to lock the hashes.
- */
-static int __udp6_lib_mcast_deliver(struct sk_buff *skb, struct in6_addr *saddr,
-			   struct in6_addr *daddr, struct hlist_head udptable[])
-{
-	struct sock *sk, *sk2;
-	const struct udphdr *uh = udp_hdr(skb);
-	int dif;
-
-	read_lock(&udp_hash_lock);
-	sk = sk_head(&udptable[ntohs(uh->dest) & (UDP_HTABLE_SIZE - 1)]);
-	dif = inet6_iif(skb);
-	sk = udp_v6_mcast_next(sk, uh->dest, daddr, uh->source, saddr, dif);
-	if (!sk) {
-		kfree_skb(skb);
-		goto out;
-	}
-
-	sk2 = sk;
-	while ((sk2 = udp_v6_mcast_next(sk_next(sk2), uh->dest, daddr,
-					uh->source, saddr, dif))) {
-		struct sk_buff *buff = skb_clone(skb, GFP_ATOMIC);
-		if (buff) {
-			bh_lock_sock_nested(sk2);
-			if (!sock_owned_by_user(sk2))
-				udpv6_queue_rcv_skb(sk2, buff);
-			else
-				sk_add_backlog(sk2, buff);
-			bh_unlock_sock(sk2);
-		}
-	}
-	bh_lock_sock_nested(sk);
-	if (!sock_owned_by_user(sk))
-		udpv6_queue_rcv_skb(sk, skb);
-	else
-		sk_add_backlog(sk, skb);
-	bh_unlock_sock(sk);
-out:
-	read_unlock(&udp_hash_lock);
-	return 0;
-}
-
-static inline int udp6_csum_init(struct sk_buff *skb, struct udphdr *uh,
-				 int proto)
-{
-	int err;
-
-	UDP_SKB_CB(skb)->partial_cov = 0;
-	UDP_SKB_CB(skb)->cscov = skb->len;
-
-	if (IS_PROTO_UDPLITE(proto)) {
-		err = udplite_checksum_init(skb, uh);
-		if (err)
-			return err;
-	}
-
-	if (uh->check == 0) {
-		/* RFC 2460 section 8.1 says that we SHOULD log
-		   this error. Well, it is reasonable.
-		 */
-		LIMIT_NETDEBUG(KERN_INFO "IPv6: udp checksum is 0\n");
-		return 1;
-	}
-	if (skb->ip_summed == CHECKSUM_COMPLETE &&
-	    !csum_ipv6_magic(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr,
-			     skb->len, proto, skb->csum))
-		skb->ip_summed = CHECKSUM_UNNECESSARY;
-
-	if (!skb_csum_unnecessary(skb))
-		skb->csum = ~csum_unfold(csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
-							 &ipv6_hdr(skb)->daddr,
-							 skb->len, proto, 0));
-
-	return 0;
-}
-
-int __udp6_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
-		   int proto)
-{
-	struct sock *sk;
-	struct udphdr *uh;
-	struct net_device *dev = skb->dev;
-	struct in6_addr *saddr, *daddr;
-	u32 ulen = 0;
-
-	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
-		goto short_packet;
-
-	saddr = &ipv6_hdr(skb)->saddr;
-	daddr = &ipv6_hdr(skb)->daddr;
-	uh = udp_hdr(skb);
-
-	ulen = ntohs(uh->len);
-	if (ulen > skb->len)
-		goto short_packet;
-
-	if (proto == IPPROTO_UDP) {
-		/* UDP validates ulen. */
-
-		/* Check for jumbo payload */
-		if (ulen == 0)
-			ulen = skb->len;
-
-		if (ulen < sizeof(*uh))
-			goto short_packet;
-
-		if (ulen < skb->len) {
-			if (pskb_trim_rcsum(skb, ulen))
-				goto short_packet;
-			saddr = &ipv6_hdr(skb)->saddr;
-			daddr = &ipv6_hdr(skb)->daddr;
-			uh = udp_hdr(skb);
-		}
-	}
-
-	if (udp6_csum_init(skb, uh, proto))
-		goto discard;
-
-	/*
-	 *	Multicast receive code
-	 */
-	if (ipv6_addr_is_multicast(daddr))
-		return __udp6_lib_mcast_deliver(skb, saddr, daddr, udptable);
-
-	/* Unicast */
-
-	/*
-	 * check socket cache ... must talk to Alan about his plans
-	 * for sock caches... i'll skip this for now.
-	 */
-	sk = __udp6_lib_lookup(skb->dev->nd_net, saddr, uh->source,
-			       daddr, uh->dest, inet6_iif(skb), udptable);
-
-	if (sk == NULL) {
-		if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
-			goto discard;
-
-		if (udp_lib_checksum_complete(skb))
-			goto discard;
-		UDP6_INC_STATS_BH(UDP_MIB_NOPORTS, IS_PROTO_UDPLITE(proto));
-
-		icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0, dev);
-
-		kfree_skb(skb);
-		return 0;
-	}
-
-	/* deliver */
-
-	bh_lock_sock_nested(sk);
-	if (!sock_owned_by_user(sk))
-		udpv6_queue_rcv_skb(sk, skb);
-	else
-		sk_add_backlog(sk, skb);
-	bh_unlock_sock(sk);
-	sock_put(sk);
-	return 0;
-
-short_packet:
-	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: %d/%u\n",
-		       IS_PROTO_UDPLITE(proto) ? "-Lite" : "",
-		       ulen, skb->len);
-
-discard:
-	UDP6_INC_STATS_BH(UDP_MIB_INERRORS, IS_PROTO_UDPLITE(proto));
-	kfree_skb(skb);
-	return 0;
-}
-
-static __inline__ int udpv6_rcv(struct sk_buff *skb)
-{
-	return __udp6_lib_rcv(skb, udp_hash, IPPROTO_UDP);
-}
-
-/*
- * Throw away all pending data and cancel the corking. Socket is locked.
- */
-static void udp_v6_flush_pending_frames(struct sock *sk)
-{
-	struct udp_sock *up = udp_sk(sk);
-
-	if (up->pending) {
-		up->len = 0;
-		up->pending = 0;
-		ip6_flush_pending_frames(sk);
-	}
-}
-
-/*
- *	Sending
- */
-
-static int udp_v6_push_pending_frames(struct sock *sk)
-{
-	struct sk_buff *skb;
-	struct udphdr *uh;
-	struct udp_sock  *up = udp_sk(sk);
-	struct inet_sock *inet = inet_sk(sk);
-	struct flowi *fl = &inet->cork.fl;
-	int err = 0;
-	int is_udplite = IS_UDPLITE(sk);
-	__wsum csum = 0;
-
-	/* Grab the skbuff where UDP header space exists. */
-	if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
-		goto out;
-
-	/*
-	 * Create a UDP header
-	 */
-	uh = udp_hdr(skb);
-	uh->source = fl->fl_ip_sport;
-	uh->dest = fl->fl_ip_dport;
-	uh->len = htons(up->len);
-	uh->check = 0;
-
-	if (is_udplite)
-		csum = udplite_csum_outgoing(sk, skb);
-	 else
-		csum = udp_csum_outgoing(sk, skb);
-
-	/* add protocol-dependent pseudo-header */
-	uh->check = csum_ipv6_magic(&fl->fl6_src, &fl->fl6_dst,
-				    up->len, fl->proto, csum   );
-	if (uh->check == 0)
-		uh->check = CSUM_MANGLED_0;
-
-	err = ip6_push_pending_frames(sk);
-out:
-	up->len = 0;
-	up->pending = 0;
-	if (!err)
-		UDP6_INC_STATS_USER(UDP_MIB_OUTDATAGRAMS, is_udplite);
-	return err;
-}
-
-int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
-		  struct msghdr *msg, size_t len)
-{
-	struct ipv6_txoptions opt_space;
-	struct udp_sock *up = udp_sk(sk);
-	struct inet_sock *inet = inet_sk(sk);
-	struct ipv6_pinfo *np = inet6_sk(sk);
-	struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) msg->msg_name;
-	struct in6_addr *daddr, *final_p = NULL, final;
-	struct ipv6_txoptions *opt = NULL;
-	struct ip6_flowlabel *flowlabel = NULL;
-	struct flowi fl;
-	struct dst_entry *dst;
-	int addr_len = msg->msg_namelen;
-	int ulen = len;
-	int hlimit = -1;
-	int tclass = -1;
-	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
-	int err;
-	int connected = 0;
-	int is_udplite = IS_UDPLITE(sk);
-	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
-
-	/* destination address check */
-	if (sin6) {
-		if (addr_len < offsetof(struct sockaddr, sa_data))
-			return -EINVAL;
-
-		switch (sin6->sin6_family) {
-		case AF_INET6:
-			if (addr_len < SIN6_LEN_RFC2133)
-				return -EINVAL;
-			daddr = &sin6->sin6_addr;
-			break;
-		case AF_INET:
-			goto do_udp_sendmsg;
-		case AF_UNSPEC:
-			msg->msg_name = sin6 = NULL;
-			msg->msg_namelen = addr_len = 0;
-			daddr = NULL;
-			break;
-		default:
-			return -EINVAL;
-		}
-	} else if (!up->pending) {
-		if (sk->sk_state != TCP_ESTABLISHED)
-			return -EDESTADDRREQ;
-		daddr = &np->daddr;
-	} else
-		daddr = NULL;
-
-	if (daddr) {
-		if (ipv6_addr_v4mapped(daddr)) {
-			struct sockaddr_in sin;
-			sin.sin_family = AF_INET;
-			sin.sin_port = sin6 ? sin6->sin6_port : inet->dport;
-			sin.sin_addr.s_addr = daddr->s6_addr32[3];
-			msg->msg_name = &sin;
-			msg->msg_namelen = sizeof(sin);
-do_udp_sendmsg:
-			if (__ipv6_only_sock(sk))
-				return -ENETUNREACH;
-			return udp_sendmsg(iocb, sk, msg, len);
-		}
-	}
-
-	if (up->pending == AF_INET)
-		return udp_sendmsg(iocb, sk, msg, len);
-
-	/* Rough check on arithmetic overflow,
-	   better check is made in ip6_append_data().
-	   */
-	if (len > INT_MAX - sizeof(struct udphdr))
-		return -EMSGSIZE;
-
-	if (up->pending) {
-		/*
-		 * There are pending frames.
-		 * The socket lock must be held while it's corked.
-		 */
-		lock_sock(sk);
-		if (likely(up->pending)) {
-			if (unlikely(up->pending != AF_INET6)) {
-				release_sock(sk);
-				return -EAFNOSUPPORT;
-			}
-			dst = NULL;
-			goto do_append_data;
-		}
-		release_sock(sk);
-	}
-	ulen += sizeof(struct udphdr);
-
-	memset(&fl, 0, sizeof(fl));
-
-	if (sin6) {
-		if (sin6->sin6_port == 0)
-			return -EINVAL;
-
-		fl.fl_ip_dport = sin6->sin6_port;
-		daddr = &sin6->sin6_addr;
-
-		if (np->sndflow) {
-			fl.fl6_flowlabel = sin6->sin6_flowinfo&IPV6_FLOWINFO_MASK;
-			if (fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) {
-				flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
-				if (flowlabel == NULL)
-					return -EINVAL;
-				daddr = &flowlabel->dst;
-			}
-		}
-
-		/*
-		 * Otherwise it will be difficult to maintain
-		 * sk->sk_dst_cache.
-		 */
-		if (sk->sk_state == TCP_ESTABLISHED &&
-		    ipv6_addr_equal(daddr, &np->daddr))
-			daddr = &np->daddr;
-
-		if (addr_len >= sizeof(struct sockaddr_in6) &&
-		    sin6->sin6_scope_id &&
-		    ipv6_addr_type(daddr)&IPV6_ADDR_LINKLOCAL)
-			fl.oif = sin6->sin6_scope_id;
-	} else {
-		if (sk->sk_state != TCP_ESTABLISHED)
-			return -EDESTADDRREQ;
-
-		fl.fl_ip_dport = inet->dport;
-		daddr = &np->daddr;
-		fl.fl6_flowlabel = np->flow_label;
-		connected = 1;
-	}
-
-	if (!fl.oif)
-		fl.oif = sk->sk_bound_dev_if;
-
-	if (msg->msg_controllen) {
-		opt = &opt_space;
-		memset(opt, 0, sizeof(struct ipv6_txoptions));
-		opt->tot_len = sizeof(*opt);
-
-		err = datagram_send_ctl(msg, &fl, opt, &hlimit, &tclass);
-		if (err < 0) {
-			fl6_sock_release(flowlabel);
-			return err;
-		}
-		if ((fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
-			flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
-			if (flowlabel == NULL)
-				return -EINVAL;
-		}
-		if (!(opt->opt_nflen|opt->opt_flen))
-			opt = NULL;
-		connected = 0;
-	}
-	if (opt == NULL)
-		opt = np->opt;
-	if (flowlabel)
-		opt = fl6_merge_options(&opt_space, flowlabel, opt);
-	opt = ipv6_fixup_options(&opt_space, opt);
-
-	fl.proto = sk->sk_protocol;
-	ipv6_addr_copy(&fl.fl6_dst, daddr);
-	if (ipv6_addr_any(&fl.fl6_src) && !ipv6_addr_any(&np->saddr))
-		ipv6_addr_copy(&fl.fl6_src, &np->saddr);
-	fl.fl_ip_sport = inet->sport;
-
-	/* merge ip6_build_xmit from ip6_output */
-	if (opt && opt->srcrt) {
-		struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt;
-		ipv6_addr_copy(&final, &fl.fl6_dst);
-		ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
-		final_p = &final;
-		connected = 0;
-	}
-
-	if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst)) {
-		fl.oif = np->mcast_oif;
-		connected = 0;
-	}
-
-	security_sk_classify_flow(sk, &fl);
-
-	err = ip6_sk_dst_lookup(sk, &dst, &fl);
-	if (err)
-		goto out;
-	if (final_p)
-		ipv6_addr_copy(&fl.fl6_dst, final_p);
-
-	if ((err = __xfrm_lookup(&dst, &fl, sk, XFRM_LOOKUP_WAIT)) < 0) {
-		if (err == -EREMOTE)
-			err = ip6_dst_blackhole(sk, &dst, &fl);
-		if (err < 0)
-			goto out;
-	}
-
-	if (hlimit < 0) {
-		if (ipv6_addr_is_multicast(&fl.fl6_dst))
-			hlimit = np->mcast_hops;
-		else
-			hlimit = np->hop_limit;
-		if (hlimit < 0)
-			hlimit = dst_metric(dst, RTAX_HOPLIMIT);
-		if (hlimit < 0)
-			hlimit = ipv6_get_hoplimit(dst->dev);
-	}
-
-	if (tclass < 0) {
-		tclass = np->tclass;
-		if (tclass < 0)
-			tclass = 0;
-	}
-
-	if (msg->msg_flags&MSG_CONFIRM)
-		goto do_confirm;
-back_from_confirm:
-
-	lock_sock(sk);
-	if (unlikely(up->pending)) {
-		/* The socket is already corked while preparing it. */
-		/* ... which is an evident application bug. --ANK */
-		release_sock(sk);
-
-		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
-		err = -EINVAL;
-		goto out;
-	}
-
-	up->pending = AF_INET6;
-
-do_append_data:
-	up->len += ulen;
-	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
-	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
-		sizeof(struct udphdr), hlimit, tclass, opt, &fl,
-		(struct rt6_info*)dst,
-		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags);
-	if (err)
-		udp_v6_flush_pending_frames(sk);
-	else if (!corkreq)
-		err = udp_v6_push_pending_frames(sk);
-	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
-		up->pending = 0;
-
-	if (dst) {
-		if (connected) {
-			ip6_dst_store(sk, dst,
-				      ipv6_addr_equal(&fl.fl6_dst, &np->daddr) ?
-				      &np->daddr : NULL,
-#ifdef CONFIG_IPV6_SUBTREES
-				      ipv6_addr_equal(&fl.fl6_src, &np->saddr) ?
-				      &np->saddr :
-#endif
-				      NULL);
-		} else {
-			dst_release(dst);
-		}
-	}
-
-	if (err > 0)
-		err = np->recverr ? net_xmit_errno(err) : 0;
-	release_sock(sk);
-out:
-	fl6_sock_release(flowlabel);
-	if (!err)
-		return len;
-	/*
-	 * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space.  Reporting
-	 * ENOBUFS might not be good (it's not tunable per se), but otherwise
-	 * we don't have a good statistic (IpOutDiscards but it can be too many
-	 * things).  We could add another new stat but at least for now that
-	 * seems like overkill.
-	 */
-	if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
-		UDP6_INC_STATS_USER(UDP_MIB_SNDBUFERRORS, is_udplite);
-	}
-	return err;
-
-do_confirm:
-	dst_confirm(dst);
-	if (!(msg->msg_flags&MSG_PROBE) || len)
-		goto back_from_confirm;
-	err = 0;
-	goto out;
-}
-
-int udpv6_destroy_sock(struct sock *sk)
-{
-	lock_sock(sk);
-	udp_v6_flush_pending_frames(sk);
-	release_sock(sk);
-
-	inet6_destroy_sock(sk);
-
-	return 0;
-}
-
-/*
- *	Socket option code for UDP
- */
-int udpv6_setsockopt(struct sock *sk, int level, int optname,
-		     char __user *optval, int optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
-					  udp_v6_push_pending_frames);
-	return ipv6_setsockopt(sk, level, optname, optval, optlen);
-}
-
-#ifdef CONFIG_COMPAT
-int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
-			    char __user *optval, int optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
-					  udp_v6_push_pending_frames);
-	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
-}
-#endif
-
-int udpv6_getsockopt(struct sock *sk, int level, int optname,
-		     char __user *optval, int __user *optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
-	return ipv6_getsockopt(sk, level, optname, optval, optlen);
-}
-
-#ifdef CONFIG_COMPAT
-int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
-			    char __user *optval, int __user *optlen)
-{
-	if (IS_SOL_UDPFAMILY(level))
-		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
-	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
-}
-#endif
-
-static struct inet6_protocol udpv6_protocol = {
-	.handler	=	udpv6_rcv,
-	.err_handler	=	udpv6_err,
-	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
-};
-
-/* ------------------------------------------------------------------------ */
-#ifdef CONFIG_PROC_FS
-
-static void udp6_sock_seq_show(struct seq_file *seq, struct sock *sp, int bucket)
-{
-	struct inet_sock *inet = inet_sk(sp);
-	struct ipv6_pinfo *np = inet6_sk(sp);
-	struct in6_addr *dest, *src;
-	__u16 destp, srcp;
-
-	dest  = &np->daddr;
-	src   = &np->rcv_saddr;
-	destp = ntohs(inet->dport);
-	srcp  = ntohs(inet->sport);
-	seq_printf(seq,
-		   "%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
-		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p\n",
-		   bucket,
-		   src->s6_addr32[0], src->s6_addr32[1],
-		   src->s6_addr32[2], src->s6_addr32[3], srcp,
-		   dest->s6_addr32[0], dest->s6_addr32[1],
-		   dest->s6_addr32[2], dest->s6_addr32[3], destp,
-		   sp->sk_state,
-		   atomic_read(&sp->sk_wmem_alloc),
-		   atomic_read(&sp->sk_rmem_alloc),
-		   0, 0L, 0,
-		   sock_i_uid(sp), 0,
-		   sock_i_ino(sp),
-		   atomic_read(&sp->sk_refcnt), sp);
-}
-
-int udp6_seq_show(struct seq_file *seq, void *v)
-{
-	if (v == SEQ_START_TOKEN)
-		seq_printf(seq,
-			   "  sl  "
-			   "local_address                         "
-			   "remote_address                        "
-			   "st tx_queue rx_queue tr tm->when retrnsmt"
-			   "   uid  timeout inode\n");
-	else
-		udp6_sock_seq_show(seq, v, ((struct udp_iter_state *)seq->private)->bucket);
-	return 0;
-}
-
-static struct file_operations udp6_seq_fops;
-static struct udp_seq_afinfo udp6_seq_afinfo = {
-	.owner		= THIS_MODULE,
-	.name		= "udp6",
-	.family		= AF_INET6,
-	.hashtable	= udp_hash,
-	.seq_show	= udp6_seq_show,
-	.seq_fops	= &udp6_seq_fops,
-};
-
-int __init udp6_proc_init(void)
-{
-	return udp_proc_register(&udp6_seq_afinfo);
-}
-
-void udp6_proc_exit(void) {
-	udp_proc_unregister(&udp6_seq_afinfo);
-}
-#endif /* CONFIG_PROC_FS */
-
-/* ------------------------------------------------------------------------ */
-
-DEFINE_PROTO_INUSE(udpv6)
-
-struct proto udpv6_prot = {
-	.name		   = "UDPv6",
-	.owner		   = THIS_MODULE,
-	.close		   = udp_lib_close,
-	.connect	   = ip6_datagram_connect,
-	.disconnect	   = udp_disconnect,
-	.ioctl		   = udp_ioctl,
-	.destroy	   = udpv6_destroy_sock,
-	.setsockopt	   = udpv6_setsockopt,
-	.getsockopt	   = udpv6_getsockopt,
-	.sendmsg	   = udpv6_sendmsg,
-	.recvmsg	   = udpv6_recvmsg,
-	.backlog_rcv	   = udpv6_queue_rcv_skb,
-	.hash		   = udp_lib_hash,
-	.unhash		   = udp_lib_unhash,
-	.get_port	   = udp_v6_get_port,
-	.memory_allocated  = &udp_memory_allocated,
-	.sysctl_mem	   = sysctl_udp_mem,
-	.sysctl_wmem	   = &sysctl_udp_wmem_min,
-	.sysctl_rmem	   = &sysctl_udp_rmem_min,
-	.obj_size	   = sizeof(struct udp6_sock),
-#ifdef CONFIG_COMPAT
-	.compat_setsockopt = compat_udpv6_setsockopt,
-	.compat_getsockopt = compat_udpv6_getsockopt,
-#endif
-	REF_PROTO_INUSE(udpv6)
-};
-
-static struct inet_protosw udpv6_protosw = {
-	.type =      SOCK_DGRAM,
-	.protocol =  IPPROTO_UDP,
-	.prot =      &udpv6_prot,
-	.ops =       &inet6_dgram_ops,
-	.capability =-1,
-	.no_check =  UDP_CSUM_DEFAULT,
-	.flags =     INET_PROTOSW_PERMANENT,
-};
-
-
-int __init udpv6_init(void)
-{
-	int ret;
-
-	ret = inet6_add_protocol(&udpv6_protocol, IPPROTO_UDP);
-	if (ret)
-		goto out;
-
-	ret = inet6_register_protosw(&udpv6_protosw);
-	if (ret)
-		goto out_udpv6_protocol;
-out:
-	return ret;
-
-out_udpv6_protocol:
-	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
-	goto out;
-}
-
-void udpv6_exit(void)
-{
-	inet6_unregister_protosw(&udpv6_protosw);
-	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
-}
diff --git a/net/ipv6/udplite.c b/net/ipv6/udplite.c
new file mode 100644
index 000000000000..87d4202522ee
--- /dev/null
+++ b/net/ipv6/udplite.c
@@ -0,0 +1,125 @@
+/*
+ *  UDPLITEv6   An implementation of the UDP-Lite protocol over IPv6.
+ *              See also net/ipv4/udplite.c
+ *
+ *  Version:    $Id: udplite.c,v 1.9 2006/10/19 08:28:10 gerrit Exp $
+ *
+ *  Authors:    Gerrit Renker       <gerrit@erg.abdn.ac.uk>
+ *
+ *  Changes:
+ *  Fixes:
+ *		This program is free software; you can redistribute it and/or
+ *		modify it under the terms of the GNU General Public License
+ *		as published by the Free Software Foundation; either version
+ *		2 of the License, or (at your option) any later version.
+ */
+#include "udp_impl.h"
+
+DEFINE_SNMP_STAT(struct udp_mib, udplite_stats_in6) __read_mostly;
+
+static int udplitev6_rcv(struct sk_buff *skb)
+{
+	return __udp6_lib_rcv(skb, udplite_hash, IPPROTO_UDPLITE);
+}
+
+static void udplitev6_err(struct sk_buff *skb,
+			  struct inet6_skb_parm *opt,
+			  int type, int code, int offset, __be32 info)
+{
+	__udp6_lib_err(skb, opt, type, code, offset, info, udplite_hash);
+}
+
+static struct inet6_protocol udplitev6_protocol = {
+	.handler	=	udplitev6_rcv,
+	.err_handler	=	udplitev6_err,
+	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
+};
+
+static int udplite_v6_get_port(struct sock *sk, unsigned short snum)
+{
+	return udplite_get_port(sk, snum, ipv6_rcv_saddr_equal);
+}
+
+DEFINE_PROTO_INUSE(udplitev6)
+
+struct proto udplitev6_prot = {
+	.name		   = "UDPLITEv6",
+	.owner		   = THIS_MODULE,
+	.close		   = udp_lib_close,
+	.connect	   = ip6_datagram_connect,
+	.disconnect	   = udp_disconnect,
+	.ioctl		   = udp_ioctl,
+	.init		   = udplite_sk_init,
+	.destroy	   = udpv6_destroy_sock,
+	.setsockopt	   = udpv6_setsockopt,
+	.getsockopt	   = udpv6_getsockopt,
+	.sendmsg	   = udpv6_sendmsg,
+	.recvmsg	   = udpv6_recvmsg,
+	.backlog_rcv	   = udpv6_queue_rcv_skb,
+	.hash		   = udp_lib_hash,
+	.unhash		   = udp_lib_unhash,
+	.get_port	   = udplite_v6_get_port,
+	.obj_size	   = sizeof(struct udp6_sock),
+#ifdef CONFIG_COMPAT
+	.compat_setsockopt = compat_udpv6_setsockopt,
+	.compat_getsockopt = compat_udpv6_getsockopt,
+#endif
+	REF_PROTO_INUSE(udplitev6)
+};
+
+static struct inet_protosw udplite6_protosw = {
+	.type		= SOCK_DGRAM,
+	.protocol	= IPPROTO_UDPLITE,
+	.prot		= &udplitev6_prot,
+	.ops		= &inet6_dgram_ops,
+	.capability	= -1,
+	.no_check	= 0,
+	.flags		= INET_PROTOSW_PERMANENT,
+};
+
+int __init udplitev6_init(void)
+{
+	int ret;
+
+	ret = inet6_add_protocol(&udplitev6_protocol, IPPROTO_UDPLITE);
+	if (ret)
+		goto out;
+
+	ret = inet6_register_protosw(&udplite6_protosw);
+	if (ret)
+		goto out_udplitev6_protocol;
+out:
+	return ret;
+
+out_udplitev6_protocol:
+	inet6_del_protocol(&udplitev6_protocol, IPPROTO_UDPLITE);
+	goto out;
+}
+
+void udplitev6_exit(void)
+{
+	inet6_unregister_protosw(&udplite6_protosw);
+	inet6_del_protocol(&udplitev6_protocol, IPPROTO_UDPLITE);
+}
+
+#ifdef CONFIG_PROC_FS
+static struct file_operations udplite6_seq_fops;
+static struct udp_seq_afinfo udplite6_seq_afinfo = {
+	.owner		= THIS_MODULE,
+	.name		= "udplite6",
+	.family		= AF_INET6,
+	.hashtable	= udplite_hash,
+	.seq_show	= udp6_seq_show,
+	.seq_fops	= &udplite6_seq_fops,
+};
+
+int __init udplite6_proc_init(void)
+{
+	return udp_proc_register(&udplite6_seq_afinfo);
+}
+
+void udplite6_proc_exit(void)
+{
+	udp_proc_unregister(&udplite6_seq_afinfo);
+}
+#endif
diff --git a/net/ipv6/udplite_ipv6.c b/net/ipv6/udplite_ipv6.c
deleted file mode 100644
index 87d4202522ee..000000000000
--- a/net/ipv6/udplite_ipv6.c
+++ /dev/null
@@ -1,125 +0,0 @@
-/*
- *  UDPLITEv6   An implementation of the UDP-Lite protocol over IPv6.
- *              See also net/ipv4/udplite.c
- *
- *  Version:    $Id: udplite.c,v 1.9 2006/10/19 08:28:10 gerrit Exp $
- *
- *  Authors:    Gerrit Renker       <gerrit@erg.abdn.ac.uk>
- *
- *  Changes:
- *  Fixes:
- *		This program is free software; you can redistribute it and/or
- *		modify it under the terms of the GNU General Public License
- *		as published by the Free Software Foundation; either version
- *		2 of the License, or (at your option) any later version.
- */
-#include "udp_impl.h"
-
-DEFINE_SNMP_STAT(struct udp_mib, udplite_stats_in6) __read_mostly;
-
-static int udplitev6_rcv(struct sk_buff *skb)
-{
-	return __udp6_lib_rcv(skb, udplite_hash, IPPROTO_UDPLITE);
-}
-
-static void udplitev6_err(struct sk_buff *skb,
-			  struct inet6_skb_parm *opt,
-			  int type, int code, int offset, __be32 info)
-{
-	__udp6_lib_err(skb, opt, type, code, offset, info, udplite_hash);
-}
-
-static struct inet6_protocol udplitev6_protocol = {
-	.handler	=	udplitev6_rcv,
-	.err_handler	=	udplitev6_err,
-	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
-};
-
-static int udplite_v6_get_port(struct sock *sk, unsigned short snum)
-{
-	return udplite_get_port(sk, snum, ipv6_rcv_saddr_equal);
-}
-
-DEFINE_PROTO_INUSE(udplitev6)
-
-struct proto udplitev6_prot = {
-	.name		   = "UDPLITEv6",
-	.owner		   = THIS_MODULE,
-	.close		   = udp_lib_close,
-	.connect	   = ip6_datagram_connect,
-	.disconnect	   = udp_disconnect,
-	.ioctl		   = udp_ioctl,
-	.init		   = udplite_sk_init,
-	.destroy	   = udpv6_destroy_sock,
-	.setsockopt	   = udpv6_setsockopt,
-	.getsockopt	   = udpv6_getsockopt,
-	.sendmsg	   = udpv6_sendmsg,
-	.recvmsg	   = udpv6_recvmsg,
-	.backlog_rcv	   = udpv6_queue_rcv_skb,
-	.hash		   = udp_lib_hash,
-	.unhash		   = udp_lib_unhash,
-	.get_port	   = udplite_v6_get_port,
-	.obj_size	   = sizeof(struct udp6_sock),
-#ifdef CONFIG_COMPAT
-	.compat_setsockopt = compat_udpv6_setsockopt,
-	.compat_getsockopt = compat_udpv6_getsockopt,
-#endif
-	REF_PROTO_INUSE(udplitev6)
-};
-
-static struct inet_protosw udplite6_protosw = {
-	.type		= SOCK_DGRAM,
-	.protocol	= IPPROTO_UDPLITE,
-	.prot		= &udplitev6_prot,
-	.ops		= &inet6_dgram_ops,
-	.capability	= -1,
-	.no_check	= 0,
-	.flags		= INET_PROTOSW_PERMANENT,
-};
-
-int __init udplitev6_init(void)
-{
-	int ret;
-
-	ret = inet6_add_protocol(&udplitev6_protocol, IPPROTO_UDPLITE);
-	if (ret)
-		goto out;
-
-	ret = inet6_register_protosw(&udplite6_protosw);
-	if (ret)
-		goto out_udplitev6_protocol;
-out:
-	return ret;
-
-out_udplitev6_protocol:
-	inet6_del_protocol(&udplitev6_protocol, IPPROTO_UDPLITE);
-	goto out;
-}
-
-void udplitev6_exit(void)
-{
-	inet6_unregister_protosw(&udplite6_protosw);
-	inet6_del_protocol(&udplitev6_protocol, IPPROTO_UDPLITE);
-}
-
-#ifdef CONFIG_PROC_FS
-static struct file_operations udplite6_seq_fops;
-static struct udp_seq_afinfo udplite6_seq_afinfo = {
-	.owner		= THIS_MODULE,
-	.name		= "udplite6",
-	.family		= AF_INET6,
-	.hashtable	= udplite_hash,
-	.seq_show	= udp6_seq_show,
-	.seq_fops	= &udplite6_seq_fops,
-};
-
-int __init udplite6_proc_init(void)
-{
-	return udp_proc_register(&udplite6_seq_afinfo);
-}
-
-void udplite6_proc_exit(void)
-{
-	udp_proc_unregister(&udplite6_seq_afinfo);
-}
-#endif
-- 
cgit v1.2.3


From 810b38179e9e4d4f57b4b733767bb08f8291a965 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Fri, 29 Feb 2008 15:21:01 -0500
Subject: sched: retain vruntime

Kei Tokunaga reported an interactivity problem when moving tasks
between control groups.

Tasks would retain their old vruntime when moved between groups, this
can cause funny lags. Re-set the vruntime on group move to fit within
the new tree.

Reported-by: Kei Tokunaga <tokunaga.keiich@jp.fujitsu.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |  4 ++++
 kernel/sched.c        |  5 +++++
 kernel/sched_fair.c   | 14 ++++++++++++++
 3 files changed, 23 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 9ae4030067a9..11d8e9a74eff 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -899,6 +899,10 @@ struct sched_class {
 			     int running);
 	void (*prio_changed) (struct rq *this_rq, struct task_struct *task,
 			     int oldprio, int running);
+
+#ifdef CONFIG_FAIR_GROUP_SCHED
+	void (*moved_group) (struct task_struct *p);
+#endif
 };
 
 struct load_weight {
diff --git a/kernel/sched.c b/kernel/sched.c
index dcd553cc4ee8..0b949c4e73ad 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -7625,6 +7625,11 @@ void sched_move_task(struct task_struct *tsk)
 
 	set_task_rq(tsk, task_cpu(tsk));
 
+#ifdef CONFIG_FAIR_GROUP_SCHED
+	if (tsk->sched_class->moved_group)
+		tsk->sched_class->moved_group(tsk);
+#endif
+
 	if (on_rq) {
 		if (unlikely(running))
 			tsk->sched_class->set_curr_task(rq);
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index 3df4d46994ca..e2a530515619 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -1353,6 +1353,16 @@ static void set_curr_task_fair(struct rq *rq)
 		set_next_entity(cfs_rq_of(se), se);
 }
 
+#ifdef CONFIG_FAIR_GROUP_SCHED
+static void moved_group_fair(struct task_struct *p)
+{
+	struct cfs_rq *cfs_rq = task_cfs_rq(p);
+
+	update_curr(cfs_rq);
+	place_entity(cfs_rq, &p->se, 1);
+}
+#endif
+
 /*
  * All the scheduling class methods:
  */
@@ -1381,6 +1391,10 @@ static const struct sched_class fair_sched_class = {
 
 	.prio_changed		= prio_changed_fair,
 	.switched_to		= switched_to_fair,
+
+#ifdef CONFIG_FAIR_GROUP_SCHED
+	.moved_group		= moved_group_fair,
+#endif
 };
 
 #ifdef CONFIG_SCHED_DEBUG
-- 
cgit v1.2.3


From e9720acd728a46cb40daa52c99a979f7c4ff195c Mon Sep 17 00:00:00 2001
From: Pavel Emelyanov <xemul@openvz.org>
Date: Fri, 7 Mar 2008 11:08:40 -0800
Subject: [NET]: Make /proc/net a symlink on /proc/self/net (v3)

Current /proc/net is done with so called "shadows", but current
implementation is broken and has little chances to get fixed.

The problem is that dentries subtree of /proc/net directory has
fancy revalidation rules to make processes living in different
net namespaces see different entries in /proc/net subtree, but
currently, tasks see in the /proc/net subdir the contents of any
other namespace, depending on who opened the file first.

The proposed fix is to turn /proc/net into a symlink, which points
to /proc/self/net, which in turn shows what previously was in
/proc/net - the network-related info, from the net namespace the
appropriate task lives in.

# ls -l /proc/net
lrwxrwxrwx  1 root root 8 Mar  5 15:17 /proc/net -> self/net

In other words - this behaves like /proc/mounts, but unlike
"mounts", "net" is not a file, but a directory.

Changes from v2:
* Fixed discrepancy of /proc/net nlink count and selinux labeling
  screwup pointed out by Stephen.

  To get the correct nlink count the ->getattr callback for /proc/net
  is overridden to read one from the net->proc_net entry.

  To make selinux still work the net->proc_net entry is initialized
  properly, i.e. with the "net" name and the proc_net parent.

Selinux fixes are
Acked-by:  Stephen Smalley <sds@tycho.nsa.gov>

Changes from v1:
* Fixed a task_struct leak in get_proc_task_net, pointed out by Paul.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 fs/proc/base.c              |   1 +
 fs/proc/generic.c           |  26 ++++++----
 fs/proc/internal.h          |   7 +++
 fs/proc/proc_net.c          | 117 +++++++++++++++++++++++++++++++++-----------
 include/linux/proc_fs.h     |   3 --
 include/net/net_namespace.h |   1 -
 6 files changed, 114 insertions(+), 41 deletions(-)

(limited to 'include/linux')

diff --git a/fs/proc/base.c b/fs/proc/base.c
index 96ee899d6502..cc43cf0c1fa5 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -2274,6 +2274,7 @@ static const struct pid_entry tgid_base_stuff[] = {
 	DIR("task",       S_IRUGO|S_IXUGO, task),
 	DIR("fd",         S_IRUSR|S_IXUSR, fd),
 	DIR("fdinfo",     S_IRUSR|S_IXUSR, fdinfo),
+	DIR("net",        S_IRUGO|S_IXUSR, net),
 	REG("environ",    S_IRUSR, environ),
 	INF("auxv",       S_IRUSR, pid_auxv),
 	ONE("status",     S_IRUGO, pid_status),
diff --git a/fs/proc/generic.c b/fs/proc/generic.c
index 68971e66cd41..a36ad3c75cf4 100644
--- a/fs/proc/generic.c
+++ b/fs/proc/generic.c
@@ -377,15 +377,14 @@ static struct dentry_operations proc_dentry_operations =
  * Don't create negative dentries here, return -ENOENT by hand
  * instead.
  */
-struct dentry *proc_lookup(struct inode * dir, struct dentry *dentry, struct nameidata *nd)
+struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *dir,
+		struct dentry *dentry)
 {
 	struct inode *inode = NULL;
-	struct proc_dir_entry * de;
 	int error = -ENOENT;
 
 	lock_kernel();
 	spin_lock(&proc_subdir_lock);
-	de = PDE(dir);
 	if (de) {
 		for (de = de->subdir; de ; de = de->next) {
 			if (de->namelen != dentry->d_name.len)
@@ -393,8 +392,6 @@ struct dentry *proc_lookup(struct inode * dir, struct dentry *dentry, struct nam
 			if (!memcmp(dentry->d_name.name, de->name, de->namelen)) {
 				unsigned int ino;
 
-				if (de->shadow_proc)
-					de = de->shadow_proc(current, de);
 				ino = de->low_ino;
 				de_get(de);
 				spin_unlock(&proc_subdir_lock);
@@ -417,6 +414,12 @@ out_unlock:
 	return ERR_PTR(error);
 }
 
+struct dentry *proc_lookup(struct inode *dir, struct dentry *dentry,
+		struct nameidata *nd)
+{
+	return proc_lookup_de(PDE(dir), dir, dentry);
+}
+
 /*
  * This returns non-zero if at EOF, so that the /proc
  * root directory can use this and check if it should
@@ -426,10 +429,9 @@ out_unlock:
  * value of the readdir() call, as long as it's non-negative
  * for success..
  */
-int proc_readdir(struct file * filp,
-	void * dirent, filldir_t filldir)
+int proc_readdir_de(struct proc_dir_entry *de, struct file *filp, void *dirent,
+		filldir_t filldir)
 {
-	struct proc_dir_entry * de;
 	unsigned int ino;
 	int i;
 	struct inode *inode = filp->f_path.dentry->d_inode;
@@ -438,7 +440,6 @@ int proc_readdir(struct file * filp,
 	lock_kernel();
 
 	ino = inode->i_ino;
-	de = PDE(inode);
 	if (!de) {
 		ret = -EINVAL;
 		goto out;
@@ -499,6 +500,13 @@ out:	unlock_kernel();
 	return ret;	
 }
 
+int proc_readdir(struct file *filp, void *dirent, filldir_t filldir)
+{
+	struct inode *inode = filp->f_path.dentry->d_inode;
+
+	return proc_readdir_de(PDE(inode), filp, dirent, filldir);
+}
+
 /*
  * These are the generic /proc directory operations. They
  * use the in-memory "struct proc_dir_entry" tree to parse
diff --git a/fs/proc/internal.h b/fs/proc/internal.h
index 1c81c8f1aeed..bc72f5c8c47d 100644
--- a/fs/proc/internal.h
+++ b/fs/proc/internal.h
@@ -64,6 +64,8 @@ extern const struct file_operations proc_numa_maps_operations;
 extern const struct file_operations proc_smaps_operations;
 extern const struct file_operations proc_clear_refs_operations;
 extern const struct file_operations proc_pagemap_operations;
+extern const struct file_operations proc_net_operations;
+extern const struct inode_operations proc_net_inode_operations;
 
 void free_proc_entry(struct proc_dir_entry *de);
 
@@ -83,3 +85,8 @@ static inline int proc_fd(struct inode *inode)
 {
 	return PROC_I(inode)->fd;
 }
+
+struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *ino,
+		struct dentry *dentry);
+int proc_readdir_de(struct proc_dir_entry *de, struct file *filp, void *dirent,
+		filldir_t filldir);
diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c
index 14e9b5aaf863..4caa5f774fb7 100644
--- a/fs/proc/proc_net.c
+++ b/fs/proc/proc_net.c
@@ -63,6 +63,82 @@ int seq_release_net(struct inode *ino, struct file *f)
 }
 EXPORT_SYMBOL_GPL(seq_release_net);
 
+static struct net *get_proc_task_net(struct inode *dir)
+{
+	struct task_struct *task;
+	struct nsproxy *ns;
+	struct net *net = NULL;
+
+	rcu_read_lock();
+	task = pid_task(proc_pid(dir), PIDTYPE_PID);
+	if (task != NULL) {
+		ns = task_nsproxy(task);
+		if (ns != NULL)
+			net = get_net(ns->net_ns);
+	}
+	rcu_read_unlock();
+
+	return net;
+}
+
+static struct dentry *proc_tgid_net_lookup(struct inode *dir,
+		struct dentry *dentry, struct nameidata *nd)
+{
+	struct dentry *de;
+	struct net *net;
+
+	de = ERR_PTR(-ENOENT);
+	net = get_proc_task_net(dir);
+	if (net != NULL) {
+		de = proc_lookup_de(net->proc_net, dir, dentry);
+		put_net(net);
+	}
+	return de;
+}
+
+static int proc_tgid_net_getattr(struct vfsmount *mnt, struct dentry *dentry,
+		struct kstat *stat)
+{
+	struct inode *inode = dentry->d_inode;
+	struct net *net;
+
+	net = get_proc_task_net(inode);
+
+	generic_fillattr(inode, stat);
+
+	if (net != NULL) {
+		stat->nlink = net->proc_net->nlink;
+		put_net(net);
+	}
+
+	return 0;
+}
+
+const struct inode_operations proc_net_inode_operations = {
+	.lookup		= proc_tgid_net_lookup,
+	.getattr	= proc_tgid_net_getattr,
+};
+
+static int proc_tgid_net_readdir(struct file *filp, void *dirent,
+		filldir_t filldir)
+{
+	int ret;
+	struct net *net;
+
+	ret = -EINVAL;
+	net = get_proc_task_net(filp->f_path.dentry->d_inode);
+	if (net != NULL) {
+		ret = proc_readdir_de(net->proc_net, filp, dirent, filldir);
+		put_net(net);
+	}
+	return ret;
+}
+
+const struct file_operations proc_net_operations = {
+	.read		= generic_read_dir,
+	.readdir	= proc_tgid_net_readdir,
+};
+
 
 struct proc_dir_entry *proc_net_fops_create(struct net *net,
 	const char *name, mode_t mode, const struct file_operations *fops)
@@ -83,14 +159,6 @@ struct net *get_proc_net(const struct inode *inode)
 }
 EXPORT_SYMBOL_GPL(get_proc_net);
 
-static struct proc_dir_entry *shadow_pde;
-
-static struct proc_dir_entry *proc_net_shadow(struct task_struct *task,
-						struct proc_dir_entry *de)
-{
-	return task->nsproxy->net_ns->proc_net;
-}
-
 struct proc_dir_entry *proc_net_mkdir(struct net *net, const char *name,
 		struct proc_dir_entry *parent)
 {
@@ -104,45 +172,39 @@ EXPORT_SYMBOL_GPL(proc_net_mkdir);
 
 static __net_init int proc_net_ns_init(struct net *net)
 {
-	struct proc_dir_entry *root, *netd, *net_statd;
+	struct proc_dir_entry *netd, *net_statd;
 	int err;
 
 	err = -ENOMEM;
-	root = kzalloc(sizeof(*root), GFP_KERNEL);
-	if (!root)
+	netd = kzalloc(sizeof(*netd), GFP_KERNEL);
+	if (!netd)
 		goto out;
 
-	err = -EEXIST;
-	netd = proc_net_mkdir(net, "net", root);
-	if (!netd)
-		goto free_root;
+	netd->data = net;
+	netd->nlink = 2;
+	netd->name = "net";
+	netd->namelen = 3;
+	netd->parent = &proc_root;
 
 	err = -EEXIST;
 	net_statd = proc_net_mkdir(net, "stat", netd);
 	if (!net_statd)
 		goto free_net;
 
-	root->data = net;
-
-	net->proc_net_root = root;
 	net->proc_net = netd;
 	net->proc_net_stat = net_statd;
-	err = 0;
+	return 0;
 
+free_net:
+	kfree(netd);
 out:
 	return err;
-free_net:
-	remove_proc_entry("net", root);
-free_root:
-	kfree(root);
-	goto out;
 }
 
 static __net_exit void proc_net_ns_exit(struct net *net)
 {
 	remove_proc_entry("stat", net->proc_net);
-	remove_proc_entry("net", net->proc_net_root);
-	kfree(net->proc_net_root);
+	kfree(net->proc_net);
 }
 
 static struct pernet_operations __net_initdata proc_net_ns_ops = {
@@ -152,8 +214,7 @@ static struct pernet_operations __net_initdata proc_net_ns_ops = {
 
 int __init proc_net_init(void)
 {
-	shadow_pde = proc_mkdir("net", NULL);
-	shadow_pde->shadow_proc = proc_net_shadow;
+	proc_symlink("net", NULL, "self/net");
 
 	return register_pernet_subsys(&proc_net_ns_ops);
 }
diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h
index d9a9e718ad19..9b6c935f69cf 100644
--- a/include/linux/proc_fs.h
+++ b/include/linux/proc_fs.h
@@ -50,8 +50,6 @@ typedef	int (read_proc_t)(char *page, char **start, off_t off,
 typedef	int (write_proc_t)(struct file *file, const char __user *buffer,
 			   unsigned long count, void *data);
 typedef int (get_info_t)(char *, char **, off_t, int);
-typedef struct proc_dir_entry *(shadow_proc_t)(struct task_struct *task,
-						struct proc_dir_entry *pde);
 
 struct proc_dir_entry {
 	unsigned int low_ino;
@@ -82,7 +80,6 @@ struct proc_dir_entry {
 	int pde_users;	/* number of callers into module in progress */
 	spinlock_t pde_unload_lock; /* proc_fops checks and pde_users bumps */
 	struct completion *pde_unload_completion;
-	shadow_proc_t *shadow_proc;
 };
 
 struct kcore_list {
diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index 28738b7d53eb..923f2b8b9096 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -31,7 +31,6 @@ struct net {
 
 	struct proc_dir_entry 	*proc_net;
 	struct proc_dir_entry 	*proc_net_stat;
-	struct proc_dir_entry 	*proc_net_root;
 
 	struct list_head	sysctl_table_headers;
 
-- 
cgit v1.2.3


From e621e69137b24fdbbe7ad28214e8d81e614c25b7 Mon Sep 17 00:00:00 2001
From: "Kirill A. Shutemov" <k.shutemov@gmail.com>
Date: Fri, 7 Mar 2008 11:11:13 -0800
Subject: [NET]: include <linux/types.h> into linux/ethtool.h for __u* typedef

Signed-off-by: Kirill A. Shutemov <k.shutemov@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/ethtool.h | 1 +
 1 file changed, 1 insertion(+)

(limited to 'include/linux')

diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h
index fcbe8b640ffb..c8d216357865 100644
--- a/include/linux/ethtool.h
+++ b/include/linux/ethtool.h
@@ -12,6 +12,7 @@
 #ifndef _LINUX_ETHTOOL_H
 #define _LINUX_ETHTOOL_H
 
+#include <linux/types.h>
 
 /* This should work for both 32 and 64 bit userland. */
 struct ethtool_cmd {
-- 
cgit v1.2.3


From c37dcd334c0b0a46a90cfa13b9f69e2aaa89bc09 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Thu, 6 Mar 2008 12:34:50 -0500
Subject: NFS: Fix the fsid revalidation in nfs_update_inode()

When we detect that we've crossed a mountpoint on the remote server, we
must take care not to use that inode to revalidate the fsid on our
current superblock. To do so, we label the inode as a remote mountpoint,
and check for that in nfs_update_inode().

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/inode.c         | 6 ++++--
 include/linux/nfs_fs.h | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 966a8850aa30..a4c7cf2bff3a 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -299,6 +299,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 				else
 					inode->i_op = &nfs_mountpoint_inode_operations;
 				inode->i_fop = NULL;
+				set_bit(NFS_INO_MOUNTPOINT, &nfsi->flags);
 			}
 		} else if (S_ISLNK(inode->i_mode))
 			inode->i_op = &nfs_symlink_inode_operations;
@@ -1003,8 +1004,9 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 
 	server = NFS_SERVER(inode);
 	/* Update the fsid? */
-	if (S_ISDIR(inode->i_mode)
-			&& !nfs_fsid_equal(&server->fsid, &fattr->fsid))
+	if (S_ISDIR(inode->i_mode) &&
+			!nfs_fsid_equal(&server->fsid, &fattr->fsid) &&
+			!test_bit(NFS_INO_MOUNTPOINT, &nfsi->flags))
 		server->fsid = fattr->fsid;
 
 	/*
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index a69ba80f2dfe..f4a0e4c218df 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -195,6 +195,7 @@ struct nfs_inode {
 #define NFS_INO_ADVISE_RDPLUS	(1)		/* advise readdirplus */
 #define NFS_INO_STALE		(2)		/* possible stale inode */
 #define NFS_INO_ACL_LRU_SET	(3)		/* Inode is on the LRU list */
+#define NFS_INO_MOUNTPOINT	(4)		/* inode is remote mountpoint */
 
 static inline struct nfs_inode *NFS_I(const struct inode *inode)
 {
-- 
cgit v1.2.3


From 38332cb98772f5ea757e6486bed7ed0381cb5f98 Mon Sep 17 00:00:00 2001
From: Segher Boessenkool <segher@kernel.crashing.org>
Date: Tue, 4 Mar 2008 14:59:54 -0800
Subject: time: prevent the loop in timespec_add_ns() from being optimised away

Since some architectures don't support __udivdi3().

Signed-off-by: Segher Boessenkool <segher@kernel.crashing.org>
Cc: john stultz <johnstul@us.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/time.h | 4 ++++
 1 file changed, 4 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/time.h b/include/linux/time.h
index 2091a19f1655..d32ef0ad4c0a 100644
--- a/include/linux/time.h
+++ b/include/linux/time.h
@@ -174,6 +174,10 @@ static inline void timespec_add_ns(struct timespec *a, u64 ns)
 {
 	ns += a->tv_nsec;
 	while(unlikely(ns >= NSEC_PER_SEC)) {
+		/* The following asm() prevents the compiler from
+		 * optimising this loop into a modulo operation.  */
+		asm("" : "+r"(ns));
+
 		ns -= NSEC_PER_SEC;
 		a->tv_sec++;
 	}
-- 
cgit v1.2.3


From 10a398d04c4a1fc395840f4d040493375f562302 Mon Sep 17 00:00:00 2001
From: Roman Zippel <zippel@linux-m68k.org>
Date: Tue, 4 Mar 2008 15:14:26 -0800
Subject: time: remove obsolete CLOCK_TICK_ADJUST

The first version of the ntp_interval/tick_length inconsistent usage patch was
recently merged as bbe4d18ac2e058c56adb0cd71f49d9ed3216a405

http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=bbe4d18ac2e058c56adb0cd71f49d9ed3216a405

While the fix did greatly improve the situation, it was correctly pointed out
by Roman that it does have a small bug: If the users change clocksources after
the system has been running and NTP has made corrections, the correctoins made
against the old clocksource will be applied against the new clocksource,
causing error.

The second attempt, which corrects the issue in the NTP_INTERVAL_LENGTH
definition has also made it up-stream as commit
e13a2e61dd5152f5499d2003470acf9c838eab84

http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=e13a2e61dd5152f5499d2003470acf9c838eab84

Roman has correctly pointed out that CLOCK_TICK_ADJUST is calculated
based on the PIT's frequency, and isn't really relevant to non-PIT
driven clocksources (that is, clocksources other then jiffies and pit).

This patch reverts both of those changes, and simply removes
CLOCK_TICK_ADJUST.

This does remove the granularity error correction for users of PIT and Jiffies
clocksource users, but the granularity error but for the majority of users, it
should be within the 500ppm range NTP can accommodate for.

For systems that have granularity errors greater then 500ppm, the
"ntp_tick_adj=" boot option can be used to compensate.

[johnstul@us.ibm.com: provided changelog]
[mattilinnanvuori@yahoo.com: maek ntp_tick_adj static]
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Acked-by: john stultz <johnstul@us.ibm.com>
Signed-off-by: Matti Linnanvuori <mattilinnanvuori@yahoo.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: mingo@elte.hu
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/timex.h     |  9 +--------
 kernel/time/ntp.c         | 11 ++++++++++-
 kernel/time/timekeeping.c |  6 ++----
 3 files changed, 13 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/timex.h b/include/linux/timex.h
index c3f374786a43..8ea3e71ba7fa 100644
--- a/include/linux/timex.h
+++ b/include/linux/timex.h
@@ -232,14 +232,7 @@ static inline int ntp_synced(void)
 #else
 #define NTP_INTERVAL_FREQ  (HZ)
 #endif
-
-#define CLOCK_TICK_OVERFLOW	(LATCH * HZ - CLOCK_TICK_RATE)
-#define CLOCK_TICK_ADJUST	(((s64)CLOCK_TICK_OVERFLOW * NSEC_PER_SEC) / \
-					(s64)CLOCK_TICK_RATE)
-
-/* Because using NSEC_PER_SEC would be too easy */
-#define NTP_INTERVAL_LENGTH ((((s64)TICK_USEC * NSEC_PER_USEC * USER_HZ) + \
-			      CLOCK_TICK_ADJUST) / NTP_INTERVAL_FREQ)
+#define NTP_INTERVAL_LENGTH (NSEC_PER_SEC/NTP_INTERVAL_FREQ)
 
 /* Returns how long ticks are at present, in ns / 2^(SHIFT_SCALE-10). */
 extern u64 current_tick_length(void);
diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c
index d4bca927f715..5fd9b9469770 100644
--- a/kernel/time/ntp.c
+++ b/kernel/time/ntp.c
@@ -42,12 +42,13 @@ long time_esterror = NTP_PHASE_LIMIT;	/* estimated error (us)		*/
 long time_freq;				/* frequency offset (scaled ppm)*/
 static long time_reftime;		/* time at last adjustment (s)	*/
 long time_adjust;
+static long ntp_tick_adj;
 
 static void ntp_update_frequency(void)
 {
 	u64 second_length = (u64)(tick_usec * NSEC_PER_USEC * USER_HZ)
 				<< TICK_LENGTH_SHIFT;
-	second_length += (s64)CLOCK_TICK_ADJUST << TICK_LENGTH_SHIFT;
+	second_length += (s64)ntp_tick_adj << TICK_LENGTH_SHIFT;
 	second_length += (s64)time_freq << (TICK_LENGTH_SHIFT - SHIFT_NSEC);
 
 	tick_length_base = second_length;
@@ -402,3 +403,11 @@ leave:	if ((time_status & (STA_UNSYNC|STA_CLOCKERR)) != 0)
 	notify_cmos_timer();
 	return(result);
 }
+
+static int __init ntp_tick_adj_setup(char *str)
+{
+	ntp_tick_adj = simple_strtol(str, NULL, 0);
+	return 1;
+}
+
+__setup("ntp_tick_adj=", ntp_tick_adj_setup);
diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 1af9fb050fe2..671af612b768 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -187,8 +187,7 @@ static void change_clocksource(void)
 
 	clock->error = 0;
 	clock->xtime_nsec = 0;
-	clocksource_calculate_interval(clock,
-		(unsigned long)(current_tick_length()>>TICK_LENGTH_SHIFT));
+	clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH);
 
 	tick_clock_notify();
 
@@ -245,8 +244,7 @@ void __init timekeeping_init(void)
 	ntp_clear();
 
 	clock = clocksource_get_next();
-	clocksource_calculate_interval(clock,
-		(unsigned long)(current_tick_length()>>TICK_LENGTH_SHIFT));
+	clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH);
 	clock->cycle_last = clocksource_read(clock);
 
 	xtime.tv_sec = sec;
-- 
cgit v1.2.3


From fbab976d7ce4556d4212d554f766dae461d22e16 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@HansenPartnership.com>
Date: Fri, 7 Mar 2008 08:57:54 -0600
Subject: firmware: provide stubs for the FW_LOADER=n case

libsas has a case where it uses the firmware loader to provide services,
but doesn't want to select it all the time.  This currently causes a
compile failure in libsas if FW_LOADER=n.  Fix this by providing error
stubs for the firmware loader API in the FW_LOADER=n case.

Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/firmware.h | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/firmware.h b/include/linux/firmware.h
index 33d8f2087b6e..4d10c7328d2d 100644
--- a/include/linux/firmware.h
+++ b/include/linux/firmware.h
@@ -10,7 +10,10 @@ struct firmware {
 	size_t size;
 	u8 *data;
 };
+
 struct device;
+
+#if defined(CONFIG_FW_LOADER) || defined(CONFIG_FW_LOADER_MODULE)
 int request_firmware(const struct firmware **fw, const char *name,
 		     struct device *device);
 int request_firmware_nowait(
@@ -19,4 +22,24 @@ int request_firmware_nowait(
 	void (*cont)(const struct firmware *fw, void *context));
 
 void release_firmware(const struct firmware *fw);
+#else
+static inline int request_firmware(const struct firmware **fw,
+				   const char *name,
+				   struct device *device)
+{
+	return -EINVAL;
+}
+static inline int request_firmware_nowait(
+	struct module *module, int uevent,
+	const char *name, struct device *device, void *context,
+	void (*cont)(const struct firmware *fw, void *context))
+{
+	return -EINVAL;
+}
+
+static inline void release_firmware(const struct firmware *fw)
+{
+}
+#endif
+
 #endif
-- 
cgit v1.2.3


From b5e85dee2a5433246d5b7488918a1a0ad22c046a Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Mon, 10 Mar 2008 16:41:06 -0700
Subject: [NETFILTER]: nfnetlink: fix ifdef in nfnetlink_compat.h

Use __KERNEL__ instead of __KERNEL to make sure the headers are not
usable by the kernel.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nfnetlink_compat.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nfnetlink_compat.h b/include/linux/netfilter/nfnetlink_compat.h
index 02a42d875cf7..e1451760c9cd 100644
--- a/include/linux/netfilter/nfnetlink_compat.h
+++ b/include/linux/netfilter/nfnetlink_compat.h
@@ -1,6 +1,6 @@
 #ifndef _NFNETLINK_COMPAT_H
 #define _NFNETLINK_COMPAT_H
-#ifndef __KERNEL
+#ifndef __KERNEL__
 /* Old nfnetlink macros for userspace */
 
 /* nfnetlink groups: Up to 32 maximum */
-- 
cgit v1.2.3


From e61062587d0484c3852e822e844416c728362438 Mon Sep 17 00:00:00 2001
From: "Robert P. J. Day" <rpjday@crashcourse.ca>
Date: Fri, 7 Mar 2008 11:02:00 -0500
Subject: USB: g_printer.h does not need to be "unifdef"ed.

Since the header file g_printer.h doesn't depend on __KERNEL__,
there's no need to unifdef it in the Kbuild file.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/usb/Kbuild | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/usb/Kbuild b/include/linux/usb/Kbuild
index b8cba1dcb2c6..42e84fc315e3 100644
--- a/include/linux/usb/Kbuild
+++ b/include/linux/usb/Kbuild
@@ -3,5 +3,5 @@ header-y += cdc.h
 header-y += ch9.h
 header-y += gadgetfs.h
 header-y += midi.h
-unifdef-y += g_printer.h
+header-y += g_printer.h
 
-- 
cgit v1.2.3


From 20f590df4fbb962d1f8fcb12c4b4e790c7054045 Mon Sep 17 00:00:00 2001
From: "Robert P. J. Day" <rpjday@crashcourse.ca>
Date: Fri, 7 Mar 2008 11:40:07 -0500
Subject: USB: Remove __KERNEL__ check from non-exported gadget.h.

Since the header file gadget.h isn't being exported to userspace,
there seems to be little point having a __KERNEL__ proprocessor check.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/usb/gadget.h | 4 ----
 1 file changed, 4 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/usb/gadget.h b/include/linux/usb/gadget.h
index aa3047ff00d1..f3295296b435 100644
--- a/include/linux/usb/gadget.h
+++ b/include/linux/usb/gadget.h
@@ -15,8 +15,6 @@
 #ifndef __LINUX_USB_GADGET_H
 #define __LINUX_USB_GADGET_H
 
-#ifdef __KERNEL__
-
 struct usb_ep;
 
 /**
@@ -848,6 +846,4 @@ extern struct usb_ep *usb_ep_autoconfig(struct usb_gadget *,
 
 extern void usb_ep_autoconfig_reset(struct usb_gadget *) __devinit;
 
-#endif  /* __KERNEL__ */
-
 #endif	/* __LINUX_USB_GADGET_H */
-- 
cgit v1.2.3


From 9f9351bbe34a9b12966b1fb6f7c21cfe128340c1 Mon Sep 17 00:00:00 2001
From: Andrew Morton <akpm@linux-foundation.org>
Date: Mon, 10 Mar 2008 11:43:34 -0700
Subject: rename DECLARE_PCI_DEVICE_TABLE to DEFINE_PCI_DEVICE_TABLE

This macro is used to define tables, not to declare them.

Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 Documentation/pci.txt | 4 ++--
 include/linux/pci.h   | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/pci.txt b/Documentation/pci.txt
index bb7bd27d4682..d2c2e6e2b224 100644
--- a/Documentation/pci.txt
+++ b/Documentation/pci.txt
@@ -123,7 +123,7 @@ initialization with a pointer to a structure describing the driver
 
 
 The ID table is an array of struct pci_device_id entries ending with an
-all-zero entry; use of the macro DECLARE_PCI_DEVICE_TABLE is the preferred
+all-zero entry; use of the macro DEFINE_PCI_DEVICE_TABLE is the preferred
 method of declaring the table.  Each entry consists of:
 
 	vendor,device	Vendor and device ID to match (or PCI_ANY_ID)
@@ -193,7 +193,7 @@ Tips on when/where to use the above attributes:
 	o Do not mark the struct pci_driver.
 
 	o The ID table array should be marked __devinitconst; this is done
-	  automatically if the table is declared with DECLARE_PCI_DEVICE_TABLE().
+	  automatically if the table is declared with DEFINE_PCI_DEVICE_TABLE().
 
 	o The probe() and remove() functions should be marked __devinit
 	  and __devexit respectively.  All initialization functions
diff --git a/include/linux/pci.h b/include/linux/pci.h
index f3165e7ac431..38eff1947750 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -389,13 +389,13 @@ struct pci_driver {
 #define	to_pci_driver(drv) container_of(drv, struct pci_driver, driver)
 
 /**
- * DECLARE_PCI_DEVICE_TABLE - macro used to describe a pci device table
+ * DEFINE_PCI_DEVICE_TABLE - macro used to describe a pci device table
  * @_table: device table name
  *
  * This macro is used to create a struct pci_device_id array (a device table)
  * in a generic manner.
  */
-#define DECLARE_PCI_DEVICE_TABLE(_table) \
+#define DEFINE_PCI_DEVICE_TABLE(_table) \
 	const struct pci_device_id _table[] __devinitconst
 
 /**
-- 
cgit v1.2.3


From e1f19995f55294fbb00ea22ba85d7b0d80ba3813 Mon Sep 17 00:00:00 2001
From: Alex Dubov <oakad@yahoo.com>
Date: Mon, 10 Mar 2008 11:43:37 -0700
Subject: memstick: introduce correct definitions in the header

Thanks to some input from kind people at JMicron it is now possible to have
more correct definitions of protocol structures and bit field semantics.

Signed-off-by: Alex Dubov <oakad@yahoo.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/memstick/core/memstick.c    |   4 +-
 drivers/memstick/core/mspro_block.c |  16 ++---
 drivers/memstick/host/tifm_ms.c     |  19 +++---
 include/linux/memstick.h            | 130 ++++++++++++++++++++++++------------
 4 files changed, 105 insertions(+), 64 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/memstick/core/memstick.c b/drivers/memstick/core/memstick.c
index bba467fe4bce..5e0e960df456 100644
--- a/drivers/memstick/core/memstick.c
+++ b/drivers/memstick/core/memstick.c
@@ -271,7 +271,7 @@ void memstick_init_req_sg(struct memstick_request *mrq, unsigned char tpc,
 		mrq->data_dir = READ;
 
 	mrq->sg = *sg;
-	mrq->io_type = MEMSTICK_IO_SG;
+	mrq->long_data = 1;
 
 	if (tpc == MS_TPC_SET_CMD || tpc == MS_TPC_EX_SET_CMD)
 		mrq->need_card_int = 1;
@@ -306,7 +306,7 @@ void memstick_init_req(struct memstick_request *mrq, unsigned char tpc,
 	if (mrq->data_dir == WRITE)
 		memcpy(mrq->data, buf, mrq->data_len);
 
-	mrq->io_type = MEMSTICK_IO_VAL;
+	mrq->long_data = 0;
 
 	if (tpc == MS_TPC_SET_CMD || tpc == MS_TPC_EX_SET_CMD)
 		mrq->need_card_int = 1;
diff --git a/drivers/memstick/core/mspro_block.c b/drivers/memstick/core/mspro_block.c
index 423ad8cf4bb9..214211c8ac9a 100644
--- a/drivers/memstick/core/mspro_block.c
+++ b/drivers/memstick/core/mspro_block.c
@@ -629,7 +629,7 @@ static void mspro_block_process_request(struct memstick_dev *card,
 			param.system = msb->system;
 			param.data_count = cpu_to_be16(page_count);
 			param.data_address = cpu_to_be32((uint32_t)t_sec);
-			param.cmd_param = 0;
+			param.tpc_param = 0;
 
 			msb->data_dir = rq_data_dir(req);
 			msb->transfer_cmd = msb->data_dir == READ
@@ -761,7 +761,7 @@ static int mspro_block_switch_to_parallel(struct memstick_dev *card)
 		.system = 0,
 		.data_count = 0,
 		.data_address = 0,
-		.cmd_param = 0
+		.tpc_param = 0
 	};
 
 	card->next_request = h_mspro_block_req_init;
@@ -773,8 +773,8 @@ static int mspro_block_switch_to_parallel(struct memstick_dev *card)
 	if (card->current_mrq.error)
 		return card->current_mrq.error;
 
-	msb->system = 0;
-	host->set_param(host, MEMSTICK_INTERFACE, MEMSTICK_PARALLEL);
+	msb->system = MEMSTICK_SYS_PAR4;
+	host->set_param(host, MEMSTICK_INTERFACE, MEMSTICK_PAR4);
 
 	card->next_request = h_mspro_block_req_init;
 	msb->mrq_handler = h_mspro_block_default;
@@ -802,7 +802,7 @@ static int mspro_block_read_attributes(struct memstick_dev *card)
 		.system = msb->system,
 		.data_count = cpu_to_be16(1),
 		.data_address = 0,
-		.cmd_param = 0
+		.tpc_param = 0
 	};
 	struct mspro_attribute *attr = NULL;
 	struct mspro_sys_attr *s_attr = NULL;
@@ -922,7 +922,7 @@ static int mspro_block_read_attributes(struct memstick_dev *card)
 		param.system = msb->system;
 		param.data_count = cpu_to_be16((rc / msb->page_size) + 1);
 		param.data_address = cpu_to_be32(addr / msb->page_size);
-		param.cmd_param = 0;
+		param.tpc_param = 0;
 
 		sg_init_one(&msb->req_sg[0], buffer,
 			    be16_to_cpu(param.data_count) * msb->page_size);
@@ -964,7 +964,7 @@ static int mspro_block_init_card(struct memstick_dev *card)
 	struct memstick_host *host = card->host;
 	int rc = 0;
 
-	msb->system = 0x80;
+	msb->system = MEMSTICK_SYS_SERIAL;
 	card->reg_addr.r_offset = offsetof(struct mspro_register, status);
 	card->reg_addr.r_length = sizeof(struct ms_status_register);
 	card->reg_addr.w_offset = offsetof(struct mspro_register, param);
@@ -973,7 +973,7 @@ static int mspro_block_init_card(struct memstick_dev *card)
 	if (memstick_set_rw_addr(card))
 		return -EIO;
 
-	if (host->caps & MEMSTICK_CAP_PARALLEL) {
+	if (host->caps & MEMSTICK_CAP_PAR4) {
 		if (mspro_block_switch_to_parallel(card))
 			printk(KERN_WARNING "%s: could not switch to "
 			       "parallel interface\n", card->dev.bus_id);
diff --git a/drivers/memstick/host/tifm_ms.c b/drivers/memstick/host/tifm_ms.c
index 4fb24215bd95..5b5bd61b3a4a 100644
--- a/drivers/memstick/host/tifm_ms.c
+++ b/drivers/memstick/host/tifm_ms.c
@@ -209,7 +209,7 @@ static int tifm_ms_issue_cmd(struct tifm_ms *host)
 
 	host->cmd_flags = 0;
 
-	if (host->req->io_type == MEMSTICK_IO_SG) {
+	if (host->req->long_data) {
 		if (!host->no_dma) {
 			if (1 != tifm_map_sg(sock, &host->req->sg, 1,
 					     host->req->data_dir == READ
@@ -248,7 +248,7 @@ static int tifm_ms_issue_cmd(struct tifm_ms *host)
 		cmd_mask = readl(sock->addr + SOCK_MS_SYSTEM);
 		cmd_mask |= TIFM_MS_SYS_DATA | TIFM_MS_SYS_NOT_RDY;
 		writel(cmd_mask, sock->addr + SOCK_MS_SYSTEM);
-	} else if (host->req->io_type == MEMSTICK_IO_VAL) {
+	} else {
 		data = host->req->data;
 		data_len = host->req->data_len;
 
@@ -294,8 +294,7 @@ static int tifm_ms_issue_cmd(struct tifm_ms *host)
 		cmd_mask |= TIFM_MS_SYS_NOT_RDY;
 		dev_dbg(&sock->dev, "mask %x\n", cmd_mask);
 		writel(cmd_mask, sock->addr + SOCK_MS_SYSTEM);
-	} else
-		BUG();
+	}
 
 	mod_timer(&host->timer, jiffies + host->timeout_jiffies);
 	writel(TIFM_CTRL_LED | readl(sock->addr + SOCK_CONTROL),
@@ -319,13 +318,13 @@ static void tifm_ms_complete_cmd(struct tifm_ms *host)
 	int rc;
 
 	del_timer(&host->timer);
-	if (host->req->io_type == MEMSTICK_IO_SG) {
+	if (host->req->long_data) {
 		if (!host->no_dma)
 			tifm_unmap_sg(sock, &host->req->sg, 1,
 				      host->req->data_dir == READ
 				      ? PCI_DMA_FROMDEVICE
 				      : PCI_DMA_TODEVICE);
-	} else if (host->req->io_type == MEMSTICK_IO_VAL) {
+	} else {
 		writel(~TIFM_MS_SYS_DATA & readl(sock->addr + SOCK_MS_SYSTEM),
 		       sock->addr + SOCK_MS_SYSTEM);
 
@@ -365,7 +364,7 @@ static int tifm_ms_check_status(struct tifm_ms *host)
 	if (!host->req->error) {
 		if (!(host->cmd_flags & CMD_READY))
 			return 1;
-		if ((host->req->io_type == MEMSTICK_IO_SG)
+		if (host->req->long_data
 		    && !(host->cmd_flags & FIFO_READY))
 			return 1;
 		if (host->req->need_card_int
@@ -505,7 +504,7 @@ static void tifm_ms_set_param(struct memstick_host *msh,
 			writel((~TIFM_CTRL_FAST_CLK)
 			       & readl(sock->addr + SOCK_CONTROL),
 			       sock->addr + SOCK_CONTROL);
-		} else if (value == MEMSTICK_PARALLEL) {
+		} else if (value == MEMSTICK_PAR4) {
 			host->mode_mask = 0;
 			writel(TIFM_CTRL_FAST_CLK
 			       | readl(sock->addr + SOCK_CONTROL),
@@ -542,7 +541,7 @@ static int tifm_ms_initialize_host(struct tifm_ms *host)
 	writel(0x0200 | TIFM_MS_SYS_NOT_RDY, sock->addr + SOCK_MS_SYSTEM);
 	writel(0xffffffff, sock->addr + SOCK_MS_STATUS);
 	if (tifm_has_ms_pif(sock))
-		msh->caps |= MEMSTICK_CAP_PARALLEL;
+		msh->caps |= MEMSTICK_CAP_PAR4;
 
 	return 0;
 }
@@ -601,7 +600,7 @@ static void tifm_ms_remove(struct tifm_dev *sock)
 		writel(TIFM_FIFO_INT_SETALL,
 		       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_CLEAR);
 		writel(TIFM_DMA_RESET, sock->addr + SOCK_DMA_CONTROL);
-		if ((host->req->io_type == MEMSTICK_IO_SG) && !host->no_dma)
+		if (host->req->long_data && !host->no_dma)
 			tifm_unmap_sg(sock, &host->req->sg, 1,
 				      host->req->data_dir == READ
 				      ? PCI_DMA_TODEVICE
diff --git a/include/linux/memstick.h b/include/linux/memstick.h
index 334d059d6794..c104e722de06 100644
--- a/include/linux/memstick.h
+++ b/include/linux/memstick.h
@@ -22,6 +22,8 @@ struct ms_status_register {
 	unsigned char reserved;
 	unsigned char interrupt;
 #define MEMSTICK_INT_CMDNAK             0x0001
+#define MEMSTICK_INT_IOREQ              0x0008
+#define MEMSTICK_INT_IOBREQ             0x0010
 #define MEMSTICK_INT_BREQ               0x0020
 #define MEMSTICK_INT_ERR                0x0040
 #define MEMSTICK_INT_CED                0x0080
@@ -47,13 +49,17 @@ struct ms_status_register {
 
 struct ms_id_register {
 	unsigned char type;
-	unsigned char reserved;
+	unsigned char if_mode;
 	unsigned char category;
 	unsigned char class;
 } __attribute__((packed));
 
 struct ms_param_register {
 	unsigned char system;
+#define MEMSTICK_SYS_ATEN 0xc0
+#define MEMSTICK_SYS_BAMD 0x80
+#define MEMSTICK_SYS_PAM  0x08
+
 	unsigned char block_address_msb;
 	unsigned short block_address;
 	unsigned char cp;
@@ -90,16 +96,48 @@ struct ms_register {
 
 struct mspro_param_register {
 	unsigned char  system;
+#define MEMSTICK_SYS_SERIAL 0x80
+#define MEMSTICK_SYS_PAR4   0x00
+#define MEMSTICK_SYS_PAR8   0x40
+
+	unsigned short data_count;
+	unsigned int   data_address;
+	unsigned char  tpc_param;
+} __attribute__((packed));
+
+struct mspro_io_info_register {
+	unsigned char version;
+	unsigned char io_category;
+	unsigned char current_req;
+	unsigned char card_opt_info;
+	unsigned char rdy_wait_time;
+} __attribute__((packed));
+
+struct mspro_io_func_register {
+	unsigned char func_enable;
+	unsigned char func_select;
+	unsigned char func_intmask;
+	unsigned char transfer_mode;
+} __attribute__((packed));
+
+struct mspro_io_cmd_register {
+	unsigned short tpc_param;
 	unsigned short data_count;
 	unsigned int   data_address;
-	unsigned char  cmd_param;
 } __attribute__((packed));
 
 struct mspro_register {
-	struct ms_status_register    status;
-	struct ms_id_register        id;
-	unsigned char                reserved[8];
-	struct mspro_param_register  param;
+	struct ms_status_register     status;
+	struct ms_id_register         id;
+	unsigned char                 reserved0[8];
+	struct mspro_param_register   param;
+	unsigned char                 reserved1[8];
+	struct mspro_io_info_register io_info;
+	struct mspro_io_func_register io_func;
+	unsigned char                 reserved2[7];
+	struct mspro_io_cmd_register  io_cmd;
+	unsigned char                 io_int;
+	unsigned char                 io_int_func;
 } __attribute__((packed));
 
 struct ms_register_addr {
@@ -110,49 +148,55 @@ struct ms_register_addr {
 } __attribute__((packed));
 
 enum {
+	MS_TPC_READ_MG_STATUS   = 0x01,
 	MS_TPC_READ_LONG_DATA   = 0x02,
 	MS_TPC_READ_SHORT_DATA  = 0x03,
+	MS_TPC_READ_MG_DATA     = 0x03,
 	MS_TPC_READ_REG         = 0x04,
-	MS_TPC_READ_IO_DATA     = 0x05, /* unverified */
+	MS_TPC_READ_QUAD_DATA   = 0x05,
+	MS_TPC_READ_IO_DATA     = 0x05,
 	MS_TPC_GET_INT          = 0x07,
 	MS_TPC_SET_RW_REG_ADRS  = 0x08,
 	MS_TPC_EX_SET_CMD       = 0x09,
-	MS_TPC_WRITE_IO_DATA    = 0x0a, /* unverified */
+	MS_TPC_WRITE_QUAD_DATA  = 0x0a,
+	MS_TPC_WRITE_IO_DATA    = 0x0a,
 	MS_TPC_WRITE_REG        = 0x0b,
 	MS_TPC_WRITE_SHORT_DATA = 0x0c,
+	MS_TPC_WRITE_MG_DATA    = 0x0c,
 	MS_TPC_WRITE_LONG_DATA  = 0x0d,
 	MS_TPC_SET_CMD          = 0x0e
 };
 
 enum {
-	MS_CMD_BLOCK_END     = 0x33,
-	MS_CMD_RESET         = 0x3c,
-	MS_CMD_BLOCK_WRITE   = 0x55,
-	MS_CMD_SLEEP         = 0x5a,
-	MS_CMD_BLOCK_ERASE   = 0x99,
-	MS_CMD_BLOCK_READ    = 0xaa,
-	MS_CMD_CLEAR_BUF     = 0xc3,
-	MS_CMD_FLASH_STOP    = 0xcc,
-	MSPRO_CMD_FORMAT     = 0x10,
-	MSPRO_CMD_SLEEP      = 0x11,
-	MSPRO_CMD_READ_DATA  = 0x20,
-	MSPRO_CMD_WRITE_DATA = 0x21,
-	MSPRO_CMD_READ_ATRB  = 0x24,
-	MSPRO_CMD_STOP       = 0x25,
-	MSPRO_CMD_ERASE      = 0x26,
-	MSPRO_CMD_SET_IBA    = 0x46,
-	MSPRO_CMD_SET_IBD    = 0x47
-/*
-	MSPRO_CMD_RESET
-	MSPRO_CMD_WAKEUP
-	MSPRO_CMD_IN_IO_DATA
-	MSPRO_CMD_OUT_IO_DATA
-	MSPRO_CMD_READ_IO_ATRB
-	MSPRO_CMD_IN_IO_FIFO
-	MSPRO_CMD_OUT_IO_FIFO
-	MSPRO_CMD_IN_IOM
-	MSPRO_CMD_OUT_IOM
-*/
+	MS_CMD_BLOCK_END       = 0x33,
+	MS_CMD_RESET           = 0x3c,
+	MS_CMD_BLOCK_WRITE     = 0x55,
+	MS_CMD_SLEEP           = 0x5a,
+	MS_CMD_BLOCK_ERASE     = 0x99,
+	MS_CMD_BLOCK_READ      = 0xaa,
+	MS_CMD_CLEAR_BUF       = 0xc3,
+	MS_CMD_FLASH_STOP      = 0xcc,
+	MS_CMD_LOAD_ID         = 0x60,
+	MS_CMD_CMP_ICV         = 0x7f,
+	MSPRO_CMD_FORMAT       = 0x10,
+	MSPRO_CMD_SLEEP        = 0x11,
+	MSPRO_CMD_WAKEUP       = 0x12,
+	MSPRO_CMD_READ_DATA    = 0x20,
+	MSPRO_CMD_WRITE_DATA   = 0x21,
+	MSPRO_CMD_READ_ATRB    = 0x24,
+	MSPRO_CMD_STOP         = 0x25,
+	MSPRO_CMD_ERASE        = 0x26,
+	MSPRO_CMD_READ_QUAD    = 0x27,
+	MSPRO_CMD_WRITE_QUAD   = 0x28,
+	MSPRO_CMD_SET_IBD      = 0x46,
+	MSPRO_CMD_GET_IBD      = 0x47,
+	MSPRO_CMD_IN_IO_DATA   = 0xb0,
+	MSPRO_CMD_OUT_IO_DATA  = 0xb1,
+	MSPRO_CMD_READ_IO_ATRB = 0xb2,
+	MSPRO_CMD_IN_IO_FIFO   = 0xb3,
+	MSPRO_CMD_OUT_IO_FIFO  = 0xb4,
+	MSPRO_CMD_IN_IOM       = 0xb5,
+	MSPRO_CMD_OUT_IOM      = 0xb6,
 };
 
 /*** Driver structures and functions ***/
@@ -165,7 +209,8 @@ enum memstick_param { MEMSTICK_POWER = 1, MEMSTICK_INTERFACE };
 #define MEMSTICK_POWER_ON  1
 
 #define MEMSTICK_SERIAL   0
-#define MEMSTICK_PARALLEL 1
+#define MEMSTICK_PAR4     1
+#define MEMSTICK_PAR8     2
 
 struct memstick_host;
 struct memstick_driver;
@@ -195,11 +240,7 @@ struct memstick_request {
 	unsigned char data_dir:1,
 		      need_card_int:1,
 		      get_int_reg:1,
-		      io_type:2;
-#define               MEMSTICK_IO_NONE 0
-#define               MEMSTICK_IO_VAL  1
-#define               MEMSTICK_IO_SG   2
-
+		      long_data:1;
 	unsigned char int_reg;
 	int           error;
 	union {
@@ -231,8 +272,9 @@ struct memstick_host {
 	struct mutex        lock;
 	unsigned int        id;
 	unsigned int        caps;
-#define MEMSTICK_CAP_PARALLEL      1
-#define MEMSTICK_CAP_AUTO_GET_INT  2
+#define MEMSTICK_CAP_AUTO_GET_INT  1
+#define MEMSTICK_CAP_PAR4          2
+#define MEMSTICK_CAP_PAR8          4
 
 	struct work_struct  media_checker;
 	struct class_device cdev;
-- 
cgit v1.2.3


From d114ad54ffb020dc781b6159c1c2f391c6ec418f Mon Sep 17 00:00:00 2001
From: Alex Dubov <oakad@yahoo.com>
Date: Mon, 10 Mar 2008 11:43:38 -0700
Subject: memstick: add memstick_suspend/resume_host methods

Bus driver may need to be informed that host is being suspended/resumed.

Signed-off-by: Alex Dubov <oakad@yahoo.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/memstick/core/memstick.c | 25 +++++++++++++++++++++++++
 drivers/memstick/host/tifm_ms.c  |  8 ++++----
 include/linux/memstick.h         |  2 ++
 3 files changed, 31 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/memstick/core/memstick.c b/drivers/memstick/core/memstick.c
index 5e0e960df456..3c97bac4e47b 100644
--- a/drivers/memstick/core/memstick.c
+++ b/drivers/memstick/core/memstick.c
@@ -561,6 +561,31 @@ void memstick_free_host(struct memstick_host *host)
 }
 EXPORT_SYMBOL(memstick_free_host);
 
+/**
+ * memstick_suspend_host - notify bus driver of host suspension
+ * @host - host to use
+ */
+void memstick_suspend_host(struct memstick_host *host)
+{
+	mutex_lock(&host->lock);
+	host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_OFF);
+	mutex_unlock(&host->lock);
+}
+EXPORT_SYMBOL(memstick_suspend_host);
+
+/**
+ * memstick_resume_host - notify bus driver of host resumption
+ * @host - host to use
+ */
+void memstick_resume_host(struct memstick_host *host)
+{
+	mutex_lock(&host->lock);
+	host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_ON);
+	mutex_unlock(&host->lock);
+	memstick_detect_change(host);
+}
+EXPORT_SYMBOL(memstick_resume_host);
+
 int memstick_register_driver(struct memstick_driver *drv)
 {
 	drv->driver.bus = &memstick_bus_type;
diff --git a/drivers/memstick/host/tifm_ms.c b/drivers/memstick/host/tifm_ms.c
index 5b5bd61b3a4a..8b1c102fc317 100644
--- a/drivers/memstick/host/tifm_ms.c
+++ b/drivers/memstick/host/tifm_ms.c
@@ -627,17 +627,17 @@ static void tifm_ms_remove(struct tifm_dev *sock)
 
 static int tifm_ms_suspend(struct tifm_dev *sock, pm_message_t state)
 {
+	struct memstick_host *msh = tifm_get_drvdata(sock);
+
+	memstick_suspend_host(msh);
 	return 0;
 }
 
 static int tifm_ms_resume(struct tifm_dev *sock)
 {
 	struct memstick_host *msh = tifm_get_drvdata(sock);
-	struct tifm_ms *host = memstick_priv(msh);
-
-	tifm_ms_initialize_host(host);
-	memstick_detect_change(msh);
 
+	memstick_resume_host(msh);
 	return 0;
 }
 
diff --git a/include/linux/memstick.h b/include/linux/memstick.h
index c104e722de06..b7ee25888836 100644
--- a/include/linux/memstick.h
+++ b/include/linux/memstick.h
@@ -312,6 +312,8 @@ int memstick_add_host(struct memstick_host *host);
 void memstick_remove_host(struct memstick_host *host);
 void memstick_free_host(struct memstick_host *host);
 void memstick_detect_change(struct memstick_host *host);
+void memstick_suspend_host(struct memstick_host *host);
+void memstick_resume_host(struct memstick_host *host);
 
 void memstick_init_req_sg(struct memstick_request *mrq, unsigned char tpc,
 			  struct scatterlist *sg);
-- 
cgit v1.2.3


From 92b22d935fed1e4d88b9b6f9a674ab2a4272ee78 Mon Sep 17 00:00:00 2001
From: Alex Dubov <oakad@yahoo.com>
Date: Mon, 10 Mar 2008 11:43:40 -0700
Subject: tifm: fix the MemoryStick host fifo handling code

Additional input received from JMicron on MemoryStick host interfaces showed
that some assumtions in fifo handling code were incorrect.  This patch also
fixes data corruption used to occure during PIO transfers.

Signed-off-by: Alex Dubov <oakad@yahoo.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/memstick/host/tifm_ms.c | 524 ++++++++++++++++++++--------------------
 include/linux/tifm.h            |   2 +-
 2 files changed, 264 insertions(+), 262 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/memstick/host/tifm_ms.c b/drivers/memstick/host/tifm_ms.c
index c62e709ca771..b88f5b30efbf 100644
--- a/drivers/memstick/host/tifm_ms.c
+++ b/drivers/memstick/host/tifm_ms.c
@@ -24,275 +24,289 @@
 static int no_dma;
 module_param(no_dma, bool, 0644);
 
-#define TIFM_MS_TIMEOUT      0x00100
-#define TIFM_MS_BADCRC       0x00200
-#define TIFM_MS_EOTPC        0x01000
-#define TIFM_MS_INT          0x02000
-
-/* The meaning of the bit majority in this constant is unknown. */
-#define TIFM_MS_SERIAL       0x04010
+/*
+ * Some control bits of TIFM appear to conform to Sony's reference design,
+ * so I'm just assuming they all are.
+ */
 
-#define TIFM_MS_SYS_LATCH    0x00100
-#define TIFM_MS_SYS_NOT_RDY  0x00800
-#define TIFM_MS_SYS_DATA     0x10000
+#define TIFM_MS_STAT_DRQ     0x04000
+#define TIFM_MS_STAT_MSINT   0x02000
+#define TIFM_MS_STAT_RDY     0x01000
+#define TIFM_MS_STAT_CRC     0x00200
+#define TIFM_MS_STAT_TOE     0x00100
+#define TIFM_MS_STAT_EMP     0x00020
+#define TIFM_MS_STAT_FUL     0x00010
+#define TIFM_MS_STAT_CED     0x00008
+#define TIFM_MS_STAT_ERR     0x00004
+#define TIFM_MS_STAT_BRQ     0x00002
+#define TIFM_MS_STAT_CNK     0x00001
+
+#define TIFM_MS_SYS_DMA      0x10000
+#define TIFM_MS_SYS_RESET    0x08000
+#define TIFM_MS_SYS_SRAC     0x04000
+#define TIFM_MS_SYS_INTEN    0x02000
+#define TIFM_MS_SYS_NOCRC    0x01000
+#define TIFM_MS_SYS_INTCLR   0x00800
+#define TIFM_MS_SYS_MSIEN    0x00400
+#define TIFM_MS_SYS_FCLR     0x00200
+#define TIFM_MS_SYS_FDIR     0x00100
+#define TIFM_MS_SYS_DAM      0x00080
+#define TIFM_MS_SYS_DRM      0x00040
+#define TIFM_MS_SYS_DRQSL    0x00020
+#define TIFM_MS_SYS_REI      0x00010
+#define TIFM_MS_SYS_REO      0x00008
+#define TIFM_MS_SYS_BSY_MASK 0x00007
+
+#define TIFM_MS_SYS_FIFO     (TIFM_MS_SYS_INTEN | TIFM_MS_SYS_MSIEN \
+			      | TIFM_MS_SYS_FCLR | TIFM_MS_SYS_BSY_MASK)
 
 /* Hardware flags */
 enum {
-	CMD_READY  = 0x0001,
-	FIFO_READY = 0x0002,
-	CARD_READY = 0x0004,
-	DATA_CARRY = 0x0008
+	CMD_READY  = 0x01,
+	FIFO_READY = 0x02,
+	CARD_INT   = 0x04
 };
 
 struct tifm_ms {
 	struct tifm_dev         *dev;
-	unsigned short          eject:1,
-				no_dma:1;
-	unsigned short          cmd_flags;
+	struct timer_list       timer;
+	struct memstick_request *req;
 	unsigned int            mode_mask;
 	unsigned int            block_pos;
 	unsigned long           timeout_jiffies;
-
-	struct timer_list       timer;
-	struct memstick_request *req;
+	unsigned char           eject:1,
+				use_dma:1;
+	unsigned char           cmd_flags;
+	unsigned char           io_pos;
 	unsigned int            io_word;
 };
 
-static void tifm_ms_read_fifo(struct tifm_ms *host, unsigned int fifo_offset,
-			      struct page *pg, unsigned int page_off,
-			      unsigned int length)
+static unsigned int tifm_ms_read_data(struct tifm_ms *host,
+				      unsigned char *buf, unsigned int length)
 {
 	struct tifm_dev *sock = host->dev;
-	unsigned int cnt = 0, off = 0;
-	unsigned char *buf = kmap_atomic(pg, KM_BIO_DST_IRQ) + page_off;
+	unsigned int off = 0;
+
+	while (host->io_pos && length) {
+		buf[off++] = host->io_word & 0xff;
+		host->io_word >>= 8;
+		length--;
+		host->io_pos--;
+	}
+
+	if (!length)
+		return off;
+
+	while (!(TIFM_MS_STAT_EMP & readl(sock->addr + SOCK_MS_STATUS))) {
+		if (length < 4)
+			break;
+		*(unsigned int *)(buf + off) = __raw_readl(sock->addr
+							   + SOCK_MS_DATA);
+		length -= 4;
+		off += 4;
+	}
 
-	if (host->cmd_flags & DATA_CARRY) {
-		while ((fifo_offset & 3) && length) {
+	if (length
+	    && !(TIFM_MS_STAT_EMP & readl(sock->addr + SOCK_MS_STATUS))) {
+		host->io_word = readl(sock->addr + SOCK_MS_DATA);
+		for (host->io_pos = 4; host->io_pos; --host->io_pos) {
 			buf[off++] = host->io_word & 0xff;
 			host->io_word >>= 8;
 			length--;
-			fifo_offset++;
+			if (!length)
+				break;
 		}
-		if (!(fifo_offset & 3))
-			host->cmd_flags &= ~DATA_CARRY;
-		if (!length)
-			return;
 	}
 
-	do {
-		host->io_word = readl(sock->addr + SOCK_FIFO_ACCESS
-				      + fifo_offset);
-		cnt = 4;
-		while (length && cnt) {
-			buf[off++] = (host->io_word >> 8) & 0xff;
-			cnt--;
-			length--;
-		}
-		fifo_offset += 4 - cnt;
-	} while (length);
-
-	if (cnt)
-		host->cmd_flags |= DATA_CARRY;
-
-	kunmap_atomic(buf - page_off, KM_BIO_DST_IRQ);
+	return off;
 }
 
-static void tifm_ms_write_fifo(struct tifm_ms *host, unsigned int fifo_offset,
-			       struct page *pg, unsigned int page_off,
-			       unsigned int length)
+static unsigned int tifm_ms_write_data(struct tifm_ms *host,
+				       unsigned char *buf, unsigned int length)
 {
 	struct tifm_dev *sock = host->dev;
-	unsigned int cnt = 0, off = 0;
-	unsigned char *buf = kmap_atomic(pg, KM_BIO_SRC_IRQ) + page_off;
+	unsigned int off = 0;
 
-	if (host->cmd_flags & DATA_CARRY) {
-		while (fifo_offset & 3) {
-			host->io_word |= buf[off++] << (8 * (fifo_offset & 3));
+	if (host->io_pos) {
+		while (host->io_pos < 4 && length) {
+			host->io_word |=  buf[off++] << (host->io_pos * 8);
+			host->io_pos++;
 			length--;
-			fifo_offset++;
-		}
-		if (!(fifo_offset & 3)) {
-			writel(host->io_word, sock->addr + SOCK_FIFO_ACCESS
-			       + fifo_offset - 4);
-
-			host->cmd_flags &= ~DATA_CARRY;
 		}
-		if (!length)
-			return;
 	}
 
-	do {
-		cnt = 4;
+	if (host->io_pos == 4
+	    && !(TIFM_MS_STAT_FUL & readl(sock->addr + SOCK_MS_STATUS))) {
+		writel(TIFM_MS_SYS_FDIR | readl(sock->addr + SOCK_MS_SYSTEM),
+		       sock->addr + SOCK_MS_SYSTEM);
+		writel(host->io_word, sock->addr + SOCK_MS_DATA);
+		host->io_pos = 0;
 		host->io_word = 0;
-		while (length && cnt) {
-			host->io_word |= buf[off++] << (4 - cnt);
-			cnt--;
-			length--;
-		}
-		fifo_offset += 4 - cnt;
-		if (!cnt)
-			writel(host->io_word, sock->addr + SOCK_FIFO_ACCESS
-					      + fifo_offset - 4);
-
-	} while (length);
-
-	if (cnt)
-		host->cmd_flags |= DATA_CARRY;
+	} else if (host->io_pos) {
+		return off;
+	}
 
-	kunmap_atomic(buf - page_off, KM_BIO_SRC_IRQ);
-}
+	if (!length)
+		return off;
 
-static void tifm_ms_move_block(struct tifm_ms *host, unsigned int length)
-{
-	unsigned int t_size;
-	unsigned int off = host->req->sg.offset + host->block_pos;
-	unsigned int p_off, p_cnt;
-	struct page *pg;
-	unsigned long flags;
+	while (!(TIFM_MS_STAT_FUL & readl(sock->addr + SOCK_MS_STATUS))) {
+		if (length < 4)
+			break;
+		writel(TIFM_MS_SYS_FDIR | readl(sock->addr + SOCK_MS_SYSTEM),
+		       sock->addr + SOCK_MS_SYSTEM);
+		__raw_writel(*(unsigned int *)(buf + off),
+			     sock->addr + SOCK_MS_DATA);
+		length -= 4;
+		off += 4;
+	}
 
-	dev_dbg(&host->dev->dev, "moving block\n");
-	local_irq_save(flags);
-	t_size = length;
-	while (t_size) {
-		pg = nth_page(sg_page(&host->req->sg), off >> PAGE_SHIFT);
-		p_off = offset_in_page(off);
-		p_cnt = PAGE_SIZE - p_off;
-		p_cnt = min(p_cnt, t_size);
+	switch (length) {
+	case 3:
+		host->io_word |= buf[off + 2] << 16;
+		host->io_pos++;
+	case 2:
+		host->io_word |= buf[off + 1] << 8;
+		host->io_pos++;
+	case 1:
+		host->io_word |= buf[off];
+		host->io_pos++;
+	}
 
-		if (host->req->data_dir == WRITE)
-			tifm_ms_write_fifo(host, length - t_size,
-					   pg, p_off, p_cnt);
-		else
-			tifm_ms_read_fifo(host, length - t_size,
-					  pg, p_off, p_cnt);
+	off += host->io_pos;
 
-		t_size -= p_cnt;
-	}
-	local_irq_restore(flags);
+	return off;
 }
 
-static int tifm_ms_transfer_data(struct tifm_ms *host, int skip)
+static unsigned int tifm_ms_transfer_data(struct tifm_ms *host)
 {
 	struct tifm_dev *sock = host->dev;
-	unsigned int length = host->req->sg.length - host->block_pos;
+	unsigned int length;
+	unsigned int off;
+	unsigned int t_size, p_off, p_cnt;
+	unsigned char *buf;
+	struct page *pg;
+	unsigned long flags = 0;
 
-	if (!length)
-		return 1;
+	if (host->req->long_data) {
+		length = host->req->sg.length - host->block_pos;
+		off = host->req->sg.offset + host->block_pos;
+	} else {
+		length = host->req->data_len - host->block_pos;
+		off = 0;
+	}
+	dev_dbg(&sock->dev, "fifo data transfer, %d, %d\n", length,
+		host->block_pos);
+
+	while (length) {
+		if (host->req->long_data) {
+			pg = nth_page(sg_page(&host->req->sg),
+				      off >> PAGE_SHIFT);
+			p_off = offset_in_page(off);
+			p_cnt = PAGE_SIZE - p_off;
+			p_cnt = min(p_cnt, length);
+
+			local_irq_save(flags);
+			buf = kmap_atomic(pg, KM_BIO_SRC_IRQ) + p_off;
+		} else {
+			buf = host->req->data + host->block_pos;
+			p_cnt = host->req->data_len - host->block_pos;
+		}
 
-	if (length > TIFM_FIFO_SIZE)
-		length = TIFM_FIFO_SIZE;
+		t_size = host->req->data_dir == WRITE
+			 ? tifm_ms_write_data(host, buf, p_cnt)
+			 : tifm_ms_read_data(host, buf, p_cnt);
 
-	if (!skip) {
-		tifm_ms_move_block(host, length);
-		host->block_pos += length;
-	}
+		if (host->req->long_data) {
+			kunmap_atomic(buf - p_off, KM_BIO_SRC_IRQ);
+			local_irq_restore(flags);
+		}
 
-	if ((host->req->data_dir == READ)
-	    && (host->block_pos == host->req->sg.length))
-		return 1;
+		if (!t_size)
+			break;
+		host->block_pos += t_size;
+		length -= t_size;
+		off += t_size;
+	}
 
-	writel(ilog2(length) - 2, sock->addr + SOCK_FIFO_PAGE_SIZE);
-	if (host->req->data_dir == WRITE)
-		writel((1 << 8) | TIFM_DMA_TX, sock->addr + SOCK_DMA_CONTROL);
-	else
-		writel((1 << 8), sock->addr + SOCK_DMA_CONTROL);
+	dev_dbg(&sock->dev, "fifo data transfer, %d remaining\n", length);
+	if (!length && (host->req->data_dir == WRITE)) {
+		if (host->io_pos) {
+			writel(TIFM_MS_SYS_FDIR
+			       | readl(sock->addr + SOCK_MS_SYSTEM),
+			       sock->addr + SOCK_MS_SYSTEM);
+			writel(host->io_word, sock->addr + SOCK_MS_DATA);
+		}
+		writel(TIFM_MS_SYS_FDIR
+		       | readl(sock->addr + SOCK_MS_SYSTEM),
+		       sock->addr + SOCK_MS_SYSTEM);
+		writel(0, sock->addr + SOCK_MS_DATA);
+	} else {
+		readl(sock->addr + SOCK_MS_DATA);
+	}
 
-	return 0;
+	return length;
 }
 
 static int tifm_ms_issue_cmd(struct tifm_ms *host)
 {
 	struct tifm_dev *sock = host->dev;
 	unsigned char *data;
-	unsigned int data_len = 0, cmd = 0, cmd_mask = 0, cnt, tval = 0;
+	unsigned int data_len, cmd, sys_param;
 
+	host->cmd_flags = 0;
+	host->block_pos = 0;
+	host->io_pos = 0;
+	host->io_word = 0;
 	host->cmd_flags = 0;
 
-	if (host->req->long_data) {
-		if (!host->no_dma) {
-			if (1 != tifm_map_sg(sock, &host->req->sg, 1,
-					     host->req->data_dir == READ
-					     ? PCI_DMA_FROMDEVICE
-					     : PCI_DMA_TODEVICE)) {
-				host->req->error = -ENOMEM;
-				return host->req->error;
-			}
-			data_len = sg_dma_len(&host->req->sg);
-		} else
-			data_len = host->req->sg.length;
-
-		writel(TIFM_FIFO_INT_SETALL,
-		       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_CLEAR);
-		writel(TIFM_FIFO_ENABLE,
-		       sock->addr + SOCK_FIFO_CONTROL);
-		writel(TIFM_FIFO_INTMASK,
-		       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_SET);
+	data = host->req->data;
 
-		if (!host->no_dma) {
-			writel(ilog2(data_len) - 2,
-			       sock->addr + SOCK_FIFO_PAGE_SIZE);
-			writel(sg_dma_address(&host->req->sg),
-			       sock->addr + SOCK_DMA_ADDRESS);
-			if (host->req->data_dir == WRITE)
-				writel((1 << 8) | TIFM_DMA_TX | TIFM_DMA_EN,
-				       sock->addr + SOCK_DMA_CONTROL);
-			else
-				writel((1 << 8) | TIFM_DMA_EN,
-				       sock->addr + SOCK_DMA_CONTROL);
-		} else {
-			tifm_ms_transfer_data(host,
-					      host->req->data_dir == READ);
-		}
+	host->use_dma = !no_dma;
 
-		cmd_mask = readl(sock->addr + SOCK_MS_SYSTEM);
-		cmd_mask |= TIFM_MS_SYS_DATA | TIFM_MS_SYS_NOT_RDY;
-		writel(cmd_mask, sock->addr + SOCK_MS_SYSTEM);
+	if (host->req->long_data) {
+		data_len = host->req->sg.length;
+		if (!is_power_of_2(data_len))
+			host->use_dma = 0;
 	} else {
-		data = host->req->data;
 		data_len = host->req->data_len;
+		host->use_dma = 0;
+	}
 
-		cmd_mask = host->mode_mask | 0x2607; /* unknown constant */
-
-		if (host->req->data_dir == WRITE) {
-			cmd_mask |= TIFM_MS_SYS_LATCH;
-			writel(cmd_mask, sock->addr + SOCK_MS_SYSTEM);
-			for (cnt = 0; (data_len - cnt) >= 4; cnt += 4) {
-				writel(TIFM_MS_SYS_LATCH
-				       | readl(sock->addr + SOCK_MS_SYSTEM),
-				       sock->addr + SOCK_MS_SYSTEM);
-				__raw_writel(*(unsigned int *)(data + cnt),
-					     sock->addr + SOCK_MS_DATA);
-				dev_dbg(&sock->dev, "writing %x\n",
-					*(int *)(data + cnt));
-			}
-			switch (data_len - cnt) {
-			case 3:
-				tval |= data[cnt + 2] << 16;
-			case 2:
-				tval |= data[cnt + 1] << 8;
-			case 1:
-				tval |= data[cnt];
-				writel(TIFM_MS_SYS_LATCH
-				       | readl(sock->addr + SOCK_MS_SYSTEM),
-				       sock->addr + SOCK_MS_SYSTEM);
-				writel(tval, sock->addr + SOCK_MS_DATA);
-				dev_dbg(&sock->dev, "writing %x\n", tval);
-			}
+	writel(TIFM_FIFO_INT_SETALL,
+	       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_CLEAR);
+	writel(TIFM_FIFO_ENABLE,
+	       sock->addr + SOCK_FIFO_CONTROL);
+
+	if (host->use_dma) {
+		if (1 != tifm_map_sg(sock, &host->req->sg, 1,
+				     host->req->data_dir == READ
+				     ? PCI_DMA_FROMDEVICE
+				     : PCI_DMA_TODEVICE)) {
+			host->req->error = -ENOMEM;
+			return host->req->error;
+		}
+		data_len = sg_dma_len(&host->req->sg);
 
-			writel(TIFM_MS_SYS_LATCH
-			       | readl(sock->addr + SOCK_MS_SYSTEM),
-			       sock->addr + SOCK_MS_SYSTEM);
-			writel(0, sock->addr + SOCK_MS_DATA);
-			dev_dbg(&sock->dev, "writing %x\n", 0);
+		writel(ilog2(data_len) - 2,
+		       sock->addr + SOCK_FIFO_PAGE_SIZE);
+		writel(TIFM_FIFO_INTMASK,
+		       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_SET);
+		sys_param = TIFM_DMA_EN | (1 << 8);
+		if (host->req->data_dir == WRITE)
+			sys_param |= TIFM_DMA_TX;
 
-		} else
-			writel(cmd_mask, sock->addr + SOCK_MS_SYSTEM);
+		writel(TIFM_FIFO_INTMASK,
+		       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_SET);
 
-		cmd_mask = readl(sock->addr + SOCK_MS_SYSTEM);
-		cmd_mask &= ~TIFM_MS_SYS_DATA;
-		cmd_mask |= TIFM_MS_SYS_NOT_RDY;
-		dev_dbg(&sock->dev, "mask %x\n", cmd_mask);
-		writel(cmd_mask, sock->addr + SOCK_MS_SYSTEM);
+		writel(sg_dma_address(&host->req->sg),
+		       sock->addr + SOCK_DMA_ADDRESS);
+		writel(sys_param, sock->addr + SOCK_DMA_CONTROL);
+	} else {
+		writel(host->mode_mask | TIFM_MS_SYS_FIFO,
+		       sock->addr + SOCK_MS_SYSTEM);
+
+		writel(TIFM_FIFO_MORE,
+		       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_SET);
 	}
 
 	mod_timer(&host->timer, jiffies + host->timeout_jiffies);
@@ -300,11 +314,21 @@ static int tifm_ms_issue_cmd(struct tifm_ms *host)
 	       sock->addr + SOCK_CONTROL);
 	host->req->error = 0;
 
+	sys_param = readl(sock->addr + SOCK_MS_SYSTEM);
+	sys_param |= TIFM_MS_SYS_INTCLR;
+
+	if (host->use_dma)
+		sys_param |= TIFM_MS_SYS_DMA;
+	else
+		sys_param &= ~TIFM_MS_SYS_DMA;
+
+	writel(sys_param, sock->addr + SOCK_MS_SYSTEM);
+
 	cmd = (host->req->tpc & 0xf) << 12;
 	cmd |= data_len;
 	writel(cmd, sock->addr + SOCK_MS_COMMAND);
 
-	dev_dbg(&sock->dev, "executing TPC %x, %x\n", cmd, cmd_mask);
+	dev_dbg(&sock->dev, "executing TPC %x, %x\n", cmd, sys_param);
 	return 0;
 }
 
@@ -312,47 +336,20 @@ static void tifm_ms_complete_cmd(struct tifm_ms *host)
 {
 	struct tifm_dev *sock = host->dev;
 	struct memstick_host *msh = tifm_get_drvdata(sock);
-	unsigned int tval = 0, data_len;
-	unsigned char *data;
 	int rc;
 
 	del_timer(&host->timer);
-	if (host->req->long_data) {
-		if (!host->no_dma)
-			tifm_unmap_sg(sock, &host->req->sg, 1,
-				      host->req->data_dir == READ
-				      ? PCI_DMA_FROMDEVICE
-				      : PCI_DMA_TODEVICE);
-	} else {
-		writel(~TIFM_MS_SYS_DATA & readl(sock->addr + SOCK_MS_SYSTEM),
-		       sock->addr + SOCK_MS_SYSTEM);
-
-		data = host->req->data;
-		data_len = host->req->data_len;
 
-		if (host->req->data_dir == READ) {
-			for (rc = 0; (data_len - rc) >= 4; rc += 4)
-				*(int *)(data + rc)
-					= __raw_readl(sock->addr
-						      + SOCK_MS_DATA);
-
-			if (data_len - rc)
-				tval = readl(sock->addr + SOCK_MS_DATA);
-			switch (data_len - rc) {
-			case 3:
-				data[rc + 2] = (tval >> 16) & 0xff;
-			case 2:
-				data[rc + 1] = (tval >> 8) & 0xff;
-			case 1:
-				data[rc] = tval & 0xff;
-			}
-			readl(sock->addr + SOCK_MS_DATA);
-		}
-	}
+	if (host->use_dma)
+		tifm_unmap_sg(sock, &host->req->sg, 1,
+			      host->req->data_dir == READ
+			      ? PCI_DMA_FROMDEVICE
+			      : PCI_DMA_TODEVICE);
 
 	writel((~TIFM_CTRL_LED) & readl(sock->addr + SOCK_CONTROL),
 	       sock->addr + SOCK_CONTROL);
 
+	dev_dbg(&sock->dev, "TPC complete\n");
 	do {
 		rc = memstick_next_req(msh, &host->req);
 	} while (!rc && tifm_ms_issue_cmd(host));
@@ -363,11 +360,10 @@ static int tifm_ms_check_status(struct tifm_ms *host)
 	if (!host->req->error) {
 		if (!(host->cmd_flags & CMD_READY))
 			return 1;
-		if (host->req->long_data
-		    && !(host->cmd_flags & FIFO_READY))
+		if (!(host->cmd_flags & FIFO_READY))
 			return 1;
 		if (host->req->need_card_int
-		    && !(host->cmd_flags & CARD_READY))
+		    && !(host->cmd_flags & CARD_INT))
 			return 1;
 	}
 	return 0;
@@ -377,18 +373,24 @@ static int tifm_ms_check_status(struct tifm_ms *host)
 static void tifm_ms_data_event(struct tifm_dev *sock)
 {
 	struct tifm_ms *host;
-	unsigned int fifo_status = 0;
+	unsigned int fifo_status = 0, host_status = 0;
 	int rc = 1;
 
 	spin_lock(&sock->lock);
 	host = memstick_priv((struct memstick_host *)tifm_get_drvdata(sock));
 	fifo_status = readl(sock->addr + SOCK_DMA_FIFO_STATUS);
-	dev_dbg(&sock->dev, "data event: fifo_status %x, flags %x\n",
-		fifo_status, host->cmd_flags);
+	host_status = readl(sock->addr + SOCK_MS_STATUS);
+	dev_dbg(&sock->dev,
+		"data event: fifo_status %x, host_status %x, flags %x\n",
+		fifo_status, host_status, host->cmd_flags);
 
 	if (host->req) {
-		if (fifo_status & TIFM_FIFO_READY) {
-			if (!host->no_dma || tifm_ms_transfer_data(host, 0)) {
+		if (host->use_dma && (fifo_status & 1)) {
+			host->cmd_flags |= FIFO_READY;
+			rc = tifm_ms_check_status(host);
+		}
+		if (!host->use_dma && (fifo_status & TIFM_FIFO_MORE)) {
+			if (!tifm_ms_transfer_data(host)) {
 				host->cmd_flags |= FIFO_READY;
 				rc = tifm_ms_check_status(host);
 			}
@@ -417,9 +419,9 @@ static void tifm_ms_card_event(struct tifm_dev *sock)
 		host_status, host->cmd_flags);
 
 	if (host->req) {
-		if (host_status & TIFM_MS_TIMEOUT)
+		if (host_status & TIFM_MS_STAT_TOE)
 			host->req->error = -ETIME;
-		else if (host_status & TIFM_MS_BADCRC)
+		else if (host_status & TIFM_MS_STAT_CRC)
 			host->req->error = -EILSEQ;
 
 		if (host->req->error) {
@@ -428,18 +430,17 @@ static void tifm_ms_card_event(struct tifm_dev *sock)
 			writel(TIFM_DMA_RESET, sock->addr + SOCK_DMA_CONTROL);
 		}
 
-		if (host_status & TIFM_MS_EOTPC)
+		if (host_status & TIFM_MS_STAT_RDY)
 			host->cmd_flags |= CMD_READY;
-		if (host_status & TIFM_MS_INT)
-			host->cmd_flags |= CARD_READY;
+
+		if (host_status & TIFM_MS_STAT_MSINT)
+			host->cmd_flags |= CARD_INT;
 
 		rc = tifm_ms_check_status(host);
 
 	}
 
-	writel(TIFM_MS_SYS_NOT_RDY | readl(sock->addr + SOCK_MS_SYSTEM),
-	       sock->addr + SOCK_MS_SYSTEM);
-	writel((~TIFM_MS_SYS_DATA) & readl(sock->addr + SOCK_MS_SYSTEM),
+	writel(TIFM_MS_SYS_INTCLR | readl(sock->addr + SOCK_MS_SYSTEM),
 	       sock->addr + SOCK_MS_SYSTEM);
 
 	if (!rc)
@@ -499,7 +500,7 @@ static void tifm_ms_set_param(struct memstick_host *msh,
 		break;
 	case MEMSTICK_INTERFACE:
 		if (value == MEMSTICK_SERIAL) {
-			host->mode_mask = TIFM_MS_SERIAL;
+			host->mode_mask = TIFM_MS_SYS_SRAC | TIFM_MS_SYS_REI;
 			writel((~TIFM_CTRL_FAST_CLK)
 			       & readl(sock->addr + SOCK_CONTROL),
 			       sock->addr + SOCK_CONTROL);
@@ -535,9 +536,10 @@ static int tifm_ms_initialize_host(struct tifm_ms *host)
 	struct tifm_dev *sock = host->dev;
 	struct memstick_host *msh = tifm_get_drvdata(sock);
 
-	host->mode_mask = TIFM_MS_SERIAL;
-	writel(0x8000, sock->addr + SOCK_MS_SYSTEM);
-	writel(0x0200 | TIFM_MS_SYS_NOT_RDY, sock->addr + SOCK_MS_SYSTEM);
+	host->mode_mask = TIFM_MS_SYS_SRAC | TIFM_MS_SYS_REI;
+	writel(TIFM_MS_SYS_RESET, sock->addr + SOCK_MS_SYSTEM);
+	writel(TIFM_MS_SYS_FCLR | TIFM_MS_SYS_INTCLR,
+	       sock->addr + SOCK_MS_SYSTEM);
 	writel(0xffffffff, sock->addr + SOCK_MS_STATUS);
 	if (tifm_has_ms_pif(sock))
 		msh->caps |= MEMSTICK_CAP_PAR4;
@@ -566,7 +568,6 @@ static int tifm_ms_probe(struct tifm_dev *sock)
 	tifm_set_drvdata(sock, msh);
 	host->dev = sock;
 	host->timeout_jiffies = msecs_to_jiffies(1000);
-	host->no_dma = no_dma;
 
 	setup_timer(&host->timer, tifm_ms_abort, (unsigned long)host);
 
@@ -599,7 +600,7 @@ static void tifm_ms_remove(struct tifm_dev *sock)
 		writel(TIFM_FIFO_INT_SETALL,
 		       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_CLEAR);
 		writel(TIFM_DMA_RESET, sock->addr + SOCK_DMA_CONTROL);
-		if (host->req->long_data && !host->no_dma)
+		if (host->use_dma)
 			tifm_unmap_sg(sock, &host->req->sg, 1,
 				      host->req->data_dir == READ
 				      ? PCI_DMA_TODEVICE
@@ -616,7 +617,8 @@ static void tifm_ms_remove(struct tifm_dev *sock)
 
 	memstick_remove_host(msh);
 
-	writel(0x0200 | TIFM_MS_SYS_NOT_RDY, sock->addr + SOCK_MS_SYSTEM);
+	writel(TIFM_MS_SYS_FCLR | TIFM_MS_SYS_INTCLR,
+	       sock->addr + SOCK_MS_SYSTEM);
 	writel(0xffffffff, sock->addr + SOCK_MS_STATUS);
 
 	memstick_free_host(msh);
diff --git a/include/linux/tifm.h b/include/linux/tifm.h
index da76ed85f595..848c0f392541 100644
--- a/include/linux/tifm.h
+++ b/include/linux/tifm.h
@@ -70,9 +70,9 @@ enum {
 
 #define TIFM_FIFO_ENABLE          0x00000001
 #define TIFM_FIFO_READY           0x00000001
+#define TIFM_FIFO_MORE            0x00000008
 #define TIFM_FIFO_INT_SETALL      0x0000ffff
 #define TIFM_FIFO_INTMASK         0x00000005
-#define TIFM_FIFO_SIZE            0x00000200
 
 #define TIFM_DMA_RESET            0x00000002
 #define TIFM_DMA_TX               0x00008000
-- 
cgit v1.2.3


From 60fdd931d577fcca351930fda4cde26ce07d35af Mon Sep 17 00:00:00 2001
From: Alex Dubov <oakad@yahoo.com>
Date: Mon, 10 Mar 2008 11:43:43 -0700
Subject: memstick: add support for JMicron jmb38x MemoryStick host controller

Signed-off-by: Alex Dubov <oakad@yahoo.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/memstick/Kconfig          |   2 +-
 drivers/memstick/host/Kconfig     |  10 +
 drivers/memstick/host/Makefile    |   6 +-
 drivers/memstick/host/jmb38x_ms.c | 945 ++++++++++++++++++++++++++++++++++++++
 include/linux/pci_ids.h           |   1 +
 5 files changed, 960 insertions(+), 4 deletions(-)
 create mode 100644 drivers/memstick/host/jmb38x_ms.c

(limited to 'include/linux')

diff --git a/drivers/memstick/Kconfig b/drivers/memstick/Kconfig
index 1093fdb07297..f0ca41c20323 100644
--- a/drivers/memstick/Kconfig
+++ b/drivers/memstick/Kconfig
@@ -8,7 +8,7 @@ menuconfig MEMSTICK
 	  Sony MemoryStick is a proprietary storage/extension card protocol.
 
 	  If you want MemoryStick support, you should say Y here and also
-	  to the specific driver for your MMC interface.
+	  to the specific driver for your MemoryStick interface.
 
 if MEMSTICK
 
diff --git a/drivers/memstick/host/Kconfig b/drivers/memstick/host/Kconfig
index c002fcc3c879..4ce5c8dffb68 100644
--- a/drivers/memstick/host/Kconfig
+++ b/drivers/memstick/host/Kconfig
@@ -20,3 +20,13 @@ config MEMSTICK_TIFM_MS
           To compile this driver as a module, choose M here: the
 	  module will be called tifm_ms.
 
+config MEMSTICK_JMICRON_38X
+	tristate "JMicron JMB38X MemoryStick interface support (EXPERIMENTAL)"
+	depends on EXPERIMENTAL && PCI
+
+	help
+	  Say Y here if you want to be able to access MemoryStick cards with
+	  the JMicron(R) JMB38X MemoryStick card reader.
+
+          To compile this driver as a module, choose M here: the
+	  module will be called jmb38x_ms.
diff --git a/drivers/memstick/host/Makefile b/drivers/memstick/host/Makefile
index ee666380efa1..12530e4311d3 100644
--- a/drivers/memstick/host/Makefile
+++ b/drivers/memstick/host/Makefile
@@ -3,8 +3,8 @@
 #
 
 ifeq ($(CONFIG_MEMSTICK_DEBUG),y)
-	EXTRA_CFLAGS		+= -DDEBUG
+	EXTRA_CFLAGS			+= -DDEBUG
 endif
 
-obj-$(CONFIG_MEMSTICK_TIFM_MS)	+= tifm_ms.o
-
+obj-$(CONFIG_MEMSTICK_TIFM_MS)		+= tifm_ms.o
+obj-$(CONFIG_MEMSTICK_JMICRON_38X)	+= jmb38x_ms.o
diff --git a/drivers/memstick/host/jmb38x_ms.c b/drivers/memstick/host/jmb38x_ms.c
new file mode 100644
index 000000000000..03fe8783b1ee
--- /dev/null
+++ b/drivers/memstick/host/jmb38x_ms.c
@@ -0,0 +1,945 @@
+/*
+ *  jmb38x_ms.c - JMicron jmb38x MemoryStick card reader
+ *
+ *  Copyright (C) 2008 Alex Dubov <oakad@yahoo.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/spinlock.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+#include <linux/highmem.h>
+#include <linux/memstick.h>
+
+#define DRIVER_NAME "jmb38x_ms"
+
+static int no_dma;
+module_param(no_dma, bool, 0644);
+
+enum {
+	DMA_ADDRESS       = 0x00,
+	BLOCK             = 0x04,
+	DMA_CONTROL       = 0x08,
+	TPC_P0            = 0x0c,
+	TPC_P1            = 0x10,
+	TPC               = 0x14,
+	HOST_CONTROL      = 0x18,
+	DATA              = 0x1c,
+	STATUS            = 0x20,
+	INT_STATUS        = 0x24,
+	INT_STATUS_ENABLE = 0x28,
+	INT_SIGNAL_ENABLE = 0x2c,
+	TIMER             = 0x30,
+	TIMER_CONTROL     = 0x34,
+	PAD_OUTPUT_ENABLE = 0x38,
+	PAD_PU_PD         = 0x3c,
+	CLOCK_DELAY       = 0x40,
+	ADMA_ADDRESS      = 0x44,
+	CLOCK_CONTROL     = 0x48,
+	LED_CONTROL       = 0x4c,
+	VERSION           = 0x50
+};
+
+struct jmb38x_ms_host {
+	struct jmb38x_ms        *chip;
+	void __iomem            *addr;
+	spinlock_t              lock;
+	int                     id;
+	char                    host_id[DEVICE_ID_SIZE];
+	int                     irq;
+	unsigned int            block_pos;
+	unsigned long           timeout_jiffies;
+	struct timer_list       timer;
+	struct memstick_request *req;
+	unsigned char           eject:1,
+				use_dma:1;
+	unsigned char           cmd_flags;
+	unsigned char           io_pos;
+	unsigned int            io_word[2];
+};
+
+struct jmb38x_ms {
+	struct pci_dev        *pdev;
+	int                   host_cnt;
+	struct memstick_host  *hosts[];
+};
+
+#define BLOCK_COUNT_MASK       0xffff0000
+#define BLOCK_SIZE_MASK        0x00000fff
+
+#define DMA_CONTROL_ENABLE     0x00000001
+
+#define TPC_DATA_SEL           0x00008000
+#define TPC_DIR                0x00004000
+#define TPC_WAIT_INT           0x00002000
+#define TPC_GET_INT            0x00000800
+#define TPC_CODE_SZ_MASK       0x00000700
+#define TPC_DATA_SZ_MASK       0x00000007
+
+#define HOST_CONTROL_RESET_REQ 0x00008000
+#define HOST_CONTROL_REI       0x00004000
+#define HOST_CONTROL_LED       0x00000400
+#define HOST_CONTROL_FAST_CLK  0x00000200
+#define HOST_CONTROL_RESET     0x00000100
+#define HOST_CONTROL_POWER_EN  0x00000080
+#define HOST_CONTROL_CLOCK_EN  0x00000040
+#define HOST_CONTROL_IF_SHIFT  4
+
+#define HOST_CONTROL_IF_SERIAL 0x0
+#define HOST_CONTROL_IF_PAR4   0x1
+#define HOST_CONTROL_IF_PAR8   0x3
+
+#define STATUS_HAS_MEDIA        0x00000400
+#define STATUS_FIFO_EMPTY       0x00000200
+#define STATUS_FIFO_FULL        0x00000100
+
+#define INT_STATUS_TPC_ERR      0x00080000
+#define INT_STATUS_CRC_ERR      0x00040000
+#define INT_STATUS_TIMER_TO     0x00020000
+#define INT_STATUS_HSK_TO       0x00010000
+#define INT_STATUS_ANY_ERR      0x00008000
+#define INT_STATUS_FIFO_WRDY    0x00000080
+#define INT_STATUS_FIFO_RRDY    0x00000040
+#define INT_STATUS_MEDIA_OUT    0x00000010
+#define INT_STATUS_MEDIA_IN     0x00000008
+#define INT_STATUS_DMA_BOUNDARY 0x00000004
+#define INT_STATUS_EOTRAN       0x00000002
+#define INT_STATUS_EOTPC        0x00000001
+
+#define INT_STATUS_ALL          0x000f801f
+
+#define PAD_OUTPUT_ENABLE_MS  0x0F3F
+
+#define PAD_PU_PD_OFF         0x7FFF0000
+#define PAD_PU_PD_ON_MS_SOCK0 0x5f8f0000
+#define PAD_PU_PD_ON_MS_SOCK1 0x0f0f0000
+
+enum {
+	CMD_READY    = 0x01,
+	FIFO_READY   = 0x02,
+	REG_DATA     = 0x04,
+	AUTO_GET_INT = 0x08
+};
+
+static unsigned int jmb38x_ms_read_data(struct jmb38x_ms_host *host,
+					unsigned char *buf, unsigned int length)
+{
+	unsigned int off = 0;
+
+	while (host->io_pos && length) {
+		buf[off++] = host->io_word[0] & 0xff;
+		host->io_word[0] >>= 8;
+		length--;
+		host->io_pos--;
+	}
+
+	if (!length)
+		return off;
+
+	while (!(STATUS_FIFO_EMPTY & readl(host->addr + STATUS))) {
+		if (length < 4)
+			break;
+		*(unsigned int *)(buf + off) = __raw_readl(host->addr + DATA);
+		length -= 4;
+		off += 4;
+	}
+
+	if (length
+	    && !(STATUS_FIFO_EMPTY & readl(host->addr + STATUS))) {
+		host->io_word[0] = readl(host->addr + DATA);
+		for (host->io_pos = 4; host->io_pos; --host->io_pos) {
+			buf[off++] = host->io_word[0] & 0xff;
+			host->io_word[0] >>= 8;
+			length--;
+			if (!length)
+				break;
+		}
+	}
+
+	return off;
+}
+
+static unsigned int jmb38x_ms_read_reg_data(struct jmb38x_ms_host *host,
+					    unsigned char *buf,
+					    unsigned int length)
+{
+	unsigned int off = 0;
+
+	while (host->io_pos > 4 && length) {
+		buf[off++] = host->io_word[0] & 0xff;
+		host->io_word[0] >>= 8;
+		length--;
+		host->io_pos--;
+	}
+
+	if (!length)
+		return off;
+
+	while (host->io_pos && length) {
+		buf[off++] = host->io_word[1] & 0xff;
+		host->io_word[1] >>= 8;
+		length--;
+		host->io_pos--;
+	}
+
+	return off;
+}
+
+static unsigned int jmb38x_ms_write_data(struct jmb38x_ms_host *host,
+					 unsigned char *buf,
+					 unsigned int length)
+{
+	unsigned int off = 0;
+
+	if (host->io_pos) {
+		while (host->io_pos < 4 && length) {
+			host->io_word[0] |=  buf[off++] << (host->io_pos * 8);
+			host->io_pos++;
+			length--;
+		}
+	}
+
+	if (host->io_pos == 4
+	    && !(STATUS_FIFO_FULL & readl(host->addr + STATUS))) {
+		writel(host->io_word[0], host->addr + DATA);
+		host->io_pos = 0;
+		host->io_word[0] = 0;
+	} else if (host->io_pos) {
+		return off;
+	}
+
+	if (!length)
+		return off;
+
+	while (!(STATUS_FIFO_FULL & readl(host->addr + STATUS))) {
+		if (length < 4)
+			break;
+
+		__raw_writel(*(unsigned int *)(buf + off),
+			     host->addr + DATA);
+		length -= 4;
+		off += 4;
+	}
+
+	switch (length) {
+	case 3:
+		host->io_word[0] |= buf[off + 2] << 16;
+		host->io_pos++;
+	case 2:
+		host->io_word[0] |= buf[off + 1] << 8;
+		host->io_pos++;
+	case 1:
+		host->io_word[0] |= buf[off];
+		host->io_pos++;
+	}
+
+	off += host->io_pos;
+
+	return off;
+}
+
+static unsigned int jmb38x_ms_write_reg_data(struct jmb38x_ms_host *host,
+					     unsigned char *buf,
+					     unsigned int length)
+{
+	unsigned int off = 0;
+
+	while (host->io_pos < 4 && length) {
+		host->io_word[0] &= ~(0xff << (host->io_pos * 8));
+		host->io_word[0] |=  buf[off++] << (host->io_pos * 8);
+		host->io_pos++;
+		length--;
+	}
+
+	if (!length)
+		return off;
+
+	while (host->io_pos < 8 && length) {
+		host->io_word[1] &= ~(0xff << (host->io_pos * 8));
+		host->io_word[1] |=  buf[off++] << (host->io_pos * 8);
+		host->io_pos++;
+		length--;
+	}
+
+	return off;
+}
+
+static int jmb38x_ms_transfer_data(struct jmb38x_ms_host *host)
+{
+	unsigned int length;
+	unsigned int off;
+	unsigned int t_size, p_off, p_cnt;
+	unsigned char *buf;
+	struct page *pg;
+	unsigned long flags = 0;
+
+	if (host->req->long_data) {
+		length = host->req->sg.length - host->block_pos;
+		off = host->req->sg.offset + host->block_pos;
+	} else {
+		length = host->req->data_len - host->block_pos;
+		off = 0;
+	}
+
+	while (length) {
+		if (host->req->long_data) {
+			pg = nth_page(sg_page(&host->req->sg),
+				      off >> PAGE_SHIFT);
+			p_off = offset_in_page(off);
+			p_cnt = PAGE_SIZE - p_off;
+			p_cnt = min(p_cnt, length);
+
+			local_irq_save(flags);
+			buf = kmap_atomic(pg, KM_BIO_SRC_IRQ) + p_off;
+		} else {
+			buf = host->req->data + host->block_pos;
+			p_cnt = host->req->data_len - host->block_pos;
+		}
+
+		if (host->req->data_dir == WRITE)
+			t_size = !(host->cmd_flags & REG_DATA)
+				 ? jmb38x_ms_write_data(host, buf, p_cnt)
+				 : jmb38x_ms_write_reg_data(host, buf, p_cnt);
+		else
+			t_size = !(host->cmd_flags & REG_DATA)
+				 ? jmb38x_ms_read_data(host, buf, p_cnt)
+				 : jmb38x_ms_read_reg_data(host, buf, p_cnt);
+
+		if (host->req->long_data) {
+			kunmap_atomic(buf - p_off, KM_BIO_SRC_IRQ);
+			local_irq_restore(flags);
+		}
+
+		if (!t_size)
+			break;
+		host->block_pos += t_size;
+		length -= t_size;
+		off += t_size;
+	}
+
+	if (!length && host->req->data_dir == WRITE) {
+		if (host->cmd_flags & REG_DATA) {
+			writel(host->io_word[0], host->addr + TPC_P0);
+			writel(host->io_word[1], host->addr + TPC_P1);
+		} else if (host->io_pos) {
+			writel(host->io_word[0], host->addr + DATA);
+		}
+	}
+
+	return length;
+}
+
+static int jmb38x_ms_issue_cmd(struct memstick_host *msh)
+{
+	struct jmb38x_ms_host *host = memstick_priv(msh);
+	unsigned char *data;
+	unsigned int data_len, cmd, t_val;
+
+	if (!(STATUS_HAS_MEDIA & readl(host->addr + STATUS))) {
+		dev_dbg(msh->cdev.dev, "no media status\n");
+		host->req->error = -ETIME;
+		return host->req->error;
+	}
+
+	dev_dbg(msh->cdev.dev, "control %08x\n",
+		readl(host->addr + HOST_CONTROL));
+	dev_dbg(msh->cdev.dev, "status %08x\n", readl(host->addr + INT_STATUS));
+	dev_dbg(msh->cdev.dev, "hstatus %08x\n", readl(host->addr + STATUS));
+
+	host->cmd_flags = 0;
+	host->block_pos = 0;
+	host->io_pos = 0;
+	host->io_word[0] = 0;
+	host->io_word[1] = 0;
+
+	cmd = host->req->tpc << 16;
+	cmd |= TPC_DATA_SEL;
+
+	if (host->req->data_dir == READ)
+		cmd |= TPC_DIR;
+	if (host->req->need_card_int)
+		cmd |= TPC_WAIT_INT;
+	if (host->req->get_int_reg)
+		cmd |= TPC_GET_INT;
+
+	data = host->req->data;
+
+	host->use_dma = !no_dma;
+
+	if (host->req->long_data) {
+		data_len = host->req->sg.length;
+	} else {
+		data_len = host->req->data_len;
+		host->use_dma = 0;
+	}
+
+	if (data_len <= 8) {
+		cmd &= ~(TPC_DATA_SEL | 0xf);
+		host->cmd_flags |= REG_DATA;
+		cmd |= data_len & 0xf;
+		host->use_dma = 0;
+	}
+
+	if (host->use_dma) {
+		if (1 != pci_map_sg(host->chip->pdev, &host->req->sg, 1,
+				    host->req->data_dir == READ
+				    ? PCI_DMA_FROMDEVICE
+				    : PCI_DMA_TODEVICE)) {
+			host->req->error = -ENOMEM;
+			return host->req->error;
+		}
+		data_len = sg_dma_len(&host->req->sg);
+		writel(sg_dma_address(&host->req->sg),
+		       host->addr + DMA_ADDRESS);
+		writel(((1 << 16) & BLOCK_COUNT_MASK)
+		       | (data_len & BLOCK_SIZE_MASK),
+		       host->addr + BLOCK);
+		writel(DMA_CONTROL_ENABLE, host->addr + DMA_CONTROL);
+	} else if (!(host->cmd_flags & REG_DATA)) {
+		writel(((1 << 16) & BLOCK_COUNT_MASK)
+		       | (data_len & BLOCK_SIZE_MASK),
+		       host->addr + BLOCK);
+			t_val = readl(host->addr + INT_STATUS_ENABLE);
+			t_val |= host->req->data_dir == READ
+				 ? INT_STATUS_FIFO_RRDY
+				 : INT_STATUS_FIFO_WRDY;
+
+			writel(t_val, host->addr + INT_STATUS_ENABLE);
+			writel(t_val, host->addr + INT_SIGNAL_ENABLE);
+	} else {
+		cmd &= ~(TPC_DATA_SEL | 0xf);
+		host->cmd_flags |= REG_DATA;
+		cmd |= data_len & 0xf;
+
+		if (host->req->data_dir == WRITE) {
+			jmb38x_ms_transfer_data(host);
+			writel(host->io_word[0], host->addr + TPC_P0);
+			writel(host->io_word[1], host->addr + TPC_P1);
+		}
+	}
+
+	mod_timer(&host->timer, jiffies + host->timeout_jiffies);
+	writel(HOST_CONTROL_LED | readl(host->addr + HOST_CONTROL),
+	       host->addr + HOST_CONTROL);
+	host->req->error = 0;
+
+	writel(cmd, host->addr + TPC);
+	dev_dbg(msh->cdev.dev, "executing TPC %08x, len %x\n", cmd, data_len);
+
+	return 0;
+}
+
+static void jmb38x_ms_complete_cmd(struct memstick_host *msh, int last)
+{
+	struct jmb38x_ms_host *host = memstick_priv(msh);
+	unsigned int t_val = 0;
+	int rc;
+
+	del_timer(&host->timer);
+
+	dev_dbg(msh->cdev.dev, "c control %08x\n",
+		readl(host->addr + HOST_CONTROL));
+	dev_dbg(msh->cdev.dev, "c status %08x\n",
+		readl(host->addr + INT_STATUS));
+	dev_dbg(msh->cdev.dev, "c hstatus %08x\n", readl(host->addr + STATUS));
+
+	if (host->req->get_int_reg) {
+		t_val = readl(host->addr + TPC_P0);
+		host->req->int_reg = (t_val & 0xff);
+	}
+
+	if (host->use_dma) {
+		writel(0, host->addr + DMA_CONTROL);
+		pci_unmap_sg(host->chip->pdev, &host->req->sg, 1,
+			     host->req->data_dir == READ
+			     ? PCI_DMA_FROMDEVICE : PCI_DMA_TODEVICE);
+	} else {
+		t_val = readl(host->addr + INT_STATUS_ENABLE);
+		if (host->req->data_dir == READ)
+			t_val &= ~INT_STATUS_FIFO_RRDY;
+		else
+			t_val &= ~INT_STATUS_FIFO_WRDY;
+
+		writel(t_val, host->addr + INT_STATUS_ENABLE);
+		writel(t_val, host->addr + INT_SIGNAL_ENABLE);
+	}
+
+	writel((~HOST_CONTROL_LED) & readl(host->addr + HOST_CONTROL),
+	       host->addr + HOST_CONTROL);
+
+	if (!last) {
+		do {
+			rc = memstick_next_req(msh, &host->req);
+		} while (!rc && jmb38x_ms_issue_cmd(msh));
+	} else {
+		do {
+			rc = memstick_next_req(msh, &host->req);
+			if (!rc)
+				host->req->error = -ETIME;
+		} while (!rc);
+	}
+}
+
+static irqreturn_t jmb38x_ms_isr(int irq, void *dev_id)
+{
+	struct memstick_host *msh = dev_id;
+	struct jmb38x_ms_host *host = memstick_priv(msh);
+	unsigned int irq_status;
+
+	spin_lock(&host->lock);
+	irq_status = readl(host->addr + INT_STATUS);
+	dev_dbg(&host->chip->pdev->dev, "irq_status = %08x\n", irq_status);
+	if (irq_status == 0 || irq_status == (~0)) {
+		spin_unlock(&host->lock);
+		return IRQ_NONE;
+	}
+
+	if (host->req) {
+		if (irq_status & INT_STATUS_ANY_ERR) {
+			if (irq_status & INT_STATUS_CRC_ERR)
+				host->req->error = -EILSEQ;
+			else
+				host->req->error = -ETIME;
+		} else {
+			if (host->use_dma) {
+				if (irq_status & INT_STATUS_EOTRAN)
+					host->cmd_flags |= FIFO_READY;
+			} else {
+				if (irq_status & (INT_STATUS_FIFO_RRDY
+						  | INT_STATUS_FIFO_WRDY))
+					jmb38x_ms_transfer_data(host);
+
+				if (irq_status & INT_STATUS_EOTRAN) {
+					jmb38x_ms_transfer_data(host);
+					host->cmd_flags |= FIFO_READY;
+				}
+			}
+
+			if (irq_status & INT_STATUS_EOTPC) {
+				host->cmd_flags |= CMD_READY;
+				if (host->cmd_flags & REG_DATA) {
+					if (host->req->data_dir == READ) {
+						host->io_word[0]
+							= readl(host->addr
+								+ TPC_P0);
+						host->io_word[1]
+							= readl(host->addr
+								+ TPC_P1);
+						host->io_pos = 8;
+
+						jmb38x_ms_transfer_data(host);
+					}
+					host->cmd_flags |= FIFO_READY;
+				}
+			}
+		}
+	}
+
+	if (irq_status & (INT_STATUS_MEDIA_IN | INT_STATUS_MEDIA_OUT)) {
+		dev_dbg(&host->chip->pdev->dev, "media changed\n");
+		memstick_detect_change(msh);
+	}
+
+	writel(irq_status, host->addr + INT_STATUS);
+
+	if (host->req
+	    && (((host->cmd_flags & CMD_READY)
+		 && (host->cmd_flags & FIFO_READY))
+		|| host->req->error))
+		jmb38x_ms_complete_cmd(msh, 0);
+
+	spin_unlock(&host->lock);
+	return IRQ_HANDLED;
+}
+
+static void jmb38x_ms_abort(unsigned long data)
+{
+	struct memstick_host *msh = (struct memstick_host *)data;
+	struct jmb38x_ms_host *host = memstick_priv(msh);
+	unsigned long flags;
+
+	dev_dbg(&host->chip->pdev->dev, "abort\n");
+	spin_lock_irqsave(&host->lock, flags);
+	if (host->req) {
+		host->req->error = -ETIME;
+		jmb38x_ms_complete_cmd(msh, 0);
+	}
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static void jmb38x_ms_request(struct memstick_host *msh)
+{
+	struct jmb38x_ms_host *host = memstick_priv(msh);
+	unsigned long flags;
+	int rc;
+
+	spin_lock_irqsave(&host->lock, flags);
+	if (host->req) {
+		spin_unlock_irqrestore(&host->lock, flags);
+		BUG();
+		return;
+	}
+
+	do {
+		rc = memstick_next_req(msh, &host->req);
+	} while (!rc && jmb38x_ms_issue_cmd(msh));
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+static void jmb38x_ms_reset(struct jmb38x_ms_host *host)
+{
+	unsigned int host_ctl = readl(host->addr + HOST_CONTROL);
+
+	writel(host_ctl | HOST_CONTROL_RESET_REQ | HOST_CONTROL_RESET,
+	       host->addr + HOST_CONTROL);
+
+	while (HOST_CONTROL_RESET_REQ
+	       & (host_ctl = readl(host->addr + HOST_CONTROL))) {
+		ndelay(100);
+		dev_dbg(&host->chip->pdev->dev, "reset\n");
+	}
+
+	writel(INT_STATUS_ALL, host->addr + INT_STATUS_ENABLE);
+	writel(INT_STATUS_ALL, host->addr + INT_SIGNAL_ENABLE);
+
+	dev_dbg(&host->chip->pdev->dev, "reset\n");
+}
+
+static void jmb38x_ms_set_param(struct memstick_host *msh,
+				enum memstick_param param,
+				int value)
+{
+	struct jmb38x_ms_host *host = memstick_priv(msh);
+	unsigned int host_ctl;
+	unsigned long flags;
+
+	spin_lock_irqsave(&host->lock, flags);
+
+	switch (param) {
+	case MEMSTICK_POWER:
+		if (value == MEMSTICK_POWER_ON) {
+			jmb38x_ms_reset(host);
+
+			writel(host->id ? PAD_PU_PD_ON_MS_SOCK1
+					  : PAD_PU_PD_ON_MS_SOCK0,
+			       host->addr + PAD_PU_PD);
+
+			writel(PAD_OUTPUT_ENABLE_MS,
+			       host->addr + PAD_OUTPUT_ENABLE);
+
+			host_ctl = readl(host->addr + HOST_CONTROL);
+			host_ctl |= 7;
+			writel(host_ctl | (HOST_CONTROL_POWER_EN
+					   | HOST_CONTROL_CLOCK_EN),
+			       host->addr + HOST_CONTROL);
+
+			dev_dbg(&host->chip->pdev->dev, "power on\n");
+		} else if (value == MEMSTICK_POWER_OFF) {
+			writel(readl(host->addr + HOST_CONTROL)
+			       & ~(HOST_CONTROL_POWER_EN
+				   | HOST_CONTROL_CLOCK_EN),
+			       host->addr +  HOST_CONTROL);
+			writel(0, host->addr + PAD_OUTPUT_ENABLE);
+			writel(PAD_PU_PD_OFF, host->addr + PAD_PU_PD);
+			dev_dbg(&host->chip->pdev->dev, "power off\n");
+		}
+		break;
+	case MEMSTICK_INTERFACE:
+		/* jmb38x_ms_reset(host); */
+
+		host_ctl = readl(host->addr + HOST_CONTROL);
+		host_ctl &= ~(3 << HOST_CONTROL_IF_SHIFT);
+		/* host_ctl |= 7; */
+
+		if (value == MEMSTICK_SERIAL) {
+			host_ctl &= ~HOST_CONTROL_FAST_CLK;
+			host_ctl |= HOST_CONTROL_IF_SERIAL
+				    << HOST_CONTROL_IF_SHIFT;
+			host_ctl |= HOST_CONTROL_REI;
+			writel(0, host->addr + CLOCK_DELAY);
+		} else if (value == MEMSTICK_PAR4) {
+			host_ctl |= HOST_CONTROL_FAST_CLK;
+			host_ctl |= HOST_CONTROL_IF_PAR4
+				    << HOST_CONTROL_IF_SHIFT;
+			host_ctl &= ~HOST_CONTROL_REI;
+			writel(4, host->addr + CLOCK_DELAY);
+		} else if (value == MEMSTICK_PAR8) {
+			host_ctl |= HOST_CONTROL_FAST_CLK;
+			host_ctl |= HOST_CONTROL_IF_PAR8
+				    << HOST_CONTROL_IF_SHIFT;
+			host_ctl &= ~HOST_CONTROL_REI;
+			writel(4, host->addr + CLOCK_DELAY);
+		}
+		writel(host_ctl, host->addr + HOST_CONTROL);
+		break;
+	};
+
+	spin_unlock_irqrestore(&host->lock, flags);
+}
+
+#ifdef CONFIG_PM
+
+static int jmb38x_ms_suspend(struct pci_dev *dev, pm_message_t state)
+{
+	struct jmb38x_ms *jm = pci_get_drvdata(dev);
+	int cnt;
+
+	for (cnt = 0; cnt < jm->host_cnt; ++cnt) {
+		if (!jm->hosts[cnt])
+			break;
+		memstick_suspend_host(jm->hosts[cnt]);
+	}
+
+	pci_save_state(dev);
+	pci_enable_wake(dev, pci_choose_state(dev, state), 0);
+	pci_disable_device(dev);
+	pci_set_power_state(dev, pci_choose_state(dev, state));
+	return 0;
+}
+
+static int jmb38x_ms_resume(struct pci_dev *dev)
+{
+	struct jmb38x_ms *jm = pci_get_drvdata(dev);
+	int rc;
+
+	pci_set_power_state(dev, PCI_D0);
+	pci_restore_state(dev);
+	rc = pci_enable_device(dev);
+	if (rc)
+		return rc;
+	pci_set_master(dev);
+
+	pci_read_config_dword(dev, 0xac, &rc);
+	pci_write_config_dword(dev, 0xac, rc | 0x00470000);
+
+	for (rc = 0; rc < jm->host_cnt; ++rc) {
+		if (!jm->hosts[rc])
+			break;
+		memstick_resume_host(jm->hosts[rc]);
+		memstick_detect_change(jm->hosts[rc]);
+	}
+
+	return 0;
+}
+
+#else
+
+#define jmb38x_ms_suspend NULL
+#define jmb38x_ms_resume NULL
+
+#endif /* CONFIG_PM */
+
+static int jmb38x_ms_count_slots(struct pci_dev *pdev)
+{
+	int cnt, rc = 0;
+
+	for (cnt = 0; cnt < PCI_ROM_RESOURCE; ++cnt) {
+		if (!(IORESOURCE_MEM & pci_resource_flags(pdev, cnt)))
+			break;
+
+		if (256 != pci_resource_len(pdev, cnt))
+			break;
+
+		++rc;
+	}
+	return rc;
+}
+
+static struct memstick_host *jmb38x_ms_alloc_host(struct jmb38x_ms *jm, int cnt)
+{
+	struct memstick_host *msh;
+	struct jmb38x_ms_host *host;
+
+	msh = memstick_alloc_host(sizeof(struct jmb38x_ms_host),
+				  &jm->pdev->dev);
+	if (!msh)
+		return NULL;
+
+	host = memstick_priv(msh);
+	host->chip = jm;
+	host->addr = ioremap(pci_resource_start(jm->pdev, cnt),
+			     pci_resource_len(jm->pdev, cnt));
+	if (!host->addr)
+		goto err_out_free;
+
+	spin_lock_init(&host->lock);
+	host->id = cnt;
+	snprintf(host->host_id, DEVICE_ID_SIZE, DRIVER_NAME ":slot%d",
+		 host->id);
+	host->irq = jm->pdev->irq;
+	host->timeout_jiffies = msecs_to_jiffies(4000);
+	msh->request = jmb38x_ms_request;
+	msh->set_param = jmb38x_ms_set_param;
+	/*
+	msh->caps = MEMSTICK_CAP_AUTO_GET_INT | MEMSTICK_CAP_PAR4
+		    | MEMSTICK_CAP_PAR8;
+	*/
+	msh->caps = MEMSTICK_CAP_PAR4 | MEMSTICK_CAP_PAR8;
+
+	setup_timer(&host->timer, jmb38x_ms_abort, (unsigned long)msh);
+
+	if (!request_irq(host->irq, jmb38x_ms_isr, IRQF_SHARED, host->host_id,
+			 msh))
+		return msh;
+
+	iounmap(host->addr);
+err_out_free:
+	kfree(msh);
+	return NULL;
+}
+
+static void jmb38x_ms_free_host(struct memstick_host *msh)
+{
+	struct jmb38x_ms_host *host = memstick_priv(msh);
+
+	free_irq(host->irq, msh);
+	iounmap(host->addr);
+	memstick_free_host(msh);
+}
+
+static int jmb38x_ms_probe(struct pci_dev *pdev,
+			   const struct pci_device_id *dev_id)
+{
+	struct jmb38x_ms *jm;
+	int pci_dev_busy = 0;
+	int rc, cnt;
+
+	rc = pci_set_dma_mask(pdev, DMA_32BIT_MASK);
+	if (rc)
+		return rc;
+
+	rc = pci_enable_device(pdev);
+	if (rc)
+		return rc;
+
+	pci_set_master(pdev);
+
+	rc = pci_request_regions(pdev, DRIVER_NAME);
+	if (rc) {
+		pci_dev_busy = 1;
+		goto err_out;
+	}
+
+	pci_read_config_dword(pdev, 0xac, &rc);
+	pci_write_config_dword(pdev, 0xac, rc | 0x00470000);
+
+	cnt = jmb38x_ms_count_slots(pdev);
+	if (!cnt) {
+		rc = -ENODEV;
+		pci_dev_busy = 1;
+		goto err_out;
+	}
+
+	jm = kzalloc(sizeof(struct jmb38x_ms)
+		     + cnt * sizeof(struct memstick_host *), GFP_KERNEL);
+	if (!jm) {
+		rc = -ENOMEM;
+		goto err_out_int;
+	}
+
+	jm->pdev = pdev;
+	jm->host_cnt = cnt;
+	pci_set_drvdata(pdev, jm);
+
+	for (cnt = 0; cnt < jm->host_cnt; ++cnt) {
+		jm->hosts[cnt] = jmb38x_ms_alloc_host(jm, cnt);
+		if (!jm->hosts[cnt])
+			break;
+
+		rc = memstick_add_host(jm->hosts[cnt]);
+
+		if (rc) {
+			jmb38x_ms_free_host(jm->hosts[cnt]);
+			jm->hosts[cnt] = NULL;
+			break;
+		}
+	}
+
+	if (cnt)
+		return 0;
+
+	rc = -ENODEV;
+
+	pci_set_drvdata(pdev, NULL);
+	kfree(jm);
+err_out_int:
+	pci_release_regions(pdev);
+err_out:
+	if (!pci_dev_busy)
+		pci_disable_device(pdev);
+	return rc;
+}
+
+static void jmb38x_ms_remove(struct pci_dev *dev)
+{
+	struct jmb38x_ms *jm = pci_get_drvdata(dev);
+	struct jmb38x_ms_host *host;
+	int cnt;
+	unsigned long flags;
+
+	for (cnt = 0; cnt < jm->host_cnt; ++cnt) {
+		if (!jm->hosts[cnt])
+			break;
+
+		host = memstick_priv(jm->hosts[cnt]);
+
+		writel(0, host->addr + INT_SIGNAL_ENABLE);
+		writel(0, host->addr + INT_STATUS_ENABLE);
+		mmiowb();
+		dev_dbg(&jm->pdev->dev, "interrupts off\n");
+		spin_lock_irqsave(&host->lock, flags);
+		if (host->req) {
+			host->req->error = -ETIME;
+			jmb38x_ms_complete_cmd(jm->hosts[cnt], 1);
+		}
+		spin_unlock_irqrestore(&host->lock, flags);
+
+		memstick_remove_host(jm->hosts[cnt]);
+		dev_dbg(&jm->pdev->dev, "host removed\n");
+
+		jmb38x_ms_free_host(jm->hosts[cnt]);
+	}
+
+	pci_set_drvdata(dev, NULL);
+	pci_release_regions(dev);
+	pci_disable_device(dev);
+	kfree(jm);
+}
+
+static struct pci_device_id jmb38x_ms_id_tbl [] = {
+	{ PCI_VENDOR_ID_JMICRON, PCI_DEVICE_ID_JMICRON_JMB38X_MS, PCI_ANY_ID,
+	  PCI_ANY_ID, 0, 0, 0 },
+	{ }
+};
+
+static struct pci_driver jmb38x_ms_driver = {
+	.name = DRIVER_NAME,
+	.id_table = jmb38x_ms_id_tbl,
+	.probe = jmb38x_ms_probe,
+	.remove = jmb38x_ms_remove,
+	.suspend = jmb38x_ms_suspend,
+	.resume = jmb38x_ms_resume
+};
+
+static int __init jmb38x_ms_init(void)
+{
+	return pci_register_driver(&jmb38x_ms_driver);
+}
+
+static void __exit jmb38x_ms_exit(void)
+{
+	pci_unregister_driver(&jmb38x_ms_driver);
+}
+
+MODULE_AUTHOR("Alex Dubov");
+MODULE_DESCRIPTION("JMicron jmb38x MemoryStick driver");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(pci, jmb38x_ms_id_tbl);
+
+module_init(jmb38x_ms_init);
+module_exit(jmb38x_ms_exit);
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index effdb558a588..70eb3c803d47 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -2184,6 +2184,7 @@
 #define PCI_DEVICE_ID_JMICRON_JMB366	0x2366
 #define PCI_DEVICE_ID_JMICRON_JMB368	0x2368
 #define PCI_DEVICE_ID_JMICRON_JMB38X_SD	0x2381
+#define PCI_DEVICE_ID_JMICRON_JMB38X_MS	0x2383
 
 #define PCI_VENDOR_ID_KORENIX		0x1982
 #define PCI_DEVICE_ID_KORENIX_JETCARDF0	0x1600
-- 
cgit v1.2.3


From 0738c4bb8f2a8bf15178f852494643b0981f578b Mon Sep 17 00:00:00 2001
From: Paul Mundt <lethal@linux-sh.org>
Date: Wed, 12 Mar 2008 16:51:31 +0900
Subject: nommu: Provide is_vmalloc_addr() stub.

Introduced in commit-id 9e2779fa281cfda13ac060753d674bbcaa23367e and
ifdef'ed out for nommu in 8ca3ed87db062201e1fa15b64a9214e193fc3a8a, both
approaches end up breaking the nommu build in different ways. An
impressive feat for a 2-liner.

Current is_vmalloc_addr() users fall in to two camps:

	- Determining whether to use vfree()/kfree()
	- Whether to do vmlist traversal (only /proc/kcore).

Since we don't support /proc/kcore on nommu, that leaves the
vfree()/kfree() determination use cases. nommu vfree() happens to be a
wrapper to kfree() anyways, so is_vmalloc_addr() can always return 0
and end up with the right behaviour.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/mm.h | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 3f3ccfe42de0..b695875d63e3 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -235,15 +235,22 @@ static inline int get_page_unless_zero(struct page *page)
 struct page *vmalloc_to_page(const void *addr);
 unsigned long vmalloc_to_pfn(const void *addr);
 
-#ifdef CONFIG_MMU
-/* Determine if an address is within the vmalloc range */
+/*
+ * Determine if an address is within the vmalloc range
+ *
+ * On nommu, vmalloc/vfree wrap through kmalloc/kfree directly, so there
+ * is no special casing required.
+ */
 static inline int is_vmalloc_addr(const void *x)
 {
+#ifdef CONFIG_MMU
 	unsigned long addr = (unsigned long)x;
 
 	return addr >= VMALLOC_START && addr < VMALLOC_END;
-}
+#else
+	return 0;
 #endif
+}
 
 static inline struct page *compound_head(struct page *page)
 {
-- 
cgit v1.2.3


From a8ae50ba9336ff77d0df0943ac27b79ba0a5a521 Mon Sep 17 00:00:00 2001
From: David Woodhouse <dwmw2@infradead.org>
Date: Wed, 12 Mar 2008 17:52:56 +0100
Subject: Remove <linux/genhd.h> from user-visible headers.

It was all wrapped in '#ifdef CONFIG_BLOCK' anyway, so userspace was
getting nothing useful out of it. And the special #ifndef __KERNEL__
version of 'struct partition' makes me inclined to promote an attitude
of violence...

Stick some comments on some of the #endifs too, while we're at it.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/Kbuild  |  1 -
 include/linux/genhd.h | 30 +++---------------------------
 2 files changed, 3 insertions(+), 28 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 994df3780007..0fac822c1157 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -205,7 +205,6 @@ unifdef-y += futex.h
 unifdef-y += fs.h
 unifdef-y += gameport.h
 unifdef-y += generic_serial.h
-unifdef-y += genhd.h
 unifdef-y += gfs2_ondisk.h
 unifdef-y += hayesesp.h
 unifdef-y += hdlcdrv.h
diff --git a/include/linux/genhd.h b/include/linux/genhd.h
index 32c2ac49a070..ecd2bf63fc84 100644
--- a/include/linux/genhd.h
+++ b/include/linux/genhd.h
@@ -55,24 +55,6 @@ enum {
 	UNIXWARE_PARTITION = 0x63,	/* Same as GNU_HURD and SCO Unix */
 };
 
-#ifndef __KERNEL__
-
-struct partition {
-	unsigned char boot_ind;		/* 0x80 - active */
-	unsigned char head;		/* starting head */
-	unsigned char sector;		/* starting sector */
-	unsigned char cyl;		/* starting cylinder */
-	unsigned char sys_ind;		/* What partition type */
-	unsigned char end_head;		/* end head */
-	unsigned char end_sector;	/* end sector */
-	unsigned char end_cyl;		/* end cylinder */
-	unsigned int start_sect;	/* starting sector counting from 0 */
-	unsigned int nr_sects;		/* nr of sectors in partition */
-} __attribute__((packed));
-
-#endif
-
-#ifdef __KERNEL__
 #include <linux/major.h>
 #include <linux/device.h>
 #include <linux/smp.h>
@@ -228,7 +210,7 @@ static inline void part_stat_set_all(struct hd_struct *part, int value)	{
 		       sizeof(struct disk_stats));
 }
 				
-#else
+#else /* !CONFIG_SMP */
 #define __disk_stat_add(gendiskp, field, addnd) \
 				(gendiskp->dkstats.field += addnd)
 #define disk_stat_read(gendiskp, field)	(gendiskp->dkstats.field)
@@ -256,7 +238,7 @@ static inline void part_stat_set_all(struct hd_struct *part, int value)
 	memset(&part->dkstats, value, sizeof(struct disk_stats));
 }
 
-#endif
+#endif /* CONFIG_SMP */
 
 #define disk_stat_add(gendiskp, field, addnd)			\
 	do {							\
@@ -395,8 +377,6 @@ static inline void set_capacity(struct gendisk *disk, sector_t size)
 	disk->capacity = size;
 }
 
-#endif  /*  __KERNEL__  */
-
 #ifdef CONFIG_SOLARIS_X86_PARTITION
 
 #define SOLARIS_X86_NUMSLICE	16
@@ -540,8 +520,6 @@ struct unixware_disklabel {
 #   define MINIX_NR_SUBPARTITIONS  4
 #endif /* CONFIG_MINIX_SUBPARTITION */
 
-#ifdef __KERNEL__
-
 #define ADDPART_FLAG_NONE	0
 #define ADDPART_FLAG_RAID	1
 #define ADDPART_FLAG_WHOLEDISK	2
@@ -570,8 +548,6 @@ static inline struct block_device *bdget_disk(struct gendisk *disk, int index)
 	return bdget(MKDEV(disk->major, disk->first_minor) + index);
 }
 
-#endif
-
 #else /* CONFIG_BLOCK */
 
 static inline void printk_all_partitions(void) { }
@@ -584,4 +560,4 @@ static inline dev_t blk_lookup_devt(const char *name)
 
 #endif /* CONFIG_BLOCK */
 
-#endif
+#endif /* _LINUX_GENHD_H */
-- 
cgit v1.2.3


From 07c941d00087581c9553661c2c4fb593da37f525 Mon Sep 17 00:00:00 2001
From: Tony Breeds <tony@bakeyournoodle.com>
Date: Wed, 12 Mar 2008 10:48:48 +1100
Subject: [POWERPC] Fix undefined pmu_sys_suspended compilation error

pmu_sys_suspended is declared extern when:
	defined(CONFIG_PM_SLEEP) && defined(CONFIG_PPC32)
but only defined when:
	defined(CONFIG_SUSPEND) && defined(CONFIG_PPC32)
which is wrong.  Let's fix that.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/linux/pmu.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/pmu.h b/include/linux/pmu.h
index 4c5f65392d36..b02b57c0fba0 100644
--- a/include/linux/pmu.h
+++ b/include/linux/pmu.h
@@ -192,7 +192,7 @@ extern unsigned int pmu_power_flags;
 extern void pmu_backlight_init(void);
 
 /* some code needs to know if the PMU was suspended for hibernation */
-#if defined(CONFIG_PM_SLEEP) && defined(CONFIG_PPC32)
+#if defined(CONFIG_SUSPEND) && defined(CONFIG_PPC32)
 extern int pmu_sys_suspended;
 #else
 /* if power management is not configured it can't be suspended */
-- 
cgit v1.2.3


From a99d9a6ebdf8328d5c61ca9f1038f4815e25720e Mon Sep 17 00:00:00 2001
From: Tony Breeds <tony@bakeyournoodle.com>
Date: Wed, 12 Mar 2008 10:48:48 +1100
Subject: [POWERPC] Fix drivers/macintosh/mediabay.c when !CONFIG_ADB_PMU

When building drivers/macintosh/mediabay.c if CONFIG_ADB_PMU isn't
defined we get:

drivers/built-in.o: In function `media_bay_step':
mediabay.c:(.text+0x92b84): undefined reference to `pmu_suspend'
mediabay.c:(.text+0x92c08): undefined reference to `pmu_resume'

Create empty place holders in that scenario.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/linux/pmu.h | 7 +++++++
 1 file changed, 7 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/pmu.h b/include/linux/pmu.h
index b02b57c0fba0..cafe98d96948 100644
--- a/include/linux/pmu.h
+++ b/include/linux/pmu.h
@@ -147,8 +147,15 @@ extern void pmu_wait_complete(struct adb_request *req);
 /* For use before switching interrupts off for a long time;
  * warning: not stackable
  */
+#if defined(CONFIG_ADB_PMU)
 extern void pmu_suspend(void);
 extern void pmu_resume(void);
+#else
+static inline void pmu_suspend(void)
+{}
+static inline void pmu_resume(void)
+{}
+#endif
 
 extern void pmu_enable_irled(int on);
 
-- 
cgit v1.2.3


From cc74d96f47b0d916840f92092595e3be9731e047 Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Wed, 12 Mar 2008 21:48:03 -0700
Subject: PCI: fix issue with busses registering multiple times in sysfs

PCI busses can be registered multiple times, so we need to detect if we
have registered our bus structure in sysfs already.  If so, don't do it
again.

Thanks to Guennadi Liakhovetski <g.liakhovetski@gmx.de> for reporting
the problem, and to Linus for poking me to get me to believe that it was
a real problem.

Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/bus.c   | 6 +++++-
 include/linux/pci.h | 1 +
 2 files changed, 6 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c
index 6a9403d79e0c..d708358326e5 100644
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -143,14 +143,18 @@ void pci_bus_add_devices(struct pci_bus *bus)
 			/* register the bus with sysfs as the parent is now
 			 * properly registered. */
 			child_bus = dev->subordinate;
+			if (child_bus->is_added)
+				continue;
 			child_bus->dev.parent = child_bus->bridge;
 			retval = device_register(&child_bus->dev);
 			if (retval)
 				dev_err(&dev->dev, "Error registering pci_bus,"
 					" continuing...\n");
-			else
+			else {
+				child_bus->is_added = 1;
 				retval = device_create_file(&child_bus->dev,
 							&dev_attr_cpuaffinity);
+			}
 			if (retval)
 				dev_err(&dev->dev, "Error creating cpuaffinity"
 					" file, continuing...\n");
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 38eff1947750..9010f5458767 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -278,6 +278,7 @@ struct pci_bus {
 	struct device		dev;
 	struct bin_attribute	*legacy_io; /* legacy I/O for this bus */
 	struct bin_attribute	*legacy_mem; /* legacy mem */
+	unsigned int		is_added:1;
 };
 
 #define pci_bus_b(n)	list_entry(n, struct pci_bus, node)
-- 
cgit v1.2.3


From e7ec2e3230633a858af1b0b359f6c4670dbeb997 Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Mon, 10 Mar 2008 17:26:32 +0100
Subject: ssb: Add SPROM/invariants support for PCMCIA devices

This adds support for reading/writing the SPROM invariants
for PCMCIA based devices.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/Kconfig       |   6 +
 drivers/ssb/Makefile      |   1 +
 drivers/ssb/main.c        |  23 +-
 drivers/ssb/pci.c         | 113 +---------
 drivers/ssb/pcmcia.c      | 518 ++++++++++++++++++++++++++++++++++++++++------
 drivers/ssb/sprom.c       | 133 ++++++++++++
 drivers/ssb/ssb_private.h |  17 ++
 include/linux/ssb/ssb.h   |   4 +-
 8 files changed, 648 insertions(+), 167 deletions(-)
 create mode 100644 drivers/ssb/sprom.c

(limited to 'include/linux')

diff --git a/drivers/ssb/Kconfig b/drivers/ssb/Kconfig
index f69ef0ba2613..0f7cce2560d1 100644
--- a/drivers/ssb/Kconfig
+++ b/drivers/ssb/Kconfig
@@ -20,6 +20,10 @@ config SSB
 
 	  If unsure, say N.
 
+# Common SPROM support routines
+config SSB_SPROM
+	bool
+
 config SSB_PCIHOST_POSSIBLE
 	bool
 	depends on SSB && (PCI = y || PCI = SSB)
@@ -28,6 +32,7 @@ config SSB_PCIHOST_POSSIBLE
 config SSB_PCIHOST
 	bool "Support for SSB on PCI-bus host"
 	depends on SSB_PCIHOST_POSSIBLE
+	select SSB_SPROM
 	default y
 	help
 	  Support for a Sonics Silicon Backplane on top
@@ -48,6 +53,7 @@ config SSB_PCMCIAHOST_POSSIBLE
 config SSB_PCMCIAHOST
 	bool "Support for SSB on PCMCIA-bus host (EXPERIMENTAL)"
 	depends on SSB_PCMCIAHOST_POSSIBLE
+	select SSB_SPROM
 	help
 	  Support for a Sonics Silicon Backplane on top
 	  of a PCMCIA device.
diff --git a/drivers/ssb/Makefile b/drivers/ssb/Makefile
index 910f35e32fc9..6f255e9c5af9 100644
--- a/drivers/ssb/Makefile
+++ b/drivers/ssb/Makefile
@@ -1,6 +1,7 @@
 # core
 ssb-y					+= main.o scan.o
 ssb-$(CONFIG_SSB_EMBEDDED)		+= embedded.o
+ssb-$(CONFIG_SSB_SPROM)			+= sprom.o
 
 # host support
 ssb-$(CONFIG_SSB_PCIHOST)		+= pci.o pcihost_wrapper.o
diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index 49d7bbb9bea7..e12371916444 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -69,6 +69,25 @@ found:
 }
 #endif /* CONFIG_SSB_PCIHOST */
 
+#ifdef CONFIG_SSB_PCMCIAHOST
+struct ssb_bus *ssb_pcmcia_dev_to_bus(struct pcmcia_device *pdev)
+{
+	struct ssb_bus *bus;
+
+	ssb_buses_lock();
+	list_for_each_entry(bus, &buses, list) {
+		if (bus->bustype == SSB_BUSTYPE_PCMCIA &&
+		    bus->host_pcmcia == pdev)
+			goto found;
+	}
+	bus = NULL;
+found:
+	ssb_buses_unlock();
+
+	return bus;
+}
+#endif /* CONFIG_SSB_PCMCIAHOST */
+
 int ssb_for_each_bus_call(unsigned long data,
 			  int (*func)(struct ssb_bus *bus, unsigned long data))
 {
@@ -398,7 +417,7 @@ void ssb_bus_unregister(struct ssb_bus *bus)
 	list_del(&bus->list);
 	ssb_buses_unlock();
 
-	/* ssb_pcmcia_exit(bus); */
+	ssb_pcmcia_exit(bus);
 	ssb_pci_exit(bus);
 	ssb_iounmap(bus);
 }
@@ -663,7 +682,7 @@ out:
 err_dequeue:
 	list_del(&bus->list);
 err_pcmcia_exit:
-/*	ssb_pcmcia_exit(bus); */
+	ssb_pcmcia_exit(bus);
 err_pci_exit:
 	ssb_pci_exit(bus);
 err_unmap:
diff --git a/drivers/ssb/pci.c b/drivers/ssb/pci.c
index 1facc7620fc8..f1514b33cfae 100644
--- a/drivers/ssb/pci.c
+++ b/drivers/ssb/pci.c
@@ -227,7 +227,7 @@ static u8 ssb_sprom_crc(const u16 *sprom, u16 size)
 	return crc;
 }
 
-static int sprom_check_crc(const u16 *sprom, u16 size)
+static int sprom_check_crc(const u16 *sprom, size_t size)
 {
 	u8 crc;
 	u8 expected_crc;
@@ -242,12 +242,14 @@ static int sprom_check_crc(const u16 *sprom, u16 size)
 	return 0;
 }
 
-static void sprom_do_read(struct ssb_bus *bus, u16 *sprom)
+static int sprom_do_read(struct ssb_bus *bus, u16 *sprom)
 {
 	int i;
 
 	for (i = 0; i < bus->sprom_size; i++)
 		sprom[i] = ioread16(bus->mmio + SSB_SPROM_BASE + (i * 2));
+
+	return 0;
 }
 
 static int sprom_do_write(struct ssb_bus *bus, const u16 *sprom)
@@ -660,71 +662,18 @@ const struct ssb_bus_ops ssb_pci_ops = {
 	.write32	= ssb_pci_write32,
 };
 
-static int sprom2hex(const u16 *sprom, char *buf, size_t buf_len, u16 size)
-{
-	int i, pos = 0;
-
-	for (i = 0; i < size; i++)
-		pos += snprintf(buf + pos, buf_len - pos - 1,
-				"%04X", swab16(sprom[i]) & 0xFFFF);
-	pos += snprintf(buf + pos, buf_len - pos - 1, "\n");
-
-	return pos + 1;
-}
-
-static int hex2sprom(u16 *sprom, const char *dump, size_t len, u16 size)
-{
-	char tmp[5] = { 0 };
-	int cnt = 0;
-	unsigned long parsed;
-
-	if (len < size * 2)
-		return -EINVAL;
-
-	while (cnt < size) {
-		memcpy(tmp, dump, 4);
-		dump += 4;
-		parsed = simple_strtoul(tmp, NULL, 16);
-		sprom[cnt++] = swab16((u16)parsed);
-	}
-
-	return 0;
-}
-
 static ssize_t ssb_pci_attr_sprom_show(struct device *pcidev,
 				       struct device_attribute *attr,
 				       char *buf)
 {
 	struct pci_dev *pdev = container_of(pcidev, struct pci_dev, dev);
 	struct ssb_bus *bus;
-	u16 *sprom;
-	int err = -ENODEV;
-	ssize_t count = 0;
 
 	bus = ssb_pci_dev_to_bus(pdev);
 	if (!bus)
-		goto out;
-	err = -ENOMEM;
-	sprom = kcalloc(bus->sprom_size, sizeof(u16), GFP_KERNEL);
-	if (!sprom)
-		goto out;
+		return -ENODEV;
 
-	/* Use interruptible locking, as the SPROM write might
-	 * be holding the lock for several seconds. So allow userspace
-	 * to cancel operation. */
-	err = -ERESTARTSYS;
-	if (mutex_lock_interruptible(&bus->pci_sprom_mutex))
-		goto out_kfree;
-	sprom_do_read(bus, sprom);
-	mutex_unlock(&bus->pci_sprom_mutex);
-
-	count = sprom2hex(sprom, buf, PAGE_SIZE, bus->sprom_size);
-	err = 0;
-
-out_kfree:
-	kfree(sprom);
-out:
-	return err ? err : count;
+	return ssb_attr_sprom_show(bus, buf, sprom_do_read);
 }
 
 static ssize_t ssb_pci_attr_sprom_store(struct device *pcidev,
@@ -733,55 +682,13 @@ static ssize_t ssb_pci_attr_sprom_store(struct device *pcidev,
 {
 	struct pci_dev *pdev = container_of(pcidev, struct pci_dev, dev);
 	struct ssb_bus *bus;
-	u16 *sprom;
-	int res = 0, err = -ENODEV;
 
 	bus = ssb_pci_dev_to_bus(pdev);
 	if (!bus)
-		goto out;
-	err = -ENOMEM;
-	sprom = kcalloc(bus->sprom_size, sizeof(u16), GFP_KERNEL);
-	if (!sprom)
-		goto out;
-	err = hex2sprom(sprom, buf, count, bus->sprom_size);
-	if (err) {
-		err = -EINVAL;
-		goto out_kfree;
-	}
-	err = sprom_check_crc(sprom, bus->sprom_size);
-	if (err) {
-		err = -EINVAL;
-		goto out_kfree;
-	}
+		return -ENODEV;
 
-	/* Use interruptible locking, as the SPROM write might
-	 * be holding the lock for several seconds. So allow userspace
-	 * to cancel operation. */
-	err = -ERESTARTSYS;
-	if (mutex_lock_interruptible(&bus->pci_sprom_mutex))
-		goto out_kfree;
-	err = ssb_devices_freeze(bus);
-	if (err == -EOPNOTSUPP) {
-		ssb_printk(KERN_ERR PFX "SPROM write: Could not freeze devices. "
-			   "No suspend support. Is CONFIG_PM enabled?\n");
-		goto out_unlock;
-	}
-	if (err) {
-		ssb_printk(KERN_ERR PFX "SPROM write: Could not freeze all devices\n");
-		goto out_unlock;
-	}
-	res = sprom_do_write(bus, sprom);
-	err = ssb_devices_thaw(bus);
-	if (err)
-		ssb_printk(KERN_ERR PFX "SPROM write: Could not thaw all devices\n");
-out_unlock:
-	mutex_unlock(&bus->pci_sprom_mutex);
-out_kfree:
-	kfree(sprom);
-out:
-	if (res)
-		return res;
-	return err ? err : count;
+	return ssb_attr_sprom_store(bus, buf, count,
+				    sprom_check_crc, sprom_do_write);
 }
 
 static DEVICE_ATTR(ssb_sprom, 0600,
@@ -808,7 +715,7 @@ int ssb_pci_init(struct ssb_bus *bus)
 		return 0;
 
 	pdev = bus->host_pci;
-	mutex_init(&bus->pci_sprom_mutex);
+	mutex_init(&bus->sprom_mutex);
 	err = device_create_file(&pdev->dev, &dev_attr_ssb_sprom);
 	if (err)
 		goto out;
diff --git a/drivers/ssb/pcmcia.c b/drivers/ssb/pcmcia.c
index 84b3a845a8a8..cd49f7c65531 100644
--- a/drivers/ssb/pcmcia.c
+++ b/drivers/ssb/pcmcia.c
@@ -3,7 +3,7 @@
  * PCMCIA-Hostbus related functions
  *
  * Copyright 2006 Johannes Berg <johannes@sipsolutions.net>
- * Copyright 2007 Michael Buesch <mb@bu3sch.de>
+ * Copyright 2007-2008 Michael Buesch <mb@bu3sch.de>
  *
  * Licensed under the GNU/GPL. See COPYING for details.
  */
@@ -11,6 +11,7 @@
 #include <linux/ssb/ssb.h>
 #include <linux/delay.h>
 #include <linux/io.h>
+#include <linux/etherdevice.h>
 
 #include <pcmcia/cs_types.h>
 #include <pcmcia/cs.h>
@@ -26,59 +27,132 @@
 #define SSB_VERBOSE_PCMCIACORESWITCH_DEBUG		0
 
 
+/* PCMCIA configuration registers */
+#define SSB_PCMCIA_CORECTL		0x00
+#define  SSB_PCMCIA_CORECTL_RESET	0x80 /* Core reset */
+#define  SSB_PCMCIA_CORECTL_IRQEN	0x04 /* IRQ enable */
+#define  SSB_PCMCIA_CORECTL_FUNCEN	0x01 /* Function enable */
+#define SSB_PCMCIA_CORECTL2		0x80
+#define SSB_PCMCIA_ADDRESS0		0x2E
+#define SSB_PCMCIA_ADDRESS1		0x30
+#define SSB_PCMCIA_ADDRESS2		0x32
+#define SSB_PCMCIA_MEMSEG		0x34
+#define SSB_PCMCIA_SPROMCTL		0x36
+#define  SSB_PCMCIA_SPROMCTL_IDLE	0
+#define  SSB_PCMCIA_SPROMCTL_WRITE	1
+#define  SSB_PCMCIA_SPROMCTL_READ	2
+#define  SSB_PCMCIA_SPROMCTL_WRITEEN	4
+#define  SSB_PCMCIA_SPROMCTL_WRITEDIS	7
+#define  SSB_PCMCIA_SPROMCTL_DONE	8
+#define SSB_PCMCIA_SPROM_DATALO		0x38
+#define SSB_PCMCIA_SPROM_DATAHI		0x3A
+#define SSB_PCMCIA_SPROM_ADDRLO		0x3C
+#define SSB_PCMCIA_SPROM_ADDRHI		0x3E
+
+/* Hardware invariants CIS tuples */
+#define SSB_PCMCIA_CIS			0x80
+#define  SSB_PCMCIA_CIS_ID		0x01
+#define  SSB_PCMCIA_CIS_BOARDREV	0x02
+#define  SSB_PCMCIA_CIS_PA		0x03
+#define   SSB_PCMCIA_CIS_PA_PA0B0_LO	0
+#define   SSB_PCMCIA_CIS_PA_PA0B0_HI	1
+#define   SSB_PCMCIA_CIS_PA_PA0B1_LO	2
+#define   SSB_PCMCIA_CIS_PA_PA0B1_HI	3
+#define   SSB_PCMCIA_CIS_PA_PA0B2_LO	4
+#define   SSB_PCMCIA_CIS_PA_PA0B2_HI	5
+#define   SSB_PCMCIA_CIS_PA_ITSSI	6
+#define   SSB_PCMCIA_CIS_PA_MAXPOW	7
+#define  SSB_PCMCIA_CIS_OEMNAME		0x04
+#define  SSB_PCMCIA_CIS_CCODE		0x05
+#define  SSB_PCMCIA_CIS_ANTENNA		0x06
+#define  SSB_PCMCIA_CIS_ANTGAIN		0x07
+#define  SSB_PCMCIA_CIS_BFLAGS		0x08
+#define  SSB_PCMCIA_CIS_LEDS		0x09
+
+/* PCMCIA SPROM size. */
+#define SSB_PCMCIA_SPROM_SIZE		256
+#define SSB_PCMCIA_SPROM_SIZE_BYTES	(SSB_PCMCIA_SPROM_SIZE * sizeof(u16))
+
+
+/* Write to a PCMCIA configuration register. */
+static int ssb_pcmcia_cfg_write(struct ssb_bus *bus, u8 offset, u8 value)
+{
+	conf_reg_t reg;
+	int res;
+
+	memset(&reg, 0, sizeof(reg));
+	reg.Offset = offset;
+	reg.Action = CS_WRITE;
+	reg.Value = value;
+	res = pcmcia_access_configuration_register(bus->host_pcmcia, &reg);
+	if (unlikely(res != CS_SUCCESS))
+		return -EBUSY;
+
+	return 0;
+}
+
+/* Read from a PCMCIA configuration register. */
+static int ssb_pcmcia_cfg_read(struct ssb_bus *bus, u8 offset, u8 *value)
+{
+	conf_reg_t reg;
+	int res;
+
+	memset(&reg, 0, sizeof(reg));
+	reg.Offset = offset;
+	reg.Action = CS_READ;
+	res = pcmcia_access_configuration_register(bus->host_pcmcia, &reg);
+	if (unlikely(res != CS_SUCCESS))
+		return -EBUSY;
+	*value = reg.Value;
+
+	return 0;
+}
+
 int ssb_pcmcia_switch_coreidx(struct ssb_bus *bus,
 			      u8 coreidx)
 {
-	struct pcmcia_device *pdev = bus->host_pcmcia;
 	int err;
 	int attempts = 0;
 	u32 cur_core;
-	conf_reg_t reg;
 	u32 addr;
 	u32 read_addr;
+	u8 val;
 
 	addr = (coreidx * SSB_CORE_SIZE) + SSB_ENUM_BASE;
 	while (1) {
-		reg.Action = CS_WRITE;
-		reg.Offset = 0x2E;
-		reg.Value = (addr & 0x0000F000) >> 12;
-		err = pcmcia_access_configuration_register(pdev, &reg);
-		if (err != CS_SUCCESS)
+		err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_ADDRESS0,
+					   (addr & 0x0000F000) >> 12);
+		if (err)
 			goto error;
-		reg.Offset = 0x30;
-		reg.Value = (addr & 0x00FF0000) >> 16;
-		err = pcmcia_access_configuration_register(pdev, &reg);
-		if (err != CS_SUCCESS)
+		err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_ADDRESS1,
+					   (addr & 0x00FF0000) >> 16);
+		if (err)
 			goto error;
-		reg.Offset = 0x32;
-		reg.Value = (addr & 0xFF000000) >> 24;
-		err = pcmcia_access_configuration_register(pdev, &reg);
-		if (err != CS_SUCCESS)
+		err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_ADDRESS2,
+					   (addr & 0xFF000000) >> 24);
+		if (err)
 			goto error;
 
 		read_addr = 0;
 
-		reg.Action = CS_READ;
-		reg.Offset = 0x2E;
-		err = pcmcia_access_configuration_register(pdev, &reg);
-		if (err != CS_SUCCESS)
+		err = ssb_pcmcia_cfg_read(bus, SSB_PCMCIA_ADDRESS0, &val);
+		if (err)
 			goto error;
-		read_addr |= ((u32)(reg.Value & 0x0F)) << 12;
-		reg.Offset = 0x30;
-		err = pcmcia_access_configuration_register(pdev, &reg);
-		if (err != CS_SUCCESS)
+		read_addr |= ((u32)(val & 0x0F)) << 12;
+		err = ssb_pcmcia_cfg_read(bus, SSB_PCMCIA_ADDRESS1, &val);
+		if (err)
 			goto error;
-		read_addr |= ((u32)reg.Value) << 16;
-		reg.Offset = 0x32;
-		err = pcmcia_access_configuration_register(pdev, &reg);
-		if (err != CS_SUCCESS)
+		read_addr |= ((u32)val) << 16;
+		err = ssb_pcmcia_cfg_read(bus, SSB_PCMCIA_ADDRESS2, &val);
+		if (err)
 			goto error;
-		read_addr |= ((u32)reg.Value) << 24;
+		read_addr |= ((u32)val) << 24;
 
 		cur_core = (read_addr - SSB_ENUM_BASE) / SSB_CORE_SIZE;
 		if (cur_core == coreidx)
 			break;
 
+		err = -ETIMEDOUT;
 		if (attempts++ > SSB_BAR0_MAX_RETRIES)
 			goto error;
 		udelay(10);
@@ -87,7 +161,7 @@ int ssb_pcmcia_switch_coreidx(struct ssb_bus *bus,
 	return 0;
 error:
 	ssb_printk(KERN_ERR PFX "Failed to switch to core %u\n", coreidx);
-	return -ENODEV;
+	return err;
 }
 
 int ssb_pcmcia_switch_core(struct ssb_bus *bus,
@@ -112,27 +186,21 @@ int ssb_pcmcia_switch_core(struct ssb_bus *bus,
 int ssb_pcmcia_switch_segment(struct ssb_bus *bus, u8 seg)
 {
 	int attempts = 0;
-	conf_reg_t reg;
-	int res;
+	int err;
+	u8 val;
 
 	SSB_WARN_ON((seg != 0) && (seg != 1));
-	reg.Offset = 0x34;
-	reg.Function = 0;
 	while (1) {
-		reg.Action = CS_WRITE;
-		reg.Value = seg;
-		res = pcmcia_access_configuration_register(bus->host_pcmcia, &reg);
-		if (unlikely(res != CS_SUCCESS))
+		err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_MEMSEG, seg);
+		if (err)
 			goto error;
-		reg.Value = 0xFF;
-		reg.Action = CS_READ;
-		res = pcmcia_access_configuration_register(bus->host_pcmcia, &reg);
-		if (unlikely(res != CS_SUCCESS))
+		err = ssb_pcmcia_cfg_read(bus, SSB_PCMCIA_MEMSEG, &val);
+		if (err)
 			goto error;
-
-		if (reg.Value == seg)
+		if (val == seg)
 			break;
 
+		err = -ETIMEDOUT;
 		if (unlikely(attempts++ > SSB_BAR0_MAX_RETRIES))
 			goto error;
 		udelay(10);
@@ -142,7 +210,7 @@ int ssb_pcmcia_switch_segment(struct ssb_bus *bus, u8 seg)
 	return 0;
 error:
 	ssb_printk(KERN_ERR PFX "Failed to switch pcmcia segment\n");
-	return -ENODEV;
+	return err;
 }
 
 static int select_core_and_segment(struct ssb_device *dev,
@@ -276,18 +344,344 @@ const struct ssb_bus_ops ssb_pcmcia_ops = {
 	.write32	= ssb_pcmcia_write32,
 };
 
-#include <linux/etherdevice.h>
+static int ssb_pcmcia_sprom_command(struct ssb_bus *bus, u8 command)
+{
+	unsigned int i;
+	int err;
+	u8 value;
+
+	err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_SPROMCTL, command);
+	if (err)
+		return err;
+	for (i = 0; i < 1000; i++) {
+		err = ssb_pcmcia_cfg_read(bus, SSB_PCMCIA_SPROMCTL, &value);
+		if (err)
+			return err;
+		if (value & SSB_PCMCIA_SPROMCTL_DONE)
+			return 0;
+		udelay(10);
+	}
+
+	return -ETIMEDOUT;
+}
+
+/* offset is the 16bit word offset */
+static int ssb_pcmcia_sprom_read(struct ssb_bus *bus, u16 offset, u16 *value)
+{
+	int err;
+	u8 lo, hi;
+
+	offset *= 2; /* Make byte offset */
+
+	err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_SPROM_ADDRLO,
+				   (offset & 0x00FF));
+	if (err)
+		return err;
+	err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_SPROM_ADDRHI,
+				   (offset & 0xFF00) >> 8);
+	if (err)
+		return err;
+	err = ssb_pcmcia_sprom_command(bus, SSB_PCMCIA_SPROMCTL_READ);
+	if (err)
+		return err;
+	err = ssb_pcmcia_cfg_read(bus, SSB_PCMCIA_SPROM_DATALO, &lo);
+	if (err)
+		return err;
+	err = ssb_pcmcia_cfg_read(bus, SSB_PCMCIA_SPROM_DATAHI, &hi);
+	if (err)
+		return err;
+	*value = (lo | (((u16)hi) << 8));
+
+	return 0;
+}
+
+/* offset is the 16bit word offset */
+static int ssb_pcmcia_sprom_write(struct ssb_bus *bus, u16 offset, u16 value)
+{
+	int err;
+
+	offset *= 2; /* Make byte offset */
+
+	err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_SPROM_ADDRLO,
+				   (offset & 0x00FF));
+	if (err)
+		return err;
+	err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_SPROM_ADDRHI,
+				   (offset & 0xFF00) >> 8);
+	if (err)
+		return err;
+	err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_SPROM_DATALO,
+				   (value & 0x00FF));
+	if (err)
+		return err;
+	err = ssb_pcmcia_cfg_write(bus, SSB_PCMCIA_SPROM_DATAHI,
+				   (value & 0xFF00) >> 8);
+	if (err)
+		return err;
+	err = ssb_pcmcia_sprom_command(bus, SSB_PCMCIA_SPROMCTL_WRITE);
+	if (err)
+		return err;
+	msleep(20);
+
+	return 0;
+}
+
+/* Read the SPROM image. bufsize is in 16bit words. */
+static int ssb_pcmcia_sprom_read_all(struct ssb_bus *bus, u16 *sprom)
+{
+	int err, i;
+
+	for (i = 0; i < SSB_PCMCIA_SPROM_SIZE; i++) {
+		err = ssb_pcmcia_sprom_read(bus, i, &sprom[i]);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+/* Write the SPROM image. size is in 16bit words. */
+static int ssb_pcmcia_sprom_write_all(struct ssb_bus *bus, const u16 *sprom)
+{
+	int i, err;
+	bool failed = 0;
+	size_t size = SSB_PCMCIA_SPROM_SIZE;
+
+	ssb_printk(KERN_NOTICE PFX
+		   "Writing SPROM. Do NOT turn off the power! "
+		   "Please stand by...\n");
+	err = ssb_pcmcia_sprom_command(bus, SSB_PCMCIA_SPROMCTL_WRITEEN);
+	if (err) {
+		ssb_printk(KERN_NOTICE PFX
+			   "Could not enable SPROM write access.\n");
+		return -EBUSY;
+	}
+	ssb_printk(KERN_NOTICE PFX "[ 0%%");
+	msleep(500);
+	for (i = 0; i < size; i++) {
+		if (i == size / 4)
+			ssb_printk("25%%");
+		else if (i == size / 2)
+			ssb_printk("50%%");
+		else if (i == (size * 3) / 4)
+			ssb_printk("75%%");
+		else if (i % 2)
+			ssb_printk(".");
+		err = ssb_pcmcia_sprom_write(bus, i, sprom[i]);
+		if (err) {
+			ssb_printk("\n" KERN_NOTICE PFX
+				   "Failed to write to SPROM.\n");
+			failed = 1;
+			break;
+		}
+	}
+	err = ssb_pcmcia_sprom_command(bus, SSB_PCMCIA_SPROMCTL_WRITEDIS);
+	if (err) {
+		ssb_printk("\n" KERN_NOTICE PFX
+			   "Could not disable SPROM write access.\n");
+		failed = 1;
+	}
+	msleep(500);
+	if (!failed) {
+		ssb_printk("100%% ]\n");
+		ssb_printk(KERN_NOTICE PFX "SPROM written.\n");
+	}
+
+	return failed ? -EBUSY : 0;
+}
+
+static int ssb_pcmcia_sprom_check_crc(const u16 *sprom, size_t size)
+{
+	//TODO
+	return 0;
+}
+
+#define GOTO_ERROR_ON(condition, description) do {	\
+	if (unlikely(condition)) {			\
+		error_description = description;	\
+		goto error;				\
+	}						\
+  } while (0)
+
 int ssb_pcmcia_get_invariants(struct ssb_bus *bus,
 			      struct ssb_init_invariants *iv)
 {
-	//TODO
-	random_ether_addr(iv->sprom.il0mac);
+	tuple_t tuple;
+	int res;
+	unsigned char buf[32];
+	struct ssb_sprom *sprom = &iv->sprom;
+	struct ssb_boardinfo *bi = &iv->boardinfo;
+	const char *error_description;
+
+	memset(sprom, 0xFF, sizeof(*sprom));
+	sprom->revision = 1;
+	sprom->boardflags_lo = 0;
+	sprom->boardflags_hi = 0;
+
+	/* First fetch the MAC address. */
+	memset(&tuple, 0, sizeof(tuple));
+	tuple.DesiredTuple = CISTPL_FUNCE;
+	tuple.TupleData = buf;
+	tuple.TupleDataMax = sizeof(buf);
+	res = pcmcia_get_first_tuple(bus->host_pcmcia, &tuple);
+	GOTO_ERROR_ON(res != CS_SUCCESS, "MAC first tpl");
+	res = pcmcia_get_tuple_data(bus->host_pcmcia, &tuple);
+	GOTO_ERROR_ON(res != CS_SUCCESS, "MAC first tpl data");
+	while (1) {
+		GOTO_ERROR_ON(tuple.TupleDataLen < 1, "MAC tpl < 1");
+		if (tuple.TupleData[0] == CISTPL_FUNCE_LAN_NODE_ID)
+			break;
+		res = pcmcia_get_next_tuple(bus->host_pcmcia, &tuple);
+		GOTO_ERROR_ON(res != CS_SUCCESS, "MAC next tpl");
+		res = pcmcia_get_tuple_data(bus->host_pcmcia, &tuple);
+		GOTO_ERROR_ON(res != CS_SUCCESS, "MAC next tpl data");
+	}
+	GOTO_ERROR_ON(tuple.TupleDataLen != ETH_ALEN + 2, "MAC tpl size");
+	memcpy(sprom->il0mac, &tuple.TupleData[2], ETH_ALEN);
+
+	/* Fetch the vendor specific tuples. */
+	memset(&tuple, 0, sizeof(tuple));
+	tuple.DesiredTuple = SSB_PCMCIA_CIS;
+	tuple.TupleData = buf;
+	tuple.TupleDataMax = sizeof(buf);
+	res = pcmcia_get_first_tuple(bus->host_pcmcia, &tuple);
+	GOTO_ERROR_ON(res != CS_SUCCESS, "VEN first tpl");
+	res = pcmcia_get_tuple_data(bus->host_pcmcia, &tuple);
+	GOTO_ERROR_ON(res != CS_SUCCESS, "VEN first tpl data");
+	while (1) {
+		GOTO_ERROR_ON(tuple.TupleDataLen < 1, "VEN tpl < 1");
+		switch (tuple.TupleData[0]) {
+		case SSB_PCMCIA_CIS_ID:
+			GOTO_ERROR_ON((tuple.TupleDataLen != 5) &&
+				      (tuple.TupleDataLen != 7),
+				      "id tpl size");
+			bi->vendor = tuple.TupleData[1] |
+			       ((u16)tuple.TupleData[2] << 8);
+			break;
+		case SSB_PCMCIA_CIS_BOARDREV:
+			GOTO_ERROR_ON(tuple.TupleDataLen != 2,
+				      "boardrev tpl size");
+			sprom->board_rev = tuple.TupleData[1];
+			break;
+		case SSB_PCMCIA_CIS_PA:
+			GOTO_ERROR_ON(tuple.TupleDataLen != 9,
+				      "pa tpl size");
+			sprom->pa0b0 = tuple.TupleData[1] |
+				 ((u16)tuple.TupleData[2] << 8);
+			sprom->pa0b1 = tuple.TupleData[3] |
+				 ((u16)tuple.TupleData[4] << 8);
+			sprom->pa0b2 = tuple.TupleData[5] |
+				 ((u16)tuple.TupleData[6] << 8);
+			sprom->itssi_a = tuple.TupleData[7];
+			sprom->itssi_bg = tuple.TupleData[7];
+			sprom->maxpwr_a = tuple.TupleData[8];
+			sprom->maxpwr_bg = tuple.TupleData[8];
+			break;
+		case SSB_PCMCIA_CIS_OEMNAME:
+			/* We ignore this. */
+			break;
+		case SSB_PCMCIA_CIS_CCODE:
+			GOTO_ERROR_ON(tuple.TupleDataLen != 2,
+				      "ccode tpl size");
+			sprom->country_code = tuple.TupleData[1];
+			break;
+		case SSB_PCMCIA_CIS_ANTENNA:
+			GOTO_ERROR_ON(tuple.TupleDataLen != 2,
+				      "ant tpl size");
+			sprom->ant_available_a = tuple.TupleData[1];
+			sprom->ant_available_bg = tuple.TupleData[1];
+			break;
+		case SSB_PCMCIA_CIS_ANTGAIN:
+			GOTO_ERROR_ON(tuple.TupleDataLen != 2,
+				      "antg tpl size");
+			sprom->antenna_gain.ghz24.a0 = tuple.TupleData[1];
+			sprom->antenna_gain.ghz24.a1 = tuple.TupleData[1];
+			sprom->antenna_gain.ghz24.a2 = tuple.TupleData[1];
+			sprom->antenna_gain.ghz24.a3 = tuple.TupleData[1];
+			sprom->antenna_gain.ghz5.a0 = tuple.TupleData[1];
+			sprom->antenna_gain.ghz5.a1 = tuple.TupleData[1];
+			sprom->antenna_gain.ghz5.a2 = tuple.TupleData[1];
+			sprom->antenna_gain.ghz5.a3 = tuple.TupleData[1];
+			break;
+		case SSB_PCMCIA_CIS_BFLAGS:
+			GOTO_ERROR_ON(tuple.TupleDataLen != 3,
+				      "bfl tpl size");
+			sprom->boardflags_lo = tuple.TupleData[1] |
+					 ((u16)tuple.TupleData[2] << 8);
+			break;
+		case SSB_PCMCIA_CIS_LEDS:
+			GOTO_ERROR_ON(tuple.TupleDataLen != 5,
+				      "leds tpl size");
+			sprom->gpio0 = tuple.TupleData[1];
+			sprom->gpio1 = tuple.TupleData[2];
+			sprom->gpio2 = tuple.TupleData[3];
+			sprom->gpio3 = tuple.TupleData[4];
+			break;
+		}
+		res = pcmcia_get_next_tuple(bus->host_pcmcia, &tuple);
+		if (res == CS_NO_MORE_ITEMS)
+			break;
+		GOTO_ERROR_ON(res != CS_SUCCESS, "VEN next tpl");
+		res = pcmcia_get_tuple_data(bus->host_pcmcia, &tuple);
+		GOTO_ERROR_ON(res != CS_SUCCESS, "VEN next tpl data");
+	}
+
 	return 0;
+error:
+	ssb_printk(KERN_ERR PFX
+		   "PCMCIA: Failed to fetch device invariants: %s\n",
+		   error_description);
+	return -ENODEV;
+}
+
+static ssize_t ssb_pcmcia_attr_sprom_show(struct device *pcmciadev,
+					  struct device_attribute *attr,
+					  char *buf)
+{
+	struct pcmcia_device *pdev =
+		container_of(pcmciadev, struct pcmcia_device, dev);
+	struct ssb_bus *bus;
+
+	bus = ssb_pcmcia_dev_to_bus(pdev);
+	if (!bus)
+		return -ENODEV;
+
+	return ssb_attr_sprom_show(bus, buf,
+				   ssb_pcmcia_sprom_read_all);
+}
+
+static ssize_t ssb_pcmcia_attr_sprom_store(struct device *pcmciadev,
+					   struct device_attribute *attr,
+					   const char *buf, size_t count)
+{
+	struct pcmcia_device *pdev =
+		container_of(pcmciadev, struct pcmcia_device, dev);
+	struct ssb_bus *bus;
+
+	bus = ssb_pcmcia_dev_to_bus(pdev);
+	if (!bus)
+		return -ENODEV;
+
+	return ssb_attr_sprom_store(bus, buf, count,
+				    ssb_pcmcia_sprom_check_crc,
+				    ssb_pcmcia_sprom_write_all);
+}
+
+static DEVICE_ATTR(ssb_sprom, 0600,
+		   ssb_pcmcia_attr_sprom_show,
+		   ssb_pcmcia_attr_sprom_store);
+
+void ssb_pcmcia_exit(struct ssb_bus *bus)
+{
+	if (bus->bustype != SSB_BUSTYPE_PCMCIA)
+		return;
+
+	device_remove_file(&bus->host_pcmcia->dev, &dev_attr_ssb_sprom);
 }
 
 int ssb_pcmcia_init(struct ssb_bus *bus)
 {
-	conf_reg_t reg;
+	u8 val, offset;
 	int err;
 
 	if (bus->bustype != SSB_BUSTYPE_PCMCIA)
@@ -298,22 +692,26 @@ int ssb_pcmcia_init(struct ssb_bus *bus)
 	ssb_pcmcia_switch_segment(bus, 0);
 
 	/* Init IRQ routing */
-	reg.Action = CS_READ;
-	reg.Function = 0;
 	if (bus->chip_id == 0x4306)
-		reg.Offset = 0x00;
+		offset = SSB_PCMCIA_CORECTL;
 	else
-		reg.Offset = 0x80;
-	err = pcmcia_access_configuration_register(bus->host_pcmcia, &reg);
-	if (err != CS_SUCCESS)
+		offset = SSB_PCMCIA_CORECTL2;
+	err = ssb_pcmcia_cfg_read(bus, offset, &val);
+	if (err)
 		goto error;
-	reg.Action = CS_WRITE;
-	reg.Value |= 0x04 | 0x01;
-	err = pcmcia_access_configuration_register(bus->host_pcmcia, &reg);
-	if (err != CS_SUCCESS)
+	val |= SSB_PCMCIA_CORECTL_IRQEN | SSB_PCMCIA_CORECTL_FUNCEN;
+	err = ssb_pcmcia_cfg_write(bus, offset, val);
+	if (err)
+		goto error;
+
+	bus->sprom_size = SSB_PCMCIA_SPROM_SIZE;
+	mutex_init(&bus->sprom_mutex);
+	err = device_create_file(&bus->host_pcmcia->dev, &dev_attr_ssb_sprom);
+	if (err)
 		goto error;
 
 	return 0;
 error:
-	return -ENODEV;
+	ssb_printk(KERN_ERR PFX "Failed to initialize PCMCIA host device\n");
+	return err;
 }
diff --git a/drivers/ssb/sprom.c b/drivers/ssb/sprom.c
new file mode 100644
index 000000000000..3668edb39315
--- /dev/null
+++ b/drivers/ssb/sprom.c
@@ -0,0 +1,133 @@
+/*
+ * Sonics Silicon Backplane
+ * Common SPROM support routines
+ *
+ * Copyright (C) 2005-2008 Michael Buesch <mb@bu3sch.de>
+ * Copyright (C) 2005 Martin Langer <martin-langer@gmx.de>
+ * Copyright (C) 2005 Stefano Brivio <st3@riseup.net>
+ * Copyright (C) 2005 Danny van Dyk <kugelfang@gentoo.org>
+ * Copyright (C) 2005 Andreas Jaggi <andreas.jaggi@waterwave.ch>
+ *
+ * Licensed under the GNU/GPL. See COPYING for details.
+ */
+
+#include "ssb_private.h"
+
+
+static int sprom2hex(const u16 *sprom, char *buf, size_t buf_len,
+		     size_t sprom_size_words)
+{
+	int i, pos = 0;
+
+	for (i = 0; i < sprom_size_words; i++)
+		pos += snprintf(buf + pos, buf_len - pos - 1,
+				"%04X", swab16(sprom[i]) & 0xFFFF);
+	pos += snprintf(buf + pos, buf_len - pos - 1, "\n");
+
+	return pos + 1;
+}
+
+static int hex2sprom(u16 *sprom, const char *dump, size_t len,
+		     size_t sprom_size_words)
+{
+	char tmp[5] = { 0 };
+	int cnt = 0;
+	unsigned long parsed;
+
+	if (len < sprom_size_words * 2)
+		return -EINVAL;
+
+	while (cnt < sprom_size_words) {
+		memcpy(tmp, dump, 4);
+		dump += 4;
+		parsed = simple_strtoul(tmp, NULL, 16);
+		sprom[cnt++] = swab16((u16)parsed);
+	}
+
+	return 0;
+}
+
+/* Common sprom device-attribute show-handler */
+ssize_t ssb_attr_sprom_show(struct ssb_bus *bus, char *buf,
+			    int (*sprom_read)(struct ssb_bus *bus, u16 *sprom))
+{
+	u16 *sprom;
+	int err = -ENOMEM;
+	ssize_t count = 0;
+	size_t sprom_size_words = bus->sprom_size;
+
+	sprom = kcalloc(sprom_size_words, sizeof(u16), GFP_KERNEL);
+	if (!sprom)
+		goto out;
+
+	/* Use interruptible locking, as the SPROM write might
+	 * be holding the lock for several seconds. So allow userspace
+	 * to cancel operation. */
+	err = -ERESTARTSYS;
+	if (mutex_lock_interruptible(&bus->sprom_mutex))
+		goto out_kfree;
+	err = sprom_read(bus, sprom);
+	mutex_unlock(&bus->sprom_mutex);
+
+	if (!err)
+		count = sprom2hex(sprom, buf, PAGE_SIZE, sprom_size_words);
+
+out_kfree:
+	kfree(sprom);
+out:
+	return err ? err : count;
+}
+
+/* Common sprom device-attribute store-handler */
+ssize_t ssb_attr_sprom_store(struct ssb_bus *bus,
+			     const char *buf, size_t count,
+			     int (*sprom_check_crc)(const u16 *sprom, size_t size),
+			     int (*sprom_write)(struct ssb_bus *bus, const u16 *sprom))
+{
+	u16 *sprom;
+	int res = 0, err = -ENOMEM;
+	size_t sprom_size_words = bus->sprom_size;
+
+	sprom = kcalloc(bus->sprom_size, sizeof(u16), GFP_KERNEL);
+	if (!sprom)
+		goto out;
+	err = hex2sprom(sprom, buf, count, sprom_size_words);
+	if (err) {
+		err = -EINVAL;
+		goto out_kfree;
+	}
+	err = sprom_check_crc(sprom, sprom_size_words);
+	if (err) {
+		err = -EINVAL;
+		goto out_kfree;
+	}
+
+	/* Use interruptible locking, as the SPROM write might
+	 * be holding the lock for several seconds. So allow userspace
+	 * to cancel operation. */
+	err = -ERESTARTSYS;
+	if (mutex_lock_interruptible(&bus->sprom_mutex))
+		goto out_kfree;
+	err = ssb_devices_freeze(bus);
+	if (err == -EOPNOTSUPP) {
+		ssb_printk(KERN_ERR PFX "SPROM write: Could not freeze devices. "
+			   "No suspend support. Is CONFIG_PM enabled?\n");
+		goto out_unlock;
+	}
+	if (err) {
+		ssb_printk(KERN_ERR PFX "SPROM write: Could not freeze all devices\n");
+		goto out_unlock;
+	}
+	res = sprom_write(bus, sprom);
+	err = ssb_devices_thaw(bus);
+	if (err)
+		ssb_printk(KERN_ERR PFX "SPROM write: Could not thaw all devices\n");
+out_unlock:
+	mutex_unlock(&bus->sprom_mutex);
+out_kfree:
+	kfree(sprom);
+out:
+	if (res)
+		return res;
+	return err ? err : count;
+}
diff --git a/drivers/ssb/ssb_private.h b/drivers/ssb/ssb_private.h
index d03b20983b1e..a83bf7a4d80b 100644
--- a/drivers/ssb/ssb_private.h
+++ b/drivers/ssb/ssb_private.h
@@ -81,6 +81,7 @@ extern int ssb_pcmcia_switch_segment(struct ssb_bus *bus,
 				     u8 seg);
 extern int ssb_pcmcia_get_invariants(struct ssb_bus *bus,
 				     struct ssb_init_invariants *iv);
+extern void ssb_pcmcia_exit(struct ssb_bus *bus);
 extern int ssb_pcmcia_init(struct ssb_bus *bus);
 extern const struct ssb_bus_ops ssb_pcmcia_ops;
 #else /* CONFIG_SSB_PCMCIAHOST */
@@ -99,6 +100,9 @@ static inline int ssb_pcmcia_switch_segment(struct ssb_bus *bus,
 {
 	return 0;
 }
+static inline void ssb_pcmcia_exit(struct ssb_bus *bus)
+{
+}
 static inline int ssb_pcmcia_init(struct ssb_bus *bus)
 {
 	return 0;
@@ -113,6 +117,17 @@ extern int ssb_bus_scan(struct ssb_bus *bus,
 extern void ssb_iounmap(struct ssb_bus *ssb);
 
 
+/* sprom.c */
+extern
+ssize_t ssb_attr_sprom_show(struct ssb_bus *bus, char *buf,
+			    int (*sprom_read)(struct ssb_bus *bus, u16 *sprom));
+extern
+ssize_t ssb_attr_sprom_store(struct ssb_bus *bus,
+			     const char *buf, size_t count,
+			     int (*sprom_check_crc)(const u16 *sprom, size_t size),
+			     int (*sprom_write)(struct ssb_bus *bus, const u16 *sprom));
+
+
 /* core.c */
 extern u32 ssb_calc_clock_rate(u32 plltype, u32 n, u32 m);
 extern int ssb_devices_freeze(struct ssb_bus *bus);
@@ -120,6 +135,8 @@ extern int ssb_devices_thaw(struct ssb_bus *bus);
 extern struct ssb_bus *ssb_pci_dev_to_bus(struct pci_dev *pdev);
 int ssb_for_each_bus_call(unsigned long data,
 			  int (*func)(struct ssb_bus *bus, unsigned long data));
+extern struct ssb_bus *ssb_pcmcia_dev_to_bus(struct pcmcia_device *pdev);
+
 
 /* b43_pci_bridge.c */
 #ifdef CONFIG_SSB_B43_PCI_BRIDGE
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index b7c388972fcf..8644e03cf588 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -245,9 +245,9 @@ struct ssb_bus {
 	/* Pointer to the PCMCIA device (only if bustype == SSB_BUSTYPE_PCMCIA). */
 	struct pcmcia_device *host_pcmcia;
 
-#ifdef CONFIG_SSB_PCIHOST
+#ifdef CONFIG_SSB_SPROM
 	/* Mutex to protect the SPROM writing. */
-	struct mutex pci_sprom_mutex;
+	struct mutex sprom_mutex;
 #endif
 
 	/* ID information about the Chip. */
-- 
cgit v1.2.3


From 9b89ca7a3847c0d5b1e86e83f4860a866f28a89b Mon Sep 17 00:00:00 2001
From: Marc Dionne <marc.c.dionne@gmail.com>
Date: Fri, 14 Mar 2008 13:11:29 +0000
Subject: struct export_operations: adjust comments to match current members

The comments in the definition of struct export_operations don't match the
current members.

Add a comment for the 2 new functions and remove 2 comments for unused ones.

Signed-off-by: Marc Dionne <marc.c.dionne@gmail.com>
Acked-by: David Howells <dhowells@redhat.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/exportfs.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/exportfs.h b/include/linux/exportfs.h
index 51d214138814..adcbb05b120b 100644
--- a/include/linux/exportfs.h
+++ b/include/linux/exportfs.h
@@ -49,11 +49,11 @@ struct fid {
 
 /**
  * struct export_operations - for nfsd to communicate with file systems
- * @decode_fh:      decode a file handle fragment and return a &struct dentry
  * @encode_fh:      encode a file handle fragment from a dentry
+ * @fh_to_dentry:   find the implied object and get a dentry for it
+ * @fh_to_parent:   find the implied object's parent and get a dentry for it
  * @get_name:       find the name for a given inode in a given directory
  * @get_parent:     find the parent of a given directory
- * @get_dentry:     find a dentry for the inode given a file handle sub-fragment
  *
  * See Documentation/filesystems/Exporting for details on how to use
  * this interface correctly.
-- 
cgit v1.2.3


From 2af3e6017e53065ddf40bb19190a29199b7ffee3 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Thu, 6 Mar 2008 16:02:42 +0100
Subject: The ps2esdi driver was marked as BROKEN more than two years ago due
 to being no longer working for some time.

A driver that had been marked as BROKEN for such a long time seems to be
unlikely to be revived in the forseeable future.

But if anyone wants to ever revive this driver, the code is still present in
the older kernel releases.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Acked-by: Alan Cox <alan@redhat.com>
Cc: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 Documentation/mca.txt   |   17 +-
 drivers/block/Kconfig   |   10 -
 drivers/block/Makefile  |    1 -
 drivers/block/ps2esdi.c | 1079 -----------------------------------------------
 include/linux/Kbuild    |    1 -
 include/linux/ps2esdi.h |   98 -----
 6 files changed, 5 insertions(+), 1201 deletions(-)
 delete mode 100644 drivers/block/ps2esdi.c
 delete mode 100644 include/linux/ps2esdi.h

(limited to 'include/linux')

diff --git a/Documentation/mca.txt b/Documentation/mca.txt
index aabce4ad90f9..510375d4209a 100644
--- a/Documentation/mca.txt
+++ b/Documentation/mca.txt
@@ -143,14 +143,7 @@ MCA Device Drivers
 
 Currently, there are a number of MCA-specific device drivers.
 
-1) PS/2 ESDI
-	drivers/block/ps2esdi.c
-	include/linux/ps2esdi.h
-   Uses major number 36, and should use /dev files /dev/eda, /dev/edb.
-   Supports two drives, but only one controller.  May use the
-   command-line args "ed=cyl,head,sec" and "tp720".
-
-2) PS/2 SCSI
+1) PS/2 SCSI
 	drivers/scsi/ibmmca.c
 	drivers/scsi/ibmmca.h
    The driver for the IBM SCSI subsystem.  Includes both integrated
@@ -159,25 +152,25 @@ Currently, there are a number of MCA-specific device drivers.
    machine with a front-panel display (i.e. model 95), you can use
    "ibmmcascsi=display" to enable a drive activity indicator.
 
-3) 3c523
+2) 3c523
 	drivers/net/3c523.c
 	drivers/net/3c523.h
    3Com 3c523 Etherlink/MC ethernet driver.
 
-4) SMC Ultra/MCA and IBM Adapter/A
+3) SMC Ultra/MCA and IBM Adapter/A
 	drivers/net/smc-mca.c
 	drivers/net/smc-mca.h
 	Driver for the MCA version of the SMC Ultra and various other
 	OEM'ed and work-alike cards (Elite, Adapter/A, etc).
 
-5) NE/2
+4) NE/2
 	driver/net/ne2.c
 	driver/net/ne2.h
 	The NE/2 is the MCA version of the NE2000.  This may not work
 	with clones that have a different adapter id than the original
 	NE/2.
 
-6) Future Domain MCS-600/700, OEM'd IBM Fast SCSI Adapter/A and
+5) Future Domain MCS-600/700, OEM'd IBM Fast SCSI Adapter/A and
    Reply Sound Blaster/SCSI (SCSI part)
 	Better support for these cards than the driver for ISA.
    Supports multiple cards with IRQ sharing.
diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
index b6d230b3209f..0d1d2133d9bc 100644
--- a/drivers/block/Kconfig
+++ b/drivers/block/Kconfig
@@ -44,16 +44,6 @@ config MAC_FLOPPY
 	  If you have a SWIM-3 (Super Woz Integrated Machine 3; from Apple)
 	  floppy controller, say Y here. Most commonly found in PowerMacs.
 
-config BLK_DEV_PS2
-	tristate "PS/2 ESDI hard disk support"
-	depends on MCA && MCA_LEGACY && BROKEN
-	help
-	  Say Y here if you have a PS/2 machine with a MCA bus and an ESDI
-	  hard disk.
-
-	  To compile this driver as a module, choose M here: the
-	  module will be called ps2esdi.
-
 config AMIGA_Z2RAM
 	tristate "Amiga Zorro II ramdisk support"
 	depends on ZORRO
diff --git a/drivers/block/Makefile b/drivers/block/Makefile
index 01c972415cb2..5e584306be99 100644
--- a/drivers/block/Makefile
+++ b/drivers/block/Makefile
@@ -13,7 +13,6 @@ obj-$(CONFIG_ATARI_FLOPPY)	+= ataflop.o
 obj-$(CONFIG_AMIGA_Z2RAM)	+= z2ram.o
 obj-$(CONFIG_BLK_DEV_RAM)	+= brd.o
 obj-$(CONFIG_BLK_DEV_LOOP)	+= loop.o
-obj-$(CONFIG_BLK_DEV_PS2)	+= ps2esdi.o
 obj-$(CONFIG_BLK_DEV_XD)	+= xd.o
 obj-$(CONFIG_BLK_CPQ_DA)	+= cpqarray.o
 obj-$(CONFIG_BLK_CPQ_CISS_DA)  += cciss.o
diff --git a/drivers/block/ps2esdi.c b/drivers/block/ps2esdi.c
deleted file mode 100644
index 3c796e236253..000000000000
--- a/drivers/block/ps2esdi.c
+++ /dev/null
@@ -1,1079 +0,0 @@
-/* ps2esdi driver based on assembler code by Arindam Banerji,
-   written by Peter De Schrijver */
-/* Reassuring note to IBM : This driver was NOT developed by vice-versa
-   engineering the PS/2's BIOS */
-/* Dedicated to Wannes, Tofke, Ykke, Godot, Killroy and all those 
-   other lovely fish out there... */
-/* This code was written during the long and boring WINA 
-   elections 1994 */
-/* Thanks to Arindam Banerij for giving me the source of his driver */
-/* This code may be freely distributed and modified in any way, 
-   as long as these notes remain intact */
-
-/*  Revised: 05/07/94 by Arindam Banerji (axb@cse.nd.edu) */
-/*  Revised: 09/08/94 by Peter De Schrijver (stud11@cc4.kuleuven.ac.be)
-   Thanks to Arindam Banerij for sending me the docs of the adapter */
-
-/* BA Modified for ThinkPad 720 by Boris Ashkinazi */
-/*                    (bash@vnet.ibm.com) 08/08/95 */
-
-/* Modified further for ThinkPad-720C by Uri Blumenthal */
-/*                    (uri@watson.ibm.com) Sep 11, 1995 */
-
-/* TODO : 
-   + Timeouts
-   + Get disk parameters
-   + DMA above 16MB
-   + reset after read/write error
- */
-
-#define DEVICE_NAME "PS/2 ESDI"
-
-#include <linux/major.h>
-#include <linux/errno.h>
-#include <linux/wait.h>
-#include <linux/interrupt.h>
-#include <linux/fs.h>
-#include <linux/kernel.h>
-#include <linux/genhd.h>
-#include <linux/ps2esdi.h>
-#include <linux/blkdev.h>
-#include <linux/mca-legacy.h>
-#include <linux/init.h>
-#include <linux/ioport.h>
-#include <linux/module.h>
-#include <linux/hdreg.h>
-
-#include <asm/system.h>
-#include <asm/io.h>
-#include <asm/dma.h>
-#include <asm/mca_dma.h>
-#include <asm/uaccess.h>
-
-#define PS2ESDI_IRQ 14
-#define MAX_HD 2
-#define MAX_RETRIES 5
-#define MAX_16BIT 65536
-#define ESDI_TIMEOUT   0xf000
-#define ESDI_STAT_TIMEOUT 4
-
-#define TYPE_0_CMD_BLK_LENGTH 2
-#define TYPE_1_CMD_BLK_LENGTH 4
-
-static void reset_ctrl(void);
-
-static int ps2esdi_geninit(void);
-
-static void do_ps2esdi_request(struct request_queue * q);
-
-static void ps2esdi_readwrite(int cmd, struct request *req);
-
-static void ps2esdi_fill_cmd_block(u_short * cmd_blk, u_short cmd,
-u_short cyl, u_short head, u_short sector, u_short length, u_char drive);
-
-static int ps2esdi_out_cmd_blk(u_short * cmd_blk);
-
-static void ps2esdi_prep_dma(char *buffer, u_short length, u_char dma_xmode);
-
-static irqreturn_t ps2esdi_interrupt_handler(int irq, void *dev_id);
-static void (*current_int_handler) (u_int) = NULL;
-static void ps2esdi_normal_interrupt_handler(u_int);
-static void ps2esdi_initial_reset_int_handler(u_int);
-static void ps2esdi_geometry_int_handler(u_int);
-static int ps2esdi_getgeo(struct block_device *bdev, struct hd_geometry *geo);
-
-static int ps2esdi_read_status_words(int num_words, int max_words, u_short * buffer);
-
-static void dump_cmd_complete_status(u_int int_ret_code);
-
-static void ps2esdi_get_device_cfg(void);
-
-static void ps2esdi_reset_timer(unsigned long unused);
-
-static u_int dma_arb_level;		/* DMA arbitration level */
-
-static DECLARE_WAIT_QUEUE_HEAD(ps2esdi_int);
-
-static int no_int_yet;
-static int ps2esdi_drives;
-static u_short io_base;
-static DEFINE_TIMER(esdi_timer, ps2esdi_reset_timer, 0, 0);
-static int reset_status;
-static int ps2esdi_slot = -1;
-static int tp720esdi = 0;	/* Is it Integrated ESDI of ThinkPad-720? */
-static int intg_esdi = 0;       /* If integrated adapter */
-struct ps2esdi_i_struct {
-	unsigned int head, sect, cyl, wpcom, lzone, ctl;
-};
-static DEFINE_SPINLOCK(ps2esdi_lock);
-static struct request_queue *ps2esdi_queue;
-static struct request *current_req;
-
-#if 0
-#if 0				/* try both - I don't know which one is better... UB */
-static struct ps2esdi_i_struct ps2esdi_info[MAX_HD] =
-{
-	{4, 48, 1553, 0, 0, 0},
-	{0, 0, 0, 0, 0, 0}};
-#else
-static struct ps2esdi_i_struct ps2esdi_info[MAX_HD] =
-{
-	{64, 32, 161, 0, 0, 0},
-	{0, 0, 0, 0, 0, 0}};
-#endif
-#endif
-static struct ps2esdi_i_struct ps2esdi_info[MAX_HD] =
-{
-	{0, 0, 0, 0, 0, 0},
-	{0, 0, 0, 0, 0, 0}};
-
-static struct block_device_operations ps2esdi_fops =
-{
-	.owner		= THIS_MODULE,
-	.getgeo		= ps2esdi_getgeo,
-};
-
-static struct gendisk *ps2esdi_gendisk[2];
-
-/* initialization routine called by ll_rw_blk.c   */
-static int __init ps2esdi_init(void)
-{
-
-	int error = 0;
-
-	/* register the device - pass the name and major number */
-	if (register_blkdev(PS2ESDI_MAJOR, "ed"))
-		return -EBUSY;
-
-	/* set up some global information - indicating device specific info */
-	ps2esdi_queue = blk_init_queue(do_ps2esdi_request, &ps2esdi_lock);
-	if (!ps2esdi_queue) {
-		unregister_blkdev(PS2ESDI_MAJOR, "ed");
-		return -ENOMEM;
-	}
-
-	/* some minor housekeeping - setup the global gendisk structure */
-	error = ps2esdi_geninit();
-	if (error) {
-		printk(KERN_WARNING "PS2ESDI: error initialising"
-			" device, releasing resources\n");
-		unregister_blkdev(PS2ESDI_MAJOR, "ed");
-		blk_cleanup_queue(ps2esdi_queue);
-		return error;
-	}
-	return 0;
-}				/* ps2esdi_init */
-
-#ifndef MODULE
-
-module_init(ps2esdi_init);
-
-#else
-
-static int cyl[MAX_HD] = {-1,-1};
-static int head[MAX_HD] = {-1, -1};
-static int sect[MAX_HD] = {-1, -1};
-
-module_param(tp720esdi, bool, 0);
-module_param_array(cyl, int, NULL, 0);
-module_param_array(head, int, NULL, 0);
-module_param_array(sect, int, NULL, 0);
-MODULE_LICENSE("GPL");
-
-int init_module(void) {
-	int drive;
-
-	for(drive = 0; drive < MAX_HD; drive++) {
-	        struct ps2esdi_i_struct *info = &ps2esdi_info[drive];
-
-        	if (cyl[drive] != -1) {
-		  	info->cyl = info->lzone = cyl[drive];
-			info->wpcom = 0;
-		}
-        	if (head[drive] != -1) {
-			info->head = head[drive];
-			info->ctl = (head[drive] > 8 ? 8 : 0);
-		}
-        	if (sect[drive] != -1) info->sect = sect[drive];
-	}
-	return ps2esdi_init();
-}
-
-void
-cleanup_module(void) {
-	int i;
-	if(ps2esdi_slot) {
-		mca_mark_as_unused(ps2esdi_slot);
-		mca_set_adapter_procfn(ps2esdi_slot, NULL, NULL);
-	}
-	release_region(io_base, 4);
-	free_dma(dma_arb_level);
-	free_irq(PS2ESDI_IRQ, &ps2esdi_gendisk);
-	unregister_blkdev(PS2ESDI_MAJOR, "ed");
-	blk_cleanup_queue(ps2esdi_queue);
-	for (i = 0; i < ps2esdi_drives; i++) {
-		del_gendisk(ps2esdi_gendisk[i]);
-		put_disk(ps2esdi_gendisk[i]);
-	}
-}
-#endif /* MODULE */
-
-/* handles boot time command line parameters */
-void __init tp720_setup(char *str, int *ints)
-{
-	/* no params, just sets the tp720esdi flag if it exists */
-
-	printk("%s: TP 720 ESDI flag set\n", DEVICE_NAME);
-	tp720esdi = 1;
-}
-
-void __init ed_setup(char *str, int *ints)
-{
-	int hdind = 0;
-
-	/* handles 3 parameters only - corresponding to
-	   1. Number of cylinders
-	   2. Number of heads
-	   3. Sectors/track
-	 */
-
-	if (ints[0] != 3)
-		return;
-
-	/* print out the information - seen at boot time */
-	printk("%s: ints[0]=%d ints[1]=%d ints[2]=%d ints[3]=%d\n",
-	       DEVICE_NAME, ints[0], ints[1], ints[2], ints[3]);
-
-	/* set the index into device specific information table */
-	if (ps2esdi_info[0].head != 0)
-		hdind = 1;
-
-	/* set up all the device information */
-	ps2esdi_info[hdind].head = ints[2];
-	ps2esdi_info[hdind].sect = ints[3];
-	ps2esdi_info[hdind].cyl = ints[1];
-	ps2esdi_info[hdind].wpcom = 0;
-	ps2esdi_info[hdind].lzone = ints[1];
-	ps2esdi_info[hdind].ctl = (ints[2] > 8 ? 8 : 0);
-#if 0				/* this may be needed for PS2/Mod.80, but it hurts ThinkPad! */
-	ps2esdi_drives = hdind + 1;	/* increment index for the next time */
-#endif
-}				/* ed_setup */
-
-static int ps2esdi_getinfo(char *buf, int slot, void *d)
-{
-	int len = 0;
-
-	len += sprintf(buf + len, "DMA Arbitration Level: %d\n",
-		       dma_arb_level);
-	len += sprintf(buf + len, "IO Port: %x\n", io_base);
-	len += sprintf(buf + len, "IRQ: 14\n");
-	len += sprintf(buf + len, "Drives: %d\n", ps2esdi_drives);
-
-	return len;
-}
-
-/* ps2 esdi specific initialization - called thru the gendisk chain */
-static int __init ps2esdi_geninit(void)
-{
-	/*
-	   The first part contains the initialization code
-	   for the ESDI disk subsystem.  All we really do
-	   is search for the POS registers of the controller
-	   to do some simple setup operations.  First, we
-	   must ensure that the controller is installed,
-	   enabled, and configured as PRIMARY.  Then we must
-	   determine the DMA arbitration level being used by
-	   the controller so we can handle data transfer
-	   operations properly.  If all of this works, then
-	   we will set the INIT_FLAG to a non-zero value.
-	 */
-
-	int slot = 0, i, reset_start, reset_end;
-	u_char status;
-	unsigned short adapterID;
-	int error = 0;
-
-	if ((slot = mca_find_adapter(INTG_ESDI_ID, 0)) != MCA_NOTFOUND) {
-		adapterID = INTG_ESDI_ID;
-		printk("%s: integrated ESDI adapter found in slot %d\n",
-		       DEVICE_NAME, slot+1);
-#ifndef MODULE
-		mca_set_adapter_name(slot, "PS/2 Integrated ESDI");
-#endif
-	} else if ((slot = mca_find_adapter(NRML_ESDI_ID, 0)) != -1) {
-		adapterID = NRML_ESDI_ID;
-		printk("%s: normal ESDI adapter found in slot %d\n",
-		       DEVICE_NAME, slot+1);
-		mca_set_adapter_name(slot, "PS/2 ESDI");
-	} else {
-		return -ENODEV;
-	}
-
-	ps2esdi_slot = slot;
-	mca_mark_as_used(slot);
-	mca_set_adapter_procfn(slot, (MCA_ProcFn) ps2esdi_getinfo, NULL);
-
-	/* Found the slot - read the POS register 2 to get the necessary
-	   configuration and status information.  POS register 2 has the
-	   following information :
-	   Bit           Function
-	   7             reserved = 0
-	   6             arbitration method
-	   0 - fairness enabled
-	   1 - fairness disabled, linear priority assignment
-	   5-2           arbitration level
-	   1             alternate address
-	   1              alternate address
-	   0 - use addresses 0x3510 - 0x3517
-	   0             adapter enable
-	 */
-
-	status = mca_read_stored_pos(slot, 2);
-	/* is it enabled ? */
-	if (!(status & STATUS_ENABLED)) {
-		printk("%s: ESDI adapter disabled\n", DEVICE_NAME);
-		error = -ENODEV;
-		goto err_out1;
-	}
-	/* try to grab IRQ, and try to grab a slow IRQ if it fails, so we can
-	   share with the SCSI driver */
-	if (request_irq(PS2ESDI_IRQ, ps2esdi_interrupt_handler,
-		  IRQF_DISABLED | IRQF_SHARED, "PS/2 ESDI", &ps2esdi_gendisk)
-	    && request_irq(PS2ESDI_IRQ, ps2esdi_interrupt_handler,
-			   IRQF_SHARED, "PS/2 ESDI", &ps2esdi_gendisk)
-	    ) {
-		printk("%s: Unable to get IRQ %d\n", DEVICE_NAME, PS2ESDI_IRQ);
-		error = -EBUSY;
-		goto err_out1;
-	}
-	if (status & STATUS_ALTERNATE)
-		io_base = ALT_IO_BASE;
-	else
-		io_base = PRIMARY_IO_BASE;
-
-	if (!request_region(io_base, 4, "ed")) {
-		printk(KERN_WARNING"Unable to request region 0x%x\n", io_base);
-		error = -EBUSY;
-		goto err_out2;
-	}
-	/* get the dma arbitration level */
-	dma_arb_level = (status >> 2) & 0xf;
-
-	/* BA */
-	printk("%s: DMA arbitration level : %d\n",
-	       DEVICE_NAME, dma_arb_level);
-
-	LITE_ON;
-	current_int_handler = ps2esdi_initial_reset_int_handler;
-	reset_ctrl();
-	reset_status = 0;
-	reset_start = jiffies;
-	while (!reset_status) {
-		init_timer(&esdi_timer);
-		esdi_timer.expires = jiffies + HZ;
-		esdi_timer.data = 0;
-		add_timer(&esdi_timer);
-		sleep_on(&ps2esdi_int);
-	}
-	reset_end = jiffies;
-	LITE_OFF;
-	printk("%s: reset interrupt after %d jiffies,  %u.%02u secs\n",
-	       DEVICE_NAME, reset_end - reset_start, (reset_end - reset_start) / HZ,
-	       (reset_end - reset_start) % HZ);
-
-
-	/* Integrated ESDI Disk and Controller has only one drive! */
-	if (adapterID == INTG_ESDI_ID) {/* if not "normal" PS2 ESDI adapter */
-		ps2esdi_drives = 1;	/* then we have only one physical disk! */		intg_esdi = 1;
-	}
-
-
-
-	/* finally this part sets up some global data structures etc. */
-
-	ps2esdi_get_device_cfg();
-
-	/* some annoyance in the above routine returns TWO drives?
-	 Is something else happining in the background?
-	 Regaurdless we fix the # of drives again. AJK */
-	/* Integrated ESDI Disk and Controller has only one drive! */
-	if (adapterID == INTG_ESDI_ID)	/* if not "normal" PS2 ESDI adapter */
-		ps2esdi_drives = 1;	/* Not three or two, ONE DAMNIT! */
-
-	current_int_handler = ps2esdi_normal_interrupt_handler;
-
-	if (request_dma(dma_arb_level, "ed") !=0) {
-		printk(KERN_WARNING "PS2ESDI: Can't request dma-channel %d\n"
-			,(int) dma_arb_level);
-		error = -EBUSY;
-		goto err_out3;
-	}
-	blk_queue_max_sectors(ps2esdi_queue, 128);
-
-	error = -ENOMEM;
-	for (i = 0; i < ps2esdi_drives; i++) {
-		struct gendisk *disk = alloc_disk(64);
-		if (!disk)
-			goto err_out4;
-		disk->major = PS2ESDI_MAJOR;
-		disk->first_minor = i<<6;
-		sprintf(disk->disk_name, "ed%c", 'a'+i);
-		disk->fops = &ps2esdi_fops;
-		ps2esdi_gendisk[i] = disk;
-	}
-
-	for (i = 0; i < ps2esdi_drives; i++) {
-		struct gendisk *disk = ps2esdi_gendisk[i];
-		set_capacity(disk, ps2esdi_info[i].head * ps2esdi_info[i].sect *
-				ps2esdi_info[i].cyl);
-		disk->queue = ps2esdi_queue;
-		disk->private_data = &ps2esdi_info[i];
-		add_disk(disk);
-	}
-	return 0;
-err_out4:
-	while (i--)
-		put_disk(ps2esdi_gendisk[i]);
-err_out3:
-	release_region(io_base, 4);
-err_out2:
-	free_irq(PS2ESDI_IRQ, &ps2esdi_gendisk);
-err_out1:
-	if(ps2esdi_slot) {
-		mca_mark_as_unused(ps2esdi_slot);
-		mca_set_adapter_procfn(ps2esdi_slot, NULL, NULL);
-	}
-	return error;
-}
-
-static void __init ps2esdi_get_device_cfg(void)
-{
-	u_short cmd_blk[TYPE_0_CMD_BLK_LENGTH];
-
-	/*BA */ printk("%s: Drive 0\n", DEVICE_NAME);
-	current_int_handler = ps2esdi_geometry_int_handler;
-	cmd_blk[0] = CMD_GET_DEV_CONFIG | 0x600;
-	cmd_blk[1] = 0;
-	no_int_yet = TRUE;
-	ps2esdi_out_cmd_blk(cmd_blk);
-	if (no_int_yet)
-		sleep_on(&ps2esdi_int);
-
-	if (ps2esdi_drives > 1) {
-		printk("%s: Drive 1\n", DEVICE_NAME);	/*BA */
-		cmd_blk[0] = CMD_GET_DEV_CONFIG | (1 << 5) | 0x600;
-		cmd_blk[1] = 0;
-		no_int_yet = TRUE;
-		ps2esdi_out_cmd_blk(cmd_blk);
-		if (no_int_yet)
-			sleep_on(&ps2esdi_int);
-	}			/* if second physical drive is present */
-	return;
-}
-
-/* strategy routine that handles most of the IO requests */
-static void do_ps2esdi_request(struct request_queue * q)
-{
-	struct request *req;
-	/* since, this routine is called with interrupts cleared - they 
-	   must be before it finishes  */
-
-	req = elv_next_request(q);
-	if (!req)
-		return;
-
-#if 0
-	printk("%s:got request. device : %s command : %d  sector : %ld count : %ld, buffer: %p\n",
-	       DEVICE_NAME,
-	       req->rq_disk->disk_name,
-	       req->cmd, req->sector,
-	       req->current_nr_sectors, req->buffer);
-#endif
-
-	/* check for above 16Mb dmas */
-	if (isa_virt_to_bus(req->buffer + req->current_nr_sectors * 512) > 16 * MB) {
-		printk("%s: DMA above 16MB not supported\n", DEVICE_NAME);
-		end_request(req, FAIL);
-		return;
-	}
-
-	if (req->sector+req->current_nr_sectors > get_capacity(req->rq_disk)) {
-		printk("Grrr. error. ps2esdi_drives: %d, %llu %llu\n",
-		    ps2esdi_drives, req->sector,
-		    (unsigned long long)get_capacity(req->rq_disk));
-		end_request(req, FAIL);
-		return;
-	}
-
-	switch (rq_data_dir(req)) {
-	case READ:
-		ps2esdi_readwrite(READ, req);
-		break;
-	case WRITE:
-		ps2esdi_readwrite(WRITE, req);
-		break;
-	default:
-		printk("%s: Unknown command\n", req->rq_disk->disk_name);
-		end_request(req, FAIL);
-		break;
-	}		/* handle different commands */
-}				/* main strategy routine */
-
-/* resets the ESDI adapter */
-static void reset_ctrl(void)
-{
-
-	u_long expire;
-	u_short status;
-
-	/* enable interrupts on the controller */
-	status = inb(ESDI_INTRPT);
-	outb((status & 0xe0) | ATT_EOI, ESDI_ATTN);	/* to be sure we don't have
-							   any interrupt pending... */
-	outb_p(CTRL_ENABLE_INTR, ESDI_CONTROL);
-
-	/* read the ESDI status port - if the controller is not busy,
-	   simply do a soft reset (fast) - otherwise we'll have to do a
-	   hard (slow) reset.  */
-	if (!(inb_p(ESDI_STATUS) & STATUS_BUSY)) {
-		/*BA */ printk("%s: soft reset...\n", DEVICE_NAME);
-		outb_p(CTRL_SOFT_RESET, ESDI_ATTN);
-	}
-	/* soft reset */ 
-	else {
-		/*BA */
-		printk("%s: hard reset...\n", DEVICE_NAME);
-		outb_p(CTRL_HARD_RESET, ESDI_CONTROL);
-		expire = jiffies + 2*HZ;
-		while (time_before(jiffies, expire));
-		outb_p(1, ESDI_CONTROL);
-	}			/* hard reset */
-
-
-}				/* reset the controller */
-
-/* called by the strategy routine to handle read and write requests */
-static void ps2esdi_readwrite(int cmd, struct request *req)
-{
-	struct ps2esdi_i_struct *p = req->rq_disk->private_data;
-	unsigned block = req->sector;
-	unsigned count = req->current_nr_sectors;
-	int drive = p - ps2esdi_info;
-	u_short track, head, cylinder, sector;
-	u_short cmd_blk[TYPE_1_CMD_BLK_LENGTH];
-
-	/* do some relevant arithmatic */
-	track = block / p->sect;
-	head = track % p->head;
-	cylinder = track / p->head;
-	sector = block % p->sect;
-
-#if 0
-	printk("%s: cyl=%d head=%d sect=%d\n", DEVICE_NAME, cylinder, head, sector);
-#endif
-	/* call the routine that actually fills out a command block */
-	ps2esdi_fill_cmd_block
-	    (cmd_blk,
-	     (cmd == READ) ? CMD_READ : CMD_WRITE,
-	     cylinder, head, sector, count, drive);
-
-	/* send the command block to the controller */
-	current_req = req;
-	spin_unlock_irq(&ps2esdi_lock);
-	if (ps2esdi_out_cmd_blk(cmd_blk)) {
-		spin_lock_irq(&ps2esdi_lock);
-		printk("%s: Controller failed\n", DEVICE_NAME);
-		if ((++req->errors) >= MAX_RETRIES)
-			end_request(req, FAIL);
-	}
-	/* check for failure to put out the command block */ 
-	else {
-		spin_lock_irq(&ps2esdi_lock);
-#if 0
-		printk("%s: waiting for xfer\n", DEVICE_NAME);
-#endif
-		/* turn disk lights on */
-		LITE_ON;
-	}
-
-}				/* ps2esdi_readwrite */
-
-/* fill out the command block */
-static void ps2esdi_fill_cmd_block(u_short * cmd_blk, u_short cmd,
- u_short cyl, u_short head, u_short sector, u_short length, u_char drive)
-{
-
-	cmd_blk[0] = (drive << 5) | cmd;
-	cmd_blk[1] = length;
-	cmd_blk[2] = ((cyl & 0x1f) << 11) | (head << 5) | sector;
-	cmd_blk[3] = (cyl & 0x3E0) >> 5;
-
-}				/* fill out the command block */
-
-/* write a command block to the controller */
-static int ps2esdi_out_cmd_blk(u_short * cmd_blk)
-{
-
-	int i;
-	unsigned long jif;
-	u_char status;
-
-	/* enable interrupts */
-	outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-
-	/* do not write to the controller, if it is busy */
-	for (jif = jiffies + ESDI_STAT_TIMEOUT;
-		time_after(jif, jiffies) &&
-			(inb(ESDI_STATUS) & STATUS_BUSY); )
-		;
-
-#if 0
-	printk("%s: i(1)=%ld\n", DEVICE_NAME, jif);
-#endif
-
-	/* if device is still busy - then just time out */
-	if (inb(ESDI_STATUS) & STATUS_BUSY) {
-		printk("%s: ps2esdi_out_cmd timed out (1)\n", DEVICE_NAME);
-		return ERROR;
-	}			/* timeout ??? */
-	/* Set up the attention register in the controller */
-	outb(((*cmd_blk) & 0xE0) | 1, ESDI_ATTN);
-
-#if 0
-	printk("%s: sending %d words to controller\n", DEVICE_NAME, (((*cmd_blk) >> 14) + 1) << 1);
-#endif
-
-	/* one by one send each word out */
-	for (i = (((*cmd_blk) >> 14) + 1) << 1; i; i--) {
-		status = inb(ESDI_STATUS);
-		for (jif = jiffies + ESDI_STAT_TIMEOUT;
-		     time_after(jif, jiffies) && (status & STATUS_BUSY) &&
-		   (status & STATUS_CMD_INF); status = inb(ESDI_STATUS));
-		if ((status & (STATUS_BUSY | STATUS_CMD_INF)) == STATUS_BUSY) {
-#if 0
-			printk("%s: sending %04X\n", DEVICE_NAME, *cmd_blk);
-#endif
-			outw(*cmd_blk++, ESDI_CMD_INT);
-		} else {
-			printk("%s: ps2esdi_out_cmd timed out while sending command (status=%02X)\n",
-			       DEVICE_NAME, status);
-			return ERROR;
-		}
-	}			/* send all words out */
-	return OK;
-}				/* send out the commands */
-
-
-/* prepare for dma - do all the necessary setup */
-static void ps2esdi_prep_dma(char *buffer, u_short length, u_char dma_xmode)
-{
-	unsigned long flags = claim_dma_lock();
-
-	mca_disable_dma(dma_arb_level);
-
-	mca_set_dma_addr(dma_arb_level, isa_virt_to_bus(buffer));
-
-	mca_set_dma_count(dma_arb_level, length * 512 / 2);
-
-	mca_set_dma_mode(dma_arb_level, dma_xmode);
-
-	mca_enable_dma(dma_arb_level);
-
-	release_dma_lock(flags);
-
-}				/* prepare for dma */
-
-
-
-static irqreturn_t ps2esdi_interrupt_handler(int irq, void *dev_id)
-{
-	u_int int_ret_code;
-
-	if (inb(ESDI_STATUS) & STATUS_INTR) {
-		int_ret_code = inb(ESDI_INTRPT);
-		if (current_int_handler) {
-			/* Disable adapter interrupts till processing is finished */
-			outb(CTRL_DISABLE_INTR, ESDI_CONTROL);
-			current_int_handler(int_ret_code);
-		} else
-			printk("%s: help ! No interrupt handler.\n", DEVICE_NAME);
-	} else {
-		return IRQ_NONE;
-	}
-	return IRQ_HANDLED;
-}
-
-static void ps2esdi_initial_reset_int_handler(u_int int_ret_code)
-{
-
-	switch (int_ret_code & 0xf) {
-	case INT_RESET:
-		/*BA */
-		printk("%s: initial reset completed.\n", DEVICE_NAME);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		wake_up(&ps2esdi_int);
-		break;
-	case INT_ATTN_ERROR:
-		printk("%s: Attention error. interrupt status : %02X\n", DEVICE_NAME,
-		       int_ret_code);
-		printk("%s: status: %02x\n", DEVICE_NAME, inb(ESDI_STATUS));
-		break;
-	default:
-		printk("%s: initial reset handler received interrupt: %02X\n",
-		       DEVICE_NAME, int_ret_code);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		break;
-	}
-	outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-}
-
-
-static void ps2esdi_geometry_int_handler(u_int int_ret_code)
-{
-	u_int status, drive_num;
-	unsigned long rba;
-	int i;
-
-	drive_num = int_ret_code >> 5;
-	switch (int_ret_code & 0xf) {
-	case INT_CMD_COMPLETE:
-		for (i = ESDI_TIMEOUT; i && !(inb(ESDI_STATUS) & STATUS_STAT_AVAIL); i--);
-		if (!(inb(ESDI_STATUS) & STATUS_STAT_AVAIL)) {
-			printk("%s: timeout reading status word\n", DEVICE_NAME);
-			outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-			break;
-		}
-		status = inw(ESDI_STT_INT);
-		if ((status & 0x1F) == CMD_GET_DEV_CONFIG) {
-#define REPLY_WORDS 5		/* we already read word 0 */
-			u_short reply[REPLY_WORDS];
-
-			if (ps2esdi_read_status_words((status >> 8) - 1, REPLY_WORDS, reply)) {
-				/*BA */
-				printk("%s: Device Configuration Status for drive %u\n",
-				       DEVICE_NAME, drive_num);
-
-				printk("%s: Spares/cyls: %u", DEVICE_NAME, reply[0] >> 8);
-
-				printk
-				    ("Config bits: %s%s%s%s%s\n",
-				     (reply[0] & CONFIG_IS) ? "Invalid Secondary, " : "",
-				     ((reply[0] & CONFIG_ZD) && !(reply[0] & CONFIG_IS))
-				 ? "Zero Defect, " : "Defects Present, ",
-				     (reply[0] & CONFIG_SF) ? "Skewed Format, " : "",
-				     (reply[0] & CONFIG_FR) ? "Removable, " : "Non-Removable, ",
-				     (reply[0] & CONFIG_RT) ? "No Retries" : "Retries");
-
-				rba = reply[1] | ((unsigned long) reply[2] << 16);
-				printk("%s: Number of RBA's: %lu\n", DEVICE_NAME, rba);
-
-				printk("%s: Physical number of cylinders: %u, Sectors/Track: %u, Heads: %u\n",
-				       DEVICE_NAME, reply[3], reply[4] >> 8, reply[4] & 0xff);
-
-				if (!ps2esdi_info[drive_num].head) {
-					ps2esdi_info[drive_num].head = 64;
-					ps2esdi_info[drive_num].sect = 32;
-					ps2esdi_info[drive_num].cyl = rba / (64 * 32);
-					ps2esdi_info[drive_num].wpcom = 0;
-					ps2esdi_info[drive_num].lzone = ps2esdi_info[drive_num].cyl;
-					ps2esdi_info[drive_num].ctl = 8;
-					if (tp720esdi) {	/* store the retrieved parameters */
-						ps2esdi_info[0].head = reply[4] & 0Xff;
-						ps2esdi_info[0].sect = reply[4] >> 8;
-						ps2esdi_info[0].cyl = reply[3];
-						ps2esdi_info[0].wpcom = 0;
-						ps2esdi_info[0].lzone = reply[3];
-					} else {
-						if (!intg_esdi)
-							ps2esdi_drives++;
-					}
-				}
-#ifdef OBSOLETE
-				if (!ps2esdi_info[drive_num].head) {
-					ps2esdi_info[drive_num].head = reply[4] & 0Xff;
-					ps2esdi_info[drive_num].sect = reply[4] >> 8;
-					ps2esdi_info[drive_num].cyl = reply[3];
-					ps2esdi_info[drive_num].wpcom = 0;
-					ps2esdi_info[drive_num].lzone = reply[3];
-					if (tp720esdi) {	/* store the retrieved parameters */
-						ps2esdi_info[0].head = reply[4] & 0Xff;
-						ps2esdi_info[0].sect = reply[4] >> 8;
-						ps2esdi_info[0].cyl = reply[3];
-						ps2esdi_info[0].wpcom = 0;
-						ps2esdi_info[0].lzone = reply[3];
-					} else {
-						ps2esdi_drives++;
-					}
-				}
-#endif
-
-			} else
-				printk("%s: failed while getting device config\n", DEVICE_NAME);
-#undef REPLY_WORDS
-		} else
-			printk("%s: command %02X unknown by geometry handler\n",
-			       DEVICE_NAME, status & 0x1f);
-
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		break;
-
-	case INT_ATTN_ERROR:
-		printk("%s: Attention error. interrupt status : %02X\n", DEVICE_NAME,
-		       int_ret_code);
-		printk("%s: Device not available\n", DEVICE_NAME);
-		break;
-	case INT_CMD_ECC:
-	case INT_CMD_RETRY:
-	case INT_CMD_ECC_RETRY:
-	case INT_CMD_WARNING:
-	case INT_CMD_ABORT:
-	case INT_CMD_FAILED:
-	case INT_DMA_ERR:
-	case INT_CMD_BLK_ERR:
-		/*BA */ printk("%s: Whaa. Error occurred...\n", DEVICE_NAME);
-		dump_cmd_complete_status(int_ret_code);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		break;
-	default:
-		printk("%s: Unknown interrupt reason: %02X\n",
-		       DEVICE_NAME, int_ret_code & 0xf);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		break;
-	}
-
-	wake_up(&ps2esdi_int);
-	no_int_yet = FALSE;
-	outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-
-}
-
-static void ps2esdi_normal_interrupt_handler(u_int int_ret_code)
-{
-	unsigned long flags;
-	u_int status;
-	u_int ending;
-	int i;
-
-	switch (int_ret_code & 0x0f) {
-	case INT_TRANSFER_REQ:
-		ps2esdi_prep_dma(current_req->buffer,
-				 current_req->current_nr_sectors,
-		    (rq_data_dir(current_req) == READ)
-		    ? MCA_DMA_MODE_16 | MCA_DMA_MODE_WRITE | MCA_DMA_MODE_XFER
-		    : MCA_DMA_MODE_16 | MCA_DMA_MODE_READ);
-		outb(CTRL_ENABLE_DMA | CTRL_ENABLE_INTR, ESDI_CONTROL);
-		ending = -1;
-		break;
-
-	case INT_ATTN_ERROR:
-		printk("%s: Attention error. interrupt status : %02X\n", DEVICE_NAME,
-		       int_ret_code);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		ending = FAIL;
-		break;
-
-	case INT_CMD_COMPLETE:
-		for (i = ESDI_TIMEOUT; i && !(inb(ESDI_STATUS) & STATUS_STAT_AVAIL); i--);
-		if (!(inb(ESDI_STATUS) & STATUS_STAT_AVAIL)) {
-			printk("%s: timeout reading status word\n", DEVICE_NAME);
-			outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-			outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-			if ((++current_req->errors) >= MAX_RETRIES)
-				ending = FAIL;
-			else
-				ending = -1;
-			break;
-		}
-		status = inw(ESDI_STT_INT);
-		switch (status & 0x1F) {
-		case (CMD_READ & 0xff):
-		case (CMD_WRITE & 0xff):
-			LITE_OFF;
-			outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-			outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-			ending = SUCCES;
-			break;
-		default:
-			printk("%s: interrupt for unknown command %02X\n",
-			       DEVICE_NAME, status & 0x1f);
-			outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-			outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-			ending = -1;
-			break;
-		}
-		break;
-	case INT_CMD_ECC:
-	case INT_CMD_RETRY:
-	case INT_CMD_ECC_RETRY:
-		LITE_OFF;
-		dump_cmd_complete_status(int_ret_code);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		ending = SUCCES;
-		break;
-	case INT_CMD_WARNING:
-	case INT_CMD_ABORT:
-	case INT_CMD_FAILED:
-	case INT_DMA_ERR:
-		LITE_OFF;
-		dump_cmd_complete_status(int_ret_code);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		if ((++current_req->errors) >= MAX_RETRIES)
-			ending = FAIL;
-		else
-			ending = -1;
-		break;
-
-	case INT_CMD_BLK_ERR:
-		dump_cmd_complete_status(int_ret_code);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		ending = FAIL;
-		break;
-
-	case INT_CMD_FORMAT:
-		printk("%s: huh ? Who issued this format command ?\n"
-		       ,DEVICE_NAME);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		ending = -1;
-		break;
-
-	case INT_RESET:
-		/* BA printk("%s: reset completed.\n", DEVICE_NAME) */ ;
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		ending = -1;
-		break;
-
-	default:
-		printk("%s: Unknown interrupt reason: %02X\n",
-		       DEVICE_NAME, int_ret_code & 0xf);
-		outb((int_ret_code & 0xe0) | ATT_EOI, ESDI_ATTN);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		ending = -1;
-		break;
-	}
-	if(ending != -1) {
-		spin_lock_irqsave(&ps2esdi_lock, flags);
-		end_request(current_req, ending);
-		current_req = NULL;
-		do_ps2esdi_request(ps2esdi_queue);
-		spin_unlock_irqrestore(&ps2esdi_lock, flags);
-	}
-}				/* handle interrupts */
-
-
-
-static int ps2esdi_read_status_words(int num_words,
-				     int max_words,
-				     u_short * buffer)
-{
-	int i;
-
-	for (; max_words && num_words; max_words--, num_words--, buffer++) {
-		for (i = ESDI_TIMEOUT; i && !(inb(ESDI_STATUS) & STATUS_STAT_AVAIL); i--);
-		if (!(inb(ESDI_STATUS) & STATUS_STAT_AVAIL)) {
-			printk("%s: timeout reading status word\n", DEVICE_NAME);
-			return FAIL;
-		}
-		*buffer = inw(ESDI_STT_INT);
-	}
-	return SUCCES;
-}
-
-
-
-
-static void dump_cmd_complete_status(u_int int_ret_code)
-{
-#define WAIT_FOR_STATUS \
-  for(i=ESDI_TIMEOUT;i && !(inb(ESDI_STATUS) & STATUS_STAT_AVAIL);i--); \
-    if(!(inb(ESDI_STATUS) & STATUS_STAT_AVAIL)) { \
-    printk("%s: timeout reading status word\n",DEVICE_NAME); \
-    return; \
-    }
-
-	int i, word_count;
-	u_short stat_word;
-	u_long rba;
-
-	printk("%s: Device: %u, interrupt ID: %02X\n",
-	       DEVICE_NAME, int_ret_code >> 5,
-	       int_ret_code & 0xf);
-
-	WAIT_FOR_STATUS;
-	stat_word = inw(ESDI_STT_INT);
-	word_count = (stat_word >> 8) - 1;
-	printk("%s: %u status words, command: %02X\n", DEVICE_NAME, word_count,
-	       stat_word & 0xff);
-
-	if (word_count--) {
-		WAIT_FOR_STATUS;
-		stat_word = inw(ESDI_STT_INT);
-		printk("%s: command status code: %02X, command error code: %02X\n",
-		       DEVICE_NAME, stat_word >> 8, stat_word & 0xff);
-	}
-	if (word_count--) {
-		WAIT_FOR_STATUS;
-		stat_word = inw(ESDI_STT_INT);
-		printk("%s: device error code: %s%s%s%s%s,%02X\n", DEVICE_NAME,
-		       (stat_word & 0x1000) ? "Ready, " : "Not Ready, ",
-		  (stat_word & 0x0800) ? "Selected, " : "Not Selected, ",
-		       (stat_word & 0x0400) ? "Write Fault, " : "",
-		       (stat_word & 0x0200) ? "Track 0, " : "",
-		(stat_word & 0x0100) ? "Seek or command complete, " : "",
-		       stat_word >> 8);
-	}
-	if (word_count--) {
-		WAIT_FOR_STATUS;
-		stat_word = inw(ESDI_STT_INT);
-		printk("%s: Blocks to do: %u", DEVICE_NAME, stat_word);
-	}
-	if (word_count -= 2) {
-		WAIT_FOR_STATUS;
-		rba = inw(ESDI_STT_INT);
-		WAIT_FOR_STATUS;
-		rba |= inw(ESDI_STT_INT) << 16;
-		printk(", Last Cyl: %u Head: %u Sector: %u\n",
-		       (u_short) ((rba & 0x1ff80000) >> 11),
-		 (u_short) ((rba & 0x7E0) >> 5), (u_short) (rba & 0x1f));
-	} else
-		printk("\n");
-
-	if (word_count--) {
-		WAIT_FOR_STATUS;
-		stat_word = inw(ESDI_STT_INT);
-		printk("%s: Blocks required ECC: %u", DEVICE_NAME, stat_word);
-	}
-	printk("\n");
-
-#undef WAIT_FOR_STATUS
-
-}
-
-static int ps2esdi_getgeo(struct block_device *bdev, struct hd_geometry *geo)
-{
-	struct ps2esdi_i_struct *p = bdev->bd_disk->private_data;
-
-	geo->heads = p->head;
-	geo->sectors = p->sect;
-	geo->cylinders = p->cyl;
-	return 0;
-}
-
-static void ps2esdi_reset_timer(unsigned long unused)
-{
-
-	int status;
-
-	status = inb(ESDI_INTRPT);
-	if ((status & 0xf) == INT_RESET) {
-		outb((status & 0xe0) | ATT_EOI, ESDI_ATTN);
-		outb(CTRL_ENABLE_INTR, ESDI_CONTROL);
-		reset_status = 1;
-	}
-	wake_up(&ps2esdi_int);
-}
diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 0fac822c1157..4108b38ebb16 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -127,7 +127,6 @@ header-y += pkt_sched.h
 header-y += posix_types.h
 header-y += ppdev.h
 header-y += prctl.h
-header-y += ps2esdi.h
 header-y += qnxtypes.h
 header-y += quotaio_v1.h
 header-y += quotaio_v2.h
diff --git a/include/linux/ps2esdi.h b/include/linux/ps2esdi.h
deleted file mode 100644
index c0e050b1dfe9..000000000000
--- a/include/linux/ps2esdi.h
+++ /dev/null
@@ -1,98 +0,0 @@
-#ifndef _PS2ESDI_H_
-#define _PS2ESDI_H_
-
-#define NRML_ESDI_ID 0xddff
-#define INTG_ESDI_ID 0xdf9f
-
-#define PRIMARY_IO_BASE 0x3510
-#define ALT_IO_BASE 0x3518
-
-#define ESDI_CMD_INT (io_base+0)
-#define ESDI_STT_INT (io_base+0)
-#define ESDI_CONTROL (io_base+2)
-#define ESDI_STATUS  (io_base+2)
-#define ESDI_ATTN    (io_base+3)
-#define ESDI_INTRPT  (io_base+3)
-
-#define STATUS_ENABLED    0x01
-#define STATUS_ALTERNATE  0x02
-#define STATUS_BUSY       0x10
-#define STATUS_STAT_AVAIL 0x08
-#define STATUS_INTR       0x01
-#define STATUS_RESET_FAIL 0xea
-#define STATUS_CMD_INF	  0x04
-
-#define CTRL_SOFT_RESET   0xe4
-#define CTRL_HARD_RESET   0x80
-#define CTRL_EOI          0xe2
-#define CTRL_ENABLE_DMA   0x02
-#define CTRL_ENABLE_INTR  0x01
-#define CTRL_DISABLE_INTR  0x00
-
-#define ATT_EOI 0x02
-
-/* bits of word 0 of configuration status block. more info see p.38 of tech ref */
-#define CONFIG_IS 0x10 /* Invalid Secondary */
-#define CONFIG_ZD 0x08 /* Zero Defect */
-#define CONFIG_SF 0x04 /* Skewed Format */
-#define CONFIG_FR 0x02 /* Removable */
-#define CONFIG_RT 0x01 /* Retries */
-
-#define PORT_SYS_A   0x92
-#define PORT_DMA_FN  0x18
-#define PORT_DMA_EX  0x1a
-
-#define ON (unsigned char)0x40
-#define OFF (unsigned char)~ON
-#define LITE_ON outb(inb(PORT_SYS_A) | ON,PORT_SYS_A)
-#define LITE_OFF outb((inb(PORT_SYS_A) & OFF),PORT_SYS_A)
-
-#define FAIL 0
-#define SUCCES 1
-
-#define INT_CMD_COMPLETE 0x01
-#define INT_CMD_ECC      0x03
-#define INT_CMD_RETRY    0x05
-#define INT_CMD_FORMAT   0x06
-#define INT_CMD_ECC_RETRY 0x07
-#define INT_CMD_WARNING  0x08
-#define INT_CMD_ABORT    0x09
-#define INT_RESET        0x0A
-#define INT_TRANSFER_REQ 0x0B
-#define INT_CMD_FAILED   0x0C
-#define INT_DMA_ERR      0x0D
-#define INT_CMD_BLK_ERR  0x0E
-#define INT_ATTN_ERROR   0x0F
-
-#define DMA_MASK_CHAN 0x90
-#define DMA_UNMASK_CHAN 0xA0
-#define DMA_WRITE_ADDR 0x20
-#define DMA_WRITE_TC 0x40
-#define DMA_WRITE_MODE 0x70
-
-#define CMD_GET_DEV_CONFIG 0x09
-#define CMD_READ 0x4601
-#define CMD_WRITE 0x4602
-#define DMA_READ_16 0x4C
-#define DMA_WRITE_16 0x44
-
-
-#define MB 1024*1024
-#define SECT_SIZE 512   
-
-#define ERROR 1
-#define OK 0
-
-#define HDIO_GETGEO 0x0301
-
-#define FALSE 0
-#define TRUE !FALSE
-
-struct ps2esdi_geometry {
-	unsigned char heads;
-	unsigned char sectors;
-	unsigned short cylinders;
-	unsigned long start;
-};
-
-#endif /* _PS2ESDI_H_ */
-- 
cgit v1.2.3


From 3e94794355724f77dc6cbb5ad956f7c72d8313a4 Mon Sep 17 00:00:00 2001
From: Magnus Damm <magnus.damm@gmail.com>
Date: Fri, 22 Feb 2008 19:55:15 +0900
Subject: smc91x: introduce platform data flags V2

This patch introduces struct smc91x_platdata and modifies the driver so
bus width is checked during run time using SMC_nBIT() instead of
SMC_CAN_USE_nBIT.

V2 keeps static configuration lean using SMC_DYNAMIC_BUS_CONFIG.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Acked-by: Nicolas Pitre <nico@cam.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/net/smc91x.c   | 34 +++++++++++++++++++++++++-----
 drivers/net/smc91x.h   | 57 +++++++++++++++++++++++++++++++-------------------
 include/linux/smc91x.h | 13 ++++++++++++
 3 files changed, 77 insertions(+), 27 deletions(-)
 create mode 100644 include/linux/smc91x.h

(limited to 'include/linux')

diff --git a/drivers/net/smc91x.c b/drivers/net/smc91x.c
index d0ef80ae018a..97bdb2a43bc8 100644
--- a/drivers/net/smc91x.c
+++ b/drivers/net/smc91x.c
@@ -1997,6 +1997,8 @@ err_out:
 
 static int smc_enable_device(struct platform_device *pdev)
 {
+	struct net_device *ndev = platform_get_drvdata(pdev);
+	struct smc_local *lp = netdev_priv(ndev);
 	unsigned long flags;
 	unsigned char ecor, ecsr;
 	void __iomem *addr;
@@ -2039,7 +2041,7 @@ static int smc_enable_device(struct platform_device *pdev)
 	 * Set the appropriate byte/word mode.
 	 */
 	ecsr = readb(addr + (ECSR << SMC_IO_SHIFT)) & ~ECSR_IOIS8;
-	if (!SMC_CAN_USE_16BIT)
+	if (!SMC_16BIT(lp))
 		ecsr |= ECSR_IOIS8;
 	writeb(ecsr, addr + (ECSR << SMC_IO_SHIFT));
 	local_irq_restore(flags);
@@ -2124,10 +2126,11 @@ static void smc_release_datacs(struct platform_device *pdev, struct net_device *
  */
 static int smc_drv_probe(struct platform_device *pdev)
 {
+	struct smc91x_platdata *pd = pdev->dev.platform_data;
+	struct smc_local *lp;
 	struct net_device *ndev;
 	struct resource *res, *ires;
 	unsigned int __iomem *addr;
-	unsigned long irq_flags = SMC_IRQ_FLAGS;
 	int ret;
 
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "smc91x-regs");
@@ -2152,6 +2155,27 @@ static int smc_drv_probe(struct platform_device *pdev)
 	}
 	SET_NETDEV_DEV(ndev, &pdev->dev);
 
+	/* get configuration from platform data, only allow use of
+	 * bus width if both SMC_CAN_USE_xxx and SMC91X_USE_xxx are set.
+	 */
+
+	lp = netdev_priv(ndev);
+	lp->cfg.irq_flags = SMC_IRQ_FLAGS;
+
+#ifdef SMC_DYNAMIC_BUS_CONFIG
+	if (pd)
+		memcpy(&lp->cfg, pd, sizeof(lp->cfg));
+	else {
+		lp->cfg.flags = SMC91X_USE_8BIT;
+		lp->cfg.flags |= SMC91X_USE_16BIT;
+		lp->cfg.flags |= SMC91X_USE_32BIT;
+	}
+
+	lp->cfg.flags &= ~(SMC_CAN_USE_8BIT ? 0 : SMC91X_USE_8BIT);
+	lp->cfg.flags &= ~(SMC_CAN_USE_16BIT ? 0 : SMC91X_USE_16BIT);
+	lp->cfg.flags &= ~(SMC_CAN_USE_32BIT ? 0 : SMC91X_USE_32BIT);
+#endif
+
 	ndev->dma = (unsigned char)-1;
 
 	ires = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
@@ -2162,7 +2186,7 @@ static int smc_drv_probe(struct platform_device *pdev)
 
 	ndev->irq = ires->start;
 	if (SMC_IRQ_FLAGS == -1)
-		irq_flags = ires->flags & IRQF_TRIGGER_MASK;
+		lp->cfg.irq_flags = ires->flags & IRQF_TRIGGER_MASK;
 
 	ret = smc_request_attrib(pdev);
 	if (ret)
@@ -2170,6 +2194,7 @@ static int smc_drv_probe(struct platform_device *pdev)
 #if defined(CONFIG_SA1100_ASSABET)
 	NCR_0 |= NCR_ENET_OSC_EN;
 #endif
+	platform_set_drvdata(pdev, ndev);
 	ret = smc_enable_device(pdev);
 	if (ret)
 		goto out_release_attrib;
@@ -2188,8 +2213,7 @@ static int smc_drv_probe(struct platform_device *pdev)
 	}
 #endif
 
-	platform_set_drvdata(pdev, ndev);
-	ret = smc_probe(ndev, addr, irq_flags);
+	ret = smc_probe(ndev, addr, lp->cfg.irq_flags);
 	if (ret != 0)
 		goto out_iounmap;
 
diff --git a/drivers/net/smc91x.h b/drivers/net/smc91x.h
index 92ff9c42367e..e044b4de1397 100644
--- a/drivers/net/smc91x.h
+++ b/drivers/net/smc91x.h
@@ -34,6 +34,7 @@
 #ifndef _SMC91X_H_
 #define _SMC91X_H_
 
+#include <linux/smc91x.h>
 
 /*
  * Define your architecture specific bus configuration parameters here.
@@ -481,6 +482,7 @@ static inline void LPD7_SMC_outsw (unsigned char* a, int r,
 #define RPC_LSA_DEFAULT		RPC_LED_100_10
 #define RPC_LSB_DEFAULT		RPC_LED_TX_RX
 
+#define SMC_DYNAMIC_BUS_CONFIG
 #endif
 
 
@@ -526,8 +528,19 @@ struct smc_local {
 #endif
 	void __iomem *base;
 	void __iomem *datacs;
+
+	struct smc91x_platdata cfg;
 };
 
+#ifdef SMC_DYNAMIC_BUS_CONFIG
+#define SMC_8BIT(p) (((p)->cfg.flags & SMC91X_USE_8BIT) && SMC_CAN_USE_8BIT)
+#define SMC_16BIT(p) (((p)->cfg.flags & SMC91X_USE_16BIT) && SMC_CAN_USE_16BIT)
+#define SMC_32BIT(p) (((p)->cfg.flags & SMC91X_USE_32BIT) && SMC_CAN_USE_32BIT)
+#else
+#define SMC_8BIT(p) SMC_CAN_USE_8BIT
+#define SMC_16BIT(p) SMC_CAN_USE_16BIT
+#define SMC_32BIT(p) SMC_CAN_USE_32BIT
+#endif
 
 #ifdef SMC_USE_PXA_DMA
 /*
@@ -1108,41 +1121,41 @@ static const char * chip_ids[ 16 ] =  {
  *
  * Enforce it on any 32-bit capable setup for now.
  */
-#define SMC_MUST_ALIGN_WRITE	SMC_CAN_USE_32BIT
+#define SMC_MUST_ALIGN_WRITE(lp)	SMC_32BIT(lp)
 
 #define SMC_GET_PN(lp)						\
-	(SMC_CAN_USE_8BIT	? (SMC_inb(ioaddr, PN_REG(lp)))	\
+	(SMC_8BIT(lp)	? (SMC_inb(ioaddr, PN_REG(lp)))	\
 				: (SMC_inw(ioaddr, PN_REG(lp)) & 0xFF))
 
 #define SMC_SET_PN(lp, x)						\
 	do {								\
-		if (SMC_MUST_ALIGN_WRITE)				\
+		if (SMC_MUST_ALIGN_WRITE(lp))				\
 			SMC_outl((x)<<16, ioaddr, SMC_REG(lp, 0, 2));	\
-		else if (SMC_CAN_USE_8BIT)				\
+		else if (SMC_8BIT(lp))				\
 			SMC_outb(x, ioaddr, PN_REG(lp));		\
 		else							\
 			SMC_outw(x, ioaddr, PN_REG(lp));		\
 	} while (0)
 
 #define SMC_GET_AR(lp)						\
-	(SMC_CAN_USE_8BIT	? (SMC_inb(ioaddr, AR_REG(lp)))	\
+	(SMC_8BIT(lp)	? (SMC_inb(ioaddr, AR_REG(lp)))	\
 				: (SMC_inw(ioaddr, PN_REG(lp)) >> 8))
 
 #define SMC_GET_TXFIFO(lp)						\
-	(SMC_CAN_USE_8BIT	? (SMC_inb(ioaddr, TXFIFO_REG(lp)))	\
+	(SMC_8BIT(lp)	? (SMC_inb(ioaddr, TXFIFO_REG(lp)))	\
 				: (SMC_inw(ioaddr, TXFIFO_REG(lp)) & 0xFF))
 
 #define SMC_GET_RXFIFO(lp)						\
-	(SMC_CAN_USE_8BIT	? (SMC_inb(ioaddr, RXFIFO_REG(lp)))	\
+	(SMC_8BIT(lp)	? (SMC_inb(ioaddr, RXFIFO_REG(lp)))	\
 				: (SMC_inw(ioaddr, TXFIFO_REG(lp)) >> 8))
 
 #define SMC_GET_INT(lp)						\
-	(SMC_CAN_USE_8BIT	? (SMC_inb(ioaddr, INT_REG(lp)))	\
+	(SMC_8BIT(lp)	? (SMC_inb(ioaddr, INT_REG(lp)))	\
 				: (SMC_inw(ioaddr, INT_REG(lp)) & 0xFF))
 
 #define SMC_ACK_INT(lp, x)						\
 	do {								\
-		if (SMC_CAN_USE_8BIT)					\
+		if (SMC_8BIT(lp))					\
 			SMC_outb(x, ioaddr, INT_REG(lp));		\
 		else {							\
 			unsigned long __flags;				\
@@ -1155,12 +1168,12 @@ static const char * chip_ids[ 16 ] =  {
 	} while (0)
 
 #define SMC_GET_INT_MASK(lp)						\
-	(SMC_CAN_USE_8BIT	? (SMC_inb(ioaddr, IM_REG(lp)))	\
+	(SMC_8BIT(lp)	? (SMC_inb(ioaddr, IM_REG(lp)))	\
 				: (SMC_inw(ioaddr, INT_REG(lp)) >> 8))
 
 #define SMC_SET_INT_MASK(lp, x)					\
 	do {								\
-		if (SMC_CAN_USE_8BIT)					\
+		if (SMC_8BIT(lp))					\
 			SMC_outb(x, ioaddr, IM_REG(lp));		\
 		else							\
 			SMC_outw((x) << 8, ioaddr, INT_REG(lp));	\
@@ -1170,7 +1183,7 @@ static const char * chip_ids[ 16 ] =  {
 
 #define SMC_SELECT_BANK(lp, x)					\
 	do {								\
-		if (SMC_MUST_ALIGN_WRITE)				\
+		if (SMC_MUST_ALIGN_WRITE(lp))				\
 			SMC_outl((x)<<16, ioaddr, 12<<SMC_IO_SHIFT);	\
 		else							\
 			SMC_outw(x, ioaddr, BANK_SELECT);		\
@@ -1208,7 +1221,7 @@ static const char * chip_ids[ 16 ] =  {
 
 #define SMC_SET_PTR(lp, x)						\
 	do {								\
-		if (SMC_MUST_ALIGN_WRITE)				\
+		if (SMC_MUST_ALIGN_WRITE(lp))				\
 			SMC_outl((x)<<16, ioaddr, SMC_REG(lp, 4, 2));	\
 		else							\
 			SMC_outw(x, ioaddr, PTR_REG(lp));		\
@@ -1226,7 +1239,7 @@ static const char * chip_ids[ 16 ] =  {
 
 #define SMC_SET_RPC(lp, x)						\
 	do {								\
-		if (SMC_MUST_ALIGN_WRITE)				\
+		if (SMC_MUST_ALIGN_WRITE(lp))				\
 			SMC_outl((x)<<16, ioaddr, SMC_REG(lp, 8, 0));	\
 		else							\
 			SMC_outw(x, ioaddr, RPC_REG(lp));		\
@@ -1267,7 +1280,7 @@ static const char * chip_ids[ 16 ] =  {
 
 #define SMC_PUT_PKT_HDR(lp, status, length)				\
 	do {								\
-		if (SMC_CAN_USE_32BIT)					\
+		if (SMC_32BIT(lp))					\
 			SMC_outl((status) | (length)<<16, ioaddr,	\
 				 DATA_REG(lp));			\
 		else {							\
@@ -1278,7 +1291,7 @@ static const char * chip_ids[ 16 ] =  {
 
 #define SMC_GET_PKT_HDR(lp, status, length)				\
 	do {								\
-		if (SMC_CAN_USE_32BIT) {				\
+		if (SMC_32BIT(lp)) {				\
 			unsigned int __val = SMC_inl(ioaddr, DATA_REG(lp)); \
 			(status) = __val & 0xffff;			\
 			(length) = __val >> 16;				\
@@ -1290,7 +1303,7 @@ static const char * chip_ids[ 16 ] =  {
 
 #define SMC_PUSH_DATA(lp, p, l)					\
 	do {								\
-		if (SMC_CAN_USE_32BIT) {				\
+		if (SMC_32BIT(lp)) {				\
 			void *__ptr = (p);				\
 			int __len = (l);				\
 			void __iomem *__ioaddr = ioaddr;		\
@@ -1308,15 +1321,15 @@ static const char * chip_ids[ 16 ] =  {
 				SMC_outw(*((u16 *)__ptr), ioaddr,	\
 					 DATA_REG(lp));		\
 			}						\
-		} else if (SMC_CAN_USE_16BIT)				\
+		} else if (SMC_16BIT(lp))				\
 			SMC_outsw(ioaddr, DATA_REG(lp), p, (l) >> 1);	\
-		else if (SMC_CAN_USE_8BIT)				\
+		else if (SMC_8BIT(lp))				\
 			SMC_outsb(ioaddr, DATA_REG(lp), p, l);	\
 	} while (0)
 
 #define SMC_PULL_DATA(lp, p, l)					\
 	do {								\
-		if (SMC_CAN_USE_32BIT) {				\
+		if (SMC_32BIT(lp)) {				\
 			void *__ptr = (p);				\
 			int __len = (l);				\
 			void __iomem *__ioaddr = ioaddr;		\
@@ -1343,9 +1356,9 @@ static const char * chip_ids[ 16 ] =  {
 				__ioaddr = lp->datacs;			\
 			__len += 2;					\
 			SMC_insl(__ioaddr, DATA_REG(lp), __ptr, __len>>2); \
-		} else if (SMC_CAN_USE_16BIT)				\
+		} else if (SMC_16BIT(lp))				\
 			SMC_insw(ioaddr, DATA_REG(lp), p, (l) >> 1);	\
-		else if (SMC_CAN_USE_8BIT)				\
+		else if (SMC_8BIT(lp))				\
 			SMC_insb(ioaddr, DATA_REG(lp), p, l);		\
 	} while (0)
 
diff --git a/include/linux/smc91x.h b/include/linux/smc91x.h
new file mode 100644
index 000000000000..8e0556b8781c
--- /dev/null
+++ b/include/linux/smc91x.h
@@ -0,0 +1,13 @@
+#ifndef __SMC91X_H__
+#define __SMC91X_H__
+
+#define SMC91X_USE_8BIT (1 << 0)
+#define SMC91X_USE_16BIT (1 << 1)
+#define SMC91X_USE_32BIT (1 << 2)
+
+struct smc91x_platdata {
+	unsigned long flags;
+	unsigned long irq_flags; /* IRQF_... */
+};
+
+#endif /* __SMC91X_H__ */
-- 
cgit v1.2.3


From 4265f161b6bb7b31163671329b1142b9023bf4e3 Mon Sep 17 00:00:00 2001
From: Christian Borntraeger <borntraeger@de.ibm.com>
Date: Fri, 14 Mar 2008 14:17:05 +0100
Subject: virtio: fix race in enable_cb

There is a race in virtio_net, dealing with disabling/enabling the callback.
I saw the following oops:

kernel BUG at /space/kvm/drivers/virtio/virtio_ring.c:218!
illegal operation: 0001 [#1] SMP
Modules linked in: sunrpc dm_mod
CPU: 2 Not tainted 2.6.25-rc1zlive-host-10623-gd358142-dirty #99
Process swapper (pid: 0, task: 000000000f85a610, ksp: 000000000f873c60)
Krnl PSW : 0404300180000000 00000000002b81a6 (vring_disable_cb+0x16/0x20)
           R:0 T:1 IO:0 EX:0 Key:0 M:1 W:0 P:0 AS:0 CC:3 PM:0 EA:3
Krnl GPRS: 0000000000000001 0000000000000001 0000000010005800 0000000000000001
           000000000f3a0900 000000000f85a610 0000000000000000 0000000000000000
           0000000000000000 000000000f870000 0000000000000000 0000000000001237
           000000000f3a0920 000000000010ff74 00000000002846f6 000000000fa0bcd8
Krnl Code: 00000000002b819a: a7110001           tmll    %r1,1
           00000000002b819e: a7840004           brc     8,2b81a6
           00000000002b81a2: a7f40001           brc     15,2b81a4
          >00000000002b81a6: a51b0001           oill    %r1,1
           00000000002b81aa: 40102000           sth     %r1,0(%r2)
           00000000002b81ae: 07fe               bcr     15,%r14
           00000000002b81b0: eb7ff0380024       stmg    %r7,%r15,56(%r15)
           00000000002b81b6: a7f13e00           tmll    %r15,15872
Call Trace:
([<000000000fa0bcd0>] 0xfa0bcd0)
 [<00000000002b8350>] vring_interrupt+0x5c/0x6c
 [<000000000010ab08>] do_extint+0xb8/0xf0
 [<0000000000110716>] ext_no_vtime+0x16/0x1a
 [<0000000000107e72>] cpu_idle+0x1c2/0x1e0

The problem can be triggered with a high amount of host->guest traffic.
I think its the following race:

poll says netif_rx_complete
poll calls enable_cb
enable_cb opens the interrupt mask
a new packet comes, an interrupt is triggered----\
enable_cb sees that there is more work           |
enable_cb disables the interrupt                 |
       .                                         V
       .                            interrupt is delivered
       .                            skb_recv_done does atomic napi test, ok
 some waiting                       disable_cb is called->check fails->bang!
       .
poll would do napi check
poll would do disable_cb

The fix is to let enable_cb not disable the interrupt again, but expect the
caller to do the cleanup if it returns false. In that case, the interrupt is
only disabled, if the napi test_set_bit was successful.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (cleaned up doco)
---
 drivers/net/virtio_net.c     | 10 +++++++---
 drivers/virtio/virtio_ring.c |  1 -
 include/linux/virtio.h       |  5 +++--
 3 files changed, 10 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index e575df83e5c2..b58472cf76f8 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -203,8 +203,11 @@ again:
 	if (received < budget) {
 		netif_rx_complete(vi->dev, napi);
 		if (unlikely(!vi->rvq->vq_ops->enable_cb(vi->rvq))
-		    && netif_rx_reschedule(vi->dev, napi))
+		    && napi_schedule_prep(napi)) {
+			vi->rvq->vq_ops->disable_cb(vi->rvq);
+			__netif_rx_schedule(vi->dev, napi);
 			goto again;
+		}
 	}
 
 	return received;
@@ -278,10 +281,11 @@ again:
 		pr_debug("%s: virtio not prepared to send\n", dev->name);
 		netif_stop_queue(dev);
 
-		/* Activate callback for using skbs: if this fails it
+		/* Activate callback for using skbs: if this returns false it
 		 * means some were used in the meantime. */
 		if (unlikely(!vi->svq->vq_ops->enable_cb(vi->svq))) {
-			printk("Unlikely: restart svq failed\n");
+			printk("Unlikely: restart svq race\n");
+			vi->svq->vq_ops->disable_cb(vi->svq);
 			netif_start_queue(dev);
 			goto again;
 		}
diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
index 3a28c1382131..aa714028641e 100644
--- a/drivers/virtio/virtio_ring.c
+++ b/drivers/virtio/virtio_ring.c
@@ -232,7 +232,6 @@ static bool vring_enable_cb(struct virtqueue *_vq)
 	vq->vring.avail->flags &= ~VRING_AVAIL_F_NO_INTERRUPT;
 	mb();
 	if (unlikely(more_used(vq))) {
-		vq->vring.avail->flags |= VRING_AVAIL_F_NO_INTERRUPT;
 		END_USE(vq);
 		return false;
 	}
diff --git a/include/linux/virtio.h b/include/linux/virtio.h
index 260d1fcf29a4..12c18ac1b973 100644
--- a/include/linux/virtio.h
+++ b/include/linux/virtio.h
@@ -43,8 +43,9 @@ struct virtqueue
  *	vq: the struct virtqueue we're talking about.
  * @enable_cb: restart callbacks after disable_cb.
  *	vq: the struct virtqueue we're talking about.
- *	This returns "false" (and doesn't re-enable) if there are pending
- *	buffers in the queue, to avoid a race.
+ *	This re-enables callbacks; it returns "false" if there are pending
+ *	buffers in the queue, to detect a possible race between the driver
+ *	checking for more work, and enabling callbacks.
  *
  * Locking rules are straightforward: the driver is responsible for
  * locking.  No two operations may be invoked simultaneously.
-- 
cgit v1.2.3


From 916fbfb7ae5f8c8f86399794d89e6d273df8826b Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Wed, 12 Mar 2008 15:26:34 +0900
Subject: devres: implement pcim_iomap_regions_request_all()

Some drivers need to reserve all PCI BARs to prevent other drivers
misusing unoccupied BARs.  pcim_iomap_regions_request_all() requests
all BARs and iomap specified BARs.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@suse.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 include/linux/pci.h |  2 ++
 lib/devres.c        | 25 +++++++++++++++++++++++++
 2 files changed, 27 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/pci.h b/include/linux/pci.h
index 9010f5458767..b7e4b633c69b 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -1045,6 +1045,8 @@ void __iomem *pcim_iomap(struct pci_dev *pdev, int bar, unsigned long maxlen);
 void pcim_iounmap(struct pci_dev *pdev, void __iomem *addr);
 void __iomem * const *pcim_iomap_table(struct pci_dev *pdev);
 int pcim_iomap_regions(struct pci_dev *pdev, u16 mask, const char *name);
+int pcim_iomap_regions_request_all(struct pci_dev *pdev, u16 mask,
+				   const char *name);
 void pcim_iounmap_regions(struct pci_dev *pdev, u16 mask);
 
 extern int pci_pci_problems;
diff --git a/lib/devres.c b/lib/devres.c
index b1d336ce7f3d..edc27a5d1b73 100644
--- a/lib/devres.c
+++ b/lib/devres.c
@@ -297,6 +297,31 @@ int pcim_iomap_regions(struct pci_dev *pdev, u16 mask, const char *name)
 }
 EXPORT_SYMBOL(pcim_iomap_regions);
 
+/**
+ * pcim_iomap_regions_request_all - Request all BARs and iomap specified ones
+ * @pdev: PCI device to map IO resources for
+ * @mask: Mask of BARs to iomap
+ * @name: Name used when requesting regions
+ *
+ * Request all PCI BARs and iomap regions specified by @mask.
+ */
+int pcim_iomap_regions_request_all(struct pci_dev *pdev, u16 mask,
+				   const char *name)
+{
+	int request_mask = ((1 << 6) - 1) & ~mask;
+	int rc;
+
+	rc = pci_request_selected_regions(pdev, request_mask, name);
+	if (rc)
+		return rc;
+
+	rc = pcim_iomap_regions(pdev, mask, name);
+	if (rc)
+		pci_release_selected_regions(pdev, request_mask);
+	return rc;
+}
+EXPORT_SYMBOL(pcim_iomap_regions_request_all);
+
 /**
  * pcim_iounmap_regions - Unmap and release PCI BARs
  * @pdev: PCI device to map IO resources for
-- 
cgit v1.2.3


From 0382b9c35469be273ed10fa374496a924055a3c8 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 17 Mar 2008 22:46:46 -0700
Subject: [PKT_SCHED]: annotate cls_u32

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/pkt_cls.h | 8 ++++----
 net/sched/cls_u32.c     | 8 ++++----
 net/sched/em_u32.c      | 2 +-
 3 files changed, 9 insertions(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/pkt_cls.h b/include/linux/pkt_cls.h
index 28dfc61cf79e..99efbed81fa2 100644
--- a/include/linux/pkt_cls.h
+++ b/include/linux/pkt_cls.h
@@ -201,8 +201,8 @@ enum
 
 struct tc_u32_key
 {
-	__u32		mask;
-	__u32		val;
+	__be32		mask;
+	__be32		val;
 	int		off;
 	int		offmask;
 };
@@ -213,12 +213,12 @@ struct tc_u32_sel
 	unsigned char		offshift;
 	unsigned char		nkeys;
 
-	__u16			offmask;
+	__be16			offmask;
 	__u16			off;
 	short			offoff;
 
 	short			hoff;
-	__u32			hmask;
+	__be32			hmask;
 	struct tc_u32_key	keys[0];
 };
 
diff --git a/net/sched/cls_u32.c b/net/sched/cls_u32.c
index b18fa95ef248..c5c16b4b6e98 100644
--- a/net/sched/cls_u32.c
+++ b/net/sched/cls_u32.c
@@ -89,7 +89,7 @@ static const struct tcf_ext_map u32_ext_map = {
 
 static struct tc_u_common *u32_list;
 
-static __inline__ unsigned u32_hash_fold(u32 key, struct tc_u32_sel *sel, u8 fshift)
+static __inline__ unsigned u32_hash_fold(__be32 key, struct tc_u32_sel *sel, u8 fshift)
 {
 	unsigned h = ntohl(key & sel->hmask)>>fshift;
 
@@ -137,7 +137,7 @@ next_knode:
 
 		for (i = n->sel.nkeys; i>0; i--, key++) {
 
-			if ((*(u32*)(ptr+key->off+(off2&key->offmask))^key->val)&key->mask) {
+			if ((*(__be32*)(ptr+key->off+(off2&key->offmask))^key->val)&key->mask) {
 				n = n->next;
 				goto next_knode;
 			}
@@ -182,7 +182,7 @@ check_terminal:
 		ht = n->ht_down;
 		sel = 0;
 		if (ht->divisor)
-			sel = ht->divisor&u32_hash_fold(*(u32*)(ptr+n->sel.hoff), &n->sel,n->fshift);
+			sel = ht->divisor&u32_hash_fold(*(__be32*)(ptr+n->sel.hoff), &n->sel,n->fshift);
 
 		if (!(n->sel.flags&(TC_U32_VAROFFSET|TC_U32_OFFSET|TC_U32_EAT)))
 			goto next_ht;
@@ -190,7 +190,7 @@ check_terminal:
 		if (n->sel.flags&(TC_U32_OFFSET|TC_U32_VAROFFSET)) {
 			off2 = n->sel.off + 3;
 			if (n->sel.flags&TC_U32_VAROFFSET)
-				off2 += ntohs(n->sel.offmask & *(u16*)(ptr+n->sel.offoff)) >>n->sel.offshift;
+				off2 += ntohs(n->sel.offmask & *(__be16*)(ptr+n->sel.offoff)) >>n->sel.offshift;
 			off2 &= ~3;
 		}
 		if (n->sel.flags&TC_U32_EAT) {
diff --git a/net/sched/em_u32.c b/net/sched/em_u32.c
index 112796e4a7c4..953f1479f7da 100644
--- a/net/sched/em_u32.c
+++ b/net/sched/em_u32.c
@@ -35,7 +35,7 @@ static int em_u32_match(struct sk_buff *skb, struct tcf_ematch *em,
 	if (!tcf_valid_offset(skb, ptr, sizeof(u32)))
 		return 0;
 
-	return !(((*(u32*) ptr)  ^ key->val) & key->mask);
+	return !(((*(__be32*) ptr)  ^ key->val) & key->mask);
 }
 
 static struct tcf_ematch_ops em_u32_ops = {
-- 
cgit v1.2.3


From 0ff9663c88ac5efdb5c8ac21c0bd7f993a4e3849 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 17 Mar 2008 22:48:46 -0700
Subject: [IPV4]: ipv4_is_lbcast() misannotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/in.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/in.h b/include/linux/in.h
index 70c6df882694..4065313cd7ee 100644
--- a/include/linux/in.h
+++ b/include/linux/in.h
@@ -265,7 +265,7 @@ static inline bool ipv4_is_local_multicast(__be32 addr)
 static inline bool ipv4_is_lbcast(__be32 addr)
 {
 	/* limited broadcast */
-	return addr == INADDR_BROADCAST;
+	return addr == htonl(INADDR_BROADCAST);
 }
 
 static inline bool ipv4_is_zeronet(__be32 addr)
-- 
cgit v1.2.3


From 4ae7d5cefd4aa3560e359a3b0f03e12adc8b5c86 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Wed, 19 Mar 2008 01:42:00 +0100
Subject: sched: improve affine wakeups

improve affine wakeups. Maintain the 'overlap' metric based on CFS's
sum_exec_runtime - which means the amount of time a task executes
after it wakes up some other task.

Use the 'overlap' for the wakeup decisions: if the 'overlap' is short,
it means there's strong workload coupling between this task and the
woken up task. If the 'overlap' is large then the workload is decoupled
and the scheduler will move them to separate CPUs more easily.

( Also slightly move the preempt_check within try_to_wake_up() - this has
  no effect on functionality but allows 'early wakeups' (for still-on-rq
  tasks) to be correctly accounted as well.)

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |  3 +++
 kernel/sched.c        |  5 ++++-
 kernel/sched_debug.c  |  1 +
 kernel/sched_fair.c   | 58 +++++++++++++++++++++++++++++++++++++--------------
 4 files changed, 50 insertions(+), 17 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 11d8e9a74eff..3625fcaf5d0f 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -929,6 +929,9 @@ struct sched_entity {
 	u64			vruntime;
 	u64			prev_sum_exec_runtime;
 
+	u64			last_wakeup;
+	u64			avg_overlap;
+
 #ifdef CONFIG_SCHEDSTATS
 	u64			wait_start;
 	u64			wait_max;
diff --git a/kernel/sched.c b/kernel/sched.c
index d1ad69b270ca..adbd475cfd25 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -1855,10 +1855,11 @@ out_activate:
 		schedstat_inc(p, se.nr_wakeups_remote);
 	update_rq_clock(rq);
 	activate_task(rq, p, 1);
-	check_preempt_curr(rq, p);
 	success = 1;
 
 out_running:
+	check_preempt_curr(rq, p);
+
 	p->state = TASK_RUNNING;
 #ifdef CONFIG_SMP
 	if (p->sched_class->task_wake_up)
@@ -1892,6 +1893,8 @@ static void __sched_fork(struct task_struct *p)
 	p->se.exec_start		= 0;
 	p->se.sum_exec_runtime		= 0;
 	p->se.prev_sum_exec_runtime	= 0;
+	p->se.last_wakeup		= 0;
+	p->se.avg_overlap		= 0;
 
 #ifdef CONFIG_SCHEDSTATS
 	p->se.wait_start		= 0;
diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c
index 4b5e24cf2f4a..ef358ba07683 100644
--- a/kernel/sched_debug.c
+++ b/kernel/sched_debug.c
@@ -288,6 +288,7 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m)
 	PN(se.exec_start);
 	PN(se.vruntime);
 	PN(se.sum_exec_runtime);
+	PN(se.avg_overlap);
 
 	nr_switches = p->nvcsw + p->nivcsw;
 
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index b5a357396b49..87c9d3a2aafa 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -556,6 +556,21 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int wakeup)
 	account_entity_enqueue(cfs_rq, se);
 }
 
+static void update_avg(u64 *avg, u64 sample)
+{
+	s64 diff = sample - *avg;
+	*avg += diff >> 3;
+}
+
+static void update_avg_stats(struct cfs_rq *cfs_rq, struct sched_entity *se)
+{
+	if (!se->last_wakeup)
+		return;
+
+	update_avg(&se->avg_overlap, se->sum_exec_runtime - se->last_wakeup);
+	se->last_wakeup = 0;
+}
+
 static void
 dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int sleep)
 {
@@ -566,6 +581,7 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int sleep)
 
 	update_stats_dequeue(cfs_rq, se);
 	if (sleep) {
+		update_avg_stats(cfs_rq, se);
 #ifdef CONFIG_SCHEDSTATS
 		if (entity_is_task(se)) {
 			struct task_struct *tsk = task_of(se);
@@ -981,12 +997,15 @@ static inline int wake_idle(int cpu, struct task_struct *p)
 
 #ifdef CONFIG_SMP
 
+static const struct sched_class fair_sched_class;
+
 static int
-wake_affine(struct rq *rq, struct sched_domain *this_sd, struct task_struct *p,
-	    int prev_cpu, int this_cpu, int sync, int idx,
-	    unsigned long load, unsigned long this_load,
+wake_affine(struct rq *rq, struct sched_domain *this_sd, struct rq *this_rq,
+	    struct task_struct *p, int prev_cpu, int this_cpu, int sync,
+	    int idx, unsigned long load, unsigned long this_load,
 	    unsigned int imbalance)
 {
+	struct task_struct *curr = this_rq->curr;
 	unsigned long tl = this_load;
 	unsigned long tl_per_task;
 
@@ -994,10 +1013,15 @@ wake_affine(struct rq *rq, struct sched_domain *this_sd, struct task_struct *p,
 		return 0;
 
 	/*
-	 * Attract cache-cold tasks on sync wakeups:
+	 * If the currently running task will sleep within
+	 * a reasonable amount of time then attract this newly
+	 * woken task:
 	 */
-	if (sync && !task_hot(p, rq->clock, this_sd))
-		return 1;
+	if (sync && curr->sched_class == &fair_sched_class) {
+		if (curr->se.avg_overlap < sysctl_sched_migration_cost &&
+				p->se.avg_overlap < sysctl_sched_migration_cost)
+			return 1;
+	}
 
 	schedstat_inc(p, se.nr_wakeups_affine_attempts);
 	tl_per_task = cpu_avg_load_per_task(this_cpu);
@@ -1030,18 +1054,16 @@ static int select_task_rq_fair(struct task_struct *p, int sync)
 	struct sched_domain *sd, *this_sd = NULL;
 	int prev_cpu, this_cpu, new_cpu;
 	unsigned long load, this_load;
+	struct rq *rq, *this_rq;
 	unsigned int imbalance;
-	struct rq *rq;
 	int idx;
 
 	prev_cpu	= task_cpu(p);
 	rq		= task_rq(p);
 	this_cpu	= smp_processor_id();
+	this_rq		= cpu_rq(this_cpu);
 	new_cpu		= prev_cpu;
 
-	if (prev_cpu == this_cpu)
-		goto out;
-
 	/*
 	 * 'this_sd' is the first domain that both
 	 * this_cpu and prev_cpu are present in:
@@ -1069,11 +1091,12 @@ static int select_task_rq_fair(struct task_struct *p, int sync)
 	load = source_load(prev_cpu, idx);
 	this_load = target_load(this_cpu, idx);
 
-	if (wake_affine(rq, this_sd, p, prev_cpu, this_cpu, sync, idx,
-				     load, this_load, imbalance)) {
-		new_cpu = this_cpu;
+	if (wake_affine(rq, this_sd, this_rq, p, prev_cpu, this_cpu, sync, idx,
+				     load, this_load, imbalance))
+		return this_cpu;
+
+	if (prev_cpu == this_cpu)
 		goto out;
-	}
 
 	/*
 	 * Start passive balancing when half the imbalance_pct
@@ -1083,8 +1106,7 @@ static int select_task_rq_fair(struct task_struct *p, int sync)
 		if (imbalance*this_load <= 100*load) {
 			schedstat_inc(this_sd, ttwu_move_balance);
 			schedstat_inc(p, se.nr_wakeups_passive);
-			new_cpu = this_cpu;
-			goto out;
+			return this_cpu;
 		}
 	}
 
@@ -1111,6 +1133,10 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p)
 		return;
 	}
 
+	se->last_wakeup = se->sum_exec_runtime;
+	if (unlikely(se == pse))
+		return;
+
 	cfs_rq_of(pse)->next = pse;
 
 	/*
-- 
cgit v1.2.3


From 33b0c4217dcd67b788318c3192a2912b530e4eef Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Sun, 16 Mar 2008 11:14:30 +0100
Subject: sched: tune multi-core idle balancing

WAKE_IDLE is too agressive on multi-core CPUs with the new
wake-affine code, keep it on for SMT/HT balancing alone
(where there's no cache affinity at all between logical CPUs).

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/topology.h | 1 -
 1 file changed, 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/topology.h b/include/linux/topology.h
index 2352f46160d3..2d8dac8799cf 100644
--- a/include/linux/topology.h
+++ b/include/linux/topology.h
@@ -138,7 +138,6 @@
 				| SD_BALANCE_FORK	\
 				| SD_BALANCE_EXEC	\
 				| SD_WAKE_AFFINE	\
-				| SD_WAKE_IDLE		\
 				| SD_SHARE_PKG_RESOURCES\
 				| BALANCE_FOR_MC_POWER,	\
 	.last_balance		= jiffies,		\
-- 
cgit v1.2.3


From ae66be9b71b12f16b84129860d06bbfe37fbec51 Mon Sep 17 00:00:00 2001
From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Date: Wed, 19 Mar 2008 17:00:57 -0700
Subject: rcu: fix misplaced mb() in rcu_enter/exit_nohz()

In the process of writing up the mechanical proof of correctness for the
dynticks/preemptable-RCU interface, I noticed misplaced memory barriers in
rcu_enter_nohz() and rcu_exit_nohz().

This patch puts them in the right place and adds a comment.  The key thing to
keep in mind is that rcu_enter_nohz() is -exiting- the mode that can legally
execute RCU read-side critical sections.

The memory barrier must be between any potential RCU read-side critical
sections and the increment of the per-CPU dynticks_progress_counter, and thus
must come -before- this increment.  And vice versa for rcu_exit_nohz().

The locking in the scheduler is probably saving us for the moment.

Also, switch to smp_mb() - we don't need a barrier for uniprocessor kernels.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Steven Rostedt <srostedt@redhat.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/rcupreempt.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h
index 01152ed532c8..d038aa6e5ee1 100644
--- a/include/linux/rcupreempt.h
+++ b/include/linux/rcupreempt.h
@@ -87,15 +87,15 @@ DECLARE_PER_CPU(long, dynticks_progress_counter);
 
 static inline void rcu_enter_nohz(void)
 {
+	smp_mb(); /* CPUs seeing ++ must see prior RCU read-side crit sects */
 	__get_cpu_var(dynticks_progress_counter)++;
 	WARN_ON(__get_cpu_var(dynticks_progress_counter) & 0x1);
-	mb();
 }
 
 static inline void rcu_exit_nohz(void)
 {
-	mb();
 	__get_cpu_var(dynticks_progress_counter)++;
+	smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */
 	WARN_ON(!(__get_cpu_var(dynticks_progress_counter) & 0x1));
 }
 
-- 
cgit v1.2.3


From a6b91919e0881a0d0a4ae5211d5c879a8c7ca92b Mon Sep 17 00:00:00 2001
From: Randy Dunlap <randy.dunlap@oracle.com>
Date: Wed, 19 Mar 2008 17:01:00 -0700
Subject: fs: fix kernel-doc notation warnings

Fix kernel-doc notation warnings in fs/.

Warning(mmotm-2008-0314-1449//fs/super.c:560): missing initial short description on line:
 *	mark_files_ro
Warning(mmotm-2008-0314-1449//fs/locks.c:1277): missing initial short description on line:
 *	lease_get_mtime
Warning(mmotm-2008-0314-1449//fs/locks.c:1277): missing initial short description on line:
 *	lease_get_mtime
Warning(mmotm-2008-0314-1449//fs/namei.c:1368): missing initial short description on line:
 * lookup_one_len:  filesystem helper to lookup single pathname component
Warning(mmotm-2008-0314-1449//fs/buffer.c:3221): missing initial short description on line:
 * bh_uptodate_or_lock: Test whether the buffer is uptodate
Warning(mmotm-2008-0314-1449//fs/buffer.c:3240): missing initial short description on line:
 * bh_submit_read: Submit a locked buffer for reading
Warning(mmotm-2008-0314-1449//fs/fs-writeback.c:30): missing initial short description on line:
 * writeback_acquire: attempt to get exclusive writeback access to a device
Warning(mmotm-2008-0314-1449//fs/fs-writeback.c:47): missing initial short description on line:
 * writeback_in_progress: determine whether there is writeback in progress
Warning(mmotm-2008-0314-1449//fs/fs-writeback.c:58): missing initial short description on line:
 * writeback_release: relinquish exclusive writeback access against a device.
Warning(mmotm-2008-0314-1449//include/linux/jbd.h:351): contents before sections
Warning(mmotm-2008-0314-1449//include/linux/jbd.h:561): contents before sections
Warning(mmotm-2008-0314-1449//fs/jbd/transaction.c:1935): missing initial short description on line:
 * void journal_invalidatepage()

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 fs/buffer.c          |  4 ++--
 fs/fs-writeback.c    |  6 +++---
 fs/jbd/transaction.c |  5 ++---
 fs/locks.c           |  4 ++--
 fs/namei.c           |  6 +++---
 fs/super.c           |  6 +++---
 include/linux/jbd.h  | 11 ++---------
 7 files changed, 17 insertions(+), 25 deletions(-)

(limited to 'include/linux')

diff --git a/fs/buffer.c b/fs/buffer.c
index ddfdd2c80bf9..7ba58386beee 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -3213,7 +3213,7 @@ static int buffer_cpu_notify(struct notifier_block *self,
 }
 
 /**
- * bh_uptodate_or_lock: Test whether the buffer is uptodate
+ * bh_uptodate_or_lock - Test whether the buffer is uptodate
  * @bh: struct buffer_head
  *
  * Return true if the buffer is up-to-date and false,
@@ -3232,7 +3232,7 @@ int bh_uptodate_or_lock(struct buffer_head *bh)
 EXPORT_SYMBOL(bh_uptodate_or_lock);
 
 /**
- * bh_submit_read: Submit a locked buffer for reading
+ * bh_submit_read - Submit a locked buffer for reading
  * @bh: struct buffer_head
  *
  * Returns zero on success and -EIO on error.
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index c0076077d338..06557679ca41 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -751,7 +751,7 @@ int generic_osync_inode(struct inode *inode, struct address_space *mapping, int
 EXPORT_SYMBOL(generic_osync_inode);
 
 /**
- * writeback_acquire: attempt to get exclusive writeback access to a device
+ * writeback_acquire - attempt to get exclusive writeback access to a device
  * @bdi: the device's backing_dev_info structure
  *
  * It is a waste of resources to have more than one pdflush thread blocked on
@@ -768,7 +768,7 @@ int writeback_acquire(struct backing_dev_info *bdi)
 }
 
 /**
- * writeback_in_progress: determine whether there is writeback in progress
+ * writeback_in_progress - determine whether there is writeback in progress
  * @bdi: the device's backing_dev_info structure.
  *
  * Determine whether there is writeback in progress against a backing device.
@@ -779,7 +779,7 @@ int writeback_in_progress(struct backing_dev_info *bdi)
 }
 
 /**
- * writeback_release: relinquish exclusive writeback access against a device.
+ * writeback_release - relinquish exclusive writeback access against a device.
  * @bdi: the device's backing_dev_info structure
  */
 void writeback_release(struct backing_dev_info *bdi)
diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c
index 4c895044c7d6..2c9e8f5d13aa 100644
--- a/fs/jbd/transaction.c
+++ b/fs/jbd/transaction.c
@@ -1904,13 +1904,12 @@ zap_buffer_unlocked:
 }
 
 /**
- * void journal_invalidatepage()
- * @journal: journal to use for flush...
+ * void journal_invalidatepage() - invalidate a journal page
+ * @journal: journal to use for flush
  * @page:    page to flush
  * @offset:  length of page to invalidate.
  *
  * Reap page buffers containing data after offset in page.
- *
  */
 void journal_invalidatepage(journal_t *journal,
 		      struct page *page,
diff --git a/fs/locks.c b/fs/locks.c
index f36f0e61558d..d83fab1b77b5 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -1275,13 +1275,13 @@ out:
 EXPORT_SYMBOL(__break_lease);
 
 /**
- *	lease_get_mtime
+ *	lease_get_mtime - get the last modified time of an inode
  *	@inode: the inode
  *      @time:  pointer to a timespec which will contain the last modified time
  *
  * This is to force NFS clients to flush their caches for files with
  * exclusive leases.  The justification is that if someone has an
- * exclusive lease, then they could be modifiying it.
+ * exclusive lease, then they could be modifying it.
  */
 void lease_get_mtime(struct inode *inode, struct timespec *time)
 {
diff --git a/fs/namei.c b/fs/namei.c
index 941c8e8228c0..6b7a0eef4090 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1364,13 +1364,13 @@ static int __lookup_one_len(const char *name, struct qstr *this,
 }
 
 /**
- * lookup_one_len:  filesystem helper to lookup single pathname component
+ * lookup_one_len - filesystem helper to lookup single pathname component
  * @name:	pathname component to lookup
  * @base:	base directory to lookup from
  * @len:	maximum length @len should be interpreted to
  *
- * Note that this routine is purely a helper for filesystem useage and should
- * not be called by generic code.  Also note that by using this function to
+ * Note that this routine is purely a helper for filesystem usage and should
+ * not be called by generic code.  Also note that by using this function the
  * nameidata argument is passed to the filesystem methods and a filesystem
  * using this helper needs to be prepared for that.
  */
diff --git a/fs/super.c b/fs/super.c
index 010446d8c40a..d0a941a4e620 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -556,11 +556,11 @@ out:
 }
 
 /**
- *	mark_files_ro
+ *	mark_files_ro - mark all files read-only
  *	@sb: superblock in question
  *
- *	All files are marked read/only.  We don't care about pending
- *	delete files so this should be used in 'force' mode only
+ *	All files are marked read-only.  We don't care about pending
+ *	delete files so this should be used in 'force' mode only.
  */
 
 static void mark_files_ro(struct super_block *sb)
diff --git a/include/linux/jbd.h b/include/linux/jbd.h
index b18fd3b9b835..423f58272188 100644
--- a/include/linux/jbd.h
+++ b/include/linux/jbd.h
@@ -348,8 +348,7 @@ static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh)
 struct jbd_revoke_table_s;
 
 /**
- * struct handle_s - The handle_s type is the concrete type associated with
- *     handle_t.
+ * struct handle_s - this is the concrete type associated with handle_t.
  * @h_transaction: Which compound transaction is this update a part of?
  * @h_buffer_credits: Number of remaining buffers we are allowed to dirty.
  * @h_ref: Reference count on this handle
@@ -358,12 +357,7 @@ struct jbd_revoke_table_s;
  * @h_jdata: flag to force data journaling
  * @h_aborted: flag indicating fatal error on handle
  * @h_lockdep_map: lockdep info for debugging lock problems
- **/
-
-/* Docbook can't yet cope with the bit fields, but will leave the documentation
- * in so it can be fixed later.
  */
-
 struct handle_s
 {
 	/* Which compound transaction is this update a part of? */
@@ -558,8 +552,7 @@ struct transaction_s
 };
 
 /**
- * struct journal_s - The journal_s type is the concrete type associated with
- *     journal_t.
+ * struct journal_s - this is the concrete type associated with journal_t.
  * @j_flags:  General journaling state flags
  * @j_errno:  Is there an outstanding uncleared error on the journal (from a
  *     prior abort)?
-- 
cgit v1.2.3


From ead70773608a5d97f81cb492f117d20b5e9f323e Mon Sep 17 00:00:00 2001
From: Alex Dubov <oakad@yahoo.com>
Date: Wed, 19 Mar 2008 17:01:06 -0700
Subject: memstick: automatically retrieve "INT" value from command response

MemoryStick storage cards, when in parallel mode, send several meaningful bits
of their "INT" register as part of command response.  This data is stored by
host and can be used to spare invocation of "GET_INT" TPC on each data page
transferred between host and card.

Signed-off-by: Alex Dubov <oakad@yahoo.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/memstick/core/memstick.c    |  9 +++----
 drivers/memstick/core/mspro_block.c | 15 +++++++-----
 drivers/memstick/host/jmb38x_ms.c   | 48 +++++++++++++++++++++----------------
 drivers/memstick/host/tifm_ms.c     | 17 +++++++------
 include/linux/memstick.h            |  1 -
 5 files changed, 49 insertions(+), 41 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/memstick/core/memstick.c b/drivers/memstick/core/memstick.c
index de80dba12f9b..946e3d3506ac 100644
--- a/drivers/memstick/core/memstick.c
+++ b/drivers/memstick/core/memstick.c
@@ -276,8 +276,6 @@ void memstick_init_req_sg(struct memstick_request *mrq, unsigned char tpc,
 		mrq->need_card_int = 1;
 	else
 		mrq->need_card_int = 0;
-
-	mrq->get_int_reg = 0;
 }
 EXPORT_SYMBOL(memstick_init_req_sg);
 
@@ -311,8 +309,6 @@ void memstick_init_req(struct memstick_request *mrq, unsigned char tpc,
 		mrq->need_card_int = 1;
 	else
 		mrq->need_card_int = 0;
-
-	mrq->get_int_reg = 0;
 }
 EXPORT_SYMBOL(memstick_init_req);
 
@@ -342,6 +338,7 @@ static int h_memstick_read_dev_id(struct memstick_dev *card,
 			card->id.class = id_reg.class;
 		}
 		complete(&card->mrq_complete);
+		dev_dbg(&card->dev, "if_mode = %02x\n", id_reg.if_mode);
 		return -EAGAIN;
 	}
 }
@@ -422,7 +419,6 @@ static void memstick_power_on(struct memstick_host *host)
 {
 	host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_ON);
 	host->set_param(host, MEMSTICK_INTERFACE, MEMSTICK_SERIAL);
-	msleep(1);
 }
 
 static void memstick_check(struct work_struct *work)
@@ -579,7 +575,8 @@ EXPORT_SYMBOL(memstick_suspend_host);
 void memstick_resume_host(struct memstick_host *host)
 {
 	mutex_lock(&host->lock);
-	host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_ON);
+	if (host->card)
+		memstick_power_on(host);
 	mutex_unlock(&host->lock);
 	memstick_detect_change(host);
 }
diff --git a/drivers/memstick/core/mspro_block.c b/drivers/memstick/core/mspro_block.c
index 1d637e4561d3..e5356f97d076 100644
--- a/drivers/memstick/core/mspro_block.c
+++ b/drivers/memstick/core/mspro_block.c
@@ -133,6 +133,7 @@ struct mspro_devinfo {
 struct mspro_block_data {
 	struct memstick_dev   *card;
 	unsigned int          usage_count;
+	unsigned int          caps;
 	struct gendisk        *disk;
 	struct request_queue  *queue;
 	spinlock_t            q_lock;
@@ -577,7 +578,6 @@ static int h_mspro_block_wait_for_ced(struct memstick_dev *card,
 static int h_mspro_block_transfer_data(struct memstick_dev *card,
 				       struct memstick_request **mrq)
 {
-	struct memstick_host *host = card->host;
 	struct mspro_block_data *msb = memstick_get_drvdata(card);
 	unsigned char t_val = 0;
 	struct scatterlist t_sg = { 0 };
@@ -591,12 +591,12 @@ static int h_mspro_block_transfer_data(struct memstick_dev *card,
 	switch ((*mrq)->tpc) {
 	case MS_TPC_WRITE_REG:
 		memstick_init_req(*mrq, MS_TPC_SET_CMD, &msb->transfer_cmd, 1);
-		(*mrq)->get_int_reg = 1;
+		(*mrq)->need_card_int = 1;
 		return 0;
 	case MS_TPC_SET_CMD:
 		t_val = (*mrq)->int_reg;
 		memstick_init_req(*mrq, MS_TPC_GET_INT, NULL, 1);
-		if (host->caps & MEMSTICK_CAP_AUTO_GET_INT)
+		if (msb->caps & MEMSTICK_CAP_AUTO_GET_INT)
 			goto has_int_reg;
 		return 0;
 	case MS_TPC_GET_INT:
@@ -646,12 +646,12 @@ has_int_reg:
 					   ? MS_TPC_READ_LONG_DATA
 					   : MS_TPC_WRITE_LONG_DATA,
 				     &t_sg);
-		(*mrq)->get_int_reg = 1;
+		(*mrq)->need_card_int = 1;
 		return 0;
 	case MS_TPC_READ_LONG_DATA:
 	case MS_TPC_WRITE_LONG_DATA:
 		msb->current_page++;
-		if (host->caps & MEMSTICK_CAP_AUTO_GET_INT) {
+		if (msb->caps & MEMSTICK_CAP_AUTO_GET_INT) {
 			t_val = (*mrq)->int_reg;
 			goto has_int_reg;
 		} else {
@@ -1052,7 +1052,8 @@ static int mspro_block_init_card(struct memstick_dev *card)
 	if (memstick_set_rw_addr(card))
 		return -EIO;
 
-	if (host->caps & MEMSTICK_CAP_PAR4) {
+	msb->caps = host->caps;
+	if (msb->caps & MEMSTICK_CAP_PAR4) {
 		if (mspro_block_switch_to_parallel(card))
 			printk(KERN_WARNING "%s: could not switch to "
 			       "parallel interface\n", card->dev.bus_id);
@@ -1062,6 +1063,8 @@ static int mspro_block_init_card(struct memstick_dev *card)
 	if (rc)
 		return rc;
 	dev_dbg(&card->dev, "card activated\n");
+	if (msb->system != MEMSTICK_SYS_SERIAL)
+		msb->caps |= MEMSTICK_CAP_AUTO_GET_INT;
 
 	card->next_request = h_mspro_block_req_init;
 	msb->mrq_handler = h_mspro_block_get_ro;
diff --git a/drivers/memstick/host/jmb38x_ms.c b/drivers/memstick/host/jmb38x_ms.c
index 9a57504183c7..f91037d50422 100644
--- a/drivers/memstick/host/jmb38x_ms.c
+++ b/drivers/memstick/host/jmb38x_ms.c
@@ -57,8 +57,6 @@ struct jmb38x_ms_host {
 	unsigned long           timeout_jiffies;
 	struct timer_list       timer;
 	struct memstick_request *req;
-	unsigned char           eject:1,
-				use_dma:1;
 	unsigned char           cmd_flags;
 	unsigned char           io_pos;
 	unsigned int            io_word[2];
@@ -95,9 +93,22 @@ struct jmb38x_ms {
 #define HOST_CONTROL_IF_PAR4   0x1
 #define HOST_CONTROL_IF_PAR8   0x3
 
+#define STATUS_BUSY             0x00080000
+#define STATUS_MS_DAT7          0x00040000
+#define STATUS_MS_DAT6          0x00020000
+#define STATUS_MS_DAT5          0x00010000
+#define STATUS_MS_DAT4          0x00008000
+#define STATUS_MS_DAT3          0x00004000
+#define STATUS_MS_DAT2          0x00002000
+#define STATUS_MS_DAT1          0x00001000
+#define STATUS_MS_DAT0          0x00000800
 #define STATUS_HAS_MEDIA        0x00000400
 #define STATUS_FIFO_EMPTY       0x00000200
 #define STATUS_FIFO_FULL        0x00000100
+#define STATUS_MS_CED           0x00000080
+#define STATUS_MS_ERR           0x00000040
+#define STATUS_MS_BRQ           0x00000020
+#define STATUS_MS_CNK           0x00000001
 
 #define INT_STATUS_TPC_ERR      0x00080000
 #define INT_STATUS_CRC_ERR      0x00040000
@@ -124,7 +135,7 @@ enum {
 	CMD_READY    = 0x01,
 	FIFO_READY   = 0x02,
 	REG_DATA     = 0x04,
-	AUTO_GET_INT = 0x08
+	DMA_DATA     = 0x08
 };
 
 static unsigned int jmb38x_ms_read_data(struct jmb38x_ms_host *host,
@@ -367,28 +378,27 @@ static int jmb38x_ms_issue_cmd(struct memstick_host *msh)
 		cmd |= TPC_DIR;
 	if (host->req->need_card_int)
 		cmd |= TPC_WAIT_INT;
-	if (host->req->get_int_reg)
-		cmd |= TPC_GET_INT;
 
 	data = host->req->data;
 
-	host->use_dma = !no_dma;
+	if (!no_dma)
+		host->cmd_flags |= DMA_DATA;
 
 	if (host->req->long_data) {
 		data_len = host->req->sg.length;
 	} else {
 		data_len = host->req->data_len;
-		host->use_dma = 0;
+		host->cmd_flags &= ~DMA_DATA;
 	}
 
 	if (data_len <= 8) {
 		cmd &= ~(TPC_DATA_SEL | 0xf);
 		host->cmd_flags |= REG_DATA;
 		cmd |= data_len & 0xf;
-		host->use_dma = 0;
+		host->cmd_flags &= ~DMA_DATA;
 	}
 
-	if (host->use_dma) {
+	if (host->cmd_flags & DMA_DATA) {
 		if (1 != pci_map_sg(host->chip->pdev, &host->req->sg, 1,
 				    host->req->data_dir == READ
 				    ? PCI_DMA_FROMDEVICE
@@ -451,13 +461,12 @@ static void jmb38x_ms_complete_cmd(struct memstick_host *msh, int last)
 		readl(host->addr + INT_STATUS));
 	dev_dbg(msh->cdev.dev, "c hstatus %08x\n", readl(host->addr + STATUS));
 
-	if (host->req->get_int_reg) {
-		t_val = readl(host->addr + TPC_P0);
-		host->req->int_reg = (t_val & 0xff);
-	}
+	host->req->int_reg = readl(host->addr + STATUS) & 0xff;
+
+	writel(0, host->addr + BLOCK);
+	writel(0, host->addr + DMA_CONTROL);
 
-	if (host->use_dma) {
-		writel(0, host->addr + DMA_CONTROL);
+	if (host->cmd_flags & DMA_DATA) {
 		pci_unmap_sg(host->chip->pdev, &host->req->sg, 1,
 			     host->req->data_dir == READ
 			     ? PCI_DMA_FROMDEVICE : PCI_DMA_TODEVICE);
@@ -509,7 +518,7 @@ static irqreturn_t jmb38x_ms_isr(int irq, void *dev_id)
 			else
 				host->req->error = -ETIME;
 		} else {
-			if (host->use_dma) {
+			if (host->cmd_flags & DMA_DATA) {
 				if (irq_status & INT_STATUS_EOTRAN)
 					host->cmd_flags |= FIFO_READY;
 			} else {
@@ -775,13 +784,10 @@ static struct memstick_host *jmb38x_ms_alloc_host(struct jmb38x_ms *jm, int cnt)
 	snprintf(host->host_id, DEVICE_ID_SIZE, DRIVER_NAME ":slot%d",
 		 host->id);
 	host->irq = jm->pdev->irq;
-	host->timeout_jiffies = msecs_to_jiffies(4000);
+	host->timeout_jiffies = msecs_to_jiffies(1000);
 	msh->request = jmb38x_ms_request;
 	msh->set_param = jmb38x_ms_set_param;
-	/*
-	msh->caps = MEMSTICK_CAP_AUTO_GET_INT | MEMSTICK_CAP_PAR4
-		    | MEMSTICK_CAP_PAR8;
-	*/
+
 	msh->caps = MEMSTICK_CAP_PAR4 | MEMSTICK_CAP_PAR8;
 
 	setup_timer(&host->timer, jmb38x_ms_abort, (unsigned long)msh);
diff --git a/drivers/memstick/host/tifm_ms.c b/drivers/memstick/host/tifm_ms.c
index 2b5bf52a8302..eb150dfb637f 100644
--- a/drivers/memstick/host/tifm_ms.c
+++ b/drivers/memstick/host/tifm_ms.c
@@ -340,11 +340,20 @@ static void tifm_ms_complete_cmd(struct tifm_ms *host)
 
 	del_timer(&host->timer);
 
-	if (host->use_dma)
+	host->req->int_reg = readl(sock->addr + SOCK_MS_STATUS) & 0xff;
+	host->req->int_reg = (host->req->int_reg & 1)
+			     | ((host->req->int_reg << 4) & 0xe0);
+
+	writel(TIFM_FIFO_INT_SETALL,
+	       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_CLEAR);
+	writel(TIFM_DMA_RESET, sock->addr + SOCK_DMA_CONTROL);
+
+	if (host->use_dma) {
 		tifm_unmap_sg(sock, &host->req->sg, 1,
 			      host->req->data_dir == READ
 			      ? PCI_DMA_FROMDEVICE
 			      : PCI_DMA_TODEVICE);
+	}
 
 	writel((~TIFM_CTRL_LED) & readl(sock->addr + SOCK_CONTROL),
 	       sock->addr + SOCK_CONTROL);
@@ -424,12 +433,6 @@ static void tifm_ms_card_event(struct tifm_dev *sock)
 		else if (host_status & TIFM_MS_STAT_CRC)
 			host->req->error = -EILSEQ;
 
-		if (host->req->error) {
-			writel(TIFM_FIFO_INT_SETALL,
-			       sock->addr + SOCK_DMA_FIFO_INT_ENABLE_CLEAR);
-			writel(TIFM_DMA_RESET, sock->addr + SOCK_DMA_CONTROL);
-		}
-
 		if (host_status & TIFM_MS_STAT_RDY)
 			host->cmd_flags |= CMD_READY;
 
diff --git a/include/linux/memstick.h b/include/linux/memstick.h
index b7ee25888836..3e686ec6a967 100644
--- a/include/linux/memstick.h
+++ b/include/linux/memstick.h
@@ -239,7 +239,6 @@ struct memstick_request {
 	unsigned char tpc;
 	unsigned char data_dir:1,
 		      need_card_int:1,
-		      get_int_reg:1,
 		      long_data:1;
 	unsigned char int_reg;
 	int           error;
-- 
cgit v1.2.3


From aedb60a67c10a0861af179725d060765262ba0fb Mon Sep 17 00:00:00 2001
From: Serge Hallyn <serge@hallyn.com>
Date: Fri, 29 Feb 2008 15:14:57 +0000
Subject: file capabilities: remove cap_task_kill()

The original justification for cap_task_kill() was as follows:

	check_kill_permission() does appropriate uid equivalence checks.
	However with file capabilities it becomes possible for an
	unprivileged user to execute a file with file capabilities
	resulting in a more privileged task with the same uid.

However now that cap_task_kill() always returns 0 (permission
granted) when p->uid==current->uid, the whole hook is worthless,
and only likely to create more subtle problems in the corner cases
where it might still be called but return -EPERM.  Those cases
are basically when uids are different but euid/suid is equivalent
as per the check in check_kill_permission().

One example of a still-broken application is 'at' for non-root users.

This patch removes cap_task_kill().

Signed-off-by: Serge Hallyn <serge@hallyn.com>
Acked-by: Andrew G. Morgan <morgan@kernel.org>
Earlier-version-tested-by: Luiz Fernando N. Capitulino <lcapitulino@mandriva.com.br>
Acked-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/security.h   |  3 +--
 security/capability.c      |  1 -
 security/commoncap.c       | 40 ----------------------------------------
 security/smack/smack_lsm.c |  5 -----
 4 files changed, 1 insertion(+), 48 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/security.h b/include/linux/security.h
index b07357ca2137..c673dfd4dffc 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -57,7 +57,6 @@ extern int cap_inode_need_killpriv(struct dentry *dentry);
 extern int cap_inode_killpriv(struct dentry *dentry);
 extern int cap_task_post_setuid (uid_t old_ruid, uid_t old_euid, uid_t old_suid, int flags);
 extern void cap_task_reparent_to_init (struct task_struct *p);
-extern int cap_task_kill(struct task_struct *p, struct siginfo *info, int sig, u32 secid);
 extern int cap_task_setscheduler (struct task_struct *p, int policy, struct sched_param *lp);
 extern int cap_task_setioprio (struct task_struct *p, int ioprio);
 extern int cap_task_setnice (struct task_struct *p, int nice);
@@ -2187,7 +2186,7 @@ static inline int security_task_kill (struct task_struct *p,
 				      struct siginfo *info, int sig,
 				      u32 secid)
 {
-	return cap_task_kill(p, info, sig, secid);
+	return 0;
 }
 
 static inline int security_task_wait (struct task_struct *p)
diff --git a/security/capability.c b/security/capability.c
index 9e99f36a8b5c..2c6e06d18fab 100644
--- a/security/capability.c
+++ b/security/capability.c
@@ -40,7 +40,6 @@ static struct security_operations capability_ops = {
 	.inode_need_killpriv =		cap_inode_need_killpriv,
 	.inode_killpriv =		cap_inode_killpriv,
 
-	.task_kill =			cap_task_kill,
 	.task_setscheduler =		cap_task_setscheduler,
 	.task_setioprio =		cap_task_setioprio,
 	.task_setnice =			cap_task_setnice,
diff --git a/security/commoncap.c b/security/commoncap.c
index bb0c095f5761..06d5c9469ba3 100644
--- a/security/commoncap.c
+++ b/security/commoncap.c
@@ -540,41 +540,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
 	return cap_safe_nice(p);
 }
 
-int cap_task_kill(struct task_struct *p, struct siginfo *info,
-				int sig, u32 secid)
-{
-	if (info != SEND_SIG_NOINFO && (is_si_special(info) || SI_FROMKERNEL(info)))
-		return 0;
-
-	/*
-	 * Running a setuid root program raises your capabilities.
-	 * Killing your own setuid root processes was previously
-	 * allowed.
-	 * We must preserve legacy signal behavior in this case.
-	 */
-	if (p->uid == current->uid)
-		return 0;
-
-	/* sigcont is permitted within same session */
-	if (sig == SIGCONT && (task_session_nr(current) == task_session_nr(p)))
-		return 0;
-
-	if (secid)
-		/*
-		 * Signal sent as a particular user.
-		 * Capabilities are ignored.  May be wrong, but it's the
-		 * only thing we can do at the moment.
-		 * Used only by usb drivers?
-		 */
-		return 0;
-	if (cap_issubset(p->cap_permitted, current->cap_permitted))
-		return 0;
-	if (capable(CAP_KILL))
-		return 0;
-
-	return -EPERM;
-}
-
 /*
  * called from kernel/sys.c for prctl(PR_CABSET_DROP)
  * done without task_capability_lock() because it introduces
@@ -605,11 +570,6 @@ int cap_task_setnice (struct task_struct *p, int nice)
 {
 	return 0;
 }
-int cap_task_kill(struct task_struct *p, struct siginfo *info,
-				int sig, u32 secid)
-{
-	return 0;
-}
 #endif
 
 void cap_task_reparent_to_init (struct task_struct *p)
diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
index 38d707593b31..732ba27923c4 100644
--- a/security/smack/smack_lsm.c
+++ b/security/smack/smack_lsm.c
@@ -1117,11 +1117,6 @@ static int smack_task_movememory(struct task_struct *p)
 static int smack_task_kill(struct task_struct *p, struct siginfo *info,
 			   int sig, u32 secid)
 {
-	int rc;
-
-	rc = cap_task_kill(p, info, sig, secid);
-	if (rc != 0)
-		return rc;
 	/*
 	 * Special cases where signals really ought to go through
 	 * in spite of policy. Stephen Smalley suggests it may
-- 
cgit v1.2.3


From 82cc1a7a56872056af0ead6c7d695aa223f36695 Mon Sep 17 00:00:00 2001
From: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Date: Fri, 21 Mar 2008 03:43:19 -0700
Subject: [NET]: Add per-connection option to set max TSO frame size

Update: My mailer ate one of Jarek's feedback mails...  Fixed the
parameter in netif_set_gso_max_size() to be u32, not u16.  Fixed the
whitespace issue due to a patch import botch.  Changed the types from
u32 to unsigned int to be more consistent with other variables in the
area.  Also brought the patch up to the latest net-2.6.26 tree.

Update: Made gso_max_size container 32 bits, not 16.  Moved the
location of gso_max_size within netdev to be less hotpath.  Made more
consistent names between the sock and netdev layers, and added a
define for the max GSO size.

Update: Respun for net-2.6.26 tree.

Update: changed max_gso_frame_size and sk_gso_max_size from signed to
unsigned - thanks Stephen!

This patch adds the ability for device drivers to control the size of
the TSO frames being sent to them, per TCP connection.  By setting the
netdevice's gso_max_size value, the socket layer will set the GSO
frame size based on that value.  This will propogate into the TCP
layer, and send TSO's of that size to the hardware.

This can be desirable to help tune the bursty nature of TSO on a
per-adapter basis, where one may have 1 GbE and 10 GbE devices
coexisting in a system, one running multiqueue and the other not, etc.

This can also be desirable for devices that cannot support full 64 KB
TSO's, but still want to benefit from some level of segmentation
offloading.

Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 10 ++++++++++
 include/net/sock.h        |  2 ++
 net/core/dev.c            |  1 +
 net/core/sock.c           |  6 ++++--
 net/ipv4/tcp_output.c     |  4 ++--
 5 files changed, 19 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index a2f003239c85..ced61f87660e 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -724,6 +724,10 @@ struct net_device
 	/* rtnetlink link ops */
 	const struct rtnl_link_ops *rtnl_link_ops;
 
+	/* for setting kernel sock attribute on TCP connection setup */
+#define GSO_MAX_SIZE		65536
+	unsigned int		gso_max_size;
+
 	/* The TX queue control structures */
 	unsigned int			egress_subqueue_count;
 	struct net_device_subqueue	egress_subqueue[1];
@@ -1475,6 +1479,12 @@ static inline int netif_needs_gso(struct net_device *dev, struct sk_buff *skb)
 		unlikely(skb->ip_summed != CHECKSUM_PARTIAL));
 }
 
+static inline void netif_set_gso_max_size(struct net_device *dev,
+					  unsigned int size)
+{
+	dev->gso_max_size = size;
+}
+
 /* On bonding slaves other than the currently active slave, suppress
  * duplicates except for 802.3ad ETH_P_SLOW, alb non-mcast/bcast, and
  * ARP on active-backup slaves with arp_validate enabled.
diff --git a/include/net/sock.h b/include/net/sock.h
index 39112e75411c..8358fff002eb 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -151,6 +151,7 @@ struct sock_common {
   *	@sk_no_check: %SO_NO_CHECK setting, wether or not checkup packets
   *	@sk_route_caps: route capabilities (e.g. %NETIF_F_TSO)
   *	@sk_gso_type: GSO type (e.g. %SKB_GSO_TCPV4)
+  *	@sk_gso_max_size: Maximum GSO segment size to build
   *	@sk_lingertime: %SO_LINGER l_linger setting
   *	@sk_backlog: always used with the per-socket spinlock held
   *	@sk_callback_lock: used with the callbacks in the end of this struct
@@ -237,6 +238,7 @@ struct sock {
 	gfp_t			sk_allocation;
 	int			sk_route_caps;
 	int			sk_gso_type;
+	unsigned int		sk_gso_max_size;
 	int			sk_rcvlowat;
 	unsigned long 		sk_flags;
 	unsigned long	        sk_lingertime;
diff --git a/net/core/dev.c b/net/core/dev.c
index fcdf03cf3b3f..f973e38b81af 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4021,6 +4021,7 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
 	}
 
 	dev->egress_subqueue_count = queue_count;
+	dev->gso_max_size = GSO_MAX_SIZE;
 
 	dev->get_stats = internal_stats;
 	netpoll_netdev_init(dev);
diff --git a/net/core/sock.c b/net/core/sock.c
index bb5236aee643..b1a6ed4d33c1 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -1095,10 +1095,12 @@ void sk_setup_caps(struct sock *sk, struct dst_entry *dst)
 	if (sk->sk_route_caps & NETIF_F_GSO)
 		sk->sk_route_caps |= NETIF_F_GSO_SOFTWARE;
 	if (sk_can_gso(sk)) {
-		if (dst->header_len)
+		if (dst->header_len) {
 			sk->sk_route_caps &= ~NETIF_F_GSO_MASK;
-		else
+		} else {
 			sk->sk_route_caps |= NETIF_F_SG | NETIF_F_HW_CSUM;
+			sk->sk_gso_max_size = dst->dev->gso_max_size;
+		}
 	}
 }
 EXPORT_SYMBOL_GPL(sk_setup_caps);
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index b4e11d834c9f..a627616314ba 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -998,7 +998,7 @@ unsigned int tcp_current_mss(struct sock *sk, int large_allowed)
 	xmit_size_goal = mss_now;
 
 	if (doing_tso) {
-		xmit_size_goal = (65535 -
+		xmit_size_goal = ((sk->sk_gso_max_size - 1) -
 				  inet_csk(sk)->icsk_af_ops->net_header_len -
 				  inet_csk(sk)->icsk_ext_hdr_len -
 				  tp->tcp_header_len);
@@ -1282,7 +1282,7 @@ static int tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb)
 	limit = min(send_win, cong_win);
 
 	/* If a full-sized TSO skb can be sent, do it. */
-	if (limit >= 65536)
+	if (limit >= sk->sk_gso_max_size)
 		goto send_now;
 
 	if (sysctl_tcp_tso_win_divisor) {
-- 
cgit v1.2.3


From 9aefd0abd8610e8f3bb097debf3afb73f8b7b210 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 12 Mar 2008 18:31:58 +0100
Subject: sched: add exported arch_reinit_sched_domains() to header file.

Needed so it can be called from outside of sched.c.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h | 1 +
 kernel/sched.c        | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 3625fcaf5d0f..fed07d03364e 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -790,6 +790,7 @@ struct sched_domain {
 };
 
 extern void partition_sched_domains(int ndoms_new, cpumask_t *doms_new);
+extern int arch_reinit_sched_domains(void);
 
 #endif	/* CONFIG_SMP */
 
diff --git a/kernel/sched.c b/kernel/sched.c
index 573179eb553e..78482e51b583 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -6920,7 +6920,7 @@ match2:
 }
 
 #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT)
-static int arch_reinit_sched_domains(void)
+int arch_reinit_sched_domains(void)
 {
 	int err;
 
-- 
cgit v1.2.3


From 22e52b072dd87faa9b2559fe89d4e8f2370f81ca Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 12 Mar 2008 18:31:59 +0100
Subject: sched: add arch_update_cpu_topology hook.

Will be called each time the scheduling domains are rebuild.
Needed for architectures that don't have a static cpu topology.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/topology.h | 2 ++
 kernel/sched.c           | 5 +++++
 2 files changed, 7 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/topology.h b/include/linux/topology.h
index 2d8dac8799cf..bd14f8b30f09 100644
--- a/include/linux/topology.h
+++ b/include/linux/topology.h
@@ -50,6 +50,8 @@
 	for_each_online_node(node)						\
 		if (nr_cpus_node(node))
 
+void arch_update_cpu_topology(void);
+
 /* Conform to ACPI 2.0 SLIT distance definitions */
 #define LOCAL_DISTANCE		10
 #define REMOTE_DISTANCE		20
diff --git a/kernel/sched.c b/kernel/sched.c
index 78482e51b583..28c73f07efb2 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -6807,6 +6807,10 @@ static int ndoms_cur;		/* number of sched domains in 'doms_cur' */
  */
 static cpumask_t fallback_doms;
 
+void __attribute__((weak)) arch_update_cpu_topology(void)
+{
+}
+
 /*
  * Set up scheduler domains and groups. Callers must hold the hotplug lock.
  * For now this just excludes isolated cpus, but could be used to
@@ -6816,6 +6820,7 @@ static int arch_init_sched_domains(const cpumask_t *cpu_map)
 {
 	int err;
 
+	arch_update_cpu_topology();
 	ndoms_cur = 1;
 	doms_cur = kmalloc(sizeof(cpumask_t), GFP_KERNEL);
 	if (!doms_cur)
-- 
cgit v1.2.3


From ec3c0982a2dd1e671bad8e9d26c28dcba0039d87 Mon Sep 17 00:00:00 2001
From: Patrick McManus <mcmanus@ducksong.com>
Date: Fri, 21 Mar 2008 16:33:01 -0700
Subject: [TCP]: TCP_DEFER_ACCEPT updates - process as established

Change TCP_DEFER_ACCEPT implementation so that it transitions a
connection to ESTABLISHED after handshake is complete instead of
leaving it in SYN-RECV until some data arrvies. Place connection in
accept queue when first data packet arrives from slow path.

Benefits:
  - established connection is now reset if it never makes it
   to the accept queue

 - diagnostic state of established matches with the packet traces
   showing completed handshake

 - TCP_DEFER_ACCEPT timeouts are expressed in seconds and can now be
   enforced with reasonable accuracy instead of rounding up to next
   exponential back-off of syn-ack retry.

Signed-off-by: Patrick McManus <mcmanus@ducksong.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/tcp.h             |  7 +++++++
 include/net/request_sock.h      |  4 ++--
 include/net/tcp.h               |  1 +
 net/ipv4/inet_connection_sock.c | 11 +++-------
 net/ipv4/tcp.c                  | 18 +++++++---------
 net/ipv4/tcp_input.c            | 46 +++++++++++++++++++++++++++++++++++++++++
 net/ipv4/tcp_ipv4.c             |  8 +++++++
 net/ipv4/tcp_minisocks.c        | 32 +++++++++++++++++-----------
 net/ipv4/tcp_timer.c            |  5 +++++
 9 files changed, 99 insertions(+), 33 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 08027f1d7f31..d96d9b122304 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -239,6 +239,11 @@ static inline struct tcp_request_sock *tcp_rsk(const struct request_sock *req)
 	return (struct tcp_request_sock *)req;
 }
 
+struct tcp_deferred_accept_info {
+	struct sock *listen_sk;
+	struct request_sock *request;
+};
+
 struct tcp_sock {
 	/* inet_connection_sock has to be the first member of tcp_sock */
 	struct inet_connection_sock	inet_conn;
@@ -374,6 +379,8 @@ struct tcp_sock {
 	unsigned int		keepalive_intvl;  /* time interval between keep alive probes */
 	int			linger2;
 
+	struct tcp_deferred_accept_info defer_tcp_accept;
+
 	unsigned long last_synq_overflow; 
 
 	u32	tso_deferred;
diff --git a/include/net/request_sock.h b/include/net/request_sock.h
index 040780add355..0369f98e9f3a 100644
--- a/include/net/request_sock.h
+++ b/include/net/request_sock.h
@@ -115,8 +115,8 @@ struct request_sock_queue {
 	struct request_sock	*rskq_accept_head;
 	struct request_sock	*rskq_accept_tail;
 	rwlock_t		syn_wait_lock;
-	u8			rskq_defer_accept;
-	/* 3 bytes hole, try to pack */
+	u16			rskq_defer_accept;
+	/* 2 bytes hole, try to pack */
 	struct listen_sock	*listen_opt;
 };
 
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 847e1634e1f4..67cc3956d29c 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -139,6 +139,7 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
 #define MAX_TCP_KEEPINTVL	32767
 #define MAX_TCP_KEEPCNT		127
 #define MAX_TCP_SYNCNT		127
+#define MAX_TCP_ACCEPT_DEFERRED 65535
 
 #define TCP_SYNQ_INTERVAL	(HZ/5)	/* Period of SYNACK timer */
 
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 8a45be988709..cc1a1859a61b 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -414,8 +414,7 @@ void inet_csk_reqsk_queue_prune(struct sock *parent,
 	struct inet_connection_sock *icsk = inet_csk(parent);
 	struct request_sock_queue *queue = &icsk->icsk_accept_queue;
 	struct listen_sock *lopt = queue->listen_opt;
-	int max_retries = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries;
-	int thresh = max_retries;
+	int thresh = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries;
 	unsigned long now = jiffies;
 	struct request_sock **reqp, *req;
 	int i, budget;
@@ -451,9 +450,6 @@ void inet_csk_reqsk_queue_prune(struct sock *parent,
 		}
 	}
 
-	if (queue->rskq_defer_accept)
-		max_retries = queue->rskq_defer_accept;
-
 	budget = 2 * (lopt->nr_table_entries / (timeout / interval));
 	i = lopt->clock_hand;
 
@@ -461,9 +457,8 @@ void inet_csk_reqsk_queue_prune(struct sock *parent,
 		reqp=&lopt->syn_table[i];
 		while ((req = *reqp) != NULL) {
 			if (time_after_eq(now, req->expires)) {
-				if ((req->retrans < (inet_rsk(req)->acked ? max_retries : thresh)) &&
-				    (inet_rsk(req)->acked ||
-				     !req->rsk_ops->rtx_syn_ack(parent, req))) {
+				if (req->retrans < thresh &&
+				    !req->rsk_ops->rtx_syn_ack(parent, req)) {
 					unsigned long timeo;
 
 					if (req->retrans++ == 0)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 071e83a894ad..e0fbc25ca816 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2105,15 +2105,12 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
 		break;
 
 	case TCP_DEFER_ACCEPT:
-		icsk->icsk_accept_queue.rskq_defer_accept = 0;
-		if (val > 0) {
-			/* Translate value in seconds to number of
-			 * retransmits */
-			while (icsk->icsk_accept_queue.rskq_defer_accept < 32 &&
-			       val > ((TCP_TIMEOUT_INIT / HZ) <<
-				       icsk->icsk_accept_queue.rskq_defer_accept))
-				icsk->icsk_accept_queue.rskq_defer_accept++;
-			icsk->icsk_accept_queue.rskq_defer_accept++;
+		if (val < 0) {
+			err = -EINVAL;
+		} else {
+			if (val > MAX_TCP_ACCEPT_DEFERRED)
+				val = MAX_TCP_ACCEPT_DEFERRED;
+			icsk->icsk_accept_queue.rskq_defer_accept = val;
 		}
 		break;
 
@@ -2295,8 +2292,7 @@ static int do_tcp_getsockopt(struct sock *sk, int level,
 			val = (val ? : sysctl_tcp_fin_timeout) / HZ;
 		break;
 	case TCP_DEFER_ACCEPT:
-		val = !icsk->icsk_accept_queue.rskq_defer_accept ? 0 :
-			((TCP_TIMEOUT_INIT / HZ) << (icsk->icsk_accept_queue.rskq_defer_accept - 1));
+		val = icsk->icsk_accept_queue.rskq_defer_accept;
 		break;
 	case TCP_WINDOW_CLAMP:
 		val = tp->window_clamp;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 9cf446427cc2..6e46b4c0f28c 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -4451,6 +4451,49 @@ static void tcp_urg(struct sock *sk, struct sk_buff *skb, struct tcphdr *th)
 	}
 }
 
+static int tcp_defer_accept_check(struct sock *sk)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+
+	if (tp->defer_tcp_accept.request) {
+		int queued_data =  tp->rcv_nxt - tp->copied_seq;
+		int hasfin =  !skb_queue_empty(&sk->sk_receive_queue) ?
+			tcp_hdr((struct sk_buff *)
+				sk->sk_receive_queue.prev)->fin : 0;
+
+		if (queued_data && hasfin)
+			queued_data--;
+
+		if (queued_data &&
+		    tp->defer_tcp_accept.listen_sk->sk_state == TCP_LISTEN) {
+			if (sock_flag(sk, SOCK_KEEPOPEN)) {
+				inet_csk_reset_keepalive_timer(sk,
+							       keepalive_time_when(tp));
+			} else {
+				inet_csk_delete_keepalive_timer(sk);
+			}
+
+			inet_csk_reqsk_queue_add(
+				tp->defer_tcp_accept.listen_sk,
+				tp->defer_tcp_accept.request,
+				sk);
+
+			tp->defer_tcp_accept.listen_sk->sk_data_ready(
+				tp->defer_tcp_accept.listen_sk, 0);
+
+			sock_put(tp->defer_tcp_accept.listen_sk);
+			sock_put(sk);
+			tp->defer_tcp_accept.listen_sk = NULL;
+			tp->defer_tcp_accept.request = NULL;
+		} else if (hasfin ||
+			   tp->defer_tcp_accept.listen_sk->sk_state != TCP_LISTEN) {
+			tcp_reset(sk);
+			return -1;
+		}
+	}
+	return 0;
+}
+
 static int tcp_copy_to_iovec(struct sock *sk, struct sk_buff *skb, int hlen)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
@@ -4811,6 +4854,9 @@ step5:
 
 	tcp_data_snd_check(sk);
 	tcp_ack_snd_check(sk);
+
+	if (tcp_defer_accept_check(sk))
+		return -1;
 	return 0;
 
 csum_error:
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 0ba6e911c979..167a0f557531 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1920,6 +1920,14 @@ int tcp_v4_destroy_sock(struct sock *sk)
 		sk->sk_sndmsg_page = NULL;
 	}
 
+	if (tp->defer_tcp_accept.request) {
+		reqsk_free(tp->defer_tcp_accept.request);
+		sock_put(tp->defer_tcp_accept.listen_sk);
+		sock_put(sk);
+		tp->defer_tcp_accept.listen_sk = NULL;
+		tp->defer_tcp_accept.request = NULL;
+	}
+
 	atomic_dec(&tcp_sockets_allocated);
 
 	return 0;
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 8245247a6ceb..019c8c16e5cc 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -571,10 +571,8 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
 	   does sequence test, SYN is truncated, and thus we consider
 	   it a bare ACK.
 
-	   If icsk->icsk_accept_queue.rskq_defer_accept, we silently drop this
-	   bare ACK.  Otherwise, we create an established connection.  Both
-	   ends (listening sockets) accept the new incoming connection and try
-	   to talk to each other. 8-)
+	   Both ends (listening sockets) accept the new incoming
+	   connection and try to talk to each other. 8-)
 
 	   Note: This case is both harmless, and rare.  Possibility is about the
 	   same as us discovering intelligent life on another plant tomorrow.
@@ -642,13 +640,6 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
 		if (!(flg & TCP_FLAG_ACK))
 			return NULL;
 
-		/* If TCP_DEFER_ACCEPT is set, drop bare ACK. */
-		if (inet_csk(sk)->icsk_accept_queue.rskq_defer_accept &&
-		    TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) {
-			inet_rsk(req)->acked = 1;
-			return NULL;
-		}
-
 		/* OK, ACK is valid, create big socket and
 		 * feed this segment to it. It will repeat all
 		 * the tests. THIS SEGMENT MUST MOVE SOCKET TO
@@ -687,7 +678,24 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
 		inet_csk_reqsk_queue_unlink(sk, req, prev);
 		inet_csk_reqsk_queue_removed(sk, req);
 
-		inet_csk_reqsk_queue_add(sk, req, child);
+		if (inet_csk(sk)->icsk_accept_queue.rskq_defer_accept &&
+		    TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) {
+
+			/* the accept queue handling is done is est recv slow
+			 * path so lets make sure to start there
+			 */
+			tcp_sk(child)->pred_flags = 0;
+			sock_hold(sk);
+			sock_hold(child);
+			tcp_sk(child)->defer_tcp_accept.listen_sk = sk;
+			tcp_sk(child)->defer_tcp_accept.request = req;
+
+			inet_csk_reset_keepalive_timer(child,
+						       inet_csk(sk)->icsk_accept_queue.rskq_defer_accept * HZ);
+		} else {
+			inet_csk_reqsk_queue_add(sk, req, child);
+		}
+
 		return child;
 
 	listen_overflow:
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index 803d758a2b12..160d16f9f4fc 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -481,6 +481,11 @@ static void tcp_keepalive_timer (unsigned long data)
 		goto death;
 	}
 
+	if (tp->defer_tcp_accept.request && sk->sk_state == TCP_ESTABLISHED) {
+		tcp_send_active_reset(sk, GFP_ATOMIC);
+		goto death;
+	}
+
 	if (!sock_flag(sk, SOCK_KEEPOPEN) || sk->sk_state == TCP_CLOSE)
 		goto out;
 
-- 
cgit v1.2.3


From 03c086a747d0b242878eb881971ec61c1555869d Mon Sep 17 00:00:00 2001
From: Darren Salt <linux@youmustbejoking.demon.co.uk>
Date: Thu, 13 Mar 2008 15:35:49 +0000
Subject: PNP: increase the number of PnP memory resources from 12 to 24

Increase the number of PnP memory resources from 12 to 24.

This removes an "exceeded the max num of mem resources" warning on boot. I
also noticed the reservation of two more iomem ranges on the computer on
which this was tested.

Signed-off-by: Darren Salt <linux@youmustbejoking.demon.co.uk>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/pnp.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/pnp.h b/include/linux/pnp.h
index cd6332b88829..29dd55838e84 100644
--- a/include/linux/pnp.h
+++ b/include/linux/pnp.h
@@ -14,7 +14,7 @@
 #include <linux/mod_devicetable.h>
 
 #define PNP_MAX_PORT		40
-#define PNP_MAX_MEM		12
+#define PNP_MAX_MEM		24
 #define PNP_MAX_IRQ		2
 #define PNP_MAX_DMA		2
 #define PNP_NAME_LEN		50
-- 
cgit v1.2.3


From 0098b7273e968fb9989a6e1e4e4c024cd081fe0d Mon Sep 17 00:00:00 2001
From: Rusty Russell <rusty@rustcorp.com.au>
Date: Sat, 22 Mar 2008 17:18:47 -0700
Subject: [NET]: NPROTO is redundant; it's equal to AF_MAX/PF_MAX.

DaveM pointed out NPROTO exposed to userspace, so keep it around,
just make sure it stays in sync.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/net.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/net.h b/include/linux/net.h
index c414d90e647b..71f7dd559285 100644
--- a/include/linux/net.h
+++ b/include/linux/net.h
@@ -19,6 +19,7 @@
 #define _LINUX_NET_H
 
 #include <linux/wait.h>
+#include <linux/socket.h>
 #include <asm/socket.h>
 
 struct poll_table_struct;
@@ -26,7 +27,7 @@ struct pipe_inode_info;
 struct inode;
 struct net;
 
-#define NPROTO		34		/* should be enough for now..	*/
+#define NPROTO		AF_MAX
 
 #define SYS_SOCKET	1		/* sys_socket(2)		*/
 #define SYS_BIND	2		/* sys_bind(2)			*/
-- 
cgit v1.2.3


From 414f69d8a6ff0b30e7ea5ce10534b19f851e172e Mon Sep 17 00:00:00 2001
From: Joe Perches <joe@perches.com>
Date: Sun, 23 Mar 2008 22:04:31 -0700
Subject: [NET]: include/linux/atalk.h - remove duplicate include

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/atalk.h | 2 --
 1 file changed, 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/atalk.h b/include/linux/atalk.h
index ced8a1ed080c..e9ebac2e2ecc 100644
--- a/include/linux/atalk.h
+++ b/include/linux/atalk.h
@@ -85,8 +85,6 @@ static inline struct atalk_sock *at_sk(struct sock *sk)
 	return (struct atalk_sock *)sk;
 }
 
-#include <asm/byteorder.h>
-
 struct ddpehdr {
 	__be16	deh_len_hops;	/* lower 10 bits are length, next 4 - hops */
 	__be16	deh_sum;
-- 
cgit v1.2.3


From cc32e05416b4023a5466a2f66e3c02236a771c5b Mon Sep 17 00:00:00 2001
From: Joe Perches <joe@perches.com>
Date: Sun, 23 Mar 2008 22:05:44 -0700
Subject: [NET]: include/linux/igmp.h - remove duplicate include

Removed duplicate #include <linux/skbuff.h>
Combined #ifdef __KERNEL__ blocks

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/igmp.h | 38 +++++++++++++++++---------------------
 1 file changed, 17 insertions(+), 21 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/igmp.h b/include/linux/igmp.h
index f510e7e382a8..f5a1a0db2e8e 100644
--- a/include/linux/igmp.h
+++ b/include/linux/igmp.h
@@ -80,27 +80,6 @@ struct igmpv3_query {
 	__be32 srcs[0];
 };
 
-#ifdef __KERNEL__
-#include <linux/skbuff.h>
-
-static inline struct igmphdr *igmp_hdr(const struct sk_buff *skb)
-{
-	return (struct igmphdr *)skb_transport_header(skb);
-}
-
-static inline struct igmpv3_report *
-			igmpv3_report_hdr(const struct sk_buff *skb)
-{
-	return (struct igmpv3_report *)skb_transport_header(skb);
-}
-
-static inline struct igmpv3_query *
-			igmpv3_query_hdr(const struct sk_buff *skb)
-{
-	return (struct igmpv3_query *)skb_transport_header(skb);
-}
-#endif
-
 #define IGMP_HOST_MEMBERSHIP_QUERY	0x11	/* From RFC1112 */
 #define IGMP_HOST_MEMBERSHIP_REPORT	0x12	/* Ditto */
 #define IGMP_DVMRP			0x13	/* DVMRP routing */
@@ -151,6 +130,23 @@ static inline struct igmpv3_query *
 #include <linux/timer.h>
 #include <linux/in.h>
 
+static inline struct igmphdr *igmp_hdr(const struct sk_buff *skb)
+{
+	return (struct igmphdr *)skb_transport_header(skb);
+}
+
+static inline struct igmpv3_report *
+			igmpv3_report_hdr(const struct sk_buff *skb)
+{
+	return (struct igmpv3_report *)skb_transport_header(skb);
+}
+
+static inline struct igmpv3_query *
+			igmpv3_query_hdr(const struct sk_buff *skb)
+{
+	return (struct igmpv3_query *)skb_transport_header(skb);
+}
+
 extern int sysctl_igmp_max_memberships;
 extern int sysctl_igmp_max_msf;
 
-- 
cgit v1.2.3


From 310afe86af8ddd96a06b75aa61ef1af233f80e89 Mon Sep 17 00:00:00 2001
From: Joe Perches <joe@perches.com>
Date: Sun, 23 Mar 2008 22:06:51 -0700
Subject: [NET]: include/linux/udp.h - remove duplicate include

Remove duplicate #include <linux/types.h>
Combine #ifdef __KERNEL__ blocks

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/udp.h | 18 +++++++-----------
 1 file changed, 7 insertions(+), 11 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/udp.h b/include/linux/udp.h
index 1e7b7cb5703b..581ca2c14c52 100644
--- a/include/linux/udp.h
+++ b/include/linux/udp.h
@@ -26,15 +26,6 @@ struct udphdr {
 	__sum16	check;
 };
 
-#ifdef __KERNEL__
-#include <linux/skbuff.h>
-
-static inline struct udphdr *udp_hdr(const struct sk_buff *skb)
-{
-	return (struct udphdr *)skb_transport_header(skb);
-}
-#endif
-
 /* UDP socket options */
 #define UDP_CORK	1	/* Never send partially complete segments */
 #define UDP_ENCAP	100	/* Set the socket to accept encapsulated packets */
@@ -45,9 +36,14 @@ static inline struct udphdr *udp_hdr(const struct sk_buff *skb)
 #define UDP_ENCAP_L2TPINUDP	3 /* rfc2661 */
 
 #ifdef __KERNEL__
-#include <linux/types.h>
-
 #include <net/inet_sock.h>
+#include <linux/skbuff.h>
+
+static inline struct udphdr *udp_hdr(const struct sk_buff *skb)
+{
+	return (struct udphdr *)skb_transport_header(skb);
+}
+
 #define UDP_HTABLE_SIZE		128
 
 struct udp_sock {
-- 
cgit v1.2.3


From c8cdaf998df221b01134a051aba38c570105061b Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Mon, 10 Mar 2008 04:30:37 -0400
Subject: [IPV4,IPV6]: Share cork.rt between IPv4 and IPv6.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/ipv6.h    |  1 -
 include/net/inet_sock.h |  2 +-
 net/ipv4/ip_output.c    | 14 ++++++--------
 net/ipv6/ip6_output.c   | 12 ++++++------
 4 files changed, 13 insertions(+), 16 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 4aaefc349a4b..2102d8b67c01 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -315,7 +315,6 @@ struct ipv6_pinfo {
 	struct sk_buff		*pktoptions;
 	struct {
 		struct ipv6_txoptions *opt;
-		struct rt6_info	*rt;
 		int hop_limit;
 		int tclass;
 	} cork;
diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h
index b6db16d2766a..a42cd63d241a 100644
--- a/include/net/inet_sock.h
+++ b/include/net/inet_sock.h
@@ -136,7 +136,7 @@ struct inet_sock {
 		unsigned int		flags;
 		unsigned int		fragsize;
 		struct ip_options	*opt;
-		struct rtable		*rt;
+		struct dst_entry	*dst;
 		int			length; /* Total length of all frames */
 		__be32			addr;
 		struct flowi		fl;
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 349fae58c1a3..913266cd9902 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -825,7 +825,7 @@ int ip_append_data(struct sock *sk,
 		inet->cork.fragsize = mtu = inet->pmtudisc == IP_PMTUDISC_PROBE ?
 					    rt->u.dst.dev->mtu :
 					    dst_mtu(rt->u.dst.path);
-		inet->cork.rt = rt;
+		inet->cork.dst = &rt->u.dst;
 		inet->cork.length = 0;
 		sk->sk_sndmsg_page = NULL;
 		sk->sk_sndmsg_off = 0;
@@ -834,7 +834,7 @@ int ip_append_data(struct sock *sk,
 			transhdrlen += exthdrlen;
 		}
 	} else {
-		rt = inet->cork.rt;
+		rt = (struct rtable *)inet->cork.dst;
 		if (inet->cork.flags & IPCORK_OPT)
 			opt = inet->cork.opt;
 
@@ -1083,7 +1083,7 @@ ssize_t	ip_append_page(struct sock *sk, struct page *page,
 	if (skb_queue_empty(&sk->sk_write_queue))
 		return -EINVAL;
 
-	rt = inet->cork.rt;
+	rt = (struct rtable *)inet->cork.dst;
 	if (inet->cork.flags & IPCORK_OPT)
 		opt = inet->cork.opt;
 
@@ -1208,10 +1208,8 @@ static void ip_cork_release(struct inet_sock *inet)
 	inet->cork.flags &= ~IPCORK_OPT;
 	kfree(inet->cork.opt);
 	inet->cork.opt = NULL;
-	if (inet->cork.rt) {
-		ip_rt_put(inet->cork.rt);
-		inet->cork.rt = NULL;
-	}
+	dst_release(inet->cork.dst);
+	inet->cork.dst = NULL;
 }
 
 /*
@@ -1224,7 +1222,7 @@ int ip_push_pending_frames(struct sock *sk)
 	struct sk_buff **tail_skb;
 	struct inet_sock *inet = inet_sk(sk);
 	struct ip_options *opt = NULL;
-	struct rtable *rt = inet->cork.rt;
+	struct rtable *rt = (struct rtable *)inet->cork.dst;
 	struct iphdr *iph;
 	__be16 df = 0;
 	__u8 ttl;
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 98762fde2b65..ed6482667a25 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -1115,7 +1115,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to,
 			/* need source address above miyazawa*/
 		}
 		dst_hold(&rt->u.dst);
-		np->cork.rt = rt;
+		inet->cork.dst = &rt->u.dst;
 		inet->cork.fl = *fl;
 		np->cork.hop_limit = hlimit;
 		np->cork.tclass = tclass;
@@ -1136,7 +1136,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to,
 		length += exthdrlen;
 		transhdrlen += exthdrlen;
 	} else {
-		rt = np->cork.rt;
+		rt = (struct rt6_info *)inet->cork.dst;
 		fl = &inet->cork.fl;
 		if (inet->cork.flags & IPCORK_OPT)
 			opt = np->cork.opt;
@@ -1381,9 +1381,9 @@ static void ip6_cork_release(struct inet_sock *inet, struct ipv6_pinfo *np)
 	inet->cork.flags &= ~IPCORK_OPT;
 	kfree(np->cork.opt);
 	np->cork.opt = NULL;
-	if (np->cork.rt) {
-		dst_release(&np->cork.rt->u.dst);
-		np->cork.rt = NULL;
+	if (inet->cork.dst) {
+		dst_release(inet->cork.dst);
+		inet->cork.dst = NULL;
 		inet->cork.flags &= ~IPCORK_ALLFRAG;
 	}
 	memset(&inet->cork.fl, 0, sizeof(inet->cork.fl));
@@ -1398,7 +1398,7 @@ int ip6_push_pending_frames(struct sock *sk)
 	struct ipv6_pinfo *np = inet6_sk(sk);
 	struct ipv6hdr *hdr;
 	struct ipv6_txoptions *opt = np->cork.opt;
-	struct rt6_info *rt = np->cork.rt;
+	struct rt6_info *rt = (struct rt6_info *)inet->cork.dst;
 	struct flowi *fl = &inet->cork.fl;
 	unsigned char proto = fl->proto;
 	int err = 0;
-- 
cgit v1.2.3


From 4725474584d6aa2f07b3d47442dfbc4f6544f65e Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Mon, 10 Mar 2008 04:41:33 -0400
Subject: [IPV6]: Convert cork.hop_limit and cork.tclass into u8 instead of
 int.

Values of those fields are always between 0 and 255 (inclusive),
so use u8 and save some memory on 32bit systems.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/ipv6.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 2102d8b67c01..9b59e37afad9 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -315,8 +315,8 @@ struct ipv6_pinfo {
 	struct sk_buff		*pktoptions;
 	struct {
 		struct ipv6_txoptions *opt;
-		int hop_limit;
-		int tclass;
+		u8 hop_limit;
+		u8 tclass;
 	} cork;
 };
 
-- 
cgit v1.2.3


From 1d5d236d309ab90fa6aedf712f586b3595721373 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Mon, 10 Mar 2008 10:56:55 -0400
Subject: [IPV6]: Use bitfields for hop_limit and mcast_hops.

Save some bits for future extensions.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/ipv6.h | 28 ++++++++++++++++++++++++----
 1 file changed, 24 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 9b59e37afad9..87ae4e389ce1 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -274,8 +274,29 @@ struct ipv6_pinfo {
 
 	__be32			flow_label;
 	__u32			frag_size;
-	__s16			hop_limit;
-	__s16			mcast_hops;
+
+	/*
+	 * Packed in 16bits.
+	 * Omit one shift by by putting the signed field at MSB.
+	 */
+#if defined(__BIG_ENDIAN_BITFIELD)
+	__s16			hop_limit:9;
+	__u16			__unused_1:7;
+#else
+	__u16			__unused_1:7;
+	__s16			hop_limit:9;
+#endif
+
+#if defined(__BIG_ENDIAN_BITFIELD)
+	/* Packed in 16bits. */
+	__s16			mcast_hops:9;
+	__u16			__unused_2:6,
+				mc_loop:1;
+#else
+	__u16			mc_loop:1,
+				__unused_2:6;
+	__s16			mcast_hops:9;
+#endif
 	int			mcast_oif;
 
 	/* pktoption flags */
@@ -298,8 +319,7 @@ struct ipv6_pinfo {
 	} rxopt;
 
 	/* sockopt flags */
-	__u8			mc_loop:1,
-	                        recverr:1,
+	__u8			recverr:1,
 	                        sndflow:1,
 				pmtudisc:2,
 				ipv6only:1;
-- 
cgit v1.2.3


From 7cbca67c073263c179f605bdbbdc565ab29d801d Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Tue, 25 Mar 2008 09:37:42 +0900
Subject: [IPV6]: Support Source Address Selection API (RFC5014).

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/in6.h      | 11 +++++++
 include/linux/ipv6.h     |  6 +++-
 include/net/addrconf.h   |  1 +
 include/net/ip6_route.h  |  9 ++++--
 net/ipv6/addrconf.c      | 17 +++++++++--
 net/ipv6/fib6_rules.c    | 12 +++++++-
 net/ipv6/ip6_output.c    |  4 ++-
 net/ipv6/ipv6_sockglue.c | 78 ++++++++++++++++++++++++++++++++++++++++++++++++
 net/ipv6/ndisc.c         |  4 ++-
 net/ipv6/route.c         | 11 ++++++-
 net/ipv6/xfrm6_policy.c  |  2 +-
 net/sctp/ipv6.c          |  4 ++-
 12 files changed, 146 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/in6.h b/include/linux/in6.h
index 2a61c82af115..f674000c6c99 100644
--- a/include/linux/in6.h
+++ b/include/linux/in6.h
@@ -249,4 +249,15 @@ struct in6_flowlabel_req
  * IP6T_SO_GET_REVISION_TARGET	69
  */
 
+/* RFC5014: Source address selection */
+#define IPV6_ADDR_PREFERENCES	72
+
+#define IPV6_PREFER_SRC_TMP		0x0001
+#define IPV6_PREFER_SRC_PUBLIC		0x0002
+#define IPV6_PREFER_SRC_PUBTMP_DEFAULT	0x0100
+#define IPV6_PREFER_SRC_COA		0x0004
+#define IPV6_PREFER_SRC_HOME		0x0400
+#define IPV6_PREFER_SRC_CGA		0x0008
+#define IPV6_PREFER_SRC_NONCGA		0x0800
+
 #endif
diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 87ae4e389ce1..c9ba0da16ce9 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -322,7 +322,11 @@ struct ipv6_pinfo {
 	__u8			recverr:1,
 	                        sndflow:1,
 				pmtudisc:2,
-				ipv6only:1;
+				ipv6only:1,
+				srcprefs:3;	/* 001: prefer temporary address
+						 * 010: prefer public address
+						 * 100: prefer care-of address
+						 */
 	__u8			tclass;
 
 	__u32			dst_cookie;
diff --git a/include/net/addrconf.h b/include/net/addrconf.h
index edcb4bbaab7d..c9276c72764d 100644
--- a/include/net/addrconf.h
+++ b/include/net/addrconf.h
@@ -78,6 +78,7 @@ extern struct inet6_ifaddr      *ipv6_get_ifaddr(struct net *net,
 
 extern int			ipv6_dev_get_saddr(struct net_device *dev, 
 					       struct in6_addr *daddr,
+					       unsigned int srcprefs,
 					       struct in6_addr *saddr);
 extern int			ipv6_get_lladdr(struct net_device *dev,
 						struct in6_addr *addr,
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 5c3b67c86aef..3ae6799c2b14 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -30,9 +30,12 @@ struct route_info {
 #include <linux/ip.h>
 #include <linux/ipv6.h>
 
-#define RT6_LOOKUP_F_IFACE	0x1
-#define RT6_LOOKUP_F_REACHABLE	0x2
-#define RT6_LOOKUP_F_HAS_SADDR	0x4
+#define RT6_LOOKUP_F_IFACE		0x00000001
+#define RT6_LOOKUP_F_REACHABLE		0x00000002
+#define RT6_LOOKUP_F_HAS_SADDR		0x00000004
+#define RT6_LOOKUP_F_SRCPREF_TMP	0x00000008
+#define RT6_LOOKUP_F_SRCPREF_PUBLIC	0x00000010
+#define RT6_LOOKUP_F_SRCPREF_COA	0x00000020
 
 extern struct rt6_info	*ip6_null_entry;
 
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 787e90af166c..89954885dee1 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -909,6 +909,7 @@ struct ipv6_saddr_dst {
 	int ifindex;
 	int scope;
 	int label;
+	unsigned int prefs;
 };
 
 static inline int ipv6_saddr_preferred(int type)
@@ -984,9 +985,12 @@ static int ipv6_get_saddr_eval(struct ipv6_saddr_score *score,
 		break;
 #ifdef CONFIG_IPV6_MIP6
 	case IPV6_SADDR_RULE_HOA:
+	    {
 		/* Rule 4: Prefer home address */
-		ret = !!(score->ifa->flags & IFA_F_HOMEADDRESS);
+		int prefhome = !(dst->prefs & IPV6_PREFER_SRC_COA);
+		ret = !(score->ifa->flags & IFA_F_HOMEADDRESS) ^ prefhome;
 		break;
+	    }
 #endif
 	case IPV6_SADDR_RULE_OIF:
 		/* Rule 5: Prefer outgoing interface */
@@ -1000,11 +1004,16 @@ static int ipv6_get_saddr_eval(struct ipv6_saddr_score *score,
 		break;
 #ifdef CONFIG_IPV6_PRIVACY
 	case IPV6_SADDR_RULE_PRIVACY:
+	    {
 		/* Rule 7: Prefer public address
 		 * Note: prefer temprary address if use_tempaddr >= 2
 		 */
-		ret = (!(score->ifa->flags & IFA_F_TEMPORARY)) ^ (score->ifa->idev->cnf.use_tempaddr >= 2);
+		int preftmp = dst->prefs & (IPV6_PREFER_SRC_PUBLIC|IPV6_PREFER_SRC_TMP) ?
+				!!(dst->prefs & IPV6_PREFER_SRC_TMP) :
+				score->ifa->idev->cnf.use_tempaddr >= 2;
+		ret = (!(score->ifa->flags & IFA_F_TEMPORARY)) ^ preftmp;
 		break;
+	    }
 #endif
 	case IPV6_SADDR_RULE_ORCHID:
 		/* Rule 8-: Prefer ORCHID vs ORCHID or
@@ -1030,7 +1039,8 @@ out:
 }
 
 int ipv6_dev_get_saddr(struct net_device *dst_dev,
-		       struct in6_addr *daddr, struct in6_addr *saddr)
+		       struct in6_addr *daddr, unsigned int prefs,
+		       struct in6_addr *saddr)
 {
 	struct ipv6_saddr_score scores[2],
 				*score = &scores[0], *hiscore = &scores[1];
@@ -1044,6 +1054,7 @@ int ipv6_dev_get_saddr(struct net_device *dst_dev,
 	dst.ifindex = dst_dev ? dst_dev->ifindex : 0;
 	dst.scope = __ipv6_addr_src_scope(dst_type);
 	dst.label = ipv6_addr_label(daddr, dst_type, dst.ifindex);
+	dst.prefs = prefs;
 
 	hiscore->rule = -1;
 	hiscore->ifa = NULL;
diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c
index 55137408f054..e7a7fe26cebf 100644
--- a/net/ipv6/fib6_rules.c
+++ b/net/ipv6/fib6_rules.c
@@ -84,8 +84,18 @@ static int fib6_rule_action(struct fib_rule *rule, struct flowi *flp,
 		if ((rule->flags & FIB_RULE_FIND_SADDR) &&
 		    r->src.plen && !(flags & RT6_LOOKUP_F_HAS_SADDR)) {
 			struct in6_addr saddr;
+			unsigned int srcprefs = 0;
+
+			if (flags & RT6_LOOKUP_F_SRCPREF_TMP)
+				srcprefs |= IPV6_PREFER_SRC_TMP;
+			if (flags & RT6_LOOKUP_F_SRCPREF_PUBLIC)
+				srcprefs |= IPV6_PREFER_SRC_PUBLIC;
+			if (flags & RT6_LOOKUP_F_SRCPREF_COA)
+				srcprefs |= IPV6_PREFER_SRC_COA;
+
 			if (ipv6_dev_get_saddr(ip6_dst_idev(&rt->u.dst)->dev,
-					       &flp->fl6_dst, &saddr))
+					       &flp->fl6_dst, srcprefs,
+					       &saddr))
 				goto again;
 			if (!ipv6_prefix_equal(&saddr, &r->src.addr,
 					       r->src.plen))
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 2a4f08c8a02d..d34aa61353bb 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -920,7 +920,9 @@ static int ip6_dst_lookup_tail(struct sock *sk,
 
 	if (ipv6_addr_any(&fl->fl6_src)) {
 		err = ipv6_dev_get_saddr(ip6_dst_idev(*dst)->dev,
-					 &fl->fl6_dst, &fl->fl6_src);
+					 &fl->fl6_dst,
+					 sk ? inet6_sk(sk)->srcprefs : 0,
+					 &fl->fl6_src);
 		if (err)
 			goto out_err_release;
 	}
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 8e29fb1d1df6..dc6695cc5767 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -617,7 +617,67 @@ done:
 		retv = xfrm_user_policy(sk, optname, optval, optlen);
 		break;
 
+	case IPV6_ADDR_PREFERENCES:
+	    {
+		unsigned int pref = 0;
+		unsigned int prefmask = ~0;
+
+		retv = -EINVAL;
+
+		/* check PUBLIC/TMP/PUBTMP_DEFAULT conflicts */
+		switch (val & (IPV6_PREFER_SRC_PUBLIC|
+			       IPV6_PREFER_SRC_TMP|
+			       IPV6_PREFER_SRC_PUBTMP_DEFAULT)) {
+		case IPV6_PREFER_SRC_PUBLIC:
+			pref |= IPV6_PREFER_SRC_PUBLIC;
+			break;
+		case IPV6_PREFER_SRC_TMP:
+			pref |= IPV6_PREFER_SRC_TMP;
+			break;
+		case IPV6_PREFER_SRC_PUBTMP_DEFAULT:
+			break;
+		case 0:
+			goto pref_skip_pubtmp;
+		default:
+			goto e_inval;
+		}
+
+		prefmask &= ~(IPV6_PREFER_SRC_PUBLIC|
+			      IPV6_PREFER_SRC_TMP);
+pref_skip_pubtmp:
+
+		/* check HOME/COA conflicts */
+		switch (val & (IPV6_PREFER_SRC_HOME|IPV6_PREFER_SRC_COA)) {
+		case IPV6_PREFER_SRC_HOME:
+			break;
+		case IPV6_PREFER_SRC_COA:
+			pref |= IPV6_PREFER_SRC_COA;
+		case 0:
+			goto pref_skip_coa;
+		default:
+			goto e_inval;
+		}
+
+		prefmask &= ~IPV6_PREFER_SRC_COA;
+pref_skip_coa:
+
+		/* check CGA/NONCGA conflicts */
+		switch (val & (IPV6_PREFER_SRC_CGA|IPV6_PREFER_SRC_NONCGA)) {
+		case IPV6_PREFER_SRC_CGA:
+		case IPV6_PREFER_SRC_NONCGA:
+		case 0:
+			break;
+		default:
+			goto e_inval;
+		}
+
+		np->srcprefs = (np->srcprefs & prefmask) | pref;
+		retv = 0;
+
+		break;
+	    }
 	}
+
 	release_sock(sk);
 
 	return retv;
@@ -932,6 +992,24 @@ static int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
 		val = np->sndflow;
 		break;
 
+	case IPV6_ADDR_PREFERENCES:
+		val = 0;
+
+		if (np->srcprefs & IPV6_PREFER_SRC_TMP)
+			val |= IPV6_PREFER_SRC_TMP;
+		else if (np->srcprefs & IPV6_PREFER_SRC_PUBLIC)
+			val |= IPV6_PREFER_SRC_PUBLIC;
+		else {
+			/* XXX: should we return system default? */
+			val |= IPV6_PREFER_SRC_PUBTMP_DEFAULT;
+		}
+
+		if (np->srcprefs & IPV6_PREFER_SRC_COA)
+			val |= IPV6_PREFER_SRC_COA;
+		else
+			val |= IPV6_PREFER_SRC_HOME;
+		break;
+
 	default:
 		return -ENOPROTOOPT;
 	}
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index e7d8e74704c1..3f68a6eae7b2 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -546,7 +546,9 @@ static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,
 			override = 0;
 		in6_ifa_put(ifp);
 	} else {
-		if (ipv6_dev_get_saddr(dev, daddr, &tmpaddr))
+		if (ipv6_dev_get_saddr(dev, daddr,
+				       inet6_sk(dev->nd_net->ipv6.ndisc_sk)->srcprefs,
+				       &tmpaddr))
 			return;
 		src_addr = &tmpaddr;
 	}
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index aa3f08718e44..06faa46920e1 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -782,6 +782,15 @@ struct dst_entry * ip6_route_output(struct net *net, struct sock *sk,
 
 	if (!ipv6_addr_any(&fl->fl6_src))
 		flags |= RT6_LOOKUP_F_HAS_SADDR;
+	else if (sk) {
+		unsigned int prefs = inet6_sk(sk)->srcprefs;
+		if (prefs & IPV6_PREFER_SRC_TMP)
+			flags |= RT6_LOOKUP_F_SRCPREF_TMP;
+		if (prefs & IPV6_PREFER_SRC_PUBLIC)
+			flags |= RT6_LOOKUP_F_SRCPREF_PUBLIC;
+		if (prefs & IPV6_PREFER_SRC_COA)
+			flags |= RT6_LOOKUP_F_SRCPREF_COA;
+	}
 
 	return fib6_rule_lookup(net, fl, flags, ip6_pol_route_output);
 }
@@ -2162,7 +2171,7 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 	else if (dst) {
 		struct in6_addr saddr_buf;
 		if (ipv6_dev_get_saddr(ip6_dst_idev(&rt->u.dst)->dev,
-				       dst, &saddr_buf) == 0)
+				       dst, 0, &saddr_buf) == 0)
 			NLA_PUT(skb, RTA_PREFSRC, 16, &saddr_buf);
 	}
 
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index e96dafdc7032..d92d1fceb8cf 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -58,7 +58,7 @@ static int xfrm6_get_saddr(xfrm_address_t *saddr, xfrm_address_t *daddr)
 		return -EHOSTUNREACH;
 
 	ipv6_dev_get_saddr(ip6_dst_idev(dst)->dev,
-			   (struct in6_addr *)&daddr->a6,
+			   (struct in6_addr *)&daddr->a6, 0,
 			   (struct in6_addr *)&saddr->a6);
 	dst_release(dst);
 	return 0;
diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c
index 46c5b3c5cb99..dc71d0d83753 100644
--- a/net/sctp/ipv6.c
+++ b/net/sctp/ipv6.c
@@ -316,7 +316,9 @@ static void sctp_v6_get_saddr(struct sctp_association *asoc,
 
 	if (!asoc) {
 		ipv6_dev_get_saddr(dst ? ip6_dst_idev(dst)->dev : NULL,
-				   &daddr->v6.sin6_addr, &saddr->v6.sin6_addr);
+				   &daddr->v6.sin6_addr,
+				   inet6_sk(asoc->base.sk)->srcprefs,
+				   &saddr->v6.sin6_addr);
 		SCTP_DEBUG_PRINTK("saddr from ipv6_get_saddr: " NIP6_FMT "\n",
 				  NIP6(saddr->v6.sin6_addr));
 		return;
-- 
cgit v1.2.3


From aacda37538e7f9cf2148eedf3766239829e51ba4 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 18 Mar 2008 17:47:43 +0900
Subject: libata: implement ata_qc_raw_nbytes()

Implement ata_qc_raw_nbytes() which determines the raw user-requested
size of a PC command.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-scsi.c | 14 +++++++++++---
 include/linux/libata.h    |  8 +++++++-
 2 files changed, 18 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 8f0e8f2bc628..15795394b0a8 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -527,6 +527,14 @@ static struct ata_queued_cmd *ata_scsi_qc_new(struct ata_device *dev,
 	return qc;
 }
 
+static void ata_qc_set_pc_nbytes(struct ata_queued_cmd *qc)
+{
+	struct scsi_cmnd *scmd = qc->scsicmd;
+
+	qc->extrabytes = scmd->request->extra_len;
+	qc->nbytes = scsi_bufflen(scmd) + qc->extrabytes;
+}
+
 /**
  *	ata_dump_status - user friendly display of error info
  *	@id: id of the port in question
@@ -2539,7 +2547,7 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc)
 	}
 
 	qc->tf.command = ATA_CMD_PACKET;
-	qc->nbytes = scsi_bufflen(scmd) + scmd->request->extra_len;
+	ata_qc_set_pc_nbytes(qc);
 
 	/* check whether ATAPI DMA is safe */
 	if (!using_pio && ata_check_atapi_dma(qc))
@@ -2550,7 +2558,7 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc)
 	 * want to set it properly, and for DMA where it is
 	 * effectively meaningless.
 	 */
-	nbytes = min(scmd->request->data_len, (unsigned int)63 * 1024);
+	nbytes = min(ata_qc_raw_nbytes(qc), (unsigned int)63 * 1024);
 
 	/* Most ATAPI devices which honor transfer chunk size don't
 	 * behave according to the spec when odd chunk size which
@@ -2876,7 +2884,7 @@ static unsigned int ata_scsi_pass_thru(struct ata_queued_cmd *qc)
 	 * TODO: find out if we need to do more here to
 	 *       cover scatter/gather case.
 	 */
-	qc->nbytes = scsi_bufflen(scmd) + scmd->request->extra_len;
+	ata_qc_set_pc_nbytes(qc);
 
 	/* request result TF and be quiet about device error */
 	qc->flags |= ATA_QCFLAG_RESULT_TF | ATA_QCFLAG_QUIET;
diff --git a/include/linux/libata.h b/include/linux/libata.h
index a05f60013642..269cdba09578 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -463,6 +463,7 @@ struct ata_queued_cmd {
 	unsigned int		sect_size;
 
 	unsigned int		nbytes;
+	unsigned int		extrabytes;
 	unsigned int		curbytes;
 
 	struct scatterlist	*cursg;
@@ -1336,6 +1337,11 @@ static inline struct ata_queued_cmd *ata_qc_from_tag(struct ata_port *ap,
 	return NULL;
 }
 
+static inline unsigned int ata_qc_raw_nbytes(struct ata_queued_cmd *qc)
+{
+	return qc->nbytes - min(qc->extrabytes, qc->nbytes);
+}
+
 static inline void ata_tf_init(struct ata_device *dev, struct ata_taskfile *tf)
 {
 	memset(tf, 0, sizeof(*tf));
@@ -1354,7 +1360,7 @@ static inline void ata_qc_reinit(struct ata_queued_cmd *qc)
 	qc->flags = 0;
 	qc->cursg = NULL;
 	qc->cursg_ofs = 0;
-	qc->nbytes = qc->curbytes = 0;
+	qc->nbytes = qc->extrabytes = qc->curbytes = 0;
 	qc->n_elem = 0;
 	qc->err_mask = 0;
 	qc->sect_size = ATA_SECT_SIZE;
-- 
cgit v1.2.3


From 392e1d9817d0024c96aae237c3c4349e47c976fd Mon Sep 17 00:00:00 2001
From: Alan Stern <stern@rowland.harvard.edu>
Date: Tue, 11 Mar 2008 10:20:12 -0400
Subject: USB: new quirk flag to avoid Set-Interface

This patch (as1057) fixes a problem with the X-Rite/Gretag-Macbeth
Eye-One Pro display colorimeter; the device crashes when it receives a
Set-Interface request.  A new quirk (USB_QUIRK_NO_SET_INTF) is
introduced and a quirks entry is created for this device.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/usb/core/message.c | 5 ++++-
 drivers/usb/core/quirks.c  | 3 +++
 include/linux/usb/quirks.h | 3 +++
 3 files changed, 10 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/usb/core/message.c b/drivers/usb/core/message.c
index fefb92296e8f..c311f67b7f08 100644
--- a/drivers/usb/core/message.c
+++ b/drivers/usb/core/message.c
@@ -1206,7 +1206,10 @@ int usb_set_interface(struct usb_device *dev, int interface, int alternate)
 		return -EINVAL;
 	}
 
-	ret = usb_control_msg(dev, usb_sndctrlpipe(dev, 0),
+	if (dev->quirks & USB_QUIRK_NO_SET_INTF)
+		ret = -EPIPE;
+	else
+		ret = usb_control_msg(dev, usb_sndctrlpipe(dev, 0),
 				   USB_REQ_SET_INTERFACE, USB_RECIP_INTERFACE,
 				   alternate, interface, NULL, 0, 5000);
 
diff --git a/drivers/usb/core/quirks.c b/drivers/usb/core/quirks.c
index d9d1eb19f2a1..dfc5418ea10c 100644
--- a/drivers/usb/core/quirks.c
+++ b/drivers/usb/core/quirks.c
@@ -50,6 +50,9 @@ static const struct usb_device_id usb_quirk_list[] = {
 	/* M-Systems Flash Disk Pioneers */
 	{ USB_DEVICE(0x08ec, 0x1000), .driver_info = USB_QUIRK_RESET_RESUME },
 
+	/* X-Rite/Gretag-Macbeth Eye-One Pro display colorimeter */
+	{ USB_DEVICE(0x0971, 0x2000), .driver_info = USB_QUIRK_NO_SET_INTF },
+
 	/* Action Semiconductor flash disk */
 	{ USB_DEVICE(0x10d6, 0x2200), .driver_info =
 			USB_QUIRK_STRING_FETCH_255 },
diff --git a/include/linux/usb/quirks.h b/include/linux/usb/quirks.h
index 2692ec9389ca..1f999ec8d08c 100644
--- a/include/linux/usb/quirks.h
+++ b/include/linux/usb/quirks.h
@@ -9,3 +9,6 @@
 
 /* device can't resume correctly so reset it instead */
 #define USB_QUIRK_RESET_RESUME		0x00000002
+
+/* device can't handle Set-Interface requests */
+#define USB_QUIRK_NO_SET_INTF		0x00000004
-- 
cgit v1.2.3


From cc36bdd47ae51b66780b317c1fa519221f894405 Mon Sep 17 00:00:00 2001
From: Constantin Baranov <const@tltsu.ru>
Date: Sun, 16 Mar 2008 20:04:23 +0000
Subject: USB: add support for Motorola ROKR Z6 cellphone in mass storage mode

Motorola ROKR Z6 cellphone has bugs in its USB, so it is impossible to use
it as mass storage. Patch describes new "unusual" USB device for it with
FIX_INQUIRY and FIX_CAPACITY flags and new BULK_IGNORE_TAG flag.
Last flag relaxes check for equality of bcs->Tag and us->tag in
usb_stor_Bulk_transport routine.

Signed-off-by: Constantin Baranov <const@tltsu.ru>
Signed-off-by: Matthew Dharm <mdharm-usb@one-eyed-alien.net>
Signed-off-by: Daniel Drake <dsd@gentoo.org>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/usb/storage/transport.c    |  3 ++-
 drivers/usb/storage/unusual_devs.h | 11 +++++++++++
 include/linux/usb_usual.h          |  4 +++-
 3 files changed, 16 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/usb/storage/transport.c b/drivers/usb/storage/transport.c
index 5780ed15f1ad..bdd4334bed5a 100644
--- a/drivers/usb/storage/transport.c
+++ b/drivers/usb/storage/transport.c
@@ -1009,7 +1009,8 @@ int usb_stor_Bulk_transport(struct scsi_cmnd *srb, struct us_data *us)
 	US_DEBUGP("Bulk Status S 0x%x T 0x%x R %u Stat 0x%x\n",
 			le32_to_cpu(bcs->Signature), bcs->Tag, 
 			residue, bcs->Status);
-	if (bcs->Tag != us->tag || bcs->Status > US_BULK_STAT_PHASE) {
+	if (!(bcs->Tag == us->tag || (us->flags & US_FL_BULK_IGNORE_TAG)) ||
+		bcs->Status > US_BULK_STAT_PHASE) {
 		US_DEBUGP("Bulk logical error\n");
 		return USB_STOR_TRANSPORT_ERROR;
 	}
diff --git a/drivers/usb/storage/unusual_devs.h b/drivers/usb/storage/unusual_devs.h
index 99679a8cfa02..e5219a56947c 100644
--- a/drivers/usb/storage/unusual_devs.h
+++ b/drivers/usb/storage/unusual_devs.h
@@ -1589,6 +1589,17 @@ UNUSUAL_DEV(  0x22b8, 0x4810, 0x0001, 0x0001,
 		US_SC_DEVICE, US_PR_DEVICE, NULL,
 		US_FL_FIX_CAPACITY),
 
+/*
+ * Patch by Constantin Baranov <const@tltsu.ru>
+ * Report by Andreas Koenecke.
+ * Motorola ROKR Z6.
+ */
+UNUSUAL_DEV(  0x22b8, 0x6426, 0x0101, 0x0101,
+		"Motorola",
+		"MSnc.",
+		US_SC_DEVICE, US_PR_DEVICE, NULL,
+		US_FL_FIX_INQUIRY | US_FL_FIX_CAPACITY | US_FL_BULK_IGNORE_TAG),
+
 /* Reported by Radovan Garabik <garabik@kassiopeia.juls.savba.sk> */
 UNUSUAL_DEV(  0x2735, 0x100b, 0x0000, 0x9999,
 		"MPIO",
diff --git a/include/linux/usb_usual.h b/include/linux/usb_usual.h
index cee0623b3c7b..0a40dfa44c9f 100644
--- a/include/linux/usb_usual.h
+++ b/include/linux/usb_usual.h
@@ -50,7 +50,9 @@
 	US_FLAG(CAPACITY_HEURISTICS,	0x00001000)		\
 		/* sometimes sizes is too big */		\
 	US_FLAG(MAX_SECTORS_MIN,0x00002000)			\
-		/* Sets max_sectors to arch min */
+		/* Sets max_sectors to arch min */		\
+	US_FLAG(BULK_IGNORE_TAG,0x00004000)			\
+		/* Ignore tag mismatch in bulk operations */
 
 
 #define US_FLAG(name, value)	US_FL_##name = value ,
-- 
cgit v1.2.3


From 49741c4d01554c2630cea02cfdf236b17062a912 Mon Sep 17 00:00:00 2001
From: Andrew Morton <akpm@linux-foundation.org>
Date: Mon, 17 Mar 2008 14:21:18 -0700
Subject: PCI: revert "pcie: utilize pcie transaction pending bit"

Revert as it is reported to cause problems for people.

commit 4348a2dc49f9baecd34a9b0904245488c6189398
Author: Shaohua Li <shaohua.li@intel.com>
Date:   Wed Oct 24 10:45:08 2007 +0800

    pcie: utilize pcie transaction pending bit

    PCIE has a mechanism to wait for Non-Posted request to complete. I think
    pci_disable_device is a good place to do this.

    Signed-off-by: Shaohua Li <shaohua.li@intel.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Due to the regression reported at
http://bugzilla.kernel.org/show_bug.cgi?id=10065

Cc: Shaohua Li <shaohua.li@intel.com>
Cc: Soeren Sonnenburg <kernel@nn7.de>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/pci.c   | 21 ---------------------
 include/linux/pci.h |  4 ----
 2 files changed, 25 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index 183fddaa38b7..a4445b7210bf 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -314,24 +314,6 @@ int pci_find_ht_capability(struct pci_dev *dev, int ht_cap)
 }
 EXPORT_SYMBOL_GPL(pci_find_ht_capability);
 
-void pcie_wait_pending_transaction(struct pci_dev *dev)
-{
-	int pos;
-	u16 reg16;
-
-	pos = pci_find_capability(dev, PCI_CAP_ID_EXP);
-	if (!pos)
-		return;
-	while (1) {
-		pci_read_config_word(dev, pos + PCI_EXP_DEVSTA, &reg16);
-		if (!(reg16 & PCI_EXP_DEVSTA_TRPND))
-			break;
-		cpu_relax();
-	}
-
-}
-EXPORT_SYMBOL_GPL(pcie_wait_pending_transaction);
-
 /**
  * pci_find_parent_resource - return resource region of parent bus of given region
  * @dev: PCI device structure contains resources to be searched
@@ -936,9 +918,6 @@ pci_disable_device(struct pci_dev *dev)
 	if (atomic_sub_return(1, &dev->enable_cnt) != 0)
 		return;
 
-	/* Wait for all transactions are finished before disabling the device */
-	pcie_wait_pending_transaction(dev);
-
 	pci_read_config_word(dev, PCI_COMMAND, &pci_command);
 	if (pci_command & PCI_COMMAND_MASTER) {
 		pci_command &= ~PCI_COMMAND_MASTER;
diff --git a/include/linux/pci.h b/include/linux/pci.h
index b7e4b633c69b..ea760e519c46 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -513,7 +513,6 @@ int pci_find_next_capability(struct pci_dev *dev, u8 pos, int cap);
 int pci_find_ext_capability(struct pci_dev *dev, int cap);
 int pci_find_ht_capability(struct pci_dev *dev, int ht_cap);
 int pci_find_next_ht_capability(struct pci_dev *dev, int pos, int ht_cap);
-void pcie_wait_pending_transaction(struct pci_dev *dev);
 struct pci_bus *pci_find_next_bus(const struct pci_bus *from);
 
 struct pci_dev *pci_get_device(unsigned int vendor, unsigned int device,
@@ -884,9 +883,6 @@ static inline int pci_find_ext_capability(struct pci_dev *dev, int cap)
 	return 0;
 }
 
-static inline void pcie_wait_pending_transaction(struct pci_dev *dev)
-{ }
-
 /* Power management related routines */
 static inline int pci_save_state(struct pci_dev *dev)
 {
-- 
cgit v1.2.3


From c346dca10840a874240c78efe3f39acf4312a1f2 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Tue, 25 Mar 2008 21:47:49 +0900
Subject: [NET] NETNS: Omit net_device->nd_net without CONFIG_NET_NS.

Introduce per-net_device inlines: dev_net(), dev_net_set().
Without CONFIG_NET_NS, no namespace other than &init_net exists.
Let's explicitly define them to help compiler optimizations.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 arch/ia64/hp/sim/simeth.c           |  2 +-
 drivers/block/aoe/aoenet.c          |  2 +-
 drivers/net/bonding/bond_3ad.c      |  2 +-
 drivers/net/bonding/bond_alb.c      |  2 +-
 drivers/net/bonding/bond_main.c     |  6 ++---
 drivers/net/hamradio/bpqether.c     |  4 ++--
 drivers/net/loopback.c              |  2 +-
 drivers/net/macvlan.c               |  2 +-
 drivers/net/pppoe.c                 |  6 ++---
 drivers/net/veth.c                  |  2 +-
 drivers/net/via-velocity.c          |  2 +-
 drivers/net/wan/dlci.c              |  2 +-
 drivers/net/wan/hdlc.c              |  4 ++--
 drivers/net/wan/lapbether.c         |  4 ++--
 drivers/net/wan/syncppp.c           |  2 +-
 drivers/s390/net/qeth_l3_main.c     |  2 +-
 include/linux/inetdevice.h          |  6 ++---
 include/linux/netdevice.h           | 25 ++++++++++++++++++++-
 net/8021q/vlan.c                    |  2 +-
 net/8021q/vlan_dev.c                |  2 +-
 net/appletalk/aarp.c                |  4 ++--
 net/appletalk/ddp.c                 |  6 ++---
 net/atm/clip.c                      |  2 +-
 net/atm/mpc.c                       |  2 +-
 net/ax25/af_ax25.c                  |  2 +-
 net/ax25/ax25_in.c                  |  2 +-
 net/bridge/br_notify.c              |  2 +-
 net/bridge/br_stp_bpdu.c            |  2 +-
 net/can/af_can.c                    |  4 ++--
 net/can/bcm.c                       |  2 +-
 net/can/raw.c                       |  2 +-
 net/core/dev.c                      | 22 +++++++++----------
 net/core/dst.c                      |  2 +-
 net/core/fib_rules.c                |  2 +-
 net/core/neighbour.c                | 12 +++++-----
 net/core/pktgen.c                   |  2 +-
 net/core/rtnetlink.c                |  4 ++--
 net/decnet/af_decnet.c              |  2 +-
 net/decnet/dn_route.c               |  2 +-
 net/econet/af_econet.c              |  4 ++--
 net/ipv4/arp.c                      | 14 ++++++------
 net/ipv4/devinet.c                  | 10 ++++-----
 net/ipv4/fib_frontend.c             | 12 +++++-----
 net/ipv4/icmp.c                     |  8 +++----
 net/ipv4/igmp.c                     | 16 +++++++-------
 net/ipv4/ip_fragment.c              |  2 +-
 net/ipv4/ip_gre.c                   |  2 +-
 net/ipv4/ip_input.c                 |  6 ++---
 net/ipv4/ip_options.c               |  2 +-
 net/ipv4/ipconfig.c                 |  4 ++--
 net/ipv4/ipmr.c                     |  2 +-
 net/ipv4/netfilter/ip_queue.c       |  2 +-
 net/ipv4/netfilter/ipt_MASQUERADE.c |  2 +-
 net/ipv4/raw.c                      |  4 ++--
 net/ipv4/route.c                    | 28 +++++++++++------------
 net/ipv4/tcp_ipv4.c                 |  6 ++---
 net/ipv4/udp.c                      |  4 ++--
 net/ipv4/xfrm4_policy.c             |  2 +-
 net/ipv6/addrconf.c                 | 44 ++++++++++++++++++-------------------
 net/ipv6/icmp.c                     |  4 ++--
 net/ipv6/ip6_output.c               |  2 +-
 net/ipv6/mcast.c                    |  6 ++---
 net/ipv6/ndisc.c                    | 24 ++++++++++----------
 net/ipv6/netfilter/ip6_queue.c      |  2 +-
 net/ipv6/proc.c                     |  2 +-
 net/ipv6/raw.c                      |  4 ++--
 net/ipv6/reassembly.c               |  2 +-
 net/ipv6/route.c                    | 40 ++++++++++++++++-----------------
 net/ipv6/tcp_ipv6.c                 | 10 ++++-----
 net/ipv6/udp.c                      |  4 ++--
 net/ipv6/xfrm6_policy.c             |  2 +-
 net/ipx/af_ipx.c                    |  4 ++--
 net/irda/irlap_frame.c              |  2 +-
 net/llc/llc_input.c                 |  2 +-
 net/netfilter/core.c                |  2 +-
 net/netfilter/nfnetlink_queue.c     |  2 +-
 net/netlabel/netlabel_unlabeled.c   |  2 +-
 net/netrom/af_netrom.c              |  2 +-
 net/packet/af_packet.c              |  8 +++----
 net/rose/af_rose.c                  |  2 +-
 net/sctp/protocol.c                 |  2 +-
 net/tipc/eth_media.c                |  4 ++--
 net/wireless/wext.c                 |  2 +-
 net/x25/af_x25.c                    |  2 +-
 net/x25/x25_dev.c                   |  2 +-
 net/xfrm/xfrm_policy.c              |  4 ++--
 security/selinux/netif.c            |  2 +-
 87 files changed, 251 insertions(+), 228 deletions(-)

(limited to 'include/linux')

diff --git a/arch/ia64/hp/sim/simeth.c b/arch/ia64/hp/sim/simeth.c
index 969fe9f443c4..3d47839a0c48 100644
--- a/arch/ia64/hp/sim/simeth.c
+++ b/arch/ia64/hp/sim/simeth.c
@@ -294,7 +294,7 @@ simeth_device_event(struct notifier_block *this,unsigned long event, void *ptr)
 		return NOTIFY_DONE;
 	}
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if ( event != NETDEV_UP && event != NETDEV_DOWN ) return NOTIFY_DONE;
diff --git a/drivers/block/aoe/aoenet.c b/drivers/block/aoe/aoenet.c
index 8460ef736d56..18d243c73eee 100644
--- a/drivers/block/aoe/aoenet.c
+++ b/drivers/block/aoe/aoenet.c
@@ -115,7 +115,7 @@ aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt,
 	struct aoe_hdr *h;
 	u32 n;
 
-	if (ifp->nd_net != &init_net)
+	if (dev_net(ifp) != &init_net)
 		goto exit;
 
 	skb = skb_share_check(skb, GFP_ATOMIC);
diff --git a/drivers/net/bonding/bond_3ad.c b/drivers/net/bonding/bond_3ad.c
index cb3c6faa7888..457d81f73e39 100644
--- a/drivers/net/bonding/bond_3ad.c
+++ b/drivers/net/bonding/bond_3ad.c
@@ -2429,7 +2429,7 @@ int bond_3ad_lacpdu_recv(struct sk_buff *skb, struct net_device *dev, struct pac
 	struct slave *slave = NULL;
 	int ret = NET_RX_DROP;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto out;
 
 	if (!(dev->flags & IFF_MASTER))
diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
index b57bc9467dbe..b986dacf5d33 100644
--- a/drivers/net/bonding/bond_alb.c
+++ b/drivers/net/bonding/bond_alb.c
@@ -345,7 +345,7 @@ static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct
 	struct arp_pkt *arp = (struct arp_pkt *)skb->data;
 	int res = NET_RX_DROP;
 
-	if (bond_dev->nd_net != &init_net)
+	if (dev_net(bond_dev) != &init_net)
 		goto out;
 
 	if (!(bond_dev->flags & IFF_MASTER))
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 5fc9d8d58ece..ac688fcb27d7 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -2629,7 +2629,7 @@ static int bond_arp_rcv(struct sk_buff *skb, struct net_device *dev, struct pack
 	unsigned char *arp_ptr;
 	__be32 sip, tip;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto out;
 
 	if (!(dev->priv_flags & IFF_BONDING) || !(dev->flags & IFF_MASTER))
@@ -3470,7 +3470,7 @@ static int bond_netdev_event(struct notifier_block *this, unsigned long event, v
 {
 	struct net_device *event_dev = (struct net_device *)ptr;
 
-	if (event_dev->nd_net != &init_net)
+	if (dev_net(event_dev) != &init_net)
 		return NOTIFY_DONE;
 
 	dprintk("event_dev: %s, event: %lx\n",
@@ -3508,7 +3508,7 @@ static int bond_inetaddr_event(struct notifier_block *this, unsigned long event,
 	struct bonding *bond, *bond_next;
 	struct vlan_entry *vlan, *vlan_next;
 
-	if (ifa->ifa_dev->dev->nd_net != &init_net)
+	if (dev_net(ifa->ifa_dev->dev) != &init_net)
 		return NOTIFY_DONE;
 
 	list_for_each_entry_safe(bond, bond_next, &bond_dev_list, bond_list) {
diff --git a/drivers/net/hamradio/bpqether.c b/drivers/net/hamradio/bpqether.c
index 5ddf8b0c34f9..5f4b4c6c9f76 100644
--- a/drivers/net/hamradio/bpqether.c
+++ b/drivers/net/hamradio/bpqether.c
@@ -172,7 +172,7 @@ static int bpq_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_ty
 	struct ethhdr *eth;
 	struct bpqdev *bpq;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
@@ -553,7 +553,7 @@ static int bpq_device_event(struct notifier_block *this,unsigned long event, voi
 {
 	struct net_device *dev = (struct net_device *)ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (!dev_is_ethdev(dev))
diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c
index f2a6e7132241..41b774baac4d 100644
--- a/drivers/net/loopback.c
+++ b/drivers/net/loopback.c
@@ -258,7 +258,7 @@ static __net_init int loopback_net_init(struct net *net)
 	if (!dev)
 		goto out;
 
-	dev->nd_net = net;
+	dev_net_set(dev, net);
 	err = register_netdev(dev);
 	if (err)
 		goto out_free_netdev;
diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c
index f651a816b280..2056cfc624dc 100644
--- a/drivers/net/macvlan.c
+++ b/drivers/net/macvlan.c
@@ -402,7 +402,7 @@ static int macvlan_newlink(struct net_device *dev,
 	if (!tb[IFLA_LINK])
 		return -EINVAL;
 
-	lowerdev = __dev_get_by_index(dev->nd_net, nla_get_u32(tb[IFLA_LINK]));
+	lowerdev = __dev_get_by_index(dev_net(dev), nla_get_u32(tb[IFLA_LINK]));
 	if (lowerdev == NULL)
 		return -ENODEV;
 
diff --git a/drivers/net/pppoe.c b/drivers/net/pppoe.c
index ac0ac98b19cd..4fad4ddb3504 100644
--- a/drivers/net/pppoe.c
+++ b/drivers/net/pppoe.c
@@ -301,7 +301,7 @@ static int pppoe_device_event(struct notifier_block *this,
 {
 	struct net_device *dev = (struct net_device *) ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* Only look at sockets that are using this specific device. */
@@ -392,7 +392,7 @@ static int pppoe_rcv(struct sk_buff *skb,
 	if (!(skb = skb_share_check(skb, GFP_ATOMIC)))
 		goto out;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	if (!pskb_may_pull(skb, sizeof(struct pppoe_hdr)))
@@ -424,7 +424,7 @@ static int pppoe_disc_rcv(struct sk_buff *skb,
 	struct pppoe_hdr *ph;
 	struct pppox_sock *po;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto abort;
 
 	if (!pskb_may_pull(skb, sizeof(struct pppoe_hdr)))
diff --git a/drivers/net/veth.c b/drivers/net/veth.c
index e2ad98bee6e7..31cd817f33f9 100644
--- a/drivers/net/veth.c
+++ b/drivers/net/veth.c
@@ -375,7 +375,7 @@ static int veth_newlink(struct net_device *dev,
 	else
 		snprintf(ifname, IFNAMSIZ, DRV_NAME "%%d");
 
-	peer = rtnl_create_link(dev->nd_net, ifname, &veth_link_ops, tbp);
+	peer = rtnl_create_link(dev_net(dev), ifname, &veth_link_ops, tbp);
 	if (IS_ERR(peer))
 		return PTR_ERR(peer);
 
diff --git a/drivers/net/via-velocity.c b/drivers/net/via-velocity.c
index 1525e8a89844..ed1afaf683a4 100644
--- a/drivers/net/via-velocity.c
+++ b/drivers/net/via-velocity.c
@@ -3464,7 +3464,7 @@ static int velocity_netdev_event(struct notifier_block *nb, unsigned long notifi
 	struct velocity_info *vptr;
 	unsigned long flags;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	spin_lock_irqsave(&velocity_dev_list_lock, flags);
diff --git a/drivers/net/wan/dlci.c b/drivers/net/wan/dlci.c
index 96b232446c0b..b14242768fad 100644
--- a/drivers/net/wan/dlci.c
+++ b/drivers/net/wan/dlci.c
@@ -517,7 +517,7 @@ static int dlci_dev_event(struct notifier_block *unused,
 {
 	struct net_device *dev = (struct net_device *) ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event == NETDEV_UNREGISTER) {
diff --git a/drivers/net/wan/hdlc.c b/drivers/net/wan/hdlc.c
index 39951d0c34d6..9a83c9d5b8cf 100644
--- a/drivers/net/wan/hdlc.c
+++ b/drivers/net/wan/hdlc.c
@@ -68,7 +68,7 @@ static int hdlc_rcv(struct sk_buff *skb, struct net_device *dev,
 {
 	struct hdlc_device *hdlc = dev_to_hdlc(dev);
 
-	if (dev->nd_net != &init_net) {
+	if (dev_net(dev) != &init_net) {
 		kfree_skb(skb);
 		return 0;
 	}
@@ -105,7 +105,7 @@ static int hdlc_device_event(struct notifier_block *this, unsigned long event,
 	unsigned long flags;
 	int on;
  
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (dev->get_stats != hdlc_get_stats)
diff --git a/drivers/net/wan/lapbether.c b/drivers/net/wan/lapbether.c
index fb37b8095231..629c909e05f9 100644
--- a/drivers/net/wan/lapbether.c
+++ b/drivers/net/wan/lapbether.c
@@ -91,7 +91,7 @@ static int lapbeth_rcv(struct sk_buff *skb, struct net_device *dev, struct packe
 	int len, err;
 	struct lapbethdev *lapbeth;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
@@ -393,7 +393,7 @@ static int lapbeth_device_event(struct notifier_block *this,
 	struct lapbethdev *lapbeth;
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (!dev_is_ethdev(dev))
diff --git a/drivers/net/wan/syncppp.c b/drivers/net/wan/syncppp.c
index 61e24b7a45a3..29b4b94e4947 100644
--- a/drivers/net/wan/syncppp.c
+++ b/drivers/net/wan/syncppp.c
@@ -1444,7 +1444,7 @@ static void sppp_print_bytes (u_char *p, u16 len)
 
 static int sppp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *p, struct net_device *orig_dev)
 {
-	if (dev->nd_net != &init_net) {
+	if (dev_net(dev) != &init_net) {
 		kfree_skb(skb);
 		return 0;
 	}
diff --git a/drivers/s390/net/qeth_l3_main.c b/drivers/s390/net/qeth_l3_main.c
index a856cb47fc78..21c439046b3c 100644
--- a/drivers/s390/net/qeth_l3_main.c
+++ b/drivers/s390/net/qeth_l3_main.c
@@ -3250,7 +3250,7 @@ static int qeth_l3_ip_event(struct notifier_block *this,
 	struct qeth_ipaddr *addr;
 	struct qeth_card *card;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	QETH_DBF_TEXT(trace, 3, "ipevent");
diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h
index da05ab47ff2f..7009b0cdd06f 100644
--- a/include/linux/inetdevice.h
+++ b/include/linux/inetdevice.h
@@ -70,13 +70,13 @@ static inline void ipv4_devconf_setall(struct in_device *in_dev)
 	ipv4_devconf_set((in_dev), NET_IPV4_CONF_ ## attr, (val))
 
 #define IN_DEV_ANDCONF(in_dev, attr) \
-	(IPV4_DEVCONF_ALL(in_dev->dev->nd_net, attr) && \
+	(IPV4_DEVCONF_ALL(dev_net(in_dev->dev), attr) && \
 	 IN_DEV_CONF_GET((in_dev), attr))
 #define IN_DEV_ORCONF(in_dev, attr) \
-	(IPV4_DEVCONF_ALL(in_dev->dev->nd_net, attr) || \
+	(IPV4_DEVCONF_ALL(dev_net(in_dev->dev), attr) || \
 	 IN_DEV_CONF_GET((in_dev), attr))
 #define IN_DEV_MAXCONF(in_dev, attr) \
-	(max(IPV4_DEVCONF_ALL(in_dev->dev->nd_net, attr), \
+	(max(IPV4_DEVCONF_ALL(dev_net(in_dev->dev), attr), \
 	     IN_DEV_CONF_GET((in_dev), attr)))
 
 #define IN_DEV_FORWARD(in_dev)		IN_DEV_CONF_GET((in_dev), FORWARDING)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index ced61f87660e..d146be40f46c 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -708,8 +708,10 @@ struct net_device
 	void                    (*poll_controller)(struct net_device *dev);
 #endif
 
+#ifdef CONFIG_NET_NS
 	/* Network namespace this network device is inside */
 	struct net		*nd_net;
+#endif
 
 	/* bridge stuff */
 	struct net_bridge_port	*br_port;
@@ -737,6 +739,27 @@ struct net_device
 #define	NETDEV_ALIGN		32
 #define	NETDEV_ALIGN_CONST	(NETDEV_ALIGN - 1)
 
+/*
+ * Net namespace inlines
+ */
+static inline
+struct net *dev_net(const struct net_device *dev)
+{
+#ifdef CONFIG_NET_NS
+	return dev->nd_net;
+#else
+	return &init_net;
+#endif
+}
+
+static inline
+void dev_net_set(struct net_device *dev, const struct net *net)
+{
+#ifdef CONFIG_NET_NS
+	dev->nd_dev = net;
+#endif
+}
+
 /**
  *	netdev_priv - access network device private data
  *	@dev: network device
@@ -813,7 +836,7 @@ static inline struct net_device *next_net_device(struct net_device *dev)
 	struct list_head *lh;
 	struct net *net;
 
-	net = dev->nd_net;
+	net = dev_net(dev);
 	lh = dev->dev_list.next;
 	return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
 }
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index dbc81b965096..c35dc230365c 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -382,7 +382,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
 	int i, flgs;
 	struct net_device *vlandev;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (!grp)
diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c
index 1e5c9904571d..e536162b1ebc 100644
--- a/net/8021q/vlan_dev.c
+++ b/net/8021q/vlan_dev.c
@@ -153,7 +153,7 @@ int vlan_skb_recv(struct sk_buff *skb, struct net_device *dev,
 	struct net_device_stats *stats;
 	unsigned short vlan_TCI;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto err_free;
 
 	skb = skb_share_check(skb, GFP_ATOMIC);
diff --git a/net/appletalk/aarp.c b/net/appletalk/aarp.c
index 61166f66479f..25aa37ce9430 100644
--- a/net/appletalk/aarp.c
+++ b/net/appletalk/aarp.c
@@ -333,7 +333,7 @@ static int aarp_device_event(struct notifier_block *this, unsigned long event,
 	struct net_device *dev = ptr;
 	int ct;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event == NETDEV_DOWN) {
@@ -716,7 +716,7 @@ static int aarp_rcv(struct sk_buff *skb, struct net_device *dev,
 	struct atalk_addr sa, *ma, da;
 	struct atalk_iface *ifa;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto out0;
 
 	/* We only do Ethernet SNAP AARP. */
diff --git a/net/appletalk/ddp.c b/net/appletalk/ddp.c
index 3be55c8ca4ef..44cd42f7786b 100644
--- a/net/appletalk/ddp.c
+++ b/net/appletalk/ddp.c
@@ -648,7 +648,7 @@ static int ddp_device_event(struct notifier_block *this, unsigned long event,
 {
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event == NETDEV_DOWN)
@@ -1405,7 +1405,7 @@ static int atalk_rcv(struct sk_buff *skb, struct net_device *dev,
 	int origlen;
 	__u16 len_hops;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto freeit;
 
 	/* Don't mangle buffer if shared */
@@ -1493,7 +1493,7 @@ freeit:
 static int ltalk_rcv(struct sk_buff *skb, struct net_device *dev,
 		     struct packet_type *pt, struct net_device *orig_dev)
 {
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto freeit;
 
 	/* Expand any short form frames */
diff --git a/net/atm/clip.c b/net/atm/clip.c
index e82da6746723..6f8223ebf551 100644
--- a/net/atm/clip.c
+++ b/net/atm/clip.c
@@ -612,7 +612,7 @@ static int clip_device_event(struct notifier_block *this, unsigned long event,
 {
 	struct net_device *dev = arg;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event == NETDEV_UNREGISTER) {
diff --git a/net/atm/mpc.c b/net/atm/mpc.c
index 9c7f712fc7e9..9db332e7a6c0 100644
--- a/net/atm/mpc.c
+++ b/net/atm/mpc.c
@@ -964,7 +964,7 @@ static int mpoa_event_listener(struct notifier_block *mpoa_notifier, unsigned lo
 
 	dev = (struct net_device *)dev_ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (dev->name == NULL || strncmp(dev->name, "lec", 3))
diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c
index 48bfcc741f25..ee9dd83e7561 100644
--- a/net/ax25/af_ax25.c
+++ b/net/ax25/af_ax25.c
@@ -116,7 +116,7 @@ static int ax25_device_event(struct notifier_block *this, unsigned long event,
 {
 	struct net_device *dev = (struct net_device *)ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* Reject non AX.25 devices */
diff --git a/net/ax25/ax25_in.c b/net/ax25/ax25_in.c
index d1be080dcb25..33790a8efbc8 100644
--- a/net/ax25/ax25_in.c
+++ b/net/ax25/ax25_in.c
@@ -451,7 +451,7 @@ int ax25_kiss_rcv(struct sk_buff *skb, struct net_device *dev,
 	skb->sk = NULL;		/* Initially we don't know who it's for */
 	skb->destructor = NULL;	/* Who initializes this, dammit?! */
 
-	if (dev->nd_net != &init_net) {
+	if (dev_net(dev) != &init_net) {
 		kfree_skb(skb);
 		return 0;
 	}
diff --git a/net/bridge/br_notify.c b/net/bridge/br_notify.c
index 07ac3ae68d8f..00644a544e3c 100644
--- a/net/bridge/br_notify.c
+++ b/net/bridge/br_notify.c
@@ -37,7 +37,7 @@ static int br_device_event(struct notifier_block *unused, unsigned long event, v
 	struct net_bridge_port *p = dev->br_port;
 	struct net_bridge *br;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* not a port of a bridge */
diff --git a/net/bridge/br_stp_bpdu.c b/net/bridge/br_stp_bpdu.c
index 0edbd2a1c3f3..8deab645ef75 100644
--- a/net/bridge/br_stp_bpdu.c
+++ b/net/bridge/br_stp_bpdu.c
@@ -142,7 +142,7 @@ int br_stp_rcv(struct sk_buff *skb, struct net_device *dev,
 	struct net_bridge *br;
 	const unsigned char *buf;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto err;
 
 	if (!p)
diff --git a/net/can/af_can.c b/net/can/af_can.c
index 36b9f22ed83a..2759b76f731c 100644
--- a/net/can/af_can.c
+++ b/net/can/af_can.c
@@ -599,7 +599,7 @@ static int can_rcv(struct sk_buff *skb, struct net_device *dev,
 	struct dev_rcv_lists *d;
 	int matches;
 
-	if (dev->type != ARPHRD_CAN || dev->nd_net != &init_net) {
+	if (dev->type != ARPHRD_CAN || dev_net(dev) != &init_net) {
 		kfree_skb(skb);
 		return 0;
 	}
@@ -710,7 +710,7 @@ static int can_notifier(struct notifier_block *nb, unsigned long msg,
 	struct net_device *dev = (struct net_device *)data;
 	struct dev_rcv_lists *d;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (dev->type != ARPHRD_CAN)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index bd4282dae754..e9f99b2c6bc9 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -1285,7 +1285,7 @@ static int bcm_notifier(struct notifier_block *nb, unsigned long msg,
 	struct bcm_op *op;
 	int notify_enodev = 0;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (dev->type != ARPHRD_CAN)
diff --git a/net/can/raw.c b/net/can/raw.c
index 94cd7f27c444..ead50c7c0d40 100644
--- a/net/can/raw.c
+++ b/net/can/raw.c
@@ -210,7 +210,7 @@ static int raw_notifier(struct notifier_block *nb,
 	struct raw_sock *ro = container_of(nb, struct raw_sock, notifier);
 	struct sock *sk = &ro->sk;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (dev->type != ARPHRD_CAN)
diff --git a/net/core/dev.c b/net/core/dev.c
index aebd08606040..812534828914 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -216,7 +216,7 @@ static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex)
 /* Device list insertion */
 static int list_netdevice(struct net_device *dev)
 {
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 
 	ASSERT_RTNL();
 
@@ -852,8 +852,8 @@ int dev_alloc_name(struct net_device *dev, const char *name)
 	struct net *net;
 	int ret;
 
-	BUG_ON(!dev->nd_net);
-	net = dev->nd_net;
+	BUG_ON(!dev_net(dev));
+	net = dev_net(dev);
 	ret = __dev_alloc_name(net, name, buf);
 	if (ret >= 0)
 		strlcpy(dev->name, buf, IFNAMSIZ);
@@ -877,9 +877,9 @@ int dev_change_name(struct net_device *dev, char *newname)
 	struct net *net;
 
 	ASSERT_RTNL();
-	BUG_ON(!dev->nd_net);
+	BUG_ON(!dev_net(dev));
 
-	net = dev->nd_net;
+	net = dev_net(dev);
 	if (dev->flags & IFF_UP)
 		return -EBUSY;
 
@@ -2615,7 +2615,7 @@ static int ptype_seq_show(struct seq_file *seq, void *v)
 
 	if (v == SEQ_START_TOKEN)
 		seq_puts(seq, "Type Device      Function\n");
-	else if (pt->dev == NULL || pt->dev->nd_net == seq_file_net(seq)) {
+	else if (pt->dev == NULL || dev_net(pt->dev) == seq_file_net(seq)) {
 		if (pt->type == htons(ETH_P_ALL))
 			seq_puts(seq, "ALL ");
 		else
@@ -3689,8 +3689,8 @@ int register_netdevice(struct net_device *dev)
 
 	/* When net_device's are persistent, this will be fatal. */
 	BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
-	BUG_ON(!dev->nd_net);
-	net = dev->nd_net;
+	BUG_ON(!dev_net(dev));
+	net = dev_net(dev);
 
 	spin_lock_init(&dev->queue_lock);
 	spin_lock_init(&dev->_xmit_lock);
@@ -4011,7 +4011,7 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
 	dev = (struct net_device *)
 		(((long)p + NETDEV_ALIGN_CONST) & ~NETDEV_ALIGN_CONST);
 	dev->padded = (char *)dev - (char *)p;
-	dev->nd_net = &init_net;
+	dev_net_set(dev, &init_net);
 
 	if (sizeof_priv) {
 		dev->priv = ((char *)dev +
@@ -4136,7 +4136,7 @@ int dev_change_net_namespace(struct net_device *dev, struct net *net, const char
 
 	/* Get out if there is nothing todo */
 	err = 0;
-	if (dev->nd_net == net)
+	if (dev_net(dev) == net)
 		goto out;
 
 	/* Pick the destination device name, and ensure
@@ -4187,7 +4187,7 @@ int dev_change_net_namespace(struct net_device *dev, struct net *net, const char
 	dev_addr_discard(dev);
 
 	/* Actually switch the network namespace */
-	dev->nd_net = net;
+	dev_net_set(dev, net);
 
 	/* Assign the new device name */
 	if (destname != dev->name)
diff --git a/net/core/dst.c b/net/core/dst.c
index 3a01a819ba47..694cd2a3f6d2 100644
--- a/net/core/dst.c
+++ b/net/core/dst.c
@@ -279,7 +279,7 @@ static inline void dst_ifdown(struct dst_entry *dst, struct net_device *dev,
 	if (!unregister) {
 		dst->input = dst->output = dst_discard;
 	} else {
-		dst->dev = dst->dev->nd_net->loopback_dev;
+		dst->dev = dev_net(dst->dev)->loopback_dev;
 		dev_hold(dst->dev);
 		dev_put(dev);
 		if (dst->neighbour && dst->neighbour->dev == dev) {
diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
index 42ccaf5b8509..942be93a2eb0 100644
--- a/net/core/fib_rules.c
+++ b/net/core/fib_rules.c
@@ -618,7 +618,7 @@ static int fib_rules_event(struct notifier_block *this, unsigned long event,
 			    void *ptr)
 {
 	struct net_device *dev = ptr;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct fib_rules_ops *ops;
 
 	ASSERT_RTNL();
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index 23c0a10c0c37..c978bd1cd659 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -388,7 +388,7 @@ struct neighbour *neigh_lookup_nodev(struct neigh_table *tbl, struct net *net,
 	hash_val = tbl->hash(pkey, NULL);
 	for (n = tbl->hash_buckets[hash_val & tbl->hash_mask]; n; n = n->next) {
 		if (!memcmp(n->primary_key, pkey, key_len) &&
-		    (net == n->dev->nd_net)) {
+		    dev_net(n->dev) == net) {
 			neigh_hold(n);
 			NEIGH_CACHE_STAT_INC(tbl, hits);
 			break;
@@ -1298,7 +1298,7 @@ struct neigh_parms *neigh_parms_alloc(struct net_device *dev,
 	struct neigh_parms *p, *ref;
 	struct net *net;
 
-	net = dev->nd_net;
+	net = dev_net(dev);
 	ref = lookup_neigh_params(tbl, net, 0);
 	if (!ref)
 		return NULL;
@@ -2050,7 +2050,7 @@ static int neigh_dump_table(struct neigh_table *tbl, struct sk_buff *skb,
 			s_idx = 0;
 		for (n = tbl->hash_buckets[h], idx = 0; n; n = n->next) {
 			int lidx;
-			if (n->dev->nd_net != net)
+			if (dev_net(n->dev) != net)
 				continue;
 			lidx = idx++;
 			if (lidx < s_idx)
@@ -2155,7 +2155,7 @@ static struct neighbour *neigh_get_first(struct seq_file *seq)
 		n = tbl->hash_buckets[bucket];
 
 		while (n) {
-			if (n->dev->nd_net != net)
+			if (dev_net(n->dev) != net)
 				goto next;
 			if (state->neigh_sub_iter) {
 				loff_t fakep = 0;
@@ -2198,7 +2198,7 @@ static struct neighbour *neigh_get_next(struct seq_file *seq,
 
 	while (1) {
 		while (n) {
-			if (n->dev->nd_net != net)
+			if (dev_net(n->dev) != net)
 				goto next;
 			if (state->neigh_sub_iter) {
 				void *v = state->neigh_sub_iter(state, n, pos);
@@ -2482,7 +2482,7 @@ static inline size_t neigh_nlmsg_size(void)
 
 static void __neigh_notify(struct neighbour *n, int type, int flags)
 {
-	struct net *net = n->dev->nd_net;
+	struct net *net = dev_net(n->dev);
 	struct sk_buff *skb;
 	int err = -ENOBUFS;
 
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index 20e63b302ba6..a803b442234c 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -1874,7 +1874,7 @@ static int pktgen_device_event(struct notifier_block *unused,
 {
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* It is OK that we do not hold the group lock right now,
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 2bd9c5f7627d..09250a0800f6 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -972,7 +972,7 @@ struct net_device *rtnl_create_link(struct net *net, char *ifname,
 			goto err_free;
 	}
 
-	dev->nd_net = net;
+	dev_net_set(dev, net);
 	dev->rtnl_link_ops = ops;
 
 	if (tb[IFLA_MTU])
@@ -1198,7 +1198,7 @@ static int rtnl_dump_all(struct sk_buff *skb, struct netlink_callback *cb)
 
 void rtmsg_ifinfo(int type, struct net_device *dev, unsigned change)
 {
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct sk_buff *skb;
 	int err = -ENOBUFS;
 
diff --git a/net/decnet/af_decnet.c b/net/decnet/af_decnet.c
index 23fd95a7ad15..3554fb3d251c 100644
--- a/net/decnet/af_decnet.c
+++ b/net/decnet/af_decnet.c
@@ -2089,7 +2089,7 @@ static int dn_device_event(struct notifier_block *this, unsigned long event,
 {
 	struct net_device *dev = (struct net_device *)ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	switch(event) {
diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c
index 9dc0abb50eaf..0a46b6c10e51 100644
--- a/net/decnet/dn_route.c
+++ b/net/decnet/dn_route.c
@@ -580,7 +580,7 @@ int dn_route_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type
 	struct dn_dev *dn = (struct dn_dev *)dev->dn_ptr;
 	unsigned char padlen = 0;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto dump_it;
 
 	if (dn == NULL)
diff --git a/net/econet/af_econet.c b/net/econet/af_econet.c
index bc0f6252613f..68d154480043 100644
--- a/net/econet/af_econet.c
+++ b/net/econet/af_econet.c
@@ -1064,7 +1064,7 @@ static int econet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet
 	struct sock *sk;
 	struct ec_device *edev = dev->ec_ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	if (skb->pkt_type == PACKET_OTHERHOST)
@@ -1121,7 +1121,7 @@ static int econet_notifier(struct notifier_block *this, unsigned long msg, void
 	struct net_device *dev = (struct net_device *)data;
 	struct ec_device *edev;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	switch (msg) {
diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c
index 832473e30b36..3ce2e137e7bc 100644
--- a/net/ipv4/arp.c
+++ b/net/ipv4/arp.c
@@ -242,7 +242,7 @@ static int arp_constructor(struct neighbour *neigh)
 		return -EINVAL;
 	}
 
-	neigh->type = inet_addr_type(dev->nd_net, addr);
+	neigh->type = inet_addr_type(dev_net(dev), addr);
 
 	parms = in_dev->arp_parms;
 	__neigh_parms_put(neigh->parms);
@@ -341,14 +341,14 @@ static void arp_solicit(struct neighbour *neigh, struct sk_buff *skb)
 	switch (IN_DEV_ARP_ANNOUNCE(in_dev)) {
 	default:
 	case 0:		/* By default announce any local IP */
-		if (skb && inet_addr_type(dev->nd_net, ip_hdr(skb)->saddr) == RTN_LOCAL)
+		if (skb && inet_addr_type(dev_net(dev), ip_hdr(skb)->saddr) == RTN_LOCAL)
 			saddr = ip_hdr(skb)->saddr;
 		break;
 	case 1:		/* Restrict announcements of saddr in same subnet */
 		if (!skb)
 			break;
 		saddr = ip_hdr(skb)->saddr;
-		if (inet_addr_type(dev->nd_net, saddr) == RTN_LOCAL) {
+		if (inet_addr_type(dev_net(dev), saddr) == RTN_LOCAL) {
 			/* saddr should be known to target */
 			if (inet_addr_onlink(in_dev, target, saddr))
 				break;
@@ -424,7 +424,7 @@ static int arp_filter(__be32 sip, __be32 tip, struct net_device *dev)
 	int flag = 0;
 	/*unsigned long now; */
 
-	if (ip_route_output_key(dev->nd_net, &rt, &fl) < 0)
+	if (ip_route_output_key(dev_net(dev), &rt, &fl) < 0)
 		return 1;
 	if (rt->u.dst.dev != dev) {
 		NET_INC_STATS_BH(LINUX_MIB_ARPFILTER);
@@ -477,7 +477,7 @@ int arp_find(unsigned char *haddr, struct sk_buff *skb)
 
 	paddr = skb->rtable->rt_gateway;
 
-	if (arp_set_predefined(inet_addr_type(dev->nd_net, paddr), haddr, paddr, dev))
+	if (arp_set_predefined(inet_addr_type(dev_net(dev), paddr), haddr, paddr, dev))
 		return 0;
 
 	n = __neigh_lookup(&arp_tbl, &paddr, dev, 1);
@@ -709,7 +709,7 @@ static int arp_process(struct sk_buff *skb)
 	u16 dev_type = dev->type;
 	int addr_type;
 	struct neighbour *n;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 
 	/* arp_rcv below verifies the ARP header and verifies the device
 	 * is ARP'able.
@@ -858,7 +858,7 @@ static int arp_process(struct sk_buff *skb)
 
 	n = __neigh_lookup(&arp_tbl, &sip, dev, 0);
 
-	if (IPV4_DEVCONF_ALL(dev->nd_net, ARP_ACCEPT)) {
+	if (IPV4_DEVCONF_ALL(dev_net(dev), ARP_ACCEPT)) {
 		/* Unsolicited ARP is not accepted by default.
 		   It is possible, that this option should be enabled for some
 		   devices (strip is candidate)
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 4a10dbbbe0a1..823c724a8593 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -165,7 +165,7 @@ static struct in_device *inetdev_init(struct net_device *dev)
 	if (!in_dev)
 		goto out;
 	INIT_RCU_HEAD(&in_dev->rcu_head);
-	memcpy(&in_dev->cnf, dev->nd_net->ipv4.devconf_dflt,
+	memcpy(&in_dev->cnf, dev_net(dev)->ipv4.devconf_dflt,
 			sizeof(in_dev->cnf));
 	in_dev->cnf.sysctl = NULL;
 	in_dev->dev = dev;
@@ -872,7 +872,7 @@ __be32 inet_select_addr(const struct net_device *dev, __be32 dst, int scope)
 {
 	__be32 addr = 0;
 	struct in_device *in_dev;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 
 	rcu_read_lock();
 	in_dev = __in_dev_get_rcu(dev);
@@ -974,7 +974,7 @@ __be32 inet_confirm_addr(struct in_device *in_dev,
 	if (scope != RT_SCOPE_LINK)
 		return confirm_addr_indev(in_dev, dst, local, scope);
 
-	net = in_dev->dev->nd_net;
+	net = dev_net(in_dev->dev);
 	read_lock(&dev_base_lock);
 	rcu_read_lock();
 	for_each_netdev(net, dev) {
@@ -1203,7 +1203,7 @@ static void rtmsg_ifa(int event, struct in_ifaddr* ifa, struct nlmsghdr *nlh,
 	int err = -ENOBUFS;
 	struct net *net;
 
-	net = ifa->ifa_dev->dev->nd_net;
+	net = dev_net(ifa->ifa_dev->dev);
 	skb = nlmsg_new(inet_nlmsg_size(), GFP_KERNEL);
 	if (skb == NULL)
 		goto errout;
@@ -1517,7 +1517,7 @@ static void devinet_sysctl_register(struct in_device *idev)
 {
 	neigh_sysctl_register(idev->dev, idev->arp_parms, NET_IPV4,
 			NET_IPV4_NEIGH, "ipv4", NULL, NULL);
-	__devinet_sysctl_register(idev->dev->nd_net, idev->dev->name,
+	__devinet_sysctl_register(dev_net(idev->dev), idev->dev->name,
 			idev->dev->ifindex, &idev->cnf);
 }
 
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index 86ff2711fc95..0e4b34b07cb5 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -257,7 +257,7 @@ int fib_validate_source(__be32 src, __be32 dst, u8 tos, int oif,
 	if (in_dev == NULL)
 		goto e_inval;
 
-	net = dev->nd_net;
+	net = dev_net(dev);
 	if (fib_lookup(net, &fl, &res))
 		goto last_resort;
 	if (res.type != RTN_UNICAST)
@@ -674,7 +674,7 @@ out:
 
 static void fib_magic(int cmd, int type, __be32 dst, int dst_len, struct in_ifaddr *ifa)
 {
-	struct net *net = ifa->ifa_dev->dev->nd_net;
+	struct net *net = dev_net(ifa->ifa_dev->dev);
 	struct fib_table *tb;
 	struct fib_config cfg = {
 		.fc_protocol = RTPROT_KERNEL,
@@ -801,15 +801,15 @@ static void fib_del_ifaddr(struct in_ifaddr *ifa)
 		fib_magic(RTM_DELROUTE, RTN_LOCAL, ifa->ifa_local, 32, prim);
 
 		/* Check, that this local address finally disappeared. */
-		if (inet_addr_type(dev->nd_net, ifa->ifa_local) != RTN_LOCAL) {
+		if (inet_addr_type(dev_net(dev), ifa->ifa_local) != RTN_LOCAL) {
 			/* And the last, but not the least thing.
 			   We must flush stray FIB entries.
 
 			   First of all, we scan fib_info list searching
 			   for stray nexthop entries, then ignite fib_flush.
 			*/
-			if (fib_sync_down_addr(dev->nd_net, ifa->ifa_local))
-				fib_flush(dev->nd_net);
+			if (fib_sync_down_addr(dev_net(dev), ifa->ifa_local))
+				fib_flush(dev_net(dev));
 		}
 	}
 #undef LOCAL_OK
@@ -899,7 +899,7 @@ static void nl_fib_lookup_exit(struct net *net)
 static void fib_disable_ip(struct net_device *dev, int force)
 {
 	if (fib_sync_down_dev(dev, force))
-		fib_flush(dev->nd_net);
+		fib_flush(dev_net(dev));
 	rt_cache_flush(0);
 	arp_ifdown(dev);
 }
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index ff9a8e643fcc..f38f093ef751 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -351,7 +351,7 @@ static void icmp_push_reply(struct icmp_bxm *icmp_param,
 	struct sock *sk;
 	struct sk_buff *skb;
 
-	sk = icmp_sk(rt->u.dst.dev->nd_net);
+	sk = icmp_sk(dev_net(rt->u.dst.dev));
 	if (ip_append_data(sk, icmp_glue_bits, icmp_param,
 			   icmp_param->data_len+icmp_param->head_len,
 			   icmp_param->head_len,
@@ -382,7 +382,7 @@ static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb)
 {
 	struct ipcm_cookie ipc;
 	struct rtable *rt = skb->rtable;
-	struct net *net = rt->u.dst.dev->nd_net;
+	struct net *net = dev_net(rt->u.dst.dev);
 	struct sock *sk = icmp_sk(net);
 	struct inet_sock *inet = inet_sk(sk);
 	__be32 daddr;
@@ -447,7 +447,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info)
 
 	if (!rt)
 		goto out;
-	net = rt->u.dst.dev->nd_net;
+	net = dev_net(rt->u.dst.dev);
 	sk = icmp_sk(net);
 
 	/*
@@ -677,7 +677,7 @@ static void icmp_unreach(struct sk_buff *skb)
 	u32 info = 0;
 	struct net *net;
 
-	net = skb->dst->dev->nd_net;
+	net = dev_net(skb->dst->dev);
 
 	/*
 	 *	Incomplete header ?
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index 6a4ee8da6994..682f632bfb77 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -130,12 +130,12 @@
  */
 
 #define IGMP_V1_SEEN(in_dev) \
-	(IPV4_DEVCONF_ALL(in_dev->dev->nd_net, FORCE_IGMP_VERSION) == 1 || \
+	(IPV4_DEVCONF_ALL(dev_net(in_dev->dev), FORCE_IGMP_VERSION) == 1 || \
 	 IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 1 || \
 	 ((in_dev)->mr_v1_seen && \
 	  time_before(jiffies, (in_dev)->mr_v1_seen)))
 #define IGMP_V2_SEEN(in_dev) \
-	(IPV4_DEVCONF_ALL(in_dev->dev->nd_net, FORCE_IGMP_VERSION) == 2 || \
+	(IPV4_DEVCONF_ALL(dev_net(in_dev->dev), FORCE_IGMP_VERSION) == 2 || \
 	 IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 2 || \
 	 ((in_dev)->mr_v2_seen && \
 	  time_before(jiffies, (in_dev)->mr_v2_seen)))
@@ -1198,7 +1198,7 @@ void ip_mc_inc_group(struct in_device *in_dev, __be32 addr)
 
 	ASSERT_RTNL();
 
-	if (in_dev->dev->nd_net != &init_net)
+	if (dev_net(in_dev->dev) != &init_net)
 		return;
 
 	for (im=in_dev->mc_list; im; im=im->next) {
@@ -1280,7 +1280,7 @@ void ip_mc_dec_group(struct in_device *in_dev, __be32 addr)
 
 	ASSERT_RTNL();
 
-	if (in_dev->dev->nd_net != &init_net)
+	if (dev_net(in_dev->dev) != &init_net)
 		return;
 
 	for (ip=&in_dev->mc_list; (i=*ip)!=NULL; ip=&i->next) {
@@ -1310,7 +1310,7 @@ void ip_mc_down(struct in_device *in_dev)
 
 	ASSERT_RTNL();
 
-	if (in_dev->dev->nd_net != &init_net)
+	if (dev_net(in_dev->dev) != &init_net)
 		return;
 
 	for (i=in_dev->mc_list; i; i=i->next)
@@ -1333,7 +1333,7 @@ void ip_mc_init_dev(struct in_device *in_dev)
 {
 	ASSERT_RTNL();
 
-	if (in_dev->dev->nd_net != &init_net)
+	if (dev_net(in_dev->dev) != &init_net)
 		return;
 
 	in_dev->mc_tomb = NULL;
@@ -1359,7 +1359,7 @@ void ip_mc_up(struct in_device *in_dev)
 
 	ASSERT_RTNL();
 
-	if (in_dev->dev->nd_net != &init_net)
+	if (dev_net(in_dev->dev) != &init_net)
 		return;
 
 	ip_mc_inc_group(in_dev, IGMP_ALL_HOSTS);
@@ -1378,7 +1378,7 @@ void ip_mc_destroy_dev(struct in_device *in_dev)
 
 	ASSERT_RTNL();
 
-	if (in_dev->dev->nd_net != &init_net)
+	if (dev_net(in_dev->dev) != &init_net)
 		return;
 
 	/* Deactivate timers */
diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c
index 8b448c4b9080..fcb60e76b234 100644
--- a/net/ipv4/ip_fragment.c
+++ b/net/ipv4/ip_fragment.c
@@ -571,7 +571,7 @@ int ip_defrag(struct sk_buff *skb, u32 user)
 
 	IP_INC_STATS_BH(IPSTATS_MIB_REASMREQDS);
 
-	net = skb->dev ? skb->dev->nd_net : skb->dst->dev->nd_net;
+	net = skb->dev ? dev_net(skb->dev) : dev_net(skb->dst->dev);
 	/* Start by cleaning up the memory. */
 	if (atomic_read(&net->ipv4.frags.mem) > net->ipv4.frags.high_thresh)
 		ip_evictor(net);
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index f9ee84420cb3..50972b397a9a 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -1190,7 +1190,7 @@ static int ipgre_close(struct net_device *dev)
 	struct ip_tunnel *t = netdev_priv(dev);
 	if (ipv4_is_multicast(t->parms.iph.daddr) && t->mlink) {
 		struct in_device *in_dev;
-		in_dev = inetdev_by_index(dev->nd_net, t->mlink);
+		in_dev = inetdev_by_index(dev_net(dev), t->mlink);
 		if (in_dev) {
 			ip_mc_dec_group(in_dev, t->parms.iph.daddr);
 			in_dev_put(in_dev);
diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c
index 2aeea5d15425..26685c83a146 100644
--- a/net/ipv4/ip_input.c
+++ b/net/ipv4/ip_input.c
@@ -172,7 +172,7 @@ int ip_call_ra_chain(struct sk_buff *skb)
 		if (sk && inet_sk(sk)->num == protocol &&
 		    (!sk->sk_bound_dev_if ||
 		     sk->sk_bound_dev_if == dev->ifindex) &&
-		    sk->sk_net == dev->nd_net) {
+		    sk->sk_net == dev_net(dev)) {
 			if (ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)) {
 				if (ip_defrag(skb, IP_DEFRAG_CALL_RA_CHAIN)) {
 					read_unlock(&ip_ra_lock);
@@ -199,7 +199,7 @@ int ip_call_ra_chain(struct sk_buff *skb)
 
 static int ip_local_deliver_finish(struct sk_buff *skb)
 {
-	struct net *net = skb->dev->nd_net;
+	struct net *net = dev_net(skb->dev);
 
 	__skb_pull(skb, ip_hdrlen(skb));
 
@@ -291,7 +291,7 @@ static inline int ip_rcv_options(struct sk_buff *skb)
 	opt = &(IPCB(skb)->opt);
 	opt->optlen = iph->ihl*4 - sizeof(struct iphdr);
 
-	if (ip_options_compile(dev->nd_net, opt, skb)) {
+	if (ip_options_compile(dev_net(dev), opt, skb)) {
 		IP_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
 		goto drop;
 	}
diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
index 87cc1222c600..d107543d3f81 100644
--- a/net/ipv4/ip_options.c
+++ b/net/ipv4/ip_options.c
@@ -145,7 +145,7 @@ int ip_options_echo(struct ip_options * dopt, struct sk_buff * skb)
 						__be32 addr;
 
 						memcpy(&addr, sptr+soffset-1, 4);
-						if (inet_addr_type(skb->dst->dev->nd_net, addr) != RTN_LOCAL) {
+						if (inet_addr_type(dev_net(skb->dst->dev), addr) != RTN_LOCAL) {
 							dopt->ts_needtime = 1;
 							soffset += 8;
 						}
diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c
index 96138b128de8..08e8fb60d315 100644
--- a/net/ipv4/ipconfig.c
+++ b/net/ipv4/ipconfig.c
@@ -434,7 +434,7 @@ ic_rarp_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt
 	unsigned char *sha, *tha;		/* s for "source", t for "target" */
 	struct ic_device *d;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
@@ -854,7 +854,7 @@ static int __init ic_bootp_recv(struct sk_buff *skb, struct net_device *dev, str
 	struct ic_device *d;
 	int len, ext_len;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	/* Perform verifications before taking the lock.  */
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 7d63d74ef62a..e54bc1364473 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -1089,7 +1089,7 @@ static int ipmr_device_event(struct notifier_block *this, unsigned long event, v
 	struct vif_device *v;
 	int ct;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event != NETDEV_UNREGISTER)
diff --git a/net/ipv4/netfilter/ip_queue.c b/net/ipv4/netfilter/ip_queue.c
index fe05da41d6ba..500998a2dec1 100644
--- a/net/ipv4/netfilter/ip_queue.c
+++ b/net/ipv4/netfilter/ip_queue.c
@@ -481,7 +481,7 @@ ipq_rcv_dev_event(struct notifier_block *this,
 {
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* Drop any packets associated with the downed device */
diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c
index c6817b18366a..84c26dd27d81 100644
--- a/net/ipv4/netfilter/ipt_MASQUERADE.c
+++ b/net/ipv4/netfilter/ipt_MASQUERADE.c
@@ -120,7 +120,7 @@ static int masq_device_event(struct notifier_block *this,
 {
 	const struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event == NETDEV_DOWN) {
diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c
index 3f68a937b602..8756d502a47f 100644
--- a/net/ipv4/raw.c
+++ b/net/ipv4/raw.c
@@ -168,7 +168,7 @@ static int raw_v4_input(struct sk_buff *skb, struct iphdr *iph, int hash)
 	if (hlist_empty(head))
 		goto out;
 
-	net = skb->dev->nd_net;
+	net = dev_net(skb->dev);
 	sk = __raw_v4_lookup(net, __sk_head(head), iph->protocol,
 			     iph->saddr, iph->daddr,
 			     skb->dev->ifindex);
@@ -276,7 +276,7 @@ void raw_icmp_error(struct sk_buff *skb, int protocol, u32 info)
 	raw_sk = sk_head(&raw_v4_hashinfo.ht[hash]);
 	if (raw_sk != NULL) {
 		iph = (struct iphdr *)skb->data;
-		net = skb->dev->nd_net;
+		net = dev_net(skb->dev);
 
 		while ((raw_sk = __raw_v4_lookup(net, raw_sk, protocol,
 						iph->daddr, iph->saddr,
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 2941ef21f203..7768d718e199 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -284,7 +284,7 @@ static struct rtable *rt_cache_get_first(struct rt_cache_iter_state *st)
 		rcu_read_lock_bh();
 		r = rcu_dereference(rt_hash_table[st->bucket].chain);
 		while (r) {
-			if (r->u.dst.dev->nd_net == st->p.net &&
+			if (dev_net(r->u.dst.dev) == st->p.net &&
 			    r->rt_genid == st->genid)
 				return r;
 			r = rcu_dereference(r->u.dst.rt_next);
@@ -312,7 +312,7 @@ static struct rtable *rt_cache_get_next(struct rt_cache_iter_state *st,
 					struct rtable *r)
 {
 	while ((r = __rt_cache_get_next(st, r)) != NULL) {
-		if (r->u.dst.dev->nd_net != st->p.net)
+		if (dev_net(r->u.dst.dev) != st->p.net)
 			continue;
 		if (r->rt_genid == st->genid)
 			break;
@@ -680,7 +680,7 @@ static inline int compare_keys(struct flowi *fl1, struct flowi *fl2)
 
 static inline int compare_netns(struct rtable *rt1, struct rtable *rt2)
 {
-	return rt1->u.dst.dev->nd_net == rt2->u.dst.dev->nd_net;
+	return dev_net(rt1->u.dst.dev) == dev_net(rt2->u.dst.dev);
 }
 
 /*
@@ -1164,7 +1164,7 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw,
 	if (!in_dev)
 		return;
 
-	net = dev->nd_net;
+	net = dev_net(dev);
 	if (new_gw == old_gw || !IN_DEV_RX_REDIRECTS(in_dev)
 	    || ipv4_is_multicast(new_gw) || ipv4_is_lbcast(new_gw)
 	    || ipv4_is_zeronet(new_gw))
@@ -1195,7 +1195,7 @@ void ip_rt_redirect(__be32 old_gw, __be32 daddr, __be32 new_gw,
 				    rth->fl.oif != ikeys[k] ||
 				    rth->fl.iif != 0 ||
 				    rth->rt_genid != atomic_read(&rt_genid) ||
-				    rth->u.dst.dev->nd_net != net) {
+				    dev_net(rth->u.dst.dev) != net) {
 					rthp = &rth->u.dst.rt_next;
 					continue;
 				}
@@ -1454,7 +1454,7 @@ unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph,
 			    rth->rt_src  == iph->saddr &&
 			    rth->fl.iif == 0 &&
 			    !(dst_metric_locked(&rth->u.dst, RTAX_MTU)) &&
-			    rth->u.dst.dev->nd_net == net &&
+			    dev_net(rth->u.dst.dev) == net &&
 			    rth->rt_genid == atomic_read(&rt_genid)) {
 				unsigned short mtu = new_mtu;
 
@@ -1530,9 +1530,9 @@ static void ipv4_dst_ifdown(struct dst_entry *dst, struct net_device *dev,
 {
 	struct rtable *rt = (struct rtable *) dst;
 	struct in_device *idev = rt->idev;
-	if (dev != dev->nd_net->loopback_dev && idev && idev->dev == dev) {
+	if (dev != dev_net(dev)->loopback_dev && idev && idev->dev == dev) {
 		struct in_device *loopback_idev =
-			in_dev_get(dev->nd_net->loopback_dev);
+			in_dev_get(dev_net(dev)->loopback_dev);
 		if (loopback_idev) {
 			rt->idev = loopback_idev;
 			in_dev_put(idev);
@@ -1576,7 +1576,7 @@ void ip_rt_get_source(u8 *addr, struct rtable *rt)
 
 	if (rt->fl.iif == 0)
 		src = rt->rt_src;
-	else if (fib_lookup(rt->u.dst.dev->nd_net, &rt->fl, &res) == 0) {
+	else if (fib_lookup(dev_net(rt->u.dst.dev), &rt->fl, &res) == 0) {
 		src = FIB_RES_PREFSRC(res);
 		fib_res_put(&res);
 	} else
@@ -1900,7 +1900,7 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 	__be32		spec_dst;
 	int		err = -EINVAL;
 	int		free_res = 0;
-	struct net    * net = dev->nd_net;
+	struct net    * net = dev_net(dev);
 
 	/* IP on this device is disabled. */
 
@@ -2071,7 +2071,7 @@ int ip_route_input(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 	int iif = dev->ifindex;
 	struct net *net;
 
-	net = dev->nd_net;
+	net = dev_net(dev);
 	tos &= IPTOS_RT_MASK;
 	hash = rt_hash(daddr, saddr, iif);
 
@@ -2084,7 +2084,7 @@ int ip_route_input(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 		    rth->fl.oif == 0 &&
 		    rth->fl.mark == skb->mark &&
 		    rth->fl.fl4_tos == tos &&
-		    rth->u.dst.dev->nd_net == net &&
+		    dev_net(rth->u.dst.dev) == net &&
 		    rth->rt_genid == atomic_read(&rt_genid)) {
 			dst_use(&rth->u.dst, jiffies);
 			RT_CACHE_STAT_INC(in_hit);
@@ -2486,7 +2486,7 @@ int __ip_route_output_key(struct net *net, struct rtable **rp,
 		    rth->fl.mark == flp->mark &&
 		    !((rth->fl.fl4_tos ^ flp->fl4_tos) &
 			    (IPTOS_RT_MASK | RTO_ONLINK)) &&
-		    rth->u.dst.dev->nd_net == net &&
+		    dev_net(rth->u.dst.dev) == net &&
 		    rth->rt_genid == atomic_read(&rt_genid)) {
 			dst_use(&rth->u.dst, jiffies);
 			RT_CACHE_STAT_INC(out_hit);
@@ -2795,7 +2795,7 @@ int ip_rt_dump(struct sk_buff *skb,  struct netlink_callback *cb)
 		rcu_read_lock_bh();
 		for (rt = rcu_dereference(rt_hash_table[h].chain), idx = 0; rt;
 		     rt = rcu_dereference(rt->u.dst.rt_next), idx++) {
-			if (rt->u.dst.dev->nd_net != net || idx < s_idx)
+			if (dev_net(rt->u.dst.dev) != net || idx < s_idx)
 				continue;
 			if (rt->rt_genid != atomic_read(&rt_genid))
 				continue;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 649d00a50cb1..28bece6f281b 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -353,7 +353,7 @@ void tcp_v4_err(struct sk_buff *skb, u32 info)
 		return;
 	}
 
-	sk = inet_lookup(skb->dev->nd_net, &tcp_hashinfo, iph->daddr, th->dest,
+	sk = inet_lookup(dev_net(skb->dev), &tcp_hashinfo, iph->daddr, th->dest,
 			iph->saddr, th->source, inet_iif(skb));
 	if (!sk) {
 		ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
@@ -1644,7 +1644,7 @@ int tcp_v4_rcv(struct sk_buff *skb)
 	TCP_SKB_CB(skb)->flags	 = iph->tos;
 	TCP_SKB_CB(skb)->sacked	 = 0;
 
-	sk = __inet_lookup(skb->dev->nd_net, &tcp_hashinfo, iph->saddr,
+	sk = __inet_lookup(dev_net(skb->dev), &tcp_hashinfo, iph->saddr,
 			th->source, iph->daddr, th->dest, inet_iif(skb));
 	if (!sk)
 		goto no_tcp_socket;
@@ -1718,7 +1718,7 @@ do_time_wait:
 	}
 	switch (tcp_timewait_state_process(inet_twsk(sk), skb, th)) {
 	case TCP_TW_SYN: {
-		struct sock *sk2 = inet_lookup_listener(skb->dev->nd_net,
+		struct sock *sk2 = inet_lookup_listener(dev_net(skb->dev),
 							&tcp_hashinfo,
 							iph->daddr, th->dest,
 							inet_iif(skb));
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index b37581dfd029..e2cd93481359 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -357,7 +357,7 @@ void __udp4_lib_err(struct sk_buff *skb, u32 info, struct hlist_head udptable[])
 	int harderr;
 	int err;
 
-	sk = __udp4_lib_lookup(skb->dev->nd_net, iph->daddr, uh->dest,
+	sk = __udp4_lib_lookup(dev_net(skb->dev), iph->daddr, uh->dest,
 			iph->saddr, uh->source, skb->dev->ifindex, udptable);
 	if (sk == NULL) {
 		ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
@@ -1181,7 +1181,7 @@ int __udp4_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 	if (rt->rt_flags & (RTCF_BROADCAST|RTCF_MULTICAST))
 		return __udp4_lib_mcast_deliver(skb, uh, saddr, daddr, udptable);
 
-	sk = __udp4_lib_lookup(skb->dev->nd_net, saddr, uh->source, daddr,
+	sk = __udp4_lib_lookup(dev_net(skb->dev), saddr, uh->source, daddr,
 			uh->dest, inet_iif(skb), udptable);
 
 	if (sk != NULL) {
diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c
index 10ed70491434..c63de0a72aba 100644
--- a/net/ipv4/xfrm4_policy.c
+++ b/net/ipv4/xfrm4_policy.c
@@ -221,7 +221,7 @@ static void xfrm4_dst_ifdown(struct dst_entry *dst, struct net_device *dev,
 	xdst = (struct xfrm_dst *)dst;
 	if (xdst->u.rt.idev->dev == dev) {
 		struct in_device *loopback_idev =
-			in_dev_get(dev->nd_net->loopback_dev);
+			in_dev_get(dev_net(dev)->loopback_dev);
 		BUG_ON(!loopback_idev);
 
 		do {
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 89954885dee1..d1de9ec74261 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -335,7 +335,7 @@ static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 
 	rwlock_init(&ndev->lock);
 	ndev->dev = dev;
-	memcpy(&ndev->cnf, dev->nd_net->ipv6.devconf_dflt, sizeof(ndev->cnf));
+	memcpy(&ndev->cnf, dev_net(dev)->ipv6.devconf_dflt, sizeof(ndev->cnf));
 	ndev->cnf.mtu6 = dev->mtu;
 	ndev->cnf.sysctl = NULL;
 	ndev->nd_parms = neigh_parms_alloc(dev, &nd_tbl);
@@ -561,7 +561,7 @@ ipv6_add_addr(struct inet6_dev *idev, const struct in6_addr *addr, int pfxlen,
 	write_lock(&addrconf_hash_lock);
 
 	/* Ignore adding duplicate addresses on an interface */
-	if (ipv6_chk_same_addr(idev->dev->nd_net, addr, idev->dev)) {
+	if (ipv6_chk_same_addr(dev_net(idev->dev), addr, idev->dev)) {
 		ADBG(("ipv6_add_addr: already assigned\n"));
 		err = -EEXIST;
 		goto out;
@@ -751,7 +751,7 @@ static void ipv6_del_addr(struct inet6_ifaddr *ifp)
 	if ((ifp->flags & IFA_F_PERMANENT) && onlink < 1) {
 		struct in6_addr prefix;
 		struct rt6_info *rt;
-		struct net *net = ifp->idev->dev->nd_net;
+		struct net *net = dev_net(ifp->idev->dev);
 		ipv6_addr_prefix(&prefix, &ifp->addr, ifp->prefix_len);
 		rt = rt6_lookup(net, &prefix, NULL, ifp->idev->dev->ifindex, 1);
 
@@ -1044,7 +1044,7 @@ int ipv6_dev_get_saddr(struct net_device *dst_dev,
 {
 	struct ipv6_saddr_score scores[2],
 				*score = &scores[0], *hiscore = &scores[1];
-	struct net *net = dst_dev->nd_net;
+	struct net *net = dev_net(dst_dev);
 	struct ipv6_saddr_dst dst;
 	struct net_device *dev;
 	int dst_type;
@@ -1217,7 +1217,7 @@ int ipv6_chk_addr(struct net *net, struct in6_addr *addr,
 
 	read_lock_bh(&addrconf_hash_lock);
 	for(ifp = inet6_addr_lst[hash]; ifp; ifp=ifp->lst_next) {
-		if (ifp->idev->dev->nd_net != net)
+		if (dev_net(ifp->idev->dev) != net)
 			continue;
 		if (ipv6_addr_equal(&ifp->addr, addr) &&
 		    !(ifp->flags&IFA_F_TENTATIVE)) {
@@ -1239,7 +1239,7 @@ int ipv6_chk_same_addr(struct net *net, const struct in6_addr *addr,
 	u8 hash = ipv6_addr_hash(addr);
 
 	for(ifp = inet6_addr_lst[hash]; ifp; ifp=ifp->lst_next) {
-		if (ifp->idev->dev->nd_net != net)
+		if (dev_net(ifp->idev->dev) != net)
 			continue;
 		if (ipv6_addr_equal(&ifp->addr, addr)) {
 			if (dev == NULL || ifp->idev->dev == dev)
@@ -1257,7 +1257,7 @@ struct inet6_ifaddr *ipv6_get_ifaddr(struct net *net, struct in6_addr *addr,
 
 	read_lock_bh(&addrconf_hash_lock);
 	for(ifp = inet6_addr_lst[hash]; ifp; ifp=ifp->lst_next) {
-		if (ifp->idev->dev->nd_net != net)
+		if (dev_net(ifp->idev->dev) != net)
 			continue;
 		if (ipv6_addr_equal(&ifp->addr, addr)) {
 			if (dev == NULL || ifp->idev->dev == dev ||
@@ -1559,7 +1559,7 @@ addrconf_prefix_route(struct in6_addr *pfx, int plen, struct net_device *dev,
 		.fc_expires = expires,
 		.fc_dst_len = plen,
 		.fc_flags = RTF_UP | flags,
-		.fc_nlinfo.nl_net = dev->nd_net,
+		.fc_nlinfo.nl_net = dev_net(dev),
 	};
 
 	ipv6_addr_copy(&cfg.fc_dst, pfx);
@@ -1586,7 +1586,7 @@ static void addrconf_add_mroute(struct net_device *dev)
 		.fc_ifindex = dev->ifindex,
 		.fc_dst_len = 8,
 		.fc_flags = RTF_UP,
-		.fc_nlinfo.nl_net = dev->nd_net,
+		.fc_nlinfo.nl_net = dev_net(dev),
 	};
 
 	ipv6_addr_set(&cfg.fc_dst, htonl(0xFF000000), 0, 0, 0);
@@ -1603,7 +1603,7 @@ static void sit_route_add(struct net_device *dev)
 		.fc_ifindex = dev->ifindex,
 		.fc_dst_len = 96,
 		.fc_flags = RTF_UP | RTF_NONEXTHOP,
-		.fc_nlinfo.nl_net = dev->nd_net,
+		.fc_nlinfo.nl_net = dev_net(dev),
 	};
 
 	/* prefix length - 96 bits "::d.d.d.d" */
@@ -1704,7 +1704,7 @@ void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len)
 
 	if (pinfo->onlink) {
 		struct rt6_info *rt;
-		rt = rt6_lookup(dev->nd_net, &pinfo->prefix, NULL,
+		rt = rt6_lookup(dev_net(dev), &pinfo->prefix, NULL,
 				dev->ifindex, 1);
 
 		if (rt && ((rt->rt6i_flags & (RTF_GATEWAY | RTF_DEFAULT)) == 0)) {
@@ -1748,7 +1748,7 @@ void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len)
 
 ok:
 
-		ifp = ipv6_get_ifaddr(dev->nd_net, &addr, dev, 1);
+		ifp = ipv6_get_ifaddr(dev_net(dev), &addr, dev, 1);
 
 		if (ifp == NULL && valid_lft) {
 			int max_addresses = in6_dev->cnf.max_addresses;
@@ -2071,7 +2071,7 @@ static void sit_add_v4_addrs(struct inet6_dev *idev)
 	struct inet6_ifaddr * ifp;
 	struct in6_addr addr;
 	struct net_device *dev;
-	struct net *net = idev->dev->nd_net;
+	struct net *net = dev_net(idev->dev);
 	int scope;
 
 	ASSERT_RTNL();
@@ -2261,7 +2261,7 @@ ipv6_inherit_linklocal(struct inet6_dev *idev, struct net_device *link_dev)
 static void ip6_tnl_add_linklocal(struct inet6_dev *idev)
 {
 	struct net_device *link_dev;
-	struct net *net = idev->dev->nd_net;
+	struct net *net = dev_net(idev->dev);
 
 	/* first try to inherit the link-local address from the link device */
 	if (idev->dev->iflink &&
@@ -2442,7 +2442,7 @@ static int addrconf_ifdown(struct net_device *dev, int how)
 {
 	struct inet6_dev *idev;
 	struct inet6_ifaddr *ifa, **bifa;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	int i;
 
 	ASSERT_RTNL();
@@ -2771,7 +2771,7 @@ static struct inet6_ifaddr *if6_get_first(struct seq_file *seq)
 	for (state->bucket = 0; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
 		ifa = inet6_addr_lst[state->bucket];
 
-		while (ifa && ifa->idev->dev->nd_net != net)
+		while (ifa && dev_net(ifa->idev->dev) != net)
 			ifa = ifa->lst_next;
 		if (ifa)
 			break;
@@ -2787,7 +2787,7 @@ static struct inet6_ifaddr *if6_get_next(struct seq_file *seq, struct inet6_ifad
 	ifa = ifa->lst_next;
 try_again:
 	if (ifa) {
-		if (ifa->idev->dev->nd_net != net) {
+		if (dev_net(ifa->idev->dev) != net) {
 			ifa = ifa->lst_next;
 			goto try_again;
 		}
@@ -2905,7 +2905,7 @@ int ipv6_chk_home_addr(struct net *net, struct in6_addr *addr)
 	u8 hash = ipv6_addr_hash(addr);
 	read_lock_bh(&addrconf_hash_lock);
 	for (ifp = inet6_addr_lst[hash]; ifp; ifp = ifp->lst_next) {
-		if (ifp->idev->dev->nd_net != net)
+		if (dev_net(ifp->idev->dev) != net)
 			continue;
 		if (ipv6_addr_cmp(&ifp->addr, addr) == 0 &&
 		    (ifp->flags & IFA_F_HOMEADDRESS)) {
@@ -3469,7 +3469,7 @@ errout:
 static void inet6_ifa_notify(int event, struct inet6_ifaddr *ifa)
 {
 	struct sk_buff *skb;
-	struct net *net = ifa->idev->dev->nd_net;
+	struct net *net = dev_net(ifa->idev->dev);
 	int err = -ENOBUFS;
 
 	skb = nlmsg_new(inet6_ifaddr_msgsize(), GFP_ATOMIC);
@@ -3675,7 +3675,7 @@ cont:
 void inet6_ifinfo_notify(int event, struct inet6_dev *idev)
 {
 	struct sk_buff *skb;
-	struct net *net = idev->dev->nd_net;
+	struct net *net = dev_net(idev->dev);
 	int err = -ENOBUFS;
 
 	skb = nlmsg_new(inet6_if_nlmsg_size(), GFP_ATOMIC);
@@ -3745,7 +3745,7 @@ static void inet6_prefix_notify(int event, struct inet6_dev *idev,
 			 struct prefix_info *pinfo)
 {
 	struct sk_buff *skb;
-	struct net *net = idev->dev->nd_net;
+	struct net *net = dev_net(idev->dev);
 	int err = -ENOBUFS;
 
 	skb = nlmsg_new(inet6_prefix_nlmsg_size(), GFP_ATOMIC);
@@ -4157,7 +4157,7 @@ static void addrconf_sysctl_register(struct inet6_dev *idev)
 			      NET_IPV6_NEIGH, "ipv6",
 			      &ndisc_ifinfo_sysctl_change,
 			      NULL);
-	__addrconf_sysctl_register(idev->dev->nd_net, idev->dev->name,
+	__addrconf_sysctl_register(dev_net(idev->dev), idev->dev->name,
 			idev->dev->ifindex, idev, &idev->cnf);
 }
 
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index 86332417b402..50857662e6b7 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -306,7 +306,7 @@ static inline void mip6_addr_swap(struct sk_buff *skb) {}
 void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
 		 struct net_device *dev)
 {
-	struct net *net = skb->dev->nd_net;
+	struct net *net = dev_net(skb->dev);
 	struct inet6_dev *idev = NULL;
 	struct ipv6hdr *hdr = ipv6_hdr(skb);
 	struct sock *sk;
@@ -507,7 +507,7 @@ EXPORT_SYMBOL(icmpv6_send);
 
 static void icmpv6_echo_reply(struct sk_buff *skb)
 {
-	struct net *net = skb->dev->nd_net;
+	struct net *net = dev_net(skb->dev);
 	struct sock *sk;
 	struct inet6_dev *idev;
 	struct ipv6_pinfo *np;
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index d34aa61353bb..556300f0eba5 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -402,7 +402,7 @@ int ip6_forward(struct sk_buff *skb)
 	struct dst_entry *dst = skb->dst;
 	struct ipv6hdr *hdr = ipv6_hdr(skb);
 	struct inet6_skb_parm *opt = IP6CB(skb);
-	struct net *net = dst->dev->nd_net;
+	struct net *net = dev_net(dst->dev);
 
 	if (ipv6_devconf.forwarding == 0)
 		goto error;
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 957ac7e9e929..0357de8e78c8 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -1400,7 +1400,7 @@ mld_scount(struct ifmcaddr6 *pmc, int type, int gdeleted, int sdeleted)
 
 static struct sk_buff *mld_newpack(struct net_device *dev, int size)
 {
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct sock *sk = net->ipv6.igmp_sk;
 	struct sk_buff *skb;
 	struct mld2_report *pmr;
@@ -1448,7 +1448,7 @@ static void mld_sendpack(struct sk_buff *skb)
 			      (struct mld2_report *)skb_transport_header(skb);
 	int payload_len, mldlen;
 	struct inet6_dev *idev = in6_dev_get(skb->dev);
-	struct net *net = skb->dev->nd_net;
+	struct net *net = dev_net(skb->dev);
 	int err;
 	struct flowi fl;
 
@@ -1762,7 +1762,7 @@ static void mld_send_cr(struct inet6_dev *idev)
 
 static void igmp6_send(struct in6_addr *addr, struct net_device *dev, int type)
 {
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct sock *sk = net->ipv6.igmp_sk;
 	struct inet6_dev *idev;
 	struct sk_buff *skb;
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 3f68a6eae7b2..79af57f586e8 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -447,7 +447,7 @@ static void __ndisc_send(struct net_device *dev,
 {
 	struct flowi fl;
 	struct dst_entry *dst;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct sock *sk = net->ipv6.ndisc_sk;
 	struct sk_buff *skb;
 	struct icmp6hdr *hdr;
@@ -539,7 +539,7 @@ static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,
 	};
 
 	/* for anycast or proxy, solicited_addr != src_addr */
-	ifp = ipv6_get_ifaddr(dev->nd_net, solicited_addr, dev, 1);
+	ifp = ipv6_get_ifaddr(dev_net(dev), solicited_addr, dev, 1);
 	if (ifp) {
 		src_addr = solicited_addr;
 		if (ifp->flags & IFA_F_OPTIMISTIC)
@@ -547,7 +547,7 @@ static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,
 		in6_ifa_put(ifp);
 	} else {
 		if (ipv6_dev_get_saddr(dev, daddr,
-				       inet6_sk(dev->nd_net->ipv6.ndisc_sk)->srcprefs,
+				       inet6_sk(dev_net(dev)->ipv6.ndisc_sk)->srcprefs,
 				       &tmpaddr))
 			return;
 		src_addr = &tmpaddr;
@@ -601,7 +601,7 @@ void ndisc_send_rs(struct net_device *dev, struct in6_addr *saddr,
 	 * suppress the inclusion of the sllao.
 	 */
 	if (send_sllao) {
-		struct inet6_ifaddr *ifp = ipv6_get_ifaddr(dev->nd_net, saddr,
+		struct inet6_ifaddr *ifp = ipv6_get_ifaddr(dev_net(dev), saddr,
 							   dev, 1);
 		if (ifp) {
 			if (ifp->flags & IFA_F_OPTIMISTIC)  {
@@ -639,7 +639,7 @@ static void ndisc_solicit(struct neighbour *neigh, struct sk_buff *skb)
 	struct in6_addr *target = (struct in6_addr *)&neigh->primary_key;
 	int probes = atomic_read(&neigh->probes);
 
-	if (skb && ipv6_chk_addr(dev->nd_net, &ipv6_hdr(skb)->saddr, dev, 1))
+	if (skb && ipv6_chk_addr(dev_net(dev), &ipv6_hdr(skb)->saddr, dev, 1))
 		saddr = &ipv6_hdr(skb)->saddr;
 
 	if ((probes -= neigh->parms->ucast_probes) < 0) {
@@ -727,7 +727,7 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 
 	inc = ipv6_addr_is_multicast(daddr);
 
-	ifp = ipv6_get_ifaddr(dev->nd_net, &msg->target, dev, 1);
+	ifp = ipv6_get_ifaddr(dev_net(dev), &msg->target, dev, 1);
 	if (ifp) {
 
 		if (ifp->flags & (IFA_F_TENTATIVE|IFA_F_OPTIMISTIC)) {
@@ -776,7 +776,7 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 		if (ipv6_chk_acast_addr(dev, &msg->target) ||
 		    (idev->cnf.forwarding &&
 		     (ipv6_devconf.proxy_ndp || idev->cnf.proxy_ndp) &&
-		     (pneigh = pneigh_lookup(&nd_tbl, dev->nd_net,
+		     (pneigh = pneigh_lookup(&nd_tbl, dev_net(dev),
 					     &msg->target, dev, 0)) != NULL)) {
 			if (!(NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED) &&
 			    skb->pkt_type != PACKET_HOST &&
@@ -886,7 +886,7 @@ static void ndisc_recv_na(struct sk_buff *skb)
 			return;
 		}
 	}
-	ifp = ipv6_get_ifaddr(dev->nd_net, &msg->target, dev, 1);
+	ifp = ipv6_get_ifaddr(dev_net(dev), &msg->target, dev, 1);
 	if (ifp) {
 		if (ifp->flags & IFA_F_TENTATIVE) {
 			addrconf_dad_failure(ifp);
@@ -918,7 +918,7 @@ static void ndisc_recv_na(struct sk_buff *skb)
 		 */
 		if (lladdr && !memcmp(lladdr, dev->dev_addr, dev->addr_len) &&
 		    ipv6_devconf.forwarding && ipv6_devconf.proxy_ndp &&
-		    pneigh_lookup(&nd_tbl, dev->nd_net, &msg->target, dev, 0)) {
+		    pneigh_lookup(&nd_tbl, dev_net(dev), &msg->target, dev, 0)) {
 			/* XXX: idev->cnf.prixy_ndp */
 			goto out;
 		}
@@ -1008,7 +1008,7 @@ static void ndisc_ra_useropt(struct sk_buff *ra, struct nd_opt_hdr *opt)
 	struct sk_buff *skb;
 	struct nlmsghdr *nlh;
 	struct nduseroptmsg *ndmsg;
-	struct net *net = ra->dev->nd_net;
+	struct net *net = dev_net(ra->dev);
 	int err;
 	int base_size = NLMSG_ALIGN(sizeof(struct nduseroptmsg)
 				    + (opt->nd_opt_len << 3));
@@ -1395,7 +1395,7 @@ void ndisc_send_redirect(struct sk_buff *skb, struct neighbour *neigh,
 			 struct in6_addr *target)
 {
 	struct net_device *dev = skb->dev;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct sock *sk = net->ipv6.ndisc_sk;
 	int len = sizeof(struct icmp6hdr) + 2 * sizeof(struct in6_addr);
 	struct sk_buff *buff;
@@ -1597,7 +1597,7 @@ int ndisc_rcv(struct sk_buff *skb)
 static int ndisc_netdev_event(struct notifier_block *this, unsigned long event, void *ptr)
 {
 	struct net_device *dev = ptr;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 
 	switch (event) {
 	case NETDEV_CHANGEADDR:
diff --git a/net/ipv6/netfilter/ip6_queue.c b/net/ipv6/netfilter/ip6_queue.c
index cc2f9afcf808..a6d30626b47c 100644
--- a/net/ipv6/netfilter/ip6_queue.c
+++ b/net/ipv6/netfilter/ip6_queue.c
@@ -484,7 +484,7 @@ ipq_rcv_dev_event(struct notifier_block *this,
 {
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* Drop any packets associated with the downed device */
diff --git a/net/ipv6/proc.c b/net/ipv6/proc.c
index 8a5be290c710..364dc332532c 100644
--- a/net/ipv6/proc.c
+++ b/net/ipv6/proc.c
@@ -214,7 +214,7 @@ int snmp6_register_dev(struct inet6_dev *idev)
 	if (!idev || !idev->dev)
 		return -EINVAL;
 
-	if (idev->dev->nd_net != &init_net)
+	if (dev_net(idev->dev) != &init_net)
 		return 0;
 
 	if (!proc_net_devsnmp6)
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index 548d0763f4d3..efb0047f6880 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -176,7 +176,7 @@ static int ipv6_raw_deliver(struct sk_buff *skb, int nexthdr)
 	if (sk == NULL)
 		goto out;
 
-	net = skb->dev->nd_net;
+	net = dev_net(skb->dev);
 	sk = __raw_v6_lookup(net, sk, nexthdr, daddr, saddr, IP6CB(skb)->iif);
 
 	while (sk) {
@@ -363,7 +363,7 @@ void raw6_icmp_error(struct sk_buff *skb, int nexthdr,
 	if (sk != NULL) {
 		saddr = &ipv6_hdr(skb)->saddr;
 		daddr = &ipv6_hdr(skb)->daddr;
-		net = skb->dev->nd_net;
+		net = dev_net(skb->dev);
 
 		while ((sk = __raw_v6_lookup(net, sk, nexthdr, saddr, daddr,
 						IP6CB(skb)->iif))) {
diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c
index f936d045a39d..4e1447634f36 100644
--- a/net/ipv6/reassembly.c
+++ b/net/ipv6/reassembly.c
@@ -600,7 +600,7 @@ static int ipv6_frag_rcv(struct sk_buff *skb)
 		return 1;
 	}
 
-	net = skb->dev->nd_net;
+	net = dev_net(skb->dev);
 	if (atomic_read(&net->ipv6.frags.mem) > net->ipv6.frags.high_thresh)
 		ip6_evictor(net, ip6_dst_idev(skb->dst));
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 06faa46920e1..65053fba8c1a 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -208,7 +208,7 @@ static void ip6_dst_ifdown(struct dst_entry *dst, struct net_device *dev,
 	struct rt6_info *rt = (struct rt6_info *)dst;
 	struct inet6_dev *idev = rt->rt6i_idev;
 	struct net_device *loopback_dev =
-		dev->nd_net->loopback_dev;
+		dev_net(dev)->loopback_dev;
 
 	if (dev != loopback_dev && idev != NULL && idev->dev == dev) {
 		struct inet6_dev *loopback_idev =
@@ -433,7 +433,7 @@ static struct rt6_info *rt6_select(struct fib6_node *fn, int oif, int strict)
 	RT6_TRACE("%s() => %p\n",
 		  __func__, match);
 
-	net = rt0->rt6i_dev->nd_net;
+	net = dev_net(rt0->rt6i_dev);
 	return (match ? match : net->ipv6.ip6_null_entry);
 }
 
@@ -441,7 +441,7 @@ static struct rt6_info *rt6_select(struct fib6_node *fn, int oif, int strict)
 int rt6_route_rcv(struct net_device *dev, u8 *opt, int len,
 		  struct in6_addr *gwaddr)
 {
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct route_info *rinfo = (struct route_info *) opt;
 	struct in6_addr prefix_buf, *prefix;
 	unsigned int pref;
@@ -607,7 +607,7 @@ static int __ip6_ins_rt(struct rt6_info *rt, struct nl_info *info)
 int ip6_ins_rt(struct rt6_info *rt)
 {
 	struct nl_info info = {
-		.nl_net = rt->rt6i_dev->nd_net,
+		.nl_net = dev_net(rt->rt6i_dev),
 	};
 	return __ip6_ins_rt(rt, &info);
 }
@@ -745,7 +745,7 @@ static struct rt6_info *ip6_pol_route_input(struct net *net, struct fib6_table *
 void ip6_route_input(struct sk_buff *skb)
 {
 	struct ipv6hdr *iph = ipv6_hdr(skb);
-	struct net *net = skb->dev->nd_net;
+	struct net *net = dev_net(skb->dev);
 	int flags = RT6_LOOKUP_F_HAS_SADDR;
 	struct flowi fl = {
 		.iif = skb->dev->ifindex,
@@ -928,7 +928,7 @@ struct dst_entry *icmp6_dst_alloc(struct net_device *dev,
 {
 	struct rt6_info *rt;
 	struct inet6_dev *idev = in6_dev_get(dev);
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 
 	if (unlikely(idev == NULL))
 		return NULL;
@@ -1252,7 +1252,7 @@ install_route:
 	rt->rt6i_idev = idev;
 	rt->rt6i_table = table;
 
-	cfg->fc_nlinfo.nl_net = dev->nd_net;
+	cfg->fc_nlinfo.nl_net = dev_net(dev);
 
 	return __ip6_ins_rt(rt, &cfg->fc_nlinfo);
 
@@ -1270,7 +1270,7 @@ static int __ip6_del_rt(struct rt6_info *rt, struct nl_info *info)
 {
 	int err;
 	struct fib6_table *table;
-	struct net *net = rt->rt6i_dev->nd_net;
+	struct net *net = dev_net(rt->rt6i_dev);
 
 	if (rt == net->ipv6.ip6_null_entry)
 		return -ENOENT;
@@ -1289,7 +1289,7 @@ static int __ip6_del_rt(struct rt6_info *rt, struct nl_info *info)
 int ip6_del_rt(struct rt6_info *rt)
 {
 	struct nl_info info = {
-		.nl_net = rt->rt6i_dev->nd_net,
+		.nl_net = dev_net(rt->rt6i_dev),
 	};
 	return __ip6_del_rt(rt, &info);
 }
@@ -1401,7 +1401,7 @@ static struct rt6_info *ip6_route_redirect(struct in6_addr *dest,
 					   struct net_device *dev)
 {
 	int flags = RT6_LOOKUP_F_HAS_SADDR;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	struct ip6rd_flowi rdfl = {
 		.fl = {
 			.oif = dev->ifindex,
@@ -1428,7 +1428,7 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *src,
 {
 	struct rt6_info *rt, *nrt = NULL;
 	struct netevent_redirect netevent;
-	struct net *net = neigh->dev->nd_net;
+	struct net *net = dev_net(neigh->dev);
 
 	rt = ip6_route_redirect(dest, src, saddr, neigh->dev);
 
@@ -1477,7 +1477,7 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *src,
 	nrt->rt6i_nexthop = neigh_clone(neigh);
 	/* Reset pmtu, it may be better */
 	nrt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(neigh->dev);
-	nrt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(neigh->dev->nd_net,
+	nrt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(dev_net(neigh->dev),
 							dst_mtu(&nrt->u.dst));
 
 	if (ip6_ins_rt(nrt))
@@ -1506,7 +1506,7 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 			struct net_device *dev, u32 pmtu)
 {
 	struct rt6_info *rt, *nrt;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 	int allfrag = 0;
 
 	rt = rt6_lookup(net, daddr, saddr, dev->ifindex, 0);
@@ -1583,7 +1583,7 @@ out:
 
 static struct rt6_info * ip6_rt_copy(struct rt6_info *ort)
 {
-	struct net *net = ort->rt6i_dev->nd_net;
+	struct net *net = dev_net(ort->rt6i_dev);
 	struct rt6_info *rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops);
 
 	if (rt) {
@@ -1682,7 +1682,7 @@ struct rt6_info *rt6_get_dflt_router(struct in6_addr *addr, struct net_device *d
 	struct rt6_info *rt;
 	struct fib6_table *table;
 
-	table = fib6_get_table(dev->nd_net, RT6_TABLE_DFLT);
+	table = fib6_get_table(dev_net(dev), RT6_TABLE_DFLT);
 	if (table == NULL)
 		return NULL;
 
@@ -1713,7 +1713,7 @@ struct rt6_info *rt6_add_dflt_router(struct in6_addr *gwaddr,
 				  RTF_UP | RTF_EXPIRES | RTF_PREF(pref),
 		.fc_nlinfo.pid = 0,
 		.fc_nlinfo.nlh = NULL,
-		.fc_nlinfo.nl_net = dev->nd_net,
+		.fc_nlinfo.nl_net = dev_net(dev),
 	};
 
 	ipv6_addr_copy(&cfg.fc_gateway, gwaddr);
@@ -1862,7 +1862,7 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
 				    const struct in6_addr *addr,
 				    int anycast)
 {
-	struct net *net = idev->dev->nd_net;
+	struct net *net = dev_net(idev->dev);
 	struct rt6_info *rt = ip6_dst_alloc(net->ipv6.ip6_dst_ops);
 
 	if (rt == NULL)
@@ -1939,7 +1939,7 @@ static int rt6_mtu_change_route(struct rt6_info *rt, void *p_arg)
 {
 	struct rt6_mtu_change_arg *arg = (struct rt6_mtu_change_arg *) p_arg;
 	struct inet6_dev *idev;
-	struct net *net = arg->dev->nd_net;
+	struct net *net = dev_net(arg->dev);
 
 	/* In IPv6 pmtu discovery is not optional,
 	   so that RTAX_MTU lock cannot disable it.
@@ -1983,7 +1983,7 @@ void rt6_mtu_change(struct net_device *dev, unsigned mtu)
 		.mtu = mtu,
 	};
 
-	fib6_clean_all(dev->nd_net, rt6_mtu_change_route, 0, &arg);
+	fib6_clean_all(dev_net(dev), rt6_mtu_change_route, 0, &arg);
 }
 
 static const struct nla_policy rtm_ipv6_policy[RTA_MAX+1] = {
@@ -2321,7 +2321,7 @@ static int ip6_route_dev_notify(struct notifier_block *this,
 				unsigned long event, void *data)
 {
 	struct net_device *dev = (struct net_device *)data;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 
 	if (event == NETDEV_REGISTER && (dev->flags & IFF_LOOPBACK)) {
 		net->ipv6.ip6_null_entry->u.dst.dev = dev;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 8dd72966ff78..086deffff9c9 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -321,7 +321,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 	struct tcp_sock *tp;
 	__u32 seq;
 
-	sk = inet6_lookup(skb->dev->nd_net, &tcp_hashinfo, &hdr->daddr,
+	sk = inet6_lookup(dev_net(skb->dev), &tcp_hashinfo, &hdr->daddr,
 			th->dest, &hdr->saddr, th->source, skb->dev->ifindex);
 
 	if (sk == NULL) {
@@ -988,7 +988,7 @@ static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb)
 	struct tcphdr *th = tcp_hdr(skb), *t1;
 	struct sk_buff *buff;
 	struct flowi fl;
-	struct net *net = skb->dst->dev->nd_net;
+	struct net *net = dev_net(skb->dst->dev);
 	struct sock *ctl_sk = net->ipv6.tcp_sk;
 	unsigned int tot_len = sizeof(*th);
 #ifdef CONFIG_TCP_MD5SIG
@@ -1093,7 +1093,7 @@ static void tcp_v6_send_ack(struct tcp_timewait_sock *tw,
 	struct tcphdr *th = tcp_hdr(skb), *t1;
 	struct sk_buff *buff;
 	struct flowi fl;
-	struct net *net = skb->dev->nd_net;
+	struct net *net = dev_net(skb->dev);
 	struct sock *ctl_sk = net->ipv6.tcp_sk;
 	unsigned int tot_len = sizeof(struct tcphdr);
 	__be32 *topt;
@@ -1739,7 +1739,7 @@ static int tcp_v6_rcv(struct sk_buff *skb)
 	TCP_SKB_CB(skb)->flags = ipv6_get_dsfield(ipv6_hdr(skb));
 	TCP_SKB_CB(skb)->sacked = 0;
 
-	sk = __inet6_lookup(skb->dev->nd_net, &tcp_hashinfo,
+	sk = __inet6_lookup(dev_net(skb->dev), &tcp_hashinfo,
 			&ipv6_hdr(skb)->saddr, th->source,
 			&ipv6_hdr(skb)->daddr, ntohs(th->dest),
 			inet6_iif(skb));
@@ -1822,7 +1822,7 @@ do_time_wait:
 	{
 		struct sock *sk2;
 
-		sk2 = inet6_lookup_listener(skb->dev->nd_net, &tcp_hashinfo,
+		sk2 = inet6_lookup_listener(dev_net(skb->dev), &tcp_hashinfo,
 					    &ipv6_hdr(skb)->daddr,
 					    ntohs(th->dest), inet6_iif(skb));
 		if (sk2 != NULL) {
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 593d3efadaf9..6683c04b427e 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -235,7 +235,7 @@ void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 	struct sock *sk;
 	int err;
 
-	sk = __udp6_lib_lookup(skb->dev->nd_net, daddr, uh->dest,
+	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
 			       saddr, uh->source, inet6_iif(skb), udptable);
 	if (sk == NULL)
 		return;
@@ -483,7 +483,7 @@ int __udp6_lib_rcv(struct sk_buff *skb, struct hlist_head udptable[],
 	 * check socket cache ... must talk to Alan about his plans
 	 * for sock caches... i'll skip this for now.
 	 */
-	sk = __udp6_lib_lookup(skb->dev->nd_net, saddr, uh->source,
+	sk = __udp6_lib_lookup(dev_net(skb->dev), saddr, uh->source,
 			       daddr, uh->dest, inet6_iif(skb), udptable);
 
 	if (sk == NULL) {
diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c
index d92d1fceb8cf..8f1e0543b3c4 100644
--- a/net/ipv6/xfrm6_policy.c
+++ b/net/ipv6/xfrm6_policy.c
@@ -247,7 +247,7 @@ static void xfrm6_dst_ifdown(struct dst_entry *dst, struct net_device *dev,
 	xdst = (struct xfrm_dst *)dst;
 	if (xdst->u.rt6.rt6i_idev->dev == dev) {
 		struct inet6_dev *loopback_idev =
-			in6_dev_get(dev->nd_net->loopback_dev);
+			in6_dev_get(dev_net(dev)->loopback_dev);
 		BUG_ON(!loopback_idev);
 
 		do {
diff --git a/net/ipx/af_ipx.c b/net/ipx/af_ipx.c
index c76a9523091b..81ae8735f5e3 100644
--- a/net/ipx/af_ipx.c
+++ b/net/ipx/af_ipx.c
@@ -335,7 +335,7 @@ static int ipxitf_device_event(struct notifier_block *notifier,
 	struct net_device *dev = ptr;
 	struct ipx_interface *i, *tmp;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event != NETDEV_DOWN && event != NETDEV_UP)
@@ -1636,7 +1636,7 @@ static int ipx_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_ty
 	u16 ipx_pktsize;
 	int rc = 0;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	/* Not ours */
diff --git a/net/irda/irlap_frame.c b/net/irda/irlap_frame.c
index a38b231c8689..90894534f3cc 100644
--- a/net/irda/irlap_frame.c
+++ b/net/irda/irlap_frame.c
@@ -1326,7 +1326,7 @@ int irlap_driver_rcv(struct sk_buff *skb, struct net_device *dev,
 	int command;
 	__u8 control;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto out;
 
 	/* FIXME: should we get our own field? */
diff --git a/net/llc/llc_input.c b/net/llc/llc_input.c
index b9143d2a04e1..a69c5c427fe3 100644
--- a/net/llc/llc_input.c
+++ b/net/llc/llc_input.c
@@ -146,7 +146,7 @@ int llc_rcv(struct sk_buff *skb, struct net_device *dev,
 	int (*rcv)(struct sk_buff *, struct net_device *,
 		   struct packet_type *, struct net_device *);
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	/*
diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index ec05684c56d7..292fa28146fb 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -168,7 +168,7 @@ int nf_hook_slow(int pf, unsigned int hook, struct sk_buff *skb,
 #ifdef CONFIG_NET_NS
 	struct net *net;
 
-	net = indev == NULL ? outdev->nd_net : indev->nd_net;
+	net = indev == NULL ? dev_net(outdev) : dev_net(indev);
 	if (net != &init_net)
 		return 1;
 #endif
diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c
index 012cb6910820..81fb048add88 100644
--- a/net/netfilter/nfnetlink_queue.c
+++ b/net/netfilter/nfnetlink_queue.c
@@ -557,7 +557,7 @@ nfqnl_rcv_dev_event(struct notifier_block *this,
 {
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* Drop any packets associated with the downed device */
diff --git a/net/netlabel/netlabel_unlabeled.c b/net/netlabel/netlabel_unlabeled.c
index 4478f2f6079d..a547c6320eb3 100644
--- a/net/netlabel/netlabel_unlabeled.c
+++ b/net/netlabel/netlabel_unlabeled.c
@@ -954,7 +954,7 @@ static int netlbl_unlhsh_netdev_handler(struct notifier_block *this,
 	struct net_device *dev = ptr;
 	struct netlbl_unlhsh_iface *iface = NULL;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	/* XXX - should this be a check for NETDEV_DOWN or _UNREGISTER? */
diff --git a/net/netrom/af_netrom.c b/net/netrom/af_netrom.c
index 972250c974f1..a270ebf9f765 100644
--- a/net/netrom/af_netrom.c
+++ b/net/netrom/af_netrom.c
@@ -106,7 +106,7 @@ static int nr_device_event(struct notifier_block *this, unsigned long event, voi
 {
 	struct net_device *dev = (struct net_device *)ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event != NETDEV_DOWN)
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index a56ed2120e07..baa290d3444a 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -263,7 +263,7 @@ static int packet_rcv_spkt(struct sk_buff *skb, struct net_device *dev,  struct
 	if (skb->pkt_type == PACKET_LOOPBACK)
 		goto out;
 
-	if (dev->nd_net != sk->sk_net)
+	if (dev_net(dev) != sk->sk_net)
 		goto out;
 
 	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
@@ -451,7 +451,7 @@ static int packet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet
 	sk = pt->af_packet_priv;
 	po = pkt_sk(sk);
 
-	if (dev->nd_net != sk->sk_net)
+	if (dev_net(dev) != sk->sk_net)
 		goto drop;
 
 	skb->dev = dev;
@@ -568,7 +568,7 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct packe
 	sk = pt->af_packet_priv;
 	po = pkt_sk(sk);
 
-	if (dev->nd_net != sk->sk_net)
+	if (dev_net(dev) != sk->sk_net)
 		goto drop;
 
 	if (dev->header_ops) {
@@ -1450,7 +1450,7 @@ static int packet_notifier(struct notifier_block *this, unsigned long msg, void
 	struct sock *sk;
 	struct hlist_node *node;
 	struct net_device *dev = data;
-	struct net *net = dev->nd_net;
+	struct net *net = dev_net(dev);
 
 	read_lock(&net->packet.sklist_lock);
 	sk_for_each(sk, node, &net->packet.sklist) {
diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c
index 4a31a81059ab..1a7f143cf741 100644
--- a/net/rose/af_rose.c
+++ b/net/rose/af_rose.c
@@ -197,7 +197,7 @@ static int rose_device_event(struct notifier_block *this, unsigned long event,
 {
 	struct net_device *dev = (struct net_device *)ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event != NETDEV_DOWN)
diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c
index beea2fb18b15..2faa0d8839eb 100644
--- a/net/sctp/protocol.c
+++ b/net/sctp/protocol.c
@@ -630,7 +630,7 @@ static int sctp_inetaddr_event(struct notifier_block *this, unsigned long ev,
 	struct sctp_sockaddr_entry *temp;
 	int found = 0;
 
-	if (ifa->ifa_dev->dev->nd_net != &init_net)
+	if (dev_net(ifa->ifa_dev->dev) != &init_net)
 		return NOTIFY_DONE;
 
 	switch (ev) {
diff --git a/net/tipc/eth_media.c b/net/tipc/eth_media.c
index 3bbef2ab22ae..9cd35eec3e7f 100644
--- a/net/tipc/eth_media.c
+++ b/net/tipc/eth_media.c
@@ -101,7 +101,7 @@ static int recv_msg(struct sk_buff *buf, struct net_device *dev,
 	struct eth_bearer *eb_ptr = (struct eth_bearer *)pt->af_packet_priv;
 	u32 size;
 
-	if (dev->nd_net != &init_net) {
+	if (dev_net(dev) != &init_net) {
 		kfree_skb(buf);
 		return 0;
 	}
@@ -198,7 +198,7 @@ static int recv_notification(struct notifier_block *nb, unsigned long evt,
 	struct eth_bearer *eb_ptr = &eth_bearers[0];
 	struct eth_bearer *stop = &eth_bearers[MAX_ETH_BEARERS];
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	while ((eb_ptr->dev != dev)) {
diff --git a/net/wireless/wext.c b/net/wireless/wext.c
index 2c569b63e7d8..947188a5b937 100644
--- a/net/wireless/wext.c
+++ b/net/wireless/wext.c
@@ -1157,7 +1157,7 @@ static void rtmsg_iwinfo(struct net_device *dev, char *event, int event_len)
 	struct sk_buff *skb;
 	int err;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return;
 
 	skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC);
diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c
index 339ca4a8e89e..7a46ea73fe2d 100644
--- a/net/x25/af_x25.c
+++ b/net/x25/af_x25.c
@@ -191,7 +191,7 @@ static int x25_device_event(struct notifier_block *this, unsigned long event,
 	struct net_device *dev = ptr;
 	struct x25_neigh *nb;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (dev->type == ARPHRD_X25
diff --git a/net/x25/x25_dev.c b/net/x25/x25_dev.c
index f0679d283110..3ff206c0ae94 100644
--- a/net/x25/x25_dev.c
+++ b/net/x25/x25_dev.c
@@ -95,7 +95,7 @@ int x25_lapb_receive_frame(struct sk_buff *skb, struct net_device *dev,
 	struct sk_buff *nskb;
 	struct x25_neigh *nb;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		goto drop;
 
 	nskb = skb_copy(skb, GFP_ATOMIC);
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 8e588f20c60c..15d73e47cc2c 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -2079,7 +2079,7 @@ static int stale_bundle(struct dst_entry *dst)
 void xfrm_dst_ifdown(struct dst_entry *dst, struct net_device *dev)
 {
 	while ((dst = dst->child) && dst->xfrm && dst->dev == dev) {
-		dst->dev = dev->nd_net->loopback_dev;
+		dst->dev = dev_net(dev)->loopback_dev;
 		dev_hold(dst->dev);
 		dev_put(dev);
 	}
@@ -2350,7 +2350,7 @@ static int xfrm_dev_event(struct notifier_block *this, unsigned long event, void
 {
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	switch (event) {
diff --git a/security/selinux/netif.c b/security/selinux/netif.c
index 013d3117a86b..9c8a82aa8baf 100644
--- a/security/selinux/netif.c
+++ b/security/selinux/netif.c
@@ -281,7 +281,7 @@ static int sel_netif_netdev_notifier_handler(struct notifier_block *this,
 {
 	struct net_device *dev = ptr;
 
-	if (dev->nd_net != &init_net)
+	if (dev_net(dev) != &init_net)
 		return NOTIFY_DONE;
 
 	if (event == NETDEV_DOWN)
-- 
cgit v1.2.3


From 3b1e0a655f8eba44ab1ee2a1068d169ccfb853b9 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Wed, 26 Mar 2008 02:26:21 +0900
Subject: [NET] NETNS: Omit sock->sk_net without CONFIG_NET_NS.

Introduce per-sock inlines: sock_net(), sock_net_set()
and per-inet_timewait_sock inlines: twsk_net(), twsk_net_set().
Without CONFIG_NET_NS, no namespace other than &init_net exists.
Let's explicitly define them to help compiler optimizations.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/ipv6.h             |  4 ++--
 include/net/inet_hashtables.h    |  8 ++++----
 include/net/inet_timewait_sock.h | 18 ++++++++++++++++++
 include/net/route.h              |  4 ++--
 include/net/sock.h               | 24 ++++++++++++++++++++++--
 net/atm/svc.c                    |  2 +-
 net/ax25/af_ax25.c               |  2 +-
 net/bluetooth/l2cap.c            |  2 +-
 net/bluetooth/rfcomm/sock.c      |  2 +-
 net/bluetooth/sco.c              |  2 +-
 net/bridge/br_netlink.c          |  4 ++--
 net/core/fib_rules.c             |  6 +++---
 net/core/neighbour.c             | 10 +++++-----
 net/core/rtnetlink.c             | 12 ++++++------
 net/core/sock.c                  | 10 +++++-----
 net/decnet/af_decnet.c           |  2 +-
 net/decnet/dn_dev.c              |  6 +++---
 net/decnet/dn_fib.c              |  4 ++--
 net/decnet/dn_route.c            |  4 ++--
 net/decnet/dn_table.c            |  2 +-
 net/ipv4/af_inet.c               |  6 +++---
 net/ipv4/devinet.c               |  6 +++---
 net/ipv4/fib_frontend.c          |  8 ++++----
 net/ipv4/fib_rules.c             |  2 +-
 net/ipv4/igmp.c                  | 14 +++++++-------
 net/ipv4/inet_connection_sock.c  |  4 ++--
 net/ipv4/inet_hashtables.c       |  8 ++++----
 net/ipv4/inet_timewait_sock.c    |  2 +-
 net/ipv4/ip_input.c              |  2 +-
 net/ipv4/ip_output.c             |  4 ++--
 net/ipv4/ip_sockglue.c           |  6 +++---
 net/ipv4/ipmr.c                  |  4 ++--
 net/ipv4/netfilter/arp_tables.c  | 16 ++++++++--------
 net/ipv4/netfilter/ip_tables.c   | 16 ++++++++--------
 net/ipv4/raw.c                   | 12 ++++++------
 net/ipv4/route.c                 |  4 ++--
 net/ipv4/tcp_ipv4.c              | 14 +++++++-------
 net/ipv4/udp.c                   | 16 ++++++++--------
 net/ipv6/addrconf.c              | 10 +++++-----
 net/ipv6/addrlabel.c             |  6 +++---
 net/ipv6/af_inet6.c              |  4 ++--
 net/ipv6/fib6_rules.c            |  2 +-
 net/ipv6/icmp.c                  |  2 +-
 net/ipv6/inet6_hashtables.c      |  4 ++--
 net/ipv6/ip6_fib.c               |  2 +-
 net/ipv6/ip6_output.c            |  2 +-
 net/ipv6/ipv6_sockglue.c         |  2 +-
 net/ipv6/mcast.c                 | 12 ++++++------
 net/ipv6/netfilter/ip6_tables.c  | 16 ++++++++--------
 net/ipv6/raw.c                   |  6 +++---
 net/ipv6/route.c                 |  4 ++--
 net/ipv6/tcp_ipv6.c              |  2 +-
 net/ipv6/udp.c                   |  4 ++--
 net/irda/af_irda.c               |  2 +-
 net/llc/llc_conn.c               |  2 +-
 net/netfilter/nf_sockopt.c       |  2 +-
 net/netlink/af_netlink.c         | 30 +++++++++++++++---------------
 net/netrom/af_netrom.c           |  2 +-
 net/packet/af_packet.c           | 28 ++++++++++++++--------------
 net/rose/af_rose.c               |  2 +-
 net/sched/act_api.c              |  4 ++--
 net/sched/cls_api.c              |  4 ++--
 net/sched/sch_api.c              | 10 +++++-----
 net/sctp/ipv6.c                  |  2 +-
 net/sctp/protocol.c              |  2 +-
 net/socket.c                     |  4 ++--
 net/tipc/socket.c                |  2 +-
 net/unix/af_unix.c               | 20 ++++++++++----------
 net/x25/af_x25.c                 |  2 +-
 69 files changed, 253 insertions(+), 215 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index c9ba0da16ce9..b90d3d461d4e 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -481,7 +481,7 @@ static inline struct raw6_sock *raw6_sk(const struct sock *sk)
 #endif /* defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) */
 
 #define INET6_MATCH(__sk, __net, __hash, __saddr, __daddr, __ports, __dif)\
-	(((__sk)->sk_hash == (__hash)) && ((__sk)->sk_net == (__net))	&& \
+	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&& \
 	 ((*((__portpair *)&(inet_sk(__sk)->dport))) == (__ports))  	&& \
 	 ((__sk)->sk_family		== AF_INET6)		&& \
 	 ipv6_addr_equal(&inet6_sk(__sk)->daddr, (__saddr))	&& \
@@ -489,7 +489,7 @@ static inline struct raw6_sock *raw6_sk(const struct sock *sk)
 	 (!((__sk)->sk_bound_dev_if) || ((__sk)->sk_bound_dev_if == (__dif))))
 
 #define INET6_TW_MATCH(__sk, __net, __hash, __saddr, __daddr, __ports, __dif) \
-	(((__sk)->sk_hash == (__hash)) && ((__sk)->sk_net == (__net))	&& \
+	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&& \
 	 (*((__portpair *)&(inet_twsk(__sk)->tw_dport)) == (__ports))	&& \
 	 ((__sk)->sk_family	       == PF_INET6)			&& \
 	 (ipv6_addr_equal(&inet6_twsk(__sk)->tw_v6_daddr, (__saddr)))	&& \
diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index d99c1ba2ece0..5525227c5e92 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -314,25 +314,25 @@ typedef __u64 __bitwise __addrpair;
 				   ((__force __u64)(__be32)(__saddr)));
 #endif /* __BIG_ENDIAN */
 #define INET_MATCH(__sk, __net, __hash, __cookie, __saddr, __daddr, __ports, __dif)\
-	(((__sk)->sk_hash == (__hash)) && ((__sk)->sk_net == (__net))	&&	\
+	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
 	 ((*((__addrpair *)&(inet_sk(__sk)->daddr))) == (__cookie))	&&	\
 	 ((*((__portpair *)&(inet_sk(__sk)->dport))) == (__ports))	&&	\
 	 (!((__sk)->sk_bound_dev_if) || ((__sk)->sk_bound_dev_if == (__dif))))
 #define INET_TW_MATCH(__sk, __net, __hash, __cookie, __saddr, __daddr, __ports, __dif)\
-	(((__sk)->sk_hash == (__hash)) && ((__sk)->sk_net == (__net))	&&	\
+	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
 	 ((*((__addrpair *)&(inet_twsk(__sk)->tw_daddr))) == (__cookie)) &&	\
 	 ((*((__portpair *)&(inet_twsk(__sk)->tw_dport))) == (__ports)) &&	\
 	 (!((__sk)->sk_bound_dev_if) || ((__sk)->sk_bound_dev_if == (__dif))))
 #else /* 32-bit arch */
 #define INET_ADDR_COOKIE(__name, __saddr, __daddr)
 #define INET_MATCH(__sk, __net, __hash, __cookie, __saddr, __daddr, __ports, __dif)	\
-	(((__sk)->sk_hash == (__hash)) && ((__sk)->sk_net == (__net))	&&	\
+	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
 	 (inet_sk(__sk)->daddr		== (__saddr))		&&	\
 	 (inet_sk(__sk)->rcv_saddr	== (__daddr))		&&	\
 	 ((*((__portpair *)&(inet_sk(__sk)->dport))) == (__ports))	&&	\
 	 (!((__sk)->sk_bound_dev_if) || ((__sk)->sk_bound_dev_if == (__dif))))
 #define INET_TW_MATCH(__sk, __net, __hash,__cookie, __saddr, __daddr, __ports, __dif)	\
-	(((__sk)->sk_hash == (__hash)) && ((__sk)->sk_net == (__net))	&&	\
+	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
 	 (inet_twsk(__sk)->tw_daddr	== (__saddr))		&&	\
 	 (inet_twsk(__sk)->tw_rcv_saddr	== (__daddr))		&&	\
 	 ((*((__portpair *)&(inet_twsk(__sk)->tw_dport))) == (__ports)) &&	\
diff --git a/include/net/inet_timewait_sock.h b/include/net/inet_timewait_sock.h
index 296547bfb0b7..07fe0d1a4f03 100644
--- a/include/net/inet_timewait_sock.h
+++ b/include/net/inet_timewait_sock.h
@@ -207,4 +207,22 @@ extern void inet_twsk_schedule(struct inet_timewait_sock *tw,
 			       const int timeo, const int timewait_len);
 extern void inet_twsk_deschedule(struct inet_timewait_sock *tw,
 				 struct inet_timewait_death_row *twdr);
+
+static inline
+struct net *twsk_net(const struct inet_timewait_sock *twsk)
+{
+#ifdef CONFIG_NET_NS
+	return twsk->tw_net;
+#else
+	return &init_net;
+#endif
+}
+
+static inline
+void twsk_net_set(struct inet_timewait_sock *twsk, const struct net *net)
+{
+#ifdef CONFIG_NET_NS
+	twsk->tw_net = net;
+#endif
+}
 #endif	/* _INET_TIMEWAIT_SOCK_ */
diff --git a/include/net/route.h b/include/net/route.h
index 28dba925663c..c6338802e8f1 100644
--- a/include/net/route.h
+++ b/include/net/route.h
@@ -160,7 +160,7 @@ static inline int ip_route_connect(struct rtable **rp, __be32 dst,
 					 .dport = dport } } };
 
 	int err;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	if (!dst || !src) {
 		err = __ip_route_output_key(net, rp, &fl);
 		if (err)
@@ -188,7 +188,7 @@ static inline int ip_route_newports(struct rtable **rp, u8 protocol,
 		ip_rt_put(*rp);
 		*rp = NULL;
 		security_sk_classify_flow(sk, &fl);
-		return ip_route_output_flow(sk->sk_net, rp, &fl, sk, 0);
+		return ip_route_output_flow(sock_net(sk), rp, &fl, sk, 0);
 	}
 	return 0;
 }
diff --git a/include/net/sock.h b/include/net/sock.h
index b433b1ed203d..7e0d4a0c4d12 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -126,7 +126,9 @@ struct sock_common {
 	atomic_t		skc_refcnt;
 	unsigned int		skc_hash;
 	struct proto		*skc_prot;
+#ifdef CONFIG_NET_NS
 	struct net	 	*skc_net;
+#endif
 };
 
 /**
@@ -1345,6 +1347,24 @@ static inline void sk_eat_skb(struct sock *sk, struct sk_buff *skb, int copied_e
 }
 #endif
 
+static inline
+struct net *sock_net(const struct sock *sk)
+{
+#ifdef CONFIG_NET_NS
+	return sk->sk_net;
+#else
+	return &init_net;
+#endif
+}
+
+static inline
+void sock_net_set(struct sock *sk, const struct net *net)
+{
+#ifdef CONFIG_NET_NS
+	sk->sk_net = net;
+#endif
+}
+
 /*
  * Kernel sockets, f.e. rtnl or icmp_socket, are a part of a namespace.
  * They should not hold a referrence to a namespace in order to allow
@@ -1353,8 +1373,8 @@ static inline void sk_eat_skb(struct sock *sk, struct sk_buff *skb, int copied_e
  */
 static inline void sk_change_net(struct sock *sk, struct net *net)
 {
-	put_net(sk->sk_net);
-	sk->sk_net = net;
+	put_net(sock_net(sk));
+	sock_net_set(sk, net);
 }
 
 extern void sock_enable_timestamp(struct sock *sk);
diff --git a/net/atm/svc.c b/net/atm/svc.c
index daf9a48a7db0..de1e4f2f3a43 100644
--- a/net/atm/svc.c
+++ b/net/atm/svc.c
@@ -326,7 +326,7 @@ static int svc_accept(struct socket *sock,struct socket *newsock,int flags)
 
 	lock_sock(sk);
 
-	error = svc_create(sk->sk_net, newsock,0);
+	error = svc_create(sock_net(sk), newsock,0);
 	if (error)
 		goto out;
 
diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c
index ee9dd83e7561..2712544cf0ca 100644
--- a/net/ax25/af_ax25.c
+++ b/net/ax25/af_ax25.c
@@ -869,7 +869,7 @@ struct sock *ax25_make_new(struct sock *osk, struct ax25_dev *ax25_dev)
 	struct sock *sk;
 	ax25_cb *ax25, *oax25;
 
-	sk = sk_alloc(osk->sk_net, PF_AX25, GFP_ATOMIC,	osk->sk_prot);
+	sk = sk_alloc(sock_net(osk), PF_AX25, GFP_ATOMIC,	osk->sk_prot);
 	if (sk == NULL)
 		return NULL;
 
diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c
index 34f8bf98bc05..6b995ac832f5 100644
--- a/net/bluetooth/l2cap.c
+++ b/net/bluetooth/l2cap.c
@@ -1499,7 +1499,7 @@ static inline int l2cap_connect_req(struct l2cap_conn *conn, struct l2cap_cmd_hd
 		goto response;
 	}
 
-	sk = l2cap_sock_alloc(parent->sk_net, NULL, BTPROTO_L2CAP, GFP_ATOMIC);
+	sk = l2cap_sock_alloc(sock_net(parent), NULL, BTPROTO_L2CAP, GFP_ATOMIC);
 	if (!sk)
 		goto response;
 
diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c
index c46d51035e77..c103fa02893b 100644
--- a/net/bluetooth/rfcomm/sock.c
+++ b/net/bluetooth/rfcomm/sock.c
@@ -868,7 +868,7 @@ int rfcomm_connect_ind(struct rfcomm_session *s, u8 channel, struct rfcomm_dlc *
 		goto done;
 	}
 
-	sk = rfcomm_sock_alloc(parent->sk_net, NULL, BTPROTO_RFCOMM, GFP_ATOMIC);
+	sk = rfcomm_sock_alloc(sock_net(parent), NULL, BTPROTO_RFCOMM, GFP_ATOMIC);
 	if (!sk)
 		goto done;
 
diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c
index b91d3c81a73c..2a5953b4405d 100644
--- a/net/bluetooth/sco.c
+++ b/net/bluetooth/sco.c
@@ -803,7 +803,7 @@ static void sco_conn_ready(struct sco_conn *conn)
 
 		bh_lock_sock(parent);
 
-		sk = sco_sock_alloc(parent->sk_net, NULL, BTPROTO_SCO, GFP_ATOMIC);
+		sk = sco_sock_alloc(sock_net(parent), NULL, BTPROTO_SCO, GFP_ATOMIC);
 		if (!sk) {
 			bh_unlock_sock(parent);
 			goto done;
diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c
index f5d69336d97b..f155e6ce8a21 100644
--- a/net/bridge/br_netlink.c
+++ b/net/bridge/br_netlink.c
@@ -108,7 +108,7 @@ errout:
  */
 static int br_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct net_device *dev;
 	int idx;
 
@@ -140,7 +140,7 @@ skip:
  */
 static int br_rtm_setlink(struct sk_buff *skb,  struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ifinfomsg *ifm;
 	struct nlattr *protinfo;
 	struct net_device *dev;
diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
index 942be93a2eb0..540c07283e31 100644
--- a/net/core/fib_rules.c
+++ b/net/core/fib_rules.c
@@ -214,7 +214,7 @@ errout:
 
 static int fib_nl_newrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct fib_rule_hdr *frh = nlmsg_data(nlh);
 	struct fib_rules_ops *ops = NULL;
 	struct fib_rule *rule, *r, *last = NULL;
@@ -352,7 +352,7 @@ errout:
 
 static int fib_nl_delrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct fib_rule_hdr *frh = nlmsg_data(nlh);
 	struct fib_rules_ops *ops = NULL;
 	struct fib_rule *rule, *tmp;
@@ -534,7 +534,7 @@ skip:
 
 static int fib_nl_dumprule(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct fib_rules_ops *ops;
 	int idx = 0, family;
 
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index c978bd1cd659..065fbac7ecd3 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -1478,7 +1478,7 @@ int neigh_table_clear(struct neigh_table *tbl)
 
 static int neigh_delete(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ndmsg *ndm;
 	struct nlattr *dst_attr;
 	struct neigh_table *tbl;
@@ -1544,7 +1544,7 @@ out:
 
 static int neigh_add(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ndmsg *ndm;
 	struct nlattr *tb[NDA_MAX+1];
 	struct neigh_table *tbl;
@@ -1812,7 +1812,7 @@ static const struct nla_policy nl_ntbl_parm_policy[NDTPA_MAX+1] = {
 
 static int neightbl_set(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct neigh_table *tbl;
 	struct ndtmsg *ndtmsg;
 	struct nlattr *tb[NDTA_MAX+1];
@@ -1937,7 +1937,7 @@ errout:
 
 static int neightbl_dump_info(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int family, tidx, nidx = 0;
 	int tbl_skip = cb->args[0];
 	int neigh_skip = cb->args[1];
@@ -2037,7 +2037,7 @@ static void neigh_update_notify(struct neighbour *neigh)
 static int neigh_dump_table(struct neigh_table *tbl, struct sk_buff *skb,
 			    struct netlink_callback *cb)
 {
-	struct net * net = skb->sk->sk_net;
+	struct net * net = sock_net(skb->sk);
 	struct neighbour *n;
 	int rc, h, s_h = cb->args[1];
 	int idx, s_idx = idx = cb->args[2];
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 09250a0800f6..da99ac0871bf 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -662,7 +662,7 @@ nla_put_failure:
 
 static int rtnl_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int idx;
 	int s_idx = cb->args[0];
 	struct net_device *dev;
@@ -879,7 +879,7 @@ errout:
 
 static int rtnl_setlink(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ifinfomsg *ifm;
 	struct net_device *dev;
 	int err;
@@ -921,7 +921,7 @@ errout:
 
 static int rtnl_dellink(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	const struct rtnl_link_ops *ops;
 	struct net_device *dev;
 	struct ifinfomsg *ifm;
@@ -1000,7 +1000,7 @@ err:
 
 static int rtnl_newlink(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	const struct rtnl_link_ops *ops;
 	struct net_device *dev;
 	struct ifinfomsg *ifm;
@@ -1132,7 +1132,7 @@ replay:
 
 static int rtnl_getlink(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ifinfomsg *ifm;
 	struct nlattr *tb[IFLA_MAX+1];
 	struct net_device *dev = NULL;
@@ -1227,7 +1227,7 @@ static int rtattr_max;
 
 static int rtnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	rtnl_doit_func doit;
 	int sz_idx, kind;
 	int min_len;
diff --git a/net/core/sock.c b/net/core/sock.c
index b1a6ed4d33c1..3ee95060dbd0 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -372,7 +372,7 @@ static int sock_bindtodevice(struct sock *sk, char __user *optval, int optlen)
 {
 	int ret = -ENOPROTOOPT;
 #ifdef CONFIG_NETDEVICES
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	char devname[IFNAMSIZ];
 	int index;
 
@@ -958,7 +958,7 @@ struct sock *sk_alloc(struct net *net, int family, gfp_t priority,
 		 */
 		sk->sk_prot = sk->sk_prot_creator = prot;
 		sock_lock_init(sk);
-		sk->sk_net = get_net(net);
+		sock_net_set(sk, get_net(net));
 	}
 
 	return sk;
@@ -983,7 +983,7 @@ void sk_free(struct sock *sk)
 		printk(KERN_DEBUG "%s: optmem leakage (%d bytes) detected.\n",
 		       __func__, atomic_read(&sk->sk_omem_alloc));
 
-	put_net(sk->sk_net);
+	put_net(sock_net(sk));
 	sk_prot_free(sk->sk_prot_creator, sk);
 }
 
@@ -1001,7 +1001,7 @@ void sk_release_kernel(struct sock *sk)
 
 	sock_hold(sk);
 	sock_release(sk->sk_socket);
-	sk->sk_net = get_net(&init_net);
+	sock_net_set(sk, get_net(&init_net));
 	sock_put(sk);
 }
 EXPORT_SYMBOL(sk_release_kernel);
@@ -1017,7 +1017,7 @@ struct sock *sk_clone(const struct sock *sk, const gfp_t priority)
 		sock_copy(newsk, sk);
 
 		/* SANITY */
-		get_net(newsk->sk_net);
+		get_net(sock_net(newsk));
 		sk_node_init(&newsk->sk_node);
 		sock_lock_init(newsk);
 		bh_lock_sock(newsk);
diff --git a/net/decnet/af_decnet.c b/net/decnet/af_decnet.c
index 3554fb3d251c..fc2efe899e91 100644
--- a/net/decnet/af_decnet.c
+++ b/net/decnet/af_decnet.c
@@ -1094,7 +1094,7 @@ static int dn_accept(struct socket *sock, struct socket *newsock, int flags)
 
 	cb = DN_SKB_CB(skb);
 	sk->sk_ack_backlog--;
-	newsk = dn_alloc_sock(sk->sk_net, newsock, sk->sk_allocation);
+	newsk = dn_alloc_sock(sock_net(sk), newsock, sk->sk_allocation);
 	if (newsk == NULL) {
 		release_sock(sk);
 		kfree_skb(skb);
diff --git a/net/decnet/dn_dev.c b/net/decnet/dn_dev.c
index 1bbfce5f7a2d..2f0ac3c3eb71 100644
--- a/net/decnet/dn_dev.c
+++ b/net/decnet/dn_dev.c
@@ -625,7 +625,7 @@ static const struct nla_policy dn_ifa_policy[IFA_MAX+1] = {
 
 static int dn_nl_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct nlattr *tb[IFA_MAX+1];
 	struct dn_dev *dn_db;
 	struct ifaddrmsg *ifm;
@@ -663,7 +663,7 @@ errout:
 
 static int dn_nl_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct nlattr *tb[IFA_MAX+1];
 	struct net_device *dev;
 	struct dn_dev *dn_db;
@@ -779,7 +779,7 @@ errout:
 
 static int dn_nl_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int idx, dn_idx = 0, skip_ndevs, skip_naddr;
 	struct net_device *dev;
 	struct dn_dev *dn_db;
diff --git a/net/decnet/dn_fib.c b/net/decnet/dn_fib.c
index 4aa9a423e606..27ea2e9b080a 100644
--- a/net/decnet/dn_fib.c
+++ b/net/decnet/dn_fib.c
@@ -504,7 +504,7 @@ static int dn_fib_check_attr(struct rtmsg *r, struct rtattr **rta)
 
 static int dn_fib_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct dn_fib_table *tb;
 	struct rtattr **rta = arg;
 	struct rtmsg *r = NLMSG_DATA(nlh);
@@ -524,7 +524,7 @@ static int dn_fib_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *
 
 static int dn_fib_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct dn_fib_table *tb;
 	struct rtattr **rta = arg;
 	struct rtmsg *r = NLMSG_DATA(nlh);
diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c
index 0a46b6c10e51..2f665a516476 100644
--- a/net/decnet/dn_route.c
+++ b/net/decnet/dn_route.c
@@ -1512,7 +1512,7 @@ rtattr_failure:
  */
 static int dn_cache_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = in_skb->sk->sk_net;
+	struct net *net = sock_net(in_skb->sk);
 	struct rtattr **rta = arg;
 	struct rtmsg *rtm = NLMSG_DATA(nlh);
 	struct dn_route *rt = NULL;
@@ -1601,7 +1601,7 @@ out_free:
  */
 int dn_cache_dump(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct dn_route *rt;
 	int h, s_h;
 	int idx, s_idx;
diff --git a/net/decnet/dn_table.c b/net/decnet/dn_table.c
index e09d915dbd77..3a2830ac89c2 100644
--- a/net/decnet/dn_table.c
+++ b/net/decnet/dn_table.c
@@ -463,7 +463,7 @@ static int dn_fib_table_dump(struct dn_fib_table *tb, struct sk_buff *skb,
 
 int dn_fib_dump(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	unsigned int h, s_h;
 	unsigned int e = 0, s_e;
 	struct dn_fib_table *tb;
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 06cfb0bed631..5882a1316441 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -464,7 +464,7 @@ int inet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
 	if (addr_len < sizeof(struct sockaddr_in))
 		goto out;
 
-	chk_addr_ret = inet_addr_type(sk->sk_net, addr->sin_addr.s_addr);
+	chk_addr_ret = inet_addr_type(sock_net(sk), addr->sin_addr.s_addr);
 
 	/* Not specified by any standard per-se, however it breaks too
 	 * many applications when removed.  It is unfortunate since
@@ -802,7 +802,7 @@ int inet_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
 {
 	struct sock *sk = sock->sk;
 	int err = 0;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	switch (cmd) {
 		case SIOCGSTAMP:
@@ -1132,7 +1132,7 @@ int inet_sk_rebuild_header(struct sock *sk)
 	};
 
 	security_sk_classify_flow(sk, &fl);
-	err = ip_route_output_flow(sk->sk_net, &rt, &fl, sk, 0);
+	err = ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 0);
 }
 	if (!err)
 		sk_setup_caps(sk, &rt->u.dst);
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 823c724a8593..6848e4760f34 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -437,7 +437,7 @@ struct in_ifaddr *inet_ifa_byprefix(struct in_device *in_dev, __be32 prefix,
 
 static int inet_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct nlattr *tb[IFA_MAX+1];
 	struct in_device *in_dev;
 	struct ifaddrmsg *ifm;
@@ -552,7 +552,7 @@ errout:
 
 static int inet_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct in_ifaddr *ifa;
 
 	ASSERT_RTNL();
@@ -1158,7 +1158,7 @@ nla_put_failure:
 
 static int inet_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int idx, ip_idx;
 	struct net_device *dev;
 	struct in_device *in_dev;
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index 0e4b34b07cb5..0f1557a4ac7a 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -583,7 +583,7 @@ errout:
 
 static int inet_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct fib_config cfg;
 	struct fib_table *tb;
 	int err;
@@ -605,7 +605,7 @@ errout:
 
 static int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct fib_config cfg;
 	struct fib_table *tb;
 	int err;
@@ -627,7 +627,7 @@ errout:
 
 static int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	unsigned int h, s_h;
 	unsigned int e = 0, s_e;
 	struct fib_table *tb;
@@ -857,7 +857,7 @@ static void nl_fib_input(struct sk_buff *skb)
 	struct fib_table *tb;
 	u32 pid;
 
-	net = skb->sk->sk_net;
+	net = sock_net(skb->sk);
 	nlh = nlmsg_hdr(skb);
 	if (skb->len < NLMSG_SPACE(0) || skb->len < nlh->nlmsg_len ||
 	    nlh->nlmsg_len < NLMSG_LENGTH(sizeof(*frn)))
diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index 19274d01afa4..1fb56876be54 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -137,7 +137,7 @@ static int fib4_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
 			       struct nlmsghdr *nlh, struct fib_rule_hdr *frh,
 			       struct nlattr **tb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int err = -EINVAL;
 	struct fib4_rule *rule4 = (struct fib4_rule *) rule;
 
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index 682f632bfb77..6250f4239b61 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -1762,7 +1762,7 @@ int ip_mc_join_group(struct sock *sk , struct ip_mreqn *imr)
 	if (!ipv4_is_multicast(addr))
 		return -EINVAL;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return -EPROTONOSUPPORT;
 
 	rtnl_lock();
@@ -1833,7 +1833,7 @@ int ip_mc_leave_group(struct sock *sk, struct ip_mreqn *imr)
 	u32 ifindex;
 	int ret = -EADDRNOTAVAIL;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return -EPROTONOSUPPORT;
 
 	rtnl_lock();
@@ -1881,7 +1881,7 @@ int ip_mc_source(int add, int omode, struct sock *sk, struct
 	if (!ipv4_is_multicast(addr))
 		return -EINVAL;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return -EPROTONOSUPPORT;
 
 	rtnl_lock();
@@ -2017,7 +2017,7 @@ int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf, int ifindex)
 	    msf->imsf_fmode != MCAST_EXCLUDE)
 		return -EINVAL;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return -EPROTONOSUPPORT;
 
 	rtnl_lock();
@@ -2100,7 +2100,7 @@ int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
 	if (!ipv4_is_multicast(addr))
 		return -EINVAL;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return -EPROTONOSUPPORT;
 
 	rtnl_lock();
@@ -2165,7 +2165,7 @@ int ip_mc_gsfget(struct sock *sk, struct group_filter *gsf,
 	if (!ipv4_is_multicast(addr))
 		return -EINVAL;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return -EPROTONOSUPPORT;
 
 	rtnl_lock();
@@ -2252,7 +2252,7 @@ void ip_mc_drop_socket(struct sock *sk)
 	if (inet->mc_list == NULL)
 		return;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return;
 
 	rtnl_lock();
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index d13c5f12bb32..a7fcaf205644 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -85,7 +85,7 @@ int inet_csk_get_port(struct sock *sk, unsigned short snum)
 	struct hlist_node *node;
 	struct inet_bind_bucket *tb;
 	int ret;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	local_bh_disable();
 	if (!snum) {
@@ -333,7 +333,7 @@ struct dst_entry* inet_csk_route_req(struct sock *sk,
 					 .dport = ireq->rmt_port } } };
 
 	security_req_classify_flow(req, &fl);
-	if (ip_route_output_flow(sk->sk_net, &rt, &fl, sk, 0)) {
+	if (ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 0)) {
 		IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
 		return NULL;
 	}
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 8cd1ad9b9111..1064111e5b96 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -139,7 +139,7 @@ static struct sock *inet_lookup_listener_slow(struct net *net,
 	sk_for_each(sk, node, head) {
 		const struct inet_sock *inet = inet_sk(sk);
 
-		if (sk->sk_net == net && inet->num == hnum &&
+		if (sock_net(sk) == net && inet->num == hnum &&
 				!ipv6_only_sock(sk)) {
 			const __be32 rcv_saddr = inet->rcv_saddr;
 			int score = sk->sk_family == PF_INET ? 1 : 0;
@@ -182,7 +182,7 @@ struct sock *__inet_lookup_listener(struct net *net,
 		if (inet->num == hnum && !sk->sk_node.next &&
 		    (!inet->rcv_saddr || inet->rcv_saddr == daddr) &&
 		    (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) &&
-		    !sk->sk_bound_dev_if && sk->sk_net == net)
+		    !sk->sk_bound_dev_if && sock_net(sk) == net)
 			goto sherry_cache;
 		sk = inet_lookup_listener_slow(net, head, daddr, hnum, dif);
 	}
@@ -254,7 +254,7 @@ static int __inet_check_established(struct inet_timewait_death_row *death_row,
 	struct sock *sk2;
 	const struct hlist_node *node;
 	struct inet_timewait_sock *tw;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	prefetch(head->chain.first);
 	write_lock(lock);
@@ -406,7 +406,7 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row,
 	struct inet_bind_hashbucket *head;
 	struct inet_bind_bucket *tb;
 	int ret;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	if (!snum) {
 		int i, remaining, low, high, port;
diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c
index 717c411a5c6b..f12bc24de46f 100644
--- a/net/ipv4/inet_timewait_sock.c
+++ b/net/ipv4/inet_timewait_sock.c
@@ -124,7 +124,7 @@ struct inet_timewait_sock *inet_twsk_alloc(const struct sock *sk, const int stat
 		tw->tw_hash	    = sk->sk_hash;
 		tw->tw_ipv6only	    = 0;
 		tw->tw_prot	    = sk->sk_prot_creator;
-		tw->tw_net          = sk->sk_net;
+		twsk_net_set(tw, sock_net(sk));
 		atomic_set(&tw->tw_refcnt, 1);
 		inet_twsk_dead_node_init(tw);
 		__module_get(tw->tw_prot->owner);
diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c
index 26685c83a146..4be00959b748 100644
--- a/net/ipv4/ip_input.c
+++ b/net/ipv4/ip_input.c
@@ -172,7 +172,7 @@ int ip_call_ra_chain(struct sk_buff *skb)
 		if (sk && inet_sk(sk)->num == protocol &&
 		    (!sk->sk_bound_dev_if ||
 		     sk->sk_bound_dev_if == dev->ifindex) &&
-		    sk->sk_net == dev_net(dev)) {
+		    sock_net(sk) == dev_net(dev)) {
 			if (ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)) {
 				if (ip_defrag(skb, IP_DEFRAG_CALL_RA_CHAIN)) {
 					read_unlock(&ip_ra_lock);
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 913266cd9902..08349267ceb4 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -351,7 +351,7 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok)
 			 * itself out.
 			 */
 			security_sk_classify_flow(sk, &fl);
-			if (ip_route_output_flow(sk->sk_net, &rt, &fl, sk, 0))
+			if (ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 0))
 				goto no_route;
 		}
 		sk_setup_caps(sk, &rt->u.dst);
@@ -1382,7 +1382,7 @@ void ip_send_reply(struct sock *sk, struct sk_buff *skb, struct ip_reply_arg *ar
 						 .dport = tcp_hdr(skb)->source } },
 				    .proto = sk->sk_protocol };
 		security_skb_classify_flow(skb, &fl);
-		if (ip_route_output_key(sk->sk_net, &rt, &fl))
+		if (ip_route_output_key(sock_net(sk), &rt, &fl))
 			return;
 	}
 
diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c
index b854431047a4..d6e76f5229cc 100644
--- a/net/ipv4/ip_sockglue.c
+++ b/net/ipv4/ip_sockglue.c
@@ -449,7 +449,7 @@ static int do_ip_setsockopt(struct sock *sk, int level,
 		struct ip_options * opt = NULL;
 		if (optlen > 40 || optlen < 0)
 			goto e_inval;
-		err = ip_options_get_from_user(sk->sk_net, &opt,
+		err = ip_options_get_from_user(sock_net(sk), &opt,
 					       optval, optlen);
 		if (err)
 			break;
@@ -590,13 +590,13 @@ static int do_ip_setsockopt(struct sock *sk, int level,
 				err = 0;
 				break;
 			}
-			dev = ip_dev_find(sk->sk_net, mreq.imr_address.s_addr);
+			dev = ip_dev_find(sock_net(sk), mreq.imr_address.s_addr);
 			if (dev) {
 				mreq.imr_ifindex = dev->ifindex;
 				dev_put(dev);
 			}
 		} else
-			dev = __dev_get_by_index(sk->sk_net, mreq.imr_ifindex);
+			dev = __dev_get_by_index(sock_net(sk), mreq.imr_ifindex);
 
 
 		err = -EADDRNOTAVAIL;
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index e54bc1364473..11700a4dcd95 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -849,7 +849,7 @@ static void mrtsock_destruct(struct sock *sk)
 {
 	rtnl_lock();
 	if (sk == mroute_socket) {
-		IPV4_DEVCONF_ALL(sk->sk_net, MC_FORWARDING)--;
+		IPV4_DEVCONF_ALL(sock_net(sk), MC_FORWARDING)--;
 
 		write_lock_bh(&mrt_lock);
 		mroute_socket=NULL;
@@ -898,7 +898,7 @@ int ip_mroute_setsockopt(struct sock *sk,int optname,char __user *optval,int opt
 			mroute_socket=sk;
 			write_unlock_bh(&mrt_lock);
 
-			IPV4_DEVCONF_ALL(sk->sk_net, MC_FORWARDING)++;
+			IPV4_DEVCONF_ALL(sock_net(sk), MC_FORWARDING)++;
 		}
 		rtnl_unlock();
 		return ret;
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index 756bc0e1a7c6..1563f29b5117 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -1496,11 +1496,11 @@ static int compat_do_arpt_set_ctl(struct sock *sk, int cmd, void __user *user,
 
 	switch (cmd) {
 	case ARPT_SO_SET_REPLACE:
-		ret = compat_do_replace(sk->sk_net, user, len);
+		ret = compat_do_replace(sock_net(sk), user, len);
 		break;
 
 	case ARPT_SO_SET_ADD_COUNTERS:
-		ret = do_add_counters(sk->sk_net, user, len, 1);
+		ret = do_add_counters(sock_net(sk), user, len, 1);
 		break;
 
 	default:
@@ -1644,10 +1644,10 @@ static int compat_do_arpt_get_ctl(struct sock *sk, int cmd, void __user *user,
 
 	switch (cmd) {
 	case ARPT_SO_GET_INFO:
-		ret = get_info(sk->sk_net, user, len, 1);
+		ret = get_info(sock_net(sk), user, len, 1);
 		break;
 	case ARPT_SO_GET_ENTRIES:
-		ret = compat_get_entries(sk->sk_net, user, len);
+		ret = compat_get_entries(sock_net(sk), user, len);
 		break;
 	default:
 		ret = do_arpt_get_ctl(sk, cmd, user, len);
@@ -1665,11 +1665,11 @@ static int do_arpt_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned
 
 	switch (cmd) {
 	case ARPT_SO_SET_REPLACE:
-		ret = do_replace(sk->sk_net, user, len);
+		ret = do_replace(sock_net(sk), user, len);
 		break;
 
 	case ARPT_SO_SET_ADD_COUNTERS:
-		ret = do_add_counters(sk->sk_net, user, len, 0);
+		ret = do_add_counters(sock_net(sk), user, len, 0);
 		break;
 
 	default:
@@ -1689,11 +1689,11 @@ static int do_arpt_get_ctl(struct sock *sk, int cmd, void __user *user, int *len
 
 	switch (cmd) {
 	case ARPT_SO_GET_INFO:
-		ret = get_info(sk->sk_net, user, len, 0);
+		ret = get_info(sock_net(sk), user, len, 0);
 		break;
 
 	case ARPT_SO_GET_ENTRIES:
-		ret = get_entries(sk->sk_net, user, len);
+		ret = get_entries(sock_net(sk), user, len);
 		break;
 
 	case ARPT_SO_GET_REVISION_TARGET: {
diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index 85a75e186b4b..a819d191e1aa 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -1852,11 +1852,11 @@ compat_do_ipt_set_ctl(struct sock *sk,	int cmd, void __user *user,
 
 	switch (cmd) {
 	case IPT_SO_SET_REPLACE:
-		ret = compat_do_replace(sk->sk_net, user, len);
+		ret = compat_do_replace(sock_net(sk), user, len);
 		break;
 
 	case IPT_SO_SET_ADD_COUNTERS:
-		ret = do_add_counters(sk->sk_net, user, len, 1);
+		ret = do_add_counters(sock_net(sk), user, len, 1);
 		break;
 
 	default:
@@ -1963,10 +1963,10 @@ compat_do_ipt_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
 
 	switch (cmd) {
 	case IPT_SO_GET_INFO:
-		ret = get_info(sk->sk_net, user, len, 1);
+		ret = get_info(sock_net(sk), user, len, 1);
 		break;
 	case IPT_SO_GET_ENTRIES:
-		ret = compat_get_entries(sk->sk_net, user, len);
+		ret = compat_get_entries(sock_net(sk), user, len);
 		break;
 	default:
 		ret = do_ipt_get_ctl(sk, cmd, user, len);
@@ -1985,11 +1985,11 @@ do_ipt_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
 
 	switch (cmd) {
 	case IPT_SO_SET_REPLACE:
-		ret = do_replace(sk->sk_net, user, len);
+		ret = do_replace(sock_net(sk), user, len);
 		break;
 
 	case IPT_SO_SET_ADD_COUNTERS:
-		ret = do_add_counters(sk->sk_net, user, len, 0);
+		ret = do_add_counters(sock_net(sk), user, len, 0);
 		break;
 
 	default:
@@ -2010,11 +2010,11 @@ do_ipt_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
 
 	switch (cmd) {
 	case IPT_SO_GET_INFO:
-		ret = get_info(sk->sk_net, user, len, 0);
+		ret = get_info(sock_net(sk), user, len, 0);
 		break;
 
 	case IPT_SO_GET_ENTRIES:
-		ret = get_entries(sk->sk_net, user, len);
+		ret = get_entries(sock_net(sk), user, len);
 		break;
 
 	case IPT_SO_GET_REVISION_MATCH:
diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c
index 8756d502a47f..be19a4048d7c 100644
--- a/net/ipv4/raw.c
+++ b/net/ipv4/raw.c
@@ -117,7 +117,7 @@ static struct sock *__raw_v4_lookup(struct net *net, struct sock *sk,
 	sk_for_each_from(sk, node) {
 		struct inet_sock *inet = inet_sk(sk);
 
-		if (sk->sk_net == net && inet->num == num 		&&
+		if (sock_net(sk) == net && inet->num == num 		&&
 		    !(inet->daddr && inet->daddr != raddr) 		&&
 		    !(inet->rcv_saddr && inet->rcv_saddr != laddr)	&&
 		    !(sk->sk_bound_dev_if && sk->sk_bound_dev_if != dif))
@@ -499,7 +499,7 @@ static int raw_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 	ipc.oif = sk->sk_bound_dev_if;
 
 	if (msg->msg_controllen) {
-		err = ip_cmsg_send(sk->sk_net, msg, &ipc);
+		err = ip_cmsg_send(sock_net(sk), msg, &ipc);
 		if (err)
 			goto out;
 		if (ipc.opt)
@@ -553,7 +553,7 @@ static int raw_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 		}
 
 		security_sk_classify_flow(sk, &fl);
-		err = ip_route_output_flow(sk->sk_net, &rt, &fl, sk, 1);
+		err = ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 1);
 	}
 	if (err)
 		goto done;
@@ -620,7 +620,7 @@ static int raw_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 
 	if (sk->sk_state != TCP_CLOSE || addr_len < sizeof(struct sockaddr_in))
 		goto out;
-	chk_addr_ret = inet_addr_type(sk->sk_net, addr->sin_addr.s_addr);
+	chk_addr_ret = inet_addr_type(sock_net(sk), addr->sin_addr.s_addr);
 	ret = -EADDRNOTAVAIL;
 	if (addr->sin_addr.s_addr && chk_addr_ret != RTN_LOCAL &&
 	    chk_addr_ret != RTN_MULTICAST && chk_addr_ret != RTN_BROADCAST)
@@ -856,7 +856,7 @@ static struct sock *raw_get_first(struct seq_file *seq)
 		struct hlist_node *node;
 
 		sk_for_each(sk, node, &state->h->ht[state->bucket])
-			if (sk->sk_net == state->p.net)
+			if (sock_net(sk) == state->p.net)
 				goto found;
 	}
 	sk = NULL;
@@ -872,7 +872,7 @@ static struct sock *raw_get_next(struct seq_file *seq, struct sock *sk)
 		sk = sk_next(sk);
 try_again:
 		;
-	} while (sk && sk->sk_net != state->p.net);
+	} while (sk && sock_net(sk) != state->p.net);
 
 	if (!sk && ++state->bucket < RAW_HTABLE_SIZE) {
 		sk = sk_head(&state->h->ht[state->bucket]);
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 7768d718e199..194f5cca3121 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -2689,7 +2689,7 @@ nla_put_failure:
 
 static int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct net *net = in_skb->sk->sk_net;
+	struct net *net = sock_net(in_skb->sk);
 	struct rtmsg *rtm;
 	struct nlattr *tb[RTA_MAX+1];
 	struct rtable *rt = NULL;
@@ -2785,7 +2785,7 @@ int ip_rt_dump(struct sk_buff *skb,  struct netlink_callback *cb)
 	int idx, s_idx;
 	struct net *net;
 
-	net = skb->sk->sk_net;
+	net = sock_net(skb->sk);
 
 	s_h = cb->args[0];
 	if (s_h < 0)
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 28bece6f281b..46847e600a46 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1486,7 +1486,7 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb)
 	if (req)
 		return tcp_check_req(sk, skb, req, prev);
 
-	nsk = inet_lookup_established(sk->sk_net, &tcp_hashinfo, iph->saddr,
+	nsk = inet_lookup_established(sock_net(sk), &tcp_hashinfo, iph->saddr,
 			th->source, iph->daddr, th->dest, inet_iif(skb));
 
 	if (nsk) {
@@ -1974,7 +1974,7 @@ static void *listening_get_next(struct seq_file *seq, void *cur)
 		while (1) {
 			while (req) {
 				if (req->rsk_ops->family == st->family &&
-				    req->sk->sk_net == net) {
+				    sock_net(req->sk) == net) {
 					cur = req;
 					goto out;
 				}
@@ -1998,7 +1998,7 @@ get_req:
 	}
 get_sk:
 	sk_for_each_from(sk, node) {
-		if (sk->sk_family == st->family && sk->sk_net == net) {
+		if (sk->sk_family == st->family && sock_net(sk) == net) {
 			cur = sk;
 			goto out;
 		}
@@ -2049,7 +2049,7 @@ static void *established_get_first(struct seq_file *seq)
 		read_lock_bh(lock);
 		sk_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) {
 			if (sk->sk_family != st->family ||
-			    sk->sk_net != net) {
+			    sock_net(sk) != net) {
 				continue;
 			}
 			rc = sk;
@@ -2059,7 +2059,7 @@ static void *established_get_first(struct seq_file *seq)
 		inet_twsk_for_each(tw, node,
 				   &tcp_hashinfo.ehash[st->bucket].twchain) {
 			if (tw->tw_family != st->family ||
-			    tw->tw_net != net) {
+			    twsk_net(tw) != net) {
 				continue;
 			}
 			rc = tw;
@@ -2086,7 +2086,7 @@ static void *established_get_next(struct seq_file *seq, void *cur)
 		tw = cur;
 		tw = tw_next(tw);
 get_tw:
-		while (tw && (tw->tw_family != st->family || tw->tw_net != net)) {
+		while (tw && (tw->tw_family != st->family || twsk_net(tw) != net)) {
 			tw = tw_next(tw);
 		}
 		if (tw) {
@@ -2107,7 +2107,7 @@ get_tw:
 		sk = sk_next(sk);
 
 	sk_for_each_from(sk, node) {
-		if (sk->sk_family == st->family && sk->sk_net == net)
+		if (sk->sk_family == st->family && sock_net(sk) == net)
 			goto found;
 	}
 
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index e2cd93481359..76d52d37d6ac 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -137,7 +137,7 @@ static inline int __udp_lib_lport_inuse(struct net *net, __u16 num,
 	struct hlist_node *node;
 
 	sk_for_each(sk, node, &udptable[num & (UDP_HTABLE_SIZE - 1)])
-		if (sk->sk_net == net && sk->sk_hash == num)
+		if (sock_net(sk) == net && sk->sk_hash == num)
 			return 1;
 	return 0;
 }
@@ -158,7 +158,7 @@ int udp_lib_get_port(struct sock *sk, unsigned short snum,
 	struct hlist_head *head;
 	struct sock *sk2;
 	int    error = 1;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	write_lock_bh(&udp_hash_lock);
 
@@ -218,7 +218,7 @@ gotit:
 		sk_for_each(sk2, node, head)
 			if (sk2->sk_hash == snum                             &&
 			    sk2 != sk                                        &&
-			    sk2->sk_net == net				     &&
+			    sock_net(sk2) == net				     &&
 			    (!sk2->sk_reuse        || !sk->sk_reuse)         &&
 			    (!sk2->sk_bound_dev_if || !sk->sk_bound_dev_if
 			     || sk2->sk_bound_dev_if == sk->sk_bound_dev_if) &&
@@ -269,7 +269,7 @@ static struct sock *__udp4_lib_lookup(struct net *net, __be32 saddr,
 	sk_for_each(sk, node, &udptable[hnum & (UDP_HTABLE_SIZE - 1)]) {
 		struct inet_sock *inet = inet_sk(sk);
 
-		if (sk->sk_net == net && sk->sk_hash == hnum &&
+		if (sock_net(sk) == net && sk->sk_hash == hnum &&
 				!ipv6_only_sock(sk)) {
 			int score = (sk->sk_family == PF_INET ? 1 : 0);
 			if (inet->rcv_saddr) {
@@ -607,7 +607,7 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 
 	ipc.oif = sk->sk_bound_dev_if;
 	if (msg->msg_controllen) {
-		err = ip_cmsg_send(sk->sk_net, msg, &ipc);
+		err = ip_cmsg_send(sock_net(sk), msg, &ipc);
 		if (err)
 			return err;
 		if (ipc.opt)
@@ -656,7 +656,7 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 					       { .sport = inet->sport,
 						 .dport = dport } } };
 		security_sk_classify_flow(sk, &fl);
-		err = ip_route_output_flow(sk->sk_net, &rt, &fl, sk, 1);
+		err = ip_route_output_flow(sock_net(sk), &rt, &fl, sk, 1);
 		if (err) {
 			if (err == -ENETUNREACH)
 				IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
@@ -1511,7 +1511,7 @@ static struct sock *udp_get_first(struct seq_file *seq)
 	for (state->bucket = 0; state->bucket < UDP_HTABLE_SIZE; ++state->bucket) {
 		struct hlist_node *node;
 		sk_for_each(sk, node, state->hashtable + state->bucket) {
-			if (sk->sk_net != net)
+			if (sock_net(sk) != net)
 				continue;
 			if (sk->sk_family == state->family)
 				goto found;
@@ -1531,7 +1531,7 @@ static struct sock *udp_get_next(struct seq_file *seq, struct sock *sk)
 		sk = sk_next(sk);
 try_again:
 		;
-	} while (sk && (sk->sk_net != net || sk->sk_family != state->family));
+	} while (sk && (sock_net(sk) != net || sk->sk_family != state->family));
 
 	if (!sk && ++state->bucket < UDP_HTABLE_SIZE) {
 		sk = sk_head(state->hashtable + state->bucket);
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index d1de9ec74261..f2c90f145cbb 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -3054,7 +3054,7 @@ static const struct nla_policy ifa_ipv6_policy[IFA_MAX+1] = {
 static int
 inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ifaddrmsg *ifm;
 	struct nlattr *tb[IFA_MAX+1];
 	struct in6_addr *pfx;
@@ -3112,7 +3112,7 @@ static int inet6_addr_modify(struct inet6_ifaddr *ifp, u8 ifa_flags,
 static int
 inet6_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ifaddrmsg *ifm;
 	struct nlattr *tb[IFA_MAX+1];
 	struct in6_addr *pfx;
@@ -3322,7 +3322,7 @@ static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
 	struct inet6_ifaddr *ifa;
 	struct ifmcaddr6 *ifmca;
 	struct ifacaddr6 *ifaca;
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 
 	s_idx = cb->args[0];
 	s_ip_idx = ip_idx = cb->args[1];
@@ -3418,7 +3418,7 @@ static int inet6_dump_ifacaddr(struct sk_buff *skb, struct netlink_callback *cb)
 static int inet6_rtm_getaddr(struct sk_buff *in_skb, struct nlmsghdr* nlh,
 			     void *arg)
 {
-	struct net *net = in_skb->sk->sk_net;
+	struct net *net = sock_net(in_skb->sk);
 	struct ifaddrmsg *ifm;
 	struct nlattr *tb[IFA_MAX+1];
 	struct in6_addr *addr = NULL;
@@ -3645,7 +3645,7 @@ nla_put_failure:
 
 static int inet6_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int idx, err;
 	int s_idx = cb->args[0];
 	struct net_device *dev;
diff --git a/net/ipv6/addrlabel.c b/net/ipv6/addrlabel.c
index de371b5997fe..9bfa8846f262 100644
--- a/net/ipv6/addrlabel.c
+++ b/net/ipv6/addrlabel.c
@@ -364,7 +364,7 @@ static const struct nla_policy ifal_policy[IFAL_MAX+1] = {
 static int ip6addrlbl_newdel(struct sk_buff *skb, struct nlmsghdr *nlh,
 			     void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ifaddrlblmsg *ifal;
 	struct nlattr *tb[IFAL_MAX+1];
 	struct in6_addr *pfx;
@@ -452,7 +452,7 @@ static int ip6addrlbl_fill(struct sk_buff *skb,
 
 static int ip6addrlbl_dump(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct ip6addrlbl_entry *p;
 	struct hlist_node *pos;
 	int idx = 0, s_idx = cb->args[0];
@@ -490,7 +490,7 @@ static inline int ip6addrlbl_msgsize(void)
 static int ip6addrlbl_get(struct sk_buff *in_skb, struct nlmsghdr* nlh,
 			  void *arg)
 {
-	struct net *net = in_skb->sk->sk_net;
+	struct net *net = sock_net(in_skb->sk);
 	struct ifaddrlblmsg *ifal;
 	struct nlattr *tb[IFAL_MAX+1];
 	struct in6_addr *addr;
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index f52bdaed8a1b..12f04e9d3e88 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -245,7 +245,7 @@ int inet6_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
 	struct sock *sk = sock->sk;
 	struct inet_sock *inet = inet_sk(sk);
 	struct ipv6_pinfo *np = inet6_sk(sk);
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	__be32 v4addr = 0;
 	unsigned short snum;
 	int addr_type = 0;
@@ -438,7 +438,7 @@ EXPORT_SYMBOL(inet6_getname);
 int inet6_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
 {
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	switch(cmd)
 	{
diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c
index e7a7fe26cebf..cac580749ebe 100644
--- a/net/ipv6/fib6_rules.c
+++ b/net/ipv6/fib6_rules.c
@@ -154,7 +154,7 @@ static int fib6_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
 			       struct nlattr **tb)
 {
 	int err = -EINVAL;
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct fib6_rule *rule6 = (struct fib6_rule *) rule;
 
 	if (rule->action == FR_ACT_TO_TBL) {
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index 50857662e6b7..63309d10df3a 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -163,7 +163,7 @@ static inline int icmpv6_xrlim_allow(struct sock *sk, int type,
 				     struct flowi *fl)
 {
 	struct dst_entry *dst;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	int res = 0;
 
 	/* Informational messages are not limited. */
diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c
index c0c8d2d17682..21c467675412 100644
--- a/net/ipv6/inet6_hashtables.c
+++ b/net/ipv6/inet6_hashtables.c
@@ -105,7 +105,7 @@ struct sock *inet6_lookup_listener(struct net *net,
 
 	read_lock(&hashinfo->lhash_lock);
 	sk_for_each(sk, node, &hashinfo->listening_hash[inet_lhashfn(hnum)]) {
-		if (sk->sk_net == net && inet_sk(sk)->num == hnum &&
+		if (sock_net(sk) == net && inet_sk(sk)->num == hnum &&
 				sk->sk_family == PF_INET6) {
 			const struct ipv6_pinfo *np = inet6_sk(sk);
 
@@ -172,7 +172,7 @@ static int __inet6_check_established(struct inet_timewait_death_row *death_row,
 	struct sock *sk2;
 	const struct hlist_node *node;
 	struct inet_timewait_sock *tw;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	prefetch(head->chain.first);
 	write_lock(lock);
diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index b0814b0082e7..b3f6e03c454c 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -346,7 +346,7 @@ end:
 
 static int inet6_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	unsigned int h, s_h;
 	unsigned int e = 0, s_e;
 	struct rt6_rtnl_dump_arg arg;
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 556300f0eba5..a8b4da25b0a7 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -910,7 +910,7 @@ static int ip6_dst_lookup_tail(struct sock *sk,
 			       struct dst_entry **dst, struct flowi *fl)
 {
 	int err;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	if (*dst == NULL)
 		*dst = ip6_route_output(net, sk, fl);
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index dc6695cc5767..d3d93d752e10 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -107,7 +107,7 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 		    char __user *optval, int optlen)
 {
 	struct ipv6_pinfo *np = inet6_sk(sk);
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	int val, valbool;
 	int retv = -ENOPROTOOPT;
 
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 0357de8e78c8..20a3d8e2f6c6 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -181,7 +181,7 @@ int ipv6_sock_mc_join(struct sock *sk, int ifindex, struct in6_addr *addr)
 	struct net_device *dev = NULL;
 	struct ipv6_mc_socklist *mc_lst;
 	struct ipv6_pinfo *np = inet6_sk(sk);
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	int err;
 
 	if (!ipv6_addr_is_multicast(addr))
@@ -255,7 +255,7 @@ int ipv6_sock_mc_drop(struct sock *sk, int ifindex, struct in6_addr *addr)
 {
 	struct ipv6_pinfo *np = inet6_sk(sk);
 	struct ipv6_mc_socklist *mc_lst, **lnk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	write_lock_bh(&ipv6_sk_mc_lock);
 	for (lnk = &np->ipv6_mc_list; (mc_lst = *lnk) !=NULL ; lnk = &mc_lst->next) {
@@ -327,7 +327,7 @@ void ipv6_sock_mc_close(struct sock *sk)
 {
 	struct ipv6_pinfo *np = inet6_sk(sk);
 	struct ipv6_mc_socklist *mc_lst;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	write_lock_bh(&ipv6_sk_mc_lock);
 	while ((mc_lst = np->ipv6_mc_list) != NULL) {
@@ -365,7 +365,7 @@ int ip6_mc_source(int add, int omode, struct sock *sk,
 	struct inet6_dev *idev;
 	struct ipv6_pinfo *inet6 = inet6_sk(sk);
 	struct ip6_sf_socklist *psl;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	int i, j, rv;
 	int leavegroup = 0;
 	int pmclocked = 0;
@@ -505,7 +505,7 @@ int ip6_mc_msfilter(struct sock *sk, struct group_filter *gsf)
 	struct inet6_dev *idev;
 	struct ipv6_pinfo *inet6 = inet6_sk(sk);
 	struct ip6_sf_socklist *newpsl, *psl;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	int leavegroup = 0;
 	int i, err;
 
@@ -598,7 +598,7 @@ int ip6_mc_msfget(struct sock *sk, struct group_filter *gsf,
 	struct net_device *dev;
 	struct ipv6_pinfo *inet6 = inet6_sk(sk);
 	struct ip6_sf_socklist *psl;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 
 	group = &((struct sockaddr_in6 *)&gsf->gf_group)->sin6_addr;
 
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index af1ec7ba757c..70ef0d276cc0 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -1879,11 +1879,11 @@ compat_do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user,
 
 	switch (cmd) {
 	case IP6T_SO_SET_REPLACE:
-		ret = compat_do_replace(sk->sk_net, user, len);
+		ret = compat_do_replace(sock_net(sk), user, len);
 		break;
 
 	case IP6T_SO_SET_ADD_COUNTERS:
-		ret = do_add_counters(sk->sk_net, user, len, 1);
+		ret = do_add_counters(sock_net(sk), user, len, 1);
 		break;
 
 	default:
@@ -1990,10 +1990,10 @@ compat_do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
 
 	switch (cmd) {
 	case IP6T_SO_GET_INFO:
-		ret = get_info(sk->sk_net, user, len, 1);
+		ret = get_info(sock_net(sk), user, len, 1);
 		break;
 	case IP6T_SO_GET_ENTRIES:
-		ret = compat_get_entries(sk->sk_net, user, len);
+		ret = compat_get_entries(sock_net(sk), user, len);
 		break;
 	default:
 		ret = do_ip6t_get_ctl(sk, cmd, user, len);
@@ -2012,11 +2012,11 @@ do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
 
 	switch (cmd) {
 	case IP6T_SO_SET_REPLACE:
-		ret = do_replace(sk->sk_net, user, len);
+		ret = do_replace(sock_net(sk), user, len);
 		break;
 
 	case IP6T_SO_SET_ADD_COUNTERS:
-		ret = do_add_counters(sk->sk_net, user, len, 0);
+		ret = do_add_counters(sock_net(sk), user, len, 0);
 		break;
 
 	default:
@@ -2037,11 +2037,11 @@ do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
 
 	switch (cmd) {
 	case IP6T_SO_GET_INFO:
-		ret = get_info(sk->sk_net, user, len, 0);
+		ret = get_info(sock_net(sk), user, len, 0);
 		break;
 
 	case IP6T_SO_GET_ENTRIES:
-		ret = get_entries(sk->sk_net, user, len);
+		ret = get_entries(sock_net(sk), user, len);
 		break;
 
 	case IP6T_SO_GET_REVISION_MATCH:
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index efb0047f6880..12c7a1560977 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -76,7 +76,7 @@ static struct sock *__raw_v6_lookup(struct net *net, struct sock *sk,
 		if (inet_sk(sk)->num == num) {
 			struct ipv6_pinfo *np = inet6_sk(sk);
 
-			if (sk->sk_net != net)
+			if (sock_net(sk) != net)
 				continue;
 
 			if (!ipv6_addr_any(&np->daddr) &&
@@ -280,7 +280,7 @@ static int rawv6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 			if (!sk->sk_bound_dev_if)
 				goto out;
 
-			dev = dev_get_by_index(sk->sk_net, sk->sk_bound_dev_if);
+			dev = dev_get_by_index(sock_net(sk), sk->sk_bound_dev_if);
 			if (!dev) {
 				err = -ENODEV;
 				goto out;
@@ -293,7 +293,7 @@ static int rawv6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 		v4addr = LOOPBACK4_IPV6;
 		if (!(addr_type & IPV6_ADDR_MULTICAST))	{
 			err = -EADDRNOTAVAIL;
-			if (!ipv6_chk_addr(sk->sk_net, &addr->sin6_addr,
+			if (!ipv6_chk_addr(sock_net(sk), &addr->sin6_addr,
 					   dev, 0)) {
 				if (dev)
 					dev_put(dev);
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 65053fba8c1a..ac4428371432 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -2020,7 +2020,7 @@ static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh,
 
 	cfg->fc_nlinfo.pid = NETLINK_CB(skb).pid;
 	cfg->fc_nlinfo.nlh = nlh;
-	cfg->fc_nlinfo.nl_net = skb->sk->sk_net;
+	cfg->fc_nlinfo.nl_net = sock_net(skb->sk);
 
 	if (tb[RTA_GATEWAY]) {
 		nla_memcpy(&cfg->fc_gateway, tb[RTA_GATEWAY], 16);
@@ -2216,7 +2216,7 @@ int rt6_dump_route(struct rt6_info *rt, void *p_arg)
 
 static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct net *net = in_skb->sk->sk_net;
+	struct net *net = sock_net(in_skb->sk);
 	struct nlattr *tb[RTA_MAX+1];
 	struct rt6_info *rt;
 	struct sk_buff *skb;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 086deffff9c9..323c7e06ef43 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1218,7 +1218,7 @@ static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb)
 	if (req)
 		return tcp_check_req(sk, skb, req, prev);
 
-	nsk = __inet6_lookup_established(sk->sk_net, &tcp_hashinfo,
+	nsk = __inet6_lookup_established(sock_net(sk), &tcp_hashinfo,
 			&ipv6_hdr(skb)->saddr, th->source,
 			&ipv6_hdr(skb)->daddr, ntohs(th->dest), inet6_iif(skb));
 
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 6683c04b427e..db266ff297e5 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -70,7 +70,7 @@ static struct sock *__udp6_lib_lookup(struct net *net,
 	sk_for_each(sk, node, &udptable[hnum & (UDP_HTABLE_SIZE - 1)]) {
 		struct inet_sock *inet = inet_sk(sk);
 
-		if (sk->sk_net == net && sk->sk_hash == hnum &&
+		if (sock_net(sk) == net && sk->sk_hash == hnum &&
 				sk->sk_family == PF_INET6) {
 			struct ipv6_pinfo *np = inet6_sk(sk);
 			int score = 0;
@@ -323,7 +323,7 @@ static struct sock *udp_v6_mcast_next(struct sock *sk,
 	sk_for_each_from(s, node) {
 		struct inet_sock *inet = inet_sk(s);
 
-		if (s->sk_net != sk->sk_net)
+		if (sock_net(s) != sock_net(sk))
 			continue;
 
 		if (s->sk_hash == num && s->sk_family == PF_INET6) {
diff --git a/net/irda/af_irda.c b/net/irda/af_irda.c
index 6f21a53cb3e7..ae54b20d0470 100644
--- a/net/irda/af_irda.c
+++ b/net/irda/af_irda.c
@@ -837,7 +837,7 @@ static int irda_accept(struct socket *sock, struct socket *newsock, int flags)
 
 	IRDA_DEBUG(2, "%s()\n", __func__);
 
-	err = irda_create(sk->sk_net, newsock, sk->sk_protocol);
+	err = irda_create(sock_net(sk), newsock, sk->sk_protocol);
 	if (err)
 		return err;
 
diff --git a/net/llc/llc_conn.c b/net/llc/llc_conn.c
index 5ebfd93ff5e7..5c6d89c6d51d 100644
--- a/net/llc/llc_conn.c
+++ b/net/llc/llc_conn.c
@@ -700,7 +700,7 @@ static struct sock *llc_create_incoming_sock(struct sock *sk,
 					     struct llc_addr *saddr,
 					     struct llc_addr *daddr)
 {
-	struct sock *newsk = llc_sk_alloc(sk->sk_net, sk->sk_family, GFP_ATOMIC,
+	struct sock *newsk = llc_sk_alloc(sock_net(sk), sk->sk_family, GFP_ATOMIC,
 					  sk->sk_prot);
 	struct llc_sock *newllc, *llc = llc_sk(sk);
 
diff --git a/net/netfilter/nf_sockopt.c b/net/netfilter/nf_sockopt.c
index 3dd4b3c76d81..69d699f95f4c 100644
--- a/net/netfilter/nf_sockopt.c
+++ b/net/netfilter/nf_sockopt.c
@@ -65,7 +65,7 @@ static struct nf_sockopt_ops *nf_sockopt_find(struct sock *sk, int pf,
 {
 	struct nf_sockopt_ops *ops;
 
-	if (sk->sk_net != &init_net)
+	if (sock_net(sk) != &init_net)
 		return ERR_PTR(-ENOPROTOOPT);
 
 	if (mutex_lock_interruptible(&nf_sockopt_mutex) != 0)
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 86bd8660a8f2..712a7bff8560 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -228,7 +228,7 @@ static inline struct sock *netlink_lookup(struct net *net, int protocol,
 	read_lock(&nl_table_lock);
 	head = nl_pid_hashfn(hash, pid);
 	sk_for_each(sk, node, head) {
-		if ((sk->sk_net == net) && (nlk_sk(sk)->pid == pid)) {
+		if (sock_net(sk) == net && (nlk_sk(sk)->pid == pid)) {
 			sock_hold(sk);
 			goto found;
 		}
@@ -348,7 +348,7 @@ static int netlink_insert(struct sock *sk, struct net *net, u32 pid)
 	head = nl_pid_hashfn(hash, pid);
 	len = 0;
 	sk_for_each(osk, node, head) {
-		if ((osk->sk_net == net) && (nlk_sk(osk)->pid == pid))
+		if (sock_net(osk) == net && (nlk_sk(osk)->pid == pid))
 			break;
 		len++;
 	}
@@ -486,7 +486,7 @@ static int netlink_release(struct socket *sock)
 
 	if (nlk->pid && !nlk->subscriptions) {
 		struct netlink_notify n = {
-						.net = sk->sk_net,
+						.net = sock_net(sk),
 						.protocol = sk->sk_protocol,
 						.pid = nlk->pid,
 					  };
@@ -518,7 +518,7 @@ static int netlink_release(struct socket *sock)
 static int netlink_autobind(struct socket *sock)
 {
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	struct nl_pid_hash *hash = &nl_table[sk->sk_protocol].hash;
 	struct hlist_head *head;
 	struct sock *osk;
@@ -532,7 +532,7 @@ retry:
 	netlink_table_grab();
 	head = nl_pid_hashfn(hash, pid);
 	sk_for_each(osk, node, head) {
-		if ((osk->sk_net != net))
+		if (sock_net(osk) != net)
 			continue;
 		if (nlk_sk(osk)->pid == pid) {
 			/* Bind collision, search negative pid values. */
@@ -611,7 +611,7 @@ static int netlink_bind(struct socket *sock, struct sockaddr *addr,
 			int addr_len)
 {
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	struct netlink_sock *nlk = nlk_sk(sk);
 	struct sockaddr_nl *nladdr = (struct sockaddr_nl *)addr;
 	int err;
@@ -720,7 +720,7 @@ static struct sock *netlink_getsockbypid(struct sock *ssk, u32 pid)
 	struct sock *sock;
 	struct netlink_sock *nlk;
 
-	sock = netlink_lookup(ssk->sk_net, ssk->sk_protocol, pid);
+	sock = netlink_lookup(sock_net(ssk), ssk->sk_protocol, pid);
 	if (!sock)
 		return ERR_PTR(-ECONNREFUSED);
 
@@ -962,7 +962,7 @@ static inline int do_one_broadcast(struct sock *sk,
 	    !test_bit(p->group - 1, nlk->groups))
 		goto out;
 
-	if ((sk->sk_net != p->net))
+	if (sock_net(sk) != p->net)
 		goto out;
 
 	if (p->failure) {
@@ -1006,7 +1006,7 @@ out:
 int netlink_broadcast(struct sock *ssk, struct sk_buff *skb, u32 pid,
 		      u32 group, gfp_t allocation)
 {
-	struct net *net = ssk->sk_net;
+	struct net *net = sock_net(ssk);
 	struct netlink_broadcast_data info;
 	struct hlist_node *node;
 	struct sock *sk;
@@ -1064,7 +1064,7 @@ static inline int do_one_set_err(struct sock *sk,
 	if (sk == p->exclude_sk)
 		goto out;
 
-	if (sk->sk_net != p->exclude_sk->sk_net)
+	if (sock_net(sk) != sock_net(p->exclude_sk))
 		goto out;
 
 	if (nlk->pid == p->pid || p->group - 1 >= nlk->ngroups ||
@@ -1601,7 +1601,7 @@ int netlink_dump_start(struct sock *ssk, struct sk_buff *skb,
 	atomic_inc(&skb->users);
 	cb->skb = skb;
 
-	sk = netlink_lookup(ssk->sk_net, ssk->sk_protocol, NETLINK_CB(skb).pid);
+	sk = netlink_lookup(sock_net(ssk), ssk->sk_protocol, NETLINK_CB(skb).pid);
 	if (sk == NULL) {
 		netlink_destroy_callback(cb);
 		return -ECONNREFUSED;
@@ -1643,7 +1643,7 @@ void netlink_ack(struct sk_buff *in_skb, struct nlmsghdr *nlh, int err)
 	if (!skb) {
 		struct sock *sk;
 
-		sk = netlink_lookup(in_skb->sk->sk_net,
+		sk = netlink_lookup(sock_net(in_skb->sk),
 				    in_skb->sk->sk_protocol,
 				    NETLINK_CB(in_skb).pid);
 		if (sk) {
@@ -1758,7 +1758,7 @@ static struct sock *netlink_seq_socket_idx(struct seq_file *seq, loff_t pos)
 
 		for (j = 0; j <= hash->mask; j++) {
 			sk_for_each(s, node, &hash->table[j]) {
-				if (iter->p.net != s->sk_net)
+				if (sock_net(s) != iter->p.net)
 					continue;
 				if (off == pos) {
 					iter->link = i;
@@ -1794,7 +1794,7 @@ static void *netlink_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 	s = v;
 	do {
 		s = sk_next(s);
-	} while (s && (iter->p.net != s->sk_net));
+	} while (s && (sock_net(s) != iter->p.net));
 	if (s)
 		return s;
 
@@ -1806,7 +1806,7 @@ static void *netlink_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 
 		for (; j <= hash->mask; j++) {
 			s = sk_head(&hash->table[j]);
-			while (s && (iter->p.net != s->sk_net))
+			while (s && sock_net(s) != iter->p.net)
 				s = sk_next(s);
 			if (s) {
 				iter->link = i;
diff --git a/net/netrom/af_netrom.c b/net/netrom/af_netrom.c
index a270ebf9f765..4bae8b998cab 100644
--- a/net/netrom/af_netrom.c
+++ b/net/netrom/af_netrom.c
@@ -466,7 +466,7 @@ static struct sock *nr_make_new(struct sock *osk)
 	if (osk->sk_type != SOCK_SEQPACKET)
 		return NULL;
 
-	sk = sk_alloc(osk->sk_net, PF_NETROM, GFP_ATOMIC, osk->sk_prot);
+	sk = sk_alloc(sock_net(osk), PF_NETROM, GFP_ATOMIC, osk->sk_prot);
 	if (sk == NULL)
 		return NULL;
 
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index baa290d3444a..25070240d4ae 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -263,7 +263,7 @@ static int packet_rcv_spkt(struct sk_buff *skb, struct net_device *dev,  struct
 	if (skb->pkt_type == PACKET_LOOPBACK)
 		goto out;
 
-	if (dev_net(dev) != sk->sk_net)
+	if (dev_net(dev) != sock_net(sk))
 		goto out;
 
 	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
@@ -337,7 +337,7 @@ static int packet_sendmsg_spkt(struct kiocb *iocb, struct socket *sock,
 	 */
 
 	saddr->spkt_device[13] = 0;
-	dev = dev_get_by_name(sk->sk_net, saddr->spkt_device);
+	dev = dev_get_by_name(sock_net(sk), saddr->spkt_device);
 	err = -ENODEV;
 	if (dev == NULL)
 		goto out_unlock;
@@ -451,7 +451,7 @@ static int packet_rcv(struct sk_buff *skb, struct net_device *dev, struct packet
 	sk = pt->af_packet_priv;
 	po = pkt_sk(sk);
 
-	if (dev_net(dev) != sk->sk_net)
+	if (dev_net(dev) != sock_net(sk))
 		goto drop;
 
 	skb->dev = dev;
@@ -568,7 +568,7 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, struct packe
 	sk = pt->af_packet_priv;
 	po = pkt_sk(sk);
 
-	if (dev_net(dev) != sk->sk_net)
+	if (dev_net(dev) != sock_net(sk))
 		goto drop;
 
 	if (dev->header_ops) {
@@ -728,7 +728,7 @@ static int packet_sendmsg(struct kiocb *iocb, struct socket *sock,
 	}
 
 
-	dev = dev_get_by_index(sk->sk_net, ifindex);
+	dev = dev_get_by_index(sock_net(sk), ifindex);
 	err = -ENXIO;
 	if (dev == NULL)
 		goto out_unlock;
@@ -800,7 +800,7 @@ static int packet_release(struct socket *sock)
 	if (!sk)
 		return 0;
 
-	net = sk->sk_net;
+	net = sock_net(sk);
 	po = pkt_sk(sk);
 
 	write_lock_bh(&net->packet.sklist_lock);
@@ -914,7 +914,7 @@ static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr, int add
 		return -EINVAL;
 	strlcpy(name,uaddr->sa_data,sizeof(name));
 
-	dev = dev_get_by_name(sk->sk_net, name);
+	dev = dev_get_by_name(sock_net(sk), name);
 	if (dev) {
 		err = packet_do_bind(sk, dev, pkt_sk(sk)->num);
 		dev_put(dev);
@@ -941,7 +941,7 @@ static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len
 
 	if (sll->sll_ifindex) {
 		err = -ENODEV;
-		dev = dev_get_by_index(sk->sk_net, sll->sll_ifindex);
+		dev = dev_get_by_index(sock_net(sk), sll->sll_ifindex);
 		if (dev == NULL)
 			goto out;
 	}
@@ -1135,7 +1135,7 @@ static int packet_getname_spkt(struct socket *sock, struct sockaddr *uaddr,
 		return -EOPNOTSUPP;
 
 	uaddr->sa_family = AF_PACKET;
-	dev = dev_get_by_index(sk->sk_net, pkt_sk(sk)->ifindex);
+	dev = dev_get_by_index(sock_net(sk), pkt_sk(sk)->ifindex);
 	if (dev) {
 		strlcpy(uaddr->sa_data, dev->name, 15);
 		dev_put(dev);
@@ -1160,7 +1160,7 @@ static int packet_getname(struct socket *sock, struct sockaddr *uaddr,
 	sll->sll_family = AF_PACKET;
 	sll->sll_ifindex = po->ifindex;
 	sll->sll_protocol = po->num;
-	dev = dev_get_by_index(sk->sk_net, po->ifindex);
+	dev = dev_get_by_index(sock_net(sk), po->ifindex);
 	if (dev) {
 		sll->sll_hatype = dev->type;
 		sll->sll_halen = dev->addr_len;
@@ -1212,7 +1212,7 @@ static int packet_mc_add(struct sock *sk, struct packet_mreq_max *mreq)
 	rtnl_lock();
 
 	err = -ENODEV;
-	dev = __dev_get_by_index(sk->sk_net, mreq->mr_ifindex);
+	dev = __dev_get_by_index(sock_net(sk), mreq->mr_ifindex);
 	if (!dev)
 		goto done;
 
@@ -1266,7 +1266,7 @@ static int packet_mc_drop(struct sock *sk, struct packet_mreq_max *mreq)
 			if (--ml->count == 0) {
 				struct net_device *dev;
 				*mlp = ml->next;
-				dev = dev_get_by_index(sk->sk_net, ml->ifindex);
+				dev = dev_get_by_index(sock_net(sk), ml->ifindex);
 				if (dev) {
 					packet_dev_mc(dev, ml, -1);
 					dev_put(dev);
@@ -1294,7 +1294,7 @@ static void packet_flush_mclist(struct sock *sk)
 		struct net_device *dev;
 
 		po->mclist = ml->next;
-		if ((dev = dev_get_by_index(sk->sk_net, ml->ifindex)) != NULL) {
+		if ((dev = dev_get_by_index(sock_net(sk), ml->ifindex)) != NULL) {
 			packet_dev_mc(dev, ml, -1);
 			dev_put(dev);
 		}
@@ -1540,7 +1540,7 @@ static int packet_ioctl(struct socket *sock, unsigned int cmd,
 		case SIOCGIFDSTADDR:
 		case SIOCSIFDSTADDR:
 		case SIOCSIFFLAGS:
-			if (sk->sk_net != &init_net)
+			if (sock_net(sk) != &init_net)
 				return -ENOIOCTLCMD;
 			return inet_dgram_ops.ioctl(sock, cmd, arg);
 #endif
diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c
index 1a7f143cf741..92d85c38e4d2 100644
--- a/net/rose/af_rose.c
+++ b/net/rose/af_rose.c
@@ -551,7 +551,7 @@ static struct sock *rose_make_new(struct sock *osk)
 	if (osk->sk_type != SOCK_SEQPACKET)
 		return NULL;
 
-	sk = sk_alloc(osk->sk_net, PF_ROSE, GFP_ATOMIC, &rose_proto);
+	sk = sk_alloc(sock_net(osk), PF_ROSE, GFP_ATOMIC, &rose_proto);
 	if (sk == NULL)
 		return NULL;
 
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index 0b8eb235bc13..74e662cbb2c5 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -951,7 +951,7 @@ done:
 
 static int tc_ctl_action(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct nlattr *tca[TCA_ACT_MAX + 1];
 	u32 pid = skb ? NETLINK_CB(skb).pid : 0;
 	int ret = 0, ovr = 0;
@@ -1029,7 +1029,7 @@ find_dump_kind(struct nlmsghdr *n)
 static int
 tc_dump_action(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct nlmsghdr *nlh;
 	unsigned char *b = skb_tail_pointer(skb);
 	struct nlattr *nest;
diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c
index 0fbedcabf111..1086df7478bc 100644
--- a/net/sched/cls_api.c
+++ b/net/sched/cls_api.c
@@ -118,7 +118,7 @@ static inline u32 tcf_auto_prio(struct tcf_proto *tp)
 
 static int tc_ctl_tfilter(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct nlattr *tca[TCA_MAX + 1];
 	struct tcmsg *t;
 	u32 protocol;
@@ -389,7 +389,7 @@ static int tcf_node_dump(struct tcf_proto *tp, unsigned long n,
 
 static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int t;
 	int s_t;
 	struct net_device *dev;
diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c
index 7e3c048ba9b1..15b91a9ee8e8 100644
--- a/net/sched/sch_api.c
+++ b/net/sched/sch_api.c
@@ -605,7 +605,7 @@ check_loop_fn(struct Qdisc *q, unsigned long cl, struct qdisc_walker *w)
 
 static int tc_get_qdisc(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct tcmsg *tcm = NLMSG_DATA(n);
 	struct nlattr *tca[TCA_MAX + 1];
 	struct net_device *dev;
@@ -674,7 +674,7 @@ static int tc_get_qdisc(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
 
 static int tc_modify_qdisc(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct tcmsg *tcm;
 	struct nlattr *tca[TCA_MAX + 1];
 	struct net_device *dev;
@@ -893,7 +893,7 @@ err_out:
 
 static int tc_dump_qdisc(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int idx, q_idx;
 	int s_idx, s_q_idx;
 	struct net_device *dev;
@@ -945,7 +945,7 @@ done:
 
 static int tc_ctl_tclass(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	struct tcmsg *tcm = NLMSG_DATA(n);
 	struct nlattr *tca[TCA_MAX + 1];
 	struct net_device *dev;
@@ -1139,7 +1139,7 @@ static int qdisc_class_dump(struct Qdisc *q, unsigned long cl, struct qdisc_walk
 
 static int tc_dump_tclass(struct sk_buff *skb, struct netlink_callback *cb)
 {
-	struct net *net = skb->sk->sk_net;
+	struct net *net = sock_net(skb->sk);
 	int t;
 	int s_t;
 	struct net_device *dev;
diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c
index dc71d0d83753..036bfcc8d15b 100644
--- a/net/sctp/ipv6.c
+++ b/net/sctp/ipv6.c
@@ -636,7 +636,7 @@ static struct sock *sctp_v6_create_accept_sk(struct sock *sk,
 	struct ipv6_pinfo *newnp, *np = inet6_sk(sk);
 	struct sctp6_sock *newsctp6sk;
 
-	newsk = sk_alloc(sk->sk_net, PF_INET6, GFP_KERNEL, sk->sk_prot);
+	newsk = sk_alloc(sock_net(sk), PF_INET6, GFP_KERNEL, sk->sk_prot);
 	if (!newsk)
 		goto out;
 
diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c
index 2faa0d8839eb..5aea91137fbb 100644
--- a/net/sctp/protocol.c
+++ b/net/sctp/protocol.c
@@ -554,7 +554,7 @@ static struct sock *sctp_v4_create_accept_sk(struct sock *sk,
 {
 	struct inet_sock *inet = inet_sk(sk);
 	struct inet_sock *newinet;
-	struct sock *newsk = sk_alloc(sk->sk_net, PF_INET, GFP_KERNEL,
+	struct sock *newsk = sk_alloc(sock_net(sk), PF_INET, GFP_KERNEL,
 			sk->sk_prot);
 
 	if (!newsk)
diff --git a/net/socket.c b/net/socket.c
index 9d3fbfbc8535..79e5382fd110 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -857,7 +857,7 @@ static long sock_ioctl(struct file *file, unsigned cmd, unsigned long arg)
 
 	sock = file->private_data;
 	sk = sock->sk;
-	net = sk->sk_net;
+	net = sock_net(sk);
 	if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15)) {
 		err = dev_ioctl(net, cmd, argp);
 	} else
@@ -1375,7 +1375,7 @@ asmlinkage long sys_listen(int fd, int backlog)
 
 	sock = sockfd_lookup_light(fd, &err, &fput_needed);
 	if (sock) {
-		somaxconn = sock->sk->sk_net->sysctl_somaxconn;
+		somaxconn = sock_net(sock->sk)->sysctl_somaxconn;
 		if ((unsigned)backlog > somaxconn)
 			backlog = somaxconn;
 
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 3220d5cb5b5d..ae45df060e3a 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -1375,7 +1375,7 @@ static int accept(struct socket *sock, struct socket *newsock, int flags)
 	}
 	buf = skb_peek(&sock->sk->sk_receive_queue);
 
-	res = tipc_create(sock->sk->sk_net, newsock, 0);
+	res = tipc_create(sock_net(sock->sk), newsock, 0);
 	if (!res) {
 		struct tipc_sock *new_tsock = tipc_sk(newsock->sk);
 		struct tipc_portid id;
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index ae584356852c..cb9d0cb5f270 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -252,7 +252,7 @@ static struct sock *__unix_find_socket_byname(struct net *net,
 	sk_for_each(s, node, &unix_socket_table[hash ^ type]) {
 		struct unix_sock *u = unix_sk(s);
 
-		if (s->sk_net != net)
+		if (sock_net(s) != net)
 			continue;
 
 		if (u->addr->len == len &&
@@ -289,7 +289,7 @@ static struct sock *unix_find_socket_byinode(struct net *net, struct inode *i)
 		    &unix_socket_table[i->i_ino & (UNIX_HASH_SIZE - 1)]) {
 		struct dentry *dentry = unix_sk(s)->dentry;
 
-		if (s->sk_net != net)
+		if (sock_net(s) != net)
 			continue;
 
 		if(dentry && dentry->d_inode == i)
@@ -654,7 +654,7 @@ static int unix_release(struct socket *sock)
 static int unix_autobind(struct socket *sock)
 {
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	struct unix_sock *u = unix_sk(sk);
 	static u32 ordernum = 1;
 	struct unix_address * addr;
@@ -758,7 +758,7 @@ fail:
 static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
 {
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	struct unix_sock *u = unix_sk(sk);
 	struct sockaddr_un *sunaddr=(struct sockaddr_un *)uaddr;
 	struct dentry * dentry = NULL;
@@ -899,7 +899,7 @@ static int unix_dgram_connect(struct socket *sock, struct sockaddr *addr,
 			      int alen, int flags)
 {
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	struct sockaddr_un *sunaddr=(struct sockaddr_un*)addr;
 	struct sock *other;
 	unsigned hash;
@@ -996,7 +996,7 @@ static int unix_stream_connect(struct socket *sock, struct sockaddr *uaddr,
 {
 	struct sockaddr_un *sunaddr=(struct sockaddr_un *)uaddr;
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	struct unix_sock *u = unix_sk(sk), *newu, *otheru;
 	struct sock *newsk = NULL;
 	struct sock *other = NULL;
@@ -1025,7 +1025,7 @@ static int unix_stream_connect(struct socket *sock, struct sockaddr *uaddr,
 	err = -ENOMEM;
 
 	/* create new sock for complete connection */
-	newsk = unix_create1(sk->sk_net, NULL);
+	newsk = unix_create1(sock_net(sk), NULL);
 	if (newsk == NULL)
 		goto out;
 
@@ -1312,7 +1312,7 @@ static int unix_dgram_sendmsg(struct kiocb *kiocb, struct socket *sock,
 {
 	struct sock_iocb *siocb = kiocb_to_siocb(kiocb);
 	struct sock *sk = sock->sk;
-	struct net *net = sk->sk_net;
+	struct net *net = sock_net(sk);
 	struct unix_sock *u = unix_sk(sk);
 	struct sockaddr_un *sunaddr=msg->msg_name;
 	struct sock *other = NULL;
@@ -2022,7 +2022,7 @@ static struct sock *unix_seq_idx(struct unix_iter_state *iter, loff_t pos)
 	struct sock *s;
 
 	for (s = first_unix_socket(&iter->i); s; s = next_unix_socket(&iter->i, s)) {
-		if (s->sk_net != iter->p.net)
+		if (sock_net(s) != iter->p.net)
 			continue;
 		if (off == pos)
 			return s;
@@ -2050,7 +2050,7 @@ static void *unix_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 		sk = first_unix_socket(&iter->i);
 	else
 		sk = next_unix_socket(&iter->i, sk);
-	while (sk && (sk->sk_net != iter->p.net))
+	while (sk && (sock_net(sk) != iter->p.net))
 		sk = next_unix_socket(&iter->i, sk);
 	return sk;
 }
diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c
index 7a46ea73fe2d..6ba67c523c16 100644
--- a/net/x25/af_x25.c
+++ b/net/x25/af_x25.c
@@ -549,7 +549,7 @@ static struct sock *x25_make_new(struct sock *osk)
 	if (osk->sk_type != SOCK_SEQPACKET)
 		goto out;
 
-	if ((sk = x25_alloc_socket(osk->sk_net)) == NULL)
+	if ((sk = x25_alloc_socket(sock_net(osk))) == NULL)
 		goto out;
 
 	x25 = x25_sk(sk);
-- 
cgit v1.2.3


From 1218854afa6f659be90b748cf1bc7badee954a35 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Wed, 26 Mar 2008 02:36:06 +0900
Subject: [NET] NETNS: Omit seq_net_private->net without CONFIG_NET_NS.

Without CONFIG_NET_NS, no namespace other than &init_net exists,
no need to store net in seq_net_private.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 fs/proc/proc_net.c       |  6 +++---
 include/linux/seq_file.h |  7 +++++++
 net/core/neighbour.c     |  8 ++++----
 net/ipv4/fib_hash.c      |  5 ++---
 net/ipv4/fib_trie.c      | 13 ++++++-------
 net/ipv4/raw.c           |  4 ++--
 net/ipv4/route.c         | 29 +++++++++++++++--------------
 net/ipv6/addrconf.c      |  4 ++--
 net/ipv6/mcast.c         |  4 ++--
 net/netfilter/x_tables.c |  4 ++--
 net/netlink/af_netlink.c |  6 +++---
 net/unix/af_unix.c       | 10 +++++-----
 12 files changed, 53 insertions(+), 47 deletions(-)

(limited to 'include/linux')

diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c
index 4caa5f774fb7..13cd7835d0df 100644
--- a/fs/proc/proc_net.c
+++ b/fs/proc/proc_net.c
@@ -44,7 +44,9 @@ int seq_open_net(struct inode *ino, struct file *f,
 		put_net(net);
 		return -ENOMEM;
 	}
+#ifdef CONFIG_NET_NS
 	p->net = net;
+#endif
 	return 0;
 }
 EXPORT_SYMBOL_GPL(seq_open_net);
@@ -52,12 +54,10 @@ EXPORT_SYMBOL_GPL(seq_open_net);
 int seq_release_net(struct inode *ino, struct file *f)
 {
 	struct seq_file *seq;
-	struct seq_net_private *p;
 
 	seq = f->private_data;
-	p = seq->private;
 
-	put_net(p->net);
+	put_net(seq_file_net(seq));
 	seq_release_private(ino, f);
 	return 0;
 }
diff --git a/include/linux/seq_file.h b/include/linux/seq_file.h
index 67c2563961f3..d870a8253769 100644
--- a/include/linux/seq_file.h
+++ b/include/linux/seq_file.h
@@ -5,6 +5,7 @@
 #include <linux/types.h>
 #include <linux/string.h>
 #include <linux/mutex.h>
+#include <net/net_namespace.h>
 
 struct seq_operations;
 struct file;
@@ -64,7 +65,9 @@ extern struct list_head *seq_list_next(void *v, struct list_head *head,
 
 struct net;
 struct seq_net_private {
+#ifdef CONFIG_NET_NS
 	struct net *net;
+#endif
 };
 
 int seq_open_net(struct inode *, struct file *,
@@ -72,7 +75,11 @@ int seq_open_net(struct inode *, struct file *,
 int seq_release_net(struct inode *, struct file *);
 static inline struct net *seq_file_net(struct seq_file *seq)
 {
+#ifdef CONFIG_NET_NS
 	return ((struct seq_net_private *)seq->private)->net;
+#else
+	return &init_net;
+#endif
 }
 
 #endif
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index 065fbac7ecd3..b8d491fb4b42 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -2145,7 +2145,7 @@ EXPORT_SYMBOL(__neigh_for_each_release);
 static struct neighbour *neigh_get_first(struct seq_file *seq)
 {
 	struct neigh_seq_state *state = seq->private;
-	struct net *net = state->p.net;
+	struct net *net = seq_file_net(seq);
 	struct neigh_table *tbl = state->tbl;
 	struct neighbour *n = NULL;
 	int bucket = state->bucket;
@@ -2186,7 +2186,7 @@ static struct neighbour *neigh_get_next(struct seq_file *seq,
 					loff_t *pos)
 {
 	struct neigh_seq_state *state = seq->private;
-	struct net *net = state->p.net;
+	struct net *net = seq_file_net(seq);
 	struct neigh_table *tbl = state->tbl;
 
 	if (state->neigh_sub_iter) {
@@ -2246,7 +2246,7 @@ static struct neighbour *neigh_get_idx(struct seq_file *seq, loff_t *pos)
 static struct pneigh_entry *pneigh_get_first(struct seq_file *seq)
 {
 	struct neigh_seq_state *state = seq->private;
-	struct net * net = state->p.net;
+	struct net *net = seq_file_net(seq);
 	struct neigh_table *tbl = state->tbl;
 	struct pneigh_entry *pn = NULL;
 	int bucket = state->bucket;
@@ -2269,7 +2269,7 @@ static struct pneigh_entry *pneigh_get_next(struct seq_file *seq,
 					    loff_t *pos)
 {
 	struct neigh_seq_state *state = seq->private;
-	struct net * net = state->p.net;
+	struct net *net = seq_file_net(seq);
 	struct neigh_table *tbl = state->tbl;
 
 	pn = pn->next;
diff --git a/net/ipv4/fib_hash.c b/net/ipv4/fib_hash.c
index 8d58d85dfac6..02088deb0461 100644
--- a/net/ipv4/fib_hash.c
+++ b/net/ipv4/fib_hash.c
@@ -821,7 +821,7 @@ static struct fib_alias *fib_get_first(struct seq_file *seq)
 	struct fib_table *main_table;
 	struct fn_hash *table;
 
-	main_table = fib_get_table(iter->p.net, RT_TABLE_MAIN);
+	main_table = fib_get_table(seq_file_net(seq), RT_TABLE_MAIN);
 	table = (struct fn_hash *)main_table->tb_data;
 
 	iter->bucket    = 0;
@@ -959,11 +959,10 @@ static struct fib_alias *fib_get_idx(struct seq_file *seq, loff_t pos)
 static void *fib_seq_start(struct seq_file *seq, loff_t *pos)
 	__acquires(fib_hash_lock)
 {
-	struct fib_iter_state *iter = seq->private;
 	void *v = NULL;
 
 	read_lock(&fib_hash_lock);
-	if (fib_get_table(iter->p.net, RT_TABLE_MAIN))
+	if (fib_get_table(seq_file_net(seq), RT_TABLE_MAIN))
 		v = *pos ? fib_get_idx(seq, *pos - 1) : SEQ_START_TOKEN;
 	return v;
 }
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index ce6cb34e28e1..9e491e70e855 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -2279,9 +2279,10 @@ static const struct file_operations fib_triestat_fops = {
 	.release = fib_triestat_seq_release,
 };
 
-static struct node *fib_trie_get_idx(struct fib_trie_iter *iter, loff_t pos)
+static struct node *fib_trie_get_idx(struct seq_file *seq, loff_t pos)
 {
-	struct net *net = iter->p.net;
+	struct fib_trie_iter *iter = seq->private;
+	struct net *net = seq_file_net(seq);
 	loff_t idx = 0;
 	unsigned int h;
 
@@ -2309,16 +2310,14 @@ static struct node *fib_trie_get_idx(struct fib_trie_iter *iter, loff_t pos)
 static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos)
 	__acquires(RCU)
 {
-	struct fib_trie_iter *iter = seq->private;
-
 	rcu_read_lock();
-	return fib_trie_get_idx(iter, *pos);
+	return fib_trie_get_idx(seq, *pos);
 }
 
 static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 {
 	struct fib_trie_iter *iter = seq->private;
-	struct net *net = iter->p.net;
+	struct net *net = seq_file_net(seq);
 	struct fib_table *tb = iter->tb;
 	struct hlist_node *tb_node;
 	unsigned int h;
@@ -2513,7 +2512,7 @@ static void *fib_route_seq_start(struct seq_file *seq, loff_t *pos)
 	struct fib_table *tb;
 
 	rcu_read_lock();
-	tb = fib_get_table(iter->p.net, RT_TABLE_MAIN);
+	tb = fib_get_table(seq_file_net(seq), RT_TABLE_MAIN);
 	if (!tb)
 		return NULL;
 
diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c
index be19a4048d7c..25dc8b38cac3 100644
--- a/net/ipv4/raw.c
+++ b/net/ipv4/raw.c
@@ -856,7 +856,7 @@ static struct sock *raw_get_first(struct seq_file *seq)
 		struct hlist_node *node;
 
 		sk_for_each(sk, node, &state->h->ht[state->bucket])
-			if (sock_net(sk) == state->p.net)
+			if (sock_net(sk) == seq_file_net(seq))
 				goto found;
 	}
 	sk = NULL;
@@ -872,7 +872,7 @@ static struct sock *raw_get_next(struct seq_file *seq, struct sock *sk)
 		sk = sk_next(sk);
 try_again:
 		;
-	} while (sk && sock_net(sk) != state->p.net);
+	} while (sk && sock_net(sk) != seq_file_net(seq));
 
 	if (!sk && ++state->bucket < RAW_HTABLE_SIZE) {
 		sk = sk_head(&state->h->ht[state->bucket]);
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 194f5cca3121..eab8d75e5222 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -276,15 +276,16 @@ struct rt_cache_iter_state {
 	int genid;
 };
 
-static struct rtable *rt_cache_get_first(struct rt_cache_iter_state *st)
+static struct rtable *rt_cache_get_first(struct seq_file *seq)
 {
+	struct rt_cache_iter_state *st = seq->private;
 	struct rtable *r = NULL;
 
 	for (st->bucket = rt_hash_mask; st->bucket >= 0; --st->bucket) {
 		rcu_read_lock_bh();
 		r = rcu_dereference(rt_hash_table[st->bucket].chain);
 		while (r) {
-			if (dev_net(r->u.dst.dev) == st->p.net &&
+			if (dev_net(r->u.dst.dev) == seq_file_net(seq) &&
 			    r->rt_genid == st->genid)
 				return r;
 			r = rcu_dereference(r->u.dst.rt_next);
@@ -294,9 +295,10 @@ static struct rtable *rt_cache_get_first(struct rt_cache_iter_state *st)
 	return r;
 }
 
-static struct rtable *__rt_cache_get_next(struct rt_cache_iter_state *st,
+static struct rtable *__rt_cache_get_next(struct seq_file *seq,
 					  struct rtable *r)
 {
+	struct rt_cache_iter_state *st = seq->private;
 	r = r->u.dst.rt_next;
 	while (!r) {
 		rcu_read_unlock_bh();
@@ -308,11 +310,12 @@ static struct rtable *__rt_cache_get_next(struct rt_cache_iter_state *st,
 	return rcu_dereference(r);
 }
 
-static struct rtable *rt_cache_get_next(struct rt_cache_iter_state *st,
+static struct rtable *rt_cache_get_next(struct seq_file *seq,
 					struct rtable *r)
 {
-	while ((r = __rt_cache_get_next(st, r)) != NULL) {
-		if (dev_net(r->u.dst.dev) != st->p.net)
+	struct rt_cache_iter_state *st = seq->private;
+	while ((r = __rt_cache_get_next(seq, r)) != NULL) {
+		if (dev_net(r->u.dst.dev) != seq_file_net(seq))
 			continue;
 		if (r->rt_genid == st->genid)
 			break;
@@ -320,12 +323,12 @@ static struct rtable *rt_cache_get_next(struct rt_cache_iter_state *st,
 	return r;
 }
 
-static struct rtable *rt_cache_get_idx(struct rt_cache_iter_state *st, loff_t pos)
+static struct rtable *rt_cache_get_idx(struct seq_file *seq, loff_t pos)
 {
-	struct rtable *r = rt_cache_get_first(st);
+	struct rtable *r = rt_cache_get_first(seq);
 
 	if (r)
-		while (pos && (r = rt_cache_get_next(st, r)))
+		while (pos && (r = rt_cache_get_next(seq, r)))
 			--pos;
 	return pos ? NULL : r;
 }
@@ -333,9 +336,8 @@ static struct rtable *rt_cache_get_idx(struct rt_cache_iter_state *st, loff_t po
 static void *rt_cache_seq_start(struct seq_file *seq, loff_t *pos)
 {
 	struct rt_cache_iter_state *st = seq->private;
-
 	if (*pos)
-		return rt_cache_get_idx(st, *pos - 1);
+		return rt_cache_get_idx(seq, *pos - 1);
 	st->genid = atomic_read(&rt_genid);
 	return SEQ_START_TOKEN;
 }
@@ -343,12 +345,11 @@ static void *rt_cache_seq_start(struct seq_file *seq, loff_t *pos)
 static void *rt_cache_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 {
 	struct rtable *r;
-	struct rt_cache_iter_state *st = seq->private;
 
 	if (v == SEQ_START_TOKEN)
-		r = rt_cache_get_first(st);
+		r = rt_cache_get_first(seq);
 	else
-		r = rt_cache_get_next(st, v);
+		r = rt_cache_get_next(seq, v);
 	++*pos;
 	return r;
 }
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index f2c90f145cbb..ac5d4f4b6312 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -2766,7 +2766,7 @@ static struct inet6_ifaddr *if6_get_first(struct seq_file *seq)
 {
 	struct inet6_ifaddr *ifa = NULL;
 	struct if6_iter_state *state = seq->private;
-	struct net *net = state->p.net;
+	struct net *net = seq_file_net(seq);
 
 	for (state->bucket = 0; state->bucket < IN6_ADDR_HSIZE; ++state->bucket) {
 		ifa = inet6_addr_lst[state->bucket];
@@ -2782,7 +2782,7 @@ static struct inet6_ifaddr *if6_get_first(struct seq_file *seq)
 static struct inet6_ifaddr *if6_get_next(struct seq_file *seq, struct inet6_ifaddr *ifa)
 {
 	struct if6_iter_state *state = seq->private;
-	struct net *net = state->p.net;
+	struct net *net = seq_file_net(seq);
 
 	ifa = ifa->lst_next;
 try_again:
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 20a3d8e2f6c6..d810cff818cf 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -2355,7 +2355,7 @@ static inline struct ifmcaddr6 *igmp6_mc_get_first(struct seq_file *seq)
 {
 	struct ifmcaddr6 *im = NULL;
 	struct igmp6_mc_iter_state *state = igmp6_mc_seq_private(seq);
-	struct net *net = state->p.net;
+	struct net *net = seq_file_net(seq);
 
 	state->idev = NULL;
 	for_each_netdev(net, state->dev) {
@@ -2486,7 +2486,7 @@ static inline struct ip6_sf_list *igmp6_mcf_get_first(struct seq_file *seq)
 	struct ip6_sf_list *psf = NULL;
 	struct ifmcaddr6 *im = NULL;
 	struct igmp6_mcf_iter_state *state = igmp6_mcf_seq_private(seq);
-	struct net *net = state->p.net;
+	struct net *net = seq_file_net(seq);
 
 	state->idev = NULL;
 	state->im = NULL;
diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index a6792089fcf9..0bd95680a494 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -727,7 +727,7 @@ struct xt_names_priv {
 static void *xt_table_seq_start(struct seq_file *seq, loff_t *pos)
 {
 	struct xt_names_priv *priv = seq->private;
-	struct net *net = priv->p.net;
+	struct net *net = seq_file_net(seq);
 	int af = priv->af;
 
 	mutex_lock(&xt[af].mutex);
@@ -737,7 +737,7 @@ static void *xt_table_seq_start(struct seq_file *seq, loff_t *pos)
 static void *xt_table_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 {
 	struct xt_names_priv *priv = seq->private;
-	struct net *net = priv->p.net;
+	struct net *net = seq_file_net(seq);
 	int af = priv->af;
 
 	return seq_list_next(v, &net->xt.tables[af], pos);
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 712a7bff8560..1d16d95dfaaf 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -1758,7 +1758,7 @@ static struct sock *netlink_seq_socket_idx(struct seq_file *seq, loff_t pos)
 
 		for (j = 0; j <= hash->mask; j++) {
 			sk_for_each(s, node, &hash->table[j]) {
-				if (sock_net(s) != iter->p.net)
+				if (sock_net(s) != seq_file_net(seq))
 					continue;
 				if (off == pos) {
 					iter->link = i;
@@ -1794,7 +1794,7 @@ static void *netlink_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 	s = v;
 	do {
 		s = sk_next(s);
-	} while (s && (sock_net(s) != iter->p.net));
+	} while (s && sock_net(s) != seq_file_net(seq));
 	if (s)
 		return s;
 
@@ -1806,7 +1806,7 @@ static void *netlink_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 
 		for (; j <= hash->mask; j++) {
 			s = sk_head(&hash->table[j]);
-			while (s && sock_net(s) != iter->p.net)
+			while (s && sock_net(s) != seq_file_net(seq))
 				s = sk_next(s);
 			if (s) {
 				iter->link = i;
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index cb9d0cb5f270..4a4793051bcb 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -2016,13 +2016,14 @@ struct unix_iter_state {
 	struct seq_net_private p;
 	int i;
 };
-static struct sock *unix_seq_idx(struct unix_iter_state *iter, loff_t pos)
+static struct sock *unix_seq_idx(struct seq_file *seq, loff_t pos)
 {
+	struct unix_iter_state *iter = seq->private;
 	loff_t off = 0;
 	struct sock *s;
 
 	for (s = first_unix_socket(&iter->i); s; s = next_unix_socket(&iter->i, s)) {
-		if (sock_net(s) != iter->p.net)
+		if (sock_net(s) != seq_file_net(seq))
 			continue;
 		if (off == pos)
 			return s;
@@ -2035,9 +2036,8 @@ static struct sock *unix_seq_idx(struct unix_iter_state *iter, loff_t pos)
 static void *unix_seq_start(struct seq_file *seq, loff_t *pos)
 	__acquires(unix_table_lock)
 {
-	struct unix_iter_state *iter = seq->private;
 	spin_lock(&unix_table_lock);
-	return *pos ? unix_seq_idx(iter, *pos - 1) : ((void *) 1);
+	return *pos ? unix_seq_idx(seq, *pos - 1) : ((void *) 1);
 }
 
 static void *unix_seq_next(struct seq_file *seq, void *v, loff_t *pos)
@@ -2050,7 +2050,7 @@ static void *unix_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 		sk = first_unix_socket(&iter->i);
 	else
 		sk = next_unix_socket(&iter->i, sk);
-	while (sk && (sock_net(sk) != iter->p.net))
+	while (sk && (sock_net(sk) != seq_file_net(seq)))
 		sk = next_unix_socket(&iter->i, sk);
 	return sk;
 }
-- 
cgit v1.2.3


From b8beedd25d3913d45b8330a08ab88fdf90eb54b8 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:09:33 -0700
Subject: [NETFILTER]: Add nf_inet_addr_cmp()

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter.h                  |  9 +++++++++
 include/net/netfilter/nf_conntrack_tuple.h | 15 +++------------
 2 files changed, 12 insertions(+), 12 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h
index f0680c2bee73..89e6c72ad295 100644
--- a/include/linux/netfilter.h
+++ b/include/linux/netfilter.h
@@ -61,6 +61,15 @@ union nf_inet_addr {
 #ifdef __KERNEL__
 #ifdef CONFIG_NETFILTER
 
+static inline int nf_inet_addr_cmp(const union nf_inet_addr *a1,
+				   const union nf_inet_addr *a2)
+{
+	return a1->all[0] == a2->all[0] &&
+	       a1->all[1] == a2->all[1] &&
+	       a1->all[2] == a2->all[2] &&
+	       a1->all[3] == a2->all[3];
+}
+
 extern void netfilter_init(void);
 
 /* Largest hook number + 1 */
diff --git a/include/net/netfilter/nf_conntrack_tuple.h b/include/net/netfilter/nf_conntrack_tuple.h
index d9b53dd601ad..168c91754d89 100644
--- a/include/net/netfilter/nf_conntrack_tuple.h
+++ b/include/net/netfilter/nf_conntrack_tuple.h
@@ -163,10 +163,7 @@ struct nf_conntrack_tuple_hash
 static inline int __nf_ct_tuple_src_equal(const struct nf_conntrack_tuple *t1,
 					  const struct nf_conntrack_tuple *t2)
 { 
-	return (t1->src.u3.all[0] == t2->src.u3.all[0] &&
-		t1->src.u3.all[1] == t2->src.u3.all[1] &&
-		t1->src.u3.all[2] == t2->src.u3.all[2] &&
-		t1->src.u3.all[3] == t2->src.u3.all[3] &&
+	return (nf_inet_addr_cmp(&t1->src.u3, &t2->src.u3) &&
 		t1->src.u.all == t2->src.u.all &&
 		t1->src.l3num == t2->src.l3num);
 }
@@ -174,10 +171,7 @@ static inline int __nf_ct_tuple_src_equal(const struct nf_conntrack_tuple *t1,
 static inline int __nf_ct_tuple_dst_equal(const struct nf_conntrack_tuple *t1,
 					  const struct nf_conntrack_tuple *t2)
 {
-	return (t1->dst.u3.all[0] == t2->dst.u3.all[0] &&
-		t1->dst.u3.all[1] == t2->dst.u3.all[1] &&
-		t1->dst.u3.all[2] == t2->dst.u3.all[2] &&
-		t1->dst.u3.all[3] == t2->dst.u3.all[3] &&
+	return (nf_inet_addr_cmp(&t1->dst.u3, &t2->dst.u3) &&
 		t1->dst.u.all == t2->dst.u.all &&
 		t1->dst.protonum == t2->dst.protonum);
 }
@@ -192,10 +186,7 @@ static inline int nf_ct_tuple_equal(const struct nf_conntrack_tuple *t1,
 static inline int nf_ct_tuple_mask_equal(const struct nf_conntrack_tuple_mask *m1,
 					 const struct nf_conntrack_tuple_mask *m2)
 {
-	return (m1->src.u3.all[0] == m2->src.u3.all[0] &&
-		m1->src.u3.all[1] == m2->src.u3.all[1] &&
-		m1->src.u3.all[2] == m2->src.u3.all[2] &&
-		m1->src.u3.all[3] == m2->src.u3.all[3] &&
+	return (nf_inet_addr_cmp(&m1->src.u3, &m2->src.u3) &&
 		m1->src.u.all == m2->src.u.all);
 }
 
-- 
cgit v1.2.3


From 2a6cfb22ae002330d445f734668d9158db9e90de Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:16:54 -0700
Subject: [NETFILTER]: nf_conntrack_sip: adjust dptr and datalen after packet
 mangling

After mangling the packet, the pointer to the data and the length of the data
portion may change and need to be adjusted.

Use double data pointers and a pointer to the length everywhere and add a
helper function to the NAT helper for performing the adjustments.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  6 +-
 net/ipv4/netfilter/nf_nat_sip.c            | 91 ++++++++++++++++--------------
 net/netfilter/nf_conntrack_sip.c           | 14 +++--
 3 files changed, 60 insertions(+), 51 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 8e5ce1ca7bfc..9d0dbfb26300 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -24,11 +24,13 @@ enum sip_header_pos {
 extern unsigned int (*nf_nat_sip_hook)(struct sk_buff *skb,
 				       enum ip_conntrack_info ctinfo,
 				       struct nf_conn *ct,
-				       const char **dptr);
+				       const char **dptr,
+				       unsigned int *datalen);
 extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				       enum ip_conntrack_info ctinfo,
 				       struct nf_conntrack_expect *exp,
-				       const char *dptr);
+				       const char **dptr,
+				       unsigned int *datalen);
 
 extern int ct_sip_get_info(const struct nf_conn *ct, const char *dptr,
                            size_t dlen, unsigned int *matchoff,
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index 84d8b4982cdf..e77122e65283 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -60,15 +60,35 @@ static void addr_map_init(const struct nf_conn *ct, struct addr_map *map)
 	}
 }
 
+static unsigned int mangle_packet(struct sk_buff *skb,
+				  const char **dptr, unsigned int *datalen,
+				  unsigned int matchoff, unsigned int matchlen,
+				  const char *buffer, unsigned int buflen)
+{
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+
+	if (!nf_nat_mangle_udp_packet(skb, ct, ctinfo, matchoff, matchlen,
+				      buffer, buflen))
+		return 0;
+
+	/* Reload data pointer and adjust datalen value */
+	*dptr = skb->data + ip_hdrlen(skb) + sizeof(struct udphdr);
+	*datalen += buflen - matchlen;
+	return 1;
+}
+
 static int map_sip_addr(struct sk_buff *skb, enum ip_conntrack_info ctinfo,
-			struct nf_conn *ct, const char **dptr, size_t dlen,
+			struct nf_conn *ct,
+			const char **dptr, unsigned int *datalen,
 			enum sip_header_pos pos, struct addr_map *map)
 {
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
 	unsigned int matchlen, matchoff, addrlen;
 	char *addr;
 
-	if (ct_sip_get_info(ct, *dptr, dlen, &matchoff, &matchlen, pos) <= 0)
+	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
+			    pos) <= 0)
 		return 1;
 
 	if ((matchlen == map->addr[dir].srciplen ||
@@ -84,26 +104,19 @@ static int map_sip_addr(struct sk_buff *skb, enum ip_conntrack_info ctinfo,
 	} else
 		return 1;
 
-	if (!nf_nat_mangle_udp_packet(skb, ct, ctinfo,
-				      matchoff, matchlen, addr, addrlen))
-		return 0;
-	*dptr = skb->data + ip_hdrlen(skb) + sizeof(struct udphdr);
-	return 1;
-
+	return mangle_packet(skb, dptr, datalen, matchoff, matchlen,
+			     addr, addrlen);
 }
 
 static unsigned int ip_nat_sip(struct sk_buff *skb,
 			       enum ip_conntrack_info ctinfo,
 			       struct nf_conn *ct,
-			       const char **dptr)
+			       const char **dptr, unsigned int *datalen)
 {
 	enum sip_header_pos pos;
 	struct addr_map map;
-	int dataoff, datalen;
 
-	dataoff = ip_hdrlen(skb) + sizeof(struct udphdr);
-	datalen = skb->len - dataoff;
-	if (datalen < sizeof("SIP/2.0") - 1)
+	if (*datalen < sizeof("SIP/2.0") - 1)
 		return NF_ACCEPT;
 
 	addr_map_init(ct, &map);
@@ -115,7 +128,7 @@ static unsigned int ip_nat_sip(struct sk_buff *skb,
 		 * The "userinfo" and "@" components of the SIP URI MUST NOT
 		 * be present.
 		 */
-		if (datalen >= sizeof("REGISTER") - 1 &&
+		if (*datalen >= sizeof("REGISTER") - 1 &&
 		    strncmp(*dptr, "REGISTER", sizeof("REGISTER") - 1) == 0)
 			pos = POS_REG_REQ_URI;
 		else
@@ -136,51 +149,45 @@ static unsigned int ip_nat_sip(struct sk_buff *skb,
 static unsigned int mangle_sip_packet(struct sk_buff *skb,
 				      enum ip_conntrack_info ctinfo,
 				      struct nf_conn *ct,
-				      const char **dptr, size_t dlen,
+				      const char **dptr, unsigned int *datalen,
 				      char *buffer, int bufflen,
 				      enum sip_header_pos pos)
 {
 	unsigned int matchlen, matchoff;
 
-	if (ct_sip_get_info(ct, *dptr, dlen, &matchoff, &matchlen, pos) <= 0)
-		return 0;
-
-	if (!nf_nat_mangle_udp_packet(skb, ct, ctinfo,
-				      matchoff, matchlen, buffer, bufflen))
+	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
+			    pos) <= 0)
 		return 0;
 
-	/* We need to reload this. Thanks Patrick. */
-	*dptr = skb->data + ip_hdrlen(skb) + sizeof(struct udphdr);
-	return 1;
+	return mangle_packet(skb, dptr, datalen, matchoff, matchlen,
+			     buffer, bufflen);
 }
 
 static int mangle_content_len(struct sk_buff *skb,
 			      enum ip_conntrack_info ctinfo,
 			      struct nf_conn *ct,
-			      const char *dptr)
+			      const char **dptr, unsigned int *datalen)
 {
-	unsigned int dataoff, matchoff, matchlen;
+	unsigned int matchoff, matchlen;
 	char buffer[sizeof("65536")];
 	int bufflen;
 
-	dataoff = ip_hdrlen(skb) + sizeof(struct udphdr);
-
 	/* Get actual SDP length */
-	if (ct_sip_get_info(ct, dptr, skb->len - dataoff, &matchoff,
+	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff,
 			    &matchlen, POS_SDP_HEADER) > 0) {
 
 		/* since ct_sip_get_info() give us a pointer passing 'v='
 		   we need to add 2 bytes in this count. */
-		int c_len = skb->len - dataoff - matchoff + 2;
+		int c_len = *datalen - matchoff + 2;
 
 		/* Now, update SDP length */
-		if (ct_sip_get_info(ct, dptr, skb->len - dataoff, &matchoff,
+		if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff,
 				    &matchlen, POS_CONTENT) > 0) {
 
 			bufflen = sprintf(buffer, "%u", c_len);
-			return nf_nat_mangle_udp_packet(skb, ct, ctinfo,
-							matchoff, matchlen,
-							buffer, bufflen);
+			return mangle_packet(skb, dptr, datalen,
+					     matchoff, matchlen,
+					     buffer, bufflen);
 		}
 	}
 	return 0;
@@ -190,30 +197,28 @@ static unsigned int mangle_sdp(struct sk_buff *skb,
 			       enum ip_conntrack_info ctinfo,
 			       struct nf_conn *ct,
 			       __be32 newip, u_int16_t port,
-			       const char *dptr)
+			       const char **dptr, unsigned int *datalen)
 {
 	char buffer[sizeof("nnn.nnn.nnn.nnn")];
-	unsigned int dataoff, bufflen;
-
-	dataoff = ip_hdrlen(skb) + sizeof(struct udphdr);
+	unsigned int bufflen;
 
 	/* Mangle owner and contact info. */
 	bufflen = sprintf(buffer, "%u.%u.%u.%u", NIPQUAD(newip));
-	if (!mangle_sip_packet(skb, ctinfo, ct, &dptr, skb->len - dataoff,
+	if (!mangle_sip_packet(skb, ctinfo, ct, dptr, datalen,
 			       buffer, bufflen, POS_OWNER_IP4))
 		return 0;
 
-	if (!mangle_sip_packet(skb, ctinfo, ct, &dptr, skb->len - dataoff,
+	if (!mangle_sip_packet(skb, ctinfo, ct, dptr, datalen,
 			       buffer, bufflen, POS_CONNECTION_IP4))
 		return 0;
 
 	/* Mangle media port. */
 	bufflen = sprintf(buffer, "%u", port);
-	if (!mangle_sip_packet(skb, ctinfo, ct, &dptr, skb->len - dataoff,
+	if (!mangle_sip_packet(skb, ctinfo, ct, dptr, datalen,
 			       buffer, bufflen, POS_MEDIA))
 		return 0;
 
-	return mangle_content_len(skb, ctinfo, ct, dptr);
+	return mangle_content_len(skb, ctinfo, ct, dptr, datalen);
 }
 
 static void ip_nat_sdp_expect(struct nf_conn *ct,
@@ -242,7 +247,7 @@ static void ip_nat_sdp_expect(struct nf_conn *ct,
 static unsigned int ip_nat_sdp(struct sk_buff *skb,
 			       enum ip_conntrack_info ctinfo,
 			       struct nf_conntrack_expect *exp,
-			       const char *dptr)
+			       const char **dptr, unsigned int *datalen)
 {
 	struct nf_conn *ct = exp->master;
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
@@ -275,7 +280,7 @@ static unsigned int ip_nat_sdp(struct sk_buff *skb,
 	if (port == 0)
 		return NF_DROP;
 
-	if (!mangle_sdp(skb, ctinfo, ct, newip, port, dptr)) {
+	if (!mangle_sdp(skb, ctinfo, ct, newip, port, dptr, datalen)) {
 		nf_ct_unexpect_related(exp);
 		return NF_DROP;
 	}
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index 016e1c1aafe4..fa0d5599ff24 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -39,13 +39,15 @@ MODULE_PARM_DESC(sip_timeout, "timeout for the master SIP session");
 unsigned int (*nf_nat_sip_hook)(struct sk_buff *skb,
 				enum ip_conntrack_info ctinfo,
 				struct nf_conn *ct,
-				const char **dptr) __read_mostly;
+				const char **dptr,
+				unsigned int *datalen) __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sip_hook);
 
 unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				enum ip_conntrack_info ctinfo,
 				struct nf_conntrack_expect *exp,
-				const char *dptr) __read_mostly;
+				const char **dptr,
+				unsigned int *datalen) __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sdp_hook);
 
 static int digits_len(const struct nf_conn *, const char *, const char *, int *);
@@ -369,7 +371,7 @@ static int set_expected_rtp(struct sk_buff *skb,
 			    enum ip_conntrack_info ctinfo,
 			    union nf_inet_addr *addr,
 			    __be16 port,
-			    const char *dptr)
+			    const char **dptr, unsigned int *datalen)
 {
 	struct nf_conntrack_expect *exp;
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
@@ -386,7 +388,7 @@ static int set_expected_rtp(struct sk_buff *skb,
 
 	nf_nat_sdp = rcu_dereference(nf_nat_sdp_hook);
 	if (nf_nat_sdp && ct->status & IPS_NAT_MASK)
-		ret = nf_nat_sdp(skb, ctinfo, exp, dptr);
+		ret = nf_nat_sdp(skb, ctinfo, exp, dptr, datalen);
 	else {
 		if (nf_ct_expect_related(exp) != 0)
 			ret = NF_DROP;
@@ -429,7 +431,7 @@ static int sip_help(struct sk_buff *skb,
 
 	nf_nat_sip = rcu_dereference(nf_nat_sip_hook);
 	if (nf_nat_sip && ct->status & IPS_NAT_MASK) {
-		if (!nf_nat_sip(skb, ctinfo, ct, &dptr)) {
+		if (!nf_nat_sip(skb, ctinfo, ct, &dptr, &datalen)) {
 			ret = NF_DROP;
 			goto out;
 		}
@@ -466,7 +468,7 @@ static int sip_help(struct sk_buff *skb,
 				goto out;
 			}
 			ret = set_expected_rtp(skb, ct, ctinfo, &addr,
-					       htons(port), dptr);
+					       htons(port), &dptr, &datalen);
 		}
 	}
 out:
-- 
cgit v1.2.3


From 212440a7d04a12ee13787afecc6c86c7fc4e6184 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:17:13 -0700
Subject: [NETFILTER]: nf_conntrack_sip: remove redundant function arguments

The conntrack reference and ctinfo can be derived from the packet.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  7 ++---
 net/ipv4/netfilter/nf_nat_sip.c            | 49 +++++++++++++++---------------
 net/netfilter/nf_conntrack_sip.c           | 24 ++++++---------
 3 files changed, 37 insertions(+), 43 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 9d0dbfb26300..b94de3d60303 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -22,15 +22,12 @@ enum sip_header_pos {
 };
 
 extern unsigned int (*nf_nat_sip_hook)(struct sk_buff *skb,
-				       enum ip_conntrack_info ctinfo,
-				       struct nf_conn *ct,
 				       const char **dptr,
 				       unsigned int *datalen);
 extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
-				       enum ip_conntrack_info ctinfo,
-				       struct nf_conntrack_expect *exp,
 				       const char **dptr,
-				       unsigned int *datalen);
+				       unsigned int *datalen,
+				       struct nf_conntrack_expect *exp);
 
 extern int ct_sip_get_info(const struct nf_conn *ct, const char *dptr,
                            size_t dlen, unsigned int *matchoff,
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index e77122e65283..acaa7d4569fa 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -78,11 +78,12 @@ static unsigned int mangle_packet(struct sk_buff *skb,
 	return 1;
 }
 
-static int map_sip_addr(struct sk_buff *skb, enum ip_conntrack_info ctinfo,
-			struct nf_conn *ct,
+static int map_sip_addr(struct sk_buff *skb,
 			const char **dptr, unsigned int *datalen,
 			enum sip_header_pos pos, struct addr_map *map)
 {
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
 	unsigned int matchlen, matchoff, addrlen;
 	char *addr;
@@ -109,10 +110,10 @@ static int map_sip_addr(struct sk_buff *skb, enum ip_conntrack_info ctinfo,
 }
 
 static unsigned int ip_nat_sip(struct sk_buff *skb,
-			       enum ip_conntrack_info ctinfo,
-			       struct nf_conn *ct,
 			       const char **dptr, unsigned int *datalen)
 {
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	enum sip_header_pos pos;
 	struct addr_map map;
 
@@ -134,25 +135,25 @@ static unsigned int ip_nat_sip(struct sk_buff *skb,
 		else
 			pos = POS_REQ_URI;
 
-		if (!map_sip_addr(skb, ctinfo, ct, dptr, datalen, pos, &map))
+		if (!map_sip_addr(skb, dptr, datalen, pos, &map))
 			return NF_DROP;
 	}
 
-	if (!map_sip_addr(skb, ctinfo, ct, dptr, datalen, POS_FROM, &map) ||
-	    !map_sip_addr(skb, ctinfo, ct, dptr, datalen, POS_TO, &map) ||
-	    !map_sip_addr(skb, ctinfo, ct, dptr, datalen, POS_VIA, &map) ||
-	    !map_sip_addr(skb, ctinfo, ct, dptr, datalen, POS_CONTACT, &map))
+	if (!map_sip_addr(skb, dptr, datalen, POS_FROM, &map) ||
+	    !map_sip_addr(skb, dptr, datalen, POS_TO, &map) ||
+	    !map_sip_addr(skb, dptr, datalen, POS_VIA, &map) ||
+	    !map_sip_addr(skb, dptr, datalen, POS_CONTACT, &map))
 		return NF_DROP;
 	return NF_ACCEPT;
 }
 
 static unsigned int mangle_sip_packet(struct sk_buff *skb,
-				      enum ip_conntrack_info ctinfo,
-				      struct nf_conn *ct,
 				      const char **dptr, unsigned int *datalen,
 				      char *buffer, int bufflen,
 				      enum sip_header_pos pos)
 {
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	unsigned int matchlen, matchoff;
 
 	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
@@ -164,10 +165,10 @@ static unsigned int mangle_sip_packet(struct sk_buff *skb,
 }
 
 static int mangle_content_len(struct sk_buff *skb,
-			      enum ip_conntrack_info ctinfo,
-			      struct nf_conn *ct,
 			      const char **dptr, unsigned int *datalen)
 {
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	unsigned int matchoff, matchlen;
 	char buffer[sizeof("65536")];
 	int bufflen;
@@ -204,21 +205,21 @@ static unsigned int mangle_sdp(struct sk_buff *skb,
 
 	/* Mangle owner and contact info. */
 	bufflen = sprintf(buffer, "%u.%u.%u.%u", NIPQUAD(newip));
-	if (!mangle_sip_packet(skb, ctinfo, ct, dptr, datalen,
-			       buffer, bufflen, POS_OWNER_IP4))
+	if (!mangle_sip_packet(skb, dptr, datalen, buffer, bufflen,
+			       POS_OWNER_IP4))
 		return 0;
 
-	if (!mangle_sip_packet(skb, ctinfo, ct, dptr, datalen,
-			       buffer, bufflen, POS_CONNECTION_IP4))
+	if (!mangle_sip_packet(skb, dptr, datalen, buffer, bufflen,
+			       POS_CONNECTION_IP4))
 		return 0;
 
 	/* Mangle media port. */
 	bufflen = sprintf(buffer, "%u", port);
-	if (!mangle_sip_packet(skb, ctinfo, ct, dptr, datalen,
-			       buffer, bufflen, POS_MEDIA))
+	if (!mangle_sip_packet(skb, dptr, datalen, buffer, bufflen,
+			       POS_MEDIA))
 		return 0;
 
-	return mangle_content_len(skb, ctinfo, ct, dptr, datalen);
+	return mangle_content_len(skb, dptr, datalen);
 }
 
 static void ip_nat_sdp_expect(struct nf_conn *ct,
@@ -245,11 +246,11 @@ static void ip_nat_sdp_expect(struct nf_conn *ct,
 /* So, this packet has hit the connection tracking matching code.
    Mangle it, and change the expectation to match the new version. */
 static unsigned int ip_nat_sdp(struct sk_buff *skb,
-			       enum ip_conntrack_info ctinfo,
-			       struct nf_conntrack_expect *exp,
-			       const char **dptr, unsigned int *datalen)
+			       const char **dptr, unsigned int *datalen,
+			       struct nf_conntrack_expect *exp)
 {
-	struct nf_conn *ct = exp->master;
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
 	__be32 newip;
 	u_int16_t port;
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index fa0d5599ff24..38e1e7a05334 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -37,17 +37,14 @@ module_param(sip_timeout, uint, 0600);
 MODULE_PARM_DESC(sip_timeout, "timeout for the master SIP session");
 
 unsigned int (*nf_nat_sip_hook)(struct sk_buff *skb,
-				enum ip_conntrack_info ctinfo,
-				struct nf_conn *ct,
 				const char **dptr,
 				unsigned int *datalen) __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sip_hook);
 
 unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
-				enum ip_conntrack_info ctinfo,
-				struct nf_conntrack_expect *exp,
 				const char **dptr,
-				unsigned int *datalen) __read_mostly;
+				unsigned int *datalen,
+				struct nf_conntrack_expect *exp) __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sdp_hook);
 
 static int digits_len(const struct nf_conn *, const char *, const char *, int *);
@@ -367,13 +364,12 @@ int ct_sip_get_info(const struct nf_conn *ct,
 EXPORT_SYMBOL_GPL(ct_sip_get_info);
 
 static int set_expected_rtp(struct sk_buff *skb,
-			    struct nf_conn *ct,
-			    enum ip_conntrack_info ctinfo,
-			    union nf_inet_addr *addr,
-			    __be16 port,
-			    const char **dptr, unsigned int *datalen)
+			    const char **dptr, unsigned int *datalen,
+			    union nf_inet_addr *addr, __be16 port)
 {
 	struct nf_conntrack_expect *exp;
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
 	int family = ct->tuplehash[!dir].tuple.src.l3num;
 	int ret;
@@ -388,7 +384,7 @@ static int set_expected_rtp(struct sk_buff *skb,
 
 	nf_nat_sdp = rcu_dereference(nf_nat_sdp_hook);
 	if (nf_nat_sdp && ct->status & IPS_NAT_MASK)
-		ret = nf_nat_sdp(skb, ctinfo, exp, dptr, datalen);
+		ret = nf_nat_sdp(skb, dptr, datalen, exp);
 	else {
 		if (nf_ct_expect_related(exp) != 0)
 			ret = NF_DROP;
@@ -431,7 +427,7 @@ static int sip_help(struct sk_buff *skb,
 
 	nf_nat_sip = rcu_dereference(nf_nat_sip_hook);
 	if (nf_nat_sip && ct->status & IPS_NAT_MASK) {
-		if (!nf_nat_sip(skb, ctinfo, ct, &dptr, &datalen)) {
+		if (!nf_nat_sip(skb, &dptr, &datalen)) {
 			ret = NF_DROP;
 			goto out;
 		}
@@ -467,8 +463,8 @@ static int sip_help(struct sk_buff *skb,
 				ret = NF_DROP;
 				goto out;
 			}
-			ret = set_expected_rtp(skb, ct, ctinfo, &addr,
-					       htons(port), &dptr, &datalen);
+			ret = set_expected_rtp(skb, &dptr, &datalen,
+					       &addr, htons(port));
 		}
 	}
 out:
-- 
cgit v1.2.3


From 3e9b4600b4e71beaa9d943251bfe9c25f6a97b8c Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:17:55 -0700
Subject: [NETFILTER]: nf_conntrack_sip: add seperate SDP header parsing
 function

SDP and SIP headers are quite different, SIP can have continuation lines,
leading and trailing whitespace after the colon and is mostly case-insensitive
while SDP headers always begin on a new line and are followed by an equal
sign and the value, without any whitespace.

Introduce new SDP header parsing function and convert all users that used
the SIP header parsing function. This will allow to properly deal with the
special SIP cases in the SIP header parsing function later.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  49 +++++++--
 net/ipv4/netfilter/nf_nat_sip.c            |  69 ++++++-------
 net/netfilter/nf_conntrack_sip.c           | 159 +++++++++++++++++------------
 3 files changed, 169 insertions(+), 108 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index b94de3d60303..9131cbc9b9de 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -13,12 +13,42 @@ enum sip_header_pos {
 	POS_VIA,
 	POS_CONTACT,
 	POS_CONTENT,
-	POS_MEDIA,
-	POS_OWNER_IP4,
-	POS_CONNECTION_IP4,
-	POS_OWNER_IP6,
-	POS_CONNECTION_IP6,
-	POS_SDP_HEADER,
+};
+
+struct sip_header {
+	const char	*name;
+	const char	*cname;
+	const char	*search;
+	unsigned int	len;
+	unsigned int	clen;
+	unsigned int	slen;
+	int		(*match_len)(const struct nf_conn *ct,
+				     const char *dptr, const char *limit,
+				     int *shift);
+};
+
+#define __SIP_HDR(__name, __cname, __search, __match)			\
+{									\
+	.name		= (__name),					\
+	.len		= sizeof(__name) - 1,				\
+	.cname		= (__cname),					\
+	.clen		= (__cname) ? sizeof(__cname) - 1 : 0,		\
+	.search		= (__search),					\
+	.slen		= (__search) ? sizeof(__search) - 1 : 0,	\
+	.match_len	= (__match),					\
+}
+
+#define SDP_HDR(__name, __search, __match) \
+	__SIP_HDR(__name, NULL, __search, __match)
+
+enum sdp_header_types {
+	SDP_HDR_UNSPEC,
+	SDP_HDR_VERSION,
+	SDP_HDR_OWNER_IP4,
+	SDP_HDR_CONNECTION_IP4,
+	SDP_HDR_OWNER_IP6,
+	SDP_HDR_CONNECTION_IP6,
+	SDP_HDR_MEDIA,
 };
 
 extern unsigned int (*nf_nat_sip_hook)(struct sk_buff *skb,
@@ -36,5 +66,12 @@ extern int ct_sip_lnlen(const char *line, const char *limit);
 extern const char *ct_sip_search(const char *needle, const char *haystack,
 				 size_t needle_len, size_t haystack_len,
 				 int case_sensitive);
+
+extern int ct_sip_get_sdp_header(const struct nf_conn *ct, const char *dptr,
+				 unsigned int dataoff, unsigned int datalen,
+				 enum sdp_header_types type,
+				 enum sdp_header_types term,
+				 unsigned int *matchoff, unsigned int *matchlen);
+
 #endif /* __KERNEL__ */
 #endif /* __NF_CONNTRACK_SIP_H__ */
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index dd1b2d86deee..aa8a4f492baf 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -147,51 +147,46 @@ static unsigned int ip_nat_sip(struct sk_buff *skb,
 	return NF_ACCEPT;
 }
 
-static unsigned int mangle_sip_packet(struct sk_buff *skb,
-				      const char **dptr, unsigned int *datalen,
-				      char *buffer, int bufflen,
-				      enum sip_header_pos pos)
+static int mangle_content_len(struct sk_buff *skb,
+			      const char **dptr, unsigned int *datalen)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
-	unsigned int matchlen, matchoff;
+	unsigned int matchoff, matchlen;
+	char buffer[sizeof("65536")];
+	int buflen, c_len;
 
+	/* Get actual SDP length */
+	if (ct_sip_get_sdp_header(ct, *dptr, 0, *datalen,
+				  SDP_HDR_VERSION, SDP_HDR_UNSPEC,
+				  &matchoff, &matchlen) <= 0)
+		return 0;
+	c_len = *datalen - matchoff + strlen("v=");
+
+	/* Now, update SDP length */
 	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
-			    pos) <= 0)
+			    POS_CONTENT) <= 0)
 		return 0;
 
+	buflen = sprintf(buffer, "%u", c_len);
 	return mangle_packet(skb, dptr, datalen, matchoff, matchlen,
-			     buffer, bufflen);
+			     buffer, buflen);
 }
 
-static int mangle_content_len(struct sk_buff *skb,
-			      const char **dptr, unsigned int *datalen)
+static unsigned mangle_sdp_packet(struct sk_buff *skb,
+				  const char **dptr, unsigned int *datalen,
+				  enum sdp_header_types type,
+				  char *buffer, int buflen)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
-	unsigned int matchoff, matchlen;
-	char buffer[sizeof("65536")];
-	int bufflen;
+	unsigned int matchlen, matchoff;
 
-	/* Get actual SDP length */
-	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff,
-			    &matchlen, POS_SDP_HEADER) > 0) {
-
-		/* since ct_sip_get_info() give us a pointer passing 'v='
-		   we need to add 2 bytes in this count. */
-		int c_len = *datalen - matchoff + 2;
-
-		/* Now, update SDP length */
-		if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff,
-				    &matchlen, POS_CONTENT) > 0) {
-
-			bufflen = sprintf(buffer, "%u", c_len);
-			return mangle_packet(skb, dptr, datalen,
-					     matchoff, matchlen,
-					     buffer, bufflen);
-		}
-	}
-	return 0;
+	if (ct_sip_get_sdp_header(ct, *dptr, 0, *datalen, type, SDP_HDR_UNSPEC,
+				  &matchoff, &matchlen) <= 0)
+		return 0;
+	return mangle_packet(skb, dptr, datalen, matchoff, matchlen,
+			     buffer, buflen);
 }
 
 static unsigned int mangle_sdp(struct sk_buff *skb,
@@ -205,18 +200,18 @@ static unsigned int mangle_sdp(struct sk_buff *skb,
 
 	/* Mangle owner and contact info. */
 	bufflen = sprintf(buffer, "%u.%u.%u.%u", NIPQUAD(newip));
-	if (!mangle_sip_packet(skb, dptr, datalen, buffer, bufflen,
-			       POS_OWNER_IP4))
+	if (!mangle_sdp_packet(skb, dptr, datalen, SDP_HDR_OWNER_IP4,
+			       buffer, bufflen))
 		return 0;
 
-	if (!mangle_sip_packet(skb, dptr, datalen, buffer, bufflen,
-			       POS_CONNECTION_IP4))
+	if (!mangle_sdp_packet(skb, dptr, datalen, SDP_HDR_CONNECTION_IP4,
+			       buffer, bufflen))
 		return 0;
 
 	/* Mangle media port. */
 	bufflen = sprintf(buffer, "%u", port);
-	if (!mangle_sip_packet(skb, dptr, datalen, buffer, bufflen,
-			       POS_MEDIA))
+	if (!mangle_sdp_packet(skb, dptr, datalen, SDP_HDR_MEDIA,
+			       buffer, bufflen))
 		return 0;
 
 	return mangle_content_len(skb, dptr, datalen);
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index cf19a7082a75..801fcb3c749f 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -124,66 +124,6 @@ static const struct sip_header_nfo ct_sip_hdrs[] = {
 		.ln_strlen	= sizeof(":") - 1,
 		.match_len	= skp_digits_len
 	},
-	[POS_MEDIA] = {		/* SDP media info */
-		.case_sensitive	= 1,
-		.lname		= "\nm=",
-		.lnlen		= sizeof("\nm=") - 1,
-		.sname		= "\rm=",
-		.snlen		= sizeof("\rm=") - 1,
-		.ln_str		= "audio ",
-		.ln_strlen	= sizeof("audio ") - 1,
-		.match_len	= digits_len
-	},
-	[POS_OWNER_IP4] = {	/* SDP owner address*/
-		.case_sensitive	= 1,
-		.lname		= "\no=",
-		.lnlen		= sizeof("\no=") - 1,
-		.sname		= "\ro=",
-		.snlen		= sizeof("\ro=") - 1,
-		.ln_str		= "IN IP4 ",
-		.ln_strlen	= sizeof("IN IP4 ") - 1,
-		.match_len	= epaddr_len
-	},
-	[POS_CONNECTION_IP4] = {/* SDP connection info */
-		.case_sensitive	= 1,
-		.lname		= "\nc=",
-		.lnlen		= sizeof("\nc=") - 1,
-		.sname		= "\rc=",
-		.snlen		= sizeof("\rc=") - 1,
-		.ln_str		= "IN IP4 ",
-		.ln_strlen	= sizeof("IN IP4 ") - 1,
-		.match_len	= epaddr_len
-	},
-	[POS_OWNER_IP6] = {	/* SDP owner address*/
-		.case_sensitive	= 1,
-		.lname		= "\no=",
-		.lnlen		= sizeof("\no=") - 1,
-		.sname		= "\ro=",
-		.snlen		= sizeof("\ro=") - 1,
-		.ln_str		= "IN IP6 ",
-		.ln_strlen	= sizeof("IN IP6 ") - 1,
-		.match_len	= epaddr_len
-	},
-	[POS_CONNECTION_IP6] = {/* SDP connection info */
-		.case_sensitive	= 1,
-		.lname		= "\nc=",
-		.lnlen		= sizeof("\nc=") - 1,
-		.sname		= "\rc=",
-		.snlen		= sizeof("\rc=") - 1,
-		.ln_str		= "IN IP6 ",
-		.ln_strlen	= sizeof("IN IP6 ") - 1,
-		.match_len	= epaddr_len
-	},
-	[POS_SDP_HEADER] = { 	/* SDP version header */
-		.case_sensitive	= 1,
-		.lname		= "\nv=",
-		.lnlen		= sizeof("\nv=") - 1,
-		.sname		= "\rv=",
-		.snlen		= sizeof("\rv=") - 1,
-		.ln_str		= "=",
-		.ln_strlen	= sizeof("=") - 1,
-		.match_len	= digits_len
-	}
 };
 
 /* get line length until first CR or LF seen. */
@@ -363,6 +303,92 @@ int ct_sip_get_info(const struct nf_conn *ct,
 }
 EXPORT_SYMBOL_GPL(ct_sip_get_info);
 
+/* SDP header parsing: a SDP session description contains an ordered set of
+ * headers, starting with a section containing general session parameters,
+ * optionally followed by multiple media descriptions.
+ *
+ * SDP headers always start at the beginning of a line. According to RFC 2327:
+ * "The sequence CRLF (0x0d0a) is used to end a record, although parsers should
+ * be tolerant and also accept records terminated with a single newline
+ * character". We handle both cases.
+ */
+static const struct sip_header ct_sdp_hdrs[] = {
+	[SDP_HDR_VERSION]		= SDP_HDR("v=", NULL, digits_len),
+	[SDP_HDR_OWNER_IP4]		= SDP_HDR("o=", "IN IP4 ", epaddr_len),
+	[SDP_HDR_CONNECTION_IP4]	= SDP_HDR("c=", "IN IP4 ", epaddr_len),
+	[SDP_HDR_OWNER_IP6]		= SDP_HDR("o=", "IN IP6 ", epaddr_len),
+	[SDP_HDR_CONNECTION_IP6]	= SDP_HDR("c=", "IN IP6 ", epaddr_len),
+	[SDP_HDR_MEDIA]			= SDP_HDR("m=", "audio ", digits_len),
+};
+
+/* Linear string search within SDP header values */
+static const char *ct_sdp_header_search(const char *dptr, const char *limit,
+					const char *needle, unsigned int len)
+{
+	for (limit -= len; dptr < limit; dptr++) {
+		if (*dptr == '\r' || *dptr == '\n')
+			break;
+		if (strncmp(dptr, needle, len) == 0)
+			return dptr;
+	}
+	return NULL;
+}
+
+/* Locate a SDP header (optionally a substring within the header value),
+ * optionally stopping at the first occurence of the term header, parse
+ * it and return the offset and length of the data we're interested in.
+ */
+int ct_sip_get_sdp_header(const struct nf_conn *ct, const char *dptr,
+			  unsigned int dataoff, unsigned int datalen,
+			  enum sdp_header_types type,
+			  enum sdp_header_types term,
+			  unsigned int *matchoff, unsigned int *matchlen)
+{
+	const struct sip_header *hdr = &ct_sdp_hdrs[type];
+	const struct sip_header *thdr = &ct_sdp_hdrs[term];
+	const char *start = dptr, *limit = dptr + datalen;
+	int shift = 0;
+
+	for (dptr += dataoff; dptr < limit; dptr++) {
+		/* Find beginning of line */
+		if (*dptr != '\r' && *dptr != '\n')
+			continue;
+		if (++dptr >= limit)
+			break;
+		if (*(dptr - 1) == '\r' && *dptr == '\n') {
+			if (++dptr >= limit)
+				break;
+		}
+
+		if (term != SDP_HDR_UNSPEC &&
+		    limit - dptr >= thdr->len &&
+		    strnicmp(dptr, thdr->name, thdr->len) == 0)
+			break;
+		else if (limit - dptr >= hdr->len &&
+			 strnicmp(dptr, hdr->name, hdr->len) == 0)
+			dptr += hdr->len;
+		else
+			continue;
+
+		*matchoff = dptr - start;
+		if (hdr->search) {
+			dptr = ct_sdp_header_search(dptr, limit, hdr->search,
+						    hdr->slen);
+			if (!dptr)
+				return -1;
+			dptr += hdr->slen;
+		}
+
+		*matchlen = hdr->match_len(ct, dptr, limit, &shift);
+		if (!*matchlen)
+			return -1;
+		*matchoff = dptr - start + shift;
+		return 1;
+	}
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ct_sip_get_sdp_header);
+
 static int set_expected_rtp(struct sk_buff *skb,
 			    const char **dptr, unsigned int *datalen,
 			    union nf_inet_addr *addr, __be16 port)
@@ -408,7 +434,7 @@ static int sip_help(struct sk_buff *skb,
 	int ret = NF_ACCEPT;
 	unsigned int matchoff, matchlen;
 	u_int16_t port;
-	enum sip_header_pos pos;
+	enum sdp_header_types type;
 	typeof(nf_nat_sip_hook) nf_nat_sip;
 
 	/* No Data ? */
@@ -446,8 +472,10 @@ static int sip_help(struct sk_buff *skb,
 		goto out;
 	}
 	/* Get address and port from SDP packet. */
-	pos = family == AF_INET ? POS_CONNECTION_IP4 : POS_CONNECTION_IP6;
-	if (ct_sip_get_info(ct, dptr, datalen, &matchoff, &matchlen, pos) > 0) {
+	type = family == AF_INET ? SDP_HDR_CONNECTION_IP4 :
+				   SDP_HDR_CONNECTION_IP6;
+	if (ct_sip_get_sdp_header(ct, dptr, 0, datalen, type, SDP_HDR_UNSPEC,
+				  &matchoff, &matchlen) > 0) {
 
 		/* We'll drop only if there are parse problems. */
 		if (!parse_addr(ct, dptr + matchoff, NULL, &addr,
@@ -455,8 +483,9 @@ static int sip_help(struct sk_buff *skb,
 			ret = NF_DROP;
 			goto out;
 		}
-		if (ct_sip_get_info(ct, dptr, datalen, &matchoff, &matchlen,
-				    POS_MEDIA) > 0) {
+		if (ct_sip_get_sdp_header(ct, dptr, 0, datalen,
+					  SDP_HDR_MEDIA, SDP_HDR_UNSPEC,
+					  &matchoff, &matchlen) > 0) {
 
 			port = simple_strtoul(dptr + matchoff, NULL, 10);
 			if (port < 1024) {
-- 
cgit v1.2.3


From ac3677406d4e36e86b1eb5a453997a3b3e0c089a Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:18:40 -0700
Subject: [NETFILTER]: nf_conntrack_sip: kill request URI "header" definitions

The request URI is not a header and needs to be treated differently than
real SIP headers. Add a seperate function for parsing it and get rid of
the POS_REQ_URI/POS_REG_REQ_URI definitions.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  5 ++-
 net/ipv4/netfilter/nf_nat_sip.c            | 46 +++++++++++----------
 net/netfilter/nf_conntrack_sip.c           | 64 +++++++++++++++++++++++-------
 3 files changed, 77 insertions(+), 38 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 9131cbc9b9de..480b26f40ce4 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -6,8 +6,6 @@
 #define SIP_TIMEOUT	3600
 
 enum sip_header_pos {
-	POS_REG_REQ_URI,
-	POS_REQ_URI,
 	POS_FROM,
 	POS_TO,
 	POS_VIA,
@@ -59,6 +57,9 @@ extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				       unsigned int *datalen,
 				       struct nf_conntrack_expect *exp);
 
+extern int ct_sip_parse_request(const struct nf_conn *ct,
+				const char *dptr, unsigned int datalen,
+				unsigned int *matchoff, unsigned int *matchlen);
 extern int ct_sip_get_info(const struct nf_conn *ct, const char *dptr,
                            size_t dlen, unsigned int *matchoff,
                            unsigned int *matchlen, enum sip_header_pos pos);
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index aa8a4f492baf..60151b5901a5 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -78,20 +78,17 @@ static unsigned int mangle_packet(struct sk_buff *skb,
 	return 1;
 }
 
-static int map_sip_addr(struct sk_buff *skb,
-			const char **dptr, unsigned int *datalen,
-			enum sip_header_pos pos, struct addr_map *map)
+static int map_addr(struct sk_buff *skb,
+		    const char **dptr, unsigned int *datalen,
+		    unsigned int matchoff, unsigned int matchlen,
+		    struct addr_map *map)
 {
 	enum ip_conntrack_info ctinfo;
-	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+	struct nf_conn *ct __maybe_unused = nf_ct_get(skb, &ctinfo);
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
-	unsigned int matchlen, matchoff, addrlen;
+	unsigned int addrlen;
 	char *addr;
 
-	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
-			    pos) <= 0)
-		return 1;
-
 	if ((matchlen == map->addr[dir].srciplen ||
 	     matchlen == map->addr[dir].srclen) &&
 	    strncmp(*dptr + matchoff, map->addr[dir].src, matchlen) == 0) {
@@ -109,13 +106,27 @@ static int map_sip_addr(struct sk_buff *skb,
 			     addr, addrlen);
 }
 
+static int map_sip_addr(struct sk_buff *skb,
+			const char **dptr, unsigned int *datalen,
+			enum sip_header_pos pos, struct addr_map *map)
+{
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+	unsigned int matchlen, matchoff;
+
+	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
+			    pos) <= 0)
+		return 1;
+	return map_addr(skb, dptr, datalen, matchoff, matchlen, map);
+}
+
 static unsigned int ip_nat_sip(struct sk_buff *skb,
 			       const char **dptr, unsigned int *datalen)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
-	enum sip_header_pos pos;
 	struct addr_map map;
+	unsigned int matchoff, matchlen;
 
 	if (*datalen < strlen("SIP/2.0"))
 		return NF_ACCEPT;
@@ -124,18 +135,9 @@ static unsigned int ip_nat_sip(struct sk_buff *skb,
 
 	/* Basic rules: requests and responses. */
 	if (strnicmp(*dptr, "SIP/2.0", strlen("SIP/2.0")) != 0) {
-		/* 10.2: Constructing the REGISTER Request:
-		 *
-		 * The "userinfo" and "@" components of the SIP URI MUST NOT
-		 * be present.
-		 */
-		if (*datalen >= strlen("REGISTER") &&
-		    strnicmp(*dptr, "REGISTER", strlen("REGISTER")) == 0)
-			pos = POS_REG_REQ_URI;
-		else
-			pos = POS_REQ_URI;
-
-		if (!map_sip_addr(skb, dptr, datalen, pos, &map))
+		if (ct_sip_parse_request(ct, *dptr, *datalen,
+					 &matchoff, &matchlen) > 0 &&
+		    !map_addr(skb, dptr, datalen, matchoff, matchlen, &map))
 			return NF_DROP;
 	}
 
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index 801fcb3c749f..bb4396155681 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -65,20 +65,6 @@ struct sip_header_nfo {
 };
 
 static const struct sip_header_nfo ct_sip_hdrs[] = {
-	[POS_REG_REQ_URI] = { 	/* SIP REGISTER request URI */
-		.lname		= "sip:",
-		.lnlen		= sizeof("sip:") - 1,
-		.ln_str		= ":",
-		.ln_strlen	= sizeof(":") - 1,
-		.match_len	= epaddr_len,
-	},
-	[POS_REQ_URI] = { 	/* SIP request URI */
-		.lname		= "sip:",
-		.lnlen		= sizeof("sip:") - 1,
-		.ln_str		= "@",
-		.ln_strlen	= sizeof("@") - 1,
-		.match_len	= epaddr_len,
-	},
 	[POS_FROM] = {		/* SIP From header */
 		.lname		= "From:",
 		.lnlen		= sizeof("From:") - 1,
@@ -164,6 +150,18 @@ const char *ct_sip_search(const char *needle, const char *haystack,
 }
 EXPORT_SYMBOL_GPL(ct_sip_search);
 
+static int string_len(const struct nf_conn *ct, const char *dptr,
+		      const char *limit, int *shift)
+{
+	int len = 0;
+
+	while (dptr < limit && isalpha(*dptr)) {
+		dptr++;
+		len++;
+	}
+	return len;
+}
+
 static int digits_len(const struct nf_conn *ct, const char *dptr,
 		      const char *limit, int *shift)
 {
@@ -258,6 +256,44 @@ static int skp_epaddr_len(const struct nf_conn *ct, const char *dptr,
 	return epaddr_len(ct, dptr, limit, shift);
 }
 
+/* Parse a SIP request line of the form:
+ *
+ * Request-Line = Method SP Request-URI SP SIP-Version CRLF
+ *
+ * and return the offset and length of the address contained in the Request-URI.
+ */
+int ct_sip_parse_request(const struct nf_conn *ct,
+			 const char *dptr, unsigned int datalen,
+			 unsigned int *matchoff, unsigned int *matchlen)
+{
+	const char *start = dptr, *limit = dptr + datalen;
+	unsigned int mlen;
+	int shift = 0;
+
+	/* Skip method and following whitespace */
+	mlen = string_len(ct, dptr, limit, NULL);
+	if (!mlen)
+		return 0;
+	dptr += mlen;
+	if (++dptr >= limit)
+		return 0;
+
+	/* Find SIP URI */
+	limit -= strlen("sip:");
+	for (; dptr < limit; dptr++) {
+		if (*dptr == '\r' || *dptr == '\n')
+			return -1;
+		if (strnicmp(dptr, "sip:", strlen("sip:")) == 0)
+			break;
+	}
+	*matchlen = skp_epaddr_len(ct, dptr, limit, &shift);
+	if (!*matchlen)
+		return 0;
+	*matchoff = dptr - start + shift;
+	return 1;
+}
+EXPORT_SYMBOL_GPL(ct_sip_parse_request);
+
 /* Returns 0 if not found, -1 error parsing. */
 int ct_sip_get_info(const struct nf_conn *ct,
 		    const char *dptr, size_t dlen,
-- 
cgit v1.2.3


From ea45f12a2766dae54e5426a23e8f4bafdbe2782e Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:18:57 -0700
Subject: [NETFILTER]: nf_conntrack_sip: parse SIP headers properly

Introduce new function for SIP header parsing that properly deals with
continuation lines and whitespace in headers and use it.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  30 ++--
 net/ipv4/netfilter/nf_nat_sip.c            |  18 +-
 net/netfilter/nf_conntrack_sip.c           | 271 ++++++++++++++---------------
 3 files changed, 151 insertions(+), 168 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 480b26f40ce4..ccc701422963 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -5,14 +5,6 @@
 #define SIP_PORT	5060
 #define SIP_TIMEOUT	3600
 
-enum sip_header_pos {
-	POS_FROM,
-	POS_TO,
-	POS_VIA,
-	POS_CONTACT,
-	POS_CONTENT,
-};
-
 struct sip_header {
 	const char	*name;
 	const char	*cname;
@@ -36,9 +28,20 @@ struct sip_header {
 	.match_len	= (__match),					\
 }
 
+#define SIP_HDR(__name, __cname, __search, __match) \
+	__SIP_HDR(__name, __cname, __search, __match)
+
 #define SDP_HDR(__name, __search, __match) \
 	__SIP_HDR(__name, NULL, __search, __match)
 
+enum sip_header_types {
+	SIP_HDR_FROM,
+	SIP_HDR_TO,
+	SIP_HDR_CONTACT,
+	SIP_HDR_VIA,
+	SIP_HDR_CONTENT_LENGTH,
+};
+
 enum sdp_header_types {
 	SDP_HDR_UNSPEC,
 	SDP_HDR_VERSION,
@@ -60,13 +63,10 @@ extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 extern int ct_sip_parse_request(const struct nf_conn *ct,
 				const char *dptr, unsigned int datalen,
 				unsigned int *matchoff, unsigned int *matchlen);
-extern int ct_sip_get_info(const struct nf_conn *ct, const char *dptr,
-                           size_t dlen, unsigned int *matchoff,
-                           unsigned int *matchlen, enum sip_header_pos pos);
-extern int ct_sip_lnlen(const char *line, const char *limit);
-extern const char *ct_sip_search(const char *needle, const char *haystack,
-				 size_t needle_len, size_t haystack_len,
-				 int case_sensitive);
+extern int ct_sip_get_header(const struct nf_conn *ct, const char *dptr,
+			     unsigned int dataoff, unsigned int datalen,
+			     enum sip_header_types type,
+			     unsigned int *matchoff, unsigned int *matchlen);
 
 extern int ct_sip_get_sdp_header(const struct nf_conn *ct, const char *dptr,
 				 unsigned int dataoff, unsigned int datalen,
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index 60151b5901a5..c13e43862361 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -108,14 +108,14 @@ static int map_addr(struct sk_buff *skb,
 
 static int map_sip_addr(struct sk_buff *skb,
 			const char **dptr, unsigned int *datalen,
-			enum sip_header_pos pos, struct addr_map *map)
+			enum sip_header_types type, struct addr_map *map)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	unsigned int matchlen, matchoff;
 
-	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
-			    pos) <= 0)
+	if (ct_sip_get_header(ct, *dptr, 0, *datalen, type,
+			      &matchoff, &matchlen) <= 0)
 		return 1;
 	return map_addr(skb, dptr, datalen, matchoff, matchlen, map);
 }
@@ -141,10 +141,10 @@ static unsigned int ip_nat_sip(struct sk_buff *skb,
 			return NF_DROP;
 	}
 
-	if (!map_sip_addr(skb, dptr, datalen, POS_FROM, &map) ||
-	    !map_sip_addr(skb, dptr, datalen, POS_TO, &map) ||
-	    !map_sip_addr(skb, dptr, datalen, POS_VIA, &map) ||
-	    !map_sip_addr(skb, dptr, datalen, POS_CONTACT, &map))
+	if (!map_sip_addr(skb, dptr, datalen, SIP_HDR_FROM, &map) ||
+	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_TO, &map) ||
+	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_VIA, &map) ||
+	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_CONTACT, &map))
 		return NF_DROP;
 	return NF_ACCEPT;
 }
@@ -166,8 +166,8 @@ static int mangle_content_len(struct sk_buff *skb,
 	c_len = *datalen - matchoff + strlen("v=");
 
 	/* Now, update SDP length */
-	if (ct_sip_get_info(ct, *dptr, *datalen, &matchoff, &matchlen,
-			    POS_CONTENT) <= 0)
+	if (ct_sip_get_header(ct, *dptr, 0, *datalen, SIP_HDR_CONTENT_LENGTH,
+			      &matchoff, &matchlen) <= 0)
 		return 0;
 
 	buflen = sprintf(buffer, "%u", c_len);
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index bb4396155681..cbc91598acee 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -47,109 +47,6 @@ unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				struct nf_conntrack_expect *exp) __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sdp_hook);
 
-static int digits_len(const struct nf_conn *, const char *, const char *, int *);
-static int epaddr_len(const struct nf_conn *, const char *, const char *, int *);
-static int skp_digits_len(const struct nf_conn *, const char *, const char *, int *);
-static int skp_epaddr_len(const struct nf_conn *, const char *, const char *, int *);
-
-struct sip_header_nfo {
-	const char	*lname;
-	const char	*sname;
-	const char	*ln_str;
-	size_t		lnlen;
-	size_t		snlen;
-	size_t		ln_strlen;
-	int		case_sensitive;
-	int		(*match_len)(const struct nf_conn *, const char *,
-				     const char *, int *);
-};
-
-static const struct sip_header_nfo ct_sip_hdrs[] = {
-	[POS_FROM] = {		/* SIP From header */
-		.lname		= "From:",
-		.lnlen		= sizeof("From:") - 1,
-		.sname		= "\r\nf:",
-		.snlen		= sizeof("\r\nf:") - 1,
-		.ln_str		= "sip:",
-		.ln_strlen	= sizeof("sip:") - 1,
-		.match_len	= skp_epaddr_len,
-	},
-	[POS_TO] = {		/* SIP To header */
-		.lname		= "To:",
-		.lnlen		= sizeof("To:") - 1,
-		.sname		= "\r\nt:",
-		.snlen		= sizeof("\r\nt:") - 1,
-		.ln_str		= "sip:",
-		.ln_strlen	= sizeof("sip:") - 1,
-		.match_len	= skp_epaddr_len
-	},
-	[POS_VIA] = { 		/* SIP Via header */
-		.lname		= "Via:",
-		.lnlen		= sizeof("Via:") - 1,
-		.sname		= "\r\nv:",
-		.snlen		= sizeof("\r\nv:") - 1, /* rfc3261 "\r\n" */
-		.ln_str		= "UDP ",
-		.ln_strlen	= sizeof("UDP ") - 1,
-		.match_len	= epaddr_len,
-	},
-	[POS_CONTACT] = { 	/* SIP Contact header */
-		.lname		= "Contact:",
-		.lnlen		= sizeof("Contact:") - 1,
-		.sname		= "\r\nm:",
-		.snlen		= sizeof("\r\nm:") - 1,
-		.ln_str		= "sip:",
-		.ln_strlen	= sizeof("sip:") - 1,
-		.match_len	= skp_epaddr_len
-	},
-	[POS_CONTENT] = { 	/* SIP Content length header */
-		.lname		= "Content-Length:",
-		.lnlen		= sizeof("Content-Length:") - 1,
-		.sname		= "\r\nl:",
-		.snlen		= sizeof("\r\nl:") - 1,
-		.ln_str		= ":",
-		.ln_strlen	= sizeof(":") - 1,
-		.match_len	= skp_digits_len
-	},
-};
-
-/* get line length until first CR or LF seen. */
-int ct_sip_lnlen(const char *line, const char *limit)
-{
-	const char *k = line;
-
-	while ((line < limit) && (*line == '\r' || *line == '\n'))
-		line++;
-
-	while (line < limit) {
-		if (*line == '\r' || *line == '\n')
-			break;
-		line++;
-	}
-	return line - k;
-}
-EXPORT_SYMBOL_GPL(ct_sip_lnlen);
-
-/* Linear string search, case sensitive. */
-const char *ct_sip_search(const char *needle, const char *haystack,
-			  size_t needle_len, size_t haystack_len,
-			  int case_sensitive)
-{
-	const char *limit = haystack + (haystack_len - needle_len);
-
-	while (haystack < limit) {
-		if (case_sensitive) {
-			if (strncmp(haystack, needle, needle_len) == 0)
-				return haystack;
-		} else {
-			if (strnicmp(haystack, needle, needle_len) == 0)
-				return haystack;
-		}
-		haystack++;
-	}
-	return NULL;
-}
-EXPORT_SYMBOL_GPL(ct_sip_search);
-
 static int string_len(const struct nf_conn *ct, const char *dptr,
 		      const char *limit, int *shift)
 {
@@ -173,16 +70,6 @@ static int digits_len(const struct nf_conn *ct, const char *dptr,
 	return len;
 }
 
-/* get digits length, skipping blank spaces. */
-static int skp_digits_len(const struct nf_conn *ct, const char *dptr,
-			  const char *limit, int *shift)
-{
-	for (; dptr < limit && *dptr == ' '; dptr++)
-		(*shift)++;
-
-	return digits_len(ct, dptr, limit, shift);
-}
-
 static int parse_addr(const struct nf_conn *ct, const char *cp,
                       const char **endp, union nf_inet_addr *addr,
                       const char *limit)
@@ -294,50 +181,146 @@ int ct_sip_parse_request(const struct nf_conn *ct,
 }
 EXPORT_SYMBOL_GPL(ct_sip_parse_request);
 
-/* Returns 0 if not found, -1 error parsing. */
-int ct_sip_get_info(const struct nf_conn *ct,
-		    const char *dptr, size_t dlen,
-		    unsigned int *matchoff,
-		    unsigned int *matchlen,
-		    enum sip_header_pos pos)
+/* SIP header parsing: SIP headers are located at the beginning of a line, but
+ * may span several lines, in which case the continuation lines begin with a
+ * whitespace character. RFC 2543 allows lines to be terminated with CR, LF or
+ * CRLF, RFC 3261 allows only CRLF, we support both.
+ *
+ * Headers are followed by (optionally) whitespace, a colon, again (optionally)
+ * whitespace and the values. Whitespace in this context means any amount of
+ * tabs, spaces and continuation lines, which are treated as a single whitespace
+ * character.
+ */
+static const struct sip_header ct_sip_hdrs[] = {
+	[SIP_HDR_FROM]			= SIP_HDR("From", "f", "sip:", skp_epaddr_len),
+	[SIP_HDR_TO]			= SIP_HDR("To", "t", "sip:", skp_epaddr_len),
+	[SIP_HDR_CONTACT]		= SIP_HDR("Contact", "m", "sip:", skp_epaddr_len),
+	[SIP_HDR_VIA]			= SIP_HDR("Via", "v", "UDP ", epaddr_len),
+	[SIP_HDR_CONTENT_LENGTH]	= SIP_HDR("Content-Length", "l", NULL, digits_len),
+};
+
+static const char *sip_follow_continuation(const char *dptr, const char *limit)
 {
-	const struct sip_header_nfo *hnfo = &ct_sip_hdrs[pos];
-	const char *limit, *aux, *k = dptr;
-	int shift = 0;
+	/* Walk past newline */
+	if (++dptr >= limit)
+		return NULL;
+
+	/* Skip '\n' in CR LF */
+	if (*(dptr - 1) == '\r' && *dptr == '\n') {
+		if (++dptr >= limit)
+			return NULL;
+	}
+
+	/* Continuation line? */
+	if (*dptr != ' ' && *dptr != '\t')
+		return NULL;
+
+	/* skip leading whitespace */
+	for (; dptr < limit; dptr++) {
+		if (*dptr != ' ' && *dptr != '\t')
+			break;
+	}
+	return dptr;
+}
+
+static const char *sip_skip_whitespace(const char *dptr, const char *limit)
+{
+	for (; dptr < limit; dptr++) {
+		if (*dptr == ' ')
+			continue;
+		if (*dptr != '\r' && *dptr != '\n')
+			break;
+		dptr = sip_follow_continuation(dptr, limit);
+		if (dptr == NULL)
+			return NULL;
+	}
+	return dptr;
+}
 
-	limit = dptr + (dlen - hnfo->lnlen);
+/* Search within a SIP header value, dealing with continuation lines */
+static const char *ct_sip_header_search(const char *dptr, const char *limit,
+					const char *needle, unsigned int len)
+{
+	for (limit -= len; dptr < limit; dptr++) {
+		if (*dptr == '\r' || *dptr == '\n') {
+			dptr = sip_follow_continuation(dptr, limit);
+			if (dptr == NULL)
+				break;
+			continue;
+		}
 
-	while (dptr < limit) {
-		if ((strncmp(dptr, hnfo->lname, hnfo->lnlen) != 0) &&
-		    (hnfo->sname == NULL ||
-		     strncmp(dptr, hnfo->sname, hnfo->snlen) != 0)) {
-			dptr++;
+		if (strnicmp(dptr, needle, len) == 0)
+			return dptr;
+	}
+	return NULL;
+}
+
+int ct_sip_get_header(const struct nf_conn *ct, const char *dptr,
+		      unsigned int dataoff, unsigned int datalen,
+		      enum sip_header_types type,
+		      unsigned int *matchoff, unsigned int *matchlen)
+{
+	const struct sip_header *hdr = &ct_sip_hdrs[type];
+	const char *start = dptr, *limit = dptr + datalen;
+	int shift = 0;
+
+	for (dptr += dataoff; dptr < limit; dptr++) {
+		/* Find beginning of line */
+		if (*dptr != '\r' && *dptr != '\n')
 			continue;
+		if (++dptr >= limit)
+			break;
+		if (*(dptr - 1) == '\r' && *dptr == '\n') {
+			if (++dptr >= limit)
+				break;
 		}
-		aux = ct_sip_search(hnfo->ln_str, dptr, hnfo->ln_strlen,
-				    ct_sip_lnlen(dptr, limit),
-				    hnfo->case_sensitive);
-		if (!aux) {
-			pr_debug("'%s' not found in '%s'.\n", hnfo->ln_str,
-				 hnfo->lname);
-			return -1;
+
+		/* Skip continuation lines */
+		if (*dptr == ' ' || *dptr == '\t')
+			continue;
+
+		/* Find header. Compact headers must be followed by a
+		 * non-alphabetic character to avoid mismatches. */
+		if (limit - dptr >= hdr->len &&
+		    strnicmp(dptr, hdr->name, hdr->len) == 0)
+			dptr += hdr->len;
+		else if (hdr->cname && limit - dptr >= hdr->clen + 1 &&
+			 strnicmp(dptr, hdr->cname, hdr->clen) == 0 &&
+			 !isalpha(*(dptr + hdr->clen + 1)))
+			dptr += hdr->clen;
+		else
+			continue;
+
+		/* Find and skip colon */
+		dptr = sip_skip_whitespace(dptr, limit);
+		if (dptr == NULL)
+			break;
+		if (*dptr != ':' || ++dptr >= limit)
+			break;
+
+		/* Skip whitespace after colon */
+		dptr = sip_skip_whitespace(dptr, limit);
+		if (dptr == NULL)
+			break;
+
+		*matchoff = dptr - start;
+		if (hdr->search) {
+			dptr = ct_sip_header_search(dptr, limit, hdr->search,
+						    hdr->slen);
+			if (!dptr)
+				return -1;
+			dptr += hdr->slen;
 		}
-		aux += hnfo->ln_strlen;
 
-		*matchlen = hnfo->match_len(ct, aux, limit, &shift);
+		*matchlen = hdr->match_len(ct, dptr, limit, &shift);
 		if (!*matchlen)
 			return -1;
-
-		*matchoff = (aux - k) + shift;
-
-		pr_debug("%s match succeeded! - len: %u\n", hnfo->lname,
-			 *matchlen);
+		*matchoff = dptr - start + shift;
 		return 1;
 	}
-	pr_debug("%s header not found.\n", hnfo->lname);
 	return 0;
 }
-EXPORT_SYMBOL_GPL(ct_sip_get_info);
+EXPORT_SYMBOL_GPL(ct_sip_get_header);
 
 /* SDP header parsing: a SDP session description contains an ordered set of
  * headers, starting with a section containing general session parameters,
-- 
cgit v1.2.3


From 05e3ced297fe755093140e7487e292fb7603316e Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:19:13 -0700
Subject: [NETFILTER]: nf_conntrack_sip: introduce SIP-URI parsing helper

Introduce a helper function to parse a SIP-URI in a header value, optionally
iterating through all headers of this kind.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |   5 ++
 net/netfilter/nf_conntrack_sip.c           | 107 +++++++++++++++++++++++++++++
 2 files changed, 112 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index ccc701422963..87bc6f79efc4 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -67,6 +67,11 @@ extern int ct_sip_get_header(const struct nf_conn *ct, const char *dptr,
 			     unsigned int dataoff, unsigned int datalen,
 			     enum sip_header_types type,
 			     unsigned int *matchoff, unsigned int *matchlen);
+extern int ct_sip_parse_header_uri(const struct nf_conn *ct, const char *dptr,
+				   unsigned int *dataoff, unsigned int datalen,
+				   enum sip_header_types type, int *in_header,
+				   unsigned int *matchoff, unsigned int *matchlen,
+				   union nf_inet_addr *addr, __be16 *port);
 
 extern int ct_sip_get_sdp_header(const struct nf_conn *ct, const char *dptr,
 				 unsigned int dataoff, unsigned int datalen,
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index cbc91598acee..a74d76a97312 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -190,6 +190,9 @@ EXPORT_SYMBOL_GPL(ct_sip_parse_request);
  * whitespace and the values. Whitespace in this context means any amount of
  * tabs, spaces and continuation lines, which are treated as a single whitespace
  * character.
+ *
+ * Some headers may appear multiple times. A comma seperated list of values is
+ * equivalent to multiple headers.
  */
 static const struct sip_header ct_sip_hdrs[] = {
 	[SIP_HDR_FROM]			= SIP_HDR("From", "f", "sip:", skp_epaddr_len),
@@ -322,6 +325,110 @@ int ct_sip_get_header(const struct nf_conn *ct, const char *dptr,
 }
 EXPORT_SYMBOL_GPL(ct_sip_get_header);
 
+/* Get next header field in a list of comma seperated values */
+static int ct_sip_next_header(const struct nf_conn *ct, const char *dptr,
+			      unsigned int dataoff, unsigned int datalen,
+			      enum sip_header_types type,
+			      unsigned int *matchoff, unsigned int *matchlen)
+{
+	const struct sip_header *hdr = &ct_sip_hdrs[type];
+	const char *start = dptr, *limit = dptr + datalen;
+	int shift = 0;
+
+	dptr += dataoff;
+
+	dptr = ct_sip_header_search(dptr, limit, ",", strlen(","));
+	if (!dptr)
+		return 0;
+
+	dptr = ct_sip_header_search(dptr, limit, hdr->search, hdr->slen);
+	if (!dptr)
+		return 0;
+	dptr += hdr->slen;
+
+	*matchoff = dptr - start;
+	*matchlen = hdr->match_len(ct, dptr, limit, &shift);
+	if (!*matchlen)
+		return -1;
+	*matchoff += shift;
+	return 1;
+}
+
+/* Walk through headers until a parsable one is found or no header of the
+ * given type is left. */
+static int ct_sip_walk_headers(const struct nf_conn *ct, const char *dptr,
+			       unsigned int dataoff, unsigned int datalen,
+			       enum sip_header_types type, int *in_header,
+			       unsigned int *matchoff, unsigned int *matchlen)
+{
+	int ret;
+
+	if (in_header && *in_header) {
+		while (1) {
+			ret = ct_sip_next_header(ct, dptr, dataoff, datalen,
+						 type, matchoff, matchlen);
+			if (ret > 0)
+				return ret;
+			if (ret == 0)
+				break;
+			dataoff += *matchoff;
+		}
+		*in_header = 0;
+	}
+
+	while (1) {
+		ret = ct_sip_get_header(ct, dptr, dataoff, datalen,
+					type, matchoff, matchlen);
+		if (ret > 0)
+			break;
+		if (ret == 0)
+			return ret;
+		dataoff += *matchoff;
+	}
+
+	if (in_header)
+		*in_header = 1;
+	return 1;
+}
+
+/* Locate a SIP header, parse the URI and return the offset and length of
+ * the address as well as the address and port themselves. A stream of
+ * headers can be parsed by handing in a non-NULL datalen and in_header
+ * pointer.
+ */
+int ct_sip_parse_header_uri(const struct nf_conn *ct, const char *dptr,
+			    unsigned int *dataoff, unsigned int datalen,
+			    enum sip_header_types type, int *in_header,
+			    unsigned int *matchoff, unsigned int *matchlen,
+			    union nf_inet_addr *addr, __be16 *port)
+{
+	const char *c, *limit = dptr + datalen;
+	unsigned int p;
+	int ret;
+
+	ret = ct_sip_walk_headers(ct, dptr, dataoff ? *dataoff : 0, datalen,
+				  type, in_header, matchoff, matchlen);
+	WARN_ON(ret < 0);
+	if (ret == 0)
+		return ret;
+
+	if (!parse_addr(ct, dptr + *matchoff, &c, addr, limit))
+		return -1;
+	if (*c == ':') {
+		c++;
+		p = simple_strtoul(c, (char **)&c, 10);
+		if (p < 1024 || p > 65535)
+			return -1;
+		*port = htons(p);
+	} else
+		*port = htons(SIP_PORT);
+
+	if (dataoff)
+		*dataoff = c - dptr;
+	return 1;
+}
+EXPORT_SYMBOL_GPL(ct_sip_parse_header_uri);
+
 /* SDP header parsing: a SDP session description contains an ordered set of
  * headers, starting with a section containing general session parameters,
  * optionally followed by multiple media descriptions.
-- 
cgit v1.2.3


From 624f8b7bba98c27a1464f5f858c4a861d5d3e8d7 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:19:30 -0700
Subject: [NETFILTER]: nf_nat_sip: get rid of text based header translation

Use the URI parsing helper to get the numerical addresses and get rid of the
text based header translation.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |   3 +-
 net/ipv4/netfilter/nf_nat_sip.c            | 100 +++++++++++------------------
 net/netfilter/nf_conntrack_sip.c           |  27 ++++++--
 3 files changed, 62 insertions(+), 68 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 87bc6f79efc4..68a0d6a41733 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -62,7 +62,8 @@ extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 
 extern int ct_sip_parse_request(const struct nf_conn *ct,
 				const char *dptr, unsigned int datalen,
-				unsigned int *matchoff, unsigned int *matchlen);
+				unsigned int *matchoff, unsigned int *matchlen,
+				union nf_inet_addr *addr, __be16 *port);
 extern int ct_sip_get_header(const struct nf_conn *ct, const char *dptr,
 			     unsigned int dataoff, unsigned int datalen,
 			     enum sip_header_types type,
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index c13e43862361..5b4a5cd23f39 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -26,39 +26,6 @@ MODULE_AUTHOR("Christian Hentschel <chentschel@arnet.com.ar>");
 MODULE_DESCRIPTION("SIP NAT helper");
 MODULE_ALIAS("ip_nat_sip");
 
-struct addr_map {
-	struct {
-		char		src[sizeof("nnn.nnn.nnn.nnn:nnnnn")];
-		char		dst[sizeof("nnn.nnn.nnn.nnn:nnnnn")];
-		unsigned int	srclen, srciplen;
-		unsigned int	dstlen, dstiplen;
-	} addr[IP_CT_DIR_MAX];
-};
-
-static void addr_map_init(const struct nf_conn *ct, struct addr_map *map)
-{
-	const struct nf_conntrack_tuple *t;
-	enum ip_conntrack_dir dir;
-	unsigned int n;
-
-	for (dir = 0; dir < IP_CT_DIR_MAX; dir++) {
-		t = &ct->tuplehash[dir].tuple;
-
-		n = sprintf(map->addr[dir].src, "%u.%u.%u.%u",
-			    NIPQUAD(t->src.u3.ip));
-		map->addr[dir].srciplen = n;
-		n += sprintf(map->addr[dir].src + n, ":%u",
-			     ntohs(t->src.u.udp.port));
-		map->addr[dir].srclen = n;
-
-		n = sprintf(map->addr[dir].dst, "%u.%u.%u.%u",
-			    NIPQUAD(t->dst.u3.ip));
-		map->addr[dir].dstiplen = n;
-		n += sprintf(map->addr[dir].dst + n, ":%u",
-			     ntohs(t->dst.u.udp.port));
-		map->addr[dir].dstlen = n;
-	}
-}
 
 static unsigned int mangle_packet(struct sk_buff *skb,
 				  const char **dptr, unsigned int *datalen,
@@ -81,43 +48,51 @@ static unsigned int mangle_packet(struct sk_buff *skb,
 static int map_addr(struct sk_buff *skb,
 		    const char **dptr, unsigned int *datalen,
 		    unsigned int matchoff, unsigned int matchlen,
-		    struct addr_map *map)
+		    union nf_inet_addr *addr, __be16 port)
 {
 	enum ip_conntrack_info ctinfo;
-	struct nf_conn *ct __maybe_unused = nf_ct_get(skb, &ctinfo);
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
-	unsigned int addrlen;
-	char *addr;
-
-	if ((matchlen == map->addr[dir].srciplen ||
-	     matchlen == map->addr[dir].srclen) &&
-	    strncmp(*dptr + matchoff, map->addr[dir].src, matchlen) == 0) {
-		addr    = map->addr[!dir].dst;
-		addrlen = map->addr[!dir].dstlen;
-	} else if ((matchlen == map->addr[dir].dstiplen ||
-		    matchlen == map->addr[dir].dstlen) &&
-		   strncmp(*dptr + matchoff, map->addr[dir].dst, matchlen) == 0) {
-		addr    = map->addr[!dir].src;
-		addrlen = map->addr[!dir].srclen;
+	char buffer[sizeof("nnn.nnn.nnn.nnn:nnnnn")];
+	unsigned int buflen;
+	__be32 newaddr;
+	__be16 newport;
+
+	if (ct->tuplehash[dir].tuple.src.u3.ip == addr->ip &&
+	    ct->tuplehash[dir].tuple.src.u.udp.port == port) {
+		newaddr = ct->tuplehash[!dir].tuple.dst.u3.ip;
+		newport = ct->tuplehash[!dir].tuple.dst.u.udp.port;
+	} else if (ct->tuplehash[dir].tuple.dst.u3.ip == addr->ip &&
+		   ct->tuplehash[dir].tuple.dst.u.udp.port == port) {
+		newaddr = ct->tuplehash[!dir].tuple.src.u3.ip;
+		newport = ct->tuplehash[!dir].tuple.src.u.udp.port;
 	} else
 		return 1;
 
+	if (newaddr == addr->ip && newport == port)
+		return 1;
+
+	buflen = sprintf(buffer, "%u.%u.%u.%u:%u",
+			 NIPQUAD(newaddr), ntohs(newport));
+
 	return mangle_packet(skb, dptr, datalen, matchoff, matchlen,
-			     addr, addrlen);
+			     buffer, buflen);
 }
 
 static int map_sip_addr(struct sk_buff *skb,
 			const char **dptr, unsigned int *datalen,
-			enum sip_header_types type, struct addr_map *map)
+			enum sip_header_types type)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	unsigned int matchlen, matchoff;
+	union nf_inet_addr addr;
+	__be16 port;
 
-	if (ct_sip_get_header(ct, *dptr, 0, *datalen, type,
-			      &matchoff, &matchlen) <= 0)
+	if (ct_sip_parse_header_uri(ct, *dptr, NULL, *datalen, type, NULL,
+				    &matchoff, &matchlen, &addr, &port) <= 0)
 		return 1;
-	return map_addr(skb, dptr, datalen, matchoff, matchlen, map);
+	return map_addr(skb, dptr, datalen, matchoff, matchlen, &addr, port);
 }
 
 static unsigned int ip_nat_sip(struct sk_buff *skb,
@@ -125,26 +100,27 @@ static unsigned int ip_nat_sip(struct sk_buff *skb,
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
-	struct addr_map map;
 	unsigned int matchoff, matchlen;
+	union nf_inet_addr addr;
+	__be16 port;
 
 	if (*datalen < strlen("SIP/2.0"))
 		return NF_ACCEPT;
 
-	addr_map_init(ct, &map);
-
 	/* Basic rules: requests and responses. */
 	if (strnicmp(*dptr, "SIP/2.0", strlen("SIP/2.0")) != 0) {
 		if (ct_sip_parse_request(ct, *dptr, *datalen,
-					 &matchoff, &matchlen) > 0 &&
-		    !map_addr(skb, dptr, datalen, matchoff, matchlen, &map))
+					 &matchoff, &matchlen,
+					 &addr, &port) > 0 &&
+		    !map_addr(skb, dptr, datalen, matchoff, matchlen,
+			      &addr, port))
 			return NF_DROP;
 	}
 
-	if (!map_sip_addr(skb, dptr, datalen, SIP_HDR_FROM, &map) ||
-	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_TO, &map) ||
-	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_VIA, &map) ||
-	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_CONTACT, &map))
+	if (!map_sip_addr(skb, dptr, datalen, SIP_HDR_FROM) ||
+	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_TO) ||
+	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_VIA) ||
+	    !map_sip_addr(skb, dptr, datalen, SIP_HDR_CONTACT))
 		return NF_DROP;
 	return NF_ACCEPT;
 }
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index a74d76a97312..f20fa2d94c0a 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -151,10 +151,12 @@ static int skp_epaddr_len(const struct nf_conn *ct, const char *dptr,
  */
 int ct_sip_parse_request(const struct nf_conn *ct,
 			 const char *dptr, unsigned int datalen,
-			 unsigned int *matchoff, unsigned int *matchlen)
+			 unsigned int *matchoff, unsigned int *matchlen,
+			 union nf_inet_addr *addr, __be16 *port)
 {
-	const char *start = dptr, *limit = dptr + datalen;
+	const char *start = dptr, *limit = dptr + datalen, *end;
 	unsigned int mlen;
+	unsigned int p;
 	int shift = 0;
 
 	/* Skip method and following whitespace */
@@ -173,10 +175,25 @@ int ct_sip_parse_request(const struct nf_conn *ct,
 		if (strnicmp(dptr, "sip:", strlen("sip:")) == 0)
 			break;
 	}
-	*matchlen = skp_epaddr_len(ct, dptr, limit, &shift);
-	if (!*matchlen)
+	if (!skp_epaddr_len(ct, dptr, limit, &shift))
 		return 0;
-	*matchoff = dptr - start + shift;
+	dptr += shift;
+
+	if (!parse_addr(ct, dptr, &end, addr, limit))
+		return -1;
+	if (end < limit && *end == ':') {
+		end++;
+		p = simple_strtoul(end, (char **)&end, 10);
+		if (p < 1024 || p > 65535)
+			return -1;
+		*port = htons(p);
+	} else
+		*port = htons(SIP_PORT);
+
+	if (end == dptr)
+		return 0;
+	*matchoff = dptr - start;
+	*matchlen = end - dptr;
 	return 1;
 }
 EXPORT_SYMBOL_GPL(ct_sip_parse_request);
-- 
cgit v1.2.3


From 30f33e6dee80c6ded917f978e4f377d1069d519d Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:22:20 -0700
Subject: [NETFILTER]: nf_conntrack_sip: support method specific
 request/response handling

Add support for per-method request/response handlers and perform SDP
parsing for INVITE/UPDATE requests and for all informational and
successful responses.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  20 ++++++
 net/netfilter/nf_conntrack_sip.c           | 107 ++++++++++++++++++++++++++---
 2 files changed, 117 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 68a0d6a41733..da93e80804c2 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -5,6 +5,25 @@
 #define SIP_PORT	5060
 #define SIP_TIMEOUT	3600
 
+struct sip_handler {
+	const char	*method;
+	unsigned int	len;
+	int		(*request)(struct sk_buff *skb,
+				   const char **dptr, unsigned int *datalen,
+				   unsigned int cseq);
+	int		(*response)(struct sk_buff *skb,
+				    const char **dptr, unsigned int *datalen,
+				    unsigned int cseq, unsigned int code);
+};
+
+#define SIP_HANDLER(__method, __request, __response)			\
+{									\
+	.method		= (__method),					\
+	.len		= sizeof(__method) - 1,				\
+	.request	= (__request),					\
+	.response	= (__response),					\
+}
+
 struct sip_header {
 	const char	*name;
 	const char	*cname;
@@ -35,6 +54,7 @@ struct sip_header {
 	__SIP_HDR(__name, NULL, __search, __match)
 
 enum sip_header_types {
+	SIP_HDR_CSEQ,
 	SIP_HDR_FROM,
 	SIP_HDR_TO,
 	SIP_HDR_CONTACT,
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index 96bedb52bd4b..1be949febab7 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -212,6 +212,7 @@ EXPORT_SYMBOL_GPL(ct_sip_parse_request);
  * equivalent to multiple headers.
  */
 static const struct sip_header ct_sip_hdrs[] = {
+	[SIP_HDR_CSEQ]			= SIP_HDR("CSeq", NULL, NULL, digits_len),
 	[SIP_HDR_FROM]			= SIP_HDR("From", "f", "sip:", skp_epaddr_len),
 	[SIP_HDR_TO]			= SIP_HDR("To", "t", "sip:", skp_epaddr_len),
 	[SIP_HDR_CONTACT]		= SIP_HDR("Contact", "m", "sip:", skp_epaddr_len),
@@ -566,7 +567,8 @@ static int set_expected_rtp(struct sk_buff *skb,
 }
 
 static int process_sdp(struct sk_buff *skb,
-		       const char **dptr, unsigned int *datalen)
+		       const char **dptr, unsigned int *datalen,
+		       unsigned int cseq)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
@@ -600,6 +602,96 @@ static int process_sdp(struct sk_buff *skb,
 
 	return set_expected_rtp(skb, dptr, datalen, &addr, htons(port));
 }
+static int process_invite_response(struct sk_buff *skb,
+				   const char **dptr, unsigned int *datalen,
+				   unsigned int cseq, unsigned int code)
+{
+	if ((code >= 100 && code <= 199) ||
+	    (code >= 200 && code <= 299))
+		return process_sdp(skb, dptr, datalen, cseq);
+
+	return NF_ACCEPT;
+}
+
+static int process_update_response(struct sk_buff *skb,
+				   const char **dptr, unsigned int *datalen,
+				   unsigned int cseq, unsigned int code)
+{
+	if ((code >= 100 && code <= 199) ||
+	    (code >= 200 && code <= 299))
+		return process_sdp(skb, dptr, datalen, cseq);
+
+	return NF_ACCEPT;
+}
+
+static const struct sip_handler sip_handlers[] = {
+	SIP_HANDLER("INVITE", process_sdp, process_invite_response),
+	SIP_HANDLER("UPDATE", process_sdp, process_update_response),
+};
+
+static int process_sip_response(struct sk_buff *skb,
+				const char **dptr, unsigned int *datalen)
+{
+	static const struct sip_handler *handler;
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+	unsigned int matchoff, matchlen;
+	unsigned int code, cseq, dataoff, i;
+
+	if (*datalen < strlen("SIP/2.0 200"))
+		return NF_ACCEPT;
+	code = simple_strtoul(*dptr + strlen("SIP/2.0 "), NULL, 10);
+	if (!code)
+		return NF_DROP;
+
+	if (ct_sip_get_header(ct, *dptr, 0, *datalen, SIP_HDR_CSEQ,
+			      &matchoff, &matchlen) <= 0)
+		return NF_DROP;
+	cseq = simple_strtoul(*dptr + matchoff, NULL, 10);
+	if (!cseq)
+		return NF_DROP;
+	dataoff = matchoff + matchlen + 1;
+
+	for (i = 0; i < ARRAY_SIZE(sip_handlers); i++) {
+		handler = &sip_handlers[i];
+		if (handler->response == NULL)
+			continue;
+		if (*datalen < dataoff + handler->len ||
+		    strnicmp(*dptr + dataoff, handler->method, handler->len))
+			continue;
+		return handler->response(skb, dptr, datalen, cseq, code);
+	}
+	return NF_ACCEPT;
+}
+
+static int process_sip_request(struct sk_buff *skb,
+			       const char **dptr, unsigned int *datalen)
+{
+	static const struct sip_handler *handler;
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+	unsigned int matchoff, matchlen;
+	unsigned int cseq, i;
+
+	for (i = 0; i < ARRAY_SIZE(sip_handlers); i++) {
+		handler = &sip_handlers[i];
+		if (handler->request == NULL)
+			continue;
+		if (*datalen < handler->len ||
+		    strnicmp(*dptr, handler->method, handler->len))
+			continue;
+
+		if (ct_sip_get_header(ct, *dptr, 0, *datalen, SIP_HDR_CSEQ,
+				      &matchoff, &matchlen) <= 0)
+			return NF_DROP;
+		cseq = simple_strtoul(*dptr + matchoff, NULL, 10);
+		if (!cseq)
+			return NF_DROP;
+
+		return handler->request(skb, dptr, datalen, cseq);
+	}
+	return NF_ACCEPT;
+}
 
 static int sip_help(struct sk_buff *skb,
 		    unsigned int protoff,
@@ -634,15 +726,10 @@ static int sip_help(struct sk_buff *skb,
 	if (datalen < strlen("SIP/2.0 200"))
 		return NF_ACCEPT;
 
-	/* RTP info only in some SDP pkts */
-	if (strnicmp(dptr, "INVITE", strlen("INVITE")) != 0 &&
-	    strnicmp(dptr, "UPDATE", strlen("UPDATE")) != 0 &&
-	    strnicmp(dptr, "SIP/2.0 180", strlen("SIP/2.0 180")) != 0 &&
-	    strnicmp(dptr, "SIP/2.0 183", strlen("SIP/2.0 183")) != 0 &&
-	    strnicmp(dptr, "SIP/2.0 200", strlen("SIP/2.0 200")) != 0)
-		return NF_ACCEPT;
-
-	return process_sdp(skb, &dptr, &datalen);
+	if (strnicmp(dptr, "SIP/2.0 ", strlen("SIP/2.0 ")) != 0)
+		return process_sip_request(skb, &dptr, &datalen);
+	else
+		return process_sip_response(skb, &dptr, &datalen);
 }
 
 static struct nf_conntrack_helper sip[MAX_PORTS][2] __read_mostly;
-- 
cgit v1.2.3


From 2bbb21168a90c788e12fe722eb66f27e611e7df7 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:24:24 -0700
Subject: [NETFILTER]: nf_conntrack_sip: introduce URI and header parameter
 parsing helpers

Introduce URI and header parameter parsing helpers. These are needed
by the conntrack helper to parse expiration values in Contact: header
parameters and by the NAT helper to properly update the Via-header
rport=, received= and maddr= parameters.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h | 10 ++++++
 net/netfilter/nf_conntrack_sip.c           | 58 ++++++++++++++++++++++++++++++
 2 files changed, 68 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index da93e80804c2..87e402825dba 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -93,6 +93,16 @@ extern int ct_sip_parse_header_uri(const struct nf_conn *ct, const char *dptr,
 				   enum sip_header_types type, int *in_header,
 				   unsigned int *matchoff, unsigned int *matchlen,
 				   union nf_inet_addr *addr, __be16 *port);
+extern int ct_sip_parse_address_param(const struct nf_conn *ct, const char *dptr,
+				      unsigned int dataoff, unsigned int datalen,
+				      const char *name,
+				      unsigned int *matchoff, unsigned int *matchlen,
+				      union nf_inet_addr *addr);
+extern int ct_sip_parse_numerical_param(const struct nf_conn *ct, const char *dptr,
+					unsigned int off, unsigned int datalen,
+					const char *name,
+					unsigned int *matchoff, unsigned int *matchen,
+					unsigned int *val);
 
 extern int ct_sip_get_sdp_header(const struct nf_conn *ct, const char *dptr,
 				 unsigned int dataoff, unsigned int datalen,
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index 8e7e5b465ffb..126f30842d60 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -448,6 +448,64 @@ int ct_sip_parse_header_uri(const struct nf_conn *ct, const char *dptr,
 }
 EXPORT_SYMBOL_GPL(ct_sip_parse_header_uri);
 
+/* Parse address from header parameter and return address, offset and length */
+int ct_sip_parse_address_param(const struct nf_conn *ct, const char *dptr,
+			       unsigned int dataoff, unsigned int datalen,
+			       const char *name,
+			       unsigned int *matchoff, unsigned int *matchlen,
+			       union nf_inet_addr *addr)
+{
+	const char *limit = dptr + datalen;
+	const char *start, *end;
+
+	limit = ct_sip_header_search(dptr + dataoff, limit, ",", strlen(","));
+	if (!limit)
+		limit = dptr + datalen;
+
+	start = ct_sip_header_search(dptr + dataoff, limit, name, strlen(name));
+	if (!start)
+		return 0;
+
+	start += strlen(name);
+	if (!parse_addr(ct, start, &end, addr, limit))
+		return 0;
+	*matchoff = start - dptr;
+	*matchlen = end - start;
+	return 1;
+}
+EXPORT_SYMBOL_GPL(ct_sip_parse_address_param);
+
+/* Parse numerical header parameter and return value, offset and length */
+int ct_sip_parse_numerical_param(const struct nf_conn *ct, const char *dptr,
+				 unsigned int dataoff, unsigned int datalen,
+				 const char *name,
+				 unsigned int *matchoff, unsigned int *matchlen,
+				 unsigned int *val)
+{
+	const char *limit = dptr + datalen;
+	const char *start;
+	char *end;
+
+	limit = ct_sip_header_search(dptr + dataoff, limit, ",", strlen(","));
+	if (!limit)
+		limit = dptr + datalen;
+
+	start = ct_sip_header_search(dptr + dataoff, limit, name, strlen(name));
+	if (!start)
+		return 0;
+
+	start += strlen(name);
+	*val = simple_strtoul(start, &end, 0);
+	if (start == end)
+		return 0;
+	if (matchoff && matchlen) {
+		*matchoff = start - dptr;
+		*matchlen = end - start;
+	}
+	return 1;
+}
+EXPORT_SYMBOL_GPL(ct_sip_parse_numerical_param);
+
 /* SDP header parsing: a SDP session description contains an ordered set of
  * headers, starting with a section containing general session parameters,
  * optionally followed by multiple media descriptions.
-- 
cgit v1.2.3


From 0f32a40fc91a9ebbbf66e826ac2a829ab37d9cf8 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:25:13 -0700
Subject: [NETFILTER]: nf_conntrack_sip: create signalling expectations

Create expectations for incoming signalling connections when seeing
a REGISTER request. This is needed when the registrar uses a
different source port number for signalling messages and for receiving
incoming calls from other endpoints than the registrar.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  18 +++
 include/net/netfilter/nf_conntrack.h       |   4 +-
 net/ipv4/netfilter/nf_nat_sip.c            | 111 +++++++++++---
 net/netfilter/nf_conntrack_sip.c           | 232 +++++++++++++++++++++++++++--
 4 files changed, 332 insertions(+), 33 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 87e402825dba..7cc84ed0c5da 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -5,6 +5,17 @@
 #define SIP_PORT	5060
 #define SIP_TIMEOUT	3600
 
+struct nf_ct_sip_master {
+	unsigned int	register_cseq;
+};
+
+enum sip_expectation_classes {
+	SIP_EXPECT_SIGNALLING,
+	SIP_EXPECT_AUDIO,
+	__SIP_EXPECT_MAX
+};
+#define SIP_EXPECT_MAX	(__SIP_EXPECT_MAX - 1)
+
 struct sip_handler {
 	const char	*method;
 	unsigned int	len;
@@ -59,6 +70,7 @@ enum sip_header_types {
 	SIP_HDR_TO,
 	SIP_HDR_CONTACT,
 	SIP_HDR_VIA,
+	SIP_HDR_EXPIRES,
 	SIP_HDR_CONTENT_LENGTH,
 };
 
@@ -75,6 +87,12 @@ enum sdp_header_types {
 extern unsigned int (*nf_nat_sip_hook)(struct sk_buff *skb,
 				       const char **dptr,
 				       unsigned int *datalen);
+extern unsigned int (*nf_nat_sip_expect_hook)(struct sk_buff *skb,
+					      const char **dptr,
+					      unsigned int *datalen,
+					      struct nf_conntrack_expect *exp,
+					      unsigned int matchoff,
+					      unsigned int matchlen);
 extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				       const char **dptr,
 				       unsigned int *datalen,
diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
index 922877133598..4a4f870d2a5e 100644
--- a/include/net/netfilter/nf_conntrack.h
+++ b/include/net/netfilter/nf_conntrack.h
@@ -46,6 +46,7 @@ union nf_conntrack_expect_proto {
 #include <linux/netfilter/nf_conntrack_pptp.h>
 #include <linux/netfilter/nf_conntrack_h323.h>
 #include <linux/netfilter/nf_conntrack_sane.h>
+#include <linux/netfilter/nf_conntrack_sip.h>
 
 /* per conntrack: application helper private data */
 union nf_conntrack_help {
@@ -54,6 +55,7 @@ union nf_conntrack_help {
 	struct nf_ct_pptp_master ct_pptp_info;
 	struct nf_ct_h323_master ct_h323_info;
 	struct nf_ct_sane_master ct_sane_info;
+	struct nf_ct_sip_master ct_sip_info;
 };
 
 #include <linux/types.h>
@@ -76,7 +78,7 @@ do {									\
 struct nf_conntrack_helper;
 
 /* Must be kept in sync with the classes defined by helpers */
-#define NF_CT_MAX_EXPECT_CLASSES	1
+#define NF_CT_MAX_EXPECT_CLASSES	2
 
 /* nf_conn feature for connections that have a helper */
 struct nf_conn_help {
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index b443618a857f..4b85e21a2a4a 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -205,6 +205,91 @@ next:
 	return NF_ACCEPT;
 }
 
+/* Handles expected signalling connections and media streams */
+static void ip_nat_sip_expected(struct nf_conn *ct,
+				struct nf_conntrack_expect *exp)
+{
+	struct nf_nat_range range;
+
+	/* This must be a fresh one. */
+	BUG_ON(ct->status & IPS_NAT_DONE_MASK);
+
+	/* For DST manip, map port here to where it's expected. */
+	range.flags = (IP_NAT_RANGE_MAP_IPS | IP_NAT_RANGE_PROTO_SPECIFIED);
+	range.min = range.max = exp->saved_proto;
+	range.min_ip = range.max_ip = exp->saved_ip;
+	nf_nat_setup_info(ct, &range, IP_NAT_MANIP_DST);
+
+	/* Change src to where master sends to, but only if the connection
+	 * actually came from the same source. */
+	if (ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.u3.ip ==
+	    ct->master->tuplehash[exp->dir].tuple.src.u3.ip) {
+		range.flags = IP_NAT_RANGE_MAP_IPS;
+		range.min_ip = range.max_ip
+			= ct->master->tuplehash[!exp->dir].tuple.dst.u3.ip;
+		nf_nat_setup_info(ct, &range, IP_NAT_MANIP_SRC);
+	}
+}
+
+static unsigned int ip_nat_sip_expect(struct sk_buff *skb,
+				      const char **dptr, unsigned int *datalen,
+				      struct nf_conntrack_expect *exp,
+				      unsigned int matchoff,
+				      unsigned int matchlen)
+{
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
+	__be32 newip;
+	u_int16_t port;
+	char buffer[sizeof("nnn.nnn.nnn.nnn:nnnnn")];
+	unsigned buflen;
+
+	/* Connection will come from reply */
+	if (ct->tuplehash[dir].tuple.src.u3.ip == ct->tuplehash[!dir].tuple.dst.u3.ip)
+		newip = exp->tuple.dst.u3.ip;
+	else
+		newip = ct->tuplehash[!dir].tuple.dst.u3.ip;
+
+	/* If the signalling port matches the connection's source port in the
+	 * original direction, try to use the destination port in the opposite
+	 * direction. */
+	if (exp->tuple.dst.u.udp.port ==
+	    ct->tuplehash[dir].tuple.src.u.udp.port)
+		port = ntohs(ct->tuplehash[!dir].tuple.dst.u.udp.port);
+	else
+		port = ntohs(exp->tuple.dst.u.udp.port);
+
+	exp->saved_ip = exp->tuple.dst.u3.ip;
+	exp->tuple.dst.u3.ip = newip;
+	exp->saved_proto.udp.port = exp->tuple.dst.u.udp.port;
+	exp->dir = !dir;
+	exp->expectfn = ip_nat_sip_expected;
+
+	for (; port != 0; port++) {
+		exp->tuple.dst.u.udp.port = htons(port);
+		if (nf_ct_expect_related(exp) == 0)
+			break;
+	}
+
+	if (port == 0)
+		return NF_DROP;
+
+	if (exp->tuple.dst.u3.ip != exp->saved_ip ||
+	    exp->tuple.dst.u.udp.port != exp->saved_proto.udp.port) {
+		buflen = sprintf(buffer, "%u.%u.%u.%u:%u",
+				 NIPQUAD(newip), port);
+		if (!mangle_packet(skb, dptr, datalen, matchoff, matchlen,
+				   buffer, buflen))
+			goto err;
+	}
+	return NF_ACCEPT;
+
+err:
+	nf_ct_unexpect_related(exp);
+	return NF_DROP;
+}
+
 static int mangle_content_len(struct sk_buff *skb,
 			      const char **dptr, unsigned int *datalen)
 {
@@ -275,27 +360,6 @@ static unsigned int mangle_sdp(struct sk_buff *skb,
 	return mangle_content_len(skb, dptr, datalen);
 }
 
-static void ip_nat_sdp_expect(struct nf_conn *ct,
-			      struct nf_conntrack_expect *exp)
-{
-	struct nf_nat_range range;
-
-	/* This must be a fresh one. */
-	BUG_ON(ct->status & IPS_NAT_DONE_MASK);
-
-	/* For DST manip, map port here to where it's expected. */
-	range.flags = (IP_NAT_RANGE_MAP_IPS | IP_NAT_RANGE_PROTO_SPECIFIED);
-	range.min = range.max = exp->saved_proto;
-	range.min_ip = range.max_ip = exp->saved_ip;
-	nf_nat_setup_info(ct, &range, IP_NAT_MANIP_DST);
-
-	/* Change src to where master sends to */
-	range.flags = IP_NAT_RANGE_MAP_IPS;
-	range.min_ip = range.max_ip
-		= ct->master->tuplehash[!exp->dir].tuple.dst.u3.ip;
-	nf_nat_setup_info(ct, &range, IP_NAT_MANIP_SRC);
-}
-
 /* So, this packet has hit the connection tracking matching code.
    Mangle it, and change the expectation to match the new version. */
 static unsigned int ip_nat_sdp(struct sk_buff *skb,
@@ -322,7 +386,7 @@ static unsigned int ip_nat_sdp(struct sk_buff *skb,
 
 	/* When you see the packet, we need to NAT it the same as the
 	   this one. */
-	exp->expectfn = ip_nat_sdp_expect;
+	exp->expectfn = ip_nat_sip_expected;
 
 	/* Try to get same port: if not, try to change it. */
 	for (port = ntohs(exp->saved_proto.udp.port); port != 0; port++) {
@@ -344,6 +408,7 @@ static unsigned int ip_nat_sdp(struct sk_buff *skb,
 static void __exit nf_nat_sip_fini(void)
 {
 	rcu_assign_pointer(nf_nat_sip_hook, NULL);
+	rcu_assign_pointer(nf_nat_sip_expect_hook, NULL);
 	rcu_assign_pointer(nf_nat_sdp_hook, NULL);
 	synchronize_rcu();
 }
@@ -351,8 +416,10 @@ static void __exit nf_nat_sip_fini(void)
 static int __init nf_nat_sip_init(void)
 {
 	BUG_ON(nf_nat_sip_hook != NULL);
+	BUG_ON(nf_nat_sip_expect_hook != NULL);
 	BUG_ON(nf_nat_sdp_hook != NULL);
 	rcu_assign_pointer(nf_nat_sip_hook, ip_nat_sip);
+	rcu_assign_pointer(nf_nat_sip_expect_hook, ip_nat_sip_expect);
 	rcu_assign_pointer(nf_nat_sdp_hook, ip_nat_sdp);
 	return 0;
 }
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index 126f30842d60..043aa557e7a8 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -37,11 +37,24 @@ static unsigned int sip_timeout __read_mostly = SIP_TIMEOUT;
 module_param(sip_timeout, uint, 0600);
 MODULE_PARM_DESC(sip_timeout, "timeout for the master SIP session");
 
+static int sip_direct_signalling __read_mostly = 1;
+module_param(sip_direct_signalling, int, 0600);
+MODULE_PARM_DESC(sip_direct_signalling, "expect incoming calls from registrar "
+					"only (default 1)");
+
 unsigned int (*nf_nat_sip_hook)(struct sk_buff *skb,
 				const char **dptr,
 				unsigned int *datalen) __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sip_hook);
 
+unsigned int (*nf_nat_sip_expect_hook)(struct sk_buff *skb,
+				       const char **dptr,
+				       unsigned int *datalen,
+				       struct nf_conntrack_expect *exp,
+				       unsigned int matchoff,
+				       unsigned int matchlen) __read_mostly;
+EXPORT_SYMBOL_GPL(nf_nat_sip_expect_hook);
+
 unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				const char **dptr,
 				unsigned int *datalen,
@@ -218,6 +231,7 @@ static const struct sip_header ct_sip_hdrs[] = {
 	[SIP_HDR_TO]			= SIP_HDR("To", "t", "sip:", skp_epaddr_len),
 	[SIP_HDR_CONTACT]		= SIP_HDR("Contact", "m", "sip:", skp_epaddr_len),
 	[SIP_HDR_VIA]			= SIP_HDR("Via", "v", "UDP ", epaddr_len),
+	[SIP_HDR_EXPIRES]		= SIP_HDR("Expires", NULL, NULL, digits_len),
 	[SIP_HDR_CONTENT_LENGTH]	= SIP_HDR("Content-Length", "l", NULL, digits_len),
 };
 
@@ -592,18 +606,50 @@ int ct_sip_get_sdp_header(const struct nf_conn *ct, const char *dptr,
 }
 EXPORT_SYMBOL_GPL(ct_sip_get_sdp_header);
 
-static void flush_expectations(struct nf_conn *ct)
+static int refresh_signalling_expectation(struct nf_conn *ct,
+					  union nf_inet_addr *addr,
+					  __be16 port,
+					  unsigned int expires)
 {
 	struct nf_conn_help *help = nfct_help(ct);
 	struct nf_conntrack_expect *exp;
 	struct hlist_node *n, *next;
+	int found = 0;
 
 	spin_lock_bh(&nf_conntrack_lock);
 	hlist_for_each_entry_safe(exp, n, next, &help->expectations, lnode) {
+		if (exp->class != SIP_EXPECT_SIGNALLING ||
+		    !nf_inet_addr_cmp(&exp->tuple.dst.u3, addr) ||
+		    exp->tuple.dst.u.udp.port != port)
+			continue;
+		if (!del_timer(&exp->timeout))
+			continue;
+		exp->flags &= ~NF_CT_EXPECT_INACTIVE;
+		exp->timeout.expires = jiffies + expires * HZ;
+		add_timer(&exp->timeout);
+		found = 1;
+		break;
+	}
+	spin_unlock_bh(&nf_conntrack_lock);
+	return found;
+}
+
+static void flush_expectations(struct nf_conn *ct, bool media)
+{
+	struct nf_conn_help *help = nfct_help(ct);
+	struct nf_conntrack_expect *exp;
+	struct hlist_node *n, *next;
+
+	spin_lock_bh(&nf_conntrack_lock);
+	hlist_for_each_entry_safe(exp, n, next, &help->expectations, lnode) {
+		if ((exp->class != SIP_EXPECT_SIGNALLING) ^ media)
+			continue;
 		if (!del_timer(&exp->timeout))
 			continue;
 		nf_ct_unlink_expect(exp);
 		nf_ct_expect_put(exp);
+		if (!media)
+			break;
 	}
 	spin_unlock_bh(&nf_conntrack_lock);
 }
@@ -623,7 +669,7 @@ static int set_expected_rtp(struct sk_buff *skb,
 	exp = nf_ct_expect_alloc(ct);
 	if (exp == NULL)
 		return NF_DROP;
-	nf_ct_expect_init(exp, NF_CT_EXPECT_CLASS_DEFAULT, family,
+	nf_ct_expect_init(exp, SIP_EXPECT_AUDIO, family,
 			  &ct->tuplehash[!dir].tuple.src.u3, addr,
 			  IPPROTO_UDP, NULL, &port);
 
@@ -688,7 +734,7 @@ static int process_invite_response(struct sk_buff *skb,
 	    (code >= 200 && code <= 299))
 		return process_sdp(skb, dptr, datalen, cseq);
 	else {
-		flush_expectations(ct);
+		flush_expectations(ct, true);
 		return NF_ACCEPT;
 	}
 }
@@ -704,7 +750,7 @@ static int process_update_response(struct sk_buff *skb,
 	    (code >= 200 && code <= 299))
 		return process_sdp(skb, dptr, datalen, cseq);
 	else {
-		flush_expectations(ct);
+		flush_expectations(ct, true);
 		return NF_ACCEPT;
 	}
 }
@@ -720,7 +766,7 @@ static int process_prack_response(struct sk_buff *skb,
 	    (code >= 200 && code <= 299))
 		return process_sdp(skb, dptr, datalen, cseq);
 	else {
-		flush_expectations(ct);
+		flush_expectations(ct, true);
 		return NF_ACCEPT;
 	}
 }
@@ -732,7 +778,165 @@ static int process_bye_request(struct sk_buff *skb,
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 
-	flush_expectations(ct);
+	flush_expectations(ct, true);
+	return NF_ACCEPT;
+}
+
+/* Parse a REGISTER request and create a permanent expectation for incoming
+ * signalling connections. The expectation is marked inactive and is activated
+ * when receiving a response indicating success from the registrar.
+ */
+static int process_register_request(struct sk_buff *skb,
+				    const char **dptr, unsigned int *datalen,
+				    unsigned int cseq)
+{
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+	struct nf_conn_help *help = nfct_help(ct);
+	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
+	int family = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num;
+	unsigned int matchoff, matchlen;
+	struct nf_conntrack_expect *exp;
+	union nf_inet_addr *saddr, daddr;
+	__be16 port;
+	unsigned int expires = 0;
+	int ret;
+	typeof(nf_nat_sip_expect_hook) nf_nat_sip_expect;
+
+	/* Expected connections can not register again. */
+	if (ct->status & IPS_EXPECTED)
+		return NF_ACCEPT;
+
+	/* We must check the expiration time: a value of zero signals the
+	 * registrar to release the binding. We'll remove our expectation
+	 * when receiving the new bindings in the response, but we don't
+	 * want to create new ones.
+	 *
+	 * The expiration time may be contained in Expires: header, the
+	 * Contact: header parameters or the URI parameters.
+	 */
+	if (ct_sip_get_header(ct, *dptr, 0, *datalen, SIP_HDR_EXPIRES,
+			      &matchoff, &matchlen) > 0)
+		expires = simple_strtoul(*dptr + matchoff, NULL, 10);
+
+	ret = ct_sip_parse_header_uri(ct, *dptr, NULL, *datalen,
+				      SIP_HDR_CONTACT, NULL,
+				      &matchoff, &matchlen, &daddr, &port);
+	if (ret < 0)
+		return NF_DROP;
+	else if (ret == 0)
+		return NF_ACCEPT;
+
+	/* We don't support third-party registrations */
+	if (!nf_inet_addr_cmp(&ct->tuplehash[dir].tuple.src.u3, &daddr))
+		return NF_ACCEPT;
+
+	if (ct_sip_parse_numerical_param(ct, *dptr,
+					 matchoff + matchlen, *datalen,
+					 "expires=", NULL, NULL, &expires) < 0)
+		return NF_DROP;
+
+	if (expires == 0) {
+		ret = NF_ACCEPT;
+		goto store_cseq;
+	}
+
+	exp = nf_ct_expect_alloc(ct);
+	if (!exp)
+		return NF_DROP;
+
+	saddr = NULL;
+	if (sip_direct_signalling)
+		saddr = &ct->tuplehash[!dir].tuple.src.u3;
+
+	nf_ct_expect_init(exp, SIP_EXPECT_SIGNALLING, family, saddr, &daddr,
+			  IPPROTO_UDP, NULL, &port);
+	exp->timeout.expires = sip_timeout * HZ;
+	exp->helper = nfct_help(ct)->helper;
+	exp->flags = NF_CT_EXPECT_PERMANENT | NF_CT_EXPECT_INACTIVE;
+
+	nf_nat_sip_expect = rcu_dereference(nf_nat_sip_expect_hook);
+	if (nf_nat_sip_expect && ct->status & IPS_NAT_MASK)
+		ret = nf_nat_sip_expect(skb, dptr, datalen, exp,
+					matchoff, matchlen);
+	else {
+		if (nf_ct_expect_related(exp) != 0)
+			ret = NF_DROP;
+		else
+			ret = NF_ACCEPT;
+	}
+	nf_ct_expect_put(exp);
+
+store_cseq:
+	if (ret == NF_ACCEPT)
+		help->help.ct_sip_info.register_cseq = cseq;
+	return ret;
+}
+
+static int process_register_response(struct sk_buff *skb,
+				     const char **dptr, unsigned int *datalen,
+				     unsigned int cseq, unsigned int code)
+{
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
+	struct nf_conn_help *help = nfct_help(ct);
+	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
+	union nf_inet_addr addr;
+	__be16 port;
+	unsigned int matchoff, matchlen, dataoff = 0;
+	unsigned int expires = 0;
+	int in_contact = 0, ret;
+
+	/* According to RFC 3261, "UAs MUST NOT send a new registration until
+	 * they have received a final response from the registrar for the
+	 * previous one or the previous REGISTER request has timed out".
+	 *
+	 * However, some servers fail to detect retransmissions and send late
+	 * responses, so we store the sequence number of the last valid
+	 * request and compare it here.
+	 */
+	if (help->help.ct_sip_info.register_cseq != cseq)
+		return NF_ACCEPT;
+
+	if (code >= 100 && code <= 199)
+		return NF_ACCEPT;
+	if (code < 200 || code > 299)
+		goto flush;
+
+	if (ct_sip_get_header(ct, *dptr, 0, *datalen, SIP_HDR_EXPIRES,
+			      &matchoff, &matchlen) > 0)
+		expires = simple_strtoul(*dptr + matchoff, NULL, 10);
+
+	while (1) {
+		unsigned int c_expires = expires;
+
+		ret = ct_sip_parse_header_uri(ct, *dptr, &dataoff, *datalen,
+					      SIP_HDR_CONTACT, &in_contact,
+					      &matchoff, &matchlen,
+					      &addr, &port);
+		if (ret < 0)
+			return NF_DROP;
+		else if (ret == 0)
+			break;
+
+		/* We don't support third-party registrations */
+		if (!nf_inet_addr_cmp(&ct->tuplehash[dir].tuple.dst.u3, &addr))
+			continue;
+
+		ret = ct_sip_parse_numerical_param(ct, *dptr,
+						   matchoff + matchlen,
+						   *datalen, "expires=",
+						   NULL, NULL, &c_expires);
+		if (ret < 0)
+			return NF_DROP;
+		if (c_expires == 0)
+			break;
+		if (refresh_signalling_expectation(ct, &addr, port, c_expires))
+			return NF_ACCEPT;
+	}
+
+flush:
+	flush_expectations(ct, false);
 	return NF_ACCEPT;
 }
 
@@ -742,6 +946,7 @@ static const struct sip_handler sip_handlers[] = {
 	SIP_HANDLER("ACK", process_sdp, NULL),
 	SIP_HANDLER("PRACK", process_sdp, process_prack_response),
 	SIP_HANDLER("BYE", process_bye_request, NULL),
+	SIP_HANDLER("REGISTER", process_register_request, process_register_response),
 };
 
 static int process_sip_response(struct sk_buff *skb,
@@ -853,9 +1058,15 @@ static int sip_help(struct sk_buff *skb,
 static struct nf_conntrack_helper sip[MAX_PORTS][2] __read_mostly;
 static char sip_names[MAX_PORTS][2][sizeof("sip-65535")] __read_mostly;
 
-static const struct nf_conntrack_expect_policy sip_exp_policy = {
-	.max_expected	= 2,
-	.timeout	= 3 * 60,
+static const struct nf_conntrack_expect_policy sip_exp_policy[SIP_EXPECT_MAX + 1] = {
+	[SIP_EXPECT_SIGNALLING] = {
+		.max_expected	= 1,
+		.timeout	= 3 * 60,
+	},
+	[SIP_EXPECT_AUDIO] = {
+		.max_expected	= IP_CT_DIR_MAX,
+		.timeout	= 3 * 60,
+	},
 };
 
 static void nf_conntrack_sip_fini(void)
@@ -887,7 +1098,8 @@ static int __init nf_conntrack_sip_init(void)
 		for (j = 0; j < 2; j++) {
 			sip[i][j].tuple.dst.protonum = IPPROTO_UDP;
 			sip[i][j].tuple.src.u.udp.port = htons(ports[i]);
-			sip[i][j].expect_policy = &sip_exp_policy;
+			sip[i][j].expect_policy = sip_exp_policy;
+			sip[i][j].expect_class_max = SIP_EXPECT_MAX;
 			sip[i][j].me = THIS_MODULE;
 			sip[i][j].help = sip_help;
 
-- 
cgit v1.2.3


From a9c1d35917c0c95c8f95a8e497fb91e301419693 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:25:49 -0700
Subject: [NETFILTER]: nf_conntrack_sip: create RTCP expectations

Create expectations for the RTCP connections in addition to RTP connections.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  3 +-
 net/ipv4/netfilter/nf_nat_sip.c            | 42 +++++++++++++---------
 net/netfilter/nf_conntrack_sip.c           | 58 +++++++++++++++++++-----------
 3 files changed, 66 insertions(+), 37 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 7cc84ed0c5da..6ddf95f51fb5 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -96,7 +96,8 @@ extern unsigned int (*nf_nat_sip_expect_hook)(struct sk_buff *skb,
 extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				       const char **dptr,
 				       unsigned int *datalen,
-				       struct nf_conntrack_expect *exp);
+				       struct nf_conntrack_expect *rtp_exp,
+				       struct nf_conntrack_expect *rtcp_exp);
 
 extern int ct_sip_parse_request(const struct nf_conn *ct,
 				const char *dptr, unsigned int datalen,
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index 4b85e21a2a4a..f73ab4883b75 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -364,7 +364,8 @@ static unsigned int mangle_sdp(struct sk_buff *skb,
    Mangle it, and change the expectation to match the new version. */
 static unsigned int ip_nat_sdp(struct sk_buff *skb,
 			       const char **dptr, unsigned int *datalen,
-			       struct nf_conntrack_expect *exp)
+			       struct nf_conntrack_expect *rtp_exp,
+			       struct nf_conntrack_expect *rtcp_exp)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
@@ -375,31 +376,40 @@ static unsigned int ip_nat_sdp(struct sk_buff *skb,
 	/* Connection will come from reply */
 	if (ct->tuplehash[dir].tuple.src.u3.ip ==
 	    ct->tuplehash[!dir].tuple.dst.u3.ip)
-		newip = exp->tuple.dst.u3.ip;
+		newip = rtp_exp->tuple.dst.u3.ip;
 	else
 		newip = ct->tuplehash[!dir].tuple.dst.u3.ip;
 
-	exp->saved_ip = exp->tuple.dst.u3.ip;
-	exp->tuple.dst.u3.ip = newip;
-	exp->saved_proto.udp.port = exp->tuple.dst.u.udp.port;
-	exp->dir = !dir;
-
-	/* When you see the packet, we need to NAT it the same as the
-	   this one. */
-	exp->expectfn = ip_nat_sip_expected;
-
-	/* Try to get same port: if not, try to change it. */
-	for (port = ntohs(exp->saved_proto.udp.port); port != 0; port++) {
-		exp->tuple.dst.u.udp.port = htons(port);
-		if (nf_ct_expect_related(exp) == 0)
+	rtp_exp->saved_ip = rtp_exp->tuple.dst.u3.ip;
+	rtp_exp->tuple.dst.u3.ip = newip;
+	rtp_exp->saved_proto.udp.port = rtp_exp->tuple.dst.u.udp.port;
+	rtp_exp->dir = !dir;
+	rtp_exp->expectfn = ip_nat_sip_expected;
+
+	rtcp_exp->saved_ip = rtcp_exp->tuple.dst.u3.ip;
+	rtcp_exp->tuple.dst.u3.ip = newip;
+	rtcp_exp->saved_proto.udp.port = rtcp_exp->tuple.dst.u.udp.port;
+	rtcp_exp->dir = !dir;
+	rtcp_exp->expectfn = ip_nat_sip_expected;
+
+	/* Try to get same pair of ports: if not, try to change them. */
+	for (port = ntohs(rtp_exp->tuple.dst.u.udp.port);
+	     port != 0; port += 2) {
+		rtp_exp->tuple.dst.u.udp.port = htons(port);
+		if (nf_ct_expect_related(rtp_exp) != 0)
+			continue;
+		rtcp_exp->tuple.dst.u.udp.port = htons(port + 1);
+		if (nf_ct_expect_related(rtcp_exp) == 0)
 			break;
+		nf_ct_unexpect_related(rtp_exp);
 	}
 
 	if (port == 0)
 		return NF_DROP;
 
 	if (!mangle_sdp(skb, ctinfo, ct, newip, port, dptr, datalen)) {
-		nf_ct_unexpect_related(exp);
+		nf_ct_unexpect_related(rtp_exp);
+		nf_ct_unexpect_related(rtcp_exp);
 		return NF_DROP;
 	}
 	return NF_ACCEPT;
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index 813aa8c67e4c..217262e23403 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -63,7 +63,9 @@ EXPORT_SYMBOL_GPL(nf_nat_sip_expect_hook);
 unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
 				const char **dptr,
 				unsigned int *datalen,
-				struct nf_conntrack_expect *exp) __read_mostly;
+				struct nf_conntrack_expect *rtp_exp,
+				struct nf_conntrack_expect *rtcp_exp)
+				__read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sdp_hook);
 
 static int string_len(const struct nf_conn *ct, const char *dptr,
@@ -659,18 +661,20 @@ static void flush_expectations(struct nf_conn *ct, bool media)
 	spin_unlock_bh(&nf_conntrack_lock);
 }
 
-static int set_expected_rtp(struct sk_buff *skb,
-			    const char **dptr, unsigned int *datalen,
-			    union nf_inet_addr *daddr, __be16 port)
+static int set_expected_rtp_rtcp(struct sk_buff *skb,
+				 const char **dptr, unsigned int *datalen,
+				 union nf_inet_addr *daddr, __be16 port)
 {
-	struct nf_conntrack_expect *exp;
+	struct nf_conntrack_expect *exp, *rtp_exp, *rtcp_exp;
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
 	union nf_inet_addr *saddr;
 	struct nf_conntrack_tuple tuple;
 	int family = ct->tuplehash[!dir].tuple.src.l3num;
-	int skip_expect = 0, ret;
+	int skip_expect = 0, ret = NF_DROP;
+	u_int16_t base_port;
+	__be16 rtp_port, rtcp_port;
 	typeof(nf_nat_sdp_hook) nf_nat_sdp;
 
 	saddr = NULL;
@@ -704,23 +708,37 @@ static int set_expected_rtp(struct sk_buff *skb,
 	if (skip_expect)
 		return NF_ACCEPT;
 
-	exp = nf_ct_expect_alloc(ct);
-	if (exp == NULL)
-		return NF_DROP;
-	nf_ct_expect_init(exp, SIP_EXPECT_AUDIO, family, saddr, daddr,
-			  IPPROTO_UDP, NULL, &port);
+	base_port = ntohs(tuple.dst.u.udp.port) & ~1;
+	rtp_port = htons(base_port);
+	rtcp_port = htons(base_port + 1);
+
+	rtp_exp = nf_ct_expect_alloc(ct);
+	if (rtp_exp == NULL)
+		goto err1;
+	nf_ct_expect_init(rtp_exp, SIP_EXPECT_AUDIO, family, saddr, daddr,
+			  IPPROTO_UDP, NULL, &rtp_port);
+
+	rtcp_exp = nf_ct_expect_alloc(ct);
+	if (rtcp_exp == NULL)
+		goto err2;
+	nf_ct_expect_init(rtcp_exp, SIP_EXPECT_AUDIO, family, saddr, daddr,
+			  IPPROTO_UDP, NULL, &rtcp_port);
 
 	nf_nat_sdp = rcu_dereference(nf_nat_sdp_hook);
 	if (nf_nat_sdp && ct->status & IPS_NAT_MASK)
-		ret = nf_nat_sdp(skb, dptr, datalen, exp);
+		ret = nf_nat_sdp(skb, dptr, datalen, rtp_exp, rtcp_exp);
 	else {
-		if (nf_ct_expect_related(exp) != 0)
-			ret = NF_DROP;
-		else
-			ret = NF_ACCEPT;
+		if (nf_ct_expect_related(rtp_exp) == 0) {
+			if (nf_ct_expect_related(rtcp_exp) != 0)
+				nf_ct_unexpect_related(rtp_exp);
+			else
+				ret = NF_ACCEPT;
+		}
 	}
-	nf_ct_expect_put(exp);
-
+	nf_ct_expect_put(rtcp_exp);
+err2:
+	nf_ct_expect_put(rtp_exp);
+err1:
 	return ret;
 }
 
@@ -758,7 +776,7 @@ static int process_sdp(struct sk_buff *skb,
 	if (port < 1024 || port > 65535)
 		return NF_DROP;
 
-	return set_expected_rtp(skb, dptr, datalen, &addr, htons(port));
+	return set_expected_rtp_rtcp(skb, dptr, datalen, &addr, htons(port));
 }
 static int process_invite_response(struct sk_buff *skb,
 				   const char **dptr, unsigned int *datalen,
@@ -1101,7 +1119,7 @@ static const struct nf_conntrack_expect_policy sip_exp_policy[SIP_EXPECT_MAX + 1
 		.timeout	= 3 * 60,
 	},
 	[SIP_EXPECT_AUDIO] = {
-		.max_expected	= IP_CT_DIR_MAX,
+		.max_expected	= 2 * IP_CT_DIR_MAX,
 		.timeout	= 3 * 60,
 	},
 };
-- 
cgit v1.2.3


From 4ab9e64e5e3c0516577818804aaf13a630d67bc9 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:26:08 -0700
Subject: [NETFILTER]: nf_nat_sip: split up SDP mangling

The SDP connection addresses may be contained in the payload multiple
times (in the session description and/or once per media description),
currently only the session description is properly updated. Split up
SDP mangling so the function setting up expectations only updates the
media port, update connection addresses from media descriptions while
parsing them and at the end update the session description when the
final addresses are known.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  25 ++++-
 net/ipv4/netfilter/nf_nat_sip.c            | 121 ++++++++++++++++--------
 net/netfilter/nf_conntrack_sip.c           | 142 +++++++++++++++++++++++------
 3 files changed, 219 insertions(+), 69 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 6ddf95f51fb5..eca3ad3f28dc 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -93,11 +93,26 @@ extern unsigned int (*nf_nat_sip_expect_hook)(struct sk_buff *skb,
 					      struct nf_conntrack_expect *exp,
 					      unsigned int matchoff,
 					      unsigned int matchlen);
-extern unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
-				       const char **dptr,
-				       unsigned int *datalen,
-				       struct nf_conntrack_expect *rtp_exp,
-				       struct nf_conntrack_expect *rtcp_exp);
+extern unsigned int (*nf_nat_sdp_addr_hook)(struct sk_buff *skb,
+					    const char **dptr,
+					    unsigned int dataoff,
+					    unsigned int *datalen,
+					    enum sdp_header_types type,
+					    enum sdp_header_types term,
+					    const union nf_inet_addr *addr);
+extern unsigned int (*nf_nat_sdp_session_hook)(struct sk_buff *skb,
+					       const char **dptr,
+					       unsigned int dataoff,
+					       unsigned int *datalen,
+					       const union nf_inet_addr *addr);
+extern unsigned int (*nf_nat_sdp_media_hook)(struct sk_buff *skb,
+					     const char **dptr,
+					     unsigned int *datalen,
+					     struct nf_conntrack_expect *rtp_exp,
+					     struct nf_conntrack_expect *rtcp_exp,
+					     unsigned int mediaoff,
+					     unsigned int medialen,
+					     union nf_inet_addr *rtp_addr);
 
 extern int ct_sip_parse_request(const struct nf_conn *ct,
 				const char *dptr, unsigned int datalen,
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index f73ab4883b75..4429069d9b42 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -316,45 +316,77 @@ static int mangle_content_len(struct sk_buff *skb,
 			     buffer, buflen);
 }
 
-static unsigned mangle_sdp_packet(struct sk_buff *skb,
-				  const char **dptr, unsigned int *datalen,
+static unsigned mangle_sdp_packet(struct sk_buff *skb, const char **dptr,
+				  unsigned int dataoff, unsigned int *datalen,
 				  enum sdp_header_types type,
+				  enum sdp_header_types term,
 				  char *buffer, int buflen)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	unsigned int matchlen, matchoff;
 
-	if (ct_sip_get_sdp_header(ct, *dptr, 0, *datalen, type, SDP_HDR_UNSPEC,
+	if (ct_sip_get_sdp_header(ct, *dptr, dataoff, *datalen, type, term,
 				  &matchoff, &matchlen) <= 0)
 		return 0;
 	return mangle_packet(skb, dptr, datalen, matchoff, matchlen,
 			     buffer, buflen);
 }
 
-static unsigned int mangle_sdp(struct sk_buff *skb,
-			       enum ip_conntrack_info ctinfo,
-			       struct nf_conn *ct,
-			       __be32 newip, u_int16_t port,
-			       const char **dptr, unsigned int *datalen)
+static unsigned int ip_nat_sdp_addr(struct sk_buff *skb, const char **dptr,
+				    unsigned int dataoff,
+				    unsigned int *datalen,
+				    enum sdp_header_types type,
+				    enum sdp_header_types term,
+				    const union nf_inet_addr *addr)
 {
 	char buffer[sizeof("nnn.nnn.nnn.nnn")];
-	unsigned int bufflen;
+	unsigned int buflen;
 
-	/* Mangle owner and contact info. */
-	bufflen = sprintf(buffer, "%u.%u.%u.%u", NIPQUAD(newip));
-	if (!mangle_sdp_packet(skb, dptr, datalen, SDP_HDR_OWNER_IP4,
-			       buffer, bufflen))
+	buflen = sprintf(buffer, NIPQUAD_FMT, NIPQUAD(addr->ip));
+	if (!mangle_sdp_packet(skb, dptr, dataoff, datalen, type, term,
+			       buffer, buflen))
 		return 0;
 
-	if (!mangle_sdp_packet(skb, dptr, datalen, SDP_HDR_CONNECTION_IP4,
-			       buffer, bufflen))
+	return mangle_content_len(skb, dptr, datalen);
+}
+
+static unsigned int ip_nat_sdp_port(struct sk_buff *skb,
+				    const char **dptr,
+				    unsigned int *datalen,
+				    unsigned int matchoff,
+				    unsigned int matchlen,
+				    u_int16_t port)
+{
+	char buffer[sizeof("nnnnn")];
+	unsigned int buflen;
+
+	buflen = sprintf(buffer, "%u", port);
+	if (!mangle_packet(skb, dptr, datalen, matchoff, matchlen,
+			   buffer, buflen))
 		return 0;
 
-	/* Mangle media port. */
-	bufflen = sprintf(buffer, "%u", port);
-	if (!mangle_sdp_packet(skb, dptr, datalen, SDP_HDR_MEDIA,
-			       buffer, bufflen))
+	return mangle_content_len(skb, dptr, datalen);
+}
+
+static unsigned int ip_nat_sdp_session(struct sk_buff *skb, const char **dptr,
+				       unsigned int dataoff,
+				       unsigned int *datalen,
+				       const union nf_inet_addr *addr)
+{
+	char buffer[sizeof("nnn.nnn.nnn.nnn")];
+	unsigned int buflen;
+
+	/* Mangle session description owner and contact addresses */
+	buflen = sprintf(buffer, "%u.%u.%u.%u", NIPQUAD(addr->ip));
+	if (!mangle_sdp_packet(skb, dptr, dataoff, datalen,
+			       SDP_HDR_OWNER_IP4, SDP_HDR_MEDIA,
+			       buffer, buflen))
+		return 0;
+
+	if (!mangle_sdp_packet(skb, dptr, dataoff, datalen,
+			       SDP_HDR_CONNECTION_IP4, SDP_HDR_MEDIA,
+			       buffer, buflen))
 		return 0;
 
 	return mangle_content_len(skb, dptr, datalen);
@@ -362,32 +394,35 @@ static unsigned int mangle_sdp(struct sk_buff *skb,
 
 /* So, this packet has hit the connection tracking matching code.
    Mangle it, and change the expectation to match the new version. */
-static unsigned int ip_nat_sdp(struct sk_buff *skb,
-			       const char **dptr, unsigned int *datalen,
-			       struct nf_conntrack_expect *rtp_exp,
-			       struct nf_conntrack_expect *rtcp_exp)
+static unsigned int ip_nat_sdp_media(struct sk_buff *skb,
+				     const char **dptr,
+				     unsigned int *datalen,
+				     struct nf_conntrack_expect *rtp_exp,
+				     struct nf_conntrack_expect *rtcp_exp,
+				     unsigned int mediaoff,
+				     unsigned int medialen,
+				     union nf_inet_addr *rtp_addr)
 {
 	enum ip_conntrack_info ctinfo;
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
-	__be32 newip;
 	u_int16_t port;
 
 	/* Connection will come from reply */
 	if (ct->tuplehash[dir].tuple.src.u3.ip ==
 	    ct->tuplehash[!dir].tuple.dst.u3.ip)
-		newip = rtp_exp->tuple.dst.u3.ip;
+		rtp_addr->ip = rtp_exp->tuple.dst.u3.ip;
 	else
-		newip = ct->tuplehash[!dir].tuple.dst.u3.ip;
+		rtp_addr->ip = ct->tuplehash[!dir].tuple.dst.u3.ip;
 
 	rtp_exp->saved_ip = rtp_exp->tuple.dst.u3.ip;
-	rtp_exp->tuple.dst.u3.ip = newip;
+	rtp_exp->tuple.dst.u3.ip = rtp_addr->ip;
 	rtp_exp->saved_proto.udp.port = rtp_exp->tuple.dst.u.udp.port;
 	rtp_exp->dir = !dir;
 	rtp_exp->expectfn = ip_nat_sip_expected;
 
 	rtcp_exp->saved_ip = rtcp_exp->tuple.dst.u3.ip;
-	rtcp_exp->tuple.dst.u3.ip = newip;
+	rtcp_exp->tuple.dst.u3.ip = rtp_addr->ip;
 	rtcp_exp->saved_proto.udp.port = rtcp_exp->tuple.dst.u.udp.port;
 	rtcp_exp->dir = !dir;
 	rtcp_exp->expectfn = ip_nat_sip_expected;
@@ -405,21 +440,29 @@ static unsigned int ip_nat_sdp(struct sk_buff *skb,
 	}
 
 	if (port == 0)
-		return NF_DROP;
+		goto err1;
+
+	/* Update media port. */
+	if (rtp_exp->tuple.dst.u.udp.port != rtp_exp->saved_proto.udp.port &&
+	    !ip_nat_sdp_port(skb, dptr, datalen, mediaoff, medialen, port))
+		goto err2;
 
-	if (!mangle_sdp(skb, ctinfo, ct, newip, port, dptr, datalen)) {
-		nf_ct_unexpect_related(rtp_exp);
-		nf_ct_unexpect_related(rtcp_exp);
-		return NF_DROP;
-	}
 	return NF_ACCEPT;
+
+err2:
+	nf_ct_unexpect_related(rtp_exp);
+	nf_ct_unexpect_related(rtcp_exp);
+err1:
+	return NF_DROP;
 }
 
 static void __exit nf_nat_sip_fini(void)
 {
 	rcu_assign_pointer(nf_nat_sip_hook, NULL);
 	rcu_assign_pointer(nf_nat_sip_expect_hook, NULL);
-	rcu_assign_pointer(nf_nat_sdp_hook, NULL);
+	rcu_assign_pointer(nf_nat_sdp_addr_hook, NULL);
+	rcu_assign_pointer(nf_nat_sdp_session_hook, NULL);
+	rcu_assign_pointer(nf_nat_sdp_media_hook, NULL);
 	synchronize_rcu();
 }
 
@@ -427,10 +470,14 @@ static int __init nf_nat_sip_init(void)
 {
 	BUG_ON(nf_nat_sip_hook != NULL);
 	BUG_ON(nf_nat_sip_expect_hook != NULL);
-	BUG_ON(nf_nat_sdp_hook != NULL);
+	BUG_ON(nf_nat_sdp_addr_hook != NULL);
+	BUG_ON(nf_nat_sdp_session_hook != NULL);
+	BUG_ON(nf_nat_sdp_media_hook != NULL);
 	rcu_assign_pointer(nf_nat_sip_hook, ip_nat_sip);
 	rcu_assign_pointer(nf_nat_sip_expect_hook, ip_nat_sip_expect);
-	rcu_assign_pointer(nf_nat_sdp_hook, ip_nat_sdp);
+	rcu_assign_pointer(nf_nat_sdp_addr_hook, ip_nat_sdp_addr);
+	rcu_assign_pointer(nf_nat_sdp_session_hook, ip_nat_sdp_session);
+	rcu_assign_pointer(nf_nat_sdp_media_hook, ip_nat_sdp_media);
 	return 0;
 }
 
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index 217262e23403..f929add324f3 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -60,13 +60,34 @@ unsigned int (*nf_nat_sip_expect_hook)(struct sk_buff *skb,
 				       unsigned int matchlen) __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sip_expect_hook);
 
-unsigned int (*nf_nat_sdp_hook)(struct sk_buff *skb,
-				const char **dptr,
-				unsigned int *datalen,
-				struct nf_conntrack_expect *rtp_exp,
-				struct nf_conntrack_expect *rtcp_exp)
-				__read_mostly;
-EXPORT_SYMBOL_GPL(nf_nat_sdp_hook);
+unsigned int (*nf_nat_sdp_addr_hook)(struct sk_buff *skb,
+				     const char **dptr,
+				     unsigned int dataoff,
+				     unsigned int *datalen,
+				     enum sdp_header_types type,
+				     enum sdp_header_types term,
+				     const union nf_inet_addr *addr)
+				     __read_mostly;
+EXPORT_SYMBOL_GPL(nf_nat_sdp_addr_hook);
+
+unsigned int (*nf_nat_sdp_session_hook)(struct sk_buff *skb,
+					const char **dptr,
+					unsigned int dataoff,
+					unsigned int *datalen,
+					const union nf_inet_addr *addr)
+					__read_mostly;
+EXPORT_SYMBOL_GPL(nf_nat_sdp_session_hook);
+
+unsigned int (*nf_nat_sdp_media_hook)(struct sk_buff *skb,
+				      const char **dptr,
+				      unsigned int *datalen,
+				      struct nf_conntrack_expect *rtp_exp,
+				      struct nf_conntrack_expect *rtcp_exp,
+				      unsigned int mediaoff,
+				      unsigned int medialen,
+				      union nf_inet_addr *rtp_addr)
+				      __read_mostly;
+EXPORT_SYMBOL_GPL(nf_nat_sdp_media_hook);
 
 static int string_len(const struct nf_conn *ct, const char *dptr,
 		      const char *limit, int *shift)
@@ -613,6 +634,26 @@ int ct_sip_get_sdp_header(const struct nf_conn *ct, const char *dptr,
 }
 EXPORT_SYMBOL_GPL(ct_sip_get_sdp_header);
 
+static int ct_sip_parse_sdp_addr(const struct nf_conn *ct, const char *dptr,
+				 unsigned int dataoff, unsigned int datalen,
+				 enum sdp_header_types type,
+				 enum sdp_header_types term,
+				 unsigned int *matchoff, unsigned int *matchlen,
+				 union nf_inet_addr *addr)
+{
+	int ret;
+
+	ret = ct_sip_get_sdp_header(ct, dptr, dataoff, datalen, type, term,
+				    matchoff, matchlen);
+	if (ret <= 0)
+		return ret;
+
+	if (!parse_addr(ct, dptr + *matchoff, NULL, addr,
+			dptr + *matchoff + *matchlen))
+		return -1;
+	return 1;
+}
+
 static int refresh_signalling_expectation(struct nf_conn *ct,
 					  union nf_inet_addr *addr,
 					  __be16 port,
@@ -663,7 +704,8 @@ static void flush_expectations(struct nf_conn *ct, bool media)
 
 static int set_expected_rtp_rtcp(struct sk_buff *skb,
 				 const char **dptr, unsigned int *datalen,
-				 union nf_inet_addr *daddr, __be16 port)
+				 union nf_inet_addr *daddr, __be16 port,
+				 unsigned int mediaoff, unsigned int medialen)
 {
 	struct nf_conntrack_expect *exp, *rtp_exp, *rtcp_exp;
 	enum ip_conntrack_info ctinfo;
@@ -675,7 +717,7 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 	int skip_expect = 0, ret = NF_DROP;
 	u_int16_t base_port;
 	__be16 rtp_port, rtcp_port;
-	typeof(nf_nat_sdp_hook) nf_nat_sdp;
+	typeof(nf_nat_sdp_media_hook) nf_nat_sdp_media;
 
 	saddr = NULL;
 	if (sip_direct_media) {
@@ -724,9 +766,10 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 	nf_ct_expect_init(rtcp_exp, SIP_EXPECT_AUDIO, family, saddr, daddr,
 			  IPPROTO_UDP, NULL, &rtcp_port);
 
-	nf_nat_sdp = rcu_dereference(nf_nat_sdp_hook);
-	if (nf_nat_sdp && ct->status & IPS_NAT_MASK)
-		ret = nf_nat_sdp(skb, dptr, datalen, rtp_exp, rtcp_exp);
+	nf_nat_sdp_media = rcu_dereference(nf_nat_sdp_media_hook);
+	if (nf_nat_sdp_media && ct->status & IPS_NAT_MASK)
+		ret = nf_nat_sdp_media(skb, dptr, datalen, rtp_exp, rtcp_exp,
+				       mediaoff, medialen, daddr);
 	else {
 		if (nf_ct_expect_related(rtp_exp) == 0) {
 			if (nf_ct_expect_related(rtcp_exp) != 0)
@@ -750,33 +793,78 @@ static int process_sdp(struct sk_buff *skb,
 	struct nf_conn *ct = nf_ct_get(skb, &ctinfo);
 	int family = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num;
 	unsigned int matchoff, matchlen;
-	union nf_inet_addr addr;
+	unsigned int mediaoff, medialen;
+	unsigned int sdpoff;
+	unsigned int caddr_len, maddr_len;
+	union nf_inet_addr caddr, maddr, rtp_addr;
 	unsigned int port;
-	enum sdp_header_types type;
+	enum sdp_header_types c_hdr;
+	int ret;
+	typeof(nf_nat_sdp_addr_hook) nf_nat_sdp_addr;
+	typeof(nf_nat_sdp_session_hook) nf_nat_sdp_session;
 
-	/* Get address and port from SDP packet. */
-	type = family == AF_INET ? SDP_HDR_CONNECTION_IP4 :
-				   SDP_HDR_CONNECTION_IP6;
+	c_hdr = family == AF_INET ? SDP_HDR_CONNECTION_IP4 :
+				    SDP_HDR_CONNECTION_IP6;
 
+	/* Find beginning of session description */
 	if (ct_sip_get_sdp_header(ct, *dptr, 0, *datalen,
-				  type, SDP_HDR_UNSPEC,
+				  SDP_HDR_VERSION, SDP_HDR_UNSPEC,
 				  &matchoff, &matchlen) <= 0)
 		return NF_ACCEPT;
-
-	/* We'll drop only if there are parse problems. */
-	if (!parse_addr(ct, *dptr + matchoff, NULL, &addr, *dptr + *datalen))
-		return NF_DROP;
-
-	if (ct_sip_get_sdp_header(ct, *dptr, 0, *datalen,
+	sdpoff = matchoff;
+
+	/* The connection information is contained in the session description
+	 * and/or once per media description. The first media description marks
+	 * the end of the session description. */
+	caddr_len = 0;
+	if (ct_sip_parse_sdp_addr(ct, *dptr, sdpoff, *datalen,
+				  c_hdr, SDP_HDR_MEDIA,
+				  &matchoff, &matchlen, &caddr) > 0)
+		caddr_len = matchlen;
+
+	if (ct_sip_get_sdp_header(ct, *dptr, sdpoff, *datalen,
 				  SDP_HDR_MEDIA, SDP_HDR_UNSPEC,
-				  &matchoff, &matchlen) <= 0)
+				  &mediaoff, &medialen) <= 0)
 		return NF_ACCEPT;
 
-	port = simple_strtoul(*dptr + matchoff, NULL, 10);
+	port = simple_strtoul(*dptr + mediaoff, NULL, 10);
 	if (port < 1024 || port > 65535)
 		return NF_DROP;
 
-	return set_expected_rtp_rtcp(skb, dptr, datalen, &addr, htons(port));
+	/* The media description overrides the session description. */
+	maddr_len = 0;
+	if (ct_sip_parse_sdp_addr(ct, *dptr, mediaoff, *datalen,
+				  c_hdr, SDP_HDR_MEDIA,
+				  &matchoff, &matchlen, &maddr) > 0) {
+		maddr_len = matchlen;
+		memcpy(&rtp_addr, &maddr, sizeof(rtp_addr));
+	} else if (caddr_len)
+		memcpy(&rtp_addr, &caddr, sizeof(rtp_addr));
+	else
+		return NF_DROP;
+
+	ret = set_expected_rtp_rtcp(skb, dptr, datalen, &rtp_addr, htons(port),
+				    mediaoff, medialen);
+	if (ret != NF_ACCEPT)
+		return ret;
+
+	/* Update media connection address if present */
+	if (maddr_len) {
+		nf_nat_sdp_addr = rcu_dereference(nf_nat_sdp_addr_hook);
+		if (nf_nat_sdp_addr && ct->status & IPS_NAT_MASK) {
+			ret = nf_nat_sdp_addr(skb, dptr, mediaoff, datalen,
+					      c_hdr, SDP_HDR_MEDIA, &rtp_addr);
+			if (ret != NF_ACCEPT)
+				return ret;
+		}
+	}
+
+	/* Update session connection and owner addresses */
+	nf_nat_sdp_session = rcu_dereference(nf_nat_sdp_session_hook);
+	if (nf_nat_sdp_session && ct->status & IPS_NAT_MASK)
+		ret = nf_nat_sdp_session(skb, dptr, sdpoff, datalen, &rtp_addr);
+
+	return ret;
 }
 static int process_invite_response(struct sk_buff *skb,
 				   const char **dptr, unsigned int *datalen,
-- 
cgit v1.2.3


From 0d0ab0378d67517a4f4ae3497706c13d9dd24af1 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:26:24 -0700
Subject: [NETFILTER]: nf_conntrack_sip: support multiple media channels

Add support for multiple media channels and use it to create
expectations for video streams when present.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  14 ++++
 include/net/netfilter/nf_conntrack.h       |   2 +-
 net/netfilter/nf_conntrack_sip.c           | 121 +++++++++++++++++++++--------
 3 files changed, 105 insertions(+), 32 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index eca3ad3f28dc..71fa3eb5f485 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -12,10 +12,24 @@ struct nf_ct_sip_master {
 enum sip_expectation_classes {
 	SIP_EXPECT_SIGNALLING,
 	SIP_EXPECT_AUDIO,
+	SIP_EXPECT_VIDEO,
 	__SIP_EXPECT_MAX
 };
 #define SIP_EXPECT_MAX	(__SIP_EXPECT_MAX - 1)
 
+struct sdp_media_type {
+	const char			*name;
+	unsigned int			len;
+	enum sip_expectation_classes	class;
+};
+
+#define SDP_MEDIA_TYPE(__name, __class)					\
+{									\
+	.name	= (__name),						\
+	.len	= sizeof(__name) - 1,					\
+	.class	= (__class),						\
+}
+
 struct sip_handler {
 	const char	*method;
 	unsigned int	len;
diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
index 4a4f870d2a5e..a3567a7a6d67 100644
--- a/include/net/netfilter/nf_conntrack.h
+++ b/include/net/netfilter/nf_conntrack.h
@@ -78,7 +78,7 @@ do {									\
 struct nf_conntrack_helper;
 
 /* Must be kept in sync with the classes defined by helpers */
-#define NF_CT_MAX_EXPECT_CLASSES	2
+#define NF_CT_MAX_EXPECT_CLASSES	3
 
 /* nf_conn feature for connections that have a helper */
 struct nf_conn_help {
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index f929add324f3..f40a525732d1 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -112,6 +112,21 @@ static int digits_len(const struct nf_conn *ct, const char *dptr,
 	return len;
 }
 
+/* get media type + port length */
+static int media_len(const struct nf_conn *ct, const char *dptr,
+		     const char *limit, int *shift)
+{
+	int len = string_len(ct, dptr, limit, shift);
+
+	dptr += len;
+	if (dptr >= limit || *dptr != ' ')
+		return 0;
+	len++;
+	dptr++;
+
+	return len + digits_len(ct, dptr, limit, shift);
+}
+
 static int parse_addr(const struct nf_conn *ct, const char *cp,
                       const char **endp, union nf_inet_addr *addr,
                       const char *limit)
@@ -563,7 +578,7 @@ static const struct sip_header ct_sdp_hdrs[] = {
 	[SDP_HDR_CONNECTION_IP4]	= SDP_HDR("c=", "IN IP4 ", epaddr_len),
 	[SDP_HDR_OWNER_IP6]		= SDP_HDR("o=", "IN IP6 ", epaddr_len),
 	[SDP_HDR_CONNECTION_IP6]	= SDP_HDR("c=", "IN IP6 ", epaddr_len),
-	[SDP_HDR_MEDIA]			= SDP_HDR("m=", "audio ", digits_len),
+	[SDP_HDR_MEDIA]			= SDP_HDR("m=", NULL, media_len),
 };
 
 /* Linear string search within SDP header values */
@@ -705,6 +720,7 @@ static void flush_expectations(struct nf_conn *ct, bool media)
 static int set_expected_rtp_rtcp(struct sk_buff *skb,
 				 const char **dptr, unsigned int *datalen,
 				 union nf_inet_addr *daddr, __be16 port,
+				 enum sip_expectation_classes class,
 				 unsigned int mediaoff, unsigned int medialen)
 {
 	struct nf_conntrack_expect *exp, *rtp_exp, *rtcp_exp;
@@ -743,7 +759,7 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 	exp = __nf_ct_expect_find(&tuple);
 	if (exp && exp->master != ct &&
 	    nfct_help(exp->master)->helper == nfct_help(ct)->helper &&
-	    exp->class == SIP_EXPECT_AUDIO)
+	    exp->class == class)
 		skip_expect = 1;
 	rcu_read_unlock();
 
@@ -757,13 +773,13 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 	rtp_exp = nf_ct_expect_alloc(ct);
 	if (rtp_exp == NULL)
 		goto err1;
-	nf_ct_expect_init(rtp_exp, SIP_EXPECT_AUDIO, family, saddr, daddr,
+	nf_ct_expect_init(rtp_exp, class, family, saddr, daddr,
 			  IPPROTO_UDP, NULL, &rtp_port);
 
 	rtcp_exp = nf_ct_expect_alloc(ct);
 	if (rtcp_exp == NULL)
 		goto err2;
-	nf_ct_expect_init(rtcp_exp, SIP_EXPECT_AUDIO, family, saddr, daddr,
+	nf_ct_expect_init(rtcp_exp, class, family, saddr, daddr,
 			  IPPROTO_UDP, NULL, &rtcp_port);
 
 	nf_nat_sdp_media = rcu_dereference(nf_nat_sdp_media_hook);
@@ -785,6 +801,28 @@ err1:
 	return ret;
 }
 
+static const struct sdp_media_type sdp_media_types[] = {
+	SDP_MEDIA_TYPE("audio ", SIP_EXPECT_AUDIO),
+	SDP_MEDIA_TYPE("video ", SIP_EXPECT_VIDEO),
+};
+
+static const struct sdp_media_type *sdp_media_type(const char *dptr,
+						   unsigned int matchoff,
+						   unsigned int matchlen)
+{
+	const struct sdp_media_type *t;
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(sdp_media_types); i++) {
+		t = &sdp_media_types[i];
+		if (matchlen < t->len ||
+		    strncmp(dptr + matchoff, t->name, t->len))
+			continue;
+		return t;
+	}
+	return NULL;
+}
+
 static int process_sdp(struct sk_buff *skb,
 		       const char **dptr, unsigned int *datalen,
 		       unsigned int cseq)
@@ -796,13 +834,16 @@ static int process_sdp(struct sk_buff *skb,
 	unsigned int mediaoff, medialen;
 	unsigned int sdpoff;
 	unsigned int caddr_len, maddr_len;
+	unsigned int i;
 	union nf_inet_addr caddr, maddr, rtp_addr;
 	unsigned int port;
 	enum sdp_header_types c_hdr;
-	int ret;
+	const struct sdp_media_type *t;
+	int ret = NF_ACCEPT;
 	typeof(nf_nat_sdp_addr_hook) nf_nat_sdp_addr;
 	typeof(nf_nat_sdp_session_hook) nf_nat_sdp_session;
 
+	nf_nat_sdp_addr = rcu_dereference(nf_nat_sdp_addr_hook);
 	c_hdr = family == AF_INET ? SDP_HDR_CONNECTION_IP4 :
 				    SDP_HDR_CONNECTION_IP6;
 
@@ -822,41 +863,55 @@ static int process_sdp(struct sk_buff *skb,
 				  &matchoff, &matchlen, &caddr) > 0)
 		caddr_len = matchlen;
 
-	if (ct_sip_get_sdp_header(ct, *dptr, sdpoff, *datalen,
-				  SDP_HDR_MEDIA, SDP_HDR_UNSPEC,
-				  &mediaoff, &medialen) <= 0)
-		return NF_ACCEPT;
+	mediaoff = sdpoff;
+	for (i = 0; i < ARRAY_SIZE(sdp_media_types); ) {
+		if (ct_sip_get_sdp_header(ct, *dptr, mediaoff, *datalen,
+					  SDP_HDR_MEDIA, SDP_HDR_UNSPEC,
+					  &mediaoff, &medialen) <= 0)
+			break;
 
-	port = simple_strtoul(*dptr + mediaoff, NULL, 10);
-	if (port < 1024 || port > 65535)
-		return NF_DROP;
+		/* Get media type and port number. A media port value of zero
+		 * indicates an inactive stream. */
+		t = sdp_media_type(*dptr, mediaoff, medialen);
+		if (!t) {
+			mediaoff += medialen;
+			continue;
+		}
+		mediaoff += t->len;
+		medialen -= t->len;
 
-	/* The media description overrides the session description. */
-	maddr_len = 0;
-	if (ct_sip_parse_sdp_addr(ct, *dptr, mediaoff, *datalen,
-				  c_hdr, SDP_HDR_MEDIA,
-				  &matchoff, &matchlen, &maddr) > 0) {
-		maddr_len = matchlen;
-		memcpy(&rtp_addr, &maddr, sizeof(rtp_addr));
-	} else if (caddr_len)
-		memcpy(&rtp_addr, &caddr, sizeof(rtp_addr));
-	else
-		return NF_DROP;
+		port = simple_strtoul(*dptr + mediaoff, NULL, 10);
+		if (port == 0)
+			continue;
+		if (port < 1024 || port > 65535)
+			return NF_DROP;
 
-	ret = set_expected_rtp_rtcp(skb, dptr, datalen, &rtp_addr, htons(port),
-				    mediaoff, medialen);
-	if (ret != NF_ACCEPT)
-		return ret;
+		/* The media description overrides the session description. */
+		maddr_len = 0;
+		if (ct_sip_parse_sdp_addr(ct, *dptr, mediaoff, *datalen,
+					  c_hdr, SDP_HDR_MEDIA,
+					  &matchoff, &matchlen, &maddr) > 0) {
+			maddr_len = matchlen;
+			memcpy(&rtp_addr, &maddr, sizeof(rtp_addr));
+		} else if (caddr_len)
+			memcpy(&rtp_addr, &caddr, sizeof(rtp_addr));
+		else
+			return NF_DROP;
+
+		ret = set_expected_rtp_rtcp(skb, dptr, datalen,
+					    &rtp_addr, htons(port), t->class,
+					    mediaoff, medialen);
+		if (ret != NF_ACCEPT)
+			return ret;
 
-	/* Update media connection address if present */
-	if (maddr_len) {
-		nf_nat_sdp_addr = rcu_dereference(nf_nat_sdp_addr_hook);
-		if (nf_nat_sdp_addr && ct->status & IPS_NAT_MASK) {
+		/* Update media connection address if present */
+		if (maddr_len && nf_nat_sdp_addr && ct->status & IPS_NAT_MASK) {
 			ret = nf_nat_sdp_addr(skb, dptr, mediaoff, datalen,
 					      c_hdr, SDP_HDR_MEDIA, &rtp_addr);
 			if (ret != NF_ACCEPT)
 				return ret;
 		}
+		i++;
 	}
 
 	/* Update session connection and owner addresses */
@@ -1210,6 +1265,10 @@ static const struct nf_conntrack_expect_policy sip_exp_policy[SIP_EXPECT_MAX + 1
 		.max_expected	= 2 * IP_CT_DIR_MAX,
 		.timeout	= 3 * 60,
 	},
+	[SIP_EXPECT_VIDEO] = {
+		.max_expected	= 2 * IP_CT_DIR_MAX,
+		.timeout	= 3 * 60,
+	},
 };
 
 static void nf_conntrack_sip_fini(void)
-- 
cgit v1.2.3


From c7f485abd618e0d249bdd1abdc586bd10fee1954 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 25 Mar 2008 20:26:43 -0700
Subject: [NETFILTER]: nf_conntrack_sip: RTP routing optimization

Optimize call routing between NATed endpoints: when an external
registrar sends a media description that contains an existing RTP
expectation from a different SNATed connection, the gatekeeper
is trying to route the call directly between the two endpoints.

We assume both endpoints can reach each other directly and
"un-NAT" the addresses, which makes the media stream go between
the two endpoints directly.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netfilter/nf_conntrack_sip.h |  6 +++
 net/ipv4/netfilter/nf_nat_sip.c            |  3 ++
 net/netfilter/nf_conntrack_sip.c           | 59 +++++++++++++++++++++++++-----
 3 files changed, 58 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h
index 71fa3eb5f485..5da04e586a3f 100644
--- a/include/linux/netfilter/nf_conntrack_sip.h
+++ b/include/linux/netfilter/nf_conntrack_sip.h
@@ -114,6 +114,12 @@ extern unsigned int (*nf_nat_sdp_addr_hook)(struct sk_buff *skb,
 					    enum sdp_header_types type,
 					    enum sdp_header_types term,
 					    const union nf_inet_addr *addr);
+extern unsigned int (*nf_nat_sdp_port_hook)(struct sk_buff *skb,
+					    const char **dptr,
+					    unsigned int *datalen,
+					    unsigned int matchoff,
+					    unsigned int matchlen,
+					    u_int16_t port);
 extern unsigned int (*nf_nat_sdp_session_hook)(struct sk_buff *skb,
 					       const char **dptr,
 					       unsigned int dataoff,
diff --git a/net/ipv4/netfilter/nf_nat_sip.c b/net/ipv4/netfilter/nf_nat_sip.c
index 4429069d9b42..bcddccddf768 100644
--- a/net/ipv4/netfilter/nf_nat_sip.c
+++ b/net/ipv4/netfilter/nf_nat_sip.c
@@ -461,6 +461,7 @@ static void __exit nf_nat_sip_fini(void)
 	rcu_assign_pointer(nf_nat_sip_hook, NULL);
 	rcu_assign_pointer(nf_nat_sip_expect_hook, NULL);
 	rcu_assign_pointer(nf_nat_sdp_addr_hook, NULL);
+	rcu_assign_pointer(nf_nat_sdp_port_hook, NULL);
 	rcu_assign_pointer(nf_nat_sdp_session_hook, NULL);
 	rcu_assign_pointer(nf_nat_sdp_media_hook, NULL);
 	synchronize_rcu();
@@ -471,11 +472,13 @@ static int __init nf_nat_sip_init(void)
 	BUG_ON(nf_nat_sip_hook != NULL);
 	BUG_ON(nf_nat_sip_expect_hook != NULL);
 	BUG_ON(nf_nat_sdp_addr_hook != NULL);
+	BUG_ON(nf_nat_sdp_port_hook != NULL);
 	BUG_ON(nf_nat_sdp_session_hook != NULL);
 	BUG_ON(nf_nat_sdp_media_hook != NULL);
 	rcu_assign_pointer(nf_nat_sip_hook, ip_nat_sip);
 	rcu_assign_pointer(nf_nat_sip_expect_hook, ip_nat_sip_expect);
 	rcu_assign_pointer(nf_nat_sdp_addr_hook, ip_nat_sdp_addr);
+	rcu_assign_pointer(nf_nat_sdp_port_hook, ip_nat_sdp_port);
 	rcu_assign_pointer(nf_nat_sdp_session_hook, ip_nat_sdp_session);
 	rcu_assign_pointer(nf_nat_sdp_media_hook, ip_nat_sdp_media);
 	return 0;
diff --git a/net/netfilter/nf_conntrack_sip.c b/net/netfilter/nf_conntrack_sip.c
index f40a525732d1..57de22c770a3 100644
--- a/net/netfilter/nf_conntrack_sip.c
+++ b/net/netfilter/nf_conntrack_sip.c
@@ -70,6 +70,14 @@ unsigned int (*nf_nat_sdp_addr_hook)(struct sk_buff *skb,
 				     __read_mostly;
 EXPORT_SYMBOL_GPL(nf_nat_sdp_addr_hook);
 
+unsigned int (*nf_nat_sdp_port_hook)(struct sk_buff *skb,
+				     const char **dptr,
+				     unsigned int *datalen,
+				     unsigned int matchoff,
+				     unsigned int matchlen,
+				     u_int16_t port) __read_mostly;
+EXPORT_SYMBOL_GPL(nf_nat_sdp_port_hook);
+
 unsigned int (*nf_nat_sdp_session_hook)(struct sk_buff *skb,
 					const char **dptr,
 					unsigned int dataoff,
@@ -730,9 +738,10 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 	union nf_inet_addr *saddr;
 	struct nf_conntrack_tuple tuple;
 	int family = ct->tuplehash[!dir].tuple.src.l3num;
-	int skip_expect = 0, ret = NF_DROP;
+	int direct_rtp = 0, skip_expect = 0, ret = NF_DROP;
 	u_int16_t base_port;
 	__be16 rtp_port, rtcp_port;
+	typeof(nf_nat_sdp_port_hook) nf_nat_sdp_port;
 	typeof(nf_nat_sdp_media_hook) nf_nat_sdp_media;
 
 	saddr = NULL;
@@ -746,6 +755,14 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 	 * to register it since we can see the same media description multiple
 	 * times on different connections in case multiple endpoints receive
 	 * the same call.
+	 *
+	 * RTP optimization: if we find a matching media channel expectation
+	 * and both the expectation and this connection are SNATed, we assume
+	 * both sides can reach each other directly and use the final
+	 * destination address from the expectation. We still need to keep
+	 * the NATed expectations for media that might arrive from the
+	 * outside, and additionally need to expect the direct RTP stream
+	 * in case it passes through us even without NAT.
 	 */
 	memset(&tuple, 0, sizeof(tuple));
 	if (saddr)
@@ -756,20 +773,42 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 	tuple.dst.u.udp.port	= port;
 
 	rcu_read_lock();
-	exp = __nf_ct_expect_find(&tuple);
-	if (exp && exp->master != ct &&
-	    nfct_help(exp->master)->helper == nfct_help(ct)->helper &&
-	    exp->class == class)
-		skip_expect = 1;
-	rcu_read_unlock();
+	do {
+		exp = __nf_ct_expect_find(&tuple);
 
-	if (skip_expect)
-		return NF_ACCEPT;
+		if (!exp || exp->master == ct ||
+		    nfct_help(exp->master)->helper != nfct_help(ct)->helper ||
+		    exp->class != class)
+			break;
+
+		if (exp->tuple.src.l3num == AF_INET && !direct_rtp &&
+		    (exp->saved_ip != exp->tuple.dst.u3.ip ||
+		     exp->saved_proto.udp.port != exp->tuple.dst.u.udp.port) &&
+		    ct->status & IPS_NAT_MASK) {
+			daddr->ip		= exp->saved_ip;
+			tuple.dst.u3.ip		= exp->saved_ip;
+			tuple.dst.u.udp.port	= exp->saved_proto.udp.port;
+			direct_rtp = 1;
+		} else
+			skip_expect = 1;
+	} while (!skip_expect);
+	rcu_read_unlock();
 
 	base_port = ntohs(tuple.dst.u.udp.port) & ~1;
 	rtp_port = htons(base_port);
 	rtcp_port = htons(base_port + 1);
 
+	if (direct_rtp) {
+		nf_nat_sdp_port = rcu_dereference(nf_nat_sdp_port_hook);
+		if (nf_nat_sdp_port &&
+		    !nf_nat_sdp_port(skb, dptr, datalen,
+				     mediaoff, medialen, ntohs(rtp_port)))
+			goto err1;
+	}
+
+	if (skip_expect)
+		return NF_ACCEPT;
+
 	rtp_exp = nf_ct_expect_alloc(ct);
 	if (rtp_exp == NULL)
 		goto err1;
@@ -783,7 +822,7 @@ static int set_expected_rtp_rtcp(struct sk_buff *skb,
 			  IPPROTO_UDP, NULL, &rtcp_port);
 
 	nf_nat_sdp_media = rcu_dereference(nf_nat_sdp_media_hook);
-	if (nf_nat_sdp_media && ct->status & IPS_NAT_MASK)
+	if (nf_nat_sdp_media && ct->status & IPS_NAT_MASK && !direct_rtp)
 		ret = nf_nat_sdp_media(skb, dptr, datalen, rtp_exp, rtcp_exp,
 				       mediaoff, medialen, daddr);
 	else {
-- 
cgit v1.2.3


From 8b78cf602fd3bd97c0080edd22fe8fd5d0fa7832 Mon Sep 17 00:00:00 2001
From: Yi Yang <yi.y.yang@intel.com>
Date: Mon, 25 Feb 2008 08:46:12 +0800
Subject: cpuidle: fix cpuidle time and usage overflow

cpuidle C-state sysfs node time and usage are very easy to overflow because
they are all of unsigned int type, time will overflow within about two hours,
usage will take longer time to overflow, but they are increasing for ever.

This patch will convert them to unsigned long long.

Signed-off-by: Yi Yang <yi.y.yang@intel.com>
Acked-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
---
 drivers/cpuidle/cpuidle.c |  2 +-
 drivers/cpuidle/sysfs.c   | 10 ++++++++--
 include/linux/cpuidle.h   |  4 ++--
 3 files changed, 11 insertions(+), 5 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c
index d73663a52324..d42deb310ac7 100644
--- a/drivers/cpuidle/cpuidle.c
+++ b/drivers/cpuidle/cpuidle.c
@@ -67,7 +67,7 @@ static void cpuidle_idle_call(void)
 	/* enter the state and update stats */
 	dev->last_residency = target_state->enter(dev, target_state);
 	dev->last_state = target_state;
-	target_state->time += dev->last_residency;
+	target_state->time += (unsigned long long)dev->last_residency;
 	target_state->usage++;
 
 	/* give the governor an opportunity to reflect on the outcome */
diff --git a/drivers/cpuidle/sysfs.c b/drivers/cpuidle/sysfs.c
index 69102ca05685..e949618b9be0 100644
--- a/drivers/cpuidle/sysfs.c
+++ b/drivers/cpuidle/sysfs.c
@@ -218,6 +218,12 @@ static ssize_t show_state_##_name(struct cpuidle_state *state, char *buf) \
 	return sprintf(buf, "%u\n", state->_name);\
 }
 
+#define define_show_state_ull_function(_name) \
+static ssize_t show_state_##_name(struct cpuidle_state *state, char *buf) \
+{ \
+	return sprintf(buf, "%llu\n", state->_name);\
+}
+
 #define define_show_state_str_function(_name) \
 static ssize_t show_state_##_name(struct cpuidle_state *state, char *buf) \
 { \
@@ -228,8 +234,8 @@ static ssize_t show_state_##_name(struct cpuidle_state *state, char *buf) \
 
 define_show_state_function(exit_latency)
 define_show_state_function(power_usage)
-define_show_state_function(usage)
-define_show_state_function(time)
+define_show_state_ull_function(usage)
+define_show_state_ull_function(time)
 define_show_state_str_function(name)
 define_show_state_str_function(desc)
 
diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h
index 6b72a4584086..51e6b1e520e6 100644
--- a/include/linux/cpuidle.h
+++ b/include/linux/cpuidle.h
@@ -38,8 +38,8 @@ struct cpuidle_state {
 	unsigned int	power_usage; /* in mW */
 	unsigned int	target_residency; /* in US */
 
-	unsigned int	usage;
-	unsigned int	time; /* in US */
+	unsigned long long	usage;
+	unsigned long long	time; /* in US */
 
 	int (*enter)	(struct cpuidle_device *dev,
 			 struct cpuidle_state *state);
-- 
cgit v1.2.3


From 06d8308c61e54346585b2691c13ee3f90cb6fb2f Mon Sep 17 00:00:00 2001
From: Thomas Gleixner <tglx@linutronix.de>
Date: Sat, 22 Mar 2008 09:20:24 +0100
Subject: NOHZ: reevaluate idle sleep length after add_timer_on()

add_timer_on() can add a timer on a CPU which is currently in a long
idle sleep, but the timer wheel is not reevaluated by the nohz code on
that CPU. So a timer can be delayed for quite a long time. This
triggered a false positive in the clocksource watchdog code.

To avoid this we need to wake up the idle CPU and enforce the
reevaluation of the timer wheel for the next timer event.

Add a function, which checks a given CPU for idle state, marks the
idle task with NEED_RESCHED and sends a reschedule IPI to notify the
other CPU of the change in the timer wheel.

Call this function from add_timer_on().

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: stable@kernel.org

--
 include/linux/sched.h |    6 ++++++
 kernel/sched.c        |   43 +++++++++++++++++++++++++++++++++++++++++++
 kernel/timer.c        |   10 +++++++++-
 3 files changed, 58 insertions(+), 1 deletion(-)
---
 include/linux/sched.h |  6 ++++++
 kernel/sched.c        | 43 +++++++++++++++++++++++++++++++++++++++++++
 kernel/timer.c        | 10 +++++++++-
 3 files changed, 58 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index fed07d03364e..6a1e7afb099b 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1541,6 +1541,12 @@ static inline void idle_task_exit(void) {}
 
 extern void sched_idle_next(void);
 
+#if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP)
+extern void wake_up_idle_cpu(int cpu);
+#else
+static inline void wake_up_idle_cpu(int cpu) { }
+#endif
+
 #ifdef CONFIG_SCHED_DEBUG
 extern unsigned int sysctl_sched_latency;
 extern unsigned int sysctl_sched_min_granularity;
diff --git a/kernel/sched.c b/kernel/sched.c
index 28c73f07efb2..8dcdec6fe0fe 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -1052,6 +1052,49 @@ static void resched_cpu(int cpu)
 	resched_task(cpu_curr(cpu));
 	spin_unlock_irqrestore(&rq->lock, flags);
 }
+
+#ifdef CONFIG_NO_HZ
+/*
+ * When add_timer_on() enqueues a timer into the timer wheel of an
+ * idle CPU then this timer might expire before the next timer event
+ * which is scheduled to wake up that CPU. In case of a completely
+ * idle system the next event might even be infinite time into the
+ * future. wake_up_idle_cpu() ensures that the CPU is woken up and
+ * leaves the inner idle loop so the newly added timer is taken into
+ * account when the CPU goes back to idle and evaluates the timer
+ * wheel for the next timer event.
+ */
+void wake_up_idle_cpu(int cpu)
+{
+	struct rq *rq = cpu_rq(cpu);
+
+	if (cpu == smp_processor_id())
+		return;
+
+	/*
+	 * This is safe, as this function is called with the timer
+	 * wheel base lock of (cpu) held. When the CPU is on the way
+	 * to idle and has not yet set rq->curr to idle then it will
+	 * be serialized on the timer wheel base lock and take the new
+	 * timer into account automatically.
+	 */
+	if (rq->curr != rq->idle)
+		return;
+
+	/*
+	 * We can set TIF_RESCHED on the idle task of the other CPU
+	 * lockless. The worst case is that the other CPU runs the
+	 * idle task through an additional NOOP schedule()
+	 */
+	set_tsk_thread_flag(rq->idle, TIF_NEED_RESCHED);
+
+	/* NEED_RESCHED must be visible before we test polling */
+	smp_mb();
+	if (!tsk_is_polling(rq->idle))
+		smp_send_reschedule(cpu);
+}
+#endif
+
 #else
 static void __resched_task(struct task_struct *p, int tif_bit)
 {
diff --git a/kernel/timer.c b/kernel/timer.c
index 99b00a25f88b..b024106daa70 100644
--- a/kernel/timer.c
+++ b/kernel/timer.c
@@ -451,10 +451,18 @@ void add_timer_on(struct timer_list *timer, int cpu)
 	spin_lock_irqsave(&base->lock, flags);
 	timer_set_base(timer, base);
 	internal_add_timer(base, timer);
+	/*
+	 * Check whether the other CPU is idle and needs to be
+	 * triggered to reevaluate the timer wheel when nohz is
+	 * active. We are protected against the other CPU fiddling
+	 * with the timer by holding the timer base lock. This also
+	 * makes sure that a CPU on the way to idle can not evaluate
+	 * the timer wheel.
+	 */
+	wake_up_idle_cpu(cpu);
 	spin_unlock_irqrestore(&base->lock, flags);
 }
 
-
 /**
  * mod_timer - modify a timer's timeout
  * @timer: the timer to be modified
-- 
cgit v1.2.3


From 9c2f5746b9cd536f0007709196d85a7e7d0070fa Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Wed, 26 Mar 2008 00:47:14 -0700
Subject: [NETNS]: Compilation fix for include/linux/netdevice.h.

Commit commit c346dca10840a874240c78efe3f39acf4312a1f2
([NET] NETNS: Omit net_device->nd_net without CONFIG_NET_NS)
breaks compilation with CONFIG_NET_NS set.

Fix the typo.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index d146be40f46c..06ca84d71db4 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -756,7 +756,7 @@ static inline
 void dev_net_set(struct net_device *dev, const struct net *net)
 {
 #ifdef CONFIG_NET_NS
-	dev->nd_dev = net;
+	dev->nd_net = net;
 #endif
 }
 
-- 
cgit v1.2.3


From f5aa23fd49063745f85644dd7a9330acd706add6 Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Wed, 26 Mar 2008 00:48:17 -0700
Subject: [NETNS]: Compilation warnings under CONFIG_NET_NS.

Recent commits from YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
have been introduced a several compilation warnings
'assignment discards qualifiers from pointer target type'
due to extra const modifier in the inline call parameters of
{dev|sock|twsk}_net_set.

Drop it.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h        | 2 +-
 include/net/inet_timewait_sock.h | 2 +-
 include/net/sock.h               | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 06ca84d71db4..15fa84a15c27 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -753,7 +753,7 @@ struct net *dev_net(const struct net_device *dev)
 }
 
 static inline
-void dev_net_set(struct net_device *dev, const struct net *net)
+void dev_net_set(struct net_device *dev, struct net *net)
 {
 #ifdef CONFIG_NET_NS
 	dev->nd_net = net;
diff --git a/include/net/inet_timewait_sock.h b/include/net/inet_timewait_sock.h
index 07fe0d1a4f03..95c660c9719b 100644
--- a/include/net/inet_timewait_sock.h
+++ b/include/net/inet_timewait_sock.h
@@ -219,7 +219,7 @@ struct net *twsk_net(const struct inet_timewait_sock *twsk)
 }
 
 static inline
-void twsk_net_set(struct inet_timewait_sock *twsk, const struct net *net)
+void twsk_net_set(struct inet_timewait_sock *twsk, struct net *net)
 {
 #ifdef CONFIG_NET_NS
 	twsk->tw_net = net;
diff --git a/include/net/sock.h b/include/net/sock.h
index 7e0d4a0c4d12..1c9d059223ee 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1358,7 +1358,7 @@ struct net *sock_net(const struct sock *sk)
 }
 
 static inline
-void sock_net_set(struct sock *sk, const struct net *net)
+void sock_net_set(struct sock *sk, struct net *net)
 {
 #ifdef CONFIG_NET_NS
 	sk->sk_net = net;
-- 
cgit v1.2.3


From 67727184f28c38d06013c6659560bb046c1d9f9c Mon Sep 17 00:00:00 2001
From: Pavel Emelyanov <xemul@openvz.org>
Date: Wed, 26 Mar 2008 16:27:22 -0700
Subject: [VLAN]: Reduce memory consumed by vlan_groups

Currently each vlan_groupd contains 8 pointers on arrays with 512
pointers on struct net_device each  :)  Such a construction "in many
cases ... wastes memory".

My proposal is to allow for some of these arrays pointers be NULL,
meaning that there are no devices in it. When a new device is added
to the vlan_group, the appropriate array is allocated.

The check in vlan_group_get_device's is safe, since the pointer
vg->vlan_devices_arrays[x] can only switch from NULL to not-NULL.
The vlan_group_prealloc_vid() is guarded with rtnl lock and is
also safe.

I've checked (I hope that) all the places, that use these arrays
and found, that the register_vlan_dev is the only place, that can
put a vlan device on an empty vlan_group.

Rough calculations shows, that after the patch a setup with a
single vlan dev (or up to 512 vlans with sequential vids) will
occupy approximately 8 times less memory.

The question I have is - does this patch makes sense, or a totally
new structures are required to store the vlan_devs?

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/if_vlan.h |  2 +-
 net/8021q/vlan.c        | 36 +++++++++++++++++++++++-------------
 2 files changed, 24 insertions(+), 14 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h
index 79504b22a932..edd55af7ebd6 100644
--- a/include/linux/if_vlan.h
+++ b/include/linux/if_vlan.h
@@ -93,7 +93,7 @@ static inline struct net_device *vlan_group_get_device(struct vlan_group *vg,
 {
 	struct net_device **array;
 	array = vg->vlan_devices_arrays[vlan_id / VLAN_GROUP_ARRAY_PART_LEN];
-	return array[vlan_id % VLAN_GROUP_ARRAY_PART_LEN];
+	return array ? array[vlan_id % VLAN_GROUP_ARRAY_PART_LEN] : NULL;
 }
 
 static inline void vlan_group_set_device(struct vlan_group *vg,
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index c35dc230365c..694be86e4490 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -106,29 +106,35 @@ static void vlan_group_free(struct vlan_group *grp)
 static struct vlan_group *vlan_group_alloc(int ifindex)
 {
 	struct vlan_group *grp;
-	unsigned int size;
-	unsigned int i;
 
 	grp = kzalloc(sizeof(struct vlan_group), GFP_KERNEL);
 	if (!grp)
 		return NULL;
 
-	size = sizeof(struct net_device *) * VLAN_GROUP_ARRAY_PART_LEN;
-
-	for (i = 0; i < VLAN_GROUP_ARRAY_SPLIT_PARTS; i++) {
-		grp->vlan_devices_arrays[i] = kzalloc(size, GFP_KERNEL);
-		if (!grp->vlan_devices_arrays[i])
-			goto err;
-	}
-
 	grp->real_dev_ifindex = ifindex;
 	hlist_add_head_rcu(&grp->hlist,
 			   &vlan_group_hash[vlan_grp_hashfn(ifindex)]);
 	return grp;
+}
 
-err:
-	vlan_group_free(grp);
-	return NULL;
+static int vlan_group_prealloc_vid(struct vlan_group *vg, int vid)
+{
+	struct net_device **array;
+	unsigned int size;
+
+	ASSERT_RTNL();
+
+	array = vg->vlan_devices_arrays[vid / VLAN_GROUP_ARRAY_PART_LEN];
+	if (array != NULL)
+		return 0;
+
+	size = sizeof(struct net_device *) * VLAN_GROUP_ARRAY_PART_LEN;
+	array = kzalloc(size, GFP_KERNEL);
+	if (array == NULL)
+		return -ENOBUFS;
+
+	vg->vlan_devices_arrays[vid / VLAN_GROUP_ARRAY_PART_LEN] = array;
+	return 0;
 }
 
 static void vlan_rcu_free(struct rcu_head *rcu)
@@ -247,6 +253,10 @@ int register_vlan_dev(struct net_device *dev)
 			return -ENOBUFS;
 	}
 
+	err = vlan_group_prealloc_vid(grp, vlan_id);
+	if (err < 0)
+		goto out_free_group;
+
 	err = register_netdevice(dev);
 	if (err < 0)
 		goto out_free_group;
-- 
cgit v1.2.3


From abc848c182960118fbb7cdae397b5608c5bcef1b Mon Sep 17 00:00:00 2001
From: Lennert Buytenhek <buytenh@wantstofly.org>
Date: Thu, 27 Mar 2008 14:51:39 -0400
Subject: introduce mbus DRAM target info abstraction

Introduce struct mbus_dram_target_info, which will be used for
passing information about the mbus target ID of the DDR unit, and
mbus target attribute, base address and size for each of the DRAM
chip selects from the platform code to peripheral drivers.

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Reviewed-by: Tzachi Perelstein <tzachi@marvell.com>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Nicolas Pitre <nico@marvell.com>
---
 include/linux/mbus.h | 36 ++++++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)
 create mode 100644 include/linux/mbus.h

(limited to 'include/linux')

diff --git a/include/linux/mbus.h b/include/linux/mbus.h
new file mode 100644
index 000000000000..c11ff2932549
--- /dev/null
+++ b/include/linux/mbus.h
@@ -0,0 +1,36 @@
+/*
+ * Marvell MBUS common definitions.
+ *
+ * Copyright (C) 2008 Marvell Semiconductor
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2.  This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#ifndef __LINUX_MBUS_H
+#define __LINUX_MBUS_H
+
+struct mbus_dram_target_info
+{
+	/*
+	 * The 4-bit MBUS target ID of the DRAM controller.
+	 */
+	u8		mbus_dram_target_id;
+
+	/*
+	 * The base address, size, and MBUS attribute ID for each
+	 * of the possible DRAM chip selects.  Peripherals are
+	 * required to support at least 4 decode windows.
+	 */
+	int		num_cs;
+	struct mbus_dram_window {
+		u8	cs_index;
+		u8	mbus_attr;
+		u32	base;
+		u32	size;
+	} cs[4];
+};
+
+
+#endif
-- 
cgit v1.2.3


From 15a32632d94011911497052a96cdbf3b905b325d Mon Sep 17 00:00:00 2001
From: Lennert Buytenhek <buytenh@wantstofly.org>
Date: Thu, 27 Mar 2008 14:51:39 -0400
Subject: sata_mv: mbus decode window support

Make it possible to pass mbus_dram_target_info to the sata_mv
driver via the platform data, make the sata_mv driver program
the window registers based on this data if it is passed in, and
make the Orion platform setup code use this method instead of
programming the SATA mbus window registers by hand.

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Reviewed-by: Tzachi Perelstein <tzachi@marvell.com>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Nicolas Pitre <nico@marvell.com>
---
 arch/arm/mach-orion/addr-map.c | 39 ---------------------------------------
 arch/arm/mach-orion/common.c   |  4 ++--
 arch/arm/mach-orion/common.h   |  1 -
 drivers/ata/sata_mv.c          | 31 +++++++++++++++++++++++++++++++
 include/linux/ata_platform.h   |  3 +++
 5 files changed, 36 insertions(+), 42 deletions(-)

(limited to 'include/linux')

diff --git a/arch/arm/mach-orion/addr-map.c b/arch/arm/mach-orion/addr-map.c
index 40bcb986ab96..3de5de9ac656 100644
--- a/arch/arm/mach-orion/addr-map.c
+++ b/arch/arm/mach-orion/addr-map.c
@@ -103,13 +103,6 @@
 #define ETH_MAX_WIN		6
 #define ETH_MAX_REMAP_WIN	4
 
-/*
- * SATA Address Decode Windows registers
- */
-#define SATA_WIN_CTRL(win)	ORION_SATA_REG(0x30 + ((win) * 0x10))
-#define SATA_WIN_BASE(win)	ORION_SATA_REG(0x34 + ((win) * 0x10))
-#define SATA_MAX_WIN		4
-
 
 struct mbus_dram_target_info orion_mbus_dram_info;
 
@@ -288,35 +281,3 @@ void __init orion_setup_eth_wins(void)
 		}
 	}
 }
-
-void __init orion_setup_sata_wins(void)
-{
-	int i;
-
-	/*
-	 * First, disable and clear windows
-	 */
-	for (i = 0; i < SATA_MAX_WIN; i++) {
-		orion_write(SATA_WIN_BASE(i), 0);
-		orion_write(SATA_WIN_CTRL(i), 0);
-	}
-
-	/*
-	 * Setup windows for DDR banks.
-	 */
-	for (i = 0; i < DDR_MAX_CS; i++) {
-		u32 base, size;
-		size = orion_read(DDR_SIZE_CS(i));
-		base = orion_read(DDR_BASE_CS(i));
-		if (size & DDR_BANK_EN) {
-			base = DDR_REG_TO_BASE(base);
-			size = DDR_REG_TO_SIZE(size);
-			orion_write(SATA_WIN_CTRL(i),
-					((size-1) & 0xffff0000) |
-					(ATTR_DDR_CS(i) << 8) |
-					(TARGET_DDR << 4) | WIN_EN);
-			orion_write(SATA_WIN_BASE(i),
-					base & 0xffff0000);
-		}
-	}
-}
diff --git a/arch/arm/mach-orion/common.c b/arch/arm/mach-orion/common.c
index d33c01dfc3f2..a32fe8e108bc 100644
--- a/arch/arm/mach-orion/common.c
+++ b/arch/arm/mach-orion/common.c
@@ -17,6 +17,7 @@
 #include <linux/mbus.h>
 #include <linux/mv643xx_eth.h>
 #include <linux/mv643xx_i2c.h>
+#include <linux/ata_platform.h>
 #include <asm/page.h>
 #include <asm/setup.h>
 #include <asm/timex.h>
@@ -289,6 +290,7 @@ static struct platform_device orion_sata = {
 
 void __init orion_sata_init(struct mv_sata_platform_data *sata_data)
 {
+	sata_data->dram = &orion_mbus_dram_info;
 	orion_sata.dev.platform_data = sata_data;
 	platform_device_register(&orion_sata);
 }
@@ -342,8 +344,6 @@ void __init orion_init(void)
 	 */
 	orion_setup_cpu_wins();
 	orion_setup_eth_wins();
-	if (dev == MV88F5182_DEV_ID)
-		orion_setup_sata_wins();
 
 	/*
 	 * REgister devices
diff --git a/arch/arm/mach-orion/common.h b/arch/arm/mach-orion/common.h
index c100355754f3..b676be0a4a86 100644
--- a/arch/arm/mach-orion/common.h
+++ b/arch/arm/mach-orion/common.h
@@ -33,7 +33,6 @@ extern struct mbus_dram_target_info orion_mbus_dram_info;
 void orion_setup_cpu_win(enum orion_target target, u32 base, u32 size, int remap);
 void orion_setup_cpu_wins(void);
 void orion_setup_eth_wins(void);
-void orion_setup_sata_wins(void);
 
 /*
  * Shared code used internally by other Orion core functions.
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 6ebebde8454a..83584b6e1ba5 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -74,6 +74,7 @@
 #include <linux/device.h>
 #include <linux/platform_device.h>
 #include <linux/ata_platform.h>
+#include <linux/mbus.h>
 #include <scsi/scsi_host.h>
 #include <scsi/scsi_cmnd.h>
 #include <scsi/scsi_device.h>
@@ -352,6 +353,9 @@ enum {
 #define IS_GEN_IIE(hpriv) ((hpriv)->hp_flags & MV_HP_GEN_IIE)
 #define HAS_PCI(host) (!((host)->ports[0]->flags & MV_FLAG_SOC))
 
+#define WINDOW_CTRL(i)		(0x20030 + ((i) << 4))
+#define WINDOW_BASE(i)		(0x20034 + ((i) << 4))
+
 enum {
 	/* DMA boundary 0xffff is required by the s/g splitting
 	 * we need on /length/ in mv_fill-sg().
@@ -2897,6 +2901,27 @@ static int mv_create_dma_pools(struct mv_host_priv *hpriv, struct device *dev)
 	return 0;
 }
 
+static void mv_conf_mbus_windows(struct mv_host_priv *hpriv,
+				 struct mbus_dram_target_info *dram)
+{
+	int i;
+
+	for (i = 0; i < 4; i++) {
+		writel(0, hpriv->base + WINDOW_CTRL(i));
+		writel(0, hpriv->base + WINDOW_BASE(i));
+	}
+
+	for (i = 0; i < dram->num_cs; i++) {
+		struct mbus_dram_window *cs = dram->cs + i;
+
+		writel(((cs->size - 1) & 0xffff0000) |
+			(cs->mbus_attr << 8) |
+			(dram->mbus_dram_target_id << 4) | 1,
+			hpriv->base + WINDOW_CTRL(i));
+		writel(cs->base, hpriv->base + WINDOW_BASE(i));
+	}
+}
+
 /**
  *      mv_platform_probe - handle a positive probe of an soc Marvell
  *      host
@@ -2951,6 +2976,12 @@ static int mv_platform_probe(struct platform_device *pdev)
 				   res->end - res->start + 1);
 	hpriv->base -= MV_SATAHC0_REG_BASE;
 
+	/*
+	 * (Re-)program MBUS remapping windows if we are asked to.
+	 */
+	if (mv_platform_data->dram != NULL)
+		mv_conf_mbus_windows(hpriv, mv_platform_data->dram);
+
 	rc = mv_create_dma_pools(hpriv, &pdev->dev);
 	if (rc)
 		return rc;
diff --git a/include/linux/ata_platform.h b/include/linux/ata_platform.h
index b856a2a590d9..9a26c83a2c9e 100644
--- a/include/linux/ata_platform.h
+++ b/include/linux/ata_platform.h
@@ -27,7 +27,10 @@ extern int __devexit __pata_platform_remove(struct device *dev);
 /*
  * Marvell SATA private data
  */
+struct mbus_dram_target_info;
+
 struct mv_sata_platform_data {
+	struct mbus_dram_target_info	*dram;
 	int	n_ports; /* number of sata ports */
 };
 
-- 
cgit v1.2.3


From 0e5f8be1388093edc324a78ebf241170b258eba3 Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Thu, 27 Mar 2008 14:25:53 -0700
Subject: [NETNS]: Compile NET /proc support only if CONFIG_NET is set.

This fix broken compilation for 'allnoconfig'. This was introduced by
Introduced by commit 1218854afa6f659be90b748cf1bc7badee954a35 ("[NET]
NETNS: Omit seq_net_private->net without CONFIG_NET_NS.")

Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 fs/proc/proc_net.c       | 2 ++
 include/linux/seq_file.h | 2 ++
 2 files changed, 4 insertions(+)

(limited to 'include/linux')

diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c
index 13cd7835d0df..7034facf8b8f 100644
--- a/fs/proc/proc_net.c
+++ b/fs/proc/proc_net.c
@@ -51,6 +51,7 @@ int seq_open_net(struct inode *ino, struct file *f,
 }
 EXPORT_SYMBOL_GPL(seq_open_net);
 
+#ifdef CONFIG_NET
 int seq_release_net(struct inode *ino, struct file *f)
 {
 	struct seq_file *seq;
@@ -218,3 +219,4 @@ int __init proc_net_init(void)
 
 	return register_pernet_subsys(&proc_net_ns_ops);
 }
+#endif /* CONFIG_NET */
diff --git a/include/linux/seq_file.h b/include/linux/seq_file.h
index d870a8253769..5da70c3f4417 100644
--- a/include/linux/seq_file.h
+++ b/include/linux/seq_file.h
@@ -63,6 +63,7 @@ extern struct list_head *seq_list_start_head(struct list_head *head,
 extern struct list_head *seq_list_next(void *v, struct list_head *head,
 		loff_t *ppos);
 
+#ifdef CONFIG_NET
 struct net;
 struct seq_net_private {
 #ifdef CONFIG_NET_NS
@@ -81,6 +82,7 @@ static inline struct net *seq_file_net(struct seq_file *seq)
 	return &init_net;
 #endif
 }
+#endif	/* CONFIG_NET */
 
 #endif
 #endif
-- 
cgit v1.2.3


From a6bd8e13034dd7d60b6f14217096efa192d0adc1 Mon Sep 17 00:00:00 2001
From: Rusty Russell <rusty@rustcorp.com.au>
Date: Fri, 28 Mar 2008 11:05:53 -0500
Subject: lguest: comment documentation update.

Took some cycles to re-read the Lguest Journey end-to-end, fix some
rot and tighten some phrases.

Only comments change.  No new jokes, but a couple of recycled old jokes.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
---
 Documentation/lguest/lguest.c         |  69 ++++++++++++----------
 arch/x86/lguest/boot.c                | 108 +++++++++++++++++++---------------
 arch/x86/lguest/i386_head.S           |  15 +++--
 drivers/lguest/core.c                 |  18 +++---
 drivers/lguest/hypercalls.c           |  11 +++-
 drivers/lguest/interrupts_and_traps.c |   7 +--
 drivers/lguest/lguest_device.c        |  11 ++--
 drivers/lguest/lguest_user.c          |  30 +++++++---
 drivers/lguest/page_tables.c          |  32 +++++-----
 drivers/lguest/x86/core.c             |  33 +++++++----
 drivers/lguest/x86/switcher_32.S      |   8 +--
 include/asm-x86/lguest_hcall.h        |   2 +-
 include/linux/lguest_launcher.h       |   6 +-
 13 files changed, 208 insertions(+), 142 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/lguest/lguest.c b/Documentation/lguest/lguest.c
index d45c7f682b1b..4c1fc65a8b3d 100644
--- a/Documentation/lguest/lguest.c
+++ b/Documentation/lguest/lguest.c
@@ -1,7 +1,7 @@
 /*P:100 This is the Launcher code, a simple program which lays out the
- * "physical" memory for the new Guest by mapping the kernel image and the
- * virtual devices, then reads repeatedly from /dev/lguest to run the Guest.
-:*/
+ * "physical" memory for the new Guest by mapping the kernel image and
+ * the virtual devices, then opens /dev/lguest to tell the kernel
+ * about the Guest and control it. :*/
 #define _LARGEFILE64_SOURCE
 #define _GNU_SOURCE
 #include <stdio.h>
@@ -43,7 +43,7 @@
 #include "linux/virtio_console.h"
 #include "linux/virtio_ring.h"
 #include "asm-x86/bootparam.h"
-/*L:110 We can ignore the 38 include files we need for this program, but I do
+/*L:110 We can ignore the 39 include files we need for this program, but I do
  * want to draw attention to the use of kernel-style types.
  *
  * As Linus said, "C is a Spartan language, and so should your naming be."  I
@@ -320,7 +320,7 @@ static unsigned long map_elf(int elf_fd, const Elf32_Ehdr *ehdr)
 		err(1, "Reading program headers");
 
 	/* Try all the headers: there are usually only three.  A read-only one,
-	 * a read-write one, and a "note" section which isn't loadable. */
+	 * a read-write one, and a "note" section which we don't load. */
 	for (i = 0; i < ehdr->e_phnum; i++) {
 		/* If this isn't a loadable segment, we ignore it */
 		if (phdr[i].p_type != PT_LOAD)
@@ -387,7 +387,7 @@ static unsigned long load_kernel(int fd)
 	if (memcmp(hdr.e_ident, ELFMAG, SELFMAG) == 0)
 		return map_elf(fd, &hdr);
 
-	/* Otherwise we assume it's a bzImage, and try to unpack it */
+	/* Otherwise we assume it's a bzImage, and try to load it. */
 	return load_bzimage(fd);
 }
 
@@ -433,12 +433,12 @@ static unsigned long load_initrd(const char *name, unsigned long mem)
 	return len;
 }
 
-/* Once we know how much memory we have, we can construct simple linear page
+/* Once we know how much memory we have we can construct simple linear page
  * tables which set virtual == physical which will get the Guest far enough
  * into the boot to create its own.
  *
  * We lay them out of the way, just below the initrd (which is why we need to
- * know its size). */
+ * know its size here). */
 static unsigned long setup_pagetables(unsigned long mem,
 				      unsigned long initrd_size)
 {
@@ -850,7 +850,8 @@ static void handle_console_output(int fd, struct virtqueue *vq)
  *
  * Handling output for network is also simple: we get all the output buffers
  * and write them (ignoring the first element) to this device's file descriptor
- * (stdout). */
+ * (/dev/net/tun).
+ */
 static void handle_net_output(int fd, struct virtqueue *vq)
 {
 	unsigned int head, out, in;
@@ -924,7 +925,7 @@ static void enable_fd(int fd, struct virtqueue *vq)
 	write(waker_fd, &vq->dev->fd, sizeof(vq->dev->fd));
 }
 
-/* Resetting a device is fairly easy. */
+/* When the Guest asks us to reset a device, it's is fairly easy. */
 static void reset_device(struct device *dev)
 {
 	struct virtqueue *vq;
@@ -1003,8 +1004,8 @@ static void handle_input(int fd)
 		if (select(devices.max_infd+1, &fds, NULL, NULL, &poll) == 0)
 			break;
 
-		/* Otherwise, call the device(s) which have readable
-		 * file descriptors and a method of handling them.  */
+		/* Otherwise, call the device(s) which have readable file
+		 * descriptors and a method of handling them.  */
 		for (i = devices.dev; i; i = i->next) {
 			if (i->handle_input && FD_ISSET(i->fd, &fds)) {
 				int dev_fd;
@@ -1015,8 +1016,7 @@ static void handle_input(int fd)
 				 * should no longer service it.  Networking and
 				 * console do this when there's no input
 				 * buffers to deliver into.  Console also uses
-				 * it when it discovers that stdin is
-				 * closed. */
+				 * it when it discovers that stdin is closed. */
 				FD_CLR(i->fd, &devices.infds);
 				/* Tell waker to ignore it too, by sending a
 				 * negative fd number (-1, since 0 is a valid
@@ -1033,7 +1033,8 @@ static void handle_input(int fd)
  *
  * All devices need a descriptor so the Guest knows it exists, and a "struct
  * device" so the Launcher can keep track of it.  We have common helper
- * routines to allocate and manage them. */
+ * routines to allocate and manage them.
+ */
 
 /* The layout of the device page is a "struct lguest_device_desc" followed by a
  * number of virtqueue descriptors, then two sets of feature bits, then an
@@ -1078,7 +1079,7 @@ static void add_virtqueue(struct device *dev, unsigned int num_descs,
 	struct virtqueue **i, *vq = malloc(sizeof(*vq));
 	void *p;
 
-	/* First we need some pages for this virtqueue. */
+	/* First we need some memory for this virtqueue. */
 	pages = (vring_size(num_descs, getpagesize()) + getpagesize() - 1)
 		/ getpagesize();
 	p = get_pages(pages);
@@ -1122,7 +1123,7 @@ static void add_virtqueue(struct device *dev, unsigned int num_descs,
 }
 
 /* The first half of the feature bitmask is for us to advertise features.  The
- * second half if for the Guest to accept features. */
+ * second half is for the Guest to accept features. */
 static void add_feature(struct device *dev, unsigned bit)
 {
 	u8 *features = get_feature_bits(dev);
@@ -1151,7 +1152,9 @@ static void set_config(struct device *dev, unsigned len, const void *conf)
 }
 
 /* This routine does all the creation and setup of a new device, including
- * calling new_dev_desc() to allocate the descriptor and device memory. */
+ * calling new_dev_desc() to allocate the descriptor and device memory.
+ *
+ * See what I mean about userspace being boring? */
 static struct device *new_device(const char *name, u16 type, int fd,
 				 bool (*handle_input)(int, struct device *))
 {
@@ -1492,7 +1495,10 @@ static int io_thread(void *_dev)
 	while (read(vblk->workpipe[0], &c, 1) == 1) {
 		/* We acknowledge each request immediately to reduce latency,
 		 * rather than waiting until we've done them all.  I haven't
-		 * measured to see if it makes any difference. */
+		 * measured to see if it makes any difference.
+		 *
+		 * That would be an interesting test, wouldn't it?  You could
+		 * also try having more than one I/O thread. */
 		while (service_io(dev))
 			write(vblk->done_fd, &c, 1);
 	}
@@ -1500,7 +1506,7 @@ static int io_thread(void *_dev)
 }
 
 /* Now we've seen the I/O thread, we return to the Launcher to see what happens
- * when the thread tells us it's completed some I/O. */
+ * when that thread tells us it's completed some I/O. */
 static bool handle_io_finish(int fd, struct device *dev)
 {
 	char c;
@@ -1572,11 +1578,12 @@ static void setup_block_file(const char *filename)
 	 * more work. */
 	pipe(vblk->workpipe);
 
-	/* Create stack for thread and run it */
+	/* Create stack for thread and run it.  Since stack grows upwards, we
+	 * point the stack pointer to the end of this region. */
 	stack = malloc(32768);
 	/* SIGCHLD - We dont "wait" for our cloned thread, so prevent it from
 	 * becoming a zombie. */
-	if (clone(io_thread, stack + 32768,  CLONE_VM | SIGCHLD, dev) == -1)
+	if (clone(io_thread, stack + 32768, CLONE_VM | SIGCHLD, dev) == -1)
 		err(1, "Creating clone");
 
 	/* We don't need to keep the I/O thread's end of the pipes open. */
@@ -1586,14 +1593,14 @@ static void setup_block_file(const char *filename)
 	verbose("device %u: virtblock %llu sectors\n",
 		devices.device_num, le64_to_cpu(conf.capacity));
 }
-/* That's the end of device setup. :*/
+/* That's the end of device setup. */
 
-/* Reboot */
+/*L:230 Reboot is pretty easy: clean up and exec() the Launcher afresh. */
 static void __attribute__((noreturn)) restart_guest(void)
 {
 	unsigned int i;
 
-	/* Closing pipes causes the waker thread and io_threads to die, and
+	/* Closing pipes causes the Waker thread and io_threads to die, and
 	 * closing /dev/lguest cleans up the Guest.  Since we don't track all
 	 * open fds, we simply close everything beyond stderr. */
 	for (i = 3; i < FD_SETSIZE; i++)
@@ -1602,7 +1609,7 @@ static void __attribute__((noreturn)) restart_guest(void)
 	err(1, "Could not exec %s", main_args[0]);
 }
 
-/*L:220 Finally we reach the core of the Launcher, which runs the Guest, serves
+/*L:220 Finally we reach the core of the Launcher which runs the Guest, serves
  * its input and output, and finally, lays it to rest. */
 static void __attribute__((noreturn)) run_guest(int lguest_fd)
 {
@@ -1643,7 +1650,7 @@ static void __attribute__((noreturn)) run_guest(int lguest_fd)
 			err(1, "Resetting break");
 	}
 }
-/*
+/*L:240
  * This is the end of the Launcher.  The good news: we are over halfway
  * through!  The bad news: the most fiendish part of the code still lies ahead
  * of us.
@@ -1690,8 +1697,8 @@ int main(int argc, char *argv[])
 	 * device receive input from a file descriptor, we keep an fdset
 	 * (infds) and the maximum fd number (max_infd) with the head of the
 	 * list.  We also keep a pointer to the last device.  Finally, we keep
-	 * the next interrupt number to hand out (1: remember that 0 is used by
-	 * the timer). */
+	 * the next interrupt number to use for devices (1: remember that 0 is
+	 * used by the timer). */
 	FD_ZERO(&devices.infds);
 	devices.max_infd = -1;
 	devices.lastdev = NULL;
@@ -1792,8 +1799,8 @@ int main(int argc, char *argv[])
 	lguest_fd = tell_kernel(pgdir, start);
 
 	/* We fork off a child process, which wakes the Launcher whenever one
-	 * of the input file descriptors needs attention.  Otherwise we would
-	 * run the Guest until it tries to output something. */
+	 * of the input file descriptors needs attention.  We call this the
+	 * Waker, and we'll cover it in a moment. */
 	waker_fd = setup_waker(lguest_fd);
 
 	/* Finally, run the Guest.  This doesn't return. */
diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c
index a104c532ff70..3335b4595efd 100644
--- a/arch/x86/lguest/boot.c
+++ b/arch/x86/lguest/boot.c
@@ -10,21 +10,19 @@
  * (such as the example in Documentation/lguest/lguest.c) is called the
  * Launcher.
  *
- * Secondly, we only run specially modified Guests, not normal kernels.  When
- * you set CONFIG_LGUEST to 'y' or 'm', this automatically sets
- * CONFIG_LGUEST_GUEST=y, which compiles this file into the kernel so it knows
- * how to be a Guest.  This means that you can use the same kernel you boot
- * normally (ie. as a Host) as a Guest.
+ * Secondly, we only run specially modified Guests, not normal kernels: setting
+ * CONFIG_LGUEST_GUEST to "y" compiles this file into the kernel so it knows
+ * how to be a Guest at boot time.  This means that you can use the same kernel
+ * you boot normally (ie. as a Host) as a Guest.
  *
  * These Guests know that they cannot do privileged operations, such as disable
  * interrupts, and that they have to ask the Host to do such things explicitly.
  * This file consists of all the replacements for such low-level native
  * hardware operations: these special Guest versions call the Host.
  *
- * So how does the kernel know it's a Guest?  The Guest starts at a special
- * entry point marked with a magic string, which sets up a few things then
- * calls here.  We replace the native functions various "paravirt" structures
- * with our Guest versions, then boot like normal. :*/
+ * So how does the kernel know it's a Guest?  We'll see that later, but let's
+ * just say that we end up here where we replace the native functions various
+ * "paravirt" structures with our Guest versions, then boot like normal. :*/
 
 /*
  * Copyright (C) 2006, Rusty Russell <rusty@rustcorp.com.au> IBM Corporation.
@@ -134,7 +132,7 @@ static void async_hcall(unsigned long call, unsigned long arg1,
  * lguest_leave_lazy_mode().
  *
  * So, when we're in lazy mode, we call async_hcall() to store the call for
- * future processing. */
+ * future processing: */
 static void lazy_hcall(unsigned long call,
 		       unsigned long arg1,
 		       unsigned long arg2,
@@ -147,7 +145,7 @@ static void lazy_hcall(unsigned long call,
 }
 
 /* When lazy mode is turned off reset the per-cpu lazy mode variable and then
- * issue a hypercall to flush any stored calls. */
+ * issue the do-nothing hypercall to flush any stored calls. */
 static void lguest_leave_lazy_mode(void)
 {
 	paravirt_leave_lazy(paravirt_get_lazy_mode());
@@ -164,7 +162,7 @@ static void lguest_leave_lazy_mode(void)
  *
  * So instead we keep an "irq_enabled" field inside our "struct lguest_data",
  * which the Guest can update with a single instruction.  The Host knows to
- * check there when it wants to deliver an interrupt.
+ * check there before it tries to deliver an interrupt.
  */
 
 /* save_flags() is expected to return the processor state (ie. "flags").  The
@@ -196,10 +194,15 @@ static void irq_enable(void)
 /*M:003 Note that we don't check for outstanding interrupts when we re-enable
  * them (or when we unmask an interrupt).  This seems to work for the moment,
  * since interrupts are rare and we'll just get the interrupt on the next timer
- * tick, but when we turn on CONFIG_NO_HZ, we should revisit this.  One way
+ * tick, but now we can run with CONFIG_NO_HZ, we should revisit this.  One way
  * would be to put the "irq_enabled" field in a page by itself, and have the
  * Host write-protect it when an interrupt comes in when irqs are disabled.
- * There will then be a page fault as soon as interrupts are re-enabled. :*/
+ * There will then be a page fault as soon as interrupts are re-enabled.
+ *
+ * A better method is to implement soft interrupt disable generally for x86:
+ * instead of disabling interrupts, we set a flag.  If an interrupt does come
+ * in, we then disable them for real.  This is uncommon, so we could simply use
+ * a hypercall for interrupt control and not worry about efficiency. :*/
 
 /*G:034
  * The Interrupt Descriptor Table (IDT).
@@ -212,6 +215,10 @@ static void irq_enable(void)
 static void lguest_write_idt_entry(gate_desc *dt,
 				   int entrynum, const gate_desc *g)
 {
+	/* The gate_desc structure is 8 bytes long: we hand it to the Host in
+	 * two 32-bit chunks.  The whole 32-bit kernel used to hand descriptors
+	 * around like this; typesafety wasn't a big concern in Linux's early
+	 * years. */
 	u32 *desc = (u32 *)g;
 	/* Keep the local copy up to date. */
 	native_write_idt_entry(dt, entrynum, g);
@@ -243,7 +250,8 @@ static void lguest_load_idt(const struct desc_ptr *desc)
  *
  * This is the opposite of the IDT code where we have a LOAD_IDT_ENTRY
  * hypercall and use that repeatedly to load a new IDT.  I don't think it
- * really matters, but wouldn't it be nice if they were the same?
+ * really matters, but wouldn't it be nice if they were the same?  Wouldn't
+ * it be even better if you were the one to send the patch to fix it?
  */
 static void lguest_load_gdt(const struct desc_ptr *desc)
 {
@@ -298,9 +306,9 @@ static void lguest_load_tr_desc(void)
 
 /* The "cpuid" instruction is a way of querying both the CPU identity
  * (manufacturer, model, etc) and its features.  It was introduced before the
- * Pentium in 1993 and keeps getting extended by both Intel and AMD.  As you
- * might imagine, after a decade and a half this treatment, it is now a giant
- * ball of hair.  Its entry in the current Intel manual runs to 28 pages.
+ * Pentium in 1993 and keeps getting extended by both Intel, AMD and others.
+ * As you might imagine, after a decade and a half this treatment, it is now a
+ * giant ball of hair.  Its entry in the current Intel manual runs to 28 pages.
  *
  * This instruction even it has its own Wikipedia entry.  The Wikipedia entry
  * has been translated into 4 languages.  I am not making this up!
@@ -594,17 +602,17 @@ static unsigned long lguest_get_wallclock(void)
 	return lguest_data.time.tv_sec;
 }
 
-/* The TSC is a Time Stamp Counter.  The Host tells us what speed it runs at,
- * or 0 if it's unusable as a reliable clock source.  This matches what we want
- * here: if we return 0 from this function, the x86 TSC clock will not register
- * itself. */
+/* The TSC is an Intel thing called the Time Stamp Counter.  The Host tells us
+ * what speed it runs at, or 0 if it's unusable as a reliable clock source.
+ * This matches what we want here: if we return 0 from this function, the x86
+ * TSC clock will give up and not register itself. */
 static unsigned long lguest_cpu_khz(void)
 {
 	return lguest_data.tsc_khz;
 }
 
-/* If we can't use the TSC, the kernel falls back to our "lguest_clock", where
- * we read the time value given to us by the Host. */
+/* If we can't use the TSC, the kernel falls back to our lower-priority
+ * "lguest_clock", where we read the time value given to us by the Host. */
 static cycle_t lguest_clock_read(void)
 {
 	unsigned long sec, nsec;
@@ -648,12 +656,16 @@ static struct clocksource lguest_clock = {
 static int lguest_clockevent_set_next_event(unsigned long delta,
                                            struct clock_event_device *evt)
 {
+	/* FIXME: I don't think this can ever happen, but James tells me he had
+	 * to put this code in.  Maybe we should remove it now.  Anyone? */
 	if (delta < LG_CLOCK_MIN_DELTA) {
 		if (printk_ratelimit())
 			printk(KERN_DEBUG "%s: small delta %lu ns\n",
 			       __FUNCTION__, delta);
 		return -ETIME;
 	}
+
+	/* Please wake us this far in the future. */
 	hcall(LHCALL_SET_CLOCKEVENT, delta, 0, 0);
 	return 0;
 }
@@ -738,7 +750,7 @@ static void lguest_time_init(void)
  * will not tolerate us trying to use that), the stack pointer, and the number
  * of pages in the stack. */
 static void lguest_load_sp0(struct tss_struct *tss,
-				     struct thread_struct *thread)
+			    struct thread_struct *thread)
 {
 	lazy_hcall(LHCALL_SET_STACK, __KERNEL_DS|0x1, thread->sp0,
 		   THREAD_SIZE/PAGE_SIZE);
@@ -786,9 +798,8 @@ static void lguest_safe_halt(void)
 	hcall(LHCALL_HALT, 0, 0, 0);
 }
 
-/* Perhaps CRASH isn't the best name for this hypercall, but we use it to get a
- * message out when we're crashing as well as elegant termination like powering
- * off.
+/* The SHUTDOWN hypercall takes a string to describe what's happening, and
+ * an argument which says whether this to restart (reboot) the Guest or not.
  *
  * Note that the Host always prefers that the Guest speak in physical addresses
  * rather than virtual addresses, so we use __pa() here. */
@@ -816,8 +827,9 @@ static struct notifier_block paniced = {
 /* Setting up memory is fairly easy. */
 static __init char *lguest_memory_setup(void)
 {
-	/* We do this here and not earlier because lockcheck barfs if we do it
-	 * before start_kernel() */
+	/* We do this here and not earlier because lockcheck used to barf if we
+	 * did it before start_kernel().  I think we fixed that, so it'd be
+	 * nice to move it back to lguest_init.  Patch welcome... */
 	atomic_notifier_chain_register(&panic_notifier_list, &paniced);
 
 	/* The Linux bootloader header contains an "e820" memory map: the
@@ -850,12 +862,19 @@ static __init int early_put_chars(u32 vtermno, const char *buf, int count)
 	return len;
 }
 
+/* Rebooting also tells the Host we're finished, but the RESTART flag tells the
+ * Launcher to reboot us. */
+static void lguest_restart(char *reason)
+{
+	hcall(LHCALL_SHUTDOWN, __pa(reason), LGUEST_SHUTDOWN_RESTART, 0);
+}
+
 /*G:050
  * Patching (Powerfully Placating Performance Pedants)
  *
- * We have already seen that pv_ops structures let us replace simple
- * native instructions with calls to the appropriate back end all throughout
- * the kernel.  This allows the same kernel to run as a Guest and as a native
+ * We have already seen that pv_ops structures let us replace simple native
+ * instructions with calls to the appropriate back end all throughout the
+ * kernel.  This allows the same kernel to run as a Guest and as a native
  * kernel, but it's slow because of all the indirect branches.
  *
  * Remember that David Wheeler quote about "Any problem in computer science can
@@ -908,14 +927,9 @@ static unsigned lguest_patch(u8 type, u16 clobber, void *ibuf,
 	return insn_len;
 }
 
-static void lguest_restart(char *reason)
-{
-	hcall(LHCALL_SHUTDOWN, __pa(reason), LGUEST_SHUTDOWN_RESTART, 0);
-}
-
-/*G:030 Once we get to lguest_init(), we know we're a Guest.  The pv_ops
- * structures in the kernel provide points for (almost) every routine we have
- * to override to avoid privileged instructions. */
+/*G:030 Once we get to lguest_init(), we know we're a Guest.  The various
+ * pv_ops structures in the kernel provide points for (almost) every routine we
+ * have to override to avoid privileged instructions. */
 __init void lguest_init(void)
 {
 	/* We're under lguest, paravirt is enabled, and we're running at
@@ -1003,9 +1017,9 @@ __init void lguest_init(void)
 	 * the normal data segment to get through booting. */
 	asm volatile ("mov %0, %%fs" : : "r" (__KERNEL_DS) : "memory");
 
-	/* The Host uses the top of the Guest's virtual address space for the
-	 * Host<->Guest Switcher, and it tells us how big that is in
-	 * lguest_data.reserve_mem, set up on the LGUEST_INIT hypercall. */
+	/* The Host<->Guest Switcher lives at the top of our address space, and
+	 * the Host told us how big it is when we made LGUEST_INIT hypercall:
+	 * it put the answer in lguest_data.reserve_mem  */
 	reserve_top_address(lguest_data.reserve_mem);
 
 	/* If we don't initialize the lock dependency checker now, it crashes
@@ -1027,6 +1041,7 @@ __init void lguest_init(void)
 	/* Math is always hard! */
 	new_cpu_data.hard_math = 1;
 
+	/* We don't have features.  We have puppies!  Puppies! */
 #ifdef CONFIG_X86_MCE
 	mce_disabled = 1;
 #endif
@@ -1044,10 +1059,11 @@ __init void lguest_init(void)
 	virtio_cons_early_init(early_put_chars);
 
 	/* Last of all, we set the power management poweroff hook to point to
-	 * the Guest routine to power off. */
+	 * the Guest routine to power off, and the reboot hook to our restart
+	 * routine. */
 	pm_power_off = lguest_power_off;
-
 	machine_ops.restart = lguest_restart;
+
 	/* Now we're set up, call start_kernel() in init/main.c and we proceed
 	 * to boot as normal.  It never returns. */
 	start_kernel();
diff --git a/arch/x86/lguest/i386_head.S b/arch/x86/lguest/i386_head.S
index 95b6fbcded63..5c7cef34c9e7 100644
--- a/arch/x86/lguest/i386_head.S
+++ b/arch/x86/lguest/i386_head.S
@@ -5,13 +5,20 @@
 #include <asm/thread_info.h>
 #include <asm/processor-flags.h>
 
-/*G:020 This is where we begin: head.S notes that the boot header's platform
- * type field is "1" (lguest), so calls us here.
+/*G:020 Our story starts with the kernel booting into startup_32 in
+ * arch/x86/kernel/head_32.S.  It expects a boot header, which is created by
+ * the bootloader (the Launcher in our case).
+ *
+ * The startup_32 function does very little: it clears the uninitialized global
+ * C variables which we expect to be zero (ie. BSS) and then copies the boot
+ * header and kernel command line somewhere safe.  Finally it checks the
+ * 'hardware_subarch' field.  This was introduced in 2.6.24 for lguest and Xen:
+ * if it's set to '1' (lguest's assigned number), then it calls us here.
  *
  * WARNING: be very careful here!  We're running at addresses equal to physical
  * addesses (around 0), not above PAGE_OFFSET as most code expectes
  * (eg. 0xC0000000).  Jumps are relative, so they're OK, but we can't touch any
- * data.
+ * data without remembering to subtract __PAGE_OFFSET!
  *
  * The .section line puts this code in .init.text so it will be discarded after
  * boot. */
@@ -24,7 +31,7 @@ ENTRY(lguest_entry)
 	int $LGUEST_TRAP_ENTRY
 
 	/* The Host put the toplevel pagetable in lguest_data.pgdir.  The movsl
-	 * instruction uses %esi implicitly as the source for the copy we'
+	 * instruction uses %esi implicitly as the source for the copy we're
 	 * about to do. */
 	movl lguest_data - __PAGE_OFFSET + LGUEST_DATA_pgdir, %esi
 
diff --git a/drivers/lguest/core.c b/drivers/lguest/core.c
index c632c08cbbdc..5eea4356d703 100644
--- a/drivers/lguest/core.c
+++ b/drivers/lguest/core.c
@@ -1,8 +1,6 @@
 /*P:400 This contains run_guest() which actually calls into the Host<->Guest
  * Switcher and analyzes the return, such as determining if the Guest wants the
- * Host to do something.  This file also contains useful helper routines, and a
- * couple of non-obvious setup and teardown pieces which were implemented after
- * days of debugging pain. :*/
+ * Host to do something.  This file also contains useful helper routines. :*/
 #include <linux/module.h>
 #include <linux/stringify.h>
 #include <linux/stddef.h>
@@ -49,8 +47,8 @@ static __init int map_switcher(void)
 	 * easy.
 	 */
 
-	/* We allocate an array of "struct page"s.  map_vm_area() wants the
-	 * pages in this form, rather than just an array of pointers. */
+	/* We allocate an array of struct page pointers.  map_vm_area() wants
+	 * this, rather than just an array of pages. */
 	switcher_page = kmalloc(sizeof(switcher_page[0])*TOTAL_SWITCHER_PAGES,
 				GFP_KERNEL);
 	if (!switcher_page) {
@@ -172,7 +170,7 @@ void __lgread(struct lg_cpu *cpu, void *b, unsigned long addr, unsigned bytes)
 	}
 }
 
-/* This is the write (copy into guest) version. */
+/* This is the write (copy into Guest) version. */
 void __lgwrite(struct lg_cpu *cpu, unsigned long addr, const void *b,
 	       unsigned bytes)
 {
@@ -209,9 +207,9 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user)
 		if (cpu->break_out)
 			return -EAGAIN;
 
-		/* Check if there are any interrupts which can be delivered
-		 * now: if so, this sets up the hander to be executed when we
-		 * next run the Guest. */
+		/* Check if there are any interrupts which can be delivered now:
+		 * if so, this sets up the hander to be executed when we next
+		 * run the Guest. */
 		maybe_do_interrupt(cpu);
 
 		/* All long-lived kernel loops need to check with this horrible
@@ -246,8 +244,10 @@ int run_guest(struct lg_cpu *cpu, unsigned long __user *user)
 		lguest_arch_handle_trap(cpu);
 	}
 
+	/* Special case: Guest is 'dead' but wants a reboot. */
 	if (cpu->lg->dead == ERR_PTR(-ERESTART))
 		return -ERESTART;
+
 	/* The Guest is dead => "No such file or directory" */
 	return -ENOENT;
 }
diff --git a/drivers/lguest/hypercalls.c b/drivers/lguest/hypercalls.c
index 0f2cb4fd7c69..54d66f05fefa 100644
--- a/drivers/lguest/hypercalls.c
+++ b/drivers/lguest/hypercalls.c
@@ -29,7 +29,7 @@
 #include "lg.h"
 
 /*H:120 This is the core hypercall routine: where the Guest gets what it wants.
- * Or gets killed.  Or, in the case of LHCALL_CRASH, both. */
+ * Or gets killed.  Or, in the case of LHCALL_SHUTDOWN, both. */
 static void do_hcall(struct lg_cpu *cpu, struct hcall_args *args)
 {
 	switch (args->arg0) {
@@ -190,6 +190,13 @@ static void initialize(struct lg_cpu *cpu)
 	 * pagetable. */
 	guest_pagetable_clear_all(cpu);
 }
+/*:*/
+
+/*M:013 If a Guest reads from a page (so creates a mapping) that it has never
+ * written to, and then the Launcher writes to it (ie. the output of a virtual
+ * device), the Guest will still see the old page.  In practice, this never
+ * happens: why would the Guest read a page which it has never written to?  But
+ * a similar scenario might one day bite us, so it's worth mentioning. :*/
 
 /*H:100
  * Hypercalls
@@ -227,7 +234,7 @@ void do_hypercalls(struct lg_cpu *cpu)
 		 * However, if we are signalled or the Guest sends I/O to the
 		 * Launcher, the run_guest() loop will exit without running the
 		 * Guest.  When it comes back it would try to re-run the
-		 * hypercall. */
+		 * hypercall.  Finding that bug sucked. */
 		cpu->hcall = NULL;
 	}
 }
diff --git a/drivers/lguest/interrupts_and_traps.c b/drivers/lguest/interrupts_and_traps.c
index 32e97c1858e5..0414ddf87587 100644
--- a/drivers/lguest/interrupts_and_traps.c
+++ b/drivers/lguest/interrupts_and_traps.c
@@ -144,7 +144,6 @@ void maybe_do_interrupt(struct lg_cpu *cpu)
 	if (copy_from_user(&blk, cpu->lg->lguest_data->blocked_interrupts,
 			   sizeof(blk)))
 		return;
-
 	bitmap_andnot(blk, cpu->irqs_pending, blk, LGUEST_IRQS);
 
 	/* Find the first interrupt. */
@@ -237,9 +236,9 @@ void free_interrupts(void)
 		clear_bit(syscall_vector, used_vectors);
 }
 
-/*H:220 Now we've got the routines to deliver interrupts, delivering traps
- * like page fault is easy.  The only trick is that Intel decided that some
- * traps should have error codes: */
+/*H:220 Now we've got the routines to deliver interrupts, delivering traps like
+ * page fault is easy.  The only trick is that Intel decided that some traps
+ * should have error codes: */
 static int has_err(unsigned int trap)
 {
 	return (trap == 8 || (trap >= 10 && trap <= 14) || trap == 17);
diff --git a/drivers/lguest/lguest_device.c b/drivers/lguest/lguest_device.c
index 1b2ec0bf5eb1..2bc9bf7e88e5 100644
--- a/drivers/lguest/lguest_device.c
+++ b/drivers/lguest/lguest_device.c
@@ -1,10 +1,10 @@
 /*P:050 Lguest guests use a very simple method to describe devices.  It's a
- * series of device descriptors contained just above the top of normal
+ * series of device descriptors contained just above the top of normal Guest
  * memory.
  *
  * We use the standard "virtio" device infrastructure, which provides us with a
  * console, a network and a block driver.  Each one expects some configuration
- * information and a "virtqueue" mechanism to send and receive data. :*/
+ * information and a "virtqueue" or two to send and receive data. :*/
 #include <linux/init.h>
 #include <linux/bootmem.h>
 #include <linux/lguest_launcher.h>
@@ -53,7 +53,7 @@ struct lguest_device {
  * Device configurations
  *
  * The configuration information for a device consists of one or more
- * virtqueues, a feature bitmaks, and some configuration bytes.  The
+ * virtqueues, a feature bitmap, and some configuration bytes.  The
  * configuration bytes don't really matter to us: the Launcher sets them up, and
  * the driver will look at them during setup.
  *
@@ -179,7 +179,7 @@ struct lguest_vq_info
 };
 
 /* When the virtio_ring code wants to prod the Host, it calls us here and we
- * make a hypercall.  We hand the page number of the virtqueue so the Host
+ * make a hypercall.  We hand the physical address of the virtqueue so the Host
  * knows which virtqueue we're talking about. */
 static void lg_notify(struct virtqueue *vq)
 {
@@ -199,7 +199,8 @@ static void lg_notify(struct virtqueue *vq)
  * allocate its own pages and tell the Host where they are, but for lguest it's
  * simpler for the Host to simply tell us where the pages are.
  *
- * So we provide devices with a "find virtqueue and set it up" function. */
+ * So we provide drivers with a "find the Nth virtqueue and set it up"
+ * function. */
 static struct virtqueue *lg_find_vq(struct virtio_device *vdev,
 				    unsigned index,
 				    void (*callback)(struct virtqueue *vq))
diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c
index 2221485b0773..564e425d71dd 100644
--- a/drivers/lguest/lguest_user.c
+++ b/drivers/lguest/lguest_user.c
@@ -73,7 +73,7 @@ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o)
 	if (current != cpu->tsk)
 		return -EPERM;
 
-	/* If the guest is already dead, we indicate why */
+	/* If the Guest is already dead, we indicate why */
 	if (lg->dead) {
 		size_t len;
 
@@ -88,7 +88,7 @@ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o)
 		return len;
 	}
 
-	/* If we returned from read() last time because the Guest notified,
+	/* If we returned from read() last time because the Guest sent I/O,
 	 * clear the flag. */
 	if (cpu->pending_notify)
 		cpu->pending_notify = 0;
@@ -97,14 +97,20 @@ static ssize_t read(struct file *file, char __user *user, size_t size,loff_t*o)
 	return run_guest(cpu, (unsigned long __user *)user);
 }
 
+/*L:025 This actually initializes a CPU.  For the moment, a Guest is only
+ * uniprocessor, so "id" is always 0. */
 static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip)
 {
+	/* We have a limited number the number of CPUs in the lguest struct. */
 	if (id >= NR_CPUS)
 		return -EINVAL;
 
+	/* Set up this CPU's id, and pointer back to the lguest struct. */
 	cpu->id = id;
 	cpu->lg = container_of((cpu - id), struct lguest, cpus[0]);
 	cpu->lg->nr_cpus++;
+
+	/* Each CPU has a timer it can set. */
 	init_clockdev(cpu);
 
 	/* We need a complete page for the Guest registers: they are accessible
@@ -120,11 +126,11 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip)
 	 * address. */
 	lguest_arch_setup_regs(cpu, start_ip);
 
-	/* Initialize the queue for the waker to wait on */
+	/* Initialize the queue for the Waker to wait on */
 	init_waitqueue_head(&cpu->break_wq);
 
 	/* We keep a pointer to the Launcher task (ie. current task) for when
-	 * other Guests want to wake this one (inter-Guest I/O). */
+	 * other Guests want to wake this one (eg. console input). */
 	cpu->tsk = current;
 
 	/* We need to keep a pointer to the Launcher's memory map, because if
@@ -136,6 +142,7 @@ static int lg_cpu_start(struct lg_cpu *cpu, unsigned id, unsigned long start_ip)
 	 * when the same Guest runs on the same CPU twice. */
 	cpu->last_pages = NULL;
 
+	/* No error == success. */
 	return 0;
 }
 
@@ -185,14 +192,13 @@ static int initialize(struct file *file, const unsigned long __user *input)
 	lg->mem_base = (void __user *)(long)args[0];
 	lg->pfn_limit = args[1];
 
-	/* This is the first cpu */
+	/* This is the first cpu (cpu 0) and it will start booting at args[3] */
 	err = lg_cpu_start(&lg->cpus[0], 0, args[3]);
 	if (err)
 		goto release_guest;
 
 	/* Initialize the Guest's shadow page tables, using the toplevel
-	 * address the Launcher gave us.  This allocates memory, so can
-	 * fail. */
+	 * address the Launcher gave us.  This allocates memory, so can fail. */
 	err = init_guest_pagetable(lg, args[2]);
 	if (err)
 		goto free_regs;
@@ -218,11 +224,16 @@ unlock:
 /*L:010 The first operation the Launcher does must be a write.  All writes
  * start with an unsigned long number: for the first write this must be
  * LHREQ_INITIALIZE to set up the Guest.  After that the Launcher can use
- * writes of other values to send interrupts. */
+ * writes of other values to send interrupts.
+ *
+ * Note that we overload the "offset" in the /dev/lguest file to indicate what
+ * CPU number we're dealing with.  Currently this is always 0, since we only
+ * support uniprocessor Guests, but you can see the beginnings of SMP support
+ * here. */
 static ssize_t write(struct file *file, const char __user *in,
 		     size_t size, loff_t *off)
 {
-	/* Once the guest is initialized, we hold the "struct lguest" in the
+	/* Once the Guest is initialized, we hold the "struct lguest" in the
 	 * file private data. */
 	struct lguest *lg = file->private_data;
 	const unsigned long __user *input = (const unsigned long __user *)in;
@@ -230,6 +241,7 @@ static ssize_t write(struct file *file, const char __user *in,
 	struct lg_cpu *uninitialized_var(cpu);
 	unsigned int cpu_id = *off;
 
+	/* The first value tells us what this request is. */
 	if (get_user(req, input) != 0)
 		return -EFAULT;
 	input++;
diff --git a/drivers/lguest/page_tables.c b/drivers/lguest/page_tables.c
index a7f64a9d67e0..d93500f24fbb 100644
--- a/drivers/lguest/page_tables.c
+++ b/drivers/lguest/page_tables.c
@@ -2,8 +2,8 @@
  * previous encounters.  It's functional, and as neat as it can be in the
  * circumstances, but be wary, for these things are subtle and break easily.
  * The Guest provides a virtual to physical mapping, but we can neither trust
- * it nor use it: we verify and convert it here to point the hardware to the
- * actual Guest pages when running the Guest. :*/
+ * it nor use it: we verify and convert it here then point the CPU to the
+ * converted Guest pages when running the Guest. :*/
 
 /* Copyright (C) Rusty Russell IBM Corporation 2006.
  * GPL v2 and any later version */
@@ -106,6 +106,11 @@ static unsigned long gpte_addr(pgd_t gpgd, unsigned long vaddr)
 	BUG_ON(!(pgd_flags(gpgd) & _PAGE_PRESENT));
 	return gpage + ((vaddr>>PAGE_SHIFT) % PTRS_PER_PTE) * sizeof(pte_t);
 }
+/*:*/
+
+/*M:014 get_pfn is slow; it takes the mmap sem and calls get_user_pages.  We
+ * could probably try to grab batches of pages here as an optimization
+ * (ie. pre-faulting). :*/
 
 /*H:350 This routine takes a page number given by the Guest and converts it to
  * an actual, physical page number.  It can fail for several reasons: the
@@ -113,8 +118,8 @@ static unsigned long gpte_addr(pgd_t gpgd, unsigned long vaddr)
  * and the page is read-only, or the write flag was set and the page was
  * shared so had to be copied, but we ran out of memory.
  *
- * This holds a reference to the page, so release_pte() is careful to
- * put that back. */
+ * This holds a reference to the page, so release_pte() is careful to put that
+ * back. */
 static unsigned long get_pfn(unsigned long virtpfn, int write)
 {
 	struct page *page;
@@ -532,13 +537,13 @@ static void do_set_pte(struct lg_cpu *cpu, int idx,
  * all processes.  So when the page table above that address changes, we update
  * all the page tables, not just the current one.  This is rare.
  *
- * The benefit is that when we have to track a new page table, we can copy keep
- * all the kernel mappings.  This speeds up context switch immensely. */
+ * The benefit is that when we have to track a new page table, we can keep all
+ * the kernel mappings.  This speeds up context switch immensely. */
 void guest_set_pte(struct lg_cpu *cpu,
 		   unsigned long gpgdir, unsigned long vaddr, pte_t gpte)
 {
-	/* Kernel mappings must be changed on all top levels.  Slow, but
-	 * doesn't happen often. */
+	/* Kernel mappings must be changed on all top levels.  Slow, but doesn't
+	 * happen often. */
 	if (vaddr >= cpu->lg->kernel_address) {
 		unsigned int i;
 		for (i = 0; i < ARRAY_SIZE(cpu->lg->pgdirs); i++)
@@ -704,12 +709,11 @@ static __init void populate_switcher_pte_page(unsigned int cpu,
 /* We've made it through the page table code.  Perhaps our tired brains are
  * still processing the details, or perhaps we're simply glad it's over.
  *
- * If nothing else, note that all this complexity in juggling shadow page
- * tables in sync with the Guest's page tables is for one reason: for most
- * Guests this page table dance determines how bad performance will be.  This
- * is why Xen uses exotic direct Guest pagetable manipulation, and why both
- * Intel and AMD have implemented shadow page table support directly into
- * hardware.
+ * If nothing else, note that all this complexity in juggling shadow page tables
+ * in sync with the Guest's page tables is for one reason: for most Guests this
+ * page table dance determines how bad performance will be.  This is why Xen
+ * uses exotic direct Guest pagetable manipulation, and why both Intel and AMD
+ * have implemented shadow page table support directly into hardware.
  *
  * There is just one file remaining in the Host. */
 
diff --git a/drivers/lguest/x86/core.c b/drivers/lguest/x86/core.c
index 635187812d52..5126d5d9ea0e 100644
--- a/drivers/lguest/x86/core.c
+++ b/drivers/lguest/x86/core.c
@@ -17,6 +17,13 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
+/*P:450 This file contains the x86-specific lguest code.  It used to be all
+ * mixed in with drivers/lguest/core.c but several foolhardy code slashers
+ * wrestled most of the dependencies out to here in preparation for porting
+ * lguest to other architectures (see what I mean by foolhardy?).
+ *
+ * This also contains a couple of non-obvious setup and teardown pieces which
+ * were implemented after days of debugging pain. :*/
 #include <linux/kernel.h>
 #include <linux/start_kernel.h>
 #include <linux/string.h>
@@ -157,6 +164,8 @@ static void run_guest_once(struct lg_cpu *cpu, struct lguest_pages *pages)
  * also simplify copy_in_guest_info().  Note that we'd still need to restore
  * things when we exit to Launcher userspace, but that's fairly easy.
  *
+ * We could also try using this hooks for PGE, but that might be too expensive.
+ *
  * The hooks were designed for KVM, but we can also put them to good use. :*/
 
 /*H:040 This is the i386-specific code to setup and run the Guest.  Interrupts
@@ -182,7 +191,7 @@ void lguest_arch_run_guest(struct lg_cpu *cpu)
 	 * was doing. */
 	run_guest_once(cpu, lguest_pages(raw_smp_processor_id()));
 
-	/* Note that the "regs" pointer contains two extra entries which are
+	/* Note that the "regs" structure contains two extra entries which are
 	 * not really registers: a trap number which says what interrupt or
 	 * trap made the switcher code come back, and an error code which some
 	 * traps set.  */
@@ -293,11 +302,10 @@ void lguest_arch_handle_trap(struct lg_cpu *cpu)
 		break;
 	case 14: /* We've intercepted a Page Fault. */
 		/* The Guest accessed a virtual address that wasn't mapped.
-		 * This happens a lot: we don't actually set up most of the
-		 * page tables for the Guest at all when we start: as it runs
-		 * it asks for more and more, and we set them up as
-		 * required. In this case, we don't even tell the Guest that
-		 * the fault happened.
+		 * This happens a lot: we don't actually set up most of the page
+		 * tables for the Guest at all when we start: as it runs it asks
+		 * for more and more, and we set them up as required. In this
+		 * case, we don't even tell the Guest that the fault happened.
 		 *
 		 * The errcode tells whether this was a read or a write, and
 		 * whether kernel or userspace code. */
@@ -342,7 +350,7 @@ void lguest_arch_handle_trap(struct lg_cpu *cpu)
 	if (!deliver_trap(cpu, cpu->regs->trapnum))
 		/* If the Guest doesn't have a handler (either it hasn't
 		 * registered any yet, or it's one of the faults we don't let
-		 * it handle), it dies with a cryptic error message. */
+		 * it handle), it dies with this cryptic error message. */
 		kill_guest(cpu, "unhandled trap %li at %#lx (%#lx)",
 			   cpu->regs->trapnum, cpu->regs->eip,
 			   cpu->regs->trapnum == 14 ? cpu->arch.last_pagefault
@@ -375,8 +383,8 @@ void __init lguest_arch_host_init(void)
 	 * The only exception is the interrupt handlers in switcher.S: their
 	 * addresses are placed in a table (default_idt_entries), so we need to
 	 * update the table with the new addresses.  switcher_offset() is a
-	 * convenience function which returns the distance between the builtin
-	 * switcher code and the high-mapped copy we just made. */
+	 * convenience function which returns the distance between the
+	 * compiled-in switcher code and the high-mapped copy we just made. */
 	for (i = 0; i < IDT_ENTRIES; i++)
 		default_idt_entries[i] += switcher_offset();
 
@@ -416,7 +424,7 @@ void __init lguest_arch_host_init(void)
 		state->guest_gdt_desc.address = (long)&state->guest_gdt;
 
 		/* We know where we want the stack to be when the Guest enters
-		 * the switcher: in pages->regs.  The stack grows upwards, so
+		 * the Switcher: in pages->regs.  The stack grows upwards, so
 		 * we start it at the end of that structure. */
 		state->guest_tss.sp0 = (long)(&pages->regs + 1);
 		/* And this is the GDT entry to use for the stack: we keep a
@@ -513,8 +521,8 @@ int lguest_arch_init_hypercalls(struct lg_cpu *cpu)
 {
 	u32 tsc_speed;
 
-	/* The pointer to the Guest's "struct lguest_data" is the only
-	 * argument.  We check that address now. */
+	/* The pointer to the Guest's "struct lguest_data" is the only argument.
+	 * We check that address now. */
 	if (!lguest_address_ok(cpu->lg, cpu->hcall->arg1,
 			       sizeof(*cpu->lg->lguest_data)))
 		return -EFAULT;
@@ -546,6 +554,7 @@ int lguest_arch_init_hypercalls(struct lg_cpu *cpu)
 
 	return 0;
 }
+/*:*/
 
 /*L:030 lguest_arch_setup_regs()
  *
diff --git a/drivers/lguest/x86/switcher_32.S b/drivers/lguest/x86/switcher_32.S
index 0af8baaa0d4a..3fc15318a80f 100644
--- a/drivers/lguest/x86/switcher_32.S
+++ b/drivers/lguest/x86/switcher_32.S
@@ -1,6 +1,6 @@
-/*P:900 This is the Switcher: code which sits at 0xFFC00000 to do the low-level
- * Guest<->Host switch.  It is as simple as it can be made, but it's naturally
- * very specific to x86.
+/*P:900 This is the Switcher: code which sits at 0xFFC00000 astride both the
+ * Host and Guest to do the low-level Guest<->Host switch.  It is as simple as
+ * it can be made, but it's naturally very specific to x86.
  *
  * You have now completed Preparation.  If this has whet your appetite; if you
  * are feeling invigorated and refreshed then the next, more challenging stage
@@ -189,7 +189,7 @@ ENTRY(switch_to_guest)
 	// Interrupts are turned back on: we are Guest.
 	iret
 
-// We treat two paths to switch back to the Host
+// We tread two paths to switch back to the Host
 // Yet both must save Guest state and restore Host
 // So we put the routine in a macro.
 #define SWITCH_TO_HOST							\
diff --git a/include/asm-x86/lguest_hcall.h b/include/asm-x86/lguest_hcall.h
index 758b9a5d4539..f239e7069cab 100644
--- a/include/asm-x86/lguest_hcall.h
+++ b/include/asm-x86/lguest_hcall.h
@@ -27,7 +27,7 @@
 #ifndef __ASSEMBLY__
 #include <asm/hw_irq.h>
 
-/*G:031 First, how does our Guest contact the Host to ask for privileged
+/*G:031 But first, how does our Guest contact the Host to ask for privileged
  * operations?  There are two ways: the direct way is to make a "hypercall",
  * to make requests of the Host Itself.
  *
diff --git a/include/linux/lguest_launcher.h b/include/linux/lguest_launcher.h
index 589be3e1f3ac..e7217dc58f39 100644
--- a/include/linux/lguest_launcher.h
+++ b/include/linux/lguest_launcher.h
@@ -16,6 +16,10 @@
  * a new device, we simply need to write a new virtio driver and create support
  * for it in the Launcher: this code won't need to change.
  *
+ * Virtio devices are also used by kvm, so we can simply reuse their optimized
+ * device drivers.  And one day when everyone uses virtio, my plan will be
+ * complete.  Bwahahahah!
+ *
  * Devices are described by a simplified ID, a status byte, and some "config"
  * bytes which describe this device's configuration.  This is placed by the
  * Launcher just above the top of physical memory:
@@ -26,7 +30,7 @@ struct lguest_device_desc {
 	/* The number of virtqueues (first in config array) */
 	__u8 num_vq;
 	/* The number of bytes of feature bits.  Multiply by 2: one for host
-	 * features and one for guest acknowledgements. */
+	 * features and one for Guest acknowledgements. */
 	__u8 feature_len;
 	/* The number of bytes of the config array after virtqueues. */
 	__u8 config_len;
-- 
cgit v1.2.3


From 50fd4407b8bfbde7c1a0bfe4f24de7df37164342 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Thu, 27 Mar 2008 17:42:50 -0700
Subject: [NET]: Use local_irq_{save,restore}() in napi_complete().

Based upon a lockdep report.

Since ->poll() can be invoked from netpoll with interrupts
disabled, we must not unconditionally enable interrupts
in napi_complete().

Instead we must use local_irq_{save,restore}().

Noticed by Peter Zijlstra:

<irqs disabled>

  netpoll_poll()
    poll_napi()
      spin_trylock(&napi->poll_lock)
      poll_one_napi()
        napi->poll() := sky2_poll()
          napi_complete()
            local_irq_disable()
            local_irq_enable() <--- *BUG*

  <irq>
    irq_exit()
      do_softirq()
        net_rx_action()
          spin_lock(&napi->poll_lock) <--- Deadlock!

Because we still hold the lock....

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index a2f003239c85..fae6a7ececdb 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -383,9 +383,11 @@ static inline void __napi_complete(struct napi_struct *n)
 
 static inline void napi_complete(struct napi_struct *n)
 {
-	local_irq_disable();
+	unsigned long flags;
+
+	local_irq_save(flags);
 	__napi_complete(n);
-	local_irq_enable();
+	local_irq_restore(flags);
 }
 
 /**
-- 
cgit v1.2.3


From 0dde3e16485dca16eb682dd59da1a598bf62e284 Mon Sep 17 00:00:00 2001
From: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Date: Thu, 27 Mar 2008 17:43:41 -0700
Subject: [NET]: uninline skb_put, de-bloats a lot
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allyesconfig (v2.6.24-mm1):

~500 files changed
...
 869 funcs, 198 +, 111003 -, diff: -110805 --- skb_put
  skb_put                       | +104

Without number of debug related CONFIGs (v2.6.25-rc2-mm1):

-60744  855 funcs, 861 +, 61605 -, diff: -60744 --- skb_put
  skb_put                       |  +57

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 21 +--------------------
 net/core/skbuff.c      | 21 +++++++++++++++++++++
 2 files changed, 22 insertions(+), 20 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 7beb239d2ee0..f085955cb5a7 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -892,6 +892,7 @@ static inline void skb_set_tail_pointer(struct sk_buff *skb, const int offset)
 /*
  *	Add data to an sk_buff
  */
+extern unsigned char *skb_put(struct sk_buff *skb, unsigned int len);
 static inline unsigned char *__skb_put(struct sk_buff *skb, unsigned int len)
 {
 	unsigned char *tmp = skb_tail_pointer(skb);
@@ -901,26 +902,6 @@ static inline unsigned char *__skb_put(struct sk_buff *skb, unsigned int len)
 	return tmp;
 }
 
-/**
- *	skb_put - add data to a buffer
- *	@skb: buffer to use
- *	@len: amount of data to add
- *
- *	This function extends the used data area of the buffer. If this would
- *	exceed the total buffer size the kernel will panic. A pointer to the
- *	first byte of the extra data is returned.
- */
-static inline unsigned char *skb_put(struct sk_buff *skb, unsigned int len)
-{
-	unsigned char *tmp = skb_tail_pointer(skb);
-	SKB_LINEAR_ASSERT(skb);
-	skb->tail += len;
-	skb->len  += len;
-	if (unlikely(skb->tail > skb->end))
-		skb_over_panic(skb, len, current_text_addr());
-	return tmp;
-}
-
 static inline unsigned char *__skb_push(struct sk_buff *skb, unsigned int len)
 {
 	skb->data -= len;
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 0d0fd28a9041..3402eca768f8 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -857,6 +857,27 @@ free_skb:
 	return err;
 }
 
+/**
+ *	skb_put - add data to a buffer
+ *	@skb: buffer to use
+ *	@len: amount of data to add
+ *
+ *	This function extends the used data area of the buffer. If this would
+ *	exceed the total buffer size the kernel will panic. A pointer to the
+ *	first byte of the extra data is returned.
+ */
+unsigned char *skb_put(struct sk_buff *skb, unsigned int len)
+{
+	unsigned char *tmp = skb_tail_pointer(skb);
+	SKB_LINEAR_ASSERT(skb);
+	skb->tail += len;
+	skb->len  += len;
+	if (unlikely(skb->tail > skb->end))
+		skb_over_panic(skb, len, __builtin_return_address(0));
+	return tmp;
+}
+EXPORT_SYMBOL(skb_put);
+
 /* Trims skb to length len. It can change skb pointers.
  */
 
-- 
cgit v1.2.3


From 6be8ac2fdc5e69dec53913a42312a92dbfbd4907 Mon Sep 17 00:00:00 2001
From: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Date: Thu, 27 Mar 2008 17:47:24 -0700
Subject: [NET]: uninline skb_pull, de-bloats a lot
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allyesconfig (v2.6.24-mm1):

-28162  354 funcs, 3005 +, 31167 -, diff: -28162 --- skb_pull

Without number of debug related CONFIGs (v2.6.25-rc2-mm1):

-9697  338 funcs, 221 +, 9918 -, diff: -9697 --- skb_pull
skb_pull                      |  +44

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 16 +---------------
 net/core/skbuff.c      | 16 ++++++++++++++++
 2 files changed, 17 insertions(+), 15 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index f085955cb5a7..6d6cde7b243c 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -927,6 +927,7 @@ static inline unsigned char *skb_push(struct sk_buff *skb, unsigned int len)
 	return skb->data;
 }
 
+extern unsigned char *skb_pull(struct sk_buff *skb, unsigned int len);
 static inline unsigned char *__skb_pull(struct sk_buff *skb, unsigned int len)
 {
 	skb->len -= len;
@@ -934,21 +935,6 @@ static inline unsigned char *__skb_pull(struct sk_buff *skb, unsigned int len)
 	return skb->data += len;
 }
 
-/**
- *	skb_pull - remove data from the start of a buffer
- *	@skb: buffer to use
- *	@len: amount of data to remove
- *
- *	This function removes data from the start of a buffer, returning
- *	the memory to the headroom. A pointer to the next data in the buffer
- *	is returned. Once the data has been pulled future pushes will overwrite
- *	the old data.
- */
-static inline unsigned char *skb_pull(struct sk_buff *skb, unsigned int len)
-{
-	return unlikely(len > skb->len) ? NULL : __skb_pull(skb, len);
-}
-
 extern unsigned char *__pskb_pull_tail(struct sk_buff *skb, int delta);
 
 static inline unsigned char *__pskb_pull(struct sk_buff *skb, unsigned int len)
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 3402eca768f8..cf489b6329e8 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -878,6 +878,22 @@ unsigned char *skb_put(struct sk_buff *skb, unsigned int len)
 }
 EXPORT_SYMBOL(skb_put);
 
+/**
+ *	skb_pull - remove data from the start of a buffer
+ *	@skb: buffer to use
+ *	@len: amount of data to remove
+ *
+ *	This function removes data from the start of a buffer, returning
+ *	the memory to the headroom. A pointer to the next data in the buffer
+ *	is returned. Once the data has been pulled future pushes will overwrite
+ *	the old data.
+ */
+unsigned char *skb_pull(struct sk_buff *skb, unsigned int len)
+{
+	return unlikely(len > skb->len) ? NULL : __skb_pull(skb, len);
+}
+EXPORT_SYMBOL(skb_pull);
+
 /* Trims skb to length len. It can change skb pointers.
  */
 
-- 
cgit v1.2.3


From 7c4b93d8269b9d35971a8239426b1f6ddc3d5ef7 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 21 Mar 2008 23:59:49 -0400
Subject: [PATCH] count ghost references to vfsmounts

make propagate_mount_busy() exclude references from the vfsmounts
that had been isolated by umount_tree() and are just waiting for
release_mounts() to dispose of their ->mnt_parent/->mnt_mountpoint.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namespace.c        | 5 ++++-
 fs/pnode.c            | 2 +-
 include/linux/mount.h | 1 +
 3 files changed, 6 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/fs/namespace.c b/fs/namespace.c
index 6324dfc80dc6..c175218ebae1 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -548,6 +548,7 @@ void release_mounts(struct list_head *head)
 			m = mnt->mnt_parent;
 			mnt->mnt_mountpoint = mnt->mnt_root;
 			mnt->mnt_parent = mnt;
+			m->mnt_ghosts--;
 			spin_unlock(&vfsmount_lock);
 			dput(dentry);
 			mntput(m);
@@ -572,8 +573,10 @@ void umount_tree(struct vfsmount *mnt, int propagate, struct list_head *kill)
 		__touch_mnt_namespace(p->mnt_ns);
 		p->mnt_ns = NULL;
 		list_del_init(&p->mnt_child);
-		if (p->mnt_parent != p)
+		if (p->mnt_parent != p) {
+			p->mnt_parent->mnt_ghosts++;
 			p->mnt_mountpoint->d_mounted--;
+		}
 		change_mnt_propagation(p, MS_PRIVATE);
 	}
 }
diff --git a/fs/pnode.c b/fs/pnode.c
index 05ba692bc540..1d8f5447f3f7 100644
--- a/fs/pnode.c
+++ b/fs/pnode.c
@@ -225,7 +225,7 @@ out:
  */
 static inline int do_refcount_check(struct vfsmount *mnt, int count)
 {
-	int mycount = atomic_read(&mnt->mnt_count);
+	int mycount = atomic_read(&mnt->mnt_count) - mnt->mnt_ghosts;
 	return (mycount > count);
 }
 
diff --git a/include/linux/mount.h b/include/linux/mount.h
index 6d3047d8c91c..dac5e67ff3ee 100644
--- a/include/linux/mount.h
+++ b/include/linux/mount.h
@@ -61,6 +61,7 @@ struct vfsmount {
 	atomic_t mnt_count;
 	int mnt_expiry_mark;		/* true if marked for expiry */
 	int mnt_pinned;
+	int mnt_ghosts;
 };
 
 static inline struct vfsmount *mntget(struct vfsmount *mnt)
-- 
cgit v1.2.3


From c35038becad0adb0e25261fff66d85b1a6ddd0c2 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 22 Mar 2008 00:46:23 -0400
Subject: [PATCH] do shrink_submounts() for all fs types

... and take it out of ->umount_begin() instances.  Call with all locks
already taken (by do_umount()) and leave calling release_mounts() to
caller (it will do release_mounts() anyway, so we can just put into
the same list).

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/afs/internal.h      |  1 -
 fs/afs/mntpt.c         |  8 --------
 fs/afs/super.c         |  1 -
 fs/cifs/cifs_dfs_ref.c |  1 -
 fs/namespace.c         | 23 ++++++++++-------------
 fs/nfs/super.c         |  2 --
 include/linux/mount.h  |  1 -
 7 files changed, 10 insertions(+), 27 deletions(-)

(limited to 'include/linux')

diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index 5ca3625cd39e..9ba16edc0af2 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -573,7 +573,6 @@ extern const struct file_operations afs_mntpt_file_operations;
 
 extern int afs_mntpt_check_symlink(struct afs_vnode *, struct key *);
 extern void afs_mntpt_kill_timer(void);
-extern void afs_umount_begin(struct vfsmount *, int);
 
 /*
  * proc.c
diff --git a/fs/afs/mntpt.c b/fs/afs/mntpt.c
index a3510b8ba3e7..2f5503902c37 100644
--- a/fs/afs/mntpt.c
+++ b/fs/afs/mntpt.c
@@ -283,11 +283,3 @@ void afs_mntpt_kill_timer(void)
 	cancel_delayed_work(&afs_mntpt_expiry_timer);
 	flush_scheduled_work();
 }
-
-/*
- * begin unmount by attempting to remove all automounted mountpoints we added
- */
-void afs_umount_begin(struct vfsmount *vfsmnt, int flags)
-{
-	shrink_submounts(vfsmnt, &afs_vfsmounts);
-}
diff --git a/fs/afs/super.c b/fs/afs/super.c
index 36bbce45f44b..4b572b801d8d 100644
--- a/fs/afs/super.c
+++ b/fs/afs/super.c
@@ -50,7 +50,6 @@ static const struct super_operations afs_super_ops = {
 	.write_inode	= afs_write_inode,
 	.destroy_inode	= afs_destroy_inode,
 	.clear_inode	= afs_clear_inode,
-	.umount_begin	= afs_umount_begin,
 	.put_super	= afs_put_super,
 	.show_options	= generic_show_options,
 };
diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c
index a1a95b027136..56c924033b78 100644
--- a/fs/cifs/cifs_dfs_ref.c
+++ b/fs/cifs/cifs_dfs_ref.c
@@ -33,7 +33,6 @@ void dfs_shrink_umount_helper(struct vfsmount *vfsmnt)
 {
 	mark_mounts_for_expiry(&cifs_dfs_automount_list);
 	mark_mounts_for_expiry(&cifs_dfs_automount_list);
-	shrink_submounts(vfsmnt, &cifs_dfs_automount_list);
 }
 
 /**
diff --git a/fs/namespace.c b/fs/namespace.c
index 1c78917ec930..7bd74b25930c 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -581,6 +581,8 @@ void umount_tree(struct vfsmount *mnt, int propagate, struct list_head *kill)
 	}
 }
 
+static void shrink_submounts(struct vfsmount *mnt, struct list_head *umounts);
+
 static int do_umount(struct vfsmount *mnt, int flags)
 {
 	struct super_block *sb = mnt->mnt_sb;
@@ -653,6 +655,9 @@ static int do_umount(struct vfsmount *mnt, int flags)
 	spin_lock(&vfsmount_lock);
 	event++;
 
+	if (!(flags & MNT_DETACH))
+		shrink_submounts(mnt, &umount_list);
+
 	retval = -EBUSY;
 	if (flags & MNT_DETACH || !propagate_mount_busy(mnt, 2)) {
 		if (!list_empty(&mnt->mnt_list))
@@ -1302,30 +1307,22 @@ resume:
  * process a list of expirable mountpoints with the intent of discarding any
  * submounts of a specific parent mountpoint
  */
-void shrink_submounts(struct vfsmount *mountpoint, struct list_head *mounts)
+static void shrink_submounts(struct vfsmount *mnt, struct list_head *umounts)
 {
 	LIST_HEAD(graveyard);
-	LIST_HEAD(umounts);
-	struct vfsmount *mnt;
+	struct vfsmount *m;
 
-	down_write(&namespace_sem);
-	spin_lock(&vfsmount_lock);
 	/* extract submounts of 'mountpoint' from the expiration list */
-	while (select_submounts(mountpoint, &graveyard)) {
+	while (select_submounts(mnt, &graveyard)) {
 		while (!list_empty(&graveyard)) {
-			mnt = list_first_entry(&graveyard, struct vfsmount,
+			m = list_first_entry(&graveyard, struct vfsmount,
 						mnt_expire);
 			touch_mnt_namespace(mnt->mnt_ns);
-			umount_tree(mnt, 1, &umounts);
+			umount_tree(mnt, 1, umounts);
 		}
 	}
-	spin_unlock(&vfsmount_lock);
-	up_write(&namespace_sem);
-	release_mounts(&umounts);
 }
 
-EXPORT_SYMBOL_GPL(shrink_submounts);
-
 /*
  * Some copy_from_user() implementations do not return the exact number of
  * bytes remaining to copy on a fault.  But copy_mount_options() requires that.
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index dd4dfcd632ec..f9219024f31a 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -589,8 +589,6 @@ static void nfs_umount_begin(struct vfsmount *vfsmnt, int flags)
 	struct nfs_server *server = NFS_SB(vfsmnt->mnt_sb);
 	struct rpc_clnt *rpc;
 
-	shrink_submounts(vfsmnt, &nfs_automount_list);
-
 	if (!(flags & MNT_FORCE))
 		return;
 	/* -EIO all pending I/O */
diff --git a/include/linux/mount.h b/include/linux/mount.h
index dac5e67ff3ee..5ee2df217cdf 100644
--- a/include/linux/mount.h
+++ b/include/linux/mount.h
@@ -99,7 +99,6 @@ extern int do_add_mount(struct vfsmount *newmnt, struct nameidata *nd,
 			int mnt_flags, struct list_head *fslist);
 
 extern void mark_mounts_for_expiry(struct list_head *mounts);
-extern void shrink_submounts(struct vfsmount *mountpoint, struct list_head *mounts);
 
 extern spinlock_t vfsmount_lock;
 extern dev_t name_to_dev_t(char *name);
-- 
cgit v1.2.3


From f58518e678e5eef430c8d5cdcc7cd28d285f1980 Mon Sep 17 00:00:00 2001
From: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Date: Thu, 27 Mar 2008 17:51:31 -0700
Subject: [NET]: uninline dev_alloc_skb, de-bloats a lot
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allyesconfig (v2.6.24-mm1):

-23668  392 funcs, 104 +, 23772 -, diff: -23668 --- dev_alloc_skb

Without many debug CONFIGs (v2.6.25-rc2-mm1):

-12178  382 funcs, 157 +, 12335 -, diff: -12178 --- dev_alloc_skb
dev_alloc_skb                 |  +37

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 17 +----------------
 net/core/skbuff.c      | 18 ++++++++++++++++++
 2 files changed, 19 insertions(+), 16 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 6d6cde7b243c..01a11b0c0291 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1272,22 +1272,7 @@ static inline struct sk_buff *__dev_alloc_skb(unsigned int length,
 	return skb;
 }
 
-/**
- *	dev_alloc_skb - allocate an skbuff for receiving
- *	@length: length to allocate
- *
- *	Allocate a new &sk_buff and assign it a usage count of one. The
- *	buffer has unspecified headroom built in. Users should allocate
- *	the headroom they think they need without accounting for the
- *	built in space. The built in space is used for optimisations.
- *
- *	%NULL is returned if there is no free memory. Although this function
- *	allocates memory it can be called from an interrupt.
- */
-static inline struct sk_buff *dev_alloc_skb(unsigned int length)
-{
-	return __dev_alloc_skb(length, GFP_ATOMIC);
-}
+extern struct sk_buff *dev_alloc_skb(unsigned int length);
 
 extern struct sk_buff *__netdev_alloc_skb(struct net_device *dev,
 		unsigned int length, gfp_t gfp_mask);
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index cf489b6329e8..0daf5c0e5b8d 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -263,6 +263,24 @@ struct sk_buff *__netdev_alloc_skb(struct net_device *dev,
 	return skb;
 }
 
+/**
+ *	dev_alloc_skb - allocate an skbuff for receiving
+ *	@length: length to allocate
+ *
+ *	Allocate a new &sk_buff and assign it a usage count of one. The
+ *	buffer has unspecified headroom built in. Users should allocate
+ *	the headroom they think they need without accounting for the
+ *	built in space. The built in space is used for optimisations.
+ *
+ *	%NULL is returned if there is no free memory. Although this function
+ *	allocates memory it can be called from an interrupt.
+ */
+struct sk_buff *dev_alloc_skb(unsigned int length)
+{
+	return __dev_alloc_skb(length, GFP_ATOMIC);
+}
+EXPORT_SYMBOL(dev_alloc_skb);
+
 static void skb_drop_list(struct sk_buff **listp)
 {
 	struct sk_buff *list = *listp;
-- 
cgit v1.2.3


From c2aa270ad73d385bd6cdebf5d741bdf18a3e17ad Mon Sep 17 00:00:00 2001
From: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Date: Thu, 27 Mar 2008 17:52:40 -0700
Subject: [NET]: uninline skb_push, de-bloats a lot
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allyesconfig (v2.6.24-mm1):

-21593  356 funcs, 2418 +, 24011 -, diff: -21593 --- skb_push

Without many debug related CONFIGs (v2.6.25-rc2-mm1):

-13890  341 funcs, 189 +, 14079 -, diff: -13890 --- skb_push
skb_push                      |  +46

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 19 +------------------
 net/core/skbuff.c      | 19 +++++++++++++++++++
 2 files changed, 20 insertions(+), 18 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 01a11b0c0291..1baf4d43bb2d 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -902,6 +902,7 @@ static inline unsigned char *__skb_put(struct sk_buff *skb, unsigned int len)
 	return tmp;
 }
 
+extern unsigned char *skb_push(struct sk_buff *skb, unsigned int len);
 static inline unsigned char *__skb_push(struct sk_buff *skb, unsigned int len)
 {
 	skb->data -= len;
@@ -909,24 +910,6 @@ static inline unsigned char *__skb_push(struct sk_buff *skb, unsigned int len)
 	return skb->data;
 }
 
-/**
- *	skb_push - add data to the start of a buffer
- *	@skb: buffer to use
- *	@len: amount of data to add
- *
- *	This function extends the used data area of the buffer at the buffer
- *	start. If this would exceed the total buffer headroom the kernel will
- *	panic. A pointer to the first byte of the extra data is returned.
- */
-static inline unsigned char *skb_push(struct sk_buff *skb, unsigned int len)
-{
-	skb->data -= len;
-	skb->len  += len;
-	if (unlikely(skb->data<skb->head))
-		skb_under_panic(skb, len, current_text_addr());
-	return skb->data;
-}
-
 extern unsigned char *skb_pull(struct sk_buff *skb, unsigned int len);
 static inline unsigned char *__skb_pull(struct sk_buff *skb, unsigned int len)
 {
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 0daf5c0e5b8d..a37127b5899c 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -896,6 +896,25 @@ unsigned char *skb_put(struct sk_buff *skb, unsigned int len)
 }
 EXPORT_SYMBOL(skb_put);
 
+/**
+ *	skb_push - add data to the start of a buffer
+ *	@skb: buffer to use
+ *	@len: amount of data to add
+ *
+ *	This function extends the used data area of the buffer at the buffer
+ *	start. If this would exceed the total buffer headroom the kernel will
+ *	panic. A pointer to the first byte of the extra data is returned.
+ */
+unsigned char *skb_push(struct sk_buff *skb, unsigned int len)
+{
+	skb->data -= len;
+	skb->len  += len;
+	if (unlikely(skb->data<skb->head))
+		skb_under_panic(skb, len, __builtin_return_address(0));
+	return skb->data;
+}
+EXPORT_SYMBOL(skb_push);
+
 /**
  *	skb_pull - remove data from the start of a buffer
  *	@skb: buffer to use
-- 
cgit v1.2.3


From 419ae74ecc9494e58928a5c6652f4c072f3ca744 Mon Sep 17 00:00:00 2001
From: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Date: Thu, 27 Mar 2008 17:54:01 -0700
Subject: [NET]: uninline skb_trim, de-bloats
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Allyesconfig (v2.6.24-mm1):
-10976  209 funcs, 123 +, 11099 -, diff: -10976 --- skb_trim

Without number of debug related CONFIGs (v2.6.25-rc2-mm1):
-7360  192 funcs, 131 +, 7491 -, diff: -7360 --- skb_trim
skb_trim                      |  +42

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 16 +---------------
 net/core/skbuff.c      | 16 ++++++++++++++++
 2 files changed, 17 insertions(+), 15 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 1baf4d43bb2d..ff72145d5d9e 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1158,21 +1158,7 @@ static inline void __skb_trim(struct sk_buff *skb, unsigned int len)
 	skb_set_tail_pointer(skb, len);
 }
 
-/**
- *	skb_trim - remove end from a buffer
- *	@skb: buffer to alter
- *	@len: new length
- *
- *	Cut the length of a buffer down by removing data from the tail. If
- *	the buffer is already under the length specified it is not modified.
- *	The skb must be linear.
- */
-static inline void skb_trim(struct sk_buff *skb, unsigned int len)
-{
-	if (skb->len > len)
-		__skb_trim(skb, len);
-}
-
+extern void skb_trim(struct sk_buff *skb, unsigned int len);
 
 static inline int __pskb_trim(struct sk_buff *skb, unsigned int len)
 {
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index a37127b5899c..86e5682728be 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -931,6 +931,22 @@ unsigned char *skb_pull(struct sk_buff *skb, unsigned int len)
 }
 EXPORT_SYMBOL(skb_pull);
 
+/**
+ *	skb_trim - remove end from a buffer
+ *	@skb: buffer to alter
+ *	@len: new length
+ *
+ *	Cut the length of a buffer down by removing data from the tail. If
+ *	the buffer is already under the length specified it is not modified.
+ *	The skb must be linear.
+ */
+void skb_trim(struct sk_buff *skb, unsigned int len)
+{
+	if (skb->len > len)
+		__skb_trim(skb, len);
+}
+EXPORT_SYMBOL(skb_trim);
+
 /* Trims skb to length len. It can change skb pointers.
  */
 
-- 
cgit v1.2.3


From 8c703d35fa91911dd92a18c31a718853f483ad80 Mon Sep 17 00:00:00 2001
From: Jonathan Corbet <corbet@lwn.net>
Date: Fri, 28 Mar 2008 14:15:49 -0700
Subject: in_atomic(): document why it is unsuitable for general use

Discourage people from inappropriately using in_atomic()

Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/hardirq.h | 7 +++++++
 1 file changed, 7 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h
index 49829988bfa0..897f723bd222 100644
--- a/include/linux/hardirq.h
+++ b/include/linux/hardirq.h
@@ -72,6 +72,13 @@
 #define in_softirq()		(softirq_count())
 #define in_interrupt()		(irq_count())
 
+/*
+ * Are we running in atomic context?  WARNING: this macro cannot
+ * always detect atomic context; in particular, it cannot know about
+ * held spinlocks in non-preemptible kernels.  Thus it should not be
+ * used in the general case to determine whether sleeping is possible.
+ * Do not use in_atomic() in driver code.
+ */
 #define in_atomic()		((preempt_count() & ~PREEMPT_ACTIVE) != 0)
 
 #ifdef CONFIG_PREEMPT
-- 
cgit v1.2.3


From 3afe3925987adc3fc052abe404e44520c2072fc8 Mon Sep 17 00:00:00 2001
From: Harvey Harrison <harvey.harrison@gmail.com>
Date: Fri, 28 Mar 2008 14:16:01 -0700
Subject: kernel: add bit rotation helpers for 16 and 8 bit

Will replace open-coded variants elsewhere.  Done in the same
style as the 32-bit versions.

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: John W. Linville <linville@tuxdriver.com>
Cc: Joe Perches <joe@perches.com>
Cc: Jiri Benc <jbenc@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/bitops.h | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/bitops.h b/include/linux/bitops.h
index 69c1edb9fe54..40d54731de7e 100644
--- a/include/linux/bitops.h
+++ b/include/linux/bitops.h
@@ -65,6 +65,46 @@ static inline __u32 ror32(__u32 word, unsigned int shift)
 	return (word >> shift) | (word << (32 - shift));
 }
 
+/**
+ * rol16 - rotate a 16-bit value left
+ * @word: value to rotate
+ * @shift: bits to roll
+ */
+static inline __u16 rol16(__u16 word, unsigned int shift)
+{
+	return (word << shift) | (word >> (16 - shift));
+}
+
+/**
+ * ror16 - rotate a 16-bit value right
+ * @word: value to rotate
+ * @shift: bits to roll
+ */
+static inline __u16 ror16(__u16 word, unsigned int shift)
+{
+	return (word >> shift) | (word << (16 - shift));
+}
+
+/**
+ * rol8 - rotate an 8-bit value left
+ * @word: value to rotate
+ * @shift: bits to roll
+ */
+static inline __u8 rol8(__u8 word, unsigned int shift)
+{
+	return (word << shift) | (word >> (8 - shift));
+}
+
+/**
+ * ror8 - rotate an 8-bit value right
+ * @word: value to rotate
+ * @shift: bits to roll
+ */
+static inline __u8 ror8(__u8 word, unsigned int shift)
+{
+	return (word >> shift) | (word << (8 - shift));
+}
+
 static inline unsigned fls_long(unsigned long l)
 {
 	if (sizeof(l) == 4)
-- 
cgit v1.2.3


From 5ac7ec85bcc70ef605657fb2d1106d27ab3bd131 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Fri, 28 Mar 2008 14:16:03 -0700
Subject: ext3: don't export ext3_fs.h and jbd.h

Neither of the headers actually compiles when included from userpsace nor
should it be made available as userspace tools should be using the libraries
or at least headers from e2fsprogs.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/Kbuild | 2 --
 1 file changed, 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 4108b38ebb16..4a446a19295e 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -195,7 +195,6 @@ unifdef-y += ethtool.h
 unifdef-y += eventpoll.h
 unifdef-y += signalfd.h
 unifdef-y += ext2_fs.h
-unifdef-y += ext3_fs.h
 unifdef-y += fb.h
 unifdef-y += fcntl.h
 unifdef-y += filter.h
@@ -248,7 +247,6 @@ unifdef-y += isdn.h
 unifdef-y += isdnif.h
 unifdef-y += isdn_divertif.h
 unifdef-y += isdn_ppp.h
-unifdef-y += jbd.h
 unifdef-y += joystick.h
 unifdef-y += kdev_t.h
 unifdef-y += kd.h
-- 
cgit v1.2.3


From 1567ca7eec7664b8be3b07755ac59dc1b1ec76cb Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Fri, 28 Mar 2008 15:53:11 -0700
Subject: [NET]: Protect device namespace inlines with CONFIG_NET

Include sites should not be bothered by whether
CONFIG_NET is set or not when trying to include
benign files like linux/etherdevice.h et al.

From a report by Stephen Rothwell.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 3b54f8a2c055..8576ca928dae 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -741,6 +741,7 @@ struct net_device
 #define	NETDEV_ALIGN		32
 #define	NETDEV_ALIGN_CONST	(NETDEV_ALIGN - 1)
 
+#ifdef CONFIG_NET
 /*
  * Net namespace inlines
  */
@@ -761,6 +762,7 @@ void dev_net_set(struct net_device *dev, struct net *net)
 	dev->nd_net = net;
 #endif
 }
+#endif
 
 /**
  *	netdev_priv - access network device private data
-- 
cgit v1.2.3


From 0ef4730927c54cac5006759fa0cf50fbeff9d191 Mon Sep 17 00:00:00 2001
From: Matti Linnanvuori <mattilinnanvuori@yahoo.com>
Date: Fri, 28 Mar 2008 16:33:00 -0700
Subject: net: Comment dev_kfree_skb_irq and dev_kfree_skb_any better

Comment dev_kfree_skb_irq and dev_kfree_skb_any better.

Signed-off-by: Matti Linnanvuori <mattilinnanvuori@yahoo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index fae6a7ececdb..ee81906b5164 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1074,12 +1074,14 @@ static inline int netif_is_multiqueue(const struct net_device *dev)
 }
 
 /* Use this variant when it is known for sure that it
- * is executing from interrupt context.
+ * is executing from hardware interrupt context or with hardware interrupts
+ * disabled.
  */
 extern void dev_kfree_skb_irq(struct sk_buff *skb);
 
 /* Use this variant in places where it could be invoked
- * either from interrupt or non-interrupt context.
+ * from either hardware interrupt or other context, with hardware interrupts
+ * either disabled or enabled.
  */
 extern void dev_kfree_skb_any(struct sk_buff *skb);
 
-- 
cgit v1.2.3


From 095d911201b0741e7f326d269a005dba55985acf Mon Sep 17 00:00:00 2001
From: Pavel Emelyanov <xemul@openvz.org>
Date: Fri, 28 Mar 2008 16:39:58 -0700
Subject: [LIB]: Drop the pcounter itself.

The knock-out. The pcounter abstraction is not used any longer in the
kernel.

Not sure whether this should go via netdev tree, but as far as I
remember it was added via this one, and besides Eric thinks that
Andrew shouldn't mind this.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/pcounter.h | 74 ------------------------------------------------
 lib/Makefile             |  1 -
 lib/pcounter.c           | 58 -------------------------------------
 3 files changed, 133 deletions(-)
 delete mode 100644 include/linux/pcounter.h
 delete mode 100644 lib/pcounter.c

(limited to 'include/linux')

diff --git a/include/linux/pcounter.h b/include/linux/pcounter.h
deleted file mode 100644
index a82d9f2628ca..000000000000
--- a/include/linux/pcounter.h
+++ /dev/null
@@ -1,74 +0,0 @@
-#ifndef __LINUX_PCOUNTER_H
-#define __LINUX_PCOUNTER_H
-/*
- * Using a dynamic percpu 'int' variable has a cost :
- * 1) Extra dereference
- * Current per_cpu_ptr() implementation uses an array per 'percpu variable'.
- * 2) memory cost of NR_CPUS*(32+sizeof(void *)) instead of num_possible_cpus()*4
- *
- * This pcounter implementation is an abstraction to be able to use
- * either a static or a dynamic per cpu variable.
- * One dynamic per cpu variable gets a fast & cheap implementation, we can
- * change pcounter implementation too.
- */
-struct pcounter {
-#ifdef CONFIG_SMP
-	void		(*add)(struct pcounter *self, int inc);
-	int		(*getval)(const struct pcounter *self, int cpu);
-	int		*per_cpu_values;
-#else
-	int		val;
-#endif
-};
-
-#ifdef CONFIG_SMP
-#include <linux/percpu.h>
-
-#define DEFINE_PCOUNTER(NAME)						\
-static DEFINE_PER_CPU(int, NAME##_pcounter_values);			\
-static void NAME##_pcounter_add(struct pcounter *self, int val)		\
-{									\
-       __get_cpu_var(NAME##_pcounter_values) += val;			\
-}									\
-static int NAME##_pcounter_getval(const struct pcounter *self, int cpu)	\
-{									\
-	return per_cpu(NAME##_pcounter_values, cpu);			\
-}									\
-
-#define PCOUNTER_MEMBER_INITIALIZER(NAME, MEMBER)		\
-	MEMBER = {						\
-		.add	= NAME##_pcounter_add,			\
-		.getval = NAME##_pcounter_getval,		\
-	}
-
-
-static inline void pcounter_add(struct pcounter *self, int inc)
-{
-	self->add(self, inc);
-}
-
-extern int pcounter_getval(const struct pcounter *self);
-extern int pcounter_alloc(struct pcounter *self);
-extern void pcounter_free(struct pcounter *self);
-
-
-#else /* CONFIG_SMP */
-
-static inline void pcounter_add(struct pcounter *self, int inc)
-{
-	self->val += inc;
-}
-
-static inline int pcounter_getval(const struct pcounter *self)
-{
-	return self->val;
-}
-
-#define DEFINE_PCOUNTER(NAME)
-#define PCOUNTER_MEMBER_INITIALIZER(NAME, MEMBER)
-#define pcounter_alloc(self) 0
-#define pcounter_free(self)
-
-#endif /* CONFIG_SMP */
-
-#endif /* __LINUX_PCOUNTER_H */
diff --git a/lib/Makefile b/lib/Makefile
index 23de261a4c83..4d059d469554 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -61,7 +61,6 @@ obj-$(CONFIG_TEXTSEARCH_KMP) += ts_kmp.o
 obj-$(CONFIG_TEXTSEARCH_BM) += ts_bm.o
 obj-$(CONFIG_TEXTSEARCH_FSM) += ts_fsm.o
 obj-$(CONFIG_SMP) += percpu_counter.o
-obj-$(CONFIG_SMP) += pcounter.o
 obj-$(CONFIG_AUDIT_GENERIC) += audit.o
 
 obj-$(CONFIG_SWIOTLB) += swiotlb.o
diff --git a/lib/pcounter.c b/lib/pcounter.c
deleted file mode 100644
index 9b56807da93b..000000000000
--- a/lib/pcounter.c
+++ /dev/null
@@ -1,58 +0,0 @@
-/*
- * Define default pcounter functions
- * Note that often used pcounters use dedicated functions to get a speed increase.
- * (see DEFINE_PCOUNTER/REF_PCOUNTER_MEMBER)
- */
-
-#include <linux/module.h>
-#include <linux/pcounter.h>
-#include <linux/smp.h>
-#include <linux/cpumask.h>
-
-static void pcounter_dyn_add(struct pcounter *self, int inc)
-{
-	per_cpu_ptr(self->per_cpu_values, smp_processor_id())[0] += inc;
-}
-
-static int pcounter_dyn_getval(const struct pcounter *self, int cpu)
-{
-	return per_cpu_ptr(self->per_cpu_values, cpu)[0];
-}
-
-int pcounter_getval(const struct pcounter *self)
-{
-	int res = 0, cpu;
-
-	for_each_possible_cpu(cpu)
-		res += self->getval(self, cpu);
-
-	return res;
-}
-EXPORT_SYMBOL_GPL(pcounter_getval);
-
-int pcounter_alloc(struct pcounter *self)
-{
-	int rc = 0;
-	if (self->add == NULL) {
-		self->per_cpu_values = alloc_percpu(int);
-		if (self->per_cpu_values != NULL) {
-			self->add    = pcounter_dyn_add;
-			self->getval = pcounter_dyn_getval;
-		} else
-			rc = 1;
-	}
-	return rc;
-}
-EXPORT_SYMBOL_GPL(pcounter_alloc);
-
-void pcounter_free(struct pcounter *self)
-{
-	if (self->per_cpu_values != NULL) {
-		free_percpu(self->per_cpu_values);
-		self->per_cpu_values = NULL;
-		self->getval = NULL;
-		self->add = NULL;
-	}
-}
-EXPORT_SYMBOL_GPL(pcounter_free);
-
-- 
cgit v1.2.3


From 9307b570a745da4f2d83195f5337927e98221bb2 Mon Sep 17 00:00:00 2001
From: "S.Caglar Onur" <caglar@pardus.org.tr>
Date: Fri, 28 Mar 2008 14:41:24 -0700
Subject: drivers/net/arcnet/arcnet.c: use time_* macros

The functions time_before, time_before_eq, time_after, and time_after_eq are
more robust for comparing jiffies against other values.

So use the time_after() macro, defined in linux/jiffies.h, which deals with
wrapping correctly.

[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: S.Caglar Onur <caglar@pardus.org.tr>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/net/arcnet/arcnet.c | 5 +++--
 include/linux/arcdevice.h   | 4 ++--
 2 files changed, 5 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/arcnet/arcnet.c b/drivers/net/arcnet/arcnet.c
index c59c8067de99..bdc4c0bb56d9 100644
--- a/drivers/net/arcnet/arcnet.c
+++ b/drivers/net/arcnet/arcnet.c
@@ -940,7 +940,7 @@ irqreturn_t arcnet_interrupt(int irq, void *dev_id)
 
 			/* is the RECON info empty or old? */
 			if (!lp->first_recon || !lp->last_recon ||
-			    jiffies - lp->last_recon > HZ * 10) {
+			    time_after(jiffies, lp->last_recon + HZ * 10)) {
 				if (lp->network_down)
 					BUGMSG(D_NORMAL, "reconfiguration detected: cabling restored?\n");
 				lp->first_recon = lp->last_recon = jiffies;
@@ -974,7 +974,8 @@ irqreturn_t arcnet_interrupt(int irq, void *dev_id)
 					lp->num_recons = 1;
 				}
 			}
-		} else if (lp->network_down && jiffies - lp->last_recon > HZ * 10) {
+		} else if (lp->network_down &&
+				time_after(jiffies, lp->last_recon + HZ * 10)) {
 			if (lp->network_down)
 				BUGMSG(D_NORMAL, "cabling restored?\n");
 			lp->first_recon = lp->last_recon = 0;
diff --git a/include/linux/arcdevice.h b/include/linux/arcdevice.h
index fde675872c56..a1916078fd08 100644
--- a/include/linux/arcdevice.h
+++ b/include/linux/arcdevice.h
@@ -283,8 +283,8 @@ struct arcnet_local {
 	int next_buf, first_free_buf;
 
 	/* network "reconfiguration" handling */
-	time_t first_recon,	/* time of "first" RECON message to count */
-		last_recon;	/* time of most recent RECON */
+	unsigned long first_recon; /* time of "first" RECON message to count */
+	unsigned long last_recon;  /* time of most recent RECON */
 	int num_recons;		/* number of RECONs between first and last. */
 	bool network_down;	/* do we think the network is down? */
 
-- 
cgit v1.2.3


From 3ec25ebd69dc120d0590e64caaf1477aa88c8a93 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 27 Mar 2008 18:37:14 +0900
Subject: libata: ATA_EHI_LPM should be ATA_EH_LPM

EH actions are ATA_EH_* not ATA_EHI_*.  Rename ATA_EHI_LPM to
ATA_EH_LPM.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 2 +-
 drivers/ata/libata-eh.c   | 2 +-
 include/linux/libata.h    | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index c4248b37ff64..48519887f94a 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -972,7 +972,7 @@ static void ata_dev_disable_pm(struct ata_device *dev)
 void ata_lpm_schedule(struct ata_port *ap, enum link_pm policy)
 {
 	ap->pm_policy = policy;
-	ap->link.eh_info.action |= ATA_EHI_LPM;
+	ap->link.eh_info.action |= ATA_EH_LPM;
 	ap->link.eh_info.flags |= ATA_EHI_NO_AUTOPSY;
 	ata_port_schedule_eh(ap);
 }
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index 681252fd8143..a5830329eda4 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -2748,7 +2748,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 			ehc->i.flags &= ~ATA_EHI_SETMODE;
 		}
 
-		if (ehc->i.action & ATA_EHI_LPM)
+		if (ehc->i.action & ATA_EH_LPM)
 			ata_link_for_each_dev(dev, link)
 				ata_dev_enable_pm(dev, ap->pm_policy);
 
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 269cdba09578..b064bfeb69ee 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -295,6 +295,7 @@ enum {
 	ATA_EH_SOFTRESET	= (1 << 1),
 	ATA_EH_HARDRESET	= (1 << 2),
 	ATA_EH_ENABLE_LINK	= (1 << 3),
+	ATA_EH_LPM		= (1 << 4),  /* link power management action */
 
 	ATA_EH_RESET_MASK	= ATA_EH_SOFTRESET | ATA_EH_HARDRESET,
 	ATA_EH_PERDEV_MASK	= ATA_EH_REVALIDATE,
@@ -304,7 +305,6 @@ enum {
 	ATA_EHI_RESUME_LINK	= (1 << 1),  /* resume link (reset modifier) */
 	ATA_EHI_NO_AUTOPSY	= (1 << 2),  /* no autopsy */
 	ATA_EHI_QUIET		= (1 << 3),  /* be quiet */
-	ATA_EHI_LPM		= (1 << 4),  /* link power management action */
 
 	ATA_EHI_DID_SOFTRESET	= (1 << 16), /* already soft-reset this port */
 	ATA_EHI_DID_HARDRESET	= (1 << 17), /* already soft-reset this port */
-- 
cgit v1.2.3


From 729d4de96a5c090e40a918a41f63b7fb1b27c240 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Sat, 29 Mar 2008 19:55:17 +0100
Subject: ide: fix defining SUPPORT_VLB_SYNC

We need to check for CONFIG_{CRIS,FRV} not {CRIS,FRV}.

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 include/linux/ide.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/ide.h b/include/linux/ide.h
index a3b69c10d667..bc26b2f27359 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -26,7 +26,7 @@
 #include <asm/semaphore.h>
 #include <asm/mutex.h>
 
-#if defined(CRIS) || defined(FRV)
+#if defined(CONFIG_CRIS) || defined(CONFIG_FRV)
 # define SUPPORT_VLB_SYNC 0
 #else
 # define SUPPORT_VLB_SYNC 1
-- 
cgit v1.2.3


From 7d61c4596d11d624efb4bbcbad01f9cf2b321162 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@ftp.linux.org.uk>
Date: Sat, 29 Mar 2008 03:09:28 +0000
Subject: compat_sys_wait4() prototype misannotation

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/compat.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/compat.h b/include/linux/compat.h
index a671dbff7a1f..8fa7857e153b 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -192,8 +192,8 @@ asmlinkage long compat_sys_select(int n, compat_ulong_t __user *inp,
 		struct compat_timeval __user *tvp);
 
 asmlinkage long compat_sys_wait4(compat_pid_t pid,
-				 compat_uint_t *stat_addr, int options,
-				 struct compat_rusage *ru);
+				 compat_uint_t __user *stat_addr, int options,
+				 struct compat_rusage __user *ru);
 
 #define BITS_PER_COMPAT_LONG    (8*sizeof(compat_long_t))
 
-- 
cgit v1.2.3


From b2ddb9019ea13fb7b62d8e45adcc468376af0de7 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@ftp.linux.org.uk>
Date: Sat, 29 Mar 2008 03:09:38 +0000
Subject: dma_page_list ->base_address is a userland pointer

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/dmaengine.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index 261e43a4c873..34d440698293 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -423,7 +423,7 @@ void dma_async_device_unregister(struct dma_device *device);
 /* --- Helper iov-locking functions --- */
 
 struct dma_page_list {
-	char *base_address;
+	char __user *base_address;
 	int nr_pages;
 	struct page **pages;
 };
-- 
cgit v1.2.3


From 3edf8fa5ccf10688a9280b5cbca8ed3947c42866 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Mon, 31 Mar 2008 00:28:14 -0700
Subject: [NET]: Fix allnoconfig build on powerpc and avr32

As reported by Haavard Skinnemoen and Stephen Rothwell:

> allnoconfig fails with
>
> include/linux/netdevice.h:843: error: implicit declaration of function 'dev_net'
>
> which seems to be because the definition of dev_net is inside #ifdef
> CONFIG_NET, while next_net_device, which calls it, is not.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 8576ca928dae..993758f924be 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -827,6 +827,7 @@ struct packet_type {
 extern rwlock_t				dev_base_lock;		/* Device list lock */
 
 
+#ifdef CONFIG_NET
 #define for_each_netdev(net, d)		\
 		list_for_each_entry(d, &(net)->dev_base_head, dev_list)
 #define for_each_netdev_safe(net, d, n)	\
@@ -850,6 +851,7 @@ static inline struct net_device *first_net_device(struct net *net)
 	return list_empty(&net->dev_base_head) ? NULL :
 		net_device_entry(net->dev_base_head.next);
 }
+#endif
 
 extern int 			netdev_boot_setup_check(struct net_device *dev);
 extern unsigned long		netdev_boot_base(const char *prefix, int unit);
-- 
cgit v1.2.3


From 58e9fee13e579df44922172dbe3c9e3ba3edf7a3 Mon Sep 17 00:00:00 2001
From: Benjamin Marzinski <bmarzins@redhat.com>
Date: Fri, 14 Mar 2008 13:52:52 -0500
Subject: [GFS2] Invalidate cache at correct point

GFS2 wasn't invalidating its cache before it called into the lock manager
with a request that could potentially drop a lock.  This was leaving a
window where the lock could be actually be held by another node, but the
file's page cache would still appear valid, causing coherency problems.
This patch moves the cache invalidation to before the lock manager call
when dropping a lock. It also adds the option to the lock_dlm lock
manager to not use conversion mode deadlock avoidance, which, on a
conversion from shared to exclusive, could internally drop the lock, and
then reacquire in. GFS2 now asks lock_dlm to not do this.  Instead, GFS2
manually drops the lock and reacquires it.

Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
---
 fs/gfs2/glock.c              | 35 +++++++++++++++++++++--------------
 fs/gfs2/incore.h             |  1 +
 fs/gfs2/locking/dlm/lock.c   |  3 ++-
 fs/gfs2/locking/dlm/thread.c | 10 +++++++++-
 fs/gfs2/ops_fstype.c         |  2 +-
 include/linux/lm_interface.h | 10 ++++++++++
 6 files changed, 44 insertions(+), 17 deletions(-)

(limited to 'include/linux')

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 63981e2fb835..d636b3e80f5d 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -764,7 +764,7 @@ static void state_change(struct gfs2_glock *gl, unsigned int new_state)
 static void drop_bh(struct gfs2_glock *gl, unsigned int ret)
 {
 	struct gfs2_sbd *sdp = gl->gl_sbd;
-	const struct gfs2_glock_operations *glops = gl->gl_ops;
+	struct gfs2_holder *gh = gl->gl_req_gh;
 
 	gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags));
 	gfs2_assert_warn(sdp, list_empty(&gl->gl_holders));
@@ -772,8 +772,14 @@ static void drop_bh(struct gfs2_glock *gl, unsigned int ret)
 
 	state_change(gl, LM_ST_UNLOCKED);
 
-	if (glops->go_inval)
-		glops->go_inval(gl, DIO_METADATA);
+	if (test_and_clear_bit(GLF_CONV_DEADLK, &gl->gl_flags)) {
+		spin_lock(&gl->gl_spin);
+		gh->gh_error = 0;
+		spin_unlock(&gl->gl_spin);
+		gfs2_glock_xmote_th(gl, gl->gl_req_gh);
+		gfs2_glock_put(gl);
+		return;
+	}
 
 	spin_lock(&gl->gl_spin);
 	gfs2_demote_wake(gl);
@@ -794,7 +800,6 @@ static void xmote_bh(struct gfs2_glock *gl, unsigned int ret)
 	struct gfs2_sbd *sdp = gl->gl_sbd;
 	const struct gfs2_glock_operations *glops = gl->gl_ops;
 	struct gfs2_holder *gh = gl->gl_req_gh;
-	int prev_state = gl->gl_state;
 	int op_done = 1;
 
 	if (!gh && (ret & LM_OUT_ST_MASK) == LM_ST_UNLOCKED) {
@@ -808,16 +813,6 @@ static void xmote_bh(struct gfs2_glock *gl, unsigned int ret)
 
 	state_change(gl, ret & LM_OUT_ST_MASK);
 
-	if (prev_state != LM_ST_UNLOCKED && !(ret & LM_OUT_CACHEABLE)) {
-		if (glops->go_inval)
-			glops->go_inval(gl, DIO_METADATA);
-	} else if (gl->gl_state == LM_ST_DEFERRED) {
-		/* We might not want to do this here.
-		   Look at moving to the inode glops. */
-		if (glops->go_inval)
-			glops->go_inval(gl, 0);
-	}
-
 	/*  Deal with each possible exit condition  */
 
 	if (!gh) {
@@ -837,6 +832,14 @@ static void xmote_bh(struct gfs2_glock *gl, unsigned int ret)
 		}
 	} else {
 		spin_lock(&gl->gl_spin);
+		if (ret & LM_OUT_CONV_DEADLK) {
+			gh->gh_error = 0;
+			set_bit(GLF_CONV_DEADLK, &gl->gl_flags);
+			spin_unlock(&gl->gl_spin);
+			gfs2_glock_drop_th(gl);
+			gfs2_glock_put(gl);
+			return;
+		}
 		list_del_init(&gh->gh_list);
 		gh->gh_error = -EIO;
 		if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) 
@@ -910,6 +913,8 @@ static void gfs2_glock_xmote_th(struct gfs2_glock *gl, struct gfs2_holder *gh)
 
 	if (glops->go_xmote_th)
 		glops->go_xmote_th(gl);
+	if (state == LM_ST_DEFERRED && glops->go_inval)
+		glops->go_inval(gl, DIO_METADATA);
 
 	gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags));
 	gfs2_assert_warn(sdp, list_empty(&gl->gl_holders));
@@ -952,6 +957,8 @@ static void gfs2_glock_drop_th(struct gfs2_glock *gl)
 
 	if (glops->go_xmote_th)
 		glops->go_xmote_th(gl);
+	if (glops->go_inval)
+		glops->go_inval(gl, DIO_METADATA);
 
 	gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags));
 	gfs2_assert_warn(sdp, list_empty(&gl->gl_holders));
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 4ba2ea63119d..9c2c0b90b22a 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -167,6 +167,7 @@ enum {
 	GLF_DEMOTE_IN_PROGRESS	= 6,
 	GLF_LFLUSH		= 7,
 	GLF_WAITERS2		= 8,
+	GLF_CONV_DEADLK		= 9,
 };
 
 struct gfs2_glock {
diff --git a/fs/gfs2/locking/dlm/lock.c b/fs/gfs2/locking/dlm/lock.c
index 542a797ac89a..53a6ab3c0919 100644
--- a/fs/gfs2/locking/dlm/lock.c
+++ b/fs/gfs2/locking/dlm/lock.c
@@ -137,7 +137,8 @@ static inline unsigned int make_flags(struct gdlm_lock *lp,
 
 		/* Conversion deadlock avoidance by DLM */
 
-		if (!test_bit(LFL_FORCE_PROMOTE, &lp->flags) &&
+		if (!(lp->ls->fsflags & LM_MFLAG_CONV_NODROP) &&
+		    !test_bit(LFL_FORCE_PROMOTE, &lp->flags) &&
 		    !(lkf & DLM_LKF_NOQUEUE) &&
 		    cur > DLM_LOCK_NL && req > DLM_LOCK_NL && cur != req)
 			lkf |= DLM_LKF_CONVDEADLK;
diff --git a/fs/gfs2/locking/dlm/thread.c b/fs/gfs2/locking/dlm/thread.c
index 521694fc19d6..e53db6fd28ab 100644
--- a/fs/gfs2/locking/dlm/thread.c
+++ b/fs/gfs2/locking/dlm/thread.c
@@ -135,7 +135,15 @@ static void process_complete(struct gdlm_lock *lp)
 			 lp->lksb.sb_status, lp->lockname.ln_type,
 			 (unsigned long long)lp->lockname.ln_number,
 			 lp->flags);
-		return;
+		if (lp->lksb.sb_status == -EDEADLOCK &&
+		    lp->ls->fsflags & LM_MFLAG_CONV_NODROP) {
+			lp->req = lp->cur;
+			acb.lc_ret |= LM_OUT_CONV_DEADLK;
+			if (lp->cur == DLM_LOCK_IV)
+				lp->lksb.sb_lkid = 0;
+			goto out;
+		} else
+			return;
 	}
 
 	/*
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 5b518f73497a..ef9c6c4f80f6 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -723,7 +723,7 @@ static int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent)
 {
 	char *proto = sdp->sd_proto_name;
 	char *table = sdp->sd_table_name;
-	int flags = 0;
+	int flags = LM_MFLAG_CONV_NODROP;
 	int error;
 
 	if (sdp->sd_args.ar_spectator)
diff --git a/include/linux/lm_interface.h b/include/linux/lm_interface.h
index 1418fdc9ac02..f274997bc283 100644
--- a/include/linux/lm_interface.h
+++ b/include/linux/lm_interface.h
@@ -21,9 +21,15 @@ typedef void (*lm_callback_t) (void *ptr, unsigned int type, void *data);
  * modify the filesystem.  The lock module shouldn't assign a journal to the FS
  * mount.  It shouldn't send recovery callbacks to the FS mount.  If the node
  * dies or withdraws, all locks can be wiped immediately.
+ *
+ * LM_MFLAG_CONV_NODROP
+ * Do not allow the dlm to internally resolve conversion deadlocks by demoting
+ * the lock to unlocked and then reacquiring it in the requested mode. Instead,
+ * it should cancel the request and return LM_OUT_CONV_DEADLK.
  */
 
 #define LM_MFLAG_SPECTATOR	0x00000001
+#define LM_MFLAG_CONV_NODROP	0x00000002
 
 /*
  * lm_lockstruct flags
@@ -110,6 +116,9 @@ typedef void (*lm_callback_t) (void *ptr, unsigned int type, void *data);
  *
  * LM_OUT_ASYNC
  * The result of the request will be returned in an LM_CB_ASYNC callback.
+ *
+ * LM_OUT_CONV_DEADLK
+ * The lock request was canceled do to a conversion deadlock.
  */
 
 #define LM_OUT_ST_MASK		0x00000003
@@ -117,6 +126,7 @@ typedef void (*lm_callback_t) (void *ptr, unsigned int type, void *data);
 #define LM_OUT_CANCELED		0x00000008
 #define LM_OUT_ASYNC		0x00000080
 #define LM_OUT_ERROR		0x00000100
+#define LM_OUT_CONV_DEADLK	0x00000200
 
 /*
  * lm_callback_t types
-- 
cgit v1.2.3


From a7097ff89c3204737a07eecbc83f9ae6002cc534 Mon Sep 17 00:00:00 2001
From: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Date: Tue, 1 Apr 2008 00:22:53 -0400
Subject: Input: make sure input interfaces pin parent input devices

Recent driver core change causes references to parent devices being
dropped early, at device_del() time, as opposed to when all children
are freed. This causes oops in evdev with grabbed devices. Take the
reference to the parent input device ourselves to ensure that it
stays around long enough.

Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
---
 drivers/input/evdev.c    | 6 ++----
 drivers/input/joydev.c   | 3 ++-
 drivers/input/mousedev.c | 3 ++-
 include/linux/input.h    | 5 +++--
 4 files changed, 9 insertions(+), 8 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/input/evdev.c b/drivers/input/evdev.c
index 99562cee827e..b32984bc516f 100644
--- a/drivers/input/evdev.c
+++ b/drivers/input/evdev.c
@@ -124,6 +124,7 @@ static void evdev_free(struct device *dev)
 {
 	struct evdev *evdev = container_of(dev, struct evdev, dev);
 
+	input_put_device(evdev->handle.dev);
 	kfree(evdev);
 }
 
@@ -853,9 +854,6 @@ static void evdev_cleanup(struct evdev *evdev)
 	evdev_hangup(evdev);
 	evdev_remove_chrdev(evdev);
 
-	if (evdev->grab)
-		evdev_ungrab(evdev, evdev->grab);
-
 	/* evdev is marked dead so no one else accesses evdev->open */
 	if (evdev->open) {
 		input_flush_device(handle, NULL);
@@ -896,7 +894,7 @@ static int evdev_connect(struct input_handler *handler, struct input_dev *dev,
 	evdev->exist = 1;
 	evdev->minor = minor;
 
-	evdev->handle.dev = dev;
+	evdev->handle.dev = input_get_device(dev);
 	evdev->handle.name = evdev->name;
 	evdev->handle.handler = handler;
 	evdev->handle.private = evdev;
diff --git a/drivers/input/joydev.c b/drivers/input/joydev.c
index 22b2789ef58a..65d7077a75a1 100644
--- a/drivers/input/joydev.c
+++ b/drivers/input/joydev.c
@@ -171,6 +171,7 @@ static void joydev_free(struct device *dev)
 {
 	struct joydev *joydev = container_of(dev, struct joydev, dev);
 
+	input_put_device(joydev->handle.dev);
 	kfree(joydev);
 }
 
@@ -750,7 +751,7 @@ static int joydev_connect(struct input_handler *handler, struct input_dev *dev,
 	joydev->minor = minor;
 
 	joydev->exist = 1;
-	joydev->handle.dev = dev;
+	joydev->handle.dev = input_get_device(dev);
 	joydev->handle.name = joydev->name;
 	joydev->handle.handler = handler;
 	joydev->handle.private = joydev;
diff --git a/drivers/input/mousedev.c b/drivers/input/mousedev.c
index bbbe5e81adc1..b989748598ae 100644
--- a/drivers/input/mousedev.c
+++ b/drivers/input/mousedev.c
@@ -414,6 +414,7 @@ static void mousedev_free(struct device *dev)
 {
 	struct mousedev *mousedev = container_of(dev, struct mousedev, dev);
 
+	input_put_device(mousedev->handle.dev);
 	kfree(mousedev);
 }
 
@@ -865,7 +866,7 @@ static struct mousedev *mousedev_create(struct input_dev *dev,
 
 	mousedev->minor = minor;
 	mousedev->exist = 1;
-	mousedev->handle.dev = dev;
+	mousedev->handle.dev = input_get_device(dev);
 	mousedev->handle.name = mousedev->name;
 	mousedev->handle.handler = handler;
 	mousedev->handle.private = mousedev;
diff --git a/include/linux/input.h b/include/linux/input.h
index 1bdc39a8c76c..cae2c35d1206 100644
--- a/include/linux/input.h
+++ b/include/linux/input.h
@@ -1227,12 +1227,13 @@ void input_free_device(struct input_dev *dev);
 
 static inline struct input_dev *input_get_device(struct input_dev *dev)
 {
-	return to_input_dev(get_device(&dev->dev));
+	return dev ? to_input_dev(get_device(&dev->dev)) : NULL;
 }
 
 static inline void input_put_device(struct input_dev *dev)
 {
-	put_device(&dev->dev);
+	if (dev)
+		put_device(&dev->dev);
 }
 
 static inline void *input_get_drvdata(struct input_dev *dev)
-- 
cgit v1.2.3


From c0f39322c335412339dec16ebfd2a05ceba5ebcf Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Wed, 2 Apr 2008 00:10:28 -0700
Subject: [NETNS]: Do not include net/net_namespace.h from seq_file.h

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/seq_file.h     | 22 ----------------------
 include/linux/seq_file_net.h | 27 +++++++++++++++++++++++++++
 include/net/net_namespace.h  |  2 ++
 3 files changed, 29 insertions(+), 22 deletions(-)
 create mode 100644 include/linux/seq_file_net.h

(limited to 'include/linux')

diff --git a/include/linux/seq_file.h b/include/linux/seq_file.h
index 5da70c3f4417..1da1e6208a0a 100644
--- a/include/linux/seq_file.h
+++ b/include/linux/seq_file.h
@@ -5,7 +5,6 @@
 #include <linux/types.h>
 #include <linux/string.h>
 #include <linux/mutex.h>
-#include <net/net_namespace.h>
 
 struct seq_operations;
 struct file;
@@ -63,26 +62,5 @@ extern struct list_head *seq_list_start_head(struct list_head *head,
 extern struct list_head *seq_list_next(void *v, struct list_head *head,
 		loff_t *ppos);
 
-#ifdef CONFIG_NET
-struct net;
-struct seq_net_private {
-#ifdef CONFIG_NET_NS
-	struct net *net;
-#endif
-};
-
-int seq_open_net(struct inode *, struct file *,
-		 const struct seq_operations *, int);
-int seq_release_net(struct inode *, struct file *);
-static inline struct net *seq_file_net(struct seq_file *seq)
-{
-#ifdef CONFIG_NET_NS
-	return ((struct seq_net_private *)seq->private)->net;
-#else
-	return &init_net;
-#endif
-}
-#endif	/* CONFIG_NET */
-
 #endif
 #endif
diff --git a/include/linux/seq_file_net.h b/include/linux/seq_file_net.h
new file mode 100644
index 000000000000..4ac52542a563
--- /dev/null
+++ b/include/linux/seq_file_net.h
@@ -0,0 +1,27 @@
+#ifndef __SEQ_FILE_NET_H__
+#define __SEQ_FILE_NET_H__
+
+#include <linux/seq_file.h>
+
+struct net;
+extern struct net init_net;
+
+struct seq_net_private {
+#ifdef CONFIG_NET_NS
+	struct net *net;
+#endif
+};
+
+int seq_open_net(struct inode *, struct file *,
+		 const struct seq_operations *, int);
+int seq_release_net(struct inode *, struct file *);
+static inline struct net *seq_file_net(struct seq_file *seq)
+{
+#ifdef CONFIG_NET_NS
+	return ((struct seq_net_private *)seq->private)->net;
+#else
+	return &init_net;
+#endif
+}
+
+#endif
diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index 4a37037b1d17..6c9a48a46685 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -61,6 +61,8 @@ struct net {
 
 
 #ifdef CONFIG_NET
+#include <linux/seq_file_net.h>
+
 /* Init's network namespace */
 extern struct net init_net;
 #define INIT_NET_NS(net_ns) .net_ns = &init_net,
-- 
cgit v1.2.3


From 34e6bbf23c8f43e8713d9bd092680f1660494b4a Mon Sep 17 00:00:00 2001
From: Fabio Checconi <fabio@gandalf.sssup.i>
Date: Wed, 2 Apr 2008 14:31:02 +0200
Subject: cfq-iosched: fix rcu freeing of cfq io contexts

SLAB_DESTROY_BY_RCU is not a direct substitute for normal call_rcu()
freeing, since it'll page freeing but NOT object freeing. So change
cfq to do the freeing on its own.

Signed-off-by: Fabio Checconi <fabio@gandalf.sssup.it>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 block/cfq-iosched.c       | 57 ++++++++++++++++++++++-------------------------
 include/linux/iocontext.h |  3 +++
 2 files changed, 30 insertions(+), 30 deletions(-)

(limited to 'include/linux')

diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index 0f962ecae91f..f26da2bfcc15 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -1143,24 +1143,37 @@ static void cfq_put_queue(struct cfq_queue *cfqq)
 }
 
 /*
- * Call func for each cic attached to this ioc. Returns number of cic's seen.
+ * Call func for each cic attached to this ioc.
  */
-static unsigned int
+static void
 call_for_each_cic(struct io_context *ioc,
 		  void (*func)(struct io_context *, struct cfq_io_context *))
 {
 	struct cfq_io_context *cic;
 	struct hlist_node *n;
-	int called = 0;
 
 	rcu_read_lock();
-	hlist_for_each_entry_rcu(cic, n, &ioc->cic_list, cic_list) {
+	hlist_for_each_entry_rcu(cic, n, &ioc->cic_list, cic_list)
 		func(ioc, cic);
-		called++;
-	}
 	rcu_read_unlock();
+}
+
+static void cfq_cic_free_rcu(struct rcu_head *head)
+{
+	struct cfq_io_context *cic;
+
+	cic = container_of(head, struct cfq_io_context, rcu_head);
+
+	kmem_cache_free(cfq_ioc_pool, cic);
+	elv_ioc_count_dec(ioc_count);
+
+	if (ioc_gone && !elv_ioc_count_read(ioc_count))
+		complete(ioc_gone);
+}
 
-	return called;
+static void cfq_cic_free(struct cfq_io_context *cic)
+{
+	call_rcu(&cic->rcu_head, cfq_cic_free_rcu);
 }
 
 static void cic_free_func(struct io_context *ioc, struct cfq_io_context *cic)
@@ -1174,24 +1187,18 @@ static void cic_free_func(struct io_context *ioc, struct cfq_io_context *cic)
 	hlist_del_rcu(&cic->cic_list);
 	spin_unlock_irqrestore(&ioc->lock, flags);
 
-	kmem_cache_free(cfq_ioc_pool, cic);
+	cfq_cic_free(cic);
 }
 
 static void cfq_free_io_context(struct io_context *ioc)
 {
-	int freed;
-
 	/*
-	 * ioc->refcount is zero here, so no more cic's are allowed to be
-	 * linked into this ioc. So it should be ok to iterate over the known
-	 * list, we will see all cic's since no new ones are added.
+	 * ioc->refcount is zero here, or we are called from elv_unregister(),
+	 * so no more cic's are allowed to be linked into this ioc.  So it
+	 * should be ok to iterate over the known list, we will see all cic's
+	 * since no new ones are added.
 	 */
-	freed = call_for_each_cic(ioc, cic_free_func);
-
-	elv_ioc_count_mod(ioc_count, -freed);
-
-	if (ioc_gone && !elv_ioc_count_read(ioc_count))
-		complete(ioc_gone);
+	call_for_each_cic(ioc, cic_free_func);
 }
 
 static void cfq_exit_cfqq(struct cfq_data *cfqd, struct cfq_queue *cfqq)
@@ -1458,15 +1465,6 @@ cfq_get_queue(struct cfq_data *cfqd, int is_sync, struct io_context *ioc,
 	return cfqq;
 }
 
-static void cfq_cic_free(struct cfq_io_context *cic)
-{
-	kmem_cache_free(cfq_ioc_pool, cic);
-	elv_ioc_count_dec(ioc_count);
-
-	if (ioc_gone && !elv_ioc_count_read(ioc_count))
-		complete(ioc_gone);
-}
-
 /*
  * We drop cfq io contexts lazily, so we may find a dead one.
  */
@@ -2138,7 +2136,7 @@ static int __init cfq_slab_setup(void)
 	if (!cfq_pool)
 		goto fail;
 
-	cfq_ioc_pool = KMEM_CACHE(cfq_io_context, SLAB_DESTROY_BY_RCU);
+	cfq_ioc_pool = KMEM_CACHE(cfq_io_context, 0);
 	if (!cfq_ioc_pool)
 		goto fail;
 
@@ -2286,7 +2284,6 @@ static void __exit cfq_exit(void)
 	smp_wmb();
 	if (elv_ioc_count_read(ioc_count))
 		wait_for_completion(ioc_gone);
-	synchronize_rcu();
 	cfq_slab_kill();
 }
 
diff --git a/include/linux/iocontext.h b/include/linux/iocontext.h
index 1b4ccf25b4d2..cac4b364cd40 100644
--- a/include/linux/iocontext.h
+++ b/include/linux/iocontext.h
@@ -2,6 +2,7 @@
 #define IOCONTEXT_H
 
 #include <linux/radix-tree.h>
+#include <linux/rcupdate.h>
 
 /*
  * This is the per-process anticipatory I/O scheduler state.
@@ -54,6 +55,8 @@ struct cfq_io_context {
 
 	void (*dtor)(struct io_context *); /* destructor */
 	void (*exit)(struct io_context *); /* called on task exit */
+
+	struct rcu_head rcu_head;
 };
 
 /*
-- 
cgit v1.2.3


From dd135ebbd2a6b5e07dadb66c4dd033bb69531051 Mon Sep 17 00:00:00 2001
From: Christian Borntraeger <borntraeger@de.ibm.com>
Date: Wed, 2 Apr 2008 13:04:40 -0700
Subject: kvm: provide kvm.h for all architecture: fixes headers_install
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Currently include/linux/kvm.h is not considered by make headers_install,
because Kbuild cannot handle " unifdef-$(CONFIG_FOO) += foo.h.  This problem
was introduced by

commit fb56dbb31c4738a3918db81fd24da732ce3b4ae6
Author: Avi Kivity <avi@qumranet.com>
Date:   Sun Dec 2 10:50:06 2007 +0200

    KVM: Export include/linux/kvm.h only if $ARCH actually supports KVM

    Currently, make headers_check barfs due to <asm/kvm.h>, which <linux/kvm.h>
    includes, not existing.  Rather than add a zillion <asm/kvm.h>s, export kvm.
    only if the arch actually supports it.

    Signed-off-by: Avi Kivity <avi@qumranet.com>

which makes this an 2.6.25 regression.

One way of solving the issue is to enhance Kbuild, but Avi and David conviced
me, that changing headers_install is not the way to go.  This patch changes
the definition for linux/kvm.h to unifdef-y.

If  unifdef-y is used for linux/kvm.h "make headers_check" will fail on all
architectures without asm/kvm.h.  Therefore, this patch also provides
asm/kvm.h on all architectures.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Avi Kivity <avi@qumranet.com>
Cc: Sam Ravnborg <sam@ravnborg.org
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/asm-alpha/kvm.h        | 6 ++++++
 include/asm-arm/kvm.h          | 6 ++++++
 include/asm-avr32/kvm.h        | 6 ++++++
 include/asm-blackfin/kvm.h     | 6 ++++++
 include/asm-cris/kvm.h         | 6 ++++++
 include/asm-frv/kvm.h          | 6 ++++++
 include/asm-generic/Kbuild.asm | 2 ++
 include/asm-h8300/kvm.h        | 6 ++++++
 include/asm-ia64/kvm.h         | 6 ++++++
 include/asm-m32r/kvm.h         | 6 ++++++
 include/asm-m68k/kvm.h         | 6 ++++++
 include/asm-m68knommu/kvm.h    | 6 ++++++
 include/asm-mips/kvm.h         | 6 ++++++
 include/asm-mn10300/kvm.h      | 6 ++++++
 include/asm-parisc/kvm.h       | 6 ++++++
 include/asm-powerpc/kvm.h      | 6 ++++++
 include/asm-s390/kvm.h         | 6 ++++++
 include/asm-sh/kvm.h           | 6 ++++++
 include/asm-sparc/kvm.h        | 6 ++++++
 include/asm-sparc64/kvm.h      | 6 ++++++
 include/asm-um/kvm.h           | 6 ++++++
 include/asm-v850/kvm.h         | 6 ++++++
 include/asm-xtensa/kvm.h       | 6 ++++++
 include/linux/Kbuild           | 2 +-
 24 files changed, 135 insertions(+), 1 deletion(-)
 create mode 100644 include/asm-alpha/kvm.h
 create mode 100644 include/asm-arm/kvm.h
 create mode 100644 include/asm-avr32/kvm.h
 create mode 100644 include/asm-blackfin/kvm.h
 create mode 100644 include/asm-cris/kvm.h
 create mode 100644 include/asm-frv/kvm.h
 create mode 100644 include/asm-h8300/kvm.h
 create mode 100644 include/asm-ia64/kvm.h
 create mode 100644 include/asm-m32r/kvm.h
 create mode 100644 include/asm-m68k/kvm.h
 create mode 100644 include/asm-m68knommu/kvm.h
 create mode 100644 include/asm-mips/kvm.h
 create mode 100644 include/asm-mn10300/kvm.h
 create mode 100644 include/asm-parisc/kvm.h
 create mode 100644 include/asm-powerpc/kvm.h
 create mode 100644 include/asm-s390/kvm.h
 create mode 100644 include/asm-sh/kvm.h
 create mode 100644 include/asm-sparc/kvm.h
 create mode 100644 include/asm-sparc64/kvm.h
 create mode 100644 include/asm-um/kvm.h
 create mode 100644 include/asm-v850/kvm.h
 create mode 100644 include/asm-xtensa/kvm.h

(limited to 'include/linux')

diff --git a/include/asm-alpha/kvm.h b/include/asm-alpha/kvm.h
new file mode 100644
index 000000000000..b9daec429689
--- /dev/null
+++ b/include/asm-alpha/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_ALPHA_H
+#define __LINUX_KVM_ALPHA_H
+
+/* alpha does not support KVM */
+
+#endif
diff --git a/include/asm-arm/kvm.h b/include/asm-arm/kvm.h
new file mode 100644
index 000000000000..cb3c08cbcb9e
--- /dev/null
+++ b/include/asm-arm/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_ARM_H
+#define __LINUX_KVM_ARM_H
+
+/* arm does not support KVM */
+
+#endif
diff --git a/include/asm-avr32/kvm.h b/include/asm-avr32/kvm.h
new file mode 100644
index 000000000000..8c5777020e2c
--- /dev/null
+++ b/include/asm-avr32/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_AVR32_H
+#define __LINUX_KVM_AVR32_H
+
+/* avr32 does not support KVM */
+
+#endif
diff --git a/include/asm-blackfin/kvm.h b/include/asm-blackfin/kvm.h
new file mode 100644
index 000000000000..e3477d77c014
--- /dev/null
+++ b/include/asm-blackfin/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_BLACKFIN_H
+#define __LINUX_KVM_BLACKFIN_H
+
+/* blackfin does not support KVM */
+
+#endif
diff --git a/include/asm-cris/kvm.h b/include/asm-cris/kvm.h
new file mode 100644
index 000000000000..c860f51149f0
--- /dev/null
+++ b/include/asm-cris/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_CRIS_H
+#define __LINUX_KVM_CRIS_H
+
+/* cris does not support KVM */
+
+#endif
diff --git a/include/asm-frv/kvm.h b/include/asm-frv/kvm.h
new file mode 100644
index 000000000000..9c8a4f08d0a9
--- /dev/null
+++ b/include/asm-frv/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_FRV_H
+#define __LINUX_KVM_FRV_H
+
+/* frv does not support KVM */
+
+#endif
diff --git a/include/asm-generic/Kbuild.asm b/include/asm-generic/Kbuild.asm
index fd9dcfd91c39..92a6d91d0c1a 100644
--- a/include/asm-generic/Kbuild.asm
+++ b/include/asm-generic/Kbuild.asm
@@ -1,3 +1,5 @@
+header-y  += kvm.h
+
 ifeq ($(wildcard include/asm-$(SRCARCH)/a.out.h),include/asm-$(SRCARCH)/a.out.h)
 unifdef-y += a.out.h
 endif
diff --git a/include/asm-h8300/kvm.h b/include/asm-h8300/kvm.h
new file mode 100644
index 000000000000..bdbed7b987e1
--- /dev/null
+++ b/include/asm-h8300/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_H8300_H
+#define __LINUX_KVM_H8300_H
+
+/* h8300 does not support KVM */
+
+#endif
diff --git a/include/asm-ia64/kvm.h b/include/asm-ia64/kvm.h
new file mode 100644
index 000000000000..030d29b4b26b
--- /dev/null
+++ b/include/asm-ia64/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_IA64_H
+#define __LINUX_KVM_IA64_H
+
+/* ia64 does not support KVM */
+
+#endif
diff --git a/include/asm-m32r/kvm.h b/include/asm-m32r/kvm.h
new file mode 100644
index 000000000000..99a40515b77e
--- /dev/null
+++ b/include/asm-m32r/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_M32R_H
+#define __LINUX_KVM_M32R_H
+
+/* m32r does not support KVM */
+
+#endif
diff --git a/include/asm-m68k/kvm.h b/include/asm-m68k/kvm.h
new file mode 100644
index 000000000000..7ed27fce5240
--- /dev/null
+++ b/include/asm-m68k/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_M68K_H
+#define __LINUX_KVM_M68K_H
+
+/* m68k does not support KVM */
+
+#endif
diff --git a/include/asm-m68knommu/kvm.h b/include/asm-m68knommu/kvm.h
new file mode 100644
index 000000000000..b49d4258dabb
--- /dev/null
+++ b/include/asm-m68knommu/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_M68KNOMMU_H
+#define __LINUX_KVM_M68KNOMMU_H
+
+/* m68knommu does not support KVM */
+
+#endif
diff --git a/include/asm-mips/kvm.h b/include/asm-mips/kvm.h
new file mode 100644
index 000000000000..093a5b7f796b
--- /dev/null
+++ b/include/asm-mips/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_MIPS_H
+#define __LINUX_KVM_MIPS_H
+
+/* mips does not support KVM */
+
+#endif
diff --git a/include/asm-mn10300/kvm.h b/include/asm-mn10300/kvm.h
new file mode 100644
index 000000000000..f6b609ff4a57
--- /dev/null
+++ b/include/asm-mn10300/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_MN10300_H
+#define __LINUX_KVM_MN10300_H
+
+/* mn10300 does not support KVM */
+
+#endif
diff --git a/include/asm-parisc/kvm.h b/include/asm-parisc/kvm.h
new file mode 100644
index 000000000000..00cc45812547
--- /dev/null
+++ b/include/asm-parisc/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_PARISC_H
+#define __LINUX_KVM_PARISC_H
+
+/* parisc does not support KVM */
+
+#endif
diff --git a/include/asm-powerpc/kvm.h b/include/asm-powerpc/kvm.h
new file mode 100644
index 000000000000..d1b530fbf8dd
--- /dev/null
+++ b/include/asm-powerpc/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_POWERPC_H
+#define __LINUX_KVM_POWERPC_H
+
+/* powerpc does not support KVM */
+
+#endif
diff --git a/include/asm-s390/kvm.h b/include/asm-s390/kvm.h
new file mode 100644
index 000000000000..573f2a351386
--- /dev/null
+++ b/include/asm-s390/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_S390_H
+#define __LINUX_KVM_S390_H
+
+/* s390 does not support KVM */
+
+#endif
diff --git a/include/asm-sh/kvm.h b/include/asm-sh/kvm.h
new file mode 100644
index 000000000000..6af51dbab2d0
--- /dev/null
+++ b/include/asm-sh/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_SH_H
+#define __LINUX_KVM_SH_H
+
+/* sh does not support KVM */
+
+#endif
diff --git a/include/asm-sparc/kvm.h b/include/asm-sparc/kvm.h
new file mode 100644
index 000000000000..2e5478da3819
--- /dev/null
+++ b/include/asm-sparc/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_SPARC_H
+#define __LINUX_KVM_SPARC_H
+
+/* sparc does not support KVM */
+
+#endif
diff --git a/include/asm-sparc64/kvm.h b/include/asm-sparc64/kvm.h
new file mode 100644
index 000000000000..380537a77bf9
--- /dev/null
+++ b/include/asm-sparc64/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_SPARC64_H
+#define __LINUX_KVM_SPARC64_H
+
+/* sparc64 does not support KVM */
+
+#endif
diff --git a/include/asm-um/kvm.h b/include/asm-um/kvm.h
new file mode 100644
index 000000000000..66aa77094551
--- /dev/null
+++ b/include/asm-um/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_UM_H
+#define __LINUX_KVM_UM_H
+
+/* um does not support KVM */
+
+#endif
diff --git a/include/asm-v850/kvm.h b/include/asm-v850/kvm.h
new file mode 100644
index 000000000000..3f729b79febc
--- /dev/null
+++ b/include/asm-v850/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_V850_H
+#define __LINUX_KVM_V850_H
+
+/* v850 does not support KVM */
+
+#endif
diff --git a/include/asm-xtensa/kvm.h b/include/asm-xtensa/kvm.h
new file mode 100644
index 000000000000..bda4e331e98c
--- /dev/null
+++ b/include/asm-xtensa/kvm.h
@@ -0,0 +1,6 @@
+#ifndef __LINUX_KVM_XTENSA_H
+#define __LINUX_KVM_XTENSA_H
+
+/* xtensa does not support KVM */
+
+#endif
diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 4a446a19295e..9cdd12a9e843 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -253,7 +253,7 @@ unifdef-y += kd.h
 unifdef-y += kernelcapi.h
 unifdef-y += kernel.h
 unifdef-y += keyboard.h
-unifdef-$(CONFIG_HAVE_KVM) += kvm.h
+unifdef-y += kvm.h
 unifdef-y += llc.h
 unifdef-y += loop.h
 unifdef-y += lp.h
-- 
cgit v1.2.3


From fadf6bf06069138f8e97c9a963be38348ba2708b Mon Sep 17 00:00:00 2001
From: "Templin, Fred L" <Fred.L.Templin@boeing.com>
Date: Tue, 11 Mar 2008 18:35:59 -0400
Subject: [IPV6] SIT: Add PRL management for ISATAP.

This patch updates the Linux the Intra-Site Automatic Tunnel Addressing
Protocol (ISATAP) implementation. It places the ISATAP potential router
list (PRL) in the kernel and adds three new private ioctls for PRL
management.

[Add several changes of structure name, constant names etc. - yoshfuji]

Signed-off-by: Fred L. Templin <fred.l.templin@boeing.com>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/if_tunnel.h |  18 ++++-
 include/linux/skbuff.h    |   3 +-
 include/net/ipip.h        |   7 ++
 include/net/ndisc.h       |   9 +++
 net/ipv6/ndisc.c          |  24 ++++++
 net/ipv6/route.c          |   2 -
 net/ipv6/sit.c            | 186 +++++++++++++++++++++++++++++++++++-----------
 7 files changed, 199 insertions(+), 50 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/if_tunnel.h b/include/linux/if_tunnel.h
index 228eb4eb3129..f20c224d544c 100644
--- a/include/linux/if_tunnel.h
+++ b/include/linux/if_tunnel.h
@@ -7,6 +7,9 @@
 #define SIOCADDTUNNEL   (SIOCDEVPRIVATE + 1)
 #define SIOCDELTUNNEL   (SIOCDEVPRIVATE + 2)
 #define SIOCCHGTUNNEL   (SIOCDEVPRIVATE + 3)
+#define SIOCADDPRL      (SIOCDEVPRIVATE + 5)
+#define SIOCDELPRL      (SIOCDEVPRIVATE + 6)
+#define SIOCCHGPRL      (SIOCDEVPRIVATE + 7)
 
 #define GRE_CSUM	__constant_htons(0x8000)
 #define GRE_ROUTING	__constant_htons(0x4000)
@@ -17,9 +20,6 @@
 #define GRE_FLAGS	__constant_htons(0x00F8)
 #define GRE_VERSION	__constant_htons(0x0007)
 
-/* i_flags values for SIT mode */
-#define	SIT_ISATAP	0x0001
-
 struct ip_tunnel_parm
 {
 	char			name[IFNAMSIZ];
@@ -31,4 +31,16 @@ struct ip_tunnel_parm
 	struct iphdr		iph;
 };
 
+/* SIT-mode i_flags */
+#define	SIT_ISATAP	0x0001
+
+struct ip_tunnel_prl {
+	__be32			addr;
+	__u16			flags;
+	__u16			__reserved;
+};
+
+/* PRL flags */
+#define	PRL_DEFAULT		0x0001
+
 #endif /* _IF_TUNNEL_H_ */
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index ff72145d5d9e..e10e55c9b081 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -313,7 +313,8 @@ struct sk_buff {
 	__u16			tc_verd;	/* traffic control verdict */
 #endif
 #endif
-	/* 2 byte hole */
+	__u8			ndisc_nodetype:2;
+	/* 14 bit hole */
 
 #ifdef CONFIG_NET_DMA
 	dma_cookie_t		dma_cookie;
diff --git a/include/net/ipip.h b/include/net/ipip.h
index 549e132bca9c..205536a014e8 100644
--- a/include/net/ipip.h
+++ b/include/net/ipip.h
@@ -24,6 +24,13 @@ struct ip_tunnel
 	int			mlink;
 
 	struct ip_tunnel_parm	parms;
+	struct ip_tunnel_prl_entry	*prl;		/* potential router list */
+};
+
+struct ip_tunnel_prl_entry
+{
+	struct ip_tunnel_prl_entry	*next;
+	struct ip_tunnel_prl		entry;
 };
 
 #define IPTUNNEL_XMIT() do {						\
diff --git a/include/net/ndisc.h b/include/net/ndisc.h
index 5aedf324de66..9f2bae68d28c 100644
--- a/include/net/ndisc.h
+++ b/include/net/ndisc.h
@@ -11,6 +11,15 @@
 #define NDISC_NEIGHBOUR_ADVERTISEMENT	136
 #define NDISC_REDIRECT			137
 
+/*
+ * Router type: cross-layer information from link-layer to
+ * IPv6 layer reported by certain link types (e.g., RFC4214).
+ */
+#define NDISC_NODETYPE_UNSPEC		0	/* unspecified (default) */
+#define NDISC_NODETYPE_HOST		1	/* host or unauthorized router */
+#define NDISC_NODETYPE_NODEFAULT	2	/* non-default router */
+#define NDISC_NODETYPE_DEFAULT		3	/* default router */
+
 /*
  *	ndisc options
  */
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 510aa747a404..53b546019fd5 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1092,6 +1092,12 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 		return;
 	}
 
+	if (skb->ndisc_nodetype == NDISC_NODETYPE_HOST) {
+		ND_PRINTK2(KERN_WARNING
+			   "ICMPv6 RA: from host or unauthorized router\n");
+		return;
+	}
+
 	/*
 	 *	set the RA_RECV flag in the interface
 	 */
@@ -1115,6 +1121,10 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 		return;
 	}
 
+	/* skip link-specific parameters from interior routers */
+	if (skb->ndisc_nodetype == NDISC_NODETYPE_NODEFAULT)
+		goto skip_linkparms;
+
 	if (in6_dev->if_flags & IF_RS_SENT) {
 		/*
 		 *	flag that an RA was received after an RS was sent
@@ -1229,6 +1239,8 @@ skip_defrtr:
 		}
 	}
 
+skip_linkparms:
+
 	/*
 	 *	Process options.
 	 */
@@ -1268,6 +1280,10 @@ skip_defrtr:
 	}
 #endif
 
+	/* skip link-specific ndopts from interior routers */
+	if (skb->ndisc_nodetype == NDISC_NODETYPE_NODEFAULT)
+		goto out;
+
 	if (in6_dev->cnf.accept_ra_pinfo && ndopts.nd_opts_pi) {
 		struct nd_opt_hdr *p;
 		for (p = ndopts.nd_opts_pi;
@@ -1331,6 +1347,14 @@ static void ndisc_redirect_rcv(struct sk_buff *skb)
 	int optlen;
 	u8 *lladdr = NULL;
 
+	switch (skb->ndisc_nodetype) {
+	case NDISC_NODETYPE_HOST:
+	case NDISC_NODETYPE_NODEFAULT:
+		ND_PRINTK2(KERN_WARNING
+			   "ICMPv6 Redirect: from host or unauthorized router\n");
+		return;
+	}
+
 	if (!(ipv6_addr_type(&ipv6_hdr(skb)->saddr) & IPV6_ADDR_LINKLOCAL)) {
 		ND_PRINTK2(KERN_WARNING
 			   "ICMPv6 Redirect: source address is not link-local.\n");
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index cd82b6db35ff..f17b2f61891e 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1699,8 +1699,6 @@ struct rt6_info *rt6_get_dflt_router(struct in6_addr *addr, struct net_device *d
 	return rt;
 }
 
-EXPORT_SYMBOL(rt6_get_dflt_router);
-
 struct rt6_info *rt6_add_dflt_router(struct in6_addr *gwaddr,
 				     struct net_device *dev,
 				     unsigned int pref)
diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c
index 1b8196c8d145..4786419ade0e 100644
--- a/net/ipv6/sit.c
+++ b/net/ipv6/sit.c
@@ -16,7 +16,7 @@
  *	Changes:
  * Roger Venning <r.venning@telstra.com>:	6to4 support
  * Nate Thompson <nate@thebog.net>:		6to4 support
- * Fred L. Templin <fltemplin@acm.org>:		isatap support
+ * Fred Templin <fred.l.templin@boeing.com>:	isatap support
  */
 
 #include <linux/module.h>
@@ -197,6 +197,119 @@ failed:
 	return NULL;
 }
 
+static struct ip_tunnel_prl_entry *
+ipip6_tunnel_locate_prl(struct ip_tunnel *t, __be32 addr)
+{
+	struct ip_tunnel_prl_entry *p = (struct ip_tunnel_prl_entry *)NULL;
+
+	for (p = t->prl; p; p = p->next)
+		if (p->entry.addr == addr)
+			break;
+	return p;
+
+}
+
+static int
+ipip6_tunnel_add_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a, int chg)
+{
+	struct ip_tunnel_prl_entry *p;
+
+	for (p = t->prl; p; p = p->next) {
+		if (p->entry.addr == a->addr) {
+			if (chg) {
+				p->entry = *a;
+				return 0;
+			}
+			return -EEXIST;
+		}
+	}
+
+	if (chg)
+		return -ENXIO;
+
+	p = kzalloc(sizeof(struct ip_tunnel_prl_entry), GFP_KERNEL);
+	if (!p)
+		return -ENOBUFS;
+
+	p->entry = *a;
+	p->next = t->prl;
+	t->prl = p;
+	return 0;
+}
+
+static int
+ipip6_tunnel_del_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a)
+{
+	struct ip_tunnel_prl_entry *x, **p;
+
+	if (a) {
+		for (p = &t->prl; *p; p = &(*p)->next) {
+			if ((*p)->entry.addr == a->addr) {
+				x = *p;
+				*p = x->next;
+				kfree(x);
+				return 0;
+			}
+		}
+		return -ENXIO;
+	} else {
+		while (t->prl) {
+			x = t->prl;
+			t->prl = t->prl->next;
+			kfree(x);
+		}
+	}
+	return 0;
+}
+
+/* copied directly from anycast.c */
+static int
+ipip6_onlink(struct in6_addr *addr, struct net_device *dev)
+{
+	struct inet6_dev	*idev;
+	struct inet6_ifaddr	*ifa;
+	int	onlink;
+
+	onlink = 0;
+	rcu_read_lock();
+	idev = __in6_dev_get(dev);
+	if (idev) {
+		read_lock_bh(&idev->lock);
+		for (ifa=idev->addr_list; ifa; ifa=ifa->if_next) {
+			onlink = ipv6_prefix_equal(addr, &ifa->addr,
+						   ifa->prefix_len);
+			if (onlink)
+				break;
+		}
+		read_unlock_bh(&idev->lock);
+	}
+	rcu_read_unlock();
+	return onlink;
+}
+
+static int
+isatap_chksrc(struct sk_buff *skb, struct iphdr *iph, struct ip_tunnel *t)
+{
+	struct ip_tunnel_prl_entry *p = ipip6_tunnel_locate_prl(t, iph->saddr);
+	int ok = 1;
+
+	if (p) {
+		if (p->entry.flags & PRL_DEFAULT)
+			skb->ndisc_nodetype = NDISC_NODETYPE_DEFAULT;
+		else
+			skb->ndisc_nodetype = NDISC_NODETYPE_NODEFAULT;
+	} else {
+		struct in6_addr *addr6 = &ipv6_hdr(skb)->saddr;
+		if (ipv6_addr_is_isatap(addr6) &&
+		    (addr6->s6_addr32[3] == iph->saddr) &&
+		    ipip6_onlink(addr6, t->dev))
+			skb->ndisc_nodetype = NDISC_NODETYPE_HOST;
+		else
+			ok = 0;
+	}
+	return ok;
+}
+
 static void ipip6_tunnel_uninit(struct net_device *dev)
 {
 	if (dev == ipip6_fb_tunnel_dev) {
@@ -206,6 +319,7 @@ static void ipip6_tunnel_uninit(struct net_device *dev)
 		dev_put(dev);
 	} else {
 		ipip6_tunnel_unlink(netdev_priv(dev));
+		ipip6_tunnel_del_prl(netdev_priv(dev), 0);
 		dev_put(dev);
 	}
 }
@@ -365,48 +479,6 @@ static inline void ipip6_ecn_decapsulate(struct iphdr *iph, struct sk_buff *skb)
 		IP6_ECN_set_ce(ipv6_hdr(skb));
 }
 
-/* ISATAP (RFC4214) - check source address */
-static int
-isatap_srcok(struct sk_buff *skb, struct iphdr *iph, struct net_device *dev)
-{
-	struct neighbour *neigh;
-	struct dst_entry *dst;
-	struct rt6_info *rt;
-	struct flowi fl;
-	struct in6_addr *addr6;
-	struct in6_addr rtr;
-	struct ipv6hdr *iph6;
-	int ok = 0;
-
-	/* from onlink default router */
-	ipv6_addr_set(&rtr,  htonl(0xFE800000), 0, 0, 0);
-	ipv6_isatap_eui64(rtr.s6_addr + 8, iph->saddr);
-	if ((rt = rt6_get_dflt_router(&rtr, dev))) {
-		dst_release(&rt->u.dst);
-		return 1;
-	}
-
-	iph6 = ipv6_hdr(skb);
-	memset(&fl, 0, sizeof(fl));
-	fl.proto = iph6->nexthdr;
-	ipv6_addr_copy(&fl.fl6_dst, &iph6->saddr);
-	fl.oif = dev->ifindex;
-	security_skb_classify_flow(skb, &fl);
-
-	dst = ip6_route_output(&init_net, NULL, &fl);
-	if (!dst->error && (dst->dev == dev) && (neigh = dst->neighbour)) {
-
-		addr6 = (struct in6_addr*)&neigh->primary_key;
-
-		/* from correct previous hop */
-		if (ipv6_addr_is_isatap(addr6) &&
-		    (addr6->s6_addr32[3] == iph->saddr))
-			ok = 1;
-	}
-	dst_release(dst);
-	return ok;
-}
-
 static int ipip6_rcv(struct sk_buff *skb)
 {
 	struct iphdr *iph;
@@ -427,7 +499,7 @@ static int ipip6_rcv(struct sk_buff *skb)
 		skb->pkt_type = PACKET_HOST;
 
 		if ((tunnel->dev->priv_flags & IFF_ISATAP) &&
-		    !isatap_srcok(skb, iph, tunnel->dev)) {
+		    !isatap_chksrc(skb, iph, tunnel)) {
 			tunnel->stat.rx_errors++;
 			read_unlock(&ipip6_lock);
 			kfree_skb(skb);
@@ -707,6 +779,7 @@ ipip6_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
 {
 	int err = 0;
 	struct ip_tunnel_parm p;
+	struct ip_tunnel_prl prl;
 	struct ip_tunnel *t;
 
 	switch (cmd) {
@@ -806,6 +879,31 @@ ipip6_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
 		err = 0;
 		break;
 
+	case SIOCADDPRL:
+	case SIOCDELPRL:
+	case SIOCCHGPRL:
+		err = -EPERM;
+		if (!capable(CAP_NET_ADMIN))
+			goto done;
+		err = -EINVAL;
+		if (dev == ipip6_fb_tunnel_dev)
+			goto done;
+		err = -EFAULT;
+		if (copy_from_user(&prl, ifr->ifr_ifru.ifru_data, sizeof(prl)))
+			goto done;
+		err = -ENOENT;
+		if (!(t = netdev_priv(dev)))
+			goto done;
+
+		ipip6_tunnel_unlink(t);
+		if (cmd == SIOCDELPRL)
+			err = ipip6_tunnel_del_prl(t, &prl);
+		else
+			err = ipip6_tunnel_add_prl(t, &prl, cmd == SIOCCHGPRL);
+		ipip6_tunnel_link(t);
+		netdev_state_change(dev);
+		break;
+
 	default:
 		err = -EINVAL;
 	}
-- 
cgit v1.2.3


From 300aaeeaab5f447fcf40e911afe96df3de28f0db Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Mon, 24 Mar 2008 18:28:39 +0900
Subject: [IPV6] SIT: Add SIOCGETPRL ioctl to get/dump PRL.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/if_tunnel.h |  4 ++
 include/net/ipip.h        |  5 ++-
 net/ipv6/sit.c            | 96 ++++++++++++++++++++++++++++++++++++++++++-----
 3 files changed, 95 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/if_tunnel.h b/include/linux/if_tunnel.h
index f20c224d544c..f1fbe9c930d7 100644
--- a/include/linux/if_tunnel.h
+++ b/include/linux/if_tunnel.h
@@ -7,6 +7,7 @@
 #define SIOCADDTUNNEL   (SIOCDEVPRIVATE + 1)
 #define SIOCDELTUNNEL   (SIOCDEVPRIVATE + 2)
 #define SIOCCHGTUNNEL   (SIOCDEVPRIVATE + 3)
+#define SIOCGETPRL      (SIOCDEVPRIVATE + 4)
 #define SIOCADDPRL      (SIOCDEVPRIVATE + 5)
 #define SIOCDELPRL      (SIOCDEVPRIVATE + 6)
 #define SIOCCHGPRL      (SIOCDEVPRIVATE + 7)
@@ -38,6 +39,9 @@ struct ip_tunnel_prl {
 	__be32			addr;
 	__u16			flags;
 	__u16			__reserved;
+	__u32			datalen;
+	__u32			__reserved2;
+	void __user		*data;
 };
 
 /* PRL flags */
diff --git a/include/net/ipip.h b/include/net/ipip.h
index 205536a014e8..633ed4def8e3 100644
--- a/include/net/ipip.h
+++ b/include/net/ipip.h
@@ -24,13 +24,16 @@ struct ip_tunnel
 	int			mlink;
 
 	struct ip_tunnel_parm	parms;
+
 	struct ip_tunnel_prl_entry	*prl;		/* potential router list */
+	unsigned int			prl_count;	/* # of entries in PRL */
 };
 
 struct ip_tunnel_prl_entry
 {
 	struct ip_tunnel_prl_entry	*next;
-	struct ip_tunnel_prl		entry;
+	__be32				addr;
+	u16				flags;
 };
 
 #define IPTUNNEL_XMIT() do {						\
diff --git a/net/ipv6/sit.c b/net/ipv6/sit.c
index 84c1ed246afb..08a483a8de50 100644
--- a/net/ipv6/sit.c
+++ b/net/ipv6/sit.c
@@ -203,12 +203,73 @@ __ipip6_tunnel_locate_prl(struct ip_tunnel *t, __be32 addr)
 	struct ip_tunnel_prl_entry *p = (struct ip_tunnel_prl_entry *)NULL;
 
 	for (p = t->prl; p; p = p->next)
-		if (p->entry.addr == addr)
+		if (p->addr == addr)
 			break;
 	return p;
 
 }
 
+static int ipip6_tunnel_get_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a)
+{
+	struct ip_tunnel_prl *kp;
+	struct ip_tunnel_prl_entry *prl;
+	unsigned int cmax, c = 0, ca, len;
+	int ret = 0;
+
+	cmax = a->datalen / sizeof(*a);
+	if (cmax > 1 && a->addr != htonl(INADDR_ANY))
+		cmax = 1;
+
+	/* For simple GET or for root users,
+	 * we try harder to allocate.
+	 */
+	kp = (cmax <= 1 || capable(CAP_NET_ADMIN)) ?
+		kcalloc(cmax, sizeof(*kp), GFP_KERNEL) :
+		NULL;
+
+	read_lock(&ipip6_lock);
+
+	ca = t->prl_count < cmax ? t->prl_count : cmax;
+
+	if (!kp) {
+		/* We don't try hard to allocate much memory for
+		 * non-root users.
+		 * For root users, retry allocating enough memory for
+		 * the answer.
+		 */
+		kp = kcalloc(ca, sizeof(*kp), GFP_ATOMIC);
+		if (!kp) {
+			ret = -ENOMEM;
+			goto out;
+		}
+	}
+
+	c = 0;
+	for (prl = t->prl; prl; prl = prl->next) {
+		if (c > cmax)
+			break;
+		if (a->addr != htonl(INADDR_ANY) && prl->addr != a->addr)
+			continue;
+		kp[c].addr = prl->addr;
+		kp[c].flags = prl->flags;
+		c++;
+		if (a->addr != htonl(INADDR_ANY))
+			break;
+	}
+out:
+	read_unlock(&ipip6_lock);
+
+	len = sizeof(*kp) * c;
+	ret = len ? copy_to_user(a->data, kp, len) : 0;
+
+	kfree(kp);
+	if (ret)
+		return -EFAULT;
+
+	a->datalen = len;
+	return 0;
+}
+
 static int
 ipip6_tunnel_add_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a, int chg)
 {
@@ -221,7 +282,7 @@ ipip6_tunnel_add_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a, int chg)
 	write_lock(&ipip6_lock);
 
 	for (p = t->prl; p; p = p->next) {
-		if (p->entry.addr == a->addr) {
+		if (p->addr == a->addr) {
 			if (chg)
 				goto update;
 			err = -EEXIST;
@@ -242,8 +303,10 @@ ipip6_tunnel_add_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a, int chg)
 
 	p->next = t->prl;
 	t->prl = p;
+	t->prl_count++;
 update:
-	p->entry = *a;
+	p->addr = a->addr;
+	p->flags = a->flags;
 out:
 	write_unlock(&ipip6_lock);
 	return err;
@@ -259,10 +322,11 @@ ipip6_tunnel_del_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a)
 
 	if (a && a->addr != htonl(INADDR_ANY)) {
 		for (p = &t->prl; *p; p = &(*p)->next) {
-			if ((*p)->entry.addr == a->addr) {
+			if ((*p)->addr == a->addr) {
 				x = *p;
 				*p = x->next;
 				kfree(x);
+				t->prl_count--;
 				goto out;
 			}
 		}
@@ -272,6 +336,7 @@ ipip6_tunnel_del_prl(struct ip_tunnel *t, struct ip_tunnel_prl *a)
 			x = t->prl;
 			t->prl = t->prl->next;
 			kfree(x);
+			t->prl_count--;
 		}
 	}
 out:
@@ -313,7 +378,7 @@ isatap_chksrc(struct sk_buff *skb, struct iphdr *iph, struct ip_tunnel *t)
 	read_lock(&ipip6_lock);
 	p = __ipip6_tunnel_locate_prl(t, iph->saddr);
 	if (p) {
-		if (p->entry.flags & PRL_DEFAULT)
+		if (p->flags & PRL_DEFAULT)
 			skb->ndisc_nodetype = NDISC_NODETYPE_DEFAULT;
 		else
 			skb->ndisc_nodetype = NDISC_NODETYPE_NODEFAULT;
@@ -899,11 +964,12 @@ ipip6_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
 		err = 0;
 		break;
 
+	case SIOCGETPRL:
 	case SIOCADDPRL:
 	case SIOCDELPRL:
 	case SIOCCHGPRL:
 		err = -EPERM;
-		if (!capable(CAP_NET_ADMIN))
+		if (cmd != SIOCGETPRL && !capable(CAP_NET_ADMIN))
 			goto done;
 		err = -EINVAL;
 		if (dev == ipip6_fb_tunnel_dev)
@@ -915,11 +981,23 @@ ipip6_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
 		if (!(t = netdev_priv(dev)))
 			goto done;
 
-		if (cmd == SIOCDELPRL)
+		switch (cmd) {
+		case SIOCGETPRL:
+			err = ipip6_tunnel_get_prl(t, &prl);
+			if (!err && copy_to_user(ifr->ifr_ifru.ifru_data,
+						 &prl, sizeof(prl)))
+				err = -EFAULT;
+			break;
+		case SIOCDELPRL:
 			err = ipip6_tunnel_del_prl(t, &prl);
-		else
+			break;
+		case SIOCADDPRL:
+		case SIOCCHGPRL:
 			err = ipip6_tunnel_add_prl(t, &prl, cmd == SIOCCHGPRL);
-		netdev_state_change(dev);
+			break;
+		}
+		if (cmd != SIOCGETPRL)
+			netdev_state_change(dev);
 		break;
 
 	default:
-- 
cgit v1.2.3


From de357cc01334a468e4d5b7ba66a17b0d3ca9d63e Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Sat, 15 Mar 2008 23:59:18 -0400
Subject: [IPV6] NDISC: Don't rely on node-type hint from L2 unless required.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/skbuff.h |  2 ++
 net/ipv6/Kconfig       |  4 ++++
 net/ipv6/ndisc.c       | 10 ++++++++++
 3 files changed, 16 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index e10e55c9b081..e517701c25ba 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -313,7 +313,9 @@ struct sk_buff {
 	__u16			tc_verd;	/* traffic control verdict */
 #endif
 #endif
+#ifdef CONFIG_IPV6_NDISC_NODETYPE
 	__u8			ndisc_nodetype:2;
+#endif
 	/* 14 bit hole */
 
 #ifdef CONFIG_NET_DMA
diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index 47263e45bacb..7d2e7f0941ac 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -168,6 +168,7 @@ config IPV6_SIT
 	tristate "IPv6: IPv6-in-IPv4 tunnel (SIT driver)"
 	depends on IPV6
 	select INET_TUNNEL
+	select IPV6_NDISC_NODETYPE
 	default y
 	---help---
 	  Tunneling means encapsulating data of one protocol type within
@@ -178,6 +179,9 @@ config IPV6_SIT
 
 	  Saying M here will produce a module called sit.ko. If unsure, say Y.
 
+config IPV6_NDISC_NODETYPE
+	bool
+
 config IPV6_TUNNEL
 	tristate "IPv6: IP-in-IPv6 tunnel (RFC2473)"
 	select INET6_TUNNEL
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 16273e11e53d..c400b874097a 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1092,11 +1092,13 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 		return;
 	}
 
+#ifdef CONFIG_IPV6_NDISC_NODETYPE
 	if (skb->ndisc_nodetype == NDISC_NODETYPE_HOST) {
 		ND_PRINTK2(KERN_WARNING
 			   "ICMPv6 RA: from host or unauthorized router\n");
 		return;
 	}
+#endif
 
 	/*
 	 *	set the RA_RECV flag in the interface
@@ -1121,9 +1123,11 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 		return;
 	}
 
+#ifdef CONFIG_IPV6_NDISC_NODETYPE
 	/* skip link-specific parameters from interior routers */
 	if (skb->ndisc_nodetype == NDISC_NODETYPE_NODEFAULT)
 		goto skip_linkparms;
+#endif
 
 	if (in6_dev->if_flags & IF_RS_SENT) {
 		/*
@@ -1239,7 +1243,9 @@ skip_defrtr:
 		}
 	}
 
+#ifdef CONFIG_IPV6_NDISC_NODETYPE
 skip_linkparms:
+#endif
 
 	/*
 	 *	Process options.
@@ -1286,9 +1292,11 @@ skip_linkparms:
 	}
 #endif
 
+#ifdef CONFIG_IPV6_NDISC_NODETYPE
 	/* skip link-specific ndopts from interior routers */
 	if (skb->ndisc_nodetype == NDISC_NODETYPE_NODEFAULT)
 		goto out;
+#endif
 
 	if (in6_dev->cnf.accept_ra_pinfo && ndopts.nd_opts_pi) {
 		struct nd_opt_hdr *p;
@@ -1353,6 +1361,7 @@ static void ndisc_redirect_rcv(struct sk_buff *skb)
 	int optlen;
 	u8 *lladdr = NULL;
 
+#ifdef CONFIG_IPV6_NDISC_NODETYPE
 	switch (skb->ndisc_nodetype) {
 	case NDISC_NODETYPE_HOST:
 	case NDISC_NODETYPE_NODEFAULT:
@@ -1360,6 +1369,7 @@ static void ndisc_redirect_rcv(struct sk_buff *skb)
 			   "ICMPv6 Redirect: from host or unauthorized router\n");
 		return;
 	}
+#endif
 
 	if (!(ipv6_addr_type(&ipv6_hdr(skb)->saddr) & IPV6_ADDR_LINKLOCAL)) {
 		ND_PRINTK2(KERN_WARNING
-- 
cgit v1.2.3


From a4aa834a9165150252c5cd953faab4de29d51b87 Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Thu, 3 Apr 2008 13:04:33 -0700
Subject: [NETNS]: Declare init_net even without CONFIG_NET defined.

This does not look good, but there is no other choice. The compilation
without CONFIG_NET is broken and can not be fixed with ease.

After that there is no need for the following commits:
1567ca7eec7664b8be3b07755ac59dc1b1ec76cb
3edf8fa5ccf10688a9280b5cbca8ed3947c42866
2d38f9a4f8d2ebdc799f03eecf82345825495711

Revert them.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h   | 4 ----
 include/net/net_namespace.h | 3 ++-
 lib/kobject_uevent.c        | 2 --
 3 files changed, 2 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index c36c76caf20b..8b17ed40dea2 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -741,7 +741,6 @@ struct net_device
 #define	NETDEV_ALIGN		32
 #define	NETDEV_ALIGN_CONST	(NETDEV_ALIGN - 1)
 
-#ifdef CONFIG_NET
 /*
  * Net namespace inlines
  */
@@ -762,7 +761,6 @@ void dev_net_set(struct net_device *dev, struct net *net)
 	dev->nd_net = net;
 #endif
 }
-#endif
 
 /**
  *	netdev_priv - access network device private data
@@ -827,7 +825,6 @@ struct packet_type {
 extern rwlock_t				dev_base_lock;		/* Device list lock */
 
 
-#ifdef CONFIG_NET
 #define for_each_netdev(net, d)		\
 		list_for_each_entry(d, &(net)->dev_base_head, dev_list)
 #define for_each_netdev_safe(net, d, n)	\
@@ -851,7 +848,6 @@ static inline struct net_device *first_net_device(struct net *net)
 	return list_empty(&net->dev_base_head) ? NULL :
 		net_device_entry(net->dev_base_head.next);
 }
-#endif
 
 extern int 			netdev_boot_setup_check(struct net_device *dev);
 extern unsigned long		netdev_boot_base(const char *prefix, int unit);
diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index 6c9a48a46685..0ab62ed2fdef 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -60,11 +60,12 @@ struct net {
 };
 
 
-#ifdef CONFIG_NET
 #include <linux/seq_file_net.h>
 
 /* Init's network namespace */
 extern struct net init_net;
+
+#ifdef CONFIG_NET
 #define INIT_NET_NS(net_ns) .net_ns = &init_net,
 
 extern struct net *copy_net_ns(unsigned long flags, struct net *net_ns);
diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
index 0d56dad319ad..b06185ed1895 100644
--- a/lib/kobject_uevent.c
+++ b/lib/kobject_uevent.c
@@ -19,12 +19,10 @@
 #include <linux/kobject.h>
 #include <linux/module.h>
 
-#ifdef CONFIG_NET
 #include <linux/socket.h>
 #include <linux/skbuff.h>
 #include <linux/netlink.h>
 #include <net/sock.h>
-#endif
 
 
 u64 uevent_seqnum;
-- 
cgit v1.2.3


From 2e8046271f68198dd37451017c1a4a2432e4ec68 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Thu, 3 Apr 2008 09:22:09 +0900
Subject: [IPV4] MROUTE: Move PIM definitions to <linux/pim.h>.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/Kbuild   |  1 +
 include/linux/mroute.h | 22 +---------------------
 include/linux/pim.h    | 29 +++++++++++++++++++++++++++++
 3 files changed, 31 insertions(+), 21 deletions(-)
 create mode 100644 include/linux/pim.h

(limited to 'include/linux')

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 9cdd12a9e843..84736acb4b99 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -289,6 +289,7 @@ unifdef-y += parport.h
 unifdef-y += patchkey.h
 unifdef-y += pci.h
 unifdef-y += personality.h
+unifdef-y += pim.h
 unifdef-y += pktcdvd.h
 unifdef-y += pmu.h
 unifdef-y += poll.h
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index 35a8277ec1bd..c41b4217ae3b 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -3,6 +3,7 @@
 
 #include <linux/sockios.h>
 #include <linux/in.h>
+#include <linux/pim.h>
 
 /*
  *	Based on the MROUTING 3.5 defines primarily to keep
@@ -210,27 +211,6 @@ struct mfc_cache
 #define IGMPMSG_WHOLEPKT	3		/* For PIM Register processing */
 
 #ifdef __KERNEL__
-
-#define PIM_V1_VERSION		__constant_htonl(0x10000000)
-#define PIM_V1_REGISTER		1
-
-#define PIM_VERSION		2
-#define PIM_REGISTER		1
-
-#define PIM_NULL_REGISTER	__constant_htonl(0x40000000)
-
-/* PIMv2 register message header layout (ietf-draft-idmr-pimvsm-v2-00.ps */
-
-struct pimreghdr
-{
-	__u8	type;
-	__u8	reserved;
-	__be16	csum;
-	__be32	flags;
-};
-
-extern int pim_rcv_v1(struct sk_buff *);
-
 struct rtmsg;
 extern int ipmr_get_route(struct sk_buff *skb, struct rtmsg *rtm, int nowait);
 #endif
diff --git a/include/linux/pim.h b/include/linux/pim.h
new file mode 100644
index 000000000000..6f689dc85503
--- /dev/null
+++ b/include/linux/pim.h
@@ -0,0 +1,29 @@
+#ifndef __LINUX_PIM_H
+#define __LINUX_PIM_H
+
+#include <asm/byteorder.h>
+
+/* Message types - V1 */
+#define PIM_V1_VERSION		__constant_htonl(0x10000000)
+#define PIM_V1_REGISTER		1
+
+/* Message types - V2 */
+#define PIM_VERSION		2
+#define PIM_REGISTER		1
+
+#if defined(__KERNEL__)
+#define PIM_NULL_REGISTER	__constant_htonl(0x40000000)
+
+/* PIMv2 register message header layout (ietf-draft-idmr-pimvsm-v2-00.ps */
+struct pimreghdr
+{
+	__u8	type;
+	__u8	reserved;
+	__be16	csum;
+	__be32	flags;
+};
+
+struct sk_buff;
+extern int pim_rcv_v1(struct sk_buff *);
+#endif
+#endif
-- 
cgit v1.2.3


From 80a9492a33dd7d852465625022d56ff76d62174d Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Thu, 3 Apr 2008 09:22:52 +0900
Subject: [IPV4] MROUTE: Adjust include files for user-space.

<linux/mroute.h> needs <linux/types.h>.
Avoid including <linux/in.h> in user-space, which conflicts with
standard <netinet/in.h>.
Add basic struct and constant in <linux/pim.h>.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/mroute.h |  3 +++
 include/linux/pim.h    | 16 ++++++++++++++++
 2 files changed, 19 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index c41b4217ae3b..de4decfa1bfc 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -2,7 +2,10 @@
 #define __LINUX_MROUTE_H
 
 #include <linux/sockios.h>
+#include <linux/types.h>
+#ifdef __KERNEL__
 #include <linux/in.h>
+#endif
 #include <linux/pim.h>
 
 /*
diff --git a/include/linux/pim.h b/include/linux/pim.h
index 6f689dc85503..236ffd317394 100644
--- a/include/linux/pim.h
+++ b/include/linux/pim.h
@@ -3,6 +3,22 @@
 
 #include <asm/byteorder.h>
 
+#ifndef __KERNEL__
+struct pim {
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u8	pim_type:4,		/* PIM message type */
+		pim_ver:4;		/* PIM version */
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	__u8	pim_ver:4;		/* PIM version */
+		pim_type:4;		/* PIM message type */
+#endif
+	__u8	pim_rsv;		/* Reserved */
+	__be16	pim_cksum;		/* Checksum */
+};
+
+#define PIM_MINLEN		8
+#endif
+
 /* Message types - V1 */
 #define PIM_V1_VERSION		__constant_htonl(0x10000000)
 #define PIM_V1_REGISTER		1
-- 
cgit v1.2.3


From 436d34b36202ef724778ded1e9cb10f8c37b32bc Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Wed, 2 Apr 2008 17:28:46 +0900
Subject: libata: uninline atapi_cmd_type()

Uninline atapi_cmd_type().  It doesn't really have to be inline and
more case will be added which need to access unexported libata
variable.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 34 ++++++++++++++++++++++++++++++++++
 include/linux/libata.h    | 22 +---------------------
 2 files changed, 35 insertions(+), 21 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 2db5c9c9ca10..e9b69ba489d2 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -325,6 +325,39 @@ static void ata_force_horkage(struct ata_device *dev)
 	}
 }
 
+/**
+ *	atapi_cmd_type - Determine ATAPI command type from SCSI opcode
+ *	@opcode: SCSI opcode
+ *
+ *	Determine ATAPI command type from @opcode.
+ *
+ *	LOCKING:
+ *	None.
+ *
+ *	RETURNS:
+ *	ATAPI_{READ|WRITE|READ_CD|PASS_THRU|MISC}
+ */
+int atapi_cmd_type(u8 opcode)
+{
+	switch (opcode) {
+	case GPCMD_READ_10:
+	case GPCMD_READ_12:
+		return ATAPI_READ;
+
+	case GPCMD_WRITE_10:
+	case GPCMD_WRITE_12:
+	case GPCMD_WRITE_AND_VERIFY_10:
+		return ATAPI_WRITE;
+
+	case GPCMD_READ_CD:
+	case GPCMD_READ_CD_MSF:
+		return ATAPI_READ_CD;
+
+	default:
+		return ATAPI_MISC;
+	}
+}
+
 /**
  *	ata_tf_to_fis - Convert ATA taskfile to SATA FIS structure
  *	@tf: Taskfile to convert
@@ -7774,6 +7807,7 @@ EXPORT_SYMBOL_GPL(ata_tf_read);
 EXPORT_SYMBOL_GPL(ata_noop_dev_select);
 EXPORT_SYMBOL_GPL(ata_std_dev_select);
 EXPORT_SYMBOL_GPL(sata_print_link_status);
+EXPORT_SYMBOL_GPL(atapi_cmd_type);
 EXPORT_SYMBOL_GPL(ata_tf_to_fis);
 EXPORT_SYMBOL_GPL(ata_tf_from_fis);
 EXPORT_SYMBOL_GPL(ata_pack_xfermask);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index b064bfeb69ee..92c64909ed25 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -849,6 +849,7 @@ extern unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
  */
 extern void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf);
 extern void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
+extern int atapi_cmd_type(u8 opcode);
 extern void ata_tf_to_fis(const struct ata_taskfile *tf,
 			  u8 pmp, int is_cmd, u8 *fis);
 extern void ata_tf_from_fis(const u8 *fis, struct ata_taskfile *tf);
@@ -1379,27 +1380,6 @@ static inline int ata_try_flush_cache(const struct ata_device *dev)
 	       ata_id_has_flush_ext(dev->id);
 }
 
-static inline int atapi_cmd_type(u8 opcode)
-{
-	switch (opcode) {
-	case GPCMD_READ_10:
-	case GPCMD_READ_12:
-		return ATAPI_READ;
-
-	case GPCMD_WRITE_10:
-	case GPCMD_WRITE_12:
-	case GPCMD_WRITE_AND_VERIFY_10:
-		return ATAPI_WRITE;
-
-	case GPCMD_READ_CD:
-	case GPCMD_READ_CD_MSF:
-		return ATAPI_READ_CD;
-
-	default:
-		return ATAPI_MISC;
-	}
-}
-
 static inline unsigned int ac_err_mask(u8 status)
 {
 	if (status & (ATA_BUSY | ATA_DRQ))
-- 
cgit v1.2.3


From e52dcc4899cf1b7601379c31542bd91cd2997a64 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Wed, 2 Apr 2008 17:35:19 +0900
Subject: libata: ATA_12/16 doesn't fall into ATAPI_MISC

SAT passthrus don't really fit into ATAPI_MISC class.  SAT passthru
commands always transfer multiple of 512 bytes and variable length
response is not allowed.  This patch creates a separate category -
ATAPI_PASS_THRU - for these.

This fixes HSM violation on "hdparm -I".

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 5 +++++
 include/linux/libata.h    | 3 ++-
 2 files changed, 7 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index e9b69ba489d2..be95fdb69726 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -353,6 +353,11 @@ int atapi_cmd_type(u8 opcode)
 	case GPCMD_READ_CD_MSF:
 		return ATAPI_READ_CD;
 
+	case ATA_16:
+	case ATA_12:
+		if (atapi_passthru16)
+			return ATAPI_PASS_THRU;
+		/* fall thru */
 	default:
 		return ATAPI_MISC;
 	}
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 92c64909ed25..37ee881c42ac 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -350,7 +350,8 @@ enum {
 	ATAPI_READ		= 0,		/* READs */
 	ATAPI_WRITE		= 1,		/* WRITEs */
 	ATAPI_READ_CD		= 2,		/* READ CD [MSF] */
-	ATAPI_MISC		= 3,		/* the rest */
+	ATAPI_PASS_THRU		= 3,		/* SAT pass-thru */
+	ATAPI_MISC		= 4,		/* the rest */
 };
 
 enum ata_xfer_mask {
-- 
cgit v1.2.3


From 5761d64b277c287a7520b868c32d656ef03374b4 Mon Sep 17 00:00:00 2001
From: Thomas Gleixner <tglx@linutronix.de>
Date: Fri, 4 Apr 2008 16:26:10 +0200
Subject: x86: revert assign IRQs to hpet timer

The commits:

commit 37a47db8d7f0f38dac5acf5a13abbc8f401707fa
Author: Balaji Rao <balajirrao@gmail.com>
Date:   Wed Jan 30 13:30:03 2008 +0100

    x86: assign IRQs to HPET timers, fix

and

commit e3f37a54f690d3e64995ea7ecea08c5ab3070faf
Author: Balaji Rao <balajirrao@gmail.com>
Date:   Wed Jan 30 13:30:03 2008 +0100

    x86: assign IRQs to HPET timers

have been identified to cause a regression on some platforms due to
the assignement of legacy IRQs which makes the legacy devices
connected to those IRQs disfunctional.

Revert them.

This fixes http://bugzilla.kernel.org/show_bug.cgi?id=10382

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/kernel/hpet.c |  9 ++++++---
 drivers/char/hpet.c    | 51 +++++++-------------------------------------------
 include/linux/hpet.h   |  2 +-
 3 files changed, 14 insertions(+), 48 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c
index 235fd6c77504..36652ea1a265 100644
--- a/arch/x86/kernel/hpet.c
+++ b/arch/x86/kernel/hpet.c
@@ -133,13 +133,16 @@ static void hpet_reserve_platform_timers(unsigned long id)
 #ifdef CONFIG_HPET_EMULATE_RTC
 	hpet_reserve_timer(&hd, 1);
 #endif
+
 	hd.hd_irq[0] = HPET_LEGACY_8254;
 	hd.hd_irq[1] = HPET_LEGACY_RTC;
 
-       for (i = 2; i < nrtimers; timer++, i++)
-	       hd.hd_irq[i] = (timer->hpet_config & Tn_INT_ROUTE_CNF_MASK) >>
-		       Tn_INT_ROUTE_CNF_SHIFT;
+	for (i = 2; i < nrtimers; timer++, i++)
+		hd.hd_irq[i] = (timer->hpet_config & Tn_INT_ROUTE_CNF_MASK) >>
+			Tn_INT_ROUTE_CNF_SHIFT;
+
 	hpet_alloc(&hd);
+
 }
 #else
 static void hpet_reserve_platform_timers(unsigned long id) { }
diff --git a/drivers/char/hpet.c b/drivers/char/hpet.c
index 465ad35ed38f..1399971be689 100644
--- a/drivers/char/hpet.c
+++ b/drivers/char/hpet.c
@@ -731,14 +731,14 @@ static unsigned long hpet_calibrate(struct hpets *hpetp)
 
 int hpet_alloc(struct hpet_data *hdp)
 {
-	u64 cap, mcfg, hpet_config;
+	u64 cap, mcfg;
 	struct hpet_dev *devp;
-	u32 i, ntimer, irq;
+	u32 i, ntimer;
 	struct hpets *hpetp;
 	size_t siz;
 	struct hpet __iomem *hpet;
 	static struct hpets *last = NULL;
-	unsigned long period, irq_bitmap;
+	unsigned long period;
 	unsigned long long temp;
 
 	/*
@@ -765,47 +765,11 @@ int hpet_alloc(struct hpet_data *hdp)
 	hpetp->hp_hpet_phys = hdp->hd_phys_address;
 
 	hpetp->hp_ntimer = hdp->hd_nirqs;
-	hpet = hpetp->hp_hpet;
-
-	/* Assign IRQs statically for legacy devices */
-	hpetp->hp_dev[0].hd_hdwirq = hdp->hd_irq[0];
-	hpetp->hp_dev[1].hd_hdwirq = hdp->hd_irq[1];
-
-	/* Assign IRQs dynamically for the others */
-	for (i = 2, devp = &hpetp->hp_dev[2]; i < hdp->hd_nirqs; i++, devp++) {
-		struct hpet_timer __iomem *timer;
 
-		timer = &hpet->hpet_timers[devp - hpetp->hp_dev];
+	for (i = 0; i < hdp->hd_nirqs; i++)
+		hpetp->hp_dev[i].hd_hdwirq = hdp->hd_irq[i];
 
-		/* Check if there's already an IRQ assigned to the timer */
-		if (hdp->hd_irq[i]) {
-			hpetp->hp_dev[i].hd_hdwirq = hdp->hd_irq[i];
-			continue;
-		}
-
-		hpet_config = readq(&timer->hpet_config);
-		irq_bitmap = (hpet_config & Tn_INT_ROUTE_CAP_MASK)
-			>> Tn_INT_ROUTE_CAP_SHIFT;
-		if (!irq_bitmap)
-			irq = 0;        /* No valid IRQ Assignable */
-		else {
-			irq = find_first_bit(&irq_bitmap, 32);
-			do {
-				hpet_config |= irq << Tn_INT_ROUTE_CNF_SHIFT;
-				writeq(hpet_config, &timer->hpet_config);
-
-				/*
-				 * Verify whether we have written a valid
-				 * IRQ number by reading it back again
-				 */
-				hpet_config = readq(&timer->hpet_config);
-				if (irq == (hpet_config & Tn_INT_ROUTE_CNF_MASK)
-						>> Tn_INT_ROUTE_CNF_SHIFT)
-					break;  /* Success */
-			} while ((irq = (find_next_bit(&irq_bitmap, 32, irq))));
-		}
-		hpetp->hp_dev[i].hd_hdwirq = irq;
-	}
+	hpet = hpetp->hp_hpet;
 
 	cap = readq(&hpet->hpet_cap);
 
@@ -836,8 +800,7 @@ int hpet_alloc(struct hpet_data *hdp)
 		hpetp->hp_which, hdp->hd_phys_address,
 		hpetp->hp_ntimer > 1 ? "s" : "");
 	for (i = 0; i < hpetp->hp_ntimer; i++)
-		printk("%s %d", i > 0 ? "," : "",
-				hpetp->hp_dev[i].hd_hdwirq);
+		printk("%s %d", i > 0 ? "," : "", hdp->hd_irq[i]);
 	printk("\n");
 
 	printk(KERN_INFO "hpet%u: %u %d-bit timers, %Lu Hz\n",
diff --git a/include/linux/hpet.h b/include/linux/hpet.h
index 9cd94bfd07e5..2dc29ce6c8e4 100644
--- a/include/linux/hpet.h
+++ b/include/linux/hpet.h
@@ -64,7 +64,7 @@ struct hpet {
  */
 
 #define	Tn_INT_ROUTE_CAP_MASK		(0xffffffff00000000ULL)
-#define	Tn_INT_ROUTE_CAP_SHIFT		(32UL)
+#define	Tn_INI_ROUTE_CAP_SHIFT		(32UL)
 #define	Tn_FSB_INT_DELCAP_MASK		(0x8000UL)
 #define	Tn_FSB_INT_DELCAP_SHIFT		(15)
 #define	Tn_FSB_EN_CNF_MASK		(0x4000UL)
-- 
cgit v1.2.3


From 8bab8dded67d026c39367bbd5e27d2f6c556c38e Mon Sep 17 00:00:00 2001
From: Paul Menage <menage@google.com>
Date: Fri, 4 Apr 2008 14:29:57 -0700
Subject: cgroups: add cgroup support for enabling controllers at boot time

The effects of cgroup_disable=foo are:

- foo isn't auto-mounted if you mount all cgroups in a single hierarchy
- foo isn't visible as an individually mountable subsystem

As a result there will only ever be one call to foo->create(), at init time;
all processes will stay in this group, and the group will never be mounted on
a visible hierarchy.  Any additional effects (e.g.  not allocating metadata)
are up to the foo subsystem.

This doesn't handle early_init subsystems (their "disabled" bit isn't set be,
but it could easily be extended to do so if any of the early_init systems
wanted it - I think it would just involve some nastier parameter processing
since it would occur before the command-line argument parser had been run.

Hugh said:

  Ballpark figures, I'm trying to get this question out rather than
  processing the exact numbers: CONFIG_CGROUP_MEM_RES_CTLR adds 15% overhead
  to the affected paths, booting with cgroup_disable=memory cuts that back to
  1% overhead (due to slightly bigger struct page).

  I'm no expert on distros, they may have no interest whatever in
  CONFIG_CGROUP_MEM_RES_CTLR=y; and the rest of us can easily build with or
  without it, or apply the cgroup_disable=memory patches.

Unix bench's execl test result on x86_64 was

== just after boot without mounting any cgroup fs.==
mem_cgorup=off : Execl Throughput       43.0     3150.1      732.6
mem_cgroup=on  : Execl Throughput       43.0     2932.6      682.0
==

[lizf@cn.fujitsu.com: fix boot option parsing]
Signed-off-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Paul Menage <menage@google.com>
Cc: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: Sudhir Kumar <skumar@linux.vnet.ibm.com>
Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 Documentation/kernel-parameters.txt |  4 ++++
 include/linux/cgroup.h              |  1 +
 kernel/cgroup.c                     | 42 ++++++++++++++++++++++++++++++++-----
 3 files changed, 42 insertions(+), 5 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 4cd1a5da80a4..32e9297ef747 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -375,6 +375,10 @@ and is between 256 and 4096 characters. It is defined in the file
 	ccw_timeout_log [S390]
 			See Documentation/s390/CommonIO for details.
 
+	cgroup_disable= [KNL] Disable a particular controller
+			Format: {name of the controller(s) to disable}
+				{Currently supported controllers - "memory"}
+
 	checkreqprot	[SELINUX] Set initial checkreqprot flag value.
 			Format: { "0" | "1" }
 			See security/selinux/Kconfig help text.
diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index 028ba3b523b1..a6a6035a4e1e 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -256,6 +256,7 @@ struct cgroup_subsys {
 	void (*bind)(struct cgroup_subsys *ss, struct cgroup *root);
 	int subsys_id;
 	int active;
+	int disabled;
 	int early_init;
 #define MAX_CGROUP_TYPE_NAMELEN 32
 	const char *name;
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 53d86b4b0ce0..62f1a5231fe9 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -782,7 +782,14 @@ static int parse_cgroupfs_options(char *data,
 		if (!*token)
 			return -EINVAL;
 		if (!strcmp(token, "all")) {
-			opts->subsys_bits = (1 << CGROUP_SUBSYS_COUNT) - 1;
+			/* Add all non-disabled subsystems */
+			int i;
+			opts->subsys_bits = 0;
+			for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+				struct cgroup_subsys *ss = subsys[i];
+				if (!ss->disabled)
+					opts->subsys_bits |= 1ul << i;
+			}
 		} else if (!strcmp(token, "noprefix")) {
 			set_bit(ROOT_NOPREFIX, &opts->flags);
 		} else if (!strncmp(token, "release_agent=", 14)) {
@@ -800,7 +807,8 @@ static int parse_cgroupfs_options(char *data,
 			for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
 				ss = subsys[i];
 				if (!strcmp(token, ss->name)) {
-					set_bit(i, &opts->subsys_bits);
+					if (!ss->disabled)
+						set_bit(i, &opts->subsys_bits);
 					break;
 				}
 			}
@@ -2600,13 +2608,13 @@ static int proc_cgroupstats_show(struct seq_file *m, void *v)
 {
 	int i;
 
-	seq_puts(m, "#subsys_name\thierarchy\tnum_cgroups\n");
+	seq_puts(m, "#subsys_name\thierarchy\tnum_cgroups\tenabled\n");
 	mutex_lock(&cgroup_mutex);
 	for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
 		struct cgroup_subsys *ss = subsys[i];
-		seq_printf(m, "%s\t%lu\t%d\n",
+		seq_printf(m, "%s\t%lu\t%d\t%d\n",
 			   ss->name, ss->root->subsys_bits,
-			   ss->root->number_of_cgroups);
+			   ss->root->number_of_cgroups, !ss->disabled);
 	}
 	mutex_unlock(&cgroup_mutex);
 	return 0;
@@ -3010,3 +3018,27 @@ static void cgroup_release_agent(struct work_struct *work)
 	spin_unlock(&release_list_lock);
 	mutex_unlock(&cgroup_mutex);
 }
+
+static int __init cgroup_disable(char *str)
+{
+	int i;
+	char *token;
+
+	while ((token = strsep(&str, ",")) != NULL) {
+		if (!*token)
+			continue;
+
+		for (i = 0; i < CGROUP_SUBSYS_COUNT; i++) {
+			struct cgroup_subsys *ss = subsys[i];
+
+			if (!strcmp(token, ss->name)) {
+				ss->disabled = 1;
+				printk(KERN_INFO "Disabling %s control group"
+					" subsystem\n", ss->name);
+				break;
+			}
+		}
+	}
+	return 1;
+}
+__setup("cgroup_disable=", cgroup_disable);
-- 
cgit v1.2.3


From 7bc570c8b4f75ddb3fd5dbeb38127cdc4acbcc9c Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Thu, 3 Apr 2008 09:22:53 +0900
Subject: [IPV6] MROUTE: Support multicast forwarding.

Based on ancient patch by Mickael Hoerdt
<hoerdt@clarinet.u-strasbg.fr>, which is available at
<http://www-r2.u-strasbg.fr/~hoerdt/dev/linux_ipv6_mforwarding/patch-linux-ipv6-mforwarding-0.1a>.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/Kbuild     |    1 +
 include/linux/ipv6.h     |    5 +
 include/linux/mroute6.h  |  227 ++++++++
 net/ipv6/Kconfig         |    7 +
 net/ipv6/Makefile        |    2 +
 net/ipv6/addrconf.c      |   15 +-
 net/ipv6/af_inet6.c      |    6 +
 net/ipv6/ip6_input.c     |   87 ++-
 net/ipv6/ip6_output.c    |    6 +-
 net/ipv6/ip6mr.c         | 1384 ++++++++++++++++++++++++++++++++++++++++++++++
 net/ipv6/ipv6_sockglue.c |    7 +
 net/ipv6/raw.c           |    7 +-
 net/ipv6/route.c         |   30 +-
 13 files changed, 1754 insertions(+), 30 deletions(-)
 create mode 100644 include/linux/mroute6.h
 create mode 100644 net/ipv6/ip6mr.c

(limited to 'include/linux')

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 84736acb4b99..29ab9b95d376 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -261,6 +261,7 @@ unifdef-y += mempolicy.h
 unifdef-y += mii.h
 unifdef-y += mman.h
 unifdef-y += mroute.h
+unifdef-y += mroute6.h
 unifdef-y += msdos_fs.h
 unifdef-y += msg.h
 unifdef-y += nbd.h
diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index b90d3d461d4e..f53e4764fc05 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -159,6 +159,9 @@ struct ipv6_devconf {
 	__s32		accept_source_route;
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
 	__s32		optimistic_dad;
+#endif
+#ifdef CONFIG_IPV6_MROUTE
+	__s32		mc_forwarding;
 #endif
 	void		*sysctl;
 };
@@ -190,6 +193,7 @@ enum {
 	DEVCONF_PROXY_NDP,
 	DEVCONF_OPTIMISTIC_DAD,
 	DEVCONF_ACCEPT_SOURCE_ROUTE,
+	DEVCONF_MC_FORWARDING,
 	DEVCONF_MAX
 };
 
@@ -230,6 +234,7 @@ struct inet6_skb_parm {
 #endif
 
 #define IP6SKB_XFRM_TRANSFORMED	1
+#define IP6SKB_FORWARDED	2
 };
 
 #define IP6CB(skb)	((struct inet6_skb_parm*)((skb)->cb))
diff --git a/include/linux/mroute6.h b/include/linux/mroute6.h
new file mode 100644
index 000000000000..b92190304e0b
--- /dev/null
+++ b/include/linux/mroute6.h
@@ -0,0 +1,227 @@
+#ifndef __LINUX_MROUTE6_H
+#define __LINUX_MROUTE6_H
+
+#include <linux/types.h>
+#include <linux/sockios.h>
+
+/*
+ *	Based on the MROUTING 3.5 defines primarily to keep
+ *	source compatibility with BSD.
+ *
+ *	See the pim6sd code for the original history.
+ *
+ *      Protocol Independent Multicast (PIM) data structures included
+ *      Carlos Picoto (cap@di.fc.ul.pt)
+ *
+ */
+
+#define MRT6_BASE	200
+#define MRT6_INIT	(MRT6_BASE)	/* Activate the kernel mroute code 	*/
+#define MRT6_DONE	(MRT6_BASE+1)	/* Shutdown the kernel mroute		*/
+#define MRT6_ADD_MIF	(MRT6_BASE+2)	/* Add a virtual interface		*/
+#define MRT6_DEL_MIF	(MRT6_BASE+3)	/* Delete a virtual interface		*/
+#define MRT6_ADD_MFC	(MRT6_BASE+4)	/* Add a multicast forwarding entry	*/
+#define MRT6_DEL_MFC	(MRT6_BASE+5)	/* Delete a multicast forwarding entry	*/
+#define MRT6_VERSION	(MRT6_BASE+6)	/* Get the kernel multicast version	*/
+
+#define SIOCGETMIFCNT_IN6	SIOCPROTOPRIVATE	/* IP protocol privates */
+#define SIOCGETSGCNT_IN6	(SIOCPROTOPRIVATE+1)
+#define SIOCGETRPF	(SIOCPROTOPRIVATE+2)
+
+#define MAXMIFS		32
+typedef unsigned long mifbitmap_t;	/* User mode code depends on this lot */
+typedef unsigned short mifi_t;
+#define ALL_MIFS	((mifi_t)(-1))
+
+#ifndef IF_SETSIZE
+#define IF_SETSIZE	256
+#endif
+
+typedef	__u32		if_mask;
+#define NIFBITS (sizeof(if_mask) * 8)        /* bits per mask */
+
+#if !defined(__KERNEL__) && !defined(DIV_ROUND_UP)
+#define	DIV_ROUND_UP(x,y)	(((x) + ((y) - 1)) / (y))
+#endif
+
+typedef struct if_set {
+	if_mask ifs_bits[DIV_ROUND_UP(IF_SETSIZE, NIFBITS)];
+} if_set;
+
+#define IF_SET(n, p)    ((p)->ifs_bits[(n)/NIFBITS] |= (1 << ((n) % NIFBITS)))
+#define IF_CLR(n, p)    ((p)->ifs_bits[(n)/NIFBITS] &= ~(1 << ((n) % NIFBITS)))
+#define IF_ISSET(n, p)  ((p)->ifs_bits[(n)/NIFBITS] & (1 << ((n) % NIFBITS)))
+#define IF_COPY(f, t)   bcopy(f, t, sizeof(*(f)))
+#define IF_ZERO(p)      bzero(p, sizeof(*(p)))
+
+/*
+ *	Passed by mrouted for an MRT_ADD_MIF - again we use the
+ *	mrouted 3.6 structures for compatibility
+ */
+
+struct mif6ctl {
+	mifi_t	mif6c_mifi;		/* Index of MIF */
+	unsigned char mif6c_flags;	/* MIFF_ flags */
+	unsigned char vifc_threshold;	/* ttl limit */
+	u_short	 mif6c_pifi;		/* the index of the physical IF */
+	unsigned int vifc_rate_limit;	/* Rate limiter values (NI) */
+};
+
+#define MIFF_REGISTER	0x1	/* register vif	*/
+
+/*
+ *	Cache manipulation structures for mrouted and PIMd
+ */
+
+struct mf6cctl
+{
+	struct sockaddr_in6 mf6cc_origin;		/* Origin of mcast	*/
+	struct sockaddr_in6 mf6cc_mcastgrp;		/* Group in question	*/
+	mifi_t	mf6cc_parent;			/* Where it arrived	*/
+	struct if_set mf6cc_ifset;		/* Where it is going */
+};
+
+/*
+ *	Group count retrieval for pim6sd
+ */
+
+struct sioc_sg_req6
+{
+	struct sockaddr_in6 src;
+	struct sockaddr_in6 grp;
+	unsigned long pktcnt;
+	unsigned long bytecnt;
+	unsigned long wrong_if;
+};
+
+/*
+ *	To get vif packet counts
+ */
+
+struct sioc_mif_req6
+{
+	mifi_t	mifi;		/* Which iface */
+	unsigned long icount;	/* In packets */
+	unsigned long ocount;	/* Out packets */
+	unsigned long ibytes;	/* In bytes */
+	unsigned long obytes;	/* Out bytes */
+};
+
+/*
+ *	That's all usermode folks
+ */
+
+#ifdef __KERNEL__
+
+#include <linux/skbuff.h>	/* for struct sk_buff_head */
+
+struct net_device;
+struct inet6_dev *ipv6_find_idev(struct net_device *dev);
+
+#ifdef CONFIG_IPV6_MROUTE
+static inline int ip6_mroute_opt(int opt)
+{
+	return (opt >= MRT6_BASE) && (opt <= MRT6_BASE + 10);
+}
+#else
+static inline int ip6_mroute_opt(int opt)
+{
+	return 0;
+}
+#endif
+
+struct sock;
+
+extern int ip6_mroute_setsockopt(struct sock *, int, char __user *, int);
+extern int ip6_mroute_getsockopt(struct sock *, int, char __user *, int __user *);
+extern int ip6_mr_input(struct sk_buff *skb);
+extern int ip6mr_ioctl(struct sock *sk, int cmd, void __user *arg);
+extern void ip6_mr_init(void);
+
+struct mif_device
+{
+	struct net_device 	*dev;			/* Device we are using */
+	unsigned long	bytes_in,bytes_out;
+	unsigned long	pkt_in,pkt_out;		/* Statistics 			*/
+	unsigned long	rate_limit;		/* Traffic shaping (NI) 	*/
+	unsigned char	threshold;		/* TTL threshold 		*/
+	unsigned short	flags;			/* Control flags 		*/
+	int		link;			/* Physical interface index	*/
+};
+
+#define VIFF_STATIC 0x8000
+
+struct mfc6_cache
+{
+	struct mfc6_cache *next;		/* Next entry on cache line 	*/
+	struct in6_addr mf6c_mcastgrp;			/* Group the entry belongs to 	*/
+	struct in6_addr mf6c_origin;			/* Source of packet 		*/
+	mifi_t mf6c_parent;			/* Source interface		*/
+	int mfc_flags;				/* Flags on line		*/
+
+	union {
+		struct {
+			unsigned long expires;
+			struct sk_buff_head unresolved;	/* Unresolved buffers		*/
+		} unres;
+		struct {
+			unsigned long last_assert;
+			int minvif;
+			int maxvif;
+			unsigned long bytes;
+			unsigned long pkt;
+			unsigned long wrong_if;
+			unsigned char ttls[MAXMIFS];	/* TTL thresholds		*/
+		} res;
+	} mfc_un;
+};
+
+#define MFC_STATIC		1
+#define MFC_NOTIFY		2
+
+#define MFC6_LINES		64
+
+#define MFC6_HASH(a, g) (((__force u32)(a)->s6_addr32[0] ^ \
+			  (__force u32)(a)->s6_addr32[1] ^ \
+			  (__force u32)(a)->s6_addr32[2] ^ \
+			  (__force u32)(a)->s6_addr32[3] ^ \
+			  (__force u32)(g)->s6_addr32[0] ^ \
+			  (__force u32)(g)->s6_addr32[1] ^ \
+			  (__force u32)(g)->s6_addr32[2] ^ \
+			  (__force u32)(g)->s6_addr32[3]) % MFC6_LINES)
+
+#define MFC_ASSERT_THRESH (3*HZ)		/* Maximal freq. of asserts */
+
+#endif
+
+#ifdef __KERNEL__
+struct rtmsg;
+extern int ip6mr_get_route(struct sk_buff *skb, struct rtmsg *rtm, int nowait);
+
+#ifdef CONFIG_IPV6_MROUTE
+extern struct sock *mroute6_socket;
+extern int ip6mr_sk_done(struct sock *sk);
+#else
+#define mroute6_socket NULL
+static inline int ip6mr_sk_done(struct sock *sk) { return 0; }
+#endif
+#endif
+
+/*
+ * Structure used to communicate from kernel to multicast router.
+ * We'll overlay the structure onto an MLD header (not an IPv6 heder like igmpmsg{}
+ * used for IPv4 implementation). This is because this structure will be passed via an
+ * IPv6 raw socket, on wich an application will only receiver the payload i.e the data after
+ * the IPv6 header and all the extension headers. (See section 3 of RFC 3542)
+ */
+
+struct mrt6msg {
+#define MRT6MSG_NOCACHE		1
+	__u8		im6_mbz;		/* must be zero		   */
+	__u8		im6_msgtype;		/* what type of message    */
+	__u16		im6_mif;		/* mif rec'd on		   */
+	__u32		im6_pad;		/* padding for 64 bit arch */
+	struct in6_addr	im6_src, im6_dst;
+};
+
+#endif
diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index 47263e45bacb..9a2ea81e499f 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -209,3 +209,10 @@ config IPV6_SUBTREES
 
 	  If unsure, say N.
 
+config IPV6_MROUTE
+	bool "IPv6: multicast routing (EXPERIMENTAL)"
+	depends on IPV6 && EXPERIMENTAL
+	---help---
+	  Experimental support for IPv6 multicast forwarding.
+	  If unsure, say N.
+
diff --git a/net/ipv6/Makefile b/net/ipv6/Makefile
index ae14617e607f..686934acfac1 100644
--- a/net/ipv6/Makefile
+++ b/net/ipv6/Makefile
@@ -11,6 +11,8 @@ ipv6-objs :=	af_inet6.o anycast.o ip6_output.o ip6_input.o addrconf.o \
 		exthdrs.o datagram.o ip6_flowlabel.o inet6_connection_sock.o
 
 ipv6-$(CONFIG_SYSCTL) = sysctl_net_ipv6.o
+ipv6-$(CONFIG_IPV6_MROUTE) += ip6mr.o
+
 ipv6-$(CONFIG_XFRM) += xfrm6_policy.o xfrm6_state.o xfrm6_input.o \
 	xfrm6_output.o
 ipv6-$(CONFIG_NETFILTER) += netfilter.o
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 11037615bc73..dbc51af69017 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -412,7 +412,7 @@ static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 	return ndev;
 }
 
-static struct inet6_dev * ipv6_find_idev(struct net_device *dev)
+struct inet6_dev * ipv6_find_idev(struct net_device *dev)
 {
 	struct inet6_dev *idev;
 
@@ -3547,6 +3547,9 @@ static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
 	array[DEVCONF_OPTIMISTIC_DAD] = cnf->optimistic_dad;
 #endif
+#ifdef CONFIG_IPV6_MROUTE
+	array[DEVCONF_MC_FORWARDING] = cnf->mc_forwarding;
+#endif
 }
 
 static inline size_t inet6_if_nlmsg_size(void)
@@ -4094,6 +4097,16 @@ static struct addrconf_sysctl_table
 			.proc_handler   =       &proc_dointvec,
 
 		},
+#endif
+#ifdef CONFIG_IPV6_MROUTE
+		{
+			.ctl_name	=	CTL_UNNUMBERED,
+			.procname	=	"mc_forwarding",
+			.data		=	&ipv6_devconf.mc_forwarding,
+			.maxlen		=	sizeof(int),
+			.mode		=	0644,
+			.proc_handler	=	&proc_dointvec,
+		},
 #endif
 		{
 			.ctl_name	=	0,	/* sentinel */
diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c
index 1731b0abf7f5..3c6aafb02183 100644
--- a/net/ipv6/af_inet6.c
+++ b/net/ipv6/af_inet6.c
@@ -61,6 +61,9 @@
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
+#ifdef CONFIG_IPV6_MROUTE
+#include <linux/mroute6.h>
+#endif
 
 MODULE_AUTHOR("Cast of dozens");
 MODULE_DESCRIPTION("IPv6 protocol stack for Linux");
@@ -953,6 +956,9 @@ static int __init inet6_init(void)
 	err = icmpv6_init();
 	if (err)
 		goto icmp_fail;
+#ifdef CONFIG_IPV6_MROUTE
+	ip6_mr_init();
+#endif
 	err = ndisc_init();
 	if (err)
 		goto ndisc_fail;
diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c
index 43a617e2268b..09a3201e408a 100644
--- a/net/ipv6/ip6_input.c
+++ b/net/ipv6/ip6_input.c
@@ -29,6 +29,7 @@
 #include <linux/netdevice.h>
 #include <linux/in6.h>
 #include <linux/icmpv6.h>
+#include <linux/mroute6.h>
 
 #include <linux/netfilter.h>
 #include <linux/netfilter_ipv6.h>
@@ -236,36 +237,84 @@ int ip6_mc_input(struct sk_buff *skb)
 	hdr = ipv6_hdr(skb);
 	deliver = ipv6_chk_mcast_addr(skb->dev, &hdr->daddr, NULL);
 
+#ifdef CONFIG_IPV6_MROUTE
 	/*
-	 *	IPv6 multicast router mode isnt currently supported.
+	 *      IPv6 multicast router mode is now supported ;)
 	 */
-#if 0
-	if (ipv6_config.multicast_route) {
-		int addr_type;
-
-		addr_type = ipv6_addr_type(&hdr->daddr);
-
-		if (!(addr_type & (IPV6_ADDR_LOOPBACK | IPV6_ADDR_LINKLOCAL))) {
-			struct sk_buff *skb2;
-			struct dst_entry *dst;
+	if (ipv6_devconf.mc_forwarding &&
+	    likely(!(IP6CB(skb)->flags & IP6SKB_FORWARDED))) {
+		/*
+		 * Okay, we try to forward - split and duplicate
+		 * packets.
+		 */
+		struct sk_buff *skb2;
+		struct inet6_skb_parm *opt = IP6CB(skb);
+
+		/* Check for MLD */
+		if (unlikely(opt->ra)) {
+			/* Check if this is a mld message */
+			u8 *ptr = skb_network_header(skb) + opt->ra;
+			struct icmp6hdr *icmp6;
+			u8 nexthdr = hdr->nexthdr;
+			int offset;
+
+			/* Check if the value of Router Alert
+			 * is for MLD (0x0000).
+			 */
+			if ((ptr[2] | ptr[3]) == 0) {
+				if (!ipv6_ext_hdr(nexthdr)) {
+					/* BUG */
+					goto discard;
+				}
+				offset = ipv6_skip_exthdr(skb, sizeof(*hdr),
+							  &nexthdr);
+				if (offset < 0)
+					goto discard;
+
+				if (nexthdr != IPPROTO_ICMPV6)
+					goto discard;
+
+				if (!pskb_may_pull(skb, (skb_network_header(skb) +
+						   offset + 1 - skb->data)))
+					goto discard;
+
+				icmp6 = (struct icmp6hdr *)(skb_network_header(skb) + offset);
+
+				switch (icmp6->icmp6_type) {
+				case ICMPV6_MGM_QUERY:
+				case ICMPV6_MGM_REPORT:
+				case ICMPV6_MGM_REDUCTION:
+				case ICMPV6_MLD2_REPORT:
+					break;
+				default:
+					/* Bogus */
+					goto discard;
+				}
+				deliver = 1;
+				goto out;
+			}
+			/* unknown RA - process it normally */
+		}
 
-			dst = skb->dst;
+		if (deliver)
+			skb2 = skb_clone(skb, GFP_ATOMIC);
+		else {
+			skb2 = skb;
+			skb = NULL;
+		}
 
-			if (deliver) {
-				skb2 = skb_clone(skb, GFP_ATOMIC);
-				dst_output(skb2);
-			} else {
-				dst_output(skb);
-				return 0;
-			}
+		if (skb2) {
+			skb2->dev = skb2->dst->dev;
+			ip6_mr_input(skb2);
 		}
 	}
 #endif
-
+out:
 	if (likely(deliver)) {
 		ip6_input(skb);
 		return 0;
 	}
+discard:
 	/* discard */
 	kfree_skb(skb);
 
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index a8b4da25b0a7..c0dbe549cc42 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -55,6 +55,7 @@
 #include <net/icmp.h>
 #include <net/xfrm.h>
 #include <net/checksum.h>
+#include <linux/mroute6.h>
 
 static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));
 
@@ -137,8 +138,9 @@ static int ip6_output2(struct sk_buff *skb)
 		struct inet6_dev *idev = ip6_dst_idev(skb->dst);
 
 		if (!(dev->flags & IFF_LOOPBACK) && (!np || np->mc_loop) &&
-		    ipv6_chk_mcast_addr(dev, &ipv6_hdr(skb)->daddr,
-					&ipv6_hdr(skb)->saddr)) {
+		    ((mroute6_socket && !(IP6CB(skb)->flags & IP6SKB_FORWARDED)) ||
+		     ipv6_chk_mcast_addr(dev, &ipv6_hdr(skb)->daddr,
+					 &ipv6_hdr(skb)->saddr))) {
 			struct sk_buff *newskb = skb_clone(skb, GFP_ATOMIC);
 
 			/* Do not check for IFF_ALLMULTI; multicast routing
diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c
new file mode 100644
index 000000000000..1bdf3c177d58
--- /dev/null
+++ b/net/ipv6/ip6mr.c
@@ -0,0 +1,1384 @@
+/*
+ *	Linux IPv6 multicast routing support for BSD pim6sd
+ *	Based on net/ipv4/ipmr.c.
+ *
+ *	(c) 2004 Mickael Hoerdt, <hoerdt@clarinet.u-strasbg.fr>
+ *		LSIIT Laboratory, Strasbourg, France
+ *	(c) 2004 Jean-Philippe Andriot, <jean-philippe.andriot@6WIND.com>
+ *		6WIND, Paris, France
+ *	Copyright (C)2007,2008 USAGI/WIDE Project
+ *		YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ *
+ */
+
+#include <asm/system.h>
+#include <asm/uaccess.h>
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/errno.h>
+#include <linux/timer.h>
+#include <linux/mm.h>
+#include <linux/kernel.h>
+#include <linux/fcntl.h>
+#include <linux/stat.h>
+#include <linux/socket.h>
+#include <linux/in.h>
+#include <linux/inet.h>
+#include <linux/netdevice.h>
+#include <linux/inetdevice.h>
+#include <linux/igmp.h>
+#include <linux/proc_fs.h>
+#include <linux/seq_file.h>
+#include <linux/mroute.h>
+#include <linux/init.h>
+#include <net/ip.h>
+#include <net/protocol.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/icmp.h>
+#include <net/udp.h>
+#include <net/raw.h>
+#include <net/route.h>
+#include <linux/notifier.h>
+#include <linux/if_arp.h>
+#include <linux/netfilter_ipv4.h>
+#include <net/ipip.h>
+#include <net/checksum.h>
+#include <net/netlink.h>
+
+#include <net/ipv6.h>
+#include <net/ip6_route.h>
+#include <linux/mroute6.h>
+#include <net/addrconf.h>
+#include <linux/netfilter_ipv6.h>
+
+struct sock *mroute6_socket;
+
+
+/* Big lock, protecting vif table, mrt cache and mroute socket state.
+   Note that the changes are semaphored via rtnl_lock.
+ */
+
+static DEFINE_RWLOCK(mrt_lock);
+
+/*
+ *	Multicast router control variables
+ */
+
+static struct mif_device vif6_table[MAXMIFS];		/* Devices 		*/
+static int maxvif;
+
+#define MIF_EXISTS(idx) (vif6_table[idx].dev != NULL)
+
+static struct mfc6_cache *mfc6_cache_array[MFC_LINES];	/* Forwarding cache	*/
+
+static struct mfc6_cache *mfc_unres_queue;		/* Queue of unresolved entries */
+static atomic_t cache_resolve_queue_len;		/* Size of unresolved	*/
+
+/* Special spinlock for queue of unresolved entries */
+static DEFINE_SPINLOCK(mfc_unres_lock);
+
+/* We return to original Alan's scheme. Hash table of resolved
+   entries is changed only in process context and protected
+   with weak lock mrt_lock. Queue of unresolved entries is protected
+   with strong spinlock mfc_unres_lock.
+
+   In this case data path is free of exclusive locks at all.
+ */
+
+static struct kmem_cache *mrt_cachep __read_mostly;
+
+static int ip6_mr_forward(struct sk_buff *skb, struct mfc6_cache *cache);
+static int ip6mr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert);
+static int ip6mr_fill_mroute(struct sk_buff *skb, struct mfc6_cache *c, struct rtmsg *rtm);
+
+static struct timer_list ipmr_expire_timer;
+
+
+#ifdef CONFIG_PROC_FS
+
+struct ipmr_mfc_iter {
+	struct mfc6_cache **cache;
+	int ct;
+};
+
+
+static struct mfc6_cache *ipmr_mfc_seq_idx(struct ipmr_mfc_iter *it, loff_t pos)
+{
+	struct mfc6_cache *mfc;
+
+	it->cache = mfc6_cache_array;
+	read_lock(&mrt_lock);
+	for (it->ct = 0; it->ct < ARRAY_SIZE(mfc6_cache_array); it->ct++)
+		for (mfc = mfc6_cache_array[it->ct]; mfc; mfc = mfc->next)
+			if (pos-- == 0)
+				return mfc;
+	read_unlock(&mrt_lock);
+
+	it->cache = &mfc_unres_queue;
+	spin_lock_bh(&mfc_unres_lock);
+	for (mfc = mfc_unres_queue; mfc; mfc = mfc->next)
+		if (pos-- == 0)
+			return mfc;
+	spin_unlock_bh(&mfc_unres_lock);
+
+	it->cache = NULL;
+	return NULL;
+}
+
+
+
+
+/*
+ *	The /proc interfaces to multicast routing /proc/ip6_mr_cache /proc/ip6_mr_vif
+ */
+
+struct ipmr_vif_iter {
+	int ct;
+};
+
+static struct mif_device *ip6mr_vif_seq_idx(struct ipmr_vif_iter *iter,
+					    loff_t pos)
+{
+	for (iter->ct = 0; iter->ct < maxvif; ++iter->ct) {
+		if (!MIF_EXISTS(iter->ct))
+			continue;
+		if (pos-- == 0)
+			return &vif6_table[iter->ct];
+	}
+	return NULL;
+}
+
+static void *ip6mr_vif_seq_start(struct seq_file *seq, loff_t *pos)
+	__acquires(mrt_lock)
+{
+	read_lock(&mrt_lock);
+	return (*pos ? ip6mr_vif_seq_idx(seq->private, *pos - 1)
+		: SEQ_START_TOKEN);
+}
+
+static void *ip6mr_vif_seq_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	struct ipmr_vif_iter *iter = seq->private;
+
+	++*pos;
+	if (v == SEQ_START_TOKEN)
+		return ip6mr_vif_seq_idx(iter, 0);
+
+	while (++iter->ct < maxvif) {
+		if (!MIF_EXISTS(iter->ct))
+			continue;
+		return &vif6_table[iter->ct];
+	}
+	return NULL;
+}
+
+static void ip6mr_vif_seq_stop(struct seq_file *seq, void *v)
+	__releases(mrt_lock)
+{
+	read_unlock(&mrt_lock);
+}
+
+static int ip6mr_vif_seq_show(struct seq_file *seq, void *v)
+{
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq,
+			 "Interface      BytesIn  PktsIn  BytesOut PktsOut Flags\n");
+	} else {
+		const struct mif_device *vif = v;
+		const char *name = vif->dev ? vif->dev->name : "none";
+
+		seq_printf(seq,
+			   "%2Zd %-10s %8ld %7ld  %8ld %7ld %05X\n",
+			   vif - vif6_table,
+			   name, vif->bytes_in, vif->pkt_in,
+			   vif->bytes_out, vif->pkt_out,
+			   vif->flags);
+	}
+	return 0;
+}
+
+static struct seq_operations ip6mr_vif_seq_ops = {
+	.start = ip6mr_vif_seq_start,
+	.next  = ip6mr_vif_seq_next,
+	.stop  = ip6mr_vif_seq_stop,
+	.show  = ip6mr_vif_seq_show,
+};
+
+static int ip6mr_vif_open(struct inode *inode, struct file *file)
+{
+	return seq_open_private(file, &ip6mr_vif_seq_ops,
+				sizeof(struct ipmr_vif_iter));
+}
+
+static struct file_operations ip6mr_vif_fops = {
+	.owner	 = THIS_MODULE,
+	.open    = ip6mr_vif_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release,
+};
+
+static void *ipmr_mfc_seq_start(struct seq_file *seq, loff_t *pos)
+{
+	return (*pos ? ipmr_mfc_seq_idx(seq->private, *pos - 1)
+		: SEQ_START_TOKEN);
+}
+
+static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	struct mfc6_cache *mfc = v;
+	struct ipmr_mfc_iter *it = seq->private;
+
+	++*pos;
+
+	if (v == SEQ_START_TOKEN)
+		return ipmr_mfc_seq_idx(seq->private, 0);
+
+	if (mfc->next)
+		return mfc->next;
+
+	if (it->cache == &mfc_unres_queue)
+		goto end_of_list;
+
+	BUG_ON(it->cache != mfc6_cache_array);
+
+	while (++it->ct < ARRAY_SIZE(mfc6_cache_array)) {
+		mfc = mfc6_cache_array[it->ct];
+		if (mfc)
+			return mfc;
+	}
+
+	/* exhausted cache_array, show unresolved */
+	read_unlock(&mrt_lock);
+	it->cache = &mfc_unres_queue;
+	it->ct = 0;
+
+	spin_lock_bh(&mfc_unres_lock);
+	mfc = mfc_unres_queue;
+	if (mfc)
+		return mfc;
+
+ end_of_list:
+	spin_unlock_bh(&mfc_unres_lock);
+	it->cache = NULL;
+
+	return NULL;
+}
+
+static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
+{
+	struct ipmr_mfc_iter *it = seq->private;
+
+	if (it->cache == &mfc_unres_queue)
+		spin_unlock_bh(&mfc_unres_lock);
+	else if (it->cache == mfc6_cache_array)
+		read_unlock(&mrt_lock);
+}
+
+static int ipmr_mfc_seq_show(struct seq_file *seq, void *v)
+{
+	int n;
+
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq,
+			 "Group                            "
+			 "Origin                           "
+			 "Iif      Pkts  Bytes     Wrong  Oifs\n");
+	} else {
+		const struct mfc6_cache *mfc = v;
+		const struct ipmr_mfc_iter *it = seq->private;
+
+		seq_printf(seq,
+			   NIP6_FMT " " NIP6_FMT " %-3d %8ld %8ld %8ld",
+			   NIP6(mfc->mf6c_mcastgrp), NIP6(mfc->mf6c_origin),
+			   mfc->mf6c_parent,
+			   mfc->mfc_un.res.pkt,
+			   mfc->mfc_un.res.bytes,
+			   mfc->mfc_un.res.wrong_if);
+
+		if (it->cache != &mfc_unres_queue) {
+			for (n = mfc->mfc_un.res.minvif;
+			     n < mfc->mfc_un.res.maxvif; n++) {
+				if (MIF_EXISTS(n) &&
+				    mfc->mfc_un.res.ttls[n] < 255)
+					seq_printf(seq,
+						   " %2d:%-3d",
+						   n, mfc->mfc_un.res.ttls[n]);
+			}
+		}
+		seq_putc(seq, '\n');
+	}
+	return 0;
+}
+
+static struct seq_operations ipmr_mfc_seq_ops = {
+	.start = ipmr_mfc_seq_start,
+	.next  = ipmr_mfc_seq_next,
+	.stop  = ipmr_mfc_seq_stop,
+	.show  = ipmr_mfc_seq_show,
+};
+
+static int ipmr_mfc_open(struct inode *inode, struct file *file)
+{
+	return seq_open_private(file, &ipmr_mfc_seq_ops,
+				sizeof(struct ipmr_mfc_iter));
+}
+
+static struct file_operations ip6mr_mfc_fops = {
+	.owner	 = THIS_MODULE,
+	.open    = ipmr_mfc_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release,
+};
+#endif
+
+/*
+ *	Delete a VIF entry
+ */
+
+static int mif6_delete(int vifi)
+{
+	struct mif_device *v;
+	struct net_device *dev;
+	if (vifi < 0 || vifi >= maxvif)
+		return -EADDRNOTAVAIL;
+
+	v = &vif6_table[vifi];
+
+	write_lock_bh(&mrt_lock);
+	dev = v->dev;
+	v->dev = NULL;
+
+	if (!dev) {
+		write_unlock_bh(&mrt_lock);
+		return -EADDRNOTAVAIL;
+	}
+
+	if (vifi + 1 == maxvif) {
+		int tmp;
+		for (tmp = vifi - 1; tmp >= 0; tmp--) {
+			if (MIF_EXISTS(tmp))
+				break;
+		}
+		maxvif = tmp + 1;
+	}
+
+	write_unlock_bh(&mrt_lock);
+
+	dev_set_allmulti(dev, -1);
+
+	if (v->flags & MIFF_REGISTER)
+		unregister_netdevice(dev);
+
+	dev_put(dev);
+	return 0;
+}
+
+/* Destroy an unresolved cache entry, killing queued skbs
+   and reporting error to netlink readers.
+ */
+
+static void ip6mr_destroy_unres(struct mfc6_cache *c)
+{
+	struct sk_buff *skb;
+
+	atomic_dec(&cache_resolve_queue_len);
+
+	while((skb = skb_dequeue(&c->mfc_un.unres.unresolved)) != NULL) {
+		if (ipv6_hdr(skb)->version == 0) {
+			struct nlmsghdr *nlh = (struct nlmsghdr *)skb_pull(skb, sizeof(struct ipv6hdr));
+			nlh->nlmsg_type = NLMSG_ERROR;
+			nlh->nlmsg_len = NLMSG_LENGTH(sizeof(struct nlmsgerr));
+			skb_trim(skb, nlh->nlmsg_len);
+			((struct nlmsgerr *)NLMSG_DATA(nlh))->error = -ETIMEDOUT;
+			rtnl_unicast(skb, &init_net, NETLINK_CB(skb).pid);
+		} else
+			kfree_skb(skb);
+	}
+
+	kmem_cache_free(mrt_cachep, c);
+}
+
+
+/* Single timer process for all the unresolved queue. */
+
+static void ipmr_do_expire_process(unsigned long dummy)
+{
+	unsigned long now = jiffies;
+	unsigned long expires = 10 * HZ;
+	struct mfc6_cache *c, **cp;
+
+	cp = &mfc_unres_queue;
+
+	while ((c = *cp) != NULL) {
+		if (time_after(c->mfc_un.unres.expires, now)) {
+			/* not yet... */
+			unsigned long interval = c->mfc_un.unres.expires - now;
+			if (interval < expires)
+				expires = interval;
+			cp = &c->next;
+			continue;
+		}
+
+		*cp = c->next;
+		ip6mr_destroy_unres(c);
+	}
+
+	if (atomic_read(&cache_resolve_queue_len))
+		mod_timer(&ipmr_expire_timer, jiffies + expires);
+}
+
+static void ipmr_expire_process(unsigned long dummy)
+{
+	if (!spin_trylock(&mfc_unres_lock)) {
+		mod_timer(&ipmr_expire_timer, jiffies + 1);
+		return;
+	}
+
+	if (atomic_read(&cache_resolve_queue_len))
+		ipmr_do_expire_process(dummy);
+
+	spin_unlock(&mfc_unres_lock);
+}
+
+/* Fill oifs list. It is called under write locked mrt_lock. */
+
+static void ip6mr_update_thresholds(struct mfc6_cache *cache, unsigned char *ttls)
+{
+	int vifi;
+
+	cache->mfc_un.res.minvif = MAXVIFS;
+	cache->mfc_un.res.maxvif = 0;
+	memset(cache->mfc_un.res.ttls, 255, MAXVIFS);
+
+	for (vifi = 0; vifi < maxvif; vifi++) {
+		if (MIF_EXISTS(vifi) && ttls[vifi] && ttls[vifi] < 255) {
+			cache->mfc_un.res.ttls[vifi] = ttls[vifi];
+			if (cache->mfc_un.res.minvif > vifi)
+				cache->mfc_un.res.minvif = vifi;
+			if (cache->mfc_un.res.maxvif <= vifi)
+				cache->mfc_un.res.maxvif = vifi + 1;
+		}
+	}
+}
+
+static int mif6_add(struct mif6ctl *vifc, int mrtsock)
+{
+	int vifi = vifc->mif6c_mifi;
+	struct mif_device *v = &vif6_table[vifi];
+	struct net_device *dev;
+
+	/* Is vif busy ? */
+	if (MIF_EXISTS(vifi))
+		return -EADDRINUSE;
+
+	switch (vifc->mif6c_flags) {
+	case 0:
+		dev = dev_get_by_index(&init_net, vifc->mif6c_pifi);
+		if (!dev)
+			return -EADDRNOTAVAIL;
+		dev_put(dev);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	dev_set_allmulti(dev, 1);
+
+	/*
+	 *	Fill in the VIF structures
+	 */
+	v->rate_limit = vifc->vifc_rate_limit;
+	v->flags = vifc->mif6c_flags;
+	if (!mrtsock)
+		v->flags |= VIFF_STATIC;
+	v->threshold = vifc->vifc_threshold;
+	v->bytes_in = 0;
+	v->bytes_out = 0;
+	v->pkt_in = 0;
+	v->pkt_out = 0;
+	v->link = dev->ifindex;
+	if (v->flags & MIFF_REGISTER)
+		v->link = dev->iflink;
+
+	/* And finish update writing critical data */
+	write_lock_bh(&mrt_lock);
+	dev_hold(dev);
+	v->dev = dev;
+	if (vifi + 1 > maxvif)
+		maxvif = vifi + 1;
+	write_unlock_bh(&mrt_lock);
+	return 0;
+}
+
+static struct mfc6_cache *ip6mr_cache_find(struct in6_addr *origin, struct in6_addr *mcastgrp)
+{
+	int line = MFC6_HASH(mcastgrp, origin);
+	struct mfc6_cache *c;
+
+	for (c = mfc6_cache_array[line]; c; c = c->next) {
+		if (ipv6_addr_equal(&c->mf6c_origin, origin) &&
+		    ipv6_addr_equal(&c->mf6c_mcastgrp, mcastgrp))
+			break;
+	}
+	return c;
+}
+
+/*
+ *	Allocate a multicast cache entry
+ */
+static struct mfc6_cache *ip6mr_cache_alloc(void)
+{
+	struct mfc6_cache *c = kmem_cache_alloc(mrt_cachep, GFP_KERNEL);
+	if (c == NULL)
+		return NULL;
+	memset(c, 0, sizeof(*c));
+	c->mfc_un.res.minvif = MAXVIFS;
+	return c;
+}
+
+static struct mfc6_cache *ip6mr_cache_alloc_unres(void)
+{
+	struct mfc6_cache *c = kmem_cache_alloc(mrt_cachep, GFP_ATOMIC);
+	if (c == NULL)
+		return NULL;
+	memset(c, 0, sizeof(*c));
+	skb_queue_head_init(&c->mfc_un.unres.unresolved);
+	c->mfc_un.unres.expires = jiffies + 10 * HZ;
+	return c;
+}
+
+/*
+ *	A cache entry has gone into a resolved state from queued
+ */
+
+static void ip6mr_cache_resolve(struct mfc6_cache *uc, struct mfc6_cache *c)
+{
+	struct sk_buff *skb;
+
+	/*
+	 *	Play the pending entries through our router
+	 */
+
+	while((skb = __skb_dequeue(&uc->mfc_un.unres.unresolved))) {
+		if (ipv6_hdr(skb)->version == 0) {
+			int err;
+			struct nlmsghdr *nlh = (struct nlmsghdr *)skb_pull(skb, sizeof(struct ipv6hdr));
+
+			if (ip6mr_fill_mroute(skb, c, NLMSG_DATA(nlh)) > 0) {
+				nlh->nlmsg_len = skb->tail - (u8 *)nlh;
+			} else {
+				nlh->nlmsg_type = NLMSG_ERROR;
+				nlh->nlmsg_len = NLMSG_LENGTH(sizeof(struct nlmsgerr));
+				skb_trim(skb, nlh->nlmsg_len);
+				((struct nlmsgerr *)NLMSG_DATA(nlh))->error = -EMSGSIZE;
+			}
+			err = rtnl_unicast(skb, &init_net, NETLINK_CB(skb).pid);
+		} else
+			ip6_mr_forward(skb, c);
+	}
+}
+
+/*
+ *	Bounce a cache query up to pim6sd. We could use netlink for this but pim6sd
+ *	expects the following bizarre scheme.
+ *
+ *	Called under mrt_lock.
+ */
+
+static int ip6mr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert)
+{
+	struct sk_buff *skb;
+	struct mrt6msg *msg;
+	int ret;
+
+	skb = alloc_skb(sizeof(struct ipv6hdr) + sizeof(*msg), GFP_ATOMIC);
+
+	if (!skb)
+		return -ENOBUFS;
+
+	/* I suppose that internal messages
+	 * do not require checksums */
+
+	skb->ip_summed = CHECKSUM_UNNECESSARY;
+
+	/*
+	 *	Copy the IP header
+	 */
+
+	skb_put(skb, sizeof(struct ipv6hdr));
+	skb_reset_network_header(skb);
+	skb_copy_to_linear_data(skb, ipv6_hdr(pkt), sizeof(struct ipv6hdr));
+
+	/*
+	 *	Add our header
+	 */
+	skb_put(skb, sizeof(*msg));
+	skb_reset_transport_header(skb);
+	msg = (struct mrt6msg *)skb_transport_header(skb);
+
+	msg->im6_mbz = 0;
+	msg->im6_msgtype = assert;
+	msg->im6_mif = vifi;
+	msg->im6_pad = 0;
+	ipv6_addr_copy(&msg->im6_src, &ipv6_hdr(pkt)->saddr);
+	ipv6_addr_copy(&msg->im6_dst, &ipv6_hdr(pkt)->daddr);
+
+	skb->dst = dst_clone(pkt->dst);
+	skb->ip_summed = CHECKSUM_UNNECESSARY;
+
+	skb_pull(skb, sizeof(struct ipv6hdr));
+
+	if (mroute6_socket == NULL) {
+		kfree_skb(skb);
+		return -EINVAL;
+	}
+
+	/*
+	 *	Deliver to user space multicast routing algorithms
+	 */
+	if ((ret = sock_queue_rcv_skb(mroute6_socket, skb)) < 0) {
+		if (net_ratelimit())
+			printk(KERN_WARNING "mroute6: pending queue full, dropping entries.\n");
+		kfree_skb(skb);
+	}
+
+	return ret;
+}
+
+/*
+ *	Queue a packet for resolution. It gets locked cache entry!
+ */
+
+static int
+ip6mr_cache_unresolved(vifi_t vifi, struct sk_buff *skb)
+{
+	int err;
+	struct mfc6_cache *c;
+
+	spin_lock_bh(&mfc_unres_lock);
+	for (c = mfc_unres_queue; c; c = c->next) {
+		if (ipv6_addr_equal(&c->mf6c_mcastgrp, &ipv6_hdr(skb)->daddr) &&
+		    ipv6_addr_equal(&c->mf6c_origin, &ipv6_hdr(skb)->saddr))
+			break;
+	}
+
+	if (c == NULL) {
+		/*
+		 *	Create a new entry if allowable
+		 */
+
+		if (atomic_read(&cache_resolve_queue_len) >= 10 ||
+		    (c = ip6mr_cache_alloc_unres()) == NULL) {
+			spin_unlock_bh(&mfc_unres_lock);
+
+			kfree_skb(skb);
+			return -ENOBUFS;
+		}
+
+		/*
+		 *	Fill in the new cache entry
+		 */
+		c->mf6c_parent = -1;
+		c->mf6c_origin = ipv6_hdr(skb)->saddr;
+		c->mf6c_mcastgrp = ipv6_hdr(skb)->daddr;
+
+		/*
+		 *	Reflect first query at pim6sd
+		 */
+		if ((err = ip6mr_cache_report(skb, vifi, MRT6MSG_NOCACHE)) < 0) {
+			/* If the report failed throw the cache entry
+			   out - Brad Parker
+			 */
+			spin_unlock_bh(&mfc_unres_lock);
+
+			kmem_cache_free(mrt_cachep, c);
+			kfree_skb(skb);
+			return err;
+		}
+
+		atomic_inc(&cache_resolve_queue_len);
+		c->next = mfc_unres_queue;
+		mfc_unres_queue = c;
+
+		ipmr_do_expire_process(1);
+	}
+
+	/*
+	 *	See if we can append the packet
+	 */
+	if (c->mfc_un.unres.unresolved.qlen > 3) {
+		kfree_skb(skb);
+		err = -ENOBUFS;
+	} else {
+		skb_queue_tail(&c->mfc_un.unres.unresolved, skb);
+		err = 0;
+	}
+
+	spin_unlock_bh(&mfc_unres_lock);
+	return err;
+}
+
+/*
+ *	MFC6 cache manipulation by user space
+ */
+
+static int ip6mr_mfc_delete(struct mf6cctl *mfc)
+{
+	int line;
+	struct mfc6_cache *c, **cp;
+
+	line = MFC6_HASH(&mfc->mf6cc_mcastgrp.sin6_addr, &mfc->mf6cc_origin.sin6_addr);
+
+	for (cp = &mfc6_cache_array[line]; (c = *cp) != NULL; cp = &c->next) {
+		if (ipv6_addr_equal(&c->mf6c_origin, &mfc->mf6cc_origin.sin6_addr) &&
+		    ipv6_addr_equal(&c->mf6c_mcastgrp, &mfc->mf6cc_mcastgrp.sin6_addr)) {
+			write_lock_bh(&mrt_lock);
+			*cp = c->next;
+			write_unlock_bh(&mrt_lock);
+
+			kmem_cache_free(mrt_cachep, c);
+			return 0;
+		}
+	}
+	return -ENOENT;
+}
+
+static int ip6mr_device_event(struct notifier_block *this,
+			      unsigned long event, void *ptr)
+{
+	struct net_device *dev = ptr;
+	struct mif_device *v;
+	int ct;
+
+	if (dev_net(dev) != &init_net)
+		return NOTIFY_DONE;
+
+	if (event != NETDEV_UNREGISTER)
+		return NOTIFY_DONE;
+
+	v = &vif6_table[0];
+	for (ct = 0; ct < maxvif; ct++, v++) {
+		if (v->dev == dev)
+			mif6_delete(ct);
+	}
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block ip6_mr_notifier = {
+	.notifier_call = ip6mr_device_event
+};
+
+/*
+ *	Setup for IP multicast routing
+ */
+
+void __init ip6_mr_init(void)
+{
+	mrt_cachep = kmem_cache_create("ip6_mrt_cache",
+				       sizeof(struct mfc6_cache),
+				       0, SLAB_HWCACHE_ALIGN,
+				       NULL);
+	if (!mrt_cachep)
+		panic("cannot allocate ip6_mrt_cache");
+
+	setup_timer(&ipmr_expire_timer, ipmr_expire_process, 0);
+	register_netdevice_notifier(&ip6_mr_notifier);
+#ifdef CONFIG_PROC_FS
+	proc_net_fops_create(&init_net, "ip6_mr_vif", 0, &ip6mr_vif_fops);
+	proc_net_fops_create(&init_net, "ip6_mr_cache", 0, &ip6mr_mfc_fops);
+#endif
+}
+
+
+static int ip6mr_mfc_add(struct mf6cctl *mfc, int mrtsock)
+{
+	int line;
+	struct mfc6_cache *uc, *c, **cp;
+	unsigned char ttls[MAXVIFS];
+	int i;
+
+	memset(ttls, 255, MAXVIFS);
+	for (i = 0; i < MAXVIFS; i++) {
+		if (IF_ISSET(i, &mfc->mf6cc_ifset))
+			ttls[i] = 1;
+
+	}
+
+	line = MFC6_HASH(&mfc->mf6cc_mcastgrp.sin6_addr, &mfc->mf6cc_origin.sin6_addr);
+
+	for (cp = &mfc6_cache_array[line]; (c = *cp) != NULL; cp = &c->next) {
+		if (ipv6_addr_equal(&c->mf6c_origin, &mfc->mf6cc_origin.sin6_addr) &&
+		    ipv6_addr_equal(&c->mf6c_mcastgrp, &mfc->mf6cc_mcastgrp.sin6_addr))
+			break;
+	}
+
+	if (c != NULL) {
+		write_lock_bh(&mrt_lock);
+		c->mf6c_parent = mfc->mf6cc_parent;
+		ip6mr_update_thresholds(c, ttls);
+		if (!mrtsock)
+			c->mfc_flags |= MFC_STATIC;
+		write_unlock_bh(&mrt_lock);
+		return 0;
+	}
+
+	if (!ipv6_addr_is_multicast(&mfc->mf6cc_mcastgrp.sin6_addr))
+		return -EINVAL;
+
+	c = ip6mr_cache_alloc();
+	if (c == NULL)
+		return -ENOMEM;
+
+	c->mf6c_origin = mfc->mf6cc_origin.sin6_addr;
+	c->mf6c_mcastgrp = mfc->mf6cc_mcastgrp.sin6_addr;
+	c->mf6c_parent = mfc->mf6cc_parent;
+	ip6mr_update_thresholds(c, ttls);
+	if (!mrtsock)
+		c->mfc_flags |= MFC_STATIC;
+
+	write_lock_bh(&mrt_lock);
+	c->next = mfc6_cache_array[line];
+	mfc6_cache_array[line] = c;
+	write_unlock_bh(&mrt_lock);
+
+	/*
+	 *	Check to see if we resolved a queued list. If so we
+	 *	need to send on the frames and tidy up.
+	 */
+	spin_lock_bh(&mfc_unres_lock);
+	for (cp = &mfc_unres_queue; (uc = *cp) != NULL;
+	     cp = &uc->next) {
+		if (ipv6_addr_equal(&uc->mf6c_origin, &c->mf6c_origin) &&
+		    ipv6_addr_equal(&uc->mf6c_mcastgrp, &c->mf6c_mcastgrp)) {
+			*cp = uc->next;
+			if (atomic_dec_and_test(&cache_resolve_queue_len))
+				del_timer(&ipmr_expire_timer);
+			break;
+		}
+	}
+	spin_unlock_bh(&mfc_unres_lock);
+
+	if (uc) {
+		ip6mr_cache_resolve(uc, c);
+		kmem_cache_free(mrt_cachep, uc);
+	}
+	return 0;
+}
+
+/*
+ *	Close the multicast socket, and clear the vif tables etc
+ */
+
+static void mroute_clean_tables(struct sock *sk)
+{
+	int i;
+
+	/*
+	 *	Shut down all active vif entries
+	 */
+	for (i = 0; i < maxvif; i++) {
+		if (!(vif6_table[i].flags & VIFF_STATIC))
+			mif6_delete(i);
+	}
+
+	/*
+	 *	Wipe the cache
+	 */
+	for (i = 0; i < ARRAY_SIZE(mfc6_cache_array); i++) {
+		struct mfc6_cache *c, **cp;
+
+		cp = &mfc6_cache_array[i];
+		while ((c = *cp) != NULL) {
+			if (c->mfc_flags & MFC_STATIC) {
+				cp = &c->next;
+				continue;
+			}
+			write_lock_bh(&mrt_lock);
+			*cp = c->next;
+			write_unlock_bh(&mrt_lock);
+
+			kmem_cache_free(mrt_cachep, c);
+		}
+	}
+
+	if (atomic_read(&cache_resolve_queue_len) != 0) {
+		struct mfc6_cache *c;
+
+		spin_lock_bh(&mfc_unres_lock);
+		while (mfc_unres_queue != NULL) {
+			c = mfc_unres_queue;
+			mfc_unres_queue = c->next;
+			spin_unlock_bh(&mfc_unres_lock);
+
+			ip6mr_destroy_unres(c);
+
+			spin_lock_bh(&mfc_unres_lock);
+		}
+		spin_unlock_bh(&mfc_unres_lock);
+	}
+}
+
+static int ip6mr_sk_init(struct sock *sk)
+{
+	int err = 0;
+
+	rtnl_lock();
+	write_lock_bh(&mrt_lock);
+	if (likely(mroute6_socket == NULL))
+		mroute6_socket = sk;
+	else
+		err = -EADDRINUSE;
+	write_unlock_bh(&mrt_lock);
+
+	rtnl_unlock();
+
+	return err;
+}
+
+int ip6mr_sk_done(struct sock *sk)
+{
+	int err = 0;
+
+	rtnl_lock();
+	if (sk == mroute6_socket) {
+		write_lock_bh(&mrt_lock);
+		mroute6_socket = NULL;
+		write_unlock_bh(&mrt_lock);
+
+		mroute_clean_tables(sk);
+	} else
+		err = -EACCES;
+	rtnl_unlock();
+
+	return err;
+}
+
+/*
+ *	Socket options and virtual interface manipulation. The whole
+ *	virtual interface system is a complete heap, but unfortunately
+ *	that's how BSD mrouted happens to think. Maybe one day with a proper
+ *	MOSPF/PIM router set up we can clean this up.
+ */
+
+int ip6_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, int optlen)
+{
+	int ret;
+	struct mif6ctl vif;
+	struct mf6cctl mfc;
+	mifi_t mifi;
+
+	if (optname != MRT6_INIT) {
+		if (sk != mroute6_socket && !capable(CAP_NET_ADMIN))
+			return -EACCES;
+	}
+
+	switch (optname) {
+	case MRT6_INIT:
+		if (sk->sk_type != SOCK_RAW ||
+		    inet_sk(sk)->num != IPPROTO_ICMPV6)
+			return -EOPNOTSUPP;
+		if (optlen < sizeof(int))
+			return -EINVAL;
+
+		return ip6mr_sk_init(sk);
+
+	case MRT6_DONE:
+		return ip6mr_sk_done(sk);
+
+	case MRT6_ADD_MIF:
+		if (optlen < sizeof(vif))
+			return -EINVAL;
+		if (copy_from_user(&vif, optval, sizeof(vif)))
+			return -EFAULT;
+		if (vif.mif6c_mifi >= MAXVIFS)
+			return -ENFILE;
+		rtnl_lock();
+		ret = mif6_add(&vif, sk == mroute6_socket);
+		rtnl_unlock();
+		return ret;
+
+	case MRT6_DEL_MIF:
+		if (optlen < sizeof(mifi_t))
+			return -EINVAL;
+		if (copy_from_user(&mifi, optval, sizeof(mifi_t)))
+			return -EFAULT;
+		rtnl_lock();
+		ret = mif6_delete(mifi);
+		rtnl_unlock();
+		return ret;
+
+	/*
+	 *	Manipulate the forwarding caches. These live
+	 *	in a sort of kernel/user symbiosis.
+	 */
+	case MRT6_ADD_MFC:
+	case MRT6_DEL_MFC:
+		if (optlen < sizeof(mfc))
+			return -EINVAL;
+		if (copy_from_user(&mfc, optval, sizeof(mfc)))
+			return -EFAULT;
+		rtnl_lock();
+		if (optname == MRT6_DEL_MFC)
+			ret = ip6mr_mfc_delete(&mfc);
+		else
+			ret = ip6mr_mfc_add(&mfc, sk == mroute6_socket);
+		rtnl_unlock();
+		return ret;
+
+	/*
+	 *	Spurious command, or MRT_VERSION which you cannot
+	 *	set.
+	 */
+	default:
+		return -ENOPROTOOPT;
+	}
+}
+
+/*
+ *	Getsock opt support for the multicast routing system.
+ */
+
+int ip6_mroute_getsockopt(struct sock *sk, int optname, char __user *optval,
+			  int __user *optlen)
+{
+	int olr;
+	int val;
+
+	switch (optname) {
+	case MRT6_VERSION:
+		val = 0x0305;
+		break;
+	default:
+		return -ENOPROTOOPT;
+	}
+
+	if (get_user(olr, optlen))
+		return -EFAULT;
+
+	olr = min_t(int, olr, sizeof(int));
+	if (olr < 0)
+		return -EINVAL;
+
+	if (put_user(olr, optlen))
+		return -EFAULT;
+	if (copy_to_user(optval, &val, olr))
+		return -EFAULT;
+	return 0;
+}
+
+/*
+ *	The IP multicast ioctl support routines.
+ */
+
+int ip6mr_ioctl(struct sock *sk, int cmd, void __user *arg)
+{
+	struct sioc_sg_req6 sr;
+	struct sioc_mif_req6 vr;
+	struct mif_device *vif;
+	struct mfc6_cache *c;
+
+	switch (cmd) {
+	case SIOCGETMIFCNT_IN6:
+		if (copy_from_user(&vr, arg, sizeof(vr)))
+			return -EFAULT;
+		if (vr.mifi >= maxvif)
+			return -EINVAL;
+		read_lock(&mrt_lock);
+		vif = &vif6_table[vr.mifi];
+		if (MIF_EXISTS(vr.mifi)) {
+			vr.icount = vif->pkt_in;
+			vr.ocount = vif->pkt_out;
+			vr.ibytes = vif->bytes_in;
+			vr.obytes = vif->bytes_out;
+			read_unlock(&mrt_lock);
+
+			if (copy_to_user(arg, &vr, sizeof(vr)))
+				return -EFAULT;
+			return 0;
+		}
+		read_unlock(&mrt_lock);
+		return -EADDRNOTAVAIL;
+	case SIOCGETSGCNT_IN6:
+		if (copy_from_user(&sr, arg, sizeof(sr)))
+			return -EFAULT;
+
+		read_lock(&mrt_lock);
+		c = ip6mr_cache_find(&sr.src.sin6_addr, &sr.grp.sin6_addr);
+		if (c) {
+			sr.pktcnt = c->mfc_un.res.pkt;
+			sr.bytecnt = c->mfc_un.res.bytes;
+			sr.wrong_if = c->mfc_un.res.wrong_if;
+			read_unlock(&mrt_lock);
+
+			if (copy_to_user(arg, &sr, sizeof(sr)))
+				return -EFAULT;
+			return 0;
+		}
+		read_unlock(&mrt_lock);
+		return -EADDRNOTAVAIL;
+	default:
+		return -ENOIOCTLCMD;
+	}
+}
+
+
+static inline int ip6mr_forward2_finish(struct sk_buff *skb)
+{
+	/* XXX stats */
+	return dst_output(skb);
+}
+
+/*
+ *	Processing handlers for ip6mr_forward
+ */
+
+static int ip6mr_forward2(struct sk_buff *skb, struct mfc6_cache *c, int vifi)
+{
+	struct ipv6hdr *ipv6h;
+	struct mif_device *vif = &vif6_table[vifi];
+	struct net_device *dev;
+	struct dst_entry *dst;
+	struct flowi fl;
+
+	if (vif->dev == NULL)
+		goto out_free;
+
+	ipv6h = ipv6_hdr(skb);
+
+	fl = (struct flowi) {
+		.oif = vif->link,
+		.nl_u = { .ip6_u =
+				{ .daddr = ipv6h->daddr, }
+		}
+	};
+
+	dst = ip6_route_output(&init_net, NULL, &fl);
+	if (!dst)
+		goto out_free;
+
+	dst_release(skb->dst);
+	skb->dst = dst;
+
+	/*
+	 * RFC1584 teaches, that DVMRP/PIM router must deliver packets locally
+	 * not only before forwarding, but after forwarding on all output
+	 * interfaces. It is clear, if mrouter runs a multicasting
+	 * program, it should receive packets not depending to what interface
+	 * program is joined.
+	 * If we will not make it, the program will have to join on all
+	 * interfaces. On the other hand, multihoming host (or router, but
+	 * not mrouter) cannot join to more than one interface - it will
+	 * result in receiving multiple packets.
+	 */
+	dev = vif->dev;
+	skb->dev = dev;
+	vif->pkt_out++;
+	vif->bytes_out += skb->len;
+
+	/* We are about to write */
+	/* XXX: extension headers? */
+	if (skb_cow(skb, sizeof(*ipv6h) + LL_RESERVED_SPACE(dev)))
+		goto out_free;
+
+	ipv6h = ipv6_hdr(skb);
+	ipv6h->hop_limit--;
+
+	IP6CB(skb)->flags |= IP6SKB_FORWARDED;
+
+	return NF_HOOK(PF_INET6, NF_INET_FORWARD, skb, skb->dev, dev,
+		       ip6mr_forward2_finish);
+
+out_free:
+	kfree_skb(skb);
+	return 0;
+}
+
+static int ip6mr_find_vif(struct net_device *dev)
+{
+	int ct;
+	for (ct = maxvif - 1; ct >= 0; ct--) {
+		if (vif6_table[ct].dev == dev)
+			break;
+	}
+	return ct;
+}
+
+static int ip6_mr_forward(struct sk_buff *skb, struct mfc6_cache *cache)
+{
+	int psend = -1;
+	int vif, ct;
+
+	vif = cache->mf6c_parent;
+	cache->mfc_un.res.pkt++;
+	cache->mfc_un.res.bytes += skb->len;
+
+	vif6_table[vif].pkt_in++;
+	vif6_table[vif].bytes_in += skb->len;
+
+	/*
+	 *	Forward the frame
+	 */
+	for (ct = cache->mfc_un.res.maxvif - 1; ct >= cache->mfc_un.res.minvif; ct--) {
+		if (ipv6_hdr(skb)->hop_limit > cache->mfc_un.res.ttls[ct]) {
+			if (psend != -1) {
+				struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
+				if (skb2)
+					ip6mr_forward2(skb2, cache, psend);
+			}
+			psend = ct;
+		}
+	}
+	if (psend != -1) {
+		ip6mr_forward2(skb, cache, psend);
+		return 0;
+	}
+
+	kfree_skb(skb);
+	return 0;
+}
+
+
+/*
+ *	Multicast packets for forwarding arrive here
+ */
+
+int ip6_mr_input(struct sk_buff *skb)
+{
+	struct mfc6_cache *cache;
+
+	read_lock(&mrt_lock);
+	cache = ip6mr_cache_find(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr);
+
+	/*
+	 *	No usable cache entry
+	 */
+	if (cache == NULL) {
+		int vif;
+
+		vif = ip6mr_find_vif(skb->dev);
+		if (vif >= 0) {
+			int err = ip6mr_cache_unresolved(vif, skb);
+			read_unlock(&mrt_lock);
+
+			return err;
+		}
+		read_unlock(&mrt_lock);
+		kfree_skb(skb);
+		return -ENODEV;
+	}
+
+	ip6_mr_forward(skb, cache);
+
+	read_unlock(&mrt_lock);
+
+	return 0;
+}
+
+
+static int
+ip6mr_fill_mroute(struct sk_buff *skb, struct mfc6_cache *c, struct rtmsg *rtm)
+{
+	int ct;
+	struct rtnexthop *nhp;
+	struct net_device *dev = vif6_table[c->mf6c_parent].dev;
+	u8 *b = skb->tail;
+	struct rtattr *mp_head;
+
+	if (dev)
+		RTA_PUT(skb, RTA_IIF, 4, &dev->ifindex);
+
+	mp_head = (struct rtattr *)skb_put(skb, RTA_LENGTH(0));
+
+	for (ct = c->mfc_un.res.minvif; ct < c->mfc_un.res.maxvif; ct++) {
+		if (c->mfc_un.res.ttls[ct] < 255) {
+			if (skb_tailroom(skb) < RTA_ALIGN(RTA_ALIGN(sizeof(*nhp)) + 4))
+				goto rtattr_failure;
+			nhp = (struct rtnexthop *)skb_put(skb, RTA_ALIGN(sizeof(*nhp)));
+			nhp->rtnh_flags = 0;
+			nhp->rtnh_hops = c->mfc_un.res.ttls[ct];
+			nhp->rtnh_ifindex = vif6_table[ct].dev->ifindex;
+			nhp->rtnh_len = sizeof(*nhp);
+		}
+	}
+	mp_head->rta_type = RTA_MULTIPATH;
+	mp_head->rta_len = skb->tail - (u8 *)mp_head;
+	rtm->rtm_type = RTN_MULTICAST;
+	return 1;
+
+rtattr_failure:
+	nlmsg_trim(skb, b);
+	return -EMSGSIZE;
+}
+
+int ip6mr_get_route(struct sk_buff *skb, struct rtmsg *rtm, int nowait)
+{
+	int err;
+	struct mfc6_cache *cache;
+	struct rt6_info *rt = (struct rt6_info *)skb->dst;
+
+	read_lock(&mrt_lock);
+	cache = ip6mr_cache_find(&rt->rt6i_src.addr, &rt->rt6i_dst.addr);
+
+	if (!cache) {
+		struct sk_buff *skb2;
+		struct ipv6hdr *iph;
+		struct net_device *dev;
+		int vif;
+
+		if (nowait) {
+			read_unlock(&mrt_lock);
+			return -EAGAIN;
+		}
+
+		dev = skb->dev;
+		if (dev == NULL || (vif = ip6mr_find_vif(dev)) < 0) {
+			read_unlock(&mrt_lock);
+			return -ENODEV;
+		}
+
+		/* really correct? */
+		skb2 = alloc_skb(sizeof(struct ipv6hdr), GFP_ATOMIC);
+		if (!skb2) {
+			read_unlock(&mrt_lock);
+			return -ENOMEM;
+		}
+
+		skb_reset_transport_header(skb2);
+
+		skb_put(skb2, sizeof(struct ipv6hdr));
+		skb_reset_network_header(skb2);
+
+		iph = ipv6_hdr(skb2);
+		iph->version = 0;
+		iph->priority = 0;
+		iph->flow_lbl[0] = 0;
+		iph->flow_lbl[1] = 0;
+		iph->flow_lbl[2] = 0;
+		iph->payload_len = 0;
+		iph->nexthdr = IPPROTO_NONE;
+		iph->hop_limit = 0;
+		ipv6_addr_copy(&iph->saddr, &rt->rt6i_src.addr);
+		ipv6_addr_copy(&iph->daddr, &rt->rt6i_dst.addr);
+
+		err = ip6mr_cache_unresolved(vif, skb2);
+		read_unlock(&mrt_lock);
+
+		return err;
+	}
+
+	if (!nowait && (rtm->rtm_flags&RTM_F_NOTIFY))
+		cache->mfc_flags |= MFC_NOTIFY;
+
+	err = ip6mr_fill_mroute(skb, cache, rtm);
+	read_unlock(&mrt_lock);
+	return err;
+}
+
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 4195ac92345e..99624109c010 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -33,6 +33,7 @@
 #include <linux/sockios.h>
 #include <linux/net.h>
 #include <linux/in6.h>
+#include <linux/mroute6.h>
 #include <linux/netdevice.h>
 #include <linux/if_arp.h>
 #include <linux/init.h>
@@ -118,6 +119,9 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
 
 	valbool = (val!=0);
 
+	if (ip6_mroute_opt(optname))
+		return ip6_mroute_setsockopt(sk, optname, optval, optlen);
+
 	lock_sock(sk);
 
 	switch (optname) {
@@ -790,6 +794,9 @@ static int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
 	int len;
 	int val;
 
+	if (ip6_mroute_opt(optname))
+		return ip6_mroute_getsockopt(sk, optname, optval, optlen);
+
 	if (get_user(len, optlen))
 		return -EFAULT;
 	switch (optname) {
diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c
index aae6cedf1709..088b80b4ce74 100644
--- a/net/ipv6/raw.c
+++ b/net/ipv6/raw.c
@@ -53,6 +53,7 @@
 #if defined(CONFIG_IPV6_MIP6) || defined(CONFIG_IPV6_MIP6_MODULE)
 #include <net/mip6.h>
 #endif
+#include <linux/mroute6.h>
 
 #include <net/raw.h>
 #include <net/rawv6.h>
@@ -1135,7 +1136,11 @@ static int rawv6_ioctl(struct sock *sk, int cmd, unsigned long arg)
 		}
 
 		default:
+#ifdef CONFIG_IPV6_MROUTE
+			return ip6mr_ioctl(sk, cmd, (void __user *)arg);
+#else
 			return -ENOIOCTLCMD;
+#endif
 	}
 }
 
@@ -1143,7 +1148,7 @@ static void rawv6_close(struct sock *sk, long timeout)
 {
 	if (inet_sk(sk)->num == IPPROTO_RAW)
 		ip6_ra_control(sk, -1, NULL);
-
+	ip6mr_sk_done(sk);
 	sk_common_release(sk);
 }
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index cd82b6db35ff..3c314d5f46c6 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -36,6 +36,7 @@
 #include <linux/route.h>
 #include <linux/netdevice.h>
 #include <linux/in6.h>
+#include <linux/mroute6.h>
 #include <linux/init.h>
 #include <linux/if_arp.h>
 #include <linux/proc_fs.h>
@@ -2106,7 +2107,7 @@ static inline size_t rt6_nlmsg_size(void)
 static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 			 struct in6_addr *dst, struct in6_addr *src,
 			 int iif, int type, u32 pid, u32 seq,
-			 int prefix, unsigned int flags)
+			 int prefix, int nowait, unsigned int flags)
 {
 	struct rtmsg *rtm;
 	struct nlmsghdr *nlh;
@@ -2166,9 +2167,24 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 	} else if (rtm->rtm_src_len)
 		NLA_PUT(skb, RTA_SRC, 16, &rt->rt6i_src.addr);
 #endif
-	if (iif)
-		NLA_PUT_U32(skb, RTA_IIF, iif);
-	else if (dst) {
+	if (iif) {
+#ifdef CONFIG_IPV6_MROUTE
+		if (ipv6_addr_is_multicast(&rt->rt6i_dst.addr)) {
+			int err = ip6mr_get_route(skb, rtm, nowait);
+			if (err <= 0) {
+				if (!nowait) {
+					if (err == 0)
+						return 0;
+					goto nla_put_failure;
+				} else {
+					if (err == -EMSGSIZE)
+						goto nla_put_failure;
+				}
+			}
+		} else
+#endif
+			NLA_PUT_U32(skb, RTA_IIF, iif);
+	} else if (dst) {
 		struct in6_addr saddr_buf;
 		if (ipv6_dev_get_saddr(ip6_dst_idev(&rt->u.dst)->dev,
 				       dst, 0, &saddr_buf) == 0)
@@ -2211,7 +2227,7 @@ int rt6_dump_route(struct rt6_info *rt, void *p_arg)
 
 	return rt6_fill_node(arg->skb, rt, NULL, NULL, 0, RTM_NEWROUTE,
 		     NETLINK_CB(arg->cb->skb).pid, arg->cb->nlh->nlmsg_seq,
-		     prefix, NLM_F_MULTI);
+		     prefix, 0, NLM_F_MULTI);
 }
 
 static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
@@ -2277,7 +2293,7 @@ static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void
 
 	err = rt6_fill_node(skb, rt, &fl.fl6_dst, &fl.fl6_src, iif,
 			    RTM_NEWROUTE, NETLINK_CB(in_skb).pid,
-			    nlh->nlmsg_seq, 0, 0);
+			    nlh->nlmsg_seq, 0, 0, 0);
 	if (err < 0) {
 		kfree_skb(skb);
 		goto errout;
@@ -2303,7 +2319,7 @@ void inet6_rt_notify(int event, struct rt6_info *rt, struct nl_info *info)
 		goto errout;
 
 	err = rt6_fill_node(skb, rt, NULL, NULL, 0,
-				event, info->pid, seq, 0, 0);
+				event, info->pid, seq, 0, 0, 0);
 	if (err < 0) {
 		/* -EMSGSIZE implies BUG in rt6_nlmsg_size() */
 		WARN_ON(err == -EMSGSIZE);
-- 
cgit v1.2.3


From 14fb64e1f449ef6666f1c3a3fa4e13aec669b98d Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Thu, 3 Apr 2008 09:22:54 +0900
Subject: [IPV6] MROUTE: Support PIM-SM (SSM).

Based on ancient patch by Mickael Hoerdt
<hoerdt@clarinet.u-strasbg.fr>, which is available at
<http://www-r2.u-strasbg.fr/~hoerdt/dev/linux_ipv6_mforwarding/patch-linux-ipv6-mforwarding-0.1a>.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/mroute6.h |   4 +
 net/ipv6/Kconfig        |   7 ++
 net/ipv6/ip6mr.c        | 275 +++++++++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 285 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/mroute6.h b/include/linux/mroute6.h
index b92190304e0b..f6469fb90840 100644
--- a/include/linux/mroute6.h
+++ b/include/linux/mroute6.h
@@ -23,6 +23,8 @@
 #define MRT6_ADD_MFC	(MRT6_BASE+4)	/* Add a multicast forwarding entry	*/
 #define MRT6_DEL_MFC	(MRT6_BASE+5)	/* Delete a multicast forwarding entry	*/
 #define MRT6_VERSION	(MRT6_BASE+6)	/* Get the kernel multicast version	*/
+#define MRT6_ASSERT	(MRT6_BASE+7)	/* Activate PIM assert mode		*/
+#define MRT6_PIM	(MRT6_BASE+8)	/* enable PIM code	*/
 
 #define SIOCGETMIFCNT_IN6	SIOCPROTOPRIVATE	/* IP protocol privates */
 #define SIOCGETSGCNT_IN6	(SIOCPROTOPRIVATE+1)
@@ -217,6 +219,8 @@ static inline int ip6mr_sk_done(struct sock *sk) { return 0; }
 
 struct mrt6msg {
 #define MRT6MSG_NOCACHE		1
+#define MRT6MSG_WRONGMIF	2
+#define MRT6MSG_WHOLEPKT	3		/* used for use level encap */
 	__u8		im6_mbz;		/* must be zero		   */
 	__u8		im6_msgtype;		/* what type of message    */
 	__u16		im6_mif;		/* mif rec'd on		   */
diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig
index 9a2ea81e499f..82f987b4ef84 100644
--- a/net/ipv6/Kconfig
+++ b/net/ipv6/Kconfig
@@ -216,3 +216,10 @@ config IPV6_MROUTE
 	  Experimental support for IPv6 multicast forwarding.
 	  If unsure, say N.
 
+config IPV6_PIMSM_V2
+	bool "IPv6: PIM-SM version 2 support (EXPERIMENTAL)"
+	depends on IPV6_MROUTE
+	---help---
+	  Support for IPv6 PIM multicast routing protocol PIM-SMv2.
+	  If unsure, say N.
+
diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c
index 1bdf3c177d58..2b70774be61f 100644
--- a/net/ipv6/ip6mr.c
+++ b/net/ipv6/ip6mr.c
@@ -54,6 +54,7 @@
 #include <net/ipv6.h>
 #include <net/ip6_route.h>
 #include <linux/mroute6.h>
+#include <linux/pim.h>
 #include <net/addrconf.h>
 #include <linux/netfilter_ipv6.h>
 
@@ -75,6 +76,13 @@ static int maxvif;
 
 #define MIF_EXISTS(idx) (vif6_table[idx].dev != NULL)
 
+static int mroute_do_assert;				/* Set in PIM assert	*/
+#ifdef CONFIG_IPV6_PIMSM_V2
+static int mroute_do_pim;
+#else
+#define mroute_do_pim 0
+#endif
+
 static struct mfc6_cache *mfc6_cache_array[MFC_LINES];	/* Forwarding cache	*/
 
 static struct mfc6_cache *mfc_unres_queue;		/* Queue of unresolved entries */
@@ -97,6 +105,10 @@ static int ip6_mr_forward(struct sk_buff *skb, struct mfc6_cache *cache);
 static int ip6mr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert);
 static int ip6mr_fill_mroute(struct sk_buff *skb, struct mfc6_cache *c, struct rtmsg *rtm);
 
+#ifdef CONFIG_IPV6_PIMSM_V2
+static struct inet6_protocol pim6_protocol;
+#endif
+
 static struct timer_list ipmr_expire_timer;
 
 
@@ -339,6 +351,132 @@ static struct file_operations ip6mr_mfc_fops = {
 };
 #endif
 
+#ifdef CONFIG_IPV6_PIMSM_V2
+static int reg_vif_num = -1;
+
+static int pim6_rcv(struct sk_buff *skb)
+{
+	struct pimreghdr *pim;
+	struct ipv6hdr   *encap;
+	struct net_device  *reg_dev = NULL;
+
+	if (!pskb_may_pull(skb, sizeof(*pim) + sizeof(*encap)))
+		goto drop;
+
+	pim = (struct pimreghdr *)skb_transport_header(skb);
+	if (pim->type != ((PIM_VERSION << 4) | PIM_REGISTER) ||
+	    (pim->flags & PIM_NULL_REGISTER) ||
+	    (ip_compute_csum((void *)pim, sizeof(*pim)) != 0 &&
+	     (u16)csum_fold(skb_checksum(skb, 0, skb->len, 0))))
+		goto drop;
+
+	/* check if the inner packet is destined to mcast group */
+	encap = (struct ipv6hdr *)(skb_transport_header(skb) +
+				   sizeof(*pim));
+
+	if (!ipv6_addr_is_multicast(&encap->daddr) ||
+	    encap->payload_len == 0 ||
+	    ntohs(encap->payload_len) + sizeof(*pim) > skb->len)
+		goto drop;
+
+	read_lock(&mrt_lock);
+	if (reg_vif_num >= 0)
+		reg_dev = vif6_table[reg_vif_num].dev;
+	if (reg_dev)
+		dev_hold(reg_dev);
+	read_unlock(&mrt_lock);
+
+	if (reg_dev == NULL)
+		goto drop;
+
+	skb->mac_header = skb->network_header;
+	skb_pull(skb, (u8 *)encap - skb->data);
+	skb_reset_network_header(skb);
+	skb->dev = reg_dev;
+	skb->protocol = htons(ETH_P_IP);
+	skb->ip_summed = 0;
+	skb->pkt_type = PACKET_HOST;
+	dst_release(skb->dst);
+	((struct net_device_stats *)netdev_priv(reg_dev))->rx_bytes += skb->len;
+	((struct net_device_stats *)netdev_priv(reg_dev))->rx_packets++;
+	skb->dst = NULL;
+	nf_reset(skb);
+	netif_rx(skb);
+	dev_put(reg_dev);
+	return 0;
+ drop:
+	kfree_skb(skb);
+	return 0;
+}
+
+static struct inet6_protocol pim6_protocol = {
+	.handler	=	pim6_rcv,
+};
+
+/* Service routines creating virtual interfaces: PIMREG */
+
+static int reg_vif_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	read_lock(&mrt_lock);
+	((struct net_device_stats *)netdev_priv(dev))->tx_bytes += skb->len;
+	((struct net_device_stats *)netdev_priv(dev))->tx_packets++;
+	ip6mr_cache_report(skb, reg_vif_num, MRT6MSG_WHOLEPKT);
+	read_unlock(&mrt_lock);
+	kfree_skb(skb);
+	return 0;
+}
+
+static struct net_device_stats *reg_vif_get_stats(struct net_device *dev)
+{
+	return (struct net_device_stats *)netdev_priv(dev);
+}
+
+static void reg_vif_setup(struct net_device *dev)
+{
+	dev->type		= ARPHRD_PIMREG;
+	dev->mtu		= 1500 - sizeof(struct ipv6hdr) - 8;
+	dev->flags		= IFF_NOARP;
+	dev->hard_start_xmit	= reg_vif_xmit;
+	dev->get_stats		= reg_vif_get_stats;
+	dev->destructor		= free_netdev;
+}
+
+static struct net_device *ip6mr_reg_vif(void)
+{
+	struct net_device *dev;
+	struct inet6_dev *in_dev;
+
+	dev = alloc_netdev(sizeof(struct net_device_stats), "pim6reg",
+			   reg_vif_setup);
+
+	if (dev == NULL)
+		return NULL;
+
+	if (register_netdevice(dev)) {
+		free_netdev(dev);
+		return NULL;
+	}
+	dev->iflink = 0;
+
+	in_dev = ipv6_find_idev(dev);
+	if (!in_dev)
+		goto failure;
+
+	if (dev_open(dev))
+		goto failure;
+
+	return dev;
+
+failure:
+	/* allow the register to be completed before unregistering. */
+	rtnl_unlock();
+	rtnl_lock();
+
+	unregister_netdevice(dev);
+	return NULL;
+}
+#endif
+
 /*
  *	Delete a VIF entry
  */
@@ -361,6 +499,11 @@ static int mif6_delete(int vifi)
 		return -EADDRNOTAVAIL;
 	}
 
+#ifdef CONFIG_IPV6_PIMSM_V2
+	if (vifi == reg_vif_num)
+		reg_vif_num = -1;
+#endif
+
 	if (vifi + 1 == maxvif) {
 		int tmp;
 		for (tmp = vifi - 1; tmp >= 0; tmp--) {
@@ -480,6 +623,19 @@ static int mif6_add(struct mif6ctl *vifc, int mrtsock)
 		return -EADDRINUSE;
 
 	switch (vifc->mif6c_flags) {
+#ifdef CONFIG_IPV6_PIMSM_V2
+	case MIFF_REGISTER:
+		/*
+		 * Special Purpose VIF in PIM
+		 * All the packets will be sent to the daemon
+		 */
+		if (reg_vif_num >= 0)
+			return -EADDRINUSE;
+		dev = ip6mr_reg_vif();
+		if (!dev)
+			return -ENOBUFS;
+		break;
+#endif
 	case 0:
 		dev = dev_get_by_index(&init_net, vifc->mif6c_pifi);
 		if (!dev)
@@ -512,6 +668,10 @@ static int mif6_add(struct mif6ctl *vifc, int mrtsock)
 	write_lock_bh(&mrt_lock);
 	dev_hold(dev);
 	v->dev = dev;
+#ifdef CONFIG_IPV6_PIMSM_V2
+	if (v->flags & MIFF_REGISTER)
+		reg_vif_num = vifi;
+#endif
 	if (vifi + 1 > maxvif)
 		maxvif = vifi + 1;
 	write_unlock_bh(&mrt_lock);
@@ -599,7 +759,13 @@ static int ip6mr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert)
 	struct mrt6msg *msg;
 	int ret;
 
-	skb = alloc_skb(sizeof(struct ipv6hdr) + sizeof(*msg), GFP_ATOMIC);
+#ifdef CONFIG_IPV6_PIMSM_V2
+	if (assert == MRT6MSG_WHOLEPKT)
+		skb = skb_realloc_headroom(pkt, -skb_network_offset(pkt)
+						+sizeof(*msg));
+	else
+#endif
+		skb = alloc_skb(sizeof(struct ipv6hdr) + sizeof(*msg), GFP_ATOMIC);
 
 	if (!skb)
 		return -ENOBUFS;
@@ -609,6 +775,29 @@ static int ip6mr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert)
 
 	skb->ip_summed = CHECKSUM_UNNECESSARY;
 
+#ifdef CONFIG_IPV6_PIMSM_V2
+	if (assert == MRT6MSG_WHOLEPKT) {
+		/* Ugly, but we have no choice with this interface.
+		   Duplicate old header, fix length etc.
+		   And all this only to mangle msg->im6_msgtype and
+		   to set msg->im6_mbz to "mbz" :-)
+		 */
+		skb_push(skb, -skb_network_offset(pkt));
+
+		skb_push(skb, sizeof(*msg));
+		skb_reset_transport_header(skb);
+		msg = (struct mrt6msg *)skb_transport_header(skb);
+		msg->im6_mbz = 0;
+		msg->im6_msgtype = MRT6MSG_WHOLEPKT;
+		msg->im6_mif = reg_vif_num;
+		msg->im6_pad = 0;
+		ipv6_addr_copy(&msg->im6_src, &ipv6_hdr(pkt)->saddr);
+		ipv6_addr_copy(&msg->im6_dst, &ipv6_hdr(pkt)->daddr);
+
+		skb->ip_summed = CHECKSUM_UNNECESSARY;
+	} else
+#endif
+	{
 	/*
 	 *	Copy the IP header
 	 */
@@ -635,6 +824,7 @@ static int ip6mr_cache_report(struct sk_buff *pkt, vifi_t vifi, int assert)
 	skb->ip_summed = CHECKSUM_UNNECESSARY;
 
 	skb_pull(skb, sizeof(struct ipv6hdr));
+	}
 
 	if (mroute6_socket == NULL) {
 		kfree_skb(skb);
@@ -1033,6 +1223,44 @@ int ip6_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, int
 		rtnl_unlock();
 		return ret;
 
+	/*
+	 *	Control PIM assert (to activate pim will activate assert)
+	 */
+	case MRT6_ASSERT:
+	{
+		int v;
+		if (get_user(v, (int __user *)optval))
+			return -EFAULT;
+		mroute_do_assert = !!v;
+		return 0;
+	}
+
+#ifdef CONFIG_IPV6_PIMSM_V2
+	case MRT6_PIM:
+	{
+		int v, ret;
+		if (get_user(v, (int __user *)optval))
+			return -EFAULT;
+		v = !!v;
+		rtnl_lock();
+		ret = 0;
+		if (v != mroute_do_pim) {
+			mroute_do_pim = v;
+			mroute_do_assert = v;
+			if (mroute_do_pim)
+				ret = inet6_add_protocol(&pim6_protocol,
+							 IPPROTO_PIM);
+			else
+				ret = inet6_del_protocol(&pim6_protocol,
+							 IPPROTO_PIM);
+			if (ret < 0)
+				ret = -EAGAIN;
+		}
+		rtnl_unlock();
+		return ret;
+	}
+
+#endif
 	/*
 	 *	Spurious command, or MRT_VERSION which you cannot
 	 *	set.
@@ -1056,6 +1284,14 @@ int ip6_mroute_getsockopt(struct sock *sk, int optname, char __user *optval,
 	case MRT6_VERSION:
 		val = 0x0305;
 		break;
+#ifdef CONFIG_IPV6_PIMSM_V2
+	case MRT6_PIM:
+		val = mroute_do_pim;
+		break;
+#endif
+	case MRT6_ASSERT:
+		val = mroute_do_assert;
+		break;
 	default:
 		return -ENOPROTOOPT;
 	}
@@ -1151,6 +1387,18 @@ static int ip6mr_forward2(struct sk_buff *skb, struct mfc6_cache *c, int vifi)
 	if (vif->dev == NULL)
 		goto out_free;
 
+#ifdef CONFIG_IPV6_PIMSM_V2
+	if (vif->flags & MIFF_REGISTER) {
+		vif->pkt_out++;
+		vif->bytes_out += skb->len;
+		((struct net_device_stats *)netdev_priv(vif->dev))->tx_bytes += skb->len;
+		((struct net_device_stats *)netdev_priv(vif->dev))->tx_packets++;
+		ip6mr_cache_report(skb, vifi, MRT6MSG_WHOLEPKT);
+		kfree_skb(skb);
+		return 0;
+	}
+#endif
+
 	ipv6h = ipv6_hdr(skb);
 
 	fl = (struct flowi) {
@@ -1220,6 +1468,30 @@ static int ip6_mr_forward(struct sk_buff *skb, struct mfc6_cache *cache)
 	cache->mfc_un.res.pkt++;
 	cache->mfc_un.res.bytes += skb->len;
 
+	/*
+	 * Wrong interface: drop packet and (maybe) send PIM assert.
+	 */
+	if (vif6_table[vif].dev != skb->dev) {
+		int true_vifi;
+
+		cache->mfc_un.res.wrong_if++;
+		true_vifi = ip6mr_find_vif(skb->dev);
+
+		if (true_vifi >= 0 && mroute_do_assert &&
+		    /* pimsm uses asserts, when switching from RPT to SPT,
+		       so that we cannot check that packet arrived on an oif.
+		       It is bad, but otherwise we would need to move pretty
+		       large chunk of pimd to kernel. Ough... --ANK
+		     */
+		    (mroute_do_pim || cache->mfc_un.res.ttls[true_vifi] < 255) &&
+		    time_after(jiffies,
+			       cache->mfc_un.res.last_assert + MFC_ASSERT_THRESH)) {
+			cache->mfc_un.res.last_assert = jiffies;
+			ip6mr_cache_report(skb, true_vifi, MRT6MSG_WRONGMIF);
+		}
+		goto dont_forward;
+	}
+
 	vif6_table[vif].pkt_in++;
 	vif6_table[vif].bytes_in += skb->len;
 
@@ -1241,6 +1513,7 @@ static int ip6_mr_forward(struct sk_buff *skb, struct mfc6_cache *cache)
 		return 0;
 	}
 
+dont_forward:
 	kfree_skb(skb);
 	return 0;
 }
-- 
cgit v1.2.3


From 12802d058a003048104fe405a8d283b94ac50801 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Thu, 3 Apr 2008 09:22:56 +0900
Subject: [IPV6]: Comment MRT6_xxx sockopts in include/linux/in6.h.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/in6.h | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/in6.h b/include/linux/in6.h
index f674000c6c99..e6aa8de2b939 100644
--- a/include/linux/in6.h
+++ b/include/linux/in6.h
@@ -260,4 +260,19 @@ struct in6_flowlabel_req
 #define IPV6_PREFER_SRC_CGA		0x0008
 #define IPV6_PREFER_SRC_NONCGA		0x0800
 
+/*
+ * Multicast Routing:
+ * see include/linux/mroute6.h.
+ *
+ * MRT6_INIT			200
+ * MRT6_DONE			201
+ * MRT6_ADD_MIF			202
+ * MRT6_DEL_MIF			203
+ * MRT6_ADD_MFC			204
+ * MRT6_DEL_MFC			205
+ * MRT6_VERSION			206
+ * MRT6_ASSERT			207
+ * MRT6_PIM			208
+ * (reserved)			209
+ */
 #endif
-- 
cgit v1.2.3


From 8d855317fcf7fd9bd900d1e5ef1bea1b14bbe6af Mon Sep 17 00:00:00 2001
From: Stelian Pop <stelian@popies.net>
Date: Wed, 5 Mar 2008 00:00:00 +0100
Subject: atmel_usba_udc: move endpoint declarations into platform data.

The atmel_usba_udc driver is being used by several platforms and arches
(avr32 and at91 ATM), and each platform may have different endpoint
settings.

The patch below moves the endpoint declarations into the platform
data and make the necessary adjustments for AVR32 (improved by
Haavard Skinnemoen <hskinnemoen@atmel.com>).

Signed-off-by: Stelian Pop <stelian@popies.net>
Acked-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
---
 arch/avr32/mach-at32ap/at32ap700x.c   | 50 +++++++++++++++++++++---
 drivers/usb/gadget/atmel_usba_udc.c   | 71 +++++++++++++++++++----------------
 include/asm-avr32/arch-at32ap/board.h |  4 +-
 include/linux/usb/atmel_usba_udc.h    | 22 +++++++++++
 4 files changed, 105 insertions(+), 42 deletions(-)
 create mode 100644 include/linux/usb/atmel_usba_udc.h

(limited to 'include/linux')

diff --git a/arch/avr32/mach-at32ap/at32ap700x.c b/arch/avr32/mach-at32ap/at32ap700x.c
index 7678fee9a885..cb47afc9fffc 100644
--- a/arch/avr32/mach-at32ap/at32ap700x.c
+++ b/arch/avr32/mach-at32ap/at32ap700x.c
@@ -11,6 +11,7 @@
 #include <linux/platform_device.h>
 #include <linux/dma-mapping.h>
 #include <linux/spi/spi.h>
+#include <linux/usb/atmel_usba_udc.h>
 
 #include <asm/io.h>
 #include <asm/irq.h>
@@ -1351,9 +1352,39 @@ static struct clk usba0_hclk = {
 	.index		= 6,
 };
 
+#define EP(nam, idx, maxpkt, maxbk, dma, isoc)			\
+	[idx] = {						\
+		.name		= nam,				\
+		.index		= idx,				\
+		.fifo_size	= maxpkt,			\
+		.nr_banks	= maxbk,			\
+		.can_dma	= dma,				\
+		.can_isoc	= isoc,				\
+	}
+
+static struct usba_ep_data at32_usba_ep[] __initdata = {
+	EP("ep0",     0,   64, 1, 0, 0),
+	EP("ep1",     1,  512, 2, 1, 1),
+	EP("ep2",     2,  512, 2, 1, 1),
+	EP("ep3-int", 3,   64, 3, 1, 0),
+	EP("ep4-int", 4,   64, 3, 1, 0),
+	EP("ep5",     5, 1024, 3, 1, 1),
+	EP("ep6",     6, 1024, 3, 1, 1),
+};
+
+#undef EP
+
 struct platform_device *__init
 at32_add_device_usba(unsigned int id, struct usba_platform_data *data)
 {
+	/*
+	 * pdata doesn't have room for any endpoints, so we need to
+	 * append room for the ones we need right after it.
+	 */
+	struct {
+		struct usba_platform_data pdata;
+		struct usba_ep_data ep[7];
+	} usba_data;
 	struct platform_device *pdev;
 
 	if (id != 0)
@@ -1367,13 +1398,20 @@ at32_add_device_usba(unsigned int id, struct usba_platform_data *data)
 					  ARRAY_SIZE(usba0_resource)))
 		goto out_free_pdev;
 
-	if (data) {
-		if (platform_device_add_data(pdev, data, sizeof(*data)))
-			goto out_free_pdev;
+	if (data)
+		usba_data.pdata.vbus_pin = data->vbus_pin;
+	else
+		usba_data.pdata.vbus_pin = -EINVAL;
 
-		if (data->vbus_pin != GPIO_PIN_NONE)
-			at32_select_gpio(data->vbus_pin, 0);
-	}
+	data = &usba_data.pdata;
+	data->num_ep = ARRAY_SIZE(at32_usba_ep);
+	memcpy(data->ep, at32_usba_ep, sizeof(at32_usba_ep));
+
+	if (platform_device_add_data(pdev, data, sizeof(usba_data)))
+		goto out_free_pdev;
+
+	if (data->vbus_pin >= 0)
+		at32_select_gpio(data->vbus_pin, 0);
 
 	usba0_pclk.dev = &pdev->dev;
 	usba0_hclk.dev = &pdev->dev;
diff --git a/drivers/usb/gadget/atmel_usba_udc.c b/drivers/usb/gadget/atmel_usba_udc.c
index 39d187fb038a..71d3c5171f89 100644
--- a/drivers/usb/gadget/atmel_usba_udc.c
+++ b/drivers/usb/gadget/atmel_usba_udc.c
@@ -18,6 +18,7 @@
 #include <linux/platform_device.h>
 #include <linux/usb/ch9.h>
 #include <linux/usb/gadget.h>
+#include <linux/usb/atmel_usba_udc.h>
 #include <linux/delay.h>
 
 #include <asm/gpio.h>
@@ -27,6 +28,7 @@
 
 
 static struct usba_udc the_udc;
+static struct usba_ep *usba_ep;
 
 #ifdef CONFIG_USB_GADGET_DEBUG_FS
 #include <linux/debugfs.h>
@@ -982,33 +984,6 @@ static const struct usb_gadget_ops usba_udc_ops = {
 	.set_selfpowered	= usba_udc_set_selfpowered,
 };
 
-#define EP(nam, idx, maxpkt, maxbk, dma, isoc)			\
-{								\
-	.ep	= {						\
-		.ops		= &usba_ep_ops,			\
-		.name		= nam,				\
-		.maxpacket	= maxpkt,			\
-	},							\
-	.udc		= &the_udc,				\
-	.queue		= LIST_HEAD_INIT(usba_ep[idx].queue),	\
-	.fifo_size	= maxpkt,				\
-	.nr_banks	= maxbk,				\
-	.index		= idx,					\
-	.can_dma	= dma,					\
-	.can_isoc	= isoc,					\
-}
-
-static struct usba_ep usba_ep[] = {
-	EP("ep0",     0,   64, 1, 0, 0),
-	EP("ep1",     1,  512, 2, 1, 1),
-	EP("ep2",     2,  512, 2, 1, 1),
-	EP("ep3-int", 3,   64, 3, 1, 0),
-	EP("ep4-int", 4,   64, 3, 1, 0),
-	EP("ep5",     5, 1024, 3, 1, 1),
-	EP("ep6",     6, 1024, 3, 1, 1),
-};
-#undef EP
-
 static struct usb_endpoint_descriptor usba_ep0_desc = {
 	.bLength = USB_DT_ENDPOINT_SIZE,
 	.bDescriptorType = USB_DT_ENDPOINT,
@@ -1027,7 +1002,6 @@ static void nop_release(struct device *dev)
 static struct usba_udc the_udc = {
 	.gadget	= {
 		.ops		= &usba_udc_ops,
-		.ep0		= &usba_ep[0].ep,
 		.ep_list	= LIST_HEAD_INIT(the_udc.gadget.ep_list),
 		.is_dualspeed	= 1,
 		.name		= "atmel_usba_udc",
@@ -1861,7 +1835,7 @@ static int __init usba_udc_probe(struct platform_device *pdev)
 
 	regs = platform_get_resource(pdev, IORESOURCE_MEM, CTRL_IOMEM_ID);
 	fifo = platform_get_resource(pdev, IORESOURCE_MEM, FIFO_IOMEM_ID);
-	if (!regs || !fifo)
+	if (!regs || !fifo || !pdata)
 		return -ENXIO;
 
 	irq = platform_get_irq(pdev, 0);
@@ -1909,16 +1883,44 @@ static int __init usba_udc_probe(struct platform_device *pdev)
 	usba_writel(udc, CTRL, 0);
 	clk_disable(pclk);
 
+	usba_ep = kmalloc(sizeof(struct usba_ep) * pdata->num_ep,
+			  GFP_KERNEL);
+	if (!usba_ep)
+		goto err_alloc_ep;
+
+	the_udc.gadget.ep0 = &usba_ep[0].ep;
+
 	INIT_LIST_HEAD(&usba_ep[0].ep.ep_list);
 	usba_ep[0].ep_regs = udc->regs + USBA_EPT_BASE(0);
 	usba_ep[0].dma_regs = udc->regs + USBA_DMA_BASE(0);
 	usba_ep[0].fifo = udc->fifo + USBA_FIFO_BASE(0);
-	for (i = 1; i < ARRAY_SIZE(usba_ep); i++) {
+	usba_ep[0].ep.ops = &usba_ep_ops;
+	usba_ep[0].ep.name = pdata->ep[0].name;
+	usba_ep[0].ep.maxpacket = pdata->ep[0].fifo_size;
+	usba_ep[0].udc = &the_udc;
+	INIT_LIST_HEAD(&usba_ep[0].queue);
+	usba_ep[0].fifo_size = pdata->ep[0].fifo_size;
+	usba_ep[0].nr_banks = pdata->ep[0].nr_banks;
+	usba_ep[0].index = pdata->ep[0].index;
+	usba_ep[0].can_dma = pdata->ep[0].can_dma;
+	usba_ep[0].can_isoc = pdata->ep[0].can_isoc;
+
+	for (i = 1; i < pdata->num_ep; i++) {
 		struct usba_ep *ep = &usba_ep[i];
 
 		ep->ep_regs = udc->regs + USBA_EPT_BASE(i);
 		ep->dma_regs = udc->regs + USBA_DMA_BASE(i);
 		ep->fifo = udc->fifo + USBA_FIFO_BASE(i);
+		ep->ep.ops = &usba_ep_ops;
+		ep->ep.name = pdata->ep[i].name;
+		ep->ep.maxpacket = pdata->ep[i].fifo_size;
+		ep->udc = &the_udc;
+		INIT_LIST_HEAD(&ep->queue);
+		ep->fifo_size = pdata->ep[i].fifo_size;
+		ep->nr_banks = pdata->ep[i].nr_banks;
+		ep->index = pdata->ep[i].index;
+		ep->can_dma = pdata->ep[i].can_dma;
+		ep->can_isoc = pdata->ep[i].can_isoc;
 
 		list_add_tail(&ep->ep.ep_list, &udc->gadget.ep_list);
 	}
@@ -1937,7 +1939,7 @@ static int __init usba_udc_probe(struct platform_device *pdev)
 		goto err_device_add;
 	}
 
-	if (pdata && pdata->vbus_pin >= 0) {
+	if (pdata->vbus_pin >= 0) {
 		if (!gpio_request(pdata->vbus_pin, "atmel_usba_udc")) {
 			udc->vbus_pin = pdata->vbus_pin;
 
@@ -1957,7 +1959,7 @@ static int __init usba_udc_probe(struct platform_device *pdev)
 	}
 
 	usba_init_debugfs(udc);
-	for (i = 1; i < ARRAY_SIZE(usba_ep); i++)
+	for (i = 1; i < pdata->num_ep; i++)
 		usba_ep_init_debugfs(udc, &usba_ep[i]);
 
 	return 0;
@@ -1965,6 +1967,8 @@ static int __init usba_udc_probe(struct platform_device *pdev)
 err_device_add:
 	free_irq(irq, udc);
 err_request_irq:
+	kfree(usba_ep);
+err_alloc_ep:
 	iounmap(udc->fifo);
 err_map_fifo:
 	iounmap(udc->regs);
@@ -1982,10 +1986,11 @@ static int __exit usba_udc_remove(struct platform_device *pdev)
 {
 	struct usba_udc *udc;
 	int i;
+	struct usba_platform_data *pdata = pdev->dev.platform_data;
 
 	udc = platform_get_drvdata(pdev);
 
-	for (i = 1; i < ARRAY_SIZE(usba_ep); i++)
+	for (i = 1; i < pdata->num_ep; i++)
 		usba_ep_cleanup_debugfs(&usba_ep[i]);
 	usba_cleanup_debugfs(udc);
 
diff --git a/include/asm-avr32/arch-at32ap/board.h b/include/asm-avr32/arch-at32ap/board.h
index 7597b0bd2f01..3fea2004f7db 100644
--- a/include/asm-avr32/arch-at32ap/board.h
+++ b/include/asm-avr32/arch-at32ap/board.h
@@ -38,9 +38,7 @@ struct platform_device *
 at32_add_device_lcdc(unsigned int id, struct atmel_lcdfb_info *data,
 		     unsigned long fbmem_start, unsigned long fbmem_len);
 
-struct usba_platform_data {
-	int vbus_pin;
-};
+struct usba_platform_data;
 struct platform_device *
 at32_add_device_usba(unsigned int id, struct usba_platform_data *data);
 
diff --git a/include/linux/usb/atmel_usba_udc.h b/include/linux/usb/atmel_usba_udc.h
new file mode 100644
index 000000000000..6311fa2d9f82
--- /dev/null
+++ b/include/linux/usb/atmel_usba_udc.h
@@ -0,0 +1,22 @@
+/*
+ * Platform data definitions for Atmel USBA gadget driver.
+ */
+#ifndef __LINUX_USB_USBA_H
+#define __LINUX_USB_USBA_H
+
+struct usba_ep_data {
+	char	*name;
+	int	index;
+	int	fifo_size;
+	int	nr_banks;
+	int	can_dma;
+	int	can_isoc;
+};
+
+struct usba_platform_data {
+	int			vbus_pin;
+	int			num_ep;
+	struct usba_ep_data	ep[0];
+};
+
+#endif /* __LINUX_USB_USBA_H */
-- 
cgit v1.2.3


From 834d97d452208279edf11c57eca150360d2dd1d6 Mon Sep 17 00:00:00 2001
From: Josh Boyer <jwboyer@linux.vnet.ibm.com>
Date: Thu, 27 Mar 2008 00:33:14 +1100
Subject: [POWERPC] Add of_device_is_available function

IEEE 1275 defined a standard "status" property to indicate the operational
status of a device.  The property has four possible values: okay, disabled,
fail, fail-xxx.  The absence of this property means the operational status
of the device is unknown or okay.

This adds a function called of_device_is_available that checks the state
of the status property of a device.  If the property is absent or set to
either "okay" or "ok", it returns 1.  Otherwise it returns 0.

Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 drivers/of/base.c  | 26 ++++++++++++++++++++++++++
 include/linux/of.h |  1 +
 2 files changed, 27 insertions(+)

(limited to 'include/linux')

diff --git a/drivers/of/base.c b/drivers/of/base.c
index 80c9deca5f35..9bd7c4a31253 100644
--- a/drivers/of/base.c
+++ b/drivers/of/base.c
@@ -116,6 +116,32 @@ int of_device_is_compatible(const struct device_node *device,
 }
 EXPORT_SYMBOL(of_device_is_compatible);
 
+/**
+ *  of_device_is_available - check if a device is available for use
+ *
+ *  @device: Node to check for availability
+ *
+ *  Returns 1 if the status property is absent or set to "okay" or "ok",
+ *  0 otherwise
+ */
+int of_device_is_available(const struct device_node *device)
+{
+	const char *status;
+	int statlen;
+
+	status = of_get_property(device, "status", &statlen);
+	if (status == NULL)
+		return 1;
+
+	if (statlen > 0) {
+		if (!strcmp(status, "okay") || !strcmp(status, "ok"))
+			return 1;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(of_device_is_available);
+
 /**
  *	of_get_parent - Get a node's parent if any
  *	@node:	Node to get parent
diff --git a/include/linux/of.h b/include/linux/of.h
index 6981016dcc25..59a61bdc98b6 100644
--- a/include/linux/of.h
+++ b/include/linux/of.h
@@ -62,6 +62,7 @@ extern struct property *of_find_property(const struct device_node *np,
 					 int *lenp);
 extern int of_device_is_compatible(const struct device_node *device,
 				   const char *);
+extern int of_device_is_available(const struct device_node *device);
 extern const void *of_get_property(const struct device_node *node,
 				const char *name,
 				int *lenp);
-- 
cgit v1.2.3


From 40f6b36c6243462fb95d0343237331c423494b03 Mon Sep 17 00:00:00 2001
From: Kai Makisara <Kai.Makisara@kolumbus.fi>
Date: Sun, 24 Feb 2008 22:23:24 +0200
Subject: [SCSI] st: add option to use SILI in variable block reads

Add new option MT_ST_SILI to enable setting the SILI bit in reads in variable
block mode. If SILI is set, reading a block shorter than the byte count does
not result in CHECK CONDITION. The length of the block is determined using the
residual count from the HBA. Avoiding the REQUEST SENSE command for every
block speeds up some real applications considerably.

Signed-off-by: Kai Makisara <kai.makisara@kolumbus.fi>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
---
 Documentation/scsi/st.txt |  7 ++++++-
 drivers/scsi/st.c         | 40 ++++++++++++++++++++++++++++++++++++----
 drivers/scsi/st.h         |  3 +++
 drivers/scsi/st_options.h |  6 +++++-
 include/linux/mtio.h      |  1 +
 5 files changed, 51 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/scsi/st.txt b/Documentation/scsi/st.txt
index b7be95b5bd24..38f81188def0 100644
--- a/Documentation/scsi/st.txt
+++ b/Documentation/scsi/st.txt
@@ -2,7 +2,7 @@ This file contains brief information about the SCSI tape driver.
 The driver is currently maintained by Kai Mäkisara (email
 Kai.Makisara@kolumbus.fi)
 
-Last modified: Mon Mar  7 21:14:44 2005 by kai.makisara
+Last modified: Thu Feb 21 21:54:16 2008 by kai.makisara
 
 
 BASICS
@@ -372,6 +372,11 @@ MTSETDRVBUFFER
 	     MT_ST_SYSV sets the SYSV semantics (mode)
 	     MT_ST_NOWAIT enables immediate mode (i.e., don't wait for
 	        the command to finish) for some commands (e.g., rewind)
+	     MT_ST_SILI enables setting the SILI bit in SCSI commands when
+		reading in variable block mode to enhance performance when
+		reading blocks shorter than the byte count; set this only
+		if you are sure that the drive supports SILI and the HBA
+		correctly returns transfer residuals
 	     MT_ST_DEBUGGING debugging (global; debugging must be
 		compiled into the driver)
 	MT_ST_SETBOOLEANS
diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c
index 0a52d9d2da2c..a4361a8c6ac6 100644
--- a/drivers/scsi/st.c
+++ b/drivers/scsi/st.c
@@ -17,7 +17,7 @@
    Last modified: 18-JAN-1998 Richard Gooch <rgooch@atnf.csiro.au> Devfs support
  */
 
-static const char *verstr = "20080221";
+static const char *verstr = "20080224";
 
 #include <linux/module.h>
 
@@ -183,6 +183,7 @@ static int modes_defined;
 
 static struct st_buffer *new_tape_buffer(int, int, int);
 static int enlarge_buffer(struct st_buffer *, int, int);
+static void clear_buffer(struct st_buffer *);
 static void normalize_buffer(struct st_buffer *);
 static int append_to_buffer(const char __user *, struct st_buffer *, int);
 static int from_buffer(struct st_buffer *, char __user *, int);
@@ -442,6 +443,7 @@ static void st_sleep_done(void *data, char *sense, int result, int resid)
 
 	memcpy(SRpnt->sense, sense, SCSI_SENSE_BUFFERSIZE);
 	(STp->buffer)->cmdstat.midlevel_result = SRpnt->result = result;
+	(STp->buffer)->cmdstat.residual = resid;
 	DEB( STp->write_pending = 0; )
 
 	if (SRpnt->waiting)
@@ -1159,6 +1161,7 @@ static int st_open(struct inode *inode, struct file *filp)
 		goto err_out;
 	}
 
+	(STp->buffer)->cleared = 0;
 	(STp->buffer)->writing = 0;
 	(STp->buffer)->syscall_result = 0;
 
@@ -1432,8 +1435,14 @@ static int setup_buffering(struct scsi_tape *STp, const char __user *buf,
 		if (STp->block_size)
 			bufsize = STp->block_size > st_fixed_buffer_size ?
 				STp->block_size : st_fixed_buffer_size;
-		else
+		else {
 			bufsize = count;
+			/* Make sure that data from previous user is not leaked even if
+			   HBA does not return correct residual */
+			if (is_read && STp->sili && !STbp->cleared)
+				clear_buffer(STbp);
+		}
+
 		if (bufsize > STbp->buffer_size &&
 		    !enlarge_buffer(STbp, bufsize, STp->restr_dma)) {
 			printk(KERN_WARNING "%s: Can't allocate %d byte tape buffer.\n",
@@ -1783,6 +1792,8 @@ static long read_tape(struct scsi_tape *STp, long count,
 	memset(cmd, 0, MAX_COMMAND_SIZE);
 	cmd[0] = READ_6;
 	cmd[1] = (STp->block_size != 0);
+	if (!cmd[1] && STp->sili)
+		cmd[1] |= 2;
 	cmd[2] = blks >> 16;
 	cmd[3] = blks >> 8;
 	cmd[4] = blks;
@@ -1911,8 +1922,11 @@ static long read_tape(struct scsi_tape *STp, long count,
 
 	}
 	/* End of error handling */ 
-	else			/* Read successful */
+	else {			/* Read successful */
 		STbp->buffer_bytes = bytes;
+		if (STp->sili) /* In fixed block mode residual is always zero here */
+			STbp->buffer_bytes -= STp->buffer->cmdstat.residual;
+	}
 
 	if (STps->drv_block >= 0) {
 		if (STp->block_size == 0)
@@ -2090,7 +2104,8 @@ static void st_log_options(struct scsi_tape * STp, struct st_modedef * STm, char
 		       name, STm->defaults_for_writes, STp->omit_blklims, STp->can_partitions,
 		       STp->scsi2_logical);
 		printk(KERN_INFO
-		       "%s:    sysv: %d nowait: %d\n", name, STm->sysv, STp->immediate);
+		       "%s:    sysv: %d nowait: %d sili: %d\n", name, STm->sysv, STp->immediate,
+			STp->sili);
 		printk(KERN_INFO "%s:    debugging: %d\n",
 		       name, debugging);
 	}
@@ -2133,6 +2148,7 @@ static int st_set_options(struct scsi_tape *STp, long options)
 		STp->scsi2_logical = (options & MT_ST_SCSI2LOGICAL) != 0;
 		STp->immediate = (options & MT_ST_NOWAIT) != 0;
 		STm->sysv = (options & MT_ST_SYSV) != 0;
+		STp->sili = (options & MT_ST_SILI) != 0;
 		DEB( debugging = (options & MT_ST_DEBUGGING) != 0;
 		     st_log_options(STp, STm, name); )
 	} else if (code == MT_ST_SETBOOLEANS || code == MT_ST_CLEARBOOLEANS) {
@@ -2164,6 +2180,8 @@ static int st_set_options(struct scsi_tape *STp, long options)
 			STp->immediate = value;
 		if ((options & MT_ST_SYSV) != 0)
 			STm->sysv = value;
+		if ((options & MT_ST_SILI) != 0)
+			STp->sili = value;
                 DEB(
 		if ((options & MT_ST_DEBUGGING) != 0)
 			debugging = value;
@@ -3655,6 +3673,8 @@ static int enlarge_buffer(struct st_buffer * STbuffer, int new_size, int need_dm
 		STbuffer->frp_segs += 1;
 		got += b_size;
 		STbuffer->buffer_size = got;
+		if (STbuffer->cleared)
+			memset(page_address(STbuffer->frp[segs].page), 0, b_size);
 		segs++;
 	}
 	STbuffer->b_data = page_address(STbuffer->frp[0].page);
@@ -3663,6 +3683,17 @@ static int enlarge_buffer(struct st_buffer * STbuffer, int new_size, int need_dm
 }
 
 
+/* Make sure that no data from previous user is in the internal buffer */
+static void clear_buffer(struct st_buffer * st_bp)
+{
+	int i;
+
+	for (i=0; i < st_bp->frp_segs; i++)
+		memset(page_address(st_bp->frp[i].page), 0, st_bp->frp[i].length);
+	st_bp->cleared = 1;
+}
+
+
 /* Release the extra buffer */
 static void normalize_buffer(struct st_buffer * STbuffer)
 {
@@ -3987,6 +4018,7 @@ static int st_probe(struct device *dev)
 	tpnt->two_fm = ST_TWO_FM;
 	tpnt->fast_mteom = ST_FAST_MTEOM;
 	tpnt->scsi2_logical = ST_SCSI2LOGICAL;
+	tpnt->sili = ST_SILI;
 	tpnt->immediate = ST_NOWAIT;
 	tpnt->default_drvbuffer = 0xff;		/* No forced buffering */
 	tpnt->partition = 0;
diff --git a/drivers/scsi/st.h b/drivers/scsi/st.h
index 5931726fcf93..b92712f95931 100644
--- a/drivers/scsi/st.h
+++ b/drivers/scsi/st.h
@@ -12,6 +12,7 @@ struct st_cmdstatus {
 	int midlevel_result;
 	struct scsi_sense_hdr sense_hdr;
 	int have_sense;
+	int residual;
 	u64 uremainder64;
 	u8 flags;
 	u8 remainder_valid;
@@ -34,6 +35,7 @@ struct st_request {
 struct st_buffer {
 	unsigned char dma;	/* DMA-able buffer */
 	unsigned char do_dio;   /* direct i/o set up? */
+	unsigned char cleared;  /* internal buffer cleared after open? */
 	int buffer_size;
 	int buffer_blocks;
 	int buffer_bytes;
@@ -122,6 +124,7 @@ struct scsi_tape {
 	unsigned char try_dio_now;		/* try direct i/o before next close? */
 	unsigned char c_algo;			/* compression algorithm */
 	unsigned char pos_unknown;			/* after reset position unknown */
+	unsigned char sili;			/* use SILI when reading in variable b mode */
 	int tape_type;
 	int long_timeout;	/* timeout for commands known to take long time */
 
diff --git a/drivers/scsi/st_options.h b/drivers/scsi/st_options.h
index b6b5c9c37677..d2f947935554 100644
--- a/drivers/scsi/st_options.h
+++ b/drivers/scsi/st_options.h
@@ -3,7 +3,7 @@
 
    Copyright 1995-2003 Kai Makisara.
 
-   Last modified: Mon Apr  7 22:49:18 2003 by makisara
+   Last modified: Thu Feb 21 21:47:07 2008 by kai.makisara
 */
 
 #ifndef _ST_OPTIONS_H
@@ -94,6 +94,10 @@
    The default is BSD semantics. */
 #define ST_SYSV 0
 
+/* If ST_SILI is non-zero, the SILI bit is set when reading in variable block
+   mode and the block size is determined using the residual returned by the HBA. */
+#define ST_SILI 0
+
 /* Time to wait for the drive to become ready if blocking open */
 #define ST_BLOCK_SECONDS     120
 
diff --git a/include/linux/mtio.h b/include/linux/mtio.h
index 6f8d2d45a8fb..ef01d6aa5934 100644
--- a/include/linux/mtio.h
+++ b/include/linux/mtio.h
@@ -192,6 +192,7 @@ struct	mtpos {
 #define MT_ST_SCSI2LOGICAL      0x800
 #define MT_ST_SYSV              0x1000
 #define MT_ST_NOWAIT            0x2000
+#define MT_ST_SILI		0x4000
 
 /* The mode parameters to be controlled. Parameter chosen with bits 20-28 */
 #define MT_ST_CLEAR_DEFAULT	0xfffff
-- 
cgit v1.2.3


From b1adaf65ba0398c9a1adc8f3a274533165a4df61 Mon Sep 17 00:00:00 2001
From: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Date: Tue, 18 Mar 2008 00:15:03 +0900
Subject: [SCSI] block: add sg buffer copy helper functions

This patch adds new three helper functions to copy data between an SG
list and a linear buffer.

- sg_copy_from_buffer copies data from linear buffer to an SG list

- sg_copy_to_buffer copies data from an SG list to a linear buffer

When the APIs copy data from a linear buffer to an SG list,
flush_kernel_dcache_page is called. It's not necessary for everyone
but it's a no-op on most architectures and in general the API is not
used in performance critical path.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
---
 include/linux/scatterlist.h |   5 +++
 lib/scatterlist.c           | 102 ++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 107 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/scatterlist.h b/include/linux/scatterlist.h
index a3d567a974e8..71fc81360048 100644
--- a/include/linux/scatterlist.h
+++ b/include/linux/scatterlist.h
@@ -213,6 +213,11 @@ int __sg_alloc_table(struct sg_table *, unsigned int, unsigned int, gfp_t,
 		     sg_alloc_fn *);
 int sg_alloc_table(struct sg_table *, unsigned int, gfp_t);
 
+size_t sg_copy_from_buffer(struct scatterlist *sgl, unsigned int nents,
+			   void *buf, size_t buflen);
+size_t sg_copy_to_buffer(struct scatterlist *sgl, unsigned int nents,
+			 void *buf, size_t buflen);
+
 /*
  * Maximum number of entries that will be allocated in one piece, if
  * a list larger than this is required then chaining will be utilized.
diff --git a/lib/scatterlist.c b/lib/scatterlist.c
index acca4901046c..b80c21100d78 100644
--- a/lib/scatterlist.c
+++ b/lib/scatterlist.c
@@ -8,6 +8,7 @@
  */
 #include <linux/module.h>
 #include <linux/scatterlist.h>
+#include <linux/highmem.h>
 
 /**
  * sg_next - return the next scatterlist entry in a list
@@ -292,3 +293,104 @@ int sg_alloc_table(struct sg_table *table, unsigned int nents, gfp_t gfp_mask)
 	return ret;
 }
 EXPORT_SYMBOL(sg_alloc_table);
+
+/**
+ * sg_copy_buffer - Copy data between a linear buffer and an SG list
+ * @sgl:		 The SG list
+ * @nents:		 Number of SG entries
+ * @buf:		 Where to copy from
+ * @buflen:		 The number of bytes to copy
+ * @to_buffer: 		 transfer direction (non zero == from an sg list to a
+ * 			 buffer, 0 == from a buffer to an sg list
+ *
+ * Returns the number of copied bytes.
+ *
+ **/
+static size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents,
+			     void *buf, size_t buflen, int to_buffer)
+{
+	struct scatterlist *sg;
+	size_t buf_off = 0;
+	int i;
+
+	WARN_ON(!irqs_disabled());
+
+	for_each_sg(sgl, sg, nents, i) {
+		struct page *page;
+		int n = 0;
+		unsigned int sg_off = sg->offset;
+		unsigned int sg_copy = sg->length;
+
+		if (sg_copy > buflen)
+			sg_copy = buflen;
+		buflen -= sg_copy;
+
+		while (sg_copy > 0) {
+			unsigned int page_copy;
+			void *p;
+
+			page_copy = PAGE_SIZE - sg_off;
+			if (page_copy > sg_copy)
+				page_copy = sg_copy;
+
+			page = nth_page(sg_page(sg), n);
+			p = kmap_atomic(page, KM_BIO_SRC_IRQ);
+
+			if (to_buffer)
+				memcpy(buf + buf_off, p + sg_off, page_copy);
+			else {
+				memcpy(p + sg_off, buf + buf_off, page_copy);
+				flush_kernel_dcache_page(page);
+			}
+
+			kunmap_atomic(p, KM_BIO_SRC_IRQ);
+
+			buf_off += page_copy;
+			sg_off += page_copy;
+			if (sg_off == PAGE_SIZE) {
+				sg_off = 0;
+				n++;
+			}
+			sg_copy -= page_copy;
+		}
+
+		if (!buflen)
+			break;
+	}
+
+	return buf_off;
+}
+
+/**
+ * sg_copy_from_buffer - Copy from a linear buffer to an SG list
+ * @sgl:		 The SG list
+ * @nents:		 Number of SG entries
+ * @buf:		 Where to copy from
+ * @buflen:		 The number of bytes to copy
+ *
+ * Returns the number of copied bytes.
+ *
+ **/
+size_t sg_copy_from_buffer(struct scatterlist *sgl, unsigned int nents,
+			   void *buf, size_t buflen)
+{
+	return sg_copy_buffer(sgl, nents, buf, buflen, 0);
+}
+EXPORT_SYMBOL(sg_copy_from_buffer);
+
+/**
+ * sg_copy_to_buffer - Copy from an SG list to a linear buffer
+ * @sgl:		 The SG list
+ * @nents:		 Number of SG entries
+ * @buf:		 Where to copy to
+ * @buflen:		 The number of bytes to copy
+ *
+ * Returns the number of copied bytes.
+ *
+ **/
+size_t sg_copy_to_buffer(struct scatterlist *sgl, unsigned int nents,
+			 void *buf, size_t buflen)
+{
+	return sg_copy_buffer(sgl, nents, buf, buflen, 1);
+}
+EXPORT_SYMBOL(sg_copy_to_buffer);
-- 
cgit v1.2.3


From 2f3edc6936e3f6be3f1df1e89c141ae028fa605e Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@HansenPartnership.com>
Date: Wed, 2 Apr 2008 10:05:48 -0500
Subject: [SCSI] transport_class: BUG if we can't release the attribute
 container

Every current transport class calls transport_container_release but
ignores the return value.  This is catastrophic if it returns an error
because the containers are part of a global list and the next action of
almost every transport class is to free the memory used by the
container.

Fix this by making transport_container_release a void, but making it BUG
if attribute_container_release returns an error ... this catches the
root cause of a system panic much earlier.  If we don't do this, we get
an eventual BUG when the attribute container list notices the corruption
caused by the freed memory it's still referencing.

Also made attribute_container_release __must_check as a reminder.

Cc: Greg KH <greg@kroah.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
---
 drivers/base/transport_class.c      | 3 ++-
 drivers/scsi/raid_class.c           | 2 +-
 include/linux/attribute_container.h | 2 +-
 include/linux/transport_class.h     | 5 +++--
 4 files changed, 7 insertions(+), 5 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/base/transport_class.c b/drivers/base/transport_class.c
index 40bca48abc12..cabd0edf2156 100644
--- a/drivers/base/transport_class.c
+++ b/drivers/base/transport_class.c
@@ -108,7 +108,8 @@ EXPORT_SYMBOL_GPL(anon_transport_class_register);
  */
 void anon_transport_class_unregister(struct anon_transport_class *atc)
 {
-	attribute_container_unregister(&atc->container);
+	if (unlikely(attribute_container_unregister(&atc->container)))
+		BUG();
 }
 EXPORT_SYMBOL_GPL(anon_transport_class_unregister);
 
diff --git a/drivers/scsi/raid_class.c b/drivers/scsi/raid_class.c
index 86e13183c9ba..52182a744ba6 100644
--- a/drivers/scsi/raid_class.c
+++ b/drivers/scsi/raid_class.c
@@ -289,7 +289,7 @@ raid_class_release(struct raid_template *r)
 {
 	struct raid_internal *i = to_raid_internal(r);
 
-	attribute_container_unregister(&i->r.raid_attrs.ac);
+	BUG_ON(attribute_container_unregister(&i->r.raid_attrs.ac));
 
 	kfree(i);
 }
diff --git a/include/linux/attribute_container.h b/include/linux/attribute_container.h
index f5582332af04..574b201b99d8 100644
--- a/include/linux/attribute_container.h
+++ b/include/linux/attribute_container.h
@@ -37,7 +37,7 @@ attribute_container_set_no_classdevs(struct attribute_container *atc)
 }
 
 int attribute_container_register(struct attribute_container *cont);
-int attribute_container_unregister(struct attribute_container *cont);
+int __must_check attribute_container_unregister(struct attribute_container *cont);
 void attribute_container_create_device(struct device *dev,
 				       int (*fn)(struct attribute_container *,
 						 struct device *,
diff --git a/include/linux/transport_class.h b/include/linux/transport_class.h
index 1d6cc22e5f42..6696cf79c4f7 100644
--- a/include/linux/transport_class.h
+++ b/include/linux/transport_class.h
@@ -86,9 +86,10 @@ static inline int transport_container_register(struct transport_container *tc)
 	return attribute_container_register(&tc->ac);
 }
 
-static inline int transport_container_unregister(struct transport_container *tc)
+static inline void transport_container_unregister(struct transport_container *tc)
 {
-	return attribute_container_unregister(&tc->ac);
+	if (unlikely(attribute_container_unregister(&tc->ac)))
+		BUG();
 }
 
 int transport_class_register(struct transport_class *);
-- 
cgit v1.2.3


From 2557a933b795c1988c721ebb871cd735128bb9cb Mon Sep 17 00:00:00 2001
From: Rusty Russell <rusty@rustcorp.com.au>
Date: Mon, 7 Apr 2008 14:30:28 +1000
Subject: virtio: remove overzealous BUG_ON.

The 'disable_cb' callback is designed as an optimization to tell the host
we don't need callbacks now.  As it is not reliable, the debug check is
overzealous: it can happen on two CPUs at the same time.  Document this.

Even if it were reliable, the virtio_net driver doesn't disable
callbacks on transmit so the START_USE/END_USE debugging reentrance
protection can be easily tripped even on UP.

Thanks to Balaji Rao for the bug report and testing.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
CC: Balaji Rao <balajirrao@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/virtio/virtio_ring.c | 3 ---
 include/linux/virtio.h       | 5 ++++-
 2 files changed, 4 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
index aa714028641e..c2fa5c630813 100644
--- a/drivers/virtio/virtio_ring.c
+++ b/drivers/virtio/virtio_ring.c
@@ -214,10 +214,7 @@ static void vring_disable_cb(struct virtqueue *_vq)
 {
 	struct vring_virtqueue *vq = to_vvq(_vq);
 
-	START_USE(vq);
-	BUG_ON(vq->vring.avail->flags & VRING_AVAIL_F_NO_INTERRUPT);
 	vq->vring.avail->flags |= VRING_AVAIL_F_NO_INTERRUPT;
-	END_USE(vq);
 }
 
 static bool vring_enable_cb(struct virtqueue *_vq)
diff --git a/include/linux/virtio.h b/include/linux/virtio.h
index 12c18ac1b973..e7d10845b3c1 100644
--- a/include/linux/virtio.h
+++ b/include/linux/virtio.h
@@ -41,6 +41,8 @@ struct virtqueue
  *	Returns NULL or the "data" token handed to add_buf.
  * @disable_cb: disable callbacks
  *	vq: the struct virtqueue we're talking about.
+ *	Note that this is not necessarily synchronous, hence unreliable and only
+ *	useful as an optimization.
  * @enable_cb: restart callbacks after disable_cb.
  *	vq: the struct virtqueue we're talking about.
  *	This re-enables callbacks; it returns "false" if there are pending
@@ -48,7 +50,8 @@ struct virtqueue
  *	checking for more work, and enabling callbacks.
  *
  * Locking rules are straightforward: the driver is responsible for
- * locking.  No two operations may be invoked simultaneously.
+ * locking.  No two operations may be invoked simultaneously, with the exception
+ * of @disable_cb.
  *
  * All operations can be called in any context.
  */
-- 
cgit v1.2.3


From 8fe2b65a18e49bfde56a59ed4ab3fc7aa0c2f325 Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Sun, 30 Mar 2008 00:10:50 +0100
Subject: ssb: Turn suspend/resume upside down

Turn the SSB bus suspend mechanism upside down.
Instead of deciding by an internal reference count when to suspend/resume,
let the parent bus call us in their suspend/resume routine.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/net/wireless/b43/pcmcia.c         | 10 ++--
 drivers/ssb/driver_chipcommon.c           |  2 +-
 drivers/ssb/main.c                        | 81 +++++++++++++------------------
 drivers/ssb/pcihost_wrapper.c             | 10 ++++
 drivers/ssb/pcmcia.c                      | 34 +++++++++----
 drivers/ssb/ssb_private.h                 |  5 ++
 include/linux/ssb/ssb.h                   | 10 ++--
 include/linux/ssb/ssb_driver_chipcommon.h |  3 +-
 8 files changed, 88 insertions(+), 67 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/wireless/b43/pcmcia.c b/drivers/net/wireless/b43/pcmcia.c
index 371e4a119511..b8aa16307f79 100644
--- a/drivers/net/wireless/b43/pcmcia.c
+++ b/drivers/net/wireless/b43/pcmcia.c
@@ -43,14 +43,16 @@ MODULE_DEVICE_TABLE(pcmcia, b43_pcmcia_tbl);
 #ifdef CONFIG_PM
 static int b43_pcmcia_suspend(struct pcmcia_device *dev)
 {
-	//TODO
-	return 0;
+	struct ssb_bus *ssb = dev->priv;
+
+	return ssb_bus_suspend(ssb);
 }
 
 static int b43_pcmcia_resume(struct pcmcia_device *dev)
 {
-	//TODO
-	return 0;
+	struct ssb_bus *ssb = dev->priv;
+
+	return ssb_bus_resume(ssb);
 }
 #else /* CONFIG_PM */
 # define b43_pcmcia_suspend		NULL
diff --git a/drivers/ssb/driver_chipcommon.c b/drivers/ssb/driver_chipcommon.c
index 45b672a69003..571f4fd55236 100644
--- a/drivers/ssb/driver_chipcommon.c
+++ b/drivers/ssb/driver_chipcommon.c
@@ -251,7 +251,7 @@ void ssb_chipcommon_init(struct ssb_chipcommon *cc)
 	calc_fast_powerup_delay(cc);
 }
 
-void ssb_chipco_suspend(struct ssb_chipcommon *cc, pm_message_t state)
+void ssb_chipco_suspend(struct ssb_chipcommon *cc)
 {
 	if (!cc->dev)
 		return;
diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index 2fcfd73b3b6e..c0cbdba07aee 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -120,35 +120,12 @@ static void ssb_device_put(struct ssb_device *dev)
 		put_device(dev->dev);
 }
 
-static int ssb_bus_resume(struct ssb_bus *bus)
-{
-	int err;
-
-	ssb_pci_xtal(bus, SSB_GPIO_XTAL | SSB_GPIO_PLL, 1);
-	err = ssb_pcmcia_init(bus);
-	if (err) {
-		/* No need to disable XTAL, as we don't have one on PCMCIA. */
-		return err;
-	}
-	ssb_chipco_resume(&bus->chipco);
-
-	return 0;
-}
-
 static int ssb_device_resume(struct device *dev)
 {
 	struct ssb_device *ssb_dev = dev_to_ssb_dev(dev);
 	struct ssb_driver *ssb_drv;
-	struct ssb_bus *bus;
 	int err = 0;
 
-	bus = ssb_dev->bus;
-	if (bus->suspend_cnt == bus->nr_devices) {
-		err = ssb_bus_resume(bus);
-		if (err)
-			return err;
-	}
-	bus->suspend_cnt--;
 	if (dev->driver) {
 		ssb_drv = drv_to_ssb_drv(dev->driver);
 		if (ssb_drv && ssb_drv->resume)
@@ -160,27 +137,10 @@ out:
 	return err;
 }
 
-static void ssb_bus_suspend(struct ssb_bus *bus, pm_message_t state)
-{
-	ssb_chipco_suspend(&bus->chipco, state);
-	ssb_pci_xtal(bus, SSB_GPIO_XTAL | SSB_GPIO_PLL, 0);
-
-	/* Reset HW state information in memory, so that HW is
-	 * completely reinitialized on resume. */
-	bus->mapped_device = NULL;
-#ifdef CONFIG_SSB_DRIVER_PCICORE
-	bus->pcicore.setup_done = 0;
-#endif
-#ifdef CONFIG_SSB_DEBUG
-	bus->powered_up = 0;
-#endif
-}
-
 static int ssb_device_suspend(struct device *dev, pm_message_t state)
 {
 	struct ssb_device *ssb_dev = dev_to_ssb_dev(dev);
 	struct ssb_driver *ssb_drv;
-	struct ssb_bus *bus;
 	int err = 0;
 
 	if (dev->driver) {
@@ -190,17 +150,44 @@ static int ssb_device_suspend(struct device *dev, pm_message_t state)
 		if (err)
 			goto out;
 	}
+out:
+	return err;
+}
+
+int ssb_bus_resume(struct ssb_bus *bus)
+{
+	int err;
 
-	bus = ssb_dev->bus;
-	bus->suspend_cnt++;
-	if (bus->suspend_cnt == bus->nr_devices) {
-		/* All devices suspended. Shutdown the bus. */
-		ssb_bus_suspend(bus, state);
+	/* Reset HW state information in memory, so that HW is
+	 * completely reinitialized. */
+	bus->mapped_device = NULL;
+#ifdef CONFIG_SSB_DRIVER_PCICORE
+	bus->pcicore.setup_done = 0;
+#endif
+
+	err = ssb_bus_powerup(bus, 0);
+	if (err)
+		return err;
+	err = ssb_pcmcia_hardware_setup(bus);
+	if (err) {
+		ssb_bus_may_powerdown(bus);
+		return err;
 	}
+	ssb_chipco_resume(&bus->chipco);
+	ssb_bus_may_powerdown(bus);
 
-out:
-	return err;
+	return 0;
+}
+EXPORT_SYMBOL(ssb_bus_resume);
+
+int ssb_bus_suspend(struct ssb_bus *bus)
+{
+	ssb_chipco_suspend(&bus->chipco);
+	ssb_pci_xtal(bus, SSB_GPIO_XTAL | SSB_GPIO_PLL, 0);
+
+	return 0;
 }
+EXPORT_SYMBOL(ssb_bus_suspend);
 
 #ifdef CONFIG_SSB_PCIHOST
 int ssb_devices_freeze(struct ssb_bus *bus)
diff --git a/drivers/ssb/pcihost_wrapper.c b/drivers/ssb/pcihost_wrapper.c
index 82a10abef640..e82db4aaa050 100644
--- a/drivers/ssb/pcihost_wrapper.c
+++ b/drivers/ssb/pcihost_wrapper.c
@@ -18,6 +18,12 @@
 #ifdef CONFIG_PM
 static int ssb_pcihost_suspend(struct pci_dev *dev, pm_message_t state)
 {
+	struct ssb_bus *ssb = pci_get_drvdata(dev);
+	int err;
+
+	err = ssb_bus_suspend(ssb);
+	if (err)
+		return err;
 	pci_save_state(dev);
 	pci_disable_device(dev);
 	pci_set_power_state(dev, pci_choose_state(dev, state));
@@ -27,6 +33,7 @@ static int ssb_pcihost_suspend(struct pci_dev *dev, pm_message_t state)
 
 static int ssb_pcihost_resume(struct pci_dev *dev)
 {
+	struct ssb_bus *ssb = pci_get_drvdata(dev);
 	int err;
 
 	pci_set_power_state(dev, 0);
@@ -34,6 +41,9 @@ static int ssb_pcihost_resume(struct pci_dev *dev)
 	if (err)
 		return err;
 	pci_restore_state(dev);
+	err = ssb_bus_resume(ssb);
+	if (err)
+		return err;
 
 	return 0;
 }
diff --git a/drivers/ssb/pcmcia.c b/drivers/ssb/pcmcia.c
index d674cef7210d..dcaf2412bea7 100644
--- a/drivers/ssb/pcmcia.c
+++ b/drivers/ssb/pcmcia.c
@@ -684,6 +684,29 @@ static int ssb_pcmcia_cor_setup(struct ssb_bus *bus, u8 cor)
 	return 0;
 }
 
+/* Initialize the PCMCIA hardware. This is called on Init and Resume. */
+int ssb_pcmcia_hardware_setup(struct ssb_bus *bus)
+{
+	int err;
+
+	if (bus->bustype != SSB_BUSTYPE_PCMCIA)
+		return 0;
+
+	/* Switch segment to a known state and sync
+	 * bus->mapped_pcmcia_seg with hardware state. */
+	ssb_pcmcia_switch_segment(bus, 0);
+	/* Init the COR register. */
+	err = ssb_pcmcia_cor_setup(bus, CISREG_COR);
+	if (err)
+		return err;
+	/* Some cards also need this register to get poked. */
+	err = ssb_pcmcia_cor_setup(bus, CISREG_COR + 0x80);
+	if (err)
+		return err;
+
+	return 0;
+}
+
 void ssb_pcmcia_exit(struct ssb_bus *bus)
 {
 	if (bus->bustype != SSB_BUSTYPE_PCMCIA)
@@ -699,16 +722,7 @@ int ssb_pcmcia_init(struct ssb_bus *bus)
 	if (bus->bustype != SSB_BUSTYPE_PCMCIA)
 		return 0;
 
-	/* Switch segment to a known state and sync
-	 * bus->mapped_pcmcia_seg with hardware state. */
-	ssb_pcmcia_switch_segment(bus, 0);
-
-	/* Init the COR register. */
-	err = ssb_pcmcia_cor_setup(bus, CISREG_COR);
-	if (err)
-		goto error;
-	/* Some cards also need this register to get poked. */
-	err = ssb_pcmcia_cor_setup(bus, CISREG_COR + 0x80);
+	err = ssb_pcmcia_hardware_setup(bus);
 	if (err)
 		goto error;
 
diff --git a/drivers/ssb/ssb_private.h b/drivers/ssb/ssb_private.h
index a83bf7a4d80b..ebc32d8fe15f 100644
--- a/drivers/ssb/ssb_private.h
+++ b/drivers/ssb/ssb_private.h
@@ -81,6 +81,7 @@ extern int ssb_pcmcia_switch_segment(struct ssb_bus *bus,
 				     u8 seg);
 extern int ssb_pcmcia_get_invariants(struct ssb_bus *bus,
 				     struct ssb_init_invariants *iv);
+extern int ssb_pcmcia_hardware_setup(struct ssb_bus *bus);
 extern void ssb_pcmcia_exit(struct ssb_bus *bus);
 extern int ssb_pcmcia_init(struct ssb_bus *bus);
 extern const struct ssb_bus_ops ssb_pcmcia_ops;
@@ -100,6 +101,10 @@ static inline int ssb_pcmcia_switch_segment(struct ssb_bus *bus,
 {
 	return 0;
 }
+static inline int ssb_pcmcia_hardware_setup(struct ssb_bus *bus)
+{
+	return 0;
+}
 static inline void ssb_pcmcia_exit(struct ssb_bus *bus)
 {
 }
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index 8644e03cf588..a8ca396f810a 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -260,9 +260,6 @@ struct ssb_bus {
 	struct ssb_device devices[SSB_MAX_NR_CORES];
 	u8 nr_devices;
 
-	/* Reference count. Number of suspended devices. */
-	u8 suspend_cnt;
-
 	/* Software ID number for this bus. */
 	unsigned int busnumber;
 
@@ -334,6 +331,13 @@ extern int ssb_bus_pcmciabus_register(struct ssb_bus *bus,
 
 extern void ssb_bus_unregister(struct ssb_bus *bus);
 
+/* Suspend a SSB bus.
+ * Call this from the parent bus suspend routine. */
+extern int ssb_bus_suspend(struct ssb_bus *bus);
+/* Resume a SSB bus.
+ * Call this from the parent bus resume routine. */
+extern int ssb_bus_resume(struct ssb_bus *bus);
+
 extern u32 ssb_clockspeed(struct ssb_bus *bus);
 
 /* Is the device enabled in hardware? */
diff --git a/include/linux/ssb/ssb_driver_chipcommon.h b/include/linux/ssb/ssb_driver_chipcommon.h
index b548a54ff1f5..7d7e03dcf77c 100644
--- a/include/linux/ssb/ssb_driver_chipcommon.h
+++ b/include/linux/ssb/ssb_driver_chipcommon.h
@@ -367,8 +367,7 @@ static inline bool ssb_chipco_available(struct ssb_chipcommon *cc)
 
 extern void ssb_chipcommon_init(struct ssb_chipcommon *cc);
 
-#include <linux/pm.h>
-extern void ssb_chipco_suspend(struct ssb_chipcommon *cc, pm_message_t state);
+extern void ssb_chipco_suspend(struct ssb_chipcommon *cc);
 extern void ssb_chipco_resume(struct ssb_chipcommon *cc);
 
 extern void ssb_chipco_get_clockcpu(struct ssb_chipcommon *cc,
-- 
cgit v1.2.3


From d625a29ba649a4df6027520ffc378f23c0e6883e Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Wed, 2 Apr 2008 19:46:56 +0200
Subject: ssb: Add support for block-I/O

This adds support for block based I/O to SSB.
This is needed in order to efficiently support PIO data
transfers to the card.
The block-I/O support is only compiled, if it's selected by the
weird driver that needs it. So there's no overhead for sane devices.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/ssb/Kconfig     |   5 ++
 drivers/ssb/main.c      | 102 +++++++++++++++++++++++++++++++++++++++++
 drivers/ssb/pci.c       |  70 ++++++++++++++++++++++++++++
 drivers/ssb/pcmcia.c    | 119 ++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/ssb/ssb.h |  19 ++++++++
 5 files changed, 315 insertions(+)

(limited to 'include/linux')

diff --git a/drivers/ssb/Kconfig b/drivers/ssb/Kconfig
index 0f7cce2560d1..cd845b8acd17 100644
--- a/drivers/ssb/Kconfig
+++ b/drivers/ssb/Kconfig
@@ -24,6 +24,11 @@ config SSB
 config SSB_SPROM
 	bool
 
+# Support for Block-I/O. SELECT this from the driver that needs it.
+config SSB_BLOCKIO
+	bool
+	depends on SSB
+
 config SSB_PCIHOST_POSSIBLE
 	bool
 	depends on SSB && (PCI = y || PCI = SSB)
diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index 3e58db7ef608..19ddd2bd1057 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -555,6 +555,55 @@ static u32 ssb_ssb_read32(struct ssb_device *dev, u16 offset)
 	return readl(bus->mmio + offset);
 }
 
+#ifdef CONFIG_SSB_BLOCKIO
+static void ssb_ssb_block_read(struct ssb_device *dev, void *buffer,
+			       size_t count, u16 offset, u8 reg_width)
+{
+	struct ssb_bus *bus = dev->bus;
+	void __iomem *addr;
+
+	offset += dev->core_index * SSB_CORE_SIZE;
+	addr = bus->mmio + offset;
+
+	switch (reg_width) {
+	case sizeof(u8): {
+		u8 *buf = buffer;
+
+		while (count) {
+			*buf = __raw_readb(addr);
+			buf++;
+			count--;
+		}
+		break;
+	}
+	case sizeof(u16): {
+		__le16 *buf = buffer;
+
+		SSB_WARN_ON(count & 1);
+		while (count) {
+			*buf = (__force __le16)__raw_readw(addr);
+			buf++;
+			count -= 2;
+		}
+		break;
+	}
+	case sizeof(u32): {
+		__le32 *buf = buffer;
+
+		SSB_WARN_ON(count & 3);
+		while (count) {
+			*buf = (__force __le32)__raw_readl(addr);
+			buf++;
+			count -= 4;
+		}
+		break;
+	}
+	default:
+		SSB_WARN_ON(1);
+	}
+}
+#endif /* CONFIG_SSB_BLOCKIO */
+
 static void ssb_ssb_write8(struct ssb_device *dev, u16 offset, u8 value)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -579,6 +628,55 @@ static void ssb_ssb_write32(struct ssb_device *dev, u16 offset, u32 value)
 	writel(value, bus->mmio + offset);
 }
 
+#ifdef CONFIG_SSB_BLOCKIO
+static void ssb_ssb_block_write(struct ssb_device *dev, const void *buffer,
+				size_t count, u16 offset, u8 reg_width)
+{
+	struct ssb_bus *bus = dev->bus;
+	void __iomem *addr;
+
+	offset += dev->core_index * SSB_CORE_SIZE;
+	addr = bus->mmio + offset;
+
+	switch (reg_width) {
+	case sizeof(u8): {
+		const u8 *buf = buffer;
+
+		while (count) {
+			__raw_writeb(*buf, addr);
+			buf++;
+			count--;
+		}
+		break;
+	}
+	case sizeof(u16): {
+		const __le16 *buf = buffer;
+
+		SSB_WARN_ON(count & 1);
+		while (count) {
+			__raw_writew((__force u16)(*buf), addr);
+			buf++;
+			count -= 2;
+		}
+		break;
+	}
+	case sizeof(u32): {
+		const __le32 *buf = buffer;
+
+		SSB_WARN_ON(count & 3);
+		while (count) {
+			__raw_writel((__force u32)(*buf), addr);
+			buf++;
+			count -= 4;
+		}
+		break;
+	}
+	default:
+		SSB_WARN_ON(1);
+	}
+}
+#endif /* CONFIG_SSB_BLOCKIO */
+
 /* Ops for the plain SSB bus without a host-device (no PCI or PCMCIA). */
 static const struct ssb_bus_ops ssb_ssb_ops = {
 	.read8		= ssb_ssb_read8,
@@ -587,6 +685,10 @@ static const struct ssb_bus_ops ssb_ssb_ops = {
 	.write8		= ssb_ssb_write8,
 	.write16	= ssb_ssb_write16,
 	.write32	= ssb_ssb_write32,
+#ifdef CONFIG_SSB_BLOCKIO
+	.block_read	= ssb_ssb_block_read,
+	.block_write	= ssb_ssb_block_write,
+#endif
 };
 
 static int ssb_fetch_invariants(struct ssb_bus *bus,
diff --git a/drivers/ssb/pci.c b/drivers/ssb/pci.c
index f1514b33cfae..904b1a8d0885 100644
--- a/drivers/ssb/pci.c
+++ b/drivers/ssb/pci.c
@@ -613,6 +613,41 @@ static u32 ssb_pci_read32(struct ssb_device *dev, u16 offset)
 	return ioread32(bus->mmio + offset);
 }
 
+#ifdef CONFIG_SSB_BLOCKIO
+static void ssb_pci_block_read(struct ssb_device *dev, void *buffer,
+			       size_t count, u16 offset, u8 reg_width)
+{
+	struct ssb_bus *bus = dev->bus;
+	void __iomem *addr = bus->mmio + offset;
+
+	if (unlikely(ssb_pci_assert_buspower(bus)))
+		goto error;
+	if (unlikely(bus->mapped_device != dev)) {
+		if (unlikely(ssb_pci_switch_core(bus, dev)))
+			goto error;
+	}
+	switch (reg_width) {
+	case sizeof(u8):
+		ioread8_rep(addr, buffer, count);
+		break;
+	case sizeof(u16):
+		SSB_WARN_ON(count & 1);
+		ioread16_rep(addr, buffer, count >> 1);
+		break;
+	case sizeof(u32):
+		SSB_WARN_ON(count & 3);
+		ioread32_rep(addr, buffer, count >> 2);
+		break;
+	default:
+		SSB_WARN_ON(1);
+	}
+
+	return;
+error:
+	memset(buffer, 0xFF, count);
+}
+#endif /* CONFIG_SSB_BLOCKIO */
+
 static void ssb_pci_write8(struct ssb_device *dev, u16 offset, u8 value)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -652,6 +687,37 @@ static void ssb_pci_write32(struct ssb_device *dev, u16 offset, u32 value)
 	iowrite32(value, bus->mmio + offset);
 }
 
+#ifdef CONFIG_SSB_BLOCKIO
+static void ssb_pci_block_write(struct ssb_device *dev, const void *buffer,
+				size_t count, u16 offset, u8 reg_width)
+{
+	struct ssb_bus *bus = dev->bus;
+	void __iomem *addr = bus->mmio + offset;
+
+	if (unlikely(ssb_pci_assert_buspower(bus)))
+		return;
+	if (unlikely(bus->mapped_device != dev)) {
+		if (unlikely(ssb_pci_switch_core(bus, dev)))
+			return;
+	}
+	switch (reg_width) {
+	case sizeof(u8):
+		iowrite8_rep(addr, buffer, count);
+		break;
+	case sizeof(u16):
+		SSB_WARN_ON(count & 1);
+		iowrite16_rep(addr, buffer, count >> 1);
+		break;
+	case sizeof(u32):
+		SSB_WARN_ON(count & 3);
+		iowrite32_rep(addr, buffer, count >> 2);
+		break;
+	default:
+		SSB_WARN_ON(1);
+	}
+}
+#endif /* CONFIG_SSB_BLOCKIO */
+
 /* Not "static", as it's used in main.c */
 const struct ssb_bus_ops ssb_pci_ops = {
 	.read8		= ssb_pci_read8,
@@ -660,6 +726,10 @@ const struct ssb_bus_ops ssb_pci_ops = {
 	.write8		= ssb_pci_write8,
 	.write16	= ssb_pci_write16,
 	.write32	= ssb_pci_write32,
+#ifdef CONFIG_SSB_BLOCKIO
+	.block_read	= ssb_pci_block_read,
+	.block_write	= ssb_pci_block_write,
+#endif
 };
 
 static ssize_t ssb_pci_attr_sprom_show(struct device *pcidev,
diff --git a/drivers/ssb/pcmcia.c b/drivers/ssb/pcmcia.c
index dcaf2412bea7..24c2a46c1476 100644
--- a/drivers/ssb/pcmcia.c
+++ b/drivers/ssb/pcmcia.c
@@ -285,6 +285,64 @@ static u32 ssb_pcmcia_read32(struct ssb_device *dev, u16 offset)
 	return (lo | (hi << 16));
 }
 
+#ifdef CONFIG_SSB_BLOCKIO
+static void ssb_pcmcia_block_read(struct ssb_device *dev, void *buffer,
+				  size_t count, u16 offset, u8 reg_width)
+{
+	struct ssb_bus *bus = dev->bus;
+	unsigned long flags;
+	void __iomem *addr = bus->mmio + offset;
+	int err;
+
+	spin_lock_irqsave(&bus->bar_lock, flags);
+	err = select_core_and_segment(dev, &offset);
+	if (unlikely(err)) {
+		memset(buffer, 0xFF, count);
+		goto unlock;
+	}
+	switch (reg_width) {
+	case sizeof(u8): {
+		u8 *buf = buffer;
+
+		while (count) {
+			*buf = __raw_readb(addr);
+			buf++;
+			count--;
+		}
+		break;
+	}
+	case sizeof(u16): {
+		__le16 *buf = buffer;
+
+		SSB_WARN_ON(count & 1);
+		while (count) {
+			*buf = (__force __le16)__raw_readw(addr);
+			buf++;
+			count -= 2;
+		}
+		break;
+	}
+	case sizeof(u32): {
+		__le16 *buf = buffer;
+
+		SSB_WARN_ON(count & 3);
+		while (count) {
+			*buf = (__force __le16)__raw_readw(addr);
+			buf++;
+			*buf = (__force __le16)__raw_readw(addr + 2);
+			buf++;
+			count -= 4;
+		}
+		break;
+	}
+	default:
+		SSB_WARN_ON(1);
+	}
+unlock:
+	spin_unlock_irqrestore(&bus->bar_lock, flags);
+}
+#endif /* CONFIG_SSB_BLOCKIO */
+
 static void ssb_pcmcia_write8(struct ssb_device *dev, u16 offset, u8 value)
 {
 	struct ssb_bus *bus = dev->bus;
@@ -329,6 +387,63 @@ static void ssb_pcmcia_write32(struct ssb_device *dev, u16 offset, u32 value)
 	spin_unlock_irqrestore(&bus->bar_lock, flags);
 }
 
+#ifdef CONFIG_SSB_BLOCKIO
+static void ssb_pcmcia_block_write(struct ssb_device *dev, const void *buffer,
+				   size_t count, u16 offset, u8 reg_width)
+{
+	struct ssb_bus *bus = dev->bus;
+	unsigned long flags;
+	void __iomem *addr = bus->mmio + offset;
+	int err;
+
+	spin_lock_irqsave(&bus->bar_lock, flags);
+	err = select_core_and_segment(dev, &offset);
+	if (unlikely(err))
+		goto unlock;
+	switch (reg_width) {
+	case sizeof(u8): {
+		const u8 *buf = buffer;
+
+		while (count) {
+			__raw_writeb(*buf, addr);
+			buf++;
+			count--;
+		}
+		break;
+	}
+	case sizeof(u16): {
+		const __le16 *buf = buffer;
+
+		SSB_WARN_ON(count & 1);
+		while (count) {
+			__raw_writew((__force u16)(*buf), addr);
+			buf++;
+			count -= 2;
+		}
+		break;
+	}
+	case sizeof(u32): {
+		const __le16 *buf = buffer;
+
+		SSB_WARN_ON(count & 3);
+		while (count) {
+			__raw_writew((__force u16)(*buf), addr);
+			buf++;
+			__raw_writew((__force u16)(*buf), addr + 2);
+			buf++;
+			count -= 4;
+		}
+		break;
+	}
+	default:
+		SSB_WARN_ON(1);
+	}
+unlock:
+	mmiowb();
+	spin_unlock_irqrestore(&bus->bar_lock, flags);
+}
+#endif /* CONFIG_SSB_BLOCKIO */
+
 /* Not "static", as it's used in main.c */
 const struct ssb_bus_ops ssb_pcmcia_ops = {
 	.read8		= ssb_pcmcia_read8,
@@ -337,6 +452,10 @@ const struct ssb_bus_ops ssb_pcmcia_ops = {
 	.write8		= ssb_pcmcia_write8,
 	.write16	= ssb_pcmcia_write16,
 	.write32	= ssb_pcmcia_write32,
+#ifdef CONFIG_SSB_BLOCKIO
+	.block_read	= ssb_pcmcia_block_read,
+	.block_write	= ssb_pcmcia_block_write,
+#endif
 };
 
 static int ssb_pcmcia_sprom_command(struct ssb_bus *bus, u8 command)
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index a8ca396f810a..9f95afd0e9e3 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -78,6 +78,12 @@ struct ssb_bus_ops {
 	void (*write8)(struct ssb_device *dev, u16 offset, u8 value);
 	void (*write16)(struct ssb_device *dev, u16 offset, u16 value);
 	void (*write32)(struct ssb_device *dev, u16 offset, u32 value);
+#ifdef CONFIG_SSB_BLOCKIO
+	void (*block_read)(struct ssb_device *dev, void *buffer,
+			   size_t count, u16 offset, u8 reg_width);
+	void (*block_write)(struct ssb_device *dev, const void *buffer,
+			    size_t count, u16 offset, u8 reg_width);
+#endif
 };
 
 
@@ -374,6 +380,19 @@ static inline void ssb_write32(struct ssb_device *dev, u16 offset, u32 value)
 {
 	dev->ops->write32(dev, offset, value);
 }
+#ifdef CONFIG_SSB_BLOCKIO
+static inline void ssb_block_read(struct ssb_device *dev, void *buffer,
+				  size_t count, u16 offset, u8 reg_width)
+{
+	dev->ops->block_read(dev, buffer, count, offset, reg_width);
+}
+
+static inline void ssb_block_write(struct ssb_device *dev, const void *buffer,
+				   size_t count, u16 offset, u8 reg_width)
+{
+	dev->ops->block_write(dev, buffer, count, offset, reg_width);
+}
+#endif /* CONFIG_SSB_BLOCKIO */
 
 
 /* Translation (routing) bits that need to be ORed to DMA
-- 
cgit v1.2.3


From b715631fad3ed320b85d386a84a6fb0b3f86b0b9 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Thu, 10 Apr 2008 01:33:47 -0700
Subject: socket: sk_filter minor cleanups

Some minor style cleanups:
  * Move __KERNEL__ definitions to one place in filter.h
  * Use const for sk_filter_len
  * Line wrapping
  * Put EXPORT_SYMBOL next to function definition

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/filter.h | 31 +++++++++++++++----------------
 net/core/filter.c      |  5 ++---
 2 files changed, 17 insertions(+), 19 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/filter.h b/include/linux/filter.h
index ddfa0372a3b7..bfc5d319b946 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -37,21 +37,6 @@ struct sock_fprog	/* Required for SO_ATTACH_FILTER. */
 	struct sock_filter __user *filter;
 };
 
-#ifdef __KERNEL__
-struct sk_filter
-{
-	atomic_t		refcnt;
-	unsigned int         	len;	/* Number of filter blocks */
-	struct rcu_head		rcu;
-	struct sock_filter     	insns[0];
-};
-
-static inline unsigned int sk_filter_len(struct sk_filter *fp)
-{
-	return fp->len*sizeof(struct sock_filter) + sizeof(*fp);
-}
-#endif
-
 /*
  * Instruction classes
  */
@@ -141,10 +126,24 @@ static inline unsigned int sk_filter_len(struct sk_filter *fp)
 #define SKF_LL_OFF    (-0x200000)
 
 #ifdef __KERNEL__
+struct sk_filter
+{
+	atomic_t		refcnt;
+	unsigned int         	len;	/* Number of filter blocks */
+	struct rcu_head		rcu;
+	struct sock_filter     	insns[0];
+};
+
+static inline unsigned int sk_filter_len(const struct sk_filter *fp)
+{
+	return fp->len * sizeof(struct sock_filter) + sizeof(*fp);
+}
+
 struct sk_buff;
 struct sock;
 
-extern unsigned int sk_run_filter(struct sk_buff *skb, struct sock_filter *filter, int flen);
+extern unsigned int sk_run_filter(struct sk_buff *skb,
+				  struct sock_filter *filter, int flen);
 extern int sk_attach_filter(struct sock_fprog *fprog, struct sock *sk);
 extern int sk_detach_filter(struct sock *sk);
 extern int sk_chk_filter(struct sock_filter *filter, int flen);
diff --git a/net/core/filter.c b/net/core/filter.c
index e0a06942c025..85a5febab567 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -275,6 +275,7 @@ load_b:
 
 	return 0;
 }
+EXPORT_SYMBOL(sk_run_filter);
 
 /**
  *	sk_chk_filter - verify socket filter code
@@ -385,6 +386,7 @@ int sk_chk_filter(struct sock_filter *filter, int flen)
 
 	return (BPF_CLASS(filter[flen - 1].code) == BPF_RET) ? 0 : -EINVAL;
 }
+EXPORT_SYMBOL(sk_chk_filter);
 
 /**
  * 	sk_filter_rcu_release: Release a socket filter by rcu_head
@@ -467,6 +469,3 @@ int sk_detach_filter(struct sock *sk)
 	rcu_read_unlock_bh();
 	return ret;
 }
-
-EXPORT_SYMBOL(sk_chk_filter);
-EXPORT_SYMBOL(sk_run_filter);
-- 
cgit v1.2.3


From 43db6d65e0ef943a361cb91f8baa49132009227b Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Thu, 10 Apr 2008 01:43:09 -0700
Subject: socket: sk_filter deinline

The sk_filter function is too big to be inlined. This saves 2296 bytes
of text on allyesconfig.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/filter.h |  1 +
 include/net/sock.h     | 35 -----------------------------------
 net/core/filter.c      | 35 +++++++++++++++++++++++++++++++++++
 3 files changed, 36 insertions(+), 35 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/filter.h b/include/linux/filter.h
index bfc5d319b946..673e5677ebcc 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -142,6 +142,7 @@ static inline unsigned int sk_filter_len(const struct sk_filter *fp)
 struct sk_buff;
 struct sock;
 
+extern int sk_filter(struct sock *sk, struct sk_buff *skb);
 extern unsigned int sk_run_filter(struct sk_buff *skb,
 				  struct sock_filter *filter, int flen);
 extern int sk_attach_filter(struct sock_fprog *fprog, struct sock *sk);
diff --git a/include/net/sock.h b/include/net/sock.h
index f4fdd101c9a2..09255eae93e9 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -927,41 +927,6 @@ extern void sk_common_release(struct sock *sk);
 /* Initialise core socket variables */
 extern void sock_init_data(struct socket *sock, struct sock *sk);
 
-/**
- *	sk_filter - run a packet through a socket filter
- *	@sk: sock associated with &sk_buff
- *	@skb: buffer to filter
- *	@needlock: set to 1 if the sock is not locked by caller.
- *
- * Run the filter code and then cut skb->data to correct size returned by
- * sk_run_filter. If pkt_len is 0 we toss packet. If skb->len is smaller
- * than pkt_len we keep whole skb->data. This is the socket level
- * wrapper to sk_run_filter. It returns 0 if the packet should
- * be accepted or -EPERM if the packet should be tossed.
- *
- */
-
-static inline int sk_filter(struct sock *sk, struct sk_buff *skb)
-{
-	int err;
-	struct sk_filter *filter;
-	
-	err = security_sock_rcv_skb(sk, skb);
-	if (err)
-		return err;
-	
-	rcu_read_lock_bh();
-	filter = rcu_dereference(sk->sk_filter);
-	if (filter) {
-		unsigned int pkt_len = sk_run_filter(skb, filter->insns,
-				filter->len);
-		err = pkt_len ? pskb_trim(skb, pkt_len) : -EPERM;
-	}
- 	rcu_read_unlock_bh();
-
-	return err;
-}
-
 /**
  *	sk_filter_release: Release a socket filter
  *	@sk: socket
diff --git a/net/core/filter.c b/net/core/filter.c
index 85a5febab567..bbb53c69857c 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -63,6 +63,41 @@ static inline void *load_pointer(struct sk_buff *skb, int k,
 	}
 }
 
+/**
+ *	sk_filter - run a packet through a socket filter
+ *	@sk: sock associated with &sk_buff
+ *	@skb: buffer to filter
+ *	@needlock: set to 1 if the sock is not locked by caller.
+ *
+ * Run the filter code and then cut skb->data to correct size returned by
+ * sk_run_filter. If pkt_len is 0 we toss packet. If skb->len is smaller
+ * than pkt_len we keep whole skb->data. This is the socket level
+ * wrapper to sk_run_filter. It returns 0 if the packet should
+ * be accepted or -EPERM if the packet should be tossed.
+ *
+ */
+int sk_filter(struct sock *sk, struct sk_buff *skb)
+{
+	int err;
+	struct sk_filter *filter;
+
+	err = security_sock_rcv_skb(sk, skb);
+	if (err)
+		return err;
+
+	rcu_read_lock_bh();
+	filter = rcu_dereference(sk->sk_filter);
+	if (filter) {
+		unsigned int pkt_len = sk_run_filter(skb, filter->insns,
+				filter->len);
+		err = pkt_len ? pskb_trim(skb, pkt_len) : -EPERM;
+	}
+	rcu_read_unlock_bh();
+
+	return err;
+}
+EXPORT_SYMBOL(sk_filter);
+
 /**
  *	sk_run_filter - run a filter on a socket
  *	@skb: buffer to run the filter on
-- 
cgit v1.2.3


From 4738c1db1593687713869fa69e733eebc7b0d6d8 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 10 Apr 2008 02:02:28 -0700
Subject: [SKFILTER]: Add SKF_ADF_NLATTR instruction

SKF_ADF_NLATTR searches for a netlink attribute, which avoids manually
parsing and walking attributes. It takes the offset at which to start
searching in the 'A' register and the attribute type in the 'X' register
and returns the offset in the 'A' register. When the attribute is not
found it returns zero.

A top-level attribute can be located using a filter like this
(example for nfnetlink, using struct nfgenmsg):

	...
	{
		/* A = offset of first attribute */
		.code	= BPF_LD | BPF_IMM,
		.k	= sizeof(struct nlmsghdr) + sizeof(struct nfgenmsg)
	},
	{
		/* X = CTA_PROTOINFO */
		.code	= BPF_LDX | BPF_IMM,
		.k	= CTA_PROTOINFO,
	},
	{
		/* A = netlink attribute offset */
		.code	= BPF_LD | BPF_B | BPF_ABS,
		.k	= SKF_AD_OFF + SKF_AD_NLATTR
	},
	{
		/* Exit if not found */
		.code   = BPF_JMP | BPF_JEQ | BPF_K,
		.k	= 0,
		.jt	= <error>
	},
	...

A nested attribute below the CTA_PROTOINFO attribute would then
be parsed like this:

	...
	{
		/* A += sizeof(struct nlattr) */
		.code	= BPF_ALU | BPF_ADD | BPF_K,
		.k	= sizeof(struct nlattr),
	},
	{
		/* X = CTA_PROTOINFO_TCP */
		.code	= BPF_LDX | BPF_IMM,
		.k	= CTA_PROTOINFO_TCP,
	},
	{
		/* A = netlink attribute offset */
		.code	= BPF_LD | BPF_B | BPF_ABS,
		.k	= SKF_AD_OFF + SKF_AD_NLATTR
	},
	...

The data of an attribute can be loaded into 'A' like this:

	...
	{
		/* X = A (attribute offset) */
		.code	= BPF_MISC | BPF_TAX,
	},
	{
		/* A = skb->data[X + k] */
		.code 	= BPF_LD | BPF_B | BPF_IND,
		.k	= sizeof(struct nlattr),
	},
	...

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/filter.h |  3 ++-
 net/core/filter.c      | 17 +++++++++++++++++
 2 files changed, 19 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/filter.h b/include/linux/filter.h
index 673e5677ebcc..b6ea9aa9e853 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -121,7 +121,8 @@ struct sock_fprog	/* Required for SO_ATTACH_FILTER. */
 #define SKF_AD_PROTOCOL 0
 #define SKF_AD_PKTTYPE 	4
 #define SKF_AD_IFINDEX 	8
-#define SKF_AD_MAX 	12
+#define SKF_AD_NLATTR	12
+#define SKF_AD_MAX 	16
 #define SKF_NET_OFF   (-0x100000)
 #define SKF_LL_OFF    (-0x200000)
 
diff --git a/net/core/filter.c b/net/core/filter.c
index bbb53c69857c..f5f3cf603064 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -27,6 +27,7 @@
 #include <linux/if_packet.h>
 #include <net/ip.h>
 #include <net/protocol.h>
+#include <net/netlink.h>
 #include <linux/skbuff.h>
 #include <net/sock.h>
 #include <linux/errno.h>
@@ -303,6 +304,22 @@ load_b:
 		case SKF_AD_IFINDEX:
 			A = skb->dev->ifindex;
 			continue;
+		case SKF_AD_NLATTR: {
+			struct nlattr *nla;
+
+			if (skb_is_nonlinear(skb))
+				return 0;
+			if (A > skb->len - sizeof(struct nlattr))
+				return 0;
+
+			nla = nla_find((struct nlattr *)&skb->data[A],
+				       skb->len - A, X);
+			if (nla)
+				A = (void *)nla - (void *)skb->data;
+			else
+				A = 0;
+			continue;
+		}
 		default:
 			return 0;
 		}
-- 
cgit v1.2.3


From 54a015104136974262afa4b8ddd943ea70dec8a2 Mon Sep 17 00:00:00 2001
From: Roland McGrath <roland@redhat.com>
Date: Thu, 10 Apr 2008 15:37:38 -0700
Subject: asmlinkage_protect replaces prevent_tail_call

The prevent_tail_call() macro works around the problem of the compiler
clobbering argument words on the stack, which for asmlinkage functions
is the caller's (user's) struct pt_regs.  The tail/sibling-call
optimization is not the only way that the compiler can decide to use
stack argument words as scratch space, which we have to prevent.
Other optimizations can do it too.

Until we have new compiler support to make "asmlinkage" binding on the
compiler's own use of the stack argument frame, we have work around all
the manifestations of this issue that crop up.

More cases seem to be prevented by also keeping the incoming argument
variables live at the end of the function.  This makes their original
stack slots attractive places to leave those variables, so the compiler
tends not clobber them for something else.  It's still no guarantee, but
it handles some observed cases that prevent_tail_call() did not.

Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/x86/kernel/tls.c     |  4 ++--
 fs/open.c                 |  8 ++++----
 include/asm-x86/linkage.h | 24 +++++++++++++++++++++++-
 include/linux/linkage.h   |  4 ++--
 kernel/exit.c             |  4 ++--
 kernel/uid16.c            | 22 +++++++++++-----------
 6 files changed, 44 insertions(+), 22 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/kernel/tls.c b/arch/x86/kernel/tls.c
index 022bcaa3b42e..ab6bf375a307 100644
--- a/arch/x86/kernel/tls.c
+++ b/arch/x86/kernel/tls.c
@@ -92,7 +92,7 @@ int do_set_thread_area(struct task_struct *p, int idx,
 asmlinkage int sys_set_thread_area(struct user_desc __user *u_info)
 {
 	int ret = do_set_thread_area(current, -1, u_info, 1);
-	prevent_tail_call(ret);
+	asmlinkage_protect(1, ret, u_info);
 	return ret;
 }
 
@@ -142,7 +142,7 @@ int do_get_thread_area(struct task_struct *p, int idx,
 asmlinkage int sys_get_thread_area(struct user_desc __user *u_info)
 {
 	int ret = do_get_thread_area(current, -1, u_info);
-	prevent_tail_call(ret);
+	asmlinkage_protect(1, ret, u_info);
 	return ret;
 }
 
diff --git a/fs/open.c b/fs/open.c
index a4b12022edaa..3fa4e4ffce4c 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -335,7 +335,7 @@ asmlinkage long sys_ftruncate(unsigned int fd, unsigned long length)
 {
 	long ret = do_sys_ftruncate(fd, length, 1);
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(2, ret, fd, length);
 	return ret;
 }
 
@@ -350,7 +350,7 @@ asmlinkage long sys_ftruncate64(unsigned int fd, loff_t length)
 {
 	long ret = do_sys_ftruncate(fd, length, 0);
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(2, ret, fd, length);
 	return ret;
 }
 #endif
@@ -1067,7 +1067,7 @@ asmlinkage long sys_open(const char __user *filename, int flags, int mode)
 
 	ret = do_sys_open(AT_FDCWD, filename, flags, mode);
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(3, ret, filename, flags, mode);
 	return ret;
 }
 
@@ -1081,7 +1081,7 @@ asmlinkage long sys_openat(int dfd, const char __user *filename, int flags,
 
 	ret = do_sys_open(dfd, filename, flags, mode);
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(4, ret, dfd, filename, flags, mode);
 	return ret;
 }
 
diff --git a/include/asm-x86/linkage.h b/include/asm-x86/linkage.h
index 31739c7d66a9..d605eeba0f70 100644
--- a/include/asm-x86/linkage.h
+++ b/include/asm-x86/linkage.h
@@ -8,12 +8,34 @@
 
 #ifdef CONFIG_X86_32
 #define asmlinkage CPP_ASMLINKAGE __attribute__((regparm(0)))
-#define prevent_tail_call(ret) __asm__ ("" : "=r" (ret) : "0" (ret))
 /*
  * For 32-bit UML - mark functions implemented in assembly that use
  * regparm input parameters:
  */
 #define asmregparm __attribute__((regparm(3)))
+
+#define asmlinkage_protect(n, ret, args...) \
+	__asmlinkage_protect##n(ret, ##args)
+#define __asmlinkage_protect_n(ret, args...) \
+	__asm__ __volatile__ ("" : "=r" (ret) : "0" (ret), ##args)
+#define __asmlinkage_protect0(ret) \
+	__asmlinkage_protect_n(ret)
+#define __asmlinkage_protect1(ret, arg1) \
+	__asmlinkage_protect_n(ret, "g" (arg1))
+#define __asmlinkage_protect2(ret, arg1, arg2) \
+	__asmlinkage_protect_n(ret, "g" (arg1), "g" (arg2))
+#define __asmlinkage_protect3(ret, arg1, arg2, arg3) \
+	__asmlinkage_protect_n(ret, "g" (arg1), "g" (arg2), "g" (arg3))
+#define __asmlinkage_protect4(ret, arg1, arg2, arg3, arg4) \
+	__asmlinkage_protect_n(ret, "g" (arg1), "g" (arg2), "g" (arg3), \
+			      "g" (arg4))
+#define __asmlinkage_protect5(ret, arg1, arg2, arg3, arg4, arg5) \
+	__asmlinkage_protect_n(ret, "g" (arg1), "g" (arg2), "g" (arg3), \
+			      "g" (arg4), "g" (arg5))
+#define __asmlinkage_protect6(ret, arg1, arg2, arg3, arg4, arg5, arg6) \
+	__asmlinkage_protect_n(ret, "g" (arg1), "g" (arg2), "g" (arg3), \
+			      "g" (arg4), "g" (arg5), "g" (arg6))
+
 #endif
 
 #ifdef CONFIG_X86_ALIGNMENT_16
diff --git a/include/linux/linkage.h b/include/linux/linkage.h
index 0592936344c4..fe2a39c489b6 100644
--- a/include/linux/linkage.h
+++ b/include/linux/linkage.h
@@ -17,8 +17,8 @@
 # define asmregparm
 #endif
 
-#ifndef prevent_tail_call
-# define prevent_tail_call(ret) do { } while (0)
+#ifndef asmlinkage_protect
+# define asmlinkage_protect(n, ret, args...)	do { } while (0)
 #endif
 
 #ifndef __ALIGN
diff --git a/kernel/exit.c b/kernel/exit.c
index 53872bf993fa..073005b1cfb2 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -1608,7 +1608,7 @@ asmlinkage long sys_waitid(int which, pid_t upid,
 	put_pid(pid);
 
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(5, ret, which, upid, infop, options, ru);
 	return ret;
 }
 
@@ -1640,7 +1640,7 @@ asmlinkage long sys_wait4(pid_t upid, int __user *stat_addr,
 	put_pid(pid);
 
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(4, ret, upid, stat_addr, options, ru);
 	return ret;
 }
 
diff --git a/kernel/uid16.c b/kernel/uid16.c
index dd308ba4e03b..3e41c1673e2f 100644
--- a/kernel/uid16.c
+++ b/kernel/uid16.c
@@ -21,7 +21,7 @@ asmlinkage long sys_chown16(const char __user * filename, old_uid_t user, old_gi
 {
 	long ret = sys_chown(filename, low2highuid(user), low2highgid(group));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(3, ret, filename, user, group);
 	return ret;
 }
 
@@ -29,7 +29,7 @@ asmlinkage long sys_lchown16(const char __user * filename, old_uid_t user, old_g
 {
 	long ret = sys_lchown(filename, low2highuid(user), low2highgid(group));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(3, ret, filename, user, group);
 	return ret;
 }
 
@@ -37,7 +37,7 @@ asmlinkage long sys_fchown16(unsigned int fd, old_uid_t user, old_gid_t group)
 {
 	long ret = sys_fchown(fd, low2highuid(user), low2highgid(group));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(3, ret, fd, user, group);
 	return ret;
 }
 
@@ -45,7 +45,7 @@ asmlinkage long sys_setregid16(old_gid_t rgid, old_gid_t egid)
 {
 	long ret = sys_setregid(low2highgid(rgid), low2highgid(egid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(2, ret, rgid, egid);
 	return ret;
 }
 
@@ -53,7 +53,7 @@ asmlinkage long sys_setgid16(old_gid_t gid)
 {
 	long ret = sys_setgid(low2highgid(gid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(1, ret, gid);
 	return ret;
 }
 
@@ -61,7 +61,7 @@ asmlinkage long sys_setreuid16(old_uid_t ruid, old_uid_t euid)
 {
 	long ret = sys_setreuid(low2highuid(ruid), low2highuid(euid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(2, ret, ruid, euid);
 	return ret;
 }
 
@@ -69,7 +69,7 @@ asmlinkage long sys_setuid16(old_uid_t uid)
 {
 	long ret = sys_setuid(low2highuid(uid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(1, ret, uid);
 	return ret;
 }
 
@@ -78,7 +78,7 @@ asmlinkage long sys_setresuid16(old_uid_t ruid, old_uid_t euid, old_uid_t suid)
 	long ret = sys_setresuid(low2highuid(ruid), low2highuid(euid),
 				 low2highuid(suid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(3, ret, ruid, euid, suid);
 	return ret;
 }
 
@@ -98,7 +98,7 @@ asmlinkage long sys_setresgid16(old_gid_t rgid, old_gid_t egid, old_gid_t sgid)
 	long ret = sys_setresgid(low2highgid(rgid), low2highgid(egid),
 				 low2highgid(sgid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(3, ret, rgid, egid, sgid);
 	return ret;
 }
 
@@ -117,7 +117,7 @@ asmlinkage long sys_setfsuid16(old_uid_t uid)
 {
 	long ret = sys_setfsuid(low2highuid(uid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(1, ret, uid);
 	return ret;
 }
 
@@ -125,7 +125,7 @@ asmlinkage long sys_setfsgid16(old_gid_t gid)
 {
 	long ret = sys_setfsgid(low2highgid(gid));
 	/* avoid REGPARM breakage on x86: */
-	prevent_tail_call(ret);
+	asmlinkage_protect(1, ret, gid);
 	return ret;
 }
 
-- 
cgit v1.2.3


From d10d89ec78114f925f63c5126a2b2490f501a462 Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Thu, 10 Apr 2008 17:35:23 -0700
Subject: Add commentary about the new "asmlinkage_protect()" macro

It's really a pretty ugly thing to need, and some day it will hopefully
be obviated by teaching gcc about the magic calling conventions for the
low-level system call code, but in the meantime we can at least add big
honking comments about why we need these insane and strange macros.

I took my comments from my version of the macro, but I ended up deciding
to just pick Roland's version of the actual code instead (with his
prettier syntax that uses vararg macros).  Thus the previous two commits
that actually implement it.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/asm-x86/linkage.h | 11 +++++++++++
 include/linux/linkage.h   | 13 +++++++++++++
 2 files changed, 24 insertions(+)

(limited to 'include/linux')

diff --git a/include/asm-x86/linkage.h b/include/asm-x86/linkage.h
index d605eeba0f70..c048353f4b85 100644
--- a/include/asm-x86/linkage.h
+++ b/include/asm-x86/linkage.h
@@ -14,6 +14,17 @@
  */
 #define asmregparm __attribute__((regparm(3)))
 
+/*
+ * Make sure the compiler doesn't do anything stupid with the
+ * arguments on the stack - they are owned by the *caller*, not
+ * the callee. This just fools gcc into not spilling into them,
+ * and keeps it from doing tailcall recursion and/or using the
+ * stack slots for temporaries, since they are live and "used"
+ * all the way to the end of the function.
+ *
+ * NOTE! On x86-64, all the arguments are in registers, so this
+ * only matters on a 32-bit kernel.
+ */
 #define asmlinkage_protect(n, ret, args...) \
 	__asmlinkage_protect##n(ret, ##args)
 #define __asmlinkage_protect_n(ret, args...) \
diff --git a/include/linux/linkage.h b/include/linux/linkage.h
index fe2a39c489b6..b163c5c40dbc 100644
--- a/include/linux/linkage.h
+++ b/include/linux/linkage.h
@@ -17,6 +17,19 @@
 # define asmregparm
 #endif
 
+/*
+ * This is used by architectures to keep arguments on the stack
+ * untouched by the compiler by keeping them live until the end.
+ * The argument stack may be owned by the assembly-language
+ * caller, not the callee, and gcc doesn't always understand
+ * that.
+ *
+ * We have the return value, and a maximum of six arguments.
+ *
+ * This should always be followed by a "return ret" for the
+ * protection to work (ie no more work that the compiler might
+ * end up needing stack temporaries for).
+ */
 #ifndef asmlinkage_protect
 # define asmlinkage_protect(n, ret, args...)	do { } while (0)
 #endif
-- 
cgit v1.2.3


From 544451a1a36b06f43fc67112ede3b92e6203eb9d Mon Sep 17 00:00:00 2001
From: Bjorn Helgaas <bjorn.helgaas@hp.com>
Date: Thu, 10 Apr 2008 21:29:28 -0700
Subject: pnp: increase number of devices supported per protocol

Increase the PNP "number of devices" limit.  We currently use an unsigned
char, which limits us to 256 devices per protocol.  This patch changes that to
an unsigned int.

Not all backends can take advantage of this: we limit ISAPNP to 10 devices in
isapnp_cfg_begin(), and PNPBIOS is limited to 256 devices because the BIOS
interfaces use a one-byte device node number.

But there is no limit on the number of PNPACPI devices we may have.  Large HP
Integrity machines have more than 256, which causes the current "unsigned char
number" to wrap around.  This causes errors like this:

    pnp: PnP ACPI init
    kobject_add failed for 00:00 with -EEXIST, don't try to register things with the same name in the same directory.

    Call Trace:
     [<a000000100010720>] show_stack+0x40/0xa0
     [<a0000001000107b0>] dump_stack+0x30/0x60
     [<a0000001001dbdf0>] kobject_add+0x290/0x2c0
     [<a0000001002bfd40>] device_add+0x160/0x860
     [<a0000001002c0470>] device_register+0x30/0x60
     [<a00000010026ba70>] __pnp_add_device+0x130/0x180
     [<a00000010026bb70>] pnp_add_device+0xb0/0xe0
     [<a0000001007f2730>] pnpacpi_add_device+0x510/0x5a0
     [<a0000001007f2810>] pnpacpi_add_device_handler+0x50/0x80

This patch increases the limit to fix this PNPACPI problem.  It should not
have any adverse effect on ISAPNP or PNPBIOS because their limits are still
enforced in the backends.

Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/pnp.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/pnp.h b/include/linux/pnp.h
index 29dd55838e84..b2f05c230f4b 100644
--- a/include/linux/pnp.h
+++ b/include/linux/pnp.h
@@ -175,7 +175,7 @@ static inline void pnp_set_card_drvdata(struct pnp_card_link *pcard, void *data)
 struct pnp_dev {
 	struct device dev;		/* Driver Model device interface */
 	u64 dma_mask;
-	unsigned char number;		/* used as an index, must be unique */
+	unsigned int number;		/* used as an index, must be unique */
 	int status;
 
 	struct list_head global_list;	/* node in global list of devices */
-- 
cgit v1.2.3


From b0fac02370cffad956ff3de5e8ed4df7e7b875d7 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Fri, 11 Apr 2008 13:46:54 +0200
Subject: Fix "$(AS) -traditional" compile breakage caused by
 asmlinkage_protect

git commit 54a015104136974262afa4b8ddd943ea70dec8a2 ("asmlinkage_protect
replaces prevent_tail_call") causes this build failure on s390:

    AS      arch/s390/kernel/entry64.o
  In file included from arch/s390/kernel/entry64.S:14:
  include/linux/linkage.h:34: error: syntax error in macro parameter list
  make[1]: *** [arch/s390/kernel/entry64.o] Error 1
  make: *** [arch/s390/kernel] Error 2

and some other architectures.  The reason is that some architectures add
the "-traditional" flag to the invocation of $(AS), which disables
variadic macro argument support.

So just surround the new define with an #ifndef __ASSEMBLY__ to prevent
any side effects on asm code.

Cc: Roland McGrath <roland@redhat.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/linkage.h | 3 +++
 1 file changed, 3 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/linkage.h b/include/linux/linkage.h
index b163c5c40dbc..2119610b24f8 100644
--- a/include/linux/linkage.h
+++ b/include/linux/linkage.h
@@ -30,9 +30,12 @@
  * protection to work (ie no more work that the compiler might
  * end up needing stack temporaries for).
  */
+/* Assembly files may be compiled with -traditional .. */
+#ifndef __ASSEMBLY__
 #ifndef asmlinkage_protect
 # define asmlinkage_protect(n, ret, args...)	do { } while (0)
 #endif
+#endif
 
 #ifndef __ALIGN
 #define __ALIGN		.align 4,0x90
-- 
cgit v1.2.3


From dc07e721a26ec7e0adb66340f1763d220cfbbd0c Mon Sep 17 00:00:00 2001
From: "J. Bruce Fields" <bfields@citi.umich.edu>
Date: Mon, 7 Apr 2008 15:59:05 -0400
Subject: Spell out behavior of atomic_dec_and_lock() in kerneldoc

A little more detail here wouldn't hurt.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
---
 include/linux/spinlock.h | 3 +++
 1 file changed, 3 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h
index 576a5f77d3bd..1129ee0a7180 100644
--- a/include/linux/spinlock.h
+++ b/include/linux/spinlock.h
@@ -341,6 +341,9 @@ static inline void double_spin_unlock(spinlock_t *l1, spinlock_t *l2,
  * atomic_dec_and_lock - lock on reaching reference count zero
  * @atomic: the atomic counter
  * @lock: the spinlock in question
+ *
+ * Decrements @atomic by 1.  If the result is 0, returns true and locks
+ * @lock.  Returns false for all other cases.
  */
 extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock);
 #define atomic_dec_and_lock(atomic, lock) \
-- 
cgit v1.2.3


From f3ee4010e84452aa133e5163e6cfabc52b194e94 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Thu, 10 Apr 2008 15:42:11 +0900
Subject: [IPV6]: Define constants for link-local multicast addresses.

- Define link-local all-node / all-router multicast addresses.
- Remove ipv6_addr_all_nodes() and ipv6_addr_all_routers().

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/linux/in6.h    |  8 ++++++++
 include/net/addrconf.h | 11 -----------
 net/ipv6/addrconf.c    | 25 +++++++------------------
 net/ipv6/mcast.c       | 23 ++++++++---------------
 net/ipv6/ndisc.c       |  5 +----
 5 files changed, 24 insertions(+), 48 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/in6.h b/include/linux/in6.h
index e6aa8de2b939..bc492048c349 100644
--- a/include/linux/in6.h
+++ b/include/linux/in6.h
@@ -48,6 +48,14 @@ extern const struct in6_addr in6addr_any;
 #define IN6ADDR_ANY_INIT { { { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } } }
 extern const struct in6_addr in6addr_loopback;
 #define IN6ADDR_LOOPBACK_INIT { { { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1 } } }
+#ifdef __KERNEL__
+extern const struct in6_addr in6addr_linklocal_allnodes;
+#define IN6ADDR_LINKLOCAL_ALLNODES_INIT	\
+		{ { { 0xff,2,0,0,0,0,0,0,0,0,0,0,0,0,0,1 } } }
+extern const struct in6_addr in6addr_linklocal_allrouters;
+#define IN6ADDR_LINKLOCAL_ALLROUTERS_INIT \
+		{ { { 0xff,2,0,0,0,0,0,0,0,0,0,0,0,0,0,2 } } }
+#endif
 
 struct sockaddr_in6 {
 	unsigned short int	sin6_family;    /* AF_INET6 */
diff --git a/include/net/addrconf.h b/include/net/addrconf.h
index 92af23d66eb9..0a2f0372df31 100644
--- a/include/net/addrconf.h
+++ b/include/net/addrconf.h
@@ -205,17 +205,6 @@ static inline void addrconf_addr_solict_mult(const struct in6_addr *addr,
 		      htonl(0xFF000000) | addr->s6_addr32[3]);
 }
 
-
-static inline void ipv6_addr_all_nodes(struct in6_addr *addr)
-{
-	ipv6_addr_set(addr, htonl(0xFF020000), 0, 0, htonl(0x1));
-}
-
-static inline void ipv6_addr_all_routers(struct in6_addr *addr)
-{
-	ipv6_addr_set(addr, htonl(0xFF020000), 0, 0, htonl(0x2));
-}
-
 static inline int ipv6_addr_is_multicast(const struct in6_addr *addr)
 {
 	return (addr->s6_addr32[0] & htonl(0xFF000000)) == htonl(0xFF000000);
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 4048c2b73b0b..7df04d294924 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -222,6 +222,8 @@ static struct ipv6_devconf ipv6_devconf_dflt __read_mostly = {
 /* IPv6 Wildcard Address and Loopback Address defined by RFC2553 */
 const struct in6_addr in6addr_any = IN6ADDR_ANY_INIT;
 const struct in6_addr in6addr_loopback = IN6ADDR_LOOPBACK_INIT;
+const struct in6_addr in6addr_linklocal_allnodes = IN6ADDR_LINKLOCAL_ALLNODES_INIT;
+const struct in6_addr in6addr_linklocal_allrouters = IN6ADDR_LINKLOCAL_ALLROUTERS_INIT;
 
 /* Check if a valid qdisc is available */
 static inline int addrconf_qdisc_ok(struct net_device *dev)
@@ -321,7 +323,6 @@ EXPORT_SYMBOL(in6_dev_finish_destroy);
 static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 {
 	struct inet6_dev *ndev;
-	struct in6_addr maddr;
 
 	ASSERT_RTNL();
 
@@ -406,8 +407,7 @@ static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 	rcu_assign_pointer(dev->ip6_ptr, ndev);
 
 	/* Join all-node multicast group */
-	ipv6_addr_all_nodes(&maddr);
-	ipv6_dev_mc_inc(dev, &maddr);
+	ipv6_dev_mc_inc(dev, &in6addr_linklocal_allnodes);
 
 	return ndev;
 }
@@ -433,18 +433,15 @@ static void dev_forward_change(struct inet6_dev *idev)
 {
 	struct net_device *dev;
 	struct inet6_ifaddr *ifa;
-	struct in6_addr addr;
 
 	if (!idev)
 		return;
 	dev = idev->dev;
 	if (dev && (dev->flags & IFF_MULTICAST)) {
-		ipv6_addr_all_routers(&addr);
-
 		if (idev->cnf.forwarding)
-			ipv6_dev_mc_inc(dev, &addr);
+			ipv6_dev_mc_inc(dev, &in6addr_linklocal_allrouters);
 		else
-			ipv6_dev_mc_dec(dev, &addr);
+			ipv6_dev_mc_dec(dev, &in6addr_linklocal_allrouters);
 	}
 	for (ifa=idev->addr_list; ifa; ifa=ifa->if_next) {
 		if (ifa->flags&IFA_F_TENTATIVE)
@@ -2654,8 +2651,6 @@ static void addrconf_rs_timer(unsigned long data)
 
 	spin_lock(&ifp->lock);
 	if (ifp->probes++ < ifp->idev->cnf.rtr_solicits) {
-		struct in6_addr all_routers;
-
 		/* The wait after the last probe can be shorter */
 		addrconf_mod_timer(ifp, AC_RS,
 				   (ifp->probes == ifp->idev->cnf.rtr_solicits) ?
@@ -2663,9 +2658,7 @@ static void addrconf_rs_timer(unsigned long data)
 				   ifp->idev->cnf.rtr_solicit_interval);
 		spin_unlock(&ifp->lock);
 
-		ipv6_addr_all_routers(&all_routers);
-
-		ndisc_send_rs(ifp->idev->dev, &ifp->addr, &all_routers);
+		ndisc_send_rs(ifp->idev->dev, &ifp->addr, &in6addr_linklocal_allrouters);
 	} else {
 		spin_unlock(&ifp->lock);
 		/*
@@ -2806,16 +2799,12 @@ static void addrconf_dad_completed(struct inet6_ifaddr *ifp)
 	    ifp->idev->cnf.rtr_solicits > 0 &&
 	    (dev->flags&IFF_LOOPBACK) == 0 &&
 	    (ipv6_addr_type(&ifp->addr) & IPV6_ADDR_LINKLOCAL)) {
-		struct in6_addr all_routers;
-
-		ipv6_addr_all_routers(&all_routers);
-
 		/*
 		 *	If a host as already performed a random delay
 		 *	[...] as part of DAD [...] there is no need
 		 *	to delay again before sending the first RS
 		 */
-		ndisc_send_rs(ifp->idev->dev, &ifp->addr, &all_routers);
+		ndisc_send_rs(ifp->idev->dev, &ifp->addr, &in6addr_linklocal_allrouters);
 
 		spin_lock_bh(&ifp->lock);
 		ifp->probes = 1;
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 0a0132a1c443..c2dc2e2b6c07 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -1766,10 +1766,9 @@ static void igmp6_send(struct in6_addr *addr, struct net_device *dev, int type)
 	struct inet6_dev *idev;
 	struct sk_buff *skb;
 	struct icmp6hdr *hdr;
-	struct in6_addr *snd_addr;
+	const struct in6_addr *snd_addr;
 	struct in6_addr *addrp;
 	struct in6_addr addr_buf;
-	struct in6_addr all_routers;
 	int err, len, payload_len, full_len;
 	u8 ra[8] = { IPPROTO_ICMPV6, 0,
 		     IPV6_TLV_ROUTERALERT, 2, 0, 0,
@@ -1780,11 +1779,10 @@ static void igmp6_send(struct in6_addr *addr, struct net_device *dev, int type)
 	IP6_INC_STATS(__in6_dev_get(dev),
 		      IPSTATS_MIB_OUTREQUESTS);
 	rcu_read_unlock();
-	snd_addr = addr;
-	if (type == ICMPV6_MGM_REDUCTION) {
-		snd_addr = &all_routers;
-		ipv6_addr_all_routers(&all_routers);
-	}
+	if (type == ICMPV6_MGM_REDUCTION)
+		snd_addr = &in6addr_linklocal_allrouters;
+	else
+		snd_addr = addr;
 
 	len = sizeof(struct icmp6hdr) + sizeof(struct in6_addr);
 	payload_len = len + sizeof(ra);
@@ -2309,24 +2307,19 @@ void ipv6_mc_init_dev(struct inet6_dev *idev)
 void ipv6_mc_destroy_dev(struct inet6_dev *idev)
 {
 	struct ifmcaddr6 *i;
-	struct in6_addr maddr;
 
 	/* Deactivate timers */
 	ipv6_mc_down(idev);
 
 	/* Delete all-nodes address. */
-	ipv6_addr_all_nodes(&maddr);
-
 	/* We cannot call ipv6_dev_mc_dec() directly, our caller in
 	 * addrconf.c has NULL'd out dev->ip6_ptr so in6_dev_get() will
 	 * fail.
 	 */
-	__ipv6_dev_mc_dec(idev, &maddr);
+	__ipv6_dev_mc_dec(idev, &in6addr_linklocal_allnodes);
 
-	if (idev->cnf.forwarding) {
-		ipv6_addr_all_routers(&maddr);
-		__ipv6_dev_mc_dec(idev, &maddr);
-	}
+	if (idev->cnf.forwarding)
+		__ipv6_dev_mc_dec(idev, &in6addr_linklocal_allrouters);
 
 	write_lock_bh(&idev->lock);
 	while ((i = idev->mc_list) != NULL) {
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 5b9ad5e2f56d..2c74885f8355 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -818,10 +818,7 @@ static void ndisc_recv_ns(struct sk_buff *skb)
 		is_router = !!idev->cnf.forwarding;
 
 	if (dad) {
-		struct in6_addr maddr;
-
-		ipv6_addr_all_nodes(&maddr);
-		ndisc_send_na(dev, NULL, &maddr, &msg->target,
+		ndisc_send_na(dev, NULL, &in6addr_linklocal_allnodes, &msg->target,
 			      is_router, 0, (ifp != NULL), 1);
 		goto out;
 	}
-- 
cgit v1.2.3


From 14daa02139dcb3193b2b0250c0720a23ef610c49 Mon Sep 17 00:00:00 2001
From: Rusty Russell <rusty@rustcorp.com.au>
Date: Sat, 12 Apr 2008 18:48:58 -0700
Subject: net: make struct tun_struct private to tun.c

There's no reason for this to be in the header, and it just hurts
recompile time.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Max Krasnyanskiy <maxk@qualcomm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/tun.c      | 33 +++++++++++++++++++++++++++++++++
 include/linux/Kbuild   |  2 +-
 include/linux/if_tun.h | 39 ---------------------------------------
 3 files changed, 34 insertions(+), 40 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 7b816a032957..970ec4793442 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -67,9 +67,42 @@
 #include <asm/system.h>
 #include <asm/uaccess.h>
 
+/* Uncomment to enable debugging */
+/* #define TUN_DEBUG 1 */
+
 #ifdef TUN_DEBUG
 static int debug;
+
+#define DBG  if(tun->debug)printk
+#define DBG1 if(debug==2)printk
+#else
+#define DBG( a... )
+#define DBG1( a... )
+#endif
+
+struct tun_struct {
+	struct list_head        list;
+	unsigned long 		flags;
+	int			attached;
+	uid_t			owner;
+	gid_t			group;
+
+	wait_queue_head_t	read_wait;
+	struct sk_buff_head	readq;
+
+	struct net_device	*dev;
+
+	struct fasync_struct    *fasync;
+
+	unsigned long if_flags;
+	u8 dev_addr[ETH_ALEN];
+	u32 chr_filter[2];
+	u32 net_filter[2];
+
+#ifdef TUN_DEBUG
+	int debug;
 #endif
+};
 
 /* Network device part of the driver */
 
diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index 9cdd12a9e843..cedbbd806bf6 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -86,6 +86,7 @@ header-y += if_plip.h
 header-y += if_ppp.h
 header-y += if_slip.h
 header-y += if_strip.h
+header-y += if_tun.h
 header-y += if_tunnel.h
 header-y += in6.h
 header-y += in_route.h
@@ -229,7 +230,6 @@ unifdef-y += if_link.h
 unifdef-y += if_pppol2tp.h
 unifdef-y += if_pppox.h
 unifdef-y += if_tr.h
-unifdef-y += if_tun.h
 unifdef-y += if_vlan.h
 unifdef-y += if_wanpipe.h
 unifdef-y += igmp.h
diff --git a/include/linux/if_tun.h b/include/linux/if_tun.h
index 72f1c5f47be3..8c71fe2fb1f5 100644
--- a/include/linux/if_tun.h
+++ b/include/linux/if_tun.h
@@ -18,47 +18,8 @@
 #ifndef __IF_TUN_H
 #define __IF_TUN_H
 
-/* Uncomment to enable debugging */
-/* #define TUN_DEBUG 1 */
-
 #include <linux/types.h>
 
-#ifdef __KERNEL__
-
-#ifdef TUN_DEBUG
-#define DBG  if(tun->debug)printk
-#define DBG1 if(debug==2)printk
-#else
-#define DBG( a... )
-#define DBG1( a... )
-#endif
-
-struct tun_struct {
-	struct list_head        list;
-	unsigned long 		flags;
-	int			attached;
-	uid_t			owner;
-	gid_t			group;
-
-	wait_queue_head_t	read_wait;
-	struct sk_buff_head	readq;
-
-	struct net_device	*dev;
-
-	struct fasync_struct    *fasync;
-
-	unsigned long if_flags;
-	u8 dev_addr[ETH_ALEN];
-	u32 chr_filter[2];
-	u32 net_filter[2];
-
-#ifdef TUN_DEBUG	
-	int debug;
-#endif  
-};
-
-#endif /* __KERNEL__ */
-
 /* Read queue size */
 #define TUN_READQ_SIZE	500
 
-- 
cgit v1.2.3


From 03e1ad7b5d871d4189b1da3125c2f12d1b5f7d0b Mon Sep 17 00:00:00 2001
From: Paul Moore <paul.moore@hp.com>
Date: Sat, 12 Apr 2008 19:07:52 -0700
Subject: LSM: Make the Labeled IPsec hooks more stack friendly

The xfrm_get_policy() and xfrm_add_pol_expire() put some rather large structs
on the stack to work around the LSM API.  This patch attempts to fix that
problem by changing the LSM API to require only the relevant "security"
pointers instead of the entire SPD entry; we do this for all of the
security_xfrm_policy*() functions to keep things consistent.

Signed-off-by: Paul Moore <paul.moore@hp.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/security.h        | 48 ++++++++++++++++++++---------------------
 net/key/af_key.c                | 23 ++++++++++----------
 net/xfrm/xfrm_policy.c          | 24 +++++++++++++--------
 net/xfrm/xfrm_user.c            | 33 ++++++++++++++--------------
 security/dummy.c                | 14 ++++++------
 security/security.c             | 21 +++++++++---------
 security/selinux/include/xfrm.h | 13 +++++------
 security/selinux/xfrm.c         | 39 ++++++++++++++-------------------
 8 files changed, 109 insertions(+), 106 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/security.h b/include/linux/security.h
index c673dfd4dffc..f5eb9ff47ac5 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -910,24 +910,24 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
  * Security hooks for XFRM operations.
  *
  * @xfrm_policy_alloc_security:
- *	@xp contains the xfrm_policy being added to Security Policy Database
- *	used by the XFRM system.
+ *	@ctxp is a pointer to the xfrm_sec_ctx being added to Security Policy
+ *	Database used by the XFRM system.
  *	@sec_ctx contains the security context information being provided by
  *	the user-level policy update program (e.g., setkey).
  *	Allocate a security structure to the xp->security field; the security
  *	field is initialized to NULL when the xfrm_policy is allocated.
  *	Return 0 if operation was successful (memory to allocate, legal context)
  * @xfrm_policy_clone_security:
- *	@old contains an existing xfrm_policy in the SPD.
- *	@new contains a new xfrm_policy being cloned from old.
- *	Allocate a security structure to the new->security field
- *	that contains the information from the old->security field.
+ *	@old_ctx contains an existing xfrm_sec_ctx.
+ *	@new_ctxp contains a new xfrm_sec_ctx being cloned from old.
+ *	Allocate a security structure in new_ctxp that contains the
+ *	information from the old_ctx structure.
  *	Return 0 if operation was successful (memory to allocate).
  * @xfrm_policy_free_security:
- *	@xp contains the xfrm_policy
+ *	@ctx contains the xfrm_sec_ctx
  *	Deallocate xp->security.
  * @xfrm_policy_delete_security:
- *	@xp contains the xfrm_policy.
+ *	@ctx contains the xfrm_sec_ctx.
  *	Authorize deletion of xp->security.
  * @xfrm_state_alloc_security:
  *	@x contains the xfrm_state being added to the Security Association
@@ -947,7 +947,7 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
  *	@x contains the xfrm_state.
  *	Authorize deletion of x->security.
  * @xfrm_policy_lookup:
- *	@xp contains the xfrm_policy for which the access control is being
+ *	@ctx contains the xfrm_sec_ctx for which the access control is being
  *	checked.
  *	@fl_secid contains the flow security label that is used to authorize
  *	access to the policy xp.
@@ -1454,17 +1454,17 @@ struct security_operations {
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
-	int (*xfrm_policy_alloc_security) (struct xfrm_policy *xp,
+	int (*xfrm_policy_alloc_security) (struct xfrm_sec_ctx **ctxp,
 			struct xfrm_user_sec_ctx *sec_ctx);
-	int (*xfrm_policy_clone_security) (struct xfrm_policy *old, struct xfrm_policy *new);
-	void (*xfrm_policy_free_security) (struct xfrm_policy *xp);
-	int (*xfrm_policy_delete_security) (struct xfrm_policy *xp);
+	int (*xfrm_policy_clone_security) (struct xfrm_sec_ctx *old_ctx, struct xfrm_sec_ctx **new_ctx);
+	void (*xfrm_policy_free_security) (struct xfrm_sec_ctx *ctx);
+	int (*xfrm_policy_delete_security) (struct xfrm_sec_ctx *ctx);
 	int (*xfrm_state_alloc_security) (struct xfrm_state *x,
 		struct xfrm_user_sec_ctx *sec_ctx,
 		u32 secid);
 	void (*xfrm_state_free_security) (struct xfrm_state *x);
 	int (*xfrm_state_delete_security) (struct xfrm_state *x);
-	int (*xfrm_policy_lookup)(struct xfrm_policy *xp, u32 fl_secid, u8 dir);
+	int (*xfrm_policy_lookup)(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir);
 	int (*xfrm_state_pol_flow_match)(struct xfrm_state *x,
 			struct xfrm_policy *xp, struct flowi *fl);
 	int (*xfrm_decode_session)(struct sk_buff *skb, u32 *secid, int ckall);
@@ -2562,16 +2562,16 @@ static inline void security_inet_conn_established(struct sock *sk,
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 
-int security_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx);
-int security_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new);
-void security_xfrm_policy_free(struct xfrm_policy *xp);
-int security_xfrm_policy_delete(struct xfrm_policy *xp);
+int security_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp, struct xfrm_user_sec_ctx *sec_ctx);
+int security_xfrm_policy_clone(struct xfrm_sec_ctx *old_ctx, struct xfrm_sec_ctx **new_ctxp);
+void security_xfrm_policy_free(struct xfrm_sec_ctx *ctx);
+int security_xfrm_policy_delete(struct xfrm_sec_ctx *ctx);
 int security_xfrm_state_alloc(struct xfrm_state *x, struct xfrm_user_sec_ctx *sec_ctx);
 int security_xfrm_state_alloc_acquire(struct xfrm_state *x,
 				      struct xfrm_sec_ctx *polsec, u32 secid);
 int security_xfrm_state_delete(struct xfrm_state *x);
 void security_xfrm_state_free(struct xfrm_state *x);
-int security_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir);
+int security_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir);
 int security_xfrm_state_pol_flow_match(struct xfrm_state *x,
 				       struct xfrm_policy *xp, struct flowi *fl);
 int security_xfrm_decode_session(struct sk_buff *skb, u32 *secid);
@@ -2579,21 +2579,21 @@ void security_skb_classify_flow(struct sk_buff *skb, struct flowi *fl);
 
 #else	/* CONFIG_SECURITY_NETWORK_XFRM */
 
-static inline int security_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx)
+static inline int security_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp, struct xfrm_user_sec_ctx *sec_ctx)
 {
 	return 0;
 }
 
-static inline int security_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new)
+static inline int security_xfrm_policy_clone(struct xfrm_sec_ctx *old, struct xfrm_sec_ctx **new_ctxp)
 {
 	return 0;
 }
 
-static inline void security_xfrm_policy_free(struct xfrm_policy *xp)
+static inline void security_xfrm_policy_free(struct xfrm_sec_ctx *ctx)
 {
 }
 
-static inline int security_xfrm_policy_delete(struct xfrm_policy *xp)
+static inline int security_xfrm_policy_delete(struct xfrm_sec_ctx *ctx)
 {
 	return 0;
 }
@@ -2619,7 +2619,7 @@ static inline int security_xfrm_state_delete(struct xfrm_state *x)
 	return 0;
 }
 
-static inline int security_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir)
+static inline int security_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir)
 {
 	return 0;
 }
diff --git a/net/key/af_key.c b/net/key/af_key.c
index 6db58924368a..1fb0fe42a72e 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -2292,7 +2292,7 @@ static int pfkey_spdadd(struct sock *sk, struct sk_buff *skb, struct sadb_msg *h
 			goto out;
 		}
 
-		err = security_xfrm_policy_alloc(xp, uctx);
+		err = security_xfrm_policy_alloc(&xp->security, uctx);
 		kfree(uctx);
 
 		if (err)
@@ -2352,10 +2352,11 @@ static int pfkey_spddelete(struct sock *sk, struct sk_buff *skb, struct sadb_msg
 	int err;
 	struct sadb_address *sa;
 	struct sadb_x_policy *pol;
-	struct xfrm_policy *xp, tmp;
+	struct xfrm_policy *xp;
 	struct xfrm_selector sel;
 	struct km_event c;
 	struct sadb_x_sec_ctx *sec_ctx;
+	struct xfrm_sec_ctx *pol_ctx;
 
 	if (!present_and_same_family(ext_hdrs[SADB_EXT_ADDRESS_SRC-1],
 				     ext_hdrs[SADB_EXT_ADDRESS_DST-1]) ||
@@ -2385,25 +2386,23 @@ static int pfkey_spddelete(struct sock *sk, struct sk_buff *skb, struct sadb_msg
 		sel.dport_mask = htons(0xffff);
 
 	sec_ctx = (struct sadb_x_sec_ctx *) ext_hdrs[SADB_X_EXT_SEC_CTX-1];
-	memset(&tmp, 0, sizeof(struct xfrm_policy));
-
 	if (sec_ctx != NULL) {
 		struct xfrm_user_sec_ctx *uctx = pfkey_sadb2xfrm_user_sec_ctx(sec_ctx);
 
 		if (!uctx)
 			return -ENOMEM;
 
-		err = security_xfrm_policy_alloc(&tmp, uctx);
+		err = security_xfrm_policy_alloc(&pol_ctx, uctx);
 		kfree(uctx);
-
 		if (err)
 			return err;
-	}
-
-	xp = xfrm_policy_bysel_ctx(XFRM_POLICY_TYPE_MAIN, pol->sadb_x_policy_dir-1,
-				   &sel, tmp.security, 1, &err);
-	security_xfrm_policy_free(&tmp);
+	} else
+		pol_ctx = NULL;
 
+	xp = xfrm_policy_bysel_ctx(XFRM_POLICY_TYPE_MAIN,
+				   pol->sadb_x_policy_dir - 1, &sel, pol_ctx,
+				   1, &err);
+	security_xfrm_policy_free(pol_ctx);
 	if (xp == NULL)
 		return -ENOENT;
 
@@ -3298,7 +3297,7 @@ static struct xfrm_policy *pfkey_compile_policy(struct sock *sk, int opt,
 		if ((*dir = verify_sec_ctx_len(p)))
 			goto out;
 		uctx = pfkey_sadb2xfrm_user_sec_ctx(sec_ctx);
-		*dir = security_xfrm_policy_alloc(xp, uctx);
+		*dir = security_xfrm_policy_alloc(&xp->security, uctx);
 		kfree(uctx);
 
 		if (*dir)
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 15d73e47cc2c..ab4d0e598a2c 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -263,7 +263,7 @@ void xfrm_policy_destroy(struct xfrm_policy *policy)
 	list_del(&policy->bytype);
 	write_unlock_bh(&xfrm_policy_lock);
 
-	security_xfrm_policy_free(policy);
+	security_xfrm_policy_free(policy->security);
 	kfree(policy);
 }
 EXPORT_SYMBOL(xfrm_policy_destroy);
@@ -676,7 +676,8 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(u8 type, int dir,
 		    xfrm_sec_ctx_match(ctx, pol->security)) {
 			xfrm_pol_hold(pol);
 			if (delete) {
-				*err = security_xfrm_policy_delete(pol);
+				*err = security_xfrm_policy_delete(
+								pol->security);
 				if (*err) {
 					write_unlock_bh(&xfrm_policy_lock);
 					return pol;
@@ -718,7 +719,8 @@ struct xfrm_policy *xfrm_policy_byid(u8 type, int dir, u32 id, int delete,
 		if (pol->type == type && pol->index == id) {
 			xfrm_pol_hold(pol);
 			if (delete) {
-				*err = security_xfrm_policy_delete(pol);
+				*err = security_xfrm_policy_delete(
+								pol->security);
 				if (*err) {
 					write_unlock_bh(&xfrm_policy_lock);
 					return pol;
@@ -756,7 +758,7 @@ xfrm_policy_flush_secctx_check(u8 type, struct xfrm_audit *audit_info)
 				     &xfrm_policy_inexact[dir], bydst) {
 			if (pol->type != type)
 				continue;
-			err = security_xfrm_policy_delete(pol);
+			err = security_xfrm_policy_delete(pol->security);
 			if (err) {
 				xfrm_audit_policy_delete(pol, 0,
 							 audit_info->loginuid,
@@ -770,7 +772,8 @@ xfrm_policy_flush_secctx_check(u8 type, struct xfrm_audit *audit_info)
 					     bydst) {
 				if (pol->type != type)
 					continue;
-				err = security_xfrm_policy_delete(pol);
+				err = security_xfrm_policy_delete(
+								pol->security);
 				if (err) {
 					xfrm_audit_policy_delete(pol, 0,
 							audit_info->loginuid,
@@ -931,7 +934,8 @@ static int xfrm_policy_match(struct xfrm_policy *pol, struct flowi *fl,
 
 	match = xfrm_selector_match(sel, fl, family);
 	if (match)
-		ret = security_xfrm_policy_lookup(pol, fl->secid, dir);
+		ret = security_xfrm_policy_lookup(pol->security, fl->secid,
+						  dir);
 
 	return ret;
 }
@@ -1048,8 +1052,9 @@ static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, struc
 		int err = 0;
 
 		if (match) {
-			err = security_xfrm_policy_lookup(pol, fl->secid,
-					policy_to_flow_dir(dir));
+			err = security_xfrm_policy_lookup(pol->security,
+						      fl->secid,
+						      policy_to_flow_dir(dir));
 			if (!err)
 				xfrm_pol_hold(pol);
 			else if (err == -ESRCH)
@@ -1138,7 +1143,8 @@ static struct xfrm_policy *clone_policy(struct xfrm_policy *old, int dir)
 
 	if (newp) {
 		newp->selector = old->selector;
-		if (security_xfrm_policy_clone(old, newp)) {
+		if (security_xfrm_policy_clone(old->security,
+					       &newp->security)) {
 			kfree(newp);
 			return NULL;  /* ENOMEM */
 		}
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 5578c909fcf6..ecf9d67daef5 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -959,7 +959,7 @@ static int copy_from_user_sec_ctx(struct xfrm_policy *pol, struct nlattr **attrs
 		return 0;
 
 	uctx = nla_data(rt);
-	return security_xfrm_policy_alloc(pol, uctx);
+	return security_xfrm_policy_alloc(&pol->security, uctx);
 }
 
 static void copy_templates(struct xfrm_policy *xp, struct xfrm_user_tmpl *ut,
@@ -1143,7 +1143,7 @@ static int xfrm_add_policy(struct sk_buff *skb, struct nlmsghdr *nlh,
 			      NETLINK_CB(skb).sid);
 
 	if (err) {
-		security_xfrm_policy_free(xp);
+		security_xfrm_policy_free(xp->security);
 		kfree(xp);
 		return err;
 	}
@@ -1337,22 +1337,23 @@ static int xfrm_get_policy(struct sk_buff *skb, struct nlmsghdr *nlh,
 		xp = xfrm_policy_byid(type, p->dir, p->index, delete, &err);
 	else {
 		struct nlattr *rt = attrs[XFRMA_SEC_CTX];
-		struct xfrm_policy tmp;
+		struct xfrm_sec_ctx *ctx;
 
 		err = verify_sec_ctx_len(attrs);
 		if (err)
 			return err;
 
-		memset(&tmp, 0, sizeof(struct xfrm_policy));
 		if (rt) {
 			struct xfrm_user_sec_ctx *uctx = nla_data(rt);
 
-			if ((err = security_xfrm_policy_alloc(&tmp, uctx)))
+			err = security_xfrm_policy_alloc(&ctx, uctx);
+			if (err)
 				return err;
-		}
-		xp = xfrm_policy_bysel_ctx(type, p->dir, &p->sel, tmp.security,
+		} else
+			ctx = NULL;
+		xp = xfrm_policy_bysel_ctx(type, p->dir, &p->sel, ctx,
 					   delete, &err);
-		security_xfrm_policy_free(&tmp);
+		security_xfrm_policy_free(ctx);
 	}
 	if (xp == NULL)
 		return -ENOENT;
@@ -1572,26 +1573,26 @@ static int xfrm_add_pol_expire(struct sk_buff *skb, struct nlmsghdr *nlh,
 		xp = xfrm_policy_byid(type, p->dir, p->index, 0, &err);
 	else {
 		struct nlattr *rt = attrs[XFRMA_SEC_CTX];
-		struct xfrm_policy tmp;
+		struct xfrm_sec_ctx *ctx;
 
 		err = verify_sec_ctx_len(attrs);
 		if (err)
 			return err;
 
-		memset(&tmp, 0, sizeof(struct xfrm_policy));
 		if (rt) {
 			struct xfrm_user_sec_ctx *uctx = nla_data(rt);
 
-			if ((err = security_xfrm_policy_alloc(&tmp, uctx)))
+			err = security_xfrm_policy_alloc(&ctx, uctx);
+			if (err)
 				return err;
-		}
-		xp = xfrm_policy_bysel_ctx(type, p->dir, &p->sel, tmp.security,
-					   0, &err);
-		security_xfrm_policy_free(&tmp);
+		} else
+			ctx = NULL;
+		xp = xfrm_policy_bysel_ctx(type, p->dir, &p->sel, ctx, 0, &err);
+		security_xfrm_policy_free(ctx);
 	}
-
 	if (xp == NULL)
 		return -ENOENT;
+
 	read_lock(&xp->lock);
 	if (xp->dead) {
 		read_unlock(&xp->lock);
diff --git a/security/dummy.c b/security/dummy.c
index 78d8f92310a4..480366f9c41d 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -876,22 +876,23 @@ static inline void dummy_req_classify_flow(const struct request_sock *req,
 #endif	/* CONFIG_SECURITY_NETWORK */
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
-static int dummy_xfrm_policy_alloc_security(struct xfrm_policy *xp,
-		struct xfrm_user_sec_ctx *sec_ctx)
+static int dummy_xfrm_policy_alloc_security(struct xfrm_sec_ctx **ctxp,
+					    struct xfrm_user_sec_ctx *sec_ctx)
 {
 	return 0;
 }
 
-static inline int dummy_xfrm_policy_clone_security(struct xfrm_policy *old, struct xfrm_policy *new)
+static inline int dummy_xfrm_policy_clone_security(struct xfrm_sec_ctx *old_ctx,
+					   struct xfrm_sec_ctx **new_ctxp)
 {
 	return 0;
 }
 
-static void dummy_xfrm_policy_free_security(struct xfrm_policy *xp)
+static void dummy_xfrm_policy_free_security(struct xfrm_sec_ctx *ctx)
 {
 }
 
-static int dummy_xfrm_policy_delete_security(struct xfrm_policy *xp)
+static int dummy_xfrm_policy_delete_security(struct xfrm_sec_ctx *ctx)
 {
 	return 0;
 }
@@ -911,7 +912,8 @@ static int dummy_xfrm_state_delete_security(struct xfrm_state *x)
 	return 0;
 }
 
-static int dummy_xfrm_policy_lookup(struct xfrm_policy *xp, u32 sk_sid, u8 dir)
+static int dummy_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx,
+				    u32 sk_sid, u8 dir)
 {
 	return 0;
 }
diff --git a/security/security.c b/security/security.c
index b1387a6b416d..c9ff7d18c2f4 100644
--- a/security/security.c
+++ b/security/security.c
@@ -1014,26 +1014,27 @@ void security_inet_conn_established(struct sock *sk,
 
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 
-int security_xfrm_policy_alloc(struct xfrm_policy *xp, struct xfrm_user_sec_ctx *sec_ctx)
+int security_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp, struct xfrm_user_sec_ctx *sec_ctx)
 {
-	return security_ops->xfrm_policy_alloc_security(xp, sec_ctx);
+	return security_ops->xfrm_policy_alloc_security(ctxp, sec_ctx);
 }
 EXPORT_SYMBOL(security_xfrm_policy_alloc);
 
-int security_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new)
+int security_xfrm_policy_clone(struct xfrm_sec_ctx *old_ctx,
+			      struct xfrm_sec_ctx **new_ctxp)
 {
-	return security_ops->xfrm_policy_clone_security(old, new);
+	return security_ops->xfrm_policy_clone_security(old_ctx, new_ctxp);
 }
 
-void security_xfrm_policy_free(struct xfrm_policy *xp)
+void security_xfrm_policy_free(struct xfrm_sec_ctx *ctx)
 {
-	security_ops->xfrm_policy_free_security(xp);
+	security_ops->xfrm_policy_free_security(ctx);
 }
 EXPORT_SYMBOL(security_xfrm_policy_free);
 
-int security_xfrm_policy_delete(struct xfrm_policy *xp)
+int security_xfrm_policy_delete(struct xfrm_sec_ctx *ctx)
 {
-	return security_ops->xfrm_policy_delete_security(xp);
+	return security_ops->xfrm_policy_delete_security(ctx);
 }
 
 int security_xfrm_state_alloc(struct xfrm_state *x, struct xfrm_user_sec_ctx *sec_ctx)
@@ -1065,9 +1066,9 @@ void security_xfrm_state_free(struct xfrm_state *x)
 	security_ops->xfrm_state_free_security(x);
 }
 
-int security_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir)
+int security_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir)
 {
-	return security_ops->xfrm_policy_lookup(xp, fl_secid, dir);
+	return security_ops->xfrm_policy_lookup(ctx, fl_secid, dir);
 }
 
 int security_xfrm_state_pol_flow_match(struct xfrm_state *x,
diff --git a/security/selinux/include/xfrm.h b/security/selinux/include/xfrm.h
index 36b0510efa7b..289e24b39e3e 100644
--- a/security/selinux/include/xfrm.h
+++ b/security/selinux/include/xfrm.h
@@ -7,16 +7,17 @@
 #ifndef _SELINUX_XFRM_H_
 #define _SELINUX_XFRM_H_
 
-int selinux_xfrm_policy_alloc(struct xfrm_policy *xp,
-		struct xfrm_user_sec_ctx *sec_ctx);
-int selinux_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new);
-void selinux_xfrm_policy_free(struct xfrm_policy *xp);
-int selinux_xfrm_policy_delete(struct xfrm_policy *xp);
+int selinux_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp,
+			      struct xfrm_user_sec_ctx *sec_ctx);
+int selinux_xfrm_policy_clone(struct xfrm_sec_ctx *old_ctx,
+			      struct xfrm_sec_ctx **new_ctxp);
+void selinux_xfrm_policy_free(struct xfrm_sec_ctx *ctx);
+int selinux_xfrm_policy_delete(struct xfrm_sec_ctx *ctx);
 int selinux_xfrm_state_alloc(struct xfrm_state *x,
 	struct xfrm_user_sec_ctx *sec_ctx, u32 secid);
 void selinux_xfrm_state_free(struct xfrm_state *x);
 int selinux_xfrm_state_delete(struct xfrm_state *x);
-int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir);
+int selinux_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir);
 int selinux_xfrm_state_pol_flow_match(struct xfrm_state *x,
 			struct xfrm_policy *xp, struct flowi *fl);
 
diff --git a/security/selinux/xfrm.c b/security/selinux/xfrm.c
index 7e158205d081..874d17c83c61 100644
--- a/security/selinux/xfrm.c
+++ b/security/selinux/xfrm.c
@@ -77,20 +77,18 @@ static inline int selinux_authorizable_xfrm(struct xfrm_state *x)
  * LSM hook implementation that authorizes that a flow can use
  * a xfrm policy rule.
  */
-int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir)
+int selinux_xfrm_policy_lookup(struct xfrm_sec_ctx *ctx, u32 fl_secid, u8 dir)
 {
 	int rc;
 	u32 sel_sid;
-	struct xfrm_sec_ctx *ctx;
 
 	/* Context sid is either set to label or ANY_ASSOC */
-	if ((ctx = xp->security)) {
+	if (ctx) {
 		if (!selinux_authorizable_ctx(ctx))
 			return -EINVAL;
 
 		sel_sid = ctx->ctx_sid;
-	}
-	else
+	} else
 		/*
 		 * All flows should be treated as polmatch'ing an
 		 * otherwise applicable "non-labeled" policy. This
@@ -103,7 +101,7 @@ int selinux_xfrm_policy_lookup(struct xfrm_policy *xp, u32 fl_secid, u8 dir)
 			  NULL);
 
 	if (rc == -EACCES)
-		rc = -ESRCH;
+		return -ESRCH;
 
 	return rc;
 }
@@ -287,15 +285,14 @@ out2:
  * LSM hook implementation that allocs and transfers uctx spec to
  * xfrm_policy.
  */
-int selinux_xfrm_policy_alloc(struct xfrm_policy *xp,
-		struct xfrm_user_sec_ctx *uctx)
+int selinux_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp,
+			      struct xfrm_user_sec_ctx *uctx)
 {
 	int err;
 
-	BUG_ON(!xp);
 	BUG_ON(!uctx);
 
-	err = selinux_xfrm_sec_ctx_alloc(&xp->security, uctx, 0);
+	err = selinux_xfrm_sec_ctx_alloc(ctxp, uctx, 0);
 	if (err == 0)
 		atomic_inc(&selinux_xfrm_refcount);
 
@@ -307,32 +304,29 @@ int selinux_xfrm_policy_alloc(struct xfrm_policy *xp,
  * LSM hook implementation that copies security data structure from old to
  * new for policy cloning.
  */
-int selinux_xfrm_policy_clone(struct xfrm_policy *old, struct xfrm_policy *new)
+int selinux_xfrm_policy_clone(struct xfrm_sec_ctx *old_ctx,
+			      struct xfrm_sec_ctx **new_ctxp)
 {
-	struct xfrm_sec_ctx *old_ctx, *new_ctx;
-
-	old_ctx = old->security;
+	struct xfrm_sec_ctx *new_ctx;
 
 	if (old_ctx) {
-		new_ctx = new->security = kmalloc(sizeof(*new_ctx) +
-						  old_ctx->ctx_len,
-						  GFP_KERNEL);
-
+		new_ctx = kmalloc(sizeof(*old_ctx) + old_ctx->ctx_len,
+				  GFP_KERNEL);
 		if (!new_ctx)
 			return -ENOMEM;
 
 		memcpy(new_ctx, old_ctx, sizeof(*new_ctx));
 		memcpy(new_ctx->ctx_str, old_ctx->ctx_str, new_ctx->ctx_len);
+		*new_ctxp = new_ctx;
 	}
 	return 0;
 }
 
 /*
- * LSM hook implementation that frees xfrm_policy security information.
+ * LSM hook implementation that frees xfrm_sec_ctx security information.
  */
-void selinux_xfrm_policy_free(struct xfrm_policy *xp)
+void selinux_xfrm_policy_free(struct xfrm_sec_ctx *ctx)
 {
-	struct xfrm_sec_ctx *ctx = xp->security;
 	if (ctx)
 		kfree(ctx);
 }
@@ -340,10 +334,9 @@ void selinux_xfrm_policy_free(struct xfrm_policy *xp)
 /*
  * LSM hook implementation that authorizes deletion of labeled policies.
  */
-int selinux_xfrm_policy_delete(struct xfrm_policy *xp)
+int selinux_xfrm_policy_delete(struct xfrm_sec_ctx *ctx)
 {
 	struct task_security_struct *tsec = current->security;
-	struct xfrm_sec_ctx *ctx = xp->security;
 	int rc = 0;
 
 	if (ctx) {
-- 
cgit v1.2.3


From cee8947338d46bccece54c752bf6cd4043035f05 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Sun, 13 Apr 2008 23:21:16 -0700
Subject: [IPV6] MROUTE: Do not call ipv6_find_idev() directly.

Since NETDEV_REGISTER notifier chain is responsible for creating
inet6_dev{}, we do not need to call ipv6_find_idev() directly here.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/mroute6.h | 3 ---
 net/ipv6/addrconf.c     | 2 +-
 net/ipv6/ip6mr.c        | 5 -----
 3 files changed, 1 insertion(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/mroute6.h b/include/linux/mroute6.h
index f6469fb90840..e7989593142b 100644
--- a/include/linux/mroute6.h
+++ b/include/linux/mroute6.h
@@ -117,9 +117,6 @@ struct sioc_mif_req6
 
 #include <linux/skbuff.h>	/* for struct sk_buff_head */
 
-struct net_device;
-struct inet6_dev *ipv6_find_idev(struct net_device *dev);
-
 #ifdef CONFIG_IPV6_MROUTE
 static inline int ip6_mroute_opt(int opt)
 {
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index e93fa62089f8..9d49ed2578d7 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -412,7 +412,7 @@ static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 	return ndev;
 }
 
-struct inet6_dev * ipv6_find_idev(struct net_device *dev)
+static struct inet6_dev * ipv6_find_idev(struct net_device *dev)
 {
 	struct inet6_dev *idev;
 
diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c
index 94ede696da2a..6e2e3c957a31 100644
--- a/net/ipv6/ip6mr.c
+++ b/net/ipv6/ip6mr.c
@@ -435,7 +435,6 @@ static void reg_vif_setup(struct net_device *dev)
 static struct net_device *ip6mr_reg_vif(void)
 {
 	struct net_device *dev;
-	struct inet6_dev *in_dev;
 
 	dev = alloc_netdev(sizeof(struct net_device_stats), "pim6reg",
 			   reg_vif_setup);
@@ -449,10 +448,6 @@ static struct net_device *ip6mr_reg_vif(void)
 	}
 	dev->iflink = 0;
 
-	in_dev = ipv6_find_idev(dev);
-	if (!in_dev)
-		goto failure;
-
 	if (dev_open(dev))
 		goto failure;
 
-- 
cgit v1.2.3


From 7cd636fe9ce5de0051c11283911baca4291868c8 Mon Sep 17 00:00:00 2001
From: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Date: Sun, 13 Apr 2008 23:33:52 -0700
Subject: [IPV6]: IPv6 extension header structures need to be packed.

struct ipv6_opt_hdr is the common structure for IPv6 extension
headers, and it is common to increment the pointer to get
the real content.  On the other hand, since the structure
consists only of 1-byte next-header field and 1-byte length
field, size of that structure depends on architecture; 2 or 4.
Add "packed" attribute to get 2.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/ipv6.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 4aaefc349a4b..134c8e5cf07c 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -53,7 +53,7 @@ struct ipv6_opt_hdr {
 	/* 
 	 * TLV encoded option data follows.
 	 */
-};
+} __attribute__ ((packed));	/* required for some archs */
 
 #define ipv6_destopt_hdr ipv6_opt_hdr
 #define ipv6_hopopt_hdr  ipv6_opt_hdr
-- 
cgit v1.2.3


From f525c06d12b72cddb085df7f6f348c3c5a39b3ce Mon Sep 17 00:00:00 2001
From: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Date: Mon, 14 Apr 2008 00:04:12 -0700
Subject: [SKB]: __skb_dequeue = skb_peek + __skb_unlink

By rearranging the order of declarations, __skb_dequeue() is expressed in terms of

 * skb_peek() and
 * __skb_unlink(),

thus in effect mirroring the analogue implementation of __skb_dequeue_tail().

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 47 ++++++++++++++++-------------------------------
 1 file changed, 16 insertions(+), 31 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index e517701c25ba..c2116200580a 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -734,35 +734,6 @@ static inline void __skb_queue_tail(struct sk_buff_head *list,
 	next->prev  = prev->next = newsk;
 }
 
-
-/**
- *	__skb_dequeue - remove from the head of the queue
- *	@list: list to dequeue from
- *
- *	Remove the head of the list. This function does not take any locks
- *	so must be used with appropriate locks held only. The head item is
- *	returned or %NULL if the list is empty.
- */
-extern struct sk_buff *skb_dequeue(struct sk_buff_head *list);
-static inline struct sk_buff *__skb_dequeue(struct sk_buff_head *list)
-{
-	struct sk_buff *next, *prev, *result;
-
-	prev = (struct sk_buff *) list;
-	next = prev->next;
-	result = NULL;
-	if (next != prev) {
-		result	     = next;
-		next	     = next->next;
-		list->qlen--;
-		next->prev   = prev;
-		prev->next   = next;
-		result->next = result->prev = NULL;
-	}
-	return result;
-}
-
-
 /*
  *	Insert a packet on a list.
  */
@@ -803,8 +774,22 @@ static inline void __skb_unlink(struct sk_buff *skb, struct sk_buff_head *list)
 	prev->next = next;
 }
 
-
-/* XXX: more streamlined implementation */
+/**
+ *	__skb_dequeue - remove from the head of the queue
+ *	@list: list to dequeue from
+ *
+ *	Remove the head of the list. This function does not take any locks
+ *	so must be used with appropriate locks held only. The head item is
+ *	returned or %NULL if the list is empty.
+ */
+extern struct sk_buff *skb_dequeue(struct sk_buff_head *list);
+static inline struct sk_buff *__skb_dequeue(struct sk_buff_head *list)
+{
+	struct sk_buff *skb = skb_peek(list);
+	if (skb)
+		__skb_unlink(skb, list);
+	return skb;
+}
 
 /**
  *	__skb_dequeue_tail - remove from the tail of the queue
-- 
cgit v1.2.3


From bf299275882624b1908521ee8074df85160e9679 Mon Sep 17 00:00:00 2001
From: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Date: Mon, 14 Apr 2008 00:04:51 -0700
Subject: [SKB]: __skb_queue_after(prev) = __skb_insert(prev, prev->next)

By reordering, __skb_queue_after() is expressed in terms of __skb_insert().

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 34 ++++++++++++----------------------
 1 file changed, 12 insertions(+), 22 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index c2116200580a..bb107ab675fc 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -663,11 +663,21 @@ static inline void skb_queue_head_init_class(struct sk_buff_head *list,
 }
 
 /*
- *	Insert an sk_buff at the start of a list.
+ *	Insert an sk_buff on a list.
  *
  *	The "__skb_xxxx()" functions are the non-atomic ones that
  *	can only be called with interrupts disabled.
  */
+extern void        skb_insert(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list);
+static inline void __skb_insert(struct sk_buff *newsk,
+				struct sk_buff *prev, struct sk_buff *next,
+				struct sk_buff_head *list)
+{
+	newsk->next = next;
+	newsk->prev = prev;
+	next->prev  = prev->next = newsk;
+	list->qlen++;
+}
 
 /**
  *	__skb_queue_after - queue a buffer at the list head
@@ -684,13 +694,7 @@ static inline void __skb_queue_after(struct sk_buff_head *list,
 				     struct sk_buff *prev,
 				     struct sk_buff *newsk)
 {
-	struct sk_buff *next;
-	list->qlen++;
-
-	next = prev->next;
-	newsk->next = next;
-	newsk->prev = prev;
-	next->prev  = prev->next = newsk;
+	__skb_insert(newsk, prev, prev->next, list);
 }
 
 /**
@@ -734,20 +738,6 @@ static inline void __skb_queue_tail(struct sk_buff_head *list,
 	next->prev  = prev->next = newsk;
 }
 
-/*
- *	Insert a packet on a list.
- */
-extern void        skb_insert(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list);
-static inline void __skb_insert(struct sk_buff *newsk,
-				struct sk_buff *prev, struct sk_buff *next,
-				struct sk_buff_head *list)
-{
-	newsk->next = next;
-	newsk->prev = prev;
-	next->prev  = prev->next = newsk;
-	list->qlen++;
-}
-
 /*
  *	Place a packet after a given packet in a list.
  */
-- 
cgit v1.2.3


From 7de6c033367ab86f39c7723392caf73325cbf286 Mon Sep 17 00:00:00 2001
From: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Date: Mon, 14 Apr 2008 00:05:09 -0700
Subject: [SKB]: __skb_append = __skb_queue_after

This expresses __skb_append in terms of __skb_queue_after, exploiting that

  __skb_append(old, new, list) = __skb_queue_after(list, old, new).

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 12 +++---------
 include/net/tcp.h      |  2 +-
 net/core/skbuff.c      |  2 +-
 net/ipv4/tcp_input.c   |  2 +-
 4 files changed, 6 insertions(+), 12 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index bb107ab675fc..83c851846829 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -697,6 +697,9 @@ static inline void __skb_queue_after(struct sk_buff_head *list,
 	__skb_insert(newsk, prev, prev->next, list);
 }
 
+extern void skb_append(struct sk_buff *old, struct sk_buff *newsk,
+		       struct sk_buff_head *list);
+
 /**
  *	__skb_queue_head - queue a buffer at the list head
  *	@list: list to use
@@ -738,15 +741,6 @@ static inline void __skb_queue_tail(struct sk_buff_head *list,
 	next->prev  = prev->next = newsk;
 }
 
-/*
- *	Place a packet after a given packet in a list.
- */
-extern void	   skb_append(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list);
-static inline void __skb_append(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list)
-{
-	__skb_insert(newsk, old, old->next, list);
-}
-
 /*
  * remove sk_buff from list. _Must_ be called atomically, and with
  * the list known..
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 58d82822414d..2ab350eca02e 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1247,7 +1247,7 @@ static inline void tcp_insert_write_queue_after(struct sk_buff *skb,
 						struct sk_buff *buff,
 						struct sock *sk)
 {
-	__skb_append(skb, buff, &sk->sk_write_queue);
+	__skb_queue_after(&sk->sk_write_queue, skb, buff);
 }
 
 /* Insert skb between prev and next on the write queue of sk.  */
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index e4259215607f..4cd12d99b12e 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -1860,7 +1860,7 @@ void skb_append(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head
 	unsigned long flags;
 
 	spin_lock_irqsave(&list->lock, flags);
-	__skb_append(old, newsk, list);
+	__skb_queue_after(list, old, newsk);
 	spin_unlock_irqrestore(&list->lock, flags);
 }
 
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 6e46b4c0f28c..743611956045 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -3968,7 +3968,7 @@ drop:
 		u32 end_seq = TCP_SKB_CB(skb)->end_seq;
 
 		if (seq == TCP_SKB_CB(skb1)->end_seq) {
-			__skb_append(skb1, skb, &tp->out_of_order_queue);
+			__skb_queue_after(&tp->out_of_order_queue, skb1, skb);
 
 			if (!tp->rx_opt.num_sacks ||
 			    tp->selective_acks[0].end_seq != seq)
-- 
cgit v1.2.3


From f5572855ec492334d8c3ec0e0e86c31865d5cf07 Mon Sep 17 00:00:00 2001
From: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Date: Mon, 14 Apr 2008 00:05:28 -0700
Subject: [SKB]: __skb_queue_tail = __skb_insert before

This expresses __skb_queue_tail() in terms of __skb_insert(),
using __skb_insert_before() as auxiliary function.

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 83c851846829..11fd9f2c4093 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -700,6 +700,13 @@ static inline void __skb_queue_after(struct sk_buff_head *list,
 extern void skb_append(struct sk_buff *old, struct sk_buff *newsk,
 		       struct sk_buff_head *list);
 
+static inline void __skb_queue_before(struct sk_buff_head *list,
+				      struct sk_buff *next,
+				      struct sk_buff *newsk)
+{
+	__skb_insert(newsk, next->prev, next, list);
+}
+
 /**
  *	__skb_queue_head - queue a buffer at the list head
  *	@list: list to use
@@ -731,14 +738,7 @@ extern void skb_queue_tail(struct sk_buff_head *list, struct sk_buff *newsk);
 static inline void __skb_queue_tail(struct sk_buff_head *list,
 				   struct sk_buff *newsk)
 {
-	struct sk_buff *prev, *next;
-
-	list->qlen++;
-	next = (struct sk_buff *)list;
-	prev = next->prev;
-	newsk->next = next;
-	newsk->prev = prev;
-	next->prev  = prev->next = newsk;
+	__skb_queue_before(list, (struct sk_buff *)list, newsk);
 }
 
 /*
-- 
cgit v1.2.3


From 666953df353194bef76086fa3f126241cbac3e3a Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@sw.ru>
Date: Mon, 14 Apr 2008 09:56:02 +0200
Subject: [NETFILTER]: ip_tables: per-netns FILTER/MANGLE/RAW tables for real

Commit 9335f047fe61587ec82ff12fbb1220bcfdd32006 aka
"[NETFILTER]: ip_tables: per-netns FILTER, MANGLE, RAW"
added per-netns _view_ of iptables rules. They were shown to user, but
ignored by filtering code. Now that it's possible to at least ping loopback,
per-netns tables can affect filtering decisions.

netns is taken in case of
	PRE_ROUTING, LOCAL_IN -- from in device,
	POST_ROUTING, LOCAL_OUT -- from out device,
	FORWARD -- from in device which should be equal to out device's netns.
		   This code is relatively new, so BUG_ON was plugged.

Wrappers were added to a) keep code the same from CONFIG_NET_NS=n users
(overwhelming majority), b) consolidate code in one place -- similar
changes will be done in ipv6 and arp netfilter code.

Signed-off-by: Alexey Dobriyan <adobriyan@sw.ru>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter.h           | 54 ++++++++++++++++++++++++++++++++++++-
 net/ipv4/netfilter/iptable_filter.c | 19 ++++++++++---
 net/ipv4/netfilter/iptable_mangle.c | 49 ++++++++++++++++++++++++++++-----
 net/ipv4/netfilter/iptable_raw.c    |  6 +++--
 4 files changed, 115 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h
index 89e6c72ad295..66bc52060fd6 100644
--- a/include/linux/netfilter.h
+++ b/include/linux/netfilter.h
@@ -6,11 +6,13 @@
 #include <linux/types.h>
 #include <linux/skbuff.h>
 #include <linux/net.h>
+#include <linux/netdevice.h>
 #include <linux/if.h>
 #include <linux/in.h>
 #include <linux/in6.h>
 #include <linux/wait.h>
 #include <linux/list.h>
+#include <net/net_namespace.h>
 #endif
 #include <linux/compiler.h>
 
@@ -76,7 +78,6 @@ extern void netfilter_init(void);
 #define NF_MAX_HOOKS 8
 
 struct sk_buff;
-struct net_device;
 
 typedef unsigned int nf_hookfn(unsigned int hooknum,
 			       struct sk_buff *skb,
@@ -320,5 +321,56 @@ extern void (*nf_ct_destroy)(struct nf_conntrack *);
 static inline void nf_ct_attach(struct sk_buff *new, struct sk_buff *skb) {}
 #endif
 
+static inline struct net *nf_pre_routing_net(const struct net_device *in,
+					     const struct net_device *out)
+{
+#ifdef CONFIG_NET_NS
+	return in->nd_net;
+#else
+	return &init_net;
+#endif
+}
+
+static inline struct net *nf_local_in_net(const struct net_device *in,
+					  const struct net_device *out)
+{
+#ifdef CONFIG_NET_NS
+	return in->nd_net;
+#else
+	return &init_net;
+#endif
+}
+
+static inline struct net *nf_forward_net(const struct net_device *in,
+					 const struct net_device *out)
+{
+#ifdef CONFIG_NET_NS
+	BUG_ON(in->nd_net != out->nd_net);
+	return in->nd_net;
+#else
+	return &init_net;
+#endif
+}
+
+static inline struct net *nf_local_out_net(const struct net_device *in,
+					   const struct net_device *out)
+{
+#ifdef CONFIG_NET_NS
+	return out->nd_net;
+#else
+	return &init_net;
+#endif
+}
+
+static inline struct net *nf_post_routing_net(const struct net_device *in,
+					      const struct net_device *out)
+{
+#ifdef CONFIG_NET_NS
+	return out->nd_net;
+#else
+	return &init_net;
+#endif
+}
+
 #endif /*__KERNEL__*/
 #endif /*__LINUX_NETFILTER_H*/
diff --git a/net/ipv4/netfilter/iptable_filter.c b/net/ipv4/netfilter/iptable_filter.c
index 69f3d7e6e96f..7fcf60adbbed 100644
--- a/net/ipv4/netfilter/iptable_filter.c
+++ b/net/ipv4/netfilter/iptable_filter.c
@@ -62,6 +62,17 @@ static struct xt_table packet_filter = {
 };
 
 /* The work comes in here from netfilter.c. */
+static unsigned int
+ipt_local_in_hook(unsigned int hook,
+		  struct sk_buff *skb,
+		  const struct net_device *in,
+		  const struct net_device *out,
+		  int (*okfn)(struct sk_buff *))
+{
+	return ipt_do_table(skb, hook, in, out,
+			    nf_local_in_net(in, out)->ipv4.iptable_filter);
+}
+
 static unsigned int
 ipt_hook(unsigned int hook,
 	 struct sk_buff *skb,
@@ -69,7 +80,8 @@ ipt_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ipt_do_table(skb, hook, in, out, init_net.ipv4.iptable_filter);
+	return ipt_do_table(skb, hook, in, out,
+			    nf_forward_net(in, out)->ipv4.iptable_filter);
 }
 
 static unsigned int
@@ -88,12 +100,13 @@ ipt_local_out_hook(unsigned int hook,
 		return NF_ACCEPT;
 	}
 
-	return ipt_do_table(skb, hook, in, out, init_net.ipv4.iptable_filter);
+	return ipt_do_table(skb, hook, in, out,
+			    nf_local_out_net(in, out)->ipv4.iptable_filter);
 }
 
 static struct nf_hook_ops ipt_ops[] __read_mostly = {
 	{
-		.hook		= ipt_hook,
+		.hook		= ipt_local_in_hook,
 		.owner		= THIS_MODULE,
 		.pf		= PF_INET,
 		.hooknum	= NF_INET_LOCAL_IN,
diff --git a/net/ipv4/netfilter/iptable_mangle.c b/net/ipv4/netfilter/iptable_mangle.c
index c55a210853a7..ba827035b691 100644
--- a/net/ipv4/netfilter/iptable_mangle.c
+++ b/net/ipv4/netfilter/iptable_mangle.c
@@ -74,13 +74,47 @@ static struct xt_table packet_mangler = {
 
 /* The work comes in here from netfilter.c. */
 static unsigned int
-ipt_route_hook(unsigned int hook,
+ipt_pre_routing_hook(unsigned int hook,
+		     struct sk_buff *skb,
+		     const struct net_device *in,
+		     const struct net_device *out,
+		     int (*okfn)(struct sk_buff *))
+{
+	return ipt_do_table(skb, hook, in, out,
+			    nf_pre_routing_net(in, out)->ipv4.iptable_mangle);
+}
+
+static unsigned int
+ipt_post_routing_hook(unsigned int hook,
+		      struct sk_buff *skb,
+		      const struct net_device *in,
+		      const struct net_device *out,
+		      int (*okfn)(struct sk_buff *))
+{
+	return ipt_do_table(skb, hook, in, out,
+			    nf_post_routing_net(in, out)->ipv4.iptable_mangle);
+}
+
+static unsigned int
+ipt_local_in_hook(unsigned int hook,
+		  struct sk_buff *skb,
+		  const struct net_device *in,
+		  const struct net_device *out,
+		  int (*okfn)(struct sk_buff *))
+{
+	return ipt_do_table(skb, hook, in, out,
+			    nf_local_in_net(in, out)->ipv4.iptable_mangle);
+}
+
+static unsigned int
+ipt_forward_hook(unsigned int hook,
 	 struct sk_buff *skb,
 	 const struct net_device *in,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ipt_do_table(skb, hook, in, out, init_net.ipv4.iptable_mangle);
+	return ipt_do_table(skb, hook, in, out,
+			    nf_forward_net(in, out)->ipv4.iptable_mangle);
 }
 
 static unsigned int
@@ -112,7 +146,8 @@ ipt_local_hook(unsigned int hook,
 	daddr = iph->daddr;
 	tos = iph->tos;
 
-	ret = ipt_do_table(skb, hook, in, out, init_net.ipv4.iptable_mangle);
+	ret = ipt_do_table(skb, hook, in, out,
+			   nf_local_out_net(in, out)->ipv4.iptable_mangle);
 	/* Reroute for ANY change. */
 	if (ret != NF_DROP && ret != NF_STOLEN && ret != NF_QUEUE) {
 		iph = ip_hdr(skb);
@@ -130,21 +165,21 @@ ipt_local_hook(unsigned int hook,
 
 static struct nf_hook_ops ipt_ops[] __read_mostly = {
 	{
-		.hook		= ipt_route_hook,
+		.hook		= ipt_pre_routing_hook,
 		.owner		= THIS_MODULE,
 		.pf		= PF_INET,
 		.hooknum	= NF_INET_PRE_ROUTING,
 		.priority	= NF_IP_PRI_MANGLE,
 	},
 	{
-		.hook		= ipt_route_hook,
+		.hook		= ipt_local_in_hook,
 		.owner		= THIS_MODULE,
 		.pf		= PF_INET,
 		.hooknum	= NF_INET_LOCAL_IN,
 		.priority	= NF_IP_PRI_MANGLE,
 	},
 	{
-		.hook		= ipt_route_hook,
+		.hook		= ipt_forward_hook,
 		.owner		= THIS_MODULE,
 		.pf		= PF_INET,
 		.hooknum	= NF_INET_FORWARD,
@@ -158,7 +193,7 @@ static struct nf_hook_ops ipt_ops[] __read_mostly = {
 		.priority	= NF_IP_PRI_MANGLE,
 	},
 	{
-		.hook		= ipt_route_hook,
+		.hook		= ipt_post_routing_hook,
 		.owner		= THIS_MODULE,
 		.pf		= PF_INET,
 		.hooknum	= NF_INET_POST_ROUTING,
diff --git a/net/ipv4/netfilter/iptable_raw.c b/net/ipv4/netfilter/iptable_raw.c
index e41fe8ca4e1c..4b689742d58b 100644
--- a/net/ipv4/netfilter/iptable_raw.c
+++ b/net/ipv4/netfilter/iptable_raw.c
@@ -52,7 +52,8 @@ ipt_hook(unsigned int hook,
 	 const struct net_device *out,
 	 int (*okfn)(struct sk_buff *))
 {
-	return ipt_do_table(skb, hook, in, out, init_net.ipv4.iptable_raw);
+	return ipt_do_table(skb, hook, in, out,
+			    nf_pre_routing_net(in, out)->ipv4.iptable_raw);
 }
 
 static unsigned int
@@ -70,7 +71,8 @@ ipt_local_hook(unsigned int hook,
 			       "packet.\n");
 		return NF_ACCEPT;
 	}
-	return ipt_do_table(skb, hook, in, out, init_net.ipv4.iptable_raw);
+	return ipt_do_table(skb, hook, in, out,
+			    nf_local_out_net(in, out)->ipv4.iptable_raw);
 }
 
 /* 'raw' is the very first table. */
-- 
cgit v1.2.3


From b9f61b160336da5eaaacb0cb41ebe32169e3bde5 Mon Sep 17 00:00:00 2001
From: Jan Engelhardt <jengelh@computergmbh.de>
Date: Mon, 14 Apr 2008 09:56:04 +0200
Subject: [NETFILTER]: xt_sctp: simplify xt_sctp.h

The use of xt_sctp.h flagged up -Wshadow warnings in userspace, which
prompted me to look at it and clean it up. Basic operations have been
directly replaced by library calls (memcpy, memset is both available
in the kernel and userspace, and usually faster than a self-made
loop). The is_set and is_clear functions now use a processing time
shortcut, too.

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter/xt_sctp.h | 84 ++++++++++++++++-----------------------
 1 file changed, 35 insertions(+), 49 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/xt_sctp.h b/include/linux/netfilter/xt_sctp.h
index dd5a4fd4cfd3..32000ba6ecef 100644
--- a/include/linux/netfilter/xt_sctp.h
+++ b/include/linux/netfilter/xt_sctp.h
@@ -37,68 +37,54 @@ struct xt_sctp_info {
 
 #define SCTP_CHUNKMAP_SET(chunkmap, type) 		\
 	do { 						\
-		chunkmap[type / bytes(u_int32_t)] |= 	\
+		(chunkmap)[type / bytes(u_int32_t)] |= 	\
 			1 << (type % bytes(u_int32_t));	\
 	} while (0)
 
 #define SCTP_CHUNKMAP_CLEAR(chunkmap, type)		 	\
 	do {							\
-		chunkmap[type / bytes(u_int32_t)] &= 		\
+		(chunkmap)[type / bytes(u_int32_t)] &= 		\
 			~(1 << (type % bytes(u_int32_t)));	\
 	} while (0)
 
 #define SCTP_CHUNKMAP_IS_SET(chunkmap, type) 			\
 ({								\
-	(chunkmap[type / bytes (u_int32_t)] & 			\
+	((chunkmap)[type / bytes (u_int32_t)] & 		\
 		(1 << (type % bytes (u_int32_t)))) ? 1: 0;	\
 })
 
-#define SCTP_CHUNKMAP_RESET(chunkmap) 				\
-	do {							\
-		int i; 						\
-		for (i = 0; i < ARRAY_SIZE(chunkmap); i++)	\
-			chunkmap[i] = 0;			\
-	} while (0)
-
-#define SCTP_CHUNKMAP_SET_ALL(chunkmap) 			\
-	do {							\
-		int i; 						\
-		for (i = 0; i < ARRAY_SIZE(chunkmap); i++) 	\
-			chunkmap[i] = ~0;			\
-	} while (0)
-
-#define SCTP_CHUNKMAP_COPY(destmap, srcmap) 			\
-	do {							\
-		int i; 						\
-		for (i = 0; i < ARRAY_SIZE(srcmap); i++) 	\
-			destmap[i] = srcmap[i];			\
-	} while (0)
-
-#define SCTP_CHUNKMAP_IS_CLEAR(chunkmap) 		\
-({							\
-	int i; 						\
-	int flag = 1;					\
-	for (i = 0; i < ARRAY_SIZE(chunkmap); i++) {	\
-		if (chunkmap[i]) {			\
-			flag = 0;			\
-			break;				\
-		}					\
-	}						\
-        flag;						\
-})
-
-#define SCTP_CHUNKMAP_IS_ALL_SET(chunkmap) 		\
-({							\
-	int i; 						\
-	int flag = 1;					\
-	for (i = 0; i < ARRAY_SIZE(chunkmap); i++) {	\
-		if (chunkmap[i] != ~0) {		\
-			flag = 0;			\
-				break;			\
-		}					\
-	}						\
-        flag;						\
-})
+#define SCTP_CHUNKMAP_RESET(chunkmap) \
+	memset((chunkmap), 0, sizeof(chunkmap))
+
+#define SCTP_CHUNKMAP_SET_ALL(chunkmap) \
+	memset((chunkmap), ~0U, sizeof(chunkmap))
+
+#define SCTP_CHUNKMAP_COPY(destmap, srcmap) \
+	memcpy((destmap), (srcmap), sizeof(srcmap))
+
+#define SCTP_CHUNKMAP_IS_CLEAR(chunkmap) \
+	__sctp_chunkmap_is_clear((chunkmap), ARRAY_SIZE(chunkmap))
+static inline bool
+__sctp_chunkmap_is_clear(const u_int32_t *chunkmap, unsigned int n)
+{
+	unsigned int i;
+	for (i = 0; i < n; ++i)
+		if (chunkmap[i])
+			return false;
+	return true;
+}
+
+#define SCTP_CHUNKMAP_IS_ALL_SET(chunkmap) \
+	__sctp_chunkmap_is_all_set((chunkmap), ARRAY_SIZE(chunkmap))
+static inline bool
+__sctp_chunkmap_is_all_set(const u_int32_t *chunkmap, unsigned int n)
+{
+	unsigned int i;
+	for (i = 0; i < n; ++i)
+		if (chunkmap[i] != ~0U)
+			return false;
+	return true;
+}
 
 #endif /* _XT_SCTP_H_ */
 
-- 
cgit v1.2.3


From 5452e425adfdfc4647b618e303f73d48f2405b0e Mon Sep 17 00:00:00 2001
From: Jan Engelhardt <jengelh@computergmbh.de>
Date: Mon, 14 Apr 2008 11:15:35 +0200
Subject: [NETFILTER]: annotate {arp,ip,ip6,x}tables with const

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter/x_tables.h |  4 ++--
 net/ipv4/netfilter/arp_tables.c    | 31 ++++++++++++++++---------------
 net/ipv4/netfilter/arpt_mangle.c   |  2 +-
 net/ipv4/netfilter/ip_tables.c     | 31 ++++++++++++++++---------------
 net/ipv6/netfilter/ip6_tables.c    | 29 +++++++++++++++--------------
 net/netfilter/x_tables.c           | 18 +++++++++---------
 6 files changed, 59 insertions(+), 56 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h
index b2c62cc618f5..2326296b6f25 100644
--- a/include/linux/netfilter/x_tables.h
+++ b/include/linux/netfilter/x_tables.h
@@ -430,13 +430,13 @@ extern int xt_compat_add_offset(int af, unsigned int offset, short delta);
 extern void xt_compat_flush_offsets(int af);
 extern short xt_compat_calc_jump(int af, unsigned int offset);
 
-extern int xt_compat_match_offset(struct xt_match *match);
+extern int xt_compat_match_offset(const struct xt_match *match);
 extern int xt_compat_match_from_user(struct xt_entry_match *m,
 				     void **dstptr, unsigned int *size);
 extern int xt_compat_match_to_user(struct xt_entry_match *m,
 				   void __user **dstptr, unsigned int *size);
 
-extern int xt_compat_target_offset(struct xt_target *target);
+extern int xt_compat_target_offset(const struct xt_target *target);
 extern void xt_compat_target_from_user(struct xt_entry_target *t,
 				       void **dstptr, unsigned int *size);
 extern int xt_compat_target_to_user(struct xt_entry_target *t,
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index 1563f29b5117..10cc442330c3 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -59,7 +59,7 @@ do {								\
 #endif
 
 static inline int arp_devaddr_compare(const struct arpt_devaddr_info *ap,
-				      char *hdr_addr, int len)
+				      const char *hdr_addr, int len)
 {
 	int i, ret;
 
@@ -80,8 +80,8 @@ static inline int arp_packet_match(const struct arphdr *arphdr,
 				   const char *outdev,
 				   const struct arpt_arp *arpinfo)
 {
-	char *arpptr = (char *)(arphdr + 1);
-	char *src_devaddr, *tgt_devaddr;
+	const char *arpptr = (char *)(arphdr + 1);
+	const char *src_devaddr, *tgt_devaddr;
 	__be32 src_ipaddr, tgt_ipaddr;
 	int i, ret;
 
@@ -226,12 +226,12 @@ unsigned int arpt_do_table(struct sk_buff *skb,
 {
 	static const char nulldevname[IFNAMSIZ];
 	unsigned int verdict = NF_DROP;
-	struct arphdr *arp;
+	const struct arphdr *arp;
 	bool hotdrop = false;
 	struct arpt_entry *e, *back;
 	const char *indev, *outdev;
 	void *table_base;
-	struct xt_table_info *private;
+	const struct xt_table_info *private;
 
 	if (!pskb_may_pull(skb, arp_hdr_len(skb->dev)))
 		return NF_DROP;
@@ -352,7 +352,7 @@ static int mark_source_chains(struct xt_table_info *newinfo,
 		e->counters.pcnt = pos;
 
 		for (;;) {
-			struct arpt_standard_target *t
+			const struct arpt_standard_target *t
 				= (void *)arpt_get_target(e);
 			int visited = e->comefrom & (1 << hook);
 
@@ -437,7 +437,7 @@ static int mark_source_chains(struct xt_table_info *newinfo,
 
 static inline int check_entry(struct arpt_entry *e, const char *name)
 {
-	struct arpt_entry_target *t;
+	const struct arpt_entry_target *t;
 
 	if (!arp_checkentry(&e->arp)) {
 		duprintf("arp_tables: arp check failed %p %s.\n", e, name);
@@ -710,7 +710,7 @@ static inline struct xt_counters *alloc_counters(struct arpt_table *table)
 {
 	unsigned int countersize;
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 
 	/* We need atomic snapshot of counters: rest doesn't change
 	 * (other than comefrom, which userspace doesn't care
@@ -737,7 +737,7 @@ static int copy_entries_to_user(unsigned int total_size,
 	unsigned int off, num;
 	struct arpt_entry *e;
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 	int ret = 0;
 	void *loc_cpu_entry;
 
@@ -872,7 +872,7 @@ static int get_info(struct net *net, void __user *user, int *len, int compat)
 				    "arptable_%s", name);
 	if (t && !IS_ERR(t)) {
 		struct arpt_getinfo info;
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
 
 #ifdef CONFIG_COMPAT
 		if (compat) {
@@ -927,7 +927,8 @@ static int get_entries(struct net *net, struct arpt_get_entries __user *uptr,
 
 	t = xt_find_table_lock(net, NF_ARP, get.name);
 	if (t && !IS_ERR(t)) {
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
+
 		duprintf("t->private->number = %u\n",
 			 private->number);
 		if (get.size == private->size)
@@ -1087,11 +1088,11 @@ static int do_add_counters(struct net *net, void __user *user, unsigned int len,
 	struct xt_counters_info tmp;
 	struct xt_counters *paddc;
 	unsigned int num_counters;
-	char *name;
+	const char *name;
 	int size;
 	void *ptmp;
 	struct arpt_table *t;
-	struct xt_table_info *private;
+	const struct xt_table_info *private;
 	int ret = 0;
 	void *loc_cpu_entry;
 #ifdef CONFIG_COMPAT
@@ -1558,7 +1559,7 @@ static int compat_copy_entries_to_user(unsigned int total_size,
 				       void __user *userptr)
 {
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 	void __user *pos;
 	unsigned int size;
 	int ret = 0;
@@ -1609,7 +1610,7 @@ static int compat_get_entries(struct net *net,
 	xt_compat_lock(NF_ARP);
 	t = xt_find_table_lock(net, NF_ARP, get.name);
 	if (t && !IS_ERR(t)) {
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
 		struct xt_table_info info;
 
 		duprintf("t->private->number = %u\n", private->number);
diff --git a/net/ipv4/netfilter/arpt_mangle.c b/net/ipv4/netfilter/arpt_mangle.c
index 3f4222b0a803..3e732c827fc2 100644
--- a/net/ipv4/netfilter/arpt_mangle.c
+++ b/net/ipv4/netfilter/arpt_mangle.c
@@ -15,7 +15,7 @@ target(struct sk_buff *skb,
        const void *targinfo)
 {
 	const struct arpt_mangle *mangle = targinfo;
-	struct arphdr *arp;
+	const struct arphdr *arp;
 	unsigned char *arpptr;
 	int pln, hln;
 
diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index a819d191e1aa..aa124b50cb4a 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -296,7 +296,7 @@ static void trace_packet(struct sk_buff *skb,
 			 struct ipt_entry *e)
 {
 	void *table_base;
-	struct ipt_entry *root;
+	const struct ipt_entry *root;
 	char *hookname, *chainname, *comment;
 	unsigned int rulenum = 0;
 
@@ -327,7 +327,7 @@ ipt_do_table(struct sk_buff *skb,
 {
 	static const char nulldevname[IFNAMSIZ] __attribute__((aligned(sizeof(long))));
 	u_int16_t offset;
-	struct iphdr *ip;
+	const struct iphdr *ip;
 	u_int16_t datalen;
 	bool hotdrop = false;
 	/* Initializing verdict to NF_DROP keeps gcc happy. */
@@ -926,7 +926,7 @@ static struct xt_counters * alloc_counters(struct xt_table *table)
 {
 	unsigned int countersize;
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 
 	/* We need atomic snapshot of counters: rest doesn't change
 	   (other than comefrom, which userspace doesn't care
@@ -953,9 +953,9 @@ copy_entries_to_user(unsigned int total_size,
 	unsigned int off, num;
 	struct ipt_entry *e;
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 	int ret = 0;
-	void *loc_cpu_entry;
+	const void *loc_cpu_entry;
 
 	counters = alloc_counters(table);
 	if (IS_ERR(counters))
@@ -975,8 +975,8 @@ copy_entries_to_user(unsigned int total_size,
 	/* ... then go back and fix counters and names */
 	for (off = 0, num = 0; off < total_size; off += e->next_offset, num++){
 		unsigned int i;
-		struct ipt_entry_match *m;
-		struct ipt_entry_target *t;
+		const struct ipt_entry_match *m;
+		const struct ipt_entry_target *t;
 
 		e = (struct ipt_entry *)(loc_cpu_entry + off);
 		if (copy_to_user(userptr + off
@@ -1116,7 +1116,7 @@ static int get_info(struct net *net, void __user *user, int *len, int compat)
 				    "iptable_%s", name);
 	if (t && !IS_ERR(t)) {
 		struct ipt_getinfo info;
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
 
 #ifdef CONFIG_COMPAT
 		if (compat) {
@@ -1172,7 +1172,7 @@ get_entries(struct net *net, struct ipt_get_entries __user *uptr, int *len)
 
 	t = xt_find_table_lock(net, AF_INET, get.name);
 	if (t && !IS_ERR(t)) {
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
 		duprintf("t->private->number = %u\n", private->number);
 		if (get.size == private->size)
 			ret = copy_entries_to_user(private->size,
@@ -1337,11 +1337,11 @@ do_add_counters(struct net *net, void __user *user, unsigned int len, int compat
 	struct xt_counters_info tmp;
 	struct xt_counters *paddc;
 	unsigned int num_counters;
-	char *name;
+	const char *name;
 	int size;
 	void *ptmp;
 	struct xt_table *t;
-	struct xt_table_info *private;
+	const struct xt_table_info *private;
 	int ret = 0;
 	void *loc_cpu_entry;
 #ifdef CONFIG_COMPAT
@@ -1878,11 +1878,11 @@ compat_copy_entries_to_user(unsigned int total_size, struct xt_table *table,
 			    void __user *userptr)
 {
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 	void __user *pos;
 	unsigned int size;
 	int ret = 0;
-	void *loc_cpu_entry;
+	const void *loc_cpu_entry;
 	unsigned int i = 0;
 
 	counters = alloc_counters(table);
@@ -1929,7 +1929,7 @@ compat_get_entries(struct net *net, struct compat_ipt_get_entries __user *uptr,
 	xt_compat_lock(AF_INET);
 	t = xt_find_table_lock(net, AF_INET, get.name);
 	if (t && !IS_ERR(t)) {
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
 		struct xt_table_info info;
 		duprintf("t->private->number = %u\n", private->number);
 		ret = compat_table_info(private, &info);
@@ -2130,7 +2130,8 @@ icmp_match(const struct sk_buff *skb,
 	   unsigned int protoff,
 	   bool *hotdrop)
 {
-	struct icmphdr _icmph, *ic;
+	const struct icmphdr *ic;
+	struct icmphdr _icmph;
 	const struct ipt_icmp *icmpinfo = matchinfo;
 
 	/* Must not be a fragment. */
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index 70ef0d276cc0..782183f63366 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -325,7 +325,7 @@ static void trace_packet(struct sk_buff *skb,
 			 struct ip6t_entry *e)
 {
 	void *table_base;
-	struct ip6t_entry *root;
+	const struct ip6t_entry *root;
 	char *hookname, *chainname, *comment;
 	unsigned int rulenum = 0;
 
@@ -952,7 +952,7 @@ static struct xt_counters *alloc_counters(struct xt_table *table)
 {
 	unsigned int countersize;
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 
 	/* We need atomic snapshot of counters: rest doesn't change
 	   (other than comefrom, which userspace doesn't care
@@ -979,9 +979,9 @@ copy_entries_to_user(unsigned int total_size,
 	unsigned int off, num;
 	struct ip6t_entry *e;
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 	int ret = 0;
-	void *loc_cpu_entry;
+	const void *loc_cpu_entry;
 
 	counters = alloc_counters(table);
 	if (IS_ERR(counters))
@@ -1001,8 +1001,8 @@ copy_entries_to_user(unsigned int total_size,
 	/* ... then go back and fix counters and names */
 	for (off = 0, num = 0; off < total_size; off += e->next_offset, num++){
 		unsigned int i;
-		struct ip6t_entry_match *m;
-		struct ip6t_entry_target *t;
+		const struct ip6t_entry_match *m;
+		const struct ip6t_entry_target *t;
 
 		e = (struct ip6t_entry *)(loc_cpu_entry + off);
 		if (copy_to_user(userptr + off
@@ -1142,7 +1142,7 @@ static int get_info(struct net *net, void __user *user, int *len, int compat)
 				    "ip6table_%s", name);
 	if (t && !IS_ERR(t)) {
 		struct ip6t_getinfo info;
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
 
 #ifdef CONFIG_COMPAT
 		if (compat) {
@@ -1225,7 +1225,7 @@ __do_replace(struct net *net, const char *name, unsigned int valid_hooks,
 	struct xt_table *t;
 	struct xt_table_info *oldinfo;
 	struct xt_counters *counters;
-	void *loc_cpu_old_entry;
+	const void *loc_cpu_old_entry;
 
 	ret = 0;
 	counters = vmalloc_node(num_counters * sizeof(struct xt_counters),
@@ -1369,9 +1369,9 @@ do_add_counters(struct net *net, void __user *user, unsigned int len,
 	int size;
 	void *ptmp;
 	struct xt_table *t;
-	struct xt_table_info *private;
+	const struct xt_table_info *private;
 	int ret = 0;
-	void *loc_cpu_entry;
+	const void *loc_cpu_entry;
 #ifdef CONFIG_COMPAT
 	struct compat_xt_counters_info compat_tmp;
 
@@ -1905,11 +1905,11 @@ compat_copy_entries_to_user(unsigned int total_size, struct xt_table *table,
 			    void __user *userptr)
 {
 	struct xt_counters *counters;
-	struct xt_table_info *private = table->private;
+	const struct xt_table_info *private = table->private;
 	void __user *pos;
 	unsigned int size;
 	int ret = 0;
-	void *loc_cpu_entry;
+	const void *loc_cpu_entry;
 	unsigned int i = 0;
 
 	counters = alloc_counters(table);
@@ -1956,7 +1956,7 @@ compat_get_entries(struct net *net, struct compat_ip6t_get_entries __user *uptr,
 	xt_compat_lock(AF_INET6);
 	t = xt_find_table_lock(net, AF_INET6, get.name);
 	if (t && !IS_ERR(t)) {
-		struct xt_table_info *private = t->private;
+		const struct xt_table_info *private = t->private;
 		struct xt_table_info info;
 		duprintf("t->private->number = %u\n", private->number);
 		ret = compat_table_info(private, &info);
@@ -2155,7 +2155,8 @@ icmp6_match(const struct sk_buff *skb,
 	   unsigned int protoff,
 	   bool *hotdrop)
 {
-	struct icmp6hdr _icmph, *ic;
+	const struct icmp6hdr *ic;
+	struct icmp6hdr _icmph;
 	const struct ip6t_icmp *icmpinfo = matchinfo;
 
 	/* Must not be a fragment. */
diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index 0bd95680a494..f52f7f810ac4 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -58,7 +58,7 @@ static struct xt_af *xt;
 #define duprintf(format, args...)
 #endif
 
-static const char *xt_prefix[NPROTO] = {
+static const char *const xt_prefix[NPROTO] = {
 	[AF_INET]	= "ip",
 	[AF_INET6]	= "ip6",
 	[NF_ARP]	= "arp",
@@ -248,7 +248,7 @@ EXPORT_SYMBOL_GPL(xt_request_find_target);
 
 static int match_revfn(int af, const char *name, u8 revision, int *bestp)
 {
-	struct xt_match *m;
+	const struct xt_match *m;
 	int have_rev = 0;
 
 	list_for_each_entry(m, &xt[af].match, list) {
@@ -264,7 +264,7 @@ static int match_revfn(int af, const char *name, u8 revision, int *bestp)
 
 static int target_revfn(int af, const char *name, u8 revision, int *bestp)
 {
-	struct xt_target *t;
+	const struct xt_target *t;
 	int have_rev = 0;
 
 	list_for_each_entry(t, &xt[af].target, list) {
@@ -385,7 +385,7 @@ short xt_compat_calc_jump(int af, unsigned int offset)
 }
 EXPORT_SYMBOL_GPL(xt_compat_calc_jump);
 
-int xt_compat_match_offset(struct xt_match *match)
+int xt_compat_match_offset(const struct xt_match *match)
 {
 	u_int16_t csize = match->compatsize ? : match->matchsize;
 	return XT_ALIGN(match->matchsize) - COMPAT_XT_ALIGN(csize);
@@ -395,7 +395,7 @@ EXPORT_SYMBOL_GPL(xt_compat_match_offset);
 int xt_compat_match_from_user(struct xt_entry_match *m, void **dstptr,
 			      unsigned int *size)
 {
-	struct xt_match *match = m->u.kernel.match;
+	const struct xt_match *match = m->u.kernel.match;
 	struct compat_xt_entry_match *cm = (struct compat_xt_entry_match *)m;
 	int pad, off = xt_compat_match_offset(match);
 	u_int16_t msize = cm->u.user.match_size;
@@ -422,7 +422,7 @@ EXPORT_SYMBOL_GPL(xt_compat_match_from_user);
 int xt_compat_match_to_user(struct xt_entry_match *m, void __user **dstptr,
 			    unsigned int *size)
 {
-	struct xt_match *match = m->u.kernel.match;
+	const struct xt_match *match = m->u.kernel.match;
 	struct compat_xt_entry_match __user *cm = *dstptr;
 	int off = xt_compat_match_offset(match);
 	u_int16_t msize = m->u.user.match_size - off;
@@ -479,7 +479,7 @@ int xt_check_target(const struct xt_target *target, unsigned short family,
 EXPORT_SYMBOL_GPL(xt_check_target);
 
 #ifdef CONFIG_COMPAT
-int xt_compat_target_offset(struct xt_target *target)
+int xt_compat_target_offset(const struct xt_target *target)
 {
 	u_int16_t csize = target->compatsize ? : target->targetsize;
 	return XT_ALIGN(target->targetsize) - COMPAT_XT_ALIGN(csize);
@@ -489,7 +489,7 @@ EXPORT_SYMBOL_GPL(xt_compat_target_offset);
 void xt_compat_target_from_user(struct xt_entry_target *t, void **dstptr,
 				unsigned int *size)
 {
-	struct xt_target *target = t->u.kernel.target;
+	const struct xt_target *target = t->u.kernel.target;
 	struct compat_xt_entry_target *ct = (struct compat_xt_entry_target *)t;
 	int pad, off = xt_compat_target_offset(target);
 	u_int16_t tsize = ct->u.user.target_size;
@@ -515,7 +515,7 @@ EXPORT_SYMBOL_GPL(xt_compat_target_from_user);
 int xt_compat_target_to_user(struct xt_entry_target *t, void __user **dstptr,
 			     unsigned int *size)
 {
-	struct xt_target *target = t->u.kernel.target;
+	const struct xt_target *target = t->u.kernel.target;
 	struct compat_xt_entry_target __user *ct = *dstptr;
 	int off = xt_compat_target_offset(target);
 	u_int16_t tsize = t->u.user.target_size - off;
-- 
cgit v1.2.3


From 4abff0775d5e4feb20b21371e1c63a1b30fc2140 Mon Sep 17 00:00:00 2001
From: Jan Engelhardt <jengelh@computergmbh.de>
Date: Mon, 14 Apr 2008 11:15:43 +0200
Subject: [NETFILTER]: remove arpt_table indirection macro

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter_arp/arp_tables.h | 11 +++++------
 net/ipv4/netfilter/arp_tables.c          | 27 +++++++++++++--------------
 net/ipv4/netfilter/arptable_filter.c     |  2 +-
 3 files changed, 19 insertions(+), 21 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter_arp/arp_tables.h b/include/linux/netfilter_arp/arp_tables.h
index db223ca92c8b..102c4134a713 100644
--- a/include/linux/netfilter_arp/arp_tables.h
+++ b/include/linux/netfilter_arp/arp_tables.h
@@ -24,7 +24,6 @@
 #define ARPT_FUNCTION_MAXNAMELEN XT_FUNCTION_MAXNAMELEN
 #define ARPT_TABLE_MAXNAMELEN XT_TABLE_MAXNAMELEN
 #define arpt_target xt_target
-#define arpt_table xt_table
 
 #define ARPT_DEV_ADDR_LEN_MAX 16
 
@@ -271,15 +270,15 @@ struct arpt_error
  	xt_register_target(tgt); })
 #define arpt_unregister_target(tgt) xt_unregister_target(tgt)
 
-extern struct arpt_table *arpt_register_table(struct net *net,
-					      struct arpt_table *table,
-					      const struct arpt_replace *repl);
-extern void arpt_unregister_table(struct arpt_table *table);
+extern struct xt_table *arpt_register_table(struct net *net,
+					    struct xt_table *table,
+					    const struct arpt_replace *repl);
+extern void arpt_unregister_table(struct xt_table *table);
 extern unsigned int arpt_do_table(struct sk_buff *skb,
 				  unsigned int hook,
 				  const struct net_device *in,
 				  const struct net_device *out,
-				  struct arpt_table *table);
+				  struct xt_table *table);
 
 #define ARPT_ALIGN(s) XT_ALIGN(s)
 
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index 10cc442330c3..34c42c831b18 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -222,7 +222,7 @@ unsigned int arpt_do_table(struct sk_buff *skb,
 			   unsigned int hook,
 			   const struct net_device *in,
 			   const struct net_device *out,
-			   struct arpt_table *table)
+			   struct xt_table *table)
 {
 	static const char nulldevname[IFNAMSIZ];
 	unsigned int verdict = NF_DROP;
@@ -706,7 +706,7 @@ static void get_counters(const struct xt_table_info *t,
 	}
 }
 
-static inline struct xt_counters *alloc_counters(struct arpt_table *table)
+static inline struct xt_counters *alloc_counters(struct xt_table *table)
 {
 	unsigned int countersize;
 	struct xt_counters *counters;
@@ -731,13 +731,13 @@ static inline struct xt_counters *alloc_counters(struct arpt_table *table)
 }
 
 static int copy_entries_to_user(unsigned int total_size,
-				struct arpt_table *table,
+				struct xt_table *table,
 				void __user *userptr)
 {
 	unsigned int off, num;
 	struct arpt_entry *e;
 	struct xt_counters *counters;
-	const struct xt_table_info *private = table->private;
+	struct xt_table_info *private = table->private;
 	int ret = 0;
 	void *loc_cpu_entry;
 
@@ -851,7 +851,7 @@ static int compat_table_info(const struct xt_table_info *info,
 static int get_info(struct net *net, void __user *user, int *len, int compat)
 {
 	char name[ARPT_TABLE_MAXNAMELEN];
-	struct arpt_table *t;
+	struct xt_table *t;
 	int ret;
 
 	if (*len != sizeof(struct arpt_getinfo)) {
@@ -911,7 +911,7 @@ static int get_entries(struct net *net, struct arpt_get_entries __user *uptr,
 {
 	int ret;
 	struct arpt_get_entries get;
-	struct arpt_table *t;
+	struct xt_table *t;
 
 	if (*len < sizeof(get)) {
 		duprintf("get_entries: %u < %Zu\n", *len, sizeof(get));
@@ -954,7 +954,7 @@ static int __do_replace(struct net *net, const char *name,
 			void __user *counters_ptr)
 {
 	int ret;
-	struct arpt_table *t;
+	struct xt_table *t;
 	struct xt_table_info *oldinfo;
 	struct xt_counters *counters;
 	void *loc_cpu_old_entry;
@@ -1091,7 +1091,7 @@ static int do_add_counters(struct net *net, void __user *user, unsigned int len,
 	const char *name;
 	int size;
 	void *ptmp;
-	struct arpt_table *t;
+	struct xt_table *t;
 	const struct xt_table_info *private;
 	int ret = 0;
 	void *loc_cpu_entry;
@@ -1555,7 +1555,7 @@ out:
 }
 
 static int compat_copy_entries_to_user(unsigned int total_size,
-				       struct arpt_table *table,
+				       struct xt_table *table,
 				       void __user *userptr)
 {
 	struct xt_counters *counters;
@@ -1593,7 +1593,7 @@ static int compat_get_entries(struct net *net,
 {
 	int ret;
 	struct compat_arpt_get_entries get;
-	struct arpt_table *t;
+	struct xt_table *t;
 
 	if (*len < sizeof(get)) {
 		duprintf("compat_get_entries: %u < %zu\n", *len, sizeof(get));
@@ -1723,9 +1723,8 @@ static int do_arpt_get_ctl(struct sock *sk, int cmd, void __user *user, int *len
 	return ret;
 }
 
-struct arpt_table *arpt_register_table(struct net *net,
-				       struct arpt_table *table,
-				       const struct arpt_replace *repl)
+struct xt_table *arpt_register_table(struct net *net, struct xt_table *table,
+				     const struct arpt_replace *repl)
 {
 	int ret;
 	struct xt_table_info *newinfo;
@@ -1767,7 +1766,7 @@ out:
 	return ERR_PTR(ret);
 }
 
-void arpt_unregister_table(struct arpt_table *table)
+void arpt_unregister_table(struct xt_table *table)
 {
 	struct xt_table_info *private;
 	void *loc_cpu_entry;
diff --git a/net/ipv4/netfilter/arptable_filter.c b/net/ipv4/netfilter/arptable_filter.c
index 629e4951a9b1..9f6526c87757 100644
--- a/net/ipv4/netfilter/arptable_filter.c
+++ b/net/ipv4/netfilter/arptable_filter.c
@@ -45,7 +45,7 @@ static struct
 	.term = ARPT_ERROR_INIT,
 };
 
-static struct arpt_table packet_filter = {
+static struct xt_table packet_filter = {
 	.name		= "filter",
 	.valid_hooks	= FILTER_VALID_HOOKS,
 	.lock		= __RW_LOCK_UNLOCKED(packet_filter.lock),
-- 
cgit v1.2.3


From 95eea855af69bfd54a7b73546190e76046ca2e07 Mon Sep 17 00:00:00 2001
From: Jan Engelhardt <jengelh@computergmbh.de>
Date: Mon, 14 Apr 2008 11:15:43 +0200
Subject: [NETFILTER]: remove arpt_target indirection macro

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter_arp/arp_tables.h | 1 -
 net/ipv4/netfilter/arp_tables.c          | 8 ++++----
 net/ipv4/netfilter/arpt_mangle.c         | 2 +-
 3 files changed, 5 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter_arp/arp_tables.h b/include/linux/netfilter_arp/arp_tables.h
index 102c4134a713..782b83e5bdb9 100644
--- a/include/linux/netfilter_arp/arp_tables.h
+++ b/include/linux/netfilter_arp/arp_tables.h
@@ -23,7 +23,6 @@
 
 #define ARPT_FUNCTION_MAXNAMELEN XT_FUNCTION_MAXNAMELEN
 #define ARPT_TABLE_MAXNAMELEN XT_TABLE_MAXNAMELEN
-#define arpt_target xt_target
 
 #define ARPT_DEV_ADDR_LEN_MAX 16
 
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index 34c42c831b18..d55f3b42eba5 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -457,7 +457,7 @@ static inline int check_entry(struct arpt_entry *e, const char *name)
 static inline int check_target(struct arpt_entry *e, const char *name)
 {
 	struct arpt_entry_target *t;
-	struct arpt_target *target;
+	struct xt_target *target;
 	int ret;
 
 	t = arpt_get_target(e);
@@ -480,7 +480,7 @@ find_check_entry(struct arpt_entry *e, const char *name, unsigned int size,
 		 unsigned int *i)
 {
 	struct arpt_entry_target *t;
-	struct arpt_target *target;
+	struct xt_target *target;
 	int ret;
 
 	ret = check_entry(e, name);
@@ -1784,7 +1784,7 @@ void arpt_unregister_table(struct xt_table *table)
 }
 
 /* The built-in targets: standard (NULL) and error. */
-static struct arpt_target arpt_standard_target __read_mostly = {
+static struct xt_target arpt_standard_target __read_mostly = {
 	.name		= ARPT_STANDARD_TARGET,
 	.targetsize	= sizeof(int),
 	.family		= NF_ARP,
@@ -1795,7 +1795,7 @@ static struct arpt_target arpt_standard_target __read_mostly = {
 #endif
 };
 
-static struct arpt_target arpt_error_target __read_mostly = {
+static struct xt_target arpt_error_target __read_mostly = {
 	.name		= ARPT_ERROR_TARGET,
 	.target		= arpt_error,
 	.targetsize	= ARPT_FUNCTION_MAXNAMELEN,
diff --git a/net/ipv4/netfilter/arpt_mangle.c b/net/ipv4/netfilter/arpt_mangle.c
index 3e732c827fc2..f9c102ab891b 100644
--- a/net/ipv4/netfilter/arpt_mangle.c
+++ b/net/ipv4/netfilter/arpt_mangle.c
@@ -73,7 +73,7 @@ checkentry(const char *tablename, const void *e, const struct xt_target *target,
 	return true;
 }
 
-static struct arpt_target arpt_mangle_reg __read_mostly = {
+static struct xt_target arpt_mangle_reg __read_mostly = {
 	.name		= "mangle",
 	.target		= target,
 	.targetsize	= sizeof(struct arpt_mangle),
-- 
cgit v1.2.3


From 3bb0362d2f53fa54a17b88c96b43fc093e47699b Mon Sep 17 00:00:00 2001
From: Jan Engelhardt <jengelh@computergmbh.de>
Date: Mon, 14 Apr 2008 11:15:44 +0200
Subject: [NETFILTER]: remove arpt_(un)register_target indirection macros

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter_arp/arp_tables.h | 5 -----
 net/ipv4/netfilter/arpt_mangle.c         | 8 +++-----
 2 files changed, 3 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter_arp/arp_tables.h b/include/linux/netfilter_arp/arp_tables.h
index 782b83e5bdb9..dd9c97f2d436 100644
--- a/include/linux/netfilter_arp/arp_tables.h
+++ b/include/linux/netfilter_arp/arp_tables.h
@@ -264,11 +264,6 @@ struct arpt_error
 	.target.errorname = "ERROR",					       \
 }
 
-#define arpt_register_target(tgt) 	\
-({	(tgt)->family = NF_ARP;		\
- 	xt_register_target(tgt); })
-#define arpt_unregister_target(tgt) xt_unregister_target(tgt)
-
 extern struct xt_table *arpt_register_table(struct net *net,
 					    struct xt_table *table,
 					    const struct arpt_replace *repl);
diff --git a/net/ipv4/netfilter/arpt_mangle.c b/net/ipv4/netfilter/arpt_mangle.c
index f9c102ab891b..a385959d2655 100644
--- a/net/ipv4/netfilter/arpt_mangle.c
+++ b/net/ipv4/netfilter/arpt_mangle.c
@@ -75,6 +75,7 @@ checkentry(const char *tablename, const void *e, const struct xt_target *target,
 
 static struct xt_target arpt_mangle_reg __read_mostly = {
 	.name		= "mangle",
+	.family		= NF_ARP,
 	.target		= target,
 	.targetsize	= sizeof(struct arpt_mangle),
 	.checkentry	= checkentry,
@@ -83,15 +84,12 @@ static struct xt_target arpt_mangle_reg __read_mostly = {
 
 static int __init arpt_mangle_init(void)
 {
-	if (arpt_register_target(&arpt_mangle_reg))
-		return -EINVAL;
-
-	return 0;
+	return xt_register_target(&arpt_mangle_reg);
 }
 
 static void __exit arpt_mangle_fini(void)
 {
-	arpt_unregister_target(&arpt_mangle_reg);
+	xt_unregister_target(&arpt_mangle_reg);
 }
 
 module_init(arpt_mangle_init);
-- 
cgit v1.2.3


From d63a650736f566a1f9e9434725d2089597c0d2cc Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 20 Mar 2008 15:15:53 +0100
Subject: [NETFILTER]: Add partial checksum validation helper

Move the UDP-Lite conntrack checksum validation to a generic helper
similar to nf_checksum() and make it fall back to nf_checksum()
in case the full packet is to be checksummed and hardware checksums
are available. This is to be used by DCCP conntrack, which also
needs to verify partial checksums.

Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter.h                  | 22 ++++++++++++++++
 net/ipv4/netfilter.c                       | 37 +++++++++++++++++++++-----
 net/ipv6/netfilter.c                       | 42 +++++++++++++++++++++++++-----
 net/netfilter/nf_conntrack_proto_udplite.c | 33 +++++------------------
 4 files changed, 94 insertions(+), 40 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h
index 66bc52060fd6..e4c66593b5c6 100644
--- a/include/linux/netfilter.h
+++ b/include/linux/netfilter.h
@@ -234,6 +234,11 @@ struct nf_afinfo {
 	unsigned short	family;
 	__sum16		(*checksum)(struct sk_buff *skb, unsigned int hook,
 				    unsigned int dataoff, u_int8_t protocol);
+	__sum16		(*checksum_partial)(struct sk_buff *skb,
+					    unsigned int hook,
+					    unsigned int dataoff,
+					    unsigned int len,
+					    u_int8_t protocol);
 	int		(*route)(struct dst_entry **dst, struct flowi *fl);
 	void		(*saveroute)(const struct sk_buff *skb,
 				     struct nf_queue_entry *entry);
@@ -263,6 +268,23 @@ nf_checksum(struct sk_buff *skb, unsigned int hook, unsigned int dataoff,
 	return csum;
 }
 
+static inline __sum16
+nf_checksum_partial(struct sk_buff *skb, unsigned int hook,
+		    unsigned int dataoff, unsigned int len,
+		    u_int8_t protocol, unsigned short family)
+{
+	const struct nf_afinfo *afinfo;
+	__sum16 csum = 0;
+
+	rcu_read_lock();
+	afinfo = nf_get_afinfo(family);
+	if (afinfo)
+		csum = afinfo->checksum_partial(skb, hook, dataoff, len,
+						protocol);
+	rcu_read_unlock();
+	return csum;
+}
+
 extern int nf_register_afinfo(const struct nf_afinfo *afinfo);
 extern void nf_unregister_afinfo(const struct nf_afinfo *afinfo);
 
diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c
index 9a904c6c0dc8..f8edacdf991d 100644
--- a/net/ipv4/netfilter.c
+++ b/net/ipv4/netfilter.c
@@ -182,21 +182,44 @@ __sum16 nf_ip_checksum(struct sk_buff *skb, unsigned int hook,
 	}
 	return csum;
 }
-
 EXPORT_SYMBOL(nf_ip_checksum);
 
+static __sum16 nf_ip_checksum_partial(struct sk_buff *skb, unsigned int hook,
+				      unsigned int dataoff, unsigned int len,
+				      u_int8_t protocol)
+{
+	const struct iphdr *iph = ip_hdr(skb);
+	__sum16 csum = 0;
+
+	switch (skb->ip_summed) {
+	case CHECKSUM_COMPLETE:
+		if (len == skb->len - dataoff)
+			return nf_ip_checksum(skb, hook, dataoff, protocol);
+		/* fall through */
+	case CHECKSUM_NONE:
+		skb->csum = csum_tcpudp_nofold(iph->saddr, iph->daddr, protocol,
+					       skb->len - dataoff, 0);
+		skb->ip_summed = CHECKSUM_NONE;
+		csum = __skb_checksum_complete_head(skb, dataoff + len);
+		if (!csum)
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+	}
+	return csum;
+}
+
 static int nf_ip_route(struct dst_entry **dst, struct flowi *fl)
 {
 	return ip_route_output_key(&init_net, (struct rtable **)dst, fl);
 }
 
 static const struct nf_afinfo nf_ip_afinfo = {
-	.family		= AF_INET,
-	.checksum	= nf_ip_checksum,
-	.route		= nf_ip_route,
-	.saveroute	= nf_ip_saveroute,
-	.reroute	= nf_ip_reroute,
-	.route_key_size	= sizeof(struct ip_rt_info),
+	.family			= AF_INET,
+	.checksum		= nf_ip_checksum,
+	.checksum_partial	= nf_ip_checksum_partial,
+	.route			= nf_ip_route,
+	.saveroute		= nf_ip_saveroute,
+	.reroute		= nf_ip_reroute,
+	.route_key_size		= sizeof(struct ip_rt_info),
 };
 
 static int ipv4_netfilter_init(void)
diff --git a/net/ipv6/netfilter.c b/net/ipv6/netfilter.c
index aed51bcc66b4..8c6c5e71f210 100644
--- a/net/ipv6/netfilter.c
+++ b/net/ipv6/netfilter.c
@@ -121,16 +121,44 @@ __sum16 nf_ip6_checksum(struct sk_buff *skb, unsigned int hook,
 	}
 	return csum;
 }
-
 EXPORT_SYMBOL(nf_ip6_checksum);
 
+static __sum16 nf_ip6_checksum_partial(struct sk_buff *skb, unsigned int hook,
+				       unsigned int dataoff, unsigned int len,
+				       u_int8_t protocol)
+{
+	struct ipv6hdr *ip6h = ipv6_hdr(skb);
+	__wsum hsum;
+	__sum16 csum = 0;
+
+	switch (skb->ip_summed) {
+	case CHECKSUM_COMPLETE:
+		if (len == skb->len - dataoff)
+			return nf_ip6_checksum(skb, hook, dataoff, protocol);
+		/* fall through */
+	case CHECKSUM_NONE:
+		hsum = skb_checksum(skb, 0, dataoff, 0);
+		skb->csum = ~csum_unfold(csum_ipv6_magic(&ip6h->saddr,
+							 &ip6h->daddr,
+							 skb->len - dataoff,
+							 protocol,
+							 csum_sub(0, hsum)));
+		skb->ip_summed = CHECKSUM_NONE;
+		csum = __skb_checksum_complete_head(skb, dataoff + len);
+		if (!csum)
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+	}
+	return csum;
+};
+
 static const struct nf_afinfo nf_ip6_afinfo = {
-	.family		= AF_INET6,
-	.checksum	= nf_ip6_checksum,
-	.route		= nf_ip6_route,
-	.saveroute	= nf_ip6_saveroute,
-	.reroute	= nf_ip6_reroute,
-	.route_key_size	= sizeof(struct ip6_rt_info),
+	.family			= AF_INET6,
+	.checksum		= nf_ip6_checksum,
+	.checksum_partial	= nf_ip6_checksum_partial,
+	.route			= nf_ip6_route,
+	.saveroute		= nf_ip6_saveroute,
+	.reroute		= nf_ip6_reroute,
+	.route_key_size		= sizeof(struct ip6_rt_info),
 };
 
 int __init ipv6_netfilter_init(void)
diff --git a/net/netfilter/nf_conntrack_proto_udplite.c b/net/netfilter/nf_conntrack_proto_udplite.c
index 9dd03c7aeac6..c3eaee6afffd 100644
--- a/net/netfilter/nf_conntrack_proto_udplite.c
+++ b/net/netfilter/nf_conntrack_proto_udplite.c
@@ -127,32 +127,13 @@ static int udplite_error(struct sk_buff *skb, unsigned int dataoff,
 	}
 
 	/* Checksum invalid? Ignore. */
-	if (nf_conntrack_checksum && !skb_csum_unnecessary(skb) &&
-	    hooknum == NF_INET_PRE_ROUTING) {
-		if (pf == PF_INET) {
-			struct iphdr *iph = ip_hdr(skb);
-
-			skb->csum = csum_tcpudp_nofold(iph->saddr, iph->daddr,
-						       udplen, IPPROTO_UDPLITE, 0);
-		} else {
-			struct ipv6hdr *ipv6h = ipv6_hdr(skb);
-			__wsum hsum = skb_checksum(skb, 0, dataoff, 0);
-
-			skb->csum = ~csum_unfold(
-				csum_ipv6_magic(&ipv6h->saddr, &ipv6h->daddr,
-						udplen, IPPROTO_UDPLITE,
-						csum_sub(0, hsum)));
-		}
-
-		skb->ip_summed = CHECKSUM_NONE;
-		if (__skb_checksum_complete_head(skb, dataoff + cscov)) {
-			if (LOG_INVALID(IPPROTO_UDPLITE))
-				nf_log_packet(pf, 0, skb, NULL, NULL, NULL,
-					      "nf_ct_udplite: bad UDPLite "
-					      "checksum ");
-			return -NF_ACCEPT;
-		}
-		skb->ip_summed = CHECKSUM_UNNECESSARY;
+	if (nf_conntrack_checksum && hooknum == NF_INET_PRE_ROUTING &&
+	    nf_checksum_partial(skb, hooknum, dataoff, cscov, IPPROTO_UDP,
+	    			pf)) {
+		if (LOG_INVALID(IPPROTO_UDPLITE))
+			nf_log_packet(pf, 0, skb, NULL, NULL, NULL,
+				      "nf_ct_udplite: bad UDPLite checksum ");
+		return -NF_ACCEPT;
 	}
 
 	return NF_ACCEPT;
-- 
cgit v1.2.3


From 2bc780499aa33311ec0f3e42624dfaa7be0ade5e Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 20 Mar 2008 15:15:55 +0100
Subject: [NETFILTER]: nf_conntrack: add DCCP protocol support

Add DCCP conntrack helper. Thanks to Gerrit Renker <gerrit@erg.abdn.ac.uk>
for review and testing.

Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter/nf_conntrack_dccp.h   |  40 ++
 include/linux/netfilter/nfnetlink_conntrack.h |   8 +
 include/net/netfilter/nf_conntrack.h          |   2 +
 include/net/netfilter/nf_conntrack_tuple.h    |   6 +
 net/netfilter/Kconfig                         |  10 +
 net/netfilter/Makefile                        |   1 +
 net/netfilter/nf_conntrack_proto_dccp.c       | 816 ++++++++++++++++++++++++++
 7 files changed, 883 insertions(+)
 create mode 100644 include/linux/netfilter/nf_conntrack_dccp.h
 create mode 100644 net/netfilter/nf_conntrack_proto_dccp.c

(limited to 'include/linux')

diff --git a/include/linux/netfilter/nf_conntrack_dccp.h b/include/linux/netfilter/nf_conntrack_dccp.h
new file mode 100644
index 000000000000..40dcc82058d1
--- /dev/null
+++ b/include/linux/netfilter/nf_conntrack_dccp.h
@@ -0,0 +1,40 @@
+#ifndef _NF_CONNTRACK_DCCP_H
+#define _NF_CONNTRACK_DCCP_H
+
+/* Exposed to userspace over nfnetlink */
+enum ct_dccp_states {
+	CT_DCCP_NONE,
+	CT_DCCP_REQUEST,
+	CT_DCCP_RESPOND,
+	CT_DCCP_PARTOPEN,
+	CT_DCCP_OPEN,
+	CT_DCCP_CLOSEREQ,
+	CT_DCCP_CLOSING,
+	CT_DCCP_TIMEWAIT,
+	CT_DCCP_IGNORE,
+	CT_DCCP_INVALID,
+	__CT_DCCP_MAX
+};
+#define CT_DCCP_MAX		(__CT_DCCP_MAX - 1)
+
+enum ct_dccp_roles {
+	CT_DCCP_ROLE_CLIENT,
+	CT_DCCP_ROLE_SERVER,
+	__CT_DCCP_ROLE_MAX
+};
+#define CT_DCCP_ROLE_MAX	(__CT_DCCP_ROLE_MAX - 1)
+
+#ifdef __KERNEL__
+#include <net/netfilter/nf_conntrack_tuple.h>
+
+struct nf_ct_dccp {
+	u_int8_t	role[IP_CT_DIR_MAX];
+	u_int8_t	state;
+	u_int8_t	last_pkt;
+	u_int8_t	last_dir;
+	u_int64_t	handshake_seq;
+};
+
+#endif /* __KERNEL__ */
+
+#endif /* _NF_CONNTRACK_DCCP_H */
diff --git a/include/linux/netfilter/nfnetlink_conntrack.h b/include/linux/netfilter/nfnetlink_conntrack.h
index e3e1533aba2d..0a383ac083cb 100644
--- a/include/linux/netfilter/nfnetlink_conntrack.h
+++ b/include/linux/netfilter/nfnetlink_conntrack.h
@@ -80,6 +80,7 @@ enum ctattr_l4proto {
 enum ctattr_protoinfo {
 	CTA_PROTOINFO_UNSPEC,
 	CTA_PROTOINFO_TCP,
+	CTA_PROTOINFO_DCCP,
 	__CTA_PROTOINFO_MAX
 };
 #define CTA_PROTOINFO_MAX (__CTA_PROTOINFO_MAX - 1)
@@ -95,6 +96,13 @@ enum ctattr_protoinfo_tcp {
 };
 #define CTA_PROTOINFO_TCP_MAX (__CTA_PROTOINFO_TCP_MAX - 1)
 
+enum ctattr_protoinfo_dccp {
+	CTA_PROTOINFO_DCCP_UNSPEC,
+	CTA_PROTOINFO_DCCP_STATE,
+	__CTA_PROTOINFO_DCCP_MAX,
+};
+#define CTA_PROTOINFO_DCCP_MAX (__CTA_PROTOINFO_DCCP_MAX - 1)
+
 enum ctattr_counters {
 	CTA_COUNTERS_UNSPEC,
 	CTA_COUNTERS_PACKETS,		/* old 64bit counters */
diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
index a3567a7a6d67..bb9fc852e973 100644
--- a/include/net/netfilter/nf_conntrack.h
+++ b/include/net/netfilter/nf_conntrack.h
@@ -20,6 +20,7 @@
 #include <asm/atomic.h>
 
 #include <linux/netfilter/nf_conntrack_tcp.h>
+#include <linux/netfilter/nf_conntrack_dccp.h>
 #include <linux/netfilter/nf_conntrack_sctp.h>
 #include <linux/netfilter/nf_conntrack_proto_gre.h>
 #include <net/netfilter/ipv4/nf_conntrack_icmp.h>
@@ -30,6 +31,7 @@
 /* per conntrack: protocol private data */
 union nf_conntrack_proto {
 	/* insert conntrack proto private data here */
+	struct nf_ct_dccp dccp;
 	struct ip_ct_sctp sctp;
 	struct ip_ct_tcp tcp;
 	struct ip_ct_icmp icmp;
diff --git a/include/net/netfilter/nf_conntrack_tuple.h b/include/net/netfilter/nf_conntrack_tuple.h
index 168c91754d89..bdeec3461384 100644
--- a/include/net/netfilter/nf_conntrack_tuple.h
+++ b/include/net/netfilter/nf_conntrack_tuple.h
@@ -39,6 +39,9 @@ union nf_conntrack_man_proto
 	struct {
 		__be16 id;
 	} icmp;
+	struct {
+		__be16 port;
+	} dccp;
 	struct {
 		__be16 port;
 	} sctp;
@@ -77,6 +80,9 @@ struct nf_conntrack_tuple
 			struct {
 				u_int8_t type, code;
 			} icmp;
+			struct {
+				__be16 port;
+			} dccp;
 			struct {
 				__be16 port;
 			} sctp;
diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig
index daf5b881064d..c1fc0f1a641c 100644
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -86,6 +86,16 @@ config NF_CONNTRACK_EVENTS
 
 	  If unsure, say `N'.
 
+config NF_CT_PROTO_DCCP
+	tristate 'DCCP protocol connection tracking support (EXPERIMENTAL)'
+	depends on EXPERIMENTAL && NF_CONNTRACK
+	depends on NETFILTER_ADVANCED
+	help
+	  With this option enabled, the layer 3 independent connection
+	  tracking code will be able to do state tracking on DCCP connections.
+
+	  If unsure, say 'N'.
+
 config NF_CT_PROTO_GRE
 	tristate
 	depends on NF_CONNTRACK
diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile
index ea7508387f95..5c4b183f6422 100644
--- a/net/netfilter/Makefile
+++ b/net/netfilter/Makefile
@@ -13,6 +13,7 @@ obj-$(CONFIG_NETFILTER_NETLINK_LOG) += nfnetlink_log.o
 obj-$(CONFIG_NF_CONNTRACK) += nf_conntrack.o
 
 # SCTP protocol connection tracking
+obj-$(CONFIG_NF_CT_PROTO_DCCP) += nf_conntrack_proto_dccp.o
 obj-$(CONFIG_NF_CT_PROTO_GRE) += nf_conntrack_proto_gre.o
 obj-$(CONFIG_NF_CT_PROTO_SCTP) += nf_conntrack_proto_sctp.o
 obj-$(CONFIG_NF_CT_PROTO_UDPLITE) += nf_conntrack_proto_udplite.o
diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c
new file mode 100644
index 000000000000..db88c5bcc5fd
--- /dev/null
+++ b/net/netfilter/nf_conntrack_proto_dccp.c
@@ -0,0 +1,816 @@
+/*
+ * DCCP connection tracking protocol helper
+ *
+ * Copyright (c) 2005, 2006, 2008 Patrick McHardy <kaber@trash.net>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/sysctl.h>
+#include <linux/spinlock.h>
+#include <linux/skbuff.h>
+#include <linux/dccp.h>
+
+#include <linux/netfilter/nfnetlink_conntrack.h>
+#include <net/netfilter/nf_conntrack.h>
+#include <net/netfilter/nf_conntrack_l4proto.h>
+#include <net/netfilter/nf_log.h>
+
+static DEFINE_RWLOCK(dccp_lock);
+
+static int nf_ct_dccp_loose __read_mostly = 1;
+
+/* Timeouts are based on values from RFC4340:
+ *
+ * - REQUEST:
+ *
+ *   8.1.2. Client Request
+ *
+ *   A client MAY give up on its DCCP-Requests after some time
+ *   (3 minutes, for example).
+ *
+ * - RESPOND:
+ *
+ *   8.1.3. Server Response
+ *
+ *   It MAY also leave the RESPOND state for CLOSED after a timeout of
+ *   not less than 4MSL (8 minutes);
+ *
+ * - PARTOPEN:
+ *
+ *   8.1.5. Handshake Completion
+ *
+ *   If the client remains in PARTOPEN for more than 4MSL (8 minutes),
+ *   it SHOULD reset the connection with Reset Code 2, "Aborted".
+ *
+ * - OPEN:
+ *
+ *   The DCCP timestamp overflows after 11.9 hours. If the connection
+ *   stays idle this long the sequence number won't be recognized
+ *   as valid anymore.
+ *
+ * - CLOSEREQ/CLOSING:
+ *
+ *   8.3. Termination
+ *
+ *   The retransmission timer should initially be set to go off in two
+ *   round-trip times and should back off to not less than once every
+ *   64 seconds ...
+ *
+ * - TIMEWAIT:
+ *
+ *   4.3. States
+ *
+ *   A server or client socket remains in this state for 2MSL (4 minutes)
+ *   after the connection has been town down, ...
+ */
+
+#define DCCP_MSL (2 * 60 * HZ)
+
+static unsigned int dccp_timeout[CT_DCCP_MAX + 1] __read_mostly = {
+	[CT_DCCP_REQUEST]	= 2 * DCCP_MSL,
+	[CT_DCCP_RESPOND]	= 4 * DCCP_MSL,
+	[CT_DCCP_PARTOPEN]	= 4 * DCCP_MSL,
+	[CT_DCCP_OPEN]		= 12 * 3600 * HZ,
+	[CT_DCCP_CLOSEREQ]	= 64 * HZ,
+	[CT_DCCP_CLOSING]	= 64 * HZ,
+	[CT_DCCP_TIMEWAIT]	= 2 * DCCP_MSL,
+};
+
+static const char * const dccp_state_names[] = {
+	[CT_DCCP_NONE]		= "NONE",
+	[CT_DCCP_REQUEST]	= "REQUEST",
+	[CT_DCCP_RESPOND]	= "RESPOND",
+	[CT_DCCP_PARTOPEN]	= "PARTOPEN",
+	[CT_DCCP_OPEN]		= "OPEN",
+	[CT_DCCP_CLOSEREQ]	= "CLOSEREQ",
+	[CT_DCCP_CLOSING]	= "CLOSING",
+	[CT_DCCP_TIMEWAIT]	= "TIMEWAIT",
+	[CT_DCCP_IGNORE]	= "IGNORE",
+	[CT_DCCP_INVALID]	= "INVALID",
+};
+
+#define sNO	CT_DCCP_NONE
+#define sRQ	CT_DCCP_REQUEST
+#define sRS	CT_DCCP_RESPOND
+#define sPO	CT_DCCP_PARTOPEN
+#define sOP	CT_DCCP_OPEN
+#define sCR	CT_DCCP_CLOSEREQ
+#define sCG	CT_DCCP_CLOSING
+#define sTW	CT_DCCP_TIMEWAIT
+#define sIG	CT_DCCP_IGNORE
+#define sIV	CT_DCCP_INVALID
+
+/*
+ * DCCP state transistion table
+ *
+ * The assumption is the same as for TCP tracking:
+ *
+ * We are the man in the middle. All the packets go through us but might
+ * get lost in transit to the destination. It is assumed that the destination
+ * can't receive segments we haven't seen.
+ *
+ * The following states exist:
+ *
+ * NONE:	Initial state, expecting Request
+ * REQUEST:	Request seen, waiting for Response from server
+ * RESPOND:	Response from server seen, waiting for Ack from client
+ * PARTOPEN:	Ack after Response seen, waiting for packet other than Response,
+ * 		Reset or Sync from server
+ * OPEN:	Packet other than Response, Reset or Sync seen
+ * CLOSEREQ:	CloseReq from server seen, expecting Close from client
+ * CLOSING:	Close seen, expecting Reset
+ * TIMEWAIT:	Reset seen
+ * IGNORE:	Not determinable whether packet is valid
+ *
+ * Some states exist only on one side of the connection: REQUEST, RESPOND,
+ * PARTOPEN, CLOSEREQ. For the other side these states are equivalent to
+ * the one it was in before.
+ *
+ * Packets are marked as ignored (sIG) if we don't know if they're valid
+ * (for example a reincarnation of a connection we didn't notice is dead
+ * already) and the server may send back a connection closing Reset or a
+ * Response. They're also used for Sync/SyncAck packets, which we don't
+ * care about.
+ */
+static const u_int8_t
+dccp_state_table[CT_DCCP_ROLE_MAX + 1][DCCP_PKT_SYNCACK + 1][CT_DCCP_MAX + 1] = {
+	[CT_DCCP_ROLE_CLIENT] = {
+		[DCCP_PKT_REQUEST] = {
+		/*
+		 * sNO -> sRQ		Regular Request
+		 * sRQ -> sRQ		Retransmitted Request or reincarnation
+		 * sRS -> sRS		Retransmitted Request (apparently Response
+		 * 			got lost after we saw it) or reincarnation
+		 * sPO -> sIG		Ignore, conntrack might be out of sync
+		 * sOP -> sIG		Ignore, conntrack might be out of sync
+		 * sCR -> sIG		Ignore, conntrack might be out of sync
+		 * sCG -> sIG		Ignore, conntrack might be out of sync
+		 * sTW -> sRQ		Reincarnation
+		 *
+		 *	sNO, sRQ, sRS, sPO. sOP, sCR, sCG, sTW, */
+			sRQ, sRQ, sRS, sIG, sIG, sIG, sIG, sRQ,
+		},
+		[DCCP_PKT_RESPONSE] = {
+		/*
+		 * sNO -> sIV		Invalid
+		 * sRQ -> sIG		Ignore, might be response to ignored Request
+		 * sRS -> sIG		Ignore, might be response to ignored Request
+		 * sPO -> sIG		Ignore, might be response to ignored Request
+		 * sOP -> sIG		Ignore, might be response to ignored Request
+		 * sCR -> sIG		Ignore, might be response to ignored Request
+		 * sCG -> sIG		Ignore, might be response to ignored Request
+		 * sTW -> sIV		Invalid, reincarnation in reverse direction
+		 *			goes through sRQ
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIG, sIG, sIG, sIG, sIG, sIG, sIV,
+		},
+		[DCCP_PKT_ACK] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sPO		Ack for Response, move to PARTOPEN (8.1.5.)
+		 * sPO -> sPO		Retransmitted Ack for Response, remain in PARTOPEN
+		 * sOP -> sOP		Regular ACK, remain in OPEN
+		 * sCR -> sCR		Ack in CLOSEREQ MAY be processed (8.3.)
+		 * sCG -> sCG		Ack in CLOSING MAY be processed (8.3.)
+		 * sTW -> sIV
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sPO, sPO, sOP, sCR, sCG, sIV
+		},
+		[DCCP_PKT_DATA] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sIV		No connection
+		 * sPO -> sIV		MUST use DataAck in PARTOPEN state (8.1.5.)
+		 * sOP -> sOP		Regular Data packet
+		 * sCR -> sCR		Data in CLOSEREQ MAY be processed (8.3.)
+		 * sCG -> sCG		Data in CLOSING MAY be processed (8.3.)
+		 * sTW -> sIV
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sIV, sOP, sCR, sCG, sIV,
+		},
+		[DCCP_PKT_DATAACK] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sPO		Ack for Response, move to PARTOPEN (8.1.5.)
+		 * sPO -> sPO		Remain in PARTOPEN state
+		 * sOP -> sOP		Regular DataAck packet in OPEN state
+		 * sCR -> sCR		DataAck in CLOSEREQ MAY be processed (8.3.)
+		 * sCG -> sCG		DataAck in CLOSING MAY be processed (8.3.)
+		 * sTW -> sIV
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sPO, sPO, sOP, sCR, sCG, sIV
+		},
+		[DCCP_PKT_CLOSEREQ] = {
+		/*
+		 * CLOSEREQ may only be sent by the server.
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV
+		},
+		[DCCP_PKT_CLOSE] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sIV		No connection
+		 * sPO -> sCG		Client-initiated close
+		 * sOP -> sCG		Client-initiated close
+		 * sCR -> sCG		Close in response to CloseReq (8.3.)
+		 * sCG -> sCG		Retransmit
+		 * sTW -> sIV		Late retransmit, already in TIME_WAIT
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sCG, sCG, sCG, sIV, sIV
+		},
+		[DCCP_PKT_RESET] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sTW		Sync received or timeout, SHOULD send Reset (8.1.1.)
+		 * sRS -> sTW		Response received without Request
+		 * sPO -> sTW		Timeout, SHOULD send Reset (8.1.5.)
+		 * sOP -> sTW		Connection reset
+		 * sCR -> sTW		Connection reset
+		 * sCG -> sTW		Connection reset
+		 * sTW -> sIG		Ignore (don't refresh timer)
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sTW, sTW, sTW, sTW, sTW, sTW, sIG
+		},
+		[DCCP_PKT_SYNC] = {
+		/*
+		 * We currently ignore Sync packets
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIG, sIG, sIG, sIG, sIG, sIG, sIG, sIG,
+		},
+		[DCCP_PKT_SYNCACK] = {
+		/*
+		 * We currently ignore SyncAck packets
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIG, sIG, sIG, sIG, sIG, sIG, sIG, sIG,
+		},
+	},
+	[CT_DCCP_ROLE_SERVER] = {
+		[DCCP_PKT_REQUEST] = {
+		/*
+		 * sNO -> sIV		Invalid
+		 * sRQ -> sIG		Ignore, conntrack might be out of sync
+		 * sRS -> sIG		Ignore, conntrack might be out of sync
+		 * sPO -> sIG		Ignore, conntrack might be out of sync
+		 * sOP -> sIG		Ignore, conntrack might be out of sync
+		 * sCR -> sIG		Ignore, conntrack might be out of sync
+		 * sCG -> sIG		Ignore, conntrack might be out of sync
+		 * sTW -> sRQ		Reincarnation, must reverse roles
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIG, sIG, sIG, sIG, sIG, sIG, sRQ
+		},
+		[DCCP_PKT_RESPONSE] = {
+		/*
+		 * sNO -> sIV		Response without Request
+		 * sRQ -> sRS		Response to clients Request
+		 * sRS -> sRS		Retransmitted Response (8.1.3. SHOULD NOT)
+		 * sPO -> sIG		Response to an ignored Request or late retransmit
+		 * sOP -> sIG		Ignore, might be response to ignored Request
+		 * sCR -> sIG		Ignore, might be response to ignored Request
+		 * sCG -> sIG		Ignore, might be response to ignored Request
+		 * sTW -> sIV		Invalid, Request from client in sTW moves to sRQ
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sRS, sRS, sIG, sIG, sIG, sIG, sIV
+		},
+		[DCCP_PKT_ACK] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sIV		No connection
+		 * sPO -> sOP		Enter OPEN state (8.1.5.)
+		 * sOP -> sOP		Regular Ack in OPEN state
+		 * sCR -> sIV		Waiting for Close from client
+		 * sCG -> sCG		Ack in CLOSING MAY be processed (8.3.)
+		 * sTW -> sIV
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sOP, sOP, sIV, sCG, sIV
+		},
+		[DCCP_PKT_DATA] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sIV		No connection
+		 * sPO -> sOP		Enter OPEN state (8.1.5.)
+		 * sOP -> sOP		Regular Data packet in OPEN state
+		 * sCR -> sIV		Waiting for Close from client
+		 * sCG -> sCG		Data in CLOSING MAY be processed (8.3.)
+		 * sTW -> sIV
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sOP, sOP, sIV, sCG, sIV
+		},
+		[DCCP_PKT_DATAACK] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sIV		No connection
+		 * sPO -> sOP		Enter OPEN state (8.1.5.)
+		 * sOP -> sOP		Regular DataAck in OPEN state
+		 * sCR -> sIV		Waiting for Close from client
+		 * sCG -> sCG		Data in CLOSING MAY be processed (8.3.)
+		 * sTW -> sIV
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sOP, sOP, sIV, sCG, sIV
+		},
+		[DCCP_PKT_CLOSEREQ] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sIV		No connection
+		 * sPO -> sOP -> sCR	Move directly to CLOSEREQ (8.1.5.)
+		 * sOP -> sCR		CloseReq in OPEN state
+		 * sCR -> sCR		Retransmit
+		 * sCG -> sCR		Simultaneous close, client sends another Close
+		 * sTW -> sIV		Already closed
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sCR, sCR, sCR, sCR, sIV
+		},
+		[DCCP_PKT_CLOSE] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sIV		No connection
+		 * sRS -> sIV		No connection
+		 * sPO -> sOP -> sCG	Move direcly to CLOSING
+		 * sOP -> sCG		Move to CLOSING
+		 * sCR -> sIV		Close after CloseReq is invalid
+		 * sCG -> sCG		Retransmit
+		 * sTW -> sIV		Already closed
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIV, sIV, sIV, sCG, sCG, sIV, sCG, sIV
+		},
+		[DCCP_PKT_RESET] = {
+		/*
+		 * sNO -> sIV		No connection
+		 * sRQ -> sTW		Reset in response to Request
+		 * sRS -> sTW		Timeout, SHOULD send Reset (8.1.3.)
+		 * sPO -> sTW		Timeout, SHOULD send Reset (8.1.3.)
+		 * sOP -> sTW
+		 * sCR -> sTW
+		 * sCG -> sTW
+		 * sTW -> sIG		Ignore (don't refresh timer)
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW, sTW */
+			sIV, sTW, sTW, sTW, sTW, sTW, sTW, sTW, sIG
+		},
+		[DCCP_PKT_SYNC] = {
+		/*
+		 * We currently ignore Sync packets
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIG, sIG, sIG, sIG, sIG, sIG, sIG, sIG,
+		},
+		[DCCP_PKT_SYNCACK] = {
+		/*
+		 * We currently ignore SyncAck packets
+		 *
+		 *	sNO, sRQ, sRS, sPO, sOP, sCR, sCG, sTW */
+			sIG, sIG, sIG, sIG, sIG, sIG, sIG, sIG,
+		},
+	},
+};
+
+static int dccp_pkt_to_tuple(const struct sk_buff *skb, unsigned int dataoff,
+			     struct nf_conntrack_tuple *tuple)
+{
+	struct dccp_hdr _hdr, *dh;
+
+	dh = skb_header_pointer(skb, dataoff, sizeof(_hdr), &_hdr);
+	if (dh == NULL)
+		return 0;
+
+	tuple->src.u.dccp.port = dh->dccph_sport;
+	tuple->dst.u.dccp.port = dh->dccph_dport;
+	return 1;
+}
+
+static int dccp_invert_tuple(struct nf_conntrack_tuple *inv,
+			     const struct nf_conntrack_tuple *tuple)
+{
+	inv->src.u.dccp.port = tuple->dst.u.dccp.port;
+	inv->dst.u.dccp.port = tuple->src.u.dccp.port;
+	return 1;
+}
+
+static int dccp_new(struct nf_conn *ct, const struct sk_buff *skb,
+		    unsigned int dataoff)
+{
+	int pf = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num;
+	struct dccp_hdr _dh, *dh;
+	const char *msg;
+	u_int8_t state;
+
+	dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &dh);
+	BUG_ON(dh == NULL);
+
+	state = dccp_state_table[CT_DCCP_ROLE_CLIENT][dh->dccph_type][CT_DCCP_NONE];
+	switch (state) {
+	default:
+		if (nf_ct_dccp_loose == 0) {
+			msg = "nf_ct_dccp: not picking up existing connection ";
+			goto out_invalid;
+		}
+	case CT_DCCP_REQUEST:
+		break;
+	case CT_DCCP_INVALID:
+		msg = "nf_ct_dccp: invalid state transition ";
+		goto out_invalid;
+	}
+
+	ct->proto.dccp.role[IP_CT_DIR_ORIGINAL] = CT_DCCP_ROLE_CLIENT;
+	ct->proto.dccp.role[IP_CT_DIR_REPLY] = CT_DCCP_ROLE_SERVER;
+	ct->proto.dccp.state = CT_DCCP_NONE;
+	return 1;
+
+out_invalid:
+	if (LOG_INVALID(IPPROTO_DCCP))
+		nf_log_packet(pf, 0, skb, NULL, NULL, NULL, msg);
+	return 0;
+}
+
+static u64 dccp_ack_seq(const struct dccp_hdr *dh)
+{
+	const struct dccp_hdr_ack_bits *dhack;
+
+	dhack = (void *)dh + __dccp_basic_hdr_len(dh);
+	return ((u64)ntohs(dhack->dccph_ack_nr_high) << 32) +
+		     ntohl(dhack->dccph_ack_nr_low);
+}
+
+static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb,
+		       unsigned int dataoff, enum ip_conntrack_info ctinfo,
+		       int pf, unsigned int hooknum)
+{
+	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
+	struct dccp_hdr _dh, *dh;
+	u_int8_t type, old_state, new_state;
+	enum ct_dccp_roles role;
+
+	dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &dh);
+	BUG_ON(dh == NULL);
+	type = dh->dccph_type;
+
+	if (type == DCCP_PKT_RESET &&
+	    !test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
+		/* Tear down connection immediately if only reply is a RESET */
+		if (del_timer(&ct->timeout))
+			ct->timeout.function((unsigned long)ct);
+		return NF_ACCEPT;
+	}
+
+	write_lock_bh(&dccp_lock);
+
+	role = ct->proto.dccp.role[dir];
+	old_state = ct->proto.dccp.state;
+	new_state = dccp_state_table[role][type][old_state];
+
+	switch (new_state) {
+	case CT_DCCP_REQUEST:
+		if (old_state == CT_DCCP_TIMEWAIT &&
+		    role == CT_DCCP_ROLE_SERVER) {
+			/* Reincarnation in the reverse direction: reopen and
+			 * reverse client/server roles. */
+			ct->proto.dccp.role[dir] = CT_DCCP_ROLE_CLIENT;
+			ct->proto.dccp.role[!dir] = CT_DCCP_ROLE_SERVER;
+		}
+		break;
+	case CT_DCCP_RESPOND:
+		if (old_state == CT_DCCP_REQUEST)
+			ct->proto.dccp.handshake_seq = dccp_hdr_seq(dh);
+		break;
+	case CT_DCCP_PARTOPEN:
+		if (old_state == CT_DCCP_RESPOND &&
+		    type == DCCP_PKT_ACK &&
+		    dccp_ack_seq(dh) == ct->proto.dccp.handshake_seq)
+			set_bit(IPS_ASSURED_BIT, &ct->status);
+		break;
+	case CT_DCCP_IGNORE:
+		/*
+		 * Connection tracking might be out of sync, so we ignore
+		 * packets that might establish a new connection and resync
+		 * if the server responds with a valid Response.
+		 */
+		if (ct->proto.dccp.last_dir == !dir &&
+		    ct->proto.dccp.last_pkt == DCCP_PKT_REQUEST &&
+		    type == DCCP_PKT_RESPONSE) {
+			ct->proto.dccp.role[!dir] = CT_DCCP_ROLE_CLIENT;
+			ct->proto.dccp.role[dir] = CT_DCCP_ROLE_SERVER;
+			ct->proto.dccp.handshake_seq = dccp_hdr_seq(dh);
+			new_state = CT_DCCP_RESPOND;
+			break;
+		}
+		ct->proto.dccp.last_dir = dir;
+		ct->proto.dccp.last_pkt = type;
+
+		write_unlock_bh(&dccp_lock);
+		if (LOG_INVALID(IPPROTO_DCCP))
+			nf_log_packet(pf, 0, skb, NULL, NULL, NULL,
+				      "nf_ct_dccp: invalid packet ignored ");
+		return NF_ACCEPT;
+	case CT_DCCP_INVALID:
+		write_unlock_bh(&dccp_lock);
+		if (LOG_INVALID(IPPROTO_DCCP))
+			nf_log_packet(pf, 0, skb, NULL, NULL, NULL,
+				      "nf_ct_dccp: invalid state transition ");
+		return -NF_ACCEPT;
+	}
+
+	ct->proto.dccp.last_dir = dir;
+	ct->proto.dccp.last_pkt = type;
+	ct->proto.dccp.state = new_state;
+	write_unlock_bh(&dccp_lock);
+	nf_ct_refresh_acct(ct, ctinfo, skb, dccp_timeout[new_state]);
+
+	return NF_ACCEPT;
+}
+
+static int dccp_error(struct sk_buff *skb, unsigned int dataoff,
+		      enum ip_conntrack_info *ctinfo, int pf,
+		      unsigned int hooknum)
+{
+	struct dccp_hdr _dh, *dh;
+	unsigned int dccp_len = skb->len - dataoff;
+	unsigned int cscov;
+	const char *msg;
+
+	dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &dh);
+	if (dh == NULL) {
+		msg = "nf_ct_dccp: short packet ";
+		goto out_invalid;
+	}
+
+	if (dh->dccph_doff * 4 < sizeof(struct dccp_hdr) ||
+	    dh->dccph_doff * 4 > dccp_len) {
+		msg = "nf_ct_dccp: truncated/malformed packet ";
+		goto out_invalid;
+	}
+
+	cscov = dccp_len;
+	if (dh->dccph_cscov) {
+		cscov = (dh->dccph_cscov - 1) * 4;
+		if (cscov > dccp_len) {
+			msg = "nf_ct_dccp: bad checksum coverage ";
+			goto out_invalid;
+		}
+	}
+
+	if (nf_conntrack_checksum && hooknum == NF_INET_PRE_ROUTING &&
+	    nf_checksum_partial(skb, hooknum, dataoff, cscov, IPPROTO_DCCP,
+				pf)) {
+		msg = "nf_ct_dccp: bad checksum ";
+		goto out_invalid;
+	}
+
+	if (dh->dccph_type >= DCCP_PKT_INVALID) {
+		msg = "nf_ct_dccp: reserved packet type ";
+		goto out_invalid;
+	}
+
+	return NF_ACCEPT;
+
+out_invalid:
+	if (LOG_INVALID(IPPROTO_DCCP))
+		nf_log_packet(pf, 0, skb, NULL, NULL, NULL, msg);
+	return -NF_ACCEPT;
+}
+
+static int dccp_print_tuple(struct seq_file *s,
+			    const struct nf_conntrack_tuple *tuple)
+{
+	return seq_printf(s, "sport=%hu dport=%hu ",
+			  ntohs(tuple->src.u.dccp.port),
+			  ntohs(tuple->dst.u.dccp.port));
+}
+
+static int dccp_print_conntrack(struct seq_file *s, const struct nf_conn *ct)
+{
+	return seq_printf(s, "%s ", dccp_state_names[ct->proto.dccp.state]);
+}
+
+#if defined(CONFIG_NF_CT_NETLINK) || defined(CONFIG_NF_CT_NETLINK_MODULE)
+static int dccp_to_nlattr(struct sk_buff *skb, struct nlattr *nla,
+			  const struct nf_conn *ct)
+{
+	struct nlattr *nest_parms;
+
+	read_lock_bh(&dccp_lock);
+	nest_parms = nla_nest_start(skb, CTA_PROTOINFO_DCCP | NLA_F_NESTED);
+	if (!nest_parms)
+		goto nla_put_failure;
+	NLA_PUT_U8(skb, CTA_PROTOINFO_DCCP_STATE, ct->proto.dccp.state);
+	nla_nest_end(skb, nest_parms);
+	read_unlock_bh(&dccp_lock);
+	return 0;
+
+nla_put_failure:
+	read_unlock_bh(&dccp_lock);
+	return -1;
+}
+
+static const struct nla_policy dccp_nla_policy[CTA_PROTOINFO_DCCP_MAX + 1] = {
+	[CTA_PROTOINFO_DCCP_STATE]	= { .type = NLA_U8 },
+};
+
+static int nlattr_to_dccp(struct nlattr *cda[], struct nf_conn *ct)
+{
+	struct nlattr *attr = cda[CTA_PROTOINFO_DCCP];
+	struct nlattr *tb[CTA_PROTOINFO_DCCP_MAX + 1];
+	int err;
+
+	if (!attr)
+		return 0;
+
+	err = nla_parse_nested(tb, CTA_PROTOINFO_DCCP_MAX, attr,
+			       dccp_nla_policy);
+	if (err < 0)
+		return err;
+
+	if (!tb[CTA_PROTOINFO_DCCP_STATE] ||
+	    nla_get_u8(tb[CTA_PROTOINFO_DCCP_STATE]) >= CT_DCCP_IGNORE)
+		return -EINVAL;
+
+	write_lock_bh(&dccp_lock);
+	ct->proto.dccp.state = nla_get_u8(tb[CTA_PROTOINFO_DCCP_STATE]);
+	write_unlock_bh(&dccp_lock);
+	return 0;
+}
+#endif
+
+#ifdef CONFIG_SYSCTL
+static unsigned int dccp_sysctl_table_users;
+static struct ctl_table_header *dccp_sysctl_header;
+static ctl_table dccp_sysctl_table[] = {
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_timeout_request",
+		.data		= &dccp_timeout[CT_DCCP_REQUEST],
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_timeout_respond",
+		.data		= &dccp_timeout[CT_DCCP_RESPOND],
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_timeout_partopen",
+		.data		= &dccp_timeout[CT_DCCP_PARTOPEN],
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_timeout_open",
+		.data		= &dccp_timeout[CT_DCCP_OPEN],
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_timeout_closereq",
+		.data		= &dccp_timeout[CT_DCCP_CLOSEREQ],
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_timeout_closing",
+		.data		= &dccp_timeout[CT_DCCP_CLOSING],
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_timeout_timewait",
+		.data		= &dccp_timeout[CT_DCCP_TIMEWAIT],
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_jiffies,
+	},
+	{
+		.ctl_name	= CTL_UNNUMBERED,
+		.procname	= "nf_conntrack_dccp_loose",
+		.data		= &nf_ct_dccp_loose,
+		.maxlen		= sizeof(nf_ct_dccp_loose),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec,
+	},
+	{
+		.ctl_name	= 0,
+	}
+};
+#endif /* CONFIG_SYSCTL */
+
+static struct nf_conntrack_l4proto dccp_proto4 __read_mostly = {
+	.l3proto		= AF_INET,
+	.l4proto		= IPPROTO_DCCP,
+	.name			= "dccp",
+	.pkt_to_tuple		= dccp_pkt_to_tuple,
+	.invert_tuple		= dccp_invert_tuple,
+	.new			= dccp_new,
+	.packet			= dccp_packet,
+	.error			= dccp_error,
+	.print_tuple		= dccp_print_tuple,
+	.print_conntrack	= dccp_print_conntrack,
+#if defined(CONFIG_NF_CT_NETLINK) || defined(CONFIG_NF_CT_NETLINK_MODULE)
+	.to_nlattr		= dccp_to_nlattr,
+	.from_nlattr		= nlattr_to_dccp,
+	.tuple_to_nlattr	= nf_ct_port_tuple_to_nlattr,
+	.nlattr_to_tuple	= nf_ct_port_nlattr_to_tuple,
+	.nla_policy		= nf_ct_port_nla_policy,
+#endif
+#ifdef CONFIG_SYSCTL
+	.ctl_table_users	= &dccp_sysctl_table_users,
+	.ctl_table_header	= &dccp_sysctl_header,
+	.ctl_table		= dccp_sysctl_table,
+#endif
+};
+
+static struct nf_conntrack_l4proto dccp_proto6 __read_mostly = {
+	.l3proto		= AF_INET6,
+	.l4proto		= IPPROTO_DCCP,
+	.name			= "dccp",
+	.pkt_to_tuple		= dccp_pkt_to_tuple,
+	.invert_tuple		= dccp_invert_tuple,
+	.new			= dccp_new,
+	.packet			= dccp_packet,
+	.error			= dccp_error,
+	.print_tuple		= dccp_print_tuple,
+	.print_conntrack	= dccp_print_conntrack,
+#if defined(CONFIG_NF_CT_NETLINK) || defined(CONFIG_NF_CT_NETLINK_MODULE)
+	.to_nlattr		= dccp_to_nlattr,
+	.from_nlattr		= nlattr_to_dccp,
+	.tuple_to_nlattr	= nf_ct_port_tuple_to_nlattr,
+	.nlattr_to_tuple	= nf_ct_port_nlattr_to_tuple,
+	.nla_policy		= nf_ct_port_nla_policy,
+#endif
+#ifdef CONFIG_SYSCTL
+	.ctl_table_users	= &dccp_sysctl_table_users,
+	.ctl_table_header	= &dccp_sysctl_header,
+	.ctl_table		= dccp_sysctl_table,
+#endif
+};
+
+static int __init nf_conntrack_proto_dccp_init(void)
+{
+	int err;
+
+	err = nf_conntrack_l4proto_register(&dccp_proto4);
+	if (err < 0)
+		goto err1;
+
+	err = nf_conntrack_l4proto_register(&dccp_proto6);
+	if (err < 0)
+		goto err2;
+	return 0;
+
+err2:
+	nf_conntrack_l4proto_unregister(&dccp_proto4);
+err1:
+	return err;
+}
+
+static void __exit nf_conntrack_proto_dccp_fini(void)
+{
+	nf_conntrack_l4proto_unregister(&dccp_proto6);
+	nf_conntrack_l4proto_unregister(&dccp_proto4);
+}
+
+module_init(nf_conntrack_proto_dccp_init);
+module_exit(nf_conntrack_proto_dccp_fini);
+
+MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");
+MODULE_DESCRIPTION("DCCP connection tracking protocol helper");
+MODULE_LICENSE("GPL");
-- 
cgit v1.2.3


From dd13b010368f85dfa59364ba87bfe8ae930b2832 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Mon, 14 Apr 2008 11:15:52 +0200
Subject: [NETFILTER]: nf_nat: kill helper and seq_adjust hooks

Connection tracking helpers (specifically FTP) need to be called
before NAT sequence numbers adjustments are performed to be able
to compare them against previously seen ones. We've introduced
two new hooks around 2.6.11 to maintain this ordering when NAT
modules were changed to get called from conntrack helpers directly.

The cost of netfilter hooks is quite high and sequence number
adjustments are only rarely needed however. Add a RCU-protected
sequence number adjustment function pointer and call it from
IPv4 conntrack after calling the helper.

Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter_ipv4.h                 |  2 -
 include/net/netfilter/nf_nat_helper.h          |  3 ++
 net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c | 56 +++++++++++++-------------
 net/ipv4/netfilter/nf_nat_core.c               |  5 +++
 net/ipv4/netfilter/nf_nat_helper.c             |  1 -
 net/ipv4/netfilter/nf_nat_standalone.c         | 35 ----------------
 6 files changed, 35 insertions(+), 67 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/netfilter_ipv4.h b/include/linux/netfilter_ipv4.h
index 9a10092e358c..650318b0c405 100644
--- a/include/linux/netfilter_ipv4.h
+++ b/include/linux/netfilter_ipv4.h
@@ -62,8 +62,6 @@ enum nf_ip_hook_priorities {
 	NF_IP_PRI_FILTER = 0,
 	NF_IP_PRI_NAT_SRC = 100,
 	NF_IP_PRI_SELINUX_LAST = 225,
-	NF_IP_PRI_CONNTRACK_HELPER = INT_MAX - 2,
-	NF_IP_PRI_NAT_SEQ_ADJUST = INT_MAX - 1,
 	NF_IP_PRI_CONNTRACK_CONFIRM = INT_MAX,
 	NF_IP_PRI_LAST = INT_MAX,
 };
diff --git a/include/net/netfilter/nf_nat_helper.h b/include/net/netfilter/nf_nat_helper.h
index 58dd22687949..237a961f40e1 100644
--- a/include/net/netfilter/nf_nat_helper.h
+++ b/include/net/netfilter/nf_nat_helper.h
@@ -24,6 +24,9 @@ extern int nf_nat_mangle_udp_packet(struct sk_buff *skb,
 extern int nf_nat_seq_adjust(struct sk_buff *skb,
 			     struct nf_conn *ct,
 			     enum ip_conntrack_info ctinfo);
+extern int (*nf_nat_seq_adjust_hook)(struct sk_buff *skb,
+				     struct nf_conn *ct,
+				     enum ip_conntrack_info ctinfo);
 
 /* Setup NAT on this expected conntrack so it follows master, but goes
  * to port ct->master->saved_proto. */
diff --git a/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c
index a65b845c5f15..41e79613eb0a 100644
--- a/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c
+++ b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c
@@ -23,6 +23,12 @@
 #include <net/netfilter/nf_conntrack_l3proto.h>
 #include <net/netfilter/nf_conntrack_core.h>
 #include <net/netfilter/ipv4/nf_conntrack_ipv4.h>
+#include <net/netfilter/nf_nat_helper.h>
+
+int (*nf_nat_seq_adjust_hook)(struct sk_buff *skb,
+			      struct nf_conn *ct,
+			      enum ip_conntrack_info ctinfo);
+EXPORT_SYMBOL_GPL(nf_nat_seq_adjust_hook);
 
 static int ipv4_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
 			     struct nf_conntrack_tuple *tuple)
@@ -100,36 +106,42 @@ static unsigned int ipv4_confirm(unsigned int hooknum,
 				 const struct net_device *in,
 				 const struct net_device *out,
 				 int (*okfn)(struct sk_buff *))
-{
-	/* We've seen it coming out the other side: confirm it */
-	return nf_conntrack_confirm(skb);
-}
-
-static unsigned int ipv4_conntrack_help(unsigned int hooknum,
-				      struct sk_buff *skb,
-				      const struct net_device *in,
-				      const struct net_device *out,
-				      int (*okfn)(struct sk_buff *))
 {
 	struct nf_conn *ct;
 	enum ip_conntrack_info ctinfo;
 	const struct nf_conn_help *help;
 	const struct nf_conntrack_helper *helper;
+	unsigned int ret;
 
 	/* This is where we call the helper: as the packet goes out. */
 	ct = nf_ct_get(skb, &ctinfo);
 	if (!ct || ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
-		return NF_ACCEPT;
+		goto out;
 
 	help = nfct_help(ct);
 	if (!help)
-		return NF_ACCEPT;
+		goto out;
+
 	/* rcu_read_lock()ed by nf_hook_slow */
 	helper = rcu_dereference(help->helper);
 	if (!helper)
-		return NF_ACCEPT;
-	return helper->help(skb, skb_network_offset(skb) + ip_hdrlen(skb),
-			    ct, ctinfo);
+		goto out;
+
+	ret = helper->help(skb, skb_network_offset(skb) + ip_hdrlen(skb),
+			   ct, ctinfo);
+	if (ret != NF_ACCEPT)
+		return ret;
+
+	if (test_bit(IPS_SEQ_ADJUST_BIT, &ct->status)) {
+		typeof(nf_nat_seq_adjust_hook) seq_adjust;
+
+		seq_adjust = rcu_dereference(nf_nat_seq_adjust_hook);
+		if (!seq_adjust || !seq_adjust(skb, ct, ctinfo))
+			return NF_DROP;
+	}
+out:
+	/* We've seen it coming out the other side: confirm it */
+	return nf_conntrack_confirm(skb);
 }
 
 static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
@@ -210,20 +222,6 @@ static struct nf_hook_ops ipv4_conntrack_ops[] __read_mostly = {
 		.hooknum	= NF_INET_LOCAL_OUT,
 		.priority	= NF_IP_PRI_CONNTRACK,
 	},
-	{
-		.hook		= ipv4_conntrack_help,
-		.owner		= THIS_MODULE,
-		.pf		= PF_INET,
-		.hooknum	= NF_INET_POST_ROUTING,
-		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
-	},
-	{
-		.hook		= ipv4_conntrack_help,
-		.owner		= THIS_MODULE,
-		.pf		= PF_INET,
-		.hooknum	= NF_INET_LOCAL_IN,
-		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
-	},
 	{
 		.hook		= ipv4_confirm,
 		.owner		= THIS_MODULE,
diff --git a/net/ipv4/netfilter/nf_nat_core.c b/net/ipv4/netfilter/nf_nat_core.c
index 9320c7ac5729..25c3efe4207e 100644
--- a/net/ipv4/netfilter/nf_nat_core.c
+++ b/net/ipv4/netfilter/nf_nat_core.c
@@ -618,6 +618,9 @@ static int __init nf_nat_init(void)
 	nf_conntrack_untracked.status |= IPS_NAT_DONE_MASK;
 
 	l3proto = nf_ct_l3proto_find_get((u_int16_t)AF_INET);
+
+	BUG_ON(nf_nat_seq_adjust_hook != NULL);
+	rcu_assign_pointer(nf_nat_seq_adjust_hook, nf_nat_seq_adjust);
 	return 0;
 
  cleanup_extend:
@@ -644,6 +647,8 @@ static void __exit nf_nat_cleanup(void)
 	nf_ct_free_hashtable(bysource, nf_nat_vmalloced, nf_nat_htable_size);
 	nf_ct_l3proto_put(l3proto);
 	nf_ct_extend_unregister(&nat_extend);
+	rcu_assign_pointer(nf_nat_seq_adjust_hook, NULL);
+	synchronize_net();
 }
 
 MODULE_LICENSE("GPL");
diff --git a/net/ipv4/netfilter/nf_nat_helper.c b/net/ipv4/netfilter/nf_nat_helper.c
index 2fca727aa8ba..11976ea29884 100644
--- a/net/ipv4/netfilter/nf_nat_helper.c
+++ b/net/ipv4/netfilter/nf_nat_helper.c
@@ -416,7 +416,6 @@ nf_nat_seq_adjust(struct sk_buff *skb,
 
 	return 1;
 }
-EXPORT_SYMBOL(nf_nat_seq_adjust);
 
 /* Setup NAT on this expected conntrack so it follows master. */
 /* If we fail to get a free NAT slot, we'll get dropped on confirm */
diff --git a/net/ipv4/netfilter/nf_nat_standalone.c b/net/ipv4/netfilter/nf_nat_standalone.c
index a366b5865b9c..b7dd695691a0 100644
--- a/net/ipv4/netfilter/nf_nat_standalone.c
+++ b/net/ipv4/netfilter/nf_nat_standalone.c
@@ -245,25 +245,6 @@ nf_nat_local_fn(unsigned int hooknum,
 	return ret;
 }
 
-static unsigned int
-nf_nat_adjust(unsigned int hooknum,
-	      struct sk_buff *skb,
-	      const struct net_device *in,
-	      const struct net_device *out,
-	      int (*okfn)(struct sk_buff *))
-{
-	struct nf_conn *ct;
-	enum ip_conntrack_info ctinfo;
-
-	ct = nf_ct_get(skb, &ctinfo);
-	if (ct && test_bit(IPS_SEQ_ADJUST_BIT, &ct->status)) {
-		pr_debug("nf_nat_standalone: adjusting sequence number\n");
-		if (!nf_nat_seq_adjust(skb, ct, ctinfo))
-			return NF_DROP;
-	}
-	return NF_ACCEPT;
-}
-
 /* We must be after connection tracking and before packet filtering. */
 
 static struct nf_hook_ops nf_nat_ops[] __read_mostly = {
@@ -283,14 +264,6 @@ static struct nf_hook_ops nf_nat_ops[] __read_mostly = {
 		.hooknum	= NF_INET_POST_ROUTING,
 		.priority	= NF_IP_PRI_NAT_SRC,
 	},
-	/* After conntrack, adjust sequence number */
-	{
-		.hook		= nf_nat_adjust,
-		.owner		= THIS_MODULE,
-		.pf		= PF_INET,
-		.hooknum	= NF_INET_POST_ROUTING,
-		.priority	= NF_IP_PRI_NAT_SEQ_ADJUST,
-	},
 	/* Before packet filtering, change destination */
 	{
 		.hook		= nf_nat_local_fn,
@@ -307,14 +280,6 @@ static struct nf_hook_ops nf_nat_ops[] __read_mostly = {
 		.hooknum	= NF_INET_LOCAL_IN,
 		.priority	= NF_IP_PRI_NAT_SRC,
 	},
-	/* After conntrack, adjust sequence number */
-	{
-		.hook		= nf_nat_adjust,
-		.owner		= THIS_MODULE,
-		.pf		= PF_INET,
-		.hooknum	= NF_INET_LOCAL_IN,
-		.priority	= NF_IP_PRI_NAT_SEQ_ADJUST,
-	},
 };
 
 static int __init nf_nat_standalone_init(void)
-- 
cgit v1.2.3


From e7bfd0a1a6c8f82977253dab19be9d9979c1ec1b Mon Sep 17 00:00:00 2001
From: Peter Warasin <peter@endian.com>
Date: Mon, 14 Apr 2008 11:15:54 +0200
Subject: [NETFILTER]: bridge: add ebt_nflog watcher

This patch adds the ebtables nflog watcher to the kernel in order to
allow ebtables log through the nfnetlink_log backend.

Signed-off-by: Peter Warasin <peter@endian.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 include/linux/netfilter_bridge/ebt_nflog.h | 21 +++++++++
 net/bridge/netfilter/Kconfig               | 14 ++++++
 net/bridge/netfilter/Makefile              |  1 +
 net/bridge/netfilter/ebt_nflog.c           | 74 ++++++++++++++++++++++++++++++
 4 files changed, 110 insertions(+)
 create mode 100644 include/linux/netfilter_bridge/ebt_nflog.h
 create mode 100644 net/bridge/netfilter/ebt_nflog.c

(limited to 'include/linux')

diff --git a/include/linux/netfilter_bridge/ebt_nflog.h b/include/linux/netfilter_bridge/ebt_nflog.h
new file mode 100644
index 000000000000..052817849b83
--- /dev/null
+++ b/include/linux/netfilter_bridge/ebt_nflog.h
@@ -0,0 +1,21 @@
+#ifndef __LINUX_BRIDGE_EBT_NFLOG_H
+#define __LINUX_BRIDGE_EBT_NFLOG_H
+
+#define EBT_NFLOG_MASK 0x0
+
+#define EBT_NFLOG_PREFIX_SIZE 64
+#define EBT_NFLOG_WATCHER "nflog"
+
+#define EBT_NFLOG_DEFAULT_GROUP		0x1
+#define EBT_NFLOG_DEFAULT_THRESHOLD	1
+
+struct ebt_nflog_info {
+	u_int32_t len;
+	u_int16_t group;
+	u_int16_t threshold;
+	u_int16_t flags;
+	u_int16_t pad;
+	char prefix[EBT_NFLOG_PREFIX_SIZE];
+};
+
+#endif				/* __LINUX_BRIDGE_EBT_NFLOG_H */
diff --git a/net/bridge/netfilter/Kconfig b/net/bridge/netfilter/Kconfig
index 4a3e2bf892c7..7beeefa0f9c0 100644
--- a/net/bridge/netfilter/Kconfig
+++ b/net/bridge/netfilter/Kconfig
@@ -212,4 +212,18 @@ config BRIDGE_EBT_ULOG
 
 	  To compile it as a module, choose M here.  If unsure, say N.
 
+config BRIDGE_EBT_NFLOG
+	tristate "ebt: nflog support"
+	depends on BRIDGE_NF_EBTABLES
+	help
+	  This option enables the nflog watcher, which allows to LOG
+	  messages through the netfilter logging API, which can use
+	  either the old LOG target, the old ULOG target or nfnetlink_log
+	  as backend.
+
+	  This option adds the ulog watcher, that you can use in any rule
+	  in any ebtables table.
+
+	  To compile it as a module, choose M here.  If unsure, say N.
+
 endmenu
diff --git a/net/bridge/netfilter/Makefile b/net/bridge/netfilter/Makefile
index 905087e0d485..83715d73a503 100644
--- a/net/bridge/netfilter/Makefile
+++ b/net/bridge/netfilter/Makefile
@@ -30,3 +30,4 @@ obj-$(CONFIG_BRIDGE_EBT_SNAT) += ebt_snat.o
 # watchers
 obj-$(CONFIG_BRIDGE_EBT_LOG) += ebt_log.o
 obj-$(CONFIG_BRIDGE_EBT_ULOG) += ebt_ulog.o
+obj-$(CONFIG_BRIDGE_EBT_NFLOG) += ebt_nflog.o
diff --git a/net/bridge/netfilter/ebt_nflog.c b/net/bridge/netfilter/ebt_nflog.c
new file mode 100644
index 000000000000..8e799aa9e560
--- /dev/null
+++ b/net/bridge/netfilter/ebt_nflog.c
@@ -0,0 +1,74 @@
+/*
+ * ebt_nflog
+ *
+ *	Author:
+ *	Peter Warasin <peter@endian.com>
+ *
+ *  February, 2008
+ *
+ * Based on:
+ *  xt_NFLOG.c, (C) 2006 by Patrick McHardy <kaber@trash.net>
+ *  ebt_ulog.c, (C) 2004 by Bart De Schuymer <bdschuym@pandora.be>
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/spinlock.h>
+#include <linux/netfilter_bridge/ebtables.h>
+#include <linux/netfilter_bridge/ebt_nflog.h>
+#include <net/netfilter/nf_log.h>
+
+static void ebt_nflog(const struct sk_buff *skb,
+		      unsigned int hooknr,
+		      const struct net_device *in,
+		      const struct net_device *out,
+		      const void *data, unsigned int datalen)
+{
+	struct ebt_nflog_info *info = (struct ebt_nflog_info *)data;
+	struct nf_loginfo li;
+
+	li.type = NF_LOG_TYPE_ULOG;
+	li.u.ulog.copy_len = info->len;
+	li.u.ulog.group = info->group;
+	li.u.ulog.qthreshold = info->threshold;
+
+	nf_log_packet(PF_BRIDGE, hooknr, skb, in, out, &li, "%s", info->prefix);
+}
+
+static int ebt_nflog_check(const char *tablename,
+			   unsigned int hookmask,
+			   const struct ebt_entry *e,
+			   void *data, unsigned int datalen)
+{
+	struct ebt_nflog_info *info = (struct ebt_nflog_info *)data;
+
+	if (datalen != EBT_ALIGN(sizeof(struct ebt_nflog_info)))
+		return -EINVAL;
+	if (info->flags & ~EBT_NFLOG_MASK)
+		return -EINVAL;
+	info->prefix[EBT_NFLOG_PREFIX_SIZE - 1] = '\0';
+	return 0;
+}
+
+static struct ebt_watcher nflog __read_mostly = {
+	.name = EBT_NFLOG_WATCHER,
+	.watcher = ebt_nflog,
+	.check = ebt_nflog_check,
+	.me = THIS_MODULE,
+};
+
+static int __init ebt_nflog_init(void)
+{
+	return ebt_register_watcher(&nflog);
+}
+
+static void __exit ebt_nflog_fini(void)
+{
+	ebt_unregister_watcher(&nflog);
+}
+
+module_init(ebt_nflog_init);
+module_exit(ebt_nflog_fini);
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Peter Warasin <peter@endian.com>");
+MODULE_DESCRIPTION("ebtables NFLOG netfilter logging module");
-- 
cgit v1.2.3


From 0f389ec63077521166f071e1e970aed36147fd45 Mon Sep 17 00:00:00 2001
From: Christoph Lameter <clameter@sgi.com>
Date: Mon, 14 Apr 2008 18:53:02 +0300
Subject: slub: No need for per node slab counters if !SLUB_DEBUG

The per node counters are used mainly for showing data through the sysfs API.
If that API is not compiled in then there is no point in keeping track of this
data. Disable counters for the number of slabs and the number of total slabs
if !SLUB_DEBUG. Incrementing the per node counters is also accessing a
potentially contended cacheline so this could actually be a performance
benefit to embedded systems.

SLABINFO support is also affected. It now must depends on SLUB_DEBUG (which
is on by default).

Patch also avoids a check for a NULL kmem_cache_node pointer in new_slab()
if the system is not compiled with NUMA support.

[penberg@cs.helsinki.fi: fix oops and move ->nr_slabs into CONFIG_SLUB_DEBUG]
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
---
 include/linux/slub_def.h |  2 +-
 init/Kconfig             |  2 +-
 mm/slub.c                | 51 +++++++++++++++++++++++++++++++++++++-----------
 3 files changed, 42 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h
index b00c1c73eb0a..79d59c937fac 100644
--- a/include/linux/slub_def.h
+++ b/include/linux/slub_def.h
@@ -45,9 +45,9 @@ struct kmem_cache_cpu {
 struct kmem_cache_node {
 	spinlock_t list_lock;	/* Protect partial list and nr_partial */
 	unsigned long nr_partial;
-	atomic_long_t nr_slabs;
 	struct list_head partial;
 #ifdef CONFIG_SLUB_DEBUG
+	atomic_long_t nr_slabs;
 	struct list_head full;
 #endif
 };
diff --git a/init/Kconfig b/init/Kconfig
index a97924bc5b8d..7fccf09bb95a 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -763,7 +763,7 @@ endmenu		# General setup
 config SLABINFO
 	bool
 	depends on PROC_FS
-	depends on SLAB || SLUB
+	depends on SLAB || SLUB_DEBUG
 	default y
 
 config RT_MUTEXES
diff --git a/mm/slub.c b/mm/slub.c
index f924cffb29e7..7f8aaa291a4e 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -837,6 +837,35 @@ static void remove_full(struct kmem_cache *s, struct page *page)
 	spin_unlock(&n->list_lock);
 }
 
+/* Tracking of the number of slabs for debugging purposes */
+static inline unsigned long slabs_node(struct kmem_cache *s, int node)
+{
+	struct kmem_cache_node *n = get_node(s, node);
+
+	return atomic_long_read(&n->nr_slabs);
+}
+
+static inline void inc_slabs_node(struct kmem_cache *s, int node)
+{
+	struct kmem_cache_node *n = get_node(s, node);
+
+	/*
+	 * May be called early in order to allocate a slab for the
+	 * kmem_cache_node structure. Solve the chicken-egg
+	 * dilemma by deferring the increment of the count during
+	 * bootstrap (see early_kmem_cache_node_alloc).
+	 */
+	if (!NUMA_BUILD || n)
+		atomic_long_inc(&n->nr_slabs);
+}
+static inline void dec_slabs_node(struct kmem_cache *s, int node)
+{
+	struct kmem_cache_node *n = get_node(s, node);
+
+	atomic_long_dec(&n->nr_slabs);
+}
+
+/* Object debug checks for alloc/free paths */
 static void setup_object_debug(struct kmem_cache *s, struct page *page,
 								void *object)
 {
@@ -1028,6 +1057,11 @@ static inline unsigned long kmem_cache_flags(unsigned long objsize,
 	return flags;
 }
 #define slub_debug 0
+
+static inline unsigned long slabs_node(struct kmem_cache *s, int node)
+							{ return 0; }
+static inline void inc_slabs_node(struct kmem_cache *s, int node) {}
+static inline void dec_slabs_node(struct kmem_cache *s, int node) {}
 #endif
 /*
  * Slab allocation and freeing
@@ -1066,7 +1100,6 @@ static void setup_object(struct kmem_cache *s, struct page *page,
 static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node)
 {
 	struct page *page;
-	struct kmem_cache_node *n;
 	void *start;
 	void *last;
 	void *p;
@@ -1078,9 +1111,7 @@ static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node)
 	if (!page)
 		goto out;
 
-	n = get_node(s, page_to_nid(page));
-	if (n)
-		atomic_long_inc(&n->nr_slabs);
+	inc_slabs_node(s, page_to_nid(page));
 	page->slab = s;
 	page->flags |= 1 << PG_slab;
 	if (s->flags & (SLAB_DEBUG_FREE | SLAB_RED_ZONE | SLAB_POISON |
@@ -1153,9 +1184,7 @@ static void free_slab(struct kmem_cache *s, struct page *page)
 
 static void discard_slab(struct kmem_cache *s, struct page *page)
 {
-	struct kmem_cache_node *n = get_node(s, page_to_nid(page));
-
-	atomic_long_dec(&n->nr_slabs);
+	dec_slabs_node(s, page_to_nid(page));
 	free_slab(s, page);
 }
 
@@ -1894,10 +1923,10 @@ static void init_kmem_cache_cpu(struct kmem_cache *s,
 static void init_kmem_cache_node(struct kmem_cache_node *n)
 {
 	n->nr_partial = 0;
-	atomic_long_set(&n->nr_slabs, 0);
 	spin_lock_init(&n->list_lock);
 	INIT_LIST_HEAD(&n->partial);
 #ifdef CONFIG_SLUB_DEBUG
+	atomic_long_set(&n->nr_slabs, 0);
 	INIT_LIST_HEAD(&n->full);
 #endif
 }
@@ -2066,7 +2095,7 @@ static struct kmem_cache_node *early_kmem_cache_node_alloc(gfp_t gfpflags,
 	init_tracking(kmalloc_caches, n);
 #endif
 	init_kmem_cache_node(n);
-	atomic_long_inc(&n->nr_slabs);
+	inc_slabs_node(kmalloc_caches, node);
 
 	/*
 	 * lockdep requires consistent irq usage for each lock
@@ -2379,7 +2408,7 @@ static inline int kmem_cache_close(struct kmem_cache *s)
 		struct kmem_cache_node *n = get_node(s, node);
 
 		n->nr_partial -= free_list(s, n, &n->partial);
-		if (atomic_long_read(&n->nr_slabs))
+		if (slabs_node(s, node))
 			return 1;
 	}
 	free_kmem_cache_nodes(s);
@@ -2801,7 +2830,7 @@ static void slab_mem_offline_callback(void *arg)
 			 * and offline_pages() function shoudn't call this
 			 * callback. So, we must fail.
 			 */
-			BUG_ON(atomic_long_read(&n->nr_slabs));
+			BUG_ON(slabs_node(s, offline_node));
 
 			s->node[offline_node] = NULL;
 			kmem_cache_free(kmalloc_caches, n);
-- 
cgit v1.2.3


From 79966fd9b4781f9bd257312489ff511f2c01f210 Mon Sep 17 00:00:00 2001
From: David Brownell <david-b@pacbell.net>
Date: Thu, 28 Feb 2008 22:07:28 -0800
Subject: ARM: OMAP: I2C: tps65010 driver converts to gpiolib

Make the tps65010 driver use gpiolib to expose its GPIOs.

Note: This patch will get merged via omap tree instead of I2C
as it will cause some board updates. This has been discussed
at on the I2C list:

http://lists.lm-sensors.org/pipermail/i2c/2008-March/003031.html

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Cc: i2c@lm-sensors.org
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
 drivers/i2c/chips/Kconfig    |   1 +
 drivers/i2c/chips/tps65010.c | 101 ++++++++++++++++++++++++++++++++++++++++++-
 include/linux/i2c/tps65010.h |  30 +++++++++++++
 3 files changed, 131 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/i2c/chips/Kconfig b/drivers/i2c/chips/Kconfig
index b21593f93586..2da2edfa68ec 100644
--- a/drivers/i2c/chips/Kconfig
+++ b/drivers/i2c/chips/Kconfig
@@ -93,6 +93,7 @@ config ISP1301_OMAP
 
 config TPS65010
 	tristate "TPS6501x Power Management chips"
+	depends on HAVE_GPIO_LIB
 	default y if MACH_OMAP_H2 || MACH_OMAP_H3 || MACH_OMAP_OSK
 	help
 	  If you say yes here you get support for the TPS6501x series of
diff --git a/drivers/i2c/chips/tps65010.c b/drivers/i2c/chips/tps65010.c
index 4154a9108859..b67f69c2e7f3 100644
--- a/drivers/i2c/chips/tps65010.c
+++ b/drivers/i2c/chips/tps65010.c
@@ -30,9 +30,13 @@
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
 #include <linux/mutex.h>
+#include <linux/platform_device.h>
 
 #include <linux/i2c/tps65010.h>
 
+#include <asm/gpio.h>
+
+
 /*-------------------------------------------------------------------------*/
 
 #define	DRIVER_VERSION	"2 May 2005"
@@ -84,7 +88,9 @@ struct tps65010 {
 	u8			chgstatus, regstatus, chgconf;
 	u8			nmask1, nmask2;
 
-	/* not currently tracking GPIO state */
+	u8			outmask;
+	struct gpio_chip	chip;
+	struct platform_device	*leds;
 };
 
 #define	POWER_POLL_DELAY	msecs_to_jiffies(5000)
@@ -447,6 +453,59 @@ static irqreturn_t tps65010_irq(int irq, void *_tps)
 	return IRQ_HANDLED;
 }
 
+/*-------------------------------------------------------------------------*/
+
+/* offsets 0..3 == GPIO1..GPIO4
+ * offsets 4..5 == LED1/nPG, LED2 (we set one of the non-BLINK modes)
+ */
+static void
+tps65010_gpio_set(struct gpio_chip *chip, unsigned offset, int value)
+{
+	if (offset < 4)
+		tps65010_set_gpio_out_value(offset + 1, value);
+	else
+		tps65010_set_led(offset - 3, value ? ON : OFF);
+}
+
+static int
+tps65010_output(struct gpio_chip *chip, unsigned offset, int value)
+{
+	/* GPIOs may be input-only */
+	if (offset < 4) {
+		struct tps65010		*tps;
+
+		tps = container_of(chip, struct tps65010, chip);
+		if (!(tps->outmask & (1 << offset)))
+			return -EINVAL;
+		tps65010_set_gpio_out_value(offset + 1, value);
+	} else
+		tps65010_set_led(offset - 3, value ? ON : OFF);
+
+	return 0;
+}
+
+static int tps65010_gpio_get(struct gpio_chip *chip, unsigned offset)
+{
+	int			value;
+	struct tps65010		*tps;
+
+	tps = container_of(chip, struct tps65010, chip);
+
+	if (offset < 4) {
+		value = i2c_smbus_read_byte_data(tps->client, TPS_DEFGPIO);
+		if (value < 0)
+			return 0;
+		if (value & (1 << (offset + 4)))	/* output */
+			return !(value & (1 << offset));
+		else					/* input */
+			return (value & (1 << offset));
+	}
+
+	/* REVISIT we *could* report LED1/nPG and LED2 state ... */
+	return 0;
+}
+
+
 /*-------------------------------------------------------------------------*/
 
 static struct tps65010 *the_tps;
@@ -454,7 +513,14 @@ static struct tps65010 *the_tps;
 static int __exit tps65010_remove(struct i2c_client *client)
 {
 	struct tps65010		*tps = i2c_get_clientdata(client);
+	struct tps65010_board	*board = client->dev.platform_data;
 
+	if (board && board->teardown) {
+		int status = board->teardown(client, board->context);
+		if (status < 0)
+			dev_dbg(&client->dev, "board %s %s err %d\n",
+				"teardown", client->name, status);
+	}
 	if (client->irq > 0)
 		free_irq(client->irq, tps);
 	cancel_delayed_work(&tps->work);
@@ -469,6 +535,7 @@ static int tps65010_probe(struct i2c_client *client)
 {
 	struct tps65010		*tps;
 	int			status;
+	struct tps65010_board	*board = client->dev.platform_data;
 
 	if (the_tps) {
 		dev_dbg(&client->dev, "only one tps6501x chip allowed\n");
@@ -577,6 +644,38 @@ static int tps65010_probe(struct i2c_client *client)
 
 	tps->file = debugfs_create_file(DRIVER_NAME, S_IRUGO, NULL,
 				tps, DEBUG_FOPS);
+
+	/* optionally register GPIOs */
+	if (board && board->base > 0) {
+		tps->outmask = board->outmask;
+
+		tps->chip.label = client->name;
+
+		tps->chip.set = tps65010_gpio_set;
+		tps->chip.direction_output = tps65010_output;
+
+		/* NOTE:  only partial support for inputs; nyet IRQs */
+		tps->chip.get = tps65010_gpio_get;
+
+		tps->chip.base = board->base;
+		tps->chip.ngpio = 6;
+		tps->chip.can_sleep = 1;
+
+		status = gpiochip_add(&tps->chip);
+		if (status < 0)
+			dev_err(&client->dev, "can't add gpiochip, err %d\n",
+					status);
+		else if (board->setup) {
+			status = board->setup(client, board->context);
+			if (status < 0) {
+				dev_dbg(&client->dev,
+					"board %s %s err %d\n",
+					"setup", client->name, status);
+				status = 0;
+			}
+		}
+	}
+
 	return 0;
 fail1:
 	kfree(tps);
diff --git a/include/linux/i2c/tps65010.h b/include/linux/i2c/tps65010.h
index 7021635ed6a0..918c5354d9b8 100644
--- a/include/linux/i2c/tps65010.h
+++ b/include/linux/i2c/tps65010.h
@@ -152,5 +152,35 @@ extern int tps65010_config_vregs1(unsigned value);
  */
 extern int tps65013_set_low_pwr(unsigned mode);
 
+
+struct i2c_client;
+
+/**
+ * struct tps65010_board - packages GPIO and LED lines
+ * @base: the GPIO number to assign to GPIO-1
+ * @outmask: bit (N-1) is set to allow GPIO-N to be used as an
+ *	(open drain) output
+ * @setup: optional callback issued once the GPIOs are valid
+ * @teardown: optional callback issued before the GPIOs are invalidated
+ * @context: optional parameter passed to setup() and teardown()
+ *
+ * Board data may be used to package the GPIO (and LED) lines for use
+ * in by the generic GPIO and LED frameworks.  The first four GPIOs
+ * starting at gpio_base are GPIO1..GPIO4.  The next two are LED1/nPG
+ * and LED2 (with hardware blinking capability, not currently exposed).
+ *
+ * The @setup callback may be used with the kind of board-specific glue
+ * which hands the (now-valid) GPIOs to other drivers, or which puts
+ * devices in their initial states using these GPIOs.
+ */
+struct tps65010_board {
+	int				base;
+	unsigned			outmask;
+
+	int		(*setup)(struct i2c_client *client, void *context);
+	int		(*teardown)(struct i2c_client *client, void *context);
+	void		*context;
+};
+
 #endif /*  __LINUX_I2C_TPS65010_H */
 
-- 
cgit v1.2.3


From d237e5c7ce82b6c7e1aa50348520a0c5a47274fa Mon Sep 17 00:00:00 2001
From: Jens Axboe <jens.axboe@oracle.com>
Date: Tue, 15 Apr 2008 09:25:33 +0200
Subject: io context: increment task attachment count in ioc_task_link()

Thanks to Nikanth Karthikesan <knikanth@suse.de> for reporting this.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 include/linux/iocontext.h | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/iocontext.h b/include/linux/iocontext.h
index cac4b364cd40..2b7a1187cb29 100644
--- a/include/linux/iocontext.h
+++ b/include/linux/iocontext.h
@@ -91,8 +91,10 @@ static inline struct io_context *ioc_task_link(struct io_context *ioc)
 	 * if ref count is zero, don't allow sharing (ioc is going away, it's
 	 * a race).
 	 */
-	if (ioc && atomic_inc_not_zero(&ioc->refcount))
+	if (ioc && atomic_inc_not_zero(&ioc->refcount)) {
+		atomic_inc(&ioc->nr_tasks);
 		return ioc;
+	}
 
 	return NULL;
 }
-- 
cgit v1.2.3


From 99971e70fdc1862e120f3319fc0a4dba8c728acf Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Tue, 15 Apr 2008 00:27:58 -0700
Subject: [WANPIPE]: Forgotten bits of Sangoma drivers removal.

Robert P. J. Day spotted that my removal of the Sangoma drivers missed
a few bits.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 Documentation/networking/00-INDEX       |   2 -
 Documentation/networking/wan-router.txt | 621 --------------------------------
 include/linux/Kbuild                    |   1 -
 include/linux/if_wanpipe.h              | 124 -------
 4 files changed, 748 deletions(-)
 delete mode 100644 Documentation/networking/wan-router.txt
 delete mode 100644 include/linux/if_wanpipe.h

(limited to 'include/linux')

diff --git a/Documentation/networking/00-INDEX b/Documentation/networking/00-INDEX
index c485ee028bd9..1634c6dcecae 100644
--- a/Documentation/networking/00-INDEX
+++ b/Documentation/networking/00-INDEX
@@ -100,8 +100,6 @@ tuntap.txt
 	- TUN/TAP device driver, allowing user space Rx/Tx of packets.
 vortex.txt
 	- info on using 3Com Vortex (3c590, 3c592, 3c595, 3c597) Ethernet cards.
-wan-router.txt
-	- WAN router documentation
 wavelan.txt
 	- AT&T GIS (nee NCR) WaveLAN card: An Ethernet-like radio transceiver
 x25.txt
diff --git a/Documentation/networking/wan-router.txt b/Documentation/networking/wan-router.txt
deleted file mode 100644
index bc2ab419a74a..000000000000
--- a/Documentation/networking/wan-router.txt
+++ /dev/null
@@ -1,621 +0,0 @@
-------------------------------------------------------------------------------
-Linux WAN Router Utilities Package
-------------------------------------------------------------------------------
-Version 2.2.1 
-Mar 28, 2001
-Author: Nenad Corbic <ncorbic@sangoma.com>
-Copyright (c) 1995-2001 Sangoma Technologies Inc.
-------------------------------------------------------------------------------
-
-INTRODUCTION
-
-Wide Area Networks (WANs) are used to interconnect Local Area Networks (LANs)
-and/or stand-alone hosts over vast distances with data transfer rates
-significantly higher than those achievable with commonly used dial-up
-connections.
-
-Usually an external device called `WAN router' sitting on your local network
-or connected to your machine's serial port provides physical connection to
-WAN.  Although router's job may be as simple as taking your local network
-traffic, converting it to WAN format and piping it through the WAN link, these
-devices are notoriously expensive, with prices as much as 2 - 5 times higher
-then the price of a typical PC box.
-
-Alternatively, considering robustness and multitasking capabilities of Linux,
-an internal router can be built (most routers use some sort of stripped down
-Unix-like operating system anyway). With a number of relatively inexpensive WAN
-interface cards available on the market, a perfectly usable router can be
-built for less than half a price of an external router.  Yet a Linux box
-acting as a router can still be used for other purposes, such as fire-walling,
-running FTP, WWW or DNS server, etc.
-
-This kernel module introduces the notion of a WAN Link Driver (WLD) to Linux
-operating system and provides generic hardware-independent services for such
-drivers.  Why can existing Linux network device interface not be used for
-this purpose?  Well, it can.  However, there are a few key differences between
-a typical network interface (e.g. Ethernet) and a WAN link.
-
-Many WAN protocols, such as X.25 and frame relay, allow for multiple logical
-connections (known as `virtual circuits' in X.25 terminology) over a single
-physical link.  Each such virtual circuit may (and almost always does) lead
-to a different geographical location and, therefore, different network.  As a
-result, it is the virtual circuit, not the physical link, that represents a
-route and, therefore, a network interface in Linux terms.
-
-To further complicate things, virtual circuits are usually volatile in nature
-(excluding so called `permanent' virtual circuits or PVCs).  With almost no
-time required to set up and tear down a virtual circuit, it is highly desirable
-to implement on-demand connections in order to minimize network charges.  So
-unlike a typical network driver, the WAN driver must be able to handle multiple
-network interfaces and cope as multiple virtual circuits come into existence
-and go away dynamically.
- 
-Last, but not least, WAN configuration is much more complex than that of say
-Ethernet and may well amount to several dozens of parameters.  Some of them
-are "link-wide"  while others are virtual circuit-specific.  The same holds
-true for WAN statistics which is by far more extensive and extremely useful
-when troubleshooting WAN connections.  Extending the ifconfig utility to suit
-these needs may be possible, but does not seem quite reasonable.  Therefore, a
-WAN configuration utility and corresponding application programmer's interface
-is needed for this purpose.
-
-Most of these problems are taken care of by this module.  Its goal is to
-provide a user with more-or-less standard look and feel for all WAN devices and
-assist a WAN device driver writer by providing common services, such as:
-
- o User-level interface via /proc file system
- o Centralized configuration
- o Device management (setup, shutdown, etc.)
- o Network interface management (dynamic creation/destruction)
- o Protocol encapsulation/decapsulation
-
-To ba able to use the Linux WAN Router you will also need a WAN Tools package
-available from
-
-	ftp.sangoma.com/pub/linux/current_wanpipe/wanpipe-X.Y.Z.tgz
-
-where vX.Y.Z represent the wanpipe version number.
-
-For technical questions and/or comments please e-mail to ncorbic@sangoma.com.
-For general inquiries please contact Sangoma Technologies Inc. by
-
-	Hotline:	1-800-388-2475	(USA and Canada, toll free)
-	Phone:		(905) 474-1990  ext: 106
-	Fax:		(905) 474-9223
-	E-mail:		dm@sangoma.com	(David Mandelstam)
-	WWW:		http://www.sangoma.com
-
-
-INSTALLATION
-
-Please read the WanpipeForLinux.pdf manual on how to 
-install the WANPIPE tools and drivers properly. 
-
-
-After installing wanpipe package: /usr/local/wanrouter/doc. 
-On the ftp.sangoma.com : /linux/current_wanpipe/doc
-
-
-COPYRIGHT AND LICENSING INFORMATION
-
-This program is free software; you can redistribute it and/or modify it under
-the terms of the GNU General Public License as published by the Free Software
-Foundation; either version 2, or (at your option) any later version.
-
-This program is distributed in the hope that it will be useful, but WITHOUT
-ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
-FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
-
-You should have received a copy of the GNU General Public License along with
-this program; if not, write to the Free Software Foundation, Inc., 675 Mass
-Ave, Cambridge, MA 02139, USA.
-
-
-
-ACKNOWLEDGEMENTS
-
-This product is based on the WANPIPE(tm) Multiprotocol WAN Router developed
-by Sangoma Technologies Inc. for Linux 2.0.x and 2.2.x.  Success of the WANPIPE
-together with the next major release of Linux kernel in summer 1996 commanded
-adequate changes to the WANPIPE code to take full advantage of new Linux
-features.
-
-Instead of continuing developing proprietary interface tied to Sangoma WAN
-cards, we decided to separate all hardware-independent code into a separate
-module and defined two levels of interfaces - one for user-level applications
-and another for kernel-level WAN drivers.  WANPIPE is now implemented as a
-WAN driver compliant with the WAN Link Driver interface.  Also a general
-purpose WAN configuration utility and a set of shell scripts was developed to 
-support WAN router at the user level.
-
-Many useful ideas concerning hardware-independent interface implementation
-were given by Mike McLagan <mike.mclagan@linux.org> and his implementation
-of the Frame Relay router and drivers for Sangoma cards (dlci/sdla).
-
-With the new implementation of the APIs being incorporated into the WANPIPE,
-a special thank goes to Alan Cox in providing insight into BSD sockets.
-
-Special thanks to all the WANPIPE users who performed field-testing, reported
-bugs and made valuable comments and suggestions that help us to improve this
-product.
-
-
-
-NEW IN THIS RELEASE
-
-	o Updated the WANCFG utility
-		Calls the pppconfig to configure the PPPD
-		for async connections.
-
-	o Added the PPPCONFIG utility
-		Used to configure the PPPD daemon for the
-		WANPIPE Async PPP and standard serial port.
-		The wancfg calls the pppconfig to configure
-		the pppd.
-
-	o Fixed the PCI autodetect feature.  
-		The SLOT 0 was used as an autodetect option
-		however, some high end PC's slot numbers start
-		from 0. 
-
-	o This release has been tested with the new backupd
-	  daemon release.
-	
-
-PRODUCT COMPONENTS AND RELATED FILES
-
-/etc: (or user defined)
-	wanpipe1.conf	default router configuration file
-
-/lib/modules/X.Y.Z/misc:
-	wanrouter.o	router kernel loadable module
-	af_wanpipe.o	wanpipe api socket module
-
-/lib/modules/X.Y.Z/net:
-	sdladrv.o	Sangoma SDLA support module
-	wanpipe.o	Sangoma WANPIPE(tm) driver module
-
-/proc/net/wanrouter
-	Config		reads current router configuration
-	Status		reads current router status
-	{name}		reads WAN driver statistics
-
-/usr/sbin:
-	wanrouter	wanrouter start-up script
-	wanconfig	wanrouter configuration utility
-	sdladump	WANPIPE adapter memory dump utility
-        fpipemon        Monitor for Frame Relay
-        cpipemon        Monitor for Cisco HDLC
-	ppipemon 	Monitor for PPP
-	xpipemon 	Monitor for X25
-	wpkbdmon        WANPIPE keyboard led monitor/debugger
-
-/usr/local/wanrouter:
-	README		this file
-	COPYING		GNU General Public License
-	Setup		installation script
-	Filelist	distribution definition file
-	wanrouter.rc	meta-configuration file 
-			(used by the Setup and wanrouter script)
-
-/usr/local/wanrouter/doc:
-	wanpipeForLinux.pdf 	WAN Router User's Manual
-
-/usr/local/wanrouter/patches:
-	wanrouter-v2213.gz  	patch for Linux kernels 2.2.11 up to 2.2.13.
-	wanrouter-v2214.gz	patch for Linux kernel 2.2.14. 
-	wanrouter-v2215.gz	patch for Linux kernels 2.2.15 to 2.2.17.
-	wanrouter-v2218.gz	patch for Linux kernels 2.2.18 and up.
-	wanrouter-v240.gz	patch for Linux kernel 2.4.0.  
-	wanrouter-v242.gz	patch for Linux kernel 2.4.2 and up.
-	wanrouter-v2034.gz	patch for Linux kernel 2.0.34
-	wanrouter-v2036.gz 	patch for Linux kernel 2.0.36 and up. 
-
-/usr/local/wanrouter/patches/kdrivers:
-	Sources of the latest WANPIPE device drivers.
-	These are used to UPGRADE the linux kernel to the newest
-	version if the kernel source has already been patched with
-	WANPIPE drivers.
-
-/usr/local/wanrouter/samples:
-	interface	sample interface configuration file
-	wanpipe1.cpri 	CHDLC primary port
-     	wanpipe2.csec 	CHDLC secondary port
-     	wanpipe1.fr   	Frame Relay protocol
-     	wanpipe1.ppp  	PPP protocol ) 
-	wanpipe1.asy	CHDLC ASYNC protocol
-	wanpipe1.x25	X25 protocol
-	wanpipe1.stty	Sync TTY driver (Used by Kernel PPPD daemon)
-	wanpipe1.atty	Async TTY driver (Used by Kernel PPPD daemon)
-	wanrouter.rc	sample meta-configuration file
-
-/usr/local/wanrouter/util:
-	*		wan-tools utilities source code
-
-/usr/local/wanrouter/api/x25:
-	*		x25 api sample programs.
-/usr/local/wanrouter/api/chdlc:
-	*		chdlc api sample programs.
-/usr/local/wanrouter/api/fr:
-	*		fr api sample programs.
-/usr/local/wanrouter/config/wancfg:
-	wancfg		WANPIPE GUI configuration program.
-                        Creates wanpipe#.conf files. 
-/usr/local/wanrouter/config/cfgft1:
-	cfgft1		GUI CSU/DSU configuration program.
-
-/usr/include/linux:
-	wanrouter.h	router API definitions
-	wanpipe.h	WANPIPE API definitions
-	sdladrv.h	SDLA support module API definitions
-	sdlasfm.h	SDLA firmware module definitions
-	if_wanpipe.h	WANPIPE Socket definitions
-	sdlapci.h	WANPIPE PCI definitions
-	
-
-/usr/src/linux/net/wanrouter:
-	*		wanrouter source code
-
-/var/log:
-	wanrouter	wanrouter start-up log (created by the Setup script)
-
-/var/lock:  (or /var/lock/subsys for RedHat)
-	wanrouter	wanrouter lock file (created by the Setup script)
-
-/usr/local/wanrouter/firmware:
-	fr514.sfm	Frame relay firmware for Sangoma S508/S514 card
-	cdual514.sfm	Dual Port Cisco HDLC firmware for Sangoma S508/S514 card
-	ppp514.sfm      PPP Firmware for Sangoma S508 and S514 cards
-	x25_508.sfm	X25 Firmware for Sangoma S508 card.
-
-
-REVISION HISTORY
-
-1.0.0	December 31, 1996	Initial version
-
-1.0.1	January 30, 1997	Status and statistics can be read via /proc
-				filesystem entries.
-
-1.0.2   April 30, 1997          Added UDP management via monitors.
-
-1.0.3	June 3, 1997		UDP management for multiple boards using Frame
-				Relay and PPP
-				Enabled continuous transmission of Configure 
-				Request Packet for PPP (for 508 only)
-				Connection Timeout for PPP changed from 900 to 0
-				Flow Control Problem fixed for Frame Relay
-
-1.0.4	July 10, 1997		S508/FT1 monitoring capability in fpipemon and
-				ppipemon utilities.
-				Configurable TTL for UDP packets.
-				Multicast and Broadcast IP source addresses are
-				silently discarded.
-
-1.0.5	July 28, 1997		Configurable T391,T392,N391,N392,N393 for Frame
-				Relay in router.conf.
-				Configurable Memory Address through router.conf 
-				for Frame Relay, PPP and X.25. (commenting this
- 				out enables auto-detection).
-				Fixed freeing up received buffers using kfree()
- 				for Frame Relay and X.25.
-				Protect sdla_peek() by calling save_flags(),
-				cli() and restore_flags().
-				Changed number of Trace elements from 32 to 20
-				Added DLCI specific data monitoring in FPIPEMON. 
-2.0.0	Nov 07, 1997		Implemented protection of RACE conditions by 
-				critical flags for FRAME RELAY and PPP.
-				DLCI List interrupt mode implemented.
-				IPX support in FRAME RELAY and PPP.
-				IPX Server Support (MARS)
-				More driver specific stats included in FPIPEMON
-				and PIPEMON.
-
-2.0.1	Nov 28, 1997		Bug Fixes for version 2.0.0.
-				Protection of "enable_irq()" while 
-				"disable_irq()" has been enabled from any other
-				routine (for Frame Relay, PPP and X25).
-				Added additional Stats for Fpipemon and Ppipemon
-				Improved Load Sharing for multiple boards
-
-2.0.2	Dec 09, 1997		Support for PAP and CHAP for ppp has been
-				implemented.
-
-2.0.3	Aug 15, 1998		New release supporting Cisco HDLC, CIR for Frame
-				relay, Dynamic IP assignment for PPP and Inverse
-				Arp support for Frame-relay.  Man Pages are 
-				included for better support and a new utility
-				for configuring FT1 cards.
-
-2.0.4	Dec 09, 1998	        Dual Port support for Cisco HDLC.
-				Support for HDLC (LAPB) API.
-				Supports BiSync Streaming code for S502E 
-				and S503 cards.
-				Support for Streaming HDLC API.
-				Provides a BSD socket interface for 
-				creating applications using BiSync
-   				streaming.        
-
-2.0.5   Aug 04, 1999 		CHDLC initialization bug fix.
-				PPP interrupt driven driver: 
-  				Fix to the PPP line hangup problem.
-				New PPP firmware
-				Added comments to the startup SYSTEM ERROR messages
-				Xpipemon debugging application for the X25 protocol
-				New USER_MANUAL.txt
-				Fixed the odd boundary 4byte writes to the board.
-				BiSync Streaming code has been taken out.  
-				 Available as a patch.
-				Streaming HDLC API has been taken out.  
-				 Available as a patch.                 
-
-2.0.6   Aug 17, 1999		Increased debugging in statup scripts
-				Fixed installation bugs from 2.0.5
-				Kernel patch works for both 2.2.10 and 2.2.11 kernels.
-				There is no functional difference between the two packages         
-
-2.0.7   Aug 26, 1999		o  Merged X25API code into WANPIPE.
-				o  Fixed a memory leak for X25API
-				o  Updated the X25API code for 2.2.X kernels.
-				o  Improved NEM handling.   
-
-2.1.0	Oct 25, 1999		o New code for S514 PCI Card
-				o New CHDLC and Frame Relay drivers
-				o PPP and X25 are not supported in this release    
-
-2.1.1	Nov 30, 1999		o PPP support for S514 PCI Cards
-
-2.1.3   Apr 06, 2000		o Socket based x25api 
-				o Socket based chdlc api
-				o Socket based fr api
-				o Dual Port Receive only CHDLC support.
-				o Asynchronous CHDLC support (Secondary Port)
-				o cfgft1 GUI csu/dsu configurator
-				o wancfg GUI configuration file 
-				  configurator.
-				o Architectural directory changes.
-
-beta-2.1.4 Jul 2000		o Dynamic interface configuration:
-					Network interfaces reflect the state
-					of protocol layer.  If the protocol becomes
-					disconnected, driver will bring down
-					the interface.  Once the protocol reconnects
-					the interface will be brought up. 
-					
-					Note: This option is turned off by default.
-
-				o Dynamic wanrouter setup using 'wanconfig':
-					wanconfig utility can be used to
-					shutdown,restart,start or reconfigure 
-					a virtual circuit dynamically.
-				     
-					Frame Relay:  Each DLCI can be: 
-						      created,stopped,restarted and reconfigured
-						      dynamically using wanconfig.
-					
-						      ex: wanconfig card wanpipe1 dev wp1_fr16 up
-				  
-				o Wanrouter startup via command line arguments:
-					wanconfig also supports wanrouter startup via command line
-					arguments.  Thus, there is no need to create a wanpipe#.conf
-					configuration file.  
-
-				o Socket based x25api update/bug fixes.
-					Added support for LCN numbers greater than 255.
-					Option to pass up modem messages.
-					Provided a PCI IRQ check, so a single S514
-					card is guaranteed to have a non-sharing interrupt.
-
-				o Fixes to the wancfg utility.
-				o New FT1 debugging support via *pipemon utilities.
-				o Frame Relay ARP support Enabled.
-
-beta3-2.1.4 Jul 2000		o X25 M_BIT Problem fix.
-				o Added the Multi-Port PPP
-				  Updated utilities for the Multi-Port PPP.
-
-2.1.4	Aut 2000
-				o In X25API:
-					Maximum packet an application can send
-					to the driver has been extended to 4096 bytes.
-
-					Fixed the x25 startup bug. Enable 
-					communications only after all interfaces
-					come up.  HIGH SVC/PVC is used to calculate
-					the number of channels.
-					Enable protocol only after all interfaces
-					are enabled.
-
-				o Added an extra state to the FT1 config, kernel module.
-				o Updated the pipemon debuggers.
-
-				o Blocked the Multi-Port PPP from running on kernels
-				  2.2.16 or greater, due to syncppp kernel module
-				  change. 
-	  
-beta1-2.1.5 	Nov 15 2000
-				o Fixed the MultiPort PPP Support for kernels 2.2.16 and above.
-				  2.2.X kernels only
-
-				o Secured the driver UDP debugging calls
-					- All illegal network debugging calls are reported to
-					  the log.
-					- Defined a set of allowed commands, all other denied.
-					
-				o Cpipemon
-					- Added set FT1 commands to the cpipemon. Thus CSU/DSU
-					  configuration can be performed using cpipemon.
-					  All systems that cannot run cfgft1 GUI utility should
-					  use cpipemon to configure the on board CSU/DSU.
-
-
-				o Keyboard Led Monitor/Debugger
-					- A new utility /usr/sbin/wpkbdmon uses keyboard leds
-					  to convey operational statistic information of the 
-					  Sangoma WANPIPE cards.
-					NUM_LOCK    = Line State  (On=connected,    Off=disconnected)
-					CAPS_LOCK   = Tx data     (On=transmitting, Off=no tx data)
-					SCROLL_LOCK = Rx data     (On=receiving,    Off=no rx data
-					
-				o Hardware probe on module load and dynamic device allocation
-					- During WANPIPE module load, all Sangoma cards are probed
-					  and found information is printed in the /var/log/messages.
-					- If no cards are found, the module load fails.
-					- Appropriate number of devices are dynamically loaded 
-					  based on the number of Sangoma cards found.
-
-					  Note: The kernel configuration option 
-						CONFIG_WANPIPE_CARDS has been taken out.
-					
-				o Fixed the Frame Relay and Chdlc network interfaces so they are
-				  compatible with libpcap libraries.  Meaning, tcpdump, snort,
-				  ethereal, and all other packet sniffers and debuggers work on
-				  all WANPIPE network interfaces.
-					- Set the network interface encoding type to ARPHRD_PPP.
-					  This tell the sniffers that data obtained from the
-					  network interface is in pure IP format.
-				  Fix for 2.2.X kernels only.
-				
-				o True interface encoding option for Frame Relay and CHDLC
-					- The above fix sets the network interface encoding
-					  type to ARPHRD_PPP, however some customers use
-					  the encoding interface type to determine the
-					  protocol running.  Therefore, the TURE ENCODING
-					  option will set the interface type back to the
-					  original value.  
-
-					  NOTE: If this option is used with Frame Relay and CHDLC
-						libpcap library support will be broken.  
-						i.e. tcpdump will not work.
-					Fix for 2.2.x Kernels only.
-						
-				o Ethernet Bridgind over Frame Relay
-					- The Frame Relay bridging has been developed by 
-					  Kristian Hoffmann and Mark Wells.  
-					- The Linux kernel bridge is used to send ethernet 
-					  data over the frame relay links.
-					For 2.2.X Kernels only.
-
-				o Added extensive 2.0.X support. Most new features of
-				  2.1.5 for protocols Frame Relay, PPP and CHDLC are
-				  supported under 2.0.X kernels. 
-
-beta1-2.2.0 	Dec 30 2000
-				o Updated drivers for 2.4.X kernels.
-				o Updated drivers for SMP support.
-				o X25API is now able to share PCI interrupts.
-				o Took out a general polling routine that was used
-				  only by X25API. 
-				o Added appropriate locks to the dynamic reconfiguration
-				  code.
-				o Fixed a bug in the keyboard debug monitor.
-
-beta2-2.2.0	Jan 8 2001
-				o Patches for 2.4.0 kernel
-				o Patches for 2.2.18 kernel
-				o Minor updates to PPP and CHLDC drivers.
-				  Note: No functional difference.
-
-beta3-2.2.9	Jan 10 2001
-				o I missed the 2.2.18 kernel patches in beta2-2.2.0
-				  release.  They are included in this release.
-
-Stable Release
-2.2.0		Feb 01 2001
-				o Bug fix in wancfg GUI configurator.
-					The edit function didn't work properly.
-
-
-bata1-2.2.1	Feb 09 2001
-			o WANPIPE TTY Driver emulation. 
-			  Two modes of operation Sync and Async.
-				Sync: Using the PPPD daemon, kernel SyncPPP layer
-				      and the Wanpipe sync TTY driver: a PPP protocol 
-				      connection can be established via Sangoma adapter, over
-				      a T1 leased line.
-			
-				      The 2.4.0 kernel PPP layer supports MULTILINK
-				      protocol, that can be used to bundle any number of Sangoma
-				      adapters (T1 lines) into one, under a single IP address.
-				      Thus, efficiently obtaining multiple T1 throughput. 
-
-				      NOTE: The remote side must also implement MULTILINK PPP
-					    protocol.
-
-				Async:Using the PPPD daemon, kernel AsyncPPP layer
-				      and the WANPIPE async TTY driver: a PPP protocol
-				      connection can be established via Sangoma adapter and
-				      a modem, over a telephone line.
-
-				      Thus, the WANPIPE async TTY driver simulates a serial
-				      TTY driver that would normally be used to interface the 
-				      MODEM to the linux kernel.
-				
-			o WANPIPE PPP Backup Utility
-				This utility will monitor the state of the PPP T1 line.
-				In case of failure, a dial up connection will be established
-				via pppd daemon, ether via a serial tty driver (serial port), 
-				or a WANPIPE async TTY driver (in case serial port is unavailable).
-				
-				Furthermore, while in dial up mode, the primary PPP T1 link
-				will be monitored for signs of life.  
-
-				If the PPP T1 link comes back to life, the dial up connection
-				will be shutdown and T1 line re-established.
-			
-
-			o New Setup installation script.
-				Option to UPGRADE device drivers if the kernel source has
-				already been patched with WANPIPE.
-
-				Option to COMPILE WANPIPE modules against the currently 
-				running kernel, thus no need for manual kernel and module
-				re-compilation.
-			
-			o Updates and Bug Fixes to wancfg utility.
-
-bata2-2.2.1	Feb 20 2001
-
-			o Bug fixes to the CHDLC device drivers.
-				The driver had compilation problems under kernels
-				2.2.14 or lower.
-
-			o Bug fixes to the Setup installation script.
-				The device drivers compilation options didn't work
-				properly.
-
-			o Update to the wpbackupd daemon.  
-				Optimized the cross-over times, between the primary
-				link and the backup dialup.
-
-beta3-2.2.1	Mar 02 2001
-			o Patches for 2.4.2 kernel.
-
-			o Bug fixes to util/ make files.
-			o Bug fixes to the Setup installation script.
-
-			o Took out the backupd support and made it into
-			  as separate package.
-			  
-beta4-2.2.1     Mar 12 2001
-
-		o Fix to the Frame Relay Device driver.
-			IPSAC sends a packet of zero length
-			header to the frame relay driver.  The
-			driver tries to push its own 2 byte header
-			into the packet, which causes the driver to
-			crash.
-
-		o Fix the WANPIPE re-configuration code.
-			Bug was found by trying to run  the cfgft1 while the
-			interface was already running.  
-
-		o Updates to cfgft1.
-			Writes a wanpipe#.cfgft1 configuration file
-			once the CSU/DSU is configured. This file can
-			holds the current CSU/DSU configuration.
-
-
-
->>>>>> END OF README <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
-
-
diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index e56b739d8e23..b3d9ccde0c27 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -231,7 +231,6 @@ unifdef-y += if_pppol2tp.h
 unifdef-y += if_pppox.h
 unifdef-y += if_tr.h
 unifdef-y += if_vlan.h
-unifdef-y += if_wanpipe.h
 unifdef-y += igmp.h
 unifdef-y += inet_diag.h
 unifdef-y += in.h
diff --git a/include/linux/if_wanpipe.h b/include/linux/if_wanpipe.h
deleted file mode 100644
index e594ca6069e5..000000000000
--- a/include/linux/if_wanpipe.h
+++ /dev/null
@@ -1,124 +0,0 @@
-/*****************************************************************************
-* if_wanpipe.h	Header file for the Sangoma AF_WANPIPE Socket 	
-*
-* Author: 	Nenad Corbic 	
-*
-* Copyright:	(c) 2000 Sangoma Technologies Inc.
-*
-*		This program is free software; you can redistribute it and/or
-*		modify it under the terms of the GNU General Public License
-*		as published by the Free Software Foundation; either version
-*		2 of the License, or (at your option) any later version.
-* ============================================================================
-*
-* Jan 28, 2000	Nenad Corbic 	Initial Version
-*
-*****************************************************************************/
-
-#ifndef __LINUX_IF_WAN_PACKET_H
-#define __LINUX_IF_WAN_PACKET_H
-
-struct wan_sockaddr_ll
-{
-	unsigned short	sll_family;
-	unsigned short	sll_protocol;
-	int		sll_ifindex;
-	unsigned short	sll_hatype;
-	unsigned char	sll_pkttype;
-	unsigned char	sll_halen;
-	unsigned char	sll_addr[8];
-	unsigned char   sll_device[14];
-	unsigned char 	sll_card[14];
-};
-
-typedef struct 
-{
-	unsigned char free;
-	unsigned char state_sk;
-	int rcvbuf;
-	int sndbuf;
-	int rmem;
-	int wmem;
-	int sk_count;
-	unsigned char bound;
-	char name[14];
-	unsigned char d_state;
-	unsigned char svc;
-	unsigned short lcn;
-	unsigned char mbox;
-	unsigned char cmd_busy;
-	unsigned char command;
-	unsigned poll;
-	unsigned poll_cnt;
-	int rblock;	
-} wan_debug_hdr_t;
-
-#define MAX_NUM_DEBUG  10
-#define X25_PROT       0x16
-#define PVC_PROT       0x17	
-
-typedef struct
-{
-	wan_debug_hdr_t debug[MAX_NUM_DEBUG];
-}wan_debug_t;
-
-#define	SIOC_WANPIPE_GET_CALL_DATA	(SIOCPROTOPRIVATE + 0)
-#define	SIOC_WANPIPE_SET_CALL_DATA	(SIOCPROTOPRIVATE + 1)
-#define SIOC_WANPIPE_ACCEPT_CALL	(SIOCPROTOPRIVATE + 2)
-#define SIOC_WANPIPE_CLEAR_CALL	        (SIOCPROTOPRIVATE + 3)
-#define SIOC_WANPIPE_RESET_CALL	        (SIOCPROTOPRIVATE + 4)
-#define SIOC_WANPIPE_DEBUG	        (SIOCPROTOPRIVATE + 5)
-#define SIOC_WANPIPE_SET_NONBLOCK	(SIOCPROTOPRIVATE + 6)
-#define SIOC_WANPIPE_CHECK_TX		(SIOCPROTOPRIVATE + 7)
-#define SIOC_WANPIPE_SOCK_STATE		(SIOCPROTOPRIVATE + 8)
-
-/* Packet types */
-
-#define WAN_PACKET_HOST		0		/* To us		*/
-#define WAN_PACKET_BROADCAST	1		/* To all		*/
-#define WAN_PACKET_MULTICAST	2		/* To group		*/
-#define WAN_PACKET_OTHERHOST	3		/* To someone else 	*/
-#define WAN_PACKET_OUTGOING		4		/* Outgoing of any type */
-/* These ones are invisible by user level */
-#define WAN_PACKET_LOOPBACK		5		/* MC/BRD frame looped back */
-#define WAN_PACKET_FASTROUTE	6		/* Fastrouted frame	*/
-
-
-/* X25 specific */
-#define WAN_PACKET_DATA 	7
-#define WAN_PACKET_CMD 		8
-#define WAN_PACKET_ASYNC	9
-#define WAN_PACKET_ERR	       10
-
-/* Packet socket options */
-
-#define WAN_PACKET_ADD_MEMBERSHIP		1
-#define WAN_PACKET_DROP_MEMBERSHIP		2
-
-#define WAN_PACKET_MR_MULTICAST	0
-#define WAN_PACKET_MR_PROMISC	1
-#define WAN_PACKET_MR_ALLMULTI	2
-
-#ifdef __KERNEL__
-
-/* Private wanpipe socket structures. */
-struct wanpipe_opt
-{
-	void   *mbox;		/* Mail box  */
-	void   *card; 		/* Card bouded to */
-	struct net_device *dev;	/* Bounded device */
-	unsigned short lcn;	/* Binded LCN */
-	unsigned char  svc;	/* 0=pvc, 1=svc */
-	unsigned char  timer;   /* flag for delayed transmit*/	
-	struct timer_list tx_timer;
-	unsigned poll_cnt;
-	unsigned char force;	/* Used to force sock release */
-	atomic_t packet_sent;   
-	unsigned short num; 
-};
-
-#define wp_sk(__sk) ((struct wanpipe_opt *)(__sk)->sk_protinfo)
-
-#endif
-
-#endif
-- 
cgit v1.2.3


From 31efdf0530b6351b0658d35a602a0f2d6bc2ed6f Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Tue, 15 Apr 2008 00:30:16 -0700
Subject: [ISDN] include/linux/isdn.h: remove dead code

This patch remove the usage of a nonexisting kconfig variable.

Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/isdn.h | 6 ------
 1 file changed, 6 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/isdn.h b/include/linux/isdn.h
index 9cb2855bb170..44cd663c53b6 100644
--- a/include/linux/isdn.h
+++ b/include/linux/isdn.h
@@ -16,14 +16,8 @@
 
 #include <linux/ioctl.h>
 
-#ifdef CONFIG_COBALT_MICRO_SERVER
-/* Save memory */
-#define ISDN_MAX_DRIVERS    2
-#define ISDN_MAX_CHANNELS   8
-#else
 #define ISDN_MAX_DRIVERS    32
 #define ISDN_MAX_CHANNELS   64
-#endif
 
 /* New ioctl-codes */
 #define IIOCNETAIF  _IO('I',1)
-- 
cgit v1.2.3


From c50f68c8aea421267ba7995b1c485c281b28add6 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Mon, 24 Mar 2008 20:50:48 +1100
Subject: [LMB] Add lmb_alloc_nid()

A variant of lmb_alloc() that tries to allocate memory on a specified
NUMA node 'nid' but falls back to normal lmb_alloc() if that fails.

The caller provides a 'nid_range' function pointer which assists the
allocator.  It is given args 'start', 'end', and pointer to integer
'this_nid'.

It places at 'this_nid' the NUMA node id that corresponds to 'start',
and returns the end address within 'start' to 'end' at which memory
assosciated with 'nid' ends.

This callback allows a platform to use lmb_alloc_nid() in just
about any context, even ones in which early_pfn_to_nid() might
not be working yet.

This function will be used by the NUMA setup code on sparc64, and also
it can be used by powerpc, replacing it's hand crafted
"careful_allocation()" function in arch/powerpc/mm/numa.c

If x86 ever converts it's NUMA support over to using the LMB helpers,
it can use this too as it has something entirely similar.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 include/linux/lmb.h |  2 ++
 lib/lmb.c           | 86 ++++++++++++++++++++++++++++++++++++++++++++++-------
 2 files changed, 78 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/lmb.h b/include/linux/lmb.h
index 632717c6a2ba..271153d27fba 100644
--- a/include/linux/lmb.h
+++ b/include/linux/lmb.h
@@ -42,6 +42,8 @@ extern void __init lmb_init(void);
 extern void __init lmb_analyze(void);
 extern long __init lmb_add(u64 base, u64 size);
 extern long __init lmb_reserve(u64 base, u64 size);
+extern u64 __init lmb_alloc_nid(u64 size, u64 align, int nid,
+				u64 (*nid_range)(u64, u64, int *));
 extern u64 __init lmb_alloc(u64 size, u64 align);
 extern u64 __init lmb_alloc_base(u64 size,
 		u64, u64 max_addr);
diff --git a/lib/lmb.c b/lib/lmb.c
index 3c43b95fef4a..549fbb3d70cf 100644
--- a/lib/lmb.c
+++ b/lib/lmb.c
@@ -232,6 +232,82 @@ long __init lmb_overlaps_region(struct lmb_region *rgn, u64 base,
 	return (i < rgn->cnt) ? i : -1;
 }
 
+static u64 lmb_align_down(u64 addr, u64 size)
+{
+	return addr & ~(size - 1);
+}
+
+static u64 lmb_align_up(u64 addr, u64 size)
+{
+	return (addr + (size - 1)) & ~(size - 1);
+}
+
+static u64 __init lmb_alloc_nid_unreserved(u64 start, u64 end,
+					   u64 size, u64 align)
+{
+	u64 base;
+	long j;
+
+	base = lmb_align_down((end - size), align);
+	while (start <= base &&
+	       ((j = lmb_overlaps_region(&lmb.reserved, base, size)) >= 0))
+		base = lmb_align_down(lmb.reserved.region[j].base - size,
+				      align);
+
+	if (base != 0 && start <= base) {
+		if (lmb_add_region(&lmb.reserved, base,
+				   lmb_align_up(size, align)) < 0)
+			base = ~(u64)0;
+		return base;
+	}
+
+	return ~(u64)0;
+}
+
+static u64 __init lmb_alloc_nid_region(struct lmb_property *mp,
+				       u64 (*nid_range)(u64, u64, int *),
+				       u64 size, u64 align, int nid)
+{
+	u64 start, end;
+
+	start = mp->base;
+	end = start + mp->size;
+
+	start = lmb_align_up(start, align);
+	while (start < end) {
+		u64 this_end;
+		int this_nid;
+
+		this_end = nid_range(start, end, &this_nid);
+		if (this_nid == nid) {
+			u64 ret = lmb_alloc_nid_unreserved(start, this_end,
+							   size, align);
+			if (ret != ~(u64)0)
+				return ret;
+		}
+		start = this_end;
+	}
+
+	return ~(u64)0;
+}
+
+u64 __init lmb_alloc_nid(u64 size, u64 align, int nid,
+			 u64 (*nid_range)(u64 start, u64 end, int *nid))
+{
+	struct lmb_region *mem = &lmb.memory;
+	int i;
+
+	for (i = 0; i < mem->cnt; i++) {
+		u64 ret = lmb_alloc_nid_region(&mem->region[i],
+					       nid_range,
+					       size, align, nid);
+		if (ret != ~(u64)0)
+			return ret;
+	}
+
+	return lmb_alloc(size, align);
+}
+
 u64 __init lmb_alloc(u64 size, u64 align)
 {
 	return lmb_alloc_base(size, align, LMB_ALLOC_ANYWHERE);
@@ -250,16 +326,6 @@ u64 __init lmb_alloc_base(u64 size, u64 align, u64 max_addr)
 	return alloc;
 }
 
-static u64 lmb_align_down(u64 addr, u64 size)
-{
-	return addr & ~(size - 1);
-}
-
-static u64 lmb_align_up(u64 addr, u64 size)
-{
-	return (addr + (size - 1)) & ~(size - 1);
-}
-
 u64 __init __lmb_alloc_base(u64 size, u64 align, u64 max_addr)
 {
 	long i, j;
-- 
cgit v1.2.3


From 4ac58469f13028e1eb97f8bc7b0fca5072591d8d Mon Sep 17 00:00:00 2001
From: Michael Buesch <mb@bu3sch.de>
Date: Fri, 11 Apr 2008 11:59:00 +0200
Subject: ssb: Fix usage of struct device used for DMAing

This fixes DMA on architectures where DMA is nontrivial, like PPC64.
We must use the host-device's (PCI) struct device for any DMA
operation instead of the SSB device. For this we add a new
struct device pointer to the SSB device structure that will always
point to the right device for DMAing.

Without this patch b43 and b44 drivers won't work on complex-DMA
architectures, that for example need dev->archdata for DMA operations.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---
 drivers/net/b44.c              | 52 +++++++++++++++++++++---------------------
 drivers/net/wireless/b43/dma.c | 27 +++++++++++-----------
 drivers/ssb/main.c             | 14 +++++++-----
 include/linux/ssb/ssb.h        |  4 ++++
 4 files changed, 52 insertions(+), 45 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/b44.c b/drivers/net/b44.c
index 25f1337cd02c..59dce6aa0865 100644
--- a/drivers/net/b44.c
+++ b/drivers/net/b44.c
@@ -148,7 +148,7 @@ static inline void b44_sync_dma_desc_for_device(struct ssb_device *sdev,
 						unsigned long offset,
 						enum dma_data_direction dir)
 {
-	dma_sync_single_range_for_device(sdev->dev, dma_base,
+	dma_sync_single_range_for_device(sdev->dma_dev, dma_base,
 					 offset & dma_desc_align_mask,
 					 dma_desc_sync_size, dir);
 }
@@ -158,7 +158,7 @@ static inline void b44_sync_dma_desc_for_cpu(struct ssb_device *sdev,
 					     unsigned long offset,
 					     enum dma_data_direction dir)
 {
-	dma_sync_single_range_for_cpu(sdev->dev, dma_base,
+	dma_sync_single_range_for_cpu(sdev->dma_dev, dma_base,
 				      offset & dma_desc_align_mask,
 				      dma_desc_sync_size, dir);
 }
@@ -613,7 +613,7 @@ static void b44_tx(struct b44 *bp)
 
 		BUG_ON(skb == NULL);
 
-		dma_unmap_single(bp->sdev->dev,
+		dma_unmap_single(bp->sdev->dma_dev,
 				 rp->mapping,
 				 skb->len,
 				 DMA_TO_DEVICE);
@@ -653,7 +653,7 @@ static int b44_alloc_rx_skb(struct b44 *bp, int src_idx, u32 dest_idx_unmasked)
 	if (skb == NULL)
 		return -ENOMEM;
 
-	mapping = dma_map_single(bp->sdev->dev, skb->data,
+	mapping = dma_map_single(bp->sdev->dma_dev, skb->data,
 				 RX_PKT_BUF_SZ,
 				 DMA_FROM_DEVICE);
 
@@ -663,19 +663,19 @@ static int b44_alloc_rx_skb(struct b44 *bp, int src_idx, u32 dest_idx_unmasked)
 		mapping + RX_PKT_BUF_SZ > DMA_30BIT_MASK) {
 		/* Sigh... */
 		if (!dma_mapping_error(mapping))
-			dma_unmap_single(bp->sdev->dev, mapping,
+			dma_unmap_single(bp->sdev->dma_dev, mapping,
 					RX_PKT_BUF_SZ, DMA_FROM_DEVICE);
 		dev_kfree_skb_any(skb);
 		skb = __netdev_alloc_skb(bp->dev, RX_PKT_BUF_SZ, GFP_ATOMIC|GFP_DMA);
 		if (skb == NULL)
 			return -ENOMEM;
-		mapping = dma_map_single(bp->sdev->dev, skb->data,
+		mapping = dma_map_single(bp->sdev->dma_dev, skb->data,
 					 RX_PKT_BUF_SZ,
 					 DMA_FROM_DEVICE);
 		if (dma_mapping_error(mapping) ||
 			mapping + RX_PKT_BUF_SZ > DMA_30BIT_MASK) {
 			if (!dma_mapping_error(mapping))
-				dma_unmap_single(bp->sdev->dev, mapping, RX_PKT_BUF_SZ,DMA_FROM_DEVICE);
+				dma_unmap_single(bp->sdev->dma_dev, mapping, RX_PKT_BUF_SZ,DMA_FROM_DEVICE);
 			dev_kfree_skb_any(skb);
 			return -ENOMEM;
 		}
@@ -750,7 +750,7 @@ static void b44_recycle_rx(struct b44 *bp, int src_idx, u32 dest_idx_unmasked)
 					     dest_idx * sizeof(dest_desc),
 					     DMA_BIDIRECTIONAL);
 
-	dma_sync_single_for_device(bp->sdev->dev, le32_to_cpu(src_desc->addr),
+	dma_sync_single_for_device(bp->sdev->dma_dev, le32_to_cpu(src_desc->addr),
 				   RX_PKT_BUF_SZ,
 				   DMA_FROM_DEVICE);
 }
@@ -772,7 +772,7 @@ static int b44_rx(struct b44 *bp, int budget)
 		struct rx_header *rh;
 		u16 len;
 
-		dma_sync_single_for_cpu(bp->sdev->dev, map,
+		dma_sync_single_for_cpu(bp->sdev->dma_dev, map,
 					    RX_PKT_BUF_SZ,
 					    DMA_FROM_DEVICE);
 		rh = (struct rx_header *) skb->data;
@@ -806,7 +806,7 @@ static int b44_rx(struct b44 *bp, int budget)
 			skb_size = b44_alloc_rx_skb(bp, cons, bp->rx_prod);
 			if (skb_size < 0)
 				goto drop_it;
-			dma_unmap_single(bp->sdev->dev, map,
+			dma_unmap_single(bp->sdev->dma_dev, map,
 					 skb_size, DMA_FROM_DEVICE);
 			/* Leave out rx_header */
                 	skb_put(skb, len + RX_PKT_OFFSET);
@@ -966,24 +966,24 @@ static int b44_start_xmit(struct sk_buff *skb, struct net_device *dev)
 		goto err_out;
 	}
 
-	mapping = dma_map_single(bp->sdev->dev, skb->data, len, DMA_TO_DEVICE);
+	mapping = dma_map_single(bp->sdev->dma_dev, skb->data, len, DMA_TO_DEVICE);
 	if (dma_mapping_error(mapping) || mapping + len > DMA_30BIT_MASK) {
 		struct sk_buff *bounce_skb;
 
 		/* Chip can't handle DMA to/from >1GB, use bounce buffer */
 		if (!dma_mapping_error(mapping))
-			dma_unmap_single(bp->sdev->dev, mapping, len,
+			dma_unmap_single(bp->sdev->dma_dev, mapping, len,
 					DMA_TO_DEVICE);
 
 		bounce_skb = __dev_alloc_skb(len, GFP_ATOMIC | GFP_DMA);
 		if (!bounce_skb)
 			goto err_out;
 
-		mapping = dma_map_single(bp->sdev->dev, bounce_skb->data,
+		mapping = dma_map_single(bp->sdev->dma_dev, bounce_skb->data,
 					 len, DMA_TO_DEVICE);
 		if (dma_mapping_error(mapping) || mapping + len > DMA_30BIT_MASK) {
 			if (!dma_mapping_error(mapping))
-				dma_unmap_single(bp->sdev->dev, mapping,
+				dma_unmap_single(bp->sdev->dma_dev, mapping,
 					 len, DMA_TO_DEVICE);
 			dev_kfree_skb_any(bounce_skb);
 			goto err_out;
@@ -1082,7 +1082,7 @@ static void b44_free_rings(struct b44 *bp)
 
 		if (rp->skb == NULL)
 			continue;
-		dma_unmap_single(bp->sdev->dev, rp->mapping, RX_PKT_BUF_SZ,
+		dma_unmap_single(bp->sdev->dma_dev, rp->mapping, RX_PKT_BUF_SZ,
 					DMA_FROM_DEVICE);
 		dev_kfree_skb_any(rp->skb);
 		rp->skb = NULL;
@@ -1094,7 +1094,7 @@ static void b44_free_rings(struct b44 *bp)
 
 		if (rp->skb == NULL)
 			continue;
-		dma_unmap_single(bp->sdev->dev, rp->mapping, rp->skb->len,
+		dma_unmap_single(bp->sdev->dma_dev, rp->mapping, rp->skb->len,
 					DMA_TO_DEVICE);
 		dev_kfree_skb_any(rp->skb);
 		rp->skb = NULL;
@@ -1117,12 +1117,12 @@ static void b44_init_rings(struct b44 *bp)
 	memset(bp->tx_ring, 0, B44_TX_RING_BYTES);
 
 	if (bp->flags & B44_FLAG_RX_RING_HACK)
-		dma_sync_single_for_device(bp->sdev->dev, bp->rx_ring_dma,
+		dma_sync_single_for_device(bp->sdev->dma_dev, bp->rx_ring_dma,
 			                  DMA_TABLE_BYTES,
 			                  DMA_BIDIRECTIONAL);
 
 	if (bp->flags & B44_FLAG_TX_RING_HACK)
-		dma_sync_single_for_device(bp->sdev->dev, bp->tx_ring_dma,
+		dma_sync_single_for_device(bp->sdev->dma_dev, bp->tx_ring_dma,
 			                  DMA_TABLE_BYTES,
 			                  DMA_TO_DEVICE);
 
@@ -1144,24 +1144,24 @@ static void b44_free_consistent(struct b44 *bp)
 	bp->tx_buffers = NULL;
 	if (bp->rx_ring) {
 		if (bp->flags & B44_FLAG_RX_RING_HACK) {
-			dma_unmap_single(bp->sdev->dev, bp->rx_ring_dma,
+			dma_unmap_single(bp->sdev->dma_dev, bp->rx_ring_dma,
 					DMA_TABLE_BYTES,
 					DMA_BIDIRECTIONAL);
 			kfree(bp->rx_ring);
 		} else
-			dma_free_coherent(bp->sdev->dev, DMA_TABLE_BYTES,
+			dma_free_coherent(bp->sdev->dma_dev, DMA_TABLE_BYTES,
 					    bp->rx_ring, bp->rx_ring_dma);
 		bp->rx_ring = NULL;
 		bp->flags &= ~B44_FLAG_RX_RING_HACK;
 	}
 	if (bp->tx_ring) {
 		if (bp->flags & B44_FLAG_TX_RING_HACK) {
-			dma_unmap_single(bp->sdev->dev, bp->tx_ring_dma,
+			dma_unmap_single(bp->sdev->dma_dev, bp->tx_ring_dma,
 					DMA_TABLE_BYTES,
 					DMA_TO_DEVICE);
 			kfree(bp->tx_ring);
 		} else
-			dma_free_coherent(bp->sdev->dev, DMA_TABLE_BYTES,
+			dma_free_coherent(bp->sdev->dma_dev, DMA_TABLE_BYTES,
 					    bp->tx_ring, bp->tx_ring_dma);
 		bp->tx_ring = NULL;
 		bp->flags &= ~B44_FLAG_TX_RING_HACK;
@@ -1187,7 +1187,7 @@ static int b44_alloc_consistent(struct b44 *bp, gfp_t gfp)
 		goto out_err;
 
 	size = DMA_TABLE_BYTES;
-	bp->rx_ring = dma_alloc_coherent(bp->sdev->dev, size, &bp->rx_ring_dma, gfp);
+	bp->rx_ring = dma_alloc_coherent(bp->sdev->dma_dev, size, &bp->rx_ring_dma, gfp);
 	if (!bp->rx_ring) {
 		/* Allocation may have failed due to pci_alloc_consistent
 		   insisting on use of GFP_DMA, which is more restrictive
@@ -1199,7 +1199,7 @@ static int b44_alloc_consistent(struct b44 *bp, gfp_t gfp)
 		if (!rx_ring)
 			goto out_err;
 
-		rx_ring_dma = dma_map_single(bp->sdev->dev, rx_ring,
+		rx_ring_dma = dma_map_single(bp->sdev->dma_dev, rx_ring,
 			                    DMA_TABLE_BYTES,
 			                    DMA_BIDIRECTIONAL);
 
@@ -1214,7 +1214,7 @@ static int b44_alloc_consistent(struct b44 *bp, gfp_t gfp)
 		bp->flags |= B44_FLAG_RX_RING_HACK;
 	}
 
-	bp->tx_ring = dma_alloc_coherent(bp->sdev->dev, size, &bp->tx_ring_dma, gfp);
+	bp->tx_ring = dma_alloc_coherent(bp->sdev->dma_dev, size, &bp->tx_ring_dma, gfp);
 	if (!bp->tx_ring) {
 		/* Allocation may have failed due to dma_alloc_coherent
 		   insisting on use of GFP_DMA, which is more restrictive
@@ -1226,7 +1226,7 @@ static int b44_alloc_consistent(struct b44 *bp, gfp_t gfp)
 		if (!tx_ring)
 			goto out_err;
 
-		tx_ring_dma = dma_map_single(bp->sdev->dev, tx_ring,
+		tx_ring_dma = dma_map_single(bp->sdev->dma_dev, tx_ring,
 			                    DMA_TABLE_BYTES,
 			                    DMA_TO_DEVICE);
 
diff --git a/drivers/net/wireless/b43/dma.c b/drivers/net/wireless/b43/dma.c
index 948eb1fe916b..48e912487b16 100644
--- a/drivers/net/wireless/b43/dma.c
+++ b/drivers/net/wireless/b43/dma.c
@@ -373,10 +373,10 @@ static inline
 	dma_addr_t dmaaddr;
 
 	if (tx) {
-		dmaaddr = dma_map_single(ring->dev->dev->dev,
+		dmaaddr = dma_map_single(ring->dev->dev->dma_dev,
 					 buf, len, DMA_TO_DEVICE);
 	} else {
-		dmaaddr = dma_map_single(ring->dev->dev->dev,
+		dmaaddr = dma_map_single(ring->dev->dev->dma_dev,
 					 buf, len, DMA_FROM_DEVICE);
 	}
 
@@ -388,9 +388,10 @@ static inline
 			  dma_addr_t addr, size_t len, int tx)
 {
 	if (tx) {
-		dma_unmap_single(ring->dev->dev->dev, addr, len, DMA_TO_DEVICE);
+		dma_unmap_single(ring->dev->dev->dma_dev,
+				 addr, len, DMA_TO_DEVICE);
 	} else {
-		dma_unmap_single(ring->dev->dev->dev,
+		dma_unmap_single(ring->dev->dev->dma_dev,
 				 addr, len, DMA_FROM_DEVICE);
 	}
 }
@@ -400,7 +401,7 @@ static inline
 				 dma_addr_t addr, size_t len)
 {
 	B43_WARN_ON(ring->tx);
-	dma_sync_single_for_cpu(ring->dev->dev->dev,
+	dma_sync_single_for_cpu(ring->dev->dev->dma_dev,
 				addr, len, DMA_FROM_DEVICE);
 }
 
@@ -409,7 +410,7 @@ static inline
 				    dma_addr_t addr, size_t len)
 {
 	B43_WARN_ON(ring->tx);
-	dma_sync_single_for_device(ring->dev->dev->dev,
+	dma_sync_single_for_device(ring->dev->dev->dma_dev,
 				   addr, len, DMA_FROM_DEVICE);
 }
 
@@ -425,7 +426,7 @@ static inline
 
 static int alloc_ringmemory(struct b43_dmaring *ring)
 {
-	struct device *dev = ring->dev->dev->dev;
+	struct device *dma_dev = ring->dev->dev->dma_dev;
 	gfp_t flags = GFP_KERNEL;
 
 	/* The specs call for 4K buffers for 30- and 32-bit DMA with 4K
@@ -439,7 +440,7 @@ static int alloc_ringmemory(struct b43_dmaring *ring)
 	 */
 	if (ring->type == B43_DMA_64BIT)
 		flags |= GFP_DMA;
-	ring->descbase = dma_alloc_coherent(dev, B43_DMA_RINGMEMSIZE,
+	ring->descbase = dma_alloc_coherent(dma_dev, B43_DMA_RINGMEMSIZE,
 					    &(ring->dmabase), flags);
 	if (!ring->descbase) {
 		b43err(ring->dev->wl, "DMA ringmemory allocation failed\n");
@@ -452,9 +453,9 @@ static int alloc_ringmemory(struct b43_dmaring *ring)
 
 static void free_ringmemory(struct b43_dmaring *ring)
 {
-	struct device *dev = ring->dev->dev->dev;
+	struct device *dma_dev = ring->dev->dev->dma_dev;
 
-	dma_free_coherent(dev, B43_DMA_RINGMEMSIZE,
+	dma_free_coherent(dma_dev, B43_DMA_RINGMEMSIZE,
 			  ring->descbase, ring->dmabase);
 }
 
@@ -854,7 +855,7 @@ struct b43_dmaring *b43_setup_dmaring(struct b43_wldev *dev,
 			goto err_kfree_meta;
 
 		/* test for ability to dma to txhdr_cache */
-		dma_test = dma_map_single(dev->dev->dev,
+		dma_test = dma_map_single(dev->dev->dma_dev,
 					  ring->txhdr_cache,
 					  b43_txhdr_size(dev),
 					  DMA_TO_DEVICE);
@@ -869,7 +870,7 @@ struct b43_dmaring *b43_setup_dmaring(struct b43_wldev *dev,
 			if (!ring->txhdr_cache)
 				goto err_kfree_meta;
 
-			dma_test = dma_map_single(dev->dev->dev,
+			dma_test = dma_map_single(dev->dev->dma_dev,
 						  ring->txhdr_cache,
 						  b43_txhdr_size(dev),
 						  DMA_TO_DEVICE);
@@ -883,7 +884,7 @@ struct b43_dmaring *b43_setup_dmaring(struct b43_wldev *dev,
 			}
 		}
 
-		dma_unmap_single(dev->dev->dev,
+		dma_unmap_single(dev->dev->dma_dev,
 				 dma_test, b43_txhdr_size(dev),
 				 DMA_TO_DEVICE);
 	}
diff --git a/drivers/ssb/main.c b/drivers/ssb/main.c
index 72017bf2e577..8003a9e55ac4 100644
--- a/drivers/ssb/main.c
+++ b/drivers/ssb/main.c
@@ -436,15 +436,18 @@ static int ssb_devices_register(struct ssb_bus *bus)
 #ifdef CONFIG_SSB_PCIHOST
 			sdev->irq = bus->host_pci->irq;
 			dev->parent = &bus->host_pci->dev;
+			sdev->dma_dev = &bus->host_pci->dev;
 #endif
 			break;
 		case SSB_BUSTYPE_PCMCIA:
 #ifdef CONFIG_SSB_PCMCIAHOST
 			sdev->irq = bus->host_pcmcia->irq.AssignedIRQ;
 			dev->parent = &bus->host_pcmcia->dev;
+			sdev->dma_dev = &bus->host_pcmcia->dev;
 #endif
 			break;
 		case SSB_BUSTYPE_SSB:
+			sdev->dma_dev = dev;
 			break;
 		}
 
@@ -1018,15 +1021,14 @@ EXPORT_SYMBOL(ssb_dma_translation);
 
 int ssb_dma_set_mask(struct ssb_device *ssb_dev, u64 mask)
 {
-	struct device *dev = ssb_dev->dev;
+	struct device *dma_dev = ssb_dev->dma_dev;
 
 #ifdef CONFIG_SSB_PCIHOST
-	if (ssb_dev->bus->bustype == SSB_BUSTYPE_PCI &&
-	    !dma_supported(dev, mask))
-		return -EIO;
+	if (ssb_dev->bus->bustype == SSB_BUSTYPE_PCI)
+		return dma_set_mask(dma_dev, mask);
 #endif
-	dev->coherent_dma_mask = mask;
-	dev->dma_mask = &dev->coherent_dma_mask;
+	dma_dev->coherent_dma_mask = mask;
+	dma_dev->dma_mask = &dma_dev->coherent_dma_mask;
 
 	return 0;
 }
diff --git a/include/linux/ssb/ssb.h b/include/linux/ssb/ssb.h
index 20add65215af..db53defde5ee 100644
--- a/include/linux/ssb/ssb.h
+++ b/include/linux/ssb/ssb.h
@@ -129,6 +129,10 @@ struct ssb_device {
 	const struct ssb_bus_ops *ops;
 
 	struct device *dev;
+	/* Pointer to the device that has to be used for
+	 * any DMA related operation. */
+	struct device *dma_dev;
+
 	struct ssb_bus *bus;
 	struct ssb_device_id id;
 
-- 
cgit v1.2.3


From 335e92e8a515420bd47a6b0f01cb9a206c0ed6e4 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Tue, 15 Apr 2008 14:34:43 -0700
Subject: vfs: fix possible deadlock in ext2, ext3, ext4 when using xattrs

mb_cache_entry_alloc() was allocating cache entries with GFP_KERNEL.  But
filesystems are calling this function while holding xattr_sem so possible
recursion into the fs violates locking ordering of xattr_sem and transaction
start / i_mutex for ext2-4.  Change mb_cache_entry_alloc() so that filesystems
can specify desired gfp mask and use GFP_NOFS from all of them.

Signed-off-by: Jan Kara <jack@suse.cz>
Reported-by: Dave Jones <davej@redhat.com>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 fs/ext2/xattr.c         | 2 +-
 fs/ext3/xattr.c         | 2 +-
 fs/ext4/xattr.c         | 2 +-
 fs/mbcache.c            | 4 ++--
 include/linux/mbcache.h | 2 +-
 5 files changed, 6 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c
index 3e8683dbb13f..a99d46f3b26e 100644
--- a/fs/ext2/xattr.c
+++ b/fs/ext2/xattr.c
@@ -835,7 +835,7 @@ ext2_xattr_cache_insert(struct buffer_head *bh)
 	struct mb_cache_entry *ce;
 	int error;
 
-	ce = mb_cache_entry_alloc(ext2_xattr_cache);
+	ce = mb_cache_entry_alloc(ext2_xattr_cache, GFP_NOFS);
 	if (!ce)
 		return -ENOMEM;
 	error = mb_cache_entry_insert(ce, bh->b_bdev, bh->b_blocknr, &hash);
diff --git a/fs/ext3/xattr.c b/fs/ext3/xattr.c
index a6ea4d6a8bb2..42856541e9a5 100644
--- a/fs/ext3/xattr.c
+++ b/fs/ext3/xattr.c
@@ -1126,7 +1126,7 @@ ext3_xattr_cache_insert(struct buffer_head *bh)
 	struct mb_cache_entry *ce;
 	int error;
 
-	ce = mb_cache_entry_alloc(ext3_xattr_cache);
+	ce = mb_cache_entry_alloc(ext3_xattr_cache, GFP_NOFS);
 	if (!ce) {
 		ea_bdebug(bh, "out of memory");
 		return;
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index d7962139c010..e9054c1c7d93 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -1386,7 +1386,7 @@ ext4_xattr_cache_insert(struct buffer_head *bh)
 	struct mb_cache_entry *ce;
 	int error;
 
-	ce = mb_cache_entry_alloc(ext4_xattr_cache);
+	ce = mb_cache_entry_alloc(ext4_xattr_cache, GFP_NOFS);
 	if (!ce) {
 		ea_bdebug(bh, "out of memory");
 		return;
diff --git a/fs/mbcache.c b/fs/mbcache.c
index eb31b73e7d69..ec88ff3d04a9 100644
--- a/fs/mbcache.c
+++ b/fs/mbcache.c
@@ -399,11 +399,11 @@ mb_cache_destroy(struct mb_cache *cache)
  * if no more memory was available.
  */
 struct mb_cache_entry *
-mb_cache_entry_alloc(struct mb_cache *cache)
+mb_cache_entry_alloc(struct mb_cache *cache, gfp_t gfp_flags)
 {
 	struct mb_cache_entry *ce;
 
-	ce = kmem_cache_alloc(cache->c_entry_cache, GFP_KERNEL);
+	ce = kmem_cache_alloc(cache->c_entry_cache, gfp_flags);
 	if (ce) {
 		atomic_inc(&cache->c_entry_count);
 		INIT_LIST_HEAD(&ce->e_lru_list);
diff --git a/include/linux/mbcache.h b/include/linux/mbcache.h
index 99e044b4efc6..a09b84e4fdb4 100644
--- a/include/linux/mbcache.h
+++ b/include/linux/mbcache.h
@@ -34,7 +34,7 @@ void mb_cache_destroy(struct mb_cache *);
 
 /* Functions on cache entries */
 
-struct mb_cache_entry *mb_cache_entry_alloc(struct mb_cache *);
+struct mb_cache_entry *mb_cache_entry_alloc(struct mb_cache *, gfp_t);
 int mb_cache_entry_insert(struct mb_cache_entry *, struct block_device *,
 			  sector_t, unsigned int[]);
 void mb_cache_entry_release(struct mb_cache_entry *);
-- 
cgit v1.2.3


From 5f1a3f2ac4ddf966cd0555dc445c6df65709c580 Mon Sep 17 00:00:00 2001
From: Krzysztof Helt <krzysztof.h1@wp.pl>
Date: Tue, 15 Apr 2008 14:34:47 -0700
Subject: acpi thermal trip points increased to 12

The THERMAL_MAX_TRIPS value is set to 10.  It is too few for the Compaq AP550
machine which has 12 trip points.

Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Cc: Len Brown <lenb@kernel.org>
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 drivers/thermal/thermal.c | 4 ++++
 include/linux/thermal.h   | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/thermal/thermal.c b/drivers/thermal/thermal.c
index 8b86e53ccf7a..7f79bbf652d7 100644
--- a/drivers/thermal/thermal.c
+++ b/drivers/thermal/thermal.c
@@ -196,6 +196,10 @@ static struct device_attribute trip_point_attrs[] = {
 	__ATTR(trip_point_8_temp, 0444, trip_point_temp_show, NULL),
 	__ATTR(trip_point_9_type, 0444, trip_point_type_show, NULL),
 	__ATTR(trip_point_9_temp, 0444, trip_point_temp_show, NULL),
+	__ATTR(trip_point_10_type, 0444, trip_point_type_show, NULL),
+	__ATTR(trip_point_10_temp, 0444, trip_point_temp_show, NULL),
+	__ATTR(trip_point_11_type, 0444, trip_point_type_show, NULL),
+	__ATTR(trip_point_11_temp, 0444, trip_point_temp_show, NULL),
 };
 
 #define TRIP_POINT_ATTR_ADD(_dev, _index, result)     \
diff --git a/include/linux/thermal.h b/include/linux/thermal.h
index 818ca1cf0b6d..90c1c191ea69 100644
--- a/include/linux/thermal.h
+++ b/include/linux/thermal.h
@@ -50,7 +50,7 @@ struct thermal_cooling_device_ops {
 };
 
 #define THERMAL_TRIPS_NONE -1
-#define THERMAL_MAX_TRIPS 10
+#define THERMAL_MAX_TRIPS 12
 #define THERMAL_NAME_LENGTH 20
 struct thermal_cooling_device {
 	int id;
-- 
cgit v1.2.3


From a9fde2607895667823e9d1172fc193087125ef68 Mon Sep 17 00:00:00 2001
From: Pavel Emelyanov <xemul@openvz.org>
Date: Wed, 16 Apr 2008 00:48:04 -0700
Subject: [VLAN]: Tag vlan_group_device with net device, not ifindex.

Currently vlan group is searched using one key - the ifindex.
We'll have to lookup the vlan_group by two keys - ifindex and
net. Turning the vlan_group lookup key to struct net_device
pointer will make this process easier.

Besides, this will eliminate one more place in the networking,
that assumes that indexes are unique in the kernel.

Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/if_vlan.h |  4 +++-
 net/8021q/vlan.c        | 22 +++++++++++-----------
 2 files changed, 14 insertions(+), 12 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h
index edd55af7ebd6..15ace02b7b24 100644
--- a/include/linux/if_vlan.h
+++ b/include/linux/if_vlan.h
@@ -81,7 +81,9 @@ extern void vlan_ioctl_set(int (*hook)(struct net *, void __user *));
 #define VLAN_GROUP_ARRAY_PART_LEN     (VLAN_GROUP_ARRAY_LEN/VLAN_GROUP_ARRAY_SPLIT_PARTS)
 
 struct vlan_group {
-	int real_dev_ifindex; /* The ifindex of the ethernet(like) device the vlan is attached to. */
+	struct net_device	*real_dev; /* The ethernet(like) device
+					    * the vlan is attached to.
+					    */
 	unsigned int		nr_vlans;
 	struct hlist_node	hlist;	/* linked list */
 	struct net_device **vlan_devices_arrays[VLAN_GROUP_ARRAY_SPLIT_PARTS];
diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
index 5975ec3be7f3..cf8d810a130d 100644
--- a/net/8021q/vlan.c
+++ b/net/8021q/vlan.c
@@ -65,14 +65,14 @@ static inline unsigned int vlan_grp_hashfn(unsigned int idx)
 }
 
 /* Must be invoked with RCU read lock (no preempt) */
-static struct vlan_group *__vlan_find_group(int real_dev_ifindex)
+static struct vlan_group *__vlan_find_group(struct net_device *real_dev)
 {
 	struct vlan_group *grp;
 	struct hlist_node *n;
-	int hash = vlan_grp_hashfn(real_dev_ifindex);
+	int hash = vlan_grp_hashfn(real_dev->ifindex);
 
 	hlist_for_each_entry_rcu(grp, n, &vlan_group_hash[hash], hlist) {
-		if (grp->real_dev_ifindex == real_dev_ifindex)
+		if (grp->real_dev == real_dev)
 			return grp;
 	}
 
@@ -86,7 +86,7 @@ static struct vlan_group *__vlan_find_group(int real_dev_ifindex)
 struct net_device *__find_vlan_dev(struct net_device *real_dev,
 				   unsigned short VID)
 {
-	struct vlan_group *grp = __vlan_find_group(real_dev->ifindex);
+	struct vlan_group *grp = __vlan_find_group(real_dev);
 
 	if (grp)
 		return vlan_group_get_device(grp, VID);
@@ -103,7 +103,7 @@ static void vlan_group_free(struct vlan_group *grp)
 	kfree(grp);
 }
 
-static struct vlan_group *vlan_group_alloc(int ifindex)
+static struct vlan_group *vlan_group_alloc(struct net_device *real_dev)
 {
 	struct vlan_group *grp;
 
@@ -111,9 +111,9 @@ static struct vlan_group *vlan_group_alloc(int ifindex)
 	if (!grp)
 		return NULL;
 
-	grp->real_dev_ifindex = ifindex;
+	grp->real_dev = real_dev;
 	hlist_add_head_rcu(&grp->hlist,
-			   &vlan_group_hash[vlan_grp_hashfn(ifindex)]);
+			&vlan_group_hash[vlan_grp_hashfn(real_dev->ifindex)]);
 	return grp;
 }
 
@@ -151,7 +151,7 @@ void unregister_vlan_dev(struct net_device *dev)
 
 	ASSERT_RTNL();
 
-	grp = __vlan_find_group(real_dev->ifindex);
+	grp = __vlan_find_group(real_dev);
 	BUG_ON(!grp);
 
 	vlan_proc_rem_dev(dev);
@@ -246,9 +246,9 @@ int register_vlan_dev(struct net_device *dev)
 	struct vlan_group *grp, *ngrp = NULL;
 	int err;
 
-	grp = __vlan_find_group(real_dev->ifindex);
+	grp = __vlan_find_group(real_dev);
 	if (!grp) {
-		ngrp = grp = vlan_group_alloc(real_dev->ifindex);
+		ngrp = grp = vlan_group_alloc(real_dev);
 		if (!grp)
 			return -ENOBUFS;
 	}
@@ -412,7 +412,7 @@ static int vlan_device_event(struct notifier_block *unused, unsigned long event,
 		goto out;
 	}
 
-	grp = __vlan_find_group(dev->ifindex);
+	grp = __vlan_find_group(dev);
 	if (!grp)
 		goto out;
 
-- 
cgit v1.2.3


From f3005d7f4abe03ad41af33b1548602cd086d86a2 Mon Sep 17 00:00:00 2001
From: "Denis V. Lunev" <den@openvz.org>
Date: Wed, 16 Apr 2008 02:02:18 -0700
Subject: [NETNS]: Add netns refcnt debug for network devices.

dev_set_net is called for
- just allocated devices
- devices moving from one namespace to another
release_net has proper check inside to distinguish these cases.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 3 ++-
 net/core/dev.c            | 2 ++
 2 files changed, 4 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 8b17ed40dea2..7c1d4466583b 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -758,7 +758,8 @@ static inline
 void dev_net_set(struct net_device *dev, struct net *net)
 {
 #ifdef CONFIG_NET_NS
-	dev->nd_net = net;
+	release_net(dev->nd_net);
+	dev->nd_net = hold_net(net);
 #endif
 }
 
diff --git a/net/core/dev.c b/net/core/dev.c
index 7aa01125287e..77530e9a34fc 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4042,6 +4042,8 @@ EXPORT_SYMBOL(alloc_netdev_mq);
  */
 void free_netdev(struct net_device *dev)
 {
+	release_net(dev_net(dev));
+
 	/*  Compatibility with error handling in drivers */
 	if (dev->reg_state == NETREG_UNINITIALIZED) {
 		kfree((char *)dev - dev->padded);
-- 
cgit v1.2.3


From 863fbf4966a7ac301a4077e4a04d73e8abfdd7b2 Mon Sep 17 00:00:00 2001
From: Anton Vorontsov <avorontsov@ru.mvista.com>
Date: Fri, 11 Apr 2008 23:06:45 +1000
Subject: [POWERPC] OF helpers for the GPIO API

This implements various helpers to support OF bindings for the GPIO
LIB API.

Previously this was PowerPC specific, but it seems this code isn't
arch-dependent anyhow, so let's place it into of/.

SPARC will not see this addition yet, real hardware seem to not use
GPIOs at all. But this might change:

   http://www.leox.org/docs/faq_MLleon.html

"16-bit I/O port" sounds promising. :-)

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 drivers/of/Kconfig      |   6 ++
 drivers/of/Makefile     |   1 +
 drivers/of/gpio.c       | 242 ++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/of_gpio.h |  69 ++++++++++++++
 4 files changed, 318 insertions(+)
 create mode 100644 drivers/of/gpio.c
 create mode 100644 include/linux/of_gpio.h

(limited to 'include/linux')

diff --git a/drivers/of/Kconfig b/drivers/of/Kconfig
index c03072b12f42..3354ad766a49 100644
--- a/drivers/of/Kconfig
+++ b/drivers/of/Kconfig
@@ -1,3 +1,9 @@
 config OF_DEVICE
 	def_bool y
 	depends on OF && (SPARC || PPC_OF)
+
+config OF_GPIO
+	def_bool y
+	depends on OF && PPC_OF && HAVE_GPIO_LIB
+	help
+	  OpenFirmware GPIO accessors
diff --git a/drivers/of/Makefile b/drivers/of/Makefile
index ab9be5d5255b..5a61f70b4027 100644
--- a/drivers/of/Makefile
+++ b/drivers/of/Makefile
@@ -1,2 +1,3 @@
 obj-y = base.o
 obj-$(CONFIG_OF_DEVICE) += device.o platform.o
+obj-$(CONFIG_OF_GPIO)   += gpio.o
diff --git a/drivers/of/gpio.c b/drivers/of/gpio.c
new file mode 100644
index 000000000000..000681e98f2c
--- /dev/null
+++ b/drivers/of/gpio.c
@@ -0,0 +1,242 @@
+/*
+ * OF helpers for the GPIO API
+ *
+ * Copyright (c) 2007-2008  MontaVista Software, Inc.
+ *
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <asm/prom.h>
+
+/**
+ * of_get_gpio - Get a GPIO number from the device tree to use with GPIO API
+ * @np:		device node to get GPIO from
+ * @index:	index of the GPIO
+ *
+ * Returns GPIO number to use with Linux generic GPIO API, or one of the errno
+ * value on the error condition.
+ */
+int of_get_gpio(struct device_node *np, int index)
+{
+	int ret = -EINVAL;
+	struct device_node *gc;
+	struct of_gpio_chip *of_gc = NULL;
+	int size;
+	const u32 *gpios;
+	u32 nr_cells;
+	int i;
+	const void *gpio_spec;
+	const u32 *gpio_cells;
+	int gpio_index = 0;
+
+	gpios = of_get_property(np, "gpios", &size);
+	if (!gpios) {
+		ret = -ENOENT;
+		goto err0;
+	}
+	nr_cells = size / sizeof(u32);
+
+	for (i = 0; i < nr_cells; gpio_index++) {
+		const phandle *gpio_phandle;
+
+		gpio_phandle = gpios + i;
+		gpio_spec = gpio_phandle + 1;
+
+		/* one cell hole in the gpios = <>; */
+		if (!*gpio_phandle) {
+			if (gpio_index == index)
+				return -ENOENT;
+			i++;
+			continue;
+		}
+
+		gc = of_find_node_by_phandle(*gpio_phandle);
+		if (!gc) {
+			pr_debug("%s: could not find phandle for gpios\n",
+				 np->full_name);
+			goto err0;
+		}
+
+		of_gc = gc->data;
+		if (!of_gc) {
+			pr_debug("%s: gpio controller %s isn't registered\n",
+				 np->full_name, gc->full_name);
+			goto err1;
+		}
+
+		gpio_cells = of_get_property(gc, "#gpio-cells", &size);
+		if (!gpio_cells || size != sizeof(*gpio_cells) ||
+				*gpio_cells != of_gc->gpio_cells) {
+			pr_debug("%s: wrong #gpio-cells for %s\n",
+				 np->full_name, gc->full_name);
+			goto err1;
+		}
+
+		/* Next phandle is at phandle cells + #gpio-cells */
+		i += sizeof(*gpio_phandle) / sizeof(u32) + *gpio_cells;
+		if (i >= nr_cells + 1) {
+			pr_debug("%s: insufficient gpio-spec length\n",
+				 np->full_name);
+			goto err1;
+		}
+
+		if (gpio_index == index)
+			break;
+
+		of_gc = NULL;
+		of_node_put(gc);
+	}
+
+	if (!of_gc) {
+		ret = -ENOENT;
+		goto err0;
+	}
+
+	ret = of_gc->xlate(of_gc, np, gpio_spec);
+	if (ret < 0)
+		goto err1;
+
+	ret += of_gc->gc.base;
+err1:
+	of_node_put(gc);
+err0:
+	pr_debug("%s exited with status %d\n", __func__, ret);
+	return ret;
+}
+EXPORT_SYMBOL(of_get_gpio);
+
+/**
+ * of_gpio_simple_xlate - translate gpio_spec to the GPIO number
+ * @of_gc:	pointer to the of_gpio_chip structure
+ * @np:		device node of the GPIO chip
+ * @gpio_spec:	gpio specifier as found in the device tree
+ *
+ * This is simple translation function, suitable for the most 1:1 mapped
+ * gpio chips. This function performs only one sanity check: whether gpio
+ * is less than ngpios (that is specified in the gpio_chip).
+ */
+int of_gpio_simple_xlate(struct of_gpio_chip *of_gc, struct device_node *np,
+			 const void *gpio_spec)
+{
+	const u32 *gpio = gpio_spec;
+
+	if (*gpio > of_gc->gc.ngpio)
+		return -EINVAL;
+
+	return *gpio;
+}
+EXPORT_SYMBOL(of_gpio_simple_xlate);
+
+/* Should be sufficient for now, later we'll use dynamic bases. */
+#if defined(CONFIG_PPC32) || defined(CONFIG_SPARC32)
+#define GPIOS_PER_CHIP 32
+#else
+#define GPIOS_PER_CHIP 64
+#endif
+
+static int of_get_gpiochip_base(struct device_node *np)
+{
+	struct device_node *gc = NULL;
+	int gpiochip_base = 0;
+
+	while ((gc = of_find_all_nodes(gc))) {
+		if (!of_get_property(gc, "gpio-controller", NULL))
+			continue;
+
+		if (gc != np) {
+			gpiochip_base += GPIOS_PER_CHIP;
+			continue;
+		}
+
+		of_node_put(gc);
+
+		if (gpiochip_base >= ARCH_NR_GPIOS)
+			return -ENOSPC;
+
+		return gpiochip_base;
+	}
+
+	return -ENOENT;
+}
+
+/**
+ * of_mm_gpiochip_add - Add memory mapped GPIO chip (bank)
+ * @np:		device node of the GPIO chip
+ * @mm_gc:	pointer to the of_mm_gpio_chip allocated structure
+ *
+ * To use this function you should allocate and fill mm_gc with:
+ *
+ * 1) In the gpio_chip structure:
+ *    - all the callbacks
+ *
+ * 2) In the of_gpio_chip structure:
+ *    - gpio_cells
+ *    - xlate callback (optional)
+ *
+ * 3) In the of_mm_gpio_chip structure:
+ *    - save_regs callback (optional)
+ *
+ * If succeeded, this function will map bank's memory and will
+ * do all necessary work for you. Then you'll able to use .regs
+ * to manage GPIOs from the callbacks.
+ */
+int of_mm_gpiochip_add(struct device_node *np,
+		       struct of_mm_gpio_chip *mm_gc)
+{
+	int ret = -ENOMEM;
+	struct of_gpio_chip *of_gc = &mm_gc->of_gc;
+	struct gpio_chip *gc = &of_gc->gc;
+
+	gc->label = kstrdup(np->full_name, GFP_KERNEL);
+	if (!gc->label)
+		goto err0;
+
+	mm_gc->regs = of_iomap(np, 0);
+	if (!mm_gc->regs)
+		goto err1;
+
+	gc->base = of_get_gpiochip_base(np);
+	if (gc->base < 0) {
+		ret = gc->base;
+		goto err1;
+	}
+
+	if (!of_gc->xlate)
+		of_gc->xlate = of_gpio_simple_xlate;
+
+	if (mm_gc->save_regs)
+		mm_gc->save_regs(mm_gc);
+
+	np->data = of_gc;
+
+	ret = gpiochip_add(gc);
+	if (ret)
+		goto err2;
+
+	/* We don't want to lose the node and its ->data */
+	of_node_get(np);
+
+	pr_debug("%s: registered as generic GPIO chip, base is %d\n",
+		 np->full_name, gc->base);
+	return 0;
+err2:
+	np->data = NULL;
+	iounmap(mm_gc->regs);
+err1:
+	kfree(gc->label);
+err0:
+	pr_err("%s: GPIO chip registration failed with status %d\n",
+	       np->full_name, ret);
+	return ret;
+}
+EXPORT_SYMBOL(of_mm_gpiochip_add);
diff --git a/include/linux/of_gpio.h b/include/linux/of_gpio.h
new file mode 100644
index 000000000000..2ee97e9877a7
--- /dev/null
+++ b/include/linux/of_gpio.h
@@ -0,0 +1,69 @@
+/*
+ * OF helpers for the GPIO API
+ *
+ * Copyright (c) 2007-2008  MontaVista Software, Inc.
+ *
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#ifndef __LINUX_OF_GPIO_H
+#define __LINUX_OF_GPIO_H
+
+#include <linux/errno.h>
+#include <asm/gpio.h>
+
+#ifdef CONFIG_OF_GPIO
+
+/*
+ * Generic OF GPIO chip
+ */
+struct of_gpio_chip {
+	struct gpio_chip gc;
+	int gpio_cells;
+	int (*xlate)(struct of_gpio_chip *of_gc, struct device_node *np,
+		     const void *gpio_spec);
+};
+
+static inline struct of_gpio_chip *to_of_gpio_chip(struct gpio_chip *gc)
+{
+	return container_of(gc, struct of_gpio_chip, gc);
+}
+
+/*
+ * OF GPIO chip for memory mapped banks
+ */
+struct of_mm_gpio_chip {
+	struct of_gpio_chip of_gc;
+	void (*save_regs)(struct of_mm_gpio_chip *mm_gc);
+	void __iomem *regs;
+};
+
+static inline struct of_mm_gpio_chip *to_of_mm_gpio_chip(struct gpio_chip *gc)
+{
+	struct of_gpio_chip *of_gc = to_of_gpio_chip(gc);
+
+	return container_of(of_gc, struct of_mm_gpio_chip, of_gc);
+}
+
+extern int of_get_gpio(struct device_node *np, int index);
+extern int of_mm_gpiochip_add(struct device_node *np,
+			      struct of_mm_gpio_chip *mm_gc);
+extern int of_gpio_simple_xlate(struct of_gpio_chip *of_gc,
+				struct device_node *np,
+				const void *gpio_spec);
+#else
+
+/* Drivers may not strictly depend on the GPIO support, so let them link. */
+static inline int of_get_gpio(struct device_node *np, int index)
+{
+	return -ENOSYS;
+}
+
+#endif /* CONFIG_OF_GPIO */
+
+#endif /* __LINUX_OF_GPIO_H */
-- 
cgit v1.2.3


From 612212a3f2f053ea68ce9cd16d3deeca7754e8c9 Mon Sep 17 00:00:00 2001
From: Jochen Friedrich <jochen@scram.de>
Date: Sat, 12 Apr 2008 05:22:35 +1000
Subject: [POWERPC] i2c: OF helpers for the i2c API

This implements various helpers to support OF bindings for the i2c
API.

Signed-off-by: Jochen Friedrich <jochen@scram.de>
Acked-by: David S. Miller <davem@davemloft.net>
Acked-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
 drivers/of/Kconfig     |   6 +++
 drivers/of/Makefile    |   1 +
 drivers/of/i2c.c       | 115 +++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/of_i2c.h |  24 +++++++++++
 4 files changed, 146 insertions(+)
 create mode 100644 drivers/of/i2c.c
 create mode 100644 include/linux/of_i2c.h

(limited to 'include/linux')

diff --git a/drivers/of/Kconfig b/drivers/of/Kconfig
index 3354ad766a49..7c305317f372 100644
--- a/drivers/of/Kconfig
+++ b/drivers/of/Kconfig
@@ -7,3 +7,9 @@ config OF_GPIO
 	depends on OF && PPC_OF && HAVE_GPIO_LIB
 	help
 	  OpenFirmware GPIO accessors
+
+config OF_I2C
+	def_bool y
+	depends on OF && I2C
+	help
+	  OpenFirmware I2C accessors
diff --git a/drivers/of/Makefile b/drivers/of/Makefile
index 5a61f70b4027..a07b95362c53 100644
--- a/drivers/of/Makefile
+++ b/drivers/of/Makefile
@@ -1,3 +1,4 @@
 obj-y = base.o
 obj-$(CONFIG_OF_DEVICE) += device.o platform.o
 obj-$(CONFIG_OF_GPIO)   += gpio.o
+obj-$(CONFIG_OF_I2C)	+= i2c.o
diff --git a/drivers/of/i2c.c b/drivers/of/i2c.c
new file mode 100644
index 000000000000..631689171159
--- /dev/null
+++ b/drivers/of/i2c.c
@@ -0,0 +1,115 @@
+/*
+ * OF helpers for the I2C API
+ *
+ * Copyright (c) 2008 Jochen Friedrich <jochen@scram.de>
+ *
+ * Based on a previous patch from Jon Smirl <jonsmirl@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/i2c.h>
+#include <linux/of.h>
+
+struct i2c_driver_device {
+	char    *of_device;
+	char    *i2c_type;
+};
+
+static struct i2c_driver_device i2c_devices[] = {
+	{ "dallas,ds1374", "rtc-ds1374" },
+};
+
+static int of_find_i2c_driver(struct device_node *node,
+			      struct i2c_board_info *info)
+{
+	int i, cplen;
+	const char *compatible;
+	const char *p;
+
+	/* 1. search for exception list entry */
+	for (i = 0; i < ARRAY_SIZE(i2c_devices); i++) {
+		if (!of_device_is_compatible(node, i2c_devices[i].of_device))
+			continue;
+		if (strlcpy(info->type, i2c_devices[i].i2c_type,
+			    I2C_NAME_SIZE) >= I2C_NAME_SIZE)
+			return -ENOMEM;
+
+		return 0;
+	}
+
+	compatible = of_get_property(node, "compatible", &cplen);
+	if (!compatible)
+		return -ENODEV;
+
+	/* 2. search for linux,<i2c-type> entry */
+	p = compatible;
+	while (cplen > 0) {
+		if (!strncmp(p, "linux,", 6)) {
+			p += 6;
+			if (strlcpy(info->type, p,
+				    I2C_NAME_SIZE) >= I2C_NAME_SIZE)
+				return -ENOMEM;
+			return 0;
+		}
+
+		i = strlen(p) + 1;
+		p += i;
+		cplen -= i;
+	}
+
+	/* 3. take fist compatible entry and strip manufacturer */
+	p = strchr(compatible, ',');
+	if (!p)
+		return -ENODEV;
+	p++;
+	if (strlcpy(info->type, p, I2C_NAME_SIZE) >= I2C_NAME_SIZE)
+		return -ENOMEM;
+	return 0;
+}
+
+void of_register_i2c_devices(struct i2c_adapter *adap,
+			     struct device_node *adap_node)
+{
+	void *result;
+	struct device_node *node;
+
+	for_each_child_of_node(adap_node, node) {
+		struct i2c_board_info info = {};
+		const u32 *addr;
+		int len;
+
+		addr = of_get_property(node, "reg", &len);
+		if (!addr || len < sizeof(int) || *addr > (1 << 10) - 1) {
+			printk(KERN_ERR
+			       "of-i2c: invalid i2c device entry\n");
+			continue;
+		}
+
+		info.irq = irq_of_parse_and_map(node, 0);
+		if (info.irq == NO_IRQ)
+			info.irq = -1;
+
+		if (of_find_i2c_driver(node, &info) < 0) {
+			irq_dispose_mapping(info.irq);
+			continue;
+		}
+
+		info.addr = *addr;
+
+		request_module(info.type);
+
+		result = i2c_new_device(adap, &info);
+		if (result == NULL) {
+			printk(KERN_ERR
+			       "of-i2c: Failed to load driver for %s\n",
+			       info.type);
+			irq_dispose_mapping(info.irq);
+			continue;
+		}
+	}
+}
+EXPORT_SYMBOL(of_register_i2c_devices);
diff --git a/include/linux/of_i2c.h b/include/linux/of_i2c.h
new file mode 100644
index 000000000000..2e5a96732042
--- /dev/null
+++ b/include/linux/of_i2c.h
@@ -0,0 +1,24 @@
+/*
+ * Generic I2C API implementation for PowerPC.
+ *
+ * Copyright (c) 2008 Jochen Friedrich <jochen@scram.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#ifndef __LINUX_OF_I2C_H
+#define __LINUX_OF_I2C_H
+
+#include <linux/i2c.h>
+
+#ifdef CONFIG_OF_I2C
+
+void of_register_i2c_devices(struct i2c_adapter *adap,
+			     struct device_node *adap_node);
+
+#endif /* CONFIG_OF_I2C */
+
+#endif /* __LINUX_OF_I2C_H */
-- 
cgit v1.2.3


From 9d9326d3bc0ea9a8bbe40bf3e5e66c7b9858caa0 Mon Sep 17 00:00:00 2001
From: Andy Fleming <afleming@freescale.com>
Date: Wed, 9 Apr 2008 19:38:13 -0500
Subject: phy: Change mii_bus id field to a string

Having the id field be an int was making more complex bus topologies
excessively difficult.  For now, just convert it to a string, and
change all instances of "bus->id = val" to
snprintf(id, MII_BUS_ID_LEN, "%x", val).

Signed-off-by: Andy Fleming <afleming@freescale.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
---
 arch/powerpc/platforms/82xx/ep8248e.c     |  2 +-
 arch/powerpc/platforms/pasemi/gpio_mdio.c |  2 +-
 arch/powerpc/sysdev/fsl_soc.c             |  5 +++--
 drivers/net/au1000_eth.c                  |  6 +++---
 drivers/net/bfin_mac.c                    |  2 +-
 drivers/net/cpmac.c                       |  5 ++---
 drivers/net/fec_mpc52xx.c                 |  2 +-
 drivers/net/fec_mpc52xx_phy.c             |  2 +-
 drivers/net/fs_enet/fs_enet-main.c        |  4 ++--
 drivers/net/fs_enet/mii-bitbang.c         |  4 ++--
 drivers/net/fs_enet/mii-fec.c             |  4 ++--
 drivers/net/gianfar_mii.c                 |  2 +-
 drivers/net/macb.c                        |  2 +-
 drivers/net/pasemi_mac.c                  |  2 +-
 drivers/net/phy/fixed.c                   |  2 +-
 drivers/net/sb1250-mac.c                  |  2 +-
 drivers/net/ucc_geth.c                    |  2 +-
 drivers/net/ucc_geth.h                    |  2 +-
 drivers/net/ucc_geth_mii.c                |  2 +-
 include/linux/fsl_devices.h               |  2 +-
 include/linux/phy.h                       | 12 ++++++++----
 21 files changed, 36 insertions(+), 32 deletions(-)

(limited to 'include/linux')

diff --git a/arch/powerpc/platforms/82xx/ep8248e.c b/arch/powerpc/platforms/82xx/ep8248e.c
index ba93d8ae9b0c..d5770fdf7f09 100644
--- a/arch/powerpc/platforms/82xx/ep8248e.c
+++ b/arch/powerpc/platforms/82xx/ep8248e.c
@@ -138,7 +138,7 @@ static int __devinit ep8248e_mdio_probe(struct of_device *ofdev,
 
 	bus->name = "ep8248e-mdio-bitbang";
 	bus->dev = &ofdev->dev;
-	bus->id = res.start;
+	snprintf(bus->id, MII_BUS_ID_SIZE, "%x", res.start);
 
 	return mdiobus_register(bus);
 }
diff --git a/arch/powerpc/platforms/pasemi/gpio_mdio.c b/arch/powerpc/platforms/pasemi/gpio_mdio.c
index b46542990cf8..ab6955412ba4 100644
--- a/arch/powerpc/platforms/pasemi/gpio_mdio.c
+++ b/arch/powerpc/platforms/pasemi/gpio_mdio.c
@@ -241,7 +241,7 @@ static int __devinit gpio_mdio_probe(struct of_device *ofdev,
 	new_bus->reset = &gpio_mdio_reset;
 
 	prop = of_get_property(np, "reg", NULL);
-	new_bus->id = *prop;
+	snprintf(new_bus->id, MII_BUS_ID_SIZE, "%x", *prop);
 	new_bus->priv = priv;
 
 	new_bus->phy_mask = 0;
diff --git a/arch/powerpc/sysdev/fsl_soc.c b/arch/powerpc/sysdev/fsl_soc.c
index 2c5388ce902a..3581416905ea 100644
--- a/arch/powerpc/sysdev/fsl_soc.c
+++ b/arch/powerpc/sysdev/fsl_soc.c
@@ -341,7 +341,7 @@ static int __init gfar_of_init(void)
 				goto unreg;
 			}
 
-			gfar_data.bus_id = 0;
+			snprintf(gfar_data.bus_id, MII_BUS_ID_SIZE, "0");
 			gfar_data.phy_id = fixed_link[0];
 		} else {
 			phy = of_find_node_by_phandle(*ph);
@@ -362,7 +362,8 @@ static int __init gfar_of_init(void)
 			}
 
 			gfar_data.phy_id = *id;
-			gfar_data.bus_id = res.start;
+			snprintf(gfar_data.bus_id, MII_BUS_ID_SIZE, "%x",
+					res.start);
 
 			of_node_put(phy);
 			of_node_put(mdio);
diff --git a/drivers/net/au1000_eth.c b/drivers/net/au1000_eth.c
index 504b7ce2747d..3634b5fd7919 100644
--- a/drivers/net/au1000_eth.c
+++ b/drivers/net/au1000_eth.c
@@ -701,7 +701,7 @@ static struct net_device * au1000_probe(int port_num)
 	aup->mii_bus.write = mdiobus_write;
 	aup->mii_bus.reset = mdiobus_reset;
 	aup->mii_bus.name = "au1000_eth_mii";
-	aup->mii_bus.id = aup->mac_id;
+	snprintf(aup->mii_bus.id, MII_BUS_ID_SIZE, "%x", aup->mac_id);
 	aup->mii_bus.irq = kmalloc(sizeof(int)*PHY_MAX_ADDR, GFP_KERNEL);
 	for(i = 0; i < PHY_MAX_ADDR; ++i)
 		aup->mii_bus.irq[i] = PHY_POLL;
@@ -709,11 +709,11 @@ static struct net_device * au1000_probe(int port_num)
 	/* if known, set corresponding PHY IRQs */
 #if defined(AU1XXX_PHY_STATIC_CONFIG)
 # if defined(AU1XXX_PHY0_IRQ)
-	if (AU1XXX_PHY0_BUSID == aup->mii_bus.id)
+	if (AU1XXX_PHY0_BUSID == aup->mac_id)
 		aup->mii_bus.irq[AU1XXX_PHY0_ADDR] = AU1XXX_PHY0_IRQ;
 # endif
 # if defined(AU1XXX_PHY1_IRQ)
-	if (AU1XXX_PHY1_BUSID == aup->mii_bus.id)
+	if (AU1XXX_PHY1_BUSID == aup->mac_id)
 		aup->mii_bus.irq[AU1XXX_PHY1_ADDR] = AU1XXX_PHY1_IRQ;
 # endif
 #endif
diff --git a/drivers/net/bfin_mac.c b/drivers/net/bfin_mac.c
index 26b2dd5016cd..717dcc1aa1e9 100644
--- a/drivers/net/bfin_mac.c
+++ b/drivers/net/bfin_mac.c
@@ -969,7 +969,7 @@ static int __init bf537mac_probe(struct net_device *dev)
 	lp->mii_bus.write = mdiobus_write;
 	lp->mii_bus.reset = mdiobus_reset;
 	lp->mii_bus.name = "bfin_mac_mdio";
-	lp->mii_bus.id = 0;
+	snprintf(lp->mii_bus.id, MII_BUS_ID_SIZE, "0");
 	lp->mii_bus.irq = kmalloc(sizeof(int)*PHY_MAX_ADDR, GFP_KERNEL);
 	for (i = 0; i < PHY_MAX_ADDR; ++i)
 		lp->mii_bus.irq[i] = PHY_POLL;
diff --git a/drivers/net/cpmac.c b/drivers/net/cpmac.c
index c85194f2cd2d..9da7ff437031 100644
--- a/drivers/net/cpmac.c
+++ b/drivers/net/cpmac.c
@@ -987,7 +987,7 @@ static int external_switch;
 static int __devinit cpmac_probe(struct platform_device *pdev)
 {
 	int rc, phy_id, i;
-	int mdio_bus_id = cpmac_mii.id;
+	char *mdio_bus_id = "0";
 	struct resource *mem;
 	struct cpmac_priv *priv;
 	struct net_device *dev;
@@ -1008,8 +1008,6 @@ static int __devinit cpmac_probe(struct platform_device *pdev)
 		if (external_switch || dumb_switch) {
 			struct fixed_phy_status status = {};
 
-			mdio_bus_id = 0;
-
 			/*
 			 * FIXME: this should be in the platform code!
 			 * Since there is not platform code at all (that is,
@@ -1143,6 +1141,7 @@ int __devinit cpmac_init(void)
 	}
 
 	cpmac_mii.phy_mask = ~(mask | 0x80000000);
+	snprintf(cpmac_mii.id, MII_BUS_ID_SIZE, "0");
 
 	res = mdiobus_register(&cpmac_mii);
 	if (res)
diff --git a/drivers/net/fec_mpc52xx.c b/drivers/net/fec_mpc52xx.c
index 58b71e60204e..43b5f30743c2 100644
--- a/drivers/net/fec_mpc52xx.c
+++ b/drivers/net/fec_mpc52xx.c
@@ -198,7 +198,7 @@ static int mpc52xx_fec_init_phy(struct net_device *dev)
 	struct phy_device *phydev;
 	char phy_id[BUS_ID_SIZE];
 
-	snprintf(phy_id, BUS_ID_SIZE, PHY_ID_FMT,
+	snprintf(phy_id, BUS_ID_SIZE, "%x:%02x",
 			(unsigned int)dev->base_addr, priv->phy_addr);
 
 	priv->link = PHY_DOWN;
diff --git a/drivers/net/fec_mpc52xx_phy.c b/drivers/net/fec_mpc52xx_phy.c
index 6a3ac4ea97e9..956836fc5ec0 100644
--- a/drivers/net/fec_mpc52xx_phy.c
+++ b/drivers/net/fec_mpc52xx_phy.c
@@ -124,7 +124,7 @@ static int mpc52xx_fec_mdio_probe(struct of_device *of, const struct of_device_i
 		goto out_free;
 	}
 
-	bus->id = res.start;
+	snprintf(bus->id, MII_BUS_ID_SIZE, "%x", res.start);
 	bus->priv = priv;
 
 	bus->dev = dev;
diff --git a/drivers/net/fs_enet/fs_enet-main.c b/drivers/net/fs_enet/fs_enet-main.c
index 940e2041ba38..67b4b0728fce 100644
--- a/drivers/net/fs_enet/fs_enet-main.c
+++ b/drivers/net/fs_enet/fs_enet-main.c
@@ -1178,7 +1178,7 @@ static int __devinit find_phy(struct device_node *np,
 
 	data  = of_get_property(np, "fixed-link", NULL);
 	if (data) {
-		snprintf(fpi->bus_id, 16, PHY_ID_FMT, 0, *data);
+		snprintf(fpi->bus_id, 16, "%x:%02x", 0, *data);
 		return 0;
 	}
 
@@ -1202,7 +1202,7 @@ static int __devinit find_phy(struct device_node *np,
 	if (!data || len != 4)
 		goto out_put_mdio;
 
-	snprintf(fpi->bus_id, 16, PHY_ID_FMT, res.start, *data);
+	snprintf(fpi->bus_id, 16, "%x:%02x", res.start, *data);
 
 out_put_mdio:
 	of_node_put(mdionode);
diff --git a/drivers/net/fs_enet/mii-bitbang.c b/drivers/net/fs_enet/mii-bitbang.c
index b8e4a736a130..1620030cd33c 100644
--- a/drivers/net/fs_enet/mii-bitbang.c
+++ b/drivers/net/fs_enet/mii-bitbang.c
@@ -130,7 +130,7 @@ static int __devinit fs_mii_bitbang_init(struct mii_bus *bus,
 	 * we get is an int, and the odds of multiple bitbang mdio buses
 	 * is low enough that it's not worth going too crazy.
 	 */
-	bus->id = res.start;
+	snprintf(bus->id, MII_BUS_ID_SIZE, "%x", res.start);
 
 	data = of_get_property(np, "fsl,mdio-pin", &len);
 	if (!data || len != 4)
@@ -307,7 +307,7 @@ static int __devinit fs_enet_mdio_probe(struct device *dev)
 		return -ENOMEM;
 
 	new_bus->name = "BB MII Bus",
-	new_bus->id = pdev->id;
+	snprintf(new_bus->id, MII_BUS_ID_SIZE, "%x", pdev->id);
 
 	new_bus->phy_mask = ~0x9;
 	pdata = (struct fs_mii_bb_platform_info *)pdev->dev.platform_data;
diff --git a/drivers/net/fs_enet/mii-fec.c b/drivers/net/fs_enet/mii-fec.c
index a89cf15090b8..ba75efc9f5b5 100644
--- a/drivers/net/fs_enet/mii-fec.c
+++ b/drivers/net/fs_enet/mii-fec.c
@@ -196,7 +196,7 @@ static int __devinit fs_enet_mdio_probe(struct of_device *ofdev,
 	if (ret)
 		return ret;
 
-	new_bus->id = res.start;
+	snprintf(new_bus->id, MII_BUS_ID_SIZE, "%x", res.start);
 
 	fec->fecp = ioremap(res.start, res.end - res.start + 1);
 	if (!fec->fecp)
@@ -309,7 +309,7 @@ static int __devinit fs_enet_fec_mdio_probe(struct device *dev)
 	new_bus->read = &fs_enet_fec_mii_read,
 	new_bus->write = &fs_enet_fec_mii_write,
 	new_bus->reset = &fs_enet_fec_mii_reset,
-	new_bus->id = pdev->id;
+	snprintf(new_bus->id, MII_BUS_ID_SIZE, "%x", pdev->id);
 
 	pdata = (struct fs_mii_fec_platform_info *)pdev->dev.platform_data;
 
diff --git a/drivers/net/gianfar_mii.c b/drivers/net/gianfar_mii.c
index 24327629bf03..b8898927236a 100644
--- a/drivers/net/gianfar_mii.c
+++ b/drivers/net/gianfar_mii.c
@@ -173,7 +173,7 @@ int gfar_mdio_probe(struct device *dev)
 	new_bus->read = &gfar_mdio_read,
 	new_bus->write = &gfar_mdio_write,
 	new_bus->reset = &gfar_mdio_reset,
-	new_bus->id = pdev->id;
+	snprintf(new_bus->id, MII_BUS_ID_SIZE, "%x", pdev->id);
 
 	pdata = (struct gianfar_mdio_data *)pdev->dev.platform_data;
 
diff --git a/drivers/net/macb.c b/drivers/net/macb.c
index 489c7c3b90d9..d513bb8a4902 100644
--- a/drivers/net/macb.c
+++ b/drivers/net/macb.c
@@ -246,7 +246,7 @@ static int macb_mii_init(struct macb *bp)
 	bp->mii_bus.read = &macb_mdio_read;
 	bp->mii_bus.write = &macb_mdio_write;
 	bp->mii_bus.reset = &macb_mdio_reset;
-	bp->mii_bus.id = bp->pdev->id;
+	snprintf(bp->mii_bus.id, MII_BUS_ID_SIZE, "%x", bp->pdev->id);
 	bp->mii_bus.priv = bp;
 	bp->mii_bus.dev = &bp->dev->dev;
 	pdata = bp->pdev->dev.platform_data;
diff --git a/drivers/net/pasemi_mac.c b/drivers/net/pasemi_mac.c
index 2e39e0285d8f..bcd7f9814ed8 100644
--- a/drivers/net/pasemi_mac.c
+++ b/drivers/net/pasemi_mac.c
@@ -1012,7 +1012,7 @@ static int pasemi_mac_phy_init(struct net_device *dev)
 		goto err;
 
 	phy_id = *prop;
-	snprintf(mac->phy_id, BUS_ID_SIZE, PHY_ID_FMT, (int)r.start, phy_id);
+	snprintf(mac->phy_id, BUS_ID_SIZE, "%x:%02x", (int)r.start, phy_id);
 
 	of_node_put(phy_dn);
 
diff --git a/drivers/net/phy/fixed.c b/drivers/net/phy/fixed.c
index ca9b040f9ad9..4e07956a483b 100644
--- a/drivers/net/phy/fixed.c
+++ b/drivers/net/phy/fixed.c
@@ -213,7 +213,7 @@ static int __init fixed_mdio_bus_init(void)
 		goto err_pdev;
 	}
 
-	fmb->mii_bus.id = 0;
+	snprintf(fmb->mii_bus.id, MII_BUS_ID_SIZE, "0");
 	fmb->mii_bus.name = "Fixed MDIO Bus";
 	fmb->mii_bus.dev = &pdev->dev;
 	fmb->mii_bus.read = &fixed_mdio_read;
diff --git a/drivers/net/sb1250-mac.c b/drivers/net/sb1250-mac.c
index 7b53d658e337..888b7dec9866 100644
--- a/drivers/net/sb1250-mac.c
+++ b/drivers/net/sb1250-mac.c
@@ -2374,7 +2374,7 @@ static int sbmac_init(struct platform_device *pldev, long long base)
 	       dev->name, base, print_mac(mac, eaddr));
 
 	sc->mii_bus.name = sbmac_mdio_string;
-	sc->mii_bus.id = idx;
+	snprintf(sc->mii_bus.id, MII_BUS_ID_SIZE, "%x", idx);
 	sc->mii_bus.priv = sc;
 	sc->mii_bus.read = sbmac_mii_read;
 	sc->mii_bus.write = sbmac_mii_write;
diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 0ee4c168e4c0..29a4d650e8a8 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3954,7 +3954,7 @@ static int ucc_geth_probe(struct of_device* ofdev, const struct of_device_id *ma
 		if (err)
 			return -1;
 
-		ug_info->mdio_bus = res.start;
+		snprintf(ug_info->mdio_bus, MII_BUS_ID_SIZE, "%x", res.start);
 	}
 
 	/* get the phy interface type, or default to MII */
diff --git a/drivers/net/ucc_geth.h b/drivers/net/ucc_geth.h
index 4fb95b3af948..9f8b7580a3a4 100644
--- a/drivers/net/ucc_geth.h
+++ b/drivers/net/ucc_geth.h
@@ -1156,7 +1156,7 @@ struct ucc_geth_info {
 	u16 pausePeriod;
 	u16 extensionField;
 	u8 phy_address;
-	u32 mdio_bus;
+	char mdio_bus[MII_BUS_ID_SIZE];
 	u8 weightfactor[NUM_TX_QUEUES];
 	u8 interruptcoalescingmaxvalue[NUM_RX_QUEUES];
 	u8 l2qt[UCC_GETH_VLAN_PRIORITY_MAX];
diff --git a/drivers/net/ucc_geth_mii.c b/drivers/net/ucc_geth_mii.c
index c69e654d539f..e4d3f330bac3 100644
--- a/drivers/net/ucc_geth_mii.c
+++ b/drivers/net/ucc_geth_mii.c
@@ -157,7 +157,7 @@ static int uec_mdio_probe(struct of_device *ofdev, const struct of_device_id *ma
 	if (err)
 		goto reg_map_fail;
 
-	new_bus->id = res.start;
+	snprintf(new_bus->id, MII_BUS_ID_SIZE, "%x", res.start);
 
 	new_bus->irq = kmalloc(32 * sizeof(int), GFP_KERNEL);
 
diff --git a/include/linux/fsl_devices.h b/include/linux/fsl_devices.h
index 1831b196c70a..2cad5c67397e 100644
--- a/include/linux/fsl_devices.h
+++ b/include/linux/fsl_devices.h
@@ -50,7 +50,7 @@ struct gianfar_platform_data {
 	u32	device_flags;
 	/* board specific information */
 	u32	board_flags;
-	u32	bus_id;
+	char	bus_id[MII_BUS_ID_SIZE];
 	u32	phy_id;
 	u8	mac_addr[6];
 	phy_interface_t interface;
diff --git a/include/linux/phy.h b/include/linux/phy.h
index 5e43ae751412..6509f377bb10 100644
--- a/include/linux/phy.h
+++ b/include/linux/phy.h
@@ -63,8 +63,6 @@ typedef enum {
 	PHY_INTERFACE_MODE_RTBI
 } phy_interface_t;
 
-#define MII_BUS_MAX 4
-
 
 #define PHY_INIT_TIMEOUT	100000
 #define PHY_STATE_TIME		1
@@ -74,13 +72,19 @@ typedef enum {
 #define PHY_MAX_ADDR	32
 
 /* Used when trying to connect to a specific phy (mii bus id:phy device id) */
-#define PHY_ID_FMT "%x:%02x"
+#define PHY_ID_FMT "%s:%02x"
+
+/*
+ * Need to be a little smaller than phydev->dev.bus_id to leave room
+ * for the ":%02x"
+ */
+#define MII_BUS_ID_SIZE	(BUS_ID_SIZE - 3)
 
 /* The Bus class for PHYs.  Devices which provide access to
  * PHYs should register using this structure */
 struct mii_bus {
 	const char *name;
-	int id;
+	char id[MII_BUS_ID_SIZE];
 	void *priv;
 	int (*read)(struct mii_bus *bus, int phy_id, int regnum);
 	int (*write)(struct mii_bus *bus, int phy_id, int regnum, u16 val);
-- 
cgit v1.2.3


From c5e38a949bfa11d10f73927fbf4fe66b73bc3001 Mon Sep 17 00:00:00 2001
From: Andy Fleming <afleming@freescale.com>
Date: Wed, 9 Apr 2008 19:38:27 -0500
Subject: phy: Clean up header style

Multi-line comments weren't all CodingStyle compliant

Signed-off-by: Andy Fleming <afleming@freescale.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
---
 include/linux/phy.h | 48 ++++++++++++++++++++++++++++++++----------------
 1 file changed, 32 insertions(+), 16 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/phy.h b/include/linux/phy.h
index 6509f377bb10..2d838448415c 100644
--- a/include/linux/phy.h
+++ b/include/linux/phy.h
@@ -39,7 +39,8 @@
 				 SUPPORTED_1000baseT_Half | \
 				 SUPPORTED_1000baseT_Full)
 
-/* Set phydev->irq to PHY_POLL if interrupts are not supported,
+/*
+ * Set phydev->irq to PHY_POLL if interrupts are not supported,
  * or not desired for this PHY.  Set to PHY_IGNORE_INTERRUPT if
  * the attached driver handles the interrupt
  */
@@ -80,8 +81,10 @@ typedef enum {
  */
 #define MII_BUS_ID_SIZE	(BUS_ID_SIZE - 3)
 
-/* The Bus class for PHYs.  Devices which provide access to
- * PHYs should register using this structure */
+/*
+ * The Bus class for PHYs.  Devices which provide access to
+ * PHYs should register using this structure
+ */
 struct mii_bus {
 	const char *name;
 	char id[MII_BUS_ID_SIZE];
@@ -90,8 +93,10 @@ struct mii_bus {
 	int (*write)(struct mii_bus *bus, int phy_id, int regnum, u16 val);
 	int (*reset)(struct mii_bus *bus);
 
-	/* A lock to ensure that only one thing can read/write
-	 * the MDIO bus at a time */
+	/*
+	 * A lock to ensure that only one thing can read/write
+	 * the MDIO bus at a time
+	 */
 	struct mutex mdio_lock;
 
 	struct device *dev;
@@ -102,8 +107,10 @@ struct mii_bus {
 	/* Phy addresses to be ignored when probing */
 	u32 phy_mask;
 
-	/* Pointer to an array of interrupts, each PHY's
-	 * interrupt at the index matching its address */
+	/*
+	 * Pointer to an array of interrupts, each PHY's
+	 * interrupt at the index matching its address
+	 */
 	int *irq;
 };
 
@@ -255,7 +262,8 @@ struct phy_device {
 	/* Bus address of the PHY (0-32) */
 	int addr;
 
-	/* forced speed & duplex (no autoneg)
+	/*
+	 * forced speed & duplex (no autoneg)
 	 * partner speed & duplex & pause (autoneg)
 	 */
 	int speed;
@@ -278,8 +286,10 @@ struct phy_device {
 
 	int link_timeout;
 
-	/* Interrupt number for this PHY
-	 * -1 means no interrupt */
+	/*
+	 * Interrupt number for this PHY
+	 * -1 means no interrupt
+	 */
 	int irq;
 
 	/* private data pointer */
@@ -329,22 +339,28 @@ struct phy_driver {
 	u32 features;
 	u32 flags;
 
-	/* Called to initialize the PHY,
-	 * including after a reset */
+	/*
+	 * Called to initialize the PHY,
+	 * including after a reset
+	 */
 	int (*config_init)(struct phy_device *phydev);
 
-	/* Called during discovery.  Used to set
-	 * up device-specific structures, if any */
+	/*
+	 * Called during discovery.  Used to set
+	 * up device-specific structures, if any
+	 */
 	int (*probe)(struct phy_device *phydev);
 
 	/* PHY Power Management */
 	int (*suspend)(struct phy_device *phydev);
 	int (*resume)(struct phy_device *phydev);
 
-	/* Configures the advertisement and resets
+	/*
+	 * Configures the advertisement and resets
 	 * autonegotiation if phydev->autoneg is on,
 	 * forces the speed to the current settings in phydev
-	 * if phydev->autoneg is off */
+	 * if phydev->autoneg is off
+	 */
 	int (*config_aneg)(struct phy_device *phydev);
 
 	/* Determines the negotiated speed and duplex */
-- 
cgit v1.2.3


From 37608eea86a358ac6a18df0af55d4f77d08a1f30 Mon Sep 17 00:00:00 2001
From: Roland Dreier <rolandd@cisco.com>
Date: Wed, 16 Apr 2008 21:01:08 -0700
Subject: mlx4_core: Fix confusion between mlx4_event and mlx4_dev_event enums

The struct mlx4_interface.event() method was supposed to get an enum
mlx4_dev_event, but the driver code was actually passing in the
hardware enum mlx4_event values.  Fix up the callers of
mlx4_dispatch_event() so that they pass in the right type of value,
and fix up the event method in mlx4_ib so that it can handle the enum
mlx4_dev_event values.

This eliminates the need for the subtype parameter to the event
method, so remove it.

This also fixes the sparse warning

    drivers/net/mlx4/intf.c:127:48: warning: mixing different enum types
    drivers/net/mlx4/intf.c:127:48:     int enum mlx4_event  versus
    drivers/net/mlx4/intf.c:127:48:     int enum mlx4_dev_event

Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/hw/mlx4/main.c | 14 ++++++++------
 drivers/net/mlx4/catas.c          |  2 +-
 drivers/net/mlx4/eq.c             |  5 ++++-
 drivers/net/mlx4/intf.c           |  8 ++------
 drivers/net/mlx4/mlx4.h           |  4 ++--
 include/linux/mlx4/driver.h       |  3 +--
 6 files changed, 18 insertions(+), 18 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 96a39b5c9254..d5512011999c 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -675,18 +675,20 @@ static void mlx4_ib_remove(struct mlx4_dev *dev, void *ibdev_ptr)
 }
 
 static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
-			  enum mlx4_dev_event event, int subtype,
-			  int port)
+			  enum mlx4_dev_event event, int port)
 {
 	struct ib_event ibev;
 
 	switch (event) {
-	case MLX4_EVENT_TYPE_PORT_CHANGE:
-		ibev.event = subtype == MLX4_PORT_CHANGE_SUBTYPE_ACTIVE ?
-			IB_EVENT_PORT_ACTIVE : IB_EVENT_PORT_ERR;
+	case MLX4_DEV_EVENT_PORT_UP:
+		ibev.event = IB_EVENT_PORT_ACTIVE;
 		break;
 
-	case MLX4_EVENT_TYPE_LOCAL_CATAS_ERROR:
+	case MLX4_DEV_EVENT_PORT_DOWN:
+		ibev.event = IB_EVENT_PORT_ERR;
+		break;
+
+	case MLX4_DEV_EVENT_CATASTROPHIC_ERROR:
 		ibev.event = IB_EVENT_DEVICE_FATAL;
 		break;
 
diff --git a/drivers/net/mlx4/catas.c b/drivers/net/mlx4/catas.c
index 6b32ec94b3a8..aa9528779044 100644
--- a/drivers/net/mlx4/catas.c
+++ b/drivers/net/mlx4/catas.c
@@ -69,7 +69,7 @@ static void poll_catas(unsigned long dev_ptr)
 	if (readl(priv->catas_err.map)) {
 		dump_err_buf(dev);
 
-		mlx4_dispatch_event(dev, MLX4_EVENT_TYPE_LOCAL_CATAS_ERROR, 0, 0);
+		mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0);
 
 		if (internal_err_reset) {
 			spin_lock(&catas_lock);
diff --git a/drivers/net/mlx4/eq.c b/drivers/net/mlx4/eq.c
index 9c36c2034030..e141a1513f07 100644
--- a/drivers/net/mlx4/eq.c
+++ b/drivers/net/mlx4/eq.c
@@ -202,7 +202,10 @@ static int mlx4_eq_int(struct mlx4_dev *dev, struct mlx4_eq *eq)
 			break;
 
 		case MLX4_EVENT_TYPE_PORT_CHANGE:
-			mlx4_dispatch_event(dev, eqe->type, eqe->subtype,
+			mlx4_dispatch_event(dev,
+					    eqe->subtype == MLX4_PORT_CHANGE_SUBTYPE_ACTIVE ?
+					    MLX4_DEV_EVENT_PORT_UP :
+					    MLX4_DEV_EVENT_PORT_DOWN,
 					    be32_to_cpu(eqe->event.port_change.port) >> 28);
 			break;
 
diff --git a/drivers/net/mlx4/intf.c b/drivers/net/mlx4/intf.c
index be5d9e90ccf2..4a6c4d526f1b 100644
--- a/drivers/net/mlx4/intf.c
+++ b/drivers/net/mlx4/intf.c
@@ -30,8 +30,6 @@
  * SOFTWARE.
  */
 
-#include <linux/mlx4/driver.h>
-
 #include "mlx4.h"
 
 struct mlx4_device_context {
@@ -113,8 +111,7 @@ void mlx4_unregister_interface(struct mlx4_interface *intf)
 }
 EXPORT_SYMBOL_GPL(mlx4_unregister_interface);
 
-void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_event type,
-			 int subtype, int port)
+void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_dev_event type, int port)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	struct mlx4_device_context *dev_ctx;
@@ -124,8 +121,7 @@ void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_event type,
 
 	list_for_each_entry(dev_ctx, &priv->ctx_list, list)
 		if (dev_ctx->intf->event)
-			dev_ctx->intf->event(dev, dev_ctx->context, type,
-					     subtype, port);
+			dev_ctx->intf->event(dev, dev_ctx->context, type, port);
 
 	spin_unlock_irqrestore(&priv->ctx_lock, flags);
 }
diff --git a/drivers/net/mlx4/mlx4.h b/drivers/net/mlx4/mlx4.h
index 53a1cdddfc13..73336810e652 100644
--- a/drivers/net/mlx4/mlx4.h
+++ b/drivers/net/mlx4/mlx4.h
@@ -42,6 +42,7 @@
 #include <linux/timer.h>
 
 #include <linux/mlx4/device.h>
+#include <linux/mlx4/driver.h>
 #include <linux/mlx4/doorbell.h>
 
 #define DRV_NAME	"mlx4_core"
@@ -313,8 +314,7 @@ void mlx4_catas_cleanup(void);
 int mlx4_restart_one(struct pci_dev *pdev);
 int mlx4_register_device(struct mlx4_dev *dev);
 void mlx4_unregister_device(struct mlx4_dev *dev);
-void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_event type,
-			 int subtype, int port);
+void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_dev_event type, int port);
 
 struct mlx4_dev_cap;
 struct mlx4_init_hca_param;
diff --git a/include/linux/mlx4/driver.h b/include/linux/mlx4/driver.h
index 1b835ca49df1..53c5fdb6eac4 100644
--- a/include/linux/mlx4/driver.h
+++ b/include/linux/mlx4/driver.h
@@ -48,8 +48,7 @@ struct mlx4_interface {
 	void *			(*add)	 (struct mlx4_dev *dev);
 	void			(*remove)(struct mlx4_dev *dev, void *context);
 	void			(*event) (struct mlx4_dev *dev, void *context,
-					  enum mlx4_dev_event event, int subtype,
-					  int port);
+					  enum mlx4_dev_event event, int port);
 	struct list_head	list;
 };
 
-- 
cgit v1.2.3


From 8ff095ec4bce7be943beff3b330562e2f0e42167 Mon Sep 17 00:00:00 2001
From: Eli Cohen <eli@dev.mellanox.co.il>
Date: Wed, 16 Apr 2008 21:01:10 -0700
Subject: IB/mlx4: Add IPoIB checksum offload support

ConnectX devices support checksum generation and verification of TCP
and UDP packets for UD IPoIB messages.  This patch checks if the HCA
supports this and sets the IB_DEVICE_UD_IP_CSUM capability flag if it
does.  It implements support for handling the IB_SEND_IP_CSUM send
flag and setting the csum_ok field in receive work completions.

Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Ali Ayub <ali@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/hw/mlx4/cq.c   | 16 ++++++++++++++++
 drivers/infiniband/hw/mlx4/main.c |  2 ++
 drivers/infiniband/hw/mlx4/qp.c   |  3 +++
 drivers/net/mlx4/fw.c             |  4 ++++
 include/linux/mlx4/cq.h           | 14 ++++++++++++--
 include/linux/mlx4/qp.h           | 10 ++++++----
 6 files changed, 43 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index 7360bbafbe84..d2e32b03e2f7 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -297,6 +297,20 @@ static void mlx4_ib_handle_error_cqe(struct mlx4_err_cqe *cqe,
 	wc->vendor_err = cqe->vendor_err_syndrome;
 }
 
+static int mlx4_ib_ipoib_csum_ok(__be32 status, __be16 checksum)
+{
+	return ((status & cpu_to_be32(MLX4_CQE_IPOIB_STATUS_IPV4	|
+				      MLX4_CQE_IPOIB_STATUS_IPV4F	|
+				      MLX4_CQE_IPOIB_STATUS_IPV4OPT	|
+				      MLX4_CQE_IPOIB_STATUS_IPV6	|
+				      MLX4_CQE_IPOIB_STATUS_IPOK)) ==
+		cpu_to_be32(MLX4_CQE_IPOIB_STATUS_IPV4	|
+			    MLX4_CQE_IPOIB_STATUS_IPOK))		&&
+		(status & cpu_to_be32(MLX4_CQE_IPOIB_STATUS_UDP	|
+				      MLX4_CQE_IPOIB_STATUS_TCP))	&&
+		checksum == cpu_to_be16(0xffff);
+}
+
 static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq,
 			    struct mlx4_ib_qp **cur_qp,
 			    struct ib_wc *wc)
@@ -434,6 +448,8 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq,
 		wc->dlid_path_bits = (g_mlpath_rqpn >> 24) & 0x7f;
 		wc->wc_flags	  |= g_mlpath_rqpn & 0x80000000 ? IB_WC_GRH : 0;
 		wc->pkey_index     = be32_to_cpu(cqe->immed_rss_invalid) & 0x7f;
+		wc->csum_ok	   = mlx4_ib_ipoib_csum_ok(cqe->ipoib_status,
+							   cqe->checksum);
 	}
 
 	return 0;
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index d5512011999c..6ea4746c2e9b 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -99,6 +99,8 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 		props->device_cap_flags |= IB_DEVICE_AUTO_PATH_MIG;
 	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_UD_AV_PORT)
 		props->device_cap_flags |= IB_DEVICE_UD_AV_PORT_ENFORCE;
+	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM)
+		props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM;
 
 	props->vendor_id	   = be32_to_cpup((__be32 *) (out_mad->data + 36)) &
 		0xffffff;
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index ac965ab28845..31b2b5b230bd 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -1436,6 +1436,9 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			 cpu_to_be32(MLX4_WQE_CTRL_CQ_UPDATE) : 0) |
 			(wr->send_flags & IB_SEND_SOLICITED ?
 			 cpu_to_be32(MLX4_WQE_CTRL_SOLICITED) : 0) |
+			((wr->send_flags & IB_SEND_IP_CSUM) ?
+			 cpu_to_be32(MLX4_WQE_CTRL_IP_CSUM |
+				     MLX4_WQE_CTRL_TCP_UDP_CSUM) : 0) |
 			qp->sq_signal_bits;
 
 		if (wr->opcode == IB_WR_SEND_WITH_IMM ||
diff --git a/drivers/net/mlx4/fw.c b/drivers/net/mlx4/fw.c
index 61dc4951d6b0..f494c3e8bce3 100644
--- a/drivers/net/mlx4/fw.c
+++ b/drivers/net/mlx4/fw.c
@@ -696,6 +696,10 @@ int mlx4_INIT_HCA(struct mlx4_dev *dev, struct mlx4_init_hca_param *param)
 	/* Check port for UD address vector: */
 	*(inbox + INIT_HCA_FLAGS_OFFSET / 4) |= cpu_to_be32(1);
 
+	/* Enable IPoIB checksumming if we can: */
+	if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM)
+		*(inbox + INIT_HCA_FLAGS_OFFSET / 4) |= cpu_to_be32(1 << 3);
+
 	/* QPC/EEC/CQC/EQC/RDMARC attributes */
 
 	MLX4_PUT(inbox, param->qpc_base,      INIT_HCA_QPC_BASE_OFFSET);
diff --git a/include/linux/mlx4/cq.h b/include/linux/mlx4/cq.h
index 0181e0a57cbf..1243ebace561 100644
--- a/include/linux/mlx4/cq.h
+++ b/include/linux/mlx4/cq.h
@@ -45,11 +45,11 @@ struct mlx4_cqe {
 	u8			sl;
 	u8			reserved1;
 	__be16			rlid;
-	u32			reserved2;
+	__be32			ipoib_status;
 	__be32			byte_cnt;
 	__be16			wqe_index;
 	__be16			checksum;
-	u8			reserved3[3];
+	u8			reserved2[3];
 	u8			owner_sr_opcode;
 };
 
@@ -85,6 +85,16 @@ enum {
 	MLX4_CQE_SYNDROME_REMOTE_ABORTED_ERR		= 0x22,
 };
 
+enum {
+	MLX4_CQE_IPOIB_STATUS_IPV4			= 1 << 22,
+	MLX4_CQE_IPOIB_STATUS_IPV4F			= 1 << 23,
+	MLX4_CQE_IPOIB_STATUS_IPV6			= 1 << 24,
+	MLX4_CQE_IPOIB_STATUS_IPV4OPT			= 1 << 25,
+	MLX4_CQE_IPOIB_STATUS_TCP			= 1 << 26,
+	MLX4_CQE_IPOIB_STATUS_UDP			= 1 << 27,
+	MLX4_CQE_IPOIB_STATUS_IPOK			= 1 << 28,
+};
+
 static inline void mlx4_cq_arm(struct mlx4_cq *cq, u32 cmd,
 			       void __iomem *uar_page,
 			       spinlock_t *doorbell_lock)
diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h
index 09a2230923f2..31f9eb3ccbb3 100644
--- a/include/linux/mlx4/qp.h
+++ b/include/linux/mlx4/qp.h
@@ -158,10 +158,12 @@ struct mlx4_qp_context {
 #define MLX4_FW_VER_WQE_CTRL_NEC mlx4_fw_ver(2, 2, 232)
 
 enum {
-	MLX4_WQE_CTRL_NEC	= 1 << 29,
-	MLX4_WQE_CTRL_FENCE	= 1 << 6,
-	MLX4_WQE_CTRL_CQ_UPDATE	= 3 << 2,
-	MLX4_WQE_CTRL_SOLICITED	= 1 << 1,
+	MLX4_WQE_CTRL_NEC		= 1 << 29,
+	MLX4_WQE_CTRL_FENCE		= 1 << 6,
+	MLX4_WQE_CTRL_CQ_UPDATE		= 3 << 2,
+	MLX4_WQE_CTRL_SOLICITED		= 1 << 1,
+	MLX4_WQE_CTRL_IP_CSUM		= 1 << 4,
+	MLX4_WQE_CTRL_TCP_UDP_CSUM	= 1 << 5,
 };
 
 struct mlx4_wqe_ctrl_seg {
-- 
cgit v1.2.3


From b832be1e4007f4a54954ec68bd865ff05d6babca Mon Sep 17 00:00:00 2001
From: Eli Cohen <eli@dev.mellanox.co.il>
Date: Wed, 16 Apr 2008 21:09:27 -0700
Subject: IB/mlx4: Add IPoIB LSO support

Add TSO support to the mlx4_ib driver.

Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/hw/mlx4/cq.c      |  3 ++
 drivers/infiniband/hw/mlx4/main.c    |  2 +
 drivers/infiniband/hw/mlx4/mlx4_ib.h |  5 +++
 drivers/infiniband/hw/mlx4/qp.c      | 72 +++++++++++++++++++++++++++++++-----
 drivers/net/mlx4/fw.c                |  9 +++++
 drivers/net/mlx4/fw.h                |  1 +
 drivers/net/mlx4/main.c              |  1 +
 include/linux/mlx4/device.h          |  1 +
 include/linux/mlx4/qp.h              |  5 +++
 9 files changed, 90 insertions(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index d2e32b03e2f7..7d70af7952b0 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -420,6 +420,9 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq,
 		case MLX4_OPCODE_BIND_MW:
 			wc->opcode    = IB_WC_BIND_MW;
 			break;
+		case MLX4_OPCODE_LSO:
+			wc->opcode    = IB_WC_LSO;
+			break;
 		}
 	} else {
 		wc->byte_len = be32_to_cpu(cqe->byte_cnt);
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 6ea4746c2e9b..e9330a0d6c03 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -101,6 +101,8 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 		props->device_cap_flags |= IB_DEVICE_UD_AV_PORT_ENFORCE;
 	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM)
 		props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM;
+	if (dev->dev->caps.max_gso_sz)
+		props->device_cap_flags |= IB_DEVICE_UD_TSO;
 
 	props->vendor_id	   = be32_to_cpup((__be32 *) (out_mad->data + 36)) &
 		0xffffff;
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 3726e451a327..3f8bd0a37b96 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -110,6 +110,10 @@ struct mlx4_ib_wq {
 	unsigned		tail;
 };
 
+enum mlx4_ib_qp_flags {
+	MLX4_IB_QP_LSO		= 1 << 0
+};
+
 struct mlx4_ib_qp {
 	struct ib_qp		ibqp;
 	struct mlx4_qp		mqp;
@@ -129,6 +133,7 @@ struct mlx4_ib_qp {
 	struct mlx4_mtt		mtt;
 	int			buf_size;
 	struct mutex		mutex;
+	u32			flags;
 	u8			port;
 	u8			alt_port;
 	u8			atomic_rd_en;
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 320c25fa74b1..2ba243084089 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -71,6 +71,7 @@ enum {
 
 static const __be32 mlx4_ib_opcode[] = {
 	[IB_WR_SEND]			= __constant_cpu_to_be32(MLX4_OPCODE_SEND),
+	[IB_WR_LSO]			= __constant_cpu_to_be32(MLX4_OPCODE_LSO),
 	[IB_WR_SEND_WITH_IMM]		= __constant_cpu_to_be32(MLX4_OPCODE_SEND_IMM),
 	[IB_WR_RDMA_WRITE]		= __constant_cpu_to_be32(MLX4_OPCODE_RDMA_WRITE),
 	[IB_WR_RDMA_WRITE_WITH_IMM]	= __constant_cpu_to_be32(MLX4_OPCODE_RDMA_WRITE_IMM),
@@ -242,7 +243,7 @@ static void mlx4_ib_qp_event(struct mlx4_qp *qp, enum mlx4_event type)
 	}
 }
 
-static int send_wqe_overhead(enum ib_qp_type type)
+static int send_wqe_overhead(enum ib_qp_type type, u32 flags)
 {
 	/*
 	 * UD WQEs must have a datagram segment.
@@ -253,7 +254,8 @@ static int send_wqe_overhead(enum ib_qp_type type)
 	switch (type) {
 	case IB_QPT_UD:
 		return sizeof (struct mlx4_wqe_ctrl_seg) +
-			sizeof (struct mlx4_wqe_datagram_seg);
+			sizeof (struct mlx4_wqe_datagram_seg) +
+			((flags & MLX4_IB_QP_LSO) ? 64 : 0);
 	case IB_QPT_UC:
 		return sizeof (struct mlx4_wqe_ctrl_seg) +
 			sizeof (struct mlx4_wqe_raddr_seg);
@@ -315,7 +317,7 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap,
 	/* Sanity check SQ size before proceeding */
 	if (cap->max_send_wr	 > dev->dev->caps.max_wqes  ||
 	    cap->max_send_sge	 > dev->dev->caps.max_sq_sg ||
-	    cap->max_inline_data + send_wqe_overhead(type) +
+	    cap->max_inline_data + send_wqe_overhead(type, qp->flags) +
 	    sizeof (struct mlx4_wqe_inline_seg) > dev->dev->caps.max_sq_desc_sz)
 		return -EINVAL;
 
@@ -329,7 +331,7 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap,
 
 	s = max(cap->max_send_sge * sizeof (struct mlx4_wqe_data_seg),
 		cap->max_inline_data + sizeof (struct mlx4_wqe_inline_seg)) +
-		send_wqe_overhead(type);
+		send_wqe_overhead(type, qp->flags);
 
 	/*
 	 * Hermon supports shrinking WQEs, such that a single work
@@ -394,7 +396,8 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap,
 	}
 
 	qp->sq.max_gs = ((qp->sq_max_wqes_per_wr << qp->sq.wqe_shift) -
-			 send_wqe_overhead(type)) / sizeof (struct mlx4_wqe_data_seg);
+			 send_wqe_overhead(type, qp->flags)) /
+		sizeof (struct mlx4_wqe_data_seg);
 
 	qp->buf_size = (qp->rq.wqe_cnt << qp->rq.wqe_shift) +
 		(qp->sq.wqe_cnt << qp->sq.wqe_shift);
@@ -503,6 +506,9 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 	} else {
 		qp->sq_no_prefetch = 0;
 
+		if (init_attr->create_flags & IB_QP_CREATE_IPOIB_UD_LSO)
+			qp->flags |= MLX4_IB_QP_LSO;
+
 		err = set_kernel_sq_size(dev, &init_attr->cap, init_attr->qp_type, qp);
 		if (err)
 			goto err;
@@ -673,7 +679,11 @@ struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
 	struct mlx4_ib_qp *qp;
 	int err;
 
-	if (init_attr->create_flags)
+	/* We only support LSO, and only for kernel UD QPs. */
+	if (init_attr->create_flags & ~IB_QP_CREATE_IPOIB_UD_LSO)
+		return ERR_PTR(-EINVAL);
+	if (init_attr->create_flags & IB_QP_CREATE_IPOIB_UD_LSO &&
+	    (pd->uobject || init_attr->qp_type != IB_QPT_UD))
 		return ERR_PTR(-EINVAL);
 
 	switch (init_attr->qp_type) {
@@ -879,10 +889,15 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 		}
 	}
 
-	if (ibqp->qp_type == IB_QPT_GSI || ibqp->qp_type == IB_QPT_SMI ||
-	    ibqp->qp_type == IB_QPT_UD)
+	if (ibqp->qp_type == IB_QPT_GSI || ibqp->qp_type == IB_QPT_SMI)
 		context->mtu_msgmax = (IB_MTU_4096 << 5) | 11;
-	else if (attr_mask & IB_QP_PATH_MTU) {
+	else if (ibqp->qp_type == IB_QPT_UD) {
+		if (qp->flags & MLX4_IB_QP_LSO)
+			context->mtu_msgmax = (IB_MTU_4096 << 5) |
+					      ilog2(dev->dev->caps.max_gso_sz);
+		else
+			context->mtu_msgmax = (IB_MTU_4096 << 5) | 11;
+	} else if (attr_mask & IB_QP_PATH_MTU) {
 		if (attr->path_mtu < IB_MTU_256 || attr->path_mtu > IB_MTU_4096) {
 			printk(KERN_ERR "path MTU (%u) is invalid\n",
 			       attr->path_mtu);
@@ -1399,6 +1414,34 @@ static void __set_data_seg(struct mlx4_wqe_data_seg *dseg, struct ib_sge *sg)
 	dseg->addr       = cpu_to_be64(sg->addr);
 }
 
+static int build_lso_seg(struct mlx4_lso_seg *wqe, struct ib_send_wr *wr,
+			 struct mlx4_ib_qp *qp, unsigned *lso_seg_len)
+{
+	unsigned halign = ALIGN(sizeof *wqe + wr->wr.ud.hlen, 16);
+
+	/*
+	 * This is a temporary limitation and will be removed in
+	 * a forthcoming FW release:
+	 */
+	if (unlikely(halign > 64))
+		return -EINVAL;
+
+	if (unlikely(!(qp->flags & MLX4_IB_QP_LSO) &&
+		     wr->num_sge > qp->sq.max_gs - (halign >> 4)))
+		return -EINVAL;
+
+	memcpy(wqe->header, wr->wr.ud.header, wr->wr.ud.hlen);
+
+	/* make sure LSO header is written before overwriting stamping */
+	wmb();
+
+	wqe->mss_hdr_size = cpu_to_be32((wr->wr.ud.mss - wr->wr.ud.hlen) << 16 |
+					wr->wr.ud.hlen);
+
+	*lso_seg_len = halign;
+	return 0;
+}
+
 int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		      struct ib_send_wr **bad_wr)
 {
@@ -1412,6 +1455,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 	unsigned ind;
 	int uninitialized_var(stamp);
 	int uninitialized_var(size);
+	unsigned seglen;
 	int i;
 
 	spin_lock_irqsave(&qp->sq.lock, flags);
@@ -1490,6 +1534,16 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			set_datagram_seg(wqe, wr);
 			wqe  += sizeof (struct mlx4_wqe_datagram_seg);
 			size += sizeof (struct mlx4_wqe_datagram_seg) / 16;
+
+			if (wr->opcode == IB_WR_LSO) {
+				err = build_lso_seg(wqe, wr, qp, &seglen);
+				if (unlikely(err)) {
+					*bad_wr = wr;
+					goto out;
+				}
+				wqe  += seglen;
+				size += seglen / 16;
+			}
 			break;
 
 		case IB_QPT_SMI:
diff --git a/drivers/net/mlx4/fw.c b/drivers/net/mlx4/fw.c
index f494c3e8bce3..d82f2751d2c7 100644
--- a/drivers/net/mlx4/fw.c
+++ b/drivers/net/mlx4/fw.c
@@ -133,6 +133,7 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 #define QUERY_DEV_CAP_MAX_AV_OFFSET		0x27
 #define QUERY_DEV_CAP_MAX_REQ_QP_OFFSET		0x29
 #define QUERY_DEV_CAP_MAX_RES_QP_OFFSET		0x2b
+#define QUERY_DEV_CAP_MAX_GSO_OFFSET		0x2d
 #define QUERY_DEV_CAP_MAX_RDMA_OFFSET		0x2f
 #define QUERY_DEV_CAP_RSZ_SRQ_OFFSET		0x33
 #define QUERY_DEV_CAP_ACK_DELAY_OFFSET		0x35
@@ -215,6 +216,13 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 	dev_cap->max_requester_per_qp = 1 << (field & 0x3f);
 	MLX4_GET(field, outbox, QUERY_DEV_CAP_MAX_RES_QP_OFFSET);
 	dev_cap->max_responder_per_qp = 1 << (field & 0x3f);
+	MLX4_GET(field, outbox, QUERY_DEV_CAP_MAX_GSO_OFFSET);
+	field &= 0x1f;
+	if (!field)
+		dev_cap->max_gso_sz = 0;
+	else
+		dev_cap->max_gso_sz = 1 << field;
+
 	MLX4_GET(field, outbox, QUERY_DEV_CAP_MAX_RDMA_OFFSET);
 	dev_cap->max_rdma_global = 1 << (field & 0x3f);
 	MLX4_GET(field, outbox, QUERY_DEV_CAP_ACK_DELAY_OFFSET);
@@ -377,6 +385,7 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 		 dev_cap->max_sq_desc_sz, dev_cap->max_sq_sg);
 	mlx4_dbg(dev, "Max RQ desc size: %d, max RQ S/G: %d\n",
 		 dev_cap->max_rq_desc_sz, dev_cap->max_rq_sg);
+	mlx4_dbg(dev, "Max GSO size: %d\n", dev_cap->max_gso_sz);
 
 	dump_dev_cap_flags(dev, dev_cap->flags);
 
diff --git a/drivers/net/mlx4/fw.h b/drivers/net/mlx4/fw.h
index e16dec890413..306cb9b0242d 100644
--- a/drivers/net/mlx4/fw.h
+++ b/drivers/net/mlx4/fw.h
@@ -96,6 +96,7 @@ struct mlx4_dev_cap {
 	u8  bmme_flags;
 	u32 reserved_lkey;
 	u64 max_icm_sz;
+	int max_gso_sz;
 };
 
 struct mlx4_adapter {
diff --git a/drivers/net/mlx4/main.c b/drivers/net/mlx4/main.c
index 08bfc130a33e..7cfbe75114d1 100644
--- a/drivers/net/mlx4/main.c
+++ b/drivers/net/mlx4/main.c
@@ -159,6 +159,7 @@ static int mlx4_dev_cap(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 	dev->caps.page_size_cap	     = ~(u32) (dev_cap->min_page_sz - 1);
 	dev->caps.flags		     = dev_cap->flags;
 	dev->caps.stat_rate_support  = dev_cap->stat_rate_support;
+	dev->caps.max_gso_sz	     = dev_cap->max_gso_sz;
 
 	return 0;
 }
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index 6cdf813cd478..ff7df1a2222f 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -186,6 +186,7 @@ struct mlx4_caps {
 	u32			flags;
 	u16			stat_rate_support;
 	u8			port_width_cap[MLX4_MAX_PORTS + 1];
+	int			max_gso_sz;
 };
 
 struct mlx4_buf_list {
diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h
index 31f9eb3ccbb3..a5e43febee4f 100644
--- a/include/linux/mlx4/qp.h
+++ b/include/linux/mlx4/qp.h
@@ -219,6 +219,11 @@ struct mlx4_wqe_datagram_seg {
 	__be32			reservd[2];
 };
 
+struct mlx4_lso_seg {
+	__be32			mss_hdr_size;
+	__be32			header[0];
+};
+
 struct mlx4_wqe_bind_seg {
 	__be32			flags1;
 	__be32			flags2;
-- 
cgit v1.2.3


From 3fdcb97f0b8d8a29117dc36acd0b15965d2a2160 Mon Sep 17 00:00:00 2001
From: Eli Cohen <eli@dev.mellanox.co.il>
Date: Wed, 16 Apr 2008 21:09:33 -0700
Subject: IB/mlx4: Add support for modifying CQ moderation parameters

Signed-off-by: Eli Cohen <eli@mellnaox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/hw/mlx4/cq.c      |  8 +++++++
 drivers/infiniband/hw/mlx4/main.c    |  1 +
 drivers/infiniband/hw/mlx4/mlx4_ib.h |  1 +
 drivers/net/mlx4/cq.c                | 44 ++++++++++++++++++++++++++++++------
 include/linux/mlx4/cmd.h             |  2 +-
 include/linux/mlx4/cq.h              |  3 +++
 6 files changed, 51 insertions(+), 8 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index 7d70af7952b0..e4fb64b118e3 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -85,6 +85,14 @@ static struct mlx4_cqe *next_cqe_sw(struct mlx4_ib_cq *cq)
 	return get_sw_cqe(cq, cq->mcq.cons_index);
 }
 
+int mlx4_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
+{
+	struct mlx4_ib_cq *mcq = to_mcq(cq);
+	struct mlx4_ib_dev *dev = to_mdev(cq->device);
+
+	return mlx4_cq_modify(dev->dev, &mcq->mcq, cq_count, cq_period);
+}
+
 struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector,
 				struct ib_ucontext *context,
 				struct ib_udata *udata)
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index e9330a0d6c03..76dd45c764b4 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -609,6 +609,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	ibdev->ib_dev.post_send		= mlx4_ib_post_send;
 	ibdev->ib_dev.post_recv		= mlx4_ib_post_recv;
 	ibdev->ib_dev.create_cq		= mlx4_ib_create_cq;
+	ibdev->ib_dev.modify_cq		= mlx4_ib_modify_cq;
 	ibdev->ib_dev.destroy_cq	= mlx4_ib_destroy_cq;
 	ibdev->ib_dev.poll_cq		= mlx4_ib_poll_cq;
 	ibdev->ib_dev.req_notify_cq	= mlx4_ib_arm_cq;
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 3f8bd0a37b96..ef8ad96e1432 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -254,6 +254,7 @@ struct ib_mr *mlx4_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 				  struct ib_udata *udata);
 int mlx4_ib_dereg_mr(struct ib_mr *mr);
 
+int mlx4_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period);
 struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector,
 				struct ib_ucontext *context,
 				struct ib_udata *udata);
diff --git a/drivers/net/mlx4/cq.c b/drivers/net/mlx4/cq.c
index d4441fee3d80..8c314341434f 100644
--- a/drivers/net/mlx4/cq.c
+++ b/drivers/net/mlx4/cq.c
@@ -38,6 +38,7 @@
 #include <linux/hardirq.h>
 
 #include <linux/mlx4/cmd.h>
+#include <linux/mlx4/cq.h>
 
 #include "mlx4.h"
 #include "icm.h"
@@ -47,21 +48,19 @@ struct mlx4_cq_context {
 	u16			reserved1[3];
 	__be16			page_offset;
 	__be32			logsize_usrpage;
-	u8			reserved2;
-	u8			cq_period;
-	u8			reserved3;
-	u8			cq_max_count;
-	u8			reserved4[3];
+	__be16			cq_period;
+	__be16			cq_max_count;
+	u8			reserved2[3];
 	u8			comp_eqn;
 	u8			log_page_size;
-	u8			reserved5[2];
+	u8			reserved3[2];
 	u8			mtt_base_addr_h;
 	__be32			mtt_base_addr_l;
 	__be32			last_notified_index;
 	__be32			solicit_producer_index;
 	__be32			consumer_index;
 	__be32			producer_index;
-	u32			reserved6[2];
+	u32			reserved4[2];
 	__be64			db_rec_addr;
 };
 
@@ -121,6 +120,13 @@ static int mlx4_SW2HW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox,
 			MLX4_CMD_TIME_CLASS_A);
 }
 
+static int mlx4_MODIFY_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox,
+			 int cq_num, u32 opmod)
+{
+	return mlx4_cmd(dev, mailbox->dma, cq_num, opmod, MLX4_CMD_MODIFY_CQ,
+			MLX4_CMD_TIME_CLASS_A);
+}
+
 static int mlx4_HW2SW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox,
 			 int cq_num)
 {
@@ -129,6 +135,30 @@ static int mlx4_HW2SW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox,
 			    MLX4_CMD_TIME_CLASS_A);
 }
 
+int mlx4_cq_modify(struct mlx4_dev *dev, struct mlx4_cq *cq,
+		   u16 count, u16 period)
+{
+	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_cq_context *cq_context;
+	int err;
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox))
+		return PTR_ERR(mailbox);
+
+	cq_context = mailbox->buf;
+	memset(cq_context, 0, sizeof *cq_context);
+
+	cq_context->cq_max_count = cpu_to_be16(count);
+	cq_context->cq_period    = cpu_to_be16(period);
+
+	err = mlx4_MODIFY_CQ(dev, mailbox, cq->cqn, 1);
+
+	mlx4_free_cmd_mailbox(dev, mailbox);
+	return err;
+}
+EXPORT_SYMBOL_GPL(mlx4_cq_modify);
+
 int mlx4_cq_alloc(struct mlx4_dev *dev, int nent, struct mlx4_mtt *mtt,
 		  struct mlx4_uar *uar, u64 db_rec, struct mlx4_cq *cq)
 {
diff --git a/include/linux/mlx4/cmd.h b/include/linux/mlx4/cmd.h
index 7d1eaa97de13..77323a72dd3c 100644
--- a/include/linux/mlx4/cmd.h
+++ b/include/linux/mlx4/cmd.h
@@ -81,7 +81,7 @@ enum {
 	MLX4_CMD_SW2HW_CQ	 = 0x16,
 	MLX4_CMD_HW2SW_CQ	 = 0x17,
 	MLX4_CMD_QUERY_CQ	 = 0x18,
-	MLX4_CMD_RESIZE_CQ	 = 0x2c,
+	MLX4_CMD_MODIFY_CQ	 = 0x2c,
 
 	/* SRQ commands */
 	MLX4_CMD_SW2HW_SRQ	 = 0x35,
diff --git a/include/linux/mlx4/cq.h b/include/linux/mlx4/cq.h
index 1243ebace561..f7c3511c594b 100644
--- a/include/linux/mlx4/cq.h
+++ b/include/linux/mlx4/cq.h
@@ -130,4 +130,7 @@ enum {
 	MLX4_CQ_DB_REQ_NOT		= 2 << 24
 };
 
+int mlx4_cq_modify(struct mlx4_dev *dev, struct mlx4_cq *cq,
+		   u16 count, u16 period);
+
 #endif /* MLX4_CQ_H */
-- 
cgit v1.2.3


From bbf8eed1a0f8949f7385146624f736f829992a70 Mon Sep 17 00:00:00 2001
From: Vladimir Sokolovsky <vlad@dev.mellanox.co.il>
Date: Wed, 16 Apr 2008 21:09:33 -0700
Subject: IB/mlx4: Add support for resizing CQs

Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
---
 drivers/infiniband/hw/mlx4/cq.c      | 292 +++++++++++++++++++++++++++++++----
 drivers/infiniband/hw/mlx4/main.c    |   2 +
 drivers/infiniband/hw/mlx4/mlx4_ib.h |   9 ++
 drivers/net/mlx4/cq.c                |  28 ++++
 include/linux/mlx4/cq.h              |   2 +
 5 files changed, 300 insertions(+), 33 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index e4fb64b118e3..3557e7edc9b6 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -93,6 +93,74 @@ int mlx4_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
 	return mlx4_cq_modify(dev->dev, &mcq->mcq, cq_count, cq_period);
 }
 
+static int mlx4_ib_alloc_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf *buf, int nent)
+{
+	int err;
+
+	err = mlx4_buf_alloc(dev->dev, nent * sizeof(struct mlx4_cqe),
+			     PAGE_SIZE * 2, &buf->buf);
+
+	if (err)
+		goto out;
+
+	err = mlx4_mtt_init(dev->dev, buf->buf.npages, buf->buf.page_shift,
+				    &buf->mtt);
+	if (err)
+		goto err_buf;
+
+	err = mlx4_buf_write_mtt(dev->dev, &buf->mtt, &buf->buf);
+	if (err)
+		goto err_mtt;
+
+	return 0;
+
+err_mtt:
+	mlx4_mtt_cleanup(dev->dev, &buf->mtt);
+
+err_buf:
+	mlx4_buf_free(dev->dev, nent * sizeof(struct mlx4_cqe),
+			      &buf->buf);
+
+out:
+	return err;
+}
+
+static void mlx4_ib_free_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf *buf, int cqe)
+{
+	mlx4_buf_free(dev->dev, (cqe + 1) * sizeof(struct mlx4_cqe), &buf->buf);
+}
+
+static int mlx4_ib_get_cq_umem(struct mlx4_ib_dev *dev, struct ib_ucontext *context,
+			       struct mlx4_ib_cq_buf *buf, struct ib_umem **umem,
+			       u64 buf_addr, int cqe)
+{
+	int err;
+
+	*umem = ib_umem_get(context, buf_addr, cqe * sizeof (struct mlx4_cqe),
+			    IB_ACCESS_LOCAL_WRITE);
+	if (IS_ERR(*umem))
+		return PTR_ERR(*umem);
+
+	err = mlx4_mtt_init(dev->dev, ib_umem_page_count(*umem),
+			    ilog2((*umem)->page_size), &buf->mtt);
+	if (err)
+		goto err_buf;
+
+	err = mlx4_ib_umem_write_mtt(dev, &buf->mtt, *umem);
+	if (err)
+		goto err_mtt;
+
+	return 0;
+
+err_mtt:
+	mlx4_mtt_cleanup(dev->dev, &buf->mtt);
+
+err_buf:
+	ib_umem_release(*umem);
+
+	return err;
+}
+
 struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector,
 				struct ib_ucontext *context,
 				struct ib_udata *udata)
@@ -100,7 +168,6 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector
 	struct mlx4_ib_dev *dev = to_mdev(ibdev);
 	struct mlx4_ib_cq *cq;
 	struct mlx4_uar *uar;
-	int buf_size;
 	int err;
 
 	if (entries < 1 || entries > dev->dev->caps.max_cqes)
@@ -112,8 +179,10 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector
 
 	entries      = roundup_pow_of_two(entries + 1);
 	cq->ibcq.cqe = entries - 1;
-	buf_size     = entries * sizeof (struct mlx4_cqe);
+	mutex_init(&cq->resize_mutex);
 	spin_lock_init(&cq->lock);
+	cq->resize_buf = NULL;
+	cq->resize_umem = NULL;
 
 	if (context) {
 		struct mlx4_ib_create_cq ucmd;
@@ -123,21 +192,10 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector
 			goto err_cq;
 		}
 
-		cq->umem = ib_umem_get(context, ucmd.buf_addr, buf_size,
-				       IB_ACCESS_LOCAL_WRITE);
-		if (IS_ERR(cq->umem)) {
-			err = PTR_ERR(cq->umem);
-			goto err_cq;
-		}
-
-		err = mlx4_mtt_init(dev->dev, ib_umem_page_count(cq->umem),
-				    ilog2(cq->umem->page_size), &cq->buf.mtt);
+		err = mlx4_ib_get_cq_umem(dev, context, &cq->buf, &cq->umem,
+					  ucmd.buf_addr, entries);
 		if (err)
-			goto err_buf;
-
-		err = mlx4_ib_umem_write_mtt(dev, &cq->buf.mtt, cq->umem);
-		if (err)
-			goto err_mtt;
+			goto err_cq;
 
 		err = mlx4_ib_db_map_user(to_mucontext(context), ucmd.db_addr,
 					  &cq->db);
@@ -155,19 +213,9 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector
 		*cq->mcq.set_ci_db = 0;
 		*cq->mcq.arm_db    = 0;
 
-		if (mlx4_buf_alloc(dev->dev, buf_size, PAGE_SIZE * 2, &cq->buf.buf)) {
-			err = -ENOMEM;
-			goto err_db;
-		}
-
-		err = mlx4_mtt_init(dev->dev, cq->buf.buf.npages, cq->buf.buf.page_shift,
-				    &cq->buf.mtt);
+		err = mlx4_ib_alloc_cq_buf(dev, &cq->buf, entries);
 		if (err)
-			goto err_buf;
-
-		err = mlx4_buf_write_mtt(dev->dev, &cq->buf.mtt, &cq->buf.buf);
-		if (err)
-			goto err_mtt;
+			goto err_db;
 
 		uar = &dev->priv_uar;
 	}
@@ -195,12 +243,10 @@ err_dbmap:
 err_mtt:
 	mlx4_mtt_cleanup(dev->dev, &cq->buf.mtt);
 
-err_buf:
 	if (context)
 		ib_umem_release(cq->umem);
 	else
-		mlx4_buf_free(dev->dev, entries * sizeof (struct mlx4_cqe),
-			      &cq->buf.buf);
+		mlx4_ib_free_cq_buf(dev, &cq->buf, entries);
 
 err_db:
 	if (!context)
@@ -212,6 +258,170 @@ err_cq:
 	return ERR_PTR(err);
 }
 
+static int mlx4_alloc_resize_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq *cq,
+				  int entries)
+{
+	int err;
+
+	if (cq->resize_buf)
+		return -EBUSY;
+
+	cq->resize_buf = kmalloc(sizeof *cq->resize_buf, GFP_ATOMIC);
+	if (!cq->resize_buf)
+		return -ENOMEM;
+
+	err = mlx4_ib_alloc_cq_buf(dev, &cq->resize_buf->buf, entries);
+	if (err) {
+		kfree(cq->resize_buf);
+		cq->resize_buf = NULL;
+		return err;
+	}
+
+	cq->resize_buf->cqe = entries - 1;
+
+	return 0;
+}
+
+static int mlx4_alloc_resize_umem(struct mlx4_ib_dev *dev, struct mlx4_ib_cq *cq,
+				   int entries, struct ib_udata *udata)
+{
+	struct mlx4_ib_resize_cq ucmd;
+	int err;
+
+	if (cq->resize_umem)
+		return -EBUSY;
+
+	if (ib_copy_from_udata(&ucmd, udata, sizeof ucmd))
+		return -EFAULT;
+
+	cq->resize_buf = kmalloc(sizeof *cq->resize_buf, GFP_ATOMIC);
+	if (!cq->resize_buf)
+		return -ENOMEM;
+
+	err = mlx4_ib_get_cq_umem(dev, cq->umem->context, &cq->resize_buf->buf,
+				  &cq->resize_umem, ucmd.buf_addr, entries);
+	if (err) {
+		kfree(cq->resize_buf);
+		cq->resize_buf = NULL;
+		return err;
+	}
+
+	cq->resize_buf->cqe = entries - 1;
+
+	return 0;
+}
+
+static int mlx4_ib_get_outstanding_cqes(struct mlx4_ib_cq *cq)
+{
+	u32 i;
+
+	i = cq->mcq.cons_index;
+	while (get_sw_cqe(cq, i & cq->ibcq.cqe))
+		++i;
+
+	return i - cq->mcq.cons_index;
+}
+
+static void mlx4_ib_cq_resize_copy_cqes(struct mlx4_ib_cq *cq)
+{
+	struct mlx4_cqe *cqe;
+	int i;
+
+	i = cq->mcq.cons_index;
+	cqe = get_cqe(cq, i & cq->ibcq.cqe);
+	while ((cqe->owner_sr_opcode & MLX4_CQE_OPCODE_MASK) != MLX4_CQE_OPCODE_RESIZE) {
+		memcpy(get_cqe_from_buf(&cq->resize_buf->buf,
+					(i + 1) & cq->resize_buf->cqe),
+			get_cqe(cq, i & cq->ibcq.cqe), sizeof(struct mlx4_cqe));
+		cqe = get_cqe(cq, ++i & cq->ibcq.cqe);
+	}
+	++cq->mcq.cons_index;
+}
+
+int mlx4_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
+{
+	struct mlx4_ib_dev *dev = to_mdev(ibcq->device);
+	struct mlx4_ib_cq *cq = to_mcq(ibcq);
+	int outst_cqe;
+	int err;
+
+	mutex_lock(&cq->resize_mutex);
+
+	if (entries < 1 || entries > dev->dev->caps.max_cqes) {
+		err = -EINVAL;
+		goto out;
+	}
+
+	entries = roundup_pow_of_two(entries + 1);
+	if (entries == ibcq->cqe + 1) {
+		err = 0;
+		goto out;
+	}
+
+	if (ibcq->uobject) {
+		err = mlx4_alloc_resize_umem(dev, cq, entries, udata);
+		if (err)
+			goto out;
+	} else {
+		/* Can't be smaller then the number of outstanding CQEs */
+		outst_cqe = mlx4_ib_get_outstanding_cqes(cq);
+		if (entries < outst_cqe + 1) {
+			err = 0;
+			goto out;
+		}
+
+		err = mlx4_alloc_resize_buf(dev, cq, entries);
+		if (err)
+			goto out;
+	}
+
+	err = mlx4_cq_resize(dev->dev, &cq->mcq, entries, &cq->resize_buf->buf.mtt);
+	if (err)
+		goto err_buf;
+
+	if (ibcq->uobject) {
+		cq->buf      = cq->resize_buf->buf;
+		cq->ibcq.cqe = cq->resize_buf->cqe;
+		ib_umem_release(cq->umem);
+		cq->umem     = cq->resize_umem;
+
+		kfree(cq->resize_buf);
+		cq->resize_buf = NULL;
+		cq->resize_umem = NULL;
+	} else {
+		spin_lock_irq(&cq->lock);
+		if (cq->resize_buf) {
+			mlx4_ib_cq_resize_copy_cqes(cq);
+			mlx4_ib_free_cq_buf(dev, &cq->buf, cq->ibcq.cqe);
+			cq->buf      = cq->resize_buf->buf;
+			cq->ibcq.cqe = cq->resize_buf->cqe;
+
+			kfree(cq->resize_buf);
+			cq->resize_buf = NULL;
+		}
+		spin_unlock_irq(&cq->lock);
+	}
+
+	goto out;
+
+err_buf:
+	if (!ibcq->uobject)
+		mlx4_ib_free_cq_buf(dev, &cq->resize_buf->buf,
+				    cq->resize_buf->cqe);
+
+	kfree(cq->resize_buf);
+	cq->resize_buf = NULL;
+
+	if (cq->resize_umem) {
+		ib_umem_release(cq->resize_umem);
+		cq->resize_umem = NULL;
+	}
+
+out:
+	mutex_unlock(&cq->resize_mutex);
+	return err;
+}
+
 int mlx4_ib_destroy_cq(struct ib_cq *cq)
 {
 	struct mlx4_ib_dev *dev = to_mdev(cq->device);
@@ -224,8 +434,7 @@ int mlx4_ib_destroy_cq(struct ib_cq *cq)
 		mlx4_ib_db_unmap_user(to_mucontext(cq->uobject->context), &mcq->db);
 		ib_umem_release(mcq->umem);
 	} else {
-		mlx4_buf_free(dev->dev, (cq->cqe + 1) * sizeof (struct mlx4_cqe),
-			      &mcq->buf.buf);
+		mlx4_ib_free_cq_buf(dev, &mcq->buf, cq->cqe + 1);
 		mlx4_ib_db_free(dev, &mcq->db);
 	}
 
@@ -332,6 +541,7 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq,
 	u32 g_mlpath_rqpn;
 	u16 wqe_ctr;
 
+repoll:
 	cqe = next_cqe_sw(cq);
 	if (!cqe)
 		return -EAGAIN;
@@ -354,6 +564,22 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq,
 		return -EINVAL;
 	}
 
+	/* Resize CQ in progress */
+	if (unlikely((cqe->owner_sr_opcode & MLX4_CQE_OPCODE_MASK) == MLX4_CQE_OPCODE_RESIZE)) {
+		if (cq->resize_buf) {
+			struct mlx4_ib_dev *dev = to_mdev(cq->ibcq.device);
+
+			mlx4_ib_free_cq_buf(dev, &cq->buf, cq->ibcq.cqe);
+			cq->buf      = cq->resize_buf->buf;
+			cq->ibcq.cqe = cq->resize_buf->cqe;
+
+			kfree(cq->resize_buf);
+			cq->resize_buf = NULL;
+		}
+
+		goto repoll;
+	}
+
 	if (!*cur_qp ||
 	    (be32_to_cpu(cqe->my_qpn) & 0xffffff) != (*cur_qp)->mqp.qpn) {
 		/*
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 76dd45c764b4..57885cd61937 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -571,6 +571,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 		(1ull << IB_USER_VERBS_CMD_DEREG_MR)		|
 		(1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL)	|
 		(1ull << IB_USER_VERBS_CMD_CREATE_CQ)		|
+		(1ull << IB_USER_VERBS_CMD_RESIZE_CQ)		|
 		(1ull << IB_USER_VERBS_CMD_DESTROY_CQ)		|
 		(1ull << IB_USER_VERBS_CMD_CREATE_QP)		|
 		(1ull << IB_USER_VERBS_CMD_MODIFY_QP)		|
@@ -610,6 +611,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	ibdev->ib_dev.post_recv		= mlx4_ib_post_recv;
 	ibdev->ib_dev.create_cq		= mlx4_ib_create_cq;
 	ibdev->ib_dev.modify_cq		= mlx4_ib_modify_cq;
+	ibdev->ib_dev.resize_cq		= mlx4_ib_resize_cq;
 	ibdev->ib_dev.destroy_cq	= mlx4_ib_destroy_cq;
 	ibdev->ib_dev.poll_cq		= mlx4_ib_poll_cq;
 	ibdev->ib_dev.req_notify_cq	= mlx4_ib_arm_cq;
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index ef8ad96e1432..9e637323c155 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -78,13 +78,21 @@ struct mlx4_ib_cq_buf {
 	struct mlx4_mtt		mtt;
 };
 
+struct mlx4_ib_cq_resize {
+	struct mlx4_ib_cq_buf	buf;
+	int			cqe;
+};
+
 struct mlx4_ib_cq {
 	struct ib_cq		ibcq;
 	struct mlx4_cq		mcq;
 	struct mlx4_ib_cq_buf	buf;
+	struct mlx4_ib_cq_resize *resize_buf;
 	struct mlx4_ib_db	db;
 	spinlock_t		lock;
+	struct mutex		resize_mutex;
 	struct ib_umem	       *umem;
+	struct ib_umem	       *resize_umem;
 };
 
 struct mlx4_ib_mr {
@@ -255,6 +263,7 @@ struct ib_mr *mlx4_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 int mlx4_ib_dereg_mr(struct ib_mr *mr);
 
 int mlx4_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period);
+int mlx4_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata);
 struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector,
 				struct ib_ucontext *context,
 				struct ib_udata *udata);
diff --git a/drivers/net/mlx4/cq.c b/drivers/net/mlx4/cq.c
index 8c314341434f..caa5bcf54e35 100644
--- a/drivers/net/mlx4/cq.c
+++ b/drivers/net/mlx4/cq.c
@@ -159,6 +159,34 @@ int mlx4_cq_modify(struct mlx4_dev *dev, struct mlx4_cq *cq,
 }
 EXPORT_SYMBOL_GPL(mlx4_cq_modify);
 
+int mlx4_cq_resize(struct mlx4_dev *dev, struct mlx4_cq *cq,
+		   int entries, struct mlx4_mtt *mtt)
+{
+	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_cq_context *cq_context;
+	u64 mtt_addr;
+	int err;
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox))
+		return PTR_ERR(mailbox);
+
+	cq_context = mailbox->buf;
+	memset(cq_context, 0, sizeof *cq_context);
+
+	cq_context->logsize_usrpage = cpu_to_be32(ilog2(entries) << 24);
+	cq_context->log_page_size   = mtt->page_shift - 12;
+	mtt_addr = mlx4_mtt_addr(dev, mtt);
+	cq_context->mtt_base_addr_h = mtt_addr >> 32;
+	cq_context->mtt_base_addr_l = cpu_to_be32(mtt_addr & 0xffffffff);
+
+	err = mlx4_MODIFY_CQ(dev, mailbox, cq->cqn, 1);
+
+	mlx4_free_cmd_mailbox(dev, mailbox);
+	return err;
+}
+EXPORT_SYMBOL_GPL(mlx4_cq_resize);
+
 int mlx4_cq_alloc(struct mlx4_dev *dev, int nent, struct mlx4_mtt *mtt,
 		  struct mlx4_uar *uar, u64 db_rec, struct mlx4_cq *cq)
 {
diff --git a/include/linux/mlx4/cq.h b/include/linux/mlx4/cq.h
index f7c3511c594b..071cf96cf01f 100644
--- a/include/linux/mlx4/cq.h
+++ b/include/linux/mlx4/cq.h
@@ -132,5 +132,7 @@ enum {
 
 int mlx4_cq_modify(struct mlx4_dev *dev, struct mlx4_cq *cq,
 		   u16 count, u16 period);
+int mlx4_cq_resize(struct mlx4_dev *dev, struct mlx4_cq *cq,
+		   int entries, struct mlx4_mtt *mtt);
 
 #endif /* MLX4_CQ_H */
-- 
cgit v1.2.3


From d7b906897e9caae452947e33674df0a2d6f7e10f Mon Sep 17 00:00:00 2001
From: Russell King <rmk+lkml@arm.linux.org.uk>
Date: Thu, 17 Apr 2008 07:46:24 +0200
Subject: [S390] genirq/clockevents: move irq affinity prototypes/inlines to
 interrupt.h

> Generic code is not supposed to include irq.h. Replace this include
> by linux/hardirq.h instead and add/replace an include of linux/irq.h
> in asm header files where necessary.
> This change should only matter for architectures that make use of
> GENERIC_CLOCKEVENTS.
> Architectures in question are mips, x86, arm, sh, powerpc, uml and sparc64.
>
> I did some cross compile tests for mips, x86_64, arm, powerpc and sparc64.
> This patch fixes also build breakages caused by the include replacement in
> tick-common.h.

I generally dislike adding optional linux/* includes in asm/* includes -
I'm nervous about this causing include loops.

However, there's a separate point to be discussed here.

That is, what interfaces are expected of every architecture in the kernel.
If generic code wants to be able to set the affinity of interrupts, then
that needs to become part of the interfaces listed in linux/interrupt.h
rather than linux/irq.h.

So what I suggest is this approach instead (against Linus' tree of a
couple of days ago) - we move irq_set_affinity() and irq_can_set_affinity()
to linux/interrupt.h, change the linux/irq.h includes to linux/interrupt.h
and include asm/irq_regs.h where needed (asm/irq_regs.h is supposed to be
rarely used include since not much touches the stacked parent context
registers.)

Build tested on ARM PXA family kernels and ARM's Realview platform
kernels which both use genirq.

[ tglx@linutronix.de: add GENERIC_HARDIRQ dependencies ]

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
---
 include/linux/interrupt.h    | 19 +++++++++++++++++++
 include/linux/irq.h          | 10 ----------
 kernel/time/tick-broadcast.c |  2 +-
 kernel/time/tick-common.c    |  4 +++-
 kernel/time/tick-oneshot.c   |  2 +-
 5 files changed, 24 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
index f8ab4ce70564..b5fef13148bd 100644
--- a/include/linux/interrupt.h
+++ b/include/linux/interrupt.h
@@ -102,6 +102,25 @@ extern void disable_irq_nosync(unsigned int irq);
 extern void disable_irq(unsigned int irq);
 extern void enable_irq(unsigned int irq);
 
+#if defined(CONFIG_SMP) && defined(CONFIG_GENERIC_HARDIRQS)
+
+extern int irq_set_affinity(unsigned int irq, cpumask_t cpumask);
+extern int irq_can_set_affinity(unsigned int irq);
+
+#else /* CONFIG_SMP */
+
+static inline int irq_set_affinity(unsigned int irq, cpumask_t cpumask)
+{
+	return -EINVAL;
+}
+
+static inline int irq_can_set_affinity(unsigned int irq)
+{
+	return 0;
+}
+
+#endif /* CONFIG_SMP && CONFIG_GENERIC_HARDIRQS */
+
 #ifdef CONFIG_GENERIC_HARDIRQS
 /*
  * Special lockdep variants of irq disabling/enabling.
diff --git a/include/linux/irq.h b/include/linux/irq.h
index 176e5e790a44..1883a85625dd 100644
--- a/include/linux/irq.h
+++ b/include/linux/irq.h
@@ -228,21 +228,11 @@ static inline void set_pending_irq(unsigned int irq, cpumask_t mask)
 
 #endif /* CONFIG_GENERIC_PENDING_IRQ */
 
-extern int irq_set_affinity(unsigned int irq, cpumask_t cpumask);
-extern int irq_can_set_affinity(unsigned int irq);
-
 #else /* CONFIG_SMP */
 
 #define move_native_irq(x)
 #define move_masked_irq(x)
 
-static inline int irq_set_affinity(unsigned int irq, cpumask_t cpumask)
-{
-	return -EINVAL;
-}
-
-static inline int irq_can_set_affinity(unsigned int irq) { return 0; }
-
 #endif /* CONFIG_SMP */
 
 #ifdef CONFIG_IRQBALANCE
diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c
index e1bd50cbbf5d..fdfa0c745bb6 100644
--- a/kernel/time/tick-broadcast.c
+++ b/kernel/time/tick-broadcast.c
@@ -14,7 +14,7 @@
 #include <linux/cpu.h>
 #include <linux/err.h>
 #include <linux/hrtimer.h>
-#include <linux/irq.h>
+#include <linux/interrupt.h>
 #include <linux/percpu.h>
 #include <linux/profile.h>
 #include <linux/sched.h>
diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c
index 1bea399a9ef0..4f3886562b8c 100644
--- a/kernel/time/tick-common.c
+++ b/kernel/time/tick-common.c
@@ -14,12 +14,14 @@
 #include <linux/cpu.h>
 #include <linux/err.h>
 #include <linux/hrtimer.h>
-#include <linux/irq.h>
+#include <linux/interrupt.h>
 #include <linux/percpu.h>
 #include <linux/profile.h>
 #include <linux/sched.h>
 #include <linux/tick.h>
 
+#include <asm/irq_regs.h>
+
 #include "tick-internal.h"
 
 /*
diff --git a/kernel/time/tick-oneshot.c b/kernel/time/tick-oneshot.c
index 0258d3115d54..450c04935b66 100644
--- a/kernel/time/tick-oneshot.c
+++ b/kernel/time/tick-oneshot.c
@@ -14,7 +14,7 @@
 #include <linux/cpu.h>
 #include <linux/err.h>
 #include <linux/hrtimer.h>
-#include <linux/irq.h>
+#include <linux/interrupt.h>
 #include <linux/percpu.h>
 #include <linux/profile.h>
 #include <linux/sched.h>
-- 
cgit v1.2.3


From a332d86d3c262cddd3de0bfa90e1910de60b4f95 Mon Sep 17 00:00:00 2001
From: Thomas Gleixner <tglx@linutronix.de>
Date: Sun, 10 Feb 2008 09:04:12 +0100
Subject: hrtimer: add nanosleep specific restart_block member

The back and forth typecasting of restart_block->args is horrible. We
added a separate union member for futex already. Do the same for
nanosleep.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/thread_info.h | 12 ++++++++++++
 1 file changed, 12 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h
index 421323e5a2d6..accd7bad35b0 100644
--- a/include/linux/thread_info.h
+++ b/include/linux/thread_info.h
@@ -9,6 +9,9 @@
 
 #include <linux/types.h>
 
+struct timespec;
+struct compat_timespec;
+
 /*
  * System call restart block.
  */
@@ -26,6 +29,15 @@ struct restart_block {
 			u32 bitset;
 			u64 time;
 		} futex;
+		/* For nanosleep */
+		struct {
+			clockid_t index;
+			struct timespec __user *rmtp;
+#ifdef CONFIG_COMPAT
+			struct compat_timespec __user *compat_rmtp;
+#endif
+			u64 expires;
+		} nanosleep;
 	};
 };
 
-- 
cgit v1.2.3


From 8e60e05fdc7344415fa69a3883b11f65db967b47 Mon Sep 17 00:00:00 2001
From: Oleg Nesterov <oleg@tv-sign.ru>
Date: Fri, 4 Apr 2008 20:54:10 +0200
Subject: hrtimers: simplify lockdep handling

In order to avoid the false positive from lockdep, each per-cpu base->lock has
the separate lock class and migrate_hrtimers() uses double_spin_lock().

This is overcomplicated: except for migrate_hrtimers() we never take 2 locks
at once, and migrate_hrtimers() can use spin_lock_nested().

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/hrtimer.h | 2 --
 kernel/hrtimer.c        | 9 ++++-----
 2 files changed, 4 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h
index 1ad56a7b2f74..56f3236da829 100644
--- a/include/linux/hrtimer.h
+++ b/include/linux/hrtimer.h
@@ -173,7 +173,6 @@ struct hrtimer_clock_base {
  * struct hrtimer_cpu_base - the per cpu clock bases
  * @lock:		lock protecting the base and associated clock bases
  *			and timers
- * @lock_key:		the lock_class_key for use with lockdep
  * @clock_base:		array of clock bases for this cpu
  * @curr_timer:		the timer which is executing a callback right now
  * @expires_next:	absolute time of the next event which was scheduled
@@ -189,7 +188,6 @@ struct hrtimer_clock_base {
  */
 struct hrtimer_cpu_base {
 	spinlock_t			lock;
-	struct lock_class_key		lock_key;
 	struct hrtimer_clock_base	clock_base[HRTIMER_MAX_CLOCK_BASES];
 	struct list_head		cb_pending;
 #ifdef CONFIG_HIGH_RES_TIMERS
diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
index 911e87d0440d..c642ef75069f 100644
--- a/kernel/hrtimer.c
+++ b/kernel/hrtimer.c
@@ -1424,7 +1424,6 @@ static void __cpuinit init_hrtimers_cpu(int cpu)
 	int i;
 
 	spin_lock_init(&cpu_base->lock);
-	lockdep_set_class(&cpu_base->lock, &cpu_base->lock_key);
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++)
 		cpu_base->clock_base[i].cpu_base = cpu_base;
@@ -1465,16 +1464,16 @@ static void migrate_hrtimers(int cpu)
 	tick_cancel_sched_timer(cpu);
 
 	local_irq_disable();
-	double_spin_lock(&new_base->lock, &old_base->lock,
-			 smp_processor_id() < cpu);
+	spin_lock(&new_base->lock);
+	spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING);
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) {
 		migrate_hrtimer_list(&old_base->clock_base[i],
 				     &new_base->clock_base[i]);
 	}
 
-	double_spin_unlock(&new_base->lock, &old_base->lock,
-			   smp_processor_id() < cpu);
+	spin_unlock(&old_base->lock);
+	spin_unlock(&new_base->lock);
 	local_irq_enable();
 	put_cpu_var(hrtimer_bases);
 }
-- 
cgit v1.2.3


From 3f3eafc921e2378954c28cfd0eb10910449f4c11 Mon Sep 17 00:00:00 2001
From: Oleg Nesterov <oleg@tv-sign.ru>
Date: Fri, 4 Apr 2008 20:54:10 +0200
Subject: locking: remove unused double_spin_lock()

double_spin_lock() has no callers, and it can't be used without additional
lockdep annotations, remove it.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/spinlock.h | 37 -------------------------------------
 1 file changed, 37 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h
index 1129ee0a7180..d311a090fae7 100644
--- a/include/linux/spinlock.h
+++ b/include/linux/spinlock.h
@@ -295,43 +295,6 @@ do {						\
 	1 : ({ local_irq_restore(flags); 0; }); \
 })
 
-/*
- * Locks two spinlocks l1 and l2.
- * l1_first indicates if spinlock l1 should be taken first.
- */
-static inline void double_spin_lock(spinlock_t *l1, spinlock_t *l2,
-				    bool l1_first)
-	__acquires(l1)
-	__acquires(l2)
-{
-	if (l1_first) {
-		spin_lock(l1);
-		spin_lock(l2);
-	} else {
-		spin_lock(l2);
-		spin_lock(l1);
-	}
-}
-
-/*
- * Unlocks two spinlocks l1 and l2.
- * l1_taken_first indicates if spinlock l1 was taken first and therefore
- * should be released after spinlock l2.
- */
-static inline void double_spin_unlock(spinlock_t *l1, spinlock_t *l2,
-				      bool l1_taken_first)
-	__releases(l1)
-	__releases(l2)
-{
-	if (l1_taken_first) {
-		spin_unlock(l2);
-		spin_unlock(l1);
-	} else {
-		spin_unlock(l1);
-		spin_unlock(l2);
-	}
-}
-
 /*
  * Pull the atomic_t declaration:
  * (asm-mips/atomic.h needs above definitions)
-- 
cgit v1.2.3


From 8b91de2e58318d1168bc13d164478c1a7217a63a Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Tue, 26 Feb 2008 09:53:20 -0500
Subject: Fix quota.h includes

quota.h currently relies on asm/semaphore.h (through some chain; it
doesn't actually include semaphore.h itself) to include wait.h.  As
well as being bad practice to rely on an implicit include, subsequent
patches will break this.  While I'm in this file, add atomic.h and
list.h, and sort the list of includes.

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
---
 include/linux/quota.h | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/quota.h b/include/linux/quota.h
index 6e0393a5b2ea..eb560d031acd 100644
--- a/include/linux/quota.h
+++ b/include/linux/quota.h
@@ -160,14 +160,18 @@ enum {
 
 
 #ifdef __KERNEL__
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
+#include <linux/list.h>
 #include <linux/mutex.h>
+#include <linux/rwsem.h>
+#include <linux/spinlock.h>
+#include <linux/wait.h>
 
 #include <linux/dqblk_xfs.h>
 #include <linux/dqblk_v1.h>
 #include <linux/dqblk_v2.h>
 
+#include <asm/atomic.h>
+
 extern spinlock_t dq_data_lock;
 
 /* Maximal numbers of writes for quota operation (insert/delete/update)
-- 
cgit v1.2.3


From 64ac24e738823161693bf791f87adc802cf529ff Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Fri, 7 Mar 2008 21:55:58 -0500
Subject: Generic semaphore implementation

Semaphores are no longer performance-critical, so a generic C
implementation is better for maintainability, debuggability and
extensibility.  Thanks to Peter Zijlstra for fixing the lockdep
warning.  Thanks to Harvey Harrison for pointing out that the
unlikely() was unnecessary.

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
---
 arch/alpha/kernel/Makefile               |   2 +-
 arch/alpha/kernel/alpha_ksyms.c          |   9 --
 arch/alpha/kernel/semaphore.c            | 224 ---------------------------
 arch/arm/kernel/Makefile                 |   2 +-
 arch/arm/kernel/semaphore.c              | 221 ---------------------------
 arch/avr32/kernel/Makefile               |   2 +-
 arch/avr32/kernel/semaphore.c            | 148 ------------------
 arch/blackfin/Kconfig                    |   4 -
 arch/blackfin/kernel/bfin_ksyms.c        |   5 -
 arch/cris/kernel/Makefile                |   3 +-
 arch/cris/kernel/crisksyms.c             |   7 -
 arch/cris/kernel/semaphore.c             | 129 ----------------
 arch/frv/kernel/Makefile                 |   2 +-
 arch/frv/kernel/frv_ksyms.c              |   1 -
 arch/frv/kernel/semaphore.c              | 155 -------------------
 arch/h8300/kernel/Makefile               |   2 +-
 arch/h8300/kernel/h8300_ksyms.c          |   1 -
 arch/h8300/kernel/semaphore.c            | 132 ----------------
 arch/ia64/kernel/Makefile                |   2 +-
 arch/ia64/kernel/ia64_ksyms.c            |   6 -
 arch/ia64/kernel/semaphore.c             | 165 --------------------
 arch/m32r/kernel/Makefile                |   2 +-
 arch/m32r/kernel/m32r_ksyms.c            |   5 -
 arch/m32r/kernel/semaphore.c             | 185 ----------------------
 arch/m68k/kernel/Makefile                |   2 +-
 arch/m68k/kernel/m68k_ksyms.c            |   6 -
 arch/m68k/kernel/semaphore.c             | 132 ----------------
 arch/m68k/lib/Makefile                   |   2 +-
 arch/m68k/lib/semaphore.S                |  53 -------
 arch/m68knommu/kernel/Makefile           |   2 +-
 arch/m68knommu/kernel/m68k_ksyms.c       |   6 -
 arch/m68knommu/kernel/semaphore.c        | 133 ----------------
 arch/m68knommu/lib/Makefile              |   2 +-
 arch/m68knommu/lib/semaphore.S           |  66 --------
 arch/mips/kernel/Makefile                |   2 +-
 arch/mips/kernel/semaphore.c             | 168 --------------------
 arch/mn10300/kernel/Makefile             |   2 +-
 arch/mn10300/kernel/semaphore.c          | 149 ------------------
 arch/parisc/kernel/Makefile              |   2 +-
 arch/parisc/kernel/parisc_ksyms.c        |   5 -
 arch/parisc/kernel/semaphore.c           | 102 -------------
 arch/powerpc/kernel/Makefile             |   2 +-
 arch/powerpc/kernel/ppc_ksyms.c          |   1 -
 arch/powerpc/kernel/semaphore.c          | 135 ----------------
 arch/ppc/kernel/semaphore.c              | 131 ----------------
 arch/s390/kernel/Makefile                |   2 +-
 arch/s390/kernel/s390_ksyms.c            |   7 -
 arch/s390/kernel/semaphore.c             | 108 -------------
 arch/sh/kernel/Makefile_32               |   2 +-
 arch/sh/kernel/Makefile_64               |   2 +-
 arch/sh/kernel/semaphore.c               | 139 -----------------
 arch/sh/kernel/sh_ksyms_32.c             |   7 -
 arch/sh/kernel/sh_ksyms_64.c             |   4 -
 arch/sparc/kernel/Makefile               |   2 +-
 arch/sparc/kernel/semaphore.c            | 155 -------------------
 arch/sparc/kernel/sparc_ksyms.c          |   5 -
 arch/sparc64/kernel/Makefile             |   2 +-
 arch/sparc64/kernel/semaphore.c          | 254 -------------------------------
 arch/sparc64/kernel/sparc64_ksyms.c      |   6 -
 arch/um/Kconfig.i386                     |   4 -
 arch/um/Kconfig.x86_64                   |   4 -
 arch/um/sys-i386/ksyms.c                 |  12 --
 arch/um/sys-ppc/Makefile                 |   8 +-
 arch/um/sys-x86_64/ksyms.c               |  13 +-
 arch/v850/kernel/Makefile                |   2 +-
 arch/v850/kernel/semaphore.c             | 166 --------------------
 arch/v850/kernel/v850_ksyms.c            |   7 -
 arch/x86/Kconfig                         |   3 -
 arch/x86/kernel/i386_ksyms_32.c          |   5 -
 arch/x86/kernel/x8664_ksyms_64.c         |   6 -
 arch/x86/lib/semaphore_32.S              |  83 ----------
 arch/x86/lib/thunk_64.S                  |   5 -
 arch/xtensa/kernel/Makefile              |   2 +-
 arch/xtensa/kernel/semaphore.c           | 226 ---------------------------
 arch/xtensa/kernel/xtensa_ksyms.c        |   9 --
 include/asm-alpha/semaphore.h            | 150 +-----------------
 include/asm-arm/semaphore-helper.h       |  84 ----------
 include/asm-arm/semaphore.h              |  99 +-----------
 include/asm-avr32/semaphore.h            | 109 +------------
 include/asm-blackfin/semaphore-helper.h  |  82 ----------
 include/asm-blackfin/semaphore.h         | 106 +------------
 include/asm-cris/semaphore-helper.h      |  78 ----------
 include/asm-cris/semaphore.h             | 134 +---------------
 include/asm-frv/semaphore.h              | 156 +------------------
 include/asm-h8300/semaphore-helper.h     |  85 -----------
 include/asm-h8300/semaphore.h            | 191 +----------------------
 include/asm-ia64/semaphore.h             | 100 +-----------
 include/asm-m32r/semaphore.h             | 145 +-----------------
 include/asm-m68k/semaphore-helper.h      | 142 -----------------
 include/asm-m68k/semaphore.h             | 164 +-------------------
 include/asm-m68knommu/semaphore-helper.h |  82 ----------
 include/asm-m68knommu/semaphore.h        | 154 +------------------
 include/asm-mips/semaphore.h             | 109 +------------
 include/asm-mn10300/semaphore.h          | 170 +--------------------
 include/asm-parisc/semaphore-helper.h    |  89 -----------
 include/asm-parisc/semaphore.h           | 146 +-----------------
 include/asm-powerpc/semaphore.h          |  95 +-----------
 include/asm-s390/semaphore.h             | 108 +------------
 include/asm-sh/semaphore-helper.h        |  89 -----------
 include/asm-sh/semaphore.h               | 116 +-------------
 include/asm-sparc/semaphore.h            | 193 +----------------------
 include/asm-sparc64/semaphore.h          |  54 +------
 include/asm-um/semaphore.h               |   7 +-
 include/asm-v850/semaphore.h             |  85 +----------
 include/asm-x86/semaphore.h              |   6 +-
 include/asm-x86/semaphore_32.h           | 175 ---------------------
 include/asm-x86/semaphore_64.h           | 180 ----------------------
 include/asm-xtensa/semaphore.h           | 100 +-----------
 include/linux/semaphore.h                |  77 ++++++++++
 kernel/Makefile                          |   2 +-
 kernel/semaphore.c                       | 187 +++++++++++++++++++++++
 lib/Makefile                             |   1 -
 lib/semaphore-sleepers.c                 | 176 ---------------------
 113 files changed, 314 insertions(+), 7679 deletions(-)
 delete mode 100644 arch/alpha/kernel/semaphore.c
 delete mode 100644 arch/arm/kernel/semaphore.c
 delete mode 100644 arch/avr32/kernel/semaphore.c
 delete mode 100644 arch/cris/kernel/semaphore.c
 delete mode 100644 arch/frv/kernel/semaphore.c
 delete mode 100644 arch/h8300/kernel/semaphore.c
 delete mode 100644 arch/ia64/kernel/semaphore.c
 delete mode 100644 arch/m32r/kernel/semaphore.c
 delete mode 100644 arch/m68k/kernel/semaphore.c
 delete mode 100644 arch/m68k/lib/semaphore.S
 delete mode 100644 arch/m68knommu/kernel/semaphore.c
 delete mode 100644 arch/m68knommu/lib/semaphore.S
 delete mode 100644 arch/mips/kernel/semaphore.c
 delete mode 100644 arch/mn10300/kernel/semaphore.c
 delete mode 100644 arch/parisc/kernel/semaphore.c
 delete mode 100644 arch/powerpc/kernel/semaphore.c
 delete mode 100644 arch/ppc/kernel/semaphore.c
 delete mode 100644 arch/s390/kernel/semaphore.c
 delete mode 100644 arch/sh/kernel/semaphore.c
 delete mode 100644 arch/sparc/kernel/semaphore.c
 delete mode 100644 arch/sparc64/kernel/semaphore.c
 delete mode 100644 arch/v850/kernel/semaphore.c
 delete mode 100644 arch/xtensa/kernel/semaphore.c
 delete mode 100644 include/asm-arm/semaphore-helper.h
 delete mode 100644 include/asm-blackfin/semaphore-helper.h
 delete mode 100644 include/asm-cris/semaphore-helper.h
 delete mode 100644 include/asm-h8300/semaphore-helper.h
 delete mode 100644 include/asm-m68k/semaphore-helper.h
 delete mode 100644 include/asm-m68knommu/semaphore-helper.h
 delete mode 100644 include/asm-parisc/semaphore-helper.h
 delete mode 100644 include/asm-sh/semaphore-helper.h
 delete mode 100644 include/asm-x86/semaphore_32.h
 delete mode 100644 include/asm-x86/semaphore_64.h
 create mode 100644 include/linux/semaphore.h
 create mode 100644 kernel/semaphore.c
 delete mode 100644 lib/semaphore-sleepers.c

(limited to 'include/linux')

diff --git a/arch/alpha/kernel/Makefile b/arch/alpha/kernel/Makefile
index dccf05245d4d..ac706c1d7ada 100644
--- a/arch/alpha/kernel/Makefile
+++ b/arch/alpha/kernel/Makefile
@@ -7,7 +7,7 @@ EXTRA_AFLAGS	:= $(KBUILD_CFLAGS)
 EXTRA_CFLAGS	:= -Werror -Wno-sign-compare
 
 obj-y    := entry.o traps.o process.o init_task.o osf_sys.o irq.o \
-	    irq_alpha.o signal.o setup.o ptrace.o time.o semaphore.o \
+	    irq_alpha.o signal.o setup.o ptrace.o time.o \
 	    alpha_ksyms.o systbls.o err_common.o io.o
 
 obj-$(CONFIG_VGA_HOSE)	+= console.o
diff --git a/arch/alpha/kernel/alpha_ksyms.c b/arch/alpha/kernel/alpha_ksyms.c
index e9762a33b043..d96e742d4dc2 100644
--- a/arch/alpha/kernel/alpha_ksyms.c
+++ b/arch/alpha/kernel/alpha_ksyms.c
@@ -77,15 +77,6 @@ EXPORT_SYMBOL(__do_clear_user);
 EXPORT_SYMBOL(__strncpy_from_user);
 EXPORT_SYMBOL(__strnlen_user);
 
-/* Semaphore helper functions.  */
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_failed_interruptible);
-EXPORT_SYMBOL(__up_wakeup);
-EXPORT_SYMBOL(down);
-EXPORT_SYMBOL(down_interruptible);
-EXPORT_SYMBOL(down_trylock);
-EXPORT_SYMBOL(up);
-
 /* 
  * SMP-specific symbols.
  */
diff --git a/arch/alpha/kernel/semaphore.c b/arch/alpha/kernel/semaphore.c
deleted file mode 100644
index 8d2982aa1b8d..000000000000
--- a/arch/alpha/kernel/semaphore.c
+++ /dev/null
@@ -1,224 +0,0 @@
-/*
- * Alpha semaphore implementation.
- *
- * (C) Copyright 1996 Linus Torvalds
- * (C) Copyright 1999, 2000 Richard Henderson
- */
-
-#include <linux/errno.h>
-#include <linux/sched.h>
-#include <linux/init.h>
-
-/*
- * This is basically the PPC semaphore scheme ported to use
- * the Alpha ll/sc sequences, so see the PPC code for
- * credits.
- */
-
-/*
- * Atomically update sem->count.
- * This does the equivalent of the following:
- *
- *	old_count = sem->count;
- *	tmp = MAX(old_count, 0) + incr;
- *	sem->count = tmp;
- *	return old_count;
- */
-static inline int __sem_update_count(struct semaphore *sem, int incr)
-{
-	long old_count, tmp = 0;
-
-	__asm__ __volatile__(
-	"1:	ldl_l	%0,%2\n"
-	"	cmovgt	%0,%0,%1\n"
-	"	addl	%1,%3,%1\n"
-	"	stl_c	%1,%2\n"
-	"	beq	%1,2f\n"
-	"	mb\n"
-	".subsection 2\n"
-	"2:	br	1b\n"
-	".previous"
-	: "=&r" (old_count), "=&r" (tmp), "=m" (sem->count)
-	: "Ir" (incr), "1" (tmp), "m" (sem->count));
-
-	return old_count;
-}
-
-/*
- * Perform the "down" function.  Return zero for semaphore acquired,
- * return negative for signalled out of the function.
- *
- * If called from down, the return is ignored and the wait loop is
- * not interruptible.  This means that a task waiting on a semaphore
- * using "down()" cannot be killed until someone does an "up()" on
- * the semaphore.
- *
- * If called from down_interruptible, the return value gets checked
- * upon return.  If the return value is negative then the task continues
- * with the negative value in the return register (it can be tested by
- * the caller).
- *
- * Either form may be used in conjunction with "up()".
- */
-
-void __sched
-__down_failed(struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): down failed(%p)\n",
-	       tsk->comm, task_pid_nr(tsk), sem);
-#endif
-
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	wmb();
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	/*
-	 * Try to get the semaphore.  If the count is > 0, then we've
-	 * got the semaphore; we decrement count and exit the loop.
-	 * If the count is 0 or negative, we set it to -1, indicating
-	 * that we are asleep, and then sleep.
-	 */
-	while (__sem_update_count(sem, -1) <= 0) {
-		schedule();
-		set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	tsk->state = TASK_RUNNING;
-
-	/*
-	 * If there are any more sleepers, wake one of them up so
-	 * that it can either get the semaphore, or set count to -1
-	 * indicating that there are still processes sleeping.
-	 */
-	wake_up(&sem->wait);
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): down acquired(%p)\n",
-	       tsk->comm, task_pid_nr(tsk), sem);
-#endif
-}
-
-int __sched
-__down_failed_interruptible(struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	long ret = 0;
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): down failed(%p)\n",
-	       tsk->comm, task_pid_nr(tsk), sem);
-#endif
-
-	tsk->state = TASK_INTERRUPTIBLE;
-	wmb();
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	while (__sem_update_count(sem, -1) <= 0) {
-		if (signal_pending(current)) {
-			/*
-			 * A signal is pending - give up trying.
-			 * Set sem->count to 0 if it is negative,
-			 * since we are no longer sleeping.
-			 */
-			__sem_update_count(sem, 0);
-			ret = -EINTR;
-			break;
-		}
-		schedule();
-		set_task_state(tsk, TASK_INTERRUPTIBLE);
-	}
-
-	remove_wait_queue(&sem->wait, &wait);
-	tsk->state = TASK_RUNNING;
-	wake_up(&sem->wait);
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): down %s(%p)\n",
-	       current->comm, task_pid_nr(current),
-	       (ret < 0 ? "interrupted" : "acquired"), sem);
-#endif
-	return ret;
-}
-
-void
-__up_wakeup(struct semaphore *sem)
-{
-	/*
-	 * Note that we incremented count in up() before we came here,
-	 * but that was ineffective since the result was <= 0, and
-	 * any negative value of count is equivalent to 0.
-	 * This ends up setting count to 1, unless count is now > 0
-	 * (i.e. because some other cpu has called up() in the meantime),
-	 * in which case we just increment count.
-	 */
-	__sem_update_count(sem, 1);
-	wake_up(&sem->wait);
-}
-
-void __sched
-down(struct semaphore *sem)
-{
-#ifdef WAITQUEUE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): down(%p) <count=%d> from %p\n",
-	       current->comm, task_pid_nr(current), sem,
-	       atomic_read(&sem->count), __builtin_return_address(0));
-#endif
-	__down(sem);
-}
-
-int __sched
-down_interruptible(struct semaphore *sem)
-{
-#ifdef WAITQUEUE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): down(%p) <count=%d> from %p\n",
-	       current->comm, task_pid_nr(current), sem,
-	       atomic_read(&sem->count), __builtin_return_address(0));
-#endif
-	return __down_interruptible(sem);
-}
-
-int
-down_trylock(struct semaphore *sem)
-{
-	int ret;
-
-#ifdef WAITQUEUE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	ret = __down_trylock(sem);
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): down_trylock %s from %p\n",
-	       current->comm, task_pid_nr(current),
-	       ret ? "failed" : "acquired",
-	       __builtin_return_address(0));
-#endif
-
-	return ret;
-}
-
-void
-up(struct semaphore *sem)
-{
-#ifdef WAITQUEUE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	printk("%s(%d): up(%p) <count=%d> from %p\n",
-	       current->comm, task_pid_nr(current), sem,
-	       atomic_read(&sem->count), __builtin_return_address(0));
-#endif
-	__up(sem);
-}
diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile
index 00d44c6fbfe9..6235f72a14f0 100644
--- a/arch/arm/kernel/Makefile
+++ b/arch/arm/kernel/Makefile
@@ -7,7 +7,7 @@ AFLAGS_head.o := -DTEXT_OFFSET=$(TEXT_OFFSET)
 # Object file lists.
 
 obj-y		:= compat.o entry-armv.o entry-common.o irq.o \
-		   process.o ptrace.o semaphore.o setup.o signal.o \
+		   process.o ptrace.o setup.o signal.o \
 		   sys_arm.o stacktrace.o time.o traps.o
 
 obj-$(CONFIG_ISA_DMA_API)	+= dma.o
diff --git a/arch/arm/kernel/semaphore.c b/arch/arm/kernel/semaphore.c
deleted file mode 100644
index 981fe5c6ccbe..000000000000
--- a/arch/arm/kernel/semaphore.c
+++ /dev/null
@@ -1,221 +0,0 @@
-/*
- *  ARM semaphore implementation, taken from
- *
- *  i386 semaphore implementation.
- *
- *  (C) Copyright 1999 Linus Torvalds
- *
- *  Modified for ARM by Russell King
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-#include <linux/module.h>
-#include <linux/sched.h>
-#include <linux/errno.h>
-#include <linux/init.h>
-
-#include <asm/semaphore.h>
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to acquire the semaphore, while the "sleeping"
- * variable is a count of such acquires.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * "sleeping" and the contention routine ordering is
- * protected by the semaphore spinlock.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-
-/*
- * Logic:
- *  - only on a boundary condition do we need to care. When we go
- *    from a negative count to a non-negative, we wake people up.
- *  - when we go from a non-negative count to a negative do we
- *    (a) synchronize with the "sleeper" count and (b) make sure
- *    that we're on the wakeup list before we synchronize so that
- *    we cannot lose wakeup events.
- */
-
-void __up(struct semaphore *sem)
-{
-	wake_up(&sem->wait);
-}
-
-static DEFINE_SPINLOCK(semaphore_lock);
-
-void __sched __down(struct semaphore * sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	spin_lock_irq(&semaphore_lock);
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irq(&semaphore_lock);
-
-		schedule();
-		tsk->state = TASK_UNINTERRUPTIBLE;
-		spin_lock_irq(&semaphore_lock);
-	}
-	spin_unlock_irq(&semaphore_lock);
-	remove_wait_queue(&sem->wait, &wait);
-	tsk->state = TASK_RUNNING;
-	wake_up(&sem->wait);
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	tsk->state = TASK_INTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	spin_lock_irq(&semaphore_lock);
-	sem->sleepers ++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * With signals pending, this turns into
-		 * the trylock failure case - we won't be
-		 * sleeping, and we* can't get the lock as
-		 * it has contention. Just correct the count
-		 * and exit.
-		 */
-		if (signal_pending(current)) {
-			retval = -EINTR;
-			sem->sleepers = 0;
-			atomic_add(sleepers, &sem->count);
-			break;
-		}
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock. The
-		 * "-1" is because we're still hoping to get
-		 * the lock.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irq(&semaphore_lock);
-
-		schedule();
-		tsk->state = TASK_INTERRUPTIBLE;
-		spin_lock_irq(&semaphore_lock);
-	}
-	spin_unlock_irq(&semaphore_lock);
-	tsk->state = TASK_RUNNING;
-	remove_wait_queue(&sem->wait, &wait);
-	wake_up(&sem->wait);
-	return retval;
-}
-
-/*
- * Trylock failed - make sure we correct for
- * having decremented the count.
- *
- * We could have done the trylock with a
- * single "cmpxchg" without failure cases,
- * but then it wouldn't work on a 386.
- */
-int __down_trylock(struct semaphore * sem)
-{
-	int sleepers;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_lock, flags);
-	sleepers = sem->sleepers + 1;
-	sem->sleepers = 0;
-
-	/*
-	 * Add "everybody else" and us into it. They aren't
-	 * playing, because we own the spinlock.
-	 */
-	if (!atomic_add_negative(sleepers, &sem->count))
-		wake_up(&sem->wait);
-
-	spin_unlock_irqrestore(&semaphore_lock, flags);
-	return 1;
-}
-
-/*
- * The semaphore operations have a special calling sequence that
- * allow us to do a simpler in-line version of them. These routines
- * need to convert that sequence back into the C sequence when
- * there is contention on the semaphore.
- *
- * ip contains the semaphore pointer on entry. Save the C-clobbered
- * registers (r0 to r3 and lr), but not ip, as we use it as a return
- * value in some cases..
- * To remain AAPCS compliant (64-bit stack align) we save r4 as well.
- */
-asm("	.section .sched.text,\"ax\",%progbits	\n\
-	.align	5				\n\
-	.globl	__down_failed			\n\
-__down_failed:					\n\
-	stmfd	sp!, {r0 - r4, lr}		\n\
-	mov	r0, ip				\n\
-	bl	__down				\n\
-	ldmfd	sp!, {r0 - r4, pc}		\n\
-						\n\
-	.align	5				\n\
-	.globl	__down_interruptible_failed	\n\
-__down_interruptible_failed:			\n\
-	stmfd	sp!, {r0 - r4, lr}		\n\
-	mov	r0, ip				\n\
-	bl	__down_interruptible		\n\
-	mov	ip, r0				\n\
-	ldmfd	sp!, {r0 - r4, pc}		\n\
-						\n\
-	.align	5				\n\
-	.globl	__down_trylock_failed		\n\
-__down_trylock_failed:				\n\
-	stmfd	sp!, {r0 - r4, lr}		\n\
-	mov	r0, ip				\n\
-	bl	__down_trylock			\n\
-	mov	ip, r0				\n\
-	ldmfd	sp!, {r0 - r4, pc}		\n\
-						\n\
-	.align	5				\n\
-	.globl	__up_wakeup			\n\
-__up_wakeup:					\n\
-	stmfd	sp!, {r0 - r4, lr}		\n\
-	mov	r0, ip				\n\
-	bl	__up				\n\
-	ldmfd	sp!, {r0 - r4, pc}		\n\
-	");
-
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_interruptible_failed);
-EXPORT_SYMBOL(__down_trylock_failed);
-EXPORT_SYMBOL(__up_wakeup);
diff --git a/arch/avr32/kernel/Makefile b/arch/avr32/kernel/Makefile
index e4b6d122b033..18229d0d1861 100644
--- a/arch/avr32/kernel/Makefile
+++ b/arch/avr32/kernel/Makefile
@@ -6,7 +6,7 @@ extra-y				:= head.o vmlinux.lds
 
 obj-$(CONFIG_SUBARCH_AVR32B)	+= entry-avr32b.o
 obj-y				+= syscall_table.o syscall-stubs.o irq.o
-obj-y				+= setup.o traps.o semaphore.o ocd.o ptrace.o
+obj-y				+= setup.o traps.o ocd.o ptrace.o
 obj-y				+= signal.o sys_avr32.o process.o time.o
 obj-y				+= init_task.o switch_to.o cpu.o
 obj-$(CONFIG_MODULES)		+= module.o avr32_ksyms.o
diff --git a/arch/avr32/kernel/semaphore.c b/arch/avr32/kernel/semaphore.c
deleted file mode 100644
index 1e2705a05016..000000000000
--- a/arch/avr32/kernel/semaphore.c
+++ /dev/null
@@ -1,148 +0,0 @@
-/*
- * AVR32 sempahore implementation.
- *
- * Copyright (C) 2004-2006 Atmel Corporation
- *
- * Based on linux/arch/i386/kernel/semaphore.c
- *  Copyright (C) 1999 Linus Torvalds
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-
-#include <linux/sched.h>
-#include <linux/errno.h>
-#include <linux/module.h>
-
-#include <asm/semaphore.h>
-#include <asm/atomic.h>
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to acquire the semaphore, while the "sleeping"
- * variable is a count of such acquires.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * "sleeping" and the contention routine ordering is protected
- * by the spinlock in the semaphore's waitqueue head.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-
-/*
- * Logic:
- *  - only on a boundary condition do we need to care. When we go
- *    from a negative count to a non-negative, we wake people up.
- *  - when we go from a non-negative count to a negative do we
- *    (a) synchronize with the "sleeper" count and (b) make sure
- *    that we're on the wakeup list before we synchronize so that
- *    we cannot lose wakeup events.
- */
-
-void __up(struct semaphore *sem)
-{
-	wake_up(&sem->wait);
-}
-EXPORT_SYMBOL(__up);
-
-void __sched __down(struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-        DECLARE_WAITQUEUE(wait, tsk);
-        unsigned long flags;
-
-        tsk->state = TASK_UNINTERRUPTIBLE;
-        spin_lock_irqsave(&sem->wait.lock, flags);
-        add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-        sem->sleepers++;
-        for (;;) {
-                int sleepers = sem->sleepers;
-
-                /*
-                 * Add "everybody else" into it. They aren't
-                 * playing, because we own the spinlock in
-                 * the wait_queue_head.
-                 */
-                if (atomic_add_return(sleepers - 1, &sem->count) >= 0) {
-                        sem->sleepers = 0;
-                        break;
-                }
-                sem->sleepers = 1;      /* us - see -1 above */
-                spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-                schedule();
-
-                spin_lock_irqsave(&sem->wait.lock, flags);
-                tsk->state = TASK_UNINTERRUPTIBLE;
-        }
-        remove_wait_queue_locked(&sem->wait, &wait);
-        wake_up_locked(&sem->wait);
-        spin_unlock_irqrestore(&sem->wait.lock, flags);
-        tsk->state = TASK_RUNNING;
-}
-EXPORT_SYMBOL(__down);
-
-int __sched __down_interruptible(struct semaphore *sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-        DECLARE_WAITQUEUE(wait, tsk);
-        unsigned long flags;
-
-        tsk->state = TASK_INTERRUPTIBLE;
-        spin_lock_irqsave(&sem->wait.lock, flags);
-        add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-        sem->sleepers++;
-        for (;;) {
-                int sleepers = sem->sleepers;
-
-		/*
-		 * With signals pending, this turns into the trylock
-		 * failure case - we won't be sleeping, and we can't
-		 * get the lock as it has contention. Just correct the
-		 * count and exit.
-		 */
-		if (signal_pending(current)) {
-			retval = -EINTR;
-			sem->sleepers = 0;
-			atomic_add(sleepers, &sem->count);
-			break;
-		}
-
-                /*
-                 * Add "everybody else" into it. They aren't
-                 * playing, because we own the spinlock in
-                 * the wait_queue_head.
-                 */
-                if (atomic_add_return(sleepers - 1, &sem->count) >= 0) {
-                        sem->sleepers = 0;
-                        break;
-                }
-                sem->sleepers = 1;      /* us - see -1 above */
-                spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-                schedule();
-
-                spin_lock_irqsave(&sem->wait.lock, flags);
-                tsk->state = TASK_INTERRUPTIBLE;
-        }
-        remove_wait_queue_locked(&sem->wait, &wait);
-        wake_up_locked(&sem->wait);
-        spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-        tsk->state = TASK_RUNNING;
-	return retval;
-}
-EXPORT_SYMBOL(__down_interruptible);
diff --git a/arch/blackfin/Kconfig b/arch/blackfin/Kconfig
index 589c6aca4803..2dd1f300a5cf 100644
--- a/arch/blackfin/Kconfig
+++ b/arch/blackfin/Kconfig
@@ -31,10 +31,6 @@ config ZONE_DMA
 	bool
 	default y
 
-config SEMAPHORE_SLEEPERS
-	bool
-	default y
-
 config GENERIC_FIND_NEXT_BIT
 	bool
 	default y
diff --git a/arch/blackfin/kernel/bfin_ksyms.c b/arch/blackfin/kernel/bfin_ksyms.c
index 0bfbb269e350..053edff6c0d8 100644
--- a/arch/blackfin/kernel/bfin_ksyms.c
+++ b/arch/blackfin/kernel/bfin_ksyms.c
@@ -42,11 +42,6 @@ EXPORT_SYMBOL(ip_fast_csum);
 
 EXPORT_SYMBOL(kernel_thread);
 
-EXPORT_SYMBOL(__up);
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_trylock);
-EXPORT_SYMBOL(__down_interruptible);
-
 EXPORT_SYMBOL(is_in_rom);
 EXPORT_SYMBOL(bfin_return_from_exception);
 
diff --git a/arch/cris/kernel/Makefile b/arch/cris/kernel/Makefile
index c8e8ea570989..ee7bcd4d20b2 100644
--- a/arch/cris/kernel/Makefile
+++ b/arch/cris/kernel/Makefile
@@ -5,8 +5,7 @@
 
 extra-y	:= vmlinux.lds
 
-obj-y   := process.o traps.o irq.o ptrace.o setup.o \
-	   time.o sys_cris.o semaphore.o
+obj-y   := process.o traps.o irq.o ptrace.o setup.o time.o sys_cris.o
 
 obj-$(CONFIG_MODULES)    += crisksyms.o
 obj-$(CONFIG_MODULES)	 += module.o
diff --git a/arch/cris/kernel/crisksyms.c b/arch/cris/kernel/crisksyms.c
index 62f0e752915a..7ac000f6a888 100644
--- a/arch/cris/kernel/crisksyms.c
+++ b/arch/cris/kernel/crisksyms.c
@@ -9,7 +9,6 @@
 #include <linux/string.h>
 #include <linux/tty.h>
 
-#include <asm/semaphore.h>
 #include <asm/processor.h>
 #include <asm/uaccess.h>
 #include <asm/checksum.h>
@@ -49,12 +48,6 @@ EXPORT_SYMBOL(__negdi2);
 EXPORT_SYMBOL(__ioremap);
 EXPORT_SYMBOL(iounmap);
 
-/* Semaphore functions */
-EXPORT_SYMBOL(__up);
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_interruptible);
-EXPORT_SYMBOL(__down_trylock);
-
 /* Userspace access functions */
 EXPORT_SYMBOL(__copy_user_zeroing);
 EXPORT_SYMBOL(__copy_user);
diff --git a/arch/cris/kernel/semaphore.c b/arch/cris/kernel/semaphore.c
deleted file mode 100644
index f137a439041f..000000000000
--- a/arch/cris/kernel/semaphore.c
+++ /dev/null
@@ -1,129 +0,0 @@
-/*
- *  Generic semaphore code. Buyer beware. Do your own
- * specific changes in <asm/semaphore-helper.h>
- */
-
-#include <linux/sched.h>
-#include <asm/semaphore-helper.h>
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to sleep, while the "waking" variable is
- * incremented when the "up()" code goes to wake up waiting
- * processes.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * waking_non_zero() (from asm/semaphore.h) must execute
- * atomically.
- *
- * When __up() is called, the count was negative before
- * incrementing it, and we need to wake up somebody.
- *
- * This routine adds one to the count of processes that need to
- * wake up and exit.  ALL waiting processes actually wake up but
- * only the one that gets to the "waking" field first will gate
- * through and acquire the semaphore.  The others will go back
- * to sleep.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-void __up(struct semaphore *sem)
-{
-	wake_one_more(sem);
-	wake_up(&sem->wait);
-}
-
-/*
- * Perform the "down" function.  Return zero for semaphore acquired,
- * return negative for signalled out of the function.
- *
- * If called from __down, the return is ignored and the wait loop is
- * not interruptible.  This means that a task waiting on a semaphore
- * using "down()" cannot be killed until someone does an "up()" on
- * the semaphore.
- *
- * If called from __down_interruptible, the return value gets checked
- * upon return.  If the return value is negative then the task continues
- * with the negative value in the return register (it can be tested by
- * the caller).
- *
- * Either form may be used in conjunction with "up()".
- *
- */
-
-#define DOWN_VAR				\
-	struct task_struct *tsk = current;	\
-	wait_queue_t wait;			\
-	init_waitqueue_entry(&wait, tsk);
-
-#define DOWN_HEAD(task_state)						\
-									\
-									\
-	tsk->state = (task_state);					\
-	add_wait_queue(&sem->wait, &wait);				\
-									\
-	/*								\
-	 * Ok, we're set up.  sem->count is known to be less than zero	\
-	 * so we must wait.						\
-	 *								\
-	 * We can let go the lock for purposes of waiting.		\
-	 * We re-acquire it after awaking so as to protect		\
-	 * all semaphore operations.					\
-	 *								\
-	 * If "up()" is called before we call waking_non_zero() then	\
-	 * we will catch it right away.  If it is called later then	\
-	 * we will have to go through a wakeup cycle to catch it.	\
-	 *								\
-	 * Multiple waiters contend for the semaphore lock to see	\
-	 * who gets to gate through and who has to wait some more.	\
-	 */								\
-	for (;;) {
-
-#define DOWN_TAIL(task_state)			\
-		tsk->state = (task_state);	\
-	}					\
-	tsk->state = TASK_RUNNING;		\
-	remove_wait_queue(&sem->wait, &wait);
-
-void __sched __down(struct semaphore * sem)
-{
-	DOWN_VAR
-	DOWN_HEAD(TASK_UNINTERRUPTIBLE)
-	if (waking_non_zero(sem))
-		break;
-	schedule();
-	DOWN_TAIL(TASK_UNINTERRUPTIBLE)
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-	DOWN_VAR
-	DOWN_HEAD(TASK_INTERRUPTIBLE)
-
-	ret = waking_non_zero_interruptible(sem, tsk);
-	if (ret)
-	{
-		if (ret == 1)
-			/* ret != 0 only if we get interrupted -arca */
-			ret = 0;
-		break;
-	}
-	schedule();
-	DOWN_TAIL(TASK_INTERRUPTIBLE)
-	return ret;
-}
-
-int __down_trylock(struct semaphore * sem)
-{
-	return waking_non_zero_trylock(sem);
-}
diff --git a/arch/frv/kernel/Makefile b/arch/frv/kernel/Makefile
index e8f73ed28b52..c36f70b6699a 100644
--- a/arch/frv/kernel/Makefile
+++ b/arch/frv/kernel/Makefile
@@ -9,7 +9,7 @@ extra-y:= head.o init_task.o vmlinux.lds
 
 obj-y := $(heads-y) entry.o entry-table.o break.o switch_to.o kernel_thread.o \
 	 kernel_execve.o process.o traps.o ptrace.o signal.o dma.o \
-	 sys_frv.o time.o semaphore.o setup.o frv_ksyms.o \
+	 sys_frv.o time.o setup.o frv_ksyms.o \
 	 debug-stub.o irq.o sleep.o uaccess.o
 
 obj-$(CONFIG_GDBSTUB)		+= gdb-stub.o gdb-io.o
diff --git a/arch/frv/kernel/frv_ksyms.c b/arch/frv/kernel/frv_ksyms.c
index f772704b3d28..0316b3c50eff 100644
--- a/arch/frv/kernel/frv_ksyms.c
+++ b/arch/frv/kernel/frv_ksyms.c
@@ -12,7 +12,6 @@
 #include <asm/pgalloc.h>
 #include <asm/irq.h>
 #include <asm/io.h>
-#include <asm/semaphore.h>
 #include <asm/checksum.h>
 #include <asm/hardirq.h>
 #include <asm/cacheflush.h>
diff --git a/arch/frv/kernel/semaphore.c b/arch/frv/kernel/semaphore.c
deleted file mode 100644
index 7ee3a147b471..000000000000
--- a/arch/frv/kernel/semaphore.c
+++ /dev/null
@@ -1,155 +0,0 @@
-/* semaphore.c: FR-V semaphores
- *
- * Copyright (C) 2003 Red Hat, Inc. All Rights Reserved.
- * Written by David Howells (dhowells@redhat.com)
- * - Derived from lib/rwsem-spinlock.c
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version
- * 2 of the License, or (at your option) any later version.
- */
-
-#include <linux/sched.h>
-#include <linux/module.h>
-#include <asm/semaphore.h>
-
-struct sem_waiter {
-	struct list_head	list;
-	struct task_struct	*task;
-};
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-void semtrace(struct semaphore *sem, const char *str)
-{
-	if (sem->debug)
-		printk("[%d] %s({%d,%d})\n",
-		       current->pid,
-		       str,
-		       sem->counter,
-		       list_empty(&sem->wait_list) ? 0 : 1);
-}
-#else
-#define semtrace(SEM,STR) do { } while(0)
-#endif
-
-/*
- * wait for a token to be granted from a semaphore
- * - entered with lock held and interrupts disabled
- */
-void __down(struct semaphore *sem, unsigned long flags)
-{
-	struct task_struct *tsk = current;
-	struct sem_waiter waiter;
-
-	semtrace(sem, "Entering __down");
-
-	/* set up my own style of waitqueue */
-	waiter.task = tsk;
-	get_task_struct(tsk);
-
-	list_add_tail(&waiter.list, &sem->wait_list);
-
-	/* we don't need to touch the semaphore struct anymore */
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-
-	/* wait to be given the semaphore */
-	set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-
-	for (;;) {
-		if (list_empty(&waiter.list))
-			break;
-		schedule();
-		set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	}
-
-	tsk->state = TASK_RUNNING;
-	semtrace(sem, "Leaving __down");
-}
-
-EXPORT_SYMBOL(__down);
-
-/*
- * interruptibly wait for a token to be granted from a semaphore
- * - entered with lock held and interrupts disabled
- */
-int __down_interruptible(struct semaphore *sem, unsigned long flags)
-{
-	struct task_struct *tsk = current;
-	struct sem_waiter waiter;
-	int ret;
-
-	semtrace(sem,"Entering __down_interruptible");
-
-	/* set up my own style of waitqueue */
-	waiter.task = tsk;
-	get_task_struct(tsk);
-
-	list_add_tail(&waiter.list, &sem->wait_list);
-
-	/* we don't need to touch the semaphore struct anymore */
-	set_task_state(tsk, TASK_INTERRUPTIBLE);
-
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-
-	/* wait to be given the semaphore */
-	ret = 0;
-	for (;;) {
-		if (list_empty(&waiter.list))
-			break;
-		if (unlikely(signal_pending(current)))
-			goto interrupted;
-		schedule();
-		set_task_state(tsk, TASK_INTERRUPTIBLE);
-	}
-
- out:
-	tsk->state = TASK_RUNNING;
-	semtrace(sem, "Leaving __down_interruptible");
-	return ret;
-
- interrupted:
-	spin_lock_irqsave(&sem->wait_lock, flags);
-
-	if (!list_empty(&waiter.list)) {
-		list_del(&waiter.list);
-		ret = -EINTR;
-	}
-
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-	if (ret == -EINTR)
-		put_task_struct(current);
-	goto out;
-}
-
-EXPORT_SYMBOL(__down_interruptible);
-
-/*
- * release a single token back to a semaphore
- * - entered with lock held and interrupts disabled
- */
-void __up(struct semaphore *sem)
-{
-	struct task_struct *tsk;
-	struct sem_waiter *waiter;
-
-	semtrace(sem,"Entering __up");
-
-	/* grant the token to the process at the front of the queue */
-	waiter = list_entry(sem->wait_list.next, struct sem_waiter, list);
-
-	/* We must be careful not to touch 'waiter' after we set ->task = NULL.
-	 * It is allocated on the waiter's stack and may become invalid at
-	 * any time after that point (due to a wakeup from another source).
-	 */
-	list_del_init(&waiter->list);
-	tsk = waiter->task;
-	mb();
-	waiter->task = NULL;
-	wake_up_process(tsk);
-	put_task_struct(tsk);
-
-	semtrace(sem,"Leaving __up");
-}
-
-EXPORT_SYMBOL(__up);
diff --git a/arch/h8300/kernel/Makefile b/arch/h8300/kernel/Makefile
index 874f6aefee65..6c248c3c5c3b 100644
--- a/arch/h8300/kernel/Makefile
+++ b/arch/h8300/kernel/Makefile
@@ -5,7 +5,7 @@
 extra-y := vmlinux.lds
 
 obj-y := process.o traps.o ptrace.o irq.o \
-	 sys_h8300.o time.o semaphore.o signal.o \
+	 sys_h8300.o time.o signal.o \
          setup.o gpio.o init_task.o syscalls.o \
 	 entry.o
 
diff --git a/arch/h8300/kernel/h8300_ksyms.c b/arch/h8300/kernel/h8300_ksyms.c
index d1b15267ac81..6866bd9c7fb4 100644
--- a/arch/h8300/kernel/h8300_ksyms.c
+++ b/arch/h8300/kernel/h8300_ksyms.c
@@ -12,7 +12,6 @@
 #include <asm/pgalloc.h>
 #include <asm/irq.h>
 #include <asm/io.h>
-#include <asm/semaphore.h>
 #include <asm/checksum.h>
 #include <asm/current.h>
 #include <asm/gpio.h>
diff --git a/arch/h8300/kernel/semaphore.c b/arch/h8300/kernel/semaphore.c
deleted file mode 100644
index d12cbbfe6ebd..000000000000
--- a/arch/h8300/kernel/semaphore.c
+++ /dev/null
@@ -1,132 +0,0 @@
-/*
- *  Generic semaphore code. Buyer beware. Do your own
- * specific changes in <asm/semaphore-helper.h>
- */
-
-#include <linux/sched.h>
-#include <linux/init.h>
-#include <asm/semaphore-helper.h>
-
-#ifndef CONFIG_RMW_INSNS
-spinlock_t semaphore_wake_lock;
-#endif
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to sleep, while the "waking" variable is
- * incremented when the "up()" code goes to wake up waiting
- * processes.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * waking_non_zero() (from asm/semaphore.h) must execute
- * atomically.
- *
- * When __up() is called, the count was negative before
- * incrementing it, and we need to wake up somebody.
- *
- * This routine adds one to the count of processes that need to
- * wake up and exit.  ALL waiting processes actually wake up but
- * only the one that gets to the "waking" field first will gate
- * through and acquire the semaphore.  The others will go back
- * to sleep.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-void __up(struct semaphore *sem)
-{
-	wake_one_more(sem);
-	wake_up(&sem->wait);
-}
-
-/*
- * Perform the "down" function.  Return zero for semaphore acquired,
- * return negative for signalled out of the function.
- *
- * If called from __down, the return is ignored and the wait loop is
- * not interruptible.  This means that a task waiting on a semaphore
- * using "down()" cannot be killed until someone does an "up()" on
- * the semaphore.
- *
- * If called from __down_interruptible, the return value gets checked
- * upon return.  If the return value is negative then the task continues
- * with the negative value in the return register (it can be tested by
- * the caller).
- *
- * Either form may be used in conjunction with "up()".
- *
- */
-
-
-#define DOWN_HEAD(task_state)						\
-									\
-									\
-	current->state = (task_state);					\
-	add_wait_queue(&sem->wait, &wait);				\
-									\
-	/*								\
-	 * Ok, we're set up.  sem->count is known to be less than zero	\
-	 * so we must wait.						\
-	 *								\
-	 * We can let go the lock for purposes of waiting.		\
-	 * We re-acquire it after awaking so as to protect		\
-	 * all semaphore operations.					\
-	 *								\
-	 * If "up()" is called before we call waking_non_zero() then	\
-	 * we will catch it right away.  If it is called later then	\
-	 * we will have to go through a wakeup cycle to catch it.	\
-	 *								\
-	 * Multiple waiters contend for the semaphore lock to see	\
-	 * who gets to gate through and who has to wait some more.	\
-	 */								\
-	for (;;) {
-
-#define DOWN_TAIL(task_state)			\
-		current->state = (task_state);	\
-	}					\
-	current->state = TASK_RUNNING;		\
-	remove_wait_queue(&sem->wait, &wait);
-
-void __sched __down(struct semaphore * sem)
-{
-	DECLARE_WAITQUEUE(wait, current);
-
-	DOWN_HEAD(TASK_UNINTERRUPTIBLE)
-	if (waking_non_zero(sem))
-		break;
-	schedule();
-	DOWN_TAIL(TASK_UNINTERRUPTIBLE)
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	DECLARE_WAITQUEUE(wait, current);
-	int ret = 0;
-
-	DOWN_HEAD(TASK_INTERRUPTIBLE)
-
-	ret = waking_non_zero_interruptible(sem, current);
-	if (ret)
-	{
-		if (ret == 1)
-			/* ret != 0 only if we get interrupted -arca */
-			ret = 0;
-		break;
-	}
-	schedule();
-	DOWN_TAIL(TASK_INTERRUPTIBLE)
-	return ret;
-}
-
-int __down_trylock(struct semaphore * sem)
-{
-	return waking_non_zero_trylock(sem);
-}
diff --git a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile
index 33e5a598672d..13fd10e8699e 100644
--- a/arch/ia64/kernel/Makefile
+++ b/arch/ia64/kernel/Makefile
@@ -6,7 +6,7 @@ extra-y	:= head.o init_task.o vmlinux.lds
 
 obj-y := acpi.o entry.o efi.o efi_stub.o gate-data.o fsys.o ia64_ksyms.o irq.o irq_ia64.o	\
 	 irq_lsapic.o ivt.o machvec.o pal.o patch.o process.o perfmon.o ptrace.o sal.o		\
-	 salinfo.o semaphore.o setup.o signal.o sys_ia64.o time.o traps.o unaligned.o \
+	 salinfo.o setup.o signal.o sys_ia64.o time.o traps.o unaligned.o \
 	 unwind.o mca.o mca_asm.o topology.o
 
 obj-$(CONFIG_IA64_BRL_EMU)	+= brl_emu.o
diff --git a/arch/ia64/kernel/ia64_ksyms.c b/arch/ia64/kernel/ia64_ksyms.c
index 8e7193d55528..6da1f20d7372 100644
--- a/arch/ia64/kernel/ia64_ksyms.c
+++ b/arch/ia64/kernel/ia64_ksyms.c
@@ -19,12 +19,6 @@ EXPORT_SYMBOL_GPL(empty_zero_page);
 EXPORT_SYMBOL(ip_fast_csum);		/* hand-coded assembly */
 EXPORT_SYMBOL(csum_ipv6_magic);
 
-#include <asm/semaphore.h>
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_interruptible);
-EXPORT_SYMBOL(__down_trylock);
-EXPORT_SYMBOL(__up);
-
 #include <asm/page.h>
 EXPORT_SYMBOL(clear_page);
 
diff --git a/arch/ia64/kernel/semaphore.c b/arch/ia64/kernel/semaphore.c
deleted file mode 100644
index 2724ef3fbae2..000000000000
--- a/arch/ia64/kernel/semaphore.c
+++ /dev/null
@@ -1,165 +0,0 @@
-/*
- * IA-64 semaphore implementation (derived from x86 version).
- *
- * Copyright (C) 1999-2000, 2002 Hewlett-Packard Co
- *	David Mosberger-Tang <davidm@hpl.hp.com>
- */
-
-/*
- * Semaphores are implemented using a two-way counter: The "count"
- * variable is decremented for each process that tries to acquire the
- * semaphore, while the "sleepers" variable is a count of such
- * acquires.
- *
- * Notably, the inline "up()" and "down()" functions can efficiently
- * test if they need to do any extra work (up needs to do something
- * only if count was negative before the increment operation.
- *
- * "sleeping" and the contention routine ordering is protected
- * by the spinlock in the semaphore's waitqueue head.
- *
- * Note that these functions are only called when there is contention
- * on the lock, and as such all this is the "non-critical" part of the
- * whole semaphore business. The critical part is the inline stuff in
- * <asm/semaphore.h> where we want to avoid any extra jumps and calls.
- */
-#include <linux/sched.h>
-#include <linux/init.h>
-
-#include <asm/errno.h>
-#include <asm/semaphore.h>
-
-/*
- * Logic:
- *  - Only on a boundary condition do we need to care. When we go
- *    from a negative count to a non-negative, we wake people up.
- *  - When we go from a non-negative count to a negative do we
- *    (a) synchronize with the "sleepers" count and (b) make sure
- *    that we're on the wakeup list before we synchronize so that
- *    we cannot lose wakeup events.
- */
-
-void
-__up (struct semaphore *sem)
-{
-	wake_up(&sem->wait);
-}
-
-void __sched __down (struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	unsigned long flags;
-
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock in
-		 * the wait_queue_head.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-		schedule();
-
-		spin_lock_irqsave(&sem->wait.lock, flags);
-		tsk->state = TASK_UNINTERRUPTIBLE;
-	}
-	remove_wait_queue_locked(&sem->wait, &wait);
-	wake_up_locked(&sem->wait);
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-	tsk->state = TASK_RUNNING;
-}
-
-int __sched __down_interruptible (struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	unsigned long flags;
-
-	tsk->state = TASK_INTERRUPTIBLE;
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-	sem->sleepers ++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * With signals pending, this turns into
-		 * the trylock failure case - we won't be
-		 * sleeping, and we* can't get the lock as
-		 * it has contention. Just correct the count
-		 * and exit.
-		 */
-		if (signal_pending(current)) {
-			retval = -EINTR;
-			sem->sleepers = 0;
-			atomic_add(sleepers, &sem->count);
-			break;
-		}
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock in
-		 * wait_queue_head. The "-1" is because we're
-		 * still hoping to get the semaphore.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-		schedule();
-
-		spin_lock_irqsave(&sem->wait.lock, flags);
-		tsk->state = TASK_INTERRUPTIBLE;
-	}
-	remove_wait_queue_locked(&sem->wait, &wait);
-	wake_up_locked(&sem->wait);
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-	tsk->state = TASK_RUNNING;
-	return retval;
-}
-
-/*
- * Trylock failed - make sure we correct for having decremented the
- * count.
- */
-int
-__down_trylock (struct semaphore *sem)
-{
-	unsigned long flags;
-	int sleepers;
-
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	sleepers = sem->sleepers + 1;
-	sem->sleepers = 0;
-
-	/*
-	 * Add "everybody else" and us into it. They aren't
-	 * playing, because we own the spinlock in the
-	 * wait_queue_head.
-	 */
-	if (!atomic_add_negative(sleepers, &sem->count)) {
-		wake_up_locked(&sem->wait);
-	}
-
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-	return 1;
-}
diff --git a/arch/m32r/kernel/Makefile b/arch/m32r/kernel/Makefile
index e97e26e87c9e..09200d4886e3 100644
--- a/arch/m32r/kernel/Makefile
+++ b/arch/m32r/kernel/Makefile
@@ -5,7 +5,7 @@
 extra-y	:= head.o init_task.o vmlinux.lds
 
 obj-y	:= process.o entry.o traps.o align.o irq.o setup.o time.o \
-	m32r_ksyms.o sys_m32r.o semaphore.o signal.o ptrace.o
+	m32r_ksyms.o sys_m32r.o signal.o ptrace.o
 
 obj-$(CONFIG_SMP)		+= smp.o smpboot.o
 obj-$(CONFIG_MODULES)		+= module.o
diff --git a/arch/m32r/kernel/m32r_ksyms.c b/arch/m32r/kernel/m32r_ksyms.c
index 41a4c95e06d6..e6709fe950ba 100644
--- a/arch/m32r/kernel/m32r_ksyms.c
+++ b/arch/m32r/kernel/m32r_ksyms.c
@@ -7,7 +7,6 @@
 #include <linux/interrupt.h>
 #include <linux/string.h>
 
-#include <asm/semaphore.h>
 #include <asm/processor.h>
 #include <asm/uaccess.h>
 #include <asm/checksum.h>
@@ -22,10 +21,6 @@ EXPORT_SYMBOL(dump_fpu);
 EXPORT_SYMBOL(__ioremap);
 EXPORT_SYMBOL(iounmap);
 EXPORT_SYMBOL(kernel_thread);
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_interruptible);
-EXPORT_SYMBOL(__up);
-EXPORT_SYMBOL(__down_trylock);
 
 /* Networking helper routines. */
 /* Delay loops */
diff --git a/arch/m32r/kernel/semaphore.c b/arch/m32r/kernel/semaphore.c
deleted file mode 100644
index 940c2d37cfd1..000000000000
--- a/arch/m32r/kernel/semaphore.c
+++ /dev/null
@@ -1,185 +0,0 @@
-/*
- *  linux/arch/m32r/semaphore.c
- *    orig : i386 2.6.4
- *
- *  M32R semaphore implementation.
- *
- *	Copyright (c) 2002 - 2004 Hitoshi Yamamoto
- */
-
-/*
- * i386 semaphore implementation.
- *
- * (C) Copyright 1999 Linus Torvalds
- *
- * Portions Copyright 1999 Red Hat, Inc.
- *
- *	This program is free software; you can redistribute it and/or
- *	modify it under the terms of the GNU General Public License
- *	as published by the Free Software Foundation; either version
- *	2 of the License, or (at your option) any later version.
- *
- * rw semaphores implemented November 1999 by Benjamin LaHaise <bcrl@kvack.org>
- */
-#include <linux/sched.h>
-#include <linux/err.h>
-#include <linux/init.h>
-#include <asm/semaphore.h>
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to acquire the semaphore, while the "sleeping"
- * variable is a count of such acquires.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * "sleeping" and the contention routine ordering is protected
- * by the spinlock in the semaphore's waitqueue head.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-
-/*
- * Logic:
- *  - only on a boundary condition do we need to care. When we go
- *    from a negative count to a non-negative, we wake people up.
- *  - when we go from a non-negative count to a negative do we
- *    (a) synchronize with the "sleeper" count and (b) make sure
- *    that we're on the wakeup list before we synchronize so that
- *    we cannot lose wakeup events.
- */
-
-asmlinkage void __up(struct semaphore *sem)
-{
-	wake_up(&sem->wait);
-}
-
-asmlinkage void __sched __down(struct semaphore * sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	unsigned long flags;
-
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock in
-		 * the wait_queue_head.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-		schedule();
-
-		spin_lock_irqsave(&sem->wait.lock, flags);
-		tsk->state = TASK_UNINTERRUPTIBLE;
-	}
-	remove_wait_queue_locked(&sem->wait, &wait);
-	wake_up_locked(&sem->wait);
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-	tsk->state = TASK_RUNNING;
-}
-
-asmlinkage int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	unsigned long flags;
-
-	tsk->state = TASK_INTERRUPTIBLE;
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * With signals pending, this turns into
-		 * the trylock failure case - we won't be
-		 * sleeping, and we* can't get the lock as
-		 * it has contention. Just correct the count
-		 * and exit.
-		 */
-		if (signal_pending(current)) {
-			retval = -EINTR;
-			sem->sleepers = 0;
-			atomic_add(sleepers, &sem->count);
-			break;
-		}
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock in
-		 * wait_queue_head. The "-1" is because we're
-		 * still hoping to get the semaphore.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-		schedule();
-
-		spin_lock_irqsave(&sem->wait.lock, flags);
-		tsk->state = TASK_INTERRUPTIBLE;
-	}
-	remove_wait_queue_locked(&sem->wait, &wait);
-	wake_up_locked(&sem->wait);
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-	tsk->state = TASK_RUNNING;
-	return retval;
-}
-
-/*
- * Trylock failed - make sure we correct for
- * having decremented the count.
- *
- * We could have done the trylock with a
- * single "cmpxchg" without failure cases,
- * but then it wouldn't work on a 386.
- */
-asmlinkage int __down_trylock(struct semaphore * sem)
-{
-	int sleepers;
-	unsigned long flags;
-
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	sleepers = sem->sleepers + 1;
-	sem->sleepers = 0;
-
-	/*
-	 * Add "everybody else" and us into it. They aren't
-	 * playing, because we own the spinlock in the
-	 * wait_queue_head.
-	 */
-	if (!atomic_add_negative(sleepers, &sem->count)) {
-		wake_up_locked(&sem->wait);
-	}
-
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-	return 1;
-}
diff --git a/arch/m68k/kernel/Makefile b/arch/m68k/kernel/Makefile
index a806208c7fb5..7a62a718143b 100644
--- a/arch/m68k/kernel/Makefile
+++ b/arch/m68k/kernel/Makefile
@@ -10,7 +10,7 @@ endif
 extra-y	+= vmlinux.lds
 
 obj-y	:= entry.o process.o traps.o ints.o signal.o ptrace.o module.o \
-	   sys_m68k.o time.o semaphore.o setup.o m68k_ksyms.o devres.o
+	   sys_m68k.o time.o setup.o m68k_ksyms.o devres.o
 
 devres-y = ../../../kernel/irq/devres.o
 
diff --git a/arch/m68k/kernel/m68k_ksyms.c b/arch/m68k/kernel/m68k_ksyms.c
index 6fc69c74fe2e..d900e77e5363 100644
--- a/arch/m68k/kernel/m68k_ksyms.c
+++ b/arch/m68k/kernel/m68k_ksyms.c
@@ -1,5 +1,4 @@
 #include <linux/module.h>
-#include <asm/semaphore.h>
 
 asmlinkage long long __ashldi3 (long long, int);
 asmlinkage long long __ashrdi3 (long long, int);
@@ -15,8 +14,3 @@ EXPORT_SYMBOL(__ashrdi3);
 EXPORT_SYMBOL(__lshrdi3);
 EXPORT_SYMBOL(__muldi3);
 
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_failed_interruptible);
-EXPORT_SYMBOL(__down_failed_trylock);
-EXPORT_SYMBOL(__up_wakeup);
-
diff --git a/arch/m68k/kernel/semaphore.c b/arch/m68k/kernel/semaphore.c
deleted file mode 100644
index d12cbbfe6ebd..000000000000
--- a/arch/m68k/kernel/semaphore.c
+++ /dev/null
@@ -1,132 +0,0 @@
-/*
- *  Generic semaphore code. Buyer beware. Do your own
- * specific changes in <asm/semaphore-helper.h>
- */
-
-#include <linux/sched.h>
-#include <linux/init.h>
-#include <asm/semaphore-helper.h>
-
-#ifndef CONFIG_RMW_INSNS
-spinlock_t semaphore_wake_lock;
-#endif
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to sleep, while the "waking" variable is
- * incremented when the "up()" code goes to wake up waiting
- * processes.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * waking_non_zero() (from asm/semaphore.h) must execute
- * atomically.
- *
- * When __up() is called, the count was negative before
- * incrementing it, and we need to wake up somebody.
- *
- * This routine adds one to the count of processes that need to
- * wake up and exit.  ALL waiting processes actually wake up but
- * only the one that gets to the "waking" field first will gate
- * through and acquire the semaphore.  The others will go back
- * to sleep.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-void __up(struct semaphore *sem)
-{
-	wake_one_more(sem);
-	wake_up(&sem->wait);
-}
-
-/*
- * Perform the "down" function.  Return zero for semaphore acquired,
- * return negative for signalled out of the function.
- *
- * If called from __down, the return is ignored and the wait loop is
- * not interruptible.  This means that a task waiting on a semaphore
- * using "down()" cannot be killed until someone does an "up()" on
- * the semaphore.
- *
- * If called from __down_interruptible, the return value gets checked
- * upon return.  If the return value is negative then the task continues
- * with the negative value in the return register (it can be tested by
- * the caller).
- *
- * Either form may be used in conjunction with "up()".
- *
- */
-
-
-#define DOWN_HEAD(task_state)						\
-									\
-									\
-	current->state = (task_state);					\
-	add_wait_queue(&sem->wait, &wait);				\
-									\
-	/*								\
-	 * Ok, we're set up.  sem->count is known to be less than zero	\
-	 * so we must wait.						\
-	 *								\
-	 * We can let go the lock for purposes of waiting.		\
-	 * We re-acquire it after awaking so as to protect		\
-	 * all semaphore operations.					\
-	 *								\
-	 * If "up()" is called before we call waking_non_zero() then	\
-	 * we will catch it right away.  If it is called later then	\
-	 * we will have to go through a wakeup cycle to catch it.	\
-	 *								\
-	 * Multiple waiters contend for the semaphore lock to see	\
-	 * who gets to gate through and who has to wait some more.	\
-	 */								\
-	for (;;) {
-
-#define DOWN_TAIL(task_state)			\
-		current->state = (task_state);	\
-	}					\
-	current->state = TASK_RUNNING;		\
-	remove_wait_queue(&sem->wait, &wait);
-
-void __sched __down(struct semaphore * sem)
-{
-	DECLARE_WAITQUEUE(wait, current);
-
-	DOWN_HEAD(TASK_UNINTERRUPTIBLE)
-	if (waking_non_zero(sem))
-		break;
-	schedule();
-	DOWN_TAIL(TASK_UNINTERRUPTIBLE)
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	DECLARE_WAITQUEUE(wait, current);
-	int ret = 0;
-
-	DOWN_HEAD(TASK_INTERRUPTIBLE)
-
-	ret = waking_non_zero_interruptible(sem, current);
-	if (ret)
-	{
-		if (ret == 1)
-			/* ret != 0 only if we get interrupted -arca */
-			ret = 0;
-		break;
-	}
-	schedule();
-	DOWN_TAIL(TASK_INTERRUPTIBLE)
-	return ret;
-}
-
-int __down_trylock(struct semaphore * sem)
-{
-	return waking_non_zero_trylock(sem);
-}
diff --git a/arch/m68k/lib/Makefile b/arch/m68k/lib/Makefile
index 6bbf19f96007..a18af095cd7c 100644
--- a/arch/m68k/lib/Makefile
+++ b/arch/m68k/lib/Makefile
@@ -5,4 +5,4 @@
 EXTRA_AFLAGS := -traditional
 
 lib-y	:= ashldi3.o ashrdi3.o lshrdi3.o muldi3.o \
-	   checksum.o string.o semaphore.o uaccess.o
+	   checksum.o string.o uaccess.o
diff --git a/arch/m68k/lib/semaphore.S b/arch/m68k/lib/semaphore.S
deleted file mode 100644
index 0215624c1602..000000000000
--- a/arch/m68k/lib/semaphore.S
+++ /dev/null
@@ -1,53 +0,0 @@
-/*
- *  linux/arch/m68k/lib/semaphore.S
- *
- *  Copyright (C) 1996  Linus Torvalds
- *
- *  m68k version by Andreas Schwab
- */
-
-#include <linux/linkage.h>
-#include <asm/semaphore.h>
-
-/*
- * The semaphore operations have a special calling sequence that
- * allow us to do a simpler in-line version of them. These routines
- * need to convert that sequence back into the C sequence when
- * there is contention on the semaphore.
- */
-ENTRY(__down_failed)
-	moveml %a0/%d0/%d1,-(%sp)
-	movel %a1,-(%sp)
-	jbsr __down
-	movel (%sp)+,%a1
-	moveml (%sp)+,%a0/%d0/%d1
-	rts
-
-ENTRY(__down_failed_interruptible)
-	movel %a0,-(%sp)
-	movel %d1,-(%sp)
-	movel %a1,-(%sp)
-	jbsr __down_interruptible
-	movel (%sp)+,%a1
-	movel (%sp)+,%d1
-	movel (%sp)+,%a0
-	rts
-
-ENTRY(__down_failed_trylock)
-	movel %a0,-(%sp)
-	movel %d1,-(%sp)
-	movel %a1,-(%sp)
-	jbsr __down_trylock
-	movel (%sp)+,%a1
-	movel (%sp)+,%d1
-	movel (%sp)+,%a0
-	rts
-
-ENTRY(__up_wakeup)
-	moveml %a0/%d0/%d1,-(%sp)
-	movel %a1,-(%sp)
-	jbsr __up
-	movel (%sp)+,%a1
-	moveml (%sp)+,%a0/%d0/%d1
-	rts
-
diff --git a/arch/m68knommu/kernel/Makefile b/arch/m68knommu/kernel/Makefile
index 1524b39ad63f..f0eab3dedb5a 100644
--- a/arch/m68knommu/kernel/Makefile
+++ b/arch/m68knommu/kernel/Makefile
@@ -5,7 +5,7 @@
 extra-y := vmlinux.lds
 
 obj-y += dma.o entry.o init_task.o irq.o m68k_ksyms.o process.o ptrace.o \
-	 semaphore.o setup.o signal.o syscalltable.o sys_m68k.o time.o traps.o
+	 setup.o signal.o syscalltable.o sys_m68k.o time.o traps.o
 
 obj-$(CONFIG_MODULES)	+= module.o
 obj-$(CONFIG_COMEMPCI)	+= comempci.o
diff --git a/arch/m68knommu/kernel/m68k_ksyms.c b/arch/m68knommu/kernel/m68k_ksyms.c
index 53fad1490282..39fe0a7aec32 100644
--- a/arch/m68knommu/kernel/m68k_ksyms.c
+++ b/arch/m68knommu/kernel/m68k_ksyms.c
@@ -13,7 +13,6 @@
 #include <asm/pgalloc.h>
 #include <asm/irq.h>
 #include <asm/io.h>
-#include <asm/semaphore.h>
 #include <asm/checksum.h>
 #include <asm/current.h>
 
@@ -39,11 +38,6 @@ EXPORT_SYMBOL(csum_partial_copy_nocheck);
 EXPORT_SYMBOL(memcpy);
 EXPORT_SYMBOL(memset);
 
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_failed_interruptible);
-EXPORT_SYMBOL(__down_failed_trylock);
-EXPORT_SYMBOL(__up_wakeup);
-
 /*
  * libgcc functions - functions that are used internally by the
  * compiler...  (prototypes are not correct though, but that
diff --git a/arch/m68knommu/kernel/semaphore.c b/arch/m68knommu/kernel/semaphore.c
deleted file mode 100644
index bce2bc7d87c6..000000000000
--- a/arch/m68knommu/kernel/semaphore.c
+++ /dev/null
@@ -1,133 +0,0 @@
-/*
- *  Generic semaphore code. Buyer beware. Do your own
- * specific changes in <asm/semaphore-helper.h>
- */
-
-#include <linux/sched.h>
-#include <linux/err.h>
-#include <linux/init.h>
-#include <asm/semaphore-helper.h>
-
-#ifndef CONFIG_RMW_INSNS
-spinlock_t semaphore_wake_lock;
-#endif
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to sleep, while the "waking" variable is
- * incremented when the "up()" code goes to wake up waiting
- * processes.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * waking_non_zero() (from asm/semaphore.h) must execute
- * atomically.
- *
- * When __up() is called, the count was negative before
- * incrementing it, and we need to wake up somebody.
- *
- * This routine adds one to the count of processes that need to
- * wake up and exit.  ALL waiting processes actually wake up but
- * only the one that gets to the "waking" field first will gate
- * through and acquire the semaphore.  The others will go back
- * to sleep.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-void __up(struct semaphore *sem)
-{
-	wake_one_more(sem);
-	wake_up(&sem->wait);
-}
-
-/*
- * Perform the "down" function.  Return zero for semaphore acquired,
- * return negative for signalled out of the function.
- *
- * If called from __down, the return is ignored and the wait loop is
- * not interruptible.  This means that a task waiting on a semaphore
- * using "down()" cannot be killed until someone does an "up()" on
- * the semaphore.
- *
- * If called from __down_interruptible, the return value gets checked
- * upon return.  If the return value is negative then the task continues
- * with the negative value in the return register (it can be tested by
- * the caller).
- *
- * Either form may be used in conjunction with "up()".
- *
- */
-
-
-#define DOWN_HEAD(task_state)						\
-									\
-									\
-	current->state = (task_state);					\
-	add_wait_queue(&sem->wait, &wait);				\
-									\
-	/*								\
-	 * Ok, we're set up.  sem->count is known to be less than zero	\
-	 * so we must wait.						\
-	 *								\
-	 * We can let go the lock for purposes of waiting.		\
-	 * We re-acquire it after awaking so as to protect		\
-	 * all semaphore operations.					\
-	 *								\
-	 * If "up()" is called before we call waking_non_zero() then	\
-	 * we will catch it right away.  If it is called later then	\
-	 * we will have to go through a wakeup cycle to catch it.	\
-	 *								\
-	 * Multiple waiters contend for the semaphore lock to see	\
-	 * who gets to gate through and who has to wait some more.	\
-	 */								\
-	for (;;) {
-
-#define DOWN_TAIL(task_state)			\
-		current->state = (task_state);	\
-	}					\
-	current->state = TASK_RUNNING;		\
-	remove_wait_queue(&sem->wait, &wait);
-
-void __sched __down(struct semaphore * sem)
-{
-	DECLARE_WAITQUEUE(wait, current);
-
-	DOWN_HEAD(TASK_UNINTERRUPTIBLE)
-	if (waking_non_zero(sem))
-		break;
-	schedule();
-	DOWN_TAIL(TASK_UNINTERRUPTIBLE)
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	DECLARE_WAITQUEUE(wait, current);
-	int ret = 0;
-
-	DOWN_HEAD(TASK_INTERRUPTIBLE)
-
-	ret = waking_non_zero_interruptible(sem, current);
-	if (ret)
-	{
-		if (ret == 1)
-			/* ret != 0 only if we get interrupted -arca */
-			ret = 0;
-		break;
-	}
-	schedule();
-	DOWN_TAIL(TASK_INTERRUPTIBLE)
-	return ret;
-}
-
-int __down_trylock(struct semaphore * sem)
-{
-	return waking_non_zero_trylock(sem);
-}
diff --git a/arch/m68knommu/lib/Makefile b/arch/m68knommu/lib/Makefile
index e051a7913987..d94d709665aa 100644
--- a/arch/m68knommu/lib/Makefile
+++ b/arch/m68knommu/lib/Makefile
@@ -4,4 +4,4 @@
 
 lib-y	:= ashldi3.o ashrdi3.o lshrdi3.o \
 	   muldi3.o mulsi3.o divsi3.o udivsi3.o modsi3.o umodsi3.o \
-	   checksum.o semaphore.o memcpy.o memset.o delay.o
+	   checksum.o memcpy.o memset.o delay.o
diff --git a/arch/m68knommu/lib/semaphore.S b/arch/m68knommu/lib/semaphore.S
deleted file mode 100644
index 87c746034376..000000000000
--- a/arch/m68knommu/lib/semaphore.S
+++ /dev/null
@@ -1,66 +0,0 @@
-/*
- *  linux/arch/m68k/lib/semaphore.S
- *
- *  Copyright (C) 1996  Linus Torvalds
- *
- *  m68k version by Andreas Schwab
- *
- *  MAR/1999 -- modified to support ColdFire (gerg@snapgear.com)
- */
-
-#include <linux/linkage.h>
-#include <asm/semaphore.h>
-
-/*
- * "down_failed" is called with the eventual return address
- * in %a0, and the address of the semaphore in %a1. We need
- * to increment the number of waiters on the semaphore,
- * call "__down()", and then eventually return to try again.
- */
-ENTRY(__down_failed)
-#ifdef CONFIG_COLDFIRE
-	subl #12,%sp
-	moveml %a0/%d0/%d1,(%sp)
-#else
-	moveml %a0/%d0/%d1,-(%sp)
-#endif
-	movel %a1,-(%sp)
-	jbsr __down
-	movel (%sp)+,%a1
-	movel (%sp)+,%d0
-	movel (%sp)+,%d1
-	rts
-
-ENTRY(__down_failed_interruptible)
-	movel %a0,-(%sp)
-	movel %d1,-(%sp)
-	movel %a1,-(%sp)
-	jbsr __down_interruptible
-	movel (%sp)+,%a1
-	movel (%sp)+,%d1
-	rts
-
-ENTRY(__up_wakeup)
-#ifdef CONFIG_COLDFIRE
-	subl #12,%sp
-	moveml %a0/%d0/%d1,(%sp)
-#else
-	moveml %a0/%d0/%d1,-(%sp)
-#endif
-	movel %a1,-(%sp)
-	jbsr __up
-	movel (%sp)+,%a1
-	movel (%sp)+,%d0
-	movel (%sp)+,%d1
-	rts
-
-ENTRY(__down_failed_trylock)
-	movel %a0,-(%sp)
-	movel %d1,-(%sp)
-	movel %a1,-(%sp)
-	jbsr __down_trylock
-	movel (%sp)+,%a1
-	movel (%sp)+,%d1
-	movel (%sp)+,%a0
-	rts
-
diff --git a/arch/mips/kernel/Makefile b/arch/mips/kernel/Makefile
index 9e78e1a4ca17..6fcdb6fda2e2 100644
--- a/arch/mips/kernel/Makefile
+++ b/arch/mips/kernel/Makefile
@@ -5,7 +5,7 @@
 extra-y		:= head.o init_task.o vmlinux.lds
 
 obj-y		+= cpu-probe.o branch.o entry.o genex.o irq.o process.o \
-		   ptrace.o reset.o semaphore.o setup.o signal.o syscall.o \
+		   ptrace.o reset.o setup.o signal.o syscall.o \
 		   time.o topology.o traps.o unaligned.o
 
 obj-$(CONFIG_CEVT_BCM1480)	+= cevt-bcm1480.o
diff --git a/arch/mips/kernel/semaphore.c b/arch/mips/kernel/semaphore.c
deleted file mode 100644
index 1265358cdca1..000000000000
--- a/arch/mips/kernel/semaphore.c
+++ /dev/null
@@ -1,168 +0,0 @@
-/*
- * MIPS-specific semaphore code.
- *
- * Copyright (C) 1999 Cort Dougan <cort@cs.nmt.edu>
- * Copyright (C) 2004 Ralf Baechle <ralf@linux-mips.org>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version
- * 2 of the License, or (at your option) any later version.
- *
- * April 2001 - Reworked by Paul Mackerras <paulus@samba.org>
- * to eliminate the SMP races in the old version between the updates
- * of `count' and `waking'.  Now we use negative `count' values to
- * indicate that some process(es) are waiting for the semaphore.
- */
-
-#include <linux/module.h>
-#include <linux/sched.h>
-#include <linux/init.h>
-#include <asm/atomic.h>
-#include <asm/cpu-features.h>
-#include <asm/errno.h>
-#include <asm/semaphore.h>
-#include <asm/war.h>
-/*
- * Atomically update sem->count.
- * This does the equivalent of the following:
- *
- *	old_count = sem->count;
- *	tmp = MAX(old_count, 0) + incr;
- *	sem->count = tmp;
- *	return old_count;
- *
- * On machines without lld/scd we need a spinlock to make the manipulation of
- * sem->count and sem->waking atomic.  Scalability isn't an issue because
- * this lock is used on UP only so it's just an empty variable.
- */
-static inline int __sem_update_count(struct semaphore *sem, int incr)
-{
-	int old_count, tmp;
-
-	if (cpu_has_llsc && R10000_LLSC_WAR) {
-		__asm__ __volatile__(
-		"	.set	mips3					\n"
-		"1:	ll	%0, %2		# __sem_update_count	\n"
-		"	sra	%1, %0, 31				\n"
-		"	not	%1					\n"
-		"	and	%1, %0, %1				\n"
-		"	addu	%1, %1, %3				\n"
-		"	sc	%1, %2					\n"
-		"	beqzl	%1, 1b					\n"
-		"	.set	mips0					\n"
-		: "=&r" (old_count), "=&r" (tmp), "=m" (sem->count)
-		: "r" (incr), "m" (sem->count));
-	} else if (cpu_has_llsc) {
-		__asm__ __volatile__(
-		"	.set	mips3					\n"
-		"1:	ll	%0, %2		# __sem_update_count	\n"
-		"	sra	%1, %0, 31				\n"
-		"	not	%1					\n"
-		"	and	%1, %0, %1				\n"
-		"	addu	%1, %1, %3				\n"
-		"	sc	%1, %2					\n"
-		"	beqz	%1, 1b					\n"
-		"	.set	mips0					\n"
-		: "=&r" (old_count), "=&r" (tmp), "=m" (sem->count)
-		: "r" (incr), "m" (sem->count));
-	} else {
-		static DEFINE_SPINLOCK(semaphore_lock);
-		unsigned long flags;
-
-		spin_lock_irqsave(&semaphore_lock, flags);
-		old_count = atomic_read(&sem->count);
-		tmp = max_t(int, old_count, 0) + incr;
-		atomic_set(&sem->count, tmp);
-		spin_unlock_irqrestore(&semaphore_lock, flags);
-	}
-
-	return old_count;
-}
-
-void __up(struct semaphore *sem)
-{
-	/*
-	 * Note that we incremented count in up() before we came here,
-	 * but that was ineffective since the result was <= 0, and
-	 * any negative value of count is equivalent to 0.
-	 * This ends up setting count to 1, unless count is now > 0
-	 * (i.e. because some other cpu has called up() in the meantime),
-	 * in which case we just increment count.
-	 */
-	__sem_update_count(sem, 1);
-	wake_up(&sem->wait);
-}
-
-EXPORT_SYMBOL(__up);
-
-/*
- * Note that when we come in to __down or __down_interruptible,
- * we have already decremented count, but that decrement was
- * ineffective since the result was < 0, and any negative value
- * of count is equivalent to 0.
- * Thus it is only when we decrement count from some value > 0
- * that we have actually got the semaphore.
- */
-void __sched __down(struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	__set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	/*
-	 * Try to get the semaphore.  If the count is > 0, then we've
-	 * got the semaphore; we decrement count and exit the loop.
-	 * If the count is 0 or negative, we set it to -1, indicating
-	 * that we are asleep, and then sleep.
-	 */
-	while (__sem_update_count(sem, -1) <= 0) {
-		schedule();
-		set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	__set_task_state(tsk, TASK_RUNNING);
-
-	/*
-	 * If there are any more sleepers, wake one of them up so
-	 * that it can either get the semaphore, or set count to -1
-	 * indicating that there are still processes sleeping.
-	 */
-	wake_up(&sem->wait);
-}
-
-EXPORT_SYMBOL(__down);
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	__set_task_state(tsk, TASK_INTERRUPTIBLE);
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	while (__sem_update_count(sem, -1) <= 0) {
-		if (signal_pending(current)) {
-			/*
-			 * A signal is pending - give up trying.
-			 * Set sem->count to 0 if it is negative,
-			 * since we are no longer sleeping.
-			 */
-			__sem_update_count(sem, 0);
-			retval = -EINTR;
-			break;
-		}
-		schedule();
-		set_task_state(tsk, TASK_INTERRUPTIBLE);
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	__set_task_state(tsk, TASK_RUNNING);
-
-	wake_up(&sem->wait);
-	return retval;
-}
-
-EXPORT_SYMBOL(__down_interruptible);
diff --git a/arch/mn10300/kernel/Makefile b/arch/mn10300/kernel/Makefile
index ef07c956170a..23f2ab67574c 100644
--- a/arch/mn10300/kernel/Makefile
+++ b/arch/mn10300/kernel/Makefile
@@ -3,7 +3,7 @@
 #
 extra-y := head.o init_task.o vmlinux.lds
 
-obj-y   := process.o semaphore.o signal.o entry.o fpu.o traps.o irq.o \
+obj-y   := process.o signal.o entry.o fpu.o traps.o irq.o \
 	   ptrace.o setup.o time.o sys_mn10300.o io.o kthread.o \
 	   switch_to.o mn10300_ksyms.o kernel_execve.o
 
diff --git a/arch/mn10300/kernel/semaphore.c b/arch/mn10300/kernel/semaphore.c
deleted file mode 100644
index 9153c4039fd2..000000000000
--- a/arch/mn10300/kernel/semaphore.c
+++ /dev/null
@@ -1,149 +0,0 @@
-/* MN10300 Semaphore implementation
- *
- * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
- * Written by David Howells (dhowells@redhat.com)
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public Licence
- * as published by the Free Software Foundation; either version
- * 2 of the Licence, or (at your option) any later version.
- */
-#include <linux/sched.h>
-#include <linux/module.h>
-#include <asm/semaphore.h>
-
-struct sem_waiter {
-	struct list_head	list;
-	struct task_struct	*task;
-};
-
-#if SEMAPHORE_DEBUG
-void semtrace(struct semaphore *sem, const char *str)
-{
-	if (sem->debug)
-		printk(KERN_DEBUG "[%d] %s({%d,%d})\n",
-		       current->pid,
-		       str,
-		       atomic_read(&sem->count),
-		       list_empty(&sem->wait_list) ? 0 : 1);
-}
-#else
-#define semtrace(SEM, STR) do { } while (0)
-#endif
-
-/*
- * wait for a token to be granted from a semaphore
- * - entered with lock held and interrupts disabled
- */
-void __down(struct semaphore *sem, unsigned long flags)
-{
-	struct task_struct *tsk = current;
-	struct sem_waiter waiter;
-
-	semtrace(sem, "Entering __down");
-
-	/* set up my own style of waitqueue */
-	waiter.task = tsk;
-	get_task_struct(tsk);
-
-	list_add_tail(&waiter.list, &sem->wait_list);
-
-	/* we don't need to touch the semaphore struct anymore */
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-
-	/* wait to be given the semaphore */
-	set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-
-	for (;;) {
-		if (!waiter.task)
-			break;
-		schedule();
-		set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	}
-
-	tsk->state = TASK_RUNNING;
-	semtrace(sem, "Leaving __down");
-}
-EXPORT_SYMBOL(__down);
-
-/*
- * interruptibly wait for a token to be granted from a semaphore
- * - entered with lock held and interrupts disabled
- */
-int __down_interruptible(struct semaphore *sem, unsigned long flags)
-{
-	struct task_struct *tsk = current;
-	struct sem_waiter waiter;
-	int ret;
-
-	semtrace(sem, "Entering __down_interruptible");
-
-	/* set up my own style of waitqueue */
-	waiter.task = tsk;
-	get_task_struct(tsk);
-
-	list_add_tail(&waiter.list, &sem->wait_list);
-
-	/* we don't need to touch the semaphore struct anymore */
-	set_task_state(tsk, TASK_INTERRUPTIBLE);
-
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-
-	/* wait to be given the semaphore */
-	ret = 0;
-	for (;;) {
-		if (!waiter.task)
-			break;
-		if (unlikely(signal_pending(current)))
-			goto interrupted;
-		schedule();
-		set_task_state(tsk, TASK_INTERRUPTIBLE);
-	}
-
- out:
-	tsk->state = TASK_RUNNING;
-	semtrace(sem, "Leaving __down_interruptible");
-	return ret;
-
- interrupted:
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	list_del(&waiter.list);
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-
-	ret = 0;
-	if (!waiter.task) {
-		put_task_struct(current);
-		ret = -EINTR;
-	}
-	goto out;
-}
-EXPORT_SYMBOL(__down_interruptible);
-
-/*
- * release a single token back to a semaphore
- * - entered with lock held and interrupts disabled
- */
-void __up(struct semaphore *sem)
-{
-	struct task_struct *tsk;
-	struct sem_waiter *waiter;
-
-	semtrace(sem, "Entering __up");
-
-	/* grant the token to the process at the front of the queue */
-	waiter = list_entry(sem->wait_list.next, struct sem_waiter, list);
-
-	/* We must be careful not to touch 'waiter' after we set ->task = NULL.
-	 * It is an allocated on the waiter's stack and may become invalid at
-	 * any time after that point (due to a wakeup from another source).
-	 */
-	list_del_init(&waiter->list);
-	tsk = waiter->task;
-	smp_mb();
-	waiter->task = NULL;
-	wake_up_process(tsk);
-	put_task_struct(tsk);
-
-	semtrace(sem, "Leaving __up");
-}
-EXPORT_SYMBOL(__up);
diff --git a/arch/parisc/kernel/Makefile b/arch/parisc/kernel/Makefile
index 27827bc3717e..1f6585a56f97 100644
--- a/arch/parisc/kernel/Makefile
+++ b/arch/parisc/kernel/Makefile
@@ -9,7 +9,7 @@ AFLAGS_pacache.o := -traditional
 
 obj-y	     	:= cache.o pacache.o setup.o traps.o time.o irq.o \
 		   pa7300lc.o syscall.o entry.o sys_parisc.o firmware.o \
-		   ptrace.o hardware.o inventory.o drivers.o semaphore.o \
+		   ptrace.o hardware.o inventory.o drivers.o \
 		   signal.o hpmc.o real2.o parisc_ksyms.o unaligned.o \
 		   process.o processor.o pdc_cons.o pdc_chassis.o unwind.o \
 		   topology.o
diff --git a/arch/parisc/kernel/parisc_ksyms.c b/arch/parisc/kernel/parisc_ksyms.c
index 7aca704e96f0..5b7fc4aa044d 100644
--- a/arch/parisc/kernel/parisc_ksyms.c
+++ b/arch/parisc/kernel/parisc_ksyms.c
@@ -69,11 +69,6 @@ EXPORT_SYMBOL(memcpy_toio);
 EXPORT_SYMBOL(memcpy_fromio);
 EXPORT_SYMBOL(memset_io);
 
-#include <asm/semaphore.h>
-EXPORT_SYMBOL(__up);
-EXPORT_SYMBOL(__down_interruptible);
-EXPORT_SYMBOL(__down);
-
 extern void $$divI(void);
 extern void $$divU(void);
 extern void $$remI(void);
diff --git a/arch/parisc/kernel/semaphore.c b/arch/parisc/kernel/semaphore.c
deleted file mode 100644
index ee806bcc3726..000000000000
--- a/arch/parisc/kernel/semaphore.c
+++ /dev/null
@@ -1,102 +0,0 @@
-/*
- * Semaphore implementation Copyright (c) 2001 Matthew Wilcox, Hewlett-Packard
- */
-
-#include <linux/sched.h>
-#include <linux/spinlock.h>
-#include <linux/errno.h>
-#include <linux/init.h>
-
-/*
- * Semaphores are complex as we wish to avoid using two variables.
- * `count' has multiple roles, depending on its value.  If it is positive
- * or zero, there are no waiters.  The functions here will never be
- * called; see <asm/semaphore.h>
- *
- * When count is -1 it indicates there is at least one task waiting
- * for the semaphore.
- *
- * When count is less than that, there are '- count - 1' wakeups
- * pending.  ie if it has value -3, there are 2 wakeups pending.
- *
- * Note that these functions are only called when there is contention
- * on the lock, and as such all this is the "non-critical" part of the
- * whole semaphore business. The critical part is the inline stuff in
- * <asm/semaphore.h> where we want to avoid any extra jumps and calls.
- */
-void __up(struct semaphore *sem)
-{
-	sem->count--;
-	wake_up(&sem->wait);
-}
-
-#define wakers(count) (-1 - count)
-
-#define DOWN_HEAD							\
-	int ret = 0;							\
-	DECLARE_WAITQUEUE(wait, current);				\
-									\
-	/* Note that someone is waiting */				\
-	if (sem->count == 0)						\
-		sem->count = -1;					\
-									\
-	/* protected by the sentry still -- use unlocked version */	\
-	wait.flags = WQ_FLAG_EXCLUSIVE;					\
-	__add_wait_queue_tail(&sem->wait, &wait);			\
- lost_race:								\
-	spin_unlock_irq(&sem->sentry);					\
-
-#define DOWN_TAIL							\
-	spin_lock_irq(&sem->sentry);					\
-	if (wakers(sem->count) == 0 && ret == 0)			\
-		goto lost_race;	/* Someone stole our wakeup */		\
-	__remove_wait_queue(&sem->wait, &wait);				\
-	current->state = TASK_RUNNING;					\
-	if (!waitqueue_active(&sem->wait) && (sem->count < 0))		\
-		sem->count = wakers(sem->count);
-
-#define UPDATE_COUNT							\
-	sem->count += (sem->count < 0) ? 1 : - 1;
-	
-
-void __sched __down(struct semaphore * sem)
-{
-	DOWN_HEAD
-
-	for(;;) {
-		set_task_state(current, TASK_UNINTERRUPTIBLE);
-		/* we can _read_ this without the sentry */
-		if (sem->count != -1)
-			break;
- 		schedule();
- 	}
-
-	DOWN_TAIL
-	UPDATE_COUNT
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	DOWN_HEAD
-
-	for(;;) {
-		set_task_state(current, TASK_INTERRUPTIBLE);
-		/* we can _read_ this without the sentry */
-		if (sem->count != -1)
-			break;
-
-		if (signal_pending(current)) {
-			ret = -EINTR;
-			break;
-		}
-		schedule();
-	}
-
-	DOWN_TAIL
-
-	if (!ret) {
-		UPDATE_COUNT
-	}
-
-	return ret;
-}
diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile
index c1baf9d5903f..b9dbfff9afe9 100644
--- a/arch/powerpc/kernel/Makefile
+++ b/arch/powerpc/kernel/Makefile
@@ -12,7 +12,7 @@ CFLAGS_prom_init.o      += -fPIC
 CFLAGS_btext.o		+= -fPIC
 endif
 
-obj-y				:= semaphore.o cputable.o ptrace.o syscalls.o \
+obj-y				:= cputable.o ptrace.o syscalls.o \
 				   irq.o align.o signal_32.o pmc.o vdso.o \
 				   init_task.o process.o systbl.o idle.o \
 				   signal.o
diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c
index 9c98424277a8..65d14e6ddc3c 100644
--- a/arch/powerpc/kernel/ppc_ksyms.c
+++ b/arch/powerpc/kernel/ppc_ksyms.c
@@ -15,7 +15,6 @@
 #include <linux/bitops.h>
 
 #include <asm/page.h>
-#include <asm/semaphore.h>
 #include <asm/processor.h>
 #include <asm/cacheflush.h>
 #include <asm/uaccess.h>
diff --git a/arch/powerpc/kernel/semaphore.c b/arch/powerpc/kernel/semaphore.c
deleted file mode 100644
index 2f8c3c951394..000000000000
--- a/arch/powerpc/kernel/semaphore.c
+++ /dev/null
@@ -1,135 +0,0 @@
-/*
- * PowerPC-specific semaphore code.
- *
- * Copyright (C) 1999 Cort Dougan <cort@cs.nmt.edu>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version
- * 2 of the License, or (at your option) any later version.
- *
- * April 2001 - Reworked by Paul Mackerras <paulus@samba.org>
- * to eliminate the SMP races in the old version between the updates
- * of `count' and `waking'.  Now we use negative `count' values to
- * indicate that some process(es) are waiting for the semaphore.
- */
-
-#include <linux/sched.h>
-#include <linux/init.h>
-#include <linux/module.h>
-
-#include <asm/atomic.h>
-#include <asm/semaphore.h>
-#include <asm/errno.h>
-
-/*
- * Atomically update sem->count.
- * This does the equivalent of the following:
- *
- *	old_count = sem->count;
- *	tmp = MAX(old_count, 0) + incr;
- *	sem->count = tmp;
- *	return old_count;
- */
-static inline int __sem_update_count(struct semaphore *sem, int incr)
-{
-	int old_count, tmp;
-
-	__asm__ __volatile__("\n"
-"1:	lwarx	%0,0,%3\n"
-"	srawi	%1,%0,31\n"
-"	andc	%1,%0,%1\n"
-"	add	%1,%1,%4\n"
-	PPC405_ERR77(0,%3)
-"	stwcx.	%1,0,%3\n"
-"	bne	1b"
-	: "=&r" (old_count), "=&r" (tmp), "=m" (sem->count)
-	: "r" (&sem->count), "r" (incr), "m" (sem->count)
-	: "cc");
-
-	return old_count;
-}
-
-void __up(struct semaphore *sem)
-{
-	/*
-	 * Note that we incremented count in up() before we came here,
-	 * but that was ineffective since the result was <= 0, and
-	 * any negative value of count is equivalent to 0.
-	 * This ends up setting count to 1, unless count is now > 0
-	 * (i.e. because some other cpu has called up() in the meantime),
-	 * in which case we just increment count.
-	 */
-	__sem_update_count(sem, 1);
-	wake_up(&sem->wait);
-}
-EXPORT_SYMBOL(__up);
-
-/*
- * Note that when we come in to __down or __down_interruptible,
- * we have already decremented count, but that decrement was
- * ineffective since the result was < 0, and any negative value
- * of count is equivalent to 0.
- * Thus it is only when we decrement count from some value > 0
- * that we have actually got the semaphore.
- */
-void __sched __down(struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	__set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	/*
-	 * Try to get the semaphore.  If the count is > 0, then we've
-	 * got the semaphore; we decrement count and exit the loop.
-	 * If the count is 0 or negative, we set it to -1, indicating
-	 * that we are asleep, and then sleep.
-	 */
-	while (__sem_update_count(sem, -1) <= 0) {
-		schedule();
-		set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	__set_task_state(tsk, TASK_RUNNING);
-
-	/*
-	 * If there are any more sleepers, wake one of them up so
-	 * that it can either get the semaphore, or set count to -1
-	 * indicating that there are still processes sleeping.
-	 */
-	wake_up(&sem->wait);
-}
-EXPORT_SYMBOL(__down);
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	__set_task_state(tsk, TASK_INTERRUPTIBLE);
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	while (__sem_update_count(sem, -1) <= 0) {
-		if (signal_pending(current)) {
-			/*
-			 * A signal is pending - give up trying.
-			 * Set sem->count to 0 if it is negative,
-			 * since we are no longer sleeping.
-			 */
-			__sem_update_count(sem, 0);
-			retval = -EINTR;
-			break;
-		}
-		schedule();
-		set_task_state(tsk, TASK_INTERRUPTIBLE);
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	__set_task_state(tsk, TASK_RUNNING);
-
-	wake_up(&sem->wait);
-	return retval;
-}
-EXPORT_SYMBOL(__down_interruptible);
diff --git a/arch/ppc/kernel/semaphore.c b/arch/ppc/kernel/semaphore.c
deleted file mode 100644
index 2fe429b27c14..000000000000
--- a/arch/ppc/kernel/semaphore.c
+++ /dev/null
@@ -1,131 +0,0 @@
-/*
- * PowerPC-specific semaphore code.
- *
- * Copyright (C) 1999 Cort Dougan <cort@cs.nmt.edu>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version
- * 2 of the License, or (at your option) any later version.
- *
- * April 2001 - Reworked by Paul Mackerras <paulus@samba.org>
- * to eliminate the SMP races in the old version between the updates
- * of `count' and `waking'.  Now we use negative `count' values to
- * indicate that some process(es) are waiting for the semaphore.
- */
-
-#include <linux/sched.h>
-#include <linux/init.h>
-#include <asm/atomic.h>
-#include <asm/semaphore.h>
-#include <asm/errno.h>
-
-/*
- * Atomically update sem->count.
- * This does the equivalent of the following:
- *
- *	old_count = sem->count;
- *	tmp = MAX(old_count, 0) + incr;
- *	sem->count = tmp;
- *	return old_count;
- */
-static inline int __sem_update_count(struct semaphore *sem, int incr)
-{
-	int old_count, tmp;
-
-	__asm__ __volatile__("\n"
-"1:	lwarx	%0,0,%3\n"
-"	srawi	%1,%0,31\n"
-"	andc	%1,%0,%1\n"
-"	add	%1,%1,%4\n"
-	PPC405_ERR77(0,%3)
-"	stwcx.	%1,0,%3\n"
-"	bne	1b"
-	: "=&r" (old_count), "=&r" (tmp), "=m" (sem->count)
-	: "r" (&sem->count), "r" (incr), "m" (sem->count)
-	: "cc");
-
-	return old_count;
-}
-
-void __up(struct semaphore *sem)
-{
-	/*
-	 * Note that we incremented count in up() before we came here,
-	 * but that was ineffective since the result was <= 0, and
-	 * any negative value of count is equivalent to 0.
-	 * This ends up setting count to 1, unless count is now > 0
-	 * (i.e. because some other cpu has called up() in the meantime),
-	 * in which case we just increment count.
-	 */
-	__sem_update_count(sem, 1);
-	wake_up(&sem->wait);
-}
-
-/*
- * Note that when we come in to __down or __down_interruptible,
- * we have already decremented count, but that decrement was
- * ineffective since the result was < 0, and any negative value
- * of count is equivalent to 0.
- * Thus it is only when we decrement count from some value > 0
- * that we have actually got the semaphore.
- */
-void __sched __down(struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-	smp_wmb();
-
-	/*
-	 * Try to get the semaphore.  If the count is > 0, then we've
-	 * got the semaphore; we decrement count and exit the loop.
-	 * If the count is 0 or negative, we set it to -1, indicating
-	 * that we are asleep, and then sleep.
-	 */
-	while (__sem_update_count(sem, -1) <= 0) {
-		schedule();
-		tsk->state = TASK_UNINTERRUPTIBLE;
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	tsk->state = TASK_RUNNING;
-
-	/*
-	 * If there are any more sleepers, wake one of them up so
-	 * that it can either get the semaphore, or set count to -1
-	 * indicating that there are still processes sleeping.
-	 */
-	wake_up(&sem->wait);
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	tsk->state = TASK_INTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-	smp_wmb();
-
-	while (__sem_update_count(sem, -1) <= 0) {
-		if (signal_pending(current)) {
-			/*
-			 * A signal is pending - give up trying.
-			 * Set sem->count to 0 if it is negative,
-			 * since we are no longer sleeping.
-			 */
-			__sem_update_count(sem, 0);
-			retval = -EINTR;
-			break;
-		}
-		schedule();
-		tsk->state = TASK_INTERRUPTIBLE;
-	}
-	tsk->state = TASK_RUNNING;
-	remove_wait_queue(&sem->wait, &wait);
-	wake_up(&sem->wait);
-	return retval;
-}
diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile
index 4d3e38392cb1..ce144b67f060 100644
--- a/arch/s390/kernel/Makefile
+++ b/arch/s390/kernel/Makefile
@@ -11,7 +11,7 @@ CFLAGS_smp.o	:= -Wno-nonnull
 
 obj-y	:=  bitmap.o traps.o time.o process.o base.o early.o \
             setup.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o \
-	    semaphore.o s390_ext.o debug.o irq.o ipl.o dis.o diag.o
+	    s390_ext.o debug.o irq.o ipl.o dis.o diag.o
 
 obj-y	+= $(if $(CONFIG_64BIT),entry64.o,entry.o)
 obj-y	+= $(if $(CONFIG_64BIT),reipl64.o,reipl.o)
diff --git a/arch/s390/kernel/s390_ksyms.c b/arch/s390/kernel/s390_ksyms.c
index 7234c737f825..48238a114ce9 100644
--- a/arch/s390/kernel/s390_ksyms.c
+++ b/arch/s390/kernel/s390_ksyms.c
@@ -26,13 +26,6 @@ EXPORT_SYMBOL(_ni_bitmap);
 EXPORT_SYMBOL(_zb_findmap);
 EXPORT_SYMBOL(_sb_findmap);
 
-/*
- * semaphore ops
- */
-EXPORT_SYMBOL(__up);
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_interruptible);
-
 /*
  * binfmt_elf loader 
  */
diff --git a/arch/s390/kernel/semaphore.c b/arch/s390/kernel/semaphore.c
deleted file mode 100644
index 191303f6c1d8..000000000000
--- a/arch/s390/kernel/semaphore.c
+++ /dev/null
@@ -1,108 +0,0 @@
-/*
- *  linux/arch/s390/kernel/semaphore.c
- *
- *  S390 version
- *    Copyright (C) 1998-2000 IBM Corporation
- *    Author(s): Martin Schwidefsky
- *
- *  Derived from "linux/arch/i386/kernel/semaphore.c
- *    Copyright (C) 1999, Linus Torvalds
- *
- */
-#include <linux/sched.h>
-#include <linux/errno.h>
-#include <linux/init.h>
-
-#include <asm/semaphore.h>
-
-/*
- * Atomically update sem->count. Equivalent to:
- *   old_val = sem->count.counter;
- *   new_val = ((old_val >= 0) ? old_val : 0) + incr;
- *   sem->count.counter = new_val;
- *   return old_val;
- */
-static inline int __sem_update_count(struct semaphore *sem, int incr)
-{
-	int old_val, new_val;
-
-	asm volatile(
-		"	l	%0,0(%3)\n"
-		"0:	ltr	%1,%0\n"
-		"	jhe	1f\n"
-		"	lhi	%1,0\n"
-		"1:	ar	%1,%4\n"
-		"	cs	%0,%1,0(%3)\n"
-		"	jl	0b\n"
-		: "=&d" (old_val), "=&d" (new_val), "=m" (sem->count)
-		: "a" (&sem->count), "d" (incr), "m" (sem->count)
-		: "cc");
-	return old_val;
-}
-
-/*
- * The inline function up() incremented count but the result
- * was <= 0. This indicates that some process is waiting on
- * the semaphore. The semaphore is free and we'll wake the
- * first sleeping process, so we set count to 1 unless some
- * other cpu has called up in the meantime in which case
- * we just increment count by 1.
- */
-void __up(struct semaphore *sem)
-{
-	__sem_update_count(sem, 1);
-	wake_up(&sem->wait);
-}
-
-/*
- * The inline function down() decremented count and the result
- * was < 0. The wait loop will atomically test and update the
- * semaphore counter following the rules:
- *   count > 0: decrement count, wake up queue and exit.
- *   count <= 0: set count to -1, go to sleep.
- */
-void __sched __down(struct semaphore * sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	__set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	add_wait_queue_exclusive(&sem->wait, &wait);
-	while (__sem_update_count(sem, -1) <= 0) {
-		schedule();
-		set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	__set_task_state(tsk, TASK_RUNNING);
-	wake_up(&sem->wait);
-}
-
-/*
- * Same as __down() with an additional test for signals.
- * If a signal is pending the count is updated as follows:
- *   count > 0: wake up queue and exit.
- *   count <= 0: set count to 0, wake up queue and exit.
- */
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	__set_task_state(tsk, TASK_INTERRUPTIBLE);
-	add_wait_queue_exclusive(&sem->wait, &wait);
-	while (__sem_update_count(sem, -1) <= 0) {
-		if (signal_pending(current)) {
-			__sem_update_count(sem, 0);
-			retval = -EINTR;
-			break;
-		}
-		schedule();
-		set_task_state(tsk, TASK_INTERRUPTIBLE);
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	__set_task_state(tsk, TASK_RUNNING);
-	wake_up(&sem->wait);
-	return retval;
-}
-
diff --git a/arch/sh/kernel/Makefile_32 b/arch/sh/kernel/Makefile_32
index 62bf373266f7..4bbdce36b92b 100644
--- a/arch/sh/kernel/Makefile_32
+++ b/arch/sh/kernel/Makefile_32
@@ -5,7 +5,7 @@
 extra-y	:= head_32.o init_task.o vmlinux.lds
 
 obj-y	:= debugtraps.o io.o io_generic.o irq.o machvec.o process_32.o \
-	   ptrace_32.o semaphore.o setup.o signal_32.o sys_sh.o sys_sh32.o \
+	   ptrace_32.o setup.o signal_32.o sys_sh.o sys_sh32.o \
 	   syscalls_32.o time_32.o topology.o traps.o traps_32.o
 
 obj-y				+= cpu/ timers/
diff --git a/arch/sh/kernel/Makefile_64 b/arch/sh/kernel/Makefile_64
index e01283d49cbf..6edf53b93d94 100644
--- a/arch/sh/kernel/Makefile_64
+++ b/arch/sh/kernel/Makefile_64
@@ -1,7 +1,7 @@
 extra-y	:= head_64.o init_task.o vmlinux.lds
 
 obj-y	:= debugtraps.o io.o io_generic.o irq.o machvec.o process_64.o \
-	   ptrace_64.o semaphore.o setup.o signal_64.o sys_sh.o sys_sh64.o \
+	   ptrace_64.o setup.o signal_64.o sys_sh.o sys_sh64.o \
 	   syscalls_64.o time_64.o topology.o traps.o traps_64.o
 
 obj-y				+= cpu/ timers/
diff --git a/arch/sh/kernel/semaphore.c b/arch/sh/kernel/semaphore.c
deleted file mode 100644
index 184119eeae56..000000000000
--- a/arch/sh/kernel/semaphore.c
+++ /dev/null
@@ -1,139 +0,0 @@
-/*
- * Just taken from alpha implementation.
- * This can't work well, perhaps.
- */
-/*
- *  Generic semaphore code. Buyer beware. Do your own
- * specific changes in <asm/semaphore-helper.h>
- */
-
-#include <linux/errno.h>
-#include <linux/sched.h>
-#include <linux/wait.h>
-#include <linux/init.h>
-#include <asm/semaphore.h>
-#include <asm/semaphore-helper.h>
-
-DEFINE_SPINLOCK(semaphore_wake_lock);
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to sleep, while the "waking" variable is
- * incremented when the "up()" code goes to wake up waiting
- * processes.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * waking_non_zero() (from asm/semaphore.h) must execute
- * atomically.
- *
- * When __up() is called, the count was negative before
- * incrementing it, and we need to wake up somebody.
- *
- * This routine adds one to the count of processes that need to
- * wake up and exit.  ALL waiting processes actually wake up but
- * only the one that gets to the "waking" field first will gate
- * through and acquire the semaphore.  The others will go back
- * to sleep.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-void __up(struct semaphore *sem)
-{
-	wake_one_more(sem);
-	wake_up(&sem->wait);
-}
-
-/*
- * Perform the "down" function.  Return zero for semaphore acquired,
- * return negative for signalled out of the function.
- *
- * If called from __down, the return is ignored and the wait loop is
- * not interruptible.  This means that a task waiting on a semaphore
- * using "down()" cannot be killed until someone does an "up()" on
- * the semaphore.
- *
- * If called from __down_interruptible, the return value gets checked
- * upon return.  If the return value is negative then the task continues
- * with the negative value in the return register (it can be tested by
- * the caller).
- *
- * Either form may be used in conjunction with "up()".
- *
- */
-
-#define DOWN_VAR				\
-	struct task_struct *tsk = current;	\
-	wait_queue_t wait;			\
-	init_waitqueue_entry(&wait, tsk);
-
-#define DOWN_HEAD(task_state)						\
-									\
-									\
-	tsk->state = (task_state);					\
-	add_wait_queue(&sem->wait, &wait);				\
-									\
-	/*								\
-	 * Ok, we're set up.  sem->count is known to be less than zero	\
-	 * so we must wait.						\
-	 *								\
-	 * We can let go the lock for purposes of waiting.		\
-	 * We re-acquire it after awaking so as to protect		\
-	 * all semaphore operations.					\
-	 *								\
-	 * If "up()" is called before we call waking_non_zero() then	\
-	 * we will catch it right away.  If it is called later then	\
-	 * we will have to go through a wakeup cycle to catch it.	\
-	 *								\
-	 * Multiple waiters contend for the semaphore lock to see	\
-	 * who gets to gate through and who has to wait some more.	\
-	 */								\
-	for (;;) {
-
-#define DOWN_TAIL(task_state)			\
-		tsk->state = (task_state);	\
-	}					\
-	tsk->state = TASK_RUNNING;		\
-	remove_wait_queue(&sem->wait, &wait);
-
-void __sched __down(struct semaphore * sem)
-{
-	DOWN_VAR
-	DOWN_HEAD(TASK_UNINTERRUPTIBLE)
-	if (waking_non_zero(sem))
-		break;
-	schedule();
-	DOWN_TAIL(TASK_UNINTERRUPTIBLE)
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-	DOWN_VAR
-	DOWN_HEAD(TASK_INTERRUPTIBLE)
-
-	ret = waking_non_zero_interruptible(sem, tsk);
-	if (ret)
-	{
-		if (ret == 1)
-			/* ret != 0 only if we get interrupted -arca */
-			ret = 0;
-		break;
-	}
-	schedule();
-	DOWN_TAIL(TASK_INTERRUPTIBLE)
-	return ret;
-}
-
-int __down_trylock(struct semaphore * sem)
-{
-	return waking_non_zero_trylock(sem);
-}
diff --git a/arch/sh/kernel/sh_ksyms_32.c b/arch/sh/kernel/sh_ksyms_32.c
index 45bb333fd9ec..6d405462cee8 100644
--- a/arch/sh/kernel/sh_ksyms_32.c
+++ b/arch/sh/kernel/sh_ksyms_32.c
@@ -9,7 +9,6 @@
 #include <linux/pci.h>
 #include <linux/irq.h>
 #include <asm/sections.h>
-#include <asm/semaphore.h>
 #include <asm/processor.h>
 #include <asm/uaccess.h>
 #include <asm/checksum.h>
@@ -48,12 +47,6 @@ EXPORT_SYMBOL(__copy_user);
 EXPORT_SYMBOL(get_vm_area);
 #endif
 
-/* semaphore exports */
-EXPORT_SYMBOL(__up);
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_interruptible);
-EXPORT_SYMBOL(__down_trylock);
-
 EXPORT_SYMBOL(__udelay);
 EXPORT_SYMBOL(__ndelay);
 EXPORT_SYMBOL(__const_udelay);
diff --git a/arch/sh/kernel/sh_ksyms_64.c b/arch/sh/kernel/sh_ksyms_64.c
index b6410ce4bd1d..a310c9707f03 100644
--- a/arch/sh/kernel/sh_ksyms_64.c
+++ b/arch/sh/kernel/sh_ksyms_64.c
@@ -16,7 +16,6 @@
 #include <linux/in6.h>
 #include <linux/interrupt.h>
 #include <linux/screen_info.h>
-#include <asm/semaphore.h>
 #include <asm/processor.h>
 #include <asm/uaccess.h>
 #include <asm/checksum.h>
@@ -37,9 +36,6 @@ EXPORT_SYMBOL(csum_partial_copy_nocheck);
 EXPORT_SYMBOL(screen_info);
 #endif
 
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_trylock);
-EXPORT_SYMBOL(__up);
 EXPORT_SYMBOL(__put_user_asm_l);
 EXPORT_SYMBOL(__get_user_asm_l);
 EXPORT_SYMBOL(copy_page);
diff --git a/arch/sparc/kernel/Makefile b/arch/sparc/kernel/Makefile
index bf1b15d3f6f5..2712bb166f6f 100644
--- a/arch/sparc/kernel/Makefile
+++ b/arch/sparc/kernel/Makefile
@@ -12,7 +12,7 @@ obj-y    := entry.o wof.o wuf.o etrap.o rtrap.o traps.o $(IRQ_OBJS) \
 	    sys_sparc.o sunos_asm.o systbls.o \
 	    time.o windows.o cpu.o devices.o sclow.o \
 	    tadpole.o tick14.o ptrace.o sys_solaris.o \
-	    unaligned.o una_asm.o muldiv.o semaphore.o \
+	    unaligned.o una_asm.o muldiv.o \
 	    prom.o of_device.o devres.o
 
 devres-y = ../../../kernel/irq/devres.o
diff --git a/arch/sparc/kernel/semaphore.c b/arch/sparc/kernel/semaphore.c
deleted file mode 100644
index 0c37c1a7cd7e..000000000000
--- a/arch/sparc/kernel/semaphore.c
+++ /dev/null
@@ -1,155 +0,0 @@
-/* $Id: semaphore.c,v 1.7 2001/04/18 21:06:05 davem Exp $ */
-
-/* sparc32 semaphore implementation, based on i386 version */
-
-#include <linux/sched.h>
-#include <linux/errno.h>
-#include <linux/init.h>
-
-#include <asm/semaphore.h>
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to acquire the semaphore, while the "sleeping"
- * variable is a count of such acquires.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * "sleeping" and the contention routine ordering is
- * protected by the semaphore spinlock.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-
-/*
- * Logic:
- *  - only on a boundary condition do we need to care. When we go
- *    from a negative count to a non-negative, we wake people up.
- *  - when we go from a non-negative count to a negative do we
- *    (a) synchronize with the "sleeper" count and (b) make sure
- *    that we're on the wakeup list before we synchronize so that
- *    we cannot lose wakeup events.
- */
-
-void __up(struct semaphore *sem)
-{
-	wake_up(&sem->wait);
-}
-
-static DEFINE_SPINLOCK(semaphore_lock);
-
-void __sched __down(struct semaphore * sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	spin_lock_irq(&semaphore_lock);
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock.
-		 */
-		if (!atomic24_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irq(&semaphore_lock);
-
-		schedule();
-		tsk->state = TASK_UNINTERRUPTIBLE;
-		spin_lock_irq(&semaphore_lock);
-	}
-	spin_unlock_irq(&semaphore_lock);
-	remove_wait_queue(&sem->wait, &wait);
-	tsk->state = TASK_RUNNING;
-	wake_up(&sem->wait);
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	tsk->state = TASK_INTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	spin_lock_irq(&semaphore_lock);
-	sem->sleepers ++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * With signals pending, this turns into
-		 * the trylock failure case - we won't be
-		 * sleeping, and we* can't get the lock as
-		 * it has contention. Just correct the count
-		 * and exit.
-		 */
-		if (signal_pending(current)) {
-			retval = -EINTR;
-			sem->sleepers = 0;
-			atomic24_add(sleepers, &sem->count);
-			break;
-		}
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock. The
-		 * "-1" is because we're still hoping to get
-		 * the lock.
-		 */
-		if (!atomic24_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irq(&semaphore_lock);
-
-		schedule();
-		tsk->state = TASK_INTERRUPTIBLE;
-		spin_lock_irq(&semaphore_lock);
-	}
-	spin_unlock_irq(&semaphore_lock);
-	tsk->state = TASK_RUNNING;
-	remove_wait_queue(&sem->wait, &wait);
-	wake_up(&sem->wait);
-	return retval;
-}
-
-/*
- * Trylock failed - make sure we correct for
- * having decremented the count.
- */
-int __down_trylock(struct semaphore * sem)
-{
-	int sleepers;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_lock, flags);
-	sleepers = sem->sleepers + 1;
-	sem->sleepers = 0;
-
-	/*
-	 * Add "everybody else" and us into it. They aren't
-	 * playing, because we own the spinlock.
-	 */
-	if (!atomic24_add_negative(sleepers, &sem->count))
-		wake_up(&sem->wait);
-
-	spin_unlock_irqrestore(&semaphore_lock, flags);
-	return 1;
-}
diff --git a/arch/sparc/kernel/sparc_ksyms.c b/arch/sparc/kernel/sparc_ksyms.c
index c1025e551650..97b1de0e9094 100644
--- a/arch/sparc/kernel/sparc_ksyms.c
+++ b/arch/sparc/kernel/sparc_ksyms.c
@@ -107,11 +107,6 @@ EXPORT_SYMBOL(___rw_read_try);
 EXPORT_SYMBOL(___rw_read_exit);
 EXPORT_SYMBOL(___rw_write_enter);
 #endif
-/* semaphores */
-EXPORT_SYMBOL(__up);
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_trylock);
-EXPORT_SYMBOL(__down_interruptible);
 
 EXPORT_SYMBOL(sparc_valid_addr_bitmap);
 EXPORT_SYMBOL(phys_base);
diff --git a/arch/sparc64/kernel/Makefile b/arch/sparc64/kernel/Makefile
index 1bf5b187de49..459462e80a12 100644
--- a/arch/sparc64/kernel/Makefile
+++ b/arch/sparc64/kernel/Makefile
@@ -10,7 +10,7 @@ extra-y		:= head.o init_task.o vmlinux.lds
 obj-y		:= process.o setup.o cpu.o idprom.o \
 		   traps.o auxio.o una_asm.o sysfs.o iommu.o \
 		   irq.o ptrace.o time.o sys_sparc.o signal.o \
-		   unaligned.o central.o pci.o starfire.o semaphore.o \
+		   unaligned.o central.o pci.o starfire.o \
 		   power.o sbus.o sparc64_ksyms.o chmc.o \
 		   visemul.o prom.o of_device.o hvapi.o sstate.o mdesc.o
 
diff --git a/arch/sparc64/kernel/semaphore.c b/arch/sparc64/kernel/semaphore.c
deleted file mode 100644
index 9974a6899551..000000000000
--- a/arch/sparc64/kernel/semaphore.c
+++ /dev/null
@@ -1,254 +0,0 @@
-/* semaphore.c: Sparc64 semaphore implementation.
- *
- * This is basically the PPC semaphore scheme ported to use
- * the sparc64 atomic instructions, so see the PPC code for
- * credits.
- */
-
-#include <linux/sched.h>
-#include <linux/errno.h>
-#include <linux/init.h>
-
-/*
- * Atomically update sem->count.
- * This does the equivalent of the following:
- *
- *	old_count = sem->count;
- *	tmp = MAX(old_count, 0) + incr;
- *	sem->count = tmp;
- *	return old_count;
- */
-static inline int __sem_update_count(struct semaphore *sem, int incr)
-{
-	int old_count, tmp;
-
-	__asm__ __volatile__("\n"
-"	! __sem_update_count old_count(%0) tmp(%1) incr(%4) &sem->count(%3)\n"
-"1:	ldsw	[%3], %0\n"
-"	mov	%0, %1\n"
-"	cmp	%0, 0\n"
-"	movl	%%icc, 0, %1\n"
-"	add	%1, %4, %1\n"
-"	cas	[%3], %0, %1\n"
-"	cmp	%0, %1\n"
-"	membar	#StoreLoad | #StoreStore\n"
-"	bne,pn	%%icc, 1b\n"
-"	 nop\n"
-	: "=&r" (old_count), "=&r" (tmp), "=m" (sem->count)
-	: "r" (&sem->count), "r" (incr), "m" (sem->count)
-	: "cc");
-
-	return old_count;
-}
-
-static void __up(struct semaphore *sem)
-{
-	__sem_update_count(sem, 1);
-	wake_up(&sem->wait);
-}
-
-void up(struct semaphore *sem)
-{
-	/* This atomically does:
-	 * 	old_val = sem->count;
-	 *	new_val = sem->count + 1;
-	 *	sem->count = new_val;
-	 *	if (old_val < 0)
-	 *		__up(sem);
-	 *
-	 * The (old_val < 0) test is equivalent to
-	 * the more straightforward (new_val <= 0),
-	 * but it is easier to test the former because
-	 * of how the CAS instruction works.
-	 */
-
-	__asm__ __volatile__("\n"
-"	! up sem(%0)\n"
-"	membar	#StoreLoad | #LoadLoad\n"
-"1:	lduw	[%0], %%g1\n"
-"	add	%%g1, 1, %%g7\n"
-"	cas	[%0], %%g1, %%g7\n"
-"	cmp	%%g1, %%g7\n"
-"	bne,pn	%%icc, 1b\n"
-"	 addcc	%%g7, 1, %%g0\n"
-"	membar	#StoreLoad | #StoreStore\n"
-"	ble,pn	%%icc, 3f\n"
-"	 nop\n"
-"2:\n"
-"	.subsection 2\n"
-"3:	mov	%0, %%g1\n"
-"	save	%%sp, -160, %%sp\n"
-"	call	%1\n"
-"	 mov	%%g1, %%o0\n"
-"	ba,pt	%%xcc, 2b\n"
-"	 restore\n"
-"	.previous\n"
-	: : "r" (sem), "i" (__up)
-	: "g1", "g2", "g3", "g7", "memory", "cc");
-}
-
-static void __sched __down(struct semaphore * sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	while (__sem_update_count(sem, -1) <= 0) {
-		schedule();
-		tsk->state = TASK_UNINTERRUPTIBLE;
-	}
-	remove_wait_queue(&sem->wait, &wait);
-	tsk->state = TASK_RUNNING;
-
-	wake_up(&sem->wait);
-}
-
-void __sched down(struct semaphore *sem)
-{
-	might_sleep();
-	/* This atomically does:
-	 * 	old_val = sem->count;
-	 *	new_val = sem->count - 1;
-	 *	sem->count = new_val;
-	 *	if (old_val < 1)
-	 *		__down(sem);
-	 *
-	 * The (old_val < 1) test is equivalent to
-	 * the more straightforward (new_val < 0),
-	 * but it is easier to test the former because
-	 * of how the CAS instruction works.
-	 */
-
-	__asm__ __volatile__("\n"
-"	! down sem(%0)\n"
-"1:	lduw	[%0], %%g1\n"
-"	sub	%%g1, 1, %%g7\n"
-"	cas	[%0], %%g1, %%g7\n"
-"	cmp	%%g1, %%g7\n"
-"	bne,pn	%%icc, 1b\n"
-"	 cmp	%%g7, 1\n"
-"	membar	#StoreLoad | #StoreStore\n"
-"	bl,pn	%%icc, 3f\n"
-"	 nop\n"
-"2:\n"
-"	.subsection 2\n"
-"3:	mov	%0, %%g1\n"
-"	save	%%sp, -160, %%sp\n"
-"	call	%1\n"
-"	 mov	%%g1, %%o0\n"
-"	ba,pt	%%xcc, 2b\n"
-"	 restore\n"
-"	.previous\n"
-	: : "r" (sem), "i" (__down)
-	: "g1", "g2", "g3", "g7", "memory", "cc");
-}
-
-int down_trylock(struct semaphore *sem)
-{
-	int ret;
-
-	/* This atomically does:
-	 * 	old_val = sem->count;
-	 *	new_val = sem->count - 1;
-	 *	if (old_val < 1) {
-	 *		ret = 1;
-	 *	} else {
-	 *		sem->count = new_val;
-	 *		ret = 0;
-	 *	}
-	 *
-	 * The (old_val < 1) test is equivalent to
-	 * the more straightforward (new_val < 0),
-	 * but it is easier to test the former because
-	 * of how the CAS instruction works.
-	 */
-
-	__asm__ __volatile__("\n"
-"	! down_trylock sem(%1) ret(%0)\n"
-"1:	lduw	[%1], %%g1\n"
-"	sub	%%g1, 1, %%g7\n"
-"	cmp	%%g1, 1\n"
-"	bl,pn	%%icc, 2f\n"
-"	 mov	1, %0\n"
-"	cas	[%1], %%g1, %%g7\n"
-"	cmp	%%g1, %%g7\n"
-"	bne,pn	%%icc, 1b\n"
-"	 mov	0, %0\n"
-"	membar	#StoreLoad | #StoreStore\n"
-"2:\n"
-	: "=&r" (ret)
-	: "r" (sem)
-	: "g1", "g7", "memory", "cc");
-
-	return ret;
-}
-
-static int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-
-	tsk->state = TASK_INTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	while (__sem_update_count(sem, -1) <= 0) {
-		if (signal_pending(current)) {
-			__sem_update_count(sem, 0);
-			retval = -EINTR;
-			break;
-		}
-		schedule();
-		tsk->state = TASK_INTERRUPTIBLE;
-	}
-	tsk->state = TASK_RUNNING;
-	remove_wait_queue(&sem->wait, &wait);
-	wake_up(&sem->wait);
-	return retval;
-}
-
-int __sched down_interruptible(struct semaphore *sem)
-{
-	int ret = 0;
-	
-	might_sleep();
-	/* This atomically does:
-	 * 	old_val = sem->count;
-	 *	new_val = sem->count - 1;
-	 *	sem->count = new_val;
-	 *	if (old_val < 1)
-	 *		ret = __down_interruptible(sem);
-	 *
-	 * The (old_val < 1) test is equivalent to
-	 * the more straightforward (new_val < 0),
-	 * but it is easier to test the former because
-	 * of how the CAS instruction works.
-	 */
-
-	__asm__ __volatile__("\n"
-"	! down_interruptible sem(%2) ret(%0)\n"
-"1:	lduw	[%2], %%g1\n"
-"	sub	%%g1, 1, %%g7\n"
-"	cas	[%2], %%g1, %%g7\n"
-"	cmp	%%g1, %%g7\n"
-"	bne,pn	%%icc, 1b\n"
-"	 cmp	%%g7, 1\n"
-"	membar	#StoreLoad | #StoreStore\n"
-"	bl,pn	%%icc, 3f\n"
-"	 nop\n"
-"2:\n"
-"	.subsection 2\n"
-"3:	mov	%2, %%g1\n"
-"	save	%%sp, -160, %%sp\n"
-"	call	%3\n"
-"	 mov	%%g1, %%o0\n"
-"	ba,pt	%%xcc, 2b\n"
-"	 restore\n"
-"	.previous\n"
-	: "=r" (ret)
-	: "0" (ret), "r" (sem), "i" (__down_interruptible)
-	: "g1", "g2", "g3", "g7", "memory", "cc");
-	return ret;
-}
diff --git a/arch/sparc64/kernel/sparc64_ksyms.c b/arch/sparc64/kernel/sparc64_ksyms.c
index 51fa773f38c9..051b8d9cb989 100644
--- a/arch/sparc64/kernel/sparc64_ksyms.c
+++ b/arch/sparc64/kernel/sparc64_ksyms.c
@@ -130,12 +130,6 @@ EXPORT_SYMBOL(_mcount);
 
 EXPORT_SYMBOL(sparc64_get_clock_tick);
 
-/* semaphores */
-EXPORT_SYMBOL(down);
-EXPORT_SYMBOL(down_trylock);
-EXPORT_SYMBOL(down_interruptible);
-EXPORT_SYMBOL(up);
-
 /* RW semaphores */
 EXPORT_SYMBOL(__down_read);
 EXPORT_SYMBOL(__down_read_trylock);
diff --git a/arch/um/Kconfig.i386 b/arch/um/Kconfig.i386
index 3cd8a04d66d8..e09edfa560da 100644
--- a/arch/um/Kconfig.i386
+++ b/arch/um/Kconfig.i386
@@ -19,10 +19,6 @@ config 64BIT
 	bool
 	default n
 
-config SEMAPHORE_SLEEPERS
-	bool
-	default y
-
 config 3_LEVEL_PGTABLES
 	bool "Three-level pagetables (EXPERIMENTAL)"
 	default n
diff --git a/arch/um/Kconfig.x86_64 b/arch/um/Kconfig.x86_64
index 6533b349f061..3fbe69e359ed 100644
--- a/arch/um/Kconfig.x86_64
+++ b/arch/um/Kconfig.x86_64
@@ -11,10 +11,6 @@ config RWSEM_GENERIC_SPINLOCK
 	bool
 	default y
 
-config SEMAPHORE_SLEEPERS
-	bool
-	default y
-
 config 3_LEVEL_PGTABLES
        bool
        default y
diff --git a/arch/um/sys-i386/ksyms.c b/arch/um/sys-i386/ksyms.c
index 2a1eac1859ce..bfbefd30db8f 100644
--- a/arch/um/sys-i386/ksyms.c
+++ b/arch/um/sys-i386/ksyms.c
@@ -1,17 +1,5 @@
 #include "linux/module.h"
-#include "linux/in6.h"
-#include "linux/rwsem.h"
-#include "asm/byteorder.h"
-#include "asm/delay.h"
-#include "asm/semaphore.h"
-#include "asm/uaccess.h"
 #include "asm/checksum.h"
-#include "asm/errno.h"
-
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_failed_interruptible);
-EXPORT_SYMBOL(__down_failed_trylock);
-EXPORT_SYMBOL(__up_wakeup);
 
 /* Networking helper routines. */
 EXPORT_SYMBOL(csum_partial);
diff --git a/arch/um/sys-ppc/Makefile b/arch/um/sys-ppc/Makefile
index 08901526e893..b8bc844fd2c4 100644
--- a/arch/um/sys-ppc/Makefile
+++ b/arch/um/sys-ppc/Makefile
@@ -3,7 +3,7 @@ OBJ = built-in.o
 .S.o:
 	$(CC) $(KBUILD_AFLAGS) -D__ASSEMBLY__ -D__UM_PPC__ -c $< -o $*.o
 
-OBJS = ptrace.o sigcontext.o semaphore.o checksum.o miscthings.o misc.o \
+OBJS = ptrace.o sigcontext.o checksum.o miscthings.o misc.o \
 	ptrace_user.o sysrq.o
 
 EXTRA_AFLAGS := -DCONFIG_PPC32 -I. -I$(srctree)/arch/ppc/kernel
@@ -20,10 +20,6 @@ ptrace_user.o: ptrace_user.c
 sigcontext.o: sigcontext.c
 	$(CC) $(USER_CFLAGS) $(EXTRA_CFLAGS) -c -o $@ $<
 
-semaphore.c:
-	rm -f $@
-	ln -s $(srctree)/arch/ppc/kernel/$@ $@
-
 checksum.S:
 	rm -f $@
 	ln -s $(srctree)/arch/ppc/lib/$@ $@
@@ -66,4 +62,4 @@ misc.o: misc.S ppc_defs.h
 	$(CC) $(EXTRA_AFLAGS) $(KBUILD_AFLAGS) -D__ASSEMBLY__ -D__UM_PPC__ -c $< -o $*.o
 	rm -f asm
 
-clean-files := $(OBJS) ppc_defs.h checksum.S semaphore.c mk_defs.c
+clean-files := $(OBJS) ppc_defs.h checksum.S mk_defs.c
diff --git a/arch/um/sys-x86_64/ksyms.c b/arch/um/sys-x86_64/ksyms.c
index 12c593607c59..4d7d1a812d8f 100644
--- a/arch/um/sys-x86_64/ksyms.c
+++ b/arch/um/sys-x86_64/ksyms.c
@@ -1,16 +1,5 @@
 #include "linux/module.h"
-#include "linux/in6.h"
-#include "linux/rwsem.h"
-#include "asm/byteorder.h"
-#include "asm/semaphore.h"
-#include "asm/uaccess.h"
-#include "asm/checksum.h"
-#include "asm/errno.h"
-
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_failed_interruptible);
-EXPORT_SYMBOL(__down_failed_trylock);
-EXPORT_SYMBOL(__up_wakeup);
+#include "asm/string.h"
 
 /*XXX: we need them because they would be exported by x86_64 */
 EXPORT_SYMBOL(__memcpy);
diff --git a/arch/v850/kernel/Makefile b/arch/v850/kernel/Makefile
index 3930482bddc4..da5889c53576 100644
--- a/arch/v850/kernel/Makefile
+++ b/arch/v850/kernel/Makefile
@@ -11,7 +11,7 @@
 
 extra-y := head.o init_task.o vmlinux.lds
 
-obj-y += intv.o entry.o process.o syscalls.o time.o semaphore.o setup.o \
+obj-y += intv.o entry.o process.o syscalls.o time.o setup.o \
 	 signal.o irq.o mach.o ptrace.o bug.o
 obj-$(CONFIG_MODULES)		+= module.o v850_ksyms.o
 # chip-specific code
diff --git a/arch/v850/kernel/semaphore.c b/arch/v850/kernel/semaphore.c
deleted file mode 100644
index fc89fd661c99..000000000000
--- a/arch/v850/kernel/semaphore.c
+++ /dev/null
@@ -1,166 +0,0 @@
-/*
- * arch/v850/kernel/semaphore.c -- Semaphore support
- *
- *  Copyright (C) 1998-2000  IBM Corporation
- *  Copyright (C) 1999  Linus Torvalds
- *
- * This file is subject to the terms and conditions of the GNU General
- * Public License.  See the file COPYING in the main directory of this
- * archive for more details.
- *
- * This file is a copy of the s390 version, arch/s390/kernel/semaphore.c
- *    Author(s): Martin Schwidefsky
- * which was derived from the i386 version, linux/arch/i386/kernel/semaphore.c
- */
-
-#include <linux/errno.h>
-#include <linux/sched.h>
-#include <linux/init.h>
-
-#include <asm/semaphore.h>
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to acquire the semaphore, while the "sleeping"
- * variable is a count of such acquires.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * "sleeping" and the contention routine ordering is
- * protected by the semaphore spinlock.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-
-/*
- * Logic:
- *  - only on a boundary condition do we need to care. When we go
- *    from a negative count to a non-negative, we wake people up.
- *  - when we go from a non-negative count to a negative do we
- *    (a) synchronize with the "sleeper" count and (b) make sure
- *    that we're on the wakeup list before we synchronize so that
- *    we cannot lose wakeup events.
- */
-
-void __up(struct semaphore *sem)
-{
-	wake_up(&sem->wait);
-}
-
-static DEFINE_SPINLOCK(semaphore_lock);
-
-void __sched __down(struct semaphore * sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	spin_lock_irq(&semaphore_lock);
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irq(&semaphore_lock);
-
-		schedule();
-		tsk->state = TASK_UNINTERRUPTIBLE;
-		spin_lock_irq(&semaphore_lock);
-	}
-	spin_unlock_irq(&semaphore_lock);
-	remove_wait_queue(&sem->wait, &wait);
-	tsk->state = TASK_RUNNING;
-	wake_up(&sem->wait);
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	tsk->state = TASK_INTERRUPTIBLE;
-	add_wait_queue_exclusive(&sem->wait, &wait);
-
-	spin_lock_irq(&semaphore_lock);
-	sem->sleepers ++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * With signals pending, this turns into
-		 * the trylock failure case - we won't be
-		 * sleeping, and we* can't get the lock as
-		 * it has contention. Just correct the count
-		 * and exit.
-		 */
-		if (signal_pending(current)) {
-			retval = -EINTR;
-			sem->sleepers = 0;
-			atomic_add(sleepers, &sem->count);
-			break;
-		}
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock. The
-		 * "-1" is because we're still hoping to get
-		 * the lock.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irq(&semaphore_lock);
-
-		schedule();
-		tsk->state = TASK_INTERRUPTIBLE;
-		spin_lock_irq(&semaphore_lock);
-	}
-	spin_unlock_irq(&semaphore_lock);
-	tsk->state = TASK_RUNNING;
-	remove_wait_queue(&sem->wait, &wait);
-	wake_up(&sem->wait);
-	return retval;
-}
-
-/*
- * Trylock failed - make sure we correct for
- * having decremented the count.
- */
-int __down_trylock(struct semaphore * sem)
-{
-        unsigned long flags;
-	int sleepers;
-
-	spin_lock_irqsave(&semaphore_lock, flags);
-	sleepers = sem->sleepers + 1;
-	sem->sleepers = 0;
-
-	/*
-	 * Add "everybody else" and us into it. They aren't
-	 * playing, because we own the spinlock.
-	 */
-	if (!atomic_add_negative(sleepers, &sem->count))
-		wake_up(&sem->wait);
-
-	spin_unlock_irqrestore(&semaphore_lock, flags);
-	return 1;
-}
diff --git a/arch/v850/kernel/v850_ksyms.c b/arch/v850/kernel/v850_ksyms.c
index 93575fdc874d..8d386a5dbc4a 100644
--- a/arch/v850/kernel/v850_ksyms.c
+++ b/arch/v850/kernel/v850_ksyms.c
@@ -11,7 +11,6 @@
 #include <asm/pgalloc.h>
 #include <asm/irq.h>
 #include <asm/io.h>
-#include <asm/semaphore.h>
 #include <asm/checksum.h>
 #include <asm/current.h>
 
@@ -34,12 +33,6 @@ EXPORT_SYMBOL (memset);
 EXPORT_SYMBOL (memcpy);
 EXPORT_SYMBOL (memmove);
 
-/* semaphores */
-EXPORT_SYMBOL (__down);
-EXPORT_SYMBOL (__down_interruptible);
-EXPORT_SYMBOL (__down_trylock);
-EXPORT_SYMBOL (__up);
-
 /*
  * libgcc functions - functions that are used internally by the
  * compiler...  (prototypes are not correct though, but that
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 6c70fed0f9a0..e4b38861ea52 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -53,9 +53,6 @@ config STACKTRACE_SUPPORT
 config HAVE_LATENCYTOP_SUPPORT
 	def_bool y
 
-config SEMAPHORE_SLEEPERS
-	def_bool y
-
 config FAST_CMPXCHG_LOCAL
 	bool
 	default y
diff --git a/arch/x86/kernel/i386_ksyms_32.c b/arch/x86/kernel/i386_ksyms_32.c
index 061627806a2d..deb43785e923 100644
--- a/arch/x86/kernel/i386_ksyms_32.c
+++ b/arch/x86/kernel/i386_ksyms_32.c
@@ -1,13 +1,8 @@
 #include <linux/module.h>
-#include <asm/semaphore.h>
 #include <asm/checksum.h>
 #include <asm/desc.h>
 #include <asm/pgtable.h>
 
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_failed_interruptible);
-EXPORT_SYMBOL(__down_failed_trylock);
-EXPORT_SYMBOL(__up_wakeup);
 /* Networking helper routines. */
 EXPORT_SYMBOL(csum_partial_copy_generic);
 
diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c
index a66e9c1a0537..95a993e18165 100644
--- a/arch/x86/kernel/x8664_ksyms_64.c
+++ b/arch/x86/kernel/x8664_ksyms_64.c
@@ -4,7 +4,6 @@
 #include <linux/module.h>
 #include <linux/smp.h>
 
-#include <asm/semaphore.h>
 #include <asm/processor.h>
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
@@ -12,11 +11,6 @@
 
 EXPORT_SYMBOL(kernel_thread);
 
-EXPORT_SYMBOL(__down_failed);
-EXPORT_SYMBOL(__down_failed_interruptible);
-EXPORT_SYMBOL(__down_failed_trylock);
-EXPORT_SYMBOL(__up_wakeup);
-
 EXPORT_SYMBOL(__get_user_1);
 EXPORT_SYMBOL(__get_user_2);
 EXPORT_SYMBOL(__get_user_4);
diff --git a/arch/x86/lib/semaphore_32.S b/arch/x86/lib/semaphore_32.S
index 3899bd37fdf0..648fe4741782 100644
--- a/arch/x86/lib/semaphore_32.S
+++ b/arch/x86/lib/semaphore_32.S
@@ -30,89 +30,6 @@
  * value or just clobbered..
  */
 	.section .sched.text, "ax"
-ENTRY(__down_failed)
-	CFI_STARTPROC
-	FRAME
-	pushl %edx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET edx,0
-	pushl %ecx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET ecx,0
-	call __down
-	popl %ecx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE ecx
-	popl %edx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE edx
-	ENDFRAME
-	ret
-	CFI_ENDPROC
-	ENDPROC(__down_failed)
-
-ENTRY(__down_failed_interruptible)
-	CFI_STARTPROC
-	FRAME
-	pushl %edx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET edx,0
-	pushl %ecx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET ecx,0
-	call __down_interruptible
-	popl %ecx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE ecx
-	popl %edx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE edx
-	ENDFRAME
-	ret
-	CFI_ENDPROC
-	ENDPROC(__down_failed_interruptible)
-
-ENTRY(__down_failed_trylock)
-	CFI_STARTPROC
-	FRAME
-	pushl %edx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET edx,0
-	pushl %ecx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET ecx,0
-	call __down_trylock
-	popl %ecx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE ecx
-	popl %edx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE edx
-	ENDFRAME
-	ret
-	CFI_ENDPROC
-	ENDPROC(__down_failed_trylock)
-
-ENTRY(__up_wakeup)
-	CFI_STARTPROC
-	FRAME
-	pushl %edx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET edx,0
-	pushl %ecx
-	CFI_ADJUST_CFA_OFFSET 4
-	CFI_REL_OFFSET ecx,0
-	call __up
-	popl %ecx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE ecx
-	popl %edx
-	CFI_ADJUST_CFA_OFFSET -4
-	CFI_RESTORE edx
-	ENDFRAME
-	ret
-	CFI_ENDPROC
-	ENDPROC(__up_wakeup)
 
 /*
  * rw spinlock fallbacks
diff --git a/arch/x86/lib/thunk_64.S b/arch/x86/lib/thunk_64.S
index 8b92d428ab02..e009251d4e9f 100644
--- a/arch/x86/lib/thunk_64.S
+++ b/arch/x86/lib/thunk_64.S
@@ -41,11 +41,6 @@
 	thunk rwsem_downgrade_thunk,rwsem_downgrade_wake
 #endif	
 	
-	thunk __down_failed,__down
-	thunk_retrax __down_failed_interruptible,__down_interruptible
-	thunk_retrax __down_failed_trylock,__down_trylock
-	thunk __up_wakeup,__up
-
 #ifdef CONFIG_TRACE_IRQFLAGS
 	thunk trace_hardirqs_on_thunk,trace_hardirqs_on
 	thunk trace_hardirqs_off_thunk,trace_hardirqs_off
diff --git a/arch/xtensa/kernel/Makefile b/arch/xtensa/kernel/Makefile
index f582d6a24ec2..7419dbccf027 100644
--- a/arch/xtensa/kernel/Makefile
+++ b/arch/xtensa/kernel/Makefile
@@ -5,7 +5,7 @@
 extra-y := head.o vmlinux.lds
 
 
-obj-y := align.o entry.o irq.o coprocessor.o process.o ptrace.o semaphore.o  \
+obj-y := align.o entry.o irq.o coprocessor.o process.o ptrace.o \
 	 setup.o signal.o syscall.o time.o traps.o vectors.o platform.o  \
 	 pci-dma.o init_task.o io.o
 
diff --git a/arch/xtensa/kernel/semaphore.c b/arch/xtensa/kernel/semaphore.c
deleted file mode 100644
index 995c6410ae10..000000000000
--- a/arch/xtensa/kernel/semaphore.c
+++ /dev/null
@@ -1,226 +0,0 @@
-/*
- * arch/xtensa/kernel/semaphore.c
- *
- * Generic semaphore code. Buyer beware. Do your own specific changes
- * in <asm/semaphore-helper.h>
- *
- * This file is subject to the terms and conditions of the GNU General Public
- * License.  See the file "COPYING" in the main directory of this archive
- * for more details.
- *
- * Copyright (C) 2001 - 2005 Tensilica Inc.
- *
- * Joe Taylor	<joe@tensilica.com, joetylr@yahoo.com>
- * Chris Zankel	<chris@zankel.net>
- * Marc Gauthier<marc@tensilica.com, marc@alumni.uwaterloo.ca>
- * Kevin Chea
- */
-
-#include <linux/sched.h>
-#include <linux/wait.h>
-#include <linux/init.h>
-#include <asm/semaphore.h>
-#include <asm/errno.h>
-
-/*
- * These two _must_ execute atomically wrt each other.
- */
-
-static __inline__ void wake_one_more(struct semaphore * sem)
-{
-	atomic_inc((atomic_t *)&sem->sleepers);
-}
-
-static __inline__ int waking_non_zero(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->sleepers > 0) {
-		sem->sleepers--;
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_interruptible:
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- *
- * We must undo the sem->count down_interruptible() increment while we are
- * protected by the spinlock in order to make atomic this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-
-static __inline__ int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->sleepers > 0) {
-		sem->sleepers--;
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_trylock:
- *	1	failed to lock
- *	0	got the lock
- *
- * We must undo the sem->count down_trylock() increment while we are
- * protected by the spinlock in order to make atomic this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-
-static __inline__ int waking_non_zero_trylock(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 1;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->sleepers <= 0)
-		atomic_inc(&sem->count);
-	else {
-		sem->sleepers--;
-		ret = 0;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-DEFINE_SPINLOCK(semaphore_wake_lock);
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to sleep, while the "waking" variable is
- * incremented when the "up()" code goes to wake up waiting
- * processes.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * waking_non_zero() (from asm/semaphore.h) must execute
- * atomically.
- *
- * When __up() is called, the count was negative before
- * incrementing it, and we need to wake up somebody.
- *
- * This routine adds one to the count of processes that need to
- * wake up and exit.  ALL waiting processes actually wake up but
- * only the one that gets to the "waking" field first will gate
- * through and acquire the semaphore.  The others will go back
- * to sleep.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-
-void __up(struct semaphore *sem)
-{
-	wake_one_more(sem);
-	wake_up(&sem->wait);
-}
-
-/*
- * Perform the "down" function.  Return zero for semaphore acquired,
- * return negative for signalled out of the function.
- *
- * If called from __down, the return is ignored and the wait loop is
- * not interruptible.  This means that a task waiting on a semaphore
- * using "down()" cannot be killed until someone does an "up()" on
- * the semaphore.
- *
- * If called from __down_interruptible, the return value gets checked
- * upon return.  If the return value is negative then the task continues
- * with the negative value in the return register (it can be tested by
- * the caller).
- *
- * Either form may be used in conjunction with "up()".
- *
- */
-
-#define DOWN_VAR				\
-	struct task_struct *tsk = current;	\
-	wait_queue_t wait;			\
-	init_waitqueue_entry(&wait, tsk);
-
-#define DOWN_HEAD(task_state)						\
-									\
-									\
-	tsk->state = (task_state);					\
-	add_wait_queue(&sem->wait, &wait);				\
-									\
-	/*								\
-	 * Ok, we're set up.  sem->count is known to be less than zero	\
-	 * so we must wait.						\
-	 *								\
-	 * We can let go the lock for purposes of waiting.		\
-	 * We re-acquire it after awaking so as to protect		\
-	 * all semaphore operations.					\
-	 *								\
-	 * If "up()" is called before we call waking_non_zero() then	\
-	 * we will catch it right away.  If it is called later then	\
-	 * we will have to go through a wakeup cycle to catch it.	\
-	 *								\
-	 * Multiple waiters contend for the semaphore lock to see	\
-	 * who gets to gate through and who has to wait some more.	\
-	 */								\
-	for (;;) {
-
-#define DOWN_TAIL(task_state)			\
-		tsk->state = (task_state);	\
-	}					\
-	tsk->state = TASK_RUNNING;		\
-	remove_wait_queue(&sem->wait, &wait);
-
-void __sched __down(struct semaphore * sem)
-{
-	DOWN_VAR
-	DOWN_HEAD(TASK_UNINTERRUPTIBLE)
-	if (waking_non_zero(sem))
-		break;
-	schedule();
-	DOWN_TAIL(TASK_UNINTERRUPTIBLE)
-}
-
-int __sched __down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-	DOWN_VAR
-	DOWN_HEAD(TASK_INTERRUPTIBLE)
-
-	ret = waking_non_zero_interruptible(sem, tsk);
-	if (ret)
-	{
-		if (ret == 1)
-			/* ret != 0 only if we get interrupted -arca */
-			ret = 0;
-		break;
-	}
-	schedule();
-	DOWN_TAIL(TASK_INTERRUPTIBLE)
-	return ret;
-}
-
-int __down_trylock(struct semaphore * sem)
-{
-	return waking_non_zero_trylock(sem);
-}
diff --git a/arch/xtensa/kernel/xtensa_ksyms.c b/arch/xtensa/kernel/xtensa_ksyms.c
index 60dbdb43fb4c..6e52cdd6166f 100644
--- a/arch/xtensa/kernel/xtensa_ksyms.c
+++ b/arch/xtensa/kernel/xtensa_ksyms.c
@@ -26,7 +26,6 @@
 #include <asm/io.h>
 #include <asm/page.h>
 #include <asm/pgalloc.h>
-#include <asm/semaphore.h>
 #ifdef CONFIG_BLK_DEV_FD
 #include <asm/floppy.h>
 #endif
@@ -71,14 +70,6 @@ EXPORT_SYMBOL(__umodsi3);
 EXPORT_SYMBOL(__udivdi3);
 EXPORT_SYMBOL(__umoddi3);
 
-/*
- * Semaphore operations
- */
-EXPORT_SYMBOL(__down);
-EXPORT_SYMBOL(__down_interruptible);
-EXPORT_SYMBOL(__down_trylock);
-EXPORT_SYMBOL(__up);
-
 #ifdef CONFIG_NET
 /*
  * Networking support
diff --git a/include/asm-alpha/semaphore.h b/include/asm-alpha/semaphore.h
index f1e9278a9fe2..d9b2034ed1d2 100644
--- a/include/asm-alpha/semaphore.h
+++ b/include/asm-alpha/semaphore.h
@@ -1,149 +1 @@
-#ifndef _ALPHA_SEMAPHORE_H
-#define _ALPHA_SEMAPHORE_H
-
-/*
- * SMP- and interrupt-safe semaphores..
- *
- * (C) Copyright 1996 Linus Torvalds
- * (C) Copyright 1996, 2000 Richard Henderson
- */
-
-#include <asm/current.h>
-#include <asm/system.h>
-#include <asm/atomic.h>
-#include <linux/compiler.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	atomic_t count;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)			\
-{								\
-	.count	= ATOMIC_INIT(n),				\
-  	.wait	= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait),	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count)		\
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name)		__DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init(struct semaphore *sem, int val)
-{
-	/*
-	 * Logically, 
-	 *   *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
-	 * except that gcc produces better initializing by parts yet.
-	 */
-
-	atomic_set(&sem->count, val);
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void down(struct semaphore *);
-extern void __down_failed(struct semaphore *);
-extern int  down_interruptible(struct semaphore *);
-extern int  __down_failed_interruptible(struct semaphore *);
-extern int  down_trylock(struct semaphore *);
-extern void up(struct semaphore *);
-extern void __up_wakeup(struct semaphore *);
-
-/*
- * Hidden out of line code is fun, but extremely messy.  Rely on newer
- * compilers to do a respectable job with this.  The contention cases
- * are handled out of line in arch/alpha/kernel/semaphore.c.
- */
-
-static inline void __down(struct semaphore *sem)
-{
-	long count;
-	might_sleep();
-	count = atomic_dec_return(&sem->count);
-	if (unlikely(count < 0))
-		__down_failed(sem);
-}
-
-static inline int __down_interruptible(struct semaphore *sem)
-{
-	long count;
-	might_sleep();
-	count = atomic_dec_return(&sem->count);
-	if (unlikely(count < 0))
-		return __down_failed_interruptible(sem);
-	return 0;
-}
-
-/*
- * down_trylock returns 0 on success, 1 if we failed to get the lock.
- */
-
-static inline int __down_trylock(struct semaphore *sem)
-{
-	long ret;
-
-	/* "Equivalent" C:
-
-	   do {
-		ret = ldl_l;
-		--ret;
-		if (ret < 0)
-			break;
-		ret = stl_c = ret;
-	   } while (ret == 0);
-	*/
-	__asm__ __volatile__(
-		"1:	ldl_l	%0,%1\n"
-		"	subl	%0,1,%0\n"
-		"	blt	%0,2f\n"
-		"	stl_c	%0,%1\n"
-		"	beq	%0,3f\n"
-		"	mb\n"
-		"2:\n"
-		".subsection 2\n"
-		"3:	br	1b\n"
-		".previous"
-		: "=&r" (ret), "=m" (sem->count)
-		: "m" (sem->count));
-
-	return ret < 0;
-}
-
-static inline void __up(struct semaphore *sem)
-{
-	if (unlikely(atomic_inc_return(&sem->count) <= 0))
-		__up_wakeup(sem);
-}
-
-#if !defined(CONFIG_DEBUG_SEMAPHORE)
-extern inline void down(struct semaphore *sem)
-{
-	__down(sem);
-}
-extern inline int down_interruptible(struct semaphore *sem)
-{
-	return __down_interruptible(sem);
-}
-extern inline int down_trylock(struct semaphore *sem)
-{
-	return __down_trylock(sem);
-}
-extern inline void up(struct semaphore *sem)
-{
-	__up(sem);
-}
-#endif
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-arm/semaphore-helper.h b/include/asm-arm/semaphore-helper.h
deleted file mode 100644
index 1d7f1987edb9..000000000000
--- a/include/asm-arm/semaphore-helper.h
+++ /dev/null
@@ -1,84 +0,0 @@
-#ifndef ASMARM_SEMAPHORE_HELPER_H
-#define ASMARM_SEMAPHORE_HELPER_H
-
-/*
- * These two _must_ execute atomically wrt each other.
- */
-static inline void wake_one_more(struct semaphore * sem)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (atomic_read(&sem->count) <= 0)
-		sem->waking++;
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-}
-
-static inline int waking_non_zero(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->waking > 0) {
-		sem->waking--;
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking non zero interruptible
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- *
- * We must undo the sem->count down_interruptible() increment while we are
- * protected by the spinlock in order to make this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-static inline int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->waking > 0) {
-		sem->waking--;
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;	
-}
-
-/*
- * waking_non_zero_try_lock:
- *	1	failed to lock
- *	0	got the lock
- *
- * We must undo the sem->count down_interruptible() increment while we are
- * protected by the spinlock in order to make this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-static inline int waking_non_zero_trylock(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 1;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->waking <= 0)
-		atomic_inc(&sem->count);
-	else {
-		sem->waking--;
-		ret = 0;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-#endif
diff --git a/include/asm-arm/semaphore.h b/include/asm-arm/semaphore.h
index 1c8b441f89e3..d9b2034ed1d2 100644
--- a/include/asm-arm/semaphore.h
+++ b/include/asm-arm/semaphore.h
@@ -1,98 +1 @@
-/*
- * linux/include/asm-arm/semaphore.h
- */
-#ifndef __ASM_ARM_SEMAPHORE_H
-#define __ASM_ARM_SEMAPHORE_H
-
-#include <linux/linkage.h>
-#include <linux/spinlock.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-#include <asm/atomic.h>
-#include <asm/locks.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INIT(name, cnt)				\
-{								\
-	.count	= ATOMIC_INIT(cnt),				\
-	.wait	= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait),	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count)	\
-	struct semaphore name = __SEMAPHORE_INIT(name,count)
-
-#define DECLARE_MUTEX(name)		__DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init(struct semaphore *sem, int val)
-{
-	atomic_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX(struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED(struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-/*
- * special register calling convention
- */
-asmlinkage void __down_failed(void);
-asmlinkage int  __down_interruptible_failed(void);
-asmlinkage int  __down_trylock_failed(void);
-asmlinkage void __up_wakeup(void);
-
-extern void __down(struct semaphore * sem);
-extern int  __down_interruptible(struct semaphore * sem);
-extern int  __down_trylock(struct semaphore * sem);
-extern void __up(struct semaphore * sem);
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "__down" is the actual routine that waits...
- */
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	__down_op(sem, __down_failed);
-}
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "__down_interruptible" is the actual routine that waits...
- */
-static inline int down_interruptible (struct semaphore * sem)
-{
-	might_sleep();
-	return __down_op_ret(sem, __down_interruptible_failed);
-}
-
-static inline int down_trylock(struct semaphore *sem)
-{
-	return __down_op_ret(sem, __down_trylock_failed);
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore * sem)
-{
-	__up_op(sem, __up_wakeup);
-}
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-avr32/semaphore.h b/include/asm-avr32/semaphore.h
index feaf1d453386..d9b2034ed1d2 100644
--- a/include/asm-avr32/semaphore.h
+++ b/include/asm-avr32/semaphore.h
@@ -1,108 +1 @@
-/*
- * SMP- and interrupt-safe semaphores.
- *
- * Copyright (C) 2006 Atmel Corporation
- *
- * Based on include/asm-i386/semaphore.h
- *   Copyright (C) 1996 Linus Torvalds
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-#ifndef __ASM_AVR32_SEMAPHORE_H
-#define __ASM_AVR32_SEMAPHORE_H
-
-#include <linux/linkage.h>
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	atomic_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-void __down(struct semaphore * sem);
-int  __down_interruptible(struct semaphore * sem);
-void __up(struct semaphore * sem);
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "__down_failed" is a special asm handler that calls the C
- * routine that actually waits. See arch/i386/kernel/semaphore.c
- */
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	if (unlikely(atomic_dec_return (&sem->count) < 0))
-		__down (sem);
-}
-
-/*
- * Interruptible try to acquire a semaphore.  If we obtained
- * it, return zero.  If we were interrupted, returns -EINTR
- */
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-
-	might_sleep();
-	if (unlikely(atomic_dec_return (&sem->count) < 0))
-		ret = __down_interruptible (sem);
-	return ret;
-}
-
-/*
- * Non-blockingly attempt to down() a semaphore.
- * Returns zero if we acquired it
- */
-static inline int down_trylock(struct semaphore * sem)
-{
-	return atomic_dec_if_positive(&sem->count) < 0;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore * sem)
-{
-	if (unlikely(atomic_inc_return (&sem->count) <= 0))
-		__up (sem);
-}
-
-#endif /*__ASM_AVR32_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-blackfin/semaphore-helper.h b/include/asm-blackfin/semaphore-helper.h
deleted file mode 100644
index 9082b0dc3eb5..000000000000
--- a/include/asm-blackfin/semaphore-helper.h
+++ /dev/null
@@ -1,82 +0,0 @@
-/* Based on M68K version,	Lineo Inc.	May 2001 */
-
-#ifndef _BFIN_SEMAPHORE_HELPER_H
-#define _BFIN_SEMAPHORE_HELPER_H
-
-/*
- * SMP- and interrupt-safe semaphores helper functions.
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- */
-
-#include <asm/errno.h>
-
-/*
- * These two _must_ execute atomically wrt each other.
- */
-static inline void wake_one_more(struct semaphore *sem)
-{
-	atomic_inc(&sem->waking);
-}
-
-static inline int waking_non_zero(struct semaphore *sem)
-{
-	int ret;
-	unsigned long flags = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 0;
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_interruptible:
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- */
-static inline int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	int ret = 0;
-	unsigned long flags = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_trylock:
- *	1	failed to lock
- *	0	got the lock
- */
-static inline int waking_non_zero_trylock(struct semaphore *sem)
-{
-	int ret = 1;
-	unsigned long flags = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 0;
-	} else
-		atomic_inc(&sem->count);
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-#endif				/* _BFIN_SEMAPHORE_HELPER_H */
diff --git a/include/asm-blackfin/semaphore.h b/include/asm-blackfin/semaphore.h
index 533f90fb2e4e..d9b2034ed1d2 100644
--- a/include/asm-blackfin/semaphore.h
+++ b/include/asm-blackfin/semaphore.h
@@ -1,105 +1 @@
-#ifndef _BFIN_SEMAPHORE_H
-#define _BFIN_SEMAPHORE_H
-
-#ifndef __ASSEMBLY__
-
-#include <linux/linkage.h>
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-#include <asm/atomic.h>
-
-/*
- * Interrupt-safe semaphores..
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * BFIN version by akbar hussain Lineo Inc  April 2001
- *
- */
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init(struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val);
-}
-
-static inline void init_MUTEX(struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED(struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down(struct semaphore *sem);
-asmlinkage int __down_interruptible(struct semaphore *sem);
-asmlinkage int __down_trylock(struct semaphore *sem);
-asmlinkage void __up(struct semaphore *sem);
-
-extern spinlock_t semaphore_wake_lock;
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "down_failed" is a special asm handler that calls the C
- * routine that actually waits.
- */
-static inline void down(struct semaphore *sem)
-{
-	might_sleep();
-	if (atomic_dec_return(&sem->count) < 0)
-		__down(sem);
-}
-
-static inline int down_interruptible(struct semaphore *sem)
-{
-	int ret = 0;
-
-	might_sleep();
-	if (atomic_dec_return(&sem->count) < 0)
-		ret = __down_interruptible(sem);
-	return (ret);
-}
-
-static inline int down_trylock(struct semaphore *sem)
-{
-	int ret = 0;
-
-	if (atomic_dec_return(&sem->count) < 0)
-		ret = __down_trylock(sem);
-	return ret;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore *sem)
-{
-	if (atomic_inc_return(&sem->count) <= 0)
-		__up(sem);
-}
-
-#endif				/* __ASSEMBLY__ */
-#endif				/* _BFIN_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-cris/semaphore-helper.h b/include/asm-cris/semaphore-helper.h
deleted file mode 100644
index 27bfeca1b981..000000000000
--- a/include/asm-cris/semaphore-helper.h
+++ /dev/null
@@ -1,78 +0,0 @@
-/* $Id: semaphore-helper.h,v 1.3 2001/03/26 15:00:33 orjanf Exp $
- *
- * SMP- and interrupt-safe semaphores helper functions. Generic versions, no
- * optimizations whatsoever... 
- *
- */
-
-#ifndef _ASM_SEMAPHORE_HELPER_H
-#define _ASM_SEMAPHORE_HELPER_H
-
-#include <asm/atomic.h>
-#include <linux/errno.h>
-
-#define read(a) ((a)->counter)
-#define inc(a) (((a)->counter)++)
-#define dec(a) (((a)->counter)--)
-
-#define count_inc(a) ((*(a))++)
-
-/*
- * These two _must_ execute atomically wrt each other.
- */
-static inline void wake_one_more(struct semaphore * sem)
-{
-	atomic_inc(&sem->waking);
-}
-
-static inline int waking_non_zero(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	local_irq_save(flags);
-	if (read(&sem->waking) > 0) {
-		dec(&sem->waking);
-		ret = 1;
-	}
-	local_irq_restore(flags);
-	return ret;
-}
-
-static inline int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	int ret = 0;
-	unsigned long flags;
-
-	local_irq_save(flags);
-	if (read(&sem->waking) > 0) {
-		dec(&sem->waking);
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		inc(&sem->count);
-		ret = -EINTR;
-	}
-	local_irq_restore(flags);
-	return ret;
-}
-
-static inline int waking_non_zero_trylock(struct semaphore *sem)
-{
-        int ret = 1;
-	unsigned long flags;
-
-	local_irq_save(flags);
-	if (read(&sem->waking) <= 0)
-		inc(&sem->count);
-	else {
-		dec(&sem->waking);
-		ret = 0;
-	}
-	local_irq_restore(flags);
-	return ret;
-}
-
-#endif /* _ASM_SEMAPHORE_HELPER_H */
-
-
diff --git a/include/asm-cris/semaphore.h b/include/asm-cris/semaphore.h
index 31a4ac448195..d9b2034ed1d2 100644
--- a/include/asm-cris/semaphore.h
+++ b/include/asm-cris/semaphore.h
@@ -1,133 +1 @@
-/* $Id: semaphore.h,v 1.3 2001/05/08 13:54:09 bjornw Exp $ */
-
-/* On the i386 these are coded in asm, perhaps we should as well. Later.. */
-
-#ifndef _CRIS_SEMAPHORE_H
-#define _CRIS_SEMAPHORE_H
-
-#define RW_LOCK_BIAS             0x01000000
-
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-
-/*
- * CRIS semaphores, implemented in C-only so far. 
- */
-
-struct semaphore {
-	atomic_t count;
-	atomic_t waking;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.waking		= ATOMIC_INIT(0),				\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)    \
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-        struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init(struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-        sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-        sema_init(sem, 0);
-}
-
-extern void __down(struct semaphore * sem);
-extern int __down_interruptible(struct semaphore * sem);
-extern int __down_trylock(struct semaphore * sem);
-extern void __up(struct semaphore * sem);
-
-/* notice - we probably can do cli/sti here instead of saving */
-
-static inline void down(struct semaphore * sem)
-{
-	unsigned long flags;
-	int failed;
-
-	might_sleep();
-
-	/* atomically decrement the semaphores count, and if its negative, we wait */
-	cris_atomic_save(sem, flags);
-	failed = --(sem->count.counter) < 0;
-	cris_atomic_restore(sem, flags);
-	if(failed) {
-		__down(sem);
-	}
-}
-
-/*
- * This version waits in interruptible state so that the waiting
- * process can be killed.  The down_interruptible routine
- * returns negative for signalled and zero for semaphore acquired.
- */
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	unsigned long flags;
-	int failed;
-
-	might_sleep();
-
-	/* atomically decrement the semaphores count, and if its negative, we wait */
-	cris_atomic_save(sem, flags);
-	failed = --(sem->count.counter) < 0;
-	cris_atomic_restore(sem, flags);
-	if(failed)
-		failed = __down_interruptible(sem);
-	return(failed);
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	unsigned long flags;
-	int failed;
-
-	cris_atomic_save(sem, flags);
-	failed = --(sem->count.counter) < 0;
-	cris_atomic_restore(sem, flags);
-	if(failed)
-		failed = __down_trylock(sem);
-	return(failed);
-
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore * sem)
-{  
-	unsigned long flags;
-	int wakeup;
-
-	/* atomically increment the semaphores count, and if it was negative, we wake people */
-	cris_atomic_save(sem, flags);
-	wakeup = ++(sem->count.counter) <= 0;
-	cris_atomic_restore(sem, flags);
-	if(wakeup) {
-		__up(sem);
-	}
-}
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-frv/semaphore.h b/include/asm-frv/semaphore.h
index d7aaa1911a1a..d9b2034ed1d2 100644
--- a/include/asm-frv/semaphore.h
+++ b/include/asm-frv/semaphore.h
@@ -1,155 +1 @@
-/* semaphore.h: semaphores for the FR-V
- *
- * Copyright (C) 2003 Red Hat, Inc. All Rights Reserved.
- * Written by David Howells (dhowells@redhat.com)
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version
- * 2 of the License, or (at your option) any later version.
- */
-#ifndef _ASM_SEMAPHORE_H
-#define _ASM_SEMAPHORE_H
-
-#define RW_LOCK_BIAS		 0x01000000
-
-#ifndef __ASSEMBLY__
-
-#include <linux/linkage.h>
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-
-/*
- * the semaphore definition
- * - if counter is >0 then there are tokens available on the semaphore for down to collect
- * - if counter is <=0 then there are no spare tokens, and anyone that wants one must wait
- * - if wait_list is not empty, then there are processes waiting for the semaphore
- */
-struct semaphore {
-	unsigned		counter;
-	spinlock_t		wait_lock;
-	struct list_head	wait_list;
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	unsigned		__magic;
-#endif
-};
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-# define __SEM_DEBUG_INIT(name) , (long)&(name).__magic
-#else
-# define __SEM_DEBUG_INIT(name)
-#endif
-
-
-#define __SEMAPHORE_INITIALIZER(name,count) \
-{ count, SPIN_LOCK_UNLOCKED, LIST_HEAD_INIT((name).wait_list) __SEM_DEBUG_INIT(name) }
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void __down(struct semaphore *sem, unsigned long flags);
-extern int  __down_interruptible(struct semaphore *sem, unsigned long flags);
-extern void __up(struct semaphore *sem);
-
-static inline void down(struct semaphore *sem)
-{
-	unsigned long flags;
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	if (likely(sem->counter > 0)) {
-		sem->counter--;
-		spin_unlock_irqrestore(&sem->wait_lock, flags);
-	}
-	else {
-		__down(sem, flags);
-	}
-}
-
-static inline int down_interruptible(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 0;
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	if (likely(sem->counter > 0)) {
-		sem->counter--;
-		spin_unlock_irqrestore(&sem->wait_lock, flags);
-	}
-	else {
-		ret = __down_interruptible(sem, flags);
-	}
-	return ret;
-}
-
-/*
- * non-blockingly attempt to down() a semaphore.
- * - returns zero if we acquired it
- */
-static inline int down_trylock(struct semaphore *sem)
-{
-	unsigned long flags;
-	int success = 0;
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	if (sem->counter > 0) {
-		sem->counter--;
-		success = 1;
-	}
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-	return !success;
-}
-
-static inline void up(struct semaphore *sem)
-{
-	unsigned long flags;
-
-#ifdef CONFIG_DEBUG_SEMAPHORE
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	if (!list_empty(&sem->wait_list))
-		__up(sem);
-	else
-		sem->counter++;
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-}
-
-static inline int sem_getcount(struct semaphore *sem)
-{
-	return sem->counter;
-}
-
-#endif /* __ASSEMBLY__ */
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-h8300/semaphore-helper.h b/include/asm-h8300/semaphore-helper.h
deleted file mode 100644
index 4fea36be5fd8..000000000000
--- a/include/asm-h8300/semaphore-helper.h
+++ /dev/null
@@ -1,85 +0,0 @@
-#ifndef _H8300_SEMAPHORE_HELPER_H
-#define _H8300_SEMAPHORE_HELPER_H
-
-/*
- * SMP- and interrupt-safe semaphores helper functions.
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * based on
- * m68k version by Andreas Schwab
- */
-
-#include <linux/errno.h>
-
-/*
- * These two _must_ execute atomically wrt each other.
- */
-static inline void wake_one_more(struct semaphore * sem)
-{
-	atomic_inc((atomic_t *)&sem->sleepers);
-}
-
-static inline int waking_non_zero(struct semaphore *sem)
-{
-	int ret;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 0;
-	if (sem->sleepers > 0) {
-		sem->sleepers--;
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_interruptible:
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- */
-static inline int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	int ret;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 0;
-	if (sem->sleepers > 0) {
-		sem->sleepers--;
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_trylock:
- *	1	failed to lock
- *	0	got the lock
- */
-static inline int waking_non_zero_trylock(struct semaphore *sem)
-{
-	int ret;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 1;
-	if (sem->sleepers <= 0)
-		atomic_inc(&sem->count);
-	else {
-		sem->sleepers--;
-		ret = 0;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-#endif
diff --git a/include/asm-h8300/semaphore.h b/include/asm-h8300/semaphore.h
index f3ffff83ff09..d9b2034ed1d2 100644
--- a/include/asm-h8300/semaphore.h
+++ b/include/asm-h8300/semaphore.h
@@ -1,190 +1 @@
-#ifndef _H8300_SEMAPHORE_H
-#define _H8300_SEMAPHORE_H
-
-#define RW_LOCK_BIAS		 0x01000000
-
-#ifndef __ASSEMBLY__
-
-#include <linux/linkage.h>
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-
-/*
- * Interrupt-safe semaphores..
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * H8/300 version by Yoshinori Sato
- */
-
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down_failed(void /* special register calling convention */);
-asmlinkage int  __down_failed_interruptible(void  /* params in registers */);
-asmlinkage int  __down_failed_trylock(void  /* params in registers */);
-asmlinkage void __up_wakeup(void /* special register calling convention */);
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-extern spinlock_t semaphore_wake_lock;
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "down_failed" is a special asm handler that calls the C
- * routine that actually waits. See arch/m68k/lib/semaphore.S
- */
-static inline void down(struct semaphore * sem)
-{
-	register atomic_t *count asm("er0");
-
-	might_sleep();
-
-	count = &(sem->count);
-	__asm__ __volatile__(
-		"stc ccr,r3l\n\t"
-		"orc #0x80,ccr\n\t"
-		"mov.l %2, er1\n\t"
-		"dec.l #1,er1\n\t"
-		"mov.l er1,%0\n\t"
-		"bpl 1f\n\t"
-		"ldc r3l,ccr\n\t"
-		"mov.l %1,er0\n\t"
-		"jsr @___down\n\t"
-		"bra 2f\n"
-		"1:\n\t"
-		"ldc r3l,ccr\n"
-		"2:"
-		: "=m"(*count)
-		: "g"(sem),"m"(*count)
-		: "cc",  "er1", "er2", "er3");
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	register atomic_t *count asm("er0");
-
-	might_sleep();
-
-	count = &(sem->count);
-	__asm__ __volatile__(
-		"stc ccr,r1l\n\t"
-		"orc #0x80,ccr\n\t"
-		"mov.l %3, er2\n\t"
-		"dec.l #1,er2\n\t"
-		"mov.l er2,%1\n\t"
-		"bpl 1f\n\t"
-		"ldc r1l,ccr\n\t"
-		"mov.l %2,er0\n\t"
-		"jsr @___down_interruptible\n\t"
-		"bra 2f\n"
-		"1:\n\t"
-		"ldc r1l,ccr\n\t"
-		"sub.l %0,%0\n\t"
-		"2:\n\t"
-		: "=r" (count),"=m" (*count)
-		: "g"(sem),"m"(*count)
-		: "cc", "er1", "er2", "er3");
-	return (int)count;
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	register atomic_t *count asm("er0");
-
-	count = &(sem->count);
-	__asm__ __volatile__(
-		"stc ccr,r3l\n\t"
-		"orc #0x80,ccr\n\t"
-		"mov.l %3,er2\n\t"
-		"dec.l #1,er2\n\t"
-		"mov.l er2,%0\n\t"
-		"bpl 1f\n\t"
-		"ldc r3l,ccr\n\t"
-		"jmp @3f\n\t"
-		LOCK_SECTION_START(".align 2\n\t")
-		"3:\n\t"
-		"mov.l %2,er0\n\t"
-		"jsr @___down_trylock\n\t"
-		"jmp @2f\n\t"
-		LOCK_SECTION_END
-		"1:\n\t"
-		"ldc r3l,ccr\n\t"
-		"sub.l %1,%1\n"
-		"2:"
-		: "=m" (*count),"=r"(count)
-		: "g"(sem),"m"(*count)
-		: "cc", "er1","er2", "er3");
-	return (int)count;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore * sem)
-{
-	register atomic_t *count asm("er0");
-
-	count = &(sem->count);
-	__asm__ __volatile__(
-		"stc ccr,r3l\n\t"
-		"orc #0x80,ccr\n\t"
-		"mov.l %2,er1\n\t"
-		"inc.l #1,er1\n\t"
-		"mov.l er1,%0\n\t"
-		"ldc r3l,ccr\n\t"
-		"sub.l er2,er2\n\t"
-		"cmp.l er2,er1\n\t"
-		"bgt 1f\n\t"
-		"mov.l %1,er0\n\t"
-		"jsr @___up\n"
-		"1:"
-		: "=m"(*count)
-		: "g"(sem),"m"(*count)
-		: "cc", "er1", "er2", "er3");
-}
-
-#endif /* __ASSEMBLY__ */
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-ia64/semaphore.h b/include/asm-ia64/semaphore.h
index d8393d11288d..d9b2034ed1d2 100644
--- a/include/asm-ia64/semaphore.h
+++ b/include/asm-ia64/semaphore.h
@@ -1,99 +1 @@
-#ifndef _ASM_IA64_SEMAPHORE_H
-#define _ASM_IA64_SEMAPHORE_H
-
-/*
- * Copyright (C) 1998-2000 Hewlett-Packard Co
- * Copyright (C) 1998-2000 David Mosberger-Tang <davidm@hpl.hp.com>
- */
-
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-#include <asm/atomic.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count)					\
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name, count)
-
-#define DECLARE_MUTEX(name)		__DECLARE_SEMAPHORE_GENERIC(name, 1)
-
-static inline void
-sema_init (struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val);
-}
-
-static inline void
-init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void
-init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void __down (struct semaphore * sem);
-extern int  __down_interruptible (struct semaphore * sem);
-extern int  __down_trylock (struct semaphore * sem);
-extern void __up (struct semaphore * sem);
-
-/*
- * Atomically decrement the semaphore's count.  If it goes negative,
- * block the calling thread in the TASK_UNINTERRUPTIBLE state.
- */
-static inline void
-down (struct semaphore *sem)
-{
-	might_sleep();
-	if (ia64_fetchadd(-1, &sem->count.counter, acq) < 1)
-		__down(sem);
-}
-
-/*
- * Atomically decrement the semaphore's count.  If it goes negative,
- * block the calling thread in the TASK_INTERRUPTIBLE state.
- */
-static inline int
-down_interruptible (struct semaphore * sem)
-{
-	int ret = 0;
-
-	might_sleep();
-	if (ia64_fetchadd(-1, &sem->count.counter, acq) < 1)
-		ret = __down_interruptible(sem);
-	return ret;
-}
-
-static inline int
-down_trylock (struct semaphore *sem)
-{
-	int ret = 0;
-
-	if (ia64_fetchadd(-1, &sem->count.counter, acq) < 1)
-		ret = __down_trylock(sem);
-	return ret;
-}
-
-static inline void
-up (struct semaphore * sem)
-{
-	if (ia64_fetchadd(1, &sem->count.counter, rel) <= -1)
-		__up(sem);
-}
-
-#endif /* _ASM_IA64_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-m32r/semaphore.h b/include/asm-m32r/semaphore.h
index b5bf95a6f2b4..d9b2034ed1d2 100644
--- a/include/asm-m32r/semaphore.h
+++ b/include/asm-m32r/semaphore.h
@@ -1,144 +1 @@
-#ifndef _ASM_M32R_SEMAPHORE_H
-#define _ASM_M32R_SEMAPHORE_H
-
-#include <linux/linkage.h>
-
-#ifdef __KERNEL__
-
-/*
- * SMP- and interrupt-safe semaphores..
- *
- * Copyright (C) 1996  Linus Torvalds
- * Copyright (C) 2004, 2006  Hirokazu Takata <takata at linux-m32r.org>
- */
-
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-#include <asm/assembler.h>
-#include <asm/system.h>
-#include <asm/atomic.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-/*
- *	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
- *
- * i'd rather use the more flexible initialization above, but sadly
- * GCC 2.7.2.3 emits a bogus warning. EGCS doesnt. Oh well.
- */
-	atomic_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down_failed(void /* special register calling convention */);
-asmlinkage int  __down_failed_interruptible(void  /* params in registers */);
-asmlinkage int  __down_failed_trylock(void  /* params in registers */);
-asmlinkage void __up_wakeup(void /* special register calling convention */);
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-/*
- * Atomically decrement the semaphore's count.  If it goes negative,
- * block the calling thread in the TASK_UNINTERRUPTIBLE state.
- */
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	if (unlikely(atomic_dec_return(&sem->count) < 0))
-		__down(sem);
-}
-
-/*
- * Interruptible try to acquire a semaphore.  If we obtained
- * it, return zero.  If we were interrupted, returns -EINTR
- */
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int result = 0;
-
-	might_sleep();
-	if (unlikely(atomic_dec_return(&sem->count) < 0))
-		result = __down_interruptible(sem);
-
-	return result;
-}
-
-/*
- * Non-blockingly attempt to down() a semaphore.
- * Returns zero if we acquired it
- */
-static inline int down_trylock(struct semaphore * sem)
-{
-	unsigned long flags;
-	long count;
-	int result = 0;
-
-	local_irq_save(flags);
-	__asm__ __volatile__ (
-		"# down_trylock			\n\t"
-		DCACHE_CLEAR("%0", "r4", "%1")
-		M32R_LOCK" %0, @%1;		\n\t"
-		"addi	%0, #-1;		\n\t"
-		M32R_UNLOCK" %0, @%1;		\n\t"
-		: "=&r" (count)
-		: "r" (&sem->count)
-		: "memory"
-#ifdef CONFIG_CHIP_M32700_TS1
-		, "r4"
-#endif	/* CONFIG_CHIP_M32700_TS1 */
-	);
-	local_irq_restore(flags);
-
-	if (unlikely(count < 0))
-		result = __down_trylock(sem);
-
-	return result;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore * sem)
-{
-	if (unlikely(atomic_inc_return(&sem->count) <= 0))
-		__up(sem);
-}
-
-#endif  /* __KERNEL__ */
-
-#endif  /* _ASM_M32R_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-m68k/semaphore-helper.h b/include/asm-m68k/semaphore-helper.h
deleted file mode 100644
index eef30ba0b499..000000000000
--- a/include/asm-m68k/semaphore-helper.h
+++ /dev/null
@@ -1,142 +0,0 @@
-#ifndef _M68K_SEMAPHORE_HELPER_H
-#define _M68K_SEMAPHORE_HELPER_H
-
-/*
- * SMP- and interrupt-safe semaphores helper functions.
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * m68k version by Andreas Schwab
- */
-
-#include <linux/errno.h>
-
-/*
- * These two _must_ execute atomically wrt each other.
- */
-static inline void wake_one_more(struct semaphore * sem)
-{
-	atomic_inc(&sem->waking);
-}
-
-#ifndef CONFIG_RMW_INSNS
-extern spinlock_t semaphore_wake_lock;
-#endif
-
-static inline int waking_non_zero(struct semaphore *sem)
-{
-	int ret;
-#ifndef CONFIG_RMW_INSNS
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 0;
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-#else
-	int tmp1, tmp2;
-
-	__asm__ __volatile__
-	  ("1:	movel	%1,%2\n"
-	   "    jle	2f\n"
-	   "	subql	#1,%2\n"
-	   "	casl	%1,%2,%3\n"
-	   "	jne	1b\n"
-	   "	moveq	#1,%0\n"
-	   "2:"
-	   : "=d" (ret), "=d" (tmp1), "=d" (tmp2)
-	   : "m" (sem->waking), "0" (0), "1" (sem->waking));
-#endif
-
-	return ret;
-}
-
-/*
- * waking_non_zero_interruptible:
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- */
-static inline int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	int ret;
-#ifndef CONFIG_RMW_INSNS
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 0;
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-#else
-	int tmp1, tmp2;
-
-	__asm__ __volatile__
-	  ("1:	movel	%1,%2\n"
-	   "	jle	2f\n"
-	   "	subql	#1,%2\n"
-	   "	casl	%1,%2,%3\n"
-	   "	jne	1b\n"
-	   "	moveq	#1,%0\n"
-	   "	jra	%a4\n"
-	   "2:"
-	   : "=d" (ret), "=d" (tmp1), "=d" (tmp2)
-	   : "m" (sem->waking), "i" (&&next), "0" (0), "1" (sem->waking));
-	if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-next:
-#endif
-
-	return ret;
-}
-
-/*
- * waking_non_zero_trylock:
- *	1	failed to lock
- *	0	got the lock
- */
-static inline int waking_non_zero_trylock(struct semaphore *sem)
-{
-	int ret;
-#ifndef CONFIG_RMW_INSNS
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 1;
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 0;
-	} else
-		atomic_inc(&sem->count);
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-#else
-	int tmp1, tmp2;
-
-	__asm__ __volatile__
-	  ("1:	movel	%1,%2\n"
-	   "    jle	2f\n"
-	   "	subql	#1,%2\n"
-	   "	casl	%1,%2,%3\n"
-	   "	jne	1b\n"
-	   "	moveq	#0,%0\n"
-	   "2:"
-	   : "=d" (ret), "=d" (tmp1), "=d" (tmp2)
-	   : "m" (sem->waking), "0" (1), "1" (sem->waking));
-	if (ret)
-		atomic_inc(&sem->count);
-#endif
-	return ret;
-}
-
-#endif
diff --git a/include/asm-m68k/semaphore.h b/include/asm-m68k/semaphore.h
index 64d6b119bb0a..d9b2034ed1d2 100644
--- a/include/asm-m68k/semaphore.h
+++ b/include/asm-m68k/semaphore.h
@@ -1,163 +1 @@
-#ifndef _M68K_SEMAPHORE_H
-#define _M68K_SEMAPHORE_H
-
-#define RW_LOCK_BIAS		 0x01000000
-
-#ifndef __ASSEMBLY__
-
-#include <linux/linkage.h>
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-#include <linux/stringify.h>
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-
-/*
- * Interrupt-safe semaphores..
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * m68k version by Andreas Schwab
- */
-
-
-struct semaphore {
-	atomic_t count;
-	atomic_t waking;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.waking		= ATOMIC_INIT(0),				\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init(struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down_failed(void /* special register calling convention */);
-asmlinkage int  __down_failed_interruptible(void  /* params in registers */);
-asmlinkage int  __down_failed_trylock(void  /* params in registers */);
-asmlinkage void __up_wakeup(void /* special register calling convention */);
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "down_failed" is a special asm handler that calls the C
- * routine that actually waits. See arch/m68k/lib/semaphore.S
- */
-static inline void down(struct semaphore *sem)
-{
-	register struct semaphore *sem1 __asm__ ("%a1") = sem;
-
-	might_sleep();
-	__asm__ __volatile__(
-		"| atomic down operation\n\t"
-		"subql #1,%0@\n\t"
-		"jmi 2f\n\t"
-		"1:\n"
-		LOCK_SECTION_START(".even\n\t")
-		"2:\tpea 1b\n\t"
-		"jbra __down_failed\n"
-		LOCK_SECTION_END
-		: /* no outputs */
-		: "a" (sem1)
-		: "memory");
-}
-
-static inline int down_interruptible(struct semaphore *sem)
-{
-	register struct semaphore *sem1 __asm__ ("%a1") = sem;
-	register int result __asm__ ("%d0");
-
-	might_sleep();
-	__asm__ __volatile__(
-		"| atomic interruptible down operation\n\t"
-		"subql #1,%1@\n\t"
-		"jmi 2f\n\t"
-		"clrl %0\n"
-		"1:\n"
-		LOCK_SECTION_START(".even\n\t")
-		"2:\tpea 1b\n\t"
-		"jbra __down_failed_interruptible\n"
-		LOCK_SECTION_END
-		: "=d" (result)
-		: "a" (sem1)
-		: "memory");
-	return result;
-}
-
-static inline int down_trylock(struct semaphore *sem)
-{
-	register struct semaphore *sem1 __asm__ ("%a1") = sem;
-	register int result __asm__ ("%d0");
-
-	__asm__ __volatile__(
-		"| atomic down trylock operation\n\t"
-		"subql #1,%1@\n\t"
-		"jmi 2f\n\t"
-		"clrl %0\n"
-		"1:\n"
-		LOCK_SECTION_START(".even\n\t")
-		"2:\tpea 1b\n\t"
-		"jbra __down_failed_trylock\n"
-		LOCK_SECTION_END
-		: "=d" (result)
-		: "a" (sem1)
-		: "memory");
-	return result;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore *sem)
-{
-	register struct semaphore *sem1 __asm__ ("%a1") = sem;
-
-	__asm__ __volatile__(
-		"| atomic up operation\n\t"
-		"addql #1,%0@\n\t"
-		"jle 2f\n"
-		"1:\n"
-		LOCK_SECTION_START(".even\n\t")
-		"2:\t"
-		"pea 1b\n\t"
-		"jbra __up_wakeup\n"
-		LOCK_SECTION_END
-		: /* no outputs */
-		: "a" (sem1)
-		: "memory");
-}
-
-#endif /* __ASSEMBLY__ */
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-m68knommu/semaphore-helper.h b/include/asm-m68knommu/semaphore-helper.h
deleted file mode 100644
index 43da7bc483c7..000000000000
--- a/include/asm-m68knommu/semaphore-helper.h
+++ /dev/null
@@ -1,82 +0,0 @@
-#ifndef _M68K_SEMAPHORE_HELPER_H
-#define _M68K_SEMAPHORE_HELPER_H
-
-/*
- * SMP- and interrupt-safe semaphores helper functions.
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * m68k version by Andreas Schwab
- */
-
-
-/*
- * These two _must_ execute atomically wrt each other.
- */
-static inline void wake_one_more(struct semaphore * sem)
-{
-	atomic_inc(&sem->waking);
-}
-
-static inline int waking_non_zero(struct semaphore *sem)
-{
-	int ret;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 0;
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_interruptible:
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- */
-static inline int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	int ret;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 0;
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_trylock:
- *	1	failed to lock
- *	0	got the lock
- */
-static inline int waking_non_zero_trylock(struct semaphore *sem)
-{
-	int ret;
-	unsigned long flags;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	ret = 1;
-	if (atomic_read(&sem->waking) > 0) {
-		atomic_dec(&sem->waking);
-		ret = 0;
-	} else
-		atomic_inc(&sem->count);
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-#endif
diff --git a/include/asm-m68knommu/semaphore.h b/include/asm-m68knommu/semaphore.h
index 5779eb6c0689..d9b2034ed1d2 100644
--- a/include/asm-m68knommu/semaphore.h
+++ b/include/asm-m68knommu/semaphore.h
@@ -1,153 +1 @@
-#ifndef _M68K_SEMAPHORE_H
-#define _M68K_SEMAPHORE_H
-
-#define RW_LOCK_BIAS		 0x01000000
-
-#ifndef __ASSEMBLY__
-
-#include <linux/linkage.h>
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-
-/*
- * Interrupt-safe semaphores..
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * m68k version by Andreas Schwab
- */
-
-
-struct semaphore {
-	atomic_t count;
-	atomic_t waking;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.waking		= ATOMIC_INIT(0),				\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down_failed(void /* special register calling convention */);
-asmlinkage int  __down_failed_interruptible(void  /* params in registers */);
-asmlinkage int  __down_failed_trylock(void  /* params in registers */);
-asmlinkage void __up_wakeup(void /* special register calling convention */);
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-extern spinlock_t semaphore_wake_lock;
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "down_failed" is a special asm handler that calls the C
- * routine that actually waits. See arch/m68k/lib/semaphore.S
- */
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	__asm__ __volatile__(
-		"| atomic down operation\n\t"
-		"movel	%0, %%a1\n\t"
-		"lea	%%pc@(1f), %%a0\n\t"
-		"subql	#1, %%a1@\n\t"
-		"jmi __down_failed\n"
-		"1:"
-		: /* no outputs */
-		: "g" (sem)
-		: "cc", "%a0", "%a1", "memory");
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret;
-
-	might_sleep();
-	__asm__ __volatile__(
-		"| atomic down operation\n\t"
-		"movel	%1, %%a1\n\t"
-		"lea	%%pc@(1f), %%a0\n\t"
-		"subql	#1, %%a1@\n\t"
-		"jmi __down_failed_interruptible\n\t"
-		"clrl	%%d0\n"
-		"1: movel	%%d0, %0\n"
-		: "=d" (ret)
-		: "g" (sem)
-		: "cc", "%d0", "%a0", "%a1", "memory");
-	return(ret);
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	register struct semaphore *sem1 __asm__ ("%a1") = sem;
-	register int result __asm__ ("%d0");
-
-	__asm__ __volatile__(
-		"| atomic down trylock operation\n\t"
-		"subql #1,%1@\n\t"
-		"jmi 2f\n\t"
-		"clrl %0\n"
-		"1:\n"
-		".section .text.lock,\"ax\"\n"
-		".even\n"
-		"2:\tpea 1b\n\t"
-		"jbra __down_failed_trylock\n"
-		".previous"
-		: "=d" (result)
-		: "a" (sem1)
-		: "memory");
-	return result;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore * sem)
-{
-	__asm__ __volatile__(
-		"| atomic up operation\n\t"
-		"movel	%0, %%a1\n\t"
-		"lea	%%pc@(1f), %%a0\n\t"
-		"addql	#1, %%a1@\n\t"
-		"jle __up_wakeup\n"
-		"1:"
-		: /* no outputs */
-		: "g" (sem)
-		: "cc", "%a0", "%a1", "memory");
-}
-
-#endif /* __ASSEMBLY__ */
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-mips/semaphore.h b/include/asm-mips/semaphore.h
index fdf8042b784b..d9b2034ed1d2 100644
--- a/include/asm-mips/semaphore.h
+++ b/include/asm-mips/semaphore.h
@@ -1,108 +1 @@
-/*
- * This file is subject to the terms and conditions of the GNU General Public
- * License.  See the file "COPYING" in the main directory of this archive
- * for more details.
- *
- * Copyright (C) 1996  Linus Torvalds
- * Copyright (C) 1998, 99, 2000, 01, 04  Ralf Baechle
- * Copyright (C) 1999, 2000, 01  Silicon Graphics, Inc.
- * Copyright (C) 2000, 01 MIPS Technologies, Inc.
- *
- * In all honesty, little of the old MIPS code left - the PPC64 variant was
- * just looking nice and portable so I ripped it.  Credits to whoever wrote
- * it.
- */
-#ifndef __ASM_SEMAPHORE_H
-#define __ASM_SEMAPHORE_H
-
-/*
- * Remove spinlock-based RW semaphores; RW semaphore definitions are
- * now in rwsem.h and we use the generic lib/rwsem.c implementation.
- * Rework semaphores to use atomic_dec_if_positive.
- * -- Paul Mackerras (paulus@samba.org)
- */
-
-#ifdef __KERNEL__
-
-#include <asm/atomic.h>
-#include <asm/system.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	/*
-	 * Note that any negative value of count is equivalent to 0,
-	 * but additionally indicates that some process(es) might be
-	 * sleeping on `wait'.
-	 */
-	atomic_t count;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name, count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name, count)
-
-#define DECLARE_MUTEX(name)		__DECLARE_SEMAPHORE_GENERIC(name, 1)
-
-static inline void sema_init(struct semaphore *sem, int val)
-{
-	atomic_set(&sem->count, val);
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX(struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED(struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void __down(struct semaphore * sem);
-extern int  __down_interruptible(struct semaphore * sem);
-extern void __up(struct semaphore * sem);
-
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-
-	/*
-	 * Try to get the semaphore, take the slow path if we fail.
-	 */
-	if (unlikely(atomic_dec_return(&sem->count) < 0))
-		__down(sem);
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-
-	might_sleep();
-
-	if (unlikely(atomic_dec_return(&sem->count) < 0))
-		ret = __down_interruptible(sem);
-	return ret;
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	return atomic_dec_if_positive(&sem->count) < 0;
-}
-
-static inline void up(struct semaphore * sem)
-{
-	if (unlikely(atomic_inc_return(&sem->count) <= 0))
-		__up(sem);
-}
-
-#endif /* __KERNEL__ */
-
-#endif /* __ASM_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-mn10300/semaphore.h b/include/asm-mn10300/semaphore.h
index 5a9e1ad0b253..d9b2034ed1d2 100644
--- a/include/asm-mn10300/semaphore.h
+++ b/include/asm-mn10300/semaphore.h
@@ -1,169 +1 @@
-/* MN10300 Semaphores
- *
- * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
- * Written by David Howells (dhowells@redhat.com)
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public Licence
- * as published by the Free Software Foundation; either version
- * 2 of the Licence, or (at your option) any later version.
- */
-#ifndef _ASM_SEMAPHORE_H
-#define _ASM_SEMAPHORE_H
-
-#ifndef __ASSEMBLY__
-
-#include <linux/linkage.h>
-#include <linux/wait.h>
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-
-#define SEMAPHORE_DEBUG		0
-
-/*
- * the semaphore definition
- * - if count is >0 then there are tokens available on the semaphore for down
- *   to collect
- * - if count is <=0 then there are no spare tokens, and anyone that wants one
- *   must wait
- * - if wait_list is not empty, then there are processes waiting for the
- *   semaphore
- */
-struct semaphore {
-	atomic_t		count;		/* it's not really atomic, it's
-						 * just that certain modules
-						 * expect to be able to access
-						 * it directly */
-	spinlock_t		wait_lock;
-	struct list_head	wait_list;
-#if SEMAPHORE_DEBUG
-	unsigned		__magic;
-#endif
-};
-
-#if SEMAPHORE_DEBUG
-# define __SEM_DEBUG_INIT(name) , (long)&(name).__magic
-#else
-# define __SEM_DEBUG_INIT(name)
-#endif
-
-
-#define __SEMAPHORE_INITIALIZER(name, init_count)			\
-{									\
-	.count		= ATOMIC_INIT(init_count),			\
-	.wait_lock	= __SPIN_LOCK_UNLOCKED((name).wait_lock),	\
-	.wait_list	= LIST_HEAD_INIT((name).wait_list)		\
-	__SEM_DEBUG_INIT(name)						\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name, count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name, 1)
-#define DECLARE_MUTEX_LOCKED(name) __DECLARE_SEMAPHORE_GENERIC(name, 0)
-
-static inline void sema_init(struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val);
-}
-
-static inline void init_MUTEX(struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED(struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void __down(struct semaphore *sem, unsigned long flags);
-extern int  __down_interruptible(struct semaphore *sem, unsigned long flags);
-extern void __up(struct semaphore *sem);
-
-static inline void down(struct semaphore *sem)
-{
-	unsigned long flags;
-	int count;
-
-#if SEMAPHORE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	count = atomic_read(&sem->count);
-	if (likely(count > 0)) {
-		atomic_set(&sem->count, count - 1);
-		spin_unlock_irqrestore(&sem->wait_lock, flags);
-	} else {
-		__down(sem, flags);
-	}
-}
-
-static inline int down_interruptible(struct semaphore *sem)
-{
-	unsigned long flags;
-	int count, ret = 0;
-
-#if SEMAPHORE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	count = atomic_read(&sem->count);
-	if (likely(count > 0)) {
-		atomic_set(&sem->count, count - 1);
-		spin_unlock_irqrestore(&sem->wait_lock, flags);
-	} else {
-		ret = __down_interruptible(sem, flags);
-	}
-	return ret;
-}
-
-/*
- * non-blockingly attempt to down() a semaphore.
- * - returns zero if we acquired it
- */
-static inline int down_trylock(struct semaphore *sem)
-{
-	unsigned long flags;
-	int count, success = 0;
-
-#if SEMAPHORE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	count = atomic_read(&sem->count);
-	if (likely(count > 0)) {
-		atomic_set(&sem->count, count - 1);
-		success = 1;
-	}
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-	return !success;
-}
-
-static inline void up(struct semaphore *sem)
-{
-	unsigned long flags;
-
-#if SEMAPHORE_DEBUG
-	CHECK_MAGIC(sem->__magic);
-#endif
-
-	spin_lock_irqsave(&sem->wait_lock, flags);
-	if (!list_empty(&sem->wait_list))
-		__up(sem);
-	else
-		atomic_set(&sem->count, atomic_read(&sem->count) + 1);
-	spin_unlock_irqrestore(&sem->wait_lock, flags);
-}
-
-static inline int sem_getcount(struct semaphore *sem)
-{
-	return atomic_read(&sem->count);
-}
-
-#endif /* __ASSEMBLY__ */
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-parisc/semaphore-helper.h b/include/asm-parisc/semaphore-helper.h
deleted file mode 100644
index 387f7c1277a2..000000000000
--- a/include/asm-parisc/semaphore-helper.h
+++ /dev/null
@@ -1,89 +0,0 @@
-#ifndef _ASM_PARISC_SEMAPHORE_HELPER_H
-#define _ASM_PARISC_SEMAPHORE_HELPER_H
-
-/*
- * SMP- and interrupt-safe semaphores helper functions.
- *
- * (C) Copyright 1996 Linus Torvalds
- * (C) Copyright 1999 Andrea Arcangeli
- */
-
-/*
- * These two _must_ execute atomically wrt each other.
- *
- * This is trivially done with load_locked/store_cond,
- * which we have.  Let the rest of the losers suck eggs.
- */
-static __inline__ void wake_one_more(struct semaphore * sem)
-{
-	atomic_inc((atomic_t *)&sem->waking);
-}
-
-static __inline__ int waking_non_zero(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->waking > 0) {
-		sem->waking--;
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_interruptible:
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- *
- * We must undo the sem->count down_interruptible() increment while we are
- * protected by the spinlock in order to make atomic this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-static __inline__ int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->waking > 0) {
-		sem->waking--;
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_trylock:
- *	1	failed to lock
- *	0	got the lock
- *
- * We must undo the sem->count down_trylock() increment while we are
- * protected by the spinlock in order to make atomic this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-static __inline__ int waking_non_zero_trylock(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 1;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->waking <= 0)
-		atomic_inc(&sem->count);
-	else {
-		sem->waking--;
-		ret = 0;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-#endif /* _ASM_PARISC_SEMAPHORE_HELPER_H */
diff --git a/include/asm-parisc/semaphore.h b/include/asm-parisc/semaphore.h
index a16271cdc748..d9b2034ed1d2 100644
--- a/include/asm-parisc/semaphore.h
+++ b/include/asm-parisc/semaphore.h
@@ -1,145 +1 @@
-/*    SMP- and interrupt-safe semaphores.
- *    PA-RISC version by Matthew Wilcox
- *
- *    Linux/PA-RISC Project (http://www.parisc-linux.org/)
- *    Copyright (C) 1996 Linus Torvalds
- *    Copyright (C) 1999-2001 Matthew Wilcox < willy at debian d0T org >
- *    Copyright (C) 2000 Grant Grundler < grundler a debian org >
- *
- *    This program is free software; you can redistribute it and/or modify
- *    it under the terms of the GNU General Public License as published by
- *    the Free Software Foundation; either version 2 of the License, or
- *    (at your option) any later version.
- *
- *    This program is distributed in the hope that it will be useful,
- *    but WITHOUT ANY WARRANTY; without even the implied warranty of
- *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- *    GNU General Public License for more details.
- *
- *    You should have received a copy of the GNU General Public License
- *    along with this program; if not, write to the Free Software
- *    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
- */
-
-#ifndef _ASM_PARISC_SEMAPHORE_H
-#define _ASM_PARISC_SEMAPHORE_H
-
-#include <linux/spinlock.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-#include <asm/system.h>
-
-/*
- * The `count' is initialised to the number of people who are allowed to
- * take the lock.  (Normally we want a mutex, so this is `1').  if
- * `count' is positive, the lock can be taken.  if it's 0, no-one is
- * waiting on it.  if it's -1, at least one task is waiting.
- */
-struct semaphore {
-	spinlock_t	sentry;
-	int		count;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.sentry		= SPIN_LOCK_UNLOCKED,				\
-	.count		= n,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-static inline int sem_getcount(struct semaphore *sem)
-{
-	return sem->count;
-}
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-/* Semaphores can be `tried' from irq context.  So we have to disable
- * interrupts while we're messing with the semaphore.  Sorry.
- */
-
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	spin_lock_irq(&sem->sentry);
-	if (sem->count > 0) {
-		sem->count--;
-	} else {
-		__down(sem);
-	}
-	spin_unlock_irq(&sem->sentry);
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-	might_sleep();
-	spin_lock_irq(&sem->sentry);
-	if (sem->count > 0) {
-		sem->count--;
-	} else {
-		ret = __down_interruptible(sem);
-	}
-	spin_unlock_irq(&sem->sentry);
-	return ret;
-}
-
-/*
- * down_trylock returns 0 on success, 1 if we failed to get the lock.
- * May not sleep, but must preserve irq state
- */
-static inline int down_trylock(struct semaphore * sem)
-{
-	unsigned long flags;
-	int count;
-
-	spin_lock_irqsave(&sem->sentry, flags);
-	count = sem->count - 1;
-	if (count >= 0)
-		sem->count = count;
-	spin_unlock_irqrestore(&sem->sentry, flags);
-	return (count < 0);
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- */
-static inline void up(struct semaphore * sem)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&sem->sentry, flags);
-	if (sem->count < 0) {
-		__up(sem);
-	} else {
-		sem->count++;
-	}
-	spin_unlock_irqrestore(&sem->sentry, flags);
-}
-
-#endif /* _ASM_PARISC_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-powerpc/semaphore.h b/include/asm-powerpc/semaphore.h
index 48dd32e07749..d9b2034ed1d2 100644
--- a/include/asm-powerpc/semaphore.h
+++ b/include/asm-powerpc/semaphore.h
@@ -1,94 +1 @@
-#ifndef _ASM_POWERPC_SEMAPHORE_H
-#define _ASM_POWERPC_SEMAPHORE_H
-
-/*
- * Remove spinlock-based RW semaphores; RW semaphore definitions are
- * now in rwsem.h and we use the generic lib/rwsem.c implementation.
- * Rework semaphores to use atomic_dec_if_positive.
- * -- Paul Mackerras (paulus@samba.org)
- */
-
-#ifdef __KERNEL__
-
-#include <asm/atomic.h>
-#include <asm/system.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	/*
-	 * Note that any negative value of count is equivalent to 0,
-	 * but additionally indicates that some process(es) might be
-	 * sleeping on `wait'.
-	 */
-	atomic_t count;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name, count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name)		__DECLARE_SEMAPHORE_GENERIC(name, 1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	atomic_set(&sem->count, val);
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void __down(struct semaphore * sem);
-extern int  __down_interruptible(struct semaphore * sem);
-extern void __up(struct semaphore * sem);
-
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-
-	/*
-	 * Try to get the semaphore, take the slow path if we fail.
-	 */
-	if (unlikely(atomic_dec_return(&sem->count) < 0))
-		__down(sem);
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-
-	might_sleep();
-
-	if (unlikely(atomic_dec_return(&sem->count) < 0))
-		ret = __down_interruptible(sem);
-	return ret;
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	return atomic_dec_if_positive(&sem->count) < 0;
-}
-
-static inline void up(struct semaphore * sem)
-{
-	if (unlikely(atomic_inc_return(&sem->count) <= 0))
-		__up(sem);
-}
-
-#endif /* __KERNEL__ */
-
-#endif /* _ASM_POWERPC_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-s390/semaphore.h b/include/asm-s390/semaphore.h
index 0e7001ad8392..d9b2034ed1d2 100644
--- a/include/asm-s390/semaphore.h
+++ b/include/asm-s390/semaphore.h
@@ -1,107 +1 @@
-/*
- *  include/asm-s390/semaphore.h
- *
- *  S390 version
- *    Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
- *
- *  Derived from "include/asm-i386/semaphore.h"
- *    (C) Copyright 1996 Linus Torvalds
- */
-
-#ifndef _S390_SEMAPHORE_H
-#define _S390_SEMAPHORE_H
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	/*
-	 * Note that any negative value of count is equivalent to 0,
-	 * but additionally indicates that some process(es) might be
-	 * sleeping on `wait'.
-	 */
-	atomic_t count;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name,count) \
-	{ ATOMIC_INIT(count), __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) }
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	atomic_set(&sem->count, val);
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	if (atomic_dec_return(&sem->count) < 0)
-		__down(sem);
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-
-	might_sleep();
-	if (atomic_dec_return(&sem->count) < 0)
-		ret = __down_interruptible(sem);
-	return ret;
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	int old_val, new_val;
-
-	/*
-	 * This inline assembly atomically implements the equivalent
-	 * to the following C code:
-	 *   old_val = sem->count.counter;
-	 *   if ((new_val = old_val) > 0)
-	 *       sem->count.counter = --new_val;
-	 * In the ppc code this is called atomic_dec_if_positive.
-	 */
-	asm volatile(
-		"	l	%0,0(%3)\n"
-		"0:	ltr	%1,%0\n"
-		"	jle	1f\n"
-		"	ahi	%1,-1\n"
-		"	cs	%0,%1,0(%3)\n"
-		"	jl	0b\n"
-		"1:"
-		: "=&d" (old_val), "=&d" (new_val), "=m" (sem->count.counter)
-		: "a" (&sem->count.counter), "m" (sem->count.counter)
-		: "cc", "memory");
-	return old_val <= 0;
-}
-
-static inline void up(struct semaphore * sem)
-{
-	if (atomic_inc_return(&sem->count) <= 0)
-		__up(sem);
-}
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-sh/semaphore-helper.h b/include/asm-sh/semaphore-helper.h
deleted file mode 100644
index bd8230c369ca..000000000000
--- a/include/asm-sh/semaphore-helper.h
+++ /dev/null
@@ -1,89 +0,0 @@
-#ifndef __ASM_SH_SEMAPHORE_HELPER_H
-#define __ASM_SH_SEMAPHORE_HELPER_H
-
-/*
- * SMP- and interrupt-safe semaphores helper functions.
- *
- * (C) Copyright 1996 Linus Torvalds
- * (C) Copyright 1999 Andrea Arcangeli
- */
-
-/*
- * These two _must_ execute atomically wrt each other.
- *
- * This is trivially done with load_locked/store_cond,
- * which we have.  Let the rest of the losers suck eggs.
- */
-static __inline__ void wake_one_more(struct semaphore * sem)
-{
-	atomic_inc((atomic_t *)&sem->sleepers);
-}
-
-static __inline__ int waking_non_zero(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->sleepers > 0) {
-		sem->sleepers--;
-		ret = 1;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_interruptible:
- *	1	got the lock
- *	0	go to sleep
- *	-EINTR	interrupted
- *
- * We must undo the sem->count down_interruptible() increment while we are
- * protected by the spinlock in order to make atomic this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-static __inline__ int waking_non_zero_interruptible(struct semaphore *sem,
-						struct task_struct *tsk)
-{
-	unsigned long flags;
-	int ret = 0;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->sleepers > 0) {
-		sem->sleepers--;
-		ret = 1;
-	} else if (signal_pending(tsk)) {
-		atomic_inc(&sem->count);
-		ret = -EINTR;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-/*
- * waking_non_zero_trylock:
- *	1	failed to lock
- *	0	got the lock
- *
- * We must undo the sem->count down_trylock() increment while we are
- * protected by the spinlock in order to make atomic this atomic_inc() with the
- * atomic_read() in wake_one_more(), otherwise we can race. -arca
- */
-static __inline__ int waking_non_zero_trylock(struct semaphore *sem)
-{
-	unsigned long flags;
-	int ret = 1;
-
-	spin_lock_irqsave(&semaphore_wake_lock, flags);
-	if (sem->sleepers <= 0)
-		atomic_inc(&sem->count);
-	else {
-		sem->sleepers--;
-		ret = 0;
-	}
-	spin_unlock_irqrestore(&semaphore_wake_lock, flags);
-	return ret;
-}
-
-#endif /* __ASM_SH_SEMAPHORE_HELPER_H */
diff --git a/include/asm-sh/semaphore.h b/include/asm-sh/semaphore.h
index 9e5a37c4dce2..d9b2034ed1d2 100644
--- a/include/asm-sh/semaphore.h
+++ b/include/asm-sh/semaphore.h
@@ -1,115 +1 @@
-#ifndef __ASM_SH_SEMAPHORE_H
-#define __ASM_SH_SEMAPHORE_H
-
-#include <linux/linkage.h>
-
-#ifdef __KERNEL__
-/*
- * SMP- and interrupt-safe semaphores.
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * SuperH verison by Niibe Yutaka
- *  (Currently no asm implementation but generic C code...)
- */
-
-#include <linux/spinlock.h>
-#include <linux/rwsem.h>
-#include <linux/wait.h>
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-/*
- *	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
- *
- * i'd rather use the more flexible initialization above, but sadly
- * GCC 2.7.2.3 emits a bogus warning. EGCS doesn't. Oh well.
- */
-	atomic_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-#if 0
-asmlinkage void __down_failed(void /* special register calling convention */);
-asmlinkage int  __down_failed_interruptible(void  /* params in registers */);
-asmlinkage int  __down_failed_trylock(void  /* params in registers */);
-asmlinkage void __up_wakeup(void /* special register calling convention */);
-#endif
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-extern spinlock_t semaphore_wake_lock;
-
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	if (atomic_dec_return(&sem->count) < 0)
-		__down(sem);
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-
-	might_sleep();
-	if (atomic_dec_return(&sem->count) < 0)
-		ret = __down_interruptible(sem);
-	return ret;
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	int ret = 0;
-
-	if (atomic_dec_return(&sem->count) < 0)
-		ret = __down_trylock(sem);
-	return ret;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- */
-static inline void up(struct semaphore * sem)
-{
-	if (atomic_inc_return(&sem->count) <= 0)
-		__up(sem);
-}
-
-#endif
-#endif /* __ASM_SH_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/asm-sparc/semaphore.h b/include/asm-sparc/semaphore.h
index 8018f9f4d497..d9b2034ed1d2 100644
--- a/include/asm-sparc/semaphore.h
+++ b/include/asm-sparc/semaphore.h
@@ -1,192 +1 @@
-#ifndef _SPARC_SEMAPHORE_H
-#define _SPARC_SEMAPHORE_H
-
-/* Dinky, good for nothing, just barely irq safe, Sparc semaphores. */
-
-#ifdef __KERNEL__
-
-#include <asm/atomic.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	atomic24_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC24_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	atomic24_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void __down(struct semaphore * sem);
-extern int __down_interruptible(struct semaphore * sem);
-extern int __down_trylock(struct semaphore * sem);
-extern void __up(struct semaphore * sem);
-
-static inline void down(struct semaphore * sem)
-{
-	register volatile int *ptr asm("g1");
-	register int increment asm("g2");
-
-	might_sleep();
-
-	ptr = &(sem->count.counter);
-	increment = 1;
-
-	__asm__ __volatile__(
-	"mov	%%o7, %%g4\n\t"
-	"call	___atomic24_sub\n\t"
-	" add	%%o7, 8, %%o7\n\t"
-	"tst	%%g2\n\t"
-	"bl	2f\n\t"
-	" nop\n"
-	"1:\n\t"
-	".subsection 2\n"
-	"2:\n\t"
-	"save	%%sp, -64, %%sp\n\t"
-	"mov	%%g1, %%l1\n\t"
-	"mov	%%g5, %%l5\n\t"
-	"call	%3\n\t"
-	" mov	%%g1, %%o0\n\t"
-	"mov	%%l1, %%g1\n\t"
-	"ba	1b\n\t"
-	" restore %%l5, %%g0, %%g5\n\t"
-	".previous\n"
-	: "=&r" (increment)
-	: "0" (increment), "r" (ptr), "i" (__down)
-	: "g3", "g4", "g7", "memory", "cc");
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	register volatile int *ptr asm("g1");
-	register int increment asm("g2");
-
-	might_sleep();
-
-	ptr = &(sem->count.counter);
-	increment = 1;
-
-	__asm__ __volatile__(
-	"mov	%%o7, %%g4\n\t"
-	"call	___atomic24_sub\n\t"
-	" add	%%o7, 8, %%o7\n\t"
-	"tst	%%g2\n\t"
-	"bl	2f\n\t"
-	" clr	%%g2\n"
-	"1:\n\t"
-	".subsection 2\n"
-	"2:\n\t"
-	"save	%%sp, -64, %%sp\n\t"
-	"mov	%%g1, %%l1\n\t"
-	"mov	%%g5, %%l5\n\t"
-	"call	%3\n\t"
-	" mov	%%g1, %%o0\n\t"
-	"mov	%%l1, %%g1\n\t"
-	"mov	%%l5, %%g5\n\t"
-	"ba	1b\n\t"
-	" restore %%o0, %%g0, %%g2\n\t"
-	".previous\n"
-	: "=&r" (increment)
-	: "0" (increment), "r" (ptr), "i" (__down_interruptible)
-	: "g3", "g4", "g7", "memory", "cc");
-
-	return increment;
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	register volatile int *ptr asm("g1");
-	register int increment asm("g2");
-
-	ptr = &(sem->count.counter);
-	increment = 1;
-
-	__asm__ __volatile__(
-	"mov	%%o7, %%g4\n\t"
-	"call	___atomic24_sub\n\t"
-	" add	%%o7, 8, %%o7\n\t"
-	"tst	%%g2\n\t"
-	"bl	2f\n\t"
-	" clr	%%g2\n"
-	"1:\n\t"
-	".subsection 2\n"
-	"2:\n\t"
-	"save	%%sp, -64, %%sp\n\t"
-	"mov	%%g1, %%l1\n\t"
-	"mov	%%g5, %%l5\n\t"
-	"call	%3\n\t"
-	" mov	%%g1, %%o0\n\t"
-	"mov	%%l1, %%g1\n\t"
-	"mov	%%l5, %%g5\n\t"
-	"ba	1b\n\t"
-	" restore %%o0, %%g0, %%g2\n\t"
-	".previous\n"
-	: "=&r" (increment)
-	: "0" (increment), "r" (ptr), "i" (__down_trylock)
-	: "g3", "g4", "g7", "memory", "cc");
-
-	return increment;
-}
-
-static inline void up(struct semaphore * sem)
-{
-	register volatile int *ptr asm("g1");
-	register int increment asm("g2");
-
-	ptr = &(sem->count.counter);
-	increment = 1;
-
-	__asm__ __volatile__(
-	"mov	%%o7, %%g4\n\t"
-	"call	___atomic24_add\n\t"
-	" add	%%o7, 8, %%o7\n\t"
-	"tst	%%g2\n\t"
-	"ble	2f\n\t"
-	" nop\n"
-	"1:\n\t"
-	".subsection 2\n"
-	"2:\n\t"
-	"save	%%sp, -64, %%sp\n\t"
-	"mov	%%g1, %%l1\n\t"
-	"mov	%%g5, %%l5\n\t"
-	"call	%3\n\t"
-	" mov	%%g1, %%o0\n\t"
-	"mov	%%l1, %%g1\n\t"
-	"ba	1b\n\t"
-	" restore %%l5, %%g0, %%g5\n\t"
-	".previous\n"
-	: "=&r" (increment)
-	: "0" (increment), "r" (ptr), "i" (__up)
-	: "g3", "g4", "g7", "memory", "cc");
-}	
-
-#endif /* __KERNEL__ */
-
-#endif /* !(_SPARC_SEMAPHORE_H) */
+#include <linux/semaphore.h>
diff --git a/include/asm-sparc64/semaphore.h b/include/asm-sparc64/semaphore.h
index 7f7c0c4e024f..d9b2034ed1d2 100644
--- a/include/asm-sparc64/semaphore.h
+++ b/include/asm-sparc64/semaphore.h
@@ -1,53 +1 @@
-#ifndef _SPARC64_SEMAPHORE_H
-#define _SPARC64_SEMAPHORE_H
-
-/* These are actually reasonable on the V9.
- *
- * See asm-ppc/semaphore.h for implementation commentary,
- * only sparc64 specific issues are commented here.
- */
-#ifdef __KERNEL__
-
-#include <asm/atomic.h>
-#include <asm/system.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	atomic_t count;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, count) \
-	{ ATOMIC_INIT(count), \
-	  __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) }
-
-#define __DECLARE_SEMAPHORE_GENERIC(name, count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name)		__DECLARE_SEMAPHORE_GENERIC(name, 1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	atomic_set(&sem->count, val);
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern void up(struct semaphore *sem);
-extern void down(struct semaphore *sem);
-extern int down_trylock(struct semaphore *sem);
-extern int down_interruptible(struct semaphore *sem);
-
-#endif /* __KERNEL__ */
-
-#endif /* !(_SPARC64_SEMAPHORE_H) */
+#include <linux/semaphore.h>
diff --git a/include/asm-um/semaphore.h b/include/asm-um/semaphore.h
index ff13c34de421..d9b2034ed1d2 100644
--- a/include/asm-um/semaphore.h
+++ b/include/asm-um/semaphore.h
@@ -1,6 +1 @@
-#ifndef __UM_SEMAPHORE_H
-#define __UM_SEMAPHORE_H
-
-#include "asm/arch/semaphore.h"
-
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-v850/semaphore.h b/include/asm-v850/semaphore.h
index 10ed0ccf37df..d9b2034ed1d2 100644
--- a/include/asm-v850/semaphore.h
+++ b/include/asm-v850/semaphore.h
@@ -1,84 +1 @@
-#ifndef __V850_SEMAPHORE_H__
-#define __V850_SEMAPHORE_H__
-
-#include <linux/linkage.h>
-#include <linux/spinlock.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-#include <asm/atomic.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name,count)				      \
-	{ ATOMIC_INIT (count), 0,					      \
-	  __WAIT_QUEUE_HEAD_INITIALIZER ((name).wait) }
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count)	\
-	struct semaphore name = __SEMAPHORE_INITIALIZER (name,count)
-
-#define DECLARE_MUTEX(name)		__DECLARE_SEMAPHORE_GENERIC (name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init (sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init (sem, 0);
-}
-
-/*
- * special register calling convention
- */
-asmlinkage void __down_failed (void);
-asmlinkage int  __down_interruptible_failed (void);
-asmlinkage int  __down_trylock_failed (void);
-asmlinkage void __up_wakeup (void);
-
-extern void __down (struct semaphore * sem);
-extern int  __down_interruptible (struct semaphore * sem);
-extern int  __down_trylock (struct semaphore * sem);
-extern void __up (struct semaphore * sem);
-
-static inline void down (struct semaphore * sem)
-{
-	might_sleep();
-	if (atomic_dec_return (&sem->count) < 0)
-		__down (sem);
-}
-
-static inline int down_interruptible (struct semaphore * sem)
-{
-	int ret = 0;
-	might_sleep();
-	if (atomic_dec_return (&sem->count) < 0)
-		ret = __down_interruptible (sem);
-	return ret;
-}
-
-static inline int down_trylock (struct semaphore *sem)
-{
-	int ret = 0;
-	if (atomic_dec_return (&sem->count) < 0)
-		ret = __down_trylock (sem);
-	return ret;
-}
-
-static inline void up (struct semaphore * sem)
-{
-	if (atomic_inc_return (&sem->count) <= 0)
-		__up (sem);
-}
-
-#endif /* __V850_SEMAPHORE_H__ */
+#include <linux/semaphore.h>
diff --git a/include/asm-x86/semaphore.h b/include/asm-x86/semaphore.h
index 572c0b67a6b0..d9b2034ed1d2 100644
--- a/include/asm-x86/semaphore.h
+++ b/include/asm-x86/semaphore.h
@@ -1,5 +1 @@
-#ifdef CONFIG_X86_32
-# include "semaphore_32.h"
-#else
-# include "semaphore_64.h"
-#endif
+#include <linux/semaphore.h>
diff --git a/include/asm-x86/semaphore_32.h b/include/asm-x86/semaphore_32.h
deleted file mode 100644
index ac96d3804d0c..000000000000
--- a/include/asm-x86/semaphore_32.h
+++ /dev/null
@@ -1,175 +0,0 @@
-#ifndef _I386_SEMAPHORE_H
-#define _I386_SEMAPHORE_H
-
-#include <linux/linkage.h>
-
-#ifdef __KERNEL__
-
-/*
- * SMP- and interrupt-safe semaphores..
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * Modified 1996-12-23 by Dave Grothe <dave@gcom.com> to fix bugs in
- *                     the original code and to make semaphore waits
- *                     interruptible so that processes waiting on
- *                     semaphores can be killed.
- * Modified 1999-02-14 by Andrea Arcangeli, split the sched.c helper
- *		       functions in asm/sempahore-helper.h while fixing a
- *		       potential and subtle race discovered by Ulrich Schmid
- *		       in down_interruptible(). Since I started to play here I
- *		       also implemented the `trylock' semaphore operation.
- *          1999-07-02 Artur Skawina <skawina@geocities.com>
- *                     Optimized "0(ecx)" -> "(ecx)" (the assembler does not
- *                     do this). Changed calling sequences from push/jmp to
- *                     traditional call/ret.
- * Modified 2001-01-01 Andreas Franck <afranck@gmx.de>
- *		       Some hacks to ensure compatibility with recent
- *		       GCC snapshots, to avoid stack corruption when compiling
- *		       with -fomit-frame-pointer. It's not sure if this will
- *		       be fixed in GCC, as our previous implementation was a
- *		       bit dubious.
- *
- * If you would like to see an analysis of this implementation, please
- * ftp to gcom.com and download the file
- * /pub/linux/src/semaphore/semaphore-2.0.24.tar.gz.
- *
- */
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-/*
- *	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
- *
- * i'd rather use the more flexible initialization above, but sadly
- * GCC 2.7.2.3 emits a bogus warning. EGCS doesn't. Oh well.
- */
-	atomic_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-extern asmregparm void __down_failed(atomic_t *count_ptr);
-extern asmregparm int  __down_failed_interruptible(atomic_t *count_ptr);
-extern asmregparm int  __down_failed_trylock(atomic_t *count_ptr);
-extern asmregparm void __up_wakeup(atomic_t *count_ptr);
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "__down_failed" is a special asm handler that calls the C
- * routine that actually waits. See arch/i386/kernel/semaphore.c
- */
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-	__asm__ __volatile__(
-		"# atomic down operation\n\t"
-		LOCK_PREFIX "decl %0\n\t"     /* --sem->count */
-		"jns 2f\n"
-		"\tlea %0,%%eax\n\t"
-		"call __down_failed\n"
-		"2:"
-		:"+m" (sem->count)
-		:
-		:"memory","ax");
-}
-
-/*
- * Interruptible try to acquire a semaphore.  If we obtained
- * it, return zero.  If we were interrupted, returns -EINTR
- */
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int result;
-
-	might_sleep();
-	__asm__ __volatile__(
-		"# atomic interruptible down operation\n\t"
-		"xorl %0,%0\n\t"
-		LOCK_PREFIX "decl %1\n\t"     /* --sem->count */
-		"jns 2f\n\t"
-		"lea %1,%%eax\n\t"
-		"call __down_failed_interruptible\n"
-		"2:"
-		:"=&a" (result), "+m" (sem->count)
-		:
-		:"memory");
-	return result;
-}
-
-/*
- * Non-blockingly attempt to down() a semaphore.
- * Returns zero if we acquired it
- */
-static inline int down_trylock(struct semaphore * sem)
-{
-	int result;
-
-	__asm__ __volatile__(
-		"# atomic interruptible down operation\n\t"
-		"xorl %0,%0\n\t"
-		LOCK_PREFIX "decl %1\n\t"     /* --sem->count */
-		"jns 2f\n\t"
-		"lea %1,%%eax\n\t"
-		"call __down_failed_trylock\n\t"
-		"2:\n"
-		:"=&a" (result), "+m" (sem->count)
-		:
-		:"memory");
-	return result;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- */
-static inline void up(struct semaphore * sem)
-{
-	__asm__ __volatile__(
-		"# atomic up operation\n\t"
-		LOCK_PREFIX "incl %0\n\t"     /* ++sem->count */
-		"jg 1f\n\t"
-		"lea %0,%%eax\n\t"
-		"call __up_wakeup\n"
-		"1:"
-		:"+m" (sem->count)
-		:
-		:"memory","ax");
-}
-
-#endif
-#endif
diff --git a/include/asm-x86/semaphore_64.h b/include/asm-x86/semaphore_64.h
deleted file mode 100644
index 79694306bf7d..000000000000
--- a/include/asm-x86/semaphore_64.h
+++ /dev/null
@@ -1,180 +0,0 @@
-#ifndef _X86_64_SEMAPHORE_H
-#define _X86_64_SEMAPHORE_H
-
-#include <linux/linkage.h>
-
-#ifdef __KERNEL__
-
-/*
- * SMP- and interrupt-safe semaphores..
- *
- * (C) Copyright 1996 Linus Torvalds
- *
- * Modified 1996-12-23 by Dave Grothe <dave@gcom.com> to fix bugs in
- *                     the original code and to make semaphore waits
- *                     interruptible so that processes waiting on
- *                     semaphores can be killed.
- * Modified 1999-02-14 by Andrea Arcangeli, split the sched.c helper
- *		       functions in asm/sempahore-helper.h while fixing a
- *		       potential and subtle race discovered by Ulrich Schmid
- *		       in down_interruptible(). Since I started to play here I
- *		       also implemented the `trylock' semaphore operation.
- *          1999-07-02 Artur Skawina <skawina@geocities.com>
- *                     Optimized "0(ecx)" -> "(ecx)" (the assembler does not
- *                     do this). Changed calling sequences from push/jmp to
- *                     traditional call/ret.
- * Modified 2001-01-01 Andreas Franck <afranck@gmx.de>
- *		       Some hacks to ensure compatibility with recent
- *		       GCC snapshots, to avoid stack corruption when compiling
- *		       with -fomit-frame-pointer. It's not sure if this will
- *		       be fixed in GCC, as our previous implementation was a
- *		       bit dubious.
- *
- * If you would like to see an analysis of this implementation, please
- * ftp to gcom.com and download the file
- * /pub/linux/src/semaphore/semaphore-2.0.24.tar.gz.
- *
- */
-
-#include <asm/system.h>
-#include <asm/atomic.h>
-#include <asm/rwlock.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-#include <linux/stringify.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name, n)				\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) \
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-/*
- *	*sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val);
- *
- * i'd rather use the more flexible initialization above, but sadly
- * GCC 2.7.2.3 emits a bogus warning. EGCS doesn't. Oh well.
- */
-	atomic_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down_failed(void /* special register calling convention */);
-asmlinkage int  __down_failed_interruptible(void  /* params in registers */);
-asmlinkage int  __down_failed_trylock(void  /* params in registers */);
-asmlinkage void __up_wakeup(void /* special register calling convention */);
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-/*
- * This is ugly, but we want the default case to fall through.
- * "__down_failed" is a special asm handler that calls the C
- * routine that actually waits. See arch/x86_64/kernel/semaphore.c
- */
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-
-	__asm__ __volatile__(
-		"# atomic down operation\n\t"
-		LOCK_PREFIX "decl %0\n\t"     /* --sem->count */
-		"jns 1f\n\t"
-		"call __down_failed\n"
-		"1:"
-		:"=m" (sem->count)
-		:"D" (sem)
-		:"memory");
-}
-
-/*
- * Interruptible try to acquire a semaphore.  If we obtained
- * it, return zero.  If we were interrupted, returns -EINTR
- */
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int result;
-
-	might_sleep();
-
-	__asm__ __volatile__(
-		"# atomic interruptible down operation\n\t"
-		"xorl %0,%0\n\t"
-		LOCK_PREFIX "decl %1\n\t"     /* --sem->count */
-		"jns 2f\n\t"
-		"call __down_failed_interruptible\n"
-		"2:\n"
-		:"=&a" (result), "=m" (sem->count)
-		:"D" (sem)
-		:"memory");
-	return result;
-}
-
-/*
- * Non-blockingly attempt to down() a semaphore.
- * Returns zero if we acquired it
- */
-static inline int down_trylock(struct semaphore * sem)
-{
-	int result;
-
-	__asm__ __volatile__(
-		"# atomic interruptible down operation\n\t"
-		"xorl %0,%0\n\t"
-		LOCK_PREFIX "decl %1\n\t"     /* --sem->count */
-		"jns 2f\n\t"
-		"call __down_failed_trylock\n\t"
-		"2:\n"
-		:"=&a" (result), "=m" (sem->count)
-		:"D" (sem)
-		:"memory","cc");
-	return result;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- * The default case (no contention) will result in NO
- * jumps for both down() and up().
- */
-static inline void up(struct semaphore * sem)
-{
-	__asm__ __volatile__(
-		"# atomic up operation\n\t"
-		LOCK_PREFIX "incl %0\n\t"     /* ++sem->count */
-		"jg 1f\n\t"
-		"call __up_wakeup\n"
-		"1:"
-		:"=m" (sem->count)
-		:"D" (sem)
-		:"memory");
-}
-#endif /* __KERNEL__ */
-#endif
diff --git a/include/asm-xtensa/semaphore.h b/include/asm-xtensa/semaphore.h
index 3e04167cd9dc..d9b2034ed1d2 100644
--- a/include/asm-xtensa/semaphore.h
+++ b/include/asm-xtensa/semaphore.h
@@ -1,99 +1 @@
-/*
- * linux/include/asm-xtensa/semaphore.h
- *
- * This file is subject to the terms and conditions of the GNU General Public
- * License.  See the file "COPYING" in the main directory of this archive
- * for more details.
- *
- * Copyright (C) 2001 - 2005 Tensilica Inc.
- */
-
-#ifndef _XTENSA_SEMAPHORE_H
-#define _XTENSA_SEMAPHORE_H
-
-#include <asm/atomic.h>
-#include <asm/system.h>
-#include <linux/wait.h>
-#include <linux/rwsem.h>
-
-struct semaphore {
-	atomic_t count;
-	int sleepers;
-	wait_queue_head_t wait;
-};
-
-#define __SEMAPHORE_INITIALIZER(name,n)					\
-{									\
-	.count		= ATOMIC_INIT(n),				\
-	.sleepers	= 0,						\
-	.wait		= __WAIT_QUEUE_HEAD_INITIALIZER((name).wait)	\
-}
-
-#define __DECLARE_SEMAPHORE_GENERIC(name,count) 			\
-	struct semaphore name = __SEMAPHORE_INITIALIZER(name,count)
-
-#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1)
-
-static inline void sema_init (struct semaphore *sem, int val)
-{
-	atomic_set(&sem->count, val);
-	sem->sleepers = 0;
-	init_waitqueue_head(&sem->wait);
-}
-
-static inline void init_MUTEX (struct semaphore *sem)
-{
-	sema_init(sem, 1);
-}
-
-static inline void init_MUTEX_LOCKED (struct semaphore *sem)
-{
-	sema_init(sem, 0);
-}
-
-asmlinkage void __down(struct semaphore * sem);
-asmlinkage int  __down_interruptible(struct semaphore * sem);
-asmlinkage int  __down_trylock(struct semaphore * sem);
-asmlinkage void __up(struct semaphore * sem);
-
-extern spinlock_t semaphore_wake_lock;
-
-static inline void down(struct semaphore * sem)
-{
-	might_sleep();
-
-	if (atomic_sub_return(1, &sem->count) < 0)
-		__down(sem);
-}
-
-static inline int down_interruptible(struct semaphore * sem)
-{
-	int ret = 0;
-
-	might_sleep();
-
-	if (atomic_sub_return(1, &sem->count) < 0)
-		ret = __down_interruptible(sem);
-	return ret;
-}
-
-static inline int down_trylock(struct semaphore * sem)
-{
-	int ret = 0;
-
-	if (atomic_sub_return(1, &sem->count) < 0)
-		ret = __down_trylock(sem);
-	return ret;
-}
-
-/*
- * Note! This is subtle. We jump to wake people up only if
- * the semaphore was negative (== somebody was waiting on it).
- */
-static inline void up(struct semaphore * sem)
-{
-	if (atomic_add_return(1, &sem->count) <= 0)
-		__up(sem);
-}
-
-#endif /* _XTENSA_SEMAPHORE_H */
+#include <linux/semaphore.h>
diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h
new file mode 100644
index 000000000000..b3c691b089b2
--- /dev/null
+++ b/include/linux/semaphore.h
@@ -0,0 +1,77 @@
+/*
+ * Copyright (c) 2008 Intel Corporation
+ * Author: Matthew Wilcox <willy@linux.intel.com>
+ *
+ * Distributed under the terms of the GNU GPL, version 2
+ *
+ * Counting semaphores allow up to <n> tasks to acquire the semaphore
+ * simultaneously.
+ */
+#ifndef __LINUX_SEMAPHORE_H
+#define __LINUX_SEMAPHORE_H
+
+#include <linux/list.h>
+#include <linux/spinlock.h>
+
+/*
+ * The spinlock controls access to the other members of the semaphore.
+ * 'count' is decremented by every task which calls down*() and incremented
+ * by every call to up().  Thus, if it is positive, it indicates how many
+ * more tasks may acquire the lock.  If it is negative, it indicates how
+ * many tasks are waiting for the lock.  Tasks waiting for the lock are
+ * kept on the wait_list.
+ */
+struct semaphore {
+	spinlock_t		lock;
+	int			count;
+	struct list_head	wait_list;
+};
+
+#define __SEMAPHORE_INITIALIZER(name, n)				\
+{									\
+	.lock		= __SPIN_LOCK_UNLOCKED((name).lock),		\
+	.count		= n,						\
+	.wait_list	= LIST_HEAD_INIT((name).wait_list),		\
+}
+
+#define __DECLARE_SEMAPHORE_GENERIC(name, count) \
+	struct semaphore name = __SEMAPHORE_INITIALIZER(name, count)
+
+#define DECLARE_MUTEX(name)	__DECLARE_SEMAPHORE_GENERIC(name, 1)
+
+static inline void sema_init(struct semaphore *sem, int val)
+{
+	static struct lock_class_key __key;
+	*sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val);
+	lockdep_init_map(&sem->lock.dep_map, "semaphore->lock", &__key, 0);
+}
+
+#define init_MUTEX(sem)		sema_init(sem, 1)
+#define init_MUTEX_LOCKED(sem)	sema_init(sem, 0)
+
+/*
+ * Attempt to acquire the semaphore.  If another task is already holding the
+ * semaphore, sleep until the semaphore is released.
+ */
+extern void down(struct semaphore *sem);
+
+/*
+ * As down(), except the sleep may be interrupted by a signal.  If it is,
+ * this function will return -EINTR.
+ */
+extern int __must_check down_interruptible(struct semaphore *sem);
+
+/*
+ * As down(), except this function will not sleep.  It will return 0 if it
+ * acquired the semaphore and 1 if the semaphore was contended.  This
+ * function may be called from any context, including interrupt and softirq.
+ */
+extern int __must_check down_trylock(struct semaphore *sem);
+
+/*
+ * Release the semaphore.  Unlike mutexes, up() may be called from any
+ * context and even by tasks which have never called down().
+ */
+extern void up(struct semaphore *sem);
+
+#endif /* __LINUX_SEMAPHORE_H */
diff --git a/kernel/Makefile b/kernel/Makefile
index 6c584c55a6e9..f45c69e69688 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -8,7 +8,7 @@ obj-y     = sched.o fork.o exec_domain.o panic.o printk.o profile.o \
 	    signal.o sys.o kmod.o workqueue.o pid.o \
 	    rcupdate.o extable.o params.o posix-timers.o \
 	    kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \
-	    hrtimer.o rwsem.o nsproxy.o srcu.o \
+	    hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \
 	    notifier.o ksysfs.o pm_qos_params.o
 
 obj-$(CONFIG_SYSCTL) += sysctl_check.o
diff --git a/kernel/semaphore.c b/kernel/semaphore.c
new file mode 100644
index 000000000000..d5a72702f261
--- /dev/null
+++ b/kernel/semaphore.c
@@ -0,0 +1,187 @@
+/*
+ * Copyright (c) 2008 Intel Corporation
+ * Author: Matthew Wilcox <willy@linux.intel.com>
+ *
+ * Distributed under the terms of the GNU GPL, version 2
+ */
+
+#include <linux/compiler.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/semaphore.h>
+#include <linux/spinlock.h>
+
+/*
+ * Some notes on the implementation:
+ *
+ * down_trylock() and up() can be called from interrupt context.
+ * So we have to disable interrupts when taking the lock.
+ *
+ * The ->count variable, if positive, defines how many more tasks can
+ * acquire the semaphore.  If negative, it represents how many tasks are
+ * waiting on the semaphore (*).  If zero, no tasks are waiting, and no more
+ * tasks can acquire the semaphore.
+ *
+ * (*) Except for the window between one task calling up() and the task
+ * sleeping in a __down_common() waking up.  In order to avoid a third task
+ * coming in and stealing the second task's wakeup, we leave the ->count
+ * negative.  If we have a more complex situation, the ->count may become
+ * zero or negative (eg a semaphore with count = 2, three tasks attempt to
+ * acquire it, one sleeps, two finish and call up(), the second task to call
+ * up() notices that the list is empty and just increments count).
+ */
+
+static noinline void __down(struct semaphore *sem);
+static noinline int __down_interruptible(struct semaphore *sem);
+static noinline void __up(struct semaphore *sem);
+
+void down(struct semaphore *sem)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&sem->lock, flags);
+	if (unlikely(sem->count-- <= 0))
+		__down(sem);
+	spin_unlock_irqrestore(&sem->lock, flags);
+}
+EXPORT_SYMBOL(down);
+
+int down_interruptible(struct semaphore *sem)
+{
+	unsigned long flags;
+	int result = 0;
+
+	spin_lock_irqsave(&sem->lock, flags);
+	if (unlikely(sem->count-- <= 0))
+		result = __down_interruptible(sem);
+	spin_unlock_irqrestore(&sem->lock, flags);
+
+	return result;
+}
+EXPORT_SYMBOL(down_interruptible);
+
+/**
+ * down_trylock - try to acquire the semaphore, without waiting
+ * @sem: the semaphore to be acquired
+ *
+ * Try to acquire the semaphore atomically.  Returns 0 if the mutex has
+ * been acquired successfully and 1 if it is contended.
+ *
+ * NOTE: This return value is inverted from both spin_trylock and
+ * mutex_trylock!  Be careful about this when converting code.
+ *
+ * Unlike mutex_trylock, this function can be used from interrupt context,
+ * and the semaphore can be released by any task or interrupt.
+ */
+int down_trylock(struct semaphore *sem)
+{
+	unsigned long flags;
+	int count;
+
+	spin_lock_irqsave(&sem->lock, flags);
+	count = sem->count - 1;
+	if (likely(count >= 0))
+		sem->count = count;
+	spin_unlock_irqrestore(&sem->lock, flags);
+
+	return (count < 0);
+}
+EXPORT_SYMBOL(down_trylock);
+
+void up(struct semaphore *sem)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&sem->lock, flags);
+	if (likely(sem->count >= 0))
+		sem->count++;
+	else
+		__up(sem);
+	spin_unlock_irqrestore(&sem->lock, flags);
+}
+EXPORT_SYMBOL(up);
+
+/* Functions for the contended case */
+
+struct semaphore_waiter {
+	struct list_head list;
+	struct task_struct *task;
+	int up;
+};
+
+/*
+ * Wake up a process waiting on a semaphore.  We need to call this from both
+ * __up and __down_common as it's possible to race a task into the semaphore
+ * if it comes in at just the right time between two tasks calling up() and
+ * a third task waking up.  This function assumes the wait_list is already
+ * checked for being non-empty.
+ */
+static noinline void __sched __up_down_common(struct semaphore *sem)
+{
+	struct semaphore_waiter *waiter = list_first_entry(&sem->wait_list,
+						struct semaphore_waiter, list);
+	list_del(&waiter->list);
+	waiter->up = 1;
+	wake_up_process(waiter->task);
+}
+
+/*
+ * Because this function is inlined, the 'state' parameter will be constant,
+ * and thus optimised away by the compiler.
+ */
+static inline int __sched __down_common(struct semaphore *sem, long state)
+{
+	int result = 0;
+	struct task_struct *task = current;
+	struct semaphore_waiter waiter;
+
+	list_add_tail(&waiter.list, &sem->wait_list);
+	waiter.task = task;
+	waiter.up = 0;
+
+	for (;;) {
+		if (state == TASK_INTERRUPTIBLE && signal_pending(task))
+			goto interrupted;
+		__set_task_state(task, state);
+		spin_unlock_irq(&sem->lock);
+		schedule();
+		spin_lock_irq(&sem->lock);
+		if (waiter.up)
+			goto woken;
+	}
+
+ interrupted:
+	list_del(&waiter.list);
+	result = -EINTR;
+ woken:
+	/*
+	 * Account for the process which woke us up.  For the case where
+	 * we're interrupted, we need to increment the count on our own
+	 * behalf.  I don't believe we can hit the case where the
+	 * sem->count hits zero, *and* there's a second task sleeping,
+	 * but it doesn't hurt, that's not a commonly exercised path and
+	 * it's not a performance path either.
+	 */
+	if (unlikely((++sem->count >= 0) && !list_empty(&sem->wait_list)))
+		__up_down_common(sem);
+	return result;
+}
+
+static noinline void __sched __down(struct semaphore *sem)
+{
+	__down_common(sem, TASK_UNINTERRUPTIBLE);
+}
+
+static noinline int __sched __down_interruptible(struct semaphore *sem)
+{
+	return __down_common(sem, TASK_INTERRUPTIBLE);
+}
+
+static noinline void __sched __up(struct semaphore *sem)
+{
+	if (unlikely(list_empty(&sem->wait_list)))
+		sem->count++;
+	else
+		__up_down_common(sem);
+}
diff --git a/lib/Makefile b/lib/Makefile
index 23de261a4c83..28dba90d5020 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -29,7 +29,6 @@ obj-$(CONFIG_DEBUG_LOCKING_API_SELFTESTS) += locking-selftest.o
 obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o
 lib-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o
 lib-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem.o
-lib-$(CONFIG_SEMAPHORE_SLEEPERS) += semaphore-sleepers.o
 lib-$(CONFIG_GENERIC_FIND_NEXT_BIT) += find_next_bit.o
 obj-$(CONFIG_GENERIC_HWEIGHT) += hweight.o
 obj-$(CONFIG_LOCK_KERNEL) += kernel_lock.o
diff --git a/lib/semaphore-sleepers.c b/lib/semaphore-sleepers.c
deleted file mode 100644
index 0198782cdacb..000000000000
--- a/lib/semaphore-sleepers.c
+++ /dev/null
@@ -1,176 +0,0 @@
-/*
- * i386 and x86-64 semaphore implementation.
- *
- * (C) Copyright 1999 Linus Torvalds
- *
- * Portions Copyright 1999 Red Hat, Inc.
- *
- *	This program is free software; you can redistribute it and/or
- *	modify it under the terms of the GNU General Public License
- *	as published by the Free Software Foundation; either version
- *	2 of the License, or (at your option) any later version.
- *
- * rw semaphores implemented November 1999 by Benjamin LaHaise <bcrl@kvack.org>
- */
-#include <linux/sched.h>
-#include <linux/err.h>
-#include <linux/init.h>
-#include <asm/semaphore.h>
-
-/*
- * Semaphores are implemented using a two-way counter:
- * The "count" variable is decremented for each process
- * that tries to acquire the semaphore, while the "sleeping"
- * variable is a count of such acquires.
- *
- * Notably, the inline "up()" and "down()" functions can
- * efficiently test if they need to do any extra work (up
- * needs to do something only if count was negative before
- * the increment operation.
- *
- * "sleeping" and the contention routine ordering is protected
- * by the spinlock in the semaphore's waitqueue head.
- *
- * Note that these functions are only called when there is
- * contention on the lock, and as such all this is the
- * "non-critical" part of the whole semaphore business. The
- * critical part is the inline stuff in <asm/semaphore.h>
- * where we want to avoid any extra jumps and calls.
- */
-
-/*
- * Logic:
- *  - only on a boundary condition do we need to care. When we go
- *    from a negative count to a non-negative, we wake people up.
- *  - when we go from a non-negative count to a negative do we
- *    (a) synchronize with the "sleeper" count and (b) make sure
- *    that we're on the wakeup list before we synchronize so that
- *    we cannot lose wakeup events.
- */
-
-void __up(struct semaphore *sem)
-{
-	wake_up(&sem->wait);
-}
-
-void __sched __down(struct semaphore *sem)
-{
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	unsigned long flags;
-
-	tsk->state = TASK_UNINTERRUPTIBLE;
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock in
-		 * the wait_queue_head.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-		schedule();
-
-		spin_lock_irqsave(&sem->wait.lock, flags);
-		tsk->state = TASK_UNINTERRUPTIBLE;
-	}
-	remove_wait_queue_locked(&sem->wait, &wait);
-	wake_up_locked(&sem->wait);
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-	tsk->state = TASK_RUNNING;
-}
-
-int __sched __down_interruptible(struct semaphore *sem)
-{
-	int retval = 0;
-	struct task_struct *tsk = current;
-	DECLARE_WAITQUEUE(wait, tsk);
-	unsigned long flags;
-
-	tsk->state = TASK_INTERRUPTIBLE;
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	add_wait_queue_exclusive_locked(&sem->wait, &wait);
-
-	sem->sleepers++;
-	for (;;) {
-		int sleepers = sem->sleepers;
-
-		/*
-		 * With signals pending, this turns into
-		 * the trylock failure case - we won't be
-		 * sleeping, and we* can't get the lock as
-		 * it has contention. Just correct the count
-		 * and exit.
-		 */
-		if (signal_pending(current)) {
-			retval = -EINTR;
-			sem->sleepers = 0;
-			atomic_add(sleepers, &sem->count);
-			break;
-		}
-
-		/*
-		 * Add "everybody else" into it. They aren't
-		 * playing, because we own the spinlock in
-		 * wait_queue_head. The "-1" is because we're
-		 * still hoping to get the semaphore.
-		 */
-		if (!atomic_add_negative(sleepers - 1, &sem->count)) {
-			sem->sleepers = 0;
-			break;
-		}
-		sem->sleepers = 1;	/* us - see -1 above */
-		spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-		schedule();
-
-		spin_lock_irqsave(&sem->wait.lock, flags);
-		tsk->state = TASK_INTERRUPTIBLE;
-	}
-	remove_wait_queue_locked(&sem->wait, &wait);
-	wake_up_locked(&sem->wait);
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-
-	tsk->state = TASK_RUNNING;
-	return retval;
-}
-
-/*
- * Trylock failed - make sure we correct for
- * having decremented the count.
- *
- * We could have done the trylock with a
- * single "cmpxchg" without failure cases,
- * but then it wouldn't work on a 386.
- */
-int __down_trylock(struct semaphore *sem)
-{
-	int sleepers;
-	unsigned long flags;
-
-	spin_lock_irqsave(&sem->wait.lock, flags);
-	sleepers = sem->sleepers + 1;
-	sem->sleepers = 0;
-
-	/*
-	 * Add "everybody else" and us into it. They aren't
-	 * playing, because we own the spinlock in the
-	 * wait_queue_head.
-	 */
-	if (!atomic_add_negative(sleepers, &sem->count)) {
-		wake_up_locked(&sem->wait);
-	}
-
-	spin_unlock_irqrestore(&sem->wait.lock, flags);
-	return 1;
-}
-- 
cgit v1.2.3


From f06d96865861c3dd01520f47e2e61c899db1631f Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Fri, 14 Mar 2008 13:19:33 -0400
Subject: Introduce down_killable()

down_killable() is the functional counterpart of mutex_lock_killable.

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
---
 include/linux/semaphore.h |  6 ++++++
 kernel/semaphore.c        | 22 ++++++++++++++++++++++
 2 files changed, 28 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h
index b3c691b089b2..88f2a28cc0f1 100644
--- a/include/linux/semaphore.h
+++ b/include/linux/semaphore.h
@@ -61,6 +61,12 @@ extern void down(struct semaphore *sem);
  */
 extern int __must_check down_interruptible(struct semaphore *sem);
 
+/*
+ * As down_interruptible(), except the sleep may only be interrupted by
+ * signals which are fatal to this process.
+ */
+extern int __must_check down_killable(struct semaphore *sem);
+
 /*
  * As down(), except this function will not sleep.  It will return 0 if it
  * acquired the semaphore and 1 if the semaphore was contended.  This
diff --git a/kernel/semaphore.c b/kernel/semaphore.c
index d5a72702f261..2da2aed950f3 100644
--- a/kernel/semaphore.c
+++ b/kernel/semaphore.c
@@ -34,6 +34,7 @@
 
 static noinline void __down(struct semaphore *sem);
 static noinline int __down_interruptible(struct semaphore *sem);
+static noinline int __down_killable(struct semaphore *sem);
 static noinline void __up(struct semaphore *sem);
 
 void down(struct semaphore *sem)
@@ -61,6 +62,20 @@ int down_interruptible(struct semaphore *sem)
 }
 EXPORT_SYMBOL(down_interruptible);
 
+int down_killable(struct semaphore *sem)
+{
+	unsigned long flags;
+	int result = 0;
+
+	spin_lock_irqsave(&sem->lock, flags);
+	if (unlikely(sem->count-- <= 0))
+		result = __down_killable(sem);
+	spin_unlock_irqrestore(&sem->lock, flags);
+
+	return result;
+}
+EXPORT_SYMBOL(down_killable);
+
 /**
  * down_trylock - try to acquire the semaphore, without waiting
  * @sem: the semaphore to be acquired
@@ -143,6 +158,8 @@ static inline int __sched __down_common(struct semaphore *sem, long state)
 	for (;;) {
 		if (state == TASK_INTERRUPTIBLE && signal_pending(task))
 			goto interrupted;
+		if (state == TASK_KILLABLE && fatal_signal_pending(task))
+			goto interrupted;
 		__set_task_state(task, state);
 		spin_unlock_irq(&sem->lock);
 		schedule();
@@ -178,6 +195,11 @@ static noinline int __sched __down_interruptible(struct semaphore *sem)
 	return __down_common(sem, TASK_INTERRUPTIBLE);
 }
 
+static noinline int __sched __down_killable(struct semaphore *sem)
+{
+	return __down_common(sem, TASK_KILLABLE);
+}
+
 static noinline void __sched __up(struct semaphore *sem)
 {
 	if (unlikely(list_empty(&sem->wait_list)))
-- 
cgit v1.2.3


From f1241c87a16c4fe9f4f51d6ed3589f031c505e8d Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Fri, 14 Mar 2008 13:43:13 -0400
Subject: Add down_timeout and change ACPI to use it

ACPI currently emulates a timeout for semaphores with calls to
down_trylock and sleep.  This produces horrible behaviour in terms of
fairness and excessive wakeups.  Now that we have a unified semaphore
implementation, adding a real down_trylock is almost trivial.

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
---
 drivers/acpi/osl.c        | 89 +++++++++++------------------------------------
 include/linux/semaphore.h |  6 ++++
 kernel/semaphore.c        | 42 ++++++++++++++++++----
 3 files changed, 62 insertions(+), 75 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c
index a697fb6cf050..a498a6cc68fe 100644
--- a/drivers/acpi/osl.c
+++ b/drivers/acpi/osl.c
@@ -4,6 +4,8 @@
  *  Copyright (C) 2000       Andrew Henroid
  *  Copyright (C) 2001, 2002 Andy Grover <andrew.grover@intel.com>
  *  Copyright (C) 2001, 2002 Paul Diefenbaugh <paul.s.diefenbaugh@intel.com>
+ *  Copyright (c) 2008 Intel Corporation
+ *   Author: Matthew Wilcox <willy@linux.intel.com>
  *
  * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  *
@@ -37,15 +39,18 @@
 #include <linux/workqueue.h>
 #include <linux/nmi.h>
 #include <linux/acpi.h>
-#include <acpi/acpi.h>
-#include <asm/io.h>
-#include <acpi/acpi_bus.h>
-#include <acpi/processor.h>
-#include <asm/uaccess.h>
-
 #include <linux/efi.h>
 #include <linux/ioport.h>
 #include <linux/list.h>
+#include <linux/jiffies.h>
+#include <linux/semaphore.h>
+
+#include <asm/io.h>
+#include <asm/uaccess.h>
+
+#include <acpi/acpi.h>
+#include <acpi/acpi_bus.h>
+#include <acpi/processor.h>
 
 #define _COMPONENT		ACPI_OS_SERVICES
 ACPI_MODULE_NAME("osl");
@@ -764,7 +769,6 @@ acpi_os_create_semaphore(u32 max_units, u32 initial_units, acpi_handle * handle)
 {
 	struct semaphore *sem = NULL;
 
-
 	sem = acpi_os_allocate(sizeof(struct semaphore));
 	if (!sem)
 		return AE_NO_MEMORY;
@@ -791,12 +795,12 @@ acpi_status acpi_os_delete_semaphore(acpi_handle handle)
 {
 	struct semaphore *sem = (struct semaphore *)handle;
 
-
 	if (!sem)
 		return AE_BAD_PARAMETER;
 
 	ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "Deleting semaphore[%p].\n", handle));
 
+	BUG_ON(!list_empty(&sem->wait_list));
 	kfree(sem);
 	sem = NULL;
 
@@ -804,21 +808,15 @@ acpi_status acpi_os_delete_semaphore(acpi_handle handle)
 }
 
 /*
- * TODO: The kernel doesn't have a 'down_timeout' function -- had to
- * improvise.  The process is to sleep for one scheduler quantum
- * until the semaphore becomes available.  Downside is that this
- * may result in starvation for timeout-based waits when there's
- * lots of semaphore activity.
- *
  * TODO: Support for units > 1?
  */
 acpi_status acpi_os_wait_semaphore(acpi_handle handle, u32 units, u16 timeout)
 {
 	acpi_status status = AE_OK;
 	struct semaphore *sem = (struct semaphore *)handle;
+	long jiffies;
 	int ret = 0;
 
-
 	if (!sem || (units < 1))
 		return AE_BAD_PARAMETER;
 
@@ -828,58 +826,14 @@ acpi_status acpi_os_wait_semaphore(acpi_handle handle, u32 units, u16 timeout)
 	ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "Waiting for semaphore[%p|%d|%d]\n",
 			  handle, units, timeout));
 
-	/*
-	 * This can be called during resume with interrupts off.
-	 * Like boot-time, we should be single threaded and will
-	 * always get the lock if we try -- timeout or not.
-	 * If this doesn't succeed, then we will oops courtesy of
-	 * might_sleep() in down().
-	 */
-	if (!down_trylock(sem))
-		return AE_OK;
-
-	switch (timeout) {
-		/*
-		 * No Wait:
-		 * --------
-		 * A zero timeout value indicates that we shouldn't wait - just
-		 * acquire the semaphore if available otherwise return AE_TIME
-		 * (a.k.a. 'would block').
-		 */
-	case 0:
-		if (down_trylock(sem))
-			status = AE_TIME;
-		break;
-
-		/*
-		 * Wait Indefinitely:
-		 * ------------------
-		 */
-	case ACPI_WAIT_FOREVER:
-		down(sem);
-		break;
-
-		/*
-		 * Wait w/ Timeout:
-		 * ----------------
-		 */
-	default:
-		// TODO: A better timeout algorithm?
-		{
-			int i = 0;
-			static const int quantum_ms = 1000 / HZ;
-
-			ret = down_trylock(sem);
-			for (i = timeout; (i > 0 && ret != 0); i -= quantum_ms) {
-				schedule_timeout_interruptible(1);
-				ret = down_trylock(sem);
-			}
-
-			if (ret != 0)
-				status = AE_TIME;
-		}
-		break;
-	}
+	if (timeout == ACPI_WAIT_FOREVER)
+		jiffies = MAX_SCHEDULE_TIMEOUT;
+	else
+		jiffies = msecs_to_jiffies(timeout);
+	
+	ret = down_timeout(sem, jiffies);
+	if (ret)
+		status = AE_TIME;
 
 	if (ACPI_FAILURE(status)) {
 		ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
@@ -902,7 +856,6 @@ acpi_status acpi_os_signal_semaphore(acpi_handle handle, u32 units)
 {
 	struct semaphore *sem = (struct semaphore *)handle;
 
-
 	if (!sem || (units < 1))
 		return AE_BAD_PARAMETER;
 
diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h
index 88f2a28cc0f1..a107aebd9148 100644
--- a/include/linux/semaphore.h
+++ b/include/linux/semaphore.h
@@ -74,6 +74,12 @@ extern int __must_check down_killable(struct semaphore *sem);
  */
 extern int __must_check down_trylock(struct semaphore *sem);
 
+/*
+ * As down(), except this function will return -ETIME if it fails to
+ * acquire the semaphore within the specified number of jiffies.
+ */
+extern int __must_check down_timeout(struct semaphore *sem, long jiffies);
+
 /*
  * Release the semaphore.  Unlike mutexes, up() may be called from any
  * context and even by tasks which have never called down().
diff --git a/kernel/semaphore.c b/kernel/semaphore.c
index 2da2aed950f3..5a12a8558982 100644
--- a/kernel/semaphore.c
+++ b/kernel/semaphore.c
@@ -35,6 +35,7 @@
 static noinline void __down(struct semaphore *sem);
 static noinline int __down_interruptible(struct semaphore *sem);
 static noinline int __down_killable(struct semaphore *sem);
+static noinline int __down_timeout(struct semaphore *sem, long jiffies);
 static noinline void __up(struct semaphore *sem);
 
 void down(struct semaphore *sem)
@@ -104,6 +105,20 @@ int down_trylock(struct semaphore *sem)
 }
 EXPORT_SYMBOL(down_trylock);
 
+int down_timeout(struct semaphore *sem, long jiffies)
+{
+	unsigned long flags;
+	int result = 0;
+
+	spin_lock_irqsave(&sem->lock, flags);
+	if (unlikely(sem->count-- <= 0))
+		result = __down_timeout(sem, jiffies);
+	spin_unlock_irqrestore(&sem->lock, flags);
+
+	return result;
+}
+EXPORT_SYMBOL(down_timeout);
+
 void up(struct semaphore *sem)
 {
 	unsigned long flags;
@@ -142,10 +157,12 @@ static noinline void __sched __up_down_common(struct semaphore *sem)
 }
 
 /*
- * Because this function is inlined, the 'state' parameter will be constant,
- * and thus optimised away by the compiler.
+ * Because this function is inlined, the 'state' parameter will be
+ * constant, and thus optimised away by the compiler.  Likewise the
+ * 'timeout' parameter for the cases without timeouts.
  */
-static inline int __sched __down_common(struct semaphore *sem, long state)
+static inline int __sched __down_common(struct semaphore *sem, long state,
+								long timeout)
 {
 	int result = 0;
 	struct task_struct *task = current;
@@ -160,14 +177,20 @@ static inline int __sched __down_common(struct semaphore *sem, long state)
 			goto interrupted;
 		if (state == TASK_KILLABLE && fatal_signal_pending(task))
 			goto interrupted;
+		if (timeout <= 0)
+			goto timed_out;
 		__set_task_state(task, state);
 		spin_unlock_irq(&sem->lock);
-		schedule();
+		timeout = schedule_timeout(timeout);
 		spin_lock_irq(&sem->lock);
 		if (waiter.up)
 			goto woken;
 	}
 
+ timed_out:
+	list_del(&waiter.list);
+	result = -ETIME;
+	goto woken;
  interrupted:
 	list_del(&waiter.list);
 	result = -EINTR;
@@ -187,17 +210,22 @@ static inline int __sched __down_common(struct semaphore *sem, long state)
 
 static noinline void __sched __down(struct semaphore *sem)
 {
-	__down_common(sem, TASK_UNINTERRUPTIBLE);
+	__down_common(sem, TASK_UNINTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT);
 }
 
 static noinline int __sched __down_interruptible(struct semaphore *sem)
 {
-	return __down_common(sem, TASK_INTERRUPTIBLE);
+	return __down_common(sem, TASK_INTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT);
 }
 
 static noinline int __sched __down_killable(struct semaphore *sem)
 {
-	return __down_common(sem, TASK_KILLABLE);
+	return __down_common(sem, TASK_KILLABLE, MAX_SCHEDULE_TIMEOUT);
+}
+
+static noinline int __sched __down_timeout(struct semaphore *sem, long jiffies)
+{
+	return __down_common(sem, TASK_UNINTERRUPTIBLE, jiffies);
 }
 
 static noinline void __sched __up(struct semaphore *sem)
-- 
cgit v1.2.3


From b17170b2fac96705db3188f093f89e8e838418e4 Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Fri, 14 Mar 2008 14:35:22 -0400
Subject: Simplify semaphore implementation

By removing the negative values of 'count' and relying on the wait_list to
indicate whether we have any waiters, we can simplify the implementation
by removing the protection against an unlikely race condition.  Thanks to
David Howells for his suggestions.

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
---
 include/linux/semaphore.h |  9 ++----
 kernel/semaphore.c        | 78 +++++++++++++++--------------------------------
 2 files changed, 27 insertions(+), 60 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h
index a107aebd9148..a7125daaff90 100644
--- a/include/linux/semaphore.h
+++ b/include/linux/semaphore.h
@@ -15,15 +15,12 @@
 
 /*
  * The spinlock controls access to the other members of the semaphore.
- * 'count' is decremented by every task which calls down*() and incremented
- * by every call to up().  Thus, if it is positive, it indicates how many
- * more tasks may acquire the lock.  If it is negative, it indicates how
- * many tasks are waiting for the lock.  Tasks waiting for the lock are
- * kept on the wait_list.
+ * 'count' represents how many more tasks can acquire this semaphore.
+ * Tasks waiting for the lock are kept on the wait_list.
  */
 struct semaphore {
 	spinlock_t		lock;
-	int			count;
+	unsigned int		count;
 	struct list_head	wait_list;
 };
 
diff --git a/kernel/semaphore.c b/kernel/semaphore.c
index 5a12a8558982..bef977b16966 100644
--- a/kernel/semaphore.c
+++ b/kernel/semaphore.c
@@ -18,18 +18,8 @@
  * down_trylock() and up() can be called from interrupt context.
  * So we have to disable interrupts when taking the lock.
  *
- * The ->count variable, if positive, defines how many more tasks can
- * acquire the semaphore.  If negative, it represents how many tasks are
- * waiting on the semaphore (*).  If zero, no tasks are waiting, and no more
- * tasks can acquire the semaphore.
- *
- * (*) Except for the window between one task calling up() and the task
- * sleeping in a __down_common() waking up.  In order to avoid a third task
- * coming in and stealing the second task's wakeup, we leave the ->count
- * negative.  If we have a more complex situation, the ->count may become
- * zero or negative (eg a semaphore with count = 2, three tasks attempt to
- * acquire it, one sleeps, two finish and call up(), the second task to call
- * up() notices that the list is empty and just increments count).
+ * The ->count variable defines how many more tasks can acquire the
+ * semaphore.  If it's zero, there may be tasks waiting on the list.
  */
 
 static noinline void __down(struct semaphore *sem);
@@ -43,7 +33,9 @@ void down(struct semaphore *sem)
 	unsigned long flags;
 
 	spin_lock_irqsave(&sem->lock, flags);
-	if (unlikely(sem->count-- <= 0))
+	if (likely(sem->count > 0))
+		sem->count--;
+	else
 		__down(sem);
 	spin_unlock_irqrestore(&sem->lock, flags);
 }
@@ -55,7 +47,9 @@ int down_interruptible(struct semaphore *sem)
 	int result = 0;
 
 	spin_lock_irqsave(&sem->lock, flags);
-	if (unlikely(sem->count-- <= 0))
+	if (likely(sem->count > 0))
+		sem->count--;
+	else
 		result = __down_interruptible(sem);
 	spin_unlock_irqrestore(&sem->lock, flags);
 
@@ -69,7 +63,9 @@ int down_killable(struct semaphore *sem)
 	int result = 0;
 
 	spin_lock_irqsave(&sem->lock, flags);
-	if (unlikely(sem->count-- <= 0))
+	if (likely(sem->count > 0))
+		sem->count--;
+	else
 		result = __down_killable(sem);
 	spin_unlock_irqrestore(&sem->lock, flags);
 
@@ -111,7 +107,9 @@ int down_timeout(struct semaphore *sem, long jiffies)
 	int result = 0;
 
 	spin_lock_irqsave(&sem->lock, flags);
-	if (unlikely(sem->count-- <= 0))
+	if (likely(sem->count > 0))
+		sem->count--;
+	else
 		result = __down_timeout(sem, jiffies);
 	spin_unlock_irqrestore(&sem->lock, flags);
 
@@ -124,7 +122,7 @@ void up(struct semaphore *sem)
 	unsigned long flags;
 
 	spin_lock_irqsave(&sem->lock, flags);
-	if (likely(sem->count >= 0))
+	if (likely(list_empty(&sem->wait_list)))
 		sem->count++;
 	else
 		__up(sem);
@@ -140,22 +138,6 @@ struct semaphore_waiter {
 	int up;
 };
 
-/*
- * Wake up a process waiting on a semaphore.  We need to call this from both
- * __up and __down_common as it's possible to race a task into the semaphore
- * if it comes in at just the right time between two tasks calling up() and
- * a third task waking up.  This function assumes the wait_list is already
- * checked for being non-empty.
- */
-static noinline void __sched __up_down_common(struct semaphore *sem)
-{
-	struct semaphore_waiter *waiter = list_first_entry(&sem->wait_list,
-						struct semaphore_waiter, list);
-	list_del(&waiter->list);
-	waiter->up = 1;
-	wake_up_process(waiter->task);
-}
-
 /*
  * Because this function is inlined, the 'state' parameter will be
  * constant, and thus optimised away by the compiler.  Likewise the
@@ -164,7 +146,6 @@ static noinline void __sched __up_down_common(struct semaphore *sem)
 static inline int __sched __down_common(struct semaphore *sem, long state,
 								long timeout)
 {
-	int result = 0;
 	struct task_struct *task = current;
 	struct semaphore_waiter waiter;
 
@@ -184,28 +165,16 @@ static inline int __sched __down_common(struct semaphore *sem, long state,
 		timeout = schedule_timeout(timeout);
 		spin_lock_irq(&sem->lock);
 		if (waiter.up)
-			goto woken;
+			return 0;
 	}
 
  timed_out:
 	list_del(&waiter.list);
-	result = -ETIME;
-	goto woken;
+	return -ETIME;
+
  interrupted:
 	list_del(&waiter.list);
-	result = -EINTR;
- woken:
-	/*
-	 * Account for the process which woke us up.  For the case where
-	 * we're interrupted, we need to increment the count on our own
-	 * behalf.  I don't believe we can hit the case where the
-	 * sem->count hits zero, *and* there's a second task sleeping,
-	 * but it doesn't hurt, that's not a commonly exercised path and
-	 * it's not a performance path either.
-	 */
-	if (unlikely((++sem->count >= 0) && !list_empty(&sem->wait_list)))
-		__up_down_common(sem);
-	return result;
+	return -EINTR;
 }
 
 static noinline void __sched __down(struct semaphore *sem)
@@ -230,8 +199,9 @@ static noinline int __sched __down_timeout(struct semaphore *sem, long jiffies)
 
 static noinline void __sched __up(struct semaphore *sem)
 {
-	if (unlikely(list_empty(&sem->wait_list)))
-		sem->count++;
-	else
-		__up_down_common(sem);
+	struct semaphore_waiter *waiter = list_first_entry(&sem->wait_list,
+						struct semaphore_waiter, list);
+	list_del(&waiter->list);
+	waiter->up = 1;
+	wake_up_process(waiter->task);
 }
-- 
cgit v1.2.3


From 714493cd5468f42ca3c4f730a9c17c203abd5059 Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Fri, 11 Apr 2008 15:23:52 -0400
Subject: Improve semaphore documentation

Move documentation from semaphore.h to semaphore.c as requested by
Andrew Morton.  Also reformat to kernel-doc style and add some more
notes about the implementation.

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
---
 include/linux/semaphore.h | 39 ++---------------------
 kernel/semaphore.c        | 79 ++++++++++++++++++++++++++++++++++++++++-------
 2 files changed, 70 insertions(+), 48 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h
index a7125daaff90..9cae64b00d6b 100644
--- a/include/linux/semaphore.h
+++ b/include/linux/semaphore.h
@@ -4,8 +4,7 @@
  *
  * Distributed under the terms of the GNU GPL, version 2
  *
- * Counting semaphores allow up to <n> tasks to acquire the semaphore
- * simultaneously.
+ * Please see kernel/semaphore.c for documentation of these functions
  */
 #ifndef __LINUX_SEMAPHORE_H
 #define __LINUX_SEMAPHORE_H
@@ -13,11 +12,7 @@
 #include <linux/list.h>
 #include <linux/spinlock.h>
 
-/*
- * The spinlock controls access to the other members of the semaphore.
- * 'count' represents how many more tasks can acquire this semaphore.
- * Tasks waiting for the lock are kept on the wait_list.
- */
+/* Please don't access any members of this structure directly */
 struct semaphore {
 	spinlock_t		lock;
 	unsigned int		count;
@@ -46,41 +41,11 @@ static inline void sema_init(struct semaphore *sem, int val)
 #define init_MUTEX(sem)		sema_init(sem, 1)
 #define init_MUTEX_LOCKED(sem)	sema_init(sem, 0)
 
-/*
- * Attempt to acquire the semaphore.  If another task is already holding the
- * semaphore, sleep until the semaphore is released.
- */
 extern void down(struct semaphore *sem);
-
-/*
- * As down(), except the sleep may be interrupted by a signal.  If it is,
- * this function will return -EINTR.
- */
 extern int __must_check down_interruptible(struct semaphore *sem);
-
-/*
- * As down_interruptible(), except the sleep may only be interrupted by
- * signals which are fatal to this process.
- */
 extern int __must_check down_killable(struct semaphore *sem);
-
-/*
- * As down(), except this function will not sleep.  It will return 0 if it
- * acquired the semaphore and 1 if the semaphore was contended.  This
- * function may be called from any context, including interrupt and softirq.
- */
 extern int __must_check down_trylock(struct semaphore *sem);
-
-/*
- * As down(), except this function will return -ETIME if it fails to
- * acquire the semaphore within the specified number of jiffies.
- */
 extern int __must_check down_timeout(struct semaphore *sem, long jiffies);
-
-/*
- * Release the semaphore.  Unlike mutexes, up() may be called from any
- * context and even by tasks which have never called down().
- */
 extern void up(struct semaphore *sem);
 
 #endif /* __LINUX_SEMAPHORE_H */
diff --git a/kernel/semaphore.c b/kernel/semaphore.c
index bef977b16966..5c2942e768cd 100644
--- a/kernel/semaphore.c
+++ b/kernel/semaphore.c
@@ -3,6 +3,26 @@
  * Author: Matthew Wilcox <willy@linux.intel.com>
  *
  * Distributed under the terms of the GNU GPL, version 2
+ *
+ * This file implements counting semaphores.
+ * A counting semaphore may be acquired 'n' times before sleeping.
+ * See mutex.c for single-acquisition sleeping locks which enforce
+ * rules which allow code to be debugged more easily.
+ */
+
+/*
+ * Some notes on the implementation:
+ *
+ * The spinlock controls access to the other members of the semaphore.
+ * down_trylock() and up() can be called from interrupt context, so we
+ * have to disable interrupts when taking the lock.  It turns out various
+ * parts of the kernel expect to be able to use down() on a semaphore in
+ * interrupt context when they know it will succeed, so we have to use
+ * irqsave variants for down(), down_interruptible() and down_killable()
+ * too.
+ *
+ * The ->count variable represents how many more tasks can acquire this
+ * semaphore.  If it's zero, there may be tasks waiting on the wait_list.
  */
 
 #include <linux/compiler.h>
@@ -12,22 +32,23 @@
 #include <linux/semaphore.h>
 #include <linux/spinlock.h>
 
-/*
- * Some notes on the implementation:
- *
- * down_trylock() and up() can be called from interrupt context.
- * So we have to disable interrupts when taking the lock.
- *
- * The ->count variable defines how many more tasks can acquire the
- * semaphore.  If it's zero, there may be tasks waiting on the list.
- */
-
 static noinline void __down(struct semaphore *sem);
 static noinline int __down_interruptible(struct semaphore *sem);
 static noinline int __down_killable(struct semaphore *sem);
 static noinline int __down_timeout(struct semaphore *sem, long jiffies);
 static noinline void __up(struct semaphore *sem);
 
+/**
+ * down - acquire the semaphore
+ * @sem: the semaphore to be acquired
+ *
+ * Acquires the semaphore.  If no more tasks are allowed to acquire the
+ * semaphore, calling this function will put the task to sleep until the
+ * semaphore is released.
+ *
+ * Use of this function is deprecated, please use down_interruptible() or
+ * down_killable() instead.
+ */
 void down(struct semaphore *sem)
 {
 	unsigned long flags;
@@ -41,6 +62,15 @@ void down(struct semaphore *sem)
 }
 EXPORT_SYMBOL(down);
 
+/**
+ * down_interruptible - acquire the semaphore unless interrupted
+ * @sem: the semaphore to be acquired
+ *
+ * Attempts to acquire the semaphore.  If no more tasks are allowed to
+ * acquire the semaphore, calling this function will put the task to sleep.
+ * If the sleep is interrupted by a signal, this function will return -EINTR.
+ * If the semaphore is successfully acquired, this function returns 0.
+ */
 int down_interruptible(struct semaphore *sem)
 {
 	unsigned long flags;
@@ -57,6 +87,16 @@ int down_interruptible(struct semaphore *sem)
 }
 EXPORT_SYMBOL(down_interruptible);
 
+/**
+ * down_killable - acquire the semaphore unless killed
+ * @sem: the semaphore to be acquired
+ *
+ * Attempts to acquire the semaphore.  If no more tasks are allowed to
+ * acquire the semaphore, calling this function will put the task to sleep.
+ * If the sleep is interrupted by a fatal signal, this function will return
+ * -EINTR.  If the semaphore is successfully acquired, this function returns
+ * 0.
+ */
 int down_killable(struct semaphore *sem)
 {
 	unsigned long flags;
@@ -78,7 +118,7 @@ EXPORT_SYMBOL(down_killable);
  * @sem: the semaphore to be acquired
  *
  * Try to acquire the semaphore atomically.  Returns 0 if the mutex has
- * been acquired successfully and 1 if it is contended.
+ * been acquired successfully or 1 if it it cannot be acquired.
  *
  * NOTE: This return value is inverted from both spin_trylock and
  * mutex_trylock!  Be careful about this when converting code.
@@ -101,6 +141,16 @@ int down_trylock(struct semaphore *sem)
 }
 EXPORT_SYMBOL(down_trylock);
 
+/**
+ * down_timeout - acquire the semaphore within a specified time
+ * @sem: the semaphore to be acquired
+ * @jiffies: how long to wait before failing
+ *
+ * Attempts to acquire the semaphore.  If no more tasks are allowed to
+ * acquire the semaphore, calling this function will put the task to sleep.
+ * If the semaphore is not released within the specified number of jiffies,
+ * this function returns -ETIME.  It returns 0 if the semaphore was acquired.
+ */
 int down_timeout(struct semaphore *sem, long jiffies)
 {
 	unsigned long flags;
@@ -117,6 +167,13 @@ int down_timeout(struct semaphore *sem, long jiffies)
 }
 EXPORT_SYMBOL(down_timeout);
 
+/**
+ * up - release the semaphore
+ * @sem: the semaphore to release
+ *
+ * Release the semaphore.  Unlike mutexes, up() may be called from any
+ * context and even by tasks which have never called down().
+ */
 void up(struct semaphore *sem)
 {
 	unsigned long flags;
-- 
cgit v1.2.3


From c33fa9f5609e918824446ef9a75319d4a802f1f4 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Thu, 17 Apr 2008 20:05:36 +0200
Subject: uaccess: add probe_kernel_write()

add probe_kernel_read() and probe_kernel_write().

Uninlined and restricted to kernel range memory only, as suggested
by Linus.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/uaccess.h | 22 ++++++++++++++++++++++
 mm/Makefile             |  2 +-
 mm/maccess.c            | 49 +++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 72 insertions(+), 1 deletion(-)
 create mode 100644 mm/maccess.c

(limited to 'include/linux')

diff --git a/include/linux/uaccess.h b/include/linux/uaccess.h
index 975c963e5789..fec6decfb983 100644
--- a/include/linux/uaccess.h
+++ b/include/linux/uaccess.h
@@ -84,4 +84,26 @@ static inline unsigned long __copy_from_user_nocache(void *to,
 		ret;					\
 	})
 
+/*
+ * probe_kernel_read(): safely attempt to read from a location
+ * @dst: pointer to the buffer that shall take the data
+ * @src: address to read from
+ * @size: size of the data chunk
+ *
+ * Safely read from address @src to the buffer at @dst.  If a kernel fault
+ * happens, handle that and return -EFAULT.
+ */
+extern long probe_kernel_read(void *dst, void *src, size_t size);
+
+/*
+ * probe_kernel_write(): safely attempt to write to a location
+ * @dst: address to write to
+ * @src: pointer to the data that shall be written
+ * @size: size of the data chunk
+ *
+ * Safely write to address @dst from the buffer at @src.  If a kernel fault
+ * happens, handle that and return -EFAULT.
+ */
+extern long probe_kernel_write(void *dst, void *src, size_t size);
+
 #endif		/* __LINUX_UACCESS_H__ */
diff --git a/mm/Makefile b/mm/Makefile
index a5b0dd93427a..18c143b3c46c 100644
--- a/mm/Makefile
+++ b/mm/Makefile
@@ -8,7 +8,7 @@ mmu-$(CONFIG_MMU)	:= fremap.o highmem.o madvise.o memory.o mincore.o \
 			   vmalloc.o
 
 obj-y			:= bootmem.o filemap.o mempool.o oom_kill.o fadvise.o \
-			   page_alloc.o page-writeback.o pdflush.o \
+			   maccess.o page_alloc.o page-writeback.o pdflush.o \
 			   readahead.o swap.o truncate.o vmscan.o \
 			   prio_tree.o util.o mmzone.o vmstat.o backing-dev.o \
 			   page_isolation.o $(mmu-y)
diff --git a/mm/maccess.c b/mm/maccess.c
new file mode 100644
index 000000000000..24f81b971403
--- /dev/null
+++ b/mm/maccess.c
@@ -0,0 +1,49 @@
+/*
+ * Access kernel memory without faulting.
+ */
+#include <linux/uaccess.h>
+#include <linux/module.h>
+#include <linux/mm.h>
+
+/**
+ * probe_kernel_read(): safely attempt to read from a location
+ * @dst: pointer to the buffer that shall take the data
+ * @src: address to read from
+ * @size: size of the data chunk
+ *
+ * Safely read from address @src to the buffer at @dst.  If a kernel fault
+ * happens, handle that and return -EFAULT.
+ */
+long probe_kernel_read(void *dst, void *src, size_t size)
+{
+	long ret;
+
+	pagefault_disable();
+	ret = __copy_from_user_inatomic(dst,
+			(__force const void __user *)src, size);
+	pagefault_enable();
+
+	return ret ? -EFAULT : 0;
+}
+EXPORT_SYMBOL_GPL(probe_kernel_read);
+
+/**
+ * probe_kernel_write(): safely attempt to write to a location
+ * @dst: address to write to
+ * @src: pointer to the data that shall be written
+ * @size: size of the data chunk
+ *
+ * Safely write to address @dst from the buffer at @src.  If a kernel fault
+ * happens, handle that and return -EFAULT.
+ */
+long probe_kernel_write(void *dst, void *src, size_t size)
+{
+	long ret;
+
+	pagefault_disable();
+	ret = __copy_to_user_inatomic((__force void __user *)dst, src, size);
+	pagefault_enable();
+
+	return ret ? -EFAULT : 0;
+}
+EXPORT_SYMBOL_GPL(probe_kernel_write);
-- 
cgit v1.2.3


From dc7d552705215ac50a0617fcf51bb9c736255b8e Mon Sep 17 00:00:00 2001
From: Jason Wessel <jason.wessel@windriver.com>
Date: Thu, 17 Apr 2008 20:05:37 +0200
Subject: kgdb: core

kgdb core code. Handles the protocol and the arch details.

[ mingo@elte.hu: heavily modified, simplified and cleaned up. ]
[ xemul@openvz.org: use find_task_by_pid_ns ]

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/kgdb.h |  271 ++++++++
 kernel/Makefile      |    1 +
 kernel/kgdb.c        | 1693 ++++++++++++++++++++++++++++++++++++++++++++++++++
 lib/Kconfig.debug    |    2 +
 lib/Kconfig.kgdb     |   27 +
 5 files changed, 1994 insertions(+)
 create mode 100644 include/linux/kgdb.h
 create mode 100644 kernel/kgdb.c
 create mode 100644 lib/Kconfig.kgdb

(limited to 'include/linux')

diff --git a/include/linux/kgdb.h b/include/linux/kgdb.h
new file mode 100644
index 000000000000..b0985b79b638
--- /dev/null
+++ b/include/linux/kgdb.h
@@ -0,0 +1,271 @@
+/*
+ * This provides the callbacks and functions that KGDB needs to share between
+ * the core, I/O and arch-specific portions.
+ *
+ * Author: Amit Kale <amitkale@linsyssoft.com> and
+ *         Tom Rini <trini@kernel.crashing.org>
+ *
+ * 2001-2004 (c) Amit S. Kale and 2003-2005 (c) MontaVista Software, Inc.
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2. This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+#ifndef _KGDB_H_
+#define _KGDB_H_
+
+#include <linux/serial_8250.h>
+#include <linux/linkage.h>
+#include <linux/init.h>
+
+#include <asm/atomic.h>
+#include <asm/kgdb.h>
+
+struct pt_regs;
+
+/*
+ *	kgdb_skipexception - Bail out of KGDB when we've been triggered.
+ *	@exception: Exception vector number
+ *	@regs: Current &struct pt_regs.
+ *
+ *	On some architectures we need to skip a breakpoint exception when
+ *	it occurs after a breakpoint has been removed.
+ */
+extern int kgdb_skipexception(int exception, struct pt_regs *regs);
+
+/*
+ *	kgdb_post_primary_code - Save error vector/code numbers.
+ *	@regs: Original pt_regs.
+ *	@e_vector: Original error vector.
+ *	@err_code: Original error code.
+ *
+ *	This is needed on architectures which support SMP and KGDB.
+ *	This function is called after all the secondary cpus have been put
+ *	to a know spin state and the primary CPU has control over KGDB.
+ */
+extern void kgdb_post_primary_code(struct pt_regs *regs, int e_vector,
+				  int err_code);
+
+/*
+ *	kgdb_disable_hw_debug - Disable hardware debugging while we in kgdb.
+ *	@regs: Current &struct pt_regs.
+ *
+ *	This function will be called if the particular architecture must
+ *	disable hardware debugging while it is processing gdb packets or
+ *	handling exception.
+ */
+extern void kgdb_disable_hw_debug(struct pt_regs *regs);
+
+struct tasklet_struct;
+struct task_struct;
+struct uart_port;
+
+/* To enter the debugger explicitly. */
+void kgdb_breakpoint(void);
+
+extern int kgdb_connected;
+
+extern atomic_t			kgdb_setting_breakpoint;
+extern atomic_t			kgdb_cpu_doing_single_step;
+
+extern struct task_struct	*kgdb_usethread;
+extern struct task_struct	*kgdb_contthread;
+
+enum kgdb_bptype {
+	BP_BREAKPOINT = 0,
+	BP_HARDWARE_BREAKPOINT,
+	BP_WRITE_WATCHPOINT,
+	BP_READ_WATCHPOINT,
+	BP_ACCESS_WATCHPOINT
+};
+
+enum kgdb_bpstate {
+	BP_UNDEFINED = 0,
+	BP_REMOVED,
+	BP_SET,
+	BP_ACTIVE
+};
+
+struct kgdb_bkpt {
+	unsigned long		bpt_addr;
+	unsigned char		saved_instr[BREAK_INSTR_SIZE];
+	enum kgdb_bptype	type;
+	enum kgdb_bpstate	state;
+};
+
+#ifndef KGDB_MAX_BREAKPOINTS
+# define KGDB_MAX_BREAKPOINTS	1000
+#endif
+
+#define KGDB_HW_BREAKPOINT	1
+
+/*
+ * Functions each KGDB-supporting architecture must provide:
+ */
+
+/*
+ *	kgdb_arch_init - Perform any architecture specific initalization.
+ *
+ *	This function will handle the initalization of any architecture
+ *	specific callbacks.
+ */
+extern int kgdb_arch_init(void);
+
+/*
+ *	kgdb_arch_exit - Perform any architecture specific uninitalization.
+ *
+ *	This function will handle the uninitalization of any architecture
+ *	specific callbacks, for dynamic registration and unregistration.
+ */
+extern void kgdb_arch_exit(void);
+
+/*
+ *	pt_regs_to_gdb_regs - Convert ptrace regs to GDB regs
+ *	@gdb_regs: A pointer to hold the registers in the order GDB wants.
+ *	@regs: The &struct pt_regs of the current process.
+ *
+ *	Convert the pt_regs in @regs into the format for registers that
+ *	GDB expects, stored in @gdb_regs.
+ */
+extern void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs);
+
+/*
+ *	sleeping_thread_to_gdb_regs - Convert ptrace regs to GDB regs
+ *	@gdb_regs: A pointer to hold the registers in the order GDB wants.
+ *	@p: The &struct task_struct of the desired process.
+ *
+ *	Convert the register values of the sleeping process in @p to
+ *	the format that GDB expects.
+ *	This function is called when kgdb does not have access to the
+ *	&struct pt_regs and therefore it should fill the gdb registers
+ *	@gdb_regs with what has	been saved in &struct thread_struct
+ *	thread field during switch_to.
+ */
+extern void
+sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p);
+
+/*
+ *	gdb_regs_to_pt_regs - Convert GDB regs to ptrace regs.
+ *	@gdb_regs: A pointer to hold the registers we've received from GDB.
+ *	@regs: A pointer to a &struct pt_regs to hold these values in.
+ *
+ *	Convert the GDB regs in @gdb_regs into the pt_regs, and store them
+ *	in @regs.
+ */
+extern void gdb_regs_to_pt_regs(unsigned long *gdb_regs, struct pt_regs *regs);
+
+/*
+ *	kgdb_arch_handle_exception - Handle architecture specific GDB packets.
+ *	@vector: The error vector of the exception that happened.
+ *	@signo: The signal number of the exception that happened.
+ *	@err_code: The error code of the exception that happened.
+ *	@remcom_in_buffer: The buffer of the packet we have read.
+ *	@remcom_out_buffer: The buffer of %BUFMAX bytes to write a packet into.
+ *	@regs: The &struct pt_regs of the current process.
+ *
+ *	This function MUST handle the 'c' and 's' command packets,
+ *	as well packets to set / remove a hardware breakpoint, if used.
+ *	If there are additional packets which the hardware needs to handle,
+ *	they are handled here.  The code should return -1 if it wants to
+ *	process more packets, and a %0 or %1 if it wants to exit from the
+ *	kgdb callback.
+ */
+extern int
+kgdb_arch_handle_exception(int vector, int signo, int err_code,
+			   char *remcom_in_buffer,
+			   char *remcom_out_buffer,
+			   struct pt_regs *regs);
+
+/*
+ *	kgdb_roundup_cpus - Get other CPUs into a holding pattern
+ *	@flags: Current IRQ state
+ *
+ *	On SMP systems, we need to get the attention of the other CPUs
+ *	and get them be in a known state.  This should do what is needed
+ *	to get the other CPUs to call kgdb_wait(). Note that on some arches,
+ *	the NMI approach is not used for rounding up all the CPUs. For example,
+ *	in case of MIPS, smp_call_function() is used to roundup CPUs. In
+ *	this case, we have to make sure that interrupts are enabled before
+ *	calling smp_call_function(). The argument to this function is
+ *	the flags that will be used when restoring the interrupts. There is
+ *	local_irq_save() call before kgdb_roundup_cpus().
+ *
+ *	On non-SMP systems, this is not called.
+ */
+extern void kgdb_roundup_cpus(unsigned long flags);
+
+/* Optional functions. */
+extern int kgdb_validate_break_address(unsigned long addr);
+extern int kgdb_arch_set_breakpoint(unsigned long addr, char *saved_instr);
+extern int kgdb_arch_remove_breakpoint(unsigned long addr, char *bundle);
+
+/*
+ * struct kgdb_arch - Describe architecture specific values.
+ * @gdb_bpt_instr: The instruction to trigger a breakpoint.
+ * @flags: Flags for the breakpoint, currently just %KGDB_HW_BREAKPOINT.
+ * @set_breakpoint: Allow an architecture to specify how to set a software
+ * breakpoint.
+ * @remove_breakpoint: Allow an architecture to specify how to remove a
+ * software breakpoint.
+ * @set_hw_breakpoint: Allow an architecture to specify how to set a hardware
+ * breakpoint.
+ * @remove_hw_breakpoint: Allow an architecture to specify how to remove a
+ * hardware breakpoint.
+ * @remove_all_hw_break: Allow an architecture to specify how to remove all
+ * hardware breakpoints.
+ * @correct_hw_break: Allow an architecture to specify how to correct the
+ * hardware debug registers.
+ */
+struct kgdb_arch {
+	unsigned char		gdb_bpt_instr[BREAK_INSTR_SIZE];
+	unsigned long		flags;
+
+	int	(*set_breakpoint)(unsigned long, char *);
+	int	(*remove_breakpoint)(unsigned long, char *);
+	int	(*set_hw_breakpoint)(unsigned long, int, enum kgdb_bptype);
+	int	(*remove_hw_breakpoint)(unsigned long, int, enum kgdb_bptype);
+	void	(*remove_all_hw_break)(void);
+	void	(*correct_hw_break)(void);
+};
+
+/*
+ * struct kgdb_io - Describe the interface for an I/O driver to talk with KGDB.
+ * @name: Name of the I/O driver.
+ * @read_char: Pointer to a function that will return one char.
+ * @write_char: Pointer to a function that will write one char.
+ * @flush: Pointer to a function that will flush any pending writes.
+ * @init: Pointer to a function that will initialize the device.
+ * @pre_exception: Pointer to a function that will do any prep work for
+ * the I/O driver.
+ * @post_exception: Pointer to a function that will do any cleanup work
+ * for the I/O driver.
+ */
+struct kgdb_io {
+	const char		*name;
+	int			(*read_char) (void);
+	void			(*write_char) (u8);
+	void			(*flush) (void);
+	int			(*init) (void);
+	void			(*pre_exception) (void);
+	void			(*post_exception) (void);
+};
+
+extern struct kgdb_arch		arch_kgdb_ops;
+
+extern int kgdb_register_io_module(struct kgdb_io *local_kgdb_io_ops);
+extern void kgdb_unregister_io_module(struct kgdb_io *local_kgdb_io_ops);
+
+extern int kgdb_hex2long(char **ptr, long *long_val);
+extern int kgdb_mem2hex(char *mem, char *buf, int count);
+extern int kgdb_hex2mem(char *buf, char *mem, int count);
+
+extern int kgdb_isremovedbreak(unsigned long addr);
+
+extern int
+kgdb_handle_exception(int ex_vector, int signo, int err_code,
+		      struct pt_regs *regs);
+extern int kgdb_nmicallback(int cpu, void *regs);
+
+extern int			kgdb_single_step;
+extern atomic_t			kgdb_active;
+
+#endif /* _KGDB_H_ */
diff --git a/kernel/Makefile b/kernel/Makefile
index 6c584c55a6e9..05c8003718ee 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -53,6 +53,7 @@ obj-$(CONFIG_AUDIT) += audit.o auditfilter.o
 obj-$(CONFIG_AUDITSYSCALL) += auditsc.o
 obj-$(CONFIG_AUDIT_TREE) += audit_tree.o
 obj-$(CONFIG_KPROBES) += kprobes.o
+obj-$(CONFIG_KGDB) += kgdb.o
 obj-$(CONFIG_DETECT_SOFTLOCKUP) += softlockup.o
 obj-$(CONFIG_GENERIC_HARDIRQS) += irq/
 obj-$(CONFIG_SECCOMP) += seccomp.o
diff --git a/kernel/kgdb.c b/kernel/kgdb.c
new file mode 100644
index 000000000000..017ee782bc08
--- /dev/null
+++ b/kernel/kgdb.c
@@ -0,0 +1,1693 @@
+/*
+ * KGDB stub.
+ *
+ * Maintainer: Jason Wessel <jason.wessel@windriver.com>
+ *
+ * Copyright (C) 2000-2001 VERITAS Software Corporation.
+ * Copyright (C) 2002-2004 Timesys Corporation
+ * Copyright (C) 2003-2004 Amit S. Kale <amitkale@linsyssoft.com>
+ * Copyright (C) 2004 Pavel Machek <pavel@suse.cz>
+ * Copyright (C) 2004-2006 Tom Rini <trini@kernel.crashing.org>
+ * Copyright (C) 2004-2006 LinSysSoft Technologies Pvt. Ltd.
+ * Copyright (C) 2005-2008 Wind River Systems, Inc.
+ * Copyright (C) 2007 MontaVista Software, Inc.
+ * Copyright (C) 2008 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
+ *
+ * Contributors at various stages not listed above:
+ *  Jason Wessel ( jason.wessel@windriver.com )
+ *  George Anzinger <george@mvista.com>
+ *  Anurekh Saxena (anurekh.saxena@timesys.com)
+ *  Lake Stevens Instrument Division (Glenn Engel)
+ *  Jim Kingdon, Cygnus Support.
+ *
+ * Original KGDB stub: David Grothe <dave@gcom.com>,
+ * Tigran Aivazian <tigran@sco.com>
+ *
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2. This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+#include <linux/pid_namespace.h>
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/console.h>
+#include <linux/threads.h>
+#include <linux/uaccess.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/ptrace.h>
+#include <linux/reboot.h>
+#include <linux/string.h>
+#include <linux/delay.h>
+#include <linux/sched.h>
+#include <linux/sysrq.h>
+#include <linux/init.h>
+#include <linux/kgdb.h>
+#include <linux/pid.h>
+#include <linux/smp.h>
+#include <linux/mm.h>
+
+#include <asm/cacheflush.h>
+#include <asm/byteorder.h>
+#include <asm/atomic.h>
+#include <asm/system.h>
+
+static int kgdb_break_asap;
+
+struct kgdb_state {
+	int			ex_vector;
+	int			signo;
+	int			err_code;
+	int			cpu;
+	int			pass_exception;
+	long			threadid;
+	long			kgdb_usethreadid;
+	struct pt_regs		*linux_regs;
+};
+
+static struct debuggerinfo_struct {
+	void			*debuggerinfo;
+	struct task_struct	*task;
+} kgdb_info[NR_CPUS];
+
+/**
+ * kgdb_connected - Is a host GDB connected to us?
+ */
+int				kgdb_connected;
+EXPORT_SYMBOL_GPL(kgdb_connected);
+
+/* All the KGDB handlers are installed */
+static int			kgdb_io_module_registered;
+
+/* Guard for recursive entry */
+static int			exception_level;
+
+static struct kgdb_io		*kgdb_io_ops;
+static DEFINE_SPINLOCK(kgdb_registration_lock);
+
+/* kgdb console driver is loaded */
+static int kgdb_con_registered;
+/* determine if kgdb console output should be used */
+static int kgdb_use_con;
+
+static int __init opt_kgdb_con(char *str)
+{
+	kgdb_use_con = 1;
+	return 0;
+}
+
+early_param("kgdbcon", opt_kgdb_con);
+
+module_param(kgdb_use_con, int, 0644);
+
+/*
+ * Holds information about breakpoints in a kernel. These breakpoints are
+ * added and removed by gdb.
+ */
+static struct kgdb_bkpt		kgdb_break[KGDB_MAX_BREAKPOINTS] = {
+	[0 ... KGDB_MAX_BREAKPOINTS-1] = { .state = BP_UNDEFINED }
+};
+
+/*
+ * The CPU# of the active CPU, or -1 if none:
+ */
+atomic_t			kgdb_active = ATOMIC_INIT(-1);
+
+/*
+ * We use NR_CPUs not PERCPU, in case kgdb is used to debug early
+ * bootup code (which might not have percpu set up yet):
+ */
+static atomic_t			passive_cpu_wait[NR_CPUS];
+static atomic_t			cpu_in_kgdb[NR_CPUS];
+atomic_t			kgdb_setting_breakpoint;
+
+struct task_struct		*kgdb_usethread;
+struct task_struct		*kgdb_contthread;
+
+int				kgdb_single_step;
+
+/* Our I/O buffers. */
+static char			remcom_in_buffer[BUFMAX];
+static char			remcom_out_buffer[BUFMAX];
+
+/* Storage for the registers, in GDB format. */
+static unsigned long		gdb_regs[(NUMREGBYTES +
+					sizeof(unsigned long) - 1) /
+					sizeof(unsigned long)];
+
+/* to keep track of the CPU which is doing the single stepping*/
+atomic_t			kgdb_cpu_doing_single_step = ATOMIC_INIT(-1);
+
+/*
+ * If you are debugging a problem where roundup (the collection of
+ * all other CPUs) is a problem [this should be extremely rare],
+ * then use the nokgdbroundup option to avoid roundup. In that case
+ * the other CPUs might interfere with your debugging context, so
+ * use this with care:
+ */
+int				kgdb_do_roundup = 1;
+
+static int __init opt_nokgdbroundup(char *str)
+{
+	kgdb_do_roundup = 0;
+
+	return 0;
+}
+
+early_param("nokgdbroundup", opt_nokgdbroundup);
+
+/*
+ * Finally, some KGDB code :-)
+ */
+
+/*
+ * Weak aliases for breakpoint management,
+ * can be overriden by architectures when needed:
+ */
+int __weak kgdb_validate_break_address(unsigned long addr)
+{
+	char tmp_variable[BREAK_INSTR_SIZE];
+
+	return probe_kernel_read(tmp_variable, (char *)addr, BREAK_INSTR_SIZE);
+}
+
+int __weak kgdb_arch_set_breakpoint(unsigned long addr, char *saved_instr)
+{
+	int err;
+
+	err = probe_kernel_read(saved_instr, (char *)addr, BREAK_INSTR_SIZE);
+	if (err)
+		return err;
+
+	return probe_kernel_write((char *)addr, arch_kgdb_ops.gdb_bpt_instr,
+				  BREAK_INSTR_SIZE);
+}
+
+int __weak kgdb_arch_remove_breakpoint(unsigned long addr, char *bundle)
+{
+	return probe_kernel_write((char *)addr,
+				  (char *)bundle, BREAK_INSTR_SIZE);
+}
+
+unsigned long __weak kgdb_arch_pc(int exception, struct pt_regs *regs)
+{
+	return instruction_pointer(regs);
+}
+
+int __weak kgdb_arch_init(void)
+{
+	return 0;
+}
+
+/**
+ *	kgdb_disable_hw_debug - Disable hardware debugging while we in kgdb.
+ *	@regs: Current &struct pt_regs.
+ *
+ *	This function will be called if the particular architecture must
+ *	disable hardware debugging while it is processing gdb packets or
+ *	handling exception.
+ */
+void __weak kgdb_disable_hw_debug(struct pt_regs *regs)
+{
+}
+
+/*
+ * GDB remote protocol parser:
+ */
+
+static const char	hexchars[] = "0123456789abcdef";
+
+static int hex(char ch)
+{
+	if ((ch >= 'a') && (ch <= 'f'))
+		return ch - 'a' + 10;
+	if ((ch >= '0') && (ch <= '9'))
+		return ch - '0';
+	if ((ch >= 'A') && (ch <= 'F'))
+		return ch - 'A' + 10;
+	return -1;
+}
+
+/* scan for the sequence $<data>#<checksum> */
+static void get_packet(char *buffer)
+{
+	unsigned char checksum;
+	unsigned char xmitcsum;
+	int count;
+	char ch;
+
+	do {
+		/*
+		 * Spin and wait around for the start character, ignore all
+		 * other characters:
+		 */
+		while ((ch = (kgdb_io_ops->read_char())) != '$')
+			/* nothing */;
+
+		kgdb_connected = 1;
+		checksum = 0;
+		xmitcsum = -1;
+
+		count = 0;
+
+		/*
+		 * now, read until a # or end of buffer is found:
+		 */
+		while (count < (BUFMAX - 1)) {
+			ch = kgdb_io_ops->read_char();
+			if (ch == '#')
+				break;
+			checksum = checksum + ch;
+			buffer[count] = ch;
+			count = count + 1;
+		}
+		buffer[count] = 0;
+
+		if (ch == '#') {
+			xmitcsum = hex(kgdb_io_ops->read_char()) << 4;
+			xmitcsum += hex(kgdb_io_ops->read_char());
+
+			if (checksum != xmitcsum)
+				/* failed checksum */
+				kgdb_io_ops->write_char('-');
+			else
+				/* successful transfer */
+				kgdb_io_ops->write_char('+');
+			if (kgdb_io_ops->flush)
+				kgdb_io_ops->flush();
+		}
+	} while (checksum != xmitcsum);
+}
+
+/*
+ * Send the packet in buffer.
+ * Check for gdb connection if asked for.
+ */
+static void put_packet(char *buffer)
+{
+	unsigned char checksum;
+	int count;
+	char ch;
+
+	/*
+	 * $<packet info>#<checksum>.
+	 */
+	while (1) {
+		kgdb_io_ops->write_char('$');
+		checksum = 0;
+		count = 0;
+
+		while ((ch = buffer[count])) {
+			kgdb_io_ops->write_char(ch);
+			checksum += ch;
+			count++;
+		}
+
+		kgdb_io_ops->write_char('#');
+		kgdb_io_ops->write_char(hexchars[checksum >> 4]);
+		kgdb_io_ops->write_char(hexchars[checksum & 0xf]);
+		if (kgdb_io_ops->flush)
+			kgdb_io_ops->flush();
+
+		/* Now see what we get in reply. */
+		ch = kgdb_io_ops->read_char();
+
+		if (ch == 3)
+			ch = kgdb_io_ops->read_char();
+
+		/* If we get an ACK, we are done. */
+		if (ch == '+')
+			return;
+
+		/*
+		 * If we get the start of another packet, this means
+		 * that GDB is attempting to reconnect.  We will NAK
+		 * the packet being sent, and stop trying to send this
+		 * packet.
+		 */
+		if (ch == '$') {
+			kgdb_io_ops->write_char('-');
+			if (kgdb_io_ops->flush)
+				kgdb_io_ops->flush();
+			return;
+		}
+	}
+}
+
+static char *pack_hex_byte(char *pkt, u8 byte)
+{
+	*pkt++ = hexchars[byte >> 4];
+	*pkt++ = hexchars[byte & 0xf];
+
+	return pkt;
+}
+
+/*
+ * Convert the memory pointed to by mem into hex, placing result in buf.
+ * Return a pointer to the last char put in buf (null). May return an error.
+ */
+int kgdb_mem2hex(char *mem, char *buf, int count)
+{
+	char *tmp;
+	int err;
+
+	/*
+	 * We use the upper half of buf as an intermediate buffer for the
+	 * raw memory copy.  Hex conversion will work against this one.
+	 */
+	tmp = buf + count;
+
+	err = probe_kernel_read(tmp, mem, count);
+	if (!err) {
+		while (count > 0) {
+			buf = pack_hex_byte(buf, *tmp);
+			tmp++;
+			count--;
+		}
+
+		*buf = 0;
+	}
+
+	return err;
+}
+
+/*
+ * Copy the binary array pointed to by buf into mem.  Fix $, #, and
+ * 0x7d escaped with 0x7d.  Return a pointer to the character after
+ * the last byte written.
+ */
+static int kgdb_ebin2mem(char *buf, char *mem, int count)
+{
+	int err = 0;
+	char c;
+
+	while (count-- > 0) {
+		c = *buf++;
+		if (c == 0x7d)
+			c = *buf++ ^ 0x20;
+
+		err = probe_kernel_write(mem, &c, 1);
+		if (err)
+			break;
+
+		mem++;
+	}
+
+	return err;
+}
+
+/*
+ * Convert the hex array pointed to by buf into binary to be placed in mem.
+ * Return a pointer to the character AFTER the last byte written.
+ * May return an error.
+ */
+int kgdb_hex2mem(char *buf, char *mem, int count)
+{
+	char *tmp_raw;
+	char *tmp_hex;
+
+	/*
+	 * We use the upper half of buf as an intermediate buffer for the
+	 * raw memory that is converted from hex.
+	 */
+	tmp_raw = buf + count * 2;
+
+	tmp_hex = tmp_raw - 1;
+	while (tmp_hex >= buf) {
+		tmp_raw--;
+		*tmp_raw = hex(*tmp_hex--);
+		*tmp_raw |= hex(*tmp_hex--) << 4;
+	}
+
+	return probe_kernel_write(mem, tmp_raw, count);
+}
+
+/*
+ * While we find nice hex chars, build a long_val.
+ * Return number of chars processed.
+ */
+int kgdb_hex2long(char **ptr, long *long_val)
+{
+	int hex_val;
+	int num = 0;
+
+	*long_val = 0;
+
+	while (**ptr) {
+		hex_val = hex(**ptr);
+		if (hex_val < 0)
+			break;
+
+		*long_val = (*long_val << 4) | hex_val;
+		num++;
+		(*ptr)++;
+	}
+
+	return num;
+}
+
+/* Write memory due to an 'M' or 'X' packet. */
+static int write_mem_msg(int binary)
+{
+	char *ptr = &remcom_in_buffer[1];
+	unsigned long addr;
+	unsigned long length;
+	int err;
+
+	if (kgdb_hex2long(&ptr, &addr) > 0 && *(ptr++) == ',' &&
+	    kgdb_hex2long(&ptr, &length) > 0 && *(ptr++) == ':') {
+		if (binary)
+			err = kgdb_ebin2mem(ptr, (char *)addr, length);
+		else
+			err = kgdb_hex2mem(ptr, (char *)addr, length);
+		if (err)
+			return err;
+		if (CACHE_FLUSH_IS_SAFE)
+			flush_icache_range(addr, addr + length + 1);
+		return 0;
+	}
+
+	return -EINVAL;
+}
+
+static void error_packet(char *pkt, int error)
+{
+	error = -error;
+	pkt[0] = 'E';
+	pkt[1] = hexchars[(error / 10)];
+	pkt[2] = hexchars[(error % 10)];
+	pkt[3] = '\0';
+}
+
+/*
+ * Thread ID accessors. We represent a flat TID space to GDB, where
+ * the per CPU idle threads (which under Linux all have PID 0) are
+ * remapped to negative TIDs.
+ */
+
+#define BUF_THREAD_ID_SIZE	16
+
+static char *pack_threadid(char *pkt, unsigned char *id)
+{
+	char *limit;
+
+	limit = pkt + BUF_THREAD_ID_SIZE;
+	while (pkt < limit)
+		pkt = pack_hex_byte(pkt, *id++);
+
+	return pkt;
+}
+
+static void int_to_threadref(unsigned char *id, int value)
+{
+	unsigned char *scan;
+	int i = 4;
+
+	scan = (unsigned char *)id;
+	while (i--)
+		*scan++ = 0;
+	*scan++ = (value >> 24) & 0xff;
+	*scan++ = (value >> 16) & 0xff;
+	*scan++ = (value >> 8) & 0xff;
+	*scan++ = (value & 0xff);
+}
+
+static struct task_struct *getthread(struct pt_regs *regs, int tid)
+{
+	/*
+	 * Non-positive TIDs are remapped idle tasks:
+	 */
+	if (tid <= 0)
+		return idle_task(-tid);
+
+	/*
+	 * find_task_by_pid_ns() does not take the tasklist lock anymore
+	 * but is nicely RCU locked - hence is a pretty resilient
+	 * thing to use:
+	 */
+	return find_task_by_pid_ns(tid, &init_pid_ns);
+}
+
+/*
+ * CPU debug state control:
+ */
+
+#ifdef CONFIG_SMP
+static void kgdb_wait(struct pt_regs *regs)
+{
+	unsigned long flags;
+	int cpu;
+
+	local_irq_save(flags);
+	cpu = raw_smp_processor_id();
+	kgdb_info[cpu].debuggerinfo = regs;
+	kgdb_info[cpu].task = current;
+	/*
+	 * Make sure the above info reaches the primary CPU before
+	 * our cpu_in_kgdb[] flag setting does:
+	 */
+	smp_wmb();
+	atomic_set(&cpu_in_kgdb[cpu], 1);
+
+	/*
+	 * The primary CPU must be active to enter here, but this is
+	 * guard in case the primary CPU had not been selected if
+	 * this was an entry via nmi.
+	 */
+	while (atomic_read(&kgdb_active) == -1)
+		cpu_relax();
+
+	/* Wait till primary CPU goes completely into the debugger. */
+	while (!atomic_read(&cpu_in_kgdb[atomic_read(&kgdb_active)]))
+		cpu_relax();
+
+	/* Wait till primary CPU is done with debugging */
+	while (atomic_read(&passive_cpu_wait[cpu]))
+		cpu_relax();
+
+	kgdb_info[cpu].debuggerinfo = NULL;
+	kgdb_info[cpu].task = NULL;
+
+	/* fix up hardware debug registers on local cpu */
+	if (arch_kgdb_ops.correct_hw_break)
+		arch_kgdb_ops.correct_hw_break();
+
+	/* Signal the primary CPU that we are done: */
+	atomic_set(&cpu_in_kgdb[cpu], 0);
+	local_irq_restore(flags);
+}
+#endif
+
+/*
+ * Some architectures need cache flushes when we set/clear a
+ * breakpoint:
+ */
+static void kgdb_flush_swbreak_addr(unsigned long addr)
+{
+	if (!CACHE_FLUSH_IS_SAFE)
+		return;
+
+	if (current->mm) {
+		flush_cache_range(current->mm->mmap_cache,
+				  addr, addr + BREAK_INSTR_SIZE);
+	} else {
+		flush_icache_range(addr, addr + BREAK_INSTR_SIZE);
+	}
+}
+
+/*
+ * SW breakpoint management:
+ */
+static int kgdb_activate_sw_breakpoints(void)
+{
+	unsigned long addr;
+	int error = 0;
+	int i;
+
+	for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+		if (kgdb_break[i].state != BP_SET)
+			continue;
+
+		addr = kgdb_break[i].bpt_addr;
+		error = kgdb_arch_set_breakpoint(addr,
+				kgdb_break[i].saved_instr);
+		if (error)
+			return error;
+
+		kgdb_flush_swbreak_addr(addr);
+		kgdb_break[i].state = BP_ACTIVE;
+	}
+	return 0;
+}
+
+static int kgdb_set_sw_break(unsigned long addr)
+{
+	int err = kgdb_validate_break_address(addr);
+	int breakno = -1;
+	int i;
+
+	if (err)
+		return err;
+
+	for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+		if ((kgdb_break[i].state == BP_SET) &&
+					(kgdb_break[i].bpt_addr == addr))
+			return -EEXIST;
+	}
+	for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+		if (kgdb_break[i].state == BP_REMOVED &&
+					kgdb_break[i].bpt_addr == addr) {
+			breakno = i;
+			break;
+		}
+	}
+
+	if (breakno == -1) {
+		for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+			if (kgdb_break[i].state == BP_UNDEFINED) {
+				breakno = i;
+				break;
+			}
+		}
+	}
+
+	if (breakno == -1)
+		return -E2BIG;
+
+	kgdb_break[breakno].state = BP_SET;
+	kgdb_break[breakno].type = BP_BREAKPOINT;
+	kgdb_break[breakno].bpt_addr = addr;
+
+	return 0;
+}
+
+static int kgdb_deactivate_sw_breakpoints(void)
+{
+	unsigned long addr;
+	int error = 0;
+	int i;
+
+	for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+		if (kgdb_break[i].state != BP_ACTIVE)
+			continue;
+		addr = kgdb_break[i].bpt_addr;
+		error = kgdb_arch_remove_breakpoint(addr,
+					kgdb_break[i].saved_instr);
+		if (error)
+			return error;
+
+		kgdb_flush_swbreak_addr(addr);
+		kgdb_break[i].state = BP_SET;
+	}
+	return 0;
+}
+
+static int kgdb_remove_sw_break(unsigned long addr)
+{
+	int i;
+
+	for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+		if ((kgdb_break[i].state == BP_SET) &&
+				(kgdb_break[i].bpt_addr == addr)) {
+			kgdb_break[i].state = BP_REMOVED;
+			return 0;
+		}
+	}
+	return -ENOENT;
+}
+
+int kgdb_isremovedbreak(unsigned long addr)
+{
+	int i;
+
+	for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+		if ((kgdb_break[i].state == BP_REMOVED) &&
+					(kgdb_break[i].bpt_addr == addr))
+			return 1;
+	}
+	return 0;
+}
+
+int remove_all_break(void)
+{
+	unsigned long addr;
+	int error;
+	int i;
+
+	/* Clear memory breakpoints. */
+	for (i = 0; i < KGDB_MAX_BREAKPOINTS; i++) {
+		if (kgdb_break[i].state != BP_SET)
+			continue;
+		addr = kgdb_break[i].bpt_addr;
+		error = kgdb_arch_remove_breakpoint(addr,
+				kgdb_break[i].saved_instr);
+		if (error)
+			return error;
+		kgdb_break[i].state = BP_REMOVED;
+	}
+
+	/* Clear hardware breakpoints. */
+	if (arch_kgdb_ops.remove_all_hw_break)
+		arch_kgdb_ops.remove_all_hw_break();
+
+	return 0;
+}
+
+/*
+ * Remap normal tasks to their real PID, idle tasks to -1 ... -NR_CPUs:
+ */
+static inline int shadow_pid(int realpid)
+{
+	if (realpid)
+		return realpid;
+
+	return -1-raw_smp_processor_id();
+}
+
+static char gdbmsgbuf[BUFMAX + 1];
+
+static void kgdb_msg_write(const char *s, int len)
+{
+	char *bufptr;
+	int wcount;
+	int i;
+
+	/* 'O'utput */
+	gdbmsgbuf[0] = 'O';
+
+	/* Fill and send buffers... */
+	while (len > 0) {
+		bufptr = gdbmsgbuf + 1;
+
+		/* Calculate how many this time */
+		if ((len << 1) > (BUFMAX - 2))
+			wcount = (BUFMAX - 2) >> 1;
+		else
+			wcount = len;
+
+		/* Pack in hex chars */
+		for (i = 0; i < wcount; i++)
+			bufptr = pack_hex_byte(bufptr, s[i]);
+		*bufptr = '\0';
+
+		/* Move up */
+		s += wcount;
+		len -= wcount;
+
+		/* Write packet */
+		put_packet(gdbmsgbuf);
+	}
+}
+
+/*
+ * Return true if there is a valid kgdb I/O module.  Also if no
+ * debugger is attached a message can be printed to the console about
+ * waiting for the debugger to attach.
+ *
+ * The print_wait argument is only to be true when called from inside
+ * the core kgdb_handle_exception, because it will wait for the
+ * debugger to attach.
+ */
+static int kgdb_io_ready(int print_wait)
+{
+	if (!kgdb_io_ops)
+		return 0;
+	if (kgdb_connected)
+		return 1;
+	if (atomic_read(&kgdb_setting_breakpoint))
+		return 1;
+	if (print_wait)
+		printk(KERN_CRIT "KGDB: Waiting for remote debugger\n");
+	return 1;
+}
+
+/*
+ * All the functions that start with gdb_cmd are the various
+ * operations to implement the handlers for the gdbserial protocol
+ * where KGDB is communicating with an external debugger
+ */
+
+/* Handle the '?' status packets */
+static void gdb_cmd_status(struct kgdb_state *ks)
+{
+	/*
+	 * We know that this packet is only sent
+	 * during initial connect.  So to be safe,
+	 * we clear out our breakpoints now in case
+	 * GDB is reconnecting.
+	 */
+	remove_all_break();
+
+	remcom_out_buffer[0] = 'S';
+	pack_hex_byte(&remcom_out_buffer[1], ks->signo);
+}
+
+/* Handle the 'g' get registers request */
+static void gdb_cmd_getregs(struct kgdb_state *ks)
+{
+	struct task_struct *thread;
+	void *local_debuggerinfo;
+	int i;
+
+	thread = kgdb_usethread;
+	if (!thread) {
+		thread = kgdb_info[ks->cpu].task;
+		local_debuggerinfo = kgdb_info[ks->cpu].debuggerinfo;
+	} else {
+		local_debuggerinfo = NULL;
+		for (i = 0; i < NR_CPUS; i++) {
+			/*
+			 * Try to find the task on some other
+			 * or possibly this node if we do not
+			 * find the matching task then we try
+			 * to approximate the results.
+			 */
+			if (thread == kgdb_info[i].task)
+				local_debuggerinfo = kgdb_info[i].debuggerinfo;
+		}
+	}
+
+	/*
+	 * All threads that don't have debuggerinfo should be
+	 * in __schedule() sleeping, since all other CPUs
+	 * are in kgdb_wait, and thus have debuggerinfo.
+	 */
+	if (local_debuggerinfo) {
+		pt_regs_to_gdb_regs(gdb_regs, local_debuggerinfo);
+	} else {
+		/*
+		 * Pull stuff saved during switch_to; nothing
+		 * else is accessible (or even particularly
+		 * relevant).
+		 *
+		 * This should be enough for a stack trace.
+		 */
+		sleeping_thread_to_gdb_regs(gdb_regs, thread);
+	}
+	kgdb_mem2hex((char *)gdb_regs, remcom_out_buffer, NUMREGBYTES);
+}
+
+/* Handle the 'G' set registers request */
+static void gdb_cmd_setregs(struct kgdb_state *ks)
+{
+	kgdb_hex2mem(&remcom_in_buffer[1], (char *)gdb_regs, NUMREGBYTES);
+
+	if (kgdb_usethread && kgdb_usethread != current) {
+		error_packet(remcom_out_buffer, -EINVAL);
+	} else {
+		gdb_regs_to_pt_regs(gdb_regs, ks->linux_regs);
+		strcpy(remcom_out_buffer, "OK");
+	}
+}
+
+/* Handle the 'm' memory read bytes */
+static void gdb_cmd_memread(struct kgdb_state *ks)
+{
+	char *ptr = &remcom_in_buffer[1];
+	unsigned long length;
+	unsigned long addr;
+	int err;
+
+	if (kgdb_hex2long(&ptr, &addr) > 0 && *ptr++ == ',' &&
+					kgdb_hex2long(&ptr, &length) > 0) {
+		err = kgdb_mem2hex((char *)addr, remcom_out_buffer, length);
+		if (err)
+			error_packet(remcom_out_buffer, err);
+	} else {
+		error_packet(remcom_out_buffer, -EINVAL);
+	}
+}
+
+/* Handle the 'M' memory write bytes */
+static void gdb_cmd_memwrite(struct kgdb_state *ks)
+{
+	int err = write_mem_msg(0);
+
+	if (err)
+		error_packet(remcom_out_buffer, err);
+	else
+		strcpy(remcom_out_buffer, "OK");
+}
+
+/* Handle the 'X' memory binary write bytes */
+static void gdb_cmd_binwrite(struct kgdb_state *ks)
+{
+	int err = write_mem_msg(1);
+
+	if (err)
+		error_packet(remcom_out_buffer, err);
+	else
+		strcpy(remcom_out_buffer, "OK");
+}
+
+/* Handle the 'D' or 'k', detach or kill packets */
+static void gdb_cmd_detachkill(struct kgdb_state *ks)
+{
+	int error;
+
+	/* The detach case */
+	if (remcom_in_buffer[0] == 'D') {
+		error = remove_all_break();
+		if (error < 0) {
+			error_packet(remcom_out_buffer, error);
+		} else {
+			strcpy(remcom_out_buffer, "OK");
+			kgdb_connected = 0;
+		}
+		put_packet(remcom_out_buffer);
+	} else {
+		/*
+		 * Assume the kill case, with no exit code checking,
+		 * trying to force detach the debugger:
+		 */
+		remove_all_break();
+		kgdb_connected = 0;
+	}
+}
+
+/* Handle the 'R' reboot packets */
+static int gdb_cmd_reboot(struct kgdb_state *ks)
+{
+	/* For now, only honor R0 */
+	if (strcmp(remcom_in_buffer, "R0") == 0) {
+		printk(KERN_CRIT "Executing emergency reboot\n");
+		strcpy(remcom_out_buffer, "OK");
+		put_packet(remcom_out_buffer);
+
+		/*
+		 * Execution should not return from
+		 * machine_emergency_restart()
+		 */
+		machine_emergency_restart();
+		kgdb_connected = 0;
+
+		return 1;
+	}
+	return 0;
+}
+
+/* Handle the 'q' query packets */
+static void gdb_cmd_query(struct kgdb_state *ks)
+{
+	struct task_struct *thread;
+	unsigned char thref[8];
+	char *ptr;
+	int i;
+
+	switch (remcom_in_buffer[1]) {
+	case 's':
+	case 'f':
+		if (memcmp(remcom_in_buffer + 2, "ThreadInfo", 10)) {
+			error_packet(remcom_out_buffer, -EINVAL);
+			break;
+		}
+
+		if (remcom_in_buffer[1] == 'f')
+			ks->threadid = 1;
+
+		remcom_out_buffer[0] = 'm';
+		ptr = remcom_out_buffer + 1;
+
+		for (i = 0; i < 17; ks->threadid++) {
+			thread = getthread(ks->linux_regs, ks->threadid);
+			if (thread) {
+				int_to_threadref(thref, ks->threadid);
+				pack_threadid(ptr, thref);
+				ptr += BUF_THREAD_ID_SIZE;
+				*(ptr++) = ',';
+				i++;
+			}
+		}
+		*(--ptr) = '\0';
+		break;
+
+	case 'C':
+		/* Current thread id */
+		strcpy(remcom_out_buffer, "QC");
+		ks->threadid = shadow_pid(current->pid);
+		int_to_threadref(thref, ks->threadid);
+		pack_threadid(remcom_out_buffer + 2, thref);
+		break;
+	case 'T':
+		if (memcmp(remcom_in_buffer + 1, "ThreadExtraInfo,", 16)) {
+			error_packet(remcom_out_buffer, -EINVAL);
+			break;
+		}
+		ks->threadid = 0;
+		ptr = remcom_in_buffer + 17;
+		kgdb_hex2long(&ptr, &ks->threadid);
+		if (!getthread(ks->linux_regs, ks->threadid)) {
+			error_packet(remcom_out_buffer, -EINVAL);
+			break;
+		}
+		if (ks->threadid > 0) {
+			kgdb_mem2hex(getthread(ks->linux_regs,
+					ks->threadid)->comm,
+					remcom_out_buffer, 16);
+		} else {
+			static char tmpstr[23 + BUF_THREAD_ID_SIZE];
+
+			sprintf(tmpstr, "Shadow task %d for pid 0",
+					(int)(-ks->threadid-1));
+			kgdb_mem2hex(tmpstr, remcom_out_buffer, strlen(tmpstr));
+		}
+		break;
+	}
+}
+
+/* Handle the 'H' task query packets */
+static void gdb_cmd_task(struct kgdb_state *ks)
+{
+	struct task_struct *thread;
+	char *ptr;
+
+	switch (remcom_in_buffer[1]) {
+	case 'g':
+		ptr = &remcom_in_buffer[2];
+		kgdb_hex2long(&ptr, &ks->threadid);
+		thread = getthread(ks->linux_regs, ks->threadid);
+		if (!thread && ks->threadid > 0) {
+			error_packet(remcom_out_buffer, -EINVAL);
+			break;
+		}
+		kgdb_usethread = thread;
+		ks->kgdb_usethreadid = ks->threadid;
+		strcpy(remcom_out_buffer, "OK");
+		break;
+	case 'c':
+		ptr = &remcom_in_buffer[2];
+		kgdb_hex2long(&ptr, &ks->threadid);
+		if (!ks->threadid) {
+			kgdb_contthread = NULL;
+		} else {
+			thread = getthread(ks->linux_regs, ks->threadid);
+			if (!thread && ks->threadid > 0) {
+				error_packet(remcom_out_buffer, -EINVAL);
+				break;
+			}
+			kgdb_contthread = thread;
+		}
+		strcpy(remcom_out_buffer, "OK");
+		break;
+	}
+}
+
+/* Handle the 'T' thread query packets */
+static void gdb_cmd_thread(struct kgdb_state *ks)
+{
+	char *ptr = &remcom_in_buffer[1];
+	struct task_struct *thread;
+
+	kgdb_hex2long(&ptr, &ks->threadid);
+	thread = getthread(ks->linux_regs, ks->threadid);
+	if (thread)
+		strcpy(remcom_out_buffer, "OK");
+	else
+		error_packet(remcom_out_buffer, -EINVAL);
+}
+
+/* Handle the 'z' or 'Z' breakpoint remove or set packets */
+static void gdb_cmd_break(struct kgdb_state *ks)
+{
+	/*
+	 * Since GDB-5.3, it's been drafted that '0' is a software
+	 * breakpoint, '1' is a hardware breakpoint, so let's do that.
+	 */
+	char *bpt_type = &remcom_in_buffer[1];
+	char *ptr = &remcom_in_buffer[2];
+	unsigned long addr;
+	unsigned long length;
+	int error = 0;
+
+	if (arch_kgdb_ops.set_hw_breakpoint && *bpt_type >= '1') {
+		/* Unsupported */
+		if (*bpt_type > '4')
+			return;
+	} else {
+		if (*bpt_type != '0' && *bpt_type != '1')
+			/* Unsupported. */
+			return;
+	}
+
+	/*
+	 * Test if this is a hardware breakpoint, and
+	 * if we support it:
+	 */
+	if (*bpt_type == '1' && !(arch_kgdb_ops.flags & KGDB_HW_BREAKPOINT))
+		/* Unsupported. */
+		return;
+
+	if (*(ptr++) != ',') {
+		error_packet(remcom_out_buffer, -EINVAL);
+		return;
+	}
+	if (!kgdb_hex2long(&ptr, &addr)) {
+		error_packet(remcom_out_buffer, -EINVAL);
+		return;
+	}
+	if (*(ptr++) != ',' ||
+		!kgdb_hex2long(&ptr, &length)) {
+		error_packet(remcom_out_buffer, -EINVAL);
+		return;
+	}
+
+	if (remcom_in_buffer[0] == 'Z' && *bpt_type == '0')
+		error = kgdb_set_sw_break(addr);
+	else if (remcom_in_buffer[0] == 'z' && *bpt_type == '0')
+		error = kgdb_remove_sw_break(addr);
+	else if (remcom_in_buffer[0] == 'Z')
+		error = arch_kgdb_ops.set_hw_breakpoint(addr,
+			(int)length, *bpt_type);
+	else if (remcom_in_buffer[0] == 'z')
+		error = arch_kgdb_ops.remove_hw_breakpoint(addr,
+			(int) length, *bpt_type);
+
+	if (error == 0)
+		strcpy(remcom_out_buffer, "OK");
+	else
+		error_packet(remcom_out_buffer, error);
+}
+
+/* Handle the 'C' signal / exception passing packets */
+static int gdb_cmd_exception_pass(struct kgdb_state *ks)
+{
+	/* C09 == pass exception
+	 * C15 == detach kgdb, pass exception
+	 */
+	if (remcom_in_buffer[1] == '0' && remcom_in_buffer[2] == '9') {
+
+		ks->pass_exception = 1;
+		remcom_in_buffer[0] = 'c';
+
+	} else if (remcom_in_buffer[1] == '1' && remcom_in_buffer[2] == '5') {
+
+		ks->pass_exception = 1;
+		remcom_in_buffer[0] = 'D';
+		remove_all_break();
+		kgdb_connected = 0;
+		return 1;
+
+	} else {
+		error_packet(remcom_out_buffer, -EINVAL);
+		return 0;
+	}
+
+	/* Indicate fall through */
+	return -1;
+}
+
+/*
+ * This function performs all gdbserial command procesing
+ */
+static int gdb_serial_stub(struct kgdb_state *ks)
+{
+	int error = 0;
+	int tmp;
+
+	/* Clear the out buffer. */
+	memset(remcom_out_buffer, 0, sizeof(remcom_out_buffer));
+
+	if (kgdb_connected) {
+		unsigned char thref[8];
+		char *ptr;
+
+		/* Reply to host that an exception has occurred */
+		ptr = remcom_out_buffer;
+		*ptr++ = 'T';
+		ptr = pack_hex_byte(ptr, ks->signo);
+		ptr += strlen(strcpy(ptr, "thread:"));
+		int_to_threadref(thref, shadow_pid(current->pid));
+		ptr = pack_threadid(ptr, thref);
+		*ptr++ = ';';
+		put_packet(remcom_out_buffer);
+	}
+
+	kgdb_usethread = kgdb_info[ks->cpu].task;
+	ks->kgdb_usethreadid = shadow_pid(kgdb_info[ks->cpu].task->pid);
+	ks->pass_exception = 0;
+
+	while (1) {
+		error = 0;
+
+		/* Clear the out buffer. */
+		memset(remcom_out_buffer, 0, sizeof(remcom_out_buffer));
+
+		get_packet(remcom_in_buffer);
+
+		switch (remcom_in_buffer[0]) {
+		case '?': /* gdbserial status */
+			gdb_cmd_status(ks);
+			break;
+		case 'g': /* return the value of the CPU registers */
+			gdb_cmd_getregs(ks);
+			break;
+		case 'G': /* set the value of the CPU registers - return OK */
+			gdb_cmd_setregs(ks);
+			break;
+		case 'm': /* mAA..AA,LLLL  Read LLLL bytes at address AA..AA */
+			gdb_cmd_memread(ks);
+			break;
+		case 'M': /* MAA..AA,LLLL: Write LLLL bytes at address AA..AA */
+			gdb_cmd_memwrite(ks);
+			break;
+		case 'X': /* XAA..AA,LLLL: Write LLLL bytes at address AA..AA */
+			gdb_cmd_binwrite(ks);
+			break;
+			/* kill or detach. KGDB should treat this like a
+			 * continue.
+			 */
+		case 'D': /* Debugger detach */
+		case 'k': /* Debugger detach via kill */
+			gdb_cmd_detachkill(ks);
+			goto default_handle;
+		case 'R': /* Reboot */
+			if (gdb_cmd_reboot(ks))
+				goto default_handle;
+			break;
+		case 'q': /* query command */
+			gdb_cmd_query(ks);
+			break;
+		case 'H': /* task related */
+			gdb_cmd_task(ks);
+			break;
+		case 'T': /* Query thread status */
+			gdb_cmd_thread(ks);
+			break;
+		case 'z': /* Break point remove */
+		case 'Z': /* Break point set */
+			gdb_cmd_break(ks);
+			break;
+		case 'C': /* Exception passing */
+			tmp = gdb_cmd_exception_pass(ks);
+			if (tmp > 0)
+				goto default_handle;
+			if (tmp == 0)
+				break;
+			/* Fall through on tmp < 0 */
+		case 'c': /* Continue packet */
+		case 's': /* Single step packet */
+			if (kgdb_contthread && kgdb_contthread != current) {
+				/* Can't switch threads in kgdb */
+				error_packet(remcom_out_buffer, -EINVAL);
+				break;
+			}
+			kgdb_activate_sw_breakpoints();
+			/* Fall through to default processing */
+		default:
+default_handle:
+			error = kgdb_arch_handle_exception(ks->ex_vector,
+						ks->signo,
+						ks->err_code,
+						remcom_in_buffer,
+						remcom_out_buffer,
+						ks->linux_regs);
+			/*
+			 * Leave cmd processing on error, detach,
+			 * kill, continue, or single step.
+			 */
+			if (error >= 0 || remcom_in_buffer[0] == 'D' ||
+			    remcom_in_buffer[0] == 'k') {
+				error = 0;
+				goto kgdb_exit;
+			}
+
+		}
+
+		/* reply to the request */
+		put_packet(remcom_out_buffer);
+	}
+
+kgdb_exit:
+	if (ks->pass_exception)
+		error = 1;
+	return error;
+}
+
+static int kgdb_reenter_check(struct kgdb_state *ks)
+{
+	unsigned long addr;
+
+	if (atomic_read(&kgdb_active) != raw_smp_processor_id())
+		return 0;
+
+	/* Panic on recursive debugger calls: */
+	exception_level++;
+	addr = kgdb_arch_pc(ks->ex_vector, ks->linux_regs);
+	kgdb_deactivate_sw_breakpoints();
+
+	/*
+	 * If the break point removed ok at the place exception
+	 * occurred, try to recover and print a warning to the end
+	 * user because the user planted a breakpoint in a place that
+	 * KGDB needs in order to function.
+	 */
+	if (kgdb_remove_sw_break(addr) == 0) {
+		exception_level = 0;
+		kgdb_skipexception(ks->ex_vector, ks->linux_regs);
+		kgdb_activate_sw_breakpoints();
+		printk(KERN_CRIT "KGDB: re-enter error: breakpoint removed\n");
+		WARN_ON_ONCE(1);
+
+		return 1;
+	}
+	remove_all_break();
+	kgdb_skipexception(ks->ex_vector, ks->linux_regs);
+
+	if (exception_level > 1) {
+		dump_stack();
+		panic("Recursive entry to debugger");
+	}
+
+	printk(KERN_CRIT "KGDB: re-enter exception: ALL breakpoints killed\n");
+	dump_stack();
+	panic("Recursive entry to debugger");
+
+	return 1;
+}
+
+/*
+ * kgdb_handle_exception() - main entry point from a kernel exception
+ *
+ * Locking hierarchy:
+ *	interface locks, if any (begin_session)
+ *	kgdb lock (kgdb_active)
+ */
+int
+kgdb_handle_exception(int evector, int signo, int ecode, struct pt_regs *regs)
+{
+	struct kgdb_state kgdb_var;
+	struct kgdb_state *ks = &kgdb_var;
+	unsigned long flags;
+	int error = 0;
+	int i, cpu;
+
+	ks->cpu			= raw_smp_processor_id();
+	ks->ex_vector		= evector;
+	ks->signo		= signo;
+	ks->ex_vector		= evector;
+	ks->err_code		= ecode;
+	ks->kgdb_usethreadid	= 0;
+	ks->linux_regs		= regs;
+
+	if (kgdb_reenter_check(ks))
+		return 0; /* Ouch, double exception ! */
+
+acquirelock:
+	/*
+	 * Interrupts will be restored by the 'trap return' code, except when
+	 * single stepping.
+	 */
+	local_irq_save(flags);
+
+	cpu = raw_smp_processor_id();
+
+	/*
+	 * Acquire the kgdb_active lock:
+	 */
+	while (atomic_cmpxchg(&kgdb_active, -1, cpu) != -1)
+		cpu_relax();
+
+	/*
+	 * Do not start the debugger connection on this CPU if the last
+	 * instance of the exception handler wanted to come into the
+	 * debugger on a different CPU via a single step
+	 */
+	if (atomic_read(&kgdb_cpu_doing_single_step) != -1 &&
+	    atomic_read(&kgdb_cpu_doing_single_step) != cpu) {
+
+		atomic_set(&kgdb_active, -1);
+		local_irq_restore(flags);
+
+		goto acquirelock;
+	}
+
+	if (!kgdb_io_ready(1)) {
+		error = 1;
+		goto kgdb_restore; /* No I/O connection, so resume the system */
+	}
+
+	/*
+	 * Don't enter if we have hit a removed breakpoint.
+	 */
+	if (kgdb_skipexception(ks->ex_vector, ks->linux_regs))
+		goto kgdb_restore;
+
+	/* Call the I/O driver's pre_exception routine */
+	if (kgdb_io_ops->pre_exception)
+		kgdb_io_ops->pre_exception();
+
+	kgdb_info[ks->cpu].debuggerinfo = ks->linux_regs;
+	kgdb_info[ks->cpu].task = current;
+
+	kgdb_disable_hw_debug(ks->linux_regs);
+
+	/*
+	 * Get the passive CPU lock which will hold all the non-primary
+	 * CPU in a spin state while the debugger is active
+	 */
+	if (!kgdb_single_step || !kgdb_contthread) {
+		for (i = 0; i < NR_CPUS; i++)
+			atomic_set(&passive_cpu_wait[i], 1);
+	}
+
+#ifdef CONFIG_SMP
+	/* Signal the other CPUs to enter kgdb_wait() */
+	if ((!kgdb_single_step || !kgdb_contthread) && kgdb_do_roundup)
+		kgdb_roundup_cpus(flags);
+#endif
+
+	/*
+	 * spin_lock code is good enough as a barrier so we don't
+	 * need one here:
+	 */
+	atomic_set(&cpu_in_kgdb[ks->cpu], 1);
+
+	/*
+	 * Wait for the other CPUs to be notified and be waiting for us:
+	 */
+	for_each_online_cpu(i) {
+		while (!atomic_read(&cpu_in_kgdb[i]))
+			cpu_relax();
+	}
+
+	/*
+	 * At this point the primary processor is completely
+	 * in the debugger and all secondary CPUs are quiescent
+	 */
+	kgdb_post_primary_code(ks->linux_regs, ks->ex_vector, ks->err_code);
+	kgdb_deactivate_sw_breakpoints();
+	kgdb_single_step = 0;
+	kgdb_contthread = NULL;
+	exception_level = 0;
+
+	/* Talk to debugger with gdbserial protocol */
+	error = gdb_serial_stub(ks);
+
+	/* Call the I/O driver's post_exception routine */
+	if (kgdb_io_ops->post_exception)
+		kgdb_io_ops->post_exception();
+
+	kgdb_info[ks->cpu].debuggerinfo = NULL;
+	kgdb_info[ks->cpu].task = NULL;
+	atomic_set(&cpu_in_kgdb[ks->cpu], 0);
+
+	if (!kgdb_single_step || !kgdb_contthread) {
+		for (i = NR_CPUS-1; i >= 0; i--)
+			atomic_set(&passive_cpu_wait[i], 0);
+		/*
+		 * Wait till all the CPUs have quit
+		 * from the debugger.
+		 */
+		for_each_online_cpu(i) {
+			while (atomic_read(&cpu_in_kgdb[i]))
+				cpu_relax();
+		}
+	}
+
+kgdb_restore:
+	/* Free kgdb_active */
+	atomic_set(&kgdb_active, -1);
+	local_irq_restore(flags);
+
+	return error;
+}
+
+int kgdb_nmicallback(int cpu, void *regs)
+{
+#ifdef CONFIG_SMP
+	if (!atomic_read(&cpu_in_kgdb[cpu]) &&
+			atomic_read(&kgdb_active) != cpu) {
+		kgdb_wait((struct pt_regs *)regs);
+		return 0;
+	}
+#endif
+	return 1;
+}
+
+void kgdb_console_write(struct console *co, const char *s, unsigned count)
+{
+	unsigned long flags;
+
+	/* If we're debugging, or KGDB has not connected, don't try
+	 * and print. */
+	if (!kgdb_connected || atomic_read(&kgdb_active) != -1)
+		return;
+
+	local_irq_save(flags);
+	kgdb_msg_write(s, count);
+	local_irq_restore(flags);
+}
+
+static struct console kgdbcons = {
+	.name		= "kgdb",
+	.write		= kgdb_console_write,
+	.flags		= CON_PRINTBUFFER | CON_ENABLED,
+	.index		= -1,
+};
+
+#ifdef CONFIG_MAGIC_SYSRQ
+static void sysrq_handle_gdb(int key, struct tty_struct *tty)
+{
+	if (!kgdb_io_ops) {
+		printk(KERN_CRIT "ERROR: No KGDB I/O module available\n");
+		return;
+	}
+	if (!kgdb_connected)
+		printk(KERN_CRIT "Entering KGDB\n");
+
+	kgdb_breakpoint();
+}
+
+static struct sysrq_key_op sysrq_gdb_op = {
+	.handler	= sysrq_handle_gdb,
+	.help_msg	= "Gdb",
+	.action_msg	= "GDB",
+};
+#endif
+
+static void kgdb_register_callbacks(void)
+{
+	if (!kgdb_io_module_registered) {
+		kgdb_io_module_registered = 1;
+		kgdb_arch_init();
+#ifdef CONFIG_MAGIC_SYSRQ
+		register_sysrq_key('g', &sysrq_gdb_op);
+#endif
+		if (kgdb_use_con && !kgdb_con_registered) {
+			register_console(&kgdbcons);
+			kgdb_con_registered = 1;
+		}
+	}
+}
+
+static void kgdb_unregister_callbacks(void)
+{
+	/*
+	 * When this routine is called KGDB should unregister from the
+	 * panic handler and clean up, making sure it is not handling any
+	 * break exceptions at the time.
+	 */
+	if (kgdb_io_module_registered) {
+		kgdb_io_module_registered = 0;
+		kgdb_arch_exit();
+#ifdef CONFIG_MAGIC_SYSRQ
+		unregister_sysrq_key('g', &sysrq_gdb_op);
+#endif
+		if (kgdb_con_registered) {
+			unregister_console(&kgdbcons);
+			kgdb_con_registered = 0;
+		}
+	}
+}
+
+static void kgdb_initial_breakpoint(void)
+{
+	kgdb_break_asap = 0;
+
+	printk(KERN_CRIT "kgdb: Waiting for connection from remote gdb...\n");
+	kgdb_breakpoint();
+}
+
+/**
+ *	kkgdb_register_io_module - register KGDB IO module
+ *	@new_kgdb_io_ops: the io ops vector
+ *
+ *	Register it with the KGDB core.
+ */
+int kgdb_register_io_module(struct kgdb_io *new_kgdb_io_ops)
+{
+	int err;
+
+	spin_lock(&kgdb_registration_lock);
+
+	if (kgdb_io_ops) {
+		spin_unlock(&kgdb_registration_lock);
+
+		printk(KERN_ERR "kgdb: Another I/O driver is already "
+				"registered with KGDB.\n");
+		return -EBUSY;
+	}
+
+	if (new_kgdb_io_ops->init) {
+		err = new_kgdb_io_ops->init();
+		if (err) {
+			spin_unlock(&kgdb_registration_lock);
+			return err;
+		}
+	}
+
+	kgdb_io_ops = new_kgdb_io_ops;
+
+	spin_unlock(&kgdb_registration_lock);
+
+	printk(KERN_INFO "kgdb: Registered I/O driver %s.\n",
+	       new_kgdb_io_ops->name);
+
+	/* Arm KGDB now. */
+	kgdb_register_callbacks();
+
+	if (kgdb_break_asap)
+		kgdb_initial_breakpoint();
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(kgdb_register_io_module);
+
+/**
+ *	kkgdb_unregister_io_module - unregister KGDB IO module
+ *	@old_kgdb_io_ops: the io ops vector
+ *
+ *	Unregister it with the KGDB core.
+ */
+void kgdb_unregister_io_module(struct kgdb_io *old_kgdb_io_ops)
+{
+	BUG_ON(kgdb_connected);
+
+	/*
+	 * KGDB is no longer able to communicate out, so
+	 * unregister our callbacks and reset state.
+	 */
+	kgdb_unregister_callbacks();
+
+	spin_lock(&kgdb_registration_lock);
+
+	WARN_ON_ONCE(kgdb_io_ops != old_kgdb_io_ops);
+	kgdb_io_ops = NULL;
+
+	spin_unlock(&kgdb_registration_lock);
+
+	printk(KERN_INFO
+		"kgdb: Unregistered I/O driver %s, debugger disabled.\n",
+		old_kgdb_io_ops->name);
+}
+EXPORT_SYMBOL_GPL(kgdb_unregister_io_module);
+
+/**
+ * kgdb_breakpoint - generate breakpoint exception
+ *
+ * This function will generate a breakpoint exception.  It is used at the
+ * beginning of a program to sync up with a debugger and can be used
+ * otherwise as a quick means to stop program execution and "break" into
+ * the debugger.
+ */
+void kgdb_breakpoint(void)
+{
+	atomic_set(&kgdb_setting_breakpoint, 1);
+	wmb(); /* Sync point before breakpoint */
+	arch_kgdb_breakpoint();
+	wmb(); /* Sync point after breakpoint */
+	atomic_set(&kgdb_setting_breakpoint, 0);
+}
+EXPORT_SYMBOL_GPL(kgdb_breakpoint);
+
+static int __init opt_kgdb_wait(char *str)
+{
+	kgdb_break_asap = 1;
+
+	if (kgdb_io_module_registered)
+		kgdb_initial_breakpoint();
+
+	return 0;
+}
+
+early_param("kgdbwait", opt_kgdb_wait);
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index 0796c1a090c0..e601d0e7ac5d 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -622,3 +622,5 @@ config PROVIDE_OHCI1394_DMA_INIT
 	  See Documentation/debugging-via-ohci1394.txt for more information.
 
 source "samples/Kconfig"
+
+source "lib/Kconfig.kgdb"
diff --git a/lib/Kconfig.kgdb b/lib/Kconfig.kgdb
new file mode 100644
index 000000000000..9631ba3baaf3
--- /dev/null
+++ b/lib/Kconfig.kgdb
@@ -0,0 +1,27 @@
+
+menuconfig KGDB
+	bool "KGDB: kernel debugging with remote gdb"
+	select FRAME_POINTER
+	depends on HAVE_ARCH_KGDB
+	depends on DEBUG_KERNEL && EXPERIMENTAL
+	help
+	  If you say Y here, it will be possible to remotely debug the
+	  kernel using gdb.  Documentation of kernel debugger is available
+	  at http://kgdb.sourceforge.net as well as in DocBook form
+	  in Documentation/DocBook/.  If unsure, say N.
+
+config HAVE_ARCH_KGDB_SHADOW_INFO
+	bool
+
+config HAVE_ARCH_KGDB
+	bool
+
+config KGDB_SERIAL_CONSOLE
+	tristate "KGDB: use kgdb over the serial console"
+	depends on KGDB
+	select CONSOLE_POLL
+	select MAGIC_SYSRQ
+	default y
+	help
+	  Share a serial console with kgdb. Sysrq-g must be used
+	  to break in initially.
-- 
cgit v1.2.3


From f2d937f3bf00665ccf048b3b6616ef95859b0945 Mon Sep 17 00:00:00 2001
From: Jason Wessel <jason.wessel@windriver.com>
Date: Thu, 17 Apr 2008 20:05:37 +0200
Subject: consoles: polling support, kgdboc

polled console handling support, to access a console in an irq-less
way while in debug or irq context.

absolutely zero impact as long as CONFIG_CONSOLE_POLL is disabled.
(which is the default)

[ jan.kiszka@siemens.com: lots of cleanups ]
[ mingo@elte.hu: redesign, splitups, cleanups. ]

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
---
 drivers/char/tty_io.c        |  47 +++++++++++++
 drivers/serial/8250.c        |  58 +++++++++++++++
 drivers/serial/Kconfig       |   3 +
 drivers/serial/Makefile      |   1 +
 drivers/serial/kgdboc.c      | 163 +++++++++++++++++++++++++++++++++++++++++++
 drivers/serial/serial_core.c |  72 ++++++++++++++++++-
 include/linux/serial_core.h  |   4 ++
 include/linux/tty_driver.h   |  12 ++++
 8 files changed, 357 insertions(+), 3 deletions(-)
 create mode 100644 drivers/serial/kgdboc.c

(limited to 'include/linux')

diff --git a/drivers/char/tty_io.c b/drivers/char/tty_io.c
index 613ec816ce60..4d3c7018f0c3 100644
--- a/drivers/char/tty_io.c
+++ b/drivers/char/tty_io.c
@@ -1155,6 +1155,48 @@ static struct tty_driver *get_tty_driver(dev_t device, int *index)
 	return NULL;
 }
 
+#ifdef CONFIG_CONSOLE_POLL
+
+/**
+ *	tty_find_polling_driver	-	find device of a polled tty
+ *	@name: name string to match
+ *	@line: pointer to resulting tty line nr
+ *
+ *	This routine returns a tty driver structure, given a name
+ *	and the condition that the tty driver is capable of polled
+ *	operation.
+ */
+struct tty_driver *tty_find_polling_driver(char *name, int *line)
+{
+	struct tty_driver *p, *res = NULL;
+	int tty_line = 0;
+	char *str;
+
+	mutex_lock(&tty_mutex);
+	/* Search through the tty devices to look for a match */
+	list_for_each_entry(p, &tty_drivers, tty_drivers) {
+		str = name + strlen(p->name);
+		tty_line = simple_strtoul(str, &str, 10);
+		if (*str == ',')
+			str++;
+		if (*str == '\0')
+			str = 0;
+
+		if (tty_line >= 0 && tty_line <= p->num && p->poll_init &&
+				!p->poll_init(p, tty_line, str)) {
+
+			res = p;
+			*line = tty_line;
+			break;
+		}
+	}
+	mutex_unlock(&tty_mutex);
+
+	return res;
+}
+EXPORT_SYMBOL_GPL(tty_find_polling_driver);
+#endif
+
 /**
  *	tty_check_change	-	check for POSIX terminal changes
  *	@tty: tty to check
@@ -3850,6 +3892,11 @@ void tty_set_operations(struct tty_driver *driver,
 	driver->write_proc = op->write_proc;
 	driver->tiocmget = op->tiocmget;
 	driver->tiocmset = op->tiocmset;
+#ifdef CONFIG_CONSOLE_POLL
+	driver->poll_init = op->poll_init;
+	driver->poll_get_char = op->poll_get_char;
+	driver->poll_put_char = op->poll_put_char;
+#endif
 }
 
 
diff --git a/drivers/serial/8250.c b/drivers/serial/8250.c
index 77f7a7f0646e..96a585e1cee8 100644
--- a/drivers/serial/8250.c
+++ b/drivers/serial/8250.c
@@ -1740,6 +1740,60 @@ static inline void wait_for_xmitr(struct uart_8250_port *up, int bits)
 	}
 }
 
+#ifdef CONFIG_CONSOLE_POLL
+/*
+ * Console polling routines for writing and reading from the uart while
+ * in an interrupt or debug context.
+ */
+
+static int serial8250_get_poll_char(struct uart_port *port)
+{
+	struct uart_8250_port *up = (struct uart_8250_port *)port;
+	unsigned char lsr = serial_inp(up, UART_LSR);
+
+	while (!(lsr & UART_LSR_DR))
+		lsr = serial_inp(up, UART_LSR);
+
+	return serial_inp(up, UART_RX);
+}
+
+
+static void serial8250_put_poll_char(struct uart_port *port,
+			 unsigned char c)
+{
+	unsigned int ier;
+	struct uart_8250_port *up = (struct uart_8250_port *)port;
+
+	/*
+	 *	First save the IER then disable the interrupts
+	 */
+	ier = serial_in(up, UART_IER);
+	if (up->capabilities & UART_CAP_UUE)
+		serial_out(up, UART_IER, UART_IER_UUE);
+	else
+		serial_out(up, UART_IER, 0);
+
+	wait_for_xmitr(up, BOTH_EMPTY);
+	/*
+	 *	Send the character out.
+	 *	If a LF, also do CR...
+	 */
+	serial_out(up, UART_TX, c);
+	if (c == 10) {
+		wait_for_xmitr(up, BOTH_EMPTY);
+		serial_out(up, UART_TX, 13);
+	}
+
+	/*
+	 *	Finally, wait for transmitter to become empty
+	 *	and restore the IER
+	 */
+	wait_for_xmitr(up, BOTH_EMPTY);
+	serial_out(up, UART_IER, ier);
+}
+
+#endif /* CONFIG_CONSOLE_POLL */
+
 static int serial8250_startup(struct uart_port *port)
 {
 	struct uart_8250_port *up = (struct uart_8250_port *)port;
@@ -2386,6 +2440,10 @@ static struct uart_ops serial8250_pops = {
 	.request_port	= serial8250_request_port,
 	.config_port	= serial8250_config_port,
 	.verify_port	= serial8250_verify_port,
+#ifdef CONFIG_CONSOLE_POLL
+	.poll_get_char = serial8250_get_poll_char,
+	.poll_put_char = serial8250_put_poll_char,
+#endif
 };
 
 static struct uart_8250_port serial8250_ports[UART_NR];
diff --git a/drivers/serial/Kconfig b/drivers/serial/Kconfig
index cf627cd1b4c8..f7cd9504d811 100644
--- a/drivers/serial/Kconfig
+++ b/drivers/serial/Kconfig
@@ -961,6 +961,9 @@ config SERIAL_CORE
 config SERIAL_CORE_CONSOLE
 	bool
 
+config CONSOLE_POLL
+	bool
+
 config SERIAL_68328
 	bool "68328 serial support"
 	depends on M68328 || M68EZ328 || M68VZ328
diff --git a/drivers/serial/Makefile b/drivers/serial/Makefile
index 640cfe44a56d..3cbea5494724 100644
--- a/drivers/serial/Makefile
+++ b/drivers/serial/Makefile
@@ -66,4 +66,5 @@ obj-$(CONFIG_SERIAL_UARTLITE) += uartlite.o
 obj-$(CONFIG_SERIAL_NETX) += netx-serial.o
 obj-$(CONFIG_SERIAL_OF_PLATFORM) += of_serial.o
 obj-$(CONFIG_SERIAL_KS8695) += serial_ks8695.o
+obj-$(CONFIG_KGDB_SERIAL_CONSOLE) += kgdboc.o
 obj-$(CONFIG_SERIAL_QE) += ucc_uart.o
diff --git a/drivers/serial/kgdboc.c b/drivers/serial/kgdboc.c
new file mode 100644
index 000000000000..341830791608
--- /dev/null
+++ b/drivers/serial/kgdboc.c
@@ -0,0 +1,163 @@
+/*
+ * Based on the same principle as kgdboe using the NETPOLL api, this
+ * driver uses a console polling api to implement a gdb serial inteface
+ * which is multiplexed on a console port.
+ *
+ * Maintainer: Jason Wessel <jason.wessel@windriver.com>
+ *
+ * 2007-2008 (c) Jason Wessel - Wind River Systems, Inc.
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+#include <linux/kernel.h>
+#include <linux/ctype.h>
+#include <linux/kgdb.h>
+#include <linux/tty.h>
+
+#define MAX_CONFIG_LEN		40
+
+static struct kgdb_io		kgdboc_io_ops;
+
+/* -1 = init not run yet, 0 = unconfigured, 1 = configured. */
+static int configured		= -1;
+
+static char config[MAX_CONFIG_LEN];
+static struct kparam_string kps = {
+	.string			= config,
+	.maxlen			= MAX_CONFIG_LEN,
+};
+
+static struct tty_driver	*kgdb_tty_driver;
+static int			kgdb_tty_line;
+
+static int kgdboc_option_setup(char *opt)
+{
+	if (strlen(opt) > MAX_CONFIG_LEN) {
+		printk(KERN_ERR "kgdboc: config string too long\n");
+		return -ENOSPC;
+	}
+	strcpy(config, opt);
+
+	return 0;
+}
+
+__setup("kgdboc=", kgdboc_option_setup);
+
+static int configure_kgdboc(void)
+{
+	struct tty_driver *p;
+	int tty_line = 0;
+	int err;
+
+	err = kgdboc_option_setup(config);
+	if (err || !strlen(config) || isspace(config[0]))
+		goto noconfig;
+
+	err = -ENODEV;
+
+	p = tty_find_polling_driver(config, &tty_line);
+	if (!p)
+		goto noconfig;
+
+	kgdb_tty_driver = p;
+	kgdb_tty_line = tty_line;
+
+	err = kgdb_register_io_module(&kgdboc_io_ops);
+	if (err)
+		goto noconfig;
+
+	configured = 1;
+
+	return 0;
+
+noconfig:
+	config[0] = 0;
+	configured = 0;
+
+	return err;
+}
+
+static int __init init_kgdboc(void)
+{
+	/* Already configured? */
+	if (configured == 1)
+		return 0;
+
+	return configure_kgdboc();
+}
+
+static void cleanup_kgdboc(void)
+{
+	if (configured == 1)
+		kgdb_unregister_io_module(&kgdboc_io_ops);
+}
+
+static int kgdboc_get_char(void)
+{
+	return kgdb_tty_driver->poll_get_char(kgdb_tty_driver, kgdb_tty_line);
+}
+
+static void kgdboc_put_char(u8 chr)
+{
+	kgdb_tty_driver->poll_put_char(kgdb_tty_driver, kgdb_tty_line, chr);
+}
+
+static int param_set_kgdboc_var(const char *kmessage, struct kernel_param *kp)
+{
+	if (strlen(kmessage) >= MAX_CONFIG_LEN) {
+		printk(KERN_ERR "kgdboc: config string too long\n");
+		return -ENOSPC;
+	}
+
+	/* Only copy in the string if the init function has not run yet */
+	if (configured < 0) {
+		strcpy(config, kmessage);
+		return 0;
+	}
+
+	if (kgdb_connected) {
+		printk(KERN_ERR
+		       "kgdboc: Cannot reconfigure while KGDB is connected.\n");
+
+		return -EBUSY;
+	}
+
+	strcpy(config, kmessage);
+
+	if (configured == 1)
+		cleanup_kgdboc();
+
+	/* Go and configure with the new params. */
+	return configure_kgdboc();
+}
+
+static void kgdboc_pre_exp_handler(void)
+{
+	/* Increment the module count when the debugger is active */
+	if (!kgdb_connected)
+		try_module_get(THIS_MODULE);
+}
+
+static void kgdboc_post_exp_handler(void)
+{
+	/* decrement the module count when the debugger detaches */
+	if (!kgdb_connected)
+		module_put(THIS_MODULE);
+}
+
+static struct kgdb_io kgdboc_io_ops = {
+	.name			= "kgdboc",
+	.read_char		= kgdboc_get_char,
+	.write_char		= kgdboc_put_char,
+	.pre_exception		= kgdboc_pre_exp_handler,
+	.post_exception		= kgdboc_post_exp_handler,
+};
+
+module_init(init_kgdboc);
+module_exit(cleanup_kgdboc);
+module_param_call(kgdboc, param_set_kgdboc_var, param_get_string, &kps, 0644);
+MODULE_PARM_DESC(kgdboc, "<serial_device>[,baud]");
+MODULE_DESCRIPTION("KGDB Console TTY Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/serial/serial_core.c b/drivers/serial/serial_core.c
index 0f5a17987cca..4d7eecbead9b 100644
--- a/drivers/serial/serial_core.c
+++ b/drivers/serial/serial_core.c
@@ -1827,7 +1827,7 @@ uart_get_console(struct uart_port *ports, int nr, struct console *co)
  *	options.  The format of the string is <baud><parity><bits><flow>,
  *	eg: 115200n8r
  */
-void __init
+void
 uart_parse_options(char *options, int *baud, int *parity, int *bits, int *flow)
 {
 	char *s = options;
@@ -1842,6 +1842,7 @@ uart_parse_options(char *options, int *baud, int *parity, int *bits, int *flow)
 	if (*s)
 		*flow = *s;
 }
+EXPORT_SYMBOL_GPL(uart_parse_options);
 
 struct baud_rates {
 	unsigned int rate;
@@ -1872,7 +1873,7 @@ static const struct baud_rates baud_rates[] = {
  *	@bits: number of data bits
  *	@flow: flow control character - 'r' (rts)
  */
-int __init
+int
 uart_set_options(struct uart_port *port, struct console *co,
 		 int baud, int parity, int bits, int flow)
 {
@@ -1924,10 +1925,16 @@ uart_set_options(struct uart_port *port, struct console *co,
 	port->mctrl |= TIOCM_DTR;
 
 	port->ops->set_termios(port, &termios, &dummy);
-	co->cflag = termios.c_cflag;
+	/*
+	 * Allow the setting of the UART parameters with a NULL console
+	 * too:
+	 */
+	if (co)
+		co->cflag = termios.c_cflag;
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(uart_set_options);
 #endif /* CONFIG_SERIAL_CORE_CONSOLE */
 
 static void uart_change_pm(struct uart_state *state, int pm_state)
@@ -2182,6 +2189,60 @@ uart_configure_port(struct uart_driver *drv, struct uart_state *state,
 	}
 }
 
+#ifdef CONFIG_CONSOLE_POLL
+
+static int uart_poll_init(struct tty_driver *driver, int line, char *options)
+{
+	struct uart_driver *drv = driver->driver_state;
+	struct uart_state *state = drv->state + line;
+	struct uart_port *port;
+	int baud = 9600;
+	int bits = 8;
+	int parity = 'n';
+	int flow = 'n';
+
+	if (!state || !state->port)
+		return -1;
+
+	port = state->port;
+	if (!(port->ops->poll_get_char && port->ops->poll_put_char))
+		return -1;
+
+	if (options) {
+		uart_parse_options(options, &baud, &parity, &bits, &flow);
+		return uart_set_options(port, NULL, baud, parity, bits, flow);
+	}
+
+	return 0;
+}
+
+static int uart_poll_get_char(struct tty_driver *driver, int line)
+{
+	struct uart_driver *drv = driver->driver_state;
+	struct uart_state *state = drv->state + line;
+	struct uart_port *port;
+
+	if (!state || !state->port)
+		return -1;
+
+	port = state->port;
+	return port->ops->poll_get_char(port);
+}
+
+static void uart_poll_put_char(struct tty_driver *driver, int line, char ch)
+{
+	struct uart_driver *drv = driver->driver_state;
+	struct uart_state *state = drv->state + line;
+	struct uart_port *port;
+
+	if (!state || !state->port)
+		return;
+
+	port = state->port;
+	port->ops->poll_put_char(port, ch);
+}
+#endif
+
 static const struct tty_operations uart_ops = {
 	.open		= uart_open,
 	.close		= uart_close,
@@ -2206,6 +2267,11 @@ static const struct tty_operations uart_ops = {
 #endif
 	.tiocmget	= uart_tiocmget,
 	.tiocmset	= uart_tiocmset,
+#ifdef CONFIG_CONSOLE_POLL
+	.poll_init	= uart_poll_init,
+	.poll_get_char	= uart_poll_get_char,
+	.poll_put_char	= uart_poll_put_char,
+#endif
 };
 
 /**
diff --git a/include/linux/serial_core.h b/include/linux/serial_core.h
index 289942fc6655..7cb094a82456 100644
--- a/include/linux/serial_core.h
+++ b/include/linux/serial_core.h
@@ -213,6 +213,10 @@ struct uart_ops {
 	void		(*config_port)(struct uart_port *, int);
 	int		(*verify_port)(struct uart_port *, struct serial_struct *);
 	int		(*ioctl)(struct uart_port *, unsigned int, unsigned long);
+#ifdef CONFIG_CONSOLE_POLL
+	void	(*poll_put_char)(struct uart_port *, unsigned char);
+	int		(*poll_get_char)(struct uart_port *);
+#endif
 };
 
 #define UART_CONFIG_TYPE	(1 << 0)
diff --git a/include/linux/tty_driver.h b/include/linux/tty_driver.h
index 85c95cd39bc3..21f69aca4505 100644
--- a/include/linux/tty_driver.h
+++ b/include/linux/tty_driver.h
@@ -125,6 +125,7 @@
 #include <linux/cdev.h>
 
 struct tty_struct;
+struct tty_driver;
 
 struct tty_operations {
 	int  (*open)(struct tty_struct * tty, struct file * filp);
@@ -157,6 +158,11 @@ struct tty_operations {
 	int (*tiocmget)(struct tty_struct *tty, struct file *file);
 	int (*tiocmset)(struct tty_struct *tty, struct file *file,
 			unsigned int set, unsigned int clear);
+#ifdef CONFIG_CONSOLE_POLL
+	int (*poll_init)(struct tty_driver *driver, int line, char *options);
+	int (*poll_get_char)(struct tty_driver *driver, int line);
+	void (*poll_put_char)(struct tty_driver *driver, int line, char ch);
+#endif
 };
 
 struct tty_driver {
@@ -220,6 +226,11 @@ struct tty_driver {
 	int (*tiocmget)(struct tty_struct *tty, struct file *file);
 	int (*tiocmset)(struct tty_struct *tty, struct file *file,
 			unsigned int set, unsigned int clear);
+#ifdef CONFIG_CONSOLE_POLL
+	int (*poll_init)(struct tty_driver *driver, int line, char *options);
+	int (*poll_get_char)(struct tty_driver *driver, int line);
+	void (*poll_put_char)(struct tty_driver *driver, int line, char ch);
+#endif
 
 	struct list_head tty_drivers;
 };
@@ -230,6 +241,7 @@ struct tty_driver *alloc_tty_driver(int lines);
 void put_tty_driver(struct tty_driver *driver);
 void tty_set_operations(struct tty_driver *driver,
 			const struct tty_operations *op);
+extern struct tty_driver *tty_find_polling_driver(char *name, int *line);
 
 /* tty driver magic number */
 #define TTY_DRIVER_MAGIC		0x5402
-- 
cgit v1.2.3


From 7c3078b637882303b1dcf6a16229d0e35f6b60a5 Mon Sep 17 00:00:00 2001
From: Jason Wessel <jason.wessel@windriver.com>
Date: Fri, 15 Feb 2008 14:55:54 -0600
Subject: kgdb: clocksource watchdog

In order to not trip the clocksource watchdog, kgdb must touch the
clocksource watchdog on the return to normal system run state.

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/clocksource.h |  1 +
 kernel/kgdb.c               |  4 ++++
 kernel/time/clocksource.c   | 12 ++++++++++++
 3 files changed, 17 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h
index 85778a4b1209..35094479ca55 100644
--- a/include/linux/clocksource.h
+++ b/include/linux/clocksource.h
@@ -216,6 +216,7 @@ static inline void clocksource_calculate_interval(struct clocksource *c,
 /* used to install a new clocksource */
 extern int clocksource_register(struct clocksource*);
 extern void clocksource_unregister(struct clocksource*);
+extern void clocksource_touch_watchdog(void);
 extern struct clocksource* clocksource_get_next(void);
 extern void clocksource_change_rating(struct clocksource *cs, int rating);
 extern void clocksource_resume(void);
diff --git a/kernel/kgdb.c b/kernel/kgdb.c
index 017ee782bc08..e3f603740425 100644
--- a/kernel/kgdb.c
+++ b/kernel/kgdb.c
@@ -28,6 +28,7 @@
  * kind, whether express or implied.
  */
 #include <linux/pid_namespace.h>
+#include <linux/clocksource.h>
 #include <linux/interrupt.h>
 #include <linux/spinlock.h>
 #include <linux/console.h>
@@ -574,6 +575,7 @@ static void kgdb_wait(struct pt_regs *regs)
 
 	/* Signal the primary CPU that we are done: */
 	atomic_set(&cpu_in_kgdb[cpu], 0);
+	clocksource_touch_watchdog();
 	local_irq_restore(flags);
 }
 #endif
@@ -1396,6 +1398,7 @@ acquirelock:
 	    atomic_read(&kgdb_cpu_doing_single_step) != cpu) {
 
 		atomic_set(&kgdb_active, -1);
+		clocksource_touch_watchdog();
 		local_irq_restore(flags);
 
 		goto acquirelock;
@@ -1487,6 +1490,7 @@ acquirelock:
 kgdb_restore:
 	/* Free kgdb_active */
 	atomic_set(&kgdb_active, -1);
+	clocksource_touch_watchdog();
 	local_irq_restore(flags);
 
 	return error;
diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index 7f60097d443a..f61402b1f2d0 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -221,6 +221,18 @@ void clocksource_resume(void)
 	spin_unlock_irqrestore(&clocksource_lock, flags);
 }
 
+/**
+ * clocksource_touch_watchdog - Update watchdog
+ *
+ * Update the watchdog after exception contexts such as kgdb so as not
+ * to incorrectly trip the watchdog.
+ *
+ */
+void clocksource_touch_watchdog(void)
+{
+	clocksource_resume_watchdog();
+}
+
 /**
  * clocksource_get_next - Returns the selected clocksource
  *
-- 
cgit v1.2.3


From e3e2aaf7dc0d82a055e084cfd48b9257c0c66b68 Mon Sep 17 00:00:00 2001
From: Jason Wessel <jason.wessel@windriver.com>
Date: Thu, 20 Mar 2008 13:43:45 -0500
Subject: kgdb: add documentation

Add in the kgdb documentation for kgdb.

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 Documentation/DocBook/Makefile  |   2 +-
 Documentation/DocBook/kgdb.tmpl | 435 ++++++++++++++++++++++++++++++++++++++++
 MAINTAINERS                     |   6 +
 include/linux/kgdb.h            |  52 +++--
 4 files changed, 473 insertions(+), 22 deletions(-)
 create mode 100644 Documentation/DocBook/kgdb.tmpl

(limited to 'include/linux')

diff --git a/Documentation/DocBook/Makefile b/Documentation/DocBook/Makefile
index 300e1707893f..e471bc466a7e 100644
--- a/Documentation/DocBook/Makefile
+++ b/Documentation/DocBook/Makefile
@@ -9,7 +9,7 @@
 DOCBOOKS := wanbook.xml z8530book.xml mcabook.xml videobook.xml \
 	    kernel-hacking.xml kernel-locking.xml deviceiobook.xml \
 	    procfs-guide.xml writing_usb_driver.xml networking.xml \
-	    kernel-api.xml filesystems.xml lsm.xml usb.xml \
+	    kernel-api.xml filesystems.xml lsm.xml usb.xml kgdb.xml \
 	    gadget.xml libata.xml mtdnand.xml librs.xml rapidio.xml \
 	    genericirq.xml s390-drivers.xml uio-howto.xml scsi.xml
 
diff --git a/Documentation/DocBook/kgdb.tmpl b/Documentation/DocBook/kgdb.tmpl
new file mode 100644
index 000000000000..95e5f84cbf56
--- /dev/null
+++ b/Documentation/DocBook/kgdb.tmpl
@@ -0,0 +1,435 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN"
+	"http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd" []>
+
+<book id="kgdbOnLinux">
+ <bookinfo>
+  <title>Using kgdb and the kgdb Internals</title>
+
+  <authorgroup>
+   <author>
+    <firstname>Jason</firstname>
+    <surname>Wessel</surname>
+    <affiliation>
+     <address>
+      <email>jason.wessel@windriver.com</email>
+     </address>
+    </affiliation>
+   </author>
+  </authorgroup>
+
+  <authorgroup>
+   <author>
+    <firstname>Tom</firstname>
+    <surname>Rini</surname>
+    <affiliation>
+     <address>
+      <email>trini@kernel.crashing.org</email>
+     </address>
+    </affiliation>
+   </author>
+  </authorgroup>
+
+  <authorgroup>
+   <author>
+    <firstname>Amit S.</firstname>
+    <surname>Kale</surname>
+    <affiliation>
+     <address>
+      <email>amitkale@linsyssoft.com</email>
+     </address>
+    </affiliation>
+   </author>
+  </authorgroup>
+
+  <copyright>
+   <year>2008</year>
+   <holder>Wind River Systems, Inc.</holder>
+  </copyright>
+  <copyright>
+   <year>2004-2005</year>
+   <holder>MontaVista Software, Inc.</holder>
+  </copyright>
+  <copyright>
+   <year>2004</year>
+   <holder>Amit S. Kale</holder>
+  </copyright>
+
+  <legalnotice>
+   <para>
+   This file is licensed under the terms of the GNU General Public License
+   version 2. This program is licensed "as is" without any warranty of any
+   kind, whether express or implied.
+   </para>
+
+  </legalnotice>
+ </bookinfo>
+
+<toc></toc>
+  <chapter id="Introduction">
+    <title>Introduction</title>
+    <para>
+    kgdb is a source level debugger for linux kernel. It is used along
+    with gdb to debug a linux kernel.  The expectation is that gdb can
+    be used to "break in" to the kernel to inspect memory, variables
+    and look through a cal stack information similar to what an
+    application developer would use gdb for.  It is possible to place
+    breakpoints in kernel code and perform some limited execution
+    stepping.
+    </para>
+    <para>
+    Two machines are required for using kgdb. One of these machines is a
+    development machine and the other is a test machine.  The kernel
+    to be debugged runs on the test machine. The development machine
+    runs an instance of gdb against the vmlinux file which contains
+    the symbols (not boot image such as bzImage, zImage, uImage...).
+    In gdb the developer specifies the connection parameters and
+    connects to kgdb.  Depending on which kgdb I/O modules exist in
+    the kernel for a given architecture, it may be possible to debug
+    the test machine's kernel with the development machine using a
+    rs232 or ethernet connection.
+    </para>
+  </chapter>
+  <chapter id="CompilingAKernel">
+    <title>Compiling a kernel</title>
+    <para>
+    To enable <symbol>CONFIG_KGDB</symbol>, look under the "Kernel debugging"
+    and then select "KGDB: kernel debugging with remote gdb".
+    </para>
+    <para>
+    Next you should choose one of more I/O drivers to interconnect debugging
+    host and debugged target.  Early boot debugging requires a KGDB
+    I/O driver that supports early debugging and the driver must be
+    built into the kernel directly. Kgdb I/O driver configuration
+    takes place via kernel or module parameters, see following
+    chapter.
+    </para>
+    <para>
+    The kgdb test compile options are described in the kgdb test suite chapter.
+    </para>
+
+  </chapter>
+  <chapter id="EnableKGDB">
+   <title>Enable kgdb for debugging</title>
+   <para>
+   In order to use kgdb you must activate it by passing configuration
+   information to one of the kgdb I/O drivers.  If you do not pass any
+   configuration information kgdb will not do anything at all.  Kgdb
+   will only actively hook up to the kernel trap hooks if a kgdb I/O
+   driver is loaded and configured.  If you unconfigure a kgdb I/O
+   driver, kgdb will unregister all the kernel hook points.
+   </para>
+   <para>
+   All drivers can be reconfigured at run time, if
+   <symbol>CONFIG_SYSFS</symbol> and <symbol>CONFIG_MODULES</symbol>
+   are enabled, by echo'ing a new config string to
+   <constant>/sys/module/&lt;driver&gt;/parameter/&lt;option&gt;</constant>.
+   The driver can be unconfigured by passing an empty string.  You cannot
+   change the configuration while the debugger is attached.  Make sure
+   to detach the debugger with the <constant>detach</constant> command
+   prior to trying unconfigure a kgdb I/O driver.
+   </para>
+   <sect1 id="kgdbwait">
+   <title>Kernel parameter: kgdbwait</title>
+   <para>
+   The Kernel command line option <constant>kgdbwait</constant> makes
+   kgdb wait for a debugger connection during booting of a kernel.  You
+   can only use this option you compiled a kgdb I/O driver into the
+   kernel and you specified the I/O driver configuration as a kernel
+   command line option.  The kgdbwait parameter should always follow the
+   configuration parameter for the kgdb I/O driver in the kernel
+   command line else the I/O driver will not be configured prior to
+   asking the kernel to use it to wait.
+   </para>
+   <para>
+   The kernel will stop and wait as early as the I/O driver and
+   architecture will allow when you use this option.  If you build the
+   kgdb I/O driver as a kernel module kgdbwait will not do anything.
+   </para>
+   </sect1>
+  <sect1 id="kgdboc">
+  <title>Kernel parameter: kgdboc</title>
+  <para>
+  The kgdboc driver was originally an abbreviation meant to stand for
+  "kgdb over console".  Kgdboc is designed to work with a single
+  serial port as example, and it was meant to cover the circumstance
+  where you wanted to use a serial console as your primary console as
+  well as using it to perform kernel debugging.
+  </para>
+  <sect2 id="UsingKgdboc">
+  <title>Using kgdboc</title>
+  <para>
+  You can configure kgdboc via sysfs or a module or kernel boot line
+  parameter depending on if you build with CONFIG_KGDBOC as a module
+  or built-in.
+  <orderedlist>
+  <listitem><para>From the module load or build-in</para>
+  <para><constant>kgdboc=&lt;tty-device&gt;,[baud]</constant></para>
+  <para>
+  The example here would be if your console port was typically ttyS0, you would use something like <constant>kgdboc=ttyS0,115200</constant> or on the ARM Versatile AB you would likely use <constant>kgdboc=ttyAMA0,115200</constant>
+  </para>
+  </listitem>
+  <listitem><para>From sysfs</para>
+  <para><constant>echo ttyS0 &gt; /sys/module/kgdboc/parameters/kgdboc</constant></para>
+  </listitem>
+  </orderedlist>
+  </para>
+  <para>
+  NOTE: Kgdboc does not support interrupting the target via the
+  gdb remote protocol.  You must manually send a sysrq-g unless you
+  have a proxy that splits console output to a terminal problem and
+  has a separate port for the debugger to connect to that sends the
+  sysrq-g for you.
+  </para>
+  <para>When using kgdboc with no debugger proxy, you can end up
+  connecting the debugger for one of two entry points.  If an
+  exception occurs after you have loaded kgdboc a message should print
+  on the console stating it is waiting for the debugger.  In case you
+  disconnect your terminal program and then connect the debugger in
+  its place.  If you want to interrupt the target system and forcibly
+  enter a debug session you have to issue a Sysrq sequence and then
+  type the letter <constant>g</constant>.  Then you disconnect the
+  terminal session and connect gdb.  Your options if you don't like
+  this are to hack gdb to send the sysrq-g for you as well as on the
+  initial connect, or to use a debugger proxy that allows an
+  unmodified gdb to do the debugging.
+  </para>
+  </sect2>
+  <sect2 id="kgdbocDesign">
+  <title>kgdboc internals</title>
+  <para>
+  The kgdboc driver is actually a very thin driver that relies on the
+  underlying low level to the hardware driver having "polling hooks"
+  which the to which the tty driver is attached.  In the initial
+  implementation of kgdboc it the serial_core was changed to expose a
+  low level uart hook for doing polled mode reading and writing of a
+  single character while in an atomic context.  When kgdb makes an I/O
+  request to the debugger, kgdboc invokes a call back in the serial
+  core which in turn uses the call back in the uart driver.  It is
+  certainly possible to extend kgdboc to work with non-uart based
+  consoles in the future.
+  </para>
+  <para>
+  When using kgdboc with a uart, the uart driver must implement two callbacks in the <constant>struct uart_ops</constant>. Example from drivers/8250.c:<programlisting>
+#ifdef CONFIG_CONSOLE_POLL
+	.poll_get_char = serial8250_get_poll_char,
+	.poll_put_char = serial8250_put_poll_char,
+#endif
+  </programlisting>
+  Any implementation specifics around creating a polling driver use the
+  <constant>#ifdef CONFIG_CONSOLE_POLL</constant>, as shown above.
+  Keep in mind that polling hooks have to be implemented in such a way
+  that they can be called from an atomic context and have to restore
+  the state of the uart chip on return such that the system can return
+  to normal when the debugger detaches.  You need to be very careful
+  with any kind of lock you consider, because failing here is most
+  going to mean pressing the reset button.
+  </para>
+  </sect2>
+  </sect1>
+  <sect1 id="kgdbcon">
+  <title>Kernel parameter: kgdbcon</title>
+  <para>
+  Kgdb supports using the gdb serial protocol to send console messages
+  to the debugger when the debugger is connected and running.  There
+  are two ways to activate this feature.
+  <orderedlist>
+  <listitem><para>Activate with the kernel command line option:</para>
+  <para><constant>kgdbcon</constant></para>
+  </listitem>
+  <listitem><para>Use sysfs before configuring an io driver</para>
+  <para>
+  <constant>echo 1 &gt; /sys/module/kgdb/parameters/kgdb_use_con</constant>
+  </para>
+  <para>
+  NOTE: If you do this after you configure the kgdb I/O driver, the
+  setting will not take effect until the next point the I/O is
+  reconfigured.
+  </para>
+  </listitem>
+  </orderedlist>
+  </para>
+  <para>
+  IMPORTANT NOTE: Using this option with kgdb over the console
+  (kgdboc) or kgdb over ethernet (kgdboe) is not supported.
+  </para>
+  </sect1>
+  </chapter>
+  <chapter id="ConnectingGDB">
+  <title>Connecting gdb</title>
+    <para>
+    If you are using kgdboc, you need to have used kgdbwait as a boot
+    argument, issued a sysrq-g, or the system you are going to debug
+    has already taken an exception and is waiting for the debugger to
+    attach before you can connect gdb.
+    </para>
+    <para>
+    If you are not using different kgdb I/O driver other than kgdboc,
+    you should be able to connect and the target will automatically
+    respond.
+    </para>
+    <para>
+    Example (using a serial port):
+    </para>
+    <programlisting>
+    % gdb ./vmlinux
+    (gdb) set remotebaud 115200
+    (gdb) target remote /dev/ttyS0
+    </programlisting>
+    <para>
+    Example (kgdb to a terminal server):
+    </para>
+    <programlisting>
+    % gdb ./vmlinux
+    (gdb) target remote udp:192.168.2.2:6443
+    </programlisting>
+    <para>
+    Example (kgdb over ethernet):
+    </para>
+    <programlisting>
+    % gdb ./vmlinux
+    (gdb) target remote udp:192.168.2.2:6443
+    </programlisting>
+    <para>
+    Once connected, you can debug a kernel the way you would debug an
+    application program.
+    </para>
+    <para>
+    If you are having problems connecting or something is going
+    seriously wrong while debugging, it will most often be the case
+    that you want to enable gdb to be verbose about its target
+    communications.  You do this prior to issuing the <constant>target
+    remote</constant> command by typing in: <constant>set remote debug 1</constant>
+    </para>
+  </chapter>
+  <chapter id="KGDBTestSuite">
+    <title>kgdb Test Suite</title>
+    <para>
+    When kgdb is enabled in the kernel config you can also elect to
+    enable the config parameter KGDB_TESTS.  Turning this on will
+    enable a special kgdb I/O module which is designed to test the
+    kgdb internal functions.
+    </para>
+    <para>
+    The kgdb tests are mainly intended for developers to test the kgdb
+    internals as well as a tool for developing a new kgdb architecture
+    specific implementation.  These tests are not really for end users
+    of the Linux kernel.  The primary source of documentation would be
+    to look in the drivers/misc/kgdbts.c file.
+    </para>
+    <para>
+    The kgdb test suite can also be configured at compile time to run
+    the core set of tests by setting the kernel config parameter
+    KGDB_TESTS_ON_BOOT.  This particular option is aimed at automated
+    regression testing and does not require modifying the kernel boot
+    config arguments.  If this is turned on, the kgdb test suite can
+    be disabled by specifying "kgdbts=" as a kernel boot argument.
+    </para>
+  </chapter>
+  <chapter id="CommonBackEndReq">
+    <title>Architecture Specifics</title>
+      <para>
+      Kgdb is organized into three basic components:
+      <orderedlist>
+      <listitem><para>kgdb core</para>
+      <para>
+      The kgdb core is found in kernel/kgdb.c.  It contains:
+      <itemizedlist>
+      <listitem><para>All the logic to implement the gdb serial protocol</para></listitem>
+      <listitem><para>A generic OS exception handler which includes sync'ing the processors into a stopped state on an multi cpu system.</para></listitem>
+      <listitem><para>The API to talk to the kgdb I/O drivers</para></listitem>
+      <listitem><para>The API to make calls to the arch specific kgdb implementation</para></listitem>
+      <listitem><para>The logic to perform safe memory reads and writes to memory while using the debugger</para></listitem>
+      <listitem><para>A full implementation for software breakpoints unless overridden by the arch</para></listitem>
+      </itemizedlist>
+      </para>
+      </listitem>
+      <listitem><para>kgdb arch specific implementation</para>
+      <para>
+      This implementation is generally found in arch/*/kernel/kgdb.c.
+      As an example, arch/x86/kernel/kgdb.c contains the specifics to
+      implement HW breakpoint as well as the initialization to
+      dynamically register and unregister for the trap handlers on
+      this architecture.  The arch specific portion implements:
+      <itemizedlist>
+      <listitem><para>contains an arch specific trap catcher which
+      invokes kgdb_handle_exception() to start kgdb about doing its
+      work</para></listitem>
+      <listitem><para>translation to and from gdb specific packet format to pt_regs</para></listitem>
+      <listitem><para>Registration and unregistration of architecture specific trap hooks</para></listitem>
+      <listitem><para>Any special exception handling and cleanup</para></listitem>
+      <listitem><para>NMI exception handling and cleanup</para></listitem>
+      <listitem><para>(optional)HW breakpoints</para></listitem>
+      </itemizedlist>
+      </para>
+      </listitem>
+      <listitem><para>kgdb I/O driver</para>
+      <para>
+      Each kgdb I/O driver has to provide an configuration
+      initialization, and cleanup handler for when it
+      unloads/unconfigures.  Any given kgdb I/O driver has to operate
+      very closely with the hardware and must do it in such a way that
+      does not enable interrupts or change other parts of the system
+      context without completely restoring them.  Every kgdb I/O
+      driver must provide a read and write character interface.  The
+      kgdb core will repeatedly "poll" a kgdb I/O driver for characters
+      when it needs input.  The I/O driver is expected to return
+      immediately if there is no data available.  Doing so allows for
+      the future possibility to touch watch dog hardware in such a way
+      as to have a target system not reset when these are enabled.
+      </para>
+      </listitem>
+      </orderedlist>
+      </para>
+      <para>
+      If you are intent on adding kgdb architecture specific support
+      for a new architecture, the architecture should define
+      <constant>HAVE_ARCH_KGDB</constant> in the architecture specific
+      Kconfig file.  This will enable kgdb for the architecture, and
+      at that point you must create an architecture specific kgdb
+      implementation.
+      </para>
+      <para>
+      There are a few flags which must be set on every architecture in
+      their &lt;asm/kgdb.h&gt; file.  These are:
+      <itemizedlist>
+        <listitem>
+	  <para>
+	  NUMREGBYTES: The size in bytes of all of the registers, so
+	  that we can ensure they will all fit into a packet.
+	  </para>
+	  <para>
+	  BUFMAX: The size in bytes of the buffer GDB will read into.
+	  This must be larger than NUMREGBYTES.
+	  </para>
+	  <para>
+	  CACHE_FLUSH_IS_SAFE: Set to 1 if it is always safe to call
+	  flush_cache_range or flush_icache_range.  On some architectures,
+	  these functions may not be safe to call on SMP since we keep other
+	  CPUs in a holding pattern.
+	  </para>
+	</listitem>
+      </itemizedlist>
+      </para>
+      <para>
+      There are also the following functions for the common backend,
+      found in kernel/kgdb.c, that must be supplied by the
+      architecture-specific backend unless marked as (optional), in
+      which case a default function maybe used if the architecture
+      does not need to provide a specific implementation.
+      </para>
+!Iinclude/linux/kgdb.h
+  </chapter>
+  <chapter id="credits">
+     <title>Credits</title>
+	<para>
+		The following people have contributed to this document:
+		<orderedlist>
+			<listitem><para>Amit Kale<email>amitkale@linsyssoft.com</email></para></listitem>
+			<listitem><para>Tom Rini<email>trini@kernel.crashing.org</email></para></listitem>
+			<listitem><para>Jason Wessel<email>jason.wessel@windriver.com</email></para></listitem>
+		</orderedlist>
+	</para>
+  </chapter>
+</book>
+
diff --git a/MAINTAINERS b/MAINTAINERS
index e46775868019..3eceebb48c92 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -2319,6 +2319,12 @@ L:	linux-kernel@vger.kernel.org
 L:	kexec@lists.infradead.org
 S:	Maintained
 
+KGDB
+P:	Jason Wessel
+M:	jason.wessel@windriver.com
+L:	kgdb-bugreport@lists.sourceforge.net
+S:	Maintained
+
 KPROBES
 P:	Ananth N Mavinakayanahalli
 M:	ananth@in.ibm.com
diff --git a/include/linux/kgdb.h b/include/linux/kgdb.h
index b0985b79b638..9757b1a6d9dc 100644
--- a/include/linux/kgdb.h
+++ b/include/linux/kgdb.h
@@ -22,31 +22,34 @@
 
 struct pt_regs;
 
-/*
- *	kgdb_skipexception - Bail out of KGDB when we've been triggered.
+/**
+ *	kgdb_skipexception - (optional) exit kgdb_handle_exception early
  *	@exception: Exception vector number
  *	@regs: Current &struct pt_regs.
  *
- *	On some architectures we need to skip a breakpoint exception when
- *	it occurs after a breakpoint has been removed.
+ *	On some architectures it is required to skip a breakpoint
+ *	exception when it occurs after a breakpoint has been removed.
+ *	This can be implemented in the architecture specific portion of
+ *	for kgdb.
  */
 extern int kgdb_skipexception(int exception, struct pt_regs *regs);
 
-/*
- *	kgdb_post_primary_code - Save error vector/code numbers.
+/**
+ *	kgdb_post_primary_code - (optional) Save error vector/code numbers.
  *	@regs: Original pt_regs.
  *	@e_vector: Original error vector.
  *	@err_code: Original error code.
  *
- *	This is needed on architectures which support SMP and KGDB.
- *	This function is called after all the secondary cpus have been put
- *	to a know spin state and the primary CPU has control over KGDB.
+ *	This is usually needed on architectures which support SMP and
+ *	KGDB.  This function is called after all the secondary cpus have
+ *	been put to a know spin state and the primary CPU has control over
+ *	KGDB.
  */
 extern void kgdb_post_primary_code(struct pt_regs *regs, int e_vector,
 				  int err_code);
 
-/*
- *	kgdb_disable_hw_debug - Disable hardware debugging while we in kgdb.
+/**
+ *	kgdb_disable_hw_debug - (optional) Disable hardware debugging hook
  *	@regs: Current &struct pt_regs.
  *
  *	This function will be called if the particular architecture must
@@ -59,7 +62,14 @@ struct tasklet_struct;
 struct task_struct;
 struct uart_port;
 
-/* To enter the debugger explicitly. */
+/**
+ *	kgdb_breakpoint - compiled in breakpoint
+ *
+ *	This will be impelmented a static inline per architecture.  This
+ *	function is called by the kgdb core to execute an architecture
+ *	specific trap to cause kgdb to enter the exception processing.
+ *
+ */
 void kgdb_breakpoint(void);
 
 extern int kgdb_connected;
@@ -102,7 +112,7 @@ struct kgdb_bkpt {
  * Functions each KGDB-supporting architecture must provide:
  */
 
-/*
+/**
  *	kgdb_arch_init - Perform any architecture specific initalization.
  *
  *	This function will handle the initalization of any architecture
@@ -110,7 +120,7 @@ struct kgdb_bkpt {
  */
 extern int kgdb_arch_init(void);
 
-/*
+/**
  *	kgdb_arch_exit - Perform any architecture specific uninitalization.
  *
  *	This function will handle the uninitalization of any architecture
@@ -118,7 +128,7 @@ extern int kgdb_arch_init(void);
  */
 extern void kgdb_arch_exit(void);
 
-/*
+/**
  *	pt_regs_to_gdb_regs - Convert ptrace regs to GDB regs
  *	@gdb_regs: A pointer to hold the registers in the order GDB wants.
  *	@regs: The &struct pt_regs of the current process.
@@ -128,7 +138,7 @@ extern void kgdb_arch_exit(void);
  */
 extern void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs);
 
-/*
+/**
  *	sleeping_thread_to_gdb_regs - Convert ptrace regs to GDB regs
  *	@gdb_regs: A pointer to hold the registers in the order GDB wants.
  *	@p: The &struct task_struct of the desired process.
@@ -143,7 +153,7 @@ extern void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs);
 extern void
 sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p);
 
-/*
+/**
  *	gdb_regs_to_pt_regs - Convert GDB regs to ptrace regs.
  *	@gdb_regs: A pointer to hold the registers we've received from GDB.
  *	@regs: A pointer to a &struct pt_regs to hold these values in.
@@ -153,7 +163,7 @@ sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p);
  */
 extern void gdb_regs_to_pt_regs(unsigned long *gdb_regs, struct pt_regs *regs);
 
-/*
+/**
  *	kgdb_arch_handle_exception - Handle architecture specific GDB packets.
  *	@vector: The error vector of the exception that happened.
  *	@signo: The signal number of the exception that happened.
@@ -175,7 +185,7 @@ kgdb_arch_handle_exception(int vector, int signo, int err_code,
 			   char *remcom_out_buffer,
 			   struct pt_regs *regs);
 
-/*
+/**
  *	kgdb_roundup_cpus - Get other CPUs into a holding pattern
  *	@flags: Current IRQ state
  *
@@ -198,7 +208,7 @@ extern int kgdb_validate_break_address(unsigned long addr);
 extern int kgdb_arch_set_breakpoint(unsigned long addr, char *saved_instr);
 extern int kgdb_arch_remove_breakpoint(unsigned long addr, char *bundle);
 
-/*
+/**
  * struct kgdb_arch - Describe architecture specific values.
  * @gdb_bpt_instr: The instruction to trigger a breakpoint.
  * @flags: Flags for the breakpoint, currently just %KGDB_HW_BREAKPOINT.
@@ -227,7 +237,7 @@ struct kgdb_arch {
 	void	(*correct_hw_break)(void);
 };
 
-/*
+/**
  * struct kgdb_io - Describe the interface for an I/O driver to talk with KGDB.
  * @name: Name of the I/O driver.
  * @read_char: Pointer to a function that will return one char.
-- 
cgit v1.2.3


From cac1f3c8a80f3fc0b4489d1d3ba29214677ffab2 Mon Sep 17 00:00:00 2001
From: Paul Gortmaker <paul.gortmaker@windriver.com>
Date: Tue, 15 Apr 2008 12:49:21 -0400
Subject: phylib: factor out get_phy_id from within get_phy_device

We were already doing what amounts to a get_phy_id from within
get_phy_device, and rather than duplicate this for the TBIPA
probing, we might as well just factor it out and make it available
instead.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Acked-by: Andy Fleming <afleming@freescale.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
---
 drivers/net/phy/phy_device.c | 38 +++++++++++++++++++++++++++++---------
 include/linux/phy.h          |  1 +
 2 files changed, 30 insertions(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c
index f4c4fd85425f..8b1121b02f98 100644
--- a/drivers/net/phy/phy_device.c
+++ b/drivers/net/phy/phy_device.c
@@ -86,35 +86,55 @@ struct phy_device* phy_device_create(struct mii_bus *bus, int addr, int phy_id)
 EXPORT_SYMBOL(phy_device_create);
 
 /**
- * get_phy_device - reads the specified PHY device and returns its @phy_device struct
+ * get_phy_id - reads the specified addr for its ID.
  * @bus: the target MII bus
  * @addr: PHY address on the MII bus
+ * @phy_id: where to store the ID retrieved.
  *
  * Description: Reads the ID registers of the PHY at @addr on the
- *   @bus, then allocates and returns the phy_device to represent it.
+ *   @bus, stores it in @phy_id and returns zero on success.
  */
-struct phy_device * get_phy_device(struct mii_bus *bus, int addr)
+int get_phy_id(struct mii_bus *bus, int addr, u32 *phy_id)
 {
 	int phy_reg;
-	u32 phy_id;
-	struct phy_device *dev = NULL;
 
 	/* Grab the bits from PHYIR1, and put them
 	 * in the upper half */
 	phy_reg = bus->read(bus, addr, MII_PHYSID1);
 
 	if (phy_reg < 0)
-		return ERR_PTR(phy_reg);
+		return -EIO;
 
-	phy_id = (phy_reg & 0xffff) << 16;
+	*phy_id = (phy_reg & 0xffff) << 16;
 
 	/* Grab the bits from PHYIR2, and put them in the lower half */
 	phy_reg = bus->read(bus, addr, MII_PHYSID2);
 
 	if (phy_reg < 0)
-		return ERR_PTR(phy_reg);
+		return -EIO;
+
+	*phy_id |= (phy_reg & 0xffff);
+
+	return 0;
+}
+
+/**
+ * get_phy_device - reads the specified PHY device and returns its @phy_device struct
+ * @bus: the target MII bus
+ * @addr: PHY address on the MII bus
+ *
+ * Description: Reads the ID registers of the PHY at @addr on the
+ *   @bus, then allocates and returns the phy_device to represent it.
+ */
+struct phy_device * get_phy_device(struct mii_bus *bus, int addr)
+{
+	struct phy_device *dev = NULL;
+	u32 phy_id;
+	int r;
 
-	phy_id |= (phy_reg & 0xffff);
+	r = get_phy_id(bus, addr, &phy_id);
+	if (r)
+		return ERR_PTR(r);
 
 	/* If the phy_id is all Fs, there is no device there */
 	if (0xffffffff == phy_id)
diff --git a/include/linux/phy.h b/include/linux/phy.h
index 2d838448415c..779cbcd65f62 100644
--- a/include/linux/phy.h
+++ b/include/linux/phy.h
@@ -381,6 +381,7 @@ struct phy_driver {
 
 int phy_read(struct phy_device *phydev, u16 regnum);
 int phy_write(struct phy_device *phydev, u16 regnum, u16 val);
+int get_phy_id(struct mii_bus *bus, int addr, u32 *phy_id);
 struct phy_device* get_phy_device(struct mii_bus *bus, int addr);
 int phy_clear_interrupt(struct phy_device *phydev);
 int phy_config_interrupt(struct phy_device *phydev, u32 interrupts);
-- 
cgit v1.2.3


From cf48062658e7ab3bc55e10c65676c3c73c16f8bf Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 24 Jan 2008 00:05:14 +0900
Subject: libata: prefer hardreset

When both soft and hard resets are available, libata preferred
softreset till now.  The logic behind it was to be softer to devices;
however, this doesn't really help much.  Rationales for the change:

* BIOS may freeze lock certain things during boot and softreset can't
  unlock those.  This by itself is okay but during operation PHY event
  or other error conditions can trigger hardreset and the device may
  end up with different configuration.

  For example, after a hardreset, previously unlockable HPA can be
  unlocked resulting in different device size and thus revalidation
  failure.  Similar condition can occur during or after resume.

* Certain ATAPI devices require hardreset to recover after certain
  error conditions.  On PATA, this is done by issuing the DEVICE RESET
  command.  On SATA, COMRESET has equivalent effect.  The problem is
  that DEVICE RESET needs its own execution protocol.

  For SFF controllers with bare TF access, it can be easily
  implemented but more advanced controllers (e.g. ahci and sata_sil24)
  require specialized implementations.  Simply using hardreset solves
  the problem nicely.

* COMRESET initialization sequence is the norm in SATA land and many
  SATA devices don't work properly if only SRST is used.  For example,
  some PMPs behave this way and libata works around by always issuing
  hardreset if the host supports PMP.

  Like the above example, libata has developed a number of mechanisms
  aiming to promote softreset to hardreset if softreset is not going
  to work.  This approach is time consuming and error prone.

  Also, note that, dependingon how you read the specs, it could be
  argued that PMP fan-out ports require COMRESET to start operation.
  In fact, all the PMPs on the market except one don't work properly
  if COMRESET is not issued to fan-out ports after PMP reset.

* COMRESET is an integral part of SATA connection and any working
  device should be able to handle COMRESET properly.  After all, it's
  the way to signal hardreset during reboot.  This is the most used
  and recommended (at least by the ahci spec) method of resetting
  devices.

So, this patch makes libata prefer hardreset over softreset by making
the following changes.

* Rename ATA_EH_RESET_MASK to ATA_EH_RESET and use it whereever
  ATA_EH_{SOFT|HARD}RESET used to be used.  ATA_EH_{SOFT|HARD}RESET is
  now only used to tell prereset whether soft or hard reset will be
  issued.

* Strip out now unneeded promote-to-hardreset logics from
  ata_eh_reset(), ata_std_prereset(), sata_pmp_std_prereset() and
  other places.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c        | 10 ++---
 drivers/ata/libata-core.c | 19 ++-------
 drivers/ata/libata-eh.c   | 99 +++++++++++++++++------------------------------
 drivers/ata/libata-pmp.c  | 28 ++++----------
 drivers/ata/libata-scsi.c |  4 +-
 drivers/ata/sata_fsl.c    |  2 +-
 drivers/ata/sata_mv.c     | 31 ++++-----------
 drivers/ata/sata_nv.c     | 12 +++---
 drivers/ata/sata_sil24.c  | 54 +++++++++++++-------------
 drivers/ata/sata_via.c    |  2 +-
 include/linux/libata.h    |  8 ++--
 11 files changed, 101 insertions(+), 168 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index b1eb4e24c86a..f6bbd52b1547 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1663,7 +1663,7 @@ static void ahci_error_intr(struct ata_port *ap, u32 irq_stat)
 		u32 *unk = (u32 *)(pp->rx_fis + RX_FIS_UNK);
 
 		active_ehi->err_mask |= AC_ERR_HSM;
-		active_ehi->action |= ATA_EH_SOFTRESET;
+		active_ehi->action |= ATA_EH_RESET;
 		ata_ehi_push_desc(active_ehi,
 				  "unknown FIS %08x %08x %08x %08x" ,
 				  unk[0], unk[1], unk[2], unk[3]);
@@ -1671,19 +1671,19 @@ static void ahci_error_intr(struct ata_port *ap, u32 irq_stat)
 
 	if (ap->nr_pmp_links && (irq_stat & PORT_IRQ_BAD_PMP)) {
 		active_ehi->err_mask |= AC_ERR_HSM;
-		active_ehi->action |= ATA_EH_SOFTRESET;
+		active_ehi->action |= ATA_EH_RESET;
 		ata_ehi_push_desc(active_ehi, "incorrect PMP");
 	}
 
 	if (irq_stat & (PORT_IRQ_HBUS_ERR | PORT_IRQ_HBUS_DATA_ERR)) {
 		host_ehi->err_mask |= AC_ERR_HOST_BUS;
-		host_ehi->action |= ATA_EH_SOFTRESET;
+		host_ehi->action |= ATA_EH_RESET;
 		ata_ehi_push_desc(host_ehi, "host bus error");
 	}
 
 	if (irq_stat & PORT_IRQ_IF_ERR) {
 		host_ehi->err_mask |= AC_ERR_ATA_BUS;
-		host_ehi->action |= ATA_EH_SOFTRESET;
+		host_ehi->action |= ATA_EH_RESET;
 		ata_ehi_push_desc(host_ehi, "interface fatal error");
 	}
 
@@ -1771,7 +1771,7 @@ static void ahci_port_intr(struct ata_port *ap)
 	/* while resetting, invalid completions are expected */
 	if (unlikely(rc < 0 && !resetting)) {
 		ehi->err_mask |= AC_ERR_HSM;
-		ehi->action |= ATA_EH_SOFTRESET;
+		ehi->action |= ATA_EH_RESET;
 		ata_port_freeze(ap);
 	}
 }
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index be95fdb69726..02e7ba43a3b3 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -3949,17 +3949,6 @@ int ata_std_prereset(struct ata_link *link, unsigned long deadline)
 	const unsigned long *timing = sata_ehc_deb_timing(ehc);
 	int rc;
 
-	/* handle link resume */
-	if ((ehc->i.flags & ATA_EHI_RESUME_LINK) &&
-	    (link->flags & ATA_LFLAG_HRST_TO_RESUME))
-		ehc->i.action |= ATA_EH_HARDRESET;
-
-	/* Some PMPs don't work with only SRST, force hardreset if PMP
-	 * is supported.
-	 */
-	if (ap->flags & ATA_FLAG_PMP)
-		ehc->i.action |= ATA_EH_HARDRESET;
-
 	/* if we're about to do hardreset, nothing more to do */
 	if (ehc->i.action & ATA_EH_HARDRESET)
 		return 0;
@@ -6055,9 +6044,9 @@ void ata_qc_issue(struct ata_queued_cmd *qc)
 		if (ata_sg_setup(qc))
 			goto sg_err;
 
-	/* if device is sleeping, schedule softreset and abort the link */
+	/* if device is sleeping, schedule reset and abort the link */
 	if (unlikely(qc->dev->flags & ATA_DFLAG_SLEEPING)) {
-		link->eh_info.action |= ATA_EH_SOFTRESET;
+		link->eh_info.action |= ATA_EH_RESET;
 		ata_ehi_push_desc(&link->eh_info, "waking up from sleep");
 		ata_link_abort(link);
 		return;
@@ -6634,7 +6623,7 @@ int ata_host_suspend(struct ata_host *host, pm_message_t mesg)
  */
 void ata_host_resume(struct ata_host *host)
 {
-	ata_host_request_pm(host, PMSG_ON, ATA_EH_SOFTRESET,
+	ata_host_request_pm(host, PMSG_ON, ATA_EH_RESET,
 			    ATA_EHI_NO_AUTOPSY | ATA_EHI_QUIET, 0);
 	host->dev->power.power_state = PMSG_ON;
 
@@ -7171,7 +7160,7 @@ int ata_host_register(struct ata_host *host, struct scsi_host_template *sht)
 
 			ehi->probe_mask =
 				(1 << ata_link_max_devices(&ap->link)) - 1;
-			ehi->action |= ATA_EH_SOFTRESET;
+			ehi->action |= ATA_EH_RESET;
 			ehi->flags |= ATA_EHI_NO_AUTOPSY | ATA_EHI_QUIET;
 
 			ap->pflags &= ~ATA_PFLAG_INITIALIZING;
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index a5830329eda4..f7cae6400155 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -1079,16 +1079,9 @@ void ata_eh_about_to_do(struct ata_link *link, struct ata_device *dev,
 
 	spin_lock_irqsave(ap->lock, flags);
 
-	/* Reset is represented by combination of actions and EHI
-	 * flags.  Suck in all related bits before clearing eh_info to
-	 * avoid losing requested action.
-	 */
-	if (action & ATA_EH_RESET_MASK) {
-		ehc->i.action |= ehi->action & ATA_EH_RESET_MASK;
+	/* suck in and clear reset modifier */
+	if (action & ATA_EH_RESET) {
 		ehc->i.flags |= ehi->flags & ATA_EHI_RESET_MODIFIER_MASK;
-
-		/* make sure all reset actions are cleared & clear EHI flags */
-		action |= ATA_EH_RESET_MASK;
 		ehi->flags &= ~ATA_EHI_RESET_MODIFIER_MASK;
 	}
 
@@ -1117,11 +1110,9 @@ void ata_eh_done(struct ata_link *link, struct ata_device *dev,
 {
 	struct ata_eh_context *ehc = &link->eh_context;
 
-	/* if reset is complete, clear all reset actions & reset modifier */
-	if (action & ATA_EH_RESET_MASK) {
-		action |= ATA_EH_RESET_MASK;
+	/* if reset is complete, clear reset modifier */
+	if (action & ATA_EH_RESET)
 		ehc->i.flags &= ~ATA_EHI_RESET_MODIFIER_MASK;
-	}
 
 	ata_eh_clear_action(link, dev, &ehc->i, action);
 }
@@ -1329,20 +1320,20 @@ static void ata_eh_analyze_serror(struct ata_link *link)
 
 	if (serror & SERR_PERSISTENT) {
 		err_mask |= AC_ERR_ATA_BUS;
-		action |= ATA_EH_HARDRESET;
+		action |= ATA_EH_RESET;
 	}
 	if (serror &
 	    (SERR_DATA_RECOVERED | SERR_COMM_RECOVERED | SERR_DATA)) {
 		err_mask |= AC_ERR_ATA_BUS;
-		action |= ATA_EH_SOFTRESET;
+		action |= ATA_EH_RESET;
 	}
 	if (serror & SERR_PROTOCOL) {
 		err_mask |= AC_ERR_HSM;
-		action |= ATA_EH_SOFTRESET;
+		action |= ATA_EH_RESET;
 	}
 	if (serror & SERR_INTERNAL) {
 		err_mask |= AC_ERR_SYSTEM;
-		action |= ATA_EH_HARDRESET;
+		action |= ATA_EH_RESET;
 	}
 
 	/* Determine whether a hotplug event has occurred.  Both
@@ -1448,7 +1439,7 @@ static unsigned int ata_eh_analyze_tf(struct ata_queued_cmd *qc,
 
 	if ((stat & (ATA_BUSY | ATA_DRQ | ATA_DRDY)) != ATA_DRDY) {
 		qc->err_mask |= AC_ERR_HSM;
-		return ATA_EH_SOFTRESET;
+		return ATA_EH_RESET;
 	}
 
 	if (stat & (ATA_ERR | ATA_DF))
@@ -1484,7 +1475,7 @@ static unsigned int ata_eh_analyze_tf(struct ata_queued_cmd *qc,
 	}
 
 	if (qc->err_mask & (AC_ERR_HSM | AC_ERR_TIMEOUT | AC_ERR_ATA_BUS))
-		action |= ATA_EH_SOFTRESET;
+		action |= ATA_EH_RESET;
 
 	return action;
 }
@@ -1685,7 +1676,7 @@ static unsigned int ata_eh_speed_down(struct ata_device *dev,
 	if (verdict & ATA_EH_SPDN_SPEED_DOWN) {
 		/* speed down SATA link speed if possible */
 		if (sata_down_spd_limit(link) == 0) {
-			action |= ATA_EH_HARDRESET;
+			action |= ATA_EH_RESET;
 			goto done;
 		}
 
@@ -1705,7 +1696,7 @@ static unsigned int ata_eh_speed_down(struct ata_device *dev,
 			dev->spdn_cnt++;
 
 			if (ata_down_xfermask_limit(dev, sel) == 0) {
-				action |= ATA_EH_SOFTRESET;
+				action |= ATA_EH_RESET;
 				goto done;
 			}
 		}
@@ -1719,7 +1710,7 @@ static unsigned int ata_eh_speed_down(struct ata_device *dev,
 	    (dev->xfer_shift != ATA_SHIFT_PIO)) {
 		if (ata_down_xfermask_limit(dev, ATA_DNXFER_FORCE_PIO) == 0) {
 			dev->spdn_cnt = 0;
-			action |= ATA_EH_SOFTRESET;
+			action |= ATA_EH_RESET;
 			goto done;
 		}
 	}
@@ -1764,9 +1755,9 @@ static void ata_eh_link_autopsy(struct ata_link *link)
 		ehc->i.serror |= serror;
 		ata_eh_analyze_serror(link);
 	} else if (rc != -EOPNOTSUPP) {
-		/* SError read failed, force hardreset and probing */
+		/* SError read failed, force reset and probing */
 		ata_ehi_schedule_probe(&ehc->i);
-		ehc->i.action |= ATA_EH_HARDRESET;
+		ehc->i.action |= ATA_EH_RESET;
 		ehc->i.err_mask |= AC_ERR_OTHER;
 	}
 
@@ -1814,7 +1805,7 @@ static void ata_eh_link_autopsy(struct ata_link *link)
 	/* enforce default EH actions */
 	if (ap->pflags & ATA_PFLAG_FROZEN ||
 	    all_err_mask & (AC_ERR_HSM | AC_ERR_TIMEOUT))
-		ehc->i.action |= ATA_EH_SOFTRESET;
+		ehc->i.action |= ATA_EH_RESET;
 	else if (((eflags & ATA_EFLAG_IS_IO) && all_err_mask) ||
 		 (!(eflags & ATA_EFLAG_IS_IO) && (all_err_mask & ~AC_ERR_DEV)))
 		ehc->i.action |= ATA_EH_REVALIDATE;
@@ -2118,7 +2109,6 @@ int ata_eh_reset(struct ata_link *link, int classify,
 	int try = 0;
 	struct ata_device *dev;
 	unsigned long deadline, now;
-	unsigned int tmp_action;
 	ata_reset_fn_t reset;
 	unsigned long flags;
 	u32 sstatus;
@@ -2129,7 +2119,7 @@ int ata_eh_reset(struct ata_link *link, int classify,
 	ap->pflags |= ATA_PFLAG_RESETTING;
 	spin_unlock_irqrestore(ap->lock, flags);
 
-	ata_eh_about_to_do(link, NULL, ehc->i.action & ATA_EH_RESET_MASK);
+	ata_eh_about_to_do(link, NULL, ATA_EH_RESET);
 
 	ata_link_for_each_dev(dev, link) {
 		/* If we issue an SRST then an ATA drive (not ATAPI)
@@ -2159,17 +2149,15 @@ int ata_eh_reset(struct ata_link *link, int classify,
 		goto done;
 	}
 
-	/* Determine which reset to use and record in ehc->i.action.
-	 * prereset() may examine and modify it.
-	 */
-	if (softreset && (!hardreset || (!(lflags & ATA_LFLAG_NO_SRST) &&
-					 !sata_set_spd_needed(link) &&
-					 !(ehc->i.action & ATA_EH_HARDRESET))))
-		tmp_action = ATA_EH_SOFTRESET;
-	else
-		tmp_action = ATA_EH_HARDRESET;
-
-	ehc->i.action = (ehc->i.action & ~ATA_EH_RESET_MASK) | tmp_action;
+	/* prefer hardreset */
+	ehc->i.action &= ~ATA_EH_RESET;
+	if (hardreset) {
+		reset = hardreset;
+		ehc->i.action = ATA_EH_HARDRESET;
+	} else {
+		reset = softreset;
+		ehc->i.action = ATA_EH_SOFTRESET;
+	}
 
 	if (prereset) {
 		rc = prereset(link, jiffies + ATA_EH_PRERESET_TIMEOUT);
@@ -2177,7 +2165,7 @@ int ata_eh_reset(struct ata_link *link, int classify,
 			if (rc == -ENOENT) {
 				ata_link_printk(link, KERN_DEBUG,
 						"port disabled. ignoring.\n");
-				ehc->i.action &= ~ATA_EH_RESET_MASK;
+				ehc->i.action &= ~ATA_EH_RESET;
 
 				ata_link_for_each_dev(dev, link)
 					classes[dev->devno] = ATA_DEV_NONE;
@@ -2190,12 +2178,8 @@ int ata_eh_reset(struct ata_link *link, int classify,
 		}
 	}
 
-	/* prereset() might have modified ehc->i.action */
-	if (ehc->i.action & ATA_EH_HARDRESET)
-		reset = hardreset;
-	else if (ehc->i.action & ATA_EH_SOFTRESET)
-		reset = softreset;
-	else {
+	/* prereset() might have cleared ATA_EH_RESET */
+	if (!(ehc->i.action & ATA_EH_RESET)) {
 		/* prereset told us not to reset, bang classes and return */
 		ata_link_for_each_dev(dev, link)
 			classes[dev->devno] = ATA_DEV_NONE;
@@ -2203,14 +2187,6 @@ int ata_eh_reset(struct ata_link *link, int classify,
 		goto out;
 	}
 
-	/* did prereset() screw up?  if so, fix up to avoid oopsing */
-	if (!reset) {
-		if (softreset)
-			reset = softreset;
-		else
-			reset = hardreset;
-	}
-
  retry:
 	deadline = jiffies + ata_eh_reset_timeouts[try++];
 
@@ -2240,7 +2216,7 @@ int ata_eh_reset(struct ata_link *link, int classify,
 			goto fail;
 		}
 
-		ata_eh_about_to_do(link, NULL, ATA_EH_RESET_MASK);
+		ata_eh_about_to_do(link, NULL, ATA_EH_RESET);
 		rc = ata_do_reset(link, reset, classes, deadline);
 	}
 
@@ -2290,7 +2266,7 @@ int ata_eh_reset(struct ata_link *link, int classify,
 		postreset(link, classes);
 
 	/* reset successful, schedule revalidation */
-	ata_eh_done(link, NULL, ehc->i.action & ATA_EH_RESET_MASK);
+	ata_eh_done(link, NULL, ATA_EH_RESET);
 	ehc->i.action |= ATA_EH_REVALIDATE;
 
 	rc = 0;
@@ -2548,7 +2524,7 @@ static int ata_eh_schedule_probe(struct ata_device *dev)
 	ata_eh_detach_dev(dev);
 	ata_dev_init(dev);
 	ehc->did_probe_mask |= (1 << dev->devno);
-	ehc->i.action |= ATA_EH_SOFTRESET;
+	ehc->i.action |= ATA_EH_RESET;
 	ehc->saved_xfer_mode[dev->devno] = 0;
 	ehc->saved_ncq_enabled &= ~(1 << dev->devno);
 
@@ -2592,12 +2568,7 @@ static int ata_eh_handle_dev_fail(struct ata_device *dev, int err)
 
 		return 1;
 	} else {
-		/* soft didn't work?  be haaaaard */
-		if (ehc->i.flags & ATA_EHI_DID_RESET)
-			ehc->i.action |= ATA_EH_HARDRESET;
-		else
-			ehc->i.action |= ATA_EH_SOFTRESET;
-
+		ehc->i.action |= ATA_EH_RESET;
 		return 0;
 	}
 }
@@ -2690,7 +2661,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 			ehc->i.action = 0;
 
 		/* do we need to reset? */
-		if (ehc->i.action & ATA_EH_RESET_MASK)
+		if (ehc->i.action & ATA_EH_RESET)
 			reset = 1;
 
 		ata_link_for_each_dev(dev, link)
@@ -2708,7 +2679,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 		ata_port_for_each_link(link, ap) {
 			struct ata_eh_context *ehc = &link->eh_context;
 
-			if (!(ehc->i.action & ATA_EH_RESET_MASK))
+			if (!(ehc->i.action & ATA_EH_RESET))
 				continue;
 
 			rc = ata_eh_reset(link, ata_link_nr_vacant(link),
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index d91f5090ba9d..8439fc8efdd6 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -194,15 +194,6 @@ int sata_pmp_std_prereset(struct ata_link *link, unsigned long deadline)
 	const unsigned long *timing = sata_ehc_deb_timing(ehc);
 	int rc;
 
-	/* force HRST? */
-	if (link->flags & ATA_LFLAG_NO_SRST)
-		ehc->i.action |= ATA_EH_HARDRESET;
-
-	/* handle link resume */
-	if ((ehc->i.flags & ATA_EHI_RESUME_LINK) &&
-	    (link->flags & ATA_LFLAG_HRST_TO_RESUME))
-		ehc->i.action |= ATA_EH_HARDRESET;
-
 	/* if we're about to do hardreset, nothing more to do */
 	if (ehc->i.action & ATA_EH_HARDRESET)
 		return 0;
@@ -445,7 +436,7 @@ static int sata_pmp_init_links(struct ata_port *ap, int nr_ports)
 
 		link->flags = 0;
 		ehc->i.probe_mask |= 1;
-		ehc->i.action |= ATA_EH_SOFTRESET;
+		ehc->i.action |= ATA_EH_RESET;
 		ehc->i.flags |= ATA_EHI_RESUME_LINK;
 	}
 
@@ -840,13 +831,12 @@ static int sata_pmp_eh_recover_pmp(struct ata_port *ap,
  retry:
 	ehc->classes[0] = ATA_DEV_UNKNOWN;
 
-	if (ehc->i.action & ATA_EH_RESET_MASK) {
+	if (ehc->i.action & ATA_EH_RESET) {
 		struct ata_link *tlink;
 
 		ata_eh_freeze_port(ap);
 
 		/* reset */
-		ehc->i.action = ATA_EH_HARDRESET;
 		rc = ata_eh_reset(link, 0, prereset, softreset, hardreset,
 				  postreset);
 		if (rc) {
@@ -890,11 +880,11 @@ static int sata_pmp_eh_recover_pmp(struct ata_port *ap,
 				reval_failed = 1;
 
 			ata_dev_printk(dev, KERN_WARNING,
-				       "retrying hardreset%s\n",
+				       "retrying reset%s\n",
 				       sleep ? " in 5 secs" : "");
 			if (sleep)
 				ssleep(5);
-			ehc->i.action |= ATA_EH_HARDRESET;
+			ehc->i.action |= ATA_EH_RESET;
 			goto retry;
 		} else {
 			ata_dev_printk(dev, KERN_ERR, "failed to recover PMP "
@@ -938,10 +928,8 @@ static int sata_pmp_eh_handle_disabled_links(struct ata_port *ap)
 		/* Some PMPs require hardreset sequence to get
 		 * SError.N working.
 		 */
-		if ((link->flags & ATA_LFLAG_HRST_TO_RESUME) &&
-		    (link->eh_context.i.flags & ATA_EHI_RESUME_LINK))
-			sata_link_hardreset(link, sata_deb_timing_normal,
-					    jiffies + ATA_TMOUT_INTERNAL_QUICK);
+		sata_link_hardreset(link, sata_deb_timing_normal,
+				    jiffies + ATA_TMOUT_INTERNAL_QUICK);
 
 		/* unconditionally clear SError.N */
 		rc = sata_scr_write(link, SCR_ERROR, SERR_PHYRDY_CHG);
@@ -1124,7 +1112,7 @@ static int sata_pmp_eh_recover(struct ata_port *ap,
 
  link_fail:
 	if (sata_pmp_handle_link_fail(link, link_tries)) {
-		pmp_ehc->i.action |= ATA_EH_HARDRESET;
+		pmp_ehc->i.action |= ATA_EH_RESET;
 		goto retry;
 	}
 
@@ -1142,7 +1130,7 @@ static int sata_pmp_eh_recover(struct ata_port *ap,
 	if (--pmp_tries) {
 		ata_port_printk(ap, KERN_WARNING,
 				"failed to recover PMP, retrying in 5 secs\n");
-		pmp_ehc->i.action |= ATA_EH_HARDRESET;
+		pmp_ehc->i.action |= ATA_EH_RESET;
 		ssleep(5);
 		goto retry;
 	}
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 15795394b0a8..1f036a7b14f3 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -3508,7 +3508,7 @@ static int ata_scsi_user_scan(struct Scsi_Host *shost, unsigned int channel,
 		ata_port_for_each_link(link, ap) {
 			struct ata_eh_info *ehi = &link->eh_info;
 			ehi->probe_mask |= (1 << ata_link_max_devices(link)) - 1;
-			ehi->action |= ATA_EH_SOFTRESET;
+			ehi->action |= ATA_EH_RESET;
 		}
 	} else {
 		struct ata_device *dev = ata_find_dev(ap, devno);
@@ -3516,7 +3516,7 @@ static int ata_scsi_user_scan(struct Scsi_Host *shost, unsigned int channel,
 		if (dev) {
 			struct ata_eh_info *ehi = &dev->link->eh_info;
 			ehi->probe_mask |= 1 << dev->devno;
-			ehi->action |= ATA_EH_SOFTRESET;
+			ehi->action |= ATA_EH_RESET;
 			ehi->flags |= ATA_EHI_RESUME_LINK;
 		} else
 			rc = -EINVAL;
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index 9d1e3cad4aa9..fbd423ad23bb 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -996,7 +996,7 @@ static void sata_fsl_error_intr(struct ata_port *ap)
 	/* handle fatal errors */
 	if (hstatus & FATAL_ERROR_DECODE) {
 		err_mask |= AC_ERR_ATA_BUS;
-		action |= ATA_EH_SOFTRESET;
+		action |= ATA_EH_RESET;
 		/* how will fatal error interrupts be completed ?? */
 		freeze = 1;
 	}
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 6ebebde8454a..a4944c8ad46d 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -1524,14 +1524,14 @@ static void mv_err_intr(struct ata_port *ap, struct ata_queued_cmd *qc)
 			EDMA_ERR_CRQB_PAR | EDMA_ERR_CRPB_PAR |
 			EDMA_ERR_INTRL_PAR)) {
 		err_mask |= AC_ERR_ATA_BUS;
-		action |= ATA_EH_HARDRESET;
+		action |= ATA_EH_RESET;
 		ata_ehi_push_desc(ehi, "parity error");
 	}
 	if (edma_err_cause & (EDMA_ERR_DEV_DCON | EDMA_ERR_DEV_CON)) {
 		ata_ehi_hotplugged(ehi);
 		ata_ehi_push_desc(ehi, edma_err_cause & EDMA_ERR_DEV_DCON ?
 			"dev disconnect" : "dev connect");
-		action |= ATA_EH_HARDRESET;
+		action |= ATA_EH_RESET;
 	}
 
 	if (IS_GEN_I(hpriv)) {
@@ -1555,7 +1555,7 @@ static void mv_err_intr(struct ata_port *ap, struct ata_queued_cmd *qc)
 			sata_scr_read(&ap->link, SCR_ERROR, &serr);
 			sata_scr_write_flush(&ap->link, SCR_ERROR, serr);
 			err_mask = AC_ERR_ATA_BUS;
-			action |= ATA_EH_HARDRESET;
+			action |= ATA_EH_RESET;
 		}
 	}
 
@@ -1564,7 +1564,7 @@ static void mv_err_intr(struct ata_port *ap, struct ata_queued_cmd *qc)
 
 	if (!err_mask) {
 		err_mask = AC_ERR_OTHER;
-		action |= ATA_EH_HARDRESET;
+		action |= ATA_EH_RESET;
 	}
 
 	ehi->serror |= serr;
@@ -1780,7 +1780,7 @@ static void mv_pci_error(struct ata_host *host, void __iomem *mmio)
 				ata_ehi_push_desc(ehi,
 					"PCI err cause 0x%08x", err_cause);
 			err_mask = AC_ERR_HOST_BUS;
-			ehi->action = ATA_EH_HARDRESET;
+			ehi->action = ATA_EH_RESET;
 			qc = ata_qc_from_tag(ap, ap->link.active_tag);
 			if (qc)
 				qc->err_mask |= err_mask;
@@ -2449,28 +2449,13 @@ static int mv_prereset(struct ata_link *link, unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
 	struct mv_port_priv *pp	= ap->private_data;
-	struct ata_eh_context *ehc = &link->eh_context;
-	int rc;
 
-	rc = mv_stop_dma(ap);
-	if (rc)
-		ehc->i.action |= ATA_EH_HARDRESET;
+	mv_stop_dma(ap);
 
-	if (!(pp->pp_flags & MV_PP_FLAG_HAD_A_RESET)) {
+	if (!(pp->pp_flags & MV_PP_FLAG_HAD_A_RESET))
 		pp->pp_flags |= MV_PP_FLAG_HAD_A_RESET;
-		ehc->i.action |= ATA_EH_HARDRESET;
-	}
-
-	/* if we're about to do hardreset, nothing more to do */
-	if (ehc->i.action & ATA_EH_HARDRESET)
-		return 0;
-
-	if (ata_link_online(link))
-		rc = ata_wait_ready(ap, deadline);
-	else
-		rc = -ENODEV;
 
-	return rc;
+	return 0;
 }
 
 static int mv_hardreset(struct ata_link *link, unsigned int *class,
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index ed5473bf7a0a..ce02e15c857c 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -929,7 +929,7 @@ static int nv_adma_check_cpb(struct ata_port *ap, int cpb_num, int force_err)
 					"notifier for tag %d with no cmd?\n",
 					cpb_num);
 			ehi->err_mask |= AC_ERR_HSM;
-			ehi->action |= ATA_EH_SOFTRESET;
+			ehi->action |= ATA_EH_RESET;
 			ata_port_freeze(ap);
 			return 1;
 		}
@@ -1892,7 +1892,7 @@ static void nv_swncq_error_handler(struct ata_port *ap)
 
 	if (ap->link.sactive) {
 		nv_swncq_ncq_stop(ap);
-		ehc->i.action |= ATA_EH_HARDRESET;
+		ehc->i.action |= ATA_EH_RESET;
 	}
 
 	ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
@@ -2173,7 +2173,7 @@ static int nv_swncq_sdbfis(struct ata_port *ap)
 		ata_ehi_clear_desc(ehi);
 		ata_ehi_push_desc(ehi, "BMDMA stat 0x%x", host_stat);
 		ehi->err_mask |= AC_ERR_HOST_BUS;
-		ehi->action |= ATA_EH_SOFTRESET;
+		ehi->action |= ATA_EH_RESET;
 		return -EINVAL;
 	}
 
@@ -2188,7 +2188,7 @@ static int nv_swncq_sdbfis(struct ata_port *ap)
 		ata_ehi_push_desc(ehi, "illegal SWNCQ:qc_active transition"
 				  "(%08x->%08x)", pp->qc_active, sactive);
 		ehi->err_mask |= AC_ERR_HSM;
-		ehi->action |= ATA_EH_HARDRESET;
+		ehi->action |= ATA_EH_RESET;
 		return -EINVAL;
 	}
 	for (i = 0; i < ATA_MAX_QUEUE; i++) {
@@ -2324,7 +2324,7 @@ static void nv_swncq_host_interrupt(struct ata_port *ap, u16 fis)
 		ata_ehi_push_desc(ehi, "Ata error. fis:0x%X", fis);
 		ehi->err_mask |= AC_ERR_DEV;
 		ehi->serror |= serror;
-		ehi->action |= ATA_EH_SOFTRESET;
+		ehi->action |= ATA_EH_RESET;
 		ata_port_freeze(ap);
 		return;
 	}
@@ -2356,7 +2356,7 @@ static void nv_swncq_host_interrupt(struct ata_port *ap, u16 fis)
 		if (pp->ncq_flags & (ncq_saw_sdb | ncq_saw_backout)) {
 			ata_ehi_push_desc(ehi, "illegal fis transaction");
 			ehi->err_mask |= AC_ERR_HSM;
-			ehi->action |= ATA_EH_HARDRESET;
+			ehi->action |= ATA_EH_RESET;
 			goto irq_error;
 		}
 
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index df7988df7908..aa8d0323c9bb 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -286,45 +286,45 @@ static struct sil24_cerr_info {
 				    "device error via D2H FIS" },
 	[PORT_CERR_SDB]		= { AC_ERR_DEV, 0,
 				    "device error via SDB FIS" },
-	[PORT_CERR_DATA]	= { AC_ERR_ATA_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_DATA]	= { AC_ERR_ATA_BUS, ATA_EH_RESET,
 				    "error in data FIS" },
-	[PORT_CERR_SEND]	= { AC_ERR_ATA_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_SEND]	= { AC_ERR_ATA_BUS, ATA_EH_RESET,
 				    "failed to transmit command FIS" },
-	[PORT_CERR_INCONSISTENT] = { AC_ERR_HSM, ATA_EH_SOFTRESET,
+	[PORT_CERR_INCONSISTENT] = { AC_ERR_HSM, ATA_EH_RESET,
 				     "protocol mismatch" },
-	[PORT_CERR_DIRECTION]	= { AC_ERR_HSM, ATA_EH_SOFTRESET,
+	[PORT_CERR_DIRECTION]	= { AC_ERR_HSM, ATA_EH_RESET,
 				    "data directon mismatch" },
-	[PORT_CERR_UNDERRUN]	= { AC_ERR_HSM, ATA_EH_SOFTRESET,
+	[PORT_CERR_UNDERRUN]	= { AC_ERR_HSM, ATA_EH_RESET,
 				    "ran out of SGEs while writing" },
-	[PORT_CERR_OVERRUN]	= { AC_ERR_HSM, ATA_EH_SOFTRESET,
+	[PORT_CERR_OVERRUN]	= { AC_ERR_HSM, ATA_EH_RESET,
 				    "ran out of SGEs while reading" },
-	[PORT_CERR_PKT_PROT]	= { AC_ERR_HSM, ATA_EH_SOFTRESET,
+	[PORT_CERR_PKT_PROT]	= { AC_ERR_HSM, ATA_EH_RESET,
 				    "invalid data directon for ATAPI CDB" },
-	[PORT_CERR_SGT_BOUNDARY] = { AC_ERR_SYSTEM, ATA_EH_SOFTRESET,
+	[PORT_CERR_SGT_BOUNDARY] = { AC_ERR_SYSTEM, ATA_EH_RESET,
 				     "SGT not on qword boundary" },
-	[PORT_CERR_SGT_TGTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_SGT_TGTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI target abort while fetching SGT" },
-	[PORT_CERR_SGT_MSTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_SGT_MSTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI master abort while fetching SGT" },
-	[PORT_CERR_SGT_PCIPERR]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_SGT_PCIPERR]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI parity error while fetching SGT" },
-	[PORT_CERR_CMD_BOUNDARY] = { AC_ERR_SYSTEM, ATA_EH_SOFTRESET,
+	[PORT_CERR_CMD_BOUNDARY] = { AC_ERR_SYSTEM, ATA_EH_RESET,
 				     "PRB not on qword boundary" },
-	[PORT_CERR_CMD_TGTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_CMD_TGTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI target abort while fetching PRB" },
-	[PORT_CERR_CMD_MSTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_CMD_MSTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI master abort while fetching PRB" },
-	[PORT_CERR_CMD_PCIPERR]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_CMD_PCIPERR]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI parity error while fetching PRB" },
-	[PORT_CERR_XFR_UNDEF]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_XFR_UNDEF]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "undefined error while transferring data" },
-	[PORT_CERR_XFR_TGTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_XFR_TGTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI target abort while transferring data" },
-	[PORT_CERR_XFR_MSTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_XFR_MSTABRT]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI master abort while transferring data" },
-	[PORT_CERR_XFR_PCIPERR]	= { AC_ERR_HOST_BUS, ATA_EH_SOFTRESET,
+	[PORT_CERR_XFR_PCIPERR]	= { AC_ERR_HOST_BUS, ATA_EH_RESET,
 				    "PCI parity error while transferring data" },
-	[PORT_CERR_SENDSERVICE]	= { AC_ERR_HSM, ATA_EH_SOFTRESET,
+	[PORT_CERR_SENDSERVICE]	= { AC_ERR_HSM, ATA_EH_RESET,
 				    "FIS received while sending service FIS" },
 };
 
@@ -616,7 +616,7 @@ static int sil24_init_port(struct ata_port *ap)
 
 	if ((tmp & (PORT_CS_INIT | PORT_CS_RDY)) != PORT_CS_RDY) {
 		pp->do_port_rst = 1;
-		ap->link.eh_context.i.action |= ATA_EH_HARDRESET;
+		ap->link.eh_context.i.action |= ATA_EH_RESET;
 		return -EIO;
 	}
 
@@ -1022,7 +1022,7 @@ static void sil24_error_intr(struct ata_port *ap)
 
 	if (irq_stat & PORT_IRQ_UNK_FIS) {
 		ehi->err_mask |= AC_ERR_HSM;
-		ehi->action |= ATA_EH_SOFTRESET;
+		ehi->action |= ATA_EH_RESET;
 		ata_ehi_push_desc(ehi, "unknown FIS");
 		freeze = 1;
 	}
@@ -1043,7 +1043,7 @@ static void sil24_error_intr(struct ata_port *ap)
 		 */
 		if (ap->nr_active_links >= 3) {
 			ehi->err_mask |= AC_ERR_OTHER;
-			ehi->action |= ATA_EH_HARDRESET;
+			ehi->action |= ATA_EH_RESET;
 			ata_ehi_push_desc(ehi, "PMP DMA CS errata");
 			pp->do_port_rst = 1;
 			freeze = 1;
@@ -1064,7 +1064,7 @@ static void sil24_error_intr(struct ata_port *ap)
 						  irq_stat);
 			} else {
 				err_mask |= AC_ERR_HSM;
-				action |= ATA_EH_HARDRESET;
+				action |= ATA_EH_RESET;
 				freeze = 1;
 			}
 		} else
@@ -1078,12 +1078,12 @@ static void sil24_error_intr(struct ata_port *ap)
 		if (ci && ci->desc) {
 			err_mask |= ci->err_mask;
 			action |= ci->action;
-			if (action & ATA_EH_RESET_MASK)
+			if (action & ATA_EH_RESET)
 				freeze = 1;
 			ata_ehi_push_desc(ehi, "%s", ci->desc);
 		} else {
 			err_mask |= AC_ERR_OTHER;
-			action |= ATA_EH_SOFTRESET;
+			action |= ATA_EH_RESET;
 			freeze = 1;
 			ata_ehi_push_desc(ehi, "unknown command error %d",
 					  cerr);
@@ -1153,7 +1153,7 @@ static inline void sil24_host_intr(struct ata_port *ap)
 	if (rc < 0) {
 		struct ata_eh_info *ehi = &ap->link.eh_info;
 		ehi->err_mask |= AC_ERR_HSM;
-		ehi->action |= ATA_EH_SOFTRESET;
+		ehi->action |= ATA_EH_RESET;
 		ata_port_freeze(ap);
 		return;
 	}
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index 0d03f44824fb..c0e0f1d18d50 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -320,7 +320,7 @@ static int vt6420_prereset(struct ata_link *link, unsigned long deadline)
 
 	if (!online) {
 		/* tell EH to bail */
-		ehc->i.action &= ~ATA_EH_RESET_MASK;
+		ehc->i.action &= ~ATA_EH_RESET;
 		return 0;
 	}
 
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 37ee881c42ac..c63cfb3b222b 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -292,12 +292,12 @@ enum {
 
 	/* reset / recovery action types */
 	ATA_EH_REVALIDATE	= (1 << 0),
-	ATA_EH_SOFTRESET	= (1 << 1),
-	ATA_EH_HARDRESET	= (1 << 2),
+	ATA_EH_SOFTRESET	= (1 << 1), /* meaningful only in ->prereset */
+	ATA_EH_HARDRESET	= (1 << 2), /* meaningful only in ->prereset */
+	ATA_EH_RESET		= ATA_EH_SOFTRESET | ATA_EH_HARDRESET,
 	ATA_EH_ENABLE_LINK	= (1 << 3),
 	ATA_EH_LPM		= (1 << 4),  /* link power management action */
 
-	ATA_EH_RESET_MASK	= ATA_EH_SOFTRESET | ATA_EH_HARDRESET,
 	ATA_EH_PERDEV_MASK	= ATA_EH_REVALIDATE,
 
 	/* ata_eh_info->flags */
@@ -1097,7 +1097,7 @@ extern void ata_ehi_clear_desc(struct ata_eh_info *ehi);
 static inline void ata_ehi_schedule_probe(struct ata_eh_info *ehi)
 {
 	ehi->flags |= ATA_EHI_RESUME_LINK;
-	ehi->action |= ATA_EH_SOFTRESET;
+	ehi->action |= ATA_EH_RESET;
 	ehi->probe_mask |= (1 << ATA_MAX_DEVICES) - 1;
 }
 
-- 
cgit v1.2.3


From d692abd92f22a81b38d52c39601871003262841c Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 24 Jan 2008 00:05:14 +0900
Subject: libata: kill ATA_LFLAG_HRST_TO_RESUME

Now that hardreset is the preferred method of resetting, there's no
need for ATA_LFLAG_HRST_TO_RESUME flag.  Kill it.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c       |  2 +-
 drivers/ata/libata-pmp.c | 16 ----------------
 drivers/ata/sata_nv.c    |  5 -----
 drivers/ata/sata_sil.c   |  5 -----
 include/linux/libata.h   |  1 -
 5 files changed, 1 insertion(+), 28 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index f6bbd52b1547..66d6c8821087 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -416,7 +416,7 @@ static const struct ata_port_info ahci_port_info[] = {
 	{
 		AHCI_HFLAGS	(AHCI_HFLAG_NO_NCQ | AHCI_HFLAG_NO_PMP),
 		.flags		= AHCI_FLAG_COMMON,
-		.link_flags	= AHCI_LFLAG_COMMON | ATA_LFLAG_HRST_TO_RESUME,
+		.link_flags	= AHCI_LFLAG_COMMON,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_vt8251_ops,
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index 8439fc8efdd6..7f8bcffa81ad 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -453,9 +453,6 @@ static void sata_pmp_quirks(struct ata_port *ap)
 	if (vendor == 0x1095 && devid == 0x3726) {
 		/* sil3726 quirks */
 		ata_port_for_each_link(link, ap) {
-			/* SError.N need a kick in the ass to get working */
-			link->flags |= ATA_LFLAG_HRST_TO_RESUME;
-
 			/* class code report is unreliable */
 			if (link->pmp < 5)
 				link->flags |= ATA_LFLAG_ASSUME_ATA;
@@ -468,9 +465,6 @@ static void sata_pmp_quirks(struct ata_port *ap)
 	} else if (vendor == 0x1095 && devid == 0x4723) {
 		/* sil4723 quirks */
 		ata_port_for_each_link(link, ap) {
-			/* SError.N need a kick in the ass to get working */
-			link->flags |= ATA_LFLAG_HRST_TO_RESUME;
-
 			/* class code report is unreliable */
 			if (link->pmp < 2)
 				link->flags |= ATA_LFLAG_ASSUME_ATA;
@@ -483,9 +477,6 @@ static void sata_pmp_quirks(struct ata_port *ap)
 	} else if (vendor == 0x1095 && devid == 0x4726) {
 		/* sil4726 quirks */
 		ata_port_for_each_link(link, ap) {
-			/* SError.N need a kick in the ass to get working */
-			link->flags |= ATA_LFLAG_HRST_TO_RESUME;
-
 			/* Class code report is unreliable and SRST
 			 * times out under certain configurations.
 			 * Config device can be at port 0 or 5 and
@@ -513,13 +504,6 @@ static void sata_pmp_quirks(struct ata_port *ap)
 		 * otherwise.  Don't try hard to recover it.
 		 */
 		ap->pmp_link[ap->nr_pmp_links - 1].flags |= ATA_LFLAG_NO_RETRY;
-	} else if (vendor == 0x11ab && devid == 0x4140) {
-		/* Marvell 88SM4140 quirks.  Fan-out ports require PHY
-		 * reset to work; other than that, it behaves very
-		 * nicely.
-		 */
-		ata_port_for_each_link(link, ap)
-			link->flags |= ATA_LFLAG_HRST_TO_RESUME;
 	}
 }
 
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index ce02e15c857c..75b76535c720 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -579,7 +579,6 @@ static const struct ata_port_info nv_port_info[] = {
 	{
 		.sht		= &nv_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
-		.link_flags	= ATA_LFLAG_HRST_TO_RESUME,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
@@ -590,7 +589,6 @@ static const struct ata_port_info nv_port_info[] = {
 	{
 		.sht		= &nv_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
-		.link_flags	= ATA_LFLAG_HRST_TO_RESUME,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
@@ -601,7 +599,6 @@ static const struct ata_port_info nv_port_info[] = {
 	{
 		.sht		= &nv_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
-		.link_flags	= ATA_LFLAG_HRST_TO_RESUME,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
@@ -613,7 +610,6 @@ static const struct ata_port_info nv_port_info[] = {
 		.sht		= &nv_adma_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_MMIO | ATA_FLAG_NCQ,
-		.link_flags	= ATA_LFLAG_HRST_TO_RESUME,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
@@ -625,7 +621,6 @@ static const struct ata_port_info nv_port_info[] = {
 		.sht		= &nv_swncq_sht,
 		.flags	        = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_NCQ,
-		.link_flags	= ATA_LFLAG_HRST_TO_RESUME,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index 0b8191b52f97..7052915a31b6 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -60,7 +60,6 @@ enum {
 
 	SIL_DFL_PORT_FLAGS	= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_MMIO,
-	SIL_DFL_LINK_FLAGS	= ATA_LFLAG_HRST_TO_RESUME,
 
 	/*
 	 * Controller IDs
@@ -215,7 +214,6 @@ static const struct ata_port_info sil_port_info[] = {
 	/* sil_3112 */
 	{
 		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_MOD15WRITE,
-		.link_flags	= SIL_DFL_LINK_FLAGS,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= ATA_UDMA5,
@@ -225,7 +223,6 @@ static const struct ata_port_info sil_port_info[] = {
 	{
 		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_MOD15WRITE |
 				  SIL_FLAG_NO_SATA_IRQ,
-		.link_flags	= SIL_DFL_LINK_FLAGS,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= ATA_UDMA5,
@@ -234,7 +231,6 @@ static const struct ata_port_info sil_port_info[] = {
 	/* sil_3512 */
 	{
 		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_RERR_ON_DMA_ACT,
-		.link_flags	= SIL_DFL_LINK_FLAGS,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= ATA_UDMA5,
@@ -243,7 +239,6 @@ static const struct ata_port_info sil_port_info[] = {
 	/* sil_3114 */
 	{
 		.flags		= SIL_DFL_PORT_FLAGS | SIL_FLAG_RERR_ON_DMA_ACT,
-		.link_flags	= SIL_DFL_LINK_FLAGS,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= ATA_UDMA5,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index c63cfb3b222b..1524af6f018b 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -163,7 +163,6 @@ enum {
 	ATA_DEV_NONE		= 9,	/* no device */
 
 	/* struct ata_link flags */
-	ATA_LFLAG_HRST_TO_RESUME = (1 << 0), /* hardreset to resume link */
 	ATA_LFLAG_SKIP_D2H_BSY	= (1 << 1), /* can't wait for the first D2H
 					     * Register FIS clearing BSY */
 	ATA_LFLAG_NO_SRST	= (1 << 2), /* avoid softreset */
-- 
cgit v1.2.3


From 672b2d65ba83a6f3f801fd3d58851aa9c0725a54 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 24 Jan 2008 00:05:14 +0900
Subject: libata: kill ATA_EHI_RESUME_LINK

ATA_EHI_RESUME_LINK has two functions - promote reset to hardreset if
ATA_LFLAG_HRST_TO_RESUME is set and preventing EH from shortcutting
reset action when probing is requested.  The former is gone now and
the latter can easily be achieved by making EH to perform at least one
reset if reset is requested, which also makes more sense than
depending on RESUME_LINK flag.

As ATA_EHI_RESUME_LINK was the only EHI reset modifier, this also
kills reset modifier handling.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-eh.c   | 21 ++++++++-------------
 drivers/ata/libata-pmp.c  |  1 -
 drivers/ata/libata-scsi.c |  1 -
 include/linux/libata.h    |  3 ---
 4 files changed, 8 insertions(+), 18 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index f7cae6400155..e6584fa7f456 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -1079,12 +1079,6 @@ void ata_eh_about_to_do(struct ata_link *link, struct ata_device *dev,
 
 	spin_lock_irqsave(ap->lock, flags);
 
-	/* suck in and clear reset modifier */
-	if (action & ATA_EH_RESET) {
-		ehc->i.flags |= ehi->flags & ATA_EHI_RESET_MODIFIER_MASK;
-		ehi->flags &= ~ATA_EHI_RESET_MODIFIER_MASK;
-	}
-
 	ata_eh_clear_action(link, dev, ehi, action);
 
 	if (!(ehc->i.flags & ATA_EHI_QUIET))
@@ -1110,10 +1104,6 @@ void ata_eh_done(struct ata_link *link, struct ata_device *dev,
 {
 	struct ata_eh_context *ehc = &link->eh_context;
 
-	/* if reset is complete, clear reset modifier */
-	if (action & ATA_EH_RESET)
-		ehc->i.flags &= ~ATA_EHI_RESET_MODIFIER_MASK;
-
 	ata_eh_clear_action(link, dev, &ehc->i, action);
 }
 
@@ -2491,6 +2481,7 @@ static int ata_link_nr_vacant(struct ata_link *link)
 
 static int ata_eh_skip_recovery(struct ata_link *link)
 {
+	struct ata_port *ap = link->ap;
 	struct ata_eh_context *ehc = &link->eh_context;
 	struct ata_device *dev;
 
@@ -2498,9 +2489,13 @@ static int ata_eh_skip_recovery(struct ata_link *link)
 	if (link->flags & ATA_LFLAG_DISABLED)
 		return 1;
 
-	/* thaw frozen port, resume link and recover failed devices */
-	if ((link->ap->pflags & ATA_PFLAG_FROZEN) ||
-	    (ehc->i.flags & ATA_EHI_RESUME_LINK) || ata_link_nr_enabled(link))
+	/* thaw frozen port and recover failed devices */
+	if ((ap->pflags & ATA_PFLAG_FROZEN) || ata_link_nr_enabled(link))
+		return 0;
+
+	/* reset at least once if reset is requested */
+	if ((ehc->i.action & ATA_EH_RESET) &&
+	    !(ehc->i.flags & ATA_EHI_DID_RESET))
 		return 0;
 
 	/* skip if class codes for all vacant slots are ATA_DEV_NONE */
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index 7f8bcffa81ad..df1d3252b9e6 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -437,7 +437,6 @@ static int sata_pmp_init_links(struct ata_port *ap, int nr_ports)
 		link->flags = 0;
 		ehc->i.probe_mask |= 1;
 		ehc->i.action |= ATA_EH_RESET;
-		ehc->i.flags |= ATA_EHI_RESUME_LINK;
 	}
 
 	return 0;
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 1f036a7b14f3..caffca7dd76f 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -3517,7 +3517,6 @@ static int ata_scsi_user_scan(struct Scsi_Host *shost, unsigned int channel,
 			struct ata_eh_info *ehi = &dev->link->eh_info;
 			ehi->probe_mask |= 1 << dev->devno;
 			ehi->action |= ATA_EH_RESET;
-			ehi->flags |= ATA_EHI_RESUME_LINK;
 		} else
 			rc = -EINVAL;
 	}
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 1524af6f018b..4093e3b6a8b7 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -301,7 +301,6 @@ enum {
 
 	/* ata_eh_info->flags */
 	ATA_EHI_HOTPLUGGED	= (1 << 0),  /* could have been hotplugged */
-	ATA_EHI_RESUME_LINK	= (1 << 1),  /* resume link (reset modifier) */
 	ATA_EHI_NO_AUTOPSY	= (1 << 2),  /* no autopsy */
 	ATA_EHI_QUIET		= (1 << 3),  /* be quiet */
 
@@ -312,7 +311,6 @@ enum {
 	ATA_EHI_POST_SETMODE	= (1 << 20), /* revaildating after setmode */
 
 	ATA_EHI_DID_RESET	= ATA_EHI_DID_SOFTRESET | ATA_EHI_DID_HARDRESET,
-	ATA_EHI_RESET_MODIFIER_MASK = ATA_EHI_RESUME_LINK,
 
 	/* max tries if error condition is still set after ->error_handler */
 	ATA_EH_MAX_TRIES	= 5,
@@ -1095,7 +1093,6 @@ extern void ata_ehi_clear_desc(struct ata_eh_info *ehi);
 
 static inline void ata_ehi_schedule_probe(struct ata_eh_info *ehi)
 {
-	ehi->flags |= ATA_EHI_RESUME_LINK;
 	ehi->action |= ATA_EH_RESET;
 	ehi->probe_mask |= (1 << ATA_MAX_DEVICES) - 1;
 }
-- 
cgit v1.2.3


From 8cebf274dd1c955a6e03385a85fd6569ce445946 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 24 Jan 2008 00:05:14 +0900
Subject: libata: kill ATA_LFLAG_SKIP_D2H_BSY

Some controllers can't reliably record the initial D2H FIS after SATA
link is brought online for whatever reason.  Advanced controllers
which don't have traditional TF register based interface often have
this problem as they don't really have the TF registers to update
while the controller and link are being initialized.

SKIP_D2H_BSY works around the problem by skipping the wait for device
readiness before issuing SRST, so for such controllers libata issues
SRST blindly and hopes for the best.

Now that libata defaults to hardreset, this workaround is no longer
necessary.  For controllers which have support for hardreset, SRST is
never issued by itself.  It is only issued as follow-up SRST for
device classification and PMP initialization, so there's no need to
wait for it from prereset.

Kill ATA_LFLAG_SKIP_D2H_BSY.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c        | 7 -------
 drivers/ata/libata-core.c | 6 ++----
 drivers/ata/sata_fsl.c    | 2 --
 drivers/ata/sata_sil24.c  | 4 ----
 include/linux/libata.h    | 2 --
 5 files changed, 2 insertions(+), 19 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 66d6c8821087..1bd258e5390f 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -198,7 +198,6 @@ enum {
 					  ATA_FLAG_MMIO | ATA_FLAG_PIO_DMA |
 					  ATA_FLAG_ACPI_SATA | ATA_FLAG_AN |
 					  ATA_FLAG_IPM,
-	AHCI_LFLAG_COMMON		= ATA_LFLAG_SKIP_D2H_BSY,
 
 	ICH_MAP				= 0x90, /* ICH MAP register */
 };
@@ -407,7 +406,6 @@ static const struct ata_port_info ahci_port_info[] = {
 	/* board_ahci */
 	{
 		.flags		= AHCI_FLAG_COMMON,
-		.link_flags	= AHCI_LFLAG_COMMON,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_ops,
@@ -416,7 +414,6 @@ static const struct ata_port_info ahci_port_info[] = {
 	{
 		AHCI_HFLAGS	(AHCI_HFLAG_NO_NCQ | AHCI_HFLAG_NO_PMP),
 		.flags		= AHCI_FLAG_COMMON,
-		.link_flags	= AHCI_LFLAG_COMMON,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_vt8251_ops,
@@ -425,7 +422,6 @@ static const struct ata_port_info ahci_port_info[] = {
 	{
 		AHCI_HFLAGS	(AHCI_HFLAG_IGN_IRQ_IF_ERR),
 		.flags		= AHCI_FLAG_COMMON,
-		.link_flags	= AHCI_LFLAG_COMMON,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_ops,
@@ -436,7 +432,6 @@ static const struct ata_port_info ahci_port_info[] = {
 				 AHCI_HFLAG_32BIT_ONLY |
 				 AHCI_HFLAG_SECT255 | AHCI_HFLAG_NO_PMP),
 		.flags		= AHCI_FLAG_COMMON,
-		.link_flags	= AHCI_LFLAG_COMMON,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_ops,
@@ -447,7 +442,6 @@ static const struct ata_port_info ahci_port_info[] = {
 				 AHCI_HFLAG_MV_PATA),
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_MMIO | ATA_FLAG_PIO_DMA,
-		.link_flags	= AHCI_LFLAG_COMMON,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_ops,
@@ -457,7 +451,6 @@ static const struct ata_port_info ahci_port_info[] = {
 		AHCI_HFLAGS	(AHCI_HFLAG_IGN_SERR_INTERNAL |
 				 AHCI_HFLAG_NO_PMP),
 		.flags		= AHCI_FLAG_COMMON,
-		.link_flags	= AHCI_LFLAG_COMMON,
 		.pio_mask	= 0x1f, /* pio0-4 */
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_ops,
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 02e7ba43a3b3..baef749ec528 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -3962,10 +3962,8 @@ int ata_std_prereset(struct ata_link *link, unsigned long deadline)
 					"link for reset (errno=%d)\n", rc);
 	}
 
-	/* Wait for !BSY if the controller can wait for the first D2H
-	 * Reg FIS and we don't know that no device is attached.
-	 */
-	if (!(link->flags & ATA_LFLAG_SKIP_D2H_BSY) && !ata_link_offline(link)) {
+	/* wait for !BSY if we don't know that no device is attached */
+	if (!ata_link_offline(link)) {
 		rc = ata_wait_ready(ap, deadline);
 		if (rc && rc != -ENODEV) {
 			ata_link_printk(link, KERN_WARNING, "device not ready "
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index fbd423ad23bb..d23b690ed1a1 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -35,7 +35,6 @@ enum {
 	SATA_FSL_HOST_FLAGS	= (ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				ATA_FLAG_MMIO | ATA_FLAG_PIO_DMA |
 				ATA_FLAG_NCQ),
-	SATA_FSL_HOST_LFLAGS	= ATA_LFLAG_SKIP_D2H_BSY,
 
 	SATA_FSL_MAX_CMDS	= SATA_FSL_QUEUE_DEPTH,
 	SATA_FSL_CMD_HDR_SIZE	= 16,	/* 4 DWORDS */
@@ -1241,7 +1240,6 @@ static const struct ata_port_operations sata_fsl_ops = {
 static const struct ata_port_info sata_fsl_port_info[] = {
 	{
 	 .flags = SATA_FSL_HOST_FLAGS,
-	 .link_flags = SATA_FSL_HOST_LFLAGS,
 	 .pio_mask = 0x1f,	/* pio 0-4 */
 	 .udma_mask = 0x7f,	/* udma 0-6 */
 	 .port_ops = &sata_fsl_ops,
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index aa8d0323c9bb..ba0c00e8ee7f 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -254,7 +254,6 @@ enum {
 				  ATA_FLAG_MMIO | ATA_FLAG_PIO_DMA |
 				  ATA_FLAG_NCQ | ATA_FLAG_ACPI_SATA |
 				  ATA_FLAG_AN | ATA_FLAG_PMP,
-	SIL24_COMMON_LFLAGS	= ATA_LFLAG_SKIP_D2H_BSY,
 	SIL24_FLAG_PCIX_IRQ_WOC	= (1 << 24), /* IRQ loss errata on PCI-X */
 
 	IRQ_STAT_4PORTS		= 0xf,
@@ -449,7 +448,6 @@ static const struct ata_port_info sil24_port_info[] = {
 	{
 		.flags		= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(4) |
 				  SIL24_FLAG_PCIX_IRQ_WOC,
-		.link_flags	= SIL24_COMMON_LFLAGS,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= ATA_UDMA5,		/* udma0-5 */
@@ -458,7 +456,6 @@ static const struct ata_port_info sil24_port_info[] = {
 	/* sil_3132 */
 	{
 		.flags		= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(2),
-		.link_flags	= SIL24_COMMON_LFLAGS,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= ATA_UDMA5,		/* udma0-5 */
@@ -467,7 +464,6 @@ static const struct ata_port_info sil24_port_info[] = {
 	/* sil_3131/sil_3531 */
 	{
 		.flags		= SIL24_COMMON_FLAGS | SIL24_NPORTS2FLAG(1),
-		.link_flags	= SIL24_COMMON_LFLAGS,
 		.pio_mask	= 0x1f,			/* pio0-4 */
 		.mwdma_mask	= 0x07,			/* mwdma0-2 */
 		.udma_mask	= ATA_UDMA5,		/* udma0-5 */
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 4093e3b6a8b7..6eec11957e54 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -163,8 +163,6 @@ enum {
 	ATA_DEV_NONE		= 9,	/* no device */
 
 	/* struct ata_link flags */
-	ATA_LFLAG_SKIP_D2H_BSY	= (1 << 1), /* can't wait for the first D2H
-					     * Register FIS clearing BSY */
 	ATA_LFLAG_NO_SRST	= (1 << 2), /* avoid softreset */
 	ATA_LFLAG_ASSUME_ATA	= (1 << 3), /* assume ATA class */
 	ATA_LFLAG_ASSUME_SEMB	= (1 << 4), /* assume SEMB class */
-- 
cgit v1.2.3


From b558edddb1c42c70a30cfe494984d4be409f7b2b Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 24 Jan 2008 00:05:14 +0900
Subject: libata: kill ata_ehi_schedule_probe()

ata_ehi_schedule_probe() was created to hide details of link-resuming
reset magic.  Now that all the softreset workarounds are gone,
scheduling probe is very simple - set probe_mask and request RESET.
Kill ata_ehi_schedule_probe() and open code it.  This also increases
consistency as ata_ehi_schedule_probe() couldn't cover individual
device probings so they were open-coded even when the helper existed.

While at it, define ATA_ALL_DEVICES as mask of all possible devices on
a link and always use it when requesting probe on link level for
simplicity and consistency.  Setting extra bits in the probe_mask
doesn't hurt anybody.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c |  3 +--
 drivers/ata/libata-eh.c   |  2 +-
 drivers/ata/libata-pmp.c  | 12 ++++++++----
 drivers/ata/libata-scsi.c |  2 +-
 include/linux/libata.h    | 12 ++++--------
 5 files changed, 15 insertions(+), 16 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index baef749ec528..4dc429fd0056 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -7156,8 +7156,7 @@ int ata_host_register(struct ata_host *host, struct scsi_host_template *sht)
 			/* kick EH for boot probing */
 			spin_lock_irqsave(ap->lock, flags);
 
-			ehi->probe_mask =
-				(1 << ata_link_max_devices(&ap->link)) - 1;
+			ehi->probe_mask |= ATA_ALL_DEVICES;
 			ehi->action |= ATA_EH_RESET;
 			ehi->flags |= ATA_EHI_NO_AUTOPSY | ATA_EHI_QUIET;
 
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index e6584fa7f456..0d0a2c0ab9e7 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -1746,7 +1746,7 @@ static void ata_eh_link_autopsy(struct ata_link *link)
 		ata_eh_analyze_serror(link);
 	} else if (rc != -EOPNOTSUPP) {
 		/* SError read failed, force reset and probing */
-		ata_ehi_schedule_probe(&ehc->i);
+		ehc->i.probe_mask |= ATA_ALL_DEVICES;
 		ehc->i.action |= ATA_EH_RESET;
 		ehc->i.err_mask |= AC_ERR_OTHER;
 	}
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index df1d3252b9e6..39e036c8a2bc 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -435,7 +435,7 @@ static int sata_pmp_init_links(struct ata_port *ap, int nr_ports)
 		struct ata_eh_context *ehc = &link->eh_context;
 
 		link->flags = 0;
-		ehc->i.probe_mask |= 1;
+		ehc->i.probe_mask |= ATA_ALL_DEVICES;
 		ehc->i.action |= ATA_EH_RESET;
 	}
 
@@ -831,8 +831,12 @@ static int sata_pmp_eh_recover_pmp(struct ata_port *ap,
 		ata_eh_thaw_port(ap);
 
 		/* PMP is reset, SErrors cannot be trusted, scan all */
-		ata_port_for_each_link(tlink, ap)
-			ata_ehi_schedule_probe(&tlink->eh_context.i);
+		ata_port_for_each_link(tlink, ap) {
+			struct ata_eh_context *ehc = &tlink->eh_context;
+
+			ehc->i.probe_mask |= ATA_ALL_DEVICES;
+			ehc->i.action |= ATA_EH_RESET;
+		}
 	}
 
 	/* If revalidation is requested, revalidate and reconfigure;
@@ -847,7 +851,7 @@ static int sata_pmp_eh_recover_pmp(struct ata_port *ap,
 		tries--;
 
 		if (rc == -ENODEV) {
-			ehc->i.probe_mask |= 1;
+			ehc->i.probe_mask |= ATA_ALL_DEVICES;
 			detach = 1;
 			/* give it just two more chances */
 			tries = min(tries, 2);
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index caffca7dd76f..798ba5e45710 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -3507,7 +3507,7 @@ static int ata_scsi_user_scan(struct Scsi_Host *shost, unsigned int channel,
 
 		ata_port_for_each_link(link, ap) {
 			struct ata_eh_info *ehi = &link->eh_info;
-			ehi->probe_mask |= (1 << ata_link_max_devices(link)) - 1;
+			ehi->probe_mask |= ATA_ALL_DEVICES;
 			ehi->action |= ATA_EH_RESET;
 		}
 	} else {
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 6eec11957e54..bc60132c7d33 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -122,6 +122,8 @@ enum {
 
 	ATAPI_MAX_DRAIN		= 16 << 10,
 
+	ATA_ALL_DEVICES		= (1 << ATA_MAX_DEVICES) - 1,
+
 	ATA_SHT_EMULATED	= 1,
 	ATA_SHT_CMD_PER_LUN	= 1,
 	ATA_SHT_THIS_ID		= -1,
@@ -1089,17 +1091,11 @@ extern void ata_ehi_push_desc(struct ata_eh_info *ehi, const char *fmt, ...)
 	__attribute__ ((format (printf, 2, 3)));
 extern void ata_ehi_clear_desc(struct ata_eh_info *ehi);
 
-static inline void ata_ehi_schedule_probe(struct ata_eh_info *ehi)
-{
-	ehi->action |= ATA_EH_RESET;
-	ehi->probe_mask |= (1 << ATA_MAX_DEVICES) - 1;
-}
-
 static inline void ata_ehi_hotplugged(struct ata_eh_info *ehi)
 {
-	ata_ehi_schedule_probe(ehi);
+	ehi->probe_mask |= (1 << ATA_MAX_DEVICES) - 1;
 	ehi->flags |= ATA_EHI_HOTPLUGGED;
-	ehi->action |= ATA_EH_ENABLE_LINK;
+	ehi->action |= ATA_EH_RESET | ATA_EH_ENABLE_LINK;
 	ehi->err_mask |= AC_ERR_ATA_BUS;
 }
 
-- 
cgit v1.2.3


From c1bc899f5805771926c9198e2ab4d77122c356a1 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:47 +0900
Subject: libata: reorganize ata_port_operations

Over the time, ops in ata_port_operations has become a bit confusing.
Reorganize.  SFF/BMDMA ops are separated into separate a group as they
will be taken out of ata_port_operations later.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 include/linux/libata.h | 117 ++++++++++++++++++++++++++-----------------------
 1 file changed, 61 insertions(+), 56 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/libata.h b/include/linux/libata.h
index bc60132c7d33..9476a479691b 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -665,69 +665,74 @@ struct ata_port {
 };
 
 struct ata_port_operations {
-	void (*dev_config) (struct ata_device *);
+	/*
+	 * Command execution
+	 */
+	int  (*qc_defer)(struct ata_queued_cmd *qc);
+	int  (*check_atapi_dma)(struct ata_queued_cmd *qc);
+	void (*qc_prep)(struct ata_queued_cmd *qc);
+	unsigned int (*qc_issue)(struct ata_queued_cmd *qc);
 
-	void (*set_piomode) (struct ata_port *, struct ata_device *);
-	void (*set_dmamode) (struct ata_port *, struct ata_device *);
-	unsigned long (*mode_filter) (struct ata_device *, unsigned long);
+	/*
+	 * Configuration and exception handling
+	 */
+	int  (*cable_detect)(struct ata_port *ap);
+	unsigned long (*mode_filter)(struct ata_device *dev, unsigned long xfer_mask);
+	void (*set_piomode)(struct ata_port *ap, struct ata_device *dev);
+	void (*set_dmamode)(struct ata_port *ap, struct ata_device *dev);
+	int  (*set_mode)(struct ata_link *link, struct ata_device **r_failed_dev);
 
-	void (*tf_load) (struct ata_port *ap, const struct ata_taskfile *tf);
-	void (*tf_read) (struct ata_port *ap, struct ata_taskfile *tf);
+	void (*dev_config)(struct ata_device *dev);
 
-	void (*exec_command)(struct ata_port *ap, const struct ata_taskfile *tf);
+	void (*freeze)(struct ata_port *ap);
+	void (*thaw)(struct ata_port *ap);
+	void (*error_handler)(struct ata_port *ap);
+	void (*post_internal_cmd)(struct ata_queued_cmd *qc);
+
+	/*
+	 * Optional features
+	 */
+	int  (*scr_read)(struct ata_port *ap, unsigned int sc_reg, u32 *val);
+	int  (*scr_write)(struct ata_port *ap, unsigned int sc_reg, u32 val);
+	void (*pmp_attach)(struct ata_port *ap);
+	void (*pmp_detach)(struct ata_port *ap);
+	int  (*enable_pm)(struct ata_port *ap, enum link_pm policy);
+	void (*disable_pm)(struct ata_port *ap);
+
+	/*
+	 * Start, stop, suspend and resume
+	 */
+	int  (*port_suspend)(struct ata_port *ap, pm_message_t mesg);
+	int  (*port_resume)(struct ata_port *ap);
+	int  (*port_start)(struct ata_port *ap);
+	void (*port_stop)(struct ata_port *ap);
+	void (*host_stop)(struct ata_host *host);
+
+	/*
+	 * SFF / taskfile oriented ops
+	 */
+	void (*dev_select)(struct ata_port *ap, unsigned int device);
 	u8   (*check_status)(struct ata_port *ap);
 	u8   (*check_altstatus)(struct ata_port *ap);
-	void (*dev_select)(struct ata_port *ap, unsigned int device);
-
-	void (*phy_reset) (struct ata_port *ap); /* obsolete */
-	int  (*set_mode) (struct ata_link *link, struct ata_device **r_failed_dev);
-
-	int (*cable_detect) (struct ata_port *ap);
-
-	int  (*check_atapi_dma) (struct ata_queued_cmd *qc);
-
-	void (*bmdma_setup) (struct ata_queued_cmd *qc);
-	void (*bmdma_start) (struct ata_queued_cmd *qc);
-
-	unsigned int (*data_xfer) (struct ata_device *dev, unsigned char *buf,
-				   unsigned int buflen, int rw);
-
-	int (*qc_defer) (struct ata_queued_cmd *qc);
-	void (*qc_prep) (struct ata_queued_cmd *qc);
-	unsigned int (*qc_issue) (struct ata_queued_cmd *qc);
-
-	/* port multiplier */
-	void (*pmp_attach) (struct ata_port *ap);
-	void (*pmp_detach) (struct ata_port *ap);
-
-	/* Error handlers.  ->error_handler overrides ->eng_timeout and
-	 * indicates that new-style EH is in place.
+	void (*tf_load)(struct ata_port *ap, const struct ata_taskfile *tf);
+	void (*tf_read)(struct ata_port *ap, struct ata_taskfile *tf);
+	void (*exec_command)(struct ata_port *ap, const struct ata_taskfile *tf);
+	unsigned int (*data_xfer)(struct ata_device *dev, unsigned char *buf,
+				  unsigned int buflen, int rw);
+	u8   (*irq_on)(struct ata_port *);
+
+	void (*irq_clear)(struct ata_port *);
+	void (*bmdma_setup)(struct ata_queued_cmd *qc);
+	void (*bmdma_start)(struct ata_queued_cmd *qc);
+	void (*bmdma_stop)(struct ata_queued_cmd *qc);
+	u8   (*bmdma_status)(struct ata_port *ap);
+
+	/*
+	 * Obsolete
 	 */
-	void (*eng_timeout) (struct ata_port *ap); /* obsolete */
-
-	void (*freeze) (struct ata_port *ap);
-	void (*thaw) (struct ata_port *ap);
-	void (*error_handler) (struct ata_port *ap);
-	void (*post_internal_cmd) (struct ata_queued_cmd *qc);
-
+	void (*phy_reset)(struct ata_port *ap);
+	void (*eng_timeout)(struct ata_port *ap);
 	irq_handler_t irq_handler;
-	void (*irq_clear) (struct ata_port *);
-	u8 (*irq_on) (struct ata_port *);
-
-	int (*scr_read) (struct ata_port *ap, unsigned int sc_reg, u32 *val);
-	int (*scr_write) (struct ata_port *ap, unsigned int sc_reg, u32 val);
-
-	int (*port_suspend) (struct ata_port *ap, pm_message_t mesg);
-	int (*port_resume) (struct ata_port *ap);
-	int (*enable_pm) (struct ata_port *ap, enum link_pm policy);
-	void (*disable_pm) (struct ata_port *ap);
-	int (*port_start) (struct ata_port *ap);
-	void (*port_stop) (struct ata_port *ap);
-
-	void (*host_stop) (struct ata_host *host);
-
-	void (*bmdma_stop) (struct ata_queued_cmd *qc);
-	u8   (*bmdma_status) (struct ata_port *ap);
 };
 
 struct ata_port_info {
-- 
cgit v1.2.3


From 358f9a77a668660729e705fde9c3cf69f013aa98 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:47 +0900
Subject: libata: implement and use ata_noop_irq_clear()

->irq_clear() is used to clear IRQ bit of a SFF controller and isn't
useful for drivers which don't use libata SFF HSM implementation.
However, it's a required callback and many drivers implement their own
noop version as placeholder.  This patch implements ata_noop_irq_clear
and use it to replace those custom placeholders.

Also, SFF drivers which don't support BMDMA don't need to use
ata_bmdma_irq_clear().  It becomes noop if BMDMA address isn't
initialized.  Convert them to use ata_noop_irq_clear().

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c           | 12 +++---------
 drivers/ata/libata-core.c    |  1 +
 drivers/ata/libata-sff.c     |  8 ++++++++
 drivers/ata/pata_ali.c       |  2 +-
 drivers/ata/pata_at32.c      |  7 +------
 drivers/ata/pata_icside.c    |  7 +------
 drivers/ata/pata_isapnp.c    |  2 +-
 drivers/ata/pata_ixp4xx_cf.c |  2 +-
 drivers/ata/pata_legacy.c    | 22 +++++++++++-----------
 drivers/ata/pata_mpc52xx.c   |  2 +-
 drivers/ata/pata_mpiix.c     |  2 +-
 drivers/ata/pata_ns87410.c   |  2 +-
 drivers/ata/pata_pcmcia.c    |  4 ++--
 drivers/ata/pata_platform.c  |  2 +-
 drivers/ata/pata_qdi.c       |  4 ++--
 drivers/ata/pata_rb500_cf.c  |  6 +-----
 drivers/ata/pata_winbond.c   |  2 +-
 drivers/ata/pdc_adma.c       |  8 +-------
 drivers/ata/sata_fsl.c       |  7 +------
 drivers/ata/sata_inic162x.c  |  7 +------
 drivers/ata/sata_mv.c        | 11 +++--------
 drivers/ata/sata_qstor.c     |  8 +-------
 drivers/ata/sata_sil24.c     |  8 +-------
 include/linux/libata.h       |  1 +
 24 files changed, 47 insertions(+), 90 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 1bd258e5390f..492e521715d6 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -244,7 +244,6 @@ static int ahci_scr_read(struct ata_port *ap, unsigned int sc_reg, u32 *val);
 static int ahci_scr_write(struct ata_port *ap, unsigned int sc_reg, u32 val);
 static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
 static unsigned int ahci_qc_issue(struct ata_queued_cmd *qc);
-static void ahci_irq_clear(struct ata_port *ap);
 static int ahci_port_start(struct ata_port *ap);
 static void ahci_port_stop(struct ata_port *ap);
 static void ahci_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
@@ -307,7 +306,7 @@ static const struct ata_port_operations ahci_ops = {
 	.qc_prep		= ahci_qc_prep,
 	.qc_issue		= ahci_qc_issue,
 
-	.irq_clear		= ahci_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 
 	.scr_read		= ahci_scr_read,
 	.scr_write		= ahci_scr_write,
@@ -343,7 +342,7 @@ static const struct ata_port_operations ahci_vt8251_ops = {
 	.qc_prep		= ahci_qc_prep,
 	.qc_issue		= ahci_qc_issue,
 
-	.irq_clear		= ahci_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 
 	.scr_read		= ahci_scr_read,
 	.scr_write		= ahci_scr_write,
@@ -377,7 +376,7 @@ static const struct ata_port_operations ahci_p5wdh_ops = {
 	.qc_prep		= ahci_qc_prep,
 	.qc_issue		= ahci_qc_issue,
 
-	.irq_clear		= ahci_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 
 	.scr_read		= ahci_scr_read,
 	.scr_write		= ahci_scr_write,
@@ -1769,11 +1768,6 @@ static void ahci_port_intr(struct ata_port *ap)
 	}
 }
 
-static void ahci_irq_clear(struct ata_port *ap)
-{
-	/* TODO */
-}
-
 static irqreturn_t ahci_interrupt(int irq, void *dev_instance)
 {
 	struct ata_host *host = dev_instance;
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 4dc429fd0056..394edf937cf2 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -7824,6 +7824,7 @@ EXPORT_SYMBOL_GPL(ata_noop_qc_prep);
 EXPORT_SYMBOL_GPL(ata_bmdma_setup);
 EXPORT_SYMBOL_GPL(ata_bmdma_start);
 EXPORT_SYMBOL_GPL(ata_bmdma_irq_clear);
+EXPORT_SYMBOL_GPL(ata_noop_irq_clear);
 EXPORT_SYMBOL_GPL(ata_bmdma_status);
 EXPORT_SYMBOL_GPL(ata_bmdma_stop);
 EXPORT_SYMBOL_GPL(ata_bmdma_freeze);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 20dc572fb45a..1cf03d41aa33 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -302,6 +302,14 @@ void ata_bmdma_irq_clear(struct ata_port *ap)
 	iowrite8(ioread8(mmio + ATA_DMA_STATUS), mmio + ATA_DMA_STATUS);
 }
 
+/**
+ *	ata_noop_irq_clear - Noop placeholder for irq_clear
+ *	@ap: Port associated with this ATA transaction.
+ */
+void ata_noop_irq_clear(struct ata_port *ap)
+{
+}
+
 /**
  *	ata_bmdma_status - Read PCI IDE BMDMA status
  *	@ap: Port associated with this ATA transaction.
diff --git a/drivers/ata/pata_ali.c b/drivers/ata/pata_ali.c
index 3814aebefb2d..84b748aef386 100644
--- a/drivers/ata/pata_ali.c
+++ b/drivers/ata/pata_ali.c
@@ -380,7 +380,7 @@ static struct ata_port_operations ali_early_port_ops = {
 	.data_xfer	= ata_data_xfer,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pata_at32.c b/drivers/ata/pata_at32.c
index db057b183d60..27c959f35c2c 100644
--- a/drivers/ata/pata_at32.c
+++ b/drivers/ata/pata_at32.c
@@ -166,11 +166,6 @@ static void pata_at32_set_piomode(struct ata_port *ap, struct ata_device *adev)
 	}
 }
 
-static void pata_at32_irq_clear(struct ata_port *ap)
-{
-	/* No DMA controller yet */
-}
-
 static struct scsi_host_template at32_sht = {
 	.module			= THIS_MODULE,
 	.name			= DRV_NAME,
@@ -208,7 +203,7 @@ static struct ata_port_operations at32_port_ops = {
 
 	.data_xfer		= ata_data_xfer,
 
-	.irq_clear		= pata_at32_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.port_start		= ata_sff_port_start,
diff --git a/drivers/ata/pata_icside.c b/drivers/ata/pata_icside.c
index f97068be2d79..e1230cae7ff1 100644
--- a/drivers/ata/pata_icside.c
+++ b/drivers/ata/pata_icside.c
@@ -322,11 +322,6 @@ static struct scsi_host_template pata_icside_sht = {
 	.bios_param		= ata_std_bios_param,
 };
 
-/* wish this was exported from libata-core */
-static void ata_dummy_noret(struct ata_port *port)
-{
-}
-
 static void pata_icside_postreset(struct ata_link *link, unsigned int *classes)
 {
 	struct ata_port *ap = link->ap;
@@ -380,7 +375,7 @@ static struct ata_port_operations pata_icside_port_ops = {
 	.error_handler		= pata_icside_error_handler,
 	.post_internal_cmd	= pata_icside_bmdma_stop,
 
-	.irq_clear		= ata_dummy_noret,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.bmdma_stop		= pata_icside_bmdma_stop,
diff --git a/drivers/ata/pata_isapnp.c b/drivers/ata/pata_isapnp.c
index 4320e7986321..ef561de0c24d 100644
--- a/drivers/ata/pata_isapnp.c
+++ b/drivers/ata/pata_isapnp.c
@@ -55,7 +55,7 @@ static struct ata_port_operations isapnp_port_ops = {
 
 	.data_xfer	= ata_data_xfer,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pata_ixp4xx_cf.c b/drivers/ata/pata_ixp4xx_cf.c
index 030878fedeb5..83e38cc077e1 100644
--- a/drivers/ata/pata_ixp4xx_cf.c
+++ b/drivers/ata/pata_ixp4xx_cf.c
@@ -126,7 +126,7 @@ static struct ata_port_operations ixp4xx_port_ops = {
 	.cable_detect		= ata_cable_40wire,
 
 	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.port_start		= ata_port_start,
diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c
index 50fe08ebe23c..6ac02f7d5289 100644
--- a/drivers/ata/pata_legacy.c
+++ b/drivers/ata/pata_legacy.c
@@ -252,7 +252,7 @@ static struct ata_port_operations simple_port_ops = {
 	.data_xfer	= ata_data_xfer_noirq,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -279,7 +279,7 @@ static struct ata_port_operations legacy_port_ops = {
 	.data_xfer	= ata_data_xfer_noirq,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -393,7 +393,7 @@ static struct ata_port_operations pdc20230_port_ops = {
 	.data_xfer	= pdc_data_xfer_vlb,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -447,7 +447,7 @@ static struct ata_port_operations ht6560a_port_ops = {
 	.data_xfer	= ata_data_xfer,	/* Check vlb/noirq */
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -512,7 +512,7 @@ static struct ata_port_operations ht6560b_port_ops = {
 	.data_xfer	= ata_data_xfer,    /* FIXME: Check 32bit and noirq */
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -633,7 +633,7 @@ static struct ata_port_operations opti82c611a_port_ops = {
 	.data_xfer	= ata_data_xfer,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -765,7 +765,7 @@ static struct ata_port_operations opti82c46x_port_ops = {
 	.data_xfer	= ata_data_xfer,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -950,7 +950,7 @@ static struct ata_port_operations qdi6500_port_ops = {
 	.data_xfer	= vlb32_data_xfer,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -977,7 +977,7 @@ static struct ata_port_operations qdi6580_port_ops = {
 	.data_xfer	= vlb32_data_xfer,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -1004,7 +1004,7 @@ static struct ata_port_operations qdi6580dp_port_ops = {
 	.data_xfer	= vlb32_data_xfer,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -1095,7 +1095,7 @@ static struct ata_port_operations winbond_port_ops = {
 
 	.data_xfer	= vlb32_data_xfer,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pata_mpc52xx.c b/drivers/ata/pata_mpc52xx.c
index 5413ebfa72e5..d84e0c8ea02f 100644
--- a/drivers/ata/pata_mpc52xx.c
+++ b/drivers/ata/pata_mpc52xx.c
@@ -296,7 +296,7 @@ static struct ata_port_operations mpc52xx_ata_port_ops = {
 	.qc_prep		= ata_qc_prep,
 	.qc_issue		= ata_qc_issue_prot,
 	.data_xfer		= ata_data_xfer,
-	.irq_clear		= ata_bmdma_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 	.port_start		= ata_port_start,
 };
diff --git a/drivers/ata/pata_mpiix.c b/drivers/ata/pata_mpiix.c
index c0d9e0cf208c..ced6372749b3 100644
--- a/drivers/ata/pata_mpiix.c
+++ b/drivers/ata/pata_mpiix.c
@@ -187,7 +187,7 @@ static struct ata_port_operations mpiix_port_ops = {
 	.qc_issue	= mpiix_qc_issue_prot,
 	.data_xfer	= ata_data_xfer,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
index 9fe66fd75017..d182bdf31ee1 100644
--- a/drivers/ata/pata_ns87410.c
+++ b/drivers/ata/pata_ns87410.c
@@ -182,7 +182,7 @@ static struct ata_port_operations ns87410_port_ops = {
 	.data_xfer	= ata_data_xfer,
 
 	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pata_pcmcia.c b/drivers/ata/pata_pcmcia.c
index 3e7f6a9da28b..9881a9e004a4 100644
--- a/drivers/ata/pata_pcmcia.c
+++ b/drivers/ata/pata_pcmcia.c
@@ -164,7 +164,7 @@ static struct ata_port_operations pcmcia_port_ops = {
 
 	.data_xfer	= ata_data_xfer_noirq,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -189,7 +189,7 @@ static struct ata_port_operations pcmcia_8bit_port_ops = {
 
 	.data_xfer	= ata_data_xfer_8bit,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pata_platform.c b/drivers/ata/pata_platform.c
index aad7adc6ea56..a8429f5c3006 100644
--- a/drivers/ata/pata_platform.c
+++ b/drivers/ata/pata_platform.c
@@ -86,7 +86,7 @@ static struct ata_port_operations pata_platform_port_ops = {
 
 	.data_xfer		= ata_data_xfer_noirq,
 
-	.irq_clear		= ata_bmdma_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.port_start		= ata_dummy_ret0,
diff --git a/drivers/ata/pata_qdi.c b/drivers/ata/pata_qdi.c
index 9f308ed76cc8..60238d5748a7 100644
--- a/drivers/ata/pata_qdi.c
+++ b/drivers/ata/pata_qdi.c
@@ -191,7 +191,7 @@ static struct ata_port_operations qdi6500_port_ops = {
 
 	.data_xfer	= qdi_data_xfer,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
@@ -217,7 +217,7 @@ static struct ata_port_operations qdi6580_port_ops = {
 
 	.data_xfer	= qdi_data_xfer,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pata_rb500_cf.c b/drivers/ata/pata_rb500_cf.c
index 4ce9b03fe6c8..08c32af1c84d 100644
--- a/drivers/ata/pata_rb500_cf.c
+++ b/drivers/ata/pata_rb500_cf.c
@@ -117,10 +117,6 @@ static irqreturn_t rb500_pata_irq_handler(int irq, void *dev_instance)
 	return IRQ_HANDLED;
 }
 
-static void rb500_pata_irq_clear(struct ata_port *ap)
-{
-}
-
 static int rb500_pata_port_start(struct ata_port *ap)
 {
 	return 0;
@@ -144,7 +140,7 @@ static struct ata_port_operations rb500_pata_port_ops = {
 	.error_handler		= ata_bmdma_error_handler,
 
 	.irq_handler		= rb500_pata_irq_handler,
-	.irq_clear		= rb500_pata_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.port_start		= rb500_pata_port_start,
diff --git a/drivers/ata/pata_winbond.c b/drivers/ata/pata_winbond.c
index 99c92eda217b..5318248782bb 100644
--- a/drivers/ata/pata_winbond.c
+++ b/drivers/ata/pata_winbond.c
@@ -159,7 +159,7 @@ static struct ata_port_operations winbond_port_ops = {
 
 	.data_xfer	= winbond_data_xfer,
 
-	.irq_clear	= ata_bmdma_irq_clear,
+	.irq_clear	= ata_noop_irq_clear,
 	.irq_on		= ata_irq_on,
 
 	.port_start	= ata_sff_port_start,
diff --git a/drivers/ata/pdc_adma.c b/drivers/ata/pdc_adma.c
index 8e1b7e9c0ae4..bc2d12a2da30 100644
--- a/drivers/ata/pdc_adma.c
+++ b/drivers/ata/pdc_adma.c
@@ -138,7 +138,6 @@ static unsigned int adma_qc_issue(struct ata_queued_cmd *qc);
 static int adma_check_atapi_dma(struct ata_queued_cmd *qc);
 static void adma_bmdma_stop(struct ata_queued_cmd *qc);
 static u8 adma_bmdma_status(struct ata_port *ap);
-static void adma_irq_clear(struct ata_port *ap);
 static void adma_freeze(struct ata_port *ap);
 static void adma_thaw(struct ata_port *ap);
 static void adma_error_handler(struct ata_port *ap);
@@ -174,7 +173,7 @@ static const struct ata_port_operations adma_ata_ops = {
 	.freeze			= adma_freeze,
 	.thaw			= adma_thaw,
 	.error_handler		= adma_error_handler,
-	.irq_clear		= adma_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 	.port_start		= adma_port_start,
 	.port_stop		= adma_port_stop,
@@ -223,11 +222,6 @@ static u8 adma_bmdma_status(struct ata_port *ap)
 	return 0;
 }
 
-static void adma_irq_clear(struct ata_port *ap)
-{
-	/* nothing */
-}
-
 static void adma_reset_engine(struct ata_port *ap)
 {
 	void __iomem *chan = ADMA_PORT_REGS(ap);
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index d23b690ed1a1..031a512cbaa8 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -933,11 +933,6 @@ static void sata_fsl_post_internal_cmd(struct ata_queued_cmd *qc)
 	}
 }
 
-static void sata_fsl_irq_clear(struct ata_port *ap)
-{
-	/* unused */
-}
-
 static void sata_fsl_error_intr(struct ata_port *ap)
 {
 	struct ata_link *link = &ap->link;
@@ -1223,7 +1218,7 @@ static const struct ata_port_operations sata_fsl_ops = {
 
 	.qc_prep = sata_fsl_qc_prep,
 	.qc_issue = sata_fsl_qc_issue,
-	.irq_clear = sata_fsl_irq_clear,
+	.irq_clear = ata_noop_irq_clear,
 
 	.scr_read = sata_fsl_scr_read,
 	.scr_write = sata_fsl_scr_write,
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index 59e65edc5820..74f14369dc8d 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -266,11 +266,6 @@ static u8 inic_bmdma_status(struct ata_port *ap)
 	return ATA_DMA_INTR;
 }
 
-static void inic_irq_clear(struct ata_port *ap)
-{
-	/* noop */
-}
-
 static void inic_host_intr(struct ata_port *ap)
 {
 	void __iomem *port_base = inic_port_base(ap);
@@ -555,7 +550,7 @@ static struct ata_port_operations inic_port_ops = {
 	.bmdma_stop		= inic_bmdma_stop,
 	.bmdma_status		= inic_bmdma_status,
 
-	.irq_clear		= inic_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.qc_prep	 	= ata_qc_prep,
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index a4944c8ad46d..b3b3da4eaa03 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -461,7 +461,6 @@ struct mv_hw_ops {
 	void (*reset_bus)(struct ata_host *host, void __iomem *mmio);
 };
 
-static void mv_irq_clear(struct ata_port *ap);
 static int mv_scr_read(struct ata_port *ap, unsigned int sc_reg_in, u32 *val);
 static int mv_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val);
 static int mv5_scr_read(struct ata_port *ap, unsigned int sc_reg_in, u32 *val);
@@ -564,7 +563,7 @@ static const struct ata_port_operations mv5_ops = {
 	.qc_issue		= mv_qc_issue,
 	.data_xfer		= ata_data_xfer,
 
-	.irq_clear		= mv_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.error_handler		= mv_error_handler,
@@ -592,7 +591,7 @@ static const struct ata_port_operations mv6_ops = {
 	.qc_issue		= mv_qc_issue,
 	.data_xfer		= ata_data_xfer,
 
-	.irq_clear		= mv_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.error_handler		= mv_error_handler,
@@ -620,7 +619,7 @@ static const struct ata_port_operations mv_iie_ops = {
 	.qc_issue		= mv_qc_issue,
 	.data_xfer		= ata_data_xfer,
 
-	.irq_clear		= mv_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 
 	.error_handler		= mv_error_handler,
@@ -801,10 +800,6 @@ static inline int mv_get_hc_count(unsigned long port_flags)
 	return ((port_flags & MV_FLAG_DUAL_HC) ? 2 : 1);
 }
 
-static void mv_irq_clear(struct ata_port *ap)
-{
-}
-
 static void mv_set_edma_ptrs(void __iomem *port_mmio,
 			     struct mv_host_priv *hpriv,
 			     struct mv_port_priv *pp)
diff --git a/drivers/ata/sata_qstor.c b/drivers/ata/sata_qstor.c
index 91cc12c82040..3c8e97f251f9 100644
--- a/drivers/ata/sata_qstor.c
+++ b/drivers/ata/sata_qstor.c
@@ -121,7 +121,6 @@ static unsigned int qs_qc_issue(struct ata_queued_cmd *qc);
 static int qs_check_atapi_dma(struct ata_queued_cmd *qc);
 static void qs_bmdma_stop(struct ata_queued_cmd *qc);
 static u8 qs_bmdma_status(struct ata_port *ap);
-static void qs_irq_clear(struct ata_port *ap);
 static void qs_freeze(struct ata_port *ap);
 static void qs_thaw(struct ata_port *ap);
 static void qs_error_handler(struct ata_port *ap);
@@ -157,7 +156,7 @@ static const struct ata_port_operations qs_ata_ops = {
 	.freeze			= qs_freeze,
 	.thaw			= qs_thaw,
 	.error_handler		= qs_error_handler,
-	.irq_clear		= qs_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 	.irq_on			= ata_irq_on,
 	.scr_read		= qs_scr_read,
 	.scr_write		= qs_scr_write,
@@ -211,11 +210,6 @@ static u8 qs_bmdma_status(struct ata_port *ap)
 	return 0;
 }
 
-static void qs_irq_clear(struct ata_port *ap)
-{
-	/* nothing */
-}
-
 static inline void qs_enter_reg_mode(struct ata_port *ap)
 {
 	u8 __iomem *chan = qs_mmio_base(ap->host) + (ap->port_no * 0x4000);
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index ba0c00e8ee7f..b85464d51f68 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -348,7 +348,6 @@ static void sil24_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
 static int sil24_qc_defer(struct ata_queued_cmd *qc);
 static void sil24_qc_prep(struct ata_queued_cmd *qc);
 static unsigned int sil24_qc_issue(struct ata_queued_cmd *qc);
-static void sil24_irq_clear(struct ata_port *ap);
 static void sil24_pmp_attach(struct ata_port *ap);
 static void sil24_pmp_detach(struct ata_port *ap);
 static void sil24_freeze(struct ata_port *ap);
@@ -416,7 +415,7 @@ static const struct ata_port_operations sil24_ops = {
 	.qc_prep		= sil24_qc_prep,
 	.qc_issue		= sil24_qc_issue,
 
-	.irq_clear		= sil24_irq_clear,
+	.irq_clear		= ata_noop_irq_clear,
 
 	.scr_read		= sil24_scr_read,
 	.scr_write		= sil24_scr_write,
@@ -921,11 +920,6 @@ static unsigned int sil24_qc_issue(struct ata_queued_cmd *qc)
 	return 0;
 }
 
-static void sil24_irq_clear(struct ata_port *ap)
-{
-	/* unused */
-}
-
 static void sil24_pmp_attach(struct ata_port *ap)
 {
 	sil24_config_pmp(ap, 1);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 9476a479691b..639298af583e 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -896,6 +896,7 @@ extern void ata_bmdma_start(struct ata_queued_cmd *qc);
 extern void ata_bmdma_stop(struct ata_queued_cmd *qc);
 extern u8   ata_bmdma_status(struct ata_port *ap);
 extern void ata_bmdma_irq_clear(struct ata_port *ap);
+extern void ata_noop_irq_clear(struct ata_port *ap);
 extern void ata_bmdma_freeze(struct ata_port *ap);
 extern void ata_bmdma_thaw(struct ata_port *ap);
 extern void ata_bmdma_drive_eh(struct ata_port *ap, ata_prereset_fn_t prereset,
-- 
cgit v1.2.3


From 68d1d07b510bb57a504588adc2bd2758adea0965 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:49 +0900
Subject: libata: implement and use SHT initializers

libata lets low level drivers build scsi_host_template and register it
to the SCSI layer.  This allows low level drivers high level of
flexibility but also burdens them with lots of boilerplate entries.

This patch implements SHT initializers which can be used to initialize
all the boilerplate entries in a sht.  Three variants of them are
implemented - BASE, BMDMA and NCQ - for different types of drivers.
Note that entries can be overriden by putting individual initializers
after the helper macro.

All sht tables are identical before and after this patch.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c              | 15 +----------
 drivers/ata/ata_generic.c       | 16 +-----------
 drivers/ata/ata_piix.c          | 16 +-----------
 drivers/ata/pata_acpi.c         | 17 +-----------
 drivers/ata/pata_ali.c          | 16 +-----------
 drivers/ata/pata_amd.c          | 16 +-----------
 drivers/ata/pata_artop.c        | 16 +-----------
 drivers/ata/pata_at32.c         | 16 +-----------
 drivers/ata/pata_atiixp.c       | 15 +----------
 drivers/ata/pata_bf54x.c        | 14 +---------
 drivers/ata/pata_cmd640.c       | 16 +-----------
 drivers/ata/pata_cmd64x.c       | 16 +-----------
 drivers/ata/pata_cs5520.c       | 15 +----------
 drivers/ata/pata_cs5530.c       | 17 ++----------
 drivers/ata/pata_cs5535.c       | 16 +-----------
 drivers/ata/pata_cs5536.c       | 16 +-----------
 drivers/ata/pata_cypress.c      | 16 +-----------
 drivers/ata/pata_efar.c         | 16 +-----------
 drivers/ata/pata_hpt366.c       | 16 +-----------
 drivers/ata/pata_hpt37x.c       | 16 +-----------
 drivers/ata/pata_hpt3x2n.c      | 16 +-----------
 drivers/ata/pata_hpt3x3.c       | 16 +-----------
 drivers/ata/pata_icside.c       | 14 +---------
 drivers/ata/pata_isapnp.c       | 16 +-----------
 drivers/ata/pata_it8213.c       | 16 +-----------
 drivers/ata/pata_it821x.c       | 16 +-----------
 drivers/ata/pata_ixp4xx_cf.c    | 16 +-----------
 drivers/ata/pata_jmicron.c      | 17 +-----------
 drivers/ata/pata_legacy.c       | 16 +-----------
 drivers/ata/pata_marvell.c      | 17 +-----------
 drivers/ata/pata_mpc52xx.c      | 16 +-----------
 drivers/ata/pata_mpiix.c        | 16 +-----------
 drivers/ata/pata_netcell.c      | 17 +-----------
 drivers/ata/pata_ninja32.c      | 16 +-----------
 drivers/ata/pata_ns87410.c      | 16 +-----------
 drivers/ata/pata_ns87415.c      | 16 +-----------
 drivers/ata/pata_oldpiix.c      | 16 +-----------
 drivers/ata/pata_opti.c         | 16 +-----------
 drivers/ata/pata_optidma.c      | 16 +-----------
 drivers/ata/pata_pcmcia.c       | 16 +-----------
 drivers/ata/pata_pdc2027x.c     | 16 +-----------
 drivers/ata/pata_pdc202xx_old.c | 16 +-----------
 drivers/ata/pata_platform.c     | 16 +-----------
 drivers/ata/pata_qdi.c          | 16 +-----------
 drivers/ata/pata_radisys.c      | 16 +-----------
 drivers/ata/pata_rb500_cf.c     | 17 +-----------
 drivers/ata/pata_rz1000.c       | 16 +-----------
 drivers/ata/pata_sc1200.c       | 17 ++----------
 drivers/ata/pata_scc.c          | 16 +-----------
 drivers/ata/pata_serverworks.c  | 16 +-----------
 drivers/ata/pata_sil680.c       | 16 +-----------
 drivers/ata/pata_sis.c          | 16 +-----------
 drivers/ata/pata_sl82c105.c     | 16 +-----------
 drivers/ata/pata_triflex.c      | 16 +-----------
 drivers/ata/pata_via.c          | 16 +-----------
 drivers/ata/pata_winbond.c      | 16 +-----------
 drivers/ata/pdc_adma.c          | 14 +---------
 drivers/ata/sata_fsl.c          | 14 +---------
 drivers/ata/sata_inic162x.c     | 16 +-----------
 drivers/ata/sata_mv.c           | 28 ++------------------
 drivers/ata/sata_nv.c           | 42 +++---------------------------
 drivers/ata/sata_promise.c      | 14 +---------
 drivers/ata/sata_qstor.c        | 14 +---------
 drivers/ata/sata_sil.c          | 16 +-----------
 drivers/ata/sata_sil24.c        | 14 +---------
 drivers/ata/sata_sis.c          | 16 +-----------
 drivers/ata/sata_svw.c          | 16 +-----------
 drivers/ata/sata_sx4.c          | 14 +---------
 drivers/ata/sata_uli.c          | 16 +-----------
 drivers/ata/sata_via.c          | 16 +-----------
 drivers/ata/sata_vsc.c          | 16 +-----------
 include/linux/libata.h          | 57 +++++++++++++++++++++++++++++++++++++++++
 72 files changed, 133 insertions(+), 1086 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index c6ea44a7f2a9..8862595cb2cf 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -62,7 +62,6 @@ enum {
 	AHCI_MAX_PORTS		= 32,
 	AHCI_MAX_SG		= 168, /* hardware max is 64K */
 	AHCI_DMA_BOUNDARY	= 0xffffffff,
-	AHCI_USE_CLUSTERING	= 1,
 	AHCI_MAX_CMDS		= 32,
 	AHCI_CMD_SZ		= 32,
 	AHCI_CMD_SLOT_SZ	= AHCI_MAX_CMDS * AHCI_CMD_SZ,
@@ -274,22 +273,10 @@ static struct class_device_attribute *ahci_shost_attrs[] = {
 };
 
 static struct scsi_host_template ahci_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.change_queue_depth	= ata_scsi_change_queue_depth,
+	ATA_NCQ_SHT(DRV_NAME),
 	.can_queue		= AHCI_MAX_CMDS - 1,
-	.this_id		= ATA_SHT_THIS_ID,
 	.sg_tablesize		= AHCI_MAX_SG,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= AHCI_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= AHCI_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 	.shost_attrs		= ahci_shost_attrs,
 };
 
diff --git a/drivers/ata/ata_generic.c b/drivers/ata/ata_generic.c
index db4c3cb78fda..5c64ce134c6c 100644
--- a/drivers/ata/ata_generic.c
+++ b/drivers/ata/ata_generic.c
@@ -95,21 +95,7 @@ static int generic_set_mode(struct ata_link *link, struct ata_device **unused)
 }
 
 static struct scsi_host_template generic_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations generic_port_ops = {
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index 067760a16889..9f887b2c92df 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -291,21 +291,7 @@ static struct pci_driver piix_pci_driver = {
 };
 
 static struct scsi_host_template piix_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations piix_pata_ops = {
diff --git a/drivers/ata/pata_acpi.c b/drivers/ata/pata_acpi.c
index bdc3b9d7395c..187545c0898a 100644
--- a/drivers/ata/pata_acpi.c
+++ b/drivers/ata/pata_acpi.c
@@ -232,22 +232,7 @@ static int pacpi_port_start(struct ata_port *ap)
 }
 
 static struct scsi_host_template pacpi_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	/* Use standard CHS mapping rules */
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations pacpi_ops = {
diff --git a/drivers/ata/pata_ali.c b/drivers/ata/pata_ali.c
index 84b748aef386..f3d6d9b345ba 100644
--- a/drivers/ata/pata_ali.c
+++ b/drivers/ata/pata_ali.c
@@ -339,21 +339,7 @@ static int ali_check_atapi_dma(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template ali_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 /*
diff --git a/drivers/ata/pata_amd.c b/drivers/ata/pata_amd.c
index 5e1bc13a756b..90d786dfbec3 100644
--- a/drivers/ata/pata_amd.c
+++ b/drivers/ata/pata_amd.c
@@ -353,21 +353,7 @@ static void nv_host_stop(struct ata_host *host)
 }
 
 static struct scsi_host_template amd_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations amd33_port_ops = {
diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c
index a238c7bd0bba..7bfb7e8bdca2 100644
--- a/drivers/ata/pata_artop.c
+++ b/drivers/ata/pata_artop.c
@@ -314,21 +314,7 @@ static void artop6260_set_dmamode (struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template artop_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations artop6210_ops = {
diff --git a/drivers/ata/pata_at32.c b/drivers/ata/pata_at32.c
index 27c959f35c2c..d7b7b7fde362 100644
--- a/drivers/ata/pata_at32.c
+++ b/drivers/ata/pata_at32.c
@@ -167,21 +167,7 @@ static void pata_at32_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template at32_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations at32_port_ops = {
diff --git a/drivers/ata/pata_atiixp.c b/drivers/ata/pata_atiixp.c
index 408bdc1a9776..645c47271ff5 100644
--- a/drivers/ata/pata_atiixp.c
+++ b/drivers/ata/pata_atiixp.c
@@ -221,21 +221,8 @@ static void atiixp_bmdma_stop(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template atiixp_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BMDMA_SHT(DRV_NAME),
 	.sg_tablesize		= LIBATA_DUMB_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static struct ata_port_operations atiixp_port_ops = {
diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c
index 146c202d5834..6c75fcac3cf4 100644
--- a/drivers/ata/pata_bf54x.c
+++ b/drivers/ata/pata_bf54x.c
@@ -1357,21 +1357,9 @@ static int bfin_port_start(struct ata_port *ap)
 }
 
 static struct scsi_host_template bfin_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BASE_SHT(DRV_NAME),
 	.sg_tablesize		= SG_NONE,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static const struct ata_port_operations bfin_pata_ops = {
diff --git a/drivers/ata/pata_cmd640.c b/drivers/ata/pata_cmd640.c
index 0ef1d1ded1f8..26562b814400 100644
--- a/drivers/ata/pata_cmd640.c
+++ b/drivers/ata/pata_cmd640.c
@@ -166,21 +166,7 @@ static int cmd640_port_start(struct ata_port *ap)
 }
 
 static struct scsi_host_template cmd640_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations cmd640_port_ops = {
diff --git a/drivers/ata/pata_cmd64x.c b/drivers/ata/pata_cmd64x.c
index e8c1262341ee..6aea05cc0940 100644
--- a/drivers/ata/pata_cmd64x.c
+++ b/drivers/ata/pata_cmd64x.c
@@ -266,21 +266,7 @@ static void cmd646r1_bmdma_stop(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template cmd64x_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations cmd64x_port_ops = {
diff --git a/drivers/ata/pata_cs5520.c b/drivers/ata/pata_cs5520.c
index 44ad2c9d488f..7e643099a444 100644
--- a/drivers/ata/pata_cs5520.c
+++ b/drivers/ata/pata_cs5520.c
@@ -140,21 +140,8 @@ static void cs5520_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template cs5520_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BMDMA_SHT(DRV_NAME),
 	.sg_tablesize		= LIBATA_DUMB_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static struct ata_port_operations cs5520_port_ops = {
diff --git a/drivers/ata/pata_cs5530.c b/drivers/ata/pata_cs5530.c
index f876aeddf1a1..0bb03dabcf18 100644
--- a/drivers/ata/pata_cs5530.c
+++ b/drivers/ata/pata_cs5530.c
@@ -161,21 +161,8 @@ static unsigned int cs5530_qc_issue_prot(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template cs5530_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_DUMB_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
+	.sg_tablesize	= LIBATA_DUMB_MAX_PRD,
 };
 
 static struct ata_port_operations cs5530_port_ops = {
diff --git a/drivers/ata/pata_cs5535.c b/drivers/ata/pata_cs5535.c
index 01324530d052..48a18349c1d8 100644
--- a/drivers/ata/pata_cs5535.c
+++ b/drivers/ata/pata_cs5535.c
@@ -158,21 +158,7 @@ static void cs5535_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template cs5535_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations cs5535_port_ops = {
diff --git a/drivers/ata/pata_cs5536.c b/drivers/ata/pata_cs5536.c
index 391aa888f8fd..f02d9107ef3b 100644
--- a/drivers/ata/pata_cs5536.c
+++ b/drivers/ata/pata_cs5536.c
@@ -221,21 +221,7 @@ static void cs5536_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template cs5536_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations cs5536_port_ops = {
diff --git a/drivers/ata/pata_cypress.c b/drivers/ata/pata_cypress.c
index fc5f9c4e5d87..07fa1ab36315 100644
--- a/drivers/ata/pata_cypress.c
+++ b/drivers/ata/pata_cypress.c
@@ -110,21 +110,7 @@ static void cy82c693_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template cy82c693_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations cy82c693_port_ops = {
diff --git a/drivers/ata/pata_efar.c b/drivers/ata/pata_efar.c
index dc33220fe5b2..8700d9dcd8c9 100644
--- a/drivers/ata/pata_efar.c
+++ b/drivers/ata/pata_efar.c
@@ -233,21 +233,7 @@ static void efar_set_dmamode (struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template efar_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations efar_ops = {
diff --git a/drivers/ata/pata_hpt366.c b/drivers/ata/pata_hpt366.c
index a82089048f58..a30028de41c0 100644
--- a/drivers/ata/pata_hpt366.c
+++ b/drivers/ata/pata_hpt366.c
@@ -290,21 +290,7 @@ static void hpt366_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template hpt36x_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 /*
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index 2ddcd07a7518..7d6fac43e2f9 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -619,21 +619,7 @@ static void hpt37x_bmdma_stop(struct ata_queued_cmd *qc)
 
 
 static struct scsi_host_template hpt37x_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 /*
diff --git a/drivers/ata/pata_hpt3x2n.c b/drivers/ata/pata_hpt3x2n.c
index 3a517cb9bd3e..aa380c46b168 100644
--- a/drivers/ata/pata_hpt3x2n.c
+++ b/drivers/ata/pata_hpt3x2n.c
@@ -339,21 +339,7 @@ static unsigned int hpt3x2n_qc_issue_prot(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template hpt3x2n_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 /*
diff --git a/drivers/ata/pata_hpt3x3.c b/drivers/ata/pata_hpt3x3.c
index c09f95a4a0d1..9837ab0181e8 100644
--- a/drivers/ata/pata_hpt3x3.c
+++ b/drivers/ata/pata_hpt3x3.c
@@ -102,21 +102,7 @@ static int hpt3x3_atapi_dma(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template hpt3x3_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations hpt3x3_port_ops = {
diff --git a/drivers/ata/pata_icside.c b/drivers/ata/pata_icside.c
index e1230cae7ff1..88a1c7ae0a4d 100644
--- a/drivers/ata/pata_icside.c
+++ b/drivers/ata/pata_icside.c
@@ -305,21 +305,9 @@ static int icside_dma_init(struct pata_icside_info *info)
 
 
 static struct scsi_host_template pata_icside_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BASE_SHT(DRV_NAME),
 	.sg_tablesize		= PATA_ICSIDE_MAX_SG,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= ~0, /* no dma boundaries */
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static void pata_icside_postreset(struct ata_link *link, unsigned int *classes)
diff --git a/drivers/ata/pata_isapnp.c b/drivers/ata/pata_isapnp.c
index ef561de0c24d..91ca4d50db04 100644
--- a/drivers/ata/pata_isapnp.c
+++ b/drivers/ata/pata_isapnp.c
@@ -20,21 +20,7 @@
 #define DRV_VERSION "0.2.2"
 
 static struct scsi_host_template isapnp_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations isapnp_port_ops = {
diff --git a/drivers/ata/pata_it8213.c b/drivers/ata/pata_it8213.c
index 25c49c2e1519..678a05b304d8 100644
--- a/drivers/ata/pata_it8213.c
+++ b/drivers/ata/pata_it8213.c
@@ -243,21 +243,7 @@ static void it8213_set_dmamode (struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template it8213_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations it8213_ops = {
diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c
index 6bdbb7140dfa..7d969c911731 100644
--- a/drivers/ata/pata_it821x.c
+++ b/drivers/ata/pata_it821x.c
@@ -632,21 +632,7 @@ static int it821x_port_start(struct ata_port *ap)
 }
 
 static struct scsi_host_template it821x_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations it821x_smart_port_ops = {
diff --git a/drivers/ata/pata_ixp4xx_cf.c b/drivers/ata/pata_ixp4xx_cf.c
index 6eb8cc9a3f12..b7e8e825a869 100644
--- a/drivers/ata/pata_ixp4xx_cf.c
+++ b/drivers/ata/pata_ixp4xx_cf.c
@@ -88,21 +88,7 @@ static unsigned int ixp4xx_mmio_data_xfer(struct ata_device *dev,
 }
 
 static struct scsi_host_template ixp4xx_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations ixp4xx_port_ops = {
diff --git a/drivers/ata/pata_jmicron.c b/drivers/ata/pata_jmicron.c
index 7d36fa85435a..69781af7b1bb 100644
--- a/drivers/ata/pata_jmicron.c
+++ b/drivers/ata/pata_jmicron.c
@@ -122,22 +122,7 @@ static void jmicron_error_handler(struct ata_port *ap)
 /* No PIO or DMA methods needed for this device */
 
 static struct scsi_host_template jmicron_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	/* Use standard CHS mapping rules */
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations jmicron_ops = {
diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c
index 6ac02f7d5289..5329b954c5f2 100644
--- a/drivers/ata/pata_legacy.c
+++ b/drivers/ata/pata_legacy.c
@@ -208,21 +208,7 @@ static int legacy_set_mode(struct ata_link *link, struct ata_device **unused)
 }
 
 static struct scsi_host_template legacy_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 /*
diff --git a/drivers/ata/pata_marvell.c b/drivers/ata/pata_marvell.c
index c4ee9b45301f..9de6e429d0d1 100644
--- a/drivers/ata/pata_marvell.c
+++ b/drivers/ata/pata_marvell.c
@@ -92,22 +92,7 @@ static void marvell_error_handler(struct ata_port *ap)
 /* No PIO or DMA methods needed for this device */
 
 static struct scsi_host_template marvell_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	/* Use standard CHS mapping rules */
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations marvell_ops = {
diff --git a/drivers/ata/pata_mpc52xx.c b/drivers/ata/pata_mpc52xx.c
index fefe71dbed1a..4117b618a9d9 100644
--- a/drivers/ata/pata_mpc52xx.c
+++ b/drivers/ata/pata_mpc52xx.c
@@ -265,21 +265,7 @@ mpc52xx_ata_error_handler(struct ata_port *ap)
 
 
 static struct scsi_host_template mpc52xx_ata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations mpc52xx_ata_port_ops = {
diff --git a/drivers/ata/pata_mpiix.c b/drivers/ata/pata_mpiix.c
index ced6372749b3..e8e6837110b4 100644
--- a/drivers/ata/pata_mpiix.c
+++ b/drivers/ata/pata_mpiix.c
@@ -151,21 +151,7 @@ static unsigned int mpiix_qc_issue_prot(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template mpiix_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations mpiix_port_ops = {
diff --git a/drivers/ata/pata_netcell.c b/drivers/ata/pata_netcell.c
index 9fd1a84c01d3..11f200a2a156 100644
--- a/drivers/ata/pata_netcell.c
+++ b/drivers/ata/pata_netcell.c
@@ -21,22 +21,7 @@
 /* No PIO or DMA methods needed for this device */
 
 static struct scsi_host_template netcell_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	/* Use standard CHS mapping rules */
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations netcell_ops = {
diff --git a/drivers/ata/pata_ninja32.c b/drivers/ata/pata_ninja32.c
index 15dd649f89ee..ce3b07cab8bc 100644
--- a/drivers/ata/pata_ninja32.c
+++ b/drivers/ata/pata_ninja32.c
@@ -79,21 +79,7 @@ static void ninja32_dev_select(struct ata_port *ap, unsigned int device)
 }
 
 static struct scsi_host_template ninja32_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations ninja32_port_ops = {
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
index d182bdf31ee1..d2f85f107d15 100644
--- a/drivers/ata/pata_ns87410.c
+++ b/drivers/ata/pata_ns87410.c
@@ -144,21 +144,7 @@ static unsigned int ns87410_qc_issue_prot(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template ns87410_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations ns87410_port_ops = {
diff --git a/drivers/ata/pata_ns87415.c b/drivers/ata/pata_ns87415.c
index 93eb958cb0c9..78d634423cbf 100644
--- a/drivers/ata/pata_ns87415.c
+++ b/drivers/ata/pata_ns87415.c
@@ -366,21 +366,7 @@ static const struct ata_port_operations ns87415_pata_ops = {
 };
 
 static struct scsi_host_template ns87415_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 
diff --git a/drivers/ata/pata_oldpiix.c b/drivers/ata/pata_oldpiix.c
index 44da09ace52c..45f9b3eb5b45 100644
--- a/drivers/ata/pata_oldpiix.c
+++ b/drivers/ata/pata_oldpiix.c
@@ -220,21 +220,7 @@ static unsigned int oldpiix_qc_issue_prot(struct ata_queued_cmd *qc)
 
 
 static struct scsi_host_template oldpiix_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations oldpiix_pata_ops = {
diff --git a/drivers/ata/pata_opti.c b/drivers/ata/pata_opti.c
index 1e865f138d1c..1deacfa0be07 100644
--- a/drivers/ata/pata_opti.c
+++ b/drivers/ata/pata_opti.c
@@ -165,21 +165,7 @@ static void opti_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template opti_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations opti_port_ops = {
diff --git a/drivers/ata/pata_optidma.c b/drivers/ata/pata_optidma.c
index 3f9d03599f23..7495758a86fe 100644
--- a/drivers/ata/pata_optidma.c
+++ b/drivers/ata/pata_optidma.c
@@ -350,21 +350,7 @@ static int optidma_set_mode(struct ata_link *link, struct ata_device **r_failed)
 }
 
 static struct scsi_host_template optidma_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations optidma_port_ops = {
diff --git a/drivers/ata/pata_pcmcia.c b/drivers/ata/pata_pcmcia.c
index 9881a9e004a4..c05b36c94d51 100644
--- a/drivers/ata/pata_pcmcia.c
+++ b/drivers/ata/pata_pcmcia.c
@@ -128,21 +128,7 @@ static unsigned int ata_data_xfer_8bit(struct ata_device *dev,
 
 
 static struct scsi_host_template pcmcia_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations pcmcia_port_ops = {
diff --git a/drivers/ata/pata_pdc2027x.c b/drivers/ata/pata_pdc2027x.c
index 511c89b9bae8..229d9acd934a 100644
--- a/drivers/ata/pata_pdc2027x.c
+++ b/drivers/ata/pata_pdc2027x.c
@@ -129,21 +129,7 @@ static struct pci_driver pdc2027x_pci_driver = {
 };
 
 static struct scsi_host_template pdc2027x_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations pdc2027x_pata100_ops = {
diff --git a/drivers/ata/pata_pdc202xx_old.c b/drivers/ata/pata_pdc202xx_old.c
index 3ed866723e0c..564ee0798ec1 100644
--- a/drivers/ata/pata_pdc202xx_old.c
+++ b/drivers/ata/pata_pdc202xx_old.c
@@ -262,21 +262,7 @@ static int pdc2026x_check_atapi_dma(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template pdc202xx_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations pdc2024x_port_ops = {
diff --git a/drivers/ata/pata_platform.c b/drivers/ata/pata_platform.c
index 602f5562d6fb..cd2d03a4591a 100644
--- a/drivers/ata/pata_platform.c
+++ b/drivers/ata/pata_platform.c
@@ -47,21 +47,7 @@ static int pata_platform_set_mode(struct ata_link *link, struct ata_device **unu
 }
 
 static struct scsi_host_template pata_platform_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations pata_platform_port_ops = {
diff --git a/drivers/ata/pata_qdi.c b/drivers/ata/pata_qdi.c
index 60238d5748a7..ccb8682300b8 100644
--- a/drivers/ata/pata_qdi.c
+++ b/drivers/ata/pata_qdi.c
@@ -154,21 +154,7 @@ static unsigned int qdi_data_xfer(struct ata_device *dev, unsigned char *buf,
 }
 
 static struct scsi_host_template qdi_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations qdi6500_port_ops = {
diff --git a/drivers/ata/pata_radisys.c b/drivers/ata/pata_radisys.c
index 8109b08fc024..3981bf84d093 100644
--- a/drivers/ata/pata_radisys.c
+++ b/drivers/ata/pata_radisys.c
@@ -185,21 +185,7 @@ static unsigned int radisys_qc_issue_prot(struct ata_queued_cmd *qc)
 
 
 static struct scsi_host_template radisys_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations radisys_pata_ops = {
diff --git a/drivers/ata/pata_rb500_cf.c b/drivers/ata/pata_rb500_cf.c
index 22cb9e1a02f5..4543c980342c 100644
--- a/drivers/ata/pata_rb500_cf.c
+++ b/drivers/ata/pata_rb500_cf.c
@@ -142,22 +142,7 @@ static struct ata_port_operations rb500_pata_port_ops = {
 /* ------------------------------------------------------------------------ */
 
 static struct scsi_host_template rb500_pata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
-	.proc_name		= DRV_NAME,
-
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 /* ------------------------------------------------------------------------ */
diff --git a/drivers/ata/pata_rz1000.c b/drivers/ata/pata_rz1000.c
index 75b252111106..80909a607d36 100644
--- a/drivers/ata/pata_rz1000.c
+++ b/drivers/ata/pata_rz1000.c
@@ -53,21 +53,7 @@ static int rz1000_set_mode(struct ata_link *link, struct ata_device **unused)
 
 
 static struct scsi_host_template rz1000_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations rz1000_port_ops = {
diff --git a/drivers/ata/pata_sc1200.c b/drivers/ata/pata_sc1200.c
index 725a8586cd6e..38ce6e12ee3d 100644
--- a/drivers/ata/pata_sc1200.c
+++ b/drivers/ata/pata_sc1200.c
@@ -179,21 +179,8 @@ static unsigned int sc1200_qc_issue_prot(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template sc1200_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_DUMB_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
+	.sg_tablesize	= LIBATA_DUMB_MAX_PRD,
 };
 
 static struct ata_port_operations sc1200_port_ops = {
diff --git a/drivers/ata/pata_scc.c b/drivers/ata/pata_scc.c
index 6c016deeaed8..1833e9ef522e 100644
--- a/drivers/ata/pata_scc.c
+++ b/drivers/ata/pata_scc.c
@@ -968,21 +968,7 @@ static void scc_port_stop (struct ata_port *ap)
 }
 
 static struct scsi_host_template scc_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations scc_pata_ops = {
diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
index 6702df37cfed..318a36988900 100644
--- a/drivers/ata/pata_serverworks.c
+++ b/drivers/ata/pata_serverworks.c
@@ -298,21 +298,7 @@ static void serverworks_set_dmamode(struct ata_port *ap, struct ata_device *adev
 }
 
 static struct scsi_host_template serverworks_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations serverworks_osb4_port_ops = {
diff --git a/drivers/ata/pata_sil680.c b/drivers/ata/pata_sil680.c
index f4dc09718cab..7812815a34c3 100644
--- a/drivers/ata/pata_sil680.c
+++ b/drivers/ata/pata_sil680.c
@@ -192,21 +192,7 @@ static void sil680_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template sil680_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations sil680_port_ops = {
diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
index abda90f51247..dcd8457a8377 100644
--- a/drivers/ata/pata_sis.c
+++ b/drivers/ata/pata_sis.c
@@ -514,21 +514,7 @@ static void sis_133_set_dmamode (struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template sis_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations sis_133_ops = {
diff --git a/drivers/ata/pata_sl82c105.c b/drivers/ata/pata_sl82c105.c
index 6c37181341ea..ece366bced0c 100644
--- a/drivers/ata/pata_sl82c105.c
+++ b/drivers/ata/pata_sl82c105.c
@@ -235,21 +235,7 @@ static int sl82c105_qc_defer(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template sl82c105_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations sl82c105_port_ops = {
diff --git a/drivers/ata/pata_triflex.c b/drivers/ata/pata_triflex.c
index 403eafcffe12..510569957d10 100644
--- a/drivers/ata/pata_triflex.c
+++ b/drivers/ata/pata_triflex.c
@@ -180,21 +180,7 @@ static void triflex_bmdma_stop(struct ata_queued_cmd *qc)
 }
 
 static struct scsi_host_template triflex_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations triflex_port_ops = {
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
index 24430f70f00e..a7bc860e1310 100644
--- a/drivers/ata/pata_via.c
+++ b/drivers/ata/pata_via.c
@@ -335,21 +335,7 @@ static void via_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct scsi_host_template via_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations via_port_ops = {
diff --git a/drivers/ata/pata_winbond.c b/drivers/ata/pata_winbond.c
index 5318248782bb..9bafae9d5fe0 100644
--- a/drivers/ata/pata_winbond.c
+++ b/drivers/ata/pata_winbond.c
@@ -122,21 +122,7 @@ static unsigned int winbond_data_xfer(struct ata_device *dev,
 }
 
 static struct scsi_host_template winbond_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_PIO_SHT(DRV_NAME),
 };
 
 static struct ata_port_operations winbond_port_ops = {
diff --git a/drivers/ata/pdc_adma.c b/drivers/ata/pdc_adma.c
index bc2d12a2da30..fdf62de57cfc 100644
--- a/drivers/ata/pdc_adma.c
+++ b/drivers/ata/pdc_adma.c
@@ -143,21 +143,9 @@ static void adma_thaw(struct ata_port *ap);
 static void adma_error_handler(struct ata_port *ap);
 
 static struct scsi_host_template adma_ata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
-	.proc_name		= DRV_NAME,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BASE_SHT(DRV_NAME),
 	.sg_tablesize		= LIBATA_MAX_PRD,
 	.dma_boundary		= ADMA_DMA_BOUNDARY,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.use_clustering		= ENABLE_CLUSTERING,
-	.emulated		= ATA_SHT_EMULATED,
 };
 
 static const struct ata_port_operations adma_ata_ops = {
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index 8389ecac62cd..f50381b4ba06 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -1191,22 +1191,10 @@ static int sata_fsl_init_controller(struct ata_host *host)
  * scsi mid-layer and libata interface structures
  */
 static struct scsi_host_template sata_fsl_sht = {
-	.module = THIS_MODULE,
-	.name = "sata_fsl",
-	.ioctl = ata_scsi_ioctl,
-	.queuecommand = ata_scsi_queuecmd,
-	.change_queue_depth = ata_scsi_change_queue_depth,
+	ATA_NCQ_SHT("sata_fsl"),
 	.can_queue = SATA_FSL_QUEUE_DEPTH,
-	.this_id = ATA_SHT_THIS_ID,
 	.sg_tablesize = SATA_FSL_MAX_PRD_USABLE,
-	.cmd_per_lun = ATA_SHT_CMD_PER_LUN,
-	.emulated = ATA_SHT_EMULATED,
-	.use_clustering = ATA_SHT_USE_CLUSTERING,
-	.proc_name = "sata_fsl",
 	.dma_boundary = ATA_DMA_BOUNDARY,
-	.slave_configure = ata_scsi_slave_config,
-	.slave_destroy = ata_scsi_slave_destroy,
-	.bios_param = ata_std_bios_param,
 };
 
 static const struct ata_port_operations sata_fsl_ops = {
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index 74f14369dc8d..bb853df865da 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -109,21 +109,7 @@ struct inic_port_priv {
 };
 
 static struct scsi_host_template inic_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const int scr_map[] = {
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 4685bce745bb..52d41edadb72 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -514,40 +514,16 @@ static int __mv_stop_dma(struct ata_port *ap);
  * PRDs for 64K boundaries in mv_fill_sg().
  */
 static struct scsi_host_template mv5_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BASE_SHT(DRV_NAME),
 	.sg_tablesize		= MV_MAX_SG_CT / 2,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= 1,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= MV_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static struct scsi_host_template mv6_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.change_queue_depth	= ata_scsi_change_queue_depth,
+	ATA_NCQ_SHT(DRV_NAME),
 	.can_queue		= MV_MAX_Q_DEPTH - 1,
-	.this_id		= ATA_SHT_THIS_ID,
 	.sg_tablesize		= MV_MAX_SG_CT / 2,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= 1,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= MV_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static const struct ata_port_operations mv5_ops = {
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index bd5b6c35ee5d..9e2b4cef48f2 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -385,59 +385,23 @@ static struct pci_driver nv_pci_driver = {
 };
 
 static struct scsi_host_template nv_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static struct scsi_host_template nv_adma_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.change_queue_depth	= ata_scsi_change_queue_depth,
+	ATA_NCQ_SHT(DRV_NAME),
 	.can_queue		= NV_ADMA_MAX_CPBS,
-	.this_id		= ATA_SHT_THIS_ID,
 	.sg_tablesize		= NV_ADMA_SGTBL_TOTAL_LEN,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= NV_ADMA_DMA_BOUNDARY,
 	.slave_configure	= nv_adma_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static struct scsi_host_template nv_swncq_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.change_queue_depth	= ata_scsi_change_queue_depth,
+	ATA_NCQ_SHT(DRV_NAME),
 	.can_queue		= ATA_MAX_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
 	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= ATA_DMA_BOUNDARY,
 	.slave_configure	= nv_swncq_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static const struct ata_port_operations nv_generic_ops = {
diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c
index 11c1afea2db2..37c32ab3b23b 100644
--- a/drivers/ata/sata_promise.c
+++ b/drivers/ata/sata_promise.c
@@ -155,21 +155,9 @@ static int pdc_pata_cable_detect(struct ata_port *ap);
 static int pdc_sata_cable_detect(struct ata_port *ap);
 
 static struct scsi_host_template pdc_ata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BASE_SHT(DRV_NAME),
 	.sg_tablesize		= PDC_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static const struct ata_port_operations pdc_sata_ops = {
diff --git a/drivers/ata/sata_qstor.c b/drivers/ata/sata_qstor.c
index 3c8e97f251f9..2566d0926aab 100644
--- a/drivers/ata/sata_qstor.c
+++ b/drivers/ata/sata_qstor.c
@@ -126,21 +126,9 @@ static void qs_thaw(struct ata_port *ap);
 static void qs_error_handler(struct ata_port *ap);
 
 static struct scsi_host_template qs_ata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BASE_SHT(DRV_NAME),
 	.sg_tablesize		= QS_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ENABLE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= QS_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static const struct ata_port_operations qs_ata_ops = {
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index 53f0bae3be43..738c1a8ae3b6 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -167,21 +167,7 @@ static struct pci_driver sil_pci_driver = {
 };
 
 static struct scsi_host_template sil_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations sil_ops = {
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index b85464d51f68..7fa63950d81a 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -384,22 +384,10 @@ static struct pci_driver sil24_pci_driver = {
 };
 
 static struct scsi_host_template sil24_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.change_queue_depth	= ata_scsi_change_queue_depth,
+	ATA_NCQ_SHT(DRV_NAME),
 	.can_queue		= SIL24_MAX_CMDS,
-	.this_id		= ATA_SHT_THIS_ID,
 	.sg_tablesize		= SIL24_MAX_SGE,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static const struct ata_port_operations sil24_ops = {
diff --git a/drivers/ata/sata_sis.c b/drivers/ata/sata_sis.c
index a8adef9786b7..4becb7fde5e7 100644
--- a/drivers/ata/sata_sis.c
+++ b/drivers/ata/sata_sis.c
@@ -86,21 +86,7 @@ static struct pci_driver sis_pci_driver = {
 };
 
 static struct scsi_host_template sis_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations sis_ops = {
diff --git a/drivers/ata/sata_svw.c b/drivers/ata/sata_svw.c
index aa690142fa90..c8768396e006 100644
--- a/drivers/ata/sata_svw.c
+++ b/drivers/ata/sata_svw.c
@@ -327,24 +327,10 @@ static int k2_sata_proc_info(struct Scsi_Host *shost, char *page, char **start,
 
 
 static struct scsi_host_template k2_sata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
+	ATA_BMDMA_SHT(DRV_NAME),
 #ifdef CONFIG_PPC_OF
 	.proc_info		= k2_sata_proc_info,
 #endif
-	.bios_param		= ata_std_bios_param,
 };
 
 
diff --git a/drivers/ata/sata_sx4.c b/drivers/ata/sata_sx4.c
index e3d56bc6726d..1802f92180e4 100644
--- a/drivers/ata/sata_sx4.c
+++ b/drivers/ata/sata_sx4.c
@@ -236,21 +236,9 @@ static unsigned int pdc20621_qc_issue_prot(struct ata_queued_cmd *qc);
 
 
 static struct scsi_host_template pdc_sata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
+	ATA_BASE_SHT(DRV_NAME),
 	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
 	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
 };
 
 static const struct ata_port_operations pdc_20621_ops = {
diff --git a/drivers/ata/sata_uli.c b/drivers/ata/sata_uli.c
index f7fc0450478b..764d7064fa59 100644
--- a/drivers/ata/sata_uli.c
+++ b/drivers/ata/sata_uli.c
@@ -76,21 +76,7 @@ static struct pci_driver uli_pci_driver = {
 };
 
 static struct scsi_host_template uli_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations uli_ops = {
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index f66ffd7da54e..9be877cb7f57 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -100,21 +100,7 @@ static struct pci_driver svia_pci_driver = {
 };
 
 static struct scsi_host_template svia_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 static const struct ata_port_operations vt6420_sata_ops = {
diff --git a/drivers/ata/sata_vsc.c b/drivers/ata/sata_vsc.c
index 099a2ba4cd4f..fd6855f0bf48 100644
--- a/drivers/ata/sata_vsc.c
+++ b/drivers/ata/sata_vsc.c
@@ -300,21 +300,7 @@ out:
 
 
 static struct scsi_host_template vsc_sata_sht = {
-	.module			= THIS_MODULE,
-	.name			= DRV_NAME,
-	.ioctl			= ata_scsi_ioctl,
-	.queuecommand		= ata_scsi_queuecmd,
-	.can_queue		= ATA_DEF_QUEUE,
-	.this_id		= ATA_SHT_THIS_ID,
-	.sg_tablesize		= LIBATA_MAX_PRD,
-	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,
-	.emulated		= ATA_SHT_EMULATED,
-	.use_clustering		= ATA_SHT_USE_CLUSTERING,
-	.proc_name		= DRV_NAME,
-	.dma_boundary		= ATA_DMA_BOUNDARY,
-	.slave_configure	= ata_scsi_slave_config,
-	.slave_destroy		= ata_scsi_slave_destroy,
-	.bios_param		= ata_std_bios_param,
+	ATA_BMDMA_SHT(DRV_NAME),
 };
 
 
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 639298af583e..eccc38e17568 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1070,6 +1070,63 @@ extern void ata_do_eh(struct ata_port *ap, ata_prereset_fn_t prereset,
 		      ata_reset_fn_t softreset, ata_reset_fn_t hardreset,
 		      ata_postreset_fn_t postreset);
 
+/*
+ * Base operations to inherit from and initializers for sht
+ *
+ * Operations
+ *
+ * base  : Common to all libata drivers.
+ * sata  : SATA controllers w/ native interface.
+ * pmp   : SATA controllers w/ PMP support.
+ * sff   : SFF ATA controllers w/o BMDMA support.
+ * bmdma : SFF ATA controllers w/ BMDMA support.
+ *
+ * sht initializers
+ *
+ * BASE  : Common to all libata drivers.  The user must set
+ *	   sg_tablesize and dma_boundary.
+ * PIO   : SFF ATA controllers w/ only PIO support.
+ * BMDMA : SFF ATA controllers w/ BMDMA support.  sg_tablesize and
+ *	   dma_boundary are set to BMDMA limits.
+ * NCQ   : SATA controllers supporting NCQ.  The user must set
+ *	   sg_tablesize, dma_boundary and can_queue.
+ */
+extern const struct ata_port_operations ata_base_port_ops;
+extern const struct ata_port_operations sata_port_ops;
+extern const struct ata_port_operations sata_pmp_port_ops;
+extern const struct ata_port_operations ata_sff_port_ops;
+extern const struct ata_port_operations ata_bmdma_port_ops;
+
+#define ATA_BASE_SHT(drv_name)					\
+	.module			= THIS_MODULE,			\
+	.name			= drv_name,			\
+	.ioctl			= ata_scsi_ioctl,		\
+	.queuecommand		= ata_scsi_queuecmd,		\
+	.can_queue		= ATA_DEF_QUEUE,		\
+	.this_id		= ATA_SHT_THIS_ID,		\
+	.cmd_per_lun		= ATA_SHT_CMD_PER_LUN,		\
+	.emulated		= ATA_SHT_EMULATED,		\
+	.use_clustering		= ATA_SHT_USE_CLUSTERING,	\
+	.proc_name		= drv_name,			\
+	.slave_configure	= ata_scsi_slave_config,	\
+	.slave_destroy		= ata_scsi_slave_destroy,	\
+	.bios_param		= ata_std_bios_param
+
+/* PIO only, sg_tablesize and dma_boundary limits can be removed */
+#define ATA_PIO_SHT(drv_name)					\
+	ATA_BASE_SHT(drv_name),					\
+	.sg_tablesize		= LIBATA_MAX_PRD,		\
+	.dma_boundary		= ATA_DMA_BOUNDARY
+
+#define ATA_BMDMA_SHT(drv_name)					\
+	ATA_BASE_SHT(drv_name),					\
+	.sg_tablesize		= LIBATA_MAX_PRD,		\
+	.dma_boundary		= ATA_DMA_BOUNDARY
+
+#define ATA_NCQ_SHT(drv_name)					\
+	ATA_BASE_SHT(drv_name),					\
+	.change_queue_depth	= ata_scsi_change_queue_depth
+
 /*
  * printk helpers
  */
-- 
cgit v1.2.3


From 029cfd6b74fc5c517865fad78cf4a3ea8d9b664a Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:49 +0900
Subject: libata: implement and use ops inheritance

libata lets low level drivers build ata_port_operations table and
register it with libata core layer.  This allows low level drivers
high level of flexibility but also burdens them with lots of
boilerplate entries.

This becomes worse for drivers which support related similar
controllers which differ slightly.  They share most of the operations
except for a few.  However, the driver still needs to list all
operations for each variant.  This results in large number of
duplicate entries, which is not only inefficient but also error-prone
as it becomes very difficult to tell what the actual differences are.

This duplicate boilerplates all over the low level drivers also make
updating the core layer exteremely difficult and error-prone.  When
compounded with multi-branched development model, it ends up
accumulating inconsistencies over time.  Some of those inconsistencies
cause immediate problems and fixed.  Others just remain there dormant
making maintenance increasingly difficult.

To rectify the problem, this patch implements ata_port_operations
inheritance.  To allow LLDs to easily re-use their own ops tables
overriding only specific methods, this patch implements poor man's
class inheritance.  An ops table has ->inherits field which can be set
to any ops table as long as it doesn't create a loop.  When the host
is started, the inheritance chain is followed and any operation which
isn't specified is taken from the nearest ancestor which has it
specified.  This operation is called finalization and done only once
per an ops table and the LLD doesn't have to do anything special about
it other than making the ops table non-const such that libata can
update it.

libata provides four base ops tables lower drivers can inherit from -
base, sata, pmp, sff and bmdma.  To avoid overriding these ops
accidentaly, these ops are declared const and LLDs should always
inherit these instead of using them directly.

After finalization, all the ops table are identical before and after
the patch except for setting .irq_handler to ata_interrupt in drivers
which didn't use to.  The .irq_handler doesn't have any actual effect
and the field will soon be removed by later patch.

* sata_sx4 is still using old style EH and currently doesn't take
  advantage of ops inheritance.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c              |  94 ++-------------
 drivers/ata/ata_generic.c       |  31 +----
 drivers/ata/ata_piix.c          | 149 +++--------------------
 drivers/ata/libata-core.c       | 113 ++++++++++++++++-
 drivers/ata/pata_acpi.c         |  35 +-----
 drivers/ata/pata_ali.c          | 122 ++-----------------
 drivers/ata/pata_amd.c          | 191 ++++-------------------------
 drivers/ata/pata_artop.c        |  61 +---------
 drivers/ata/pata_at32.c         |  23 +---
 drivers/ata/pata_atiixp.c       |  32 +----
 drivers/ata/pata_bf54x.c        |   4 +-
 drivers/ata/pata_cmd640.c       |  33 +----
 drivers/ata/pata_cmd64x.c       |  96 ++-------------
 drivers/ata/pata_cs5520.c       |  29 +----
 drivers/ata/pata_cs5530.c       |  31 +----
 drivers/ata/pata_cs5535.c       |  31 +----
 drivers/ata/pata_cs5536.c       |  31 +----
 drivers/ata/pata_cypress.c      |  31 +----
 drivers/ata/pata_efar.c         |  30 +----
 drivers/ata/pata_hpt366.c       |  32 +----
 drivers/ata/pata_hpt37x.c       | 122 ++-----------------
 drivers/ata/pata_hpt3x2n.c      |  32 +----
 drivers/ata/pata_hpt3x3.c       |  33 +----
 drivers/ata/pata_icside.c       |  32 ++---
 drivers/ata/pata_isapnp.c       |  21 +---
 drivers/ata/pata_it8213.c       |  31 +----
 drivers/ata/pata_it821x.c       |  57 ++-------
 drivers/ata/pata_ixp4xx_cf.c    |  23 +---
 drivers/ata/pata_jmicron.c      |  31 +----
 drivers/ata/pata_legacy.c       | 260 +++-------------------------------------
 drivers/ata/pata_marvell.c      |  33 +----
 drivers/ata/pata_mpc52xx.c      |  18 +--
 drivers/ata/pata_mpiix.c        |  23 +---
 drivers/ata/pata_netcell.c      |  32 +----
 drivers/ata/pata_ninja32.c      |  30 +----
 drivers/ata/pata_ns87410.c      |  25 +---
 drivers/ata/pata_ns87415.c      |  68 ++---------
 drivers/ata/pata_oldpiix.c      |  31 +----
 drivers/ata/pata_opti.c         |  23 +---
 drivers/ata/pata_optidma.c      |  63 +---------
 drivers/ata/pata_pcmcia.c       |  48 +-------
 drivers/ata/pata_pdc2027x.c     |  56 +--------
 drivers/ata/pata_pdc202xx_old.c |  75 +++---------
 drivers/ata/pata_platform.c     |  24 +---
 drivers/ata/pata_qdi.c          |  47 +-------
 drivers/ata/pata_radisys.c      |  32 +----
 drivers/ata/pata_rb500_cf.c     |  16 +--
 drivers/ata/pata_rz1000.c       |  25 +---
 drivers/ata/pata_sc1200.c       |  33 +----
 drivers/ata/pata_scc.c          |   7 +-
 drivers/ata/pata_serverworks.c  |  63 +---------
 drivers/ata/pata_sil680.c       |  30 +----
 drivers/ata/pata_sis.c          | 177 +++------------------------
 drivers/ata/pata_sl82c105.c     |  34 +-----
 drivers/ata/pata_triflex.c      |  32 +----
 drivers/ata/pata_via.c          |  61 +---------
 drivers/ata/pata_winbond.c      |  25 +---
 drivers/ata/pdc_adma.c          |  19 +--
 drivers/ata/sata_fsl.c          |   5 +-
 drivers/ata/sata_inic162x.c     |  19 +--
 drivers/ata/sata_mv.c           |  66 ++--------
 drivers/ata/sata_nv.c           | 113 +++--------------
 drivers/ata/sata_promise.c      |  62 +++-------
 drivers/ata/sata_qstor.c        |  20 ++--
 drivers/ata/sata_sil.c          |  21 +---
 drivers/ata/sata_sil24.c        |  20 ++--
 drivers/ata/sata_sis.c          |  23 +---
 drivers/ata/sata_svw.c          |  18 +--
 drivers/ata/sata_sx4.c          |   3 +-
 drivers/ata/sata_uli.c          |  28 +----
 drivers/ata/sata_via.c          |  88 ++------------
 drivers/ata/sata_vsc.c          |  19 +--
 include/linux/libata.h          |  23 +++-
 73 files changed, 523 insertions(+), 2996 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 8862595cb2cf..dacb3ef0c3e6 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -280,118 +280,46 @@ static struct scsi_host_template ahci_sht = {
 	.shost_attrs		= ahci_shost_attrs,
 };
 
-static const struct ata_port_operations ahci_ops = {
+static struct ata_port_operations ahci_ops = {
+	.inherits		= &sata_pmp_port_ops,
+
 	.check_status		= ahci_check_status,
 	.check_altstatus	= ahci_check_status,
-	.dev_select		= ata_noop_dev_select,
-
-	.dev_config		= ahci_dev_config,
 
 	.tf_read		= ahci_tf_read,
-
 	.qc_defer		= sata_pmp_qc_defer_cmd_switch,
 	.qc_prep		= ahci_qc_prep,
 	.qc_issue		= ahci_qc_issue,
 
-	.irq_clear		= ata_noop_irq_clear,
-
-	.scr_read		= ahci_scr_read,
-	.scr_write		= ahci_scr_write,
-
 	.freeze			= ahci_freeze,
 	.thaw			= ahci_thaw,
-
 	.error_handler		= ahci_error_handler,
 	.post_internal_cmd	= ahci_post_internal_cmd,
-
-	.pmp_attach		= ahci_pmp_attach,
-	.pmp_detach		= ahci_pmp_detach,
-
-#ifdef CONFIG_PM
-	.port_suspend		= ahci_port_suspend,
-	.port_resume		= ahci_port_resume,
-#endif
-	.enable_pm		= ahci_enable_alpm,
-	.disable_pm		= ahci_disable_alpm,
-
-	.port_start		= ahci_port_start,
-	.port_stop		= ahci_port_stop,
-};
-
-static const struct ata_port_operations ahci_vt8251_ops = {
-	.check_status		= ahci_check_status,
-	.check_altstatus	= ahci_check_status,
-	.dev_select		= ata_noop_dev_select,
-
 	.dev_config		= ahci_dev_config,
 
-	.tf_read		= ahci_tf_read,
-
-	.qc_defer		= sata_pmp_qc_defer_cmd_switch,
-	.qc_prep		= ahci_qc_prep,
-	.qc_issue		= ahci_qc_issue,
-
-	.irq_clear		= ata_noop_irq_clear,
-
 	.scr_read		= ahci_scr_read,
 	.scr_write		= ahci_scr_write,
-
-	.freeze			= ahci_freeze,
-	.thaw			= ahci_thaw,
-
-	.error_handler		= ahci_vt8251_error_handler,
-	.post_internal_cmd	= ahci_post_internal_cmd,
-
 	.pmp_attach		= ahci_pmp_attach,
 	.pmp_detach		= ahci_pmp_detach,
 
+	.enable_pm		= ahci_enable_alpm,
+	.disable_pm		= ahci_disable_alpm,
 #ifdef CONFIG_PM
 	.port_suspend		= ahci_port_suspend,
 	.port_resume		= ahci_port_resume,
 #endif
-	.enable_pm		= ahci_enable_alpm,
-	.disable_pm		= ahci_disable_alpm,
-
 	.port_start		= ahci_port_start,
 	.port_stop		= ahci_port_stop,
 };
 
-static const struct ata_port_operations ahci_p5wdh_ops = {
-	.check_status		= ahci_check_status,
-	.check_altstatus	= ahci_check_status,
-	.dev_select		= ata_noop_dev_select,
-
-	.dev_config		= ahci_dev_config,
-
-	.tf_read		= ahci_tf_read,
-
-	.qc_defer		= sata_pmp_qc_defer_cmd_switch,
-	.qc_prep		= ahci_qc_prep,
-	.qc_issue		= ahci_qc_issue,
-
-	.irq_clear		= ata_noop_irq_clear,
-
-	.scr_read		= ahci_scr_read,
-	.scr_write		= ahci_scr_write,
-
-	.freeze			= ahci_freeze,
-	.thaw			= ahci_thaw,
+static struct ata_port_operations ahci_vt8251_ops = {
+	.inherits		= &ahci_ops,
+	.error_handler		= ahci_vt8251_error_handler,
+};
 
+static struct ata_port_operations ahci_p5wdh_ops = {
+	.inherits		= &ahci_ops,
 	.error_handler		= ahci_p5wdh_error_handler,
-	.post_internal_cmd	= ahci_post_internal_cmd,
-
-	.pmp_attach		= ahci_pmp_attach,
-	.pmp_detach		= ahci_pmp_detach,
-
-#ifdef CONFIG_PM
-	.port_suspend		= ahci_port_suspend,
-	.port_resume		= ahci_port_resume,
-#endif
-	.enable_pm		= ahci_enable_alpm,
-	.disable_pm		= ahci_disable_alpm,
-
-	.port_start		= ahci_port_start,
-	.port_stop		= ahci_port_stop,
 };
 
 #define AHCI_HFLAGS(flags)	.private_data	= (void *)(flags)
diff --git a/drivers/ata/ata_generic.c b/drivers/ata/ata_generic.c
index 5c64ce134c6c..0b5b515ae159 100644
--- a/drivers/ata/ata_generic.c
+++ b/drivers/ata/ata_generic.c
@@ -99,36 +99,9 @@ static struct scsi_host_template generic_sht = {
 };
 
 static struct ata_port_operations generic_port_ops = {
-	.set_mode	= generic_set_mode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.data_xfer	= ata_data_xfer,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.inherits	= &ata_bmdma_port_ops,
 	.cable_detect	= ata_cable_unknown,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.set_mode	= generic_set_mode,
 };
 
 static int all_generic_ide;		/* Set to claim all devices */
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index 9f887b2c92df..bb46b61a7c6b 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -294,155 +294,34 @@ static struct scsi_host_template piix_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations piix_pata_ops = {
+static struct ata_port_operations piix_pata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= piix_set_piomode,
 	.set_dmamode		= piix_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= piix_pata_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_40wire,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
+};
 
-	.port_start		= ata_sff_port_start,
+static struct ata_port_operations piix_vmw_ops = {
+	.inherits		= &piix_pata_ops,
+	.bmdma_status		= piix_vmw_bmdma_status,
 };
 
-static const struct ata_port_operations ich_pata_ops = {
-	.set_piomode		= piix_set_piomode,
-	.set_dmamode		= ich_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= piix_pata_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+static struct ata_port_operations ich_pata_ops = {
+	.inherits		= &piix_pata_ops,
 	.cable_detect		= ich_pata_cable_detect,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
+	.set_dmamode		= ich_set_dmamode,
 };
 
-static const struct ata_port_operations piix_sata_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
+static struct ata_port_operations piix_sata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 };
 
-static const struct ata_port_operations piix_vmw_ops = {
-	.set_piomode		= piix_set_piomode,
-	.set_dmamode		= piix_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= piix_vmw_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= piix_pata_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_40wire,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
-};
-
-static const struct ata_port_operations piix_sidpr_sata_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
+static struct ata_port_operations piix_sidpr_sata_ops = {
+	.inherits		= &piix_sata_ops,
 	.scr_read		= piix_sidpr_scr_read,
 	.scr_write		= piix_sidpr_scr_write,
-
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= piix_sidpr_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 static const struct piix_map_db ich5_map_db = {
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 394edf937cf2..32fa9ee397b6 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -74,6 +74,56 @@ const unsigned long sata_deb_timing_normal[]		= {   5,  100, 2000 };
 const unsigned long sata_deb_timing_hotplug[]		= {  25,  500, 2000 };
 const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
+const struct ata_port_operations ata_base_port_ops = {
+	.irq_clear		= ata_noop_irq_clear,
+};
+
+const struct ata_port_operations sata_port_ops = {
+	.inherits		= &ata_base_port_ops,
+
+	.qc_defer		= ata_std_qc_defer,
+	.dev_select		= ata_noop_dev_select,
+};
+
+const struct ata_port_operations sata_pmp_port_ops = {
+	.inherits		= &sata_port_ops,
+};
+
+const struct ata_port_operations ata_sff_port_ops = {
+	.inherits		= &ata_base_port_ops,
+
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.error_handler		= ata_bmdma_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.dev_select		= ata_std_dev_select,
+	.check_status		= ata_check_status,
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.exec_command		= ata_exec_command,
+	.data_xfer		= ata_data_xfer,
+	.irq_on			= ata_irq_on,
+
+	.port_start		= ata_sff_port_start,
+	.irq_handler		= ata_interrupt,
+};
+
+const struct ata_port_operations ata_bmdma_port_ops = {
+	.inherits		= &ata_sff_port_ops,
+
+	.mode_filter		= ata_pci_default_filter,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.irq_clear		= ata_bmdma_irq_clear,
+};
+
 static unsigned int ata_dev_init_params(struct ata_device *dev,
 					u16 heads, u16 sectors);
 static unsigned int ata_dev_set_xfermode(struct ata_device *dev);
@@ -6971,6 +7021,56 @@ static void ata_host_stop(struct device *gendev, void *res)
 		host->ops->host_stop(host);
 }
 
+/**
+ *	ata_finalize_port_ops - finalize ata_port_operations
+ *	@ops: ata_port_operations to finalize
+ *
+ *	An ata_port_operations can inherit from another ops and that
+ *	ops can again inherit from another.  This can go on as many
+ *	times as necessary as long as there is no loop in the
+ *	inheritance chain.
+ *
+ *	Ops tables are finalized when the host is started.  NULL or
+ *	unspecified entries are inherited from the closet ancestor
+ *	which has the method and the entry is populated with it.
+ *	After finalization, the ops table directly points to all the
+ *	methods and ->inherits is no longer necessary and cleared.
+ *
+ *	Using ATA_OP_NULL, inheriting ops can force a method to NULL.
+ *
+ *	LOCKING:
+ *	None.
+ */
+static void ata_finalize_port_ops(struct ata_port_operations *ops)
+{
+	static spinlock_t lock = SPIN_LOCK_UNLOCKED;
+	const struct ata_port_operations *cur;
+	void **begin = (void **)ops;
+	void **end = (void **)&ops->inherits;
+	void **pp;
+
+	if (!ops || !ops->inherits)
+		return;
+
+	spin_lock(&lock);
+
+	for (cur = ops->inherits; cur; cur = cur->inherits) {
+		void **inherit = (void **)cur;
+
+		for (pp = begin; pp < end; pp++, inherit++)
+			if (!*pp)
+				*pp = *inherit;
+	}
+
+	for (pp = begin; pp < end; pp++)
+		if (IS_ERR(*pp))
+			*pp = NULL;
+
+	ops->inherits = NULL;
+
+	spin_unlock(&lock);
+}
+
 /**
  *	ata_host_start - start and freeze ports of an ATA host
  *	@host: ATA host to start ports for
@@ -6996,9 +7096,13 @@ int ata_host_start(struct ata_host *host)
 	if (host->flags & ATA_HOST_STARTED)
 		return 0;
 
+	ata_finalize_port_ops(host->ops);
+
 	for (i = 0; i < host->n_ports; i++) {
 		struct ata_port *ap = host->ports[i];
 
+		ata_finalize_port_ops(ap->ops);
+
 		if (!host->ops && !ata_port_is_dummy(ap))
 			host->ops = ap->ops;
 
@@ -7060,7 +7164,7 @@ int ata_host_start(struct ata_host *host)
  */
 /* KILLME - the only user left is ipr */
 void ata_host_init(struct ata_host *host, struct device *dev,
-		   unsigned long flags, const struct ata_port_operations *ops)
+		   unsigned long flags, struct ata_port_operations *ops)
 {
 	spin_lock_init(&host->lock);
 	host->dev = dev;
@@ -7749,7 +7853,7 @@ static unsigned int ata_dummy_qc_issue(struct ata_queued_cmd *qc)
 	return AC_ERR_SYSTEM;
 }
 
-const struct ata_port_operations ata_dummy_port_ops = {
+struct ata_port_operations ata_dummy_port_ops = {
 	.check_status		= ata_dummy_check_status,
 	.check_altstatus	= ata_dummy_check_status,
 	.dev_select		= ata_noop_dev_select,
@@ -7777,6 +7881,11 @@ const struct ata_port_info ata_dummy_port_info = {
 EXPORT_SYMBOL_GPL(sata_deb_timing_normal);
 EXPORT_SYMBOL_GPL(sata_deb_timing_hotplug);
 EXPORT_SYMBOL_GPL(sata_deb_timing_long);
+EXPORT_SYMBOL_GPL(ata_base_port_ops);
+EXPORT_SYMBOL_GPL(sata_port_ops);
+EXPORT_SYMBOL_GPL(sata_pmp_port_ops);
+EXPORT_SYMBOL_GPL(ata_sff_port_ops);
+EXPORT_SYMBOL_GPL(ata_bmdma_port_ops);
 EXPORT_SYMBOL_GPL(ata_dummy_port_ops);
 EXPORT_SYMBOL_GPL(ata_dummy_port_info);
 EXPORT_SYMBOL_GPL(ata_std_bios_param);
diff --git a/drivers/ata/pata_acpi.c b/drivers/ata/pata_acpi.c
index 187545c0898a..35ad488db6ed 100644
--- a/drivers/ata/pata_acpi.c
+++ b/drivers/ata/pata_acpi.c
@@ -235,39 +235,14 @@ static struct scsi_host_template pacpi_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations pacpi_ops = {
+static struct ata_port_operations pacpi_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.qc_issue		= pacpi_qc_issue_prot,
+	.cable_detect		= pacpi_cable_detect,
+	.mode_filter		= pacpi_mode_filter,
 	.set_piomode		= pacpi_set_piomode,
 	.set_dmamode		= pacpi_set_dmamode,
-	.mode_filter		= pacpi_mode_filter,
-
-	/* Task file is PCI ATA format, use helpers */
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= pacpi_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= pacpi_cable_detect,
-
-	/* BMDMA handling is PCI ATA format, use helpers */
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= pacpi_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	/* Timeout handling */
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	/* Generic PATA PCI ATA helpers */
 	.port_start		= pacpi_port_start,
 };
 
diff --git a/drivers/ata/pata_ali.c b/drivers/ata/pata_ali.c
index f3d6d9b345ba..b00a9cf72c31 100644
--- a/drivers/ata/pata_ali.c
+++ b/drivers/ata/pata_ali.c
@@ -347,29 +347,15 @@ static struct scsi_host_template ali_sht = {
  */
 
 static struct ata_port_operations ali_early_port_ops = {
-	.set_piomode	= ali_set_piomode,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.inherits	= &ata_sff_port_ops,
 	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= ali_set_piomode,
+};
 
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+static const struct ata_port_operations ali_dma_base_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.set_piomode	= ali_set_piomode,
+	.set_dmamode	= ali_set_dmamode,
 };
 
 /*
@@ -377,115 +363,31 @@ static struct ata_port_operations ali_early_port_ops = {
  *	detect
  */
 static struct ata_port_operations ali_20_port_ops = {
-	.set_piomode	= ali_set_piomode,
-	.set_dmamode	= ali_set_dmamode,
+	.inherits	= &ali_dma_base_ops,
+	.cable_detect	= ata_cable_40wire,
 	.mode_filter	= ali_20_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
 	.check_atapi_dma = ali_check_atapi_dma,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
 	.dev_config	= ali_lock_sectors,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /*
  *	Port operations for DMA capable ALi with cable detect
  */
 static struct ata_port_operations ali_c2_port_ops = {
-	.set_piomode	= ali_set_piomode,
-	.set_dmamode	= ali_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
+	.inherits	= &ali_dma_base_ops,
 	.check_atapi_dma = ali_check_atapi_dma,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-	.dev_config	= ali_lock_sectors,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
 	.cable_detect	= ali_c2_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.dev_config	= ali_lock_sectors,
 };
 
 /*
  *	Port operations for DMA capable ALi with cable detect and LBA48
  */
 static struct ata_port_operations ali_c5_port_ops = {
-	.set_piomode	= ali_set_piomode,
-	.set_dmamode	= ali_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
+	.inherits	= &ali_dma_base_ops,
 	.check_atapi_dma = ali_check_atapi_dma,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
 	.dev_config	= ali_warn_atapi_dma,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
 	.cable_detect	= ali_c2_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_amd.c b/drivers/ata/pata_amd.c
index 90d786dfbec3..b0cb4eaf273c 100644
--- a/drivers/ata/pata_amd.c
+++ b/drivers/ata/pata_amd.c
@@ -356,204 +356,57 @@ static struct scsi_host_template amd_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
+static const struct ata_port_operations amd_base_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.error_handler	= amd_error_handler,
+};
+
 static struct ata_port_operations amd33_port_ops = {
+	.inherits	= &amd_base_port_ops,
+	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= amd33_set_piomode,
 	.set_dmamode	= amd33_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= amd_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations amd66_port_ops = {
+	.inherits	= &amd_base_port_ops,
+	.cable_detect	= ata_cable_unknown,
 	.set_piomode	= amd66_set_piomode,
 	.set_dmamode	= amd66_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= amd_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_unknown,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations amd100_port_ops = {
+	.inherits	= &amd_base_port_ops,
+	.cable_detect	= ata_cable_unknown,
 	.set_piomode	= amd100_set_piomode,
 	.set_dmamode	= amd100_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= amd_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_unknown,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations amd133_port_ops = {
+	.inherits	= &amd_base_port_ops,
+	.cable_detect	= amd_cable_detect,
 	.set_piomode	= amd133_set_piomode,
 	.set_dmamode	= amd133_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= amd_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= amd_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
+};
 
-	.port_start	= ata_sff_port_start,
+static const struct ata_port_operations nv_base_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= ata_cable_ignore,
+	.mode_filter	= nv_mode_filter,
+	.error_handler	= nv_error_handler,
+	.host_stop	= nv_host_stop,
 };
 
 static struct ata_port_operations nv100_port_ops = {
+	.inherits	= &nv_base_port_ops,
 	.set_piomode	= nv100_set_piomode,
 	.set_dmamode	= nv100_set_dmamode,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= nv_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_ignore,
-	.mode_filter	= nv_mode_filter,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
-	.host_stop	= nv_host_stop,
 };
 
 static struct ata_port_operations nv133_port_ops = {
+	.inherits	= &nv_base_port_ops,
 	.set_piomode	= nv133_set_piomode,
 	.set_dmamode	= nv133_set_dmamode,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= nv_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_ignore,
-	.mode_filter	= nv_mode_filter,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
-	.host_stop	= nv_host_stop,
 };
 
 static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c
index 7bfb7e8bdca2..0101e5aef3e0 100644
--- a/drivers/ata/pata_artop.c
+++ b/drivers/ata/pata_artop.c
@@ -317,69 +317,20 @@ static struct scsi_host_template artop_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations artop6210_ops = {
+static struct ata_port_operations artop6210_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= artop6210_set_piomode,
 	.set_dmamode		= artop6210_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= artop6210_error_handler,
-	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_40wire,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations artop6260_ops = {
+static struct ata_port_operations artop6260_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= artop6260_cable_detect,
 	.set_piomode		= artop6260_set_piomode,
 	.set_dmamode		= artop6260_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= artop6260_error_handler,
-	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= artop6260_cable_detect,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_at32.c b/drivers/ata/pata_at32.c
index d7b7b7fde362..528315587532 100644
--- a/drivers/ata/pata_at32.c
+++ b/drivers/ata/pata_at32.c
@@ -171,28 +171,9 @@ static struct scsi_host_template at32_sht = {
 };
 
 static struct ata_port_operations at32_port_ops = {
-	.set_piomode		= pata_at32_set_piomode,
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+	.inherits		= &ata_sff_port_ops,
 	.cable_detect		= ata_cable_40wire,
-
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
-	.data_xfer		= ata_data_xfer,
-
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
+	.set_piomode		= pata_at32_set_piomode,
 };
 
 static int __init pata_at32_init_one(struct device *dev,
diff --git a/drivers/ata/pata_atiixp.c b/drivers/ata/pata_atiixp.c
index 645c47271ff5..2655f6a17ad3 100644
--- a/drivers/ata/pata_atiixp.c
+++ b/drivers/ata/pata_atiixp.c
@@ -226,36 +226,16 @@ static struct scsi_host_template atiixp_sht = {
 };
 
 static struct ata_port_operations atiixp_port_ops = {
-	.set_piomode	= atiixp_set_piomode,
-	.set_dmamode	= atiixp_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= atiixp_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= atiixp_cable_detect,
+	.inherits	= &ata_bmdma_port_ops,
 
-	.bmdma_setup 	= ata_bmdma_setup,
+	.qc_prep 	= ata_dumb_qc_prep,
 	.bmdma_start 	= atiixp_bmdma_start,
 	.bmdma_stop	= atiixp_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_dumb_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
 
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= atiixp_cable_detect,
+	.set_piomode	= atiixp_set_piomode,
+	.set_dmamode	= atiixp_set_dmamode,
+	.error_handler	= atiixp_error_handler,
 };
 
 static int atiixp_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c
index 6c75fcac3cf4..7a22ef483061 100644
--- a/drivers/ata/pata_bf54x.c
+++ b/drivers/ata/pata_bf54x.c
@@ -1363,6 +1363,8 @@ static struct scsi_host_template bfin_sht = {
 };
 
 static const struct ata_port_operations bfin_pata_ops = {
+	.inherits		= &ata_sff_port_ops,
+
 	.set_piomode		= bfin_set_piomode,
 	.set_dmamode		= bfin_set_dmamode,
 
@@ -1380,14 +1382,12 @@ static const struct ata_port_operations bfin_pata_ops = {
 	.data_xfer		= bfin_data_xfer,
 
 	.qc_prep		= ata_noop_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
 
 	.freeze			= bfin_bmdma_freeze,
 	.thaw			= bfin_bmdma_thaw,
 	.error_handler		= bfin_error_handler,
 	.post_internal_cmd	= bfin_bmdma_stop,
 
-	.irq_handler		= ata_interrupt,
 	.irq_clear		= bfin_irq_clear,
 	.irq_on			= bfin_irq_on,
 
diff --git a/drivers/ata/pata_cmd640.c b/drivers/ata/pata_cmd640.c
index 26562b814400..061c891c8a66 100644
--- a/drivers/ata/pata_cmd640.c
+++ b/drivers/ata/pata_cmd640.c
@@ -170,35 +170,12 @@ static struct scsi_host_template cmd640_sht = {
 };
 
 static struct ata_port_operations cmd640_port_ops = {
-	.set_piomode	= cmd640_set_piomode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= cmd640_qc_issue_prot,
-
-	/* In theory this is not needed once we kill the prefetcher */
+	.inherits	= &ata_bmdma_port_ops,
+	/* In theory xfer_noirq is not needed once we kill the prefetcher */
 	.data_xfer	= ata_data_xfer_noirq,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
+	.qc_issue	= cmd640_qc_issue_prot,
+	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= cmd640_set_piomode,
 	.port_start	= cmd640_port_start,
 };
 
diff --git a/drivers/ata/pata_cmd64x.c b/drivers/ata/pata_cmd64x.c
index 6aea05cc0940..1ac8ecfb97e2 100644
--- a/drivers/ata/pata_cmd64x.c
+++ b/drivers/ata/pata_cmd64x.c
@@ -269,103 +269,27 @@ static struct scsi_host_template cmd64x_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static struct ata_port_operations cmd64x_port_ops = {
+static const struct ata_port_operations cmd64x_base_ops = {
+	.inherits	= &ata_bmdma_port_ops,
 	.set_piomode	= cmd64x_set_piomode,
 	.set_dmamode	= cmd64x_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
-static struct ata_port_operations cmd646r1_port_ops = {
-	.set_piomode	= cmd64x_set_piomode,
-	.set_dmamode	= cmd64x_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+static struct ata_port_operations cmd64x_port_ops = {
+	.inherits	= &cmd64x_base_ops,
 	.cable_detect	= ata_cable_40wire,
+};
 
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
+static struct ata_port_operations cmd646r1_port_ops = {
+	.inherits	= &cmd64x_base_ops,
 	.bmdma_stop	= cmd646r1_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
 };
 
 static struct ata_port_operations cmd648_port_ops = {
-	.set_piomode	= cmd64x_set_piomode,
-	.set_dmamode	= cmd64x_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= cmd648_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
+	.inherits	= &cmd64x_base_ops,
 	.bmdma_stop	= cmd648_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= cmd648_cable_detect,
 };
 
 static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_cs5520.c b/drivers/ata/pata_cs5520.c
index 7e643099a444..46d0ce32ee5a 100644
--- a/drivers/ata/pata_cs5520.c
+++ b/drivers/ata/pata_cs5520.c
@@ -145,34 +145,11 @@ static struct scsi_host_template cs5520_sht = {
 };
 
 static struct ata_port_operations cs5520_port_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.qc_prep		= ata_dumb_qc_prep,
+	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= cs5520_set_piomode,
 	.set_dmamode		= cs5520_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_40wire,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_dumb_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 static int __devinit cs5520_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_cs5530.c b/drivers/ata/pata_cs5530.c
index 0bb03dabcf18..e4a16a578cac 100644
--- a/drivers/ata/pata_cs5530.c
+++ b/drivers/ata/pata_cs5530.c
@@ -166,37 +166,14 @@ static struct scsi_host_template cs5530_sht = {
 };
 
 static struct ata_port_operations cs5530_port_ops = {
-	.set_piomode	= cs5530_set_piomode,
-	.set_dmamode	= cs5530_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
+	.inherits	= &ata_bmdma_port_ops,
 
 	.qc_prep 	= ata_dumb_qc_prep,
 	.qc_issue	= cs5530_qc_issue_prot,
 
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= cs5530_set_piomode,
+	.set_dmamode	= cs5530_set_dmamode,
 };
 
 static const struct dmi_system_id palmax_dmi_table[] = {
diff --git a/drivers/ata/pata_cs5535.c b/drivers/ata/pata_cs5535.c
index 48a18349c1d8..f910a8aa7437 100644
--- a/drivers/ata/pata_cs5535.c
+++ b/drivers/ata/pata_cs5535.c
@@ -162,37 +162,10 @@ static struct scsi_host_template cs5535_sht = {
 };
 
 static struct ata_port_operations cs5535_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= cs5535_cable_detect,
 	.set_piomode	= cs5535_set_piomode,
 	.set_dmamode	= cs5535_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= cs5535_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_cs5536.c b/drivers/ata/pata_cs5536.c
index f02d9107ef3b..075ee6a7be39 100644
--- a/drivers/ata/pata_cs5536.c
+++ b/drivers/ata/pata_cs5536.c
@@ -225,37 +225,10 @@ static struct scsi_host_template cs5536_sht = {
 };
 
 static struct ata_port_operations cs5536_port_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= cs5536_cable_detect,
 	.set_piomode		= cs5536_set_piomode,
 	.set_dmamode		= cs5536_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= cs5536_cable_detect,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_cypress.c b/drivers/ata/pata_cypress.c
index 07fa1ab36315..c459553e7d1e 100644
--- a/drivers/ata/pata_cypress.c
+++ b/drivers/ata/pata_cypress.c
@@ -114,37 +114,10 @@ static struct scsi_host_template cy82c693_sht = {
 };
 
 static struct ata_port_operations cy82c693_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= cy82c693_set_piomode,
 	.set_dmamode	= cy82c693_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static int cy82c693_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_efar.c b/drivers/ata/pata_efar.c
index 8700d9dcd8c9..ef62fc642c17 100644
--- a/drivers/ata/pata_efar.c
+++ b/drivers/ata/pata_efar.c
@@ -236,36 +236,12 @@ static struct scsi_host_template efar_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations efar_ops = {
+static struct ata_port_operations efar_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= efar_cable_detect,
 	.set_piomode		= efar_set_piomode,
 	.set_dmamode		= efar_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= efar_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= efar_cable_detect,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_hpt366.c b/drivers/ata/pata_hpt366.c
index a30028de41c0..788955f57ff8 100644
--- a/drivers/ata/pata_hpt366.c
+++ b/drivers/ata/pata_hpt366.c
@@ -298,37 +298,11 @@ static struct scsi_host_template hpt36x_sht = {
  */
 
 static struct ata_port_operations hpt366_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= hpt36x_cable_detect,
+	.mode_filter	= hpt366_filter,
 	.set_piomode	= hpt366_set_piomode,
 	.set_dmamode	= hpt366_set_dmamode,
-	.mode_filter	= hpt366_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= hpt36x_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index 7d6fac43e2f9..c42eec70d297 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -627,36 +627,15 @@ static struct scsi_host_template hpt37x_sht = {
  */
 
 static struct ata_port_operations hpt370_port_ops = {
-	.set_piomode	= hpt370_set_piomode,
-	.set_dmamode	= hpt370_set_dmamode,
-	.mode_filter	= hpt370_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= hpt37x_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.inherits	= &ata_bmdma_port_ops,
 
-	.bmdma_setup 	= ata_bmdma_setup,
 	.bmdma_start 	= hpt370_bmdma_start,
 	.bmdma_stop	= hpt370_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
 
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.mode_filter	= hpt370_filter,
+	.set_piomode	= hpt370_set_piomode,
+	.set_dmamode	= hpt370_set_dmamode,
+	.error_handler	= hpt37x_error_handler,
 };
 
 /*
@@ -664,36 +643,8 @@ static struct ata_port_operations hpt370_port_ops = {
  */
 
 static struct ata_port_operations hpt370a_port_ops = {
-	.set_piomode	= hpt370_set_piomode,
-	.set_dmamode	= hpt370_set_dmamode,
+	.inherits	= &hpt370_port_ops,
 	.mode_filter	= hpt370a_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= hpt37x_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= hpt370_bmdma_start,
-	.bmdma_stop	= hpt370_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /*
@@ -702,36 +653,13 @@ static struct ata_port_operations hpt370a_port_ops = {
  */
 
 static struct ata_port_operations hpt372_port_ops = {
-	.set_piomode	= hpt372_set_piomode,
-	.set_dmamode	= hpt372_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= hpt37x_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.inherits	= &ata_bmdma_port_ops,
 
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
 	.bmdma_stop	= hpt37x_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
 
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.set_piomode	= hpt372_set_piomode,
+	.set_dmamode	= hpt372_set_dmamode,
+	.error_handler	= hpt37x_error_handler,
 };
 
 /*
@@ -740,36 +668,8 @@ static struct ata_port_operations hpt372_port_ops = {
  */
 
 static struct ata_port_operations hpt374_port_ops = {
-	.set_piomode	= hpt372_set_piomode,
-	.set_dmamode	= hpt372_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
+	.inherits	= &hpt372_port_ops,
 	.error_handler	= hpt374_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= hpt37x_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_hpt3x2n.c b/drivers/ata/pata_hpt3x2n.c
index aa380c46b168..b77b1279d757 100644
--- a/drivers/ata/pata_hpt3x2n.c
+++ b/drivers/ata/pata_hpt3x2n.c
@@ -347,37 +347,15 @@ static struct scsi_host_template hpt3x2n_sht = {
  */
 
 static struct ata_port_operations hpt3x2n_port_ops = {
-	.set_piomode	= hpt3x2n_set_piomode,
-	.set_dmamode	= hpt3x2n_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
+	.inherits	= &ata_bmdma_port_ops,
 
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= hpt3x2n_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= hpt3x2n_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
 	.bmdma_stop	= hpt3x2n_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
 	.qc_issue	= hpt3x2n_qc_issue_prot,
 
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= hpt3x2n_cable_detect,
+	.set_piomode	= hpt3x2n_set_piomode,
+	.set_dmamode	= hpt3x2n_set_dmamode,
+	.error_handler	= hpt3x2n_error_handler,
 };
 
 /**
diff --git a/drivers/ata/pata_hpt3x3.c b/drivers/ata/pata_hpt3x3.c
index 9837ab0181e8..8857d029ac2e 100644
--- a/drivers/ata/pata_hpt3x3.c
+++ b/drivers/ata/pata_hpt3x3.c
@@ -106,40 +106,13 @@ static struct scsi_host_template hpt3x3_sht = {
 };
 
 static struct ata_port_operations hpt3x3_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.check_atapi_dma= hpt3x3_atapi_dma,
+	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= hpt3x3_set_piomode,
 #if defined(CONFIG_PATA_HPT3X3_DMA)
 	.set_dmamode	= hpt3x3_set_dmamode,
 #endif
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-	.check_atapi_dma= hpt3x3_atapi_dma,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_icside.c b/drivers/ata/pata_icside.c
index 88a1c7ae0a4d..ff16b0eaa2c2 100644
--- a/drivers/ata/pata_icside.c
+++ b/drivers/ata/pata_icside.c
@@ -339,35 +339,19 @@ static void pata_icside_error_handler(struct ata_port *ap)
 }
 
 static struct ata_port_operations pata_icside_port_ops = {
-	.set_dmamode		= pata_icside_set_dmamode,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-
-	.cable_detect		= ata_cable_40wire,
-
-	.bmdma_setup		= pata_icside_bmdma_setup,
-	.bmdma_start		= pata_icside_bmdma_start,
-
-	.data_xfer		= ata_data_xfer_noirq,
-
+	.inherits		= &ata_sff_port_ops,
 	/* no need to build any PRD tables for DMA */
 	.qc_prep		= ata_noop_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
+	.data_xfer		= ata_data_xfer_noirq,
+	.bmdma_setup		= pata_icside_bmdma_setup,
+	.bmdma_start		= pata_icside_bmdma_start,
+	.bmdma_stop		= pata_icside_bmdma_stop,
+	.bmdma_status		= pata_icside_bmdma_status,
 
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
+	.cable_detect		= ata_cable_40wire,
+	.set_dmamode		= pata_icside_set_dmamode,
 	.error_handler		= pata_icside_error_handler,
 	.post_internal_cmd	= pata_icside_bmdma_stop,
-
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.bmdma_stop		= pata_icside_bmdma_stop,
-	.bmdma_status		= pata_icside_bmdma_status,
 };
 
 static void __devinit
diff --git a/drivers/ata/pata_isapnp.c b/drivers/ata/pata_isapnp.c
index 91ca4d50db04..085913ec6f68 100644
--- a/drivers/ata/pata_isapnp.c
+++ b/drivers/ata/pata_isapnp.c
@@ -24,27 +24,8 @@ static struct scsi_host_template isapnp_sht = {
 };
 
 static struct ata_port_operations isapnp_port_ops = {
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.inherits	= &ata_sff_port_ops,
 	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_it8213.c b/drivers/ata/pata_it8213.c
index 678a05b304d8..9ce89522e764 100644
--- a/drivers/ata/pata_it8213.c
+++ b/drivers/ata/pata_it8213.c
@@ -246,36 +246,13 @@ static struct scsi_host_template it8213_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations it8213_ops = {
+
+static struct ata_port_operations it8213_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= it8213_cable_detect,
 	.set_piomode		= it8213_set_piomode,
 	.set_dmamode		= it8213_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= it8213_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= it8213_cable_detect,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c
index 7d969c911731..669d224d30ca 100644
--- a/drivers/ata/pata_it821x.c
+++ b/drivers/ata/pata_it821x.c
@@ -636,71 +636,30 @@ static struct scsi_host_template it821x_sht = {
 };
 
 static struct ata_port_operations it821x_smart_port_ops = {
-	.set_mode	= it821x_smart_set_mode,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.mode_filter	= ata_pci_default_filter,
+	.inherits	= &ata_bmdma_port_ops,
 
-	.check_status 	= ata_check_status,
 	.check_atapi_dma= it821x_check_atapi_dma,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-	.dev_config	= it821x_dev_config,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= it821x_ident_hack,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
 	.qc_issue	= it821x_smart_qc_issue_prot,
 
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
+	.cable_detect	= it821x_ident_hack,
+	.set_mode	= it821x_smart_set_mode,
+	.dev_config	= it821x_dev_config,
 
 	.port_start	= it821x_port_start,
 };
 
 static struct ata_port_operations it821x_passthru_port_ops = {
-	.set_piomode	= it821x_passthru_set_piomode,
-	.set_dmamode	= it821x_passthru_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
+	.inherits	= &ata_bmdma_port_ops,
 
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
 	.check_atapi_dma= it821x_check_atapi_dma,
 	.dev_select 	= it821x_passthru_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_unknown,
-
-	.bmdma_setup 	= ata_bmdma_setup,
 	.bmdma_start 	= it821x_passthru_bmdma_start,
 	.bmdma_stop	= it821x_passthru_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
 	.qc_issue	= it821x_passthru_qc_issue_prot,
 
-	.data_xfer	= ata_data_xfer,
-
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_handler	= ata_interrupt,
-	.irq_on		= ata_irq_on,
+	.cable_detect	= ata_cable_unknown,
+	.set_piomode	= it821x_passthru_set_piomode,
+	.set_dmamode	= it821x_passthru_set_dmamode,
 
 	.port_start	= it821x_port_start,
 };
diff --git a/drivers/ata/pata_ixp4xx_cf.c b/drivers/ata/pata_ixp4xx_cf.c
index b7e8e825a869..d02629aa20da 100644
--- a/drivers/ata/pata_ixp4xx_cf.c
+++ b/drivers/ata/pata_ixp4xx_cf.c
@@ -92,29 +92,10 @@ static struct scsi_host_template ixp4xx_sht = {
 };
 
 static struct ata_port_operations ixp4xx_port_ops = {
-	.set_mode		= ixp4xx_set_mode,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status 		= ata_check_status,
-	.dev_select 		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.qc_prep 		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
+	.inherits		= &ata_sff_port_ops,
 	.data_xfer		= ixp4xx_mmio_data_xfer,
 	.cable_detect		= ata_cable_40wire,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
+	.set_mode		= ixp4xx_set_mode,
 };
 
 static void ixp4xx_setup_port(struct ata_port *ap,
diff --git a/drivers/ata/pata_jmicron.c b/drivers/ata/pata_jmicron.c
index 69781af7b1bb..61ff5c6b4568 100644
--- a/drivers/ata/pata_jmicron.c
+++ b/drivers/ata/pata_jmicron.c
@@ -125,36 +125,9 @@ static struct scsi_host_template jmicron_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations jmicron_ops = {
-	/* Task file is PCI ATA format, use helpers */
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
+static struct ata_port_operations jmicron_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.error_handler		= jmicron_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	/* BMDMA handling is PCI ATA format, use helpers */
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	/* IRQ-related hooks */
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	/* Generic PATA PCI ATA helpers */
-	.port_start		= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c
index 5329b954c5f2..2474068596f4 100644
--- a/drivers/ata/pata_legacy.c
+++ b/drivers/ata/pata_legacy.c
@@ -211,6 +211,11 @@ static struct scsi_host_template legacy_sht = {
 	ATA_PIO_SHT(DRV_NAME),
 };
 
+static const struct ata_port_operations legacy_base_port_ops = {
+	.inherits	= &ata_sff_port_ops,
+	.cable_detect	= ata_cable_40wire,
+};
+
 /*
  *	These ops are used if the user indicates the hardware
  *	snoops the commands to decide on the mode and handles the
@@ -220,55 +225,14 @@ static struct scsi_host_template legacy_sht = {
  */
 
 static struct ata_port_operations simple_port_ops = {
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
+	.inherits	= &legacy_base_port_ops,
 	.data_xfer	= ata_data_xfer_noirq,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations legacy_port_ops = {
-	.set_mode	= legacy_set_mode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-	.cable_detect	= ata_cable_40wire,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
+	.inherits	= &legacy_base_port_ops,
 	.data_xfer	= ata_data_xfer_noirq,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.set_mode	= legacy_set_mode,
 };
 
 /*
@@ -359,30 +323,9 @@ static unsigned int pdc_data_xfer_vlb(struct ata_device *dev,
 }
 
 static struct ata_port_operations pdc20230_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= pdc20230_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
 	.data_xfer	= pdc_data_xfer_vlb,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /*
@@ -413,30 +356,8 @@ static void ht6560a_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct ata_port_operations ht6560a_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= ht6560a_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,	/* Check vlb/noirq */
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /*
@@ -478,30 +399,8 @@ static void ht6560b_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 static struct ata_port_operations ht6560b_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= ht6560b_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,    /* FIXME: Check 32bit and noirq */
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /*
@@ -599,30 +498,8 @@ static void opti82c611a_set_piomode(struct ata_port *ap,
 
 
 static struct ata_port_operations opti82c611a_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= opti82c611a_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /*
@@ -731,30 +608,9 @@ static unsigned int opti82c46x_qc_issue_prot(struct ata_queued_cmd *qc)
 }
 
 static struct ata_port_operations opti82c46x_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= opti82c46x_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
 	.qc_issue	= opti82c46x_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static void qdi6500_set_piomode(struct ata_port *ap, struct ata_device *adev)
@@ -916,84 +772,22 @@ static int qdi_port(struct platform_device *dev,
 }
 
 static struct ata_port_operations qdi6500_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= qdi6500_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
 	.qc_issue	= qdi_qc_issue_prot,
-
 	.data_xfer	= vlb32_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations qdi6580_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= qdi6580_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
 	.data_xfer	= vlb32_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations qdi6580dp_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= qdi6580dp_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= qdi_qc_issue_prot,
-
 	.data_xfer	= vlb32_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static DEFINE_SPINLOCK(winbond_lock);
@@ -1062,29 +856,9 @@ static int winbond_port(struct platform_device *dev,
 }
 
 static struct ata_port_operations winbond_port_ops = {
+	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= winbond_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
 	.data_xfer	= vlb32_data_xfer,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct legacy_controller controllers[] = {
diff --git a/drivers/ata/pata_marvell.c b/drivers/ata/pata_marvell.c
index 9de6e429d0d1..286310fc5910 100644
--- a/drivers/ata/pata_marvell.c
+++ b/drivers/ata/pata_marvell.c
@@ -95,37 +95,10 @@ static struct scsi_host_template marvell_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations marvell_ops = {
-	/* Task file is PCI ATA format, use helpers */
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= marvell_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+static struct ata_port_operations marvell_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.cable_detect		= marvell_cable_detect,
-
-	/* BMDMA handling is PCI ATA format, use helpers */
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	/* Timeout handling */
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	/* Generic PATA PCI ATA helpers */
-	.port_start		= ata_sff_port_start,
+	.error_handler		= marvell_error_handler,
 };
 
 
diff --git a/drivers/ata/pata_mpc52xx.c b/drivers/ata/pata_mpc52xx.c
index 4117b618a9d9..ac7c0822b1a7 100644
--- a/drivers/ata/pata_mpc52xx.c
+++ b/drivers/ata/pata_mpc52xx.c
@@ -269,22 +269,12 @@ static struct scsi_host_template mpc52xx_ata_sht = {
 };
 
 static struct ata_port_operations mpc52xx_ata_port_ops = {
-	.set_piomode		= mpc52xx_ata_set_piomode,
+	.inherits		= &ata_sff_port_ops,
 	.dev_select		= mpc52xx_ata_dev_select,
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= mpc52xx_ata_error_handler,
 	.cable_detect		= ata_cable_40wire,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-	.port_start		= ata_sff_port_start,
+	.set_piomode		= mpc52xx_ata_set_piomode,
+	.error_handler		= mpc52xx_ata_error_handler,
+	.post_internal_cmd	= ATA_OP_NULL,
 };
 
 static int __devinit
diff --git a/drivers/ata/pata_mpiix.c b/drivers/ata/pata_mpiix.c
index e8e6837110b4..dab54f8a272d 100644
--- a/drivers/ata/pata_mpiix.c
+++ b/drivers/ata/pata_mpiix.c
@@ -155,28 +155,11 @@ static struct scsi_host_template mpiix_sht = {
 };
 
 static struct ata_port_operations mpiix_port_ops = {
+	.inherits	= &ata_sff_port_ops,
+	.qc_issue	= mpiix_qc_issue_prot,
+	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= mpiix_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
 	.error_handler	= mpiix_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= mpiix_qc_issue_prot,
-	.data_xfer	= ata_data_xfer,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static int mpiix_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_netcell.c b/drivers/ata/pata_netcell.c
index 11f200a2a156..65389d1837b3 100644
--- a/drivers/ata/pata_netcell.c
+++ b/drivers/ata/pata_netcell.c
@@ -24,37 +24,9 @@ static struct scsi_host_template netcell_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations netcell_ops = {
-	/* Task file is PCI ATA format, use helpers */
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+static struct ata_port_operations netcell_ops = {
+	.inherits	= &ata_bmdma_port_ops,
 	.cable_detect		= ata_cable_80wire,
-
-	/* BMDMA handling is PCI ATA format, use helpers */
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	/* IRQ-related hooks */
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	/* Generic PATA PCI ATA helpers */
-	.port_start		= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_ninja32.c b/drivers/ata/pata_ninja32.c
index ce3b07cab8bc..8213d081f313 100644
--- a/drivers/ata/pata_ninja32.c
+++ b/drivers/ata/pata_ninja32.c
@@ -83,36 +83,10 @@ static struct scsi_host_template ninja32_sht = {
 };
 
 static struct ata_port_operations ninja32_port_ops = {
-	.set_piomode	= ninja32_set_piomode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
+	.inherits	= &ata_bmdma_port_ops,
 	.dev_select 	= ninja32_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
 	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.set_piomode	= ninja32_set_piomode,
 };
 
 static int ninja32_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
index d2f85f107d15..5b1982fa0be1 100644
--- a/drivers/ata/pata_ns87410.c
+++ b/drivers/ata/pata_ns87410.c
@@ -148,30 +148,11 @@ static struct scsi_host_template ns87410_sht = {
 };
 
 static struct ata_port_operations ns87410_port_ops = {
+	.inherits	= &ata_sff_port_ops,
+	.qc_issue	= ns87410_qc_issue_prot,
+	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= ns87410_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
 	.error_handler	= ns87410_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ns87410_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static int ns87410_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_ns87415.c b/drivers/ata/pata_ns87415.c
index 78d634423cbf..38d86a262dbb 100644
--- a/drivers/ata/pata_ns87415.c
+++ b/drivers/ata/pata_ns87415.c
@@ -297,73 +297,29 @@ static u8 ns87560_bmdma_status(struct ata_port *ap)
 {
 	return ns87560_read_buggy(ap->ioaddr.bmdma_addr + ATA_DMA_STATUS);
 }
+#endif		/* 87560 SuperIO Support */
 
-static const struct ata_port_operations ns87560_pata_ops = {
-	.set_piomode		= ns87415_set_piomode,
-	.mode_filter		= ata_pci_default_filter,
+static struct ata_port_operations ns87415_pata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 
-	.tf_load		= ata_tf_load,
-	.tf_read		= ns87560_tf_read,
-	.check_status		= ns87560_check_status,
 	.check_atapi_dma	= ns87415_check_atapi_dma,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_40wire,
-
 	.bmdma_setup		= ns87415_bmdma_setup,
 	.bmdma_start		= ns87415_bmdma_start,
 	.bmdma_stop		= ns87415_bmdma_stop,
-	.bmdma_status		= ns87560_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
 	.irq_clear		= ns87415_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
-};
-
-#endif		/* 87560 SuperIO Support */
-
-
-static const struct ata_port_operations ns87415_pata_ops = {
-	.set_piomode		= ns87415_set_piomode,
-	.mode_filter		= ata_pci_default_filter,
 
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.check_atapi_dma	= ns87415_check_atapi_dma,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
 	.cable_detect		= ata_cable_40wire,
+	.set_piomode		= ns87415_set_piomode,
+};
 
-	.bmdma_setup		= ns87415_bmdma_setup,
-	.bmdma_start		= ns87415_bmdma_start,
-	.bmdma_stop		= ns87415_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ns87415_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
+#if defined(CONFIG_SUPERIO)
+static struct ata_port_operations ns87560_pata_ops = {
+	.inherits		= &ns87415_pata_ops,
+	.tf_read		= ns87560_tf_read,
+	.check_status		= ns87560_check_status,
+	.bmdma_status		= ns87560_bmdma_status,
 };
+#endif
 
 static struct scsi_host_template ns87415_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
diff --git a/drivers/ata/pata_oldpiix.c b/drivers/ata/pata_oldpiix.c
index 45f9b3eb5b45..f6062b37310d 100644
--- a/drivers/ata/pata_oldpiix.c
+++ b/drivers/ata/pata_oldpiix.c
@@ -223,36 +223,13 @@ static struct scsi_host_template oldpiix_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations oldpiix_pata_ops = {
+static struct ata_port_operations oldpiix_pata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.qc_issue		= oldpiix_qc_issue_prot,
+	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= oldpiix_set_piomode,
 	.set_dmamode		= oldpiix_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= oldpiix_pata_error_handler,
-	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_40wire,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= oldpiix_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_opti.c b/drivers/ata/pata_opti.c
index 1deacfa0be07..c4a0795c3ff4 100644
--- a/drivers/ata/pata_opti.c
+++ b/drivers/ata/pata_opti.c
@@ -169,29 +169,10 @@ static struct scsi_host_template opti_sht = {
 };
 
 static struct ata_port_operations opti_port_ops = {
+	.inherits	= &ata_sff_port_ops,
+	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= opti_set_piomode,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
 	.error_handler	= opti_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static int opti_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_optidma.c b/drivers/ata/pata_optidma.c
index 7495758a86fe..eb4b08190e3a 100644
--- a/drivers/ata/pata_optidma.c
+++ b/drivers/ata/pata_optidma.c
@@ -354,73 +354,18 @@ static struct scsi_host_template optidma_sht = {
 };
 
 static struct ata_port_operations optidma_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= optidma_set_pio_mode,
 	.set_dmamode	= optidma_set_dma_mode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.error_handler	= optidma_error_handler,
 	.set_mode	= optidma_set_mode,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.error_handler	= optidma_error_handler,
 };
 
 static struct ata_port_operations optiplus_port_ops = {
+	.inherits	= &optidma_port_ops,
 	.set_piomode	= optiplus_set_pio_mode,
 	.set_dmamode	= optiplus_set_dma_mode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.error_handler	= optidma_error_handler,
-	.set_mode	= optidma_set_mode,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_pcmcia.c b/drivers/ata/pata_pcmcia.c
index c05b36c94d51..57efbf05c95f 100644
--- a/drivers/ata/pata_pcmcia.c
+++ b/drivers/ata/pata_pcmcia.c
@@ -132,53 +132,17 @@ static struct scsi_host_template pcmcia_sht = {
 };
 
 static struct ata_port_operations pcmcia_port_ops = {
-	.set_mode	= pcmcia_set_mode,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
+	.inherits	= &ata_sff_port_ops,
 	.data_xfer	= ata_data_xfer_noirq,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_mode	= pcmcia_set_mode,
 };
 
 static struct ata_port_operations pcmcia_8bit_port_ops = {
-	.set_mode	= pcmcia_set_mode_8bit,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
+	.inherits	= &ata_sff_port_ops,
 	.data_xfer	= ata_data_xfer_8bit,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_mode	= pcmcia_set_mode_8bit,
 };
 
 #define CS_CHECK(fn, ret) \
diff --git a/drivers/ata/pata_pdc2027x.c b/drivers/ata/pata_pdc2027x.c
index 229d9acd934a..f619c20dd192 100644
--- a/drivers/ata/pata_pdc2027x.c
+++ b/drivers/ata/pata_pdc2027x.c
@@ -133,66 +133,18 @@ static struct scsi_host_template pdc2027x_sht = {
 };
 
 static struct ata_port_operations pdc2027x_pata100_ops = {
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
+	.inherits		= &ata_bmdma_port_ops,
 	.check_atapi_dma	= pdc2027x_check_atapi_dma,
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= pdc2027x_error_handler,
-	.post_internal_cmd 	= ata_bmdma_post_internal_cmd,
 	.cable_detect		= pdc2027x_cable_detect,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
+	.error_handler		= pdc2027x_error_handler,
 };
 
 static struct ata_port_operations pdc2027x_pata133_ops = {
+	.inherits		= &pdc2027x_pata100_ops,
+	.mode_filter		= pdc2027x_mode_filter,
 	.set_piomode		= pdc2027x_set_piomode,
 	.set_dmamode		= pdc2027x_set_dmamode,
 	.set_mode		= pdc2027x_set_mode,
-	.mode_filter		= pdc2027x_mode_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.check_atapi_dma	= pdc2027x_check_atapi_dma,
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= pdc2027x_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= pdc2027x_cable_detect,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 static struct ata_port_info pdc2027x_port_info[] = {
diff --git a/drivers/ata/pata_pdc202xx_old.c b/drivers/ata/pata_pdc202xx_old.c
index 564ee0798ec1..4daac20df0bc 100644
--- a/drivers/ata/pata_pdc202xx_old.c
+++ b/drivers/ata/pata_pdc202xx_old.c
@@ -266,69 +266,24 @@ static struct scsi_host_template pdc202xx_sht = {
 };
 
 static struct ata_port_operations pdc2024x_port_ops = {
-	.set_piomode	= pdc202xx_set_piomode,
-	.set_dmamode	= pdc202xx_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.inherits		= &ata_bmdma_port_ops,
+
+	.cable_detect		= ata_cable_40wire,
+	.set_piomode		= pdc202xx_set_piomode,
+	.set_dmamode		= pdc202xx_set_dmamode,
 };
 
 static struct ata_port_operations pdc2026x_port_ops = {
-	.set_piomode	= pdc202xx_set_piomode,
-	.set_dmamode	= pdc202xx_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-	.dev_config	= pdc2026x_dev_config,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= pdc2026x_cable_detect,
-
-	.check_atapi_dma= pdc2026x_check_atapi_dma,
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= pdc2026x_bmdma_start,
-	.bmdma_stop	= pdc2026x_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= pdc2026x_port_start,
+	.inherits		= &pdc2024x_port_ops,
+
+	.check_atapi_dma	= pdc2026x_check_atapi_dma,
+	.bmdma_start		= pdc2026x_bmdma_start,
+	.bmdma_stop		= pdc2026x_bmdma_stop,
+
+	.cable_detect		= pdc2026x_cable_detect,
+	.dev_config		= pdc2026x_dev_config,
+
+	.port_start		= pdc2026x_port_start,
 };
 
 static int pdc202xx_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_platform.c b/drivers/ata/pata_platform.c
index cd2d03a4591a..0588c9b7e73e 100644
--- a/drivers/ata/pata_platform.c
+++ b/drivers/ata/pata_platform.c
@@ -51,27 +51,11 @@ static struct scsi_host_template pata_platform_sht = {
 };
 
 static struct ata_port_operations pata_platform_port_ops = {
-	.set_mode		= pata_platform_set_mode,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_unknown,
-
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
+	.inherits		= &ata_sff_port_ops,
 	.data_xfer		= ata_data_xfer_noirq,
-
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
+	.cable_detect		= ata_cable_unknown,
+	.set_mode		= pata_platform_set_mode,
+	.port_start		= ATA_OP_NULL,
 };
 
 static void pata_platform_setup_port(struct ata_ioports *ioaddr,
diff --git a/drivers/ata/pata_qdi.c b/drivers/ata/pata_qdi.c
index ccb8682300b8..d16b343d2a62 100644
--- a/drivers/ata/pata_qdi.c
+++ b/drivers/ata/pata_qdi.c
@@ -158,55 +158,16 @@ static struct scsi_host_template qdi_sht = {
 };
 
 static struct ata_port_operations qdi6500_port_ops = {
-	.set_piomode	= qdi6500_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
+	.inherits	= &ata_sff_port_ops,
 	.qc_issue	= qdi_qc_issue_prot,
-
 	.data_xfer	= qdi_data_xfer,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= qdi6500_set_piomode,
 };
 
 static struct ata_port_operations qdi6580_port_ops = {
+	.inherits	= &qdi6500_port_ops,
 	.set_piomode	= qdi6580_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= qdi_qc_issue_prot,
-
-	.data_xfer	= qdi_data_xfer,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_radisys.c b/drivers/ata/pata_radisys.c
index 3981bf84d093..94e60b3a1ec6 100644
--- a/drivers/ata/pata_radisys.c
+++ b/drivers/ata/pata_radisys.c
@@ -188,36 +188,12 @@ static struct scsi_host_template radisys_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations radisys_pata_ops = {
+static struct ata_port_operations radisys_pata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.qc_issue		= radisys_qc_issue_prot,
+	.cable_detect		= ata_cable_unknown,
 	.set_piomode		= radisys_set_piomode,
 	.set_dmamode		= radisys_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= ata_cable_unknown,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= radisys_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 
diff --git a/drivers/ata/pata_rb500_cf.c b/drivers/ata/pata_rb500_cf.c
index 4543c980342c..7affceec1c29 100644
--- a/drivers/ata/pata_rb500_cf.c
+++ b/drivers/ata/pata_rb500_cf.c
@@ -118,25 +118,11 @@ static irqreturn_t rb500_pata_irq_handler(int irq, void *dev_instance)
 }
 
 static struct ata_port_operations rb500_pata_port_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-
+	.inherits		= &ata_sff_port_ops,
 	.exec_command		= rb500_pata_exec_command,
-	.check_status 		= ata_check_status,
-	.dev_select 		= ata_std_dev_select,
-
 	.data_xfer		= rb500_pata_data_xfer,
-
-	.qc_prep 		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
 	.freeze			= rb500_pata_freeze,
 	.thaw			= rb500_pata_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
 };
 
 /* ------------------------------------------------------------------------ */
diff --git a/drivers/ata/pata_rz1000.c b/drivers/ata/pata_rz1000.c
index 80909a607d36..a2aef7328bfc 100644
--- a/drivers/ata/pata_rz1000.c
+++ b/drivers/ata/pata_rz1000.c
@@ -57,30 +57,9 @@ static struct scsi_host_template rz1000_sht = {
 };
 
 static struct ata_port_operations rz1000_port_ops = {
-	.set_mode	= rz1000_set_mode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
+	.inherits	= &ata_sff_port_ops,
 	.cable_detect	= ata_cable_40wire,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.set_mode	= rz1000_set_mode,
 };
 
 static int rz1000_fifo_disable(struct pci_dev *pdev)
diff --git a/drivers/ata/pata_sc1200.c b/drivers/ata/pata_sc1200.c
index 38ce6e12ee3d..362b7f829d8e 100644
--- a/drivers/ata/pata_sc1200.c
+++ b/drivers/ata/pata_sc1200.c
@@ -184,37 +184,12 @@ static struct scsi_host_template sc1200_sht = {
 };
 
 static struct ata_port_operations sc1200_port_ops = {
-	.set_piomode	= sc1200_set_piomode,
-	.set_dmamode	= sc1200_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
+	.inherits	= &ata_bmdma_port_ops,
 	.qc_prep 	= ata_dumb_qc_prep,
 	.qc_issue	= sc1200_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= sc1200_set_piomode,
+	.set_dmamode	= sc1200_set_dmamode,
 };
 
 /**
diff --git a/drivers/ata/pata_scc.c b/drivers/ata/pata_scc.c
index 1833e9ef522e..033d1f3a82de 100644
--- a/drivers/ata/pata_scc.c
+++ b/drivers/ata/pata_scc.c
@@ -972,6 +972,8 @@ static struct scsi_host_template scc_sht = {
 };
 
 static const struct ata_port_operations scc_pata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+
 	.set_piomode		= scc_set_piomode,
 	.set_dmamode		= scc_set_dmamode,
 	.mode_filter		= scc_mode_filter,
@@ -989,12 +991,7 @@ static const struct ata_port_operations scc_pata_ops = {
 	.bmdma_status		= scc_bmdma_status,
 	.data_xfer		= scc_data_xfer,
 
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
 	.freeze			= scc_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-
 	.error_handler		= scc_error_handler,
 	.post_internal_cmd	= scc_bmdma_stop,
 
diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
index 318a36988900..627abcf85c6e 100644
--- a/drivers/ata/pata_serverworks.c
+++ b/drivers/ata/pata_serverworks.c
@@ -302,71 +302,16 @@ static struct scsi_host_template serverworks_sht = {
 };
 
 static struct ata_port_operations serverworks_osb4_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= serverworks_cable_detect,
+	.mode_filter	= serverworks_osb4_filter,
 	.set_piomode	= serverworks_set_piomode,
 	.set_dmamode	= serverworks_set_dmamode,
-	.mode_filter	= serverworks_osb4_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= serverworks_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations serverworks_csb_port_ops = {
-	.set_piomode	= serverworks_set_piomode,
-	.set_dmamode	= serverworks_set_dmamode,
+	.inherits	= &serverworks_osb4_port_ops,
 	.mode_filter	= serverworks_csb_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= serverworks_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static int serverworks_fixup_osb4(struct pci_dev *pdev)
diff --git a/drivers/ata/pata_sil680.c b/drivers/ata/pata_sil680.c
index 7812815a34c3..0936f534d9c7 100644
--- a/drivers/ata/pata_sil680.c
+++ b/drivers/ata/pata_sil680.c
@@ -196,36 +196,10 @@ static struct scsi_host_template sil680_sht = {
 };
 
 static struct ata_port_operations sil680_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= sil680_cable_detect,
 	.set_piomode	= sil680_set_piomode,
 	.set_dmamode	= sil680_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= sil680_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
index dcd8457a8377..3ed628670cd7 100644
--- a/drivers/ata/pata_sis.c
+++ b/drivers/ata/pata_sis.c
@@ -517,196 +517,51 @@ static struct scsi_host_template sis_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations sis_133_ops = {
+static struct ata_port_operations sis_133_for_sata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.set_piomode		= sis_133_set_piomode,
 	.set_dmamode		= sis_133_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= sis_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
 	.cable_detect		= sis_133_cable_detect,
+};
 
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
+static struct ata_port_operations sis_base_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.error_handler		= sis_error_handler,
 };
 
-static const struct ata_port_operations sis_133_for_sata_ops = {
+static struct ata_port_operations sis_133_ops = {
+	.inherits		= &sis_base_ops,
 	.set_piomode		= sis_133_set_piomode,
 	.set_dmamode		= sis_133_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
 	.cable_detect		= sis_133_cable_detect,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations sis_133_early_ops = {
+static struct ata_port_operations sis_133_early_ops = {
+	.inherits		= &sis_base_ops,
 	.set_piomode		= sis_100_set_piomode,
 	.set_dmamode		= sis_133_early_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= sis_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
 	.cable_detect		= sis_66_cable_detect,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations sis_100_ops = {
+static struct ata_port_operations sis_100_ops = {
+	.inherits		= &sis_base_ops,
 	.set_piomode		= sis_100_set_piomode,
 	.set_dmamode		= sis_100_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= sis_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
 	.cable_detect		= sis_66_cable_detect,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations sis_66_ops = {
+static struct ata_port_operations sis_66_ops = {
+	.inherits		= &sis_base_ops,
 	.set_piomode		= sis_old_set_piomode,
 	.set_dmamode		= sis_66_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
 	.cable_detect		= sis_66_cable_detect,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= sis_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations sis_old_ops = {
+static struct ata_port_operations sis_old_ops = {
+	.inherits		= &sis_base_ops,
 	.set_piomode		= sis_old_set_piomode,
 	.set_dmamode		= sis_old_set_dmamode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= sis_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
 	.cable_detect		= ata_cable_40wire,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_handler		= ata_interrupt,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
 static const struct ata_port_info sis_info = {
diff --git a/drivers/ata/pata_sl82c105.c b/drivers/ata/pata_sl82c105.c
index ece366bced0c..0dd8e2f69558 100644
--- a/drivers/ata/pata_sl82c105.c
+++ b/drivers/ata/pata_sl82c105.c
@@ -239,37 +239,13 @@ static struct scsi_host_template sl82c105_sht = {
 };
 
 static struct ata_port_operations sl82c105_port_ops = {
-	.set_piomode	= sl82c105_set_piomode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= sl82c105_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
+	.inherits	= &ata_bmdma_port_ops,
+	.qc_defer	= sl82c105_qc_defer,
 	.bmdma_start 	= sl82c105_bmdma_start,
 	.bmdma_stop	= sl82c105_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_defer	= sl82c105_qc_defer,
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= sl82c105_set_piomode,
+	.error_handler	= sl82c105_error_handler,
 };
 
 /**
diff --git a/drivers/ata/pata_triflex.c b/drivers/ata/pata_triflex.c
index 510569957d10..bc4956ef0931 100644
--- a/drivers/ata/pata_triflex.c
+++ b/drivers/ata/pata_triflex.c
@@ -184,36 +184,12 @@ static struct scsi_host_template triflex_sht = {
 };
 
 static struct ata_port_operations triflex_port_ops = {
-	.set_piomode	= triflex_set_piomode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= triflex_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.bmdma_setup 	= ata_bmdma_setup,
+	.inherits	= &ata_bmdma_port_ops,
 	.bmdma_start 	= triflex_bmdma_start,
 	.bmdma_stop	= triflex_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= triflex_set_piomode,
+	.error_handler	= triflex_error_handler,
 };
 
 static int triflex_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
index a7bc860e1310..d1edb1b27480 100644
--- a/drivers/ata/pata_via.c
+++ b/drivers/ata/pata_via.c
@@ -339,71 +339,16 @@ static struct scsi_host_template via_sht = {
 };
 
 static struct ata_port_operations via_port_ops = {
+	.inherits	= &ata_bmdma_port_ops,
+	.cable_detect	= via_cable_detect,
 	.set_piomode	= via_set_piomode,
 	.set_dmamode	= via_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
 	.error_handler	= via_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= via_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
-	.data_xfer	= ata_data_xfer,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 static struct ata_port_operations via_port_ops_noirq = {
-	.set_piomode	= via_set_piomode,
-	.set_dmamode	= via_set_dmamode,
-	.mode_filter	= ata_pci_default_filter,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= via_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= via_cable_detect,
-
-	.bmdma_setup 	= ata_bmdma_setup,
-	.bmdma_start 	= ata_bmdma_start,
-	.bmdma_stop	= ata_bmdma_stop,
-	.bmdma_status 	= ata_bmdma_status,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
+	.inherits	= &via_port_ops,
 	.data_xfer	= ata_data_xfer_noirq,
-
-	.irq_handler	= ata_interrupt,
-	.irq_clear	= ata_bmdma_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
 };
 
 /**
diff --git a/drivers/ata/pata_winbond.c b/drivers/ata/pata_winbond.c
index 9bafae9d5fe0..f235bb0d6139 100644
--- a/drivers/ata/pata_winbond.c
+++ b/drivers/ata/pata_winbond.c
@@ -126,29 +126,10 @@ static struct scsi_host_template winbond_sht = {
 };
 
 static struct ata_port_operations winbond_port_ops = {
-	.set_piomode	= winbond_set_piomode,
-
-	.tf_load	= ata_tf_load,
-	.tf_read	= ata_tf_read,
-	.check_status 	= ata_check_status,
-	.exec_command	= ata_exec_command,
-	.dev_select 	= ata_std_dev_select,
-
-	.freeze		= ata_bmdma_freeze,
-	.thaw		= ata_bmdma_thaw,
-	.error_handler	= ata_bmdma_error_handler,
-	.post_internal_cmd = ata_bmdma_post_internal_cmd,
-	.cable_detect	= ata_cable_40wire,
-
-	.qc_prep 	= ata_qc_prep,
-	.qc_issue	= ata_qc_issue_prot,
-
+	.inherits	= &ata_sff_port_ops,
 	.data_xfer	= winbond_data_xfer,
-
-	.irq_clear	= ata_noop_irq_clear,
-	.irq_on		= ata_irq_on,
-
-	.port_start	= ata_sff_port_start,
+	.cable_detect	= ata_cable_40wire,
+	.set_piomode	= winbond_set_piomode,
 };
 
 /**
diff --git a/drivers/ata/pdc_adma.c b/drivers/ata/pdc_adma.c
index fdf62de57cfc..a5706149af6b 100644
--- a/drivers/ata/pdc_adma.c
+++ b/drivers/ata/pdc_adma.c
@@ -148,26 +148,29 @@ static struct scsi_host_template adma_ata_sht = {
 	.dma_boundary		= ADMA_DMA_BOUNDARY,
 };
 
-static const struct ata_port_operations adma_ata_ops = {
+static struct ata_port_operations adma_ata_ops = {
+	.inherits		= &ata_base_port_ops,
+
+	.dev_select		= ata_std_dev_select,
 	.tf_load		= ata_tf_load,
 	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
 	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-	.check_atapi_dma	= adma_check_atapi_dma,
+	.exec_command		= ata_exec_command,
 	.data_xfer		= ata_data_xfer,
+	.check_atapi_dma	= adma_check_atapi_dma,
+	.bmdma_stop		= adma_bmdma_stop,
+	.bmdma_status		= adma_bmdma_status,
 	.qc_prep		= adma_qc_prep,
 	.qc_issue		= adma_qc_issue,
+	.irq_on			= ata_irq_on,
+
 	.freeze			= adma_freeze,
 	.thaw			= adma_thaw,
 	.error_handler		= adma_error_handler,
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
+
 	.port_start		= adma_port_start,
 	.port_stop		= adma_port_stop,
 	.host_stop		= adma_host_stop,
-	.bmdma_stop		= adma_bmdma_stop,
-	.bmdma_status		= adma_bmdma_status,
 };
 
 static struct ata_port_info adma_port_info[] = {
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index f50381b4ba06..865030ae8f8a 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -1198,16 +1198,15 @@ static struct scsi_host_template sata_fsl_sht = {
 };
 
 static const struct ata_port_operations sata_fsl_ops = {
+	.inherits = &sata_port_ops,
+
 	.check_status = sata_fsl_check_status,
 	.check_altstatus = sata_fsl_check_status,
-	.dev_select = ata_noop_dev_select,
 
 	.tf_read = sata_fsl_tf_read,
 
-	.qc_defer = ata_std_qc_defer,
 	.qc_prep = sata_fsl_qc_prep,
 	.qc_issue = sata_fsl_qc_issue,
-	.irq_clear = ata_noop_irq_clear,
 
 	.scr_read = sata_fsl_scr_read,
 	.scr_write = sata_fsl_scr_write,
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index bb853df865da..047f80f5825c 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -522,26 +522,13 @@ static int inic_port_start(struct ata_port *ap)
 }
 
 static struct ata_port_operations inic_port_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.scr_read		= inic_scr_read,
-	.scr_write		= inic_scr_write,
+	.inherits		= &ata_sff_port_ops,
 
 	.bmdma_setup		= inic_bmdma_setup,
 	.bmdma_start		= inic_bmdma_start,
 	.bmdma_stop		= inic_bmdma_stop,
 	.bmdma_status		= inic_bmdma_status,
-
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.qc_prep	 	= ata_qc_prep,
 	.qc_issue		= inic_qc_issue,
-	.data_xfer		= ata_data_xfer,
 
 	.freeze			= inic_freeze,
 	.thaw			= inic_thaw,
@@ -549,8 +536,10 @@ static struct ata_port_operations inic_port_ops = {
 	.post_internal_cmd	= inic_post_internal_cmd,
 	.dev_config		= inic_dev_config,
 
-	.port_resume		= inic_port_resume,
+	.scr_read		= inic_scr_read,
+	.scr_write		= inic_scr_write,
 
+	.port_resume		= inic_port_resume,
 	.port_start		= inic_port_start,
 };
 
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 52d41edadb72..f341a82d27bf 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -526,23 +526,16 @@ static struct scsi_host_template mv6_sht = {
 	.dma_boundary		= MV_DMA_BOUNDARY,
 };
 
-static const struct ata_port_operations mv5_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
+static struct ata_port_operations mv5_ops = {
+	.inherits		= &ata_sff_port_ops,
 
 	.qc_prep		= mv_qc_prep,
 	.qc_issue		= mv_qc_issue,
-	.data_xfer		= ata_data_xfer,
 
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.error_handler		= mv_error_handler,
 	.freeze			= mv_eh_freeze,
 	.thaw			= mv_eh_thaw,
+	.error_handler		= mv_error_handler,
+	.post_internal_cmd	= ATA_OP_NULL,
 
 	.scr_read		= mv5_scr_read,
 	.scr_write		= mv5_scr_write,
@@ -551,57 +544,18 @@ static const struct ata_port_operations mv5_ops = {
 	.port_stop		= mv_port_stop,
 };
 
-static const struct ata_port_operations mv6_ops = {
-	.dev_config             = mv6_dev_config,
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.qc_prep		= mv_qc_prep,
-	.qc_issue		= mv_qc_issue,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.error_handler		= mv_error_handler,
-	.freeze			= mv_eh_freeze,
-	.thaw			= mv_eh_thaw,
+static struct ata_port_operations mv6_ops = {
+	.inherits		= &mv5_ops,
 	.qc_defer		= ata_std_qc_defer,
-
+	.dev_config             = mv6_dev_config,
 	.scr_read		= mv_scr_read,
 	.scr_write		= mv_scr_write,
-
-	.port_start		= mv_port_start,
-	.port_stop		= mv_port_stop,
 };
 
-static const struct ata_port_operations mv_iie_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
+static struct ata_port_operations mv_iie_ops = {
+	.inherits		= &mv6_ops,
+	.dev_config		= ATA_OP_NULL,
 	.qc_prep		= mv_qc_prep_iie,
-	.qc_issue		= mv_qc_issue,
-	.data_xfer		= ata_data_xfer,
-
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.error_handler		= mv_error_handler,
-	.freeze			= mv_eh_freeze,
-	.thaw			= mv_eh_thaw,
-	.qc_defer		= ata_std_qc_defer,
-
-	.scr_read		= mv_scr_read,
-	.scr_write		= mv_scr_write,
-
-	.port_start		= mv_port_start,
-	.port_stop		= mv_port_stop,
 };
 
 static const struct ata_port_info mv_port_info[] = {
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index 9e2b4cef48f2..7b7ba0e26903 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -404,106 +404,41 @@ static struct scsi_host_template nv_swncq_sht = {
 	.slave_configure	= nv_swncq_slave_config,
 };
 
-static const struct ata_port_operations nv_generic_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
+static struct ata_port_operations nv_generic_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.error_handler		= nv_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
 	.scr_read		= nv_scr_read,
 	.scr_write		= nv_scr_write,
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations nv_nf2_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.mode_filter		= ata_pci_default_filter,
+static struct ata_port_operations nv_nf2_ops = {
+	.inherits		= &nv_generic_ops,
 	.freeze			= nv_nf2_freeze,
 	.thaw			= nv_nf2_thaw,
-	.error_handler		= nv_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-	.scr_read		= nv_scr_read,
-	.scr_write		= nv_scr_write,
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations nv_ck804_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.mode_filter		= ata_pci_default_filter,
+static struct ata_port_operations nv_ck804_ops = {
+	.inherits		= &nv_generic_ops,
 	.freeze			= nv_ck804_freeze,
 	.thaw			= nv_ck804_thaw,
-	.error_handler		= nv_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-	.scr_read		= nv_scr_read,
-	.scr_write		= nv_scr_write,
-	.port_start		= ata_sff_port_start,
 	.host_stop		= nv_ck804_host_stop,
 };
 
-static const struct ata_port_operations nv_adma_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= nv_adma_tf_read,
+static struct ata_port_operations nv_adma_ops = {
+	.inherits		= &nv_generic_ops,
+
 	.check_atapi_dma	= nv_adma_check_atapi_dma,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
+	.tf_read		= nv_adma_tf_read,
 	.qc_defer		= ata_std_qc_defer,
 	.qc_prep		= nv_adma_qc_prep,
 	.qc_issue		= nv_adma_qc_issue,
-	.mode_filter		= ata_pci_default_filter,
+	.irq_clear		= nv_adma_irq_clear,
+
 	.freeze			= nv_adma_freeze,
 	.thaw			= nv_adma_thaw,
 	.error_handler		= nv_adma_error_handler,
 	.post_internal_cmd	= nv_adma_post_internal_cmd,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= nv_adma_irq_clear,
-	.irq_on			= ata_irq_on,
-	.scr_read		= nv_scr_read,
-	.scr_write		= nv_scr_write,
+
 	.port_start		= nv_adma_port_start,
 	.port_stop		= nv_adma_port_stop,
 #ifdef CONFIG_PM
@@ -513,29 +448,17 @@ static const struct ata_port_operations nv_adma_ops = {
 	.host_stop		= nv_adma_host_stop,
 };
 
-static const struct ata_port_operations nv_swncq_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
+static struct ata_port_operations nv_swncq_ops = {
+	.inherits		= &nv_generic_ops,
+
 	.qc_defer		= ata_std_qc_defer,
 	.qc_prep		= nv_swncq_qc_prep,
 	.qc_issue		= nv_swncq_qc_issue,
-	.mode_filter		= ata_pci_default_filter,
+
 	.freeze			= nv_mcp55_freeze,
 	.thaw			= nv_mcp55_thaw,
 	.error_handler		= nv_swncq_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-	.scr_read		= nv_scr_read,
-	.scr_write		= nv_scr_write,
+
 #ifdef CONFIG_PM
 	.port_suspend		= nv_swncq_port_suspend,
 	.port_resume		= nv_swncq_port_resume,
diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c
index 37c32ab3b23b..e09b975c973d 100644
--- a/drivers/ata/sata_promise.c
+++ b/drivers/ata/sata_promise.c
@@ -160,74 +160,42 @@ static struct scsi_host_template pdc_ata_sht = {
 	.dma_boundary		= ATA_DMA_BOUNDARY,
 };
 
-static const struct ata_port_operations pdc_sata_ops = {
+static const struct ata_port_operations pdc_common_ops = {
+	.inherits		= &ata_sff_port_ops,
+
 	.tf_load		= pdc_tf_load_mmio,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
 	.exec_command		= pdc_exec_command_mmio,
-	.dev_select		= ata_std_dev_select,
 	.check_atapi_dma	= pdc_check_atapi_dma,
-
 	.qc_prep		= pdc_qc_prep,
 	.qc_issue		= pdc_qc_issue_prot,
-	.freeze			= pdc_sata_freeze,
-	.thaw			= pdc_sata_thaw,
-	.error_handler		= pdc_sata_error_handler,
-	.post_internal_cmd	= pdc_post_internal_cmd,
-	.cable_detect		= pdc_sata_cable_detect,
-	.data_xfer		= ata_data_xfer,
 	.irq_clear		= pdc_irq_clear,
-	.irq_on			= ata_irq_on,
 
-	.scr_read		= pdc_sata_scr_read,
-	.scr_write		= pdc_sata_scr_write,
-	.port_start		= pdc_sata_port_start,
+	.post_internal_cmd	= pdc_post_internal_cmd,
 };
 
-/* First-generation chips need a more restrictive ->check_atapi_dma op */
-static const struct ata_port_operations pdc_old_sata_ops = {
-	.tf_load		= pdc_tf_load_mmio,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= pdc_exec_command_mmio,
-	.dev_select		= ata_std_dev_select,
-	.check_atapi_dma	= pdc_old_sata_check_atapi_dma,
-
-	.qc_prep		= pdc_qc_prep,
-	.qc_issue		= pdc_qc_issue_prot,
+static struct ata_port_operations pdc_sata_ops = {
+	.inherits		= &pdc_common_ops,
+	.cable_detect		= pdc_sata_cable_detect,
 	.freeze			= pdc_sata_freeze,
 	.thaw			= pdc_sata_thaw,
 	.error_handler		= pdc_sata_error_handler,
-	.post_internal_cmd	= pdc_post_internal_cmd,
-	.cable_detect		= pdc_sata_cable_detect,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= pdc_irq_clear,
-	.irq_on			= ata_irq_on,
-
 	.scr_read		= pdc_sata_scr_read,
 	.scr_write		= pdc_sata_scr_write,
 	.port_start		= pdc_sata_port_start,
 };
 
-static const struct ata_port_operations pdc_pata_ops = {
-	.tf_load		= pdc_tf_load_mmio,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= pdc_exec_command_mmio,
-	.dev_select		= ata_std_dev_select,
-	.check_atapi_dma	= pdc_check_atapi_dma,
+/* First-generation chips need a more restrictive ->check_atapi_dma op */
+static struct ata_port_operations pdc_old_sata_ops = {
+	.inherits		= &pdc_sata_ops,
+	.check_atapi_dma	= pdc_old_sata_check_atapi_dma,
+};
 
-	.qc_prep		= pdc_qc_prep,
-	.qc_issue		= pdc_qc_issue_prot,
+static struct ata_port_operations pdc_pata_ops = {
+	.inherits		= &pdc_common_ops,
+	.cable_detect		= pdc_pata_cable_detect,
 	.freeze			= pdc_freeze,
 	.thaw			= pdc_thaw,
 	.error_handler		= pdc_pata_error_handler,
-	.post_internal_cmd	= pdc_post_internal_cmd,
-	.cable_detect		= pdc_pata_cable_detect,
-	.data_xfer		= ata_data_xfer,
-	.irq_clear		= pdc_irq_clear,
-	.irq_on			= ata_irq_on,
-
 	.port_start		= pdc_common_port_start,
 };
 
diff --git a/drivers/ata/sata_qstor.c b/drivers/ata/sata_qstor.c
index 2566d0926aab..107ef09814de 100644
--- a/drivers/ata/sata_qstor.c
+++ b/drivers/ata/sata_qstor.c
@@ -131,27 +131,25 @@ static struct scsi_host_template qs_ata_sht = {
 	.dma_boundary		= QS_DMA_BOUNDARY,
 };
 
-static const struct ata_port_operations qs_ata_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
+static struct ata_port_operations qs_ata_ops = {
+	.inherits		= &ata_sff_port_ops,
+
 	.check_atapi_dma	= qs_check_atapi_dma,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
+	.bmdma_stop		= qs_bmdma_stop,
+	.bmdma_status		= qs_bmdma_status,
 	.qc_prep		= qs_qc_prep,
 	.qc_issue		= qs_qc_issue,
-	.data_xfer		= ata_data_xfer,
+
 	.freeze			= qs_freeze,
 	.thaw			= qs_thaw,
 	.error_handler		= qs_error_handler,
-	.irq_clear		= ata_noop_irq_clear,
-	.irq_on			= ata_irq_on,
+	.post_internal_cmd	= ATA_OP_NULL,
+
 	.scr_read		= qs_scr_read,
 	.scr_write		= qs_scr_write,
+
 	.port_start		= qs_port_start,
 	.host_stop		= qs_host_stop,
-	.bmdma_stop		= qs_bmdma_stop,
-	.bmdma_status		= qs_bmdma_status,
 };
 
 static const struct ata_port_info qs_port_info[] = {
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index 738c1a8ae3b6..eac7ca73cfa0 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -170,31 +170,14 @@ static struct scsi_host_template sil_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations sil_ops = {
+static struct ata_port_operations sil_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.dev_config		= sil_dev_config,
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
 	.set_mode		= sil_set_mode,
-	.mode_filter		= ata_pci_default_filter,
-	.bmdma_setup            = ata_bmdma_setup,
-	.bmdma_start            = ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
 	.freeze			= sil_freeze,
 	.thaw			= sil_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
 	.scr_read		= sil_scr_read,
 	.scr_write		= sil_scr_write,
-	.port_start		= ata_sff_port_start,
 };
 
 static const struct ata_port_info sil_port_info[] = {
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index 7fa63950d81a..363fb90e1047 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -390,34 +390,28 @@ static struct scsi_host_template sil24_sht = {
 	.dma_boundary		= ATA_DMA_BOUNDARY,
 };
 
-static const struct ata_port_operations sil24_ops = {
-	.dev_config		= sil24_dev_config,
+static struct ata_port_operations sil24_ops = {
+	.inherits		= &sata_pmp_port_ops,
 
 	.check_status		= sil24_check_status,
 	.check_altstatus	= sil24_check_status,
-	.dev_select		= ata_noop_dev_select,
-
 	.tf_read		= sil24_tf_read,
-
 	.qc_defer		= sil24_qc_defer,
 	.qc_prep		= sil24_qc_prep,
 	.qc_issue		= sil24_qc_issue,
 
-	.irq_clear		= ata_noop_irq_clear,
+	.freeze			= sil24_freeze,
+	.thaw			= sil24_thaw,
+	.error_handler		= sil24_error_handler,
+	.post_internal_cmd	= sil24_post_internal_cmd,
+	.dev_config		= sil24_dev_config,
 
 	.scr_read		= sil24_scr_read,
 	.scr_write		= sil24_scr_write,
-
 	.pmp_attach		= sil24_pmp_attach,
 	.pmp_detach		= sil24_pmp_detach,
 
-	.freeze			= sil24_freeze,
-	.thaw			= sil24_thaw,
-	.error_handler		= sil24_error_handler,
-	.post_internal_cmd	= sil24_post_internal_cmd,
-
 	.port_start		= sil24_port_start,
-
 #ifdef CONFIG_PM
 	.port_resume		= sil24_port_resume,
 #endif
diff --git a/drivers/ata/sata_sis.c b/drivers/ata/sata_sis.c
index 4becb7fde5e7..9089c7ab5000 100644
--- a/drivers/ata/sata_sis.c
+++ b/drivers/ata/sata_sis.c
@@ -89,29 +89,10 @@ static struct scsi_host_template sis_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations sis_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-	.bmdma_setup            = ata_bmdma_setup,
-	.bmdma_start            = ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
+static struct ata_port_operations sis_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.scr_read		= sis_scr_read,
 	.scr_write		= sis_scr_write,
-	.port_start		= ata_sff_port_start,
 };
 
 static const struct ata_port_info sis_port_info = {
diff --git a/drivers/ata/sata_svw.c b/drivers/ata/sata_svw.c
index c8768396e006..8636f164256e 100644
--- a/drivers/ata/sata_svw.c
+++ b/drivers/ata/sata_svw.c
@@ -334,30 +334,16 @@ static struct scsi_host_template k2_sata_sht = {
 };
 
 
-static const struct ata_port_operations k2_sata_ops = {
+static struct ata_port_operations k2_sata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.tf_load		= k2_sata_tf_load,
 	.tf_read		= k2_sata_tf_read,
 	.check_status		= k2_stat_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
 	.check_atapi_dma	= k2_sata_check_atapi_dma,
 	.bmdma_setup		= k2_bmdma_setup_mmio,
 	.bmdma_start		= k2_bmdma_start_mmio,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
 	.scr_read		= k2_sata_scr_read,
 	.scr_write		= k2_sata_scr_write,
-	.port_start		= ata_sff_port_start,
 };
 
 static const struct ata_port_info k2_port_info[] = {
diff --git a/drivers/ata/sata_sx4.c b/drivers/ata/sata_sx4.c
index 1802f92180e4..8138cda86a66 100644
--- a/drivers/ata/sata_sx4.c
+++ b/drivers/ata/sata_sx4.c
@@ -241,7 +241,8 @@ static struct scsi_host_template pdc_sata_sht = {
 	.dma_boundary		= ATA_DMA_BOUNDARY,
 };
 
-static const struct ata_port_operations pdc_20621_ops = {
+/* TODO: inherit from base port_ops after converting to new EH */
+static struct ata_port_operations pdc_20621_ops = {
 	.tf_load		= pdc_tf_load_mmio,
 	.tf_read		= ata_tf_read,
 	.check_status		= ata_check_status,
diff --git a/drivers/ata/sata_uli.c b/drivers/ata/sata_uli.c
index 764d7064fa59..6ecd13fefa1a 100644
--- a/drivers/ata/sata_uli.c
+++ b/drivers/ata/sata_uli.c
@@ -79,34 +79,10 @@ static struct scsi_host_template uli_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations uli_ops = {
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup            = ata_bmdma_setup,
-	.bmdma_start            = ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.mode_filter		= ata_pci_default_filter,
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
+static struct ata_port_operations uli_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.scr_read		= uli_scr_read,
 	.scr_write		= uli_scr_write,
-
-	.port_start		= ata_sff_port_start,
 };
 
 static const struct ata_port_info uli_port_info = {
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index 9be877cb7f57..6326bcf8ea5d 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -103,97 +103,23 @@ static struct scsi_host_template svia_sht = {
 	ATA_BMDMA_SHT(DRV_NAME),
 };
 
-static const struct ata_port_operations vt6420_sata_ops = {
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup            = ata_bmdma_setup,
-	.bmdma_start            = ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
+static struct ata_port_operations vt6420_sata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.freeze			= svia_noop_freeze,
-	.thaw			= ata_bmdma_thaw,
 	.error_handler		= vt6420_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations vt6421_pata_ops = {
+static struct ata_port_operations vt6421_pata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= vt6421_pata_cable_detect,
 	.set_piomode		= vt6421_set_pio_mode,
 	.set_dmamode		= vt6421_set_dma_mode,
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup            = ata_bmdma_setup,
-	.bmdma_start            = ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.cable_detect		= vt6421_pata_cable_detect,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
 };
 
-static const struct ata_port_operations vt6421_sata_ops = {
-	.mode_filter		= ata_pci_default_filter,
-
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
-	.exec_command		= ata_exec_command,
-	.dev_select		= ata_std_dev_select,
-
-	.bmdma_setup            = ata_bmdma_setup,
-	.bmdma_start            = ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
-
+static struct ata_port_operations vt6421_sata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.scr_read		= svia_scr_read,
 	.scr_write		= svia_scr_write,
-
-	.port_start		= ata_sff_port_start,
 };
 
 static const struct ata_port_info vt6420_port_info = {
diff --git a/drivers/ata/sata_vsc.c b/drivers/ata/sata_vsc.c
index fd6855f0bf48..8045a72dc559 100644
--- a/drivers/ata/sata_vsc.c
+++ b/drivers/ata/sata_vsc.c
@@ -304,29 +304,14 @@ static struct scsi_host_template vsc_sata_sht = {
 };
 
 
-static const struct ata_port_operations vsc_sata_ops = {
+static struct ata_port_operations vsc_sata_ops = {
+	.inherits		= &ata_bmdma_port_ops,
 	.tf_load		= vsc_sata_tf_load,
 	.tf_read		= vsc_sata_tf_read,
-	.exec_command		= ata_exec_command,
-	.check_status		= ata_check_status,
-	.dev_select		= ata_std_dev_select,
-	.bmdma_setup            = ata_bmdma_setup,
-	.bmdma_start            = ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
-	.mode_filter		= ata_pci_default_filter,
 	.freeze			= vsc_freeze,
 	.thaw			= vsc_thaw,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-	.irq_clear		= ata_bmdma_irq_clear,
-	.irq_on			= ata_irq_on,
 	.scr_read		= vsc_sata_scr_read,
 	.scr_write		= vsc_sata_scr_write,
-	.port_start		= ata_sff_port_start,
 };
 
 static void __devinit vsc_sata_setup_port(struct ata_ioports *port,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index eccc38e17568..46aa4ab64891 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -433,7 +433,7 @@ struct ata_host {
 	void __iomem * const	*iomap;
 	unsigned int		n_ports;
 	void			*private_data;
-	const struct ata_port_operations *ops;
+	struct ata_port_operations *ops;
 	unsigned long		flags;
 #ifdef CONFIG_ATA_ACPI
 	acpi_handle		acpi_handle;
@@ -602,7 +602,7 @@ struct ata_link {
 
 struct ata_port {
 	struct Scsi_Host	*scsi_host; /* our co-allocated scsi host */
-	const struct ata_port_operations *ops;
+	struct ata_port_operations *ops;
 	spinlock_t		*lock;
 	unsigned long		flags;	/* ATA_FLAG_xxx */
 	unsigned int		pflags; /* ATA_PFLAG_xxx */
@@ -664,6 +664,13 @@ struct ata_port {
 	u8			sector_buf[ATA_SECT_SIZE]; /* owned by EH */
 };
 
+/* The following initializer overrides a method to NULL whether one of
+ * its parent has the method defined or not.  This is equivalent to
+ * ERR_PTR(-ENOENT).  Unfortunately, ERR_PTR doesn't render a constant
+ * expression and thus can't be used as an initializer.
+ */
+#define ATA_OP_NULL		(void *)(unsigned long)(-ENOENT)
+
 struct ata_port_operations {
 	/*
 	 * Command execution
@@ -733,6 +740,12 @@ struct ata_port_operations {
 	void (*phy_reset)(struct ata_port *ap);
 	void (*eng_timeout)(struct ata_port *ap);
 	irq_handler_t irq_handler;
+
+	/*
+	 * ->inherits must be the last field and all the preceding
+	 * fields must be pointers.
+	 */
+	const struct ata_port_operations	*inherits;
 };
 
 struct ata_port_info {
@@ -742,7 +755,7 @@ struct ata_port_info {
 	unsigned long		pio_mask;
 	unsigned long		mwdma_mask;
 	unsigned long		udma_mask;
-	const struct ata_port_operations *port_ops;
+	struct ata_port_operations *port_ops;
 	irq_handler_t		irq_handler;
 	void 			*private_data;
 };
@@ -765,7 +778,7 @@ extern const unsigned long sata_deb_timing_normal[];
 extern const unsigned long sata_deb_timing_hotplug[];
 extern const unsigned long sata_deb_timing_long[];
 
-extern const struct ata_port_operations ata_dummy_port_ops;
+extern struct ata_port_operations ata_dummy_port_ops;
 extern const struct ata_port_info ata_dummy_port_info;
 
 static inline const unsigned long *
@@ -812,7 +825,7 @@ extern int ata_host_activate(struct ata_host *host, int irq,
 			     struct scsi_host_template *sht);
 extern void ata_host_detach(struct ata_host *host);
 extern void ata_host_init(struct ata_host *, struct device *,
-			  unsigned long, const struct ata_port_operations *);
+			  unsigned long, struct ata_port_operations *);
 extern int ata_scsi_detect(struct scsi_host_template *sht);
 extern int ata_scsi_ioctl(struct scsi_device *dev, int cmd, void __user *arg);
 extern int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *));
-- 
cgit v1.2.3


From 1bd5b715a305f6f13455e89becbd839010dd14b5 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:49 +0900
Subject: libata: make ata_pci_init_one() not use ops->irq_handler and pi->sht

ata_pci_init_one() is the only function which uses ops->irq_handler
and pi->sht.  Other initialization functions take the same information
as arguments.  This causes confusion and duplicate unused entries in
structures.

Make ata_pci_init_one() take sht as an argument and use ata_interrupt
implicitly.  All current users use ata_interrupt and if different irq
handler is necessary open coding ata_pci_init_one() using
ata_prepare_sff_host() and ata_activate_sff_host can be done under ten
lines including error handling and driver which requires custom
interrupt handler is likely to require custom initialization anyway.

As ata_pci_init_one() was the last user of ops->irq_handler, this
patch also kills the field.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ata_generic.c       |  3 +--
 drivers/ata/libata-core.c       |  1 -
 drivers/ata/libata-sff.c        |  7 ++++---
 drivers/ata/pata_acpi.c         |  3 +--
 drivers/ata/pata_ali.c          |  9 +--------
 drivers/ata/pata_amd.c          | 12 +-----------
 drivers/ata/pata_artop.c        |  6 +-----
 drivers/ata/pata_atiixp.c       |  3 +--
 drivers/ata/pata_cmd640.c       |  3 +--
 drivers/ata/pata_cmd64x.c       |  8 +-------
 drivers/ata/pata_cs5530.c       |  4 +---
 drivers/ata/pata_cs5535.c       |  3 +--
 drivers/ata/pata_cs5536.c       |  3 +--
 drivers/ata/pata_cypress.c      |  3 +--
 drivers/ata/pata_efar.c         |  3 +--
 drivers/ata/pata_hpt366.c       |  3 +--
 drivers/ata/pata_hpt37x.c       |  8 +-------
 drivers/ata/pata_hpt3x2n.c      |  3 +--
 drivers/ata/pata_it8213.c       |  3 +--
 drivers/ata/pata_it821x.c       |  4 +---
 drivers/ata/pata_jmicron.c      |  3 +--
 drivers/ata/pata_marvell.c      |  4 +---
 drivers/ata/pata_netcell.c      |  3 +--
 drivers/ata/pata_ns87410.c      |  3 +--
 drivers/ata/pata_ns87415.c      |  4 +---
 drivers/ata/pata_oldpiix.c      |  3 +--
 drivers/ata/pata_opti.c         |  3 +--
 drivers/ata/pata_optidma.c      |  4 +---
 drivers/ata/pata_pdc202xx_old.c |  5 +----
 drivers/ata/pata_radisys.c      |  3 +--
 drivers/ata/pata_rz1000.c       |  3 +--
 drivers/ata/pata_sc1200.c       |  3 +--
 drivers/ata/pata_serverworks.c  |  6 +-----
 drivers/ata/pata_sil680.c       |  4 +---
 drivers/ata/pata_sis.c          | 10 +---------
 drivers/ata/pata_sl82c105.c     |  4 +---
 drivers/ata/pata_triflex.c      |  3 +--
 drivers/ata/pata_via.c          |  8 +-------
 include/linux/libata.h          |  4 ++--
 39 files changed, 42 insertions(+), 130 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ata_generic.c b/drivers/ata/ata_generic.c
index 0b5b515ae159..a912ee01a47c 100644
--- a/drivers/ata/ata_generic.c
+++ b/drivers/ata/ata_generic.c
@@ -120,7 +120,6 @@ static int ata_generic_init_one(struct pci_dev *dev, const struct pci_device_id
 {
 	u16 command;
 	static const struct ata_port_info info = {
-		.sht = &generic_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -153,7 +152,7 @@ static int ata_generic_init_one(struct pci_dev *dev, const struct pci_device_id
 	if (dev->vendor == PCI_VENDOR_ID_AL)
 	    	ata_pci_clear_simplex(dev);
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &generic_sht);
 }
 
 static struct pci_device_id ata_generic[] = {
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 32fa9ee397b6..abbe3229480c 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -109,7 +109,6 @@ const struct ata_port_operations ata_sff_port_ops = {
 	.irq_on			= ata_irq_on,
 
 	.port_start		= ata_sff_port_start,
-	.irq_handler		= ata_interrupt,
 };
 
 const struct ata_port_operations ata_bmdma_port_ops = {
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 1cf03d41aa33..a9d5898cbbc4 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -826,6 +826,7 @@ int ata_pci_activate_sff_host(struct ata_host *host,
  *	ata_pci_init_one - Initialize/register PCI IDE host controller
  *	@pdev: Controller to be initialized
  *	@ppi: array of port_info, must be enough for two ports
+ *	@sht: scsi_host_template to use when registering the host
  *
  *	This is a helper function which can be called from a driver's
  *	xxx_init_one() probe function if the hardware uses traditional
@@ -846,7 +847,8 @@ int ata_pci_activate_sff_host(struct ata_host *host,
  *	Zero on success, negative on errno-based value on error.
  */
 int ata_pci_init_one(struct pci_dev *pdev,
-		     const struct ata_port_info * const * ppi)
+		     const struct ata_port_info * const * ppi,
+		     struct scsi_host_template *sht)
 {
 	struct device *dev = &pdev->dev;
 	const struct ata_port_info *pi = NULL;
@@ -882,8 +884,7 @@ int ata_pci_init_one(struct pci_dev *pdev,
 		goto out;
 
 	pci_set_master(pdev);
-	rc = ata_pci_activate_sff_host(host, pi->port_ops->irq_handler,
-				       pi->sht);
+	rc = ata_pci_activate_sff_host(host, ata_interrupt, sht);
  out:
 	if (rc == 0)
 		devres_remove_group(&pdev->dev, NULL);
diff --git a/drivers/ata/pata_acpi.c b/drivers/ata/pata_acpi.c
index 35ad488db6ed..3edde51750da 100644
--- a/drivers/ata/pata_acpi.c
+++ b/drivers/ata/pata_acpi.c
@@ -264,7 +264,6 @@ static struct ata_port_operations pacpi_ops = {
 static int pacpi_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht		= &pacpi_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
 
 		.pio_mask	= 0x1f,
@@ -274,7 +273,7 @@ static int pacpi_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 		.port_ops	= &pacpi_ops,
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &pacpi_sht);
 }
 
 static const struct pci_device_id pacpi_pci_tbl[] = {
diff --git a/drivers/ata/pata_ali.c b/drivers/ata/pata_ali.c
index b00a9cf72c31..f2924996f6e3 100644
--- a/drivers/ata/pata_ali.c
+++ b/drivers/ata/pata_ali.c
@@ -463,14 +463,12 @@ static void ali_init_chipset(struct pci_dev *pdev)
 static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info_early = {
-		.sht = &ali_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.port_ops = &ali_early_port_ops
 	};
 	/* Revision 0x20 added DMA */
 	static const struct ata_port_info info_20 = {
-		.sht = &ali_sht,
 		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_PIO_LBA48,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -478,7 +476,6 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* Revision 0x20 with support logic added UDMA */
 	static const struct ata_port_info info_20_udma = {
-		.sht = &ali_sht,
 		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_PIO_LBA48,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -487,7 +484,6 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* Revision 0xC2 adds UDMA66 */
 	static const struct ata_port_info info_c2 = {
-		.sht = &ali_sht,
 		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_PIO_LBA48,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -496,7 +492,6 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* Revision 0xC3 is UDMA66 for now */
 	static const struct ata_port_info info_c3 = {
-		.sht = &ali_sht,
 		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_PIO_LBA48,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -505,7 +500,6 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* Revision 0xC4 is UDMA100 */
 	static const struct ata_port_info info_c4 = {
-		.sht = &ali_sht,
 		.flags = ATA_FLAG_SLAVE_POSS | ATA_FLAG_PIO_LBA48,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -514,7 +508,6 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* Revision 0xC5 is UDMA133 with LBA48 DMA */
 	static const struct ata_port_info info_c5 = {
-		.sht = &ali_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -559,7 +552,7 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	        	ppi[0] = &info_20_udma;
 		pci_dev_put(isa_bridge);
 	}
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &ali_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_amd.c b/drivers/ata/pata_amd.c
index b0cb4eaf273c..644702cac6ee 100644
--- a/drivers/ata/pata_amd.c
+++ b/drivers/ata/pata_amd.c
@@ -413,7 +413,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info[10] = {
 		{	/* 0: AMD 7401 */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,	/* No SWDMA */
@@ -421,7 +420,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd33_port_ops
 		},
 		{	/* 1: Early AMD7409 - no swdma */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -429,7 +427,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd66_port_ops
 		},
 		{	/* 2: AMD 7409, no swdma errata */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -437,7 +434,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd66_port_ops
 		},
 		{	/* 3: AMD 7411 */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -445,7 +441,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd100_port_ops
 		},
 		{	/* 4: AMD 7441 */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -453,7 +448,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd100_port_ops
 		},
 		{	/* 5: AMD 8111*/
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -461,7 +455,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd133_port_ops
 		},
 		{	/* 6: AMD 8111 UDMA 100 (Serenade) */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -469,7 +462,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd133_port_ops
 		},
 		{	/* 7: Nvidia Nforce */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -477,7 +469,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &nv100_port_ops
 		},
 		{	/* 8: Nvidia Nforce2 and later */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -485,7 +476,6 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &nv133_port_ops
 		},
 		{	/* 9: AMD CS5536 (Geode companion) */
-			.sht = &amd_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -544,7 +534,7 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	}
 
 	/* And fire it up */
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &amd_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c
index 0101e5aef3e0..698a53c96111 100644
--- a/drivers/ata/pata_artop.c
+++ b/drivers/ata/pata_artop.c
@@ -352,7 +352,6 @@ static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static int printed_version;
 	static const struct ata_port_info info_6210 = {
-		.sht		= &artop_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -360,7 +359,6 @@ static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 		.port_ops	= &artop6210_ops,
 	};
 	static const struct ata_port_info info_626x = {
-		.sht		= &artop_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -368,7 +366,6 @@ static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 		.port_ops	= &artop6260_ops,
 	};
 	static const struct ata_port_info info_628x = {
-		.sht		= &artop_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -376,7 +373,6 @@ static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 		.port_ops	= &artop6260_ops,
 	};
 	static const struct ata_port_info info_628x_fast = {
-		.sht		= &artop_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -434,7 +430,7 @@ static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 
 	BUG_ON(ppi[0] == NULL);
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &artop_sht);
 }
 
 static const struct pci_device_id artop_pci_tbl[] = {
diff --git a/drivers/ata/pata_atiixp.c b/drivers/ata/pata_atiixp.c
index 2655f6a17ad3..6fe433ba62bd 100644
--- a/drivers/ata/pata_atiixp.c
+++ b/drivers/ata/pata_atiixp.c
@@ -241,7 +241,6 @@ static struct ata_port_operations atiixp_port_ops = {
 static int atiixp_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &atiixp_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x06,	/* No MWDMA0 support */
@@ -249,7 +248,7 @@ static int atiixp_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.port_ops = &atiixp_port_ops
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &atiixp_sht);
 }
 
 static const struct pci_device_id atiixp[] = {
diff --git a/drivers/ata/pata_cmd640.c b/drivers/ata/pata_cmd640.c
index 061c891c8a66..efd2bb5747b4 100644
--- a/drivers/ata/pata_cmd640.c
+++ b/drivers/ata/pata_cmd640.c
@@ -211,7 +211,6 @@ static void cmd640_hardware_init(struct pci_dev *pdev)
 static int cmd640_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &cmd640_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.port_ops = &cmd640_port_ops
@@ -225,7 +224,7 @@ static int cmd640_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	cmd640_hardware_init(pdev);
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &cmd640_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cmd64x.c b/drivers/ata/pata_cmd64x.c
index 1ac8ecfb97e2..bfd72ef9cd31 100644
--- a/drivers/ata/pata_cmd64x.c
+++ b/drivers/ata/pata_cmd64x.c
@@ -298,21 +298,18 @@ static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	static const struct ata_port_info cmd_info[6] = {
 		{	/* CMD 643 - no UDMA */
-			.sht = &cmd64x_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
 			.port_ops = &cmd64x_port_ops
 		},
 		{	/* CMD 646 with broken UDMA */
-			.sht = &cmd64x_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
 			.port_ops = &cmd64x_port_ops
 		},
 		{	/* CMD 646 with working UDMA */
-			.sht = &cmd64x_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -320,14 +317,12 @@ static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &cmd64x_port_ops
 		},
 		{	/* CMD 646 rev 1  */
-			.sht = &cmd64x_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
 			.port_ops = &cmd646r1_port_ops
 		},
 		{	/* CMD 648 */
-			.sht = &cmd64x_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -335,7 +330,6 @@ static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &cmd648_port_ops
 		},
 		{	/* CMD 649 */
-			.sht = &cmd64x_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -379,7 +373,7 @@ static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	pci_write_config_byte(pdev, UDIDETCR0, 0xF0);
 #endif
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &cmd64x_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cs5530.c b/drivers/ata/pata_cs5530.c
index e4a16a578cac..c632ce499d33 100644
--- a/drivers/ata/pata_cs5530.c
+++ b/drivers/ata/pata_cs5530.c
@@ -298,7 +298,6 @@ fail_put:
 static int cs5530_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &cs5530_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -307,7 +306,6 @@ static int cs5530_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* The docking connector doesn't do UDMA, and it seems not MWDMA */
 	static const struct ata_port_info info_palmax_secondary = {
-		.sht = &cs5530_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.port_ops = &cs5530_port_ops
@@ -327,7 +325,7 @@ static int cs5530_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 		ppi[1] = &info_palmax_secondary;
 
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &cs5530_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cs5535.c b/drivers/ata/pata_cs5535.c
index f910a8aa7437..d78cf95cbe45 100644
--- a/drivers/ata/pata_cs5535.c
+++ b/drivers/ata/pata_cs5535.c
@@ -181,7 +181,6 @@ static struct ata_port_operations cs5535_port_ops = {
 static int cs5535_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &cs5535_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -200,7 +199,7 @@ static int cs5535_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	rdmsr(ATAC_CH0D1_PIO, timings, dummy);
 	if (CS5535_BAD_PIO(timings))
 		wrmsr(ATAC_CH0D1_PIO, 0xF7F4F7F4UL, 0);
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &cs5535_sht);
 }
 
 static const struct pci_device_id cs5535[] = {
diff --git a/drivers/ata/pata_cs5536.c b/drivers/ata/pata_cs5536.c
index 075ee6a7be39..f7c0e4e319ed 100644
--- a/drivers/ata/pata_cs5536.c
+++ b/drivers/ata/pata_cs5536.c
@@ -241,7 +241,6 @@ static struct ata_port_operations cs5536_port_ops = {
 static int cs5536_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &cs5536_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -262,7 +261,7 @@ static int cs5536_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		return -ENODEV;
 	}
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &cs5536_sht);
 }
 
 static const struct pci_device_id cs5536[] = {
diff --git a/drivers/ata/pata_cypress.c b/drivers/ata/pata_cypress.c
index c459553e7d1e..cbd6670ea0de 100644
--- a/drivers/ata/pata_cypress.c
+++ b/drivers/ata/pata_cypress.c
@@ -123,7 +123,6 @@ static struct ata_port_operations cy82c693_port_ops = {
 static int cy82c693_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &cy82c693_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -137,7 +136,7 @@ static int cy82c693_init_one(struct pci_dev *pdev, const struct pci_device_id *i
 	if (PCI_FUNC(pdev->devfn) != 1)
 		return -ENODEV;
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &cy82c693_sht);
 }
 
 static const struct pci_device_id cy82c693[] = {
diff --git a/drivers/ata/pata_efar.c b/drivers/ata/pata_efar.c
index ef62fc642c17..0260edac2370 100644
--- a/drivers/ata/pata_efar.c
+++ b/drivers/ata/pata_efar.c
@@ -263,7 +263,6 @@ static int efar_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	static int printed_version;
 	static const struct ata_port_info info = {
-		.sht		= &efar_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma1-2 */
@@ -276,7 +275,7 @@ static int efar_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &efar_sht);
 }
 
 static const struct pci_device_id efar_pci_tbl[] = {
diff --git a/drivers/ata/pata_hpt366.c b/drivers/ata/pata_hpt366.c
index 788955f57ff8..b62d398ed84b 100644
--- a/drivers/ata/pata_hpt366.c
+++ b/drivers/ata/pata_hpt366.c
@@ -350,7 +350,6 @@ static void hpt36x_init_chipset(struct pci_dev *dev)
 static int hpt36x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info_hpt366 = {
-		.sht = &hpt36x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -394,7 +393,7 @@ static int hpt36x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 			break;
 	}
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &hpt36x_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index c42eec70d297..a43c19753669 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -783,7 +783,6 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	/* HPT370 - UDMA100 */
 	static const struct ata_port_info info_hpt370 = {
-		.sht = &hpt37x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -792,7 +791,6 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	};
 	/* HPT370A - UDMA100 */
 	static const struct ata_port_info info_hpt370a = {
-		.sht = &hpt37x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -801,7 +799,6 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	};
 	/* HPT370 - UDMA100 */
 	static const struct ata_port_info info_hpt370_33 = {
-		.sht = &hpt37x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -810,7 +807,6 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	};
 	/* HPT370A - UDMA100 */
 	static const struct ata_port_info info_hpt370a_33 = {
-		.sht = &hpt37x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -819,7 +815,6 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	};
 	/* HPT371, 372 and friends - UDMA133 */
 	static const struct ata_port_info info_hpt372 = {
-		.sht = &hpt37x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -828,7 +823,6 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	};
 	/* HPT374 - UDMA100 */
 	static const struct ata_port_info info_hpt374 = {
-		.sht = &hpt37x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -1051,7 +1045,7 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	port_info = *port;
 	port_info.private_data = private_data;
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &hpt37x_sht);
 }
 
 static const struct pci_device_id hpt37x[] = {
diff --git a/drivers/ata/pata_hpt3x2n.c b/drivers/ata/pata_hpt3x2n.c
index b77b1279d757..2c178c30116c 100644
--- a/drivers/ata/pata_hpt3x2n.c
+++ b/drivers/ata/pata_hpt3x2n.c
@@ -452,7 +452,6 @@ static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	/* HPT372N and friends - UDMA133 */
 	static const struct ata_port_info info = {
-		.sht = &hpt3x2n_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -568,7 +567,7 @@ static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	}
 
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &hpt3x2n_sht);
 }
 
 static const struct pci_device_id hpt3x2n[] = {
diff --git a/drivers/ata/pata_it8213.c b/drivers/ata/pata_it8213.c
index 9ce89522e764..291a0d6e2434 100644
--- a/drivers/ata/pata_it8213.c
+++ b/drivers/ata/pata_it8213.c
@@ -274,7 +274,6 @@ static int it8213_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 {
 	static int printed_version;
 	static const struct ata_port_info info = {
-		.sht		= &it8213_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -288,7 +287,7 @@ static int it8213_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &it8213_sht);
 }
 
 static const struct pci_device_id it8213_pci_tbl[] = {
diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c
index 669d224d30ca..63c5cf0d1fee 100644
--- a/drivers/ata/pata_it821x.c
+++ b/drivers/ata/pata_it821x.c
@@ -687,14 +687,12 @@ static int it821x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	u8 conf;
 
 	static const struct ata_port_info info_smart = {
-		.sht = &it821x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
 		.port_ops = &it821x_smart_port_ops
 	};
 	static const struct ata_port_info info_passthru = {
-		.sht = &it821x_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -724,7 +722,7 @@ static int it821x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	else
 		ppi[0] = &info_smart;
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &it821x_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_jmicron.c b/drivers/ata/pata_jmicron.c
index 61ff5c6b4568..859e47a600cc 100644
--- a/drivers/ata/pata_jmicron.c
+++ b/drivers/ata/pata_jmicron.c
@@ -148,7 +148,6 @@ static struct ata_port_operations jmicron_ops = {
 static int jmicron_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht		= &jmicron_sht,
 		.flags	= ATA_FLAG_SLAVE_POSS,
 
 		.pio_mask	= 0x1f,
@@ -159,7 +158,7 @@ static int jmicron_init_one (struct pci_dev *pdev, const struct pci_device_id *i
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &jmicron_sht);
 }
 
 static const struct pci_device_id jmicron_pci_tbl[] = {
diff --git a/drivers/ata/pata_marvell.c b/drivers/ata/pata_marvell.c
index 286310fc5910..d8da4f344c0a 100644
--- a/drivers/ata/pata_marvell.c
+++ b/drivers/ata/pata_marvell.c
@@ -119,7 +119,6 @@ static struct ata_port_operations marvell_ops = {
 static int marvell_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht		= &marvell_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 
 		.pio_mask	= 0x1f,
@@ -129,7 +128,6 @@ static int marvell_init_one (struct pci_dev *pdev, const struct pci_device_id *i
 		.port_ops	= &marvell_ops,
 	};
 	static const struct ata_port_info info_sata = {
-		.sht		= &marvell_sht,
 		/* Slave possible as its magically mapped not real */
 		.flags		= ATA_FLAG_SLAVE_POSS,
 
@@ -144,7 +142,7 @@ static int marvell_init_one (struct pci_dev *pdev, const struct pci_device_id *i
 	if (pdev->device == 0x6101)
 		ppi[1] = &ata_dummy_port_info;
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &marvell_sht);
 }
 
 static const struct pci_device_id marvell_pci_tbl[] = {
diff --git a/drivers/ata/pata_netcell.c b/drivers/ata/pata_netcell.c
index 65389d1837b3..ae50a5e85cf1 100644
--- a/drivers/ata/pata_netcell.c
+++ b/drivers/ata/pata_netcell.c
@@ -48,7 +48,6 @@ static int netcell_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 {
 	static int printed_version;
 	static const struct ata_port_info info = {
-		.sht		= &netcell_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		/* Actually we don't really care about these as the
 		   firmware deals with it */
@@ -72,7 +71,7 @@ static int netcell_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	ata_pci_clear_simplex(pdev);
 
 	/* And let the library code do the work */
-	return ata_pci_init_one(pdev, port_info);
+	return ata_pci_init_one(pdev, port_info, &netcell_sht);
 }
 
 static const struct pci_device_id netcell_pci_tbl[] = {
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
index 5b1982fa0be1..1bdca8f1e767 100644
--- a/drivers/ata/pata_ns87410.c
+++ b/drivers/ata/pata_ns87410.c
@@ -158,13 +158,12 @@ static struct ata_port_operations ns87410_port_ops = {
 static int ns87410_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &ns87410_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x0F,
 		.port_ops = &ns87410_port_ops
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &ns87410_sht);
 }
 
 static const struct pci_device_id ns87410[] = {
diff --git a/drivers/ata/pata_ns87415.c b/drivers/ata/pata_ns87415.c
index 38d86a262dbb..42508940e4a9 100644
--- a/drivers/ata/pata_ns87415.c
+++ b/drivers/ata/pata_ns87415.c
@@ -345,7 +345,6 @@ static int ns87415_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 {
 	static int printed_version;
 	static const struct ata_port_info info = {
-		.sht		= &ns87415_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -355,7 +354,6 @@ static int ns87415_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	int rc;
 #if defined(CONFIG_SUPERIO)
 	static const struct ata_port_info info87560 = {
-		.sht		= &ns87415_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma0-2 */
@@ -377,7 +375,7 @@ static int ns87415_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	pci_write_config_byte(pdev, 0x55, 0xEE);
 	/* Select PIO0 8bit clocking */
 	pci_write_config_byte(pdev, 0x54, 0xB7);
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &ns87415_sht);
 }
 
 static const struct pci_device_id ns87415_pci_tbl[] = {
diff --git a/drivers/ata/pata_oldpiix.c b/drivers/ata/pata_oldpiix.c
index f6062b37310d..9e3afadbd04a 100644
--- a/drivers/ata/pata_oldpiix.c
+++ b/drivers/ata/pata_oldpiix.c
@@ -252,7 +252,6 @@ static int oldpiix_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 {
 	static int printed_version;
 	static const struct ata_port_info info = {
-		.sht		= &oldpiix_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma1-2 */
@@ -264,7 +263,7 @@ static int oldpiix_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &oldpiix_sht);
 }
 
 static const struct pci_device_id oldpiix_pci_tbl[] = {
diff --git a/drivers/ata/pata_opti.c b/drivers/ata/pata_opti.c
index c4a0795c3ff4..8601d9c3cb39 100644
--- a/drivers/ata/pata_opti.c
+++ b/drivers/ata/pata_opti.c
@@ -178,7 +178,6 @@ static struct ata_port_operations opti_port_ops = {
 static int opti_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &opti_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.port_ops = &opti_port_ops
@@ -189,7 +188,7 @@ static int opti_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (!printed_version++)
 		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &opti_sht);
 }
 
 static const struct pci_device_id opti[] = {
diff --git a/drivers/ata/pata_optidma.c b/drivers/ata/pata_optidma.c
index eb4b08190e3a..c376f9ef77c8 100644
--- a/drivers/ata/pata_optidma.c
+++ b/drivers/ata/pata_optidma.c
@@ -414,14 +414,12 @@ done_nomsg:		/* Wrong chip revision */
 static int optidma_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info_82c700 = {
-		.sht = &optidma_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
 		.port_ops = &optidma_port_ops
 	};
 	static const struct ata_port_info info_82c700_udma = {
-		.sht = &optidma_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -447,7 +445,7 @@ static int optidma_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (optiplus_with_udma(dev))
 		ppi[0] = &info_82c700_udma;
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &optidma_sht);
 }
 
 static const struct pci_device_id optidma[] = {
diff --git a/drivers/ata/pata_pdc202xx_old.c b/drivers/ata/pata_pdc202xx_old.c
index 4daac20df0bc..5545fbab6a7e 100644
--- a/drivers/ata/pata_pdc202xx_old.c
+++ b/drivers/ata/pata_pdc202xx_old.c
@@ -290,7 +290,6 @@ static int pdc202xx_init_one(struct pci_dev *dev, const struct pci_device_id *id
 {
 	static const struct ata_port_info info[3] = {
 		{
-			.sht = &pdc202xx_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -298,7 +297,6 @@ static int pdc202xx_init_one(struct pci_dev *dev, const struct pci_device_id *id
 			.port_ops = &pdc2024x_port_ops
 		},
 		{
-			.sht = &pdc202xx_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -306,7 +304,6 @@ static int pdc202xx_init_one(struct pci_dev *dev, const struct pci_device_id *id
 			.port_ops = &pdc2026x_port_ops
 		},
 		{
-			.sht = &pdc202xx_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -327,7 +324,7 @@ static int pdc202xx_init_one(struct pci_dev *dev, const struct pci_device_id *id
 				return -ENODEV;
 		}
 	}
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &pdc202xx_sht);
 }
 
 static const struct pci_device_id pdc202xx[] = {
diff --git a/drivers/ata/pata_radisys.c b/drivers/ata/pata_radisys.c
index 94e60b3a1ec6..145d5ba92795 100644
--- a/drivers/ata/pata_radisys.c
+++ b/drivers/ata/pata_radisys.c
@@ -216,7 +216,6 @@ static int radisys_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 {
 	static int printed_version;
 	static const struct ata_port_info info = {
-		.sht		= &radisys_sht,
 		.flags		= ATA_FLAG_SLAVE_POSS,
 		.pio_mask	= 0x1f,	/* pio0-4 */
 		.mwdma_mask	= 0x07, /* mwdma1-2 */
@@ -229,7 +228,7 @@ static int radisys_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &radisys_sht);
 }
 
 static const struct pci_device_id radisys_pci_tbl[] = {
diff --git a/drivers/ata/pata_rz1000.c b/drivers/ata/pata_rz1000.c
index a2aef7328bfc..04be6aee4354 100644
--- a/drivers/ata/pata_rz1000.c
+++ b/drivers/ata/pata_rz1000.c
@@ -89,7 +89,6 @@ static int rz1000_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 {
 	static int printed_version;
 	static const struct ata_port_info info = {
-		.sht = &rz1000_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.port_ops = &rz1000_port_ops
@@ -100,7 +99,7 @@ static int rz1000_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 		printk(KERN_DEBUG DRV_NAME " version " DRV_VERSION "\n");
 
 	if (rz1000_fifo_disable(pdev) == 0)
-		return ata_pci_init_one(pdev, ppi);
+		return ata_pci_init_one(pdev, ppi, &rz1000_sht);
 
 	printk(KERN_ERR DRV_NAME ": failed to disable read-ahead on chipset..\n");
 	/* Not safe to use so skip */
diff --git a/drivers/ata/pata_sc1200.c b/drivers/ata/pata_sc1200.c
index 362b7f829d8e..38c7fb0bebe9 100644
--- a/drivers/ata/pata_sc1200.c
+++ b/drivers/ata/pata_sc1200.c
@@ -204,7 +204,6 @@ static struct ata_port_operations sc1200_port_ops = {
 static int sc1200_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &sc1200_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -214,7 +213,7 @@ static int sc1200_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	/* Can't enable port 2 yet, see top comments */
 	const struct ata_port_info *ppi[] = { &info, &ata_dummy_port_info };
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &sc1200_sht);
 }
 
 static const struct pci_device_id sc1200[] = {
diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
index 627abcf85c6e..515b5b70a555 100644
--- a/drivers/ata/pata_serverworks.c
+++ b/drivers/ata/pata_serverworks.c
@@ -399,28 +399,24 @@ static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id
 {
 	static const struct ata_port_info info[4] = {
 		{ /* OSB4 */
-			.sht = &serverworks_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
 			.udma_mask = 0x07,
 			.port_ops = &serverworks_osb4_port_ops
 		}, { /* OSB4 no UDMA */
-			.sht = &serverworks_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
 			.udma_mask = 0x00,
 			.port_ops = &serverworks_osb4_port_ops
 		}, { /* CSB5 */
-			.sht = &serverworks_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
 			.udma_mask = ATA_UDMA4,
 			.port_ops = &serverworks_csb_port_ops
 		}, { /* CSB5 - later revisions*/
-			.sht = &serverworks_sht,
 			.flags = ATA_FLAG_SLAVE_POSS,
 			.pio_mask = 0x1f,
 			.mwdma_mask = 0x07,
@@ -465,7 +461,7 @@ static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id
 	if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE)
 		ata_pci_clear_simplex(pdev);
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &serverworks_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_sil680.c b/drivers/ata/pata_sil680.c
index 0936f534d9c7..5313deeffa6d 100644
--- a/drivers/ata/pata_sil680.c
+++ b/drivers/ata/pata_sil680.c
@@ -282,7 +282,6 @@ static int __devinit sil680_init_one(struct pci_dev *pdev,
 				     const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &sil680_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -290,7 +289,6 @@ static int __devinit sil680_init_one(struct pci_dev *pdev,
 		.port_ops = &sil680_port_ops
 	};
 	static const struct ata_port_info info_slow = {
-		.sht = &sil680_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -364,7 +362,7 @@ static int __devinit sil680_init_one(struct pci_dev *pdev,
 				 &sil680_sht);
 
 use_ioports:
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &sil680_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
index 3ed628670cd7..32be13ba5f06 100644
--- a/drivers/ata/pata_sis.c
+++ b/drivers/ata/pata_sis.c
@@ -565,7 +565,6 @@ static struct ata_port_operations sis_old_ops = {
 };
 
 static const struct ata_port_info sis_info = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.mwdma_mask	= 0x07,
@@ -573,7 +572,6 @@ static const struct ata_port_info sis_info = {
 	.port_ops	= &sis_old_ops,
 };
 static const struct ata_port_info sis_info33 = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.mwdma_mask	= 0x07,
@@ -581,42 +579,36 @@ static const struct ata_port_info sis_info33 = {
 	.port_ops	= &sis_old_ops,
 };
 static const struct ata_port_info sis_info66 = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.udma_mask	= ATA_UDMA4,	/* UDMA 66 */
 	.port_ops	= &sis_66_ops,
 };
 static const struct ata_port_info sis_info100 = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.udma_mask	= ATA_UDMA5,
 	.port_ops	= &sis_100_ops,
 };
 static const struct ata_port_info sis_info100_early = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS,
 	.udma_mask	= ATA_UDMA5,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.port_ops	= &sis_66_ops,
 };
 static const struct ata_port_info sis_info133 = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.udma_mask	= ATA_UDMA6,
 	.port_ops	= &sis_133_ops,
 };
 const struct ata_port_info sis_info133_for_sata = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS | ATA_FLAG_SRST,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.udma_mask	= ATA_UDMA6,
 	.port_ops	= &sis_133_for_sata_ops,
 };
 static const struct ata_port_info sis_info133_early = {
-	.sht		= &sis_sht,
 	.flags		= ATA_FLAG_SLAVE_POSS,
 	.pio_mask	= 0x1f,	/* pio0-4 */
 	.udma_mask	= ATA_UDMA6,
@@ -844,7 +836,7 @@ static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	sis_fixup(pdev, chipset);
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &sis_sht);
 }
 
 static const struct pci_device_id sis_pci_tbl[] = {
diff --git a/drivers/ata/pata_sl82c105.c b/drivers/ata/pata_sl82c105.c
index 0dd8e2f69558..2d14b2505c7d 100644
--- a/drivers/ata/pata_sl82c105.c
+++ b/drivers/ata/pata_sl82c105.c
@@ -289,14 +289,12 @@ static int sl82c105_bridge_revision(struct pci_dev *pdev)
 static int sl82c105_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info_dma = {
-		.sht = &sl82c105_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
 		.port_ops = &sl82c105_port_ops
 	};
 	static const struct ata_port_info info_early = {
-		.sht = &sl82c105_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.port_ops = &sl82c105_port_ops
@@ -325,7 +323,7 @@ static int sl82c105_init_one(struct pci_dev *dev, const struct pci_device_id *id
 	val |= CTRL_P0EN | CTRL_P0F16 | CTRL_P1F16;
 	pci_write_config_dword(dev, 0x40, val);
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &sl82c105_sht);
 }
 
 static const struct pci_device_id sl82c105[] = {
diff --git a/drivers/ata/pata_triflex.c b/drivers/ata/pata_triflex.c
index bc4956ef0931..86dc66c37389 100644
--- a/drivers/ata/pata_triflex.c
+++ b/drivers/ata/pata_triflex.c
@@ -195,7 +195,6 @@ static struct ata_port_operations triflex_port_ops = {
 static int triflex_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 {
 	static const struct ata_port_info info = {
-		.sht = &triflex_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -207,7 +206,7 @@ static int triflex_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (!printed_version++)
 		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(dev, ppi);
+	return ata_pci_init_one(dev, ppi, &triflex_sht);
 }
 
 static const struct pci_device_id triflex[] = {
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
index d1edb1b27480..e66bb85ad3d1 100644
--- a/drivers/ata/pata_via.c
+++ b/drivers/ata/pata_via.c
@@ -398,7 +398,6 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 {
 	/* Early VIA without UDMA support */
 	static const struct ata_port_info via_mwdma_info = {
-		.sht = &via_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -406,7 +405,6 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* Ditto with IRQ masking required */
 	static const struct ata_port_info via_mwdma_info_borked = {
-		.sht = &via_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -414,7 +412,6 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* VIA UDMA 33 devices (and borked 66) */
 	static const struct ata_port_info via_udma33_info = {
-		.sht = &via_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -423,7 +420,6 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* VIA UDMA 66 devices */
 	static const struct ata_port_info via_udma66_info = {
-		.sht = &via_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -432,7 +428,6 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* VIA UDMA 100 devices */
 	static const struct ata_port_info via_udma100_info = {
-		.sht = &via_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -441,7 +436,6 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	};
 	/* UDMA133 with bad AST (All current 133) */
 	static const struct ata_port_info via_udma133_info = {
-		.sht = &via_sht,
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
@@ -532,7 +526,7 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	/* We have established the device type, now fire it up */
 	type.private_data = (void *)config;
 
-	return ata_pci_init_one(pdev, ppi);
+	return ata_pci_init_one(pdev, ppi, &via_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 46aa4ab64891..5494119854de 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -739,7 +739,6 @@ struct ata_port_operations {
 	 */
 	void (*phy_reset)(struct ata_port *ap);
 	void (*eng_timeout)(struct ata_port *ap);
-	irq_handler_t irq_handler;
 
 	/*
 	 * ->inherits must be the last field and all the preceding
@@ -1020,7 +1019,8 @@ static inline int ata_acpi_cbl_80wire(struct ata_port *ap,
 struct pci_dev;
 
 extern int ata_pci_init_one(struct pci_dev *pdev,
-			     const struct ata_port_info * const * ppi);
+			    const struct ata_port_info * const * ppi,
+			    struct scsi_host_template *sht);
 extern void ata_pci_remove_one(struct pci_dev *pdev);
 #ifdef CONFIG_PM
 extern void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t mesg);
-- 
cgit v1.2.3


From 887125e3740283be25564bfc6fb5d24974b651ab Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:49 +0900
Subject: libata: stop overloading port_info->private_data

port_info->private_data is currently used for two purposes - to record
private data about the port_info or to specify host->private_data to
use when allocating ata_host.

This overloading is confusing and counter-intuitive in that
port_info->private_data becomes host->private_data instead of
port->private_data.  In addition, port_info and host don't correspond
to each other 1-to-1.  Currently, the first non-NULL
port_info->private_data is used.

This patch makes port_info->private_data just be what it is -
private_data for the port_info where LLD can jot down extra info.
libata no longer sets host->private_data to the first non-NULL
port_info->private_data, @host_priv argument is added to
ata_pci_init_one() instead.  LLDs which use ata_pci_init_one() can use
this argument to pass in pointer to host private data.  LLDs which
don't should use init-register model anyway and can initialize
host->private_data directly.

Adding @host_priv instead of using init-register model for LLDs which
use ata_pci_init_one() is suggested by Alan Cox.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
---
 drivers/ata/ata_generic.c       |  2 +-
 drivers/ata/libata-core.c       |  2 --
 drivers/ata/libata-sff.c        |  4 +++-
 drivers/ata/pata_acpi.c         | 18 ++----------------
 drivers/ata/pata_ali.c          |  2 +-
 drivers/ata/pata_amd.c          | 27 +++++++--------------------
 drivers/ata/pata_artop.c        |  2 +-
 drivers/ata/pata_atiixp.c       |  2 +-
 drivers/ata/pata_cmd640.c       |  2 +-
 drivers/ata/pata_cmd64x.c       |  2 +-
 drivers/ata/pata_cs5530.c       |  2 +-
 drivers/ata/pata_cs5535.c       |  2 +-
 drivers/ata/pata_cs5536.c       |  2 +-
 drivers/ata/pata_cypress.c      |  2 +-
 drivers/ata/pata_efar.c         |  2 +-
 drivers/ata/pata_hpt366.c       | 12 ++++++------
 drivers/ata/pata_hpt37x.c       | 33 ++++++++++++++-------------------
 drivers/ata/pata_hpt3x2n.c      |  9 ++++-----
 drivers/ata/pata_it8213.c       |  2 +-
 drivers/ata/pata_it821x.c       |  2 +-
 drivers/ata/pata_jmicron.c      | 18 ++----------------
 drivers/ata/pata_marvell.c      | 18 ++----------------
 drivers/ata/pata_netcell.c      |  2 +-
 drivers/ata/pata_ns87410.c      |  2 +-
 drivers/ata/pata_ns87415.c      |  2 +-
 drivers/ata/pata_oldpiix.c      |  2 +-
 drivers/ata/pata_opti.c         |  2 +-
 drivers/ata/pata_optidma.c      |  2 +-
 drivers/ata/pata_pdc202xx_old.c |  2 +-
 drivers/ata/pata_radisys.c      |  2 +-
 drivers/ata/pata_rz1000.c       |  2 +-
 drivers/ata/pata_sc1200.c       |  2 +-
 drivers/ata/pata_serverworks.c  |  2 +-
 drivers/ata/pata_sil680.c       |  2 +-
 drivers/ata/pata_sis.c          |  8 +++-----
 drivers/ata/pata_sl82c105.c     |  2 +-
 drivers/ata/pata_triflex.c      |  2 +-
 drivers/ata/pata_via.c          | 19 ++++++++-----------
 include/linux/libata.h          |  2 +-
 39 files changed, 79 insertions(+), 145 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ata_generic.c b/drivers/ata/ata_generic.c
index a912ee01a47c..b23e2a1099c5 100644
--- a/drivers/ata/ata_generic.c
+++ b/drivers/ata/ata_generic.c
@@ -152,7 +152,7 @@ static int ata_generic_init_one(struct pci_dev *dev, const struct pci_device_id
 	if (dev->vendor == PCI_VENDOR_ID_AL)
 	    	ata_pci_clear_simplex(dev);
 
-	return ata_pci_init_one(dev, ppi, &generic_sht);
+	return ata_pci_init_one(dev, ppi, &generic_sht, NULL);
 }
 
 static struct pci_device_id ata_generic[] = {
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index abbe3229480c..aa6bcd79d60a 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -6995,8 +6995,6 @@ struct ata_host *ata_host_alloc_pinfo(struct device *dev,
 
 		if (!host->ops && (pi->port_ops != &ata_dummy_port_ops))
 			host->ops = pi->port_ops;
-		if (!host->private_data && pi->private_data)
-			host->private_data = pi->private_data;
 	}
 
 	return host;
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index a9d5898cbbc4..6223ec042c80 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -827,6 +827,7 @@ int ata_pci_activate_sff_host(struct ata_host *host,
  *	@pdev: Controller to be initialized
  *	@ppi: array of port_info, must be enough for two ports
  *	@sht: scsi_host_template to use when registering the host
+ *	@host_priv: host private_data
  *
  *	This is a helper function which can be called from a driver's
  *	xxx_init_one() probe function if the hardware uses traditional
@@ -848,7 +849,7 @@ int ata_pci_activate_sff_host(struct ata_host *host,
  */
 int ata_pci_init_one(struct pci_dev *pdev,
 		     const struct ata_port_info * const * ppi,
-		     struct scsi_host_template *sht)
+		     struct scsi_host_template *sht, void *host_priv)
 {
 	struct device *dev = &pdev->dev;
 	const struct ata_port_info *pi = NULL;
@@ -882,6 +883,7 @@ int ata_pci_init_one(struct pci_dev *pdev,
 	rc = ata_pci_prepare_sff_host(pdev, ppi, &host);
 	if (rc)
 		goto out;
+	host->private_data = host_priv;
 
 	pci_set_master(pdev);
 	rc = ata_pci_activate_sff_host(host, ata_interrupt, sht);
diff --git a/drivers/ata/pata_acpi.c b/drivers/ata/pata_acpi.c
index 3edde51750da..d337f3209caf 100644
--- a/drivers/ata/pata_acpi.c
+++ b/drivers/ata/pata_acpi.c
@@ -67,20 +67,6 @@ static int pacpi_cable_detect(struct ata_port *ap)
 		return ATA_CBL_PATA40;
 }
 
-/**
- *	pacpi_error_handler - Setup and error handler
- *	@ap: Port to handle
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void pacpi_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, pacpi_pre_reset, ata_std_softreset, NULL,
-			   ata_std_postreset);
-}
-
 /**
  *	pacpi_discover_modes	-	filter non ACPI modes
  *	@adev: ATA device
@@ -242,7 +228,7 @@ static struct ata_port_operations pacpi_ops = {
 	.mode_filter		= pacpi_mode_filter,
 	.set_piomode		= pacpi_set_piomode,
 	.set_dmamode		= pacpi_set_dmamode,
-	.error_handler		= pacpi_error_handler,
+	.prereset		= pacpi_pre_reset,
 	.port_start		= pacpi_port_start,
 };
 
@@ -273,7 +259,7 @@ static int pacpi_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 		.port_ops	= &pacpi_ops,
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(pdev, ppi, &pacpi_sht);
+	return ata_pci_init_one(pdev, ppi, &pacpi_sht, NULL);
 }
 
 static const struct pci_device_id pacpi_pci_tbl[] = {
diff --git a/drivers/ata/pata_ali.c b/drivers/ata/pata_ali.c
index f2924996f6e3..43c558f20f92 100644
--- a/drivers/ata/pata_ali.c
+++ b/drivers/ata/pata_ali.c
@@ -552,7 +552,7 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	        	ppi[0] = &info_20_udma;
 		pci_dev_put(isa_bridge);
 	}
-	return ata_pci_init_one(pdev, ppi, &ali_sht);
+	return ata_pci_init_one(pdev, ppi, &ali_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_amd.c b/drivers/ata/pata_amd.c
index 644702cac6ee..09c8286b6890 100644
--- a/drivers/ata/pata_amd.c
+++ b/drivers/ata/pata_amd.c
@@ -144,12 +144,6 @@ static int amd_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-static void amd_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, amd_pre_reset, ata_std_softreset, NULL,
-			   ata_std_postreset);
-}
-
 static int amd_cable_detect(struct ata_port *ap)
 {
 	static const u32 bitmask[2] = {0x03, 0x0C};
@@ -300,13 +294,6 @@ static int nv_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-static void nv_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, nv_pre_reset,
-			       ata_std_softreset, NULL,
-			       ata_std_postreset);
-}
-
 /**
  *	nv100_set_piomode	-	set initial PIO mode data
  *	@ap: ATA interface
@@ -358,7 +345,7 @@ static struct scsi_host_template amd_sht = {
 
 static const struct ata_port_operations amd_base_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
-	.error_handler	= amd_error_handler,
+	.prereset	= amd_pre_reset,
 };
 
 static struct ata_port_operations amd33_port_ops = {
@@ -393,7 +380,7 @@ static const struct ata_port_operations nv_base_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 	.cable_detect	= ata_cable_ignore,
 	.mode_filter	= nv_mode_filter,
-	.error_handler	= nv_error_handler,
+	.prereset	= nv_pre_reset,
 	.host_stop	= nv_host_stop,
 };
 
@@ -483,10 +470,10 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 			.port_ops = &amd100_port_ops
 		}
 	};
-	struct ata_port_info pi;
-	const struct ata_port_info *ppi[] = { &pi, NULL };
+	const struct ata_port_info *ppi[] = { NULL, NULL };
 	static int printed_version;
 	int type = id->driver_data;
+	void *hpriv = NULL;
 	u8 fifo;
 	int rc;
 
@@ -511,7 +498,7 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	/*
 	 * Okay, type is determined now.  Apply type-specific workarounds.
 	 */
-	pi = info[type];
+	ppi[0] = &info[type];
 
 	if (type < 3)
 		ata_pci_clear_simplex(pdev);
@@ -530,11 +517,11 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 		u32 udma;
 
 		pci_read_config_dword(pdev, 0x60, &udma);
-		pi.private_data = (void *)(unsigned long)udma;
+		hpriv = (void *)(unsigned long)udma;
 	}
 
 	/* And fire it up */
-	return ata_pci_init_one(pdev, ppi, &amd_sht);
+	return ata_pci_init_one(pdev, ppi, &amd_sht, hpriv);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c
index 698a53c96111..ebd15cadf15f 100644
--- a/drivers/ata/pata_artop.c
+++ b/drivers/ata/pata_artop.c
@@ -430,7 +430,7 @@ static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 
 	BUG_ON(ppi[0] == NULL);
 
-	return ata_pci_init_one(pdev, ppi, &artop_sht);
+	return ata_pci_init_one(pdev, ppi, &artop_sht, NULL);
 }
 
 static const struct pci_device_id artop_pci_tbl[] = {
diff --git a/drivers/ata/pata_atiixp.c b/drivers/ata/pata_atiixp.c
index 6fe433ba62bd..0bea7e75d2d6 100644
--- a/drivers/ata/pata_atiixp.c
+++ b/drivers/ata/pata_atiixp.c
@@ -248,7 +248,7 @@ static int atiixp_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.port_ops = &atiixp_port_ops
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(dev, ppi, &atiixp_sht);
+	return ata_pci_init_one(dev, ppi, &atiixp_sht, NULL);
 }
 
 static const struct pci_device_id atiixp[] = {
diff --git a/drivers/ata/pata_cmd640.c b/drivers/ata/pata_cmd640.c
index efd2bb5747b4..27219b00edf4 100644
--- a/drivers/ata/pata_cmd640.c
+++ b/drivers/ata/pata_cmd640.c
@@ -224,7 +224,7 @@ static int cmd640_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	cmd640_hardware_init(pdev);
 
-	return ata_pci_init_one(pdev, ppi, &cmd640_sht);
+	return ata_pci_init_one(pdev, ppi, &cmd640_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cmd64x.c b/drivers/ata/pata_cmd64x.c
index bfd72ef9cd31..f0e566623614 100644
--- a/drivers/ata/pata_cmd64x.c
+++ b/drivers/ata/pata_cmd64x.c
@@ -373,7 +373,7 @@ static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	pci_write_config_byte(pdev, UDIDETCR0, 0xF0);
 #endif
 
-	return ata_pci_init_one(pdev, ppi, &cmd64x_sht);
+	return ata_pci_init_one(pdev, ppi, &cmd64x_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cs5530.c b/drivers/ata/pata_cs5530.c
index c632ce499d33..ac3ad55d7c3c 100644
--- a/drivers/ata/pata_cs5530.c
+++ b/drivers/ata/pata_cs5530.c
@@ -325,7 +325,7 @@ static int cs5530_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 		ppi[1] = &info_palmax_secondary;
 
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(pdev, ppi, &cs5530_sht);
+	return ata_pci_init_one(pdev, ppi, &cs5530_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cs5535.c b/drivers/ata/pata_cs5535.c
index d78cf95cbe45..5c0762ebf58c 100644
--- a/drivers/ata/pata_cs5535.c
+++ b/drivers/ata/pata_cs5535.c
@@ -199,7 +199,7 @@ static int cs5535_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	rdmsr(ATAC_CH0D1_PIO, timings, dummy);
 	if (CS5535_BAD_PIO(timings))
 		wrmsr(ATAC_CH0D1_PIO, 0xF7F4F7F4UL, 0);
-	return ata_pci_init_one(dev, ppi, &cs5535_sht);
+	return ata_pci_init_one(dev, ppi, &cs5535_sht, NULL);
 }
 
 static const struct pci_device_id cs5535[] = {
diff --git a/drivers/ata/pata_cs5536.c b/drivers/ata/pata_cs5536.c
index f7c0e4e319ed..2d34b9145dcb 100644
--- a/drivers/ata/pata_cs5536.c
+++ b/drivers/ata/pata_cs5536.c
@@ -261,7 +261,7 @@ static int cs5536_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		return -ENODEV;
 	}
 
-	return ata_pci_init_one(dev, ppi, &cs5536_sht);
+	return ata_pci_init_one(dev, ppi, &cs5536_sht, NULL);
 }
 
 static const struct pci_device_id cs5536[] = {
diff --git a/drivers/ata/pata_cypress.c b/drivers/ata/pata_cypress.c
index cbd6670ea0de..ae14969e1dfe 100644
--- a/drivers/ata/pata_cypress.c
+++ b/drivers/ata/pata_cypress.c
@@ -136,7 +136,7 @@ static int cy82c693_init_one(struct pci_dev *pdev, const struct pci_device_id *i
 	if (PCI_FUNC(pdev->devfn) != 1)
 		return -ENODEV;
 
-	return ata_pci_init_one(pdev, ppi, &cy82c693_sht);
+	return ata_pci_init_one(pdev, ppi, &cy82c693_sht, NULL);
 }
 
 static const struct pci_device_id cy82c693[] = {
diff --git a/drivers/ata/pata_efar.c b/drivers/ata/pata_efar.c
index 0260edac2370..2f5b4848456a 100644
--- a/drivers/ata/pata_efar.c
+++ b/drivers/ata/pata_efar.c
@@ -275,7 +275,7 @@ static int efar_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &efar_sht);
+	return ata_pci_init_one(pdev, ppi, &efar_sht, NULL);
 }
 
 static const struct pci_device_id efar_pci_tbl[] = {
diff --git a/drivers/ata/pata_hpt366.c b/drivers/ata/pata_hpt366.c
index b62d398ed84b..c2d4923d4db7 100644
--- a/drivers/ata/pata_hpt366.c
+++ b/drivers/ata/pata_hpt366.c
@@ -356,9 +356,9 @@ static int hpt36x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.udma_mask = ATA_UDMA4,
 		.port_ops = &hpt366_port_ops
 	};
-	struct ata_port_info info = info_hpt366;
-	const struct ata_port_info *ppi[] = { &info, NULL };
+	const struct ata_port_info *ppi[] = { &info_hpt366, NULL };
 
+	void *hpriv = NULL;
 	u32 class_rev;
 	u32 reg1;
 	int rc;
@@ -383,17 +383,17 @@ static int hpt36x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	/* info_hpt366 is safe against re-entry so we can scribble on it */
 	switch((reg1 & 0x700) >> 8) {
 		case 5:
-			info.private_data = &hpt366_40;
+			hpriv = &hpt366_40;
 			break;
 		case 9:
-			info.private_data = &hpt366_25;
+			hpriv = &hpt366_25;
 			break;
 		default:
-			info.private_data = &hpt366_33;
+			hpriv = &hpt366_33;
 			break;
 	}
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(dev, ppi, &hpt36x_sht);
+	return ata_pci_init_one(dev, ppi, &hpt36x_sht, hpriv);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index a43c19753669..fb37e3a161fc 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -831,10 +831,8 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	};
 
 	static const int MHz[4] = { 33, 40, 50, 66 };
-	const struct ata_port_info *port;
 	void *private_data = NULL;
-	struct ata_port_info port_info;
-	const struct ata_port_info *ppi[] = { &port_info, NULL };
+	const struct ata_port_info *ppi[] = { NULL, NULL };
 
 	u8 irqmask;
 	u32 class_rev;
@@ -866,17 +864,17 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 
 		switch(class_rev) {
 			case 3:
-				port = &info_hpt370;
+				ppi[0] = &info_hpt370;
 				chip_table = &hpt370;
 				prefer_dpll = 0;
 				break;
 			case 4:
-				port = &info_hpt370a;
+				ppi[0] = &info_hpt370a;
 				chip_table = &hpt370a;
 				prefer_dpll = 0;
 				break;
 			case 5:
-				port = &info_hpt372;
+				ppi[0] = &info_hpt372;
 				chip_table = &hpt372;
 				break;
 			default:
@@ -889,21 +887,21 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 				/* 372N if rev >= 2*/
 				if (class_rev >= 2)
 					return -ENODEV;
-				port = &info_hpt372;
+				ppi[0] = &info_hpt372;
 				chip_table = &hpt372a;
 				break;
 			case PCI_DEVICE_ID_TTI_HPT302:
 				/* 302N if rev > 1 */
 				if (class_rev > 1)
 					return -ENODEV;
-				port = &info_hpt372;
+				ppi[0] = &info_hpt372;
 				/* Check this */
 				chip_table = &hpt302;
 				break;
 			case PCI_DEVICE_ID_TTI_HPT371:
 				if (class_rev > 1)
 					return -ENODEV;
-				port = &info_hpt372;
+				ppi[0] = &info_hpt372;
 				chip_table = &hpt371;
 				/* Single channel device, master is not present
 				   but the BIOS (or us for non x86) must mark it
@@ -914,7 +912,7 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 				break;
 			case PCI_DEVICE_ID_TTI_HPT374:
 				chip_table = &hpt374;
-				port = &info_hpt374;
+				ppi[0] = &info_hpt374;
 				break;
 			default:
 				printk(KERN_ERR "pata_hpt37x: PCI table is bogus please report (%d).\n", dev->device);
@@ -993,7 +991,7 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		int dpll, adjust;
 
 		/* Compute DPLL */
-		dpll = (port->udma_mask & 0xC0) ? 3 : 2;
+		dpll = (ppi[0]->udma_mask & 0xC0) ? 3 : 2;
 
 		f_low = (MHz[clock_slot] * 48) / MHz[dpll];
 		f_high = f_low + 2;
@@ -1033,19 +1031,16 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		 *	about lack of UDMA133 support on lower clocks
  		 */
 
-		if (clock_slot < 2 && port == &info_hpt370)
-			port = &info_hpt370_33;
-		if (clock_slot < 2 && port == &info_hpt370a)
-			port = &info_hpt370a_33;
+		if (clock_slot < 2 && ppi[0] == &info_hpt370)
+			ppi[0] = &info_hpt370_33;
+		if (clock_slot < 2 && ppi[0] == &info_hpt370a)
+			ppi[0] = &info_hpt370a_33;
 		printk(KERN_INFO "pata_hpt37x: %s using %dMHz bus clock.\n",
 		       chip_table->name, MHz[clock_slot]);
 	}
 
 	/* Now kick off ATA set up */
-	port_info = *port;
-	port_info.private_data = private_data;
-
-	return ata_pci_init_one(dev, ppi, &hpt37x_sht);
+	return ata_pci_init_one(dev, ppi, &hpt37x_sht, private_data);
 }
 
 static const struct pci_device_id hpt37x[] = {
diff --git a/drivers/ata/pata_hpt3x2n.c b/drivers/ata/pata_hpt3x2n.c
index 2c178c30116c..c774be93ae04 100644
--- a/drivers/ata/pata_hpt3x2n.c
+++ b/drivers/ata/pata_hpt3x2n.c
@@ -458,8 +458,7 @@ static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.udma_mask = ATA_UDMA6,
 		.port_ops = &hpt3x2n_port_ops
 	};
-	struct ata_port_info port = info;
-	const struct ata_port_info *ppi[] = { &port, NULL };
+	const struct ata_port_info *ppi[] = { &info, NULL };
 
 	u8 irqmask;
 	u32 class_rev;
@@ -468,6 +467,7 @@ static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	unsigned int f_low, f_high;
 	int adjust;
 	unsigned long iobase = pci_resource_start(dev, 4);
+	void *hpriv = NULL;
 	int rc;
 
 	rc = pcim_enable_device(dev);
@@ -554,9 +554,8 @@ static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	       pci_mhz);
 	/* Set our private data up. We only need a few flags so we use
 	   it directly */
-	port.private_data = NULL;
 	if (pci_mhz > 60) {
-		port.private_data = (void *)PCI66;
+		hpriv = (void *)PCI66;
 		/*
 		 * On  HPT371N, if ATA clock is 66 MHz we must set bit 2 in
 		 * the MISC. register to stretch the UltraDMA Tss timing.
@@ -567,7 +566,7 @@ static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	}
 
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(dev, ppi, &hpt3x2n_sht);
+	return ata_pci_init_one(dev, ppi, &hpt3x2n_sht, hpriv);
 }
 
 static const struct pci_device_id hpt3x2n[] = {
diff --git a/drivers/ata/pata_it8213.c b/drivers/ata/pata_it8213.c
index 291a0d6e2434..d23a46b75028 100644
--- a/drivers/ata/pata_it8213.c
+++ b/drivers/ata/pata_it8213.c
@@ -287,7 +287,7 @@ static int it8213_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &it8213_sht);
+	return ata_pci_init_one(pdev, ppi, &it8213_sht, NULL);
 }
 
 static const struct pci_device_id it8213_pci_tbl[] = {
diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c
index 63c5cf0d1fee..6a8a4ddf5bfe 100644
--- a/drivers/ata/pata_it821x.c
+++ b/drivers/ata/pata_it821x.c
@@ -722,7 +722,7 @@ static int it821x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	else
 		ppi[0] = &info_smart;
 
-	return ata_pci_init_one(pdev, ppi, &it821x_sht);
+	return ata_pci_init_one(pdev, ppi, &it821x_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_jmicron.c b/drivers/ata/pata_jmicron.c
index 859e47a600cc..317f3474e0ba 100644
--- a/drivers/ata/pata_jmicron.c
+++ b/drivers/ata/pata_jmicron.c
@@ -105,20 +105,6 @@ static int jmicron_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	jmicron_error_handler - Setup and error handler
- *	@ap: Port to handle
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void jmicron_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, jmicron_pre_reset, ata_std_softreset, NULL,
-			   ata_std_postreset);
-}
-
 /* No PIO or DMA methods needed for this device */
 
 static struct scsi_host_template jmicron_sht = {
@@ -127,7 +113,7 @@ static struct scsi_host_template jmicron_sht = {
 
 static struct ata_port_operations jmicron_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.error_handler		= jmicron_error_handler,
+	.prereset		= jmicron_pre_reset,
 };
 
 
@@ -158,7 +144,7 @@ static int jmicron_init_one (struct pci_dev *pdev, const struct pci_device_id *i
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
 
-	return ata_pci_init_one(pdev, ppi, &jmicron_sht);
+	return ata_pci_init_one(pdev, ppi, &jmicron_sht, NULL);
 }
 
 static const struct pci_device_id jmicron_pci_tbl[] = {
diff --git a/drivers/ata/pata_marvell.c b/drivers/ata/pata_marvell.c
index d8da4f344c0a..d38e64cd6097 100644
--- a/drivers/ata/pata_marvell.c
+++ b/drivers/ata/pata_marvell.c
@@ -75,20 +75,6 @@ static int marvell_cable_detect(struct ata_port *ap)
 	return 0;	/* Our BUG macro needs the right markup */
 }
 
-/**
- *	marvell_error_handler - Setup and error handler
- *	@ap: Port to handle
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void marvell_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, marvell_pre_reset, ata_std_softreset, NULL,
-			   ata_std_postreset);
-}
-
 /* No PIO or DMA methods needed for this device */
 
 static struct scsi_host_template marvell_sht = {
@@ -98,7 +84,7 @@ static struct scsi_host_template marvell_sht = {
 static struct ata_port_operations marvell_ops = {
 	.inherits		= &ata_bmdma_port_ops,
 	.cable_detect		= marvell_cable_detect,
-	.error_handler		= marvell_error_handler,
+	.prereset		= marvell_pre_reset,
 };
 
 
@@ -142,7 +128,7 @@ static int marvell_init_one (struct pci_dev *pdev, const struct pci_device_id *i
 	if (pdev->device == 0x6101)
 		ppi[1] = &ata_dummy_port_info;
 
-	return ata_pci_init_one(pdev, ppi, &marvell_sht);
+	return ata_pci_init_one(pdev, ppi, &marvell_sht, NULL);
 }
 
 static const struct pci_device_id marvell_pci_tbl[] = {
diff --git a/drivers/ata/pata_netcell.c b/drivers/ata/pata_netcell.c
index ae50a5e85cf1..349182840d24 100644
--- a/drivers/ata/pata_netcell.c
+++ b/drivers/ata/pata_netcell.c
@@ -71,7 +71,7 @@ static int netcell_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	ata_pci_clear_simplex(pdev);
 
 	/* And let the library code do the work */
-	return ata_pci_init_one(pdev, port_info, &netcell_sht);
+	return ata_pci_init_one(pdev, port_info, &netcell_sht, NULL);
 }
 
 static const struct pci_device_id netcell_pci_tbl[] = {
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
index 1bdca8f1e767..5a043e426480 100644
--- a/drivers/ata/pata_ns87410.c
+++ b/drivers/ata/pata_ns87410.c
@@ -163,7 +163,7 @@ static int ns87410_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.port_ops = &ns87410_port_ops
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(dev, ppi, &ns87410_sht);
+	return ata_pci_init_one(dev, ppi, &ns87410_sht, NULL);
 }
 
 static const struct pci_device_id ns87410[] = {
diff --git a/drivers/ata/pata_ns87415.c b/drivers/ata/pata_ns87415.c
index 42508940e4a9..cdd79d6fc0ee 100644
--- a/drivers/ata/pata_ns87415.c
+++ b/drivers/ata/pata_ns87415.c
@@ -375,7 +375,7 @@ static int ns87415_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	pci_write_config_byte(pdev, 0x55, 0xEE);
 	/* Select PIO0 8bit clocking */
 	pci_write_config_byte(pdev, 0x54, 0xB7);
-	return ata_pci_init_one(pdev, ppi, &ns87415_sht);
+	return ata_pci_init_one(pdev, ppi, &ns87415_sht, NULL);
 }
 
 static const struct pci_device_id ns87415_pci_tbl[] = {
diff --git a/drivers/ata/pata_oldpiix.c b/drivers/ata/pata_oldpiix.c
index 9e3afadbd04a..7001b756819e 100644
--- a/drivers/ata/pata_oldpiix.c
+++ b/drivers/ata/pata_oldpiix.c
@@ -263,7 +263,7 @@ static int oldpiix_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &oldpiix_sht);
+	return ata_pci_init_one(pdev, ppi, &oldpiix_sht, NULL);
 }
 
 static const struct pci_device_id oldpiix_pci_tbl[] = {
diff --git a/drivers/ata/pata_opti.c b/drivers/ata/pata_opti.c
index 8601d9c3cb39..5a5f20e03fc0 100644
--- a/drivers/ata/pata_opti.c
+++ b/drivers/ata/pata_opti.c
@@ -188,7 +188,7 @@ static int opti_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (!printed_version++)
 		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(dev, ppi, &opti_sht);
+	return ata_pci_init_one(dev, ppi, &opti_sht, NULL);
 }
 
 static const struct pci_device_id opti[] = {
diff --git a/drivers/ata/pata_optidma.c b/drivers/ata/pata_optidma.c
index c376f9ef77c8..ba2819ff964b 100644
--- a/drivers/ata/pata_optidma.c
+++ b/drivers/ata/pata_optidma.c
@@ -445,7 +445,7 @@ static int optidma_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (optiplus_with_udma(dev))
 		ppi[0] = &info_82c700_udma;
 
-	return ata_pci_init_one(dev, ppi, &optidma_sht);
+	return ata_pci_init_one(dev, ppi, &optidma_sht, NULL);
 }
 
 static const struct pci_device_id optidma[] = {
diff --git a/drivers/ata/pata_pdc202xx_old.c b/drivers/ata/pata_pdc202xx_old.c
index 5545fbab6a7e..8214100e3ac1 100644
--- a/drivers/ata/pata_pdc202xx_old.c
+++ b/drivers/ata/pata_pdc202xx_old.c
@@ -324,7 +324,7 @@ static int pdc202xx_init_one(struct pci_dev *dev, const struct pci_device_id *id
 				return -ENODEV;
 		}
 	}
-	return ata_pci_init_one(dev, ppi, &pdc202xx_sht);
+	return ata_pci_init_one(dev, ppi, &pdc202xx_sht, NULL);
 }
 
 static const struct pci_device_id pdc202xx[] = {
diff --git a/drivers/ata/pata_radisys.c b/drivers/ata/pata_radisys.c
index 145d5ba92795..9ab84fc3798d 100644
--- a/drivers/ata/pata_radisys.c
+++ b/drivers/ata/pata_radisys.c
@@ -228,7 +228,7 @@ static int radisys_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &radisys_sht);
+	return ata_pci_init_one(pdev, ppi, &radisys_sht, NULL);
 }
 
 static const struct pci_device_id radisys_pci_tbl[] = {
diff --git a/drivers/ata/pata_rz1000.c b/drivers/ata/pata_rz1000.c
index 04be6aee4354..462b72a31280 100644
--- a/drivers/ata/pata_rz1000.c
+++ b/drivers/ata/pata_rz1000.c
@@ -99,7 +99,7 @@ static int rz1000_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 		printk(KERN_DEBUG DRV_NAME " version " DRV_VERSION "\n");
 
 	if (rz1000_fifo_disable(pdev) == 0)
-		return ata_pci_init_one(pdev, ppi, &rz1000_sht);
+		return ata_pci_init_one(pdev, ppi, &rz1000_sht, NULL);
 
 	printk(KERN_ERR DRV_NAME ": failed to disable read-ahead on chipset..\n");
 	/* Not safe to use so skip */
diff --git a/drivers/ata/pata_sc1200.c b/drivers/ata/pata_sc1200.c
index 38c7fb0bebe9..42efacf73c79 100644
--- a/drivers/ata/pata_sc1200.c
+++ b/drivers/ata/pata_sc1200.c
@@ -213,7 +213,7 @@ static int sc1200_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	/* Can't enable port 2 yet, see top comments */
 	const struct ata_port_info *ppi[] = { &info, &ata_dummy_port_info };
 
-	return ata_pci_init_one(dev, ppi, &sc1200_sht);
+	return ata_pci_init_one(dev, ppi, &sc1200_sht, NULL);
 }
 
 static const struct pci_device_id sc1200[] = {
diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
index 515b5b70a555..2f4f9b0f89de 100644
--- a/drivers/ata/pata_serverworks.c
+++ b/drivers/ata/pata_serverworks.c
@@ -461,7 +461,7 @@ static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id
 	if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE)
 		ata_pci_clear_simplex(pdev);
 
-	return ata_pci_init_one(pdev, ppi, &serverworks_sht);
+	return ata_pci_init_one(pdev, ppi, &serverworks_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_sil680.c b/drivers/ata/pata_sil680.c
index 5313deeffa6d..63fafc6d6da3 100644
--- a/drivers/ata/pata_sil680.c
+++ b/drivers/ata/pata_sil680.c
@@ -362,7 +362,7 @@ static int __devinit sil680_init_one(struct pci_dev *pdev,
 				 &sil680_sht);
 
 use_ioports:
-	return ata_pci_init_one(pdev, ppi, &sil680_sht);
+	return ata_pci_init_one(pdev, ppi, &sil680_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
index 32be13ba5f06..28abfc26e7a4 100644
--- a/drivers/ata/pata_sis.c
+++ b/drivers/ata/pata_sis.c
@@ -690,8 +690,7 @@ static void sis_fixup(struct pci_dev *pdev, struct sis_chipset *sis)
 static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	static int printed_version;
-	struct ata_port_info port;
-	const struct ata_port_info *ppi[] = { &port, NULL };
+	const struct ata_port_info *ppi[] = { NULL, NULL };
 	struct pci_dev *host = NULL;
 	struct sis_chipset *chipset = NULL;
 	struct sis_chipset *sets;
@@ -831,12 +830,11 @@ static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (chipset == NULL)
 		return -ENODEV;
 
-	port = *chipset->info;
-	port.private_data = chipset;
+	ppi[0] = chipset->info;
 
 	sis_fixup(pdev, chipset);
 
-	return ata_pci_init_one(pdev, ppi, &sis_sht);
+	return ata_pci_init_one(pdev, ppi, &sis_sht, chipset);
 }
 
 static const struct pci_device_id sis_pci_tbl[] = {
diff --git a/drivers/ata/pata_sl82c105.c b/drivers/ata/pata_sl82c105.c
index 2d14b2505c7d..1d97f920bd2b 100644
--- a/drivers/ata/pata_sl82c105.c
+++ b/drivers/ata/pata_sl82c105.c
@@ -323,7 +323,7 @@ static int sl82c105_init_one(struct pci_dev *dev, const struct pci_device_id *id
 	val |= CTRL_P0EN | CTRL_P0F16 | CTRL_P1F16;
 	pci_write_config_dword(dev, 0x40, val);
 
-	return ata_pci_init_one(dev, ppi, &sl82c105_sht);
+	return ata_pci_init_one(dev, ppi, &sl82c105_sht, NULL);
 }
 
 static const struct pci_device_id sl82c105[] = {
diff --git a/drivers/ata/pata_triflex.c b/drivers/ata/pata_triflex.c
index 86dc66c37389..f07b0e5df222 100644
--- a/drivers/ata/pata_triflex.c
+++ b/drivers/ata/pata_triflex.c
@@ -206,7 +206,7 @@ static int triflex_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (!printed_version++)
 		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(dev, ppi, &triflex_sht);
+	return ata_pci_init_one(dev, ppi, &triflex_sht, NULL);
 }
 
 static const struct pci_device_id triflex[] = {
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
index e66bb85ad3d1..f4092cbd566f 100644
--- a/drivers/ata/pata_via.c
+++ b/drivers/ata/pata_via.c
@@ -442,8 +442,7 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 		.udma_mask = ATA_UDMA6,	/* FIXME: should check north bridge */
 		.port_ops = &via_port_ops
 	};
-	struct ata_port_info type;
-	const struct ata_port_info *ppi[] = { &type, NULL };
+	const struct ata_port_info *ppi[] = { NULL, NULL };
 	struct pci_dev *isa = NULL;
 	const struct via_isa_bridge *config;
 	static int printed_version;
@@ -491,25 +490,25 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	switch(config->flags & VIA_UDMA) {
 		case VIA_UDMA_NONE:
 			if (config->flags & VIA_NO_UNMASK)
-				type = via_mwdma_info_borked;
+				ppi[0] = &via_mwdma_info_borked;
 			else
-				type = via_mwdma_info;
+				ppi[0] = &via_mwdma_info;
 			break;
 		case VIA_UDMA_33:
-			type = via_udma33_info;
+			ppi[0] = &via_udma33_info;
 			break;
 		case VIA_UDMA_66:
-			type = via_udma66_info;
+			ppi[0] = &via_udma66_info;
 			/* The 66 MHz devices require we enable the clock */
 			pci_read_config_dword(pdev, 0x50, &timing);
 			timing |= 0x80008;
 			pci_write_config_dword(pdev, 0x50, timing);
 			break;
 		case VIA_UDMA_100:
-			type = via_udma100_info;
+			ppi[0] = &via_udma100_info;
 			break;
 		case VIA_UDMA_133:
-			type = via_udma133_info;
+			ppi[0] = &via_udma133_info;
 			break;
 		default:
 			WARN_ON(1);
@@ -524,9 +523,7 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	}
 
 	/* We have established the device type, now fire it up */
-	type.private_data = (void *)config;
-
-	return ata_pci_init_one(pdev, ppi, &via_sht);
+	return ata_pci_init_one(pdev, ppi, &via_sht, (void *)config);
 }
 
 #ifdef CONFIG_PM
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 5494119854de..e7f10a88efe1 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1020,7 +1020,7 @@ struct pci_dev;
 
 extern int ata_pci_init_one(struct pci_dev *pdev,
 			    const struct ata_port_info * const * ppi,
-			    struct scsi_host_template *sht);
+			    struct scsi_host_template *sht, void *host_priv);
 extern void ata_pci_remove_one(struct pci_dev *pdev);
 #ifdef CONFIG_PM
 extern void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t mesg);
-- 
cgit v1.2.3


From 959471936241bd83da7d0a76411cef6772140fe6 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:49 +0900
Subject: libata: kill port_info->sht and ->irq_handler

libata core layer doesn't care about sht or ->irq_handler.  Those are
only of interest to the LLD during initialization.  This is confusing
and has caused several drivers to have duplicate unused initializers
for these fields.

Currently only sata_nv uses these fields.  Make sata_nv use
->private_data, which is supposed to carry LLD-specific information,
instead and kill ->sht and ->irq_handler.  nv_pi_priv structure is
defined and struct literals are used to initialize private_data.
Notational overhead is negligible.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/sata_nv.c  | 29 +++++++++++++++++------------
 include/linux/libata.h |  2 --
 2 files changed, 17 insertions(+), 14 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index 7b7ba0e26903..5637b082bc85 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -466,58 +466,61 @@ static struct ata_port_operations nv_swncq_ops = {
 	.port_start		= nv_swncq_port_start,
 };
 
+struct nv_pi_priv {
+	irq_handler_t			irq_handler;
+	struct scsi_host_template	*sht;
+};
+
+#define NV_PI_PRIV(_irq_handler, _sht) \
+	&(struct nv_pi_priv){ .irq_handler = _irq_handler, .sht = _sht }
+
 static const struct ata_port_info nv_port_info[] = {
 	/* generic */
 	{
-		.sht		= &nv_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
 		.port_ops	= &nv_generic_ops,
-		.irq_handler	= nv_generic_interrupt,
+		.private_data	= NV_PI_PRIV(nv_generic_interrupt, &nv_sht),
 	},
 	/* nforce2/3 */
 	{
-		.sht		= &nv_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
 		.port_ops	= &nv_nf2_ops,
-		.irq_handler	= nv_nf2_interrupt,
+		.private_data	= NV_PI_PRIV(nv_nf2_interrupt, &nv_sht),
 	},
 	/* ck804 */
 	{
-		.sht		= &nv_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
 		.port_ops	= &nv_ck804_ops,
-		.irq_handler	= nv_ck804_interrupt,
+		.private_data	= NV_PI_PRIV(nv_ck804_interrupt, &nv_sht),
 	},
 	/* ADMA */
 	{
-		.sht		= &nv_adma_sht,
 		.flags		= ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_MMIO | ATA_FLAG_NCQ,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
 		.port_ops	= &nv_adma_ops,
-		.irq_handler	= nv_adma_interrupt,
+		.private_data	= NV_PI_PRIV(nv_adma_interrupt, &nv_adma_sht),
 	},
 	/* SWNCQ */
 	{
-		.sht		= &nv_swncq_sht,
 		.flags	        = ATA_FLAG_SATA | ATA_FLAG_NO_LEGACY |
 				  ATA_FLAG_NCQ,
 		.pio_mask	= NV_PIO_MASK,
 		.mwdma_mask	= NV_MWDMA_MASK,
 		.udma_mask	= NV_UDMA_MASK,
 		.port_ops	= &nv_swncq_ops,
-		.irq_handler	= nv_swncq_interrupt,
+		.private_data	= NV_PI_PRIV(nv_swncq_interrupt, &nv_swncq_sht),
 	},
 };
 
@@ -2316,6 +2319,7 @@ static int nv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	static int printed_version;
 	const struct ata_port_info *ppi[] = { NULL, NULL };
+	struct nv_pi_priv *ipriv;
 	struct ata_host *host;
 	struct nv_host_priv *hpriv;
 	int rc;
@@ -2352,6 +2356,7 @@ static int nv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	}
 
 	ppi[0] = &nv_port_info[type];
+	ipriv = ppi[0]->private_data;
 	rc = ata_pci_prepare_sff_host(pdev, ppi, &host);
 	if (rc)
 		return rc;
@@ -2390,8 +2395,8 @@ static int nv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 		nv_swncq_host_init(host);
 
 	pci_set_master(pdev);
-	return ata_host_activate(host, pdev->irq, ppi[0]->irq_handler,
-				 IRQF_SHARED, ppi[0]->sht);
+	return ata_host_activate(host, pdev->irq, ipriv->irq_handler,
+				 IRQF_SHARED, ipriv->sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/include/linux/libata.h b/include/linux/libata.h
index e7f10a88efe1..88c6fa84ed74 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -748,14 +748,12 @@ struct ata_port_operations {
 };
 
 struct ata_port_info {
-	struct scsi_host_template	*sht;
 	unsigned long		flags;
 	unsigned long		link_flags;
 	unsigned long		pio_mask;
 	unsigned long		mwdma_mask;
 	unsigned long		udma_mask;
 	struct ata_port_operations *port_ops;
-	irq_handler_t		irq_handler;
 	void 			*private_data;
 };
 
-- 
cgit v1.2.3


From a1efdaba2dbd6fb89e23a87b66d3f4dd92c9f5af Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 12:22:50 +0900
Subject: libata: make reset related methods proper port operations

Currently reset methods are not specified directly in the
ata_port_operations table.  If a LLD wants to use custom reset
methods, it should construct and use a error_handler which uses those
reset methods.  It's done this way for two reasons.

First, the ops table already contained too many methods and adding
four more of them would noticeably increase the amount of necessary
boilerplate code all over low level drivers.

Second, as ->error_handler uses those reset methods, it can get
confusing.  ie. By overriding ->error_handler, those reset ops can be
made useless making layering a bit hazy.

Now that ops table uses inheritance, the first problem doesn't exist
anymore.  The second isn't completely solved but is relieved by
providing default values - most drivers can just override what it has
implemented and don't have to concern itself about higher level
callbacks.  In fact, there currently is no driver which actually
modifies error handling behavior.  Drivers which override
->error_handler just wraps the standard error handler only to prepare
the controller for EH.  I don't think making ops layering strict has
any noticeable benefit.

This patch makes ->prereset, ->softreset, ->hardreset, ->postreset and
their PMP counterparts propoer ops.  Default ops are provided in the
base ops tables and drivers are converted to override individual reset
methods instead of creating custom error_handler.

* ata_std_error_handler() doesn't use sata_std_hardreset() if SCRs
  aren't accessible.  sata_promise doesn't need to use separate
  error_handlers for PATA and SATA anymore.

* softreset is broken for sata_inic162x and sata_sx4.  As libata now
  always prefers hardreset, this doesn't really matter but the ops are
  forced to NULL using ATA_OP_NULL for documentation purpose.

* pata_hpt374 needs to use different prereset for the first and second
  PCI functions.  This used to be done by branching from
  hpt374_error_handler().  The proper way to do this is to use
  separate ops and port_info tables for each function.  Converted.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c          | 53 +++++++++++++--------------------------
 drivers/ata/ata_piix.c      | 21 ++++------------
 drivers/ata/libata-core.c   | 14 +++++++++--
 drivers/ata/libata-eh.c     | 25 +++++++++++++++++++
 drivers/ata/libata-pmp.c    | 47 ++++++++--------------------------
 drivers/ata/libata-sff.c    | 49 ++++++++++++------------------------
 drivers/ata/pata_artop.c    | 34 ++-----------------------
 drivers/ata/pata_atiixp.c   |  7 +-----
 drivers/ata/pata_bf54x.c    | 14 ++---------
 drivers/ata/pata_efar.c     | 15 +----------
 drivers/ata/pata_hpt37x.c   | 61 +++++++++++++++------------------------------
 drivers/ata/pata_hpt3x2n.c  | 16 ++----------
 drivers/ata/pata_icside.c   |  8 +-----
 drivers/ata/pata_it8213.c   | 15 +----------
 drivers/ata/pata_mpc52xx.c  | 10 --------
 drivers/ata/pata_mpiix.c    | 16 +-----------
 drivers/ata/pata_ns87410.c  | 16 +-----------
 drivers/ata/pata_oldpiix.c  | 16 +-----------
 drivers/ata/pata_opti.c     | 17 +------------
 drivers/ata/pata_optidma.c  | 17 +------------
 drivers/ata/pata_pdc2027x.c | 19 ++------------
 drivers/ata/pata_scc.c      | 15 +++--------
 drivers/ata/pata_sis.c      | 15 +----------
 drivers/ata/pata_sl82c105.c |  8 +-----
 drivers/ata/pata_triflex.c  |  7 +-----
 drivers/ata/pata_via.c      | 14 +----------
 drivers/ata/pdc_adma.c      | 11 +++-----
 drivers/ata/sata_fsl.c      | 12 +--------
 drivers/ata/sata_inic162x.c |  5 ++--
 drivers/ata/sata_mv.c       | 16 ++++++------
 drivers/ata/sata_nv.c       | 17 ++++---------
 drivers/ata/sata_promise.c  | 22 +++-------------
 drivers/ata/sata_qstor.c    |  6 +++--
 drivers/ata/sata_sil24.c    | 18 +++++++++----
 drivers/ata/sata_via.c      | 10 ++------
 drivers/scsi/ipr.c          |  3 ++-
 include/linux/libata.h      | 19 +++++++-------
 37 files changed, 186 insertions(+), 502 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index dacb3ef0c3e6..3efa9904f7a0 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -252,9 +252,18 @@ static void ahci_freeze(struct ata_port *ap);
 static void ahci_thaw(struct ata_port *ap);
 static void ahci_pmp_attach(struct ata_port *ap);
 static void ahci_pmp_detach(struct ata_port *ap);
+static int ahci_softreset(struct ata_link *link, unsigned int *class,
+			  unsigned long deadline);
+static int ahci_hardreset(struct ata_link *link, unsigned int *class,
+			  unsigned long deadline);
+static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
+				 unsigned long deadline);
+static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
+				unsigned long deadline);
+static void ahci_postreset(struct ata_link *link, unsigned int *class);
+static int ahci_pmp_softreset(struct ata_link *link, unsigned int *class,
+			      unsigned long deadline);
 static void ahci_error_handler(struct ata_port *ap);
-static void ahci_vt8251_error_handler(struct ata_port *ap);
-static void ahci_p5wdh_error_handler(struct ata_port *ap);
 static void ahci_post_internal_cmd(struct ata_queued_cmd *qc);
 static int ahci_port_resume(struct ata_port *ap);
 static void ahci_dev_config(struct ata_device *dev);
@@ -293,6 +302,10 @@ static struct ata_port_operations ahci_ops = {
 
 	.freeze			= ahci_freeze,
 	.thaw			= ahci_thaw,
+	.softreset		= ahci_softreset,
+	.hardreset		= ahci_hardreset,
+	.postreset		= ahci_postreset,
+	.pmp_softreset		= ahci_pmp_softreset,
 	.error_handler		= ahci_error_handler,
 	.post_internal_cmd	= ahci_post_internal_cmd,
 	.dev_config		= ahci_dev_config,
@@ -314,12 +327,12 @@ static struct ata_port_operations ahci_ops = {
 
 static struct ata_port_operations ahci_vt8251_ops = {
 	.inherits		= &ahci_ops,
-	.error_handler		= ahci_vt8251_error_handler,
+	.hardreset		= ahci_vt8251_hardreset,
 };
 
 static struct ata_port_operations ahci_p5wdh_ops = {
 	.inherits		= &ahci_ops,
-	.error_handler		= ahci_p5wdh_error_handler,
+	.hardreset		= ahci_p5wdh_hardreset,
 };
 
 #define AHCI_HFLAGS(flags)	.private_data	= (void *)(flags)
@@ -1796,37 +1809,7 @@ static void ahci_error_handler(struct ata_port *ap)
 		ahci_start_engine(ap);
 	}
 
-	/* perform recovery */
-	sata_pmp_do_eh(ap, ata_std_prereset, ahci_softreset,
-		       ahci_hardreset, ahci_postreset,
-		       sata_pmp_std_prereset, ahci_pmp_softreset,
-		       sata_pmp_std_hardreset, sata_pmp_std_postreset);
-}
-
-static void ahci_vt8251_error_handler(struct ata_port *ap)
-{
-	if (!(ap->pflags & ATA_PFLAG_FROZEN)) {
-		/* restart engine */
-		ahci_stop_engine(ap);
-		ahci_start_engine(ap);
-	}
-
-	/* perform recovery */
-	ata_do_eh(ap, ata_std_prereset, ahci_softreset, ahci_vt8251_hardreset,
-		  ahci_postreset);
-}
-
-static void ahci_p5wdh_error_handler(struct ata_port *ap)
-{
-	if (!(ap->pflags & ATA_PFLAG_FROZEN)) {
-		/* restart engine */
-		ahci_stop_engine(ap);
-		ahci_start_engine(ap);
-	}
-
-	/* perform recovery */
-	ata_do_eh(ap, ata_std_prereset, ahci_softreset, ahci_p5wdh_hardreset,
-		  ahci_postreset);
+	sata_pmp_error_handler(ap);
 }
 
 static void ahci_post_internal_cmd(struct ata_queued_cmd *qc)
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index bb46b61a7c6b..eafb984313f6 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -162,15 +162,16 @@ struct piix_host_priv {
 
 static int piix_init_one(struct pci_dev *pdev,
 			 const struct pci_device_id *ent);
-static void piix_pata_error_handler(struct ata_port *ap);
+static int piix_pata_prereset(struct ata_link *link, unsigned long deadline);
 static void piix_set_piomode(struct ata_port *ap, struct ata_device *adev);
 static void piix_set_dmamode(struct ata_port *ap, struct ata_device *adev);
 static void ich_set_dmamode(struct ata_port *ap, struct ata_device *adev);
 static int ich_pata_cable_detect(struct ata_port *ap);
 static u8 piix_vmw_bmdma_status(struct ata_port *ap);
+static int piix_sidpr_hardreset(struct ata_link *link, unsigned int *class,
+				unsigned long deadline);
 static int piix_sidpr_scr_read(struct ata_port *ap, unsigned int reg, u32 *val);
 static int piix_sidpr_scr_write(struct ata_port *ap, unsigned int reg, u32 val);
-static void piix_sidpr_error_handler(struct ata_port *ap);
 #ifdef CONFIG_PM
 static int piix_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg);
 static int piix_pci_device_resume(struct pci_dev *pdev);
@@ -299,7 +300,7 @@ static struct ata_port_operations piix_pata_ops = {
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= piix_set_piomode,
 	.set_dmamode		= piix_set_dmamode,
-	.error_handler		= piix_pata_error_handler,
+	.prereset		= piix_pata_prereset,
 };
 
 static struct ata_port_operations piix_vmw_ops = {
@@ -319,9 +320,9 @@ static struct ata_port_operations piix_sata_ops = {
 
 static struct ata_port_operations piix_sidpr_sata_ops = {
 	.inherits		= &piix_sata_ops,
+	.hardreset		= piix_sidpr_hardreset,
 	.scr_read		= piix_sidpr_scr_read,
 	.scr_write		= piix_sidpr_scr_write,
-	.error_handler		= piix_sidpr_error_handler,
 };
 
 static const struct piix_map_db ich5_map_db = {
@@ -645,12 +646,6 @@ static int piix_pata_prereset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-static void piix_pata_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, piix_pata_prereset, ata_std_softreset, NULL,
-			   ata_std_postreset);
-}
-
 /**
  *	piix_set_piomode - Initialize host controller PATA PIO timings
  *	@ap: Port whose timings we are configuring
@@ -1057,12 +1052,6 @@ static int piix_sidpr_hardreset(struct ata_link *link, unsigned int *class,
 	return -EAGAIN;
 }
 
-static void piix_sidpr_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
-			   piix_sidpr_hardreset, ata_std_postreset);
-}
-
 #ifdef CONFIG_PM
 static int piix_broken_suspend(void)
 {
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index aa6bcd79d60a..345f4a6865a8 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -76,6 +76,10 @@ const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
 const struct ata_port_operations ata_base_port_ops = {
 	.irq_clear		= ata_noop_irq_clear,
+	.prereset		= ata_std_prereset,
+	.hardreset		= sata_std_hardreset,
+	.postreset		= ata_std_postreset,
+	.error_handler		= ata_std_error_handler,
 };
 
 const struct ata_port_operations sata_port_ops = {
@@ -87,6 +91,11 @@ const struct ata_port_operations sata_port_ops = {
 
 const struct ata_port_operations sata_pmp_port_ops = {
 	.inherits		= &sata_port_ops,
+
+	.pmp_prereset		= sata_pmp_std_prereset,
+	.pmp_hardreset		= sata_pmp_std_hardreset,
+	.pmp_postreset		= sata_pmp_std_postreset,
+	.error_handler		= sata_pmp_error_handler,
 };
 
 const struct ata_port_operations ata_sff_port_ops = {
@@ -97,6 +106,7 @@ const struct ata_port_operations ata_sff_port_ops = {
 
 	.freeze			= ata_bmdma_freeze,
 	.thaw			= ata_bmdma_thaw,
+	.softreset		= ata_std_softreset,
 	.error_handler		= ata_bmdma_error_handler,
 	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
 
@@ -7935,7 +7945,6 @@ EXPORT_SYMBOL_GPL(ata_bmdma_status);
 EXPORT_SYMBOL_GPL(ata_bmdma_stop);
 EXPORT_SYMBOL_GPL(ata_bmdma_freeze);
 EXPORT_SYMBOL_GPL(ata_bmdma_thaw);
-EXPORT_SYMBOL_GPL(ata_bmdma_drive_eh);
 EXPORT_SYMBOL_GPL(ata_bmdma_error_handler);
 EXPORT_SYMBOL_GPL(ata_bmdma_post_internal_cmd);
 EXPORT_SYMBOL_GPL(ata_port_probe);
@@ -8005,7 +8014,7 @@ EXPORT_SYMBOL_GPL(sata_pmp_qc_defer_cmd_switch);
 EXPORT_SYMBOL_GPL(sata_pmp_std_prereset);
 EXPORT_SYMBOL_GPL(sata_pmp_std_hardreset);
 EXPORT_SYMBOL_GPL(sata_pmp_std_postreset);
-EXPORT_SYMBOL_GPL(sata_pmp_do_eh);
+EXPORT_SYMBOL_GPL(sata_pmp_error_handler);
 
 EXPORT_SYMBOL_GPL(__ata_ehi_push_desc);
 EXPORT_SYMBOL_GPL(ata_ehi_push_desc);
@@ -8024,6 +8033,7 @@ EXPORT_SYMBOL_GPL(ata_eh_thaw_port);
 EXPORT_SYMBOL_GPL(ata_eh_qc_complete);
 EXPORT_SYMBOL_GPL(ata_eh_qc_retry);
 EXPORT_SYMBOL_GPL(ata_do_eh);
+EXPORT_SYMBOL_GPL(ata_std_error_handler);
 EXPORT_SYMBOL_GPL(ata_irq_on);
 EXPORT_SYMBOL_GPL(ata_dev_try_classify);
 
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index 0d0a2c0ab9e7..ec32082356cb 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -2814,6 +2814,7 @@ void ata_eh_finish(struct ata_port *ap)
 /**
  *	ata_do_eh - do standard error handling
  *	@ap: host port to handle error for
+ *
  *	@prereset: prereset method (can be NULL)
  *	@softreset: softreset method (can be NULL)
  *	@hardreset: hardreset method (can be NULL)
@@ -2844,6 +2845,30 @@ void ata_do_eh(struct ata_port *ap, ata_prereset_fn_t prereset,
 	ata_eh_finish(ap);
 }
 
+/**
+ *	ata_std_error_handler - standard error handler
+ *	@ap: host port to handle error for
+ *
+ *	Standard error handler
+ *
+ *	LOCKING:
+ *	Kernel thread context (may sleep).
+ */
+void ata_std_error_handler(struct ata_port *ap)
+{
+	struct ata_port_operations *ops = ap->ops;
+	ata_reset_fn_t hardreset = ops->hardreset;
+
+	/* sata_std_hardreset is inherited to all drivers from
+	 * ata_base_port_ops.  Ignore it if SCR access is not
+	 * available.
+	 */
+	if (hardreset == sata_std_hardreset && !sata_scr_valid(&ap->link))
+		hardreset = NULL;
+
+	ata_do_eh(ap, ops->prereset, ops->softreset, hardreset, ops->postreset);
+}
+
 #ifdef CONFIG_PM
 /**
  *	ata_eh_handle_port_suspend - perform port suspend operation
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index 39e036c8a2bc..a7cb1498c9b2 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -962,14 +962,6 @@ static int sata_pmp_handle_link_fail(struct ata_link *link, int *link_tries)
 /**
  *	sata_pmp_eh_recover - recover PMP-enabled port
  *	@ap: ATA port to recover
- *	@prereset: prereset method (can be NULL)
- *	@softreset: softreset method
- *	@hardreset: hardreset method
- *	@postreset: postreset method (can be NULL)
- *	@pmp_prereset: PMP prereset method (can be NULL)
- *	@pmp_softreset: PMP softreset method (can be NULL)
- *	@pmp_hardreset: PMP hardreset method (can be NULL)
- *	@pmp_postreset: PMP postreset method (can be NULL)
  *
  *	Drive EH recovery operation for PMP enabled port @ap.  This
  *	function recovers host and PMP ports with proper retrials and
@@ -982,12 +974,9 @@ static int sata_pmp_handle_link_fail(struct ata_link *link, int *link_tries)
  *	RETURNS:
  *	0 on success, -errno on failure.
  */
-static int sata_pmp_eh_recover(struct ata_port *ap,
-		ata_prereset_fn_t prereset, ata_reset_fn_t softreset,
-		ata_reset_fn_t hardreset, ata_postreset_fn_t postreset,
-		ata_prereset_fn_t pmp_prereset, ata_reset_fn_t pmp_softreset,
-		ata_reset_fn_t pmp_hardreset, ata_postreset_fn_t pmp_postreset)
+static int sata_pmp_eh_recover(struct ata_port *ap)
 {
+	struct ata_port_operations *ops = ap->ops;
 	int pmp_tries, link_tries[SATA_PMP_MAX_PORTS];
 	struct ata_link *pmp_link = &ap->link;
 	struct ata_device *pmp_dev = pmp_link->device;
@@ -1005,8 +994,8 @@ static int sata_pmp_eh_recover(struct ata_port *ap,
  retry:
 	/* PMP attached? */
 	if (!ap->nr_pmp_links) {
-		rc = ata_eh_recover(ap, prereset, softreset, hardreset,
-				    postreset, NULL);
+		rc = ata_eh_recover(ap, ops->prereset, ops->softreset,
+				    ops->hardreset, ops->postreset, NULL);
 		if (rc) {
 			ata_link_for_each_dev(dev, &ap->link)
 				ata_dev_disable(dev);
@@ -1024,8 +1013,8 @@ static int sata_pmp_eh_recover(struct ata_port *ap,
 	}
 
 	/* recover pmp */
-	rc = sata_pmp_eh_recover_pmp(ap, prereset, softreset, hardreset,
-				     postreset);
+	rc = sata_pmp_eh_recover_pmp(ap, ops->prereset, ops->softreset,
+				     ops->hardreset, ops->postreset);
 	if (rc)
 		goto pmp_fail;
 
@@ -1035,8 +1024,8 @@ static int sata_pmp_eh_recover(struct ata_port *ap,
 		goto pmp_fail;
 
 	/* recover links */
-	rc = ata_eh_recover(ap, pmp_prereset, pmp_softreset, pmp_hardreset,
-			    pmp_postreset, &link);
+	rc = ata_eh_recover(ap, ops->pmp_prereset, ops->pmp_softreset,
+			    ops->pmp_hardreset, ops->pmp_postreset, &link);
 	if (rc)
 		goto link_fail;
 
@@ -1132,16 +1121,8 @@ static int sata_pmp_eh_recover(struct ata_port *ap,
 }
 
 /**
- *	sata_pmp_do_eh - do standard error handling for PMP-enabled host
+ *	sata_pmp_error_handler - do standard error handling for PMP-enabled host
  *	@ap: host port to handle error for
- *	@prereset: prereset method (can be NULL)
- *	@softreset: softreset method
- *	@hardreset: hardreset method
- *	@postreset: postreset method (can be NULL)
- *	@pmp_prereset: PMP prereset method (can be NULL)
- *	@pmp_softreset: PMP softreset method (can be NULL)
- *	@pmp_hardreset: PMP hardreset method (can be NULL)
- *	@pmp_postreset: PMP postreset method (can be NULL)
  *
  *	Perform standard error handling sequence for PMP-enabled host
  *	@ap.
@@ -1149,16 +1130,10 @@ static int sata_pmp_eh_recover(struct ata_port *ap,
  *	LOCKING:
  *	Kernel thread context (may sleep).
  */
-void sata_pmp_do_eh(struct ata_port *ap,
-		ata_prereset_fn_t prereset, ata_reset_fn_t softreset,
-		ata_reset_fn_t hardreset, ata_postreset_fn_t postreset,
-		ata_prereset_fn_t pmp_prereset, ata_reset_fn_t pmp_softreset,
-		ata_reset_fn_t pmp_hardreset, ata_postreset_fn_t pmp_postreset)
+void sata_pmp_error_handler(struct ata_port *ap)
 {
 	ata_eh_autopsy(ap);
 	ata_eh_report(ap);
-	sata_pmp_eh_recover(ap, prereset, softreset, hardreset, postreset,
-			    pmp_prereset, pmp_softreset, pmp_hardreset,
-			    pmp_postreset);
+	sata_pmp_eh_recover(ap);
 	ata_eh_finish(ap);
 }
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 6223ec042c80..2a229a1d3211 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -396,28 +396,21 @@ void ata_bmdma_thaw(struct ata_port *ap)
 }
 
 /**
- *	ata_bmdma_drive_eh - Perform EH with given methods for BMDMA controller
+ *	ata_bmdma_error_handler - Stock error handler for BMDMA controller
  *	@ap: port to handle error for
- *	@prereset: prereset method (can be NULL)
- *	@softreset: softreset method (can be NULL)
- *	@hardreset: hardreset method (can be NULL)
- *	@postreset: postreset method (can be NULL)
  *
- *	Handle error for ATA BMDMA controller.  It can handle both
+ *	Stock error handler for BMDMA controller.  It can handle both
  *	PATA and SATA controllers.  Many controllers should be able to
  *	use this EH as-is or with some added handling before and
  *	after.
  *
- *	This function is intended to be used for constructing
- *	->error_handler callback by low level drivers.
- *
  *	LOCKING:
  *	Kernel thread context (may sleep)
  */
-void ata_bmdma_drive_eh(struct ata_port *ap, ata_prereset_fn_t prereset,
-			ata_reset_fn_t softreset, ata_reset_fn_t hardreset,
-			ata_postreset_fn_t postreset)
+void ata_bmdma_error_handler(struct ata_port *ap)
 {
+	ata_reset_fn_t softreset = ap->ops->softreset;
+	ata_reset_fn_t hardreset = ap->ops->hardreset;
 	struct ata_queued_cmd *qc;
 	unsigned long flags;
 	int thaw = 0;
@@ -460,29 +453,19 @@ void ata_bmdma_drive_eh(struct ata_port *ap, ata_prereset_fn_t prereset,
 		ata_eh_thaw_port(ap);
 
 	/* PIO and DMA engines have been stopped, perform recovery */
-	ata_do_eh(ap, prereset, softreset, hardreset, postreset);
-}
-
-/**
- *	ata_bmdma_error_handler - Stock error handler for BMDMA controller
- *	@ap: port to handle error for
- *
- *	Stock error handler for BMDMA controller.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep)
- */
-void ata_bmdma_error_handler(struct ata_port *ap)
-{
-	ata_reset_fn_t softreset = NULL, hardreset = NULL;
 
-	if (ap->ioaddr.ctl_addr)
-		softreset = ata_std_softreset;
-	if (sata_scr_valid(&ap->link))
-		hardreset = sata_std_hardreset;
+	/* ata_std_softreset and sata_std_hardreset are inherited to
+	 * all SFF drivers from ata_sff_port_ops.  Ignore softreset if
+	 * ctl isn't accessible.  Ignore hardreset if SCR access isn't
+	 * available.
+	 */
+	if (softreset == ata_std_softreset && !ap->ioaddr.ctl_addr)
+		softreset = NULL;
+	if (hardreset == sata_std_hardreset && !sata_scr_valid(&ap->link))
+		hardreset = NULL;
 
-	ata_bmdma_drive_eh(ap, ata_std_prereset, softreset, hardreset,
-			   ata_std_postreset);
+	ata_do_eh(ap, ap->ops->prereset, softreset, hardreset,
+		  ap->ops->postreset);
 }
 
 /**
diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c
index ebd15cadf15f..b6d8c4d0e6c2 100644
--- a/drivers/ata/pata_artop.c
+++ b/drivers/ata/pata_artop.c
@@ -55,21 +55,6 @@ static int artop6210_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	artop6210_error_handler - Probe specified port on PATA host controller
- *	@ap: Port to probe
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void artop6210_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, artop6210_pre_reset,
-				    ata_std_softreset, NULL,
-				    ata_std_postreset);
-}
-
 /**
  *	artop6260_pre_reset	-	check for 40/80 pin
  *	@link: link
@@ -113,21 +98,6 @@ static int artop6260_cable_detect(struct ata_port *ap)
 	return ATA_CBL_PATA80;
 }
 
-/**
- *	artop6260_error_handler - Probe specified port on PATA host controller
- *	@ap: Port to probe
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void artop6260_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, artop6260_pre_reset,
-				    ata_std_softreset, NULL,
-				    ata_std_postreset);
-}
-
 /**
  *	artop6210_load_piomode - Load a set of PATA PIO timings
  *	@ap: Port whose timings we are configuring
@@ -322,7 +292,7 @@ static struct ata_port_operations artop6210_ops = {
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= artop6210_set_piomode,
 	.set_dmamode		= artop6210_set_dmamode,
-	.error_handler		= artop6210_error_handler,
+	.prereset		= artop6210_pre_reset,
 };
 
 static struct ata_port_operations artop6260_ops = {
@@ -330,7 +300,7 @@ static struct ata_port_operations artop6260_ops = {
 	.cable_detect		= artop6260_cable_detect,
 	.set_piomode		= artop6260_set_piomode,
 	.set_dmamode		= artop6260_set_dmamode,
-	.error_handler		= artop6260_error_handler,
+	.prereset		= artop6260_pre_reset,
 };
 
 
diff --git a/drivers/ata/pata_atiixp.c b/drivers/ata/pata_atiixp.c
index 0bea7e75d2d6..56a65baddd4a 100644
--- a/drivers/ata/pata_atiixp.c
+++ b/drivers/ata/pata_atiixp.c
@@ -48,11 +48,6 @@ static int atiixp_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-static void atiixp_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, atiixp_pre_reset, ata_std_softreset, NULL,   ata_std_postreset);
-}
-
 static int atiixp_cable_detect(struct ata_port *ap)
 {
 	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
@@ -235,7 +230,7 @@ static struct ata_port_operations atiixp_port_ops = {
 	.cable_detect	= atiixp_cable_detect,
 	.set_piomode	= atiixp_set_piomode,
 	.set_dmamode	= atiixp_set_dmamode,
-	.error_handler	= atiixp_error_handler,
+	.prereset	= atiixp_pre_reset,
 };
 
 static int atiixp_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c
index 7a22ef483061..eea275acb2a8 100644
--- a/drivers/ata/pata_bf54x.c
+++ b/drivers/ata/pata_bf54x.c
@@ -1314,17 +1314,6 @@ static void bfin_std_postreset(struct ata_link *link, unsigned int *classes)
 	write_atapi_register(base, ATA_REG_CTRL, ap->ctl);
 }
 
-/**
- *	bfin_error_handler - Stock error handler for DMA controller
- *	@ap: port to handle error for
- */
-
-static void bfin_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, ata_std_prereset, bfin_std_softreset, NULL,
-			   bfin_std_postreset);
-}
-
 static void bfin_port_stop(struct ata_port *ap)
 {
 	dev_dbg(ap->dev, "in atapi port stop\n");
@@ -1385,7 +1374,8 @@ static const struct ata_port_operations bfin_pata_ops = {
 
 	.freeze			= bfin_bmdma_freeze,
 	.thaw			= bfin_bmdma_thaw,
-	.error_handler		= bfin_error_handler,
+	.softreset		= bfin_std_softreset,
+	.postreset		= bfin_std_postreset,
 	.post_internal_cmd	= bfin_bmdma_stop,
 
 	.irq_clear		= bfin_irq_clear,
diff --git a/drivers/ata/pata_efar.c b/drivers/ata/pata_efar.c
index 2f5b4848456a..1d839a57068e 100644
--- a/drivers/ata/pata_efar.c
+++ b/drivers/ata/pata_efar.c
@@ -48,19 +48,6 @@ static int efar_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	efar_probe_reset - Probe specified port on PATA host controller
- *	@ap: Port to probe
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void efar_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, efar_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	efar_cable_detect	-	check for 40/80 pin
  *	@ap: Port
@@ -241,7 +228,7 @@ static struct ata_port_operations efar_ops = {
 	.cable_detect		= efar_cable_detect,
 	.set_piomode		= efar_set_piomode,
 	.set_dmamode		= efar_set_dmamode,
-	.error_handler		= efar_error_handler,
+	.prereset		= efar_pre_reset,
 };
 
 
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index fb37e3a161fc..c10fcd31418d 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -341,19 +341,7 @@ static int hpt37x_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	hpt37x_error_handler	-	reset the hpt374
- *	@ap: ATA port to reset
- *
- *	Perform probe for HPT37x, except for HPT374 channel 2
- */
-
-static void hpt37x_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, hpt37x_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
-static int hpt374_pre_reset(struct ata_link *link, unsigned long deadline)
+static int hpt374_fn1_pre_reset(struct ata_link *link, unsigned long deadline)
 {
 	static const struct pci_bits hpt37x_enable_bits[] = {
 		{ 0x50, 1, 0x04, 0x04 },
@@ -389,25 +377,6 @@ static int hpt374_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	hpt374_error_handler	-	reset the hpt374
- *	@classes:
- *
- *	The 374 cable detect is a little different due to the extra
- *	channels. The function 0 channels work like usual but function 1
- *	is special
- */
-
-static void hpt374_error_handler(struct ata_port *ap)
-{
-	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
-
-	if (!(PCI_FUNC(pdev->devfn) & 1))
-		hpt37x_error_handler(ap);
-	else
-		ata_bmdma_drive_eh(ap, hpt374_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	hpt370_set_piomode		-	PIO setup
  *	@ap: ATA interface
@@ -635,7 +604,7 @@ static struct ata_port_operations hpt370_port_ops = {
 	.mode_filter	= hpt370_filter,
 	.set_piomode	= hpt370_set_piomode,
 	.set_dmamode	= hpt370_set_dmamode,
-	.error_handler	= hpt37x_error_handler,
+	.prereset	= hpt37x_pre_reset,
 };
 
 /*
@@ -659,17 +628,17 @@ static struct ata_port_operations hpt372_port_ops = {
 
 	.set_piomode	= hpt372_set_piomode,
 	.set_dmamode	= hpt372_set_dmamode,
-	.error_handler	= hpt37x_error_handler,
+	.prereset	= hpt37x_pre_reset,
 };
 
 /*
  *	Configuration for HPT374. Mode setting works like 372 and friends
- *	but we have a different cable detection procedure.
+ *	but we have a different cable detection procedure for function 1.
  */
 
-static struct ata_port_operations hpt374_port_ops = {
+static struct ata_port_operations hpt374_fn1_port_ops = {
 	.inherits	= &hpt372_port_ops,
-	.error_handler	= hpt374_error_handler,
+	.prereset	= hpt374_fn1_pre_reset,
 };
 
 /**
@@ -821,13 +790,20 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.udma_mask = ATA_UDMA6,
 		.port_ops = &hpt372_port_ops
 	};
-	/* HPT374 - UDMA100 */
-	static const struct ata_port_info info_hpt374 = {
+	/* HPT374 - UDMA100, function 1 uses different prereset method */
+	static const struct ata_port_info info_hpt374_fn0 = {
+		.flags = ATA_FLAG_SLAVE_POSS,
+		.pio_mask = 0x1f,
+		.mwdma_mask = 0x07,
+		.udma_mask = ATA_UDMA5,
+		.port_ops = &hpt372_port_ops
+	};
+	static const struct ata_port_info info_hpt374_fn1 = {
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = 0x1f,
 		.mwdma_mask = 0x07,
 		.udma_mask = ATA_UDMA5,
-		.port_ops = &hpt374_port_ops
+		.port_ops = &hpt374_fn1_port_ops
 	};
 
 	static const int MHz[4] = { 33, 40, 50, 66 };
@@ -912,7 +888,10 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 				break;
 			case PCI_DEVICE_ID_TTI_HPT374:
 				chip_table = &hpt374;
-				ppi[0] = &info_hpt374;
+				if (!(PCI_FUNC(dev->devfn) & 1))
+					*ppi = &info_hpt374_fn0;
+				else
+					*ppi = &info_hpt374_fn1;
 				break;
 			default:
 				printk(KERN_ERR "pata_hpt37x: PCI table is bogus please report (%d).\n", dev->device);
diff --git a/drivers/ata/pata_hpt3x2n.c b/drivers/ata/pata_hpt3x2n.c
index c774be93ae04..cd44ee3d3cc1 100644
--- a/drivers/ata/pata_hpt3x2n.c
+++ b/drivers/ata/pata_hpt3x2n.c
@@ -148,7 +148,7 @@ static int hpt3x2n_cable_detect(struct ata_port *ap)
  *	Reset the hardware and state machine,
  */
 
-static int hpt3xn_pre_reset(struct ata_link *link, unsigned long deadline)
+static int hpt3x2n_pre_reset(struct ata_link *link, unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
 	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
@@ -159,18 +159,6 @@ static int hpt3xn_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	hpt3x2n_error_handler	-	probe the hpt3x2n bus
- *	@ap: ATA port to reset
- *
- *	Perform the probe reset handling for the 3x2N
- */
-
-static void hpt3x2n_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, hpt3xn_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	hpt3x2n_set_piomode		-	PIO setup
  *	@ap: ATA interface
@@ -355,7 +343,7 @@ static struct ata_port_operations hpt3x2n_port_ops = {
 	.cable_detect	= hpt3x2n_cable_detect,
 	.set_piomode	= hpt3x2n_set_piomode,
 	.set_dmamode	= hpt3x2n_set_dmamode,
-	.error_handler	= hpt3x2n_error_handler,
+	.prereset	= hpt3x2n_pre_reset,
 };
 
 /**
diff --git a/drivers/ata/pata_icside.c b/drivers/ata/pata_icside.c
index ff16b0eaa2c2..13d43e9dd279 100644
--- a/drivers/ata/pata_icside.c
+++ b/drivers/ata/pata_icside.c
@@ -332,12 +332,6 @@ static void pata_icside_postreset(struct ata_link *link, unsigned int *classes)
 	}
 }
 
-static void pata_icside_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset, NULL,
-			   pata_icside_postreset);
-}
-
 static struct ata_port_operations pata_icside_port_ops = {
 	.inherits		= &ata_sff_port_ops,
 	/* no need to build any PRD tables for DMA */
@@ -350,7 +344,7 @@ static struct ata_port_operations pata_icside_port_ops = {
 
 	.cable_detect		= ata_cable_40wire,
 	.set_dmamode		= pata_icside_set_dmamode,
-	.error_handler		= pata_icside_error_handler,
+	.postreset		= pata_icside_postreset,
 	.post_internal_cmd	= pata_icside_bmdma_stop,
 };
 
diff --git a/drivers/ata/pata_it8213.c b/drivers/ata/pata_it8213.c
index d23a46b75028..84ab89e8a247 100644
--- a/drivers/ata/pata_it8213.c
+++ b/drivers/ata/pata_it8213.c
@@ -43,19 +43,6 @@ static int it8213_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	it8213_error_handler - Probe specified port on PATA host controller
- *	@ap: Port to probe
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void it8213_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, it8213_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	it8213_cable_detect	-	check for 40/80 pin
  *	@ap: Port
@@ -252,7 +239,7 @@ static struct ata_port_operations it8213_ops = {
 	.cable_detect		= it8213_cable_detect,
 	.set_piomode		= it8213_set_piomode,
 	.set_dmamode		= it8213_set_dmamode,
-	.error_handler		= it8213_error_handler,
+	.prereset		= it8213_pre_reset,
 };
 
 
diff --git a/drivers/ata/pata_mpc52xx.c b/drivers/ata/pata_mpc52xx.c
index ac7c0822b1a7..fec93196710e 100644
--- a/drivers/ata/pata_mpc52xx.c
+++ b/drivers/ata/pata_mpc52xx.c
@@ -255,15 +255,6 @@ mpc52xx_ata_dev_select(struct ata_port *ap, unsigned int device)
 	ata_std_dev_select(ap,device);
 }
 
-static void
-mpc52xx_ata_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset, NULL,
-			ata_std_postreset);
-}
-
-
-
 static struct scsi_host_template mpc52xx_ata_sht = {
 	ATA_PIO_SHT(DRV_NAME),
 };
@@ -273,7 +264,6 @@ static struct ata_port_operations mpc52xx_ata_port_ops = {
 	.dev_select		= mpc52xx_ata_dev_select,
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= mpc52xx_ata_set_piomode,
-	.error_handler		= mpc52xx_ata_error_handler,
 	.post_internal_cmd	= ATA_OP_NULL,
 };
 
diff --git a/drivers/ata/pata_mpiix.c b/drivers/ata/pata_mpiix.c
index dab54f8a272d..1b9d0d412ebf 100644
--- a/drivers/ata/pata_mpiix.c
+++ b/drivers/ata/pata_mpiix.c
@@ -58,20 +58,6 @@ static int mpiix_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	mpiix_error_handler		-	probe reset
- *	@ap: ATA port
- *
- *	Perform the ATA probe and bus reset sequence plus specific handling
- *	for this hardware. The MPIIX has the enable bits in a different place
- *	to PIIX4 and friends. As a pure PIO device it has no cable detect
- */
-
-static void mpiix_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, mpiix_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	mpiix_set_piomode	-	set initial PIO mode data
  *	@ap: ATA interface
@@ -159,7 +145,7 @@ static struct ata_port_operations mpiix_port_ops = {
 	.qc_issue	= mpiix_qc_issue_prot,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= mpiix_set_piomode,
-	.error_handler	= mpiix_error_handler,
+	.prereset	= mpiix_pre_reset,
 };
 
 static int mpiix_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
index 5a043e426480..4d2eefee7387 100644
--- a/drivers/ata/pata_ns87410.c
+++ b/drivers/ata/pata_ns87410.c
@@ -53,20 +53,6 @@ static int ns87410_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	ns87410_error_handler		-	probe reset
- *	@ap: ATA port
- *
- *	Perform the ATA probe and bus reset sequence plus specific handling
- *	for this hardware. The MPIIX has the enable bits in a different place
- *	to PIIX4 and friends. As a pure PIO device it has no cable detect
- */
-
-static void ns87410_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, ns87410_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	ns87410_set_piomode	-	set initial PIO mode data
  *	@ap: ATA interface
@@ -152,7 +138,7 @@ static struct ata_port_operations ns87410_port_ops = {
 	.qc_issue	= ns87410_qc_issue_prot,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= ns87410_set_piomode,
-	.error_handler	= ns87410_error_handler,
+	.prereset	= ns87410_pre_reset,
 };
 
 static int ns87410_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_oldpiix.c b/drivers/ata/pata_oldpiix.c
index 7001b756819e..c1da79a76439 100644
--- a/drivers/ata/pata_oldpiix.c
+++ b/drivers/ata/pata_oldpiix.c
@@ -50,20 +50,6 @@ static int oldpiix_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	oldpiix_pata_error_handler - Probe specified port on PATA host controller
- *	@ap: Port to probe
- *	@classes:
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void oldpiix_pata_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, oldpiix_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	oldpiix_set_piomode - Initialize host controller PATA PIO timings
  *	@ap: Port whose timings we are configuring
@@ -229,7 +215,7 @@ static struct ata_port_operations oldpiix_pata_ops = {
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= oldpiix_set_piomode,
 	.set_dmamode		= oldpiix_set_dmamode,
-	.error_handler		= oldpiix_pata_error_handler,
+	.prereset		= oldpiix_pre_reset,
 };
 
 
diff --git a/drivers/ata/pata_opti.c b/drivers/ata/pata_opti.c
index 5a5f20e03fc0..4ddd03a67775 100644
--- a/drivers/ata/pata_opti.c
+++ b/drivers/ata/pata_opti.c
@@ -67,21 +67,6 @@ static int opti_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	opti_probe_reset		-	probe reset
- *	@ap: ATA port
- *
- *	Perform the ATA probe and bus reset sequence plus specific handling
- *	for this hardware. The Opti needs little handling - we have no UDMA66
- *	capability that needs cable detection. All we must do is check the port
- *	is enabled.
- */
-
-static void opti_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, opti_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	opti_write_reg		-	control register setup
  *	@ap: ATA port
@@ -172,7 +157,7 @@ static struct ata_port_operations opti_port_ops = {
 	.inherits	= &ata_sff_port_ops,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= opti_set_piomode,
-	.error_handler	= opti_error_handler,
+	.prereset	= opti_pre_reset,
 };
 
 static int opti_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_optidma.c b/drivers/ata/pata_optidma.c
index ba2819ff964b..36ac147de178 100644
--- a/drivers/ata/pata_optidma.c
+++ b/drivers/ata/pata_optidma.c
@@ -67,21 +67,6 @@ static int optidma_pre_reset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	optidma_probe_reset		-	probe reset
- *	@ap: ATA port
- *
- *	Perform the ATA probe and bus reset sequence plus specific handling
- *	for this hardware. The Opti needs little handling - we have no UDMA66
- *	capability that needs cable detection. All we must do is check the port
- *	is enabled.
- */
-
-static void optidma_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, optidma_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	optidma_unlock		-	unlock control registers
  *	@ap: ATA port
@@ -359,7 +344,7 @@ static struct ata_port_operations optidma_port_ops = {
 	.set_piomode	= optidma_set_pio_mode,
 	.set_dmamode	= optidma_set_dma_mode,
 	.set_mode	= optidma_set_mode,
-	.error_handler	= optidma_error_handler,
+	.prereset	= optidma_pre_reset,
 };
 
 static struct ata_port_operations optiplus_port_ops = {
diff --git a/drivers/ata/pata_pdc2027x.c b/drivers/ata/pata_pdc2027x.c
index f619c20dd192..d235c9f92d09 100644
--- a/drivers/ata/pata_pdc2027x.c
+++ b/drivers/ata/pata_pdc2027x.c
@@ -63,7 +63,7 @@ enum {
 };
 
 static int pdc2027x_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
-static void pdc2027x_error_handler(struct ata_port *ap);
+static int pdc2027x_prereset(struct ata_link *link, unsigned long deadline);
 static void pdc2027x_set_piomode(struct ata_port *ap, struct ata_device *adev);
 static void pdc2027x_set_dmamode(struct ata_port *ap, struct ata_device *adev);
 static int pdc2027x_check_atapi_dma(struct ata_queued_cmd *qc);
@@ -136,7 +136,7 @@ static struct ata_port_operations pdc2027x_pata100_ops = {
 	.inherits		= &ata_bmdma_port_ops,
 	.check_atapi_dma	= pdc2027x_check_atapi_dma,
 	.cable_detect		= pdc2027x_cable_detect,
-	.error_handler		= pdc2027x_error_handler,
+	.prereset		= pdc2027x_prereset,
 };
 
 static struct ata_port_operations pdc2027x_pata133_ops = {
@@ -251,21 +251,6 @@ static int pdc2027x_prereset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-/**
- *	pdc2027x_error_handler - Perform reset on PATA port and classify
- *	@ap: Port to reset
- *
- *	Reset PATA phy and classify attached devices.
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void pdc2027x_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, pdc2027x_prereset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	pdc2720x_mode_filter	-	mode selection filter
  *	@adev: ATA device
diff --git a/drivers/ata/pata_scc.c b/drivers/ata/pata_scc.c
index 033d1f3a82de..07f2d7a6f1a8 100644
--- a/drivers/ata/pata_scc.c
+++ b/drivers/ata/pata_scc.c
@@ -904,17 +904,6 @@ static void scc_std_postreset(struct ata_link *link, unsigned int *classes)
 	DPRINTK("EXIT\n");
 }
 
-/**
- *	scc_error_handler - Stock error handler for BMDMA controller
- *	@ap: port to handle error for
- */
-
-static void scc_error_handler (struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, scc_pata_prereset, scc_std_softreset, NULL,
-			   scc_std_postreset);
-}
-
 /**
  *	scc_bmdma_irq_clear - Clear PCI IDE BMDMA interrupt.
  *	@ap: Port associated with this ATA transaction.
@@ -992,7 +981,9 @@ static const struct ata_port_operations scc_pata_ops = {
 	.data_xfer		= scc_data_xfer,
 
 	.freeze			= scc_bmdma_freeze,
-	.error_handler		= scc_error_handler,
+	.prereset		= scc_pata_prereset,
+	.softreset		= scc_std_softreset,
+	.postreset		= scc_std_postreset,
 	.post_internal_cmd	= scc_bmdma_stop,
 
 	.irq_clear		= scc_bmdma_irq_clear,
diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
index 28abfc26e7a4..793e6714df8c 100644
--- a/drivers/ata/pata_sis.c
+++ b/drivers/ata/pata_sis.c
@@ -160,19 +160,6 @@ static int sis_pre_reset(struct ata_link *link, unsigned long deadline)
 }
 
 
-/**
- *	sis_error_handler - Probe specified port on PATA host controller
- *	@ap: Port to probe
- *
- *	LOCKING:
- *	None (inherited from caller).
- */
-
-static void sis_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, sis_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	sis_set_fifo	-	Set RWP fifo bits for this device
  *	@ap: Port
@@ -526,7 +513,7 @@ static struct ata_port_operations sis_133_for_sata_ops = {
 
 static struct ata_port_operations sis_base_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.error_handler		= sis_error_handler,
+	.prereset		= sis_pre_reset,
 };
 
 static struct ata_port_operations sis_133_ops = {
diff --git a/drivers/ata/pata_sl82c105.c b/drivers/ata/pata_sl82c105.c
index 1d97f920bd2b..bee11ca8f55a 100644
--- a/drivers/ata/pata_sl82c105.c
+++ b/drivers/ata/pata_sl82c105.c
@@ -64,12 +64,6 @@ static int sl82c105_pre_reset(struct ata_link *link, unsigned long deadline)
 }
 
 
-static void sl82c105_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, sl82c105_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
-
 /**
  *	sl82c105_configure_piomode	-	set chip PIO timing
  *	@ap: ATA interface
@@ -245,7 +239,7 @@ static struct ata_port_operations sl82c105_port_ops = {
 	.bmdma_stop	= sl82c105_bmdma_stop,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= sl82c105_set_piomode,
-	.error_handler	= sl82c105_error_handler,
+	.prereset	= sl82c105_pre_reset,
 };
 
 /**
diff --git a/drivers/ata/pata_triflex.c b/drivers/ata/pata_triflex.c
index f07b0e5df222..bd546a389ce1 100644
--- a/drivers/ata/pata_triflex.c
+++ b/drivers/ata/pata_triflex.c
@@ -71,11 +71,6 @@ static int triflex_prereset(struct ata_link *link, unsigned long deadline)
 
 
 
-static void triflex_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, triflex_prereset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	triflex_load_timing		-	timing configuration
  *	@ap: ATA interface
@@ -189,7 +184,7 @@ static struct ata_port_operations triflex_port_ops = {
 	.bmdma_stop	= triflex_bmdma_stop,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= triflex_set_piomode,
-	.error_handler	= triflex_error_handler,
+	.prereset	= triflex_prereset,
 };
 
 static int triflex_init_one(struct pci_dev *dev, const struct pci_device_id *id)
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
index f4092cbd566f..2928fa173132 100644
--- a/drivers/ata/pata_via.c
+++ b/drivers/ata/pata_via.c
@@ -214,18 +214,6 @@ static int via_pre_reset(struct ata_link *link, unsigned long deadline)
 }
 
 
-/**
- *	via_error_handler		-	reset for VIA chips
- *	@ap: ATA port
- *
- *	Handle the reset callback for the later chips with cable detect
- */
-
-static void via_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, via_pre_reset, ata_std_softreset, NULL, ata_std_postreset);
-}
-
 /**
  *	via_do_set_mode	-	set initial PIO mode data
  *	@ap: ATA interface
@@ -343,7 +331,7 @@ static struct ata_port_operations via_port_ops = {
 	.cable_detect	= via_cable_detect,
 	.set_piomode	= via_set_piomode,
 	.set_dmamode	= via_set_dmamode,
-	.error_handler	= via_error_handler,
+	.prereset	= via_pre_reset,
 };
 
 static struct ata_port_operations via_port_ops_noirq = {
diff --git a/drivers/ata/pdc_adma.c b/drivers/ata/pdc_adma.c
index a5706149af6b..5ed065d0ab4c 100644
--- a/drivers/ata/pdc_adma.c
+++ b/drivers/ata/pdc_adma.c
@@ -140,7 +140,7 @@ static void adma_bmdma_stop(struct ata_queued_cmd *qc);
 static u8 adma_bmdma_status(struct ata_port *ap);
 static void adma_freeze(struct ata_port *ap);
 static void adma_thaw(struct ata_port *ap);
-static void adma_error_handler(struct ata_port *ap);
+static int adma_prereset(struct ata_link *link, unsigned long deadline);
 
 static struct scsi_host_template adma_ata_sht = {
 	ATA_BASE_SHT(DRV_NAME),
@@ -166,7 +166,8 @@ static struct ata_port_operations adma_ata_ops = {
 
 	.freeze			= adma_freeze,
 	.thaw			= adma_thaw,
-	.error_handler		= adma_error_handler,
+	.prereset		= adma_prereset,
+	.softreset		= ata_std_softreset,
 
 	.port_start		= adma_port_start,
 	.port_stop		= adma_port_stop,
@@ -292,12 +293,6 @@ static int adma_prereset(struct ata_link *link, unsigned long deadline)
 	return ata_std_prereset(link, deadline);
 }
 
-static void adma_error_handler(struct ata_port *ap)
-{
-	ata_do_eh(ap, adma_prereset, ata_std_softreset, NULL,
-		  ata_std_postreset);
-}
-
 static int adma_fill_sg(struct ata_queued_cmd *qc)
 {
 	struct scatterlist *sg;
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index 865030ae8f8a..676302fdaa97 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -912,16 +912,6 @@ err:
 	return -EIO;
 }
 
-static void sata_fsl_error_handler(struct ata_port *ap)
-{
-
-	DPRINTK("in xx_error_handler\n");
-
-	/* perform recovery */
-	ata_do_eh(ap, ata_std_prereset, sata_fsl_softreset, sata_std_hardreset,
-		  ata_std_postreset);
-}
-
 static void sata_fsl_post_internal_cmd(struct ata_queued_cmd *qc)
 {
 	if (qc->flags & ATA_QCFLAG_FAILED)
@@ -1213,7 +1203,7 @@ static const struct ata_port_operations sata_fsl_ops = {
 
 	.freeze = sata_fsl_freeze,
 	.thaw = sata_fsl_thaw,
-	.error_handler = sata_fsl_error_handler,
+	.softreset = sata_fsl_softreset,
 	.post_internal_cmd = sata_fsl_post_internal_cmd,
 
 	.port_start = sata_fsl_port_start,
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index 047f80f5825c..ba1c09953517 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -452,8 +452,7 @@ static void inic_error_handler(struct ata_port *ap)
 	spin_unlock_irqrestore(ap->lock, flags);
 
 	/* PIO and DMA engines have been stopped, perform recovery */
-	ata_do_eh(ap, ata_std_prereset, NULL, inic_hardreset,
-		  ata_std_postreset);
+	ata_std_error_handler(ap);
 }
 
 static void inic_post_internal_cmd(struct ata_queued_cmd *qc)
@@ -532,6 +531,8 @@ static struct ata_port_operations inic_port_ops = {
 
 	.freeze			= inic_freeze,
 	.thaw			= inic_thaw,
+	.softreset		= ATA_OP_NULL,	/* softreset is broken */
+	.hardreset		= inic_hardreset,
 	.error_handler		= inic_error_handler,
 	.post_internal_cmd	= inic_post_internal_cmd,
 	.dev_config		= inic_dev_config,
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index f341a82d27bf..9a89390531b1 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -470,7 +470,10 @@ static void mv_port_stop(struct ata_port *ap);
 static void mv_qc_prep(struct ata_queued_cmd *qc);
 static void mv_qc_prep_iie(struct ata_queued_cmd *qc);
 static unsigned int mv_qc_issue(struct ata_queued_cmd *qc);
-static void mv_error_handler(struct ata_port *ap);
+static int mv_prereset(struct ata_link *link, unsigned long deadline);
+static int mv_hardreset(struct ata_link *link, unsigned int *class,
+			unsigned long deadline);
+static void mv_postreset(struct ata_link *link, unsigned int *classes);
 static void mv_eh_freeze(struct ata_port *ap);
 static void mv_eh_thaw(struct ata_port *ap);
 static void mv6_dev_config(struct ata_device *dev);
@@ -534,7 +537,10 @@ static struct ata_port_operations mv5_ops = {
 
 	.freeze			= mv_eh_freeze,
 	.thaw			= mv_eh_thaw,
-	.error_handler		= mv_error_handler,
+	.prereset		= mv_prereset,
+	.hardreset		= mv_hardreset,
+	.postreset		= mv_postreset,
+	.error_handler		= ata_std_error_handler, /* avoid SFF EH */
 	.post_internal_cmd	= ATA_OP_NULL,
 
 	.scr_read		= mv5_scr_read,
@@ -2415,12 +2421,6 @@ static void mv_postreset(struct ata_link *link, unsigned int *classes)
 	iowrite8(ap->ctl, ap->ioaddr.ctl_addr);
 }
 
-static void mv_error_handler(struct ata_port *ap)
-{
-	ata_do_eh(ap, mv_prereset, ata_std_softreset,
-		  mv_hardreset, mv_postreset);
-}
-
 static void mv_eh_freeze(struct ata_port *ap)
 {
 	struct mv_host_priv *hpriv = ap->host->private_data;
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index 5637b082bc85..b2eb5724cf85 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -309,7 +309,8 @@ static void nv_nf2_freeze(struct ata_port *ap);
 static void nv_nf2_thaw(struct ata_port *ap);
 static void nv_ck804_freeze(struct ata_port *ap);
 static void nv_ck804_thaw(struct ata_port *ap);
-static void nv_error_handler(struct ata_port *ap);
+static int nv_hardreset(struct ata_link *link, unsigned int *class,
+			unsigned long deadline);
 static int nv_adma_slave_config(struct scsi_device *sdev);
 static int nv_adma_check_atapi_dma(struct ata_queued_cmd *qc);
 static void nv_adma_qc_prep(struct ata_queued_cmd *qc);
@@ -406,7 +407,7 @@ static struct scsi_host_template nv_swncq_sht = {
 
 static struct ata_port_operations nv_generic_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.error_handler		= nv_error_handler,
+	.hardreset		= nv_hardreset,
 	.scr_read		= nv_scr_read,
 	.scr_write		= nv_scr_write,
 };
@@ -1599,12 +1600,6 @@ static int nv_hardreset(struct ata_link *link, unsigned int *class,
 	return sata_std_hardreset(link, &dummy, deadline);
 }
 
-static void nv_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
-			   nv_hardreset, ata_std_postreset);
-}
-
 static void nv_adma_error_handler(struct ata_port *ap)
 {
 	struct nv_adma_port_priv *pp = ap->private_data;
@@ -1658,8 +1653,7 @@ static void nv_adma_error_handler(struct ata_port *ap)
 		readw(mmio + NV_ADMA_CTL);	/* flush posted write */
 	}
 
-	ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
-			   nv_hardreset, ata_std_postreset);
+	ata_bmdma_error_handler(ap);
 }
 
 static void nv_swncq_qc_to_dq(struct ata_port *ap, struct ata_queued_cmd *qc)
@@ -1785,8 +1779,7 @@ static void nv_swncq_error_handler(struct ata_port *ap)
 		ehc->i.action |= ATA_EH_RESET;
 	}
 
-	ata_bmdma_drive_eh(ap, ata_std_prereset, ata_std_softreset,
-			   nv_hardreset, ata_std_postreset);
+	ata_bmdma_error_handler(ap);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c
index e09b975c973d..91659dc15caf 100644
--- a/drivers/ata/sata_promise.c
+++ b/drivers/ata/sata_promise.c
@@ -148,8 +148,7 @@ static void pdc_freeze(struct ata_port *ap);
 static void pdc_sata_freeze(struct ata_port *ap);
 static void pdc_thaw(struct ata_port *ap);
 static void pdc_sata_thaw(struct ata_port *ap);
-static void pdc_pata_error_handler(struct ata_port *ap);
-static void pdc_sata_error_handler(struct ata_port *ap);
+static void pdc_error_handler(struct ata_port *ap);
 static void pdc_post_internal_cmd(struct ata_queued_cmd *qc);
 static int pdc_pata_cable_detect(struct ata_port *ap);
 static int pdc_sata_cable_detect(struct ata_port *ap);
@@ -171,6 +170,7 @@ static const struct ata_port_operations pdc_common_ops = {
 	.irq_clear		= pdc_irq_clear,
 
 	.post_internal_cmd	= pdc_post_internal_cmd,
+	.error_handler		= pdc_error_handler,
 };
 
 static struct ata_port_operations pdc_sata_ops = {
@@ -178,7 +178,6 @@ static struct ata_port_operations pdc_sata_ops = {
 	.cable_detect		= pdc_sata_cable_detect,
 	.freeze			= pdc_sata_freeze,
 	.thaw			= pdc_sata_thaw,
-	.error_handler		= pdc_sata_error_handler,
 	.scr_read		= pdc_sata_scr_read,
 	.scr_write		= pdc_sata_scr_write,
 	.port_start		= pdc_sata_port_start,
@@ -195,7 +194,6 @@ static struct ata_port_operations pdc_pata_ops = {
 	.cable_detect		= pdc_pata_cable_detect,
 	.freeze			= pdc_freeze,
 	.thaw			= pdc_thaw,
-	.error_handler		= pdc_pata_error_handler,
 	.port_start		= pdc_common_port_start,
 };
 
@@ -694,24 +692,12 @@ static void pdc_sata_thaw(struct ata_port *ap)
 	readl(host_mmio + hotplug_offset); /* flush */
 }
 
-static void pdc_common_error_handler(struct ata_port *ap, ata_reset_fn_t hardreset)
+static void pdc_error_handler(struct ata_port *ap)
 {
 	if (!(ap->pflags & ATA_PFLAG_FROZEN))
 		pdc_reset_port(ap);
 
-	/* perform recovery */
-	ata_do_eh(ap, ata_std_prereset, ata_std_softreset, hardreset,
-		  ata_std_postreset);
-}
-
-static void pdc_pata_error_handler(struct ata_port *ap)
-{
-	pdc_common_error_handler(ap, NULL);
-}
-
-static void pdc_sata_error_handler(struct ata_port *ap)
-{
-	pdc_common_error_handler(ap, sata_std_hardreset);
+	ata_std_error_handler(ap);
 }
 
 static void pdc_post_internal_cmd(struct ata_queued_cmd *qc)
diff --git a/drivers/ata/sata_qstor.c b/drivers/ata/sata_qstor.c
index 107ef09814de..2ceb0990bcd8 100644
--- a/drivers/ata/sata_qstor.c
+++ b/drivers/ata/sata_qstor.c
@@ -123,6 +123,7 @@ static void qs_bmdma_stop(struct ata_queued_cmd *qc);
 static u8 qs_bmdma_status(struct ata_port *ap);
 static void qs_freeze(struct ata_port *ap);
 static void qs_thaw(struct ata_port *ap);
+static int qs_prereset(struct ata_link *link, unsigned long deadline);
 static void qs_error_handler(struct ata_port *ap);
 
 static struct scsi_host_template qs_ata_sht = {
@@ -142,6 +143,8 @@ static struct ata_port_operations qs_ata_ops = {
 
 	.freeze			= qs_freeze,
 	.thaw			= qs_thaw,
+	.prereset		= qs_prereset,
+	.softreset		= ATA_OP_NULL,
 	.error_handler		= qs_error_handler,
 	.post_internal_cmd	= ATA_OP_NULL,
 
@@ -250,8 +253,7 @@ static int qs_scr_read(struct ata_port *ap, unsigned int sc_reg, u32 *val)
 static void qs_error_handler(struct ata_port *ap)
 {
 	qs_enter_reg_mode(ap);
-	ata_do_eh(ap, qs_prereset, NULL, sata_std_hardreset,
-		  ata_std_postreset);
+	ata_std_error_handler(ap);
 }
 
 static int qs_scr_write(struct ata_port *ap, unsigned int sc_reg, u32 val)
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index 363fb90e1047..67df1d753305 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -352,6 +352,14 @@ static void sil24_pmp_attach(struct ata_port *ap);
 static void sil24_pmp_detach(struct ata_port *ap);
 static void sil24_freeze(struct ata_port *ap);
 static void sil24_thaw(struct ata_port *ap);
+static int sil24_softreset(struct ata_link *link, unsigned int *class,
+			   unsigned long deadline);
+static int sil24_hardreset(struct ata_link *link, unsigned int *class,
+			   unsigned long deadline);
+static int sil24_pmp_softreset(struct ata_link *link, unsigned int *class,
+			       unsigned long deadline);
+static int sil24_pmp_hardreset(struct ata_link *link, unsigned int *class,
+			       unsigned long deadline);
 static void sil24_error_handler(struct ata_port *ap);
 static void sil24_post_internal_cmd(struct ata_queued_cmd *qc);
 static int sil24_port_start(struct ata_port *ap);
@@ -402,6 +410,10 @@ static struct ata_port_operations sil24_ops = {
 
 	.freeze			= sil24_freeze,
 	.thaw			= sil24_thaw,
+	.softreset		= sil24_softreset,
+	.hardreset		= sil24_hardreset,
+	.pmp_softreset		= sil24_pmp_softreset,
+	.pmp_hardreset		= sil24_pmp_hardreset,
 	.error_handler		= sil24_error_handler,
 	.post_internal_cmd	= sil24_post_internal_cmd,
 	.dev_config		= sil24_dev_config,
@@ -1181,11 +1193,7 @@ static void sil24_error_handler(struct ata_port *ap)
 	if (sil24_init_port(ap))
 		ata_eh_freeze_port(ap);
 
-	/* perform recovery */
-	sata_pmp_do_eh(ap, ata_std_prereset, sil24_softreset, sil24_hardreset,
-		       ata_std_postreset, sata_pmp_std_prereset,
-		       sil24_pmp_softreset, sil24_pmp_hardreset,
-		       sata_pmp_std_postreset);
+	sata_pmp_error_handler(ap);
 
 	pp->do_port_rst = 0;
 }
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index 6326bcf8ea5d..402fd7333d48 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -71,7 +71,7 @@ static int svia_init_one(struct pci_dev *pdev, const struct pci_device_id *ent);
 static int svia_scr_read(struct ata_port *ap, unsigned int sc_reg, u32 *val);
 static int svia_scr_write(struct ata_port *ap, unsigned int sc_reg, u32 val);
 static void svia_noop_freeze(struct ata_port *ap);
-static void vt6420_error_handler(struct ata_port *ap);
+static int vt6420_prereset(struct ata_link *link, unsigned long deadline);
 static int vt6421_pata_cable_detect(struct ata_port *ap);
 static void vt6421_set_pio_mode(struct ata_port *ap, struct ata_device *adev);
 static void vt6421_set_dma_mode(struct ata_port *ap, struct ata_device *adev);
@@ -106,7 +106,7 @@ static struct scsi_host_template svia_sht = {
 static struct ata_port_operations vt6420_sata_ops = {
 	.inherits		= &ata_bmdma_port_ops,
 	.freeze			= svia_noop_freeze,
-	.error_handler		= vt6420_error_handler,
+	.prereset		= vt6420_prereset,
 };
 
 static struct ata_port_operations vt6421_pata_ops = {
@@ -247,12 +247,6 @@ static int vt6420_prereset(struct ata_link *link, unsigned long deadline)
 	return 0;
 }
 
-static void vt6420_error_handler(struct ata_port *ap)
-{
-	ata_bmdma_drive_eh(ap, vt6420_prereset, ata_std_softreset, NULL,
-			   ata_std_postreset);
-}
-
 static int vt6421_pata_cable_detect(struct ata_port *ap)
 {
 	struct pci_dev *pdev = to_pci_dev(ap->host->dev);
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index c72014a3e7d4..79fd2436bd70 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -3937,7 +3937,7 @@ static int __ipr_eh_dev_reset(struct scsi_cmnd * scsi_cmd)
 	if (ipr_is_gata(res) && res->sata_port) {
 		ap = res->sata_port->ap;
 		spin_unlock_irq(scsi_cmd->device->host->host_lock);
-		ata_do_eh(ap, NULL, NULL, ipr_sata_reset, NULL);
+		ata_std_error_handler(ap);
 		spin_lock_irq(scsi_cmd->device->host->host_lock);
 
 		list_for_each_entry(ipr_cmd, &ioa_cfg->pending_q, queue) {
@@ -5275,6 +5275,7 @@ static struct ata_port_operations ipr_sata_ops = {
 	.check_altstatus = ipr_ata_check_altstatus,
 	.dev_select = ata_noop_dev_select,
 	.phy_reset = ipr_ata_phy_reset,
+	.hardreset = ipr_sata_reset,
 	.post_internal_cmd = ipr_ata_post_internal,
 	.tf_read = ipr_tf_read,
 	.qc_prep = ata_noop_qc_prep,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 88c6fa84ed74..01c233303aee 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -693,6 +693,14 @@ struct ata_port_operations {
 
 	void (*freeze)(struct ata_port *ap);
 	void (*thaw)(struct ata_port *ap);
+	ata_prereset_fn_t	prereset;
+	ata_reset_fn_t		softreset;
+	ata_reset_fn_t		hardreset;
+	ata_postreset_fn_t	postreset;
+	ata_prereset_fn_t	pmp_prereset;
+	ata_reset_fn_t		pmp_softreset;
+	ata_reset_fn_t		pmp_hardreset;
+	ata_postreset_fn_t	pmp_postreset;
 	void (*error_handler)(struct ata_port *ap);
 	void (*post_internal_cmd)(struct ata_queued_cmd *qc);
 
@@ -909,10 +917,6 @@ extern void ata_bmdma_irq_clear(struct ata_port *ap);
 extern void ata_noop_irq_clear(struct ata_port *ap);
 extern void ata_bmdma_freeze(struct ata_port *ap);
 extern void ata_bmdma_thaw(struct ata_port *ap);
-extern void ata_bmdma_drive_eh(struct ata_port *ap, ata_prereset_fn_t prereset,
-			       ata_reset_fn_t softreset,
-			       ata_reset_fn_t hardreset,
-			       ata_postreset_fn_t postreset);
 extern void ata_bmdma_error_handler(struct ata_port *ap);
 extern void ata_bmdma_post_internal_cmd(struct ata_queued_cmd *qc);
 extern int ata_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
@@ -1056,11 +1060,7 @@ extern int sata_pmp_std_prereset(struct ata_link *link, unsigned long deadline);
 extern int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
 				  unsigned long deadline);
 extern void sata_pmp_std_postreset(struct ata_link *link, unsigned int *class);
-extern void sata_pmp_do_eh(struct ata_port *ap,
-		ata_prereset_fn_t prereset, ata_reset_fn_t softreset,
-		ata_reset_fn_t hardreset, ata_postreset_fn_t postreset,
-		ata_prereset_fn_t pmp_prereset, ata_reset_fn_t pmp_softreset,
-		ata_reset_fn_t pmp_hardreset, ata_postreset_fn_t pmp_postreset);
+extern void sata_pmp_error_handler(struct ata_port *ap);
 
 /*
  * EH
@@ -1080,6 +1080,7 @@ extern void ata_eh_qc_retry(struct ata_queued_cmd *qc);
 extern void ata_do_eh(struct ata_port *ap, ata_prereset_fn_t prereset,
 		      ata_reset_fn_t softreset, ata_reset_fn_t hardreset,
 		      ata_postreset_fn_t postreset);
+extern void ata_std_error_handler(struct ata_port *ap);
 
 /*
  * Base operations to inherit from and initializers for sht
-- 
cgit v1.2.3


From 624d5c514eed18d5a93062e9d86d67065175f30a Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 22:16:41 +0900
Subject: libata: reorganize SFF related stuff

* Move SFF related functions from libata-core.c to libata-sff.c.

  ata_[bmdma_]sff_port_ops, ata_devchk(), ata_dev_try_classify(),
  ata_std_dev_select(), ata_tf_to_host(), ata_busy_sleep(),
  ata_wait_after_reset(), ata_wait_ready(), ata_bus_post_reset(),
  ata_bus_softreset(), ata_bus_reset(), ata_std_softreset(),
  sata_std_hardreset(), ata_fill_sg(), ata_fill_sg_dumb(),
  ata_qc_prep(), ata_dump_qc_prep(), ata_data_xfer(),
  ata_data_xfer_noirq(), ata_pio_sector(), ata_pio_sectors(),
  atapi_send_cdb(), __atapi_pio_bytes(), atapi_pio_bytes(),
  ata_hsm_ok_in_wq(), ata_hsm_qc_complete(), ata_hsm_move(),
  ata_pio_task(), ata_qc_issue_prot(), ata_host_intr(),
  ata_interrupt(), ata_std_ports()

* Make ata_pio_queue_task() global as it's now called from
  libata-sff.c.

* Move SFF related stuff in include/linux/libata.h and
  drivers/ata/libata.h into one place.  While at it, move timing
  constants into the global enum definition and fortify comments a
  bit.

This patch strictly moves stuff around and as such doesn't cause any
functional difference.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-core.c | 2114 +++-----------------------------------------
 drivers/ata/libata-sff.c  | 2151 +++++++++++++++++++++++++++++++++++++++++----
 drivers/ata/libata.h      |    8 +-
 include/linux/libata.h    |  379 ++++----
 4 files changed, 2321 insertions(+), 2331 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 7860d9f60ae4..34c068f18350 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -46,7 +46,6 @@
 #include <linux/init.h>
 #include <linux/list.h>
 #include <linux/mm.h>
-#include <linux/highmem.h>
 #include <linux/spinlock.h>
 #include <linux/blkdev.h>
 #include <linux/delay.h>
@@ -98,41 +97,6 @@ const struct ata_port_operations sata_pmp_port_ops = {
 	.error_handler		= sata_pmp_error_handler,
 };
 
-const struct ata_port_operations ata_sff_port_ops = {
-	.inherits		= &ata_base_port_ops,
-
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.softreset		= ata_std_softreset,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.dev_select		= ata_std_dev_select,
-	.check_status		= ata_check_status,
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.data_xfer		= ata_data_xfer,
-	.irq_on			= ata_irq_on,
-
-	.port_start		= ata_sff_port_start,
-};
-
-const struct ata_port_operations ata_bmdma_port_ops = {
-	.inherits		= &ata_sff_port_ops,
-
-	.mode_filter		= ata_pci_default_filter,
-
-	.bmdma_setup		= ata_bmdma_setup,
-	.bmdma_start		= ata_bmdma_start,
-	.bmdma_stop		= ata_bmdma_stop,
-	.bmdma_status		= ata_bmdma_status,
-	.irq_clear		= ata_bmdma_irq_clear,
-};
-
 static unsigned int ata_dev_init_params(struct ata_device *dev,
 					u16 heads, u16 sectors);
 static unsigned int ata_dev_set_xfermode(struct ata_device *dev);
@@ -422,6 +386,14 @@ int atapi_cmd_type(u8 opcode)
 	}
 }
 
+/**
+ *	ata_noop_irq_clear - Noop placeholder for irq_clear
+ *	@ap: Port associated with this ATA transaction.
+ */
+void ata_noop_irq_clear(struct ata_port *ap)
+{
+}
+
 /**
  *	ata_tf_to_fis - Convert ATA taskfile to SATA FIS structure
  *	@tf: Taskfile to convert
@@ -1102,50 +1074,6 @@ static void ata_lpm_disable(struct ata_host *host)
 }
 #endif	/* CONFIG_PM */
 
-
-/**
- *	ata_devchk - PATA device presence detection
- *	@ap: ATA channel to examine
- *	@device: Device to examine (starting at zero)
- *
- *	This technique was originally described in
- *	Hale Landis's ATADRVR (www.ata-atapi.com), and
- *	later found its way into the ATA/ATAPI spec.
- *
- *	Write a pattern to the ATA shadow registers,
- *	and if a device is present, it will respond by
- *	correctly storing and echoing back the
- *	ATA shadow register contents.
- *
- *	LOCKING:
- *	caller.
- */
-
-static unsigned int ata_devchk(struct ata_port *ap, unsigned int device)
-{
-	struct ata_ioports *ioaddr = &ap->ioaddr;
-	u8 nsect, lbal;
-
-	ap->ops->dev_select(ap, device);
-
-	iowrite8(0x55, ioaddr->nsect_addr);
-	iowrite8(0xaa, ioaddr->lbal_addr);
-
-	iowrite8(0xaa, ioaddr->nsect_addr);
-	iowrite8(0x55, ioaddr->lbal_addr);
-
-	iowrite8(0x55, ioaddr->nsect_addr);
-	iowrite8(0xaa, ioaddr->lbal_addr);
-
-	nsect = ioread8(ioaddr->nsect_addr);
-	lbal = ioread8(ioaddr->lbal_addr);
-
-	if ((nsect == 0x55) && (lbal == 0xaa))
-		return 1;	/* we found a device */
-
-	return 0;		/* nothing found */
-}
-
 /**
  *	ata_dev_classify - determine device type based on ATA-spec signature
  *	@tf: ATA taskfile register set for device to be identified
@@ -1205,75 +1133,6 @@ unsigned int ata_dev_classify(const struct ata_taskfile *tf)
 	return ATA_DEV_UNKNOWN;
 }
 
-/**
- *	ata_dev_try_classify - Parse returned ATA device signature
- *	@dev: ATA device to classify (starting at zero)
- *	@present: device seems present
- *	@r_err: Value of error register on completion
- *
- *	After an event -- SRST, E.D.D., or SATA COMRESET -- occurs,
- *	an ATA/ATAPI-defined set of values is placed in the ATA
- *	shadow registers, indicating the results of device detection
- *	and diagnostics.
- *
- *	Select the ATA device, and read the values from the ATA shadow
- *	registers.  Then parse according to the Error register value,
- *	and the spec-defined values examined by ata_dev_classify().
- *
- *	LOCKING:
- *	caller.
- *
- *	RETURNS:
- *	Device type - %ATA_DEV_ATA, %ATA_DEV_ATAPI or %ATA_DEV_NONE.
- */
-unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
-				  u8 *r_err)
-{
-	struct ata_port *ap = dev->link->ap;
-	struct ata_taskfile tf;
-	unsigned int class;
-	u8 err;
-
-	ap->ops->dev_select(ap, dev->devno);
-
-	memset(&tf, 0, sizeof(tf));
-
-	ap->ops->tf_read(ap, &tf);
-	err = tf.feature;
-	if (r_err)
-		*r_err = err;
-
-	/* see if device passed diags: continue and warn later */
-	if (err == 0)
-		/* diagnostic fail : do nothing _YET_ */
-		dev->horkage |= ATA_HORKAGE_DIAGNOSTIC;
-	else if (err == 1)
-		/* do nothing */ ;
-	else if ((dev->devno == 0) && (err == 0x81))
-		/* do nothing */ ;
-	else
-		return ATA_DEV_NONE;
-
-	/* determine if device is ATA or ATAPI */
-	class = ata_dev_classify(&tf);
-
-	if (class == ATA_DEV_UNKNOWN) {
-		/* If the device failed diagnostic, it's likely to
-		 * have reported incorrect device signature too.
-		 * Assume ATA device if the device seems present but
-		 * device signature is invalid with diagnostic
-		 * failure.
-		 */
-		if (present && (dev->horkage & ATA_HORKAGE_DIAGNOSTIC))
-			class = ATA_DEV_ATA;
-		else
-			class = ATA_DEV_NONE;
-	} else if ((class == ATA_DEV_ATA) && (ata_chk_status(ap) == 0))
-		class = ATA_DEV_NONE;
-
-	return class;
-}
-
 /**
  *	ata_id_string - Convert IDENTIFY DEVICE page into string
  *	@id: IDENTIFY DEVICE results we will examine
@@ -1597,73 +1456,6 @@ void ata_noop_dev_select(struct ata_port *ap, unsigned int device)
 {
 }
 
-
-/**
- *	ata_std_dev_select - Select device 0/1 on ATA bus
- *	@ap: ATA channel to manipulate
- *	@device: ATA device (numbered from zero) to select
- *
- *	Use the method defined in the ATA specification to
- *	make either device 0, or device 1, active on the
- *	ATA channel.  Works with both PIO and MMIO.
- *
- *	May be used as the dev_select() entry in ata_port_operations.
- *
- *	LOCKING:
- *	caller.
- */
-
-void ata_std_dev_select(struct ata_port *ap, unsigned int device)
-{
-	u8 tmp;
-
-	if (device == 0)
-		tmp = ATA_DEVICE_OBS;
-	else
-		tmp = ATA_DEVICE_OBS | ATA_DEV1;
-
-	iowrite8(tmp, ap->ioaddr.device_addr);
-	ata_pause(ap);		/* needed; also flushes, for mmio */
-}
-
-/**
- *	ata_dev_select - Select device 0/1 on ATA bus
- *	@ap: ATA channel to manipulate
- *	@device: ATA device (numbered from zero) to select
- *	@wait: non-zero to wait for Status register BSY bit to clear
- *	@can_sleep: non-zero if context allows sleeping
- *
- *	Use the method defined in the ATA specification to
- *	make either device 0, or device 1, active on the
- *	ATA channel.
- *
- *	This is a high-level version of ata_std_dev_select(),
- *	which additionally provides the services of inserting
- *	the proper pauses and status polling, where needed.
- *
- *	LOCKING:
- *	caller.
- */
-
-void ata_dev_select(struct ata_port *ap, unsigned int device,
-			   unsigned int wait, unsigned int can_sleep)
-{
-	if (ata_msg_probe(ap))
-		ata_port_printk(ap, KERN_INFO, "ata_dev_select: ENTER, "
-				"device %u, wait %u\n", device, wait);
-
-	if (wait)
-		ata_wait_idle(ap);
-
-	ap->ops->dev_select(ap, device);
-
-	if (wait) {
-		if (can_sleep && ap->link.device[device].class == ATA_DEV_ATAPI)
-			msleep(150);
-		ata_wait_idle(ap);
-	}
-}
-
 /**
  *	ata_dump_id - IDENTIFY DEVICE info debugging output
  *	@id: IDENTIFY DEVICE page to dump
@@ -1791,8 +1583,7 @@ unsigned long ata_id_xfermask(const u16 *id)
  *	LOCKING:
  *	Inherited from caller.
  */
-static void ata_pio_queue_task(struct ata_port *ap, void *data,
-			       unsigned long delay)
+void ata_pio_queue_task(struct ata_port *ap, void *data, unsigned long delay)
 {
 	ap->port_task_data = data;
 
@@ -3531,353 +3322,6 @@ int ata_do_set_mode(struct ata_link *link, struct ata_device **r_failed_dev)
 	return rc;
 }
 
-/**
- *	ata_tf_to_host - issue ATA taskfile to host controller
- *	@ap: port to which command is being issued
- *	@tf: ATA taskfile register set
- *
- *	Issues ATA taskfile register set to ATA host controller,
- *	with proper synchronization with interrupt handler and
- *	other threads.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- */
-
-static inline void ata_tf_to_host(struct ata_port *ap,
-				  const struct ata_taskfile *tf)
-{
-	ap->ops->tf_load(ap, tf);
-	ap->ops->exec_command(ap, tf);
-}
-
-/**
- *	ata_busy_sleep - sleep until BSY clears, or timeout
- *	@ap: port containing status register to be polled
- *	@tmout_pat: impatience timeout
- *	@tmout: overall timeout
- *
- *	Sleep until ATA Status register bit BSY clears,
- *	or a timeout occurs.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep).
- *
- *	RETURNS:
- *	0 on success, -errno otherwise.
- */
-int ata_busy_sleep(struct ata_port *ap,
-		   unsigned long tmout_pat, unsigned long tmout)
-{
-	unsigned long timer_start, timeout;
-	u8 status;
-
-	status = ata_busy_wait(ap, ATA_BUSY, 300);
-	timer_start = jiffies;
-	timeout = timer_start + tmout_pat;
-	while (status != 0xff && (status & ATA_BUSY) &&
-	       time_before(jiffies, timeout)) {
-		msleep(50);
-		status = ata_busy_wait(ap, ATA_BUSY, 3);
-	}
-
-	if (status != 0xff && (status & ATA_BUSY))
-		ata_port_printk(ap, KERN_WARNING,
-				"port is slow to respond, please be patient "
-				"(Status 0x%x)\n", status);
-
-	timeout = timer_start + tmout;
-	while (status != 0xff && (status & ATA_BUSY) &&
-	       time_before(jiffies, timeout)) {
-		msleep(50);
-		status = ata_chk_status(ap);
-	}
-
-	if (status == 0xff)
-		return -ENODEV;
-
-	if (status & ATA_BUSY) {
-		ata_port_printk(ap, KERN_ERR, "port failed to respond "
-				"(%lu secs, Status 0x%x)\n",
-				tmout / HZ, status);
-		return -EBUSY;
-	}
-
-	return 0;
-}
-
-/**
- *	ata_wait_after_reset - wait before checking status after reset
- *	@ap: port containing status register to be polled
- *	@deadline: deadline jiffies for the operation
- *
- *	After reset, we need to pause a while before reading status.
- *	Also, certain combination of controller and device report 0xff
- *	for some duration (e.g. until SATA PHY is up and running)
- *	which is interpreted as empty port in ATA world.  This
- *	function also waits for such devices to get out of 0xff
- *	status.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep).
- */
-void ata_wait_after_reset(struct ata_port *ap, unsigned long deadline)
-{
-	unsigned long until = jiffies + ATA_TMOUT_FF_WAIT;
-
-	if (time_before(until, deadline))
-		deadline = until;
-
-	/* Spec mandates ">= 2ms" before checking status.  We wait
-	 * 150ms, because that was the magic delay used for ATAPI
-	 * devices in Hale Landis's ATADRVR, for the period of time
-	 * between when the ATA command register is written, and then
-	 * status is checked.  Because waiting for "a while" before
-	 * checking status is fine, post SRST, we perform this magic
-	 * delay here as well.
-	 *
-	 * Old drivers/ide uses the 2mS rule and then waits for ready.
-	 */
-	msleep(150);
-
-	/* Wait for 0xff to clear.  Some SATA devices take a long time
-	 * to clear 0xff after reset.  For example, HHD424020F7SV00
-	 * iVDR needs >= 800ms while.  Quantum GoVault needs even more
-	 * than that.
-	 *
-	 * Note that some PATA controllers (pata_ali) explode if
-	 * status register is read more than once when there's no
-	 * device attached.
-	 */
-	if (ap->flags & ATA_FLAG_SATA) {
-		while (1) {
-			u8 status = ata_chk_status(ap);
-
-			if (status != 0xff || time_after(jiffies, deadline))
-				return;
-
-			msleep(50);
-		}
-	}
-}
-
-/**
- *	ata_wait_ready - sleep until BSY clears, or timeout
- *	@ap: port containing status register to be polled
- *	@deadline: deadline jiffies for the operation
- *
- *	Sleep until ATA Status register bit BSY clears, or timeout
- *	occurs.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep).
- *
- *	RETURNS:
- *	0 on success, -errno otherwise.
- */
-int ata_wait_ready(struct ata_port *ap, unsigned long deadline)
-{
-	unsigned long start = jiffies;
-	int warned = 0;
-
-	while (1) {
-		u8 status = ata_chk_status(ap);
-		unsigned long now = jiffies;
-
-		if (!(status & ATA_BUSY))
-			return 0;
-		if (!ata_link_online(&ap->link) && status == 0xff)
-			return -ENODEV;
-		if (time_after(now, deadline))
-			return -EBUSY;
-
-		if (!warned && time_after(now, start + 5 * HZ) &&
-		    (deadline - now > 3 * HZ)) {
-			ata_port_printk(ap, KERN_WARNING,
-				"port is slow to respond, please be patient "
-				"(Status 0x%x)\n", status);
-			warned = 1;
-		}
-
-		msleep(50);
-	}
-}
-
-static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
-			      unsigned long deadline)
-{
-	struct ata_ioports *ioaddr = &ap->ioaddr;
-	unsigned int dev0 = devmask & (1 << 0);
-	unsigned int dev1 = devmask & (1 << 1);
-	int rc, ret = 0;
-
-	/* if device 0 was found in ata_devchk, wait for its
-	 * BSY bit to clear
-	 */
-	if (dev0) {
-		rc = ata_wait_ready(ap, deadline);
-		if (rc) {
-			if (rc != -ENODEV)
-				return rc;
-			ret = rc;
-		}
-	}
-
-	/* if device 1 was found in ata_devchk, wait for register
-	 * access briefly, then wait for BSY to clear.
-	 */
-	if (dev1) {
-		int i;
-
-		ap->ops->dev_select(ap, 1);
-
-		/* Wait for register access.  Some ATAPI devices fail
-		 * to set nsect/lbal after reset, so don't waste too
-		 * much time on it.  We're gonna wait for !BSY anyway.
-		 */
-		for (i = 0; i < 2; i++) {
-			u8 nsect, lbal;
-
-			nsect = ioread8(ioaddr->nsect_addr);
-			lbal = ioread8(ioaddr->lbal_addr);
-			if ((nsect == 1) && (lbal == 1))
-				break;
-			msleep(50);	/* give drive a breather */
-		}
-
-		rc = ata_wait_ready(ap, deadline);
-		if (rc) {
-			if (rc != -ENODEV)
-				return rc;
-			ret = rc;
-		}
-	}
-
-	/* is all this really necessary? */
-	ap->ops->dev_select(ap, 0);
-	if (dev1)
-		ap->ops->dev_select(ap, 1);
-	if (dev0)
-		ap->ops->dev_select(ap, 0);
-
-	return ret;
-}
-
-static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
-			     unsigned long deadline)
-{
-	struct ata_ioports *ioaddr = &ap->ioaddr;
-
-	DPRINTK("ata%u: bus reset via SRST\n", ap->print_id);
-
-	/* software reset.  causes dev0 to be selected */
-	iowrite8(ap->ctl, ioaddr->ctl_addr);
-	udelay(20);	/* FIXME: flush */
-	iowrite8(ap->ctl | ATA_SRST, ioaddr->ctl_addr);
-	udelay(20);	/* FIXME: flush */
-	iowrite8(ap->ctl, ioaddr->ctl_addr);
-
-	/* wait a while before checking status */
-	ata_wait_after_reset(ap, deadline);
-
-	/* Before we perform post reset processing we want to see if
-	 * the bus shows 0xFF because the odd clown forgets the D7
-	 * pulldown resistor.
-	 */
-	if (ata_chk_status(ap) == 0xFF)
-		return -ENODEV;
-
-	return ata_bus_post_reset(ap, devmask, deadline);
-}
-
-/**
- *	ata_bus_reset - reset host port and associated ATA channel
- *	@ap: port to reset
- *
- *	This is typically the first time we actually start issuing
- *	commands to the ATA channel.  We wait for BSY to clear, then
- *	issue EXECUTE DEVICE DIAGNOSTIC command, polling for its
- *	result.  Determine what devices, if any, are on the channel
- *	by looking at the device 0/1 error register.  Look at the signature
- *	stored in each device's taskfile registers, to determine if
- *	the device is ATA or ATAPI.
- *
- *	LOCKING:
- *	PCI/etc. bus probe sem.
- *	Obtains host lock.
- *
- *	SIDE EFFECTS:
- *	Sets ATA_FLAG_DISABLED if bus reset fails.
- */
-
-void ata_bus_reset(struct ata_port *ap)
-{
-	struct ata_device *device = ap->link.device;
-	struct ata_ioports *ioaddr = &ap->ioaddr;
-	unsigned int slave_possible = ap->flags & ATA_FLAG_SLAVE_POSS;
-	u8 err;
-	unsigned int dev0, dev1 = 0, devmask = 0;
-	int rc;
-
-	DPRINTK("ENTER, host %u, port %u\n", ap->print_id, ap->port_no);
-
-	/* determine if device 0/1 are present */
-	if (ap->flags & ATA_FLAG_SATA_RESET)
-		dev0 = 1;
-	else {
-		dev0 = ata_devchk(ap, 0);
-		if (slave_possible)
-			dev1 = ata_devchk(ap, 1);
-	}
-
-	if (dev0)
-		devmask |= (1 << 0);
-	if (dev1)
-		devmask |= (1 << 1);
-
-	/* select device 0 again */
-	ap->ops->dev_select(ap, 0);
-
-	/* issue bus reset */
-	if (ap->flags & ATA_FLAG_SRST) {
-		rc = ata_bus_softreset(ap, devmask, jiffies + 40 * HZ);
-		if (rc && rc != -ENODEV)
-			goto err_out;
-	}
-
-	/*
-	 * determine by signature whether we have ATA or ATAPI devices
-	 */
-	device[0].class = ata_dev_try_classify(&device[0], dev0, &err);
-	if ((slave_possible) && (err != 0x81))
-		device[1].class = ata_dev_try_classify(&device[1], dev1, &err);
-
-	/* is double-select really necessary? */
-	if (device[1].class != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 1);
-	if (device[0].class != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 0);
-
-	/* if no devices were detected, disable this port */
-	if ((device[0].class == ATA_DEV_NONE) &&
-	    (device[1].class == ATA_DEV_NONE))
-		goto err_out;
-
-	if (ap->flags & (ATA_FLAG_SATA_RESET | ATA_FLAG_SRST)) {
-		/* set up device control for ATA_FLAG_SATA_RESET */
-		iowrite8(ap->ctl, ioaddr->ctl_addr);
-	}
-
-	DPRINTK("EXIT\n");
-	return;
-
-err_out:
-	ata_port_printk(ap, KERN_ERR, "disabling port\n");
-	ata_port_disable(ap);
-
-	DPRINTK("EXIT\n");
-}
-
 /**
  *	sata_link_debounce - debounce SATA phy status
  *	@link: ATA link to debounce SATA phy status for
@@ -4034,12 +3478,12 @@ int ata_std_prereset(struct ata_link *link, unsigned long deadline)
 }
 
 /**
- *	ata_std_softreset - reset host port via ATA SRST
- *	@link: ATA link to reset
- *	@classes: resulting classes of attached devices
+ *	sata_link_hardreset - reset link via SATA phy reset
+ *	@link: link to reset
+ *	@timing: timing parameters { interval, duratinon, timeout } in msec
  *	@deadline: deadline jiffies for the operation
  *
- *	Reset host port using ATA SRST.
+ *	SATA phy-reset @link using DET bits of SControl register.
  *
  *	LOCKING:
  *	Kernel thread context (may sleep)
@@ -4047,70 +3491,10 @@ int ata_std_prereset(struct ata_link *link, unsigned long deadline)
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_std_softreset(struct ata_link *link, unsigned int *classes,
-		      unsigned long deadline)
+int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
+			unsigned long deadline)
 {
-	struct ata_port *ap = link->ap;
-	unsigned int slave_possible = ap->flags & ATA_FLAG_SLAVE_POSS;
-	unsigned int devmask = 0;
-	int rc;
-	u8 err;
-
-	DPRINTK("ENTER\n");
-
-	if (ata_link_offline(link)) {
-		classes[0] = ATA_DEV_NONE;
-		goto out;
-	}
-
-	/* determine if device 0/1 are present */
-	if (ata_devchk(ap, 0))
-		devmask |= (1 << 0);
-	if (slave_possible && ata_devchk(ap, 1))
-		devmask |= (1 << 1);
-
-	/* select device 0 again */
-	ap->ops->dev_select(ap, 0);
-
-	/* issue bus reset */
-	DPRINTK("about to softreset, devmask=%x\n", devmask);
-	rc = ata_bus_softreset(ap, devmask, deadline);
-	/* if link is occupied, -ENODEV too is an error */
-	if (rc && (rc != -ENODEV || sata_scr_valid(link))) {
-		ata_link_printk(link, KERN_ERR, "SRST failed (errno=%d)\n", rc);
-		return rc;
-	}
-
-	/* determine by signature whether we have ATA or ATAPI devices */
-	classes[0] = ata_dev_try_classify(&link->device[0],
-					  devmask & (1 << 0), &err);
-	if (slave_possible && err != 0x81)
-		classes[1] = ata_dev_try_classify(&link->device[1],
-						  devmask & (1 << 1), &err);
-
- out:
-	DPRINTK("EXIT, classes[0]=%u [1]=%u\n", classes[0], classes[1]);
-	return 0;
-}
-
-/**
- *	sata_link_hardreset - reset link via SATA phy reset
- *	@link: link to reset
- *	@timing: timing parameters { interval, duratinon, timeout } in msec
- *	@deadline: deadline jiffies for the operation
- *
- *	SATA phy-reset @link using DET bits of SControl register.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep)
- *
- *	RETURNS:
- *	0 on success, -errno otherwise.
- */
-int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
-			unsigned long deadline)
-{
-	u32 scontrol;
+	u32 scontrol;
 	int rc;
 
 	DPRINTK("ENTER\n");
@@ -4153,74 +3537,6 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
 	return rc;
 }
 
-/**
- *	sata_std_hardreset - reset host port via SATA phy reset
- *	@link: link to reset
- *	@class: resulting class of attached device
- *	@deadline: deadline jiffies for the operation
- *
- *	SATA phy-reset host port using DET bits of SControl register,
- *	wait for !BSY and classify the attached device.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep)
- *
- *	RETURNS:
- *	0 on success, -errno otherwise.
- */
-int sata_std_hardreset(struct ata_link *link, unsigned int *class,
-		       unsigned long deadline)
-{
-	struct ata_port *ap = link->ap;
-	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
-	int rc;
-
-	DPRINTK("ENTER\n");
-
-	/* do hardreset */
-	rc = sata_link_hardreset(link, timing, deadline);
-	if (rc) {
-		ata_link_printk(link, KERN_ERR,
-				"COMRESET failed (errno=%d)\n", rc);
-		return rc;
-	}
-
-	/* TODO: phy layer with polling, timeouts, etc. */
-	if (ata_link_offline(link)) {
-		*class = ATA_DEV_NONE;
-		DPRINTK("EXIT, link offline\n");
-		return 0;
-	}
-
-	/* wait a while before checking status */
-	ata_wait_after_reset(ap, deadline);
-
-	/* If PMP is supported, we have to do follow-up SRST.  Note
-	 * that some PMPs don't send D2H Reg FIS after hardreset at
-	 * all if the first port is empty.  Wait for it just for a
-	 * second and request follow-up SRST.
-	 */
-	if (ap->flags & ATA_FLAG_PMP) {
-		ata_wait_ready(ap, jiffies + HZ);
-		return -EAGAIN;
-	}
-
-	rc = ata_wait_ready(ap, deadline);
-	/* link occupied, -ENODEV too is an error */
-	if (rc) {
-		ata_link_printk(link, KERN_ERR,
-				"COMRESET failed (errno=%d)\n", rc);
-		return rc;
-	}
-
-	ap->ops->dev_select(ap, 0);	/* probably unnecessary */
-
-	*class = ata_dev_try_classify(link->device, 1, NULL);
-
-	DPRINTK("EXIT, class=%u\n", *class);
-	return 0;
-}
-
 /**
  *	ata_std_postreset - standard postreset callback
  *	@link: the target ata_link
@@ -4803,112 +4119,6 @@ void ata_sg_clean(struct ata_queued_cmd *qc)
 	qc->sg = NULL;
 }
 
-/**
- *	ata_fill_sg - Fill PCI IDE PRD table
- *	@qc: Metadata associated with taskfile to be transferred
- *
- *	Fill PCI IDE PRD (scatter-gather) table with segments
- *	associated with the current disk command.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- *
- */
-static void ata_fill_sg(struct ata_queued_cmd *qc)
-{
-	struct ata_port *ap = qc->ap;
-	struct scatterlist *sg;
-	unsigned int si, pi;
-
-	pi = 0;
-	for_each_sg(qc->sg, sg, qc->n_elem, si) {
-		u32 addr, offset;
-		u32 sg_len, len;
-
-		/* determine if physical DMA addr spans 64K boundary.
-		 * Note h/w doesn't support 64-bit, so we unconditionally
-		 * truncate dma_addr_t to u32.
-		 */
-		addr = (u32) sg_dma_address(sg);
-		sg_len = sg_dma_len(sg);
-
-		while (sg_len) {
-			offset = addr & 0xffff;
-			len = sg_len;
-			if ((offset + sg_len) > 0x10000)
-				len = 0x10000 - offset;
-
-			ap->prd[pi].addr = cpu_to_le32(addr);
-			ap->prd[pi].flags_len = cpu_to_le32(len & 0xffff);
-			VPRINTK("PRD[%u] = (0x%X, 0x%X)\n", pi, addr, len);
-
-			pi++;
-			sg_len -= len;
-			addr += len;
-		}
-	}
-
-	ap->prd[pi - 1].flags_len |= cpu_to_le32(ATA_PRD_EOT);
-}
-
-/**
- *	ata_fill_sg_dumb - Fill PCI IDE PRD table
- *	@qc: Metadata associated with taskfile to be transferred
- *
- *	Fill PCI IDE PRD (scatter-gather) table with segments
- *	associated with the current disk command. Perform the fill
- *	so that we avoid writing any length 64K records for
- *	controllers that don't follow the spec.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- *
- */
-static void ata_fill_sg_dumb(struct ata_queued_cmd *qc)
-{
-	struct ata_port *ap = qc->ap;
-	struct scatterlist *sg;
-	unsigned int si, pi;
-
-	pi = 0;
-	for_each_sg(qc->sg, sg, qc->n_elem, si) {
-		u32 addr, offset;
-		u32 sg_len, len, blen;
-
-		/* determine if physical DMA addr spans 64K boundary.
-		 * Note h/w doesn't support 64-bit, so we unconditionally
-		 * truncate dma_addr_t to u32.
-		 */
-		addr = (u32) sg_dma_address(sg);
-		sg_len = sg_dma_len(sg);
-
-		while (sg_len) {
-			offset = addr & 0xffff;
-			len = sg_len;
-			if ((offset + sg_len) > 0x10000)
-				len = 0x10000 - offset;
-
-			blen = len & 0xffff;
-			ap->prd[pi].addr = cpu_to_le32(addr);
-			if (blen == 0) {
-			   /* Some PATA chipsets like the CS5530 can't
-			      cope with 0x0000 meaning 64K as the spec says */
-				ap->prd[pi].flags_len = cpu_to_le32(0x8000);
-				blen = 0x8000;
-				ap->prd[++pi].addr = cpu_to_le32(addr + 0x8000);
-			}
-			ap->prd[pi].flags_len = cpu_to_le32(blen);
-			VPRINTK("PRD[%u] = (0x%X, 0x%X)\n", pi, addr, len);
-
-			pi++;
-			sg_len -= len;
-			addr += len;
-		}
-	}
-
-	ap->prd[pi - 1].flags_len |= cpu_to_le32(ATA_PRD_EOT);
-}
-
 /**
  *	ata_check_atapi_dma - Check whether ATAPI DMA can be supported
  *	@qc: Metadata associated with taskfile to check
@@ -4918,858 +4128,132 @@ static void ata_fill_sg_dumb(struct ata_queued_cmd *qc)
  *	supplied PACKET command.
  *
  *	LOCKING:
- *	spin_lock_irqsave(host lock)
- *
- *	RETURNS: 0 when ATAPI DMA can be used
- *               nonzero otherwise
- */
-int ata_check_atapi_dma(struct ata_queued_cmd *qc)
-{
-	struct ata_port *ap = qc->ap;
-
-	/* Don't allow DMA if it isn't multiple of 16 bytes.  Quite a
-	 * few ATAPI devices choke on such DMA requests.
-	 */
-	if (unlikely(qc->nbytes & 15))
-		return 1;
-
-	if (ap->ops->check_atapi_dma)
-		return ap->ops->check_atapi_dma(qc);
-
-	return 0;
-}
-
-/**
- *	ata_std_qc_defer - Check whether a qc needs to be deferred
- *	@qc: ATA command in question
- *
- *	Non-NCQ commands cannot run with any other command, NCQ or
- *	not.  As upper layer only knows the queue depth, we are
- *	responsible for maintaining exclusion.  This function checks
- *	whether a new command @qc can be issued.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- *
- *	RETURNS:
- *	ATA_DEFER_* if deferring is needed, 0 otherwise.
- */
-int ata_std_qc_defer(struct ata_queued_cmd *qc)
-{
-	struct ata_link *link = qc->dev->link;
-
-	if (qc->tf.protocol == ATA_PROT_NCQ) {
-		if (!ata_tag_valid(link->active_tag))
-			return 0;
-	} else {
-		if (!ata_tag_valid(link->active_tag) && !link->sactive)
-			return 0;
-	}
-
-	return ATA_DEFER_LINK;
-}
-
-/**
- *	ata_qc_prep - Prepare taskfile for submission
- *	@qc: Metadata associated with taskfile to be prepared
- *
- *	Prepare ATA taskfile for submission.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- */
-void ata_qc_prep(struct ata_queued_cmd *qc)
-{
-	if (!(qc->flags & ATA_QCFLAG_DMAMAP))
-		return;
-
-	ata_fill_sg(qc);
-}
-
-/**
- *	ata_dumb_qc_prep - Prepare taskfile for submission
- *	@qc: Metadata associated with taskfile to be prepared
- *
- *	Prepare ATA taskfile for submission.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- */
-void ata_dumb_qc_prep(struct ata_queued_cmd *qc)
-{
-	if (!(qc->flags & ATA_QCFLAG_DMAMAP))
-		return;
-
-	ata_fill_sg_dumb(qc);
-}
-
-void ata_noop_qc_prep(struct ata_queued_cmd *qc) { }
-
-/**
- *	ata_sg_init - Associate command with scatter-gather table.
- *	@qc: Command to be associated
- *	@sg: Scatter-gather table.
- *	@n_elem: Number of elements in s/g table.
- *
- *	Initialize the data-related elements of queued_cmd @qc
- *	to point to a scatter-gather table @sg, containing @n_elem
- *	elements.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- */
-void ata_sg_init(struct ata_queued_cmd *qc, struct scatterlist *sg,
-		 unsigned int n_elem)
-{
-	qc->sg = sg;
-	qc->n_elem = n_elem;
-	qc->cursg = qc->sg;
-}
-
-/**
- *	ata_sg_setup - DMA-map the scatter-gather table associated with a command.
- *	@qc: Command with scatter-gather table to be mapped.
- *
- *	DMA-map the scatter-gather table associated with queued_cmd @qc.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- *
- *	RETURNS:
- *	Zero on success, negative on error.
- *
- */
-static int ata_sg_setup(struct ata_queued_cmd *qc)
-{
-	struct ata_port *ap = qc->ap;
-	unsigned int n_elem;
-
-	VPRINTK("ENTER, ata%u\n", ap->print_id);
-
-	n_elem = dma_map_sg(ap->dev, qc->sg, qc->n_elem, qc->dma_dir);
-	if (n_elem < 1)
-		return -1;
-
-	DPRINTK("%d sg elements mapped\n", n_elem);
-
-	qc->n_elem = n_elem;
-	qc->flags |= ATA_QCFLAG_DMAMAP;
-
-	return 0;
-}
-
-/**
- *	swap_buf_le16 - swap halves of 16-bit words in place
- *	@buf:  Buffer to swap
- *	@buf_words:  Number of 16-bit words in buffer.
- *
- *	Swap halves of 16-bit words if needed to convert from
- *	little-endian byte order to native cpu byte order, or
- *	vice-versa.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-void swap_buf_le16(u16 *buf, unsigned int buf_words)
-{
-#ifdef __BIG_ENDIAN
-	unsigned int i;
-
-	for (i = 0; i < buf_words; i++)
-		buf[i] = le16_to_cpu(buf[i]);
-#endif /* __BIG_ENDIAN */
-}
-
-/**
- *	ata_data_xfer - Transfer data by PIO
- *	@dev: device to target
- *	@buf: data buffer
- *	@buflen: buffer length
- *	@rw: read/write
- *
- *	Transfer data from/to the device data register by PIO.
- *
- *	LOCKING:
- *	Inherited from caller.
- *
- *	RETURNS:
- *	Bytes consumed.
- */
-unsigned int ata_data_xfer(struct ata_device *dev, unsigned char *buf,
-			   unsigned int buflen, int rw)
-{
-	struct ata_port *ap = dev->link->ap;
-	void __iomem *data_addr = ap->ioaddr.data_addr;
-	unsigned int words = buflen >> 1;
-
-	/* Transfer multiple of 2 bytes */
-	if (rw == READ)
-		ioread16_rep(data_addr, buf, words);
-	else
-		iowrite16_rep(data_addr, buf, words);
-
-	/* Transfer trailing 1 byte, if any. */
-	if (unlikely(buflen & 0x01)) {
-		__le16 align_buf[1] = { 0 };
-		unsigned char *trailing_buf = buf + buflen - 1;
-
-		if (rw == READ) {
-			align_buf[0] = cpu_to_le16(ioread16(data_addr));
-			memcpy(trailing_buf, align_buf, 1);
-		} else {
-			memcpy(align_buf, trailing_buf, 1);
-			iowrite16(le16_to_cpu(align_buf[0]), data_addr);
-		}
-		words++;
-	}
-
-	return words << 1;
-}
-
-/**
- *	ata_data_xfer_noirq - Transfer data by PIO
- *	@dev: device to target
- *	@buf: data buffer
- *	@buflen: buffer length
- *	@rw: read/write
- *
- *	Transfer data from/to the device data register by PIO. Do the
- *	transfer with interrupts disabled.
- *
- *	LOCKING:
- *	Inherited from caller.
- *
- *	RETURNS:
- *	Bytes consumed.
- */
-unsigned int ata_data_xfer_noirq(struct ata_device *dev, unsigned char *buf,
-				 unsigned int buflen, int rw)
-{
-	unsigned long flags;
-	unsigned int consumed;
-
-	local_irq_save(flags);
-	consumed = ata_data_xfer(dev, buf, buflen, rw);
-	local_irq_restore(flags);
-
-	return consumed;
-}
-
-
-/**
- *	ata_pio_sector - Transfer a sector of data.
- *	@qc: Command on going
- *
- *	Transfer qc->sect_size bytes of data from/to the ATA device.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-
-static void ata_pio_sector(struct ata_queued_cmd *qc)
-{
-	int do_write = (qc->tf.flags & ATA_TFLAG_WRITE);
-	struct ata_port *ap = qc->ap;
-	struct page *page;
-	unsigned int offset;
-	unsigned char *buf;
-
-	if (qc->curbytes == qc->nbytes - qc->sect_size)
-		ap->hsm_task_state = HSM_ST_LAST;
-
-	page = sg_page(qc->cursg);
-	offset = qc->cursg->offset + qc->cursg_ofs;
-
-	/* get the current page and offset */
-	page = nth_page(page, (offset >> PAGE_SHIFT));
-	offset %= PAGE_SIZE;
-
-	DPRINTK("data %s\n", qc->tf.flags & ATA_TFLAG_WRITE ? "write" : "read");
-
-	if (PageHighMem(page)) {
-		unsigned long flags;
-
-		/* FIXME: use a bounce buffer */
-		local_irq_save(flags);
-		buf = kmap_atomic(page, KM_IRQ0);
-
-		/* do the actual data transfer */
-		ap->ops->data_xfer(qc->dev, buf + offset, qc->sect_size, do_write);
-
-		kunmap_atomic(buf, KM_IRQ0);
-		local_irq_restore(flags);
-	} else {
-		buf = page_address(page);
-		ap->ops->data_xfer(qc->dev, buf + offset, qc->sect_size, do_write);
-	}
-
-	qc->curbytes += qc->sect_size;
-	qc->cursg_ofs += qc->sect_size;
-
-	if (qc->cursg_ofs == qc->cursg->length) {
-		qc->cursg = sg_next(qc->cursg);
-		qc->cursg_ofs = 0;
-	}
-}
-
-/**
- *	ata_pio_sectors - Transfer one or many sectors.
- *	@qc: Command on going
- *
- *	Transfer one or many sectors of data from/to the
- *	ATA device for the DRQ request.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-
-static void ata_pio_sectors(struct ata_queued_cmd *qc)
-{
-	if (is_multi_taskfile(&qc->tf)) {
-		/* READ/WRITE MULTIPLE */
-		unsigned int nsect;
-
-		WARN_ON(qc->dev->multi_count == 0);
-
-		nsect = min((qc->nbytes - qc->curbytes) / qc->sect_size,
-			    qc->dev->multi_count);
-		while (nsect--)
-			ata_pio_sector(qc);
-	} else
-		ata_pio_sector(qc);
-
-	ata_altstatus(qc->ap); /* flush */
-}
-
-/**
- *	atapi_send_cdb - Write CDB bytes to hardware
- *	@ap: Port to which ATAPI device is attached.
- *	@qc: Taskfile currently active
- *
- *	When device has indicated its readiness to accept
- *	a CDB, this function is called.  Send the CDB.
- *
- *	LOCKING:
- *	caller.
- */
-
-static void atapi_send_cdb(struct ata_port *ap, struct ata_queued_cmd *qc)
-{
-	/* send SCSI cdb */
-	DPRINTK("send cdb\n");
-	WARN_ON(qc->dev->cdb_len < 12);
-
-	ap->ops->data_xfer(qc->dev, qc->cdb, qc->dev->cdb_len, 1);
-	ata_altstatus(ap); /* flush */
-
-	switch (qc->tf.protocol) {
-	case ATAPI_PROT_PIO:
-		ap->hsm_task_state = HSM_ST;
-		break;
-	case ATAPI_PROT_NODATA:
-		ap->hsm_task_state = HSM_ST_LAST;
-		break;
-	case ATAPI_PROT_DMA:
-		ap->hsm_task_state = HSM_ST_LAST;
-		/* initiate bmdma */
-		ap->ops->bmdma_start(qc);
-		break;
-	}
-}
-
-/**
- *	__atapi_pio_bytes - Transfer data from/to the ATAPI device.
- *	@qc: Command on going
- *	@bytes: number of bytes
- *
- *	Transfer Transfer data from/to the ATAPI device.
- *
- *	LOCKING:
- *	Inherited from caller.
- *
- */
-static int __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes)
-{
-	int rw = (qc->tf.flags & ATA_TFLAG_WRITE) ? WRITE : READ;
-	struct ata_port *ap = qc->ap;
-	struct ata_device *dev = qc->dev;
-	struct ata_eh_info *ehi = &dev->link->eh_info;
-	struct scatterlist *sg;
-	struct page *page;
-	unsigned char *buf;
-	unsigned int offset, count, consumed;
-
-next_sg:
-	sg = qc->cursg;
-	if (unlikely(!sg)) {
-		ata_ehi_push_desc(ehi, "unexpected or too much trailing data "
-				  "buf=%u cur=%u bytes=%u",
-				  qc->nbytes, qc->curbytes, bytes);
-		return -1;
-	}
-
-	page = sg_page(sg);
-	offset = sg->offset + qc->cursg_ofs;
-
-	/* get the current page and offset */
-	page = nth_page(page, (offset >> PAGE_SHIFT));
-	offset %= PAGE_SIZE;
-
-	/* don't overrun current sg */
-	count = min(sg->length - qc->cursg_ofs, bytes);
-
-	/* don't cross page boundaries */
-	count = min(count, (unsigned int)PAGE_SIZE - offset);
-
-	DPRINTK("data %s\n", qc->tf.flags & ATA_TFLAG_WRITE ? "write" : "read");
-
-	if (PageHighMem(page)) {
-		unsigned long flags;
-
-		/* FIXME: use bounce buffer */
-		local_irq_save(flags);
-		buf = kmap_atomic(page, KM_IRQ0);
-
-		/* do the actual data transfer */
-		consumed = ap->ops->data_xfer(dev,  buf + offset, count, rw);
-
-		kunmap_atomic(buf, KM_IRQ0);
-		local_irq_restore(flags);
-	} else {
-		buf = page_address(page);
-		consumed = ap->ops->data_xfer(dev,  buf + offset, count, rw);
-	}
-
-	bytes -= min(bytes, consumed);
-	qc->curbytes += count;
-	qc->cursg_ofs += count;
-
-	if (qc->cursg_ofs == sg->length) {
-		qc->cursg = sg_next(qc->cursg);
-		qc->cursg_ofs = 0;
-	}
-
-	/* consumed can be larger than count only for the last transfer */
-	WARN_ON(qc->cursg && count != consumed);
-
-	if (bytes)
-		goto next_sg;
-	return 0;
-}
-
-/**
- *	atapi_pio_bytes - Transfer data from/to the ATAPI device.
- *	@qc: Command on going
- *
- *	Transfer Transfer data from/to the ATAPI device.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-
-static void atapi_pio_bytes(struct ata_queued_cmd *qc)
-{
-	struct ata_port *ap = qc->ap;
-	struct ata_device *dev = qc->dev;
-	struct ata_eh_info *ehi = &dev->link->eh_info;
-	unsigned int ireason, bc_lo, bc_hi, bytes;
-	int i_write, do_write = (qc->tf.flags & ATA_TFLAG_WRITE) ? 1 : 0;
-
-	/* Abuse qc->result_tf for temp storage of intermediate TF
-	 * here to save some kernel stack usage.
-	 * For normal completion, qc->result_tf is not relevant. For
-	 * error, qc->result_tf is later overwritten by ata_qc_complete().
-	 * So, the correctness of qc->result_tf is not affected.
-	 */
-	ap->ops->tf_read(ap, &qc->result_tf);
-	ireason = qc->result_tf.nsect;
-	bc_lo = qc->result_tf.lbam;
-	bc_hi = qc->result_tf.lbah;
-	bytes = (bc_hi << 8) | bc_lo;
-
-	/* shall be cleared to zero, indicating xfer of data */
-	if (unlikely(ireason & (1 << 0)))
-		goto atapi_check;
-
-	/* make sure transfer direction matches expected */
-	i_write = ((ireason & (1 << 1)) == 0) ? 1 : 0;
-	if (unlikely(do_write != i_write))
-		goto atapi_check;
-
-	if (unlikely(!bytes))
-		goto atapi_check;
-
-	VPRINTK("ata%u: xfering %d bytes\n", ap->print_id, bytes);
-
-	if (unlikely(__atapi_pio_bytes(qc, bytes)))
-		goto err_out;
-	ata_altstatus(ap); /* flush */
-
-	return;
-
- atapi_check:
-	ata_ehi_push_desc(ehi, "ATAPI check failed (ireason=0x%x bytes=%u)",
-			  ireason, bytes);
- err_out:
-	qc->err_mask |= AC_ERR_HSM;
-	ap->hsm_task_state = HSM_ST_ERR;
-}
-
-/**
- *	ata_hsm_ok_in_wq - Check if the qc can be handled in the workqueue.
- *	@ap: the target ata_port
- *	@qc: qc on going
- *
- *	RETURNS:
- *	1 if ok in workqueue, 0 otherwise.
- */
-
-static inline int ata_hsm_ok_in_wq(struct ata_port *ap, struct ata_queued_cmd *qc)
-{
-	if (qc->tf.flags & ATA_TFLAG_POLLING)
-		return 1;
-
-	if (ap->hsm_task_state == HSM_ST_FIRST) {
-		if (qc->tf.protocol == ATA_PROT_PIO &&
-		    (qc->tf.flags & ATA_TFLAG_WRITE))
-		    return 1;
-
-		if (ata_is_atapi(qc->tf.protocol) &&
-		    !(qc->dev->flags & ATA_DFLAG_CDB_INTR))
-			return 1;
-	}
-
-	return 0;
-}
-
-/**
- *	ata_hsm_qc_complete - finish a qc running on standard HSM
- *	@qc: Command to complete
- *	@in_wq: 1 if called from workqueue, 0 otherwise
- *
- *	Finish @qc which is running on standard HSM.
- *
- *	LOCKING:
- *	If @in_wq is zero, spin_lock_irqsave(host lock).
- *	Otherwise, none on entry and grabs host lock.
- */
-static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
-{
-	struct ata_port *ap = qc->ap;
-	unsigned long flags;
-
-	if (ap->ops->error_handler) {
-		if (in_wq) {
-			spin_lock_irqsave(ap->lock, flags);
-
-			/* EH might have kicked in while host lock is
-			 * released.
-			 */
-			qc = ata_qc_from_tag(ap, qc->tag);
-			if (qc) {
-				if (likely(!(qc->err_mask & AC_ERR_HSM))) {
-					ap->ops->irq_on(ap);
-					ata_qc_complete(qc);
-				} else
-					ata_port_freeze(ap);
-			}
-
-			spin_unlock_irqrestore(ap->lock, flags);
-		} else {
-			if (likely(!(qc->err_mask & AC_ERR_HSM)))
-				ata_qc_complete(qc);
-			else
-				ata_port_freeze(ap);
-		}
-	} else {
-		if (in_wq) {
-			spin_lock_irqsave(ap->lock, flags);
-			ap->ops->irq_on(ap);
-			ata_qc_complete(qc);
-			spin_unlock_irqrestore(ap->lock, flags);
-		} else
-			ata_qc_complete(qc);
-	}
-}
-
-/**
- *	ata_hsm_move - move the HSM to the next state.
- *	@ap: the target ata_port
- *	@qc: qc on going
- *	@status: current device status
- *	@in_wq: 1 if called from workqueue, 0 otherwise
- *
- *	RETURNS:
- *	1 when poll next status needed, 0 otherwise.
- */
-int ata_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
-		 u8 status, int in_wq)
-{
-	unsigned long flags = 0;
-	int poll_next;
-
-	WARN_ON((qc->flags & ATA_QCFLAG_ACTIVE) == 0);
-
-	/* Make sure ata_qc_issue_prot() does not throw things
-	 * like DMA polling into the workqueue. Notice that
-	 * in_wq is not equivalent to (qc->tf.flags & ATA_TFLAG_POLLING).
-	 */
-	WARN_ON(in_wq != ata_hsm_ok_in_wq(ap, qc));
-
-fsm_start:
-	DPRINTK("ata%u: protocol %d task_state %d (dev_stat 0x%X)\n",
-		ap->print_id, qc->tf.protocol, ap->hsm_task_state, status);
-
-	switch (ap->hsm_task_state) {
-	case HSM_ST_FIRST:
-		/* Send first data block or PACKET CDB */
-
-		/* If polling, we will stay in the work queue after
-		 * sending the data. Otherwise, interrupt handler
-		 * takes over after sending the data.
-		 */
-		poll_next = (qc->tf.flags & ATA_TFLAG_POLLING);
-
-		/* check device status */
-		if (unlikely((status & ATA_DRQ) == 0)) {
-			/* handle BSY=0, DRQ=0 as error */
-			if (likely(status & (ATA_ERR | ATA_DF)))
-				/* device stops HSM for abort/error */
-				qc->err_mask |= AC_ERR_DEV;
-			else
-				/* HSM violation. Let EH handle this */
-				qc->err_mask |= AC_ERR_HSM;
-
-			ap->hsm_task_state = HSM_ST_ERR;
-			goto fsm_start;
-		}
-
-		/* Device should not ask for data transfer (DRQ=1)
-		 * when it finds something wrong.
-		 * We ignore DRQ here and stop the HSM by
-		 * changing hsm_task_state to HSM_ST_ERR and
-		 * let the EH abort the command or reset the device.
-		 */
-		if (unlikely(status & (ATA_ERR | ATA_DF))) {
-			/* Some ATAPI tape drives forget to clear the ERR bit
-			 * when doing the next command (mostly request sense).
-			 * We ignore ERR here to workaround and proceed sending
-			 * the CDB.
-			 */
-			if (!(qc->dev->horkage & ATA_HORKAGE_STUCK_ERR)) {
-				ata_port_printk(ap, KERN_WARNING,
-						"DRQ=1 with device error, "
-						"dev_stat 0x%X\n", status);
-				qc->err_mask |= AC_ERR_HSM;
-				ap->hsm_task_state = HSM_ST_ERR;
-				goto fsm_start;
-			}
-		}
-
-		/* Send the CDB (atapi) or the first data block (ata pio out).
-		 * During the state transition, interrupt handler shouldn't
-		 * be invoked before the data transfer is complete and
-		 * hsm_task_state is changed. Hence, the following locking.
-		 */
-		if (in_wq)
-			spin_lock_irqsave(ap->lock, flags);
-
-		if (qc->tf.protocol == ATA_PROT_PIO) {
-			/* PIO data out protocol.
-			 * send first data block.
-			 */
-
-			/* ata_pio_sectors() might change the state
-			 * to HSM_ST_LAST. so, the state is changed here
-			 * before ata_pio_sectors().
-			 */
-			ap->hsm_task_state = HSM_ST;
-			ata_pio_sectors(qc);
-		} else
-			/* send CDB */
-			atapi_send_cdb(ap, qc);
-
-		if (in_wq)
-			spin_unlock_irqrestore(ap->lock, flags);
-
-		/* if polling, ata_pio_task() handles the rest.
-		 * otherwise, interrupt handler takes over from here.
-		 */
-		break;
-
-	case HSM_ST:
-		/* complete command or read/write the data register */
-		if (qc->tf.protocol == ATAPI_PROT_PIO) {
-			/* ATAPI PIO protocol */
-			if ((status & ATA_DRQ) == 0) {
-				/* No more data to transfer or device error.
-				 * Device error will be tagged in HSM_ST_LAST.
-				 */
-				ap->hsm_task_state = HSM_ST_LAST;
-				goto fsm_start;
-			}
-
-			/* Device should not ask for data transfer (DRQ=1)
-			 * when it finds something wrong.
-			 * We ignore DRQ here and stop the HSM by
-			 * changing hsm_task_state to HSM_ST_ERR and
-			 * let the EH abort the command or reset the device.
-			 */
-			if (unlikely(status & (ATA_ERR | ATA_DF))) {
-				ata_port_printk(ap, KERN_WARNING, "DRQ=1 with "
-						"device error, dev_stat 0x%X\n",
-						status);
-				qc->err_mask |= AC_ERR_HSM;
-				ap->hsm_task_state = HSM_ST_ERR;
-				goto fsm_start;
-			}
-
-			atapi_pio_bytes(qc);
-
-			if (unlikely(ap->hsm_task_state == HSM_ST_ERR))
-				/* bad ireason reported by device */
-				goto fsm_start;
-
-		} else {
-			/* ATA PIO protocol */
-			if (unlikely((status & ATA_DRQ) == 0)) {
-				/* handle BSY=0, DRQ=0 as error */
-				if (likely(status & (ATA_ERR | ATA_DF)))
-					/* device stops HSM for abort/error */
-					qc->err_mask |= AC_ERR_DEV;
-				else
-					/* HSM violation. Let EH handle this.
-					 * Phantom devices also trigger this
-					 * condition.  Mark hint.
-					 */
-					qc->err_mask |= AC_ERR_HSM |
-							AC_ERR_NODEV_HINT;
-
-				ap->hsm_task_state = HSM_ST_ERR;
-				goto fsm_start;
-			}
-
-			/* For PIO reads, some devices may ask for
-			 * data transfer (DRQ=1) alone with ERR=1.
-			 * We respect DRQ here and transfer one
-			 * block of junk data before changing the
-			 * hsm_task_state to HSM_ST_ERR.
-			 *
-			 * For PIO writes, ERR=1 DRQ=1 doesn't make
-			 * sense since the data block has been
-			 * transferred to the device.
-			 */
-			if (unlikely(status & (ATA_ERR | ATA_DF))) {
-				/* data might be corrputed */
-				qc->err_mask |= AC_ERR_DEV;
-
-				if (!(qc->tf.flags & ATA_TFLAG_WRITE)) {
-					ata_pio_sectors(qc);
-					status = ata_wait_idle(ap);
-				}
-
-				if (status & (ATA_BUSY | ATA_DRQ))
-					qc->err_mask |= AC_ERR_HSM;
-
-				/* ata_pio_sectors() might change the
-				 * state to HSM_ST_LAST. so, the state
-				 * is changed after ata_pio_sectors().
-				 */
-				ap->hsm_task_state = HSM_ST_ERR;
-				goto fsm_start;
-			}
+ *	spin_lock_irqsave(host lock)
+ *
+ *	RETURNS: 0 when ATAPI DMA can be used
+ *               nonzero otherwise
+ */
+int ata_check_atapi_dma(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
 
-			ata_pio_sectors(qc);
+	/* Don't allow DMA if it isn't multiple of 16 bytes.  Quite a
+	 * few ATAPI devices choke on such DMA requests.
+	 */
+	if (unlikely(qc->nbytes & 15))
+		return 1;
 
-			if (ap->hsm_task_state == HSM_ST_LAST &&
-			    (!(qc->tf.flags & ATA_TFLAG_WRITE))) {
-				/* all data read */
-				status = ata_wait_idle(ap);
-				goto fsm_start;
-			}
-		}
+	if (ap->ops->check_atapi_dma)
+		return ap->ops->check_atapi_dma(qc);
 
-		poll_next = 1;
-		break;
+	return 0;
+}
 
-	case HSM_ST_LAST:
-		if (unlikely(!ata_ok(status))) {
-			qc->err_mask |= __ac_err_mask(status);
-			ap->hsm_task_state = HSM_ST_ERR;
-			goto fsm_start;
-		}
+/**
+ *	ata_std_qc_defer - Check whether a qc needs to be deferred
+ *	@qc: ATA command in question
+ *
+ *	Non-NCQ commands cannot run with any other command, NCQ or
+ *	not.  As upper layer only knows the queue depth, we are
+ *	responsible for maintaining exclusion.  This function checks
+ *	whether a new command @qc can be issued.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ *
+ *	RETURNS:
+ *	ATA_DEFER_* if deferring is needed, 0 otherwise.
+ */
+int ata_std_qc_defer(struct ata_queued_cmd *qc)
+{
+	struct ata_link *link = qc->dev->link;
 
-		/* no more data to transfer */
-		DPRINTK("ata%u: dev %u command complete, drv_stat 0x%x\n",
-			ap->print_id, qc->dev->devno, status);
+	if (qc->tf.protocol == ATA_PROT_NCQ) {
+		if (!ata_tag_valid(link->active_tag))
+			return 0;
+	} else {
+		if (!ata_tag_valid(link->active_tag) && !link->sactive)
+			return 0;
+	}
 
-		WARN_ON(qc->err_mask);
+	return ATA_DEFER_LINK;
+}
 
-		ap->hsm_task_state = HSM_ST_IDLE;
+void ata_noop_qc_prep(struct ata_queued_cmd *qc) { }
 
-		/* complete taskfile transaction */
-		ata_hsm_qc_complete(qc, in_wq);
+/**
+ *	ata_sg_init - Associate command with scatter-gather table.
+ *	@qc: Command to be associated
+ *	@sg: Scatter-gather table.
+ *	@n_elem: Number of elements in s/g table.
+ *
+ *	Initialize the data-related elements of queued_cmd @qc
+ *	to point to a scatter-gather table @sg, containing @n_elem
+ *	elements.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ */
+void ata_sg_init(struct ata_queued_cmd *qc, struct scatterlist *sg,
+		 unsigned int n_elem)
+{
+	qc->sg = sg;
+	qc->n_elem = n_elem;
+	qc->cursg = qc->sg;
+}
 
-		poll_next = 0;
-		break;
+/**
+ *	ata_sg_setup - DMA-map the scatter-gather table associated with a command.
+ *	@qc: Command with scatter-gather table to be mapped.
+ *
+ *	DMA-map the scatter-gather table associated with queued_cmd @qc.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ *
+ *	RETURNS:
+ *	Zero on success, negative on error.
+ *
+ */
+static int ata_sg_setup(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	unsigned int n_elem;
 
-	case HSM_ST_ERR:
-		/* make sure qc->err_mask is available to
-		 * know what's wrong and recover
-		 */
-		WARN_ON(qc->err_mask == 0);
+	VPRINTK("ENTER, ata%u\n", ap->print_id);
 
-		ap->hsm_task_state = HSM_ST_IDLE;
+	n_elem = dma_map_sg(ap->dev, qc->sg, qc->n_elem, qc->dma_dir);
+	if (n_elem < 1)
+		return -1;
 
-		/* complete taskfile transaction */
-		ata_hsm_qc_complete(qc, in_wq);
+	DPRINTK("%d sg elements mapped\n", n_elem);
 
-		poll_next = 0;
-		break;
-	default:
-		poll_next = 0;
-		BUG();
-	}
+	qc->n_elem = n_elem;
+	qc->flags |= ATA_QCFLAG_DMAMAP;
 
-	return poll_next;
+	return 0;
 }
 
-static void ata_pio_task(struct work_struct *work)
+/**
+ *	swap_buf_le16 - swap halves of 16-bit words in place
+ *	@buf:  Buffer to swap
+ *	@buf_words:  Number of 16-bit words in buffer.
+ *
+ *	Swap halves of 16-bit words if needed to convert from
+ *	little-endian byte order to native cpu byte order, or
+ *	vice-versa.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+void swap_buf_le16(u16 *buf, unsigned int buf_words)
 {
-	struct ata_port *ap =
-		container_of(work, struct ata_port, port_task.work);
-	struct ata_queued_cmd *qc = ap->port_task_data;
-	u8 status;
-	int poll_next;
-
-fsm_start:
-	WARN_ON(ap->hsm_task_state == HSM_ST_IDLE);
-
-	/*
-	 * This is purely heuristic.  This is a fast path.
-	 * Sometimes when we enter, BSY will be cleared in
-	 * a chk-status or two.  If not, the drive is probably seeking
-	 * or something.  Snooze for a couple msecs, then
-	 * chk-status again.  If still busy, queue delayed work.
-	 */
-	status = ata_busy_wait(ap, ATA_BUSY, 5);
-	if (status & ATA_BUSY) {
-		msleep(2);
-		status = ata_busy_wait(ap, ATA_BUSY, 10);
-		if (status & ATA_BUSY) {
-			ata_pio_queue_task(ap, qc, ATA_SHORT_PAUSE);
-			return;
-		}
-	}
-
-	/* move the HSM */
-	poll_next = ata_hsm_move(ap, qc, status, 1);
+#ifdef __BIG_ENDIAN
+	unsigned int i;
 
-	/* another command or interrupt handler
-	 * may be running at this point.
-	 */
-	if (poll_next)
-		goto fsm_start;
+	for (i = 0; i < buf_words; i++)
+		buf[i] = le16_to_cpu(buf[i]);
+#endif /* __BIG_ENDIAN */
 }
 
 /**
@@ -6121,285 +4605,6 @@ err:
 	ata_qc_complete(qc);
 }
 
-/**
- *	ata_qc_issue_prot - issue taskfile to device in proto-dependent manner
- *	@qc: command to issue to device
- *
- *	Using various libata functions and hooks, this function
- *	starts an ATA command.  ATA commands are grouped into
- *	classes called "protocols", and issuing each type of protocol
- *	is slightly different.
- *
- *	May be used as the qc_issue() entry in ata_port_operations.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- *
- *	RETURNS:
- *	Zero on success, AC_ERR_* mask on failure
- */
-
-unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc)
-{
-	struct ata_port *ap = qc->ap;
-
-	/* Use polling pio if the LLD doesn't handle
-	 * interrupt driven pio and atapi CDB interrupt.
-	 */
-	if (ap->flags & ATA_FLAG_PIO_POLLING) {
-		switch (qc->tf.protocol) {
-		case ATA_PROT_PIO:
-		case ATA_PROT_NODATA:
-		case ATAPI_PROT_PIO:
-		case ATAPI_PROT_NODATA:
-			qc->tf.flags |= ATA_TFLAG_POLLING;
-			break;
-		case ATAPI_PROT_DMA:
-			if (qc->dev->flags & ATA_DFLAG_CDB_INTR)
-				/* see ata_dma_blacklisted() */
-				BUG();
-			break;
-		default:
-			break;
-		}
-	}
-
-	/* select the device */
-	ata_dev_select(ap, qc->dev->devno, 1, 0);
-
-	/* start the command */
-	switch (qc->tf.protocol) {
-	case ATA_PROT_NODATA:
-		if (qc->tf.flags & ATA_TFLAG_POLLING)
-			ata_qc_set_polling(qc);
-
-		ata_tf_to_host(ap, &qc->tf);
-		ap->hsm_task_state = HSM_ST_LAST;
-
-		if (qc->tf.flags & ATA_TFLAG_POLLING)
-			ata_pio_queue_task(ap, qc, 0);
-
-		break;
-
-	case ATA_PROT_DMA:
-		WARN_ON(qc->tf.flags & ATA_TFLAG_POLLING);
-
-		ap->ops->tf_load(ap, &qc->tf);	 /* load tf registers */
-		ap->ops->bmdma_setup(qc);	    /* set up bmdma */
-		ap->ops->bmdma_start(qc);	    /* initiate bmdma */
-		ap->hsm_task_state = HSM_ST_LAST;
-		break;
-
-	case ATA_PROT_PIO:
-		if (qc->tf.flags & ATA_TFLAG_POLLING)
-			ata_qc_set_polling(qc);
-
-		ata_tf_to_host(ap, &qc->tf);
-
-		if (qc->tf.flags & ATA_TFLAG_WRITE) {
-			/* PIO data out protocol */
-			ap->hsm_task_state = HSM_ST_FIRST;
-			ata_pio_queue_task(ap, qc, 0);
-
-			/* always send first data block using
-			 * the ata_pio_task() codepath.
-			 */
-		} else {
-			/* PIO data in protocol */
-			ap->hsm_task_state = HSM_ST;
-
-			if (qc->tf.flags & ATA_TFLAG_POLLING)
-				ata_pio_queue_task(ap, qc, 0);
-
-			/* if polling, ata_pio_task() handles the rest.
-			 * otherwise, interrupt handler takes over from here.
-			 */
-		}
-
-		break;
-
-	case ATAPI_PROT_PIO:
-	case ATAPI_PROT_NODATA:
-		if (qc->tf.flags & ATA_TFLAG_POLLING)
-			ata_qc_set_polling(qc);
-
-		ata_tf_to_host(ap, &qc->tf);
-
-		ap->hsm_task_state = HSM_ST_FIRST;
-
-		/* send cdb by polling if no cdb interrupt */
-		if ((!(qc->dev->flags & ATA_DFLAG_CDB_INTR)) ||
-		    (qc->tf.flags & ATA_TFLAG_POLLING))
-			ata_pio_queue_task(ap, qc, 0);
-		break;
-
-	case ATAPI_PROT_DMA:
-		WARN_ON(qc->tf.flags & ATA_TFLAG_POLLING);
-
-		ap->ops->tf_load(ap, &qc->tf);	 /* load tf registers */
-		ap->ops->bmdma_setup(qc);	    /* set up bmdma */
-		ap->hsm_task_state = HSM_ST_FIRST;
-
-		/* send cdb by polling if no cdb interrupt */
-		if (!(qc->dev->flags & ATA_DFLAG_CDB_INTR))
-			ata_pio_queue_task(ap, qc, 0);
-		break;
-
-	default:
-		WARN_ON(1);
-		return AC_ERR_SYSTEM;
-	}
-
-	return 0;
-}
-
-/**
- *	ata_host_intr - Handle host interrupt for given (port, task)
- *	@ap: Port on which interrupt arrived (possibly...)
- *	@qc: Taskfile currently active in engine
- *
- *	Handle host interrupt for given queued command.  Currently,
- *	only DMA interrupts are handled.  All other commands are
- *	handled via polling with interrupts disabled (nIEN bit).
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- *
- *	RETURNS:
- *	One if interrupt was handled, zero if not (shared irq).
- */
-
-inline unsigned int ata_host_intr(struct ata_port *ap,
-				  struct ata_queued_cmd *qc)
-{
-	struct ata_eh_info *ehi = &ap->link.eh_info;
-	u8 status, host_stat = 0;
-
-	VPRINTK("ata%u: protocol %d task_state %d\n",
-		ap->print_id, qc->tf.protocol, ap->hsm_task_state);
-
-	/* Check whether we are expecting interrupt in this state */
-	switch (ap->hsm_task_state) {
-	case HSM_ST_FIRST:
-		/* Some pre-ATAPI-4 devices assert INTRQ
-		 * at this state when ready to receive CDB.
-		 */
-
-		/* Check the ATA_DFLAG_CDB_INTR flag is enough here.
-		 * The flag was turned on only for atapi devices.  No
-		 * need to check ata_is_atapi(qc->tf.protocol) again.
-		 */
-		if (!(qc->dev->flags & ATA_DFLAG_CDB_INTR))
-			goto idle_irq;
-		break;
-	case HSM_ST_LAST:
-		if (qc->tf.protocol == ATA_PROT_DMA ||
-		    qc->tf.protocol == ATAPI_PROT_DMA) {
-			/* check status of DMA engine */
-			host_stat = ap->ops->bmdma_status(ap);
-			VPRINTK("ata%u: host_stat 0x%X\n",
-				ap->print_id, host_stat);
-
-			/* if it's not our irq... */
-			if (!(host_stat & ATA_DMA_INTR))
-				goto idle_irq;
-
-			/* before we do anything else, clear DMA-Start bit */
-			ap->ops->bmdma_stop(qc);
-
-			if (unlikely(host_stat & ATA_DMA_ERR)) {
-				/* error when transfering data to/from memory */
-				qc->err_mask |= AC_ERR_HOST_BUS;
-				ap->hsm_task_state = HSM_ST_ERR;
-			}
-		}
-		break;
-	case HSM_ST:
-		break;
-	default:
-		goto idle_irq;
-	}
-
-	/* check altstatus */
-	status = ata_altstatus(ap);
-	if (status & ATA_BUSY)
-		goto idle_irq;
-
-	/* check main status, clearing INTRQ */
-	status = ata_chk_status(ap);
-	if (unlikely(status & ATA_BUSY))
-		goto idle_irq;
-
-	/* ack bmdma irq events */
-	ap->ops->irq_clear(ap);
-
-	ata_hsm_move(ap, qc, status, 0);
-
-	if (unlikely(qc->err_mask) && (qc->tf.protocol == ATA_PROT_DMA ||
-				       qc->tf.protocol == ATAPI_PROT_DMA))
-		ata_ehi_push_desc(ehi, "BMDMA stat 0x%x", host_stat);
-
-	return 1;	/* irq handled */
-
-idle_irq:
-	ap->stats.idle_irq++;
-
-#ifdef ATA_IRQ_TRAP
-	if ((ap->stats.idle_irq % 1000) == 0) {
-		ata_chk_status(ap);
-		ap->ops->irq_clear(ap);
-		ata_port_printk(ap, KERN_WARNING, "irq trap\n");
-		return 1;
-	}
-#endif
-	return 0;	/* irq not handled */
-}
-
-/**
- *	ata_interrupt - Default ATA host interrupt handler
- *	@irq: irq line (unused)
- *	@dev_instance: pointer to our ata_host information structure
- *
- *	Default interrupt handler for PCI IDE devices.  Calls
- *	ata_host_intr() for each port that is not disabled.
- *
- *	LOCKING:
- *	Obtains host lock during operation.
- *
- *	RETURNS:
- *	IRQ_NONE or IRQ_HANDLED.
- */
-
-irqreturn_t ata_interrupt(int irq, void *dev_instance)
-{
-	struct ata_host *host = dev_instance;
-	unsigned int i;
-	unsigned int handled = 0;
-	unsigned long flags;
-
-	/* TODO: make _irqsave conditional on x86 PCI IDE legacy mode */
-	spin_lock_irqsave(&host->lock, flags);
-
-	for (i = 0; i < host->n_ports; i++) {
-		struct ata_port *ap;
-
-		ap = host->ports[i];
-		if (ap &&
-		    !(ap->flags & ATA_FLAG_DISABLED)) {
-			struct ata_queued_cmd *qc;
-
-			qc = ata_qc_from_tag(ap, ap->link.active_tag);
-			if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)) &&
-			    (qc->flags & ATA_QCFLAG_ACTIVE))
-				handled |= ata_host_intr(ap, qc);
-		}
-	}
-
-	spin_unlock_irqrestore(&host->lock, flags);
-
-	return IRQ_RETVAL(handled);
-}
-
 /**
  *	sata_scr_valid - test whether SCRs are accessible
  *	@link: ATA link to test SCR accessibility for
@@ -7432,33 +5637,6 @@ void ata_host_detach(struct ata_host *host)
 	ata_acpi_dissociate(host);
 }
 
-/**
- *	ata_std_ports - initialize ioaddr with standard port offsets.
- *	@ioaddr: IO address structure to be initialized
- *
- *	Utility function which initializes data_addr, error_addr,
- *	feature_addr, nsect_addr, lbal_addr, lbam_addr, lbah_addr,
- *	device_addr, status_addr, and command_addr to standard offsets
- *	relative to cmd_addr.
- *
- *	Does not set ctl_addr, altstatus_addr, bmdma_addr, or scr_addr.
- */
-
-void ata_std_ports(struct ata_ioports *ioaddr)
-{
-	ioaddr->data_addr = ioaddr->cmd_addr + ATA_REG_DATA;
-	ioaddr->error_addr = ioaddr->cmd_addr + ATA_REG_ERR;
-	ioaddr->feature_addr = ioaddr->cmd_addr + ATA_REG_FEATURE;
-	ioaddr->nsect_addr = ioaddr->cmd_addr + ATA_REG_NSECT;
-	ioaddr->lbal_addr = ioaddr->cmd_addr + ATA_REG_LBAL;
-	ioaddr->lbam_addr = ioaddr->cmd_addr + ATA_REG_LBAM;
-	ioaddr->lbah_addr = ioaddr->cmd_addr + ATA_REG_LBAH;
-	ioaddr->device_addr = ioaddr->cmd_addr + ATA_REG_DEVICE;
-	ioaddr->status_addr = ioaddr->cmd_addr + ATA_REG_STATUS;
-	ioaddr->command_addr = ioaddr->cmd_addr + ATA_REG_CMD;
-}
-
-
 #ifdef CONFIG_PCI
 
 /**
@@ -7890,12 +6068,9 @@ EXPORT_SYMBOL_GPL(sata_deb_timing_long);
 EXPORT_SYMBOL_GPL(ata_base_port_ops);
 EXPORT_SYMBOL_GPL(sata_port_ops);
 EXPORT_SYMBOL_GPL(sata_pmp_port_ops);
-EXPORT_SYMBOL_GPL(ata_sff_port_ops);
-EXPORT_SYMBOL_GPL(ata_bmdma_port_ops);
 EXPORT_SYMBOL_GPL(ata_dummy_port_ops);
 EXPORT_SYMBOL_GPL(ata_dummy_port_info);
 EXPORT_SYMBOL_GPL(ata_std_bios_param);
-EXPORT_SYMBOL_GPL(ata_std_ports);
 EXPORT_SYMBOL_GPL(ata_host_init);
 EXPORT_SYMBOL_GPL(ata_host_alloc);
 EXPORT_SYMBOL_GPL(ata_host_alloc_pinfo);
@@ -7904,14 +6079,9 @@ EXPORT_SYMBOL_GPL(ata_host_register);
 EXPORT_SYMBOL_GPL(ata_host_activate);
 EXPORT_SYMBOL_GPL(ata_host_detach);
 EXPORT_SYMBOL_GPL(ata_sg_init);
-EXPORT_SYMBOL_GPL(ata_hsm_move);
 EXPORT_SYMBOL_GPL(ata_qc_complete);
 EXPORT_SYMBOL_GPL(ata_qc_complete_multiple);
-EXPORT_SYMBOL_GPL(ata_qc_issue_prot);
-EXPORT_SYMBOL_GPL(ata_tf_load);
-EXPORT_SYMBOL_GPL(ata_tf_read);
 EXPORT_SYMBOL_GPL(ata_noop_dev_select);
-EXPORT_SYMBOL_GPL(ata_std_dev_select);
 EXPORT_SYMBOL_GPL(sata_print_link_status);
 EXPORT_SYMBOL_GPL(atapi_cmd_type);
 EXPORT_SYMBOL_GPL(ata_tf_to_fis);
@@ -7923,54 +6093,27 @@ EXPORT_SYMBOL_GPL(ata_xfer_mode2mask);
 EXPORT_SYMBOL_GPL(ata_xfer_mode2shift);
 EXPORT_SYMBOL_GPL(ata_mode_string);
 EXPORT_SYMBOL_GPL(ata_id_xfermask);
-EXPORT_SYMBOL_GPL(ata_check_status);
-EXPORT_SYMBOL_GPL(ata_altstatus);
-EXPORT_SYMBOL_GPL(ata_exec_command);
 EXPORT_SYMBOL_GPL(ata_port_start);
-EXPORT_SYMBOL_GPL(ata_sff_port_start);
-EXPORT_SYMBOL_GPL(ata_interrupt);
 EXPORT_SYMBOL_GPL(ata_do_set_mode);
-EXPORT_SYMBOL_GPL(ata_data_xfer);
-EXPORT_SYMBOL_GPL(ata_data_xfer_noirq);
 EXPORT_SYMBOL_GPL(ata_std_qc_defer);
-EXPORT_SYMBOL_GPL(ata_qc_prep);
-EXPORT_SYMBOL_GPL(ata_dumb_qc_prep);
 EXPORT_SYMBOL_GPL(ata_noop_qc_prep);
-EXPORT_SYMBOL_GPL(ata_bmdma_setup);
-EXPORT_SYMBOL_GPL(ata_bmdma_start);
-EXPORT_SYMBOL_GPL(ata_bmdma_irq_clear);
 EXPORT_SYMBOL_GPL(ata_noop_irq_clear);
-EXPORT_SYMBOL_GPL(ata_bmdma_status);
-EXPORT_SYMBOL_GPL(ata_bmdma_stop);
-EXPORT_SYMBOL_GPL(ata_bmdma_freeze);
-EXPORT_SYMBOL_GPL(ata_bmdma_thaw);
-EXPORT_SYMBOL_GPL(ata_bmdma_error_handler);
-EXPORT_SYMBOL_GPL(ata_bmdma_post_internal_cmd);
 EXPORT_SYMBOL_GPL(ata_port_probe);
 EXPORT_SYMBOL_GPL(ata_dev_disable);
 EXPORT_SYMBOL_GPL(sata_set_spd);
 EXPORT_SYMBOL_GPL(sata_link_debounce);
 EXPORT_SYMBOL_GPL(sata_link_resume);
-EXPORT_SYMBOL_GPL(ata_bus_reset);
-EXPORT_SYMBOL_GPL(ata_std_prereset);
-EXPORT_SYMBOL_GPL(ata_std_softreset);
 EXPORT_SYMBOL_GPL(sata_link_hardreset);
-EXPORT_SYMBOL_GPL(sata_std_hardreset);
-EXPORT_SYMBOL_GPL(ata_std_postreset);
 EXPORT_SYMBOL_GPL(ata_dev_classify);
 EXPORT_SYMBOL_GPL(ata_dev_pair);
 EXPORT_SYMBOL_GPL(ata_port_disable);
 EXPORT_SYMBOL_GPL(ata_ratelimit);
 EXPORT_SYMBOL_GPL(ata_wait_register);
-EXPORT_SYMBOL_GPL(ata_busy_sleep);
-EXPORT_SYMBOL_GPL(ata_wait_after_reset);
-EXPORT_SYMBOL_GPL(ata_wait_ready);
 EXPORT_SYMBOL_GPL(ata_scsi_ioctl);
 EXPORT_SYMBOL_GPL(ata_scsi_queuecmd);
 EXPORT_SYMBOL_GPL(ata_scsi_slave_config);
 EXPORT_SYMBOL_GPL(ata_scsi_slave_destroy);
 EXPORT_SYMBOL_GPL(ata_scsi_change_queue_depth);
-EXPORT_SYMBOL_GPL(ata_host_intr);
 EXPORT_SYMBOL_GPL(sata_scr_valid);
 EXPORT_SYMBOL_GPL(sata_scr_read);
 EXPORT_SYMBOL_GPL(sata_scr_write);
@@ -7993,11 +6136,6 @@ EXPORT_SYMBOL_GPL(ata_timing_cycle2mode);
 
 #ifdef CONFIG_PCI
 EXPORT_SYMBOL_GPL(pci_test_config_bits);
-EXPORT_SYMBOL_GPL(ata_pci_init_sff_host);
-EXPORT_SYMBOL_GPL(ata_pci_init_bmdma);
-EXPORT_SYMBOL_GPL(ata_pci_prepare_sff_host);
-EXPORT_SYMBOL_GPL(ata_pci_activate_sff_host);
-EXPORT_SYMBOL_GPL(ata_pci_init_one);
 EXPORT_SYMBOL_GPL(ata_pci_remove_one);
 #ifdef CONFIG_PM
 EXPORT_SYMBOL_GPL(ata_pci_device_do_suspend);
@@ -8005,8 +6143,6 @@ EXPORT_SYMBOL_GPL(ata_pci_device_do_resume);
 EXPORT_SYMBOL_GPL(ata_pci_device_suspend);
 EXPORT_SYMBOL_GPL(ata_pci_device_resume);
 #endif /* CONFIG_PM */
-EXPORT_SYMBOL_GPL(ata_pci_default_filter);
-EXPORT_SYMBOL_GPL(ata_pci_clear_simplex);
 #endif /* CONFIG_PCI */
 
 EXPORT_SYMBOL_GPL(sata_pmp_qc_defer_cmd_switch);
@@ -8033,8 +6169,6 @@ EXPORT_SYMBOL_GPL(ata_eh_qc_complete);
 EXPORT_SYMBOL_GPL(ata_eh_qc_retry);
 EXPORT_SYMBOL_GPL(ata_do_eh);
 EXPORT_SYMBOL_GPL(ata_std_error_handler);
-EXPORT_SYMBOL_GPL(ata_irq_on);
-EXPORT_SYMBOL_GPL(ata_dev_try_classify);
 
 EXPORT_SYMBOL_GPL(ata_cable_40wire);
 EXPORT_SYMBOL_GPL(ata_cable_80wire);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 40645ed125b1..840ae6da59bc 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -35,9 +35,185 @@
 #include <linux/kernel.h>
 #include <linux/pci.h>
 #include <linux/libata.h>
+#include <linux/highmem.h>
 
 #include "libata.h"
 
+const struct ata_port_operations ata_sff_port_ops = {
+	.inherits		= &ata_base_port_ops,
+
+	.qc_prep		= ata_qc_prep,
+	.qc_issue		= ata_qc_issue_prot,
+
+	.freeze			= ata_bmdma_freeze,
+	.thaw			= ata_bmdma_thaw,
+	.softreset		= ata_std_softreset,
+	.error_handler		= ata_bmdma_error_handler,
+	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
+
+	.dev_select		= ata_std_dev_select,
+	.check_status		= ata_check_status,
+	.tf_load		= ata_tf_load,
+	.tf_read		= ata_tf_read,
+	.exec_command		= ata_exec_command,
+	.data_xfer		= ata_data_xfer,
+	.irq_on			= ata_irq_on,
+
+	.port_start		= ata_sff_port_start,
+};
+
+const struct ata_port_operations ata_bmdma_port_ops = {
+	.inherits		= &ata_sff_port_ops,
+
+	.mode_filter		= ata_pci_default_filter,
+
+	.bmdma_setup		= ata_bmdma_setup,
+	.bmdma_start		= ata_bmdma_start,
+	.bmdma_stop		= ata_bmdma_stop,
+	.bmdma_status		= ata_bmdma_status,
+	.irq_clear		= ata_bmdma_irq_clear,
+};
+
+/**
+ *	ata_fill_sg - Fill PCI IDE PRD table
+ *	@qc: Metadata associated with taskfile to be transferred
+ *
+ *	Fill PCI IDE PRD (scatter-gather) table with segments
+ *	associated with the current disk command.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ *
+ */
+static void ata_fill_sg(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct scatterlist *sg;
+	unsigned int si, pi;
+
+	pi = 0;
+	for_each_sg(qc->sg, sg, qc->n_elem, si) {
+		u32 addr, offset;
+		u32 sg_len, len;
+
+		/* determine if physical DMA addr spans 64K boundary.
+		 * Note h/w doesn't support 64-bit, so we unconditionally
+		 * truncate dma_addr_t to u32.
+		 */
+		addr = (u32) sg_dma_address(sg);
+		sg_len = sg_dma_len(sg);
+
+		while (sg_len) {
+			offset = addr & 0xffff;
+			len = sg_len;
+			if ((offset + sg_len) > 0x10000)
+				len = 0x10000 - offset;
+
+			ap->prd[pi].addr = cpu_to_le32(addr);
+			ap->prd[pi].flags_len = cpu_to_le32(len & 0xffff);
+			VPRINTK("PRD[%u] = (0x%X, 0x%X)\n", pi, addr, len);
+
+			pi++;
+			sg_len -= len;
+			addr += len;
+		}
+	}
+
+	ap->prd[pi - 1].flags_len |= cpu_to_le32(ATA_PRD_EOT);
+}
+
+/**
+ *	ata_fill_sg_dumb - Fill PCI IDE PRD table
+ *	@qc: Metadata associated with taskfile to be transferred
+ *
+ *	Fill PCI IDE PRD (scatter-gather) table with segments
+ *	associated with the current disk command. Perform the fill
+ *	so that we avoid writing any length 64K records for
+ *	controllers that don't follow the spec.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ *
+ */
+static void ata_fill_sg_dumb(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct scatterlist *sg;
+	unsigned int si, pi;
+
+	pi = 0;
+	for_each_sg(qc->sg, sg, qc->n_elem, si) {
+		u32 addr, offset;
+		u32 sg_len, len, blen;
+
+		/* determine if physical DMA addr spans 64K boundary.
+		 * Note h/w doesn't support 64-bit, so we unconditionally
+		 * truncate dma_addr_t to u32.
+		 */
+		addr = (u32) sg_dma_address(sg);
+		sg_len = sg_dma_len(sg);
+
+		while (sg_len) {
+			offset = addr & 0xffff;
+			len = sg_len;
+			if ((offset + sg_len) > 0x10000)
+				len = 0x10000 - offset;
+
+			blen = len & 0xffff;
+			ap->prd[pi].addr = cpu_to_le32(addr);
+			if (blen == 0) {
+			   /* Some PATA chipsets like the CS5530 can't
+			      cope with 0x0000 meaning 64K as the spec says */
+				ap->prd[pi].flags_len = cpu_to_le32(0x8000);
+				blen = 0x8000;
+				ap->prd[++pi].addr = cpu_to_le32(addr + 0x8000);
+			}
+			ap->prd[pi].flags_len = cpu_to_le32(blen);
+			VPRINTK("PRD[%u] = (0x%X, 0x%X)\n", pi, addr, len);
+
+			pi++;
+			sg_len -= len;
+			addr += len;
+		}
+	}
+
+	ap->prd[pi - 1].flags_len |= cpu_to_le32(ATA_PRD_EOT);
+}
+
+/**
+ *	ata_qc_prep - Prepare taskfile for submission
+ *	@qc: Metadata associated with taskfile to be prepared
+ *
+ *	Prepare ATA taskfile for submission.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ */
+void ata_qc_prep(struct ata_queued_cmd *qc)
+{
+	if (!(qc->flags & ATA_QCFLAG_DMAMAP))
+		return;
+
+	ata_fill_sg(qc);
+}
+
+/**
+ *	ata_dumb_qc_prep - Prepare taskfile for submission
+ *	@qc: Metadata associated with taskfile to be prepared
+ *
+ *	Prepare ATA taskfile for submission.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ */
+void ata_dumb_qc_prep(struct ata_queued_cmd *qc)
+{
+	if (!(qc->flags & ATA_QCFLAG_DMAMAP))
+		return;
+
+	ata_fill_sg_dumb(qc);
+}
+
 /**
  *	ata_check_status - Read device status reg & clear interrupt
  *	@ap: port where the device is
@@ -67,223 +243,1746 @@ u8 ata_check_status(struct ata_port *ap)
  *	LOCKING:
  *	Inherited from caller.
  */
-u8 ata_altstatus(struct ata_port *ap)
+u8 ata_altstatus(struct ata_port *ap)
+{
+	if (ap->ops->check_altstatus)
+		return ap->ops->check_altstatus(ap);
+
+	return ioread8(ap->ioaddr.altstatus_addr);
+}
+
+/**
+ *	ata_busy_sleep - sleep until BSY clears, or timeout
+ *	@ap: port containing status register to be polled
+ *	@tmout_pat: impatience timeout
+ *	@tmout: overall timeout
+ *
+ *	Sleep until ATA Status register bit BSY clears,
+ *	or a timeout occurs.
+ *
+ *	LOCKING:
+ *	Kernel thread context (may sleep).
+ *
+ *	RETURNS:
+ *	0 on success, -errno otherwise.
+ */
+int ata_busy_sleep(struct ata_port *ap,
+		   unsigned long tmout_pat, unsigned long tmout)
+{
+	unsigned long timer_start, timeout;
+	u8 status;
+
+	status = ata_busy_wait(ap, ATA_BUSY, 300);
+	timer_start = jiffies;
+	timeout = timer_start + tmout_pat;
+	while (status != 0xff && (status & ATA_BUSY) &&
+	       time_before(jiffies, timeout)) {
+		msleep(50);
+		status = ata_busy_wait(ap, ATA_BUSY, 3);
+	}
+
+	if (status != 0xff && (status & ATA_BUSY))
+		ata_port_printk(ap, KERN_WARNING,
+				"port is slow to respond, please be patient "
+				"(Status 0x%x)\n", status);
+
+	timeout = timer_start + tmout;
+	while (status != 0xff && (status & ATA_BUSY) &&
+	       time_before(jiffies, timeout)) {
+		msleep(50);
+		status = ata_chk_status(ap);
+	}
+
+	if (status == 0xff)
+		return -ENODEV;
+
+	if (status & ATA_BUSY) {
+		ata_port_printk(ap, KERN_ERR, "port failed to respond "
+				"(%lu secs, Status 0x%x)\n",
+				tmout / HZ, status);
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+/**
+ *	ata_wait_ready - sleep until BSY clears, or timeout
+ *	@ap: port containing status register to be polled
+ *	@deadline: deadline jiffies for the operation
+ *
+ *	Sleep until ATA Status register bit BSY clears, or timeout
+ *	occurs.
+ *
+ *	LOCKING:
+ *	Kernel thread context (may sleep).
+ *
+ *	RETURNS:
+ *	0 on success, -errno otherwise.
+ */
+int ata_wait_ready(struct ata_port *ap, unsigned long deadline)
+{
+	unsigned long start = jiffies;
+	int warned = 0;
+
+	while (1) {
+		u8 status = ata_chk_status(ap);
+		unsigned long now = jiffies;
+
+		if (!(status & ATA_BUSY))
+			return 0;
+		if (!ata_link_online(&ap->link) && status == 0xff)
+			return -ENODEV;
+		if (time_after(now, deadline))
+			return -EBUSY;
+
+		if (!warned && time_after(now, start + 5 * HZ) &&
+		    (deadline - now > 3 * HZ)) {
+			ata_port_printk(ap, KERN_WARNING,
+				"port is slow to respond, please be patient "
+				"(Status 0x%x)\n", status);
+			warned = 1;
+		}
+
+		msleep(50);
+	}
+}
+
+/**
+ *	ata_std_dev_select - Select device 0/1 on ATA bus
+ *	@ap: ATA channel to manipulate
+ *	@device: ATA device (numbered from zero) to select
+ *
+ *	Use the method defined in the ATA specification to
+ *	make either device 0, or device 1, active on the
+ *	ATA channel.  Works with both PIO and MMIO.
+ *
+ *	May be used as the dev_select() entry in ata_port_operations.
+ *
+ *	LOCKING:
+ *	caller.
+ */
+void ata_std_dev_select(struct ata_port *ap, unsigned int device)
+{
+	u8 tmp;
+
+	if (device == 0)
+		tmp = ATA_DEVICE_OBS;
+	else
+		tmp = ATA_DEVICE_OBS | ATA_DEV1;
+
+	iowrite8(tmp, ap->ioaddr.device_addr);
+	ata_pause(ap);		/* needed; also flushes, for mmio */
+}
+
+/**
+ *	ata_dev_select - Select device 0/1 on ATA bus
+ *	@ap: ATA channel to manipulate
+ *	@device: ATA device (numbered from zero) to select
+ *	@wait: non-zero to wait for Status register BSY bit to clear
+ *	@can_sleep: non-zero if context allows sleeping
+ *
+ *	Use the method defined in the ATA specification to
+ *	make either device 0, or device 1, active on the
+ *	ATA channel.
+ *
+ *	This is a high-level version of ata_std_dev_select(),
+ *	which additionally provides the services of inserting
+ *	the proper pauses and status polling, where needed.
+ *
+ *	LOCKING:
+ *	caller.
+ */
+void ata_dev_select(struct ata_port *ap, unsigned int device,
+			   unsigned int wait, unsigned int can_sleep)
+{
+	if (ata_msg_probe(ap))
+		ata_port_printk(ap, KERN_INFO, "ata_dev_select: ENTER, "
+				"device %u, wait %u\n", device, wait);
+
+	if (wait)
+		ata_wait_idle(ap);
+
+	ap->ops->dev_select(ap, device);
+
+	if (wait) {
+		if (can_sleep && ap->link.device[device].class == ATA_DEV_ATAPI)
+			msleep(150);
+		ata_wait_idle(ap);
+	}
+}
+
+/**
+ *	ata_irq_on - Enable interrupts on a port.
+ *	@ap: Port on which interrupts are enabled.
+ *
+ *	Enable interrupts on a legacy IDE device using MMIO or PIO,
+ *	wait for idle, clear any pending interrupts.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+u8 ata_irq_on(struct ata_port *ap)
+{
+	struct ata_ioports *ioaddr = &ap->ioaddr;
+	u8 tmp;
+
+	ap->ctl &= ~ATA_NIEN;
+	ap->last_ctl = ap->ctl;
+
+	if (ioaddr->ctl_addr)
+		iowrite8(ap->ctl, ioaddr->ctl_addr);
+	tmp = ata_wait_idle(ap);
+
+	ap->ops->irq_clear(ap);
+
+	return tmp;
+}
+
+/**
+ *	ata_bmdma_irq_clear - Clear PCI IDE BMDMA interrupt.
+ *	@ap: Port associated with this ATA transaction.
+ *
+ *	Clear interrupt and error flags in DMA status register.
+ *
+ *	May be used as the irq_clear() entry in ata_port_operations.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ */
+void ata_bmdma_irq_clear(struct ata_port *ap)
+{
+	void __iomem *mmio = ap->ioaddr.bmdma_addr;
+
+	if (!mmio)
+		return;
+
+	iowrite8(ioread8(mmio + ATA_DMA_STATUS), mmio + ATA_DMA_STATUS);
+}
+
+/**
+ *	ata_tf_load - send taskfile registers to host controller
+ *	@ap: Port to which output is sent
+ *	@tf: ATA taskfile register set
+ *
+ *	Outputs ATA taskfile to standard ATA host controller.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
+{
+	struct ata_ioports *ioaddr = &ap->ioaddr;
+	unsigned int is_addr = tf->flags & ATA_TFLAG_ISADDR;
+
+	if (tf->ctl != ap->last_ctl) {
+		if (ioaddr->ctl_addr)
+			iowrite8(tf->ctl, ioaddr->ctl_addr);
+		ap->last_ctl = tf->ctl;
+		ata_wait_idle(ap);
+	}
+
+	if (is_addr && (tf->flags & ATA_TFLAG_LBA48)) {
+		WARN_ON(!ioaddr->ctl_addr);
+		iowrite8(tf->hob_feature, ioaddr->feature_addr);
+		iowrite8(tf->hob_nsect, ioaddr->nsect_addr);
+		iowrite8(tf->hob_lbal, ioaddr->lbal_addr);
+		iowrite8(tf->hob_lbam, ioaddr->lbam_addr);
+		iowrite8(tf->hob_lbah, ioaddr->lbah_addr);
+		VPRINTK("hob: feat 0x%X nsect 0x%X, lba 0x%X 0x%X 0x%X\n",
+			tf->hob_feature,
+			tf->hob_nsect,
+			tf->hob_lbal,
+			tf->hob_lbam,
+			tf->hob_lbah);
+	}
+
+	if (is_addr) {
+		iowrite8(tf->feature, ioaddr->feature_addr);
+		iowrite8(tf->nsect, ioaddr->nsect_addr);
+		iowrite8(tf->lbal, ioaddr->lbal_addr);
+		iowrite8(tf->lbam, ioaddr->lbam_addr);
+		iowrite8(tf->lbah, ioaddr->lbah_addr);
+		VPRINTK("feat 0x%X nsect 0x%X lba 0x%X 0x%X 0x%X\n",
+			tf->feature,
+			tf->nsect,
+			tf->lbal,
+			tf->lbam,
+			tf->lbah);
+	}
+
+	if (tf->flags & ATA_TFLAG_DEVICE) {
+		iowrite8(tf->device, ioaddr->device_addr);
+		VPRINTK("device 0x%X\n", tf->device);
+	}
+
+	ata_wait_idle(ap);
+}
+
+/**
+ *	ata_tf_read - input device's ATA taskfile shadow registers
+ *	@ap: Port from which input is read
+ *	@tf: ATA taskfile register set for storing input
+ *
+ *	Reads ATA taskfile registers for currently-selected device
+ *	into @tf. Assumes the device has a fully SFF compliant task file
+ *	layout and behaviour. If you device does not (eg has a different
+ *	status method) then you will need to provide a replacement tf_read
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
+{
+	struct ata_ioports *ioaddr = &ap->ioaddr;
+
+	tf->command = ata_check_status(ap);
+	tf->feature = ioread8(ioaddr->error_addr);
+	tf->nsect = ioread8(ioaddr->nsect_addr);
+	tf->lbal = ioread8(ioaddr->lbal_addr);
+	tf->lbam = ioread8(ioaddr->lbam_addr);
+	tf->lbah = ioread8(ioaddr->lbah_addr);
+	tf->device = ioread8(ioaddr->device_addr);
+
+	if (tf->flags & ATA_TFLAG_LBA48) {
+		if (likely(ioaddr->ctl_addr)) {
+			iowrite8(tf->ctl | ATA_HOB, ioaddr->ctl_addr);
+			tf->hob_feature = ioread8(ioaddr->error_addr);
+			tf->hob_nsect = ioread8(ioaddr->nsect_addr);
+			tf->hob_lbal = ioread8(ioaddr->lbal_addr);
+			tf->hob_lbam = ioread8(ioaddr->lbam_addr);
+			tf->hob_lbah = ioread8(ioaddr->lbah_addr);
+			iowrite8(tf->ctl, ioaddr->ctl_addr);
+			ap->last_ctl = tf->ctl;
+		} else
+			WARN_ON(1);
+	}
+}
+
+/**
+ *	ata_exec_command - issue ATA command to host controller
+ *	@ap: port to which command is being issued
+ *	@tf: ATA taskfile register set
+ *
+ *	Issues ATA command, with proper synchronization with interrupt
+ *	handler / other threads.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ */
+void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf)
+{
+	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
+
+	iowrite8(tf->command, ap->ioaddr.command_addr);
+	ata_pause(ap);
+}
+
+/**
+ *	ata_tf_to_host - issue ATA taskfile to host controller
+ *	@ap: port to which command is being issued
+ *	@tf: ATA taskfile register set
+ *
+ *	Issues ATA taskfile register set to ATA host controller,
+ *	with proper synchronization with interrupt handler and
+ *	other threads.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ */
+static inline void ata_tf_to_host(struct ata_port *ap,
+				  const struct ata_taskfile *tf)
+{
+	ap->ops->tf_load(ap, tf);
+	ap->ops->exec_command(ap, tf);
+}
+
+/**
+ *	ata_data_xfer - Transfer data by PIO
+ *	@dev: device to target
+ *	@buf: data buffer
+ *	@buflen: buffer length
+ *	@rw: read/write
+ *
+ *	Transfer data from/to the device data register by PIO.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ *
+ *	RETURNS:
+ *	Bytes consumed.
+ */
+unsigned int ata_data_xfer(struct ata_device *dev, unsigned char *buf,
+			   unsigned int buflen, int rw)
+{
+	struct ata_port *ap = dev->link->ap;
+	void __iomem *data_addr = ap->ioaddr.data_addr;
+	unsigned int words = buflen >> 1;
+
+	/* Transfer multiple of 2 bytes */
+	if (rw == READ)
+		ioread16_rep(data_addr, buf, words);
+	else
+		iowrite16_rep(data_addr, buf, words);
+
+	/* Transfer trailing 1 byte, if any. */
+	if (unlikely(buflen & 0x01)) {
+		__le16 align_buf[1] = { 0 };
+		unsigned char *trailing_buf = buf + buflen - 1;
+
+		if (rw == READ) {
+			align_buf[0] = cpu_to_le16(ioread16(data_addr));
+			memcpy(trailing_buf, align_buf, 1);
+		} else {
+			memcpy(align_buf, trailing_buf, 1);
+			iowrite16(le16_to_cpu(align_buf[0]), data_addr);
+		}
+		words++;
+	}
+
+	return words << 1;
+}
+
+/**
+ *	ata_data_xfer_noirq - Transfer data by PIO
+ *	@dev: device to target
+ *	@buf: data buffer
+ *	@buflen: buffer length
+ *	@rw: read/write
+ *
+ *	Transfer data from/to the device data register by PIO. Do the
+ *	transfer with interrupts disabled.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ *
+ *	RETURNS:
+ *	Bytes consumed.
+ */
+unsigned int ata_data_xfer_noirq(struct ata_device *dev, unsigned char *buf,
+				 unsigned int buflen, int rw)
+{
+	unsigned long flags;
+	unsigned int consumed;
+
+	local_irq_save(flags);
+	consumed = ata_data_xfer(dev, buf, buflen, rw);
+	local_irq_restore(flags);
+
+	return consumed;
+}
+
+/**
+ *	ata_pio_sector - Transfer a sector of data.
+ *	@qc: Command on going
+ *
+ *	Transfer qc->sect_size bytes of data from/to the ATA device.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+static void ata_pio_sector(struct ata_queued_cmd *qc)
+{
+	int do_write = (qc->tf.flags & ATA_TFLAG_WRITE);
+	struct ata_port *ap = qc->ap;
+	struct page *page;
+	unsigned int offset;
+	unsigned char *buf;
+
+	if (qc->curbytes == qc->nbytes - qc->sect_size)
+		ap->hsm_task_state = HSM_ST_LAST;
+
+	page = sg_page(qc->cursg);
+	offset = qc->cursg->offset + qc->cursg_ofs;
+
+	/* get the current page and offset */
+	page = nth_page(page, (offset >> PAGE_SHIFT));
+	offset %= PAGE_SIZE;
+
+	DPRINTK("data %s\n", qc->tf.flags & ATA_TFLAG_WRITE ? "write" : "read");
+
+	if (PageHighMem(page)) {
+		unsigned long flags;
+
+		/* FIXME: use a bounce buffer */
+		local_irq_save(flags);
+		buf = kmap_atomic(page, KM_IRQ0);
+
+		/* do the actual data transfer */
+		ap->ops->data_xfer(qc->dev, buf + offset, qc->sect_size, do_write);
+
+		kunmap_atomic(buf, KM_IRQ0);
+		local_irq_restore(flags);
+	} else {
+		buf = page_address(page);
+		ap->ops->data_xfer(qc->dev, buf + offset, qc->sect_size, do_write);
+	}
+
+	qc->curbytes += qc->sect_size;
+	qc->cursg_ofs += qc->sect_size;
+
+	if (qc->cursg_ofs == qc->cursg->length) {
+		qc->cursg = sg_next(qc->cursg);
+		qc->cursg_ofs = 0;
+	}
+}
+
+/**
+ *	ata_pio_sectors - Transfer one or many sectors.
+ *	@qc: Command on going
+ *
+ *	Transfer one or many sectors of data from/to the
+ *	ATA device for the DRQ request.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+static void ata_pio_sectors(struct ata_queued_cmd *qc)
+{
+	if (is_multi_taskfile(&qc->tf)) {
+		/* READ/WRITE MULTIPLE */
+		unsigned int nsect;
+
+		WARN_ON(qc->dev->multi_count == 0);
+
+		nsect = min((qc->nbytes - qc->curbytes) / qc->sect_size,
+			    qc->dev->multi_count);
+		while (nsect--)
+			ata_pio_sector(qc);
+	} else
+		ata_pio_sector(qc);
+
+	ata_altstatus(qc->ap); /* flush */
+}
+
+/**
+ *	atapi_send_cdb - Write CDB bytes to hardware
+ *	@ap: Port to which ATAPI device is attached.
+ *	@qc: Taskfile currently active
+ *
+ *	When device has indicated its readiness to accept
+ *	a CDB, this function is called.  Send the CDB.
+ *
+ *	LOCKING:
+ *	caller.
+ */
+static void atapi_send_cdb(struct ata_port *ap, struct ata_queued_cmd *qc)
+{
+	/* send SCSI cdb */
+	DPRINTK("send cdb\n");
+	WARN_ON(qc->dev->cdb_len < 12);
+
+	ap->ops->data_xfer(qc->dev, qc->cdb, qc->dev->cdb_len, 1);
+	ata_altstatus(ap); /* flush */
+
+	switch (qc->tf.protocol) {
+	case ATAPI_PROT_PIO:
+		ap->hsm_task_state = HSM_ST;
+		break;
+	case ATAPI_PROT_NODATA:
+		ap->hsm_task_state = HSM_ST_LAST;
+		break;
+	case ATAPI_PROT_DMA:
+		ap->hsm_task_state = HSM_ST_LAST;
+		/* initiate bmdma */
+		ap->ops->bmdma_start(qc);
+		break;
+	}
+}
+
+/**
+ *	__atapi_pio_bytes - Transfer data from/to the ATAPI device.
+ *	@qc: Command on going
+ *	@bytes: number of bytes
+ *
+ *	Transfer Transfer data from/to the ATAPI device.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ *
+ */
+static int __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes)
+{
+	int rw = (qc->tf.flags & ATA_TFLAG_WRITE) ? WRITE : READ;
+	struct ata_port *ap = qc->ap;
+	struct ata_device *dev = qc->dev;
+	struct ata_eh_info *ehi = &dev->link->eh_info;
+	struct scatterlist *sg;
+	struct page *page;
+	unsigned char *buf;
+	unsigned int offset, count, consumed;
+
+next_sg:
+	sg = qc->cursg;
+	if (unlikely(!sg)) {
+		ata_ehi_push_desc(ehi, "unexpected or too much trailing data "
+				  "buf=%u cur=%u bytes=%u",
+				  qc->nbytes, qc->curbytes, bytes);
+		return -1;
+	}
+
+	page = sg_page(sg);
+	offset = sg->offset + qc->cursg_ofs;
+
+	/* get the current page and offset */
+	page = nth_page(page, (offset >> PAGE_SHIFT));
+	offset %= PAGE_SIZE;
+
+	/* don't overrun current sg */
+	count = min(sg->length - qc->cursg_ofs, bytes);
+
+	/* don't cross page boundaries */
+	count = min(count, (unsigned int)PAGE_SIZE - offset);
+
+	DPRINTK("data %s\n", qc->tf.flags & ATA_TFLAG_WRITE ? "write" : "read");
+
+	if (PageHighMem(page)) {
+		unsigned long flags;
+
+		/* FIXME: use bounce buffer */
+		local_irq_save(flags);
+		buf = kmap_atomic(page, KM_IRQ0);
+
+		/* do the actual data transfer */
+		consumed = ap->ops->data_xfer(dev,  buf + offset, count, rw);
+
+		kunmap_atomic(buf, KM_IRQ0);
+		local_irq_restore(flags);
+	} else {
+		buf = page_address(page);
+		consumed = ap->ops->data_xfer(dev,  buf + offset, count, rw);
+	}
+
+	bytes -= min(bytes, consumed);
+	qc->curbytes += count;
+	qc->cursg_ofs += count;
+
+	if (qc->cursg_ofs == sg->length) {
+		qc->cursg = sg_next(qc->cursg);
+		qc->cursg_ofs = 0;
+	}
+
+	/* consumed can be larger than count only for the last transfer */
+	WARN_ON(qc->cursg && count != consumed);
+
+	if (bytes)
+		goto next_sg;
+	return 0;
+}
+
+/**
+ *	atapi_pio_bytes - Transfer data from/to the ATAPI device.
+ *	@qc: Command on going
+ *
+ *	Transfer Transfer data from/to the ATAPI device.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+static void atapi_pio_bytes(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+	struct ata_device *dev = qc->dev;
+	struct ata_eh_info *ehi = &dev->link->eh_info;
+	unsigned int ireason, bc_lo, bc_hi, bytes;
+	int i_write, do_write = (qc->tf.flags & ATA_TFLAG_WRITE) ? 1 : 0;
+
+	/* Abuse qc->result_tf for temp storage of intermediate TF
+	 * here to save some kernel stack usage.
+	 * For normal completion, qc->result_tf is not relevant. For
+	 * error, qc->result_tf is later overwritten by ata_qc_complete().
+	 * So, the correctness of qc->result_tf is not affected.
+	 */
+	ap->ops->tf_read(ap, &qc->result_tf);
+	ireason = qc->result_tf.nsect;
+	bc_lo = qc->result_tf.lbam;
+	bc_hi = qc->result_tf.lbah;
+	bytes = (bc_hi << 8) | bc_lo;
+
+	/* shall be cleared to zero, indicating xfer of data */
+	if (unlikely(ireason & (1 << 0)))
+		goto atapi_check;
+
+	/* make sure transfer direction matches expected */
+	i_write = ((ireason & (1 << 1)) == 0) ? 1 : 0;
+	if (unlikely(do_write != i_write))
+		goto atapi_check;
+
+	if (unlikely(!bytes))
+		goto atapi_check;
+
+	VPRINTK("ata%u: xfering %d bytes\n", ap->print_id, bytes);
+
+	if (unlikely(__atapi_pio_bytes(qc, bytes)))
+		goto err_out;
+	ata_altstatus(ap); /* flush */
+
+	return;
+
+ atapi_check:
+	ata_ehi_push_desc(ehi, "ATAPI check failed (ireason=0x%x bytes=%u)",
+			  ireason, bytes);
+ err_out:
+	qc->err_mask |= AC_ERR_HSM;
+	ap->hsm_task_state = HSM_ST_ERR;
+}
+
+/**
+ *	ata_hsm_ok_in_wq - Check if the qc can be handled in the workqueue.
+ *	@ap: the target ata_port
+ *	@qc: qc on going
+ *
+ *	RETURNS:
+ *	1 if ok in workqueue, 0 otherwise.
+ */
+static inline int ata_hsm_ok_in_wq(struct ata_port *ap, struct ata_queued_cmd *qc)
+{
+	if (qc->tf.flags & ATA_TFLAG_POLLING)
+		return 1;
+
+	if (ap->hsm_task_state == HSM_ST_FIRST) {
+		if (qc->tf.protocol == ATA_PROT_PIO &&
+		    (qc->tf.flags & ATA_TFLAG_WRITE))
+		    return 1;
+
+		if (ata_is_atapi(qc->tf.protocol) &&
+		    !(qc->dev->flags & ATA_DFLAG_CDB_INTR))
+			return 1;
+	}
+
+	return 0;
+}
+
+/**
+ *	ata_hsm_qc_complete - finish a qc running on standard HSM
+ *	@qc: Command to complete
+ *	@in_wq: 1 if called from workqueue, 0 otherwise
+ *
+ *	Finish @qc which is running on standard HSM.
+ *
+ *	LOCKING:
+ *	If @in_wq is zero, spin_lock_irqsave(host lock).
+ *	Otherwise, none on entry and grabs host lock.
+ */
+static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
+{
+	struct ata_port *ap = qc->ap;
+	unsigned long flags;
+
+	if (ap->ops->error_handler) {
+		if (in_wq) {
+			spin_lock_irqsave(ap->lock, flags);
+
+			/* EH might have kicked in while host lock is
+			 * released.
+			 */
+			qc = ata_qc_from_tag(ap, qc->tag);
+			if (qc) {
+				if (likely(!(qc->err_mask & AC_ERR_HSM))) {
+					ap->ops->irq_on(ap);
+					ata_qc_complete(qc);
+				} else
+					ata_port_freeze(ap);
+			}
+
+			spin_unlock_irqrestore(ap->lock, flags);
+		} else {
+			if (likely(!(qc->err_mask & AC_ERR_HSM)))
+				ata_qc_complete(qc);
+			else
+				ata_port_freeze(ap);
+		}
+	} else {
+		if (in_wq) {
+			spin_lock_irqsave(ap->lock, flags);
+			ap->ops->irq_on(ap);
+			ata_qc_complete(qc);
+			spin_unlock_irqrestore(ap->lock, flags);
+		} else
+			ata_qc_complete(qc);
+	}
+}
+
+/**
+ *	ata_hsm_move - move the HSM to the next state.
+ *	@ap: the target ata_port
+ *	@qc: qc on going
+ *	@status: current device status
+ *	@in_wq: 1 if called from workqueue, 0 otherwise
+ *
+ *	RETURNS:
+ *	1 when poll next status needed, 0 otherwise.
+ */
+int ata_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
+		 u8 status, int in_wq)
+{
+	unsigned long flags = 0;
+	int poll_next;
+
+	WARN_ON((qc->flags & ATA_QCFLAG_ACTIVE) == 0);
+
+	/* Make sure ata_qc_issue_prot() does not throw things
+	 * like DMA polling into the workqueue. Notice that
+	 * in_wq is not equivalent to (qc->tf.flags & ATA_TFLAG_POLLING).
+	 */
+	WARN_ON(in_wq != ata_hsm_ok_in_wq(ap, qc));
+
+fsm_start:
+	DPRINTK("ata%u: protocol %d task_state %d (dev_stat 0x%X)\n",
+		ap->print_id, qc->tf.protocol, ap->hsm_task_state, status);
+
+	switch (ap->hsm_task_state) {
+	case HSM_ST_FIRST:
+		/* Send first data block or PACKET CDB */
+
+		/* If polling, we will stay in the work queue after
+		 * sending the data. Otherwise, interrupt handler
+		 * takes over after sending the data.
+		 */
+		poll_next = (qc->tf.flags & ATA_TFLAG_POLLING);
+
+		/* check device status */
+		if (unlikely((status & ATA_DRQ) == 0)) {
+			/* handle BSY=0, DRQ=0 as error */
+			if (likely(status & (ATA_ERR | ATA_DF)))
+				/* device stops HSM for abort/error */
+				qc->err_mask |= AC_ERR_DEV;
+			else
+				/* HSM violation. Let EH handle this */
+				qc->err_mask |= AC_ERR_HSM;
+
+			ap->hsm_task_state = HSM_ST_ERR;
+			goto fsm_start;
+		}
+
+		/* Device should not ask for data transfer (DRQ=1)
+		 * when it finds something wrong.
+		 * We ignore DRQ here and stop the HSM by
+		 * changing hsm_task_state to HSM_ST_ERR and
+		 * let the EH abort the command or reset the device.
+		 */
+		if (unlikely(status & (ATA_ERR | ATA_DF))) {
+			/* Some ATAPI tape drives forget to clear the ERR bit
+			 * when doing the next command (mostly request sense).
+			 * We ignore ERR here to workaround and proceed sending
+			 * the CDB.
+			 */
+			if (!(qc->dev->horkage & ATA_HORKAGE_STUCK_ERR)) {
+				ata_port_printk(ap, KERN_WARNING,
+						"DRQ=1 with device error, "
+						"dev_stat 0x%X\n", status);
+				qc->err_mask |= AC_ERR_HSM;
+				ap->hsm_task_state = HSM_ST_ERR;
+				goto fsm_start;
+			}
+		}
+
+		/* Send the CDB (atapi) or the first data block (ata pio out).
+		 * During the state transition, interrupt handler shouldn't
+		 * be invoked before the data transfer is complete and
+		 * hsm_task_state is changed. Hence, the following locking.
+		 */
+		if (in_wq)
+			spin_lock_irqsave(ap->lock, flags);
+
+		if (qc->tf.protocol == ATA_PROT_PIO) {
+			/* PIO data out protocol.
+			 * send first data block.
+			 */
+
+			/* ata_pio_sectors() might change the state
+			 * to HSM_ST_LAST. so, the state is changed here
+			 * before ata_pio_sectors().
+			 */
+			ap->hsm_task_state = HSM_ST;
+			ata_pio_sectors(qc);
+		} else
+			/* send CDB */
+			atapi_send_cdb(ap, qc);
+
+		if (in_wq)
+			spin_unlock_irqrestore(ap->lock, flags);
+
+		/* if polling, ata_pio_task() handles the rest.
+		 * otherwise, interrupt handler takes over from here.
+		 */
+		break;
+
+	case HSM_ST:
+		/* complete command or read/write the data register */
+		if (qc->tf.protocol == ATAPI_PROT_PIO) {
+			/* ATAPI PIO protocol */
+			if ((status & ATA_DRQ) == 0) {
+				/* No more data to transfer or device error.
+				 * Device error will be tagged in HSM_ST_LAST.
+				 */
+				ap->hsm_task_state = HSM_ST_LAST;
+				goto fsm_start;
+			}
+
+			/* Device should not ask for data transfer (DRQ=1)
+			 * when it finds something wrong.
+			 * We ignore DRQ here and stop the HSM by
+			 * changing hsm_task_state to HSM_ST_ERR and
+			 * let the EH abort the command or reset the device.
+			 */
+			if (unlikely(status & (ATA_ERR | ATA_DF))) {
+				ata_port_printk(ap, KERN_WARNING, "DRQ=1 with "
+						"device error, dev_stat 0x%X\n",
+						status);
+				qc->err_mask |= AC_ERR_HSM;
+				ap->hsm_task_state = HSM_ST_ERR;
+				goto fsm_start;
+			}
+
+			atapi_pio_bytes(qc);
+
+			if (unlikely(ap->hsm_task_state == HSM_ST_ERR))
+				/* bad ireason reported by device */
+				goto fsm_start;
+
+		} else {
+			/* ATA PIO protocol */
+			if (unlikely((status & ATA_DRQ) == 0)) {
+				/* handle BSY=0, DRQ=0 as error */
+				if (likely(status & (ATA_ERR | ATA_DF)))
+					/* device stops HSM for abort/error */
+					qc->err_mask |= AC_ERR_DEV;
+				else
+					/* HSM violation. Let EH handle this.
+					 * Phantom devices also trigger this
+					 * condition.  Mark hint.
+					 */
+					qc->err_mask |= AC_ERR_HSM |
+							AC_ERR_NODEV_HINT;
+
+				ap->hsm_task_state = HSM_ST_ERR;
+				goto fsm_start;
+			}
+
+			/* For PIO reads, some devices may ask for
+			 * data transfer (DRQ=1) alone with ERR=1.
+			 * We respect DRQ here and transfer one
+			 * block of junk data before changing the
+			 * hsm_task_state to HSM_ST_ERR.
+			 *
+			 * For PIO writes, ERR=1 DRQ=1 doesn't make
+			 * sense since the data block has been
+			 * transferred to the device.
+			 */
+			if (unlikely(status & (ATA_ERR | ATA_DF))) {
+				/* data might be corrputed */
+				qc->err_mask |= AC_ERR_DEV;
+
+				if (!(qc->tf.flags & ATA_TFLAG_WRITE)) {
+					ata_pio_sectors(qc);
+					status = ata_wait_idle(ap);
+				}
+
+				if (status & (ATA_BUSY | ATA_DRQ))
+					qc->err_mask |= AC_ERR_HSM;
+
+				/* ata_pio_sectors() might change the
+				 * state to HSM_ST_LAST. so, the state
+				 * is changed after ata_pio_sectors().
+				 */
+				ap->hsm_task_state = HSM_ST_ERR;
+				goto fsm_start;
+			}
+
+			ata_pio_sectors(qc);
+
+			if (ap->hsm_task_state == HSM_ST_LAST &&
+			    (!(qc->tf.flags & ATA_TFLAG_WRITE))) {
+				/* all data read */
+				status = ata_wait_idle(ap);
+				goto fsm_start;
+			}
+		}
+
+		poll_next = 1;
+		break;
+
+	case HSM_ST_LAST:
+		if (unlikely(!ata_ok(status))) {
+			qc->err_mask |= __ac_err_mask(status);
+			ap->hsm_task_state = HSM_ST_ERR;
+			goto fsm_start;
+		}
+
+		/* no more data to transfer */
+		DPRINTK("ata%u: dev %u command complete, drv_stat 0x%x\n",
+			ap->print_id, qc->dev->devno, status);
+
+		WARN_ON(qc->err_mask);
+
+		ap->hsm_task_state = HSM_ST_IDLE;
+
+		/* complete taskfile transaction */
+		ata_hsm_qc_complete(qc, in_wq);
+
+		poll_next = 0;
+		break;
+
+	case HSM_ST_ERR:
+		/* make sure qc->err_mask is available to
+		 * know what's wrong and recover
+		 */
+		WARN_ON(qc->err_mask == 0);
+
+		ap->hsm_task_state = HSM_ST_IDLE;
+
+		/* complete taskfile transaction */
+		ata_hsm_qc_complete(qc, in_wq);
+
+		poll_next = 0;
+		break;
+	default:
+		poll_next = 0;
+		BUG();
+	}
+
+	return poll_next;
+}
+
+void ata_pio_task(struct work_struct *work)
+{
+	struct ata_port *ap =
+		container_of(work, struct ata_port, port_task.work);
+	struct ata_queued_cmd *qc = ap->port_task_data;
+	u8 status;
+	int poll_next;
+
+fsm_start:
+	WARN_ON(ap->hsm_task_state == HSM_ST_IDLE);
+
+	/*
+	 * This is purely heuristic.  This is a fast path.
+	 * Sometimes when we enter, BSY will be cleared in
+	 * a chk-status or two.  If not, the drive is probably seeking
+	 * or something.  Snooze for a couple msecs, then
+	 * chk-status again.  If still busy, queue delayed work.
+	 */
+	status = ata_busy_wait(ap, ATA_BUSY, 5);
+	if (status & ATA_BUSY) {
+		msleep(2);
+		status = ata_busy_wait(ap, ATA_BUSY, 10);
+		if (status & ATA_BUSY) {
+			ata_pio_queue_task(ap, qc, ATA_SHORT_PAUSE);
+			return;
+		}
+	}
+
+	/* move the HSM */
+	poll_next = ata_hsm_move(ap, qc, status, 1);
+
+	/* another command or interrupt handler
+	 * may be running at this point.
+	 */
+	if (poll_next)
+		goto fsm_start;
+}
+
+/**
+ *	ata_qc_issue_prot - issue taskfile to device in proto-dependent manner
+ *	@qc: command to issue to device
+ *
+ *	Using various libata functions and hooks, this function
+ *	starts an ATA command.  ATA commands are grouped into
+ *	classes called "protocols", and issuing each type of protocol
+ *	is slightly different.
+ *
+ *	May be used as the qc_issue() entry in ata_port_operations.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ *
+ *	RETURNS:
+ *	Zero on success, AC_ERR_* mask on failure
+ */
+unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc)
+{
+	struct ata_port *ap = qc->ap;
+
+	/* Use polling pio if the LLD doesn't handle
+	 * interrupt driven pio and atapi CDB interrupt.
+	 */
+	if (ap->flags & ATA_FLAG_PIO_POLLING) {
+		switch (qc->tf.protocol) {
+		case ATA_PROT_PIO:
+		case ATA_PROT_NODATA:
+		case ATAPI_PROT_PIO:
+		case ATAPI_PROT_NODATA:
+			qc->tf.flags |= ATA_TFLAG_POLLING;
+			break;
+		case ATAPI_PROT_DMA:
+			if (qc->dev->flags & ATA_DFLAG_CDB_INTR)
+				/* see ata_dma_blacklisted() */
+				BUG();
+			break;
+		default:
+			break;
+		}
+	}
+
+	/* select the device */
+	ata_dev_select(ap, qc->dev->devno, 1, 0);
+
+	/* start the command */
+	switch (qc->tf.protocol) {
+	case ATA_PROT_NODATA:
+		if (qc->tf.flags & ATA_TFLAG_POLLING)
+			ata_qc_set_polling(qc);
+
+		ata_tf_to_host(ap, &qc->tf);
+		ap->hsm_task_state = HSM_ST_LAST;
+
+		if (qc->tf.flags & ATA_TFLAG_POLLING)
+			ata_pio_queue_task(ap, qc, 0);
+
+		break;
+
+	case ATA_PROT_DMA:
+		WARN_ON(qc->tf.flags & ATA_TFLAG_POLLING);
+
+		ap->ops->tf_load(ap, &qc->tf);	 /* load tf registers */
+		ap->ops->bmdma_setup(qc);	    /* set up bmdma */
+		ap->ops->bmdma_start(qc);	    /* initiate bmdma */
+		ap->hsm_task_state = HSM_ST_LAST;
+		break;
+
+	case ATA_PROT_PIO:
+		if (qc->tf.flags & ATA_TFLAG_POLLING)
+			ata_qc_set_polling(qc);
+
+		ata_tf_to_host(ap, &qc->tf);
+
+		if (qc->tf.flags & ATA_TFLAG_WRITE) {
+			/* PIO data out protocol */
+			ap->hsm_task_state = HSM_ST_FIRST;
+			ata_pio_queue_task(ap, qc, 0);
+
+			/* always send first data block using
+			 * the ata_pio_task() codepath.
+			 */
+		} else {
+			/* PIO data in protocol */
+			ap->hsm_task_state = HSM_ST;
+
+			if (qc->tf.flags & ATA_TFLAG_POLLING)
+				ata_pio_queue_task(ap, qc, 0);
+
+			/* if polling, ata_pio_task() handles the rest.
+			 * otherwise, interrupt handler takes over from here.
+			 */
+		}
+
+		break;
+
+	case ATAPI_PROT_PIO:
+	case ATAPI_PROT_NODATA:
+		if (qc->tf.flags & ATA_TFLAG_POLLING)
+			ata_qc_set_polling(qc);
+
+		ata_tf_to_host(ap, &qc->tf);
+
+		ap->hsm_task_state = HSM_ST_FIRST;
+
+		/* send cdb by polling if no cdb interrupt */
+		if ((!(qc->dev->flags & ATA_DFLAG_CDB_INTR)) ||
+		    (qc->tf.flags & ATA_TFLAG_POLLING))
+			ata_pio_queue_task(ap, qc, 0);
+		break;
+
+	case ATAPI_PROT_DMA:
+		WARN_ON(qc->tf.flags & ATA_TFLAG_POLLING);
+
+		ap->ops->tf_load(ap, &qc->tf);	 /* load tf registers */
+		ap->ops->bmdma_setup(qc);	    /* set up bmdma */
+		ap->hsm_task_state = HSM_ST_FIRST;
+
+		/* send cdb by polling if no cdb interrupt */
+		if (!(qc->dev->flags & ATA_DFLAG_CDB_INTR))
+			ata_pio_queue_task(ap, qc, 0);
+		break;
+
+	default:
+		WARN_ON(1);
+		return AC_ERR_SYSTEM;
+	}
+
+	return 0;
+}
+
+/**
+ *	ata_host_intr - Handle host interrupt for given (port, task)
+ *	@ap: Port on which interrupt arrived (possibly...)
+ *	@qc: Taskfile currently active in engine
+ *
+ *	Handle host interrupt for given queued command.  Currently,
+ *	only DMA interrupts are handled.  All other commands are
+ *	handled via polling with interrupts disabled (nIEN bit).
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ *
+ *	RETURNS:
+ *	One if interrupt was handled, zero if not (shared irq).
+ */
+inline unsigned int ata_host_intr(struct ata_port *ap,
+				  struct ata_queued_cmd *qc)
+{
+	struct ata_eh_info *ehi = &ap->link.eh_info;
+	u8 status, host_stat = 0;
+
+	VPRINTK("ata%u: protocol %d task_state %d\n",
+		ap->print_id, qc->tf.protocol, ap->hsm_task_state);
+
+	/* Check whether we are expecting interrupt in this state */
+	switch (ap->hsm_task_state) {
+	case HSM_ST_FIRST:
+		/* Some pre-ATAPI-4 devices assert INTRQ
+		 * at this state when ready to receive CDB.
+		 */
+
+		/* Check the ATA_DFLAG_CDB_INTR flag is enough here.
+		 * The flag was turned on only for atapi devices.  No
+		 * need to check ata_is_atapi(qc->tf.protocol) again.
+		 */
+		if (!(qc->dev->flags & ATA_DFLAG_CDB_INTR))
+			goto idle_irq;
+		break;
+	case HSM_ST_LAST:
+		if (qc->tf.protocol == ATA_PROT_DMA ||
+		    qc->tf.protocol == ATAPI_PROT_DMA) {
+			/* check status of DMA engine */
+			host_stat = ap->ops->bmdma_status(ap);
+			VPRINTK("ata%u: host_stat 0x%X\n",
+				ap->print_id, host_stat);
+
+			/* if it's not our irq... */
+			if (!(host_stat & ATA_DMA_INTR))
+				goto idle_irq;
+
+			/* before we do anything else, clear DMA-Start bit */
+			ap->ops->bmdma_stop(qc);
+
+			if (unlikely(host_stat & ATA_DMA_ERR)) {
+				/* error when transfering data to/from memory */
+				qc->err_mask |= AC_ERR_HOST_BUS;
+				ap->hsm_task_state = HSM_ST_ERR;
+			}
+		}
+		break;
+	case HSM_ST:
+		break;
+	default:
+		goto idle_irq;
+	}
+
+	/* check altstatus */
+	status = ata_altstatus(ap);
+	if (status & ATA_BUSY)
+		goto idle_irq;
+
+	/* check main status, clearing INTRQ */
+	status = ata_chk_status(ap);
+	if (unlikely(status & ATA_BUSY))
+		goto idle_irq;
+
+	/* ack bmdma irq events */
+	ap->ops->irq_clear(ap);
+
+	ata_hsm_move(ap, qc, status, 0);
+
+	if (unlikely(qc->err_mask) && (qc->tf.protocol == ATA_PROT_DMA ||
+				       qc->tf.protocol == ATAPI_PROT_DMA))
+		ata_ehi_push_desc(ehi, "BMDMA stat 0x%x", host_stat);
+
+	return 1;	/* irq handled */
+
+idle_irq:
+	ap->stats.idle_irq++;
+
+#ifdef ATA_IRQ_TRAP
+	if ((ap->stats.idle_irq % 1000) == 0) {
+		ata_chk_status(ap);
+		ap->ops->irq_clear(ap);
+		ata_port_printk(ap, KERN_WARNING, "irq trap\n");
+		return 1;
+	}
+#endif
+	return 0;	/* irq not handled */
+}
+
+/**
+ *	ata_interrupt - Default ATA host interrupt handler
+ *	@irq: irq line (unused)
+ *	@dev_instance: pointer to our ata_host information structure
+ *
+ *	Default interrupt handler for PCI IDE devices.  Calls
+ *	ata_host_intr() for each port that is not disabled.
+ *
+ *	LOCKING:
+ *	Obtains host lock during operation.
+ *
+ *	RETURNS:
+ *	IRQ_NONE or IRQ_HANDLED.
+ */
+irqreturn_t ata_interrupt(int irq, void *dev_instance)
+{
+	struct ata_host *host = dev_instance;
+	unsigned int i;
+	unsigned int handled = 0;
+	unsigned long flags;
+
+	/* TODO: make _irqsave conditional on x86 PCI IDE legacy mode */
+	spin_lock_irqsave(&host->lock, flags);
+
+	for (i = 0; i < host->n_ports; i++) {
+		struct ata_port *ap;
+
+		ap = host->ports[i];
+		if (ap &&
+		    !(ap->flags & ATA_FLAG_DISABLED)) {
+			struct ata_queued_cmd *qc;
+
+			qc = ata_qc_from_tag(ap, ap->link.active_tag);
+			if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)) &&
+			    (qc->flags & ATA_QCFLAG_ACTIVE))
+				handled |= ata_host_intr(ap, qc);
+		}
+	}
+
+	spin_unlock_irqrestore(&host->lock, flags);
+
+	return IRQ_RETVAL(handled);
+}
+
+/**
+ *	ata_bmdma_freeze - Freeze BMDMA controller port
+ *	@ap: port to freeze
+ *
+ *	Freeze BMDMA controller port.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+void ata_bmdma_freeze(struct ata_port *ap)
+{
+	struct ata_ioports *ioaddr = &ap->ioaddr;
+
+	ap->ctl |= ATA_NIEN;
+	ap->last_ctl = ap->ctl;
+
+	if (ioaddr->ctl_addr)
+		iowrite8(ap->ctl, ioaddr->ctl_addr);
+
+	/* Under certain circumstances, some controllers raise IRQ on
+	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
+	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
+	 */
+	ata_chk_status(ap);
+
+	ap->ops->irq_clear(ap);
+}
+
+/**
+ *	ata_bmdma_thaw - Thaw BMDMA controller port
+ *	@ap: port to thaw
+ *
+ *	Thaw BMDMA controller port.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+void ata_bmdma_thaw(struct ata_port *ap)
 {
-	if (ap->ops->check_altstatus)
-		return ap->ops->check_altstatus(ap);
-
-	return ioread8(ap->ioaddr.altstatus_addr);
+	/* clear & re-enable interrupts */
+	ata_chk_status(ap);
+	ap->ops->irq_clear(ap);
+	ap->ops->irq_on(ap);
 }
 
 /**
- *	ata_irq_on - Enable interrupts on a port.
- *	@ap: Port on which interrupts are enabled.
+ *	ata_devchk - PATA device presence detection
+ *	@ap: ATA channel to examine
+ *	@device: Device to examine (starting at zero)
  *
- *	Enable interrupts on a legacy IDE device using MMIO or PIO,
- *	wait for idle, clear any pending interrupts.
+ *	This technique was originally described in
+ *	Hale Landis's ATADRVR (www.ata-atapi.com), and
+ *	later found its way into the ATA/ATAPI spec.
+ *
+ *	Write a pattern to the ATA shadow registers,
+ *	and if a device is present, it will respond by
+ *	correctly storing and echoing back the
+ *	ATA shadow register contents.
  *
  *	LOCKING:
- *	Inherited from caller.
+ *	caller.
  */
-u8 ata_irq_on(struct ata_port *ap)
+static unsigned int ata_devchk(struct ata_port *ap, unsigned int device)
 {
 	struct ata_ioports *ioaddr = &ap->ioaddr;
-	u8 tmp;
+	u8 nsect, lbal;
 
-	ap->ctl &= ~ATA_NIEN;
-	ap->last_ctl = ap->ctl;
+	ap->ops->dev_select(ap, device);
 
-	if (ioaddr->ctl_addr)
-		iowrite8(ap->ctl, ioaddr->ctl_addr);
-	tmp = ata_wait_idle(ap);
+	iowrite8(0x55, ioaddr->nsect_addr);
+	iowrite8(0xaa, ioaddr->lbal_addr);
 
-	ap->ops->irq_clear(ap);
+	iowrite8(0xaa, ioaddr->nsect_addr);
+	iowrite8(0x55, ioaddr->lbal_addr);
 
-	return tmp;
+	iowrite8(0x55, ioaddr->nsect_addr);
+	iowrite8(0xaa, ioaddr->lbal_addr);
+
+	nsect = ioread8(ioaddr->nsect_addr);
+	lbal = ioread8(ioaddr->lbal_addr);
+
+	if ((nsect == 0x55) && (lbal == 0xaa))
+		return 1;	/* we found a device */
+
+	return 0;		/* nothing found */
 }
 
 /**
- *	ata_bmdma_irq_clear - Clear PCI IDE BMDMA interrupt.
- *	@ap: Port associated with this ATA transaction.
+ *	ata_dev_try_classify - Parse returned ATA device signature
+ *	@dev: ATA device to classify (starting at zero)
+ *	@present: device seems present
+ *	@r_err: Value of error register on completion
  *
- *	Clear interrupt and error flags in DMA status register.
+ *	After an event -- SRST, E.D.D., or SATA COMRESET -- occurs,
+ *	an ATA/ATAPI-defined set of values is placed in the ATA
+ *	shadow registers, indicating the results of device detection
+ *	and diagnostics.
  *
- *	May be used as the irq_clear() entry in ata_port_operations.
+ *	Select the ATA device, and read the values from the ATA shadow
+ *	registers.  Then parse according to the Error register value,
+ *	and the spec-defined values examined by ata_dev_classify().
  *
  *	LOCKING:
- *	spin_lock_irqsave(host lock)
+ *	caller.
+ *
+ *	RETURNS:
+ *	Device type - %ATA_DEV_ATA, %ATA_DEV_ATAPI or %ATA_DEV_NONE.
  */
-void ata_bmdma_irq_clear(struct ata_port *ap)
+unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
+				  u8 *r_err)
 {
-	void __iomem *mmio = ap->ioaddr.bmdma_addr;
+	struct ata_port *ap = dev->link->ap;
+	struct ata_taskfile tf;
+	unsigned int class;
+	u8 err;
+
+	ap->ops->dev_select(ap, dev->devno);
+
+	memset(&tf, 0, sizeof(tf));
+
+	ap->ops->tf_read(ap, &tf);
+	err = tf.feature;
+	if (r_err)
+		*r_err = err;
+
+	/* see if device passed diags: continue and warn later */
+	if (err == 0)
+		/* diagnostic fail : do nothing _YET_ */
+		dev->horkage |= ATA_HORKAGE_DIAGNOSTIC;
+	else if (err == 1)
+		/* do nothing */ ;
+	else if ((dev->devno == 0) && (err == 0x81))
+		/* do nothing */ ;
+	else
+		return ATA_DEV_NONE;
 
-	if (!mmio)
-		return;
+	/* determine if device is ATA or ATAPI */
+	class = ata_dev_classify(&tf);
 
-	iowrite8(ioread8(mmio + ATA_DMA_STATUS), mmio + ATA_DMA_STATUS);
+	if (class == ATA_DEV_UNKNOWN) {
+		/* If the device failed diagnostic, it's likely to
+		 * have reported incorrect device signature too.
+		 * Assume ATA device if the device seems present but
+		 * device signature is invalid with diagnostic
+		 * failure.
+		 */
+		if (present && (dev->horkage & ATA_HORKAGE_DIAGNOSTIC))
+			class = ATA_DEV_ATA;
+		else
+			class = ATA_DEV_NONE;
+	} else if ((class == ATA_DEV_ATA) && (ata_chk_status(ap) == 0))
+		class = ATA_DEV_NONE;
+
+	return class;
 }
 
-/**
- *	ata_tf_load - send taskfile registers to host controller
- *	@ap: Port to which output is sent
- *	@tf: ATA taskfile register set
- *
- *	Outputs ATA taskfile to standard ATA host controller.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
+static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
+			      unsigned long deadline)
 {
 	struct ata_ioports *ioaddr = &ap->ioaddr;
-	unsigned int is_addr = tf->flags & ATA_TFLAG_ISADDR;
+	unsigned int dev0 = devmask & (1 << 0);
+	unsigned int dev1 = devmask & (1 << 1);
+	int rc, ret = 0;
 
-	if (tf->ctl != ap->last_ctl) {
-		if (ioaddr->ctl_addr)
-			iowrite8(tf->ctl, ioaddr->ctl_addr);
-		ap->last_ctl = tf->ctl;
-		ata_wait_idle(ap);
+	/* if device 0 was found in ata_devchk, wait for its
+	 * BSY bit to clear
+	 */
+	if (dev0) {
+		rc = ata_wait_ready(ap, deadline);
+		if (rc) {
+			if (rc != -ENODEV)
+				return rc;
+			ret = rc;
+		}
 	}
 
-	if (is_addr && (tf->flags & ATA_TFLAG_LBA48)) {
-		WARN_ON(!ioaddr->ctl_addr);
-		iowrite8(tf->hob_feature, ioaddr->feature_addr);
-		iowrite8(tf->hob_nsect, ioaddr->nsect_addr);
-		iowrite8(tf->hob_lbal, ioaddr->lbal_addr);
-		iowrite8(tf->hob_lbam, ioaddr->lbam_addr);
-		iowrite8(tf->hob_lbah, ioaddr->lbah_addr);
-		VPRINTK("hob: feat 0x%X nsect 0x%X, lba 0x%X 0x%X 0x%X\n",
-			tf->hob_feature,
-			tf->hob_nsect,
-			tf->hob_lbal,
-			tf->hob_lbam,
-			tf->hob_lbah);
-	}
+	/* if device 1 was found in ata_devchk, wait for register
+	 * access briefly, then wait for BSY to clear.
+	 */
+	if (dev1) {
+		int i;
 
-	if (is_addr) {
-		iowrite8(tf->feature, ioaddr->feature_addr);
-		iowrite8(tf->nsect, ioaddr->nsect_addr);
-		iowrite8(tf->lbal, ioaddr->lbal_addr);
-		iowrite8(tf->lbam, ioaddr->lbam_addr);
-		iowrite8(tf->lbah, ioaddr->lbah_addr);
-		VPRINTK("feat 0x%X nsect 0x%X lba 0x%X 0x%X 0x%X\n",
-			tf->feature,
-			tf->nsect,
-			tf->lbal,
-			tf->lbam,
-			tf->lbah);
-	}
+		ap->ops->dev_select(ap, 1);
 
-	if (tf->flags & ATA_TFLAG_DEVICE) {
-		iowrite8(tf->device, ioaddr->device_addr);
-		VPRINTK("device 0x%X\n", tf->device);
+		/* Wait for register access.  Some ATAPI devices fail
+		 * to set nsect/lbal after reset, so don't waste too
+		 * much time on it.  We're gonna wait for !BSY anyway.
+		 */
+		for (i = 0; i < 2; i++) {
+			u8 nsect, lbal;
+
+			nsect = ioread8(ioaddr->nsect_addr);
+			lbal = ioread8(ioaddr->lbal_addr);
+			if ((nsect == 1) && (lbal == 1))
+				break;
+			msleep(50);	/* give drive a breather */
+		}
+
+		rc = ata_wait_ready(ap, deadline);
+		if (rc) {
+			if (rc != -ENODEV)
+				return rc;
+			ret = rc;
+		}
 	}
 
-	ata_wait_idle(ap);
+	/* is all this really necessary? */
+	ap->ops->dev_select(ap, 0);
+	if (dev1)
+		ap->ops->dev_select(ap, 1);
+	if (dev0)
+		ap->ops->dev_select(ap, 0);
+
+	return ret;
 }
 
 /**
- *	ata_tf_read - input device's ATA taskfile shadow registers
- *	@ap: Port from which input is read
- *	@tf: ATA taskfile register set for storing input
+ *	ata_wait_after_reset - wait before checking status after reset
+ *	@ap: port containing status register to be polled
+ *	@deadline: deadline jiffies for the operation
  *
- *	Reads ATA taskfile registers for currently-selected device
- *	into @tf. Assumes the device has a fully SFF compliant task file
- *	layout and behaviour. If you device does not (eg has a different
- *	status method) then you will need to provide a replacement tf_read
+ *	After reset, we need to pause a while before reading status.
+ *	Also, certain combination of controller and device report 0xff
+ *	for some duration (e.g. until SATA PHY is up and running)
+ *	which is interpreted as empty port in ATA world.  This
+ *	function also waits for such devices to get out of 0xff
+ *	status.
  *
  *	LOCKING:
- *	Inherited from caller.
+ *	Kernel thread context (may sleep).
  */
-void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
+void ata_wait_after_reset(struct ata_port *ap, unsigned long deadline)
 {
-	struct ata_ioports *ioaddr = &ap->ioaddr;
+	unsigned long until = jiffies + ATA_TMOUT_FF_WAIT;
+
+	if (time_before(until, deadline))
+		deadline = until;
+
+	/* Spec mandates ">= 2ms" before checking status.  We wait
+	 * 150ms, because that was the magic delay used for ATAPI
+	 * devices in Hale Landis's ATADRVR, for the period of time
+	 * between when the ATA command register is written, and then
+	 * status is checked.  Because waiting for "a while" before
+	 * checking status is fine, post SRST, we perform this magic
+	 * delay here as well.
+	 *
+	 * Old drivers/ide uses the 2mS rule and then waits for ready.
+	 */
+	msleep(150);
 
-	tf->command = ata_check_status(ap);
-	tf->feature = ioread8(ioaddr->error_addr);
-	tf->nsect = ioread8(ioaddr->nsect_addr);
-	tf->lbal = ioread8(ioaddr->lbal_addr);
-	tf->lbam = ioread8(ioaddr->lbam_addr);
-	tf->lbah = ioread8(ioaddr->lbah_addr);
-	tf->device = ioread8(ioaddr->device_addr);
+	/* Wait for 0xff to clear.  Some SATA devices take a long time
+	 * to clear 0xff after reset.  For example, HHD424020F7SV00
+	 * iVDR needs >= 800ms while.  Quantum GoVault needs even more
+	 * than that.
+	 *
+	 * Note that some PATA controllers (pata_ali) explode if
+	 * status register is read more than once when there's no
+	 * device attached.
+	 */
+	if (ap->flags & ATA_FLAG_SATA) {
+		while (1) {
+			u8 status = ata_chk_status(ap);
 
-	if (tf->flags & ATA_TFLAG_LBA48) {
-		if (likely(ioaddr->ctl_addr)) {
-			iowrite8(tf->ctl | ATA_HOB, ioaddr->ctl_addr);
-			tf->hob_feature = ioread8(ioaddr->error_addr);
-			tf->hob_nsect = ioread8(ioaddr->nsect_addr);
-			tf->hob_lbal = ioread8(ioaddr->lbal_addr);
-			tf->hob_lbam = ioread8(ioaddr->lbam_addr);
-			tf->hob_lbah = ioread8(ioaddr->lbah_addr);
-			iowrite8(tf->ctl, ioaddr->ctl_addr);
-			ap->last_ctl = tf->ctl;
-		} else
-			WARN_ON(1);
+			if (status != 0xff || time_after(jiffies, deadline))
+				return;
+
+			msleep(50);
+		}
 	}
 }
 
-/**
- *	ata_exec_command - issue ATA command to host controller
- *	@ap: port to which command is being issued
- *	@tf: ATA taskfile register set
- *
- *	Issues ATA command, with proper synchronization with interrupt
- *	handler / other threads.
- *
- *	LOCKING:
- *	spin_lock_irqsave(host lock)
- */
-void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf)
+static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
+			     unsigned long deadline)
 {
-	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
+	struct ata_ioports *ioaddr = &ap->ioaddr;
 
-	iowrite8(tf->command, ap->ioaddr.command_addr);
-	ata_pause(ap);
+	DPRINTK("ata%u: bus reset via SRST\n", ap->print_id);
+
+	/* software reset.  causes dev0 to be selected */
+	iowrite8(ap->ctl, ioaddr->ctl_addr);
+	udelay(20);	/* FIXME: flush */
+	iowrite8(ap->ctl | ATA_SRST, ioaddr->ctl_addr);
+	udelay(20);	/* FIXME: flush */
+	iowrite8(ap->ctl, ioaddr->ctl_addr);
+
+	/* wait a while before checking status */
+	ata_wait_after_reset(ap, deadline);
+
+	/* Before we perform post reset processing we want to see if
+	 * the bus shows 0xFF because the odd clown forgets the D7
+	 * pulldown resistor.
+	 */
+	if (ata_chk_status(ap) == 0xFF)
+		return -ENODEV;
+
+	return ata_bus_post_reset(ap, devmask, deadline);
 }
 
 /**
- *	ata_bmdma_freeze - Freeze BMDMA controller port
- *	@ap: port to freeze
+ *	ata_std_softreset - reset host port via ATA SRST
+ *	@link: ATA link to reset
+ *	@classes: resulting classes of attached devices
+ *	@deadline: deadline jiffies for the operation
  *
- *	Freeze BMDMA controller port.
+ *	Reset host port using ATA SRST.
  *
  *	LOCKING:
- *	Inherited from caller.
+ *	Kernel thread context (may sleep)
+ *
+ *	RETURNS:
+ *	0 on success, -errno otherwise.
  */
-void ata_bmdma_freeze(struct ata_port *ap)
+int ata_std_softreset(struct ata_link *link, unsigned int *classes,
+		      unsigned long deadline)
 {
-	struct ata_ioports *ioaddr = &ap->ioaddr;
+	struct ata_port *ap = link->ap;
+	unsigned int slave_possible = ap->flags & ATA_FLAG_SLAVE_POSS;
+	unsigned int devmask = 0;
+	int rc;
+	u8 err;
 
-	ap->ctl |= ATA_NIEN;
-	ap->last_ctl = ap->ctl;
+	DPRINTK("ENTER\n");
 
-	if (ioaddr->ctl_addr)
-		iowrite8(ap->ctl, ioaddr->ctl_addr);
+	if (ata_link_offline(link)) {
+		classes[0] = ATA_DEV_NONE;
+		goto out;
+	}
 
-	/* Under certain circumstances, some controllers raise IRQ on
-	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
-	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
-	 */
-	ata_chk_status(ap);
+	/* determine if device 0/1 are present */
+	if (ata_devchk(ap, 0))
+		devmask |= (1 << 0);
+	if (slave_possible && ata_devchk(ap, 1))
+		devmask |= (1 << 1);
+
+	/* select device 0 again */
+	ap->ops->dev_select(ap, 0);
+
+	/* issue bus reset */
+	DPRINTK("about to softreset, devmask=%x\n", devmask);
+	rc = ata_bus_softreset(ap, devmask, deadline);
+	/* if link is occupied, -ENODEV too is an error */
+	if (rc && (rc != -ENODEV || sata_scr_valid(link))) {
+		ata_link_printk(link, KERN_ERR, "SRST failed (errno=%d)\n", rc);
+		return rc;
+	}
 
-	ap->ops->irq_clear(ap);
+	/* determine by signature whether we have ATA or ATAPI devices */
+	classes[0] = ata_dev_try_classify(&link->device[0],
+					  devmask & (1 << 0), &err);
+	if (slave_possible && err != 0x81)
+		classes[1] = ata_dev_try_classify(&link->device[1],
+						  devmask & (1 << 1), &err);
+
+ out:
+	DPRINTK("EXIT, classes[0]=%u [1]=%u\n", classes[0], classes[1]);
+	return 0;
 }
 
 /**
- *	ata_bmdma_thaw - Thaw BMDMA controller port
- *	@ap: port to thaw
+ *	sata_std_hardreset - reset host port via SATA phy reset
+ *	@link: link to reset
+ *	@class: resulting class of attached device
+ *	@deadline: deadline jiffies for the operation
  *
- *	Thaw BMDMA controller port.
+ *	SATA phy-reset host port using DET bits of SControl register,
+ *	wait for !BSY and classify the attached device.
  *
  *	LOCKING:
- *	Inherited from caller.
+ *	Kernel thread context (may sleep)
+ *
+ *	RETURNS:
+ *	0 on success, -errno otherwise.
  */
-void ata_bmdma_thaw(struct ata_port *ap)
+int sata_std_hardreset(struct ata_link *link, unsigned int *class,
+		       unsigned long deadline)
 {
-	/* clear & re-enable interrupts */
-	ata_chk_status(ap);
-	ap->ops->irq_clear(ap);
-	ap->ops->irq_on(ap);
+	struct ata_port *ap = link->ap;
+	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
+	int rc;
+
+	DPRINTK("ENTER\n");
+
+	/* do hardreset */
+	rc = sata_link_hardreset(link, timing, deadline);
+	if (rc) {
+		ata_link_printk(link, KERN_ERR,
+				"COMRESET failed (errno=%d)\n", rc);
+		return rc;
+	}
+
+	/* TODO: phy layer with polling, timeouts, etc. */
+	if (ata_link_offline(link)) {
+		*class = ATA_DEV_NONE;
+		DPRINTK("EXIT, link offline\n");
+		return 0;
+	}
+
+	/* wait a while before checking status */
+	ata_wait_after_reset(ap, deadline);
+
+	/* If PMP is supported, we have to do follow-up SRST.  Note
+	 * that some PMPs don't send D2H Reg FIS after hardreset at
+	 * all if the first port is empty.  Wait for it just for a
+	 * second and request follow-up SRST.
+	 */
+	if (ap->flags & ATA_FLAG_PMP) {
+		ata_wait_ready(ap, jiffies + HZ);
+		return -EAGAIN;
+	}
+
+	rc = ata_wait_ready(ap, deadline);
+	/* link occupied, -ENODEV too is an error */
+	if (rc) {
+		ata_link_printk(link, KERN_ERR,
+				"COMRESET failed (errno=%d)\n", rc);
+		return rc;
+	}
+
+	ap->ops->dev_select(ap, 0);	/* probably unnecessary */
+
+	*class = ata_dev_try_classify(link->device, 1, NULL);
+
+	DPRINTK("EXIT, class=%u\n", *class);
+	return 0;
 }
 
 /**
@@ -393,6 +2092,31 @@ int ata_sff_port_start(struct ata_port *ap)
 	return 0;
 }
 
+/**
+ *	ata_std_ports - initialize ioaddr with standard port offsets.
+ *	@ioaddr: IO address structure to be initialized
+ *
+ *	Utility function which initializes data_addr, error_addr,
+ *	feature_addr, nsect_addr, lbal_addr, lbam_addr, lbah_addr,
+ *	device_addr, status_addr, and command_addr to standard offsets
+ *	relative to cmd_addr.
+ *
+ *	Does not set ctl_addr, altstatus_addr, bmdma_addr, or scr_addr.
+ */
+void ata_std_ports(struct ata_ioports *ioaddr)
+{
+	ioaddr->data_addr = ioaddr->cmd_addr + ATA_REG_DATA;
+	ioaddr->error_addr = ioaddr->cmd_addr + ATA_REG_ERR;
+	ioaddr->feature_addr = ioaddr->cmd_addr + ATA_REG_FEATURE;
+	ioaddr->nsect_addr = ioaddr->cmd_addr + ATA_REG_NSECT;
+	ioaddr->lbal_addr = ioaddr->cmd_addr + ATA_REG_LBAL;
+	ioaddr->lbam_addr = ioaddr->cmd_addr + ATA_REG_LBAM;
+	ioaddr->lbah_addr = ioaddr->cmd_addr + ATA_REG_LBAH;
+	ioaddr->device_addr = ioaddr->cmd_addr + ATA_REG_DEVICE;
+	ioaddr->status_addr = ioaddr->cmd_addr + ATA_REG_STATUS;
+	ioaddr->command_addr = ioaddr->cmd_addr + ATA_REG_CMD;
+}
+
 /**
  *	ata_bmdma_setup - Set up PCI IDE BMDMA transaction
  *	@qc: Info associated with this ATA transaction.
@@ -494,11 +2218,94 @@ u8 ata_bmdma_status(struct ata_port *ap)
 }
 
 /**
- *	ata_noop_irq_clear - Noop placeholder for irq_clear
- *	@ap: Port associated with this ATA transaction.
+ *	ata_bus_reset - reset host port and associated ATA channel
+ *	@ap: port to reset
+ *
+ *	This is typically the first time we actually start issuing
+ *	commands to the ATA channel.  We wait for BSY to clear, then
+ *	issue EXECUTE DEVICE DIAGNOSTIC command, polling for its
+ *	result.  Determine what devices, if any, are on the channel
+ *	by looking at the device 0/1 error register.  Look at the signature
+ *	stored in each device's taskfile registers, to determine if
+ *	the device is ATA or ATAPI.
+ *
+ *	LOCKING:
+ *	PCI/etc. bus probe sem.
+ *	Obtains host lock.
+ *
+ *	SIDE EFFECTS:
+ *	Sets ATA_FLAG_DISABLED if bus reset fails.
+ *
+ *	DEPRECATED:
+ *	This function is only for drivers which still use old EH and
+ *	will be removed soon.
  */
-void ata_noop_irq_clear(struct ata_port *ap)
+void ata_bus_reset(struct ata_port *ap)
 {
+	struct ata_device *device = ap->link.device;
+	struct ata_ioports *ioaddr = &ap->ioaddr;
+	unsigned int slave_possible = ap->flags & ATA_FLAG_SLAVE_POSS;
+	u8 err;
+	unsigned int dev0, dev1 = 0, devmask = 0;
+	int rc;
+
+	DPRINTK("ENTER, host %u, port %u\n", ap->print_id, ap->port_no);
+
+	/* determine if device 0/1 are present */
+	if (ap->flags & ATA_FLAG_SATA_RESET)
+		dev0 = 1;
+	else {
+		dev0 = ata_devchk(ap, 0);
+		if (slave_possible)
+			dev1 = ata_devchk(ap, 1);
+	}
+
+	if (dev0)
+		devmask |= (1 << 0);
+	if (dev1)
+		devmask |= (1 << 1);
+
+	/* select device 0 again */
+	ap->ops->dev_select(ap, 0);
+
+	/* issue bus reset */
+	if (ap->flags & ATA_FLAG_SRST) {
+		rc = ata_bus_softreset(ap, devmask, jiffies + 40 * HZ);
+		if (rc && rc != -ENODEV)
+			goto err_out;
+	}
+
+	/*
+	 * determine by signature whether we have ATA or ATAPI devices
+	 */
+	device[0].class = ata_dev_try_classify(&device[0], dev0, &err);
+	if ((slave_possible) && (err != 0x81))
+		device[1].class = ata_dev_try_classify(&device[1], dev1, &err);
+
+	/* is double-select really necessary? */
+	if (device[1].class != ATA_DEV_NONE)
+		ap->ops->dev_select(ap, 1);
+	if (device[0].class != ATA_DEV_NONE)
+		ap->ops->dev_select(ap, 0);
+
+	/* if no devices were detected, disable this port */
+	if ((device[0].class == ATA_DEV_NONE) &&
+	    (device[1].class == ATA_DEV_NONE))
+		goto err_out;
+
+	if (ap->flags & (ATA_FLAG_SATA_RESET | ATA_FLAG_SRST)) {
+		/* set up device control for ATA_FLAG_SATA_RESET */
+		iowrite8(ap->ctl, ioaddr->ctl_addr);
+	}
+
+	DPRINTK("EXIT\n");
+	return;
+
+err_out:
+	ata_port_printk(ap, KERN_ERR, "disabling port\n");
+	ata_port_disable(ap);
+
+	DPRINTK("EXIT\n");
 }
 
 #ifdef CONFIG_PCI
@@ -914,3 +2721,49 @@ int ata_pci_init_one(struct pci_dev *pdev,
 
 #endif /* CONFIG_PCI */
 
+EXPORT_SYMBOL_GPL(ata_sff_port_ops);
+EXPORT_SYMBOL_GPL(ata_bmdma_port_ops);
+EXPORT_SYMBOL_GPL(ata_qc_prep);
+EXPORT_SYMBOL_GPL(ata_dumb_qc_prep);
+EXPORT_SYMBOL_GPL(ata_std_dev_select);
+EXPORT_SYMBOL_GPL(ata_check_status);
+EXPORT_SYMBOL_GPL(ata_altstatus);
+EXPORT_SYMBOL_GPL(ata_busy_sleep);
+EXPORT_SYMBOL_GPL(ata_wait_ready);
+EXPORT_SYMBOL_GPL(ata_tf_load);
+EXPORT_SYMBOL_GPL(ata_tf_read);
+EXPORT_SYMBOL_GPL(ata_exec_command);
+EXPORT_SYMBOL_GPL(ata_data_xfer);
+EXPORT_SYMBOL_GPL(ata_data_xfer_noirq);
+EXPORT_SYMBOL_GPL(ata_irq_on);
+EXPORT_SYMBOL_GPL(ata_bmdma_irq_clear);
+EXPORT_SYMBOL_GPL(ata_hsm_move);
+EXPORT_SYMBOL_GPL(ata_qc_issue_prot);
+EXPORT_SYMBOL_GPL(ata_host_intr);
+EXPORT_SYMBOL_GPL(ata_interrupt);
+EXPORT_SYMBOL_GPL(ata_bmdma_freeze);
+EXPORT_SYMBOL_GPL(ata_bmdma_thaw);
+EXPORT_SYMBOL_GPL(ata_std_prereset);
+EXPORT_SYMBOL_GPL(ata_dev_try_classify);
+EXPORT_SYMBOL_GPL(ata_wait_after_reset);
+EXPORT_SYMBOL_GPL(ata_std_softreset);
+EXPORT_SYMBOL_GPL(sata_std_hardreset);
+EXPORT_SYMBOL_GPL(ata_std_postreset);
+EXPORT_SYMBOL_GPL(ata_bmdma_error_handler);
+EXPORT_SYMBOL_GPL(ata_bmdma_post_internal_cmd);
+EXPORT_SYMBOL_GPL(ata_sff_port_start);
+EXPORT_SYMBOL_GPL(ata_std_ports);
+EXPORT_SYMBOL_GPL(ata_bmdma_setup);
+EXPORT_SYMBOL_GPL(ata_bmdma_start);
+EXPORT_SYMBOL_GPL(ata_bmdma_stop);
+EXPORT_SYMBOL_GPL(ata_bmdma_status);
+EXPORT_SYMBOL_GPL(ata_bus_reset);
+#ifdef CONFIG_PCI
+EXPORT_SYMBOL_GPL(ata_pci_clear_simplex);
+EXPORT_SYMBOL_GPL(ata_pci_default_filter);
+EXPORT_SYMBOL_GPL(ata_pci_init_bmdma);
+EXPORT_SYMBOL_GPL(ata_pci_init_sff_host);
+EXPORT_SYMBOL_GPL(ata_pci_prepare_sff_host);
+EXPORT_SYMBOL_GPL(ata_pci_activate_sff_host);
+EXPORT_SYMBOL_GPL(ata_pci_init_one);
+#endif /* CONFIG_PCI */
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index aa884f71a12a..a69f663c7402 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -67,6 +67,8 @@ extern int ata_build_rw_tf(struct ata_taskfile *tf, struct ata_device *dev,
 			   unsigned int tag);
 extern u64 ata_tf_read_block(struct ata_taskfile *tf, struct ata_device *dev);
 extern void ata_dev_disable(struct ata_device *dev);
+extern void ata_pio_queue_task(struct ata_port *ap, void *data,
+			       unsigned long delay);
 extern void ata_port_flush_task(struct ata_port *ap);
 extern unsigned ata_exec_internal(struct ata_device *dev,
 				  struct ata_taskfile *tf, const u8 *cdb,
@@ -91,8 +93,6 @@ extern void ata_qc_free(struct ata_queued_cmd *qc);
 extern void ata_qc_issue(struct ata_queued_cmd *qc);
 extern void __ata_qc_complete(struct ata_queued_cmd *qc);
 extern int ata_check_atapi_dma(struct ata_queued_cmd *qc);
-extern void ata_dev_select(struct ata_port *ap, unsigned int device,
-                           unsigned int wait, unsigned int can_sleep);
 extern void swap_buf_le16(u16 *buf, unsigned int buf_words);
 extern int ata_flush_cache(struct ata_device *dev);
 extern void ata_dev_init(struct ata_device *dev);
@@ -194,7 +194,9 @@ extern int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 extern void ata_eh_finish(struct ata_port *ap);
 
 /* libata-sff.c */
+extern void ata_dev_select(struct ata_port *ap, unsigned int device,
+                           unsigned int wait, unsigned int can_sleep);
 extern u8 ata_irq_on(struct ata_port *ap);
-
+extern void ata_pio_task(struct work_struct *work);
 
 #endif /* __LIBATA_H__ */
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 01c233303aee..673f34b256ba 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -349,6 +349,22 @@ enum {
 	ATAPI_READ_CD		= 2,		/* READ CD [MSF] */
 	ATAPI_PASS_THRU		= 3,		/* SAT pass-thru */
 	ATAPI_MISC		= 4,		/* the rest */
+
+	/* Timing constants */
+	ATA_TIMING_SETUP	= (1 << 0),
+	ATA_TIMING_ACT8B	= (1 << 1),
+	ATA_TIMING_REC8B	= (1 << 2),
+	ATA_TIMING_CYC8B	= (1 << 3),
+	ATA_TIMING_8BIT		= ATA_TIMING_ACT8B | ATA_TIMING_REC8B |
+				  ATA_TIMING_CYC8B,
+	ATA_TIMING_ACTIVE	= (1 << 4),
+	ATA_TIMING_RECOVER	= (1 << 5),
+	ATA_TIMING_CYCLE	= (1 << 6),
+	ATA_TIMING_UDMA		= (1 << 7),
+	ATA_TIMING_ALL		= ATA_TIMING_SETUP | ATA_TIMING_ACT8B |
+				  ATA_TIMING_REC8B | ATA_TIMING_CYC8B |
+				  ATA_TIMING_ACTIVE | ATA_TIMING_RECOVER |
+				  ATA_TIMING_CYCLE | ATA_TIMING_UDMA,
 };
 
 enum ata_xfer_mask {
@@ -779,6 +795,9 @@ struct ata_timing {
 
 #define FIT(v, vmin, vmax)	max_t(short, min_t(short, v, vmax), vmin)
 
+/*
+ * Core layer - drivers/ata/libata-core.c
+ */
 extern const unsigned long sata_deb_timing_normal[];
 extern const unsigned long sata_deb_timing_hotplug[];
 extern const unsigned long sata_deb_timing_long[];
@@ -802,22 +821,14 @@ static inline int ata_port_is_dummy(struct ata_port *ap)
 
 extern void sata_print_link_status(struct ata_link *link);
 extern void ata_port_probe(struct ata_port *);
-extern void ata_bus_reset(struct ata_port *ap);
 extern int sata_set_spd(struct ata_link *link);
 extern int sata_link_debounce(struct ata_link *link,
 			const unsigned long *params, unsigned long deadline);
 extern int sata_link_resume(struct ata_link *link, const unsigned long *params,
 			    unsigned long deadline);
-extern int ata_std_prereset(struct ata_link *link, unsigned long deadline);
-extern int ata_std_softreset(struct ata_link *link, unsigned int *classes,
-			     unsigned long deadline);
 extern int sata_link_hardreset(struct ata_link *link,
 			const unsigned long *timing, unsigned long deadline);
-extern int sata_std_hardreset(struct ata_link *link, unsigned int *class,
-			      unsigned long deadline);
-extern void ata_std_postreset(struct ata_link *link, unsigned int *classes);
 extern void ata_port_disable(struct ata_port *);
-extern void ata_std_ports(struct ata_ioports *ioaddr);
 
 extern struct ata_host *ata_host_alloc(struct device *dev, int max_ports);
 extern struct ata_host *ata_host_alloc_pinfo(struct device *dev,
@@ -843,7 +854,6 @@ extern void ata_sas_port_stop(struct ata_port *ap);
 extern int ata_sas_slave_configure(struct scsi_device *, struct ata_port *);
 extern int ata_sas_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *),
 			    struct ata_port *ap);
-extern unsigned int ata_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc);
 extern int sata_scr_valid(struct ata_link *link);
 extern int sata_scr_read(struct ata_link *link, int reg, u32 *val);
 extern int sata_scr_write(struct ata_link *link, int reg, u32 val);
@@ -855,21 +865,9 @@ extern int ata_host_suspend(struct ata_host *host, pm_message_t mesg);
 extern void ata_host_resume(struct ata_host *host);
 #endif
 extern int ata_ratelimit(void);
-extern int ata_busy_sleep(struct ata_port *ap,
-			  unsigned long timeout_pat, unsigned long timeout);
-extern void ata_wait_after_reset(struct ata_port *ap, unsigned long deadline);
-extern int ata_wait_ready(struct ata_port *ap, unsigned long deadline);
 extern u32 ata_wait_register(void __iomem *reg, u32 mask, u32 val,
 			     unsigned long interval_msec,
 			     unsigned long timeout_msec);
-extern unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
-					 u8 *r_err);
-
-/*
- * Default driver ops implementations
- */
-extern void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf);
-extern void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
 extern int atapi_cmd_type(u8 opcode);
 extern void ata_tf_to_fis(const struct ata_taskfile *tf,
 			  u8 pmp, int is_cmd, u8 *fis);
@@ -885,22 +883,9 @@ extern int ata_xfer_mode2shift(unsigned long xfer_mode);
 extern const char *ata_mode_string(unsigned long xfer_mask);
 extern unsigned long ata_id_xfermask(const u16 *id);
 extern void ata_noop_dev_select(struct ata_port *ap, unsigned int device);
-extern void ata_std_dev_select(struct ata_port *ap, unsigned int device);
-extern u8 ata_check_status(struct ata_port *ap);
-extern u8 ata_altstatus(struct ata_port *ap);
-extern void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf);
 extern int ata_port_start(struct ata_port *ap);
-extern int ata_sff_port_start(struct ata_port *ap);
-extern irqreturn_t ata_interrupt(int irq, void *dev_instance);
-extern unsigned int ata_data_xfer(struct ata_device *dev,
-			unsigned char *buf, unsigned int buflen, int rw);
-extern unsigned int ata_data_xfer_noirq(struct ata_device *dev,
-			unsigned char *buf, unsigned int buflen, int rw);
 extern int ata_std_qc_defer(struct ata_queued_cmd *qc);
-extern void ata_dumb_qc_prep(struct ata_queued_cmd *qc);
-extern void ata_qc_prep(struct ata_queued_cmd *qc);
 extern void ata_noop_qc_prep(struct ata_queued_cmd *qc);
-extern unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc);
 extern void ata_sg_init(struct ata_queued_cmd *qc, struct scatterlist *sg,
 		 unsigned int n_elem);
 extern unsigned int ata_dev_classify(const struct ata_taskfile *tf);
@@ -909,18 +894,7 @@ extern void ata_id_string(const u16 *id, unsigned char *s,
 			  unsigned int ofs, unsigned int len);
 extern void ata_id_c_string(const u16 *id, unsigned char *s,
 			    unsigned int ofs, unsigned int len);
-extern void ata_bmdma_setup(struct ata_queued_cmd *qc);
-extern void ata_bmdma_start(struct ata_queued_cmd *qc);
-extern void ata_bmdma_stop(struct ata_queued_cmd *qc);
-extern u8   ata_bmdma_status(struct ata_port *ap);
-extern void ata_bmdma_irq_clear(struct ata_port *ap);
 extern void ata_noop_irq_clear(struct ata_port *ap);
-extern void ata_bmdma_freeze(struct ata_port *ap);
-extern void ata_bmdma_thaw(struct ata_port *ap);
-extern void ata_bmdma_error_handler(struct ata_port *ap);
-extern void ata_bmdma_post_internal_cmd(struct ata_queued_cmd *qc);
-extern int ata_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
-			u8 status, int in_wq);
 extern void ata_qc_complete(struct ata_queued_cmd *qc);
 extern int ata_qc_complete_multiple(struct ata_port *ap, u32 qc_active,
 				    void (*finish_qc)(struct ata_queued_cmd *));
@@ -935,7 +909,6 @@ extern int ata_scsi_change_queue_depth(struct scsi_device *sdev,
 				       int queue_depth);
 extern struct ata_device *ata_dev_pair(struct ata_device *adev);
 extern int ata_do_set_mode(struct ata_link *link, struct ata_device **r_failed_dev);
-extern u8 ata_irq_on(struct ata_port *ap);
 
 extern int ata_cable_40wire(struct ata_port *ap);
 extern int ata_cable_80wire(struct ata_port *ap);
@@ -943,10 +916,7 @@ extern int ata_cable_sata(struct ata_port *ap);
 extern int ata_cable_ignore(struct ata_port *ap);
 extern int ata_cable_unknown(struct ata_port *ap);
 
-/*
- * Timing helpers
- */
-
+/* Timing helpers */
 extern unsigned int ata_pio_need_iordy(const struct ata_device *);
 extern const struct ata_timing *ata_timing_find_mode(u8 xfer_mode);
 extern int ata_timing_compute(struct ata_device *, unsigned short,
@@ -956,24 +926,31 @@ extern void ata_timing_merge(const struct ata_timing *,
 			     unsigned int);
 extern u8 ata_timing_cycle2mode(unsigned int xfer_shift, int cycle);
 
-enum {
-	ATA_TIMING_SETUP	= (1 << 0),
-	ATA_TIMING_ACT8B	= (1 << 1),
-	ATA_TIMING_REC8B	= (1 << 2),
-	ATA_TIMING_CYC8B	= (1 << 3),
-	ATA_TIMING_8BIT		= ATA_TIMING_ACT8B | ATA_TIMING_REC8B |
-				  ATA_TIMING_CYC8B,
-	ATA_TIMING_ACTIVE	= (1 << 4),
-	ATA_TIMING_RECOVER	= (1 << 5),
-	ATA_TIMING_CYCLE	= (1 << 6),
-	ATA_TIMING_UDMA		= (1 << 7),
-	ATA_TIMING_ALL		= ATA_TIMING_SETUP | ATA_TIMING_ACT8B |
-				  ATA_TIMING_REC8B | ATA_TIMING_CYC8B |
-				  ATA_TIMING_ACTIVE | ATA_TIMING_RECOVER |
-				  ATA_TIMING_CYCLE | ATA_TIMING_UDMA,
+/* PCI */
+#ifdef CONFIG_PCI
+struct pci_dev;
+
+struct pci_bits {
+	unsigned int		reg;	/* PCI config register to read */
+	unsigned int		width;	/* 1 (8 bit), 2 (16 bit), 4 (32 bit) */
+	unsigned long		mask;
+	unsigned long		val;
 };
 
-/* libata-acpi.c */
+extern int pci_test_config_bits(struct pci_dev *pdev, const struct pci_bits *bits);
+extern void ata_pci_remove_one(struct pci_dev *pdev);
+
+#ifdef CONFIG_PM
+extern void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t mesg);
+extern int __must_check ata_pci_device_do_resume(struct pci_dev *pdev);
+extern int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg);
+extern int ata_pci_device_resume(struct pci_dev *pdev);
+#endif /* CONFIG_PM */
+#endif /* CONFIG_PCI */
+
+/*
+ * ACPI - drivers/ata/libata-acpi.c
+ */
 #ifdef CONFIG_ATA_ACPI
 static inline const struct ata_acpi_gtm *ata_acpi_init_gtm(struct ata_port *ap)
 {
@@ -1017,43 +994,8 @@ static inline int ata_acpi_cbl_80wire(struct ata_port *ap,
 }
 #endif
 
-#ifdef CONFIG_PCI
-struct pci_dev;
-
-extern int ata_pci_init_one(struct pci_dev *pdev,
-			    const struct ata_port_info * const * ppi,
-			    struct scsi_host_template *sht, void *host_priv);
-extern void ata_pci_remove_one(struct pci_dev *pdev);
-#ifdef CONFIG_PM
-extern void ata_pci_device_do_suspend(struct pci_dev *pdev, pm_message_t mesg);
-extern int __must_check ata_pci_device_do_resume(struct pci_dev *pdev);
-extern int ata_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg);
-extern int ata_pci_device_resume(struct pci_dev *pdev);
-#endif
-extern int ata_pci_clear_simplex(struct pci_dev *pdev);
-
-struct pci_bits {
-	unsigned int		reg;	/* PCI config register to read */
-	unsigned int		width;	/* 1 (8 bit), 2 (16 bit), 4 (32 bit) */
-	unsigned long		mask;
-	unsigned long		val;
-};
-
-extern int ata_pci_init_sff_host(struct ata_host *host);
-extern int ata_pci_init_bmdma(struct ata_host *host);
-extern int ata_pci_prepare_sff_host(struct pci_dev *pdev,
-				    const struct ata_port_info * const * ppi,
-				    struct ata_host **r_host);
-extern int ata_pci_activate_sff_host(struct ata_host *host,
-				     irq_handler_t irq_handler,
-				     struct scsi_host_template *sht);
-extern int pci_test_config_bits(struct pci_dev *pdev, const struct pci_bits *bits);
-extern unsigned long ata_pci_default_filter(struct ata_device *dev,
-					    unsigned long xfer_mask);
-#endif /* CONFIG_PCI */
-
 /*
- * PMP
+ * PMP - drivers/ata/libata-pmp.c
  */
 extern int sata_pmp_qc_defer_cmd_switch(struct ata_queued_cmd *qc);
 extern int sata_pmp_std_prereset(struct ata_link *link, unsigned long deadline);
@@ -1063,7 +1005,7 @@ extern void sata_pmp_std_postreset(struct ata_link *link, unsigned int *class);
 extern void sata_pmp_error_handler(struct ata_port *ap);
 
 /*
- * EH
+ * EH - drivers/ata/libata-eh.c
  */
 extern void ata_port_schedule_eh(struct ata_port *ap);
 extern int ata_link_abort(struct ata_link *link);
@@ -1106,8 +1048,6 @@ extern void ata_std_error_handler(struct ata_port *ap);
 extern const struct ata_port_operations ata_base_port_ops;
 extern const struct ata_port_operations sata_port_ops;
 extern const struct ata_port_operations sata_pmp_port_ops;
-extern const struct ata_port_operations ata_sff_port_ops;
-extern const struct ata_port_operations ata_bmdma_port_ops;
 
 #define ATA_BASE_SHT(drv_name)					\
 	.module			= THIS_MODULE,			\
@@ -1124,17 +1064,6 @@ extern const struct ata_port_operations ata_bmdma_port_ops;
 	.slave_destroy		= ata_scsi_slave_destroy,	\
 	.bios_param		= ata_std_bios_param
 
-/* PIO only, sg_tablesize and dma_boundary limits can be removed */
-#define ATA_PIO_SHT(drv_name)					\
-	ATA_BASE_SHT(drv_name),					\
-	.sg_tablesize		= LIBATA_MAX_PRD,		\
-	.dma_boundary		= ATA_DMA_BOUNDARY
-
-#define ATA_BMDMA_SHT(drv_name)					\
-	ATA_BASE_SHT(drv_name),					\
-	.sg_tablesize		= LIBATA_MAX_PRD,		\
-	.dma_boundary		= ATA_DMA_BOUNDARY
-
 #define ATA_NCQ_SHT(drv_name)					\
 	ATA_BASE_SHT(drv_name),					\
 	.change_queue_depth	= ata_scsi_change_queue_depth
@@ -1287,11 +1216,6 @@ static inline struct ata_link *ata_port_next_link(struct ata_link *link)
 	for ((dev) = (link)->device + ata_link_max_devices(link) - 1; \
 	     (dev) >= (link)->device || ((dev) = NULL); (dev)--)
 
-static inline u8 ata_chk_status(struct ata_port *ap)
-{
-	return ap->ops->check_status(ap);
-}
-
 /**
  *	ata_ncq_enabled - Test whether NCQ is enabled
  *	@dev: ATA device to test for
@@ -1308,74 +1232,6 @@ static inline int ata_ncq_enabled(struct ata_device *dev)
 			      ATA_DFLAG_NCQ)) == ATA_DFLAG_NCQ;
 }
 
-/**
- *	ata_pause - Flush writes and pause 400 nanoseconds.
- *	@ap: Port to wait for.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-
-static inline void ata_pause(struct ata_port *ap)
-{
-	ata_altstatus(ap);
-	ndelay(400);
-}
-
-
-/**
- *	ata_busy_wait - Wait for a port status register
- *	@ap: Port to wait for.
- *	@bits: bits that must be clear
- *	@max: number of 10uS waits to perform
- *
- *	Waits up to max*10 microseconds for the selected bits in the port's
- *	status register to be cleared.
- *	Returns final value of status register.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-
-static inline u8 ata_busy_wait(struct ata_port *ap, unsigned int bits,
-			       unsigned int max)
-{
-	u8 status;
-
-	do {
-		udelay(10);
-		status = ata_chk_status(ap);
-		max--;
-	} while (status != 0xff && (status & bits) && (max > 0));
-
-	return status;
-}
-
-
-/**
- *	ata_wait_idle - Wait for a port to be idle.
- *	@ap: Port to wait for.
- *
- *	Waits up to 10ms for port's BUSY and DRQ signals to clear.
- *	Returns final value of status register.
- *
- *	LOCKING:
- *	Inherited from caller.
- */
-
-static inline u8 ata_wait_idle(struct ata_port *ap)
-{
-	u8 status = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
-
-#ifdef ATA_DEBUG
-	if (status != 0xff && (status & (ATA_BUSY | ATA_DRQ)))
-		ata_port_printk(ap, KERN_DEBUG, "abnormal Status 0x%X\n",
-				status);
-#endif
-
-	return status;
-}
-
 static inline void ata_qc_set_polling(struct ata_queued_cmd *qc)
 {
 	qc->tf.ctl |= ATA_NIEN;
@@ -1468,4 +1324,149 @@ static inline struct ata_port *ata_shost_to_port(struct Scsi_Host *host)
 	return *(struct ata_port **)&host->hostdata[0];
 }
 
+/**************************************************************************
+ * SFF - drivers/ata/libata-sff.c
+ */
+extern const struct ata_port_operations ata_sff_port_ops;
+extern const struct ata_port_operations ata_bmdma_port_ops;
+
+/* PIO only, sg_tablesize and dma_boundary limits can be removed */
+#define ATA_PIO_SHT(drv_name)					\
+	ATA_BASE_SHT(drv_name),					\
+	.sg_tablesize		= LIBATA_MAX_PRD,		\
+	.dma_boundary		= ATA_DMA_BOUNDARY
+
+#define ATA_BMDMA_SHT(drv_name)					\
+	ATA_BASE_SHT(drv_name),					\
+	.sg_tablesize		= LIBATA_MAX_PRD,		\
+	.dma_boundary		= ATA_DMA_BOUNDARY
+
+extern void ata_qc_prep(struct ata_queued_cmd *qc);
+extern void ata_dumb_qc_prep(struct ata_queued_cmd *qc);
+extern void ata_std_dev_select(struct ata_port *ap, unsigned int device);
+extern u8 ata_check_status(struct ata_port *ap);
+extern u8 ata_altstatus(struct ata_port *ap);
+extern int ata_busy_sleep(struct ata_port *ap,
+			  unsigned long timeout_pat, unsigned long timeout);
+extern int ata_wait_ready(struct ata_port *ap, unsigned long deadline);
+extern void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf);
+extern void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
+extern void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf);
+extern unsigned int ata_data_xfer(struct ata_device *dev,
+			unsigned char *buf, unsigned int buflen, int rw);
+extern unsigned int ata_data_xfer_noirq(struct ata_device *dev,
+			unsigned char *buf, unsigned int buflen, int rw);
+extern u8 ata_irq_on(struct ata_port *ap);
+extern void ata_bmdma_irq_clear(struct ata_port *ap);
+extern int ata_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
+			u8 status, int in_wq);
+extern unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc);
+extern unsigned int ata_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc);
+extern irqreturn_t ata_interrupt(int irq, void *dev_instance);
+extern void ata_bmdma_freeze(struct ata_port *ap);
+extern void ata_bmdma_thaw(struct ata_port *ap);
+extern int ata_std_prereset(struct ata_link *link, unsigned long deadline);
+extern unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
+					 u8 *r_err);
+extern void ata_wait_after_reset(struct ata_port *ap, unsigned long deadline);
+extern int ata_std_softreset(struct ata_link *link, unsigned int *classes,
+			     unsigned long deadline);
+extern int sata_std_hardreset(struct ata_link *link, unsigned int *class,
+			      unsigned long deadline);
+extern void ata_std_postreset(struct ata_link *link, unsigned int *classes);
+extern void ata_bmdma_error_handler(struct ata_port *ap);
+extern void ata_bmdma_post_internal_cmd(struct ata_queued_cmd *qc);
+extern int ata_sff_port_start(struct ata_port *ap);
+extern void ata_std_ports(struct ata_ioports *ioaddr);
+extern void ata_bmdma_setup(struct ata_queued_cmd *qc);
+extern void ata_bmdma_start(struct ata_queued_cmd *qc);
+extern void ata_bmdma_stop(struct ata_queued_cmd *qc);
+extern u8 ata_bmdma_status(struct ata_port *ap);
+extern void ata_bus_reset(struct ata_port *ap);
+
+#ifdef CONFIG_PCI
+extern int ata_pci_clear_simplex(struct pci_dev *pdev);
+extern unsigned long ata_pci_default_filter(struct ata_device *dev,
+					    unsigned long xfer_mask);
+extern int ata_pci_init_bmdma(struct ata_host *host);
+extern int ata_pci_init_sff_host(struct ata_host *host);
+extern int ata_pci_prepare_sff_host(struct pci_dev *pdev,
+				    const struct ata_port_info * const * ppi,
+				    struct ata_host **r_host);
+extern int ata_pci_activate_sff_host(struct ata_host *host,
+				     irq_handler_t irq_handler,
+				     struct scsi_host_template *sht);
+extern int ata_pci_init_one(struct pci_dev *pdev,
+			    const struct ata_port_info * const * ppi,
+			    struct scsi_host_template *sht, void *host_priv);
+#endif /* CONFIG_PCI */
+
+static inline u8 ata_chk_status(struct ata_port *ap)
+{
+	return ap->ops->check_status(ap);
+}
+
+/**
+ *	ata_pause - Flush writes and pause 400 nanoseconds.
+ *	@ap: Port to wait for.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+static inline void ata_pause(struct ata_port *ap)
+{
+	ata_altstatus(ap);
+	ndelay(400);
+}
+
+/**
+ *	ata_busy_wait - Wait for a port status register
+ *	@ap: Port to wait for.
+ *	@bits: bits that must be clear
+ *	@max: number of 10uS waits to perform
+ *
+ *	Waits up to max*10 microseconds for the selected bits in the port's
+ *	status register to be cleared.
+ *	Returns final value of status register.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+static inline u8 ata_busy_wait(struct ata_port *ap, unsigned int bits,
+			       unsigned int max)
+{
+	u8 status;
+
+	do {
+		udelay(10);
+		status = ata_chk_status(ap);
+		max--;
+	} while (status != 0xff && (status & bits) && (max > 0));
+
+	return status;
+}
+
+/**
+ *	ata_wait_idle - Wait for a port to be idle.
+ *	@ap: Port to wait for.
+ *
+ *	Waits up to 10ms for port's BUSY and DRQ signals to clear.
+ *	Returns final value of status register.
+ *
+ *	LOCKING:
+ *	Inherited from caller.
+ */
+static inline u8 ata_wait_idle(struct ata_port *ap)
+{
+	u8 status = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
+
+#ifdef ATA_DEBUG
+	if (status != 0xff && (status & (ATA_BUSY | ATA_DRQ)))
+		ata_port_printk(ap, KERN_DEBUG, "abnormal Status 0x%X\n",
+				status);
+#endif
+
+	return status;
+}
+
 #endif /* __LINUX_LIBATA_H__ */
-- 
cgit v1.2.3


From 071ce34d57924edb76b76f7de460eb4991463959 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 22:16:42 +0900
Subject: libata: move ata_pci_default_filter() out of CONFIG_PCI

ata_pci_default_filter() doesn't really have anything to do with PCI.
It's generally applicable to BMDMA controllers.  Move it out of
CONFIG_PCI.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-sff.c | 22 +++++++++++-----------
 include/linux/libata.h   |  4 ++--
 2 files changed, 13 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 840ae6da59bc..ebdd46bc13c4 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -2117,6 +2117,16 @@ void ata_std_ports(struct ata_ioports *ioaddr)
 	ioaddr->command_addr = ioaddr->cmd_addr + ATA_REG_CMD;
 }
 
+unsigned long ata_pci_default_filter(struct ata_device *adev, unsigned long xfer_mask)
+{
+	/* Filter out DMA modes if the device has been configured by
+	   the BIOS as PIO only */
+
+	if (adev->link->ap->ioaddr.bmdma_addr == NULL)
+		xfer_mask &= ~(ATA_MASK_MWDMA | ATA_MASK_UDMA);
+	return xfer_mask;
+}
+
 /**
  *	ata_bmdma_setup - Set up PCI IDE BMDMA transaction
  *	@qc: Info associated with this ATA transaction.
@@ -2335,16 +2345,6 @@ int ata_pci_clear_simplex(struct pci_dev *pdev)
 	return 0;
 }
 
-unsigned long ata_pci_default_filter(struct ata_device *adev, unsigned long xfer_mask)
-{
-	/* Filter out DMA modes if the device has been configured by
-	   the BIOS as PIO only */
-
-	if (adev->link->ap->ioaddr.bmdma_addr == NULL)
-		xfer_mask &= ~(ATA_MASK_MWDMA | ATA_MASK_UDMA);
-	return xfer_mask;
-}
-
 /**
  *	ata_pci_init_bmdma - acquire PCI BMDMA resources and init ATA host
  *	@host: target ATA host
@@ -2725,6 +2725,7 @@ EXPORT_SYMBOL_GPL(ata_sff_port_ops);
 EXPORT_SYMBOL_GPL(ata_bmdma_port_ops);
 EXPORT_SYMBOL_GPL(ata_qc_prep);
 EXPORT_SYMBOL_GPL(ata_dumb_qc_prep);
+EXPORT_SYMBOL_GPL(ata_pci_default_filter);
 EXPORT_SYMBOL_GPL(ata_std_dev_select);
 EXPORT_SYMBOL_GPL(ata_check_status);
 EXPORT_SYMBOL_GPL(ata_altstatus);
@@ -2760,7 +2761,6 @@ EXPORT_SYMBOL_GPL(ata_bmdma_status);
 EXPORT_SYMBOL_GPL(ata_bus_reset);
 #ifdef CONFIG_PCI
 EXPORT_SYMBOL_GPL(ata_pci_clear_simplex);
-EXPORT_SYMBOL_GPL(ata_pci_default_filter);
 EXPORT_SYMBOL_GPL(ata_pci_init_bmdma);
 EXPORT_SYMBOL_GPL(ata_pci_init_sff_host);
 EXPORT_SYMBOL_GPL(ata_pci_prepare_sff_host);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 673f34b256ba..53b8db05a1fb 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1378,6 +1378,8 @@ extern void ata_bmdma_error_handler(struct ata_port *ap);
 extern void ata_bmdma_post_internal_cmd(struct ata_queued_cmd *qc);
 extern int ata_sff_port_start(struct ata_port *ap);
 extern void ata_std_ports(struct ata_ioports *ioaddr);
+extern unsigned long ata_pci_default_filter(struct ata_device *dev,
+					    unsigned long xfer_mask);
 extern void ata_bmdma_setup(struct ata_queued_cmd *qc);
 extern void ata_bmdma_start(struct ata_queued_cmd *qc);
 extern void ata_bmdma_stop(struct ata_queued_cmd *qc);
@@ -1386,8 +1388,6 @@ extern void ata_bus_reset(struct ata_port *ap);
 
 #ifdef CONFIG_PCI
 extern int ata_pci_clear_simplex(struct pci_dev *pdev);
-extern unsigned long ata_pci_default_filter(struct ata_device *dev,
-					    unsigned long xfer_mask);
 extern int ata_pci_init_bmdma(struct ata_host *host);
 extern int ata_pci_init_sff_host(struct ata_host *host);
 extern int ata_pci_prepare_sff_host(struct pci_dev *pdev,
-- 
cgit v1.2.3


From 6fd36390117f7844ad147377878ddb52088f583a Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Tue, 25 Mar 2008 22:16:44 +0900
Subject: libata: kill ata_chk_status()

ata_chk_status() just calls ops->check_status and it only adds
confusion with other status functions.  Kill it.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-sff.c    | 22 +++++++++++-----------
 drivers/ata/pata_bf54x.c    |  2 +-
 drivers/ata/pata_scc.c      |  2 +-
 drivers/ata/sata_inic162x.c | 10 +++++-----
 drivers/ata/sata_sil.c      |  6 +++---
 drivers/ata/sata_via.c      |  2 +-
 drivers/ata/sata_vsc.c      |  2 +-
 include/linux/libata.h      |  7 +------
 8 files changed, 24 insertions(+), 29 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index ebdd46bc13c4..c601dcef2925 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -290,7 +290,7 @@ int ata_busy_sleep(struct ata_port *ap,
 	while (status != 0xff && (status & ATA_BUSY) &&
 	       time_before(jiffies, timeout)) {
 		msleep(50);
-		status = ata_chk_status(ap);
+		status = ap->ops->check_status(ap);
 	}
 
 	if (status == 0xff)
@@ -326,7 +326,7 @@ int ata_wait_ready(struct ata_port *ap, unsigned long deadline)
 	int warned = 0;
 
 	while (1) {
-		u8 status = ata_chk_status(ap);
+		u8 status = ap->ops->check_status(ap);
 		unsigned long now = jiffies;
 
 		if (!(status & ATA_BUSY))
@@ -1486,7 +1486,7 @@ inline unsigned int ata_host_intr(struct ata_port *ap,
 		goto idle_irq;
 
 	/* check main status, clearing INTRQ */
-	status = ata_chk_status(ap);
+	status = ap->ops->check_status(ap);
 	if (unlikely(status & ATA_BUSY))
 		goto idle_irq;
 
@@ -1506,7 +1506,7 @@ idle_irq:
 
 #ifdef ATA_IRQ_TRAP
 	if ((ap->stats.idle_irq % 1000) == 0) {
-		ata_chk_status(ap);
+		ap->ops->check_status(ap);
 		ap->ops->irq_clear(ap);
 		ata_port_printk(ap, KERN_WARNING, "irq trap\n");
 		return 1;
@@ -1582,7 +1582,7 @@ void ata_bmdma_freeze(struct ata_port *ap)
 	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
 	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
 	 */
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 
 	ap->ops->irq_clear(ap);
 }
@@ -1599,7 +1599,7 @@ void ata_bmdma_freeze(struct ata_port *ap)
 void ata_bmdma_thaw(struct ata_port *ap)
 {
 	/* clear & re-enable interrupts */
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 	ap->ops->irq_clear(ap);
 	ap->ops->irq_on(ap);
 }
@@ -1709,7 +1709,7 @@ unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
 			class = ATA_DEV_ATA;
 		else
 			class = ATA_DEV_NONE;
-	} else if ((class == ATA_DEV_ATA) && (ata_chk_status(ap) == 0))
+	} else if ((class == ATA_DEV_ATA) && (ap->ops->check_status(ap) == 0))
 		class = ATA_DEV_NONE;
 
 	return class;
@@ -1820,7 +1820,7 @@ void ata_wait_after_reset(struct ata_port *ap, unsigned long deadline)
 	 */
 	if (ap->flags & ATA_FLAG_SATA) {
 		while (1) {
-			u8 status = ata_chk_status(ap);
+			u8 status = ap->ops->check_status(ap);
 
 			if (status != 0xff || time_after(jiffies, deadline))
 				return;
@@ -1851,7 +1851,7 @@ static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
 	 * the bus shows 0xFF because the odd clown forgets the D7
 	 * pulldown resistor.
 	 */
-	if (ata_chk_status(ap) == 0xFF)
+	if (ap->ops->check_status(ap) == 0xFF)
 		return -ENODEV;
 
 	return ata_bus_post_reset(ap, devmask, deadline);
@@ -2034,7 +2034,7 @@ void ata_bmdma_error_handler(struct ata_port *ap)
 	}
 
 	ata_altstatus(ap);
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 	ap->ops->irq_clear(ap);
 
 	spin_unlock_irqrestore(ap->lock, flags);
@@ -2725,7 +2725,6 @@ EXPORT_SYMBOL_GPL(ata_sff_port_ops);
 EXPORT_SYMBOL_GPL(ata_bmdma_port_ops);
 EXPORT_SYMBOL_GPL(ata_qc_prep);
 EXPORT_SYMBOL_GPL(ata_dumb_qc_prep);
-EXPORT_SYMBOL_GPL(ata_pci_default_filter);
 EXPORT_SYMBOL_GPL(ata_std_dev_select);
 EXPORT_SYMBOL_GPL(ata_check_status);
 EXPORT_SYMBOL_GPL(ata_altstatus);
@@ -2754,6 +2753,7 @@ EXPORT_SYMBOL_GPL(ata_bmdma_error_handler);
 EXPORT_SYMBOL_GPL(ata_bmdma_post_internal_cmd);
 EXPORT_SYMBOL_GPL(ata_sff_port_start);
 EXPORT_SYMBOL_GPL(ata_std_ports);
+EXPORT_SYMBOL_GPL(ata_pci_default_filter);
 EXPORT_SYMBOL_GPL(ata_bmdma_setup);
 EXPORT_SYMBOL_GPL(ata_bmdma_start);
 EXPORT_SYMBOL_GPL(ata_bmdma_stop);
diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c
index eea275acb2a8..457ac800cd5e 100644
--- a/drivers/ata/pata_bf54x.c
+++ b/drivers/ata/pata_bf54x.c
@@ -1264,7 +1264,7 @@ static void bfin_bmdma_freeze(struct ata_port *ap)
 	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
 	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
 	 */
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 
 	bfin_irq_clear(ap);
 }
diff --git a/drivers/ata/pata_scc.c b/drivers/ata/pata_scc.c
index 07f2d7a6f1a8..85d33637dff9 100644
--- a/drivers/ata/pata_scc.c
+++ b/drivers/ata/pata_scc.c
@@ -854,7 +854,7 @@ static void scc_bmdma_freeze (struct ata_port *ap)
 	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
 	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
 	 */
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 
 	ap->ops->irq_clear(ap);
 }
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index ba1c09953517..12fbf3868fe2 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -267,14 +267,14 @@ static void inic_host_intr(struct ata_port *ap)
 			ata_qc_from_tag(ap, ap->link.active_tag);
 
 		if (unlikely(!qc || (qc->tf.flags & ATA_TFLAG_POLLING))) {
-			ata_chk_status(ap);	/* clear ATA interrupt */
+			ap->ops->check_status(ap); /* clear ATA interrupt */
 			return;
 		}
 
 		if (likely(ata_host_intr(ap, qc)))
 			return;
 
-		ata_chk_status(ap);	/* clear ATA interrupt */
+		ap->ops->check_status(ap); /* clear ATA interrupt */
 		ata_port_printk(ap, KERN_WARNING, "unhandled "
 				"interrupt, irq_stat=%x\n", irq_stat);
 		return;
@@ -351,7 +351,7 @@ static unsigned int inic_qc_issue(struct ata_queued_cmd *qc)
 	 */
 	if (unlikely(qc->tf.command == ATA_CMD_ID_ATA ||
 		     qc->tf.command == ATA_CMD_ID_ATAPI)) {
-		u8 stat = ata_chk_status(ap);
+		u8 stat = ap->ops->check_status(ap);
 		if (stat == 0x7f || stat == 0xff)
 			return AC_ERR_HSM;
 	}
@@ -365,7 +365,7 @@ static void inic_freeze(struct ata_port *ap)
 
 	__inic_set_pirq_mask(ap, PIRQ_MASK_FREEZE);
 
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 	writeb(0xff, port_base + PORT_IRQ_STAT);
 
 	readb(port_base + PORT_IRQ_STAT); /* flush */
@@ -375,7 +375,7 @@ static void inic_thaw(struct ata_port *ap)
 {
 	void __iomem *port_base = inic_port_base(ap);
 
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 	writeb(0xff, port_base + PORT_IRQ_STAT);
 
 	__inic_set_pirq_mask(ap, PIRQ_MASK_OTHER);
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index eac7ca73cfa0..659dfcbdc1b2 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -369,7 +369,7 @@ static void sil_host_intr(struct ata_port *ap, u32 bmdma2)
 
 	if (unlikely(!qc || (qc->tf.flags & ATA_TFLAG_POLLING))) {
 		/* this sometimes happens, just clear IRQ */
-		ata_chk_status(ap);
+		ap->ops->check_status(ap);
 		return;
 	}
 
@@ -405,7 +405,7 @@ static void sil_host_intr(struct ata_port *ap, u32 bmdma2)
 	}
 
 	/* check main status, clearing INTRQ */
-	status = ata_chk_status(ap);
+	status = ap->ops->check_status(ap);
 	if (unlikely(status & ATA_BUSY))
 		goto err_hsm;
 
@@ -480,7 +480,7 @@ static void sil_thaw(struct ata_port *ap)
 	u32 tmp;
 
 	/* clear IRQ */
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 	ata_bmdma_irq_clear(ap);
 
 	/* turn on SATA IRQ if supported */
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index 402fd7333d48..4bc6e849af2c 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -173,7 +173,7 @@ static void svia_noop_freeze(struct ata_port *ap)
 	/* Some VIA controllers choke if ATA_NIEN is manipulated in
 	 * certain way.  Leave it alone and just clear pending IRQ.
 	 */
-	ata_chk_status(ap);
+	ap->ops->check_status(ap);
 	ata_bmdma_irq_clear(ap);
 }
 
diff --git a/drivers/ata/sata_vsc.c b/drivers/ata/sata_vsc.c
index 8045a72dc559..fb3a88722664 100644
--- a/drivers/ata/sata_vsc.c
+++ b/drivers/ata/sata_vsc.c
@@ -251,7 +251,7 @@ static void vsc_port_intr(u8 port_status, struct ata_port *ap)
 	 * simply clear the interrupt
 	 */
 	if (unlikely(!handled))
-		ata_chk_status(ap);
+		ap->ops->check_status(ap);
 }
 
 /*
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 53b8db05a1fb..61a7f8d06971 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1401,11 +1401,6 @@ extern int ata_pci_init_one(struct pci_dev *pdev,
 			    struct scsi_host_template *sht, void *host_priv);
 #endif /* CONFIG_PCI */
 
-static inline u8 ata_chk_status(struct ata_port *ap)
-{
-	return ap->ops->check_status(ap);
-}
-
 /**
  *	ata_pause - Flush writes and pause 400 nanoseconds.
  *	@ap: Port to wait for.
@@ -1439,7 +1434,7 @@ static inline u8 ata_busy_wait(struct ata_port *ap, unsigned int bits,
 
 	do {
 		udelay(10);
-		status = ata_chk_status(ap);
+		status = ap->ops->check_status(ap);
 		max--;
 	} while (status != 0xff && (status & bits) && (max > 0));
 
-- 
cgit v1.2.3


From 03faab7827e4e45823fd27c47b84c133e20a0cd0 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Thu, 27 Mar 2008 19:14:24 +0900
Subject: libata: implement ATA_QCFLAG_RETRY

Currently whether a command should be retried after failure is
determined inside ata_eh_finish().  Add ATA_QCFLAG_RETRY and move the
logic into ata_eh_autopsy().  This makes things clearer and helps
extending retry determination logic.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 drivers/ata/libata-eh.c | 18 ++++++++----------
 include/linux/libata.h  |  1 +
 2 files changed, 9 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index ec32082356cb..cc8548e1572a 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -1785,6 +1785,11 @@ static void ata_eh_link_autopsy(struct ata_link *link)
 		if (qc->flags & ATA_QCFLAG_SENSE_VALID)
 			qc->err_mask &= ~(AC_ERR_DEV | AC_ERR_OTHER);
 
+		/* determine whether the command is worth retrying */
+		if (!(qc->err_mask & AC_ERR_INVALID) &&
+		    ((qc->flags & ATA_QCFLAG_IO) || qc->err_mask != AC_ERR_DEV))
+			qc->flags |= ATA_QCFLAG_RETRY;
+
 		/* accumulate error info */
 		ehc->i.dev = qc->dev;
 		all_err_mask |= qc->err_mask;
@@ -2783,18 +2788,11 @@ void ata_eh_finish(struct ata_port *ap)
 			/* FIXME: Once EH migration is complete,
 			 * generate sense data in this function,
 			 * considering both err_mask and tf.
-			 *
-			 * There's no point in retrying invalid
-			 * (detected by libata) and non-IO device
-			 * errors (rejected by device).  Finish them
-			 * immediately.
 			 */
-			if ((qc->err_mask & AC_ERR_INVALID) ||
-			    (!(qc->flags & ATA_QCFLAG_IO) &&
-			     qc->err_mask == AC_ERR_DEV))
-				ata_eh_qc_complete(qc);
-			else
+			if (qc->flags & ATA_QCFLAG_RETRY)
 				ata_eh_qc_retry(qc);
+			else
+				ata_eh_qc_complete(qc);
 		} else {
 			if (qc->flags & ATA_QCFLAG_SENSE_VALID) {
 				ata_eh_qc_complete(qc);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 61a7f8d06971..b25ea6ab1be9 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -224,6 +224,7 @@ enum {
 	ATA_QCFLAG_RESULT_TF	= (1 << 4), /* result TF requested */
 	ATA_QCFLAG_CLEAR_EXCL	= (1 << 5), /* clear excl_link on completion */
 	ATA_QCFLAG_QUIET	= (1 << 6), /* don't report device error */
+	ATA_QCFLAG_RETRY	= (1 << 7), /* retry after failure */
 
 	ATA_QCFLAG_FAILED	= (1 << 16), /* cmd failed and is owned by EH */
 	ATA_QCFLAG_SENSE_VALID	= (1 << 17), /* sense data valid */
-- 
cgit v1.2.3


From 83c063dd730cb56bf3fc89b70250ff9a398fec1e Mon Sep 17 00:00:00 2001
From: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Date: Thu, 28 Feb 2008 21:43:13 +0900
Subject: use ATA_TAG_INTERNAL in ata_tag_internal()

It should be ATA_TAG_INTERNAL.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
---
 include/linux/libata.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/libata.h b/include/linux/libata.h
index b25ea6ab1be9..a05de2ba7a72 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1121,7 +1121,7 @@ static inline unsigned int ata_tag_valid(unsigned int tag)
 
 static inline unsigned int ata_tag_internal(unsigned int tag)
 {
-	return tag == ATA_MAX_QUEUE - 1;
+	return tag == ATA_TAG_INTERNAL;
 }
 
 /*
-- 
cgit v1.2.3


From 9363c3825ea9ad76561eb48a395349dd29211ed6 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:16 +0900
Subject: libata: rename SFF functions

SFF functions have confusing names.  Some have sff prefix, some have
bmdma, some std, some pci and some none.  Unify the naming by...

* SFF functions which are common to both BMDMA and non-BMDMA are
  prefixed with ata_sff_.

* SFF functions which are specific to BMDMA are prefixed with
  ata_bmdma_.

* SFF functions which are specific to PCI but apply to both BMDMA and
  non-BMDMA are prefixed with ata_pci_sff_.

* SFF functions which are specific to PCI and BMDMA are prefixed with
  ata_pci_bmdma_.

* Drop generic prefixes from LLD specific routines.  For example,
  bfin_std_dev_select -> bfin_dev_select.

The following renames are noteworthy.

  ata_qc_issue_prot() -> ata_sff_qc_issue()
  ata_pci_default_filter() -> ata_bmdma_mode_filter()
  ata_dev_try_classify() -> ata_sff_dev_classify()

This rename is in preparation of separating SFF support out of libata
core layer.  This patch strictly renames functions and doesn't
introduce any behavior difference.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c              |  10 +-
 drivers/ata/ata_generic.c       |   4 +-
 drivers/ata/ata_piix.c          |   6 +-
 drivers/ata/libata-core.c       |  16 +-
 drivers/ata/libata-eh.c         |   2 +-
 drivers/ata/libata-sff.c        | 344 ++++++++++++++++++++--------------------
 drivers/ata/pata_acpi.c         |  16 +-
 drivers/ata/pata_ali.c          |   6 +-
 drivers/ata/pata_amd.c          |  10 +-
 drivers/ata/pata_artop.c        |   6 +-
 drivers/ata/pata_at32.c         |   2 +-
 drivers/ata/pata_atiixp.c       |   6 +-
 drivers/ata/pata_bf54x.c        |  84 +++++-----
 drivers/ata/pata_cmd640.c       |  16 +-
 drivers/ata/pata_cmd64x.c       |   4 +-
 drivers/ata/pata_cs5520.c       |   8 +-
 drivers/ata/pata_cs5530.c       |  12 +-
 drivers/ata/pata_cs5535.c       |   2 +-
 drivers/ata/pata_cs5536.c       |   2 +-
 drivers/ata/pata_cypress.c      |   2 +-
 drivers/ata/pata_efar.c         |   4 +-
 drivers/ata/pata_hpt366.c       |   4 +-
 drivers/ata/pata_hpt37x.c       |  10 +-
 drivers/ata/pata_hpt3x2n.c      |  10 +-
 drivers/ata/pata_hpt3x3.c       |   6 +-
 drivers/ata/pata_icside.c       |   8 +-
 drivers/ata/pata_isapnp.c       |   4 +-
 drivers/ata/pata_it8213.c       |   4 +-
 drivers/ata/pata_it821x.c       |  20 +--
 drivers/ata/pata_ixp4xx_cf.c    |   4 +-
 drivers/ata/pata_jmicron.c      |   4 +-
 drivers/ata/pata_legacy.c       |  32 ++--
 drivers/ata/pata_marvell.c      |   4 +-
 drivers/ata/pata_mpc52xx.c      |   4 +-
 drivers/ata/pata_mpiix.c        |  18 +--
 drivers/ata/pata_netcell.c      |   4 +-
 drivers/ata/pata_ninja32.c      |   6 +-
 drivers/ata/pata_ns87410.c      |  12 +-
 drivers/ata/pata_ns87415.c      |   8 +-
 drivers/ata/pata_oldpiix.c      |  12 +-
 drivers/ata/pata_opti.c         |   4 +-
 drivers/ata/pata_optidma.c      |   4 +-
 drivers/ata/pata_pcmcia.c       |   6 +-
 drivers/ata/pata_pdc2027x.c     |  10 +-
 drivers/ata/pata_pdc202xx_old.c |   2 +-
 drivers/ata/pata_platform.c     |   4 +-
 drivers/ata/pata_qdi.c          |  14 +-
 drivers/ata/pata_radisys.c      |  10 +-
 drivers/ata/pata_rb500_cf.c     |   6 +-
 drivers/ata/pata_rz1000.c       |   2 +-
 drivers/ata/pata_sc1200.c       |  12 +-
 drivers/ata/pata_scc.c          |  80 +++++-----
 drivers/ata/pata_serverworks.c  |  12 +-
 drivers/ata/pata_sil680.c       |  10 +-
 drivers/ata/pata_sis.c          |   4 +-
 drivers/ata/pata_sl82c105.c     |   4 +-
 drivers/ata/pata_triflex.c      |   4 +-
 drivers/ata/pata_via.c          |   6 +-
 drivers/ata/pata_winbond.c      |   6 +-
 drivers/ata/pdc_adma.c          |  12 +-
 drivers/ata/sata_inic162x.c     |  12 +-
 drivers/ata/sata_mv.c           |   6 +-
 drivers/ata/sata_nv.c           |  32 ++--
 drivers/ata/sata_promise.c      |  12 +-
 drivers/ata/sata_qstor.c        |  10 +-
 drivers/ata/sata_sil.c          |   8 +-
 drivers/ata/sata_sis.c          |   6 +-
 drivers/ata/sata_svw.c          |   4 +-
 drivers/ata/sata_sx4.c          |  26 +--
 drivers/ata/sata_uli.c          |  12 +-
 drivers/ata/sata_via.c          |  12 +-
 drivers/ata/sata_vsc.c          |   4 +-
 include/linux/libata.h          | 103 ++++++------
 73 files changed, 599 insertions(+), 596 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 3efa9904f7a0..771509c9a3fb 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1303,9 +1303,9 @@ static int ahci_do_softreset(struct ata_link *link, unsigned int *class,
 	ahci_exec_polled_cmd(ap, pmp, &tf, 0, 0, 0);
 
 	/* wait a while before checking status */
-	ata_wait_after_reset(ap, deadline);
+	ata_sff_wait_after_reset(ap, deadline);
 
-	rc = ata_wait_ready(ap, deadline);
+	rc = ata_sff_wait_ready(ap, deadline);
 	/* link occupied, -ENODEV too is an error */
 	if (rc) {
 		reason = "device not ready";
@@ -1350,7 +1350,7 @@ static int ahci_hardreset(struct ata_link *link, unsigned int *class,
 	tf.command = 0x80;
 	ata_tf_to_fis(&tf, 0, 0, d2h_fis);
 
-	rc = sata_std_hardreset(link, class, deadline);
+	rc = sata_sff_hardreset(link, class, deadline);
 
 	ahci_start_engine(ap);
 
@@ -1431,7 +1431,7 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 	 * have to be reset again.  For most cases, this should
 	 * suffice while making probing snappish enough.
 	 */
-	rc = ata_wait_ready(ap, jiffies + 2 * HZ);
+	rc = ata_sff_wait_ready(ap, jiffies + 2 * HZ);
 	if (rc)
 		ahci_kick_engine(ap, 0);
 
@@ -1444,7 +1444,7 @@ static void ahci_postreset(struct ata_link *link, unsigned int *class)
 	void __iomem *port_mmio = ahci_port_base(ap);
 	u32 new_tmp, tmp;
 
-	ata_std_postreset(link, class);
+	ata_sff_postreset(link, class);
 
 	/* Make sure port's ATAPI bit is set appropriately */
 	new_tmp = tmp = readl(port_mmio + PORT_CMD);
diff --git a/drivers/ata/ata_generic.c b/drivers/ata/ata_generic.c
index b23e2a1099c5..47aeccd52fa9 100644
--- a/drivers/ata/ata_generic.c
+++ b/drivers/ata/ata_generic.c
@@ -150,9 +150,9 @@ static int ata_generic_init_one(struct pci_dev *dev, const struct pci_device_id
 		return -ENODEV;
 
 	if (dev->vendor == PCI_VENDOR_ID_AL)
-	    	ata_pci_clear_simplex(dev);
+		ata_pci_bmdma_clear_simplex(dev);
 
-	return ata_pci_init_one(dev, ppi, &generic_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &generic_sht, NULL);
 }
 
 static struct pci_device_id ata_generic[] = {
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index e113f2f80275..7ab76a413cdf 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -629,7 +629,7 @@ static int piix_pata_prereset(struct ata_link *link, unsigned long deadline)
 
 	if (!pci_test_config_bits(pdev, &piix_enable_bits[ap->port_no]))
 		return -ENOENT;
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -1493,7 +1493,7 @@ static int __devinit piix_init_one(struct pci_dev *pdev,
 		hpriv->map = piix_init_sata_map(pdev, port_info,
 					piix_map_db_table[ent->driver_data]);
 
-	rc = ata_pci_prepare_sff_host(pdev, ppi, &host);
+	rc = ata_pci_sff_prepare_host(pdev, ppi, &host);
 	if (rc)
 		return rc;
 	host->private_data = hpriv;
@@ -1527,7 +1527,7 @@ static int __devinit piix_init_one(struct pci_dev *pdev,
 	}
 
 	pci_set_master(pdev);
-	return ata_pci_activate_sff_host(host, ata_interrupt, &piix_sht);
+	return ata_pci_sff_activate_host(host, ata_sff_interrupt, &piix_sht);
 }
 
 static int __init piix_init(void)
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index fa205d7c5bea..cbdbfb5eaeaa 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -75,9 +75,9 @@ const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
 const struct ata_port_operations ata_base_port_ops = {
 	.irq_clear		= ata_noop_irq_clear,
-	.prereset		= ata_std_prereset,
-	.hardreset		= sata_std_hardreset,
-	.postreset		= ata_std_postreset,
+	.prereset		= ata_sff_prereset,
+	.hardreset		= sata_sff_hardreset,
+	.postreset		= ata_sff_postreset,
 	.error_handler		= ata_std_error_handler,
 };
 
@@ -3425,7 +3425,7 @@ int sata_link_resume(struct ata_link *link, const unsigned long *params,
 }
 
 /**
- *	ata_std_prereset - prepare for reset
+ *	ata_sff_prereset - prepare for reset
  *	@link: ATA link to be reset
  *	@deadline: deadline jiffies for the operation
  *
@@ -3441,7 +3441,7 @@ int sata_link_resume(struct ata_link *link, const unsigned long *params,
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_std_prereset(struct ata_link *link, unsigned long deadline)
+int ata_sff_prereset(struct ata_link *link, unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
 	struct ata_eh_context *ehc = &link->eh_context;
@@ -3463,7 +3463,7 @@ int ata_std_prereset(struct ata_link *link, unsigned long deadline)
 
 	/* wait for !BSY if we don't know that no device is attached */
 	if (!ata_link_offline(link)) {
-		rc = ata_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(ap, deadline);
 		if (rc && rc != -ENODEV) {
 			ata_link_printk(link, KERN_WARNING, "device not ready "
 					"(errno=%d), forcing hardreset\n", rc);
@@ -3535,7 +3535,7 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
 }
 
 /**
- *	ata_std_postreset - standard postreset callback
+ *	ata_sff_postreset - standard postreset callback
  *	@link: the target ata_link
  *	@classes: classes of attached devices
  *
@@ -3546,7 +3546,7 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
  *	LOCKING:
  *	Kernel thread context (may sleep)
  */
-void ata_std_postreset(struct ata_link *link, unsigned int *classes)
+void ata_sff_postreset(struct ata_link *link, unsigned int *classes)
 {
 	struct ata_port *ap = link->ap;
 	u32 serror;
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index cc8548e1572a..f4f9c2783821 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -2861,7 +2861,7 @@ void ata_std_error_handler(struct ata_port *ap)
 	 * ata_base_port_ops.  Ignore it if SCR access is not
 	 * available.
 	 */
-	if (hardreset == sata_std_hardreset && !sata_scr_valid(&ap->link))
+	if (hardreset == sata_sff_hardreset && !sata_scr_valid(&ap->link))
 		hardreset = NULL;
 
 	ata_do_eh(ap, ops->prereset, ops->softreset, hardreset, ops->postreset);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 5208ca21f634..8544321293d4 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -42,22 +42,22 @@
 const struct ata_port_operations ata_sff_port_ops = {
 	.inherits		= &ata_base_port_ops,
 
-	.qc_prep		= ata_qc_prep,
-	.qc_issue		= ata_qc_issue_prot,
-
-	.freeze			= ata_bmdma_freeze,
-	.thaw			= ata_bmdma_thaw,
-	.softreset		= ata_std_softreset,
-	.error_handler		= ata_bmdma_error_handler,
-	.post_internal_cmd	= ata_bmdma_post_internal_cmd,
-
-	.dev_select		= ata_std_dev_select,
-	.check_status		= ata_check_status,
-	.tf_load		= ata_tf_load,
-	.tf_read		= ata_tf_read,
-	.exec_command		= ata_exec_command,
-	.data_xfer		= ata_data_xfer,
-	.irq_on			= ata_irq_on,
+	.qc_prep		= ata_sff_qc_prep,
+	.qc_issue		= ata_sff_qc_issue,
+
+	.freeze			= ata_sff_freeze,
+	.thaw			= ata_sff_thaw,
+	.softreset		= ata_sff_softreset,
+	.error_handler		= ata_sff_error_handler,
+	.post_internal_cmd	= ata_sff_post_internal_cmd,
+
+	.dev_select		= ata_sff_dev_select,
+	.check_status		= ata_sff_check_status,
+	.tf_load		= ata_sff_tf_load,
+	.tf_read		= ata_sff_tf_read,
+	.exec_command		= ata_sff_exec_command,
+	.data_xfer		= ata_sff_data_xfer,
+	.irq_on			= ata_sff_irq_on,
 
 	.port_start		= ata_sff_port_start,
 };
@@ -65,13 +65,13 @@ const struct ata_port_operations ata_sff_port_ops = {
 const struct ata_port_operations ata_bmdma_port_ops = {
 	.inherits		= &ata_sff_port_ops,
 
-	.mode_filter		= ata_pci_default_filter,
+	.mode_filter		= ata_bmdma_mode_filter,
 
 	.bmdma_setup		= ata_bmdma_setup,
 	.bmdma_start		= ata_bmdma_start,
 	.bmdma_stop		= ata_bmdma_stop,
 	.bmdma_status		= ata_bmdma_status,
-	.irq_clear		= ata_bmdma_irq_clear,
+	.irq_clear		= ata_sff_irq_clear,
 };
 
 /**
@@ -181,7 +181,7 @@ static void ata_fill_sg_dumb(struct ata_queued_cmd *qc)
 }
 
 /**
- *	ata_qc_prep - Prepare taskfile for submission
+ *	ata_sff_qc_prep - Prepare taskfile for submission
  *	@qc: Metadata associated with taskfile to be prepared
  *
  *	Prepare ATA taskfile for submission.
@@ -189,7 +189,7 @@ static void ata_fill_sg_dumb(struct ata_queued_cmd *qc)
  *	LOCKING:
  *	spin_lock_irqsave(host lock)
  */
-void ata_qc_prep(struct ata_queued_cmd *qc)
+void ata_sff_qc_prep(struct ata_queued_cmd *qc)
 {
 	if (!(qc->flags & ATA_QCFLAG_DMAMAP))
 		return;
@@ -198,7 +198,7 @@ void ata_qc_prep(struct ata_queued_cmd *qc)
 }
 
 /**
- *	ata_dumb_qc_prep - Prepare taskfile for submission
+ *	ata_sff_dumb_qc_prep - Prepare taskfile for submission
  *	@qc: Metadata associated with taskfile to be prepared
  *
  *	Prepare ATA taskfile for submission.
@@ -206,7 +206,7 @@ void ata_qc_prep(struct ata_queued_cmd *qc)
  *	LOCKING:
  *	spin_lock_irqsave(host lock)
  */
-void ata_dumb_qc_prep(struct ata_queued_cmd *qc)
+void ata_sff_dumb_qc_prep(struct ata_queued_cmd *qc)
 {
 	if (!(qc->flags & ATA_QCFLAG_DMAMAP))
 		return;
@@ -215,7 +215,7 @@ void ata_dumb_qc_prep(struct ata_queued_cmd *qc)
 }
 
 /**
- *	ata_check_status - Read device status reg & clear interrupt
+ *	ata_sff_check_status - Read device status reg & clear interrupt
  *	@ap: port where the device is
  *
  *	Reads ATA taskfile status register for currently-selected device
@@ -225,13 +225,13 @@ void ata_dumb_qc_prep(struct ata_queued_cmd *qc)
  *	LOCKING:
  *	Inherited from caller.
  */
-u8 ata_check_status(struct ata_port *ap)
+u8 ata_sff_check_status(struct ata_port *ap)
 {
 	return ioread8(ap->ioaddr.status_addr);
 }
 
 /**
- *	ata_altstatus - Read device alternate status reg
+ *	ata_sff_altstatus - Read device alternate status reg
  *	@ap: port where the device is
  *
  *	Reads ATA taskfile alternate status register for
@@ -243,7 +243,7 @@ u8 ata_check_status(struct ata_port *ap)
  *	LOCKING:
  *	Inherited from caller.
  */
-u8 ata_altstatus(struct ata_port *ap)
+u8 ata_sff_altstatus(struct ata_port *ap)
 {
 	if (ap->ops->check_altstatus)
 		return ap->ops->check_altstatus(ap);
@@ -252,7 +252,7 @@ u8 ata_altstatus(struct ata_port *ap)
 }
 
 /**
- *	ata_busy_sleep - sleep until BSY clears, or timeout
+ *	ata_sff_busy_sleep - sleep until BSY clears, or timeout
  *	@ap: port containing status register to be polled
  *	@tmout_pat: impatience timeout
  *	@tmout: overall timeout
@@ -266,19 +266,19 @@ u8 ata_altstatus(struct ata_port *ap)
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_busy_sleep(struct ata_port *ap,
-		   unsigned long tmout_pat, unsigned long tmout)
+int ata_sff_busy_sleep(struct ata_port *ap,
+		       unsigned long tmout_pat, unsigned long tmout)
 {
 	unsigned long timer_start, timeout;
 	u8 status;
 
-	status = ata_busy_wait(ap, ATA_BUSY, 300);
+	status = ata_sff_busy_wait(ap, ATA_BUSY, 300);
 	timer_start = jiffies;
 	timeout = timer_start + tmout_pat;
 	while (status != 0xff && (status & ATA_BUSY) &&
 	       time_before(jiffies, timeout)) {
 		msleep(50);
-		status = ata_busy_wait(ap, ATA_BUSY, 3);
+		status = ata_sff_busy_wait(ap, ATA_BUSY, 3);
 	}
 
 	if (status != 0xff && (status & ATA_BUSY))
@@ -307,7 +307,7 @@ int ata_busy_sleep(struct ata_port *ap,
 }
 
 /**
- *	ata_wait_ready - sleep until BSY clears, or timeout
+ *	ata_sff_wait_ready - sleep until BSY clears, or timeout
  *	@ap: port containing status register to be polled
  *	@deadline: deadline jiffies for the operation
  *
@@ -320,7 +320,7 @@ int ata_busy_sleep(struct ata_port *ap,
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_wait_ready(struct ata_port *ap, unsigned long deadline)
+int ata_sff_wait_ready(struct ata_port *ap, unsigned long deadline)
 {
 	unsigned long start = jiffies;
 	int warned = 0;
@@ -349,7 +349,7 @@ int ata_wait_ready(struct ata_port *ap, unsigned long deadline)
 }
 
 /**
- *	ata_std_dev_select - Select device 0/1 on ATA bus
+ *	ata_sff_dev_select - Select device 0/1 on ATA bus
  *	@ap: ATA channel to manipulate
  *	@device: ATA device (numbered from zero) to select
  *
@@ -362,7 +362,7 @@ int ata_wait_ready(struct ata_port *ap, unsigned long deadline)
  *	LOCKING:
  *	caller.
  */
-void ata_std_dev_select(struct ata_port *ap, unsigned int device)
+void ata_sff_dev_select(struct ata_port *ap, unsigned int device)
 {
 	u8 tmp;
 
@@ -372,7 +372,7 @@ void ata_std_dev_select(struct ata_port *ap, unsigned int device)
 		tmp = ATA_DEVICE_OBS | ATA_DEV1;
 
 	iowrite8(tmp, ap->ioaddr.device_addr);
-	ata_pause(ap);		/* needed; also flushes, for mmio */
+	ata_sff_pause(ap);	/* needed; also flushes, for mmio */
 }
 
 /**
@@ -386,9 +386,9 @@ void ata_std_dev_select(struct ata_port *ap, unsigned int device)
  *	make either device 0, or device 1, active on the
  *	ATA channel.
  *
- *	This is a high-level version of ata_std_dev_select(),
- *	which additionally provides the services of inserting
- *	the proper pauses and status polling, where needed.
+ *	This is a high-level version of ata_sff_dev_select(), which
+ *	additionally provides the services of inserting the proper
+ *	pauses and status polling, where needed.
  *
  *	LOCKING:
  *	caller.
@@ -413,7 +413,7 @@ void ata_dev_select(struct ata_port *ap, unsigned int device,
 }
 
 /**
- *	ata_irq_on - Enable interrupts on a port.
+ *	ata_sff_irq_on - Enable interrupts on a port.
  *	@ap: Port on which interrupts are enabled.
  *
  *	Enable interrupts on a legacy IDE device using MMIO or PIO,
@@ -422,7 +422,7 @@ void ata_dev_select(struct ata_port *ap, unsigned int device,
  *	LOCKING:
  *	Inherited from caller.
  */
-u8 ata_irq_on(struct ata_port *ap)
+u8 ata_sff_irq_on(struct ata_port *ap)
 {
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 	u8 tmp;
@@ -440,7 +440,7 @@ u8 ata_irq_on(struct ata_port *ap)
 }
 
 /**
- *	ata_bmdma_irq_clear - Clear PCI IDE BMDMA interrupt.
+ *	ata_sff_irq_clear - Clear PCI IDE BMDMA interrupt.
  *	@ap: Port associated with this ATA transaction.
  *
  *	Clear interrupt and error flags in DMA status register.
@@ -450,7 +450,7 @@ u8 ata_irq_on(struct ata_port *ap)
  *	LOCKING:
  *	spin_lock_irqsave(host lock)
  */
-void ata_bmdma_irq_clear(struct ata_port *ap)
+void ata_sff_irq_clear(struct ata_port *ap)
 {
 	void __iomem *mmio = ap->ioaddr.bmdma_addr;
 
@@ -461,7 +461,7 @@ void ata_bmdma_irq_clear(struct ata_port *ap)
 }
 
 /**
- *	ata_tf_load - send taskfile registers to host controller
+ *	ata_sff_tf_load - send taskfile registers to host controller
  *	@ap: Port to which output is sent
  *	@tf: ATA taskfile register set
  *
@@ -470,7 +470,7 @@ void ata_bmdma_irq_clear(struct ata_port *ap)
  *	LOCKING:
  *	Inherited from caller.
  */
-void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
+void ata_sff_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
 {
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 	unsigned int is_addr = tf->flags & ATA_TFLAG_ISADDR;
@@ -520,7 +520,7 @@ void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
 }
 
 /**
- *	ata_tf_read - input device's ATA taskfile shadow registers
+ *	ata_sff_tf_read - input device's ATA taskfile shadow registers
  *	@ap: Port from which input is read
  *	@tf: ATA taskfile register set for storing input
  *
@@ -532,11 +532,11 @@ void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
  *	LOCKING:
  *	Inherited from caller.
  */
-void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
+void ata_sff_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
 {
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 
-	tf->command = ata_check_status(ap);
+	tf->command = ata_sff_check_status(ap);
 	tf->feature = ioread8(ioaddr->error_addr);
 	tf->nsect = ioread8(ioaddr->nsect_addr);
 	tf->lbal = ioread8(ioaddr->lbal_addr);
@@ -560,7 +560,7 @@ void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
 }
 
 /**
- *	ata_exec_command - issue ATA command to host controller
+ *	ata_sff_exec_command - issue ATA command to host controller
  *	@ap: port to which command is being issued
  *	@tf: ATA taskfile register set
  *
@@ -570,12 +570,12 @@ void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
  *	LOCKING:
  *	spin_lock_irqsave(host lock)
  */
-void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf)
+void ata_sff_exec_command(struct ata_port *ap, const struct ata_taskfile *tf)
 {
 	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
 
 	iowrite8(tf->command, ap->ioaddr.command_addr);
-	ata_pause(ap);
+	ata_sff_pause(ap);
 }
 
 /**
@@ -598,7 +598,7 @@ static inline void ata_tf_to_host(struct ata_port *ap,
 }
 
 /**
- *	ata_data_xfer - Transfer data by PIO
+ *	ata_sff_data_xfer - Transfer data by PIO
  *	@dev: device to target
  *	@buf: data buffer
  *	@buflen: buffer length
@@ -612,8 +612,8 @@ static inline void ata_tf_to_host(struct ata_port *ap,
  *	RETURNS:
  *	Bytes consumed.
  */
-unsigned int ata_data_xfer(struct ata_device *dev, unsigned char *buf,
-			   unsigned int buflen, int rw)
+unsigned int ata_sff_data_xfer(struct ata_device *dev, unsigned char *buf,
+			       unsigned int buflen, int rw)
 {
 	struct ata_port *ap = dev->link->ap;
 	void __iomem *data_addr = ap->ioaddr.data_addr;
@@ -644,7 +644,7 @@ unsigned int ata_data_xfer(struct ata_device *dev, unsigned char *buf,
 }
 
 /**
- *	ata_data_xfer_noirq - Transfer data by PIO
+ *	ata_sff_data_xfer_noirq - Transfer data by PIO
  *	@dev: device to target
  *	@buf: data buffer
  *	@buflen: buffer length
@@ -659,14 +659,14 @@ unsigned int ata_data_xfer(struct ata_device *dev, unsigned char *buf,
  *	RETURNS:
  *	Bytes consumed.
  */
-unsigned int ata_data_xfer_noirq(struct ata_device *dev, unsigned char *buf,
-				 unsigned int buflen, int rw)
+unsigned int ata_sff_data_xfer_noirq(struct ata_device *dev, unsigned char *buf,
+				     unsigned int buflen, int rw)
 {
 	unsigned long flags;
 	unsigned int consumed;
 
 	local_irq_save(flags);
-	consumed = ata_data_xfer(dev, buf, buflen, rw);
+	consumed = ata_sff_data_xfer(dev, buf, buflen, rw);
 	local_irq_restore(flags);
 
 	return consumed;
@@ -752,7 +752,7 @@ static void ata_pio_sectors(struct ata_queued_cmd *qc)
 	} else
 		ata_pio_sector(qc);
 
-	ata_altstatus(qc->ap); /* flush */
+	ata_sff_altstatus(qc->ap); /* flush */
 }
 
 /**
@@ -773,7 +773,7 @@ static void atapi_send_cdb(struct ata_port *ap, struct ata_queued_cmd *qc)
 	WARN_ON(qc->dev->cdb_len < 12);
 
 	ap->ops->data_xfer(qc->dev, qc->cdb, qc->dev->cdb_len, 1);
-	ata_altstatus(ap); /* flush */
+	ata_sff_altstatus(ap); /* flush */
 
 	switch (qc->tf.protocol) {
 	case ATAPI_PROT_PIO:
@@ -915,7 +915,7 @@ static void atapi_pio_bytes(struct ata_queued_cmd *qc)
 
 	if (unlikely(__atapi_pio_bytes(qc, bytes)))
 		goto err_out;
-	ata_altstatus(ap); /* flush */
+	ata_sff_altstatus(ap); /* flush */
 
 	return;
 
@@ -1004,7 +1004,7 @@ static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
 }
 
 /**
- *	ata_hsm_move - move the HSM to the next state.
+ *	ata_sff_hsm_move - move the HSM to the next state.
  *	@ap: the target ata_port
  *	@qc: qc on going
  *	@status: current device status
@@ -1013,15 +1013,15 @@ static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
  *	RETURNS:
  *	1 when poll next status needed, 0 otherwise.
  */
-int ata_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
-		 u8 status, int in_wq)
+int ata_sff_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
+		     u8 status, int in_wq)
 {
 	unsigned long flags = 0;
 	int poll_next;
 
 	WARN_ON((qc->flags & ATA_QCFLAG_ACTIVE) == 0);
 
-	/* Make sure ata_qc_issue_prot() does not throw things
+	/* Make sure ata_sff_qc_issue() does not throw things
 	 * like DMA polling into the workqueue. Notice that
 	 * in_wq is not equivalent to (qc->tf.flags & ATA_TFLAG_POLLING).
 	 */
@@ -1263,10 +1263,10 @@ fsm_start:
 	 * or something.  Snooze for a couple msecs, then
 	 * chk-status again.  If still busy, queue delayed work.
 	 */
-	status = ata_busy_wait(ap, ATA_BUSY, 5);
+	status = ata_sff_busy_wait(ap, ATA_BUSY, 5);
 	if (status & ATA_BUSY) {
 		msleep(2);
-		status = ata_busy_wait(ap, ATA_BUSY, 10);
+		status = ata_sff_busy_wait(ap, ATA_BUSY, 10);
 		if (status & ATA_BUSY) {
 			ata_pio_queue_task(ap, qc, ATA_SHORT_PAUSE);
 			return;
@@ -1274,7 +1274,7 @@ fsm_start:
 	}
 
 	/* move the HSM */
-	poll_next = ata_hsm_move(ap, qc, status, 1);
+	poll_next = ata_sff_hsm_move(ap, qc, status, 1);
 
 	/* another command or interrupt handler
 	 * may be running at this point.
@@ -1284,7 +1284,7 @@ fsm_start:
 }
 
 /**
- *	ata_qc_issue_prot - issue taskfile to device in proto-dependent manner
+ *	ata_sff_qc_issue - issue taskfile to device in proto-dependent manner
  *	@qc: command to issue to device
  *
  *	Using various libata functions and hooks, this function
@@ -1300,7 +1300,7 @@ fsm_start:
  *	RETURNS:
  *	Zero on success, AC_ERR_* mask on failure
  */
-unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc)
+unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 
@@ -1415,7 +1415,7 @@ unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc)
 }
 
 /**
- *	ata_host_intr - Handle host interrupt for given (port, task)
+ *	ata_sff_host_intr - Handle host interrupt for given (port, task)
  *	@ap: Port on which interrupt arrived (possibly...)
  *	@qc: Taskfile currently active in engine
  *
@@ -1429,8 +1429,8 @@ unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc)
  *	RETURNS:
  *	One if interrupt was handled, zero if not (shared irq).
  */
-inline unsigned int ata_host_intr(struct ata_port *ap,
-				  struct ata_queued_cmd *qc)
+inline unsigned int ata_sff_host_intr(struct ata_port *ap,
+				      struct ata_queued_cmd *qc)
 {
 	struct ata_eh_info *ehi = &ap->link.eh_info;
 	u8 status, host_stat = 0;
@@ -1481,7 +1481,7 @@ inline unsigned int ata_host_intr(struct ata_port *ap,
 	}
 
 	/* check altstatus */
-	status = ata_altstatus(ap);
+	status = ata_sff_altstatus(ap);
 	if (status & ATA_BUSY)
 		goto idle_irq;
 
@@ -1493,7 +1493,7 @@ inline unsigned int ata_host_intr(struct ata_port *ap,
 	/* ack bmdma irq events */
 	ap->ops->irq_clear(ap);
 
-	ata_hsm_move(ap, qc, status, 0);
+	ata_sff_hsm_move(ap, qc, status, 0);
 
 	if (unlikely(qc->err_mask) && (qc->tf.protocol == ATA_PROT_DMA ||
 				       qc->tf.protocol == ATAPI_PROT_DMA))
@@ -1516,12 +1516,12 @@ idle_irq:
 }
 
 /**
- *	ata_interrupt - Default ATA host interrupt handler
+ *	ata_sff_interrupt - Default ATA host interrupt handler
  *	@irq: irq line (unused)
  *	@dev_instance: pointer to our ata_host information structure
  *
  *	Default interrupt handler for PCI IDE devices.  Calls
- *	ata_host_intr() for each port that is not disabled.
+ *	ata_sff_host_intr() for each port that is not disabled.
  *
  *	LOCKING:
  *	Obtains host lock during operation.
@@ -1529,7 +1529,7 @@ idle_irq:
  *	RETURNS:
  *	IRQ_NONE or IRQ_HANDLED.
  */
-irqreturn_t ata_interrupt(int irq, void *dev_instance)
+irqreturn_t ata_sff_interrupt(int irq, void *dev_instance)
 {
 	struct ata_host *host = dev_instance;
 	unsigned int i;
@@ -1550,7 +1550,7 @@ irqreturn_t ata_interrupt(int irq, void *dev_instance)
 			qc = ata_qc_from_tag(ap, ap->link.active_tag);
 			if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)) &&
 			    (qc->flags & ATA_QCFLAG_ACTIVE))
-				handled |= ata_host_intr(ap, qc);
+				handled |= ata_sff_host_intr(ap, qc);
 		}
 	}
 
@@ -1560,7 +1560,7 @@ irqreturn_t ata_interrupt(int irq, void *dev_instance)
 }
 
 /**
- *	ata_bmdma_freeze - Freeze BMDMA controller port
+ *	ata_sff_freeze - Freeze SFF controller port
  *	@ap: port to freeze
  *
  *	Freeze BMDMA controller port.
@@ -1568,7 +1568,7 @@ irqreturn_t ata_interrupt(int irq, void *dev_instance)
  *	LOCKING:
  *	Inherited from caller.
  */
-void ata_bmdma_freeze(struct ata_port *ap)
+void ata_sff_freeze(struct ata_port *ap)
 {
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 
@@ -1588,15 +1588,15 @@ void ata_bmdma_freeze(struct ata_port *ap)
 }
 
 /**
- *	ata_bmdma_thaw - Thaw BMDMA controller port
+ *	ata_sff_thaw - Thaw SFF controller port
  *	@ap: port to thaw
  *
- *	Thaw BMDMA controller port.
+ *	Thaw SFF controller port.
  *
  *	LOCKING:
  *	Inherited from caller.
  */
-void ata_bmdma_thaw(struct ata_port *ap)
+void ata_sff_thaw(struct ata_port *ap)
 {
 	/* clear & re-enable interrupts */
 	ap->ops->check_status(ap);
@@ -1647,7 +1647,7 @@ static unsigned int ata_devchk(struct ata_port *ap, unsigned int device)
 }
 
 /**
- *	ata_dev_try_classify - Parse returned ATA device signature
+ *	ata_sff_dev_classify - Parse returned ATA device signature
  *	@dev: ATA device to classify (starting at zero)
  *	@present: device seems present
  *	@r_err: Value of error register on completion
@@ -1667,7 +1667,7 @@ static unsigned int ata_devchk(struct ata_port *ap, unsigned int device)
  *	RETURNS:
  *	Device type - %ATA_DEV_ATA, %ATA_DEV_ATAPI or %ATA_DEV_NONE.
  */
-unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
+unsigned int ata_sff_dev_classify(struct ata_device *dev, int present,
 				  u8 *r_err)
 {
 	struct ata_port *ap = dev->link->ap;
@@ -1727,7 +1727,7 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	 * BSY bit to clear
 	 */
 	if (dev0) {
-		rc = ata_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(ap, deadline);
 		if (rc) {
 			if (rc != -ENODEV)
 				return rc;
@@ -1757,7 +1757,7 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 			msleep(50);	/* give drive a breather */
 		}
 
-		rc = ata_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(ap, deadline);
 		if (rc) {
 			if (rc != -ENODEV)
 				return rc;
@@ -1776,7 +1776,7 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 }
 
 /**
- *	ata_wait_after_reset - wait before checking status after reset
+ *	ata_sff_wait_after_reset - wait before checking status after reset
  *	@ap: port containing status register to be polled
  *	@deadline: deadline jiffies for the operation
  *
@@ -1790,7 +1790,7 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
  *	LOCKING:
  *	Kernel thread context (may sleep).
  */
-void ata_wait_after_reset(struct ata_port *ap, unsigned long deadline)
+void ata_sff_wait_after_reset(struct ata_port *ap, unsigned long deadline)
 {
 	unsigned long until = jiffies + ATA_TMOUT_FF_WAIT;
 
@@ -1845,7 +1845,7 @@ static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
 	iowrite8(ap->ctl, ioaddr->ctl_addr);
 
 	/* wait a while before checking status */
-	ata_wait_after_reset(ap, deadline);
+	ata_sff_wait_after_reset(ap, deadline);
 
 	/* Before we perform post reset processing we want to see if
 	 * the bus shows 0xFF because the odd clown forgets the D7
@@ -1858,7 +1858,7 @@ static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
 }
 
 /**
- *	ata_std_softreset - reset host port via ATA SRST
+ *	ata_sff_softreset - reset host port via ATA SRST
  *	@link: ATA link to reset
  *	@classes: resulting classes of attached devices
  *	@deadline: deadline jiffies for the operation
@@ -1871,7 +1871,7 @@ static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_std_softreset(struct ata_link *link, unsigned int *classes,
+int ata_sff_softreset(struct ata_link *link, unsigned int *classes,
 		      unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
@@ -1906,10 +1906,10 @@ int ata_std_softreset(struct ata_link *link, unsigned int *classes,
 	}
 
 	/* determine by signature whether we have ATA or ATAPI devices */
-	classes[0] = ata_dev_try_classify(&link->device[0],
+	classes[0] = ata_sff_dev_classify(&link->device[0],
 					  devmask & (1 << 0), &err);
 	if (slave_possible && err != 0x81)
-		classes[1] = ata_dev_try_classify(&link->device[1],
+		classes[1] = ata_sff_dev_classify(&link->device[1],
 						  devmask & (1 << 1), &err);
 
  out:
@@ -1918,7 +1918,7 @@ int ata_std_softreset(struct ata_link *link, unsigned int *classes,
 }
 
 /**
- *	sata_std_hardreset - reset host port via SATA phy reset
+ *	sata_sff_hardreset - reset host port via SATA phy reset
  *	@link: link to reset
  *	@class: resulting class of attached device
  *	@deadline: deadline jiffies for the operation
@@ -1932,7 +1932,7 @@ int ata_std_softreset(struct ata_link *link, unsigned int *classes,
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int sata_std_hardreset(struct ata_link *link, unsigned int *class,
+int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
 		       unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
@@ -1957,7 +1957,7 @@ int sata_std_hardreset(struct ata_link *link, unsigned int *class,
 	}
 
 	/* wait a while before checking status */
-	ata_wait_after_reset(ap, deadline);
+	ata_sff_wait_after_reset(ap, deadline);
 
 	/* If PMP is supported, we have to do follow-up SRST.  Note
 	 * that some PMPs don't send D2H Reg FIS after hardreset at
@@ -1965,11 +1965,11 @@ int sata_std_hardreset(struct ata_link *link, unsigned int *class,
 	 * second and request follow-up SRST.
 	 */
 	if (ap->flags & ATA_FLAG_PMP) {
-		ata_wait_ready(ap, jiffies + HZ);
+		ata_sff_wait_ready(ap, jiffies + HZ);
 		return -EAGAIN;
 	}
 
-	rc = ata_wait_ready(ap, deadline);
+	rc = ata_sff_wait_ready(ap, deadline);
 	/* link occupied, -ENODEV too is an error */
 	if (rc) {
 		ata_link_printk(link, KERN_ERR,
@@ -1979,17 +1979,17 @@ int sata_std_hardreset(struct ata_link *link, unsigned int *class,
 
 	ap->ops->dev_select(ap, 0);	/* probably unnecessary */
 
-	*class = ata_dev_try_classify(link->device, 1, NULL);
+	*class = ata_sff_dev_classify(link->device, 1, NULL);
 
 	DPRINTK("EXIT, class=%u\n", *class);
 	return 0;
 }
 
 /**
- *	ata_bmdma_error_handler - Stock error handler for BMDMA controller
+ *	ata_sff_error_handler - Stock error handler for BMDMA controller
  *	@ap: port to handle error for
  *
- *	Stock error handler for BMDMA controller.  It can handle both
+ *	Stock error handler for SFF controller.  It can handle both
  *	PATA and SATA controllers.  Many controllers should be able to
  *	use this EH as-is or with some added handling before and
  *	after.
@@ -1997,7 +1997,7 @@ int sata_std_hardreset(struct ata_link *link, unsigned int *class,
  *	LOCKING:
  *	Kernel thread context (may sleep)
  */
-void ata_bmdma_error_handler(struct ata_port *ap)
+void ata_sff_error_handler(struct ata_port *ap)
 {
 	ata_reset_fn_t softreset = ap->ops->softreset;
 	ata_reset_fn_t hardreset = ap->ops->hardreset;
@@ -2034,7 +2034,7 @@ void ata_bmdma_error_handler(struct ata_port *ap)
 		ap->ops->bmdma_stop(qc);
 	}
 
-	ata_altstatus(ap);
+	ata_sff_altstatus(ap);
 	ap->ops->check_status(ap);
 	ap->ops->irq_clear(ap);
 
@@ -2045,14 +2045,14 @@ void ata_bmdma_error_handler(struct ata_port *ap)
 
 	/* PIO and DMA engines have been stopped, perform recovery */
 
-	/* ata_std_softreset and sata_std_hardreset are inherited to
+	/* ata_sff_softreset and sata_sff_hardreset are inherited to
 	 * all SFF drivers from ata_sff_port_ops.  Ignore softreset if
 	 * ctl isn't accessible.  Ignore hardreset if SCR access isn't
 	 * available.
 	 */
-	if (softreset == ata_std_softreset && !ap->ioaddr.ctl_addr)
+	if (softreset == ata_sff_softreset && !ap->ioaddr.ctl_addr)
 		softreset = NULL;
-	if (hardreset == sata_std_hardreset && !sata_scr_valid(&ap->link))
+	if (hardreset == sata_sff_hardreset && !sata_scr_valid(&ap->link))
 		hardreset = NULL;
 
 	ata_do_eh(ap, ap->ops->prereset, softreset, hardreset,
@@ -2060,14 +2060,13 @@ void ata_bmdma_error_handler(struct ata_port *ap)
 }
 
 /**
- *	ata_bmdma_post_internal_cmd - Stock post_internal_cmd for
- *				      BMDMA controller
+ *	ata_sff_post_internal_cmd - Stock post_internal_cmd for SFF controller
  *	@qc: internal command to clean up
  *
  *	LOCKING:
  *	Kernel thread context (may sleep)
  */
-void ata_bmdma_post_internal_cmd(struct ata_queued_cmd *qc)
+void ata_sff_post_internal_cmd(struct ata_queued_cmd *qc)
 {
 	if (qc->ap->ioaddr.bmdma_addr)
 		ata_bmdma_stop(qc);
@@ -2094,7 +2093,7 @@ int ata_sff_port_start(struct ata_port *ap)
 }
 
 /**
- *	ata_std_ports - initialize ioaddr with standard port offsets.
+ *	ata_sff_std_ports - initialize ioaddr with standard port offsets.
  *	@ioaddr: IO address structure to be initialized
  *
  *	Utility function which initializes data_addr, error_addr,
@@ -2104,7 +2103,7 @@ int ata_sff_port_start(struct ata_port *ap)
  *
  *	Does not set ctl_addr, altstatus_addr, bmdma_addr, or scr_addr.
  */
-void ata_std_ports(struct ata_ioports *ioaddr)
+void ata_sff_std_ports(struct ata_ioports *ioaddr)
 {
 	ioaddr->data_addr = ioaddr->cmd_addr + ATA_REG_DATA;
 	ioaddr->error_addr = ioaddr->cmd_addr + ATA_REG_ERR;
@@ -2118,7 +2117,8 @@ void ata_std_ports(struct ata_ioports *ioaddr)
 	ioaddr->command_addr = ioaddr->cmd_addr + ATA_REG_CMD;
 }
 
-unsigned long ata_pci_default_filter(struct ata_device *adev, unsigned long xfer_mask)
+unsigned long ata_bmdma_mode_filter(struct ata_device *adev,
+				    unsigned long xfer_mask)
 {
 	/* Filter out DMA modes if the device has been configured by
 	   the BIOS as PIO only */
@@ -2209,7 +2209,7 @@ void ata_bmdma_stop(struct ata_queued_cmd *qc)
 		 mmio + ATA_DMA_CMD);
 
 	/* one-PIO-cycle guaranteed wait, per spec, for HDMA1:0 transition */
-	ata_altstatus(ap);        /* dummy read */
+	ata_sff_altstatus(ap);        /* dummy read */
 }
 
 /**
@@ -2289,9 +2289,9 @@ void ata_bus_reset(struct ata_port *ap)
 	/*
 	 * determine by signature whether we have ATA or ATAPI devices
 	 */
-	device[0].class = ata_dev_try_classify(&device[0], dev0, &err);
+	device[0].class = ata_sff_dev_classify(&device[0], dev0, &err);
 	if ((slave_possible) && (err != 0x81))
-		device[1].class = ata_dev_try_classify(&device[1], dev1, &err);
+		device[1].class = ata_sff_dev_classify(&device[1], dev1, &err);
 
 	/* is double-select really necessary? */
 	if (device[1].class != ATA_DEV_NONE)
@@ -2322,7 +2322,7 @@ err_out:
 #ifdef CONFIG_PCI
 
 /**
- *	ata_pci_clear_simplex	-	attempt to kick device out of simplex
+ *	ata_pci_bmdma_clear_simplex -	attempt to kick device out of simplex
  *	@pdev: PCI device
  *
  *	Some PCI ATA devices report simplex mode but in fact can be told to
@@ -2330,7 +2330,7 @@ err_out:
  *	perform the task on such devices. Calling it on other devices will
  *	have -undefined- behaviour.
  */
-int ata_pci_clear_simplex(struct pci_dev *pdev)
+int ata_pci_bmdma_clear_simplex(struct pci_dev *pdev)
 {
 	unsigned long bmdma = pci_resource_start(pdev, 4);
 	u8 simplex;
@@ -2347,7 +2347,7 @@ int ata_pci_clear_simplex(struct pci_dev *pdev)
 }
 
 /**
- *	ata_pci_init_bmdma - acquire PCI BMDMA resources and init ATA host
+ *	ata_pci_bmdma_init - acquire PCI BMDMA resources and init ATA host
  *	@host: target ATA host
  *
  *	Acquire PCI BMDMA resources and initialize @host accordingly.
@@ -2358,7 +2358,7 @@ int ata_pci_clear_simplex(struct pci_dev *pdev)
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_pci_init_bmdma(struct ata_host *host)
+int ata_pci_bmdma_init(struct ata_host *host)
 {
 	struct device *gdev = host->dev;
 	struct pci_dev *pdev = to_pci_dev(gdev);
@@ -2418,7 +2418,7 @@ static int ata_resources_present(struct pci_dev *pdev, int port)
 }
 
 /**
- *	ata_pci_init_sff_host - acquire native PCI ATA resources and init host
+ *	ata_pci_sff_init_host - acquire native PCI ATA resources and init host
  *	@host: target ATA host
  *
  *	Acquire native PCI ATA resources for @host and initialize the
@@ -2436,7 +2436,7 @@ static int ata_resources_present(struct pci_dev *pdev, int port)
  *	0 if at least one port is initialized, -ENODEV if no port is
  *	available.
  */
-int ata_pci_init_sff_host(struct ata_host *host)
+int ata_pci_sff_init_host(struct ata_host *host)
 {
 	struct device *gdev = host->dev;
 	struct pci_dev *pdev = to_pci_dev(gdev);
@@ -2478,7 +2478,7 @@ int ata_pci_init_sff_host(struct ata_host *host)
 		ap->ioaddr.altstatus_addr =
 		ap->ioaddr.ctl_addr = (void __iomem *)
 			((unsigned long)iomap[base + 1] | ATA_PCI_CTL_OFS);
-		ata_std_ports(&ap->ioaddr);
+		ata_sff_std_ports(&ap->ioaddr);
 
 		ata_port_desc(ap, "cmd 0x%llx ctl 0x%llx",
 			(unsigned long long)pci_resource_start(pdev, base),
@@ -2496,7 +2496,7 @@ int ata_pci_init_sff_host(struct ata_host *host)
 }
 
 /**
- *	ata_pci_prepare_sff_host - helper to prepare native PCI ATA host
+ *	ata_pci_sff_prepare_host - helper to prepare native PCI ATA host
  *	@pdev: target PCI device
  *	@ppi: array of port_info, must be enough for two ports
  *	@r_host: out argument for the initialized ATA host
@@ -2510,7 +2510,7 @@ int ata_pci_init_sff_host(struct ata_host *host)
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_pci_prepare_sff_host(struct pci_dev *pdev,
+int ata_pci_sff_prepare_host(struct pci_dev *pdev,
 			     const struct ata_port_info * const * ppi,
 			     struct ata_host **r_host)
 {
@@ -2528,12 +2528,12 @@ int ata_pci_prepare_sff_host(struct pci_dev *pdev,
 		goto err_out;
 	}
 
-	rc = ata_pci_init_sff_host(host);
+	rc = ata_pci_sff_init_host(host);
 	if (rc)
 		goto err_out;
 
 	/* init DMA related stuff */
-	rc = ata_pci_init_bmdma(host);
+	rc = ata_pci_bmdma_init(host);
 	if (rc)
 		goto err_bmdma;
 
@@ -2554,7 +2554,7 @@ int ata_pci_prepare_sff_host(struct pci_dev *pdev,
 }
 
 /**
- *	ata_pci_activate_sff_host - start SFF host, request IRQ and register it
+ *	ata_pci_sff_activate_host - start SFF host, request IRQ and register it
  *	@host: target SFF ATA host
  *	@irq_handler: irq_handler used when requesting IRQ(s)
  *	@sht: scsi_host_template to use when registering the host
@@ -2569,7 +2569,7 @@ int ata_pci_prepare_sff_host(struct pci_dev *pdev,
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_pci_activate_sff_host(struct ata_host *host,
+int ata_pci_sff_activate_host(struct ata_host *host,
 			      irq_handler_t irq_handler,
 			      struct scsi_host_template *sht)
 {
@@ -2647,7 +2647,7 @@ int ata_pci_activate_sff_host(struct ata_host *host,
 }
 
 /**
- *	ata_pci_init_one - Initialize/register PCI IDE host controller
+ *	ata_pci_sff_init_one - Initialize/register PCI IDE host controller
  *	@pdev: Controller to be initialized
  *	@ppi: array of port_info, must be enough for two ports
  *	@sht: scsi_host_template to use when registering the host
@@ -2671,9 +2671,9 @@ int ata_pci_activate_sff_host(struct ata_host *host,
  *	RETURNS:
  *	Zero on success, negative on errno-based value on error.
  */
-int ata_pci_init_one(struct pci_dev *pdev,
-		     const struct ata_port_info * const * ppi,
-		     struct scsi_host_template *sht, void *host_priv)
+int ata_pci_sff_init_one(struct pci_dev *pdev,
+			 const struct ata_port_info * const * ppi,
+			 struct scsi_host_template *sht, void *host_priv)
 {
 	struct device *dev = &pdev->dev;
 	const struct ata_port_info *pi = NULL;
@@ -2704,13 +2704,13 @@ int ata_pci_init_one(struct pci_dev *pdev,
 		goto out;
 
 	/* prepare and activate SFF host */
-	rc = ata_pci_prepare_sff_host(pdev, ppi, &host);
+	rc = ata_pci_sff_prepare_host(pdev, ppi, &host);
 	if (rc)
 		goto out;
 	host->private_data = host_priv;
 
 	pci_set_master(pdev);
-	rc = ata_pci_activate_sff_host(host, ata_interrupt, sht);
+	rc = ata_pci_sff_activate_host(host, ata_sff_interrupt, sht);
  out:
 	if (rc == 0)
 		devres_remove_group(&pdev->dev, NULL);
@@ -2724,47 +2724,47 @@ int ata_pci_init_one(struct pci_dev *pdev,
 
 EXPORT_SYMBOL_GPL(ata_sff_port_ops);
 EXPORT_SYMBOL_GPL(ata_bmdma_port_ops);
-EXPORT_SYMBOL_GPL(ata_qc_prep);
-EXPORT_SYMBOL_GPL(ata_dumb_qc_prep);
-EXPORT_SYMBOL_GPL(ata_std_dev_select);
-EXPORT_SYMBOL_GPL(ata_check_status);
-EXPORT_SYMBOL_GPL(ata_altstatus);
-EXPORT_SYMBOL_GPL(ata_busy_sleep);
-EXPORT_SYMBOL_GPL(ata_wait_ready);
-EXPORT_SYMBOL_GPL(ata_tf_load);
-EXPORT_SYMBOL_GPL(ata_tf_read);
-EXPORT_SYMBOL_GPL(ata_exec_command);
-EXPORT_SYMBOL_GPL(ata_data_xfer);
-EXPORT_SYMBOL_GPL(ata_data_xfer_noirq);
-EXPORT_SYMBOL_GPL(ata_irq_on);
-EXPORT_SYMBOL_GPL(ata_bmdma_irq_clear);
-EXPORT_SYMBOL_GPL(ata_hsm_move);
-EXPORT_SYMBOL_GPL(ata_qc_issue_prot);
-EXPORT_SYMBOL_GPL(ata_host_intr);
-EXPORT_SYMBOL_GPL(ata_interrupt);
-EXPORT_SYMBOL_GPL(ata_bmdma_freeze);
-EXPORT_SYMBOL_GPL(ata_bmdma_thaw);
-EXPORT_SYMBOL_GPL(ata_std_prereset);
-EXPORT_SYMBOL_GPL(ata_dev_try_classify);
-EXPORT_SYMBOL_GPL(ata_wait_after_reset);
-EXPORT_SYMBOL_GPL(ata_std_softreset);
-EXPORT_SYMBOL_GPL(sata_std_hardreset);
-EXPORT_SYMBOL_GPL(ata_std_postreset);
-EXPORT_SYMBOL_GPL(ata_bmdma_error_handler);
-EXPORT_SYMBOL_GPL(ata_bmdma_post_internal_cmd);
+EXPORT_SYMBOL_GPL(ata_sff_qc_prep);
+EXPORT_SYMBOL_GPL(ata_sff_dumb_qc_prep);
+EXPORT_SYMBOL_GPL(ata_sff_dev_select);
+EXPORT_SYMBOL_GPL(ata_sff_check_status);
+EXPORT_SYMBOL_GPL(ata_sff_altstatus);
+EXPORT_SYMBOL_GPL(ata_sff_busy_sleep);
+EXPORT_SYMBOL_GPL(ata_sff_wait_ready);
+EXPORT_SYMBOL_GPL(ata_sff_tf_load);
+EXPORT_SYMBOL_GPL(ata_sff_tf_read);
+EXPORT_SYMBOL_GPL(ata_sff_exec_command);
+EXPORT_SYMBOL_GPL(ata_sff_data_xfer);
+EXPORT_SYMBOL_GPL(ata_sff_data_xfer_noirq);
+EXPORT_SYMBOL_GPL(ata_sff_irq_on);
+EXPORT_SYMBOL_GPL(ata_sff_irq_clear);
+EXPORT_SYMBOL_GPL(ata_sff_hsm_move);
+EXPORT_SYMBOL_GPL(ata_sff_qc_issue);
+EXPORT_SYMBOL_GPL(ata_sff_host_intr);
+EXPORT_SYMBOL_GPL(ata_sff_interrupt);
+EXPORT_SYMBOL_GPL(ata_sff_freeze);
+EXPORT_SYMBOL_GPL(ata_sff_thaw);
+EXPORT_SYMBOL_GPL(ata_sff_prereset);
+EXPORT_SYMBOL_GPL(ata_sff_dev_classify);
+EXPORT_SYMBOL_GPL(ata_sff_wait_after_reset);
+EXPORT_SYMBOL_GPL(ata_sff_softreset);
+EXPORT_SYMBOL_GPL(sata_sff_hardreset);
+EXPORT_SYMBOL_GPL(ata_sff_postreset);
+EXPORT_SYMBOL_GPL(ata_sff_error_handler);
+EXPORT_SYMBOL_GPL(ata_sff_post_internal_cmd);
 EXPORT_SYMBOL_GPL(ata_sff_port_start);
-EXPORT_SYMBOL_GPL(ata_std_ports);
-EXPORT_SYMBOL_GPL(ata_pci_default_filter);
+EXPORT_SYMBOL_GPL(ata_sff_std_ports);
+EXPORT_SYMBOL_GPL(ata_bmdma_mode_filter);
 EXPORT_SYMBOL_GPL(ata_bmdma_setup);
 EXPORT_SYMBOL_GPL(ata_bmdma_start);
 EXPORT_SYMBOL_GPL(ata_bmdma_stop);
 EXPORT_SYMBOL_GPL(ata_bmdma_status);
 EXPORT_SYMBOL_GPL(ata_bus_reset);
 #ifdef CONFIG_PCI
-EXPORT_SYMBOL_GPL(ata_pci_clear_simplex);
-EXPORT_SYMBOL_GPL(ata_pci_init_bmdma);
-EXPORT_SYMBOL_GPL(ata_pci_init_sff_host);
-EXPORT_SYMBOL_GPL(ata_pci_prepare_sff_host);
-EXPORT_SYMBOL_GPL(ata_pci_activate_sff_host);
-EXPORT_SYMBOL_GPL(ata_pci_init_one);
+EXPORT_SYMBOL_GPL(ata_pci_bmdma_clear_simplex);
+EXPORT_SYMBOL_GPL(ata_pci_bmdma_init);
+EXPORT_SYMBOL_GPL(ata_pci_sff_init_host);
+EXPORT_SYMBOL_GPL(ata_pci_sff_prepare_host);
+EXPORT_SYMBOL_GPL(ata_pci_sff_activate_host);
+EXPORT_SYMBOL_GPL(ata_pci_sff_init_one);
 #endif /* CONFIG_PCI */
diff --git a/drivers/ata/pata_acpi.c b/drivers/ata/pata_acpi.c
index d337f3209caf..c5f91e629945 100644
--- a/drivers/ata/pata_acpi.c
+++ b/drivers/ata/pata_acpi.c
@@ -47,7 +47,7 @@ static int pacpi_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (ap->acpi_handle == NULL || ata_acpi_gtm(ap, &acpi->gtm) < 0)
 		return -ENODEV;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -106,7 +106,7 @@ static unsigned long pacpi_discover_modes(struct ata_port *ap, struct ata_device
 static unsigned long pacpi_mode_filter(struct ata_device *adev, unsigned long mask)
 {
 	struct pata_acpi *acpi = adev->link->ap->private_data;
-	return ata_pci_default_filter(adev, mask & acpi->mask[adev->devno]);
+	return ata_bmdma_mode_filter(adev, mask & acpi->mask[adev->devno]);
 }
 
 /**
@@ -162,7 +162,7 @@ static void pacpi_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	pacpi_qc_issue_prot	-	command issue
+ *	pacpi_qc_issue	-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -170,14 +170,14 @@ static void pacpi_set_dmamode(struct ata_port *ap, struct ata_device *adev)
  *	neccessary.
  */
 
-static unsigned int pacpi_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int pacpi_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
 	struct pata_acpi *acpi = ap->private_data;
 
 	if (acpi->gtm.flags & 0x10)
-		return ata_qc_issue_prot(qc);
+		return ata_sff_qc_issue(qc);
 
 	if (adev != acpi->last) {
 		pacpi_set_piomode(ap, adev);
@@ -185,7 +185,7 @@ static unsigned int pacpi_qc_issue_prot(struct ata_queued_cmd *qc)
 			pacpi_set_dmamode(ap, adev);
 		acpi->last = adev;
 	}
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 /**
@@ -223,7 +223,7 @@ static struct scsi_host_template pacpi_sht = {
 
 static struct ata_port_operations pacpi_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.qc_issue		= pacpi_qc_issue_prot,
+	.qc_issue		= pacpi_qc_issue,
 	.cable_detect		= pacpi_cable_detect,
 	.mode_filter		= pacpi_mode_filter,
 	.set_piomode		= pacpi_set_piomode,
@@ -259,7 +259,7 @@ static int pacpi_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 		.port_ops	= &pacpi_ops,
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(pdev, ppi, &pacpi_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &pacpi_sht, NULL);
 }
 
 static const struct pci_device_id pacpi_pci_tbl[] = {
diff --git a/drivers/ata/pata_ali.c b/drivers/ata/pata_ali.c
index 43c558f20f92..fcabe46f262b 100644
--- a/drivers/ata/pata_ali.c
+++ b/drivers/ata/pata_ali.c
@@ -121,7 +121,7 @@ static unsigned long ali_20_filter(struct ata_device *adev, unsigned long mask)
 	ata_id_c_string(adev->id, model_num, ATA_ID_PROD, sizeof(model_num));
 	if (strstr(model_num, "WDC"))
 		return mask &= ~ATA_MASK_UDMA;
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 /**
@@ -449,7 +449,7 @@ static void ali_init_chipset(struct pci_dev *pdev)
 	}
 	pci_dev_put(isa_bridge);
 	pci_dev_put(north);
-	ata_pci_clear_simplex(pdev);
+	ata_pci_bmdma_clear_simplex(pdev);
 }
 /**
  *	ali_init_one		-	discovery callback
@@ -552,7 +552,7 @@ static int ali_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	        	ppi[0] = &info_20_udma;
 		pci_dev_put(isa_bridge);
 	}
-	return ata_pci_init_one(pdev, ppi, &ali_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &ali_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_amd.c b/drivers/ata/pata_amd.c
index 33074c34105c..26665c396485 100644
--- a/drivers/ata/pata_amd.c
+++ b/drivers/ata/pata_amd.c
@@ -143,7 +143,7 @@ static int amd_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &amd_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 static int amd_cable_detect(struct ata_port *ap)
@@ -293,7 +293,7 @@ static int nv_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &nv_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -503,7 +503,7 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	ppi[0] = &info[type];
 
 	if (type < 3)
-		ata_pci_clear_simplex(pdev);
+		ata_pci_bmdma_clear_simplex(pdev);
 
 	/* Check for AMD7411 */
 	if (type == 3)
@@ -523,7 +523,7 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	}
 
 	/* And fire it up */
-	return ata_pci_init_one(pdev, ppi, &amd_sht, hpriv);
+	return ata_pci_sff_init_one(pdev, ppi, &amd_sht, hpriv);
 }
 
 #ifdef CONFIG_PM
@@ -546,7 +546,7 @@ static int amd_reinit_one(struct pci_dev *pdev)
 			pci_write_config_byte(pdev, 0x41, fifo | 0xF0);
 		if (pdev->device == PCI_DEVICE_ID_AMD_VIPER_7409 ||
 		    pdev->device == PCI_DEVICE_ID_AMD_COBRA_7401)
-		    	ata_pci_clear_simplex(pdev);
+			ata_pci_bmdma_clear_simplex(pdev);
 	}
 
 	ata_host_resume(host);
diff --git a/drivers/ata/pata_artop.c b/drivers/ata/pata_artop.c
index b6d8c4d0e6c2..0f513bc11193 100644
--- a/drivers/ata/pata_artop.c
+++ b/drivers/ata/pata_artop.c
@@ -52,7 +52,7 @@ static int artop6210_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &artop_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -78,7 +78,7 @@ static int artop6260_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (pdev->device % 1 && !pci_test_config_bits(pdev, &artop_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -400,7 +400,7 @@ static int artop_init_one (struct pci_dev *pdev, const struct pci_device_id *id)
 
 	BUG_ON(ppi[0] == NULL);
 
-	return ata_pci_init_one(pdev, ppi, &artop_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &artop_sht, NULL);
 }
 
 static const struct pci_device_id artop_pci_tbl[] = {
diff --git a/drivers/ata/pata_at32.c b/drivers/ata/pata_at32.c
index 528315587532..3e8651d78952 100644
--- a/drivers/ata/pata_at32.c
+++ b/drivers/ata/pata_at32.c
@@ -223,7 +223,7 @@ static int __init pata_at32_init_one(struct device *dev,
 	host->private_data = info;
 
 	/* Register ATA device and return */
-	return ata_host_activate(host, info->irq, ata_interrupt,
+	return ata_host_activate(host, info->irq, ata_sff_interrupt,
 				 IRQF_SHARED | IRQF_TRIGGER_RISING,
 				 &at32_sht);
 }
diff --git a/drivers/ata/pata_atiixp.c b/drivers/ata/pata_atiixp.c
index 56a65baddd4a..78738fb4223b 100644
--- a/drivers/ata/pata_atiixp.c
+++ b/drivers/ata/pata_atiixp.c
@@ -45,7 +45,7 @@ static int atiixp_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &atiixp_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 static int atiixp_cable_detect(struct ata_port *ap)
@@ -223,7 +223,7 @@ static struct scsi_host_template atiixp_sht = {
 static struct ata_port_operations atiixp_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 
-	.qc_prep 	= ata_dumb_qc_prep,
+	.qc_prep 	= ata_sff_dumb_qc_prep,
 	.bmdma_start 	= atiixp_bmdma_start,
 	.bmdma_stop	= atiixp_bmdma_stop,
 
@@ -243,7 +243,7 @@ static int atiixp_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.port_ops = &atiixp_port_ops
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(dev, ppi, &atiixp_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &atiixp_sht, NULL);
 }
 
 static const struct pci_device_id atiixp[] = {
diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c
index 457ac800cd5e..c854e882d4a9 100644
--- a/drivers/ata/pata_bf54x.c
+++ b/drivers/ata/pata_bf54x.c
@@ -674,7 +674,7 @@ static void read_atapi_data(void __iomem *base,
  *	@ap: Port to which output is sent
  *	@tf: ATA taskfile register set
  *
- *	Note: Original code is ata_tf_load().
+ *	Note: Original code is ata_sff_tf_load().
  */
 
 static void bfin_tf_load(struct ata_port *ap, const struct ata_taskfile *tf)
@@ -745,7 +745,7 @@ static u8 bfin_check_status(struct ata_port *ap)
  *	@ap: Port from which input is read
  *	@tf: ATA taskfile register set for storing input
  *
- *	Note: Original code is ata_tf_read().
+ *	Note: Original code is ata_sff_tf_read().
  */
 
 static void bfin_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
@@ -775,7 +775,7 @@ static void bfin_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
  *	@ap: port to which command is being issued
  *	@tf: ATA taskfile register set
  *
- *	Note: Original code is ata_exec_command().
+ *	Note: Original code is ata_sff_exec_command().
  */
 
 static void bfin_exec_command(struct ata_port *ap,
@@ -785,7 +785,7 @@ static void bfin_exec_command(struct ata_port *ap,
 	dev_dbg(ap->dev, "ata%u: cmd 0x%X\n", ap->print_id, tf->command);
 
 	write_atapi_register(base, ATA_REG_CMD, tf->command);
-	ata_pause(ap);
+	ata_sff_pause(ap);
 }
 
 /**
@@ -800,14 +800,14 @@ static u8 bfin_check_altstatus(struct ata_port *ap)
 }
 
 /**
- *	bfin_std_dev_select - Select device 0/1 on ATA bus
+ *	bfin_dev_select - Select device 0/1 on ATA bus
  *	@ap: ATA channel to manipulate
  *	@device: ATA device (numbered from zero) to select
  *
- *	Note: Original code is ata_std_dev_select().
+ *	Note: Original code is ata_sff_dev_select().
  */
 
-static void bfin_std_dev_select(struct ata_port *ap, unsigned int device)
+static void bfin_dev_select(struct ata_port *ap, unsigned int device)
 {
 	void __iomem *base = (void __iomem *)ap->ioaddr.ctl_addr;
 	u8 tmp;
@@ -818,7 +818,7 @@ static void bfin_std_dev_select(struct ata_port *ap, unsigned int device)
 		tmp = ATA_DEVICE_OBS | ATA_DEV1;
 
 	write_atapi_register(base, ATA_REG_DEVICE, tmp);
-	ata_pause(ap);
+	ata_sff_pause(ap);
 }
 
 /**
@@ -977,7 +977,7 @@ static unsigned int bfin_devchk(struct ata_port *ap,
 	void __iomem *base = (void __iomem *)ap->ioaddr.ctl_addr;
 	u8 nsect, lbal;
 
-	bfin_std_dev_select(ap, device);
+	bfin_dev_select(ap, device);
 
 	write_atapi_register(base, ATA_REG_NSECT, 0x55);
 	write_atapi_register(base, ATA_REG_LBAL, 0xaa);
@@ -1014,7 +1014,7 @@ static void bfin_bus_post_reset(struct ata_port *ap, unsigned int devmask)
 	 * BSY bit to clear
 	 */
 	if (dev0)
-		ata_busy_sleep(ap, ATA_TMOUT_BOOT_QUICK, ATA_TMOUT_BOOT);
+		ata_sff_busy_sleep(ap, ATA_TMOUT_BOOT_QUICK, ATA_TMOUT_BOOT);
 
 	/* if device 1 was found in ata_devchk, wait for
 	 * register access, then wait for BSY to clear
@@ -1023,7 +1023,7 @@ static void bfin_bus_post_reset(struct ata_port *ap, unsigned int devmask)
 	while (dev1) {
 		u8 nsect, lbal;
 
-		bfin_std_dev_select(ap, 1);
+		bfin_dev_select(ap, 1);
 		nsect = read_atapi_register(base, ATA_REG_NSECT);
 		lbal = read_atapi_register(base, ATA_REG_LBAL);
 		if ((nsect == 1) && (lbal == 1))
@@ -1035,14 +1035,14 @@ static void bfin_bus_post_reset(struct ata_port *ap, unsigned int devmask)
 		msleep(50);	/* give drive a breather */
 	}
 	if (dev1)
-		ata_busy_sleep(ap, ATA_TMOUT_BOOT_QUICK, ATA_TMOUT_BOOT);
+		ata_sff_busy_sleep(ap, ATA_TMOUT_BOOT_QUICK, ATA_TMOUT_BOOT);
 
 	/* is all this really necessary? */
-	bfin_std_dev_select(ap, 0);
+	bfin_dev_select(ap, 0);
 	if (dev1)
-		bfin_std_dev_select(ap, 1);
+		bfin_dev_select(ap, 1);
 	if (dev0)
-		bfin_std_dev_select(ap, 0);
+		bfin_dev_select(ap, 0);
 }
 
 /**
@@ -1088,15 +1088,15 @@ static unsigned int bfin_bus_softreset(struct ata_port *ap,
 }
 
 /**
- *	bfin_std_softreset - reset host port via ATA SRST
+ *	bfin_softreset - reset host port via ATA SRST
  *	@ap: port to reset
  *	@classes: resulting classes of attached devices
  *
- *	Note: Original code is ata_std_softreset().
+ *	Note: Original code is ata_sff_softreset().
  */
 
-static int bfin_std_softreset(struct ata_link *link, unsigned int *classes,
-		unsigned long deadline)
+static int bfin_softreset(struct ata_link *link, unsigned int *classes,
+			  unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
 	unsigned int slave_possible = ap->flags & ATA_FLAG_SLAVE_POSS;
@@ -1115,7 +1115,7 @@ static int bfin_std_softreset(struct ata_link *link, unsigned int *classes,
 		devmask |= (1 << 1);
 
 	/* select device 0 again */
-	bfin_std_dev_select(ap, 0);
+	bfin_dev_select(ap, 0);
 
 	/* issue bus reset */
 	err_mask = bfin_bus_softreset(ap, devmask);
@@ -1126,10 +1126,10 @@ static int bfin_std_softreset(struct ata_link *link, unsigned int *classes,
 	}
 
 	/* determine by signature whether we have ATA or ATAPI devices */
-	classes[0] = ata_dev_try_classify(&ap->link.device[0],
+	classes[0] = ata_sff_dev_classify(&ap->link.device[0],
 				devmask & (1 << 0), &err);
 	if (slave_possible && err != 0x81)
-		classes[1] = ata_dev_try_classify(&ap->link.device[1],
+		classes[1] = ata_sff_dev_classify(&ap->link.device[1],
 					devmask & (1 << 1), &err);
 
  out:
@@ -1167,7 +1167,7 @@ static unsigned char bfin_bmdma_status(struct ata_port *ap)
  *	@buflen: buffer length
  *	@write_data: read/write
  *
- *	Note: Original code is ata_data_xfer().
+ *	Note: Original code is ata_sff_data_xfer().
  */
 
 static unsigned int bfin_data_xfer(struct ata_device *dev, unsigned char *buf,
@@ -1206,7 +1206,7 @@ static unsigned int bfin_data_xfer(struct ata_device *dev, unsigned char *buf,
  *	bfin_irq_clear - Clear ATAPI interrupt.
  *	@ap: Port associated with this ATA transaction.
  *
- *	Note: Original code is ata_bmdma_irq_clear().
+ *	Note: Original code is ata_sff_irq_clear().
  */
 
 static void bfin_irq_clear(struct ata_port *ap)
@@ -1223,7 +1223,7 @@ static void bfin_irq_clear(struct ata_port *ap)
  *	bfin_irq_on - Enable interrupts on a port.
  *	@ap: Port on which interrupts are enabled.
  *
- *	Note: Original code is ata_irq_on().
+ *	Note: Original code is ata_sff_irq_on().
  */
 
 static unsigned char bfin_irq_on(struct ata_port *ap)
@@ -1244,13 +1244,13 @@ static unsigned char bfin_irq_on(struct ata_port *ap)
 }
 
 /**
- *	bfin_bmdma_freeze - Freeze DMA controller port
+ *	bfin_freeze - Freeze DMA controller port
  *	@ap: port to freeze
  *
- *	Note: Original code is ata_bmdma_freeze().
+ *	Note: Original code is ata_sff_freeze().
  */
 
-static void bfin_bmdma_freeze(struct ata_port *ap)
+static void bfin_freeze(struct ata_port *ap)
 {
 	void __iomem *base = (void __iomem *)ap->ioaddr.ctl_addr;
 
@@ -1270,13 +1270,13 @@ static void bfin_bmdma_freeze(struct ata_port *ap)
 }
 
 /**
- *	bfin_bmdma_thaw - Thaw DMA controller port
+ *	bfin_thaw - Thaw DMA controller port
  *	@ap: port to thaw
  *
- *	Note: Original code is ata_bmdma_thaw().
+ *	Note: Original code is ata_sff_thaw().
  */
 
-void bfin_bmdma_thaw(struct ata_port *ap)
+void bfin_thaw(struct ata_port *ap)
 {
 	bfin_check_status(ap);
 	bfin_irq_clear(ap);
@@ -1284,14 +1284,14 @@ void bfin_bmdma_thaw(struct ata_port *ap)
 }
 
 /**
- *	bfin_std_postreset - standard postreset callback
+ *	bfin_postreset - standard postreset callback
  *	@ap: the target ata_port
  *	@classes: classes of attached devices
  *
- *	Note: Original code is ata_std_postreset().
+ *	Note: Original code is ata_sff_postreset().
  */
 
-static void bfin_std_postreset(struct ata_link *link, unsigned int *classes)
+static void bfin_postreset(struct ata_link *link, unsigned int *classes)
 {
 	struct ata_port *ap = link->ap;
 	void __iomem *base = (void __iomem *)ap->ioaddr.ctl_addr;
@@ -1301,9 +1301,9 @@ static void bfin_std_postreset(struct ata_link *link, unsigned int *classes)
 
 	/* is double-select really necessary? */
 	if (classes[0] != ATA_DEV_NONE)
-		bfin_std_dev_select(ap, 1);
+		bfin_dev_select(ap, 1);
 	if (classes[1] != ATA_DEV_NONE)
-		bfin_std_dev_select(ap, 0);
+		bfin_dev_select(ap, 0);
 
 	/* bail out if no device is present */
 	if (classes[0] == ATA_DEV_NONE && classes[1] == ATA_DEV_NONE) {
@@ -1362,7 +1362,7 @@ static const struct ata_port_operations bfin_pata_ops = {
 	.exec_command		= bfin_exec_command,
 	.check_status		= bfin_check_status,
 	.check_altstatus	= bfin_check_altstatus,
-	.dev_select		= bfin_std_dev_select,
+	.dev_select		= bfin_dev_select,
 
 	.bmdma_setup		= bfin_bmdma_setup,
 	.bmdma_start		= bfin_bmdma_start,
@@ -1372,10 +1372,10 @@ static const struct ata_port_operations bfin_pata_ops = {
 
 	.qc_prep		= ata_noop_qc_prep,
 
-	.freeze			= bfin_bmdma_freeze,
-	.thaw			= bfin_bmdma_thaw,
-	.softreset		= bfin_std_softreset,
-	.postreset		= bfin_std_postreset,
+	.freeze			= bfin_freeze,
+	.thaw			= bfin_thaw,
+	.softreset		= bfin_softreset,
+	.postreset		= bfin_postreset,
 	.post_internal_cmd	= bfin_bmdma_stop,
 
 	.irq_clear		= bfin_irq_clear,
@@ -1513,7 +1513,7 @@ static int __devinit bfin_atapi_probe(struct platform_device *pdev)
 	}
 
 	if (ata_host_activate(host, platform_get_irq(pdev, 0),
-		ata_interrupt, IRQF_SHARED, &bfin_sht) != 0) {
+		ata_sff_interrupt, IRQF_SHARED, &bfin_sht) != 0) {
 		peripheral_free_list(atapi_io_port);
 		dev_err(&pdev->dev, "Fail to attach ATAPI device\n");
 		return -ENODEV;
diff --git a/drivers/ata/pata_cmd640.c b/drivers/ata/pata_cmd640.c
index 27219b00edf4..a907cf478891 100644
--- a/drivers/ata/pata_cmd640.c
+++ b/drivers/ata/pata_cmd640.c
@@ -107,8 +107,8 @@ static void cmd640_set_piomode(struct ata_port *ap, struct ata_device *adev)
 		pci_write_config_byte(pdev, arttim + 1, (t.active << 4) | t.recover);
 	} else {
 		/* Save the shared timings for channel, they will be loaded
-		   by qc_issue_prot. Reloading the setup time is expensive
-		   so we keep a merged one loaded */
+		   by qc_issue. Reloading the setup time is expensive so we
+		   keep a merged one loaded */
 		pci_read_config_byte(pdev, ARTIM23, &reg);
 		reg &= 0x3F;
 		reg |= t.setup;
@@ -119,14 +119,14 @@ static void cmd640_set_piomode(struct ata_port *ap, struct ata_device *adev)
 
 
 /**
- *	cmd640_qc_issue_prot	-	command preparation hook
+ *	cmd640_qc_issue	-	command preparation hook
  *	@qc: Command to be issued
  *
  *	Channel 1 has shared timings. We must reprogram the
  *	clock each drive 2/3 switch we do.
  */
 
-static unsigned int cmd640_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int cmd640_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -137,7 +137,7 @@ static unsigned int cmd640_qc_issue_prot(struct ata_queued_cmd *qc)
 		pci_write_config_byte(pdev, DRWTIM23, timing->reg58[adev->devno]);
 		timing->last = adev->devno;
 	}
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 /**
@@ -172,8 +172,8 @@ static struct scsi_host_template cmd640_sht = {
 static struct ata_port_operations cmd640_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 	/* In theory xfer_noirq is not needed once we kill the prefetcher */
-	.data_xfer	= ata_data_xfer_noirq,
-	.qc_issue	= cmd640_qc_issue_prot,
+	.data_xfer	= ata_sff_data_xfer_noirq,
+	.qc_issue	= cmd640_qc_issue,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= cmd640_set_piomode,
 	.port_start	= cmd640_port_start,
@@ -224,7 +224,7 @@ static int cmd640_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	cmd640_hardware_init(pdev);
 
-	return ata_pci_init_one(pdev, ppi, &cmd640_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &cmd640_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cmd64x.c b/drivers/ata/pata_cmd64x.c
index f0e566623614..ddd09b7d98c9 100644
--- a/drivers/ata/pata_cmd64x.c
+++ b/drivers/ata/pata_cmd64x.c
@@ -349,7 +349,7 @@ static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	class_rev &= 0xFF;
 
 	if (id->driver_data == 0)	/* 643 */
-		ata_pci_clear_simplex(pdev);
+		ata_pci_bmdma_clear_simplex(pdev);
 
 	if (pdev->device == PCI_DEVICE_ID_CMD_646) {
 		/* Does UDMA work ? */
@@ -373,7 +373,7 @@ static int cmd64x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	pci_write_config_byte(pdev, UDIDETCR0, 0xF0);
 #endif
 
-	return ata_pci_init_one(pdev, ppi, &cmd64x_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &cmd64x_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cs5520.c b/drivers/ata/pata_cs5520.c
index 46d0ce32ee5a..1186bcd2781c 100644
--- a/drivers/ata/pata_cs5520.c
+++ b/drivers/ata/pata_cs5520.c
@@ -146,7 +146,7 @@ static struct scsi_host_template cs5520_sht = {
 
 static struct ata_port_operations cs5520_port_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.qc_prep		= ata_dumb_qc_prep,
+	.qc_prep		= ata_sff_dumb_qc_prep,
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= cs5520_set_piomode,
 	.set_dmamode		= cs5520_set_dmamode,
@@ -227,7 +227,7 @@ static int __devinit cs5520_init_one(struct pci_dev *pdev, const struct pci_devi
 	ioaddr->ctl_addr = iomap[1];
 	ioaddr->altstatus_addr = iomap[1];
 	ioaddr->bmdma_addr = iomap[4];
-	ata_std_ports(ioaddr);
+	ata_sff_std_ports(ioaddr);
 
 	ata_port_desc(host->ports[0],
 		      "cmd 0x%x ctl 0x%x", cmd_port[0], ctl_port[0]);
@@ -238,7 +238,7 @@ static int __devinit cs5520_init_one(struct pci_dev *pdev, const struct pci_devi
 	ioaddr->ctl_addr = iomap[3];
 	ioaddr->altstatus_addr = iomap[3];
 	ioaddr->bmdma_addr = iomap[4] + 8;
-	ata_std_ports(ioaddr);
+	ata_sff_std_ports(ioaddr);
 
 	ata_port_desc(host->ports[1],
 		      "cmd 0x%x ctl 0x%x", cmd_port[1], ctl_port[1]);
@@ -258,7 +258,7 @@ static int __devinit cs5520_init_one(struct pci_dev *pdev, const struct pci_devi
 			continue;
 
 		rc = devm_request_irq(&pdev->dev, irq[ap->port_no],
-				      ata_interrupt, 0, DRV_NAME, host);
+				      ata_sff_interrupt, 0, DRV_NAME, host);
 		if (rc)
 			return rc;
 
diff --git a/drivers/ata/pata_cs5530.c b/drivers/ata/pata_cs5530.c
index ac3ad55d7c3c..744beebaaf49 100644
--- a/drivers/ata/pata_cs5530.c
+++ b/drivers/ata/pata_cs5530.c
@@ -133,7 +133,7 @@ static void cs5530_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	cs5530_qc_issue_prot	-	command issue
+ *	cs5530_qc_issue		-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -142,7 +142,7 @@ static void cs5530_set_dmamode(struct ata_port *ap, struct ata_device *adev)
  *	one MWDMA/UDMA bit.
  */
 
-static unsigned int cs5530_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int cs5530_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -157,7 +157,7 @@ static unsigned int cs5530_qc_issue_prot(struct ata_queued_cmd *qc)
 		    	cs5530_set_dmamode(ap, adev);
 	}
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static struct scsi_host_template cs5530_sht = {
@@ -168,8 +168,8 @@ static struct scsi_host_template cs5530_sht = {
 static struct ata_port_operations cs5530_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 
-	.qc_prep 	= ata_dumb_qc_prep,
-	.qc_issue	= cs5530_qc_issue_prot,
+	.qc_prep 	= ata_sff_dumb_qc_prep,
+	.qc_issue	= cs5530_qc_issue,
 
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= cs5530_set_piomode,
@@ -325,7 +325,7 @@ static int cs5530_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 		ppi[1] = &info_palmax_secondary;
 
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(pdev, ppi, &cs5530_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &cs5530_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_cs5535.c b/drivers/ata/pata_cs5535.c
index 5c0762ebf58c..f1b6556f0483 100644
--- a/drivers/ata/pata_cs5535.c
+++ b/drivers/ata/pata_cs5535.c
@@ -199,7 +199,7 @@ static int cs5535_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	rdmsr(ATAC_CH0D1_PIO, timings, dummy);
 	if (CS5535_BAD_PIO(timings))
 		wrmsr(ATAC_CH0D1_PIO, 0xF7F4F7F4UL, 0);
-	return ata_pci_init_one(dev, ppi, &cs5535_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &cs5535_sht, NULL);
 }
 
 static const struct pci_device_id cs5535[] = {
diff --git a/drivers/ata/pata_cs5536.c b/drivers/ata/pata_cs5536.c
index 2d34b9145dcb..73f8332cb679 100644
--- a/drivers/ata/pata_cs5536.c
+++ b/drivers/ata/pata_cs5536.c
@@ -261,7 +261,7 @@ static int cs5536_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		return -ENODEV;
 	}
 
-	return ata_pci_init_one(dev, ppi, &cs5536_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &cs5536_sht, NULL);
 }
 
 static const struct pci_device_id cs5536[] = {
diff --git a/drivers/ata/pata_cypress.c b/drivers/ata/pata_cypress.c
index ae14969e1dfe..a9c3218e22fd 100644
--- a/drivers/ata/pata_cypress.c
+++ b/drivers/ata/pata_cypress.c
@@ -136,7 +136,7 @@ static int cy82c693_init_one(struct pci_dev *pdev, const struct pci_device_id *i
 	if (PCI_FUNC(pdev->devfn) != 1)
 		return -ENODEV;
 
-	return ata_pci_init_one(pdev, ppi, &cy82c693_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &cy82c693_sht, NULL);
 }
 
 static const struct pci_device_id cy82c693[] = {
diff --git a/drivers/ata/pata_efar.c b/drivers/ata/pata_efar.c
index 1d839a57068e..9fba82976ba6 100644
--- a/drivers/ata/pata_efar.c
+++ b/drivers/ata/pata_efar.c
@@ -45,7 +45,7 @@ static int efar_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &efar_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -262,7 +262,7 @@ static int efar_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &efar_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &efar_sht, NULL);
 }
 
 static const struct pci_device_id efar_pci_tbl[] = {
diff --git a/drivers/ata/pata_hpt366.c b/drivers/ata/pata_hpt366.c
index c2d4923d4db7..f2b83eabc7c7 100644
--- a/drivers/ata/pata_hpt366.c
+++ b/drivers/ata/pata_hpt366.c
@@ -184,7 +184,7 @@ static unsigned long hpt366_filter(struct ata_device *adev, unsigned long mask)
 		if (hpt_dma_blacklisted(adev, "UDMA4", bad_ata66_4))
 			mask &= ~(0xF0 << ATA_SHIFT_UDMA);
 	}
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 /**
@@ -393,7 +393,7 @@ static int hpt36x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 			break;
 	}
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(dev, ppi, &hpt36x_sht, hpriv);
+	return ata_pci_sff_init_one(dev, ppi, &hpt36x_sht, hpriv);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index c10fcd31418d..42163998de9a 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -283,7 +283,7 @@ static unsigned long hpt370_filter(struct ata_device *adev, unsigned long mask)
 		if (hpt_dma_blacklisted(adev, "UDMA100", bad_ata100_5))
 			mask &= ~(0xE0 << ATA_SHIFT_UDMA);
 	}
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 /**
@@ -299,7 +299,7 @@ static unsigned long hpt370a_filter(struct ata_device *adev, unsigned long mask)
 		if (hpt_dma_blacklisted(adev, "UDMA100", bad_ata100_5))
 			mask &= ~(0xE0 << ATA_SHIFT_UDMA);
 	}
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 /**
@@ -338,7 +338,7 @@ static int hpt37x_pre_reset(struct ata_link *link, unsigned long deadline)
 	pci_write_config_byte(pdev, 0x50 + 4 * ap->port_no, 0x37);
 	udelay(100);
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 static int hpt374_fn1_pre_reset(struct ata_link *link, unsigned long deadline)
@@ -374,7 +374,7 @@ static int hpt374_fn1_pre_reset(struct ata_link *link, unsigned long deadline)
 	pci_write_config_byte(pdev, 0x50 + 4 * ap->port_no, 0x37);
 	udelay(100);
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -1019,7 +1019,7 @@ static int hpt37x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	}
 
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(dev, ppi, &hpt37x_sht, private_data);
+	return ata_pci_sff_init_one(dev, ppi, &hpt37x_sht, private_data);
 }
 
 static const struct pci_device_id hpt37x[] = {
diff --git a/drivers/ata/pata_hpt3x2n.c b/drivers/ata/pata_hpt3x2n.c
index cd44ee3d3cc1..d5c9fd7b82bb 100644
--- a/drivers/ata/pata_hpt3x2n.c
+++ b/drivers/ata/pata_hpt3x2n.c
@@ -156,7 +156,7 @@ static int hpt3x2n_pre_reset(struct ata_link *link, unsigned long deadline)
 	pci_write_config_byte(pdev, 0x50 + 4 * ap->port_no, 0x37);
 	udelay(100);
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -308,7 +308,7 @@ static int hpt3x2n_use_dpll(struct ata_port *ap, int writing)
 	return 0;
 }
 
-static unsigned int hpt3x2n_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int hpt3x2n_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_taskfile *tf = &qc->tf;
 	struct ata_port *ap = qc->ap;
@@ -323,7 +323,7 @@ static unsigned int hpt3x2n_qc_issue_prot(struct ata_queued_cmd *qc)
 				hpt3x2n_set_clock(ap, 0x23);
 		}
 	}
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static struct scsi_host_template hpt3x2n_sht = {
@@ -338,7 +338,7 @@ static struct ata_port_operations hpt3x2n_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 
 	.bmdma_stop	= hpt3x2n_bmdma_stop,
-	.qc_issue	= hpt3x2n_qc_issue_prot,
+	.qc_issue	= hpt3x2n_qc_issue,
 
 	.cable_detect	= hpt3x2n_cable_detect,
 	.set_piomode	= hpt3x2n_set_piomode,
@@ -554,7 +554,7 @@ static int hpt3x2n_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	}
 
 	/* Now kick off ATA set up */
-	return ata_pci_init_one(dev, ppi, &hpt3x2n_sht, hpriv);
+	return ata_pci_sff_init_one(dev, ppi, &hpt3x2n_sht, hpriv);
 }
 
 static const struct pci_device_id hpt3x2n[] = {
diff --git a/drivers/ata/pata_hpt3x3.c b/drivers/ata/pata_hpt3x3.c
index 8857d029ac2e..f11a320337c0 100644
--- a/drivers/ata/pata_hpt3x3.c
+++ b/drivers/ata/pata_hpt3x3.c
@@ -202,15 +202,15 @@ static int hpt3x3_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 		ioaddr->altstatus_addr =
 		ioaddr->ctl_addr = base + offset_ctl[i];
 		ioaddr->scr_addr = NULL;
-		ata_std_ports(ioaddr);
+		ata_sff_std_ports(ioaddr);
 		ioaddr->bmdma_addr = base + 8 * i;
 
 		ata_port_pbar_desc(ap, 4, -1, "ioport");
 		ata_port_pbar_desc(ap, 4, offset_cmd[i], "cmd");
 	}
 	pci_set_master(pdev);
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &hpt3x3_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &hpt3x3_sht);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_icside.c b/drivers/ata/pata_icside.c
index 13d43e9dd279..52de9f908fb0 100644
--- a/drivers/ata/pata_icside.c
+++ b/drivers/ata/pata_icside.c
@@ -270,7 +270,7 @@ static void pata_icside_bmdma_stop(struct ata_queued_cmd *qc)
 	disable_dma(state->dma);
 
 	/* see ata_bmdma_stop */
-	ata_altstatus(ap);
+	ata_sff_altstatus(ap);
 }
 
 static u8 pata_icside_bmdma_status(struct ata_port *ap)
@@ -316,7 +316,7 @@ static void pata_icside_postreset(struct ata_link *link, unsigned int *classes)
 	struct pata_icside_state *state = ap->host->private_data;
 
 	if (classes[0] != ATA_DEV_NONE || classes[1] != ATA_DEV_NONE)
-		return ata_std_postreset(link, classes);
+		return ata_sff_postreset(link, classes);
 
 	state->port[ap->port_no].disabled = 1;
 
@@ -336,7 +336,7 @@ static struct ata_port_operations pata_icside_port_ops = {
 	.inherits		= &ata_sff_port_ops,
 	/* no need to build any PRD tables for DMA */
 	.qc_prep		= ata_noop_qc_prep,
-	.data_xfer		= ata_data_xfer_noirq,
+	.data_xfer		= ata_sff_data_xfer_noirq,
 	.bmdma_setup		= pata_icside_bmdma_setup,
 	.bmdma_start		= pata_icside_bmdma_start,
 	.bmdma_stop		= pata_icside_bmdma_stop,
@@ -481,7 +481,7 @@ static int __devinit pata_icside_add_ports(struct pata_icside_info *info)
 		pata_icside_setup_ioaddr(ap, info->base, info, info->port[i]);
 	}
 
-	return ata_host_activate(host, ec->irq, ata_interrupt, 0,
+	return ata_host_activate(host, ec->irq, ata_sff_interrupt, 0,
 				 &pata_icside_sht);
 }
 
diff --git a/drivers/ata/pata_isapnp.c b/drivers/ata/pata_isapnp.c
index 085913ec6f68..6a111baab523 100644
--- a/drivers/ata/pata_isapnp.c
+++ b/drivers/ata/pata_isapnp.c
@@ -50,7 +50,7 @@ static int isapnp_init_one(struct pnp_dev *idev, const struct pnp_device_id *dev
 
 	if (pnp_irq_valid(idev, 0)) {
 		irq = pnp_irq(idev, 0);
-		handler = ata_interrupt;
+		handler = ata_sff_interrupt;
 	}
 
 	/* allocate host */
@@ -78,7 +78,7 @@ static int isapnp_init_one(struct pnp_dev *idev, const struct pnp_device_id *dev
 		ap->ioaddr.ctl_addr = ctl_addr;
 	}
 
-	ata_std_ports(&ap->ioaddr);
+	ata_sff_std_ports(&ap->ioaddr);
 
 	ata_port_desc(ap, "cmd 0x%llx ctl 0x%llx",
 		      (unsigned long long)pnp_port_start(idev, 0),
diff --git a/drivers/ata/pata_it8213.c b/drivers/ata/pata_it8213.c
index 84ab89e8a247..c113d7c079c8 100644
--- a/drivers/ata/pata_it8213.c
+++ b/drivers/ata/pata_it8213.c
@@ -40,7 +40,7 @@ static int it8213_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &it8213_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -274,7 +274,7 @@ static int it8213_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &it8213_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &it8213_sht, NULL);
 }
 
 static const struct pci_device_id it8213_pci_tbl[] = {
diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c
index 6a8a4ddf5bfe..88e37cfcfc4f 100644
--- a/drivers/ata/pata_it821x.c
+++ b/drivers/ata/pata_it821x.c
@@ -395,11 +395,11 @@ static void it821x_passthru_dev_select(struct ata_port *ap,
 		it821x_program(ap, adev, itdev->pio[adev->devno]);
 		itdev->last_device = device;
 	}
-	ata_std_dev_select(ap, device);
+	ata_sff_dev_select(ap, device);
 }
 
 /**
- *	it821x_smart_qc_issue_prot	-	wrap qc issue prot
+ *	it821x_smart_qc_issue		-	wrap qc issue prot
  *	@qc: command
  *
  *	Wrap the command issue sequence for the IT821x. We need to
@@ -407,7 +407,7 @@ static void it821x_passthru_dev_select(struct ata_port *ap,
  *	usual happenings kick off
  */
 
-static unsigned int it821x_smart_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int it821x_smart_qc_issue(struct ata_queued_cmd *qc)
 {
 	switch(qc->tf.command)
 	{
@@ -427,14 +427,14 @@ static unsigned int it821x_smart_qc_issue_prot(struct ata_queued_cmd *qc)
 		case ATA_CMD_ID_ATA:
 		/* Arguably should just no-op this one */
 		case ATA_CMD_SET_FEATURES:
-			return ata_qc_issue_prot(qc);
+			return ata_sff_qc_issue(qc);
 	}
 	printk(KERN_DEBUG "it821x: can't process command 0x%02X\n", qc->tf.command);
 	return AC_ERR_DEV;
 }
 
 /**
- *	it821x_passthru_qc_issue_prot	-	wrap qc issue prot
+ *	it821x_passthru_qc_issue	-	wrap qc issue prot
  *	@qc: command
  *
  *	Wrap the command issue sequence for the IT821x. We need to
@@ -442,10 +442,10 @@ static unsigned int it821x_smart_qc_issue_prot(struct ata_queued_cmd *qc)
  *	usual happenings kick off
  */
 
-static unsigned int it821x_passthru_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int it821x_passthru_qc_issue(struct ata_queued_cmd *qc)
 {
 	it821x_passthru_dev_select(qc->ap, qc->dev->devno);
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 /**
@@ -639,7 +639,7 @@ static struct ata_port_operations it821x_smart_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 
 	.check_atapi_dma= it821x_check_atapi_dma,
-	.qc_issue	= it821x_smart_qc_issue_prot,
+	.qc_issue	= it821x_smart_qc_issue,
 
 	.cable_detect	= it821x_ident_hack,
 	.set_mode	= it821x_smart_set_mode,
@@ -655,7 +655,7 @@ static struct ata_port_operations it821x_passthru_port_ops = {
 	.dev_select 	= it821x_passthru_dev_select,
 	.bmdma_start 	= it821x_passthru_bmdma_start,
 	.bmdma_stop	= it821x_passthru_bmdma_stop,
-	.qc_issue	= it821x_passthru_qc_issue_prot,
+	.qc_issue	= it821x_passthru_qc_issue,
 
 	.cable_detect	= ata_cable_unknown,
 	.set_piomode	= it821x_passthru_set_piomode,
@@ -722,7 +722,7 @@ static int it821x_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	else
 		ppi[0] = &info_smart;
 
-	return ata_pci_init_one(pdev, ppi, &it821x_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &it821x_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_ixp4xx_cf.c b/drivers/ata/pata_ixp4xx_cf.c
index d02629aa20da..283a8fba79b3 100644
--- a/drivers/ata/pata_ixp4xx_cf.c
+++ b/drivers/ata/pata_ixp4xx_cf.c
@@ -110,7 +110,7 @@ static void ixp4xx_setup_port(struct ata_port *ap,
 	ioaddr->altstatus_addr	= data->cs1 + 0x06;
 	ioaddr->ctl_addr	= data->cs1 + 0x06;
 
-	ata_std_ports(ioaddr);
+	ata_sff_std_ports(ioaddr);
 
 #ifndef __ARMEB__
 
@@ -186,7 +186,7 @@ static __devinit int ixp4xx_pata_probe(struct platform_device *pdev)
 	dev_printk(KERN_INFO, &pdev->dev, "version " DRV_VERSION "\n");
 
 	/* activate host */
-	return ata_host_activate(host, irq, ata_interrupt, 0, &ixp4xx_sht);
+	return ata_host_activate(host, irq, ata_sff_interrupt, 0, &ixp4xx_sht);
 }
 
 static __devexit int ixp4xx_pata_remove(struct platform_device *dev)
diff --git a/drivers/ata/pata_jmicron.c b/drivers/ata/pata_jmicron.c
index 317f3474e0ba..73b7596816b4 100644
--- a/drivers/ata/pata_jmicron.c
+++ b/drivers/ata/pata_jmicron.c
@@ -102,7 +102,7 @@ static int jmicron_pre_reset(struct ata_link *link, unsigned long deadline)
 		ap->cbl = ATA_CBL_SATA;
 		break;
 	}
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /* No PIO or DMA methods needed for this device */
@@ -144,7 +144,7 @@ static int jmicron_init_one (struct pci_dev *pdev, const struct pci_device_id *i
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
 
-	return ata_pci_init_one(pdev, ppi, &jmicron_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &jmicron_sht, NULL);
 }
 
 static const struct pci_device_id jmicron_pci_tbl[] = {
diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c
index 2474068596f4..f13f10a55ef3 100644
--- a/drivers/ata/pata_legacy.c
+++ b/drivers/ata/pata_legacy.c
@@ -226,12 +226,12 @@ static const struct ata_port_operations legacy_base_port_ops = {
 
 static struct ata_port_operations simple_port_ops = {
 	.inherits	= &legacy_base_port_ops,
-	.data_xfer	= ata_data_xfer_noirq,
+	.data_xfer	= ata_sff_data_xfer_noirq,
 };
 
 static struct ata_port_operations legacy_port_ops = {
 	.inherits	= &legacy_base_port_ops,
-	.data_xfer	= ata_data_xfer_noirq,
+	.data_xfer	= ata_sff_data_xfer_noirq,
 	.set_mode	= legacy_set_mode,
 };
 
@@ -317,7 +317,7 @@ static unsigned int pdc_data_xfer_vlb(struct ata_device *dev,
 		}
 		local_irq_restore(flags);
 	} else
-		buflen = ata_data_xfer_noirq(dev, buf, buflen, rw);
+		buflen = ata_sff_data_xfer_noirq(dev, buf, buflen, rw);
 
 	return buflen;
 }
@@ -579,7 +579,7 @@ static void opti82c46x_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	opt82c465mv_qc_issue_prot	-	command issue
+ *	opt82c465mv_qc_issue		-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -593,7 +593,7 @@ static void opti82c46x_set_piomode(struct ata_port *ap, struct ata_device *adev)
  *	FIXME: dual channel needs ->serialize support
  */
 
-static unsigned int opti82c46x_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int opti82c46x_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -604,13 +604,13 @@ static unsigned int opti82c46x_qc_issue_prot(struct ata_queued_cmd *qc)
 	    && ap->host->private_data != NULL)
 		opti82c46x_set_piomode(ap, adev);
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static struct ata_port_operations opti82c46x_port_ops = {
 	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= opti82c46x_set_piomode,
-	.qc_issue	= opti82c46x_qc_issue_prot,
+	.qc_issue	= opti82c46x_qc_issue,
 };
 
 static void qdi6500_set_piomode(struct ata_port *ap, struct ata_device *adev)
@@ -644,7 +644,7 @@ static void qdi6500_set_piomode(struct ata_port *ap, struct ata_device *adev)
  *	@irq: interrupt line
  *
  *	In dual channel mode the 6580 has one clock per channel and we have
- *	to software clockswitch in qc_issue_prot.
+ *	to software clockswitch in qc_issue.
  */
 
 static void qdi6580dp_set_piomode(struct ata_port *ap, struct ata_device *adev)
@@ -710,14 +710,14 @@ static void qdi6580_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	qdi_qc_issue_prot	-	command issue
+ *	qdi_qc_issue		-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
  *	this interface so that we can load the correct ATA timings.
  */
 
-static unsigned int qdi_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int qdi_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -730,7 +730,7 @@ static unsigned int qdi_qc_issue_prot(struct ata_queued_cmd *qc)
 							2 * ap->port_no);
 		}
 	}
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static unsigned int vlb32_data_xfer(struct ata_device *adev, unsigned char *buf,
@@ -759,7 +759,7 @@ static unsigned int vlb32_data_xfer(struct ata_device *adev, unsigned char *buf,
 		}
 		return (buflen + 3) & ~3;
 	} else
-		return ata_data_xfer(adev, buf, buflen, rw);
+		return ata_sff_data_xfer(adev, buf, buflen, rw);
 }
 
 static int qdi_port(struct platform_device *dev,
@@ -774,7 +774,7 @@ static int qdi_port(struct platform_device *dev,
 static struct ata_port_operations qdi6500_port_ops = {
 	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= qdi6500_set_piomode,
-	.qc_issue	= qdi_qc_issue_prot,
+	.qc_issue	= qdi_qc_issue,
 	.data_xfer	= vlb32_data_xfer,
 };
 
@@ -1016,13 +1016,13 @@ static __init int legacy_init_one(struct legacy_probe *probe)
 	ap->ioaddr.cmd_addr = io_addr;
 	ap->ioaddr.altstatus_addr = ctrl_addr;
 	ap->ioaddr.ctl_addr = ctrl_addr;
-	ata_std_ports(&ap->ioaddr);
+	ata_sff_std_ports(&ap->ioaddr);
 	ap->host->private_data = ld;
 
 	ata_port_desc(ap, "cmd 0x%lx ctl 0x%lx", io, io + 0x0206);
 
-	ret = ata_host_activate(host, probe->irq, ata_interrupt, 0,
-								&legacy_sht);
+	ret = ata_host_activate(host, probe->irq, ata_sff_interrupt, 0,
+				&legacy_sht);
 	if (ret)
 		goto fail;
 	ld->platform_dev = pdev;
diff --git a/drivers/ata/pata_marvell.c b/drivers/ata/pata_marvell.c
index d38e64cd6097..24a011b25024 100644
--- a/drivers/ata/pata_marvell.c
+++ b/drivers/ata/pata_marvell.c
@@ -55,7 +55,7 @@ static int marvell_pre_reset(struct ata_link *link, unsigned long deadline)
 	    (!(devices & 0x10)))	/* PATA enable ? */
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 static int marvell_cable_detect(struct ata_port *ap)
@@ -128,7 +128,7 @@ static int marvell_init_one (struct pci_dev *pdev, const struct pci_device_id *i
 	if (pdev->device == 0x6101)
 		ppi[1] = &ata_dummy_port_info;
 
-	return ata_pci_init_one(pdev, ppi, &marvell_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &marvell_sht, NULL);
 }
 
 static const struct pci_device_id marvell_pci_tbl[] = {
diff --git a/drivers/ata/pata_mpc52xx.c b/drivers/ata/pata_mpc52xx.c
index fec93196710e..5d1d32a39c4d 100644
--- a/drivers/ata/pata_mpc52xx.c
+++ b/drivers/ata/pata_mpc52xx.c
@@ -252,7 +252,7 @@ mpc52xx_ata_dev_select(struct ata_port *ap, unsigned int device)
 	if (device != priv->csel)
 		mpc52xx_ata_apply_timings(priv, device);
 
-	ata_std_dev_select(ap,device);
+	ata_sff_dev_select(ap,device);
 }
 
 static struct scsi_host_template mpc52xx_ata_sht = {
@@ -305,7 +305,7 @@ mpc52xx_ata_init_one(struct device *dev, struct mpc52xx_ata_priv *priv,
 	ata_port_desc(ap, "ata_regs 0x%lx", raw_ata_regs);
 
 	/* activate host */
-	return ata_host_activate(host, priv->ata_irq, ata_interrupt, 0,
+	return ata_host_activate(host, priv->ata_irq, ata_sff_interrupt, 0,
 				 &mpc52xx_ata_sht);
 }
 
diff --git a/drivers/ata/pata_mpiix.c b/drivers/ata/pata_mpiix.c
index 1b9d0d412ebf..7d7e3fdab71f 100644
--- a/drivers/ata/pata_mpiix.c
+++ b/drivers/ata/pata_mpiix.c
@@ -55,7 +55,7 @@ static int mpiix_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &mpiix_enable_bits))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -69,8 +69,8 @@ static int mpiix_pre_reset(struct ata_link *link, unsigned long deadline)
  *
  *	This would get very ugly because we can only program timing for one
  *	device at a time, the other gets PIO0. Fortunately libata calls
- *	our qc_issue_prot command before a command is issued so we can
- *	flip the timings back and forth to reduce the pain.
+ *	our qc_issue command before a command is issued so we can flip the
+ *	timings back and forth to reduce the pain.
  */
 
 static void mpiix_set_piomode(struct ata_port *ap, struct ata_device *adev)
@@ -110,7 +110,7 @@ static void mpiix_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	mpiix_qc_issue_prot	-	command issue
+ *	mpiix_qc_issue		-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -120,7 +120,7 @@ static void mpiix_set_piomode(struct ata_port *ap, struct ata_device *adev)
  *	be made PIO0.
  */
 
-static unsigned int mpiix_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int mpiix_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -133,7 +133,7 @@ static unsigned int mpiix_qc_issue_prot(struct ata_queued_cmd *qc)
 	if (adev->pio_mode && adev != ap->private_data)
 		mpiix_set_piomode(ap, adev);
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static struct scsi_host_template mpiix_sht = {
@@ -142,7 +142,7 @@ static struct scsi_host_template mpiix_sht = {
 
 static struct ata_port_operations mpiix_port_ops = {
 	.inherits	= &ata_sff_port_ops,
-	.qc_issue	= mpiix_qc_issue_prot,
+	.qc_issue	= mpiix_qc_issue,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= mpiix_set_piomode,
 	.prereset	= mpiix_pre_reset,
@@ -207,10 +207,10 @@ static int mpiix_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	ap->ioaddr.altstatus_addr = ctl_addr;
 
 	/* Let libata fill in the port details */
-	ata_std_ports(&ap->ioaddr);
+	ata_sff_std_ports(&ap->ioaddr);
 
 	/* activate host */
-	return ata_host_activate(host, irq, ata_interrupt, IRQF_SHARED,
+	return ata_host_activate(host, irq, ata_sff_interrupt, IRQF_SHARED,
 				 &mpiix_sht);
 }
 
diff --git a/drivers/ata/pata_netcell.c b/drivers/ata/pata_netcell.c
index 349182840d24..d9719c8b9dbe 100644
--- a/drivers/ata/pata_netcell.c
+++ b/drivers/ata/pata_netcell.c
@@ -68,10 +68,10 @@ static int netcell_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 		return rc;
 
 	/* Any chip specific setup/optimisation/messages here */
-	ata_pci_clear_simplex(pdev);
+	ata_pci_bmdma_clear_simplex(pdev);
 
 	/* And let the library code do the work */
-	return ata_pci_init_one(pdev, port_info, &netcell_sht, NULL);
+	return ata_pci_sff_init_one(pdev, port_info, &netcell_sht, NULL);
 }
 
 static const struct pci_device_id netcell_pci_tbl[] = {
diff --git a/drivers/ata/pata_ninja32.c b/drivers/ata/pata_ninja32.c
index 8213d081f313..c9c0ea67c71a 100644
--- a/drivers/ata/pata_ninja32.c
+++ b/drivers/ata/pata_ninja32.c
@@ -73,7 +73,7 @@ static void ninja32_dev_select(struct ata_port *ap, unsigned int device)
 	struct ata_device *adev = &ap->link.device[device];
 	if (ap->private_data != adev) {
 		iowrite8(0xd6, ap->ioaddr.bmdma_addr + 0x1f);
-		ata_std_dev_select(ap, device);
+		ata_sff_dev_select(ap, device);
 		ninja32_set_piomode(ap, adev);
 	}
 }
@@ -132,7 +132,7 @@ static int ninja32_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	ap->ioaddr.ctl_addr = base + 0x1E;
 	ap->ioaddr.altstatus_addr = base + 0x1E;
 	ap->ioaddr.bmdma_addr = base;
-	ata_std_ports(&ap->ioaddr);
+	ata_sff_std_ports(&ap->ioaddr);
 
 	iowrite8(0x05, base + 0x01);	/* Enable interrupt lines */
 	iowrite8(0xBE, base + 0x02);	/* Burst, ?? setup */
@@ -142,7 +142,7 @@ static int ninja32_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	iowrite8(0xa4, base + 0x1c);	/* Unknown */
 	iowrite8(0x83, base + 0x1d);	/* BMDMA control: WAIT0 */
 	/* FIXME: Should we disable them at remove ? */
-	return ata_host_activate(host, dev->irq, ata_interrupt,
+	return ata_host_activate(host, dev->irq, ata_sff_interrupt,
 				 IRQF_SHARED, &ninja32_sht);
 }
 
diff --git a/drivers/ata/pata_ns87410.c b/drivers/ata/pata_ns87410.c
index 4d2eefee7387..76d2455bc453 100644
--- a/drivers/ata/pata_ns87410.c
+++ b/drivers/ata/pata_ns87410.c
@@ -50,7 +50,7 @@ static int ns87410_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &ns87410_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -105,7 +105,7 @@ static void ns87410_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	ns87410_qc_issue_prot	-	command issue
+ *	ns87410_qc_issue	-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -113,7 +113,7 @@ static void ns87410_set_piomode(struct ata_port *ap, struct ata_device *adev)
  *	necessary.
  */
 
-static unsigned int ns87410_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int ns87410_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -126,7 +126,7 @@ static unsigned int ns87410_qc_issue_prot(struct ata_queued_cmd *qc)
 	if (adev->pio_mode && adev != ap->private_data)
 		ns87410_set_piomode(ap, adev);
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static struct scsi_host_template ns87410_sht = {
@@ -135,7 +135,7 @@ static struct scsi_host_template ns87410_sht = {
 
 static struct ata_port_operations ns87410_port_ops = {
 	.inherits	= &ata_sff_port_ops,
-	.qc_issue	= ns87410_qc_issue_prot,
+	.qc_issue	= ns87410_qc_issue,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= ns87410_set_piomode,
 	.prereset	= ns87410_pre_reset,
@@ -149,7 +149,7 @@ static int ns87410_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 		.port_ops = &ns87410_port_ops
 	};
 	const struct ata_port_info *ppi[] = { &info, NULL };
-	return ata_pci_init_one(dev, ppi, &ns87410_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &ns87410_sht, NULL);
 }
 
 static const struct pci_device_id ns87410[] = {
diff --git a/drivers/ata/pata_ns87415.c b/drivers/ata/pata_ns87415.c
index cdd79d6fc0ee..03a52cefc010 100644
--- a/drivers/ata/pata_ns87415.c
+++ b/drivers/ata/pata_ns87415.c
@@ -172,14 +172,14 @@ static void ns87415_bmdma_stop(struct ata_queued_cmd *qc)
 }
 
 /**
- *	ns87415_bmdma_irq_clear		-	Clear interrupt
+ *	ns87415_irq_clear		-	Clear interrupt
  *	@ap: Channel to clear
  *
  *	Erratum: Due to a chip bug regisers 02 and 0A bit 1 and 2 (the
  *	error bits) are reset by writing to register 00 or 08.
  */
 
-static void ns87415_bmdma_irq_clear(struct ata_port *ap)
+static void ns87415_irq_clear(struct ata_port *ap)
 {
 	void __iomem *mmio = ap->ioaddr.bmdma_addr;
 
@@ -306,7 +306,7 @@ static struct ata_port_operations ns87415_pata_ops = {
 	.bmdma_setup		= ns87415_bmdma_setup,
 	.bmdma_start		= ns87415_bmdma_start,
 	.bmdma_stop		= ns87415_bmdma_stop,
-	.irq_clear		= ns87415_bmdma_irq_clear,
+	.irq_clear		= ns87415_irq_clear,
 
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= ns87415_set_piomode,
@@ -375,7 +375,7 @@ static int ns87415_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 	pci_write_config_byte(pdev, 0x55, 0xEE);
 	/* Select PIO0 8bit clocking */
 	pci_write_config_byte(pdev, 0x54, 0xB7);
-	return ata_pci_init_one(pdev, ppi, &ns87415_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &ns87415_sht, NULL);
 }
 
 static const struct pci_device_id ns87415_pci_tbl[] = {
diff --git a/drivers/ata/pata_oldpiix.c b/drivers/ata/pata_oldpiix.c
index c1da79a76439..e678af383d13 100644
--- a/drivers/ata/pata_oldpiix.c
+++ b/drivers/ata/pata_oldpiix.c
@@ -47,7 +47,7 @@ static int oldpiix_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &oldpiix_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -181,7 +181,7 @@ static void oldpiix_set_dmamode (struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	oldpiix_qc_issue_prot	-	command issue
+ *	oldpiix_qc_issue	-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -191,7 +191,7 @@ static void oldpiix_set_dmamode (struct ata_port *ap, struct ata_device *adev)
  *	be made PIO0.
  */
 
-static unsigned int oldpiix_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int oldpiix_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -201,7 +201,7 @@ static unsigned int oldpiix_qc_issue_prot(struct ata_queued_cmd *qc)
 		if (adev->dma_mode)
 			oldpiix_set_dmamode(ap, adev);
 	}
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 
@@ -211,7 +211,7 @@ static struct scsi_host_template oldpiix_sht = {
 
 static struct ata_port_operations oldpiix_pata_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.qc_issue		= oldpiix_qc_issue_prot,
+	.qc_issue		= oldpiix_qc_issue,
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= oldpiix_set_piomode,
 	.set_dmamode		= oldpiix_set_dmamode,
@@ -249,7 +249,7 @@ static int oldpiix_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &oldpiix_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &oldpiix_sht, NULL);
 }
 
 static const struct pci_device_id oldpiix_pci_tbl[] = {
diff --git a/drivers/ata/pata_opti.c b/drivers/ata/pata_opti.c
index 4ddd03a67775..fb2cf661b0e8 100644
--- a/drivers/ata/pata_opti.c
+++ b/drivers/ata/pata_opti.c
@@ -64,7 +64,7 @@ static int opti_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &opti_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -173,7 +173,7 @@ static int opti_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (!printed_version++)
 		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(dev, ppi, &opti_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &opti_sht, NULL);
 }
 
 static const struct pci_device_id opti[] = {
diff --git a/drivers/ata/pata_optidma.c b/drivers/ata/pata_optidma.c
index 36ac147de178..4cd744456313 100644
--- a/drivers/ata/pata_optidma.c
+++ b/drivers/ata/pata_optidma.c
@@ -64,7 +64,7 @@ static int optidma_pre_reset(struct ata_link *link, unsigned long deadline)
 	if (ap->port_no && !pci_test_config_bits(pdev, &optidma_enable_bits))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -430,7 +430,7 @@ static int optidma_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (optiplus_with_udma(dev))
 		ppi[0] = &info_82c700_udma;
 
-	return ata_pci_init_one(dev, ppi, &optidma_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &optidma_sht, NULL);
 }
 
 static const struct pci_device_id optidma[] = {
diff --git a/drivers/ata/pata_pcmcia.c b/drivers/ata/pata_pcmcia.c
index 57efbf05c95f..2e206c5f869a 100644
--- a/drivers/ata/pata_pcmcia.c
+++ b/drivers/ata/pata_pcmcia.c
@@ -133,7 +133,7 @@ static struct scsi_host_template pcmcia_sht = {
 
 static struct ata_port_operations pcmcia_port_ops = {
 	.inherits	= &ata_sff_port_ops,
-	.data_xfer	= ata_data_xfer_noirq,
+	.data_xfer	= ata_sff_data_xfer_noirq,
 	.cable_detect	= ata_cable_40wire,
 	.set_mode	= pcmcia_set_mode,
 };
@@ -323,13 +323,13 @@ next_entry:
 		ap->ioaddr.cmd_addr = io_addr + 0x10 * p;
 		ap->ioaddr.altstatus_addr = ctl_addr + 0x10 * p;
 		ap->ioaddr.ctl_addr = ctl_addr + 0x10 * p;
-		ata_std_ports(&ap->ioaddr);
+		ata_sff_std_ports(&ap->ioaddr);
 
 		ata_port_desc(ap, "cmd 0x%lx ctl 0x%lx", io_base, ctl_base);
 	}
 
 	/* activate */
-	ret = ata_host_activate(host, pdev->irq.AssignedIRQ, ata_interrupt,
+	ret = ata_host_activate(host, pdev->irq.AssignedIRQ, ata_sff_interrupt,
 				IRQF_SHARED, &pcmcia_sht);
 	if (ret)
 		goto failed;
diff --git a/drivers/ata/pata_pdc2027x.c b/drivers/ata/pata_pdc2027x.c
index d235c9f92d09..0e1c2c1134d3 100644
--- a/drivers/ata/pata_pdc2027x.c
+++ b/drivers/ata/pata_pdc2027x.c
@@ -248,7 +248,7 @@ static int pdc2027x_prereset(struct ata_link *link, unsigned long deadline)
 	/* Check whether port enabled */
 	if (!pdc2027x_port_enabled(link->ap))
 		return -ENOENT;
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
@@ -265,7 +265,7 @@ static unsigned long pdc2027x_mode_filter(struct ata_device *adev, unsigned long
 	struct ata_device *pair = ata_dev_pair(adev);
 
 	if (adev->class != ATA_DEV_ATA || adev->devno == 0 || pair == NULL)
-		return ata_pci_default_filter(adev, mask);
+		return ata_bmdma_mode_filter(adev, mask);
 
 	/* Check for slave of a Maxtor at UDMA6 */
 	ata_id_c_string(pair->id, model_num, ATA_ID_PROD,
@@ -274,7 +274,7 @@ static unsigned long pdc2027x_mode_filter(struct ata_device *adev, unsigned long
 	if (strstr(model_num, "Maxtor") == NULL && pair->dma_mode == XFER_UDMA_6)
 		mask &= ~ (1 << (6 + ATA_SHIFT_UDMA));
 
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 /**
@@ -759,8 +759,8 @@ static int __devinit pdc2027x_init_one(struct pci_dev *pdev, const struct pci_de
 		return -EIO;
 
 	pci_set_master(pdev);
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &pdc2027x_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &pdc2027x_sht);
 }
 
 /**
diff --git a/drivers/ata/pata_pdc202xx_old.c b/drivers/ata/pata_pdc202xx_old.c
index 8214100e3ac1..d2673060bc8d 100644
--- a/drivers/ata/pata_pdc202xx_old.c
+++ b/drivers/ata/pata_pdc202xx_old.c
@@ -324,7 +324,7 @@ static int pdc202xx_init_one(struct pci_dev *dev, const struct pci_device_id *id
 				return -ENODEV;
 		}
 	}
-	return ata_pci_init_one(dev, ppi, &pdc202xx_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &pdc202xx_sht, NULL);
 }
 
 static const struct pci_device_id pdc202xx[] = {
diff --git a/drivers/ata/pata_platform.c b/drivers/ata/pata_platform.c
index 0588c9b7e73e..1edfc13d05d2 100644
--- a/drivers/ata/pata_platform.c
+++ b/drivers/ata/pata_platform.c
@@ -52,7 +52,7 @@ static struct scsi_host_template pata_platform_sht = {
 
 static struct ata_port_operations pata_platform_port_ops = {
 	.inherits		= &ata_sff_port_ops,
-	.data_xfer		= ata_data_xfer_noirq,
+	.data_xfer		= ata_sff_data_xfer_noirq,
 	.cable_detect		= ata_cable_unknown,
 	.set_mode		= pata_platform_set_mode,
 	.port_start		= ATA_OP_NULL,
@@ -176,7 +176,7 @@ int __devinit __pata_platform_probe(struct device *dev,
 		      (unsigned long long)ctl_res->start);
 
 	/* activate */
-	return ata_host_activate(host, irq, irq ? ata_interrupt : NULL,
+	return ata_host_activate(host, irq, irq ? ata_sff_interrupt : NULL,
 				 irq_flags, &pata_platform_sht);
 }
 EXPORT_SYMBOL_GPL(__pata_platform_probe);
diff --git a/drivers/ata/pata_qdi.c b/drivers/ata/pata_qdi.c
index d16b343d2a62..0d81dd5fdc38 100644
--- a/drivers/ata/pata_qdi.c
+++ b/drivers/ata/pata_qdi.c
@@ -102,14 +102,14 @@ static void qdi6580_set_piomode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	qdi_qc_issue_prot	-	command issue
+ *	qdi_qc_issue		-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
  *	this interface so that we can load the correct ATA timings.
  */
 
-static unsigned int qdi_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int qdi_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -121,7 +121,7 @@ static unsigned int qdi_qc_issue_prot(struct ata_queued_cmd *qc)
 			outb(qdi->clock[adev->devno], qdi->timing);
 		}
 	}
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static unsigned int qdi_data_xfer(struct ata_device *dev, unsigned char *buf,
@@ -148,7 +148,7 @@ static unsigned int qdi_data_xfer(struct ata_device *dev, unsigned char *buf,
 			buflen += 4 - slop;
 		}
 	} else
-		buflen = ata_data_xfer(dev, buf, buflen, rw);
+		buflen = ata_sff_data_xfer(dev, buf, buflen, rw);
 
 	return buflen;
 }
@@ -159,7 +159,7 @@ static struct scsi_host_template qdi_sht = {
 
 static struct ata_port_operations qdi6500_port_ops = {
 	.inherits	= &ata_sff_port_ops,
-	.qc_issue	= qdi_qc_issue_prot,
+	.qc_issue	= qdi_qc_issue,
 	.data_xfer	= qdi_data_xfer,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= qdi6500_set_piomode,
@@ -223,7 +223,7 @@ static __init int qdi_init_one(unsigned long port, int type, unsigned long io, i
 	ap->ioaddr.cmd_addr = io_addr;
 	ap->ioaddr.altstatus_addr = ctl_addr;
 	ap->ioaddr.ctl_addr = ctl_addr;
-	ata_std_ports(&ap->ioaddr);
+	ata_sff_std_ports(&ap->ioaddr);
 
 	ata_port_desc(ap, "cmd %lx ctl %lx", io, ctl);
 
@@ -239,7 +239,7 @@ static __init int qdi_init_one(unsigned long port, int type, unsigned long io, i
 	printk(KERN_INFO DRV_NAME": qd%d at 0x%lx.\n", type, io);
 
 	/* activate */
-	ret = ata_host_activate(host, irq, ata_interrupt, 0, &qdi_sht);
+	ret = ata_host_activate(host, irq, ata_sff_interrupt, 0, &qdi_sht);
 	if (ret)
 		goto fail;
 
diff --git a/drivers/ata/pata_radisys.c b/drivers/ata/pata_radisys.c
index 9ab84fc3798d..1c0d9fa7ee54 100644
--- a/drivers/ata/pata_radisys.c
+++ b/drivers/ata/pata_radisys.c
@@ -156,7 +156,7 @@ static void radisys_set_dmamode (struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	radisys_qc_issue_prot	-	command issue
+ *	radisys_qc_issue	-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -166,7 +166,7 @@ static void radisys_set_dmamode (struct ata_port *ap, struct ata_device *adev)
  *	be made PIO0.
  */
 
-static unsigned int radisys_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int radisys_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -180,7 +180,7 @@ static unsigned int radisys_qc_issue_prot(struct ata_queued_cmd *qc)
 				radisys_set_piomode(ap, adev);
 		}
 	}
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 
@@ -190,7 +190,7 @@ static struct scsi_host_template radisys_sht = {
 
 static struct ata_port_operations radisys_pata_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.qc_issue		= radisys_qc_issue_prot,
+	.qc_issue		= radisys_qc_issue,
 	.cable_detect		= ata_cable_unknown,
 	.set_piomode		= radisys_set_piomode,
 	.set_dmamode		= radisys_set_dmamode,
@@ -228,7 +228,7 @@ static int radisys_init_one (struct pci_dev *pdev, const struct pci_device_id *e
 		dev_printk(KERN_DEBUG, &pdev->dev,
 			   "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(pdev, ppi, &radisys_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &radisys_sht, NULL);
 }
 
 static const struct pci_device_id radisys_pci_tbl[] = {
diff --git a/drivers/ata/pata_rb500_cf.c b/drivers/ata/pata_rb500_cf.c
index 7affceec1c29..5b23d79bc92b 100644
--- a/drivers/ata/pata_rb500_cf.c
+++ b/drivers/ata/pata_rb500_cf.c
@@ -57,7 +57,7 @@ static inline void rb500_pata_finish_io(struct ata_port *ap)
 	struct ata_host *ah = ap->host;
 	struct rb500_cf_info *info = ah->private_data;
 
-	ata_altstatus(ap);
+	ata_sff_altstatus(ap);
 	ndelay(RB500_CF_IO_DELAY);
 
 	set_irq_type(info->irq, IRQ_TYPE_LEVEL_HIGH);
@@ -109,7 +109,7 @@ static irqreturn_t rb500_pata_irq_handler(int irq, void *dev_instance)
 	if (gpio_get_value(info->gpio_line)) {
 		set_irq_type(info->irq, IRQ_TYPE_LEVEL_LOW);
 		if (!info->frozen)
-			ata_interrupt(info->irq, dev_instance);
+			ata_sff_interrupt(info->irq, dev_instance);
 	} else {
 		set_irq_type(info->irq, IRQ_TYPE_LEVEL_HIGH);
 	}
@@ -148,7 +148,7 @@ static void rb500_pata_setup_ports(struct ata_host *ah)
 	ap->ioaddr.ctl_addr	= info->iobase + RB500_CF_REG_CTRL;
 	ap->ioaddr.altstatus_addr = info->iobase + RB500_CF_REG_CTRL;
 
-	ata_std_ports(&ap->ioaddr);
+	ata_sff_std_ports(&ap->ioaddr);
 
 	ap->ioaddr.data_addr	= info->iobase + RB500_CF_REG_DATA;
 }
diff --git a/drivers/ata/pata_rz1000.c b/drivers/ata/pata_rz1000.c
index 462b72a31280..7dfd1f3f6f3a 100644
--- a/drivers/ata/pata_rz1000.c
+++ b/drivers/ata/pata_rz1000.c
@@ -99,7 +99,7 @@ static int rz1000_init_one (struct pci_dev *pdev, const struct pci_device_id *en
 		printk(KERN_DEBUG DRV_NAME " version " DRV_VERSION "\n");
 
 	if (rz1000_fifo_disable(pdev) == 0)
-		return ata_pci_init_one(pdev, ppi, &rz1000_sht, NULL);
+		return ata_pci_sff_init_one(pdev, ppi, &rz1000_sht, NULL);
 
 	printk(KERN_ERR DRV_NAME ": failed to disable read-ahead on chipset..\n");
 	/* Not safe to use so skip */
diff --git a/drivers/ata/pata_sc1200.c b/drivers/ata/pata_sc1200.c
index 42efacf73c79..cbab397e3db7 100644
--- a/drivers/ata/pata_sc1200.c
+++ b/drivers/ata/pata_sc1200.c
@@ -151,7 +151,7 @@ static void sc1200_set_dmamode(struct ata_port *ap, struct ata_device *adev)
 }
 
 /**
- *	sc1200_qc_issue_prot	-	command issue
+ *	sc1200_qc_issue		-	command issue
  *	@qc: command pending
  *
  *	Called when the libata layer is about to issue a command. We wrap
@@ -160,7 +160,7 @@ static void sc1200_set_dmamode(struct ata_port *ap, struct ata_device *adev)
  *	one MWDMA/UDMA bit.
  */
 
-static unsigned int sc1200_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int sc1200_qc_issue(struct ata_queued_cmd *qc)
 {
 	struct ata_port *ap = qc->ap;
 	struct ata_device *adev = qc->dev;
@@ -175,7 +175,7 @@ static unsigned int sc1200_qc_issue_prot(struct ata_queued_cmd *qc)
 		    	sc1200_set_dmamode(ap, adev);
 	}
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static struct scsi_host_template sc1200_sht = {
@@ -185,8 +185,8 @@ static struct scsi_host_template sc1200_sht = {
 
 static struct ata_port_operations sc1200_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
-	.qc_prep 	= ata_dumb_qc_prep,
-	.qc_issue	= sc1200_qc_issue_prot,
+	.qc_prep 	= ata_sff_dumb_qc_prep,
+	.qc_issue	= sc1200_qc_issue,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= sc1200_set_piomode,
 	.set_dmamode	= sc1200_set_dmamode,
@@ -213,7 +213,7 @@ static int sc1200_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	/* Can't enable port 2 yet, see top comments */
 	const struct ata_port_info *ppi[] = { &info, &ata_dummy_port_info };
 
-	return ata_pci_init_one(dev, ppi, &sc1200_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &sc1200_sht, NULL);
 }
 
 static const struct pci_device_id sc1200[] = {
diff --git a/drivers/ata/pata_scc.c b/drivers/ata/pata_scc.c
index fba5bed0a641..701d0addae73 100644
--- a/drivers/ata/pata_scc.c
+++ b/drivers/ata/pata_scc.c
@@ -266,7 +266,7 @@ unsigned long scc_mode_filter(struct ata_device *adev, unsigned long mask)
 		printk(KERN_INFO "%s: limit ATAPI UDMA to UDMA4\n", DRV_NAME);
 		mask &= ~(0xE0 << ATA_SHIFT_UDMA);
 	}
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 /**
@@ -274,7 +274,7 @@ unsigned long scc_mode_filter(struct ata_device *adev, unsigned long mask)
  *	@ap: Port to which output is sent
  *	@tf: ATA taskfile register set
  *
- *	Note: Original code is ata_tf_load().
+ *	Note: Original code is ata_sff_tf_load().
  */
 
 static void scc_tf_load (struct ata_port *ap, const struct ata_taskfile *tf)
@@ -341,7 +341,7 @@ static u8 scc_check_status (struct ata_port *ap)
  *	@ap: Port from which input is read
  *	@tf: ATA taskfile register set for storing input
  *
- *	Note: Original code is ata_tf_read().
+ *	Note: Original code is ata_sff_tf_read().
  */
 
 static void scc_tf_read (struct ata_port *ap, struct ata_taskfile *tf)
@@ -373,7 +373,7 @@ static void scc_tf_read (struct ata_port *ap, struct ata_taskfile *tf)
  *	@ap: port to which command is being issued
  *	@tf: ATA taskfile register set
  *
- *	Note: Original code is ata_exec_command().
+ *	Note: Original code is ata_sff_exec_command().
  */
 
 static void scc_exec_command (struct ata_port *ap,
@@ -382,7 +382,7 @@ static void scc_exec_command (struct ata_port *ap,
 	DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command);
 
 	out_be32(ap->ioaddr.command_addr, tf->command);
-	ata_pause(ap);
+	ata_sff_pause(ap);
 }
 
 /**
@@ -396,14 +396,14 @@ static u8 scc_check_altstatus (struct ata_port *ap)
 }
 
 /**
- *	scc_std_dev_select - Select device 0/1 on ATA bus
+ *	scc_dev_select - Select device 0/1 on ATA bus
  *	@ap: ATA channel to manipulate
  *	@device: ATA device (numbered from zero) to select
  *
- *	Note: Original code is ata_std_dev_select().
+ *	Note: Original code is ata_sff_dev_select().
  */
 
-static void scc_std_dev_select (struct ata_port *ap, unsigned int device)
+static void scc_dev_select (struct ata_port *ap, unsigned int device)
 {
 	u8 tmp;
 
@@ -413,7 +413,7 @@ static void scc_std_dev_select (struct ata_port *ap, unsigned int device)
 		tmp = ATA_DEVICE_OBS | ATA_DEV1;
 
 	out_be32(ap->ioaddr.device_addr, tmp);
-	ata_pause(ap);
+	ata_sff_pause(ap);
 }
 
 /**
@@ -514,7 +514,7 @@ static int scc_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	 * BSY bit to clear
 	 */
 	if (dev0) {
-		rc = ata_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(ap, deadline);
 		if (rc && rc != -ENODEV)
 			return rc;
 	}
@@ -535,7 +535,7 @@ static int scc_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 		msleep(50);	/* give drive a breather */
 	}
 	if (dev1) {
-		rc = ata_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(ap, deadline);
 		if (rc && rc != -ENODEV)
 			return rc;
 	}
@@ -571,7 +571,7 @@ static unsigned int scc_bus_softreset(struct ata_port *ap, unsigned int devmask,
 	out_be32(ioaddr->ctl_addr, ap->ctl);
 
 	/* wait a while before checking status */
-	ata_wait_after_reset(ap, deadline);
+	ata_sff_wait_after_reset(ap, deadline);
 
 	/* Before we perform post reset processing we want to see if
 	 * the bus shows 0xFF because the odd clown forgets the D7
@@ -586,16 +586,16 @@ static unsigned int scc_bus_softreset(struct ata_port *ap, unsigned int devmask,
 }
 
 /**
- *	scc_std_softreset - reset host port via ATA SRST
+ *	scc_softreset - reset host port via ATA SRST
  *	@ap: port to reset
  *	@classes: resulting classes of attached devices
  *	@deadline: deadline jiffies for the operation
  *
- *	Note: Original code is ata_std_softreset().
+ *	Note: Original code is ata_sff_softreset().
  */
 
-static int scc_std_softreset(struct ata_link *link, unsigned int *classes,
-                             unsigned long deadline)
+static int scc_softreset(struct ata_link *link, unsigned int *classes,
+			 unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
 	unsigned int slave_possible = ap->flags & ATA_FLAG_SLAVE_POSS;
@@ -628,10 +628,10 @@ static int scc_std_softreset(struct ata_link *link, unsigned int *classes,
 	}
 
 	/* determine by signature whether we have ATA or ATAPI devices */
-	classes[0] = ata_dev_try_classify(&ap->link.device[0],
+	classes[0] = ata_sff_dev_classify(&ap->link.device[0],
 					  devmask & (1 << 0), &err);
 	if (slave_possible && err != 0x81)
-		classes[1] = ata_dev_try_classify(&ap->link.device[1],
+		classes[1] = ata_sff_dev_classify(&ap->link.device[1],
 						  devmask & (1 << 1), &err);
 
  out:
@@ -695,7 +695,7 @@ static void scc_bmdma_stop (struct ata_queued_cmd *qc)
 			printk(KERN_WARNING "%s: Internal Bus Error\n", DRV_NAME);
 			out_be32(bmid_base + SCC_DMA_INTST, INTSTS_BMSINT);
 			/* TBD: SW reset */
-			scc_std_softreset(&ap->link, &classes, deadline);
+			scc_softreset(&ap->link, &classes, deadline);
 			continue;
 		}
 
@@ -721,7 +721,7 @@ static void scc_bmdma_stop (struct ata_queued_cmd *qc)
 		 in_be32(bmid_base + SCC_DMA_CMD) & ~ATA_DMA_START);
 
 	/* one-PIO-cycle guaranteed wait, per spec, for HDMA1:0 transition */
-	ata_altstatus(ap);	/* dummy read */
+	ata_sff_altstatus(ap);	/* dummy read */
 }
 
 /**
@@ -742,7 +742,7 @@ static u8 scc_bmdma_status (struct ata_port *ap)
 		return host_stat;
 
 	/* errata A252,A308 workaround: Step4 */
-	if ((ata_altstatus(ap) & ATA_ERR) && (int_status & INTSTS_INTRQ))
+	if ((ata_sff_altstatus(ap) & ATA_ERR) && (int_status & INTSTS_INTRQ))
 		return (host_stat | ATA_DMA_INTR);
 
 	/* errata A308 workaround Step5 */
@@ -773,7 +773,7 @@ static u8 scc_bmdma_status (struct ata_port *ap)
  *	@buflen: buffer length
  *	@rw: read/write
  *
- *	Note: Original code is ata_data_xfer().
+ *	Note: Original code is ata_sff_data_xfer().
  */
 
 static unsigned int scc_data_xfer (struct ata_device *dev, unsigned char *buf,
@@ -815,7 +815,7 @@ static unsigned int scc_data_xfer (struct ata_device *dev, unsigned char *buf,
  *	scc_irq_on - Enable interrupts on a port.
  *	@ap: Port on which interrupts are enabled.
  *
- *	Note: Original code is ata_irq_on().
+ *	Note: Original code is ata_sff_irq_on().
  */
 
 static u8 scc_irq_on (struct ata_port *ap)
@@ -835,13 +835,13 @@ static u8 scc_irq_on (struct ata_port *ap)
 }
 
 /**
- *	scc_bmdma_freeze - Freeze BMDMA controller port
+ *	scc_freeze - Freeze BMDMA controller port
  *	@ap: port to freeze
  *
- *	Note: Original code is ata_bmdma_freeze().
+ *	Note: Original code is ata_sff_freeze().
  */
 
-static void scc_bmdma_freeze (struct ata_port *ap)
+static void scc_freeze (struct ata_port *ap)
 {
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 
@@ -868,18 +868,18 @@ static void scc_bmdma_freeze (struct ata_port *ap)
 static int scc_pata_prereset(struct ata_link *link, unsigned long deadline)
 {
 	link->ap->cbl = ATA_CBL_PATA80;
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 /**
- *	scc_std_postreset - standard postreset callback
+ *	scc_postreset - standard postreset callback
  *	@ap: the target ata_port
  *	@classes: classes of attached devices
  *
- *	Note: Original code is ata_std_postreset().
+ *	Note: Original code is ata_sff_postreset().
  */
 
-static void scc_std_postreset(struct ata_link *link, unsigned int *classes)
+static void scc_postreset(struct ata_link *link, unsigned int *classes)
 {
 	struct ata_port *ap = link->ap;
 
@@ -905,13 +905,13 @@ static void scc_std_postreset(struct ata_link *link, unsigned int *classes)
 }
 
 /**
- *	scc_bmdma_irq_clear - Clear PCI IDE BMDMA interrupt.
+ *	scc_irq_clear - Clear PCI IDE BMDMA interrupt.
  *	@ap: Port associated with this ATA transaction.
  *
- *	Note: Original code is ata_bmdma_irq_clear().
+ *	Note: Original code is ata_sff_irq_clear().
  */
 
-static void scc_bmdma_irq_clear (struct ata_port *ap)
+static void scc_irq_clear (struct ata_port *ap)
 {
 	void __iomem *mmio = ap->ioaddr.bmdma_addr;
 
@@ -972,7 +972,7 @@ static struct ata_port_operations scc_pata_ops = {
 	.exec_command		= scc_exec_command,
 	.check_status		= scc_check_status,
 	.check_altstatus	= scc_check_altstatus,
-	.dev_select		= scc_std_dev_select,
+	.dev_select		= scc_dev_select,
 
 	.bmdma_setup		= scc_bmdma_setup,
 	.bmdma_start		= scc_bmdma_start,
@@ -980,13 +980,13 @@ static struct ata_port_operations scc_pata_ops = {
 	.bmdma_status		= scc_bmdma_status,
 	.data_xfer		= scc_data_xfer,
 
-	.freeze			= scc_bmdma_freeze,
+	.freeze			= scc_freeze,
 	.prereset		= scc_pata_prereset,
-	.softreset		= scc_std_softreset,
-	.postreset		= scc_std_postreset,
+	.softreset		= scc_softreset,
+	.postreset		= scc_postreset,
 	.post_internal_cmd	= scc_bmdma_stop,
 
-	.irq_clear		= scc_bmdma_irq_clear,
+	.irq_clear		= scc_irq_clear,
 	.irq_on			= scc_irq_on,
 
 	.port_start		= scc_port_start,
@@ -1140,8 +1140,8 @@ static int scc_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		return rc;
 
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &scc_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &scc_sht);
 }
 
 static struct pci_driver scc_pci_driver = {
diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
index 2f4f9b0f89de..ffd26d0dc50d 100644
--- a/drivers/ata/pata_serverworks.c
+++ b/drivers/ata/pata_serverworks.c
@@ -199,7 +199,7 @@ static unsigned long serverworks_osb4_filter(struct ata_device *adev, unsigned l
 {
 	if (adev->class == ATA_DEV_ATA)
 		mask &= ~ATA_MASK_UDMA;
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 
@@ -219,7 +219,7 @@ static unsigned long serverworks_csb_filter(struct ata_device *adev, unsigned lo
 
 	/* Disk, UDMA */
 	if (adev->class != ATA_DEV_ATA)
-		return ata_pci_default_filter(adev, mask);
+		return ata_bmdma_mode_filter(adev, mask);
 
 	/* Actually do need to check */
 	ata_id_c_string(adev->id, model_num, ATA_ID_PROD, sizeof(model_num));
@@ -228,7 +228,7 @@ static unsigned long serverworks_csb_filter(struct ata_device *adev, unsigned lo
 		if (!strcmp(p, model_num))
 			mask &= ~(0xE0 << ATA_SHIFT_UDMA);
 	}
-	return ata_pci_default_filter(adev, mask);
+	return ata_bmdma_mode_filter(adev, mask);
 }
 
 /**
@@ -459,9 +459,9 @@ static int serverworks_init_one(struct pci_dev *pdev, const struct pci_device_id
 		serverworks_fixup_ht1000(pdev);
 
 	if (pdev->device == PCI_DEVICE_ID_SERVERWORKS_CSB5IDE)
-		ata_pci_clear_simplex(pdev);
+		ata_pci_bmdma_clear_simplex(pdev);
 
-	return ata_pci_init_one(pdev, ppi, &serverworks_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &serverworks_sht, NULL);
 }
 
 #ifdef CONFIG_PM
@@ -482,7 +482,7 @@ static int serverworks_reinit_one(struct pci_dev *pdev)
 			serverworks_fixup_osb4(pdev);
 			break;
 		case PCI_DEVICE_ID_SERVERWORKS_CSB5IDE:
-			ata_pci_clear_simplex(pdev);
+			ata_pci_bmdma_clear_simplex(pdev);
 			/* fall through */
 		case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE:
 		case PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2:
diff --git a/drivers/ata/pata_sil680.c b/drivers/ata/pata_sil680.c
index 63fafc6d6da3..720b8645f58a 100644
--- a/drivers/ata/pata_sil680.c
+++ b/drivers/ata/pata_sil680.c
@@ -350,19 +350,19 @@ static int __devinit sil680_init_one(struct pci_dev *pdev,
 	host->ports[0]->ioaddr.cmd_addr = mmio_base + 0x80;
 	host->ports[0]->ioaddr.ctl_addr = mmio_base + 0x8a;
 	host->ports[0]->ioaddr.altstatus_addr = mmio_base + 0x8a;
-	ata_std_ports(&host->ports[0]->ioaddr);
+	ata_sff_std_ports(&host->ports[0]->ioaddr);
 	host->ports[1]->ioaddr.bmdma_addr = mmio_base + 0x08;
 	host->ports[1]->ioaddr.cmd_addr = mmio_base + 0xc0;
 	host->ports[1]->ioaddr.ctl_addr = mmio_base + 0xca;
 	host->ports[1]->ioaddr.altstatus_addr = mmio_base + 0xca;
-	ata_std_ports(&host->ports[1]->ioaddr);
+	ata_sff_std_ports(&host->ports[1]->ioaddr);
 
 	/* Register & activate */
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &sil680_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &sil680_sht);
 
 use_ioports:
-	return ata_pci_init_one(pdev, ppi, &sil680_sht, NULL);
+	return ata_pci_sff_init_one(pdev, ppi, &sil680_sht, NULL);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_sis.c b/drivers/ata/pata_sis.c
index 793e6714df8c..e82c66e8d31b 100644
--- a/drivers/ata/pata_sis.c
+++ b/drivers/ata/pata_sis.c
@@ -156,7 +156,7 @@ static int sis_pre_reset(struct ata_link *link, unsigned long deadline)
 	/* Clear the FIFO settings. We can't enable the FIFO until
 	   we know we are poking at a disk */
 	pci_write_config_byte(pdev, 0x4B, 0);
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 
@@ -821,7 +821,7 @@ static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	sis_fixup(pdev, chipset);
 
-	return ata_pci_init_one(pdev, ppi, &sis_sht, chipset);
+	return ata_pci_sff_init_one(pdev, ppi, &sis_sht, chipset);
 }
 
 static const struct pci_device_id sis_pci_tbl[] = {
diff --git a/drivers/ata/pata_sl82c105.c b/drivers/ata/pata_sl82c105.c
index bee11ca8f55a..70d94fb28a5f 100644
--- a/drivers/ata/pata_sl82c105.c
+++ b/drivers/ata/pata_sl82c105.c
@@ -60,7 +60,7 @@ static int sl82c105_pre_reset(struct ata_link *link, unsigned long deadline)
 
 	if (ap->port_no && !pci_test_config_bits(pdev, &sl82c105_enable_bits[ap->port_no]))
 		return -ENOENT;
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 
@@ -317,7 +317,7 @@ static int sl82c105_init_one(struct pci_dev *dev, const struct pci_device_id *id
 	val |= CTRL_P0EN | CTRL_P0F16 | CTRL_P1F16;
 	pci_write_config_dword(dev, 0x40, val);
 
-	return ata_pci_init_one(dev, ppi, &sl82c105_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &sl82c105_sht, NULL);
 }
 
 static const struct pci_device_id sl82c105[] = {
diff --git a/drivers/ata/pata_triflex.c b/drivers/ata/pata_triflex.c
index bd546a389ce1..b181261f2743 100644
--- a/drivers/ata/pata_triflex.c
+++ b/drivers/ata/pata_triflex.c
@@ -66,7 +66,7 @@ static int triflex_prereset(struct ata_link *link, unsigned long deadline)
 	if (!pci_test_config_bits(pdev, &triflex_enable_bits[ap->port_no]))
 		return -ENOENT;
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 
@@ -201,7 +201,7 @@ static int triflex_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 	if (!printed_version++)
 		dev_printk(KERN_DEBUG, &dev->dev, "version " DRV_VERSION "\n");
 
-	return ata_pci_init_one(dev, ppi, &triflex_sht, NULL);
+	return ata_pci_sff_init_one(dev, ppi, &triflex_sht, NULL);
 }
 
 static const struct pci_device_id triflex[] = {
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
index 2928fa173132..4b85f84fbe76 100644
--- a/drivers/ata/pata_via.c
+++ b/drivers/ata/pata_via.c
@@ -210,7 +210,7 @@ static int via_pre_reset(struct ata_link *link, unsigned long deadline)
 			return -ENOENT;
 	}
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 
@@ -336,7 +336,7 @@ static struct ata_port_operations via_port_ops = {
 
 static struct ata_port_operations via_port_ops_noirq = {
 	.inherits	= &via_port_ops,
-	.data_xfer	= ata_data_xfer_noirq,
+	.data_xfer	= ata_sff_data_xfer_noirq,
 };
 
 /**
@@ -511,7 +511,7 @@ static int via_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	}
 
 	/* We have established the device type, now fire it up */
-	return ata_pci_init_one(pdev, ppi, &via_sht, (void *)config);
+	return ata_pci_sff_init_one(pdev, ppi, &via_sht, (void *)config);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/ata/pata_winbond.c b/drivers/ata/pata_winbond.c
index f235bb0d6139..cc18231e9334 100644
--- a/drivers/ata/pata_winbond.c
+++ b/drivers/ata/pata_winbond.c
@@ -116,7 +116,7 @@ static unsigned int winbond_data_xfer(struct ata_device *dev,
 			buflen += 4 - slop;
 		}
 	} else
-		buflen = ata_data_xfer(dev, buf, buflen, rw);
+		buflen = ata_sff_data_xfer(dev, buf, buflen, rw);
 
 	return buflen;
 }
@@ -198,7 +198,7 @@ static __init int winbond_init_one(unsigned long port)
 		ap->ioaddr.cmd_addr = cmd_addr;
 		ap->ioaddr.altstatus_addr = ctl_addr;
 		ap->ioaddr.ctl_addr = ctl_addr;
-		ata_std_ports(&ap->ioaddr);
+		ata_sff_std_ports(&ap->ioaddr);
 
 		/* hook in a private data structure per channel */
 		host->private_data = &winbond_data[nr_winbond_host];
@@ -206,7 +206,7 @@ static __init int winbond_init_one(unsigned long port)
 		winbond_data[nr_winbond_host].platform_dev = pdev;
 
 		/* activate */
-		rc = ata_host_activate(host, 14 + i, ata_interrupt, 0,
+		rc = ata_host_activate(host, 14 + i, ata_sff_interrupt, 0,
 				       &winbond_sht);
 		if (rc)
 			goto err_unregister;
diff --git a/drivers/ata/pdc_adma.c b/drivers/ata/pdc_adma.c
index c431bf36f9ba..be53545c9f64 100644
--- a/drivers/ata/pdc_adma.c
+++ b/drivers/ata/pdc_adma.c
@@ -208,7 +208,7 @@ static void adma_reinit_engine(struct ata_port *ap)
 
 	/* mask/clear ATA interrupts */
 	writeb(ATA_NIEN, ap->ioaddr.ctl_addr);
-	ata_check_status(ap);
+	ata_sff_check_status(ap);
 
 	/* reset the ADMA engine */
 	adma_reset_engine(ap);
@@ -243,7 +243,7 @@ static void adma_freeze(struct ata_port *ap)
 
 	/* mask/clear ATA interrupts */
 	writeb(ATA_NIEN, ap->ioaddr.ctl_addr);
-	ata_check_status(ap);
+	ata_sff_check_status(ap);
 
 	/* reset ADMA to idle state */
 	writew(aPIOMD4 | aNIEN | aRSTADM, chan + ADMA_CONTROL);
@@ -266,7 +266,7 @@ static int adma_prereset(struct ata_link *link, unsigned long deadline)
 		pp->state = adma_state_mmio;
 	adma_reinit_engine(ap);
 
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 static int adma_fill_sg(struct ata_queued_cmd *qc)
@@ -322,7 +322,7 @@ static void adma_qc_prep(struct ata_queued_cmd *qc)
 
 	adma_enter_reg_mode(qc->ap);
 	if (qc->tf.protocol != ATA_PROT_DMA) {
-		ata_qc_prep(qc);
+		ata_sff_qc_prep(qc);
 		return;
 	}
 
@@ -421,7 +421,7 @@ static unsigned int adma_qc_issue(struct ata_queued_cmd *qc)
 	}
 
 	pp->state = adma_state_mmio;
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static inline unsigned int adma_intr_pkt(struct ata_host *host)
@@ -492,7 +492,7 @@ static inline unsigned int adma_intr_mmio(struct ata_host *host)
 			if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING))) {
 
 				/* check main status, clearing INTRQ */
-				u8 status = ata_check_status(ap);
+				u8 status = ata_sff_check_status(ap);
 				if ((status & ATA_BUSY))
 					continue;
 				DPRINTK("ata%u: protocol %d (dev_stat 0x%X)\n",
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index 12fbf3868fe2..6e6fca4c20b6 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -271,7 +271,7 @@ static void inic_host_intr(struct ata_port *ap)
 			return;
 		}
 
-		if (likely(ata_host_intr(ap, qc)))
+		if (likely(ata_sff_host_intr(ap, qc)))
 			return;
 
 		ap->ops->check_status(ap); /* clear ATA interrupt */
@@ -356,7 +356,7 @@ static unsigned int inic_qc_issue(struct ata_queued_cmd *qc)
 			return AC_ERR_HSM;
 	}
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static void inic_freeze(struct ata_port *ap)
@@ -418,9 +418,9 @@ static int inic_hardreset(struct ata_link *link, unsigned int *class,
 		struct ata_taskfile tf;
 
 		/* wait a while before checking status */
-		ata_wait_after_reset(ap, deadline);
+		ata_sff_wait_after_reset(ap, deadline);
 
-		rc = ata_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(ap, deadline);
 		/* link occupied, -ENODEV too is an error */
 		if (rc) {
 			ata_link_printk(link, KERN_WARNING, "device not ready "
@@ -428,7 +428,7 @@ static int inic_hardreset(struct ata_link *link, unsigned int *class,
 			return rc;
 		}
 
-		ata_tf_read(ap, &tf);
+		ata_sff_tf_read(ap, &tf);
 		*class = ata_dev_classify(&tf);
 		if (*class == ATA_DEV_UNKNOWN)
 			*class = ATA_DEV_NONE;
@@ -663,7 +663,7 @@ static int inic_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 			((unsigned long)iomap[2 * i + 1] | ATA_PCI_CTL_OFS);
 		port->scr_addr = iomap[MMIO_BAR] + offset + PORT_SCR;
 
-		ata_std_ports(port);
+		ata_sff_std_ports(port);
 
 		ata_port_pbar_desc(ap, MMIO_BAR, -1, "mmio");
 		ata_port_pbar_desc(ap, MMIO_BAR, offset, "port");
diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c
index 16c15ed3536e..fa75df634c75 100644
--- a/drivers/ata/sata_mv.c
+++ b/drivers/ata/sata_mv.c
@@ -1386,7 +1386,7 @@ static unsigned int mv_qc_issue(struct ata_queued_cmd *qc)
 		 * shadow block, etc registers.
 		 */
 		mv_stop_edma(ap);
-		return ata_qc_issue_prot(qc);
+		return ata_sff_qc_issue(qc);
 	}
 
 	mv_start_dma(ap, port_mmio, pp, qc->tf.protocol);
@@ -2362,7 +2362,7 @@ comreset_retry:
 	 */
 	retry = 20;
 	while (1) {
-		u8 drv_stat = ata_check_status(ap);
+		u8 drv_stat = ata_sff_check_status(ap);
 		if ((drv_stat != 0x80) && (drv_stat != 0x7f))
 			break;
 		msleep(500);
@@ -2377,7 +2377,7 @@ comreset_retry:
 	 */
 
 	/* finally, read device signature from TF registers */
-	*class = ata_dev_try_classify(ap->link.device, 1, NULL);
+	*class = ata_sff_dev_classify(ap->link.device, 1, NULL);
 
 	writelfl(0, port_mmio + EDMA_ERR_IRQ_CAUSE_OFS);
 
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index 95fd0cd28b49..63cc43765f04 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -730,7 +730,7 @@ static void nv_adma_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
 	   ADMA mode could abort outstanding commands. */
 	nv_adma_register_mode(ap);
 
-	ata_tf_read(ap, tf);
+	ata_sff_tf_read(ap, tf);
 }
 
 static unsigned int nv_adma_tf_to_cpb(struct ata_taskfile *tf, __le16 *cpb)
@@ -844,12 +844,12 @@ static int nv_host_intr(struct ata_port *ap, u8 irq_stat)
 
 	/* DEV interrupt w/ no active qc? */
 	if (unlikely(!qc || (qc->tf.flags & ATA_TFLAG_POLLING))) {
-		ata_check_status(ap);
+		ata_sff_check_status(ap);
 		return 1;
 	}
 
 	/* handle interrupt */
-	return ata_host_intr(ap, qc);
+	return ata_sff_host_intr(ap, qc);
 }
 
 static irqreturn_t nv_adma_interrupt(int irq, void *dev_instance)
@@ -1028,7 +1028,7 @@ static void nv_adma_irq_clear(struct ata_port *ap)
 	u32 notifier_clears[2];
 
 	if (pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) {
-		ata_bmdma_irq_clear(ap);
+		ata_sff_irq_clear(ap);
 		return;
 	}
 
@@ -1059,7 +1059,7 @@ static void nv_adma_post_internal_cmd(struct ata_queued_cmd *qc)
 	struct nv_adma_port_priv *pp = qc->ap->private_data;
 
 	if (pp->flags & NV_ADMA_PORT_REGISTER_MODE)
-		ata_bmdma_post_internal_cmd(qc);
+		ata_sff_post_internal_cmd(qc);
 }
 
 static int nv_adma_port_start(struct ata_port *ap)
@@ -1336,7 +1336,7 @@ static void nv_adma_qc_prep(struct ata_queued_cmd *qc)
 		BUG_ON(!(pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) &&
 			(qc->flags & ATA_QCFLAG_DMAMAP));
 		nv_adma_register_mode(qc->ap);
-		ata_qc_prep(qc);
+		ata_sff_qc_prep(qc);
 		return;
 	}
 
@@ -1395,7 +1395,7 @@ static unsigned int nv_adma_qc_issue(struct ata_queued_cmd *qc)
 		BUG_ON(!(pp->flags & NV_ADMA_ATAPI_SETUP_COMPLETE) &&
 			(qc->flags & ATA_QCFLAG_DMAMAP));
 		nv_adma_register_mode(qc->ap);
-		return ata_qc_issue_prot(qc);
+		return ata_sff_qc_issue(qc);
 	} else
 		nv_adma_mode(qc->ap);
 
@@ -1436,7 +1436,7 @@ static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance)
 
 			qc = ata_qc_from_tag(ap, ap->link.active_tag);
 			if (qc && (!(qc->tf.flags & ATA_TFLAG_POLLING)))
-				handled += ata_host_intr(ap, qc);
+				handled += ata_sff_host_intr(ap, qc);
 			else
 				// No request pending?  Clear interrupt status
 				// anyway, in case there's one pending.
@@ -1571,7 +1571,7 @@ static void nv_mcp55_freeze(struct ata_port *ap)
 	mask = readl(mmio_base + NV_INT_ENABLE_MCP55);
 	mask &= ~(NV_INT_ALL_MCP55 << shift);
 	writel(mask, mmio_base + NV_INT_ENABLE_MCP55);
-	ata_bmdma_freeze(ap);
+	ata_sff_freeze(ap);
 }
 
 static void nv_mcp55_thaw(struct ata_port *ap)
@@ -1585,7 +1585,7 @@ static void nv_mcp55_thaw(struct ata_port *ap)
 	mask = readl(mmio_base + NV_INT_ENABLE_MCP55);
 	mask |= (NV_INT_MASK_MCP55 << shift);
 	writel(mask, mmio_base + NV_INT_ENABLE_MCP55);
-	ata_bmdma_thaw(ap);
+	ata_sff_thaw(ap);
 }
 
 static int nv_hardreset(struct ata_link *link, unsigned int *class,
@@ -1597,7 +1597,7 @@ static int nv_hardreset(struct ata_link *link, unsigned int *class,
 	 * some controllers.  Don't classify on hardreset.  For more
 	 * info, see http://bugzilla.kernel.org/show_bug.cgi?id=3352
 	 */
-	return sata_std_hardreset(link, &dummy, deadline);
+	return sata_sff_hardreset(link, &dummy, deadline);
 }
 
 static void nv_adma_error_handler(struct ata_port *ap)
@@ -1653,7 +1653,7 @@ static void nv_adma_error_handler(struct ata_port *ap)
 		readw(mmio + NV_ADMA_CTL);	/* flush posted write */
 	}
 
-	ata_bmdma_error_handler(ap);
+	ata_sff_error_handler(ap);
 }
 
 static void nv_swncq_qc_to_dq(struct ata_port *ap, struct ata_queued_cmd *qc)
@@ -1779,7 +1779,7 @@ static void nv_swncq_error_handler(struct ata_port *ap)
 		ehc->i.action |= ATA_EH_RESET;
 	}
 
-	ata_bmdma_error_handler(ap);
+	ata_sff_error_handler(ap);
 }
 
 #ifdef CONFIG_PM
@@ -1925,7 +1925,7 @@ static int nv_swncq_port_start(struct ata_port *ap)
 static void nv_swncq_qc_prep(struct ata_queued_cmd *qc)
 {
 	if (qc->tf.protocol != ATA_PROT_NCQ) {
-		ata_qc_prep(qc);
+		ata_sff_qc_prep(qc);
 		return;
 	}
 
@@ -2001,7 +2001,7 @@ static unsigned int nv_swncq_qc_issue(struct ata_queued_cmd *qc)
 	struct nv_swncq_port_priv *pp = ap->private_data;
 
 	if (qc->tf.protocol != ATA_PROT_NCQ)
-		return ata_qc_issue_prot(qc);
+		return ata_sff_qc_issue(qc);
 
 	DPRINTK("Enter\n");
 
@@ -2350,7 +2350,7 @@ static int nv_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	ppi[0] = &nv_port_info[type];
 	ipriv = ppi[0]->private_data;
-	rc = ata_pci_prepare_sff_host(pdev, ppi, &host);
+	rc = ata_pci_sff_prepare_host(pdev, ppi, &host);
 	if (rc)
 		return rc;
 
diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c
index 5a0a03a08abc..9923e860eae3 100644
--- a/drivers/ata/sata_promise.c
+++ b/drivers/ata/sata_promise.c
@@ -143,7 +143,7 @@ static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile
 static int pdc_check_atapi_dma(struct ata_queued_cmd *qc);
 static int pdc_old_sata_check_atapi_dma(struct ata_queued_cmd *qc);
 static void pdc_irq_clear(struct ata_port *ap);
-static unsigned int pdc_qc_issue_prot(struct ata_queued_cmd *qc);
+static unsigned int pdc_qc_issue(struct ata_queued_cmd *qc);
 static void pdc_freeze(struct ata_port *ap);
 static void pdc_sata_freeze(struct ata_port *ap);
 static void pdc_thaw(struct ata_port *ap);
@@ -166,7 +166,7 @@ static const struct ata_port_operations pdc_common_ops = {
 	.exec_command		= pdc_exec_command_mmio,
 	.check_atapi_dma	= pdc_check_atapi_dma,
 	.qc_prep		= pdc_qc_prep,
-	.qc_issue		= pdc_qc_issue_prot,
+	.qc_issue		= pdc_qc_issue,
 	.irq_clear		= pdc_irq_clear,
 
 	.post_internal_cmd	= pdc_post_internal_cmd,
@@ -894,7 +894,7 @@ static inline void pdc_packet_start(struct ata_queued_cmd *qc)
 	readl(ap->ioaddr.cmd_addr + PDC_PKT_SUBMIT); /* flush */
 }
 
-static unsigned int pdc_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int pdc_qc_issue(struct ata_queued_cmd *qc)
 {
 	switch (qc->tf.protocol) {
 	case ATAPI_PROT_NODATA:
@@ -914,20 +914,20 @@ static unsigned int pdc_qc_issue_prot(struct ata_queued_cmd *qc)
 		break;
 	}
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf)
 {
 	WARN_ON(tf->protocol == ATA_PROT_DMA || tf->protocol == ATAPI_PROT_DMA);
-	ata_tf_load(ap, tf);
+	ata_sff_tf_load(ap, tf);
 }
 
 static void pdc_exec_command_mmio(struct ata_port *ap,
 				  const struct ata_taskfile *tf)
 {
 	WARN_ON(tf->protocol == ATA_PROT_DMA || tf->protocol == ATAPI_PROT_DMA);
-	ata_exec_command(ap, tf);
+	ata_sff_exec_command(ap, tf);
 }
 
 static int pdc_check_atapi_dma(struct ata_queued_cmd *qc)
diff --git a/drivers/ata/sata_qstor.c b/drivers/ata/sata_qstor.c
index 2ceb0990bcd8..1600107047cf 100644
--- a/drivers/ata/sata_qstor.c
+++ b/drivers/ata/sata_qstor.c
@@ -239,7 +239,7 @@ static int qs_prereset(struct ata_link *link, unsigned long deadline)
 	struct ata_port *ap = link->ap;
 
 	qs_reset_channel_logic(ap);
-	return ata_std_prereset(link, deadline);
+	return ata_sff_prereset(link, deadline);
 }
 
 static int qs_scr_read(struct ata_port *ap, unsigned int sc_reg, u32 *val)
@@ -303,7 +303,7 @@ static void qs_qc_prep(struct ata_queued_cmd *qc)
 
 	qs_enter_reg_mode(qc->ap);
 	if (qc->tf.protocol != ATA_PROT_DMA) {
-		ata_qc_prep(qc);
+		ata_sff_qc_prep(qc);
 		return;
 	}
 
@@ -362,7 +362,7 @@ static unsigned int qs_qc_issue(struct ata_queued_cmd *qc)
 	}
 
 	pp->state = qs_state_mmio;
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static void qs_do_or_die(struct ata_queued_cmd *qc, u8 status)
@@ -451,7 +451,7 @@ static inline unsigned int qs_intr_mmio(struct ata_host *host)
 				 * and pretend we knew it was ours.. (ugh).
 				 * This does not affect packet mode.
 				 */
-				ata_check_status(ap);
+				ata_sff_check_status(ap);
 				handled = 1;
 				continue;
 			}
@@ -459,7 +459,7 @@ static inline unsigned int qs_intr_mmio(struct ata_host *host)
 			if (!pp || pp->state != qs_state_mmio)
 				continue;
 			if (!(qc->tf.flags & ATA_TFLAG_POLLING))
-				handled |= ata_host_intr(ap, qc);
+				handled |= ata_sff_host_intr(ap, qc);
 		}
 	}
 	return handled;
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index 659dfcbdc1b2..987313b68f3b 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -410,10 +410,10 @@ static void sil_host_intr(struct ata_port *ap, u32 bmdma2)
 		goto err_hsm;
 
 	/* ack bmdma irq events */
-	ata_bmdma_irq_clear(ap);
+	ata_sff_irq_clear(ap);
 
 	/* kick HSM in the ass */
-	ata_hsm_move(ap, qc, status, 0);
+	ata_sff_hsm_move(ap, qc, status, 0);
 
 	if (unlikely(qc->err_mask) && ata_is_dma(qc->tf.protocol))
 		ata_ehi_push_desc(ehi, "BMDMA2 stat 0x%x", bmdma2);
@@ -481,7 +481,7 @@ static void sil_thaw(struct ata_port *ap)
 
 	/* clear IRQ */
 	ap->ops->check_status(ap);
-	ata_bmdma_irq_clear(ap);
+	ata_sff_irq_clear(ap);
 
 	/* turn on SATA IRQ if supported */
 	if (!(ap->flags & SIL_FLAG_NO_SATA_IRQ))
@@ -655,7 +655,7 @@ static int sil_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 		ioaddr->ctl_addr = mmio_base + sil_port[i].ctl;
 		ioaddr->bmdma_addr = mmio_base + sil_port[i].bmdma;
 		ioaddr->scr_addr = mmio_base + sil_port[i].scr;
-		ata_std_ports(ioaddr);
+		ata_sff_std_ports(ioaddr);
 
 		ata_port_pbar_desc(ap, SIL_MMIO_BAR, -1, "mmio");
 		ata_port_pbar_desc(ap, SIL_MMIO_BAR, sil_port[i].tf, "tf");
diff --git a/drivers/ata/sata_sis.c b/drivers/ata/sata_sis.c
index 9089c7ab5000..6b8e45ba32e8 100644
--- a/drivers/ata/sata_sis.c
+++ b/drivers/ata/sata_sis.c
@@ -309,7 +309,7 @@ static int sis_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 		break;
 	}
 
-	rc = ata_pci_prepare_sff_host(pdev, ppi, &host);
+	rc = ata_pci_sff_prepare_host(pdev, ppi, &host);
 	if (rc)
 		return rc;
 
@@ -327,8 +327,8 @@ static int sis_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	pci_set_master(pdev);
 	pci_intx(pdev, 1);
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &sis_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &sis_sht);
 }
 
 static int __init sis_init(void)
diff --git a/drivers/ata/sata_svw.c b/drivers/ata/sata_svw.c
index 8636f164256e..7b941106f7de 100644
--- a/drivers/ata/sata_svw.c
+++ b/drivers/ata/sata_svw.c
@@ -492,8 +492,8 @@ static int k2_sata_init_one(struct pci_dev *pdev, const struct pci_device_id *en
 	writel(0x0, mmio_base + K2_SATA_SIM_OFFSET);
 
 	pci_set_master(pdev);
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &k2_sata_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &k2_sata_sht);
 }
 
 /* 0x240 is device ID for Apple K2 device
diff --git a/drivers/ata/sata_sx4.c b/drivers/ata/sata_sx4.c
index 8d7f39532d4a..40be2ff60902 100644
--- a/drivers/ata/sata_sx4.c
+++ b/drivers/ata/sata_sx4.c
@@ -232,7 +232,7 @@ static void pdc20621_get_from_dimm(struct ata_host *host,
 static void pdc20621_put_to_dimm(struct ata_host *host,
 				 void *psource, u32 offset, u32 size);
 static void pdc20621_irq_clear(struct ata_port *ap);
-static unsigned int pdc20621_qc_issue_prot(struct ata_queued_cmd *qc);
+static unsigned int pdc20621_qc_issue(struct ata_queued_cmd *qc);
 
 
 static struct scsi_host_template pdc_sata_sht = {
@@ -244,17 +244,17 @@ static struct scsi_host_template pdc_sata_sht = {
 /* TODO: inherit from base port_ops after converting to new EH */
 static struct ata_port_operations pdc_20621_ops = {
 	.tf_load		= pdc_tf_load_mmio,
-	.tf_read		= ata_tf_read,
-	.check_status		= ata_check_status,
+	.tf_read		= ata_sff_tf_read,
+	.check_status		= ata_sff_check_status,
 	.exec_command		= pdc_exec_command_mmio,
-	.dev_select		= ata_std_dev_select,
+	.dev_select		= ata_sff_dev_select,
 	.phy_reset		= pdc_20621_phy_reset,
 	.qc_prep		= pdc20621_qc_prep,
-	.qc_issue		= pdc20621_qc_issue_prot,
-	.data_xfer		= ata_data_xfer,
+	.qc_issue		= pdc20621_qc_issue,
+	.data_xfer		= ata_sff_data_xfer,
 	.eng_timeout		= pdc_eng_timeout,
 	.irq_clear		= pdc20621_irq_clear,
-	.irq_on			= ata_irq_on,
+	.irq_on			= ata_sff_irq_on,
 	.port_start		= pdc_port_start,
 };
 
@@ -682,7 +682,7 @@ static void pdc20621_packet_start(struct ata_queued_cmd *qc)
 	}
 }
 
-static unsigned int pdc20621_qc_issue_prot(struct ata_queued_cmd *qc)
+static unsigned int pdc20621_qc_issue(struct ata_queued_cmd *qc)
 {
 	switch (qc->tf.protocol) {
 	case ATA_PROT_DMA:
@@ -698,7 +698,7 @@ static unsigned int pdc20621_qc_issue_prot(struct ata_queued_cmd *qc)
 		break;
 	}
 
-	return ata_qc_issue_prot(qc);
+	return ata_sff_qc_issue(qc);
 }
 
 static inline unsigned int pdc20621_host_intr(struct ata_port *ap,
@@ -770,7 +770,7 @@ static inline unsigned int pdc20621_host_intr(struct ata_port *ap,
 	/* command completion, but no data xfer */
 	} else if (qc->tf.protocol == ATA_PROT_NODATA) {
 
-		status = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
+		status = ata_sff_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
 		DPRINTK("BUS_NODATA (drv_stat 0x%X)\n", status);
 		qc->err_mask |= ac_err_mask(status);
 		ata_qc_complete(qc);
@@ -879,7 +879,7 @@ static void pdc_eng_timeout(struct ata_port *ap)
 		break;
 
 	default:
-		drv_stat = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
+		drv_stat = ata_sff_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
 
 		ata_port_printk(ap, KERN_ERR,
 				"unknown timeout, cmd 0x%x stat 0x%x\n",
@@ -898,7 +898,7 @@ static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf)
 {
 	WARN_ON(tf->protocol == ATA_PROT_DMA ||
 		tf->protocol == ATA_PROT_NODATA);
-	ata_tf_load(ap, tf);
+	ata_sff_tf_load(ap, tf);
 }
 
 
@@ -906,7 +906,7 @@ static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile
 {
 	WARN_ON(tf->protocol == ATA_PROT_DMA ||
 		tf->protocol == ATA_PROT_NODATA);
-	ata_exec_command(ap, tf);
+	ata_sff_exec_command(ap, tf);
 }
 
 
diff --git a/drivers/ata/sata_uli.c b/drivers/ata/sata_uli.c
index 6ecd13fefa1a..f277cea904ce 100644
--- a/drivers/ata/sata_uli.c
+++ b/drivers/ata/sata_uli.c
@@ -175,11 +175,11 @@ static int uli_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	host->private_data = hpriv;
 
 	/* the first two ports are standard SFF */
-	rc = ata_pci_init_sff_host(host);
+	rc = ata_pci_sff_init_host(host);
 	if (rc)
 		return rc;
 
-	rc = ata_pci_init_bmdma(host);
+	rc = ata_pci_bmdma_init(host);
 	if (rc)
 		return rc;
 
@@ -200,7 +200,7 @@ static int uli_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 			((unsigned long)iomap[1] | ATA_PCI_CTL_OFS) + 4;
 		ioaddr->bmdma_addr = iomap[4] + 16;
 		hpriv->scr_cfg_addr[2] = ULI5287_BASE + ULI5287_OFFS*4;
-		ata_std_ports(ioaddr);
+		ata_sff_std_ports(ioaddr);
 
 		ata_port_desc(host->ports[2],
 			"cmd 0x%llx ctl 0x%llx bmdma 0x%llx",
@@ -215,7 +215,7 @@ static int uli_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 			((unsigned long)iomap[3] | ATA_PCI_CTL_OFS) + 4;
 		ioaddr->bmdma_addr = iomap[4] + 24;
 		hpriv->scr_cfg_addr[3] = ULI5287_BASE + ULI5287_OFFS*5;
-		ata_std_ports(ioaddr);
+		ata_sff_std_ports(ioaddr);
 
 		ata_port_desc(host->ports[2],
 			"cmd 0x%llx ctl 0x%llx bmdma 0x%llx",
@@ -242,8 +242,8 @@ static int uli_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	pci_set_master(pdev);
 	pci_intx(pdev, 1);
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &uli_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &uli_sht);
 }
 
 static int __init uli_init(void)
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index 4bc6e849af2c..9323bc2a2785 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -174,7 +174,7 @@ static void svia_noop_freeze(struct ata_port *ap)
 	 * certain way.  Leave it alone and just clear pending IRQ.
 	 */
 	ap->ops->check_status(ap);
-	ata_bmdma_irq_clear(ap);
+	ata_sff_irq_clear(ap);
 }
 
 /**
@@ -242,7 +242,7 @@ static int vt6420_prereset(struct ata_link *link, unsigned long deadline)
 
  skip_scr:
 	/* wait for !BSY */
-	ata_wait_ready(ap, deadline);
+	ata_sff_wait_ready(ap, deadline);
 
 	return 0;
 }
@@ -304,7 +304,7 @@ static void vt6421_init_addrs(struct ata_port *ap)
 	ioaddr->bmdma_addr = bmdma_addr;
 	ioaddr->scr_addr = vt6421_scr_addr(iomap[5], ap->port_no);
 
-	ata_std_ports(ioaddr);
+	ata_sff_std_ports(ioaddr);
 
 	ata_port_pbar_desc(ap, ap->port_no, -1, "port");
 	ata_port_pbar_desc(ap, 4, ap->port_no * 8, "bmdma");
@@ -316,7 +316,7 @@ static int vt6420_prepare_host(struct pci_dev *pdev, struct ata_host **r_host)
 	struct ata_host *host;
 	int rc;
 
-	rc = ata_pci_prepare_sff_host(pdev, ppi, &host);
+	rc = ata_pci_sff_prepare_host(pdev, ppi, &host);
 	if (rc)
 		return rc;
 	*r_host = host;
@@ -448,8 +448,8 @@ static int svia_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	svia_configure(pdev);
 
 	pci_set_master(pdev);
-	return ata_host_activate(host, pdev->irq, ata_interrupt, IRQF_SHARED,
-				 &svia_sht);
+	return ata_host_activate(host, pdev->irq, ata_sff_interrupt,
+				 IRQF_SHARED, &svia_sht);
 }
 
 static int __init svia_init(void)
diff --git a/drivers/ata/sata_vsc.c b/drivers/ata/sata_vsc.c
index fb3a88722664..2c3c7693c750 100644
--- a/drivers/ata/sata_vsc.c
+++ b/drivers/ata/sata_vsc.c
@@ -200,7 +200,7 @@ static void vsc_sata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 	u16 nsect, lbal, lbam, lbah, feature;
 
-	tf->command = ata_check_status(ap);
+	tf->command = ata_sff_check_status(ap);
 	tf->device = readw(ioaddr->device_addr);
 	feature = readw(ioaddr->error_addr);
 	nsect = readw(ioaddr->nsect_addr);
@@ -243,7 +243,7 @@ static void vsc_port_intr(u8 port_status, struct ata_port *ap)
 
 	qc = ata_qc_from_tag(ap, ap->link.active_tag);
 	if (qc && likely(!(qc->tf.flags & ATA_TFLAG_POLLING)))
-		handled = ata_host_intr(ap, qc);
+		handled = ata_sff_host_intr(ap, qc);
 
 	/* We received an interrupt during a polled command,
 	 * or some other spurious condition.  Interrupt reporting
diff --git a/include/linux/libata.h b/include/linux/libata.h
index a05de2ba7a72..66663bfe2c71 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1342,45 +1342,48 @@ extern const struct ata_port_operations ata_bmdma_port_ops;
 	.sg_tablesize		= LIBATA_MAX_PRD,		\
 	.dma_boundary		= ATA_DMA_BOUNDARY
 
-extern void ata_qc_prep(struct ata_queued_cmd *qc);
-extern void ata_dumb_qc_prep(struct ata_queued_cmd *qc);
-extern void ata_std_dev_select(struct ata_port *ap, unsigned int device);
-extern u8 ata_check_status(struct ata_port *ap);
-extern u8 ata_altstatus(struct ata_port *ap);
-extern int ata_busy_sleep(struct ata_port *ap,
-			  unsigned long timeout_pat, unsigned long timeout);
-extern int ata_wait_ready(struct ata_port *ap, unsigned long deadline);
-extern void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf);
-extern void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
-extern void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf);
-extern unsigned int ata_data_xfer(struct ata_device *dev,
+extern void ata_sff_qc_prep(struct ata_queued_cmd *qc);
+extern void ata_sff_dumb_qc_prep(struct ata_queued_cmd *qc);
+extern void ata_sff_dev_select(struct ata_port *ap, unsigned int device);
+extern u8 ata_sff_check_status(struct ata_port *ap);
+extern u8 ata_sff_altstatus(struct ata_port *ap);
+extern int ata_sff_busy_sleep(struct ata_port *ap,
+			      unsigned long timeout_pat, unsigned long timeout);
+extern int ata_sff_wait_ready(struct ata_port *ap, unsigned long deadline);
+extern void ata_sff_tf_load(struct ata_port *ap, const struct ata_taskfile *tf);
+extern void ata_sff_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
+extern void ata_sff_exec_command(struct ata_port *ap,
+				 const struct ata_taskfile *tf);
+extern unsigned int ata_sff_data_xfer(struct ata_device *dev,
 			unsigned char *buf, unsigned int buflen, int rw);
-extern unsigned int ata_data_xfer_noirq(struct ata_device *dev,
+extern unsigned int ata_sff_data_xfer_noirq(struct ata_device *dev,
 			unsigned char *buf, unsigned int buflen, int rw);
-extern u8 ata_irq_on(struct ata_port *ap);
-extern void ata_bmdma_irq_clear(struct ata_port *ap);
-extern int ata_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
-			u8 status, int in_wq);
-extern unsigned int ata_qc_issue_prot(struct ata_queued_cmd *qc);
-extern unsigned int ata_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc);
-extern irqreturn_t ata_interrupt(int irq, void *dev_instance);
-extern void ata_bmdma_freeze(struct ata_port *ap);
-extern void ata_bmdma_thaw(struct ata_port *ap);
-extern int ata_std_prereset(struct ata_link *link, unsigned long deadline);
-extern unsigned int ata_dev_try_classify(struct ata_device *dev, int present,
-					 u8 *r_err);
-extern void ata_wait_after_reset(struct ata_port *ap, unsigned long deadline);
-extern int ata_std_softreset(struct ata_link *link, unsigned int *classes,
+extern u8 ata_sff_irq_on(struct ata_port *ap);
+extern void ata_sff_irq_clear(struct ata_port *ap);
+extern int ata_sff_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
+			    u8 status, int in_wq);
+extern unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc);
+extern unsigned int ata_sff_host_intr(struct ata_port *ap,
+				      struct ata_queued_cmd *qc);
+extern irqreturn_t ata_sff_interrupt(int irq, void *dev_instance);
+extern void ata_sff_freeze(struct ata_port *ap);
+extern void ata_sff_thaw(struct ata_port *ap);
+extern int ata_sff_prereset(struct ata_link *link, unsigned long deadline);
+extern unsigned int ata_sff_dev_classify(struct ata_device *dev, int present,
+					  u8 *r_err);
+extern void ata_sff_wait_after_reset(struct ata_port *ap,
+				     unsigned long deadline);
+extern int ata_sff_softreset(struct ata_link *link, unsigned int *classes,
 			     unsigned long deadline);
-extern int sata_std_hardreset(struct ata_link *link, unsigned int *class,
-			      unsigned long deadline);
-extern void ata_std_postreset(struct ata_link *link, unsigned int *classes);
-extern void ata_bmdma_error_handler(struct ata_port *ap);
-extern void ata_bmdma_post_internal_cmd(struct ata_queued_cmd *qc);
+extern int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
+			       unsigned long deadline);
+extern void ata_sff_postreset(struct ata_link *link, unsigned int *classes);
+extern void ata_sff_error_handler(struct ata_port *ap);
+extern void ata_sff_post_internal_cmd(struct ata_queued_cmd *qc);
 extern int ata_sff_port_start(struct ata_port *ap);
-extern void ata_std_ports(struct ata_ioports *ioaddr);
-extern unsigned long ata_pci_default_filter(struct ata_device *dev,
-					    unsigned long xfer_mask);
+extern void ata_sff_std_ports(struct ata_ioports *ioaddr);
+extern unsigned long ata_bmdma_mode_filter(struct ata_device *dev,
+					   unsigned long xfer_mask);
 extern void ata_bmdma_setup(struct ata_queued_cmd *qc);
 extern void ata_bmdma_start(struct ata_queued_cmd *qc);
 extern void ata_bmdma_stop(struct ata_queued_cmd *qc);
@@ -1388,35 +1391,35 @@ extern u8 ata_bmdma_status(struct ata_port *ap);
 extern void ata_bus_reset(struct ata_port *ap);
 
 #ifdef CONFIG_PCI
-extern int ata_pci_clear_simplex(struct pci_dev *pdev);
-extern int ata_pci_init_bmdma(struct ata_host *host);
-extern int ata_pci_init_sff_host(struct ata_host *host);
-extern int ata_pci_prepare_sff_host(struct pci_dev *pdev,
+extern int ata_pci_bmdma_clear_simplex(struct pci_dev *pdev);
+extern int ata_pci_bmdma_init(struct ata_host *host);
+extern int ata_pci_sff_init_host(struct ata_host *host);
+extern int ata_pci_sff_prepare_host(struct pci_dev *pdev,
 				    const struct ata_port_info * const * ppi,
 				    struct ata_host **r_host);
-extern int ata_pci_activate_sff_host(struct ata_host *host,
+extern int ata_pci_sff_activate_host(struct ata_host *host,
 				     irq_handler_t irq_handler,
 				     struct scsi_host_template *sht);
-extern int ata_pci_init_one(struct pci_dev *pdev,
-			    const struct ata_port_info * const * ppi,
-			    struct scsi_host_template *sht, void *host_priv);
+extern int ata_pci_sff_init_one(struct pci_dev *pdev,
+				const struct ata_port_info * const * ppi,
+				struct scsi_host_template *sht, void *host_priv);
 #endif /* CONFIG_PCI */
 
 /**
- *	ata_pause - Flush writes and pause 400 nanoseconds.
+ *	ata_sff_pause - Flush writes and pause 400 nanoseconds.
  *	@ap: Port to wait for.
  *
  *	LOCKING:
  *	Inherited from caller.
  */
-static inline void ata_pause(struct ata_port *ap)
+static inline void ata_sff_pause(struct ata_port *ap)
 {
-	ata_altstatus(ap);
+	ata_sff_altstatus(ap);
 	ndelay(400);
 }
 
 /**
- *	ata_busy_wait - Wait for a port status register
+ *	ata_sff_busy_wait - Wait for a port status register
  *	@ap: Port to wait for.
  *	@bits: bits that must be clear
  *	@max: number of 10uS waits to perform
@@ -1428,8 +1431,8 @@ static inline void ata_pause(struct ata_port *ap)
  *	LOCKING:
  *	Inherited from caller.
  */
-static inline u8 ata_busy_wait(struct ata_port *ap, unsigned int bits,
-			       unsigned int max)
+static inline u8 ata_sff_busy_wait(struct ata_port *ap, unsigned int bits,
+				   unsigned int max)
 {
 	u8 status;
 
@@ -1454,7 +1457,7 @@ static inline u8 ata_busy_wait(struct ata_port *ap, unsigned int bits,
  */
 static inline u8 ata_wait_idle(struct ata_port *ap)
 {
-	u8 status = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
+	u8 status = ata_sff_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000);
 
 #ifdef ATA_DEBUG
 	if (status != 0xff && (status & (ATA_BUSY | ATA_DRQ)))
-- 
cgit v1.2.3


From 5682ed33aae05d10a25c95633ef9d9c062825888 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:16 +0900
Subject: libata: rename SFF port ops

Add sff_ prefix to SFF specific port ops.

This rename is in preparation of separating SFF support out of libata
core layer.  This patch strictly renames ops and doesn't introduce any
behavior difference.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c            |   6 +--
 drivers/ata/libata-core.c     |  18 +++----
 drivers/ata/libata-scsi.c     |   2 +-
 drivers/ata/libata-sff.c      | 109 ++++++++++++++++++++++--------------------
 drivers/ata/pata_bf54x.c      |  20 ++++----
 drivers/ata/pata_cmd640.c     |   2 +-
 drivers/ata/pata_icside.c     |   4 +-
 drivers/ata/pata_it821x.c     |   2 +-
 drivers/ata/pata_ixp4xx_cf.c  |   2 +-
 drivers/ata/pata_legacy.c     |  14 +++---
 drivers/ata/pata_mpc52xx.c    |   2 +-
 drivers/ata/pata_ninja32.c    |   2 +-
 drivers/ata/pata_ns87415.c    |   8 ++--
 drivers/ata/pata_pcmcia.c     |   4 +-
 drivers/ata/pata_platform.c   |   2 +-
 drivers/ata/pata_qdi.c        |   2 +-
 drivers/ata/pata_rb500_cf.c   |   4 +-
 drivers/ata/pata_scc.c        |  42 ++++++++--------
 drivers/ata/pata_via.c        |   2 +-
 drivers/ata/pata_winbond.c    |   2 +-
 drivers/ata/sata_fsl.c        |   6 +--
 drivers/ata/sata_inic162x.c   |  12 ++---
 drivers/ata/sata_nv.c         |  20 ++++----
 drivers/ata/sata_promise.c    |   6 +--
 drivers/ata/sata_sil.c        |   6 +--
 drivers/ata/sata_sil24.c      |   6 +--
 drivers/ata/sata_svw.c        |  10 ++--
 drivers/ata/sata_sx4.c        |  16 +++----
 drivers/ata/sata_via.c        |   2 +-
 drivers/ata/sata_vsc.c        |   6 +--
 drivers/scsi/ipr.c            |   8 ++--
 drivers/scsi/libsas/sas_ata.c |   8 ++--
 include/linux/libata.h        |  25 +++++-----
 33 files changed, 192 insertions(+), 188 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 771509c9a3fb..c5e4501daa74 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -292,10 +292,10 @@ static struct scsi_host_template ahci_sht = {
 static struct ata_port_operations ahci_ops = {
 	.inherits		= &sata_pmp_port_ops,
 
-	.check_status		= ahci_check_status,
-	.check_altstatus	= ahci_check_status,
+	.sff_check_status	= ahci_check_status,
+	.sff_check_altstatus	= ahci_check_status,
 
-	.tf_read		= ahci_tf_read,
+	.sff_tf_read		= ahci_tf_read,
 	.qc_defer		= sata_pmp_qc_defer_cmd_switch,
 	.qc_prep		= ahci_qc_prep,
 	.qc_issue		= ahci_qc_issue,
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index cbdbfb5eaeaa..a01e02c5ce7a 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -74,7 +74,7 @@ const unsigned long sata_deb_timing_hotplug[]		= {  25,  500, 2000 };
 const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
 const struct ata_port_operations ata_base_port_ops = {
-	.irq_clear		= ata_noop_irq_clear,
+	.sff_irq_clear		= ata_noop_irq_clear,
 	.prereset		= ata_sff_prereset,
 	.hardreset		= sata_sff_hardreset,
 	.postreset		= ata_sff_postreset,
@@ -85,7 +85,7 @@ const struct ata_port_operations sata_port_ops = {
 	.inherits		= &ata_base_port_ops,
 
 	.qc_defer		= ata_std_qc_defer,
-	.dev_select		= ata_noop_dev_select,
+	.sff_dev_select		= ata_noop_dev_select,
 };
 
 const struct ata_port_operations sata_pmp_port_ops = {
@@ -3563,9 +3563,9 @@ void ata_sff_postreset(struct ata_link *link, unsigned int *classes)
 
 	/* is double-select really necessary? */
 	if (classes[0] != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 1);
+		ap->ops->sff_dev_select(ap, 1);
 	if (classes[1] != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 0);
+		ap->ops->sff_dev_select(ap, 0);
 
 	/* bail out if no device is present */
 	if (classes[0] == ATA_DEV_NONE && classes[1] == ATA_DEV_NONE) {
@@ -4416,7 +4416,7 @@ static void fill_result_tf(struct ata_queued_cmd *qc)
 	struct ata_port *ap = qc->ap;
 
 	qc->result_tf.flags = qc->tf.flags;
-	ap->ops->tf_read(ap, &qc->result_tf);
+	ap->ops->sff_tf_read(ap, &qc->result_tf);
 }
 
 static void ata_verify_xfer(struct ata_queued_cmd *qc)
@@ -6049,16 +6049,16 @@ static unsigned int ata_dummy_qc_issue(struct ata_queued_cmd *qc)
 }
 
 struct ata_port_operations ata_dummy_port_ops = {
-	.check_status		= ata_dummy_check_status,
-	.check_altstatus	= ata_dummy_check_status,
-	.dev_select		= ata_noop_dev_select,
+	.sff_check_status	= ata_dummy_check_status,
+	.sff_check_altstatus	= ata_dummy_check_status,
+	.sff_dev_select		= ata_noop_dev_select,
 	.qc_prep		= ata_noop_qc_prep,
 	.qc_issue		= ata_dummy_qc_issue,
 	.freeze			= ata_dummy_noret,
 	.thaw			= ata_dummy_noret,
 	.error_handler		= ata_dummy_noret,
 	.post_internal_cmd	= ata_dummy_qc_noret,
-	.irq_clear		= ata_dummy_noret,
+	.sff_irq_clear		= ata_dummy_noret,
 	.port_start		= ata_dummy_ret0,
 	.port_stop		= ata_dummy_noret,
 };
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 798ba5e45710..f8be92836a6e 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -2393,7 +2393,7 @@ static void atapi_request_sense(struct ata_queued_cmd *qc)
 	/* FIXME: is this needed? */
 	memset(cmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
 
-	ap->ops->tf_read(ap, &qc->tf);
+	ap->ops->sff_tf_read(ap, &qc->tf);
 
 	/* fill these in, for the case where they are -not- overwritten */
 	cmd->sense_buffer[0] = 0x70;
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 8544321293d4..04024a556660 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -51,13 +51,13 @@ const struct ata_port_operations ata_sff_port_ops = {
 	.error_handler		= ata_sff_error_handler,
 	.post_internal_cmd	= ata_sff_post_internal_cmd,
 
-	.dev_select		= ata_sff_dev_select,
-	.check_status		= ata_sff_check_status,
-	.tf_load		= ata_sff_tf_load,
-	.tf_read		= ata_sff_tf_read,
-	.exec_command		= ata_sff_exec_command,
-	.data_xfer		= ata_sff_data_xfer,
-	.irq_on			= ata_sff_irq_on,
+	.sff_dev_select		= ata_sff_dev_select,
+	.sff_check_status	= ata_sff_check_status,
+	.sff_tf_load		= ata_sff_tf_load,
+	.sff_tf_read		= ata_sff_tf_read,
+	.sff_exec_command	= ata_sff_exec_command,
+	.sff_data_xfer		= ata_sff_data_xfer,
+	.sff_irq_on		= ata_sff_irq_on,
 
 	.port_start		= ata_sff_port_start,
 };
@@ -71,7 +71,7 @@ const struct ata_port_operations ata_bmdma_port_ops = {
 	.bmdma_start		= ata_bmdma_start,
 	.bmdma_stop		= ata_bmdma_stop,
 	.bmdma_status		= ata_bmdma_status,
-	.irq_clear		= ata_sff_irq_clear,
+	.sff_irq_clear		= ata_sff_irq_clear,
 };
 
 /**
@@ -245,8 +245,8 @@ u8 ata_sff_check_status(struct ata_port *ap)
  */
 u8 ata_sff_altstatus(struct ata_port *ap)
 {
-	if (ap->ops->check_altstatus)
-		return ap->ops->check_altstatus(ap);
+	if (ap->ops->sff_check_altstatus)
+		return ap->ops->sff_check_altstatus(ap);
 
 	return ioread8(ap->ioaddr.altstatus_addr);
 }
@@ -290,7 +290,7 @@ int ata_sff_busy_sleep(struct ata_port *ap,
 	while (status != 0xff && (status & ATA_BUSY) &&
 	       time_before(jiffies, timeout)) {
 		msleep(50);
-		status = ap->ops->check_status(ap);
+		status = ap->ops->sff_check_status(ap);
 	}
 
 	if (status == 0xff)
@@ -326,7 +326,7 @@ int ata_sff_wait_ready(struct ata_port *ap, unsigned long deadline)
 	int warned = 0;
 
 	while (1) {
-		u8 status = ap->ops->check_status(ap);
+		u8 status = ap->ops->sff_check_status(ap);
 		unsigned long now = jiffies;
 
 		if (!(status & ATA_BUSY))
@@ -403,7 +403,7 @@ void ata_dev_select(struct ata_port *ap, unsigned int device,
 	if (wait)
 		ata_wait_idle(ap);
 
-	ap->ops->dev_select(ap, device);
+	ap->ops->sff_dev_select(ap, device);
 
 	if (wait) {
 		if (can_sleep && ap->link.device[device].class == ATA_DEV_ATAPI)
@@ -434,7 +434,7 @@ u8 ata_sff_irq_on(struct ata_port *ap)
 		iowrite8(ap->ctl, ioaddr->ctl_addr);
 	tmp = ata_wait_idle(ap);
 
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_irq_clear(ap);
 
 	return tmp;
 }
@@ -593,8 +593,8 @@ void ata_sff_exec_command(struct ata_port *ap, const struct ata_taskfile *tf)
 static inline void ata_tf_to_host(struct ata_port *ap,
 				  const struct ata_taskfile *tf)
 {
-	ap->ops->tf_load(ap, tf);
-	ap->ops->exec_command(ap, tf);
+	ap->ops->sff_tf_load(ap, tf);
+	ap->ops->sff_exec_command(ap, tf);
 }
 
 /**
@@ -709,13 +709,15 @@ static void ata_pio_sector(struct ata_queued_cmd *qc)
 		buf = kmap_atomic(page, KM_IRQ0);
 
 		/* do the actual data transfer */
-		ap->ops->data_xfer(qc->dev, buf + offset, qc->sect_size, do_write);
+		ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size,
+				       do_write);
 
 		kunmap_atomic(buf, KM_IRQ0);
 		local_irq_restore(flags);
 	} else {
 		buf = page_address(page);
-		ap->ops->data_xfer(qc->dev, buf + offset, qc->sect_size, do_write);
+		ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size,
+				       do_write);
 	}
 
 	qc->curbytes += qc->sect_size;
@@ -772,7 +774,7 @@ static void atapi_send_cdb(struct ata_port *ap, struct ata_queued_cmd *qc)
 	DPRINTK("send cdb\n");
 	WARN_ON(qc->dev->cdb_len < 12);
 
-	ap->ops->data_xfer(qc->dev, qc->cdb, qc->dev->cdb_len, 1);
+	ap->ops->sff_data_xfer(qc->dev, qc->cdb, qc->dev->cdb_len, 1);
 	ata_sff_altstatus(ap); /* flush */
 
 	switch (qc->tf.protocol) {
@@ -844,13 +846,13 @@ next_sg:
 		buf = kmap_atomic(page, KM_IRQ0);
 
 		/* do the actual data transfer */
-		consumed = ap->ops->data_xfer(dev,  buf + offset, count, rw);
+		consumed = ap->ops->sff_data_xfer(dev,  buf + offset, count, rw);
 
 		kunmap_atomic(buf, KM_IRQ0);
 		local_irq_restore(flags);
 	} else {
 		buf = page_address(page);
-		consumed = ap->ops->data_xfer(dev,  buf + offset, count, rw);
+		consumed = ap->ops->sff_data_xfer(dev,  buf + offset, count, rw);
 	}
 
 	bytes -= min(bytes, consumed);
@@ -893,7 +895,7 @@ static void atapi_pio_bytes(struct ata_queued_cmd *qc)
 	 * error, qc->result_tf is later overwritten by ata_qc_complete().
 	 * So, the correctness of qc->result_tf is not affected.
 	 */
-	ap->ops->tf_read(ap, &qc->result_tf);
+	ap->ops->sff_tf_read(ap, &qc->result_tf);
 	ireason = qc->result_tf.nsect;
 	bc_lo = qc->result_tf.lbam;
 	bc_hi = qc->result_tf.lbah;
@@ -979,7 +981,7 @@ static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
 			qc = ata_qc_from_tag(ap, qc->tag);
 			if (qc) {
 				if (likely(!(qc->err_mask & AC_ERR_HSM))) {
-					ap->ops->irq_on(ap);
+					ap->ops->sff_irq_on(ap);
 					ata_qc_complete(qc);
 				} else
 					ata_port_freeze(ap);
@@ -995,7 +997,7 @@ static void ata_hsm_qc_complete(struct ata_queued_cmd *qc, int in_wq)
 	} else {
 		if (in_wq) {
 			spin_lock_irqsave(ap->lock, flags);
-			ap->ops->irq_on(ap);
+			ap->ops->sff_irq_on(ap);
 			ata_qc_complete(qc);
 			spin_unlock_irqrestore(ap->lock, flags);
 		} else
@@ -1345,7 +1347,7 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc)
 	case ATA_PROT_DMA:
 		WARN_ON(qc->tf.flags & ATA_TFLAG_POLLING);
 
-		ap->ops->tf_load(ap, &qc->tf);	 /* load tf registers */
+		ap->ops->sff_tf_load(ap, &qc->tf);  /* load tf registers */
 		ap->ops->bmdma_setup(qc);	    /* set up bmdma */
 		ap->ops->bmdma_start(qc);	    /* initiate bmdma */
 		ap->hsm_task_state = HSM_ST_LAST;
@@ -1397,7 +1399,7 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc)
 	case ATAPI_PROT_DMA:
 		WARN_ON(qc->tf.flags & ATA_TFLAG_POLLING);
 
-		ap->ops->tf_load(ap, &qc->tf);	 /* load tf registers */
+		ap->ops->sff_tf_load(ap, &qc->tf);  /* load tf registers */
 		ap->ops->bmdma_setup(qc);	    /* set up bmdma */
 		ap->hsm_task_state = HSM_ST_FIRST;
 
@@ -1486,12 +1488,12 @@ inline unsigned int ata_sff_host_intr(struct ata_port *ap,
 		goto idle_irq;
 
 	/* check main status, clearing INTRQ */
-	status = ap->ops->check_status(ap);
+	status = ap->ops->sff_check_status(ap);
 	if (unlikely(status & ATA_BUSY))
 		goto idle_irq;
 
 	/* ack bmdma irq events */
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_irq_clear(ap);
 
 	ata_sff_hsm_move(ap, qc, status, 0);
 
@@ -1506,8 +1508,8 @@ idle_irq:
 
 #ifdef ATA_IRQ_TRAP
 	if ((ap->stats.idle_irq % 1000) == 0) {
-		ap->ops->check_status(ap);
-		ap->ops->irq_clear(ap);
+		ap->ops->sff_check_status(ap);
+		ap->ops->sff_irq_clear(ap);
 		ata_port_printk(ap, KERN_WARNING, "irq trap\n");
 		return 1;
 	}
@@ -1582,9 +1584,9 @@ void ata_sff_freeze(struct ata_port *ap)
 	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
 	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
 	 */
-	ap->ops->check_status(ap);
+	ap->ops->sff_check_status(ap);
 
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_irq_clear(ap);
 }
 
 /**
@@ -1599,9 +1601,9 @@ void ata_sff_freeze(struct ata_port *ap)
 void ata_sff_thaw(struct ata_port *ap)
 {
 	/* clear & re-enable interrupts */
-	ap->ops->check_status(ap);
-	ap->ops->irq_clear(ap);
-	ap->ops->irq_on(ap);
+	ap->ops->sff_check_status(ap);
+	ap->ops->sff_irq_clear(ap);
+	ap->ops->sff_irq_on(ap);
 }
 
 /**
@@ -1626,7 +1628,7 @@ static unsigned int ata_devchk(struct ata_port *ap, unsigned int device)
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 	u8 nsect, lbal;
 
-	ap->ops->dev_select(ap, device);
+	ap->ops->sff_dev_select(ap, device);
 
 	iowrite8(0x55, ioaddr->nsect_addr);
 	iowrite8(0xaa, ioaddr->lbal_addr);
@@ -1675,11 +1677,11 @@ unsigned int ata_sff_dev_classify(struct ata_device *dev, int present,
 	unsigned int class;
 	u8 err;
 
-	ap->ops->dev_select(ap, dev->devno);
+	ap->ops->sff_dev_select(ap, dev->devno);
 
 	memset(&tf, 0, sizeof(tf));
 
-	ap->ops->tf_read(ap, &tf);
+	ap->ops->sff_tf_read(ap, &tf);
 	err = tf.feature;
 	if (r_err)
 		*r_err = err;
@@ -1709,7 +1711,8 @@ unsigned int ata_sff_dev_classify(struct ata_device *dev, int present,
 			class = ATA_DEV_ATA;
 		else
 			class = ATA_DEV_NONE;
-	} else if ((class == ATA_DEV_ATA) && (ap->ops->check_status(ap) == 0))
+	} else if ((class == ATA_DEV_ATA) &&
+		   (ap->ops->sff_check_status(ap) == 0))
 		class = ATA_DEV_NONE;
 
 	return class;
@@ -1741,7 +1744,7 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	if (dev1) {
 		int i;
 
-		ap->ops->dev_select(ap, 1);
+		ap->ops->sff_dev_select(ap, 1);
 
 		/* Wait for register access.  Some ATAPI devices fail
 		 * to set nsect/lbal after reset, so don't waste too
@@ -1766,11 +1769,11 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	}
 
 	/* is all this really necessary? */
-	ap->ops->dev_select(ap, 0);
+	ap->ops->sff_dev_select(ap, 0);
 	if (dev1)
-		ap->ops->dev_select(ap, 1);
+		ap->ops->sff_dev_select(ap, 1);
 	if (dev0)
-		ap->ops->dev_select(ap, 0);
+		ap->ops->sff_dev_select(ap, 0);
 
 	return ret;
 }
@@ -1820,7 +1823,7 @@ void ata_sff_wait_after_reset(struct ata_port *ap, unsigned long deadline)
 	 */
 	if (ap->flags & ATA_FLAG_SATA) {
 		while (1) {
-			u8 status = ap->ops->check_status(ap);
+			u8 status = ap->ops->sff_check_status(ap);
 
 			if (status != 0xff || time_after(jiffies, deadline))
 				return;
@@ -1851,7 +1854,7 @@ static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
 	 * the bus shows 0xFF because the odd clown forgets the D7
 	 * pulldown resistor.
 	 */
-	if (ap->ops->check_status(ap) == 0xFF)
+	if (ap->ops->sff_check_status(ap) == 0xFF)
 		return -ENODEV;
 
 	return ata_bus_post_reset(ap, devmask, deadline);
@@ -1894,7 +1897,7 @@ int ata_sff_softreset(struct ata_link *link, unsigned int *classes,
 		devmask |= (1 << 1);
 
 	/* select device 0 again */
-	ap->ops->dev_select(ap, 0);
+	ap->ops->sff_dev_select(ap, 0);
 
 	/* issue bus reset */
 	DPRINTK("about to softreset, devmask=%x\n", devmask);
@@ -1977,7 +1980,7 @@ int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
 		return rc;
 	}
 
-	ap->ops->dev_select(ap, 0);	/* probably unnecessary */
+	ap->ops->sff_dev_select(ap, 0);	/* probably unnecessary */
 
 	*class = ata_sff_dev_classify(link->device, 1, NULL);
 
@@ -2035,8 +2038,8 @@ void ata_sff_error_handler(struct ata_port *ap)
 	}
 
 	ata_sff_altstatus(ap);
-	ap->ops->check_status(ap);
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_check_status(ap);
+	ap->ops->sff_irq_clear(ap);
 
 	spin_unlock_irqrestore(ap->lock, flags);
 
@@ -2153,7 +2156,7 @@ void ata_bmdma_setup(struct ata_queued_cmd *qc)
 	iowrite8(dmactl, ap->ioaddr.bmdma_addr + ATA_DMA_CMD);
 
 	/* issue r/w command */
-	ap->ops->exec_command(ap, &qc->tf);
+	ap->ops->sff_exec_command(ap, &qc->tf);
 }
 
 /**
@@ -2277,7 +2280,7 @@ void ata_bus_reset(struct ata_port *ap)
 		devmask |= (1 << 1);
 
 	/* select device 0 again */
-	ap->ops->dev_select(ap, 0);
+	ap->ops->sff_dev_select(ap, 0);
 
 	/* issue bus reset */
 	if (ap->flags & ATA_FLAG_SRST) {
@@ -2295,9 +2298,9 @@ void ata_bus_reset(struct ata_port *ap)
 
 	/* is double-select really necessary? */
 	if (device[1].class != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 1);
+		ap->ops->sff_dev_select(ap, 1);
 	if (device[0].class != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 0);
+		ap->ops->sff_dev_select(ap, 0);
 
 	/* if no devices were detected, disable this port */
 	if ((device[0].class == ATA_DEV_NONE) &&
diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c
index c854e882d4a9..d98bd7455e6b 100644
--- a/drivers/ata/pata_bf54x.c
+++ b/drivers/ata/pata_bf54x.c
@@ -1264,7 +1264,7 @@ static void bfin_freeze(struct ata_port *ap)
 	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
 	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
 	 */
-	ap->ops->check_status(ap);
+	ap->ops->sff_check_status(ap);
 
 	bfin_irq_clear(ap);
 }
@@ -1357,18 +1357,18 @@ static const struct ata_port_operations bfin_pata_ops = {
 	.set_piomode		= bfin_set_piomode,
 	.set_dmamode		= bfin_set_dmamode,
 
-	.tf_load		= bfin_tf_load,
-	.tf_read		= bfin_tf_read,
-	.exec_command		= bfin_exec_command,
-	.check_status		= bfin_check_status,
-	.check_altstatus	= bfin_check_altstatus,
-	.dev_select		= bfin_dev_select,
+	.sff_tf_load		= bfin_tf_load,
+	.sff_tf_read		= bfin_tf_read,
+	.sff_exec_command	= bfin_exec_command,
+	.sff_check_status	= bfin_check_status,
+	.sff_check_altstatus	= bfin_check_altstatus,
+	.sff_dev_select		= bfin_dev_select,
 
 	.bmdma_setup		= bfin_bmdma_setup,
 	.bmdma_start		= bfin_bmdma_start,
 	.bmdma_stop		= bfin_bmdma_stop,
 	.bmdma_status		= bfin_bmdma_status,
-	.data_xfer		= bfin_data_xfer,
+	.sff_data_xfer		= bfin_data_xfer,
 
 	.qc_prep		= ata_noop_qc_prep,
 
@@ -1378,8 +1378,8 @@ static const struct ata_port_operations bfin_pata_ops = {
 	.postreset		= bfin_postreset,
 	.post_internal_cmd	= bfin_bmdma_stop,
 
-	.irq_clear		= bfin_irq_clear,
-	.irq_on			= bfin_irq_on,
+	.sff_irq_clear		= bfin_irq_clear,
+	.sff_irq_on		= bfin_irq_on,
 
 	.port_start		= bfin_port_start,
 	.port_stop		= bfin_port_stop,
diff --git a/drivers/ata/pata_cmd640.c b/drivers/ata/pata_cmd640.c
index a907cf478891..2de30b990278 100644
--- a/drivers/ata/pata_cmd640.c
+++ b/drivers/ata/pata_cmd640.c
@@ -172,7 +172,7 @@ static struct scsi_host_template cmd640_sht = {
 static struct ata_port_operations cmd640_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 	/* In theory xfer_noirq is not needed once we kill the prefetcher */
-	.data_xfer	= ata_sff_data_xfer_noirq,
+	.sff_data_xfer	= ata_sff_data_xfer_noirq,
 	.qc_issue	= cmd640_qc_issue,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= cmd640_set_piomode,
diff --git a/drivers/ata/pata_icside.c b/drivers/ata/pata_icside.c
index 52de9f908fb0..17138436423d 100644
--- a/drivers/ata/pata_icside.c
+++ b/drivers/ata/pata_icside.c
@@ -250,7 +250,7 @@ static void pata_icside_bmdma_setup(struct ata_queued_cmd *qc)
 	set_dma_mode(state->dma, write ? DMA_MODE_WRITE : DMA_MODE_READ);
 
 	/* issue r/w command */
-	ap->ops->exec_command(ap, &qc->tf);
+	ap->ops->sff_exec_command(ap, &qc->tf);
 }
 
 static void pata_icside_bmdma_start(struct ata_queued_cmd *qc)
@@ -336,7 +336,7 @@ static struct ata_port_operations pata_icside_port_ops = {
 	.inherits		= &ata_sff_port_ops,
 	/* no need to build any PRD tables for DMA */
 	.qc_prep		= ata_noop_qc_prep,
-	.data_xfer		= ata_sff_data_xfer_noirq,
+	.sff_data_xfer		= ata_sff_data_xfer_noirq,
 	.bmdma_setup		= pata_icside_bmdma_setup,
 	.bmdma_start		= pata_icside_bmdma_start,
 	.bmdma_stop		= pata_icside_bmdma_stop,
diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c
index 88e37cfcfc4f..e10816931b2f 100644
--- a/drivers/ata/pata_it821x.c
+++ b/drivers/ata/pata_it821x.c
@@ -652,7 +652,7 @@ static struct ata_port_operations it821x_passthru_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
 
 	.check_atapi_dma= it821x_check_atapi_dma,
-	.dev_select 	= it821x_passthru_dev_select,
+	.sff_dev_select	= it821x_passthru_dev_select,
 	.bmdma_start 	= it821x_passthru_bmdma_start,
 	.bmdma_stop	= it821x_passthru_bmdma_stop,
 	.qc_issue	= it821x_passthru_qc_issue,
diff --git a/drivers/ata/pata_ixp4xx_cf.c b/drivers/ata/pata_ixp4xx_cf.c
index 283a8fba79b3..8a175f23b907 100644
--- a/drivers/ata/pata_ixp4xx_cf.c
+++ b/drivers/ata/pata_ixp4xx_cf.c
@@ -93,7 +93,7 @@ static struct scsi_host_template ixp4xx_sht = {
 
 static struct ata_port_operations ixp4xx_port_ops = {
 	.inherits		= &ata_sff_port_ops,
-	.data_xfer		= ixp4xx_mmio_data_xfer,
+	.sff_data_xfer		= ixp4xx_mmio_data_xfer,
 	.cable_detect		= ata_cable_40wire,
 	.set_mode		= ixp4xx_set_mode,
 };
diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c
index f13f10a55ef3..7af4b29cc422 100644
--- a/drivers/ata/pata_legacy.c
+++ b/drivers/ata/pata_legacy.c
@@ -226,12 +226,12 @@ static const struct ata_port_operations legacy_base_port_ops = {
 
 static struct ata_port_operations simple_port_ops = {
 	.inherits	= &legacy_base_port_ops,
-	.data_xfer	= ata_sff_data_xfer_noirq,
+	.sff_data_xfer	= ata_sff_data_xfer_noirq,
 };
 
 static struct ata_port_operations legacy_port_ops = {
 	.inherits	= &legacy_base_port_ops,
-	.data_xfer	= ata_sff_data_xfer_noirq,
+	.sff_data_xfer	= ata_sff_data_xfer_noirq,
 	.set_mode	= legacy_set_mode,
 };
 
@@ -325,7 +325,7 @@ static unsigned int pdc_data_xfer_vlb(struct ata_device *dev,
 static struct ata_port_operations pdc20230_port_ops = {
 	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= pdc20230_set_piomode,
-	.data_xfer	= pdc_data_xfer_vlb,
+	.sff_data_xfer	= pdc_data_xfer_vlb,
 };
 
 /*
@@ -775,19 +775,19 @@ static struct ata_port_operations qdi6500_port_ops = {
 	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= qdi6500_set_piomode,
 	.qc_issue	= qdi_qc_issue,
-	.data_xfer	= vlb32_data_xfer,
+	.sff_data_xfer	= vlb32_data_xfer,
 };
 
 static struct ata_port_operations qdi6580_port_ops = {
 	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= qdi6580_set_piomode,
-	.data_xfer	= vlb32_data_xfer,
+	.sff_data_xfer	= vlb32_data_xfer,
 };
 
 static struct ata_port_operations qdi6580dp_port_ops = {
 	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= qdi6580dp_set_piomode,
-	.data_xfer	= vlb32_data_xfer,
+	.sff_data_xfer	= vlb32_data_xfer,
 };
 
 static DEFINE_SPINLOCK(winbond_lock);
@@ -858,7 +858,7 @@ static int winbond_port(struct platform_device *dev,
 static struct ata_port_operations winbond_port_ops = {
 	.inherits	= &legacy_base_port_ops,
 	.set_piomode	= winbond_set_piomode,
-	.data_xfer	= vlb32_data_xfer,
+	.sff_data_xfer	= vlb32_data_xfer,
 };
 
 static struct legacy_controller controllers[] = {
diff --git a/drivers/ata/pata_mpc52xx.c b/drivers/ata/pata_mpc52xx.c
index 5d1d32a39c4d..bc79df6e7cb0 100644
--- a/drivers/ata/pata_mpc52xx.c
+++ b/drivers/ata/pata_mpc52xx.c
@@ -261,7 +261,7 @@ static struct scsi_host_template mpc52xx_ata_sht = {
 
 static struct ata_port_operations mpc52xx_ata_port_ops = {
 	.inherits		= &ata_sff_port_ops,
-	.dev_select		= mpc52xx_ata_dev_select,
+	.sff_dev_select		= mpc52xx_ata_dev_select,
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= mpc52xx_ata_set_piomode,
 	.post_internal_cmd	= ATA_OP_NULL,
diff --git a/drivers/ata/pata_ninja32.c b/drivers/ata/pata_ninja32.c
index c9c0ea67c71a..565e67cd13fa 100644
--- a/drivers/ata/pata_ninja32.c
+++ b/drivers/ata/pata_ninja32.c
@@ -84,7 +84,7 @@ static struct scsi_host_template ninja32_sht = {
 
 static struct ata_port_operations ninja32_port_ops = {
 	.inherits	= &ata_bmdma_port_ops,
-	.dev_select 	= ninja32_dev_select,
+	.sff_dev_select = ninja32_dev_select,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= ninja32_set_piomode,
 };
diff --git a/drivers/ata/pata_ns87415.c b/drivers/ata/pata_ns87415.c
index 03a52cefc010..ae92b0049bd5 100644
--- a/drivers/ata/pata_ns87415.c
+++ b/drivers/ata/pata_ns87415.c
@@ -138,7 +138,7 @@ static void ns87415_bmdma_setup(struct ata_queued_cmd *qc)
 		dmactl |= ATA_DMA_WR;
 	iowrite8(dmactl, ap->ioaddr.bmdma_addr + ATA_DMA_CMD);
 	/* issue r/w command */
-	ap->ops->exec_command(ap, &qc->tf);
+	ap->ops->sff_exec_command(ap, &qc->tf);
 }
 
 /**
@@ -306,7 +306,7 @@ static struct ata_port_operations ns87415_pata_ops = {
 	.bmdma_setup		= ns87415_bmdma_setup,
 	.bmdma_start		= ns87415_bmdma_start,
 	.bmdma_stop		= ns87415_bmdma_stop,
-	.irq_clear		= ns87415_irq_clear,
+	.sff_irq_clear		= ns87415_irq_clear,
 
 	.cable_detect		= ata_cable_40wire,
 	.set_piomode		= ns87415_set_piomode,
@@ -315,8 +315,8 @@ static struct ata_port_operations ns87415_pata_ops = {
 #if defined(CONFIG_SUPERIO)
 static struct ata_port_operations ns87560_pata_ops = {
 	.inherits		= &ns87415_pata_ops,
-	.tf_read		= ns87560_tf_read,
-	.check_status		= ns87560_check_status,
+	.sff_tf_read		= ns87560_tf_read,
+	.sff_check_status	= ns87560_check_status,
 	.bmdma_status		= ns87560_bmdma_status,
 };
 #endif
diff --git a/drivers/ata/pata_pcmcia.c b/drivers/ata/pata_pcmcia.c
index 2e206c5f869a..3d39f9dfec5a 100644
--- a/drivers/ata/pata_pcmcia.c
+++ b/drivers/ata/pata_pcmcia.c
@@ -133,14 +133,14 @@ static struct scsi_host_template pcmcia_sht = {
 
 static struct ata_port_operations pcmcia_port_ops = {
 	.inherits	= &ata_sff_port_ops,
-	.data_xfer	= ata_sff_data_xfer_noirq,
+	.sff_data_xfer	= ata_sff_data_xfer_noirq,
 	.cable_detect	= ata_cable_40wire,
 	.set_mode	= pcmcia_set_mode,
 };
 
 static struct ata_port_operations pcmcia_8bit_port_ops = {
 	.inherits	= &ata_sff_port_ops,
-	.data_xfer	= ata_data_xfer_8bit,
+	.sff_data_xfer	= ata_data_xfer_8bit,
 	.cable_detect	= ata_cable_40wire,
 	.set_mode	= pcmcia_set_mode_8bit,
 };
diff --git a/drivers/ata/pata_platform.c b/drivers/ata/pata_platform.c
index 1edfc13d05d2..6527c56c34a3 100644
--- a/drivers/ata/pata_platform.c
+++ b/drivers/ata/pata_platform.c
@@ -52,7 +52,7 @@ static struct scsi_host_template pata_platform_sht = {
 
 static struct ata_port_operations pata_platform_port_ops = {
 	.inherits		= &ata_sff_port_ops,
-	.data_xfer		= ata_sff_data_xfer_noirq,
+	.sff_data_xfer		= ata_sff_data_xfer_noirq,
 	.cable_detect		= ata_cable_unknown,
 	.set_mode		= pata_platform_set_mode,
 	.port_start		= ATA_OP_NULL,
diff --git a/drivers/ata/pata_qdi.c b/drivers/ata/pata_qdi.c
index 0d81dd5fdc38..bf45cf017753 100644
--- a/drivers/ata/pata_qdi.c
+++ b/drivers/ata/pata_qdi.c
@@ -160,7 +160,7 @@ static struct scsi_host_template qdi_sht = {
 static struct ata_port_operations qdi6500_port_ops = {
 	.inherits	= &ata_sff_port_ops,
 	.qc_issue	= qdi_qc_issue,
-	.data_xfer	= qdi_data_xfer,
+	.sff_data_xfer	= qdi_data_xfer,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= qdi6500_set_piomode,
 };
diff --git a/drivers/ata/pata_rb500_cf.c b/drivers/ata/pata_rb500_cf.c
index 5b23d79bc92b..800ae4601f44 100644
--- a/drivers/ata/pata_rb500_cf.c
+++ b/drivers/ata/pata_rb500_cf.c
@@ -119,8 +119,8 @@ static irqreturn_t rb500_pata_irq_handler(int irq, void *dev_instance)
 
 static struct ata_port_operations rb500_pata_port_ops = {
 	.inherits		= &ata_sff_port_ops,
-	.exec_command		= rb500_pata_exec_command,
-	.data_xfer		= rb500_pata_data_xfer,
+	.sff_exec_command	= rb500_pata_exec_command,
+	.sff_data_xfer		= rb500_pata_data_xfer,
 	.freeze			= rb500_pata_freeze,
 	.thaw			= rb500_pata_thaw,
 };
diff --git a/drivers/ata/pata_scc.c b/drivers/ata/pata_scc.c
index 701d0addae73..2b9da715c704 100644
--- a/drivers/ata/pata_scc.c
+++ b/drivers/ata/pata_scc.c
@@ -441,7 +441,7 @@ static void scc_bmdma_setup (struct ata_queued_cmd *qc)
 	out_be32(mmio + SCC_DMA_CMD, dmactl);
 
 	/* issue r/w command */
-	ap->ops->exec_command(ap, &qc->tf);
+	ap->ops->sff_exec_command(ap, &qc->tf);
 }
 
 /**
@@ -476,7 +476,7 @@ static unsigned int scc_devchk (struct ata_port *ap,
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 	u8 nsect, lbal;
 
-	ap->ops->dev_select(ap, device);
+	ap->ops->sff_dev_select(ap, device);
 
 	out_be32(ioaddr->nsect_addr, 0x55);
 	out_be32(ioaddr->lbal_addr, 0xaa);
@@ -525,7 +525,7 @@ static int scc_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	while (dev1) {
 		u8 nsect, lbal;
 
-		ap->ops->dev_select(ap, 1);
+		ap->ops->sff_dev_select(ap, 1);
 		nsect = in_be32(ioaddr->nsect_addr);
 		lbal = in_be32(ioaddr->lbal_addr);
 		if ((nsect == 1) && (lbal == 1))
@@ -541,11 +541,11 @@ static int scc_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	}
 
 	/* is all this really necessary? */
-	ap->ops->dev_select(ap, 0);
+	ap->ops->sff_dev_select(ap, 0);
 	if (dev1)
-		ap->ops->dev_select(ap, 1);
+		ap->ops->sff_dev_select(ap, 1);
 	if (dev0)
-		ap->ops->dev_select(ap, 0);
+		ap->ops->sff_dev_select(ap, 0);
 
 	return 0;
 }
@@ -616,7 +616,7 @@ static int scc_softreset(struct ata_link *link, unsigned int *classes,
 		devmask |= (1 << 1);
 
 	/* select device 0 again */
-	ap->ops->dev_select(ap, 0);
+	ap->ops->sff_dev_select(ap, 0);
 
 	/* issue bus reset */
 	DPRINTK("about to softreset, devmask=%x\n", devmask);
@@ -829,7 +829,7 @@ static u8 scc_irq_on (struct ata_port *ap)
 	out_be32(ioaddr->ctl_addr, ap->ctl);
 	tmp = ata_wait_idle(ap);
 
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_irq_clear(ap);
 
 	return tmp;
 }
@@ -854,9 +854,9 @@ static void scc_freeze (struct ata_port *ap)
 	 * ATA_NIEN manipulation.  Also, many controllers fail to mask
 	 * previously pending IRQ on ATA_NIEN assertion.  Clear it.
 	 */
-	ap->ops->check_status(ap);
+	ap->ops->sff_check_status(ap);
 
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_irq_clear(ap);
 }
 
 /**
@@ -887,9 +887,9 @@ static void scc_postreset(struct ata_link *link, unsigned int *classes)
 
 	/* is double-select really necessary? */
 	if (classes[0] != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 1);
+		ap->ops->sff_dev_select(ap, 1);
 	if (classes[1] != ATA_DEV_NONE)
-		ap->ops->dev_select(ap, 0);
+		ap->ops->sff_dev_select(ap, 0);
 
 	/* bail out if no device is present */
 	if (classes[0] == ATA_DEV_NONE && classes[1] == ATA_DEV_NONE) {
@@ -967,18 +967,18 @@ static struct ata_port_operations scc_pata_ops = {
 	.set_dmamode		= scc_set_dmamode,
 	.mode_filter		= scc_mode_filter,
 
-	.tf_load		= scc_tf_load,
-	.tf_read		= scc_tf_read,
-	.exec_command		= scc_exec_command,
-	.check_status		= scc_check_status,
-	.check_altstatus	= scc_check_altstatus,
-	.dev_select		= scc_dev_select,
+	.sff_tf_load		= scc_tf_load,
+	.sff_tf_read		= scc_tf_read,
+	.sff_exec_command	= scc_exec_command,
+	.sff_check_status	= scc_check_status,
+	.sff_check_altstatus	= scc_check_altstatus,
+	.sff_dev_select		= scc_dev_select,
 
 	.bmdma_setup		= scc_bmdma_setup,
 	.bmdma_start		= scc_bmdma_start,
 	.bmdma_stop		= scc_bmdma_stop,
 	.bmdma_status		= scc_bmdma_status,
-	.data_xfer		= scc_data_xfer,
+	.sff_data_xfer		= scc_data_xfer,
 
 	.freeze			= scc_freeze,
 	.prereset		= scc_pata_prereset,
@@ -986,8 +986,8 @@ static struct ata_port_operations scc_pata_ops = {
 	.postreset		= scc_postreset,
 	.post_internal_cmd	= scc_bmdma_stop,
 
-	.irq_clear		= scc_irq_clear,
-	.irq_on			= scc_irq_on,
+	.sff_irq_clear		= scc_irq_clear,
+	.sff_irq_on		= scc_irq_on,
 
 	.port_start		= scc_port_start,
 	.port_stop		= scc_port_stop,
diff --git a/drivers/ata/pata_via.c b/drivers/ata/pata_via.c
index 4b85f84fbe76..d4840748fb5c 100644
--- a/drivers/ata/pata_via.c
+++ b/drivers/ata/pata_via.c
@@ -336,7 +336,7 @@ static struct ata_port_operations via_port_ops = {
 
 static struct ata_port_operations via_port_ops_noirq = {
 	.inherits	= &via_port_ops,
-	.data_xfer	= ata_sff_data_xfer_noirq,
+	.sff_data_xfer	= ata_sff_data_xfer_noirq,
 };
 
 /**
diff --git a/drivers/ata/pata_winbond.c b/drivers/ata/pata_winbond.c
index cc18231e9334..6e52a3573fbf 100644
--- a/drivers/ata/pata_winbond.c
+++ b/drivers/ata/pata_winbond.c
@@ -127,7 +127,7 @@ static struct scsi_host_template winbond_sht = {
 
 static struct ata_port_operations winbond_port_ops = {
 	.inherits	= &ata_sff_port_ops,
-	.data_xfer	= winbond_data_xfer,
+	.sff_data_xfer	= winbond_data_xfer,
 	.cable_detect	= ata_cable_40wire,
 	.set_piomode	= winbond_set_piomode,
 };
diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c
index 676302fdaa97..37482d4e883f 100644
--- a/drivers/ata/sata_fsl.c
+++ b/drivers/ata/sata_fsl.c
@@ -1190,10 +1190,10 @@ static struct scsi_host_template sata_fsl_sht = {
 static const struct ata_port_operations sata_fsl_ops = {
 	.inherits = &sata_port_ops,
 
-	.check_status = sata_fsl_check_status,
-	.check_altstatus = sata_fsl_check_status,
+	.sff_check_status = sata_fsl_check_status,
+	.sff_check_altstatus = sata_fsl_check_status,
 
-	.tf_read = sata_fsl_tf_read,
+	.sff_tf_read = sata_fsl_tf_read,
 
 	.qc_prep = sata_fsl_qc_prep,
 	.qc_issue = sata_fsl_qc_issue,
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index 6e6fca4c20b6..0b5a736a45e3 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -222,7 +222,7 @@ static void inic_bmdma_setup(struct ata_queued_cmd *qc)
 	writeb(pp->cached_prdctl, port_base + PORT_PRD_CTL);
 
 	/* issue r/w command */
-	ap->ops->exec_command(ap, &qc->tf);
+	ap->ops->sff_exec_command(ap, &qc->tf);
 }
 
 static void inic_bmdma_start(struct ata_queued_cmd *qc)
@@ -267,14 +267,14 @@ static void inic_host_intr(struct ata_port *ap)
 			ata_qc_from_tag(ap, ap->link.active_tag);
 
 		if (unlikely(!qc || (qc->tf.flags & ATA_TFLAG_POLLING))) {
-			ap->ops->check_status(ap); /* clear ATA interrupt */
+			ap->ops->sff_check_status(ap); /* clear ATA interrupt */
 			return;
 		}
 
 		if (likely(ata_sff_host_intr(ap, qc)))
 			return;
 
-		ap->ops->check_status(ap); /* clear ATA interrupt */
+		ap->ops->sff_check_status(ap); /* clear ATA interrupt */
 		ata_port_printk(ap, KERN_WARNING, "unhandled "
 				"interrupt, irq_stat=%x\n", irq_stat);
 		return;
@@ -351,7 +351,7 @@ static unsigned int inic_qc_issue(struct ata_queued_cmd *qc)
 	 */
 	if (unlikely(qc->tf.command == ATA_CMD_ID_ATA ||
 		     qc->tf.command == ATA_CMD_ID_ATAPI)) {
-		u8 stat = ap->ops->check_status(ap);
+		u8 stat = ap->ops->sff_check_status(ap);
 		if (stat == 0x7f || stat == 0xff)
 			return AC_ERR_HSM;
 	}
@@ -365,7 +365,7 @@ static void inic_freeze(struct ata_port *ap)
 
 	__inic_set_pirq_mask(ap, PIRQ_MASK_FREEZE);
 
-	ap->ops->check_status(ap);
+	ap->ops->sff_check_status(ap);
 	writeb(0xff, port_base + PORT_IRQ_STAT);
 
 	readb(port_base + PORT_IRQ_STAT); /* flush */
@@ -375,7 +375,7 @@ static void inic_thaw(struct ata_port *ap)
 {
 	void __iomem *port_base = inic_port_base(ap);
 
-	ap->ops->check_status(ap);
+	ap->ops->sff_check_status(ap);
 	writeb(0xff, port_base + PORT_IRQ_STAT);
 
 	__inic_set_pirq_mask(ap, PIRQ_MASK_OTHER);
diff --git a/drivers/ata/sata_nv.c b/drivers/ata/sata_nv.c
index 63cc43765f04..109b07495721 100644
--- a/drivers/ata/sata_nv.c
+++ b/drivers/ata/sata_nv.c
@@ -429,11 +429,11 @@ static struct ata_port_operations nv_adma_ops = {
 	.inherits		= &nv_generic_ops,
 
 	.check_atapi_dma	= nv_adma_check_atapi_dma,
-	.tf_read		= nv_adma_tf_read,
+	.sff_tf_read		= nv_adma_tf_read,
 	.qc_defer		= ata_std_qc_defer,
 	.qc_prep		= nv_adma_qc_prep,
 	.qc_issue		= nv_adma_qc_issue,
-	.irq_clear		= nv_adma_irq_clear,
+	.sff_irq_clear		= nv_adma_irq_clear,
 
 	.freeze			= nv_adma_freeze,
 	.thaw			= nv_adma_thaw,
@@ -1440,7 +1440,7 @@ static irqreturn_t nv_generic_interrupt(int irq, void *dev_instance)
 			else
 				// No request pending?  Clear interrupt status
 				// anyway, in case there's one pending.
-				ap->ops->check_status(ap);
+				ap->ops->sff_check_status(ap);
 		}
 
 	}
@@ -1739,7 +1739,7 @@ static void nv_swncq_ncq_stop(struct ata_port *ap)
 		pp->dhfis_bits, pp->dmafis_bits, pp->sdbfis_bits);
 
 	ata_port_printk(ap, KERN_ERR, "ATA_REG 0x%X ERR_REG 0x%X\n",
-			ap->ops->check_status(ap),
+			ap->ops->sff_check_status(ap),
 			ioread8(ap->ioaddr.error_addr));
 
 	sactive = readl(pp->sactive_block);
@@ -1765,7 +1765,7 @@ static void nv_swncq_ncq_stop(struct ata_port *ap)
 	}
 
 	nv_swncq_pp_reinit(ap);
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_irq_clear(ap);
 	__ata_bmdma_stop(ap);
 	nv_swncq_irq_clear(ap, 0xffff);
 }
@@ -1987,8 +1987,8 @@ static unsigned int nv_swncq_issue_atacmd(struct ata_port *ap,
 	pp->dmafis_bits &= ~(1 << qc->tag);
 	pp->qc_active |= (0x1 << qc->tag);
 
-	ap->ops->tf_load(ap, &qc->tf);	 /* load tf registers */
-	ap->ops->exec_command(ap, &qc->tf);
+	ap->ops->sff_tf_load(ap, &qc->tf);	 /* load tf registers */
+	ap->ops->sff_exec_command(ap, &qc->tf);
 
 	DPRINTK("Issued tag %u\n", qc->tag);
 
@@ -2060,7 +2060,7 @@ static int nv_swncq_sdbfis(struct ata_port *ap)
 		return -EINVAL;
 	}
 
-	ap->ops->irq_clear(ap);
+	ap->ops->sff_irq_clear(ap);
 	__ata_bmdma_stop(ap);
 
 	sactive = readl(pp->sactive_block);
@@ -2182,7 +2182,7 @@ static void nv_swncq_host_interrupt(struct ata_port *ap, u16 fis)
 	u8 ata_stat;
 	int rc = 0;
 
-	ata_stat = ap->ops->check_status(ap);
+	ata_stat = ap->ops->sff_check_status(ap);
 	nv_swncq_irq_clear(ap, fis);
 	if (!fis)
 		return;
@@ -2245,7 +2245,7 @@ static void nv_swncq_host_interrupt(struct ata_port *ap, u16 fis)
 
 		if (!(fis & NV_SWNCQ_IRQ_DMASETUP) &&
 		    !(pp->ncq_flags & ncq_saw_dmas)) {
-			ata_stat = ap->ops->check_status(ap);
+			ata_stat = ap->ops->sff_check_status(ap);
 			if (ata_stat & ATA_BUSY)
 				goto irq_exit;
 
diff --git a/drivers/ata/sata_promise.c b/drivers/ata/sata_promise.c
index 9923e860eae3..5a10dc5048ad 100644
--- a/drivers/ata/sata_promise.c
+++ b/drivers/ata/sata_promise.c
@@ -162,12 +162,12 @@ static struct scsi_host_template pdc_ata_sht = {
 static const struct ata_port_operations pdc_common_ops = {
 	.inherits		= &ata_sff_port_ops,
 
-	.tf_load		= pdc_tf_load_mmio,
-	.exec_command		= pdc_exec_command_mmio,
+	.sff_tf_load		= pdc_tf_load_mmio,
+	.sff_exec_command	= pdc_exec_command_mmio,
 	.check_atapi_dma	= pdc_check_atapi_dma,
 	.qc_prep		= pdc_qc_prep,
 	.qc_issue		= pdc_qc_issue,
-	.irq_clear		= pdc_irq_clear,
+	.sff_irq_clear		= pdc_irq_clear,
 
 	.post_internal_cmd	= pdc_post_internal_cmd,
 	.error_handler		= pdc_error_handler,
diff --git a/drivers/ata/sata_sil.c b/drivers/ata/sata_sil.c
index 987313b68f3b..88bf4212590f 100644
--- a/drivers/ata/sata_sil.c
+++ b/drivers/ata/sata_sil.c
@@ -369,7 +369,7 @@ static void sil_host_intr(struct ata_port *ap, u32 bmdma2)
 
 	if (unlikely(!qc || (qc->tf.flags & ATA_TFLAG_POLLING))) {
 		/* this sometimes happens, just clear IRQ */
-		ap->ops->check_status(ap);
+		ap->ops->sff_check_status(ap);
 		return;
 	}
 
@@ -405,7 +405,7 @@ static void sil_host_intr(struct ata_port *ap, u32 bmdma2)
 	}
 
 	/* check main status, clearing INTRQ */
-	status = ap->ops->check_status(ap);
+	status = ap->ops->sff_check_status(ap);
 	if (unlikely(status & ATA_BUSY))
 		goto err_hsm;
 
@@ -480,7 +480,7 @@ static void sil_thaw(struct ata_port *ap)
 	u32 tmp;
 
 	/* clear IRQ */
-	ap->ops->check_status(ap);
+	ap->ops->sff_check_status(ap);
 	ata_sff_irq_clear(ap);
 
 	/* turn on SATA IRQ if supported */
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index 67df1d753305..b83851f6e068 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -401,9 +401,9 @@ static struct scsi_host_template sil24_sht = {
 static struct ata_port_operations sil24_ops = {
 	.inherits		= &sata_pmp_port_ops,
 
-	.check_status		= sil24_check_status,
-	.check_altstatus	= sil24_check_status,
-	.tf_read		= sil24_tf_read,
+	.sff_check_status	= sil24_check_status,
+	.sff_check_altstatus	= sil24_check_status,
+	.sff_tf_read		= sil24_tf_read,
 	.qc_defer		= sil24_qc_defer,
 	.qc_prep		= sil24_qc_prep,
 	.qc_issue		= sil24_qc_issue,
diff --git a/drivers/ata/sata_svw.c b/drivers/ata/sata_svw.c
index 7b941106f7de..16aa6839aa5a 100644
--- a/drivers/ata/sata_svw.c
+++ b/drivers/ata/sata_svw.c
@@ -233,7 +233,7 @@ static void k2_bmdma_setup_mmio(struct ata_queued_cmd *qc)
 
 	/* issue r/w command if this is not a ATA DMA command*/
 	if (qc->tf.protocol != ATA_PROT_DMA)
-		ap->ops->exec_command(ap, &qc->tf);
+		ap->ops->sff_exec_command(ap, &qc->tf);
 }
 
 /**
@@ -269,7 +269,7 @@ static void k2_bmdma_start_mmio(struct ata_queued_cmd *qc)
 	   and the start command. */
 	/* issue r/w command if the access is to ATA*/
 	if (qc->tf.protocol == ATA_PROT_DMA)
-		ap->ops->exec_command(ap, &qc->tf);
+		ap->ops->sff_exec_command(ap, &qc->tf);
 }
 
 
@@ -336,9 +336,9 @@ static struct scsi_host_template k2_sata_sht = {
 
 static struct ata_port_operations k2_sata_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.tf_load		= k2_sata_tf_load,
-	.tf_read		= k2_sata_tf_read,
-	.check_status		= k2_stat_check_status,
+	.sff_tf_load		= k2_sata_tf_load,
+	.sff_tf_read		= k2_sata_tf_read,
+	.sff_check_status	= k2_stat_check_status,
 	.check_atapi_dma	= k2_sata_check_atapi_dma,
 	.bmdma_setup		= k2_bmdma_setup_mmio,
 	.bmdma_start		= k2_bmdma_start_mmio,
diff --git a/drivers/ata/sata_sx4.c b/drivers/ata/sata_sx4.c
index 40be2ff60902..6107eff731f3 100644
--- a/drivers/ata/sata_sx4.c
+++ b/drivers/ata/sata_sx4.c
@@ -243,18 +243,18 @@ static struct scsi_host_template pdc_sata_sht = {
 
 /* TODO: inherit from base port_ops after converting to new EH */
 static struct ata_port_operations pdc_20621_ops = {
-	.tf_load		= pdc_tf_load_mmio,
-	.tf_read		= ata_sff_tf_read,
-	.check_status		= ata_sff_check_status,
-	.exec_command		= pdc_exec_command_mmio,
-	.dev_select		= ata_sff_dev_select,
+	.sff_tf_load		= pdc_tf_load_mmio,
+	.sff_tf_read		= ata_sff_tf_read,
+	.sff_check_status	= ata_sff_check_status,
+	.sff_exec_command	= pdc_exec_command_mmio,
+	.sff_dev_select		= ata_sff_dev_select,
 	.phy_reset		= pdc_20621_phy_reset,
 	.qc_prep		= pdc20621_qc_prep,
 	.qc_issue		= pdc20621_qc_issue,
-	.data_xfer		= ata_sff_data_xfer,
+	.sff_data_xfer		= ata_sff_data_xfer,
 	.eng_timeout		= pdc_eng_timeout,
-	.irq_clear		= pdc20621_irq_clear,
-	.irq_on			= ata_sff_irq_on,
+	.sff_irq_clear		= pdc20621_irq_clear,
+	.sff_irq_on		= ata_sff_irq_on,
 	.port_start		= pdc_port_start,
 };
 
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index 9323bc2a2785..e5df37689740 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -173,7 +173,7 @@ static void svia_noop_freeze(struct ata_port *ap)
 	/* Some VIA controllers choke if ATA_NIEN is manipulated in
 	 * certain way.  Leave it alone and just clear pending IRQ.
 	 */
-	ap->ops->check_status(ap);
+	ap->ops->sff_check_status(ap);
 	ata_sff_irq_clear(ap);
 }
 
diff --git a/drivers/ata/sata_vsc.c b/drivers/ata/sata_vsc.c
index 2c3c7693c750..f3d635c0a2e9 100644
--- a/drivers/ata/sata_vsc.c
+++ b/drivers/ata/sata_vsc.c
@@ -251,7 +251,7 @@ static void vsc_port_intr(u8 port_status, struct ata_port *ap)
 	 * simply clear the interrupt
 	 */
 	if (unlikely(!handled))
-		ap->ops->check_status(ap);
+		ap->ops->sff_check_status(ap);
 }
 
 /*
@@ -306,8 +306,8 @@ static struct scsi_host_template vsc_sata_sht = {
 
 static struct ata_port_operations vsc_sata_ops = {
 	.inherits		= &ata_bmdma_port_ops,
-	.tf_load		= vsc_sata_tf_load,
-	.tf_read		= vsc_sata_tf_read,
+	.sff_tf_load		= vsc_sata_tf_load,
+	.sff_tf_read		= vsc_sata_tf_read,
 	.freeze			= vsc_freeze,
 	.thaw			= vsc_thaw,
 	.scr_read		= vsc_sata_scr_read,
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index 79fd2436bd70..5bddae19d1fb 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -5271,13 +5271,13 @@ static u8 ipr_ata_check_altstatus(struct ata_port *ap)
 }
 
 static struct ata_port_operations ipr_sata_ops = {
-	.check_status = ipr_ata_check_status,
-	.check_altstatus = ipr_ata_check_altstatus,
-	.dev_select = ata_noop_dev_select,
+	.sff_check_status = ipr_ata_check_status,
+	.sff_check_altstatus = ipr_ata_check_altstatus,
+	.sff_dev_select = ata_noop_dev_select,
 	.phy_reset = ipr_ata_phy_reset,
 	.hardreset = ipr_sata_reset,
 	.post_internal_cmd = ipr_ata_post_internal,
-	.tf_read = ipr_tf_read,
+	.sff_tf_read = ipr_tf_read,
 	.qc_prep = ata_noop_qc_prep,
 	.qc_issue = ipr_qc_issue,
 	.port_start = ata_sas_port_start,
diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index b0e5ac372a32..a1664b87927a 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -348,12 +348,12 @@ static int sas_ata_scr_read(struct ata_port *ap, unsigned int sc_reg_in,
 }
 
 static struct ata_port_operations sas_sata_ops = {
-	.check_status		= sas_ata_check_status,
-	.check_altstatus	= sas_ata_check_status,
-	.dev_select		= ata_noop_dev_select,
+	.sff_check_status	= sas_ata_check_status,
+	.sff_check_altstatus	= sas_ata_check_status,
+	.sff_dev_select		= ata_noop_dev_select,
 	.phy_reset		= sas_ata_phy_reset,
 	.post_internal_cmd	= sas_ata_post_internal,
-	.tf_read		= sas_ata_tf_read,
+	.sff_tf_read		= sas_ata_tf_read,
 	.qc_prep		= ata_noop_qc_prep,
 	.qc_issue		= sas_ata_qc_issue,
 	.port_start		= ata_sas_port_start,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 66663bfe2c71..584eca19b8f1 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -743,17 +743,18 @@ struct ata_port_operations {
 	/*
 	 * SFF / taskfile oriented ops
 	 */
-	void (*dev_select)(struct ata_port *ap, unsigned int device);
-	u8   (*check_status)(struct ata_port *ap);
-	u8   (*check_altstatus)(struct ata_port *ap);
-	void (*tf_load)(struct ata_port *ap, const struct ata_taskfile *tf);
-	void (*tf_read)(struct ata_port *ap, struct ata_taskfile *tf);
-	void (*exec_command)(struct ata_port *ap, const struct ata_taskfile *tf);
-	unsigned int (*data_xfer)(struct ata_device *dev, unsigned char *buf,
-				  unsigned int buflen, int rw);
-	u8   (*irq_on)(struct ata_port *);
-
-	void (*irq_clear)(struct ata_port *);
+	void (*sff_dev_select)(struct ata_port *ap, unsigned int device);
+	u8   (*sff_check_status)(struct ata_port *ap);
+	u8   (*sff_check_altstatus)(struct ata_port *ap);
+	void (*sff_tf_load)(struct ata_port *ap, const struct ata_taskfile *tf);
+	void (*sff_tf_read)(struct ata_port *ap, struct ata_taskfile *tf);
+	void (*sff_exec_command)(struct ata_port *ap,
+				 const struct ata_taskfile *tf);
+	unsigned int (*sff_data_xfer)(struct ata_device *dev,
+			unsigned char *buf, unsigned int buflen, int rw);
+	u8   (*sff_irq_on)(struct ata_port *);
+	void (*sff_irq_clear)(struct ata_port *);
+
 	void (*bmdma_setup)(struct ata_queued_cmd *qc);
 	void (*bmdma_start)(struct ata_queued_cmd *qc);
 	void (*bmdma_stop)(struct ata_queued_cmd *qc);
@@ -1438,7 +1439,7 @@ static inline u8 ata_sff_busy_wait(struct ata_port *ap, unsigned int bits,
 
 	do {
 		udelay(10);
-		status = ap->ops->check_status(ap);
+		status = ap->ops->sff_check_status(ap);
 		max--;
 	} while (status != 0xff && (status & bits) && (max > 0));
 
-- 
cgit v1.2.3


From 288623a06c652239d2f57d271af12bb024cf7218 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:17 +0900
Subject: libata: clean up port_ops->sff_irq_clear()

->sff_irq_clear() is called only from SFF interrupt handler, so there
is no reason to initialize it for non-SFF controllers.  Also,
ata_sff_irq_clear() can handle both BMDMA and non-BMDMA SFF
controllers.

This patch kills ata_noop_irq_clear() and removes it from base
port_ops and sets ->sff_irq_clear to ata_sff_irq_clear() in sff
port_ops instead of bmdma port_ops.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c | 10 ----------
 drivers/ata/libata-sff.c  |  2 +-
 include/linux/libata.h    |  1 -
 3 files changed, 1 insertion(+), 12 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index a01e02c5ce7a..fd912ccb90f8 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -74,7 +74,6 @@ const unsigned long sata_deb_timing_hotplug[]		= {  25,  500, 2000 };
 const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
 const struct ata_port_operations ata_base_port_ops = {
-	.sff_irq_clear		= ata_noop_irq_clear,
 	.prereset		= ata_sff_prereset,
 	.hardreset		= sata_sff_hardreset,
 	.postreset		= ata_sff_postreset,
@@ -386,14 +385,6 @@ int atapi_cmd_type(u8 opcode)
 	}
 }
 
-/**
- *	ata_noop_irq_clear - Noop placeholder for irq_clear
- *	@ap: Port associated with this ATA transaction.
- */
-void ata_noop_irq_clear(struct ata_port *ap)
-{
-}
-
 /**
  *	ata_tf_to_fis - Convert ATA taskfile to SATA FIS structure
  *	@tf: Taskfile to convert
@@ -6108,7 +6099,6 @@ EXPORT_SYMBOL_GPL(ata_port_start);
 EXPORT_SYMBOL_GPL(ata_do_set_mode);
 EXPORT_SYMBOL_GPL(ata_std_qc_defer);
 EXPORT_SYMBOL_GPL(ata_noop_qc_prep);
-EXPORT_SYMBOL_GPL(ata_noop_irq_clear);
 EXPORT_SYMBOL_GPL(ata_port_probe);
 EXPORT_SYMBOL_GPL(ata_dev_disable);
 EXPORT_SYMBOL_GPL(sata_set_spd);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 04024a556660..c11601617134 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -58,6 +58,7 @@ const struct ata_port_operations ata_sff_port_ops = {
 	.sff_exec_command	= ata_sff_exec_command,
 	.sff_data_xfer		= ata_sff_data_xfer,
 	.sff_irq_on		= ata_sff_irq_on,
+	.sff_irq_clear		= ata_sff_irq_clear,
 
 	.port_start		= ata_sff_port_start,
 };
@@ -71,7 +72,6 @@ const struct ata_port_operations ata_bmdma_port_ops = {
 	.bmdma_start		= ata_bmdma_start,
 	.bmdma_stop		= ata_bmdma_stop,
 	.bmdma_status		= ata_bmdma_status,
-	.sff_irq_clear		= ata_sff_irq_clear,
 };
 
 /**
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 584eca19b8f1..603712b59cf3 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -896,7 +896,6 @@ extern void ata_id_string(const u16 *id, unsigned char *s,
 			  unsigned int ofs, unsigned int len);
 extern void ata_id_c_string(const u16 *id, unsigned char *s,
 			    unsigned int ofs, unsigned int len);
-extern void ata_noop_irq_clear(struct ata_port *ap);
 extern void ata_qc_complete(struct ata_queued_cmd *qc);
 extern int ata_qc_complete_multiple(struct ata_port *ap, u32 qc_active,
 				    void (*finish_qc)(struct ata_queued_cmd *));
-- 
cgit v1.2.3


From 0aa1113d544226bc2c4a20d6ac1d71170512a361 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:18 +0900
Subject: libata: separate out ata_std_prereset() from ata_sff_prereset()

Separate out generic ATA portion from ata_sff_prereset() into
ata_std_prereset() and implement ata_sff_prereset() using the std
version.  Waiting for device readiness is the only SFF specific part.

ata_base_port_ops now has ata_std_prereset() for its prereset and
ata_sff_port_ops overrides it to ata_sff_prereset().  This change can
affect pdc_adma, ahci, sata_fsl and sata_sil24.  pdc_adma implements
its own prereset using ata_sff_prereset() and the rest has hardreset
and thus are unaffected by this change.

This change reflects real world situation.  There is no generic way to
wait for device readiness for non-SFF controllers and some of them
don't have any mechanism for that.  Non-sff drivers which don't have
hardreset should wrap ata_std_prereset() and wait for device readiness
itself but there's no such driver now and isn't likely to be popular
in the future either.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c | 17 ++++-------------
 drivers/ata/libata-sff.c  | 43 +++++++++++++++++++++++++++++++++++++++++++
 include/linux/libata.h    |  1 +
 3 files changed, 48 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index fd912ccb90f8..5f771bb44332 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -74,7 +74,7 @@ const unsigned long sata_deb_timing_hotplug[]		= {  25,  500, 2000 };
 const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
 const struct ata_port_operations ata_base_port_ops = {
-	.prereset		= ata_sff_prereset,
+	.prereset		= ata_std_prereset,
 	.hardreset		= sata_sff_hardreset,
 	.postreset		= ata_sff_postreset,
 	.error_handler		= ata_std_error_handler,
@@ -3416,7 +3416,7 @@ int sata_link_resume(struct ata_link *link, const unsigned long *params,
 }
 
 /**
- *	ata_sff_prereset - prepare for reset
+ *	ata_std_prereset - prepare for reset
  *	@link: ATA link to be reset
  *	@deadline: deadline jiffies for the operation
  *
@@ -3432,7 +3432,7 @@ int sata_link_resume(struct ata_link *link, const unsigned long *params,
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_sff_prereset(struct ata_link *link, unsigned long deadline)
+int ata_std_prereset(struct ata_link *link, unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
 	struct ata_eh_context *ehc = &link->eh_context;
@@ -3452,16 +3452,6 @@ int ata_sff_prereset(struct ata_link *link, unsigned long deadline)
 					"link for reset (errno=%d)\n", rc);
 	}
 
-	/* wait for !BSY if we don't know that no device is attached */
-	if (!ata_link_offline(link)) {
-		rc = ata_sff_wait_ready(ap, deadline);
-		if (rc && rc != -ENODEV) {
-			ata_link_printk(link, KERN_WARNING, "device not ready "
-					"(errno=%d), forcing hardreset\n", rc);
-			ehc->i.action |= ATA_EH_HARDRESET;
-		}
-	}
-
 	return 0;
 }
 
@@ -6104,6 +6094,7 @@ EXPORT_SYMBOL_GPL(ata_dev_disable);
 EXPORT_SYMBOL_GPL(sata_set_spd);
 EXPORT_SYMBOL_GPL(sata_link_debounce);
 EXPORT_SYMBOL_GPL(sata_link_resume);
+EXPORT_SYMBOL_GPL(ata_std_prereset);
 EXPORT_SYMBOL_GPL(sata_link_hardreset);
 EXPORT_SYMBOL_GPL(ata_dev_classify);
 EXPORT_SYMBOL_GPL(ata_dev_pair);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index c11601617134..9234bc047956 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -47,6 +47,7 @@ const struct ata_port_operations ata_sff_port_ops = {
 
 	.freeze			= ata_sff_freeze,
 	.thaw			= ata_sff_thaw,
+	.prereset		= ata_sff_prereset,
 	.softreset		= ata_sff_softreset,
 	.error_handler		= ata_sff_error_handler,
 	.post_internal_cmd	= ata_sff_post_internal_cmd,
@@ -1606,6 +1607,48 @@ void ata_sff_thaw(struct ata_port *ap)
 	ap->ops->sff_irq_on(ap);
 }
 
+/**
+ *	ata_sff_prereset - prepare SFF link for reset
+ *	@link: SFF link to be reset
+ *	@deadline: deadline jiffies for the operation
+ *
+ *	SFF link @link is about to be reset.  Initialize it.  It first
+ *	calls ata_std_prereset() and wait for !BSY if the port is
+ *	being softreset.
+ *
+ *	LOCKING:
+ *	Kernel thread context (may sleep)
+ *
+ *	RETURNS:
+ *	0 on success, -errno otherwise.
+ */
+int ata_sff_prereset(struct ata_link *link, unsigned long deadline)
+{
+	struct ata_port *ap = link->ap;
+	struct ata_eh_context *ehc = &link->eh_context;
+	int rc;
+
+	rc = ata_std_prereset(link, deadline);
+	if (rc)
+		return rc;
+
+	/* if we're about to do hardreset, nothing more to do */
+	if (ehc->i.action & ATA_EH_HARDRESET)
+		return 0;
+
+	/* wait for !BSY if we don't know that no device is attached */
+	if (!ata_link_offline(link)) {
+		rc = ata_sff_wait_ready(ap, deadline);
+		if (rc && rc != -ENODEV) {
+			ata_link_printk(link, KERN_WARNING, "device not ready "
+					"(errno=%d), forcing hardreset\n", rc);
+			ehc->i.action |= ATA_EH_HARDRESET;
+		}
+	}
+
+	return 0;
+}
+
 /**
  *	ata_devchk - PATA device presence detection
  *	@ap: ATA channel to examine
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 603712b59cf3..595ede55fe4c 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -824,6 +824,7 @@ static inline int ata_port_is_dummy(struct ata_port *ap)
 extern void sata_print_link_status(struct ata_link *link);
 extern void ata_port_probe(struct ata_port *);
 extern int sata_set_spd(struct ata_link *link);
+extern int ata_std_prereset(struct ata_link *link, unsigned long deadline);
 extern int sata_link_debounce(struct ata_link *link,
 			const unsigned long *params, unsigned long deadline);
 extern int sata_link_resume(struct ata_link *link, const unsigned long *params,
-- 
cgit v1.2.3


From 203c75b8245c5386044721d9c5eda5c6b71b3d14 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:18 +0900
Subject: libata: separate out ata_std_postreset() from ata_sff_postreset()

Separate out generic ATA portion from ata_sff_postreset() into
ata_std_postreset() and implement ata_sff_postreset() using the std
version.

ata_base_port_ops now has ata_std_postreset() for its postreset and
ata_sff_port_ops overrides it to ata_sff_postreset().

This change affects pdc_adma, ahci, sata_fsl and sata_sil24.  pdc_adma
now specifies postreset to ata_sff_postreset() explicitly.  sata_fsl
and sata_sil24 now use ata_std_postreset() which makes no difference
to them.  ahci now calls ata_std_postreset() from its own postreset
method, which causes no behavior difference.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c        |  2 +-
 drivers/ata/libata-core.c | 24 ++++--------------------
 drivers/ata/libata-sff.c  | 36 ++++++++++++++++++++++++++++++++++++
 include/linux/libata.h    |  1 +
 4 files changed, 42 insertions(+), 21 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index c5e4501daa74..939dc1d4e50d 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1444,7 +1444,7 @@ static void ahci_postreset(struct ata_link *link, unsigned int *class)
 	void __iomem *port_mmio = ahci_port_base(ap);
 	u32 new_tmp, tmp;
 
-	ata_sff_postreset(link, class);
+	ata_std_postreset(link, class);
 
 	/* Make sure port's ATAPI bit is set appropriately */
 	new_tmp = tmp = readl(port_mmio + PORT_CMD);
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 5f771bb44332..eaead76c9443 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -76,7 +76,7 @@ const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 const struct ata_port_operations ata_base_port_ops = {
 	.prereset		= ata_std_prereset,
 	.hardreset		= sata_sff_hardreset,
-	.postreset		= ata_sff_postreset,
+	.postreset		= ata_std_postreset,
 	.error_handler		= ata_std_error_handler,
 };
 
@@ -3516,7 +3516,7 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
 }
 
 /**
- *	ata_sff_postreset - standard postreset callback
+ *	ata_std_postreset - standard postreset callback
  *	@link: the target ata_link
  *	@classes: classes of attached devices
  *
@@ -3527,9 +3527,8 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
  *	LOCKING:
  *	Kernel thread context (may sleep)
  */
-void ata_sff_postreset(struct ata_link *link, unsigned int *classes)
+void ata_std_postreset(struct ata_link *link, unsigned int *classes)
 {
-	struct ata_port *ap = link->ap;
 	u32 serror;
 
 	DPRINTK("ENTER\n");
@@ -3542,22 +3541,6 @@ void ata_sff_postreset(struct ata_link *link, unsigned int *classes)
 		sata_scr_write(link, SCR_ERROR, serror);
 	link->eh_info.serror = 0;
 
-	/* is double-select really necessary? */
-	if (classes[0] != ATA_DEV_NONE)
-		ap->ops->sff_dev_select(ap, 1);
-	if (classes[1] != ATA_DEV_NONE)
-		ap->ops->sff_dev_select(ap, 0);
-
-	/* bail out if no device is present */
-	if (classes[0] == ATA_DEV_NONE && classes[1] == ATA_DEV_NONE) {
-		DPRINTK("EXIT, no device\n");
-		return;
-	}
-
-	/* set up device control */
-	if (ap->ioaddr.ctl_addr)
-		iowrite8(ap->ctl, ap->ioaddr.ctl_addr);
-
 	DPRINTK("EXIT\n");
 }
 
@@ -6096,6 +6079,7 @@ EXPORT_SYMBOL_GPL(sata_link_debounce);
 EXPORT_SYMBOL_GPL(sata_link_resume);
 EXPORT_SYMBOL_GPL(ata_std_prereset);
 EXPORT_SYMBOL_GPL(sata_link_hardreset);
+EXPORT_SYMBOL_GPL(ata_std_postreset);
 EXPORT_SYMBOL_GPL(ata_dev_classify);
 EXPORT_SYMBOL_GPL(ata_dev_pair);
 EXPORT_SYMBOL_GPL(ata_port_disable);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 9234bc047956..e530baccc9cb 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -49,6 +49,7 @@ const struct ata_port_operations ata_sff_port_ops = {
 	.thaw			= ata_sff_thaw,
 	.prereset		= ata_sff_prereset,
 	.softreset		= ata_sff_softreset,
+	.postreset		= ata_sff_postreset,
 	.error_handler		= ata_sff_error_handler,
 	.post_internal_cmd	= ata_sff_post_internal_cmd,
 
@@ -2031,6 +2032,41 @@ int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
 	return 0;
 }
 
+/**
+ *	ata_sff_postreset - SFF postreset callback
+ *	@link: the target SFF ata_link
+ *	@classes: classes of attached devices
+ *
+ *	This function is invoked after a successful reset.  It first
+ *	calls ata_std_postreset() and performs SFF specific postreset
+ *	processing.
+ *
+ *	LOCKING:
+ *	Kernel thread context (may sleep)
+ */
+void ata_sff_postreset(struct ata_link *link, unsigned int *classes)
+{
+	struct ata_port *ap = link->ap;
+
+	ata_std_postreset(link, classes);
+
+	/* is double-select really necessary? */
+	if (classes[0] != ATA_DEV_NONE)
+		ap->ops->sff_dev_select(ap, 1);
+	if (classes[1] != ATA_DEV_NONE)
+		ap->ops->sff_dev_select(ap, 0);
+
+	/* bail out if no device is present */
+	if (classes[0] == ATA_DEV_NONE && classes[1] == ATA_DEV_NONE) {
+		DPRINTK("EXIT, no device\n");
+		return;
+	}
+
+	/* set up device control */
+	if (ap->ioaddr.ctl_addr)
+		iowrite8(ap->ctl, ap->ioaddr.ctl_addr);
+}
+
 /**
  *	ata_sff_error_handler - Stock error handler for BMDMA controller
  *	@ap: port to handle error for
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 595ede55fe4c..6e14c27319d5 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -831,6 +831,7 @@ extern int sata_link_resume(struct ata_link *link, const unsigned long *params,
 			    unsigned long deadline);
 extern int sata_link_hardreset(struct ata_link *link,
 			const unsigned long *timing, unsigned long deadline);
+extern void ata_std_postreset(struct ata_link *link, unsigned int *classes);
 extern void ata_port_disable(struct ata_port *);
 
 extern struct ata_host *ata_host_alloc(struct device *dev, int max_ports);
-- 
cgit v1.2.3


From 705e76beb90b97421e1f61e857c4246799781bb5 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:19 +0900
Subject: libata: restructure SFF post-reset readiness waits

Previously, post-softreset readiness is waited as follows.

1. ata_sff_wait_after_reset() waits for 150ms and then for
   ATA_TMOUT_FF_WAIT if status is 0xff and other conditions meet.

2. ata_bus_softreset() finishes with -ENODEV if status is still 0xff.
   If not, continue to #3.

3. ata_bus_post_reset() waits readiness of dev0 and/or dev1 depending
   on devmask using ata_sff_wait_ready().

And for post-hardreset readiness,

1. ata_sff_wait_after_reset() waits for 150ms and then for
   ATA_TMOUT_FF_WAIT if status is 0xff and other conditions meet.

2. sata_sff_hardreset waits for device readiness using
   ata_sff_wait_ready().

This patch merges and unifies post-reset readiness waits into
ata_sff_wait_ready() and ata_sff_wait_after_reset().

ATA_TMOUT_FF_WAIT handling is merged into ata_sff_wait_ready().  If TF
status is 0xff, link status is unknown and the port is SATA, it will
continue polling till ATA_TMOUT_FF_WAIT.

ata_sff_wait_after_reset() is updated to perform the following steps.

1. waits for 150ms.

2. waits for dev0 readiness using ata_sff_wait_ready().  Note that
   this is done regardless of devmask, as ata_sff_wait_ready() handles
   0xff status correctly, this preserves the original behavior except
   that it may wait longer after softreset if link is online but
   status is 0xff.  This behavior change is very unlikely to cause any
   actual difference and is intended.  It brings softreset behavior to
   that of hardreset.

3. waits for dev1 readiness just the same way ata_bus_post_reset() did.

Now both soft and hard resets call ata_sff_wait_after_reset() after
reset to wait for readiness after resets.  As
ata_sff_wait_after_reset() contains calls to ->sff_dev_select(),
explicit call near the end of sata_sff_hardreset() is removed.

This change makes reset implementation simpler and more consistent.

While at it, make the magical 150ms wait post-reset wait duration a
constant and ata_sff_wait_ready() and ata_sff_wait_after_reset() take
@link instead of @ap.  This is to make them consistent with other
reset helpers and ease core changes.

pata_scc is updated accordingly.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c          |  11 +--
 drivers/ata/libata-sff.c    | 161 ++++++++++++++++++--------------------------
 drivers/ata/pata_scc.c      |  91 ++++++++++++++-----------
 drivers/ata/sata_inic162x.c |   6 +-
 drivers/ata/sata_via.c      |   2 +-
 include/linux/libata.h      |  18 ++++-
 6 files changed, 139 insertions(+), 150 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 939dc1d4e50d..45a67a9ad8ab 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1302,10 +1302,8 @@ static int ahci_do_softreset(struct ata_link *link, unsigned int *class,
 	tf.ctl &= ~ATA_SRST;
 	ahci_exec_polled_cmd(ap, pmp, &tf, 0, 0, 0);
 
-	/* wait a while before checking status */
-	ata_sff_wait_after_reset(ap, deadline);
-
-	rc = ata_sff_wait_ready(ap, deadline);
+	/* wait for link to become ready */
+	rc = ata_sff_wait_after_reset(link, 1, deadline);
 	/* link occupied, -ENODEV too is an error */
 	if (rc) {
 		reason = "device not ready";
@@ -1415,9 +1413,6 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 	if (rc || ata_link_offline(link))
 		return rc;
 
-	/* spec mandates ">= 2ms" before checking status */
-	msleep(150);
-
 	/* The pseudo configuration device on SIMG4726 attached to
 	 * ASUS P5W-DH Deluxe doesn't send signature FIS after
 	 * hardreset if no device is attached to the first downstream
@@ -1431,7 +1426,7 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 	 * have to be reset again.  For most cases, this should
 	 * suffice while making probing snappish enough.
 	 */
-	rc = ata_sff_wait_ready(ap, jiffies + 2 * HZ);
+	rc = ata_sff_wait_after_reset(link, 1, jiffies + 2 * HZ);
 	if (rc)
 		ahci_kick_engine(ap, 0);
 
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index e530baccc9cb..6e8de3c1595e 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -310,7 +310,7 @@ int ata_sff_busy_sleep(struct ata_port *ap,
 
 /**
  *	ata_sff_wait_ready - sleep until BSY clears, or timeout
- *	@ap: port containing status register to be polled
+ *	@link: SFF link to wait ready status for
  *	@deadline: deadline jiffies for the operation
  *
  *	Sleep until ATA Status register bit BSY clears, or timeout
@@ -322,26 +322,52 @@ int ata_sff_busy_sleep(struct ata_port *ap,
  *	RETURNS:
  *	0 on success, -errno otherwise.
  */
-int ata_sff_wait_ready(struct ata_port *ap, unsigned long deadline)
+int ata_sff_wait_ready(struct ata_link *link, unsigned long deadline)
 {
+	struct ata_port *ap = link->ap;
 	unsigned long start = jiffies;
+	unsigned long nodev_deadline = start + ATA_TMOUT_FF_WAIT;
 	int warned = 0;
 
+	if (time_after(nodev_deadline, deadline))
+		nodev_deadline = deadline;
+
 	while (1) {
 		u8 status = ap->ops->sff_check_status(ap);
 		unsigned long now = jiffies;
 
 		if (!(status & ATA_BUSY))
 			return 0;
-		if (!ata_link_online(&ap->link) && status == 0xff)
-			return -ENODEV;
+
+		/* No device status could be transient.  Ignore it if
+		 * link is online.  Also, some SATA devices take a
+		 * long time to clear 0xff after reset.  For example,
+		 * HHD424020F7SV00 iVDR needs >= 800ms while Quantum
+		 * GoVault needs even more than that.  Wait for
+		 * ATA_TMOUT_FF_WAIT on -ENODEV if link isn't offline.
+		 *
+		 * Note that some PATA controllers (pata_ali) explode
+		 * if status register is read more than once when
+		 * there's no device attached.
+		 */
+		if (status == 0xff) {
+			if (ata_link_online(link))
+				status = ATA_BUSY;
+			else if ((link->ap->flags & ATA_FLAG_SATA) &&
+				 !ata_link_offline(link) &&
+				 time_before(now, nodev_deadline))
+				status = ATA_BUSY;
+			if (status == 0xff)
+				return -ENODEV;
+		}
+
 		if (time_after(now, deadline))
 			return -EBUSY;
 
 		if (!warned && time_after(now, start + 5 * HZ) &&
 		    (deadline - now > 3 * HZ)) {
-			ata_port_printk(ap, KERN_WARNING,
-				"port is slow to respond, please be patient "
+			ata_link_printk(link, KERN_WARNING,
+				"link is slow to respond, please be patient "
 				"(Status 0x%x)\n", status);
 			warned = 1;
 		}
@@ -1625,7 +1651,6 @@ void ata_sff_thaw(struct ata_port *ap)
  */
 int ata_sff_prereset(struct ata_link *link, unsigned long deadline)
 {
-	struct ata_port *ap = link->ap;
 	struct ata_eh_context *ehc = &link->eh_context;
 	int rc;
 
@@ -1639,7 +1664,7 @@ int ata_sff_prereset(struct ata_link *link, unsigned long deadline)
 
 	/* wait for !BSY if we don't know that no device is attached */
 	if (!ata_link_offline(link)) {
-		rc = ata_sff_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(link, deadline);
 		if (rc && rc != -ENODEV) {
 			ata_link_printk(link, KERN_WARNING, "device not ready "
 					"(errno=%d), forcing hardreset\n", rc);
@@ -1762,25 +1787,41 @@ unsigned int ata_sff_dev_classify(struct ata_device *dev, int present,
 	return class;
 }
 
-static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
-			      unsigned long deadline)
+/**
+ *	ata_sff_wait_after_reset - wait for devices to become ready after reset
+ *	@link: SFF link which is just reset
+ *	@devmask: mask of present devices
+ *	@deadline: deadline jiffies for the operation
+ *
+ *	Wait devices attached to SFF @link to become ready after
+ *	reset.  It contains preceding 150ms wait to avoid accessing TF
+ *	status register too early.
+ *
+ *	LOCKING:
+ *	Kernel thread context (may sleep).
+ *
+ *	RETURNS:
+ *	0 on success, -ENODEV if some or all of devices in @devmask
+ *	don't seem to exist.  -errno on other errors.
+ */
+int ata_sff_wait_after_reset(struct ata_link *link, unsigned int devmask,
+			     unsigned long deadline)
 {
+	struct ata_port *ap = link->ap;
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 	unsigned int dev0 = devmask & (1 << 0);
 	unsigned int dev1 = devmask & (1 << 1);
 	int rc, ret = 0;
 
-	/* if device 0 was found in ata_devchk, wait for its
-	 * BSY bit to clear
+	msleep(ATA_WAIT_AFTER_RESET_MSECS);
+
+	/* always check readiness of the master device */
+	rc = ata_sff_wait_ready(link, deadline);
+	/* -ENODEV means the odd clown forgot the D7 pulldown resistor
+	 * and TF status is 0xff, bail out on it too.
 	 */
-	if (dev0) {
-		rc = ata_sff_wait_ready(ap, deadline);
-		if (rc) {
-			if (rc != -ENODEV)
-				return rc;
-			ret = rc;
-		}
-	}
+	if (rc)
+		return rc;
 
 	/* if device 1 was found in ata_devchk, wait for register
 	 * access briefly, then wait for BSY to clear.
@@ -1804,7 +1845,7 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 			msleep(50);	/* give drive a breather */
 		}
 
-		rc = ata_sff_wait_ready(ap, deadline);
+		rc = ata_sff_wait_ready(link, deadline);
 		if (rc) {
 			if (rc != -ENODEV)
 				return rc;
@@ -1822,61 +1863,6 @@ static int ata_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	return ret;
 }
 
-/**
- *	ata_sff_wait_after_reset - wait before checking status after reset
- *	@ap: port containing status register to be polled
- *	@deadline: deadline jiffies for the operation
- *
- *	After reset, we need to pause a while before reading status.
- *	Also, certain combination of controller and device report 0xff
- *	for some duration (e.g. until SATA PHY is up and running)
- *	which is interpreted as empty port in ATA world.  This
- *	function also waits for such devices to get out of 0xff
- *	status.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep).
- */
-void ata_sff_wait_after_reset(struct ata_port *ap, unsigned long deadline)
-{
-	unsigned long until = jiffies + ATA_TMOUT_FF_WAIT;
-
-	if (time_before(until, deadline))
-		deadline = until;
-
-	/* Spec mandates ">= 2ms" before checking status.  We wait
-	 * 150ms, because that was the magic delay used for ATAPI
-	 * devices in Hale Landis's ATADRVR, for the period of time
-	 * between when the ATA command register is written, and then
-	 * status is checked.  Because waiting for "a while" before
-	 * checking status is fine, post SRST, we perform this magic
-	 * delay here as well.
-	 *
-	 * Old drivers/ide uses the 2mS rule and then waits for ready.
-	 */
-	msleep(150);
-
-	/* Wait for 0xff to clear.  Some SATA devices take a long time
-	 * to clear 0xff after reset.  For example, HHD424020F7SV00
-	 * iVDR needs >= 800ms while.  Quantum GoVault needs even more
-	 * than that.
-	 *
-	 * Note that some PATA controllers (pata_ali) explode if
-	 * status register is read more than once when there's no
-	 * device attached.
-	 */
-	if (ap->flags & ATA_FLAG_SATA) {
-		while (1) {
-			u8 status = ap->ops->sff_check_status(ap);
-
-			if (status != 0xff || time_after(jiffies, deadline))
-				return;
-
-			msleep(50);
-		}
-	}
-}
-
 static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
 			     unsigned long deadline)
 {
@@ -1891,17 +1877,8 @@ static int ata_bus_softreset(struct ata_port *ap, unsigned int devmask,
 	udelay(20);	/* FIXME: flush */
 	iowrite8(ap->ctl, ioaddr->ctl_addr);
 
-	/* wait a while before checking status */
-	ata_sff_wait_after_reset(ap, deadline);
-
-	/* Before we perform post reset processing we want to see if
-	 * the bus shows 0xFF because the odd clown forgets the D7
-	 * pulldown resistor.
-	 */
-	if (ap->ops->sff_check_status(ap) == 0xFF)
-		return -ENODEV;
-
-	return ata_bus_post_reset(ap, devmask, deadline);
+	/* wait the port to become ready */
+	return ata_sff_wait_after_reset(&ap->link, devmask, deadline);
 }
 
 /**
@@ -2003,20 +1980,18 @@ int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
 		return 0;
 	}
 
-	/* wait a while before checking status */
-	ata_sff_wait_after_reset(ap, deadline);
-
 	/* If PMP is supported, we have to do follow-up SRST.  Note
 	 * that some PMPs don't send D2H Reg FIS after hardreset at
 	 * all if the first port is empty.  Wait for it just for a
 	 * second and request follow-up SRST.
 	 */
 	if (ap->flags & ATA_FLAG_PMP) {
-		ata_sff_wait_ready(ap, jiffies + HZ);
+		ata_sff_wait_after_reset(link, 1, jiffies + HZ);
 		return -EAGAIN;
 	}
 
-	rc = ata_sff_wait_ready(ap, deadline);
+	/* wait for the link to become online */
+	rc = ata_sff_wait_after_reset(link, 1, deadline);
 	/* link occupied, -ENODEV too is an error */
 	if (rc) {
 		ata_link_printk(link, KERN_ERR,
@@ -2024,8 +1999,6 @@ int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
 		return rc;
 	}
 
-	ap->ops->sff_dev_select(ap, 0);	/* probably unnecessary */
-
 	*class = ata_sff_dev_classify(link->device, 1, NULL);
 
 	DPRINTK("EXIT, class=%u\n", *class);
diff --git a/drivers/ata/pata_scc.c b/drivers/ata/pata_scc.c
index 2b9da715c704..accc275e74cc 100644
--- a/drivers/ata/pata_scc.c
+++ b/drivers/ata/pata_scc.c
@@ -497,47 +497,68 @@ static unsigned int scc_devchk (struct ata_port *ap,
 }
 
 /**
- *	scc_bus_post_reset - PATA device post reset
+ *	scc_wait_after_reset - wait for devices to become ready after reset
  *
- *	Note: Original code is ata_bus_post_reset().
+ *	Note: Original code is ata_sff_wait_after_reset
  */
 
-static int scc_bus_post_reset(struct ata_port *ap, unsigned int devmask,
-                              unsigned long deadline)
+int scc_wait_after_reset(struct ata_link *link, unsigned int devmask,
+			 unsigned long deadline)
 {
+	struct ata_port *ap = link->ap;
 	struct ata_ioports *ioaddr = &ap->ioaddr;
 	unsigned int dev0 = devmask & (1 << 0);
 	unsigned int dev1 = devmask & (1 << 1);
-	int rc;
+	int rc, ret = 0;
+
+	/* Spec mandates ">= 2ms" before checking status.  We wait
+	 * 150ms, because that was the magic delay used for ATAPI
+	 * devices in Hale Landis's ATADRVR, for the period of time
+	 * between when the ATA command register is written, and then
+	 * status is checked.  Because waiting for "a while" before
+	 * checking status is fine, post SRST, we perform this magic
+	 * delay here as well.
+	 *
+	 * Old drivers/ide uses the 2mS rule and then waits for ready.
+	 */
+	msleep(150);
 
-	/* if device 0 was found in ata_devchk, wait for its
-	 * BSY bit to clear
+	/* always check readiness of the master device */
+	rc = ata_sff_wait_ready(link, deadline);
+	/* -ENODEV means the odd clown forgot the D7 pulldown resistor
+	 * and TF status is 0xff, bail out on it too.
 	 */
-	if (dev0) {
-		rc = ata_sff_wait_ready(ap, deadline);
-		if (rc && rc != -ENODEV)
-			return rc;
-	}
+	if (rc)
+		return rc;
 
-	/* if device 1 was found in ata_devchk, wait for
-	 * register access, then wait for BSY to clear
+	/* if device 1 was found in ata_devchk, wait for register
+	 * access briefly, then wait for BSY to clear.
 	 */
-	while (dev1) {
-		u8 nsect, lbal;
+	if (dev1) {
+		int i;
 
 		ap->ops->sff_dev_select(ap, 1);
-		nsect = in_be32(ioaddr->nsect_addr);
-		lbal = in_be32(ioaddr->lbal_addr);
-		if ((nsect == 1) && (lbal == 1))
-			break;
-		if (time_after(jiffies, deadline))
-			return -EBUSY;
-		msleep(50);	/* give drive a breather */
-	}
-	if (dev1) {
-		rc = ata_sff_wait_ready(ap, deadline);
-		if (rc && rc != -ENODEV)
-			return rc;
+
+		/* Wait for register access.  Some ATAPI devices fail
+		 * to set nsect/lbal after reset, so don't waste too
+		 * much time on it.  We're gonna wait for !BSY anyway.
+		 */
+		for (i = 0; i < 2; i++) {
+			u8 nsect, lbal;
+
+			nsect = in_be32(ioaddr->nsect_addr);
+			lbal = in_be32(ioaddr->lbal_addr);
+			if ((nsect == 1) && (lbal == 1))
+				break;
+			msleep(50);	/* give drive a breather */
+		}
+
+		rc = ata_sff_wait_ready(link, deadline);
+		if (rc) {
+			if (rc != -ENODEV)
+				return rc;
+			ret = rc;
+		}
 	}
 
 	/* is all this really necessary? */
@@ -547,7 +568,7 @@ static int scc_bus_post_reset(struct ata_port *ap, unsigned int devmask,
 	if (dev0)
 		ap->ops->sff_dev_select(ap, 0);
 
-	return 0;
+	return ret;
 }
 
 /**
@@ -570,17 +591,7 @@ static unsigned int scc_bus_softreset(struct ata_port *ap, unsigned int devmask,
 	udelay(20);
 	out_be32(ioaddr->ctl_addr, ap->ctl);
 
-	/* wait a while before checking status */
-	ata_sff_wait_after_reset(ap, deadline);
-
-	/* Before we perform post reset processing we want to see if
-	 * the bus shows 0xFF because the odd clown forgets the D7
-	 * pulldown resistor.
-	 */
-	if (scc_check_status(ap) == 0xFF)
-		return 0;
-
-	scc_bus_post_reset(ap, devmask, deadline);
+	scc_wait_after_reset(&ap->link, devmask, deadlien);
 
 	return 0;
 }
diff --git a/drivers/ata/sata_inic162x.c b/drivers/ata/sata_inic162x.c
index 0b5a736a45e3..9f47d0022453 100644
--- a/drivers/ata/sata_inic162x.c
+++ b/drivers/ata/sata_inic162x.c
@@ -417,10 +417,8 @@ static int inic_hardreset(struct ata_link *link, unsigned int *class,
 	if (ata_link_online(link)) {
 		struct ata_taskfile tf;
 
-		/* wait a while before checking status */
-		ata_sff_wait_after_reset(ap, deadline);
-
-		rc = ata_sff_wait_ready(ap, deadline);
+		/* wait for link to become ready */
+		rc = ata_sff_wait_after_reset(link, 1, deadline);
 		/* link occupied, -ENODEV too is an error */
 		if (rc) {
 			ata_link_printk(link, KERN_WARNING, "device not ready "
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index e5df37689740..96deeb354e16 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -242,7 +242,7 @@ static int vt6420_prereset(struct ata_link *link, unsigned long deadline)
 
  skip_scr:
 	/* wait for !BSY */
-	ata_sff_wait_ready(ap, deadline);
+	ata_sff_wait_ready(link, deadline);
 
 	return 0;
 }
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 6e14c27319d5..da5560244787 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -249,6 +249,18 @@ enum {
 	 */
 	ATA_TMOUT_FF_WAIT	= 4 * HZ / 5,
 
+	/* Spec mandates to wait for ">= 2ms" before checking status
+	 * after reset.  We wait 150ms, because that was the magic
+	 * delay used for ATAPI devices in Hale Landis's ATADRVR, for
+	 * the period of time between when the ATA command register is
+	 * written, and then status is checked.  Because waiting for
+	 * "a while" before checking status is fine, post SRST, we
+	 * perform this magic delay here as well.
+	 *
+	 * Old drivers/ide uses the 2mS rule and then waits for ready.
+	 */
+	ATA_WAIT_AFTER_RESET_MSECS = 150,
+
 	/* ATA bus states */
 	BUS_UNKNOWN		= 0,
 	BUS_DMA			= 1,
@@ -1351,7 +1363,7 @@ extern u8 ata_sff_check_status(struct ata_port *ap);
 extern u8 ata_sff_altstatus(struct ata_port *ap);
 extern int ata_sff_busy_sleep(struct ata_port *ap,
 			      unsigned long timeout_pat, unsigned long timeout);
-extern int ata_sff_wait_ready(struct ata_port *ap, unsigned long deadline);
+extern int ata_sff_wait_ready(struct ata_link *link, unsigned long deadline);
 extern void ata_sff_tf_load(struct ata_port *ap, const struct ata_taskfile *tf);
 extern void ata_sff_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
 extern void ata_sff_exec_command(struct ata_port *ap,
@@ -1373,8 +1385,8 @@ extern void ata_sff_thaw(struct ata_port *ap);
 extern int ata_sff_prereset(struct ata_link *link, unsigned long deadline);
 extern unsigned int ata_sff_dev_classify(struct ata_device *dev, int present,
 					  u8 *r_err);
-extern void ata_sff_wait_after_reset(struct ata_port *ap,
-				     unsigned long deadline);
+extern int ata_sff_wait_after_reset(struct ata_link *link, unsigned int devmask,
+				    unsigned long deadline);
 extern int ata_sff_softreset(struct ata_link *link, unsigned int *classes,
 			     unsigned long deadline);
 extern int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
-- 
cgit v1.2.3


From aa2731ad9ad80ac3fca48bd1c4cf0eceede4810e Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:19 +0900
Subject: libata: separate out ata_wait_ready() and implement
 ata_wait_after_reset()

Factor out waiting logic (which is common to all ATA controllers) from
ata_sff_wait_ready() into ata_wait_ready().  ata_wait_ready() takes
@check_ready function pointer and uses it to poll for readiness.  This
allows non-SFF controllers to use ata_wait_ready() to wait for link
readiness.

This patch also implements ata_wait_after_reset() - generic version of
ata_sff_wait_after_reset() - using ata_wait_ready().

ata_sff_wait_ready() is reimplemented using ata_wait_ready() and
ata_sff_check_ready().  Functionality remains the same.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c | 98 +++++++++++++++++++++++++++++++++++++++++++++++
 drivers/ata/libata-sff.c  | 62 ++++++------------------------
 drivers/ata/libata.h      |  2 +
 include/linux/libata.h    |  2 +
 4 files changed, 114 insertions(+), 50 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index eaead76c9443..3bad6f189190 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -3310,6 +3310,103 @@ int ata_do_set_mode(struct ata_link *link, struct ata_device **r_failed_dev)
 	return rc;
 }
 
+/**
+ *	ata_wait_ready - wait for link to become ready
+ *	@link: link to be waited on
+ *	@deadline: deadline jiffies for the operation
+ *	@check_ready: callback to check link readiness
+ *
+ *	Wait for @link to become ready.  @check_ready should return
+ *	positive number if @link is ready, 0 if it isn't, -ENODEV if
+ *	link doesn't seem to be occupied, other errno for other error
+ *	conditions.
+ *
+ *	Transient -ENODEV conditions are allowed for
+ *	ATA_TMOUT_FF_WAIT.
+ *
+ *	LOCKING:
+ *	EH context.
+ *
+ *	RETURNS:
+ *	0 if @linke is ready before @deadline; otherwise, -errno.
+ */
+int ata_wait_ready(struct ata_link *link, unsigned long deadline,
+		   int (*check_ready)(struct ata_link *link))
+{
+	unsigned long start = jiffies;
+	unsigned long nodev_deadline = start + ATA_TMOUT_FF_WAIT;
+	int warned = 0;
+
+	if (time_after(nodev_deadline, deadline))
+		nodev_deadline = deadline;
+
+	while (1) {
+		unsigned long now = jiffies;
+		int ready, tmp;
+
+		ready = tmp = check_ready(link);
+		if (ready > 0)
+			return 0;
+
+		/* -ENODEV could be transient.  Ignore -ENODEV if link
+		 * is online.  Also, some SATA devices take a long
+		 * time to clear 0xff after reset.  For example,
+		 * HHD424020F7SV00 iVDR needs >= 800ms while Quantum
+		 * GoVault needs even more than that.  Wait for
+		 * ATA_TMOUT_FF_WAIT on -ENODEV if link isn't offline.
+		 *
+		 * Note that some PATA controllers (pata_ali) explode
+		 * if status register is read more than once when
+		 * there's no device attached.
+		 */
+		if (ready == -ENODEV) {
+			if (ata_link_online(link))
+				ready = 0;
+			else if ((link->ap->flags & ATA_FLAG_SATA) &&
+				 !ata_link_offline(link) &&
+				 time_before(now, nodev_deadline))
+				ready = 0;
+		}
+
+		if (ready)
+			return ready;
+		if (time_after(now, deadline))
+			return -EBUSY;
+
+		if (!warned && time_after(now, start + 5 * HZ) &&
+		    (deadline - now > 3 * HZ)) {
+			ata_link_printk(link, KERN_WARNING,
+				"link is slow to respond, please be patient "
+				"(ready=%d)\n", tmp);
+			warned = 1;
+		}
+
+		msleep(50);
+	}
+}
+
+/**
+ *	ata_wait_after_reset - wait for link to become ready after reset
+ *	@link: link to be waited on
+ *	@deadline: deadline jiffies for the operation
+ *	@check_ready: callback to check link readiness
+ *
+ *	Wait for @link to become ready after reset.
+ *
+ *	LOCKING:
+ *	EH context.
+ *
+ *	RETURNS:
+ *	0 if @linke is ready before @deadline; otherwise, -errno.
+ */
+extern int ata_wait_after_reset(struct ata_link *link, unsigned long deadline,
+				int (*check_ready)(struct ata_link *link))
+{
+	msleep(ATA_WAIT_AFTER_RESET_MSECS);
+
+	return ata_wait_ready(link, deadline, check_ready);
+}
+
 /**
  *	sata_link_debounce - debounce SATA phy status
  *	@link: ATA link to debounce SATA phy status for
@@ -6075,6 +6172,7 @@ EXPORT_SYMBOL_GPL(ata_noop_qc_prep);
 EXPORT_SYMBOL_GPL(ata_port_probe);
 EXPORT_SYMBOL_GPL(ata_dev_disable);
 EXPORT_SYMBOL_GPL(sata_set_spd);
+EXPORT_SYMBOL_GPL(ata_wait_after_reset);
 EXPORT_SYMBOL_GPL(sata_link_debounce);
 EXPORT_SYMBOL_GPL(sata_link_resume);
 EXPORT_SYMBOL_GPL(ata_std_prereset);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 6e8de3c1595e..78912c5011ad 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -308,6 +308,17 @@ int ata_sff_busy_sleep(struct ata_port *ap,
 	return 0;
 }
 
+static int ata_sff_check_ready(struct ata_link *link)
+{
+	u8 status = link->ap->ops->sff_check_status(link->ap);
+
+	if (!(status & ATA_BUSY))
+		return 1;
+	if (status == 0xff)
+		return -ENODEV;
+	return 0;
+}
+
 /**
  *	ata_sff_wait_ready - sleep until BSY clears, or timeout
  *	@link: SFF link to wait ready status for
@@ -324,56 +335,7 @@ int ata_sff_busy_sleep(struct ata_port *ap,
  */
 int ata_sff_wait_ready(struct ata_link *link, unsigned long deadline)
 {
-	struct ata_port *ap = link->ap;
-	unsigned long start = jiffies;
-	unsigned long nodev_deadline = start + ATA_TMOUT_FF_WAIT;
-	int warned = 0;
-
-	if (time_after(nodev_deadline, deadline))
-		nodev_deadline = deadline;
-
-	while (1) {
-		u8 status = ap->ops->sff_check_status(ap);
-		unsigned long now = jiffies;
-
-		if (!(status & ATA_BUSY))
-			return 0;
-
-		/* No device status could be transient.  Ignore it if
-		 * link is online.  Also, some SATA devices take a
-		 * long time to clear 0xff after reset.  For example,
-		 * HHD424020F7SV00 iVDR needs >= 800ms while Quantum
-		 * GoVault needs even more than that.  Wait for
-		 * ATA_TMOUT_FF_WAIT on -ENODEV if link isn't offline.
-		 *
-		 * Note that some PATA controllers (pata_ali) explode
-		 * if status register is read more than once when
-		 * there's no device attached.
-		 */
-		if (status == 0xff) {
-			if (ata_link_online(link))
-				status = ATA_BUSY;
-			else if ((link->ap->flags & ATA_FLAG_SATA) &&
-				 !ata_link_offline(link) &&
-				 time_before(now, nodev_deadline))
-				status = ATA_BUSY;
-			if (status == 0xff)
-				return -ENODEV;
-		}
-
-		if (time_after(now, deadline))
-			return -EBUSY;
-
-		if (!warned && time_after(now, start + 5 * HZ) &&
-		    (deadline - now > 3 * HZ)) {
-			ata_link_printk(link, KERN_WARNING,
-				"link is slow to respond, please be patient "
-				"(Status 0x%x)\n", status);
-			warned = 1;
-		}
-
-		msleep(50);
-	}
+	return ata_wait_ready(link, deadline, ata_sff_check_ready);
 }
 
 /**
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 11b5f67a19d4..08af43e2c081 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -81,6 +81,8 @@ extern unsigned ata_exec_internal_sg(struct ata_device *dev,
 				     int dma_dir, struct scatterlist *sg,
 				     unsigned int n_elem, unsigned long timeout);
 extern unsigned int ata_do_simple_cmd(struct ata_device *dev, u8 cmd);
+extern int ata_wait_ready(struct ata_link *link, unsigned long deadline,
+			  int (*check_ready)(struct ata_link *link));
 extern int ata_dev_read_id(struct ata_device *dev, unsigned int *p_class,
 			   unsigned int flags, u16 *id);
 extern int ata_dev_reread_id(struct ata_device *dev, unsigned int readid_flags);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index da5560244787..4bbf2524e473 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -837,6 +837,8 @@ extern void sata_print_link_status(struct ata_link *link);
 extern void ata_port_probe(struct ata_port *);
 extern int sata_set_spd(struct ata_link *link);
 extern int ata_std_prereset(struct ata_link *link, unsigned long deadline);
+extern int ata_wait_after_reset(struct ata_link *link, unsigned long deadline,
+				int (*check_ready)(struct ata_link *link));
 extern int sata_link_debounce(struct ata_link *link,
 			const unsigned long *params, unsigned long deadline);
 extern int sata_link_resume(struct ata_link *link, const unsigned long *params,
-- 
cgit v1.2.3


From 9dadd45b24145d6aee2fabb28d7aef972301892b Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:19 +0900
Subject: libata: move generic hardreset code from sata_sff_hardreset() to
 sata_link_hardreset()

sata_sff_hardreset() contains link readiness wait logic which isn't
SFF specific.  Move that part into sata_link_hardreset(), which now
takes two more parameters - @online and @check_ready.  Both are
optional.  The former is out parameter for link onlineness after
reset.  The latter is used to wait for link readiness after hardreset.

Users of sata_link_hardreset() is updated to use new funtionality and
ahci_hardreset() is updated to use sata_link_hardreset() instead of
sata_sff_hardreset().  This doesn't really cause any behavior change.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c        | 35 +++++++++++++++++++--------------
 drivers/ata/ata_piix.c    |  2 +-
 drivers/ata/libata-core.c | 50 ++++++++++++++++++++++++++++++++++++++++++++++-
 drivers/ata/libata-pmp.c  |  7 ++++---
 drivers/ata/libata-sff.c  | 49 +++++++++-------------------------------------
 include/linux/libata.h    | 10 +++++++++-
 6 files changed, 92 insertions(+), 61 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 7e251a2cbda5..0f553aaa6f79 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1343,10 +1343,12 @@ static int ahci_softreset(struct ata_link *link, unsigned int *class,
 static int ahci_hardreset(struct ata_link *link, unsigned int *class,
 			  unsigned long deadline)
 {
+	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
 	struct ata_port *ap = link->ap;
 	struct ahci_port_priv *pp = ap->private_data;
 	u8 *d2h_fis = pp->rx_fis + RX_FIS_D2H_REG;
 	struct ata_taskfile tf;
+	bool online;
 	int rc;
 
 	DPRINTK("ENTER\n");
@@ -1358,14 +1360,14 @@ static int ahci_hardreset(struct ata_link *link, unsigned int *class,
 	tf.command = 0x80;
 	ata_tf_to_fis(&tf, 0, 0, d2h_fis);
 
-	rc = sata_sff_hardreset(link, class, deadline);
+	rc = sata_link_hardreset(link, timing, deadline, &online,
+				 ahci_check_ready);
 
 	ahci_start_engine(ap);
 
-	if (rc == 0 && ata_link_online(link))
+	*class = ATA_DEV_NONE;
+	if (online)
 		*class = ahci_dev_classify(ap);
-	if (rc != -EAGAIN && *class == ATA_DEV_UNKNOWN)
-		*class = ATA_DEV_NONE;
 
 	DPRINTK("EXIT, rc=%d, class=%u\n", rc, *class);
 	return rc;
@@ -1376,6 +1378,7 @@ static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
 {
 	struct ata_port *ap = link->ap;
 	u32 serror;
+	bool online;
 	int rc;
 
 	DPRINTK("ENTER\n");
@@ -1383,7 +1386,7 @@ static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
 	ahci_stop_engine(ap);
 
 	rc = sata_link_hardreset(link, sata_ehc_deb_timing(&link->eh_context),
-				 deadline);
+				 deadline, &online, NULL);
 
 	/* vt8251 needs SError cleared for the port to operate */
 	ahci_scr_read(ap, SCR_ERROR, &serror);
@@ -1396,7 +1399,8 @@ static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
 	/* vt8251 doesn't clear BSY on signature FIS reception,
 	 * request follow-up softreset.
 	 */
-	return rc ?: -EAGAIN;
+	*class = ATA_DEV_NONE;
+	return online ? -EAGAIN : rc;
 }
 
 static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
@@ -1406,6 +1410,7 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 	struct ahci_port_priv *pp = ap->private_data;
 	u8 *d2h_fis = pp->rx_fis + RX_FIS_D2H_REG;
 	struct ata_taskfile tf;
+	bool online;
 	int rc;
 
 	ahci_stop_engine(ap);
@@ -1416,13 +1421,10 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 	ata_tf_to_fis(&tf, 0, 0, d2h_fis);
 
 	rc = sata_link_hardreset(link, sata_ehc_deb_timing(&link->eh_context),
-				 deadline);
+				 deadline, &online, NULL);
 
 	ahci_start_engine(ap);
 
-	if (rc || ata_link_offline(link))
-		return rc;
-
 	/* The pseudo configuration device on SIMG4726 attached to
 	 * ASUS P5W-DH Deluxe doesn't send signature FIS after
 	 * hardreset if no device is attached to the first downstream
@@ -1436,11 +1438,14 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 	 * have to be reset again.  For most cases, this should
 	 * suffice while making probing snappish enough.
 	 */
-	rc = ata_wait_after_reset(link, jiffies + 2 * HZ, ahci_check_ready);
-	if (rc)
-		ahci_kick_engine(ap, 0);
-
-	return 0;
+	if (online) {
+		rc = ata_wait_after_reset(link, jiffies + 2 * HZ,
+					  ahci_check_ready);
+		if (rc)
+			ahci_kick_engine(ap, 0);
+	}
+	*class = ATA_DEV_NONE;
+	return rc;
 }
 
 static void ahci_postreset(struct ata_link *link, unsigned int *class)
diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index 7ab76a413cdf..f59a55bfade4 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -1022,7 +1022,7 @@ static int piix_sidpr_hardreset(struct ata_link *link, unsigned int *class,
 	int rc;
 
 	/* do hardreset */
-	rc = sata_link_hardreset(link, timing, deadline);
+	rc = sata_link_hardreset(link, timing, deadline, NULL, NULL);
 	if (rc) {
 		ata_link_printk(link, KERN_ERR,
 				"COMRESET failed (errno=%d)\n", rc);
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 3bad6f189190..b607292b6480 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -3557,8 +3557,18 @@ int ata_std_prereset(struct ata_link *link, unsigned long deadline)
  *	@link: link to reset
  *	@timing: timing parameters { interval, duratinon, timeout } in msec
  *	@deadline: deadline jiffies for the operation
+ *	@online: optional out parameter indicating link onlineness
+ *	@check_ready: optional callback to check link readiness
  *
  *	SATA phy-reset @link using DET bits of SControl register.
+ *	After hardreset, link readiness is waited upon using
+ *	ata_wait_ready() if @check_ready is specified.  LLDs are
+ *	allowed to not specify @check_ready and wait itself after this
+ *	function returns.  Device classification is LLD's
+ *	responsibility.
+ *
+ *	*@online is set to one iff reset succeeded and @link is online
+ *	after reset.
  *
  *	LOCKING:
  *	Kernel thread context (may sleep)
@@ -3567,13 +3577,17 @@ int ata_std_prereset(struct ata_link *link, unsigned long deadline)
  *	0 on success, -errno otherwise.
  */
 int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
-			unsigned long deadline)
+			unsigned long deadline,
+			bool *online, int (*check_ready)(struct ata_link *))
 {
 	u32 scontrol;
 	int rc;
 
 	DPRINTK("ENTER\n");
 
+	if (online)
+		*online = false;
+
 	if (sata_set_spd_needed(link)) {
 		/* SATA spec says nothing about how to reconfigure
 		 * spd.  To be on the safe side, turn off phy during
@@ -3607,7 +3621,41 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
 
 	/* bring link back */
 	rc = sata_link_resume(link, timing, deadline);
+	if (rc)
+		goto out;
+	/* if link is offline nothing more to do */
+	if (ata_link_offline(link))
+		goto out;
+
+	/* Link is online.  From this point, -ENODEV too is an error. */
+	if (online)
+		*online = true;
+
+	if ((link->ap->flags & ATA_FLAG_PMP) && ata_is_host_link(link)) {
+		/* If PMP is supported, we have to do follow-up SRST.
+		 * Some PMPs don't send D2H Reg FIS after hardreset if
+		 * the first port is empty.  Wait only for
+		 * ATA_TMOUT_PMP_SRST_WAIT.
+		 */
+		if (check_ready) {
+			unsigned long pmp_deadline;
+
+			pmp_deadline = jiffies + ATA_TMOUT_PMP_SRST_WAIT;
+			if (time_after(pmp_deadline, deadline))
+				pmp_deadline = deadline;
+			ata_wait_ready(link, pmp_deadline, check_ready);
+		}
+		rc = -EAGAIN;
+		goto out;
+	}
+
+	rc = 0;
+	if (check_ready)
+		rc = ata_wait_ready(link, deadline, check_ready);
  out:
+	if (rc && rc != -EAGAIN)
+		ata_link_printk(link, KERN_ERR,
+				"COMRESET failed (errno=%d)\n", rc);
 	DPRINTK("EXIT, rc=%d\n", rc);
 	return rc;
 }
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index a7cb1498c9b2..7f1a87f01ab2 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -239,13 +239,14 @@ int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
 			   unsigned long deadline)
 {
 	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
+	bool online;
 	u32 tmp;
 	int rc;
 
 	DPRINTK("ENTER\n");
 
 	/* do hardreset */
-	rc = sata_link_hardreset(link, timing, deadline);
+	rc = sata_link_hardreset(link, timing, deadline, &online, NULL);
 	if (rc) {
 		ata_link_printk(link, KERN_ERR,
 				"COMRESET failed (errno=%d)\n", rc);
@@ -261,7 +262,7 @@ int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
 	}
 
 	/* if device is present, follow up with srst to wait for !BSY */
-	if (ata_link_online(link))
+	if (online)
 		rc = -EAGAIN;
  out:
 	/* if SCR isn't accessible, we need to reset the PMP */
@@ -916,7 +917,7 @@ static int sata_pmp_eh_handle_disabled_links(struct ata_port *ap)
 		 * SError.N working.
 		 */
 		sata_link_hardreset(link, sata_deb_timing_normal,
-				    jiffies + ATA_TMOUT_INTERNAL_QUICK);
+				jiffies + ATA_TMOUT_INTERNAL_QUICK, NULL, NULL);
 
 		/* unconditionally clear SError.N */
 		rc = sata_scr_write(link, SCR_ERROR, SERR_PHYRDY_CHG);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 78912c5011ad..0b97e84d3af6 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -1921,50 +1921,19 @@ int ata_sff_softreset(struct ata_link *link, unsigned int *classes,
 int sata_sff_hardreset(struct ata_link *link, unsigned int *class,
 		       unsigned long deadline)
 {
-	struct ata_port *ap = link->ap;
-	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
+	struct ata_eh_context *ehc = &link->eh_context;
+	const unsigned long *timing = sata_ehc_deb_timing(ehc);
+	bool online;
 	int rc;
 
-	DPRINTK("ENTER\n");
-
-	/* do hardreset */
-	rc = sata_link_hardreset(link, timing, deadline);
-	if (rc) {
-		ata_link_printk(link, KERN_ERR,
-				"COMRESET failed (errno=%d)\n", rc);
-		return rc;
-	}
-
-	/* TODO: phy layer with polling, timeouts, etc. */
-	if (ata_link_offline(link)) {
-		*class = ATA_DEV_NONE;
-		DPRINTK("EXIT, link offline\n");
-		return 0;
-	}
-
-	/* If PMP is supported, we have to do follow-up SRST.  Note
-	 * that some PMPs don't send D2H Reg FIS after hardreset at
-	 * all if the first port is empty.  Wait for it just for a
-	 * second and request follow-up SRST.
-	 */
-	if (ap->flags & ATA_FLAG_PMP) {
-		ata_sff_wait_after_reset(link, 1, jiffies + HZ);
-		return -EAGAIN;
-	}
-
-	/* wait for the link to become online */
-	rc = ata_sff_wait_after_reset(link, 1, deadline);
-	/* link occupied, -ENODEV too is an error */
-	if (rc) {
-		ata_link_printk(link, KERN_ERR,
-				"COMRESET failed (errno=%d)\n", rc);
-		return rc;
-	}
-
-	*class = ata_sff_dev_classify(link->device, 1, NULL);
+	rc = sata_link_hardreset(link, timing, deadline, &online,
+				 ata_sff_check_ready);
+	*class = ATA_DEV_NONE;
+	if (online)
+		*class = ata_sff_dev_classify(link->device, 1, NULL);
 
 	DPRINTK("EXIT, class=%u\n", *class);
-	return 0;
+	return rc;
 }
 
 /**
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 4bbf2524e473..d9ebce2bf5e7 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -261,6 +261,13 @@ enum {
 	 */
 	ATA_WAIT_AFTER_RESET_MSECS = 150,
 
+	/* If PMP is supported, we have to do follow-up SRST.  As some
+	 * PMPs don't send D2H Reg FIS after hardreset, LLDs are
+	 * advised to wait only for the following duration before
+	 * doing SRST.
+	 */
+	ATA_TMOUT_PMP_SRST_WAIT	= 1 * HZ,
+
 	/* ATA bus states */
 	BUS_UNKNOWN		= 0,
 	BUS_DMA			= 1,
@@ -844,7 +851,8 @@ extern int sata_link_debounce(struct ata_link *link,
 extern int sata_link_resume(struct ata_link *link, const unsigned long *params,
 			    unsigned long deadline);
 extern int sata_link_hardreset(struct ata_link *link,
-			const unsigned long *timing, unsigned long deadline);
+			const unsigned long *timing, unsigned long deadline,
+			bool *online, int (*check_ready)(struct ata_link *));
 extern void ata_std_postreset(struct ata_link *link, unsigned int *classes);
 extern void ata_port_disable(struct ata_port *);
 
-- 
cgit v1.2.3


From 57c9efdfb3cee5d4564fcb5f70555e2edb1bc52a Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:19 +0900
Subject: libata: implement and use sata_std_hardreset()

Implement sata_std_hardreset(), which simply wraps around
sata_link_hardreset().  sata_std_hardreset() becomes new standard
hardreset method for sata_port_ops and sata_sff_hardreset() moves from
ata_base_port_ops to ata_sff_port_ops, which is where it really
belongs.

ata_is_builtin_hardreset() is added so that both
ata_std_error_handler() and ata_sff_error_handler() skip both builtin
hardresets if SCR isn't accessible.

piix_sidpr_hardreset() in ata_piix.c is identical to
sata_std_hardreset() in functionality and got replaced with the
standard function.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ata_piix.c    | 27 +--------------------------
 drivers/ata/libata-core.c | 30 +++++++++++++++++++++++++++++-
 drivers/ata/libata-eh.c   |  7 ++-----
 drivers/ata/libata-sff.c  |  9 ++++-----
 drivers/ata/libata.h      |  9 +++++++++
 include/linux/libata.h    |  2 ++
 6 files changed, 47 insertions(+), 37 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c
index f59a55bfade4..b7c38eeb498f 100644
--- a/drivers/ata/ata_piix.c
+++ b/drivers/ata/ata_piix.c
@@ -165,8 +165,6 @@ static void piix_set_dmamode(struct ata_port *ap, struct ata_device *adev);
 static void ich_set_dmamode(struct ata_port *ap, struct ata_device *adev);
 static int ich_pata_cable_detect(struct ata_port *ap);
 static u8 piix_vmw_bmdma_status(struct ata_port *ap);
-static int piix_sidpr_hardreset(struct ata_link *link, unsigned int *class,
-				unsigned long deadline);
 static int piix_sidpr_scr_read(struct ata_port *ap, unsigned int reg, u32 *val);
 static int piix_sidpr_scr_write(struct ata_port *ap, unsigned int reg, u32 val);
 #ifdef CONFIG_PM
@@ -319,7 +317,7 @@ static struct ata_port_operations piix_sata_ops = {
 
 static struct ata_port_operations piix_sidpr_sata_ops = {
 	.inherits		= &piix_sata_ops,
-	.hardreset		= piix_sidpr_hardreset,
+	.hardreset		= sata_std_hardreset,
 	.scr_read		= piix_sidpr_scr_read,
 	.scr_write		= piix_sidpr_scr_write,
 };
@@ -1015,29 +1013,6 @@ static int piix_sidpr_scr_write(struct ata_port *ap, unsigned int reg, u32 val)
 	return 0;
 }
 
-static int piix_sidpr_hardreset(struct ata_link *link, unsigned int *class,
-				unsigned long deadline)
-{
-	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
-	int rc;
-
-	/* do hardreset */
-	rc = sata_link_hardreset(link, timing, deadline, NULL, NULL);
-	if (rc) {
-		ata_link_printk(link, KERN_ERR,
-				"COMRESET failed (errno=%d)\n", rc);
-		return rc;
-	}
-
-	/* TODO: phy layer with polling, timeouts, etc. */
-	if (ata_link_offline(link)) {
-		*class = ATA_DEV_NONE;
-		return 0;
-	}
-
-	return -EAGAIN;
-}
-
 #ifdef CONFIG_PM
 static int piix_broken_suspend(void)
 {
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index b607292b6480..c4fd4afbf349 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -75,7 +75,6 @@ const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
 const struct ata_port_operations ata_base_port_ops = {
 	.prereset		= ata_std_prereset,
-	.hardreset		= sata_sff_hardreset,
 	.postreset		= ata_std_postreset,
 	.error_handler		= ata_std_error_handler,
 };
@@ -84,6 +83,7 @@ const struct ata_port_operations sata_port_ops = {
 	.inherits		= &ata_base_port_ops,
 
 	.qc_defer		= ata_std_qc_defer,
+	.hardreset		= sata_std_hardreset,
 	.sff_dev_select		= ata_noop_dev_select,
 };
 
@@ -3660,6 +3660,33 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
 	return rc;
 }
 
+/**
+ *	sata_std_hardreset - COMRESET w/o waiting or classification
+ *	@link: link to reset
+ *	@class: resulting class of attached device
+ *	@deadline: deadline jiffies for the operation
+ *
+ *	Standard SATA COMRESET w/o waiting or classification.
+ *
+ *	LOCKING:
+ *	Kernel thread context (may sleep)
+ *
+ *	RETURNS:
+ *	0 if link offline, -EAGAIN if link online, -errno on errors.
+ */
+int sata_std_hardreset(struct ata_link *link, unsigned int *class,
+		       unsigned long deadline)
+{
+	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
+	bool online;
+	int rc;
+
+	/* do hardreset */
+	rc = sata_link_hardreset(link, timing, deadline, &online, NULL);
+	*class = ATA_DEV_NONE;
+	return online ? -EAGAIN : rc;
+}
+
 /**
  *	ata_std_postreset - standard postreset callback
  *	@link: the target ata_link
@@ -6225,6 +6252,7 @@ EXPORT_SYMBOL_GPL(sata_link_debounce);
 EXPORT_SYMBOL_GPL(sata_link_resume);
 EXPORT_SYMBOL_GPL(ata_std_prereset);
 EXPORT_SYMBOL_GPL(sata_link_hardreset);
+EXPORT_SYMBOL_GPL(sata_std_hardreset);
 EXPORT_SYMBOL_GPL(ata_std_postreset);
 EXPORT_SYMBOL_GPL(ata_dev_classify);
 EXPORT_SYMBOL_GPL(ata_dev_pair);
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index f4f9c2783821..21687bbd9a70 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -2857,11 +2857,8 @@ void ata_std_error_handler(struct ata_port *ap)
 	struct ata_port_operations *ops = ap->ops;
 	ata_reset_fn_t hardreset = ops->hardreset;
 
-	/* sata_std_hardreset is inherited to all drivers from
-	 * ata_base_port_ops.  Ignore it if SCR access is not
-	 * available.
-	 */
-	if (hardreset == sata_sff_hardreset && !sata_scr_valid(&ap->link))
+	/* ignore built-in hardreset if SCR access is not available */
+	if (ata_is_builtin_hardreset(hardreset) && !sata_scr_valid(&ap->link))
 		hardreset = NULL;
 
 	ata_do_eh(ap, ops->prereset, ops->softreset, hardreset, ops->postreset);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 0b97e84d3af6..f464ca1fa261 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -49,6 +49,7 @@ const struct ata_port_operations ata_sff_port_ops = {
 	.thaw			= ata_sff_thaw,
 	.prereset		= ata_sff_prereset,
 	.softreset		= ata_sff_softreset,
+	.hardreset		= sata_sff_hardreset,
 	.postreset		= ata_sff_postreset,
 	.error_handler		= ata_sff_error_handler,
 	.post_internal_cmd	= ata_sff_post_internal_cmd,
@@ -2031,14 +2032,12 @@ void ata_sff_error_handler(struct ata_port *ap)
 
 	/* PIO and DMA engines have been stopped, perform recovery */
 
-	/* ata_sff_softreset and sata_sff_hardreset are inherited to
-	 * all SFF drivers from ata_sff_port_ops.  Ignore softreset if
-	 * ctl isn't accessible.  Ignore hardreset if SCR access isn't
-	 * available.
+	/* Ignore ata_sff_softreset if ctl isn't accessible and
+	 * built-in hardresets if SCR access isn't available.
 	 */
 	if (softreset == ata_sff_softreset && !ap->ioaddr.ctl_addr)
 		softreset = NULL;
-	if (hardreset == sata_sff_hardreset && !sata_scr_valid(&ap->link))
+	if (ata_is_builtin_hardreset(hardreset) && !sata_scr_valid(&ap->link))
 		hardreset = NULL;
 
 	ata_do_eh(ap, ap->ops->prereset, softreset, hardreset,
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 08af43e2c081..87f54a1db3b0 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -38,6 +38,15 @@ struct ata_scsi_args {
 	void			(*done)(struct scsi_cmnd *);
 };
 
+static inline int ata_is_builtin_hardreset(ata_reset_fn_t reset)
+{
+	if (reset == sata_std_hardreset)
+		return 1;
+	if (reset == sata_sff_hardreset)
+		return 1;
+	return 0;
+}
+
 /* libata-core.c */
 enum {
 	/* flags for ata_dev_read_id() */
diff --git a/include/linux/libata.h b/include/linux/libata.h
index d9ebce2bf5e7..c060cd3cba66 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -853,6 +853,8 @@ extern int sata_link_resume(struct ata_link *link, const unsigned long *params,
 extern int sata_link_hardreset(struct ata_link *link,
 			const unsigned long *timing, unsigned long deadline,
 			bool *online, int (*check_ready)(struct ata_link *));
+extern int sata_std_hardreset(struct ata_link *link, unsigned int *class,
+			      unsigned long deadline);
 extern void ata_std_postreset(struct ata_link *link, unsigned int *classes);
 extern void ata_port_disable(struct ata_port *);
 
-- 
cgit v1.2.3


From ac371987a81c61c2efbd6931245cdcaf43baad89 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:19 +0900
Subject: libata: clear SError after link resume

SError used to be cleared in ->postreset.  This has small hotplug race
condition.  If a device is plugged in after reset is complete but
postreset hasn't run yet, its hotplug event gets lost when SError is
cleared.  This patch makes sata_link_resume() clear SError.  This
kills the race condition and makes a lot of sense as some PMP and host
PHYs don't work properly without SError cleared.

This change makes sata_pmp_std_{pre|post}_reset()'s unnecessary as
they become identical to ata_std counterparts.  It also simplifies
sata_pmp_hardreset() and ahci_vt8251_hardreset().

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c        |  5 ---
 drivers/ata/libata-core.c | 35 +++++++++++-------
 drivers/ata/libata-pmp.c  | 93 +----------------------------------------------
 include/linux/libata.h    |  2 -
 4 files changed, 23 insertions(+), 112 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 0f553aaa6f79..a69bcca4eb1b 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1377,7 +1377,6 @@ static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
 				 unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
-	u32 serror;
 	bool online;
 	int rc;
 
@@ -1388,10 +1387,6 @@ static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
 	rc = sata_link_hardreset(link, sata_ehc_deb_timing(&link->eh_context),
 				 deadline, &online, NULL);
 
-	/* vt8251 needs SError cleared for the port to operate */
-	ahci_scr_read(ap, SCR_ERROR, &serror);
-	ahci_scr_write(ap, SCR_ERROR, serror);
-
 	ahci_start_engine(ap);
 
 	DPRINTK("EXIT, rc=%d, class=%u\n", rc, *class);
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index c4fd4afbf349..e00b620f161a 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -90,9 +90,9 @@ const struct ata_port_operations sata_port_ops = {
 const struct ata_port_operations sata_pmp_port_ops = {
 	.inherits		= &sata_port_ops,
 
-	.pmp_prereset		= sata_pmp_std_prereset,
+	.pmp_prereset		= ata_std_prereset,
 	.pmp_hardreset		= sata_pmp_std_hardreset,
-	.pmp_postreset		= sata_pmp_std_postreset,
+	.pmp_postreset		= ata_std_postreset,
 	.error_handler		= sata_pmp_error_handler,
 };
 
@@ -3493,7 +3493,7 @@ int sata_link_debounce(struct ata_link *link, const unsigned long *params,
 int sata_link_resume(struct ata_link *link, const unsigned long *params,
 		     unsigned long deadline)
 {
-	u32 scontrol;
+	u32 scontrol, serror;
 	int rc;
 
 	if ((rc = sata_scr_read(link, SCR_CONTROL, &scontrol)))
@@ -3509,7 +3509,25 @@ int sata_link_resume(struct ata_link *link, const unsigned long *params,
 	 */
 	msleep(200);
 
-	return sata_link_debounce(link, params, deadline);
+	if ((rc = sata_link_debounce(link, params, deadline)))
+		return rc;
+
+	/* Clear SError.  PMP and some host PHYs require this to
+	 * operate and clearing should be done before checking PHY
+	 * online status to avoid race condition (hotplugging between
+	 * link resume and status check).
+	 */
+	if (!(rc = sata_scr_read(link, SCR_ERROR, &serror)))
+		rc = sata_scr_write(link, SCR_ERROR, serror);
+	if (rc == 0 || rc == -EINVAL) {
+		unsigned long flags;
+
+		spin_lock_irqsave(link->ap->lock, flags);
+		link->eh_info.serror = 0;
+		spin_unlock_irqrestore(link->ap->lock, flags);
+		rc = 0;
+	}
+	return rc;
 }
 
 /**
@@ -3701,18 +3719,11 @@ int sata_std_hardreset(struct ata_link *link, unsigned int *class,
  */
 void ata_std_postreset(struct ata_link *link, unsigned int *classes)
 {
-	u32 serror;
-
 	DPRINTK("ENTER\n");
 
 	/* print link status */
 	sata_print_link_status(link);
 
-	/* clear SError */
-	if (sata_scr_read(link, SCR_ERROR, &serror) == 0)
-		sata_scr_write(link, SCR_ERROR, serror);
-	link->eh_info.serror = 0;
-
 	DPRINTK("EXIT\n");
 }
 
@@ -6296,9 +6307,7 @@ EXPORT_SYMBOL_GPL(ata_pci_device_resume);
 #endif /* CONFIG_PCI */
 
 EXPORT_SYMBOL_GPL(sata_pmp_qc_defer_cmd_switch);
-EXPORT_SYMBOL_GPL(sata_pmp_std_prereset);
 EXPORT_SYMBOL_GPL(sata_pmp_std_hardreset);
-EXPORT_SYMBOL_GPL(sata_pmp_std_postreset);
 EXPORT_SYMBOL_GPL(sata_pmp_error_handler);
 
 EXPORT_SYMBOL_GPL(__ata_ehi_push_desc);
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index 7f1a87f01ab2..2f8a9577c26d 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -175,49 +175,6 @@ int sata_pmp_scr_write(struct ata_link *link, int reg, u32 val)
 	return 0;
 }
 
-/**
- *	sata_pmp_std_prereset - prepare PMP link for reset
- *	@link: link to be reset
- *	@deadline: deadline jiffies for the operation
- *
- *	@link is about to be reset.  Initialize it.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep)
- *
- *	RETURNS:
- *	0 on success, -errno otherwise.
- */
-int sata_pmp_std_prereset(struct ata_link *link, unsigned long deadline)
-{
-	struct ata_eh_context *ehc = &link->eh_context;
-	const unsigned long *timing = sata_ehc_deb_timing(ehc);
-	int rc;
-
-	/* if we're about to do hardreset, nothing more to do */
-	if (ehc->i.action & ATA_EH_HARDRESET)
-		return 0;
-
-	/* resume link */
-	rc = sata_link_resume(link, timing, deadline);
-	if (rc) {
-		/* phy resume failed */
-		ata_link_printk(link, KERN_WARNING, "failed to resume link "
-				"for reset (errno=%d)\n", rc);
-		return rc;
-	}
-
-	/* clear SError bits including .X which blocks the port when set */
-	rc = sata_scr_write(link, SCR_ERROR, 0xffffffff);
-	if (rc) {
-		ata_link_printk(link, KERN_ERR,
-				"failed to clear SError (errno=%d)\n", rc);
-		return rc;
-	}
-
-	return 0;
-}
-
 /**
  *	sata_pmp_std_hardreset - standard hardreset method for PMP link
  *	@link: link to be reset
@@ -238,33 +195,13 @@ int sata_pmp_std_prereset(struct ata_link *link, unsigned long deadline)
 int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
 			   unsigned long deadline)
 {
-	const unsigned long *timing = sata_ehc_deb_timing(&link->eh_context);
-	bool online;
 	u32 tmp;
 	int rc;
 
 	DPRINTK("ENTER\n");
 
-	/* do hardreset */
-	rc = sata_link_hardreset(link, timing, deadline, &online, NULL);
-	if (rc) {
-		ata_link_printk(link, KERN_ERR,
-				"COMRESET failed (errno=%d)\n", rc);
-		goto out;
-	}
-
-	/* clear SError bits including .X which blocks the port when set */
-	rc = sata_scr_write(link, SCR_ERROR, 0xffffffff);
-	if (rc) {
-		ata_link_printk(link, KERN_ERR, "failed to clear SError "
-				"during hardreset (errno=%d)\n", rc);
-		goto out;
-	}
+	rc = sata_std_hardreset(link, class, deadline);
 
-	/* if device is present, follow up with srst to wait for !BSY */
-	if (online)
-		rc = -EAGAIN;
- out:
 	/* if SCR isn't accessible, we need to reset the PMP */
 	if (rc && rc != -EAGAIN && sata_scr_read(link, SCR_STATUS, &tmp))
 		rc = -ERESTART;
@@ -273,34 +210,6 @@ int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
 	return rc;
 }
 
-/**
- *	ata_std_postreset - standard postreset method for PMP link
- *	@link: the target ata_link
- *	@classes: classes of attached devices
- *
- *	This function is invoked after a successful reset.  Note that
- *	the device might have been reset more than once using
- *	different reset methods before postreset is invoked.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep)
- */
-void sata_pmp_std_postreset(struct ata_link *link, unsigned int *class)
-{
-	u32 serror;
-
-	DPRINTK("ENTER\n");
-
-	/* clear SError */
-	if (sata_scr_read(link, SCR_ERROR, &serror) == 0)
-		sata_scr_write(link, SCR_ERROR, serror);
-
-	/* print link status */
-	sata_print_link_status(link);
-
-	DPRINTK("EXIT\n");
-}
-
 /**
  *	sata_pmp_read_gscr - read GSCR block of SATA PMP
  *	@dev: PMP device
diff --git a/include/linux/libata.h b/include/linux/libata.h
index c060cd3cba66..b9188371b12a 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1025,10 +1025,8 @@ static inline int ata_acpi_cbl_80wire(struct ata_port *ap,
  * PMP - drivers/ata/libata-pmp.c
  */
 extern int sata_pmp_qc_defer_cmd_switch(struct ata_queued_cmd *qc);
-extern int sata_pmp_std_prereset(struct ata_link *link, unsigned long deadline);
 extern int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
 				  unsigned long deadline);
-extern void sata_pmp_std_postreset(struct ata_link *link, unsigned int *class);
 extern void sata_pmp_error_handler(struct ata_port *ap);
 
 /*
-- 
cgit v1.2.3


From 5958e3025fd9d97429163e074d9cfa3848f51f28 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:20 +0900
Subject: libata: move PMP SCR access failure during reset to ata_eh_reset()

If PMP fan-out reset fails and SCR isn't accessible, PMP should be
reset.  This used to be tested by sata_pmp_std_hardreset() and
communicated to EH by -ERESTART.  However, this logic is generic and
doesn't really have much to do with specific hardreset implementation.

This patch moves SCR access failure detection logic to ata_eh_reset()
where it belongs.  As this makes sata_pmp_std_hardreset() identical to
sata_std_hardreset(), the function is killed and replaced with the
standard method.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c |  3 +--
 drivers/ata/libata-eh.c   |  5 +++++
 drivers/ata/libata-pmp.c  | 35 -----------------------------------
 drivers/ata/sata_sil24.c  |  2 +-
 include/linux/libata.h    |  2 --
 5 files changed, 7 insertions(+), 40 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index e00b620f161a..2da579b46bdd 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -91,7 +91,7 @@ const struct ata_port_operations sata_pmp_port_ops = {
 	.inherits		= &sata_port_ops,
 
 	.pmp_prereset		= ata_std_prereset,
-	.pmp_hardreset		= sata_pmp_std_hardreset,
+	.pmp_hardreset		= sata_std_hardreset,
 	.pmp_postreset		= ata_std_postreset,
 	.error_handler		= sata_pmp_error_handler,
 };
@@ -6307,7 +6307,6 @@ EXPORT_SYMBOL_GPL(ata_pci_device_resume);
 #endif /* CONFIG_PCI */
 
 EXPORT_SYMBOL_GPL(sata_pmp_qc_defer_cmd_switch);
-EXPORT_SYMBOL_GPL(sata_pmp_std_hardreset);
 EXPORT_SYMBOL_GPL(sata_pmp_error_handler);
 
 EXPORT_SYMBOL_GPL(__ata_ehi_push_desc);
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index 21687bbd9a70..d8c4a45dcf26 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -2276,6 +2276,11 @@ int ata_eh_reset(struct ata_link *link, int classify,
 	return rc;
 
  fail:
+	/* if SCR isn't accessible on a fan-out port, PMP needs to be reset */
+	if (!ata_is_host_link(link) &&
+	    sata_scr_read(link, SCR_STATUS, &sstatus))
+		rc = -ERESTART;
+
 	if (rc == -ERESTART || try >= max_tries)
 		goto out;
 
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index 2f8a9577c26d..9c998611b644 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -175,41 +175,6 @@ int sata_pmp_scr_write(struct ata_link *link, int reg, u32 val)
 	return 0;
 }
 
-/**
- *	sata_pmp_std_hardreset - standard hardreset method for PMP link
- *	@link: link to be reset
- *	@class: resulting class of attached device
- *	@deadline: deadline jiffies for the operation
- *
- *	Hardreset PMP port @link.  Note that this function doesn't
- *	wait for BSY clearance.  There simply isn't a generic way to
- *	wait the event.  Instead, this function return -EAGAIN thus
- *	telling libata-EH to followup with softreset.
- *
- *	LOCKING:
- *	Kernel thread context (may sleep)
- *
- *	RETURNS:
- *	0 on success, -errno otherwise.
- */
-int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
-			   unsigned long deadline)
-{
-	u32 tmp;
-	int rc;
-
-	DPRINTK("ENTER\n");
-
-	rc = sata_std_hardreset(link, class, deadline);
-
-	/* if SCR isn't accessible, we need to reset the PMP */
-	if (rc && rc != -EAGAIN && sata_scr_read(link, SCR_STATUS, &tmp))
-		rc = -ERESTART;
-
-	DPRINTK("EXIT, rc=%d\n", rc);
-	return rc;
-}
-
 /**
  *	sata_pmp_read_gscr - read GSCR block of SATA PMP
  *	@dev: PMP device
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index b83851f6e068..fc9d48cd8122 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -944,7 +944,7 @@ static int sil24_pmp_hardreset(struct ata_link *link, unsigned int *class,
 		return rc;
 	}
 
-	return sata_pmp_std_hardreset(link, class, deadline);
+	return sata_std_hardreset(link, class, deadline);
 }
 
 static void sil24_freeze(struct ata_port *ap)
diff --git a/include/linux/libata.h b/include/linux/libata.h
index b9188371b12a..2b5a0b77e179 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1025,8 +1025,6 @@ static inline int ata_acpi_cbl_80wire(struct ata_port *ap,
  * PMP - drivers/ata/libata-pmp.c
  */
 extern int sata_pmp_qc_defer_cmd_switch(struct ata_queued_cmd *qc);
-extern int sata_pmp_std_hardreset(struct ata_link *link, unsigned int *class,
-				  unsigned long deadline);
 extern void sata_pmp_error_handler(struct ata_port *ap);
 
 /*
-- 
cgit v1.2.3


From 22183bf569c8600ff414ac25f23134044e0ef453 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:20 +0900
Subject: libata: add qc_fill_rtf port operation

On command completion, ata_qc_complete() directly called ops->tf_read
to fill qc->result_tf.  This patch adds ops->qc_fill_rtf to replace
hardcoded ops->tf_read usage.

ata_sff_qc_fill_rtf() which uses ops->tf_read to fill result_tf is
implemented and set in ata_base_port_ops and other ops tables which
don't inherit from ata_base_port_ops, so this patch doesn't introduce
any behavior change.

ops->qc_fill_rtf() is similar to ops->sff_tf_read() but can only be
called when a command finishes.  As some non-SFF controllers don't
have TF registers defined unless they're associated with in-flight
commands, this limited operation makes life easier for those drivers
and help lifting SFF assumptions from libata core layer.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c     |  3 ++-
 drivers/ata/libata-sff.c      | 20 ++++++++++++++++++++
 drivers/ata/sata_sx4.c        |  1 +
 drivers/scsi/ipr.c            |  1 +
 drivers/scsi/libsas/sas_ata.c |  1 +
 include/linux/libata.h        |  2 ++
 6 files changed, 27 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 51876b93c1b7..3b822124e97e 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -74,6 +74,7 @@ const unsigned long sata_deb_timing_hotplug[]		= {  25,  500, 2000 };
 const unsigned long sata_deb_timing_long[]		= { 100, 2000, 5000 };
 
 const struct ata_port_operations ata_base_port_ops = {
+	.qc_fill_rtf		= ata_sff_qc_fill_rtf,
 	.prereset		= ata_std_prereset,
 	.postreset		= ata_std_postreset,
 	.error_handler		= ata_std_error_handler,
@@ -4562,7 +4563,7 @@ static void fill_result_tf(struct ata_queued_cmd *qc)
 	struct ata_port *ap = qc->ap;
 
 	qc->result_tf.flags = qc->tf.flags;
-	ap->ops->sff_tf_read(ap, &qc->result_tf);
+	ap->ops->qc_fill_rtf(qc);
 }
 
 static void ata_verify_xfer(struct ata_queued_cmd *qc)
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 5be8a6058dac..5ae813f54420 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -1407,6 +1407,25 @@ unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc)
 	return 0;
 }
 
+/**
+ *	ata_sff_qc_fill_rtf - fill result TF using ->sff_tf_read
+ *	@qc: qc to fill result TF for
+ *
+ *	@qc is finished and result TF needs to be filled.  Fill it
+ *	using ->sff_tf_read.
+ *
+ *	LOCKING:
+ *	spin_lock_irqsave(host lock)
+ *
+ *	RETURNS:
+ *	true indicating that result TF is successfully filled.
+ */
+bool ata_sff_qc_fill_rtf(struct ata_queued_cmd *qc)
+{
+	qc->ap->ops->sff_tf_read(qc->ap, &qc->result_tf);
+	return true;
+}
+
 /**
  *	ata_sff_host_intr - Handle host interrupt for given (port, task)
  *	@ap: Port on which interrupt arrived (possibly...)
@@ -2724,6 +2743,7 @@ EXPORT_SYMBOL_GPL(ata_sff_irq_on);
 EXPORT_SYMBOL_GPL(ata_sff_irq_clear);
 EXPORT_SYMBOL_GPL(ata_sff_hsm_move);
 EXPORT_SYMBOL_GPL(ata_sff_qc_issue);
+EXPORT_SYMBOL_GPL(ata_sff_qc_fill_rtf);
 EXPORT_SYMBOL_GPL(ata_sff_host_intr);
 EXPORT_SYMBOL_GPL(ata_sff_interrupt);
 EXPORT_SYMBOL_GPL(ata_sff_freeze);
diff --git a/drivers/ata/sata_sx4.c b/drivers/ata/sata_sx4.c
index 6107eff731f3..ec04b8d3c791 100644
--- a/drivers/ata/sata_sx4.c
+++ b/drivers/ata/sata_sx4.c
@@ -251,6 +251,7 @@ static struct ata_port_operations pdc_20621_ops = {
 	.phy_reset		= pdc_20621_phy_reset,
 	.qc_prep		= pdc20621_qc_prep,
 	.qc_issue		= pdc20621_qc_issue,
+	.qc_fill_rtf		= ata_sff_qc_fill_rtf,
 	.sff_data_xfer		= ata_sff_data_xfer,
 	.eng_timeout		= pdc_eng_timeout,
 	.sff_irq_clear		= pdc20621_irq_clear,
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index 5bddae19d1fb..2ecd32991522 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -5280,6 +5280,7 @@ static struct ata_port_operations ipr_sata_ops = {
 	.sff_tf_read = ipr_tf_read,
 	.qc_prep = ata_noop_qc_prep,
 	.qc_issue = ipr_qc_issue,
+	.qc_fill_rtf = ata_sff_qc_fill_rtf,
 	.port_start = ata_sas_port_start,
 	.port_stop = ata_sas_port_stop
 };
diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index a1664b87927a..2ec255839dcd 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -356,6 +356,7 @@ static struct ata_port_operations sas_sata_ops = {
 	.sff_tf_read		= sas_ata_tf_read,
 	.qc_prep		= ata_noop_qc_prep,
 	.qc_issue		= sas_ata_qc_issue,
+	.qc_fill_rtf		= ata_sff_qc_fill_rtf,
 	.port_start		= ata_sas_port_start,
 	.port_stop		= ata_sas_port_stop,
 	.scr_read		= sas_ata_scr_read,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 2b5a0b77e179..bb4200d42f0d 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -715,6 +715,7 @@ struct ata_port_operations {
 	int  (*check_atapi_dma)(struct ata_queued_cmd *qc);
 	void (*qc_prep)(struct ata_queued_cmd *qc);
 	unsigned int (*qc_issue)(struct ata_queued_cmd *qc);
+	bool (*qc_fill_rtf)(struct ata_queued_cmd *qc);
 
 	/*
 	 * Configuration and exception handling
@@ -1385,6 +1386,7 @@ extern void ata_sff_irq_clear(struct ata_port *ap);
 extern int ata_sff_hsm_move(struct ata_port *ap, struct ata_queued_cmd *qc,
 			    u8 status, int in_wq);
 extern unsigned int ata_sff_qc_issue(struct ata_queued_cmd *qc);
+extern bool ata_sff_qc_fill_rtf(struct ata_queued_cmd *qc);
 extern unsigned int ata_sff_host_intr(struct ata_port *ap,
 				      struct ata_queued_cmd *qc);
 extern irqreturn_t ata_sff_interrupt(int irq, void *dev_instance);
-- 
cgit v1.2.3


From 79f97dadfe9b4b561634d202225ba2fa910dc225 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:20 +0900
Subject: libata: drop @finish_qc from ata_qc_complete_multiple()

ata_qc_complete_multiple() took @finish_qc and called it on every qc
before completing it.  This was to give opportunity to update TF cache
before ata_qc_complete() tries to fill result_tf.  Now that result TF
is a separate operation, this is no longer necessary.

Update sata_sil24, which was the only user of this mechanism, such
that it implements its own ops->qc_fill_rtf() and drop @finish_qc from
ata_qc_complete_multiple().

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c        |  2 +-
 drivers/ata/libata-core.c |  6 +-----
 drivers/ata/sata_sil24.c  | 19 +++++++++----------
 include/linux/libata.h    |  3 +--
 4 files changed, 12 insertions(+), 18 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 3071a2341be3..1389c64e0027 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -1696,7 +1696,7 @@ static void ahci_port_intr(struct ata_port *ap)
 	else
 		qc_active = readl(port_mmio + PORT_CMD_ISSUE);
 
-	rc = ata_qc_complete_multiple(ap, qc_active, NULL);
+	rc = ata_qc_complete_multiple(ap, qc_active);
 
 	/* while resetting, invalid completions are expected */
 	if (unlikely(rc < 0 && !resetting)) {
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 3b822124e97e..dcc5a28e26c1 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -4673,7 +4673,6 @@ void ata_qc_complete(struct ata_queued_cmd *qc)
  *	ata_qc_complete_multiple - Complete multiple qcs successfully
  *	@ap: port in question
  *	@qc_active: new qc_active mask
- *	@finish_qc: LLDD callback invoked before completing a qc
  *
  *	Complete in-flight commands.  This functions is meant to be
  *	called from low-level driver's interrupt routine to complete
@@ -4686,8 +4685,7 @@ void ata_qc_complete(struct ata_queued_cmd *qc)
  *	RETURNS:
  *	Number of completed commands on success, -errno otherwise.
  */
-int ata_qc_complete_multiple(struct ata_port *ap, u32 qc_active,
-			     void (*finish_qc)(struct ata_queued_cmd *))
+int ata_qc_complete_multiple(struct ata_port *ap, u32 qc_active)
 {
 	int nr_done = 0;
 	u32 done_mask;
@@ -4708,8 +4706,6 @@ int ata_qc_complete_multiple(struct ata_port *ap, u32 qc_active,
 			continue;
 
 		if ((qc = ata_qc_from_tag(ap, i))) {
-			if (finish_qc)
-				finish_qc(qc);
 			ata_qc_complete(qc);
 			nr_done++;
 		}
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index fc9d48cd8122..79952f825b40 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -348,6 +348,7 @@ static void sil24_tf_read(struct ata_port *ap, struct ata_taskfile *tf);
 static int sil24_qc_defer(struct ata_queued_cmd *qc);
 static void sil24_qc_prep(struct ata_queued_cmd *qc);
 static unsigned int sil24_qc_issue(struct ata_queued_cmd *qc);
+static bool sil24_qc_fill_rtf(struct ata_queued_cmd *qc);
 static void sil24_pmp_attach(struct ata_port *ap);
 static void sil24_pmp_detach(struct ata_port *ap);
 static void sil24_freeze(struct ata_port *ap);
@@ -407,6 +408,7 @@ static struct ata_port_operations sil24_ops = {
 	.qc_defer		= sil24_qc_defer,
 	.qc_prep		= sil24_qc_prep,
 	.qc_issue		= sil24_qc_issue,
+	.qc_fill_rtf		= sil24_qc_fill_rtf,
 
 	.freeze			= sil24_freeze,
 	.thaw			= sil24_thaw,
@@ -914,6 +916,12 @@ static unsigned int sil24_qc_issue(struct ata_queued_cmd *qc)
 	return 0;
 }
 
+static bool sil24_qc_fill_rtf(struct ata_queued_cmd *qc)
+{
+	sil24_read_tf(qc->ap, qc->tag, &qc->result_tf);
+	return true;
+}
+
 static void sil24_pmp_attach(struct ata_port *ap)
 {
 	sil24_config_pmp(ap, 1);
@@ -1098,15 +1106,6 @@ static void sil24_error_intr(struct ata_port *ap)
 	}
 }
 
-static void sil24_finish_qc(struct ata_queued_cmd *qc)
-{
-	struct ata_port *ap = qc->ap;
-	struct sil24_port_priv *pp = ap->private_data;
-
-	if (qc->flags & ATA_QCFLAG_RESULT_TF)
-		sil24_read_tf(ap, qc->tag, &pp->tf);
-}
-
 static inline void sil24_host_intr(struct ata_port *ap)
 {
 	void __iomem *port = ap->ioaddr.cmd_addr;
@@ -1131,7 +1130,7 @@ static inline void sil24_host_intr(struct ata_port *ap)
 	}
 
 	qc_active = slot_stat & ~HOST_SSTAT_ATTN;
-	rc = ata_qc_complete_multiple(ap, qc_active, sil24_finish_qc);
+	rc = ata_qc_complete_multiple(ap, qc_active);
 	if (rc > 0)
 		return;
 	if (rc < 0) {
diff --git a/include/linux/libata.h b/include/linux/libata.h
index bb4200d42f0d..1d8b6b7de0a2 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -924,8 +924,7 @@ extern void ata_id_string(const u16 *id, unsigned char *s,
 extern void ata_id_c_string(const u16 *id, unsigned char *s,
 			    unsigned int ofs, unsigned int len);
 extern void ata_qc_complete(struct ata_queued_cmd *qc);
-extern int ata_qc_complete_multiple(struct ata_port *ap, u32 qc_active,
-				    void (*finish_qc)(struct ata_queued_cmd *));
+extern int ata_qc_complete_multiple(struct ata_port *ap, u32 qc_active);
 extern void ata_scsi_simulate(struct ata_device *dev, struct scsi_cmnd *cmd,
 			      void (*done)(struct scsi_cmnd *));
 extern int ata_std_bios_param(struct scsi_device *sdev,
-- 
cgit v1.2.3


From c9f75b04ed5ed65a058d18a8a8dda50632a96de8 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:21 +0900
Subject: libata: kill ata_noop_dev_select()

Now that SFF assumptions are separated out from non-SFF reset
sequence, port_ops->sff_dev_select() is no longer necessary for
non-SFF controllers.  Kill ata_noop_dev_select() and ->sff_dev_select
initialization from base and other non-SFF port_ops.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c     | 19 -------------------
 drivers/scsi/ipr.c            |  1 -
 drivers/scsi/libsas/sas_ata.c |  1 -
 include/linux/libata.h        |  1 -
 4 files changed, 22 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index e8598eeeec39..f23f50af551c 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -84,7 +84,6 @@ const struct ata_port_operations sata_port_ops = {
 
 	.qc_defer		= ata_std_qc_defer,
 	.hardreset		= sata_std_hardreset,
-	.sff_dev_select		= ata_noop_dev_select,
 };
 
 const struct ata_port_operations sata_pmp_port_ops = {
@@ -1431,22 +1430,6 @@ static int ata_hpa_resize(struct ata_device *dev)
 	return 0;
 }
 
-/**
- *	ata_noop_dev_select - Select device 0/1 on ATA bus
- *	@ap: ATA channel to manipulate
- *	@device: ATA device (numbered from zero) to select
- *
- *	This function performs no actual function.
- *
- *	May be used as the dev_select() entry in ata_port_operations.
- *
- *	LOCKING:
- *	caller.
- */
-void ata_noop_dev_select(struct ata_port *ap, unsigned int device)
-{
-}
-
 /**
  *	ata_dump_id - IDENTIFY DEVICE info debugging output
  *	@id: IDENTIFY DEVICE page to dump
@@ -6193,7 +6176,6 @@ static unsigned int ata_dummy_qc_issue(struct ata_queued_cmd *qc)
 struct ata_port_operations ata_dummy_port_ops = {
 	.sff_check_status	= ata_dummy_check_status,
 	.sff_check_altstatus	= ata_dummy_check_status,
-	.sff_dev_select		= ata_noop_dev_select,
 	.qc_prep		= ata_noop_qc_prep,
 	.qc_issue		= ata_dummy_qc_issue,
 	.freeze			= ata_dummy_noret,
@@ -6234,7 +6216,6 @@ EXPORT_SYMBOL_GPL(ata_host_detach);
 EXPORT_SYMBOL_GPL(ata_sg_init);
 EXPORT_SYMBOL_GPL(ata_qc_complete);
 EXPORT_SYMBOL_GPL(ata_qc_complete_multiple);
-EXPORT_SYMBOL_GPL(ata_noop_dev_select);
 EXPORT_SYMBOL_GPL(sata_print_link_status);
 EXPORT_SYMBOL_GPL(atapi_cmd_type);
 EXPORT_SYMBOL_GPL(ata_tf_to_fis);
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index 5a95ea7c4dba..65dc18dea845 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -5247,7 +5247,6 @@ static bool ipr_qc_fill_rtf(struct ata_queued_cmd *qc)
 }
 
 static struct ata_port_operations ipr_sata_ops = {
-	.sff_dev_select = ata_noop_dev_select,
 	.phy_reset = ipr_ata_phy_reset,
 	.hardreset = ipr_sata_reset,
 	.post_internal_cmd = ipr_ata_post_internal,
diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index 5c114bb7e907..a4811e4106df 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -344,7 +344,6 @@ static int sas_ata_scr_read(struct ata_port *ap, unsigned int sc_reg_in,
 }
 
 static struct ata_port_operations sas_sata_ops = {
-	.sff_dev_select		= ata_noop_dev_select,
 	.phy_reset		= sas_ata_phy_reset,
 	.post_internal_cmd	= sas_ata_post_internal,
 	.qc_prep		= ata_noop_qc_prep,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 1d8b6b7de0a2..037db1883bae 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -911,7 +911,6 @@ extern unsigned long ata_xfer_mode2mask(u8 xfer_mode);
 extern int ata_xfer_mode2shift(unsigned long xfer_mode);
 extern const char *ata_mode_string(unsigned long xfer_mask);
 extern unsigned long ata_id_xfermask(const u16 *id);
-extern void ata_noop_dev_select(struct ata_port *ap, unsigned int device);
 extern int ata_port_start(struct ata_port *ap);
 extern int ata_std_qc_defer(struct ata_queued_cmd *qc);
 extern void ata_noop_qc_prep(struct ata_queued_cmd *qc);
-- 
cgit v1.2.3


From 127102aea2ea9ec4e9ca233e2b1a75c8d3b058c4 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:21 +0900
Subject: libata: make SFF support optional

Now that SFF support is completely separated out from the core layer,
it can be made optional.  Add CONFIG_ATA_SFF and let SFF drivers
depend on it.  If CONFIG_ATA_SFF isn't set, all codes in libata-sff.c
and data structures for SFF support are disabled.  This saves good
number of bytes for small systems.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/Kconfig       | 44 +++++++++++++++++++++++++++-----------------
 drivers/ata/Makefile      |  4 ++--
 drivers/ata/libata-core.c |  2 ++
 drivers/ata/libata-scsi.c |  2 ++
 drivers/ata/libata.h      |  4 ++++
 include/linux/libata.h    |  9 +++++++++
 6 files changed, 46 insertions(+), 19 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig
index ea665c249035..48c8fc55391e 100644
--- a/drivers/ata/Kconfig
+++ b/drivers/ata/Kconfig
@@ -49,6 +49,32 @@ config SATA_AHCI
 
 	  If unsure, say N.
 
+config SATA_SIL24
+	tristate "Silicon Image 3124/3132 SATA support"
+	depends on PCI
+	help
+	  This option enables support for Silicon Image 3124/3132 Serial ATA.
+
+	  If unsure, say N.
+
+config SATA_FSL
+	tristate "Freescale 3.0Gbps SATA support"
+	depends on FSL_SOC
+	help
+	  This option enables support for Freescale 3.0Gbps SATA controller.
+	  It can be found on MPC837x and MPC8315.
+
+	  If unsure, say N.
+
+config ATA_SFF
+	bool "ATA SFF support"
+	default y
+	help
+	  This option adds support for ATA controllers with SFF
+	  compliant or similar programming interface.
+
+if ATA_SFF
+
 config SATA_SVW
 	tristate "ServerWorks Frodo / Apple K2 SATA support"
 	depends on PCI
@@ -125,14 +151,6 @@ config SATA_SIL
 
 	  If unsure, say N.
 
-config SATA_SIL24
-	tristate "Silicon Image 3124/3132 SATA support"
-	depends on PCI
-	help
-	  This option enables support for Silicon Image 3124/3132 Serial ATA.
-
-	  If unsure, say N.
-
 config SATA_SIS
 	tristate "SiS 964/965/966/180 SATA support"
 	depends on PCI
@@ -183,15 +201,6 @@ config PATA_ACPI
 	  firmware in the BIOS. This driver can sometimes handle
 	  otherwise unsupported hardware.
 
-config SATA_FSL
-	tristate "Freescale 3.0Gbps SATA support"
-	depends on FSL_SOC
-	help
-	  This option enables support for Freescale 3.0Gbps SATA controller.
-	  It can be found on MPC837x and MPC8315.
-
-	  If unsure, say N.
-
 config PATA_ALI
 	tristate "ALi PATA support (Experimental)"
 	depends on PCI && EXPERIMENTAL
@@ -679,4 +688,5 @@ config PATA_BF54X
 
 	  If unsure, say N.
 
+endif # ATA_SFF
 endif # ATA
diff --git a/drivers/ata/Makefile b/drivers/ata/Makefile
index 0511e6f0bb58..e6e41b2c731c 100644
--- a/drivers/ata/Makefile
+++ b/drivers/ata/Makefile
@@ -78,6 +78,6 @@ obj-$(CONFIG_ATA_GENERIC)	+= ata_generic.o
 # Should be last libata driver
 obj-$(CONFIG_PATA_LEGACY)	+= pata_legacy.o
 
-libata-objs	:= libata-core.o libata-scsi.o libata-sff.o libata-eh.o \
-		   libata-pmp.o
+libata-objs	:= libata-core.o libata-scsi.o libata-eh.o libata-pmp.o
+libata-$(CONFIG_ATA_SFF)	+= libata-sff.o
 libata-$(CONFIG_ATA_ACPI)	+= libata-acpi.o
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 3a94c69c7fe7..ca60af0cb051 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -5195,7 +5195,9 @@ struct ata_port *ata_port_alloc(struct ata_host *host)
 	ap->msg_enable = ATA_MSG_DRV | ATA_MSG_ERR | ATA_MSG_WARN;
 #endif
 
+#ifdef CONFIG_ATA_SFF
 	INIT_DELAYED_WORK(&ap->port_task, ata_pio_task);
+#endif
 	INIT_DELAYED_WORK(&ap->hotplug_task, ata_scsi_hotplug);
 	INIT_WORK(&ap->scsi_rescan_task, ata_scsi_dev_rescan);
 	INIT_LIST_HEAD(&ap->eh_done_q);
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index f8be92836a6e..a70881c408e5 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -2393,7 +2393,9 @@ static void atapi_request_sense(struct ata_queued_cmd *qc)
 	/* FIXME: is this needed? */
 	memset(cmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE);
 
+#ifdef CONFIG_ATA_SFF
 	ap->ops->sff_tf_read(ap, &qc->tf);
+#endif
 
 	/* fill these in, for the case where they are -not- overwritten */
 	cmd->sense_buffer[0] = 0x70;
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 87f54a1db3b0..6b70a624828c 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -42,8 +42,10 @@ static inline int ata_is_builtin_hardreset(ata_reset_fn_t reset)
 {
 	if (reset == sata_std_hardreset)
 		return 1;
+#ifdef CONFIG_ATA_SFF
 	if (reset == sata_sff_hardreset)
 		return 1;
+#endif
 	return 0;
 }
 
@@ -206,9 +208,11 @@ extern int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 extern void ata_eh_finish(struct ata_port *ap);
 
 /* libata-sff.c */
+#ifdef CONFIG_ATA_SFF
 extern void ata_dev_select(struct ata_port *ap, unsigned int device,
                            unsigned int wait, unsigned int can_sleep);
 extern u8 ata_irq_on(struct ata_port *ap);
 extern void ata_pio_task(struct work_struct *work);
+#endif /* CONFIG_ATA_SFF */
 
 #endif /* __LIBATA_H__ */
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 037db1883bae..db77b90003fd 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -445,6 +445,7 @@ enum link_pm {
 };
 extern struct class_device_attribute class_device_attr_link_power_management_policy;
 
+#ifdef CONFIG_ATA_SFF
 struct ata_ioports {
 	void __iomem		*cmd_addr;
 	void __iomem		*data_addr;
@@ -462,6 +463,7 @@ struct ata_ioports {
 	void __iomem		*bmdma_addr;
 	void __iomem		*scr_addr;
 };
+#endif /* CONFIG_ATA_SFF */
 
 struct ata_host {
 	spinlock_t		lock;
@@ -648,7 +650,9 @@ struct ata_port {
 	struct ata_prd		*prd;	 /* our SG list */
 	dma_addr_t		prd_dma; /* and its DMA mapping */
 
+#ifdef CONFIG_ATA_SFF
 	struct ata_ioports	ioaddr;	/* ATA cmd/ctl/dma register blocks */
+#endif /* CONFIG_ATA_SFF */
 
 	u8			ctl;	/* cache of ATA control register */
 	u8			last_ctl;	/* Cache last written value */
@@ -760,6 +764,7 @@ struct ata_port_operations {
 	void (*port_stop)(struct ata_port *ap);
 	void (*host_stop)(struct ata_host *host);
 
+#ifdef CONFIG_ATA_SFF
 	/*
 	 * SFF / taskfile oriented ops
 	 */
@@ -779,6 +784,7 @@ struct ata_port_operations {
 	void (*bmdma_start)(struct ata_queued_cmd *qc);
 	void (*bmdma_stop)(struct ata_queued_cmd *qc);
 	u8   (*bmdma_status)(struct ata_port *ap);
+#endif /* CONFIG_ATA_SFF */
 
 	/*
 	 * Obsolete
@@ -1349,6 +1355,8 @@ static inline struct ata_port *ata_shost_to_port(struct Scsi_Host *host)
 /**************************************************************************
  * SFF - drivers/ata/libata-sff.c
  */
+#ifdef CONFIG_ATA_SFF
+
 extern const struct ata_port_operations ata_sff_port_ops;
 extern const struct ata_port_operations ata_bmdma_port_ops;
 
@@ -1489,5 +1497,6 @@ static inline u8 ata_wait_idle(struct ata_port *ap)
 
 	return status;
 }
+#endif /* CONFIG_ATA_SFF */
 
 #endif /* __LINUX_LIBATA_H__ */
-- 
cgit v1.2.3


From 48515f6c006c2a9d7b624ee8ad068018c2d3fe0e Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:21 +0900
Subject: libata: separate PMP support code from core code

Most of PMP support code is already in libata-pmp.c.  All that are in
libata-core.c are sata_pmp_port_ops and EXPORTs.  Move them to
libata-pmp.c.  Also, collect PMP related prototypes and declarations
in header files and move them right above of SFF stuff.

This change is to make PMP support optional.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/libata-core.c | 13 -------------
 drivers/ata/libata-pmp.c  | 12 ++++++++++++
 drivers/ata/libata.h      | 10 +++++-----
 include/linux/libata.h    | 17 ++++++++++-------
 4 files changed, 27 insertions(+), 25 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index ca60af0cb051..b2d5d63fb6c9 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -86,15 +86,6 @@ const struct ata_port_operations sata_port_ops = {
 	.hardreset		= sata_std_hardreset,
 };
 
-const struct ata_port_operations sata_pmp_port_ops = {
-	.inherits		= &sata_port_ops,
-
-	.pmp_prereset		= ata_std_prereset,
-	.pmp_hardreset		= sata_std_hardreset,
-	.pmp_postreset		= ata_std_postreset,
-	.error_handler		= sata_pmp_error_handler,
-};
-
 static unsigned int ata_dev_init_params(struct ata_device *dev,
 					u16 heads, u16 sectors);
 static unsigned int ata_dev_set_xfermode(struct ata_device *dev);
@@ -6192,7 +6183,6 @@ EXPORT_SYMBOL_GPL(sata_deb_timing_hotplug);
 EXPORT_SYMBOL_GPL(sata_deb_timing_long);
 EXPORT_SYMBOL_GPL(ata_base_port_ops);
 EXPORT_SYMBOL_GPL(sata_port_ops);
-EXPORT_SYMBOL_GPL(sata_pmp_port_ops);
 EXPORT_SYMBOL_GPL(ata_dummy_port_ops);
 EXPORT_SYMBOL_GPL(ata_dummy_port_info);
 EXPORT_SYMBOL_GPL(ata_std_bios_param);
@@ -6272,9 +6262,6 @@ EXPORT_SYMBOL_GPL(ata_pci_device_resume);
 #endif /* CONFIG_PM */
 #endif /* CONFIG_PCI */
 
-EXPORT_SYMBOL_GPL(sata_pmp_qc_defer_cmd_switch);
-EXPORT_SYMBOL_GPL(sata_pmp_error_handler);
-
 EXPORT_SYMBOL_GPL(__ata_ehi_push_desc);
 EXPORT_SYMBOL_GPL(ata_ehi_push_desc);
 EXPORT_SYMBOL_GPL(ata_ehi_clear_desc);
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index 9c998611b644..bb10c0630791 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -11,6 +11,14 @@
 #include <linux/libata.h>
 #include "libata.h"
 
+const struct ata_port_operations sata_pmp_port_ops = {
+	.inherits		= &sata_port_ops,
+	.pmp_prereset		= ata_std_prereset,
+	.pmp_hardreset		= sata_std_hardreset,
+	.pmp_postreset		= ata_std_postreset,
+	.error_handler		= sata_pmp_error_handler,
+};
+
 /**
  *	sata_pmp_read - read PMP register
  *	@link: link to read PMP register for
@@ -1012,3 +1020,7 @@ void sata_pmp_error_handler(struct ata_port *ap)
 	sata_pmp_eh_recover(ap);
 	ata_eh_finish(ap);
 }
+
+EXPORT_SYMBOL_GPL(sata_pmp_port_ops);
+EXPORT_SYMBOL_GPL(sata_pmp_qc_defer_cmd_switch);
+EXPORT_SYMBOL_GPL(sata_pmp_error_handler);
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 6b70a624828c..42b30e38495d 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -179,11 +179,6 @@ extern void ata_schedule_scsi_eh(struct Scsi_Host *shost);
 extern void ata_scsi_dev_rescan(struct work_struct *work);
 extern int ata_bus_probe(struct ata_port *ap);
 
-/* libata-pmp.c */
-extern int sata_pmp_scr_read(struct ata_link *link, int reg, u32 *val);
-extern int sata_pmp_scr_write(struct ata_link *link, int reg, u32 val);
-extern int sata_pmp_attach(struct ata_device *dev);
-
 /* libata-eh.c */
 extern enum scsi_eh_timer_return ata_scsi_timed_out(struct scsi_cmnd *cmd);
 extern void ata_scsi_error(struct Scsi_Host *host);
@@ -207,6 +202,11 @@ extern int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 			  struct ata_link **r_failed_disk);
 extern void ata_eh_finish(struct ata_port *ap);
 
+/* libata-pmp.c */
+extern int sata_pmp_scr_read(struct ata_link *link, int reg, u32 *val);
+extern int sata_pmp_scr_write(struct ata_link *link, int reg, u32 val);
+extern int sata_pmp_attach(struct ata_device *dev);
+
 /* libata-sff.c */
 #ifdef CONFIG_ATA_SFF
 extern void ata_dev_select(struct ata_port *ap, unsigned int device,
diff --git a/include/linux/libata.h b/include/linux/libata.h
index db77b90003fd..eb86d6f39635 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1026,12 +1026,6 @@ static inline int ata_acpi_cbl_80wire(struct ata_port *ap,
 }
 #endif
 
-/*
- * PMP - drivers/ata/libata-pmp.c
- */
-extern int sata_pmp_qc_defer_cmd_switch(struct ata_queued_cmd *qc);
-extern void sata_pmp_error_handler(struct ata_port *ap);
-
 /*
  * EH - drivers/ata/libata-eh.c
  */
@@ -1075,7 +1069,6 @@ extern void ata_std_error_handler(struct ata_port *ap);
  */
 extern const struct ata_port_operations ata_base_port_ops;
 extern const struct ata_port_operations sata_port_ops;
-extern const struct ata_port_operations sata_pmp_port_ops;
 
 #define ATA_BASE_SHT(drv_name)					\
 	.module			= THIS_MODULE,			\
@@ -1352,6 +1345,16 @@ static inline struct ata_port *ata_shost_to_port(struct Scsi_Host *host)
 	return *(struct ata_port **)&host->hostdata[0];
 }
 
+
+/**************************************************************************
+ * PMP - drivers/ata/libata-pmp.c
+ */
+extern const struct ata_port_operations sata_pmp_port_ops;
+
+extern int sata_pmp_qc_defer_cmd_switch(struct ata_queued_cmd *qc);
+extern void sata_pmp_error_handler(struct ata_port *ap);
+
+
 /**************************************************************************
  * SFF - drivers/ata/libata-sff.c
  */
-- 
cgit v1.2.3


From 071f44b1d2c051641b62a3571223314737ccbe59 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:22 +0900
Subject: libata: implement PMP helpers

Implement helpers to test whether PMP is supported, attached and
determine pmp number to use when issuing SRST to a link.  While at it,
move ata_is_host_link() so that it's together with the two new PMP
helpers.

This change simplifies LLDs and helps making PMP support optional.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/ahci.c        | 30 ++++++------------------------
 drivers/ata/libata-acpi.c |  2 +-
 drivers/ata/libata-core.c |  4 ++--
 drivers/ata/libata-eh.c   | 14 +++++++-------
 drivers/ata/libata-pmp.c  |  6 +++---
 drivers/ata/libata-scsi.c |  6 +++---
 drivers/ata/sata_sil24.c  | 27 +++++++--------------------
 include/linux/libata.h    | 38 +++++++++++++++++++++++++++++---------
 8 files changed, 58 insertions(+), 69 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 6281f7f9eae6..0de6432ee026 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -260,8 +260,6 @@ static int ahci_vt8251_hardreset(struct ata_link *link, unsigned int *class,
 static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class,
 				unsigned long deadline);
 static void ahci_postreset(struct ata_link *link, unsigned int *class);
-static int ahci_pmp_softreset(struct ata_link *link, unsigned int *class,
-			      unsigned long deadline);
 static void ahci_error_handler(struct ata_port *ap);
 static void ahci_post_internal_cmd(struct ata_queued_cmd *qc);
 static int ahci_port_resume(struct ata_port *ap);
@@ -301,7 +299,7 @@ static struct ata_port_operations ahci_ops = {
 	.softreset		= ahci_softreset,
 	.hardreset		= ahci_hardreset,
 	.postreset		= ahci_postreset,
-	.pmp_softreset		= ahci_pmp_softreset,
+	.pmp_softreset		= ahci_softreset,
 	.error_handler		= ahci_error_handler,
 	.post_internal_cmd	= ahci_post_internal_cmd,
 	.dev_config		= ahci_dev_config,
@@ -1263,10 +1261,11 @@ static int ahci_check_ready(struct ata_link *link)
 	return 0;
 }
 
-static int ahci_do_softreset(struct ata_link *link, unsigned int *class,
-			     int pmp, unsigned long deadline)
+static int ahci_softreset(struct ata_link *link, unsigned int *class,
+			  unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
+	int pmp = sata_srst_pmp(link);
 	const char *reason = NULL;
 	unsigned long now, msecs;
 	struct ata_taskfile tf;
@@ -1326,17 +1325,6 @@ static int ahci_do_softreset(struct ata_link *link, unsigned int *class,
 	return rc;
 }
 
-static int ahci_softreset(struct ata_link *link, unsigned int *class,
-			  unsigned long deadline)
-{
-	int pmp = 0;
-
-	if (link->ap->flags & ATA_FLAG_PMP)
-		pmp = SATA_PMP_CTRL_PORT;
-
-	return ahci_do_softreset(link, class, pmp, deadline);
-}
-
 static int ahci_hardreset(struct ata_link *link, unsigned int *class,
 			  unsigned long deadline)
 {
@@ -1457,12 +1445,6 @@ static void ahci_postreset(struct ata_link *link, unsigned int *class)
 	}
 }
 
-static int ahci_pmp_softreset(struct ata_link *link, unsigned int *class,
-			      unsigned long deadline)
-{
-	return ahci_do_softreset(link, class, link->pmp, deadline);
-}
-
 static unsigned int ahci_fill_sg(struct ata_queued_cmd *qc, void *cmd_tbl)
 {
 	struct scatterlist *sg;
@@ -1581,7 +1563,7 @@ static void ahci_error_intr(struct ata_port *ap, u32 irq_stat)
 				  unk[0], unk[1], unk[2], unk[3]);
 	}
 
-	if (ap->nr_pmp_links && (irq_stat & PORT_IRQ_BAD_PMP)) {
+	if (sata_pmp_attached(ap) && (irq_stat & PORT_IRQ_BAD_PMP)) {
 		active_ehi->err_mask |= AC_ERR_HSM;
 		active_ehi->action |= ATA_EH_RESET;
 		ata_ehi_push_desc(active_ehi, "incorrect PMP");
@@ -1847,7 +1829,7 @@ static int ahci_port_resume(struct ata_port *ap)
 	ahci_power_up(ap);
 	ahci_start_port(ap);
 
-	if (ap->nr_pmp_links)
+	if (sata_pmp_attached(ap))
 		ahci_pmp_attach(ap);
 	else
 		ahci_pmp_detach(ap);
diff --git a/drivers/ata/libata-acpi.c b/drivers/ata/libata-acpi.c
index bf98a566adac..f88a4f940e11 100644
--- a/drivers/ata/libata-acpi.c
+++ b/drivers/ata/libata-acpi.c
@@ -77,7 +77,7 @@ void ata_acpi_associate_sata_port(struct ata_port *ap)
 {
 	WARN_ON(!(ap->flags & ATA_FLAG_ACPI_SATA));
 
-	if (!ap->nr_pmp_links) {
+	if (!sata_pmp_attached(ap)) {
 		acpi_integer adr = SATA_ADR(ap->port_no, NO_PORT_MULT);
 
 		ap->link.device->acpi_handle =
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index b2d5d63fb6c9..3401248180c9 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -2278,7 +2278,7 @@ int ata_dev_configure(struct ata_device *dev)
 		 * changed notifications and ATAPI ANs.
 		 */
 		if ((ap->flags & ATA_FLAG_AN) && ata_id_has_atapi_AN(id) &&
-		    (!ap->nr_pmp_links ||
+		    (!sata_pmp_attached(ap) ||
 		     sata_scr_read(&ap->link, SCR_NOTIFICATION, &sntf) == 0)) {
 			unsigned int err_mask;
 
@@ -3623,7 +3623,7 @@ int sata_link_hardreset(struct ata_link *link, const unsigned long *timing,
 	if (online)
 		*online = true;
 
-	if ((link->ap->flags & ATA_FLAG_PMP) && ata_is_host_link(link)) {
+	if (sata_pmp_supported(link->ap) && ata_is_host_link(link)) {
 		/* If PMP is supported, we have to do follow-up SRST.
 		 * Some PMPs don't send D2H Reg FIS after hardreset if
 		 * the first port is empty.  Wait only for
diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c
index 4ec1397434c0..99f83bdc572b 100644
--- a/drivers/ata/libata-eh.c
+++ b/drivers/ata/libata-eh.c
@@ -873,9 +873,9 @@ int sata_async_notification(struct ata_port *ap)
 	if (rc == 0)
 		sata_scr_write(&ap->link, SCR_NOTIFICATION, sntf);
 
-	if (!ap->nr_pmp_links || rc) {
+	if (!sata_pmp_attached(ap) || rc) {
 		/* PMP is not attached or SNTF is not available */
-		if (!ap->nr_pmp_links) {
+		if (!sata_pmp_attached(ap)) {
 			/* PMP is not attached.  Check whether ATAPI
 			 * AN is configured.  If so, notify media
 			 * change.
@@ -1853,7 +1853,7 @@ void ata_eh_autopsy(struct ata_port *ap)
 	/* Autopsy of fanout ports can affect host link autopsy.
 	 * Perform host link autopsy last.
 	 */
-	if (ap->nr_pmp_links)
+	if (sata_pmp_attached(ap))
 		ata_eh_link_autopsy(&ap->link);
 }
 
@@ -2076,7 +2076,7 @@ static int ata_eh_followup_srst_needed(struct ata_link *link,
 	}
 	if (rc != 0)
 		return 0;
-	if ((link->ap->flags & ATA_FLAG_PMP) && ata_is_host_link(link))
+	if (sata_pmp_supported(link->ap) && ata_is_host_link(link))
 		return 1;
 	return 0;
 }
@@ -2668,7 +2668,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 		/* if PMP is attached, this function only deals with
 		 * downstream links, port should stay thawed.
 		 */
-		if (!ap->nr_pmp_links)
+		if (!sata_pmp_attached(ap))
 			ata_eh_freeze_port(ap);
 
 		ata_port_for_each_link(link, ap) {
@@ -2687,7 +2687,7 @@ int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 			}
 		}
 
-		if (!ap->nr_pmp_links)
+		if (!sata_pmp_attached(ap))
 			ata_eh_thaw_port(ap);
 	}
 
@@ -2731,7 +2731,7 @@ dev_fail:
 			/* PMP reset requires working host port.
 			 * Can't retry if it's frozen.
 			 */
-			if (ap->nr_pmp_links)
+			if (sata_pmp_attached(ap))
 				goto out;
 			break;
 		}
diff --git a/drivers/ata/libata-pmp.c b/drivers/ata/libata-pmp.c
index bb10c0630791..ff1822a7da38 100644
--- a/drivers/ata/libata-pmp.c
+++ b/drivers/ata/libata-pmp.c
@@ -411,7 +411,7 @@ int sata_pmp_attach(struct ata_device *dev)
 	int rc;
 
 	/* is it hanging off the right place? */
-	if (!(ap->flags & ATA_FLAG_PMP)) {
+	if (!sata_pmp_supported(ap)) {
 		ata_dev_printk(dev, KERN_ERR,
 			       "host does not support Port Multiplier\n");
 		return -EINVAL;
@@ -876,7 +876,7 @@ static int sata_pmp_eh_recover(struct ata_port *ap)
 
  retry:
 	/* PMP attached? */
-	if (!ap->nr_pmp_links) {
+	if (!sata_pmp_attached(ap)) {
 		rc = ata_eh_recover(ap, ops->prereset, ops->softreset,
 				    ops->hardreset, ops->postreset, NULL);
 		if (rc) {
@@ -983,7 +983,7 @@ static int sata_pmp_eh_recover(struct ata_port *ap)
 	if (ap->pflags & ATA_PFLAG_UNLOADING)
 		return rc;
 
-	if (!ap->nr_pmp_links)
+	if (!sata_pmp_attached(ap))
 		goto retry;
 
 	if (--pmp_tries) {
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index a70881c408e5..fedf62de9460 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -2617,7 +2617,7 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc)
 
 static struct ata_device *ata_find_dev(struct ata_port *ap, int devno)
 {
-	if (ap->nr_pmp_links == 0) {
+	if (!sata_pmp_attached(ap)) {
 		if (likely(devno < ata_link_max_devices(&ap->link)))
 			return &ap->link.device[devno];
 	} else {
@@ -2634,7 +2634,7 @@ static struct ata_device *__ata_scsi_find_dev(struct ata_port *ap,
 	int devno;
 
 	/* skip commands not addressed to targets we simulate */
-	if (ap->nr_pmp_links == 0) {
+	if (!sata_pmp_attached(ap)) {
 		if (unlikely(scsidev->channel || scsidev->lun))
 			return NULL;
 		devno = scsidev->id;
@@ -3492,7 +3492,7 @@ static int ata_scsi_user_scan(struct Scsi_Host *shost, unsigned int channel,
 	if (lun != SCAN_WILD_CARD && lun)
 		return -EINVAL;
 
-	if (ap->nr_pmp_links == 0) {
+	if (!sata_pmp_attached(ap)) {
 		if (channel != SCAN_WILD_CARD && channel)
 			return -EINVAL;
 		devno = id;
diff --git a/drivers/ata/sata_sil24.c b/drivers/ata/sata_sil24.c
index 6039614e956c..068789361895 100644
--- a/drivers/ata/sata_sil24.c
+++ b/drivers/ata/sata_sil24.c
@@ -354,8 +354,6 @@ static int sil24_softreset(struct ata_link *link, unsigned int *class,
 			   unsigned long deadline);
 static int sil24_hardreset(struct ata_link *link, unsigned int *class,
 			   unsigned long deadline);
-static int sil24_pmp_softreset(struct ata_link *link, unsigned int *class,
-			       unsigned long deadline);
 static int sil24_pmp_hardreset(struct ata_link *link, unsigned int *class,
 			       unsigned long deadline);
 static void sil24_error_handler(struct ata_port *ap);
@@ -408,7 +406,7 @@ static struct ata_port_operations sil24_ops = {
 	.thaw			= sil24_thaw,
 	.softreset		= sil24_softreset,
 	.hardreset		= sil24_hardreset,
-	.pmp_softreset		= sil24_pmp_softreset,
+	.pmp_softreset		= sil24_softreset,
 	.pmp_hardreset		= sil24_pmp_hardreset,
 	.error_handler		= sil24_error_handler,
 	.post_internal_cmd	= sil24_post_internal_cmd,
@@ -588,7 +586,7 @@ static int sil24_init_port(struct ata_port *ap)
 	u32 tmp;
 
 	/* clear PMP error status */
-	if (ap->nr_pmp_links)
+	if (sata_pmp_attached(ap))
 		sil24_clear_pmp(ap);
 
 	writel(PORT_CS_INIT, port + PORT_CTRL_STAT);
@@ -653,10 +651,11 @@ static int sil24_exec_polled_cmd(struct ata_port *ap, int pmp,
 	return rc;
 }
 
-static int sil24_do_softreset(struct ata_link *link, unsigned int *class,
-			      int pmp, unsigned long deadline)
+static int sil24_softreset(struct ata_link *link, unsigned int *class,
+			   unsigned long deadline)
 {
 	struct ata_port *ap = link->ap;
+	int pmp = sata_srst_pmp(link);
 	unsigned long timeout_msec = 0;
 	struct ata_taskfile tf;
 	const char *reason;
@@ -706,12 +705,6 @@ static int sil24_do_softreset(struct ata_link *link, unsigned int *class,
 	return -EIO;
 }
 
-static int sil24_softreset(struct ata_link *link, unsigned int *class,
-			   unsigned long deadline)
-{
-	return sil24_do_softreset(link, class, SATA_PMP_CTRL_PORT, deadline);
-}
-
 static int sil24_hardreset(struct ata_link *link, unsigned int *class,
 			   unsigned long deadline)
 {
@@ -926,12 +919,6 @@ static void sil24_pmp_detach(struct ata_port *ap)
 	sil24_config_pmp(ap, 0);
 }
 
-static int sil24_pmp_softreset(struct ata_link *link, unsigned int *class,
-			       unsigned long deadline)
-{
-	return sil24_do_softreset(link, class, link->pmp, deadline);
-}
-
 static int sil24_pmp_hardreset(struct ata_link *link, unsigned int *class,
 			       unsigned long deadline)
 {
@@ -1034,7 +1021,7 @@ static void sil24_error_intr(struct ata_port *ap)
 		}
 
 		/* find out the offending link and qc */
-		if (ap->nr_pmp_links) {
+		if (sata_pmp_attached(ap)) {
 			context = readl(port + PORT_CONTEXT);
 			pmp = (context >> 5) & 0xf;
 
@@ -1082,7 +1069,7 @@ static void sil24_error_intr(struct ata_port *ap)
 		ehi->action |= action;
 
 		/* if PMP, resume */
-		if (ap->nr_pmp_links)
+		if (sata_pmp_attached(ap))
 			writel(PORT_CS_PMP_RESUME, port + PORT_CTRL_STAT);
 	}
 
diff --git a/include/linux/libata.h b/include/linux/libata.h
index eb86d6f39635..1908bf484743 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1089,6 +1089,31 @@ extern const struct ata_port_operations sata_port_ops;
 	ATA_BASE_SHT(drv_name),					\
 	.change_queue_depth	= ata_scsi_change_queue_depth
 
+/*
+ * PMP helpers
+ */
+static inline bool sata_pmp_supported(struct ata_port *ap)
+{
+	return ap->flags & ATA_FLAG_PMP;
+}
+
+static inline bool sata_pmp_attached(struct ata_port *ap)
+{
+	return ap->nr_pmp_links != 0;
+}
+
+static inline int ata_is_host_link(const struct ata_link *link)
+{
+	return link == &link->ap->link;
+}
+
+static inline int sata_srst_pmp(struct ata_link *link)
+{
+	if (sata_pmp_supported(link->ap) && ata_is_host_link(link))
+		return SATA_PMP_CTRL_PORT;
+	return link->pmp;
+}
+
 /*
  * printk helpers
  */
@@ -1096,7 +1121,7 @@ extern const struct ata_port_operations sata_port_ops;
 	printk("%sata%u: "fmt, lv, (ap)->print_id , ##args)
 
 #define ata_link_printk(link, lv, fmt, args...) do { \
-	if ((link)->ap->nr_pmp_links) \
+	if (sata_pmp_attached((link)->ap)) \
 		printk("%sata%u.%02u: "fmt, lv, (link)->ap->print_id,	\
 		       (link)->pmp , ##args); \
 	else \
@@ -1182,11 +1207,6 @@ static inline unsigned int ata_dev_absent(const struct ata_device *dev)
 /*
  * link helpers
  */
-static inline int ata_is_host_link(const struct ata_link *link)
-{
-	return link == &link->ap->link;
-}
-
 static inline int ata_link_max_devices(const struct ata_link *link)
 {
 	if (ata_is_host_link(link) && link->ap->flags & ATA_FLAG_SLAVE_POSS)
@@ -1201,7 +1221,7 @@ static inline int ata_link_active(struct ata_link *link)
 
 static inline struct ata_link *ata_port_first_link(struct ata_port *ap)
 {
-	if (ap->nr_pmp_links)
+	if (sata_pmp_attached(ap))
 		return ap->pmp_link;
 	return &ap->link;
 }
@@ -1210,8 +1230,8 @@ static inline struct ata_link *ata_port_next_link(struct ata_link *link)
 {
 	struct ata_port *ap = link->ap;
 
-	if (link == &ap->link) {
-		if (!ap->nr_pmp_links)
+	if (ata_is_host_link(link)) {
+		if (!sata_pmp_attached(ap))
 			return NULL;
 		return ap->pmp_link;
 	}
-- 
cgit v1.2.3


From 88fcd5627563722483427a55113c0a83f56e8080 Mon Sep 17 00:00:00 2001
From: Tejun Heo <htejun@gmail.com>
Date: Mon, 7 Apr 2008 22:47:22 +0900
Subject: libata: make PMP support optional

Make PMP support optional by adding CONFIG_SATA_PMP and leaving out
libata-pmp.c if it isn't set.  PMP helpers return constant values if
PMP support is not enabled and PMP declarations alias non-PMP
counterparts.  This makes the compiler to leave out PMP related part
out and LLDs to use non-PMP counterparts automatically.

Signed-off-by: Tejun Heo <htejun@gmail.com>
---
 drivers/ata/Kconfig    |  6 ++++++
 drivers/ata/Makefile   |  3 ++-
 drivers/ata/libata.h   | 17 +++++++++++++++++
 include/linux/libata.h | 27 +++++++++++++++++++++++++++
 4 files changed, 52 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig
index 48c8fc55391e..3eb6035b61b8 100644
--- a/drivers/ata/Kconfig
+++ b/drivers/ata/Kconfig
@@ -41,6 +41,12 @@ config ATA_ACPI
 	  You can disable this at kernel boot time by using the
 	  option libata.noacpi=1
 
+config SATA_PMP
+	bool "SATA Port Multiplier support"
+	default y
+	help
+	  This option adds support for SATA Port Multipliers.
+
 config SATA_AHCI
 	tristate "AHCI SATA support"
 	depends on PCI
diff --git a/drivers/ata/Makefile b/drivers/ata/Makefile
index e6e41b2c731c..1fbc2aa648b7 100644
--- a/drivers/ata/Makefile
+++ b/drivers/ata/Makefile
@@ -78,6 +78,7 @@ obj-$(CONFIG_ATA_GENERIC)	+= ata_generic.o
 # Should be last libata driver
 obj-$(CONFIG_PATA_LEGACY)	+= pata_legacy.o
 
-libata-objs	:= libata-core.o libata-scsi.o libata-eh.o libata-pmp.o
+libata-objs	:= libata-core.o libata-scsi.o libata-eh.o
 libata-$(CONFIG_ATA_SFF)	+= libata-sff.o
+libata-$(CONFIG_SATA_PMP)	+= libata-pmp.o
 libata-$(CONFIG_ATA_ACPI)	+= libata-acpi.o
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 42b30e38495d..4aeeabb10a47 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -203,9 +203,26 @@ extern int ata_eh_recover(struct ata_port *ap, ata_prereset_fn_t prereset,
 extern void ata_eh_finish(struct ata_port *ap);
 
 /* libata-pmp.c */
+#ifdef CONFIG_SATA_PMP
 extern int sata_pmp_scr_read(struct ata_link *link, int reg, u32 *val);
 extern int sata_pmp_scr_write(struct ata_link *link, int reg, u32 val);
 extern int sata_pmp_attach(struct ata_device *dev);
+#else /* CONFIG_SATA_PMP */
+static inline int sata_pmp_scr_read(struct ata_link *link, int reg, u32 *val)
+{
+	return -EINVAL;
+}
+
+static inline int sata_pmp_scr_write(struct ata_link *link, int reg, u32 val)
+{
+	return -EINVAL;
+}
+
+static inline int sata_pmp_attach(struct ata_device *dev)
+{
+	return -EINVAL;
+}
+#endif /* CONFIG_SATA_PMP */
 
 /* libata-sff.c */
 #ifdef CONFIG_ATA_SFF
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 1908bf484743..165734a2dd47 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -1092,6 +1092,7 @@ extern const struct ata_port_operations sata_port_ops;
 /*
  * PMP helpers
  */
+#ifdef CONFIG_SATA_PMP
 static inline bool sata_pmp_supported(struct ata_port *ap)
 {
 	return ap->flags & ATA_FLAG_PMP;
@@ -1106,6 +1107,22 @@ static inline int ata_is_host_link(const struct ata_link *link)
 {
 	return link == &link->ap->link;
 }
+#else /* CONFIG_SATA_PMP */
+static inline bool sata_pmp_supported(struct ata_port *ap)
+{
+	return false;
+}
+
+static inline bool sata_pmp_attached(struct ata_port *ap)
+{
+	return false;
+}
+
+static inline int ata_is_host_link(const struct ata_link *link)
+{
+	return 1;
+}
+#endif /* CONFIG_SATA_PMP */
 
 static inline int sata_srst_pmp(struct ata_link *link)
 {
@@ -1369,11 +1386,21 @@ static inline struct ata_port *ata_shost_to_port(struct Scsi_Host *host)
 /**************************************************************************
  * PMP - drivers/ata/libata-pmp.c
  */
+#ifdef CONFIG_SATA_PMP
+
 extern const struct ata_port_operations sata_pmp_port_ops;
 
 extern int sata_pmp_qc_defer_cmd_switch(struct ata_queued_cmd *qc);
 extern void sata_pmp_error_handler(struct ata_port *ap);
 
+#else /* CONFIG_SATA_PMP */
+
+#define sata_pmp_port_ops		sata_port_ops
+#define sata_pmp_qc_defer_cmd_switch	ata_std_qc_defer
+#define sata_pmp_error_handler		ata_std_error_handler
+
+#endif /* CONFIG_SATA_PMP */
+
 
 /**************************************************************************
  * SFF - drivers/ata/libata-sff.c
-- 
cgit v1.2.3


From 19242d7233df7d658405d4b7ee1758d21414cfaa Mon Sep 17 00:00:00 2001
From: Dan Williams <dan.j.williams@intel.com>
Date: Thu, 17 Apr 2008 20:17:25 -0700
Subject: async_tx: fix multiple dependency submission

Shrink struct dma_async_tx_descriptor and introduce
async_tx_channel_switch to properly inject a channel switch interrupt in
the descriptor stream.  This simplifies the locking model as drivers no
longer need to handle dma_async_tx_descriptor.lock.

Acked-by: Shannon Nelson <shannon.nelson@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 crypto/async_tx/async_tx.c | 197 +++++++++++++++++++++++++++++++++++++--------
 drivers/dma/dmaengine.c    |   2 -
 drivers/dma/iop-adma.c     |   9 ++-
 include/linux/dmaengine.h  |   9 +--
 4 files changed, 170 insertions(+), 47 deletions(-)

(limited to 'include/linux')

diff --git a/crypto/async_tx/async_tx.c b/crypto/async_tx/async_tx.c
index 2be3bae89930..69756164b61d 100644
--- a/crypto/async_tx/async_tx.c
+++ b/crypto/async_tx/async_tx.c
@@ -89,13 +89,19 @@ dma_wait_for_async_tx(struct dma_async_tx_descriptor *tx)
 		iter = tx;
 
 		/* find the root of the unsubmitted dependency chain */
-		while (iter->cookie == -EBUSY) {
+		do {
 			parent = iter->parent;
-			if (parent && parent->cookie == -EBUSY)
-				iter = iter->parent;
-			else
+			if (!parent)
 				break;
-		}
+			else
+				iter = parent;
+		} while (parent);
+
+		/* there is a small window for ->parent == NULL and
+		 * ->cookie == -EBUSY
+		 */
+		while (iter->cookie == -EBUSY)
+			cpu_relax();
 
 		status = dma_sync_wait(iter->chan, iter->cookie);
 	} while (status == DMA_IN_PROGRESS || (iter != tx));
@@ -111,24 +117,33 @@ EXPORT_SYMBOL_GPL(dma_wait_for_async_tx);
 void
 async_tx_run_dependencies(struct dma_async_tx_descriptor *tx)
 {
-	struct dma_async_tx_descriptor *dep_tx, *_dep_tx;
-	struct dma_device *dev;
+	struct dma_async_tx_descriptor *next = tx->next;
 	struct dma_chan *chan;
 
-	list_for_each_entry_safe(dep_tx, _dep_tx, &tx->depend_list,
-		depend_node) {
-		chan = dep_tx->chan;
-		dev = chan->device;
-		/* we can't depend on ourselves */
-		BUG_ON(chan == tx->chan);
-		list_del(&dep_tx->depend_node);
-		tx->tx_submit(dep_tx);
-
-		/* we need to poke the engine as client code does not
-		 * know about dependency submission events
-		 */
-		dev->device_issue_pending(chan);
+	if (!next)
+		return;
+
+	tx->next = NULL;
+	chan = next->chan;
+
+	/* keep submitting up until a channel switch is detected
+	 * in that case we will be called again as a result of
+	 * processing the interrupt from async_tx_channel_switch
+	 */
+	while (next && next->chan == chan) {
+		struct dma_async_tx_descriptor *_next;
+
+		spin_lock_bh(&next->lock);
+		next->parent = NULL;
+		_next = next->next;
+		next->next = NULL;
+		spin_unlock_bh(&next->lock);
+
+		next->tx_submit(next);
+		next = _next;
 	}
+
+	chan->device->device_issue_pending(chan);
 }
 EXPORT_SYMBOL_GPL(async_tx_run_dependencies);
 
@@ -397,6 +412,92 @@ static void __exit async_tx_exit(void)
 }
 #endif
 
+
+/**
+ * async_tx_channel_switch - queue an interrupt descriptor with a dependency
+ * 	pre-attached.
+ * @depend_tx: the operation that must finish before the new operation runs
+ * @tx: the new operation
+ */
+static void
+async_tx_channel_switch(struct dma_async_tx_descriptor *depend_tx,
+			struct dma_async_tx_descriptor *tx)
+{
+	struct dma_chan *chan;
+	struct dma_device *device;
+	struct dma_async_tx_descriptor *intr_tx = (void *) ~0;
+
+	/* first check to see if we can still append to depend_tx */
+	spin_lock_bh(&depend_tx->lock);
+	if (depend_tx->parent && depend_tx->chan == tx->chan) {
+		tx->parent = depend_tx;
+		depend_tx->next = tx;
+		intr_tx = NULL;
+	}
+	spin_unlock_bh(&depend_tx->lock);
+
+	if (!intr_tx)
+		return;
+
+	chan = depend_tx->chan;
+	device = chan->device;
+
+	/* see if we can schedule an interrupt
+	 * otherwise poll for completion
+	 */
+	if (dma_has_cap(DMA_INTERRUPT, device->cap_mask))
+		intr_tx = device->device_prep_dma_interrupt(chan);
+	else
+		intr_tx = NULL;
+
+	if (intr_tx) {
+		intr_tx->callback = NULL;
+		intr_tx->callback_param = NULL;
+		tx->parent = intr_tx;
+		/* safe to set ->next outside the lock since we know we are
+		 * not submitted yet
+		 */
+		intr_tx->next = tx;
+
+		/* check if we need to append */
+		spin_lock_bh(&depend_tx->lock);
+		if (depend_tx->parent) {
+			intr_tx->parent = depend_tx;
+			depend_tx->next = intr_tx;
+			async_tx_ack(intr_tx);
+			intr_tx = NULL;
+		}
+		spin_unlock_bh(&depend_tx->lock);
+
+		if (intr_tx) {
+			intr_tx->parent = NULL;
+			intr_tx->tx_submit(intr_tx);
+			async_tx_ack(intr_tx);
+		}
+	} else {
+		if (dma_wait_for_async_tx(depend_tx) == DMA_ERROR)
+			panic("%s: DMA_ERROR waiting for depend_tx\n",
+			      __func__);
+		tx->tx_submit(tx);
+	}
+}
+
+
+/**
+ * submit_disposition - while holding depend_tx->lock we must avoid submitting
+ * 	new operations to prevent a circular locking dependency with
+ * 	drivers that already hold a channel lock when calling
+ * 	async_tx_run_dependencies.
+ * @ASYNC_TX_SUBMITTED: we were able to append the new operation under the lock
+ * @ASYNC_TX_CHANNEL_SWITCH: when the lock is dropped schedule a channel switch
+ * @ASYNC_TX_DIRECT_SUBMIT: when the lock is dropped submit directly
+ */
+enum submit_disposition {
+	ASYNC_TX_SUBMITTED,
+	ASYNC_TX_CHANNEL_SWITCH,
+	ASYNC_TX_DIRECT_SUBMIT,
+};
+
 void
 async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx,
 	enum async_tx_flags flags, struct dma_async_tx_descriptor *depend_tx,
@@ -405,28 +506,54 @@ async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx,
 	tx->callback = cb_fn;
 	tx->callback_param = cb_param;
 
-	/* set this new tx to run after depend_tx if:
-	 * 1/ a dependency exists (depend_tx is !NULL)
-	 * 2/ the tx can not be submitted to the current channel
-	 */
-	if (depend_tx && depend_tx->chan != chan) {
-		/* if ack is already set then we cannot be sure
+	if (depend_tx) {
+		enum submit_disposition s;
+
+		/* sanity check the dependency chain:
+		 * 1/ if ack is already set then we cannot be sure
 		 * we are referring to the correct operation
+		 * 2/ dependencies are 1:1 i.e. two transactions can
+		 * not depend on the same parent
 		 */
-		BUG_ON(depend_tx->ack);
+		BUG_ON(depend_tx->ack || depend_tx->next || tx->parent);
 
-		tx->parent = depend_tx;
+		/* the lock prevents async_tx_run_dependencies from missing
+		 * the setting of ->next when ->parent != NULL
+		 */
 		spin_lock_bh(&depend_tx->lock);
-		list_add_tail(&tx->depend_node, &depend_tx->depend_list);
-		if (depend_tx->cookie == 0) {
-			struct dma_chan *dep_chan = depend_tx->chan;
-			struct dma_device *dep_dev = dep_chan->device;
-			dep_dev->device_dependency_added(dep_chan);
+		if (depend_tx->parent) {
+			/* we have a parent so we can not submit directly
+			 * if we are staying on the same channel: append
+			 * else: channel switch
+			 */
+			if (depend_tx->chan == chan) {
+				tx->parent = depend_tx;
+				depend_tx->next = tx;
+				s = ASYNC_TX_SUBMITTED;
+			} else
+				s = ASYNC_TX_CHANNEL_SWITCH;
+		} else {
+			/* we do not have a parent so we may be able to submit
+			 * directly if we are staying on the same channel
+			 */
+			if (depend_tx->chan == chan)
+				s = ASYNC_TX_DIRECT_SUBMIT;
+			else
+				s = ASYNC_TX_CHANNEL_SWITCH;
 		}
 		spin_unlock_bh(&depend_tx->lock);
 
-		/* schedule an interrupt to trigger the channel switch */
-		async_trigger_callback(ASYNC_TX_ACK, depend_tx, NULL, NULL);
+		switch (s) {
+		case ASYNC_TX_SUBMITTED:
+			break;
+		case ASYNC_TX_CHANNEL_SWITCH:
+			async_tx_channel_switch(depend_tx, tx);
+			break;
+		case ASYNC_TX_DIRECT_SUBMIT:
+			tx->parent = NULL;
+			tx->tx_submit(tx);
+			break;
+		}
 	} else {
 		tx->parent = NULL;
 		tx->tx_submit(tx);
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index 8db0e7f9d3f4..9cb898a76bb3 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -600,8 +600,6 @@ void dma_async_tx_descriptor_init(struct dma_async_tx_descriptor *tx,
 {
 	tx->chan = chan;
 	spin_lock_init(&tx->lock);
-	INIT_LIST_HEAD(&tx->depend_node);
-	INIT_LIST_HEAD(&tx->depend_list);
 }
 EXPORT_SYMBOL(dma_async_tx_descriptor_init);
 
diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c
index f82b0906d466..21854cd7190f 100644
--- a/drivers/dma/iop-adma.c
+++ b/drivers/dma/iop-adma.c
@@ -63,7 +63,6 @@ iop_adma_run_tx_complete_actions(struct iop_adma_desc_slot *desc,
 	struct iop_adma_chan *iop_chan, dma_cookie_t cookie)
 {
 	BUG_ON(desc->async_tx.cookie < 0);
-	spin_lock_bh(&desc->async_tx.lock);
 	if (desc->async_tx.cookie > 0) {
 		cookie = desc->async_tx.cookie;
 		desc->async_tx.cookie = 0;
@@ -101,7 +100,6 @@ iop_adma_run_tx_complete_actions(struct iop_adma_desc_slot *desc,
 
 	/* run dependent operations */
 	async_tx_run_dependencies(&desc->async_tx);
-	spin_unlock_bh(&desc->async_tx.lock);
 
 	return cookie;
 }
@@ -275,8 +273,11 @@ iop_adma_slot_cleanup(struct iop_adma_chan *iop_chan)
 
 static void iop_adma_tasklet(unsigned long data)
 {
-	struct iop_adma_chan *chan = (struct iop_adma_chan *) data;
-	__iop_adma_slot_cleanup(chan);
+	struct iop_adma_chan *iop_chan = (struct iop_adma_chan *) data;
+
+	spin_lock(&iop_chan->lock);
+	__iop_adma_slot_cleanup(iop_chan);
+	spin_unlock(&iop_chan->lock);
 }
 
 static struct iop_adma_desc_slot *
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index 34d440698293..91252a7e4d03 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -221,11 +221,9 @@ typedef void (*dma_async_tx_callback)(void *dma_async_param);
  * @callback: routine to call after this operation is complete
  * @callback_param: general parameter to pass to the callback routine
  * ---async_tx api specific fields---
- * @depend_list: at completion this list of transactions are submitted
- * @depend_node: allow this transaction to be executed after another
- *	transaction has completed, possibly on another channel
+ * @next: at completion submit this descriptor
  * @parent: pointer to the next level up in the dependency chain
- * @lock: protect the dependency list
+ * @lock: protect the parent and next pointers
  */
 struct dma_async_tx_descriptor {
 	dma_cookie_t cookie;
@@ -236,8 +234,7 @@ struct dma_async_tx_descriptor {
 	dma_cookie_t (*tx_submit)(struct dma_async_tx_descriptor *tx);
 	dma_async_tx_callback callback;
 	void *callback_param;
-	struct list_head depend_list;
-	struct list_head depend_node;
+	struct dma_async_tx_descriptor *next;
 	struct dma_async_tx_descriptor *parent;
 	spinlock_t lock;
 };
-- 
cgit v1.2.3


From ce4d65a5db77e1568c82d5151a746f627c4f6ed5 Mon Sep 17 00:00:00 2001
From: Dan Williams <dan.j.williams@intel.com>
Date: Thu, 17 Apr 2008 20:17:26 -0700
Subject: async_tx: kill ->device_dependency_added

DMA drivers no longer need to be notified of dependency submission
events as async_tx_run_dependencies and async_tx_channel_switch will
handle the scheduling and execution of dependent operations.

[sfr@canb.auug.org.au: extend this for fsldma]
Acked-by: Shannon Nelson <shannon.nelson@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 drivers/dma/dmaengine.c   |  1 -
 drivers/dma/fsldma.c      |  8 --------
 drivers/dma/ioat_dma.c    | 12 ------------
 drivers/dma/iop-adma.c    |  7 -------
 include/linux/dmaengine.h |  2 --
 5 files changed, 30 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index 9cb898a76bb3..af6911a75dae 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -362,7 +362,6 @@ int dma_async_device_register(struct dma_device *device)
 
 	BUG_ON(!device->device_alloc_chan_resources);
 	BUG_ON(!device->device_free_chan_resources);
-	BUG_ON(!device->device_dependency_added);
 	BUG_ON(!device->device_is_tx_complete);
 	BUG_ON(!device->device_issue_pending);
 	BUG_ON(!device->dev);
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index d8ae18dbf1a7..95b36b7934a5 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -658,13 +658,6 @@ static void fsl_dma_memcpy_issue_pending(struct dma_chan *chan)
 	fsl_chan_xfer_ld_queue(fsl_chan);
 }
 
-static void fsl_dma_dependency_added(struct dma_chan *chan)
-{
-	struct fsl_dma_chan *fsl_chan = to_fsl_chan(chan);
-
-	fsl_chan_ld_cleanup(fsl_chan);
-}
-
 /**
  * fsl_dma_is_complete - Determine the DMA status
  * @fsl_chan : Freescale DMA channel
@@ -1089,7 +1082,6 @@ static int __devinit of_fsl_dma_probe(struct of_device *dev,
 	fdev->common.device_prep_dma_memcpy = fsl_dma_prep_memcpy;
 	fdev->common.device_is_tx_complete = fsl_dma_is_complete;
 	fdev->common.device_issue_pending = fsl_dma_memcpy_issue_pending;
-	fdev->common.device_dependency_added = fsl_dma_dependency_added;
 	fdev->common.dev = &dev->dev;
 
 	irq = irq_of_parse_and_map(dev->node, 0);
diff --git a/drivers/dma/ioat_dma.c b/drivers/dma/ioat_dma.c
index 4017d9e7acd2..1517fe4e2d14 100644
--- a/drivers/dma/ioat_dma.c
+++ b/drivers/dma/ioat_dma.c
@@ -924,17 +924,6 @@ static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *ioat_chan)
 	spin_unlock_bh(&ioat_chan->cleanup_lock);
 }
 
-static void ioat_dma_dependency_added(struct dma_chan *chan)
-{
-	struct ioat_dma_chan *ioat_chan = to_ioat_chan(chan);
-	spin_lock_bh(&ioat_chan->desc_lock);
-	if (ioat_chan->pending == 0) {
-		spin_unlock_bh(&ioat_chan->desc_lock);
-		ioat_dma_memcpy_cleanup(ioat_chan);
-	} else
-		spin_unlock_bh(&ioat_chan->desc_lock);
-}
-
 /**
  * ioat_dma_is_complete - poll the status of a IOAT DMA transaction
  * @chan: IOAT DMA channel handle
@@ -1316,7 +1305,6 @@ struct ioatdma_device *ioat_dma_probe(struct pci_dev *pdev,
 
 	dma_cap_set(DMA_MEMCPY, device->common.cap_mask);
 	device->common.device_is_tx_complete = ioat_dma_is_complete;
-	device->common.device_dependency_added = ioat_dma_dependency_added;
 	switch (device->version) {
 	case IOAT_VER_1_2:
 		device->common.device_prep_dma_memcpy = ioat1_dma_prep_memcpy;
diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c
index 21854cd7190f..2aa3df50c842 100644
--- a/drivers/dma/iop-adma.c
+++ b/drivers/dma/iop-adma.c
@@ -672,12 +672,6 @@ iop_adma_prep_dma_zero_sum(struct dma_chan *chan, dma_addr_t *dma_src,
 	return sw_desc ? &sw_desc->async_tx : NULL;
 }
 
-static void iop_adma_dependency_added(struct dma_chan *chan)
-{
-	struct iop_adma_chan *iop_chan = to_iop_adma_chan(chan);
-	tasklet_schedule(&iop_chan->irq_tasklet);
-}
-
 static void iop_adma_free_chan_resources(struct dma_chan *chan)
 {
 	struct iop_adma_chan *iop_chan = to_iop_adma_chan(chan);
@@ -1178,7 +1172,6 @@ static int __devinit iop_adma_probe(struct platform_device *pdev)
 	dma_dev->device_free_chan_resources = iop_adma_free_chan_resources;
 	dma_dev->device_is_tx_complete = iop_adma_is_complete;
 	dma_dev->device_issue_pending = iop_adma_issue_pending;
-	dma_dev->device_dependency_added = iop_adma_dependency_added;
 	dma_dev->dev = &pdev->dev;
 
 	/* set prep routines based on capability */
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index 91252a7e4d03..cd34df78c6aa 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -258,7 +258,6 @@ struct dma_async_tx_descriptor {
  * @device_prep_dma_zero_sum: prepares a zero_sum operation
  * @device_prep_dma_memset: prepares a memset operation
  * @device_prep_dma_interrupt: prepares an end of chain interrupt operation
- * @device_dependency_added: async_tx notifies the channel about new deps
  * @device_issue_pending: push pending transactions to hardware
  */
 struct dma_device {
@@ -293,7 +292,6 @@ struct dma_device {
 	struct dma_async_tx_descriptor *(*device_prep_dma_interrupt)(
 		struct dma_chan *chan);
 
-	void (*device_dependency_added)(struct dma_chan *chan);
 	enum dma_status (*device_is_tx_complete)(struct dma_chan *chan,
 			dma_cookie_t cookie, dma_cookie_t *last,
 			dma_cookie_t *used);
-- 
cgit v1.2.3


From 636bdeaa1243327501edfd2a597ed7443eb4239a Mon Sep 17 00:00:00 2001
From: Dan Williams <dan.j.williams@intel.com>
Date: Thu, 17 Apr 2008 20:17:26 -0700
Subject: dmaengine: ack to flags: make use of the unused bits in the 'ack'
 field

'ack' is currently a simple integer that flags whether or not a client is done
touching fields in the given descriptor.  It is effectively just a single bit
of information.  Converting this to a flags parameter allows the other bits to
be put to use to control completion actions, like dma-unmap, and capture
results, like xor-zero-sum == 0.

Changes are one of:
1/ convert all open-coded ->ack manipulations to use async_tx_ack
   and async_tx_test_ack.
2/ set the ack bit at prep time where possible
3/ make drivers store the flags at prep time
4/ add flags to the device_prep_dma_interrupt prototype

Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 crypto/async_tx/async_memcpy.c |  2 +-
 crypto/async_tx/async_tx.c     |  9 +++++----
 crypto/async_tx/async_xor.c    |  2 +-
 drivers/dma/dmaengine.c        | 12 ++++++------
 drivers/dma/fsldma.c           | 10 +++++-----
 drivers/dma/ioat_dma.c         | 24 ++++++++++++------------
 drivers/dma/iop-adma.c         | 39 +++++++++++++++++++++------------------
 include/linux/dmaengine.h      | 25 ++++++++++++++++++-------
 8 files changed, 69 insertions(+), 54 deletions(-)

(limited to 'include/linux')

diff --git a/crypto/async_tx/async_memcpy.c b/crypto/async_tx/async_memcpy.c
index 84caa4efc0d4..a5eda80e8427 100644
--- a/crypto/async_tx/async_memcpy.c
+++ b/crypto/async_tx/async_memcpy.c
@@ -77,7 +77,7 @@ async_memcpy(struct page *dest, struct page *src, unsigned int dest_offset,
 			/* if ack is already set then we cannot be sure
 			 * we are referring to the correct operation
 			 */
-			BUG_ON(depend_tx->ack);
+			BUG_ON(async_tx_test_ack(depend_tx));
 			if (dma_wait_for_async_tx(depend_tx) == DMA_ERROR)
 				panic("%s: DMA_ERROR waiting for depend_tx\n",
 					__func__);
diff --git a/crypto/async_tx/async_tx.c b/crypto/async_tx/async_tx.c
index 69756164b61d..c6e772fc5ccd 100644
--- a/crypto/async_tx/async_tx.c
+++ b/crypto/async_tx/async_tx.c
@@ -446,7 +446,7 @@ async_tx_channel_switch(struct dma_async_tx_descriptor *depend_tx,
 	 * otherwise poll for completion
 	 */
 	if (dma_has_cap(DMA_INTERRUPT, device->cap_mask))
-		intr_tx = device->device_prep_dma_interrupt(chan);
+		intr_tx = device->device_prep_dma_interrupt(chan, 0);
 	else
 		intr_tx = NULL;
 
@@ -515,7 +515,8 @@ async_tx_submit(struct dma_chan *chan, struct dma_async_tx_descriptor *tx,
 		 * 2/ dependencies are 1:1 i.e. two transactions can
 		 * not depend on the same parent
 		 */
-		BUG_ON(depend_tx->ack || depend_tx->next || tx->parent);
+		BUG_ON(async_tx_test_ack(depend_tx) || depend_tx->next ||
+		       tx->parent);
 
 		/* the lock prevents async_tx_run_dependencies from missing
 		 * the setting of ->next when ->parent != NULL
@@ -594,7 +595,7 @@ async_trigger_callback(enum async_tx_flags flags,
 		if (device && !dma_has_cap(DMA_INTERRUPT, device->cap_mask))
 			device = NULL;
 
-		tx = device ? device->device_prep_dma_interrupt(chan) : NULL;
+		tx = device ? device->device_prep_dma_interrupt(chan, 0) : NULL;
 	} else
 		tx = NULL;
 
@@ -610,7 +611,7 @@ async_trigger_callback(enum async_tx_flags flags,
 			/* if ack is already set then we cannot be sure
 			 * we are referring to the correct operation
 			 */
-			BUG_ON(depend_tx->ack);
+			BUG_ON(async_tx_test_ack(depend_tx));
 			if (dma_wait_for_async_tx(depend_tx) == DMA_ERROR)
 				panic("%s: DMA_ERROR waiting for depend_tx\n",
 					__func__);
diff --git a/crypto/async_tx/async_xor.c b/crypto/async_tx/async_xor.c
index 1c445c7bdab7..3a0dddca5a10 100644
--- a/crypto/async_tx/async_xor.c
+++ b/crypto/async_tx/async_xor.c
@@ -191,7 +191,7 @@ async_xor(struct page *dest, struct page **src_list, unsigned int offset,
 				/* if ack is already set then we cannot be sure
 				 * we are referring to the correct operation
 				 */
-				BUG_ON(depend_tx->ack);
+				BUG_ON(async_tx_test_ack(depend_tx));
 				if (dma_wait_for_async_tx(depend_tx) ==
 					DMA_ERROR)
 					panic("%s: DMA_ERROR waiting for "
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index af6911a75dae..d6dc70fd7527 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -478,7 +478,8 @@ dma_async_memcpy_buf_to_buf(struct dma_chan *chan, void *dest,
 
 	dma_src = dma_map_single(dev->dev, src, len, DMA_TO_DEVICE);
 	dma_dest = dma_map_single(dev->dev, dest, len, DMA_FROM_DEVICE);
-	tx = dev->device_prep_dma_memcpy(chan, dma_dest, dma_src, len, 0);
+	tx = dev->device_prep_dma_memcpy(chan, dma_dest, dma_src, len,
+					 DMA_CTRL_ACK);
 
 	if (!tx) {
 		dma_unmap_single(dev->dev, dma_src, len, DMA_TO_DEVICE);
@@ -486,7 +487,6 @@ dma_async_memcpy_buf_to_buf(struct dma_chan *chan, void *dest,
 		return -ENOMEM;
 	}
 
-	tx->ack = 1;
 	tx->callback = NULL;
 	cookie = tx->tx_submit(tx);
 
@@ -524,7 +524,8 @@ dma_async_memcpy_buf_to_pg(struct dma_chan *chan, struct page *page,
 
 	dma_src = dma_map_single(dev->dev, kdata, len, DMA_TO_DEVICE);
 	dma_dest = dma_map_page(dev->dev, page, offset, len, DMA_FROM_DEVICE);
-	tx = dev->device_prep_dma_memcpy(chan, dma_dest, dma_src, len, 0);
+	tx = dev->device_prep_dma_memcpy(chan, dma_dest, dma_src, len,
+					 DMA_CTRL_ACK);
 
 	if (!tx) {
 		dma_unmap_single(dev->dev, dma_src, len, DMA_TO_DEVICE);
@@ -532,7 +533,6 @@ dma_async_memcpy_buf_to_pg(struct dma_chan *chan, struct page *page,
 		return -ENOMEM;
 	}
 
-	tx->ack = 1;
 	tx->callback = NULL;
 	cookie = tx->tx_submit(tx);
 
@@ -573,7 +573,8 @@ dma_async_memcpy_pg_to_pg(struct dma_chan *chan, struct page *dest_pg,
 	dma_src = dma_map_page(dev->dev, src_pg, src_off, len, DMA_TO_DEVICE);
 	dma_dest = dma_map_page(dev->dev, dest_pg, dest_off, len,
 				DMA_FROM_DEVICE);
-	tx = dev->device_prep_dma_memcpy(chan, dma_dest, dma_src, len, 0);
+	tx = dev->device_prep_dma_memcpy(chan, dma_dest, dma_src, len,
+					 DMA_CTRL_ACK);
 
 	if (!tx) {
 		dma_unmap_page(dev->dev, dma_src, len, DMA_TO_DEVICE);
@@ -581,7 +582,6 @@ dma_async_memcpy_pg_to_pg(struct dma_chan *chan, struct page *dest_pg,
 		return -ENOMEM;
 	}
 
-	tx->ack = 1;
 	tx->callback = NULL;
 	cookie = tx->tx_submit(tx);
 
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index 95b36b7934a5..054eabffc185 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -412,7 +412,7 @@ static void fsl_dma_free_chan_resources(struct dma_chan *chan)
 }
 
 static struct dma_async_tx_descriptor *
-fsl_dma_prep_interrupt(struct dma_chan *chan)
+fsl_dma_prep_interrupt(struct dma_chan *chan, unsigned long flags)
 {
 	struct fsl_dma_chan *fsl_chan;
 	struct fsl_desc_sw *new;
@@ -429,7 +429,7 @@ fsl_dma_prep_interrupt(struct dma_chan *chan)
 	}
 
 	new->async_tx.cookie = -EBUSY;
-	new->async_tx.ack = 0;
+	new->async_tx.flags = flags;
 
 	/* Insert the link descriptor to the LD ring */
 	list_add_tail(&new->node, &new->async_tx.tx_list);
@@ -482,7 +482,7 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_memcpy(
 			set_desc_next(fsl_chan, &prev->hw, new->async_tx.phys);
 
 		new->async_tx.cookie = 0;
-		new->async_tx.ack = 1;
+		async_tx_ack(&new->async_tx);
 
 		prev = new;
 		len -= copy;
@@ -493,7 +493,7 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_memcpy(
 		list_add_tail(&new->node, &first->async_tx.tx_list);
 	} while (len);
 
-	new->async_tx.ack = 0; /* client is in control of this ack */
+	new->async_tx.flags = flags; /* client is in control of this ack */
 	new->async_tx.cookie = -EBUSY;
 
 	/* Set End-of-link to the last link descriptor of new list*/
@@ -874,7 +874,7 @@ static int fsl_dma_self_test(struct fsl_dma_chan *fsl_chan)
 	async_tx_ack(tx3);
 
 	/* Interrupt tx test */
-	tx1 = fsl_dma_prep_interrupt(chan);
+	tx1 = fsl_dma_prep_interrupt(chan, 0);
 	async_tx_ack(tx1);
 	cookie = fsl_dma_tx_submit(tx1);
 
diff --git a/drivers/dma/ioat_dma.c b/drivers/dma/ioat_dma.c
index 1517fe4e2d14..318e8a22d814 100644
--- a/drivers/dma/ioat_dma.c
+++ b/drivers/dma/ioat_dma.c
@@ -212,14 +212,14 @@ static dma_cookie_t ioat1_tx_submit(struct dma_async_tx_descriptor *tx)
 	u32 copy;
 	size_t len;
 	dma_addr_t src, dst;
-	int orig_ack;
+	unsigned long orig_flags;
 	unsigned int desc_count = 0;
 
 	/* src and dest and len are stored in the initial descriptor */
 	len = first->len;
 	src = first->src;
 	dst = first->dst;
-	orig_ack = first->async_tx.ack;
+	orig_flags = first->async_tx.flags;
 	new = first;
 
 	spin_lock_bh(&ioat_chan->desc_lock);
@@ -228,7 +228,7 @@ static dma_cookie_t ioat1_tx_submit(struct dma_async_tx_descriptor *tx)
 	do {
 		copy = min_t(size_t, len, ioat_chan->xfercap);
 
-		new->async_tx.ack = 1;
+		async_tx_ack(&new->async_tx);
 
 		hw = new->hw;
 		hw->size = copy;
@@ -264,7 +264,7 @@ static dma_cookie_t ioat1_tx_submit(struct dma_async_tx_descriptor *tx)
 	}
 
 	new->tx_cnt = desc_count;
-	new->async_tx.ack = orig_ack; /* client is in control of this ack */
+	new->async_tx.flags = orig_flags; /* client is in control of this ack */
 
 	/* store the original values for use in later cleanup */
 	if (new != first) {
@@ -304,14 +304,14 @@ static dma_cookie_t ioat2_tx_submit(struct dma_async_tx_descriptor *tx)
 	u32 copy;
 	size_t len;
 	dma_addr_t src, dst;
-	int orig_ack;
+	unsigned long orig_flags;
 	unsigned int desc_count = 0;
 
 	/* src and dest and len are stored in the initial descriptor */
 	len = first->len;
 	src = first->src;
 	dst = first->dst;
-	orig_ack = first->async_tx.ack;
+	orig_flags = first->async_tx.flags;
 	new = first;
 
 	/*
@@ -321,7 +321,7 @@ static dma_cookie_t ioat2_tx_submit(struct dma_async_tx_descriptor *tx)
 	do {
 		copy = min_t(size_t, len, ioat_chan->xfercap);
 
-		new->async_tx.ack = 1;
+		async_tx_ack(&new->async_tx);
 
 		hw = new->hw;
 		hw->size = copy;
@@ -349,7 +349,7 @@ static dma_cookie_t ioat2_tx_submit(struct dma_async_tx_descriptor *tx)
 	}
 
 	new->tx_cnt = desc_count;
-	new->async_tx.ack = orig_ack; /* client is in control of this ack */
+	new->async_tx.flags = orig_flags; /* client is in control of this ack */
 
 	/* store the original values for use in later cleanup */
 	if (new != first) {
@@ -714,7 +714,7 @@ static struct dma_async_tx_descriptor *ioat1_dma_prep_memcpy(
 		new->len = len;
 		new->dst = dma_dest;
 		new->src = dma_src;
-		new->async_tx.ack = 0;
+		new->async_tx.flags = flags;
 		return &new->async_tx;
 	} else
 		return NULL;
@@ -742,7 +742,7 @@ static struct dma_async_tx_descriptor *ioat2_dma_prep_memcpy(
 		new->len = len;
 		new->dst = dma_dest;
 		new->src = dma_src;
-		new->async_tx.ack = 0;
+		new->async_tx.flags = flags;
 		return &new->async_tx;
 	} else
 		return NULL;
@@ -842,7 +842,7 @@ static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *ioat_chan)
 				 * a completed entry, but not the last, so clean
 				 * up if the client is done with the descriptor
 				 */
-				if (desc->async_tx.ack) {
+				if (async_tx_test_ack(&desc->async_tx)) {
 					list_del(&desc->node);
 					list_add_tail(&desc->node,
 						      &ioat_chan->free_desc);
@@ -979,7 +979,7 @@ static void ioat_dma_start_null_desc(struct ioat_dma_chan *ioat_chan)
 	desc->hw->size = 0;
 	desc->hw->src_addr = 0;
 	desc->hw->dst_addr = 0;
-	desc->async_tx.ack = 1;
+	async_tx_ack(&desc->async_tx);
 	switch (ioat_chan->device->version) {
 	case IOAT_VER_1_2:
 		desc->hw->next = 0;
diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c
index 93252294f32b..762b729672e0 100644
--- a/drivers/dma/iop-adma.c
+++ b/drivers/dma/iop-adma.c
@@ -111,7 +111,7 @@ iop_adma_clean_slot(struct iop_adma_desc_slot *desc,
 	/* the client is allowed to attach dependent operations
 	 * until 'ack' is set
 	 */
-	if (!desc->async_tx.ack)
+	if (!async_tx_test_ack(&desc->async_tx))
 		return 0;
 
 	/* leave the last descriptor in the chain
@@ -148,7 +148,7 @@ static void __iop_adma_slot_cleanup(struct iop_adma_chan *iop_chan)
 			"this_desc: %#x next_desc: %#x ack: %d\n",
 			iter->async_tx.cookie, iter->idx, busy,
 			iter->async_tx.phys, iop_desc_get_next_desc(iter),
-			iter->async_tx.ack);
+			async_tx_test_ack(&iter->async_tx));
 		prefetch(_iter);
 		prefetch(&_iter->async_tx);
 
@@ -338,9 +338,7 @@ retry:
 
 				/* pre-ack all but the last descriptor */
 				if (num_slots != slots_per_op)
-					iter->async_tx.ack = 1;
-				else
-					iter->async_tx.ack = 0;
+					async_tx_ack(&iter->async_tx);
 
 				list_add_tail(&iter->chain_node, &chain);
 				alloc_tail = iter;
@@ -513,7 +511,7 @@ static int iop_adma_alloc_chan_resources(struct dma_chan *chan)
 }
 
 static struct dma_async_tx_descriptor *
-iop_adma_prep_dma_interrupt(struct dma_chan *chan)
+iop_adma_prep_dma_interrupt(struct dma_chan *chan, unsigned long flags)
 {
 	struct iop_adma_chan *iop_chan = to_iop_adma_chan(chan);
 	struct iop_adma_desc_slot *sw_desc, *grp_start;
@@ -528,6 +526,7 @@ iop_adma_prep_dma_interrupt(struct dma_chan *chan)
 		grp_start = sw_desc->group_head;
 		iop_desc_init_interrupt(grp_start, iop_chan);
 		grp_start->unmap_len = 0;
+		sw_desc->async_tx.flags = flags;
 	}
 	spin_unlock_bh(&iop_chan->lock);
 
@@ -560,6 +559,7 @@ iop_adma_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dma_dest,
 		iop_desc_set_memcpy_src_addr(grp_start, dma_src);
 		sw_desc->unmap_src_cnt = 1;
 		sw_desc->unmap_len = len;
+		sw_desc->async_tx.flags = flags;
 	}
 	spin_unlock_bh(&iop_chan->lock);
 
@@ -592,6 +592,7 @@ iop_adma_prep_dma_memset(struct dma_chan *chan, dma_addr_t dma_dest,
 		iop_desc_set_dest_addr(grp_start, iop_chan, dma_dest);
 		sw_desc->unmap_src_cnt = 1;
 		sw_desc->unmap_len = len;
+		sw_desc->async_tx.flags = flags;
 	}
 	spin_unlock_bh(&iop_chan->lock);
 
@@ -625,6 +626,7 @@ iop_adma_prep_dma_xor(struct dma_chan *chan, dma_addr_t dma_dest,
 		iop_desc_set_dest_addr(grp_start, iop_chan, dma_dest);
 		sw_desc->unmap_src_cnt = src_cnt;
 		sw_desc->unmap_len = len;
+		sw_desc->async_tx.flags = flags;
 		while (src_cnt--)
 			iop_desc_set_xor_src_addr(grp_start, src_cnt,
 						  dma_src[src_cnt]);
@@ -661,6 +663,7 @@ iop_adma_prep_dma_zero_sum(struct dma_chan *chan, dma_addr_t *dma_src,
 			__func__, grp_start->xor_check_result);
 		sw_desc->unmap_src_cnt = src_cnt;
 		sw_desc->unmap_len = len;
+		sw_desc->async_tx.flags = flags;
 		while (src_cnt--)
 			iop_desc_set_zero_sum_src_addr(grp_start, src_cnt,
 						       dma_src[src_cnt]);
@@ -847,11 +850,11 @@ static int __devinit iop_adma_memcpy_self_test(struct iop_adma_device *device)
 	src_dma = dma_map_single(dma_chan->device->dev, src,
 				IOP_ADMA_TEST_SIZE, DMA_TO_DEVICE);
 	tx = iop_adma_prep_dma_memcpy(dma_chan, dest_dma, src_dma,
-				      IOP_ADMA_TEST_SIZE, 1);
+				      IOP_ADMA_TEST_SIZE,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
 	cookie = iop_adma_tx_submit(tx);
 	iop_adma_issue_pending(dma_chan);
-	async_tx_ack(tx);
 	msleep(1);
 
 	if (iop_adma_is_complete(dma_chan, cookie, NULL, NULL) !=
@@ -947,11 +950,11 @@ iop_adma_xor_zero_sum_self_test(struct iop_adma_device *device)
 		dma_srcs[i] = dma_map_page(dma_chan->device->dev, xor_srcs[i],
 					   0, PAGE_SIZE, DMA_TO_DEVICE);
 	tx = iop_adma_prep_dma_xor(dma_chan, dest_dma, dma_srcs,
-				   IOP_ADMA_NUM_SRC_TEST, PAGE_SIZE, 1);
+				   IOP_ADMA_NUM_SRC_TEST, PAGE_SIZE,
+				   DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
 	cookie = iop_adma_tx_submit(tx);
 	iop_adma_issue_pending(dma_chan);
-	async_tx_ack(tx);
 	msleep(8);
 
 	if (iop_adma_is_complete(dma_chan, cookie, NULL, NULL) !=
@@ -994,11 +997,11 @@ iop_adma_xor_zero_sum_self_test(struct iop_adma_device *device)
 					   DMA_TO_DEVICE);
 	tx = iop_adma_prep_dma_zero_sum(dma_chan, dma_srcs,
 					IOP_ADMA_NUM_SRC_TEST + 1, PAGE_SIZE,
-					&zero_sum_result, 1);
+					&zero_sum_result,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
 	cookie = iop_adma_tx_submit(tx);
 	iop_adma_issue_pending(dma_chan);
-	async_tx_ack(tx);
 	msleep(8);
 
 	if (iop_adma_is_complete(dma_chan, cookie, NULL, NULL) != DMA_SUCCESS) {
@@ -1018,11 +1021,11 @@ iop_adma_xor_zero_sum_self_test(struct iop_adma_device *device)
 	/* test memset */
 	dma_addr = dma_map_page(dma_chan->device->dev, dest, 0,
 			PAGE_SIZE, DMA_FROM_DEVICE);
-	tx = iop_adma_prep_dma_memset(dma_chan, dma_addr, 0, PAGE_SIZE, 1);
+	tx = iop_adma_prep_dma_memset(dma_chan, dma_addr, 0, PAGE_SIZE,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
 	cookie = iop_adma_tx_submit(tx);
 	iop_adma_issue_pending(dma_chan);
-	async_tx_ack(tx);
 	msleep(8);
 
 	if (iop_adma_is_complete(dma_chan, cookie, NULL, NULL) != DMA_SUCCESS) {
@@ -1050,11 +1053,11 @@ iop_adma_xor_zero_sum_self_test(struct iop_adma_device *device)
 					   DMA_TO_DEVICE);
 	tx = iop_adma_prep_dma_zero_sum(dma_chan, dma_srcs,
 					IOP_ADMA_NUM_SRC_TEST + 1, PAGE_SIZE,
-					&zero_sum_result, 1);
+					&zero_sum_result,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
 	cookie = iop_adma_tx_submit(tx);
 	iop_adma_issue_pending(dma_chan);
-	async_tx_ack(tx);
 	msleep(8);
 
 	if (iop_adma_is_complete(dma_chan, cookie, NULL, NULL) != DMA_SUCCESS) {
@@ -1287,7 +1290,7 @@ static void iop_chan_start_null_memcpy(struct iop_adma_chan *iop_chan)
 		grp_start = sw_desc->group_head;
 
 		list_splice_init(&sw_desc->async_tx.tx_list, &iop_chan->chain);
-		sw_desc->async_tx.ack = 1;
+		async_tx_ack(&sw_desc->async_tx);
 		iop_desc_init_memcpy(grp_start, 0);
 		iop_desc_set_byte_count(grp_start, iop_chan, 0);
 		iop_desc_set_dest_addr(grp_start, iop_chan, 0);
@@ -1343,7 +1346,7 @@ static void iop_chan_start_null_xor(struct iop_adma_chan *iop_chan)
 	if (sw_desc) {
 		grp_start = sw_desc->group_head;
 		list_splice_init(&sw_desc->async_tx.tx_list, &iop_chan->chain);
-		sw_desc->async_tx.ack = 1;
+		async_tx_ack(&sw_desc->async_tx);
 		iop_desc_init_null_xor(grp_start, 2, 0);
 		iop_desc_set_byte_count(grp_start, iop_chan, 0);
 		iop_desc_set_dest_addr(grp_start, iop_chan, 0);
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index cd34df78c6aa..b4d84ed6187d 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -95,12 +95,17 @@ enum dma_transaction_type {
 #define DMA_TX_TYPE_END (DMA_INTERRUPT + 1)
 
 /**
- * enum dma_prep_flags - DMA flags to augment operation preparation
+ * enum dma_ctrl_flags - DMA flags to augment operation preparation,
+ * 	control completion, and communicate status.
  * @DMA_PREP_INTERRUPT - trigger an interrupt (callback) upon completion of
  * 	this transaction
+ * @DMA_CTRL_ACK - the descriptor cannot be reused until the client
+ * 	acknowledges receipt, i.e. has has a chance to establish any
+ * 	dependency chains
  */
-enum dma_prep_flags {
+enum dma_ctrl_flags {
 	DMA_PREP_INTERRUPT = (1 << 0),
+	DMA_CTRL_ACK = (1 << 1),
 };
 
 /**
@@ -211,8 +216,8 @@ typedef void (*dma_async_tx_callback)(void *dma_async_param);
  * ---dma generic offload fields---
  * @cookie: tracking cookie for this transaction, set to -EBUSY if
  *	this tx is sitting on a dependency list
- * @ack: the descriptor can not be reused until the client acknowledges
- *	receipt, i.e. has has a chance to establish any dependency chains
+ * @flags: flags to augment operation preparation, control completion, and
+ * 	communicate status
  * @phys: physical address of the descriptor
  * @tx_list: driver common field for operations that require multiple
  *	descriptors
@@ -227,7 +232,7 @@ typedef void (*dma_async_tx_callback)(void *dma_async_param);
  */
 struct dma_async_tx_descriptor {
 	dma_cookie_t cookie;
-	int ack;
+	enum dma_ctrl_flags flags; /* not a 'long' to pack with cookie */
 	dma_addr_t phys;
 	struct list_head tx_list;
 	struct dma_chan *chan;
@@ -290,7 +295,7 @@ struct dma_device {
 		struct dma_chan *chan, dma_addr_t dest, int value, size_t len,
 		unsigned long flags);
 	struct dma_async_tx_descriptor *(*device_prep_dma_interrupt)(
-		struct dma_chan *chan);
+		struct dma_chan *chan, unsigned long flags);
 
 	enum dma_status (*device_is_tx_complete)(struct dma_chan *chan,
 			dma_cookie_t cookie, dma_cookie_t *last,
@@ -316,7 +321,13 @@ void dma_async_tx_descriptor_init(struct dma_async_tx_descriptor *tx,
 static inline void
 async_tx_ack(struct dma_async_tx_descriptor *tx)
 {
-	tx->ack = 1;
+	tx->flags |= DMA_CTRL_ACK;
+}
+
+static inline int
+async_tx_test_ack(struct dma_async_tx_descriptor *tx)
+{
+	return tx->flags & DMA_CTRL_ACK;
 }
 
 #define first_dma_cap(mask) __first_dma_cap(&(mask))
-- 
cgit v1.2.3


From a594eeb1a1d320981fccc29584b6f21fcebd765f Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Fri, 18 Apr 2008 00:46:20 +0200
Subject: IDE: remove ide=reverse IDE core

This option is obsolete and can be removed safely.

It allows us to remove the pci_get_device_reverse() function from the
PCI core.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 Documentation/ide/ide.txt           |  2 --
 Documentation/kernel-parameters.txt |  2 +-
 drivers/ide/Kconfig                 |  9 +--------
 drivers/ide/ide-scan-pci.c          |  9 ++-------
 drivers/ide/ide.c                   | 12 ------------
 include/linux/ide.h                 |  1 -
 6 files changed, 4 insertions(+), 31 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/ide/ide.txt b/Documentation/ide/ide.txt
index 818676aad45a..ce7c01af30f6 100644
--- a/Documentation/ide/ide.txt
+++ b/Documentation/ide/ide.txt
@@ -269,8 +269,6 @@ Summary of ide driver parameters for kernel command line
 			  ability to bit test for detection is currently
 			  unknown.
 
- "ide=reverse"		: formerly called to pci sub-system, but now local.
-
  "ide=doubler"		: probe/support IDE doublers on Amiga
 
 There may be more options than shown -- use the source, Luke!
diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index dafd001bf833..228d85a3f319 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -763,7 +763,7 @@ and is between 256 and 4096 characters. It is defined in the file
 			Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]]
 
 	ide=		[HW] (E)IDE subsystem
-			Format: ide=nodma or ide=doubler or ide=reverse
+			Format: ide=nodma or ide=doubler
 			See Documentation/ide/ide.txt.
 
 	ide?=		[HW] (E)IDE subsystem
diff --git a/drivers/ide/Kconfig b/drivers/ide/Kconfig
index eed6d8e1b5c7..ac5875783430 100644
--- a/drivers/ide/Kconfig
+++ b/drivers/ide/Kconfig
@@ -416,12 +416,6 @@ config BLK_DEV_OFFBOARD
 	  This can improve the usability of some boot managers such as lilo
 	  when booting from a drive on an off-board controller.
 
-	  If you say Y here, and you actually want to reverse the device scan
-	  order as explained above, you also need to issue the kernel command
-	  line option "ide=reverse". (Try "man bootparam" or see the
-	  documentation of your boot loader (lilo or loadlin) about how to
-	  pass options to the kernel at boot time.)
-
 	  Note that, if you do this, the order of the hd* devices will be
 	  rearranged which may require modification of fstab and other files.
 
@@ -615,8 +609,7 @@ config BLK_DEV_HPT366
 	  reference to device 0x80. The other solution is to say Y to "Boot
 	  off-board chipsets first support" (CONFIG_BLK_DEV_OFFBOARD) unless
 	  your mother board has the chipset natively mounted. Regardless one
-	  should use the fore mentioned option and call at LILO or include
-	  "ide=reverse" in LILO's append-line.
+	  should use the fore mentioned option and call at LILO.
 
 	  This driver requires dynamic tuning of the chipset during the
 	  ide-probe at boot. It is reported to support DVD II drives, by the
diff --git a/drivers/ide/ide-scan-pci.c b/drivers/ide/ide-scan-pci.c
index 93d2e41be853..98888da1b600 100644
--- a/drivers/ide/ide-scan-pci.c
+++ b/drivers/ide/ide-scan-pci.c
@@ -88,13 +88,8 @@ static int __init ide_scan_pcibus(void)
 	struct list_head *l, *n;
 
 	pre_init = 0;
-	if (!ide_scan_direction)
-		while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)))
-			ide_scan_pcidev(dev);
-	else
-		while ((dev = pci_get_device_reverse(PCI_ANY_ID, PCI_ANY_ID,
-						     dev)))
-			ide_scan_pcidev(dev);
+	while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)))
+		ide_scan_pcidev(dev);
 
 	/*
 	 *	Hand the drivers over to the PCI layer now we
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index fc69fe2e3ec0..5f545153391a 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -90,10 +90,6 @@ static int system_bus_speed;	/* holds what we think is VESA/PCI bus speed */
 DEFINE_MUTEX(ide_cfg_mtx);
  __cacheline_aligned_in_smp DEFINE_SPINLOCK(ide_lock);
 
-#ifdef CONFIG_IDEPCI_PCIBUS_ORDER
-int ide_scan_direction; /* THIS was formerly 2.2.x pci=reverse */
-#endif
-
 int noautodma = 0;
 
 #ifdef CONFIG_BLK_DEV_IDEACPI
@@ -1217,14 +1213,6 @@ static int __init ide_setup(char *s)
 		goto obsolete_option;
 	}
 
-#ifdef CONFIG_IDEPCI_PCIBUS_ORDER
-	if (!strcmp(s, "ide=reverse")) {
-		ide_scan_direction = 1;
-		printk(" : Enabled support for IDE inverse scan order.\n");
-		goto obsolete_option;
-	}
-#endif
-
 #ifdef CONFIG_BLK_DEV_IDEACPI
 	if (!strcmp(s, "ide=noacpi")) {
 		//printk(" : Disable IDE ACPI support.\n");
diff --git a/include/linux/ide.h b/include/linux/ide.h
index bc26b2f27359..43d2968a4e13 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -990,7 +990,6 @@ extern void do_ide_request(struct request_queue *);
 void ide_init_disk(struct gendisk *, ide_drive_t *);
 
 #ifdef CONFIG_IDEPCI_PCIBUS_ORDER
-extern int ide_scan_direction;
 extern int __ide_pci_register_driver(struct pci_driver *driver, struct module *owner, const char *mod_name);
 #define ide_pci_register_driver(d) __ide_pci_register_driver(d, THIS_MODULE, KBUILD_MODNAME)
 #else
-- 
cgit v1.2.3


From 4f0eee4d877e3b617b6a22d209d52b3dfca2b2a7 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:21 +0200
Subject: ide: use ide_find_port() instead of ide_deprecated_find_port()

* Use ide_find_port() instead of ide_deprecated_find_port() in bast-ide/
  palm_bk3710/ide-cs/delkin_cb host drivers and in ide_register_hw().

* Remove no longer needed ide_deprecated_find_port().

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/arm/bast-ide.c    |  2 +-
 drivers/ide/arm/palm_bk3710.c |  2 +-
 drivers/ide/ide.c             | 27 +--------------------------
 drivers/ide/legacy/ide-cs.c   |  2 +-
 drivers/ide/pci/delkin_cb.c   |  2 +-
 include/linux/ide.h           |  1 -
 6 files changed, 5 insertions(+), 31 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/arm/bast-ide.c b/drivers/ide/arm/bast-ide.c
index 161d30c8481e..a22da7ab2b57 100644
--- a/drivers/ide/arm/bast-ide.c
+++ b/drivers/ide/arm/bast-ide.c
@@ -41,7 +41,7 @@ static int __init bastide_register(unsigned int base, unsigned int aux, int irq)
 	hw.io_ports[IDE_CONTROL_OFFSET] = aux + (6 * 0x20);
 	hw.irq = irq;
 
-	hwif = ide_deprecated_find_port(hw.io_ports[IDE_DATA_OFFSET]);
+	hwif = ide_find_port(hw.io_ports[IDE_DATA_OFFSET]);
 	if (hwif == NULL)
 		goto out;
 
diff --git a/drivers/ide/arm/palm_bk3710.c b/drivers/ide/arm/palm_bk3710.c
index 8e1f6bd33887..0a722503c102 100644
--- a/drivers/ide/arm/palm_bk3710.c
+++ b/drivers/ide/arm/palm_bk3710.c
@@ -378,7 +378,7 @@ static int __devinit palm_bk3710_probe(struct platform_device *pdev)
 	hw.irq = irq->start;
 	hw.chipset = ide_palm3710;
 
-	hwif = ide_deprecated_find_port(hw.io_ports[IDE_DATA_OFFSET]);
+	hwif = ide_find_port(hw.io_ports[IDE_DATA_OFFSET]);
 	if (hwif == NULL)
 		goto out;
 
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index 67ce697a9d51..bd5ff7d987d0 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -618,31 +618,6 @@ void ide_init_port_hw(ide_hwif_t *hwif, hw_regs_t *hw)
 }
 EXPORT_SYMBOL_GPL(ide_init_port_hw);
 
-ide_hwif_t *ide_deprecated_find_port(unsigned long base)
-{
-	ide_hwif_t *hwif;
-	int i;
-
-	for (i = 0; i < MAX_HWIFS; i++) {
-		hwif = &ide_hwifs[i];
-		if (hwif->io_ports[IDE_DATA_OFFSET] == base)
-			goto found;
-	}
-
-	for (i = 0; i < MAX_HWIFS; i++) {
-		hwif = &ide_hwifs[i];
-		if (hwif->hold)
-			continue;
-		if (!hwif->present && hwif->mate == NULL)
-			goto found;
-	}
-
-	hwif = NULL;
-found:
-	return hwif;
-}
-EXPORT_SYMBOL_GPL(ide_deprecated_find_port);
-
 /**
  *	ide_register_hw		-	register IDE interface
  *	@hw: hardware registers
@@ -662,7 +637,7 @@ int ide_register_hw(hw_regs_t *hw, void (*quirkproc)(ide_drive_t *),
 	u8 idx[4] = { 0xff, 0xff, 0xff, 0xff };
 
 	do {
-		hwif = ide_deprecated_find_port(hw->io_ports[IDE_DATA_OFFSET]);
+		hwif = ide_find_port(hw->io_ports[IDE_DATA_OFFSET]);
 		if (hwif)
 			goto found;
 		for (index = 0; index < MAX_HWIFS; index++)
diff --git a/drivers/ide/legacy/ide-cs.c b/drivers/ide/legacy/ide-cs.c
index 15ccf6944ae2..de2e5944809e 100644
--- a/drivers/ide/legacy/ide-cs.c
+++ b/drivers/ide/legacy/ide-cs.c
@@ -156,7 +156,7 @@ static int idecs_register(unsigned long io, unsigned long ctl, unsigned long irq
     hw.chipset = ide_pci;
     hw.dev = &handle->dev;
 
-    hwif = ide_deprecated_find_port(hw.io_ports[IDE_DATA_OFFSET]);
+    hwif = ide_find_port(hw.io_ports[IDE_DATA_OFFSET]);
     if (hwif == NULL)
 	return -1;
 
diff --git a/drivers/ide/pci/delkin_cb.c b/drivers/ide/pci/delkin_cb.c
index 3f9cd64c26a6..e08e13a0bb6e 100644
--- a/drivers/ide/pci/delkin_cb.c
+++ b/drivers/ide/pci/delkin_cb.c
@@ -78,7 +78,7 @@ delkin_cb_probe (struct pci_dev *dev, const struct pci_device_id *id)
 	hw.irq = dev->irq;
 	hw.chipset = ide_pci;		/* this enables IRQ sharing */
 
-	hwif = ide_deprecated_find_port(hw.io_ports[IDE_DATA_OFFSET]);
+	hwif = ide_find_port(hw.io_ports[IDE_DATA_OFFSET]);
 	if (hwif == NULL)
 		goto out_disable;
 
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 43d2968a4e13..1b423958a894 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -186,7 +186,6 @@ typedef struct hw_regs_s {
 } hw_regs_t;
 
 struct hwif_s * ide_find_port(unsigned long);
-struct hwif_s *ide_deprecated_find_port(unsigned long);
 void ide_init_port_data(struct hwif_s *, unsigned int);
 void ide_init_port_hw(struct hwif_s *, hw_regs_t *);
 
-- 
cgit v1.2.3


From 5b0c4b30a625927340a3e7f565aa4de8b60489cc Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:22 +0200
Subject: ide: remove IDE devices from /proc/ide/ before unregistering them

IDE devices need to be removed from /proc/ide/ _before_ being unregistered:

* Drop 'ide_hwif_t *hwif' argument from destroy_proc_ide_device()
  and use drive->hwif instead.

* Rename destroy_proc_ide_device() to ide_proc_unregister_device().

* Call ide_proc_unregister_device() in drive_release_dev().

* Remove no longer needed destroy_proc_ide_drives().

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/ide-probe.c |  2 ++
 drivers/ide/ide-proc.c  | 16 ++--------------
 include/linux/ide.h     |  2 ++
 3 files changed, 6 insertions(+), 14 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c
index 9db8978ad394..dea314ce33d3 100644
--- a/drivers/ide/ide-probe.c
+++ b/drivers/ide/ide-probe.c
@@ -1199,6 +1199,8 @@ static void drive_release_dev (struct device *dev)
 {
 	ide_drive_t *drive = container_of(dev, ide_drive_t, gendev);
 
+	ide_proc_unregister_device(drive);
+
 	spin_lock_irq(&ide_lock);
 	ide_remove_drive_from_hwgroup(drive);
 	kfree(drive->id);
diff --git a/drivers/ide/ide-proc.c b/drivers/ide/ide-proc.c
index bab88ca7f7ec..77025d1057b5 100644
--- a/drivers/ide/ide-proc.c
+++ b/drivers/ide/ide-proc.c
@@ -764,27 +764,16 @@ void ide_proc_port_register_devices(ide_hwif_t *hwif)
 	}
 }
 
-static void destroy_proc_ide_device(ide_hwif_t *hwif, ide_drive_t *drive)
+void ide_proc_unregister_device(ide_drive_t *drive)
 {
 	if (drive->proc) {
 		ide_remove_proc_entries(drive->proc, generic_drive_entries);
 		remove_proc_entry(drive->name, proc_ide_root);
-		remove_proc_entry(drive->name, hwif->proc);
+		remove_proc_entry(drive->name, drive->hwif->proc);
 		drive->proc = NULL;
 	}
 }
 
-static void destroy_proc_ide_drives(ide_hwif_t *hwif)
-{
-	int	d;
-
-	for (d = 0; d < MAX_DRIVES; d++) {
-		ide_drive_t *drive = &hwif->drives[d];
-		if (drive->proc)
-			destroy_proc_ide_device(hwif, drive);
-	}
-}
-
 static ide_proc_entry_t hwif_entries[] = {
 	{ "channel",	S_IFREG|S_IRUGO,	proc_ide_read_channel,	NULL },
 	{ "mate",	S_IFREG|S_IRUGO,	proc_ide_read_mate,	NULL },
@@ -816,7 +805,6 @@ EXPORT_SYMBOL_GPL(ide_pci_create_host_proc);
 void ide_proc_unregister_port(ide_hwif_t *hwif)
 {
 	if (hwif->proc) {
-		destroy_proc_ide_drives(hwif);
 		ide_remove_proc_entries(hwif->proc, hwif_entries);
 		remove_proc_entry(hwif->name, proc_ide_root);
 		hwif->proc = NULL;
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 1b423958a894..f9449ecd79d9 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -690,6 +690,7 @@ void proc_ide_create(void);
 void proc_ide_destroy(void);
 void ide_proc_register_port(ide_hwif_t *);
 void ide_proc_port_register_devices(ide_hwif_t *);
+void ide_proc_unregister_device(ide_drive_t *);
 void ide_proc_unregister_port(ide_hwif_t *);
 void ide_proc_register_driver(ide_drive_t *, ide_driver_t *);
 void ide_proc_unregister_driver(ide_drive_t *, ide_driver_t *);
@@ -723,6 +724,7 @@ static inline void proc_ide_create(void) { ; }
 static inline void proc_ide_destroy(void) { ; }
 static inline void ide_proc_register_port(ide_hwif_t *hwif) { ; }
 static inline void ide_proc_port_register_devices(ide_hwif_t *hwif) { ; }
+static inline void ide_proc_unregister_device(ide_drive_t *drive) { ; }
 static inline void ide_proc_unregister_port(ide_hwif_t *hwif) { ; }
 static inline void ide_proc_register_driver(ide_drive_t *drive, ide_driver_t *driver) { ; }
 static inline void ide_proc_unregister_driver(ide_drive_t *drive, ide_driver_t *driver) { ; }
-- 
cgit v1.2.3


From 2dde7861afa23cd59db83515cb0b810b92b220aa Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:23 +0200
Subject: ide: rework PowerMac media-bay support (take 2)

Rework PowerMac media-bay support in such way that instead of
un/registering the IDE interface we un/register IDE devices:

* Add ide_port_scan() helper for probing+registerering devices on a port.

* Rename ide_port_unregister_devices() to __ide_port_unregister_devices().

* Add ide_port_unregister_devices() helper for unregistering devices on a port.

* Add 'ide_hwif_t *cd_port' to 'struct media_bay_info', pass 'hwif' instead
  of hwif->index to media_bay_set_ide_infos() and use it to setup 'cd_port'.

* Use ide_port_unregister_devices() instead of ide_unregister()
  and ide_port_scan() instead of ide_register_hw() in media_bay_step().

* Unexport ide_register_hw() and make it static.

v2:
* Fix build by adding <linux/ide.h> include to <asm-powerpc/mediabay.h>.
  (Reported by Michael/Kamalesh/Andrew).

Cc: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>
Cc: Michael Ellerman <michael@ellerman.id.au>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/ide-probe.c        | 18 ++++++++++++++++++
 drivers/ide/ide.c              | 22 ++++++++++++++++------
 drivers/ide/ppc/pmac.c         |  3 ++-
 drivers/macintosh/mediabay.c   | 17 +++++++----------
 include/asm-powerpc/mediabay.h |  6 +++++-
 include/linux/ide.h            |  6 ++----
 6 files changed, 50 insertions(+), 22 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c
index f3ee098b69c7..468c4ac4181d 100644
--- a/drivers/ide/ide-probe.c
+++ b/drivers/ide/ide-probe.c
@@ -1494,3 +1494,21 @@ int ide_device_add(u8 idx[4], const struct ide_port_info *d)
 	return ide_device_add_all(idx_all, d);
 }
 EXPORT_SYMBOL_GPL(ide_device_add);
+
+void ide_port_scan(ide_hwif_t *hwif)
+{
+	ide_port_cable_detect(hwif);
+	ide_port_init_devices(hwif);
+
+	if (ide_probe_port(hwif) < 0)
+		return;
+
+	hwif->present = 1;
+
+	ide_port_tune_devices(hwif);
+	ide_acpi_port_init_devices(hwif);
+	ide_port_setup_devices(hwif);
+	hwif_register_devices(hwif);
+	ide_proc_port_register_devices(hwif);
+}
+EXPORT_SYMBOL_GPL(ide_port_scan);
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index 47c44d15ad4a..a8b5b08dd6e2 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -502,7 +502,7 @@ void ide_remove_port_from_hwgroup(ide_hwif_t *hwif)
 }
 
 /* Called with ide_lock held. */
-static void ide_port_unregister_devices(ide_hwif_t *hwif)
+static void __ide_port_unregister_devices(ide_hwif_t *hwif)
 {
 	int i;
 
@@ -518,6 +518,18 @@ static void ide_port_unregister_devices(ide_hwif_t *hwif)
 	}
 }
 
+void ide_port_unregister_devices(ide_hwif_t *hwif)
+{
+	mutex_lock(&ide_cfg_mtx);
+	spin_lock_irq(&ide_lock);
+	__ide_port_unregister_devices(hwif);
+	hwif->present = 0;
+	ide_port_init_devices_data(hwif);
+	spin_unlock_irq(&ide_lock);
+	mutex_unlock(&ide_cfg_mtx);
+}
+EXPORT_SYMBOL_GPL(ide_port_unregister_devices);
+
 /**
  *	ide_unregister		-	free an IDE interface
  *	@index: index of interface (will change soon to a pointer)
@@ -558,7 +570,7 @@ void ide_unregister(unsigned int index, int init_default, int restore)
 	hwif = &ide_hwifs[index];
 	if (!hwif->present)
 		goto abort;
-	ide_port_unregister_devices(hwif);
+	__ide_port_unregister_devices(hwif);
 	hwif->present = 0;
 
 	spin_unlock_irq(&ide_lock);
@@ -648,8 +660,8 @@ EXPORT_SYMBOL_GPL(ide_init_port_hw);
  *	Returns -1 on error.
  */
 
-int ide_register_hw(hw_regs_t *hw, void (*quirkproc)(ide_drive_t *),
-		    ide_hwif_t **hwifp)
+static int ide_register_hw(hw_regs_t *hw, void (*quirkproc)(ide_drive_t *),
+			   ide_hwif_t **hwifp)
 {
 	int index, retry = 1;
 	ide_hwif_t *hwif;
@@ -683,8 +695,6 @@ found:
 	return hwif->present ? index : -1;
 }
 
-EXPORT_SYMBOL(ide_register_hw);
-
 /*
  *	Locks for IDE setting functionality
  */
diff --git a/drivers/ide/ppc/pmac.c b/drivers/ide/ppc/pmac.c
index d9ca52e6cdab..7889f5f5c49a 100644
--- a/drivers/ide/ppc/pmac.c
+++ b/drivers/ide/ppc/pmac.c
@@ -1088,7 +1088,8 @@ pmac_ide_setup_device(pmac_ide_hwif_t *pmif, ide_hwif_t *hwif, hw_regs_t *hw)
 	if (np->parent && np->parent->name
 	    && strcasecmp(np->parent->name, "media-bay") == 0) {
 #ifdef CONFIG_PMAC_MEDIABAY
-		media_bay_set_ide_infos(np->parent, pmif->regbase, pmif->irq, hwif->index);
+		media_bay_set_ide_infos(np->parent, pmif->regbase, pmif->irq,
+					hwif);
 #endif /* CONFIG_PMAC_MEDIABAY */
 		pmif->mediabay = 1;
 		if (!bidp)
diff --git a/drivers/macintosh/mediabay.c b/drivers/macintosh/mediabay.c
index bd8a1d14b45d..82add26cc665 100644
--- a/drivers/macintosh/mediabay.c
+++ b/drivers/macintosh/mediabay.c
@@ -79,6 +79,7 @@ struct media_bay_info {
 	int				sleeping;
 	struct semaphore		lock;
 #ifdef CONFIG_BLK_DEV_IDE_PMAC
+	ide_hwif_t			*cd_port;
 	void __iomem			*cd_base;
 	int				cd_irq;
 	int				cd_retry;
@@ -448,7 +449,7 @@ int check_media_bay_by_base(unsigned long base, int what)
 }
 
 int media_bay_set_ide_infos(struct device_node* which_bay, unsigned long base,
-			    int irq, int index)
+			    int irq, ide_hwif_t *hwif)
 {
 	int	i;
 
@@ -456,10 +457,11 @@ int media_bay_set_ide_infos(struct device_node* which_bay, unsigned long base,
 		struct media_bay_info* bay = &media_bays[i];
 
 		if (bay->mdev && which_bay == bay->mdev->ofdev.node) {
-			int timeout = 5000;
+			int timeout = 5000, index = hwif->index;
 			
 			down(&bay->lock);
 
+			bay->cd_port	= hwif;
  			bay->cd_base	= (void __iomem *) base;
 			bay->cd_irq	= irq;
 
@@ -551,15 +553,10 @@ static void media_bay_step(int i)
 			bay->timer = 0;
 			bay->state = mb_up;
 			if (bay->cd_index < 0) {
-				hw_regs_t hw;
-
 				printk("mediabay %d, registering IDE...\n", i);
 				pmu_suspend();
-				ide_init_hwif_ports(&hw, (unsigned long) bay->cd_base, (unsigned long) 0, NULL);
-				hw.irq = bay->cd_irq;
-				hw.chipset = ide_pmac;
-				bay->cd_index =
-					ide_register_hw(&hw, NULL, NULL);
+				ide_port_scan(bay->cd_port);
+				bay->cd_index = bay->cd_port->index;
 				pmu_resume();
 			}
 			if (bay->cd_index == -1) {
@@ -589,7 +586,7 @@ static void media_bay_step(int i)
     	        if (bay->cd_index >= 0) {
 			printk(KERN_DEBUG "Unregistering mb %d ide, index:%d\n", i,
 			       bay->cd_index);
-			ide_unregister(bay->cd_index, 1, 1);
+			ide_port_unregister_devices(bay->cd_port);
 			bay->cd_index = -1;
 		}
 	    	if (bay->cd_retry) {
diff --git a/include/asm-powerpc/mediabay.h b/include/asm-powerpc/mediabay.h
index de83fe196309..df111c362a7f 100644
--- a/include/asm-powerpc/mediabay.h
+++ b/include/asm-powerpc/mediabay.h
@@ -22,10 +22,14 @@ int check_media_bay(struct device_node *which_bay, int what);
 /* Number of bays in the machine or 0 */
 extern int media_bay_count;
 
+#ifdef CONFIG_BLK_DEV_IDE_PMAC
+#include <linux/ide.h>
+
 int check_media_bay_by_base(unsigned long base, int what);
 /* called by IDE PMAC host driver to register IDE controller for media bay */
 int media_bay_set_ide_infos(struct device_node *which_bay, unsigned long base,
-			    int irq, int index);
+			    int irq, ide_hwif_t *hwif);
+#endif
 
 #endif /* __KERNEL__ */
 #endif /* _PPC_MEDIABAY_H */
diff --git a/include/linux/ide.h b/include/linux/ide.h
index f9449ecd79d9..9aaad7e70593 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -189,10 +189,6 @@ struct hwif_s * ide_find_port(unsigned long);
 void ide_init_port_data(struct hwif_s *, unsigned int);
 void ide_init_port_hw(struct hwif_s *, hw_regs_t *);
 
-struct ide_drive_s;
-int ide_register_hw(hw_regs_t *, void (*)(struct ide_drive_s *),
-		    struct hwif_s **);
-
 static inline void ide_std_init_ports(hw_regs_t *hw,
 				      unsigned long io_addr,
 				      unsigned long ctl_addr)
@@ -1204,6 +1200,8 @@ void ide_undecoded_slave(ide_drive_t *);
 
 int ide_device_add_all(u8 *idx, const struct ide_port_info *);
 int ide_device_add(u8 idx[4], const struct ide_port_info *);
+void ide_port_unregister_devices(ide_hwif_t *);
+void ide_port_scan(ide_hwif_t *);
 
 static inline void *ide_get_hwifdata (ide_hwif_t * hwif)
 {
-- 
cgit v1.2.3


From 50672e5d7486c9ab312432cbe180ac071f1de8e0 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:23 +0200
Subject: ide: remove dead/obsolete ->busproc method

->busproc method is used by HDIO_SET_BUSSTATE ioctl but it has no chance
of working as intended (in 2.4.x days) because to issue an ioctl there
is a device node needed and:

- for BUSSTATE_TRISTATE+OFF it is too late (devices are already gone)

- for BUSSTATE_TRISTATE+ON it is too early (devices are not registered yet)

Just remove ->busproc method for now (it was only implemented by hpt366,
siimage and tc86c001 host drivers).

Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/ide.c          |  3 ---
 drivers/ide/pci/hpt366.c   | 59 ----------------------------------------------
 drivers/ide/pci/sgiioc4.c  |  1 -
 drivers/ide/pci/siimage.c  | 43 ---------------------------------
 drivers/ide/pci/tc86c001.c | 36 ----------------------------
 include/linux/ide.h        |  2 --
 6 files changed, 144 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index a8b5b08dd6e2..cb18ba8de22d 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -414,7 +414,6 @@ static void ide_hwif_restore(ide_hwif_t *hwif, ide_hwif_t *tmp_hwif)
 	hwif->resetproc			= tmp_hwif->resetproc;
 	hwif->maskproc			= tmp_hwif->maskproc;
 	hwif->quirkproc			= tmp_hwif->quirkproc;
-	hwif->busproc			= tmp_hwif->busproc;
 
 	hwif->ata_input_data		= tmp_hwif->ata_input_data;
 	hwif->ata_output_data		= tmp_hwif->ata_output_data;
@@ -1071,8 +1070,6 @@ int generic_ide_ioctl(ide_drive_t *drive, struct file *file, struct block_device
 		case HDIO_SET_BUSSTATE:
 			if (!capable(CAP_SYS_ADMIN))
 				return -EACCES;
-			if (HWIF(drive)->busproc)
-				return HWIF(drive)->busproc(drive, (int)arg);
 			return -EOPNOTSUPP;
 		default:
 			return -EINVAL;
diff --git a/drivers/ide/pci/hpt366.c b/drivers/ide/pci/hpt366.c
index 6357bb6269ab..d03a231d965e 100644
--- a/drivers/ide/pci/hpt366.c
+++ b/drivers/ide/pci/hpt366.c
@@ -929,64 +929,6 @@ static void hpt3xxn_rw_disk(ide_drive_t *drive, struct request *rq)
 	hpt3xxn_set_clock(HWIF(drive), rq_data_dir(rq) ? 0x23 : 0x21);
 }
 
-/* 
- * Set/get power state for a drive.
- * NOTE: affects both drives on each channel.
- *
- * When we turn the power back on, we need to re-initialize things.
- */
-#define TRISTATE_BIT  0x8000
-
-static int hpt3xx_busproc(ide_drive_t *drive, int state)
-{
-	ide_hwif_t *hwif	= HWIF(drive);
-	struct pci_dev *dev	= to_pci_dev(hwif->dev);
-	u8  mcr_addr		= hwif->select_data + 2;
-	u8  resetmask		= hwif->channel ? 0x80 : 0x40;
-	u8  bsr2		= 0;
-	u16 mcr			= 0;
-
-	hwif->bus_state = state;
-
-	/* Grab the status. */
-	pci_read_config_word(dev, mcr_addr, &mcr);
-	pci_read_config_byte(dev, 0x59, &bsr2);
-
-	/*
-	 * Set the state. We don't set it if we don't need to do so.
-	 * Make sure that the drive knows that it has failed if it's off.
-	 */
-	switch (state) {
-	case BUSSTATE_ON:
-		if (!(bsr2 & resetmask))
-			return 0;
-		hwif->drives[0].failures = hwif->drives[1].failures = 0;
-
-		pci_write_config_byte(dev, 0x59, bsr2 & ~resetmask);
-		pci_write_config_word(dev, mcr_addr, mcr & ~TRISTATE_BIT);
-		return 0;
-	case BUSSTATE_OFF:
-		if ((bsr2 & resetmask) && !(mcr & TRISTATE_BIT))
-			return 0;
-		mcr &= ~TRISTATE_BIT;
-		break;
-	case BUSSTATE_TRISTATE:
-		if ((bsr2 & resetmask) &&  (mcr & TRISTATE_BIT))
-			return 0;
-		mcr |= TRISTATE_BIT;
-		break;
-	default:
-		return -EINVAL;
-	}
-
-	hwif->drives[0].failures = hwif->drives[0].max_failures + 1;
-	hwif->drives[1].failures = hwif->drives[1].max_failures + 1;
-
-	pci_write_config_word(dev, mcr_addr, mcr);
-	pci_write_config_byte(dev, 0x59, bsr2 | resetmask);
-	return 0;
-}
-
 /**
  *	hpt37x_calibrate_dpll	-	calibrate the DPLL
  *	@dev: PCI device
@@ -1334,7 +1276,6 @@ static void __devinit init_hwif_hpt366(ide_hwif_t *hwif)
 
 	hwif->quirkproc		= &hpt3xx_quirkproc;
 	hwif->maskproc		= &hpt3xx_maskproc;
-	hwif->busproc		= &hpt3xx_busproc;
 
 	hwif->udma_filter	= &hpt3xx_udma_filter;
 	hwif->mdma_filter	= &hpt3xx_mdma_filter;
diff --git a/drivers/ide/pci/sgiioc4.c b/drivers/ide/pci/sgiioc4.c
index 054626497be4..9046a69117ff 100644
--- a/drivers/ide/pci/sgiioc4.c
+++ b/drivers/ide/pci/sgiioc4.c
@@ -562,7 +562,6 @@ ide_init_sgiioc4(ide_hwif_t * hwif)
 						clear interrupts */
 	hwif->maskproc = &sgiioc4_maskproc;	/* Mask on/off NIEN register */
 	hwif->quirkproc = NULL;
-	hwif->busproc = NULL;
 
 	hwif->INB = &sgiioc4_INB;
 
diff --git a/drivers/ide/pci/siimage.c b/drivers/ide/pci/siimage.c
index 8d624afe8529..b6be1b45f329 100644
--- a/drivers/ide/pci/siimage.c
+++ b/drivers/ide/pci/siimage.c
@@ -369,48 +369,6 @@ static int siimage_mmio_ide_dma_test_irq (ide_drive_t *drive)
 	return 0;
 }
 
-/**
- *	sil_sata_busproc	-	bus isolation IOCTL
- *	@drive: drive to isolate/restore
- *	@state: bus state to set
- *
- *	Used by the SII3112 to handle bus isolation. As this is a 
- *	SATA controller the work required is quite limited, we 
- *	just have to clean up the statistics
- */
-
-static int sil_sata_busproc(ide_drive_t * drive, int state)
-{
-	ide_hwif_t *hwif	= HWIF(drive);
-	struct pci_dev *dev	= to_pci_dev(hwif->dev);
-	u32 stat_config		= 0;
-	unsigned long addr	= siimage_selreg(hwif, 0);
-
-	if (hwif->mmio)
-		stat_config = readl((void __iomem *)addr);
-	else
-		pci_read_config_dword(dev, addr, &stat_config);
-
-	switch (state) {
-		case BUSSTATE_ON:
-			hwif->drives[0].failures = 0;
-			hwif->drives[1].failures = 0;
-			break;
-		case BUSSTATE_OFF:
-			hwif->drives[0].failures = hwif->drives[0].max_failures + 1;
-			hwif->drives[1].failures = hwif->drives[1].max_failures + 1;
-			break;
-		case BUSSTATE_TRISTATE:
-			hwif->drives[0].failures = hwif->drives[0].max_failures + 1;
-			hwif->drives[1].failures = hwif->drives[1].max_failures + 1;
-			break;
-		default:
-			return -EINVAL;
-	}
-	hwif->bus_state = state;
-	return 0;
-}
-
 /**
  *	sil_sata_reset_poll	-	wait for SATA reset
  *	@drive: drive we are resetting
@@ -818,7 +776,6 @@ static void __devinit init_hwif_siimage(ide_hwif_t *hwif)
 	if (sata) {
 		static int first = 1;
 
-		hwif->busproc = &sil_sata_busproc;
 		hwif->reset_poll = &sil_sata_reset_poll;
 		hwif->pre_reset = &sil_sata_pre_reset;
 		hwif->udma_filter = &sil_sata_udma_filter;
diff --git a/drivers/ide/pci/tc86c001.c b/drivers/ide/pci/tc86c001.c
index 2ef2ed2f2b32..1e4a6262bcef 100644
--- a/drivers/ide/pci/tc86c001.c
+++ b/drivers/ide/pci/tc86c001.c
@@ -126,40 +126,6 @@ static void tc86c001_dma_start(ide_drive_t *drive)
 	ide_dma_start(drive);
 }
 
-static int tc86c001_busproc(ide_drive_t *drive, int state)
-{
-	ide_hwif_t *hwif	= HWIF(drive);
-	unsigned long sc_base	= hwif->config_data;
-	u16 scr1;
-
-	/* System Control 1 Register bit 11 (ATA Hard Reset) read */
-	scr1 = inw(sc_base + 0x00);
-
-	switch (state) {
-		case BUSSTATE_ON:
-			if (!(scr1 & 0x0800))
-				return 0;
-			scr1 &= ~0x0800;
-
-			hwif->drives[0].failures = hwif->drives[1].failures = 0;
-			break;
-		case BUSSTATE_OFF:
-			if (scr1 & 0x0800)
-				return 0;
-			scr1 |= 0x0800;
-
-			hwif->drives[0].failures = hwif->drives[0].max_failures + 1;
-			hwif->drives[1].failures = hwif->drives[1].max_failures + 1;
-			break;
-		default:
-			return -EINVAL;
-	}
-
-	/* System Control 1 Register bit 11 (ATA Hard Reset) write */
-	outw(scr1, sc_base + 0x00);
-	return 0;
-}
-
 static u8 __devinit tc86c001_cable_detect(ide_hwif_t *hwif)
 {
 	struct pci_dev *dev = to_pci_dev(hwif->dev);
@@ -194,8 +160,6 @@ static void __devinit init_hwif_tc86c001(ide_hwif_t *hwif)
 	hwif->set_pio_mode = &tc86c001_set_pio_mode;
 	hwif->set_dma_mode = &tc86c001_set_mode;
 
-	hwif->busproc	= &tc86c001_busproc;
-
 	hwif->cable_detect = tc86c001_cable_detect;
 
 	if (!hwif->dma_base)
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 9aaad7e70593..e43570a19200 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -502,8 +502,6 @@ typedef struct hwif_s {
 	void	(*maskproc)(ide_drive_t *, int);
 	/* check host's drive quirk list */
 	void	(*quirkproc)(ide_drive_t *);
-	/* driver soft-power interface */
-	int	(*busproc)(ide_drive_t *, int);
 #endif
 	u8 (*mdma_filter)(ide_drive_t *);
 	u8 (*udma_filter)(ide_drive_t *);
-- 
cgit v1.2.3


From f74c91413ec6140ee0553180c5f56fdd27c22a2e Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:23 +0200
Subject: ide: add warm-plug support for IDE devices (take 2)

* Add 'struct class ide_port_class' ('ide_port' class) and a 'struct
  device *portdev' ('ide_port' class device) in ide_hwif_t.

* Register 'ide_port' class in ide_init() and unregister it in
  cleanup_module().

* Create ->portdev in ide_register_port () and unregister it in
  ide_unregister().

* Add "delete_devices" class device attribute for unregistering IDE devices
  on a port and "scan" one for probing+registering IDE devices on a port.

* Add ide_sysfs_register_port() helper for registering "delete_devices"
  and "scan" attributes with ->portdev.  Call it in ide_device_add_all().

* Document IDE warm-plug support in Documentation/ide/warm-plug-howto.txt.

v2:
* Convert patch from using 'struct class_device' to use 'struct device'.
  (thanks to Kay Sievers for doing it)

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 Documentation/ide/warm-plug-howto.txt | 13 +++++++
 drivers/ide/ide-probe.c               | 71 ++++++++++++++++++++++++++++++++++-
 drivers/ide/ide.c                     | 24 ++++++++++++
 include/linux/ide.h                   |  5 ++-
 4 files changed, 110 insertions(+), 3 deletions(-)
 create mode 100644 Documentation/ide/warm-plug-howto.txt

(limited to 'include/linux')

diff --git a/Documentation/ide/warm-plug-howto.txt b/Documentation/ide/warm-plug-howto.txt
new file mode 100644
index 000000000000..d5885468b072
--- /dev/null
+++ b/Documentation/ide/warm-plug-howto.txt
@@ -0,0 +1,13 @@
+
+IDE warm-plug HOWTO
+===================
+
+To warm-plug devices on a port 'idex':
+
+# echo -n "1" > /sys/class/ide_port/idex/delete_devices
+
+unplug old device(s) and plug new device(s)
+
+# echo -n "1" > /sys/class/ide_port/idex/scan
+
+done
diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c
index 468c4ac4181d..510254ab3c9b 100644
--- a/drivers/ide/ide-probe.c
+++ b/drivers/ide/ide-probe.c
@@ -623,7 +623,7 @@ static void hwif_release_dev (struct device *dev)
 	complete(&hwif->gendev_rel_comp);
 }
 
-static void ide_register_port(ide_hwif_t *hwif)
+static int ide_register_port(ide_hwif_t *hwif)
 {
 	int ret;
 
@@ -639,9 +639,23 @@ static void ide_register_port(ide_hwif_t *hwif)
 	}
 	hwif->gendev.release = hwif_release_dev;
 	ret = device_register(&hwif->gendev);
-	if (ret < 0)
+	if (ret < 0) {
 		printk(KERN_WARNING "IDE: %s: device_register error: %d\n",
 			__FUNCTION__, ret);
+		goto out;
+	}
+
+	get_device(&hwif->gendev);
+
+	hwif->portdev = device_create(ide_port_class, &hwif->gendev,
+				      MKDEV(0, 0), hwif->name);
+	if (IS_ERR(hwif->portdev)) {
+		ret = PTR_ERR(hwif->portdev);
+		device_unregister(&hwif->gendev);
+	}
+	dev_set_drvdata(hwif->portdev, hwif);
+out:
+	return ret;
 }
 
 /**
@@ -1378,6 +1392,58 @@ static void ide_port_cable_detect(ide_hwif_t *hwif)
 	}
 }
 
+static ssize_t store_delete_devices(struct device *portdev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t n)
+{
+	ide_hwif_t *hwif = dev_get_drvdata(portdev);
+
+	if (strncmp(buf, "1", n))
+		return -EINVAL;
+
+	ide_port_unregister_devices(hwif);
+
+	return n;
+};
+
+static DEVICE_ATTR(delete_devices, S_IWUSR, NULL, store_delete_devices);
+
+static ssize_t store_scan(struct device *portdev,
+			  struct device_attribute *attr,
+			  const char *buf, size_t n)
+{
+	ide_hwif_t *hwif = dev_get_drvdata(portdev);
+
+	if (strncmp(buf, "1", n))
+		return -EINVAL;
+
+	ide_port_unregister_devices(hwif);
+	ide_port_scan(hwif);
+
+	return n;
+};
+
+static DEVICE_ATTR(scan, S_IWUSR, NULL, store_scan);
+
+static struct device_attribute *ide_port_attrs[] = {
+	&dev_attr_delete_devices,
+	&dev_attr_scan,
+	NULL
+};
+
+static int ide_sysfs_register_port(ide_hwif_t *hwif)
+{
+	int i, rc;
+
+	for (i = 0; ide_port_attrs[i]; i++) {
+		rc = device_create_file(hwif->portdev, ide_port_attrs[i]);
+		if (rc)
+			break;
+	}
+
+	return rc;
+}
+
 int ide_device_add_all(u8 *idx, const struct ide_port_info *d)
 {
 	ide_hwif_t *hwif, *mate = NULL;
@@ -1474,6 +1540,7 @@ int ide_device_add_all(u8 *idx, const struct ide_port_info *d)
 		hwif = &ide_hwifs[idx[i]];
 
 		if (hwif->present) {
+			ide_sysfs_register_port(hwif);
 			ide_proc_register_port(hwif);
 			ide_proc_port_register_devices(hwif);
 		}
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index cb18ba8de22d..d791b1ffb586 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -78,6 +78,8 @@
 /* default maximum number of failures */
 #define IDE_DEFAULT_MAX_FAILURES 	1
 
+struct class *ide_port_class;
+
 static const u8 ide_hwif_to_major[] = { IDE0_MAJOR, IDE1_MAJOR,
 					IDE2_MAJOR, IDE3_MAJOR,
 					IDE4_MAJOR, IDE5_MAJOR,
@@ -591,6 +593,7 @@ void ide_unregister(unsigned int index, int init_default, int restore)
 
 	ide_remove_port_from_hwgroup(hwif);
 
+	device_unregister(hwif->portdev);
 	device_unregister(&hwif->gendev);
 	wait_for_completion(&hwif->gendev_rel_comp);
 
@@ -1590,6 +1593,13 @@ struct bus_type ide_bus_type = {
 
 EXPORT_SYMBOL_GPL(ide_bus_type);
 
+static void ide_port_class_release(struct device *portdev)
+{
+	ide_hwif_t *hwif = dev_get_drvdata(portdev);
+
+	put_device(&hwif->gendev);
+}
+
 /*
  * This is gets invoked once during initialization, to set *everything* up
  */
@@ -1610,11 +1620,23 @@ static int __init ide_init(void)
 		return ret;
 	}
 
+	ide_port_class = class_create(THIS_MODULE, "ide_port");
+	if (IS_ERR(ide_port_class)) {
+		ret = PTR_ERR(ide_port_class);
+		goto out_port_class;
+	}
+	ide_port_class->dev_release = ide_port_class_release;
+
 	init_ide_data();
 
 	proc_ide_create();
 
 	return 0;
+
+out_port_class:
+	bus_unregister(&ide_bus_type);
+
+	return ret;
 }
 
 #ifdef MODULE
@@ -1651,6 +1673,8 @@ void __exit cleanup_module (void)
 
 	proc_ide_destroy();
 
+	class_destroy(ide_port_class);
+
 	bus_unregister(&ide_bus_type);
 }
 
diff --git a/include/linux/ide.h b/include/linux/ide.h
index e43570a19200..9cebf3054080 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -579,7 +579,9 @@ typedef struct hwif_s {
 	unsigned	mmio       : 1; /* host uses MMIO */
 	unsigned	straight8  : 1;	/* Alan's straight 8 check */
 
-	struct device	gendev;
+	struct device		gendev;
+	struct device		*portdev;
+
 	struct completion gendev_rel_comp; /* To deal with device release() */
 
 	void		*hwif_data;	/* extra hwif data */
@@ -1275,6 +1277,7 @@ extern struct mutex ide_cfg_mtx;
 #define local_irq_set(flags)	do { local_save_flags((flags)); local_irq_enable_in_hardirq(); } while (0)
 
 extern struct bus_type ide_bus_type;
+extern struct class *ide_port_class;
 
 /* check if CACHE FLUSH (EXT) command is supported (bits defined in ATA-6) */
 #define ide_id_has_flush_cache(id)	((id)->cfs_enable_2 & 0x3000)
-- 
cgit v1.2.3


From 9a0e77f28b50128df0c9e26ae489e44e29a7270a Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:24 +0200
Subject: ide: remove obsoleted "idex=base[,ctl[,irq]]" kernel parameters (take
 2)

* Remove obsoleted "idex=base[,ctl[,irq]]" kernel parameters
  and update Documentation/ide/ide.txt.

* Remove no longer needed ide_forced chipset type.

v2:
* is_chipset_set[] -> is_chipset_set in ide.c.

* Documentation/ide/ide.txt fix.

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 Documentation/ide/ide.txt | 33 +--------------------------------
 drivers/ide/ide-generic.c |  3 +--
 drivers/ide/ide-probe.c   |  3 +--
 drivers/ide/ide-proc.c    |  3 ---
 drivers/ide/ide.c         | 41 ++++++++++++-----------------------------
 drivers/ide/pci/cmd640.c  |  2 +-
 drivers/ide/setup-pci.c   | 11 -----------
 include/linux/ide.h       |  2 +-
 8 files changed, 17 insertions(+), 81 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/ide/ide.txt b/Documentation/ide/ide.txt
index ae4f4f43c25c..18c02df2f78f 100644
--- a/Documentation/ide/ide.txt
+++ b/Documentation/ide/ide.txt
@@ -71,29 +71,6 @@ This driver automatically probes for most IDE interfaces (including all PCI
 ones), for the drives/geometries attached to those interfaces, and for the IRQ
 lines being used by the interfaces (normally 14, 15 for ide0/ide1).
 
-For special cases, interfaces may be specified using kernel "command line"
-options.  For example,
-
-	ide3=0x168,0x36e,10	/* ioports 0x168-0x16f,0x36e, irq 10 */
-
-Normally the irq number need not be specified, as ide.c will probe for it:
-
-	ide3=0x168,0x36e	/* ioports 0x168-0x16f,0x36e */
-
-The standard port, and irq values are these:
-
-	ide0=0x1f0,0x3f6,14
-	ide1=0x170,0x376,15
-	ide2=0x1e8,0x3ee,11
-	ide3=0x168,0x36e,10
-
-Note that the first parameter reserves 8 contiguous ioports, whereas the
-second value denotes a single ioport. If in doubt, do a 'cat /proc/ioports'.
-
-In all probability the device uses these ports and IRQs if it is attached
-to the appropriate ide channel.  Pass the parameter for the correct ide
-channel to the kernel, as explained above.
-
 Any number of interfaces may share a single IRQ if necessary, at a slight
 performance penalty, whether on separate cards or a single VLB card.
 The IDE driver automatically detects and handles this.  However, this may
@@ -199,7 +176,7 @@ When ide.c is used as a module, you can pass command line parameters to the
 driver using the "options=" keyword to insmod, while replacing any ',' with
 ';'.  For example:
 
-	insmod ide.o options="ide0=serialize ide1=serialize ide2=0x1e8;0x3ee;11"
+	insmod ide.o options="hda=nodma hdb=nodma"
 
 
 ================================================================================
@@ -240,14 +217,6 @@ Summary of ide driver parameters for kernel command line
 			  As for VLB, it is safest to not specify it.
 			  Bigger values are safer than smaller ones.
 
- "idex=base"		: probe for an interface at the addr specified,
-			  where "base" is usually 0x1f0 or 0x170
-			  and "ctl" is assumed to be "base"+0x206
-
- "idex=base,ctl"	: specify both base and ctl
-
- "idex=base,ctl,irq"	: specify base, ctl, and irq number
-
  "idex=serialize"	: do not overlap operations on idex. Please note
 			  that you will have to specify this option for
 			  both the respective primary and secondary channel
diff --git a/drivers/ide/ide-generic.c b/drivers/ide/ide-generic.c
index 387574fe4b2b..bae41459192d 100644
--- a/drivers/ide/ide-generic.c
+++ b/drivers/ide/ide-generic.c
@@ -93,8 +93,7 @@ static int __init ide_generic_init(void)
 		ide_hwif_t *hwif = &ide_hwifs[i];
 
 		if (hwif->io_ports[IDE_DATA_OFFSET] &&
-		    (hwif->chipset == ide_unknown ||
-		     hwif->chipset == ide_forced))
+		    hwif->chipset == ide_unknown)
 			idx[i] = i;
 		else
 			idx[i] = 0xff;
diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c
index 8ef5194f6d47..33cb5e5a249b 100644
--- a/drivers/ide/ide-probe.c
+++ b/drivers/ide/ide-probe.c
@@ -1519,8 +1519,7 @@ int ide_device_add_all(u8 *idx, const struct ide_port_info *d)
 		hwif = &ide_hwifs[idx[i]];
 
 		if (hwif->present) {
-			if (hwif->chipset == ide_unknown ||
-			    hwif->chipset == ide_forced)
+			if (hwif->chipset == ide_unknown)
 				hwif->chipset = ide_generic;
 			hwif_register_devices(hwif);
 		}
diff --git a/drivers/ide/ide-proc.c b/drivers/ide/ide-proc.c
index 77025d1057b5..edd7f186dc4d 100644
--- a/drivers/ide/ide-proc.c
+++ b/drivers/ide/ide-proc.c
@@ -46,9 +46,6 @@ static int proc_ide_read_imodel
 	int		len;
 	const char	*name;
 
-	/*
-	 * Neither ide_unknown nor ide_forced should be set at this point.
-	 */
 	switch (hwif->chipset) {
 		case ide_generic:	name = "generic";	break;
 		case ide_pci:		name = "pci";		break;
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index 96126b3b12e9..7e789c97a8b8 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -1170,7 +1170,7 @@ extern int probe_ht6560b;
 extern int probe_qd65xx;
 extern int cmd640_vlb;
 
-static int __initdata is_chipset_set[MAX_HWIFS];
+static int __initdata is_chipset_set;
 
 /*
  * ide_setup() gets called VERY EARLY during initialization,
@@ -1328,8 +1328,6 @@ static int __init ide_setup(char *s)
 			"minus10", "four", "qd65xx", "ht6560b", "cmd640_vlb",
 			"dtc2278", "umc8672", "ali14xx", NULL };
 
-		hw_regs_t hwregs;
-
 		hw = s[3] - '0';
 		hwif = &ide_hwifs[hw];
 		i = match_parm(&s[4], ide_words, vals, 3);
@@ -1338,19 +1336,14 @@ static int __init ide_setup(char *s)
 		 * Cryptic check to ensure chipset not already set for hwif.
 		 * Note: we can't depend on hwif->chipset here.
 		 */
-		if ((i >= -18 && i <= -11) || (i > 0 && i <= 3)) {
+		if (i >= -18 && i <= -11) {
 			/* chipset already specified */
-			if (is_chipset_set[hw])
+			if (is_chipset_set)
 				goto bad_option;
-			if (i > -18 && i <= -11) {
-				/* these drivers are for "ide0=" only */
-				if (hw != 0)
-					goto bad_hwif;
-				/* chipset already specified for 2nd port */
-				if (is_chipset_set[hw+1])
-					goto bad_option;
-			}
-			is_chipset_set[hw] = 1;
+			/* these drivers are for "ide0=" only */
+			if (hw != 0)
+				goto bad_hwif;
+			is_chipset_set = 1;
 			printk("\n");
 		}
 
@@ -1430,21 +1423,11 @@ static int __init ide_setup(char *s)
 			case -1: /* "noprobe" */
 				hwif->noprobe = 1;
 				goto obsolete_option;
-
-			case 1:	/* base */
-				vals[1] = vals[0] + 0x206; /* default ctl */
-			case 2: /* base,ctl */
-				vals[2] = 0;	/* default irq = probe for it */
-			case 3: /* base,ctl,irq */
-				memset(&hwregs, 0, sizeof(hwregs));
-				ide_init_hwif_ports(&hwregs, vals[0], vals[1], &hwif->irq);
-				memcpy(hwif->io_ports, hwregs.io_ports, sizeof(hwif->io_ports));
-				hwif->irq      = vals[2];
-				hwif->noprobe  = 0;
-				hwif->chipset  = ide_forced;
-				goto obsolete_option;
-
-			case 0: goto bad_option;
+			case 0:
+			case 1:
+			case 2:
+			case 3:
+				goto bad_option;
 			default:
 				printk(" -- SUPPORT NOT CONFIGURED IN THIS KERNEL\n");
 				return 1;
diff --git a/drivers/ide/pci/cmd640.c b/drivers/ide/pci/cmd640.c
index 29fbc5ead03b..58a95f62e383 100644
--- a/drivers/ide/pci/cmd640.c
+++ b/drivers/ide/pci/cmd640.c
@@ -415,7 +415,7 @@ static void __init setup_device_ptrs (void)
 	cmd_hwif1 = &ide_hwifs[1]; /* default, if not found below */
 	for (i = 0; i < MAX_HWIFS; i++) {
 		ide_hwif_t *hwif = &ide_hwifs[i];
-		if (hwif->chipset == ide_unknown || hwif->chipset == ide_forced) {
+		if (hwif->chipset == ide_unknown) {
 			if (hwif->io_ports[IDE_DATA_OFFSET] == 0x1f0)
 				cmd_hwif0 = hwif;
 			else if (hwif->io_ports[IDE_DATA_OFFSET] == 0x170)
diff --git a/drivers/ide/setup-pci.c b/drivers/ide/setup-pci.c
index 634e3f6a9608..ea66c996e4ec 100644
--- a/drivers/ide/setup-pci.c
+++ b/drivers/ide/setup-pci.c
@@ -40,17 +40,6 @@ static ide_hwif_t *ide_match_hwif(unsigned long io_base, u8 bootable, const char
 	int h;
 	ide_hwif_t *hwif;
 
-	/*
-	 * Look for a hwif with matching io_base specified using
-	 * parameters to ide_setup().
-	 */
-	for (h = 0; h < MAX_HWIFS; ++h) {
-		hwif = &ide_hwifs[h];
-		if (hwif->io_ports[IDE_DATA_OFFSET] == io_base) {
-			if (hwif->chipset == ide_forced)
-				return hwif; /* a perfect match */
-		}
-	}
 	/*
 	 * Look for a hwif with matching io_base default value.
 	 * If chipset is "ide_unknown", then claim that hwif slot.
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 9cebf3054080..67f83c60845f 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -169,7 +169,7 @@ enum {		ide_unknown,	ide_generic,	ide_pci,
 		ide_rz1000,	ide_trm290,
 		ide_cmd646,	ide_cy82c693,	ide_4drives,
 		ide_pmac,	ide_etrax100,	ide_acorn,
-		ide_au1xxx,	ide_palm3710,	ide_forced
+		ide_au1xxx,	ide_palm3710
 };
 
 typedef u8 hwif_chipset_t;
-- 
cgit v1.2.3


From 93de00fd1c70e1a23a73a865e0f9abfe74a7a719 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:24 +0200
Subject: ide: remove broken/dangerous HDIO_[UNREGISTER,SCAN]_HWIF ioctls (take
 3)

hdparm explicitely marks HDIO_[UNREGISTER,SCAN]_HWIF ioctls as DANGEROUS
and given the number of bugs we can assume that there are no real users:

* DMA has no chance of working because DMA resources are released by
  ide_unregister() and they are never allocated again.

* Since ide_init_hwif_ports() is used for ->io_ports[] setup the ioctls
  don't work for almost all hosts with "non-standard" (== non ISA-like)
  layout of IDE taskfile registers (there is a lot of such host drivers).

* ide_port_init_devices() is not called when probing IDE devices so:
  - drive->autotune is never set and IDE host/devices are not programmed
    for the correct PIO/DMA transfer modes (=> possible data corruption)
  - host specific I/O 32-bit and IRQ unmasking settings are not applied
    (=> possible data corruption)
  - host specific ->port_init_devs method is not called (=> no luck with
    ht6560b, qd65xx and opti621 host drivers)

* ->rw_disk method is not preserved (=> no HPT3xxN chipsets support).

* ->serialized flag is not preserved (=> possible data corruption when
   using icside, aec62xx (ATP850UF chipset), cmd640, cs5530, hpt366
   (HPT3xxN chipsets), rz1000, sc1200, dtc2278 and ht6560b host drivers).

* ->ack_intr method is not preserved (=> needed by ide-cris, buddha,
  gayle and macide host drivers).

* ->sata_scr[] and sata_misc[] is cleared by ide_unregister() and it
  isn't initialized again (SiI3112 support needs them).

* To issue an ioctl() there need to be at least one IDE device present
  in the system.

* ->cable_detect method is not preserved + it is not called when probing
  IDE devices so cable detection is broken (however since DMA support is
  also broken it doesn't really matter ;-).

* Some objects which may have already been freed in ide_unregister()
  are restored by ide_hwif_restore() (i.e. ->hwgroup).

* ide_register_hw() may unregister unrelated IDE ports if free ide_hwifs[]
  slot cannot be found.

* When IDE host drivers are modular unregistered port may be re-used by
  different host driver that owned it first causing subtle bugs.

Since we now have a proper warm-plug support remove these ioctls,
then remove no longer needed:
- ide_register_hw() and ide_hwif_restore() functions
- 'init_default' and 'restore' arguments of ide_unregister()
- zeroeing of hwif->{dma,extra}_* fields in ide_unregister()

As an added bonus IDE core code size shrinks by ~3kB (x86-32).

v2:
* fix ide_unregister() arguments in cleanup_module() (Andrew Morton).

v3:
* fix ide_unregister() arguments in palm_bk3710.c.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 block/compat_ioctl.c              |   1 -
 drivers/ide/arm/bast-ide.c        |   2 +-
 drivers/ide/arm/palm_bk3710.c     |   2 +-
 drivers/ide/arm/rapide.c          |   2 +-
 drivers/ide/ide-pnp.c             |   2 +-
 drivers/ide/ide.c                 | 199 +-------------------------------------
 drivers/ide/legacy/ide-cs.c       |   4 +-
 drivers/ide/legacy/ide_platform.c |   2 +-
 drivers/ide/mips/au1xxx-ide.c     |   2 +-
 drivers/ide/pci/delkin_cb.c       |   4 +-
 drivers/ide/pci/scc_pata.c        |   2 +-
 include/linux/hdreg.h             |   4 +-
 include/linux/ide.h               |   2 +-
 13 files changed, 19 insertions(+), 209 deletions(-)

(limited to 'include/linux')

diff --git a/block/compat_ioctl.c b/block/compat_ioctl.c
index b73373216b0e..c70d0b6f666f 100644
--- a/block/compat_ioctl.c
+++ b/block/compat_ioctl.c
@@ -624,7 +624,6 @@ static int compat_blkdev_driver_ioctl(struct inode *inode, struct file *file,
 	case HDIO_GET_IDENTITY:
 	case HDIO_DRIVE_TASK:
 	case HDIO_DRIVE_CMD:
-	case HDIO_SCAN_HWIF:
 	/* 0x330 is reserved -- it used to be HDIO_GETGEO_BIG */
 	case 0x330:
 	/* 0x02 -- Floppy ioctls */
diff --git a/drivers/ide/arm/bast-ide.c b/drivers/ide/arm/bast-ide.c
index a22da7ab2b57..d2196436788b 100644
--- a/drivers/ide/arm/bast-ide.c
+++ b/drivers/ide/arm/bast-ide.c
@@ -48,7 +48,7 @@ static int __init bastide_register(unsigned int base, unsigned int aux, int irq)
 	i = hwif->index;
 
 	if (hwif->present)
-		ide_unregister(i, 0, 0);
+		ide_unregister(i);
 	else if (!hwif->hold)
 		ide_init_port_data(hwif, i);
 
diff --git a/drivers/ide/arm/palm_bk3710.c b/drivers/ide/arm/palm_bk3710.c
index 0a722503c102..73899ef4ab0d 100644
--- a/drivers/ide/arm/palm_bk3710.c
+++ b/drivers/ide/arm/palm_bk3710.c
@@ -385,7 +385,7 @@ static int __devinit palm_bk3710_probe(struct platform_device *pdev)
 	i = hwif->index;
 
 	if (hwif->present)
-		ide_unregister(i, 0, 0);
+		ide_unregister(i);
 	else if (!hwif->hold)
 		ide_init_port_data(hwif, i);
 
diff --git a/drivers/ide/arm/rapide.c b/drivers/ide/arm/rapide.c
index efba00d2fc37..b30adcf321c3 100644
--- a/drivers/ide/arm/rapide.c
+++ b/drivers/ide/arm/rapide.c
@@ -76,7 +76,7 @@ static void __devexit rapide_remove(struct expansion_card *ec)
 
 	ecard_set_drvdata(ec, NULL);
 
-	ide_unregister(hwif->index, 0, 0);
+	ide_unregister(hwif->index);
 
 	ecard_release_resources(ec);
 }
diff --git a/drivers/ide/ide-pnp.c b/drivers/ide/ide-pnp.c
index c14bb5380c25..34c2ad36ce54 100644
--- a/drivers/ide/ide-pnp.c
+++ b/drivers/ide/ide-pnp.c
@@ -62,7 +62,7 @@ static void idepnp_remove(struct pnp_dev * dev)
 	ide_hwif_t *hwif = pnp_get_drvdata(dev);
 
 	if (hwif)
-		ide_unregister(hwif->index, 0, 0);
+		ide_unregister(hwif->index);
 	else
 		printk(KERN_ERR "idepnp: Unable to remove device, please report.\n");
 }
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index 7e789c97a8b8..1121d9cb2a9b 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -362,107 +362,6 @@ void ide_hwif_release_regions(ide_hwif_t *hwif)
 			release_region(hwif->io_ports[i], 1);
 }
 
-/**
- *	ide_hwif_restore	-	restore hwif to template
- *	@hwif: hwif to update
- *	@tmp_hwif: template
- *
- *	Restore hwif to a previous state by copying most settings
- *	from the template.
- */
-
-static void ide_hwif_restore(ide_hwif_t *hwif, ide_hwif_t *tmp_hwif)
-{
-	hwif->hwgroup			= tmp_hwif->hwgroup;
-
-	hwif->gendev.parent		= tmp_hwif->gendev.parent;
-
-	hwif->proc			= tmp_hwif->proc;
-
-	hwif->major			= tmp_hwif->major;
-	hwif->straight8			= tmp_hwif->straight8;
-	hwif->bus_state			= tmp_hwif->bus_state;
-
-	hwif->host_flags		= tmp_hwif->host_flags;
-
-	hwif->pio_mask			= tmp_hwif->pio_mask;
-
-	hwif->ultra_mask		= tmp_hwif->ultra_mask;
-	hwif->mwdma_mask		= tmp_hwif->mwdma_mask;
-	hwif->swdma_mask		= tmp_hwif->swdma_mask;
-
-	hwif->cbl			= tmp_hwif->cbl;
-
-	hwif->chipset			= tmp_hwif->chipset;
-	hwif->hold			= tmp_hwif->hold;
-
-	hwif->dev			= tmp_hwif->dev;
-
-#ifdef CONFIG_BLK_DEV_IDEPCI
-	hwif->cds			= tmp_hwif->cds;
-#endif
-
-	hwif->set_pio_mode		= tmp_hwif->set_pio_mode;
-	hwif->set_dma_mode		= tmp_hwif->set_dma_mode;
-	hwif->mdma_filter		= tmp_hwif->mdma_filter;
-	hwif->udma_filter		= tmp_hwif->udma_filter;
-	hwif->selectproc		= tmp_hwif->selectproc;
-	hwif->reset_poll		= tmp_hwif->reset_poll;
-	hwif->pre_reset			= tmp_hwif->pre_reset;
-	hwif->resetproc			= tmp_hwif->resetproc;
-	hwif->maskproc			= tmp_hwif->maskproc;
-	hwif->quirkproc			= tmp_hwif->quirkproc;
-
-	hwif->ata_input_data		= tmp_hwif->ata_input_data;
-	hwif->ata_output_data		= tmp_hwif->ata_output_data;
-	hwif->atapi_input_bytes		= tmp_hwif->atapi_input_bytes;
-	hwif->atapi_output_bytes	= tmp_hwif->atapi_output_bytes;
-
-	hwif->dma_host_set		= tmp_hwif->dma_host_set;
-	hwif->dma_setup			= tmp_hwif->dma_setup;
-	hwif->dma_exec_cmd		= tmp_hwif->dma_exec_cmd;
-	hwif->dma_start			= tmp_hwif->dma_start;
-	hwif->ide_dma_end		= tmp_hwif->ide_dma_end;
-	hwif->ide_dma_test_irq		= tmp_hwif->ide_dma_test_irq;
-	hwif->ide_dma_clear_irq		= tmp_hwif->ide_dma_clear_irq;
-	hwif->dma_lost_irq		= tmp_hwif->dma_lost_irq;
-	hwif->dma_timeout		= tmp_hwif->dma_timeout;
-
-	hwif->OUTB			= tmp_hwif->OUTB;
-	hwif->OUTBSYNC			= tmp_hwif->OUTBSYNC;
-	hwif->OUTW			= tmp_hwif->OUTW;
-	hwif->OUTSW			= tmp_hwif->OUTSW;
-	hwif->OUTSL			= tmp_hwif->OUTSL;
-
-	hwif->INB			= tmp_hwif->INB;
-	hwif->INW			= tmp_hwif->INW;
-	hwif->INSW			= tmp_hwif->INSW;
-	hwif->INSL			= tmp_hwif->INSL;
-
-	hwif->sg_max_nents		= tmp_hwif->sg_max_nents;
-
-	hwif->mmio			= tmp_hwif->mmio;
-	hwif->rqsize			= tmp_hwif->rqsize;
-
-#ifndef CONFIG_BLK_DEV_IDECS
-	hwif->irq			= tmp_hwif->irq;
-#endif
-
-	hwif->dma_base			= tmp_hwif->dma_base;
-	hwif->dma_command		= tmp_hwif->dma_command;
-	hwif->dma_vendor1		= tmp_hwif->dma_vendor1;
-	hwif->dma_status		= tmp_hwif->dma_status;
-	hwif->dma_vendor3		= tmp_hwif->dma_vendor3;
-	hwif->dma_prdtable		= tmp_hwif->dma_prdtable;
-
-	hwif->config_data		= tmp_hwif->config_data;
-	hwif->select_data		= tmp_hwif->select_data;
-	hwif->extra_base		= tmp_hwif->extra_base;
-	hwif->extra_ports		= tmp_hwif->extra_ports;
-
-	hwif->hwif_data			= tmp_hwif->hwif_data;
-}
-
 void ide_remove_port_from_hwgroup(ide_hwif_t *hwif)
 {
 	ide_hwgroup_t *hwgroup = hwif->hwgroup;
@@ -530,8 +429,6 @@ EXPORT_SYMBOL_GPL(ide_port_unregister_devices);
 /**
  *	ide_unregister		-	free an IDE interface
  *	@index: index of interface (will change soon to a pointer)
- *	@init_default: init default hwif flag
- *	@restore: restore hwif flag
  *
  *	Perform the final unregister of an IDE interface. At the moment
  *	we don't refcount interfaces so this will also get split up.
@@ -551,10 +448,9 @@ EXPORT_SYMBOL_GPL(ide_port_unregister_devices);
  *	This is raving bonkers.
  */
 
-void ide_unregister(unsigned int index, int init_default, int restore)
+void ide_unregister(unsigned int index)
 {
 	ide_hwif_t *hwif, *g;
-	static ide_hwif_t tmp_hwif; /* protected by ide_cfg_mtx */
 	ide_hwgroup_t *hwgroup;
 	int irq_count = 0;
 
@@ -601,34 +497,14 @@ void ide_unregister(unsigned int index, int init_default, int restore)
 	unregister_blkdev(hwif->major, hwif->name);
 	spin_lock_irq(&ide_lock);
 
-	if (hwif->dma_base) {
-		(void) ide_release_dma(hwif);
-
-		hwif->dma_base = 0;
-		hwif->dma_command = 0;
-		hwif->dma_vendor1 = 0;
-		hwif->dma_status = 0;
-		hwif->dma_vendor3 = 0;
-		hwif->dma_prdtable = 0;
-
-		hwif->extra_base  = 0;
-		hwif->extra_ports = 0;
-	}
+	if (hwif->dma_base)
+		(void)ide_release_dma(hwif);
 
 	ide_hwif_release_regions(hwif);
 
-	/* copy original settings */
-	tmp_hwif = *hwif;
-
 	/* restore hwif data to pristine status */
 	ide_init_port_data(hwif, index);
 
-	if (init_default)
-		init_hwif_default(hwif, index);
-
-	if (restore)
-		ide_hwif_restore(hwif, &tmp_hwif);
-
 abort:
 	spin_unlock_irq(&ide_lock);
 	mutex_unlock(&ide_cfg_mtx);
@@ -647,52 +523,6 @@ void ide_init_port_hw(ide_hwif_t *hwif, hw_regs_t *hw)
 }
 EXPORT_SYMBOL_GPL(ide_init_port_hw);
 
-/**
- *	ide_register_hw		-	register IDE interface
- *	@hw: hardware registers
- *	@quirkproc: quirkproc function
- *	@hwifp: pointer to returned hwif
- *
- *	Register an IDE interface, specifying exactly the registers etc.
- *
- *	Returns -1 on error.
- */
-
-static int ide_register_hw(hw_regs_t *hw, void (*quirkproc)(ide_drive_t *),
-			   ide_hwif_t **hwifp)
-{
-	int index, retry = 1;
-	ide_hwif_t *hwif;
-	u8 idx[4] = { 0xff, 0xff, 0xff, 0xff };
-
-	do {
-		hwif = ide_find_port(hw->io_ports[IDE_DATA_OFFSET]);
-		if (hwif)
-			goto found;
-		for (index = 0; index < MAX_HWIFS; index++)
-			ide_unregister(index, 1, 1);
-	} while (retry--);
-	return -1;
-found:
-	index = hwif->index;
-	if (hwif->present)
-		ide_unregister(index, 0, 1);
-	else if (!hwif->hold)
-		ide_init_port_data(hwif, index);
-
-	ide_init_port_hw(hwif, hw);
-	hwif->quirkproc = quirkproc;
-
-	idx[0] = index;
-
-	ide_device_add(idx, NULL);
-
-	if (hwifp)
-		*hwifp = hwif;
-
-	return hwif->present ? index : -1;
-}
-
 /*
  *	Locks for IDE setting functionality
  */
@@ -995,27 +825,6 @@ int generic_ide_ioctl(ide_drive_t *drive, struct file *file, struct block_device
 			if (!capable(CAP_SYS_RAWIO))
 				return -EACCES;
 			return ide_task_ioctl(drive, cmd, arg);
-
-		case HDIO_SCAN_HWIF:
-		{
-			hw_regs_t hw;
-			int args[3];
-			if (!capable(CAP_SYS_RAWIO)) return -EACCES;
-			if (copy_from_user(args, p, 3 * sizeof(int)))
-				return -EFAULT;
-			memset(&hw, 0, sizeof(hw));
-			ide_init_hwif_ports(&hw, (unsigned long) args[0],
-					    (unsigned long) args[1], NULL);
-			hw.irq = args[2];
-			if (ide_register_hw(&hw, NULL, NULL) == -1)
-				return -EIO;
-			return 0;
-		}
-	        case HDIO_UNREGISTER_HWIF:
-			if (!capable(CAP_SYS_RAWIO)) return -EACCES;
-			/* (arg > MAX_HWIFS) checked in function */
-			ide_unregister(arg, 1, 1);
-			return 0;
 		case HDIO_SET_NICE:
 			if (!capable(CAP_SYS_ADMIN)) return -EACCES;
 			if (arg != (arg & ((1 << IDE_NICE_DSC_OVERLAP) | (1 << IDE_NICE_1))))
@@ -1648,7 +1457,7 @@ void __exit cleanup_module (void)
 	int index;
 
 	for (index = 0; index < MAX_HWIFS; ++index)
-		ide_unregister(index, 0, 0);
+		ide_unregister(index);
 
 	proc_ide_destroy();
 
diff --git a/drivers/ide/legacy/ide-cs.c b/drivers/ide/legacy/ide-cs.c
index de2e5944809e..2b0b4958881a 100644
--- a/drivers/ide/legacy/ide-cs.c
+++ b/drivers/ide/legacy/ide-cs.c
@@ -163,7 +163,7 @@ static int idecs_register(unsigned long io, unsigned long ctl, unsigned long irq
     i = hwif->index;
 
     if (hwif->present)
-	ide_unregister(i, 0, 0);
+	ide_unregister(i);
     else if (!hwif->hold)
 	ide_init_port_data(hwif, i);
 
@@ -360,7 +360,7 @@ void ide_release(struct pcmcia_device *link)
     if (info->ndev) {
 	/* FIXME: if this fails we need to queue the cleanup somehow
 	   -- need to investigate the required PCMCIA magic */
-	ide_unregister(info->hd, 0, 0);
+	ide_unregister(info->hd);
     }
     info->ndev = 0;
 
diff --git a/drivers/ide/legacy/ide_platform.c b/drivers/ide/legacy/ide_platform.c
index 688fcae17488..249651e2da42 100644
--- a/drivers/ide/legacy/ide_platform.c
+++ b/drivers/ide/legacy/ide_platform.c
@@ -122,7 +122,7 @@ static int __devexit plat_ide_remove(struct platform_device *pdev)
 {
 	ide_hwif_t *hwif = pdev->dev.driver_data;
 
-	ide_unregister(hwif->index, 0, 0);
+	ide_unregister(hwif->index);
 
 	return 0;
 }
diff --git a/drivers/ide/mips/au1xxx-ide.c b/drivers/ide/mips/au1xxx-ide.c
index 85c016bdfd38..ee76023f3737 100644
--- a/drivers/ide/mips/au1xxx-ide.c
+++ b/drivers/ide/mips/au1xxx-ide.c
@@ -673,7 +673,7 @@ static int au_ide_remove(struct device *dev)
 	ide_hwif_t *hwif = dev_get_drvdata(dev);
 	_auide_hwif *ahwif = &auide_hwif;
 
-	ide_unregister(hwif->index, 0, 0);
+	ide_unregister(hwif->index);
 
 	iounmap((void *)ahwif->regbase);
 
diff --git a/drivers/ide/pci/delkin_cb.c b/drivers/ide/pci/delkin_cb.c
index e08e13a0bb6e..89570df52f0a 100644
--- a/drivers/ide/pci/delkin_cb.c
+++ b/drivers/ide/pci/delkin_cb.c
@@ -85,7 +85,7 @@ delkin_cb_probe (struct pci_dev *dev, const struct pci_device_id *id)
 	i = hwif->index;
 
 	if (hwif->present)
-		ide_unregister(i, 0, 0);
+		ide_unregister(i);
 	else if (!hwif->hold)
 		ide_init_port_data(hwif, i);
 
@@ -120,7 +120,7 @@ delkin_cb_remove (struct pci_dev *dev)
 	ide_hwif_t *hwif = pci_get_drvdata(dev);
 
 	if (hwif)
-		ide_unregister(hwif->index, 0, 0);
+		ide_unregister(hwif->index);
 
 	pci_disable_device(dev);
 }
diff --git a/drivers/ide/pci/scc_pata.c b/drivers/ide/pci/scc_pata.c
index 238e3e181e87..085c1b58a99c 100644
--- a/drivers/ide/pci/scc_pata.c
+++ b/drivers/ide/pci/scc_pata.c
@@ -736,7 +736,7 @@ static void __devexit scc_remove(struct pci_dev *dev)
 		hwif->dmatable_cpu = NULL;
 	}
 
-	ide_unregister(hwif->index, 0, 0);
+	ide_unregister(hwif->index);
 
 	hwif->chipset = ide_unknown;
 	iounmap((void*)ports->dma);
diff --git a/include/linux/hdreg.h b/include/linux/hdreg.h
index e38e75967e74..c37e9241fae7 100644
--- a/include/linux/hdreg.h
+++ b/include/linux/hdreg.h
@@ -422,9 +422,11 @@ struct hd_geometry {
 #define HDIO_SET_NOWERR		0x0325	/* change ignore-write-error flag */
 #define HDIO_SET_DMA		0x0326	/* change use-dma flag */
 #define HDIO_SET_PIO_MODE	0x0327	/* reconfig interface to new speed */
+#ifndef __KERNEL__
 #define HDIO_SCAN_HWIF		0x0328	/* register and (re)scan interface */
-#define HDIO_SET_NICE		0x0329	/* set nice flags */
 #define HDIO_UNREGISTER_HWIF	0x032a  /* unregister interface */
+#endif
+#define HDIO_SET_NICE		0x0329	/* set nice flags */
 #define HDIO_SET_WCACHE		0x032b	/* change write cache enable-disable */
 #define HDIO_SET_ACOUSTIC	0x032c	/* change acoustic behavior */
 #define HDIO_SET_BUSSTATE	0x032d	/* set the bus state of the hwif */
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 67f83c60845f..478ddf7e21d5 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -1191,7 +1191,7 @@ static inline void ide_acpi_set_state(ide_hwif_t *hwif, int on) {}
 void ide_remove_port_from_hwgroup(ide_hwif_t *);
 extern int ide_hwif_request_regions(ide_hwif_t *hwif);
 extern void ide_hwif_release_regions(ide_hwif_t* hwif);
-void ide_unregister(unsigned int, int, int);
+void ide_unregister(unsigned int);
 
 void ide_register_region(struct gendisk *);
 void ide_unregister_region(struct gendisk *);
-- 
cgit v1.2.3


From 2304dc6481f9d4fb4f0cb5b72497dfe1694cef9c Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:24 +0200
Subject: ide: remove ->hold field from ide_hwif_t (take 2)

->hold is write-only now, remove it.

v2:
* v1 missed bast-ide, palm_bk3710, ide-cs and delkin_cb host drivers.

Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/arm/bast-ide.c    | 2 +-
 drivers/ide/arm/palm_bk3710.c | 2 +-
 drivers/ide/legacy/ide-cs.c   | 2 +-
 drivers/ide/mips/au1xxx-ide.c | 3 ---
 drivers/ide/pci/delkin_cb.c   | 2 +-
 drivers/ide/ppc/pmac.c        | 1 -
 include/linux/ide.h           | 1 -
 7 files changed, 4 insertions(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/arm/bast-ide.c b/drivers/ide/arm/bast-ide.c
index d2196436788b..ec46c44b061c 100644
--- a/drivers/ide/arm/bast-ide.c
+++ b/drivers/ide/arm/bast-ide.c
@@ -49,7 +49,7 @@ static int __init bastide_register(unsigned int base, unsigned int aux, int irq)
 
 	if (hwif->present)
 		ide_unregister(i);
-	else if (!hwif->hold)
+	else
 		ide_init_port_data(hwif, i);
 
 	ide_init_port_hw(hwif, &hw);
diff --git a/drivers/ide/arm/palm_bk3710.c b/drivers/ide/arm/palm_bk3710.c
index 73899ef4ab0d..474162cdf665 100644
--- a/drivers/ide/arm/palm_bk3710.c
+++ b/drivers/ide/arm/palm_bk3710.c
@@ -386,7 +386,7 @@ static int __devinit palm_bk3710_probe(struct platform_device *pdev)
 
 	if (hwif->present)
 		ide_unregister(i);
-	else if (!hwif->hold)
+	else
 		ide_init_port_data(hwif, i);
 
 	ide_init_port_hw(hwif, &hw);
diff --git a/drivers/ide/legacy/ide-cs.c b/drivers/ide/legacy/ide-cs.c
index 2b0b4958881a..9a23b94f2939 100644
--- a/drivers/ide/legacy/ide-cs.c
+++ b/drivers/ide/legacy/ide-cs.c
@@ -164,7 +164,7 @@ static int idecs_register(unsigned long io, unsigned long ctl, unsigned long irq
 
     if (hwif->present)
 	ide_unregister(i);
-    else if (!hwif->hold)
+    else
 	ide_init_port_data(hwif, i);
 
     ide_init_port_hw(hwif, &hw);
diff --git a/drivers/ide/mips/au1xxx-ide.c b/drivers/ide/mips/au1xxx-ide.c
index ee76023f3737..9b628248f2f4 100644
--- a/drivers/ide/mips/au1xxx-ide.c
+++ b/drivers/ide/mips/au1xxx-ide.c
@@ -613,9 +613,6 @@ static int au_ide_probe(struct device *dev)
 
 	hwif->dev = dev;
 
-	/* hold should be on in all cases */
-	hwif->hold                      = 1;
-
 	hwif->mmio  = 1;
 
 	/* If the user has selected DDMA assisted copies,
diff --git a/drivers/ide/pci/delkin_cb.c b/drivers/ide/pci/delkin_cb.c
index 89570df52f0a..961698d655eb 100644
--- a/drivers/ide/pci/delkin_cb.c
+++ b/drivers/ide/pci/delkin_cb.c
@@ -86,7 +86,7 @@ delkin_cb_probe (struct pci_dev *dev, const struct pci_device_id *id)
 
 	if (hwif->present)
 		ide_unregister(i);
-	else if (!hwif->hold)
+	else
 		ide_init_port_data(hwif, i);
 
 	ide_init_port_hw(hwif, &hw);
diff --git a/drivers/ide/ppc/pmac.c b/drivers/ide/ppc/pmac.c
index 7889f5f5c49a..b37dcfbdaad6 100644
--- a/drivers/ide/ppc/pmac.c
+++ b/drivers/ide/ppc/pmac.c
@@ -1120,7 +1120,6 @@ pmac_ide_setup_device(pmac_ide_hwif_t *pmif, ide_hwif_t *hwif, hw_regs_t *hw)
 	hwif->hwif_data = pmif;
 	ide_init_port_hw(hwif, hw);
 	hwif->noprobe = pmif->mediabay;
-	hwif->hold = pmif->mediabay;
 	hwif->cbl = pmif->cable_80 ? ATA_CBL_PATA80 : ATA_CBL_PATA40;
 	hwif->set_pio_mode = pmac_ide_set_pio_mode;
 	if (pmif->kind == controller_un_ata6
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 478ddf7e21d5..7744ac9d1ff9 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -571,7 +571,6 @@ typedef struct hwif_s {
 
 	unsigned	noprobe    : 1;	/* don't probe for this interface */
 	unsigned	present    : 1;	/* this interface exists */
-	unsigned	hold       : 1; /* this interface is always present */
 	unsigned	serialized : 1;	/* serialized all channel operation */
 	unsigned	sharing_irq: 1;	/* 1 = sharing irq with another hwif */
 	unsigned	reset      : 1;	/* reset after probe */
-- 
cgit v1.2.3


From e6bfa38a4803646e212ef542b957344e790c3733 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:25 +0200
Subject: ide: remove ide_init_hwif_ports()

ide_init_hwif_ports() is only used by init_ide_data() now, inline it there.

Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/ide.c   | 11 +++++++++--
 include/linux/ide.h | 32 --------------------------------
 2 files changed, 9 insertions(+), 34 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index a60b956db519..9e4d503bc716 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -197,13 +197,20 @@ static void __init init_ide_data (void)
 	/* Initialise all interface structures */
 	for (index = 0; index < MAX_HWIFS; ++index) {
 		ide_hwif_t *hwif = &ide_hwifs[index];
+		unsigned long io_addr = ide_default_io_base(index);
+		unsigned long ctl_addr = ide_default_io_ctl(io_addr);
 
 		ide_init_port_data(hwif, index);
 
+#ifdef CONFIG_IDE_ARCH_OBSOLETE_INIT
 		memset(&hw, 0, sizeof(hw));
-		ide_init_hwif_ports(&hw, ide_default_io_base(index), 0,
-				    &hwif->irq);
+		ide_std_init_ports(&hw, io_addr, ctl_addr);
+# ifdef CONFIG_PPC32
+		if (ppc_ide_md.ide_init_hwif)
+			ppc_ide_md.ide_init_hwif(&hw, io_addr, 0, &hwif->irq);
+# endif
 		memcpy(hwif->io_ports, hw.io_ports, sizeof(hw.io_ports));
+#endif
 		hwif->noprobe = !hwif->io_ports[IDE_DATA_OFFSET];
 #if !defined(CONFIG_PPC32) || !defined(CONFIG_PCI)
 		hwif->irq =
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 7744ac9d1ff9..f4c7db572a16 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -215,38 +215,6 @@ static inline void ide_std_init_ports(hw_regs_t *hw,
 # define ide_init_default_irq(base)	(0)
 #endif
 
-#ifdef CONFIG_IDE_ARCH_OBSOLETE_INIT
-static inline void ide_init_hwif_ports(hw_regs_t *hw,
-				       unsigned long io_addr,
-				       unsigned long ctl_addr,
-				       int *irq)
-{
-	if (!ctl_addr)
-		ide_std_init_ports(hw, io_addr, ide_default_io_ctl(io_addr));
-	else
-		ide_std_init_ports(hw, io_addr, ctl_addr);
-
-	if (irq)
-		*irq = 0;
-
-	hw->io_ports[IDE_IRQ_OFFSET] = 0;
-
-#ifdef CONFIG_PPC32
-	if (ppc_ide_md.ide_init_hwif)
-		ppc_ide_md.ide_init_hwif(hw, io_addr, ctl_addr, irq);
-#endif
-}
-#else
-static inline void ide_init_hwif_ports(hw_regs_t *hw,
-				       unsigned long io_addr,
-				       unsigned long ctl_addr,
-				       int *irq)
-{
-	if (io_addr || ctl_addr)
-		printk(KERN_WARNING "%s: must not be called\n", __FUNCTION__);
-}
-#endif /* CONFIG_IDE_ARCH_OBSOLETE_INIT */
-
 /* Currently only m68k, apus and m8xx need it */
 #ifndef IDE_ARCH_ACK_INTR
 # define ide_ack_intr(hwif) (1)
-- 
cgit v1.2.3


From 7616c0ad2087c7d244b8985390c63059a6223c45 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:26 +0200
Subject: ide: add ide_atapi_{discard_data,write_zeros} inline helpers

Add ide_atapi_{discard_data,write_zeros} inline helpers to <linux/ide.h>
and use them instead of home-brewn helpers in ide-{floppy,tape,scsi}.

There should be no functional changes caused by this patch.

Cc: Borislav Petkov <petkovbb@gmail.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/ide-floppy.c | 24 +++---------------------
 drivers/ide/ide-tape.c   | 14 ++------------
 drivers/scsi/ide-scsi.c  | 18 +++---------------
 include/linux/ide.h      | 22 ++++++++++++++++++++++
 4 files changed, 30 insertions(+), 48 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/ide-floppy.c b/drivers/ide/ide-floppy.c
index 973f5f6c815e..170c60d93f55 100644
--- a/drivers/ide/ide-floppy.c
+++ b/drivers/ide/ide-floppy.c
@@ -229,23 +229,6 @@ static void ide_floppy_put(struct ide_floppy_obj *floppy)
 	mutex_unlock(&idefloppy_ref_mutex);
 }
 
-/*
- * Too bad. The drive wants to send us data which we are not ready to accept.
- * Just throw it away.
- */
-static void idefloppy_discard_data(ide_drive_t *drive, unsigned int bcount)
-{
-	while (bcount--)
-		(void) HWIF(drive)->INB(IDE_DATA_REG);
-}
-
-static void idefloppy_write_zeros(ide_drive_t *drive, unsigned int bcount)
-{
-	while (bcount--)
-		HWIF(drive)->OUTB(0, IDE_DATA_REG);
-}
-
-
 /*
  * Used to finish servicing a request. For read/write requests, we will call
  * ide_end_request to pass to the next buffer.
@@ -313,10 +296,9 @@ static void ide_floppy_io_buffers(ide_drive_t *drive, idefloppy_pc_t *pc,
 		printk(KERN_ERR "%s: leftover data in %s, bcount == %d\n",
 				drive->name, __func__, bcount);
 		if (direction)
-			idefloppy_write_zeros(drive, bcount);
+			ide_atapi_write_zeros(drive, bcount);
 		else
-			idefloppy_discard_data(drive, bcount);
-
+			ide_atapi_discard_data(drive, bcount);
 	}
 }
 
@@ -541,7 +523,7 @@ static ide_startstop_t idefloppy_pc_intr (ide_drive_t *drive)
 				printk(KERN_ERR "ide-floppy: The floppy wants "
 					"to send us more data than expected "
 					"- discarding data\n");
-				idefloppy_discard_data(drive, bcount);
+				ide_atapi_discard_data(drive, bcount);
 
 				ide_set_handler(drive,
 						&idefloppy_pc_intr,
diff --git a/drivers/ide/ide-tape.c b/drivers/ide/ide-tape.c
index 0598ecfd5f37..bfdc4f449797 100644
--- a/drivers/ide/ide-tape.c
+++ b/drivers/ide/ide-tape.c
@@ -518,16 +518,6 @@ static struct ide_tape_obj *ide_tape_chrdev_get(unsigned int i)
 	return tape;
 }
 
-/*
- * Too bad. The drive wants to send us data which we are not ready to accept.
- * Just throw it away.
- */
-static void idetape_discard_data(ide_drive_t *drive, unsigned int bcount)
-{
-	while (bcount--)
-		(void) HWIF(drive)->INB(IDE_DATA_REG);
-}
-
 static void idetape_input_buffers(ide_drive_t *drive, idetape_pc_t *pc,
 				  unsigned int bcount)
 {
@@ -538,7 +528,7 @@ static void idetape_input_buffers(ide_drive_t *drive, idetape_pc_t *pc,
 		if (bh == NULL) {
 			printk(KERN_ERR "ide-tape: bh == NULL in "
 				"idetape_input_buffers\n");
-			idetape_discard_data(drive, bcount);
+			ide_atapi_discard_data(drive, bcount);
 			return;
 		}
 		count = min(
@@ -1152,7 +1142,7 @@ static ide_startstop_t idetape_pc_intr(ide_drive_t *drive)
 				printk(KERN_ERR "ide-tape: The tape wants to "
 					"send us more data than expected "
 					"- discarding data\n");
-				idetape_discard_data(drive, bcount);
+				ide_atapi_discard_data(drive, bcount);
 				ide_set_handler(drive, &idetape_pc_intr,
 						IDETAPE_WAIT_CMD, NULL);
 				return ide_started;
diff --git a/drivers/scsi/ide-scsi.c b/drivers/scsi/ide-scsi.c
index 68e5c632c5d5..3c3b3502c4d4 100644
--- a/drivers/scsi/ide-scsi.c
+++ b/drivers/scsi/ide-scsi.c
@@ -152,18 +152,6 @@ static inline idescsi_scsi_t *drive_to_idescsi(ide_drive_t *ide_drive)
  */
 #define IDESCSI_PC_RQ			90
 
-static void idescsi_discard_data (ide_drive_t *drive, unsigned int bcount)
-{
-	while (bcount--)
-		(void) HWIF(drive)->INB(IDE_DATA_REG);
-}
-
-static void idescsi_output_zeros (ide_drive_t *drive, unsigned int bcount)
-{
-	while (bcount--)
-		HWIF(drive)->OUTB(0, IDE_DATA_REG);
-}
-
 /*
  *	PIO data transfer routines using the scatter gather table.
  */
@@ -200,7 +188,7 @@ static void idescsi_input_buffers (ide_drive_t *drive, idescsi_pc_t *pc, unsigne
 
 	if (bcount) {
 		printk (KERN_ERR "ide-scsi: scatter gather table too small, discarding data\n");
-		idescsi_discard_data (drive, bcount);
+		ide_atapi_discard_data(drive, bcount);
 	}
 }
 
@@ -237,7 +225,7 @@ static void idescsi_output_buffers (ide_drive_t *drive, idescsi_pc_t *pc, unsign
 
 	if (bcount) {
 		printk (KERN_ERR "ide-scsi: scatter gather table too small, padding with zeros\n");
-		idescsi_output_zeros (drive, bcount);
+		ide_atapi_write_zeros(drive, bcount);
 	}
 }
 
@@ -463,7 +451,7 @@ static ide_startstop_t idescsi_pc_intr (ide_drive_t *drive)
 				}
 				pc->actually_transferred += temp;
 				pc->current_position += temp;
-				idescsi_discard_data(drive, bcount - temp);
+				ide_atapi_discard_data(drive, bcount - temp);
 				ide_set_handler(drive, &idescsi_pc_intr, get_timeout(pc), idescsi_expiry);
 				return ide_started;
 			}
diff --git a/include/linux/ide.h b/include/linux/ide.h
index f4c7db572a16..2eb99cab4a3d 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -1297,4 +1297,26 @@ static inline u8 ide_read_error(ide_drive_t *drive)
 	return hwif->INB(hwif->io_ports[IDE_ERROR_OFFSET]);
 }
 
+/*
+ * Too bad. The drive wants to send us data which we are not ready to accept.
+ * Just throw it away.
+ */
+static inline void ide_atapi_discard_data(ide_drive_t *drive, unsigned bcount)
+{
+	ide_hwif_t *hwif = drive->hwif;
+
+	/* FIXME: use ->atapi_input_bytes */
+	while (bcount--)
+		(void)hwif->INB(hwif->io_ports[IDE_DATA_OFFSET]);
+}
+
+static inline void ide_atapi_write_zeros(ide_drive_t *drive, unsigned bcount)
+{
+	ide_hwif_t *hwif = drive->hwif;
+
+	/* FIXME: use ->atapi_output_bytes */
+	while (bcount--)
+		hwif->OUTB(0, hwif->io_ports[IDE_DATA_OFFSET]);
+}
+
 #endif /* _IDE_H */
-- 
cgit v1.2.3


From 23579a2a170265aacf78069f4817a41c1d6e9323 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:26 +0200
Subject: ide: remove IDE_*_REG macros

* Add IDE_{ALTSTATUS,IREASON,BCOUNTL,BCOUNTH}_OFFSET defines.

* Remove IDE_*_REG macros - this results in more readable
  and slightly smaller code.

There should be no functional changes caused by this patch.

Cc: Borislav Petkov <petkovbb@gmail.com>
Acked-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/cris/ide-cris.c  | 18 +++++++---
 drivers/ide/ide-cd.c         | 10 +++---
 drivers/ide/ide-floppy.c     | 14 ++++----
 drivers/ide/ide-io.c         | 38 ++++++++++++--------
 drivers/ide/ide-iops.c       | 86 +++++++++++++++++++++++++-------------------
 drivers/ide/ide-probe.c      | 19 +++++-----
 drivers/ide/ide-tape.c       | 12 +++----
 drivers/ide/ide-taskfile.c   | 29 ++++++++-------
 drivers/ide/legacy/ht6560b.c | 13 +++----
 drivers/ide/pci/hpt366.c     |  2 +-
 drivers/ide/pci/scc_pata.c   |  6 ++--
 drivers/ide/pci/sgiioc4.c    | 19 +++++-----
 drivers/ide/ppc/pmac.c       |  3 +-
 drivers/scsi/ide-scsi.c      | 15 ++++----
 include/linux/ide.h          | 29 +++++----------
 15 files changed, 175 insertions(+), 138 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/cris/ide-cris.c b/drivers/ide/cris/ide-cris.c
index c8ffbaf29a88..31266d278095 100644
--- a/drivers/ide/cris/ide-cris.c
+++ b/drivers/ide/cris/ide-cris.c
@@ -228,7 +228,10 @@ cris_ide_fill_descriptor(cris_dma_descr_type *d, void* buf, unsigned int len, in
 static void
 cris_ide_start_dma(ide_drive_t *drive, cris_dma_descr_type *d, int dir,int type,int len)
 {
-	reg_ata_rw_ctrl2 ctrl2 = REG_TYPE_CONV(reg_ata_rw_ctrl2, int, IDE_DATA_REG);
+	ide_hwif_t *hwif = drive->hwif;
+
+	reg_ata_rw_ctrl2 ctrl2 = REG_TYPE_CONV(reg_ata_rw_ctrl2, int,
+					       hwif->io_ports[IDE_DATA_OFFSET]);
 	reg_ata_rw_trf_cnt trf_cnt = {0};
 
 	mycontext.saved_data = (dma_descr_data*)virt_to_phys(d);
@@ -264,8 +267,12 @@ cris_ide_wait_dma(int dir)
 
 static int cris_dma_test_irq(ide_drive_t *drive)
 {
+	ide_hwif_t *hwif = drive->hwif;
 	int intr = REG_RD_INT(ata, regi_ata, r_intr);
-	reg_ata_rw_ctrl2 ctrl2 = REG_TYPE_CONV(reg_ata_rw_ctrl2, int, IDE_DATA_REG);
+
+	reg_ata_rw_ctrl2 ctrl2 = REG_TYPE_CONV(reg_ata_rw_ctrl2, int,
+					       hwif->io_ports[IDE_DATA_OFFSET]);
+
 	return intr & (1 << ctrl2.sel) ? 1 : 0;
 }
 
@@ -523,7 +530,8 @@ static void cris_ide_start_dma(ide_drive_t *drive, cris_dma_descr_type *d, int d
 	                          IO_STATE(R_ATA_CTRL_DATA, handsh, dma);
 	*R_ATA_CTRL_DATA =
 		cmd |
-		IO_FIELD(R_ATA_CTRL_DATA, data, IDE_DATA_REG) |
+		IO_FIELD(R_ATA_CTRL_DATA, data,
+			 drive->hwif->io_ports[IDE_DATA_OFFSET]) |
 		IO_STATE(R_ATA_CTRL_DATA, src_dst,  dma)  |
 		IO_STATE(R_ATA_CTRL_DATA, multi,    on)   |
 		IO_STATE(R_ATA_CTRL_DATA, dma_size, word);
@@ -541,7 +549,9 @@ cris_ide_wait_dma(int dir)
 static int cris_dma_test_irq(ide_drive_t *drive)
 {
 	int intr = *R_IRQ_MASK0_RD;
-	int bus = IO_EXTRACT(R_ATA_CTRL_DATA, sel, IDE_DATA_REG);
+	int bus = IO_EXTRACT(R_ATA_CTRL_DATA, sel,
+			     drive->hwif->io_ports[IDE_DATA_OFFSET]);
+
 	return intr & (1 << (bus + IO_BITNR(R_IRQ_MASK0_RD, ata_irq0))) ? 1 : 0;
 }
 
diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c
index c8d0e8715997..396000208f81 100644
--- a/drivers/ide/ide-cd.c
+++ b/drivers/ide/ide-cd.c
@@ -542,7 +542,8 @@ static ide_startstop_t cdrom_start_packet_command(ide_drive_t *drive,
 
 		/* packet command */
 		spin_lock_irqsave(&ide_lock, flags);
-		hwif->OUTBSYNC(drive, WIN_PACKETCMD, IDE_COMMAND_REG);
+		hwif->OUTBSYNC(drive, WIN_PACKETCMD,
+			       hwif->io_ports[IDE_COMMAND_OFFSET]);
 		ndelay(400);
 		spin_unlock_irqrestore(&ide_lock, flags);
 
@@ -992,6 +993,7 @@ static int cdrom_newpc_intr_dummy_cb(struct request *rq)
 
 static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive)
 {
+	ide_hwif_t *hwif = drive->hwif;
 	struct cdrom_info *info = drive->driver_data;
 	struct request *rq = HWGROUP(drive)->rq;
 	xfer_func_t *xferfunc;
@@ -1032,9 +1034,9 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive)
 	/*
 	 * ok we fall to pio :/
 	 */
-	ireason = HWIF(drive)->INB(IDE_IREASON_REG) & 0x3;
-	lowcyl  = HWIF(drive)->INB(IDE_BCOUNTL_REG);
-	highcyl = HWIF(drive)->INB(IDE_BCOUNTH_REG);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]) & 0x3;
+	lowcyl  = hwif->INB(hwif->io_ports[IDE_BCOUNTL_OFFSET]);
+	highcyl = hwif->INB(hwif->io_ports[IDE_BCOUNTH_OFFSET]);
 
 	len = lowcyl + (256 * highcyl);
 
diff --git a/drivers/ide/ide-floppy.c b/drivers/ide/ide-floppy.c
index 170c60d93f55..4ce67bdb5d5e 100644
--- a/drivers/ide/ide-floppy.c
+++ b/drivers/ide/ide-floppy.c
@@ -498,10 +498,10 @@ static ide_startstop_t idefloppy_pc_intr (ide_drive_t *drive)
 	}
 
 	/* Get the number of bytes to transfer */
-	bcount = (hwif->INB(IDE_BCOUNTH_REG) << 8) |
-		  hwif->INB(IDE_BCOUNTL_REG);
+	bcount = (hwif->INB(hwif->io_ports[IDE_BCOUNTH_OFFSET]) << 8) |
+		  hwif->INB(hwif->io_ports[IDE_BCOUNTL_OFFSET]);
 	/* on this interrupt */
-	ireason = hwif->INB(IDE_IREASON_REG);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 
 	if (ireason & CD) {
 		printk(KERN_ERR "ide-floppy: CoD != 0 in %s\n", __func__);
@@ -562,6 +562,7 @@ static ide_startstop_t idefloppy_pc_intr (ide_drive_t *drive)
  */
 static ide_startstop_t idefloppy_transfer_pc(ide_drive_t *drive)
 {
+	ide_hwif_t *hwif = drive->hwif;
 	ide_startstop_t startstop;
 	idefloppy_floppy_t *floppy = drive->driver_data;
 	u8 ireason;
@@ -571,7 +572,7 @@ static ide_startstop_t idefloppy_transfer_pc(ide_drive_t *drive)
 				"initiated yet DRQ isn't asserted\n");
 		return startstop;
 	}
-	ireason = drive->hwif->INB(IDE_IREASON_REG);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 	if ((ireason & CD) == 0 || (ireason & IO)) {
 		printk(KERN_ERR "ide-floppy: (IO,CoD) != (0,1) while "
 				"issuing a packet command\n");
@@ -608,6 +609,7 @@ static int idefloppy_transfer_pc2(ide_drive_t *drive)
 
 static ide_startstop_t idefloppy_transfer_pc1(ide_drive_t *drive)
 {
+	ide_hwif_t *hwif = drive->hwif;
 	idefloppy_floppy_t *floppy = drive->driver_data;
 	ide_startstop_t startstop;
 	u8 ireason;
@@ -617,7 +619,7 @@ static ide_startstop_t idefloppy_transfer_pc1(ide_drive_t *drive)
 				"initiated yet DRQ isn't asserted\n");
 		return startstop;
 	}
-	ireason = drive->hwif->INB(IDE_IREASON_REG);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 	if ((ireason & CD) == 0 || (ireason & IO)) {
 		printk(KERN_ERR "ide-floppy: (IO,CoD) != (0,1) "
 				"while issuing a packet command\n");
@@ -723,7 +725,7 @@ static ide_startstop_t idefloppy_issue_pc(ide_drive_t *drive,
 		return ide_started;
 	} else {
 		/* Issue the packet command */
-		HWIF(drive)->OUTB(WIN_PACKETCMD, IDE_COMMAND_REG);
+		hwif->OUTB(WIN_PACKETCMD, hwif->io_ports[IDE_COMMAND_OFFSET]);
 		return (*pkt_xfer_routine) (drive);
 	}
 }
diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c
index 715379605a7b..31e5afadb7e9 100644
--- a/drivers/ide/ide-io.c
+++ b/drivers/ide/ide-io.c
@@ -301,39 +301,45 @@ void ide_tf_read(ide_drive_t *drive, ide_task_t *task)
 	struct ide_taskfile *tf = &task->tf;
 
 	if (task->tf_flags & IDE_TFLAG_IN_DATA) {
-		u16 data = hwif->INW(IDE_DATA_REG);
+		u16 data = hwif->INW(hwif->io_ports[IDE_DATA_OFFSET]);
 
 		tf->data = data & 0xff;
 		tf->hob_data = (data >> 8) & 0xff;
 	}
 
 	/* be sure we're looking at the low order bits */
-	hwif->OUTB(drive->ctl & ~0x80, IDE_CONTROL_REG);
+	hwif->OUTB(drive->ctl & ~0x80, hwif->io_ports[IDE_CONTROL_OFFSET]);
 
 	if (task->tf_flags & IDE_TFLAG_IN_NSECT)
-		tf->nsect  = hwif->INB(IDE_NSECTOR_REG);
+		tf->nsect  = hwif->INB(hwif->io_ports[IDE_NSECTOR_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_IN_LBAL)
-		tf->lbal   = hwif->INB(IDE_SECTOR_REG);
+		tf->lbal   = hwif->INB(hwif->io_ports[IDE_SECTOR_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_IN_LBAM)
-		tf->lbam   = hwif->INB(IDE_LCYL_REG);
+		tf->lbam   = hwif->INB(hwif->io_ports[IDE_LCYL_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_IN_LBAH)
-		tf->lbah   = hwif->INB(IDE_HCYL_REG);
+		tf->lbah   = hwif->INB(hwif->io_ports[IDE_HCYL_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_IN_DEVICE)
-		tf->device = hwif->INB(IDE_SELECT_REG);
+		tf->device = hwif->INB(hwif->io_ports[IDE_SELECT_OFFSET]);
 
 	if (task->tf_flags & IDE_TFLAG_LBA48) {
-		hwif->OUTB(drive->ctl | 0x80, IDE_CONTROL_REG);
+		hwif->OUTB(drive->ctl | 0x80,
+			   hwif->io_ports[IDE_CONTROL_OFFSET]);
 
 		if (task->tf_flags & IDE_TFLAG_IN_HOB_FEATURE)
-			tf->hob_feature = hwif->INB(IDE_FEATURE_REG);
+			tf->hob_feature =
+				hwif->INB(hwif->io_ports[IDE_FEATURE_OFFSET]);
 		if (task->tf_flags & IDE_TFLAG_IN_HOB_NSECT)
-			tf->hob_nsect   = hwif->INB(IDE_NSECTOR_REG);
+			tf->hob_nsect   =
+				hwif->INB(hwif->io_ports[IDE_NSECTOR_OFFSET]);
 		if (task->tf_flags & IDE_TFLAG_IN_HOB_LBAL)
-			tf->hob_lbal    = hwif->INB(IDE_SECTOR_REG);
+			tf->hob_lbal    =
+				hwif->INB(hwif->io_ports[IDE_SECTOR_OFFSET]);
 		if (task->tf_flags & IDE_TFLAG_IN_HOB_LBAM)
-			tf->hob_lbam    = hwif->INB(IDE_LCYL_REG);
+			tf->hob_lbam    =
+				hwif->INB(hwif->io_ports[IDE_LCYL_OFFSET]);
 		if (task->tf_flags & IDE_TFLAG_IN_HOB_LBAH)
-			tf->hob_lbah    = hwif->INB(IDE_HCYL_REG);
+			tf->hob_lbah    =
+				hwif->INB(hwif->io_ports[IDE_HCYL_OFFSET]);
 	}
 }
 
@@ -448,7 +454,8 @@ static ide_startstop_t ide_ata_error(ide_drive_t *drive, struct request *rq, u8
 		if (err == ABRT_ERR) {
 			if (drive->select.b.lba &&
 			    /* some newer drives don't support WIN_SPECIFY */
-			    hwif->INB(IDE_COMMAND_REG) == WIN_SPECIFY)
+			    hwif->INB(hwif->io_ports[IDE_COMMAND_OFFSET]) ==
+				WIN_SPECIFY)
 				return ide_stopped;
 		} else if ((err & BAD_CRC) == BAD_CRC) {
 			/* UDMA crc error, just retry the operation */
@@ -500,7 +507,8 @@ static ide_startstop_t ide_atapi_error(ide_drive_t *drive, struct request *rq, u
 
 	if (ide_read_status(drive) & (BUSY_STAT | DRQ_STAT))
 		/* force an abort */
-		hwif->OUTB(WIN_IDLEIMMEDIATE, IDE_COMMAND_REG);
+		hwif->OUTB(WIN_IDLEIMMEDIATE,
+			   hwif->io_ports[IDE_COMMAND_OFFSET]);
 
 	if (rq->errors >= ERROR_MAX) {
 		ide_kill_rq(drive, rq);
diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c
index e77cee0e5d65..45944219eea0 100644
--- a/drivers/ide/ide-iops.c
+++ b/drivers/ide/ide-iops.c
@@ -158,9 +158,12 @@ EXPORT_SYMBOL(default_hwif_mmiops);
 
 void SELECT_DRIVE (ide_drive_t *drive)
 {
-	if (HWIF(drive)->selectproc)
-		HWIF(drive)->selectproc(drive);
-	HWIF(drive)->OUTB(drive->select.all, IDE_SELECT_REG);
+	ide_hwif_t *hwif = drive->hwif;
+
+	if (hwif->selectproc)
+		hwif->selectproc(drive);
+
+	hwif->OUTB(drive->select.all, hwif->io_ports[IDE_SELECT_OFFSET]);
 }
 
 void SELECT_MASK (ide_drive_t *drive, int mask)
@@ -194,15 +197,18 @@ static void ata_input_data(ide_drive_t *drive, void *buffer, u32 wcount)
 	if (io_32bit) {
 		if (io_32bit & 2) {
 			unsigned long flags;
+
 			local_irq_save(flags);
-			ata_vlb_sync(drive, IDE_NSECTOR_REG);
-			hwif->INSL(IDE_DATA_REG, buffer, wcount);
+			ata_vlb_sync(drive, hwif->io_ports[IDE_NSECTOR_OFFSET]);
+			hwif->INSL(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+				   wcount);
 			local_irq_restore(flags);
 		} else
-			hwif->INSL(IDE_DATA_REG, buffer, wcount);
-	} else {
-		hwif->INSW(IDE_DATA_REG, buffer, wcount<<1);
-	}
+			hwif->INSL(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+				   wcount);
+	} else
+		hwif->INSW(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+			   wcount << 1);
 }
 
 /*
@@ -216,15 +222,18 @@ static void ata_output_data(ide_drive_t *drive, void *buffer, u32 wcount)
 	if (io_32bit) {
 		if (io_32bit & 2) {
 			unsigned long flags;
+
 			local_irq_save(flags);
-			ata_vlb_sync(drive, IDE_NSECTOR_REG);
-			hwif->OUTSL(IDE_DATA_REG, buffer, wcount);
+			ata_vlb_sync(drive, hwif->io_ports[IDE_NSECTOR_OFFSET]);
+			hwif->OUTSL(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+				    wcount);
 			local_irq_restore(flags);
 		} else
-			hwif->OUTSL(IDE_DATA_REG, buffer, wcount);
-	} else {
-		hwif->OUTSW(IDE_DATA_REG, buffer, wcount<<1);
-	}
+			hwif->OUTSL(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+				    wcount);
+	} else
+		hwif->OUTSW(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+			    wcount << 1);
 }
 
 /*
@@ -243,13 +252,15 @@ static void atapi_input_bytes(ide_drive_t *drive, void *buffer, u32 bytecount)
 #if defined(CONFIG_ATARI) || defined(CONFIG_Q40)
 	if (MACH_IS_ATARI || MACH_IS_Q40) {
 		/* Atari has a byte-swapped IDE interface */
-		insw_swapw(IDE_DATA_REG, buffer, bytecount / 2);
+		insw_swapw(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+			   bytecount / 2);
 		return;
 	}
 #endif /* CONFIG_ATARI || CONFIG_Q40 */
 	hwif->ata_input_data(drive, buffer, bytecount / 4);
 	if ((bytecount & 0x03) >= 2)
-		hwif->INSW(IDE_DATA_REG, ((u8 *)buffer)+(bytecount & ~0x03), 1);
+		hwif->INSW(hwif->io_ports[IDE_DATA_OFFSET],
+			   (u8 *)buffer + (bytecount & ~0x03), 1);
 }
 
 static void atapi_output_bytes(ide_drive_t *drive, void *buffer, u32 bytecount)
@@ -260,13 +271,15 @@ static void atapi_output_bytes(ide_drive_t *drive, void *buffer, u32 bytecount)
 #if defined(CONFIG_ATARI) || defined(CONFIG_Q40)
 	if (MACH_IS_ATARI || MACH_IS_Q40) {
 		/* Atari has a byte-swapped IDE interface */
-		outsw_swapw(IDE_DATA_REG, buffer, bytecount / 2);
+		outsw_swapw(hwif->io_ports[IDE_DATA_OFFSET], buffer,
+			    bytecount / 2);
 		return;
 	}
 #endif /* CONFIG_ATARI || CONFIG_Q40 */
 	hwif->ata_output_data(drive, buffer, bytecount / 4);
 	if ((bytecount & 0x03) >= 2)
-		hwif->OUTSW(IDE_DATA_REG, ((u8*)buffer)+(bytecount & ~0x03), 1);
+		hwif->OUTSW(hwif->io_ports[IDE_DATA_OFFSET],
+			    (u8 *)buffer + (bytecount & ~0x03), 1);
 }
 
 void default_hwif_transport(ide_hwif_t *hwif)
@@ -429,7 +442,7 @@ int drive_is_ready (ide_drive_t *drive)
 	 * an interrupt with another pci card/device.  We make no assumptions
 	 * about possible isa-pnp and pci-pnp issues yet.
 	 */
-	if (IDE_CONTROL_REG)
+	if (hwif->io_ports[IDE_CONTROL_OFFSET])
 		stat = ide_read_altstatus(drive);
 	else
 		/* Note: this may clear a pending IRQ!! */
@@ -631,7 +644,7 @@ int ide_driveid_update(ide_drive_t *drive)
 	SELECT_MASK(drive, 1);
 	ide_set_irq(drive, 1);
 	msleep(50);
-	hwif->OUTB(WIN_IDENTIFY, IDE_COMMAND_REG);
+	hwif->OUTB(WIN_IDENTIFY, hwif->io_ports[IDE_COMMAND_OFFSET]);
 	timeout = jiffies + WAIT_WORSTCASE;
 	do {
 		if (time_after(jiffies, timeout)) {
@@ -718,9 +731,10 @@ int ide_config_drive_speed(ide_drive_t *drive, u8 speed)
 	SELECT_MASK(drive, 0);
 	udelay(1);
 	ide_set_irq(drive, 0);
-	hwif->OUTB(speed, IDE_NSECTOR_REG);
-	hwif->OUTB(SETFEATURES_XFER, IDE_FEATURE_REG);
-	hwif->OUTBSYNC(drive, WIN_SETFEATURES, IDE_COMMAND_REG);
+	hwif->OUTB(speed, hwif->io_ports[IDE_NSECTOR_OFFSET]);
+	hwif->OUTB(SETFEATURES_XFER, hwif->io_ports[IDE_FEATURE_OFFSET]);
+	hwif->OUTBSYNC(drive, WIN_SETFEATURES,
+		       hwif->io_ports[IDE_COMMAND_OFFSET]);
 	if (drive->quirk_list == 2)
 		ide_set_irq(drive, 1);
 
@@ -828,7 +842,7 @@ void ide_execute_command(ide_drive_t *drive, u8 cmd, ide_handler_t *handler,
 
 	spin_lock_irqsave(&ide_lock, flags);
 	__ide_set_handler(drive, handler, timeout, expiry);
-	hwif->OUTBSYNC(drive, cmd, IDE_COMMAND_REG);
+	hwif->OUTBSYNC(drive, cmd, hwif->io_ports[IDE_COMMAND_OFFSET]);
 	/*
 	 * Drive takes 400nS to respond, we must avoid the IRQ being
 	 * serviced before that.
@@ -1009,7 +1023,8 @@ static ide_startstop_t do_reset1 (ide_drive_t *drive, int do_not_try_atapi)
 	unsigned long flags;
 	ide_hwif_t *hwif;
 	ide_hwgroup_t *hwgroup;
-	
+	u8 ctl;
+
 	spin_lock_irqsave(&ide_lock, flags);
 	hwif = HWIF(drive);
 	hwgroup = HWGROUP(drive);
@@ -1023,7 +1038,8 @@ static ide_startstop_t do_reset1 (ide_drive_t *drive, int do_not_try_atapi)
 		pre_reset(drive);
 		SELECT_DRIVE(drive);
 		udelay (20);
-		hwif->OUTBSYNC(drive, WIN_SRST, IDE_COMMAND_REG);
+		hwif->OUTBSYNC(drive, WIN_SRST,
+			       hwif->io_ports[IDE_COMMAND_OFFSET]);
 		ndelay(400);
 		hwgroup->poll_timeout = jiffies + WAIT_WORSTCASE;
 		hwgroup->polling = 1;
@@ -1039,7 +1055,7 @@ static ide_startstop_t do_reset1 (ide_drive_t *drive, int do_not_try_atapi)
 	for (unit = 0; unit < MAX_DRIVES; ++unit)
 		pre_reset(&hwif->drives[unit]);
 
-	if (!IDE_CONTROL_REG) {
+	if (hwif->io_ports[IDE_CONTROL_OFFSET] == 0) {
 		spin_unlock_irqrestore(&ide_lock, flags);
 		return ide_stopped;
 	}
@@ -1054,16 +1070,14 @@ static ide_startstop_t do_reset1 (ide_drive_t *drive, int do_not_try_atapi)
 	 * recover from reset very quickly, saving us the first 50ms wait time.
 	 */
 	/* set SRST and nIEN */
-	hwif->OUTBSYNC(drive, drive->ctl|6,IDE_CONTROL_REG);
+	hwif->OUTBSYNC(drive, drive->ctl|6, hwif->io_ports[IDE_CONTROL_OFFSET]);
 	/* more than enough time */
 	udelay(10);
-	if (drive->quirk_list == 2) {
-		/* clear SRST and nIEN */
-		hwif->OUTBSYNC(drive, drive->ctl, IDE_CONTROL_REG);
-	} else {
-		/* clear SRST, leave nIEN */
-		hwif->OUTBSYNC(drive, drive->ctl|2, IDE_CONTROL_REG);
-	}
+	if (drive->quirk_list == 2)
+		ctl = drive->ctl;	/* clear SRST and nIEN */
+	else
+		ctl = drive->ctl | 2;	/* clear SRST, leave nIEN */
+	hwif->OUTBSYNC(drive, ctl, hwif->io_ports[IDE_CONTROL_OFFSET]);
 	/* more than enough time */
 	udelay(10);
 	hwgroup->poll_timeout = jiffies + WAIT_WORSTCASE;
diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c
index 33cb5e5a249b..10ccf278d5be 100644
--- a/drivers/ide/ide-probe.c
+++ b/drivers/ide/ide-probe.c
@@ -271,7 +271,7 @@ static int actual_try_to_identify (ide_drive_t *drive, u8 cmd)
 	/* take a deep breath */
 	msleep(50);
 
-	if (IDE_CONTROL_REG) {
+	if (hwif->io_ports[IDE_CONTROL_OFFSET]) {
 		a = ide_read_altstatus(drive);
 		s = ide_read_status(drive);
 		if ((a ^ s) & ~INDEX_STAT)
@@ -289,10 +289,10 @@ static int actual_try_to_identify (ide_drive_t *drive, u8 cmd)
 	 */
 	if ((cmd == WIN_PIDENTIFY))
 		/* disable dma & overlap */
-		hwif->OUTB(0, IDE_FEATURE_REG);
+		hwif->OUTB(0, hwif->io_ports[IDE_FEATURE_OFFSET]);
 
 	/* ask drive for ID */
-	hwif->OUTB(cmd, IDE_COMMAND_REG);
+	hwif->OUTB(cmd, hwif->io_ports[IDE_COMMAND_OFFSET]);
 
 	timeout = ((cmd == WIN_IDENTIFY) ? WAIT_WORSTCASE : WAIT_PIDENTIFY) / 2;
 	timeout += jiffies;
@@ -353,7 +353,7 @@ static int try_to_identify (ide_drive_t *drive, u8 cmd)
 	 * interrupts during the identify-phase that
 	 * the irq handler isn't expecting.
 	 */
-	if (IDE_CONTROL_REG) {
+	if (hwif->io_ports[IDE_CONTROL_OFFSET]) {
 		if (!hwif->irq) {
 			autoprobe = 1;
 			cookie = probe_irq_on();
@@ -445,7 +445,8 @@ static int do_probe (ide_drive_t *drive, u8 cmd)
 	msleep(50);
 	SELECT_DRIVE(drive);
 	msleep(50);
-	if (hwif->INB(IDE_SELECT_REG) != drive->select.all && !drive->present) {
+	if (hwif->INB(hwif->io_ports[IDE_SELECT_OFFSET]) != drive->select.all &&
+	    !drive->present) {
 		if (drive->select.b.unit != 0) {
 			/* exit with drive0 selected */
 			SELECT_DRIVE(&hwif->drives[0]);
@@ -477,9 +478,11 @@ static int do_probe (ide_drive_t *drive, u8 cmd)
 			printk(KERN_ERR "%s: no response (status = 0x%02x), "
 					"resetting drive\n", drive->name, stat);
 			msleep(50);
-			hwif->OUTB(drive->select.all, IDE_SELECT_REG);
+			hwif->OUTB(drive->select.all,
+				   hwif->io_ports[IDE_SELECT_OFFSET]);
 			msleep(50);
-			hwif->OUTB(WIN_SRST, IDE_COMMAND_REG);
+			hwif->OUTB(WIN_SRST,
+				   hwif->io_ports[IDE_COMMAND_OFFSET]);
 			(void)ide_busy_sleep(hwif);
 			rc = try_to_identify(drive, cmd);
 		}
@@ -515,7 +518,7 @@ static void enable_nest (ide_drive_t *drive)
 	printk("%s: enabling %s -- ", hwif->name, drive->id->model);
 	SELECT_DRIVE(drive);
 	msleep(50);
-	hwif->OUTB(EXABYTE_ENABLE_NEST, IDE_COMMAND_REG);
+	hwif->OUTB(EXABYTE_ENABLE_NEST, hwif->io_ports[IDE_COMMAND_OFFSET]);
 
 	if (ide_busy_sleep(hwif)) {
 		printk(KERN_CONT "failed (timeout)\n");
diff --git a/drivers/ide/ide-tape.c b/drivers/ide/ide-tape.c
index bfdc4f449797..8c39146b6088 100644
--- a/drivers/ide/ide-tape.c
+++ b/drivers/ide/ide-tape.c
@@ -1117,10 +1117,10 @@ static ide_startstop_t idetape_pc_intr(ide_drive_t *drive)
 		return ide_do_reset(drive);
 	}
 	/* Get the number of bytes to transfer on this interrupt. */
-	bcount = (hwif->INB(IDE_BCOUNTH_REG) << 8) |
-		  hwif->INB(IDE_BCOUNTL_REG);
+	bcount = (hwif->INB(hwif->io_ports[IDE_BCOUNTH_OFFSET]) << 8) |
+		  hwif->INB(hwif->io_ports[IDE_BCOUNTL_OFFSET]);
 
-	ireason = hwif->INB(IDE_IREASON_REG);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 
 	if (ireason & CD) {
 		printk(KERN_ERR "ide-tape: CoD != 0 in %s\n", __func__);
@@ -1224,12 +1224,12 @@ static ide_startstop_t idetape_transfer_pc(ide_drive_t *drive)
 				"yet DRQ isn't asserted\n");
 		return startstop;
 	}
-	ireason = hwif->INB(IDE_IREASON_REG);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 	while (retries-- && ((ireason & CD) == 0 || (ireason & IO))) {
 		printk(KERN_ERR "ide-tape: (IO,CoD != (0,1) while issuing "
 				"a packet command, retrying\n");
 		udelay(100);
-		ireason = hwif->INB(IDE_IREASON_REG);
+		ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 		if (retries == 0) {
 			printk(KERN_ERR "ide-tape: (IO,CoD != (0,1) while "
 					"issuing a packet command, ignoring\n");
@@ -1323,7 +1323,7 @@ static ide_startstop_t idetape_issue_pc(ide_drive_t *drive, idetape_pc_t *pc)
 				    IDETAPE_WAIT_CMD, NULL);
 		return ide_started;
 	} else {
-		hwif->OUTB(WIN_PACKETCMD, IDE_COMMAND_REG);
+		hwif->OUTB(WIN_PACKETCMD, hwif->io_ports[IDE_COMMAND_OFFSET]);
 		return idetape_transfer_pc(drive);
 	}
 }
diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c
index 4c86a8d84b4c..155cc904f4eb 100644
--- a/drivers/ide/ide-taskfile.c
+++ b/drivers/ide/ide-taskfile.c
@@ -59,32 +59,34 @@ void ide_tf_load(ide_drive_t *drive, ide_task_t *task)
 		SELECT_MASK(drive, 0);
 
 	if (task->tf_flags & IDE_TFLAG_OUT_DATA)
-		hwif->OUTW((tf->hob_data << 8) | tf->data, IDE_DATA_REG);
+		hwif->OUTW((tf->hob_data << 8) | tf->data,
+			   hwif->io_ports[IDE_DATA_OFFSET]);
 
 	if (task->tf_flags & IDE_TFLAG_OUT_HOB_FEATURE)
-		hwif->OUTB(tf->hob_feature, IDE_FEATURE_REG);
+		hwif->OUTB(tf->hob_feature, hwif->io_ports[IDE_FEATURE_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_HOB_NSECT)
-		hwif->OUTB(tf->hob_nsect, IDE_NSECTOR_REG);
+		hwif->OUTB(tf->hob_nsect, hwif->io_ports[IDE_NSECTOR_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_HOB_LBAL)
-		hwif->OUTB(tf->hob_lbal, IDE_SECTOR_REG);
+		hwif->OUTB(tf->hob_lbal, hwif->io_ports[IDE_SECTOR_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_HOB_LBAM)
-		hwif->OUTB(tf->hob_lbam, IDE_LCYL_REG);
+		hwif->OUTB(tf->hob_lbam, hwif->io_ports[IDE_LCYL_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_HOB_LBAH)
-		hwif->OUTB(tf->hob_lbah, IDE_HCYL_REG);
+		hwif->OUTB(tf->hob_lbah, hwif->io_ports[IDE_HCYL_OFFSET]);
 
 	if (task->tf_flags & IDE_TFLAG_OUT_FEATURE)
-		hwif->OUTB(tf->feature, IDE_FEATURE_REG);
+		hwif->OUTB(tf->feature, hwif->io_ports[IDE_FEATURE_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_NSECT)
-		hwif->OUTB(tf->nsect, IDE_NSECTOR_REG);
+		hwif->OUTB(tf->nsect, hwif->io_ports[IDE_NSECTOR_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_LBAL)
-		hwif->OUTB(tf->lbal, IDE_SECTOR_REG);
+		hwif->OUTB(tf->lbal, hwif->io_ports[IDE_SECTOR_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_LBAM)
-		hwif->OUTB(tf->lbam, IDE_LCYL_REG);
+		hwif->OUTB(tf->lbam, hwif->io_ports[IDE_LCYL_OFFSET]);
 	if (task->tf_flags & IDE_TFLAG_OUT_LBAH)
-		hwif->OUTB(tf->lbah, IDE_HCYL_REG);
+		hwif->OUTB(tf->lbah, hwif->io_ports[IDE_HCYL_OFFSET]);
 
 	if (task->tf_flags & IDE_TFLAG_OUT_DEVICE)
-		hwif->OUTB((tf->device & HIHI) | drive->select.all, IDE_SELECT_REG);
+		hwif->OUTB((tf->device & HIHI) | drive->select.all,
+			   hwif->io_ports[IDE_SELECT_OFFSET]);
 }
 
 int taskfile_lib_get_identify (ide_drive_t *drive, u8 *buf)
@@ -152,7 +154,8 @@ ide_startstop_t do_rw_taskfile (ide_drive_t *drive, ide_task_t *task)
 	switch (task->data_phase) {
 	case TASKFILE_MULTI_OUT:
 	case TASKFILE_OUT:
-		hwif->OUTBSYNC(drive, tf->command, IDE_COMMAND_REG);
+		hwif->OUTBSYNC(drive, tf->command,
+			       hwif->io_ports[IDE_COMMAND_OFFSET]);
 		ndelay(400);	/* FIXME */
 		return pre_task_out_intr(drive, task->rq);
 	case TASKFILE_MULTI_IN:
diff --git a/drivers/ide/legacy/ht6560b.c b/drivers/ide/legacy/ht6560b.c
index 78ca68e60f97..314e6c6aeb6c 100644
--- a/drivers/ide/legacy/ht6560b.c
+++ b/drivers/ide/legacy/ht6560b.c
@@ -82,7 +82,7 @@
  * out how they setup those cycle time interfacing values, as they at Holtek
  * call them. IDESETUP.COM that is supplied with the drivers figures out
  * optimal values and fetches those values to drivers. I found out that
- * they use IDE_SELECT_REG to fetch timings to the ide board right after
+ * they use Select register to fetch timings to the ide board right after
  * interface switching. After that it was quite easy to add code to
  * ht6560b.c.
  *
@@ -127,6 +127,7 @@
  */
 static void ht6560b_selectproc (ide_drive_t *drive)
 {
+	ide_hwif_t *hwif = drive->hwif;
 	unsigned long flags;
 	static u8 current_select = 0;
 	static u8 current_timing = 0;
@@ -155,8 +156,8 @@ static void ht6560b_selectproc (ide_drive_t *drive)
 		/*
 		 * Set timing for this drive:
 		 */
-		outb(timing, IDE_SELECT_REG);
-		(void)inb(IDE_STATUS_REG);
+		outb(timing, hwif->io_ports[IDE_SELECT_OFFSET]);
+		(void)inb(hwif->io_ports[IDE_STATUS_OFFSET]);
 #ifdef DEBUG
 		printk("ht6560b: %s: select=%#x timing=%#x\n",
 			drive->name, select, timing);
@@ -193,9 +194,9 @@ static int __init try_to_init_ht6560b(void)
 	 * Ht6560b autodetected
 	 */
 	outb(HT_CONFIG_DEFAULT, HT_CONFIG_PORT);
-	outb(HT_TIMING_DEFAULT, 0x1f6);  /* IDE_SELECT_REG */
-	(void) inb(0x1f7);               /* IDE_STATUS_REG */
-	
+	outb(HT_TIMING_DEFAULT, 0x1f6);	/* Select register */
+	(void)inb(0x1f7);		/* Status register */
+
 	printk("ht6560b " HT6560B_VERSION
 	       ": chipset detected and initialized"
 #ifdef DEBUG
diff --git a/drivers/ide/pci/hpt366.c b/drivers/ide/pci/hpt366.c
index d03a231d965e..82d0e318a1fe 100644
--- a/drivers/ide/pci/hpt366.c
+++ b/drivers/ide/pci/hpt366.c
@@ -760,7 +760,7 @@ static void hpt3xx_maskproc(ide_drive_t *drive, int mask)
 		}
 	} else
 		outb(mask ? (drive->ctl | 2) : (drive->ctl & ~2),
-		     IDE_CONTROL_REG);
+		     hwif->io_ports[IDE_CONTROL_OFFSET]);
 }
 
 /*
diff --git a/drivers/ide/pci/scc_pata.c b/drivers/ide/pci/scc_pata.c
index 1a560dc1eac1..ef07c7a8b97a 100644
--- a/drivers/ide/pci/scc_pata.c
+++ b/drivers/ide/pci/scc_pata.c
@@ -334,7 +334,8 @@ static int scc_ide_dma_end(ide_drive_t * drive)
 
 	/* errata A308 workaround: Step5 (check data loss) */
 	/* We don't check non ide_disk because it is limited to UDMA4 */
-	if (!(in_be32((void __iomem *)IDE_ALTSTATUS_REG) & ERR_STAT) &&
+	if (!(in_be32((void __iomem *)hwif->io_ports[IDE_ALTSTATUS_OFFSET])
+	      & ERR_STAT) &&
 	    drive->media == ide_disk && drive->current_speed > XFER_UDMA_4) {
 		reg = in_be32((void __iomem *)intsts_port);
 		if (!(reg & INTSTS_ACTEINT)) {
@@ -437,7 +438,8 @@ static int scc_dma_test_irq(ide_drive_t *drive)
 	u32 int_stat = in_be32((void __iomem *)hwif->dma_base + 0x014);
 
 	/* SCC errata A252,A308 workaround: Step4 */
-	if ((in_be32((void __iomem *)IDE_ALTSTATUS_REG) & ERR_STAT) &&
+	if ((in_be32((void __iomem *)hwif->io_ports[IDE_ALTSTATUS_OFFSET])
+	     & ERR_STAT) &&
 	    (int_stat & INTSTS_INTRQ))
 		return 1;
 
diff --git a/drivers/ide/pci/sgiioc4.c b/drivers/ide/pci/sgiioc4.c
index 9046a69117ff..9d1a3038af9b 100644
--- a/drivers/ide/pci/sgiioc4.c
+++ b/drivers/ide/pci/sgiioc4.c
@@ -112,10 +112,9 @@ static void
 sgiioc4_maskproc(ide_drive_t * drive, int mask)
 {
 	writeb(mask ? (drive->ctl | 2) : (drive->ctl & ~2),
-	       (void __iomem *)IDE_CONTROL_REG);
+	       (void __iomem *)drive->hwif->io_ports[IDE_CONTROL_OFFSET]);
 }
 
-
 static int
 sgiioc4_checkirq(ide_hwif_t * hwif)
 {
@@ -142,18 +141,18 @@ sgiioc4_clearirq(ide_drive_t * drive)
 	intr_reg = readl((void __iomem *)other_ir);
 	if (intr_reg & 0x03) { /* Valid IOC4-IDE interrupt */
 		/*
-		 * Using sgiioc4_INB to read the IDE_STATUS_REG has a side effect
-		 * of clearing the interrupt.  The first read should clear it
-		 * if it is set.  The second read should return a "clear" status
-		 * if it got cleared.  If not, then spin for a bit trying to
-		 * clear it.
+		 * Using sgiioc4_INB to read the Status register has a side
+		 * effect of clearing the interrupt.  The first read should
+		 * clear it if it is set.  The second read should return
+		 * a "clear" status if it got cleared.  If not, then spin
+		 * for a bit trying to clear it.
 		 */
-		u8 stat = sgiioc4_INB(IDE_STATUS_REG);
+		u8 stat = sgiioc4_INB(hwif->io_ports[IDE_STATUS_OFFSET]);
 		int count = 0;
-		stat = sgiioc4_INB(IDE_STATUS_REG);
+		stat = sgiioc4_INB(hwif->io_ports[IDE_STATUS_OFFSET]);
 		while ((stat & 0x80) && (count++ < 100)) {
 			udelay(1);
-			stat = sgiioc4_INB(IDE_STATUS_REG);
+			stat = sgiioc4_INB(hwif->io_ports[IDE_STATUS_OFFSET]);
 		}
 
 		if (intr_reg & 0x02) {
diff --git a/drivers/ide/ppc/pmac.c b/drivers/ide/ppc/pmac.c
index b37dcfbdaad6..03a77713caf2 100644
--- a/drivers/ide/ppc/pmac.c
+++ b/drivers/ide/ppc/pmac.c
@@ -450,7 +450,8 @@ pmac_ide_init_hwif_ports(hw_regs_t *hw,
 	hw->dev = &pmac_ide[ix].mdev->ofdev.dev;
 }
 
-#define PMAC_IDE_REG(x) ((void __iomem *)(IDE_DATA_REG+(x)))
+#define PMAC_IDE_REG(x) \
+	((void __iomem *)((drive)->hwif->io_ports[IDE_DATA_OFFSET] + (x)))
 
 /*
  * Apply the timings of the proper unit (master/slave) to the shared
diff --git a/drivers/scsi/ide-scsi.c b/drivers/scsi/ide-scsi.c
index 3c3b3502c4d4..7fea769cf291 100644
--- a/drivers/scsi/ide-scsi.c
+++ b/drivers/scsi/ide-scsi.c
@@ -275,9 +275,12 @@ static int idescsi_end_request(ide_drive_t *, int, int);
 static ide_startstop_t
 idescsi_atapi_error(ide_drive_t *drive, struct request *rq, u8 stat, u8 err)
 {
+	ide_hwif_t *hwif = drive->hwif;
+
 	if (ide_read_status(drive) & (BUSY_STAT | DRQ_STAT))
 		/* force an abort */
-		HWIF(drive)->OUTB(WIN_IDLEIMMEDIATE,IDE_COMMAND_REG);
+		hwif->OUTB(WIN_IDLEIMMEDIATE,
+			   hwif->io_ports[IDE_COMMAND_OFFSET]);
 
 	rq->errors++;
 
@@ -423,9 +426,9 @@ static ide_startstop_t idescsi_pc_intr (ide_drive_t *drive)
 		idescsi_end_request (drive, 1, 0);
 		return ide_stopped;
 	}
-	bcount = (hwif->INB(IDE_BCOUNTH_REG) << 8) |
-		  hwif->INB(IDE_BCOUNTL_REG);
-	ireason = hwif->INB(IDE_IREASON_REG);
+	bcount = (hwif->INB(hwif->io_ports[IDE_BCOUNTH_OFFSET]) << 8) |
+		  hwif->INB(hwif->io_ports[IDE_BCOUNTL_OFFSET]);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 
 	if (ireason & CD) {
 		printk(KERN_ERR "ide-scsi: CoD != 0 in idescsi_pc_intr\n");
@@ -497,7 +500,7 @@ static ide_startstop_t idescsi_transfer_pc(ide_drive_t *drive)
 			"initiated yet DRQ isn't asserted\n");
 		return startstop;
 	}
-	ireason = hwif->INB(IDE_IREASON_REG);
+	ireason = hwif->INB(hwif->io_ports[IDE_IREASON_OFFSET]);
 	if ((ireason & CD) == 0 || (ireason & IO)) {
 		printk(KERN_ERR "ide-scsi: (IO,CoD) != (0,1) while "
 				"issuing a packet command\n");
@@ -587,7 +590,7 @@ static ide_startstop_t idescsi_issue_pc (ide_drive_t *drive, idescsi_pc_t *pc)
 		return ide_started;
 	} else {
 		/* Issue the packet command */
-		HWIF(drive)->OUTB(WIN_PACKETCMD, IDE_COMMAND_REG);
+		hwif->OUTB(WIN_PACKETCMD, hwif->io_ports[IDE_COMMAND_OFFSET]);
 		return idescsi_transfer_pc(drive);
 	}
 }
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 2eb99cab4a3d..3b691cce00e1 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -82,24 +82,10 @@ typedef unsigned char	byte;	/* used everywhere */
 
 #define IDE_FEATURE_OFFSET	IDE_ERROR_OFFSET
 #define IDE_COMMAND_OFFSET	IDE_STATUS_OFFSET
-
-#define IDE_DATA_REG		(HWIF(drive)->io_ports[IDE_DATA_OFFSET])
-#define IDE_ERROR_REG		(HWIF(drive)->io_ports[IDE_ERROR_OFFSET])
-#define IDE_NSECTOR_REG		(HWIF(drive)->io_ports[IDE_NSECTOR_OFFSET])
-#define IDE_SECTOR_REG		(HWIF(drive)->io_ports[IDE_SECTOR_OFFSET])
-#define IDE_LCYL_REG		(HWIF(drive)->io_ports[IDE_LCYL_OFFSET])
-#define IDE_HCYL_REG		(HWIF(drive)->io_ports[IDE_HCYL_OFFSET])
-#define IDE_SELECT_REG		(HWIF(drive)->io_ports[IDE_SELECT_OFFSET])
-#define IDE_STATUS_REG		(HWIF(drive)->io_ports[IDE_STATUS_OFFSET])
-#define IDE_CONTROL_REG		(HWIF(drive)->io_ports[IDE_CONTROL_OFFSET])
-#define IDE_IRQ_REG		(HWIF(drive)->io_ports[IDE_IRQ_OFFSET])
-
-#define IDE_FEATURE_REG		IDE_ERROR_REG
-#define IDE_COMMAND_REG		IDE_STATUS_REG
-#define IDE_ALTSTATUS_REG	IDE_CONTROL_REG
-#define IDE_IREASON_REG		IDE_NSECTOR_REG
-#define IDE_BCOUNTL_REG		IDE_LCYL_REG
-#define IDE_BCOUNTH_REG		IDE_HCYL_REG
+#define IDE_ALTSTATUS_OFFSET	IDE_CONTROL_OFFSET
+#define IDE_IREASON_OFFSET	IDE_NSECTOR_OFFSET
+#define IDE_BCOUNTL_OFFSET	IDE_LCYL_OFFSET
+#define IDE_BCOUNTH_OFFSET	IDE_HCYL_OFFSET
 
 #define OK_STAT(stat,good,bad)	(((stat)&((good)|(bad)))==(good))
 #define BAD_R_STAT		(BUSY_STAT   | ERR_STAT)
@@ -369,7 +355,7 @@ typedef struct ide_drive_s {
         u8	wcache;		/* status of write cache */
 	u8	acoustic;	/* acoustic management */
 	u8	media;		/* disk, cdrom, tape, floppy, ... */
-	u8	ctl;		/* "normal" value for IDE_CONTROL_REG */
+	u8	ctl;		/* "normal" value for Control register */
 	u8	ready_stat;	/* min status value for drive ready */
 	u8	mult_count;	/* current multiple sector setting */
 	u8	mult_req;	/* requested multiple sector setting */
@@ -1273,7 +1259,10 @@ static inline ide_drive_t *ide_get_paired_drive(ide_drive_t *drive)
 
 static inline void ide_set_irq(ide_drive_t *drive, int on)
 {
-	drive->hwif->OUTB(drive->ctl | (on ? 0 : 2), IDE_CONTROL_REG);
+	ide_hwif_t *hwif = drive->hwif;
+
+	hwif->OUTB(drive->ctl | (on ? 0 : 2),
+		   hwif->io_ports[IDE_CONTROL_OFFSET]);
 }
 
 static inline u8 ide_read_status(ide_drive_t *drive)
-- 
cgit v1.2.3


From 8303b46e18b58b2d0257e6842e60b50ac880a6d1 Mon Sep 17 00:00:00 2001
From: Borislav Petkov <petkovbb@googlemail.com>
Date: Fri, 18 Apr 2008 00:46:26 +0200
Subject: ide: add generic packet command representation ide_atapi_pc

This new struct unifies ide{-floppy,-tape,-scsi}'s view of a packet command. For now,
it represents the common denominator between the three drivers while adding driver-
specific members at the end of the struct which will be merged/simplified into the
generic ATAPI handling code in later steps, or removed completely.

Bart:
- move struct ide_atapi_pc outside of #ifdef/#endif CONFIG_IDE_PROC_FS

Signed-off-by: Borislav Petkov <petkovbb@gmail.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 include/linux/ide.h | 47 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/ide.h b/include/linux/ide.h
index 3b691cce00e1..e9b1def1e66a 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -595,6 +595,53 @@ int set_io_32bit(ide_drive_t *, int);
 int set_pio_mode(ide_drive_t *, int);
 int set_using_dma(ide_drive_t *, int);
 
+struct ide_atapi_pc {
+	/* actual packet bytes */
+	u8 c[12];
+	/* incremented on each retry */
+	int retries;
+	int error;
+
+	/* bytes to transfer */
+	int req_xfer;
+	/* bytes actually transferred */
+	int xferred;
+
+	/* data buffer */
+	u8 *buf;
+	/* current buffer position */
+	u8 *cur_pos;
+	int buf_size;
+	/* missing/available data on the current buffer */
+	int b_count;
+
+	/* the corresponding request */
+	struct request *rq;
+
+	unsigned long flags;
+
+	/*
+	 * those are more or less driver-specific and some of them are subject
+	 * to change/removal later.
+	 */
+	u8 pc_buf[256];
+	void (*idefloppy_callback) (ide_drive_t *);
+	ide_startstop_t (*idetape_callback) (ide_drive_t *);
+
+	/* idetape only */
+	struct idetape_bh *bh;
+	char *b_data;
+
+	/* idescsi only for now */
+	struct scatterlist *sg;
+	unsigned int sg_cnt;
+
+	struct scsi_cmnd *scsi_cmd;
+	void (*done) (struct scsi_cmnd *);
+
+	unsigned long timeout;
+};
+
 #ifdef CONFIG_IDE_PROC_FS
 /*
  * configurable drive settings
-- 
cgit v1.2.3


From eaec3e7ded9dbc88bad393c076b65f4b7b11d30d Mon Sep 17 00:00:00 2001
From: Borislav Petkov <petkovbb@googlemail.com>
Date: Fri, 18 Apr 2008 00:46:27 +0200
Subject: ide: use generic ATAPI packet command flags in ide-{floppy,tape}

Signed-off-by: Borislav Petkov <petkovbb@gmail.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/ide-floppy.c | 14 --------------
 drivers/ide/ide-tape.c   | 16 ----------------
 include/linux/ide.h      | 15 +++++++++++++++
 3 files changed, 15 insertions(+), 30 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/ide-floppy.c b/drivers/ide/ide-floppy.c
index bf1ef60a5a07..5f133dfb541c 100644
--- a/drivers/ide/ide-floppy.c
+++ b/drivers/ide/ide-floppy.c
@@ -78,20 +78,6 @@
  */
 #define IDEFLOPPY_PC_STACK		(10 + IDEFLOPPY_MAX_PC_RETRIES)
 
-/* Packet command flag bits. */
-enum {
-	/* 1 when we prefer to use DMA if possible */
-	PC_FLAG_DMA_RECOMMENDED	= (1 << 0),
-	/* 1 while DMA in progress */
-	PC_FLAG_DMA_IN_PROGRESS	= (1 << 1),
-	/* 1 when encountered problem during DMA */
-	PC_FLAG_DMA_ERROR	= (1 << 2),
-	/* Data direction */
-	PC_FLAG_WRITING		= (1 << 3),
-	/* Suppress error reporting */
-	PC_FLAG_SUPPRESS_ERROR	= (1 << 4),
-};
-
 /* format capacities descriptor codes */
 #define CAPACITY_INVALID	0x00
 #define CAPACITY_UNFORMATTED	0x01
diff --git a/drivers/ide/ide-tape.c b/drivers/ide/ide-tape.c
index 3f9dcca6f092..f43fd070f1b6 100644
--- a/drivers/ide/ide-tape.c
+++ b/drivers/ide/ide-tape.c
@@ -181,22 +181,6 @@ struct idetape_bh {
 	char *b_data;
 };
 
-/* Packet command flag bits. */
-enum {
-	/* Set when an error is considered normal - We won't retry */
-	PC_FLAG_ABORT		= (1 << 0),
-	/* 1 When polling for DSC on a media access command */
-	PC_FLAG_WAIT_FOR_DSC	= (1 << 1),
-	/* 1 when we prefer to use DMA if possible */
-	PC_FLAG_DMA_RECOMMENDED	= (1 << 2),
-	/* 1 while DMA in progress */
-	PC_FLAG_DMA_IN_PROGRESS	= (1 << 3),
-	/* 1 when encountered problem during DMA */
-	PC_FLAG_DMA_ERROR	= (1 << 4),
-	/* Data direction */
-	PC_FLAG_WRITING		= (1 <<	5),
-};
-
 /* Tape door status */
 #define DOOR_UNLOCKED			0
 #define DOOR_LOCKED			1
diff --git a/include/linux/ide.h b/include/linux/ide.h
index e9b1def1e66a..65445b7efc63 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -595,6 +595,21 @@ int set_io_32bit(ide_drive_t *, int);
 int set_pio_mode(ide_drive_t *, int);
 int set_using_dma(ide_drive_t *, int);
 
+/* ATAPI packet command flags */
+enum {
+	/* set when an error is considered normal - no retry (ide-tape) */
+	PC_FLAG_ABORT			= (1 << 0),
+	PC_FLAG_SUPPRESS_ERROR		= (1 << 1),
+	PC_FLAG_WAIT_FOR_DSC		= (1 << 2),
+	PC_FLAG_DMA_OK			= (1 << 3),
+	PC_FLAG_DMA_RECOMMENDED		= (1 << 4),
+	PC_FLAG_DMA_IN_PROGRESS		= (1 << 5),
+	PC_FLAG_DMA_ERROR		= (1 << 6),
+	PC_FLAG_WRITING			= (1 << 7),
+	/* command timed out */
+	PC_FLAG_TIMEDOUT		= (1 << 8),
+};
+
 struct ide_atapi_pc {
 	/* actual packet bytes */
 	u8 c[12];
-- 
cgit v1.2.3


From 0e33555fffdc8490630d98070e76e5fe031bcac2 Mon Sep 17 00:00:00 2001
From: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Date: Fri, 18 Apr 2008 00:46:33 +0200
Subject: ide: add CONFIG_IDE_ARCH_OBSOLETE_DEFAULTS (take 2)

* Add CONFIG_IDE_ARCH_OBSOLETE_DEFAULTS to drivers/ide/Kconfig and use
  it instead of defining IDE_ARCH_OBSOLETE_DEFAULTS in <arch/ide.h>.

v2:
* Define ide_default_irq() in ide-probe.c/ns87415.c if not already defined
  and drop defining ide_default_irq() for CONFIG_IDE_ARCH_OBSOLETE_DEFAULTS=n.

  [ Thanks to Stephen Rothwell and David Miller for noticing the problem. ]

Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: David Miller <davem@davemloft.net>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
---
 drivers/ide/Kconfig                 | 3 +++
 drivers/ide/ide-probe.c             | 4 ++++
 drivers/ide/ide.c                   | 4 ++++
 drivers/ide/pci/ns87415.c           | 4 ++++
 include/asm-alpha/ide.h             | 3 ---
 include/asm-ia64/ide.h              | 2 --
 include/asm-m32r/ide.h              | 2 --
 include/asm-mips/mach-generic/ide.h | 2 --
 include/asm-powerpc/ide.h           | 2 --
 include/asm-x86/ide.h               | 2 --
 include/linux/ide.h                 | 7 -------
 11 files changed, 15 insertions(+), 20 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/ide/Kconfig b/drivers/ide/Kconfig
index 4dc2761e9704..a57893c03b7a 100644
--- a/drivers/ide/Kconfig
+++ b/drivers/ide/Kconfig
@@ -1092,6 +1092,9 @@ config BLK_DEV_IDEDMA
 config IDE_ARCH_OBSOLETE_INIT
 	def_bool ALPHA || (ARM && !ARCH_L7200) || BLACKFIN || X86 || IA64 || M32R || MIPS || PARISC || PPC || (SUPERH64 && BLK_DEV_IDEPCI) || SPARC
 
+config IDE_ARCH_OBSOLETE_DEFAULTS
+	def_bool ALPHA || X86 || IA64 || M32R || MIPS || PPC32
+
 endif
 
 config BLK_DEV_HD_ONLY
diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c
index 10ccf278d5be..6a196c27b0aa 100644
--- a/drivers/ide/ide-probe.c
+++ b/drivers/ide/ide-probe.c
@@ -1233,6 +1233,10 @@ static void drive_release_dev (struct device *dev)
 	complete(&drive->gendev_rel_comp);
 }
 
+#ifndef ide_default_irq
+#define ide_default_irq(irq) 0
+#endif
+
 static int hwif_init(ide_hwif_t *hwif)
 {
 	int old_irq;
diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c
index c2fb5c964a51..a1a02c74d77f 100644
--- a/drivers/ide/ide.c
+++ b/drivers/ide/ide.c
@@ -165,6 +165,10 @@ static void ide_port_init_devices_data(ide_hwif_t *hwif)
 	}
 }
 
+#ifndef CONFIG_IDE_ARCH_OBSOLETE_DEFAULTS
+# define ide_default_io_base(index)	(0)
+# define ide_init_default_irq(base)	(0)
+#endif
 
 /*
  * init_ide_data() sets reasonable default values into all fields
diff --git a/drivers/ide/pci/ns87415.c b/drivers/ide/pci/ns87415.c
index bf0d3b2931f1..75513320aad9 100644
--- a/drivers/ide/pci/ns87415.c
+++ b/drivers/ide/pci/ns87415.c
@@ -181,6 +181,10 @@ static int ns87415_ide_dma_setup(ide_drive_t *drive)
 	return 1;
 }
 
+#ifndef ide_default_irq
+#define ide_default_irq(irq) 0
+#endif
+
 static void __devinit init_hwif_ns87415 (ide_hwif_t *hwif)
 {
 	struct pci_dev *dev = to_pci_dev(hwif->dev);
diff --git a/include/asm-alpha/ide.h b/include/asm-alpha/ide.h
index b7bf68d0407b..a2feed30bb68 100644
--- a/include/asm-alpha/ide.h
+++ b/include/asm-alpha/ide.h
@@ -13,9 +13,6 @@
 
 #ifdef __KERNEL__
 
-
-#define IDE_ARCH_OBSOLETE_DEFAULTS
-
 static inline int ide_default_irq(unsigned long base)
 {
 	switch (base) {
diff --git a/include/asm-ia64/ide.h b/include/asm-ia64/ide.h
index 1ccf23809329..09c2a05e1c8a 100644
--- a/include/asm-ia64/ide.h
+++ b/include/asm-ia64/ide.h
@@ -16,8 +16,6 @@
 
 #include <linux/irq.h>
 
-#define IDE_ARCH_OBSOLETE_DEFAULTS
-
 static inline int ide_default_irq(unsigned long base)
 {
 	switch (base) {
diff --git a/include/asm-m32r/ide.h b/include/asm-m32r/ide.h
index 5d2044e529ab..feb7f0d7aca9 100644
--- a/include/asm-m32r/ide.h
+++ b/include/asm-m32r/ide.h
@@ -23,8 +23,6 @@
 # endif
 #endif
 
-#define IDE_ARCH_OBSOLETE_DEFAULTS
-
 static __inline__ int ide_default_irq(unsigned long base)
 {
 	switch (base) {
diff --git a/include/asm-mips/mach-generic/ide.h b/include/asm-mips/mach-generic/ide.h
index 4ec2b930dfbb..45e24474cf43 100644
--- a/include/asm-mips/mach-generic/ide.h
+++ b/include/asm-mips/mach-generic/ide.h
@@ -27,8 +27,6 @@
 # endif
 #endif
 
-#define IDE_ARCH_OBSOLETE_DEFAULTS
-
 static __inline__ int ide_probe_legacy(void)
 {
 #ifdef CONFIG_PCI
diff --git a/include/asm-powerpc/ide.h b/include/asm-powerpc/ide.h
index 06549456c953..fef2ef1dbe86 100644
--- a/include/asm-powerpc/ide.h
+++ b/include/asm-powerpc/ide.h
@@ -31,8 +31,6 @@
 #include <linux/hdreg.h>
 #include <linux/ioport.h>
 
-#define IDE_ARCH_OBSOLETE_DEFAULTS
-
 /* FIXME: use ide_platform host driver */
 static __inline__ int ide_default_irq(unsigned long base)
 {
diff --git a/include/asm-x86/ide.h b/include/asm-x86/ide.h
index c2552d8bebf7..58080a7111de 100644
--- a/include/asm-x86/ide.h
+++ b/include/asm-x86/ide.h
@@ -20,8 +20,6 @@
 # endif
 #endif
 
-#define IDE_ARCH_OBSOLETE_DEFAULTS
-
 static __inline__ int ide_default_irq(unsigned long base)
 {
 	switch (base) {
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 65445b7efc63..6c39482fd1a1 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -194,13 +194,6 @@ static inline void ide_std_init_ports(hw_regs_t *hw,
 #define MAX_HWIFS	CONFIG_IDE_MAX_HWIFS
 #endif
 
-/* needed on alpha, x86/x86_64, ia64, mips, ppc32 and sh */
-#ifndef IDE_ARCH_OBSOLETE_DEFAULTS
-# define ide_default_io_base(index)	(0)
-# define ide_default_irq(base)		(0)
-# define ide_init_default_irq(base)	(0)
-#endif
-
 /* Currently only m68k, apus and m8xx need it */
 #ifndef IDE_ARCH_ACK_INTR
 # define ide_ack_intr(hwif) (1)
-- 
cgit v1.2.3


From 8a076191f373abaeb4aa5f6755d22e49db98940f Mon Sep 17 00:00:00 2001
From: "Ahmed S. Darwish" <darwish.07@gmail.com>
Date: Sat, 1 Mar 2008 21:51:09 +0200
Subject: LSM: Introduce inode_getsecid and ipc_getsecid hooks

Introduce inode_getsecid(inode, secid) and ipc_getsecid(ipcp, secid)
LSM hooks. These hooks will be used instead of similar exported
SELinux interfaces.

Let {inode,ipc,task}_getsecid hooks set the secid to 0 by default
if CONFIG_SECURITY is not defined or if the hook is set to
NULL (dummy). This is done to notify the caller that no valid
secid exists.

Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
---
 include/linux/security.h | 30 +++++++++++++++++++++++++++++-
 security/dummy.c         | 16 +++++++++++++++-
 security/security.c      | 10 ++++++++++
 3 files changed, 54 insertions(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/security.h b/include/linux/security.h
index c673dfd4dffc..45717d9d9656 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -468,6 +468,11 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
  *	@dentry is the dentry being changed.
  *	Return 0 on success.  If error is returned, then the operation
  *	causing setuid bit removal is failed.
+ * @inode_getsecid:
+ *	Get the secid associated with the node.
+ *	@inode contains a pointer to the inode.
+ *	@secid contains a pointer to the location where result will be saved.
+ *	In case of failure, @secid will be set to zero.
  *
  * Security hooks for file operations
  *
@@ -636,6 +641,8 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
  * @task_getsecid:
  *	Retrieve the security identifier of the process @p.
  *	@p contains the task_struct for the process and place is into @secid.
+ *	In case of failure, @secid will be set to zero.
+ *
  * @task_setgroups:
  *	Check permission before setting the supplementary group set of the
  *	current process.
@@ -997,6 +1004,11 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
  *	@ipcp contains the kernel IPC permission structure
  *	@flag contains the desired (requested) permission set
  *	Return 0 if permission is granted.
+ * @ipc_getsecid:
+ *	Get the secid associated with the ipc object.
+ *	@ipcp contains the kernel IPC permission structure.
+ *	@secid contains a pointer to the location where result will be saved.
+ *	In case of failure, @secid will be set to zero.
  *
  * Security hooks for individual messages held in System V IPC message queues
  * @msg_msg_alloc_security:
@@ -1317,6 +1329,7 @@ struct security_operations {
 	int (*inode_getsecurity)(const struct inode *inode, const char *name, void **buffer, bool alloc);
   	int (*inode_setsecurity)(struct inode *inode, const char *name, const void *value, size_t size, int flags);
   	int (*inode_listsecurity)(struct inode *inode, char *buffer, size_t buffer_size);
+	void (*inode_getsecid)(const struct inode *inode, u32 *secid);
 
 	int (*file_permission) (struct file * file, int mask);
 	int (*file_alloc_security) (struct file * file);
@@ -1369,6 +1382,7 @@ struct security_operations {
 	void (*task_to_inode)(struct task_struct *p, struct inode *inode);
 
 	int (*ipc_permission) (struct kern_ipc_perm * ipcp, short flag);
+	void (*ipc_getsecid) (struct kern_ipc_perm *ipcp, u32 *secid);
 
 	int (*msg_msg_alloc_security) (struct msg_msg * msg);
 	void (*msg_msg_free_security) (struct msg_msg * msg);
@@ -1578,6 +1592,7 @@ int security_inode_killpriv(struct dentry *dentry);
 int security_inode_getsecurity(const struct inode *inode, const char *name, void **buffer, bool alloc);
 int security_inode_setsecurity(struct inode *inode, const char *name, const void *value, size_t size, int flags);
 int security_inode_listsecurity(struct inode *inode, char *buffer, size_t buffer_size);
+void security_inode_getsecid(const struct inode *inode, u32 *secid);
 int security_file_permission(struct file *file, int mask);
 int security_file_alloc(struct file *file);
 void security_file_free(struct file *file);
@@ -1622,6 +1637,7 @@ int security_task_prctl(int option, unsigned long arg2, unsigned long arg3,
 void security_task_reparent_to_init(struct task_struct *p);
 void security_task_to_inode(struct task_struct *p, struct inode *inode);
 int security_ipc_permission(struct kern_ipc_perm *ipcp, short flag);
+void security_ipc_getsecid(struct kern_ipc_perm *ipcp, u32 *secid);
 int security_msg_msg_alloc(struct msg_msg *msg);
 void security_msg_msg_free(struct msg_msg *msg);
 int security_msg_queue_alloc(struct msg_queue *msq);
@@ -2022,6 +2038,11 @@ static inline int security_inode_listsecurity(struct inode *inode, char *buffer,
 	return 0;
 }
 
+static inline void security_inode_getsecid(const struct inode *inode, u32 *secid)
+{
+	*secid = 0;
+}
+
 static inline int security_file_permission (struct file *file, int mask)
 {
 	return 0;
@@ -2137,7 +2158,9 @@ static inline int security_task_getsid (struct task_struct *p)
 }
 
 static inline void security_task_getsecid (struct task_struct *p, u32 *secid)
-{ }
+{
+	*secid = 0;
+}
 
 static inline int security_task_setgroups (struct group_info *group_info)
 {
@@ -2216,6 +2239,11 @@ static inline int security_ipc_permission (struct kern_ipc_perm *ipcp,
 	return 0;
 }
 
+static inline void security_ipc_getsecid(struct kern_ipc_perm *ipcp, u32 *secid)
+{
+	*secid = 0;
+}
+
 static inline int security_msg_msg_alloc (struct msg_msg * msg)
 {
 	return 0;
diff --git a/security/dummy.c b/security/dummy.c
index 78d8f92310a4..fb2e942efbb6 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -424,6 +424,11 @@ static int dummy_inode_listsecurity(struct inode *inode, char *buffer, size_t bu
 	return 0;
 }
 
+static void dummy_inode_getsecid(const struct inode *inode, u32 *secid)
+{
+	*secid = 0;
+}
+
 static int dummy_file_permission (struct file *file, int mask)
 {
 	return 0;
@@ -542,7 +547,9 @@ static int dummy_task_getsid (struct task_struct *p)
 }
 
 static void dummy_task_getsecid (struct task_struct *p, u32 *secid)
-{ }
+{
+	*secid = 0;
+}
 
 static int dummy_task_setgroups (struct group_info *group_info)
 {
@@ -616,6 +623,11 @@ static int dummy_ipc_permission (struct kern_ipc_perm *ipcp, short flag)
 	return 0;
 }
 
+static void dummy_ipc_getsecid(struct kern_ipc_perm *ipcp, u32 *secid)
+{
+	*secid = 0;
+}
+
 static int dummy_msg_msg_alloc_security (struct msg_msg *msg)
 {
 	return 0;
@@ -1058,6 +1070,7 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, inode_getsecurity);
 	set_to_dummy_if_null(ops, inode_setsecurity);
 	set_to_dummy_if_null(ops, inode_listsecurity);
+	set_to_dummy_if_null(ops, inode_getsecid);
 	set_to_dummy_if_null(ops, file_permission);
 	set_to_dummy_if_null(ops, file_alloc_security);
 	set_to_dummy_if_null(ops, file_free_security);
@@ -1094,6 +1107,7 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, task_reparent_to_init);
  	set_to_dummy_if_null(ops, task_to_inode);
 	set_to_dummy_if_null(ops, ipc_permission);
+	set_to_dummy_if_null(ops, ipc_getsecid);
 	set_to_dummy_if_null(ops, msg_msg_alloc_security);
 	set_to_dummy_if_null(ops, msg_msg_free_security);
 	set_to_dummy_if_null(ops, msg_queue_alloc_security);
diff --git a/security/security.c b/security/security.c
index 9beecac933b4..290482bdbbb0 100644
--- a/security/security.c
+++ b/security/security.c
@@ -523,6 +523,11 @@ int security_inode_listsecurity(struct inode *inode, char *buffer, size_t buffer
 	return security_ops->inode_listsecurity(inode, buffer, buffer_size);
 }
 
+void security_inode_getsecid(const struct inode *inode, u32 *secid)
+{
+	security_ops->inode_getsecid(inode, secid);
+}
+
 int security_file_permission(struct file *file, int mask)
 {
 	return security_ops->file_permission(file, mask);
@@ -712,6 +717,11 @@ int security_ipc_permission(struct kern_ipc_perm *ipcp, short flag)
 	return security_ops->ipc_permission(ipcp, flag);
 }
 
+void security_ipc_getsecid(struct kern_ipc_perm *ipcp, u32 *secid)
+{
+	security_ops->ipc_getsecid(ipcp, secid);
+}
+
 int security_msg_msg_alloc(struct msg_msg *msg)
 {
 	return security_ops->msg_msg_alloc_security(msg);
-- 
cgit v1.2.3


From 6b89a74be0fbbc6cc639d5cf7dcf8e6ee0f120a7 Mon Sep 17 00:00:00 2001
From: "Ahmed S. Darwish" <darwish.07@gmail.com>
Date: Sat, 1 Mar 2008 21:58:32 +0200
Subject: SELinux: remove redundant exports

Remove the following exported SELinux interfaces:
selinux_get_inode_sid(inode, sid)
selinux_get_ipc_sid(ipcp, sid)
selinux_get_task_sid(tsk, sid)
selinux_sid_to_string(sid, ctx, len)

They can be substitued with the following generic equivalents
respectively:
new LSM hook, inode_getsecid(inode, secid)
new LSM hook, ipc_getsecid*(ipcp, secid)
LSM hook, task_getsecid(tsk, secid)
LSM hook, sid_to_secctx(sid, ctx, len)

Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
---
 include/linux/selinux.h    | 62 ----------------------------------------------
 security/selinux/exports.c | 42 -------------------------------
 2 files changed, 104 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/selinux.h b/include/linux/selinux.h
index 8c2cc4c02526..24b0af1c4cac 100644
--- a/include/linux/selinux.h
+++ b/include/linux/selinux.h
@@ -16,7 +16,6 @@
 
 struct selinux_audit_rule;
 struct audit_context;
-struct inode;
 struct kern_ipc_perm;
 
 #ifdef CONFIG_SECURITY_SELINUX
@@ -69,45 +68,6 @@ int selinux_audit_rule_match(u32 sid, u32 field, u32 op,
  */
 void selinux_audit_set_callback(int (*callback)(void));
 
-/**
- *     selinux_sid_to_string - map a security context ID to a string
- *     @sid: security context ID to be converted.
- *     @ctx: address of context string to be returned
- *     @ctxlen: length of returned context string.
- *
- *     Returns 0 if successful, -errno if not.  On success, the context
- *     string will be allocated internally, and the caller must call
- *     kfree() on it after use.
- */
-int selinux_sid_to_string(u32 sid, char **ctx, u32 *ctxlen);
-
-/**
- *     selinux_get_inode_sid - get the inode's security context ID
- *     @inode: inode structure to get the sid from.
- *     @sid: pointer to security context ID to be filled in.
- *
- *     Returns nothing
- */
-void selinux_get_inode_sid(const struct inode *inode, u32 *sid);
-
-/**
- *     selinux_get_ipc_sid - get the ipc security context ID
- *     @ipcp: ipc structure to get the sid from.
- *     @sid: pointer to security context ID to be filled in.
- *
- *     Returns nothing
- */
-void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid);
-
-/**
- *     selinux_get_task_sid - return the SID of task
- *     @tsk: the task whose SID will be returned
- *     @sid: pointer to security context ID to be filled in.
- *
- *     Returns nothing
- */
-void selinux_get_task_sid(struct task_struct *tsk, u32 *sid);
-
 /**
  *     selinux_string_to_sid - map a security context string to a security ID
  *     @str: the security context string to be mapped
@@ -175,28 +135,6 @@ static inline void selinux_audit_set_callback(int (*callback)(void))
 	return;
 }
 
-static inline int selinux_sid_to_string(u32 sid, char **ctx, u32 *ctxlen)
-{
-       *ctx = NULL;
-       *ctxlen = 0;
-       return 0;
-}
-
-static inline void selinux_get_inode_sid(const struct inode *inode, u32 *sid)
-{
-	*sid = 0;
-}
-
-static inline void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid)
-{
-	*sid = 0;
-}
-
-static inline void selinux_get_task_sid(struct task_struct *tsk, u32 *sid)
-{
-	*sid = 0;
-}
-
 static inline int selinux_string_to_sid(const char *str, u32 *sid)
 {
        *sid = 0;
diff --git a/security/selinux/exports.c b/security/selinux/exports.c
index 87d2bb3ea355..64af2d3409ef 100644
--- a/security/selinux/exports.c
+++ b/security/selinux/exports.c
@@ -25,48 +25,6 @@
 /* SECMARK reference count */
 extern atomic_t selinux_secmark_refcount;
 
-int selinux_sid_to_string(u32 sid, char **ctx, u32 *ctxlen)
-{
-	if (selinux_enabled)
-		return security_sid_to_context(sid, ctx, ctxlen);
-	else {
-		*ctx = NULL;
-		*ctxlen = 0;
-	}
-
-	return 0;
-}
-
-void selinux_get_inode_sid(const struct inode *inode, u32 *sid)
-{
-	if (selinux_enabled) {
-		struct inode_security_struct *isec = inode->i_security;
-		*sid = isec->sid;
-		return;
-	}
-	*sid = 0;
-}
-
-void selinux_get_ipc_sid(const struct kern_ipc_perm *ipcp, u32 *sid)
-{
-	if (selinux_enabled) {
-		struct ipc_security_struct *isec = ipcp->security;
-		*sid = isec->sid;
-		return;
-	}
-	*sid = 0;
-}
-
-void selinux_get_task_sid(struct task_struct *tsk, u32 *sid)
-{
-	if (selinux_enabled) {
-		struct task_security_struct *tsec = tsk->security;
-		*sid = tsec->sid;
-		return;
-	}
-	*sid = 0;
-}
-
 int selinux_string_to_sid(char *str, u32 *sid)
 {
 	if (selinux_enabled)
-- 
cgit v1.2.3


From 03d37d25e0f91b28c4b6d002be6221f1af4b19d8 Mon Sep 17 00:00:00 2001
From: "Ahmed S. Darwish" <darwish.07@gmail.com>
Date: Sat, 1 Mar 2008 22:00:05 +0200
Subject: LSM/Audit: Introduce generic Audit LSM hooks

Introduce a generic Audit interface for security modules
by adding the following new LSM hooks:

audit_rule_init(field, op, rulestr, lsmrule)
audit_rule_known(krule)
audit_rule_match(secid, field, op, rule, actx)
audit_rule_free(rule)

Those hooks are only available if CONFIG_AUDIT is enabled.

Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
---
 include/linux/security.h | 72 ++++++++++++++++++++++++++++++++++++++++++++++++
 security/dummy.c         | 31 ++++++++++++++++++++-
 security/security.c      | 25 +++++++++++++++++
 3 files changed, 127 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/security.h b/include/linux/security.h
index 45717d9d9656..697f228daf19 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -37,6 +37,7 @@
 extern unsigned securebits;
 
 struct ctl_table;
+struct audit_krule;
 
 /*
  * These functions are in security/capability.c and are used
@@ -1235,6 +1236,37 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
  *	@secdata contains the security context.
  *	@seclen contains the length of the security context.
  *
+ * Security hooks for Audit
+ *
+ * @audit_rule_init:
+ *	Allocate and initialize an LSM audit rule structure.
+ *	@field contains the required Audit action. Fields flags are defined in include/linux/audit.h
+ *	@op contains the operator the rule uses.
+ *	@rulestr contains the context where the rule will be applied to.
+ *	@lsmrule contains a pointer to receive the result.
+ *	Return 0 if @lsmrule has been successfully set,
+ *	-EINVAL in case of an invalid rule.
+ *
+ * @audit_rule_known:
+ *	Specifies whether given @rule contains any fields related to current LSM.
+ *	@rule contains the audit rule of interest.
+ *	Return 1 in case of relation found, 0 otherwise.
+ *
+ * @audit_rule_match:
+ *	Determine if given @secid matches a rule previously approved
+ *	by @audit_rule_known.
+ *	@secid contains the security id in question.
+ *	@field contains the field which relates to current LSM.
+ *	@op contains the operator that will be used for matching.
+ *	@rule points to the audit rule that will be checked against.
+ *	@actx points to the audit context associated with the check.
+ *	Return 1 if secid matches the rule, 0 if it does not, -ERRNO on failure.
+ *
+ * @audit_rule_free:
+ *	Deallocate the LSM audit rule structure previously allocated by
+ *	audit_rule_init.
+ *	@rule contains the allocated rule
+ *
  * This is the main security structure.
  */
 struct security_operations {
@@ -1494,6 +1526,13 @@ struct security_operations {
 
 #endif	/* CONFIG_KEYS */
 
+#ifdef CONFIG_AUDIT
+	int (*audit_rule_init)(u32 field, u32 op, char *rulestr, void **lsmrule);
+	int (*audit_rule_known)(struct audit_krule *krule);
+	int (*audit_rule_match)(u32 secid, u32 field, u32 op, void *lsmrule,
+				struct audit_context *actx);
+	void (*audit_rule_free)(void *lsmrule);
+#endif /* CONFIG_AUDIT */
 };
 
 /* prototypes */
@@ -2700,5 +2739,38 @@ static inline int security_key_permission(key_ref_t key_ref,
 #endif
 #endif /* CONFIG_KEYS */
 
+#ifdef CONFIG_AUDIT
+#ifdef CONFIG_SECURITY
+int security_audit_rule_init(u32 field, u32 op, char *rulestr, void **lsmrule);
+int security_audit_rule_known(struct audit_krule *krule);
+int security_audit_rule_match(u32 secid, u32 field, u32 op, void *lsmrule,
+			      struct audit_context *actx);
+void security_audit_rule_free(void *lsmrule);
+
+#else
+
+static inline int security_audit_rule_init(u32 field, u32 op, char *rulestr,
+					   void **lsmrule)
+{
+	return 0;
+}
+
+static inline int security_audit_rule_known(struct audit_krule *krule)
+{
+	return 0;
+}
+
+static inline int security_audit_rule_match(u32 secid, u32 field, u32 op,
+				   void *lsmrule, struct audit_context *actx)
+{
+	return 0;
+}
+
+static inline void security_audit_rule_free(void *lsmrule)
+{ }
+
+#endif /* CONFIG_SECURITY */
+#endif /* CONFIG_AUDIT */
+
 #endif /* ! __LINUX_SECURITY_H */
 
diff --git a/security/dummy.c b/security/dummy.c
index fb2e942efbb6..1ac9f8e66aa2 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -993,6 +993,30 @@ static inline int dummy_key_permission(key_ref_t key_ref,
 }
 #endif /* CONFIG_KEYS */
 
+#ifdef CONFIG_AUDIT
+static inline int dummy_audit_rule_init(u32 field, u32 op, char *rulestr,
+					void **lsmrule)
+{
+	return 0;
+}
+
+static inline int dummy_audit_rule_known(struct audit_krule *krule)
+{
+	return 0;
+}
+
+static inline int dummy_audit_rule_match(u32 secid, u32 field, u32 op,
+					 void *lsmrule,
+					 struct audit_context *actx)
+{
+	return 0;
+}
+
+static inline void dummy_audit_rule_free(void *lsmrule)
+{ }
+
+#endif /* CONFIG_AUDIT */
+
 struct security_operations dummy_security_ops;
 
 #define set_to_dummy_if_null(ops, function)				\
@@ -1182,6 +1206,11 @@ void security_fixup_ops (struct security_operations *ops)
 	set_to_dummy_if_null(ops, key_free);
 	set_to_dummy_if_null(ops, key_permission);
 #endif	/* CONFIG_KEYS */
-
+#ifdef CONFIG_AUDIT
+	set_to_dummy_if_null(ops, audit_rule_init);
+	set_to_dummy_if_null(ops, audit_rule_known);
+	set_to_dummy_if_null(ops, audit_rule_match);
+	set_to_dummy_if_null(ops, audit_rule_free);
+#endif
 }
 
diff --git a/security/security.c b/security/security.c
index 290482bdbbb0..2ef593ec70f3 100644
--- a/security/security.c
+++ b/security/security.c
@@ -1120,3 +1120,28 @@ int security_key_permission(key_ref_t key_ref,
 }
 
 #endif	/* CONFIG_KEYS */
+
+#ifdef CONFIG_AUDIT
+
+int security_audit_rule_init(u32 field, u32 op, char *rulestr, void **lsmrule)
+{
+	return security_ops->audit_rule_init(field, op, rulestr, lsmrule);
+}
+
+int security_audit_rule_known(struct audit_krule *krule)
+{
+	return security_ops->audit_rule_known(krule);
+}
+
+void security_audit_rule_free(void *lsmrule)
+{
+	security_ops->audit_rule_free(lsmrule);
+}
+
+int security_audit_rule_match(u32 secid, u32 field, u32 op, void *lsmrule,
+			      struct audit_context *actx)
+{
+	return security_ops->audit_rule_match(secid, field, op, lsmrule, actx);
+}
+
+#endif /* CONFIG_AUDIT */
-- 
cgit v1.2.3


From 9d57a7f9e23dc30783d245280fc9907cf2c87837 Mon Sep 17 00:00:00 2001
From: "Ahmed S. Darwish" <darwish.07@gmail.com>
Date: Sat, 1 Mar 2008 22:03:14 +0200
Subject: SELinux: use new audit hooks, remove redundant exports

Setup the new Audit LSM hooks for SELinux.
Remove the now redundant exported SELinux Audit interface.

Audit: Export 'audit_krule' and 'audit_field' to the public
since their internals are needed by the implementation of the
new LSM hook 'audit_rule_known'.

Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
---
 include/linux/audit.h          | 29 +++++++++++++++++
 include/linux/selinux.h        | 72 ------------------------------------------
 kernel/audit.h                 | 25 ---------------
 security/selinux/hooks.c       |  8 +++++
 security/selinux/ss/services.c | 45 +++++++++++++++++++-------
 5 files changed, 71 insertions(+), 108 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/audit.h b/include/linux/audit.h
index 2af9ec025015..04869c96016b 100644
--- a/include/linux/audit.h
+++ b/include/linux/audit.h
@@ -353,6 +353,33 @@ struct netlink_skb_parms;
 struct linux_binprm;
 struct mq_attr;
 struct mqstat;
+struct audit_watch;
+struct audit_tree;
+
+struct audit_krule {
+	int			vers_ops;
+	u32			flags;
+	u32			listnr;
+	u32			action;
+	u32			mask[AUDIT_BITMASK_SIZE];
+	u32			buflen; /* for data alloc on list rules */
+	u32			field_count;
+	char			*filterkey; /* ties events to rules */
+	struct audit_field	*fields;
+	struct audit_field	*arch_f; /* quick access to arch field */
+	struct audit_field	*inode_f; /* quick access to an inode field */
+	struct audit_watch	*watch;	/* associated watch */
+	struct audit_tree	*tree;	/* associated watched tree */
+	struct list_head	rlist;	/* entry in audit_{watch,tree}.rules list */
+};
+
+struct audit_field {
+	u32				type;
+	u32				val;
+	u32				op;
+	char				*se_str;
+	void				*se_rule;
+};
 
 #define AUDITSC_INVALID 0
 #define AUDITSC_SUCCESS 1
@@ -536,6 +563,8 @@ extern void		    audit_log_d_path(struct audit_buffer *ab,
 					     const char *prefix,
 					     struct path *path);
 extern void		    audit_log_lost(const char *message);
+extern int		    audit_update_lsm_rules(void);
+
 				/* Private API (for audit.c only) */
 extern int audit_filter_user(struct netlink_skb_parms *cb, int type);
 extern int audit_filter_type(int type);
diff --git a/include/linux/selinux.h b/include/linux/selinux.h
index 24b0af1c4cac..20f965d4b041 100644
--- a/include/linux/selinux.h
+++ b/include/linux/selinux.h
@@ -20,54 +20,6 @@ struct kern_ipc_perm;
 
 #ifdef CONFIG_SECURITY_SELINUX
 
-/**
- *	selinux_audit_rule_init - alloc/init an selinux audit rule structure.
- *	@field: the field this rule refers to
- *	@op: the operater the rule uses
- *	@rulestr: the text "target" of the rule
- *	@rule: pointer to the new rule structure returned via this
- *
- *	Returns 0 if successful, -errno if not.  On success, the rule structure
- *	will be allocated internally.  The caller must free this structure with
- *	selinux_audit_rule_free() after use.
- */
-int selinux_audit_rule_init(u32 field, u32 op, char *rulestr,
-                            struct selinux_audit_rule **rule);
-
-/**
- *	selinux_audit_rule_free - free an selinux audit rule structure.
- *	@rule: pointer to the audit rule to be freed
- *
- *	This will free all memory associated with the given rule.
- *	If @rule is NULL, no operation is performed.
- */
-void selinux_audit_rule_free(struct selinux_audit_rule *rule);
-
-/**
- *	selinux_audit_rule_match - determine if a context ID matches a rule.
- *	@sid: the context ID to check
- *	@field: the field this rule refers to
- *	@op: the operater the rule uses
- *	@rule: pointer to the audit rule to check against
- *	@actx: the audit context (can be NULL) associated with the check
- *
- *	Returns 1 if the context id matches the rule, 0 if it does not, and
- *	-errno on failure.
- */
-int selinux_audit_rule_match(u32 sid, u32 field, u32 op,
-                             struct selinux_audit_rule *rule,
-                             struct audit_context *actx);
-
-/**
- *	selinux_audit_set_callback - set the callback for policy reloads.
- *	@callback: the function to call when the policy is reloaded
- *
- *	This sets the function callback function that will update the rules
- *	upon policy reloads.  This callback should rebuild all existing rules
- *	using selinux_audit_rule_init().
- */
-void selinux_audit_set_callback(int (*callback)(void));
-
 /**
  *     selinux_string_to_sid - map a security context string to a security ID
  *     @str: the security context string to be mapped
@@ -111,30 +63,6 @@ void selinux_secmark_refcount_inc(void);
 void selinux_secmark_refcount_dec(void);
 #else
 
-static inline int selinux_audit_rule_init(u32 field, u32 op,
-                                          char *rulestr,
-                                          struct selinux_audit_rule **rule)
-{
-	return -EOPNOTSUPP;
-}
-
-static inline void selinux_audit_rule_free(struct selinux_audit_rule *rule)
-{
-	return;
-}
-
-static inline int selinux_audit_rule_match(u32 sid, u32 field, u32 op,
-                                           struct selinux_audit_rule *rule,
-                                           struct audit_context *actx)
-{
-	return 0;
-}
-
-static inline void selinux_audit_set_callback(int (*callback)(void))
-{
-	return;
-}
-
 static inline int selinux_string_to_sid(const char *str, u32 *sid)
 {
        *sid = 0;
diff --git a/kernel/audit.h b/kernel/audit.h
index 2554bd524fd1..3cfc54ee3e1f 100644
--- a/kernel/audit.h
+++ b/kernel/audit.h
@@ -65,34 +65,9 @@ struct audit_watch {
 	struct list_head	rules;	/* associated rules */
 };
 
-struct audit_field {
-	u32				type;
-	u32				val;
-	u32				op;
-	char				*se_str;
-	struct selinux_audit_rule	*se_rule;
-};
-
 struct audit_tree;
 struct audit_chunk;
 
-struct audit_krule {
-	int			vers_ops;
-	u32			flags;
-	u32			listnr;
-	u32			action;
-	u32			mask[AUDIT_BITMASK_SIZE];
-	u32			buflen; /* for data alloc on list rules */
-	u32			field_count;
-	char			*filterkey; /* ties events to rules */
-	struct audit_field	*fields;
-	struct audit_field	*arch_f; /* quick access to arch field */
-	struct audit_field	*inode_f; /* quick access to an inode field */
-	struct audit_watch	*watch;	/* associated watch */
-	struct audit_tree	*tree;	/* associated watched tree */
-	struct list_head	rlist;	/* entry in audit_{watch,tree}.rules list */
-};
-
 struct audit_entry {
 	struct list_head	list;
 	struct rcu_head		rcu;
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index bfffaa52e0cb..a2f7e9cf78c5 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -83,6 +83,7 @@
 #include "netport.h"
 #include "xfrm.h"
 #include "netlabel.h"
+#include "audit.h"
 
 #define XATTR_SELINUX_SUFFIX "selinux"
 #define XATTR_NAME_SELINUX XATTR_SECURITY_PREFIX XATTR_SELINUX_SUFFIX
@@ -5478,6 +5479,13 @@ static struct security_operations selinux_ops = {
 	.key_free =                     selinux_key_free,
 	.key_permission =               selinux_key_permission,
 #endif
+
+#ifdef CONFIG_AUDIT
+	.audit_rule_init =		selinux_audit_rule_init,
+	.audit_rule_known =		selinux_audit_rule_known,
+	.audit_rule_match =		selinux_audit_rule_match,
+	.audit_rule_free =		selinux_audit_rule_free,
+#endif
 };
 
 static __init int selinux_init(void)
diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c
index d75050819b06..1e0df5ec1bcd 100644
--- a/security/selinux/ss/services.c
+++ b/security/selinux/ss/services.c
@@ -57,6 +57,7 @@
 #include "netlabel.h"
 #include "xfrm.h"
 #include "ebitmap.h"
+#include "audit.h"
 
 extern void selnl_notify_policyload(u32 seqno);
 unsigned int policydb_loaded_version;
@@ -2296,21 +2297,23 @@ struct selinux_audit_rule {
 	struct context au_ctxt;
 };
 
-void selinux_audit_rule_free(struct selinux_audit_rule *rule)
+void selinux_audit_rule_free(void *vrule)
 {
+	struct selinux_audit_rule *rule = vrule;
+
 	if (rule) {
 		context_destroy(&rule->au_ctxt);
 		kfree(rule);
 	}
 }
 
-int selinux_audit_rule_init(u32 field, u32 op, char *rulestr,
-                            struct selinux_audit_rule **rule)
+int selinux_audit_rule_init(u32 field, u32 op, char *rulestr, void **vrule)
 {
 	struct selinux_audit_rule *tmprule;
 	struct role_datum *roledatum;
 	struct type_datum *typedatum;
 	struct user_datum *userdatum;
+	struct selinux_audit_rule **rule = (struct selinux_audit_rule **)vrule;
 	int rc = 0;
 
 	*rule = NULL;
@@ -2397,12 +2400,37 @@ int selinux_audit_rule_init(u32 field, u32 op, char *rulestr,
 	return rc;
 }
 
-int selinux_audit_rule_match(u32 sid, u32 field, u32 op,
-                             struct selinux_audit_rule *rule,
+/* Check to see if the rule contains any selinux fields */
+int selinux_audit_rule_known(struct audit_krule *rule)
+{
+	int i;
+
+	for (i = 0; i < rule->field_count; i++) {
+		struct audit_field *f = &rule->fields[i];
+		switch (f->type) {
+		case AUDIT_SUBJ_USER:
+		case AUDIT_SUBJ_ROLE:
+		case AUDIT_SUBJ_TYPE:
+		case AUDIT_SUBJ_SEN:
+		case AUDIT_SUBJ_CLR:
+		case AUDIT_OBJ_USER:
+		case AUDIT_OBJ_ROLE:
+		case AUDIT_OBJ_TYPE:
+		case AUDIT_OBJ_LEV_LOW:
+		case AUDIT_OBJ_LEV_HIGH:
+			return 1;
+		}
+	}
+
+	return 0;
+}
+
+int selinux_audit_rule_match(u32 sid, u32 field, u32 op, void *vrule,
                              struct audit_context *actx)
 {
 	struct context *ctxt;
 	struct mls_level *level;
+	struct selinux_audit_rule *rule = vrule;
 	int match = 0;
 
 	if (!rule) {
@@ -2509,7 +2537,7 @@ out:
 	return match;
 }
 
-static int (*aurule_callback)(void) = NULL;
+static int (*aurule_callback)(void) = audit_update_lsm_rules;
 
 static int aurule_avc_callback(u32 event, u32 ssid, u32 tsid,
                                u16 class, u32 perms, u32 *retained)
@@ -2534,11 +2562,6 @@ static int __init aurule_init(void)
 }
 __initcall(aurule_init);
 
-void selinux_audit_set_callback(int (*callback)(void))
-{
-	aurule_callback = callback;
-}
-
 #ifdef CONFIG_NETLABEL
 /**
  * security_netlbl_cache_add - Add an entry to the NetLabel cache
-- 
cgit v1.2.3


From 04305e4aff8b0533dc05f9f6f1a34d0796bd985f Mon Sep 17 00:00:00 2001
From: "Ahmed S. Darwish" <darwish.07@gmail.com>
Date: Sat, 19 Apr 2008 09:59:43 +1000
Subject: Audit: Final renamings and cleanup

Rename the se_str and se_rule audit fields elements to
lsm_str and lsm_rule to avoid confusion.

Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
---
 include/linux/audit.h            |  4 +--
 kernel/auditfilter.c             | 40 ++++++++++++-------------
 kernel/auditsc.c                 | 12 ++++----
 security/selinux/include/audit.h | 65 ++++++++++++++++++++++++++++++++++++++++
 4 files changed, 93 insertions(+), 28 deletions(-)
 create mode 100644 security/selinux/include/audit.h

(limited to 'include/linux')

diff --git a/include/linux/audit.h b/include/linux/audit.h
index 04869c96016b..4ccb048cae1d 100644
--- a/include/linux/audit.h
+++ b/include/linux/audit.h
@@ -377,8 +377,8 @@ struct audit_field {
 	u32				type;
 	u32				val;
 	u32				op;
-	char				*se_str;
-	void				*se_rule;
+	char				*lsm_str;
+	void				*lsm_rule;
 };
 
 #define AUDITSC_INVALID 0
diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c
index 7c69cb5e44fb..28fef6bf8534 100644
--- a/kernel/auditfilter.c
+++ b/kernel/auditfilter.c
@@ -139,8 +139,8 @@ static inline void audit_free_rule(struct audit_entry *e)
 	if (e->rule.fields)
 		for (i = 0; i < e->rule.field_count; i++) {
 			struct audit_field *f = &e->rule.fields[i];
-			kfree(f->se_str);
-			security_audit_rule_free(f->se_rule);
+			kfree(f->lsm_str);
+			security_audit_rule_free(f->lsm_rule);
 		}
 	kfree(e->rule.fields);
 	kfree(e->rule.filterkey);
@@ -554,8 +554,8 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data,
 		f->op = data->fieldflags[i] & AUDIT_OPERATORS;
 		f->type = data->fields[i];
 		f->val = data->values[i];
-		f->se_str = NULL;
-		f->se_rule = NULL;
+		f->lsm_str = NULL;
+		f->lsm_rule = NULL;
 		switch(f->type) {
 		case AUDIT_PID:
 		case AUDIT_UID:
@@ -598,7 +598,7 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data,
 			entry->rule.buflen += f->val;
 
 			err = security_audit_rule_init(f->type, f->op, str,
-						       (void **)&f->se_rule);
+						       (void **)&f->lsm_rule);
 			/* Keep currently invalid fields around in case they
 			 * become valid after a policy reload. */
 			if (err == -EINVAL) {
@@ -610,7 +610,7 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data,
 				kfree(str);
 				goto exit_free;
 			} else
-				f->se_str = str;
+				f->lsm_str = str;
 			break;
 		case AUDIT_WATCH:
 			str = audit_unpack_string(&bufp, &remain, f->val);
@@ -754,7 +754,7 @@ static struct audit_rule_data *audit_krule_to_data(struct audit_krule *krule)
 		case AUDIT_OBJ_LEV_LOW:
 		case AUDIT_OBJ_LEV_HIGH:
 			data->buflen += data->values[i] =
-				audit_pack_string(&bufp, f->se_str);
+				audit_pack_string(&bufp, f->lsm_str);
 			break;
 		case AUDIT_WATCH:
 			data->buflen += data->values[i] =
@@ -806,7 +806,7 @@ static int audit_compare_rule(struct audit_krule *a, struct audit_krule *b)
 		case AUDIT_OBJ_TYPE:
 		case AUDIT_OBJ_LEV_LOW:
 		case AUDIT_OBJ_LEV_HIGH:
-			if (strcmp(a->fields[i].se_str, b->fields[i].se_str))
+			if (strcmp(a->fields[i].lsm_str, b->fields[i].lsm_str))
 				return 1;
 			break;
 		case AUDIT_WATCH:
@@ -862,28 +862,28 @@ out:
 	return new;
 }
 
-/* Duplicate LSM field information.  The se_rule is opaque, so must be
+/* Duplicate LSM field information.  The lsm_rule is opaque, so must be
  * re-initialized. */
 static inline int audit_dupe_lsm_field(struct audit_field *df,
 					   struct audit_field *sf)
 {
 	int ret = 0;
-	char *se_str;
+	char *lsm_str;
 
-	/* our own copy of se_str */
-	se_str = kstrdup(sf->se_str, GFP_KERNEL);
-	if (unlikely(!se_str))
+	/* our own copy of lsm_str */
+	lsm_str = kstrdup(sf->lsm_str, GFP_KERNEL);
+	if (unlikely(!lsm_str))
 		return -ENOMEM;
-	df->se_str = se_str;
+	df->lsm_str = lsm_str;
 
-	/* our own (refreshed) copy of se_rule */
-	ret = security_audit_rule_init(df->type, df->op, df->se_str,
-				       (void **)&df->se_rule);
+	/* our own (refreshed) copy of lsm_rule */
+	ret = security_audit_rule_init(df->type, df->op, df->lsm_str,
+				       (void **)&df->lsm_rule);
 	/* Keep currently invalid fields around in case they
 	 * become valid after a policy reload. */
 	if (ret == -EINVAL) {
 		printk(KERN_WARNING "audit rule for LSM \'%s\' is "
-		       "invalid\n", df->se_str);
+		       "invalid\n", df->lsm_str);
 		ret = 0;
 	}
 
@@ -930,7 +930,7 @@ static struct audit_entry *audit_dupe_rule(struct audit_krule *old,
 	new->tree = old->tree;
 	memcpy(new->fields, old->fields, sizeof(struct audit_field) * fcount);
 
-	/* deep copy this information, updating the se_rule fields, because
+	/* deep copy this information, updating the lsm_rule fields, because
 	 * the originals will all be freed when the old rule is freed. */
 	for (i = 0; i < fcount; i++) {
 		switch (new->fields[i].type) {
@@ -1762,7 +1762,7 @@ unlock_and_return:
 	return result;
 }
 
-/* This function will re-initialize the se_rule field of all applicable rules.
+/* This function will re-initialize the lsm_rule field of all applicable rules.
  * It will traverse the filter lists serarching for rules that contain LSM
  * specific filter fields.  When such a rule is found, it is copied, the
  * LSM field is re-initialized, and the old rule is replaced with the
diff --git a/kernel/auditsc.c b/kernel/auditsc.c
index c0700535e5c5..56e56ed594a8 100644
--- a/kernel/auditsc.c
+++ b/kernel/auditsc.c
@@ -527,14 +527,14 @@ static int audit_filter_rules(struct task_struct *tsk,
 			   match for now to avoid losing information that
 			   may be wanted.   An error message will also be
 			   logged upon error */
-			if (f->se_rule) {
+			if (f->lsm_rule) {
 				if (need_sid) {
 					security_task_getsecid(tsk, &sid);
 					need_sid = 0;
 				}
 				result = security_audit_rule_match(sid, f->type,
 				                                  f->op,
-				                                  f->se_rule,
+				                                  f->lsm_rule,
 				                                  ctx);
 			}
 			break;
@@ -545,18 +545,18 @@ static int audit_filter_rules(struct task_struct *tsk,
 		case AUDIT_OBJ_LEV_HIGH:
 			/* The above note for AUDIT_SUBJ_USER...AUDIT_SUBJ_CLR
 			   also applies here */
-			if (f->se_rule) {
+			if (f->lsm_rule) {
 				/* Find files that match */
 				if (name) {
 					result = security_audit_rule_match(
 					           name->osid, f->type, f->op,
-					           f->se_rule, ctx);
+					           f->lsm_rule, ctx);
 				} else if (ctx) {
 					for (j = 0; j < ctx->name_count; j++) {
 						if (security_audit_rule_match(
 						      ctx->names[j].osid,
 						      f->type, f->op,
-						      f->se_rule, ctx)) {
+						      f->lsm_rule, ctx)) {
 							++result;
 							break;
 						}
@@ -569,7 +569,7 @@ static int audit_filter_rules(struct task_struct *tsk,
 					     aux = aux->next) {
 						if (aux->type == AUDIT_IPC) {
 							struct audit_aux_data_ipcctl *axi = (void *)aux;
-							if (security_audit_rule_match(axi->osid, f->type, f->op, f->se_rule, ctx)) {
+							if (security_audit_rule_match(axi->osid, f->type, f->op, f->lsm_rule, ctx)) {
 								++result;
 								break;
 							}
diff --git a/security/selinux/include/audit.h b/security/selinux/include/audit.h
new file mode 100644
index 000000000000..6c8b9ef15579
--- /dev/null
+++ b/security/selinux/include/audit.h
@@ -0,0 +1,65 @@
+/*
+ * SELinux support for the Audit LSM hooks
+ *
+ * Most of below header was moved from include/linux/selinux.h which 
+ * is released under below copyrights:
+ *
+ * Author: James Morris <jmorris@redhat.com>
+ *
+ * Copyright (C) 2005 Red Hat, Inc., James Morris <jmorris@redhat.com>
+ * Copyright (C) 2006 Trusted Computer Solutions, Inc. <dgoeddel@trustedcs.com>
+ * Copyright (C) 2006 IBM Corporation, Timothy R. Chavez <tinytim@us.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2,
+ * as published by the Free Software Foundation.
+ */
+
+#ifndef _SELINUX_AUDIT_H
+#define _SELINUX_AUDIT_H
+
+/**
+ *	selinux_audit_rule_init - alloc/init an selinux audit rule structure.
+ *	@field: the field this rule refers to
+ *	@op: the operater the rule uses
+ *	@rulestr: the text "target" of the rule
+ *	@rule: pointer to the new rule structure returned via this
+ *
+ *	Returns 0 if successful, -errno if not.  On success, the rule structure
+ *	will be allocated internally.  The caller must free this structure with
+ *	selinux_audit_rule_free() after use.
+ */
+int selinux_audit_rule_init(u32 field, u32 op, char *rulestr, void **rule);
+
+/**
+ *	selinux_audit_rule_free - free an selinux audit rule structure.
+ *	@rule: pointer to the audit rule to be freed
+ *
+ *	This will free all memory associated with the given rule.
+ *	If @rule is NULL, no operation is performed.
+ */
+void selinux_audit_rule_free(void *rule);
+
+/**
+ *	selinux_audit_rule_match - determine if a context ID matches a rule.
+ *	@sid: the context ID to check
+ *	@field: the field this rule refers to
+ *	@op: the operater the rule uses
+ *	@rule: pointer to the audit rule to check against
+ *	@actx: the audit context (can be NULL) associated with the check
+ *
+ *	Returns 1 if the context id matches the rule, 0 if it does not, and
+ *	-errno on failure.
+ */
+int selinux_audit_rule_match(u32 sid, u32 field, u32 op, void *rule,
+                             struct audit_context *actx);
+
+/**
+ *	selinux_audit_rule_known - check to see if rule contains selinux fields.
+ *	@rule: rule to be checked
+ *	Returns 1 if there are selinux fields specified in the rule, 0 otherwise.
+ */
+int selinux_audit_rule_known(struct audit_krule *krule);
+
+#endif /* _SELINUX_AUDIT_H */
+
-- 
cgit v1.2.3


From 076c54c5bcaed2081c0cba94a6f77c4d470236ad Mon Sep 17 00:00:00 2001
From: "Ahmed S. Darwish" <darwish.07@gmail.com>
Date: Thu, 6 Mar 2008 18:09:10 +0200
Subject: Security: Introduce security= boot parameter

Add the security= boot parameter. This is done to avoid LSM
registration clashes in case of more than one bult-in module.

User can choose a security module to enable at boot. If no
security= boot parameter is specified, only the first LSM
asking for registration will be loaded. An invalid security
module name will be treated as if no module has been chosen.

LSM modules must check now if they are allowed to register
by calling security_module_enable(ops) first. Modify SELinux
and SMACK to do so.

Do not let SMACK register smackfs if it was not chosen on
boot. Smackfs assumes that smack hooks are registered and
the initial task security setup (swapper->security) is done.

Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
---
 Documentation/kernel-parameters.txt |  6 ++++++
 include/linux/security.h            | 12 ++++++++++++
 security/dummy.c                    |  4 +++-
 security/security.c                 | 38 ++++++++++++++++++++++++++++++++++++-
 security/selinux/hooks.c            |  7 +++++++
 security/smack/smack.h              |  2 ++
 security/smack/smack_lsm.c          |  7 ++++++-
 security/smack/smackfs.c            | 11 ++++++++++-
 8 files changed, 83 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 256a2162503c..4b0f1ae31a4c 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -366,6 +366,12 @@ and is between 256 and 4096 characters. It is defined in the file
 			possible to determine what the correct size should be.
 			This option provides an override for these situations.
 
+	security=	[SECURITY] Choose a security module to enable at boot.
+			If this boot parameter is not specified, only the first
+			security module asking for security registration will be
+			loaded. An invalid security module name will be treated
+			as if no module has been chosen.
+
 	capability.disable=
 			[SECURITY] Disable capabilities.  This would normally
 			be used only if an alternative security model is to be
diff --git a/include/linux/security.h b/include/linux/security.h
index 697f228daf19..f4116d6ed64b 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -36,6 +36,9 @@
 
 extern unsigned securebits;
 
+/* Maximum number of letters for an LSM name string */
+#define SECURITY_NAME_MAX	10
+
 struct ctl_table;
 struct audit_krule;
 
@@ -137,6 +140,12 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
 /**
  * struct security_operations - main security structure
  *
+ * Security module identifier.
+ *
+ * @name:
+ *	A string that acts as a unique identifeir for the LSM with max number
+ *	of characters = SECURITY_NAME_MAX.
+ *
  * Security hooks for program execution operations.
  *
  * @bprm_alloc_security:
@@ -1270,6 +1279,8 @@ static inline void security_free_mnt_opts(struct security_mnt_opts *opts)
  * This is the main security structure.
  */
 struct security_operations {
+	char name[SECURITY_NAME_MAX + 1];
+
 	int (*ptrace) (struct task_struct * parent, struct task_struct * child);
 	int (*capget) (struct task_struct * target,
 		       kernel_cap_t * effective,
@@ -1537,6 +1548,7 @@ struct security_operations {
 
 /* prototypes */
 extern int security_init	(void);
+extern int security_module_enable(struct security_operations *ops);
 extern int register_security	(struct security_operations *ops);
 extern int mod_reg_security	(const char *name, struct security_operations *ops);
 extern struct dentry *securityfs_create_file(const char *name, mode_t mode,
diff --git a/security/dummy.c b/security/dummy.c
index 1ac9f8e66aa2..d797a4196b89 100644
--- a/security/dummy.c
+++ b/security/dummy.c
@@ -1017,7 +1017,9 @@ static inline void dummy_audit_rule_free(void *lsmrule)
 
 #endif /* CONFIG_AUDIT */
 
-struct security_operations dummy_security_ops;
+struct security_operations dummy_security_ops = {
+	.name = "dummy",
+};
 
 #define set_to_dummy_if_null(ops, function)				\
 	do {								\
diff --git a/security/security.c b/security/security.c
index 2ef593ec70f3..dd0c6baed494 100644
--- a/security/security.c
+++ b/security/security.c
@@ -17,6 +17,8 @@
 #include <linux/kernel.h>
 #include <linux/security.h>
 
+/* Boot-time LSM user choice */
+static __initdata char chosen_lsm[SECURITY_NAME_MAX + 1];
 
 /* things that live in dummy.c */
 extern struct security_operations dummy_security_ops;
@@ -67,13 +69,47 @@ int __init security_init(void)
 	return 0;
 }
 
+/* Save user chosen LSM */
+static int __init choose_lsm(char *str)
+{
+	strncpy(chosen_lsm, str, SECURITY_NAME_MAX);
+	return 1;
+}
+__setup("security=", choose_lsm);
+
+/**
+ * security_module_enable - Load given security module on boot ?
+ * @ops: a pointer to the struct security_operations that is to be checked.
+ *
+ * Each LSM must pass this method before registering its own operations
+ * to avoid security registration races. This method may also be used
+ * to check if your LSM is currently loaded.
+ *
+ * Return true if:
+ *	-The passed LSM is the one chosen by user at boot time,
+ *	-or user didsn't specify a specific LSM and we're the first to ask
+ *	 for registeration permissoin,
+ *	-or the passed LSM is currently loaded.
+ * Otherwise, return false.
+ */
+int __init security_module_enable(struct security_operations *ops)
+{
+	if (!*chosen_lsm)
+		strncpy(chosen_lsm, ops->name, SECURITY_NAME_MAX);
+	else if (strncmp(ops->name, chosen_lsm, SECURITY_NAME_MAX))
+		return 0;
+
+	return 1;
+}
+
 /**
  * register_security - registers a security framework with the kernel
  * @ops: a pointer to the struct security_options that is to be registered
  *
  * This function is to allow a security module to register itself with the
  * kernel security subsystem.  Some rudimentary checking is done on the @ops
- * value passed to this function.
+ * value passed to this function. You'll need to check first if your LSM
+ * is allowed to register its @ops by calling security_module_enable(@ops).
  *
  * If there is already a security module registered with the kernel,
  * an error will be returned.  Otherwise 0 is returned on success.
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index a2f7e9cf78c5..f9927f02bc3d 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -5295,6 +5295,8 @@ static int selinux_key_permission(key_ref_t key_ref,
 #endif
 
 static struct security_operations selinux_ops = {
+	.name =				"selinux",
+
 	.ptrace =			selinux_ptrace,
 	.capget =			selinux_capget,
 	.capset_check =			selinux_capset_check,
@@ -5492,6 +5494,11 @@ static __init int selinux_init(void)
 {
 	struct task_security_struct *tsec;
 
+	if (!security_module_enable(&selinux_ops)) {
+		selinux_enabled = 0;
+		return 0;
+	}
+
 	if (!selinux_enabled) {
 		printk(KERN_INFO "SELinux:  Disabled at boot.\n");
 		return 0;
diff --git a/security/smack/smack.h b/security/smack/smack.h
index 62c1e982849d..4a4477f5afdc 100644
--- a/security/smack/smack.h
+++ b/security/smack/smack.h
@@ -15,6 +15,7 @@
 
 #include <linux/capability.h>
 #include <linux/spinlock.h>
+#include <linux/security.h>
 #include <net/netlabel.h>
 
 /*
@@ -187,6 +188,7 @@ extern struct smack_known smack_known_star;
 extern struct smack_known smack_known_unset;
 
 extern struct smk_list_entry *smack_list;
+extern struct security_operations smack_ops;
 
 /*
  * Stricly for CIPSO level manipulation.
diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
index 732ba27923c4..904bdc01a12b 100644
--- a/security/smack/smack_lsm.c
+++ b/security/smack/smack_lsm.c
@@ -2424,7 +2424,9 @@ static void smack_release_secctx(char *secdata, u32 seclen)
 {
 }
 
-static struct security_operations smack_ops = {
+struct security_operations smack_ops = {
+	.name =				"smack",
+
 	.ptrace = 			smack_ptrace,
 	.capget = 			cap_capget,
 	.capset_check = 		cap_capset_check,
@@ -2557,6 +2559,9 @@ static struct security_operations smack_ops = {
  */
 static __init int smack_init(void)
 {
+	if (!security_module_enable(&smack_ops))
+		return 0;
+
 	printk(KERN_INFO "Smack:  Initializing.\n");
 
 	/*
diff --git a/security/smack/smackfs.c b/security/smack/smackfs.c
index cfae8afcc262..6ba283783b70 100644
--- a/security/smack/smackfs.c
+++ b/security/smack/smackfs.c
@@ -965,12 +965,21 @@ static struct vfsmount *smackfs_mount;
  *
  * register the smackfs
  *
- * Returns 0 unless the registration fails.
+ * Do not register smackfs if Smack wasn't enabled
+ * on boot. We can not put this method normally under the
+ * smack_init() code path since the security subsystem get
+ * initialized before the vfs caches.
+ *
+ * Returns true if we were not chosen on boot or if
+ * we were chosen and filesystem registration succeeded.
  */
 static int __init init_smk_fs(void)
 {
 	int err;
 
+	if (!security_module_enable(&smack_ops))
+		return 0;
+
 	err = register_filesystem(&smk_fs_type);
 	if (!err) {
 		smackfs_mount = kern_mount(&smk_fs_type);
-- 
cgit v1.2.3


From 5a6483feb0c5193519625d0ea8c4254364d423cc Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Tue, 26 Feb 2008 10:00:17 -0500
Subject: include: Remove unnecessary inclusions of asm/semaphore.h

None of these files use any of the functionality promised by
asm/semaphore.h.  It's possible that they (or some user of them) rely
on it dragging in some unrelated header file, but I can't build all
these files, so we'll have to fix any build failures as they come up.

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
---
 include/asm-blackfin/dma.h    | 1 -
 include/asm-ia64/sn/nodepda.h | 1 -
 include/asm-ppc/ocp.h         | 1 -
 include/linux/cpu.h           | 1 -
 include/linux/ide.h           | 1 -
 include/linux/if_pppox.h      | 1 -
 include/linux/jbd.h           | 2 --
 include/linux/jbd2.h          | 2 --
 include/linux/kernelcapi.h    | 1 -
 include/linux/raid/md.h       | 1 -
 include/linux/sched.h         | 1 -
 include/linux/syscalls.h      | 1 -
 include/scsi/libsas.h         | 1 -
 13 files changed, 15 deletions(-)

(limited to 'include/linux')

diff --git a/include/asm-blackfin/dma.h b/include/asm-blackfin/dma.h
index 5abaa2cee8db..16d493574ba8 100644
--- a/include/asm-blackfin/dma.h
+++ b/include/asm-blackfin/dma.h
@@ -33,7 +33,6 @@
 #include <linux/slab.h>
 #include <asm/irq.h>
 #include <asm/signal.h>
-#include <asm/semaphore.h>
 
 #include <linux/kernel.h>
 #include <asm/mach/dma.h>
diff --git a/include/asm-ia64/sn/nodepda.h b/include/asm-ia64/sn/nodepda.h
index 6f6d69e39ff5..ee118b901de4 100644
--- a/include/asm-ia64/sn/nodepda.h
+++ b/include/asm-ia64/sn/nodepda.h
@@ -9,7 +9,6 @@
 #define _ASM_IA64_SN_NODEPDA_H
 
 
-#include <asm/semaphore.h>
 #include <asm/irq.h>
 #include <asm/sn/arch.h>
 #include <asm/sn/intr.h>
diff --git a/include/asm-ppc/ocp.h b/include/asm-ppc/ocp.h
index 1379a4f76de3..3909a2eec286 100644
--- a/include/asm-ppc/ocp.h
+++ b/include/asm-ppc/ocp.h
@@ -31,7 +31,6 @@
 
 #include <asm/mmu.h>
 #include <asm/ocp_ids.h>
-#include <asm/semaphore.h>
 
 #ifdef CONFIG_PPC_OCP
 
diff --git a/include/linux/cpu.h b/include/linux/cpu.h
index 0be8d65bc3c8..f212fa98283e 100644
--- a/include/linux/cpu.h
+++ b/include/linux/cpu.h
@@ -23,7 +23,6 @@
 #include <linux/node.h>
 #include <linux/compiler.h>
 #include <linux/cpumask.h>
-#include <asm/semaphore.h>
 #include <linux/mutex.h>
 
 struct cpu {
diff --git a/include/linux/ide.h b/include/linux/ide.h
index 6c39482fd1a1..5f3e82ae901a 100644
--- a/include/linux/ide.h
+++ b/include/linux/ide.h
@@ -23,7 +23,6 @@
 #include <asm/byteorder.h>
 #include <asm/system.h>
 #include <asm/io.h>
-#include <asm/semaphore.h>
 #include <asm/mutex.h>
 
 #if defined(CONFIG_CRIS) || defined(CONFIG_FRV)
diff --git a/include/linux/if_pppox.h b/include/linux/if_pppox.h
index 40743e032845..6fb7f1788570 100644
--- a/include/linux/if_pppox.h
+++ b/include/linux/if_pppox.h
@@ -24,7 +24,6 @@
 #include <linux/if_ether.h>
 #include <linux/if.h>
 #include <linux/netdevice.h>
-#include <asm/semaphore.h>
 #include <linux/ppp_channel.h>
 #endif /* __KERNEL__ */
 #include <linux/if_pppol2tp.h>
diff --git a/include/linux/jbd.h b/include/linux/jbd.h
index 423f58272188..07a9b52a2654 100644
--- a/include/linux/jbd.h
+++ b/include/linux/jbd.h
@@ -32,8 +32,6 @@
 #include <linux/timer.h>
 #include <linux/lockdep.h>
 
-#include <asm/semaphore.h>
-
 #define journal_oom_retry 1
 
 /*
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 2cbf6fdb1799..05e2b307161a 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -30,8 +30,6 @@
 #include <linux/bit_spinlock.h>
 #include <linux/mutex.h>
 #include <linux/timer.h>
-
-#include <asm/semaphore.h>
 #endif
 
 #define journal_oom_retry 1
diff --git a/include/linux/kernelcapi.h b/include/linux/kernelcapi.h
index 8c4350a9ed87..a53e932f80fb 100644
--- a/include/linux/kernelcapi.h
+++ b/include/linux/kernelcapi.h
@@ -48,7 +48,6 @@ typedef struct kcapi_carddef {
 #include <linux/list.h>
 #include <linux/skbuff.h>
 #include <linux/workqueue.h>
-#include <asm/semaphore.h>
 
 #define	KCI_CONTRUP	0	/* arg: struct capi_profile */
 #define	KCI_CONTRDOWN	1	/* arg: NULL */
diff --git a/include/linux/raid/md.h b/include/linux/raid/md.h
index fbaeda79b2e9..8ab630b67fcc 100644
--- a/include/linux/raid/md.h
+++ b/include/linux/raid/md.h
@@ -19,7 +19,6 @@
 #define _MD_H
 
 #include <linux/blkdev.h>
-#include <asm/semaphore.h>
 #include <linux/major.h>
 #include <linux/ioctl.h>
 #include <linux/types.h>
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 6a1e7afb099b..a37b5964828a 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -61,7 +61,6 @@ struct sched_param {
 #include <linux/mm_types.h>
 
 #include <asm/system.h>
-#include <asm/semaphore.h>
 #include <asm/page.h>
 #include <asm/ptrace.h>
 #include <asm/cputime.h>
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index 4c2577bd1c85..8df6d1382ac8 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -60,7 +60,6 @@ struct getcpu_cache;
 #include <linux/capability.h>
 #include <linux/list.h>
 #include <linux/sem.h>
-#include <asm/semaphore.h>
 #include <asm/siginfo.h>
 #include <asm/signal.h>
 #include <linux/quota.h>
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
index 98724ba65a79..e78d3b62d8ec 100644
--- a/include/scsi/libsas.h
+++ b/include/scsi/libsas.h
@@ -32,7 +32,6 @@
 #include <scsi/sas.h>
 #include <linux/libata.h>
 #include <linux/list.h>
-#include <asm/semaphore.h>
 #include <scsi/scsi_device.h>
 #include <scsi/scsi_cmnd.h>
 #include <scsi/scsi_transport_sas.h>
-- 
cgit v1.2.3


From 6188e10d38b8d7244ee7776d5f1f88c837b4b93f Mon Sep 17 00:00:00 2001
From: Matthew Wilcox <matthew@wil.cx>
Date: Fri, 18 Apr 2008 22:21:05 -0400
Subject: Convert asm/semaphore.h users to linux/semaphore.h

Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
---
 Documentation/DocBook/kernel-locking.tmpl | 6 +++---
 arch/ia64/kernel/salinfo.c                | 2 +-
 drivers/base/core.c                       | 2 +-
 drivers/char/snsc.h                       | 2 +-
 drivers/firewire/fw-device.c              | 3 ++-
 drivers/i2c/i2c-core.c                    | 2 +-
 drivers/ieee1394/nodemgr.c                | 2 +-
 drivers/infiniband/core/user_mad.c        | 2 +-
 drivers/infiniband/hw/mthca/mthca_dev.h   | 3 +--
 drivers/input/serio/hp_sdc_mlc.c          | 2 +-
 drivers/macintosh/adb.c                   | 2 +-
 drivers/macintosh/windfarm_smu_sat.c      | 2 +-
 drivers/net/3c527.c                       | 2 +-
 drivers/net/hamradio/6pack.c              | 2 +-
 drivers/s390/cio/qdio.c                   | 2 +-
 drivers/scsi/aacraid/commctrl.c           | 2 +-
 drivers/scsi/aacraid/commsup.c            | 2 +-
 drivers/scsi/aacraid/dpcsup.c             | 2 +-
 drivers/scsi/megaraid/megaraid_ioctl.h    | 2 +-
 drivers/scsi/qla2xxx/qla_def.h            | 2 +-
 drivers/watchdog/sc1200wdt.c              | 2 +-
 fs/jffs2/jffs2_fs_i.h                     | 2 +-
 fs/jffs2/jffs2_fs_sb.h                    | 2 +-
 fs/reiserfs/journal.c                     | 2 +-
 fs/xfs/linux-2.6/sema.h                   | 2 +-
 include/linux/device.h                    | 2 +-
 include/linux/fs.h                        | 2 +-
 include/linux/hil_mlc.h                   | 2 +-
 include/linux/i2o.h                       | 2 +-
 include/linux/memory.h                    | 3 +--
 include/linux/parport.h                   | 2 +-
 lib/kernel_lock.c                         | 2 +-
 32 files changed, 35 insertions(+), 36 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/DocBook/kernel-locking.tmpl b/Documentation/DocBook/kernel-locking.tmpl
index 2e9d6b41f034..435413ca40dc 100644
--- a/Documentation/DocBook/kernel-locking.tmpl
+++ b/Documentation/DocBook/kernel-locking.tmpl
@@ -241,7 +241,7 @@
    </para>
    <para>
      The third type is a semaphore
-     (<filename class="headerfile">include/asm/semaphore.h</filename>): it
+     (<filename class="headerfile">include/linux/semaphore.h</filename>): it
      can have more than one holder at any time (the number decided at
      initialization time), although it is most commonly used as a
      single-holder lock (a mutex).  If you can't get a semaphore, your
@@ -290,7 +290,7 @@
      <para>
        If you have a data structure which is only ever accessed from
        user context, then you can use a simple semaphore
-       (<filename>linux/asm/semaphore.h</filename>) to protect it.  This 
+       (<filename>linux/linux/semaphore.h</filename>) to protect it.  This
        is the most trivial case: you initialize the semaphore to the number 
        of resources available (usually 1), and call
        <function>down_interruptible()</function> to grab the semaphore, and 
@@ -1656,7 +1656,7 @@ the amount of locking which needs to be done.
  #include &lt;linux/slab.h&gt;
  #include &lt;linux/string.h&gt;
 +#include &lt;linux/rcupdate.h&gt;
- #include &lt;asm/semaphore.h&gt;
+ #include &lt;linux/semaphore.h&gt;
  #include &lt;asm/errno.h&gt;
 
  struct object
diff --git a/arch/ia64/kernel/salinfo.c b/arch/ia64/kernel/salinfo.c
index 779c3cca206c..b11bb50a197a 100644
--- a/arch/ia64/kernel/salinfo.c
+++ b/arch/ia64/kernel/salinfo.c
@@ -44,8 +44,8 @@
 #include <linux/smp.h>
 #include <linux/timer.h>
 #include <linux/vmalloc.h>
+#include <linux/semaphore.h>
 
-#include <asm/semaphore.h>
 #include <asm/sal.h>
 #include <asm/uaccess.h>
 
diff --git a/drivers/base/core.c b/drivers/base/core.c
index 24198ad01976..7c4b36ccb1a0 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -20,7 +20,7 @@
 #include <linux/notifier.h>
 #include <linux/genhd.h>
 #include <linux/kallsyms.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 #include "base.h"
 #include "power/power.h"
diff --git a/drivers/char/snsc.h b/drivers/char/snsc.h
index 8a98169b60c1..4be62eda9fbc 100644
--- a/drivers/char/snsc.h
+++ b/drivers/char/snsc.h
@@ -22,8 +22,8 @@
 #include <linux/kobject.h>
 #include <linux/fs.h>
 #include <linux/cdev.h>
+#include <linux/semaphore.h>
 #include <asm/sn/types.h>
-#include <asm/semaphore.h>
 
 #define CHUNKSIZE 127
 
diff --git a/drivers/firewire/fw-device.c b/drivers/firewire/fw-device.c
index 2d01bc1b9752..d9c8daf7ae7d 100644
--- a/drivers/firewire/fw-device.c
+++ b/drivers/firewire/fw-device.c
@@ -26,7 +26,8 @@
 #include <linux/delay.h>
 #include <linux/idr.h>
 #include <linux/string.h>
-#include <asm/semaphore.h>
+#include <linux/rwsem.h>
+#include <linux/semaphore.h>
 #include <asm/system.h>
 #include <linux/ctype.h>
 #include "fw-transaction.h"
diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
index 8b645c6b2cb5..e186df657119 100644
--- a/drivers/i2c/i2c-core.c
+++ b/drivers/i2c/i2c-core.c
@@ -35,8 +35,8 @@
 #include <linux/completion.h>
 #include <linux/hardirq.h>
 #include <linux/irqflags.h>
+#include <linux/semaphore.h>
 #include <asm/uaccess.h>
-#include <asm/semaphore.h>
 
 #include "i2c-core.h"
 
diff --git a/drivers/ieee1394/nodemgr.c b/drivers/ieee1394/nodemgr.c
index 70afa3786f3f..29d833e71cbf 100644
--- a/drivers/ieee1394/nodemgr.c
+++ b/drivers/ieee1394/nodemgr.c
@@ -18,8 +18,8 @@
 #include <linux/moduleparam.h>
 #include <linux/mutex.h>
 #include <linux/freezer.h>
+#include <linux/semaphore.h>
 #include <asm/atomic.h>
-#include <asm/semaphore.h>
 
 #include "csr.h"
 #include "highlevel.h"
diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c
index 4e915104ac4c..be953e87bf93 100644
--- a/drivers/infiniband/core/user_mad.c
+++ b/drivers/infiniband/core/user_mad.c
@@ -46,9 +46,9 @@
 #include <linux/mutex.h>
 #include <linux/kref.h>
 #include <linux/compat.h>
+#include <linux/semaphore.h>
 
 #include <asm/uaccess.h>
-#include <asm/semaphore.h>
 
 #include <rdma/ib_mad.h>
 #include <rdma/ib_user_mad.h>
diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h
index 0e842e023400..7bc32f8e377e 100644
--- a/drivers/infiniband/hw/mthca/mthca_dev.h
+++ b/drivers/infiniband/hw/mthca/mthca_dev.h
@@ -46,8 +46,7 @@
 #include <linux/timer.h>
 #include <linux/mutex.h>
 #include <linux/list.h>
-
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 #include "mthca_provider.h"
 #include "mthca_doorbell.h"
diff --git a/drivers/input/serio/hp_sdc_mlc.c b/drivers/input/serio/hp_sdc_mlc.c
index c45ea74d53e4..f1fd3b638a37 100644
--- a/drivers/input/serio/hp_sdc_mlc.c
+++ b/drivers/input/serio/hp_sdc_mlc.c
@@ -40,7 +40,7 @@
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/string.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 #define PREFIX "HP SDC MLC: "
 
diff --git a/drivers/macintosh/adb.c b/drivers/macintosh/adb.c
index 28958101061f..20978205cd02 100644
--- a/drivers/macintosh/adb.c
+++ b/drivers/macintosh/adb.c
@@ -37,9 +37,9 @@
 #include <linux/device.h>
 #include <linux/kthread.h>
 #include <linux/platform_device.h>
+#include <linux/semaphore.h>
 
 #include <asm/uaccess.h>
-#include <asm/semaphore.h>
 #ifdef CONFIG_PPC
 #include <asm/prom.h>
 #include <asm/machdep.h>
diff --git a/drivers/macintosh/windfarm_smu_sat.c b/drivers/macintosh/windfarm_smu_sat.c
index f449d775cdf4..797918d0e59c 100644
--- a/drivers/macintosh/windfarm_smu_sat.c
+++ b/drivers/macintosh/windfarm_smu_sat.c
@@ -13,7 +13,7 @@
 #include <linux/init.h>
 #include <linux/wait.h>
 #include <linux/i2c.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 #include <asm/prom.h>
 #include <asm/smu.h>
 #include <asm/pmac_low_i2c.h>
diff --git a/drivers/net/3c527.c b/drivers/net/3c527.c
index b72b89d53ec8..fae295b6809c 100644
--- a/drivers/net/3c527.c
+++ b/drivers/net/3c527.c
@@ -103,8 +103,8 @@ DRV_NAME ".c:v" DRV_VERSION " " DRV_RELDATE " Richard Procter <rnp@paradise.net.
 #include <linux/ethtool.h>
 #include <linux/completion.h>
 #include <linux/bitops.h>
+#include <linux/semaphore.h>
 
-#include <asm/semaphore.h>
 #include <asm/uaccess.h>
 #include <asm/system.h>
 #include <asm/io.h>
diff --git a/drivers/net/hamradio/6pack.c b/drivers/net/hamradio/6pack.c
index 0a9b75139e0f..1da55dd2a5a0 100644
--- a/drivers/net/hamradio/6pack.c
+++ b/drivers/net/hamradio/6pack.c
@@ -33,7 +33,7 @@
 #include <linux/init.h>
 #include <linux/ip.h>
 #include <linux/tcp.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 #include <asm/atomic.h>
 
 #define SIXPACK_VERSION    "Revision: 0.3.0"
diff --git a/drivers/s390/cio/qdio.c b/drivers/s390/cio/qdio.c
index c359386708e9..10aa1e780801 100644
--- a/drivers/s390/cio/qdio.c
+++ b/drivers/s390/cio/qdio.c
@@ -38,11 +38,11 @@
 #include <linux/proc_fs.h>
 #include <linux/timer.h>
 #include <linux/mempool.h>
+#include <linux/semaphore.h>
 
 #include <asm/ccwdev.h>
 #include <asm/io.h>
 #include <asm/atomic.h>
-#include <asm/semaphore.h>
 #include <asm/timex.h>
 
 #include <asm/debug.h>
diff --git a/drivers/scsi/aacraid/commctrl.c b/drivers/scsi/aacraid/commctrl.c
index abef05146d75..5fd83deab36c 100644
--- a/drivers/scsi/aacraid/commctrl.c
+++ b/drivers/scsi/aacraid/commctrl.c
@@ -39,7 +39,7 @@
 #include <linux/blkdev.h>
 #include <linux/delay.h> /* ssleep prototype */
 #include <linux/kthread.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 #include <asm/uaccess.h>
 
 #include "aacraid.h"
diff --git a/drivers/scsi/aacraid/commsup.c b/drivers/scsi/aacraid/commsup.c
index 23a8e9f8dcb4..ef67816a6fe5 100644
--- a/drivers/scsi/aacraid/commsup.c
+++ b/drivers/scsi/aacraid/commsup.c
@@ -41,11 +41,11 @@
 #include <linux/delay.h>
 #include <linux/kthread.h>
 #include <linux/interrupt.h>
+#include <linux/semaphore.h>
 #include <scsi/scsi.h>
 #include <scsi/scsi_host.h>
 #include <scsi/scsi_device.h>
 #include <scsi/scsi_cmnd.h>
-#include <asm/semaphore.h>
 
 #include "aacraid.h"
 
diff --git a/drivers/scsi/aacraid/dpcsup.c b/drivers/scsi/aacraid/dpcsup.c
index d1163ded132b..933f208eedba 100644
--- a/drivers/scsi/aacraid/dpcsup.c
+++ b/drivers/scsi/aacraid/dpcsup.c
@@ -36,7 +36,7 @@
 #include <linux/slab.h>
 #include <linux/completion.h>
 #include <linux/blkdev.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 #include "aacraid.h"
 
diff --git a/drivers/scsi/megaraid/megaraid_ioctl.h b/drivers/scsi/megaraid/megaraid_ioctl.h
index 706fa05a187a..05f6e4ec3453 100644
--- a/drivers/scsi/megaraid/megaraid_ioctl.h
+++ b/drivers/scsi/megaraid/megaraid_ioctl.h
@@ -18,7 +18,7 @@
 #define _MEGARAID_IOCTL_H_
 
 #include <linux/types.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 #include "mbox_defs.h"
 
diff --git a/drivers/scsi/qla2xxx/qla_def.h b/drivers/scsi/qla2xxx/qla_def.h
index 094d95f0764c..299eccf6cabd 100644
--- a/drivers/scsi/qla2xxx/qla_def.h
+++ b/drivers/scsi/qla2xxx/qla_def.h
@@ -25,7 +25,7 @@
 #include <linux/firmware.h>
 #include <linux/aer.h>
 #include <linux/mutex.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 #include <scsi/scsi.h>
 #include <scsi/scsi_host.h>
diff --git a/drivers/watchdog/sc1200wdt.c b/drivers/watchdog/sc1200wdt.c
index 32ccd7c89c7d..35cddff7020f 100644
--- a/drivers/watchdog/sc1200wdt.c
+++ b/drivers/watchdog/sc1200wdt.c
@@ -38,8 +38,8 @@
 #include <linux/init.h>
 #include <linux/pnp.h>
 #include <linux/fs.h>
+#include <linux/semaphore.h>
 
-#include <asm/semaphore.h>
 #include <asm/io.h>
 #include <asm/uaccess.h>
 
diff --git a/fs/jffs2/jffs2_fs_i.h b/fs/jffs2/jffs2_fs_i.h
index 0b78fdc9773b..a841f4973a74 100644
--- a/fs/jffs2/jffs2_fs_i.h
+++ b/fs/jffs2/jffs2_fs_i.h
@@ -15,7 +15,7 @@
 #include <linux/version.h>
 #include <linux/rbtree.h>
 #include <linux/posix_acl.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 struct jffs2_inode_info {
 	/* We need an internal mutex similar to inode->i_mutex.
diff --git a/fs/jffs2/jffs2_fs_sb.h b/fs/jffs2/jffs2_fs_sb.h
index 3a2197f3c812..18fca2b9e531 100644
--- a/fs/jffs2/jffs2_fs_sb.h
+++ b/fs/jffs2/jffs2_fs_sb.h
@@ -16,7 +16,7 @@
 #include <linux/spinlock.h>
 #include <linux/workqueue.h>
 #include <linux/completion.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 #include <linux/timer.h>
 #include <linux/wait.h>
 #include <linux/list.h>
diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c
index bb05a3e51b93..060eb3f598e7 100644
--- a/fs/reiserfs/journal.c
+++ b/fs/reiserfs/journal.c
@@ -38,7 +38,7 @@
 #include <asm/system.h>
 
 #include <linux/time.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 #include <linux/vmalloc.h>
 #include <linux/reiserfs_fs.h>
diff --git a/fs/xfs/linux-2.6/sema.h b/fs/xfs/linux-2.6/sema.h
index 2009e6d922ce..3abe7e9ceb33 100644
--- a/fs/xfs/linux-2.6/sema.h
+++ b/fs/xfs/linux-2.6/sema.h
@@ -20,8 +20,8 @@
 
 #include <linux/time.h>
 #include <linux/wait.h>
+#include <linux/semaphore.h>
 #include <asm/atomic.h>
-#include <asm/semaphore.h>
 
 /*
  * sema_t structure just maps to struct semaphore in Linux kernel.
diff --git a/include/linux/device.h b/include/linux/device.h
index 2258d89bf523..c79b93e56fa0 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -20,7 +20,7 @@
 #include <linux/types.h>
 #include <linux/module.h>
 #include <linux/pm.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 #include <asm/atomic.h>
 #include <asm/device.h>
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index b84b848431f2..91e8dec9e42b 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -287,9 +287,9 @@ extern int dir_notify_enable;
 #include <linux/pid.h>
 #include <linux/mutex.h>
 #include <linux/capability.h>
+#include <linux/semaphore.h>
 
 #include <asm/atomic.h>
-#include <asm/semaphore.h>
 #include <asm/byteorder.h>
 
 struct export_operations;
diff --git a/include/linux/hil_mlc.h b/include/linux/hil_mlc.h
index 8df29ca48a13..394a8405dd74 100644
--- a/include/linux/hil_mlc.h
+++ b/include/linux/hil_mlc.h
@@ -34,7 +34,7 @@
 #include <linux/hil.h>
 #include <linux/time.h>
 #include <linux/interrupt.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 #include <linux/serio.h>
 #include <linux/list.h>
 
diff --git a/include/linux/i2o.h b/include/linux/i2o.h
index 7da5b98d90e6..e92170dda245 100644
--- a/include/linux/i2o.h
+++ b/include/linux/i2o.h
@@ -33,9 +33,9 @@
 #include <linux/mempool.h>
 #include <linux/mutex.h>
 #include <linux/scatterlist.h>
+#include <linux/semaphore.h>	/* Needed for MUTEX init macros */
 
 #include <asm/io.h>
-#include <asm/semaphore.h>	/* Needed for MUTEX init macros */
 
 /* message queue empty */
 #define I2O_QUEUE_EMPTY		0xffffffff
diff --git a/include/linux/memory.h b/include/linux/memory.h
index 33f0ff0cf634..54d7866d9c0e 100644
--- a/include/linux/memory.h
+++ b/include/linux/memory.h
@@ -18,8 +18,7 @@
 #include <linux/sysdev.h>
 #include <linux/node.h>
 #include <linux/compiler.h>
-
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 struct memory_block {
 	unsigned long phys_index;
diff --git a/include/linux/parport.h b/include/linux/parport.h
index d1ad546c8c9e..dcb9e01a69ca 100644
--- a/include/linux/parport.h
+++ b/include/linux/parport.h
@@ -101,9 +101,9 @@ typedef enum {
 #include <linux/spinlock.h>
 #include <linux/wait.h>
 #include <linux/irqreturn.h>
+#include <linux/semaphore.h>
 #include <asm/system.h>
 #include <asm/ptrace.h>
-#include <asm/semaphore.h>
 
 /* Define this later. */
 struct parport;
diff --git a/lib/kernel_lock.c b/lib/kernel_lock.c
index fbc11a336bc5..cd3e82530b03 100644
--- a/lib/kernel_lock.c
+++ b/lib/kernel_lock.c
@@ -8,7 +8,7 @@
 #include <linux/smp_lock.h>
 #include <linux/module.h>
 #include <linux/kallsyms.h>
-#include <asm/semaphore.h>
+#include <linux/semaphore.h>
 
 /*
  * The 'big kernel semaphore'
-- 
cgit v1.2.3


From a70e65df8812c52252fa07a2eb92a46451a4427f Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Fri, 15 Feb 2008 14:37:28 -0800
Subject: [PATCH] merge open_namei() and do_filp_open()

open_namei() will, in the future, need to take mount write counts
over its creation and truncation (via may_open()) operations.  It
needs to keep these write counts until any potential filp that is
created gets __fput()'d.

This gets complicated in the error handling and becomes very murky
as to how far open_namei() actually got, and whether or not that
mount write count was taken.  That makes it a bad interface.

All that the current do_filp_open() really does is allocate the
nameidata on the stack, then call open_namei().

So, this merges those two functions and moves filp_open() over
to namei.c so it can be close to its buddy: do_filp_open().  It
also gets a kerneldoc comment in the process.

Acked-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c         | 100 ++++++++++++++++++++++++++++++-----------------------
 fs/open.c          |  19 ----------
 include/linux/fs.h |   3 +-
 3 files changed, 59 insertions(+), 63 deletions(-)

(limited to 'include/linux')

diff --git a/fs/namei.c b/fs/namei.c
index c70dbf720109..a1f8bbbd58e5 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1725,17 +1725,13 @@ static inline int open_to_namei_flags(int flag)
 }
 
 /*
- *	open_namei()
- *
- * namei for open - this is in fact almost the whole open-routine.
- *
  * Note that the low bits of "flag" aren't the same as in the open
  * system call.  See open_to_namei_flags().
- * SMP-safe
  */
-int open_namei(int dfd, const char *pathname, int open_flag,
-		int mode, struct nameidata *nd)
+struct file *do_filp_open(int dfd, const char *pathname,
+		int open_flag, int mode)
 {
+	struct nameidata nd;
 	int acc_mode, error;
 	struct path path;
 	struct dentry *dir;
@@ -1758,18 +1754,19 @@ int open_namei(int dfd, const char *pathname, int open_flag,
 	 */
 	if (!(flag & O_CREAT)) {
 		error = path_lookup_open(dfd, pathname, lookup_flags(flag),
-					 nd, flag);
+					 &nd, flag);
 		if (error)
-			return error;
+			return ERR_PTR(error);
 		goto ok;
 	}
 
 	/*
 	 * Create - we need to know the parent.
 	 */
-	error = path_lookup_create(dfd,pathname,LOOKUP_PARENT,nd,flag,mode);
+	error = path_lookup_create(dfd, pathname, LOOKUP_PARENT,
+				   &nd, flag, mode);
 	if (error)
-		return error;
+		return ERR_PTR(error);
 
 	/*
 	 * We have the parent and last component. First of all, check
@@ -1777,14 +1774,14 @@ int open_namei(int dfd, const char *pathname, int open_flag,
 	 * will not do.
 	 */
 	error = -EISDIR;
-	if (nd->last_type != LAST_NORM || nd->last.name[nd->last.len])
+	if (nd.last_type != LAST_NORM || nd.last.name[nd.last.len])
 		goto exit;
 
-	dir = nd->path.dentry;
-	nd->flags &= ~LOOKUP_PARENT;
+	dir = nd.path.dentry;
+	nd.flags &= ~LOOKUP_PARENT;
 	mutex_lock(&dir->d_inode->i_mutex);
-	path.dentry = lookup_hash(nd);
-	path.mnt = nd->path.mnt;
+	path.dentry = lookup_hash(&nd);
+	path.mnt = nd.path.mnt;
 
 do_last:
 	error = PTR_ERR(path.dentry);
@@ -1793,18 +1790,18 @@ do_last:
 		goto exit;
 	}
 
-	if (IS_ERR(nd->intent.open.file)) {
+	if (IS_ERR(nd.intent.open.file)) {
 		mutex_unlock(&dir->d_inode->i_mutex);
-		error = PTR_ERR(nd->intent.open.file);
+		error = PTR_ERR(nd.intent.open.file);
 		goto exit_dput;
 	}
 
 	/* Negative dentry, just create the file */
 	if (!path.dentry->d_inode) {
-		error = __open_namei_create(nd, &path, flag, mode);
+		error = __open_namei_create(&nd, &path, flag, mode);
 		if (error)
 			goto exit;
-		return 0;
+		return nameidata_to_filp(&nd, open_flag);
 	}
 
 	/*
@@ -1829,23 +1826,23 @@ do_last:
 	if (path.dentry->d_inode->i_op && path.dentry->d_inode->i_op->follow_link)
 		goto do_link;
 
-	path_to_nameidata(&path, nd);
+	path_to_nameidata(&path, &nd);
 	error = -EISDIR;
 	if (path.dentry->d_inode && S_ISDIR(path.dentry->d_inode->i_mode))
 		goto exit;
 ok:
-	error = may_open(nd, acc_mode, flag);
+	error = may_open(&nd, acc_mode, flag);
 	if (error)
 		goto exit;
-	return 0;
+	return nameidata_to_filp(&nd, open_flag);
 
 exit_dput:
-	path_put_conditional(&path, nd);
+	path_put_conditional(&path, &nd);
 exit:
-	if (!IS_ERR(nd->intent.open.file))
-		release_open_intent(nd);
-	path_put(&nd->path);
-	return error;
+	if (!IS_ERR(nd.intent.open.file))
+		release_open_intent(&nd);
+	path_put(&nd.path);
+	return ERR_PTR(error);
 
 do_link:
 	error = -ELOOP;
@@ -1861,42 +1858,59 @@ do_link:
 	 * stored in nd->last.name and we will have to putname() it when we
 	 * are done. Procfs-like symlinks just set LAST_BIND.
 	 */
-	nd->flags |= LOOKUP_PARENT;
-	error = security_inode_follow_link(path.dentry, nd);
+	nd.flags |= LOOKUP_PARENT;
+	error = security_inode_follow_link(path.dentry, &nd);
 	if (error)
 		goto exit_dput;
-	error = __do_follow_link(&path, nd);
+	error = __do_follow_link(&path, &nd);
 	if (error) {
 		/* Does someone understand code flow here? Or it is only
 		 * me so stupid? Anathema to whoever designed this non-sense
 		 * with "intent.open".
 		 */
-		release_open_intent(nd);
-		return error;
+		release_open_intent(&nd);
+		return ERR_PTR(error);
 	}
-	nd->flags &= ~LOOKUP_PARENT;
-	if (nd->last_type == LAST_BIND)
+	nd.flags &= ~LOOKUP_PARENT;
+	if (nd.last_type == LAST_BIND)
 		goto ok;
 	error = -EISDIR;
-	if (nd->last_type != LAST_NORM)
+	if (nd.last_type != LAST_NORM)
 		goto exit;
-	if (nd->last.name[nd->last.len]) {
-		__putname(nd->last.name);
+	if (nd.last.name[nd.last.len]) {
+		__putname(nd.last.name);
 		goto exit;
 	}
 	error = -ELOOP;
 	if (count++==32) {
-		__putname(nd->last.name);
+		__putname(nd.last.name);
 		goto exit;
 	}
-	dir = nd->path.dentry;
+	dir = nd.path.dentry;
 	mutex_lock(&dir->d_inode->i_mutex);
-	path.dentry = lookup_hash(nd);
-	path.mnt = nd->path.mnt;
-	__putname(nd->last.name);
+	path.dentry = lookup_hash(&nd);
+	path.mnt = nd.path.mnt;
+	__putname(nd.last.name);
 	goto do_last;
 }
 
+/**
+ * filp_open - open file and return file pointer
+ *
+ * @filename:	path to open
+ * @flags:	open flags as per the open(2) second argument
+ * @mode:	mode for the new file if O_CREAT is set, else ignored
+ *
+ * This is the helper to open a file from kernelspace if you really
+ * have to.  But in generally you should not do this, so please move
+ * along, nothing to see here..
+ */
+struct file *filp_open(const char *filename, int flags, int mode)
+{
+	return do_filp_open(AT_FDCWD, filename, flags, mode);
+}
+EXPORT_SYMBOL(filp_open);
+
 /**
  * lookup_create - lookup a dentry, creating it if it doesn't exist
  * @nd: nameidata info
diff --git a/fs/open.c b/fs/open.c
index 5ab3f3f079c0..8111947905d8 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -796,25 +796,6 @@ cleanup_file:
 	return ERR_PTR(error);
 }
 
-static struct file *do_filp_open(int dfd, const char *filename, int flags,
-				 int mode)
-{
-	int error;
-	struct nameidata nd;
-
-	error = open_namei(dfd, filename, flags, mode, &nd);
-	if (!error)
-		return nameidata_to_filp(&nd, flags);
-
-	return ERR_PTR(error);
-}
-
-struct file *filp_open(const char *filename, int flags, int mode)
-{
-	return do_filp_open(AT_FDCWD, filename, flags, mode);
-}
-EXPORT_SYMBOL(filp_open);
-
 /**
  * lookup_instantiate_filp - instantiates the open intent filp
  * @nd: pointer to nameidata
diff --git a/include/linux/fs.h b/include/linux/fs.h
index b84b848431f2..013b9c2b88e6 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1735,7 +1735,8 @@ extern struct file *create_read_pipe(struct file *f);
 extern struct file *create_write_pipe(void);
 extern void free_write_pipe(struct file *);
 
-extern int open_namei(int dfd, const char *, int, int, struct nameidata *);
+extern struct file *do_filp_open(int dfd, const char *pathname,
+		int open_flag, int mode);
 extern int may_open(struct nameidata *, int, int);
 
 extern int kernel_read(struct file *, unsigned long, char *, unsigned long);
-- 
cgit v1.2.3


From 8366025eb80dfa0d8d94b286d53027081c280ef1 Mon Sep 17 00:00:00 2001
From: Dave Hansen <haveblue@us.ibm.com>
Date: Fri, 15 Feb 2008 14:37:30 -0800
Subject: [PATCH] r/o bind mounts: stub functions

This patch adds two function mnt_want_write() and mnt_drop_write().  These are
used like a lock pair around and fs operations that might cause a write to the
filesystem.

Before these can become useful, we must first cover each place in the VFS
where writes are performed with a want/drop pair.  When that is complete, we
can actually introduce code that will safely check the counts before allowing
r/w<->r/o transitions to occur.

Acked-by: Serge Hallyn <serue@us.ibm.com>
Acked-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namespace.c        | 54 +++++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/mount.h |  3 +++
 2 files changed, 57 insertions(+)

(limited to 'include/linux')

diff --git a/fs/namespace.c b/fs/namespace.c
index 94f026ec990a..066b393578c1 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -80,6 +80,60 @@ struct vfsmount *alloc_vfsmnt(const char *name)
 	return mnt;
 }
 
+/*
+ * Most r/o checks on a fs are for operations that take
+ * discrete amounts of time, like a write() or unlink().
+ * We must keep track of when those operations start
+ * (for permission checks) and when they end, so that
+ * we can determine when writes are able to occur to
+ * a filesystem.
+ */
+/**
+ * mnt_want_write - get write access to a mount
+ * @mnt: the mount on which to take a write
+ *
+ * This tells the low-level filesystem that a write is
+ * about to be performed to it, and makes sure that
+ * writes are allowed before returning success.  When
+ * the write operation is finished, mnt_drop_write()
+ * must be called.  This is effectively a refcount.
+ */
+int mnt_want_write(struct vfsmount *mnt)
+{
+	if (__mnt_is_readonly(mnt))
+		return -EROFS;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(mnt_want_write);
+
+/**
+ * mnt_drop_write - give up write access to a mount
+ * @mnt: the mount on which to give up write access
+ *
+ * Tells the low-level filesystem that we are done
+ * performing writes to it.  Must be matched with
+ * mnt_want_write() call above.
+ */
+void mnt_drop_write(struct vfsmount *mnt)
+{
+}
+EXPORT_SYMBOL_GPL(mnt_drop_write);
+
+/*
+ * __mnt_is_readonly: check whether a mount is read-only
+ * @mnt: the mount to check for its write status
+ *
+ * This shouldn't be used directly ouside of the VFS.
+ * It does not guarantee that the filesystem will stay
+ * r/w, just that it is right *now*.  This can not and
+ * should not be used in place of IS_RDONLY(inode).
+ */
+int __mnt_is_readonly(struct vfsmount *mnt)
+{
+	return (mnt->mnt_sb->s_flags & MS_RDONLY);
+}
+EXPORT_SYMBOL_GPL(__mnt_is_readonly);
+
 int simple_set_mnt(struct vfsmount *mnt, struct super_block *sb)
 {
 	mnt->mnt_sb = sb;
diff --git a/include/linux/mount.h b/include/linux/mount.h
index 5ee2df217cdf..2eecd2c8c760 100644
--- a/include/linux/mount.h
+++ b/include/linux/mount.h
@@ -71,9 +71,12 @@ static inline struct vfsmount *mntget(struct vfsmount *mnt)
 	return mnt;
 }
 
+extern int mnt_want_write(struct vfsmount *mnt);
+extern void mnt_drop_write(struct vfsmount *mnt);
 extern void mntput_no_expire(struct vfsmount *mnt);
 extern void mnt_pin(struct vfsmount *mnt);
 extern void mnt_unpin(struct vfsmount *mnt);
+extern int __mnt_is_readonly(struct vfsmount *mnt);
 
 static inline void mntput(struct vfsmount *mnt)
 {
-- 
cgit v1.2.3


From aceaf78da92a53f5e1b105649a1b8c0afdb2135c Mon Sep 17 00:00:00 2001
From: Dave Hansen <haveblue@us.ibm.com>
Date: Fri, 15 Feb 2008 14:37:31 -0800
Subject: [PATCH] r/o bind mounts: create helper to drop file write access

If someone decides to demote a file from r/w to just
r/o, they can use this same code as __fput().

NFS does just that, and will use this in the next
patch.

AV: drop write access in __fput() only after we evict from file list.

Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Cc: Erez Zadok <ezk@cs.sunysb.edu>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: "J Bruce Fields" <bfields@fieldses.org>
Acked-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/file_table.c      | 21 +++++++++++++++++++--
 fs/nfsd/nfs4state.c  |  3 ++-
 include/linux/file.h |  1 +
 3 files changed, 22 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/fs/file_table.c b/fs/file_table.c
index 986ff4ed0a7c..3f73eb1f195a 100644
--- a/fs/file_table.c
+++ b/fs/file_table.c
@@ -211,6 +211,23 @@ void fput(struct file *file)
 
 EXPORT_SYMBOL(fput);
 
+/**
+ * drop_file_write_access - give up ability to write to a file
+ * @file: the file to which we will stop writing
+ *
+ * This is a central place which will give up the ability
+ * to write to @file, along with access to write through
+ * its vfsmount.
+ */
+void drop_file_write_access(struct file *file)
+{
+	struct dentry *dentry = file->f_path.dentry;
+	struct inode *inode = dentry->d_inode;
+
+	put_write_access(inode);
+}
+EXPORT_SYMBOL_GPL(drop_file_write_access);
+
 /* __fput is called from task context when aio completion releases the last
  * last use of a struct file *.  Do not use otherwise.
  */
@@ -236,10 +253,10 @@ void __fput(struct file *file)
 	if (unlikely(S_ISCHR(inode->i_mode) && inode->i_cdev != NULL))
 		cdev_put(inode->i_cdev);
 	fops_put(file->f_op);
-	if (file->f_mode & FMODE_WRITE)
-		put_write_access(inode);
 	put_pid(file->f_owner.pid);
 	file_kill(file);
+	if (file->f_mode & FMODE_WRITE)
+		drop_file_write_access(file);
 	file->f_path.dentry = NULL;
 	file->f_path.mnt = NULL;
 	file_free(file);
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index bcb97d8e8b8b..81a75f3081f4 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -41,6 +41,7 @@
 #include <linux/sunrpc/svc.h>
 #include <linux/nfsd/nfsd.h>
 #include <linux/nfsd/cache.h>
+#include <linux/file.h>
 #include <linux/mount.h>
 #include <linux/workqueue.h>
 #include <linux/smp_lock.h>
@@ -1239,7 +1240,7 @@ static inline void
 nfs4_file_downgrade(struct file *filp, unsigned int share_access)
 {
 	if (share_access & NFS4_SHARE_ACCESS_WRITE) {
-		put_write_access(filp->f_path.dentry->d_inode);
+		drop_file_write_access(filp);
 		filp->f_mode = (filp->f_mode | FMODE_READ) & ~FMODE_WRITE;
 	}
 }
diff --git a/include/linux/file.h b/include/linux/file.h
index 7239baac81a9..653477021e4c 100644
--- a/include/linux/file.h
+++ b/include/linux/file.h
@@ -61,6 +61,7 @@ extern struct kmem_cache *filp_cachep;
 
 extern void __fput(struct file *);
 extern void fput(struct file *);
+extern void drop_file_write_access(struct file *file);
 
 struct file_operations;
 struct vfsmount;
-- 
cgit v1.2.3


From 3d733633a633065729c9e4e254b2e5442c00ef7e Mon Sep 17 00:00:00 2001
From: Dave Hansen <haveblue@us.ibm.com>
Date: Fri, 15 Feb 2008 14:37:59 -0800
Subject: [PATCH] r/o bind mounts: track numbers of writers to mounts

This is the real meat of the entire series.  It actually
implements the tracking of the number of writers to a mount.
However, it causes scalability problems because there can be
hundreds of cpus doing open()/close() on files on the same mnt at
the same time.  Even an atomic_t in the mnt has massive scalaing
problems because the cacheline gets so terribly contended.

This uses a statically-allocated percpu variable.  All want/drop
operations are local to a cpu as long that cpu operates on the same
mount, and there are no writer count imbalances.  Writer count
imbalances happen when a write is taken on one cpu, and released
on another, like when an open/close pair is performed on two

Upon a remount,ro request, all of the data from the percpu
variables is collected (expensive, but very rare) and we determine
if there are any outstanding writers to the mount.

I've written a little benchmark to sit in a loop for a couple of
seconds in several cpus in parallel doing open/write/close loops.

http://sr71.net/~dave/linux/openbench.c

The code in here is a a worst-possible case for this patch.  It
does opens on a _pair_ of files in two different mounts in parallel.
This should cause my code to lose its "operate on the same mount"
optimization completely.  This worst-case scenario causes a 3%
degredation in the benchmark.

I could probably get rid of even this 3%, but it would be more
complex than what I have here, and I think this is getting into
acceptable territory.  In practice, I expect writing more than 3
bytes to a file, as well as disk I/O to mask any effects that this
has.

(To get rid of that 3%, we could have an #defined number of mounts
in the percpu variable.  So, instead of a CPU getting operate only
on percpu data when it accesses only one mount, it could stay on
percpu data when it only accesses N or fewer mounts.)

[AV] merged fix for __clear_mnt_mount() stepping on freed vfsmount

Acked-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namespace.c        | 252 +++++++++++++++++++++++++++++++++++++++++++++++---
 include/linux/mount.h |   7 ++
 2 files changed, 244 insertions(+), 15 deletions(-)

(limited to 'include/linux')

diff --git a/fs/namespace.c b/fs/namespace.c
index 066b393578c1..e3ce18d91aad 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -17,6 +17,7 @@
 #include <linux/quotaops.h>
 #include <linux/acct.h>
 #include <linux/capability.h>
+#include <linux/cpumask.h>
 #include <linux/module.h>
 #include <linux/sysfs.h>
 #include <linux/seq_file.h>
@@ -55,6 +56,8 @@ static inline unsigned long hash(struct vfsmount *mnt, struct dentry *dentry)
 	return tmp & (HASH_SIZE - 1);
 }
 
+#define MNT_WRITER_UNDERFLOW_LIMIT -(1<<16)
+
 struct vfsmount *alloc_vfsmnt(const char *name)
 {
 	struct vfsmount *mnt = kmem_cache_zalloc(mnt_cache, GFP_KERNEL);
@@ -68,6 +71,7 @@ struct vfsmount *alloc_vfsmnt(const char *name)
 		INIT_LIST_HEAD(&mnt->mnt_share);
 		INIT_LIST_HEAD(&mnt->mnt_slave_list);
 		INIT_LIST_HEAD(&mnt->mnt_slave);
+		atomic_set(&mnt->__mnt_writers, 0);
 		if (name) {
 			int size = strlen(name) + 1;
 			char *newname = kmalloc(size, GFP_KERNEL);
@@ -80,6 +84,92 @@ struct vfsmount *alloc_vfsmnt(const char *name)
 	return mnt;
 }
 
+/*
+ * Most r/o checks on a fs are for operations that take
+ * discrete amounts of time, like a write() or unlink().
+ * We must keep track of when those operations start
+ * (for permission checks) and when they end, so that
+ * we can determine when writes are able to occur to
+ * a filesystem.
+ */
+/*
+ * __mnt_is_readonly: check whether a mount is read-only
+ * @mnt: the mount to check for its write status
+ *
+ * This shouldn't be used directly ouside of the VFS.
+ * It does not guarantee that the filesystem will stay
+ * r/w, just that it is right *now*.  This can not and
+ * should not be used in place of IS_RDONLY(inode).
+ * mnt_want/drop_write() will _keep_ the filesystem
+ * r/w.
+ */
+int __mnt_is_readonly(struct vfsmount *mnt)
+{
+	return (mnt->mnt_sb->s_flags & MS_RDONLY);
+}
+EXPORT_SYMBOL_GPL(__mnt_is_readonly);
+
+struct mnt_writer {
+	/*
+	 * If holding multiple instances of this lock, they
+	 * must be ordered by cpu number.
+	 */
+	spinlock_t lock;
+	struct lock_class_key lock_class; /* compiles out with !lockdep */
+	unsigned long count;
+	struct vfsmount *mnt;
+} ____cacheline_aligned_in_smp;
+static DEFINE_PER_CPU(struct mnt_writer, mnt_writers);
+
+static int __init init_mnt_writers(void)
+{
+	int cpu;
+	for_each_possible_cpu(cpu) {
+		struct mnt_writer *writer = &per_cpu(mnt_writers, cpu);
+		spin_lock_init(&writer->lock);
+		lockdep_set_class(&writer->lock, &writer->lock_class);
+		writer->count = 0;
+	}
+	return 0;
+}
+fs_initcall(init_mnt_writers);
+
+static void unlock_mnt_writers(void)
+{
+	int cpu;
+	struct mnt_writer *cpu_writer;
+
+	for_each_possible_cpu(cpu) {
+		cpu_writer = &per_cpu(mnt_writers, cpu);
+		spin_unlock(&cpu_writer->lock);
+	}
+}
+
+static inline void __clear_mnt_count(struct mnt_writer *cpu_writer)
+{
+	if (!cpu_writer->mnt)
+		return;
+	/*
+	 * This is in case anyone ever leaves an invalid,
+	 * old ->mnt and a count of 0.
+	 */
+	if (!cpu_writer->count)
+		return;
+	atomic_add(cpu_writer->count, &cpu_writer->mnt->__mnt_writers);
+	cpu_writer->count = 0;
+}
+ /*
+ * must hold cpu_writer->lock
+ */
+static inline void use_cpu_writer_for_mount(struct mnt_writer *cpu_writer,
+					  struct vfsmount *mnt)
+{
+	if (cpu_writer->mnt == mnt)
+		return;
+	__clear_mnt_count(cpu_writer);
+	cpu_writer->mnt = mnt;
+}
+
 /*
  * Most r/o checks on a fs are for operations that take
  * discrete amounts of time, like a write() or unlink().
@@ -100,12 +190,77 @@ struct vfsmount *alloc_vfsmnt(const char *name)
  */
 int mnt_want_write(struct vfsmount *mnt)
 {
-	if (__mnt_is_readonly(mnt))
-		return -EROFS;
-	return 0;
+	int ret = 0;
+	struct mnt_writer *cpu_writer;
+
+	cpu_writer = &get_cpu_var(mnt_writers);
+	spin_lock(&cpu_writer->lock);
+	if (__mnt_is_readonly(mnt)) {
+		ret = -EROFS;
+		goto out;
+	}
+	use_cpu_writer_for_mount(cpu_writer, mnt);
+	cpu_writer->count++;
+out:
+	spin_unlock(&cpu_writer->lock);
+	put_cpu_var(mnt_writers);
+	return ret;
 }
 EXPORT_SYMBOL_GPL(mnt_want_write);
 
+static void lock_mnt_writers(void)
+{
+	int cpu;
+	struct mnt_writer *cpu_writer;
+
+	for_each_possible_cpu(cpu) {
+		cpu_writer = &per_cpu(mnt_writers, cpu);
+		spin_lock(&cpu_writer->lock);
+		__clear_mnt_count(cpu_writer);
+		cpu_writer->mnt = NULL;
+	}
+}
+
+/*
+ * These per-cpu write counts are not guaranteed to have
+ * matched increments and decrements on any given cpu.
+ * A file open()ed for write on one cpu and close()d on
+ * another cpu will imbalance this count.  Make sure it
+ * does not get too far out of whack.
+ */
+static void handle_write_count_underflow(struct vfsmount *mnt)
+{
+	if (atomic_read(&mnt->__mnt_writers) >=
+	    MNT_WRITER_UNDERFLOW_LIMIT)
+		return;
+	/*
+	 * It isn't necessary to hold all of the locks
+	 * at the same time, but doing it this way makes
+	 * us share a lot more code.
+	 */
+	lock_mnt_writers();
+	/*
+	 * vfsmount_lock is for mnt_flags.
+	 */
+	spin_lock(&vfsmount_lock);
+	/*
+	 * If coalescing the per-cpu writer counts did not
+	 * get us back to a positive writer count, we have
+	 * a bug.
+	 */
+	if ((atomic_read(&mnt->__mnt_writers) < 0) &&
+	    !(mnt->mnt_flags & MNT_IMBALANCED_WRITE_COUNT)) {
+		printk(KERN_DEBUG "leak detected on mount(%p) writers "
+				"count: %d\n",
+			mnt, atomic_read(&mnt->__mnt_writers));
+		WARN_ON(1);
+		/* use the flag to keep the dmesg spam down */
+		mnt->mnt_flags |= MNT_IMBALANCED_WRITE_COUNT;
+	}
+	spin_unlock(&vfsmount_lock);
+	unlock_mnt_writers();
+}
+
 /**
  * mnt_drop_write - give up write access to a mount
  * @mnt: the mount on which to give up write access
@@ -116,23 +271,61 @@ EXPORT_SYMBOL_GPL(mnt_want_write);
  */
 void mnt_drop_write(struct vfsmount *mnt)
 {
+	int must_check_underflow = 0;
+	struct mnt_writer *cpu_writer;
+
+	cpu_writer = &get_cpu_var(mnt_writers);
+	spin_lock(&cpu_writer->lock);
+
+	use_cpu_writer_for_mount(cpu_writer, mnt);
+	if (cpu_writer->count > 0) {
+		cpu_writer->count--;
+	} else {
+		must_check_underflow = 1;
+		atomic_dec(&mnt->__mnt_writers);
+	}
+
+	spin_unlock(&cpu_writer->lock);
+	/*
+	 * Logically, we could call this each time,
+	 * but the __mnt_writers cacheline tends to
+	 * be cold, and makes this expensive.
+	 */
+	if (must_check_underflow)
+		handle_write_count_underflow(mnt);
+	/*
+	 * This could be done right after the spinlock
+	 * is taken because the spinlock keeps us on
+	 * the cpu, and disables preemption.  However,
+	 * putting it here bounds the amount that
+	 * __mnt_writers can underflow.  Without it,
+	 * we could theoretically wrap __mnt_writers.
+	 */
+	put_cpu_var(mnt_writers);
 }
 EXPORT_SYMBOL_GPL(mnt_drop_write);
 
-/*
- * __mnt_is_readonly: check whether a mount is read-only
- * @mnt: the mount to check for its write status
- *
- * This shouldn't be used directly ouside of the VFS.
- * It does not guarantee that the filesystem will stay
- * r/w, just that it is right *now*.  This can not and
- * should not be used in place of IS_RDONLY(inode).
- */
-int __mnt_is_readonly(struct vfsmount *mnt)
+int mnt_make_readonly(struct vfsmount *mnt)
 {
-	return (mnt->mnt_sb->s_flags & MS_RDONLY);
+	int ret = 0;
+
+	lock_mnt_writers();
+	/*
+	 * With all the locks held, this value is stable
+	 */
+	if (atomic_read(&mnt->__mnt_writers) > 0) {
+		ret = -EBUSY;
+		goto out;
+	}
+	/*
+	 * actually set mount's r/o flag here to make
+	 * __mnt_is_readonly() true, which keeps anyone
+	 * from doing a successful mnt_want_write().
+	 */
+out:
+	unlock_mnt_writers();
+	return ret;
 }
-EXPORT_SYMBOL_GPL(__mnt_is_readonly);
 
 int simple_set_mnt(struct vfsmount *mnt, struct super_block *sb)
 {
@@ -325,7 +518,36 @@ static struct vfsmount *clone_mnt(struct vfsmount *old, struct dentry *root,
 
 static inline void __mntput(struct vfsmount *mnt)
 {
+	int cpu;
 	struct super_block *sb = mnt->mnt_sb;
+	/*
+	 * We don't have to hold all of the locks at the
+	 * same time here because we know that we're the
+	 * last reference to mnt and that no new writers
+	 * can come in.
+	 */
+	for_each_possible_cpu(cpu) {
+		struct mnt_writer *cpu_writer = &per_cpu(mnt_writers, cpu);
+		if (cpu_writer->mnt != mnt)
+			continue;
+		spin_lock(&cpu_writer->lock);
+		atomic_add(cpu_writer->count, &mnt->__mnt_writers);
+		cpu_writer->count = 0;
+		/*
+		 * Might as well do this so that no one
+		 * ever sees the pointer and expects
+		 * it to be valid.
+		 */
+		cpu_writer->mnt = NULL;
+		spin_unlock(&cpu_writer->lock);
+	}
+	/*
+	 * This probably indicates that somebody messed
+	 * up a mnt_want/drop_write() pair.  If this
+	 * happens, the filesystem was probably unable
+	 * to make r/w->r/o transitions.
+	 */
+	WARN_ON(atomic_read(&mnt->__mnt_writers));
 	dput(mnt->mnt_root);
 	free_vfsmnt(mnt);
 	deactivate_super(sb);
diff --git a/include/linux/mount.h b/include/linux/mount.h
index 2eecd2c8c760..8c8e94369ac8 100644
--- a/include/linux/mount.h
+++ b/include/linux/mount.h
@@ -14,6 +14,7 @@
 
 #include <linux/types.h>
 #include <linux/list.h>
+#include <linux/nodemask.h>
 #include <linux/spinlock.h>
 #include <asm/atomic.h>
 
@@ -30,6 +31,7 @@ struct mnt_namespace;
 #define MNT_RELATIME	0x20
 
 #define MNT_SHRINKABLE	0x100
+#define MNT_IMBALANCED_WRITE_COUNT	0x200 /* just for debugging */
 
 #define MNT_SHARED	0x1000	/* if the vfsmount is a shared mount */
 #define MNT_UNBINDABLE	0x2000	/* if the vfsmount is a unbindable mount */
@@ -62,6 +64,11 @@ struct vfsmount {
 	int mnt_expiry_mark;		/* true if marked for expiry */
 	int mnt_pinned;
 	int mnt_ghosts;
+	/*
+	 * This value is not stable unless all of the mnt_writers[] spinlocks
+	 * are held, and all mnt_writer[]s on this mount have 0 as their ->count
+	 */
+	atomic_t __mnt_writers;
 };
 
 static inline struct vfsmount *mntget(struct vfsmount *mnt)
-- 
cgit v1.2.3


From 2e4b7fcd926006531935a4c79a5e9349fe51125b Mon Sep 17 00:00:00 2001
From: Dave Hansen <haveblue@us.ibm.com>
Date: Fri, 15 Feb 2008 14:38:00 -0800
Subject: [PATCH] r/o bind mounts: honor mount writer counts at remount

Originally from: Herbert Poetzl <herbert@13thfloor.at>

This is the core of the read-only bind mount patch set.

Note that this does _not_ add a "ro" option directly to the bind mount
operation.  If you require such a mount, you must first do the bind, then
follow it up with a 'mount -o remount,ro' operation:

If you wish to have a r/o bind mount of /foo on bar:

	mount --bind /foo /bar
	mount -o remount,ro /bar

Acked-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namespace.c        | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 include/linux/mount.h |  1 +
 2 files changed, 44 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/fs/namespace.c b/fs/namespace.c
index e3ce18d91aad..678f7ce060f2 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -105,7 +105,11 @@ struct vfsmount *alloc_vfsmnt(const char *name)
  */
 int __mnt_is_readonly(struct vfsmount *mnt)
 {
-	return (mnt->mnt_sb->s_flags & MS_RDONLY);
+	if (mnt->mnt_flags & MNT_READONLY)
+		return 1;
+	if (mnt->mnt_sb->s_flags & MS_RDONLY)
+		return 1;
+	return 0;
 }
 EXPORT_SYMBOL_GPL(__mnt_is_readonly);
 
@@ -305,7 +309,7 @@ void mnt_drop_write(struct vfsmount *mnt)
 }
 EXPORT_SYMBOL_GPL(mnt_drop_write);
 
-int mnt_make_readonly(struct vfsmount *mnt)
+static int mnt_make_readonly(struct vfsmount *mnt)
 {
 	int ret = 0;
 
@@ -318,15 +322,25 @@ int mnt_make_readonly(struct vfsmount *mnt)
 		goto out;
 	}
 	/*
-	 * actually set mount's r/o flag here to make
-	 * __mnt_is_readonly() true, which keeps anyone
-	 * from doing a successful mnt_want_write().
+	 * nobody can do a successful mnt_want_write() with all
+	 * of the counts in MNT_DENIED_WRITE and the locks held.
 	 */
+	spin_lock(&vfsmount_lock);
+	if (!ret)
+		mnt->mnt_flags |= MNT_READONLY;
+	spin_unlock(&vfsmount_lock);
 out:
 	unlock_mnt_writers();
 	return ret;
 }
 
+static void __mnt_unmake_readonly(struct vfsmount *mnt)
+{
+	spin_lock(&vfsmount_lock);
+	mnt->mnt_flags &= ~MNT_READONLY;
+	spin_unlock(&vfsmount_lock);
+}
+
 int simple_set_mnt(struct vfsmount *mnt, struct super_block *sb)
 {
 	mnt->mnt_sb = sb;
@@ -693,7 +707,7 @@ static int show_vfsmnt(struct seq_file *m, void *v)
 		seq_putc(m, '.');
 		mangle(m, mnt->mnt_sb->s_subtype);
 	}
-	seq_puts(m, mnt->mnt_sb->s_flags & MS_RDONLY ? " ro" : " rw");
+	seq_puts(m, __mnt_is_readonly(mnt) ? " ro" : " rw");
 	for (fs_infop = fs_info; fs_infop->flag; fs_infop++) {
 		if (mnt->mnt_sb->s_flags & fs_infop->flag)
 			seq_puts(m, fs_infop->str);
@@ -1295,6 +1309,23 @@ out:
 	return err;
 }
 
+static int change_mount_flags(struct vfsmount *mnt, int ms_flags)
+{
+	int error = 0;
+	int readonly_request = 0;
+
+	if (ms_flags & MS_RDONLY)
+		readonly_request = 1;
+	if (readonly_request == __mnt_is_readonly(mnt))
+		return 0;
+
+	if (readonly_request)
+		error = mnt_make_readonly(mnt);
+	else
+		__mnt_unmake_readonly(mnt);
+	return error;
+}
+
 /*
  * change filesystem flags. dir should be a physical root of filesystem.
  * If you've mounted a non-root directory somewhere and want to do remount
@@ -1317,7 +1348,10 @@ static noinline int do_remount(struct nameidata *nd, int flags, int mnt_flags,
 		return -EINVAL;
 
 	down_write(&sb->s_umount);
-	err = do_remount_sb(sb, flags, data, 0);
+	if (flags & MS_BIND)
+		err = change_mount_flags(nd->path.mnt, flags);
+	else
+		err = do_remount_sb(sb, flags, data, 0);
 	if (!err)
 		nd->path.mnt->mnt_flags = mnt_flags;
 	up_write(&sb->s_umount);
@@ -1701,6 +1735,8 @@ long do_mount(char *dev_name, char *dir_name, char *type_page,
 		mnt_flags |= MNT_NODIRATIME;
 	if (flags & MS_RELATIME)
 		mnt_flags |= MNT_RELATIME;
+	if (flags & MS_RDONLY)
+		mnt_flags |= MNT_READONLY;
 
 	flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | MS_ACTIVE |
 		   MS_NOATIME | MS_NODIRATIME | MS_RELATIME| MS_KERNMOUNT);
diff --git a/include/linux/mount.h b/include/linux/mount.h
index 8c8e94369ac8..d6600e3f7e45 100644
--- a/include/linux/mount.h
+++ b/include/linux/mount.h
@@ -29,6 +29,7 @@ struct mnt_namespace;
 #define MNT_NOATIME	0x08
 #define MNT_NODIRATIME	0x10
 #define MNT_RELATIME	0x20
+#define MNT_READONLY	0x40	/* does the user want this to be r/o? */
 
 #define MNT_SHRINKABLE	0x100
 #define MNT_IMBALANCED_WRITE_COUNT	0x200 /* just for debugging */
-- 
cgit v1.2.3


From ad775f5a8faa5845377f093ca11caf577404add9 Mon Sep 17 00:00:00 2001
From: Dave Hansen <haveblue@us.ibm.com>
Date: Fri, 15 Feb 2008 14:38:01 -0800
Subject: [PATCH] r/o bind mounts: debugging for missed calls

There have been a few oopses caused by 'struct file's with NULL f_vfsmnts.
There was also a set of potentially missed mnt_want_write()s from
dentry_open() calls.

This patch provides a very simple debugging framework to catch these kinds of
bugs.  It will WARN_ON() them, but should stop us from having any oopses or
mnt_writer count imbalances.

I'm quite convinced that this is a good thing because it found bugs in the
stuff I was working on as soon as I wrote it.

[hch: made it conditional on a debug option.
      But it's still a little bit too ugly]

[hch: merged forced remount r/o fix from Dave and akpm's fix for the fix]

Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Acked-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/file_table.c    | 11 +++++++++--
 fs/open.c          | 12 +++++++++++-
 fs/super.c         |  3 +++
 include/linux/fs.h | 49 +++++++++++++++++++++++++++++++++++++++++++++++++
 lib/Kconfig.debug  | 10 ++++++++++
 5 files changed, 82 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/fs/file_table.c b/fs/file_table.c
index 71efc7000226..7a0a9b872251 100644
--- a/fs/file_table.c
+++ b/fs/file_table.c
@@ -42,6 +42,7 @@ static inline void file_free_rcu(struct rcu_head *head)
 static inline void file_free(struct file *f)
 {
 	percpu_counter_dec(&nr_files);
+	file_check_state(f);
 	call_rcu(&f->f_u.fu_rcuhead, file_free_rcu);
 }
 
@@ -207,6 +208,7 @@ int init_file(struct file *file, struct vfsmount *mnt, struct dentry *dentry,
 	 * that we can do debugging checks at __fput()
 	 */
 	if ((mode & FMODE_WRITE) && !special_file(dentry->d_inode->i_mode)) {
+		file_take_write(file);
 		error = mnt_want_write(mnt);
 		WARN_ON(error);
 	}
@@ -237,8 +239,13 @@ void drop_file_write_access(struct file *file)
 	struct inode *inode = dentry->d_inode;
 
 	put_write_access(inode);
-	if (!special_file(inode->i_mode))
-		mnt_drop_write(mnt);
+
+	if (special_file(inode->i_mode))
+		return;
+	if (file_check_writeable(file) != 0)
+		return;
+	mnt_drop_write(mnt);
+	file_release_write(file);
 }
 EXPORT_SYMBOL_GPL(drop_file_write_access);
 
diff --git a/fs/open.c b/fs/open.c
index e58382d57e72..b70e7666bb2c 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -806,6 +806,8 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt,
 		error = __get_file_write_access(inode, mnt);
 		if (error)
 			goto cleanup_file;
+		if (!special_file(inode->i_mode))
+			file_take_write(f);
 	}
 
 	f->f_mapping = inode->i_mapping;
@@ -847,8 +849,16 @@ cleanup_all:
 	fops_put(f->f_op);
 	if (f->f_mode & FMODE_WRITE) {
 		put_write_access(inode);
-		if (!special_file(inode->i_mode))
+		if (!special_file(inode->i_mode)) {
+			/*
+			 * We don't consider this a real
+			 * mnt_want/drop_write() pair
+			 * because it all happenend right
+			 * here, so just reset the state.
+			 */
+			file_reset_write(f);
 			mnt_drop_write(mnt);
+		}
 	}
 	file_kill(f);
 	f->f_path.dentry = NULL;
diff --git a/fs/super.c b/fs/super.c
index 01d5c40e9119..1f8f05ede437 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -579,6 +579,9 @@ retry:
 		if (!(f->f_mode & FMODE_WRITE))
 			continue;
 		f->f_mode &= ~FMODE_WRITE;
+		if (file_check_writeable(f) != 0)
+			continue;
+		file_release_write(f);
 		mnt = mntget(f->f_path.mnt);
 		file_list_unlock();
 		/*
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 013b9c2b88e6..d1eeea669d2c 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -776,6 +776,9 @@ static inline int ra_has_index(struct file_ra_state *ra, pgoff_t index)
 		index <  ra->start + ra->size);
 }
 
+#define FILE_MNT_WRITE_TAKEN	1
+#define FILE_MNT_WRITE_RELEASED	2
+
 struct file {
 	/*
 	 * fu_list becomes invalid after file_free is called and queued via
@@ -810,6 +813,9 @@ struct file {
 	spinlock_t		f_ep_lock;
 #endif /* #ifdef CONFIG_EPOLL */
 	struct address_space	*f_mapping;
+#ifdef CONFIG_DEBUG_WRITECOUNT
+	unsigned long f_mnt_write_state;
+#endif
 };
 extern spinlock_t files_lock;
 #define file_list_lock() spin_lock(&files_lock);
@@ -818,6 +824,49 @@ extern spinlock_t files_lock;
 #define get_file(x)	atomic_inc(&(x)->f_count)
 #define file_count(x)	atomic_read(&(x)->f_count)
 
+#ifdef CONFIG_DEBUG_WRITECOUNT
+static inline void file_take_write(struct file *f)
+{
+	WARN_ON(f->f_mnt_write_state != 0);
+	f->f_mnt_write_state = FILE_MNT_WRITE_TAKEN;
+}
+static inline void file_release_write(struct file *f)
+{
+	f->f_mnt_write_state |= FILE_MNT_WRITE_RELEASED;
+}
+static inline void file_reset_write(struct file *f)
+{
+	f->f_mnt_write_state = 0;
+}
+static inline void file_check_state(struct file *f)
+{
+	/*
+	 * At this point, either both or neither of these bits
+	 * should be set.
+	 */
+	WARN_ON(f->f_mnt_write_state == FILE_MNT_WRITE_TAKEN);
+	WARN_ON(f->f_mnt_write_state == FILE_MNT_WRITE_RELEASED);
+}
+static inline int file_check_writeable(struct file *f)
+{
+	if (f->f_mnt_write_state == FILE_MNT_WRITE_TAKEN)
+		return 0;
+	printk(KERN_WARNING "writeable file with no "
+			    "mnt_want_write()\n");
+	WARN_ON(1);
+	return -EINVAL;
+}
+#else /* !CONFIG_DEBUG_WRITECOUNT */
+static inline void file_take_write(struct file *filp) {}
+static inline void file_release_write(struct file *filp) {}
+static inline void file_reset_write(struct file *filp) {}
+static inline void file_check_state(struct file *filp) {}
+static inline int file_check_writeable(struct file *filp)
+{
+	return 0;
+}
+#endif /* CONFIG_DEBUG_WRITECOUNT */
+
 #define	MAX_NON_LFS	((1UL<<31) - 1)
 
 /* Page cache limit. The filesystems should put that into their s_maxbytes 
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index 95de3102bc87..623ef24c2381 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -427,6 +427,16 @@ config DEBUG_VM
 
 	  If unsure, say N.
 
+config DEBUG_WRITECOUNT
+	bool "Debug filesystem writers count"
+	depends on DEBUG_KERNEL
+	help
+	  Enable this to catch wrong use of the writers count in struct
+	  vfsmount.  This will increase the size of each file struct by
+	  32 bits.
+
+	  If unsure, say N.
+
 config DEBUG_LIST
 	bool "Debug linked list manipulation"
 	depends on DEBUG_KERNEL
-- 
cgit v1.2.3


From a1635b8fe59de2c5223cda5ca8397b875c901904 Mon Sep 17 00:00:00 2001
From: Philipp Zabel <philipp.zabel@gmail.com>
Date: Wed, 9 Apr 2008 19:20:34 +0100
Subject: [ARM] 4947/1: htc-egpio, a driver for GPIO/IRQ expanders with fixed
 input/output pins

implemented in CPLD chips on several HTC devices.

The original driver was written by Kevin O'Connor, I have adapted it to
use gpiolib and made the bus/register widths configurable.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
---
 drivers/mfd/Kconfig           |   8 +
 drivers/mfd/Makefile          |   2 +
 drivers/mfd/htc-egpio.c       | 440 ++++++++++++++++++++++++++++++++++++++++++
 include/linux/mfd/htc-egpio.h |  57 ++++++
 4 files changed, 507 insertions(+)
 create mode 100644 drivers/mfd/htc-egpio.c
 create mode 100644 include/linux/mfd/htc-egpio.h

(limited to 'include/linux')

diff --git a/drivers/mfd/Kconfig b/drivers/mfd/Kconfig
index 0c886c882385..284b2dc03444 100644
--- a/drivers/mfd/Kconfig
+++ b/drivers/mfd/Kconfig
@@ -22,6 +22,14 @@ config MFD_ASIC3
 	  This driver supports the ASIC3 multifunction chip found on many
 	  PDAs (mainly iPAQ and HTC based ones)
 
+config HTC_EGPIO
+	bool "HTC EGPIO support"
+	depends on GENERIC_HARDIRQS && HAVE_GPIO_LIB
+	help
+	    This driver supports the CPLD egpio chip present on
+	    several HTC phones.  It provides basic support for input
+	    pins, output pins, and irqs.
+
 endmenu
 
 menu "Multimedia Capabilities Port drivers"
diff --git a/drivers/mfd/Makefile b/drivers/mfd/Makefile
index 521cd5cb68af..f81f8d2e2c03 100644
--- a/drivers/mfd/Makefile
+++ b/drivers/mfd/Makefile
@@ -5,6 +5,8 @@
 obj-$(CONFIG_MFD_SM501)		+= sm501.o
 obj-$(CONFIG_MFD_ASIC3)		+= asic3.o
 
+obj-$(CONFIG_HTC_EGPIO)		+= htc-egpio.o
+
 obj-$(CONFIG_MCP)		+= mcp-core.o
 obj-$(CONFIG_MCP_SA11X0)	+= mcp-sa11x0.o
 obj-$(CONFIG_MCP_UCB1200)	+= ucb1x00-core.o
diff --git a/drivers/mfd/htc-egpio.c b/drivers/mfd/htc-egpio.c
new file mode 100644
index 000000000000..8872cc077519
--- /dev/null
+++ b/drivers/mfd/htc-egpio.c
@@ -0,0 +1,440 @@
+/*
+ * Support for the GPIO/IRQ expander chips present on several HTC phones.
+ * These are implemented in CPLD chips present on the board.
+ *
+ * Copyright (c) 2007 Kevin O'Connor <kevin@koconnor.net>
+ * Copyright (c) 2007 Philipp Zabel <philipp.zabel@gmail.com>
+ *
+ * This file may be distributed under the terms of the GNU GPL license.
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/io.h>
+#include <linux/spinlock.h>
+#include <linux/platform_device.h>
+#include <linux/module.h>
+#include <linux/mfd/htc-egpio.h>
+
+struct egpio_chip {
+	int              reg_start;
+	int              cached_values;
+	unsigned long    is_out;
+	struct device    *dev;
+	struct gpio_chip chip;
+};
+
+struct egpio_info {
+	spinlock_t        lock;
+
+	/* iomem info */
+	void __iomem      *base_addr;
+	int               bus_shift;	/* byte shift */
+	int               reg_shift;	/* bit shift */
+	int               reg_mask;
+
+	/* irq info */
+	int               ack_register;
+	int               ack_write;
+	u16               irqs_enabled;
+	uint              irq_start;
+	int               nirqs;
+	uint              chained_irq;
+
+	/* egpio info */
+	struct egpio_chip *chip;
+	int               nchips;
+};
+
+static inline void egpio_writew(u16 value, struct egpio_info *ei, int reg)
+{
+	writew(value, ei->base_addr + (reg << ei->bus_shift));
+}
+
+static inline u16 egpio_readw(struct egpio_info *ei, int reg)
+{
+	return readw(ei->base_addr + (reg << ei->bus_shift));
+}
+
+/*
+ * IRQs
+ */
+
+static inline void ack_irqs(struct egpio_info *ei)
+{
+	egpio_writew(ei->ack_write, ei, ei->ack_register);
+	pr_debug("EGPIO ack - write %x to base+%x\n",
+			ei->ack_write, ei->ack_register << ei->bus_shift);
+}
+
+static void egpio_ack(unsigned int irq)
+{
+}
+
+/* There does not appear to be a way to proactively mask interrupts
+ * on the egpio chip itself.  So, we simply ignore interrupts that
+ * aren't desired. */
+static void egpio_mask(unsigned int irq)
+{
+	struct egpio_info *ei = get_irq_chip_data(irq);
+	ei->irqs_enabled &= ~(1 << (irq - ei->irq_start));
+	pr_debug("EGPIO mask %d %04x\n", irq, ei->irqs_enabled);
+}
+static void egpio_unmask(unsigned int irq)
+{
+	struct egpio_info *ei = get_irq_chip_data(irq);
+	ei->irqs_enabled |= 1 << (irq - ei->irq_start);
+	pr_debug("EGPIO unmask %d %04x\n", irq, ei->irqs_enabled);
+}
+
+static struct irq_chip egpio_muxed_chip = {
+	.name   = "htc-egpio",
+	.ack	= egpio_ack,
+	.mask   = egpio_mask,
+	.unmask = egpio_unmask,
+};
+
+static void egpio_handler(unsigned int irq, struct irq_desc *desc)
+{
+	struct egpio_info *ei = get_irq_data(irq);
+	int irqpin;
+
+	/* Read current pins. */
+	unsigned long readval = egpio_readw(ei, ei->ack_register);
+	pr_debug("IRQ reg: %x\n", (unsigned int)readval);
+	/* Ack/unmask interrupts. */
+	ack_irqs(ei);
+	/* Process all set pins. */
+	readval &= ei->irqs_enabled;
+	for_each_bit(irqpin, &readval, ei->nirqs) {
+		/* Run irq handler */
+		pr_debug("got IRQ %d\n", irqpin);
+		irq = ei->irq_start + irqpin;
+		desc = &irq_desc[irq];
+		desc->handle_irq(irq, desc);
+	}
+}
+
+int htc_egpio_get_wakeup_irq(struct device *dev)
+{
+	struct egpio_info *ei = dev_get_drvdata(dev);
+
+	/* Read current pins. */
+	u16 readval = egpio_readw(ei, ei->ack_register);
+	/* Ack/unmask interrupts. */
+	ack_irqs(ei);
+	/* Return first set pin. */
+	readval &= ei->irqs_enabled;
+	return ei->irq_start + ffs(readval) - 1;
+}
+EXPORT_SYMBOL(htc_egpio_get_wakeup_irq);
+
+static inline int egpio_pos(struct egpio_info *ei, int bit)
+{
+	return bit >> ei->reg_shift;
+}
+
+static inline int egpio_bit(struct egpio_info *ei, int bit)
+{
+	return 1 << (bit & ((1 << ei->reg_shift)-1));
+}
+
+/*
+ * Input pins
+ */
+
+static int egpio_get(struct gpio_chip *chip, unsigned offset)
+{
+	struct egpio_chip *egpio;
+	struct egpio_info *ei;
+	unsigned           bit;
+	int                reg;
+	int                value;
+
+	pr_debug("egpio_get_value(%d)\n", chip->base + offset);
+
+	egpio = container_of(chip, struct egpio_chip, chip);
+	ei    = dev_get_drvdata(egpio->dev);
+	bit   = egpio_bit(ei, offset);
+	reg   = egpio->reg_start + egpio_pos(ei, offset);
+
+	value = egpio_readw(ei, reg);
+	pr_debug("readw(%p + %x) = %x\n",
+			ei->base_addr, reg << ei->bus_shift, value);
+	return value & bit;
+}
+
+static int egpio_direction_input(struct gpio_chip *chip, unsigned offset)
+{
+	struct egpio_chip *egpio;
+
+	egpio = container_of(chip, struct egpio_chip, chip);
+	return test_bit(offset, &egpio->is_out) ? -EINVAL : 0;
+}
+
+
+/*
+ * Output pins
+ */
+
+static void egpio_set(struct gpio_chip *chip, unsigned offset, int value)
+{
+	unsigned long     flag;
+	struct egpio_chip *egpio;
+	struct egpio_info *ei;
+	unsigned          bit;
+	int               pos;
+	int               reg;
+	int               shift;
+
+	pr_debug("egpio_set(%s, %d(%d), %d)\n",
+			chip->label, offset, offset+chip->base, value);
+
+	egpio = container_of(chip, struct egpio_chip, chip);
+	ei    = dev_get_drvdata(egpio->dev);
+	bit   = egpio_bit(ei, offset);
+	pos   = egpio_pos(ei, offset);
+	reg   = egpio->reg_start + pos;
+	shift = pos << ei->reg_shift;
+
+	pr_debug("egpio %s: reg %d = 0x%04x\n", value ? "set" : "clear",
+			reg, (egpio->cached_values >> shift) & ei->reg_mask);
+
+	spin_lock_irqsave(&ei->lock, flag);
+	if (value)
+		egpio->cached_values |= (1 << offset);
+	else
+		egpio->cached_values &= ~(1 << offset);
+	egpio_writew((egpio->cached_values >> shift) & ei->reg_mask, ei, reg);
+	spin_unlock_irqrestore(&ei->lock, flag);
+}
+
+static int egpio_direction_output(struct gpio_chip *chip,
+					unsigned offset, int value)
+{
+	struct egpio_chip *egpio;
+
+	egpio = container_of(chip, struct egpio_chip, chip);
+	if (test_bit(offset, &egpio->is_out)) {
+		egpio_set(chip, offset, value);
+		return 0;
+	} else {
+		return -EINVAL;
+	}
+}
+
+static void egpio_write_cache(struct egpio_info *ei)
+{
+	int               i;
+	struct egpio_chip *egpio;
+	int               shift;
+
+	for (i = 0; i < ei->nchips; i++) {
+		egpio = &(ei->chip[i]);
+		if (!egpio->is_out)
+			continue;
+
+		for (shift = 0; shift < egpio->chip.ngpio;
+				shift += (1<<ei->reg_shift)) {
+
+			int reg = egpio->reg_start + egpio_pos(ei, shift);
+
+			if (!((egpio->is_out >> shift) & ei->reg_mask))
+				continue;
+
+			pr_debug("EGPIO: setting %x to %x, was %x\n", reg,
+				(egpio->cached_values >> shift) & ei->reg_mask,
+				egpio_readw(ei, reg));
+
+			egpio_writew((egpio->cached_values >> shift)
+					& ei->reg_mask, ei, reg);
+		}
+	}
+}
+
+
+/*
+ * Setup
+ */
+
+static int __init egpio_probe(struct platform_device *pdev)
+{
+	struct htc_egpio_platform_data *pdata = pdev->dev.platform_data;
+	struct resource   *res;
+	struct egpio_info *ei;
+	struct gpio_chip  *chip;
+	unsigned int      irq, irq_end;
+	int               i;
+	int               ret;
+
+	/* Initialize ei data structure. */
+	ei = kzalloc(sizeof(*ei), GFP_KERNEL);
+	if (!ei)
+		return -ENOMEM;
+
+	spin_lock_init(&ei->lock);
+
+	/* Find chained irq */
+	ret = -EINVAL;
+	res = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
+	if (res)
+		ei->chained_irq = res->start;
+
+	/* Map egpio chip into virtual address space. */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		goto fail;
+	ei->base_addr = ioremap_nocache(res->start, res->end - res->start);
+	if (!ei->base_addr)
+		goto fail;
+	pr_debug("EGPIO phys=%08x virt=%p\n", res->start, ei->base_addr);
+
+	if ((pdata->bus_width != 16) && (pdata->bus_width != 32))
+		goto fail;
+	ei->bus_shift = fls(pdata->bus_width - 1) - 3;
+	pr_debug("bus_shift = %d\n", ei->bus_shift);
+
+	if ((pdata->reg_width != 8) && (pdata->reg_width != 16))
+		goto fail;
+	ei->reg_shift = fls(pdata->reg_width - 1);
+	pr_debug("reg_shift = %d\n", ei->reg_shift);
+
+	ei->reg_mask = (1 << pdata->reg_width) - 1;
+
+	platform_set_drvdata(pdev, ei);
+
+	ei->nchips = pdata->num_chips;
+	ei->chip = kzalloc(sizeof(struct egpio_chip) * ei->nchips, GFP_KERNEL);
+	if (!ei) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+	for (i = 0; i < ei->nchips; i++) {
+		ei->chip[i].reg_start = pdata->chip[i].reg_start;
+		ei->chip[i].cached_values = pdata->chip[i].initial_values;
+		ei->chip[i].is_out = pdata->chip[i].direction;
+		ei->chip[i].dev = &(pdev->dev);
+		chip = &(ei->chip[i].chip);
+		chip->label           = "htc-egpio";
+		chip->get             = egpio_get;
+		chip->set             = egpio_set;
+		chip->direction_input = egpio_direction_input;
+		chip->direction_output = egpio_direction_output;
+		chip->base            = pdata->chip[i].gpio_base;
+		chip->ngpio           = pdata->chip[i].num_gpios;
+
+		gpiochip_add(chip);
+	}
+
+	/* Set initial pin values */
+	egpio_write_cache(ei);
+
+	ei->irq_start = pdata->irq_base;
+	ei->nirqs = pdata->num_irqs;
+	ei->ack_register = pdata->ack_register;
+
+	if (ei->chained_irq) {
+		/* Setup irq handlers */
+		ei->ack_write = 0xFFFF;
+		if (pdata->invert_acks)
+			ei->ack_write = 0;
+		irq_end = ei->irq_start + ei->nirqs;
+		for (irq = ei->irq_start; irq < irq_end; irq++) {
+			set_irq_chip(irq, &egpio_muxed_chip);
+			set_irq_chip_data(irq, ei);
+			set_irq_handler(irq, handle_simple_irq);
+			set_irq_flags(irq, IRQF_VALID | IRQF_PROBE);
+		}
+		set_irq_type(ei->chained_irq, IRQ_TYPE_EDGE_RISING);
+		set_irq_data(ei->chained_irq, ei);
+		set_irq_chained_handler(ei->chained_irq, egpio_handler);
+		ack_irqs(ei);
+
+		device_init_wakeup(&pdev->dev, 1);
+	}
+
+	return 0;
+
+fail:
+	printk(KERN_ERR "EGPIO failed to setup\n");
+	kfree(ei);
+	return ret;
+}
+
+static int __exit egpio_remove(struct platform_device *pdev)
+{
+	struct egpio_info *ei = platform_get_drvdata(pdev);
+	unsigned int      irq, irq_end;
+
+	if (ei->chained_irq) {
+		irq_end = ei->irq_start + ei->nirqs;
+		for (irq = ei->irq_start; irq < irq_end; irq++) {
+			set_irq_chip(irq, NULL);
+			set_irq_handler(irq, NULL);
+			set_irq_flags(irq, 0);
+		}
+		set_irq_chained_handler(ei->chained_irq, NULL);
+		device_init_wakeup(&pdev->dev, 0);
+	}
+	iounmap(ei->base_addr);
+	kfree(ei->chip);
+	kfree(ei);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int egpio_suspend(struct platform_device *pdev, pm_message_t state)
+{
+	struct egpio_info *ei = platform_get_drvdata(pdev);
+
+	if (ei->chained_irq && device_may_wakeup(&pdev->dev))
+		enable_irq_wake(ei->chained_irq);
+	return 0;
+}
+
+static int egpio_resume(struct platform_device *pdev)
+{
+	struct egpio_info *ei = platform_get_drvdata(pdev);
+
+	if (ei->chained_irq && device_may_wakeup(&pdev->dev))
+		disable_irq_wake(ei->chained_irq);
+
+	/* Update registers from the cache, in case
+	   the CPLD was powered off during suspend */
+	egpio_write_cache(ei);
+	return 0;
+}
+#else
+#define egpio_suspend NULL
+#define egpio_resume NULL
+#endif
+
+
+static struct platform_driver egpio_driver = {
+	.driver = {
+		.name = "htc-egpio",
+	},
+	.remove       = __exit_p(egpio_remove),
+	.suspend      = egpio_suspend,
+	.resume       = egpio_resume,
+};
+
+static int __init egpio_init(void)
+{
+	return platform_driver_probe(&egpio_driver, egpio_probe);
+}
+
+static void __exit egpio_exit(void)
+{
+	platform_driver_unregister(&egpio_driver);
+}
+
+/* start early for dependencies */
+subsys_initcall(egpio_init);
+module_exit(egpio_exit)
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Kevin O'Connor <kevin@koconnor.net>");
diff --git a/include/linux/mfd/htc-egpio.h b/include/linux/mfd/htc-egpio.h
new file mode 100644
index 000000000000..b4201c971367
--- /dev/null
+++ b/include/linux/mfd/htc-egpio.h
@@ -0,0 +1,57 @@
+/*
+ * HTC simple EGPIO irq and gpio extender
+ */
+
+#ifndef __HTC_EGPIO_H__
+#define __HTC_EGPIO_H__
+
+#include <linux/gpio.h>
+
+/* Descriptive values for all-in or all-out htc_egpio_chip descriptors. */
+#define HTC_EGPIO_OUTPUT (~0)
+#define HTC_EGPIO_INPUT  0
+
+/**
+ * struct htc_egpio_chip - descriptor to create gpio_chip for register range
+ * @reg_start: index of first register
+ * @gpio_base: gpio number of first pin in this register range
+ * @num_gpios: number of gpios in this register range, max BITS_PER_LONG
+ *    (number of registers = DIV_ROUND_UP(num_gpios, reg_width))
+ * @direction: bitfield, '0' = input, '1' = output,
+ */
+struct htc_egpio_chip {
+	int           reg_start;
+	int           gpio_base;
+	int           num_gpios;
+	unsigned long direction;
+	unsigned long initial_values;
+};
+
+/**
+ * struct htc_egpio_platform_data - description provided by the arch
+ * @irq_base: beginning of available IRQs (eg, IRQ_BOARD_START)
+ * @num_irqs: number of irqs
+ * @reg_width: number of bits per register, either 8 or 16 bit
+ * @bus_width: alignment of the registers, either 16 or 32 bit
+ * @invert_acks: set if chip requires writing '0' to ack an irq, instead of '1'
+ * @ack_register: location of the irq/ack register
+ * @chip: pointer to array of htc_egpio_chip descriptors
+ * @num_chips: number of egpio chip descriptors
+ */
+struct htc_egpio_platform_data {
+	int                   bus_width;
+	int                   reg_width;
+
+	int                   irq_base;
+	int                   num_irqs;
+	int                   invert_acks;
+	int                   ack_register;
+
+	struct htc_egpio_chip *chip;
+	int                   num_chips;
+};
+
+/* Determine the wakeup irq, to be called during early resume */
+extern int htc_egpio_get_wakeup_irq(struct device *dev);
+
+#endif
-- 
cgit v1.2.3


From 5dc3339aa5ba29593ea57814049ddca8c12831c8 Mon Sep 17 00:00:00 2001
From: Philipp Zabel <philipp.zabel@gmail.com>
Date: Sat, 12 Apr 2008 13:25:41 +0100
Subject: [ARM] 4964/1: htc-pasic3: MFD driver for PASIC3 LED control + DS1WM
 chip

This driver will provide registers, clocks and GPIOs of
the HTC PASIC3 (AIC3) and PASIC2 (AIC2) chips to the
ds1wm and leds-pasic3 drivers.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
---
 drivers/mfd/Kconfig            |   8 ++
 drivers/mfd/Makefile           |   1 +
 drivers/mfd/htc-pasic3.c       | 265 +++++++++++++++++++++++++++++++++++++++++
 include/linux/mfd/htc-pasic3.h |  55 +++++++++
 4 files changed, 329 insertions(+)
 create mode 100644 drivers/mfd/htc-pasic3.c
 create mode 100644 include/linux/mfd/htc-pasic3.h

(limited to 'include/linux')

diff --git a/drivers/mfd/Kconfig b/drivers/mfd/Kconfig
index 284b2dc03444..2566479937c9 100644
--- a/drivers/mfd/Kconfig
+++ b/drivers/mfd/Kconfig
@@ -30,6 +30,14 @@ config HTC_EGPIO
 	    several HTC phones.  It provides basic support for input
 	    pins, output pins, and irqs.
 
+config HTC_PASIC3
+	tristate "HTC PASIC3 LED/DS1WM chip support"
+	help
+	  This core driver provides register access for the LED/DS1WM
+	  chips labeled "AIC2" and "AIC3", found on HTC Blueangel and
+	  HTC Magician devices, respectively. Actual functionality is
+	  handled by the leds-pasic3 and ds1wm drivers.
+
 endmenu
 
 menu "Multimedia Capabilities Port drivers"
diff --git a/drivers/mfd/Makefile b/drivers/mfd/Makefile
index f81f8d2e2c03..eef4e26807df 100644
--- a/drivers/mfd/Makefile
+++ b/drivers/mfd/Makefile
@@ -6,6 +6,7 @@ obj-$(CONFIG_MFD_SM501)		+= sm501.o
 obj-$(CONFIG_MFD_ASIC3)		+= asic3.o
 
 obj-$(CONFIG_HTC_EGPIO)		+= htc-egpio.o
+obj-$(CONFIG_HTC_PASIC3)	+= htc-pasic3.o
 
 obj-$(CONFIG_MCP)		+= mcp-core.o
 obj-$(CONFIG_MCP_SA11X0)	+= mcp-sa11x0.o
diff --git a/drivers/mfd/htc-pasic3.c b/drivers/mfd/htc-pasic3.c
new file mode 100644
index 000000000000..af66f4f28300
--- /dev/null
+++ b/drivers/mfd/htc-pasic3.c
@@ -0,0 +1,265 @@
+/*
+ * Core driver for HTC PASIC3 LED/DS1WM chip.
+ *
+ * Copyright (C) 2006 Philipp Zabel <philipp.zabel@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+
+#include <linux/ds1wm.h>
+#include <linux/gpio.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/interrupt.h>
+#include <linux/mfd/htc-pasic3.h>
+
+#include <asm/arch/pxa-regs.h>
+
+struct pasic3_data {
+	void __iomem *mapping;
+	unsigned int bus_shift;
+	struct platform_device *ds1wm_pdev;
+	struct platform_device *led_pdev;
+};
+
+#define REG_ADDR  5
+#define REG_DATA  6
+#define NUM_REGS  7
+
+#define READ_MODE 0x80
+
+/*
+ * write to a secondary register on the PASIC3
+ */
+void pasic3_write_register(struct device *dev, u32 reg, u8 val)
+{
+	struct pasic3_data *asic = dev->driver_data;
+	int bus_shift = asic->bus_shift;
+	void __iomem *addr = asic->mapping + (REG_ADDR << bus_shift);
+	void __iomem *data = asic->mapping + (REG_DATA << bus_shift);
+
+	__raw_writeb(~READ_MODE & reg, addr);
+	__raw_writeb(val, data);
+}
+EXPORT_SYMBOL(pasic3_write_register); /* for leds-pasic3 */
+
+/*
+ * read from a secondary register on the PASIC3
+ */
+u8 pasic3_read_register(struct device *dev, u32 reg)
+{
+	struct pasic3_data *asic = dev->driver_data;
+	int bus_shift = asic->bus_shift;
+	void __iomem *addr = asic->mapping + (REG_ADDR << bus_shift);
+	void __iomem *data = asic->mapping + (REG_DATA << bus_shift);
+
+	__raw_writeb(READ_MODE | reg, addr);
+	return __raw_readb(data);
+}
+EXPORT_SYMBOL(pasic3_read_register); /* for leds-pasic3 */
+
+/*
+ * LEDs
+ */
+
+static int led_device_add(struct device *pasic3_dev,
+				const struct pasic3_leds_machinfo *pdata)
+{
+	struct pasic3_data *asic = pasic3_dev->driver_data;
+	struct platform_device *pdev;
+	int ret;
+
+	pdev = platform_device_alloc("pasic3-led", -1);
+	if (!pdev) {
+		dev_dbg(pasic3_dev, "failed to allocate LED platform device\n");
+		return -ENOMEM;
+	}
+
+	ret = platform_device_add_data(pdev, pdata,
+					sizeof(struct pasic3_leds_machinfo));
+	if (ret < 0) {
+		dev_dbg(pasic3_dev, "failed to add LED platform data\n");
+		goto exit_pdev_put;
+	}
+
+	pdev->dev.parent = pasic3_dev;
+	ret = platform_device_add(pdev);
+	if (ret < 0) {
+		dev_dbg(pasic3_dev, "failed to add LED platform device\n");
+		goto exit_pdev_put;
+	}
+
+	asic->led_pdev = pdev;
+	return 0;
+
+exit_pdev_put:
+	platform_device_put(pdev);
+	return ret;
+}
+
+/*
+ * DS1WM
+ */
+
+static void ds1wm_enable(struct platform_device *pdev)
+{
+	struct device *dev = pdev->dev.parent;
+	int c;
+
+	c = pasic3_read_register(dev, 0x28);
+	pasic3_write_register(dev, 0x28, c & 0x7f);
+
+	dev_dbg(dev, "DS1WM OWM_EN low (active) %02x\n", c & 0x7f);
+}
+
+static void ds1wm_disable(struct platform_device *pdev)
+{
+	struct device *dev = pdev->dev.parent;
+	int c;
+
+	c = pasic3_read_register(dev, 0x28);
+	pasic3_write_register(dev, 0x28, c | 0x80);
+
+	dev_dbg(dev, "DS1WM OWM_EN high (inactive) %02x\n", c | 0x80);
+}
+
+static struct ds1wm_platform_data ds1wm_pdata = {
+	.bus_shift = 2,
+	.enable    = ds1wm_enable,
+	.disable   = ds1wm_disable,
+};
+
+static int ds1wm_device_add(struct device *pasic3_dev, int bus_shift)
+{
+	struct pasic3_data *asic = pasic3_dev->driver_data;
+	struct platform_device *pdev;
+	int ret;
+
+	pdev = platform_device_alloc("ds1wm", -1);
+	if (!pdev) {
+		dev_dbg(pasic3_dev, "failed to allocate DS1WM platform device\n");
+		return -ENOMEM;
+	}
+
+	ret = platform_device_add_resources(pdev, pdev->resource,
+						pdev->num_resources);
+	if (ret < 0) {
+		dev_dbg(pasic3_dev, "failed to add DS1WM resources\n");
+		goto exit_pdev_put;
+	}
+
+	ds1wm_pdata.bus_shift = asic->bus_shift;
+	ret = platform_device_add_data(pdev, &ds1wm_pdata,
+					sizeof(struct ds1wm_platform_data));
+	if (ret < 0) {
+		dev_dbg(pasic3_dev, "failed to add DS1WM platform data\n");
+		goto exit_pdev_put;
+	}
+
+	pdev->dev.parent = pasic3_dev;
+	ret = platform_device_add(pdev);
+	if (ret < 0) {
+		dev_dbg(pasic3_dev, "failed to add DS1WM platform device\n");
+		goto exit_pdev_put;
+	}
+
+	asic->ds1wm_pdev = pdev;
+	return 0;
+
+exit_pdev_put:
+	platform_device_put(pdev);
+	return ret;
+}
+
+static int __init pasic3_probe(struct platform_device *pdev)
+{
+	struct pasic3_platform_data *pdata = pdev->dev.platform_data;
+	struct device *dev = &pdev->dev;
+	struct pasic3_data *asic;
+	struct resource *r;
+	int ret;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!r)
+		return -ENXIO;
+
+	if (!request_mem_region(r->start, r->end - r->start + 1, "pasic3"))
+		return -EBUSY;
+
+	asic = kzalloc(sizeof(struct pasic3_data), GFP_KERNEL);
+	if (!asic)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, asic);
+
+	if (pdata && pdata->bus_shift)
+		asic->bus_shift = pdata->bus_shift;
+	else
+		asic->bus_shift = 2;
+
+	asic->mapping = ioremap(r->start, r->end - r->start + 1);
+	if (!asic->mapping) {
+		dev_err(dev, "couldn't ioremap PASIC3\n");
+		kfree(asic);
+		return -ENOMEM;
+	}
+
+	ret = ds1wm_device_add(dev, asic->bus_shift);
+	if (ret < 0)
+		dev_warn(dev, "failed to register DS1WM\n");
+
+	if (pdata->led_pdata) {
+		ret = led_device_add(dev, pdata->led_pdata);
+		if (ret < 0)
+			dev_warn(dev, "failed to register LED device\n");
+	}
+
+	return 0;
+}
+
+static int pasic3_remove(struct platform_device *pdev)
+{
+	struct pasic3_data *asic = platform_get_drvdata(pdev);
+	struct resource *r;
+
+	if (asic->led_pdev)
+		platform_device_unregister(asic->led_pdev);
+	if (asic->ds1wm_pdev)
+		platform_device_unregister(asic->ds1wm_pdev);
+
+	iounmap(asic->mapping);
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	release_mem_region(r->start, r->end - r->start + 1);
+	kfree(asic);
+	return 0;
+}
+
+static struct platform_driver pasic3_driver = {
+	.driver		= {
+		.name	= "pasic3",
+	},
+	.remove		= pasic3_remove,
+};
+
+static int __init pasic3_base_init(void)
+{
+	return platform_driver_probe(&pasic3_driver, pasic3_probe);
+}
+
+static void __exit pasic3_base_exit(void)
+{
+	platform_driver_unregister(&pasic3_driver);
+}
+
+module_init(pasic3_base_init);
+module_exit(pasic3_base_exit);
+
+MODULE_AUTHOR("Philipp Zabel <philipp.zabel@gmail.com>");
+MODULE_DESCRIPTION("Core driver for HTC PASIC3");
+MODULE_LICENSE("GPL");
diff --git a/include/linux/mfd/htc-pasic3.h b/include/linux/mfd/htc-pasic3.h
new file mode 100644
index 000000000000..b4294f12c4f8
--- /dev/null
+++ b/include/linux/mfd/htc-pasic3.h
@@ -0,0 +1,55 @@
+/*
+ * HTC PASIC3 driver - LEDs and DS1WM
+ *
+ * Copyright (c) 2007 Philipp Zabel <philipp.zabel@gmail.com>
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive for
+ * more details.
+ *
+ */
+
+#ifndef __PASIC3_H
+#define __PASIC3_H
+
+#include <linux/platform_device.h>
+#include <linux/leds.h>
+
+extern void pasic3_write_register(struct device *dev, u32 reg, u8 val);
+extern u8 pasic3_read_register(struct device *dev, u32 reg);
+
+/*
+ * mask for registers 0x20,0x21,0x22
+ */
+#define PASIC3_MASK_LED0 0x04
+#define PASIC3_MASK_LED1 0x08
+#define PASIC3_MASK_LED2 0x40
+
+/*
+ * bits in register 0x06
+ */
+#define PASIC3_BIT2_LED0 0x08
+#define PASIC3_BIT2_LED1 0x10
+#define PASIC3_BIT2_LED2 0x20
+
+struct pasic3_led {
+	struct led_classdev         led;
+	unsigned int                hw_num;
+	unsigned int                bit2;
+	unsigned int                mask;
+	struct pasic3_leds_machinfo *pdata;
+};
+
+struct pasic3_leds_machinfo {
+	unsigned int      num_leds;
+	unsigned int      power_gpio;
+	struct pasic3_led *leds;
+};
+
+struct pasic3_platform_data {
+	struct pasic3_leds_machinfo *led_pdata;
+	unsigned int                 bus_shift;
+	unsigned int                 clock_rate;
+};
+
+#endif
-- 
cgit v1.2.3


From 4a3575fd436aa98957184afd745e4ada8f1542d8 Mon Sep 17 00:00:00 2001
From: "Huang, Ying" <ying.huang@intel.com>
Date: Mon, 25 Feb 2008 15:18:37 +0800
Subject: x86: EFI_PAGE_SHIFT fix

Make x86 EFI code works when EFI_PAGE_SHIFT != PAGE_SHIFT. The
memrage_efi_to_native() provided in this patch can be used on other
EFI platform such as IA64 too.

This patch has been tested on Intel x86_64 platform with EFI 64/32
firmware.

Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 arch/x86/kernel/efi.c    | 18 +++++++++++++-----
 arch/x86/kernel/efi_64.c | 12 ++++++------
 include/linux/efi.h      |  7 +++++++
 3 files changed, 26 insertions(+), 11 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/kernel/efi.c b/arch/x86/kernel/efi.c
index 759e02bec070..77d424cf68b3 100644
--- a/arch/x86/kernel/efi.c
+++ b/arch/x86/kernel/efi.c
@@ -383,6 +383,7 @@ static void __init runtime_code_page_mkexec(void)
 {
 	efi_memory_desc_t *md;
 	void *p;
+	u64 addr, npages;
 
 	/* Make EFI runtime service code area executable */
 	for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) {
@@ -391,7 +392,10 @@ static void __init runtime_code_page_mkexec(void)
 		if (md->type != EFI_RUNTIME_SERVICES_CODE)
 			continue;
 
-		set_memory_x(md->virt_addr, md->num_pages);
+		addr = md->virt_addr;
+		npages = md->num_pages;
+		memrange_efi_to_native(&addr, &npages);
+		set_memory_x(addr, npages);
 	}
 }
 
@@ -408,7 +412,7 @@ void __init efi_enter_virtual_mode(void)
 	efi_memory_desc_t *md;
 	efi_status_t status;
 	unsigned long size;
-	u64 end, systab;
+	u64 end, systab, addr, npages;
 	void *p, *va;
 
 	efi.systab = NULL;
@@ -420,7 +424,7 @@ void __init efi_enter_virtual_mode(void)
 		size = md->num_pages << EFI_PAGE_SHIFT;
 		end = md->phys_addr + size;
 
-		if ((end >> PAGE_SHIFT) <= max_pfn_mapped)
+		if (PFN_UP(end) <= max_pfn_mapped)
 			va = __va(md->phys_addr);
 		else
 			va = efi_ioremap(md->phys_addr, size);
@@ -433,8 +437,12 @@ void __init efi_enter_virtual_mode(void)
 			continue;
 		}
 
-		if (!(md->attribute & EFI_MEMORY_WB))
-			set_memory_uc(md->virt_addr, md->num_pages);
+		if (!(md->attribute & EFI_MEMORY_WB)) {
+			addr = md->virt_addr;
+			npages = md->num_pages;
+			memrange_efi_to_native(&addr, &npages);
+			set_memory_uc(addr, npages);
+		}
 
 		systab = (u64) (unsigned long) efi_phys.systab;
 		if (md->phys_addr <= systab && systab < end) {
diff --git a/arch/x86/kernel/efi_64.c b/arch/x86/kernel/efi_64.c
index d143a1e76b30..d0060fdcccac 100644
--- a/arch/x86/kernel/efi_64.c
+++ b/arch/x86/kernel/efi_64.c
@@ -105,14 +105,14 @@ void __init efi_reserve_bootmem(void)
 
 void __iomem * __init efi_ioremap(unsigned long phys_addr, unsigned long size)
 {
-	static unsigned pages_mapped;
+	static unsigned pages_mapped __initdata;
 	unsigned i, pages;
+	unsigned long offset;
 
-	/* phys_addr and size must be page aligned */
-	if ((phys_addr & ~PAGE_MASK) || (size & ~PAGE_MASK))
-		return NULL;
+	pages = PFN_UP(phys_addr + size) - PFN_DOWN(phys_addr);
+	offset = phys_addr & ~PAGE_MASK;
+	phys_addr &= PAGE_MASK;
 
-	pages = size >> PAGE_SHIFT;
 	if (pages_mapped + pages > MAX_EFI_IO_PAGES)
 		return NULL;
 
@@ -124,5 +124,5 @@ void __iomem * __init efi_ioremap(unsigned long phys_addr, unsigned long size)
 	}
 
 	return (void __iomem *)__fix_to_virt(FIX_EFI_IO_MAP_FIRST_PAGE - \
-					     (pages_mapped - pages));
+					     (pages_mapped - pages)) + offset;
 }
diff --git a/include/linux/efi.h b/include/linux/efi.h
index 14813b595802..a5f359a7ad0e 100644
--- a/include/linux/efi.h
+++ b/include/linux/efi.h
@@ -18,6 +18,7 @@
 #include <linux/proc_fs.h>
 #include <linux/rtc.h>
 #include <linux/ioport.h>
+#include <linux/pfn.h>
 
 #include <asm/page.h>
 #include <asm/system.h>
@@ -394,4 +395,10 @@ struct efi_generic_dev_path {
 	u16 length;
 } __attribute ((packed));
 
+static inline void memrange_efi_to_native(u64 *addr, u64 *npages)
+{
+	*npages = PFN_UP(*addr + (*npages<<EFI_PAGE_SHIFT)) - PFN_DOWN(*addr);
+	*addr &= PAGE_MASK;
+}
+
 #endif /* _LINUX_EFI_H */
-- 
cgit v1.2.3


From a7d5ac87b220829bb077cdc8e01c4fd4714ae41e Mon Sep 17 00:00:00 2001
From: Harvey Harrison <harvey.harrison@gmail.com>
Date: Tue, 4 Mar 2008 22:05:27 -0800
Subject: x86: pageattr.c fix shadowed variable warning

irqs_disabled() uses flags internally, use _flags to avoid shadowing
code calling into this macro.

Introduced between 2.6.25-rc3 and -rc4

Fixes the sparse warning:
arch/x86/mm/pageattr.c:383:21: warning: symbol 'flags' shadows an earlier one
arch/x86/mm/pageattr.c:369:16: originally declared here

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/irqflags.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/irqflags.h b/include/linux/irqflags.h
index 412e025bc5c7..e600c4e9b8c5 100644
--- a/include/linux/irqflags.h
+++ b/include/linux/irqflags.h
@@ -84,10 +84,10 @@
 
 #define irqs_disabled()						\
 ({								\
-	unsigned long flags;					\
+	unsigned long _flags;					\
 								\
-	raw_local_save_flags(flags);				\
-	raw_irqs_disabled_flags(flags);				\
+	raw_local_save_flags(_flags);				\
+	raw_irqs_disabled_flags(_flags);			\
 })
 
 #define irqs_disabled_flags(flags)	raw_irqs_disabled_flags(flags)
-- 
cgit v1.2.3


From 8fb402bccf203ecca8f9e0202b8fd3c937dece6f Mon Sep 17 00:00:00 2001
From: Erik Bosman <ebn310@few.vu.nl>
Date: Fri, 11 Apr 2008 18:54:17 +0200
Subject: generic, x86: add prctl commands PR_GET_TSC and PR_SET_TSC

This patch adds prctl commands that make it possible
to deny the execution of timestamp counters in userspace.
If this is not implemented on a specific architecture,
prctl will return -EINVAL.

ned-off-by: Erik Bosman <ejbosman@cs.vu.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/prctl.h |  6 ++++++
 kernel/sys.c          | 13 ++++++++++++-
 2 files changed, 18 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/prctl.h b/include/linux/prctl.h
index 3800639775ae..5c80b1939636 100644
--- a/include/linux/prctl.h
+++ b/include/linux/prctl.h
@@ -67,4 +67,10 @@
 #define PR_CAPBSET_READ 23
 #define PR_CAPBSET_DROP 24
 
+/* Get/set the process' ability to use the timestamp counter instruction */
+#define PR_GET_TSC 25
+#define PR_SET_TSC 26
+# define PR_TSC_ENABLE		1	/* allow the use of the timestamp counter */
+# define PR_TSC_SIGSEGV		2	/* throw a SIGSEGV instead of reading the TSC */
+
 #endif /* _LINUX_PRCTL_H */
diff --git a/kernel/sys.c b/kernel/sys.c
index a626116af5db..6a0cc71ee88d 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -67,6 +67,12 @@
 #ifndef SET_ENDIAN
 # define SET_ENDIAN(a,b)	(-EINVAL)
 #endif
+#ifndef GET_TSC_CTL
+# define GET_TSC_CTL(a)		(-EINVAL)
+#endif
+#ifndef SET_TSC_CTL
+# define SET_TSC_CTL(a)		(-EINVAL)
+#endif
 
 /*
  * this is where the system-wide overflow UID and GID are defined, for
@@ -1737,7 +1743,12 @@ asmlinkage long sys_prctl(int option, unsigned long arg2, unsigned long arg3,
 #else
 			return -EINVAL;
 #endif
-
+		case PR_GET_TSC:
+			error = GET_TSC_CTL(arg2);
+			break;
+		case PR_SET_TSC:
+			error = SET_TSC_CTL(arg2);
+			break;
 		default:
 			error = -EINVAL;
 			break;
-- 
cgit v1.2.3


From 50df5d6aea6694ca481b8005900401e8c95c2603 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Fri, 14 Mar 2008 16:09:59 +0100
Subject: sched: remove sysctl_sched_batch_wakeup_granularity

it's unused.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |  1 -
 kernel/sched.c        |  1 -
 kernel/sched_debug.c  |  1 -
 kernel/sched_fair.c   | 10 ----------
 kernel/sysctl.c       | 11 -----------
 5 files changed, 24 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 6a1e7afb099b..15f05ff453d8 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1551,7 +1551,6 @@ static inline void wake_up_idle_cpu(int cpu) { }
 extern unsigned int sysctl_sched_latency;
 extern unsigned int sysctl_sched_min_granularity;
 extern unsigned int sysctl_sched_wakeup_granularity;
-extern unsigned int sysctl_sched_batch_wakeup_granularity;
 extern unsigned int sysctl_sched_child_runs_first;
 extern unsigned int sysctl_sched_features;
 extern unsigned int sysctl_sched_migration_cost;
diff --git a/kernel/sched.c b/kernel/sched.c
index 770449bee6da..e813e845d9cf 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -5396,7 +5396,6 @@ static inline void sched_init_granularity(void)
 		sysctl_sched_latency = limit;
 
 	sysctl_sched_wakeup_granularity *= factor;
-	sysctl_sched_batch_wakeup_granularity *= factor;
 }
 
 #ifdef CONFIG_SMP
diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c
index ef358ba07683..3d09106990cb 100644
--- a/kernel/sched_debug.c
+++ b/kernel/sched_debug.c
@@ -214,7 +214,6 @@ static int sched_debug_show(struct seq_file *m, void *v)
 	PN(sysctl_sched_latency);
 	PN(sysctl_sched_min_granularity);
 	PN(sysctl_sched_wakeup_granularity);
-	PN(sysctl_sched_batch_wakeup_granularity);
 	PN(sysctl_sched_child_runs_first);
 	P(sysctl_sched_features);
 #undef PN
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index b01f8e77f2ac..bedda18f37a5 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -61,16 +61,6 @@ const_debug unsigned int sysctl_sched_child_runs_first = 1;
  */
 unsigned int __read_mostly sysctl_sched_compat_yield;
 
-/*
- * SCHED_BATCH wake-up granularity.
- * (default: 10 msec * (1 + ilog(ncpus)), units: nanoseconds)
- *
- * This option delays the preemption effects of decoupled workloads
- * and reduces their over-scheduling. Synchronous workloads will still
- * have immediate wakeup/sleep latencies.
- */
-unsigned int sysctl_sched_batch_wakeup_granularity = 10000000UL;
-
 /*
  * SCHED_OTHER wake-up granularity.
  * (default: 10 msec * (1 + ilog(ncpus)), units: nanoseconds)
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index b2a2d6889bab..be332e1a0c29 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -268,17 +268,6 @@ static struct ctl_table kern_table[] = {
 		.extra1		= &min_wakeup_granularity_ns,
 		.extra2		= &max_wakeup_granularity_ns,
 	},
-	{
-		.ctl_name	= CTL_UNNUMBERED,
-		.procname	= "sched_batch_wakeup_granularity_ns",
-		.data		= &sysctl_sched_batch_wakeup_granularity,
-		.maxlen		= sizeof(unsigned int),
-		.mode		= 0644,
-		.proc_handler	= &proc_dointvec_minmax,
-		.strategy	= &sysctl_intvec,
-		.extra1		= &min_wakeup_granularity_ns,
-		.extra2		= &max_wakeup_granularity_ns,
-	},
 	{
 		.ctl_name	= CTL_UNNUMBERED,
 		.procname	= "sched_child_runs_first",
-- 
cgit v1.2.3


From 57d3da2911787a101a384532f4519f9640bae883 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Wed, 27 Feb 2008 14:05:10 +0100
Subject: time: add ns_to_ktime()

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/ktime.h | 6 ++++++
 1 file changed, 6 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/ktime.h b/include/linux/ktime.h
index 2cd7fa73d1af..ce5983225be4 100644
--- a/include/linux/ktime.h
+++ b/include/linux/ktime.h
@@ -327,4 +327,10 @@ extern void ktime_get_ts(struct timespec *ts);
 /* Get the real (wall-) time in timespec format: */
 #define ktime_get_real_ts(ts)	getnstimeofday(ts)
 
+static inline ktime_t ns_to_ktime(u64 ns)
+{
+	static const ktime_t ktime_zero = { .tv64 = 0 };
+	return ktime_add_ns(ktime_zero, ns);
+}
+
 #endif
-- 
cgit v1.2.3


From d0b27fa77854b149ad4af08b0fe47fe712a47ade Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Sat, 19 Apr 2008 19:44:57 +0200
Subject: sched: rt-group: synchonised bandwidth period

Various SMP balancing algorithms require that the bandwidth period
run in sync.

Possible improvements are moving the rt_bandwidth thing into root_domain
and keeping a span per rt_bandwidth which marks throttled cpus.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h    |   7 ++
 kernel/sched.c           | 260 ++++++++++++++++++++++++++++++++++++++---------
 kernel/sched_rt.c        | 104 +++++++++++++------
 kernel/sysctl.c          |   4 +-
 kernel/time/tick-sched.c |   5 -
 kernel/user.c            |  28 +++++
 6 files changed, 320 insertions(+), 88 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 15f05ff453d8..be5d31752dbd 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1563,6 +1563,10 @@ int sched_nr_latency_handler(struct ctl_table *table, int write,
 extern unsigned int sysctl_sched_rt_period;
 extern int sysctl_sched_rt_runtime;
 
+int sched_rt_handler(struct ctl_table *table, int write,
+		struct file *filp, void __user *buffer, size_t *lenp,
+		loff_t *ppos);
+
 extern unsigned int sysctl_sched_compat_yield;
 
 #ifdef CONFIG_RT_MUTEXES
@@ -2052,6 +2056,9 @@ extern unsigned long sched_group_shares(struct task_group *tg);
 extern int sched_group_set_rt_runtime(struct task_group *tg,
 				      long rt_runtime_us);
 extern long sched_group_rt_runtime(struct task_group *tg);
+extern int sched_group_set_rt_period(struct task_group *tg,
+				      long rt_period_us);
+extern long sched_group_rt_period(struct task_group *tg);
 #endif
 #endif
 
diff --git a/kernel/sched.c b/kernel/sched.c
index e813e845d9cf..bb20323f7d09 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -115,6 +115,11 @@ unsigned long long __attribute__((weak)) sched_clock(void)
  */
 #define DEF_TIMESLICE		(100 * HZ / 1000)
 
+/*
+ * single value that denotes runtime == period, ie unlimited time.
+ */
+#define RUNTIME_INF	((u64)~0ULL)
+
 #ifdef CONFIG_SMP
 /*
  * Divide a load by a sched group cpu_power : (load / sg->__cpu_power)
@@ -156,6 +161,80 @@ struct rt_prio_array {
 	struct list_head queue[MAX_RT_PRIO];
 };
 
+struct rt_bandwidth {
+	ktime_t rt_period;
+	u64 rt_runtime;
+	struct hrtimer rt_period_timer;
+};
+
+static struct rt_bandwidth def_rt_bandwidth;
+
+static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun);
+
+static enum hrtimer_restart sched_rt_period_timer(struct hrtimer *timer)
+{
+	struct rt_bandwidth *rt_b =
+		container_of(timer, struct rt_bandwidth, rt_period_timer);
+	ktime_t now;
+	int overrun;
+	int idle = 0;
+
+	for (;;) {
+		now = hrtimer_cb_get_time(timer);
+		overrun = hrtimer_forward(timer, now, rt_b->rt_period);
+
+		if (!overrun)
+			break;
+
+		idle = do_sched_rt_period_timer(rt_b, overrun);
+	}
+
+	return idle ? HRTIMER_NORESTART : HRTIMER_RESTART;
+}
+
+static
+void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime)
+{
+	rt_b->rt_period = ns_to_ktime(period);
+	rt_b->rt_runtime = runtime;
+
+	hrtimer_init(&rt_b->rt_period_timer,
+			CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+	rt_b->rt_period_timer.function = sched_rt_period_timer;
+	rt_b->rt_period_timer.cb_mode = HRTIMER_CB_IRQSAFE_NO_SOFTIRQ;
+}
+
+static void start_rt_bandwidth(struct rt_bandwidth *rt_b)
+{
+	ktime_t now;
+
+	if (rt_b->rt_runtime == RUNTIME_INF)
+		return;
+
+	if (hrtimer_active(&rt_b->rt_period_timer))
+		return;
+
+	spin_lock(&rt_b->rt_runtime_lock);
+	for (;;) {
+		if (hrtimer_active(&rt_b->rt_period_timer))
+			break;
+
+		now = hrtimer_cb_get_time(&rt_b->rt_period_timer);
+		hrtimer_forward(&rt_b->rt_period_timer, now, rt_b->rt_period);
+		hrtimer_start(&rt_b->rt_period_timer,
+			      rt_b->rt_period_timer.expires,
+			      HRTIMER_MODE_ABS);
+	}
+	spin_unlock(&rt_b->rt_runtime_lock);
+}
+
+#ifdef CONFIG_RT_GROUP_SCHED
+static void destroy_rt_bandwidth(struct rt_bandwidth *rt_b)
+{
+	hrtimer_cancel(&rt_b->rt_period_timer);
+}
+#endif
+
 #ifdef CONFIG_GROUP_SCHED
 
 #include <linux/cgroup.h>
@@ -182,7 +261,7 @@ struct task_group {
 	struct sched_rt_entity **rt_se;
 	struct rt_rq **rt_rq;
 
-	u64 rt_runtime;
+	struct rt_bandwidth rt_bandwidth;
 #endif
 
 	struct rcu_head rcu;
@@ -407,8 +486,6 @@ struct rq {
 
 	struct cfs_rq cfs;
 	struct rt_rq rt;
-	u64 rt_period_expire;
-	int rt_throttled;
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
 	/* list of leaf cfs_rq on this cpu: */
@@ -592,23 +669,6 @@ static void update_rq_clock(struct rq *rq)
 #define task_rq(p)		cpu_rq(task_cpu(p))
 #define cpu_curr(cpu)		(cpu_rq(cpu)->curr)
 
-unsigned long rt_needs_cpu(int cpu)
-{
-	struct rq *rq = cpu_rq(cpu);
-	u64 delta;
-
-	if (!rq->rt_throttled)
-		return 0;
-
-	if (rq->clock > rq->rt_period_expire)
-		return 1;
-
-	delta = rq->rt_period_expire - rq->clock;
-	do_div(delta, NSEC_PER_SEC / HZ);
-
-	return (unsigned long)delta;
-}
-
 /*
  * Tunables that become constants when CONFIG_SCHED_DEBUG is off:
  */
@@ -664,10 +724,18 @@ static __read_mostly int scheduler_running;
  */
 int sysctl_sched_rt_runtime = 950000;
 
-/*
- * single value that denotes runtime == period, ie unlimited time.
- */
-#define RUNTIME_INF	((u64)~0ULL)
+static inline u64 global_rt_period(void)
+{
+	return (u64)sysctl_sched_rt_period * NSEC_PER_USEC;
+}
+
+static inline u64 global_rt_runtime(void)
+{
+	if (sysctl_sched_rt_period < 0)
+		return RUNTIME_INF;
+
+	return (u64)sysctl_sched_rt_runtime * NSEC_PER_USEC;
+}
 
 static const unsigned long long time_sync_thresh = 100000;
 
@@ -3854,7 +3922,6 @@ void scheduler_tick(void)
 	update_last_tick_seen(rq);
 	update_cpu_load(rq);
 	curr->sched_class->task_tick(rq, curr, 0);
-	update_sched_rt_period(rq);
 	spin_unlock(&rq->lock);
 
 #ifdef CONFIG_SMP
@@ -4689,7 +4756,7 @@ recheck:
 	 * Do not allow realtime tasks into groups that have no runtime
 	 * assigned.
 	 */
-	if (rt_policy(policy) && task_group(p)->rt_runtime == 0)
+	if (rt_policy(policy) && task_group(p)->rt_bandwidth.rt_runtime == 0)
 		return -EPERM;
 #endif
 
@@ -7288,6 +7355,14 @@ void __init sched_init(void)
 	init_defrootdomain();
 #endif
 
+	init_rt_bandwidth(&def_rt_bandwidth,
+			global_rt_period(), global_rt_runtime());
+
+#ifdef CONFIG_RT_GROUP_SCHED
+	init_rt_bandwidth(&init_task_group.rt_bandwidth,
+			global_rt_period(), global_rt_runtime());
+#endif
+
 #ifdef CONFIG_GROUP_SCHED
 	list_add(&init_task_group.list, &task_groups);
 #endif
@@ -7312,15 +7387,11 @@ void __init sched_init(void)
 
 #endif
 #ifdef CONFIG_RT_GROUP_SCHED
-		init_task_group.rt_runtime =
-			sysctl_sched_rt_runtime * NSEC_PER_USEC;
 		INIT_LIST_HEAD(&rq->leaf_rt_rq_list);
 		init_tg_rt_entry(rq, &init_task_group,
 				&per_cpu(init_rt_rq, i),
 				&per_cpu(init_sched_rt_entity, i), i, 1);
 #endif
-		rq->rt_period_expire = 0;
-		rq->rt_throttled = 0;
 
 		for (j = 0; j < CPU_LOAD_IDX_MAX; j++)
 			rq->cpu_load[j] = 0;
@@ -7506,8 +7577,6 @@ void set_curr_task(int cpu, struct task_struct *p)
 
 #endif
 
-#ifdef CONFIG_GROUP_SCHED
-
 #ifdef CONFIG_FAIR_GROUP_SCHED
 static void free_fair_sched_group(struct task_group *tg)
 {
@@ -7596,6 +7665,8 @@ static void free_rt_sched_group(struct task_group *tg)
 {
 	int i;
 
+	destroy_rt_bandwidth(&tg->rt_bandwidth);
+
 	for_each_possible_cpu(i) {
 		if (tg->rt_rq)
 			kfree(tg->rt_rq[i]);
@@ -7621,7 +7692,8 @@ static int alloc_rt_sched_group(struct task_group *tg)
 	if (!tg->rt_se)
 		goto err;
 
-	tg->rt_runtime = 0;
+	init_rt_bandwidth(&tg->rt_bandwidth,
+			ktime_to_ns(def_rt_bandwidth.rt_period), 0);
 
 	for_each_possible_cpu(i) {
 		rq = cpu_rq(i);
@@ -7674,6 +7746,7 @@ static inline void unregister_rt_sched_group(struct task_group *tg, int cpu)
 }
 #endif
 
+#ifdef CONFIG_GROUP_SCHED
 static void free_sched_group(struct task_group *tg)
 {
 	free_fair_sched_group(tg);
@@ -7775,6 +7848,7 @@ void sched_move_task(struct task_struct *tsk)
 
 	task_rq_unlock(rq, &flags);
 }
+#endif
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
 static void set_se_shares(struct sched_entity *se, unsigned long shares)
@@ -7871,16 +7945,15 @@ static int __rt_schedulable(struct task_group *tg, u64 period, u64 runtime)
 	struct task_group *tgi;
 	unsigned long total = 0;
 	unsigned long global_ratio =
-		to_ratio(sysctl_sched_rt_period,
-			 sysctl_sched_rt_runtime < 0 ?
-				RUNTIME_INF : sysctl_sched_rt_runtime);
+		to_ratio(global_rt_period(), global_rt_runtime());
 
 	rcu_read_lock();
 	list_for_each_entry_rcu(tgi, &task_groups, list) {
 		if (tgi == tg)
 			continue;
 
-		total += to_ratio(period, tgi->rt_runtime);
+		total += to_ratio(ktime_to_ns(tgi->rt_bandwidth.rt_period),
+				tgi->rt_bandwidth.rt_runtime);
 	}
 	rcu_read_unlock();
 
@@ -7898,16 +7971,11 @@ static inline int tg_has_rt_tasks(struct task_group *tg)
 	return 0;
 }
 
-int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
+static int tg_set_bandwidth(struct task_group *tg,
+		u64 rt_period, u64 rt_runtime)
 {
-	u64 rt_runtime, rt_period;
 	int err = 0;
 
-	rt_period = (u64)sysctl_sched_rt_period * NSEC_PER_USEC;
-	rt_runtime = (u64)rt_runtime_us * NSEC_PER_USEC;
-	if (rt_runtime_us == -1)
-		rt_runtime = RUNTIME_INF;
-
 	mutex_lock(&rt_constraints_mutex);
 	read_lock(&tasklist_lock);
 	if (rt_runtime_us == 0 && tg_has_rt_tasks(tg)) {
@@ -7918,7 +7986,8 @@ int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
 		err = -EINVAL;
 		goto unlock;
 	}
-	tg->rt_runtime = rt_runtime;
+	tg->rt_bandwidth.rt_period = ns_to_ktime(rt_period);
+	tg->rt_bandwidth.rt_runtime = rt_runtime;
  unlock:
 	read_unlock(&tasklist_lock);
 	mutex_unlock(&rt_constraints_mutex);
@@ -7926,19 +7995,96 @@ int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
 	return err;
 }
 
+int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
+{
+	u64 rt_runtime, rt_period;
+
+	rt_period = ktime_to_ns(tg->rt_bandwidth.rt_period);
+	rt_runtime = (u64)rt_runtime_us * NSEC_PER_USEC;
+	if (rt_runtime_us < 0)
+		rt_runtime = RUNTIME_INF;
+
+	return tg_set_bandwidth(tg, rt_period, rt_runtime);
+}
+
 long sched_group_rt_runtime(struct task_group *tg)
 {
 	u64 rt_runtime_us;
 
-	if (tg->rt_runtime == RUNTIME_INF)
+	if (tg->rt_bandwidth.rt_runtime == RUNTIME_INF)
 		return -1;
 
-	rt_runtime_us = tg->rt_runtime;
+	rt_runtime_us = tg->rt_bandwidth.rt_runtime;
 	do_div(rt_runtime_us, NSEC_PER_USEC);
 	return rt_runtime_us;
 }
+
+int sched_group_set_rt_period(struct task_group *tg, long rt_period_us)
+{
+	u64 rt_runtime, rt_period;
+
+	rt_period = (u64)rt_period_us * NSEC_PER_USEC;
+	rt_runtime = tg->rt_bandwidth.rt_runtime;
+
+	return tg_set_bandwidth(tg, rt_period, rt_runtime);
+}
+
+long sched_group_rt_period(struct task_group *tg)
+{
+	u64 rt_period_us;
+
+	rt_period_us = ktime_to_ns(tg->rt_bandwidth.rt_period);
+	do_div(rt_period_us, NSEC_PER_USEC);
+	return rt_period_us;
+}
+
+static int sched_rt_global_constraints(void)
+{
+	int ret = 0;
+
+	mutex_lock(&rt_constraints_mutex);
+	if (!__rt_schedulable(NULL, 1, 0))
+		ret = -EINVAL;
+	mutex_unlock(&rt_constraints_mutex);
+
+	return ret;
+}
+#else
+static int sched_rt_global_constraints(void)
+{
+	return 0;
+}
 #endif
-#endif	/* CONFIG_GROUP_SCHED */
+
+int sched_rt_handler(struct ctl_table *table, int write,
+		struct file *filp, void __user *buffer, size_t *lenp,
+		loff_t *ppos)
+{
+	int ret;
+	int old_period, old_runtime;
+	static DEFINE_MUTEX(mutex);
+
+	mutex_lock(&mutex);
+	old_period = sysctl_sched_rt_period;
+	old_runtime = sysctl_sched_rt_runtime;
+
+	ret = proc_dointvec(table, write, filp, buffer, lenp, ppos);
+
+	if (!ret && write) {
+		ret = sched_rt_global_constraints();
+		if (ret) {
+			sysctl_sched_rt_period = old_period;
+			sysctl_sched_rt_runtime = old_runtime;
+		} else {
+			def_rt_bandwidth.rt_runtime = global_rt_runtime();
+			def_rt_bandwidth.rt_period =
+				ns_to_ktime(global_rt_period());
+		}
+	}
+	mutex_unlock(&mutex);
+
+	return ret;
+}
 
 #ifdef CONFIG_CGROUP_SCHED
 
@@ -7988,7 +8134,7 @@ cpu_cgroup_can_attach(struct cgroup_subsys *ss, struct cgroup *cgrp,
 {
 #ifdef CONFIG_RT_GROUP_SCHED
 	/* Don't accept realtime tasks when there is no way for them to run */
-	if (rt_task(tsk) && cgroup_tg(cgrp)->rt_runtime == 0)
+	if (rt_task(tsk) && cgroup_tg(cgrp)->rt_bandwidth.rt_runtime == 0)
 		return -EINVAL;
 #else
 	/* We don't support RT-tasks being in separate groups */
@@ -8066,6 +8212,17 @@ static ssize_t cpu_rt_runtime_read(struct cgroup *cgrp, struct cftype *cft,
 
 	return simple_read_from_buffer(buf, nbytes, ppos, tmp, len);
 }
+
+static int cpu_rt_period_write_uint(struct cgroup *cgrp, struct cftype *cftype,
+		u64 rt_period_us)
+{
+	return sched_group_set_rt_period(cgroup_tg(cgrp), rt_period_us);
+}
+
+static u64 cpu_rt_period_read_uint(struct cgroup *cgrp, struct cftype *cft)
+{
+	return sched_group_rt_period(cgroup_tg(cgrp));
+}
 #endif
 
 static struct cftype cpu_files[] = {
@@ -8082,6 +8239,11 @@ static struct cftype cpu_files[] = {
 		.read = cpu_rt_runtime_read,
 		.write = cpu_rt_runtime_write,
 	},
+	{
+		.name = "rt_period_us",
+		.read_uint = cpu_rt_period_read_uint,
+		.write_uint = cpu_rt_period_write_uint,
+	},
 #endif
 };
 
diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 0a6d2e516420..8bc176136666 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -62,7 +62,7 @@ static inline u64 sched_rt_runtime(struct rt_rq *rt_rq)
 	if (!rt_rq->tg)
 		return RUNTIME_INF;
 
-	return rt_rq->tg->rt_runtime;
+	return rt_rq->tg->rt_bandwidth.rt_runtime;
 }
 
 #define for_each_leaf_rt_rq(rt_rq, rq) \
@@ -127,14 +127,29 @@ static int rt_se_boosted(struct sched_rt_entity *rt_se)
 	return p->prio != p->normal_prio;
 }
 
+#ifdef CONFIG_SMP
+static inline cpumask_t sched_rt_period_mask(void)
+{
+	return cpu_rq(smp_processor_id())->rd->span;
+}
 #else
+static inline cpumask_t sched_rt_period_mask(void)
+{
+	return cpu_online_map;
+}
+#endif
 
-static inline u64 sched_rt_runtime(struct rt_rq *rt_rq)
+static inline
+struct rt_rq *sched_rt_period_rt_rq(struct rt_bandwidth *rt_b, int cpu)
 {
-	if (sysctl_sched_rt_runtime == -1)
-		return RUNTIME_INF;
+	return container_of(rt_b, struct task_group, rt_bandwidth)->rt_rq[cpu];
+}
 
-	return (u64)sysctl_sched_rt_runtime * NSEC_PER_USEC;
+#else
+
+static inline u64 sched_rt_runtime(struct rt_rq *rt_rq)
+{
+	return def_rt_bandwidth.rt_runtime;
 }
 
 #define for_each_leaf_rt_rq(rt_rq, rq) \
@@ -173,8 +188,55 @@ static inline int rt_rq_throttled(struct rt_rq *rt_rq)
 {
 	return rt_rq->rt_throttled;
 }
+
+static inline cpumask_t sched_rt_period_mask(void)
+{
+	return cpu_online_map;
+}
+
+static inline
+struct rt_rq *sched_rt_period_rt_rq(struct rt_bandwidth *rt_b, int cpu)
+{
+	return &cpu_rq(cpu)->rt;
+}
+
 #endif
 
+static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
+{
+	int i, idle = 1;
+	cpumask_t span;
+
+	if (rt_b->rt_runtime == RUNTIME_INF)
+		return 1;
+
+	span = sched_rt_period_mask();
+	for_each_cpu_mask(i, span) {
+		int enqueue = 0;
+		struct rt_rq *rt_rq = sched_rt_period_rt_rq(rt_b, i);
+		struct rq *rq = rq_of_rt_rq(rt_rq);
+
+		spin_lock(&rq->lock);
+		if (rt_rq->rt_time) {
+			u64 runtime = rt_b->rt_runtime;
+
+			rt_rq->rt_time -= min(rt_rq->rt_time, overrun*runtime);
+			if (rt_rq->rt_throttled && rt_rq->rt_time < runtime) {
+				rt_rq->rt_throttled = 0;
+				enqueue = 1;
+			}
+			if (rt_rq->rt_time || rt_rq->rt_nr_running)
+				idle = 0;
+		}
+
+		if (enqueue)
+			sched_rt_rq_enqueue(rt_rq);
+		spin_unlock(&rq->lock);
+	}
+
+	return idle;
+}
+
 static inline int rt_se_prio(struct sched_rt_entity *rt_se)
 {
 #ifdef CONFIG_RT_GROUP_SCHED
@@ -198,11 +260,7 @@ static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq)
 		return rt_rq_throttled(rt_rq);
 
 	if (rt_rq->rt_time > runtime) {
-		struct rq *rq = rq_of_rt_rq(rt_rq);
-
-		rq->rt_throttled = 1;
 		rt_rq->rt_throttled = 1;
-
 		if (rt_rq_throttled(rt_rq)) {
 			sched_rt_rq_dequeue(rt_rq);
 			return 1;
@@ -212,29 +270,6 @@ static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq)
 	return 0;
 }
 
-static void update_sched_rt_period(struct rq *rq)
-{
-	struct rt_rq *rt_rq;
-	u64 period;
-
-	while (rq->clock > rq->rt_period_expire) {
-		period = (u64)sysctl_sched_rt_period * NSEC_PER_USEC;
-		rq->rt_period_expire += period;
-
-		for_each_leaf_rt_rq(rt_rq, rq) {
-			u64 runtime = sched_rt_runtime(rt_rq);
-
-			rt_rq->rt_time -= min(rt_rq->rt_time, runtime);
-			if (rt_rq->rt_throttled && rt_rq->rt_time < runtime) {
-				rt_rq->rt_throttled = 0;
-				sched_rt_rq_enqueue(rt_rq);
-			}
-		}
-
-		rq->rt_throttled = 0;
-	}
-}
-
 /*
  * Update the current task's runtime statistics. Skip current tasks that
  * are not in our scheduling class.
@@ -284,6 +319,11 @@ void inc_rt_tasks(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 #ifdef CONFIG_RT_GROUP_SCHED
 	if (rt_se_boosted(rt_se))
 		rt_rq->rt_nr_boosted++;
+
+	if (rt_rq->tg)
+		start_rt_bandwidth(&rt_rq->tg->rt_bandwidth);
+#else
+	start_rt_bandwidth(&def_rt_bandwidth);
 #endif
 }
 
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index be332e1a0c29..fd3364827ccf 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -307,7 +307,7 @@ static struct ctl_table kern_table[] = {
 		.data		= &sysctl_sched_rt_period,
 		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec,
+		.proc_handler	= &sched_rt_handler,
 	},
 	{
 		.ctl_name	= CTL_UNNUMBERED,
@@ -315,7 +315,7 @@ static struct ctl_table kern_table[] = {
 		.data		= &sysctl_sched_rt_runtime,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec,
+		.proc_handler	= &sched_rt_handler,
 	},
 	{
 		.ctl_name	= CTL_UNNUMBERED,
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 69dba0c71727..d358d4e3a958 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -191,7 +191,6 @@ u64 get_cpu_idle_time_us(int cpu, u64 *last_update_time)
 void tick_nohz_stop_sched_tick(void)
 {
 	unsigned long seq, last_jiffies, next_jiffies, delta_jiffies, flags;
-	unsigned long rt_jiffies;
 	struct tick_sched *ts;
 	ktime_t last_update, expires, now;
 	struct clock_event_device *dev = __get_cpu_var(tick_cpu_device).evtdev;
@@ -243,10 +242,6 @@ void tick_nohz_stop_sched_tick(void)
 	next_jiffies = get_next_timer_interrupt(last_jiffies);
 	delta_jiffies = next_jiffies - last_jiffies;
 
-	rt_jiffies = rt_needs_cpu(cpu);
-	if (rt_jiffies && rt_jiffies < delta_jiffies)
-		delta_jiffies = rt_jiffies;
-
 	if (rcu_needs_cpu(cpu))
 		delta_jiffies = 1;
 	/*
diff --git a/kernel/user.c b/kernel/user.c
index 7132022a040c..5925c6887c10 100644
--- a/kernel/user.c
+++ b/kernel/user.c
@@ -193,6 +193,33 @@ static ssize_t cpu_rt_runtime_store(struct kobject *kobj,
 
 static struct kobj_attribute cpu_rt_runtime_attr =
 	__ATTR(cpu_rt_runtime, 0644, cpu_rt_runtime_show, cpu_rt_runtime_store);
+
+static ssize_t cpu_rt_period_show(struct kobject *kobj,
+				   struct kobj_attribute *attr,
+				   char *buf)
+{
+	struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+
+	return sprintf(buf, "%lu\n", sched_group_rt_period(up->tg));
+}
+
+static ssize_t cpu_rt_period_store(struct kobject *kobj,
+				    struct kobj_attribute *attr,
+				    const char *buf, size_t size)
+{
+	struct user_struct *up = container_of(kobj, struct user_struct, kobj);
+	unsigned long rt_period;
+	int rc;
+
+	sscanf(buf, "%lu", &rt_period);
+
+	rc = sched_group_set_rt_period(up->tg, rt_period);
+
+	return (rc ? rc : size);
+}
+
+static struct kobj_attribute cpu_rt_period_attr =
+	__ATTR(cpu_rt_period, 0644, cpu_rt_period_show, cpu_rt_period_store);
 #endif
 
 /* default attributes per uid directory */
@@ -202,6 +229,7 @@ static struct attribute *uids_attributes[] = {
 #endif
 #ifdef CONFIG_RT_GROUP_SCHED
 	&cpu_rt_runtime_attr.attr,
+	&cpu_rt_period_attr.attr,
 #endif
 	NULL
 };
-- 
cgit v1.2.3


From 30ca60c15a725f655e5d3f14e0238a066bc5aeb7 Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Tue, 25 Mar 2008 15:06:55 -0700
Subject: cpumask: add cpumask_scnprintf_len function

Add a new function cpumask_scnprintf_len() to return the number of
characters needed to display "len" cpumask bits.  The current method
of allocating NR_CPUS bytes is incorrect as what's really needed is
9 characters per 32-bit word of cpumask bits (8 hex digits plus the
seperator [','] or the terminating NULL.)  This function provides the
caller the means to allocate the correct string length.

Cc: Paul Jackson <pj@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/bitmap.h  |  1 +
 include/linux/cpumask.h |  7 +++++++
 lib/bitmap.c            | 16 ++++++++++++++++
 3 files changed, 24 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/bitmap.h b/include/linux/bitmap.h
index acad1105d942..1dbe074f1c64 100644
--- a/include/linux/bitmap.h
+++ b/include/linux/bitmap.h
@@ -108,6 +108,7 @@ extern int __bitmap_weight(const unsigned long *bitmap, int bits);
 
 extern int bitmap_scnprintf(char *buf, unsigned int len,
 			const unsigned long *src, int nbits);
+extern int bitmap_scnprintf_len(unsigned int len);
 extern int __bitmap_parse(const char *buf, unsigned int buflen, int is_user,
 			unsigned long *dst, int nbits);
 extern int bitmap_parse_user(const char __user *ubuf, unsigned int ulen,
diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h
index 7047f58306a7..67e0e38d32b1 100644
--- a/include/linux/cpumask.h
+++ b/include/linux/cpumask.h
@@ -273,6 +273,13 @@ static inline int __cpumask_scnprintf(char *buf, int len,
 	return bitmap_scnprintf(buf, len, srcp->bits, nbits);
 }
 
+#define cpumask_scnprintf_len(len) \
+			__cpumask_scnprintf_len((len))
+static inline int __cpumask_scnprintf_len(int len)
+{
+	return bitmap_scnprintf_len(len);
+}
+
 #define cpumask_parse_user(ubuf, ulen, dst) \
 			__cpumask_parse_user((ubuf), (ulen), &(dst), NR_CPUS)
 static inline int __cpumask_parse_user(const char __user *buf, int len,
diff --git a/lib/bitmap.c b/lib/bitmap.c
index 2c9242e3fed0..a6939e18d7bb 100644
--- a/lib/bitmap.c
+++ b/lib/bitmap.c
@@ -315,6 +315,22 @@ int bitmap_scnprintf(char *buf, unsigned int buflen,
 }
 EXPORT_SYMBOL(bitmap_scnprintf);
 
+/**
+ * bitmap_scnprintf_len - return buffer length needed to convert
+ * bitmap to an ASCII hex string.
+ * @len: number of bits to be converted
+ */
+int bitmap_scnprintf_len(unsigned int len)
+{
+	/* we need 9 chars per word for 32 bit words (8 hexdigits + sep/null) */
+	int bitslen = ALIGN(len, CHUNKSZ);
+	int wordlen = CHUNKSZ / 4;
+	int buflen = (bitslen / wordlen) * (wordlen + 1) * sizeof(char);
+
+	return buflen;
+}
+EXPORT_SYMBOL(bitmap_scnprintf_len);
+
 /**
  * __bitmap_parse - convert an ASCII hex string into a bitmap.
  * @buf: pointer to buffer containing string.
-- 
cgit v1.2.3


From f9a86fcbbb1e5542eabf45c9144ac4b6330861a4 Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Fri, 4 Apr 2008 18:11:07 -0700
Subject: cpuset: modify cpuset_set_cpus_allowed to use cpumask pointer

  * Modify cpuset_cpus_allowed to return the currently allowed cpuset
    via a pointer argument instead of as the function return value.

  * Use new set_cpus_allowed_ptr function.

  * Cleanup CPU_MASK_ALL and NODE_MASK_ALL uses.

Depends on:
	[sched-devel]: sched: add new set_cpus_allowed_ptr function

Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/cpuset.h | 13 +++++++------
 kernel/cpuset.c        | 31 ++++++++++++-------------------
 kernel/sched.c         |  8 +++++---
 mm/pdflush.c           |  4 ++--
 4 files changed, 26 insertions(+), 30 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/cpuset.h b/include/linux/cpuset.h
index 0a26be353cb3..726761e24003 100644
--- a/include/linux/cpuset.h
+++ b/include/linux/cpuset.h
@@ -20,8 +20,8 @@ extern int number_of_cpusets;	/* How many cpusets are defined in system? */
 extern int cpuset_init_early(void);
 extern int cpuset_init(void);
 extern void cpuset_init_smp(void);
-extern cpumask_t cpuset_cpus_allowed(struct task_struct *p);
-extern cpumask_t cpuset_cpus_allowed_locked(struct task_struct *p);
+extern void cpuset_cpus_allowed(struct task_struct *p, cpumask_t *mask);
+extern void cpuset_cpus_allowed_locked(struct task_struct *p, cpumask_t *mask);
 extern nodemask_t cpuset_mems_allowed(struct task_struct *p);
 #define cpuset_current_mems_allowed (current->mems_allowed)
 void cpuset_init_current_mems_allowed(void);
@@ -84,13 +84,14 @@ static inline int cpuset_init_early(void) { return 0; }
 static inline int cpuset_init(void) { return 0; }
 static inline void cpuset_init_smp(void) {}
 
-static inline cpumask_t cpuset_cpus_allowed(struct task_struct *p)
+static inline void cpuset_cpus_allowed(struct task_struct *p, cpumask_t *mask)
 {
-	return cpu_possible_map;
+	*mask = cpu_possible_map;
 }
-static inline cpumask_t cpuset_cpus_allowed_locked(struct task_struct *p)
+static inline void cpuset_cpus_allowed_locked(struct task_struct *p,
+								cpumask_t *mask)
 {
-	return cpu_possible_map;
+	*mask = cpu_possible_map;
 }
 
 static inline nodemask_t cpuset_mems_allowed(struct task_struct *p)
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index a1b61f414228..6b9ac296a05c 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -729,7 +729,7 @@ int cpuset_test_cpumask(struct task_struct *tsk, struct cgroup_scanner *scan)
  */
 void cpuset_change_cpumask(struct task_struct *tsk, struct cgroup_scanner *scan)
 {
-	set_cpus_allowed(tsk, (cgroup_cs(scan->cg))->cpus_allowed);
+	set_cpus_allowed_ptr(tsk, &((cgroup_cs(scan->cg))->cpus_allowed));
 }
 
 /**
@@ -1178,7 +1178,7 @@ static void cpuset_attach(struct cgroup_subsys *ss,
 
 	mutex_lock(&callback_mutex);
 	guarantee_online_cpus(cs, &cpus);
-	set_cpus_allowed(tsk, cpus);
+	set_cpus_allowed_ptr(tsk, &cpus);
 	mutex_unlock(&callback_mutex);
 
 	from = oldcs->mems_allowed;
@@ -1555,8 +1555,8 @@ static struct cgroup_subsys_state *cpuset_create(
 	if (is_spread_slab(parent))
 		set_bit(CS_SPREAD_SLAB, &cs->flags);
 	set_bit(CS_SCHED_LOAD_BALANCE, &cs->flags);
-	cs->cpus_allowed = CPU_MASK_NONE;
-	cs->mems_allowed = NODE_MASK_NONE;
+	cpus_clear(cs->cpus_allowed);
+	nodes_clear(cs->mems_allowed);
 	cs->mems_generation = cpuset_mems_generation++;
 	fmeter_init(&cs->fmeter);
 
@@ -1625,8 +1625,8 @@ int __init cpuset_init(void)
 {
 	int err = 0;
 
-	top_cpuset.cpus_allowed = CPU_MASK_ALL;
-	top_cpuset.mems_allowed = NODE_MASK_ALL;
+	cpus_setall(top_cpuset.cpus_allowed);
+	nodes_setall(top_cpuset.mems_allowed);
 
 	fmeter_init(&top_cpuset.fmeter);
 	top_cpuset.mems_generation = cpuset_mems_generation++;
@@ -1844,6 +1844,7 @@ void __init cpuset_init_smp(void)
 
  * cpuset_cpus_allowed - return cpus_allowed mask from a tasks cpuset.
  * @tsk: pointer to task_struct from which to obtain cpuset->cpus_allowed.
+ * @pmask: pointer to cpumask_t variable to receive cpus_allowed set.
  *
  * Description: Returns the cpumask_t cpus_allowed of the cpuset
  * attached to the specified @tsk.  Guaranteed to return some non-empty
@@ -1851,35 +1852,27 @@ void __init cpuset_init_smp(void)
  * tasks cpuset.
  **/
 
-cpumask_t cpuset_cpus_allowed(struct task_struct *tsk)
+void cpuset_cpus_allowed(struct task_struct *tsk, cpumask_t *pmask)
 {
-	cpumask_t mask;
-
 	mutex_lock(&callback_mutex);
-	mask = cpuset_cpus_allowed_locked(tsk);
+	cpuset_cpus_allowed_locked(tsk, pmask);
 	mutex_unlock(&callback_mutex);
-
-	return mask;
 }
 
 /**
  * cpuset_cpus_allowed_locked - return cpus_allowed mask from a tasks cpuset.
  * Must be called with callback_mutex held.
  **/
-cpumask_t cpuset_cpus_allowed_locked(struct task_struct *tsk)
+void cpuset_cpus_allowed_locked(struct task_struct *tsk, cpumask_t *pmask)
 {
-	cpumask_t mask;
-
 	task_lock(tsk);
-	guarantee_online_cpus(task_cs(tsk), &mask);
+	guarantee_online_cpus(task_cs(tsk), pmask);
 	task_unlock(tsk);
-
-	return mask;
 }
 
 void cpuset_init_current_mems_allowed(void)
 {
-	current->mems_allowed = NODE_MASK_ALL;
+	nodes_setall(current->mems_allowed);
 }
 
 /**
diff --git a/kernel/sched.c b/kernel/sched.c
index ef3f28b334ea..ccc23a9cd264 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -4941,13 +4941,13 @@ long sched_setaffinity(pid_t pid, cpumask_t new_mask)
 	if (retval)
 		goto out_unlock;
 
-	cpus_allowed = cpuset_cpus_allowed(p);
+	cpuset_cpus_allowed(p, &cpus_allowed);
 	cpus_and(new_mask, new_mask, cpus_allowed);
  again:
 	retval = set_cpus_allowed(p, new_mask);
 
 	if (!retval) {
-		cpus_allowed = cpuset_cpus_allowed(p);
+		cpuset_cpus_allowed(p, &cpus_allowed);
 		if (!cpus_subset(new_mask, cpus_allowed)) {
 			/*
 			 * We must have raced with a concurrent cpuset
@@ -5661,7 +5661,9 @@ static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p)
 
 		/* No more Mr. Nice Guy. */
 		if (dest_cpu >= nr_cpu_ids) {
-			cpumask_t cpus_allowed = cpuset_cpus_allowed_locked(p);
+			cpumask_t cpus_allowed;
+
+			cpuset_cpus_allowed_locked(p, &cpus_allowed);
 			/*
 			 * Try to stay on the same cpuset, where the
 			 * current cpuset may be a subset of all cpus.
diff --git a/mm/pdflush.c b/mm/pdflush.c
index 8f6ee073c0e3..0ceacff56457 100644
--- a/mm/pdflush.c
+++ b/mm/pdflush.c
@@ -187,8 +187,8 @@ static int pdflush(void *dummy)
 	 * This is needed as pdflush's are dynamically created and destroyed.
 	 * The boottime pdflush's are easily placed w/o these 2 lines.
 	 */
-	cpus_allowed = cpuset_cpus_allowed(current);
-	set_cpus_allowed(current, cpus_allowed);
+	cpuset_cpus_allowed(current, &cpus_allowed);
+	set_cpus_allowed_ptr(current, &cpus_allowed);
 
 	return __pdflush(&my_work);
 }
-- 
cgit v1.2.3


From b53e921ba1cff8453dc9a87a84052fa12d5b30bd Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Fri, 4 Apr 2008 18:11:08 -0700
Subject: generic: reduce stack pressure in sched_affinity

  * Modify sched_affinity functions to pass cpumask_t variables by reference
    instead of by value.

  * Use new set_cpus_allowed_ptr function.

Depends on:
	[sched-devel]: sched: add new set_cpus_allowed_ptr function

Cc: Paul Jackson <pj@sgi.com>
Cc: Cliff Wickman <cpw@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/kernel/cpu/mcheck/mce_amd_64.c | 46 ++++++++++++++++-----------------
 include/linux/sched.h                   |  2 +-
 kernel/compat.c                         |  2 +-
 kernel/rcupreempt.c                     |  4 +--
 kernel/sched.c                          |  5 ++--
 5 files changed, 30 insertions(+), 29 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c b/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
index 32671da8184e..7c9a813e1193 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd_64.c
@@ -251,18 +251,18 @@ struct threshold_attr {
 	ssize_t(*store) (struct threshold_block *, const char *, size_t count);
 };
 
-static cpumask_t affinity_set(unsigned int cpu)
+static void affinity_set(unsigned int cpu, cpumask_t *oldmask,
+					   cpumask_t *newmask)
 {
-	cpumask_t oldmask = current->cpus_allowed;
-	cpumask_t newmask = CPU_MASK_NONE;
-	cpu_set(cpu, newmask);
-	set_cpus_allowed(current, newmask);
-	return oldmask;
+	*oldmask = current->cpus_allowed;
+	cpus_clear(*newmask);
+	cpu_set(cpu, *newmask);
+	set_cpus_allowed_ptr(current, newmask);
 }
 
-static void affinity_restore(cpumask_t oldmask)
+static void affinity_restore(const cpumask_t *oldmask)
 {
-	set_cpus_allowed(current, oldmask);
+	set_cpus_allowed_ptr(current, oldmask);
 }
 
 #define SHOW_FIELDS(name)                                           \
@@ -277,15 +277,15 @@ static ssize_t store_interrupt_enable(struct threshold_block *b,
 				      const char *buf, size_t count)
 {
 	char *end;
-	cpumask_t oldmask;
+	cpumask_t oldmask, newmask;
 	unsigned long new = simple_strtoul(buf, &end, 0);
 	if (end == buf)
 		return -EINVAL;
 	b->interrupt_enable = !!new;
 
-	oldmask = affinity_set(b->cpu);
+	affinity_set(b->cpu, &oldmask, &newmask);
 	threshold_restart_bank(b, 0, 0);
-	affinity_restore(oldmask);
+	affinity_restore(&oldmask);
 
 	return end - buf;
 }
@@ -294,7 +294,7 @@ static ssize_t store_threshold_limit(struct threshold_block *b,
 				     const char *buf, size_t count)
 {
 	char *end;
-	cpumask_t oldmask;
+	cpumask_t oldmask, newmask;
 	u16 old;
 	unsigned long new = simple_strtoul(buf, &end, 0);
 	if (end == buf)
@@ -306,9 +306,9 @@ static ssize_t store_threshold_limit(struct threshold_block *b,
 	old = b->threshold_limit;
 	b->threshold_limit = new;
 
-	oldmask = affinity_set(b->cpu);
+	affinity_set(b->cpu, &oldmask, &newmask);
 	threshold_restart_bank(b, 0, old);
-	affinity_restore(oldmask);
+	affinity_restore(&oldmask);
 
 	return end - buf;
 }
@@ -316,10 +316,10 @@ static ssize_t store_threshold_limit(struct threshold_block *b,
 static ssize_t show_error_count(struct threshold_block *b, char *buf)
 {
 	u32 high, low;
-	cpumask_t oldmask;
-	oldmask = affinity_set(b->cpu);
+	cpumask_t oldmask, newmask;
+	affinity_set(b->cpu, &oldmask, &newmask);
 	rdmsr(b->address, low, high);
-	affinity_restore(oldmask);
+	affinity_restore(&oldmask);
 	return sprintf(buf, "%x\n",
 		       (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit));
 }
@@ -327,10 +327,10 @@ static ssize_t show_error_count(struct threshold_block *b, char *buf)
 static ssize_t store_error_count(struct threshold_block *b,
 				 const char *buf, size_t count)
 {
-	cpumask_t oldmask;
-	oldmask = affinity_set(b->cpu);
+	cpumask_t oldmask, newmask;
+	affinity_set(b->cpu, &oldmask, &newmask);
 	threshold_restart_bank(b, 1, 0);
-	affinity_restore(oldmask);
+	affinity_restore(&oldmask);
 	return 1;
 }
 
@@ -468,7 +468,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
 {
 	int i, err = 0;
 	struct threshold_bank *b = NULL;
-	cpumask_t oldmask = CPU_MASK_NONE;
+	cpumask_t oldmask, newmask;
 	char name[32];
 
 	sprintf(name, "threshold_bank%i", bank);
@@ -519,10 +519,10 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank)
 
 	per_cpu(threshold_banks, cpu)[bank] = b;
 
-	oldmask = affinity_set(cpu);
+	affinity_set(cpu, &oldmask, &newmask);
 	err = allocate_threshold_blocks(cpu, bank, 0,
 					MSR_IA32_MC0_MISC + bank * 4);
-	affinity_restore(oldmask);
+	affinity_restore(&oldmask);
 
 	if (err)
 		goto out_free;
diff --git a/include/linux/sched.h b/include/linux/sched.h
index be5d31752dbd..383502dfda17 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2034,7 +2034,7 @@ static inline void arch_pick_mmap_layout(struct mm_struct *mm)
 }
 #endif
 
-extern long sched_setaffinity(pid_t pid, cpumask_t new_mask);
+extern long sched_setaffinity(pid_t pid, const cpumask_t *new_mask);
 extern long sched_getaffinity(pid_t pid, cpumask_t *mask);
 
 extern int sched_mc_power_savings, sched_smt_power_savings;
diff --git a/kernel/compat.c b/kernel/compat.c
index 9c48abfcd4a5..e1ef04870c2a 100644
--- a/kernel/compat.c
+++ b/kernel/compat.c
@@ -445,7 +445,7 @@ asmlinkage long compat_sys_sched_setaffinity(compat_pid_t pid,
 	if (retval)
 		return retval;
 
-	return sched_setaffinity(pid, new_mask);
+	return sched_setaffinity(pid, &new_mask);
 }
 
 asmlinkage long compat_sys_sched_getaffinity(compat_pid_t pid, unsigned int len,
diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c
index e9517014b57c..e1cdf196a515 100644
--- a/kernel/rcupreempt.c
+++ b/kernel/rcupreempt.c
@@ -1007,10 +1007,10 @@ void __synchronize_sched(void)
 	if (sched_getaffinity(0, &oldmask) < 0)
 		oldmask = cpu_possible_map;
 	for_each_online_cpu(cpu) {
-		sched_setaffinity(0, cpumask_of_cpu(cpu));
+		sched_setaffinity(0, &cpumask_of_cpu(cpu));
 		schedule();
 	}
-	sched_setaffinity(0, oldmask);
+	sched_setaffinity(0, &oldmask);
 }
 EXPORT_SYMBOL_GPL(__synchronize_sched);
 
diff --git a/kernel/sched.c b/kernel/sched.c
index ccc23a9cd264..1a8252385c4d 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -4908,9 +4908,10 @@ out_unlock:
 	return retval;
 }
 
-long sched_setaffinity(pid_t pid, cpumask_t new_mask)
+long sched_setaffinity(pid_t pid, const cpumask_t *in_mask)
 {
 	cpumask_t cpus_allowed;
+	cpumask_t new_mask = *in_mask;
 	struct task_struct *p;
 	int retval;
 
@@ -4991,7 +4992,7 @@ asmlinkage long sys_sched_setaffinity(pid_t pid, unsigned int len,
 	if (retval)
 		return retval;
 
-	return sched_setaffinity(pid, new_mask);
+	return sched_setaffinity(pid, &new_mask);
 }
 
 /*
-- 
cgit v1.2.3


From 7c16ec585c558960a508ccf9a08fcb9ed49b3754 Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Fri, 4 Apr 2008 18:11:11 -0700
Subject: cpumask: reduce stack usage in SD_x_INIT initializers

  * Remove empty cpumask_t (and all non-zero/non-null) variables
    in SD_*_INIT macros.  Use memset(0) to clear.  Also, don't
    inline the initializer functions to save on stack space in
    build_sched_domains().

  * Merge change to include/linux/topology.h that uses the new
    node_to_cpumask_ptr function in the nr_cpus_node macro into
    this patch.

Depends on:
	[mm-patch]: asm-generic-add-node_to_cpumask_ptr-macro.patch
	[sched-devel]: sched: add new set_cpus_allowed_ptr function

Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/asm-x86/topology.h |   5 -
 include/linux/topology.h   |  46 +-----
 kernel/sched.c             | 368 ++++++++++++++++++++++++++++++---------------
 3 files changed, 256 insertions(+), 163 deletions(-)

(limited to 'include/linux')

diff --git a/include/asm-x86/topology.h b/include/asm-x86/topology.h
index b167ca90f96f..9ef74c5d5ad6 100644
--- a/include/asm-x86/topology.h
+++ b/include/asm-x86/topology.h
@@ -154,10 +154,6 @@ extern unsigned long node_remap_size[];
 
 /* sched_domains SD_NODE_INIT for NUMAQ machines */
 #define SD_NODE_INIT (struct sched_domain) {		\
-	.span			= CPU_MASK_NONE,	\
-	.parent			= NULL,			\
-	.child			= NULL,			\
-	.groups			= NULL,			\
 	.min_interval		= 8,			\
 	.max_interval		= 32,			\
 	.busy_factor		= 32,			\
@@ -175,7 +171,6 @@ extern unsigned long node_remap_size[];
 				| SD_WAKE_BALANCE,	\
 	.last_balance		= jiffies,		\
 	.balance_interval	= 1,			\
-	.nr_balance_failed	= 0,			\
 }
 
 #ifdef CONFIG_X86_64_ACPI_NUMA
diff --git a/include/linux/topology.h b/include/linux/topology.h
index bd14f8b30f09..4bb7074a2c3a 100644
--- a/include/linux/topology.h
+++ b/include/linux/topology.h
@@ -38,16 +38,15 @@
 #endif
 
 #ifndef nr_cpus_node
-#define nr_cpus_node(node)							\
-	({									\
-		cpumask_t __tmp__;						\
-		__tmp__ = node_to_cpumask(node);				\
-		cpus_weight(__tmp__);						\
+#define nr_cpus_node(node)				\
+	({						\
+		node_to_cpumask_ptr(__tmp__, node);	\
+		cpus_weight(*__tmp__);			\
 	})
 #endif
 
-#define for_each_node_with_cpus(node)						\
-	for_each_online_node(node)						\
+#define for_each_node_with_cpus(node)			\
+	for_each_online_node(node)			\
 		if (nr_cpus_node(node))
 
 void arch_update_cpu_topology(void);
@@ -80,7 +79,9 @@ void arch_update_cpu_topology(void);
  * by defining their own arch-specific initializer in include/asm/topology.h.
  * A definition there will automagically override these default initializers
  * and allow arch-specific performance tuning of sched_domains.
+ * (Only non-zero and non-null fields need be specified.)
  */
+
 #ifdef CONFIG_SCHED_SMT
 /* MCD - Do we really need this?  It is always on if CONFIG_SCHED_SMT is,
  * so can't we drop this in favor of CONFIG_SCHED_SMT?
@@ -89,20 +90,10 @@ void arch_update_cpu_topology(void);
 /* Common values for SMT siblings */
 #ifndef SD_SIBLING_INIT
 #define SD_SIBLING_INIT (struct sched_domain) {		\
-	.span			= CPU_MASK_NONE,	\
-	.parent			= NULL,			\
-	.child			= NULL,			\
-	.groups			= NULL,			\
 	.min_interval		= 1,			\
 	.max_interval		= 2,			\
 	.busy_factor		= 64,			\
 	.imbalance_pct		= 110,			\
-	.cache_nice_tries	= 0,			\
-	.busy_idx		= 0,			\
-	.idle_idx		= 0,			\
-	.newidle_idx		= 0,			\
-	.wake_idx		= 0,			\
-	.forkexec_idx		= 0,			\
 	.flags			= SD_LOAD_BALANCE	\
 				| SD_BALANCE_NEWIDLE	\
 				| SD_BALANCE_FORK	\
@@ -112,7 +103,6 @@ void arch_update_cpu_topology(void);
 				| SD_SHARE_CPUPOWER,	\
 	.last_balance		= jiffies,		\
 	.balance_interval	= 1,			\
-	.nr_balance_failed	= 0,			\
 }
 #endif
 #endif /* CONFIG_SCHED_SMT */
@@ -121,18 +111,12 @@ void arch_update_cpu_topology(void);
 /* Common values for MC siblings. for now mostly derived from SD_CPU_INIT */
 #ifndef SD_MC_INIT
 #define SD_MC_INIT (struct sched_domain) {		\
-	.span			= CPU_MASK_NONE,	\
-	.parent			= NULL,			\
-	.child			= NULL,			\
-	.groups			= NULL,			\
 	.min_interval		= 1,			\
 	.max_interval		= 4,			\
 	.busy_factor		= 64,			\
 	.imbalance_pct		= 125,			\
 	.cache_nice_tries	= 1,			\
 	.busy_idx		= 2,			\
-	.idle_idx		= 0,			\
-	.newidle_idx		= 0,			\
 	.wake_idx		= 1,			\
 	.forkexec_idx		= 1,			\
 	.flags			= SD_LOAD_BALANCE	\
@@ -144,7 +128,6 @@ void arch_update_cpu_topology(void);
 				| BALANCE_FOR_MC_POWER,	\
 	.last_balance		= jiffies,		\
 	.balance_interval	= 1,			\
-	.nr_balance_failed	= 0,			\
 }
 #endif
 #endif /* CONFIG_SCHED_MC */
@@ -152,10 +135,6 @@ void arch_update_cpu_topology(void);
 /* Common values for CPUs */
 #ifndef SD_CPU_INIT
 #define SD_CPU_INIT (struct sched_domain) {		\
-	.span			= CPU_MASK_NONE,	\
-	.parent			= NULL,			\
-	.child			= NULL,			\
-	.groups			= NULL,			\
 	.min_interval		= 1,			\
 	.max_interval		= 4,			\
 	.busy_factor		= 64,			\
@@ -174,16 +153,11 @@ void arch_update_cpu_topology(void);
 				| BALANCE_FOR_PKG_POWER,\
 	.last_balance		= jiffies,		\
 	.balance_interval	= 1,			\
-	.nr_balance_failed	= 0,			\
 }
 #endif
 
 /* sched_domains SD_ALLNODES_INIT for NUMA machines */
 #define SD_ALLNODES_INIT (struct sched_domain) {	\
-	.span			= CPU_MASK_NONE,	\
-	.parent			= NULL,			\
-	.child			= NULL,			\
-	.groups			= NULL,			\
 	.min_interval		= 64,			\
 	.max_interval		= 64*num_online_cpus(),	\
 	.busy_factor		= 128,			\
@@ -191,14 +165,10 @@ void arch_update_cpu_topology(void);
 	.cache_nice_tries	= 1,			\
 	.busy_idx		= 3,			\
 	.idle_idx		= 3,			\
-	.newidle_idx		= 0, /* unused */	\
-	.wake_idx		= 0, /* unused */	\
-	.forkexec_idx		= 0, /* unused */	\
 	.flags			= SD_LOAD_BALANCE	\
 				| SD_SERIALIZE,	\
 	.last_balance		= jiffies,		\
 	.balance_interval	= 64,			\
-	.nr_balance_failed	= 0,			\
 }
 
 #ifdef CONFIG_NUMA
diff --git a/kernel/sched.c b/kernel/sched.c
index 9f7980f8ec00..6809178eaa9d 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -1869,17 +1869,17 @@ find_idlest_group(struct sched_domain *sd, struct task_struct *p, int this_cpu)
  * find_idlest_cpu - find the idlest cpu among the cpus in group.
  */
 static int
-find_idlest_cpu(struct sched_group *group, struct task_struct *p, int this_cpu)
+find_idlest_cpu(struct sched_group *group, struct task_struct *p, int this_cpu,
+		cpumask_t *tmp)
 {
-	cpumask_t tmp;
 	unsigned long load, min_load = ULONG_MAX;
 	int idlest = -1;
 	int i;
 
 	/* Traverse only the allowed CPUs */
-	cpus_and(tmp, group->cpumask, p->cpus_allowed);
+	cpus_and(*tmp, group->cpumask, p->cpus_allowed);
 
-	for_each_cpu_mask(i, tmp) {
+	for_each_cpu_mask(i, *tmp) {
 		load = weighted_cpuload(i);
 
 		if (load < min_load || (load == min_load && i == this_cpu)) {
@@ -1918,7 +1918,7 @@ static int sched_balance_self(int cpu, int flag)
 	}
 
 	while (sd) {
-		cpumask_t span;
+		cpumask_t span, tmpmask;
 		struct sched_group *group;
 		int new_cpu, weight;
 
@@ -1934,7 +1934,7 @@ static int sched_balance_self(int cpu, int flag)
 			continue;
 		}
 
-		new_cpu = find_idlest_cpu(group, t, cpu);
+		new_cpu = find_idlest_cpu(group, t, cpu, &tmpmask);
 		if (new_cpu == -1 || new_cpu == cpu) {
 			/* Now try balancing at a lower domain level of cpu */
 			sd = sd->child;
@@ -2818,7 +2818,7 @@ static int move_one_task(struct rq *this_rq, int this_cpu, struct rq *busiest,
 static struct sched_group *
 find_busiest_group(struct sched_domain *sd, int this_cpu,
 		   unsigned long *imbalance, enum cpu_idle_type idle,
-		   int *sd_idle, cpumask_t *cpus, int *balance)
+		   int *sd_idle, const cpumask_t *cpus, int *balance)
 {
 	struct sched_group *busiest = NULL, *this = NULL, *group = sd->groups;
 	unsigned long max_load, avg_load, total_load, this_load, total_pwr;
@@ -3119,7 +3119,7 @@ ret:
  */
 static struct rq *
 find_busiest_queue(struct sched_group *group, enum cpu_idle_type idle,
-		   unsigned long imbalance, cpumask_t *cpus)
+		   unsigned long imbalance, const cpumask_t *cpus)
 {
 	struct rq *busiest = NULL, *rq;
 	unsigned long max_load = 0;
@@ -3158,15 +3158,16 @@ find_busiest_queue(struct sched_group *group, enum cpu_idle_type idle,
  */
 static int load_balance(int this_cpu, struct rq *this_rq,
 			struct sched_domain *sd, enum cpu_idle_type idle,
-			int *balance)
+			int *balance, cpumask_t *cpus)
 {
 	int ld_moved, all_pinned = 0, active_balance = 0, sd_idle = 0;
 	struct sched_group *group;
 	unsigned long imbalance;
 	struct rq *busiest;
-	cpumask_t cpus = CPU_MASK_ALL;
 	unsigned long flags;
 
+	cpus_setall(*cpus);
+
 	/*
 	 * When power savings policy is enabled for the parent domain, idle
 	 * sibling can pick up load irrespective of busy siblings. In this case,
@@ -3181,7 +3182,7 @@ static int load_balance(int this_cpu, struct rq *this_rq,
 
 redo:
 	group = find_busiest_group(sd, this_cpu, &imbalance, idle, &sd_idle,
-				   &cpus, balance);
+				   cpus, balance);
 
 	if (*balance == 0)
 		goto out_balanced;
@@ -3191,7 +3192,7 @@ redo:
 		goto out_balanced;
 	}
 
-	busiest = find_busiest_queue(group, idle, imbalance, &cpus);
+	busiest = find_busiest_queue(group, idle, imbalance, cpus);
 	if (!busiest) {
 		schedstat_inc(sd, lb_nobusyq[idle]);
 		goto out_balanced;
@@ -3224,8 +3225,8 @@ redo:
 
 		/* All tasks on this runqueue were pinned by CPU affinity */
 		if (unlikely(all_pinned)) {
-			cpu_clear(cpu_of(busiest), cpus);
-			if (!cpus_empty(cpus))
+			cpu_clear(cpu_of(busiest), *cpus);
+			if (!cpus_empty(*cpus))
 				goto redo;
 			goto out_balanced;
 		}
@@ -3310,7 +3311,8 @@ out_one_pinned:
  * this_rq is locked.
  */
 static int
-load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd)
+load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd,
+			cpumask_t *cpus)
 {
 	struct sched_group *group;
 	struct rq *busiest = NULL;
@@ -3318,7 +3320,8 @@ load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd)
 	int ld_moved = 0;
 	int sd_idle = 0;
 	int all_pinned = 0;
-	cpumask_t cpus = CPU_MASK_ALL;
+
+	cpus_setall(*cpus);
 
 	/*
 	 * When power savings policy is enabled for the parent domain, idle
@@ -3333,14 +3336,13 @@ load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd)
 	schedstat_inc(sd, lb_count[CPU_NEWLY_IDLE]);
 redo:
 	group = find_busiest_group(sd, this_cpu, &imbalance, CPU_NEWLY_IDLE,
-				   &sd_idle, &cpus, NULL);
+				   &sd_idle, cpus, NULL);
 	if (!group) {
 		schedstat_inc(sd, lb_nobusyg[CPU_NEWLY_IDLE]);
 		goto out_balanced;
 	}
 
-	busiest = find_busiest_queue(group, CPU_NEWLY_IDLE, imbalance,
-				&cpus);
+	busiest = find_busiest_queue(group, CPU_NEWLY_IDLE, imbalance, cpus);
 	if (!busiest) {
 		schedstat_inc(sd, lb_nobusyq[CPU_NEWLY_IDLE]);
 		goto out_balanced;
@@ -3362,8 +3364,8 @@ redo:
 		spin_unlock(&busiest->lock);
 
 		if (unlikely(all_pinned)) {
-			cpu_clear(cpu_of(busiest), cpus);
-			if (!cpus_empty(cpus))
+			cpu_clear(cpu_of(busiest), *cpus);
+			if (!cpus_empty(*cpus))
 				goto redo;
 		}
 	}
@@ -3397,6 +3399,7 @@ static void idle_balance(int this_cpu, struct rq *this_rq)
 	struct sched_domain *sd;
 	int pulled_task = -1;
 	unsigned long next_balance = jiffies + HZ;
+	cpumask_t tmpmask;
 
 	for_each_domain(this_cpu, sd) {
 		unsigned long interval;
@@ -3406,8 +3409,8 @@ static void idle_balance(int this_cpu, struct rq *this_rq)
 
 		if (sd->flags & SD_BALANCE_NEWIDLE)
 			/* If we've pulled tasks over stop searching: */
-			pulled_task = load_balance_newidle(this_cpu,
-								this_rq, sd);
+			pulled_task = load_balance_newidle(this_cpu, this_rq,
+							   sd, &tmpmask);
 
 		interval = msecs_to_jiffies(sd->balance_interval);
 		if (time_after(next_balance, sd->last_balance + interval))
@@ -3566,6 +3569,7 @@ static void rebalance_domains(int cpu, enum cpu_idle_type idle)
 	/* Earliest time when we have to do rebalance again */
 	unsigned long next_balance = jiffies + 60*HZ;
 	int update_next_balance = 0;
+	cpumask_t tmp;
 
 	for_each_domain(cpu, sd) {
 		if (!(sd->flags & SD_LOAD_BALANCE))
@@ -3589,7 +3593,7 @@ static void rebalance_domains(int cpu, enum cpu_idle_type idle)
 		}
 
 		if (time_after_eq(jiffies, sd->last_balance + interval)) {
-			if (load_balance(cpu, rq, sd, idle, &balance)) {
+			if (load_balance(cpu, rq, sd, idle, &balance, &tmp)) {
 				/*
 				 * We've pulled tasks over so either we're no
 				 * longer idle, or one of our SMT siblings is
@@ -4945,7 +4949,7 @@ long sched_setaffinity(pid_t pid, const cpumask_t *in_mask)
 	cpuset_cpus_allowed(p, &cpus_allowed);
 	cpus_and(new_mask, new_mask, cpus_allowed);
  again:
-	retval = set_cpus_allowed(p, new_mask);
+	retval = set_cpus_allowed_ptr(p, &new_mask);
 
 	if (!retval) {
 		cpuset_cpus_allowed(p, &cpus_allowed);
@@ -5700,7 +5704,7 @@ static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p)
  */
 static void migrate_nr_uninterruptible(struct rq *rq_src)
 {
-	struct rq *rq_dest = cpu_rq(any_online_cpu(CPU_MASK_ALL));
+	struct rq *rq_dest = cpu_rq(any_online_cpu(*CPU_MASK_ALL_PTR));
 	unsigned long flags;
 
 	local_irq_save(flags);
@@ -6118,14 +6122,14 @@ EXPORT_SYMBOL(nr_cpu_ids);
 
 #ifdef CONFIG_SCHED_DEBUG
 
-static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level)
+static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level,
+				  cpumask_t *groupmask)
 {
 	struct sched_group *group = sd->groups;
-	cpumask_t groupmask;
 	char str[256];
 
 	cpulist_scnprintf(str, sizeof(str), sd->span);
-	cpus_clear(groupmask);
+	cpus_clear(*groupmask);
 
 	printk(KERN_DEBUG "%*s domain %d: ", level, "", level);
 
@@ -6169,13 +6173,13 @@ static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level)
 			break;
 		}
 
-		if (cpus_intersects(groupmask, group->cpumask)) {
+		if (cpus_intersects(*groupmask, group->cpumask)) {
 			printk(KERN_CONT "\n");
 			printk(KERN_ERR "ERROR: repeated CPUs\n");
 			break;
 		}
 
-		cpus_or(groupmask, groupmask, group->cpumask);
+		cpus_or(*groupmask, *groupmask, group->cpumask);
 
 		cpulist_scnprintf(str, sizeof(str), group->cpumask);
 		printk(KERN_CONT " %s", str);
@@ -6184,10 +6188,10 @@ static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level)
 	} while (group != sd->groups);
 	printk(KERN_CONT "\n");
 
-	if (!cpus_equal(sd->span, groupmask))
+	if (!cpus_equal(sd->span, *groupmask))
 		printk(KERN_ERR "ERROR: groups don't span domain->span\n");
 
-	if (sd->parent && !cpus_subset(groupmask, sd->parent->span))
+	if (sd->parent && !cpus_subset(*groupmask, sd->parent->span))
 		printk(KERN_ERR "ERROR: parent span is not a superset "
 			"of domain->span\n");
 	return 0;
@@ -6195,6 +6199,7 @@ static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level)
 
 static void sched_domain_debug(struct sched_domain *sd, int cpu)
 {
+	cpumask_t *groupmask;
 	int level = 0;
 
 	if (!sd) {
@@ -6204,14 +6209,21 @@ static void sched_domain_debug(struct sched_domain *sd, int cpu)
 
 	printk(KERN_DEBUG "CPU%d attaching sched-domain:\n", cpu);
 
+	groupmask = kmalloc(sizeof(cpumask_t), GFP_KERNEL);
+	if (!groupmask) {
+		printk(KERN_DEBUG "Cannot load-balance (out of memory)\n");
+		return;
+	}
+
 	for (;;) {
-		if (sched_domain_debug_one(sd, cpu, level))
+		if (sched_domain_debug_one(sd, cpu, level, groupmask))
 			break;
 		level++;
 		sd = sd->parent;
 		if (!sd)
 			break;
 	}
+	kfree(groupmask);
 }
 #else
 # define sched_domain_debug(sd, cpu) do { } while (0)
@@ -6399,30 +6411,33 @@ __setup("isolcpus=", isolated_cpu_setup);
  * and ->cpu_power to 0.
  */
 static void
-init_sched_build_groups(cpumask_t span, const cpumask_t *cpu_map,
+init_sched_build_groups(const cpumask_t *span, const cpumask_t *cpu_map,
 			int (*group_fn)(int cpu, const cpumask_t *cpu_map,
-					struct sched_group **sg))
+					struct sched_group **sg,
+					cpumask_t *tmpmask),
+			cpumask_t *covered, cpumask_t *tmpmask)
 {
 	struct sched_group *first = NULL, *last = NULL;
-	cpumask_t covered = CPU_MASK_NONE;
 	int i;
 
-	for_each_cpu_mask(i, span) {
+	cpus_clear(*covered);
+
+	for_each_cpu_mask(i, *span) {
 		struct sched_group *sg;
-		int group = group_fn(i, cpu_map, &sg);
+		int group = group_fn(i, cpu_map, &sg, tmpmask);
 		int j;
 
-		if (cpu_isset(i, covered))
+		if (cpu_isset(i, *covered))
 			continue;
 
-		sg->cpumask = CPU_MASK_NONE;
+		cpus_clear(sg->cpumask);
 		sg->__cpu_power = 0;
 
-		for_each_cpu_mask(j, span) {
-			if (group_fn(j, cpu_map, NULL) != group)
+		for_each_cpu_mask(j, *span) {
+			if (group_fn(j, cpu_map, NULL, tmpmask) != group)
 				continue;
 
-			cpu_set(j, covered);
+			cpu_set(j, *covered);
 			cpu_set(j, sg->cpumask);
 		}
 		if (!first)
@@ -6520,7 +6535,8 @@ static DEFINE_PER_CPU(struct sched_domain, cpu_domains);
 static DEFINE_PER_CPU(struct sched_group, sched_group_cpus);
 
 static int
-cpu_to_cpu_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg)
+cpu_to_cpu_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg,
+		 cpumask_t *unused)
 {
 	if (sg)
 		*sg = &per_cpu(sched_group_cpus, cpu);
@@ -6538,19 +6554,22 @@ static DEFINE_PER_CPU(struct sched_group, sched_group_core);
 
 #if defined(CONFIG_SCHED_MC) && defined(CONFIG_SCHED_SMT)
 static int
-cpu_to_core_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg)
+cpu_to_core_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg,
+		  cpumask_t *mask)
 {
 	int group;
-	cpumask_t mask = per_cpu(cpu_sibling_map, cpu);
-	cpus_and(mask, mask, *cpu_map);
-	group = first_cpu(mask);
+
+	*mask = per_cpu(cpu_sibling_map, cpu);
+	cpus_and(*mask, *mask, *cpu_map);
+	group = first_cpu(*mask);
 	if (sg)
 		*sg = &per_cpu(sched_group_core, group);
 	return group;
 }
 #elif defined(CONFIG_SCHED_MC)
 static int
-cpu_to_core_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg)
+cpu_to_core_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg,
+		  cpumask_t *unused)
 {
 	if (sg)
 		*sg = &per_cpu(sched_group_core, cpu);
@@ -6562,17 +6581,18 @@ static DEFINE_PER_CPU(struct sched_domain, phys_domains);
 static DEFINE_PER_CPU(struct sched_group, sched_group_phys);
 
 static int
-cpu_to_phys_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg)
+cpu_to_phys_group(int cpu, const cpumask_t *cpu_map, struct sched_group **sg,
+		  cpumask_t *mask)
 {
 	int group;
 #ifdef CONFIG_SCHED_MC
-	cpumask_t mask = cpu_coregroup_map(cpu);
-	cpus_and(mask, mask, *cpu_map);
-	group = first_cpu(mask);
+	*mask = cpu_coregroup_map(cpu);
+	cpus_and(*mask, *mask, *cpu_map);
+	group = first_cpu(*mask);
 #elif defined(CONFIG_SCHED_SMT)
-	cpumask_t mask = per_cpu(cpu_sibling_map, cpu);
-	cpus_and(mask, mask, *cpu_map);
-	group = first_cpu(mask);
+	*mask = per_cpu(cpu_sibling_map, cpu);
+	cpus_and(*mask, *mask, *cpu_map);
+	group = first_cpu(*mask);
 #else
 	group = cpu;
 #endif
@@ -6594,13 +6614,13 @@ static DEFINE_PER_CPU(struct sched_domain, allnodes_domains);
 static DEFINE_PER_CPU(struct sched_group, sched_group_allnodes);
 
 static int cpu_to_allnodes_group(int cpu, const cpumask_t *cpu_map,
-				 struct sched_group **sg)
+				 struct sched_group **sg, cpumask_t *nodemask)
 {
-	cpumask_t nodemask = node_to_cpumask(cpu_to_node(cpu));
 	int group;
 
-	cpus_and(nodemask, nodemask, *cpu_map);
-	group = first_cpu(nodemask);
+	*nodemask = node_to_cpumask(cpu_to_node(cpu));
+	cpus_and(*nodemask, *nodemask, *cpu_map);
+	group = first_cpu(*nodemask);
 
 	if (sg)
 		*sg = &per_cpu(sched_group_allnodes, group);
@@ -6636,7 +6656,7 @@ static void init_numa_sched_groups_power(struct sched_group *group_head)
 
 #ifdef CONFIG_NUMA
 /* Free memory allocated for various sched_group structures */
-static void free_sched_groups(const cpumask_t *cpu_map)
+static void free_sched_groups(const cpumask_t *cpu_map, cpumask_t *nodemask)
 {
 	int cpu, i;
 
@@ -6648,11 +6668,11 @@ static void free_sched_groups(const cpumask_t *cpu_map)
 			continue;
 
 		for (i = 0; i < MAX_NUMNODES; i++) {
-			cpumask_t nodemask = node_to_cpumask(i);
 			struct sched_group *oldsg, *sg = sched_group_nodes[i];
 
-			cpus_and(nodemask, nodemask, *cpu_map);
-			if (cpus_empty(nodemask))
+			*nodemask = node_to_cpumask(i);
+			cpus_and(*nodemask, *nodemask, *cpu_map);
+			if (cpus_empty(*nodemask))
 				continue;
 
 			if (sg == NULL)
@@ -6670,7 +6690,7 @@ next_sg:
 	}
 }
 #else
-static void free_sched_groups(const cpumask_t *cpu_map)
+static void free_sched_groups(const cpumask_t *cpu_map, cpumask_t *nodemask)
 {
 }
 #endif
@@ -6727,6 +6747,65 @@ static void init_sched_groups_power(int cpu, struct sched_domain *sd)
 	} while (group != child->groups);
 }
 
+/*
+ * Initializers for schedule domains
+ * Non-inlined to reduce accumulated stack pressure in build_sched_domains()
+ */
+
+#define	SD_INIT(sd, type)	sd_init_##type(sd)
+#define SD_INIT_FUNC(type)	\
+static noinline void sd_init_##type(struct sched_domain *sd)	\
+{								\
+	memset(sd, 0, sizeof(*sd));				\
+	*sd = SD_##type##_INIT;					\
+}
+
+SD_INIT_FUNC(CPU)
+#ifdef CONFIG_NUMA
+ SD_INIT_FUNC(ALLNODES)
+ SD_INIT_FUNC(NODE)
+#endif
+#ifdef CONFIG_SCHED_SMT
+ SD_INIT_FUNC(SIBLING)
+#endif
+#ifdef CONFIG_SCHED_MC
+ SD_INIT_FUNC(MC)
+#endif
+
+/*
+ * To minimize stack usage kmalloc room for cpumasks and share the
+ * space as the usage in build_sched_domains() dictates.  Used only
+ * if the amount of space is significant.
+ */
+struct allmasks {
+	cpumask_t tmpmask;			/* make this one first */
+	union {
+		cpumask_t nodemask;
+		cpumask_t this_sibling_map;
+		cpumask_t this_core_map;
+	};
+	cpumask_t send_covered;
+
+#ifdef CONFIG_NUMA
+	cpumask_t domainspan;
+	cpumask_t covered;
+	cpumask_t notcovered;
+#endif
+};
+
+#if	NR_CPUS > 128
+#define	SCHED_CPUMASK_ALLOC		1
+#define	SCHED_CPUMASK_FREE(v)		kfree(v)
+#define	SCHED_CPUMASK_DECLARE(v)	struct allmasks *v
+#else
+#define	SCHED_CPUMASK_ALLOC		0
+#define	SCHED_CPUMASK_FREE(v)
+#define	SCHED_CPUMASK_DECLARE(v)	struct allmasks _v, *v = &_v
+#endif
+
+#define	SCHED_CPUMASK_VAR(v, a) 	cpumask_t *v = (cpumask_t *) \
+			((unsigned long)(a) + offsetof(struct allmasks, v))
+
 /*
  * Build sched domains for a given set of cpus and attach the sched domains
  * to the individual cpus
@@ -6735,6 +6814,8 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 {
 	int i;
 	struct root_domain *rd;
+	SCHED_CPUMASK_DECLARE(allmasks);
+	cpumask_t *tmpmask;
 #ifdef CONFIG_NUMA
 	struct sched_group **sched_group_nodes = NULL;
 	int sd_allnodes = 0;
@@ -6748,38 +6829,60 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 		printk(KERN_WARNING "Can not alloc sched group node list\n");
 		return -ENOMEM;
 	}
-	sched_group_nodes_bycpu[first_cpu(*cpu_map)] = sched_group_nodes;
 #endif
 
 	rd = alloc_rootdomain();
 	if (!rd) {
 		printk(KERN_WARNING "Cannot alloc root domain\n");
+#ifdef CONFIG_NUMA
+		kfree(sched_group_nodes);
+#endif
 		return -ENOMEM;
 	}
 
+#if SCHED_CPUMASK_ALLOC
+	/* get space for all scratch cpumask variables */
+	allmasks = kmalloc(sizeof(*allmasks), GFP_KERNEL);
+	if (!allmasks) {
+		printk(KERN_WARNING "Cannot alloc cpumask array\n");
+		kfree(rd);
+#ifdef CONFIG_NUMA
+		kfree(sched_group_nodes);
+#endif
+		return -ENOMEM;
+	}
+#endif
+	tmpmask = (cpumask_t *)allmasks;
+
+
+#ifdef CONFIG_NUMA
+	sched_group_nodes_bycpu[first_cpu(*cpu_map)] = sched_group_nodes;
+#endif
+
 	/*
 	 * Set up domains for cpus specified by the cpu_map.
 	 */
 	for_each_cpu_mask(i, *cpu_map) {
 		struct sched_domain *sd = NULL, *p;
-		cpumask_t nodemask = node_to_cpumask(cpu_to_node(i));
+		SCHED_CPUMASK_VAR(nodemask, allmasks);
 
-		cpus_and(nodemask, nodemask, *cpu_map);
+		*nodemask = node_to_cpumask(cpu_to_node(i));
+		cpus_and(*nodemask, *nodemask, *cpu_map);
 
 #ifdef CONFIG_NUMA
 		if (cpus_weight(*cpu_map) >
-				SD_NODES_PER_DOMAIN*cpus_weight(nodemask)) {
+				SD_NODES_PER_DOMAIN*cpus_weight(*nodemask)) {
 			sd = &per_cpu(allnodes_domains, i);
-			*sd = SD_ALLNODES_INIT;
+			SD_INIT(sd, ALLNODES);
 			sd->span = *cpu_map;
-			cpu_to_allnodes_group(i, cpu_map, &sd->groups);
+			cpu_to_allnodes_group(i, cpu_map, &sd->groups, tmpmask);
 			p = sd;
 			sd_allnodes = 1;
 		} else
 			p = NULL;
 
 		sd = &per_cpu(node_domains, i);
-		*sd = SD_NODE_INIT;
+		SD_INIT(sd, NODE);
 		sd->span = sched_domain_node_span(cpu_to_node(i));
 		sd->parent = p;
 		if (p)
@@ -6789,94 +6892,114 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 
 		p = sd;
 		sd = &per_cpu(phys_domains, i);
-		*sd = SD_CPU_INIT;
-		sd->span = nodemask;
+		SD_INIT(sd, CPU);
+		sd->span = *nodemask;
 		sd->parent = p;
 		if (p)
 			p->child = sd;
-		cpu_to_phys_group(i, cpu_map, &sd->groups);
+		cpu_to_phys_group(i, cpu_map, &sd->groups, tmpmask);
 
 #ifdef CONFIG_SCHED_MC
 		p = sd;
 		sd = &per_cpu(core_domains, i);
-		*sd = SD_MC_INIT;
+		SD_INIT(sd, MC);
 		sd->span = cpu_coregroup_map(i);
 		cpus_and(sd->span, sd->span, *cpu_map);
 		sd->parent = p;
 		p->child = sd;
-		cpu_to_core_group(i, cpu_map, &sd->groups);
+		cpu_to_core_group(i, cpu_map, &sd->groups, tmpmask);
 #endif
 
 #ifdef CONFIG_SCHED_SMT
 		p = sd;
 		sd = &per_cpu(cpu_domains, i);
-		*sd = SD_SIBLING_INIT;
+		SD_INIT(sd, SIBLING);
 		sd->span = per_cpu(cpu_sibling_map, i);
 		cpus_and(sd->span, sd->span, *cpu_map);
 		sd->parent = p;
 		p->child = sd;
-		cpu_to_cpu_group(i, cpu_map, &sd->groups);
+		cpu_to_cpu_group(i, cpu_map, &sd->groups, tmpmask);
 #endif
 	}
 
 #ifdef CONFIG_SCHED_SMT
 	/* Set up CPU (sibling) groups */
 	for_each_cpu_mask(i, *cpu_map) {
-		cpumask_t this_sibling_map = per_cpu(cpu_sibling_map, i);
-		cpus_and(this_sibling_map, this_sibling_map, *cpu_map);
-		if (i != first_cpu(this_sibling_map))
+		SCHED_CPUMASK_VAR(this_sibling_map, allmasks);
+		SCHED_CPUMASK_VAR(send_covered, allmasks);
+
+		*this_sibling_map = per_cpu(cpu_sibling_map, i);
+		cpus_and(*this_sibling_map, *this_sibling_map, *cpu_map);
+		if (i != first_cpu(*this_sibling_map))
 			continue;
 
 		init_sched_build_groups(this_sibling_map, cpu_map,
-					&cpu_to_cpu_group);
+					&cpu_to_cpu_group,
+					send_covered, tmpmask);
 	}
 #endif
 
 #ifdef CONFIG_SCHED_MC
 	/* Set up multi-core groups */
 	for_each_cpu_mask(i, *cpu_map) {
-		cpumask_t this_core_map = cpu_coregroup_map(i);
-		cpus_and(this_core_map, this_core_map, *cpu_map);
-		if (i != first_cpu(this_core_map))
+		SCHED_CPUMASK_VAR(this_core_map, allmasks);
+		SCHED_CPUMASK_VAR(send_covered, allmasks);
+
+		*this_core_map = cpu_coregroup_map(i);
+		cpus_and(*this_core_map, *this_core_map, *cpu_map);
+		if (i != first_cpu(*this_core_map))
 			continue;
+
 		init_sched_build_groups(this_core_map, cpu_map,
-					&cpu_to_core_group);
+					&cpu_to_core_group,
+					send_covered, tmpmask);
 	}
 #endif
 
 	/* Set up physical groups */
 	for (i = 0; i < MAX_NUMNODES; i++) {
-		cpumask_t nodemask = node_to_cpumask(i);
+		SCHED_CPUMASK_VAR(nodemask, allmasks);
+		SCHED_CPUMASK_VAR(send_covered, allmasks);
 
-		cpus_and(nodemask, nodemask, *cpu_map);
-		if (cpus_empty(nodemask))
+		*nodemask = node_to_cpumask(i);
+		cpus_and(*nodemask, *nodemask, *cpu_map);
+		if (cpus_empty(*nodemask))
 			continue;
 
-		init_sched_build_groups(nodemask, cpu_map, &cpu_to_phys_group);
+		init_sched_build_groups(nodemask, cpu_map,
+					&cpu_to_phys_group,
+					send_covered, tmpmask);
 	}
 
 #ifdef CONFIG_NUMA
 	/* Set up node groups */
-	if (sd_allnodes)
-		init_sched_build_groups(*cpu_map, cpu_map,
-					&cpu_to_allnodes_group);
+	if (sd_allnodes) {
+		SCHED_CPUMASK_VAR(send_covered, allmasks);
+
+		init_sched_build_groups(cpu_map, cpu_map,
+					&cpu_to_allnodes_group,
+					send_covered, tmpmask);
+	}
 
 	for (i = 0; i < MAX_NUMNODES; i++) {
 		/* Set up node groups */
 		struct sched_group *sg, *prev;
-		cpumask_t nodemask = node_to_cpumask(i);
-		cpumask_t domainspan;
-		cpumask_t covered = CPU_MASK_NONE;
+		SCHED_CPUMASK_VAR(nodemask, allmasks);
+		SCHED_CPUMASK_VAR(domainspan, allmasks);
+		SCHED_CPUMASK_VAR(covered, allmasks);
 		int j;
 
-		cpus_and(nodemask, nodemask, *cpu_map);
-		if (cpus_empty(nodemask)) {
+		*nodemask = node_to_cpumask(i);
+		cpus_clear(*covered);
+
+		cpus_and(*nodemask, *nodemask, *cpu_map);
+		if (cpus_empty(*nodemask)) {
 			sched_group_nodes[i] = NULL;
 			continue;
 		}
 
-		domainspan = sched_domain_node_span(i);
-		cpus_and(domainspan, domainspan, *cpu_map);
+		*domainspan = sched_domain_node_span(i);
+		cpus_and(*domainspan, *domainspan, *cpu_map);
 
 		sg = kmalloc_node(sizeof(struct sched_group), GFP_KERNEL, i);
 		if (!sg) {
@@ -6885,31 +7008,31 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 			goto error;
 		}
 		sched_group_nodes[i] = sg;
-		for_each_cpu_mask(j, nodemask) {
+		for_each_cpu_mask(j, *nodemask) {
 			struct sched_domain *sd;
 
 			sd = &per_cpu(node_domains, j);
 			sd->groups = sg;
 		}
 		sg->__cpu_power = 0;
-		sg->cpumask = nodemask;
+		sg->cpumask = *nodemask;
 		sg->next = sg;
-		cpus_or(covered, covered, nodemask);
+		cpus_or(*covered, *covered, *nodemask);
 		prev = sg;
 
 		for (j = 0; j < MAX_NUMNODES; j++) {
-			cpumask_t tmp, notcovered;
+			SCHED_CPUMASK_VAR(notcovered, allmasks);
 			int n = (i + j) % MAX_NUMNODES;
 			node_to_cpumask_ptr(pnodemask, n);
 
-			cpus_complement(notcovered, covered);
-			cpus_and(tmp, notcovered, *cpu_map);
-			cpus_and(tmp, tmp, domainspan);
-			if (cpus_empty(tmp))
+			cpus_complement(*notcovered, *covered);
+			cpus_and(*tmpmask, *notcovered, *cpu_map);
+			cpus_and(*tmpmask, *tmpmask, *domainspan);
+			if (cpus_empty(*tmpmask))
 				break;
 
-			cpus_and(tmp, tmp, *pnodemask);
-			if (cpus_empty(tmp))
+			cpus_and(*tmpmask, *tmpmask, *pnodemask);
+			if (cpus_empty(*tmpmask))
 				continue;
 
 			sg = kmalloc_node(sizeof(struct sched_group),
@@ -6920,9 +7043,9 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 				goto error;
 			}
 			sg->__cpu_power = 0;
-			sg->cpumask = tmp;
+			sg->cpumask = *tmpmask;
 			sg->next = prev->next;
-			cpus_or(covered, covered, tmp);
+			cpus_or(*covered, *covered, *tmpmask);
 			prev->next = sg;
 			prev = sg;
 		}
@@ -6958,7 +7081,8 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 	if (sd_allnodes) {
 		struct sched_group *sg;
 
-		cpu_to_allnodes_group(first_cpu(*cpu_map), cpu_map, &sg);
+		cpu_to_allnodes_group(first_cpu(*cpu_map), cpu_map, &sg,
+								tmpmask);
 		init_numa_sched_groups_power(sg);
 	}
 #endif
@@ -6976,11 +7100,13 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 		cpu_attach_domain(sd, rd, i);
 	}
 
+	SCHED_CPUMASK_FREE((void *)allmasks);
 	return 0;
 
 #ifdef CONFIG_NUMA
 error:
-	free_sched_groups(cpu_map);
+	free_sched_groups(cpu_map, tmpmask);
+	SCHED_CPUMASK_FREE((void *)allmasks);
 	return -ENOMEM;
 #endif
 }
@@ -7020,9 +7146,10 @@ static int arch_init_sched_domains(const cpumask_t *cpu_map)
 	return err;
 }
 
-static void arch_destroy_sched_domains(const cpumask_t *cpu_map)
+static void arch_destroy_sched_domains(const cpumask_t *cpu_map,
+				       cpumask_t *tmpmask)
 {
-	free_sched_groups(cpu_map);
+	free_sched_groups(cpu_map, tmpmask);
 }
 
 /*
@@ -7031,6 +7158,7 @@ static void arch_destroy_sched_domains(const cpumask_t *cpu_map)
  */
 static void detach_destroy_domains(const cpumask_t *cpu_map)
 {
+	cpumask_t tmpmask;
 	int i;
 
 	unregister_sched_domain_sysctl();
@@ -7038,7 +7166,7 @@ static void detach_destroy_domains(const cpumask_t *cpu_map)
 	for_each_cpu_mask(i, *cpu_map)
 		cpu_attach_domain(NULL, &def_root_domain, i);
 	synchronize_sched();
-	arch_destroy_sched_domains(cpu_map);
+	arch_destroy_sched_domains(cpu_map, &tmpmask);
 }
 
 /*
@@ -7246,7 +7374,7 @@ void __init sched_init_smp(void)
 	hotcpu_notifier(update_sched_domains, 0);
 
 	/* Move init over to a non-isolated CPU */
-	if (set_cpus_allowed(current, non_isolated_cpus) < 0)
+	if (set_cpus_allowed_ptr(current, &non_isolated_cpus) < 0)
 		BUG();
 	sched_init_granularity();
 }
-- 
cgit v1.2.3


From 321a8e9dcb714f3c350ba55e41ed447bf3f05fac Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Fri, 4 Apr 2008 18:11:02 -0700
Subject: cpumask: add CPU_MASK_ALL_PTR macro

  * Add a static cpumask_t variable "CPU_MASK_ALL_PTR" to use as
    a pointer reference to CPU_MASK_ALL.  This reduces where possible
    the instances where CPU_MASK_ALL allocates and fills a large
    array on the stack.  Used only if NR_CPUS > BITS_PER_LONG.

  * Change init/main.c to use new set_cpus_allowed_ptr().

Depends on:
	[sched-devel]: sched: add new set_cpus_allowed_ptr function

Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/cpumask.h | 6 ++++++
 init/main.c             | 7 ++++++-
 2 files changed, 12 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h
index 67e0e38d32b1..629102feaa66 100644
--- a/include/linux/cpumask.h
+++ b/include/linux/cpumask.h
@@ -243,6 +243,8 @@ int __next_cpu(int n, const cpumask_t *srcp);
 	[BITS_TO_LONGS(NR_CPUS)-1] = CPU_MASK_LAST_WORD			\
 } }
 
+#define CPU_MASK_ALL_PTR	(&CPU_MASK_ALL)
+
 #else
 
 #define CPU_MASK_ALL							\
@@ -251,6 +253,10 @@ int __next_cpu(int n, const cpumask_t *srcp);
 	[BITS_TO_LONGS(NR_CPUS)-1] = CPU_MASK_LAST_WORD			\
 } }
 
+/* cpu_mask_all is in init/main.c */
+extern cpumask_t cpu_mask_all;
+#define CPU_MASK_ALL_PTR	(&cpu_mask_all)
+
 #endif
 
 #define CPU_MASK_NONE							\
diff --git a/init/main.c b/init/main.c
index 99ce94930b09..2df3f0617fdc 100644
--- a/init/main.c
+++ b/init/main.c
@@ -363,6 +363,11 @@ static inline void smp_prepare_cpus(unsigned int maxcpus) { }
 
 #else
 
+#if NR_CPUS > BITS_PER_LONG
+cpumask_t cpu_mask_all __read_mostly = CPU_MASK_ALL;
+EXPORT_SYMBOL(cpu_mask_all);
+#endif
+
 #ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA
 unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
 
@@ -811,7 +816,7 @@ static int __init kernel_init(void * unused)
 	/*
 	 * init can run on any cpu.
 	 */
-	set_cpus_allowed(current, CPU_MASK_ALL);
+	set_cpus_allowed_ptr(current, CPU_MASK_ALL_PTR);
 	/*
 	 * Tell the world that we're going to be the grim
 	 * reaper of innocent orphaned children.
-- 
cgit v1.2.3


From 9f0e8d0400d925c3acd5f4e01dbeb736e4011882 Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Fri, 4 Apr 2008 18:11:01 -0700
Subject: x86: convert cpumask_of_cpu macro to allocated array

  * Here is a simple patch to use an allocated array of cpumasks to
    represent cpumask_of_cpu() instead of constructing one on the stack.
    It's based on the Kconfig option "HAVE_CPUMASK_OF_CPU_MAP" which is
    currently only set for x86_64 SMP.  Otherwise the the existing
    cpumask_of_cpu() is used but has been changed to produce an lvalue
    so a pointer to it can be used.

Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/Kconfig        |  3 +++
 arch/x86/kernel/setup.c | 28 +++++++++++++++++++++++++++-
 include/linux/cpumask.h | 12 +++++++++---
 3 files changed, 39 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 2a59dbb28248..7f30b754bfc3 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -117,6 +117,9 @@ config ARCH_HAS_CPU_RELAX
 config HAVE_SETUP_PER_CPU_AREA
 	def_bool X86_64 || (X86_SMP && !X86_VOYAGER)
 
+config HAVE_CPUMASK_OF_CPU_MAP
+	def_bool X86_64_SMP
+
 config ARCH_HIBERNATION_POSSIBLE
 	def_bool y
 	depends on !SMP || !X86_VOYAGER
diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
index ed157c90412e..0d1f44ae6eea 100644
--- a/arch/x86/kernel/setup.c
+++ b/arch/x86/kernel/setup.c
@@ -54,6 +54,24 @@ static void __init setup_per_cpu_maps(void)
 #endif
 }
 
+#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP
+cpumask_t *cpumask_of_cpu_map __read_mostly;
+EXPORT_SYMBOL(cpumask_of_cpu_map);
+
+/* requires nr_cpu_ids to be initialized */
+static void __init setup_cpumask_of_cpu(void)
+{
+	int i;
+
+	/* alloc_bootmem zeroes memory */
+	cpumask_of_cpu_map = alloc_bootmem_low(sizeof(cpumask_t) * nr_cpu_ids);
+	for (i = 0; i < nr_cpu_ids; i++)
+		cpu_set(i, cpumask_of_cpu_map[i]);
+}
+#else
+static inline void setup_cpumask_of_cpu(void) { }
+#endif
+
 #ifdef CONFIG_X86_32
 /*
  * Great future not-so-futuristic plan: make i386 and x86_64 do it
@@ -70,7 +88,7 @@ EXPORT_SYMBOL(__per_cpu_offset);
  */
 void __init setup_per_cpu_areas(void)
 {
-	int i;
+	int i, highest_cpu = 0;
 	unsigned long size;
 
 #ifdef CONFIG_HOTPLUG_CPU
@@ -104,10 +122,18 @@ void __init setup_per_cpu_areas(void)
 		__per_cpu_offset[i] = ptr - __per_cpu_start;
 #endif
 		memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
+
+		highest_cpu = i;
 	}
 
+	nr_cpu_ids = highest_cpu + 1;
+	printk(KERN_DEBUG "NR_CPUS: %d, nr_cpu_ids: %d\n", NR_CPUS, nr_cpu_ids);
+
 	/* Setup percpu data maps */
 	setup_per_cpu_maps();
+
+	/* Setup cpumask_of_cpu map */
+	setup_cpumask_of_cpu();
 }
 
 #endif
diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h
index 629102feaa66..259c8051155d 100644
--- a/include/linux/cpumask.h
+++ b/include/linux/cpumask.h
@@ -222,8 +222,13 @@ int __next_cpu(int n, const cpumask_t *srcp);
 #define next_cpu(n, src)	({ (void)(src); 1; })
 #endif
 
+#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP
+extern cpumask_t *cpumask_of_cpu_map;
+#define cpumask_of_cpu(cpu)    (cpumask_of_cpu_map[cpu])
+
+#else
 #define cpumask_of_cpu(cpu)						\
-({									\
+(*({									\
 	typeof(_unused_cpumask_arg_) m;					\
 	if (sizeof(m) == sizeof(unsigned long)) {			\
 		m.bits[0] = 1UL<<(cpu);					\
@@ -231,8 +236,9 @@ int __next_cpu(int n, const cpumask_t *srcp);
 		cpus_clear(m);						\
 		cpu_set((cpu), m);					\
 	}								\
-	m;								\
-})
+	&m;								\
+}))
+#endif
 
 #define CPU_MASK_LAST_WORD BITMAP_LAST_WORD_MASK(NR_CPUS)
 
-- 
cgit v1.2.3


From 9d1fe3236a1d64ab687e16b4cbbaa1383352a2c1 Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Tue, 8 Apr 2008 11:43:04 -0700
Subject: cpumask: add show cpu map functions

  * Add cpu_sysdev_class functions to display the following maps
    with cpulist_scnprintf().

	cpu_online_map
	cpu_present_map
	cpu_possible_map

  * Small change to include/linux/sysdev.h to allow the attribute
    name and label to be different (to avoid collision with the
    "attr_online" entry for bringing cpus on- and off-line.)

Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 drivers/base/cpu.c     | 48 ++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/sysdev.h | 17 +++++++++++------
 2 files changed, 59 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
index 499b003f9278..2c76afff3b15 100644
--- a/drivers/base/cpu.c
+++ b/drivers/base/cpu.c
@@ -102,6 +102,51 @@ static ssize_t show_crash_notes(struct sys_device *dev, char *buf)
 static SYSDEV_ATTR(crash_notes, 0400, show_crash_notes, NULL);
 #endif
 
+/*
+ * Print cpu online, possible, present, and system maps
+ */
+static ssize_t print_cpus_map(char *buf, cpumask_t *map)
+{
+	int n = cpulist_scnprintf(buf, PAGE_SIZE-2, *map);
+
+	buf[n++] = '\n';
+	buf[n] = '\0';
+	return n;
+}
+
+#define	print_cpus_func(type) \
+static ssize_t print_cpus_##type(struct sysdev_class *class, char *buf)	\
+{									\
+	return print_cpus_map(buf, &cpu_##type##_map);			\
+}									\
+struct sysdev_class_attribute attr_##type##_map = 			\
+	_SYSDEV_CLASS_ATTR(type, 0444, print_cpus_##type, NULL)
+
+print_cpus_func(online);
+print_cpus_func(possible);
+print_cpus_func(present);
+
+struct sysdev_class_attribute *cpu_state_attr[] = {
+	&attr_online_map,
+	&attr_possible_map,
+	&attr_present_map,
+};
+
+static int cpu_states_init(void)
+{
+	int i;
+	int err = 0;
+
+	for (i = 0;  i < ARRAY_SIZE(cpu_state_attr); i++) {
+		int ret;
+		ret = sysdev_class_create_file(&cpu_sysdev_class,
+						cpu_state_attr[i]);
+		if (!err)
+			err = ret;
+	}
+	return err;
+}
+
 /*
  * register_cpu - Setup a sysfs device for a CPU.
  * @cpu - cpu->hotpluggable field set to 1 will generate a control file in
@@ -147,6 +192,9 @@ int __init cpu_dev_init(void)
 	int err;
 
 	err = sysdev_class_register(&cpu_sysdev_class);
+	if (!err)
+		err = cpu_states_init();
+
 #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT)
 	if (!err)
 		err = sched_create_sysfs_power_savings_entries(&cpu_sysdev_class);
diff --git a/include/linux/sysdev.h b/include/linux/sysdev.h
index f752e73bf977..f2767bc6b735 100644
--- a/include/linux/sysdev.h
+++ b/include/linux/sysdev.h
@@ -45,12 +45,16 @@ struct sysdev_class_attribute {
 	ssize_t (*store)(struct sysdev_class *, const char *, size_t);
 };
 
-#define SYSDEV_CLASS_ATTR(_name,_mode,_show,_store) 		\
-struct sysdev_class_attribute attr_##_name = { 			\
+#define _SYSDEV_CLASS_ATTR(_name,_mode,_show,_store) 		\
+{					 			\
 	.attr = {.name = __stringify(_name), .mode = _mode },	\
 	.show	= _show,					\
 	.store	= _store,					\
-};
+}
+
+#define SYSDEV_CLASS_ATTR(_name,_mode,_show,_store) 		\
+	struct sysdev_class_attribute attr_##_name = 		\
+		_SYSDEV_CLASS_ATTR(_name,_mode,_show,_store)
 
 
 extern int sysdev_class_register(struct sysdev_class *);
@@ -100,15 +104,16 @@ struct sysdev_attribute {
 };
 
 
-#define _SYSDEV_ATTR(_name,_mode,_show,_store)			\
+#define _SYSDEV_ATTR(_name, _mode, _show, _store)		\
 {								\
 	.attr = { .name = __stringify(_name), .mode = _mode },	\
 	.show	= _show,					\
 	.store	= _store,					\
 }
 
-#define SYSDEV_ATTR(_name,_mode,_show,_store)		\
-struct sysdev_attribute attr_##_name = _SYSDEV_ATTR(_name,_mode,_show,_store);
+#define SYSDEV_ATTR(_name, _mode, _show, _store)		\
+	struct sysdev_attribute attr_##_name =			\
+		_SYSDEV_ATTR(_name, _mode, _show, _store);
 
 extern int sysdev_create_file(struct sys_device *, struct sysdev_attribute *);
 extern void sysdev_remove_file(struct sys_device *, struct sysdev_attribute *);
-- 
cgit v1.2.3


From cd8ba7cd9be0192348c2836cb6645d9b2cd2bfd2 Mon Sep 17 00:00:00 2001
From: Mike Travis <travis@sgi.com>
Date: Wed, 26 Mar 2008 14:23:49 -0700
Subject: sched: add new set_cpus_allowed_ptr function

Add a new function that accepts a pointer to the "newly allowed cpus"
cpumask argument.

int set_cpus_allowed_ptr(struct task_struct *p, const cpumask_t *new_mask)

The current set_cpus_allowed() function is modified to use the above
but this does not result in an ABI change.  And with some compiler
optimization help, it may not introduce any additional overhead.

Additionally, to enforce the read only nature of the new_mask arg, the
"const" property is migrated to sub-functions called by set_cpus_allowed.
This silences compiler warnings.

Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h | 15 +++++++++++----
 kernel/sched.c        | 16 ++++++++--------
 kernel/sched_rt.c     |  3 ++-
 3 files changed, 21 insertions(+), 13 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 383502dfda17..79c025c3b627 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -889,7 +889,8 @@ struct sched_class {
 	void (*set_curr_task) (struct rq *rq);
 	void (*task_tick) (struct rq *rq, struct task_struct *p, int queued);
 	void (*task_new) (struct rq *rq, struct task_struct *p);
-	void (*set_cpus_allowed)(struct task_struct *p, cpumask_t *newmask);
+	void (*set_cpus_allowed)(struct task_struct *p,
+				 const cpumask_t *newmask);
 
 	void (*join_domain)(struct rq *rq);
 	void (*leave_domain)(struct rq *rq);
@@ -1502,15 +1503,21 @@ static inline void put_task_struct(struct task_struct *t)
 #define used_math() tsk_used_math(current)
 
 #ifdef CONFIG_SMP
-extern int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask);
+extern int set_cpus_allowed_ptr(struct task_struct *p,
+				const cpumask_t *new_mask);
 #else
-static inline int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask)
+static inline int set_cpus_allowed_ptr(struct task_struct *p,
+				       const cpumask_t *new_mask)
 {
-	if (!cpu_isset(0, new_mask))
+	if (!cpu_isset(0, *new_mask))
 		return -EINVAL;
 	return 0;
 }
 #endif
+static inline int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask)
+{
+	return set_cpus_allowed_ptr(p, &new_mask);
+}
 
 extern unsigned long long sched_clock(void);
 
diff --git a/kernel/sched.c b/kernel/sched.c
index 6ab0fcbf26e9..521b89b01480 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -5486,7 +5486,7 @@ static inline void sched_init_granularity(void)
  * task must not exit() & deallocate itself prematurely. The
  * call is not atomic; no spinlocks may be held.
  */
-int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask)
+int set_cpus_allowed_ptr(struct task_struct *p, const cpumask_t *new_mask)
 {
 	struct migration_req req;
 	unsigned long flags;
@@ -5494,23 +5494,23 @@ int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask)
 	int ret = 0;
 
 	rq = task_rq_lock(p, &flags);
-	if (!cpus_intersects(new_mask, cpu_online_map)) {
+	if (!cpus_intersects(*new_mask, cpu_online_map)) {
 		ret = -EINVAL;
 		goto out;
 	}
 
 	if (p->sched_class->set_cpus_allowed)
-		p->sched_class->set_cpus_allowed(p, &new_mask);
+		p->sched_class->set_cpus_allowed(p, new_mask);
 	else {
-		p->cpus_allowed = new_mask;
-		p->rt.nr_cpus_allowed = cpus_weight(new_mask);
+		p->cpus_allowed = *new_mask;
+		p->rt.nr_cpus_allowed = cpus_weight(*new_mask);
 	}
 
 	/* Can the task run on the task's current CPU? If so, we're done */
-	if (cpu_isset(task_cpu(p), new_mask))
+	if (cpu_isset(task_cpu(p), *new_mask))
 		goto out;
 
-	if (migrate_task(p, any_online_cpu(new_mask), &req)) {
+	if (migrate_task(p, any_online_cpu(*new_mask), &req)) {
 		/* Need help from migration thread: drop lock and wait. */
 		task_rq_unlock(rq, &flags);
 		wake_up_process(rq->migration_thread);
@@ -5523,7 +5523,7 @@ out:
 
 	return ret;
 }
-EXPORT_SYMBOL_GPL(set_cpus_allowed);
+EXPORT_SYMBOL_GPL(set_cpus_allowed_ptr);
 
 /*
  * Move (not current) task off this cpu, onto dest cpu. We're doing
diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 6928ded24da1..8ff824565e06 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -1123,7 +1123,8 @@ move_one_task_rt(struct rq *this_rq, int this_cpu, struct rq *busiest,
 	return 0;
 }
 
-static void set_cpus_allowed_rt(struct task_struct *p, cpumask_t *new_mask)
+static void set_cpus_allowed_rt(struct task_struct *p,
+				const cpumask_t *new_mask)
 {
 	int weight = cpus_weight(*new_mask);
 
-- 
cgit v1.2.3


From ec7dc8ac73e4a56ed03b673f026f08c0d547f597 Mon Sep 17 00:00:00 2001
From: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Date: Sat, 19 Apr 2008 19:44:59 +0200
Subject: sched: allow the group scheduler to have multiple levels

This patch makes the group scheduler multi hierarchy aware.

[a.p.zijlstra@chello.nl: rt-parts and assorted fixes]
Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |  2 +-
 kernel/sched.c        | 85 ++++++++++++++++++++++++++++++++-------------------
 kernel/user.c         |  2 +-
 3 files changed, 55 insertions(+), 34 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 79c025c3b627..fa14781747cb 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2052,7 +2052,7 @@ extern void normalize_rt_tasks(void);
 
 extern struct task_group init_task_group;
 
-extern struct task_group *sched_create_group(void);
+extern struct task_group *sched_create_group(struct task_group *parent);
 extern void sched_destroy_group(struct task_group *tg);
 extern void sched_move_task(struct task_struct *tsk);
 #ifdef CONFIG_FAIR_GROUP_SCHED
diff --git a/kernel/sched.c b/kernel/sched.c
index 1b7399dfa361..f9c8da798bbf 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -7438,10 +7438,11 @@ static void init_rt_rq(struct rt_rq *rt_rq, struct rq *rq)
 }
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
-static void init_tg_cfs_entry(struct rq *rq, struct task_group *tg,
-		struct cfs_rq *cfs_rq, struct sched_entity *se,
-		int cpu, int add)
+static void init_tg_cfs_entry(struct task_group *tg, struct cfs_rq *cfs_rq,
+				struct sched_entity *se, int cpu, int add,
+				struct sched_entity *parent)
 {
+	struct rq *rq = cpu_rq(cpu);
 	tg->cfs_rq[cpu] = cfs_rq;
 	init_cfs_rq(cfs_rq, rq);
 	cfs_rq->tg = tg;
@@ -7453,19 +7454,25 @@ static void init_tg_cfs_entry(struct rq *rq, struct task_group *tg,
 	if (!se)
 		return;
 
-	se->cfs_rq = &rq->cfs;
+	if (!parent)
+		se->cfs_rq = &rq->cfs;
+	else
+		se->cfs_rq = parent->my_q;
+
 	se->my_q = cfs_rq;
 	se->load.weight = tg->shares;
 	se->load.inv_weight = div64_64(1ULL<<32, se->load.weight);
-	se->parent = NULL;
+	se->parent = parent;
 }
 #endif
 
 #ifdef CONFIG_RT_GROUP_SCHED
-static void init_tg_rt_entry(struct rq *rq, struct task_group *tg,
-		struct rt_rq *rt_rq, struct sched_rt_entity *rt_se,
-		int cpu, int add)
+static void init_tg_rt_entry(struct task_group *tg, struct rt_rq *rt_rq,
+		struct sched_rt_entity *rt_se, int cpu, int add,
+		struct sched_rt_entity *parent)
 {
+	struct rq *rq = cpu_rq(cpu);
+
 	tg->rt_rq[cpu] = rt_rq;
 	init_rt_rq(rt_rq, rq);
 	rt_rq->tg = tg;
@@ -7478,9 +7485,14 @@ static void init_tg_rt_entry(struct rq *rq, struct task_group *tg,
 	if (!rt_se)
 		return;
 
+	if (!parent)
+		rt_se->rt_rq = &rq->rt;
+	else
+		rt_se->rt_rq = parent->my_q;
+
 	rt_se->rt_rq = &rq->rt;
 	rt_se->my_q = rt_rq;
-	rt_se->parent = NULL;
+	rt_se->parent = parent;
 	INIT_LIST_HEAD(&rt_se->run_list);
 }
 #endif
@@ -7568,7 +7580,7 @@ void __init sched_init(void)
 		 * We achieve this by letting init_task_group's tasks sit
 		 * directly in rq->cfs (i.e init_task_group->se[] = NULL).
 		 */
-		init_tg_cfs_entry(rq, &init_task_group, &rq->cfs, NULL, i, 1);
+		init_tg_cfs_entry(&init_task_group, &rq->cfs, NULL, i, 1, NULL);
 #elif defined CONFIG_USER_SCHED
 		/*
 		 * In case of task-groups formed thr' the user id of tasks,
@@ -7581,9 +7593,9 @@ void __init sched_init(void)
 		 * (init_cfs_rq) and having one entity represent this group of
 		 * tasks in rq->cfs (i.e init_task_group->se[] != NULL).
 		 */
-		init_tg_cfs_entry(rq, &init_task_group,
+		init_tg_cfs_entry(&init_task_group,
 				&per_cpu(init_cfs_rq, i),
-				&per_cpu(init_sched_entity, i), i, 1);
+				&per_cpu(init_sched_entity, i), i, 1, NULL);
 
 #endif
 #endif /* CONFIG_FAIR_GROUP_SCHED */
@@ -7592,11 +7604,11 @@ void __init sched_init(void)
 #ifdef CONFIG_RT_GROUP_SCHED
 		INIT_LIST_HEAD(&rq->leaf_rt_rq_list);
 #ifdef CONFIG_CGROUP_SCHED
-		init_tg_rt_entry(rq, &init_task_group, &rq->rt, NULL, i, 1);
+		init_tg_rt_entry(&init_task_group, &rq->rt, NULL, i, 1, NULL);
 #elif defined CONFIG_USER_SCHED
-		init_tg_rt_entry(rq, &init_task_group,
+		init_tg_rt_entry(&init_task_group,
 				&per_cpu(init_rt_rq, i),
-				&per_cpu(init_sched_rt_entity, i), i, 1);
+				&per_cpu(init_sched_rt_entity, i), i, 1, NULL);
 #endif
 #endif
 
@@ -7798,10 +7810,11 @@ static void free_fair_sched_group(struct task_group *tg)
 	kfree(tg->se);
 }
 
-static int alloc_fair_sched_group(struct task_group *tg)
+static
+int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent)
 {
 	struct cfs_rq *cfs_rq;
-	struct sched_entity *se;
+	struct sched_entity *se, *parent_se;
 	struct rq *rq;
 	int i;
 
@@ -7827,7 +7840,8 @@ static int alloc_fair_sched_group(struct task_group *tg)
 		if (!se)
 			goto err;
 
-		init_tg_cfs_entry(rq, tg, cfs_rq, se, i, 0);
+		parent_se = parent ? parent->se[i] : NULL;
+		init_tg_cfs_entry(tg, cfs_rq, se, i, 0, parent_se);
 	}
 
 	return 1;
@@ -7851,7 +7865,8 @@ static inline void free_fair_sched_group(struct task_group *tg)
 {
 }
 
-static inline int alloc_fair_sched_group(struct task_group *tg)
+static inline
+int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent)
 {
 	return 1;
 }
@@ -7883,10 +7898,11 @@ static void free_rt_sched_group(struct task_group *tg)
 	kfree(tg->rt_se);
 }
 
-static int alloc_rt_sched_group(struct task_group *tg)
+static
+int alloc_rt_sched_group(struct task_group *tg, struct task_group *parent)
 {
 	struct rt_rq *rt_rq;
-	struct sched_rt_entity *rt_se;
+	struct sched_rt_entity *rt_se, *parent_se;
 	struct rq *rq;
 	int i;
 
@@ -7913,7 +7929,8 @@ static int alloc_rt_sched_group(struct task_group *tg)
 		if (!rt_se)
 			goto err;
 
-		init_tg_rt_entry(rq, tg, rt_rq, rt_se, i, 0);
+		parent_se = parent ? parent->rt_se[i] : NULL;
+		init_tg_rt_entry(tg, rt_rq, rt_se, i, 0, parent_se);
 	}
 
 	return 1;
@@ -7937,7 +7954,8 @@ static inline void free_rt_sched_group(struct task_group *tg)
 {
 }
 
-static inline int alloc_rt_sched_group(struct task_group *tg)
+static inline
+int alloc_rt_sched_group(struct task_group *tg, struct task_group *parent)
 {
 	return 1;
 }
@@ -7960,7 +7978,7 @@ static void free_sched_group(struct task_group *tg)
 }
 
 /* allocate runqueue etc for a new task group */
-struct task_group *sched_create_group(void)
+struct task_group *sched_create_group(struct task_group *parent)
 {
 	struct task_group *tg;
 	unsigned long flags;
@@ -7970,10 +7988,10 @@ struct task_group *sched_create_group(void)
 	if (!tg)
 		return ERR_PTR(-ENOMEM);
 
-	if (!alloc_fair_sched_group(tg))
+	if (!alloc_fair_sched_group(tg, parent))
 		goto err;
 
-	if (!alloc_rt_sched_group(tg))
+	if (!alloc_rt_sched_group(tg, parent))
 		goto err;
 
 	spin_lock_irqsave(&task_group_lock, flags);
@@ -8084,6 +8102,12 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares)
 	int i;
 	unsigned long flags;
 
+	/*
+	 * We can't change the weight of the root cgroup.
+	 */
+	if (!tg->se[0])
+		return -EINVAL;
+
 	/*
 	 * A weight of 0 or 1 can cause arithmetics problems.
 	 * (The default weight is 1024 - so there's no practical
@@ -8327,7 +8351,7 @@ static inline struct task_group *cgroup_tg(struct cgroup *cgrp)
 static struct cgroup_subsys_state *
 cpu_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cgrp)
 {
-	struct task_group *tg;
+	struct task_group *tg, *parent;
 
 	if (!cgrp->parent) {
 		/* This is early initialization for the top cgroup */
@@ -8335,11 +8359,8 @@ cpu_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cgrp)
 		return &init_task_group.css;
 	}
 
-	/* we support only 1-level deep hierarchical scheduler atm */
-	if (cgrp->parent->parent)
-		return ERR_PTR(-EINVAL);
-
-	tg = sched_create_group();
+	parent = cgroup_tg(cgrp->parent);
+	tg = sched_create_group(parent);
 	if (IS_ERR(tg))
 		return ERR_PTR(-ENOMEM);
 
diff --git a/kernel/user.c b/kernel/user.c
index 5925c6887c10..a28d9f992468 100644
--- a/kernel/user.c
+++ b/kernel/user.c
@@ -101,7 +101,7 @@ static int sched_create_user(struct user_struct *up)
 {
 	int rc = 0;
 
-	up->tg = sched_create_group();
+	up->tg = sched_create_group(NULL);
 	if (IS_ERR(up->tg))
 		rc = -ENOMEM;
 
-- 
cgit v1.2.3


From eff766a65c60237bfa865160c3129de31fab591b Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Sat, 19 Apr 2008 19:45:00 +0200
Subject: sched: fix the task_group hierarchy for UID grouping

UID grouping doesn't actually have a task_group representing the root of
the task_group tree. Add one.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |  3 +++
 kernel/sched.c        | 43 +++++++++++++++++++++++++++++++++++++++++--
 kernel/user.c         |  2 +-
 3 files changed, 45 insertions(+), 3 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index fa14781747cb..ada24022d230 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2051,6 +2051,9 @@ extern void normalize_rt_tasks(void);
 #ifdef CONFIG_GROUP_SCHED
 
 extern struct task_group init_task_group;
+#ifdef CONFIG_USER_SCHED
+extern struct task_group root_task_group;
+#endif
 
 extern struct task_group *sched_create_group(struct task_group *parent);
 extern void sched_destroy_group(struct task_group *tg);
diff --git a/kernel/sched.c b/kernel/sched.c
index f9c8da798bbf..e03b45ccf789 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -274,6 +274,14 @@ struct task_group {
 };
 
 #ifdef CONFIG_USER_SCHED
+
+/*
+ * Root task group.
+ * 	Every UID task group (including init_task_group aka UID-0) will
+ * 	be a child to this group.
+ */
+struct task_group root_task_group;
+
 #ifdef CONFIG_FAIR_GROUP_SCHED
 /* Default task group's sched entity on each cpu */
 static DEFINE_PER_CPU(struct sched_entity, init_sched_entity);
@@ -285,6 +293,8 @@ static DEFINE_PER_CPU(struct cfs_rq, init_cfs_rq) ____cacheline_aligned_in_smp;
 static DEFINE_PER_CPU(struct sched_rt_entity, init_sched_rt_entity);
 static DEFINE_PER_CPU(struct rt_rq, init_rt_rq) ____cacheline_aligned_in_smp;
 #endif
+#else
+#define root_task_group init_task_group
 #endif
 
 /* task_group_lock serializes add/remove of task groups and also changes to
@@ -7507,6 +7517,9 @@ void __init sched_init(void)
 #endif
 #ifdef CONFIG_RT_GROUP_SCHED
 	alloc_size += 2 * nr_cpu_ids * sizeof(void **);
+#endif
+#ifdef CONFIG_USER_SCHED
+	alloc_size *= 2;
 #endif
 	/*
 	 * As sched_init() is called before page_alloc is setup,
@@ -7521,12 +7534,29 @@ void __init sched_init(void)
 
 		init_task_group.cfs_rq = (struct cfs_rq **)ptr;
 		ptr += nr_cpu_ids * sizeof(void **);
+
+#ifdef CONFIG_USER_SCHED
+		root_task_group.se = (struct sched_entity **)ptr;
+		ptr += nr_cpu_ids * sizeof(void **);
+
+		root_task_group.cfs_rq = (struct cfs_rq **)ptr;
+		ptr += nr_cpu_ids * sizeof(void **);
+#endif
 #endif
 #ifdef CONFIG_RT_GROUP_SCHED
 		init_task_group.rt_se = (struct sched_rt_entity **)ptr;
 		ptr += nr_cpu_ids * sizeof(void **);
 
 		init_task_group.rt_rq = (struct rt_rq **)ptr;
+		ptr += nr_cpu_ids * sizeof(void **);
+
+#ifdef CONFIG_USER_SCHED
+		root_task_group.rt_se = (struct sched_rt_entity **)ptr;
+		ptr += nr_cpu_ids * sizeof(void **);
+
+		root_task_group.rt_rq = (struct rt_rq **)ptr;
+		ptr += nr_cpu_ids * sizeof(void **);
+#endif
 #endif
 	}
 
@@ -7540,6 +7570,10 @@ void __init sched_init(void)
 #ifdef CONFIG_RT_GROUP_SCHED
 	init_rt_bandwidth(&init_task_group.rt_bandwidth,
 			global_rt_period(), global_rt_runtime());
+#ifdef CONFIG_USER_SCHED
+	init_rt_bandwidth(&root_task_group.rt_bandwidth,
+			global_rt_period(), RUNTIME_INF);
+#endif
 #endif
 
 #ifdef CONFIG_GROUP_SCHED
@@ -7582,6 +7616,8 @@ void __init sched_init(void)
 		 */
 		init_tg_cfs_entry(&init_task_group, &rq->cfs, NULL, i, 1, NULL);
 #elif defined CONFIG_USER_SCHED
+		root_task_group.shares = NICE_0_LOAD;
+		init_tg_cfs_entry(&root_task_group, &rq->cfs, NULL, i, 0, NULL);
 		/*
 		 * In case of task-groups formed thr' the user id of tasks,
 		 * init_task_group represents tasks belonging to root user.
@@ -7595,7 +7631,8 @@ void __init sched_init(void)
 		 */
 		init_tg_cfs_entry(&init_task_group,
 				&per_cpu(init_cfs_rq, i),
-				&per_cpu(init_sched_entity, i), i, 1, NULL);
+				&per_cpu(init_sched_entity, i), i, 1,
+				root_task_group.se[i]);
 
 #endif
 #endif /* CONFIG_FAIR_GROUP_SCHED */
@@ -7606,9 +7643,11 @@ void __init sched_init(void)
 #ifdef CONFIG_CGROUP_SCHED
 		init_tg_rt_entry(&init_task_group, &rq->rt, NULL, i, 1, NULL);
 #elif defined CONFIG_USER_SCHED
+		init_tg_rt_entry(&root_task_group, &rq->rt, NULL, i, 0, NULL);
 		init_tg_rt_entry(&init_task_group,
 				&per_cpu(init_rt_rq, i),
-				&per_cpu(init_sched_rt_entity, i), i, 1, NULL);
+				&per_cpu(init_sched_rt_entity, i), i, 1,
+				root_task_group.rt_se[i]);
 #endif
 #endif
 
diff --git a/kernel/user.c b/kernel/user.c
index a28d9f992468..debce602bfdd 100644
--- a/kernel/user.c
+++ b/kernel/user.c
@@ -101,7 +101,7 @@ static int sched_create_user(struct user_struct *up)
 {
 	int rc = 0;
 
-	up->tg = sched_create_group(NULL);
+	up->tg = sched_create_group(&root_task_group);
 	if (IS_ERR(up->tg))
 		rc = -ENOMEM;
 
-- 
cgit v1.2.3


From 1d3504fcf5606579d60b649d19f44b3871c1ddae Mon Sep 17 00:00:00 2001
From: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Date: Tue, 15 Apr 2008 14:04:23 +0900
Subject: sched, cpuset: customize sched domains, core

[rebased for sched-devel/latest]

 - Add a new cpuset file, having levels:
     sched_relax_domain_level

 - Modify partition_sched_domains() and build_sched_domains()
   to take attributes parameter passed from cpuset.

 - Fill newidle_idx for node domains which currently unused but
   might be required if sched_relax_domain_level become higher.

 - We can change the default level by boot option 'relax_domain_level='.

Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/asm-ia64/topology.h |  2 +-
 include/asm-sh/topology.h   |  2 +-
 include/asm-x86/topology.h  |  2 +-
 include/linux/sched.h       | 23 ++++++++++++-
 kernel/cpuset.c             | 61 ++++++++++++++++++++++++++++++++++-
 kernel/sched.c              | 78 ++++++++++++++++++++++++++++++++++++++++++---
 kernel/sched_fair.c         |  4 ++-
 7 files changed, 161 insertions(+), 11 deletions(-)

(limited to 'include/linux')

diff --git a/include/asm-ia64/topology.h b/include/asm-ia64/topology.h
index f929dde85343..f2f72ef2a897 100644
--- a/include/asm-ia64/topology.h
+++ b/include/asm-ia64/topology.h
@@ -93,7 +93,7 @@ void build_cpu_to_node_map(void);
 	.cache_nice_tries	= 2,			\
 	.busy_idx		= 3,			\
 	.idle_idx		= 2,			\
-	.newidle_idx		= 0, /* unused */	\
+	.newidle_idx		= 2,			\
 	.wake_idx		= 1,			\
 	.forkexec_idx		= 1,			\
 	.flags			= SD_LOAD_BALANCE	\
diff --git a/include/asm-sh/topology.h b/include/asm-sh/topology.h
index f402a3b1cfa4..34cdb28e8f44 100644
--- a/include/asm-sh/topology.h
+++ b/include/asm-sh/topology.h
@@ -16,7 +16,7 @@
 	.cache_nice_tries	= 2,			\
 	.busy_idx		= 3,			\
 	.idle_idx		= 2,			\
-	.newidle_idx		= 0,			\
+	.newidle_idx		= 2,			\
 	.wake_idx		= 1,			\
 	.forkexec_idx		= 1,			\
 	.flags			= SD_LOAD_BALANCE	\
diff --git a/include/asm-x86/topology.h b/include/asm-x86/topology.h
index 9ef74c5d5ad6..22073268b481 100644
--- a/include/asm-x86/topology.h
+++ b/include/asm-x86/topology.h
@@ -147,7 +147,7 @@ extern unsigned long node_remap_size[];
 
 # define SD_CACHE_NICE_TRIES	2
 # define SD_IDLE_IDX		2
-# define SD_NEWIDLE_IDX		0
+# define SD_NEWIDLE_IDX		2
 # define SD_FORKEXEC_IDX	1
 
 #endif
diff --git a/include/linux/sched.h b/include/linux/sched.h
index ada24022d230..11f47249cdd2 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -704,6 +704,7 @@ enum cpu_idle_type {
 #define SD_POWERSAVINGS_BALANCE	256	/* Balance for power savings */
 #define SD_SHARE_PKG_RESOURCES	512	/* Domain members share cpu pkg resources */
 #define SD_SERIALIZE		1024	/* Only a single load balancing instance */
+#define SD_WAKE_IDLE_FAR	2048	/* Gain latency sacrificing cache hit */
 
 #define BALANCE_FOR_MC_POWER	\
 	(sched_smt_power_savings ? SD_POWERSAVINGS_BALANCE : 0)
@@ -733,6 +734,24 @@ struct sched_group {
 	u32 reciprocal_cpu_power;
 };
 
+enum sched_domain_level {
+	SD_LV_NONE = 0,
+	SD_LV_SIBLING,
+	SD_LV_MC,
+	SD_LV_CPU,
+	SD_LV_NODE,
+	SD_LV_ALLNODES,
+	SD_LV_MAX
+};
+
+struct sched_domain_attr {
+	int relax_domain_level;
+};
+
+#define SD_ATTR_INIT	(struct sched_domain_attr) {	\
+	.relax_domain_level = -1,			\
+}
+
 struct sched_domain {
 	/* These fields must be setup */
 	struct sched_domain *parent;	/* top domain must be null terminated */
@@ -750,6 +769,7 @@ struct sched_domain {
 	unsigned int wake_idx;
 	unsigned int forkexec_idx;
 	int flags;			/* See SD_* */
+	enum sched_domain_level level;
 
 	/* Runtime fields. */
 	unsigned long last_balance;	/* init to jiffies. units in jiffies */
@@ -789,7 +809,8 @@ struct sched_domain {
 #endif
 };
 
-extern void partition_sched_domains(int ndoms_new, cpumask_t *doms_new);
+extern void partition_sched_domains(int ndoms_new, cpumask_t *doms_new,
+				    struct sched_domain_attr *dattr_new);
 extern int arch_reinit_sched_domains(void);
 
 #endif	/* CONFIG_SMP */
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index b0c870b2ac30..8b35fbd8292f 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -98,6 +98,9 @@ struct cpuset {
 	/* partition number for rebuild_sched_domains() */
 	int pn;
 
+	/* for custom sched domain */
+	int relax_domain_level;
+
 	/* used for walking a cpuset heirarchy */
 	struct list_head stack_list;
 };
@@ -478,6 +481,16 @@ static int cpusets_overlap(struct cpuset *a, struct cpuset *b)
 	return cpus_intersects(a->cpus_allowed, b->cpus_allowed);
 }
 
+static void
+update_domain_attr(struct sched_domain_attr *dattr, struct cpuset *c)
+{
+	if (!dattr)
+		return;
+	if (dattr->relax_domain_level < c->relax_domain_level)
+		dattr->relax_domain_level = c->relax_domain_level;
+	return;
+}
+
 /*
  * rebuild_sched_domains()
  *
@@ -553,12 +566,14 @@ static void rebuild_sched_domains(void)
 	int csn;		/* how many cpuset ptrs in csa so far */
 	int i, j, k;		/* indices for partition finding loops */
 	cpumask_t *doms;	/* resulting partition; i.e. sched domains */
+	struct sched_domain_attr *dattr;  /* attributes for custom domains */
 	int ndoms;		/* number of sched domains in result */
 	int nslot;		/* next empty doms[] cpumask_t slot */
 
 	q = NULL;
 	csa = NULL;
 	doms = NULL;
+	dattr = NULL;
 
 	/* Special case for the 99% of systems with one, full, sched domain */
 	if (is_sched_load_balance(&top_cpuset)) {
@@ -566,6 +581,11 @@ static void rebuild_sched_domains(void)
 		doms = kmalloc(sizeof(cpumask_t), GFP_KERNEL);
 		if (!doms)
 			goto rebuild;
+		dattr = kmalloc(sizeof(struct sched_domain_attr), GFP_KERNEL);
+		if (dattr) {
+			*dattr = SD_ATTR_INIT;
+			update_domain_attr(dattr, &top_cpuset);
+		}
 		*doms = top_cpuset.cpus_allowed;
 		goto rebuild;
 	}
@@ -622,6 +642,7 @@ restart:
 	doms = kmalloc(ndoms * sizeof(cpumask_t), GFP_KERNEL);
 	if (!doms)
 		goto rebuild;
+	dattr = kmalloc(ndoms * sizeof(struct sched_domain_attr), GFP_KERNEL);
 
 	for (nslot = 0, i = 0; i < csn; i++) {
 		struct cpuset *a = csa[i];
@@ -644,12 +665,15 @@ restart:
 			}
 
 			cpus_clear(*dp);
+			if (dattr)
+				*(dattr + nslot) = SD_ATTR_INIT;
 			for (j = i; j < csn; j++) {
 				struct cpuset *b = csa[j];
 
 				if (apn == b->pn) {
 					cpus_or(*dp, *dp, b->cpus_allowed);
 					b->pn = -1;
+					update_domain_attr(dattr, b);
 				}
 			}
 			nslot++;
@@ -660,7 +684,7 @@ restart:
 rebuild:
 	/* Have scheduler rebuild sched domains */
 	get_online_cpus();
-	partition_sched_domains(ndoms, doms);
+	partition_sched_domains(ndoms, doms, dattr);
 	put_online_cpus();
 
 done:
@@ -668,6 +692,7 @@ done:
 		kfifo_free(q);
 	kfree(csa);
 	/* Don't kfree(doms) -- partition_sched_domains() does that. */
+	/* Don't kfree(dattr) -- partition_sched_domains() does that. */
 }
 
 static inline int started_after_time(struct task_struct *t1,
@@ -1011,6 +1036,21 @@ static int update_memory_pressure_enabled(struct cpuset *cs, char *buf)
 	return 0;
 }
 
+static int update_relax_domain_level(struct cpuset *cs, char *buf)
+{
+	int val = simple_strtol(buf, NULL, 10);
+
+	if (val < 0)
+		val = -1;
+
+	if (val != cs->relax_domain_level) {
+		cs->relax_domain_level = val;
+		rebuild_sched_domains();
+	}
+
+	return 0;
+}
+
 /*
  * update_flag - read a 0 or a 1 in a file and update associated flag
  * bit:	the bit to update (CS_CPU_EXCLUSIVE, CS_MEM_EXCLUSIVE,
@@ -1202,6 +1242,7 @@ typedef enum {
 	FILE_CPU_EXCLUSIVE,
 	FILE_MEM_EXCLUSIVE,
 	FILE_SCHED_LOAD_BALANCE,
+	FILE_SCHED_RELAX_DOMAIN_LEVEL,
 	FILE_MEMORY_PRESSURE_ENABLED,
 	FILE_MEMORY_PRESSURE,
 	FILE_SPREAD_PAGE,
@@ -1256,6 +1297,9 @@ static ssize_t cpuset_common_file_write(struct cgroup *cont,
 	case FILE_SCHED_LOAD_BALANCE:
 		retval = update_flag(CS_SCHED_LOAD_BALANCE, cs, buffer);
 		break;
+	case FILE_SCHED_RELAX_DOMAIN_LEVEL:
+		retval = update_relax_domain_level(cs, buffer);
+		break;
 	case FILE_MEMORY_MIGRATE:
 		retval = update_flag(CS_MEMORY_MIGRATE, cs, buffer);
 		break;
@@ -1354,6 +1398,9 @@ static ssize_t cpuset_common_file_read(struct cgroup *cont,
 	case FILE_SCHED_LOAD_BALANCE:
 		*s++ = is_sched_load_balance(cs) ? '1' : '0';
 		break;
+	case FILE_SCHED_RELAX_DOMAIN_LEVEL:
+		s += sprintf(s, "%d", cs->relax_domain_level);
+		break;
 	case FILE_MEMORY_MIGRATE:
 		*s++ = is_memory_migrate(cs) ? '1' : '0';
 		break;
@@ -1424,6 +1471,13 @@ static struct cftype cft_sched_load_balance = {
 	.private = FILE_SCHED_LOAD_BALANCE,
 };
 
+static struct cftype cft_sched_relax_domain_level = {
+	.name = "sched_relax_domain_level",
+	.read = cpuset_common_file_read,
+	.write = cpuset_common_file_write,
+	.private = FILE_SCHED_RELAX_DOMAIN_LEVEL,
+};
+
 static struct cftype cft_memory_migrate = {
 	.name = "memory_migrate",
 	.read = cpuset_common_file_read,
@@ -1475,6 +1529,9 @@ static int cpuset_populate(struct cgroup_subsys *ss, struct cgroup *cont)
 		return err;
 	if ((err = cgroup_add_file(cont, ss, &cft_sched_load_balance)) < 0)
 		return err;
+	if ((err = cgroup_add_file(cont, ss,
+					&cft_sched_relax_domain_level)) < 0)
+		return err;
 	if ((err = cgroup_add_file(cont, ss, &cft_memory_pressure)) < 0)
 		return err;
 	if ((err = cgroup_add_file(cont, ss, &cft_spread_page)) < 0)
@@ -1559,6 +1616,7 @@ static struct cgroup_subsys_state *cpuset_create(
 	nodes_clear(cs->mems_allowed);
 	cs->mems_generation = cpuset_mems_generation++;
 	fmeter_init(&cs->fmeter);
+	cs->relax_domain_level = -1;
 
 	cs->parent = parent;
 	number_of_cpusets++;
@@ -1631,6 +1689,7 @@ int __init cpuset_init(void)
 	fmeter_init(&top_cpuset.fmeter);
 	top_cpuset.mems_generation = cpuset_mems_generation++;
 	set_bit(CS_SCHED_LOAD_BALANCE, &top_cpuset.flags);
+	top_cpuset.relax_domain_level = -1;
 
 	err = register_filesystem(&cpuset_fs_type);
 	if (err < 0)
diff --git a/kernel/sched.c b/kernel/sched.c
index 475e3fcab738..62d7481caca5 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -6771,6 +6771,7 @@ static noinline void sd_init_##type(struct sched_domain *sd)	\
 {								\
 	memset(sd, 0, sizeof(*sd));				\
 	*sd = SD_##type##_INIT;					\
+	sd->level = SD_LV_##type;				\
 }
 
 SD_INIT_FUNC(CPU)
@@ -6819,11 +6820,42 @@ struct allmasks {
 #define	SCHED_CPUMASK_VAR(v, a) 	cpumask_t *v = (cpumask_t *) \
 			((unsigned long)(a) + offsetof(struct allmasks, v))
 
+static int default_relax_domain_level = -1;
+
+static int __init setup_relax_domain_level(char *str)
+{
+	default_relax_domain_level = simple_strtoul(str, NULL, 0);
+	return 1;
+}
+__setup("relax_domain_level=", setup_relax_domain_level);
+
+static void set_domain_attribute(struct sched_domain *sd,
+				 struct sched_domain_attr *attr)
+{
+	int request;
+
+	if (!attr || attr->relax_domain_level < 0) {
+		if (default_relax_domain_level < 0)
+			return;
+		else
+			request = default_relax_domain_level;
+	} else
+		request = attr->relax_domain_level;
+	if (request < sd->level) {
+		/* turn off idle balance on this domain */
+		sd->flags &= ~(SD_WAKE_IDLE|SD_BALANCE_NEWIDLE);
+	} else {
+		/* turn on idle balance on this domain */
+		sd->flags |= (SD_WAKE_IDLE_FAR|SD_BALANCE_NEWIDLE);
+	}
+}
+
 /*
  * Build sched domains for a given set of cpus and attach the sched domains
  * to the individual cpus
  */
-static int build_sched_domains(const cpumask_t *cpu_map)
+static int __build_sched_domains(const cpumask_t *cpu_map,
+				 struct sched_domain_attr *attr)
 {
 	int i;
 	struct root_domain *rd;
@@ -6887,6 +6919,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 				SD_NODES_PER_DOMAIN*cpus_weight(*nodemask)) {
 			sd = &per_cpu(allnodes_domains, i);
 			SD_INIT(sd, ALLNODES);
+			set_domain_attribute(sd, attr);
 			sd->span = *cpu_map;
 			cpu_to_allnodes_group(i, cpu_map, &sd->groups, tmpmask);
 			p = sd;
@@ -6896,6 +6929,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 
 		sd = &per_cpu(node_domains, i);
 		SD_INIT(sd, NODE);
+		set_domain_attribute(sd, attr);
 		sched_domain_node_span(cpu_to_node(i), &sd->span);
 		sd->parent = p;
 		if (p)
@@ -6906,6 +6940,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 		p = sd;
 		sd = &per_cpu(phys_domains, i);
 		SD_INIT(sd, CPU);
+		set_domain_attribute(sd, attr);
 		sd->span = *nodemask;
 		sd->parent = p;
 		if (p)
@@ -6916,6 +6951,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 		p = sd;
 		sd = &per_cpu(core_domains, i);
 		SD_INIT(sd, MC);
+		set_domain_attribute(sd, attr);
 		sd->span = cpu_coregroup_map(i);
 		cpus_and(sd->span, sd->span, *cpu_map);
 		sd->parent = p;
@@ -6927,6 +6963,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
 		p = sd;
 		sd = &per_cpu(cpu_domains, i);
 		SD_INIT(sd, SIBLING);
+		set_domain_attribute(sd, attr);
 		sd->span = per_cpu(cpu_sibling_map, i);
 		cpus_and(sd->span, sd->span, *cpu_map);
 		sd->parent = p;
@@ -7124,8 +7161,15 @@ error:
 #endif
 }
 
+static int build_sched_domains(const cpumask_t *cpu_map)
+{
+	return __build_sched_domains(cpu_map, NULL);
+}
+
 static cpumask_t *doms_cur;	/* current sched domains */
 static int ndoms_cur;		/* number of sched domains in 'doms_cur' */
+static struct sched_domain_attr *dattr_cur;	/* attribues of custom domains
+						   in 'doms_cur' */
 
 /*
  * Special case: If a kmalloc of a doms_cur partition (array of
@@ -7153,6 +7197,7 @@ static int arch_init_sched_domains(const cpumask_t *cpu_map)
 	if (!doms_cur)
 		doms_cur = &fallback_doms;
 	cpus_andnot(*doms_cur, *cpu_map, cpu_isolated_map);
+	dattr_cur = NULL;
 	err = build_sched_domains(doms_cur);
 	register_sched_domain_sysctl();
 
@@ -7182,6 +7227,22 @@ static void detach_destroy_domains(const cpumask_t *cpu_map)
 	arch_destroy_sched_domains(cpu_map, &tmpmask);
 }
 
+/* handle null as "default" */
+static int dattrs_equal(struct sched_domain_attr *cur, int idx_cur,
+			struct sched_domain_attr *new, int idx_new)
+{
+	struct sched_domain_attr tmp;
+
+	/* fast path */
+	if (!new && !cur)
+		return 1;
+
+	tmp = SD_ATTR_INIT;
+	return !memcmp(cur ? (cur + idx_cur) : &tmp,
+			new ? (new + idx_new) : &tmp,
+			sizeof(struct sched_domain_attr));
+}
+
 /*
  * Partition sched domains as specified by the 'ndoms_new'
  * cpumasks in the array doms_new[] of cpumasks. This compares
@@ -7203,7 +7264,8 @@ static void detach_destroy_domains(const cpumask_t *cpu_map)
  *
  * Call with hotplug lock held
  */
-void partition_sched_domains(int ndoms_new, cpumask_t *doms_new)
+void partition_sched_domains(int ndoms_new, cpumask_t *doms_new,
+			     struct sched_domain_attr *dattr_new)
 {
 	int i, j;
 
@@ -7216,12 +7278,14 @@ void partition_sched_domains(int ndoms_new, cpumask_t *doms_new)
 		ndoms_new = 1;
 		doms_new = &fallback_doms;
 		cpus_andnot(doms_new[0], cpu_online_map, cpu_isolated_map);
+		dattr_new = NULL;
 	}
 
 	/* Destroy deleted domains */
 	for (i = 0; i < ndoms_cur; i++) {
 		for (j = 0; j < ndoms_new; j++) {
-			if (cpus_equal(doms_cur[i], doms_new[j]))
+			if (cpus_equal(doms_cur[i], doms_new[j])
+			    && dattrs_equal(dattr_cur, i, dattr_new, j))
 				goto match1;
 		}
 		/* no match - a current sched domain not in new doms_new[] */
@@ -7233,11 +7297,13 @@ match1:
 	/* Build new domains */
 	for (i = 0; i < ndoms_new; i++) {
 		for (j = 0; j < ndoms_cur; j++) {
-			if (cpus_equal(doms_new[i], doms_cur[j]))
+			if (cpus_equal(doms_new[i], doms_cur[j])
+			    && dattrs_equal(dattr_new, i, dattr_cur, j))
 				goto match2;
 		}
 		/* no match - add a new doms_new */
-		build_sched_domains(doms_new + i);
+		__build_sched_domains(doms_new + i,
+					dattr_new ? dattr_new + i : NULL);
 match2:
 		;
 	}
@@ -7245,7 +7311,9 @@ match2:
 	/* Remember the new sched domains */
 	if (doms_cur != &fallback_doms)
 		kfree(doms_cur);
+	kfree(dattr_cur);	/* kfree(NULL) is safe */
 	doms_cur = doms_new;
+	dattr_cur = dattr_new;
 	ndoms_cur = ndoms_new;
 
 	register_sched_domain_sysctl();
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index de4250c53a19..b43748efaa7f 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -940,7 +940,9 @@ static int wake_idle(int cpu, struct task_struct *p)
 		return cpu;
 
 	for_each_domain(cpu, sd) {
-		if (sd->flags & SD_WAKE_IDLE) {
+		if ((sd->flags & SD_WAKE_IDLE)
+		    || ((sd->flags & SD_WAKE_IDLE_FAR)
+			&& !task_hot(p, task_rq(p)->clock, sd))) {
 			cpus_and(tmp, sd->span, p->cpus_allowed);
 			for_each_cpu_mask(i, tmp) {
 				if (idle_cpu(i)) {
-- 
cgit v1.2.3


From 18d95a2832c1392a2d63227a7a6d433cb9f2037e Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Sat, 19 Apr 2008 19:45:00 +0200
Subject: sched: fair-group: SMP-nice for group scheduling

Implement SMP nice support for the full group hierarchy.

On each load-balance action, compile a sched_domain wide view of the full
task_group tree. We compute the domain wide view when walking down the
hierarchy, and readjust the weights when walking back up.

After collecting and readjusting the domain wide view, we try to balance the
tasks within the task_groups. The current approach is a naively balance each
task group until we've moved the targeted amount of load.

Inspired by Srivatsa Vaddsgiri's previous code and Abhishek Chandra's H-SMP
paper.

XXX: there will be some numerical issues due to the limited nature of
     SCHED_LOAD_SCALE wrt to representing a task_groups influence on the
     total weight. When the tree is deep enough, or the task weight small
     enough, we'll run out of bits.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
CC: Abhishek Chandra <chandra@cs.umn.edu>
CC: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |   1 +
 kernel/sched.c        | 497 ++++++++++++++++++++++++++++++++++++++++++++++----
 kernel/sched_fair.c   | 124 ++++++++-----
 kernel/sched_rt.c     |   4 +
 4 files changed, 548 insertions(+), 78 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 11f47249cdd2..0a32059e6ed4 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -758,6 +758,7 @@ struct sched_domain {
 	struct sched_domain *child;	/* bottom domain must be null terminated */
 	struct sched_group *groups;	/* the balancing groups of the domain */
 	cpumask_t span;			/* span of all CPUs in this domain */
+	int first_cpu;			/* cache of the first cpu in this domain */
 	unsigned long min_interval;	/* Minimum balance interval ms */
 	unsigned long max_interval;	/* Maximum balance interval ms */
 	unsigned int busy_factor;	/* less balancing by factor if busy */
diff --git a/kernel/sched.c b/kernel/sched.c
index 62d7481caca5..ae1a3e936d28 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -316,6 +316,8 @@ static DEFINE_MUTEX(doms_cur_mutex);
 # define INIT_TASK_GROUP_LOAD	NICE_0_LOAD
 #endif
 
+#define MIN_SHARES	2
+
 static int init_task_group_load = INIT_TASK_GROUP_LOAD;
 #endif
 
@@ -403,6 +405,43 @@ struct cfs_rq {
 	 */
 	struct list_head leaf_cfs_rq_list;
 	struct task_group *tg;	/* group that "owns" this runqueue */
+
+#ifdef CONFIG_SMP
+	unsigned long task_weight;
+	unsigned long shares;
+	/*
+	 * We need space to build a sched_domain wide view of the full task
+	 * group tree, in order to avoid depending on dynamic memory allocation
+	 * during the load balancing we place this in the per cpu task group
+	 * hierarchy. This limits the load balancing to one instance per cpu,
+	 * but more should not be needed anyway.
+	 */
+	struct aggregate_struct {
+		/*
+		 *   load = weight(cpus) * f(tg)
+		 *
+		 * Where f(tg) is the recursive weight fraction assigned to
+		 * this group.
+		 */
+		unsigned long load;
+
+		/*
+		 * part of the group weight distributed to this span.
+		 */
+		unsigned long shares;
+
+		/*
+		 * The sum of all runqueue weights within this span.
+		 */
+		unsigned long rq_weight;
+
+		/*
+		 * Weight contributed by tasks; this is the part we can
+		 * influence by moving tasks around.
+		 */
+		unsigned long task_weight;
+	} aggregate;
+#endif
 #endif
 };
 
@@ -1402,11 +1441,390 @@ static void cpuacct_charge(struct task_struct *tsk, u64 cputime);
 static inline void cpuacct_charge(struct task_struct *tsk, u64 cputime) {}
 #endif
 
+static inline void inc_cpu_load(struct rq *rq, unsigned long load)
+{
+	update_load_add(&rq->load, load);
+}
+
+static inline void dec_cpu_load(struct rq *rq, unsigned long load)
+{
+	update_load_sub(&rq->load, load);
+}
+
 #ifdef CONFIG_SMP
 static unsigned long source_load(int cpu, int type);
 static unsigned long target_load(int cpu, int type);
 static unsigned long cpu_avg_load_per_task(int cpu);
 static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd);
+
+#ifdef CONFIG_FAIR_GROUP_SCHED
+
+/*
+ * Group load balancing.
+ *
+ * We calculate a few balance domain wide aggregate numbers; load and weight.
+ * Given the pictures below, and assuming each item has equal weight:
+ *
+ *         root          1 - thread
+ *         / | \         A - group
+ *        A  1  B
+ *       /|\   / \
+ *      C 2 D 3   4
+ *      |   |
+ *      5   6
+ *
+ * load:
+ *    A and B get 1/3-rd of the total load. C and D get 1/3-rd of A's 1/3-rd,
+ *    which equals 1/9-th of the total load.
+ *
+ * shares:
+ *    The weight of this group on the selected cpus.
+ *
+ * rq_weight:
+ *    Direct sum of all the cpu's their rq weight, e.g. A would get 3 while
+ *    B would get 2.
+ *
+ * task_weight:
+ *    Part of the rq_weight contributed by tasks; all groups except B would
+ *    get 1, B gets 2.
+ */
+
+static inline struct aggregate_struct *
+aggregate(struct task_group *tg, struct sched_domain *sd)
+{
+	return &tg->cfs_rq[sd->first_cpu]->aggregate;
+}
+
+typedef void (*aggregate_func)(struct task_group *, struct sched_domain *);
+
+/*
+ * Iterate the full tree, calling @down when first entering a node and @up when
+ * leaving it for the final time.
+ */
+static
+void aggregate_walk_tree(aggregate_func down, aggregate_func up,
+			 struct sched_domain *sd)
+{
+	struct task_group *parent, *child;
+
+	rcu_read_lock();
+	parent = &root_task_group;
+down:
+	(*down)(parent, sd);
+	list_for_each_entry_rcu(child, &parent->children, siblings) {
+		parent = child;
+		goto down;
+
+up:
+		continue;
+	}
+	(*up)(parent, sd);
+
+	child = parent;
+	parent = parent->parent;
+	if (parent)
+		goto up;
+	rcu_read_unlock();
+}
+
+/*
+ * Calculate the aggregate runqueue weight.
+ */
+static
+void aggregate_group_weight(struct task_group *tg, struct sched_domain *sd)
+{
+	unsigned long rq_weight = 0;
+	unsigned long task_weight = 0;
+	int i;
+
+	for_each_cpu_mask(i, sd->span) {
+		rq_weight += tg->cfs_rq[i]->load.weight;
+		task_weight += tg->cfs_rq[i]->task_weight;
+	}
+
+	aggregate(tg, sd)->rq_weight = rq_weight;
+	aggregate(tg, sd)->task_weight = task_weight;
+}
+
+/*
+ * Redistribute tg->shares amongst all tg->cfs_rq[]s.
+ */
+static void __aggregate_redistribute_shares(struct task_group *tg)
+{
+	int i, max_cpu = smp_processor_id();
+	unsigned long rq_weight = 0;
+	unsigned long shares, max_shares = 0, shares_rem = tg->shares;
+
+	for_each_possible_cpu(i)
+		rq_weight += tg->cfs_rq[i]->load.weight;
+
+	for_each_possible_cpu(i) {
+		/*
+		 * divide shares proportional to the rq_weights.
+		 */
+		shares = tg->shares * tg->cfs_rq[i]->load.weight;
+		shares /= rq_weight + 1;
+
+		tg->cfs_rq[i]->shares = shares;
+
+		if (shares > max_shares) {
+			max_shares = shares;
+			max_cpu = i;
+		}
+		shares_rem -= shares;
+	}
+
+	/*
+	 * Ensure it all adds up to tg->shares; we can loose a few
+	 * due to rounding down when computing the per-cpu shares.
+	 */
+	if (shares_rem)
+		tg->cfs_rq[max_cpu]->shares += shares_rem;
+}
+
+/*
+ * Compute the weight of this group on the given cpus.
+ */
+static
+void aggregate_group_shares(struct task_group *tg, struct sched_domain *sd)
+{
+	unsigned long shares = 0;
+	int i;
+
+again:
+	for_each_cpu_mask(i, sd->span)
+		shares += tg->cfs_rq[i]->shares;
+
+	/*
+	 * When the span doesn't have any shares assigned, but does have
+	 * tasks to run do a machine wide rebalance (should be rare).
+	 */
+	if (unlikely(!shares && aggregate(tg, sd)->rq_weight)) {
+		__aggregate_redistribute_shares(tg);
+		goto again;
+	}
+
+	aggregate(tg, sd)->shares = shares;
+}
+
+/*
+ * Compute the load fraction assigned to this group, relies on the aggregate
+ * weight and this group's parent's load, i.e. top-down.
+ */
+static
+void aggregate_group_load(struct task_group *tg, struct sched_domain *sd)
+{
+	unsigned long load;
+
+	if (!tg->parent) {
+		int i;
+
+		load = 0;
+		for_each_cpu_mask(i, sd->span)
+			load += cpu_rq(i)->load.weight;
+
+	} else {
+		load = aggregate(tg->parent, sd)->load;
+
+		/*
+		 * shares is our weight in the parent's rq so
+		 * shares/parent->rq_weight gives our fraction of the load
+		 */
+		load *= aggregate(tg, sd)->shares;
+		load /= aggregate(tg->parent, sd)->rq_weight + 1;
+	}
+
+	aggregate(tg, sd)->load = load;
+}
+
+static void __set_se_shares(struct sched_entity *se, unsigned long shares);
+
+/*
+ * Calculate and set the cpu's group shares.
+ */
+static void
+__update_group_shares_cpu(struct task_group *tg, struct sched_domain *sd,
+			  int tcpu)
+{
+	int boost = 0;
+	unsigned long shares;
+	unsigned long rq_weight;
+
+	if (!tg->se[tcpu])
+		return;
+
+	rq_weight = tg->cfs_rq[tcpu]->load.weight;
+
+	/*
+	 * If there are currently no tasks on the cpu pretend there is one of
+	 * average load so that when a new task gets to run here it will not
+	 * get delayed by group starvation.
+	 */
+	if (!rq_weight) {
+		boost = 1;
+		rq_weight = NICE_0_LOAD;
+	}
+
+	/*
+	 *           \Sum shares * rq_weight
+	 * shares =  -----------------------
+	 *               \Sum rq_weight
+	 *
+	 */
+	shares = aggregate(tg, sd)->shares * rq_weight;
+	shares /= aggregate(tg, sd)->rq_weight + 1;
+
+	/*
+	 * record the actual number of shares, not the boosted amount.
+	 */
+	tg->cfs_rq[tcpu]->shares = boost ? 0 : shares;
+
+	if (shares < MIN_SHARES)
+		shares = MIN_SHARES;
+
+	__set_se_shares(tg->se[tcpu], shares);
+}
+
+/*
+ * Re-adjust the weights on the cpu the task came from and on the cpu the
+ * task went to.
+ */
+static void
+__move_group_shares(struct task_group *tg, struct sched_domain *sd,
+		    int scpu, int dcpu)
+{
+	unsigned long shares;
+
+	shares = tg->cfs_rq[scpu]->shares + tg->cfs_rq[dcpu]->shares;
+
+	__update_group_shares_cpu(tg, sd, scpu);
+	__update_group_shares_cpu(tg, sd, dcpu);
+
+	/*
+	 * ensure we never loose shares due to rounding errors in the
+	 * above redistribution.
+	 */
+	shares -= tg->cfs_rq[scpu]->shares + tg->cfs_rq[dcpu]->shares;
+	if (shares)
+		tg->cfs_rq[dcpu]->shares += shares;
+}
+
+/*
+ * Because changing a group's shares changes the weight of the super-group
+ * we need to walk up the tree and change all shares until we hit the root.
+ */
+static void
+move_group_shares(struct task_group *tg, struct sched_domain *sd,
+		  int scpu, int dcpu)
+{
+	while (tg) {
+		__move_group_shares(tg, sd, scpu, dcpu);
+		tg = tg->parent;
+	}
+}
+
+static
+void aggregate_group_set_shares(struct task_group *tg, struct sched_domain *sd)
+{
+	unsigned long shares = aggregate(tg, sd)->shares;
+	int i;
+
+	for_each_cpu_mask(i, sd->span) {
+		struct rq *rq = cpu_rq(i);
+		unsigned long flags;
+
+		spin_lock_irqsave(&rq->lock, flags);
+		__update_group_shares_cpu(tg, sd, i);
+		spin_unlock_irqrestore(&rq->lock, flags);
+	}
+
+	aggregate_group_shares(tg, sd);
+
+	/*
+	 * ensure we never loose shares due to rounding errors in the
+	 * above redistribution.
+	 */
+	shares -= aggregate(tg, sd)->shares;
+	if (shares) {
+		tg->cfs_rq[sd->first_cpu]->shares += shares;
+		aggregate(tg, sd)->shares += shares;
+	}
+}
+
+/*
+ * Calculate the accumulative weight and recursive load of each task group
+ * while walking down the tree.
+ */
+static
+void aggregate_get_down(struct task_group *tg, struct sched_domain *sd)
+{
+	aggregate_group_weight(tg, sd);
+	aggregate_group_shares(tg, sd);
+	aggregate_group_load(tg, sd);
+}
+
+/*
+ * Rebalance the cpu shares while walking back up the tree.
+ */
+static
+void aggregate_get_up(struct task_group *tg, struct sched_domain *sd)
+{
+	aggregate_group_set_shares(tg, sd);
+}
+
+static DEFINE_PER_CPU(spinlock_t, aggregate_lock);
+
+static void __init init_aggregate(void)
+{
+	int i;
+
+	for_each_possible_cpu(i)
+		spin_lock_init(&per_cpu(aggregate_lock, i));
+}
+
+static int get_aggregate(struct sched_domain *sd)
+{
+	if (!spin_trylock(&per_cpu(aggregate_lock, sd->first_cpu)))
+		return 0;
+
+	aggregate_walk_tree(aggregate_get_down, aggregate_get_up, sd);
+	return 1;
+}
+
+static void put_aggregate(struct sched_domain *sd)
+{
+	spin_unlock(&per_cpu(aggregate_lock, sd->first_cpu));
+}
+
+static void cfs_rq_set_shares(struct cfs_rq *cfs_rq, unsigned long shares)
+{
+	cfs_rq->shares = shares;
+}
+
+#else
+
+static inline void init_aggregate(void)
+{
+}
+
+static inline int get_aggregate(struct sched_domain *sd)
+{
+	return 0;
+}
+
+static inline void put_aggregate(struct sched_domain *sd)
+{
+}
+#endif
+
+#else /* CONFIG_SMP */
+
+#ifdef CONFIG_FAIR_GROUP_SCHED
+static void cfs_rq_set_shares(struct cfs_rq *cfs_rq, unsigned long shares)
+{
+}
+#endif
+
 #endif /* CONFIG_SMP */
 
 #include "sched_stats.h"
@@ -1419,26 +1837,14 @@ static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd);
 
 #define sched_class_highest (&rt_sched_class)
 
-static inline void inc_load(struct rq *rq, const struct task_struct *p)
-{
-	update_load_add(&rq->load, p->se.load.weight);
-}
-
-static inline void dec_load(struct rq *rq, const struct task_struct *p)
-{
-	update_load_sub(&rq->load, p->se.load.weight);
-}
-
-static void inc_nr_running(struct task_struct *p, struct rq *rq)
+static void inc_nr_running(struct rq *rq)
 {
 	rq->nr_running++;
-	inc_load(rq, p);
 }
 
-static void dec_nr_running(struct task_struct *p, struct rq *rq)
+static void dec_nr_running(struct rq *rq)
 {
 	rq->nr_running--;
-	dec_load(rq, p);
 }
 
 static void set_load_weight(struct task_struct *p)
@@ -1530,7 +1936,7 @@ static void activate_task(struct rq *rq, struct task_struct *p, int wakeup)
 		rq->nr_uninterruptible--;
 
 	enqueue_task(rq, p, wakeup);
-	inc_nr_running(p, rq);
+	inc_nr_running(rq);
 }
 
 /*
@@ -1542,7 +1948,7 @@ static void deactivate_task(struct rq *rq, struct task_struct *p, int sleep)
 		rq->nr_uninterruptible++;
 
 	dequeue_task(rq, p, sleep);
-	dec_nr_running(p, rq);
+	dec_nr_running(rq);
 }
 
 /**
@@ -2194,7 +2600,7 @@ void wake_up_new_task(struct task_struct *p, unsigned long clone_flags)
 		 * management (if any):
 		 */
 		p->sched_class->task_new(rq, p);
-		inc_nr_running(p, rq);
+		inc_nr_running(rq);
 	}
 	check_preempt_curr(rq, p);
 #ifdef CONFIG_SMP
@@ -3185,9 +3591,12 @@ static int load_balance(int this_cpu, struct rq *this_rq,
 	unsigned long imbalance;
 	struct rq *busiest;
 	unsigned long flags;
+	int unlock_aggregate;
 
 	cpus_setall(*cpus);
 
+	unlock_aggregate = get_aggregate(sd);
+
 	/*
 	 * When power savings policy is enabled for the parent domain, idle
 	 * sibling can pick up load irrespective of busy siblings. In this case,
@@ -3303,8 +3712,9 @@ redo:
 
 	if (!ld_moved && !sd_idle && sd->flags & SD_SHARE_CPUPOWER &&
 	    !test_sd_parent(sd, SD_POWERSAVINGS_BALANCE))
-		return -1;
-	return ld_moved;
+		ld_moved = -1;
+
+	goto out;
 
 out_balanced:
 	schedstat_inc(sd, lb_balanced[idle]);
@@ -3319,8 +3729,13 @@ out_one_pinned:
 
 	if (!sd_idle && sd->flags & SD_SHARE_CPUPOWER &&
 	    !test_sd_parent(sd, SD_POWERSAVINGS_BALANCE))
-		return -1;
-	return 0;
+		ld_moved = -1;
+	else
+		ld_moved = 0;
+out:
+	if (unlock_aggregate)
+		put_aggregate(sd);
+	return ld_moved;
 }
 
 /*
@@ -4535,10 +4950,8 @@ void set_user_nice(struct task_struct *p, long nice)
 		goto out_unlock;
 	}
 	on_rq = p->se.on_rq;
-	if (on_rq) {
+	if (on_rq)
 		dequeue_task(rq, p, 0);
-		dec_load(rq, p);
-	}
 
 	p->static_prio = NICE_TO_PRIO(nice);
 	set_load_weight(p);
@@ -4548,7 +4961,6 @@ void set_user_nice(struct task_struct *p, long nice)
 
 	if (on_rq) {
 		enqueue_task(rq, p, 0);
-		inc_load(rq, p);
 		/*
 		 * If the task increased its priority or is running and
 		 * lowered its priority, then reschedule its CPU:
@@ -6921,6 +7333,7 @@ static int __build_sched_domains(const cpumask_t *cpu_map,
 			SD_INIT(sd, ALLNODES);
 			set_domain_attribute(sd, attr);
 			sd->span = *cpu_map;
+			sd->first_cpu = first_cpu(sd->span);
 			cpu_to_allnodes_group(i, cpu_map, &sd->groups, tmpmask);
 			p = sd;
 			sd_allnodes = 1;
@@ -6931,6 +7344,7 @@ static int __build_sched_domains(const cpumask_t *cpu_map,
 		SD_INIT(sd, NODE);
 		set_domain_attribute(sd, attr);
 		sched_domain_node_span(cpu_to_node(i), &sd->span);
+		sd->first_cpu = first_cpu(sd->span);
 		sd->parent = p;
 		if (p)
 			p->child = sd;
@@ -6942,6 +7356,7 @@ static int __build_sched_domains(const cpumask_t *cpu_map,
 		SD_INIT(sd, CPU);
 		set_domain_attribute(sd, attr);
 		sd->span = *nodemask;
+		sd->first_cpu = first_cpu(sd->span);
 		sd->parent = p;
 		if (p)
 			p->child = sd;
@@ -6953,6 +7368,7 @@ static int __build_sched_domains(const cpumask_t *cpu_map,
 		SD_INIT(sd, MC);
 		set_domain_attribute(sd, attr);
 		sd->span = cpu_coregroup_map(i);
+		sd->first_cpu = first_cpu(sd->span);
 		cpus_and(sd->span, sd->span, *cpu_map);
 		sd->parent = p;
 		p->child = sd;
@@ -6965,6 +7381,7 @@ static int __build_sched_domains(const cpumask_t *cpu_map,
 		SD_INIT(sd, SIBLING);
 		set_domain_attribute(sd, attr);
 		sd->span = per_cpu(cpu_sibling_map, i);
+		sd->first_cpu = first_cpu(sd->span);
 		cpus_and(sd->span, sd->span, *cpu_map);
 		sd->parent = p;
 		p->child = sd;
@@ -7633,6 +8050,7 @@ void __init sched_init(void)
 	}
 
 #ifdef CONFIG_SMP
+	init_aggregate();
 	init_defrootdomain();
 #endif
 
@@ -8199,14 +8617,11 @@ void sched_move_task(struct task_struct *tsk)
 #endif
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
-static void set_se_shares(struct sched_entity *se, unsigned long shares)
+static void __set_se_shares(struct sched_entity *se, unsigned long shares)
 {
 	struct cfs_rq *cfs_rq = se->cfs_rq;
-	struct rq *rq = cfs_rq->rq;
 	int on_rq;
 
-	spin_lock_irq(&rq->lock);
-
 	on_rq = se->on_rq;
 	if (on_rq)
 		dequeue_entity(cfs_rq, se, 0);
@@ -8216,8 +8631,17 @@ static void set_se_shares(struct sched_entity *se, unsigned long shares)
 
 	if (on_rq)
 		enqueue_entity(cfs_rq, se, 0);
+}
 
-	spin_unlock_irq(&rq->lock);
+static void set_se_shares(struct sched_entity *se, unsigned long shares)
+{
+	struct cfs_rq *cfs_rq = se->cfs_rq;
+	struct rq *rq = cfs_rq->rq;
+	unsigned long flags;
+
+	spin_lock_irqsave(&rq->lock, flags);
+	__set_se_shares(se, shares);
+	spin_unlock_irqrestore(&rq->lock, flags);
 }
 
 static DEFINE_MUTEX(shares_mutex);
@@ -8238,8 +8662,8 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares)
 	 * (The default weight is 1024 - so there's no practical
 	 *  limitation from this.)
 	 */
-	if (shares < 2)
-		shares = 2;
+	if (shares < MIN_SHARES)
+		shares = MIN_SHARES;
 
 	mutex_lock(&shares_mutex);
 	if (tg->shares == shares)
@@ -8259,8 +8683,13 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares)
 	 * w/o tripping rebalance_share or load_balance_fair.
 	 */
 	tg->shares = shares;
-	for_each_possible_cpu(i)
-		set_se_shares(tg->se[i], shares);
+	for_each_possible_cpu(i) {
+		/*
+		 * force a rebalance
+		 */
+		cfs_rq_set_shares(tg->cfs_rq[i], 0);
+		set_se_shares(tg->se[i], shares/nr_cpu_ids);
+	}
 
 	/*
 	 * Enable load balance activity on this group, by inserting it back on
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index b43748efaa7f..b89fec93a237 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -492,10 +492,27 @@ update_stats_curr_start(struct cfs_rq *cfs_rq, struct sched_entity *se)
  * Scheduling class queueing methods:
  */
 
+#if defined CONFIG_SMP && defined CONFIG_FAIR_GROUP_SCHED
+static void
+add_cfs_task_weight(struct cfs_rq *cfs_rq, unsigned long weight)
+{
+	cfs_rq->task_weight += weight;
+}
+#else
+static inline void
+add_cfs_task_weight(struct cfs_rq *cfs_rq, unsigned long weight)
+{
+}
+#endif
+
 static void
 account_entity_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se)
 {
 	update_load_add(&cfs_rq->load, se->load.weight);
+	if (!parent_entity(se))
+		inc_cpu_load(rq_of(cfs_rq), se->load.weight);
+	if (entity_is_task(se))
+		add_cfs_task_weight(cfs_rq, se->load.weight);
 	cfs_rq->nr_running++;
 	se->on_rq = 1;
 }
@@ -504,6 +521,10 @@ static void
 account_entity_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se)
 {
 	update_load_sub(&cfs_rq->load, se->load.weight);
+	if (!parent_entity(se))
+		dec_cpu_load(rq_of(cfs_rq), se->load.weight);
+	if (entity_is_task(se))
+		add_cfs_task_weight(cfs_rq, -se->load.weight);
 	cfs_rq->nr_running--;
 	se->on_rq = 0;
 }
@@ -1286,75 +1307,90 @@ static struct task_struct *load_balance_next_fair(void *arg)
 	return __load_balance_iterator(cfs_rq, cfs_rq->rb_load_balance_curr);
 }
 
-#ifdef CONFIG_FAIR_GROUP_SCHED
-static int cfs_rq_best_prio(struct cfs_rq *cfs_rq)
+static unsigned long
+__load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
+		unsigned long max_load_move, struct sched_domain *sd,
+		enum cpu_idle_type idle, int *all_pinned, int *this_best_prio,
+		struct cfs_rq *cfs_rq)
 {
-	struct sched_entity *curr;
-	struct task_struct *p;
-
-	if (!cfs_rq->nr_running || !first_fair(cfs_rq))
-		return MAX_PRIO;
-
-	curr = cfs_rq->curr;
-	if (!curr)
-		curr = __pick_next_entity(cfs_rq);
+	struct rq_iterator cfs_rq_iterator;
 
-	p = task_of(curr);
+	cfs_rq_iterator.start = load_balance_start_fair;
+	cfs_rq_iterator.next = load_balance_next_fair;
+	cfs_rq_iterator.arg = cfs_rq;
 
-	return p->prio;
+	return balance_tasks(this_rq, this_cpu, busiest,
+			max_load_move, sd, idle, all_pinned,
+			this_best_prio, &cfs_rq_iterator);
 }
-#endif
 
+#ifdef CONFIG_FAIR_GROUP_SCHED
 static unsigned long
 load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
 		  unsigned long max_load_move,
 		  struct sched_domain *sd, enum cpu_idle_type idle,
 		  int *all_pinned, int *this_best_prio)
 {
-	struct cfs_rq *busy_cfs_rq;
 	long rem_load_move = max_load_move;
-	struct rq_iterator cfs_rq_iterator;
+	int busiest_cpu = cpu_of(busiest);
+	struct task_group *tg;
 
-	cfs_rq_iterator.start = load_balance_start_fair;
-	cfs_rq_iterator.next = load_balance_next_fair;
-
-	for_each_leaf_cfs_rq(busiest, busy_cfs_rq) {
-#ifdef CONFIG_FAIR_GROUP_SCHED
-		struct cfs_rq *this_cfs_rq;
+	rcu_read_lock();
+	list_for_each_entry(tg, &task_groups, list) {
 		long imbalance;
-		unsigned long maxload;
+		unsigned long this_weight, busiest_weight;
+		long rem_load, max_load, moved_load;
+
+		/*
+		 * empty group
+		 */
+		if (!aggregate(tg, sd)->task_weight)
+			continue;
+
+		rem_load = rem_load_move * aggregate(tg, sd)->rq_weight;
+		rem_load /= aggregate(tg, sd)->load + 1;
 
-		this_cfs_rq = cpu_cfs_rq(busy_cfs_rq, this_cpu);
+		this_weight = tg->cfs_rq[this_cpu]->task_weight;
+		busiest_weight = tg->cfs_rq[busiest_cpu]->task_weight;
 
-		imbalance = busy_cfs_rq->load.weight - this_cfs_rq->load.weight;
-		/* Don't pull if this_cfs_rq has more load than busy_cfs_rq */
-		if (imbalance <= 0)
+		imbalance = (busiest_weight - this_weight) / 2;
+
+		if (imbalance < 0)
+			imbalance = busiest_weight;
+
+		max_load = max(rem_load, imbalance);
+		moved_load = __load_balance_fair(this_rq, this_cpu, busiest,
+				max_load, sd, idle, all_pinned, this_best_prio,
+				tg->cfs_rq[busiest_cpu]);
+
+		if (!moved_load)
 			continue;
 
-		/* Don't pull more than imbalance/2 */
-		imbalance /= 2;
-		maxload = min(rem_load_move, imbalance);
+		move_group_shares(tg, sd, busiest_cpu, this_cpu);
 
-		*this_best_prio = cfs_rq_best_prio(this_cfs_rq);
-#else
-# define maxload rem_load_move
-#endif
-		/*
-		 * pass busy_cfs_rq argument into
-		 * load_balance_[start|next]_fair iterators
-		 */
-		cfs_rq_iterator.arg = busy_cfs_rq;
-		rem_load_move -= balance_tasks(this_rq, this_cpu, busiest,
-					       maxload, sd, idle, all_pinned,
-					       this_best_prio,
-					       &cfs_rq_iterator);
+		moved_load *= aggregate(tg, sd)->load;
+		moved_load /= aggregate(tg, sd)->rq_weight + 1;
 
-		if (rem_load_move <= 0)
+		rem_load_move -= moved_load;
+		if (rem_load_move < 0)
 			break;
 	}
+	rcu_read_unlock();
 
 	return max_load_move - rem_load_move;
 }
+#else
+static unsigned long
+load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
+		  unsigned long max_load_move,
+		  struct sched_domain *sd, enum cpu_idle_type idle,
+		  int *all_pinned, int *this_best_prio)
+{
+	return __load_balance_fair(this_rq, this_cpu, busiest,
+			max_load_move, sd, idle, all_pinned,
+			this_best_prio, &busiest->cfs);
+}
+#endif
 
 static int
 move_one_task_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 201a69382a42..736fb8fd8977 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -518,6 +518,8 @@ static void enqueue_task_rt(struct rq *rq, struct task_struct *p, int wakeup)
 	 */
 	for_each_sched_rt_entity(rt_se)
 		enqueue_rt_entity(rt_se);
+
+	inc_cpu_load(rq, p->se.load.weight);
 }
 
 static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int sleep)
@@ -537,6 +539,8 @@ static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int sleep)
 		if (rt_rq && rt_rq->rt_nr_running)
 			enqueue_rt_entity(rt_se);
 	}
+
+	dec_cpu_load(rq, p->se.load.weight);
 }
 
 /*
-- 
cgit v1.2.3


From 58d6c2d72f8628f39e8689fbde8aa177fcf00a37 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Sat, 19 Apr 2008 19:45:00 +0200
Subject: sched: rt-group: optimize dequeue_rt_stack

Now that the group hierarchy can have an arbitrary depth the O(n^2) nature
of RT task dequeues will really hurt. Optimize this by providing space to
store the tree path, so we can walk it the other way.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/sched.h |  1 +
 kernel/sched_rt.c     | 27 +++++++++++----------------
 2 files changed, 12 insertions(+), 16 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 0a32059e6ed4..887f5db8942d 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1005,6 +1005,7 @@ struct sched_rt_entity {
 	unsigned long timeout;
 	int nr_cpus_allowed;
 
+	struct sched_rt_entity *back;
 #ifdef CONFIG_RT_GROUP_SCHED
 	struct sched_rt_entity	*parent;
 	/* rq on which this entity is (to be) queued: */
diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 736fb8fd8977..c2730a5a4f05 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -479,26 +479,21 @@ static void dequeue_rt_entity(struct sched_rt_entity *rt_se)
 /*
  * Because the prio of an upper entry depends on the lower
  * entries, we must remove entries top - down.
- *
- * XXX: O(1/2 h^2) because we can only walk up, not down the chain.
  */
 static void dequeue_rt_stack(struct task_struct *p)
 {
-	struct sched_rt_entity *rt_se, *top_se;
+	struct sched_rt_entity *rt_se, *back = NULL;
 
-	/*
-	 * dequeue all, top - down.
-	 */
-	do {
-		rt_se = &p->rt;
-		top_se = NULL;
-		for_each_sched_rt_entity(rt_se) {
-			if (on_rt_rq(rt_se))
-				top_se = rt_se;
-		}
-		if (top_se)
-			dequeue_rt_entity(top_se);
-	} while (top_se);
+	rt_se = &p->rt;
+	for_each_sched_rt_entity(rt_se) {
+		rt_se->back = back;
+		back = rt_se;
+	}
+
+	for (rt_se = back; rt_se; rt_se = rt_se->back) {
+		if (on_rt_rq(rt_se))
+			dequeue_rt_entity(rt_se);
+	}
 }
 
 /*
-- 
cgit v1.2.3


From 4a55bd5e97b1775913f88f11108a4f144f590e89 Mon Sep 17 00:00:00 2001
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Sat, 19 Apr 2008 19:45:00 +0200
Subject: sched: fair-group: de-couple load-balancing from the rb-trees

De-couple load-balancing from the rb-trees, so that I can change their
organization.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 include/linux/init_task.h |  3 +++
 include/linux/sched.h     |  1 +
 kernel/sched.c            | 10 ++++++++--
 kernel/sched_fair.c       | 21 +++++++++++++--------
 4 files changed, 25 insertions(+), 10 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/init_task.h b/include/linux/init_task.h
index 1f74e1d7415f..37a6f5bc4a92 100644
--- a/include/linux/init_task.h
+++ b/include/linux/init_task.h
@@ -151,6 +151,9 @@ extern struct group_info init_groups;
 	.cpus_allowed	= CPU_MASK_ALL,					\
 	.mm		= NULL,						\
 	.active_mm	= &init_mm,					\
+	.se		= {						\
+		.group_node 	= LIST_HEAD_INIT(tsk.se.group_node),	\
+	},								\
 	.rt		= {						\
 		.run_list	= LIST_HEAD_INIT(tsk.rt.run_list),	\
 		.time_slice	= HZ, 					\
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 887f5db8942d..be6914014c70 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -946,6 +946,7 @@ struct load_weight {
 struct sched_entity {
 	struct load_weight	load;		/* for load-balancing */
 	struct rb_node		run_node;
+	struct list_head	group_node;
 	unsigned int		on_rq;
 
 	u64			exec_start;
diff --git a/kernel/sched.c b/kernel/sched.c
index ae1a3e936d28..3202462109f5 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -384,8 +384,12 @@ struct cfs_rq {
 
 	struct rb_root tasks_timeline;
 	struct rb_node *rb_leftmost;
-	struct rb_node *rb_load_balance_curr;
-	/* 'curr' points to currently running entity on this cfs_rq.
+
+	struct list_head tasks;
+	struct list_head *balance_iterator;
+
+	/*
+	 * 'curr' points to currently running entity on this cfs_rq.
 	 * It is set to NULL otherwise (i.e when none are currently running).
 	 */
 	struct sched_entity *curr, *next;
@@ -2525,6 +2529,7 @@ static void __sched_fork(struct task_struct *p)
 
 	INIT_LIST_HEAD(&p->rt.run_list);
 	p->se.on_rq = 0;
+	INIT_LIST_HEAD(&p->se.group_node);
 
 #ifdef CONFIG_PREEMPT_NOTIFIERS
 	INIT_HLIST_HEAD(&p->preempt_notifiers);
@@ -7898,6 +7903,7 @@ int in_sched_functions(unsigned long addr)
 static void init_cfs_rq(struct cfs_rq *cfs_rq, struct rq *rq)
 {
 	cfs_rq->tasks_timeline = RB_ROOT;
+	INIT_LIST_HEAD(&cfs_rq->tasks);
 #ifdef CONFIG_FAIR_GROUP_SCHED
 	cfs_rq->rq = rq;
 #endif
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index 9e301a2bab6f..ed8ce329899b 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -533,6 +533,7 @@ account_entity_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se)
 		add_cfs_task_weight(cfs_rq, se->load.weight);
 	cfs_rq->nr_running++;
 	se->on_rq = 1;
+	list_add(&se->group_node, &cfs_rq->tasks);
 }
 
 static void
@@ -545,6 +546,7 @@ account_entity_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se)
 		add_cfs_task_weight(cfs_rq, -se->load.weight);
 	cfs_rq->nr_running--;
 	se->on_rq = 0;
+	list_del_init(&se->group_node);
 }
 
 static void enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se)
@@ -1289,21 +1291,24 @@ static void put_prev_task_fair(struct rq *rq, struct task_struct *prev)
  * the current task:
  */
 static struct task_struct *
-__load_balance_iterator(struct cfs_rq *cfs_rq, struct rb_node *curr)
+__load_balance_iterator(struct cfs_rq *cfs_rq, struct list_head *next)
 {
 	struct task_struct *p = NULL;
 	struct sched_entity *se;
 
-	if (!curr)
+	if (next == &cfs_rq->tasks)
 		return NULL;
 
 	/* Skip over entities that are not tasks */
 	do {
-		se = rb_entry(curr, struct sched_entity, run_node);
-		curr = rb_next(curr);
-	} while (curr && !entity_is_task(se));
+		se = list_entry(next, struct sched_entity, group_node);
+		next = next->next;
+	} while (next != &cfs_rq->tasks && !entity_is_task(se));
 
-	cfs_rq->rb_load_balance_curr = curr;
+	if (next == &cfs_rq->tasks)
+		return NULL;
+
+	cfs_rq->balance_iterator = next;
 
 	if (entity_is_task(se))
 		p = task_of(se);
@@ -1315,14 +1320,14 @@ static struct task_struct *load_balance_start_fair(void *arg)
 {
 	struct cfs_rq *cfs_rq = arg;
 
-	return __load_balance_iterator(cfs_rq, first_fair(cfs_rq));
+	return __load_balance_iterator(cfs_rq, cfs_rq->tasks.next);
 }
 
 static struct task_struct *load_balance_next_fair(void *arg)
 {
 	struct cfs_rq *cfs_rq = arg;
 
-	return __load_balance_iterator(cfs_rq, cfs_rq->rb_load_balance_curr);
+	return __load_balance_iterator(cfs_rq, cfs_rq->balance_iterator);
 }
 
 static unsigned long
-- 
cgit v1.2.3


From da19cbcf71cde3c09587b5924d113f0c7f1fd23a Mon Sep 17 00:00:00 2001
From: Daniel Walker <dwalker@mvista.com>
Date: Mon, 4 Feb 2008 23:35:47 -0800
Subject: driver core: memory: semaphore to mutex

Signed-off-by: Daniel Walker <dwalker@mvista.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/base/memory.c  | 7 ++++---
 include/linux/memory.h | 5 ++---
 mm/memory_hotplug.c    | 2 +-
 3 files changed, 7 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/base/memory.c b/drivers/base/memory.c
index 7ae413fdd5fc..1f3801a8184d 100644
--- a/drivers/base/memory.c
+++ b/drivers/base/memory.c
@@ -20,6 +20,7 @@
 #include <linux/kobject.h>
 #include <linux/memory_hotplug.h>
 #include <linux/mm.h>
+#include <linux/mutex.h>
 #include <asm/atomic.h>
 #include <asm/uaccess.h>
 
@@ -205,7 +206,7 @@ static int memory_block_change_state(struct memory_block *mem,
 		unsigned long to_state, unsigned long from_state_req)
 {
 	int ret = 0;
-	down(&mem->state_sem);
+	mutex_lock(&mem->state_mutex);
 
 	if (mem->state != from_state_req) {
 		ret = -EINVAL;
@@ -217,7 +218,7 @@ static int memory_block_change_state(struct memory_block *mem,
 		mem->state = to_state;
 
 out:
-	up(&mem->state_sem);
+	mutex_unlock(&mem->state_mutex);
 	return ret;
 }
 
@@ -341,7 +342,7 @@ static int add_memory_block(unsigned long node_id, struct mem_section *section,
 
 	mem->phys_index = __section_nr(section);
 	mem->state = state;
-	init_MUTEX(&mem->state_sem);
+	mutex_init(&mem->state_mutex);
 	mem->phys_device = phys_device;
 
 	ret = register_memory(mem, section, NULL);
diff --git a/include/linux/memory.h b/include/linux/memory.h
index 33f0ff0cf634..f80e0e331cb7 100644
--- a/include/linux/memory.h
+++ b/include/linux/memory.h
@@ -18,8 +18,7 @@
 #include <linux/sysdev.h>
 #include <linux/node.h>
 #include <linux/compiler.h>
-
-#include <asm/semaphore.h>
+#include <linux/mutex.h>
 
 struct memory_block {
 	unsigned long phys_index;
@@ -30,7 +29,7 @@ struct memory_block {
 	 * created long after the critical areas during
 	 * initialization.
 	 */
-	struct semaphore state_sem;
+	struct mutex state_mutex;
 	int phys_device;		/* to which fru does this belong? */
 	void *hw;			/* optional pointer to fw/hw data */
 	int (*phys_callback)(struct memory_block *);
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index 7469c503580d..0fb330271271 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -208,7 +208,7 @@ int online_pages(unsigned long pfn, unsigned long nr_pages)
 	/*
 	 * This doesn't need a lock to do pfn_to_page().
 	 * The section can't be removed here because of the
-	 * memory_block->state_sem.
+	 * memory_block->state_mutex.
 	 */
 	zone = page_zone(pfn_to_page(pfn));
 	pgdat_resize_lock(zone->zone_pgdat, &flags);
-- 
cgit v1.2.3


From 1429db83e276c2a16c7ea83bdcf0dcd3a36e406d Mon Sep 17 00:00:00 2001
From: Joe Perches <joe@perches.com>
Date: Tue, 26 Feb 2008 19:08:42 -0800
Subject: driver core: Convert debug functions declared inline
 __attribute__((format (printf,x,y) to statement expression macros

When DEBUG is not defined, pr_debug and dev_dbg and some
other local debugging functions are specified as:

"inline __attribute__((format (printf, x, y)))"

This is done to validate printk arguments when not debugging.

Converting these functions to macros or statement expressions
"do { if (0) printk(fmt, ##arg); } while (0)"
or
"({ if (0) printk(fmt, ##arg); 0; })
makes at least gcc 4.2.2 produce smaller objects.

This has the additional benefit of allowing the optimizer to
avoid calling functions like print_mac that might have been
arguments to the printk.

defconfig x86 current:

$ size vmlinux
   text    data     bss     dec     hex filename
4716770  474560  618496 5809826  58a6a2 vmlinux

all converted: (More patches follow)

$ size vmlinux
   text    data     bss     dec     hex filename
4716642  474560  618496 5809698  58a622 vmlinux

Even kernel/sched.o, which doesn't even use these
functions, becomes smaller.

It appears that merely having an indirect include
of <linux/device.h> can cause bigger objects.

$ size sched.inline.o sched.if0.o
   text    data     bss     dec     hex filename
  31385    2854     328   34567    8707 sched.inline.o
  31366    2854     328   34548    86f4 sched.if0.o

The current preprocessed only kernel/sched.i file contains:

# 612 "include/linux/device.h"
static inline __attribute__((always_inline)) int __attribute__ ((format (printf, 2, 3)))
dev_dbg(struct device *dev, const char *fmt, ...)
{
 return 0;
}
# 628 "include/linux/device.h"
static inline __attribute__((always_inline)) int __attribute__ ((format (printf, 2, 3)))
dev_vdbg(struct device *dev, const char *fmt, ...)
{
 return 0;
}

Removing these unused inlines from sched.i shrinks sched.o

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/device.h | 15 +++++----------
 include/linux/kernel.h |  6 ++----
 2 files changed, 7 insertions(+), 14 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/device.h b/include/linux/device.h
index 2258d89bf523..d57661129cb2 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -608,21 +608,16 @@ extern const char *dev_driver_string(struct device *dev);
 #define dev_dbg(dev, format, arg...)		\
 	dev_printk(KERN_DEBUG , dev , format , ## arg)
 #else
-static inline int __attribute__ ((format (printf, 2, 3)))
-dev_dbg(struct device *dev, const char *fmt, ...)
-{
-	return 0;
-}
+#define dev_dbg(dev, format, arg...)		\
+	({ if (0) dev_printk(KERN_DEBUG, dev, format, ##arg); 0; })
 #endif
 
 #ifdef VERBOSE_DEBUG
 #define dev_vdbg	dev_dbg
 #else
-static inline int __attribute__ ((format (printf, 2, 3)))
-dev_vdbg(struct device *dev, const char *fmt, ...)
-{
-	return 0;
-}
+
+#define dev_vdbg(dev, format, arg...)		\
+	({ if (0) dev_printk(KERN_DEBUG, dev, format, ##arg); 0; })
 #endif
 
 /* Create alias, so I can be autoloaded. */
diff --git a/include/linux/kernel.h b/include/linux/kernel.h
index 2df44e773270..cd6d02cf854d 100644
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -293,10 +293,8 @@ extern void print_hex_dump_bytes(const char *prefix_str, int prefix_type,
 #define pr_debug(fmt, arg...) \
 	printk(KERN_DEBUG fmt, ##arg)
 #else
-static inline int __attribute__ ((format (printf, 1, 2))) pr_debug(const char * fmt, ...)
-{
-	return 0;
-}
+#define pr_debug(fmt, arg...) \
+	({ if (0) printk(KERN_DEBUG fmt, ##arg); 0; })
 #endif
 
 /*
-- 
cgit v1.2.3


From 3612e06b2c1cec41e9a59da3eec673a206af4643 Mon Sep 17 00:00:00 2001
From: David Rientjes <rientjes@google.com>
Date: Tue, 19 Feb 2008 17:39:02 -0800
Subject: sysfs: small header file cleanup for SYSFS=n

Convert sysfs_remove_bin_file() to have a return type of 'void' for
!CONFIG_SYSFS configurations.  Also removes unnecessary colons from empty
void functions.

Signed-off-by: David Rientjes <rientjes@google.com>
Reviewed-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/sysfs.h | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/sysfs.h b/include/linux/sysfs.h
index 802710438a9e..03378e3515b3 100644
--- a/include/linux/sysfs.h
+++ b/include/linux/sysfs.h
@@ -131,7 +131,6 @@ static inline int sysfs_create_dir(struct kobject *kobj)
 
 static inline void sysfs_remove_dir(struct kobject *kobj)
 {
-	;
 }
 
 static inline int sysfs_rename_dir(struct kobject *kobj, const char *new_name)
@@ -160,7 +159,6 @@ static inline int sysfs_chmod_file(struct kobject *kobj,
 static inline void sysfs_remove_file(struct kobject *kobj,
 				     const struct attribute *attr)
 {
-	;
 }
 
 static inline int sysfs_create_bin_file(struct kobject *kobj,
@@ -169,10 +167,9 @@ static inline int sysfs_create_bin_file(struct kobject *kobj,
 	return 0;
 }
 
-static inline int sysfs_remove_bin_file(struct kobject *kobj,
-					struct bin_attribute *attr)
+static inline void sysfs_remove_bin_file(struct kobject *kobj,
+					 struct bin_attribute *attr)
 {
-	return 0;
 }
 
 static inline int sysfs_create_link(struct kobject *kobj,
@@ -183,7 +180,6 @@ static inline int sysfs_create_link(struct kobject *kobj,
 
 static inline void sysfs_remove_link(struct kobject *kobj, const char *name)
 {
-	;
 }
 
 static inline int sysfs_create_group(struct kobject *kobj,
@@ -195,7 +191,6 @@ static inline int sysfs_create_group(struct kobject *kobj,
 static inline void sysfs_remove_group(struct kobject *kobj,
 				      const struct attribute_group *grp)
 {
-	;
 }
 
 static inline int sysfs_add_file_to_group(struct kobject *kobj,
-- 
cgit v1.2.3


From 58aca23226a19983571bd3b65167521fc64f5869 Mon Sep 17 00:00:00 2001
From: "Rafael J. Wysocki" <rjw@sisk.pl>
Date: Wed, 12 Mar 2008 00:57:22 +0100
Subject: PM: Handle device registrations during suspend/resume

Modify the PM core to protect its data structures, specifically the
dpm_active list, from being corrupted if a child of the currently
suspending device is registered concurrently with its ->suspend()
callback.  In that case, since the new device (the child) is added
to dpm_active after its parent, the PM core will attempt to
suspend it after the parent, which is wrong.

Introduce a new member of struct dev_pm_info, called 'sleeping',
and use it to check if the parent of the device being added to
dpm_active has been suspended, in which case the device registration
fails.  Also, use 'sleeping' for checking if the ordering of devices
on dpm_active is correct.

Introduce variable 'all_sleeping' that will be set to 'true' once all
devices have been suspended and make new device registrations fail
until 'all_sleeping' is reset to 'false', in order to avoid having
unsuspended devices around while the system is going into a sleep state.

Remove pm_sleep_rwsem which is not necessary any more.

Special thanks to Alan Stern for discussions and suggestions that
lead to the creation of this patch.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 Documentation/power/devices.txt |  5 ++++
 drivers/base/core.c             |  6 ++++-
 drivers/base/power/main.c       | 57 +++++++++++++++++++----------------------
 drivers/base/power/power.h      | 23 +++--------------
 include/linux/pm.h              |  1 +
 5 files changed, 40 insertions(+), 52 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/power/devices.txt b/Documentation/power/devices.txt
index 461e4f1dbec4..421e7d00ffd0 100644
--- a/Documentation/power/devices.txt
+++ b/Documentation/power/devices.txt
@@ -196,6 +196,11 @@ its parent; and can't be removed or suspended after that parent.
 
 The policy is that the device tree should match hardware bus topology.
 (Or at least the control bus, for devices which use multiple busses.)
+In particular, this means that a device registration may fail if the parent of
+the device is suspending (ie. has been chosen by the PM core as the next
+device to suspend) or has already suspended, as well as after all of the other
+devices have been suspended.  Device drivers must be prepared to cope with such
+situations.
 
 
 Suspending Devices
diff --git a/drivers/base/core.c b/drivers/base/core.c
index 24198ad01976..79848e6c5db5 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -820,7 +820,11 @@ int device_add(struct device *dev)
 	error = dpm_sysfs_add(dev);
 	if (error)
 		goto PMError;
-	device_pm_add(dev);
+	error = device_pm_add(dev);
+	if (error) {
+		dpm_sysfs_remove(dev);
+		goto PMError;
+	}
 	error = bus_add_device(dev);
 	if (error)
 		goto BusError;
diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 26de2c0fda80..0e3991a437c6 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -54,7 +54,8 @@ static LIST_HEAD(dpm_destroy);
 
 static DEFINE_MUTEX(dpm_list_mtx);
 
-static DECLARE_RWSEM(pm_sleep_rwsem);
+/* 'true' if all devices have been suspended, protected by dpm_list_mtx */
+static bool all_sleeping;
 
 int (*platform_enable_wakeup)(struct device *dev, int is_on);
 
@@ -62,14 +63,28 @@ int (*platform_enable_wakeup)(struct device *dev, int is_on);
  *	device_pm_add - add a device to the list of active devices
  *	@dev:	Device to be added to the list
  */
-void device_pm_add(struct device *dev)
+int device_pm_add(struct device *dev)
 {
+	int error = 0;
+
 	pr_debug("PM: Adding info for %s:%s\n",
 		 dev->bus ? dev->bus->name : "No Bus",
 		 kobject_name(&dev->kobj));
 	mutex_lock(&dpm_list_mtx);
-	list_add_tail(&dev->power.entry, &dpm_active);
+	if ((dev->parent && dev->parent->power.sleeping) || all_sleeping) {
+		if (dev->parent->power.sleeping)
+			dev_warn(dev,
+				"parent %s is sleeping, will not add\n",
+				dev->parent->bus_id);
+		else
+			dev_warn(dev, "devices are sleeping, will not add\n");
+		WARN_ON(true);
+		error = -EBUSY;
+	} else {
+		list_add_tail(&dev->power.entry, &dpm_active);
+	}
 	mutex_unlock(&dpm_list_mtx);
+	return error;
 }
 
 /**
@@ -107,32 +122,6 @@ void device_pm_schedule_removal(struct device *dev)
 }
 EXPORT_SYMBOL_GPL(device_pm_schedule_removal);
 
-/**
- *	pm_sleep_lock - mutual exclusion for registration and suspend
- *
- *	Returns 0 if no suspend is underway and device registration
- *	may proceed, otherwise -EBUSY.
- */
-int pm_sleep_lock(void)
-{
-	if (down_read_trylock(&pm_sleep_rwsem))
-		return 0;
-
-	return -EBUSY;
-}
-
-/**
- *	pm_sleep_unlock - mutual exclusion for registration and suspend
- *
- *	This routine undoes the effect of device_pm_add_lock
- *	when a device's registration is complete.
- */
-void pm_sleep_unlock(void)
-{
-	up_read(&pm_sleep_rwsem);
-}
-
-
 /*------------------------- Resume routines -------------------------*/
 
 /**
@@ -242,11 +231,13 @@ static int resume_device(struct device *dev)
 static void dpm_resume(void)
 {
 	mutex_lock(&dpm_list_mtx);
+	all_sleeping = false;
 	while(!list_empty(&dpm_off)) {
 		struct list_head *entry = dpm_off.next;
 		struct device *dev = to_device(entry);
 
 		list_move_tail(entry, &dpm_active);
+		dev->power.sleeping = false;
 		mutex_unlock(&dpm_list_mtx);
 		resume_device(dev);
 		mutex_lock(&dpm_list_mtx);
@@ -285,7 +276,6 @@ void device_resume(void)
 	might_sleep();
 	dpm_resume();
 	unregister_dropped_devices();
-	up_write(&pm_sleep_rwsem);
 }
 EXPORT_SYMBOL_GPL(device_resume);
 
@@ -421,6 +411,9 @@ static int dpm_suspend(pm_message_t state)
 		struct list_head *entry = dpm_active.prev;
 		struct device *dev = to_device(entry);
 
+		WARN_ON(dev->parent && dev->parent->power.sleeping);
+
+		dev->power.sleeping = true;
 		mutex_unlock(&dpm_list_mtx);
 		error = suspend_device(dev, state);
 		mutex_lock(&dpm_list_mtx);
@@ -432,11 +425,14 @@ static int dpm_suspend(pm_message_t state)
 					(error == -EAGAIN ?
 					" (please convert to suspend_late)" :
 					""));
+			dev->power.sleeping = false;
 			break;
 		}
 		if (!list_empty(&dev->power.entry))
 			list_move(&dev->power.entry, &dpm_off);
 	}
+	if (!error)
+		all_sleeping = true;
 	mutex_unlock(&dpm_list_mtx);
 
 	return error;
@@ -454,7 +450,6 @@ int device_suspend(pm_message_t state)
 	int error;
 
 	might_sleep();
-	down_write(&pm_sleep_rwsem);
 	error = dpm_suspend(state);
 	if (error)
 		device_resume();
diff --git a/drivers/base/power/power.h b/drivers/base/power/power.h
index e32d3bdb92c1..a6894f2a4b99 100644
--- a/drivers/base/power/power.h
+++ b/drivers/base/power/power.h
@@ -11,30 +11,13 @@ static inline struct device *to_device(struct list_head *entry)
 	return container_of(entry, struct device, power.entry);
 }
 
-extern void device_pm_add(struct device *);
+extern int device_pm_add(struct device *);
 extern void device_pm_remove(struct device *);
-extern int pm_sleep_lock(void);
-extern void pm_sleep_unlock(void);
 
 #else /* CONFIG_PM_SLEEP */
 
-
-static inline void device_pm_add(struct device *dev)
-{
-}
-
-static inline void device_pm_remove(struct device *dev)
-{
-}
-
-static inline int pm_sleep_lock(void)
-{
-	return 0;
-}
-
-static inline void pm_sleep_unlock(void)
-{
-}
+static inline int device_pm_add(struct device *dev) { return 0; }
+static inline void device_pm_remove(struct device *dev) {}
 
 #endif
 
diff --git a/include/linux/pm.h b/include/linux/pm.h
index 015b735811b4..e6b9f29e27d7 100644
--- a/include/linux/pm.h
+++ b/include/linux/pm.h
@@ -183,6 +183,7 @@ typedef struct pm_message {
 struct dev_pm_info {
 	pm_message_t		power_state;
 	unsigned		can_wakeup:1;
+	bool			sleeping:1;	/* Owned by the PM core */
 #ifdef	CONFIG_PM_SLEEP
 	unsigned		should_wakeup:1;
 	struct list_head	entry;
-- 
cgit v1.2.3


From d288e47c471e1090e80c62ad95882fafbf3f499d Mon Sep 17 00:00:00 2001
From: Alan Stern <stern@rowland.harvard.edu>
Date: Wed, 19 Mar 2008 22:37:42 +0100
Subject: PM: Make wakeup flags available whenever CONFIG_PM is set

The various wakeup flags and their accessor macros in struct
dev_pm_info should be available whenever CONFIG_PM is enabled, not
just when CONFIG_PM_SLEEP is on.  Otherwise remote wakeup won't always
be configurable for runtime power management.  This patch (as1056b)
fixes the oversight.

David Brownell adds:
	More accurately, fixes the "regression" ... as noted sometime
	last summer, after 296699de6bdc717189a331ab6bbe90e05c94db06
	introduced CONFIG_SUSPEND.  But that didn't make the regression
	list for that kernel, ergo the delay in fixing it.

[rjw: rebased]

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/base/power/main.c  |  2 --
 drivers/base/power/sysfs.c |  2 ++
 include/linux/pm.h         | 36 +++++++++++++++++++++---------------
 3 files changed, 23 insertions(+), 17 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 93a146940b91..5630af302b2f 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -57,8 +57,6 @@ static DEFINE_MUTEX(dpm_list_mtx);
 /* 'true' if all devices have been suspended, protected by dpm_list_mtx */
 static bool all_sleeping;
 
-int (*platform_enable_wakeup)(struct device *dev, int is_on);
-
 /**
  *	device_pm_add - add a device to the list of active devices
  *	@dev:	Device to be added to the list
diff --git a/drivers/base/power/sysfs.c b/drivers/base/power/sysfs.c
index f2ed179cd695..d11f74b038db 100644
--- a/drivers/base/power/sysfs.c
+++ b/drivers/base/power/sysfs.c
@@ -6,6 +6,8 @@
 #include <linux/string.h>
 #include "power.h"
 
+int (*platform_enable_wakeup)(struct device *dev, int is_on);
+
 
 /*
  *	wakeup - Report/change current wakeup option for device
diff --git a/include/linux/pm.h b/include/linux/pm.h
index e6b9f29e27d7..3342627e2bd6 100644
--- a/include/linux/pm.h
+++ b/include/linux/pm.h
@@ -183,9 +183,9 @@ typedef struct pm_message {
 struct dev_pm_info {
 	pm_message_t		power_state;
 	unsigned		can_wakeup:1;
+	unsigned		should_wakeup:1;
 	bool			sleeping:1;	/* Owned by the PM core */
 #ifdef	CONFIG_PM_SLEEP
-	unsigned		should_wakeup:1;
 	struct list_head	entry;
 #endif
 };
@@ -198,11 +198,6 @@ extern void device_resume(void);
 extern int device_suspend(pm_message_t state);
 extern int device_prepare_suspend(pm_message_t state);
 
-#define device_set_wakeup_enable(dev,val) \
-	((dev)->power.should_wakeup = !!(val))
-#define device_may_wakeup(dev) \
-	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
-
 extern void __suspend_report_result(const char *function, void *fn, int ret);
 
 #define suspend_report_result(fn, ret)					\
@@ -210,6 +205,24 @@ extern void __suspend_report_result(const char *function, void *fn, int ret);
 		__suspend_report_result(__FUNCTION__, fn, ret);		\
 	} while (0)
 
+#else /* !CONFIG_PM_SLEEP */
+
+static inline int device_suspend(pm_message_t state)
+{
+	return 0;
+}
+
+#define suspend_report_result(fn, ret) do { } while (0)
+
+#endif /* !CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+
+#define device_set_wakeup_enable(dev,val) \
+	((dev)->power.should_wakeup = !!(val))
+#define device_may_wakeup(dev) \
+	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
+
 /*
  * Platform hook to activate device wakeup capability, if that's not already
  * handled by enable_irq_wake() etc.
@@ -224,24 +237,17 @@ static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
 	return 0;
 }
 
-#else /* !CONFIG_PM_SLEEP */
-
-static inline int device_suspend(pm_message_t state)
-{
-	return 0;
-}
+#else /* !CONFIG_PM */
 
 #define device_set_wakeup_enable(dev,val)	do{}while(0)
 #define device_may_wakeup(dev)			(0)
 
-#define suspend_report_result(fn, ret) do { } while (0)
-
 static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
 {
 	return 0;
 }
 
-#endif /* !CONFIG_PM_SLEEP */
+#endif /* !CONFIG_PM */
 
 /* changes to device_may_wakeup take effect on the next pm state change.
  * by default, devices should wakeup if they can.
-- 
cgit v1.2.3


From 9a3df1f7de0ecaf77a1dde86f2a4dc020f37f87e Mon Sep 17 00:00:00 2001
From: Alan Stern <stern@rowland.harvard.edu>
Date: Wed, 19 Mar 2008 22:39:13 +0100
Subject: PM: Convert wakeup flag accessors to inline functions

This patch (as1058) improves the wakeup macros in include/linux/pm.h.
All but the trivial ones are converted to inline routines, which
requires moving them to a separate header file since they depend on
the definition of struct device.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/device.h    |  3 ++
 include/linux/pm.h        | 46 +-----------------------
 include/linux/pm_wakeup.h | 90 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 94 insertions(+), 45 deletions(-)
 create mode 100644 include/linux/pm_wakeup.h

(limited to 'include/linux')

diff --git a/include/linux/device.h b/include/linux/device.h
index d57661129cb2..d7a1ae063b65 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -475,6 +475,9 @@ struct device {
 	void	(*release)(struct device *dev);
 };
 
+/* Get the wakeup routines, which depend on struct device */
+#include <linux/pm_wakeup.h>
+
 #ifdef CONFIG_NUMA
 static inline int dev_to_node(struct device *dev)
 {
diff --git a/include/linux/pm.h b/include/linux/pm.h
index 3342627e2bd6..1de72cbbe0d1 100644
--- a/include/linux/pm.h
+++ b/include/linux/pm.h
@@ -212,54 +212,10 @@ static inline int device_suspend(pm_message_t state)
 	return 0;
 }
 
-#define suspend_report_result(fn, ret) do { } while (0)
+#define suspend_report_result(fn, ret)		do {} while (0)
 
 #endif /* !CONFIG_PM_SLEEP */
 
-#ifdef CONFIG_PM
-
-#define device_set_wakeup_enable(dev,val) \
-	((dev)->power.should_wakeup = !!(val))
-#define device_may_wakeup(dev) \
-	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
-
-/*
- * Platform hook to activate device wakeup capability, if that's not already
- * handled by enable_irq_wake() etc.
- * Returns zero on success, else negative errno
- */
-extern int (*platform_enable_wakeup)(struct device *dev, int is_on);
-
-static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
-{
-	if (platform_enable_wakeup)
-		return (*platform_enable_wakeup)(dev, is_on);
-	return 0;
-}
-
-#else /* !CONFIG_PM */
-
-#define device_set_wakeup_enable(dev,val)	do{}while(0)
-#define device_may_wakeup(dev)			(0)
-
-static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
-{
-	return 0;
-}
-
-#endif /* !CONFIG_PM */
-
-/* changes to device_may_wakeup take effect on the next pm state change.
- * by default, devices should wakeup if they can.
- */
-#define device_can_wakeup(dev) \
-	((dev)->power.can_wakeup)
-#define device_init_wakeup(dev,val) \
-	do { \
-		device_can_wakeup(dev) = !!(val); \
-		device_set_wakeup_enable(dev,val); \
-	} while(0)
-
 /*
  * Global Power Management flags
  * Used to keep APM and ACPI from both being active
diff --git a/include/linux/pm_wakeup.h b/include/linux/pm_wakeup.h
new file mode 100644
index 000000000000..f0d0b2cb8d20
--- /dev/null
+++ b/include/linux/pm_wakeup.h
@@ -0,0 +1,90 @@
+/*
+ *  pm_wakeup.h - Power management wakeup interface
+ *
+ *  Copyright (C) 2008 Alan Stern
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef _LINUX_PM_WAKEUP_H
+#define _LINUX_PM_WAKEUP_H
+
+#ifndef _DEVICE_H_
+# error "please don't include this file directly"
+#endif
+
+#ifdef CONFIG_PM
+
+/* changes to device_may_wakeup take effect on the next pm state change.
+ * by default, devices should wakeup if they can.
+ */
+static inline void device_init_wakeup(struct device *dev, int val)
+{
+	dev->power.can_wakeup = dev->power.should_wakeup = !!val;
+}
+
+static inline int device_can_wakeup(struct device *dev)
+{
+	return dev->power.can_wakeup;
+}
+
+static inline void device_set_wakeup_enable(struct device *dev, int val)
+{
+	dev->power.should_wakeup = !!val;
+}
+
+static inline int device_may_wakeup(struct device *dev)
+{
+	return dev->power.can_wakeup & dev->power.should_wakeup;
+}
+
+/*
+ * Platform hook to activate device wakeup capability, if that's not already
+ * handled by enable_irq_wake() etc.
+ * Returns zero on success, else negative errno
+ */
+extern int (*platform_enable_wakeup)(struct device *dev, int is_on);
+
+static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
+{
+	if (platform_enable_wakeup)
+		return (*platform_enable_wakeup)(dev, is_on);
+	return 0;
+}
+
+#else /* !CONFIG_PM */
+
+/* For some reason the next two routines work even without CONFIG_PM */
+static inline void device_init_wakeup(struct device *dev, int val)
+{
+	dev->power.can_wakeup = !!val;
+}
+
+static inline int device_can_wakeup(struct device *dev)
+{
+	return dev->power.can_wakeup;
+}
+
+#define device_set_wakeup_enable(dev, val)	do {} while (0)
+#define device_may_wakeup(dev)			0
+
+static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
+{
+	return 0;
+}
+
+#endif /* !CONFIG_PM */
+
+#endif /* _LINUX_PM_WAKEUP_H */
-- 
cgit v1.2.3


From 3f62e5700b2a679ae987b32a68126dd6dcf2488f Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Thu, 13 Mar 2008 17:07:03 -0400
Subject: Driver core: make device_is_registered() work for class devices

device_is_registered() can use the kobject value for this, so it will
now work with devices that are associated with only a class, not a bus
and a driver.

Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/base/bus.c     | 9 ++-------
 include/linux/device.h | 3 +--
 2 files changed, 3 insertions(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/base/bus.c b/drivers/base/bus.c
index 2d207ad30336..450942acca1a 100644
--- a/drivers/base/bus.c
+++ b/drivers/base/bus.c
@@ -505,14 +505,11 @@ void bus_attach_device(struct device *dev)
 	int ret = 0;
 
 	if (bus) {
-		dev->is_registered = 1;
 		if (bus->p->drivers_autoprobe)
 			ret = device_attach(dev);
 		WARN_ON(ret < 0);
 		if (ret >= 0)
 			klist_add_tail(&dev->knode_bus, &bus->p->klist_devices);
-		else
-			dev->is_registered = 0;
 	}
 }
 
@@ -533,10 +530,8 @@ void bus_remove_device(struct device *dev)
 		sysfs_remove_link(&dev->bus->p->devices_kset->kobj,
 				  dev->bus_id);
 		device_remove_attrs(dev->bus, dev);
-		if (dev->is_registered) {
-			dev->is_registered = 0;
-			klist_del(&dev->knode_bus);
-		}
+		klist_del(&dev->knode_bus);
+
 		pr_debug("bus: '%s': remove device %s\n",
 			 dev->bus->name, dev->bus_id);
 		device_release_driver(dev);
diff --git a/include/linux/device.h b/include/linux/device.h
index d7a1ae063b65..441461f5ee20 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -429,7 +429,6 @@ struct device {
 	struct kobject kobj;
 	char	bus_id[BUS_ID_SIZE];	/* position on parent bus */
 	struct device_type	*type;
-	unsigned		is_registered:1;
 	unsigned		uevent_suppress:1;
 
 	struct semaphore	sem;	/* semaphore to synchronize calls to
@@ -509,7 +508,7 @@ static inline void dev_set_drvdata(struct device *dev, void *data)
 
 static inline int device_is_registered(struct device *dev)
 {
-	return dev->is_registered;
+	return dev->kobj.state_in_sysfs;
 }
 
 void driver_init(void);
-- 
cgit v1.2.3


From 138fe4e069798d9aa948a5402ff15e58f483ee4e Mon Sep 17 00:00:00 2001
From: Konrad Rzeszutek <ketuzsezr@darnok.org>
Date: Wed, 9 Apr 2008 19:50:41 -0700
Subject: Firmware: add iSCSI iBFT Support

Add /sysfs/firmware/ibft/[initiator|targetX|ethernetX] directories along with
text properties which export the the iSCSI Boot Firmware Table (iBFT)
structure.

What is iSCSI Boot Firmware Table?  It is a mechanism for the iSCSI tools to
extract from the machine NICs the iSCSI connection information so that they
can automagically mount the iSCSI share/target.  Currently the iSCSI
information is hard-coded in the initrd.  The /sysfs entries are read-only
one-name-and-value fields.

The usual set of data exposed is:

# for a in `find /sys/firmware/ibft/ -type f -print`; do  echo -n "$a: ";  cat $a; done
/sys/firmware/ibft/target0/target-name: iqn.2007.com.intel-sbx44:storage-10gb
/sys/firmware/ibft/target0/nic-assoc: 0
/sys/firmware/ibft/target0/chap-type: 0
/sys/firmware/ibft/target0/lun: 00000000
/sys/firmware/ibft/target0/port: 3260
/sys/firmware/ibft/target0/ip-addr: 192.168.79.116
/sys/firmware/ibft/target0/flags: 3
/sys/firmware/ibft/target0/index: 0
/sys/firmware/ibft/ethernet0/mac: 00:11:25:9d:8b:01
/sys/firmware/ibft/ethernet0/vlan: 0
/sys/firmware/ibft/ethernet0/gateway: 192.168.79.254
/sys/firmware/ibft/ethernet0/origin: 0
/sys/firmware/ibft/ethernet0/subnet-mask: 255.255.252.0
/sys/firmware/ibft/ethernet0/ip-addr: 192.168.77.41
/sys/firmware/ibft/ethernet0/flags: 7
/sys/firmware/ibft/ethernet0/index: 0
/sys/firmware/ibft/initiator/initiator-name: iqn.2007-07.com:konrad.initiator
/sys/firmware/ibft/initiator/flags: 3
/sys/firmware/ibft/initiator/index: 0

For full details of the IBFT structure please take a look at:
ftp://ftp.software.ibm.com/systems/support/system_x_pdf/ibm_iscsi_boot_firmware_table_v1.02.pdf

[akpm@linux-foundation.org: fix build]
Signed-off-by: Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Peter Jones <pjones@redhat.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 Documentation/ABI/testing/sysfs-ibft |  23 +
 arch/x86/kernel/setup_32.c           |   3 +
 arch/x86/kernel/setup_64.c           |   4 +
 drivers/firmware/Kconfig             |  20 +
 drivers/firmware/Makefile            |   2 +
 drivers/firmware/iscsi_ibft.c        | 982 +++++++++++++++++++++++++++++++++++
 drivers/firmware/iscsi_ibft_find.c   |  84 +++
 include/linux/iscsi_ibft.h           |  50 ++
 8 files changed, 1168 insertions(+)
 create mode 100644 Documentation/ABI/testing/sysfs-ibft
 create mode 100644 drivers/firmware/iscsi_ibft.c
 create mode 100644 drivers/firmware/iscsi_ibft_find.c
 create mode 100644 include/linux/iscsi_ibft.h

(limited to 'include/linux')

diff --git a/Documentation/ABI/testing/sysfs-ibft b/Documentation/ABI/testing/sysfs-ibft
new file mode 100644
index 000000000000..c2b7d1154bec
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-ibft
@@ -0,0 +1,23 @@
+What:		/sys/firmware/ibft/initiator
+Date:		November 2007
+Contact:	Konrad Rzeszutek <ketuzsezr@darnok.org>
+Description:	The /sys/firmware/ibft/initiator directory will contain
+		files that expose the iSCSI Boot Firmware Table initiator data.
+		Usually this contains the Initiator name.
+
+What:		/sys/firmware/ibft/targetX
+Date:		November 2007
+Contact:	Konrad Rzeszutek <ketuzsezr@darnok.org>
+Description:	The /sys/firmware/ibft/targetX directory will contain
+		files that expose the iSCSI Boot Firmware Table target data.
+		Usually this contains the target's IP address, boot LUN,
+		target name, and what NIC it is associated with. It can also
+		contain the CHAP name (and password), the reverse CHAP
+		name (and password)
+
+What:		/sys/firmware/ibft/ethernetX
+Date:		November 2007
+Contact:	Konrad Rzeszutek <ketuzsezr@darnok.org>
+Description:	The /sys/firmware/ibft/ethernetX directory will contain
+		files that expose the iSCSI Boot Firmware Table NIC data.
+		This can this can the IP address, MAC, and gateway of the NIC.
diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c
index 5b0bffb7fcc9..4ef91749959e 100644
--- a/arch/x86/kernel/setup_32.c
+++ b/arch/x86/kernel/setup_32.c
@@ -39,6 +39,7 @@
 #include <linux/efi.h>
 #include <linux/init.h>
 #include <linux/edd.h>
+#include <linux/iscsi_ibft.h>
 #include <linux/nodemask.h>
 #include <linux/kexec.h>
 #include <linux/crash_dump.h>
@@ -689,6 +690,8 @@ void __init setup_bootmem_allocator(void)
 #endif
 	numa_kva_reserve();
 	reserve_crashkernel();
+
+	reserve_ibft_region();
 }
 
 /*
diff --git a/arch/x86/kernel/setup_64.c b/arch/x86/kernel/setup_64.c
index 674ef3510cdf..216c93bd9993 100644
--- a/arch/x86/kernel/setup_64.c
+++ b/arch/x86/kernel/setup_64.c
@@ -33,6 +33,7 @@
 #include <linux/acpi.h>
 #include <linux/kallsyms.h>
 #include <linux/edd.h>
+#include <linux/iscsi_ibft.h>
 #include <linux/mmzone.h>
 #include <linux/kexec.h>
 #include <linux/cpufreq.h>
@@ -434,6 +435,9 @@ void __init setup_arch(char **cmdline_p)
 	}
 #endif
 	reserve_crashkernel();
+
+	reserve_ibft_region();
+
 	paging_init();
 	map_vsyscall();
 
diff --git a/drivers/firmware/Kconfig b/drivers/firmware/Kconfig
index 05f02a326f1c..40ffd767647d 100644
--- a/drivers/firmware/Kconfig
+++ b/drivers/firmware/Kconfig
@@ -93,4 +93,24 @@ config DMIID
 	  information from userspace through /sys/class/dmi/id/ or if you want
 	  DMI-based module auto-loading.
 
+config ISCSI_IBFT_FIND
+	bool "iSCSI Boot Firmware Table Attributes"
+	depends on X86
+	default n
+	help
+	  This option enables the kernel to find the region of memory
+	  in which the ISCSI Boot Firmware Table (iBFT) resides. This
+	  is necessary for iSCSI Boot Firmware Table Attributes module to work
+	  properly.
+
+config ISCSI_IBFT
+	tristate "iSCSI Boot Firmware Table Attributes module"
+	depends on ISCSI_IBFT_FIND
+	default	n
+	help
+	  This option enables support for detection and exposing of iSCSI
+	  Boot Firmware Table (iBFT) via sysfs to userspace. If you wish to
+	  detect iSCSI boot parameters dynamically during system boot, say Y.
+	  Otherwise, say N.
+
 endmenu
diff --git a/drivers/firmware/Makefile b/drivers/firmware/Makefile
index 8d4ebc805a50..4c9147154df8 100644
--- a/drivers/firmware/Makefile
+++ b/drivers/firmware/Makefile
@@ -8,3 +8,5 @@ obj-$(CONFIG_EFI_PCDP)		+= pcdp.o
 obj-$(CONFIG_DELL_RBU)          += dell_rbu.o
 obj-$(CONFIG_DCDBAS)		+= dcdbas.o
 obj-$(CONFIG_DMIID)		+= dmi-id.o
+obj-$(CONFIG_ISCSI_IBFT_FIND)	+= iscsi_ibft_find.o
+obj-$(CONFIG_ISCSI_IBFT)	+= iscsi_ibft.o
diff --git a/drivers/firmware/iscsi_ibft.c b/drivers/firmware/iscsi_ibft.c
new file mode 100644
index 000000000000..8024e3bfd877
--- /dev/null
+++ b/drivers/firmware/iscsi_ibft.c
@@ -0,0 +1,982 @@
+/*
+ *  Copyright 2007 Red Hat, Inc.
+ *  by Peter Jones <pjones@redhat.com>
+ *  Copyright 2008 IBM, Inc.
+ *  by Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *  Copyright 2008
+ *  by Konrad Rzeszutek <ketuzsezr@darnok.org>
+ *
+ * This code exposes the iSCSI Boot Format Table to userland via sysfs.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License v2.0 as published by
+ * the Free Software Foundation
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * Changelog:
+ *
+ *  14 Mar 2008 - Konrad Rzeszutek <ketuzsezr@darnok.org>
+ *    Updated comments and copyrights. (v0.4.9)
+ *
+ *  11 Feb 2008 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *    Converted to using ibft_addr. (v0.4.8)
+ *
+ *   8 Feb 2008 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *    Combined two functions in one: reserve_ibft_region. (v0.4.7)
+ *
+ *  30 Jan 2008 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Added logic to handle IPv6 addresses. (v0.4.6)
+ *
+ *  25 Jan 2008 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Added logic to handle badly not-to-spec iBFT. (v0.4.5)
+ *
+ *   4 Jan 2008 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Added __init to function declarations. (v0.4.4)
+ *
+ *  21 Dec 2007 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Updated kobject registration, combined unregister functions in one
+ *   and code and style cleanup. (v0.4.3)
+ *
+ *   5 Dec 2007 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Added end-markers to enums and re-organized kobject registration. (v0.4.2)
+ *
+ *   4 Dec 2007 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Created 'device' sysfs link to the NIC and style cleanup. (v0.4.1)
+ *
+ *  28 Nov 2007 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Added sysfs-ibft documentation, moved 'find_ibft' function to
+ *   in its own file and added text attributes for every struct field.  (v0.4)
+ *
+ *  21 Nov 2007 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Added text attributes emulating OpenFirmware /proc/device-tree naming.
+ *   Removed binary /sysfs interface (v0.3)
+ *
+ *  29 Aug 2007 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   Added functionality in setup.c to reserve iBFT region. (v0.2)
+ *
+ *  27 Aug 2007 - Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *   First version exposing iBFT data via a binary /sysfs. (v0.1)
+ *
+ */
+
+
+#include <linux/blkdev.h>
+#include <linux/capability.h>
+#include <linux/ctype.h>
+#include <linux/device.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/iscsi_ibft.h>
+#include <linux/limits.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+#include <linux/stat.h>
+#include <linux/string.h>
+#include <linux/types.h>
+
+#define IBFT_ISCSI_VERSION "0.4.9"
+#define IBFT_ISCSI_DATE "2008-Mar-14"
+
+MODULE_AUTHOR("Peter Jones <pjones@redhat.com> and \
+Konrad Rzeszutek <ketuzsezr@darnok.org>");
+MODULE_DESCRIPTION("sysfs interface to BIOS iBFT information");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(IBFT_ISCSI_VERSION);
+
+struct ibft_hdr {
+	u8 id;
+	u8 version;
+	u16 length;
+	u8 index;
+	u8 flags;
+} __attribute__((__packed__));
+
+struct ibft_control {
+	struct ibft_hdr hdr;
+	u16 extensions;
+	u16 initiator_off;
+	u16 nic0_off;
+	u16 tgt0_off;
+	u16 nic1_off;
+	u16 tgt1_off;
+} __attribute__((__packed__));
+
+struct ibft_initiator {
+	struct ibft_hdr hdr;
+	char isns_server[16];
+	char slp_server[16];
+	char pri_radius_server[16];
+	char sec_radius_server[16];
+	u16 initiator_name_len;
+	u16 initiator_name_off;
+} __attribute__((__packed__));
+
+struct ibft_nic {
+	struct ibft_hdr hdr;
+	char ip_addr[16];
+	u8 subnet_mask_prefix;
+	u8 origin;
+	char gateway[16];
+	char primary_dns[16];
+	char secondary_dns[16];
+	char dhcp[16];
+	u16 vlan;
+	char mac[6];
+	u16 pci_bdf;
+	u16 hostname_len;
+	u16 hostname_off;
+} __attribute__((__packed__));
+
+struct ibft_tgt {
+	struct ibft_hdr hdr;
+	char ip_addr[16];
+	u16 port;
+	char lun[8];
+	u8 chap_type;
+	u8 nic_assoc;
+	u16 tgt_name_len;
+	u16 tgt_name_off;
+	u16 chap_name_len;
+	u16 chap_name_off;
+	u16 chap_secret_len;
+	u16 chap_secret_off;
+	u16 rev_chap_name_len;
+	u16 rev_chap_name_off;
+	u16 rev_chap_secret_len;
+	u16 rev_chap_secret_off;
+} __attribute__((__packed__));
+
+/*
+ * The kobject different types and its names.
+ *
+*/
+enum ibft_id {
+	id_reserved = 0, /* We don't support. */
+	id_control = 1, /* Should show up only once and is not exported. */
+	id_initiator = 2,
+	id_nic = 3,
+	id_target = 4,
+	id_extensions = 5, /* We don't support. */
+	id_end_marker,
+};
+
+/*
+ * We do not support the other types, hence the usage of NULL.
+ * This maps to the enum ibft_id.
+ */
+static const char *ibft_id_names[] =
+	{NULL, NULL, "initiator", "ethernet%d", "target%d", NULL, NULL};
+
+/*
+ * The text attributes names for each of the kobjects.
+*/
+enum ibft_eth_properties_enum {
+	ibft_eth_index,
+	ibft_eth_flags,
+	ibft_eth_ip_addr,
+	ibft_eth_subnet_mask,
+	ibft_eth_origin,
+	ibft_eth_gateway,
+	ibft_eth_primary_dns,
+	ibft_eth_secondary_dns,
+	ibft_eth_dhcp,
+	ibft_eth_vlan,
+	ibft_eth_mac,
+	/* ibft_eth_pci_bdf - this is replaced by link to the device itself. */
+	ibft_eth_hostname,
+	ibft_eth_end_marker,
+};
+
+static const char *ibft_eth_properties[] =
+	{"index", "flags", "ip-addr", "subnet-mask", "origin", "gateway",
+	"primary-dns", "secondary-dns", "dhcp", "vlan", "mac", "hostname",
+	NULL};
+
+enum ibft_tgt_properties_enum {
+	ibft_tgt_index,
+	ibft_tgt_flags,
+	ibft_tgt_ip_addr,
+	ibft_tgt_port,
+	ibft_tgt_lun,
+	ibft_tgt_chap_type,
+	ibft_tgt_nic_assoc,
+	ibft_tgt_name,
+	ibft_tgt_chap_name,
+	ibft_tgt_chap_secret,
+	ibft_tgt_rev_chap_name,
+	ibft_tgt_rev_chap_secret,
+	ibft_tgt_end_marker,
+};
+
+static const char *ibft_tgt_properties[] =
+	{"index", "flags", "ip-addr", "port", "lun", "chap-type", "nic-assoc",
+	"target-name", "chap-name", "chap-secret", "rev-chap-name",
+	"rev-chap-name-secret", NULL};
+
+enum ibft_initiator_properties_enum {
+	ibft_init_index,
+	ibft_init_flags,
+	ibft_init_isns_server,
+	ibft_init_slp_server,
+	ibft_init_pri_radius_server,
+	ibft_init_sec_radius_server,
+	ibft_init_initiator_name,
+	ibft_init_end_marker,
+};
+
+static const char *ibft_initiator_properties[] =
+	{"index", "flags", "isns-server", "slp-server", "pri-radius-server",
+	"sec-radius-server", "initiator-name", NULL};
+
+/*
+ * The kobject and attribute structures.
+ */
+
+struct ibft_kobject {
+	struct ibft_table_header *header;
+	union {
+		struct ibft_initiator *initiator;
+		struct ibft_nic *nic;
+		struct ibft_tgt *tgt;
+		struct ibft_hdr *hdr;
+	};
+	struct kobject kobj;
+	struct list_head node;
+};
+
+struct ibft_attribute {
+	struct attribute attr;
+	ssize_t (*show) (struct  ibft_kobject *entry,
+			 struct ibft_attribute *attr, char *buf);
+	union {
+		struct ibft_initiator *initiator;
+		struct ibft_nic *nic;
+		struct ibft_tgt *tgt;
+		struct ibft_hdr *hdr;
+	};
+	struct kobject *kobj;
+	int type; /* The enum of the type. This can be any value of:
+		ibft_eth_properties_enum, ibft_tgt_properties_enum,
+		or ibft_initiator_properties_enum. */
+	struct list_head node;
+};
+
+static LIST_HEAD(ibft_attr_list);
+static LIST_HEAD(ibft_kobject_list);
+
+static const char nulls[16];
+
+/*
+ * Helper functions to parse data properly.
+ */
+static ssize_t sprintf_ipaddr(char *buf, u8 *ip)
+{
+	char *str = buf;
+
+	if (ip[0] == 0 && ip[1] == 0 && ip[2] == 0 && ip[3] == 0 &&
+	    ip[4] == 0 && ip[5] == 0 && ip[6] == 0 && ip[7] == 0 &&
+	    ip[8] == 0 && ip[9] == 0 && ip[10] == 0xff && ip[11] == 0xff) {
+		/*
+		 * IPV4
+		 */
+		str += sprintf(buf, NIPQUAD_FMT, ip[12],
+			       ip[13], ip[14], ip[15]);
+	} else {
+		/*
+		 * IPv6
+		 */
+		str += sprintf(str, NIP6_FMT, ntohs(ip[0]), ntohs(ip[1]),
+			       ntohs(ip[2]), ntohs(ip[3]), ntohs(ip[4]),
+			       ntohs(ip[5]), ntohs(ip[6]), ntohs(ip[7]));
+	}
+	str += sprintf(str, "\n");
+	return str - buf;
+}
+
+static ssize_t sprintf_string(char *str, int len, char *buf)
+{
+	return sprintf(str, "%.*s\n", len, buf);
+}
+
+/*
+ * Helper function to verify the IBFT header.
+ */
+static int ibft_verify_hdr(char *t, struct ibft_hdr *hdr, int id, int length)
+{
+	if (hdr->id != id) {
+		printk(KERN_ERR "iBFT error: We expected the " \
+				"field header.id to have %d but " \
+				"found %d instead!\n", id, hdr->id);
+		return -ENODEV;
+	}
+	if (hdr->length != length) {
+		printk(KERN_ERR "iBFT error: We expected the " \
+				"field header.length to have %d but " \
+				"found %d instead!\n", length, hdr->length);
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
+static void ibft_release(struct kobject *kobj)
+{
+	struct ibft_kobject *ibft =
+		container_of(kobj, struct ibft_kobject, kobj);
+	kfree(ibft);
+}
+
+/*
+ *  Routines for parsing the iBFT data to be human readable.
+ */
+ssize_t ibft_attr_show_initiator(struct ibft_kobject *entry,
+				  struct ibft_attribute *attr,
+				  char *buf)
+{
+	struct ibft_initiator *initiator = entry->initiator;
+	void *ibft_loc = entry->header;
+	char *str = buf;
+
+	if (!initiator)
+		return 0;
+
+	switch (attr->type) {
+	case ibft_init_index:
+		str += sprintf(str, "%d\n", initiator->hdr.index);
+		break;
+	case ibft_init_flags:
+		str += sprintf(str, "%d\n", initiator->hdr.flags);
+		break;
+	case ibft_init_isns_server:
+		str += sprintf_ipaddr(str, initiator->isns_server);
+		break;
+	case ibft_init_slp_server:
+		str += sprintf_ipaddr(str, initiator->slp_server);
+		break;
+	case ibft_init_pri_radius_server:
+		str += sprintf_ipaddr(str, initiator->pri_radius_server);
+		break;
+	case ibft_init_sec_radius_server:
+		str += sprintf_ipaddr(str, initiator->sec_radius_server);
+		break;
+	case ibft_init_initiator_name:
+		str += sprintf_string(str, initiator->initiator_name_len,
+				      (char *)ibft_loc +
+				      initiator->initiator_name_off);
+		break;
+	default:
+		break;
+	}
+
+	return str - buf;
+}
+
+ssize_t ibft_attr_show_nic(struct ibft_kobject *entry,
+			    struct ibft_attribute *attr,
+			    char *buf)
+{
+	struct ibft_nic *nic = entry->nic;
+	void *ibft_loc = entry->header;
+	char *str = buf;
+	char *mac;
+	int val;
+
+	if (!nic)
+		return 0;
+
+	switch (attr->type) {
+	case ibft_eth_index:
+		str += sprintf(str, "%d\n", nic->hdr.index);
+		break;
+	case ibft_eth_flags:
+		str += sprintf(str, "%d\n", nic->hdr.flags);
+		break;
+	case ibft_eth_ip_addr:
+		str += sprintf_ipaddr(str, nic->ip_addr);
+		break;
+	case ibft_eth_subnet_mask:
+		val = ~((1 << (32-nic->subnet_mask_prefix))-1);
+		str += sprintf(str, NIPQUAD_FMT,
+			       (u8)(val >> 24), (u8)(val >> 16),
+			       (u8)(val >> 8), (u8)(val));
+		break;
+	case ibft_eth_origin:
+		str += sprintf(str, "%d\n", nic->origin);
+		break;
+	case ibft_eth_gateway:
+		str += sprintf_ipaddr(str, nic->gateway);
+		break;
+	case ibft_eth_primary_dns:
+		str += sprintf_ipaddr(str, nic->primary_dns);
+		break;
+	case ibft_eth_secondary_dns:
+		str += sprintf_ipaddr(str, nic->secondary_dns);
+		break;
+	case ibft_eth_dhcp:
+		str += sprintf_ipaddr(str, nic->dhcp);
+		break;
+	case ibft_eth_vlan:
+		str += sprintf(str, "%d\n", nic->vlan);
+		break;
+	case ibft_eth_mac:
+		mac = nic->mac;
+		str += sprintf(str, "%02x:%02x:%02x:%02x:%02x:%02x\n",
+			       (u8)mac[0], (u8)mac[1], (u8)mac[2],
+			       (u8)mac[3], (u8)mac[4], (u8)mac[5]);
+		break;
+	case ibft_eth_hostname:
+		str += sprintf_string(str, nic->hostname_len,
+				      (char *)ibft_loc + nic->hostname_off);
+		break;
+	default:
+		break;
+	}
+
+	return str - buf;
+};
+
+ssize_t ibft_attr_show_target(struct ibft_kobject *entry,
+			       struct ibft_attribute *attr,
+			       char *buf)
+{
+	struct ibft_tgt *tgt = entry->tgt;
+	void *ibft_loc = entry->header;
+	char *str = buf;
+	int i;
+
+	if (!tgt)
+		return 0;
+
+	switch (attr->type) {
+	case ibft_tgt_index:
+		str += sprintf(str, "%d\n", tgt->hdr.index);
+		break;
+	case ibft_tgt_flags:
+		str += sprintf(str, "%d\n", tgt->hdr.flags);
+		break;
+	case ibft_tgt_ip_addr:
+		str += sprintf_ipaddr(str, tgt->ip_addr);
+		break;
+	case ibft_tgt_port:
+		str += sprintf(str, "%d\n", tgt->port);
+		break;
+	case ibft_tgt_lun:
+		for (i = 0; i < 8; i++)
+			str += sprintf(str, "%x", (u8)tgt->lun[i]);
+		str += sprintf(str, "\n");
+		break;
+	case ibft_tgt_nic_assoc:
+		str += sprintf(str, "%d\n", tgt->nic_assoc);
+		break;
+	case ibft_tgt_chap_type:
+		str += sprintf(str, "%d\n", tgt->chap_type);
+		break;
+	case ibft_tgt_name:
+		str += sprintf_string(str, tgt->tgt_name_len,
+				      (char *)ibft_loc + tgt->tgt_name_off);
+		break;
+	case ibft_tgt_chap_name:
+		str += sprintf_string(str, tgt->chap_name_len,
+				      (char *)ibft_loc + tgt->chap_name_off);
+		break;
+	case ibft_tgt_chap_secret:
+		str += sprintf_string(str, tgt->chap_secret_len,
+				      (char *)ibft_loc + tgt->chap_secret_off);
+		break;
+	case ibft_tgt_rev_chap_name:
+		str += sprintf_string(str, tgt->rev_chap_name_len,
+				      (char *)ibft_loc +
+				      tgt->rev_chap_name_off);
+		break;
+	case ibft_tgt_rev_chap_secret:
+		str += sprintf_string(str, tgt->rev_chap_secret_len,
+				      (char *)ibft_loc +
+				      tgt->rev_chap_secret_off);
+		break;
+	default:
+		break;
+	}
+
+	return str - buf;
+}
+
+/*
+ * The routine called for all sysfs attributes.
+ */
+static ssize_t ibft_show_attribute(struct kobject *kobj,
+				    struct attribute *attr,
+				    char *buf)
+{
+	struct ibft_kobject *dev =
+		container_of(kobj, struct ibft_kobject, kobj);
+	struct ibft_attribute *ibft_attr =
+		container_of(attr, struct ibft_attribute, attr);
+	ssize_t ret = -EIO;
+	char *str = buf;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	if (ibft_attr->show)
+		ret = ibft_attr->show(dev, ibft_attr, str);
+
+	return ret;
+}
+
+static struct sysfs_ops ibft_attr_ops = {
+	.show = ibft_show_attribute,
+};
+
+static struct kobj_type ibft_ktype = {
+	.release = ibft_release,
+	.sysfs_ops = &ibft_attr_ops,
+};
+
+static struct kset *ibft_kset;
+
+static int __init ibft_check_device(void)
+{
+	int len;
+	u8 *pos;
+	u8 csum = 0;
+
+	len = ibft_addr->length;
+
+	/* Sanity checking of iBFT. */
+	if (ibft_addr->revision != 1) {
+		printk(KERN_ERR "iBFT module supports only revision 1, " \
+				"while this is %d.\n", ibft_addr->revision);
+		return -ENOENT;
+	}
+	for (pos = (u8 *)ibft_addr; pos < (u8 *)ibft_addr + len; pos++)
+		csum += *pos;
+
+	if (csum) {
+		printk(KERN_ERR "iBFT has incorrect checksum (0x%x)!\n", csum);
+		return -ENOENT;
+	}
+
+	return 0;
+}
+
+/*
+ * Helper function for ibft_register_kobjects.
+ */
+static int __init ibft_create_kobject(struct ibft_table_header *header,
+				       struct ibft_hdr *hdr,
+				       struct list_head *list)
+{
+	struct ibft_kobject *ibft_kobj = NULL;
+	struct ibft_nic *nic = (struct ibft_nic *)hdr;
+	struct pci_dev *pci_dev;
+	int rc = 0;
+
+	ibft_kobj = kzalloc(sizeof(*ibft_kobj), GFP_KERNEL);
+	if (!ibft_kobj)
+		return -ENOMEM;
+
+	ibft_kobj->header = header;
+	ibft_kobj->hdr = hdr;
+
+	switch (hdr->id) {
+	case id_initiator:
+		rc = ibft_verify_hdr("initiator", hdr, id_initiator,
+				     sizeof(*ibft_kobj->initiator));
+		break;
+	case id_nic:
+		rc = ibft_verify_hdr("ethernet", hdr, id_nic,
+				     sizeof(*ibft_kobj->nic));
+		break;
+	case id_target:
+		rc = ibft_verify_hdr("target", hdr, id_target,
+				     sizeof(*ibft_kobj->tgt));
+		break;
+	case id_reserved:
+	case id_control:
+	case id_extensions:
+		/* Fields which we don't support. Ignore them */
+		rc = 1;
+		break;
+	default:
+		printk(KERN_ERR "iBFT has unknown structure type (%d). " \
+				"Report this bug to %.6s!\n", hdr->id,
+				header->oem_id);
+		rc = 1;
+		break;
+	}
+
+	if (rc) {
+		/* Skip adding this kobject, but exit with non-fatal error. */
+		kfree(ibft_kobj);
+		goto out_invalid_struct;
+	}
+
+	ibft_kobj->kobj.kset = ibft_kset;
+
+	rc = kobject_init_and_add(&ibft_kobj->kobj, &ibft_ktype,
+				  NULL, ibft_id_names[hdr->id], hdr->index);
+
+	if (rc) {
+		kfree(ibft_kobj);
+		goto out;
+	}
+
+	kobject_uevent(&ibft_kobj->kobj, KOBJ_ADD);
+
+	if (hdr->id == id_nic) {
+		/*
+		* We don't search for the device in other domains than
+		* zero. This is because on x86 platforms the BIOS
+		* executes only devices which are in domain 0. Furthermore, the
+		* iBFT spec doesn't have a domain id field :-(
+		*/
+		pci_dev = pci_get_bus_and_slot((nic->pci_bdf & 0xff00) >> 8,
+					       (nic->pci_bdf & 0xff));
+		if (pci_dev) {
+			rc = sysfs_create_link(&ibft_kobj->kobj,
+					       &pci_dev->dev.kobj, "device");
+			pci_dev_put(pci_dev);
+		}
+	}
+
+	/* Nothing broke so lets add it to the list. */
+	list_add_tail(&ibft_kobj->node, list);
+out:
+	return rc;
+out_invalid_struct:
+	/* Unsupported structs are skipped. */
+	return 0;
+}
+
+/*
+ * Scan the IBFT table structure for the NIC and Target fields. When
+ * found add them on the passed-in list. We do not support the other
+ * fields at this point, so they are skipped.
+ */
+static int __init ibft_register_kobjects(struct ibft_table_header *header,
+					  struct list_head *list)
+{
+	struct ibft_control *control = NULL;
+	void *ptr, *end;
+	int rc = 0;
+	u16 offset;
+	u16 eot_offset;
+
+	control = (void *)header + sizeof(*header);
+	end = (void *)control + control->hdr.length;
+	eot_offset = (void *)header + header->length -
+		     (void *)control - sizeof(*header);
+	rc = ibft_verify_hdr("control", (struct ibft_hdr *)control, id_control,
+			     sizeof(*control));
+
+	/* iBFT table safety checking */
+	rc |= ((control->hdr.index) ? -ENODEV : 0);
+	if (rc) {
+		printk(KERN_ERR "iBFT error: Control header is invalid!\n");
+		return rc;
+	}
+	for (ptr = &control->initiator_off; ptr < end; ptr += sizeof(u16)) {
+		offset = *(u16 *)ptr;
+		if (offset && offset < header->length && offset < eot_offset) {
+			rc = ibft_create_kobject(header,
+						 (void *)header + offset,
+						 list);
+			if (rc)
+				break;
+		}
+	}
+
+	return rc;
+}
+
+static void ibft_unregister(struct list_head *attr_list,
+			     struct list_head *kobj_list)
+{
+	struct ibft_kobject *data = NULL, *n;
+	struct ibft_attribute *attr = NULL, *m;
+
+	list_for_each_entry_safe(attr, m, attr_list, node) {
+		sysfs_remove_file(attr->kobj, &attr->attr);
+		list_del(&attr->node);
+		kfree(attr);
+	};
+	list_del_init(attr_list);
+
+	list_for_each_entry_safe(data, n, kobj_list, node) {
+		list_del(&data->node);
+		if (data->hdr->id == id_nic)
+			sysfs_remove_link(&data->kobj, "device");
+		kobject_put(&data->kobj);
+	};
+	list_del_init(kobj_list);
+}
+
+static int __init ibft_create_attribute(struct ibft_kobject *kobj_data,
+					 int type,
+					 const char *name,
+					 ssize_t (*show)(struct ibft_kobject *,
+							 struct ibft_attribute*,
+							 char *buf),
+					 struct list_head *list)
+{
+	struct ibft_attribute *attr = NULL;
+	struct ibft_hdr *hdr = kobj_data->hdr;
+
+	attr = kmalloc(sizeof(*attr), GFP_KERNEL);
+	if (!attr)
+		return -ENOMEM;
+
+	attr->attr.name = name;
+	attr->attr.mode = S_IRUSR;
+	attr->attr.owner = THIS_MODULE;
+
+	attr->hdr = hdr;
+	attr->show = show;
+	attr->kobj = &kobj_data->kobj;
+	attr->type = type;
+
+	list_add_tail(&attr->node, list);
+
+	return 0;
+}
+
+/*
+ * Helper routiners to check to determine if the entry is valid
+ * in the proper iBFT structure.
+ */
+static int __init ibft_check_nic_for(struct ibft_nic *nic, int entry)
+{
+	int rc = 0;
+
+	switch (entry) {
+	case ibft_eth_index:
+	case ibft_eth_flags:
+		rc = 1;
+		break;
+	case ibft_eth_ip_addr:
+		if (!memcmp(nic->dhcp, nulls, sizeof(nic->dhcp)))
+			rc = 1;
+		break;
+	case ibft_eth_subnet_mask:
+		if (!memcmp(nic->dhcp, nulls, sizeof(nic->dhcp)))
+			rc = 1;
+		break;
+	case ibft_eth_origin:
+		rc = 1;
+		break;
+	case ibft_eth_gateway:
+		if (memcmp(nic->gateway, nulls, sizeof(nic->gateway)))
+			rc = 1;
+		break;
+	case ibft_eth_primary_dns:
+		if (memcmp(nic->primary_dns, nulls,
+			   sizeof(nic->primary_dns)))
+			rc = 1;
+		break;
+	case ibft_eth_secondary_dns:
+		if (memcmp(nic->secondary_dns, nulls,
+			   sizeof(nic->secondary_dns)))
+			rc = 1;
+		break;
+	case ibft_eth_dhcp:
+		if (memcmp(nic->dhcp, nulls, sizeof(nic->dhcp)))
+			rc = 1;
+		break;
+	case ibft_eth_vlan:
+	case ibft_eth_mac:
+		rc = 1;
+		break;
+	case ibft_eth_hostname:
+		if (nic->hostname_off)
+			rc = 1;
+		break;
+	default:
+		break;
+	}
+
+	return rc;
+}
+
+static int __init ibft_check_tgt_for(struct ibft_tgt *tgt, int entry)
+{
+	int rc = 0;
+
+	switch (entry) {
+	case ibft_tgt_index:
+	case ibft_tgt_flags:
+	case ibft_tgt_ip_addr:
+	case ibft_tgt_port:
+	case ibft_tgt_lun:
+	case ibft_tgt_nic_assoc:
+	case ibft_tgt_chap_type:
+		rc = 1;
+	case ibft_tgt_name:
+		if (tgt->tgt_name_len)
+			rc = 1;
+		break;
+	case ibft_tgt_chap_name:
+	case ibft_tgt_chap_secret:
+		if (tgt->chap_name_len)
+			rc = 1;
+		break;
+	case ibft_tgt_rev_chap_name:
+	case ibft_tgt_rev_chap_secret:
+		if (tgt->rev_chap_name_len)
+			rc = 1;
+		break;
+	default:
+		break;
+	}
+
+	return rc;
+}
+
+static int __init ibft_check_initiator_for(struct ibft_initiator *init,
+					    int entry)
+{
+	int rc = 0;
+
+	switch (entry) {
+	case ibft_init_index:
+	case ibft_init_flags:
+		rc = 1;
+		break;
+	case ibft_init_isns_server:
+		if (memcmp(init->isns_server, nulls,
+			   sizeof(init->isns_server)))
+			rc = 1;
+		break;
+	case ibft_init_slp_server:
+		if (memcmp(init->slp_server, nulls,
+			   sizeof(init->slp_server)))
+			rc = 1;
+		break;
+	case ibft_init_pri_radius_server:
+		if (memcmp(init->pri_radius_server, nulls,
+			   sizeof(init->pri_radius_server)))
+			rc = 1;
+		break;
+	case ibft_init_sec_radius_server:
+		if (memcmp(init->sec_radius_server, nulls,
+			   sizeof(init->sec_radius_server)))
+			rc = 1;
+		break;
+	case ibft_init_initiator_name:
+		if (init->initiator_name_len)
+			rc = 1;
+		break;
+	default:
+		break;
+	}
+
+	return rc;
+}
+
+/*
+ *  Register the attributes for all of the kobjects.
+ */
+static int __init ibft_register_attributes(struct list_head *kobject_list,
+					    struct list_head *attr_list)
+{
+	int rc = 0, i = 0;
+	struct ibft_kobject *data = NULL;
+	struct ibft_attribute *attr = NULL, *m;
+
+	list_for_each_entry(data, kobject_list, node) {
+		switch (data->hdr->id) {
+		case id_nic:
+			for (i = 0; i < ibft_eth_end_marker && !rc; i++)
+				if (ibft_check_nic_for(data->nic, i))
+					rc = ibft_create_attribute(data, i,
+						ibft_eth_properties[i],
+						ibft_attr_show_nic, attr_list);
+			break;
+		case id_target:
+			for (i = 0; i < ibft_tgt_end_marker && !rc; i++)
+				if (ibft_check_tgt_for(data->tgt, i))
+					rc = ibft_create_attribute(data, i,
+						ibft_tgt_properties[i],
+						ibft_attr_show_target,
+						attr_list);
+			break;
+		case id_initiator:
+			for (i = 0; i < ibft_init_end_marker && !rc; i++)
+				if (ibft_check_initiator_for(
+					data->initiator, i))
+					rc = ibft_create_attribute(data, i,
+						ibft_initiator_properties[i],
+						ibft_attr_show_initiator,
+						attr_list);
+			break;
+		default:
+			break;
+		}
+		if (rc)
+			break;
+	}
+	list_for_each_entry_safe(attr, m, attr_list, node) {
+		rc = sysfs_create_file(attr->kobj, &attr->attr);
+		if (rc) {
+			list_del(&attr->node);
+			kfree(attr);
+			break;
+		}
+	}
+
+	return rc;
+}
+
+/*
+ * ibft_init() - creates sysfs tree entries for the iBFT data.
+ */
+static int __init ibft_init(void)
+{
+	int rc = 0;
+
+	ibft_kset = kset_create_and_add("ibft", NULL, firmware_kobj);
+	if (!ibft_kset)
+		return -ENOMEM;
+
+	if (ibft_addr) {
+		printk(KERN_INFO "iBFT detected at 0x%lx.\n",
+		       virt_to_phys((void *)ibft_addr));
+
+		rc = ibft_check_device();
+		if (rc)
+			goto out_firmware_unregister;
+
+		/* Scan the IBFT for data and register the kobjects. */
+		rc = ibft_register_kobjects(ibft_addr, &ibft_kobject_list);
+		if (rc)
+			goto out_free;
+
+		/* Register the attributes */
+		rc = ibft_register_attributes(&ibft_kobject_list,
+					      &ibft_attr_list);
+		if (rc)
+			goto out_free;
+	} else
+		printk(KERN_INFO "No iBFT detected.\n");
+
+	return 0;
+
+out_free:
+	ibft_unregister(&ibft_attr_list, &ibft_kobject_list);
+out_firmware_unregister:
+	kset_unregister(ibft_kset);
+	return rc;
+}
+
+static void __exit ibft_exit(void)
+{
+	ibft_unregister(&ibft_attr_list, &ibft_kobject_list);
+	kset_unregister(ibft_kset);
+}
+
+module_init(ibft_init);
+module_exit(ibft_exit);
diff --git a/drivers/firmware/iscsi_ibft_find.c b/drivers/firmware/iscsi_ibft_find.c
new file mode 100644
index 000000000000..d0e5fa4ea51b
--- /dev/null
+++ b/drivers/firmware/iscsi_ibft_find.c
@@ -0,0 +1,84 @@
+/*
+ *  Copyright 2007 Red Hat, Inc.
+ *  by Peter Jones <pjones@redhat.com>
+ *  Copyright 2007 IBM, Inc.
+ *  by Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *  Copyright 2008
+ *  by Konrad Rzeszutek <ketuzsezr@darnok.org>
+ *
+ * This code finds the iSCSI Boot Format Table.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License v2.0 as published by
+ * the Free Software Foundation
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/bootmem.h>
+#include <linux/blkdev.h>
+#include <linux/ctype.h>
+#include <linux/device.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/limits.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+#include <linux/stat.h>
+#include <linux/string.h>
+#include <linux/types.h>
+
+#include <asm/mmzone.h>
+
+/*
+ * Physical location of iSCSI Boot Format Table.
+ */
+struct ibft_table_header *ibft_addr;
+EXPORT_SYMBOL_GPL(ibft_addr);
+
+#define IBFT_SIGN "iBFT"
+#define IBFT_SIGN_LEN 4
+#define IBFT_START 0x80000 /* 512kB */
+#define IBFT_END 0x100000 /* 1MB */
+#define VGA_MEM 0xA0000 /* VGA buffer */
+#define VGA_SIZE 0x20000 /* 128kB */
+
+
+/*
+ * Routine used to find the iSCSI Boot Format Table. The logical
+ * kernel address is set in the ibft_addr global variable.
+ */
+void __init reserve_ibft_region(void)
+{
+	unsigned long pos;
+	unsigned int len = 0;
+	void *virt;
+
+	ibft_addr = 0;
+
+	for (pos = IBFT_START; pos < IBFT_END; pos += 16) {
+		/* The table can't be inside the VGA BIOS reserved space,
+		 * so skip that area */
+		if (pos == VGA_MEM)
+			pos += VGA_SIZE;
+		virt = phys_to_virt(pos);
+		if (memcmp(virt, IBFT_SIGN, IBFT_SIGN_LEN) == 0) {
+			unsigned long *addr =
+			    (unsigned long *)phys_to_virt(pos + 4);
+			len = *addr;
+			/* if the length of the table extends past 1M,
+			 * the table cannot be valid. */
+			if (pos + len <= (IBFT_END-1)) {
+				ibft_addr = (struct ibft_table_header *)virt;
+				break;
+			}
+		}
+	}
+	if (ibft_addr)
+		reserve_bootmem(pos, PAGE_ALIGN(len), BOOTMEM_DEFAULT);
+}
+EXPORT_SYMBOL_GPL(reserve_ibft_region);
diff --git a/include/linux/iscsi_ibft.h b/include/linux/iscsi_ibft.h
new file mode 100644
index 000000000000..6092487e2950
--- /dev/null
+++ b/include/linux/iscsi_ibft.h
@@ -0,0 +1,50 @@
+/*
+ *  Copyright 2007 Red Hat, Inc.
+ *  by Peter Jones <pjones@redhat.com>
+ *  Copyright 2007 IBM, Inc.
+ *  by Konrad Rzeszutek <konradr@linux.vnet.ibm.com>
+ *  Copyright 2008
+ *  by Konrad Rzeszutek <ketuzsezr@darnok.org>
+ *
+ * This code exposes the iSCSI Boot Format Table to userland via sysfs.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License v2.0 as published by
+ * the Free Software Foundation
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef ISCSI_IBFT_H
+#define ISCSI_IBFT_H
+
+struct ibft_table_header {
+	char signature[4];
+	u32 length;
+	u8 revision;
+	u8 checksum;
+	char oem_id[6];
+	char oem_table_id[8];
+	char reserved[24];
+} __attribute__((__packed__));
+
+/*
+ * Logical location of iSCSI Boot Format Table.
+ * If the value is NULL there is no iBFT on the machine.
+ */
+extern struct ibft_table_header *ibft_addr;
+
+/*
+ * Routine used to find and reserve the iSCSI Boot Format Table. The
+ * mapped address is set in the ibft_addr variable.
+ */
+#ifdef CONFIG_ISCSI_IBFT_FIND
+extern void __init reserve_ibft_region(void);
+#else
+static inline void reserve_ibft_region(void) { }
+#endif
+
+#endif /* ISCSI_IBFT_H */
-- 
cgit v1.2.3


From b844eba292b477cda14582bfc6f535deed57a82d Mon Sep 17 00:00:00 2001
From: "Rafael J. Wysocki" <rjw@sisk.pl>
Date: Sun, 23 Mar 2008 20:28:24 +0100
Subject: PM: Remove destroy_suspended_device()

After 2.6.24 there was a plan to make the PM core acquire all device
semaphores during a suspend/hibernation to protect itself from
concurrent operations involving device objects.  That proved to be
too heavy-handed and we found a better way to achieve the goal, but
before it happened, we had introduced the functions
device_pm_schedule_removal() and destroy_suspended_device() to allow
drivers to "safely" destroy a suspended device and we had adapted some
drivers to use them.  Now that these functions are no longer necessary,
it seems reasonable to remove them and modify their users to use the
normal device unregistration instead.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 arch/x86/kernel/cpuid.c         |  4 +---
 arch/x86/kernel/msr.c           |  4 +---
 drivers/base/core.c             | 29 -----------------------------
 drivers/base/power/main.c       | 40 ----------------------------------------
 drivers/char/hw_random/core.c   | 10 +++++-----
 drivers/char/misc.c             | 13 ++++---------
 drivers/leds/led-class.c        | 11 +++--------
 drivers/net/wireless/b43/leds.c |  5 +----
 drivers/net/wireless/b43/main.c |  8 ++++----
 include/linux/device.h          | 14 --------------
 include/linux/hw_random.h       | 10 +---------
 include/linux/leds.h            | 10 +---------
 include/linux/miscdevice.h      | 10 +---------
 13 files changed, 22 insertions(+), 146 deletions(-)

(limited to 'include/linux')

diff --git a/arch/x86/kernel/cpuid.c b/arch/x86/kernel/cpuid.c
index 288e7a6598ac..daff52a62248 100644
--- a/arch/x86/kernel/cpuid.c
+++ b/arch/x86/kernel/cpuid.c
@@ -154,12 +154,10 @@ static int __cpuinit cpuid_class_cpu_callback(struct notifier_block *nfb,
 		err = cpuid_device_create(cpu);
 		break;
 	case CPU_UP_CANCELED:
+	case CPU_UP_CANCELED_FROZEN:
 	case CPU_DEAD:
 		cpuid_device_destroy(cpu);
 		break;
-	case CPU_UP_CANCELED_FROZEN:
-		destroy_suspended_device(cpuid_class, MKDEV(CPUID_MAJOR, cpu));
-		break;
 	}
 	return err ? NOTIFY_BAD : NOTIFY_OK;
 }
diff --git a/arch/x86/kernel/msr.c b/arch/x86/kernel/msr.c
index 4dfb40530057..1f3abe048e93 100644
--- a/arch/x86/kernel/msr.c
+++ b/arch/x86/kernel/msr.c
@@ -162,12 +162,10 @@ static int __cpuinit msr_class_cpu_callback(struct notifier_block *nfb,
 		err = msr_device_create(cpu);
 		break;
 	case CPU_UP_CANCELED:
+	case CPU_UP_CANCELED_FROZEN:
 	case CPU_DEAD:
 		msr_device_destroy(cpu);
 		break;
-	case CPU_UP_CANCELED_FROZEN:
-		destroy_suspended_device(msr_class, MKDEV(MSR_MAJOR, cpu));
-		break;
 	}
 	return err ? NOTIFY_BAD : NOTIFY_OK;
 }
diff --git a/drivers/base/core.c b/drivers/base/core.c
index adbc01788447..0262fc7c45fc 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -1163,35 +1163,6 @@ void device_destroy(struct class *class, dev_t devt)
 }
 EXPORT_SYMBOL_GPL(device_destroy);
 
-#ifdef CONFIG_PM_SLEEP
-/**
- * destroy_suspended_device - asks the PM core to remove a suspended device
- * @class: pointer to the struct class that this device was registered with
- * @devt: the dev_t of the device that was previously registered
- *
- * This call notifies the PM core of the necessity to unregister a suspended
- * device created with a call to device_create() (devices cannot be
- * unregistered directly while suspended, since the PM core holds their
- * semaphores at that time).
- *
- * It can only be called within the scope of a system sleep transition.  In
- * practice this means it has to be directly or indirectly invoked either by
- * a suspend or resume method, or by the PM core (e.g. via
- * disable_nonboot_cpus() or enable_nonboot_cpus()).
- */
-void destroy_suspended_device(struct class *class, dev_t devt)
-{
-	struct device *dev;
-
-	dev = class_find_device(class, &devt, __match_devt);
-	if (dev) {
-		device_pm_schedule_removal(dev);
-		put_device(dev);
-	}
-}
-EXPORT_SYMBOL_GPL(destroy_suspended_device);
-#endif /* CONFIG_PM_SLEEP */
-
 /**
  * device_rename - renames a device
  * @dev: the pointer to the struct device to be renamed
diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 5630af302b2f..c4568b82875b 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -50,7 +50,6 @@
 LIST_HEAD(dpm_active);
 static LIST_HEAD(dpm_off);
 static LIST_HEAD(dpm_off_irq);
-static LIST_HEAD(dpm_destroy);
 
 static DEFINE_MUTEX(dpm_list_mtx);
 
@@ -104,24 +103,6 @@ void device_pm_remove(struct device *dev)
 	mutex_unlock(&dpm_list_mtx);
 }
 
-/**
- *	device_pm_schedule_removal - schedule the removal of a suspended device
- *	@dev:	Device to destroy
- *
- *	Moves the device to the dpm_destroy list for further processing by
- *	unregister_dropped_devices().
- */
-void device_pm_schedule_removal(struct device *dev)
-{
-	pr_debug("PM: Preparing for removal: %s:%s\n",
-		dev->bus ? dev->bus->name : "No Bus",
-		kobject_name(&dev->kobj));
-	mutex_lock(&dpm_list_mtx);
-	list_move_tail(&dev->power.entry, &dpm_destroy);
-	mutex_unlock(&dpm_list_mtx);
-}
-EXPORT_SYMBOL_GPL(device_pm_schedule_removal);
-
 /*------------------------- Resume routines -------------------------*/
 
 /**
@@ -245,26 +226,6 @@ static void dpm_resume(void)
 	mutex_unlock(&dpm_list_mtx);
 }
 
-/**
- *	unregister_dropped_devices - Unregister devices scheduled for removal
- *
- *	Unregister all devices on the dpm_destroy list.
- */
-static void unregister_dropped_devices(void)
-{
-	mutex_lock(&dpm_list_mtx);
-	while (!list_empty(&dpm_destroy)) {
-		struct list_head *entry = dpm_destroy.next;
-		struct device *dev = to_device(entry);
-
-		mutex_unlock(&dpm_list_mtx);
-		/* This also removes the device from the list */
-		device_unregister(dev);
-		mutex_lock(&dpm_list_mtx);
-	}
-	mutex_unlock(&dpm_list_mtx);
-}
-
 /**
  *	device_resume - Restore state of each device in system.
  *
@@ -275,7 +236,6 @@ void device_resume(void)
 {
 	might_sleep();
 	dpm_resume();
-	unregister_dropped_devices();
 }
 EXPORT_SYMBOL_GPL(device_resume);
 
diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c
index 349b6edc5794..662d60e44e9a 100644
--- a/drivers/char/hw_random/core.c
+++ b/drivers/char/hw_random/core.c
@@ -238,11 +238,11 @@ static DEVICE_ATTR(rng_available, S_IRUGO,
 		   NULL);
 
 
-static void unregister_miscdev(bool suspended)
+static void unregister_miscdev(void)
 {
 	device_remove_file(rng_miscdev.this_device, &dev_attr_rng_available);
 	device_remove_file(rng_miscdev.this_device, &dev_attr_rng_current);
-	__misc_deregister(&rng_miscdev, suspended);
+	misc_deregister(&rng_miscdev);
 }
 
 static int register_miscdev(void)
@@ -317,7 +317,7 @@ out:
 }
 EXPORT_SYMBOL_GPL(hwrng_register);
 
-void __hwrng_unregister(struct hwrng *rng, bool suspended)
+void hwrng_unregister(struct hwrng *rng)
 {
 	int err;
 
@@ -336,11 +336,11 @@ void __hwrng_unregister(struct hwrng *rng, bool suspended)
 		}
 	}
 	if (list_empty(&rng_list))
-		unregister_miscdev(suspended);
+		unregister_miscdev();
 
 	mutex_unlock(&rng_mutex);
 }
-EXPORT_SYMBOL_GPL(__hwrng_unregister);
+EXPORT_SYMBOL_GPL(hwrng_unregister);
 
 
 MODULE_DESCRIPTION("H/W Random Number Generator (RNG) driver");
diff --git a/drivers/char/misc.c b/drivers/char/misc.c
index a39101feb2ed..4d058dadbfcc 100644
--- a/drivers/char/misc.c
+++ b/drivers/char/misc.c
@@ -232,9 +232,8 @@ int misc_register(struct miscdevice * misc)
 }
 
 /**
- *	__misc_deregister - unregister a miscellaneous device
+ *	misc_deregister - unregister a miscellaneous device
  *	@misc: device to unregister
- *	@suspended: to be set if the function is used during suspend/resume
  *
  *	Unregister a miscellaneous device that was previously
  *	successfully registered with misc_register(). Success
@@ -242,7 +241,7 @@ int misc_register(struct miscdevice * misc)
  *	indicates an error.
  */
 
-int __misc_deregister(struct miscdevice *misc, bool suspended)
+int misc_deregister(struct miscdevice *misc)
 {
 	int i = misc->minor;
 
@@ -251,11 +250,7 @@ int __misc_deregister(struct miscdevice *misc, bool suspended)
 
 	mutex_lock(&misc_mtx);
 	list_del(&misc->list);
-	if (suspended)
-		destroy_suspended_device(misc_class,
-					MKDEV(MISC_MAJOR, misc->minor));
-	else
-		device_destroy(misc_class, MKDEV(MISC_MAJOR, misc->minor));
+	device_destroy(misc_class, MKDEV(MISC_MAJOR, misc->minor));
 	if (i < DYNAMIC_MINORS && i>0) {
 		misc_minors[i>>3] &= ~(1 << (misc->minor & 7));
 	}
@@ -264,7 +259,7 @@ int __misc_deregister(struct miscdevice *misc, bool suspended)
 }
 
 EXPORT_SYMBOL(misc_register);
-EXPORT_SYMBOL(__misc_deregister);
+EXPORT_SYMBOL(misc_deregister);
 
 static int __init misc_init(void)
 {
diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c
index 4a938780dfc3..63aad90247c4 100644
--- a/drivers/leds/led-class.c
+++ b/drivers/leds/led-class.c
@@ -139,12 +139,10 @@ EXPORT_SYMBOL_GPL(led_classdev_register);
 /**
  * __led_classdev_unregister - unregisters a object of led_properties class.
  * @led_cdev: the led device to unregister
- * @suspended: indicates whether system-wide suspend or resume is in progress
  *
  * Unregisters a previously registered via led_classdev_register object.
  */
-void __led_classdev_unregister(struct led_classdev *led_cdev,
-				      bool suspended)
+void led_classdev_unregister(struct led_classdev *led_cdev)
 {
 	device_remove_file(led_cdev->dev, &dev_attr_brightness);
 #ifdef CONFIG_LEDS_TRIGGERS
@@ -155,16 +153,13 @@ void __led_classdev_unregister(struct led_classdev *led_cdev,
 	up_write(&led_cdev->trigger_lock);
 #endif
 
-	if (suspended)
-		device_pm_schedule_removal(led_cdev->dev);
-	else
-		device_unregister(led_cdev->dev);
+	device_unregister(led_cdev->dev);
 
 	down_write(&leds_list_lock);
 	list_del(&led_cdev->node);
 	up_write(&leds_list_lock);
 }
-EXPORT_SYMBOL_GPL(__led_classdev_unregister);
+EXPORT_SYMBOL_GPL(led_classdev_unregister);
 
 static int __init leds_init(void)
 {
diff --git a/drivers/net/wireless/b43/leds.c b/drivers/net/wireless/b43/leds.c
index 0aac1ff511df..36a9c42df835 100644
--- a/drivers/net/wireless/b43/leds.c
+++ b/drivers/net/wireless/b43/leds.c
@@ -116,10 +116,7 @@ static void b43_unregister_led(struct b43_led *led)
 {
 	if (!led->dev)
 		return;
-	if (led->dev->suspend_in_progress)
-		led_classdev_unregister_suspended(&led->led_dev);
-	else
-		led_classdev_unregister(&led->led_dev);
+	led_classdev_unregister(&led->led_dev);
 	b43_led_turn_off(led->dev, led->index, led->activelow);
 	led->dev = NULL;
 }
diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c
index cf5c046c9fa8..943cc851c504 100644
--- a/drivers/net/wireless/b43/main.c
+++ b/drivers/net/wireless/b43/main.c
@@ -2804,10 +2804,10 @@ static int b43_rng_read(struct hwrng *rng, u32 * data)
 	return (sizeof(u16));
 }
 
-static void b43_rng_exit(struct b43_wl *wl, bool suspended)
+static void b43_rng_exit(struct b43_wl *wl)
 {
 	if (wl->rng_initialized)
-		__hwrng_unregister(&wl->rng, suspended);
+		hwrng_unregister(&wl->rng);
 }
 
 static int b43_rng_init(struct b43_wl *wl)
@@ -3824,7 +3824,7 @@ static void b43_wireless_core_exit(struct b43_wldev *dev)
 
 	if (!dev->suspend_in_progress) {
 		b43_leds_exit(dev);
-		b43_rng_exit(dev->wl, false);
+		b43_rng_exit(dev->wl);
 	}
 	b43_dma_free(dev);
 	b43_pio_free(dev);
@@ -4589,7 +4589,7 @@ static int b43_resume(struct ssb_device *dev)
 		err = b43_wireless_core_start(wldev);
 		if (err) {
 			b43_leds_exit(wldev);
-			b43_rng_exit(wldev->wl, true);
+			b43_rng_exit(wldev->wl);
 			b43_wireless_core_exit(wldev);
 			b43err(wl, "Resume failed at core start\n");
 			goto out;
diff --git a/include/linux/device.h b/include/linux/device.h
index 441461f5ee20..dc3429e2eb0f 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -545,20 +545,6 @@ extern struct device *device_create(struct class *cls, struct device *parent,
 				    dev_t devt, const char *fmt, ...)
 				    __attribute__((format(printf, 4, 5)));
 extern void device_destroy(struct class *cls, dev_t devt);
-#ifdef CONFIG_PM_SLEEP
-extern void destroy_suspended_device(struct class *cls, dev_t devt);
-extern void device_pm_schedule_removal(struct device *);
-#else /* !CONFIG_PM_SLEEP */
-static inline void destroy_suspended_device(struct class *cls, dev_t devt)
-{
-	device_destroy(cls, devt);
-}
-
-static inline void device_pm_schedule_removal(struct device *dev)
-{
-	device_unregister(dev);
-}
-#endif /* !CONFIG_PM_SLEEP */
 
 /*
  * Platform "fixup" functions - allow the platform to have their say
diff --git a/include/linux/hw_random.h b/include/linux/hw_random.h
index 42131820bb89..85d11916e9ea 100644
--- a/include/linux/hw_random.h
+++ b/include/linux/hw_random.h
@@ -44,15 +44,7 @@ struct hwrng {
 /** Register a new Hardware Random Number Generator driver. */
 extern int hwrng_register(struct hwrng *rng);
 /** Unregister a Hardware Random Number Generator driver. */
-extern void __hwrng_unregister(struct hwrng *rng, bool suspended);
-static inline void hwrng_unregister(struct hwrng *rng)
-{
-	__hwrng_unregister(rng, false);
-}
-static inline void hwrng_unregister_suspended(struct hwrng *rng)
-{
-	__hwrng_unregister(rng, true);
-}
+extern void hwrng_unregister(struct hwrng *rng);
 
 #endif /* __KERNEL__ */
 #endif /* LINUX_HWRANDOM_H_ */
diff --git a/include/linux/leds.h b/include/linux/leds.h
index 0201f6f51cea..b07e3d400bd6 100644
--- a/include/linux/leds.h
+++ b/include/linux/leds.h
@@ -59,15 +59,7 @@ struct led_classdev {
 
 extern int led_classdev_register(struct device *parent,
 				 struct led_classdev *led_cdev);
-extern void __led_classdev_unregister(struct led_classdev *led_cdev, bool sus);
-static inline void led_classdev_unregister(struct led_classdev *lcd)
-{
-	__led_classdev_unregister(lcd, false);
-}
-static inline void led_classdev_unregister_suspended(struct led_classdev *lcd)
-{
-	__led_classdev_unregister(lcd, true);
-}
+extern void led_classdev_unregister(struct led_classdev *lcd);
 extern void led_classdev_suspend(struct led_classdev *led_cdev);
 extern void led_classdev_resume(struct led_classdev *led_cdev);
 
diff --git a/include/linux/miscdevice.h b/include/linux/miscdevice.h
index 24b30b9b4f8a..26433ec520b3 100644
--- a/include/linux/miscdevice.h
+++ b/include/linux/miscdevice.h
@@ -43,15 +43,7 @@ struct miscdevice  {
 };
 
 extern int misc_register(struct miscdevice * misc);
-extern int __misc_deregister(struct miscdevice *misc, bool suspended);
-static inline int misc_deregister(struct miscdevice *misc)
-{
-	return __misc_deregister(misc, false);
-}
-static inline int misc_deregister_suspended(struct miscdevice *misc)
-{
-	return __misc_deregister(misc, true);
-}
+extern int misc_deregister(struct miscdevice *misc);
 
 #define MODULE_ALIAS_MISCDEV(minor)				\
 	MODULE_ALIAS("char-major-" __stringify(MISC_MAJOR)	\
-- 
cgit v1.2.3


From c4c66cf1787d408066fbfc69209185701f5df15f Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Tue, 4 Mar 2008 00:13:36 +0100
Subject: memstick: convert struct class_device to struct device

struct class_device is going away, struct device should be used instead.

Signed-off-by: Tony Jones <tonyj@suse.de>
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Cc: Alex Dubov <oakad@yahoo.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/memstick/core/memstick.c    | 33 ++++++++++++++++-----------------
 drivers/memstick/core/mspro_block.c |  4 ++--
 drivers/memstick/host/jmb38x_ms.c   | 16 ++++++++--------
 include/linux/memstick.h            |  2 +-
 4 files changed, 27 insertions(+), 28 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/memstick/core/memstick.c b/drivers/memstick/core/memstick.c
index 946e3d3506ac..61b98c333cb0 100644
--- a/drivers/memstick/core/memstick.c
+++ b/drivers/memstick/core/memstick.c
@@ -177,16 +177,16 @@ static struct bus_type memstick_bus_type = {
 	.resume         = memstick_device_resume
 };
 
-static void memstick_free(struct class_device *cdev)
+static void memstick_free(struct device *dev)
 {
-	struct memstick_host *host = container_of(cdev, struct memstick_host,
-						  cdev);
+	struct memstick_host *host = container_of(dev, struct memstick_host,
+						  dev);
 	kfree(host);
 }
 
 static struct class memstick_host_class = {
 	.name       = "memstick_host",
-	.release    = memstick_free
+	.dev_release = memstick_free
 };
 
 static void memstick_free_card(struct device *dev)
@@ -383,8 +383,8 @@ static struct memstick_dev *memstick_alloc_card(struct memstick_host *host)
 	if (card) {
 		card->host = host;
 		snprintf(card->dev.bus_id, sizeof(card->dev.bus_id),
-			 "%s", host->cdev.class_id);
-		card->dev.parent = host->cdev.dev;
+			 "%s", host->dev.bus_id);
+		card->dev.parent = &host->dev;
 		card->dev.bus = &memstick_bus_type;
 		card->dev.release = memstick_free_card;
 		card->check = memstick_dummy_check;
@@ -427,7 +427,7 @@ static void memstick_check(struct work_struct *work)
 						  media_checker);
 	struct memstick_dev *card;
 
-	dev_dbg(host->cdev.dev, "memstick_check started\n");
+	dev_dbg(&host->dev, "memstick_check started\n");
 	mutex_lock(&host->lock);
 	if (!host->card)
 		memstick_power_on(host);
@@ -440,7 +440,7 @@ static void memstick_check(struct work_struct *work)
 			host->card = NULL;
 		}
 	} else {
-		dev_dbg(host->cdev.dev, "new card %02x, %02x, %02x\n",
+		dev_dbg(&host->dev, "new card %02x, %02x, %02x\n",
 			card->id.type, card->id.category, card->id.class);
 		if (host->card) {
 			if (memstick_set_rw_addr(host->card)
@@ -465,7 +465,7 @@ static void memstick_check(struct work_struct *work)
 		host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_OFF);
 
 	mutex_unlock(&host->lock);
-	dev_dbg(host->cdev.dev, "memstick_check finished\n");
+	dev_dbg(&host->dev, "memstick_check finished\n");
 }
 
 /**
@@ -482,9 +482,9 @@ struct memstick_host *memstick_alloc_host(unsigned int extra,
 	if (host) {
 		mutex_init(&host->lock);
 		INIT_WORK(&host->media_checker, memstick_check);
-		host->cdev.class = &memstick_host_class;
-		host->cdev.dev = dev;
-		class_device_initialize(&host->cdev);
+		host->dev.class = &memstick_host_class;
+		host->dev.parent = dev;
+		device_initialize(&host->dev);
 	}
 	return host;
 }
@@ -507,10 +507,9 @@ int memstick_add_host(struct memstick_host *host)
 	if (rc)
 		return rc;
 
-	snprintf(host->cdev.class_id, BUS_ID_SIZE,
-		 "memstick%u", host->id);
+	snprintf(host->dev.bus_id, BUS_ID_SIZE, "memstick%u", host->id);
 
-	rc = class_device_add(&host->cdev);
+	rc = device_add(&host->dev);
 	if (rc) {
 		spin_lock(&memstick_host_lock);
 		idr_remove(&memstick_host_idr, host->id);
@@ -541,7 +540,7 @@ void memstick_remove_host(struct memstick_host *host)
 	spin_lock(&memstick_host_lock);
 	idr_remove(&memstick_host_idr, host->id);
 	spin_unlock(&memstick_host_lock);
-	class_device_del(&host->cdev);
+	device_del(&host->dev);
 }
 EXPORT_SYMBOL(memstick_remove_host);
 
@@ -552,7 +551,7 @@ EXPORT_SYMBOL(memstick_remove_host);
 void memstick_free_host(struct memstick_host *host)
 {
 	mutex_destroy(&host->lock);
-	class_device_put(&host->cdev);
+	put_device(&host->dev);
 }
 EXPORT_SYMBOL(memstick_free_host);
 
diff --git a/drivers/memstick/core/mspro_block.c b/drivers/memstick/core/mspro_block.c
index 557dbbba5cb2..477d0fb6e588 100644
--- a/drivers/memstick/core/mspro_block.c
+++ b/drivers/memstick/core/mspro_block.c
@@ -1127,8 +1127,8 @@ static int mspro_block_init_disk(struct memstick_dev *card)
 	u64 limit = BLK_BOUNCE_HIGH;
 	unsigned long capacity;
 
-	if (host->cdev.dev->dma_mask && *(host->cdev.dev->dma_mask))
-		limit = *(host->cdev.dev->dma_mask);
+	if (host->dev.dma_mask && *(host->dev.dma_mask))
+		limit = *(host->dev.dma_mask);
 
 	for (rc = 0; msb->attr_group.attrs[rc]; ++rc) {
 		s_attr = mspro_from_sysfs_attr(msb->attr_group.attrs[rc]);
diff --git a/drivers/memstick/host/jmb38x_ms.c b/drivers/memstick/host/jmb38x_ms.c
index 8770a5fac3b6..a054668eda16 100644
--- a/drivers/memstick/host/jmb38x_ms.c
+++ b/drivers/memstick/host/jmb38x_ms.c
@@ -361,15 +361,15 @@ static int jmb38x_ms_issue_cmd(struct memstick_host *msh)
 	unsigned int data_len, cmd, t_val;
 
 	if (!(STATUS_HAS_MEDIA & readl(host->addr + STATUS))) {
-		dev_dbg(msh->cdev.dev, "no media status\n");
+		dev_dbg(&msh->dev, "no media status\n");
 		host->req->error = -ETIME;
 		return host->req->error;
 	}
 
-	dev_dbg(msh->cdev.dev, "control %08x\n",
+	dev_dbg(&msh->dev, "control %08x\n",
 		readl(host->addr + HOST_CONTROL));
-	dev_dbg(msh->cdev.dev, "status %08x\n", readl(host->addr + INT_STATUS));
-	dev_dbg(msh->cdev.dev, "hstatus %08x\n", readl(host->addr + STATUS));
+	dev_dbg(&msh->dev, "status %08x\n", readl(host->addr + INT_STATUS));
+	dev_dbg(&msh->dev, "hstatus %08x\n", readl(host->addr + STATUS));
 
 	host->cmd_flags = 0;
 	host->block_pos = 0;
@@ -448,7 +448,7 @@ static int jmb38x_ms_issue_cmd(struct memstick_host *msh)
 	host->req->error = 0;
 
 	writel(cmd, host->addr + TPC);
-	dev_dbg(msh->cdev.dev, "executing TPC %08x, len %x\n", cmd, data_len);
+	dev_dbg(&msh->dev, "executing TPC %08x, len %x\n", cmd, data_len);
 
 	return 0;
 }
@@ -461,11 +461,11 @@ static void jmb38x_ms_complete_cmd(struct memstick_host *msh, int last)
 
 	del_timer(&host->timer);
 
-	dev_dbg(msh->cdev.dev, "c control %08x\n",
+	dev_dbg(&msh->dev, "c control %08x\n",
 		readl(host->addr + HOST_CONTROL));
-	dev_dbg(msh->cdev.dev, "c status %08x\n",
+	dev_dbg(&msh->dev, "c status %08x\n",
 		readl(host->addr + INT_STATUS));
-	dev_dbg(msh->cdev.dev, "c hstatus %08x\n", readl(host->addr + STATUS));
+	dev_dbg(&msh->dev, "c hstatus %08x\n", readl(host->addr + STATUS));
 
 	host->req->int_reg = readl(host->addr + STATUS) & 0xff;
 
diff --git a/include/linux/memstick.h b/include/linux/memstick.h
index 3e686ec6a967..37a5cdb03918 100644
--- a/include/linux/memstick.h
+++ b/include/linux/memstick.h
@@ -276,7 +276,7 @@ struct memstick_host {
 #define MEMSTICK_CAP_PAR8          4
 
 	struct work_struct  media_checker;
-	struct class_device cdev;
+	struct device       dev;
 
 	struct memstick_dev *card;
 	unsigned int        retries;
-- 
cgit v1.2.3


From ee959b00c335d7780136c5abda37809191fe52c3 Mon Sep 17 00:00:00 2001
From: Tony Jones <tonyj@suse.de>
Date: Fri, 22 Feb 2008 00:13:36 +0100
Subject: SCSI: convert struct class_device to struct device

It's big, but there doesn't seem to be a way to split it up smaller...

Signed-off-by: Tony Jones <tonyj@suse.de>
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 block/bsg.c                           |  11 +-
 drivers/ata/ahci.c                    |   4 +-
 drivers/ata/libata-scsi.c             |  15 +-
 drivers/base/attribute_container.c    |  77 +++---
 drivers/base/transport_class.c        |  14 +-
 drivers/infiniband/ulp/srp/ib_srp.c   | 132 +++++-----
 drivers/infiniband/ulp/srp/ib_srp.h   |   2 +-
 drivers/message/fusion/mptscsih.c     | 122 +++++----
 drivers/message/fusion/mptscsih.h     |   2 +-
 drivers/misc/enclosure.c              | 118 +++++----
 drivers/scsi/3w-9xxx.c                |   9 +-
 drivers/scsi/3w-xxxx.c                |   9 +-
 drivers/scsi/aacraid/aachba.c         |   2 +-
 drivers/scsi/aacraid/aacraid.h        |   4 +-
 drivers/scsi/aacraid/linit.c          |  93 ++++---
 drivers/scsi/arcmsr/arcmsr.h          |   4 +-
 drivers/scsi/arcmsr/arcmsr_attr.c     | 163 +++++++-----
 drivers/scsi/ch.c                     |  13 +-
 drivers/scsi/hosts.c                  |  34 +--
 drivers/scsi/hptiop.c                 |  14 +-
 drivers/scsi/ibmvscsi/ibmvscsi.c      |  44 ++--
 drivers/scsi/ibmvscsi/ibmvstgt.c      |  25 +-
 drivers/scsi/ipr.c                    | 140 +++++-----
 drivers/scsi/lpfc/lpfc_attr.c         | 484 ++++++++++++++++++----------------
 drivers/scsi/lpfc/lpfc_crtn.h         |   4 +-
 drivers/scsi/megaraid/megaraid_mbox.c |  13 +-
 drivers/scsi/ncr53c8xx.c              |   7 +-
 drivers/scsi/osst.c                   |  76 +++---
 drivers/scsi/pcmcia/sym53c500_cs.c    |  14 +-
 drivers/scsi/qla2xxx/qla_attr.c       | 177 +++++++------
 drivers/scsi/qla2xxx/qla_gbl.h        |   4 +-
 drivers/scsi/raid_class.c             |  73 ++---
 drivers/scsi/scsi_sas_internal.h      |  24 +-
 drivers/scsi/scsi_sysfs.c             | 150 ++++++-----
 drivers/scsi/scsi_transport_fc.c      | 385 ++++++++++++++-------------
 drivers/scsi/scsi_transport_iscsi.c   | 107 ++++----
 drivers/scsi/scsi_transport_sas.c     | 157 ++++++-----
 drivers/scsi/scsi_transport_spi.c     | 185 +++++++------
 drivers/scsi/scsi_transport_srp.c     |  26 +-
 drivers/scsi/sd.c                     |  78 +++---
 drivers/scsi/ses.c                    |  28 +-
 drivers/scsi/sg.c                     |  36 +--
 drivers/scsi/st.c                     |  82 +++---
 include/linux/attribute_container.h   |  28 +-
 include/linux/bsg.h                   |   2 +-
 include/linux/enclosure.h             |  11 +-
 include/linux/libata.h                |   2 +-
 include/linux/raid_class.h            |  12 +-
 include/linux/transport_class.h       |   6 +-
 include/scsi/scsi_device.h            |  10 +-
 include/scsi/scsi_host.h              |   7 +-
 include/scsi/scsi_transport.h         |   2 +-
 include/scsi/scsi_transport_fc.h      |  14 +-
 include/scsi/scsi_transport_sas.h     |  12 +-
 include/scsi/sd.h                     |   4 +-
 55 files changed, 1761 insertions(+), 1510 deletions(-)

(limited to 'include/linux')

diff --git a/block/bsg.c b/block/bsg.c
index 302ac1f5af39..f51172ed27c2 100644
--- a/block/bsg.c
+++ b/block/bsg.c
@@ -758,7 +758,7 @@ static struct bsg_device *bsg_add_device(struct inode *inode,
 	mutex_lock(&bsg_mutex);
 	hlist_add_head(&bd->dev_list, bsg_dev_idx_hash(iminor(inode)));
 
-	strncpy(bd->name, rq->bsg_dev.class_dev->class_id, sizeof(bd->name) - 1);
+	strncpy(bd->name, rq->bsg_dev.class_dev->bus_id, sizeof(bd->name) - 1);
 	dprintk("bound to <%s>, max queue %d\n",
 		format_dev_t(buf, inode->i_rdev), bd->max_queue);
 
@@ -946,7 +946,7 @@ void bsg_unregister_queue(struct request_queue *q)
 	mutex_lock(&bsg_mutex);
 	idr_remove(&bsg_minor_idr, bcd->minor);
 	sysfs_remove_link(&q->kobj, "bsg");
-	class_device_unregister(bcd->class_dev);
+	device_unregister(bcd->class_dev);
 	put_device(bcd->dev);
 	bcd->class_dev = NULL;
 	mutex_unlock(&bsg_mutex);
@@ -959,7 +959,7 @@ int bsg_register_queue(struct request_queue *q, struct device *gdev,
 	struct bsg_class_device *bcd;
 	dev_t dev;
 	int ret, minor;
-	struct class_device *class_dev = NULL;
+	struct device *class_dev = NULL;
 	const char *devname;
 
 	if (name)
@@ -998,8 +998,7 @@ int bsg_register_queue(struct request_queue *q, struct device *gdev,
 	bcd->queue = q;
 	bcd->dev = get_device(gdev);
 	dev = MKDEV(bsg_major, bcd->minor);
-	class_dev = class_device_create(bsg_class, NULL, dev, gdev, "%s",
-					devname);
+	class_dev = device_create(bsg_class, gdev, dev, "%s", devname);
 	if (IS_ERR(class_dev)) {
 		ret = PTR_ERR(class_dev);
 		goto put_dev;
@@ -1016,7 +1015,7 @@ int bsg_register_queue(struct request_queue *q, struct device *gdev,
 	return 0;
 
 unregister_class_dev:
-	class_device_unregister(class_dev);
+	device_unregister(class_dev);
 put_dev:
 	put_device(gdev);
 remove_idr:
diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c
index 739ba3f222e8..986e3324e302 100644
--- a/drivers/ata/ahci.c
+++ b/drivers/ata/ahci.c
@@ -273,8 +273,8 @@ static int ahci_pci_device_suspend(struct pci_dev *pdev, pm_message_t mesg);
 static int ahci_pci_device_resume(struct pci_dev *pdev);
 #endif
 
-static struct class_device_attribute *ahci_shost_attrs[] = {
-	&class_device_attr_link_power_management_policy,
+static struct device_attribute *ahci_shost_attrs[] = {
+	&dev_attr_link_power_management_policy,
 	NULL
 };
 
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index f3c69a8c1103..a34f32442edf 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -131,10 +131,11 @@ static const char *ata_scsi_lpm_get(enum link_pm policy)
 	return NULL;
 }
 
-static ssize_t ata_scsi_lpm_put(struct class_device *class_dev,
-	const char *buf, size_t count)
+static ssize_t ata_scsi_lpm_put(struct device *dev,
+				struct device_attribute *attr,
+				const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ata_port *ap = ata_shost_to_port(shost);
 	enum link_pm policy = 0;
 	int i;
@@ -162,9 +163,9 @@ static ssize_t ata_scsi_lpm_put(struct class_device *class_dev,
 }
 
 static ssize_t
-ata_scsi_lpm_show(struct class_device *class_dev, char *buf)
+ata_scsi_lpm_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ata_port *ap = ata_shost_to_port(shost);
 	const char *policy =
 		ata_scsi_lpm_get(ap->pm_policy);
@@ -174,9 +175,9 @@ ata_scsi_lpm_show(struct class_device *class_dev, char *buf)
 
 	return snprintf(buf, 23, "%s\n", policy);
 }
-CLASS_DEVICE_ATTR(link_power_management_policy, S_IRUGO | S_IWUSR,
+DEVICE_ATTR(link_power_management_policy, S_IRUGO | S_IWUSR,
 		ata_scsi_lpm_show, ata_scsi_lpm_put);
-EXPORT_SYMBOL_GPL(class_device_attr_link_power_management_policy);
+EXPORT_SYMBOL_GPL(dev_attr_link_power_management_policy);
 
 static void ata_scsi_invalid_field(struct scsi_cmnd *cmd,
 				   void (*done)(struct scsi_cmnd *))
diff --git a/drivers/base/attribute_container.c b/drivers/base/attribute_container.c
index 3b43e8a9f87e..f57652db0a2a 100644
--- a/drivers/base/attribute_container.c
+++ b/drivers/base/attribute_container.c
@@ -27,21 +27,21 @@
 struct internal_container {
 	struct klist_node node;
 	struct attribute_container *cont;
-	struct class_device classdev;
+	struct device classdev;
 };
 
 static void internal_container_klist_get(struct klist_node *n)
 {
 	struct internal_container *ic =
 		container_of(n, struct internal_container, node);
-	class_device_get(&ic->classdev);
+	get_device(&ic->classdev);
 }
 
 static void internal_container_klist_put(struct klist_node *n)
 {
 	struct internal_container *ic =
 		container_of(n, struct internal_container, node);
-	class_device_put(&ic->classdev);
+	put_device(&ic->classdev);
 }
 
 
@@ -53,7 +53,7 @@ static void internal_container_klist_put(struct klist_node *n)
  * Returns the container associated with this classdev.
  */
 struct attribute_container *
-attribute_container_classdev_to_container(struct class_device *classdev)
+attribute_container_classdev_to_container(struct device *classdev)
 {
 	struct internal_container *ic =
 		container_of(classdev, struct internal_container, classdev);
@@ -110,11 +110,11 @@ attribute_container_unregister(struct attribute_container *cont)
 EXPORT_SYMBOL_GPL(attribute_container_unregister);
 
 /* private function used as class release */
-static void attribute_container_release(struct class_device *classdev)
+static void attribute_container_release(struct device *classdev)
 {
 	struct internal_container *ic 
 		= container_of(classdev, struct internal_container, classdev);
-	struct device *dev = classdev->dev;
+	struct device *dev = classdev->parent;
 
 	kfree(ic);
 	put_device(dev);
@@ -129,12 +129,12 @@ static void attribute_container_release(struct class_device *classdev)
  * This function allocates storage for the class device(s) to be
  * attached to dev (one for each matching attribute_container).  If no
  * fn is provided, the code will simply register the class device via
- * class_device_add.  If a function is provided, it is expected to add
+ * device_add.  If a function is provided, it is expected to add
  * the class device at the appropriate time.  One of the things that
  * might be necessary is to allocate and initialise the classdev and
  * then add it a later time.  To do this, call this routine for
  * allocation and initialisation and then use
- * attribute_container_device_trigger() to call class_device_add() on
+ * attribute_container_device_trigger() to call device_add() on
  * it.  Note: after this, the class device contains a reference to dev
  * which is not relinquished until the release of the classdev.
  */
@@ -142,7 +142,7 @@ void
 attribute_container_add_device(struct device *dev,
 			       int (*fn)(struct attribute_container *,
 					 struct device *,
-					 struct class_device *))
+					 struct device *))
 {
 	struct attribute_container *cont;
 
@@ -163,11 +163,11 @@ attribute_container_add_device(struct device *dev,
 		}
 
 		ic->cont = cont;
-		class_device_initialize(&ic->classdev);
-		ic->classdev.dev = get_device(dev);
+		device_initialize(&ic->classdev);
+		ic->classdev.parent = get_device(dev);
 		ic->classdev.class = cont->class;
-		cont->class->release = attribute_container_release;
-		strcpy(ic->classdev.class_id, dev->bus_id);
+		cont->class->dev_release = attribute_container_release;
+		strcpy(ic->classdev.bus_id, dev->bus_id);
 		if (fn)
 			fn(cont, dev, &ic->classdev);
 		else
@@ -195,20 +195,19 @@ attribute_container_add_device(struct device *dev,
  * @fn:	  A function to call to remove the device
  *
  * This routine triggers device removal.  If fn is NULL, then it is
- * simply done via class_device_unregister (note that if something
+ * simply done via device_unregister (note that if something
  * still has a reference to the classdev, then the memory occupied
  * will not be freed until the classdev is released).  If you want a
  * two phase release: remove from visibility and then delete the
  * device, then you should use this routine with a fn that calls
- * class_device_del() and then use
- * attribute_container_device_trigger() to do the final put on the
- * classdev.
+ * device_del() and then use attribute_container_device_trigger()
+ * to do the final put on the classdev.
  */
 void
 attribute_container_remove_device(struct device *dev,
 				  void (*fn)(struct attribute_container *,
 					     struct device *,
-					     struct class_device *))
+					     struct device *))
 {
 	struct attribute_container *cont;
 
@@ -224,14 +223,14 @@ attribute_container_remove_device(struct device *dev,
 			continue;
 
 		klist_for_each_entry(ic, &cont->containers, node, &iter) {
-			if (dev != ic->classdev.dev)
+			if (dev != ic->classdev.parent)
 				continue;
 			klist_del(&ic->node);
 			if (fn)
 				fn(cont, dev, &ic->classdev);
 			else {
 				attribute_container_remove_attrs(&ic->classdev);
-				class_device_unregister(&ic->classdev);
+				device_unregister(&ic->classdev);
 			}
 		}
 	}
@@ -252,7 +251,7 @@ void
 attribute_container_device_trigger(struct device *dev, 
 				   int (*fn)(struct attribute_container *,
 					     struct device *,
-					     struct class_device *))
+					     struct device *))
 {
 	struct attribute_container *cont;
 
@@ -270,7 +269,7 @@ attribute_container_device_trigger(struct device *dev,
 		}
 
 		klist_for_each_entry(ic, &cont->containers, node, &iter) {
-			if (dev == ic->classdev.dev)
+			if (dev == ic->classdev.parent)
 				fn(cont, dev, &ic->classdev);
 		}
 	}
@@ -313,11 +312,11 @@ attribute_container_trigger(struct device *dev,
  * attributes listed in the container
  */
 int
-attribute_container_add_attrs(struct class_device *classdev)
+attribute_container_add_attrs(struct device *classdev)
 {
 	struct attribute_container *cont =
 		attribute_container_classdev_to_container(classdev);
-	struct class_device_attribute **attrs =	cont->attrs;
+	struct device_attribute **attrs = cont->attrs;
 	int i, error;
 
 	BUG_ON(attrs && cont->grp);
@@ -329,7 +328,7 @@ attribute_container_add_attrs(struct class_device *classdev)
 		return sysfs_create_group(&classdev->kobj, cont->grp);
 
 	for (i = 0; attrs[i]; i++) {
-		error = class_device_create_file(classdev, attrs[i]);
+		error = device_create_file(classdev, attrs[i]);
 		if (error)
 			return error;
 	}
@@ -338,18 +337,18 @@ attribute_container_add_attrs(struct class_device *classdev)
 }
 
 /**
- * attribute_container_add_class_device - same function as class_device_add
+ * attribute_container_add_class_device - same function as device_add
  *
  * @classdev:	the class device to add
  *
- * This performs essentially the same function as class_device_add except for
+ * This performs essentially the same function as device_add except for
  * attribute containers, namely add the classdev to the system and then
  * create the attribute files
  */
 int
-attribute_container_add_class_device(struct class_device *classdev)
+attribute_container_add_class_device(struct device *classdev)
 {
-	int error = class_device_add(classdev);
+	int error = device_add(classdev);
 	if (error)
 		return error;
 	return attribute_container_add_attrs(classdev);
@@ -364,7 +363,7 @@ attribute_container_add_class_device(struct class_device *classdev)
 int
 attribute_container_add_class_device_adapter(struct attribute_container *cont,
 					     struct device *dev,
-					     struct class_device *classdev)
+					     struct device *classdev)
 {
 	return attribute_container_add_class_device(classdev);
 }
@@ -376,11 +375,11 @@ attribute_container_add_class_device_adapter(struct attribute_container *cont,
  *
  */
 void
-attribute_container_remove_attrs(struct class_device *classdev)
+attribute_container_remove_attrs(struct device *classdev)
 {
 	struct attribute_container *cont =
 		attribute_container_classdev_to_container(classdev);
-	struct class_device_attribute **attrs =	cont->attrs;
+	struct device_attribute **attrs = cont->attrs;
 	int i;
 
 	if (!attrs && !cont->grp)
@@ -392,7 +391,7 @@ attribute_container_remove_attrs(struct class_device *classdev)
 	}
 
 	for (i = 0; attrs[i]; i++)
-		class_device_remove_file(classdev, attrs[i]);
+		device_remove_file(classdev, attrs[i]);
 }
 
 /**
@@ -401,13 +400,13 @@ attribute_container_remove_attrs(struct class_device *classdev)
  * @classdev: the class device
  *
  * This function simply removes all the attribute files and then calls
- * class_device_del.
+ * device_del.
  */
 void
-attribute_container_class_device_del(struct class_device *classdev)
+attribute_container_class_device_del(struct device *classdev)
 {
 	attribute_container_remove_attrs(classdev);
-	class_device_del(classdev);
+	device_del(classdev);
 }
 
 /**
@@ -419,16 +418,16 @@ attribute_container_class_device_del(struct class_device *classdev)
  * Looks up the device in the container's list of class devices and returns
  * the corresponding class_device.
  */
-struct class_device *
+struct device *
 attribute_container_find_class_device(struct attribute_container *cont,
 				      struct device *dev)
 {
-	struct class_device *cdev = NULL;
+	struct device *cdev = NULL;
 	struct internal_container *ic;
 	struct klist_iter iter;
 
 	klist_for_each_entry(ic, &cont->containers, node, &iter) {
-		if (ic->classdev.dev == dev) {
+		if (ic->classdev.parent == dev) {
 			cdev = &ic->classdev;
 			/* FIXME: must exit iterator then break */
 			klist_iter_exit(&iter);
diff --git a/drivers/base/transport_class.c b/drivers/base/transport_class.c
index cabd0edf2156..84997efdb23d 100644
--- a/drivers/base/transport_class.c
+++ b/drivers/base/transport_class.c
@@ -66,7 +66,7 @@ EXPORT_SYMBOL_GPL(transport_class_unregister);
 
 static int anon_transport_dummy_function(struct transport_container *tc,
 					 struct device *dev,
-					 struct class_device *cdev)
+					 struct device *cdev)
 {
 	/* do nothing */
 	return 0;
@@ -115,7 +115,7 @@ EXPORT_SYMBOL_GPL(anon_transport_class_unregister);
 
 static int transport_setup_classdev(struct attribute_container *cont,
 				    struct device *dev,
-				    struct class_device *classdev)
+				    struct device *classdev)
 {
 	struct transport_class *tclass = class_to_transport_class(cont->class);
 	struct transport_container *tcont = attribute_container_to_transport_container(cont);
@@ -149,7 +149,7 @@ EXPORT_SYMBOL_GPL(transport_setup_device);
 
 static int transport_add_class_device(struct attribute_container *cont,
 				      struct device *dev,
-				      struct class_device *classdev)
+				      struct device *classdev)
 {
 	int error = attribute_container_add_class_device(classdev);
 	struct transport_container *tcont = 
@@ -181,7 +181,7 @@ EXPORT_SYMBOL_GPL(transport_add_device);
 
 static int transport_configure(struct attribute_container *cont,
 			       struct device *dev,
-			       struct class_device *cdev)
+			       struct device *cdev)
 {
 	struct transport_class *tclass = class_to_transport_class(cont->class);
 	struct transport_container *tcont = attribute_container_to_transport_container(cont);
@@ -212,7 +212,7 @@ EXPORT_SYMBOL_GPL(transport_configure_device);
 
 static int transport_remove_classdev(struct attribute_container *cont,
 				     struct device *dev,
-				     struct class_device *classdev)
+				     struct device *classdev)
 {
 	struct transport_container *tcont = 
 		attribute_container_to_transport_container(cont);
@@ -251,12 +251,12 @@ EXPORT_SYMBOL_GPL(transport_remove_device);
 
 static void transport_destroy_classdev(struct attribute_container *cont,
 				      struct device *dev,
-				      struct class_device *classdev)
+				      struct device *classdev)
 {
 	struct transport_class *tclass = class_to_transport_class(cont->class);
 
 	if (tclass->remove != anon_transport_dummy_function)
-		class_device_put(classdev);
+		put_device(classdev);
 }
 
 
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 99a110660040..435145709dd6 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -1458,9 +1458,10 @@ static int srp_reset_host(struct scsi_cmnd *scmnd)
 	return ret;
 }
 
-static ssize_t show_id_ext(struct class_device *cdev, char *buf)
+static ssize_t show_id_ext(struct device *dev, struct device_attribute *attr,
+			   char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	if (target->state == SRP_TARGET_DEAD ||
 	    target->state == SRP_TARGET_REMOVED)
@@ -1470,9 +1471,10 @@ static ssize_t show_id_ext(struct class_device *cdev, char *buf)
 		       (unsigned long long) be64_to_cpu(target->id_ext));
 }
 
-static ssize_t show_ioc_guid(struct class_device *cdev, char *buf)
+static ssize_t show_ioc_guid(struct device *dev, struct device_attribute *attr,
+			     char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	if (target->state == SRP_TARGET_DEAD ||
 	    target->state == SRP_TARGET_REMOVED)
@@ -1482,9 +1484,10 @@ static ssize_t show_ioc_guid(struct class_device *cdev, char *buf)
 		       (unsigned long long) be64_to_cpu(target->ioc_guid));
 }
 
-static ssize_t show_service_id(struct class_device *cdev, char *buf)
+static ssize_t show_service_id(struct device *dev,
+			       struct device_attribute *attr, char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	if (target->state == SRP_TARGET_DEAD ||
 	    target->state == SRP_TARGET_REMOVED)
@@ -1494,9 +1497,10 @@ static ssize_t show_service_id(struct class_device *cdev, char *buf)
 		       (unsigned long long) be64_to_cpu(target->service_id));
 }
 
-static ssize_t show_pkey(struct class_device *cdev, char *buf)
+static ssize_t show_pkey(struct device *dev, struct device_attribute *attr,
+			 char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	if (target->state == SRP_TARGET_DEAD ||
 	    target->state == SRP_TARGET_REMOVED)
@@ -1505,9 +1509,10 @@ static ssize_t show_pkey(struct class_device *cdev, char *buf)
 	return sprintf(buf, "0x%04x\n", be16_to_cpu(target->path.pkey));
 }
 
-static ssize_t show_dgid(struct class_device *cdev, char *buf)
+static ssize_t show_dgid(struct device *dev, struct device_attribute *attr,
+			 char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	if (target->state == SRP_TARGET_DEAD ||
 	    target->state == SRP_TARGET_REMOVED)
@@ -1524,9 +1529,10 @@ static ssize_t show_dgid(struct class_device *cdev, char *buf)
 		       be16_to_cpu(((__be16 *) target->path.dgid.raw)[7]));
 }
 
-static ssize_t show_orig_dgid(struct class_device *cdev, char *buf)
+static ssize_t show_orig_dgid(struct device *dev,
+			      struct device_attribute *attr, char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	if (target->state == SRP_TARGET_DEAD ||
 	    target->state == SRP_TARGET_REMOVED)
@@ -1543,9 +1549,10 @@ static ssize_t show_orig_dgid(struct class_device *cdev, char *buf)
 		       be16_to_cpu(target->orig_dgid[7]));
 }
 
-static ssize_t show_zero_req_lim(struct class_device *cdev, char *buf)
+static ssize_t show_zero_req_lim(struct device *dev,
+				 struct device_attribute *attr, char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	if (target->state == SRP_TARGET_DEAD ||
 	    target->state == SRP_TARGET_REMOVED)
@@ -1554,40 +1561,42 @@ static ssize_t show_zero_req_lim(struct class_device *cdev, char *buf)
 	return sprintf(buf, "%d\n", target->zero_req_lim);
 }
 
-static ssize_t show_local_ib_port(struct class_device *cdev, char *buf)
+static ssize_t show_local_ib_port(struct device *dev,
+				  struct device_attribute *attr, char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	return sprintf(buf, "%d\n", target->srp_host->port);
 }
 
-static ssize_t show_local_ib_device(struct class_device *cdev, char *buf)
+static ssize_t show_local_ib_device(struct device *dev,
+				    struct device_attribute *attr, char *buf)
 {
-	struct srp_target_port *target = host_to_target(class_to_shost(cdev));
+	struct srp_target_port *target = host_to_target(class_to_shost(dev));
 
 	return sprintf(buf, "%s\n", target->srp_host->srp_dev->dev->name);
 }
 
-static CLASS_DEVICE_ATTR(id_ext,	  S_IRUGO, show_id_ext,		 NULL);
-static CLASS_DEVICE_ATTR(ioc_guid,	  S_IRUGO, show_ioc_guid,	 NULL);
-static CLASS_DEVICE_ATTR(service_id,	  S_IRUGO, show_service_id,	 NULL);
-static CLASS_DEVICE_ATTR(pkey,		  S_IRUGO, show_pkey,		 NULL);
-static CLASS_DEVICE_ATTR(dgid,		  S_IRUGO, show_dgid,		 NULL);
-static CLASS_DEVICE_ATTR(orig_dgid,	  S_IRUGO, show_orig_dgid,	 NULL);
-static CLASS_DEVICE_ATTR(zero_req_lim,	  S_IRUGO, show_zero_req_lim,	 NULL);
-static CLASS_DEVICE_ATTR(local_ib_port,   S_IRUGO, show_local_ib_port,	 NULL);
-static CLASS_DEVICE_ATTR(local_ib_device, S_IRUGO, show_local_ib_device, NULL);
-
-static struct class_device_attribute *srp_host_attrs[] = {
-	&class_device_attr_id_ext,
-	&class_device_attr_ioc_guid,
-	&class_device_attr_service_id,
-	&class_device_attr_pkey,
-	&class_device_attr_dgid,
-	&class_device_attr_orig_dgid,
-	&class_device_attr_zero_req_lim,
-	&class_device_attr_local_ib_port,
-	&class_device_attr_local_ib_device,
+static DEVICE_ATTR(id_ext,	    S_IRUGO, show_id_ext,	   NULL);
+static DEVICE_ATTR(ioc_guid,	    S_IRUGO, show_ioc_guid,	   NULL);
+static DEVICE_ATTR(service_id,	    S_IRUGO, show_service_id,	   NULL);
+static DEVICE_ATTR(pkey,	    S_IRUGO, show_pkey,		   NULL);
+static DEVICE_ATTR(dgid,	    S_IRUGO, show_dgid,		   NULL);
+static DEVICE_ATTR(orig_dgid,	    S_IRUGO, show_orig_dgid,	   NULL);
+static DEVICE_ATTR(zero_req_lim,    S_IRUGO, show_zero_req_lim,	   NULL);
+static DEVICE_ATTR(local_ib_port,   S_IRUGO, show_local_ib_port,   NULL);
+static DEVICE_ATTR(local_ib_device, S_IRUGO, show_local_ib_device, NULL);
+
+static struct device_attribute *srp_host_attrs[] = {
+	&dev_attr_id_ext,
+	&dev_attr_ioc_guid,
+	&dev_attr_service_id,
+	&dev_attr_pkey,
+	&dev_attr_dgid,
+	&dev_attr_orig_dgid,
+	&dev_attr_zero_req_lim,
+	&dev_attr_local_ib_port,
+	&dev_attr_local_ib_device,
 	NULL
 };
 
@@ -1639,17 +1648,17 @@ static int srp_add_target(struct srp_host *host, struct srp_target_port *target)
 	return 0;
 }
 
-static void srp_release_class_dev(struct class_device *class_dev)
+static void srp_release_dev(struct device *dev)
 {
 	struct srp_host *host =
-		container_of(class_dev, struct srp_host, class_dev);
+		container_of(dev, struct srp_host, dev);
 
 	complete(&host->released);
 }
 
 static struct class srp_class = {
 	.name    = "infiniband_srp",
-	.release = srp_release_class_dev
+	.dev_release = srp_release_dev
 };
 
 /*
@@ -1837,11 +1846,12 @@ out:
 	return ret;
 }
 
-static ssize_t srp_create_target(struct class_device *class_dev,
+static ssize_t srp_create_target(struct device *dev,
+				 struct device_attribute *attr,
 				 const char *buf, size_t count)
 {
 	struct srp_host *host =
-		container_of(class_dev, struct srp_host, class_dev);
+		container_of(dev, struct srp_host, dev);
 	struct Scsi_Host *target_host;
 	struct srp_target_port *target;
 	int ret;
@@ -1929,27 +1939,27 @@ err:
 	return ret;
 }
 
-static CLASS_DEVICE_ATTR(add_target, S_IWUSR, NULL, srp_create_target);
+static DEVICE_ATTR(add_target, S_IWUSR, NULL, srp_create_target);
 
-static ssize_t show_ibdev(struct class_device *class_dev, char *buf)
+static ssize_t show_ibdev(struct device *dev, struct device_attribute *attr,
+			  char *buf)
 {
-	struct srp_host *host =
-		container_of(class_dev, struct srp_host, class_dev);
+	struct srp_host *host = container_of(dev, struct srp_host, dev);
 
 	return sprintf(buf, "%s\n", host->srp_dev->dev->name);
 }
 
-static CLASS_DEVICE_ATTR(ibdev, S_IRUGO, show_ibdev, NULL);
+static DEVICE_ATTR(ibdev, S_IRUGO, show_ibdev, NULL);
 
-static ssize_t show_port(struct class_device *class_dev, char *buf)
+static ssize_t show_port(struct device *dev, struct device_attribute *attr,
+			 char *buf)
 {
-	struct srp_host *host =
-		container_of(class_dev, struct srp_host, class_dev);
+	struct srp_host *host = container_of(dev, struct srp_host, dev);
 
 	return sprintf(buf, "%d\n", host->port);
 }
 
-static CLASS_DEVICE_ATTR(port, S_IRUGO, show_port, NULL);
+static DEVICE_ATTR(port, S_IRUGO, show_port, NULL);
 
 static struct srp_host *srp_add_port(struct srp_device *device, u8 port)
 {
@@ -1965,24 +1975,24 @@ static struct srp_host *srp_add_port(struct srp_device *device, u8 port)
 	host->srp_dev = device;
 	host->port = port;
 
-	host->class_dev.class = &srp_class;
-	host->class_dev.dev   = device->dev->dma_device;
-	snprintf(host->class_dev.class_id, BUS_ID_SIZE, "srp-%s-%d",
+	host->dev.class = &srp_class;
+	host->dev.parent = device->dev->dma_device;
+	snprintf(host->dev.bus_id, BUS_ID_SIZE, "srp-%s-%d",
 		 device->dev->name, port);
 
-	if (class_device_register(&host->class_dev))
+	if (device_register(&host->dev))
 		goto free_host;
-	if (class_device_create_file(&host->class_dev, &class_device_attr_add_target))
+	if (device_create_file(&host->dev, &dev_attr_add_target))
 		goto err_class;
-	if (class_device_create_file(&host->class_dev, &class_device_attr_ibdev))
+	if (device_create_file(&host->dev, &dev_attr_ibdev))
 		goto err_class;
-	if (class_device_create_file(&host->class_dev, &class_device_attr_port))
+	if (device_create_file(&host->dev, &dev_attr_port))
 		goto err_class;
 
 	return host;
 
 err_class:
-	class_device_unregister(&host->class_dev);
+	device_unregister(&host->dev);
 
 free_host:
 	kfree(host);
@@ -2087,7 +2097,7 @@ static void srp_remove_one(struct ib_device *device)
 	srp_dev = ib_get_client_data(device, &srp_client);
 
 	list_for_each_entry_safe(host, tmp_host, &srp_dev->dev_list, list) {
-		class_device_unregister(&host->class_dev);
+		device_unregister(&host->dev);
 		/*
 		 * Wait for the sysfs entry to go away, so that no new
 		 * target ports can be created.
diff --git a/drivers/infiniband/ulp/srp/ib_srp.h b/drivers/infiniband/ulp/srp/ib_srp.h
index 67e17c336a50..63d2ae724061 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.h
+++ b/drivers/infiniband/ulp/srp/ib_srp.h
@@ -99,7 +99,7 @@ struct srp_device {
 struct srp_host {
 	struct srp_device      *srp_dev;
 	u8			port;
-	struct class_device	class_dev;
+	struct device		dev;
 	struct list_head	target_list;
 	spinlock_t		target_lock;
 	struct completion	released;
diff --git a/drivers/message/fusion/mptscsih.c b/drivers/message/fusion/mptscsih.c
index 89c63147a15d..b109bd8a4d19 100644
--- a/drivers/message/fusion/mptscsih.c
+++ b/drivers/message/fusion/mptscsih.c
@@ -3300,9 +3300,10 @@ mptscsih_synchronize_cache(MPT_SCSI_HOST *hd, VirtDevice *vdevice)
 }
 
 static ssize_t
-mptscsih_version_fw_show(struct class_device *cdev, char *buf)
+mptscsih_version_fw_show(struct device *dev, struct device_attribute *attr,
+			 char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
@@ -3312,12 +3313,13 @@ mptscsih_version_fw_show(struct class_device *cdev, char *buf)
 	    (ioc->facts.FWVersion.Word & 0x0000FF00) >> 8,
 	    ioc->facts.FWVersion.Word & 0x000000FF);
 }
-static CLASS_DEVICE_ATTR(version_fw, S_IRUGO, mptscsih_version_fw_show, NULL);
+static DEVICE_ATTR(version_fw, S_IRUGO, mptscsih_version_fw_show, NULL);
 
 static ssize_t
-mptscsih_version_bios_show(struct class_device *cdev, char *buf)
+mptscsih_version_bios_show(struct device *dev, struct device_attribute *attr,
+			   char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
@@ -3327,129 +3329,141 @@ mptscsih_version_bios_show(struct class_device *cdev, char *buf)
 	    (ioc->biosVersion & 0x0000FF00) >> 8,
 	    ioc->biosVersion & 0x000000FF);
 }
-static CLASS_DEVICE_ATTR(version_bios, S_IRUGO, mptscsih_version_bios_show, NULL);
+static DEVICE_ATTR(version_bios, S_IRUGO, mptscsih_version_bios_show, NULL);
 
 static ssize_t
-mptscsih_version_mpi_show(struct class_device *cdev, char *buf)
+mptscsih_version_mpi_show(struct device *dev, struct device_attribute *attr,
+			  char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%03x\n", ioc->facts.MsgVersion);
 }
-static CLASS_DEVICE_ATTR(version_mpi, S_IRUGO, mptscsih_version_mpi_show, NULL);
+static DEVICE_ATTR(version_mpi, S_IRUGO, mptscsih_version_mpi_show, NULL);
 
 static ssize_t
-mptscsih_version_product_show(struct class_device *cdev, char *buf)
+mptscsih_version_product_show(struct device *dev,
+			      struct device_attribute *attr,
+char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%s\n", ioc->prod_name);
 }
-static CLASS_DEVICE_ATTR(version_product, S_IRUGO,
+static DEVICE_ATTR(version_product, S_IRUGO,
     mptscsih_version_product_show, NULL);
 
 static ssize_t
-mptscsih_version_nvdata_persistent_show(struct class_device *cdev, char *buf)
+mptscsih_version_nvdata_persistent_show(struct device *dev,
+					struct device_attribute *attr,
+					char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%02xh\n",
 	    ioc->nvdata_version_persistent);
 }
-static CLASS_DEVICE_ATTR(version_nvdata_persistent, S_IRUGO,
+static DEVICE_ATTR(version_nvdata_persistent, S_IRUGO,
     mptscsih_version_nvdata_persistent_show, NULL);
 
 static ssize_t
-mptscsih_version_nvdata_default_show(struct class_device *cdev, char *buf)
+mptscsih_version_nvdata_default_show(struct device *dev,
+				     struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%02xh\n",ioc->nvdata_version_default);
 }
-static CLASS_DEVICE_ATTR(version_nvdata_default, S_IRUGO,
+static DEVICE_ATTR(version_nvdata_default, S_IRUGO,
     mptscsih_version_nvdata_default_show, NULL);
 
 static ssize_t
-mptscsih_board_name_show(struct class_device *cdev, char *buf)
+mptscsih_board_name_show(struct device *dev, struct device_attribute *attr,
+			 char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%s\n", ioc->board_name);
 }
-static CLASS_DEVICE_ATTR(board_name, S_IRUGO, mptscsih_board_name_show, NULL);
+static DEVICE_ATTR(board_name, S_IRUGO, mptscsih_board_name_show, NULL);
 
 static ssize_t
-mptscsih_board_assembly_show(struct class_device *cdev, char *buf)
+mptscsih_board_assembly_show(struct device *dev,
+			     struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%s\n", ioc->board_assembly);
 }
-static CLASS_DEVICE_ATTR(board_assembly, S_IRUGO,
+static DEVICE_ATTR(board_assembly, S_IRUGO,
     mptscsih_board_assembly_show, NULL);
 
 static ssize_t
-mptscsih_board_tracer_show(struct class_device *cdev, char *buf)
+mptscsih_board_tracer_show(struct device *dev, struct device_attribute *attr,
+			   char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%s\n", ioc->board_tracer);
 }
-static CLASS_DEVICE_ATTR(board_tracer, S_IRUGO,
+static DEVICE_ATTR(board_tracer, S_IRUGO,
     mptscsih_board_tracer_show, NULL);
 
 static ssize_t
-mptscsih_io_delay_show(struct class_device *cdev, char *buf)
+mptscsih_io_delay_show(struct device *dev, struct device_attribute *attr,
+		       char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%02d\n", ioc->io_missing_delay);
 }
-static CLASS_DEVICE_ATTR(io_delay, S_IRUGO,
+static DEVICE_ATTR(io_delay, S_IRUGO,
     mptscsih_io_delay_show, NULL);
 
 static ssize_t
-mptscsih_device_delay_show(struct class_device *cdev, char *buf)
+mptscsih_device_delay_show(struct device *dev, struct device_attribute *attr,
+			   char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%02d\n", ioc->device_missing_delay);
 }
-static CLASS_DEVICE_ATTR(device_delay, S_IRUGO,
+static DEVICE_ATTR(device_delay, S_IRUGO,
     mptscsih_device_delay_show, NULL);
 
 static ssize_t
-mptscsih_debug_level_show(struct class_device *cdev, char *buf)
+mptscsih_debug_level_show(struct device *dev, struct device_attribute *attr,
+			  char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 
 	return snprintf(buf, PAGE_SIZE, "%08xh\n", ioc->debug_level);
 }
 static ssize_t
-mptscsih_debug_level_store(struct class_device *cdev, const char *buf,
-								size_t count)
+mptscsih_debug_level_store(struct device *dev, struct device_attribute *attr,
+			   const char *buf, size_t count)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	MPT_SCSI_HOST	*hd = shost_priv(host);
 	MPT_ADAPTER *ioc = hd->ioc;
 	int val = 0;
@@ -3462,22 +3476,22 @@ mptscsih_debug_level_store(struct class_device *cdev, const char *buf,
 				ioc->name, ioc->debug_level);
 	return strlen(buf);
 }
-static CLASS_DEVICE_ATTR(debug_level, S_IRUGO | S_IWUSR,
-    mptscsih_debug_level_show, mptscsih_debug_level_store);
-
-struct class_device_attribute *mptscsih_host_attrs[] = {
-	&class_device_attr_version_fw,
-	&class_device_attr_version_bios,
-	&class_device_attr_version_mpi,
-	&class_device_attr_version_product,
-	&class_device_attr_version_nvdata_persistent,
-	&class_device_attr_version_nvdata_default,
-	&class_device_attr_board_name,
-	&class_device_attr_board_assembly,
-	&class_device_attr_board_tracer,
-	&class_device_attr_io_delay,
-	&class_device_attr_device_delay,
-	&class_device_attr_debug_level,
+static DEVICE_ATTR(debug_level, S_IRUGO | S_IWUSR,
+	mptscsih_debug_level_show, mptscsih_debug_level_store);
+
+struct device_attribute *mptscsih_host_attrs[] = {
+	&dev_attr_version_fw,
+	&dev_attr_version_bios,
+	&dev_attr_version_mpi,
+	&dev_attr_version_product,
+	&dev_attr_version_nvdata_persistent,
+	&dev_attr_version_nvdata_default,
+	&dev_attr_board_name,
+	&dev_attr_board_assembly,
+	&dev_attr_board_tracer,
+	&dev_attr_io_delay,
+	&dev_attr_device_delay,
+	&dev_attr_debug_level,
 	NULL,
 };
 EXPORT_SYMBOL(mptscsih_host_attrs);
diff --git a/drivers/message/fusion/mptscsih.h b/drivers/message/fusion/mptscsih.h
index d289e97cfe8b..7ea7da0e090c 100644
--- a/drivers/message/fusion/mptscsih.h
+++ b/drivers/message/fusion/mptscsih.h
@@ -129,4 +129,4 @@ extern void mptscsih_timer_expired(unsigned long data);
 extern int mptscsih_TMHandler(MPT_SCSI_HOST *hd, u8 type, u8 channel, u8 id, int lun, int ctx2abort, ulong timeout);
 extern u8 mptscsih_raid_id_to_num(MPT_ADAPTER *ioc, u8 channel, u8 id);
 extern int mptscsih_is_phys_disk(MPT_ADAPTER *ioc, u8 channel, u8 id);
-extern struct class_device_attribute *mptscsih_host_attrs[];
+extern struct device_attribute *mptscsih_host_attrs[];
diff --git a/drivers/misc/enclosure.c b/drivers/misc/enclosure.c
index 6fcb0e96adf4..fafb57fed761 100644
--- a/drivers/misc/enclosure.c
+++ b/drivers/misc/enclosure.c
@@ -40,16 +40,16 @@ static struct class enclosure_component_class;
  * Looks through the list of registered enclosures to see
  * if it can find a match for a device.  Returns NULL if no
  * enclosure is found. Obtains a reference to the enclosure class
- * device which must be released with class_device_put().
+ * device which must be released with device_put().
  */
 struct enclosure_device *enclosure_find(struct device *dev)
 {
-	struct enclosure_device *edev = NULL;
+	struct enclosure_device *edev;
 
 	mutex_lock(&container_list_lock);
 	list_for_each_entry(edev, &container_list, node) {
-		if (edev->cdev.dev == dev) {
-			class_device_get(&edev->cdev);
+		if (edev->edev.parent == dev) {
+			get_device(&edev->edev);
 			mutex_unlock(&container_list_lock);
 			return edev;
 		}
@@ -117,11 +117,11 @@ enclosure_register(struct device *dev, const char *name, int components,
 
 	edev->components = components;
 
-	edev->cdev.class = &enclosure_class;
-	edev->cdev.dev = get_device(dev);
+	edev->edev.class = &enclosure_class;
+	edev->edev.parent = get_device(dev);
 	edev->cb = cb;
-	snprintf(edev->cdev.class_id, BUS_ID_SIZE, "%s", name);
-	err = class_device_register(&edev->cdev);
+	snprintf(edev->edev.bus_id, BUS_ID_SIZE, "%s", name);
+	err = device_register(&edev->edev);
 	if (err)
 		goto err;
 
@@ -135,7 +135,7 @@ enclosure_register(struct device *dev, const char *name, int components,
 	return edev;
 
  err:
-	put_device(edev->cdev.dev);
+	put_device(edev->edev.parent);
 	kfree(edev);
 	return ERR_PTR(err);
 }
@@ -158,27 +158,28 @@ void enclosure_unregister(struct enclosure_device *edev)
 
 	for (i = 0; i < edev->components; i++)
 		if (edev->component[i].number != -1)
-			class_device_unregister(&edev->component[i].cdev);
+			device_unregister(&edev->component[i].cdev);
 
 	/* prevent any callbacks into service user */
 	edev->cb = &enclosure_null_callbacks;
-	class_device_unregister(&edev->cdev);
+	device_unregister(&edev->edev);
 }
 EXPORT_SYMBOL_GPL(enclosure_unregister);
 
-static void enclosure_release(struct class_device *cdev)
+static void enclosure_release(struct device *cdev)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev);
 
-	put_device(cdev->dev);
+	put_device(cdev->parent);
 	kfree(edev);
 }
 
-static void enclosure_component_release(struct class_device *cdev)
+static void enclosure_component_release(struct device *dev)
 {
-	if (cdev->dev)
-		put_device(cdev->dev);
-	class_device_put(cdev->parent);
+	struct enclosure_component *cdev = to_enclosure_component(dev);
+
+	put_device(cdev->dev);
+	put_device(dev->parent);
 }
 
 /**
@@ -201,7 +202,7 @@ enclosure_component_register(struct enclosure_device *edev,
 			     const char *name)
 {
 	struct enclosure_component *ecomp;
-	struct class_device *cdev;
+	struct device *cdev;
 	int err;
 
 	if (number >= edev->components)
@@ -215,14 +216,14 @@ enclosure_component_register(struct enclosure_device *edev,
 	ecomp->type = type;
 	ecomp->number = number;
 	cdev = &ecomp->cdev;
-	cdev->parent = class_device_get(&edev->cdev);
+	cdev->parent = get_device(&edev->edev);
 	cdev->class = &enclosure_component_class;
 	if (name)
-		snprintf(cdev->class_id, BUS_ID_SIZE, "%s", name);
+		snprintf(cdev->bus_id, BUS_ID_SIZE, "%s", name);
 	else
-		snprintf(cdev->class_id, BUS_ID_SIZE, "%u", number);
+		snprintf(cdev->bus_id, BUS_ID_SIZE, "%u", number);
 
-	err = class_device_register(cdev);
+	err = device_register(cdev);
 	if (err)
 		ERR_PTR(err);
 
@@ -247,18 +248,17 @@ EXPORT_SYMBOL_GPL(enclosure_component_register);
 int enclosure_add_device(struct enclosure_device *edev, int component,
 			 struct device *dev)
 {
-	struct class_device *cdev;
+	struct enclosure_component *cdev;
 
 	if (!edev || component >= edev->components)
 		return -EINVAL;
 
-	cdev = &edev->component[component].cdev;
+	cdev = &edev->component[component];
 
-	class_device_del(cdev);
-	if (cdev->dev)
-		put_device(cdev->dev);
+	device_del(&cdev->cdev);
+	put_device(cdev->dev);
 	cdev->dev = get_device(dev);
-	return class_device_add(cdev);
+	return device_add(&cdev->cdev);
 }
 EXPORT_SYMBOL_GPL(enclosure_add_device);
 
@@ -272,18 +272,17 @@ EXPORT_SYMBOL_GPL(enclosure_add_device);
  */
 int enclosure_remove_device(struct enclosure_device *edev, int component)
 {
-	struct class_device *cdev;
+	struct enclosure_component *cdev;
 
 	if (!edev || component >= edev->components)
 		return -EINVAL;
 
-	cdev = &edev->component[component].cdev;
+	cdev = &edev->component[component];
 
-	class_device_del(cdev);
-	if (cdev->dev)
-		put_device(cdev->dev);
+	device_del(&cdev->cdev);
+	put_device(cdev->dev);
 	cdev->dev = NULL;
-	return class_device_add(cdev);
+	return device_add(&cdev->cdev);
 }
 EXPORT_SYMBOL_GPL(enclosure_remove_device);
 
@@ -291,14 +290,16 @@ EXPORT_SYMBOL_GPL(enclosure_remove_device);
  * sysfs pieces below
  */
 
-static ssize_t enclosure_show_components(struct class_device *cdev, char *buf)
+static ssize_t enclosure_show_components(struct device *cdev,
+					 struct device_attribute *attr,
+					 char *buf)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev);
 
 	return snprintf(buf, 40, "%d\n", edev->components);
 }
 
-static struct class_device_attribute enclosure_attrs[] = {
+static struct device_attribute enclosure_attrs[] = {
 	__ATTR(components, S_IRUGO, enclosure_show_components, NULL),
 	__ATTR_NULL
 };
@@ -306,8 +307,8 @@ static struct class_device_attribute enclosure_attrs[] = {
 static struct class enclosure_class = {
 	.name			= "enclosure",
 	.owner			= THIS_MODULE,
-	.release		= enclosure_release,
-	.class_dev_attrs	= enclosure_attrs,
+	.dev_release		= enclosure_release,
+	.dev_attrs		= enclosure_attrs,
 };
 
 static const char *const enclosure_status [] = {
@@ -326,7 +327,8 @@ static const char *const enclosure_type [] = {
 	[ENCLOSURE_COMPONENT_ARRAY_DEVICE] = "array device",
 };
 
-static ssize_t get_component_fault(struct class_device *cdev, char *buf)
+static ssize_t get_component_fault(struct device *cdev,
+				   struct device_attribute *attr, char *buf)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -336,8 +338,9 @@ static ssize_t get_component_fault(struct class_device *cdev, char *buf)
 	return snprintf(buf, 40, "%d\n", ecomp->fault);
 }
 
-static ssize_t set_component_fault(struct class_device *cdev, const char *buf,
-				   size_t count)
+static ssize_t set_component_fault(struct device *cdev,
+				   struct device_attribute *attr,
+				   const char *buf, size_t count)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -348,7 +351,8 @@ static ssize_t set_component_fault(struct class_device *cdev, const char *buf,
 	return count;
 }
 
-static ssize_t get_component_status(struct class_device *cdev, char *buf)
+static ssize_t get_component_status(struct device *cdev,
+				    struct device_attribute *attr,char *buf)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -358,8 +362,9 @@ static ssize_t get_component_status(struct class_device *cdev, char *buf)
 	return snprintf(buf, 40, "%s\n", enclosure_status[ecomp->status]);
 }
 
-static ssize_t set_component_status(struct class_device *cdev, const char *buf,
-				   size_t count)
+static ssize_t set_component_status(struct device *cdev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t count)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -380,7 +385,8 @@ static ssize_t set_component_status(struct class_device *cdev, const char *buf,
 		return -EINVAL;
 }
 
-static ssize_t get_component_active(struct class_device *cdev, char *buf)
+static ssize_t get_component_active(struct device *cdev,
+				    struct device_attribute *attr, char *buf)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -390,8 +396,9 @@ static ssize_t get_component_active(struct class_device *cdev, char *buf)
 	return snprintf(buf, 40, "%d\n", ecomp->active);
 }
 
-static ssize_t set_component_active(struct class_device *cdev, const char *buf,
-				   size_t count)
+static ssize_t set_component_active(struct device *cdev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t count)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -402,7 +409,8 @@ static ssize_t set_component_active(struct class_device *cdev, const char *buf,
 	return count;
 }
 
-static ssize_t get_component_locate(struct class_device *cdev, char *buf)
+static ssize_t get_component_locate(struct device *cdev,
+				    struct device_attribute *attr, char *buf)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -412,8 +420,9 @@ static ssize_t get_component_locate(struct class_device *cdev, char *buf)
 	return snprintf(buf, 40, "%d\n", ecomp->locate);
 }
 
-static ssize_t set_component_locate(struct class_device *cdev, const char *buf,
-				   size_t count)
+static ssize_t set_component_locate(struct device *cdev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t count)
 {
 	struct enclosure_device *edev = to_enclosure_device(cdev->parent);
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
@@ -424,7 +433,8 @@ static ssize_t set_component_locate(struct class_device *cdev, const char *buf,
 	return count;
 }
 
-static ssize_t get_component_type(struct class_device *cdev, char *buf)
+static ssize_t get_component_type(struct device *cdev,
+				  struct device_attribute *attr, char *buf)
 {
 	struct enclosure_component *ecomp = to_enclosure_component(cdev);
 
@@ -432,7 +442,7 @@ static ssize_t get_component_type(struct class_device *cdev, char *buf)
 }
 
 
-static struct class_device_attribute enclosure_component_attrs[] = {
+static struct device_attribute enclosure_component_attrs[] = {
 	__ATTR(fault, S_IRUGO | S_IWUSR, get_component_fault,
 	       set_component_fault),
 	__ATTR(status, S_IRUGO | S_IWUSR, get_component_status,
@@ -448,8 +458,8 @@ static struct class_device_attribute enclosure_component_attrs[] = {
 static struct class enclosure_component_class =  {
 	.name			= "enclosure_component",
 	.owner			= THIS_MODULE,
-	.class_dev_attrs	= enclosure_component_attrs,
-	.release		= enclosure_component_release,
+	.dev_attrs	= enclosure_component_attrs,
+	.dev_release		= enclosure_component_release,
 };
 
 static int __init enclosure_init(void)
diff --git a/drivers/scsi/3w-9xxx.c b/drivers/scsi/3w-9xxx.c
index 51c3ebf1c7d1..b31faeccb9cd 100644
--- a/drivers/scsi/3w-9xxx.c
+++ b/drivers/scsi/3w-9xxx.c
@@ -140,9 +140,10 @@ static void twa_unmap_scsi_data(TW_Device_Extension *tw_dev, int request_id);
 /* Functions */
 
 /* Show some statistics about the card */
-static ssize_t twa_show_stats(struct class_device *class_dev, char *buf)
+static ssize_t twa_show_stats(struct device *dev,
+			      struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(class_dev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	TW_Device_Extension *tw_dev = (TW_Device_Extension *)host->hostdata;
 	unsigned long flags = 0;
 	ssize_t len;
@@ -184,7 +185,7 @@ static int twa_change_queue_depth(struct scsi_device *sdev, int queue_depth)
 } /* End twa_change_queue_depth() */
 
 /* Create sysfs 'stats' entry */
-static struct class_device_attribute twa_host_stats_attr = {
+static struct device_attribute twa_host_stats_attr = {
 	.attr = {
 		.name = 	"stats",
 		.mode =		S_IRUGO,
@@ -193,7 +194,7 @@ static struct class_device_attribute twa_host_stats_attr = {
 };
 
 /* Host attributes initializer */
-static struct class_device_attribute *twa_host_attrs[] = {
+static struct device_attribute *twa_host_attrs[] = {
 	&twa_host_stats_attr,
 	NULL,
 };
diff --git a/drivers/scsi/3w-xxxx.c b/drivers/scsi/3w-xxxx.c
index adb98a297210..8c22329aa85e 100644
--- a/drivers/scsi/3w-xxxx.c
+++ b/drivers/scsi/3w-xxxx.c
@@ -484,9 +484,10 @@ static void tw_state_request_start(TW_Device_Extension *tw_dev, int *request_id)
 } /* End tw_state_request_start() */
 
 /* Show some statistics about the card */
-static ssize_t tw_show_stats(struct class_device *class_dev, char *buf)
+static ssize_t tw_show_stats(struct device *dev, struct device_attribute *attr,
+			     char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(class_dev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	TW_Device_Extension *tw_dev = (TW_Device_Extension *)host->hostdata;
 	unsigned long flags = 0;
 	ssize_t len;
@@ -528,7 +529,7 @@ static int tw_change_queue_depth(struct scsi_device *sdev, int queue_depth)
 } /* End tw_change_queue_depth() */
 
 /* Create sysfs 'stats' entry */
-static struct class_device_attribute tw_host_stats_attr = {
+static struct device_attribute tw_host_stats_attr = {
 	.attr = {
 		.name = 	"stats",
 		.mode =		S_IRUGO,
@@ -537,7 +538,7 @@ static struct class_device_attribute tw_host_stats_attr = {
 };
 
 /* Host attributes initializer */
-static struct class_device_attribute *tw_host_attrs[] = {
+static struct device_attribute *tw_host_attrs[] = {
 	&tw_host_stats_attr,
 	NULL,
 };
diff --git a/drivers/scsi/aacraid/aachba.c b/drivers/scsi/aacraid/aachba.c
index 369fcf78f396..439fd8146657 100644
--- a/drivers/scsi/aacraid/aachba.c
+++ b/drivers/scsi/aacraid/aachba.c
@@ -1316,7 +1316,7 @@ int aac_get_adapter_info(struct aac_dev* dev)
 			tmp>>24,(tmp>>16)&0xff,tmp&0xff,
 			le32_to_cpu(dev->adapter_info.biosbuild));
 		buffer[0] = '\0';
-		if (aac_show_serial_number(
+		if (aac_get_serial_number(
 		  shost_to_class(dev->scsi_host_ptr), buffer))
 			printk(KERN_INFO "%s%d: serial %s",
 			  dev->name, dev->id, buffer);
diff --git a/drivers/scsi/aacraid/aacraid.h b/drivers/scsi/aacraid/aacraid.h
index ace0b751c131..113ca9c8934c 100644
--- a/drivers/scsi/aacraid/aacraid.h
+++ b/drivers/scsi/aacraid/aacraid.h
@@ -1850,9 +1850,9 @@ int aac_get_containers(struct aac_dev *dev);
 int aac_scsi_cmd(struct scsi_cmnd *cmd);
 int aac_dev_ioctl(struct aac_dev *dev, int cmd, void __user *arg);
 #ifndef shost_to_class
-#define shost_to_class(shost) &shost->shost_classdev
+#define shost_to_class(shost) &shost->shost_dev
 #endif
-ssize_t aac_show_serial_number(struct class_device *class_dev, char *buf);
+ssize_t aac_get_serial_number(struct device *dev, char *buf);
 int aac_do_ioctl(struct aac_dev * dev, int cmd, void __user *arg);
 int aac_rx_init(struct aac_dev *dev);
 int aac_rkt_init(struct aac_dev *dev);
diff --git a/drivers/scsi/aacraid/linit.c b/drivers/scsi/aacraid/linit.c
index ae5f74fb62d5..c6391838b2cd 100644
--- a/drivers/scsi/aacraid/linit.c
+++ b/drivers/scsi/aacraid/linit.c
@@ -755,10 +755,10 @@ static long aac_compat_cfg_ioctl(struct file *file, unsigned cmd, unsigned long
 }
 #endif
 
-static ssize_t aac_show_model(struct class_device *class_dev,
-		char *buf)
+static ssize_t aac_show_model(struct device *device,
+			      struct device_attribute *attr, char *buf)
 {
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(device)->hostdata;
 	int len;
 
 	if (dev->supplement_adapter_info.AdapterTypeText[0]) {
@@ -774,10 +774,10 @@ static ssize_t aac_show_model(struct class_device *class_dev,
 	return len;
 }
 
-static ssize_t aac_show_vendor(struct class_device *class_dev,
-		char *buf)
+static ssize_t aac_show_vendor(struct device *device,
+			       struct device_attribute *attr, char *buf)
 {
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(device)->hostdata;
 	int len;
 
 	if (dev->supplement_adapter_info.AdapterTypeText[0]) {
@@ -793,10 +793,11 @@ static ssize_t aac_show_vendor(struct class_device *class_dev,
 	return len;
 }
 
-static ssize_t aac_show_flags(struct class_device *class_dev, char *buf)
+static ssize_t aac_show_flags(struct device *cdev,
+			      struct device_attribute *attr, char *buf)
 {
 	int len = 0;
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(cdev)->hostdata;
 
 	if (nblank(dprintk(x)))
 		len = snprintf(buf, PAGE_SIZE, "dprintk\n");
@@ -812,10 +813,11 @@ static ssize_t aac_show_flags(struct class_device *class_dev, char *buf)
 	return len;
 }
 
-static ssize_t aac_show_kernel_version(struct class_device *class_dev,
-		char *buf)
+static ssize_t aac_show_kernel_version(struct device *device,
+				       struct device_attribute *attr,
+				       char *buf)
 {
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(device)->hostdata;
 	int len, tmp;
 
 	tmp = le32_to_cpu(dev->adapter_info.kernelrev);
@@ -825,10 +827,11 @@ static ssize_t aac_show_kernel_version(struct class_device *class_dev,
 	return len;
 }
 
-static ssize_t aac_show_monitor_version(struct class_device *class_dev,
-		char *buf)
+static ssize_t aac_show_monitor_version(struct device *device,
+					struct device_attribute *attr,
+					char *buf)
 {
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(device)->hostdata;
 	int len, tmp;
 
 	tmp = le32_to_cpu(dev->adapter_info.monitorrev);
@@ -838,10 +841,11 @@ static ssize_t aac_show_monitor_version(struct class_device *class_dev,
 	return len;
 }
 
-static ssize_t aac_show_bios_version(struct class_device *class_dev,
-		char *buf)
+static ssize_t aac_show_bios_version(struct device *device,
+				     struct device_attribute *attr,
+				     char *buf)
 {
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(device)->hostdata;
 	int len, tmp;
 
 	tmp = le32_to_cpu(dev->adapter_info.biosrev);
@@ -851,9 +855,10 @@ static ssize_t aac_show_bios_version(struct class_device *class_dev,
 	return len;
 }
 
-ssize_t aac_show_serial_number(struct class_device *class_dev, char *buf)
+ssize_t aac_show_serial_number(struct device *device,
+			       struct device_attribute *attr, char *buf)
 {
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(device)->hostdata;
 	int len = 0;
 
 	if (le32_to_cpu(dev->adapter_info.serial[0]) != 0xBAD0)
@@ -869,35 +874,39 @@ ssize_t aac_show_serial_number(struct class_device *class_dev, char *buf)
 	return len;
 }
 
-static ssize_t aac_show_max_channel(struct class_device *class_dev, char *buf)
+static ssize_t aac_show_max_channel(struct device *device,
+				    struct device_attribute *attr, char *buf)
 {
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-	  class_to_shost(class_dev)->max_channel);
+	  class_to_shost(device)->max_channel);
 }
 
-static ssize_t aac_show_max_id(struct class_device *class_dev, char *buf)
+static ssize_t aac_show_max_id(struct device *device,
+			       struct device_attribute *attr, char *buf)
 {
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-	  class_to_shost(class_dev)->max_id);
+	  class_to_shost(device)->max_id);
 }
 
-static ssize_t aac_store_reset_adapter(struct class_device *class_dev,
-		const char *buf, size_t count)
+static ssize_t aac_store_reset_adapter(struct device *device,
+				       struct device_attribute *attr,
+				       const char *buf, size_t count)
 {
 	int retval = -EACCES;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return retval;
-	retval = aac_reset_adapter((struct aac_dev*)class_to_shost(class_dev)->hostdata, buf[0] == '!');
+	retval = aac_reset_adapter((struct aac_dev*)class_to_shost(device)->hostdata, buf[0] == '!');
 	if (retval >= 0)
 		retval = count;
 	return retval;
 }
 
-static ssize_t aac_show_reset_adapter(struct class_device *class_dev,
-		char *buf)
+static ssize_t aac_show_reset_adapter(struct device *device,
+				      struct device_attribute *attr,
+				      char *buf)
 {
-	struct aac_dev *dev = (struct aac_dev*)class_to_shost(class_dev)->hostdata;
+	struct aac_dev *dev = (struct aac_dev*)class_to_shost(device)->hostdata;
 	int len, tmp;
 
 	tmp = aac_adapter_check_health(dev);
@@ -907,70 +916,70 @@ static ssize_t aac_show_reset_adapter(struct class_device *class_dev,
 	return len;
 }
 
-static struct class_device_attribute aac_model = {
+static struct device_attribute aac_model = {
 	.attr = {
 		.name = "model",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_model,
 };
-static struct class_device_attribute aac_vendor = {
+static struct device_attribute aac_vendor = {
 	.attr = {
 		.name = "vendor",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_vendor,
 };
-static struct class_device_attribute aac_flags = {
+static struct device_attribute aac_flags = {
 	.attr = {
 		.name = "flags",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_flags,
 };
-static struct class_device_attribute aac_kernel_version = {
+static struct device_attribute aac_kernel_version = {
 	.attr = {
 		.name = "hba_kernel_version",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_kernel_version,
 };
-static struct class_device_attribute aac_monitor_version = {
+static struct device_attribute aac_monitor_version = {
 	.attr = {
 		.name = "hba_monitor_version",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_monitor_version,
 };
-static struct class_device_attribute aac_bios_version = {
+static struct device_attribute aac_bios_version = {
 	.attr = {
 		.name = "hba_bios_version",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_bios_version,
 };
-static struct class_device_attribute aac_serial_number = {
+static struct device_attribute aac_serial_number = {
 	.attr = {
 		.name = "serial_number",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_serial_number,
 };
-static struct class_device_attribute aac_max_channel = {
+static struct device_attribute aac_max_channel = {
 	.attr = {
 		.name = "max_channel",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_max_channel,
 };
-static struct class_device_attribute aac_max_id = {
+static struct device_attribute aac_max_id = {
 	.attr = {
 		.name = "max_id",
 		.mode = S_IRUGO,
 	},
 	.show = aac_show_max_id,
 };
-static struct class_device_attribute aac_reset = {
+static struct device_attribute aac_reset = {
 	.attr = {
 		.name = "reset_host",
 		.mode = S_IWUSR|S_IRUGO,
@@ -979,7 +988,7 @@ static struct class_device_attribute aac_reset = {
 	.show = aac_show_reset_adapter,
 };
 
-static struct class_device_attribute *aac_attrs[] = {
+static struct device_attribute *aac_attrs[] = {
 	&aac_model,
 	&aac_vendor,
 	&aac_flags,
@@ -993,6 +1002,10 @@ static struct class_device_attribute *aac_attrs[] = {
 	NULL
 };
 
+ssize_t aac_get_serial_number(struct device *device, char *buf)
+{
+	return aac_show_serial_number(device, &aac_serial_number, buf);
+}
 
 static const struct file_operations aac_cfg_fops = {
 	.owner		= THIS_MODULE,
diff --git a/drivers/scsi/arcmsr/arcmsr.h b/drivers/scsi/arcmsr/arcmsr.h
index 3288be2e49f8..ab646e580d64 100644
--- a/drivers/scsi/arcmsr/arcmsr.h
+++ b/drivers/scsi/arcmsr/arcmsr.h
@@ -44,7 +44,7 @@
 */
 #include <linux/interrupt.h>
 
-struct class_device_attribute;
+struct device_attribute;
 /*The limit of outstanding scsi command that firmware can handle*/
 #define ARCMSR_MAX_OUTSTANDING_CMD						256
 #define ARCMSR_MAX_FREECCB_NUM							320
@@ -556,6 +556,6 @@ struct SENSE_DATA
 extern void arcmsr_post_ioctldata2iop(struct AdapterControlBlock *);
 extern void arcmsr_iop_message_read(struct AdapterControlBlock *);
 extern struct QBUFFER __iomem *arcmsr_get_iop_rqbuffer(struct AdapterControlBlock *);
-extern struct class_device_attribute *arcmsr_host_attrs[];
+extern struct device_attribute *arcmsr_host_attrs[];
 extern int arcmsr_alloc_sysfs_attr(struct AdapterControlBlock *);
 void arcmsr_free_sysfs_attr(struct AdapterControlBlock *acb);
diff --git a/drivers/scsi/arcmsr/arcmsr_attr.c b/drivers/scsi/arcmsr/arcmsr_attr.c
index 7d7b0a554276..69f8346aa288 100644
--- a/drivers/scsi/arcmsr/arcmsr_attr.c
+++ b/drivers/scsi/arcmsr/arcmsr_attr.c
@@ -57,15 +57,15 @@
 #include <scsi/scsi_transport.h>
 #include "arcmsr.h"
 
-struct class_device_attribute *arcmsr_host_attrs[];
+struct device_attribute *arcmsr_host_attrs[];
 
 static ssize_t arcmsr_sysfs_iop_message_read(struct kobject *kobj,
 					     struct bin_attribute *bin,
 					     char *buf, loff_t off,
 					     size_t count)
 {
-	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct device *dev = container_of(kobj,struct device,kobj);
+	struct Scsi_Host *host = class_to_shost(dev);
 	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
 	uint8_t *pQbuffer,*ptmpQbuffer;
 	int32_t allxfer_len = 0;
@@ -110,8 +110,8 @@ static ssize_t arcmsr_sysfs_iop_message_write(struct kobject *kobj,
 					      char *buf, loff_t off,
 					      size_t count)
 {
-	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct device *dev = container_of(kobj,struct device,kobj);
+	struct Scsi_Host *host = class_to_shost(dev);
 	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
 	int32_t my_empty_len, user_len, wqbuf_firstindex, wqbuf_lastindex;
 	uint8_t *pQbuffer, *ptmpuserbuffer;
@@ -158,8 +158,8 @@ static ssize_t arcmsr_sysfs_iop_message_clear(struct kobject *kobj,
 					      char *buf, loff_t off,
 					      size_t count)
 {
-	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct device *dev = container_of(kobj,struct device,kobj);
+	struct Scsi_Host *host = class_to_shost(dev);
 	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
 	uint8_t *pQbuffer;
 
@@ -220,87 +220,104 @@ int arcmsr_alloc_sysfs_attr(struct AdapterControlBlock *acb)
 	struct Scsi_Host *host = acb->host;
 	int error;
 
-	error = sysfs_create_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_read_attr);
+	error = sysfs_create_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_read_attr);
 	if (error) {
 		printk(KERN_ERR "arcmsr: alloc sysfs mu_read failed\n");
 		goto error_bin_file_message_read;
 	}
-	error = sysfs_create_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_write_attr);
+	error = sysfs_create_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_write_attr);
 	if (error) {
 		printk(KERN_ERR "arcmsr: alloc sysfs mu_write failed\n");
 		goto error_bin_file_message_write;
 	}
-	error = sysfs_create_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_clear_attr);
+	error = sysfs_create_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_clear_attr);
 	if (error) {
 		printk(KERN_ERR "arcmsr: alloc sysfs mu_clear failed\n");
 		goto error_bin_file_message_clear;
 	}
 	return 0;
 error_bin_file_message_clear:
-	sysfs_remove_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_write_attr);
+	sysfs_remove_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_write_attr);
 error_bin_file_message_write:
-	sysfs_remove_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_read_attr);
+	sysfs_remove_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_read_attr);
 error_bin_file_message_read:
 	return error;
 }
 
-void
-arcmsr_free_sysfs_attr(struct AdapterControlBlock *acb) {
+void arcmsr_free_sysfs_attr(struct AdapterControlBlock *acb)
+{
 	struct Scsi_Host *host = acb->host;
 
-	sysfs_remove_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_clear_attr);
-	sysfs_remove_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_write_attr);
-	sysfs_remove_bin_file(&host->shost_classdev.kobj, &arcmsr_sysfs_message_read_attr);
+	sysfs_remove_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_clear_attr);
+	sysfs_remove_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_write_attr);
+	sysfs_remove_bin_file(&host->shost_dev.kobj, &arcmsr_sysfs_message_read_attr);
 }
 
 
 static ssize_t
-arcmsr_attr_host_driver_version(struct class_device *cdev, char *buf) {
+arcmsr_attr_host_driver_version(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
 	return snprintf(buf, PAGE_SIZE,
 			"%s\n",
 			ARCMSR_DRIVER_VERSION);
 }
 
 static ssize_t
-arcmsr_attr_host_driver_posted_cmd(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_driver_posted_cmd(struct device *dev,
+				   struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 	return snprintf(buf, PAGE_SIZE,
 			"%4d\n",
 			atomic_read(&acb->ccboutstandingcount));
 }
 
 static ssize_t
-arcmsr_attr_host_driver_reset(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_driver_reset(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 	return snprintf(buf, PAGE_SIZE,
 			"%4d\n",
 			acb->num_resets);
 }
 
 static ssize_t
-arcmsr_attr_host_driver_abort(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_driver_abort(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 	return snprintf(buf, PAGE_SIZE,
 			"%4d\n",
 			acb->num_aborts);
 }
 
 static ssize_t
-arcmsr_attr_host_fw_model(struct class_device *cdev, char *buf) {
-    struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_fw_model(struct device *dev, struct device_attribute *attr,
+			  char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 	return snprintf(buf, PAGE_SIZE,
 			"%s\n",
 			acb->firm_model);
 }
 
 static ssize_t
-arcmsr_attr_host_fw_version(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_fw_version(struct device *dev,
+			    struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+			(struct AdapterControlBlock *) host->hostdata;
 
 	return snprintf(buf, PAGE_SIZE,
 			"%s\n",
@@ -308,9 +325,12 @@ arcmsr_attr_host_fw_version(struct class_device *cdev, char *buf) {
 }
 
 static ssize_t
-arcmsr_attr_host_fw_request_len(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_fw_request_len(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 
 	return snprintf(buf, PAGE_SIZE,
 			"%4d\n",
@@ -318,9 +338,12 @@ arcmsr_attr_host_fw_request_len(struct class_device *cdev, char *buf) {
 }
 
 static ssize_t
-arcmsr_attr_host_fw_numbers_queue(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_fw_numbers_queue(struct device *dev,
+				  struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 
 	return snprintf(buf, PAGE_SIZE,
 			"%4d\n",
@@ -328,9 +351,12 @@ arcmsr_attr_host_fw_numbers_queue(struct class_device *cdev, char *buf) {
 }
 
 static ssize_t
-arcmsr_attr_host_fw_sdram_size(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_fw_sdram_size(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 
 	return snprintf(buf, PAGE_SIZE,
 			"%4d\n",
@@ -338,36 +364,39 @@ arcmsr_attr_host_fw_sdram_size(struct class_device *cdev, char *buf) {
 }
 
 static ssize_t
-arcmsr_attr_host_fw_hd_channels(struct class_device *cdev, char *buf) {
-	struct Scsi_Host *host = class_to_shost(cdev);
-	struct AdapterControlBlock *acb = (struct AdapterControlBlock *) host->hostdata;
+arcmsr_attr_host_fw_hd_channels(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	struct Scsi_Host *host = class_to_shost(dev);
+	struct AdapterControlBlock *acb =
+		(struct AdapterControlBlock *) host->hostdata;
 
 	return snprintf(buf, PAGE_SIZE,
 			"%4d\n",
 			acb->firm_hd_channels);
 }
 
-static CLASS_DEVICE_ATTR(host_driver_version, S_IRUGO, arcmsr_attr_host_driver_version, NULL);
-static CLASS_DEVICE_ATTR(host_driver_posted_cmd, S_IRUGO, arcmsr_attr_host_driver_posted_cmd, NULL);
-static CLASS_DEVICE_ATTR(host_driver_reset, S_IRUGO, arcmsr_attr_host_driver_reset, NULL);
-static CLASS_DEVICE_ATTR(host_driver_abort, S_IRUGO, arcmsr_attr_host_driver_abort, NULL);
-static CLASS_DEVICE_ATTR(host_fw_model, S_IRUGO, arcmsr_attr_host_fw_model, NULL);
-static CLASS_DEVICE_ATTR(host_fw_version, S_IRUGO, arcmsr_attr_host_fw_version, NULL);
-static CLASS_DEVICE_ATTR(host_fw_request_len, S_IRUGO, arcmsr_attr_host_fw_request_len, NULL);
-static CLASS_DEVICE_ATTR(host_fw_numbers_queue, S_IRUGO, arcmsr_attr_host_fw_numbers_queue, NULL);
-static CLASS_DEVICE_ATTR(host_fw_sdram_size, S_IRUGO, arcmsr_attr_host_fw_sdram_size, NULL);
-static CLASS_DEVICE_ATTR(host_fw_hd_channels, S_IRUGO, arcmsr_attr_host_fw_hd_channels, NULL);
-
-struct class_device_attribute *arcmsr_host_attrs[] = {
-	&class_device_attr_host_driver_version,
-	&class_device_attr_host_driver_posted_cmd,
-	&class_device_attr_host_driver_reset,
-	&class_device_attr_host_driver_abort,
-	&class_device_attr_host_fw_model,
-	&class_device_attr_host_fw_version,
-	&class_device_attr_host_fw_request_len,
-	&class_device_attr_host_fw_numbers_queue,
-	&class_device_attr_host_fw_sdram_size,
-	&class_device_attr_host_fw_hd_channels,
+static DEVICE_ATTR(host_driver_version, S_IRUGO, arcmsr_attr_host_driver_version, NULL);
+static DEVICE_ATTR(host_driver_posted_cmd, S_IRUGO, arcmsr_attr_host_driver_posted_cmd, NULL);
+static DEVICE_ATTR(host_driver_reset, S_IRUGO, arcmsr_attr_host_driver_reset, NULL);
+static DEVICE_ATTR(host_driver_abort, S_IRUGO, arcmsr_attr_host_driver_abort, NULL);
+static DEVICE_ATTR(host_fw_model, S_IRUGO, arcmsr_attr_host_fw_model, NULL);
+static DEVICE_ATTR(host_fw_version, S_IRUGO, arcmsr_attr_host_fw_version, NULL);
+static DEVICE_ATTR(host_fw_request_len, S_IRUGO, arcmsr_attr_host_fw_request_len, NULL);
+static DEVICE_ATTR(host_fw_numbers_queue, S_IRUGO, arcmsr_attr_host_fw_numbers_queue, NULL);
+static DEVICE_ATTR(host_fw_sdram_size, S_IRUGO, arcmsr_attr_host_fw_sdram_size, NULL);
+static DEVICE_ATTR(host_fw_hd_channels, S_IRUGO, arcmsr_attr_host_fw_hd_channels, NULL);
+
+struct device_attribute *arcmsr_host_attrs[] = {
+	&dev_attr_host_driver_version,
+	&dev_attr_host_driver_posted_cmd,
+	&dev_attr_host_driver_reset,
+	&dev_attr_host_driver_abort,
+	&dev_attr_host_fw_model,
+	&dev_attr_host_fw_version,
+	&dev_attr_host_fw_request_len,
+	&dev_attr_host_fw_numbers_queue,
+	&dev_attr_host_fw_sdram_size,
+	&dev_attr_host_fw_hd_channels,
 	NULL,
 };
diff --git a/drivers/scsi/ch.c b/drivers/scsi/ch.c
index 92d1cb1b21cb..75c84d7b9ce8 100644
--- a/drivers/scsi/ch.c
+++ b/drivers/scsi/ch.c
@@ -881,7 +881,7 @@ static long ch_ioctl_compat(struct file * file,
 static int ch_probe(struct device *dev)
 {
 	struct scsi_device *sd = to_scsi_device(dev);
-	struct class_device *class_dev;
+	struct device *class_dev;
 	int minor, ret = -ENOMEM;
 	scsi_changer *ch;
 
@@ -910,11 +910,11 @@ static int ch_probe(struct device *dev)
 	ch->minor = minor;
 	sprintf(ch->name,"ch%d",ch->minor);
 
-	class_dev = class_device_create(ch_sysfs_class, NULL,
-					MKDEV(SCSI_CHANGER_MAJOR, ch->minor),
-					dev, "s%s", ch->name);
+	class_dev = device_create(ch_sysfs_class, dev,
+				  MKDEV(SCSI_CHANGER_MAJOR,ch->minor),
+				  "s%s", ch->name);
 	if (IS_ERR(class_dev)) {
-		printk(KERN_WARNING "ch%d: class_device_create failed\n",
+		printk(KERN_WARNING "ch%d: device_create failed\n",
 		       ch->minor);
 		ret = PTR_ERR(class_dev);
 		goto remove_idr;
@@ -945,8 +945,7 @@ static int ch_remove(struct device *dev)
 	idr_remove(&ch_index_idr, ch->minor);
 	spin_unlock(&ch_index_lock);
 
-	class_device_destroy(ch_sysfs_class,
-			     MKDEV(SCSI_CHANGER_MAJOR,ch->minor));
+	device_destroy(ch_sysfs_class, MKDEV(SCSI_CHANGER_MAJOR,ch->minor));
 	kfree(ch->dt);
 	kfree(ch);
 	return 0;
diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c
index 1592640a87b5..c264a8c5f01e 100644
--- a/drivers/scsi/hosts.c
+++ b/drivers/scsi/hosts.c
@@ -43,14 +43,14 @@
 static int scsi_host_next_hn;		/* host_no for next new host */
 
 
-static void scsi_host_cls_release(struct class_device *class_dev)
+static void scsi_host_cls_release(struct device *dev)
 {
-	put_device(&class_to_shost(class_dev)->shost_gendev);
+	put_device(&class_to_shost(dev)->shost_gendev);
 }
 
 static struct class shost_class = {
 	.name		= "scsi_host",
-	.release	= scsi_host_cls_release,
+	.dev_release	= scsi_host_cls_release,
 };
 
 /**
@@ -174,7 +174,7 @@ void scsi_remove_host(struct Scsi_Host *shost)
 	spin_unlock_irqrestore(shost->host_lock, flags);
 
 	transport_unregister_device(&shost->shost_gendev);
-	class_device_unregister(&shost->shost_classdev);
+	device_unregister(&shost->shost_dev);
 	device_del(&shost->shost_gendev);
 	scsi_proc_hostdir_rm(shost->hostt);
 }
@@ -212,7 +212,7 @@ int scsi_add_host(struct Scsi_Host *shost, struct device *dev)
 	scsi_host_set_state(shost, SHOST_RUNNING);
 	get_device(shost->shost_gendev.parent);
 
-	error = class_device_add(&shost->shost_classdev);
+	error = device_add(&shost->shost_dev);
 	if (error)
 		goto out_del_gendev;
 
@@ -223,7 +223,7 @@ int scsi_add_host(struct Scsi_Host *shost, struct device *dev)
 					 GFP_KERNEL);
 		if (shost->shost_data == NULL) {
 			error = -ENOMEM;
-			goto out_del_classdev;
+			goto out_del_dev;
 		}
 	}
 
@@ -250,8 +250,8 @@ int scsi_add_host(struct Scsi_Host *shost, struct device *dev)
 		destroy_workqueue(shost->work_q);
  out_free_shost_data:
 	kfree(shost->shost_data);
- out_del_classdev:
-	class_device_del(&shost->shost_classdev);
+ out_del_dev:
+	device_del(&shost->shost_dev);
  out_del_gendev:
 	device_del(&shost->shost_gendev);
  out:
@@ -385,11 +385,11 @@ struct Scsi_Host *scsi_host_alloc(struct scsi_host_template *sht, int privsize)
 		shost->host_no);
 	shost->shost_gendev.release = scsi_host_dev_release;
 
-	class_device_initialize(&shost->shost_classdev);
-	shost->shost_classdev.dev = &shost->shost_gendev;
-	shost->shost_classdev.class = &shost_class;
-	snprintf(shost->shost_classdev.class_id, BUS_ID_SIZE, "host%d",
-		  shost->host_no);
+	device_initialize(&shost->shost_dev);
+	shost->shost_dev.parent = &shost->shost_gendev;
+	shost->shost_dev.class = &shost_class;
+	snprintf(shost->shost_dev.bus_id, BUS_ID_SIZE, "host%d",
+		 shost->host_no);
 
 	shost->ehandler = kthread_run(scsi_error_handler, shost,
 			"scsi_eh_%d", shost->host_no);
@@ -432,12 +432,12 @@ void scsi_unregister(struct Scsi_Host *shost)
 }
 EXPORT_SYMBOL(scsi_unregister);
 
-static int __scsi_host_match(struct class_device *cdev, void *data)
+static int __scsi_host_match(struct device *dev, void *data)
 {
 	struct Scsi_Host *p;
 	unsigned short *hostnum = (unsigned short *)data;
 
-	p = class_to_shost(cdev);
+	p = class_to_shost(dev);
 	return p->host_no == *hostnum;
 }
 
@@ -450,10 +450,10 @@ static int __scsi_host_match(struct class_device *cdev, void *data)
  **/
 struct Scsi_Host *scsi_host_lookup(unsigned short hostnum)
 {
-	struct class_device *cdev;
+	struct device *cdev;
 	struct Scsi_Host *shost = ERR_PTR(-ENXIO);
 
-	cdev = class_find_child(&shost_class, &hostnum, __scsi_host_match);
+	cdev = class_find_device(&shost_class, &hostnum, __scsi_host_match);
 	if (cdev)
 		shost = scsi_host_get(class_to_shost(cdev));
 
diff --git a/drivers/scsi/hptiop.c b/drivers/scsi/hptiop.c
index beecda991682..5b7be1e9841c 100644
--- a/drivers/scsi/hptiop.c
+++ b/drivers/scsi/hptiop.c
@@ -859,14 +859,16 @@ static int hptiop_adjust_disk_queue_depth(struct scsi_device *sdev,
 	return queue_depth;
 }
 
-static ssize_t hptiop_show_version(struct class_device *class_dev, char *buf)
+static ssize_t hptiop_show_version(struct device *dev,
+				   struct device_attribute *attr, char *buf)
 {
 	return snprintf(buf, PAGE_SIZE, "%s\n", driver_ver);
 }
 
-static ssize_t hptiop_show_fw_version(struct class_device *class_dev, char *buf)
+static ssize_t hptiop_show_fw_version(struct device *dev,
+				      struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(class_dev);
+	struct Scsi_Host *host = class_to_shost(dev);
 	struct hptiop_hba *hba = (struct hptiop_hba *)host->hostdata;
 
 	return snprintf(buf, PAGE_SIZE, "%d.%d.%d.%d\n",
@@ -876,7 +878,7 @@ static ssize_t hptiop_show_fw_version(struct class_device *class_dev, char *buf)
 				hba->firmware_version & 0xff);
 }
 
-static struct class_device_attribute hptiop_attr_version = {
+static struct device_attribute hptiop_attr_version = {
 	.attr = {
 		.name = "driver-version",
 		.mode = S_IRUGO,
@@ -884,7 +886,7 @@ static struct class_device_attribute hptiop_attr_version = {
 	.show = hptiop_show_version,
 };
 
-static struct class_device_attribute hptiop_attr_fw_version = {
+static struct device_attribute hptiop_attr_fw_version = {
 	.attr = {
 		.name = "firmware-version",
 		.mode = S_IRUGO,
@@ -892,7 +894,7 @@ static struct class_device_attribute hptiop_attr_fw_version = {
 	.show = hptiop_show_fw_version,
 };
 
-static struct class_device_attribute *hptiop_attrs[] = {
+static struct device_attribute *hptiop_attrs[] = {
 	&hptiop_attr_version,
 	&hptiop_attr_fw_version,
 	NULL
diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index 78d46a900bb5..4a922c57125e 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c
@@ -1456,9 +1456,10 @@ static int ibmvscsi_change_queue_depth(struct scsi_device *sdev, int qdepth)
 /* ------------------------------------------------------------
  * sysfs attributes
  */
-static ssize_t show_host_srp_version(struct class_device *class_dev, char *buf)
+static ssize_t show_host_srp_version(struct device *dev,
+				     struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
 	int len;
 
@@ -1467,7 +1468,7 @@ static ssize_t show_host_srp_version(struct class_device *class_dev, char *buf)
 	return len;
 }
 
-static struct class_device_attribute ibmvscsi_host_srp_version = {
+static struct device_attribute ibmvscsi_host_srp_version = {
 	.attr = {
 		 .name = "srp_version",
 		 .mode = S_IRUGO,
@@ -1475,10 +1476,11 @@ static struct class_device_attribute ibmvscsi_host_srp_version = {
 	.show = show_host_srp_version,
 };
 
-static ssize_t show_host_partition_name(struct class_device *class_dev,
+static ssize_t show_host_partition_name(struct device *dev,
+					struct device_attribute *attr,
 					char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
 	int len;
 
@@ -1487,7 +1489,7 @@ static ssize_t show_host_partition_name(struct class_device *class_dev,
 	return len;
 }
 
-static struct class_device_attribute ibmvscsi_host_partition_name = {
+static struct device_attribute ibmvscsi_host_partition_name = {
 	.attr = {
 		 .name = "partition_name",
 		 .mode = S_IRUGO,
@@ -1495,10 +1497,11 @@ static struct class_device_attribute ibmvscsi_host_partition_name = {
 	.show = show_host_partition_name,
 };
 
-static ssize_t show_host_partition_number(struct class_device *class_dev,
+static ssize_t show_host_partition_number(struct device *dev,
+					  struct device_attribute *attr,
 					  char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
 	int len;
 
@@ -1507,7 +1510,7 @@ static ssize_t show_host_partition_number(struct class_device *class_dev,
 	return len;
 }
 
-static struct class_device_attribute ibmvscsi_host_partition_number = {
+static struct device_attribute ibmvscsi_host_partition_number = {
 	.attr = {
 		 .name = "partition_number",
 		 .mode = S_IRUGO,
@@ -1515,9 +1518,10 @@ static struct class_device_attribute ibmvscsi_host_partition_number = {
 	.show = show_host_partition_number,
 };
 
-static ssize_t show_host_mad_version(struct class_device *class_dev, char *buf)
+static ssize_t show_host_mad_version(struct device *dev,
+				     struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
 	int len;
 
@@ -1526,7 +1530,7 @@ static ssize_t show_host_mad_version(struct class_device *class_dev, char *buf)
 	return len;
 }
 
-static struct class_device_attribute ibmvscsi_host_mad_version = {
+static struct device_attribute ibmvscsi_host_mad_version = {
 	.attr = {
 		 .name = "mad_version",
 		 .mode = S_IRUGO,
@@ -1534,9 +1538,10 @@ static struct class_device_attribute ibmvscsi_host_mad_version = {
 	.show = show_host_mad_version,
 };
 
-static ssize_t show_host_os_type(struct class_device *class_dev, char *buf)
+static ssize_t show_host_os_type(struct device *dev,
+				 struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
 	int len;
 
@@ -1544,7 +1549,7 @@ static ssize_t show_host_os_type(struct class_device *class_dev, char *buf)
 	return len;
 }
 
-static struct class_device_attribute ibmvscsi_host_os_type = {
+static struct device_attribute ibmvscsi_host_os_type = {
 	.attr = {
 		 .name = "os_type",
 		 .mode = S_IRUGO,
@@ -1552,9 +1557,10 @@ static struct class_device_attribute ibmvscsi_host_os_type = {
 	.show = show_host_os_type,
 };
 
-static ssize_t show_host_config(struct class_device *class_dev, char *buf)
+static ssize_t show_host_config(struct device *dev,
+				struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
 
 	/* returns null-terminated host config data */
@@ -1564,7 +1570,7 @@ static ssize_t show_host_config(struct class_device *class_dev, char *buf)
 		return 0;
 }
 
-static struct class_device_attribute ibmvscsi_host_config = {
+static struct device_attribute ibmvscsi_host_config = {
 	.attr = {
 		 .name = "config",
 		 .mode = S_IRUGO,
@@ -1572,7 +1578,7 @@ static struct class_device_attribute ibmvscsi_host_config = {
 	.show = show_host_config,
 };
 
-static struct class_device_attribute *ibmvscsi_attrs[] = {
+static struct device_attribute *ibmvscsi_attrs[] = {
 	&ibmvscsi_host_srp_version,
 	&ibmvscsi_host_partition_name,
 	&ibmvscsi_host_partition_number,
diff --git a/drivers/scsi/ibmvscsi/ibmvstgt.c b/drivers/scsi/ibmvscsi/ibmvstgt.c
index e5881e92d0fb..3b9514c8f1f1 100644
--- a/drivers/scsi/ibmvscsi/ibmvstgt.c
+++ b/drivers/scsi/ibmvscsi/ibmvstgt.c
@@ -780,32 +780,35 @@ static int ibmvstgt_it_nexus_response(struct Scsi_Host *shost, u64 itn_id,
 	return 0;
 }
 
-static ssize_t system_id_show(struct class_device *cdev, char *buf)
+static ssize_t system_id_show(struct device *dev,
+			      struct device_attribute *attr, char *buf)
 {
 	return snprintf(buf, PAGE_SIZE, "%s\n", system_id);
 }
 
-static ssize_t partition_number_show(struct class_device *cdev, char *buf)
+static ssize_t partition_number_show(struct device *dev,
+				     struct device_attribute *attr, char *buf)
 {
 	return snprintf(buf, PAGE_SIZE, "%x\n", partition_number);
 }
 
-static ssize_t unit_address_show(struct class_device *cdev, char *buf)
+static ssize_t unit_address_show(struct device *dev,
+				  struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(cdev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct srp_target *target = host_to_srp_target(shost);
 	struct vio_port *vport = target_to_port(target);
 	return snprintf(buf, PAGE_SIZE, "%x\n", vport->dma_dev->unit_address);
 }
 
-static CLASS_DEVICE_ATTR(system_id, S_IRUGO, system_id_show, NULL);
-static CLASS_DEVICE_ATTR(partition_number, S_IRUGO, partition_number_show, NULL);
-static CLASS_DEVICE_ATTR(unit_address, S_IRUGO, unit_address_show, NULL);
+static DEVICE_ATTR(system_id, S_IRUGO, system_id_show, NULL);
+static DEVICE_ATTR(partition_number, S_IRUGO, partition_number_show, NULL);
+static DEVICE_ATTR(unit_address, S_IRUGO, unit_address_show, NULL);
 
-static struct class_device_attribute *ibmvstgt_attrs[] = {
-	&class_device_attr_system_id,
-	&class_device_attr_partition_number,
-	&class_device_attr_unit_address,
+static struct device_attribute *ibmvstgt_attrs[] = {
+	&dev_attr_system_id,
+	&dev_attr_partition_number,
+	&dev_attr_unit_address,
 	NULL,
 };
 
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index 65dc18dea845..de5ae6a65029 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -2431,7 +2431,7 @@ restart:
 	}
 
 	spin_unlock_irqrestore(ioa_cfg->host->host_lock, lock_flags);
-	kobject_uevent(&ioa_cfg->host->shost_classdev.kobj, KOBJ_CHANGE);
+	kobject_uevent(&ioa_cfg->host->shost_dev.kobj, KOBJ_CHANGE);
 	LEAVE;
 }
 
@@ -2451,8 +2451,8 @@ static ssize_t ipr_read_trace(struct kobject *kobj,
 			      struct bin_attribute *bin_attr,
 			      char *buf, loff_t off, size_t count)
 {
-	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
-	struct Scsi_Host *shost = class_to_shost(cdev);
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags = 0;
 	int size = IPR_TRACE_SIZE;
@@ -2492,15 +2492,16 @@ static const struct {
 
 /**
  * ipr_show_write_caching - Show the write caching attribute
- * @class_dev:	class device struct
- * @buf:		buffer
+ * @dev:	device struct
+ * @buf:	buffer
  *
  * Return value:
  *	number of bytes printed to buffer
  **/
-static ssize_t ipr_show_write_caching(struct class_device *class_dev, char *buf)
+static ssize_t ipr_show_write_caching(struct device *dev,
+				      struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags = 0;
 	int i, len = 0;
@@ -2519,19 +2520,20 @@ static ssize_t ipr_show_write_caching(struct class_device *class_dev, char *buf)
 
 /**
  * ipr_store_write_caching - Enable/disable adapter write cache
- * @class_dev:	class_device struct
- * @buf:		buffer
- * @count:		buffer size
+ * @dev:	device struct
+ * @buf:	buffer
+ * @count:	buffer size
  *
  * This function will enable/disable adapter write cache.
  *
  * Return value:
  * 	count on success / other on failure
  **/
-static ssize_t ipr_store_write_caching(struct class_device *class_dev,
-					const char *buf, size_t count)
+static ssize_t ipr_store_write_caching(struct device *dev,
+				       struct device_attribute *attr,
+				       const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags = 0;
 	enum ipr_cache_state new_state = CACHE_INVALID;
@@ -2569,7 +2571,7 @@ static ssize_t ipr_store_write_caching(struct class_device *class_dev,
 	return count;
 }
 
-static struct class_device_attribute ipr_ioa_cache_attr = {
+static struct device_attribute ipr_ioa_cache_attr = {
 	.attr = {
 		.name =		"write_cache",
 		.mode =		S_IRUGO | S_IWUSR,
@@ -2580,15 +2582,16 @@ static struct class_device_attribute ipr_ioa_cache_attr = {
 
 /**
  * ipr_show_fw_version - Show the firmware version
- * @class_dev:	class device struct
- * @buf:		buffer
+ * @dev:	class device struct
+ * @buf:	buffer
  *
  * Return value:
  *	number of bytes printed to buffer
  **/
-static ssize_t ipr_show_fw_version(struct class_device *class_dev, char *buf)
+static ssize_t ipr_show_fw_version(struct device *dev,
+				   struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	struct ipr_inquiry_page3 *ucode_vpd = &ioa_cfg->vpd_cbs->page3_data;
 	unsigned long lock_flags = 0;
@@ -2603,7 +2606,7 @@ static ssize_t ipr_show_fw_version(struct class_device *class_dev, char *buf)
 	return len;
 }
 
-static struct class_device_attribute ipr_fw_version_attr = {
+static struct device_attribute ipr_fw_version_attr = {
 	.attr = {
 		.name =		"fw_version",
 		.mode =		S_IRUGO,
@@ -2613,15 +2616,16 @@ static struct class_device_attribute ipr_fw_version_attr = {
 
 /**
  * ipr_show_log_level - Show the adapter's error logging level
- * @class_dev:	class device struct
- * @buf:		buffer
+ * @dev:	class device struct
+ * @buf:	buffer
  *
  * Return value:
  * 	number of bytes printed to buffer
  **/
-static ssize_t ipr_show_log_level(struct class_device *class_dev, char *buf)
+static ssize_t ipr_show_log_level(struct device *dev,
+				   struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags = 0;
 	int len;
@@ -2634,16 +2638,17 @@ static ssize_t ipr_show_log_level(struct class_device *class_dev, char *buf)
 
 /**
  * ipr_store_log_level - Change the adapter's error logging level
- * @class_dev:	class device struct
- * @buf:		buffer
+ * @dev:	class device struct
+ * @buf:	buffer
  *
  * Return value:
  * 	number of bytes printed to buffer
  **/
-static ssize_t ipr_store_log_level(struct class_device *class_dev,
+static ssize_t ipr_store_log_level(struct device *dev,
+			           struct device_attribute *attr,
 				   const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags = 0;
 
@@ -2653,7 +2658,7 @@ static ssize_t ipr_store_log_level(struct class_device *class_dev,
 	return strlen(buf);
 }
 
-static struct class_device_attribute ipr_log_level_attr = {
+static struct device_attribute ipr_log_level_attr = {
 	.attr = {
 		.name =		"log_level",
 		.mode =		S_IRUGO | S_IWUSR,
@@ -2664,9 +2669,9 @@ static struct class_device_attribute ipr_log_level_attr = {
 
 /**
  * ipr_store_diagnostics - IOA Diagnostics interface
- * @class_dev:	class_device struct
- * @buf:		buffer
- * @count:		buffer size
+ * @dev:	device struct
+ * @buf:	buffer
+ * @count:	buffer size
  *
  * This function will reset the adapter and wait a reasonable
  * amount of time for any errors that the adapter might log.
@@ -2674,10 +2679,11 @@ static struct class_device_attribute ipr_log_level_attr = {
  * Return value:
  * 	count on success / other on failure
  **/
-static ssize_t ipr_store_diagnostics(struct class_device *class_dev,
+static ssize_t ipr_store_diagnostics(struct device *dev,
+				     struct device_attribute *attr,
 				     const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags = 0;
 	int rc = count;
@@ -2714,7 +2720,7 @@ static ssize_t ipr_store_diagnostics(struct class_device *class_dev,
 	return rc;
 }
 
-static struct class_device_attribute ipr_diagnostics_attr = {
+static struct device_attribute ipr_diagnostics_attr = {
 	.attr = {
 		.name =		"run_diagnostics",
 		.mode =		S_IWUSR,
@@ -2724,15 +2730,16 @@ static struct class_device_attribute ipr_diagnostics_attr = {
 
 /**
  * ipr_show_adapter_state - Show the adapter's state
- * @class_dev:	class device struct
- * @buf:		buffer
+ * @class_dev:	device struct
+ * @buf:	buffer
  *
  * Return value:
  * 	number of bytes printed to buffer
  **/
-static ssize_t ipr_show_adapter_state(struct class_device *class_dev, char *buf)
+static ssize_t ipr_show_adapter_state(struct device *dev,
+				      struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags = 0;
 	int len;
@@ -2748,19 +2755,20 @@ static ssize_t ipr_show_adapter_state(struct class_device *class_dev, char *buf)
 
 /**
  * ipr_store_adapter_state - Change adapter state
- * @class_dev:	class_device struct
- * @buf:		buffer
- * @count:		buffer size
+ * @dev:	device struct
+ * @buf:	buffer
+ * @count:	buffer size
  *
  * This function will change the adapter's state.
  *
  * Return value:
  * 	count on success / other on failure
  **/
-static ssize_t ipr_store_adapter_state(struct class_device *class_dev,
+static ssize_t ipr_store_adapter_state(struct device *dev,
+				       struct device_attribute *attr,
 				       const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags;
 	int result = count;
@@ -2781,7 +2789,7 @@ static ssize_t ipr_store_adapter_state(struct class_device *class_dev,
 	return result;
 }
 
-static struct class_device_attribute ipr_ioa_state_attr = {
+static struct device_attribute ipr_ioa_state_attr = {
 	.attr = {
 		.name =		"state",
 		.mode =		S_IRUGO | S_IWUSR,
@@ -2792,19 +2800,20 @@ static struct class_device_attribute ipr_ioa_state_attr = {
 
 /**
  * ipr_store_reset_adapter - Reset the adapter
- * @class_dev:	class_device struct
- * @buf:		buffer
- * @count:		buffer size
+ * @dev:	device struct
+ * @buf:	buffer
+ * @count:	buffer size
  *
  * This function will reset the adapter.
  *
  * Return value:
  * 	count on success / other on failure
  **/
-static ssize_t ipr_store_reset_adapter(struct class_device *class_dev,
+static ssize_t ipr_store_reset_adapter(struct device *dev,
+				       struct device_attribute *attr,
 				       const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	unsigned long lock_flags;
 	int result = count;
@@ -2821,7 +2830,7 @@ static ssize_t ipr_store_reset_adapter(struct class_device *class_dev,
 	return result;
 }
 
-static struct class_device_attribute ipr_ioa_reset_attr = {
+static struct device_attribute ipr_ioa_reset_attr = {
 	.attr = {
 		.name =		"reset_host",
 		.mode =		S_IWUSR,
@@ -3054,19 +3063,20 @@ static int ipr_update_ioa_ucode(struct ipr_ioa_cfg *ioa_cfg,
 
 /**
  * ipr_store_update_fw - Update the firmware on the adapter
- * @class_dev:	class_device struct
- * @buf:		buffer
- * @count:		buffer size
+ * @class_dev:	device struct
+ * @buf:	buffer
+ * @count:	buffer size
  *
  * This function will update the firmware on the adapter.
  *
  * Return value:
  * 	count on success / other on failure
  **/
-static ssize_t ipr_store_update_fw(struct class_device *class_dev,
-				       const char *buf, size_t count)
+static ssize_t ipr_store_update_fw(struct device *dev,
+				   struct device_attribute *attr,
+				   const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	struct ipr_ucode_image_header *image_hdr;
 	const struct firmware *fw_entry;
@@ -3124,7 +3134,7 @@ out:
 	return result;
 }
 
-static struct class_device_attribute ipr_update_fw_attr = {
+static struct device_attribute ipr_update_fw_attr = {
 	.attr = {
 		.name =		"update_fw",
 		.mode =		S_IWUSR,
@@ -3132,7 +3142,7 @@ static struct class_device_attribute ipr_update_fw_attr = {
 	.store = ipr_store_update_fw
 };
 
-static struct class_device_attribute *ipr_ioa_attrs[] = {
+static struct device_attribute *ipr_ioa_attrs[] = {
 	&ipr_fw_version_attr,
 	&ipr_log_level_attr,
 	&ipr_diagnostics_attr,
@@ -3159,7 +3169,7 @@ static ssize_t ipr_read_dump(struct kobject *kobj,
 			     struct bin_attribute *bin_attr,
 			     char *buf, loff_t off, size_t count)
 {
-	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
+	struct device *cdev = container_of(kobj, struct device, kobj);
 	struct Scsi_Host *shost = class_to_shost(cdev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	struct ipr_dump *dump;
@@ -3322,7 +3332,7 @@ static ssize_t ipr_write_dump(struct kobject *kobj,
 			      struct bin_attribute *bin_attr,
 			      char *buf, loff_t off, size_t count)
 {
-	struct class_device *cdev = container_of(kobj,struct class_device,kobj);
+	struct device *cdev = container_of(kobj, struct device, kobj);
 	struct Scsi_Host *shost = class_to_shost(cdev);
 	struct ipr_ioa_cfg *ioa_cfg = (struct ipr_ioa_cfg *)shost->hostdata;
 	int rc;
@@ -7671,9 +7681,9 @@ static void ipr_remove(struct pci_dev *pdev)
 
 	ENTER;
 
-	ipr_remove_trace_file(&ioa_cfg->host->shost_classdev.kobj,
+	ipr_remove_trace_file(&ioa_cfg->host->shost_dev.kobj,
 			      &ipr_trace_attr);
-	ipr_remove_dump_file(&ioa_cfg->host->shost_classdev.kobj,
+	ipr_remove_dump_file(&ioa_cfg->host->shost_dev.kobj,
 			     &ipr_dump_attr);
 	scsi_remove_host(ioa_cfg->host);
 
@@ -7714,7 +7724,7 @@ static int __devinit ipr_probe(struct pci_dev *pdev,
 		return rc;
 	}
 
-	rc = ipr_create_trace_file(&ioa_cfg->host->shost_classdev.kobj,
+	rc = ipr_create_trace_file(&ioa_cfg->host->shost_dev.kobj,
 				   &ipr_trace_attr);
 
 	if (rc) {
@@ -7723,11 +7733,11 @@ static int __devinit ipr_probe(struct pci_dev *pdev,
 		return rc;
 	}
 
-	rc = ipr_create_dump_file(&ioa_cfg->host->shost_classdev.kobj,
+	rc = ipr_create_dump_file(&ioa_cfg->host->shost_dev.kobj,
 				   &ipr_dump_attr);
 
 	if (rc) {
-		ipr_remove_trace_file(&ioa_cfg->host->shost_classdev.kobj,
+		ipr_remove_trace_file(&ioa_cfg->host->shost_dev.kobj,
 				      &ipr_trace_attr);
 		scsi_remove_host(ioa_cfg->host);
 		__ipr_remove(pdev);
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 74c9fc204211..a9fbb3f88659 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -66,23 +66,26 @@ lpfc_jedec_to_ascii(int incr, char hdw[])
 }
 
 static ssize_t
-lpfc_drvr_version_show(struct class_device *cdev, char *buf)
+lpfc_drvr_version_show(struct device *dev, struct device_attribute *attr,
+		       char *buf)
 {
 	return snprintf(buf, PAGE_SIZE, LPFC_MODULE_DESC "\n");
 }
 
 static ssize_t
-lpfc_info_show(struct class_device *cdev, char *buf)
+lpfc_info_show(struct device *dev, struct device_attribute *attr,
+	       char *buf)
 {
-	struct Scsi_Host *host = class_to_shost(cdev);
+	struct Scsi_Host *host = class_to_shost(dev);
 
 	return snprintf(buf, PAGE_SIZE, "%s\n",lpfc_info(host));
 }
 
 static ssize_t
-lpfc_serialnum_show(struct class_device *cdev, char *buf)
+lpfc_serialnum_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -90,18 +93,20 @@ lpfc_serialnum_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_temp_sensor_show(struct class_device *cdev, char *buf)
+lpfc_temp_sensor_show(struct device *dev, struct device_attribute *attr,
+		      char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(cdev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	return snprintf(buf, PAGE_SIZE, "%d\n",phba->temp_sensor_support);
 }
 
 static ssize_t
-lpfc_modeldesc_show(struct class_device *cdev, char *buf)
+lpfc_modeldesc_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -109,9 +114,10 @@ lpfc_modeldesc_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_modelname_show(struct class_device *cdev, char *buf)
+lpfc_modelname_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -119,9 +125,10 @@ lpfc_modelname_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_programtype_show(struct class_device *cdev, char *buf)
+lpfc_programtype_show(struct device *dev, struct device_attribute *attr,
+		      char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -129,9 +136,10 @@ lpfc_programtype_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_vportnum_show(struct class_device *cdev, char *buf)
+lpfc_vportnum_show(struct device *dev, struct device_attribute *attr,
+		   char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -139,9 +147,10 @@ lpfc_vportnum_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_fwrev_show(struct class_device *cdev, char *buf)
+lpfc_fwrev_show(struct device *dev, struct device_attribute *attr,
+		char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	char fwrev[32];
@@ -151,10 +160,10 @@ lpfc_fwrev_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_hdw_show(struct class_device *cdev, char *buf)
+lpfc_hdw_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
 	char hdw[9];
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	lpfc_vpd_t *vp = &phba->vpd;
@@ -163,18 +172,20 @@ lpfc_hdw_show(struct class_device *cdev, char *buf)
 	return snprintf(buf, PAGE_SIZE, "%s\n", hdw);
 }
 static ssize_t
-lpfc_option_rom_version_show(struct class_device *cdev, char *buf)
+lpfc_option_rom_version_show(struct device *dev, struct device_attribute *attr,
+			     char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
 	return snprintf(buf, PAGE_SIZE, "%s\n", phba->OptionROMVersion);
 }
 static ssize_t
-lpfc_state_show(struct class_device *cdev, char *buf)
+lpfc_state_show(struct device *dev, struct device_attribute *attr,
+		char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	int  len = 0;
@@ -243,9 +254,10 @@ lpfc_state_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_num_discovered_ports_show(struct class_device *cdev, char *buf)
+lpfc_num_discovered_ports_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
@@ -367,9 +379,10 @@ lpfc_selective_reset(struct lpfc_hba *phba)
 }
 
 static ssize_t
-lpfc_issue_reset(struct class_device *cdev, const char *buf, size_t count)
+lpfc_issue_reset(struct device *dev, struct device_attribute *attr,
+		 const char *buf, size_t count)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -385,9 +398,10 @@ lpfc_issue_reset(struct class_device *cdev, const char *buf, size_t count)
 }
 
 static ssize_t
-lpfc_nport_evt_cnt_show(struct class_device *cdev, char *buf)
+lpfc_nport_evt_cnt_show(struct device *dev, struct device_attribute *attr,
+			char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -395,9 +409,10 @@ lpfc_nport_evt_cnt_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_board_mode_show(struct class_device *cdev, char *buf)
+lpfc_board_mode_show(struct device *dev, struct device_attribute *attr,
+		     char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	char  * state;
@@ -415,9 +430,10 @@ lpfc_board_mode_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_board_mode_store(struct class_device *cdev, const char *buf, size_t count)
+lpfc_board_mode_store(struct device *dev, struct device_attribute *attr,
+		      const char *buf, size_t count)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	struct completion online_compl;
@@ -509,9 +525,10 @@ lpfc_get_hba_info(struct lpfc_hba *phba,
 }
 
 static ssize_t
-lpfc_max_rpi_show(struct class_device *cdev, char *buf)
+lpfc_max_rpi_show(struct device *dev, struct device_attribute *attr,
+		  char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	uint32_t cnt;
@@ -522,9 +539,10 @@ lpfc_max_rpi_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_used_rpi_show(struct class_device *cdev, char *buf)
+lpfc_used_rpi_show(struct device *dev, struct device_attribute *attr,
+		   char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	uint32_t cnt, acnt;
@@ -535,9 +553,10 @@ lpfc_used_rpi_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_max_xri_show(struct class_device *cdev, char *buf)
+lpfc_max_xri_show(struct device *dev, struct device_attribute *attr,
+		  char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	uint32_t cnt;
@@ -548,9 +567,10 @@ lpfc_max_xri_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_used_xri_show(struct class_device *cdev, char *buf)
+lpfc_used_xri_show(struct device *dev, struct device_attribute *attr,
+		   char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	uint32_t cnt, acnt;
@@ -561,9 +581,10 @@ lpfc_used_xri_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_max_vpi_show(struct class_device *cdev, char *buf)
+lpfc_max_vpi_show(struct device *dev, struct device_attribute *attr,
+		  char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	uint32_t cnt;
@@ -574,9 +595,10 @@ lpfc_max_vpi_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_used_vpi_show(struct class_device *cdev, char *buf)
+lpfc_used_vpi_show(struct device *dev, struct device_attribute *attr,
+		   char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	uint32_t cnt, acnt;
@@ -587,9 +609,10 @@ lpfc_used_vpi_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_npiv_info_show(struct class_device *cdev, char *buf)
+lpfc_npiv_info_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -601,9 +624,10 @@ lpfc_npiv_info_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_poll_show(struct class_device *cdev, char *buf)
+lpfc_poll_show(struct device *dev, struct device_attribute *attr,
+	       char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -611,10 +635,10 @@ lpfc_poll_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-lpfc_poll_store(struct class_device *cdev, const char *buf,
-		size_t count)
+lpfc_poll_store(struct device *dev, struct device_attribute *attr,
+		const char *buf, size_t count)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	uint32_t creg_val;
@@ -670,9 +694,10 @@ lpfc_poll_store(struct class_device *cdev, const char *buf,
 
 #define lpfc_param_show(attr)	\
 static ssize_t \
-lpfc_##attr##_show(struct class_device *cdev, char *buf) \
+lpfc_##attr##_show(struct device *dev, struct device_attribute *attr, \
+		   char *buf) \
 { \
-	struct Scsi_Host  *shost = class_to_shost(cdev);\
+	struct Scsi_Host  *shost = class_to_shost(dev);\
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;\
 	struct lpfc_hba   *phba = vport->phba;\
 	int val = 0;\
@@ -683,9 +708,10 @@ lpfc_##attr##_show(struct class_device *cdev, char *buf) \
 
 #define lpfc_param_hex_show(attr)	\
 static ssize_t \
-lpfc_##attr##_show(struct class_device *cdev, char *buf) \
+lpfc_##attr##_show(struct device *dev, struct device_attribute *attr, \
+		   char *buf) \
 { \
-	struct Scsi_Host  *shost = class_to_shost(cdev);\
+	struct Scsi_Host  *shost = class_to_shost(dev);\
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;\
 	struct lpfc_hba   *phba = vport->phba;\
 	int val = 0;\
@@ -725,9 +751,10 @@ lpfc_##attr##_set(struct lpfc_hba *phba, int val) \
 
 #define lpfc_param_store(attr)	\
 static ssize_t \
-lpfc_##attr##_store(struct class_device *cdev, const char *buf, size_t count) \
+lpfc_##attr##_store(struct device *dev, struct device_attribute *attr, \
+		    const char *buf, size_t count) \
 { \
-	struct Scsi_Host  *shost = class_to_shost(cdev);\
+	struct Scsi_Host  *shost = class_to_shost(dev);\
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;\
 	struct lpfc_hba   *phba = vport->phba;\
 	int val=0;\
@@ -743,9 +770,10 @@ lpfc_##attr##_store(struct class_device *cdev, const char *buf, size_t count) \
 
 #define lpfc_vport_param_show(attr)	\
 static ssize_t \
-lpfc_##attr##_show(struct class_device *cdev, char *buf) \
+lpfc_##attr##_show(struct device *dev, struct device_attribute *attr, \
+		   char *buf) \
 { \
-	struct Scsi_Host  *shost = class_to_shost(cdev);\
+	struct Scsi_Host  *shost = class_to_shost(dev);\
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;\
 	int val = 0;\
 	val = vport->cfg_##attr;\
@@ -754,9 +782,10 @@ lpfc_##attr##_show(struct class_device *cdev, char *buf) \
 
 #define lpfc_vport_param_hex_show(attr)	\
 static ssize_t \
-lpfc_##attr##_show(struct class_device *cdev, char *buf) \
+lpfc_##attr##_show(struct device *dev, struct device_attribute *attr, \
+		   char *buf) \
 { \
-	struct Scsi_Host  *shost = class_to_shost(cdev);\
+	struct Scsi_Host  *shost = class_to_shost(dev);\
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;\
 	int val = 0;\
 	val = vport->cfg_##attr;\
@@ -794,9 +823,10 @@ lpfc_##attr##_set(struct lpfc_vport *vport, int val) \
 
 #define lpfc_vport_param_store(attr)	\
 static ssize_t \
-lpfc_##attr##_store(struct class_device *cdev, const char *buf, size_t count) \
+lpfc_##attr##_store(struct device *dev, struct device_attribute *attr, \
+		    const char *buf, size_t count) \
 { \
-	struct Scsi_Host  *shost = class_to_shost(cdev);\
+	struct Scsi_Host  *shost = class_to_shost(dev);\
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;\
 	int val=0;\
 	if (!isdigit(buf[0]))\
@@ -822,7 +852,7 @@ module_param(lpfc_##name, int, 0);\
 MODULE_PARM_DESC(lpfc_##name, desc);\
 lpfc_param_show(name)\
 lpfc_param_init(name, defval, minval, maxval)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
+static DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
 
 #define LPFC_ATTR_RW(name, defval, minval, maxval, desc) \
 static int lpfc_##name = defval;\
@@ -832,8 +862,8 @@ lpfc_param_show(name)\
 lpfc_param_init(name, defval, minval, maxval)\
 lpfc_param_set(name, defval, minval, maxval)\
 lpfc_param_store(name)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
-			 lpfc_##name##_show, lpfc_##name##_store)
+static DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
+		   lpfc_##name##_show, lpfc_##name##_store)
 
 #define LPFC_ATTR_HEX_R(name, defval, minval, maxval, desc) \
 static int lpfc_##name = defval;\
@@ -841,7 +871,7 @@ module_param(lpfc_##name, int, 0);\
 MODULE_PARM_DESC(lpfc_##name, desc);\
 lpfc_param_hex_show(name)\
 lpfc_param_init(name, defval, minval, maxval)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
+static DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
 
 #define LPFC_ATTR_HEX_RW(name, defval, minval, maxval, desc) \
 static int lpfc_##name = defval;\
@@ -851,8 +881,8 @@ lpfc_param_hex_show(name)\
 lpfc_param_init(name, defval, minval, maxval)\
 lpfc_param_set(name, defval, minval, maxval)\
 lpfc_param_store(name)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
-			 lpfc_##name##_show, lpfc_##name##_store)
+static DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
+		   lpfc_##name##_show, lpfc_##name##_store)
 
 #define LPFC_VPORT_ATTR(name, defval, minval, maxval, desc) \
 static int lpfc_##name = defval;\
@@ -866,7 +896,7 @@ module_param(lpfc_##name, int, 0);\
 MODULE_PARM_DESC(lpfc_##name, desc);\
 lpfc_vport_param_show(name)\
 lpfc_vport_param_init(name, defval, minval, maxval)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
+static DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
 
 #define LPFC_VPORT_ATTR_RW(name, defval, minval, maxval, desc) \
 static int lpfc_##name = defval;\
@@ -876,8 +906,8 @@ lpfc_vport_param_show(name)\
 lpfc_vport_param_init(name, defval, minval, maxval)\
 lpfc_vport_param_set(name, defval, minval, maxval)\
 lpfc_vport_param_store(name)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
-			 lpfc_##name##_show, lpfc_##name##_store)
+static DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
+		   lpfc_##name##_show, lpfc_##name##_store)
 
 #define LPFC_VPORT_ATTR_HEX_R(name, defval, minval, maxval, desc) \
 static int lpfc_##name = defval;\
@@ -885,7 +915,7 @@ module_param(lpfc_##name, int, 0);\
 MODULE_PARM_DESC(lpfc_##name, desc);\
 lpfc_vport_param_hex_show(name)\
 lpfc_vport_param_init(name, defval, minval, maxval)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
+static DEVICE_ATTR(lpfc_##name, S_IRUGO , lpfc_##name##_show, NULL)
 
 #define LPFC_VPORT_ATTR_HEX_RW(name, defval, minval, maxval, desc) \
 static int lpfc_##name = defval;\
@@ -895,46 +925,44 @@ lpfc_vport_param_hex_show(name)\
 lpfc_vport_param_init(name, defval, minval, maxval)\
 lpfc_vport_param_set(name, defval, minval, maxval)\
 lpfc_vport_param_store(name)\
-static CLASS_DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
-			 lpfc_##name##_show, lpfc_##name##_store)
-
-static CLASS_DEVICE_ATTR(info, S_IRUGO, lpfc_info_show, NULL);
-static CLASS_DEVICE_ATTR(serialnum, S_IRUGO, lpfc_serialnum_show, NULL);
-static CLASS_DEVICE_ATTR(modeldesc, S_IRUGO, lpfc_modeldesc_show, NULL);
-static CLASS_DEVICE_ATTR(modelname, S_IRUGO, lpfc_modelname_show, NULL);
-static CLASS_DEVICE_ATTR(programtype, S_IRUGO, lpfc_programtype_show, NULL);
-static CLASS_DEVICE_ATTR(portnum, S_IRUGO, lpfc_vportnum_show, NULL);
-static CLASS_DEVICE_ATTR(fwrev, S_IRUGO, lpfc_fwrev_show, NULL);
-static CLASS_DEVICE_ATTR(hdw, S_IRUGO, lpfc_hdw_show, NULL);
-static CLASS_DEVICE_ATTR(state, S_IRUGO, lpfc_state_show, NULL);
-static CLASS_DEVICE_ATTR(option_rom_version, S_IRUGO,
-					lpfc_option_rom_version_show, NULL);
-static CLASS_DEVICE_ATTR(num_discovered_ports, S_IRUGO,
-					lpfc_num_discovered_ports_show, NULL);
-static CLASS_DEVICE_ATTR(nport_evt_cnt, S_IRUGO, lpfc_nport_evt_cnt_show, NULL);
-static CLASS_DEVICE_ATTR(lpfc_drvr_version, S_IRUGO, lpfc_drvr_version_show,
-			 NULL);
-static CLASS_DEVICE_ATTR(board_mode, S_IRUGO | S_IWUSR,
-			 lpfc_board_mode_show, lpfc_board_mode_store);
-static CLASS_DEVICE_ATTR(issue_reset, S_IWUSR, NULL, lpfc_issue_reset);
-static CLASS_DEVICE_ATTR(max_vpi, S_IRUGO, lpfc_max_vpi_show, NULL);
-static CLASS_DEVICE_ATTR(used_vpi, S_IRUGO, lpfc_used_vpi_show, NULL);
-static CLASS_DEVICE_ATTR(max_rpi, S_IRUGO, lpfc_max_rpi_show, NULL);
-static CLASS_DEVICE_ATTR(used_rpi, S_IRUGO, lpfc_used_rpi_show, NULL);
-static CLASS_DEVICE_ATTR(max_xri, S_IRUGO, lpfc_max_xri_show, NULL);
-static CLASS_DEVICE_ATTR(used_xri, S_IRUGO, lpfc_used_xri_show, NULL);
-static CLASS_DEVICE_ATTR(npiv_info, S_IRUGO, lpfc_npiv_info_show, NULL);
-static CLASS_DEVICE_ATTR(lpfc_temp_sensor, S_IRUGO, lpfc_temp_sensor_show,
-			 NULL);
+static DEVICE_ATTR(lpfc_##name, S_IRUGO | S_IWUSR,\
+		   lpfc_##name##_show, lpfc_##name##_store)
+
+static DEVICE_ATTR(info, S_IRUGO, lpfc_info_show, NULL);
+static DEVICE_ATTR(serialnum, S_IRUGO, lpfc_serialnum_show, NULL);
+static DEVICE_ATTR(modeldesc, S_IRUGO, lpfc_modeldesc_show, NULL);
+static DEVICE_ATTR(modelname, S_IRUGO, lpfc_modelname_show, NULL);
+static DEVICE_ATTR(programtype, S_IRUGO, lpfc_programtype_show, NULL);
+static DEVICE_ATTR(portnum, S_IRUGO, lpfc_vportnum_show, NULL);
+static DEVICE_ATTR(fwrev, S_IRUGO, lpfc_fwrev_show, NULL);
+static DEVICE_ATTR(hdw, S_IRUGO, lpfc_hdw_show, NULL);
+static DEVICE_ATTR(state, S_IRUGO, lpfc_state_show, NULL);
+static DEVICE_ATTR(option_rom_version, S_IRUGO,
+		   lpfc_option_rom_version_show, NULL);
+static DEVICE_ATTR(num_discovered_ports, S_IRUGO,
+		   lpfc_num_discovered_ports_show, NULL);
+static DEVICE_ATTR(nport_evt_cnt, S_IRUGO, lpfc_nport_evt_cnt_show, NULL);
+static DEVICE_ATTR(lpfc_drvr_version, S_IRUGO, lpfc_drvr_version_show, NULL);
+static DEVICE_ATTR(board_mode, S_IRUGO | S_IWUSR,
+		   lpfc_board_mode_show, lpfc_board_mode_store);
+static DEVICE_ATTR(issue_reset, S_IWUSR, NULL, lpfc_issue_reset);
+static DEVICE_ATTR(max_vpi, S_IRUGO, lpfc_max_vpi_show, NULL);
+static DEVICE_ATTR(used_vpi, S_IRUGO, lpfc_used_vpi_show, NULL);
+static DEVICE_ATTR(max_rpi, S_IRUGO, lpfc_max_rpi_show, NULL);
+static DEVICE_ATTR(used_rpi, S_IRUGO, lpfc_used_rpi_show, NULL);
+static DEVICE_ATTR(max_xri, S_IRUGO, lpfc_max_xri_show, NULL);
+static DEVICE_ATTR(used_xri, S_IRUGO, lpfc_used_xri_show, NULL);
+static DEVICE_ATTR(npiv_info, S_IRUGO, lpfc_npiv_info_show, NULL);
+static DEVICE_ATTR(lpfc_temp_sensor, S_IRUGO, lpfc_temp_sensor_show, NULL);
 
 
 static char *lpfc_soft_wwn_key = "C99G71SL8032A";
 
 static ssize_t
-lpfc_soft_wwn_enable_store(struct class_device *cdev, const char *buf,
-				size_t count)
+lpfc_soft_wwn_enable_store(struct device *dev, struct device_attribute *attr,
+			   const char *buf, size_t count)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	unsigned int cnt = count;
@@ -963,13 +991,14 @@ lpfc_soft_wwn_enable_store(struct class_device *cdev, const char *buf,
 	phba->soft_wwn_enable = 1;
 	return count;
 }
-static CLASS_DEVICE_ATTR(lpfc_soft_wwn_enable, S_IWUSR, NULL,
-				lpfc_soft_wwn_enable_store);
+static DEVICE_ATTR(lpfc_soft_wwn_enable, S_IWUSR, NULL,
+		   lpfc_soft_wwn_enable_store);
 
 static ssize_t
-lpfc_soft_wwpn_show(struct class_device *cdev, char *buf)
+lpfc_soft_wwpn_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -979,9 +1008,10 @@ lpfc_soft_wwpn_show(struct class_device *cdev, char *buf)
 
 
 static ssize_t
-lpfc_soft_wwpn_store(struct class_device *cdev, const char *buf, size_t count)
+lpfc_soft_wwpn_store(struct device *dev, struct device_attribute *attr,
+		     const char *buf, size_t count)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	struct completion online_compl;
@@ -1047,13 +1077,14 @@ lpfc_soft_wwpn_store(struct class_device *cdev, const char *buf, size_t count)
 				"reinit adapter - %d\n", stat2);
 	return (stat1 || stat2) ? -EIO : count;
 }
-static CLASS_DEVICE_ATTR(lpfc_soft_wwpn, S_IRUGO | S_IWUSR,\
-			 lpfc_soft_wwpn_show, lpfc_soft_wwpn_store);
+static DEVICE_ATTR(lpfc_soft_wwpn, S_IRUGO | S_IWUSR,\
+		   lpfc_soft_wwpn_show, lpfc_soft_wwpn_store);
 
 static ssize_t
-lpfc_soft_wwnn_show(struct class_device *cdev, char *buf)
+lpfc_soft_wwnn_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(cdev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct lpfc_hba *phba = ((struct lpfc_vport *)shost->hostdata)->phba;
 	return snprintf(buf, PAGE_SIZE, "0x%llx\n",
 			(unsigned long long)phba->cfg_soft_wwnn);
@@ -1061,9 +1092,10 @@ lpfc_soft_wwnn_show(struct class_device *cdev, char *buf)
 
 
 static ssize_t
-lpfc_soft_wwnn_store(struct class_device *cdev, const char *buf, size_t count)
+lpfc_soft_wwnn_store(struct device *dev, struct device_attribute *attr,
+		     const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(cdev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	struct lpfc_hba *phba = ((struct lpfc_vport *)shost->hostdata)->phba;
 	unsigned int i, j, cnt=count;
 	u8 wwnn[8];
@@ -1107,8 +1139,8 @@ lpfc_soft_wwnn_store(struct class_device *cdev, const char *buf, size_t count)
 
 	return count;
 }
-static CLASS_DEVICE_ATTR(lpfc_soft_wwnn, S_IRUGO | S_IWUSR,\
-			 lpfc_soft_wwnn_show, lpfc_soft_wwnn_store);
+static DEVICE_ATTR(lpfc_soft_wwnn, S_IRUGO | S_IWUSR,\
+		   lpfc_soft_wwnn_show, lpfc_soft_wwnn_store);
 
 
 static int lpfc_poll = 0;
@@ -1118,8 +1150,8 @@ MODULE_PARM_DESC(lpfc_poll, "FCP ring polling mode control:"
 		 " 1 - poll with interrupts enabled"
 		 " 3 - poll and disable FCP ring interrupts");
 
-static CLASS_DEVICE_ATTR(lpfc_poll, S_IRUGO | S_IWUSR,
-			 lpfc_poll_show, lpfc_poll_store);
+static DEVICE_ATTR(lpfc_poll, S_IRUGO | S_IWUSR,
+		   lpfc_poll_show, lpfc_poll_store);
 
 int  lpfc_sli_mode = 0;
 module_param(lpfc_sli_mode, int, 0);
@@ -1133,7 +1165,7 @@ module_param(lpfc_enable_npiv, int, 0);
 MODULE_PARM_DESC(lpfc_enable_npiv, "Enable NPIV functionality");
 lpfc_param_show(enable_npiv);
 lpfc_param_init(enable_npiv, 0, 0, 1);
-static CLASS_DEVICE_ATTR(lpfc_enable_npiv, S_IRUGO,
+static DEVICE_ATTR(lpfc_enable_npiv, S_IRUGO,
 			 lpfc_enable_npiv_show, NULL);
 
 /*
@@ -1147,9 +1179,10 @@ MODULE_PARM_DESC(lpfc_nodev_tmo,
 		 "Seconds driver will hold I/O waiting "
 		 "for a device to come back");
 static ssize_t
-lpfc_nodev_tmo_show(struct class_device *cdev, char *buf)
+lpfc_nodev_tmo_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	int val = 0;
 	val = vport->cfg_devloss_tmo;
@@ -1221,8 +1254,8 @@ lpfc_nodev_tmo_set(struct lpfc_vport *vport, int val)
 
 lpfc_vport_param_store(nodev_tmo)
 
-static CLASS_DEVICE_ATTR(lpfc_nodev_tmo, S_IRUGO | S_IWUSR,
-			 lpfc_nodev_tmo_show, lpfc_nodev_tmo_store);
+static DEVICE_ATTR(lpfc_nodev_tmo, S_IRUGO | S_IWUSR,
+		   lpfc_nodev_tmo_show, lpfc_nodev_tmo_store);
 
 /*
 # lpfc_devloss_tmo: If set, it will hold all I/O errors on devices that
@@ -1255,8 +1288,8 @@ lpfc_devloss_tmo_set(struct lpfc_vport *vport, int val)
 }
 
 lpfc_vport_param_store(devloss_tmo)
-static CLASS_DEVICE_ATTR(lpfc_devloss_tmo, S_IRUGO | S_IWUSR,
-	lpfc_devloss_tmo_show, lpfc_devloss_tmo_store);
+static DEVICE_ATTR(lpfc_devloss_tmo, S_IRUGO | S_IWUSR,
+		   lpfc_devloss_tmo_show, lpfc_devloss_tmo_store);
 
 /*
 # lpfc_log_verbose: Only turn this flag on if you are willing to risk being
@@ -1374,8 +1407,8 @@ lpfc_restrict_login_set(struct lpfc_vport *vport, int val)
 	return 0;
 }
 lpfc_vport_param_store(restrict_login);
-static CLASS_DEVICE_ATTR(lpfc_restrict_login, S_IRUGO | S_IWUSR,
-			 lpfc_restrict_login_show, lpfc_restrict_login_store);
+static DEVICE_ATTR(lpfc_restrict_login, S_IRUGO | S_IWUSR,
+		   lpfc_restrict_login_show, lpfc_restrict_login_store);
 
 /*
 # Some disk devices have a "select ID" or "select Target" capability.
@@ -1433,7 +1466,7 @@ MODULE_PARM_DESC(lpfc_topology, "Select Fibre Channel topology");
 lpfc_param_show(topology)
 lpfc_param_init(topology, 0, 0, 6)
 lpfc_param_store(topology)
-static CLASS_DEVICE_ATTR(lpfc_topology, S_IRUGO | S_IWUSR,
+static DEVICE_ATTR(lpfc_topology, S_IRUGO | S_IWUSR,
 		lpfc_topology_show, lpfc_topology_store);
 
 /*
@@ -1497,7 +1530,7 @@ lpfc_link_speed_init(struct lpfc_hba *phba, int val)
 }
 
 lpfc_param_store(link_speed)
-static CLASS_DEVICE_ATTR(lpfc_link_speed, S_IRUGO | S_IWUSR,
+static DEVICE_ATTR(lpfc_link_speed, S_IRUGO | S_IWUSR,
 		lpfc_link_speed_show, lpfc_link_speed_store);
 
 /*
@@ -1623,82 +1656,81 @@ LPFC_ATTR_R(enable_hba_heartbeat, 1, 0, 1, "Enable HBA Heartbeat.");
 LPFC_ATTR_R(sg_seg_cnt, LPFC_DEFAULT_SG_SEG_CNT, LPFC_DEFAULT_SG_SEG_CNT,
 	    LPFC_MAX_SG_SEG_CNT, "Max Scatter Gather Segment Count");
 
-struct class_device_attribute *lpfc_hba_attrs[] = {
-	&class_device_attr_info,
-	&class_device_attr_serialnum,
-	&class_device_attr_modeldesc,
-	&class_device_attr_modelname,
-	&class_device_attr_programtype,
-	&class_device_attr_portnum,
-	&class_device_attr_fwrev,
-	&class_device_attr_hdw,
-	&class_device_attr_option_rom_version,
-	&class_device_attr_state,
-	&class_device_attr_num_discovered_ports,
-	&class_device_attr_lpfc_drvr_version,
-	&class_device_attr_lpfc_temp_sensor,
-	&class_device_attr_lpfc_log_verbose,
-	&class_device_attr_lpfc_lun_queue_depth,
-	&class_device_attr_lpfc_hba_queue_depth,
-	&class_device_attr_lpfc_peer_port_login,
-	&class_device_attr_lpfc_nodev_tmo,
-	&class_device_attr_lpfc_devloss_tmo,
-	&class_device_attr_lpfc_fcp_class,
-	&class_device_attr_lpfc_use_adisc,
-	&class_device_attr_lpfc_ack0,
-	&class_device_attr_lpfc_topology,
-	&class_device_attr_lpfc_scan_down,
-	&class_device_attr_lpfc_link_speed,
-	&class_device_attr_lpfc_cr_delay,
-	&class_device_attr_lpfc_cr_count,
-	&class_device_attr_lpfc_multi_ring_support,
-	&class_device_attr_lpfc_multi_ring_rctl,
-	&class_device_attr_lpfc_multi_ring_type,
-	&class_device_attr_lpfc_fdmi_on,
-	&class_device_attr_lpfc_max_luns,
-	&class_device_attr_lpfc_enable_npiv,
-	&class_device_attr_nport_evt_cnt,
-	&class_device_attr_board_mode,
-	&class_device_attr_max_vpi,
-	&class_device_attr_used_vpi,
-	&class_device_attr_max_rpi,
-	&class_device_attr_used_rpi,
-	&class_device_attr_max_xri,
-	&class_device_attr_used_xri,
-	&class_device_attr_npiv_info,
-	&class_device_attr_issue_reset,
-	&class_device_attr_lpfc_poll,
-	&class_device_attr_lpfc_poll_tmo,
-	&class_device_attr_lpfc_use_msi,
-	&class_device_attr_lpfc_soft_wwnn,
-	&class_device_attr_lpfc_soft_wwpn,
-	&class_device_attr_lpfc_soft_wwn_enable,
-	&class_device_attr_lpfc_enable_hba_reset,
-	&class_device_attr_lpfc_enable_hba_heartbeat,
-	&class_device_attr_lpfc_sg_seg_cnt,
+struct device_attribute *lpfc_hba_attrs[] = {
+	&dev_attr_info,
+	&dev_attr_serialnum,
+	&dev_attr_modeldesc,
+	&dev_attr_modelname,
+	&dev_attr_programtype,
+	&dev_attr_portnum,
+	&dev_attr_fwrev,
+	&dev_attr_hdw,
+	&dev_attr_option_rom_version,
+	&dev_attr_state,
+	&dev_attr_num_discovered_ports,
+	&dev_attr_lpfc_drvr_version,
+	&dev_attr_lpfc_temp_sensor,
+	&dev_attr_lpfc_log_verbose,
+	&dev_attr_lpfc_lun_queue_depth,
+	&dev_attr_lpfc_hba_queue_depth,
+	&dev_attr_lpfc_peer_port_login,
+	&dev_attr_lpfc_nodev_tmo,
+	&dev_attr_lpfc_devloss_tmo,
+	&dev_attr_lpfc_fcp_class,
+	&dev_attr_lpfc_use_adisc,
+	&dev_attr_lpfc_ack0,
+	&dev_attr_lpfc_topology,
+	&dev_attr_lpfc_scan_down,
+	&dev_attr_lpfc_link_speed,
+	&dev_attr_lpfc_cr_delay,
+	&dev_attr_lpfc_cr_count,
+	&dev_attr_lpfc_multi_ring_support,
+	&dev_attr_lpfc_multi_ring_rctl,
+	&dev_attr_lpfc_multi_ring_type,
+	&dev_attr_lpfc_fdmi_on,
+	&dev_attr_lpfc_max_luns,
+	&dev_attr_lpfc_enable_npiv,
+	&dev_attr_nport_evt_cnt,
+	&dev_attr_board_mode,
+	&dev_attr_max_vpi,
+	&dev_attr_used_vpi,
+	&dev_attr_max_rpi,
+	&dev_attr_used_rpi,
+	&dev_attr_max_xri,
+	&dev_attr_used_xri,
+	&dev_attr_npiv_info,
+	&dev_attr_issue_reset,
+	&dev_attr_lpfc_poll,
+	&dev_attr_lpfc_poll_tmo,
+	&dev_attr_lpfc_use_msi,
+	&dev_attr_lpfc_soft_wwnn,
+	&dev_attr_lpfc_soft_wwpn,
+	&dev_attr_lpfc_soft_wwn_enable,
+	&dev_attr_lpfc_enable_hba_reset,
+	&dev_attr_lpfc_enable_hba_heartbeat,
+	&dev_attr_lpfc_sg_seg_cnt,
 	NULL,
 };
 
-struct class_device_attribute *lpfc_vport_attrs[] = {
-	&class_device_attr_info,
-	&class_device_attr_state,
-	&class_device_attr_num_discovered_ports,
-	&class_device_attr_lpfc_drvr_version,
-
-	&class_device_attr_lpfc_log_verbose,
-	&class_device_attr_lpfc_lun_queue_depth,
-	&class_device_attr_lpfc_nodev_tmo,
-	&class_device_attr_lpfc_devloss_tmo,
-	&class_device_attr_lpfc_hba_queue_depth,
-	&class_device_attr_lpfc_peer_port_login,
-	&class_device_attr_lpfc_restrict_login,
-	&class_device_attr_lpfc_fcp_class,
-	&class_device_attr_lpfc_use_adisc,
-	&class_device_attr_lpfc_fdmi_on,
-	&class_device_attr_lpfc_max_luns,
-	&class_device_attr_nport_evt_cnt,
-	&class_device_attr_npiv_info,
-	&class_device_attr_lpfc_enable_da_id,
+struct device_attribute *lpfc_vport_attrs[] = {
+	&dev_attr_info,
+	&dev_attr_state,
+	&dev_attr_num_discovered_ports,
+	&dev_attr_lpfc_drvr_version,
+	&dev_attr_lpfc_log_verbose,
+	&dev_attr_lpfc_lun_queue_depth,
+	&dev_attr_lpfc_nodev_tmo,
+	&dev_attr_lpfc_devloss_tmo,
+	&dev_attr_lpfc_hba_queue_depth,
+	&dev_attr_lpfc_peer_port_login,
+	&dev_attr_lpfc_restrict_login,
+	&dev_attr_lpfc_fcp_class,
+	&dev_attr_lpfc_use_adisc,
+	&dev_attr_lpfc_fdmi_on,
+	&dev_attr_lpfc_max_luns,
+	&dev_attr_nport_evt_cnt,
+	&dev_attr_npiv_info,
+	&dev_attr_lpfc_enable_da_id,
 	NULL,
 };
 
@@ -1707,9 +1739,8 @@ sysfs_ctlreg_write(struct kobject *kobj, struct bin_attribute *bin_attr,
 		   char *buf, loff_t off, size_t count)
 {
 	size_t buf_off;
-	struct class_device *cdev = container_of(kobj, struct class_device,
-						 kobj);
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -1741,9 +1772,8 @@ sysfs_ctlreg_read(struct kobject *kobj, struct bin_attribute *bin_attr,
 {
 	size_t buf_off;
 	uint32_t * tmp_ptr;
-	struct class_device *cdev = container_of(kobj, struct class_device,
-						 kobj);
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 
@@ -1798,9 +1828,8 @@ static ssize_t
 sysfs_mbox_write(struct kobject *kobj, struct bin_attribute *bin_attr,
 		 char *buf, loff_t off, size_t count)
 {
-	struct class_device *cdev = container_of(kobj, struct class_device,
-						 kobj);
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	struct lpfcMboxq  *mbox = NULL;
@@ -1853,9 +1882,8 @@ static ssize_t
 sysfs_mbox_read(struct kobject *kobj, struct bin_attribute *bin_attr,
 		char *buf, loff_t off, size_t count)
 {
-	struct class_device *cdev = container_of(kobj, struct class_device,
-						 kobj);
-	struct Scsi_Host  *shost = class_to_shost(cdev);
+	struct device *dev = container_of(kobj, struct device, kobj);
+	struct Scsi_Host  *shost = class_to_shost(dev);
 	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
 	struct lpfc_hba   *phba = vport->phba;
 	int rc;
@@ -2038,19 +2066,19 @@ lpfc_alloc_sysfs_attr(struct lpfc_vport *vport)
 	struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
 	int error;
 
-	error = sysfs_create_bin_file(&shost->shost_classdev.kobj,
+	error = sysfs_create_bin_file(&shost->shost_dev.kobj,
 				      &sysfs_ctlreg_attr);
 	if (error)
 		goto out;
 
-	error = sysfs_create_bin_file(&shost->shost_classdev.kobj,
+	error = sysfs_create_bin_file(&shost->shost_dev.kobj,
 				      &sysfs_mbox_attr);
 	if (error)
 		goto out_remove_ctlreg_attr;
 
 	return 0;
 out_remove_ctlreg_attr:
-	sysfs_remove_bin_file(&shost->shost_classdev.kobj, &sysfs_ctlreg_attr);
+	sysfs_remove_bin_file(&shost->shost_dev.kobj, &sysfs_ctlreg_attr);
 out:
 	return error;
 }
@@ -2060,8 +2088,8 @@ lpfc_free_sysfs_attr(struct lpfc_vport *vport)
 {
 	struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
 
-	sysfs_remove_bin_file(&shost->shost_classdev.kobj, &sysfs_mbox_attr);
-	sysfs_remove_bin_file(&shost->shost_classdev.kobj, &sysfs_ctlreg_attr);
+	sysfs_remove_bin_file(&shost->shost_dev.kobj, &sysfs_mbox_attr);
+	sysfs_remove_bin_file(&shost->shost_dev.kobj, &sysfs_ctlreg_attr);
 }
 
 
@@ -2443,9 +2471,11 @@ lpfc_set_rport_loss_tmo(struct fc_rport *rport, uint32_t timeout)
 
 #define lpfc_rport_show_function(field, format_string, sz, cast)	\
 static ssize_t								\
-lpfc_show_rport_##field (struct class_device *cdev, char *buf)		\
+lpfc_show_rport_##field (struct device *dev,				\
+			 struct device_attribute *attr,			\
+			 char *buf)					\
 {									\
-	struct fc_rport *rport = transport_class_to_rport(cdev);	\
+	struct fc_rport *rport = transport_class_to_rport(dev);		\
 	struct lpfc_rport_data *rdata = rport->hostdata;		\
 	return snprintf(buf, sz, format_string,				\
 		(rdata->target) ? cast rdata->target->field : 0);	\
diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index 0819f5f39de5..7c9f8317d972 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -253,8 +253,8 @@ void lpfc_get_cfgparam(struct lpfc_hba *);
 void lpfc_get_vport_cfgparam(struct lpfc_vport *);
 int lpfc_alloc_sysfs_attr(struct lpfc_vport *);
 void lpfc_free_sysfs_attr(struct lpfc_vport *);
-extern struct class_device_attribute *lpfc_hba_attrs[];
-extern struct class_device_attribute *lpfc_vport_attrs[];
+extern struct device_attribute *lpfc_hba_attrs[];
+extern struct device_attribute *lpfc_vport_attrs[];
 extern struct scsi_host_template lpfc_template;
 extern struct scsi_host_template lpfc_vport_template;
 extern struct fc_function_template lpfc_transport_functions;
diff --git a/drivers/scsi/megaraid/megaraid_mbox.c b/drivers/scsi/megaraid/megaraid_mbox.c
index 9f041929aca5..820f91fb63ba 100644
--- a/drivers/scsi/megaraid/megaraid_mbox.c
+++ b/drivers/scsi/megaraid/megaraid_mbox.c
@@ -125,7 +125,7 @@ static irqreturn_t megaraid_isr(int, void *);
 
 static void megaraid_mbox_dpc(unsigned long);
 
-static ssize_t megaraid_sysfs_show_app_hndl(struct class_device *, char *);
+static ssize_t megaraid_sysfs_show_app_hndl(struct device *, struct device_attribute *attr, char *);
 static ssize_t megaraid_sysfs_show_ldnum(struct device *, struct device_attribute *attr, char *);
 
 static int megaraid_cmm_register(adapter_t *);
@@ -313,12 +313,12 @@ static struct pci_driver megaraid_pci_driver = {
 // definitions for the device attributes for exporting logical drive number
 // for a scsi address (Host, Channel, Id, Lun)
 
-CLASS_DEVICE_ATTR(megaraid_mbox_app_hndl, S_IRUSR, megaraid_sysfs_show_app_hndl,
+DEVICE_ATTR(megaraid_mbox_app_hndl, S_IRUSR, megaraid_sysfs_show_app_hndl,
 		NULL);
 
 // Host template initializer for megaraid mbox sysfs device attributes
-static struct class_device_attribute *megaraid_shost_attrs[] = {
-	&class_device_attr_megaraid_mbox_app_hndl,
+static struct device_attribute *megaraid_shost_attrs[] = {
+	&dev_attr_megaraid_mbox_app_hndl,
 	NULL,
 };
 
@@ -4063,9 +4063,10 @@ megaraid_sysfs_get_ldmap(adapter_t *adapter)
  * handle, since we do not interface with applications directly.
  */
 static ssize_t
-megaraid_sysfs_show_app_hndl(struct class_device *cdev, char *buf)
+megaraid_sysfs_show_app_hndl(struct device *dev, struct device_attribute *attr,
+			     char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(cdev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	adapter_t	*adapter = (adapter_t *)SCSIHOST2ADAP(shost);
 	uint32_t	app_hndl;
 
diff --git a/drivers/scsi/ncr53c8xx.c b/drivers/scsi/ncr53c8xx.c
index c5ebf018b378..d89289400425 100644
--- a/drivers/scsi/ncr53c8xx.c
+++ b/drivers/scsi/ncr53c8xx.c
@@ -8243,7 +8243,8 @@ static void process_waiting_list(struct ncb *np, int sts)
 
 #undef next_wcmd
 
-static ssize_t show_ncr53c8xx_revision(struct class_device *dev, char *buf)
+static ssize_t show_ncr53c8xx_revision(struct device *dev,
+				       struct device_attribute *attr, char *buf)
 {
 	struct Scsi_Host *host = class_to_shost(dev);
 	struct host_data *host_data = (struct host_data *)host->hostdata;
@@ -8251,12 +8252,12 @@ static ssize_t show_ncr53c8xx_revision(struct class_device *dev, char *buf)
 	return snprintf(buf, 20, "0x%x\n", host_data->ncb->revision_id);
 }
   
-static struct class_device_attribute ncr53c8xx_revision_attr = {
+static struct device_attribute ncr53c8xx_revision_attr = {
 	.attr	= { .name = "revision", .mode = S_IRUGO, },
 	.show	= show_ncr53c8xx_revision,
 };
   
-static struct class_device_attribute *ncr53c8xx_host_attrs[] = {
+static struct device_attribute *ncr53c8xx_host_attrs[] = {
 	&ncr53c8xx_revision_attr,
 	NULL
 };
diff --git a/drivers/scsi/osst.c b/drivers/scsi/osst.c
index abef7048f25b..31f7aec44d90 100644
--- a/drivers/scsi/osst.c
+++ b/drivers/scsi/osst.c
@@ -5591,9 +5591,10 @@ static void osst_remove_sysfs_files(struct device_driver *sysfs)
  * sysfs support for accessing ADR header information
  */
 
-static ssize_t osst_adr_rev_show(struct class_device *class_dev, char *buf)
+static ssize_t osst_adr_rev_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
 {
-	struct osst_tape * STp = (struct osst_tape *) class_get_devdata (class_dev);
+	struct osst_tape * STp = (struct osst_tape *) dev_get_drvdata (dev);
 	ssize_t l = 0;
 
 	if (STp && STp->header_ok && STp->linux_media)
@@ -5601,11 +5602,13 @@ static ssize_t osst_adr_rev_show(struct class_device *class_dev, char *buf)
 	return l;
 }
 
-CLASS_DEVICE_ATTR(ADR_rev, S_IRUGO, osst_adr_rev_show, NULL);
+DEVICE_ATTR(ADR_rev, S_IRUGO, osst_adr_rev_show, NULL);
 
-static ssize_t osst_linux_media_version_show(struct class_device *class_dev, char *buf)
+static ssize_t osst_linux_media_version_show(struct device *dev,
+					     struct device_attribute *attr,
+					     char *buf)
 {
-	struct osst_tape * STp = (struct osst_tape *) class_get_devdata (class_dev);
+	struct osst_tape * STp = (struct osst_tape *) dev_get_drvdata (dev);
 	ssize_t l = 0;
 
 	if (STp && STp->header_ok && STp->linux_media)
@@ -5613,11 +5616,12 @@ static ssize_t osst_linux_media_version_show(struct class_device *class_dev, cha
 	return l;
 }
 
-CLASS_DEVICE_ATTR(media_version, S_IRUGO, osst_linux_media_version_show, NULL);
+DEVICE_ATTR(media_version, S_IRUGO, osst_linux_media_version_show, NULL);
 
-static ssize_t osst_capacity_show(struct class_device *class_dev, char *buf)
+static ssize_t osst_capacity_show(struct device *dev,
+				  struct device_attribute *attr, char *buf)
 {
-	struct osst_tape * STp = (struct osst_tape *) class_get_devdata (class_dev);
+	struct osst_tape * STp = (struct osst_tape *) dev_get_drvdata (dev);
 	ssize_t l = 0;
 
 	if (STp && STp->header_ok && STp->linux_media)
@@ -5625,11 +5629,13 @@ static ssize_t osst_capacity_show(struct class_device *class_dev, char *buf)
 	return l;
 }
 
-CLASS_DEVICE_ATTR(capacity, S_IRUGO, osst_capacity_show, NULL);
+DEVICE_ATTR(capacity, S_IRUGO, osst_capacity_show, NULL);
 
-static ssize_t osst_first_data_ppos_show(struct class_device *class_dev, char *buf)
+static ssize_t osst_first_data_ppos_show(struct device *dev,
+					 struct device_attribute *attr,
+					 char *buf)
 {
-	struct osst_tape * STp = (struct osst_tape *) class_get_devdata (class_dev);
+	struct osst_tape * STp = (struct osst_tape *) dev_get_drvdata (dev);
 	ssize_t l = 0;
 
 	if (STp && STp->header_ok && STp->linux_media)
@@ -5637,11 +5643,13 @@ static ssize_t osst_first_data_ppos_show(struct class_device *class_dev, char *b
 	return l;
 }
 
-CLASS_DEVICE_ATTR(BOT_frame, S_IRUGO, osst_first_data_ppos_show, NULL);
+DEVICE_ATTR(BOT_frame, S_IRUGO, osst_first_data_ppos_show, NULL);
 
-static ssize_t osst_eod_frame_ppos_show(struct class_device *class_dev, char *buf)
+static ssize_t osst_eod_frame_ppos_show(struct device *dev,
+					struct device_attribute *attr,
+					char *buf)
 {
-	struct osst_tape * STp = (struct osst_tape *) class_get_devdata (class_dev);
+	struct osst_tape * STp = (struct osst_tape *) dev_get_drvdata (dev);
 	ssize_t l = 0;
 
 	if (STp && STp->header_ok && STp->linux_media)
@@ -5649,11 +5657,12 @@ static ssize_t osst_eod_frame_ppos_show(struct class_device *class_dev, char *bu
 	return l;
 }
 
-CLASS_DEVICE_ATTR(EOD_frame, S_IRUGO, osst_eod_frame_ppos_show, NULL);
+DEVICE_ATTR(EOD_frame, S_IRUGO, osst_eod_frame_ppos_show, NULL);
 
-static ssize_t osst_filemark_cnt_show(struct class_device *class_dev, char *buf)
+static ssize_t osst_filemark_cnt_show(struct device *dev,
+				      struct device_attribute *attr, char *buf)
 {
-	struct osst_tape * STp = (struct osst_tape *) class_get_devdata (class_dev);
+	struct osst_tape * STp = (struct osst_tape *) dev_get_drvdata (dev);
 	ssize_t l = 0;
 
 	if (STp && STp->header_ok && STp->linux_media)
@@ -5661,7 +5670,7 @@ static ssize_t osst_filemark_cnt_show(struct class_device *class_dev, char *buf)
 	return l;
 }
 
-CLASS_DEVICE_ATTR(file_count, S_IRUGO, osst_filemark_cnt_show, NULL);
+DEVICE_ATTR(file_count, S_IRUGO, osst_filemark_cnt_show, NULL);
 
 static struct class *osst_sysfs_class;
 
@@ -5678,44 +5687,37 @@ static int osst_sysfs_init(void)
 
 static void osst_sysfs_destroy(dev_t dev)
 {
-	class_device_destroy(osst_sysfs_class, dev);
+	device_destroy(osst_sysfs_class, dev);
 }
 
 static int osst_sysfs_add(dev_t dev, struct device *device, struct osst_tape * STp, char * name)
 {
-	struct class_device *osst_class_member;
+	struct device *osst_member;
 	int err;
 
-	osst_class_member = class_device_create(osst_sysfs_class, NULL, dev,
-						device, "%s", name);
-	if (IS_ERR(osst_class_member)) {
+	osst_member = device_create(osst_sysfs_class, device, dev, "%s", name);
+	if (IS_ERR(osst_member)) {
 		printk(KERN_WARNING "osst :W: Unable to add sysfs class member %s\n", name);
-		return PTR_ERR(osst_class_member);
+		return PTR_ERR(osst_member);
 	}
 
-	class_set_devdata(osst_class_member, STp);
-	err = class_device_create_file(osst_class_member,
-				       &class_device_attr_ADR_rev);
+	dev_set_drvdata(osst_member, STp);
+	err = device_create_file(osst_member, &dev_attr_ADR_rev);
 	if (err)
 		goto err_out;
-	err = class_device_create_file(osst_class_member,
-				       &class_device_attr_media_version);
+	err = device_create_file(osst_member, &dev_attr_media_version);
 	if (err)
 		goto err_out;
-	err = class_device_create_file(osst_class_member,
-				       &class_device_attr_capacity);
+	err = device_create_file(osst_member, &dev_attr_capacity);
 	if (err)
 		goto err_out;
-	err = class_device_create_file(osst_class_member,
-				       &class_device_attr_BOT_frame);
+	err = device_create_file(osst_member, &dev_attr_BOT_frame);
 	if (err)
 		goto err_out;
-	err = class_device_create_file(osst_class_member,
-				       &class_device_attr_EOD_frame);
+	err = device_create_file(osst_member, &dev_attr_EOD_frame);
 	if (err)
 		goto err_out;
-	err = class_device_create_file(osst_class_member,
-				       &class_device_attr_file_count);
+	err = device_create_file(osst_member, &dev_attr_file_count);
 	if (err)
 		goto err_out;
 
diff --git a/drivers/scsi/pcmcia/sym53c500_cs.c b/drivers/scsi/pcmcia/sym53c500_cs.c
index 3454a5714749..0be232b58ffb 100644
--- a/drivers/scsi/pcmcia/sym53c500_cs.c
+++ b/drivers/scsi/pcmcia/sym53c500_cs.c
@@ -632,9 +632,10 @@ SYM53C500_biosparm(struct scsi_device *disk,
 }
 
 static ssize_t
-SYM53C500_show_pio(struct class_device *cdev, char *buf)
+SYM53C500_show_pio(struct device *dev, struct device_attribute *attr,
+		   char *buf)
 {
-	struct Scsi_Host *SHp = class_to_shost(cdev);
+	struct Scsi_Host *SHp = class_to_shost(dev);
 	struct sym53c500_data *data =
 	    (struct sym53c500_data *)SHp->hostdata;
 
@@ -642,10 +643,11 @@ SYM53C500_show_pio(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-SYM53C500_store_pio(struct class_device *cdev, const char *buf, size_t count)
+SYM53C500_store_pio(struct device *dev, struct device_attribute *attr,
+		    const char *buf, size_t count)
 {
 	int pio;
-	struct Scsi_Host *SHp = class_to_shost(cdev);
+	struct Scsi_Host *SHp = class_to_shost(dev);
 	struct sym53c500_data *data =
 	    (struct sym53c500_data *)SHp->hostdata;
 
@@ -662,7 +664,7 @@ SYM53C500_store_pio(struct class_device *cdev, const char *buf, size_t count)
 *  SCSI HBA device attributes we want to
 *  make available via sysfs.
 */
-static struct class_device_attribute SYM53C500_pio_attr = {
+static struct device_attribute SYM53C500_pio_attr = {
 	.attr = {
 		.name = "fast_pio",
 		.mode = (S_IRUGO | S_IWUSR),
@@ -671,7 +673,7 @@ static struct class_device_attribute SYM53C500_pio_attr = {
 	.store = SYM53C500_store_pio,
 };
 
-static struct class_device_attribute *SYM53C500_shost_attrs[] = {
+static struct device_attribute *SYM53C500_shost_attrs[] = {
 	&SYM53C500_pio_attr,
 	NULL,
 };
diff --git a/drivers/scsi/qla2xxx/qla_attr.c b/drivers/scsi/qla2xxx/qla_attr.c
index 413d8cd6a324..d61df036910c 100644
--- a/drivers/scsi/qla2xxx/qla_attr.c
+++ b/drivers/scsi/qla2xxx/qla_attr.c
@@ -530,15 +530,17 @@ qla2x00_free_sysfs_attr(scsi_qla_host_t *ha)
 /* Scsi_Host attributes. */
 
 static ssize_t
-qla2x00_drvr_version_show(struct class_device *cdev, char *buf)
+qla2x00_drvr_version_show(struct device *dev,
+			  struct device_attribute *attr, char *buf)
 {
 	return snprintf(buf, PAGE_SIZE, "%s\n", qla2x00_version_str);
 }
 
 static ssize_t
-qla2x00_fw_version_show(struct class_device *cdev, char *buf)
+qla2x00_fw_version_show(struct device *dev,
+			struct device_attribute *attr, char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	char fw_str[30];
 
 	return snprintf(buf, PAGE_SIZE, "%s\n",
@@ -546,9 +548,10 @@ qla2x00_fw_version_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-qla2x00_serial_num_show(struct class_device *cdev, char *buf)
+qla2x00_serial_num_show(struct device *dev, struct device_attribute *attr,
+			char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	uint32_t sn;
 
 	if (IS_FWI2_CAPABLE(ha))
@@ -560,40 +563,45 @@ qla2x00_serial_num_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-qla2x00_isp_name_show(struct class_device *cdev, char *buf)
+qla2x00_isp_name_show(struct device *dev, struct device_attribute *attr,
+		      char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	return snprintf(buf, PAGE_SIZE, "ISP%04X\n", ha->pdev->device);
 }
 
 static ssize_t
-qla2x00_isp_id_show(struct class_device *cdev, char *buf)
+qla2x00_isp_id_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	return snprintf(buf, PAGE_SIZE, "%04x %04x %04x %04x\n",
 	    ha->product_id[0], ha->product_id[1], ha->product_id[2],
 	    ha->product_id[3]);
 }
 
 static ssize_t
-qla2x00_model_name_show(struct class_device *cdev, char *buf)
+qla2x00_model_name_show(struct device *dev, struct device_attribute *attr,
+			char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	return snprintf(buf, PAGE_SIZE, "%s\n", ha->model_number);
 }
 
 static ssize_t
-qla2x00_model_desc_show(struct class_device *cdev, char *buf)
+qla2x00_model_desc_show(struct device *dev, struct device_attribute *attr,
+			char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	return snprintf(buf, PAGE_SIZE, "%s\n",
 	    ha->model_desc ? ha->model_desc: "");
 }
 
 static ssize_t
-qla2x00_pci_info_show(struct class_device *cdev, char *buf)
+qla2x00_pci_info_show(struct device *dev, struct device_attribute *attr,
+		      char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	char pci_info[30];
 
 	return snprintf(buf, PAGE_SIZE, "%s\n",
@@ -601,9 +609,10 @@ qla2x00_pci_info_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-qla2x00_state_show(struct class_device *cdev, char *buf)
+qla2x00_state_show(struct device *dev, struct device_attribute *attr,
+		   char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	int len = 0;
 
 	if (atomic_read(&ha->loop_state) == LOOP_DOWN ||
@@ -639,9 +648,10 @@ qla2x00_state_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-qla2x00_zio_show(struct class_device *cdev, char *buf)
+qla2x00_zio_show(struct device *dev, struct device_attribute *attr,
+		 char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	int len = 0;
 
 	switch (ha->zio_mode) {
@@ -656,9 +666,10 @@ qla2x00_zio_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-qla2x00_zio_store(struct class_device *cdev, const char *buf, size_t count)
+qla2x00_zio_store(struct device *dev, struct device_attribute *attr,
+		  const char *buf, size_t count)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	int val = 0;
 	uint16_t zio_mode;
 
@@ -682,18 +693,19 @@ qla2x00_zio_store(struct class_device *cdev, const char *buf, size_t count)
 }
 
 static ssize_t
-qla2x00_zio_timer_show(struct class_device *cdev, char *buf)
+qla2x00_zio_timer_show(struct device *dev, struct device_attribute *attr,
+		       char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 
 	return snprintf(buf, PAGE_SIZE, "%d us\n", ha->zio_timer * 100);
 }
 
 static ssize_t
-qla2x00_zio_timer_store(struct class_device *cdev, const char *buf,
-    size_t count)
+qla2x00_zio_timer_store(struct device *dev, struct device_attribute *attr,
+			const char *buf, size_t count)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	int val = 0;
 	uint16_t zio_timer;
 
@@ -709,9 +721,10 @@ qla2x00_zio_timer_store(struct class_device *cdev, const char *buf,
 }
 
 static ssize_t
-qla2x00_beacon_show(struct class_device *cdev, char *buf)
+qla2x00_beacon_show(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	int len = 0;
 
 	if (ha->beacon_blink_led)
@@ -722,10 +735,10 @@ qla2x00_beacon_show(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-qla2x00_beacon_store(struct class_device *cdev, const char *buf,
-    size_t count)
+qla2x00_beacon_store(struct device *dev, struct device_attribute *attr,
+		     const char *buf, size_t count)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 	int val = 0;
 	int rval;
 
@@ -753,84 +766,86 @@ qla2x00_beacon_store(struct class_device *cdev, const char *buf,
 }
 
 static ssize_t
-qla2x00_optrom_bios_version_show(struct class_device *cdev, char *buf)
+qla2x00_optrom_bios_version_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 
 	return snprintf(buf, PAGE_SIZE, "%d.%02d\n", ha->bios_revision[1],
 	    ha->bios_revision[0]);
 }
 
 static ssize_t
-qla2x00_optrom_efi_version_show(struct class_device *cdev, char *buf)
+qla2x00_optrom_efi_version_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 
 	return snprintf(buf, PAGE_SIZE, "%d.%02d\n", ha->efi_revision[1],
 	    ha->efi_revision[0]);
 }
 
 static ssize_t
-qla2x00_optrom_fcode_version_show(struct class_device *cdev, char *buf)
+qla2x00_optrom_fcode_version_show(struct device *dev,
+				  struct device_attribute *attr, char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 
 	return snprintf(buf, PAGE_SIZE, "%d.%02d\n", ha->fcode_revision[1],
 	    ha->fcode_revision[0]);
 }
 
 static ssize_t
-qla2x00_optrom_fw_version_show(struct class_device *cdev, char *buf)
+qla2x00_optrom_fw_version_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
 {
-	scsi_qla_host_t *ha = shost_priv(class_to_shost(cdev));
+	scsi_qla_host_t *ha = shost_priv(class_to_shost(dev));
 
 	return snprintf(buf, PAGE_SIZE, "%d.%02d.%02d %d\n",
 	    ha->fw_revision[0], ha->fw_revision[1], ha->fw_revision[2],
 	    ha->fw_revision[3]);
 }
 
-static CLASS_DEVICE_ATTR(driver_version, S_IRUGO, qla2x00_drvr_version_show,
-	NULL);
-static CLASS_DEVICE_ATTR(fw_version, S_IRUGO, qla2x00_fw_version_show, NULL);
-static CLASS_DEVICE_ATTR(serial_num, S_IRUGO, qla2x00_serial_num_show, NULL);
-static CLASS_DEVICE_ATTR(isp_name, S_IRUGO, qla2x00_isp_name_show, NULL);
-static CLASS_DEVICE_ATTR(isp_id, S_IRUGO, qla2x00_isp_id_show, NULL);
-static CLASS_DEVICE_ATTR(model_name, S_IRUGO, qla2x00_model_name_show, NULL);
-static CLASS_DEVICE_ATTR(model_desc, S_IRUGO, qla2x00_model_desc_show, NULL);
-static CLASS_DEVICE_ATTR(pci_info, S_IRUGO, qla2x00_pci_info_show, NULL);
-static CLASS_DEVICE_ATTR(state, S_IRUGO, qla2x00_state_show, NULL);
-static CLASS_DEVICE_ATTR(zio, S_IRUGO | S_IWUSR, qla2x00_zio_show,
-    qla2x00_zio_store);
-static CLASS_DEVICE_ATTR(zio_timer, S_IRUGO | S_IWUSR, qla2x00_zio_timer_show,
-    qla2x00_zio_timer_store);
-static CLASS_DEVICE_ATTR(beacon, S_IRUGO | S_IWUSR, qla2x00_beacon_show,
-    qla2x00_beacon_store);
-static CLASS_DEVICE_ATTR(optrom_bios_version, S_IRUGO,
-    qla2x00_optrom_bios_version_show, NULL);
-static CLASS_DEVICE_ATTR(optrom_efi_version, S_IRUGO,
-    qla2x00_optrom_efi_version_show, NULL);
-static CLASS_DEVICE_ATTR(optrom_fcode_version, S_IRUGO,
-    qla2x00_optrom_fcode_version_show, NULL);
-static CLASS_DEVICE_ATTR(optrom_fw_version, S_IRUGO,
-    qla2x00_optrom_fw_version_show, NULL);
-
-struct class_device_attribute *qla2x00_host_attrs[] = {
-	&class_device_attr_driver_version,
-	&class_device_attr_fw_version,
-	&class_device_attr_serial_num,
-	&class_device_attr_isp_name,
-	&class_device_attr_isp_id,
-	&class_device_attr_model_name,
-	&class_device_attr_model_desc,
-	&class_device_attr_pci_info,
-	&class_device_attr_state,
-	&class_device_attr_zio,
-	&class_device_attr_zio_timer,
-	&class_device_attr_beacon,
-	&class_device_attr_optrom_bios_version,
-	&class_device_attr_optrom_efi_version,
-	&class_device_attr_optrom_fcode_version,
-	&class_device_attr_optrom_fw_version,
+static DEVICE_ATTR(driver_version, S_IRUGO, qla2x00_drvr_version_show, NULL);
+static DEVICE_ATTR(fw_version, S_IRUGO, qla2x00_fw_version_show, NULL);
+static DEVICE_ATTR(serial_num, S_IRUGO, qla2x00_serial_num_show, NULL);
+static DEVICE_ATTR(isp_name, S_IRUGO, qla2x00_isp_name_show, NULL);
+static DEVICE_ATTR(isp_id, S_IRUGO, qla2x00_isp_id_show, NULL);
+static DEVICE_ATTR(model_name, S_IRUGO, qla2x00_model_name_show, NULL);
+static DEVICE_ATTR(model_desc, S_IRUGO, qla2x00_model_desc_show, NULL);
+static DEVICE_ATTR(pci_info, S_IRUGO, qla2x00_pci_info_show, NULL);
+static DEVICE_ATTR(state, S_IRUGO, qla2x00_state_show, NULL);
+static DEVICE_ATTR(zio, S_IRUGO | S_IWUSR, qla2x00_zio_show, qla2x00_zio_store);
+static DEVICE_ATTR(zio_timer, S_IRUGO | S_IWUSR, qla2x00_zio_timer_show,
+		   qla2x00_zio_timer_store);
+static DEVICE_ATTR(beacon, S_IRUGO | S_IWUSR, qla2x00_beacon_show,
+		   qla2x00_beacon_store);
+static DEVICE_ATTR(optrom_bios_version, S_IRUGO,
+		   qla2x00_optrom_bios_version_show, NULL);
+static DEVICE_ATTR(optrom_efi_version, S_IRUGO,
+		   qla2x00_optrom_efi_version_show, NULL);
+static DEVICE_ATTR(optrom_fcode_version, S_IRUGO,
+		   qla2x00_optrom_fcode_version_show, NULL);
+static DEVICE_ATTR(optrom_fw_version, S_IRUGO, qla2x00_optrom_fw_version_show,
+		   NULL);
+
+struct device_attribute *qla2x00_host_attrs[] = {
+	&dev_attr_driver_version,
+	&dev_attr_fw_version,
+	&dev_attr_serial_num,
+	&dev_attr_isp_name,
+	&dev_attr_isp_id,
+	&dev_attr_model_name,
+	&dev_attr_model_desc,
+	&dev_attr_pci_info,
+	&dev_attr_state,
+	&dev_attr_zio,
+	&dev_attr_zio_timer,
+	&dev_attr_beacon,
+	&dev_attr_optrom_bios_version,
+	&dev_attr_optrom_efi_version,
+	&dev_attr_optrom_fcode_version,
+	&dev_attr_optrom_fw_version,
 	NULL,
 };
 
diff --git a/drivers/scsi/qla2xxx/qla_gbl.h b/drivers/scsi/qla2xxx/qla_gbl.h
index a9571c214a9e..76eb4fecce65 100644
--- a/drivers/scsi/qla2xxx/qla_gbl.h
+++ b/drivers/scsi/qla2xxx/qla_gbl.h
@@ -347,8 +347,8 @@ extern void qla2x00_get_sym_node_name(scsi_qla_host_t *, uint8_t *);
 /*
  * Global Function Prototypes in qla_attr.c source file.
  */
-struct class_device_attribute;
-extern struct class_device_attribute *qla2x00_host_attrs[];
+struct device_attribute;
+extern struct device_attribute *qla2x00_host_attrs[];
 struct fc_function_template;
 extern struct fc_function_template qla2xxx_transport_functions;
 extern struct fc_function_template qla2xxx_transport_vport_functions;
diff --git a/drivers/scsi/raid_class.c b/drivers/scsi/raid_class.c
index 52182a744ba6..913a931176ef 100644
--- a/drivers/scsi/raid_class.c
+++ b/drivers/scsi/raid_class.c
@@ -24,15 +24,15 @@ struct raid_internal {
 	struct raid_template r;
 	struct raid_function_template *f;
 	/* The actual attributes */
-	struct class_device_attribute private_attrs[RAID_NUM_ATTRS];
+	struct device_attribute private_attrs[RAID_NUM_ATTRS];
 	/* The array of null terminated pointers to attributes 
 	 * needed by scsi_sysfs.c */
-	struct class_device_attribute *attrs[RAID_NUM_ATTRS + 1];
+	struct device_attribute *attrs[RAID_NUM_ATTRS + 1];
 };
 
 struct raid_component {
 	struct list_head node;
-	struct class_device cdev;
+	struct device dev;
 	int num;
 };
 
@@ -50,9 +50,9 @@ struct raid_component {
 	tc_to_raid_internal(tc);					\
 })
 
-#define class_device_to_raid_internal(cdev) ({				\
+#define device_to_raid_internal(dev) ({				\
 	struct attribute_container *ac =				\
-		attribute_container_classdev_to_container(cdev);	\
+		attribute_container_classdev_to_container(dev);	\
 	ac_to_raid_internal(ac);					\
 })
 	
@@ -76,33 +76,33 @@ static int raid_match(struct attribute_container *cont, struct device *dev)
 }
 
 static int raid_setup(struct transport_container *tc, struct device *dev,
-		       struct class_device *cdev)
+		       struct device *cdev)
 {
 	struct raid_data *rd;
 
-	BUG_ON(class_get_devdata(cdev));
+	BUG_ON(dev_get_drvdata(cdev));
 
 	rd = kzalloc(sizeof(*rd), GFP_KERNEL);
 	if (!rd)
 		return -ENOMEM;
 
 	INIT_LIST_HEAD(&rd->component_list);
-	class_set_devdata(cdev, rd);
+	dev_set_drvdata(cdev, rd);
 		
 	return 0;
 }
 
 static int raid_remove(struct transport_container *tc, struct device *dev,
-		       struct class_device *cdev)
+		       struct device *cdev)
 {
-	struct raid_data *rd = class_get_devdata(cdev);
+	struct raid_data *rd = dev_get_drvdata(cdev);
 	struct raid_component *rc, *next;
 	dev_printk(KERN_ERR, dev, "RAID REMOVE\n");
-	class_set_devdata(cdev, NULL);
+	dev_set_drvdata(cdev, NULL);
 	list_for_each_entry_safe(rc, next, &rd->component_list, node) {
 		list_del(&rc->node);
-		dev_printk(KERN_ERR, rc->cdev.dev, "RAID COMPONENT REMOVE\n");
-		class_device_unregister(&rc->cdev);
+		dev_printk(KERN_ERR, rc->dev.parent, "RAID COMPONENT REMOVE\n");
+		device_unregister(&rc->dev);
 	}
 	dev_printk(KERN_ERR, dev, "RAID REMOVE DONE\n");
 	kfree(rd);
@@ -171,9 +171,11 @@ static const char *raid_level_name(enum raid_level level)
 }
 
 #define raid_attr_show_internal(attr, fmt, var, code)			\
-static ssize_t raid_show_##attr(struct class_device *cdev, char *buf)	\
+static ssize_t raid_show_##attr(struct device *dev, 			\
+				struct device_attribute *attr, 		\
+				char *buf)				\
 {									\
-	struct raid_data *rd = class_get_devdata(cdev);			\
+	struct raid_data *rd = dev_get_drvdata(dev);			\
 	code								\
 	return snprintf(buf, 20, #fmt "\n", var);			\
 }
@@ -184,17 +186,17 @@ raid_attr_show_internal(attr, %s, name,					\
 	code								\
 	name = raid_##states##_name(rd->attr);				\
 )									\
-static CLASS_DEVICE_ATTR(attr, S_IRUGO, raid_show_##attr, NULL)
+static DEVICE_ATTR(attr, S_IRUGO, raid_show_##attr, NULL)
 
 
 #define raid_attr_ro_internal(attr, code)				\
 raid_attr_show_internal(attr, %d, rd->attr, code)			\
-static CLASS_DEVICE_ATTR(attr, S_IRUGO, raid_show_##attr, NULL)
+static DEVICE_ATTR(attr, S_IRUGO, raid_show_##attr, NULL)
 
 #define ATTR_CODE(attr)							\
-	struct raid_internal *i = class_device_to_raid_internal(cdev);	\
+	struct raid_internal *i = device_to_raid_internal(dev);		\
 	if (i->f->get_##attr)						\
-		i->f->get_##attr(cdev->dev);
+		i->f->get_##attr(dev->parent);
 
 #define raid_attr_ro(attr)	raid_attr_ro_internal(attr, )
 #define raid_attr_ro_fn(attr)	raid_attr_ro_internal(attr, ATTR_CODE(attr))
@@ -206,23 +208,23 @@ raid_attr_ro_state(level);
 raid_attr_ro_fn(resync);
 raid_attr_ro_state_fn(state);
 
-static void raid_component_release(struct class_device *cdev)
+static void raid_component_release(struct device *dev)
 {
-	struct raid_component *rc = container_of(cdev, struct raid_component,
-						 cdev);
-	dev_printk(KERN_ERR, rc->cdev.dev, "COMPONENT RELEASE\n");
-	put_device(rc->cdev.dev);
+	struct raid_component *rc =
+		container_of(dev, struct raid_component, dev);
+	dev_printk(KERN_ERR, rc->dev.parent, "COMPONENT RELEASE\n");
+	put_device(rc->dev.parent);
 	kfree(rc);
 }
 
 int raid_component_add(struct raid_template *r,struct device *raid_dev,
 		       struct device *component_dev)
 {
-	struct class_device *cdev =
+	struct device *cdev =
 		attribute_container_find_class_device(&r->raid_attrs.ac,
 						      raid_dev);
 	struct raid_component *rc;
-	struct raid_data *rd = class_get_devdata(cdev);
+	struct raid_data *rd = dev_get_drvdata(cdev);
 	int err;
 
 	rc = kzalloc(sizeof(*rc), GFP_KERNEL);
@@ -230,17 +232,16 @@ int raid_component_add(struct raid_template *r,struct device *raid_dev,
 		return -ENOMEM;
 
 	INIT_LIST_HEAD(&rc->node);
-	class_device_initialize(&rc->cdev);
-	rc->cdev.release = raid_component_release;
-	rc->cdev.dev = get_device(component_dev);
+	device_initialize(&rc->dev);
+	rc->dev.release = raid_component_release;
+	rc->dev.parent = get_device(component_dev);
 	rc->num = rd->component_count++;
 
-	snprintf(rc->cdev.class_id, sizeof(rc->cdev.class_id),
+	snprintf(rc->dev.bus_id, sizeof(rc->dev.bus_id),
 		 "component-%d", rc->num);
 	list_add_tail(&rc->node, &rd->component_list);
-	rc->cdev.parent = cdev;
-	rc->cdev.class = &raid_class.class;
-	err = class_device_add(&rc->cdev);
+	rc->dev.class = &raid_class.class;
+	err = device_add(&rc->dev);
 	if (err)
 		goto err_out;
 
@@ -273,9 +274,9 @@ raid_class_attach(struct raid_function_template *ft)
 
 	attribute_container_register(&i->r.raid_attrs.ac);
 
-	i->attrs[count++] = &class_device_attr_level;
-	i->attrs[count++] = &class_device_attr_resync;
-	i->attrs[count++] = &class_device_attr_state;
+	i->attrs[count++] = &dev_attr_level;
+	i->attrs[count++] = &dev_attr_resync;
+	i->attrs[count++] = &dev_attr_state;
 
 	i->attrs[count] = NULL;
 	BUG_ON(count > RAID_NUM_ATTRS);
diff --git a/drivers/scsi/scsi_sas_internal.h b/drivers/scsi/scsi_sas_internal.h
index e1edab45a37b..998cb5be6833 100644
--- a/drivers/scsi/scsi_sas_internal.h
+++ b/drivers/scsi/scsi_sas_internal.h
@@ -13,12 +13,12 @@ struct sas_internal {
 	struct sas_function_template *f;
 	struct sas_domain_function_template *dft;
 
-	struct class_device_attribute private_host_attrs[SAS_HOST_ATTRS];
-	struct class_device_attribute private_phy_attrs[SAS_PHY_ATTRS];
-	struct class_device_attribute private_port_attrs[SAS_PORT_ATTRS];
-	struct class_device_attribute private_rphy_attrs[SAS_RPORT_ATTRS];
-	struct class_device_attribute private_end_dev_attrs[SAS_END_DEV_ATTRS];
-	struct class_device_attribute private_expander_attrs[SAS_EXPANDER_ATTRS];
+	struct device_attribute private_host_attrs[SAS_HOST_ATTRS];
+	struct device_attribute private_phy_attrs[SAS_PHY_ATTRS];
+	struct device_attribute private_port_attrs[SAS_PORT_ATTRS];
+	struct device_attribute private_rphy_attrs[SAS_RPORT_ATTRS];
+	struct device_attribute private_end_dev_attrs[SAS_END_DEV_ATTRS];
+	struct device_attribute private_expander_attrs[SAS_EXPANDER_ATTRS];
 
 	struct transport_container phy_attr_cont;
 	struct transport_container port_attr_cont;
@@ -30,12 +30,12 @@ struct sas_internal {
 	 * The array of null terminated pointers to attributes
 	 * needed by scsi_sysfs.c
 	 */
-	struct class_device_attribute *host_attrs[SAS_HOST_ATTRS + 1];
-	struct class_device_attribute *phy_attrs[SAS_PHY_ATTRS + 1];
-	struct class_device_attribute *port_attrs[SAS_PORT_ATTRS + 1];
-	struct class_device_attribute *rphy_attrs[SAS_RPORT_ATTRS + 1];
-	struct class_device_attribute *end_dev_attrs[SAS_END_DEV_ATTRS + 1];
-	struct class_device_attribute *expander_attrs[SAS_EXPANDER_ATTRS + 1];
+	struct device_attribute *host_attrs[SAS_HOST_ATTRS + 1];
+	struct device_attribute *phy_attrs[SAS_PHY_ATTRS + 1];
+	struct device_attribute *port_attrs[SAS_PORT_ATTRS + 1];
+	struct device_attribute *rphy_attrs[SAS_RPORT_ATTRS + 1];
+	struct device_attribute *end_dev_attrs[SAS_END_DEV_ATTRS + 1];
+	struct device_attribute *expander_attrs[SAS_EXPANDER_ATTRS + 1];
 };
 #define to_sas_internal(tmpl)	container_of(tmpl, struct sas_internal, t)
 
diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c
index ed83cdb6e67d..67bb20ed45d2 100644
--- a/drivers/scsi/scsi_sysfs.c
+++ b/drivers/scsi/scsi_sysfs.c
@@ -119,9 +119,10 @@ static int scsi_scan(struct Scsi_Host *shost, const char *str)
  */
 #define shost_show_function(name, field, format_string)			\
 static ssize_t								\
-show_##name (struct class_device *class_dev, char *buf)			\
+show_##name (struct device *dev, struct device_attribute *attr, 	\
+	     char *buf)							\
 {									\
-	struct Scsi_Host *shost = class_to_shost(class_dev);		\
+	struct Scsi_Host *shost = class_to_shost(dev);			\
 	return snprintf (buf, 20, format_string, shost->field);		\
 }
 
@@ -131,7 +132,7 @@ show_##name (struct class_device *class_dev, char *buf)			\
  */
 #define shost_rd_attr2(name, field, format_string)			\
 	shost_show_function(name, field, format_string)			\
-static CLASS_DEVICE_ATTR(name, S_IRUGO, show_##name, NULL);
+static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL);
 
 #define shost_rd_attr(field, format_string) \
 shost_rd_attr2(field, field, format_string)
@@ -140,10 +141,11 @@ shost_rd_attr2(field, field, format_string)
  * Create the actual show/store functions and data structures.
  */
 
-static ssize_t store_scan(struct class_device *class_dev, const char *buf,
-			  size_t count)
+static ssize_t
+store_scan(struct device *dev, struct device_attribute *attr,
+	   const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	int res;
 
 	res = scsi_scan(shost, buf);
@@ -151,13 +153,14 @@ static ssize_t store_scan(struct class_device *class_dev, const char *buf,
 		res = count;
 	return res;
 };
-static CLASS_DEVICE_ATTR(scan, S_IWUSR, NULL, store_scan);
+static DEVICE_ATTR(scan, S_IWUSR, NULL, store_scan);
 
 static ssize_t
-store_shost_state(struct class_device *class_dev, const char *buf, size_t count)
+store_shost_state(struct device *dev, struct device_attribute *attr,
+		  const char *buf, size_t count)
 {
 	int i;
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	enum scsi_host_state state = 0;
 
 	for (i = 0; i < ARRAY_SIZE(shost_states); i++) {
@@ -177,9 +180,9 @@ store_shost_state(struct class_device *class_dev, const char *buf, size_t count)
 }
 
 static ssize_t
-show_shost_state(struct class_device *class_dev, char *buf)
+show_shost_state(struct device *dev, struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	const char *name = scsi_host_state_name(shost->shost_state);
 
 	if (!name)
@@ -188,7 +191,9 @@ show_shost_state(struct class_device *class_dev, char *buf)
 	return snprintf(buf, 20, "%s\n", name);
 }
 
-static CLASS_DEVICE_ATTR(state, S_IRUGO | S_IWUSR, show_shost_state, store_shost_state);
+/* DEVICE_ATTR(state) clashes with dev_attr_state for sdev */
+struct device_attribute dev_attr_hstate =
+	__ATTR(state, S_IRUGO | S_IWUSR, show_shost_state, store_shost_state);
 
 static ssize_t
 show_shost_mode(unsigned int mode, char *buf)
@@ -206,9 +211,11 @@ show_shost_mode(unsigned int mode, char *buf)
 	return len;
 }
 
-static ssize_t show_shost_supported_mode(struct class_device *class_dev, char *buf)
+static ssize_t
+show_shost_supported_mode(struct device *dev, struct device_attribute *attr,
+			  char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 	unsigned int supported_mode = shost->hostt->supported_mode;
 
 	if (supported_mode == MODE_UNKNOWN)
@@ -218,11 +225,13 @@ static ssize_t show_shost_supported_mode(struct class_device *class_dev, char *b
 	return show_shost_mode(supported_mode, buf);
 }
 
-static CLASS_DEVICE_ATTR(supported_mode, S_IRUGO | S_IWUSR, show_shost_supported_mode, NULL);
+static DEVICE_ATTR(supported_mode, S_IRUGO | S_IWUSR, show_shost_supported_mode, NULL);
 
-static ssize_t show_shost_active_mode(struct class_device *class_dev, char *buf)
+static ssize_t
+show_shost_active_mode(struct device *dev,
+		       struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = class_to_shost(class_dev);
+	struct Scsi_Host *shost = class_to_shost(dev);
 
 	if (shost->active_mode == MODE_UNKNOWN)
 		return snprintf(buf, 20, "unknown\n");
@@ -230,7 +239,7 @@ static ssize_t show_shost_active_mode(struct class_device *class_dev, char *buf)
 		return show_shost_mode(shost->active_mode, buf);
 }
 
-static CLASS_DEVICE_ATTR(active_mode, S_IRUGO | S_IWUSR, show_shost_active_mode, NULL);
+static DEVICE_ATTR(active_mode, S_IRUGO | S_IWUSR, show_shost_active_mode, NULL);
 
 shost_rd_attr(unique_id, "%u\n");
 shost_rd_attr(host_busy, "%hu\n");
@@ -240,22 +249,22 @@ shost_rd_attr(sg_tablesize, "%hu\n");
 shost_rd_attr(unchecked_isa_dma, "%d\n");
 shost_rd_attr2(proc_name, hostt->proc_name, "%s\n");
 
-static struct class_device_attribute *scsi_sysfs_shost_attrs[] = {
-	&class_device_attr_unique_id,
-	&class_device_attr_host_busy,
-	&class_device_attr_cmd_per_lun,
-	&class_device_attr_can_queue,
-	&class_device_attr_sg_tablesize,
-	&class_device_attr_unchecked_isa_dma,
-	&class_device_attr_proc_name,
-	&class_device_attr_scan,
-	&class_device_attr_state,
-	&class_device_attr_supported_mode,
-	&class_device_attr_active_mode,
+static struct device_attribute *scsi_sysfs_shost_attrs[] = {
+	&dev_attr_unique_id,
+	&dev_attr_host_busy,
+	&dev_attr_cmd_per_lun,
+	&dev_attr_can_queue,
+	&dev_attr_sg_tablesize,
+	&dev_attr_unchecked_isa_dma,
+	&dev_attr_proc_name,
+	&dev_attr_scan,
+	&dev_attr_hstate,
+	&dev_attr_supported_mode,
+	&dev_attr_active_mode,
 	NULL
 };
 
-static void scsi_device_cls_release(struct class_device *class_dev)
+static void scsi_device_cls_release(struct device *class_dev)
 {
 	struct scsi_device *sdev;
 
@@ -320,7 +329,7 @@ static void scsi_device_dev_release(struct device *dev)
 
 static struct class sdev_class = {
 	.name		= "scsi_device",
-	.release	= scsi_device_cls_release,
+	.dev_release	= scsi_device_cls_release,
 };
 
 /* all probing is done in the individual ->probe routines */
@@ -424,7 +433,8 @@ void scsi_sysfs_unregister(void)
  */
 #define sdev_show_function(field, format_string)				\
 static ssize_t								\
-sdev_show_##field (struct device *dev, struct device_attribute *attr, char *buf)				\
+sdev_show_##field (struct device *dev, struct device_attribute *attr,	\
+		   char *buf)						\
 {									\
 	struct scsi_device *sdev;					\
 	sdev = to_scsi_device(dev);					\
@@ -448,7 +458,8 @@ static DEVICE_ATTR(field, S_IRUGO, sdev_show_##field, NULL);
 	sdev_show_function(field, format_string)				\
 									\
 static ssize_t								\
-sdev_store_##field (struct device *dev, struct device_attribute *attr, const char *buf, size_t count)	\
+sdev_store_##field (struct device *dev, struct device_attribute *attr,	\
+		    const char *buf, size_t count)			\
 {									\
 	struct scsi_device *sdev;					\
 	sdev = to_scsi_device(dev);					\
@@ -468,7 +479,8 @@ static DEVICE_ATTR(field, S_IRUGO | S_IWUSR, sdev_show_##field, sdev_store_##fie
 	sdev_show_function(field, "%d\n")					\
 									\
 static ssize_t								\
-sdev_store_##field (struct device *dev, struct device_attribute *attr, const char *buf, size_t count)	\
+sdev_store_##field (struct device *dev, struct device_attribute *attr,	\
+		    const char *buf, size_t count)			\
 {									\
 	int ret;							\
 	struct scsi_device *sdev;					\
@@ -519,7 +531,8 @@ sdev_show_timeout (struct device *dev, struct device_attribute *attr, char *buf)
 }
 
 static ssize_t
-sdev_store_timeout (struct device *dev, struct device_attribute *attr, const char *buf, size_t count)
+sdev_store_timeout (struct device *dev, struct device_attribute *attr,
+		    const char *buf, size_t count)
 {
 	struct scsi_device *sdev;
 	int timeout;
@@ -531,7 +544,8 @@ sdev_store_timeout (struct device *dev, struct device_attribute *attr, const cha
 static DEVICE_ATTR(timeout, S_IRUGO | S_IWUSR, sdev_show_timeout, sdev_store_timeout);
 
 static ssize_t
-store_rescan_field (struct device *dev, struct device_attribute *attr, const char *buf, size_t count)
+store_rescan_field (struct device *dev, struct device_attribute *attr,
+		    const char *buf, size_t count)
 {
 	scsi_rescan_device(dev);
 	return count;
@@ -543,8 +557,9 @@ static void sdev_store_delete_callback(struct device *dev)
 	scsi_remove_device(to_scsi_device(dev));
 }
 
-static ssize_t sdev_store_delete(struct device *dev, struct device_attribute *attr, const char *buf,
-				 size_t count)
+static ssize_t
+sdev_store_delete(struct device *dev, struct device_attribute *attr,
+		  const char *buf, size_t count)
 {
 	int rc;
 
@@ -559,7 +574,8 @@ static ssize_t sdev_store_delete(struct device *dev, struct device_attribute *at
 static DEVICE_ATTR(delete, S_IWUSR, NULL, sdev_store_delete);
 
 static ssize_t
-store_state_field(struct device *dev, struct device_attribute *attr, const char *buf, size_t count)
+store_state_field(struct device *dev, struct device_attribute *attr,
+		  const char *buf, size_t count)
 {
 	int i;
 	struct scsi_device *sdev = to_scsi_device(dev);
@@ -596,7 +612,8 @@ show_state_field(struct device *dev, struct device_attribute *attr, char *buf)
 static DEVICE_ATTR(state, S_IRUGO | S_IWUSR, show_state_field, store_state_field);
 
 static ssize_t
-show_queue_type_field(struct device *dev, struct device_attribute *attr, char *buf)
+show_queue_type_field(struct device *dev, struct device_attribute *attr,
+		      char *buf)
 {
 	struct scsi_device *sdev = to_scsi_device(dev);
 	const char *name = "none";
@@ -612,7 +629,7 @@ show_queue_type_field(struct device *dev, struct device_attribute *attr, char *b
 static DEVICE_ATTR(queue_type, S_IRUGO, show_queue_type_field, NULL);
 
 static ssize_t
-show_iostat_counterbits(struct device *dev, struct device_attribute *attr, char *buf)
+show_iostat_counterbits(struct device *dev, struct device_attribute *attr, 				char *buf)
 {
 	return snprintf(buf, 20, "%d\n", (int)sizeof(atomic_t) * 8);
 }
@@ -621,7 +638,8 @@ static DEVICE_ATTR(iocounterbits, S_IRUGO, show_iostat_counterbits, NULL);
 
 #define show_sdev_iostat(field)						\
 static ssize_t								\
-show_iostat_##field(struct device *dev, struct device_attribute *attr, char *buf)			\
+show_iostat_##field(struct device *dev, struct device_attribute *attr,	\
+		    char *buf)						\
 {									\
 	struct scsi_device *sdev = to_scsi_device(dev);			\
 	unsigned long long count = atomic_read(&sdev->field);		\
@@ -645,7 +663,7 @@ static DEVICE_ATTR(modalias, S_IRUGO, sdev_show_modalias, NULL);
 #define DECLARE_EVT_SHOW(name, Cap_name)				\
 static ssize_t								\
 sdev_show_evt_##name(struct device *dev, struct device_attribute *attr,	\
-				char *buf)				\
+		     char *buf)						\
 {									\
 	struct scsi_device *sdev = to_scsi_device(dev);			\
 	int val = test_bit(SDEV_EVT_##Cap_name, sdev->supported_events);\
@@ -654,7 +672,7 @@ sdev_show_evt_##name(struct device *dev, struct device_attribute *attr,	\
 
 #define DECLARE_EVT_STORE(name, Cap_name)				\
 static ssize_t								\
-sdev_store_evt_##name(struct device *dev, struct device_attribute *attr, \
+sdev_store_evt_##name(struct device *dev, struct device_attribute *attr,\
 		      const char *buf, size_t count)			\
 {									\
 	struct scsi_device *sdev = to_scsi_device(dev);			\
@@ -707,8 +725,9 @@ static struct attribute_group *scsi_sdev_attr_groups[] = {
 	NULL
 };
 
-static ssize_t sdev_store_queue_depth_rw(struct device *dev, struct device_attribute *attr, const char *buf,
-					 size_t count)
+static ssize_t
+sdev_store_queue_depth_rw(struct device *dev, struct device_attribute *attr,
+			  const char *buf, size_t count)
 {
 	int depth, retval;
 	struct scsi_device *sdev = to_scsi_device(dev);
@@ -733,8 +752,9 @@ static struct device_attribute sdev_attr_queue_depth_rw =
 	__ATTR(queue_depth, S_IRUGO | S_IWUSR, sdev_show_queue_depth,
 	       sdev_store_queue_depth_rw);
 
-static ssize_t sdev_store_queue_type_rw(struct device *dev, struct device_attribute *attr, const char *buf,
-					size_t count)
+static ssize_t
+sdev_store_queue_type_rw(struct device *dev, struct device_attribute *attr,
+			 const char *buf, size_t count)
 {
 	struct scsi_device *sdev = to_scsi_device(dev);
 	struct scsi_host_template *sht = sdev->host->hostt;
@@ -786,13 +806,13 @@ int scsi_sysfs_add_sdev(struct scsi_device *sdev)
 		printk(KERN_INFO "error 1\n");
 		return error;
 	}
-	error = class_device_add(&sdev->sdev_classdev);
+	error = device_add(&sdev->sdev_dev);
 	if (error) {
 		printk(KERN_INFO "error 2\n");
 		goto clean_device;
 	}
 
-	/* take a reference for the sdev_classdev; this is
+	/* take a reference for the sdev_dev; this is
 	 * released by the sdev_class .release */
 	get_device(&sdev->sdev_gendev);
 
@@ -858,7 +878,7 @@ void __scsi_remove_device(struct scsi_device *sdev)
 		return;
 
 	bsg_unregister_queue(sdev->request_queue);
-	class_device_unregister(&sdev->sdev_classdev);
+	device_unregister(&sdev->sdev_dev);
 	transport_remove_device(dev);
 	device_del(dev);
 	scsi_device_set_state(sdev, SDEV_DEL);
@@ -952,9 +972,9 @@ int scsi_register_interface(struct class_interface *intf)
 EXPORT_SYMBOL(scsi_register_interface);
 
 
-static struct class_device_attribute *class_attr_overridden(
-		struct class_device_attribute **attrs,
-		struct class_device_attribute *attr)
+static struct device_attribute *class_attr_overridden(
+		struct device_attribute **attrs,
+		struct device_attribute *attr)
 {
 	int i;
 
@@ -966,10 +986,10 @@ static struct class_device_attribute *class_attr_overridden(
 	return NULL;
 }
 
-static int class_attr_add(struct class_device *classdev,
-		struct class_device_attribute *attr)
+static int class_attr_add(struct device *classdev,
+		struct device_attribute *attr)
 {
-	struct class_device_attribute *base_attr;
+	struct device_attribute *base_attr;
 
 	/*
 	 * Spare the caller from having to copy things it's not interested in.
@@ -986,7 +1006,7 @@ static int class_attr_add(struct class_device *classdev,
 			attr->store = base_attr->store;
 	}
 
-	return class_device_create_file(classdev, attr);
+	return device_create_file(classdev, attr);
 }
 
 /**
@@ -1000,7 +1020,7 @@ int scsi_sysfs_add_host(struct Scsi_Host *shost)
 
 	if (shost->hostt->shost_attrs) {
 		for (i = 0; shost->hostt->shost_attrs[i]; i++) {
-			error = class_attr_add(&shost->shost_classdev,
+			error = class_attr_add(&shost->shost_dev,
 					shost->hostt->shost_attrs[i]);
 			if (error)
 				return error;
@@ -1010,7 +1030,7 @@ int scsi_sysfs_add_host(struct Scsi_Host *shost)
 	for (i = 0; scsi_sysfs_shost_attrs[i]; i++) {
 		if (!class_attr_overridden(shost->hostt->shost_attrs,
 					scsi_sysfs_shost_attrs[i])) {
-			error = class_device_create_file(&shost->shost_classdev,
+			error = device_create_file(&shost->shost_dev,
 					scsi_sysfs_shost_attrs[i]);
 			if (error)
 				return error;
@@ -1041,10 +1061,10 @@ void scsi_sysfs_device_initialize(struct scsi_device *sdev)
 		sdev->host->host_no, sdev->channel, sdev->id,
 		sdev->lun);
 	
-	class_device_initialize(&sdev->sdev_classdev);
-	sdev->sdev_classdev.dev = &sdev->sdev_gendev;
-	sdev->sdev_classdev.class = &sdev_class;
-	snprintf(sdev->sdev_classdev.class_id, BUS_ID_SIZE,
+	device_initialize(&sdev->sdev_dev);
+	sdev->sdev_dev.parent = &sdev->sdev_gendev;
+	sdev->sdev_dev.class = &sdev_class;
+	snprintf(sdev->sdev_dev.bus_id, BUS_ID_SIZE,
 		 "%d:%d:%d:%d", sdev->host->host_no,
 		 sdev->channel, sdev->id, sdev->lun);
 	sdev->scsi_level = starget->scsi_level;
diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index b1119da6e88c..6b092a6c295d 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -72,8 +72,8 @@ static int fc_vport_create(struct Scsi_Host *shost, int channel,
  * Redefine so that we can have same named attributes in the
  * sdev/starget/host objects.
  */
-#define FC_CLASS_DEVICE_ATTR(_prefix,_name,_mode,_show,_store)		\
-struct class_device_attribute class_device_attr_##_prefix##_##_name = 	\
+#define FC_DEVICE_ATTR(_prefix,_name,_mode,_show,_store)		\
+struct device_attribute device_attr_##_prefix##_##_name = 	\
 	__ATTR(_name,_mode,_show,_store)
 
 #define fc_enum_name_search(title, table_type, table)			\
@@ -326,26 +326,26 @@ struct fc_internal {
 	 * part of the midlayer. As the remote port is specific to the
 	 * fc transport, we must provide the attribute container.
 	 */
-	struct class_device_attribute private_starget_attrs[
+	struct device_attribute private_starget_attrs[
 							FC_STARGET_NUM_ATTRS];
-	struct class_device_attribute *starget_attrs[FC_STARGET_NUM_ATTRS + 1];
+	struct device_attribute *starget_attrs[FC_STARGET_NUM_ATTRS + 1];
 
-	struct class_device_attribute private_host_attrs[FC_HOST_NUM_ATTRS];
-	struct class_device_attribute *host_attrs[FC_HOST_NUM_ATTRS + 1];
+	struct device_attribute private_host_attrs[FC_HOST_NUM_ATTRS];
+	struct device_attribute *host_attrs[FC_HOST_NUM_ATTRS + 1];
 
 	struct transport_container rport_attr_cont;
-	struct class_device_attribute private_rport_attrs[FC_RPORT_NUM_ATTRS];
-	struct class_device_attribute *rport_attrs[FC_RPORT_NUM_ATTRS + 1];
+	struct device_attribute private_rport_attrs[FC_RPORT_NUM_ATTRS];
+	struct device_attribute *rport_attrs[FC_RPORT_NUM_ATTRS + 1];
 
 	struct transport_container vport_attr_cont;
-	struct class_device_attribute private_vport_attrs[FC_VPORT_NUM_ATTRS];
-	struct class_device_attribute *vport_attrs[FC_VPORT_NUM_ATTRS + 1];
+	struct device_attribute private_vport_attrs[FC_VPORT_NUM_ATTRS];
+	struct device_attribute *vport_attrs[FC_VPORT_NUM_ATTRS + 1];
 };
 
 #define to_fc_internal(tmpl)	container_of(tmpl, struct fc_internal, t)
 
 static int fc_target_setup(struct transport_container *tc, struct device *dev,
-			   struct class_device *cdev)
+			   struct device *cdev)
 {
 	struct scsi_target *starget = to_scsi_target(dev);
 	struct fc_rport *rport = starget_to_rport(starget);
@@ -375,7 +375,7 @@ static DECLARE_TRANSPORT_CLASS(fc_transport_class,
 			       NULL);
 
 static int fc_host_setup(struct transport_container *tc, struct device *dev,
-			 struct class_device *cdev)
+			 struct device *cdev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 	struct fc_host_attrs *fc_host = shost_to_fc_host(shost);
@@ -682,9 +682,10 @@ static void __exit fc_transport_exit(void)
 
 #define fc_rport_show_function(field, format_string, sz, cast)		\
 static ssize_t								\
-show_fc_rport_##field (struct class_device *cdev, char *buf)		\
+show_fc_rport_##field (struct device *dev, 				\
+		       struct device_attribute *attr, char *buf)	\
 {									\
-	struct fc_rport *rport = transport_class_to_rport(cdev);	\
+	struct fc_rport *rport = transport_class_to_rport(dev);		\
 	struct Scsi_Host *shost = rport_to_shost(rport);		\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	if ((i->f->get_rport_##field) &&				\
@@ -697,11 +698,12 @@ show_fc_rport_##field (struct class_device *cdev, char *buf)		\
 
 #define fc_rport_store_function(field)					\
 static ssize_t								\
-store_fc_rport_##field(struct class_device *cdev, const char *buf,	\
-			   size_t count)				\
+store_fc_rport_##field(struct device *dev,				\
+		       struct device_attribute *attr,			\
+		       const char *buf,	size_t count)			\
 {									\
 	int val;							\
-	struct fc_rport *rport = transport_class_to_rport(cdev);	\
+	struct fc_rport *rport = transport_class_to_rport(dev);		\
 	struct Scsi_Host *shost = rport_to_shost(rport);		\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	char *cp;							\
@@ -718,58 +720,60 @@ store_fc_rport_##field(struct class_device *cdev, const char *buf,	\
 
 #define fc_rport_rd_attr(field, format_string, sz)			\
 	fc_rport_show_function(field, format_string, sz, )		\
-static FC_CLASS_DEVICE_ATTR(rport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(rport, field, S_IRUGO,			\
 			 show_fc_rport_##field, NULL)
 
 #define fc_rport_rd_attr_cast(field, format_string, sz, cast)		\
 	fc_rport_show_function(field, format_string, sz, (cast))	\
-static FC_CLASS_DEVICE_ATTR(rport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(rport, field, S_IRUGO,			\
 			  show_fc_rport_##field, NULL)
 
 #define fc_rport_rw_attr(field, format_string, sz)			\
 	fc_rport_show_function(field, format_string, sz, )		\
 	fc_rport_store_function(field)					\
-static FC_CLASS_DEVICE_ATTR(rport, field, S_IRUGO | S_IWUSR,		\
+static FC_DEVICE_ATTR(rport, field, S_IRUGO | S_IWUSR,		\
 			show_fc_rport_##field,				\
 			store_fc_rport_##field)
 
 
 #define fc_private_rport_show_function(field, format_string, sz, cast)	\
 static ssize_t								\
-show_fc_rport_##field (struct class_device *cdev, char *buf)		\
+show_fc_rport_##field (struct device *dev, 				\
+		       struct device_attribute *attr, char *buf)	\
 {									\
-	struct fc_rport *rport = transport_class_to_rport(cdev);	\
+	struct fc_rport *rport = transport_class_to_rport(dev);		\
 	return snprintf(buf, sz, format_string, cast rport->field); 	\
 }
 
 #define fc_private_rport_rd_attr(field, format_string, sz)		\
 	fc_private_rport_show_function(field, format_string, sz, )	\
-static FC_CLASS_DEVICE_ATTR(rport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(rport, field, S_IRUGO,			\
 			 show_fc_rport_##field, NULL)
 
 #define fc_private_rport_rd_attr_cast(field, format_string, sz, cast)	\
 	fc_private_rport_show_function(field, format_string, sz, (cast)) \
-static FC_CLASS_DEVICE_ATTR(rport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(rport, field, S_IRUGO,			\
 			  show_fc_rport_##field, NULL)
 
 
 #define fc_private_rport_rd_enum_attr(title, maxlen)			\
 static ssize_t								\
-show_fc_rport_##title (struct class_device *cdev, char *buf)		\
+show_fc_rport_##title (struct device *dev,				\
+		       struct device_attribute *attr, char *buf)	\
 {									\
-	struct fc_rport *rport = transport_class_to_rport(cdev);	\
+	struct fc_rport *rport = transport_class_to_rport(dev);		\
 	const char *name;						\
 	name = get_fc_##title##_name(rport->title);			\
 	if (!name)							\
 		return -EINVAL;						\
 	return snprintf(buf, maxlen, "%s\n", name);			\
 }									\
-static FC_CLASS_DEVICE_ATTR(rport, title, S_IRUGO,			\
+static FC_DEVICE_ATTR(rport, title, S_IRUGO,			\
 			show_fc_rport_##title, NULL)
 
 
 #define SETUP_RPORT_ATTRIBUTE_RD(field)					\
-	i->private_rport_attrs[count] = class_device_attr_rport_##field; \
+	i->private_rport_attrs[count] = device_attr_rport_##field; \
 	i->private_rport_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_rport_attrs[count].store = NULL;			\
 	i->rport_attrs[count] = &i->private_rport_attrs[count];		\
@@ -777,14 +781,14 @@ static FC_CLASS_DEVICE_ATTR(rport, title, S_IRUGO,			\
 		count++
 
 #define SETUP_PRIVATE_RPORT_ATTRIBUTE_RD(field)				\
-	i->private_rport_attrs[count] = class_device_attr_rport_##field; \
+	i->private_rport_attrs[count] = device_attr_rport_##field; \
 	i->private_rport_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_rport_attrs[count].store = NULL;			\
 	i->rport_attrs[count] = &i->private_rport_attrs[count];		\
 	count++
 
 #define SETUP_RPORT_ATTRIBUTE_RW(field)					\
-	i->private_rport_attrs[count] = class_device_attr_rport_##field; \
+	i->private_rport_attrs[count] = device_attr_rport_##field; \
 	if (!i->f->set_rport_##field) {					\
 		i->private_rport_attrs[count].attr.mode = S_IRUGO;	\
 		i->private_rport_attrs[count].store = NULL;		\
@@ -795,7 +799,7 @@ static FC_CLASS_DEVICE_ATTR(rport, title, S_IRUGO,			\
 
 #define SETUP_PRIVATE_RPORT_ATTRIBUTE_RW(field)				\
 {									\
-	i->private_rport_attrs[count] = class_device_attr_rport_##field; \
+	i->private_rport_attrs[count] = device_attr_rport_##field; \
 	i->rport_attrs[count] = &i->private_rport_attrs[count];		\
 	count++;							\
 }
@@ -808,14 +812,15 @@ static FC_CLASS_DEVICE_ATTR(rport, title, S_IRUGO,			\
 fc_private_rport_rd_attr(maxframe_size, "%u bytes\n", 20);
 
 static ssize_t
-show_fc_rport_supported_classes (struct class_device *cdev, char *buf)
+show_fc_rport_supported_classes (struct device *dev,
+				 struct device_attribute *attr, char *buf)
 {
-	struct fc_rport *rport = transport_class_to_rport(cdev);
+	struct fc_rport *rport = transport_class_to_rport(dev);
 	if (rport->supported_classes == FC_COS_UNSPECIFIED)
 		return snprintf(buf, 20, "unspecified\n");
 	return get_fc_cos_names(rport->supported_classes, buf);
 }
-static FC_CLASS_DEVICE_ATTR(rport, supported_classes, S_IRUGO,
+static FC_DEVICE_ATTR(rport, supported_classes, S_IRUGO,
 		show_fc_rport_supported_classes, NULL);
 
 /* Dynamic Remote Port Attributes */
@@ -825,11 +830,11 @@ static FC_CLASS_DEVICE_ATTR(rport, supported_classes, S_IRUGO,
  */
 fc_rport_show_function(dev_loss_tmo, "%d\n", 20, )
 static ssize_t
-store_fc_rport_dev_loss_tmo(struct class_device *cdev, const char *buf,
-			   size_t count)
+store_fc_rport_dev_loss_tmo(struct device *dev, struct device_attribute *attr,
+			    const char *buf, size_t count)
 {
 	int val;
-	struct fc_rport *rport = transport_class_to_rport(cdev);
+	struct fc_rport *rport = transport_class_to_rport(dev);
 	struct Scsi_Host *shost = rport_to_shost(rport);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
 	char *cp;
@@ -844,7 +849,7 @@ store_fc_rport_dev_loss_tmo(struct class_device *cdev, const char *buf,
 	i->f->set_rport_dev_loss_tmo(rport, val);
 	return count;
 }
-static FC_CLASS_DEVICE_ATTR(rport, dev_loss_tmo, S_IRUGO | S_IWUSR,
+static FC_DEVICE_ATTR(rport, dev_loss_tmo, S_IRUGO | S_IWUSR,
 		show_fc_rport_dev_loss_tmo, store_fc_rport_dev_loss_tmo);
 
 
@@ -855,9 +860,10 @@ fc_private_rport_rd_attr_cast(port_name, "0x%llx\n", 20, unsigned long long);
 fc_private_rport_rd_attr(port_id, "0x%06x\n", 20);
 
 static ssize_t
-show_fc_rport_roles (struct class_device *cdev, char *buf)
+show_fc_rport_roles (struct device *dev, struct device_attribute *attr,
+		     char *buf)
 {
-	struct fc_rport *rport = transport_class_to_rport(cdev);
+	struct fc_rport *rport = transport_class_to_rport(dev);
 
 	/* identify any roles that are port_id specific */
 	if ((rport->port_id != -1) &&
@@ -883,7 +889,7 @@ show_fc_rport_roles (struct class_device *cdev, char *buf)
 		return get_fc_port_roles_names(rport->roles, buf);
 	}
 }
-static FC_CLASS_DEVICE_ATTR(rport, roles, S_IRUGO,
+static FC_DEVICE_ATTR(rport, roles, S_IRUGO,
 		show_fc_rport_roles, NULL);
 
 fc_private_rport_rd_enum_attr(port_state, FC_PORTSTATE_MAX_NAMELEN);
@@ -893,9 +899,10 @@ fc_private_rport_rd_attr(scsi_target_id, "%d\n", 20);
  * fast_io_fail_tmo attribute
  */
 static ssize_t
-show_fc_rport_fast_io_fail_tmo (struct class_device *cdev, char *buf)
+show_fc_rport_fast_io_fail_tmo (struct device *dev,
+				struct device_attribute *attr, char *buf)
 {
-	struct fc_rport *rport = transport_class_to_rport(cdev);
+	struct fc_rport *rport = transport_class_to_rport(dev);
 
 	if (rport->fast_io_fail_tmo == -1)
 		return snprintf(buf, 5, "off\n");
@@ -903,12 +910,13 @@ show_fc_rport_fast_io_fail_tmo (struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-store_fc_rport_fast_io_fail_tmo(struct class_device *cdev, const char *buf,
-			   size_t count)
+store_fc_rport_fast_io_fail_tmo(struct device *dev,
+				struct device_attribute *attr, const char *buf,
+				size_t count)
 {
 	int val;
 	char *cp;
-	struct fc_rport *rport = transport_class_to_rport(cdev);
+	struct fc_rport *rport = transport_class_to_rport(dev);
 
 	if ((rport->port_state == FC_PORTSTATE_BLOCKED) ||
 	    (rport->port_state == FC_PORTSTATE_DELETED) ||
@@ -925,7 +933,7 @@ store_fc_rport_fast_io_fail_tmo(struct class_device *cdev, const char *buf,
 	}
 	return count;
 }
-static FC_CLASS_DEVICE_ATTR(rport, fast_io_fail_tmo, S_IRUGO | S_IWUSR,
+static FC_DEVICE_ATTR(rport, fast_io_fail_tmo, S_IRUGO | S_IWUSR,
 	show_fc_rport_fast_io_fail_tmo, store_fc_rport_fast_io_fail_tmo);
 
 
@@ -941,9 +949,10 @@ static FC_CLASS_DEVICE_ATTR(rport, fast_io_fail_tmo, S_IRUGO | S_IWUSR,
  */
 #define fc_starget_show_function(field, format_string, sz, cast)	\
 static ssize_t								\
-show_fc_starget_##field (struct class_device *cdev, char *buf)		\
+show_fc_starget_##field (struct device *dev, 				\
+			 struct device_attribute *attr, char *buf)	\
 {									\
-	struct scsi_target *starget = transport_class_to_starget(cdev);	\
+	struct scsi_target *starget = transport_class_to_starget(dev);	\
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);	\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	struct fc_rport *rport = starget_to_rport(starget);		\
@@ -957,16 +966,16 @@ show_fc_starget_##field (struct class_device *cdev, char *buf)		\
 
 #define fc_starget_rd_attr(field, format_string, sz)			\
 	fc_starget_show_function(field, format_string, sz, )		\
-static FC_CLASS_DEVICE_ATTR(starget, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(starget, field, S_IRUGO,			\
 			 show_fc_starget_##field, NULL)
 
 #define fc_starget_rd_attr_cast(field, format_string, sz, cast)		\
 	fc_starget_show_function(field, format_string, sz, (cast))	\
-static FC_CLASS_DEVICE_ATTR(starget, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(starget, field, S_IRUGO,			\
 			  show_fc_starget_##field, NULL)
 
 #define SETUP_STARGET_ATTRIBUTE_RD(field)				\
-	i->private_starget_attrs[count] = class_device_attr_starget_##field; \
+	i->private_starget_attrs[count] = device_attr_starget_##field; \
 	i->private_starget_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_starget_attrs[count].store = NULL;			\
 	i->starget_attrs[count] = &i->private_starget_attrs[count];	\
@@ -974,7 +983,7 @@ static FC_CLASS_DEVICE_ATTR(starget, field, S_IRUGO,			\
 		count++
 
 #define SETUP_STARGET_ATTRIBUTE_RW(field)				\
-	i->private_starget_attrs[count] = class_device_attr_starget_##field; \
+	i->private_starget_attrs[count] = device_attr_starget_##field; \
 	if (!i->f->set_starget_##field) {				\
 		i->private_starget_attrs[count].attr.mode = S_IRUGO;	\
 		i->private_starget_attrs[count].store = NULL;		\
@@ -995,9 +1004,10 @@ fc_starget_rd_attr(port_id, "0x%06x\n", 20);
 
 #define fc_vport_show_function(field, format_string, sz, cast)		\
 static ssize_t								\
-show_fc_vport_##field (struct class_device *cdev, char *buf)		\
+show_fc_vport_##field (struct device *dev, 				\
+		       struct device_attribute *attr, char *buf)	\
 {									\
-	struct fc_vport *vport = transport_class_to_vport(cdev);	\
+	struct fc_vport *vport = transport_class_to_vport(dev);		\
 	struct Scsi_Host *shost = vport_to_shost(vport);		\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	if ((i->f->get_vport_##field) &&				\
@@ -1008,11 +1018,12 @@ show_fc_vport_##field (struct class_device *cdev, char *buf)		\
 
 #define fc_vport_store_function(field)					\
 static ssize_t								\
-store_fc_vport_##field(struct class_device *cdev, const char *buf,	\
-			   size_t count)				\
+store_fc_vport_##field(struct device *dev,				\
+		       struct device_attribute *attr,			\
+		       const char *buf,	size_t count)			\
 {									\
 	int val;							\
-	struct fc_vport *vport = transport_class_to_vport(cdev);	\
+	struct fc_vport *vport = transport_class_to_vport(dev);		\
 	struct Scsi_Host *shost = vport_to_shost(vport);		\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	char *cp;							\
@@ -1027,10 +1038,11 @@ store_fc_vport_##field(struct class_device *cdev, const char *buf,	\
 
 #define fc_vport_store_str_function(field, slen)			\
 static ssize_t								\
-store_fc_vport_##field(struct class_device *cdev, const char *buf,	\
-			   size_t count)				\
+store_fc_vport_##field(struct device *dev,				\
+		       struct device_attribute *attr, 			\
+		       const char *buf,	size_t count)			\
 {									\
-	struct fc_vport *vport = transport_class_to_vport(cdev);	\
+	struct fc_vport *vport = transport_class_to_vport(dev);		\
 	struct Scsi_Host *shost = vport_to_shost(vport);		\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	unsigned int cnt=count;						\
@@ -1047,36 +1059,38 @@ store_fc_vport_##field(struct class_device *cdev, const char *buf,	\
 
 #define fc_vport_rd_attr(field, format_string, sz)			\
 	fc_vport_show_function(field, format_string, sz, )		\
-static FC_CLASS_DEVICE_ATTR(vport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(vport, field, S_IRUGO,			\
 			 show_fc_vport_##field, NULL)
 
 #define fc_vport_rd_attr_cast(field, format_string, sz, cast)		\
 	fc_vport_show_function(field, format_string, sz, (cast))	\
-static FC_CLASS_DEVICE_ATTR(vport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(vport, field, S_IRUGO,			\
 			  show_fc_vport_##field, NULL)
 
 #define fc_vport_rw_attr(field, format_string, sz)			\
 	fc_vport_show_function(field, format_string, sz, )		\
 	fc_vport_store_function(field)					\
-static FC_CLASS_DEVICE_ATTR(vport, field, S_IRUGO | S_IWUSR,		\
+static FC_DEVICE_ATTR(vport, field, S_IRUGO | S_IWUSR,		\
 			show_fc_vport_##field,				\
 			store_fc_vport_##field)
 
 #define fc_private_vport_show_function(field, format_string, sz, cast)	\
 static ssize_t								\
-show_fc_vport_##field (struct class_device *cdev, char *buf)		\
+show_fc_vport_##field (struct device *dev,				\
+		       struct device_attribute *attr, char *buf)	\
 {									\
-	struct fc_vport *vport = transport_class_to_vport(cdev);	\
+	struct fc_vport *vport = transport_class_to_vport(dev);		\
 	return snprintf(buf, sz, format_string, cast vport->field); 	\
 }
 
 #define fc_private_vport_store_u32_function(field)			\
 static ssize_t								\
-store_fc_vport_##field(struct class_device *cdev, const char *buf,	\
-			   size_t count)				\
+store_fc_vport_##field(struct device *dev,				\
+		       struct device_attribute *attr,			\
+		       const char *buf,	size_t count)			\
 {									\
 	u32 val;							\
-	struct fc_vport *vport = transport_class_to_vport(cdev);	\
+	struct fc_vport *vport = transport_class_to_vport(dev);		\
 	char *cp;							\
 	if (vport->flags & (FC_VPORT_DEL | FC_VPORT_CREATING))		\
 		return -EBUSY;						\
@@ -1090,39 +1104,41 @@ store_fc_vport_##field(struct class_device *cdev, const char *buf,	\
 
 #define fc_private_vport_rd_attr(field, format_string, sz)		\
 	fc_private_vport_show_function(field, format_string, sz, )	\
-static FC_CLASS_DEVICE_ATTR(vport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(vport, field, S_IRUGO,			\
 			 show_fc_vport_##field, NULL)
 
 #define fc_private_vport_rd_attr_cast(field, format_string, sz, cast)	\
 	fc_private_vport_show_function(field, format_string, sz, (cast)) \
-static FC_CLASS_DEVICE_ATTR(vport, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(vport, field, S_IRUGO,			\
 			  show_fc_vport_##field, NULL)
 
 #define fc_private_vport_rw_u32_attr(field, format_string, sz)		\
 	fc_private_vport_show_function(field, format_string, sz, )	\
 	fc_private_vport_store_u32_function(field)			\
-static FC_CLASS_DEVICE_ATTR(vport, field, S_IRUGO | S_IWUSR,		\
+static FC_DEVICE_ATTR(vport, field, S_IRUGO | S_IWUSR,		\
 			show_fc_vport_##field,				\
 			store_fc_vport_##field)
 
 
 #define fc_private_vport_rd_enum_attr(title, maxlen)			\
 static ssize_t								\
-show_fc_vport_##title (struct class_device *cdev, char *buf)		\
+show_fc_vport_##title (struct device *dev,				\
+		       struct device_attribute *attr,			\
+		       char *buf)					\
 {									\
-	struct fc_vport *vport = transport_class_to_vport(cdev);	\
+	struct fc_vport *vport = transport_class_to_vport(dev);		\
 	const char *name;						\
 	name = get_fc_##title##_name(vport->title);			\
 	if (!name)							\
 		return -EINVAL;						\
 	return snprintf(buf, maxlen, "%s\n", name);			\
 }									\
-static FC_CLASS_DEVICE_ATTR(vport, title, S_IRUGO,			\
+static FC_DEVICE_ATTR(vport, title, S_IRUGO,			\
 			show_fc_vport_##title, NULL)
 
 
 #define SETUP_VPORT_ATTRIBUTE_RD(field)					\
-	i->private_vport_attrs[count] = class_device_attr_vport_##field; \
+	i->private_vport_attrs[count] = device_attr_vport_##field; \
 	i->private_vport_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_vport_attrs[count].store = NULL;			\
 	i->vport_attrs[count] = &i->private_vport_attrs[count];		\
@@ -1131,21 +1147,21 @@ static FC_CLASS_DEVICE_ATTR(vport, title, S_IRUGO,			\
 	/* NOTE: Above MACRO differs: checks function not show bit */
 
 #define SETUP_PRIVATE_VPORT_ATTRIBUTE_RD(field)				\
-	i->private_vport_attrs[count] = class_device_attr_vport_##field; \
+	i->private_vport_attrs[count] = device_attr_vport_##field; \
 	i->private_vport_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_vport_attrs[count].store = NULL;			\
 	i->vport_attrs[count] = &i->private_vport_attrs[count];		\
 	count++
 
 #define SETUP_VPORT_ATTRIBUTE_WR(field)					\
-	i->private_vport_attrs[count] = class_device_attr_vport_##field; \
+	i->private_vport_attrs[count] = device_attr_vport_##field; \
 	i->vport_attrs[count] = &i->private_vport_attrs[count];		\
 	if (i->f->field)						\
 		count++
 	/* NOTE: Above MACRO differs: checks function */
 
 #define SETUP_VPORT_ATTRIBUTE_RW(field)					\
-	i->private_vport_attrs[count] = class_device_attr_vport_##field; \
+	i->private_vport_attrs[count] = device_attr_vport_##field; \
 	if (!i->f->set_vport_##field) {					\
 		i->private_vport_attrs[count].attr.mode = S_IRUGO;	\
 		i->private_vport_attrs[count].store = NULL;		\
@@ -1156,7 +1172,7 @@ static FC_CLASS_DEVICE_ATTR(vport, title, S_IRUGO,			\
 
 #define SETUP_PRIVATE_VPORT_ATTRIBUTE_RW(field)				\
 {									\
-	i->private_vport_attrs[count] = class_device_attr_vport_##field; \
+	i->private_vport_attrs[count] = device_attr_vport_##field; \
 	i->vport_attrs[count] = &i->private_vport_attrs[count];		\
 	count++;							\
 }
@@ -1176,35 +1192,36 @@ fc_private_vport_rd_attr_cast(node_name, "0x%llx\n", 20, unsigned long long);
 fc_private_vport_rd_attr_cast(port_name, "0x%llx\n", 20, unsigned long long);
 
 static ssize_t
-show_fc_vport_roles (struct class_device *cdev, char *buf)
+show_fc_vport_roles (struct device *dev, struct device_attribute *attr,
+		     char *buf)
 {
-	struct fc_vport *vport = transport_class_to_vport(cdev);
+	struct fc_vport *vport = transport_class_to_vport(dev);
 
 	if (vport->roles == FC_PORT_ROLE_UNKNOWN)
 		return snprintf(buf, 20, "unknown\n");
 	return get_fc_port_roles_names(vport->roles, buf);
 }
-static FC_CLASS_DEVICE_ATTR(vport, roles, S_IRUGO, show_fc_vport_roles, NULL);
+static FC_DEVICE_ATTR(vport, roles, S_IRUGO, show_fc_vport_roles, NULL);
 
 fc_private_vport_rd_enum_attr(vport_type, FC_PORTTYPE_MAX_NAMELEN);
 
 fc_private_vport_show_function(symbolic_name, "%s\n",
 		FC_VPORT_SYMBOLIC_NAMELEN + 1, )
 fc_vport_store_str_function(symbolic_name, FC_VPORT_SYMBOLIC_NAMELEN)
-static FC_CLASS_DEVICE_ATTR(vport, symbolic_name, S_IRUGO | S_IWUSR,
+static FC_DEVICE_ATTR(vport, symbolic_name, S_IRUGO | S_IWUSR,
 		show_fc_vport_symbolic_name, store_fc_vport_symbolic_name);
 
 static ssize_t
-store_fc_vport_delete(struct class_device *cdev, const char *buf,
-			   size_t count)
+store_fc_vport_delete(struct device *dev, struct device_attribute *attr,
+		      const char *buf, size_t count)
 {
-	struct fc_vport *vport = transport_class_to_vport(cdev);
+	struct fc_vport *vport = transport_class_to_vport(dev);
 	struct Scsi_Host *shost = vport_to_shost(vport);
 
 	fc_queue_work(shost, &vport->vport_delete_work);
 	return count;
 }
-static FC_CLASS_DEVICE_ATTR(vport, vport_delete, S_IWUSR,
+static FC_DEVICE_ATTR(vport, vport_delete, S_IWUSR,
 			NULL, store_fc_vport_delete);
 
 
@@ -1213,10 +1230,11 @@ static FC_CLASS_DEVICE_ATTR(vport, vport_delete, S_IWUSR,
  *  Write "1" to disable, write "0" to enable
  */
 static ssize_t
-store_fc_vport_disable(struct class_device *cdev, const char *buf,
+store_fc_vport_disable(struct device *dev, struct device_attribute *attr,
+		       const char *buf,
 			   size_t count)
 {
-	struct fc_vport *vport = transport_class_to_vport(cdev);
+	struct fc_vport *vport = transport_class_to_vport(dev);
 	struct Scsi_Host *shost = vport_to_shost(vport);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
 	int stat;
@@ -1236,7 +1254,7 @@ store_fc_vport_disable(struct class_device *cdev, const char *buf,
 	stat = i->f->vport_disable(vport, ((*buf == '0') ? false : true));
 	return stat ? stat : count;
 }
-static FC_CLASS_DEVICE_ATTR(vport, vport_disable, S_IWUSR,
+static FC_DEVICE_ATTR(vport, vport_disable, S_IWUSR,
 			NULL, store_fc_vport_disable);
 
 
@@ -1246,9 +1264,10 @@ static FC_CLASS_DEVICE_ATTR(vport, vport_disable, S_IWUSR,
 
 #define fc_host_show_function(field, format_string, sz, cast)		\
 static ssize_t								\
-show_fc_host_##field (struct class_device *cdev, char *buf)		\
+show_fc_host_##field (struct device *dev,				\
+		      struct device_attribute *attr, char *buf)		\
 {									\
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
+	struct Scsi_Host *shost = transport_class_to_shost(dev);	\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	if (i->f->get_host_##field)					\
 		i->f->get_host_##field(shost);				\
@@ -1257,11 +1276,12 @@ show_fc_host_##field (struct class_device *cdev, char *buf)		\
 
 #define fc_host_store_function(field)					\
 static ssize_t								\
-store_fc_host_##field(struct class_device *cdev, const char *buf,	\
-			   size_t count)				\
+store_fc_host_##field(struct device *dev, 				\
+		      struct device_attribute *attr,			\
+		      const char *buf,	size_t count)			\
 {									\
 	int val;							\
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
+	struct Scsi_Host *shost = transport_class_to_shost(dev);	\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	char *cp;							\
 									\
@@ -1274,10 +1294,11 @@ store_fc_host_##field(struct class_device *cdev, const char *buf,	\
 
 #define fc_host_store_str_function(field, slen)				\
 static ssize_t								\
-store_fc_host_##field(struct class_device *cdev, const char *buf,	\
-			   size_t count)				\
+store_fc_host_##field(struct device *dev,				\
+		      struct device_attribute *attr,			\
+		      const char *buf, size_t count)			\
 {									\
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
+	struct Scsi_Host *shost = transport_class_to_shost(dev);	\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	unsigned int cnt=count;						\
 									\
@@ -1293,26 +1314,27 @@ store_fc_host_##field(struct class_device *cdev, const char *buf,	\
 
 #define fc_host_rd_attr(field, format_string, sz)			\
 	fc_host_show_function(field, format_string, sz, )		\
-static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(host, field, S_IRUGO,			\
 			 show_fc_host_##field, NULL)
 
 #define fc_host_rd_attr_cast(field, format_string, sz, cast)		\
 	fc_host_show_function(field, format_string, sz, (cast))		\
-static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(host, field, S_IRUGO,			\
 			  show_fc_host_##field, NULL)
 
 #define fc_host_rw_attr(field, format_string, sz)			\
 	fc_host_show_function(field, format_string, sz, )		\
 	fc_host_store_function(field)					\
-static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO | S_IWUSR,		\
+static FC_DEVICE_ATTR(host, field, S_IRUGO | S_IWUSR,		\
 			show_fc_host_##field,				\
 			store_fc_host_##field)
 
 #define fc_host_rd_enum_attr(title, maxlen)				\
 static ssize_t								\
-show_fc_host_##title (struct class_device *cdev, char *buf)		\
+show_fc_host_##title (struct device *dev,				\
+		      struct device_attribute *attr, char *buf)		\
 {									\
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
+	struct Scsi_Host *shost = transport_class_to_shost(dev);	\
 	struct fc_internal *i = to_fc_internal(shost->transportt);	\
 	const char *name;						\
 	if (i->f->get_host_##title)					\
@@ -1322,10 +1344,10 @@ show_fc_host_##title (struct class_device *cdev, char *buf)		\
 		return -EINVAL;						\
 	return snprintf(buf, maxlen, "%s\n", name);			\
 }									\
-static FC_CLASS_DEVICE_ATTR(host, title, S_IRUGO, show_fc_host_##title, NULL)
+static FC_DEVICE_ATTR(host, title, S_IRUGO, show_fc_host_##title, NULL)
 
 #define SETUP_HOST_ATTRIBUTE_RD(field)					\
-	i->private_host_attrs[count] = class_device_attr_host_##field;	\
+	i->private_host_attrs[count] = device_attr_host_##field;	\
 	i->private_host_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_host_attrs[count].store = NULL;			\
 	i->host_attrs[count] = &i->private_host_attrs[count];		\
@@ -1333,14 +1355,14 @@ static FC_CLASS_DEVICE_ATTR(host, title, S_IRUGO, show_fc_host_##title, NULL)
 		count++
 
 #define SETUP_HOST_ATTRIBUTE_RD_NS(field)				\
-	i->private_host_attrs[count] = class_device_attr_host_##field;	\
+	i->private_host_attrs[count] = device_attr_host_##field;	\
 	i->private_host_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_host_attrs[count].store = NULL;			\
 	i->host_attrs[count] = &i->private_host_attrs[count];		\
 	count++
 
 #define SETUP_HOST_ATTRIBUTE_RW(field)					\
-	i->private_host_attrs[count] = class_device_attr_host_##field;	\
+	i->private_host_attrs[count] = device_attr_host_##field;	\
 	if (!i->f->set_host_##field) {					\
 		i->private_host_attrs[count].attr.mode = S_IRUGO;	\
 		i->private_host_attrs[count].store = NULL;		\
@@ -1352,24 +1374,25 @@ static FC_CLASS_DEVICE_ATTR(host, title, S_IRUGO, show_fc_host_##title, NULL)
 
 #define fc_private_host_show_function(field, format_string, sz, cast)	\
 static ssize_t								\
-show_fc_host_##field (struct class_device *cdev, char *buf)		\
+show_fc_host_##field (struct device *dev,				\
+		      struct device_attribute *attr, char *buf)		\
 {									\
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
+	struct Scsi_Host *shost = transport_class_to_shost(dev);	\
 	return snprintf(buf, sz, format_string, cast fc_host_##field(shost)); \
 }
 
 #define fc_private_host_rd_attr(field, format_string, sz)		\
 	fc_private_host_show_function(field, format_string, sz, )	\
-static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(host, field, S_IRUGO,			\
 			 show_fc_host_##field, NULL)
 
 #define fc_private_host_rd_attr_cast(field, format_string, sz, cast)	\
 	fc_private_host_show_function(field, format_string, sz, (cast)) \
-static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO,			\
+static FC_DEVICE_ATTR(host, field, S_IRUGO,			\
 			  show_fc_host_##field, NULL)
 
 #define SETUP_PRIVATE_HOST_ATTRIBUTE_RD(field)			\
-	i->private_host_attrs[count] = class_device_attr_host_##field;	\
+	i->private_host_attrs[count] = device_attr_host_##field;	\
 	i->private_host_attrs[count].attr.mode = S_IRUGO;		\
 	i->private_host_attrs[count].store = NULL;			\
 	i->host_attrs[count] = &i->private_host_attrs[count];		\
@@ -1377,7 +1400,7 @@ static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO,			\
 
 #define SETUP_PRIVATE_HOST_ATTRIBUTE_RW(field)			\
 {									\
-	i->private_host_attrs[count] = class_device_attr_host_##field;	\
+	i->private_host_attrs[count] = device_attr_host_##field;	\
 	i->host_attrs[count] = &i->private_host_attrs[count];		\
 	count++;							\
 }
@@ -1386,38 +1409,41 @@ static FC_CLASS_DEVICE_ATTR(host, field, S_IRUGO,			\
 /* Fixed Host Attributes */
 
 static ssize_t
-show_fc_host_supported_classes (struct class_device *cdev, char *buf)
+show_fc_host_supported_classes (struct device *dev,
+			        struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 
 	if (fc_host_supported_classes(shost) == FC_COS_UNSPECIFIED)
 		return snprintf(buf, 20, "unspecified\n");
 
 	return get_fc_cos_names(fc_host_supported_classes(shost), buf);
 }
-static FC_CLASS_DEVICE_ATTR(host, supported_classes, S_IRUGO,
+static FC_DEVICE_ATTR(host, supported_classes, S_IRUGO,
 		show_fc_host_supported_classes, NULL);
 
 static ssize_t
-show_fc_host_supported_fc4s (struct class_device *cdev, char *buf)
+show_fc_host_supported_fc4s (struct device *dev,
+			     struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	return (ssize_t)show_fc_fc4s(buf, fc_host_supported_fc4s(shost));
 }
-static FC_CLASS_DEVICE_ATTR(host, supported_fc4s, S_IRUGO,
+static FC_DEVICE_ATTR(host, supported_fc4s, S_IRUGO,
 		show_fc_host_supported_fc4s, NULL);
 
 static ssize_t
-show_fc_host_supported_speeds (struct class_device *cdev, char *buf)
+show_fc_host_supported_speeds (struct device *dev,
+			       struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 
 	if (fc_host_supported_speeds(shost) == FC_PORTSPEED_UNKNOWN)
 		return snprintf(buf, 20, "unknown\n");
 
 	return get_fc_port_speed_names(fc_host_supported_speeds(shost), buf);
 }
-static FC_CLASS_DEVICE_ATTR(host, supported_speeds, S_IRUGO,
+static FC_DEVICE_ATTR(host, supported_speeds, S_IRUGO,
 		show_fc_host_supported_speeds, NULL);
 
 
@@ -1433,9 +1459,10 @@ fc_private_host_rd_attr(serial_number, "%s\n", (FC_SERIAL_NUMBER_SIZE +1));
 /* Dynamic Host Attributes */
 
 static ssize_t
-show_fc_host_active_fc4s (struct class_device *cdev, char *buf)
+show_fc_host_active_fc4s (struct device *dev,
+			  struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
 
 	if (i->f->get_host_active_fc4s)
@@ -1443,13 +1470,14 @@ show_fc_host_active_fc4s (struct class_device *cdev, char *buf)
 
 	return (ssize_t)show_fc_fc4s(buf, fc_host_active_fc4s(shost));
 }
-static FC_CLASS_DEVICE_ATTR(host, active_fc4s, S_IRUGO,
+static FC_DEVICE_ATTR(host, active_fc4s, S_IRUGO,
 		show_fc_host_active_fc4s, NULL);
 
 static ssize_t
-show_fc_host_speed (struct class_device *cdev, char *buf)
+show_fc_host_speed (struct device *dev,
+		    struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
 
 	if (i->f->get_host_speed)
@@ -1460,7 +1488,7 @@ show_fc_host_speed (struct class_device *cdev, char *buf)
 
 	return get_fc_port_speed_names(fc_host_speed(shost), buf);
 }
-static FC_CLASS_DEVICE_ATTR(host, speed, S_IRUGO,
+static FC_DEVICE_ATTR(host, speed, S_IRUGO,
 		show_fc_host_speed, NULL);
 
 
@@ -1473,16 +1501,17 @@ fc_host_rd_attr(symbolic_name, "%s\n", FC_SYMBOLIC_NAME_SIZE + 1);
 fc_private_host_show_function(system_hostname, "%s\n",
 		FC_SYMBOLIC_NAME_SIZE + 1, )
 fc_host_store_str_function(system_hostname, FC_SYMBOLIC_NAME_SIZE)
-static FC_CLASS_DEVICE_ATTR(host, system_hostname, S_IRUGO | S_IWUSR,
+static FC_DEVICE_ATTR(host, system_hostname, S_IRUGO | S_IWUSR,
 		show_fc_host_system_hostname, store_fc_host_system_hostname);
 
 
 /* Private Host Attributes */
 
 static ssize_t
-show_fc_private_host_tgtid_bind_type(struct class_device *cdev, char *buf)
+show_fc_private_host_tgtid_bind_type(struct device *dev,
+				     struct device_attribute *attr, char *buf)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	const char *name;
 
 	name = get_fc_tgtid_bind_type_name(fc_host_tgtid_bind_type(shost));
@@ -1495,10 +1524,10 @@ show_fc_private_host_tgtid_bind_type(struct class_device *cdev, char *buf)
 	pos = list_entry((head)->next, typeof(*pos), member)
 
 static ssize_t
-store_fc_private_host_tgtid_bind_type(struct class_device *cdev,
-	const char *buf, size_t count)
+store_fc_private_host_tgtid_bind_type(struct device *dev,
+	struct device_attribute *attr, const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_rport *rport;
  	enum fc_tgtid_binding_type val;
 	unsigned long flags;
@@ -1523,15 +1552,15 @@ store_fc_private_host_tgtid_bind_type(struct class_device *cdev,
 	return count;
 }
 
-static FC_CLASS_DEVICE_ATTR(host, tgtid_bind_type, S_IRUGO | S_IWUSR,
+static FC_DEVICE_ATTR(host, tgtid_bind_type, S_IRUGO | S_IWUSR,
 			show_fc_private_host_tgtid_bind_type,
 			store_fc_private_host_tgtid_bind_type);
 
 static ssize_t
-store_fc_private_host_issue_lip(struct class_device *cdev,
-	const char *buf, size_t count)
+store_fc_private_host_issue_lip(struct device *dev,
+	struct device_attribute *attr, const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
 	int ret;
 
@@ -1544,7 +1573,7 @@ store_fc_private_host_issue_lip(struct class_device *cdev,
 	return -ENOENT;
 }
 
-static FC_CLASS_DEVICE_ATTR(host, issue_lip, S_IWUSR, NULL,
+static FC_DEVICE_ATTR(host, issue_lip, S_IWUSR, NULL,
 			store_fc_private_host_issue_lip);
 
 fc_private_host_rd_attr(npiv_vports_inuse, "%u\n", 20);
@@ -1556,9 +1585,9 @@ fc_private_host_rd_attr(npiv_vports_inuse, "%u\n", 20);
 
 /* Show a given an attribute in the statistics group */
 static ssize_t
-fc_stat_show(const struct class_device *cdev, char *buf, unsigned long offset)
+fc_stat_show(const struct device *dev, char *buf, unsigned long offset)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
 	struct fc_host_statistics *stats;
 	ssize_t ret = -ENOENT;
@@ -1579,12 +1608,14 @@ fc_stat_show(const struct class_device *cdev, char *buf, unsigned long offset)
 
 /* generate a read-only statistics attribute */
 #define fc_host_statistic(name)						\
-static ssize_t show_fcstat_##name(struct class_device *cd, char *buf) 	\
+static ssize_t show_fcstat_##name(struct device *cd,			\
+				  struct device_attribute *attr,	\
+				  char *buf)				\
 {									\
 	return fc_stat_show(cd, buf, 					\
 			    offsetof(struct fc_host_statistics, name));	\
 }									\
-static FC_CLASS_DEVICE_ATTR(host, name, S_IRUGO, show_fcstat_##name, NULL)
+static FC_DEVICE_ATTR(host, name, S_IRUGO, show_fcstat_##name, NULL)
 
 fc_host_statistic(seconds_since_last_reset);
 fc_host_statistic(tx_frames);
@@ -1608,10 +1639,10 @@ fc_host_statistic(fcp_input_megabytes);
 fc_host_statistic(fcp_output_megabytes);
 
 static ssize_t
-fc_reset_statistics(struct class_device *cdev, const char *buf,
-			   size_t count)
+fc_reset_statistics(struct device *dev, struct device_attribute *attr,
+		    const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_internal *i = to_fc_internal(shost->transportt);
 
 	/* ignore any data value written to the attribute */
@@ -1622,31 +1653,31 @@ fc_reset_statistics(struct class_device *cdev, const char *buf,
 
 	return -ENOENT;
 }
-static FC_CLASS_DEVICE_ATTR(host, reset_statistics, S_IWUSR, NULL,
+static FC_DEVICE_ATTR(host, reset_statistics, S_IWUSR, NULL,
 				fc_reset_statistics);
 
 static struct attribute *fc_statistics_attrs[] = {
-	&class_device_attr_host_seconds_since_last_reset.attr,
-	&class_device_attr_host_tx_frames.attr,
-	&class_device_attr_host_tx_words.attr,
-	&class_device_attr_host_rx_frames.attr,
-	&class_device_attr_host_rx_words.attr,
-	&class_device_attr_host_lip_count.attr,
-	&class_device_attr_host_nos_count.attr,
-	&class_device_attr_host_error_frames.attr,
-	&class_device_attr_host_dumped_frames.attr,
-	&class_device_attr_host_link_failure_count.attr,
-	&class_device_attr_host_loss_of_sync_count.attr,
-	&class_device_attr_host_loss_of_signal_count.attr,
-	&class_device_attr_host_prim_seq_protocol_err_count.attr,
-	&class_device_attr_host_invalid_tx_word_count.attr,
-	&class_device_attr_host_invalid_crc_count.attr,
-	&class_device_attr_host_fcp_input_requests.attr,
-	&class_device_attr_host_fcp_output_requests.attr,
-	&class_device_attr_host_fcp_control_requests.attr,
-	&class_device_attr_host_fcp_input_megabytes.attr,
-	&class_device_attr_host_fcp_output_megabytes.attr,
-	&class_device_attr_host_reset_statistics.attr,
+	&device_attr_host_seconds_since_last_reset.attr,
+	&device_attr_host_tx_frames.attr,
+	&device_attr_host_tx_words.attr,
+	&device_attr_host_rx_frames.attr,
+	&device_attr_host_rx_words.attr,
+	&device_attr_host_lip_count.attr,
+	&device_attr_host_nos_count.attr,
+	&device_attr_host_error_frames.attr,
+	&device_attr_host_dumped_frames.attr,
+	&device_attr_host_link_failure_count.attr,
+	&device_attr_host_loss_of_sync_count.attr,
+	&device_attr_host_loss_of_signal_count.attr,
+	&device_attr_host_prim_seq_protocol_err_count.attr,
+	&device_attr_host_invalid_tx_word_count.attr,
+	&device_attr_host_invalid_crc_count.attr,
+	&device_attr_host_fcp_input_requests.attr,
+	&device_attr_host_fcp_output_requests.attr,
+	&device_attr_host_fcp_control_requests.attr,
+	&device_attr_host_fcp_input_megabytes.attr,
+	&device_attr_host_fcp_output_megabytes.attr,
+	&device_attr_host_reset_statistics.attr,
 	NULL
 };
 
@@ -1695,10 +1726,10 @@ fc_parse_wwn(const char *ns, u64 *nm)
  * as hex characters, and may *not* contain any prefixes (e.g. 0x, x, etc)
  */
 static ssize_t
-store_fc_host_vport_create(struct class_device *cdev, const char *buf,
-			   size_t count)
+store_fc_host_vport_create(struct device *dev, struct device_attribute *attr,
+			   const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_vport_identifiers vid;
 	struct fc_vport *vport;
 	unsigned int cnt=count;
@@ -1731,7 +1762,7 @@ store_fc_host_vport_create(struct class_device *cdev, const char *buf,
 	stat = fc_vport_create(shost, 0, &shost->shost_gendev, &vid, &vport);
 	return stat ? stat : count;
 }
-static FC_CLASS_DEVICE_ATTR(host, vport_create, S_IWUSR, NULL,
+static FC_DEVICE_ATTR(host, vport_create, S_IWUSR, NULL,
 			store_fc_host_vport_create);
 
 
@@ -1742,10 +1773,10 @@ static FC_CLASS_DEVICE_ATTR(host, vport_create, S_IWUSR, NULL,
  * any prefixes (e.g. 0x, x, etc)
  */
 static ssize_t
-store_fc_host_vport_delete(struct class_device *cdev, const char *buf,
-			   size_t count)
+store_fc_host_vport_delete(struct device *dev, struct device_attribute *attr,
+			   const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct fc_host_attrs *fc_host = shost_to_fc_host(shost);
 	struct fc_vport *vport;
 	u64 wwpn, wwnn;
@@ -1787,7 +1818,7 @@ store_fc_host_vport_delete(struct class_device *cdev, const char *buf,
 	stat = fc_vport_terminate(vport);
 	return stat ? stat : count;
 }
-static FC_CLASS_DEVICE_ATTR(host, vport_delete, S_IWUSR, NULL,
+static FC_DEVICE_ATTR(host, vport_delete, S_IWUSR, NULL,
 			store_fc_host_vport_delete);
 
 
diff --git a/drivers/scsi/scsi_transport_iscsi.c b/drivers/scsi/scsi_transport_iscsi.c
index ca7bb6f63bde..65d1737eb664 100644
--- a/drivers/scsi/scsi_transport_iscsi.c
+++ b/drivers/scsi/scsi_transport_iscsi.c
@@ -40,13 +40,13 @@ struct iscsi_internal {
 	struct scsi_transport_template t;
 	struct iscsi_transport *iscsi_transport;
 	struct list_head list;
-	struct class_device cdev;
+	struct device dev;
 
-	struct class_device_attribute *host_attrs[ISCSI_HOST_ATTRS + 1];
+	struct device_attribute *host_attrs[ISCSI_HOST_ATTRS + 1];
 	struct transport_container conn_cont;
-	struct class_device_attribute *conn_attrs[ISCSI_CONN_ATTRS + 1];
+	struct device_attribute *conn_attrs[ISCSI_CONN_ATTRS + 1];
 	struct transport_container session_cont;
-	struct class_device_attribute *session_attrs[ISCSI_SESSION_ATTRS + 1];
+	struct device_attribute *session_attrs[ISCSI_SESSION_ATTRS + 1];
 };
 
 static atomic_t iscsi_session_nr; /* sysfs session id for next new session */
@@ -63,12 +63,12 @@ static DEFINE_SPINLOCK(iscsi_transport_lock);
 #define to_iscsi_internal(tmpl) \
 	container_of(tmpl, struct iscsi_internal, t)
 
-#define cdev_to_iscsi_internal(_cdev) \
-	container_of(_cdev, struct iscsi_internal, cdev)
+#define dev_to_iscsi_internal(_dev) \
+	container_of(_dev, struct iscsi_internal, dev)
 
-static void iscsi_transport_release(struct class_device *cdev)
+static void iscsi_transport_release(struct device *dev)
 {
-	struct iscsi_internal *priv = cdev_to_iscsi_internal(cdev);
+	struct iscsi_internal *priv = dev_to_iscsi_internal(dev);
 	kfree(priv);
 }
 
@@ -78,25 +78,27 @@ static void iscsi_transport_release(struct class_device *cdev)
  */
 static struct class iscsi_transport_class = {
 	.name = "iscsi_transport",
-	.release = iscsi_transport_release,
+	.dev_release = iscsi_transport_release,
 };
 
 static ssize_t
-show_transport_handle(struct class_device *cdev, char *buf)
+show_transport_handle(struct device *dev, struct device_attribute *attr,
+		      char *buf)
 {
-	struct iscsi_internal *priv = cdev_to_iscsi_internal(cdev);
+	struct iscsi_internal *priv = dev_to_iscsi_internal(dev);
 	return sprintf(buf, "%llu\n", (unsigned long long)iscsi_handle(priv->iscsi_transport));
 }
-static CLASS_DEVICE_ATTR(handle, S_IRUGO, show_transport_handle, NULL);
+static DEVICE_ATTR(handle, S_IRUGO, show_transport_handle, NULL);
 
 #define show_transport_attr(name, format)				\
 static ssize_t								\
-show_transport_##name(struct class_device *cdev, char *buf)		\
+show_transport_##name(struct device *dev, 				\
+		      struct device_attribute *attr,char *buf)		\
 {									\
-	struct iscsi_internal *priv = cdev_to_iscsi_internal(cdev);	\
+	struct iscsi_internal *priv = dev_to_iscsi_internal(dev);	\
 	return sprintf(buf, format"\n", priv->iscsi_transport->name);	\
 }									\
-static CLASS_DEVICE_ATTR(name, S_IRUGO, show_transport_##name, NULL);
+static DEVICE_ATTR(name, S_IRUGO, show_transport_##name, NULL);
 
 show_transport_attr(caps, "0x%x");
 show_transport_attr(max_lun, "%d");
@@ -104,11 +106,11 @@ show_transport_attr(max_conn, "%d");
 show_transport_attr(max_cmd_len, "%d");
 
 static struct attribute *iscsi_transport_attrs[] = {
-	&class_device_attr_handle.attr,
-	&class_device_attr_caps.attr,
-	&class_device_attr_max_lun.attr,
-	&class_device_attr_max_conn.attr,
-	&class_device_attr_max_cmd_len.attr,
+	&dev_attr_handle.attr,
+	&dev_attr_caps.attr,
+	&dev_attr_max_lun.attr,
+	&dev_attr_max_conn.attr,
+	&dev_attr_max_cmd_len.attr,
 	NULL,
 };
 
@@ -119,7 +121,7 @@ static struct attribute_group iscsi_transport_group = {
 
 
 static int iscsi_setup_host(struct transport_container *tc, struct device *dev,
-			    struct class_device *cdev)
+			    struct device *cdev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 	struct iscsi_host *ihost = shost->shost_data;
@@ -139,7 +141,7 @@ static int iscsi_setup_host(struct transport_container *tc, struct device *dev,
 }
 
 static int iscsi_remove_host(struct transport_container *tc, struct device *dev,
-			     struct class_device *cdev)
+			     struct device *cdev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 	struct iscsi_host *ihost = shost->shost_data;
@@ -1337,11 +1339,8 @@ iscsi_if_rx(struct sk_buff *skb)
 	mutex_unlock(&rx_queue_mutex);
 }
 
-#define iscsi_cdev_to_conn(_cdev) \
-	iscsi_dev_to_conn(_cdev->dev)
-
 #define ISCSI_CLASS_ATTR(_prefix,_name,_mode,_show,_store)		\
-struct class_device_attribute class_device_attr_##_prefix##_##_name =	\
+struct device_attribute dev_attr_##_prefix##_##_name =	\
 	__ATTR(_name,_mode,_show,_store)
 
 /*
@@ -1349,9 +1348,10 @@ struct class_device_attribute class_device_attr_##_prefix##_##_name =	\
  */
 #define iscsi_conn_attr_show(param)					\
 static ssize_t								\
-show_conn_param_##param(struct class_device *cdev, char *buf)		\
+show_conn_param_##param(struct device *dev, 				\
+			struct device_attribute *attr, char *buf)	\
 {									\
-	struct iscsi_cls_conn *conn = iscsi_cdev_to_conn(cdev);		\
+	struct iscsi_cls_conn *conn = iscsi_dev_to_conn(dev->parent);	\
 	struct iscsi_transport *t = conn->transport;			\
 	return t->get_conn_param(conn, param, buf);			\
 }
@@ -1375,17 +1375,16 @@ iscsi_conn_attr(address, ISCSI_PARAM_CONN_ADDRESS);
 iscsi_conn_attr(ping_tmo, ISCSI_PARAM_PING_TMO);
 iscsi_conn_attr(recv_tmo, ISCSI_PARAM_RECV_TMO);
 
-#define iscsi_cdev_to_session(_cdev) \
-	iscsi_dev_to_session(_cdev->dev)
-
 /*
  * iSCSI session attrs
  */
 #define iscsi_session_attr_show(param, perm)				\
 static ssize_t								\
-show_session_param_##param(struct class_device *cdev, char *buf)	\
+show_session_param_##param(struct device *dev,				\
+			   struct device_attribute *attr, char *buf)	\
 {									\
-	struct iscsi_cls_session *session = iscsi_cdev_to_session(cdev); \
+	struct iscsi_cls_session *session = 				\
+		iscsi_dev_to_session(dev->parent);			\
 	struct iscsi_transport *t = session->transport;			\
 									\
 	if (perm && !capable(CAP_SYS_ADMIN))				\
@@ -1417,9 +1416,10 @@ iscsi_session_attr(abort_tmo, ISCSI_PARAM_ABORT_TMO, 0);
 iscsi_session_attr(lu_reset_tmo, ISCSI_PARAM_LU_RESET_TMO, 0);
 
 static ssize_t
-show_priv_session_state(struct class_device *cdev, char *buf)
+show_priv_session_state(struct device *dev, struct device_attribute *attr,
+			char *buf)
 {
-	struct iscsi_cls_session *session = iscsi_cdev_to_session(cdev);
+	struct iscsi_cls_session *session = iscsi_dev_to_session(dev->parent);
 	return sprintf(buf, "%s\n", iscsi_session_state_name(session->state));
 }
 static ISCSI_CLASS_ATTR(priv_sess, state, S_IRUGO, show_priv_session_state,
@@ -1427,9 +1427,11 @@ static ISCSI_CLASS_ATTR(priv_sess, state, S_IRUGO, show_priv_session_state,
 
 #define iscsi_priv_session_attr_show(field, format)			\
 static ssize_t								\
-show_priv_session_##field(struct class_device *cdev, char *buf)		\
+show_priv_session_##field(struct device *dev, 				\
+			  struct device_attribute *attr, char *buf)	\
 {									\
-	struct iscsi_cls_session *session = iscsi_cdev_to_session(cdev);\
+	struct iscsi_cls_session *session = 				\
+			iscsi_dev_to_session(dev->parent);		\
 	return sprintf(buf, format"\n", session->field);		\
 }
 
@@ -1444,9 +1446,10 @@ iscsi_priv_session_attr(recovery_tmo, "%d");
  */
 #define iscsi_host_attr_show(param)					\
 static ssize_t								\
-show_host_param_##param(struct class_device *cdev, char *buf)		\
+show_host_param_##param(struct device *dev, 				\
+			struct device_attribute *attr, char *buf)	\
 {									\
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);	\
+	struct Scsi_Host *shost = transport_class_to_shost(dev);	\
 	struct iscsi_internal *priv = to_iscsi_internal(shost->transportt); \
 	return priv->iscsi_transport->get_host_param(shost, param, buf); \
 }
@@ -1463,7 +1466,7 @@ iscsi_host_attr(initiatorname, ISCSI_HOST_PARAM_INITIATOR_NAME);
 
 #define SETUP_PRIV_SESSION_RD_ATTR(field)				\
 do {									\
-	priv->session_attrs[count] = &class_device_attr_priv_sess_##field; \
+	priv->session_attrs[count] = &dev_attr_priv_sess_##field; \
 	count++;							\
 } while (0)
 
@@ -1471,7 +1474,7 @@ do {									\
 #define SETUP_SESSION_RD_ATTR(field, param_flag)			\
 do {									\
 	if (tt->param_mask & param_flag) {				\
-		priv->session_attrs[count] = &class_device_attr_sess_##field; \
+		priv->session_attrs[count] = &dev_attr_sess_##field; \
 		count++;						\
 	}								\
 } while (0)
@@ -1479,7 +1482,7 @@ do {									\
 #define SETUP_CONN_RD_ATTR(field, param_flag)				\
 do {									\
 	if (tt->param_mask & param_flag) {				\
-		priv->conn_attrs[count] = &class_device_attr_conn_##field; \
+		priv->conn_attrs[count] = &dev_attr_conn_##field; \
 		count++;						\
 	}								\
 } while (0)
@@ -1487,7 +1490,7 @@ do {									\
 #define SETUP_HOST_RD_ATTR(field, param_flag)				\
 do {									\
 	if (tt->host_param_mask & param_flag) {				\
-		priv->host_attrs[count] = &class_device_attr_host_##field; \
+		priv->host_attrs[count] = &dev_attr_host_##field; \
 		count++;						\
 	}								\
 } while (0)
@@ -1578,15 +1581,15 @@ iscsi_register_transport(struct iscsi_transport *tt)
 	priv->iscsi_transport = tt;
 	priv->t.user_scan = iscsi_user_scan;
 
-	priv->cdev.class = &iscsi_transport_class;
-	snprintf(priv->cdev.class_id, BUS_ID_SIZE, "%s", tt->name);
-	err = class_device_register(&priv->cdev);
+	priv->dev.class = &iscsi_transport_class;
+	snprintf(priv->dev.bus_id, BUS_ID_SIZE, "%s", tt->name);
+	err = device_register(&priv->dev);
 	if (err)
 		goto free_priv;
 
-	err = sysfs_create_group(&priv->cdev.kobj, &iscsi_transport_group);
+	err = sysfs_create_group(&priv->dev.kobj, &iscsi_transport_group);
 	if (err)
-		goto unregister_cdev;
+		goto unregister_dev;
 
 	/* host parameters */
 	priv->t.host_attrs.ac.attrs = &priv->host_attrs[0];
@@ -1663,8 +1666,8 @@ iscsi_register_transport(struct iscsi_transport *tt)
 	printk(KERN_NOTICE "iscsi: registered transport (%s)\n", tt->name);
 	return &priv->t;
 
-unregister_cdev:
-	class_device_unregister(&priv->cdev);
+unregister_dev:
+	device_unregister(&priv->dev);
 free_priv:
 	kfree(priv);
 	return NULL;
@@ -1691,8 +1694,8 @@ int iscsi_unregister_transport(struct iscsi_transport *tt)
 	transport_container_unregister(&priv->session_cont);
 	transport_container_unregister(&priv->t.host_attrs);
 
-	sysfs_remove_group(&priv->cdev.kobj, &iscsi_transport_group);
-	class_device_unregister(&priv->cdev);
+	sysfs_remove_group(&priv->dev.kobj, &iscsi_transport_group);
+	device_unregister(&priv->dev);
 	mutex_unlock(&rx_queue_mutex);
 
 	return 0;
diff --git a/drivers/scsi/scsi_transport_sas.c b/drivers/scsi/scsi_transport_sas.c
index 43a964d635b4..27ec625ab771 100644
--- a/drivers/scsi/scsi_transport_sas.c
+++ b/drivers/scsi/scsi_transport_sas.c
@@ -53,8 +53,8 @@ struct sas_host_attrs {
 /*
  * Hack to allow attributes of the same name in different objects.
  */
-#define SAS_CLASS_DEVICE_ATTR(_prefix,_name,_mode,_show,_store) \
-	struct class_device_attribute class_device_attr_##_prefix##_##_name = \
+#define SAS_DEVICE_ATTR(_prefix,_name,_mode,_show,_store) \
+	struct device_attribute dev_attr_##_prefix##_##_name = \
 	__ATTR(_name,_mode,_show,_store)
 
 
@@ -261,7 +261,7 @@ static void sas_bsg_remove(struct Scsi_Host *shost, struct sas_rphy *rphy)
  */
 
 static int sas_host_setup(struct transport_container *tc, struct device *dev,
-			  struct class_device *cdev)
+			  struct device *cdev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 	struct sas_host_attrs *sas_host = to_sas_host_attrs(shost);
@@ -280,7 +280,7 @@ static int sas_host_setup(struct transport_container *tc, struct device *dev,
 }
 
 static int sas_host_remove(struct transport_container *tc, struct device *dev,
-			   struct class_device *cdev)
+			   struct device *cdev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 
@@ -356,22 +356,24 @@ EXPORT_SYMBOL(sas_remove_host);
 
 #define sas_phy_show_simple(field, name, format_string, cast)		\
 static ssize_t								\
-show_sas_phy_##name(struct class_device *cdev, char *buf)		\
+show_sas_phy_##name(struct device *dev, 				\
+		    struct device_attribute *attr, char *buf)		\
 {									\
-	struct sas_phy *phy = transport_class_to_phy(cdev);		\
+	struct sas_phy *phy = transport_class_to_phy(dev);		\
 									\
 	return snprintf(buf, 20, format_string, cast phy->field);	\
 }
 
 #define sas_phy_simple_attr(field, name, format_string, type)		\
 	sas_phy_show_simple(field, name, format_string, (type))	\
-static CLASS_DEVICE_ATTR(name, S_IRUGO, show_sas_phy_##name, NULL)
+static DEVICE_ATTR(name, S_IRUGO, show_sas_phy_##name, NULL)
 
 #define sas_phy_show_protocol(field, name)				\
 static ssize_t								\
-show_sas_phy_##name(struct class_device *cdev, char *buf)		\
+show_sas_phy_##name(struct device *dev, 				\
+		    struct device_attribute *attr, char *buf)		\
 {									\
-	struct sas_phy *phy = transport_class_to_phy(cdev);		\
+	struct sas_phy *phy = transport_class_to_phy(dev);		\
 									\
 	if (!phy->field)						\
 		return snprintf(buf, 20, "none\n");			\
@@ -380,13 +382,14 @@ show_sas_phy_##name(struct class_device *cdev, char *buf)		\
 
 #define sas_phy_protocol_attr(field, name)				\
 	sas_phy_show_protocol(field, name)				\
-static CLASS_DEVICE_ATTR(name, S_IRUGO, show_sas_phy_##name, NULL)
+static DEVICE_ATTR(name, S_IRUGO, show_sas_phy_##name, NULL)
 
 #define sas_phy_show_linkspeed(field)					\
 static ssize_t								\
-show_sas_phy_##field(struct class_device *cdev, char *buf)		\
+show_sas_phy_##field(struct device *dev, 				\
+		     struct device_attribute *attr, char *buf)		\
 {									\
-	struct sas_phy *phy = transport_class_to_phy(cdev);		\
+	struct sas_phy *phy = transport_class_to_phy(dev);		\
 									\
 	return get_sas_linkspeed_names(phy->field, buf);		\
 }
@@ -394,10 +397,11 @@ show_sas_phy_##field(struct class_device *cdev, char *buf)		\
 /* Fudge to tell if we're minimum or maximum */
 #define sas_phy_store_linkspeed(field)					\
 static ssize_t								\
-store_sas_phy_##field(struct class_device *cdev, const char *buf,	\
-		      size_t count)					\
+store_sas_phy_##field(struct device *dev, 				\
+		      struct device_attribute *attr, 			\
+		      const char *buf,	size_t count)			\
 {									\
-	struct sas_phy *phy = transport_class_to_phy(cdev);		\
+	struct sas_phy *phy = transport_class_to_phy(dev);		\
 	struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);	\
 	struct sas_internal *i = to_sas_internal(shost->transportt);	\
 	u32 value;							\
@@ -416,19 +420,20 @@ store_sas_phy_##field(struct class_device *cdev, const char *buf,	\
 #define sas_phy_linkspeed_rw_attr(field)				\
 	sas_phy_show_linkspeed(field)					\
 	sas_phy_store_linkspeed(field)					\
-static CLASS_DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field,		\
+static DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field,		\
 	store_sas_phy_##field)
 
 #define sas_phy_linkspeed_attr(field)					\
 	sas_phy_show_linkspeed(field)					\
-static CLASS_DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field, NULL)
+static DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field, NULL)
 
 
 #define sas_phy_show_linkerror(field)					\
 static ssize_t								\
-show_sas_phy_##field(struct class_device *cdev, char *buf)		\
+show_sas_phy_##field(struct device *dev, 				\
+		     struct device_attribute *attr, char *buf)		\
 {									\
-	struct sas_phy *phy = transport_class_to_phy(cdev);		\
+	struct sas_phy *phy = transport_class_to_phy(dev);		\
 	struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);	\
 	struct sas_internal *i = to_sas_internal(shost->transportt);	\
 	int error;							\
@@ -441,24 +446,25 @@ show_sas_phy_##field(struct class_device *cdev, char *buf)		\
 
 #define sas_phy_linkerror_attr(field)					\
 	sas_phy_show_linkerror(field)					\
-static CLASS_DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field, NULL)
+static DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field, NULL)
 
 
 static ssize_t
-show_sas_device_type(struct class_device *cdev, char *buf)
+show_sas_device_type(struct device *dev,
+		     struct device_attribute *attr, char *buf)
 {
-	struct sas_phy *phy = transport_class_to_phy(cdev);
+	struct sas_phy *phy = transport_class_to_phy(dev);
 
 	if (!phy->identify.device_type)
 		return snprintf(buf, 20, "none\n");
 	return get_sas_device_type_names(phy->identify.device_type, buf);
 }
-static CLASS_DEVICE_ATTR(device_type, S_IRUGO, show_sas_device_type, NULL);
+static DEVICE_ATTR(device_type, S_IRUGO, show_sas_device_type, NULL);
 
-static ssize_t do_sas_phy_enable(struct class_device *cdev,
+static ssize_t do_sas_phy_enable(struct device *dev,
 		size_t count, int enable)
 {
-	struct sas_phy *phy = transport_class_to_phy(cdev);
+	struct sas_phy *phy = transport_class_to_phy(dev);
 	struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);
 	struct sas_internal *i = to_sas_internal(shost->transportt);
 	int error;
@@ -470,18 +476,19 @@ static ssize_t do_sas_phy_enable(struct class_device *cdev,
 	return count;
 };
 
-static ssize_t store_sas_phy_enable(struct class_device *cdev,
-		const char *buf, size_t count)
+static ssize_t
+store_sas_phy_enable(struct device *dev, struct device_attribute *attr,
+		     const char *buf, size_t count)
 {
 	if (count < 1)
 		return -EINVAL;
 
 	switch (buf[0]) {
 	case '0':
-		do_sas_phy_enable(cdev, count, 0);
+		do_sas_phy_enable(dev, count, 0);
 		break;
 	case '1':
-		do_sas_phy_enable(cdev, count, 1);
+		do_sas_phy_enable(dev, count, 1);
 		break;
 	default:
 		return -EINVAL;
@@ -490,20 +497,22 @@ static ssize_t store_sas_phy_enable(struct class_device *cdev,
 	return count;
 }
 
-static ssize_t show_sas_phy_enable(struct class_device *cdev, char *buf)
+static ssize_t
+show_sas_phy_enable(struct device *dev, struct device_attribute *attr,
+		    char *buf)
 {
-	struct sas_phy *phy = transport_class_to_phy(cdev);
+	struct sas_phy *phy = transport_class_to_phy(dev);
 
 	return snprintf(buf, 20, "%d", phy->enabled);
 }
 
-static CLASS_DEVICE_ATTR(enable, S_IRUGO | S_IWUSR, show_sas_phy_enable,
+static DEVICE_ATTR(enable, S_IRUGO | S_IWUSR, show_sas_phy_enable,
 			 store_sas_phy_enable);
 
-static ssize_t do_sas_phy_reset(struct class_device *cdev,
-		size_t count, int hard_reset)
+static ssize_t
+do_sas_phy_reset(struct device *dev, size_t count, int hard_reset)
 {
-	struct sas_phy *phy = transport_class_to_phy(cdev);
+	struct sas_phy *phy = transport_class_to_phy(dev);
 	struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);
 	struct sas_internal *i = to_sas_internal(shost->transportt);
 	int error;
@@ -514,19 +523,21 @@ static ssize_t do_sas_phy_reset(struct class_device *cdev,
 	return count;
 };
 
-static ssize_t store_sas_link_reset(struct class_device *cdev,
-		const char *buf, size_t count)
+static ssize_t
+store_sas_link_reset(struct device *dev, struct device_attribute *attr,
+		     const char *buf, size_t count)
 {
-	return do_sas_phy_reset(cdev, count, 0);
+	return do_sas_phy_reset(dev, count, 0);
 }
-static CLASS_DEVICE_ATTR(link_reset, S_IWUSR, NULL, store_sas_link_reset);
+static DEVICE_ATTR(link_reset, S_IWUSR, NULL, store_sas_link_reset);
 
-static ssize_t store_sas_hard_reset(struct class_device *cdev,
-		const char *buf, size_t count)
+static ssize_t
+store_sas_hard_reset(struct device *dev, struct device_attribute *attr,
+		     const char *buf, size_t count)
 {
-	return do_sas_phy_reset(cdev, count, 1);
+	return do_sas_phy_reset(dev, count, 1);
 }
-static CLASS_DEVICE_ATTR(hard_reset, S_IWUSR, NULL, store_sas_hard_reset);
+static DEVICE_ATTR(hard_reset, S_IWUSR, NULL, store_sas_hard_reset);
 
 sas_phy_protocol_attr(identify.initiator_port_protocols,
 		initiator_port_protocols);
@@ -695,16 +706,17 @@ EXPORT_SYMBOL(scsi_is_sas_phy);
  */
 #define sas_port_show_simple(field, name, format_string, cast)		\
 static ssize_t								\
-show_sas_port_##name(struct class_device *cdev, char *buf)		\
+show_sas_port_##name(struct device *dev, 				\
+		     struct device_attribute *attr, char *buf)		\
 {									\
-	struct sas_port *port = transport_class_to_sas_port(cdev);	\
+	struct sas_port *port = transport_class_to_sas_port(dev);	\
 									\
 	return snprintf(buf, 20, format_string, cast port->field);	\
 }
 
 #define sas_port_simple_attr(field, name, format_string, type)		\
 	sas_port_show_simple(field, name, format_string, (type))	\
-static CLASS_DEVICE_ATTR(name, S_IRUGO, show_sas_port_##name, NULL)
+static DEVICE_ATTR(name, S_IRUGO, show_sas_port_##name, NULL)
 
 sas_port_simple_attr(num_phys, num_phys, "%d\n", int);
 
@@ -1017,23 +1029,25 @@ EXPORT_SYMBOL(sas_port_mark_backlink);
 
 #define sas_rphy_show_simple(field, name, format_string, cast)		\
 static ssize_t								\
-show_sas_rphy_##name(struct class_device *cdev, char *buf)		\
+show_sas_rphy_##name(struct device *dev, 				\
+		     struct device_attribute *attr, char *buf)		\
 {									\
-	struct sas_rphy *rphy = transport_class_to_rphy(cdev);	\
+	struct sas_rphy *rphy = transport_class_to_rphy(dev);		\
 									\
 	return snprintf(buf, 20, format_string, cast rphy->field);	\
 }
 
 #define sas_rphy_simple_attr(field, name, format_string, type)		\
 	sas_rphy_show_simple(field, name, format_string, (type))	\
-static SAS_CLASS_DEVICE_ATTR(rphy, name, S_IRUGO, 			\
+static SAS_DEVICE_ATTR(rphy, name, S_IRUGO, 			\
 		show_sas_rphy_##name, NULL)
 
 #define sas_rphy_show_protocol(field, name)				\
 static ssize_t								\
-show_sas_rphy_##name(struct class_device *cdev, char *buf)		\
+show_sas_rphy_##name(struct device *dev, 				\
+		     struct device_attribute *attr, char *buf)		\
 {									\
-	struct sas_rphy *rphy = transport_class_to_rphy(cdev);	\
+	struct sas_rphy *rphy = transport_class_to_rphy(dev);		\
 									\
 	if (!rphy->field)					\
 		return snprintf(buf, 20, "none\n");			\
@@ -1042,13 +1056,14 @@ show_sas_rphy_##name(struct class_device *cdev, char *buf)		\
 
 #define sas_rphy_protocol_attr(field, name)				\
 	sas_rphy_show_protocol(field, name)				\
-static SAS_CLASS_DEVICE_ATTR(rphy, name, S_IRUGO,			\
+static SAS_DEVICE_ATTR(rphy, name, S_IRUGO,			\
 		show_sas_rphy_##name, NULL)
 
 static ssize_t
-show_sas_rphy_device_type(struct class_device *cdev, char *buf)
+show_sas_rphy_device_type(struct device *dev,
+			  struct device_attribute *attr, char *buf)
 {
-	struct sas_rphy *rphy = transport_class_to_rphy(cdev);
+	struct sas_rphy *rphy = transport_class_to_rphy(dev);
 
 	if (!rphy->identify.device_type)
 		return snprintf(buf, 20, "none\n");
@@ -1056,13 +1071,14 @@ show_sas_rphy_device_type(struct class_device *cdev, char *buf)
 			rphy->identify.device_type, buf);
 }
 
-static SAS_CLASS_DEVICE_ATTR(rphy, device_type, S_IRUGO,
+static SAS_DEVICE_ATTR(rphy, device_type, S_IRUGO,
 		show_sas_rphy_device_type, NULL);
 
 static ssize_t
-show_sas_rphy_enclosure_identifier(struct class_device *cdev, char *buf)
+show_sas_rphy_enclosure_identifier(struct device *dev,
+				   struct device_attribute *attr, char *buf)
 {
-	struct sas_rphy *rphy = transport_class_to_rphy(cdev);
+	struct sas_rphy *rphy = transport_class_to_rphy(dev);
 	struct sas_phy *phy = dev_to_phy(rphy->dev.parent);
 	struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);
 	struct sas_internal *i = to_sas_internal(shost->transportt);
@@ -1082,13 +1098,14 @@ show_sas_rphy_enclosure_identifier(struct class_device *cdev, char *buf)
 	return sprintf(buf, "0x%llx\n", (unsigned long long)identifier);
 }
 
-static SAS_CLASS_DEVICE_ATTR(rphy, enclosure_identifier, S_IRUGO,
+static SAS_DEVICE_ATTR(rphy, enclosure_identifier, S_IRUGO,
 		show_sas_rphy_enclosure_identifier, NULL);
 
 static ssize_t
-show_sas_rphy_bay_identifier(struct class_device *cdev, char *buf)
+show_sas_rphy_bay_identifier(struct device *dev,
+			     struct device_attribute *attr, char *buf)
 {
-	struct sas_rphy *rphy = transport_class_to_rphy(cdev);
+	struct sas_rphy *rphy = transport_class_to_rphy(dev);
 	struct sas_phy *phy = dev_to_phy(rphy->dev.parent);
 	struct Scsi_Host *shost = dev_to_shost(phy->dev.parent);
 	struct sas_internal *i = to_sas_internal(shost->transportt);
@@ -1103,7 +1120,7 @@ show_sas_rphy_bay_identifier(struct class_device *cdev, char *buf)
 	return sprintf(buf, "%d\n", val);
 }
 
-static SAS_CLASS_DEVICE_ATTR(rphy, bay_identifier, S_IRUGO,
+static SAS_DEVICE_ATTR(rphy, bay_identifier, S_IRUGO,
 		show_sas_rphy_bay_identifier, NULL);
 
 sas_rphy_protocol_attr(identify.initiator_port_protocols,
@@ -1161,9 +1178,10 @@ static DECLARE_TRANSPORT_CLASS(sas_end_dev_class,
 
 #define sas_end_dev_show_simple(field, name, format_string, cast)	\
 static ssize_t								\
-show_sas_end_dev_##name(struct class_device *cdev, char *buf)		\
+show_sas_end_dev_##name(struct device *dev, 				\
+			struct device_attribute *attr, char *buf)	\
 {									\
-	struct sas_rphy *rphy = transport_class_to_rphy(cdev);		\
+	struct sas_rphy *rphy = transport_class_to_rphy(dev);		\
 	struct sas_end_device *rdev = rphy_to_end_device(rphy);		\
 									\
 	return snprintf(buf, 20, format_string, cast rdev->field);	\
@@ -1171,7 +1189,7 @@ show_sas_end_dev_##name(struct class_device *cdev, char *buf)		\
 
 #define sas_end_dev_simple_attr(field, name, format_string, type)	\
 	sas_end_dev_show_simple(field, name, format_string, (type))	\
-static SAS_CLASS_DEVICE_ATTR(end_dev, name, S_IRUGO, 			\
+static SAS_DEVICE_ATTR(end_dev, name, S_IRUGO, 			\
 		show_sas_end_dev_##name, NULL)
 
 sas_end_dev_simple_attr(ready_led_meaning, ready_led_meaning, "%d\n", int);
@@ -1185,9 +1203,10 @@ static DECLARE_TRANSPORT_CLASS(sas_expander_class,
 
 #define sas_expander_show_simple(field, name, format_string, cast)	\
 static ssize_t								\
-show_sas_expander_##name(struct class_device *cdev, char *buf)		\
+show_sas_expander_##name(struct device *dev, 				\
+			 struct device_attribute *attr, char *buf)	\
 {									\
-	struct sas_rphy *rphy = transport_class_to_rphy(cdev);		\
+	struct sas_rphy *rphy = transport_class_to_rphy(dev);		\
 	struct sas_expander_device *edev = rphy_to_expander_device(rphy); \
 									\
 	return snprintf(buf, 20, format_string, cast edev->field);	\
@@ -1195,7 +1214,7 @@ show_sas_expander_##name(struct class_device *cdev, char *buf)		\
 
 #define sas_expander_simple_attr(field, name, format_string, type)	\
 	sas_expander_show_simple(field, name, format_string, (type))	\
-static SAS_CLASS_DEVICE_ATTR(expander, name, S_IRUGO, 			\
+static SAS_DEVICE_ATTR(expander, name, S_IRUGO, 			\
 		show_sas_expander_##name, NULL)
 
 sas_expander_simple_attr(vendor_id, vendor_id, "%s\n", char *);
@@ -1554,14 +1573,14 @@ static int sas_user_scan(struct Scsi_Host *shost, uint channel,
  */
 
 #define SETUP_TEMPLATE(attrb, field, perm, test)			\
-	i->private_##attrb[count] = class_device_attr_##field;		\
+	i->private_##attrb[count] = dev_attr_##field;		\
 	i->private_##attrb[count].attr.mode = perm;			\
 	i->attrb[count] = &i->private_##attrb[count];			\
 	if (test)							\
 		count++
 
 #define SETUP_TEMPLATE_RW(attrb, field, perm, test, ro_test, ro_perm)	\
-	i->private_##attrb[count] = class_device_attr_##field;		\
+	i->private_##attrb[count] = dev_attr_##field;		\
 	i->private_##attrb[count].attr.mode = perm;			\
 	if (ro_test) {							\
 		i->private_##attrb[count].attr.mode = ro_perm;		\
diff --git a/drivers/scsi/scsi_transport_spi.c b/drivers/scsi/scsi_transport_spi.c
index 1fb60313a516..bc12b5d5d676 100644
--- a/drivers/scsi/scsi_transport_spi.c
+++ b/drivers/scsi/scsi_transport_spi.c
@@ -158,7 +158,7 @@ static inline enum spi_signal_type spi_signal_to_value(const char *name)
 }
 
 static int spi_host_setup(struct transport_container *tc, struct device *dev,
-			  struct class_device *cdev)
+			  struct device *cdev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 
@@ -169,7 +169,7 @@ static int spi_host_setup(struct transport_container *tc, struct device *dev,
 
 static int spi_host_configure(struct transport_container *tc,
 			      struct device *dev,
-			      struct class_device *cdev);
+			      struct device *cdev);
 
 static DECLARE_TRANSPORT_CLASS(spi_host_class,
 			       "spi_host",
@@ -195,11 +195,11 @@ static int spi_host_match(struct attribute_container *cont,
 
 static int spi_target_configure(struct transport_container *tc,
 				struct device *dev,
-				struct class_device *cdev);
+				struct device *cdev);
 
 static int spi_device_configure(struct transport_container *tc,
 				struct device *dev,
-				struct class_device *cdev)
+				struct device *cdev)
 {
 	struct scsi_device *sdev = to_scsi_device(dev);
 	struct scsi_target *starget = sdev->sdev_target;
@@ -219,7 +219,7 @@ static int spi_device_configure(struct transport_container *tc,
 
 static int spi_setup_transport_attrs(struct transport_container *tc,
 				     struct device *dev,
-				     struct class_device *cdev)
+				     struct device *cdev)
 {
 	struct scsi_target *starget = to_scsi_target(dev);
 
@@ -248,9 +248,10 @@ static int spi_setup_transport_attrs(struct transport_container *tc,
 #define spi_transport_show_simple(field, format_string)			\
 									\
 static ssize_t								\
-show_spi_transport_##field(struct class_device *cdev, char *buf)	\
+show_spi_transport_##field(struct device *dev, 			\
+			   struct device_attribute *attr, char *buf)	\
 {									\
-	struct scsi_target *starget = transport_class_to_starget(cdev);	\
+	struct scsi_target *starget = transport_class_to_starget(dev);	\
 	struct spi_transport_attrs *tp;					\
 									\
 	tp = (struct spi_transport_attrs *)&starget->starget_data;	\
@@ -260,11 +261,12 @@ show_spi_transport_##field(struct class_device *cdev, char *buf)	\
 #define spi_transport_store_simple(field, format_string)		\
 									\
 static ssize_t								\
-store_spi_transport_##field(struct class_device *cdev, const char *buf, \
-			    size_t count)				\
+store_spi_transport_##field(struct device *dev, 			\
+			    struct device_attribute *attr, 		\
+			    const char *buf, size_t count)		\
 {									\
 	int val;							\
-	struct scsi_target *starget = transport_class_to_starget(cdev);	\
+	struct scsi_target *starget = transport_class_to_starget(dev);	\
 	struct spi_transport_attrs *tp;					\
 									\
 	tp = (struct spi_transport_attrs *)&starget->starget_data;	\
@@ -276,9 +278,10 @@ store_spi_transport_##field(struct class_device *cdev, const char *buf, \
 #define spi_transport_show_function(field, format_string)		\
 									\
 static ssize_t								\
-show_spi_transport_##field(struct class_device *cdev, char *buf)	\
+show_spi_transport_##field(struct device *dev, 			\
+			   struct device_attribute *attr, char *buf)	\
 {									\
-	struct scsi_target *starget = transport_class_to_starget(cdev);	\
+	struct scsi_target *starget = transport_class_to_starget(dev);	\
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);	\
 	struct spi_transport_attrs *tp;					\
 	struct spi_internal *i = to_spi_internal(shost->transportt);	\
@@ -290,11 +293,12 @@ show_spi_transport_##field(struct class_device *cdev, char *buf)	\
 
 #define spi_transport_store_function(field, format_string)		\
 static ssize_t								\
-store_spi_transport_##field(struct class_device *cdev, const char *buf, \
-			    size_t count)				\
+store_spi_transport_##field(struct device *dev, 			\
+			    struct device_attribute *attr,		\
+			    const char *buf, size_t count)		\
 {									\
 	int val;							\
-	struct scsi_target *starget = transport_class_to_starget(cdev);	\
+	struct scsi_target *starget = transport_class_to_starget(dev);	\
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);	\
 	struct spi_internal *i = to_spi_internal(shost->transportt);	\
 									\
@@ -307,11 +311,12 @@ store_spi_transport_##field(struct class_device *cdev, const char *buf, \
 
 #define spi_transport_store_max(field, format_string)			\
 static ssize_t								\
-store_spi_transport_##field(struct class_device *cdev, const char *buf, \
-			    size_t count)				\
+store_spi_transport_##field(struct device *dev, 			\
+			    struct device_attribute *attr,		\
+			    const char *buf, size_t count)		\
 {									\
 	int val;							\
-	struct scsi_target *starget = transport_class_to_starget(cdev);	\
+	struct scsi_target *starget = transport_class_to_starget(dev);	\
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);	\
 	struct spi_internal *i = to_spi_internal(shost->transportt);	\
 	struct spi_transport_attrs *tp					\
@@ -329,24 +334,24 @@ store_spi_transport_##field(struct class_device *cdev, const char *buf, \
 #define spi_transport_rd_attr(field, format_string)			\
 	spi_transport_show_function(field, format_string)		\
 	spi_transport_store_function(field, format_string)		\
-static CLASS_DEVICE_ATTR(field, S_IRUGO,				\
-			 show_spi_transport_##field,			\
-			 store_spi_transport_##field);
+static DEVICE_ATTR(field, S_IRUGO,				\
+		   show_spi_transport_##field,			\
+		   store_spi_transport_##field);
 
 #define spi_transport_simple_attr(field, format_string)			\
 	spi_transport_show_simple(field, format_string)			\
 	spi_transport_store_simple(field, format_string)		\
-static CLASS_DEVICE_ATTR(field, S_IRUGO,				\
-			 show_spi_transport_##field,			\
-			 store_spi_transport_##field);
+static DEVICE_ATTR(field, S_IRUGO,				\
+		   show_spi_transport_##field,			\
+		   store_spi_transport_##field);
 
 #define spi_transport_max_attr(field, format_string)			\
 	spi_transport_show_function(field, format_string)		\
 	spi_transport_store_max(field, format_string)			\
 	spi_transport_simple_attr(max_##field, format_string)		\
-static CLASS_DEVICE_ATTR(field, S_IRUGO,				\
-			 show_spi_transport_##field,			\
-			 store_spi_transport_##field);
+static DEVICE_ATTR(field, S_IRUGO,				\
+		   show_spi_transport_##field,			\
+		   store_spi_transport_##field);
 
 /* The Parallel SCSI Tranport Attributes: */
 spi_transport_max_attr(offset, "%d\n");
@@ -370,14 +375,15 @@ static int child_iter(struct device *dev, void *data)
 }
 
 static ssize_t
-store_spi_revalidate(struct class_device *cdev, const char *buf, size_t count)
+store_spi_revalidate(struct device *dev, struct device_attribute *attr,
+		     const char *buf, size_t count)
 {
-	struct scsi_target *starget = transport_class_to_starget(cdev);
+	struct scsi_target *starget = transport_class_to_starget(dev);
 
 	device_for_each_child(&starget->dev, NULL, child_iter);
 	return count;
 }
-static CLASS_DEVICE_ATTR(revalidate, S_IWUSR, NULL, store_spi_revalidate);
+static DEVICE_ATTR(revalidate, S_IWUSR, NULL, store_spi_revalidate);
 
 /* Translate the period into ns according to the current spec
  * for SDTR/PPR messages */
@@ -412,7 +418,7 @@ show_spi_transport_period_helper(char *buf, int period)
 }
 
 static ssize_t
-store_spi_transport_period_helper(struct class_device *cdev, const char *buf,
+store_spi_transport_period_helper(struct device *dev, const char *buf,
 				  size_t count, int *periodp)
 {
 	int j, picosec, period = -1;
@@ -449,9 +455,10 @@ store_spi_transport_period_helper(struct class_device *cdev, const char *buf,
 }
 
 static ssize_t
-show_spi_transport_period(struct class_device *cdev, char *buf)
+show_spi_transport_period(struct device *dev,
+			  struct device_attribute *attr, char *buf)
 {
-	struct scsi_target *starget = transport_class_to_starget(cdev);
+	struct scsi_target *starget = transport_class_to_starget(dev);
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);
 	struct spi_internal *i = to_spi_internal(shost->transportt);
 	struct spi_transport_attrs *tp =
@@ -464,8 +471,8 @@ show_spi_transport_period(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-store_spi_transport_period(struct class_device *cdev, const char *buf,
-			    size_t count)
+store_spi_transport_period(struct device *cdev, struct device_attribute *attr,
+			   const char *buf, size_t count)
 {
 	struct scsi_target *starget = transport_class_to_starget(cdev);
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);
@@ -487,12 +494,13 @@ store_spi_transport_period(struct class_device *cdev, const char *buf,
 	return retval;
 }
 
-static CLASS_DEVICE_ATTR(period, S_IRUGO,
-			 show_spi_transport_period,
-			 store_spi_transport_period);
+static DEVICE_ATTR(period, S_IRUGO,
+		   show_spi_transport_period,
+		   store_spi_transport_period);
 
 static ssize_t
-show_spi_transport_min_period(struct class_device *cdev, char *buf)
+show_spi_transport_min_period(struct device *cdev,
+			      struct device_attribute *attr, char *buf)
 {
 	struct scsi_target *starget = transport_class_to_starget(cdev);
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);
@@ -507,8 +515,9 @@ show_spi_transport_min_period(struct class_device *cdev, char *buf)
 }
 
 static ssize_t
-store_spi_transport_min_period(struct class_device *cdev, const char *buf,
-			    size_t count)
+store_spi_transport_min_period(struct device *cdev,
+			       struct device_attribute *attr,
+			       const char *buf, size_t count)
 {
 	struct scsi_target *starget = transport_class_to_starget(cdev);
 	struct spi_transport_attrs *tp =
@@ -519,12 +528,14 @@ store_spi_transport_min_period(struct class_device *cdev, const char *buf,
 }
 
 
-static CLASS_DEVICE_ATTR(min_period, S_IRUGO,
-			 show_spi_transport_min_period,
-			 store_spi_transport_min_period);
+static DEVICE_ATTR(min_period, S_IRUGO,
+		   show_spi_transport_min_period,
+		   store_spi_transport_min_period);
 
 
-static ssize_t show_spi_host_signalling(struct class_device *cdev, char *buf)
+static ssize_t show_spi_host_signalling(struct device *cdev,
+					struct device_attribute *attr,
+					char *buf)
 {
 	struct Scsi_Host *shost = transport_class_to_shost(cdev);
 	struct spi_internal *i = to_spi_internal(shost->transportt);
@@ -534,10 +545,11 @@ static ssize_t show_spi_host_signalling(struct class_device *cdev, char *buf)
 
 	return sprintf(buf, "%s\n", spi_signal_to_string(spi_signalling(shost)));
 }
-static ssize_t store_spi_host_signalling(struct class_device *cdev,
+static ssize_t store_spi_host_signalling(struct device *dev,
+					 struct device_attribute *attr,
 					 const char *buf, size_t count)
 {
-	struct Scsi_Host *shost = transport_class_to_shost(cdev);
+	struct Scsi_Host *shost = transport_class_to_shost(dev);
 	struct spi_internal *i = to_spi_internal(shost->transportt);
 	enum spi_signal_type type = spi_signal_to_value(buf);
 
@@ -549,9 +561,9 @@ static ssize_t store_spi_host_signalling(struct class_device *cdev,
 
 	return count;
 }
-static CLASS_DEVICE_ATTR(signalling, S_IRUGO,
-			 show_spi_host_signalling,
-			 store_spi_host_signalling);
+static DEVICE_ATTR(signalling, S_IRUGO,
+		   show_spi_host_signalling,
+		   store_spi_host_signalling);
 
 #define DV_SET(x, y)			\
 	if(i->f->set_##x)		\
@@ -1334,7 +1346,7 @@ static DECLARE_ANON_TRANSPORT_CLASS(spi_device_class,
 				    spi_device_configure);
 
 static struct attribute *host_attributes[] = {
-	&class_device_attr_signalling.attr,
+	&dev_attr_signalling.attr,
 	NULL
 };
 
@@ -1344,12 +1356,12 @@ static struct attribute_group host_attribute_group = {
 
 static int spi_host_configure(struct transport_container *tc,
 			      struct device *dev,
-			      struct class_device *cdev)
+			      struct device *cdev)
 {
 	struct kobject *kobj = &cdev->kobj;
 	struct Scsi_Host *shost = transport_class_to_shost(cdev);
 	struct spi_internal *si = to_spi_internal(shost->transportt);
-	struct attribute *attr = &class_device_attr_signalling.attr;
+	struct attribute *attr = &dev_attr_signalling.attr;
 	int rc = 0;
 
 	if (si->f->set_signalling)
@@ -1368,76 +1380,75 @@ static int spi_host_configure(struct transport_container *tc,
 static int target_attribute_is_visible(struct kobject *kobj,
 				       struct attribute *attr, int i)
 {
-	struct class_device *cdev =
-		container_of(kobj, struct class_device, kobj);
+	struct device *cdev = container_of(kobj, struct device, kobj);
 	struct scsi_target *starget = transport_class_to_starget(cdev);
 	struct Scsi_Host *shost = transport_class_to_shost(cdev);
 	struct spi_internal *si = to_spi_internal(shost->transportt);
 
-	if (attr == &class_device_attr_period.attr &&
+	if (attr == &dev_attr_period.attr &&
 	    spi_support_sync(starget))
 		return TARGET_ATTRIBUTE_HELPER(period);
-	else if (attr == &class_device_attr_min_period.attr &&
+	else if (attr == &dev_attr_min_period.attr &&
 		 spi_support_sync(starget))
 		return TARGET_ATTRIBUTE_HELPER(period);
-	else if (attr == &class_device_attr_offset.attr &&
+	else if (attr == &dev_attr_offset.attr &&
 		 spi_support_sync(starget))
 		return TARGET_ATTRIBUTE_HELPER(offset);
-	else if (attr == &class_device_attr_max_offset.attr &&
+	else if (attr == &dev_attr_max_offset.attr &&
 		 spi_support_sync(starget))
 		return TARGET_ATTRIBUTE_HELPER(offset);
-	else if (attr == &class_device_attr_width.attr &&
+	else if (attr == &dev_attr_width.attr &&
 		 spi_support_wide(starget))
 		return TARGET_ATTRIBUTE_HELPER(width);
-	else if (attr == &class_device_attr_max_width.attr &&
+	else if (attr == &dev_attr_max_width.attr &&
 		 spi_support_wide(starget))
 		return TARGET_ATTRIBUTE_HELPER(width);
-	else if (attr == &class_device_attr_iu.attr &&
+	else if (attr == &dev_attr_iu.attr &&
 		 spi_support_ius(starget))
 		return TARGET_ATTRIBUTE_HELPER(iu);
-	else if (attr == &class_device_attr_dt.attr &&
+	else if (attr == &dev_attr_dt.attr &&
 		 spi_support_dt(starget))
 		return TARGET_ATTRIBUTE_HELPER(dt);
-	else if (attr == &class_device_attr_qas.attr &&
+	else if (attr == &dev_attr_qas.attr &&
 		 spi_support_qas(starget))
 		return TARGET_ATTRIBUTE_HELPER(qas);
-	else if (attr == &class_device_attr_wr_flow.attr &&
+	else if (attr == &dev_attr_wr_flow.attr &&
 		 spi_support_ius(starget))
 		return TARGET_ATTRIBUTE_HELPER(wr_flow);
-	else if (attr == &class_device_attr_rd_strm.attr &&
+	else if (attr == &dev_attr_rd_strm.attr &&
 		 spi_support_ius(starget))
 		return TARGET_ATTRIBUTE_HELPER(rd_strm);
-	else if (attr == &class_device_attr_rti.attr &&
+	else if (attr == &dev_attr_rti.attr &&
 		 spi_support_ius(starget))
 		return TARGET_ATTRIBUTE_HELPER(rti);
-	else if (attr == &class_device_attr_pcomp_en.attr &&
+	else if (attr == &dev_attr_pcomp_en.attr &&
 		 spi_support_ius(starget))
 		return TARGET_ATTRIBUTE_HELPER(pcomp_en);
-	else if (attr == &class_device_attr_hold_mcs.attr &&
+	else if (attr == &dev_attr_hold_mcs.attr &&
 		 spi_support_ius(starget))
 		return TARGET_ATTRIBUTE_HELPER(hold_mcs);
-	else if (attr == &class_device_attr_revalidate.attr)
+	else if (attr == &dev_attr_revalidate.attr)
 		return 1;
 
 	return 0;
 }
 
 static struct attribute *target_attributes[] = {
-	&class_device_attr_period.attr,
-	&class_device_attr_min_period.attr,
-	&class_device_attr_offset.attr,
-	&class_device_attr_max_offset.attr,
-	&class_device_attr_width.attr,
-	&class_device_attr_max_width.attr,
-	&class_device_attr_iu.attr,
-	&class_device_attr_dt.attr,
-	&class_device_attr_qas.attr,
-	&class_device_attr_wr_flow.attr,
-	&class_device_attr_rd_strm.attr,
-	&class_device_attr_rti.attr,
-	&class_device_attr_pcomp_en.attr,
-	&class_device_attr_hold_mcs.attr,
-	&class_device_attr_revalidate.attr,
+	&dev_attr_period.attr,
+	&dev_attr_min_period.attr,
+	&dev_attr_offset.attr,
+	&dev_attr_max_offset.attr,
+	&dev_attr_width.attr,
+	&dev_attr_max_width.attr,
+	&dev_attr_iu.attr,
+	&dev_attr_dt.attr,
+	&dev_attr_qas.attr,
+	&dev_attr_wr_flow.attr,
+	&dev_attr_rd_strm.attr,
+	&dev_attr_rti.attr,
+	&dev_attr_pcomp_en.attr,
+	&dev_attr_hold_mcs.attr,
+	&dev_attr_revalidate.attr,
 	NULL
 };
 
@@ -1448,7 +1459,7 @@ static struct attribute_group target_attribute_group = {
 
 static int spi_target_configure(struct transport_container *tc,
 				struct device *dev,
-				struct class_device *cdev)
+				struct device *cdev)
 {
 	struct kobject *kobj = &cdev->kobj;
 	int i;
@@ -1462,7 +1473,7 @@ static int spi_target_configure(struct transport_container *tc,
 		 * to ignore, sysfs also does a WARN_ON and dumps a trace,
 		 * which is bad, so temporarily, skip attributes that are
 		 * already visible (the revalidate one) */
-		if (j && attr != &class_device_attr_revalidate.attr)
+		if (j && attr != &dev_attr_revalidate.attr)
 			rc = sysfs_add_file_to_group(kobj, attr,
 						target_attribute_group.name);
 		/* and make the attribute writeable if we have a set
diff --git a/drivers/scsi/scsi_transport_srp.c b/drivers/scsi/scsi_transport_srp.c
index 2445c98ae95e..8a7af951d98a 100644
--- a/drivers/scsi/scsi_transport_srp.c
+++ b/drivers/scsi/scsi_transport_srp.c
@@ -44,20 +44,20 @@ struct srp_internal {
 	struct scsi_transport_template t;
 	struct srp_function_template *f;
 
-	struct class_device_attribute *host_attrs[SRP_HOST_ATTRS + 1];
+	struct device_attribute *host_attrs[SRP_HOST_ATTRS + 1];
 
-	struct class_device_attribute *rport_attrs[SRP_RPORT_ATTRS + 1];
-	struct class_device_attribute private_rport_attrs[SRP_RPORT_ATTRS];
+	struct device_attribute *rport_attrs[SRP_RPORT_ATTRS + 1];
+	struct device_attribute private_rport_attrs[SRP_RPORT_ATTRS];
 	struct transport_container rport_attr_cont;
 };
 
 #define to_srp_internal(tmpl) container_of(tmpl, struct srp_internal, t)
 
 #define	dev_to_rport(d)	container_of(d, struct srp_rport, dev)
-#define transport_class_to_srp_rport(cdev) dev_to_rport((cdev)->dev)
+#define transport_class_to_srp_rport(dev) dev_to_rport((dev)->parent)
 
 static int srp_host_setup(struct transport_container *tc, struct device *dev,
-			  struct class_device *cdev)
+			  struct device *cdev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 	struct srp_host_attrs *srp_host = to_srp_host_attrs(shost);
@@ -73,7 +73,7 @@ static DECLARE_TRANSPORT_CLASS(srp_rport_class, "srp_remote_ports",
 			       NULL, NULL, NULL);
 
 #define SETUP_TEMPLATE(attrb, field, perm, test, ro_test, ro_perm)	\
-	i->private_##attrb[count] = class_device_attr_##field;		\
+	i->private_##attrb[count] = dev_attr_##field;		\
 	i->private_##attrb[count].attr.mode = perm;			\
 	if (ro_test) {							\
 		i->private_##attrb[count].attr.mode = ro_perm;		\
@@ -100,13 +100,14 @@ static DECLARE_TRANSPORT_CLASS(srp_rport_class, "srp_remote_ports",
 	"%02x:%02x:%02x:%02x:%02x:%02x:%02x:%02x"
 
 static ssize_t
-show_srp_rport_id(struct class_device *cdev, char *buf)
+show_srp_rport_id(struct device *dev, struct device_attribute *attr,
+		  char *buf)
 {
-	struct srp_rport *rport = transport_class_to_srp_rport(cdev);
+	struct srp_rport *rport = transport_class_to_srp_rport(dev);
 	return sprintf(buf, SRP_PID_FMT "\n", SRP_PID(rport));
 }
 
-static CLASS_DEVICE_ATTR(port_id, S_IRUGO, show_srp_rport_id, NULL);
+static DEVICE_ATTR(port_id, S_IRUGO, show_srp_rport_id, NULL);
 
 static const struct {
 	u32 value;
@@ -117,9 +118,10 @@ static const struct {
 };
 
 static ssize_t
-show_srp_rport_roles(struct class_device *cdev, char *buf)
+show_srp_rport_roles(struct device *dev, struct device_attribute *attr,
+		     char *buf)
 {
-	struct srp_rport *rport = transport_class_to_srp_rport(cdev);
+	struct srp_rport *rport = transport_class_to_srp_rport(dev);
 	int i;
 	char *name = NULL;
 
@@ -131,7 +133,7 @@ show_srp_rport_roles(struct class_device *cdev, char *buf)
 	return sprintf(buf, "%s\n", name ? : "unknown");
 }
 
-static CLASS_DEVICE_ATTR(roles, S_IRUGO, show_srp_rport_roles, NULL);
+static DEVICE_ATTR(roles, S_IRUGO, show_srp_rport_roles, NULL);
 
 static void srp_rport_release(struct device *dev)
 {
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index 5fe7aaed904c..3cea17dd5dba 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -95,7 +95,7 @@ static int sd_resume(struct device *);
 static void sd_rescan(struct device *);
 static int sd_done(struct scsi_cmnd *);
 static void sd_read_capacity(struct scsi_disk *sdkp, unsigned char *buffer);
-static void scsi_disk_release(struct class_device *cdev);
+static void scsi_disk_release(struct device *cdev);
 static void sd_print_sense_hdr(struct scsi_disk *, struct scsi_sense_hdr *);
 static void sd_print_result(struct scsi_disk *, int);
 
@@ -112,11 +112,12 @@ static const char *sd_cache_types[] = {
 	"write back, no read (daft)"
 };
 
-static ssize_t sd_store_cache_type(struct class_device *cdev, const char *buf,
-				   size_t count)
+static ssize_t
+sd_store_cache_type(struct device *dev, struct device_attribute *attr,
+		    const char *buf, size_t count)
 {
 	int i, ct = -1, rcd, wce, sp;
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 	struct scsi_device *sdp = sdkp->device;
 	char buffer[64];
 	char *buffer_data;
@@ -163,10 +164,11 @@ static ssize_t sd_store_cache_type(struct class_device *cdev, const char *buf,
 	return count;
 }
 
-static ssize_t sd_store_manage_start_stop(struct class_device *cdev,
-					  const char *buf, size_t count)
+static ssize_t
+sd_store_manage_start_stop(struct device *dev, struct device_attribute *attr,
+			   const char *buf, size_t count)
 {
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 	struct scsi_device *sdp = sdkp->device;
 
 	if (!capable(CAP_SYS_ADMIN))
@@ -177,10 +179,11 @@ static ssize_t sd_store_manage_start_stop(struct class_device *cdev,
 	return count;
 }
 
-static ssize_t sd_store_allow_restart(struct class_device *cdev, const char *buf,
-				      size_t count)
+static ssize_t
+sd_store_allow_restart(struct device *dev, struct device_attribute *attr,
+		       const char *buf, size_t count)
 {
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 	struct scsi_device *sdp = sdkp->device;
 
 	if (!capable(CAP_SYS_ADMIN))
@@ -194,37 +197,44 @@ static ssize_t sd_store_allow_restart(struct class_device *cdev, const char *buf
 	return count;
 }
 
-static ssize_t sd_show_cache_type(struct class_device *cdev, char *buf)
+static ssize_t
+sd_show_cache_type(struct device *dev, struct device_attribute *attr,
+		   char *buf)
 {
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 	int ct = sdkp->RCD + 2*sdkp->WCE;
 
 	return snprintf(buf, 40, "%s\n", sd_cache_types[ct]);
 }
 
-static ssize_t sd_show_fua(struct class_device *cdev, char *buf)
+static ssize_t
+sd_show_fua(struct device *dev, struct device_attribute *attr, char *buf)
 {
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 
 	return snprintf(buf, 20, "%u\n", sdkp->DPOFUA);
 }
 
-static ssize_t sd_show_manage_start_stop(struct class_device *cdev, char *buf)
+static ssize_t
+sd_show_manage_start_stop(struct device *dev, struct device_attribute *attr,
+			  char *buf)
 {
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 	struct scsi_device *sdp = sdkp->device;
 
 	return snprintf(buf, 20, "%u\n", sdp->manage_start_stop);
 }
 
-static ssize_t sd_show_allow_restart(struct class_device *cdev, char *buf)
+static ssize_t
+sd_show_allow_restart(struct device *dev, struct device_attribute *attr,
+		      char *buf)
 {
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 
 	return snprintf(buf, 40, "%d\n", sdkp->device->allow_restart);
 }
 
-static struct class_device_attribute sd_disk_attrs[] = {
+static struct device_attribute sd_disk_attrs[] = {
 	__ATTR(cache_type, S_IRUGO|S_IWUSR, sd_show_cache_type,
 	       sd_store_cache_type),
 	__ATTR(FUA, S_IRUGO, sd_show_fua, NULL),
@@ -238,8 +248,8 @@ static struct class_device_attribute sd_disk_attrs[] = {
 static struct class sd_disk_class = {
 	.name		= "scsi_disk",
 	.owner		= THIS_MODULE,
-	.release	= scsi_disk_release,
-	.class_dev_attrs = sd_disk_attrs,
+	.dev_release	= scsi_disk_release,
+	.dev_attrs	= sd_disk_attrs,
 };
 
 static struct scsi_driver sd_template = {
@@ -297,7 +307,7 @@ static struct scsi_disk *__scsi_disk_get(struct gendisk *disk)
 	if (disk->private_data) {
 		sdkp = scsi_disk(disk);
 		if (scsi_device_get(sdkp->device) == 0)
-			class_device_get(&sdkp->cdev);
+			get_device(&sdkp->dev);
 		else
 			sdkp = NULL;
 	}
@@ -331,7 +341,7 @@ static void scsi_disk_put(struct scsi_disk *sdkp)
 	struct scsi_device *sdev = sdkp->device;
 
 	mutex_lock(&sd_ref_mutex);
-	class_device_put(&sdkp->cdev);
+	put_device(&sdkp->dev);
 	scsi_device_put(sdev);
 	mutex_unlock(&sd_ref_mutex);
 }
@@ -1663,12 +1673,12 @@ static int sd_probe(struct device *dev)
 			sdp->timeout = SD_MOD_TIMEOUT;
 	}
 
-	class_device_initialize(&sdkp->cdev);
-	sdkp->cdev.dev = &sdp->sdev_gendev;
-	sdkp->cdev.class = &sd_disk_class;
-	strncpy(sdkp->cdev.class_id, sdp->sdev_gendev.bus_id, BUS_ID_SIZE);
+	device_initialize(&sdkp->dev);
+	sdkp->dev.parent = &sdp->sdev_gendev;
+	sdkp->dev.class = &sd_disk_class;
+	strncpy(sdkp->dev.bus_id, sdp->sdev_gendev.bus_id, BUS_ID_SIZE);
 
-	if (class_device_add(&sdkp->cdev))
+	if (device_add(&sdkp->dev))
 		goto out_put;
 
 	get_device(&sdp->sdev_gendev);
@@ -1734,13 +1744,13 @@ static int sd_remove(struct device *dev)
 {
 	struct scsi_disk *sdkp = dev_get_drvdata(dev);
 
-	class_device_del(&sdkp->cdev);
+	device_del(&sdkp->dev);
 	del_gendisk(sdkp->disk);
 	sd_shutdown(dev);
 
 	mutex_lock(&sd_ref_mutex);
 	dev_set_drvdata(dev, NULL);
-	class_device_put(&sdkp->cdev);
+	put_device(&sdkp->dev);
 	mutex_unlock(&sd_ref_mutex);
 
 	return 0;
@@ -1748,16 +1758,16 @@ static int sd_remove(struct device *dev)
 
 /**
  *	scsi_disk_release - Called to free the scsi_disk structure
- *	@cdev: pointer to embedded class device
+ *	@dev: pointer to embedded class device
  *
  *	sd_ref_mutex must be held entering this routine.  Because it is
  *	called on last put, you should always use the scsi_disk_get()
  *	scsi_disk_put() helpers which manipulate the semaphore directly
- *	and never do a direct class_device_put().
+ *	and never do a direct put_device.
  **/
-static void scsi_disk_release(struct class_device *cdev)
+static void scsi_disk_release(struct device *dev)
 {
-	struct scsi_disk *sdkp = to_scsi_disk(cdev);
+	struct scsi_disk *sdkp = to_scsi_disk(dev);
 	struct gendisk *disk = sdkp->disk;
 	
 	spin_lock(&sd_index_lock);
diff --git a/drivers/scsi/ses.c b/drivers/scsi/ses.c
index a6d96694d0a5..45df83b9d847 100644
--- a/drivers/scsi/ses.c
+++ b/drivers/scsi/ses.c
@@ -107,7 +107,7 @@ static int ses_set_page2_descriptor(struct enclosure_device *edev,
 				      unsigned char *desc)
 {
 	int i, j, count = 0, descriptor = ecomp->number;
-	struct scsi_device *sdev = to_scsi_device(edev->cdev.dev);
+	struct scsi_device *sdev = to_scsi_device(edev->edev.parent);
 	struct ses_device *ses_dev = edev->scratch;
 	unsigned char *type_ptr = ses_dev->page1 + 12 + ses_dev->page1[11];
 	unsigned char *desc_ptr = ses_dev->page2 + 8;
@@ -137,7 +137,7 @@ static unsigned char *ses_get_page2_descriptor(struct enclosure_device *edev,
 				      struct enclosure_component *ecomp)
 {
 	int i, j, count = 0, descriptor = ecomp->number;
-	struct scsi_device *sdev = to_scsi_device(edev->cdev.dev);
+	struct scsi_device *sdev = to_scsi_device(edev->edev.parent);
 	struct ses_device *ses_dev = edev->scratch;
 	unsigned char *type_ptr = ses_dev->page1 + 12 + ses_dev->page1[11];
 	unsigned char *desc_ptr = ses_dev->page2 + 8;
@@ -269,10 +269,10 @@ int ses_match_host(struct enclosure_device *edev, void *data)
 	struct ses_host_edev *sed = data;
 	struct scsi_device *sdev;
 
-	if (!scsi_is_sdev_device(edev->cdev.dev))
+	if (!scsi_is_sdev_device(edev->edev.parent))
 		return 0;
 
-	sdev = to_scsi_device(edev->cdev.dev);
+	sdev = to_scsi_device(edev->edev.parent);
 
 	if (sdev->host != sed->shost)
 		return 0;
@@ -407,10 +407,10 @@ static void ses_match_to_enclosure(struct enclosure_device *edev,
 
 #define INIT_ALLOC_SIZE 32
 
-static int ses_intf_add(struct class_device *cdev,
+static int ses_intf_add(struct device *cdev,
 			struct class_interface *intf)
 {
-	struct scsi_device *sdev = to_scsi_device(cdev->dev);
+	struct scsi_device *sdev = to_scsi_device(cdev->parent);
 	struct scsi_device *tmp_sdev;
 	unsigned char *buf = NULL, *hdr_buf, *type_ptr, *desc_ptr = NULL,
 		*addl_desc_ptr = NULL;
@@ -426,7 +426,7 @@ static int ses_intf_add(struct class_device *cdev,
 		edev = enclosure_find(&sdev->host->shost_gendev);
 		if (edev) {
 			ses_match_to_enclosure(edev, sdev);
-			class_device_put(&edev->cdev);
+			put_device(&edev->edev);
 		}
 		return -ENODEV;
 	}
@@ -515,7 +515,7 @@ static int ses_intf_add(struct class_device *cdev,
 	if (!scomp)
 		goto err_free;
 
-	edev = enclosure_register(cdev->dev, sdev->sdev_gendev.bus_id,
+	edev = enclosure_register(cdev->parent, sdev->sdev_gendev.bus_id,
 				  components, &ses_enclosure_callbacks);
 	if (IS_ERR(edev)) {
 		err = PTR_ERR(edev);
@@ -625,17 +625,17 @@ static int ses_remove(struct device *dev)
 	return 0;
 }
 
-static void ses_intf_remove(struct class_device *cdev,
+static void ses_intf_remove(struct device *cdev,
 			    struct class_interface *intf)
 {
-	struct scsi_device *sdev = to_scsi_device(cdev->dev);
+	struct scsi_device *sdev = to_scsi_device(cdev->parent);
 	struct enclosure_device *edev;
 	struct ses_device *ses_dev;
 
 	if (!scsi_device_enclosure(sdev))
 		return;
 
-	edev = enclosure_find(cdev->dev);
+	edev = enclosure_find(cdev->parent);
 	if (!edev)
 		return;
 
@@ -649,13 +649,13 @@ static void ses_intf_remove(struct class_device *cdev,
 
 	kfree(edev->component[0].scratch);
 
-	class_device_put(&edev->cdev);
+	put_device(&edev->edev);
 	enclosure_unregister(edev);
 }
 
 static struct class_interface ses_interface = {
-	.add	= ses_intf_add,
-	.remove	= ses_intf_remove,
+	.add_dev	= ses_intf_add,
+	.remove_dev	= ses_intf_remove,
 };
 
 static struct scsi_driver ses_template = {
diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c
index e5156aa6dd20..2029422bc04d 100644
--- a/drivers/scsi/sg.c
+++ b/drivers/scsi/sg.c
@@ -101,16 +101,16 @@ static int scatter_elem_sz_prev = SG_SCATTER_SZ;
 #define SG_SECTOR_SZ 512
 #define SG_SECTOR_MSK (SG_SECTOR_SZ - 1)
 
-static int sg_add(struct class_device *, struct class_interface *);
-static void sg_remove(struct class_device *, struct class_interface *);
+static int sg_add(struct device *, struct class_interface *);
+static void sg_remove(struct device *, struct class_interface *);
 
 static DEFINE_IDR(sg_index_idr);
 static DEFINE_RWLOCK(sg_index_lock);	/* Also used to lock
 							   file descriptor list for device */
 
 static struct class_interface sg_interface = {
-	.add		= sg_add,
-	.remove		= sg_remove,
+	.add_dev	= sg_add,
+	.remove_dev	= sg_remove,
 };
 
 typedef struct sg_scatter_hold { /* holding area for scsi scatter gather info */
@@ -1401,9 +1401,9 @@ static Sg_device *sg_alloc(struct gendisk *disk, struct scsi_device *scsidp)
 }
 
 static int
-sg_add(struct class_device *cl_dev, struct class_interface *cl_intf)
+sg_add(struct device *cl_dev, struct class_interface *cl_intf)
 {
-	struct scsi_device *scsidp = to_scsi_device(cl_dev->dev);
+	struct scsi_device *scsidp = to_scsi_device(cl_dev->parent);
 	struct gendisk *disk;
 	Sg_device *sdp = NULL;
 	struct cdev * cdev = NULL;
@@ -1439,19 +1439,19 @@ sg_add(struct class_device *cl_dev, struct class_interface *cl_intf)
 
 	sdp->cdev = cdev;
 	if (sg_sysfs_valid) {
-		struct class_device * sg_class_member;
+		struct device *sg_class_member;
 
-		sg_class_member = class_device_create(sg_sysfs_class, NULL,
-				MKDEV(SCSI_GENERIC_MAJOR, sdp->index),
-				cl_dev->dev, "%s",
-				disk->disk_name);
+		sg_class_member = device_create(sg_sysfs_class, cl_dev->parent,
+						MKDEV(SCSI_GENERIC_MAJOR,
+						      sdp->index),
+						"%s", disk->disk_name);
 		if (IS_ERR(sg_class_member)) {
 			printk(KERN_ERR "sg_add: "
-			       "class_device_create failed\n");
+			       "device_create failed\n");
 			error = PTR_ERR(sg_class_member);
 			goto cdev_add_err;
 		}
-		class_set_devdata(sg_class_member, sdp);
+		dev_set_drvdata(sg_class_member, sdp);
 		error = sysfs_create_link(&scsidp->sdev_gendev.kobj,
 					  &sg_class_member->kobj, "generic");
 		if (error)
@@ -1464,7 +1464,7 @@ sg_add(struct class_device *cl_dev, struct class_interface *cl_intf)
 		    "Attached scsi generic sg%d type %d\n", sdp->index,
 		    scsidp->type);
 
-	class_set_devdata(cl_dev, sdp);
+	dev_set_drvdata(cl_dev, sdp);
 
 	return 0;
 
@@ -1482,10 +1482,10 @@ out:
 }
 
 static void
-sg_remove(struct class_device *cl_dev, struct class_interface *cl_intf)
+sg_remove(struct device *cl_dev, struct class_interface *cl_intf)
 {
-	struct scsi_device *scsidp = to_scsi_device(cl_dev->dev);
-	Sg_device *sdp = class_get_devdata(cl_dev);
+	struct scsi_device *scsidp = to_scsi_device(cl_dev->parent);
+	Sg_device *sdp = dev_get_drvdata(cl_dev);
 	unsigned long iflags;
 	Sg_fd *sfp;
 	Sg_fd *tsfp;
@@ -1528,7 +1528,7 @@ sg_remove(struct class_device *cl_dev, struct class_interface *cl_intf)
 	write_unlock_irqrestore(&sg_index_lock, iflags);
 
 	sysfs_remove_link(&scsidp->sdev_gendev.kobj, "generic");
-	class_device_destroy(sg_sysfs_class, MKDEV(SCSI_GENERIC_MAJOR, sdp->index));
+	device_destroy(sg_sysfs_class, MKDEV(SCSI_GENERIC_MAJOR, sdp->index));
 	cdev_del(sdp->cdev);
 	sdp->cdev = NULL;
 	put_disk(sdp->disk);
diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c
index df83bea2c620..a860c3a9ae99 100644
--- a/drivers/scsi/st.c
+++ b/drivers/scsi/st.c
@@ -4108,9 +4108,9 @@ out_free_tape:
 			if (STm->cdevs[j]) {
 				if (cdev == STm->cdevs[j])
 					cdev = NULL;
-				class_device_destroy(st_sysfs_class,
-						     MKDEV(SCSI_TAPE_MAJOR,
-							   TAPE_MINOR(i, mode, j)));
+					device_destroy(st_sysfs_class,
+						       MKDEV(SCSI_TAPE_MAJOR,
+							     TAPE_MINOR(i, mode, j)));
 				cdev_del(STm->cdevs[j]);
 			}
 		}
@@ -4148,9 +4148,9 @@ static int st_remove(struct device *dev)
 					  "tape");
 			for (mode = 0; mode < ST_NBR_MODES; ++mode) {
 				for (j=0; j < 2; j++) {
-					class_device_destroy(st_sysfs_class,
-							     MKDEV(SCSI_TAPE_MAJOR,
-								   TAPE_MINOR(i, mode, j)));
+					device_destroy(st_sysfs_class,
+						       MKDEV(SCSI_TAPE_MAJOR,
+							     TAPE_MINOR(i, mode, j)));
 					cdev_del(tpnt->modes[mode].cdevs[j]);
 					tpnt->modes[mode].cdevs[j] = NULL;
 				}
@@ -4319,31 +4319,34 @@ static void do_remove_sysfs_files(void)
 
 
 /* The sysfs simple class interface */
-static ssize_t st_defined_show(struct class_device *class_dev, char *buf)
+static ssize_t
+st_defined_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
-	struct st_modedef *STm = (struct st_modedef *)class_get_devdata(class_dev);
+	struct st_modedef *STm = (struct st_modedef *)dev_get_drvdata(dev);
 	ssize_t l = 0;
 
 	l = snprintf(buf, PAGE_SIZE, "%d\n", STm->defined);
 	return l;
 }
 
-CLASS_DEVICE_ATTR(defined, S_IRUGO, st_defined_show, NULL);
+DEVICE_ATTR(defined, S_IRUGO, st_defined_show, NULL);
 
-static ssize_t st_defblk_show(struct class_device *class_dev, char *buf)
+static ssize_t
+st_defblk_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
-	struct st_modedef *STm = (struct st_modedef *)class_get_devdata(class_dev);
+	struct st_modedef *STm = (struct st_modedef *)dev_get_drvdata(dev);
 	ssize_t l = 0;
 
 	l = snprintf(buf, PAGE_SIZE, "%d\n", STm->default_blksize);
 	return l;
 }
 
-CLASS_DEVICE_ATTR(default_blksize, S_IRUGO, st_defblk_show, NULL);
+DEVICE_ATTR(default_blksize, S_IRUGO, st_defblk_show, NULL);
 
-static ssize_t st_defdensity_show(struct class_device *class_dev, char *buf)
+static ssize_t
+st_defdensity_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
-	struct st_modedef *STm = (struct st_modedef *)class_get_devdata(class_dev);
+	struct st_modedef *STm = (struct st_modedef *)dev_get_drvdata(dev);
 	ssize_t l = 0;
 	char *fmt;
 
@@ -4352,22 +4355,25 @@ static ssize_t st_defdensity_show(struct class_device *class_dev, char *buf)
 	return l;
 }
 
-CLASS_DEVICE_ATTR(default_density, S_IRUGO, st_defdensity_show, NULL);
+DEVICE_ATTR(default_density, S_IRUGO, st_defdensity_show, NULL);
 
-static ssize_t st_defcompression_show(struct class_device *class_dev, char *buf)
+static ssize_t
+st_defcompression_show(struct device *dev, struct device_attribute *attr,
+		       char *buf)
 {
-	struct st_modedef *STm = (struct st_modedef *)class_get_devdata(class_dev);
+	struct st_modedef *STm = (struct st_modedef *)dev_get_drvdata(dev);
 	ssize_t l = 0;
 
 	l = snprintf(buf, PAGE_SIZE, "%d\n", STm->default_compression - 1);
 	return l;
 }
 
-CLASS_DEVICE_ATTR(default_compression, S_IRUGO, st_defcompression_show, NULL);
+DEVICE_ATTR(default_compression, S_IRUGO, st_defcompression_show, NULL);
 
-static ssize_t st_options_show(struct class_device *class_dev, char *buf)
+static ssize_t
+st_options_show(struct device *dev, struct device_attribute *attr, char *buf)
 {
-	struct st_modedef *STm = (struct st_modedef *)class_get_devdata(class_dev);
+	struct st_modedef *STm = (struct st_modedef *)dev_get_drvdata(dev);
 	struct scsi_tape *STp;
 	int i, j, options;
 	ssize_t l = 0;
@@ -4403,13 +4409,13 @@ static ssize_t st_options_show(struct class_device *class_dev, char *buf)
 	return l;
 }
 
-CLASS_DEVICE_ATTR(options, S_IRUGO, st_options_show, NULL);
+DEVICE_ATTR(options, S_IRUGO, st_options_show, NULL);
 
 static int do_create_class_files(struct scsi_tape *STp, int dev_num, int mode)
 {
 	int i, rew, error;
 	char name[10];
-	struct class_device *st_class_member;
+	struct device *st_class_member;
 
 	for (rew=0; rew < 2; rew++) {
 		/* Make sure that the minor numbers corresponding to the four
@@ -4418,32 +4424,32 @@ static int do_create_class_files(struct scsi_tape *STp, int dev_num, int mode)
 		snprintf(name, 10, "%s%s%s", rew ? "n" : "",
 			 STp->disk->disk_name, st_formats[i]);
 		st_class_member =
-			class_device_create(st_sysfs_class, NULL,
-					    MKDEV(SCSI_TAPE_MAJOR,
-						  TAPE_MINOR(dev_num, mode, rew)),
-					    &STp->device->sdev_gendev, "%s", name);
+			device_create(st_sysfs_class, &STp->device->sdev_gendev,
+				      MKDEV(SCSI_TAPE_MAJOR,
+						TAPE_MINOR(dev_num, mode, rew)),
+				      "%s", name);
 		if (IS_ERR(st_class_member)) {
-			printk(KERN_WARNING "st%d: class_device_create failed\n",
+			printk(KERN_WARNING "st%d: device_create failed\n",
 			       dev_num);
 			error = PTR_ERR(st_class_member);
 			goto out;
 		}
-		class_set_devdata(st_class_member, &STp->modes[mode]);
+		dev_set_drvdata(st_class_member, &STp->modes[mode]);
 
-		error = class_device_create_file(st_class_member,
-					       &class_device_attr_defined);
+		error = device_create_file(st_class_member,
+					   &dev_attr_defined);
 		if (error) goto out;
-		error = class_device_create_file(st_class_member,
-					    &class_device_attr_default_blksize);
+		error = device_create_file(st_class_member,
+					   &dev_attr_default_blksize);
 		if (error) goto out;
-		error = class_device_create_file(st_class_member,
-					    &class_device_attr_default_density);
+		error = device_create_file(st_class_member,
+					   &dev_attr_default_density);
 		if (error) goto out;
-		error = class_device_create_file(st_class_member,
-				        &class_device_attr_default_compression);
+		error = device_create_file(st_class_member,
+					   &dev_attr_default_compression);
 		if (error) goto out;
-		error = class_device_create_file(st_class_member,
-				        &class_device_attr_options);
+		error = device_create_file(st_class_member,
+					   &dev_attr_options);
 		if (error) goto out;
 
 		if (mode == 0 && rew == 0) {
diff --git a/include/linux/attribute_container.h b/include/linux/attribute_container.h
index 574b201b99d8..794ad74b1d61 100644
--- a/include/linux/attribute_container.h
+++ b/include/linux/attribute_container.h
@@ -1,5 +1,5 @@
 /*
- * class_container.h - a generic container for all classes
+ * attribute_container.h - a generic container for all classes
  *
  * Copyright (c) 2005 - James Bottomley <James.Bottomley@steeleye.com>
  *
@@ -18,7 +18,7 @@ struct attribute_container {
 	struct klist		containers;
 	struct class		*class;
 	struct attribute_group	*grp;
-	struct class_device_attribute **attrs;
+	struct device_attribute **attrs;
 	int (*match)(struct attribute_container *, struct device *);
 #define	ATTRIBUTE_CONTAINER_NO_CLASSDEVS	0x01
 	unsigned long		flags;
@@ -41,31 +41,31 @@ int __must_check attribute_container_unregister(struct attribute_container *cont
 void attribute_container_create_device(struct device *dev,
 				       int (*fn)(struct attribute_container *,
 						 struct device *,
-						 struct class_device *));
+						 struct device *));
 void attribute_container_add_device(struct device *dev,
 				    int (*fn)(struct attribute_container *,
 					      struct device *,
-					      struct class_device *));
+					      struct device *));
 void attribute_container_remove_device(struct device *dev,
 				       void (*fn)(struct attribute_container *,
 						  struct device *,
-						  struct class_device *));
+						  struct device *));
 void attribute_container_device_trigger(struct device *dev, 
 					int (*fn)(struct attribute_container *,
 						  struct device *,
-						  struct class_device *));
+						  struct device *));
 void attribute_container_trigger(struct device *dev, 
 				 int (*fn)(struct attribute_container *,
 					   struct device *));
-int attribute_container_add_attrs(struct class_device *classdev);
-int attribute_container_add_class_device(struct class_device *classdev);
+int attribute_container_add_attrs(struct device *classdev);
+int attribute_container_add_class_device(struct device *classdev);
 int attribute_container_add_class_device_adapter(struct attribute_container *cont,
 						 struct device *dev,
-						 struct class_device *classdev);
-void attribute_container_remove_attrs(struct class_device *classdev);
-void attribute_container_class_device_del(struct class_device *classdev);
-struct attribute_container *attribute_container_classdev_to_container(struct class_device *);
-struct class_device *attribute_container_find_class_device(struct attribute_container *, struct device *);
-struct class_device_attribute **attribute_container_classdev_to_attrs(const struct class_device *classdev);
+						 struct device *classdev);
+void attribute_container_remove_attrs(struct device *classdev);
+void attribute_container_class_device_del(struct device *classdev);
+struct attribute_container *attribute_container_classdev_to_container(struct device *);
+struct device *attribute_container_find_class_device(struct attribute_container *, struct device *);
+struct device_attribute **attribute_container_classdev_to_attrs(const struct device *classdev);
 
 #endif
diff --git a/include/linux/bsg.h b/include/linux/bsg.h
index 60e377b520f8..e8406c55c6d3 100644
--- a/include/linux/bsg.h
+++ b/include/linux/bsg.h
@@ -55,7 +55,7 @@ struct sg_io_v4 {
 
 #if defined(CONFIG_BLK_DEV_BSG)
 struct bsg_class_device {
-	struct class_device *class_dev;
+	struct device *class_dev;
 	struct device *dev;
 	int minor;
 	struct request_queue *queue;
diff --git a/include/linux/enclosure.h b/include/linux/enclosure.h
index a5978f18ca40..4332442b1b57 100644
--- a/include/linux/enclosure.h
+++ b/include/linux/enclosure.h
@@ -82,7 +82,8 @@ struct enclosure_component_callbacks {
 
 struct enclosure_component {
 	void *scratch;
-	struct class_device cdev;
+	struct device cdev;
+	struct device *dev;
 	enum enclosure_component_type type;
 	int number;
 	int fault;
@@ -94,20 +95,20 @@ struct enclosure_component {
 struct enclosure_device {
 	void *scratch;
 	struct list_head node;
-	struct class_device cdev;
+	struct device edev;
 	struct enclosure_component_callbacks *cb;
 	int components;
 	struct enclosure_component component[0];
 };
 
 static inline struct enclosure_device *
-to_enclosure_device(struct class_device *dev)
+to_enclosure_device(struct device *dev)
 {
-	return container_of(dev, struct enclosure_device, cdev);
+	return container_of(dev, struct enclosure_device, edev);
 }
 
 static inline struct enclosure_component *
-to_enclosure_component(struct class_device *dev)
+to_enclosure_component(struct device *dev)
 {
 	return container_of(dev, struct enclosure_component, cdev);
 }
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 165734a2dd47..07ed56f7a767 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -443,7 +443,7 @@ enum link_pm {
 	MAX_PERFORMANCE,
 	MEDIUM_POWER,
 };
-extern struct class_device_attribute class_device_attr_link_power_management_policy;
+extern struct device_attribute dev_attr_link_power_management_policy;
 
 #ifdef CONFIG_ATA_SFF
 struct ata_ioports {
diff --git a/include/linux/raid_class.h b/include/linux/raid_class.h
index d22ad392242a..6b537f1ac96c 100644
--- a/include/linux/raid_class.h
+++ b/include/linux/raid_class.h
@@ -53,20 +53,20 @@ struct raid_data {
 #define DEFINE_RAID_ATTRIBUTE(type, attr)				      \
 static inline void							      \
 raid_set_##attr(struct raid_template *r, struct device *dev, type value) {    \
-	struct class_device *cdev =					      \
+	struct device *device =						      \
 		attribute_container_find_class_device(&r->raid_attrs.ac, dev);\
 	struct raid_data *rd;						      \
-	BUG_ON(!cdev);							      \
-	rd = class_get_devdata(cdev);					      \
+	BUG_ON(!device);						      \
+	rd = dev_get_drvdata(device);					      \
 	rd->attr = value;						      \
 }									      \
 static inline type							      \
 raid_get_##attr(struct raid_template *r, struct device *dev) {		      \
-	struct class_device *cdev =					      \
+	struct device *device =						      \
 		attribute_container_find_class_device(&r->raid_attrs.ac, dev);\
 	struct raid_data *rd;						      \
-	BUG_ON(!cdev);							      \
-	rd = class_get_devdata(cdev);					      \
+	BUG_ON(!device);						      \
+	rd = dev_get_drvdata(device);					      \
 	return rd->attr;						      \
 }
 
diff --git a/include/linux/transport_class.h b/include/linux/transport_class.h
index 6696cf79c4f7..eaec1ea9558e 100644
--- a/include/linux/transport_class.h
+++ b/include/linux/transport_class.h
@@ -17,11 +17,11 @@ struct transport_container;
 struct transport_class {
 	struct class class;
 	int (*setup)(struct transport_container *, struct device *,
-		     struct class_device *);
+		     struct device *);
 	int (*configure)(struct transport_container *, struct device *,
-			 struct class_device *);
+			 struct device *);
 	int (*remove)(struct transport_container *, struct device *,
-		      struct class_device *);
+		      struct device *);
 };
 
 #define DECLARE_TRANSPORT_CLASS(cls, nm, su, rm, cfg)			\
diff --git a/include/scsi/scsi_device.h b/include/scsi/scsi_device.h
index ab7acbe80960..b8b19e2f57bb 100644
--- a/include/scsi/scsi_device.h
+++ b/include/scsi/scsi_device.h
@@ -156,8 +156,8 @@ struct scsi_device {
 
 	int timeout;
 
-	struct device		sdev_gendev;
-	struct class_device	sdev_classdev;
+	struct device		sdev_gendev,
+				sdev_dev;
 
 	struct execute_work	ew; /* used to get process context on put */
 
@@ -167,9 +167,9 @@ struct scsi_device {
 #define	to_scsi_device(d)	\
 	container_of(d, struct scsi_device, sdev_gendev)
 #define	class_to_sdev(d)	\
-	container_of(d, struct scsi_device, sdev_classdev)
+	container_of(d, struct scsi_device, sdev_dev)
 #define transport_class_to_sdev(class_dev) \
-	to_scsi_device(class_dev->dev)
+	to_scsi_device(class_dev->parent)
 
 #define sdev_printk(prefix, sdev, fmt, a...)	\
 	dev_printk(prefix, &(sdev)->sdev_gendev, fmt, ##a)
@@ -220,7 +220,7 @@ static inline struct scsi_target *scsi_target(struct scsi_device *sdev)
 	return to_scsi_target(sdev->sdev_gendev.parent);
 }
 #define transport_class_to_starget(class_dev) \
-	to_scsi_target(class_dev->dev)
+	to_scsi_target(class_dev->parent)
 
 #define starget_printk(prefix, starget, fmt, a...)	\
 	dev_printk(prefix, &(starget)->dev, fmt, ##a)
diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h
index 49132862bfaa..d967d6dc7a28 100644
--- a/include/scsi/scsi_host.h
+++ b/include/scsi/scsi_host.h
@@ -470,7 +470,7 @@ struct scsi_host_template {
 	/*
 	 * Pointer to the sysfs class properties for this host, NULL terminated.
 	 */
-	struct class_device_attribute **shost_attrs;
+	struct device_attribute **shost_attrs;
 
 	/*
 	 * Pointer to the SCSI device properties for this host, NULL terminated.
@@ -655,8 +655,7 @@ struct Scsi_Host {
 	enum scsi_host_state shost_state;
 
 	/* ldm bits */
-	struct device		shost_gendev;
-	struct class_device	shost_classdev;
+	struct device		shost_gendev, shost_dev;
 
 	/*
 	 * List of hosts per template.
@@ -683,7 +682,7 @@ struct Scsi_Host {
 };
 
 #define		class_to_shost(d)	\
-	container_of(d, struct Scsi_Host, shost_classdev)
+	container_of(d, struct Scsi_Host, shost_dev)
 
 #define shost_printk(prefix, shost, fmt, a...)	\
 	dev_printk(prefix, &(shost)->shost_gendev, fmt, ##a)
diff --git a/include/scsi/scsi_transport.h b/include/scsi/scsi_transport.h
index 0dfef752f0e2..490bd13a634c 100644
--- a/include/scsi/scsi_transport.h
+++ b/include/scsi/scsi_transport.h
@@ -80,7 +80,7 @@ struct scsi_transport_template {
 };
 
 #define transport_class_to_shost(tc) \
-	dev_to_shost((tc)->dev)
+	dev_to_shost((tc)->parent)
 
 
 /* Private area maintenance. The driver requested allocations come
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index 4769efd4db24..06f72bab9df0 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -163,8 +163,8 @@ enum fc_tgtid_binding_type  {
 
 
 /* Macro for use in defining Virtual Port attributes */
-#define FC_VPORT_ATTR(_name,_mode,_show,_store)				\
-struct class_device_attribute class_device_attr_vport_##_name = 	\
+#define FC_VPORT_ATTR(_name,_mode,_show,_store)		\
+struct device_attribute dev_attr_vport_##_name = 	\
 	__ATTR(_name,_mode,_show,_store)
 
 
@@ -234,8 +234,8 @@ struct fc_vport {
 
 #define	dev_to_vport(d)				\
 	container_of(d, struct fc_vport, dev)
-#define transport_class_to_vport(classdev)	\
-	dev_to_vport(classdev->dev)
+#define transport_class_to_vport(dev)		\
+	dev_to_vport(dev->parent)
 #define vport_to_shost(v)			\
 	(v->shost)
 #define vport_to_shost_channel(v)		\
@@ -271,7 +271,7 @@ struct fc_rport_identifiers {
 
 /* Macro for use in defining Remote Port attributes */
 #define FC_RPORT_ATTR(_name,_mode,_show,_store)				\
-struct class_device_attribute class_device_attr_rport_##_name = 	\
+struct device_attribute dev_attr_rport_##_name = 	\
 	__ATTR(_name,_mode,_show,_store)
 
 
@@ -341,8 +341,8 @@ struct fc_rport {	/* aka fc_starget_attrs */
 
 #define	dev_to_rport(d)				\
 	container_of(d, struct fc_rport, dev)
-#define transport_class_to_rport(classdev)	\
-	dev_to_rport(classdev->dev)
+#define transport_class_to_rport(dev)	\
+	dev_to_rport(dev->parent)
 #define rport_to_shost(r)			\
 	dev_to_shost(r->dev.parent)
 
diff --git a/include/scsi/scsi_transport_sas.h b/include/scsi/scsi_transport_sas.h
index 09125fa95b93..61ad3594aad6 100644
--- a/include/scsi/scsi_transport_sas.h
+++ b/include/scsi/scsi_transport_sas.h
@@ -80,8 +80,8 @@ struct sas_phy {
 
 #define dev_to_phy(d) \
 	container_of((d), struct sas_phy, dev)
-#define transport_class_to_phy(cdev) \
-	dev_to_phy((cdev)->dev)
+#define transport_class_to_phy(dev) \
+	dev_to_phy((dev)->parent)
 #define phy_to_shost(phy) \
 	dev_to_shost((phy)->dev.parent)
 
@@ -96,8 +96,8 @@ struct sas_rphy {
 
 #define dev_to_rphy(d) \
 	container_of((d), struct sas_rphy, dev)
-#define transport_class_to_rphy(cdev) \
-	dev_to_rphy((cdev)->dev)
+#define transport_class_to_rphy(dev) \
+	dev_to_rphy((dev)->parent)
 #define rphy_to_shost(rphy) \
 	dev_to_shost((rphy)->dev.parent)
 #define target_to_rphy(targ) \
@@ -152,8 +152,8 @@ struct sas_port {
 
 #define dev_to_sas_port(d) \
 	container_of((d), struct sas_port, dev)
-#define transport_class_to_sas_port(cdev) \
-	dev_to_sas_port((cdev)->dev)
+#define transport_class_to_sas_port(dev) \
+	dev_to_sas_port((dev)->parent)
 
 struct sas_phy_linkrates {
 	enum sas_linkrate maximum_linkrate;
diff --git a/include/scsi/sd.h b/include/scsi/sd.h
index 8ea9f7358ac1..4f032d48cb6e 100644
--- a/include/scsi/sd.h
+++ b/include/scsi/sd.h
@@ -34,7 +34,7 @@
 struct scsi_disk {
 	struct scsi_driver *driver;	/* always &sd_template */
 	struct scsi_device *device;
-	struct class_device cdev;
+	struct device	dev;
 	struct gendisk	*disk;
 	unsigned int	openers;	/* protected by BKL for now, yuck */
 	sector_t	capacity;	/* size in 512-byte sectors */
@@ -46,7 +46,7 @@ struct scsi_disk {
 	unsigned	RCD : 1;	/* state of disk RCD bit, unused */
 	unsigned	DPOFUA : 1;	/* state of disk DPOFUA bit */
 };
-#define to_scsi_disk(obj) container_of(obj,struct scsi_disk,cdev)
+#define to_scsi_disk(obj) container_of(obj,struct scsi_disk,dev)
 
 #define sd_printk(prefix, sdsk, fmt, a...)				\
         (sdsk)->disk ?							\
-- 
cgit v1.2.3


From c3715cb90f722b1cf5f6f073be02cc8a49659b90 Mon Sep 17 00:00:00 2001
From: Sebastian Siewior <sebastian@breakpoint.cc>
Date: Sun, 30 Mar 2008 16:36:09 +0800
Subject: [CRYPTO] api: Make the crypto subsystem fully modular

Signed-off-by: Sebastian Siewior <sebastian@breakpoint.cc>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
 crypto/Kconfig         | 2 +-
 crypto/Makefile        | 3 ++-
 crypto/api.c           | 3 +++
 include/linux/crypto.h | 7 -------
 4 files changed, 6 insertions(+), 9 deletions(-)

(limited to 'include/linux')

diff --git a/crypto/Kconfig b/crypto/Kconfig
index e14ff1275018..edd00c5307a4 100644
--- a/crypto/Kconfig
+++ b/crypto/Kconfig
@@ -13,7 +13,7 @@ source "crypto/async_tx/Kconfig"
 # Cryptographic API Configuration
 #
 menuconfig CRYPTO
-	bool "Cryptographic API"
+	tristate "Cryptographic API"
 	help
 	  This option provides the core Cryptographic API.
 
diff --git a/crypto/Makefile b/crypto/Makefile
index 6d34bf7ecf8d..ca024418f4fb 100644
--- a/crypto/Makefile
+++ b/crypto/Makefile
@@ -2,7 +2,8 @@
 # Cryptographic API
 #
 
-obj-$(CONFIG_CRYPTO) += api.o cipher.o digest.o compress.o
+obj-$(CONFIG_CRYPTO) += crypto.o
+crypto-objs := api.o cipher.o digest.o compress.o
 
 crypto_algapi-$(CONFIG_PROC_FS) += proc.o
 crypto_algapi-objs := algapi.o scatterwalk.o $(crypto_algapi-y)
diff --git a/crypto/api.c b/crypto/api.c
index a2496d1bc6d4..0a0f41ef255f 100644
--- a/crypto/api.c
+++ b/crypto/api.c
@@ -445,3 +445,6 @@ int crypto_has_alg(const char *name, u32 type, u32 mask)
 	return ret;
 }
 EXPORT_SYMBOL_GPL(crypto_has_alg);
+
+MODULE_DESCRIPTION("Cryptographic core API");
+MODULE_LICENSE("GPL");
diff --git a/include/linux/crypto.h b/include/linux/crypto.h
index 5e02d1b46370..425824bd49f3 100644
--- a/include/linux/crypto.h
+++ b/include/linux/crypto.h
@@ -317,14 +317,7 @@ int crypto_unregister_alg(struct crypto_alg *alg);
 /*
  * Algorithm query interface.
  */
-#ifdef CONFIG_CRYPTO
 int crypto_has_alg(const char *name, u32 type, u32 mask);
-#else
-static inline int crypto_has_alg(const char *name, u32 type, u32 mask)
-{
-	return 0;
-}
-#endif
 
 /*
  * Transforms: user-instantiated objects which encapsulate algorithms
-- 
cgit v1.2.3


From 2baad5f96b498812626eadb6f6af3eb41d8656a3 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Wed, 13 Feb 2008 23:30:12 +0200
Subject: PCI: #if 0 pci_assign_resource_fixed()

An unused function that bloated the kernel only when CONFIG_EMBEDDED was
enabled...

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/setup-res.c | 2 +-
 include/linux/pci.h     | 1 -
 2 files changed, 1 insertion(+), 2 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/setup-res.c b/drivers/pci/setup-res.c
index 4be7ccf7e3ae..9e4d485ba9cd 100644
--- a/drivers/pci/setup-res.c
+++ b/drivers/pci/setup-res.c
@@ -171,7 +171,7 @@ int pci_assign_resource(struct pci_dev *dev, int resno)
 	return ret;
 }
 
-#ifdef CONFIG_EMBEDDED
+#if 0
 int pci_assign_resource_fixed(struct pci_dev *dev, int resno)
 {
 	struct pci_bus *bus = dev->bus;
diff --git a/include/linux/pci.h b/include/linux/pci.h
index ea760e519c46..3a2b9fbdb379 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -601,7 +601,6 @@ int pcie_get_readrq(struct pci_dev *dev);
 int pcie_set_readrq(struct pci_dev *dev, int rq);
 void pci_update_resource(struct pci_dev *dev, struct resource *res, int resno);
 int __must_check pci_assign_resource(struct pci_dev *dev, int i);
-int __must_check pci_assign_resource_fixed(struct pci_dev *dev, int i);
 int pci_select_bars(struct pci_dev *dev, unsigned long flags);
 
 /* ROM control related routines */
-- 
cgit v1.2.3


From 448432c4b8e2e3189177d6dbd16b8a8d83c5c11c Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Tue, 12 Feb 2008 13:36:20 -0800
Subject: PCI: remove pci_find_present

No one is using this function anymore for quite some time, so remove it.
Everyone calls pci_dev_present() instead anyway...

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/search.c | 35 +++++++++++++++--------------------
 include/linux/pci.h  |  2 --
 2 files changed, 15 insertions(+), 22 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/search.c b/drivers/pci/search.c
index 8541034021f0..1aabe3dbc7c3 100644
--- a/drivers/pci/search.c
+++ b/drivers/pci/search.c
@@ -436,7 +436,18 @@ exit:
 	return dev;
 }
 
-const struct pci_device_id *pci_find_present(const struct pci_device_id *ids)
+/**
+ * pci_dev_present - Returns 1 if device matching the device list is present, 0 if not.
+ * @ids: A pointer to a null terminated list of struct pci_device_id structures
+ * that describe the type of PCI device the caller is trying to find.
+ *
+ * Obvious fact: You do not have a reference to any device that might be found
+ * by this function, so if that device is removed from the system right after
+ * this function is finished, the value will be stale.  Use this function to
+ * find devices that are usually built into a system, or for a general hint as
+ * to if another device happens to be present at this specific moment in time.
+ */
+int pci_dev_present(const struct pci_device_id *ids)
 {
 	struct pci_dev *dev;
 	const struct pci_device_id *found = NULL;
@@ -452,27 +463,11 @@ const struct pci_device_id *pci_find_present(const struct pci_device_id *ids)
 	}
 exit:
 	up_read(&pci_bus_sem);
-	return found;
-}
-
-/**
- * pci_dev_present - Returns 1 if device matching the device list is present, 0 if not.
- * @ids: A pointer to a null terminated list of struct pci_device_id structures
- * that describe the type of PCI device the caller is trying to find.
- *
- * Obvious fact: You do not have a reference to any device that might be found
- * by this function, so if that device is removed from the system right after
- * this function is finished, the value will be stale.  Use this function to
- * find devices that are usually built into a system, or for a general hint as
- * to if another device happens to be present at this specific moment in time.
- */
-int pci_dev_present(const struct pci_device_id *ids)
-{
-	return pci_find_present(ids) == NULL ? 0 : 1;
+	if (found)
+		return 1;
+	return 0;
 }
-
 EXPORT_SYMBOL(pci_dev_present);
-EXPORT_SYMBOL(pci_find_present);
 
 #ifdef CONFIG_PCI_LEGACY
 EXPORT_SYMBOL(pci_find_device);
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 3a2b9fbdb379..b39f2abbea17 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -527,7 +527,6 @@ struct pci_dev *pci_get_slot(struct pci_bus *bus, unsigned int devfn);
 struct pci_dev *pci_get_bus_and_slot(unsigned int bus, unsigned int devfn);
 struct pci_dev *pci_get_class(unsigned int class, struct pci_dev *from);
 int pci_dev_present(const struct pci_device_id *ids);
-const struct pci_device_id *pci_find_present(const struct pci_device_id *ids);
 
 int pci_bus_read_config_byte(struct pci_bus *bus, unsigned int devfn,
 			     int where, u8 *val);
@@ -816,7 +815,6 @@ static inline struct pci_dev *pci_get_class(unsigned int class,
 
 #define pci_dev_present(ids)	(0)
 #define no_pci_devices()	(1)
-#define pci_find_present(ids)	(NULL)
 #define pci_dev_put(dev)	do { } while (0)
 
 static inline void pci_set_master(struct pci_dev *dev)
-- 
cgit v1.2.3


From 34220909a26b7f7cfc71e88ce01856c2563fe1d4 Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Wed, 13 Feb 2008 09:32:03 -0800
Subject: PCI: remove pci_get_device_reverse

This removes the pci_get_device_reverse function as there should not be
any need to walk pci devices backwards anymore.  All users of this call
are now gone from the tree, so it is safe to remove it.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/search.c | 41 -----------------------------------------
 include/linux/pci.h  | 10 ----------
 2 files changed, 51 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/search.c b/drivers/pci/search.c
index 1aabe3dbc7c3..a04c43ffce4c 100644
--- a/drivers/pci/search.c
+++ b/drivers/pci/search.c
@@ -359,46 +359,6 @@ pci_get_device(unsigned int vendor, unsigned int device, struct pci_dev *from)
 	return pci_get_subsys(vendor, device, PCI_ANY_ID, PCI_ANY_ID, from);
 }
 
-/**
- * pci_get_device_reverse - begin or continue searching for a PCI device by vendor/device id
- * @vendor: PCI vendor id to match, or %PCI_ANY_ID to match all vendor ids
- * @device: PCI device id to match, or %PCI_ANY_ID to match all device ids
- * @from: Previous PCI device found in search, or %NULL for new search.
- *
- * Iterates through the list of known PCI devices in the reverse order of
- * pci_get_device.
- * If a PCI device is found with a matching @vendor and @device, the reference
- * count to the  device is incremented and a pointer to its device structure
- * is returned Otherwise, %NULL is returned.  A new search is initiated by
- * passing %NULL as the @from argument.  Otherwise if @from is not %NULL,
- * searches continue from next device on the global list.  The reference
- * count for @from is always decremented if it is not %NULL.
- */
-struct pci_dev *
-pci_get_device_reverse(unsigned int vendor, unsigned int device, struct pci_dev *from)
-{
-	struct list_head *n;
-	struct pci_dev *dev;
-
-	WARN_ON(in_interrupt());
-	down_read(&pci_bus_sem);
-	n = from ? from->global_list.prev : pci_devices.prev;
-
-	while (n && (n != &pci_devices)) {
-		dev = pci_dev_g(n);
-		if ((vendor == PCI_ANY_ID || dev->vendor == vendor) &&
-		    (device == PCI_ANY_ID || dev->device == device))
-			goto exit;
-		n = n->prev;
-	}
-	dev = NULL;
-exit:
-	dev = pci_dev_get(dev);
-	up_read(&pci_bus_sem);
-	pci_dev_put(from);
-	return dev;
-}
-
 /**
  * pci_get_class - begin or continue searching for a PCI device by class
  * @class: search for a PCI device with this class designation
@@ -479,7 +439,6 @@ EXPORT_SYMBOL(pci_find_bus);
 EXPORT_SYMBOL(pci_find_next_bus);
 /* For everyone */
 EXPORT_SYMBOL(pci_get_device);
-EXPORT_SYMBOL(pci_get_device_reverse);
 EXPORT_SYMBOL(pci_get_subsys);
 EXPORT_SYMBOL(pci_get_slot);
 EXPORT_SYMBOL(pci_get_bus_and_slot);
diff --git a/include/linux/pci.h b/include/linux/pci.h
index b39f2abbea17..39ecf48ffa3b 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -517,9 +517,6 @@ struct pci_bus *pci_find_next_bus(const struct pci_bus *from);
 
 struct pci_dev *pci_get_device(unsigned int vendor, unsigned int device,
 				struct pci_dev *from);
-struct pci_dev *pci_get_device_reverse(unsigned int vendor, unsigned int device,
-				struct pci_dev *from);
-
 struct pci_dev *pci_get_subsys(unsigned int vendor, unsigned int device,
 				unsigned int ss_vendor, unsigned int ss_device,
 				struct pci_dev *from);
@@ -791,13 +788,6 @@ static inline struct pci_dev *pci_get_device(unsigned int vendor,
 	return NULL;
 }
 
-static inline struct pci_dev *pci_get_device_reverse(unsigned int vendor,
-						     unsigned int device,
-						     struct pci_dev *from)
-{
-	return NULL;
-}
-
 static inline struct pci_dev *pci_get_subsys(unsigned int vendor,
 					     unsigned int device,
 					     unsigned int ss_vendor,
-- 
cgit v1.2.3


From 95247b57ed844511a212265b45cf9a919753aea1 Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Wed, 13 Feb 2008 11:03:58 -0800
Subject: PCI: clean up search.c a lot

This cleans up the search.c file, now using the pci list of devices that
are created for the driver core, instead of relying on our separate list
of devices.  It's better to use the functions already created for this
kind of thing, instead of rolling our own all the time.

This work is done in anticipation of getting rid of that second list of
pci devices all together.

And it ends up saving code, always a nice benefit.

This also removes one compiler warning for when CONFIG_PCI_LEGACY is
enabled as we no longer internally use the deprecated functions anymore.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/search.c | 249 +++++++++++++++++++++++----------------------------
 include/linux/pci.h  |   4 +-
 2 files changed, 114 insertions(+), 139 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/search.c b/drivers/pci/search.c
index a04c43ffce4c..217814fef4ef 100644
--- a/drivers/pci/search.c
+++ b/drivers/pci/search.c
@@ -114,31 +114,63 @@ pci_find_next_bus(const struct pci_bus *from)
 }
 
 #ifdef CONFIG_PCI_LEGACY
-
 /**
  * pci_find_slot - locate PCI device from a given PCI slot
  * @bus: number of PCI bus on which desired PCI device resides
- * @devfn: encodes number of PCI slot in which the desired PCI 
- * device resides and the logical device number within that slot 
+ * @devfn: encodes number of PCI slot in which the desired PCI
+ * device resides and the logical device number within that slot
  * in case of multi-function devices.
  *
- * Given a PCI bus and slot/function number, the desired PCI device 
+ * Given a PCI bus and slot/function number, the desired PCI device
  * is located in system global list of PCI devices.  If the device
- * is found, a pointer to its data structure is returned.  If no 
+ * is found, a pointer to its data structure is returned.  If no
  * device is found, %NULL is returned.
+ *
+ * NOTE: Do not use this function any more; use pci_get_slot() instead, as
+ * the PCI device returned by this function can disappear at any moment in
+ * time.
  */
-struct pci_dev *
-pci_find_slot(unsigned int bus, unsigned int devfn)
+struct pci_dev *pci_find_slot(unsigned int bus, unsigned int devfn)
 {
 	struct pci_dev *dev = NULL;
 
-	while ((dev = pci_find_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
-		if (dev->bus->number == bus && dev->devfn == devfn)
+	while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
+		if (dev->bus->number == bus && dev->devfn == devfn) {
+			pci_dev_put(dev);
 			return dev;
+		}
 	}
 	return NULL;
 }
+EXPORT_SYMBOL(pci_find_slot);
 
+/**
+ * pci_find_device - begin or continue searching for a PCI device by vendor/device id
+ * @vendor: PCI vendor id to match, or %PCI_ANY_ID to match all vendor ids
+ * @device: PCI device id to match, or %PCI_ANY_ID to match all device ids
+ * @from: Previous PCI device found in search, or %NULL for new search.
+ *
+ * Iterates through the list of known PCI devices.  If a PCI device is found
+ * with a matching @vendor and @device, a pointer to its device structure is
+ * returned.  Otherwise, %NULL is returned.
+ * A new search is initiated by passing %NULL as the @from argument.
+ * Otherwise if @from is not %NULL, searches continue from next device
+ * on the global list.
+ *
+ * NOTE: Do not use this function any more; use pci_get_device() instead, as
+ * the PCI device returned by this function can disappear at any moment in
+ * time.
+ */
+struct pci_dev *pci_find_device(unsigned int vendor, unsigned int device,
+				const struct pci_dev *from)
+{
+	struct pci_dev *pdev;
+
+	pdev = pci_get_subsys(vendor, device, PCI_ANY_ID, PCI_ANY_ID, from);
+	pci_dev_put(pdev);
+	return pdev;
+}
+EXPORT_SYMBOL(pci_find_device);
 #endif /* CONFIG_PCI_LEGACY */
 
 /**
@@ -204,86 +236,52 @@ struct pci_dev * pci_get_bus_and_slot(unsigned int bus, unsigned int devfn)
 	return NULL;
 }
 
-#ifdef CONFIG_PCI_LEGACY
-/**
- * pci_find_subsys - begin or continue searching for a PCI device by vendor/subvendor/device/subdevice id
- * @vendor: PCI vendor id to match, or %PCI_ANY_ID to match all vendor ids
- * @device: PCI device id to match, or %PCI_ANY_ID to match all device ids
- * @ss_vendor: PCI subsystem vendor id to match, or %PCI_ANY_ID to match all vendor ids
- * @ss_device: PCI subsystem device id to match, or %PCI_ANY_ID to match all device ids
- * @from: Previous PCI device found in search, or %NULL for new search.
- *
- * Iterates through the list of known PCI devices.  If a PCI device is
- * found with a matching @vendor, @device, @ss_vendor and @ss_device, a
- * pointer to its device structure is returned.  Otherwise, %NULL is returned.
- * A new search is initiated by passing %NULL as the @from argument.
- * Otherwise if @from is not %NULL, searches continue from next device
- * on the global list.
- *
- * NOTE: Do not use this function any more; use pci_get_subsys() instead, as
- * the PCI device returned by this function can disappear at any moment in
- * time.
- */
-static struct pci_dev * pci_find_subsys(unsigned int vendor,
-				        unsigned int device,
-					unsigned int ss_vendor,
-					unsigned int ss_device,
-					const struct pci_dev *from)
+static int match_pci_dev_by_id(struct device *dev, void *data)
 {
-	struct list_head *n;
-	struct pci_dev *dev;
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct pci_device_id *id = data;
 
-	WARN_ON(in_interrupt());
-
-	/*
-	 * pci_find_subsys() can be called on the ide_setup() path, super-early
-	 * in boot.  But the down_read() will enable local interrupts, which
-	 * can cause some machines to crash.  So here we detect and flag that
-	 * situation and bail out early.
-	 */
-	if (unlikely(no_pci_devices()))
-		return NULL;
-	down_read(&pci_bus_sem);
-	n = from ? from->global_list.next : pci_devices.next;
-
-	while (n && (n != &pci_devices)) {
-		dev = pci_dev_g(n);
-		if ((vendor == PCI_ANY_ID || dev->vendor == vendor) &&
-		    (device == PCI_ANY_ID || dev->device == device) &&
-		    (ss_vendor == PCI_ANY_ID || dev->subsystem_vendor == ss_vendor) &&
-		    (ss_device == PCI_ANY_ID || dev->subsystem_device == ss_device))
-			goto exit;
-		n = n->next;
-	}
-	dev = NULL;
-exit:
-	up_read(&pci_bus_sem);
-	return dev;
+	if (pci_match_one_device(id, pdev))
+		return 1;
+	return 0;
 }
 
-/**
- * pci_find_device - begin or continue searching for a PCI device by vendor/device id
- * @vendor: PCI vendor id to match, or %PCI_ANY_ID to match all vendor ids
- * @device: PCI device id to match, or %PCI_ANY_ID to match all device ids
+/*
+ * pci_get_dev_by_id - begin or continue searching for a PCI device by id
+ * @id: pointer to struct pci_device_id to match for the device
  * @from: Previous PCI device found in search, or %NULL for new search.
  *
  * Iterates through the list of known PCI devices.  If a PCI device is found
- * with a matching @vendor and @device, a pointer to its device structure is
- * returned.  Otherwise, %NULL is returned.
- * A new search is initiated by passing %NULL as the @from argument.
- * Otherwise if @from is not %NULL, searches continue from next device
- * on the global list.
- * 
- * NOTE: Do not use this function any more; use pci_get_device() instead, as
- * the PCI device returned by this function can disappear at any moment in
- * time.
+ * with a matching id a pointer to its device structure is returned, and the
+ * reference count to the device is incremented.  Otherwise, %NULL is returned.
+ * A new search is initiated by passing %NULL as the @from argument.  Otherwise
+ * if @from is not %NULL, searches continue from next device on the global
+ * list.  The reference count for @from is always decremented if it is not
+ * %NULL.
+ *
+ * This is an internal function for use by the other search functions in
+ * this file.
  */
-struct pci_dev *
-pci_find_device(unsigned int vendor, unsigned int device, const struct pci_dev *from)
+static struct pci_dev *pci_get_dev_by_id(const struct pci_device_id *id,
+					 const struct pci_dev *from)
 {
-	return pci_find_subsys(vendor, device, PCI_ANY_ID, PCI_ANY_ID, from);
+	struct device *dev;
+	struct device *dev_start = NULL;
+	struct pci_dev *pdev = NULL;
+
+	WARN_ON(in_interrupt());
+	if (from) {
+		/* FIXME
+		 * take the cast off, when bus_find_device is made const.
+		 */
+		dev_start = (struct device *)&from->dev;
+	}
+	dev = bus_find_device(&pci_bus_type, dev_start, (void *)id,
+			      match_pci_dev_by_id);
+	if (dev)
+		pdev = to_pci_dev(dev);
+	return pdev;
 }
-#endif /* CONFIG_PCI_LEGACY */
 
 /**
  * pci_get_subsys - begin or continue searching for a PCI device by vendor/subvendor/device/subdevice id
@@ -301,42 +299,34 @@ pci_find_device(unsigned int vendor, unsigned int device, const struct pci_dev *
  * searches continue from next device on the global list.
  * The reference count for @from is always decremented if it is not %NULL.
  */
-struct pci_dev * 
-pci_get_subsys(unsigned int vendor, unsigned int device,
-	       unsigned int ss_vendor, unsigned int ss_device,
-	       struct pci_dev *from)
+struct pci_dev *pci_get_subsys(unsigned int vendor, unsigned int device,
+			       unsigned int ss_vendor, unsigned int ss_device,
+			       const struct pci_dev *from)
 {
-	struct list_head *n;
-	struct pci_dev *dev;
-
-	WARN_ON(in_interrupt());
+	struct pci_dev *pdev;
+	struct pci_device_id *id;
 
 	/*
-	 * pci_get_subsys() can potentially be called by drivers super-early
-	 * in boot.  But the down_read() will enable local interrupts, which
-	 * can cause some machines to crash.  So here we detect and flag that
-	 * situation and bail out early.
+	 * pci_find_subsys() can be called on the ide_setup() path,
+	 * super-early in boot.  But the down_read() will enable local
+	 * interrupts, which can cause some machines to crash.  So here we
+	 * detect and flag that situation and bail out early.
 	 */
 	if (unlikely(no_pci_devices()))
 		return NULL;
-	down_read(&pci_bus_sem);
-	n = from ? from->global_list.next : pci_devices.next;
-
-	while (n && (n != &pci_devices)) {
-		dev = pci_dev_g(n);
-		if ((vendor == PCI_ANY_ID || dev->vendor == vendor) &&
-		    (device == PCI_ANY_ID || dev->device == device) &&
-		    (ss_vendor == PCI_ANY_ID || dev->subsystem_vendor == ss_vendor) &&
-		    (ss_device == PCI_ANY_ID || dev->subsystem_device == ss_device))
-			goto exit;
-		n = n->next;
-	}
-	dev = NULL;
-exit:
-	dev = pci_dev_get(dev);
-	up_read(&pci_bus_sem);
-	pci_dev_put(from);
-	return dev;
+
+	id = kzalloc(sizeof(*id), GFP_KERNEL);
+	if (!id)
+		return NULL;
+	id->vendor = vendor;
+	id->device = device;
+	id->subvendor = ss_vendor;
+	id->subdevice = ss_device;
+
+	pdev = pci_get_dev_by_id(id, from);
+	kfree(id);
+
+	return pdev;
 }
 
 /**
@@ -375,24 +365,18 @@ pci_get_device(unsigned int vendor, unsigned int device, struct pci_dev *from)
  */
 struct pci_dev *pci_get_class(unsigned int class, struct pci_dev *from)
 {
-	struct list_head *n;
 	struct pci_dev *dev;
+	struct pci_device_id *id;
 
-	WARN_ON(in_interrupt());
-	down_read(&pci_bus_sem);
-	n = from ? from->global_list.next : pci_devices.next;
+	id = kzalloc(sizeof(*id), GFP_KERNEL);
+	if (!id)
+		return NULL;
+	id->vendor = id->device = id->subvendor = id->subdevice = PCI_ANY_ID;
+	id->class_mask = PCI_ANY_ID;
+	id->class = class;
 
-	while (n && (n != &pci_devices)) {
-		dev = pci_dev_g(n);
-		if (dev->class == class)
-			goto exit;
-		n = n->next;
-	}
-	dev = NULL;
-exit:
-	dev = pci_dev_get(dev);
-	up_read(&pci_bus_sem);
-	pci_dev_put(from);
+	dev = pci_get_dev_by_id(id, from);
+	kfree(id);
 	return dev;
 }
 
@@ -409,31 +393,22 @@ exit:
  */
 int pci_dev_present(const struct pci_device_id *ids)
 {
-	struct pci_dev *dev;
-	const struct pci_device_id *found = NULL;
+	struct pci_dev *found = NULL;
 
 	WARN_ON(in_interrupt());
-	down_read(&pci_bus_sem);
 	while (ids->vendor || ids->subvendor || ids->class_mask) {
-		list_for_each_entry(dev, &pci_devices, global_list) {
-			if ((found = pci_match_one_device(ids, dev)) != NULL)
-				goto exit;
-		}
+		found = pci_get_dev_by_id(ids, NULL);
+		if (found)
+			goto exit;
 		ids++;
 	}
 exit:
-	up_read(&pci_bus_sem);
 	if (found)
 		return 1;
 	return 0;
 }
 EXPORT_SYMBOL(pci_dev_present);
 
-#ifdef CONFIG_PCI_LEGACY
-EXPORT_SYMBOL(pci_find_device);
-EXPORT_SYMBOL(pci_find_slot);
-#endif /* CONFIG_PCI_LEGACY */
-
 /* For boot time work */
 EXPORT_SYMBOL(pci_find_bus);
 EXPORT_SYMBOL(pci_find_next_bus);
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 39ecf48ffa3b..5f79c72bae63 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -519,7 +519,7 @@ struct pci_dev *pci_get_device(unsigned int vendor, unsigned int device,
 				struct pci_dev *from);
 struct pci_dev *pci_get_subsys(unsigned int vendor, unsigned int device,
 				unsigned int ss_vendor, unsigned int ss_device,
-				struct pci_dev *from);
+				const struct pci_dev *from);
 struct pci_dev *pci_get_slot(struct pci_bus *bus, unsigned int devfn);
 struct pci_dev *pci_get_bus_and_slot(unsigned int bus, unsigned int devfn);
 struct pci_dev *pci_get_class(unsigned int class, struct pci_dev *from);
@@ -792,7 +792,7 @@ static inline struct pci_dev *pci_get_subsys(unsigned int vendor,
 					     unsigned int device,
 					     unsigned int ss_vendor,
 					     unsigned int ss_device,
-					     struct pci_dev *from)
+					     const struct pci_dev *from)
 {
 	return NULL;
 }
-- 
cgit v1.2.3


From 8a1bc9013a03d41a0e36ee413bb6f97281b30bd1 Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Thu, 14 Feb 2008 14:56:56 -0800
Subject: PCI: add is_added flag to struct pci_dev

This lets us check if the device is really added to the driver core or
not, which is what we need when walking some of the bus lists.  The flag
is there in anticipation of getting rid of the other PCI device list,
which is what we used to check in this situation.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 arch/powerpc/platforms/pseries/pci_dlpar.c |  7 ++-----
 drivers/pci/bus.c                          | 11 ++++-------
 drivers/pci/probe.c                        |  2 +-
 drivers/pci/remove.c                       |  6 ++----
 include/linux/pci.h                        |  1 +
 5 files changed, 10 insertions(+), 17 deletions(-)

(limited to 'include/linux')

diff --git a/arch/powerpc/platforms/pseries/pci_dlpar.c b/arch/powerpc/platforms/pseries/pci_dlpar.c
index 5a5a19e40bb4..d26a7bcad6b6 100644
--- a/arch/powerpc/platforms/pseries/pci_dlpar.c
+++ b/arch/powerpc/platforms/pseries/pci_dlpar.c
@@ -88,11 +88,8 @@ pcibios_fixup_new_pci_devices(struct pci_bus *bus)
 	struct pci_dev *dev;
 
 	list_for_each_entry(dev, &bus->devices, bus_list) {
-		/*
-		 * Skip already-present devices (which are on the
-		 * global device list.)
-		 */
-		if (list_empty(&dev->global_list)) {
+		/* Skip already-added devices */
+		if (!dev->is_added) {
 			int i;
 
 			/* Fill device archdata and setup iommu table */
diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c
index d708358326e5..e1c079aa0e82 100644
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -84,6 +84,7 @@ int pci_bus_add_device(struct pci_dev *dev)
 	if (retval)
 		return retval;
 
+	dev->is_added = 1;
 	down_write(&pci_bus_sem);
 	list_add_tail(&dev->global_list, &pci_devices);
 	up_write(&pci_bus_sem);
@@ -112,11 +113,8 @@ void pci_bus_add_devices(struct pci_bus *bus)
 	int retval;
 
 	list_for_each_entry(dev, &bus->devices, bus_list) {
-		/*
-		 * Skip already-present devices (which are on the
-		 * global device list.)
-		 */
-		if (!list_empty(&dev->global_list))
+		/* Skip already-added devices */
+		if (dev->is_added)
 			continue;
 		retval = pci_bus_add_device(dev);
 		if (retval)
@@ -124,8 +122,7 @@ void pci_bus_add_devices(struct pci_bus *bus)
 	}
 
 	list_for_each_entry(dev, &bus->devices, bus_list) {
-
-		BUG_ON(list_empty(&dev->global_list));
+		BUG_ON(!dev->is_added);
 
 		/*
 		 * If there is an unattached subordinate bus, attach
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 387fbbb97431..7217f4283ce8 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -984,7 +984,7 @@ EXPORT_SYMBOL(pci_scan_single_device);
  *
  * Scan a PCI slot on the specified PCI bus for devices, adding
  * discovered devices to the @bus->devices list.  New devices
- * will have an empty dev->global_list head.
+ * will not have is_added set.
  */
 int pci_scan_slot(struct pci_bus *bus, int devfn)
 {
diff --git a/drivers/pci/remove.c b/drivers/pci/remove.c
index 9684e1bde277..d3c77cbe3279 100644
--- a/drivers/pci/remove.c
+++ b/drivers/pci/remove.c
@@ -18,13 +18,11 @@ static void pci_free_resources(struct pci_dev *dev)
 
 static void pci_stop_dev(struct pci_dev *dev)
 {
-	if (!dev->global_list.next)
-		return;
-
-	if (!list_empty(&dev->global_list)) {
+	if (dev->is_added) {
 		pci_proc_detach_device(dev);
 		pci_remove_sysfs_dev_files(dev);
 		device_unregister(&dev->dev);
+		dev->is_added = 0;
 		down_write(&pci_bus_sem);
 		list_del(&dev->global_list);
 		dev->global_list.next = dev->global_list.prev = NULL;
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 5f79c72bae63..5e6d0f413fb9 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -181,6 +181,7 @@ struct pci_dev {
 	unsigned int	transparent:1;	/* Transparent PCI bridge */
 	unsigned int	multifunction:1;/* Part of multi-function device */
 	/* keep track of device state */
+	unsigned int	is_added:1;
 	unsigned int	is_busmaster:1; /* device is busmaster */
 	unsigned int	no_msi:1;	/* device may not use msi */
 	unsigned int	no_d1d2:1;   /* only allow d0 or d3 */
-- 
cgit v1.2.3


From 5ff580c10ec06fd296bd23d4570c1a95194094a0 Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Thu, 14 Feb 2008 14:56:56 -0800
Subject: PCI: remove global list of PCI devices

This patch finally removes the global list of PCI devices.  We are
relying entirely on the list held in the driver core now, and do not
need a separate "shadow" list as no one uses it.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/bus.c    |  4 ----
 drivers/pci/probe.c  | 39 +--------------------------------------
 drivers/pci/remove.c |  4 ----
 include/linux/pci.h  |  3 ---
 4 files changed, 1 insertion(+), 49 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c
index e1c079aa0e82..529d9d7727b0 100644
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -85,10 +85,6 @@ int pci_bus_add_device(struct pci_dev *dev)
 		return retval;
 
 	dev->is_added = 1;
-	down_write(&pci_bus_sem);
-	list_add_tail(&dev->global_list, &pci_devices);
-	up_write(&pci_bus_sem);
-
 	pci_proc_attach_device(dev);
 	pci_create_sysfs_dev_files(dev);
 	return 0;
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 7217f4283ce8..504f19b2af45 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -20,8 +20,6 @@
 LIST_HEAD(pci_root_buses);
 EXPORT_SYMBOL(pci_root_buses);
 
-LIST_HEAD(pci_devices);
-
 
 static int find_anything(struct device *dev, void *data)
 {
@@ -860,7 +858,6 @@ struct pci_dev *alloc_pci_dev(void)
 	if (!dev)
 		return NULL;
 
-	INIT_LIST_HEAD(&dev->global_list);
 	INIT_LIST_HEAD(&dev->bus_list);
 
 	pci_msi_init_pci_dev(dev);
@@ -957,7 +954,6 @@ void pci_device_add(struct pci_dev *dev, struct pci_bus *bus)
 	 * Add the device to our list of discovered devices
 	 * and the bus list for fixup functions, etc.
 	 */
-	INIT_LIST_HEAD(&dev->global_list);
 	down_write(&pci_bus_sem);
 	list_add_tail(&dev->bus_list, &bus->devices);
 	up_write(&pci_bus_sem);
@@ -1186,7 +1182,7 @@ static void __init pci_insertion_sort_klist(struct pci_dev *a, struct list_head
 	list_move_tail(&a->dev.knode_bus.n_node, list);
 }
 
-static void __init pci_sort_breadthfirst_klist(void)
+void __init pci_sort_breadthfirst(void)
 {
 	LIST_HEAD(sorted_devices);
 	struct list_head *pos, *tmp;
@@ -1207,36 +1203,3 @@ static void __init pci_sort_breadthfirst_klist(void)
 	list_splice(&sorted_devices, &device_klist->k_list);
 	spin_unlock(&device_klist->k_lock);
 }
-
-static void __init pci_insertion_sort_devices(struct pci_dev *a, struct list_head *list)
-{
-	struct pci_dev *b;
-
-	list_for_each_entry(b, list, global_list) {
-		if (pci_sort_bf_cmp(a, b) <= 0) {
-			list_move_tail(&a->global_list, &b->global_list);
-			return;
-		}
-	}
-	list_move_tail(&a->global_list, list);
-}
-
-static void __init pci_sort_breadthfirst_devices(void)
-{
-	LIST_HEAD(sorted_devices);
-	struct pci_dev *dev, *tmp;
-
-	down_write(&pci_bus_sem);
-	list_for_each_entry_safe(dev, tmp, &pci_devices, global_list) {
-		pci_insertion_sort_devices(dev, &sorted_devices);
-	}
-	list_splice(&sorted_devices, &pci_devices);
-	up_write(&pci_bus_sem);
-}
-
-void __init pci_sort_breadthfirst(void)
-{
-	pci_sort_breadthfirst_devices();
-	pci_sort_breadthfirst_klist();
-}
-
diff --git a/drivers/pci/remove.c b/drivers/pci/remove.c
index d3c77cbe3279..b6824833343f 100644
--- a/drivers/pci/remove.c
+++ b/drivers/pci/remove.c
@@ -23,10 +23,6 @@ static void pci_stop_dev(struct pci_dev *dev)
 		pci_remove_sysfs_dev_files(dev);
 		device_unregister(&dev->dev);
 		dev->is_added = 0;
-		down_write(&pci_bus_sem);
-		list_del(&dev->global_list);
-		dev->global_list.next = dev->global_list.prev = NULL;
-		up_write(&pci_bus_sem);
 	}
 }
 
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 5e6d0f413fb9..3b8a4e17052f 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -132,7 +132,6 @@ struct pci_cap_saved_state {
  * The pci_dev structure is used to describe PCI devices.
  */
 struct pci_dev {
-	struct list_head global_list;	/* node in list of all PCI devices */
 	struct list_head bus_list;	/* node in per-bus list */
 	struct pci_bus	*bus;		/* bus this device is on */
 	struct pci_bus	*subordinate;	/* bus this device bridges to */
@@ -206,7 +205,6 @@ struct pci_dev {
 
 extern struct pci_dev *alloc_pci_dev(void);
 
-#define pci_dev_g(n) list_entry(n, struct pci_dev, global_list)
 #define pci_dev_b(n) list_entry(n, struct pci_dev, bus_list)
 #define	to_pci_dev(n) container_of(n, struct pci_dev, dev)
 #define for_each_pci_dev(d) while ((d = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, d)) != NULL)
@@ -450,7 +448,6 @@ extern struct bus_type pci_bus_type;
 /* Do NOT directly access these two variables, unless you are arch specific pci
  * code, or pci core code. */
 extern struct list_head pci_root_buses;	/* list of all known PCI buses */
-extern struct list_head pci_devices;	/* list of all devices */
 /* Some device drivers need know if pci is initiated */
 extern int no_pci_devices(void);
 
-- 
cgit v1.2.3


From 21c6847406784fde73ad5ea47c2c3434714d58d1 Mon Sep 17 00:00:00 2001
From: Adrian Bunk <bunk@kernel.org>
Date: Mon, 4 Feb 2008 23:50:11 -0800
Subject: PCI: #if 0 pci_cleanup_aer_correct_error_status()

#if 0 the no longer used pci_cleanup_aer_correct_error_status().

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Cc: Stephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/pcie/aer/aerdrv_core.c | 3 ++-
 include/linux/aer.h                | 5 -----
 2 files changed, 2 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/pcie/aer/aerdrv_core.c b/drivers/pci/pcie/aer/aerdrv_core.c
index 3c0d8d138f5a..e84dfc8be0e9 100644
--- a/drivers/pci/pcie/aer/aerdrv_core.c
+++ b/drivers/pci/pcie/aer/aerdrv_core.c
@@ -117,6 +117,7 @@ int pci_cleanup_aer_uncorrect_error_status(struct pci_dev *dev)
 	return 0;
 }
 
+#if 0
 int pci_cleanup_aer_correct_error_status(struct pci_dev *dev)
 {
 	int pos;
@@ -131,6 +132,7 @@ int pci_cleanup_aer_correct_error_status(struct pci_dev *dev)
 
 	return 0;
 }
+#endif  /*  0  */
 
 static int find_device_iter(struct device *device, void *data)
 {
@@ -757,5 +759,4 @@ EXPORT_SYMBOL_GPL(pci_find_aer_capability);
 EXPORT_SYMBOL_GPL(pci_enable_pcie_error_reporting);
 EXPORT_SYMBOL_GPL(pci_disable_pcie_error_reporting);
 EXPORT_SYMBOL_GPL(pci_cleanup_aer_uncorrect_error_status);
-EXPORT_SYMBOL_GPL(pci_cleanup_aer_correct_error_status);
 
diff --git a/include/linux/aer.h b/include/linux/aer.h
index bcf236d825e8..f2518141de88 100644
--- a/include/linux/aer.h
+++ b/include/linux/aer.h
@@ -13,7 +13,6 @@ extern int pci_enable_pcie_error_reporting(struct pci_dev *dev);
 extern int pci_find_aer_capability(struct pci_dev *dev);
 extern int pci_disable_pcie_error_reporting(struct pci_dev *dev);
 extern int pci_cleanup_aer_uncorrect_error_status(struct pci_dev *dev);
-extern int pci_cleanup_aer_correct_error_status(struct pci_dev *dev);
 #else
 static inline int pci_enable_pcie_error_reporting(struct pci_dev *dev)
 {
@@ -31,10 +30,6 @@ static inline int pci_cleanup_aer_uncorrect_error_status(struct pci_dev *dev)
 {
 	return -EINVAL;
 }
-static inline int pci_cleanup_aer_correct_error_status(struct pci_dev *dev)
-{
-	return -EINVAL;
-}
 #endif
 
 #endif //_AER_H_
-- 
cgit v1.2.3


From 7d715a6c1ae5785d00fb9a876b5abdfc43abc44b Mon Sep 17 00:00:00 2001
From: Shaohua Li <shaohua.li@intel.com>
Date: Mon, 25 Feb 2008 09:46:41 +0800
Subject: PCI: add PCI Express ASPM support

PCI Express ASPM defines a protocol for PCI Express components in the D0
state to reduce Link power by placing their Links into a low power state
and instructing the other end of the Link to do likewise. This
capability allows hardware-autonomous, dynamic Link power reduction
beyond what is achievable by software-only controlled power management.
However, The device should be configured by software appropriately.
Enabling ASPM will save power, but will introduce device latency.

This patch adds ASPM support in Linux. It introduces a global policy for
ASPM, a sysfs file /sys/module/pcie_aspm/parameters/policy can control
it. The interface can be used as a boot option too. Currently we have
below setting:
        -default, BIOS default setting
        -powersave, highest power saving mode, enable all available ASPM
state and clock power management
        -performance, highest performance, disable ASPM and clock power
management
By default, the 'default' policy is used currently.

In my test, power difference between powersave mode and performance mode
is about 1.3w in a system with 3 PCIE links.

Note: some devices might not work well with aspm, either because chipset
issue or device issue. The patch provide API (pci_disable_link_state),
driver can disable ASPM for specific device.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/pci-sysfs.c   |   5 +
 drivers/pci/pci.c         |   4 +
 drivers/pci/pcie/Kconfig  |  20 ++
 drivers/pci/pcie/Makefile |   3 +
 drivers/pci/pcie/aspm.c   | 811 ++++++++++++++++++++++++++++++++++++++++++++++
 drivers/pci/probe.c       |   5 +
 drivers/pci/remove.c      |   4 +
 include/linux/pci-aspm.h  |  56 ++++
 include/linux/pci.h       |   5 +
 include/linux/pci_regs.h  |   8 +
 10 files changed, 921 insertions(+)
 create mode 100644 drivers/pci/pcie/aspm.c
 create mode 100644 include/linux/pci-aspm.h

(limited to 'include/linux')

diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c
index 8dcf1458aa2f..f5b0b622c189 100644
--- a/drivers/pci/pci-sysfs.c
+++ b/drivers/pci/pci-sysfs.c
@@ -21,6 +21,7 @@
 #include <linux/topology.h>
 #include <linux/mm.h>
 #include <linux/capability.h>
+#include <linux/pci-aspm.h>
 #include "pci.h"
 
 static int sysfs_initialized;	/* = 0 */
@@ -650,6 +651,8 @@ int __must_check pci_create_sysfs_dev_files (struct pci_dev *pdev)
 	if (pcibios_add_platform_entries(pdev))
 		goto err_rom_file;
 
+	pcie_aspm_create_sysfs_dev_files(pdev);
+
 	return 0;
 
 err_rom_file:
@@ -679,6 +682,8 @@ void pci_remove_sysfs_dev_files(struct pci_dev *pdev)
 	if (!sysfs_initialized)
 		return;
 
+	pcie_aspm_remove_sysfs_dev_files(pdev);
+
 	if (pdev->cfg_size < 4096)
 		sysfs_remove_bin_file(&pdev->dev.kobj, &pci_config_attr);
 	else
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index a4445b7210bf..f331feb4eb8d 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -18,6 +18,7 @@
 #include <linux/spinlock.h>
 #include <linux/string.h>
 #include <linux/log2.h>
+#include <linux/pci-aspm.h>
 #include <asm/dma.h>	/* isa_dma_bridge_buggy */
 #include "pci.h"
 
@@ -501,6 +502,9 @@ pci_set_power_state(struct pci_dev *dev, pci_power_t state)
 	if (need_restore)
 		pci_restore_bars(dev);
 
+	if (dev->bus->self)
+		pcie_aspm_pm_state_change(dev->bus->self);
+
 	return 0;
 }
 
diff --git a/drivers/pci/pcie/Kconfig b/drivers/pci/pcie/Kconfig
index 287a9311716c..25b04fb2517d 100644
--- a/drivers/pci/pcie/Kconfig
+++ b/drivers/pci/pcie/Kconfig
@@ -26,3 +26,23 @@ config HOTPLUG_PCI_PCIE
 	  When in doubt, say N.
 
 source "drivers/pci/pcie/aer/Kconfig"
+
+#
+# PCI Express ASPM
+#
+config PCIEASPM
+	bool "PCI Express ASPM support(Experimental)"
+	depends on PCI && EXPERIMENTAL && PCIEPORTBUS
+	default y
+	help
+	  This enables PCI Express ASPM (Active State Power Management) and
+	  Clock Power Management. ASPM supports state L0/L0s/L1.
+
+	  When in doubt, say N.
+config PCIEASPM_DEBUG
+	bool "Debug PCI Express ASPM"
+	depends on PCIEASPM
+	default n
+	help
+	  This enables PCI Express ASPM debug support. It will add per-device
+	  interface to control ASPM.
diff --git a/drivers/pci/pcie/Makefile b/drivers/pci/pcie/Makefile
index e00fb99acf44..11f6bb1eae24 100644
--- a/drivers/pci/pcie/Makefile
+++ b/drivers/pci/pcie/Makefile
@@ -2,6 +2,9 @@
 # Makefile for PCI-Express PORT Driver
 #
 
+# Build PCI Express ASPM if needed
+obj-$(CONFIG_PCIEASPM)		+= aspm.o
+
 pcieportdrv-y			:= portdrv_core.o portdrv_pci.o portdrv_bus.o
 
 obj-$(CONFIG_PCIEPORTBUS)	+= pcieportdrv.o
diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c
new file mode 100644
index 000000000000..61fedb2448b6
--- /dev/null
+++ b/drivers/pci/pcie/aspm.c
@@ -0,0 +1,811 @@
+/*
+ * File:	drivers/pci/pcie/aspm.c
+ * Enabling PCIE link L0s/L1 state and Clock Power Management
+ *
+ * Copyright (C) 2007 Intel
+ * Copyright (C) Zhang Yanmin (yanmin.zhang@intel.com)
+ * Copyright (C) Shaohua Li (shaohua.li@intel.com)
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/pci.h>
+#include <linux/pci_regs.h>
+#include <linux/errno.h>
+#include <linux/pm.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/pci-aspm.h>
+#include "../pci.h"
+
+#ifdef MODULE_PARAM_PREFIX
+#undef MODULE_PARAM_PREFIX
+#endif
+#define MODULE_PARAM_PREFIX "pcie_aspm."
+
+struct endpoint_state {
+	unsigned int l0s_acceptable_latency;
+	unsigned int l1_acceptable_latency;
+};
+
+struct pcie_link_state {
+	struct list_head sibiling;
+	struct pci_dev *pdev;
+
+	/* ASPM state */
+	unsigned int support_state;
+	unsigned int enabled_state;
+	unsigned int bios_aspm_state;
+	/* upstream component */
+	unsigned int l0s_upper_latency;
+	unsigned int l1_upper_latency;
+	/* downstream component */
+	unsigned int l0s_down_latency;
+	unsigned int l1_down_latency;
+	/* Clock PM state*/
+	unsigned int clk_pm_capable;
+	unsigned int clk_pm_enabled;
+	unsigned int bios_clk_state;
+
+	/*
+	 * A pcie downstream port only has one slot under it, so at most there
+	 * are 8 functions
+	 */
+	struct endpoint_state endpoints[8];
+};
+
+static int aspm_disabled;
+static DEFINE_MUTEX(aspm_lock);
+static LIST_HEAD(link_list);
+
+#define POLICY_DEFAULT 0	/* BIOS default setting */
+#define POLICY_PERFORMANCE 1	/* high performance */
+#define POLICY_POWERSAVE 2	/* high power saving */
+static int aspm_policy;
+static const char *policy_str[] = {
+	[POLICY_DEFAULT] = "default",
+	[POLICY_PERFORMANCE] = "performance",
+	[POLICY_POWERSAVE] = "powersave"
+};
+
+static int policy_to_aspm_state(struct pci_dev *pdev)
+{
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	switch (aspm_policy) {
+	case POLICY_PERFORMANCE:
+		/* Disable ASPM and Clock PM */
+		return 0;
+	case POLICY_POWERSAVE:
+		/* Enable ASPM L0s/L1 */
+		return PCIE_LINK_STATE_L0S|PCIE_LINK_STATE_L1;
+	case POLICY_DEFAULT:
+		return link_state->bios_aspm_state;
+	}
+	return 0;
+}
+
+static int policy_to_clkpm_state(struct pci_dev *pdev)
+{
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	switch (aspm_policy) {
+	case POLICY_PERFORMANCE:
+		/* Disable ASPM and Clock PM */
+		return 0;
+	case POLICY_POWERSAVE:
+		/* Disable Clock PM */
+		return 1;
+	case POLICY_DEFAULT:
+		return link_state->bios_clk_state;
+	}
+	return 0;
+}
+
+static void pcie_set_clock_pm(struct pci_dev *pdev, int enable)
+{
+	struct pci_dev *child_dev;
+	int pos;
+	u16 reg16;
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	list_for_each_entry(child_dev, &pdev->subordinate->devices, bus_list) {
+		pos = pci_find_capability(child_dev, PCI_CAP_ID_EXP);
+		if (!pos)
+			return;
+		pci_read_config_word(child_dev, pos + PCI_EXP_LNKCTL, &reg16);
+		if (enable)
+			reg16 |= PCI_EXP_LNKCTL_CLKREQ_EN;
+		else
+			reg16 &= ~PCI_EXP_LNKCTL_CLKREQ_EN;
+		pci_write_config_word(child_dev, pos + PCI_EXP_LNKCTL, reg16);
+	}
+	link_state->clk_pm_enabled = !!enable;
+}
+
+static void pcie_check_clock_pm(struct pci_dev *pdev)
+{
+	int pos;
+	u32 reg32;
+	u16 reg16;
+	int capable = 1, enabled = 1;
+	struct pci_dev *child_dev;
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	/* All functions should have the same cap and state, take the worst */
+	list_for_each_entry(child_dev, &pdev->subordinate->devices, bus_list) {
+		pos = pci_find_capability(child_dev, PCI_CAP_ID_EXP);
+		if (!pos)
+			return;
+		pci_read_config_dword(child_dev, pos + PCI_EXP_LNKCAP, &reg32);
+		if (!(reg32 & PCI_EXP_LNKCAP_CLKPM)) {
+			capable = 0;
+			enabled = 0;
+			break;
+		}
+		pci_read_config_word(child_dev, pos + PCI_EXP_LNKCTL, &reg16);
+		if (!(reg16 & PCI_EXP_LNKCTL_CLKREQ_EN))
+			enabled = 0;
+	}
+	link_state->clk_pm_capable = capable;
+	link_state->clk_pm_enabled = enabled;
+	link_state->bios_clk_state = enabled;
+	pcie_set_clock_pm(pdev, policy_to_clkpm_state(pdev));
+}
+
+/*
+ * pcie_aspm_configure_common_clock: check if the 2 ends of a link
+ *   could use common clock. If they are, configure them to use the
+ *   common clock. That will reduce the ASPM state exit latency.
+ */
+static void pcie_aspm_configure_common_clock(struct pci_dev *pdev)
+{
+	int pos, child_pos;
+	u16 reg16 = 0;
+	struct pci_dev *child_dev;
+	int same_clock = 1;
+
+	/*
+	 * all functions of a slot should have the same Slot Clock
+	 * Configuration, so just check one function
+	 * */
+	child_dev = list_entry(pdev->subordinate->devices.next, struct pci_dev,
+		bus_list);
+	BUG_ON(!child_dev->is_pcie);
+
+	/* Check downstream component if bit Slot Clock Configuration is 1 */
+	child_pos = pci_find_capability(child_dev, PCI_CAP_ID_EXP);
+	pci_read_config_word(child_dev, child_pos + PCI_EXP_LNKSTA, &reg16);
+	if (!(reg16 & PCI_EXP_LNKSTA_SLC))
+		same_clock = 0;
+
+	/* Check upstream component if bit Slot Clock Configuration is 1 */
+	pos = pci_find_capability(pdev, PCI_CAP_ID_EXP);
+	pci_read_config_word(pdev, pos + PCI_EXP_LNKSTA, &reg16);
+	if (!(reg16 & PCI_EXP_LNKSTA_SLC))
+		same_clock = 0;
+
+	/* Configure downstream component, all functions */
+	list_for_each_entry(child_dev, &pdev->subordinate->devices, bus_list) {
+		child_pos = pci_find_capability(child_dev, PCI_CAP_ID_EXP);
+		pci_read_config_word(child_dev, child_pos + PCI_EXP_LNKCTL,
+			&reg16);
+		if (same_clock)
+			reg16 |= PCI_EXP_LNKCTL_CCC;
+		else
+			reg16 &= ~PCI_EXP_LNKCTL_CCC;
+		pci_write_config_word(child_dev, child_pos + PCI_EXP_LNKCTL,
+			reg16);
+	}
+
+	/* Configure upstream component */
+	pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
+	if (same_clock)
+		reg16 |= PCI_EXP_LNKCTL_CCC;
+	else
+		reg16 &= ~PCI_EXP_LNKCTL_CCC;
+	pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);
+
+	/* retrain link */
+	reg16 |= PCI_EXP_LNKCTL_RL;
+	pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);
+
+	/* Wait for link training end */
+	while (1) {
+		pci_read_config_word(pdev, pos + PCI_EXP_LNKSTA, &reg16);
+		if (!(reg16 & PCI_EXP_LNKSTA_LT))
+			break;
+		cpu_relax();
+	}
+}
+
+/*
+ * calc_L0S_latency: Convert L0s latency encoding to ns
+ */
+static unsigned int calc_L0S_latency(unsigned int latency_encoding, int ac)
+{
+	unsigned int ns = 64;
+
+	if (latency_encoding == 0x7) {
+		if (ac)
+			ns = -1U;
+		else
+			ns = 5*1000; /* > 4us */
+	} else
+		ns *= (1 << latency_encoding);
+	return ns;
+}
+
+/*
+ * calc_L1_latency: Convert L1 latency encoding to ns
+ */
+static unsigned int calc_L1_latency(unsigned int latency_encoding, int ac)
+{
+	unsigned int ns = 1000;
+
+	if (latency_encoding == 0x7) {
+		if (ac)
+			ns = -1U;
+		else
+			ns = 65*1000; /* > 64us */
+	} else
+		ns *= (1 << latency_encoding);
+	return ns;
+}
+
+static void pcie_aspm_get_cap_device(struct pci_dev *pdev, u32 *state,
+	unsigned int *l0s, unsigned int *l1, unsigned int *enabled)
+{
+	int pos;
+	u16 reg16;
+	u32 reg32;
+	unsigned int latency;
+
+	pos = pci_find_capability(pdev, PCI_CAP_ID_EXP);
+	pci_read_config_dword(pdev, pos + PCI_EXP_LNKCAP, &reg32);
+	*state = (reg32 & PCI_EXP_LNKCAP_ASPMS) >> 10;
+	if (*state != PCIE_LINK_STATE_L0S &&
+		*state != (PCIE_LINK_STATE_L1|PCIE_LINK_STATE_L0S))
+		*state = 0;
+	if (*state == 0)
+		return;
+
+	latency = (reg32 & PCI_EXP_LNKCAP_L0SEL) >> 12;
+	*l0s = calc_L0S_latency(latency, 0);
+	if (*state & PCIE_LINK_STATE_L1) {
+		latency = (reg32 & PCI_EXP_LNKCAP_L1EL) >> 15;
+		*l1 = calc_L1_latency(latency, 0);
+	}
+	pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
+	*enabled = reg16 & (PCIE_LINK_STATE_L0S|PCIE_LINK_STATE_L1);
+}
+
+static void pcie_aspm_cap_init(struct pci_dev *pdev)
+{
+	struct pci_dev *child_dev;
+	u32 state, tmp;
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	/* upstream component states */
+	pcie_aspm_get_cap_device(pdev, &link_state->support_state,
+		&link_state->l0s_upper_latency,
+		&link_state->l1_upper_latency,
+		&link_state->enabled_state);
+	/* downstream component states, all functions have the same setting */
+	child_dev = list_entry(pdev->subordinate->devices.next, struct pci_dev,
+		bus_list);
+	pcie_aspm_get_cap_device(child_dev, &state,
+		&link_state->l0s_down_latency,
+		&link_state->l1_down_latency,
+		&tmp);
+	link_state->support_state &= state;
+	if (!link_state->support_state)
+		return;
+	link_state->enabled_state &= link_state->support_state;
+	link_state->bios_aspm_state = link_state->enabled_state;
+
+	/* ENDPOINT states*/
+	list_for_each_entry(child_dev, &pdev->subordinate->devices, bus_list) {
+		int pos;
+		u32 reg32;
+		unsigned int latency;
+		struct endpoint_state *ep_state =
+			&link_state->endpoints[PCI_FUNC(child_dev->devfn)];
+
+		if (child_dev->pcie_type != PCI_EXP_TYPE_ENDPOINT &&
+			child_dev->pcie_type != PCI_EXP_TYPE_LEG_END)
+			continue;
+
+		pos = pci_find_capability(child_dev, PCI_CAP_ID_EXP);
+		pci_read_config_dword(child_dev, pos + PCI_EXP_DEVCAP, &reg32);
+		latency = (reg32 & PCI_EXP_DEVCAP_L0S) >> 6;
+		latency = calc_L0S_latency(latency, 1);
+		ep_state->l0s_acceptable_latency = latency;
+		if (link_state->support_state & PCIE_LINK_STATE_L1) {
+			latency = (reg32 & PCI_EXP_DEVCAP_L1) >> 9;
+			latency = calc_L1_latency(latency, 1);
+			ep_state->l1_acceptable_latency = latency;
+		}
+	}
+}
+
+static unsigned int __pcie_aspm_check_state_one(struct pci_dev *pdev,
+	unsigned int state)
+{
+	struct pci_dev *parent_dev, *tmp_dev;
+	unsigned int latency, l1_latency = 0;
+	struct pcie_link_state *link_state;
+	struct endpoint_state *ep_state;
+
+	parent_dev = pdev->bus->self;
+	link_state = parent_dev->link_state;
+	state &= link_state->support_state;
+	if (state == 0)
+		return 0;
+	ep_state = &link_state->endpoints[PCI_FUNC(pdev->devfn)];
+
+	/*
+	 * Check latency for endpoint device.
+	 * TBD: The latency from the endpoint to root complex vary per
+	 * switch's upstream link state above the device. Here we just do a
+	 * simple check which assumes all links above the device can be in L1
+	 * state, that is we just consider the worst case. If switch's upstream
+	 * link can't be put into L0S/L1, then our check is too strictly.
+	 */
+	tmp_dev = pdev;
+	while (state & (PCIE_LINK_STATE_L0S | PCIE_LINK_STATE_L1)) {
+		parent_dev = tmp_dev->bus->self;
+		link_state = parent_dev->link_state;
+		if (state & PCIE_LINK_STATE_L0S) {
+			latency = max_t(unsigned int,
+					link_state->l0s_upper_latency,
+					link_state->l0s_down_latency);
+			if (latency > ep_state->l0s_acceptable_latency)
+				state &= ~PCIE_LINK_STATE_L0S;
+		}
+		if (state & PCIE_LINK_STATE_L1) {
+			latency = max_t(unsigned int,
+					link_state->l1_upper_latency,
+					link_state->l1_down_latency);
+			if (latency + l1_latency >
+					ep_state->l1_acceptable_latency)
+				state &= ~PCIE_LINK_STATE_L1;
+		}
+		if (!parent_dev->bus->self) /* parent_dev is a root port */
+			break;
+		else {
+			/*
+			 * parent_dev is the downstream port of a switch, make
+			 * tmp_dev the upstream port of the switch
+			 */
+			tmp_dev = parent_dev->bus->self;
+			/*
+			 * every switch on the path to root complex need 1 more
+			 * microsecond for L1. Spec doesn't mention L0S.
+			 */
+			if (state & PCIE_LINK_STATE_L1)
+				l1_latency += 1000;
+		}
+	}
+	return state;
+}
+
+static unsigned int pcie_aspm_check_state(struct pci_dev *pdev,
+	unsigned int state)
+{
+	struct pci_dev *child_dev;
+
+	/* If no child, disable the link */
+	if (list_empty(&pdev->subordinate->devices))
+		return 0;
+	list_for_each_entry(child_dev, &pdev->subordinate->devices, bus_list) {
+		if (child_dev->pcie_type == PCI_EXP_TYPE_PCI_BRIDGE) {
+			/*
+			 * If downstream component of a link is pci bridge, we
+			 * disable ASPM for now for the link
+			 * */
+			state = 0;
+			break;
+		}
+		if ((child_dev->pcie_type != PCI_EXP_TYPE_ENDPOINT &&
+			child_dev->pcie_type != PCI_EXP_TYPE_LEG_END))
+			continue;
+		/* Device not in D0 doesn't need check latency */
+		if (child_dev->current_state == PCI_D1 ||
+			child_dev->current_state == PCI_D2 ||
+			child_dev->current_state == PCI_D3hot ||
+			child_dev->current_state == PCI_D3cold)
+			continue;
+		state = __pcie_aspm_check_state_one(child_dev, state);
+	}
+	return state;
+}
+
+static void __pcie_aspm_config_one_dev(struct pci_dev *pdev, unsigned int state)
+{
+	u16 reg16;
+	int pos = pci_find_capability(pdev, PCI_CAP_ID_EXP);
+
+	pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
+	reg16 &= ~0x3;
+	reg16 |= state;
+	pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);
+}
+
+static void __pcie_aspm_config_link(struct pci_dev *pdev, unsigned int state)
+{
+	struct pci_dev *child_dev;
+	int valid = 1;
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	/*
+	 * if the downstream component has pci bridge function, don't do ASPM
+	 * now
+	 */
+	list_for_each_entry(child_dev, &pdev->subordinate->devices, bus_list) {
+		if (child_dev->pcie_type == PCI_EXP_TYPE_PCI_BRIDGE) {
+			valid = 0;
+			break;
+		}
+	}
+	if (!valid)
+		return;
+
+	/*
+	 * spec 2.0 suggests all functions should be configured the same
+	 * setting for ASPM. Enabling ASPM L1 should be done in upstream
+	 * component first and then downstream, and vice versa for disabling
+	 * ASPM L1. Spec doesn't mention L0S.
+	 */
+	if (state & PCIE_LINK_STATE_L1)
+		__pcie_aspm_config_one_dev(pdev, state);
+
+	list_for_each_entry(child_dev, &pdev->subordinate->devices, bus_list)
+		__pcie_aspm_config_one_dev(child_dev, state);
+
+	if (!(state & PCIE_LINK_STATE_L1))
+		__pcie_aspm_config_one_dev(pdev, state);
+
+	link_state->enabled_state = state;
+}
+
+static void __pcie_aspm_configure_link_state(struct pci_dev *pdev,
+	unsigned int state)
+{
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	if (link_state->support_state == 0)
+		return;
+	state &= PCIE_LINK_STATE_L0S|PCIE_LINK_STATE_L1;
+
+	/* state 0 means disabling aspm */
+	state = pcie_aspm_check_state(pdev, state);
+	if (link_state->enabled_state == state)
+		return;
+	__pcie_aspm_config_link(pdev, state);
+}
+
+/*
+ * pcie_aspm_configure_link_state: enable/disable PCI express link state
+ * @pdev: the root port or switch downstream port
+ */
+static void pcie_aspm_configure_link_state(struct pci_dev *pdev,
+	unsigned int state)
+{
+	down_read(&pci_bus_sem);
+	mutex_lock(&aspm_lock);
+	__pcie_aspm_configure_link_state(pdev, state);
+	mutex_unlock(&aspm_lock);
+	up_read(&pci_bus_sem);
+}
+
+static void free_link_state(struct pci_dev *pdev)
+{
+	kfree(pdev->link_state);
+	pdev->link_state = NULL;
+}
+
+/*
+ * pcie_aspm_init_link_state: Initiate PCI express link state.
+ * It is called after the pcie and its children devices are scaned.
+ * @pdev: the root port or switch downstream port
+ */
+void pcie_aspm_init_link_state(struct pci_dev *pdev)
+{
+	unsigned int state;
+	struct pcie_link_state *link_state;
+	int error = 0;
+
+	if (aspm_disabled || !pdev->is_pcie || pdev->link_state)
+		return;
+	if (pdev->pcie_type != PCI_EXP_TYPE_ROOT_PORT &&
+		pdev->pcie_type != PCI_EXP_TYPE_DOWNSTREAM)
+		return;
+	down_read(&pci_bus_sem);
+	if (list_empty(&pdev->subordinate->devices))
+		goto out;
+
+	mutex_lock(&aspm_lock);
+
+	link_state = kzalloc(sizeof(*link_state), GFP_KERNEL);
+	if (!link_state)
+		goto unlock_out;
+	pdev->link_state = link_state;
+
+	pcie_aspm_configure_common_clock(pdev);
+
+	pcie_aspm_cap_init(pdev);
+
+	/* config link state to avoid BIOS error */
+	state = pcie_aspm_check_state(pdev, policy_to_aspm_state(pdev));
+	__pcie_aspm_config_link(pdev, state);
+
+	pcie_check_clock_pm(pdev);
+
+	link_state->pdev = pdev;
+	list_add(&link_state->sibiling, &link_list);
+
+unlock_out:
+	if (error)
+		free_link_state(pdev);
+	mutex_unlock(&aspm_lock);
+out:
+	up_read(&pci_bus_sem);
+}
+
+/* @pdev: the endpoint device */
+void pcie_aspm_exit_link_state(struct pci_dev *pdev)
+{
+	struct pci_dev *parent = pdev->bus->self;
+	struct pcie_link_state *link_state = parent->link_state;
+
+	if (aspm_disabled || !pdev->is_pcie || !parent || !link_state)
+		return;
+	if (parent->pcie_type != PCI_EXP_TYPE_ROOT_PORT &&
+		parent->pcie_type != PCI_EXP_TYPE_DOWNSTREAM)
+		return;
+	down_read(&pci_bus_sem);
+	mutex_lock(&aspm_lock);
+
+	/*
+	 * All PCIe functions are in one slot, remove one function will remove
+	 * the the whole slot, so just wait
+	 */
+	if (!list_empty(&parent->subordinate->devices))
+		goto out;
+
+	/* All functions are removed, so just disable ASPM for the link */
+	__pcie_aspm_config_one_dev(parent, 0);
+	list_del(&link_state->sibiling);
+	/* Clock PM is for endpoint device */
+
+	free_link_state(parent);
+out:
+	mutex_unlock(&aspm_lock);
+	up_read(&pci_bus_sem);
+}
+
+/* @pdev: the root port or switch downstream port */
+void pcie_aspm_pm_state_change(struct pci_dev *pdev)
+{
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	if (aspm_disabled || !pdev->is_pcie || !pdev->link_state)
+		return;
+	if (pdev->pcie_type != PCI_EXP_TYPE_ROOT_PORT &&
+		pdev->pcie_type != PCI_EXP_TYPE_DOWNSTREAM)
+		return;
+	/*
+	 * devices changed PM state, we should recheck if latency meets all
+	 * functions' requirement
+	 */
+	pcie_aspm_configure_link_state(pdev, link_state->enabled_state);
+}
+
+/*
+ * pci_disable_link_state - disable pci device's link state, so the link will
+ * never enter specific states
+ */
+void pci_disable_link_state(struct pci_dev *pdev, int state)
+{
+	struct pci_dev *parent = pdev->bus->self;
+	struct pcie_link_state *link_state;
+
+	if (aspm_disabled || !pdev->is_pcie)
+		return;
+	if (pdev->pcie_type == PCI_EXP_TYPE_ROOT_PORT ||
+	    pdev->pcie_type == PCI_EXP_TYPE_DOWNSTREAM)
+		parent = pdev;
+	if (!parent || !parent->link_state)
+		return;
+
+	down_read(&pci_bus_sem);
+	mutex_lock(&aspm_lock);
+	link_state = parent->link_state;
+	link_state->support_state &=
+		~(state & (PCIE_LINK_STATE_L0S|PCIE_LINK_STATE_L1));
+	if (state & PCIE_LINK_STATE_CLKPM)
+		link_state->clk_pm_capable = 0;
+
+	__pcie_aspm_configure_link_state(parent, link_state->enabled_state);
+	if (!link_state->clk_pm_capable && link_state->clk_pm_enabled)
+		pcie_set_clock_pm(parent, 0);
+	mutex_unlock(&aspm_lock);
+	up_read(&pci_bus_sem);
+}
+EXPORT_SYMBOL(pci_disable_link_state);
+
+static int pcie_aspm_set_policy(const char *val, struct kernel_param *kp)
+{
+	int i;
+	struct pci_dev *pdev;
+	struct pcie_link_state *link_state;
+
+	for (i = 0; i < ARRAY_SIZE(policy_str); i++)
+		if (!strncmp(val, policy_str[i], strlen(policy_str[i])))
+			break;
+	if (i >= ARRAY_SIZE(policy_str))
+		return -EINVAL;
+	if (i == aspm_policy)
+		return 0;
+
+	down_read(&pci_bus_sem);
+	mutex_lock(&aspm_lock);
+	aspm_policy = i;
+	list_for_each_entry(link_state, &link_list, sibiling) {
+		pdev = link_state->pdev;
+		__pcie_aspm_configure_link_state(pdev,
+			policy_to_aspm_state(pdev));
+		if (link_state->clk_pm_capable &&
+		    link_state->clk_pm_enabled != policy_to_clkpm_state(pdev))
+			pcie_set_clock_pm(pdev, policy_to_clkpm_state(pdev));
+
+	}
+	mutex_unlock(&aspm_lock);
+	up_read(&pci_bus_sem);
+	return 0;
+}
+
+static int pcie_aspm_get_policy(char *buffer, struct kernel_param *kp)
+{
+	int i, cnt = 0;
+	for (i = 0; i < ARRAY_SIZE(policy_str); i++)
+		if (i == aspm_policy)
+			cnt += sprintf(buffer + cnt, "[%s] ", policy_str[i]);
+		else
+			cnt += sprintf(buffer + cnt, "%s ", policy_str[i]);
+	return cnt;
+}
+
+module_param_call(policy, pcie_aspm_set_policy, pcie_aspm_get_policy,
+	NULL, 0644);
+
+#ifdef CONFIG_PCIEASPM_DEBUG
+static ssize_t link_state_show(struct device *dev,
+		struct device_attribute *attr,
+		char *buf)
+{
+	struct pci_dev *pci_device = to_pci_dev(dev);
+	struct pcie_link_state *link_state = pci_device->link_state;
+
+	return sprintf(buf, "%d\n", link_state->enabled_state);
+}
+
+static ssize_t link_state_store(struct device *dev,
+		struct device_attribute *attr,
+		const char *buf,
+		size_t n)
+{
+	struct pci_dev *pci_device = to_pci_dev(dev);
+	int state;
+
+	if (n < 1)
+		return -EINVAL;
+	state = buf[0]-'0';
+	if (state >= 0 && state <= 3) {
+		/* setup link aspm state */
+		pcie_aspm_configure_link_state(pci_device, state);
+		return n;
+	}
+
+	return -EINVAL;
+}
+
+static ssize_t clk_ctl_show(struct device *dev,
+		struct device_attribute *attr,
+		char *buf)
+{
+	struct pci_dev *pci_device = to_pci_dev(dev);
+	struct pcie_link_state *link_state = pci_device->link_state;
+
+	return sprintf(buf, "%d\n", link_state->clk_pm_enabled);
+}
+
+static ssize_t clk_ctl_store(struct device *dev,
+		struct device_attribute *attr,
+		const char *buf,
+		size_t n)
+{
+	struct pci_dev *pci_device = to_pci_dev(dev);
+	int state;
+
+	if (n < 1)
+		return -EINVAL;
+	state = buf[0]-'0';
+
+	down_read(&pci_bus_sem);
+	mutex_lock(&aspm_lock);
+	pcie_set_clock_pm(pci_device, !!state);
+	mutex_unlock(&aspm_lock);
+	up_read(&pci_bus_sem);
+
+	return n;
+}
+
+static DEVICE_ATTR(link_state, 0644, link_state_show, link_state_store);
+static DEVICE_ATTR(clk_ctl, 0644, clk_ctl_show, clk_ctl_store);
+
+static char power_group[] = "power";
+void pcie_aspm_create_sysfs_dev_files(struct pci_dev *pdev)
+{
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	if (!pdev->is_pcie || (pdev->pcie_type != PCI_EXP_TYPE_ROOT_PORT &&
+		pdev->pcie_type != PCI_EXP_TYPE_DOWNSTREAM) || !link_state)
+		return;
+
+	if (link_state->support_state)
+		sysfs_add_file_to_group(&pdev->dev.kobj,
+			&dev_attr_link_state.attr, power_group);
+	if (link_state->clk_pm_capable)
+		sysfs_add_file_to_group(&pdev->dev.kobj,
+			&dev_attr_clk_ctl.attr, power_group);
+}
+
+void pcie_aspm_remove_sysfs_dev_files(struct pci_dev *pdev)
+{
+	struct pcie_link_state *link_state = pdev->link_state;
+
+	if (!pdev->is_pcie || (pdev->pcie_type != PCI_EXP_TYPE_ROOT_PORT &&
+		pdev->pcie_type != PCI_EXP_TYPE_DOWNSTREAM) || !link_state)
+		return;
+
+	if (link_state->support_state)
+		sysfs_remove_file_from_group(&pdev->dev.kobj,
+			&dev_attr_link_state.attr, power_group);
+	if (link_state->clk_pm_capable)
+		sysfs_remove_file_from_group(&pdev->dev.kobj,
+			&dev_attr_clk_ctl.attr, power_group);
+}
+#endif
+
+static int __init pcie_aspm_disable(char *str)
+{
+	aspm_disabled = 1;
+	return 1;
+}
+
+__setup("pcie_noaspm", pcie_aspm_disable);
+
+#ifdef CONFIG_ACPI
+#include <acpi/acpi_bus.h>
+#include <linux/pci-acpi.h>
+static void pcie_aspm_platform_init(void)
+{
+	pcie_osc_support_set(OSC_ACTIVE_STATE_PWR_SUPPORT|
+		OSC_CLOCK_PWR_CAPABILITY_SUPPORT);
+}
+#else
+static inline void pcie_aspm_platform_init(void) { }
+#endif
+
+static int __init pcie_aspm_init(void)
+{
+	if (aspm_disabled)
+		return 0;
+	pcie_aspm_platform_init();
+	return 0;
+}
+
+fs_initcall(pcie_aspm_init);
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 07d5c7424b01..284ef392c3ea 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -9,6 +9,7 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/cpumask.h>
+#include <linux/pci-aspm.h>
 #include "pci.h"
 
 #define CARDBUS_LATENCY_TIMER	176	/* secondary latency timer */
@@ -1014,6 +1015,10 @@ int pci_scan_slot(struct pci_bus *bus, int devfn)
 				break;
 		}
 	}
+
+	if (bus->self)
+		pcie_aspm_init_link_state(bus->self);
+
 	return nr;
 }
 
diff --git a/drivers/pci/remove.c b/drivers/pci/remove.c
index b6824833343f..bdc2a44d68e1 100644
--- a/drivers/pci/remove.c
+++ b/drivers/pci/remove.c
@@ -1,5 +1,6 @@
 #include <linux/pci.h>
 #include <linux/module.h>
+#include <linux/pci-aspm.h>
 #include "pci.h"
 
 static void pci_free_resources(struct pci_dev *dev)
@@ -24,6 +25,9 @@ static void pci_stop_dev(struct pci_dev *dev)
 		device_unregister(&dev->dev);
 		dev->is_added = 0;
 	}
+
+	if (dev->bus->self)
+		pcie_aspm_exit_link_state(dev);
 }
 
 static void pci_destroy_dev(struct pci_dev *dev)
diff --git a/include/linux/pci-aspm.h b/include/linux/pci-aspm.h
new file mode 100644
index 000000000000..a1a1e618e996
--- /dev/null
+++ b/include/linux/pci-aspm.h
@@ -0,0 +1,56 @@
+/*
+ *	aspm.h
+ *
+ *	PCI Express ASPM defines and function prototypes
+ *
+ *	Copyright (C) 2007 Intel Corp.
+ *		Zhang Yanmin (yanmin.zhang@intel.com)
+ *		Shaohua Li (shaohua.li@intel.com)
+ *
+ *	For more information, please consult the following manuals (look at
+ *	http://www.pcisig.com/ for how to get them):
+ *
+ *	PCI Express Specification
+ */
+
+#ifndef LINUX_ASPM_H
+#define LINUX_ASPM_H
+
+#include <linux/pci.h>
+
+#define PCIE_LINK_STATE_L0S	1
+#define PCIE_LINK_STATE_L1	2
+#define PCIE_LINK_STATE_CLKPM	4
+
+#ifdef CONFIG_PCIEASPM
+extern void pcie_aspm_init_link_state(struct pci_dev *pdev);
+extern void pcie_aspm_exit_link_state(struct pci_dev *pdev);
+extern void pcie_aspm_pm_state_change(struct pci_dev *pdev);
+extern void pci_disable_link_state(struct pci_dev *pdev, int state);
+#else
+static inline void pcie_aspm_init_link_state(struct pci_dev *pdev)
+{
+}
+static inline void pcie_aspm_exit_link_state(struct pci_dev *pdev)
+{
+}
+static inline void pcie_aspm_pm_state_change(struct pci_dev *pdev)
+{
+}
+static inline void pci_disable_link_state(struct pci_dev *pdev, int state)
+{
+}
+#endif
+
+#ifdef CONFIG_PCIEASPM_DEBUG /* this depends on CONFIG_PCIEASPM */
+extern void pcie_aspm_create_sysfs_dev_files(struct pci_dev *pdev);
+extern void pcie_aspm_remove_sysfs_dev_files(struct pci_dev *pdev);
+#else
+static inline void pcie_aspm_create_sysfs_dev_files(struct pci_dev *pdev)
+{
+}
+static inline void pcie_aspm_remove_sysfs_dev_files(struct pci_dev *pdev)
+{
+}
+#endif
+#endif /* LINUX_ASPM_H */
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 3b8a4e17052f..14bf3d236d19 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -128,6 +128,7 @@ struct pci_cap_saved_state {
 	u32 data[0];
 };
 
+struct pcie_link_state;
 /*
  * The pci_dev structure is used to describe PCI devices.
  */
@@ -164,6 +165,10 @@ struct pci_dev {
 					   this is D0-D3, D0 being fully functional,
 					   and D3 being off. */
 
+#ifdef CONFIG_PCIEASPM
+	struct pcie_link_state	*link_state;	/* ASPM link state. */
+#endif
+
 	pci_channel_state_t error_state;	/* current connectivity state */
 	struct	device	dev;		/* Generic device interface */
 
diff --git a/include/linux/pci_regs.h b/include/linux/pci_regs.h
index c1914a8b94a9..c0c1223c9194 100644
--- a/include/linux/pci_regs.h
+++ b/include/linux/pci_regs.h
@@ -395,9 +395,17 @@
 #define  PCI_EXP_DEVSTA_AUXPD	0x10	/* AUX Power Detected */
 #define  PCI_EXP_DEVSTA_TRPND	0x20	/* Transactions Pending */
 #define PCI_EXP_LNKCAP		12	/* Link Capabilities */
+#define  PCI_EXP_LNKCAP_ASPMS	0xc00	/* ASPM Support */
+#define  PCI_EXP_LNKCAP_L0SEL	0x7000	/* L0s Exit Latency */
+#define  PCI_EXP_LNKCAP_L1EL	0x38000	/* L1 Exit Latency */
+#define  PCI_EXP_LNKCAP_CLKPM	0x40000	/* L1 Clock Power Management */
 #define PCI_EXP_LNKCTL		16	/* Link Control */
+#define  PCI_EXP_LNKCTL_RL	0x20	/* Retrain Link */
+#define  PCI_EXP_LNKCTL_CCC	0x40	/* Common Clock COnfiguration */
 #define  PCI_EXP_LNKCTL_CLKREQ_EN 0x100	/* Enable clkreq */
 #define PCI_EXP_LNKSTA		18	/* Link Status */
+#define  PCI_EXP_LNKSTA_LT	0x800	/* Link Training */
+#define  PCI_EXP_LNKSTA_SLC	0x1000	/* Slot Clock Configuration */
 #define PCI_EXP_SLTCAP		20	/* Slot Capabilities */
 #define PCI_EXP_SLTCTL		24	/* Slot Control */
 #define PCI_EXP_SLTSTA		26	/* Slot Status */
-- 
cgit v1.2.3


From 842de40d93e00a5c40a1a7f520a6fbe422994e99 Mon Sep 17 00:00:00 2001
From: Bjorn Helgaas <bjorn.helgaas@hp.com>
Date: Tue, 4 Mar 2008 11:56:47 -0700
Subject: PCI: add generic pci_enable_resources()

Each architecture has its own pcibios_enable_resources() implementation.
These differ in many minor ways that have nothing to do with actual
architectural differences.  Follow-on patches will make most arches
use this generic version instead.

This version is based on powerpc, which seemed most up-to-date.  The only
functional difference from the x86 version is that this uses "!r->parent"
to check for resource collisions instead of "!r->start && r->end".

Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/setup-res.c | 43 +++++++++++++++++++++++++++++++++++++++++++
 include/linux/pci.h     |  1 +
 2 files changed, 44 insertions(+)

(limited to 'include/linux')

diff --git a/drivers/pci/setup-res.c b/drivers/pci/setup-res.c
index 9e4d485ba9cd..bad509e40fbc 100644
--- a/drivers/pci/setup-res.c
+++ b/drivers/pci/setup-res.c
@@ -263,3 +263,46 @@ void pdev_sort_resources(struct pci_dev *dev, struct resource_list *head)
 		}
 	}
 }
+
+int pci_enable_resources(struct pci_dev *dev, int mask)
+{
+	u16 cmd, old_cmd;
+	int i;
+	struct resource *r;
+
+	pci_read_config_word(dev, PCI_COMMAND, &cmd);
+	old_cmd = cmd;
+
+	for (i = 0; i < PCI_NUM_RESOURCES; i++) {
+		if (!(mask & (1 << i)))
+			continue;
+
+		r = &dev->resource[i];
+
+		if (!(r->flags & (IORESOURCE_IO | IORESOURCE_MEM)))
+			continue;
+		if ((i == PCI_ROM_RESOURCE) &&
+				(!(r->flags & IORESOURCE_ROM_ENABLE)))
+			continue;
+
+		if (!r->parent) {
+			dev_err(&dev->dev, "device not available because of "
+				"BAR %d [%llx:%llx] collisions\n", i,
+				(unsigned long long) r->start,
+				(unsigned long long) r->end);
+			return -EINVAL;
+		}
+
+		if (r->flags & IORESOURCE_IO)
+			cmd |= PCI_COMMAND_IO;
+		if (r->flags & IORESOURCE_MEM)
+			cmd |= PCI_COMMAND_MEMORY;
+	}
+
+	if (cmd != old_cmd) {
+		dev_info(&dev->dev, "enabling device (%04x -> %04x)\n",
+			 old_cmd, cmd);
+		pci_write_config_word(dev, PCI_COMMAND, cmd);
+	}
+	return 0;
+}
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 14bf3d236d19..e2f46b05cf8b 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -624,6 +624,7 @@ int pci_claim_resource(struct pci_dev *, int);
 void pci_assign_unassigned_resources(void);
 void pdev_enable_device(struct pci_dev *);
 void pdev_sort_resources(struct pci_dev *, struct resource_list *);
+int pci_enable_resources(struct pci_dev *, int mask);
 void pci_fixup_irqs(u8 (*)(struct pci_dev *, u8 *),
 		    int (*)(struct pci_dev *, u8, u8));
 #define HAVE_PCI_REQ_REGIONS	2
-- 
cgit v1.2.3


From 94e6108803469a37ee1e3c92dafdd1d59298602f Mon Sep 17 00:00:00 2001
From: Ben Hutchings <bhutchings@solarflare.com>
Date: Wed, 5 Mar 2008 16:52:39 +0000
Subject: PCI: Expose PCI VPD through sysfs

Vital Product Data (VPD) may be exposed by PCI devices in several
ways.  It is generally unsafe to read this information through the
existing interfaces to user-land because of stateful interfaces.

This adds:
- abstract operations for VPD access (struct pci_vpd_ops)
- VPD state information in struct pci_dev (struct pci_vpd)
- an implementation of the VPD access method specified in PCI 2.2
  (in access.c)
- a 'vpd' binary file in sysfs directories for PCI devices with VPD
  operations defined

It adds a probe for PCI 2.2 VPD in pci_scan_device() and release of
VPD state in pci_release_dev().

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 Documentation/ABI/testing/sysfs-bus-pci |  11 +++
 drivers/pci/access.c                    | 166 ++++++++++++++++++++++++++++++++
 drivers/pci/pci-sysfs.c                 | 109 ++++++++++++++++++---
 drivers/pci/pci.h                       |  19 ++++
 drivers/pci/probe.c                     |   3 +
 include/linux/pci.h                     |   3 +
 6 files changed, 297 insertions(+), 14 deletions(-)
 create mode 100644 Documentation/ABI/testing/sysfs-bus-pci

(limited to 'include/linux')

diff --git a/Documentation/ABI/testing/sysfs-bus-pci b/Documentation/ABI/testing/sysfs-bus-pci
new file mode 100644
index 000000000000..ceddcff4082a
--- /dev/null
+++ b/Documentation/ABI/testing/sysfs-bus-pci
@@ -0,0 +1,11 @@
+What:		/sys/bus/pci/devices/.../vpd
+Date:		February 2008
+Contact:	Ben Hutchings <bhutchings@solarflare.com>
+Description:
+		A file named vpd in a device directory will be a
+		binary file containing the Vital Product Data for the
+		device.  It should follow the VPD format defined in
+		PCI Specification 2.1 or 2.2, but users should consider
+		that some devices may have malformatted data.  If the
+		underlying VPD has a writable section then the
+		corresponding section of this file will be writable.
diff --git a/drivers/pci/access.c b/drivers/pci/access.c
index fc405f0165d9..ec8f7002b09d 100644
--- a/drivers/pci/access.c
+++ b/drivers/pci/access.c
@@ -1,3 +1,4 @@
+#include <linux/delay.h>
 #include <linux/pci.h>
 #include <linux/module.h>
 #include <linux/sched.h>
@@ -126,6 +127,171 @@ PCI_USER_WRITE_CONFIG(byte, u8)
 PCI_USER_WRITE_CONFIG(word, u16)
 PCI_USER_WRITE_CONFIG(dword, u32)
 
+/* VPD access through PCI 2.2+ VPD capability */
+
+#define PCI_VPD_PCI22_SIZE (PCI_VPD_ADDR_MASK + 1)
+
+struct pci_vpd_pci22 {
+	struct pci_vpd base;
+	spinlock_t lock; /* controls access to hardware and the flags */
+	u8	cap;
+	bool	busy;
+	bool	flag; /* value of F bit to wait for */
+};
+
+/* Wait for last operation to complete */
+static int pci_vpd_pci22_wait(struct pci_dev *dev)
+{
+	struct pci_vpd_pci22 *vpd =
+		container_of(dev->vpd, struct pci_vpd_pci22, base);
+	u16 flag, status;
+	int wait;
+	int ret;
+
+	if (!vpd->busy)
+		return 0;
+
+	flag = vpd->flag ? PCI_VPD_ADDR_F : 0;
+	wait = vpd->flag ? 10 : 1000; /* read: 100 us; write: 10 ms */
+	for (;;) {
+		ret = pci_user_read_config_word(dev,
+						vpd->cap + PCI_VPD_ADDR,
+						&status);
+		if (ret < 0)
+			return ret;
+		if ((status & PCI_VPD_ADDR_F) == flag) {
+			vpd->busy = false;
+			return 0;
+		}
+		if (wait-- == 0)
+			return -ETIMEDOUT;
+		udelay(10);
+	}
+}
+
+static int pci_vpd_pci22_read(struct pci_dev *dev, int pos, int size,
+			      char *buf)
+{
+	struct pci_vpd_pci22 *vpd =
+		container_of(dev->vpd, struct pci_vpd_pci22, base);
+	u32 val;
+	int ret;
+	int begin, end, i;
+
+	if (pos < 0 || pos > PCI_VPD_PCI22_SIZE ||
+	    size > PCI_VPD_PCI22_SIZE  - pos)
+		return -EINVAL;
+	if (size == 0)
+		return 0;
+
+	spin_lock_irq(&vpd->lock);
+	ret = pci_vpd_pci22_wait(dev);
+	if (ret < 0)
+		goto out;
+	ret = pci_user_write_config_word(dev, vpd->cap + PCI_VPD_ADDR,
+					 pos & ~3);
+	if (ret < 0)
+		goto out;
+	vpd->busy = true;
+	vpd->flag = 1;
+	ret = pci_vpd_pci22_wait(dev);
+	if (ret < 0)
+		goto out;
+	ret = pci_user_read_config_dword(dev, vpd->cap + PCI_VPD_DATA,
+					 &val);
+out:
+	spin_unlock_irq(&vpd->lock);
+	if (ret < 0)
+		return ret;
+
+	/* Convert to bytes */
+	begin = pos & 3;
+	end = min(4, begin + size);
+	for (i = 0; i < end; ++i) {
+		if (i >= begin)
+			*buf++ = val;
+		val >>= 8;
+	}
+	return end - begin;
+}
+
+static int pci_vpd_pci22_write(struct pci_dev *dev, int pos, int size,
+			       const char *buf)
+{
+	struct pci_vpd_pci22 *vpd =
+		container_of(dev->vpd, struct pci_vpd_pci22, base);
+	u32 val;
+	int ret;
+
+	if (pos < 0 || pos > PCI_VPD_PCI22_SIZE || pos & 3 ||
+	    size > PCI_VPD_PCI22_SIZE - pos || size < 4)
+		return -EINVAL;
+
+	val = (u8) *buf++;
+	val |= ((u8) *buf++) << 8;
+	val |= ((u8) *buf++) << 16;
+	val |= ((u32)(u8) *buf++) << 24;
+
+	spin_lock_irq(&vpd->lock);
+	ret = pci_vpd_pci22_wait(dev);
+	if (ret < 0)
+		goto out;
+	ret = pci_user_write_config_dword(dev, vpd->cap + PCI_VPD_DATA,
+					  val);
+	if (ret < 0)
+		goto out;
+	ret = pci_user_write_config_word(dev, vpd->cap + PCI_VPD_ADDR,
+					 pos | PCI_VPD_ADDR_F);
+	if (ret < 0)
+		goto out;
+	vpd->busy = true;
+	vpd->flag = 0;
+	ret = pci_vpd_pci22_wait(dev);
+out:
+	spin_unlock_irq(&vpd->lock);
+	if (ret < 0)
+		return ret;
+
+	return 4;
+}
+
+static int pci_vpd_pci22_get_size(struct pci_dev *dev)
+{
+	return PCI_VPD_PCI22_SIZE;
+}
+
+static void pci_vpd_pci22_release(struct pci_dev *dev)
+{
+	kfree(container_of(dev->vpd, struct pci_vpd_pci22, base));
+}
+
+static struct pci_vpd_ops pci_vpd_pci22_ops = {
+	.read = pci_vpd_pci22_read,
+	.write = pci_vpd_pci22_write,
+	.get_size = pci_vpd_pci22_get_size,
+	.release = pci_vpd_pci22_release,
+};
+
+int pci_vpd_pci22_init(struct pci_dev *dev)
+{
+	struct pci_vpd_pci22 *vpd;
+	u8 cap;
+
+	cap = pci_find_capability(dev, PCI_CAP_ID_VPD);
+	if (!cap)
+		return -ENODEV;
+	vpd = kzalloc(sizeof(*vpd), GFP_ATOMIC);
+	if (!vpd)
+		return -ENOMEM;
+
+	vpd->base.ops = &pci_vpd_pci22_ops;
+	spin_lock_init(&vpd->lock);
+	vpd->cap = cap;
+	vpd->busy = false;
+	dev->vpd = &vpd->base;
+	return 0;
+}
+
 /**
  * pci_block_user_cfg_access - Block userspace PCI config reads/writes
  * @dev:	pci device struct
diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c
index f5b0b622c189..ae9a7695be97 100644
--- a/drivers/pci/pci-sysfs.c
+++ b/drivers/pci/pci-sysfs.c
@@ -343,6 +343,58 @@ pci_write_config(struct kobject *kobj, struct bin_attribute *bin_attr,
 	return count;
 }
 
+static ssize_t
+pci_read_vpd(struct kobject *kobj, struct bin_attribute *bin_attr,
+	     char *buf, loff_t off, size_t count)
+{
+	struct pci_dev *dev =
+		to_pci_dev(container_of(kobj, struct device, kobj));
+	int end;
+	int ret;
+
+	if (off > bin_attr->size)
+		count = 0;
+	else if (count > bin_attr->size - off)
+		count = bin_attr->size - off;
+	end = off + count;
+
+	while (off < end) {
+		ret = dev->vpd->ops->read(dev, off, end - off, buf);
+		if (ret < 0)
+			return ret;
+		buf += ret;
+		off += ret;
+	}
+
+	return count;
+}
+
+static ssize_t
+pci_write_vpd(struct kobject *kobj, struct bin_attribute *bin_attr,
+	      char *buf, loff_t off, size_t count)
+{
+	struct pci_dev *dev =
+		to_pci_dev(container_of(kobj, struct device, kobj));
+	int end;
+	int ret;
+
+	if (off > bin_attr->size)
+		count = 0;
+	else if (count > bin_attr->size - off)
+		count = bin_attr->size - off;
+	end = off + count;
+
+	while (off < end) {
+		ret = dev->vpd->ops->write(dev, off, end - off, buf);
+		if (ret < 0)
+			return ret;
+		buf += ret;
+		off += ret;
+	}
+
+	return count;
+}
+
 #ifdef HAVE_PCI_LEGACY
 /**
  * pci_read_legacy_io - read byte(s) from legacy I/O port space
@@ -611,7 +663,7 @@ int __attribute__ ((weak)) pcibios_add_platform_entries(struct pci_dev *dev)
 
 int __must_check pci_create_sysfs_dev_files (struct pci_dev *pdev)
 {
-	struct bin_attribute *rom_attr = NULL;
+	struct bin_attribute *attr = NULL;
 	int retval;
 
 	if (!sysfs_initialized)
@@ -624,22 +676,41 @@ int __must_check pci_create_sysfs_dev_files (struct pci_dev *pdev)
 	if (retval)
 		goto err;
 
+	/* If the device has VPD, try to expose it in sysfs. */
+	if (pdev->vpd) {
+		attr = kzalloc(sizeof(*attr), GFP_ATOMIC);
+		if (attr) {
+			pdev->vpd->attr = attr;
+			attr->size = pdev->vpd->ops->get_size(pdev);
+			attr->attr.name = "vpd";
+			attr->attr.mode = S_IRUGO | S_IWUSR;
+			attr->read = pci_read_vpd;
+			attr->write = pci_write_vpd;
+			retval = sysfs_create_bin_file(&pdev->dev.kobj, attr);
+			if (retval)
+				goto err_vpd;
+		} else {
+			retval = -ENOMEM;
+			goto err_config_file;
+		}
+	}
+
 	retval = pci_create_resource_files(pdev);
 	if (retval)
-		goto err_bin_file;
+		goto err_vpd_file;
 
 	/* If the device has a ROM, try to expose it in sysfs. */
 	if (pci_resource_len(pdev, PCI_ROM_RESOURCE) ||
 	    (pdev->resource[PCI_ROM_RESOURCE].flags & IORESOURCE_ROM_SHADOW)) {
-		rom_attr = kzalloc(sizeof(*rom_attr), GFP_ATOMIC);
-		if (rom_attr) {
-			pdev->rom_attr = rom_attr;
-			rom_attr->size = pci_resource_len(pdev, PCI_ROM_RESOURCE);
-			rom_attr->attr.name = "rom";
-			rom_attr->attr.mode = S_IRUSR;
-			rom_attr->read = pci_read_rom;
-			rom_attr->write = pci_write_rom;
-			retval = sysfs_create_bin_file(&pdev->dev.kobj, rom_attr);
+		attr = kzalloc(sizeof(*attr), GFP_ATOMIC);
+		if (attr) {
+			pdev->rom_attr = attr;
+			attr->size = pci_resource_len(pdev, PCI_ROM_RESOURCE);
+			attr->attr.name = "rom";
+			attr->attr.mode = S_IRUSR;
+			attr->read = pci_read_rom;
+			attr->write = pci_write_rom;
+			retval = sysfs_create_bin_file(&pdev->dev.kobj, attr);
 			if (retval)
 				goto err_rom;
 		} else {
@@ -657,12 +728,18 @@ int __must_check pci_create_sysfs_dev_files (struct pci_dev *pdev)
 
 err_rom_file:
 	if (pci_resource_len(pdev, PCI_ROM_RESOURCE))
-		sysfs_remove_bin_file(&pdev->dev.kobj, rom_attr);
+		sysfs_remove_bin_file(&pdev->dev.kobj, pdev->rom_attr);
 err_rom:
-	kfree(rom_attr);
+	kfree(pdev->rom_attr);
 err_resource_files:
 	pci_remove_resource_files(pdev);
-err_bin_file:
+err_vpd_file:
+	if (pdev->vpd) {
+		sysfs_remove_bin_file(&pdev->dev.kobj, pdev->vpd->attr);
+err_vpd:
+		kfree(pdev->vpd->attr);
+	}
+err_config_file:
 	if (pdev->cfg_size < 4096)
 		sysfs_remove_bin_file(&pdev->dev.kobj, &pci_config_attr);
 	else
@@ -684,6 +761,10 @@ void pci_remove_sysfs_dev_files(struct pci_dev *pdev)
 
 	pcie_aspm_remove_sysfs_dev_files(pdev);
 
+	if (pdev->vpd) {
+		sysfs_remove_bin_file(&pdev->dev.kobj, pdev->vpd->attr);
+		kfree(pdev->vpd->attr);
+	}
 	if (pdev->cfg_size < 4096)
 		sysfs_remove_bin_file(&pdev->dev.kobj, &pci_config_attr);
 	else
diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h
index eabeb1f2ec99..0a497c1b4227 100644
--- a/drivers/pci/pci.h
+++ b/drivers/pci/pci.h
@@ -18,6 +18,25 @@ extern int pci_user_write_config_byte(struct pci_dev *dev, int where, u8 val);
 extern int pci_user_write_config_word(struct pci_dev *dev, int where, u16 val);
 extern int pci_user_write_config_dword(struct pci_dev *dev, int where, u32 val);
 
+struct pci_vpd_ops {
+	int (*read)(struct pci_dev *dev, int pos, int size, char *buf);
+	int (*write)(struct pci_dev *dev, int pos, int size, const char *buf);
+	int (*get_size)(struct pci_dev *dev);
+	void (*release)(struct pci_dev *dev);
+};
+
+struct pci_vpd {
+	struct pci_vpd_ops *ops;
+	struct bin_attribute *attr; /* descriptor for sysfs VPD entry */
+};
+
+extern int pci_vpd_pci22_init(struct pci_dev *dev);
+static inline void pci_vpd_release(struct pci_dev *dev)
+{
+	if (dev->vpd)
+		dev->vpd->ops->release(dev);
+}
+
 /* PCI /proc functions */
 #ifdef CONFIG_PROC_FS
 extern int pci_proc_attach_device(struct pci_dev *dev);
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 284ef392c3ea..c2e99fd87faf 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -794,6 +794,7 @@ static void pci_release_dev(struct device *dev)
 	struct pci_dev *pci_dev;
 
 	pci_dev = to_pci_dev(dev);
+	pci_vpd_release(pci_dev);
 	kfree(pci_dev);
 }
 
@@ -933,6 +934,8 @@ pci_scan_device(struct pci_bus *bus, int devfn)
 		return NULL;
 	}
 
+	pci_vpd_pci22_init(dev);
+
 	return dev;
 }
 
diff --git a/include/linux/pci.h b/include/linux/pci.h
index e2f46b05cf8b..292491324b01 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -20,6 +20,8 @@
 /* Include the pci register defines */
 #include <linux/pci_regs.h>
 
+struct pci_vpd;
+
 /*
  * The PCI interface treats multi-function devices as independent
  * devices.  The slot/function address of each device is encoded
@@ -206,6 +208,7 @@ struct pci_dev {
 #ifdef CONFIG_PCI_MSI
 	struct list_head msi_list;
 #endif
+	struct pci_vpd *vpd;
 };
 
 extern struct pci_dev *alloc_pci_dev(void);
-- 
cgit v1.2.3


From 884525655d07fdee9245716b998ecdc45cdd8007 Mon Sep 17 00:00:00 2001
From: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Date: Sun, 30 Mar 2008 19:50:14 +0400
Subject: PCI: clean up resource alignment management

Done per Linus' request and suggestions. Linus has explained that
better than I'll be able to explain:

On Thu, Mar 27, 2008 at 10:12:10AM -0700, Linus Torvalds wrote:
> Actually, before we go any further, there might be a less intrusive
> alternative: add just a couple of flags to the resource flags field (we
> still have something like 8 unused bits on 32-bit), and use those to
> implement a generic "resource_alignment()" routine.
>
> Two flags would do it:
>
>  - IORESOURCE_SIZEALIGN: size indicates alignment (regular PCI device
>    resources)
>
>  - IORESOURCE_STARTALIGN: start field is alignment (PCI bus resources
>    during probing)
>
> and then the case of both flags zero (or both bits set) would actually be
> "invalid", and we would also clear the IORESOURCE_STARTALIGN flag when we
> actually allocate the resource (so that we don't use the "start" field as
> alignment incorrectly when it no longer indicates alignment).
>
> That wouldn't be totally generic, but it would have the nice property of
> automatically at least add sanity checking for that whole "res->start has
> the odd meaning of 'alignment' during probing" and remove the need for a
> new field, and it would allow us to have a generic "resource_alignment()"
> routine that just gets a resource pointer.

Besides, I removed IORESOURCE_BUS_HAS_VGA flag which was unused for ages.

Signed-off-by: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Gary Hade <garyhade@us.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 drivers/pci/probe.c     |  5 +++--
 drivers/pci/setup-bus.c |  3 +++
 drivers/pci/setup-res.c | 42 +++++++++++++++++++++++-------------------
 include/linux/ioport.h  |  5 ++++-
 kernel/resource.c       | 18 ++++++++++++++++++
 5 files changed, 51 insertions(+), 22 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index c2e99fd87faf..33d9b8bea6e0 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -235,7 +235,7 @@ static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom)
 			res->flags |= l & ~PCI_BASE_ADDRESS_IO_MASK;
 		}
 		res->end = res->start + (unsigned long) sz;
-		res->flags |= pci_calc_resource_flags(l);
+		res->flags |= pci_calc_resource_flags(l) | IORESOURCE_SIZEALIGN;
 		if (is_64bit_memory(l)) {
 			u32 szhi, lhi;
 
@@ -288,7 +288,8 @@ static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom)
 			if (sz) {
 				res->flags = (l & IORESOURCE_ROM_ENABLE) |
 				  IORESOURCE_MEM | IORESOURCE_PREFETCH |
-				  IORESOURCE_READONLY | IORESOURCE_CACHEABLE;
+				  IORESOURCE_READONLY | IORESOURCE_CACHEABLE |
+				  IORESOURCE_SIZEALIGN;
 				res->start = l & PCI_ROM_ADDRESS_MASK;
 				res->end = res->start + (unsigned long) sz;
 			}
diff --git a/drivers/pci/setup-bus.c b/drivers/pci/setup-bus.c
index f7cb8e0758b4..5cf84568c9e4 100644
--- a/drivers/pci/setup-bus.c
+++ b/drivers/pci/setup-bus.c
@@ -65,6 +65,7 @@ static void pbus_assign_resources_sorted(struct pci_bus *bus)
 		res = list->res;
 		idx = res - &list->dev->resource[0];
 		if (pci_assign_resource(list->dev, idx)) {
+			/* FIXME: get rid of this */
 			res->start = 0;
 			res->end = 0;
 			res->flags = 0;
@@ -327,6 +328,7 @@ static void pbus_size_io(struct pci_bus *bus)
 	/* Alignment of the IO window is always 4K */
 	b_res->start = 4096;
 	b_res->end = b_res->start + size - 1;
+	b_res->flags |= IORESOURCE_STARTALIGN;
 }
 
 /* Calculate the size of the bus and minimal alignment which
@@ -401,6 +403,7 @@ static int pbus_size_mem(struct pci_bus *bus, unsigned long mask, unsigned long
 	}
 	b_res->start = min_align;
 	b_res->end = size + min_align - 1;
+	b_res->flags |= IORESOURCE_STARTALIGN;
 	return 1;
 }
 
diff --git a/drivers/pci/setup-res.c b/drivers/pci/setup-res.c
index bad509e40fbc..7d35cdf4579f 100644
--- a/drivers/pci/setup-res.c
+++ b/drivers/pci/setup-res.c
@@ -137,10 +137,16 @@ int pci_assign_resource(struct pci_dev *dev, int resno)
 
 	size = res->end - res->start + 1;
 	min = (res->flags & IORESOURCE_IO) ? PCIBIOS_MIN_IO : PCIBIOS_MIN_MEM;
-	/* The bridge resources are special, as their
-	   size != alignment. Sizing routines return
-	   required alignment in the "start" field. */
-	align = (resno < PCI_BRIDGE_RESOURCES) ? size : res->start;
+
+	align = resource_alignment(res);
+	if (!align) {
+		printk(KERN_ERR "PCI: Cannot allocate resource (bogus "
+			"alignment) %d [%llx:%llx] (flags %lx) of %s\n",
+			resno, (unsigned long long)res->start,
+			(unsigned long long)res->end, res->flags,
+			pci_name(dev));
+		return -EINVAL;
+	}
 
 	/* First, try exact prefetching match.. */
 	ret = pci_bus_alloc_resource(bus, res, size, align, min,
@@ -164,8 +170,10 @@ int pci_assign_resource(struct pci_dev *dev, int resno)
 			res->flags & IORESOURCE_IO ? "I/O" : "mem",
 			resno, (unsigned long long)size,
 			(unsigned long long)res->start, pci_name(dev));
-	} else if (resno < PCI_BRIDGE_RESOURCES) {
-		pci_update_resource(dev, res, resno);
+	} else {
+		res->flags &= ~IORESOURCE_STARTALIGN;
+		if (resno < PCI_BRIDGE_RESOURCES)
+			pci_update_resource(dev, res, resno);
 	}
 
 	return ret;
@@ -226,29 +234,25 @@ void pdev_sort_resources(struct pci_dev *dev, struct resource_list *head)
 		if (r->flags & IORESOURCE_PCI_FIXED)
 			continue;
 
-		r_align = r->end - r->start;
-		
 		if (!(r->flags) || r->parent)
 			continue;
+
+		r_align = resource_alignment(r);
 		if (!r_align) {
-			printk(KERN_WARNING "PCI: Ignore bogus resource %d "
-				"[%llx:%llx] of %s\n",
+			printk(KERN_WARNING "PCI: bogus alignment of resource "
+				"%d [%llx:%llx] (flags %lx) of %s\n",
 				i, (unsigned long long)r->start,
-				(unsigned long long)r->end, pci_name(dev));
+				(unsigned long long)r->end, r->flags,
+				pci_name(dev));
 			continue;
 		}
-		r_align = (i < PCI_BRIDGE_RESOURCES) ? r_align + 1 : r->start;
 		for (list = head; ; list = list->next) {
 			resource_size_t align = 0;
 			struct resource_list *ln = list->next;
-			int idx;
 
-			if (ln) {
-				idx = ln->res - &ln->dev->resource[0];
-				align = (idx < PCI_BRIDGE_RESOURCES) ?
-					ln->res->end - ln->res->start + 1 :
-					ln->res->start;
-			}
+			if (ln)
+				align = resource_alignment(ln->res);
+
 			if (r_align > align) {
 				tmp = kmalloc(sizeof(*tmp), GFP_KERNEL);
 				if (!tmp)
diff --git a/include/linux/ioport.h b/include/linux/ioport.h
index 605d237364d2..d5d40a9f7929 100644
--- a/include/linux/ioport.h
+++ b/include/linux/ioport.h
@@ -44,7 +44,9 @@ struct resource_list {
 #define IORESOURCE_CACHEABLE	0x00004000
 #define IORESOURCE_RANGELENGTH	0x00008000
 #define IORESOURCE_SHADOWABLE	0x00010000
-#define IORESOURCE_BUS_HAS_VGA	0x00080000
+
+#define IORESOURCE_SIZEALIGN	0x00020000	/* size indicates alignment */
+#define IORESOURCE_STARTALIGN	0x00040000	/* start field is alignment */
 
 #define IORESOURCE_DISABLED	0x10000000
 #define IORESOURCE_UNSET	0x20000000
@@ -110,6 +112,7 @@ extern int allocate_resource(struct resource *root, struct resource *new,
 			     void *alignf_data);
 int adjust_resource(struct resource *res, resource_size_t start,
 		    resource_size_t size);
+resource_size_t resource_alignment(struct resource *res);
 
 /* Convenience shorthand with allocation */
 #define request_region(start,n,name)	__request_region(&ioport_resource, (start), (n), (name))
diff --git a/kernel/resource.c b/kernel/resource.c
index 82aea814d409..cee12cc47cab 100644
--- a/kernel/resource.c
+++ b/kernel/resource.c
@@ -486,6 +486,24 @@ int adjust_resource(struct resource *res, resource_size_t start, resource_size_t
 
 EXPORT_SYMBOL(adjust_resource);
 
+/**
+ * resource_alignment - calculate resource's alignment
+ * @res: resource pointer
+ *
+ * Returns alignment on success, 0 (invalid alignment) on failure.
+ */
+resource_size_t resource_alignment(struct resource *res)
+{
+	switch (res->flags & (IORESOURCE_SIZEALIGN | IORESOURCE_STARTALIGN)) {
+	case IORESOURCE_SIZEALIGN:
+		return res->end - res->start + 1;
+	case IORESOURCE_STARTALIGN:
+		return res->start;
+	default:
+		return 0;
+	}
+}
+
 /*
  * This is compatibility stuff for IO resources.
  *
-- 
cgit v1.2.3


From 3f34d024c12e49fbce4009d094ae1d287084e511 Mon Sep 17 00:00:00 2001
From: Dave Young <hidave.darkstar@gmail.com>
Date: Fri, 18 Apr 2008 13:38:57 -0700
Subject: jiffies: add time_is_after_jiffies and others which compare with
 jiffies

Most of time_after like macros usages just compare jiffies and another number,
so here add some time_is_* macros for convenience.

Signed-off-by: Dave Young <hidave.darkstar@gmail.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 include/linux/jiffies.h | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

(limited to 'include/linux')

diff --git a/include/linux/jiffies.h b/include/linux/jiffies.h
index e0b5b684d83f..e377e34e589e 100644
--- a/include/linux/jiffies.h
+++ b/include/linux/jiffies.h
@@ -134,6 +134,22 @@ static inline u64 get_jiffies_64(void)
 	 ((__s64)(a) - (__s64)(b) >= 0))
 #define time_before_eq64(a,b)	time_after_eq64(b,a)
 
+/*
+ * These four macros compare jiffies and 'a' for convenience.
+ */
+
+/* time_is_before_jiffies(a) return true if a is before jiffies */
+#define time_is_before_jiffies(a) time_after(jiffies, a)
+
+/* time_is_after_jiffies(a) return true if a is after jiffies */
+#define time_is_after_jiffies(a) time_before(jiffies, a)
+
+/* time_is_before_eq_jiffies(a) return true if a is before or equal to jiffies*/
+#define time_is_before_eq_jiffies(a) time_after_eq(jiffies, a)
+
+/* time_is_after_eq_jiffies(a) return true if a is after or equal to jiffies*/
+#define time_is_after_eq_jiffies(a) time_before_eq(jiffies, a)
+
 /*
  * Have the 32 bit jiffies value wrap 5 minutes after boot
  * so jiffies wrap bugs show up earlier.
-- 
cgit v1.2.3


From 7fd097d42b90afadae4867db5d580bcd7b3b596d Mon Sep 17 00:00:00 2001
From: Akinobu Mita <akinobu.mita@gmail.com>
Date: Wed, 26 Mar 2008 12:09:02 +0100
Subject: cdrom: use list_head for cdrom_device_info list

Use list_head for cdrom_device_info list instead of opencoded
singly list handling.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 drivers/cdrom/cdrom.c | 29 ++++++-----------------------
 include/linux/cdrom.h |  3 ++-
 2 files changed, 8 insertions(+), 24 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/cdrom/cdrom.c b/drivers/cdrom/cdrom.c
index 326ef1b4a4ea..c4213b7d0b20 100644
--- a/drivers/cdrom/cdrom.c
+++ b/drivers/cdrom/cdrom.c
@@ -362,7 +362,7 @@ static int cdrom_get_disc_info(struct cdrom_device_info *cdi, disc_information *
 
 static void cdrom_sysctl_register(void);
 
-static struct cdrom_device_info *topCdromPtr;
+static LIST_HEAD(cdrom_list);
 
 static int cdrom_dummy_generic_packet(struct cdrom_device_info *cdi,
 				      struct packet_command *cgc)
@@ -436,35 +436,18 @@ int register_cdrom(struct cdrom_device_info *cdi)
 
 	cdinfo(CD_REG_UNREG, "drive \"/dev/%s\" registered\n", cdi->name);
 	mutex_lock(&cdrom_mutex);
-	cdi->next = topCdromPtr; 	
-	topCdromPtr = cdi;
+	list_add(&cdi->list, &cdrom_list);
 	mutex_unlock(&cdrom_mutex);
 	return 0;
 }
 #undef ENSURE
 
-int unregister_cdrom(struct cdrom_device_info *unreg)
+int unregister_cdrom(struct cdrom_device_info *cdi)
 {
-	struct cdrom_device_info *cdi, *prev;
 	cdinfo(CD_OPEN, "entering unregister_cdrom\n"); 
 
-	prev = NULL;
 	mutex_lock(&cdrom_mutex);
-	cdi = topCdromPtr;
-	while (cdi && cdi != unreg) {
-		prev = cdi;
-		cdi = cdi->next;
-	}
-
-	if (cdi == NULL) {
-		mutex_unlock(&cdrom_mutex);
-		return -2;
-	}
-	if (prev)
-		prev->next = cdi->next;
-	else
-		topCdromPtr = cdi->next;
-
+	list_del(&cdi->list);
 	mutex_unlock(&cdrom_mutex);
 
 	if (cdi->exit)
@@ -3306,7 +3289,7 @@ static int cdrom_print_info(const char *header, int val, char *info,
 
 	*pos += ret;
 
-	for (cdi = topCdromPtr; cdi; cdi = cdi->next) {
+	list_for_each_entry(cdi, &cdrom_list, list) {
 		switch (option) {
 		case CTL_NAME:
 			ret = scnprintf(info + *pos, max_size - *pos,
@@ -3428,7 +3411,7 @@ static void cdrom_update_settings(void)
 	struct cdrom_device_info *cdi;
 
 	mutex_lock(&cdrom_mutex);
-	for (cdi = topCdromPtr; cdi != NULL; cdi = cdi->next) {
+	list_for_each_entry(cdi, &cdrom_list, list) {
 		if (autoclose && CDROM_CAN(CDC_CLOSE_TRAY))
 			cdi->options |= CDO_AUTO_CLOSE;
 		else if (!autoclose)
diff --git a/include/linux/cdrom.h b/include/linux/cdrom.h
index a5cd2047624e..40e05d0a6e45 100644
--- a/include/linux/cdrom.h
+++ b/include/linux/cdrom.h
@@ -910,6 +910,7 @@ struct mode_page_header {
 #ifdef __KERNEL__
 #include <linux/fs.h>		/* not really needed, later.. */
 #include <linux/device.h>
+#include <linux/list.h>
 
 struct packet_command
 {
@@ -934,7 +935,7 @@ struct packet_command
 /* Uniform cdrom data structures for cdrom.c */
 struct cdrom_device_info {
 	struct cdrom_device_ops  *ops;  /* link to device_ops */
-	struct cdrom_device_info *next; /* next device_info for this major */
+	struct list_head list;		/* linked list of all device_info */
 	struct gendisk *disk;		/* matching block layer disk */
 	void *handle;		        /* driver-dependent data */
 /* specifications */
-- 
cgit v1.2.3


From 0a0c4114df4a6903bccb65b06cabb6ddc968f877 Mon Sep 17 00:00:00 2001
From: Akinobu Mita <akinobu.mita@gmail.com>
Date: Wed, 26 Mar 2008 12:09:02 +0100
Subject: cdrom: make unregister_cdrom() return void

Now unregister_cdrom() always returns 0.
Make it return void and update all callers that check the return value.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Adrian McMenamin <adrian@mcmen.demon.co.uk>
Cc: Borislav Petkov <petkovbb@gmail.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 Documentation/cdrom/cdrom-standard.tex | 2 +-
 drivers/cdrom/cdrom.c                  | 3 +--
 drivers/cdrom/gdrom.c                  | 4 +++-
 drivers/cdrom/viocd.c                  | 5 +----
 drivers/ide/ide-cd.c                   | 5 ++---
 include/linux/cdrom.h                  | 2 +-
 6 files changed, 9 insertions(+), 12 deletions(-)

(limited to 'include/linux')

diff --git a/Documentation/cdrom/cdrom-standard.tex b/Documentation/cdrom/cdrom-standard.tex
index c713aeb020c4..c06233fe52ac 100644
--- a/Documentation/cdrom/cdrom-standard.tex
+++ b/Documentation/cdrom/cdrom-standard.tex
@@ -777,7 +777,7 @@ Note that a driver must have one static structure, $<device>_dops$, while
 it may have as many structures $<device>_info$ as there are minor devices
 active. $Register_cdrom()$ builds a linked list from these. 
 
-\subsection{$Int\ unregister_cdrom(struct\ cdrom_device_info * cdi)$}
+\subsection{$Void\ unregister_cdrom(struct\ cdrom_device_info * cdi)$}
 
 Unregistering device $cdi$ with minor number $MINOR(cdi\to dev)$ removes
 the minor device from the list. If it was the last registered minor for
diff --git a/drivers/cdrom/cdrom.c b/drivers/cdrom/cdrom.c
index c4213b7d0b20..663a7f7dc580 100644
--- a/drivers/cdrom/cdrom.c
+++ b/drivers/cdrom/cdrom.c
@@ -442,7 +442,7 @@ int register_cdrom(struct cdrom_device_info *cdi)
 }
 #undef ENSURE
 
-int unregister_cdrom(struct cdrom_device_info *cdi)
+void unregister_cdrom(struct cdrom_device_info *cdi)
 {
 	cdinfo(CD_OPEN, "entering unregister_cdrom\n"); 
 
@@ -455,7 +455,6 @@ int unregister_cdrom(struct cdrom_device_info *cdi)
 
 	cdi->ops->n_minors--;
 	cdinfo(CD_REG_UNREG, "drive \"/dev/%s\" unregistered\n", cdi->name);
-	return 0;
 }
 
 int cdrom_get_media_event(struct cdrom_device_info *cdi,
diff --git a/drivers/cdrom/gdrom.c b/drivers/cdrom/gdrom.c
index 4e2bbcccc064..71ec426ecffc 100644
--- a/drivers/cdrom/gdrom.c
+++ b/drivers/cdrom/gdrom.c
@@ -827,7 +827,9 @@ static int __devexit remove_gdrom(struct platform_device *devptr)
 	del_gendisk(gd.disk);
 	if (gdrom_major)
 		unregister_blkdev(gdrom_major, GDROM_DEV_NAME);
-	return unregister_cdrom(gd.cd_info);
+	unregister_cdrom(gd.cd_info);
+
+	return 0;
 }
 
 static struct platform_driver gdrom_driver = {
diff --git a/drivers/cdrom/viocd.c b/drivers/cdrom/viocd.c
index cac06bc1754b..b74b6c2768a8 100644
--- a/drivers/cdrom/viocd.c
+++ b/drivers/cdrom/viocd.c
@@ -650,10 +650,7 @@ static int viocd_remove(struct vio_dev *vdev)
 {
 	struct disk_info *d = &viocd_diskinfo[vdev->unit_address];
 
-	if (unregister_cdrom(&d->viocd_info) != 0)
-		printk(VIOCD_KERN_WARNING
-				"Cannot unregister viocd CD-ROM %s!\n",
-				d->viocd_info.name);
+	unregister_cdrom(&d->viocd_info);
 	del_gendisk(d->viocd_disk);
 	blk_cleanup_queue(d->viocd_disk->queue);
 	put_disk(d->viocd_disk);
diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c
index 396000208f81..fe5aefbf8339 100644
--- a/drivers/ide/ide-cd.c
+++ b/drivers/ide/ide-cd.c
@@ -2032,9 +2032,8 @@ static void ide_cd_release(struct kref *kref)
 
 	kfree(info->buffer);
 	kfree(info->toc);
-	if (devinfo->handle == drive && unregister_cdrom(devinfo))
-		printk(KERN_ERR "%s: %s failed to unregister device from the cdrom "
-				"driver.\n", __FUNCTION__, drive->name);
+	if (devinfo->handle == drive)
+		unregister_cdrom(devinfo);
 	drive->dsc_overlap = 0;
 	drive->driver_data = NULL;
 	blk_queue_prep_rq(drive->queue, NULL);
diff --git a/include/linux/cdrom.h b/include/linux/cdrom.h
index 40e05d0a6e45..5db265ea60f6 100644
--- a/include/linux/cdrom.h
+++ b/include/linux/cdrom.h
@@ -995,7 +995,7 @@ extern int cdrom_ioctl(struct file *file, struct cdrom_device_info *cdi,
 extern int cdrom_media_changed(struct cdrom_device_info *);
 
 extern int register_cdrom(struct cdrom_device_info *cdi);
-extern int unregister_cdrom(struct cdrom_device_info *cdi);
+extern void unregister_cdrom(struct cdrom_device_info *cdi);
 
 typedef struct {
     int data;
-- 
cgit v1.2.3


From c5dec1c3034f1ae3503efbf641ff3b0273b64797 Mon Sep 17 00:00:00 2001
From: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Date: Fri, 11 Apr 2008 12:56:49 +0200
Subject: block: convert bio_copy_user to bio_copy_user_iov

This patch enables bio_copy_user to take struct sg_iovec (renamed
bio_copy_user_iov). bio_copy_user uses bio_copy_user_iov internally as
bio_map_user uses bio_map_user_iov.

The major changes are:

- adds sg_iovec array to struct bio_map_data

- adds __bio_copy_iov that copy data between bio and
sg_iovec. bio_copy_user_iov and bio_uncopy_user use it.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Tejun Heo <htejun@gmail.com>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 fs/bio.c            | 158 ++++++++++++++++++++++++++++++++++++++--------------
 include/linux/bio.h |   2 +
 2 files changed, 119 insertions(+), 41 deletions(-)

(limited to 'include/linux')

diff --git a/fs/bio.c b/fs/bio.c
index 553b5b7960ad..6e0b6f66df03 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -444,22 +444,27 @@ int bio_add_page(struct bio *bio, struct page *page, unsigned int len,
 
 struct bio_map_data {
 	struct bio_vec *iovecs;
-	void __user *userptr;
+	int nr_sgvecs;
+	struct sg_iovec *sgvecs;
 };
 
-static void bio_set_map_data(struct bio_map_data *bmd, struct bio *bio)
+static void bio_set_map_data(struct bio_map_data *bmd, struct bio *bio,
+			     struct sg_iovec *iov, int iov_count)
 {
 	memcpy(bmd->iovecs, bio->bi_io_vec, sizeof(struct bio_vec) * bio->bi_vcnt);
+	memcpy(bmd->sgvecs, iov, sizeof(struct sg_iovec) * iov_count);
+	bmd->nr_sgvecs = iov_count;
 	bio->bi_private = bmd;
 }
 
 static void bio_free_map_data(struct bio_map_data *bmd)
 {
 	kfree(bmd->iovecs);
+	kfree(bmd->sgvecs);
 	kfree(bmd);
 }
 
-static struct bio_map_data *bio_alloc_map_data(int nr_segs)
+static struct bio_map_data *bio_alloc_map_data(int nr_segs, int iov_count)
 {
 	struct bio_map_data *bmd = kmalloc(sizeof(*bmd), GFP_KERNEL);
 
@@ -467,13 +472,71 @@ static struct bio_map_data *bio_alloc_map_data(int nr_segs)
 		return NULL;
 
 	bmd->iovecs = kmalloc(sizeof(struct bio_vec) * nr_segs, GFP_KERNEL);
-	if (bmd->iovecs)
+	if (!bmd->iovecs) {
+		kfree(bmd);
+		return NULL;
+	}
+
+	bmd->sgvecs = kmalloc(sizeof(struct sg_iovec) * iov_count, GFP_KERNEL);
+	if (bmd->sgvecs)
 		return bmd;
 
+	kfree(bmd->iovecs);
 	kfree(bmd);
 	return NULL;
 }
 
+static int __bio_copy_iov(struct bio *bio, struct sg_iovec *iov, int iov_count,
+			  int uncopy)
+{
+	int ret = 0, i;
+	struct bio_vec *bvec;
+	int iov_idx = 0;
+	unsigned int iov_off = 0;
+	int read = bio_data_dir(bio) == READ;
+
+	__bio_for_each_segment(bvec, bio, i, 0) {
+		char *bv_addr = page_address(bvec->bv_page);
+		unsigned int bv_len = bvec->bv_len;
+
+		while (bv_len && iov_idx < iov_count) {
+			unsigned int bytes;
+			char *iov_addr;
+
+			bytes = min_t(unsigned int,
+				      iov[iov_idx].iov_len - iov_off, bv_len);
+			iov_addr = iov[iov_idx].iov_base + iov_off;
+
+			if (!ret) {
+				if (!read && !uncopy)
+					ret = copy_from_user(bv_addr, iov_addr,
+							     bytes);
+				if (read && uncopy)
+					ret = copy_to_user(iov_addr, bv_addr,
+							   bytes);
+
+				if (ret)
+					ret = -EFAULT;
+			}
+
+			bv_len -= bytes;
+			bv_addr += bytes;
+			iov_addr += bytes;
+			iov_off += bytes;
+
+			if (iov[iov_idx].iov_len == iov_off) {
+				iov_idx++;
+				iov_off = 0;
+			}
+		}
+
+		if (uncopy)
+			__free_page(bvec->bv_page);
+	}
+
+	return ret;
+}
+
 /**
  *	bio_uncopy_user	-	finish previously mapped bio
  *	@bio: bio being terminated
@@ -484,55 +547,56 @@ static struct bio_map_data *bio_alloc_map_data(int nr_segs)
 int bio_uncopy_user(struct bio *bio)
 {
 	struct bio_map_data *bmd = bio->bi_private;
-	const int read = bio_data_dir(bio) == READ;
-	struct bio_vec *bvec;
-	int i, ret = 0;
+	int ret;
 
-	__bio_for_each_segment(bvec, bio, i, 0) {
-		char *addr = page_address(bvec->bv_page);
-		unsigned int len = bmd->iovecs[i].bv_len;
+	ret = __bio_copy_iov(bio, bmd->sgvecs, bmd->nr_sgvecs, 1);
 
-		if (read && !ret && copy_to_user(bmd->userptr, addr, len))
-			ret = -EFAULT;
-
-		__free_page(bvec->bv_page);
-		bmd->userptr += len;
-	}
 	bio_free_map_data(bmd);
 	bio_put(bio);
 	return ret;
 }
 
 /**
- *	bio_copy_user	-	copy user data to bio
+ *	bio_copy_user_iov	-	copy user data to bio
  *	@q: destination block queue
- *	@uaddr: start of user address
- *	@len: length in bytes
+ *	@iov:	the iovec.
+ *	@iov_count: number of elements in the iovec
  *	@write_to_vm: bool indicating writing to pages or not
  *
  *	Prepares and returns a bio for indirect user io, bouncing data
  *	to/from kernel pages as necessary. Must be paired with
  *	call bio_uncopy_user() on io completion.
  */
-struct bio *bio_copy_user(struct request_queue *q, unsigned long uaddr,
-			  unsigned int len, int write_to_vm)
+struct bio *bio_copy_user_iov(struct request_queue *q, struct sg_iovec *iov,
+			      int iov_count, int write_to_vm)
 {
-	unsigned long end = (uaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT;
-	unsigned long start = uaddr >> PAGE_SHIFT;
 	struct bio_map_data *bmd;
 	struct bio_vec *bvec;
 	struct page *page;
 	struct bio *bio;
 	int i, ret;
+	int nr_pages = 0;
+	unsigned int len = 0;
 
-	bmd = bio_alloc_map_data(end - start);
+	for (i = 0; i < iov_count; i++) {
+		unsigned long uaddr;
+		unsigned long end;
+		unsigned long start;
+
+		uaddr = (unsigned long)iov[i].iov_base;
+		end = (uaddr + iov[i].iov_len + PAGE_SIZE - 1) >> PAGE_SHIFT;
+		start = uaddr >> PAGE_SHIFT;
+
+		nr_pages += end - start;
+		len += iov[i].iov_len;
+	}
+
+	bmd = bio_alloc_map_data(nr_pages, iov_count);
 	if (!bmd)
 		return ERR_PTR(-ENOMEM);
 
-	bmd->userptr = (void __user *) uaddr;
-
 	ret = -ENOMEM;
-	bio = bio_alloc(GFP_KERNEL, end - start);
+	bio = bio_alloc(GFP_KERNEL, nr_pages);
 	if (!bio)
 		goto out_bmd;
 
@@ -564,22 +628,12 @@ struct bio *bio_copy_user(struct request_queue *q, unsigned long uaddr,
 	 * success
 	 */
 	if (!write_to_vm) {
-		char __user *p = (char __user *) uaddr;
-
-		/*
-		 * for a write, copy in data to kernel pages
-		 */
-		ret = -EFAULT;
-		bio_for_each_segment(bvec, bio, i) {
-			char *addr = page_address(bvec->bv_page);
-
-			if (copy_from_user(addr, p, bvec->bv_len))
-				goto cleanup;
-			p += bvec->bv_len;
-		}
+		ret = __bio_copy_iov(bio, iov, iov_count, 0);
+		if (ret)
+			goto cleanup;
 	}
 
-	bio_set_map_data(bmd, bio);
+	bio_set_map_data(bmd, bio, iov, iov_count);
 	return bio;
 cleanup:
 	bio_for_each_segment(bvec, bio, i)
@@ -591,6 +645,28 @@ out_bmd:
 	return ERR_PTR(ret);
 }
 
+/**
+ *	bio_copy_user	-	copy user data to bio
+ *	@q: destination block queue
+ *	@uaddr: start of user address
+ *	@len: length in bytes
+ *	@write_to_vm: bool indicating writing to pages or not
+ *
+ *	Prepares and returns a bio for indirect user io, bouncing data
+ *	to/from kernel pages as necessary. Must be paired with
+ *	call bio_uncopy_user() on io completion.
+ */
+struct bio *bio_copy_user(struct request_queue *q, unsigned long uaddr,
+			  unsigned int len, int write_to_vm)
+{
+	struct sg_iovec iov;
+
+	iov.iov_base = (void __user *)uaddr;
+	iov.iov_len = len;
+
+	return bio_copy_user_iov(q, &iov, 1, write_to_vm);
+}
+
 static struct bio *__bio_map_user_iov(struct request_queue *q,
 				      struct block_device *bdev,
 				      struct sg_iovec *iov, int iov_count,
diff --git a/include/linux/bio.h b/include/linux/bio.h
index 4c59bdccd3ee..d259690863fb 100644
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -327,6 +327,8 @@ extern struct bio *bio_map_kern(struct request_queue *, void *, unsigned int,
 extern void bio_set_pages_dirty(struct bio *bio);
 extern void bio_check_pages_dirty(struct bio *bio);
 extern struct bio *bio_copy_user(struct request_queue *, unsigned long, unsigned int, int);
+extern struct bio *bio_copy_user_iov(struct request_queue *, struct sg_iovec *,
+				     int, int);
 extern int bio_uncopy_user(struct bio *);
 void zero_fill_bio(struct bio *bio);
 
-- 
cgit v1.2.3


From f18573abcc57844a7c3c12699d40eead8728cd8a Mon Sep 17 00:00:00 2001
From: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Date: Fri, 11 Apr 2008 12:56:52 +0200
Subject: block: move the padding adjustment to blk_rq_map_sg

blk_rq_map_user adjusts bi_size of the last bio. It breaks the rule
that req->data_len (the true data length) is equal to sum(bio). It
broke the scsi command completion code.

commit e97a294ef6938512b655b1abf17656cf2b26f709 was introduced to fix
the above issue. However, the partial completion code doesn't work
with it. The commit is also a layer violation (scsi mid-layer should
not know about the block layer's padding).

This patch moves the padding adjustment to blk_rq_map_sg (suggested by
James). The padding works like the drain buffer. This patch breaks the
rule that req->data_len is equal to sum(sg), however, the drain buffer
already broke it. So this patch just restores the rule that
req->data_len is equal to sub(bio) without breaking anything new.

Now when a low level driver needs padding, blk_rq_map_user and
blk_rq_map_user_iov guarantee there's enough room for padding.
blk_rq_map_sg can safely extend the last entry of a scatter list.

blk_rq_map_sg must extend the last entry of a scatter list only for a
request that got through bio_copy_user_iov. This patches introduces
new REQ_COPY_USER flag.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Tejun Heo <htejun@gmail.com>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 block/blk-map.c        | 24 +++++-------------------
 block/blk-merge.c      |  9 +++++++++
 drivers/scsi/scsi.c    |  2 +-
 include/linux/blkdev.h |  2 ++
 4 files changed, 17 insertions(+), 20 deletions(-)

(limited to 'include/linux')

diff --git a/block/blk-map.c b/block/blk-map.c
index ab43533ba641..3c942bd6422a 100644
--- a/block/blk-map.c
+++ b/block/blk-map.c
@@ -141,25 +141,8 @@ int blk_rq_map_user(struct request_queue *q, struct request *rq,
 		ubuf += ret;
 	}
 
-	/*
-	 * __blk_rq_map_user() copies the buffers if starting address
-	 * or length isn't aligned to dma_pad_mask.  As the copied
-	 * buffer is always page aligned, we know that there's enough
-	 * room for padding.  Extend the last bio and update
-	 * rq->data_len accordingly.
-	 *
-	 * On unmap, bio_uncopy_user() will use unmodified
-	 * bio_map_data pointed to by bio->bi_private.
-	 */
-	if (len & q->dma_pad_mask) {
-		unsigned int pad_len = (q->dma_pad_mask & ~len) + 1;
-		struct bio *tail = rq->biotail;
-
-		tail->bi_io_vec[tail->bi_vcnt - 1].bv_len += pad_len;
-		tail->bi_size += pad_len;
-
-		rq->extra_len += pad_len;
-	}
+	if (!bio_flagged(bio, BIO_USER_MAPPED))
+		rq->cmd_flags |= REQ_COPY_USER;
 
 	rq->buffer = rq->data = NULL;
 	return 0;
@@ -224,6 +207,9 @@ int blk_rq_map_user_iov(struct request_queue *q, struct request *rq,
 		return -EINVAL;
 	}
 
+	if (!bio_flagged(bio, BIO_USER_MAPPED))
+		rq->cmd_flags |= REQ_COPY_USER;
+
 	bio_get(bio);
 	blk_rq_bio_prep(q, rq, bio);
 	rq->buffer = rq->data = NULL;
diff --git a/block/blk-merge.c b/block/blk-merge.c
index 0f58616bcd7f..b5c5c4a9e3f0 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -220,6 +220,15 @@ new_segment:
 		bvprv = bvec;
 	} /* segments in rq */
 
+
+	if (unlikely(rq->cmd_flags & REQ_COPY_USER) &&
+	    (rq->data_len & q->dma_pad_mask)) {
+		unsigned int pad_len = (q->dma_pad_mask & ~rq->data_len) + 1;
+
+		sg->length += pad_len;
+		rq->extra_len += pad_len;
+	}
+
 	if (q->dma_drain_size && q->dma_drain_needed(rq)) {
 		if (rq->cmd_flags & REQ_RW)
 			memset(q->dma_drain_buffer, 0, q->dma_drain_size);
diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c
index f6980bd9d8f9..12d69d7c8577 100644
--- a/drivers/scsi/scsi.c
+++ b/drivers/scsi/scsi.c
@@ -852,7 +852,7 @@ void scsi_finish_command(struct scsi_cmnd *cmd)
 				"Notifying upper driver of completion "
 				"(result %x)\n", cmd->result));
 
-	good_bytes = scsi_bufflen(cmd) + cmd->request->extra_len;
+	good_bytes = scsi_bufflen(cmd);
         if (cmd->request->cmd_type != REQ_TYPE_BLOCK_PC) {
 		drv = scsi_cmd_to_driver(cmd);
 		if (drv->done)
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 6f79d40dd3c0..b3a58adc4352 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -112,6 +112,7 @@ enum rq_flag_bits {
 	__REQ_RW_SYNC,		/* request is sync (O_DIRECT) */
 	__REQ_ALLOCED,		/* request came from our alloc pool */
 	__REQ_RW_META,		/* metadata io request */
+	__REQ_COPY_USER,	/* contains copies of user pages */
 	__REQ_NR_BITS,		/* stops here */
 };
 
@@ -133,6 +134,7 @@ enum rq_flag_bits {
 #define REQ_RW_SYNC	(1 << __REQ_RW_SYNC)
 #define REQ_ALLOCED	(1 << __REQ_ALLOCED)
 #define REQ_RW_META	(1 << __REQ_RW_META)
+#define REQ_COPY_USER	(1 << __REQ_COPY_USER)
 
 #define BLK_MAX_CDB	16
 
-- 
cgit v1.2.3


From 2472892a3ce17b177cc0d8099a6391949c75abf2 Mon Sep 17 00:00:00 2001
From: Andi Kleen <andi@firstfloor.org>
Date: Mon, 21 Apr 2008 09:51:05 +0200
Subject: block: fix memory hotplug and bouncing in block layer

Only noticed this while hacking something else, no test case.

blk_max_low_pfn is initialized once at bootup by the block layer from
max_low_pfn.  But max_low_pfn is not necessarily constant over the runtime of
the system when you consider memory hotplug.  What could happen if that
someone adds memory later the block layer wouldn't get updated and then start
bouncing memory unnecessarily.

Also on 64bit blk_max_low_pfn actually isn't needed because it just disables
bouncing essentially and there is no highmem.  And nobody can pass pfns >
max_low_pfn to the block layer, because those wouldn't have a struct page and
I suspect block layer wouldn't be very happy without that.

So set BLK_BOUNCE_HIGH to infinity (-1ULL) on 64bit.  That avoids the problem
of having to update it on memory hotadd.

On 32bit I kept the same behaviour because at least on i386
memory hotadd only adds HIGHMEM, never lowmem.

BLK_BOUNCE_ANY is always set to infinity on both 32 and 64bit.

Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Jens Axboe <jens.axboe@oracle.com>
Acked-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 include/linux/blkdev.h | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index b3a58adc4352..c5065e3d2ca9 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -535,8 +535,13 @@ extern unsigned long blk_max_low_pfn, blk_max_pfn;
  * BLK_BOUNCE_ANY	: don't bounce anything
  * BLK_BOUNCE_ISA	: bounce pages above ISA DMA boundary
  */
+
+#if BITS_PER_LONG == 32
 #define BLK_BOUNCE_HIGH		((u64)blk_max_low_pfn << PAGE_SHIFT)
-#define BLK_BOUNCE_ANY		((u64)blk_max_pfn << PAGE_SHIFT)
+#else
+#define BLK_BOUNCE_HIGH		-1ULL
+#endif
+#define BLK_BOUNCE_ANY		(-1ULL)
 #define BLK_BOUNCE_ISA		(ISA_DMA_THRESHOLD)
 
 /*
-- 
cgit v1.2.3


From cc216c5d429892872f70f76975e243aef7ad9db1 Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Sun, 20 Apr 2008 21:59:13 -0700
Subject: Fix RCU list iterator use of 'rcu_dereference()'

The RCU iterators used 'rcu_dereference()' on an already-fetched RCU
pointer value, which defeats the whole point of the exercise.

When we dereference a pointer protected by RCU, we need to make sure
that we only fetch the value _once_, because if the compiler ends up
re-loading it due to register pressure, the newly reloaded value could
be different from the previously fetched one, and you get inconsistent
results.

Cleaned-up, fixed, and the pointless list_for_each_safe_rcu #define
deleted by Paul Kenney.

Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/list.h | 48 +++++++++++++++---------------------------------
 1 file changed, 15 insertions(+), 33 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/list.h b/include/linux/list.h
index 75ce2cb4ff6e..dac16f99c701 100644
--- a/include/linux/list.h
+++ b/include/linux/list.h
@@ -631,31 +631,14 @@ static inline void list_splice_init_rcu(struct list_head *list,
  * as long as the traversal is guarded by rcu_read_lock().
  */
 #define list_for_each_rcu(pos, head) \
-	for (pos = (head)->next; \
-		prefetch(rcu_dereference(pos)->next), pos != (head); \
-        	pos = pos->next)
+	for (pos = rcu_dereference((head)->next); \
+		prefetch(pos->next), pos != (head); \
+		pos = rcu_dereference(pos->next))
 
 #define __list_for_each_rcu(pos, head) \
-	for (pos = (head)->next; \
-		rcu_dereference(pos) != (head); \
-        	pos = pos->next)
-
-/**
- * list_for_each_safe_rcu
- * @pos:	the &struct list_head to use as a loop cursor.
- * @n:		another &struct list_head to use as temporary storage
- * @head:	the head for your list.
- *
- * Iterate over an rcu-protected list, safe against removal of list entry.
- *
- * This list-traversal primitive may safely run concurrently with
- * the _rcu list-mutation primitives such as list_add_rcu()
- * as long as the traversal is guarded by rcu_read_lock().
- */
-#define list_for_each_safe_rcu(pos, n, head) \
-	for (pos = (head)->next; \
-		n = rcu_dereference(pos)->next, pos != (head); \
-		pos = n)
+	for (pos = rcu_dereference((head)->next); \
+		pos != (head); \
+		pos = rcu_dereference(pos->next))
 
 /**
  * list_for_each_entry_rcu	-	iterate over rcu list of given type
@@ -668,10 +651,9 @@ static inline void list_splice_init_rcu(struct list_head *list,
  * as long as the traversal is guarded by rcu_read_lock().
  */
 #define list_for_each_entry_rcu(pos, head, member) \
-	for (pos = list_entry((head)->next, typeof(*pos), member); \
-		prefetch(rcu_dereference(pos)->member.next), \
-			&pos->member != (head); \
-		pos = list_entry(pos->member.next, typeof(*pos), member))
+	for (pos = list_entry(rcu_dereference((head)->next), typeof(*pos), member); \
+		prefetch(pos->member.next), &pos->member != (head); \
+		pos = list_entry(rcu_dereference(pos->member.next), typeof(*pos), member))
 
 
 /**
@@ -686,9 +668,9 @@ static inline void list_splice_init_rcu(struct list_head *list,
  * as long as the traversal is guarded by rcu_read_lock().
  */
 #define list_for_each_continue_rcu(pos, head) \
-	for ((pos) = (pos)->next; \
-		prefetch(rcu_dereference((pos))->next), (pos) != (head); \
-        	(pos) = (pos)->next)
+	for ((pos) = rcu_dereference((pos)->next); \
+		prefetch((pos)->next), (pos) != (head); \
+		(pos) = rcu_dereference((pos)->next))
 
 /*
  * Double linked lists with a single pointer list head.
@@ -986,10 +968,10 @@ static inline void hlist_add_after_rcu(struct hlist_node *prev,
  * as long as the traversal is guarded by rcu_read_lock().
  */
 #define hlist_for_each_entry_rcu(tpos, pos, head, member)		 \
-	for (pos = (head)->first;					 \
-	     rcu_dereference(pos) && ({ prefetch(pos->next); 1;}) &&	 \
+	for (pos = rcu_dereference((head)->first);			 \
+	        pos && ({ prefetch(pos->next); 1;}) &&			 \
 		({ tpos = hlist_entry(pos, typeof(*tpos), member); 1;}); \
-	     pos = pos->next)
+	     pos = rcu_dereference(pos->next))
 
 #else
 #warning "don't include kernel headers in userspace"
-- 
cgit v1.2.3


From 218ff137bc67252694420563d23d051ab9227f17 Mon Sep 17 00:00:00 2001
From: Johannes Weiner <hannes@saeurebad.de>
Date: Mon, 21 Apr 2008 22:35:29 +0000
Subject: Remove unused MAX_NODES_SHIFT

MAX_NODES_SHIFT is not referenced anywhere in the tree, so dump it.

Signed-off-by: Johannes Weiner <hannes@saeurebad.de>
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
---
 include/linux/mmzone.h | 1 -
 1 file changed, 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index 8d8d1977736e..9f274a687c7e 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -699,7 +699,6 @@ extern char numa_zonelist_order[];
 extern struct pglist_data contig_page_data;
 #define NODE_DATA(nid)		(&contig_page_data)
 #define NODE_MEM_MAP(nid)	mem_map
-#define MAX_NODES_SHIFT		1
 
 #else /* CONFIG_NEED_MULTIPLE_NODES */
 
-- 
cgit v1.2.3


From 8a5703f846e2363fc466aff3f53608340a1ae33f Mon Sep 17 00:00:00 2001
From: Sebastian Siewior <bigeasy@tglx.de>
Date: Mon, 21 Apr 2008 22:38:45 +0000
Subject: DMA engine: typo fixes

Spelling fixes for dmaengine.[ch]

Signed-off-by: Sebastian Siewior <bigeasy@linutronix.de>
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
---
 drivers/dma/dmaengine.c   | 6 +++---
 include/linux/dmaengine.h | 2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index d6dc70fd7527..97b329e76798 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -42,9 +42,9 @@
  *
  * Each device has a kref, which is initialized to 1 when the device is
  * registered. A kref_get is done for each device registered.  When the
- * device is released, the coresponding kref_put is done in the release
+ * device is released, the corresponding kref_put is done in the release
  * method. Every time one of the device's channels is allocated to a client,
- * a kref_get occurs.  When the channel is freed, the coresponding kref_put
+ * a kref_get occurs.  When the channel is freed, the corresponding kref_put
  * happens. The device's release function does a completion, so
  * unregister_device does a remove event, device_unregister, a kref_put
  * for the first reference, then waits on the completion for all other
@@ -53,7 +53,7 @@
  * Each channel has an open-coded implementation of Rusty Russell's "bigref,"
  * with a kref and a per_cpu local_t.  A dma_chan_get is called when a client
  * signals that it wants to use a channel, and dma_chan_put is called when
- * a channel is removed or a client using it is unregesitered.  A client can
+ * a channel is removed or a client using it is unregistered.  A client can
  * take extra references per outstanding transaction, as is the case with
  * the NET DMA client.  The release function does a kref_put on the device.
  *	-ChrisL, DanW
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index b4d84ed6187d..d08a5c5eb928 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -404,7 +404,7 @@ static inline enum dma_status dma_async_is_tx_complete(struct dma_chan *chan,
  * @last_used: last cookie value handed out
  *
  * dma_async_is_complete() is used in dma_async_memcpy_complete()
- * the test logic is seperated for lightweight testing of multiple cookies
+ * the test logic is separated for lightweight testing of multiple cookies
  */
 static inline enum dma_status dma_async_is_complete(dma_cookie_t cookie,
 			dma_cookie_t last_complete, dma_cookie_t last_used)
-- 
cgit v1.2.3


From 553a56726be86c09cfa53c84da1ea0e2043e364e Mon Sep 17 00:00:00 2001
From: Randy Dunlap <randy.dunlap@oracle.com>
Date: Sun, 20 Apr 2008 10:51:01 -0700
Subject: skbuff: fix missing kernel-doc notation

Add kernel-doc notation for ndisc_nodetype:

Warning(linux-2.6.25-git2//include/linux/skbuff.h:340): No description found for parameter 'ndisc_nodetype'

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 include/linux/skbuff.h | 1 +
 1 file changed, 1 insertion(+)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 11fd9f2c4093..299ec4b31412 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -242,6 +242,7 @@ typedef unsigned char *sk_buff_data_t;
  *	@queue_mapping: Queue mapping for multiqueue devices
  *	@tc_index: Traffic control index
  *	@tc_verd: traffic control verdict
+ *	@ndisc_nodetype: router type (from link layer)
  *	@dma_cookie: a cookie to one of several possible DMA operations
  *		done by skb DMA functions
  *	@secmark: security marking
-- 
cgit v1.2.3


From 37679011c5a674eb80bff5c2b9b067bf16011d46 Mon Sep 17 00:00:00 2001
From: "Robert P. J. Day" <rpjday@crashcourse.ca>
Date: Mon, 21 Apr 2008 22:56:14 +0000
Subject: Generate a slightly more informative error msg for bad HZ

Generate a slightly more informative error msg for bad HZ
 in include/linux/jiffies.h

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
---
 include/linux/jiffies.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/jiffies.h b/include/linux/jiffies.h
index e0b5b684d83f..8f4aa7280ca9 100644
--- a/include/linux/jiffies.h
+++ b/include/linux/jiffies.h
@@ -36,7 +36,7 @@
 #elif HZ >= 6144 && HZ < 12288
 # define SHIFT_HZ	13
 #else
-# error You lose.
+# error Invalid value of HZ.
 #endif
 
 /* LATCH is used in the interval timer and ftape setup. */
-- 
cgit v1.2.3