Merge branch 'master' into x86/memblock

Conflicts & resolutions:

* arch/x86/xen/setup.c

	dc91c728fd "xen: allow extra memory to be in multiple regions"
	24aa07882b "memblock, x86: Replace memblock_x86_reserve/free..."

	conflicted on xen_add_extra_mem() updates.  The resolution is
	trivial as the latter just want to replace
	memblock_x86_reserve_range() with memblock_reserve().

* drivers/pci/intel-iommu.c

	166e9278a3f "x86/ia64: intel-iommu: move to drivers/iommu/"
	5dfe8660a3d "bootmem: Replace work_with_active_regions() with..."

	conflicted as the former moved the file under drivers/iommu/.
	Resolved by applying the chnages from the latter on the moved
	file.

* mm/Kconfig

	6661672053a "memblock: add NO_BOOTMEM config symbol"
	c378ddd53f9 "memblock, x86: Make ARCH_DISCARD_MEMBLOCK a config option"

	conflicted trivially.  Both added config options.  Just
	letting both add their own options resolves the conflict.

* mm/memblock.c

	d1f0ece6cdc "mm/memblock.c: small function definition fixes"
	ed7b56a799c "memblock: Remove memblock_memory_can_coalesce()"

	confliected.  The former updates function removed by the
	latter.  Resolution is trivial.

Signed-off-by: Tejun Heo <tj@kernel.org>

13 files changed, 985 insertions, 67 deletions

diff --git a/arch/arm/boot/compressed/.gitignore b/arch/arm/boot/compressed/.gitignore
index c6028967d336..e0936a148516 100644
--- a/arch/arm/boot/compressed/.gitignore
+++ b/arch/arm/boot/compressed/.gitignore
@@ -5,3 +5,12 @@ piggy.lzo
 piggy.lzma
 vmlinux
 vmlinux.lds
+
+# borrowed libfdt files
+fdt.c
+fdt.h
+fdt_ro.c
+fdt_rw.c
+fdt_wip.c
+libfdt.h
+libfdt_internal.h
diff --git a/arch/arm/boot/compressed/Makefile b/arch/arm/boot/compressed/Makefile
index 23aad0722303..21f56ff32797 100644
--- a/arch/arm/boot/compressed/Makefile
+++ b/arch/arm/boot/compressed/Makefile
@@ -6,13 +6,19 @@
 
 OBJS		=
 
-# Ensure that mmcif loader code appears early in the image
+# Ensure that MMCIF loader code appears early in the image
 # to minimise that number of bocks that have to be read in
 # order to load it.
 ifeq ($(CONFIG_ZBOOT_ROM_MMCIF),y)
-ifeq ($(CONFIG_ARCH_SH7372),y)
 OBJS		+= mmcif-sh7372.o
 endif
+
+# Ensure that SDHI loader code appears early in the image
+# to minimise that number of bocks that have to be read in
+# order to load it.
+ifeq ($(CONFIG_ZBOOT_ROM_SH_MOBILE_SDHI),y)
+OBJS		+= sdhi-shmobile.o
+OBJS		+= sdhi-sh7372.o
 endif
 
 AFLAGS_head.o += -DTEXT_OFFSET=$(TEXT_OFFSET)
@@ -20,6 +26,10 @@ HEAD	= head.o
 OBJS	+= misc.o decompress.o
 FONTC	= $(srctree)/drivers/video/console/font_acorn_8x8.c
 
+# string library code (-Os is enforced to keep it much smaller)
+OBJS		+= string.o
+CFLAGS_string.o	:= -Os
+
 #
 # Architecture dependencies
 #
@@ -83,21 +93,41 @@ suffix_$(CONFIG_KERNEL_GZIP) = gzip
 suffix_$(CONFIG_KERNEL_LZO)  = lzo
 suffix_$(CONFIG_KERNEL_LZMA) = lzma
 
+# Borrowed libfdt files for the ATAG compatibility mode
+
+libfdt		:= fdt_rw.c fdt_ro.c fdt_wip.c fdt.c
+libfdt_hdrs	:= fdt.h libfdt.h libfdt_internal.h
+
+libfdt_objs	:= $(addsuffix .o, $(basename $(libfdt)))
+
+$(addprefix $(obj)/,$(libfdt) $(libfdt_hdrs)): $(obj)/%: $(srctree)/scripts/dtc/libfdt/%
+	$(call cmd,shipped)
+
+$(addprefix $(obj)/,$(libfdt_objs) atags_to_fdt.o): \
+	$(addprefix $(obj)/,$(libfdt_hdrs))
+
+ifeq ($(CONFIG_ARM_ATAG_DTB_COMPAT),y)
+OBJS	+= $(libfdt_objs) atags_to_fdt.o
+endif
+
 targets       := vmlinux vmlinux.lds \
 		 piggy.$(suffix_y) piggy.$(suffix_y).o \
-		 font.o font.c head.o misc.o $(OBJS)
+		 lib1funcs.o lib1funcs.S font.o font.c head.o misc.o $(OBJS)
 
 # Make sure files are removed during clean
-extra-y       += piggy.gzip piggy.lzo piggy.lzma lib1funcs.S
+extra-y       += piggy.gzip piggy.lzo piggy.lzma lib1funcs.S $(libfdt) $(libfdt_hdrs)
 
 ifeq ($(CONFIG_FUNCTION_TRACER),y)
 ORIG_CFLAGS := $(KBUILD_CFLAGS)
 KBUILD_CFLAGS = $(subst -pg, , $(ORIG_CFLAGS))
 endif
 
-ccflags-y := -fpic -fno-builtin
+ccflags-y := -fpic -fno-builtin -I$(obj)
 asflags-y := -Wa,-march=all
 
+# Supply kernel BSS size to the decompressor via a linker symbol.
+KBSS_SZ = $(shell size $(obj)/../../../../vmlinux | awk 'END{print $$3}')
+LDFLAGS_vmlinux = --defsym _kernel_bss_size=$(KBSS_SZ)
 # Supply ZRELADDR to the decompressor via a linker symbol.
 ifneq ($(CONFIG_AUTO_ZRELADDR),y)
 LDFLAGS_vmlinux += --defsym zreladdr=$(ZRELADDR)
@@ -117,7 +147,7 @@ LDFLAGS_vmlinux += -T
 # For __aeabi_uidivmod
 lib1funcs = $(obj)/lib1funcs.o
 
-$(obj)/lib1funcs.S: $(srctree)/arch/$(SRCARCH)/lib/lib1funcs.S FORCE
+$(obj)/lib1funcs.S: $(srctree)/arch/$(SRCARCH)/lib/lib1funcs.S
 	$(call cmd,shipped)
 
 # We need to prevent any GOTOFF relocs being used with references
@@ -133,8 +163,16 @@ bad_syms=$$($(CROSS_COMPILE)nm $@ | sed -n 's/^.\{8\} [bc] \(.*\)/\1/p') && \
   ( echo "following symbols must have non local/private scope:" >&2; \
     echo "$$bad_syms" >&2; rm -f $@; false )
 
+check_for_multiple_zreladdr = \
+if [ $(words $(ZRELADDR)) -gt 1 -a "$(CONFIG_AUTO_ZRELADDR)" = "" ]; then \
+	echo 'multiple zreladdrs: $(ZRELADDR)'; \
+	echo 'This needs CONFIG_AUTO_ZRELADDR to be set'; \
+	false; \
+fi
+
 $(obj)/vmlinux: $(obj)/vmlinux.lds $(obj)/$(HEAD) $(obj)/piggy.$(suffix_y).o \
 	 	$(addprefix $(obj)/, $(OBJS)) $(lib1funcs) FORCE
+	@$(check_for_multiple_zreladdr)
 	$(call if_changed,ld)
 	@$(check_for_bad_syms)
 
diff --git a/arch/arm/boot/compressed/atags_to_fdt.c b/arch/arm/boot/compressed/atags_to_fdt.c
new file mode 100644
index 000000000000..6ce11c481178
--- /dev/null
+++ b/arch/arm/boot/compressed/atags_to_fdt.c
@@ -0,0 +1,97 @@
+#include <asm/setup.h>
+#include <libfdt.h>
+
+static int node_offset(void *fdt, const char *node_path)
+{
+	int offset = fdt_path_offset(fdt, node_path);
+	if (offset == -FDT_ERR_NOTFOUND)
+		offset = fdt_add_subnode(fdt, 0, node_path);
+	return offset;
+}
+
+static int setprop(void *fdt, const char *node_path, const char *property,
+		   uint32_t *val_array, int size)
+{
+	int offset = node_offset(fdt, node_path);
+	if (offset < 0)
+		return offset;
+	return fdt_setprop(fdt, offset, property, val_array, size);
+}
+
+static int setprop_string(void *fdt, const char *node_path,
+			  const char *property, const char *string)
+{
+	int offset = node_offset(fdt, node_path);
+	if (offset < 0)
+		return offset;
+	return fdt_setprop_string(fdt, offset, property, string);
+}
+
+static int setprop_cell(void *fdt, const char *node_path,
+			const char *property, uint32_t val)
+{
+	int offset = node_offset(fdt, node_path);
+	if (offset < 0)
+		return offset;
+	return fdt_setprop_cell(fdt, offset, property, val);
+}
+
+/*
+ * Convert and fold provided ATAGs into the provided FDT.
+ *
+ * REturn values:
+ *    = 0 -> pretend success
+ *    = 1 -> bad ATAG (may retry with another possible ATAG pointer)
+ *    < 0 -> error from libfdt
+ */
+int atags_to_fdt(void *atag_list, void *fdt, int total_space)
+{
+	struct tag *atag = atag_list;
+	uint32_t mem_reg_property[2 * NR_BANKS];
+	int memcount = 0;
+	int ret;
+
+	/* make sure we've got an aligned pointer */
+	if ((u32)atag_list & 0x3)
+		return 1;
+
+	/* if we get a DTB here we're done already */
+	if (*(u32 *)atag_list == fdt32_to_cpu(FDT_MAGIC))
+	       return 0;
+
+	/* validate the ATAG */
+	if (atag->hdr.tag != ATAG_CORE ||
+	    (atag->hdr.size != tag_size(tag_core) &&
+	     atag->hdr.size != 2))
+		return 1;
+
+	/* let's give it all the room it could need */
+	ret = fdt_open_into(fdt, fdt, total_space);
+	if (ret < 0)
+		return ret;
+
+	for_each_tag(atag, atag_list) {
+		if (atag->hdr.tag == ATAG_CMDLINE) {
+			setprop_string(fdt, "/chosen", "bootargs",
+					atag->u.cmdline.cmdline);
+		} else if (atag->hdr.tag == ATAG_MEM) {
+			if (memcount >= sizeof(mem_reg_property)/4)
+				continue;
+			mem_reg_property[memcount++] = cpu_to_fdt32(atag->u.mem.start);
+			mem_reg_property[memcount++] = cpu_to_fdt32(atag->u.mem.size);
+		} else if (atag->hdr.tag == ATAG_INITRD2) {
+			uint32_t initrd_start, initrd_size;
+			initrd_start = atag->u.initrd.start;
+			initrd_size = atag->u.initrd.size;
+			setprop_cell(fdt, "/chosen", "linux,initrd-start",
+					initrd_start);
+			setprop_cell(fdt, "/chosen", "linux,initrd-end",
+					initrd_start + initrd_size);
+		}
+	}
+
+	if (memcount)
+		setprop(fdt, "/memory", "reg", mem_reg_property, 4*memcount);
+
+	return fdt_pack(fdt);
+}
diff --git a/arch/arm/boot/compressed/head-shmobile.S b/arch/arm/boot/compressed/head-shmobile.S
index c943d2e7da9d..fe3719b516fd 100644
--- a/arch/arm/boot/compressed/head-shmobile.S
+++ b/arch/arm/boot/compressed/head-shmobile.S
@@ -25,14 +25,14 @@
 	/* load board-specific initialization code */
 #include <mach/zboot.h>
 
-#ifdef CONFIG_ZBOOT_ROM_MMCIF
-	/* Load image from MMC */
-	adr	sp, __tmp_stack + 128
+#if defined(CONFIG_ZBOOT_ROM_MMCIF) || defined(CONFIG_ZBOOT_ROM_SH_MOBILE_SDHI)
+	/* Load image from MMC/SD */
+	adr	sp, __tmp_stack + 256
 	ldr	r0, __image_start
 	ldr	r1, __image_end
 	subs	r1, r1, r0
 	ldr	r0, __load_base
-	bl	mmcif_loader
+	bl	mmc_loader
 
 	/* Jump to loaded code */
 	ldr	r0, __loaded
@@ -51,9 +51,9 @@ __loaded:
 	.long	__continue
 	.align
 __tmp_stack:
-	.space	128
+	.space	256
 __continue:
-#endif /* CONFIG_ZBOOT_ROM_MMCIF */
+#endif /* CONFIG_ZBOOT_ROM_MMC || CONFIG_ZBOOT_ROM_SH_MOBILE_SDHI */
 
 	b	1f
 __atags:@ tag #1
diff --git a/arch/arm/boot/compressed/head.S b/arch/arm/boot/compressed/head.S
index 940b20178107..c2effc917254 100644
--- a/arch/arm/boot/compressed/head.S
+++ b/arch/arm/boot/compressed/head.S
@@ -216,6 +216,104 @@ restart:	adr	r0, LC0
 		mov	r10, r6
 #endif
 
+		mov	r5, #0			@ init dtb size to 0
+#ifdef CONFIG_ARM_APPENDED_DTB
+/*
+ *   r0  = delta
+ *   r2  = BSS start
+ *   r3  = BSS end
+ *   r4  = final kernel address
+ *   r5  = appended dtb size (still unknown)
+ *   r6  = _edata
+ *   r7  = architecture ID
+ *   r8  = atags/device tree pointer
+ *   r9  = size of decompressed image
+ *   r10 = end of this image, including  bss/stack/malloc space if non XIP
+ *   r11 = GOT start
+ *   r12 = GOT end
+ *   sp  = stack pointer
+ *
+ * if there are device trees (dtb) appended to zImage, advance r10 so that the
+ * dtb data will get relocated along with the kernel if necessary.
+ */
+
+		ldr	lr, [r6, #0]
+#ifndef __ARMEB__
+		ldr	r1, =0xedfe0dd0		@ sig is 0xd00dfeed big endian
+#else
+		ldr	r1, =0xd00dfeed
+#endif
+		cmp	lr, r1
+		bne	dtb_check_done		@ not found
+
+#ifdef CONFIG_ARM_ATAG_DTB_COMPAT
+		/*
+		 * OK... Let's do some funky business here.
+		 * If we do have a DTB appended to zImage, and we do have
+		 * an ATAG list around, we want the later to be translated
+		 * and folded into the former here.  To be on the safe side,
+		 * let's temporarily move  the stack away into the malloc
+		 * area.  No GOT fixup has occurred yet, but none of the
+		 * code we're about to call uses any global variable.
+		*/
+		add	sp, sp, #0x10000
+		stmfd	sp!, {r0-r3, ip, lr}
+		mov	r0, r8
+		mov	r1, r6
+		sub	r2, sp, r6
+		bl	atags_to_fdt
+
+		/*
+		 * If returned value is 1, there is no ATAG at the location
+		 * pointed by r8.  Try the typical 0x100 offset from start
+		 * of RAM and hope for the best.
+		 */
+		cmp	r0, #1
+		sub	r0, r4, #TEXT_OFFSET
+		add	r0, r0, #0x100
+		mov	r1, r6
+		sub	r2, sp, r6
+		blne	atags_to_fdt
+
+		ldmfd	sp!, {r0-r3, ip, lr}
+		sub	sp, sp, #0x10000
+#endif
+
+		mov	r8, r6			@ use the appended device tree
+
+		/*
+		 * Make sure that the DTB doesn't end up in the final
+		 * kernel's .bss area. To do so, we adjust the decompressed
+		 * kernel size to compensate if that .bss size is larger
+		 * than the relocated code.
+		 */
+		ldr	r5, =_kernel_bss_size
+		adr	r1, wont_overwrite
+		sub	r1, r6, r1
+		subs	r1, r5, r1
+		addhi	r9, r9, r1
+
+		/* Get the dtb's size */
+		ldr	r5, [r6, #4]
+#ifndef __ARMEB__
+		/* convert r5 (dtb size) to little endian */
+		eor	r1, r5, r5, ror #16
+		bic	r1, r1, #0x00ff0000
+		mov	r5, r5, ror #8
+		eor	r5, r5, r1, lsr #8
+#endif
+
+		/* preserve 64-bit alignment */
+		add	r5, r5, #7
+		bic	r5, r5, #7
+
+		/* relocate some pointers past the appended dtb */
+		add	r6, r6, r5
+		add	r10, r10, r5
+		add	sp, sp, r5
+dtb_check_done:
+#endif
+
 /*
  * Check to see if we will overwrite ourselves.
  *   r4  = final kernel address
@@ -223,15 +321,14 @@ restart:	adr	r0, LC0
  *   r10 = end of this image, including  bss/stack/malloc space if non XIP
  * We basically want:
  *   r4 - 16k page directory >= r10 -> OK
- *   r4 + image length <= current position (pc) -> OK
+ *   r4 + image length <= address of wont_overwrite -> OK
  */
 		add	r10, r10, #16384
 		cmp	r4, r10
 		bhs	wont_overwrite
 		add	r10, r4, r9
-   ARM(		cmp	r10, pc		)
- THUMB(		mov	lr, pc		)
- THUMB(		cmp	r10, lr		)
+		adr	r9, wont_overwrite
+		cmp	r10, r9
 		bls	wont_overwrite
 
 /*
@@ -285,14 +382,16 @@ wont_overwrite:
  *   r2  = BSS start
  *   r3  = BSS end
  *   r4  = kernel execution address
+ *   r5  = appended dtb size (0 if not present)
  *   r7  = architecture ID
  *   r8  = atags pointer
  *   r11 = GOT start
  *   r12 = GOT end
  *   sp  = stack pointer
  */
-		teq	r0, #0
+		orrs	r1, r0, r5
 		beq	not_relocated
+
 		add	r11, r11, r0
 		add	r12, r12, r0
 
@@ -307,12 +406,21 @@ wont_overwrite:
 
 		/*
 		 * Relocate all entries in the GOT table.
+		 * Bump bss entries to _edata + dtb size
 		 */
 1:		ldr	r1, [r11, #0]		@ relocate entries in the GOT
-		add	r1, r1, r0		@ table.  This fixes up the
-		str	r1, [r11], #4		@ C references.
+		add	r1, r1, r0		@ This fixes up C references
+		cmp	r1, r2			@ if entry >= bss_start &&
+		cmphs	r3, r1			@       bss_end > entry
+		addhi	r1, r1, r5		@    entry += dtb size
+		str	r1, [r11], #4		@ next entry
 		cmp	r11, r12
 		blo	1b
+
+		/* bump our bss pointers too */
+		add	r2, r2, r5
+		add	r3, r3, r5
+
 #else
 
 		/*
@@ -353,7 +461,8 @@ not_relocated:	mov	r0, #0
 		mov	r0, #0			@ must be zero
 		mov	r1, r7			@ restore architecture number
 		mov	r2, r8			@ restore atags pointer
-		mov	pc, r4			@ call kernel
+ ARM(		mov	pc, r4	)		@ call kernel
+ THUMB(		bx	r4	)		@ entry point is always ARM
 
 		.align	2
 		.type	LC0, #object
diff --git a/arch/arm/boot/compressed/libfdt_env.h b/arch/arm/boot/compressed/libfdt_env.h
new file mode 100644
index 000000000000..1f4e71876b00
--- /dev/null
+++ b/arch/arm/boot/compressed/libfdt_env.h
@@ -0,0 +1,15 @@
+#ifndef _ARM_LIBFDT_ENV_H
+#define _ARM_LIBFDT_ENV_H
+
+#include <linux/types.h>
+#include <linux/string.h>
+#include <asm/byteorder.h>
+
+#define fdt16_to_cpu(x)		be16_to_cpu(x)
+#define cpu_to_fdt16(x)		cpu_to_be16(x)
+#define fdt32_to_cpu(x)		be32_to_cpu(x)
+#define cpu_to_fdt32(x)		cpu_to_be32(x)
+#define fdt64_to_cpu(x)		be64_to_cpu(x)
+#define cpu_to_fdt64(x)		cpu_to_be64(x)
+
+#endif
diff --git a/arch/arm/boot/compressed/misc.c b/arch/arm/boot/compressed/misc.c
index 832d37236c59..8e2a8fca5ed2 100644
--- a/arch/arm/boot/compressed/misc.c
+++ b/arch/arm/boot/compressed/misc.c
@@ -18,14 +18,9 @@
 
 unsigned int __machine_arch_type;
 
-#define _LINUX_STRING_H_
-
 #include <linux/compiler.h>	/* for inline */
-#include <linux/types.h>	/* for size_t */
-#include <linux/stddef.h>	/* for NULL */
+#include <linux/types.h>
 #include <linux/linkage.h>
-#include <asm/string.h>
-
 
 static void putstr(const char *ptr);
 extern void error(char *x);
@@ -101,41 +96,6 @@ static void putstr(const char *ptr)
 	flush();
 }
 
-
-void *memcpy(void *__dest, __const void *__src, size_t __n)
-{
-	int i = 0;
-	unsigned char *d = (unsigned char *)__dest, *s = (unsigned char *)__src;
-
-	for (i = __n >> 3; i > 0; i--) {
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-	}
-
-	if (__n & 1 << 2) {
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-	}
-
-	if (__n & 1 << 1) {
-		*d++ = *s++;
-		*d++ = *s++;
-	}
-
-	if (__n & 1)
-		*d++ = *s++;
-
-	return __dest;
-}
-
 /*
  * gzip declarations
  */
diff --git a/arch/arm/boot/compressed/mmcif-sh7372.c b/arch/arm/boot/compressed/mmcif-sh7372.c
index 7453c8337b83..672ae95db5c3 100644
--- a/arch/arm/boot/compressed/mmcif-sh7372.c
+++ b/arch/arm/boot/compressed/mmcif-sh7372.c
@@ -40,7 +40,7 @@
  * to an MMC card
  * # dd if=vrl4.out of=/dev/sdx bs=512 seek=1
  */
-asmlinkage void mmcif_loader(unsigned char *buf, unsigned long len)
+asmlinkage void mmc_loader(unsigned char *buf, unsigned long len)
 {
 	mmc_init_progress();
 	mmc_update_progress(MMC_PROGRESS_ENTER);
@@ -82,7 +82,7 @@ asmlinkage void mmcif_loader(unsigned char *buf, unsigned long len)
 
 
 	/* Disable clock to MMC hardware block */
-	__raw_writel(__raw_readl(SMSTPCR3) & (1 << 12), SMSTPCR3);
+	__raw_writel(__raw_readl(SMSTPCR3) | (1 << 12), SMSTPCR3);
 
 	mmc_update_progress(MMC_PROGRESS_DONE);
 }
diff --git a/arch/arm/boot/compressed/sdhi-sh7372.c b/arch/arm/boot/compressed/sdhi-sh7372.c
new file mode 100644
index 000000000000..d279294f2381
--- /dev/null
+++ b/arch/arm/boot/compressed/sdhi-sh7372.c
@@ -0,0 +1,95 @@
+/*
+ * SuperH Mobile SDHI
+ *
+ * Copyright (C) 2010 Magnus Damm
+ * Copyright (C) 2010 Kuninori Morimoto
+ * Copyright (C) 2010 Simon Horman
+ *
+ * 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.
+ *
+ * Parts inspired by u-boot
+ */
+
+#include <linux/io.h>
+#include <mach/mmc.h>
+#include <linux/mmc/boot.h>
+#include <linux/mmc/tmio.h>
+
+#include "sdhi-shmobile.h"
+
+#define PORT179CR       0xe60520b3
+#define PORT180CR       0xe60520b4
+#define PORT181CR       0xe60520b5
+#define PORT182CR       0xe60520b6
+#define PORT183CR       0xe60520b7
+#define PORT184CR       0xe60520b8
+
+#define SMSTPCR3        0xe615013c
+
+#define CR_INPUT_ENABLE 0x10
+#define CR_FUNCTION1    0x01
+
+#define SDHI1_BASE	(void __iomem *)0xe6860000
+#define SDHI_BASE	SDHI1_BASE
+
+/*  SuperH Mobile SDHI loader
+ *
+ * loads the zImage from an SD card starting from block 0
+ * on physical partition 1
+ *
+ * The image must be start with a vrl4 header and
+ * the zImage must start at offset 512 of the image. That is,
+ * at block 1 (=byte 512) of physical partition 1
+ *
+ * Use the following line to write the vrl4 formated zImage
+ * to an SD card
+ * # dd if=vrl4.out of=/dev/sdx bs=512
+ */
+asmlinkage void mmc_loader(unsigned short *buf, unsigned long len)
+{
+	int high_capacity;
+
+	mmc_init_progress();
+
+	mmc_update_progress(MMC_PROGRESS_ENTER);
+        /* Initialise SDHI1 */
+        /* PORT184CR: GPIO_FN_SDHICMD1 Control */
+        __raw_writeb(CR_FUNCTION1, PORT184CR);
+        /* PORT179CR: GPIO_FN_SDHICLK1 Control */
+        __raw_writeb(CR_INPUT_ENABLE|CR_FUNCTION1, PORT179CR);
+        /* PORT181CR: GPIO_FN_SDHID1_3 Control */
+        __raw_writeb(CR_FUNCTION1, PORT183CR);
+        /* PORT182CR: GPIO_FN_SDHID1_2 Control */
+        __raw_writeb(CR_FUNCTION1, PORT182CR);
+        /* PORT183CR: GPIO_FN_SDHID1_1 Control */
+        __raw_writeb(CR_FUNCTION1, PORT181CR);
+        /* PORT180CR: GPIO_FN_SDHID1_0 Control */
+        __raw_writeb(CR_FUNCTION1, PORT180CR);
+
+        /* Enable clock to SDHI1 hardware block */
+        __raw_writel(__raw_readl(SMSTPCR3) & ~(1 << 13), SMSTPCR3);
+
+	/* setup SDHI hardware */
+	mmc_update_progress(MMC_PROGRESS_INIT);
+	high_capacity = sdhi_boot_init(SDHI_BASE);
+	if (high_capacity < 0)
+		goto err;
+
+	mmc_update_progress(MMC_PROGRESS_LOAD);
+	/* load kernel */
+	if (sdhi_boot_do_read(SDHI_BASE, high_capacity,
+			      0, /* Kernel is at block 1 */
+			      (len + TMIO_BBS - 1) / TMIO_BBS, buf))
+		goto err;
+
+        /* Disable clock to SDHI1 hardware block */
+        __raw_writel(__raw_readl(SMSTPCR3) | (1 << 13), SMSTPCR3);
+
+	mmc_update_progress(MMC_PROGRESS_DONE);
+
+	return;
+err:
+	for(;;);
+}
diff --git a/arch/arm/boot/compressed/sdhi-shmobile.c b/arch/arm/boot/compressed/sdhi-shmobile.c
new file mode 100644
index 000000000000..bd3d46980955
--- /dev/null
+++ b/arch/arm/boot/compressed/sdhi-shmobile.c
@@ -0,0 +1,449 @@
+/*
+ * SuperH Mobile SDHI
+ *
+ * Copyright (C) 2010 Magnus Damm
+ * Copyright (C) 2010 Kuninori Morimoto
+ * Copyright (C) 2010 Simon Horman
+ *
+ * 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.
+ *
+ * Parts inspired by u-boot
+ */
+
+#include <linux/io.h>
+#include <linux/mmc/host.h>
+#include <linux/mmc/core.h>
+#include <linux/mmc/mmc.h>
+#include <linux/mmc/sd.h>
+#include <linux/mmc/tmio.h>
+#include <mach/sdhi.h>
+
+#define OCR_FASTBOOT		(1<<29)
+#define OCR_HCS			(1<<30)
+#define OCR_BUSY		(1<<31)
+
+#define RESP_CMD12		0x00000030
+
+static inline u16 sd_ctrl_read16(void __iomem *base, int addr)
+{
+        return __raw_readw(base + addr);
+}
+
+static inline u32 sd_ctrl_read32(void __iomem *base, int addr)
+{
+	return __raw_readw(base + addr) |
+	       __raw_readw(base + addr + 2) << 16;
+}
+
+static inline void sd_ctrl_write16(void __iomem *base, int addr, u16 val)
+{
+	__raw_writew(val, base + addr);
+}
+
+static inline void sd_ctrl_write32(void __iomem *base, int addr, u32 val)
+{
+	__raw_writew(val, base + addr);
+	__raw_writew(val >> 16, base + addr + 2);
+}
+
+#define ALL_ERROR (TMIO_STAT_CMD_IDX_ERR | TMIO_STAT_CRCFAIL |		\
+		   TMIO_STAT_STOPBIT_ERR | TMIO_STAT_DATATIMEOUT |	\
+		   TMIO_STAT_RXOVERFLOW | TMIO_STAT_TXUNDERRUN |	\
+		   TMIO_STAT_CMDTIMEOUT | TMIO_STAT_ILL_ACCESS |	\
+		   TMIO_STAT_ILL_FUNC)
+
+static int sdhi_intr(void __iomem *base)
+{
+	unsigned long state = sd_ctrl_read32(base, CTL_STATUS);
+
+	if (state & ALL_ERROR) {
+		sd_ctrl_write32(base, CTL_STATUS, ~ALL_ERROR);
+		sd_ctrl_write32(base, CTL_IRQ_MASK,
+				ALL_ERROR |
+				sd_ctrl_read32(base, CTL_IRQ_MASK));
+		return -EINVAL;
+	}
+	if (state & TMIO_STAT_CMDRESPEND) {
+		sd_ctrl_write32(base, CTL_STATUS, ~TMIO_STAT_CMDRESPEND);
+		sd_ctrl_write32(base, CTL_IRQ_MASK,
+				TMIO_STAT_CMDRESPEND |
+				sd_ctrl_read32(base, CTL_IRQ_MASK));
+		return 0;
+	}
+	if (state & TMIO_STAT_RXRDY) {
+		sd_ctrl_write32(base, CTL_STATUS, ~TMIO_STAT_RXRDY);
+		sd_ctrl_write32(base, CTL_IRQ_MASK,
+				TMIO_STAT_RXRDY | TMIO_STAT_TXUNDERRUN |
+				sd_ctrl_read32(base, CTL_IRQ_MASK));
+		return 0;
+	}
+	if (state & TMIO_STAT_DATAEND) {
+		sd_ctrl_write32(base, CTL_STATUS, ~TMIO_STAT_DATAEND);
+		sd_ctrl_write32(base, CTL_IRQ_MASK,
+				TMIO_STAT_DATAEND |
+				sd_ctrl_read32(base, CTL_IRQ_MASK));
+		return 0;
+	}
+
+	return -EAGAIN;
+}
+
+static int sdhi_boot_wait_resp_end(void __iomem *base)
+{
+	int err = -EAGAIN, timeout = 10000000;
+
+	while (timeout--) {
+		err = sdhi_intr(base);
+		if (err != -EAGAIN)
+			break;
+		udelay(1);
+	}
+
+	return err;
+}
+
+/* SDHI_CLK_CTRL */
+#define CLK_MMC_ENABLE                 (1 << 8)
+#define CLK_MMC_INIT                   (1 << 6)        /* clk / 256 */
+
+static void sdhi_boot_mmc_clk_stop(void __iomem *base)
+{
+	sd_ctrl_write16(base, CTL_CLK_AND_WAIT_CTL, 0x0000);
+	msleep(10);
+	sd_ctrl_write16(base, CTL_SD_CARD_CLK_CTL, ~CLK_MMC_ENABLE &
+		sd_ctrl_read16(base, CTL_SD_CARD_CLK_CTL));
+	msleep(10);
+}
+
+static void sdhi_boot_mmc_clk_start(void __iomem *base)
+{
+	sd_ctrl_write16(base, CTL_SD_CARD_CLK_CTL, CLK_MMC_ENABLE |
+		sd_ctrl_read16(base, CTL_SD_CARD_CLK_CTL));
+	msleep(10);
+	sd_ctrl_write16(base, CTL_CLK_AND_WAIT_CTL, CLK_MMC_ENABLE);
+	msleep(10);
+}
+
+static void sdhi_boot_reset(void __iomem *base)
+{
+	sd_ctrl_write16(base, CTL_RESET_SD, 0x0000);
+	msleep(10);
+	sd_ctrl_write16(base, CTL_RESET_SD, 0x0001);
+	msleep(10);
+}
+
+/* Set MMC clock / power.
+ * Note: This controller uses a simple divider scheme therefore it cannot
+ * run a MMC card at full speed (20MHz). The max clock is 24MHz on SD, but as
+ * MMC wont run that fast, it has to be clocked at 12MHz which is the next
+ * slowest setting.
+ */
+static int sdhi_boot_mmc_set_ios(void __iomem *base, struct mmc_ios *ios)
+{
+	if (sd_ctrl_read32(base, CTL_STATUS) & TMIO_STAT_CMD_BUSY)
+		return -EBUSY;
+
+	if (ios->clock)
+		sd_ctrl_write16(base, CTL_SD_CARD_CLK_CTL,
+				ios->clock | CLK_MMC_ENABLE);
+
+	/* Power sequence - OFF -> ON -> UP */
+	switch (ios->power_mode) {
+	case MMC_POWER_OFF: /* power down SD bus */
+		sdhi_boot_mmc_clk_stop(base);
+		break;
+	case MMC_POWER_ON: /* power up SD bus */
+		break;
+	case MMC_POWER_UP: /* start bus clock */
+		sdhi_boot_mmc_clk_start(base);
+		break;
+	}
+
+	switch (ios->bus_width) {
+	case MMC_BUS_WIDTH_1:
+		sd_ctrl_write16(base, CTL_SD_MEM_CARD_OPT, 0x80e0);
+	break;
+	case MMC_BUS_WIDTH_4:
+		sd_ctrl_write16(base, CTL_SD_MEM_CARD_OPT, 0x00e0);
+	break;
+	}
+
+	/* Let things settle. delay taken from winCE driver */
+	udelay(140);
+
+	return 0;
+}
+
+/* These are the bitmasks the tmio chip requires to implement the MMC response
+ * types. Note that R1 and R6 are the same in this scheme. */
+#define RESP_NONE      0x0300
+#define RESP_R1        0x0400
+#define RESP_R1B       0x0500
+#define RESP_R2        0x0600
+#define RESP_R3        0x0700
+#define DATA_PRESENT   0x0800
+#define TRANSFER_READ  0x1000
+
+static int sdhi_boot_request(void __iomem *base, struct mmc_command *cmd)
+{
+	int err, c = cmd->opcode;
+
+	switch (mmc_resp_type(cmd)) {
+	case MMC_RSP_NONE: c |= RESP_NONE; break;
+	case MMC_RSP_R1:   c |= RESP_R1;   break;
+	case MMC_RSP_R1B:  c |= RESP_R1B;  break;
+	case MMC_RSP_R2:   c |= RESP_R2;   break;
+	case MMC_RSP_R3:   c |= RESP_R3;   break;
+	default:
+		return -EINVAL;
+	}
+
+	/* No interrupts so this may not be cleared */
+	sd_ctrl_write32(base, CTL_STATUS, ~TMIO_STAT_CMDRESPEND);
+
+	sd_ctrl_write32(base, CTL_IRQ_MASK, TMIO_STAT_CMDRESPEND |
+			sd_ctrl_read32(base, CTL_IRQ_MASK));
+	sd_ctrl_write32(base, CTL_ARG_REG, cmd->arg);
+	sd_ctrl_write16(base, CTL_SD_CMD, c);
+
+
+	sd_ctrl_write32(base, CTL_IRQ_MASK,
+			~(TMIO_STAT_CMDRESPEND | ALL_ERROR) &
+			sd_ctrl_read32(base, CTL_IRQ_MASK));
+
+	err = sdhi_boot_wait_resp_end(base);
+	if (err)
+		return err;
+
+	cmd->resp[0] = sd_ctrl_read32(base, CTL_RESPONSE);
+
+	return 0;
+}
+
+static int sdhi_boot_do_read_single(void __iomem *base, int high_capacity,
+				    unsigned long block, unsigned short *buf)
+{
+	int err, i;
+
+	/* CMD17 - Read */
+	{
+		struct mmc_command cmd;
+
+		cmd.opcode = MMC_READ_SINGLE_BLOCK | \
+			     TRANSFER_READ | DATA_PRESENT;
+		if (high_capacity)
+			cmd.arg = block;
+		else
+			cmd.arg = block * TMIO_BBS;
+		cmd.flags = MMC_RSP_R1;
+		err = sdhi_boot_request(base, &cmd);
+		if (err)
+			return err;
+	}
+
+	sd_ctrl_write32(base, CTL_IRQ_MASK,
+			~(TMIO_STAT_DATAEND | TMIO_STAT_RXRDY |
+			  TMIO_STAT_TXUNDERRUN) &
+			sd_ctrl_read32(base, CTL_IRQ_MASK));
+	err = sdhi_boot_wait_resp_end(base);
+	if (err)
+		return err;
+
+	sd_ctrl_write16(base, CTL_SD_XFER_LEN, TMIO_BBS);
+	for (i = 0; i < TMIO_BBS / sizeof(*buf); i++)
+		*buf++ = sd_ctrl_read16(base, RESP_CMD12);
+
+	err = sdhi_boot_wait_resp_end(base);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+int sdhi_boot_do_read(void __iomem *base, int high_capacity,
+		      unsigned long offset, unsigned short count,
+		      unsigned short *buf)
+{
+	unsigned long i;
+	int err = 0;
+
+	for (i = 0; i < count; i++) {
+		err = sdhi_boot_do_read_single(base, high_capacity, offset + i,
+					       buf + (i * TMIO_BBS /
+						      sizeof(*buf)));
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+#define VOLTAGES (MMC_VDD_32_33 | MMC_VDD_33_34)
+
+int sdhi_boot_init(void __iomem *base)
+{
+	bool sd_v2 = false, sd_v1_0 = false;
+	unsigned short cid;
+	int err, high_capacity = 0;
+
+	sdhi_boot_mmc_clk_stop(base);
+	sdhi_boot_reset(base);
+
+	/* mmc0: clock 400000Hz busmode 1 powermode 2 cs 0 Vdd 21 width 0 timing 0 */
+	{
+		struct mmc_ios ios;
+		ios.power_mode = MMC_POWER_ON;
+		ios.bus_width = MMC_BUS_WIDTH_1;
+		ios.clock = CLK_MMC_INIT;
+		err = sdhi_boot_mmc_set_ios(base, &ios);
+		if (err)
+			return err;
+	}
+
+	/* CMD0 */
+	{
+		struct mmc_command cmd;
+		msleep(1);
+		cmd.opcode = MMC_GO_IDLE_STATE;
+		cmd.arg = 0;
+		cmd.flags = MMC_RSP_NONE;
+		err = sdhi_boot_request(base, &cmd);
+		if (err)
+			return err;
+		msleep(2);
+	}
+
+	/* CMD8 - Test for SD version 2 */
+	{
+		struct mmc_command cmd;
+		cmd.opcode = SD_SEND_IF_COND;
+		cmd.arg = (VOLTAGES != 0) << 8 | 0xaa;
+		cmd.flags = MMC_RSP_R1;
+		err = sdhi_boot_request(base, &cmd); /* Ignore error */
+		if ((cmd.resp[0] & 0xff) == 0xaa)
+			sd_v2 = true;
+	}
+
+	/* CMD55 - Get OCR (SD) */
+	{
+		int timeout = 1000;
+		struct mmc_command cmd;
+
+		cmd.arg = 0;
+
+		do {
+			cmd.opcode = MMC_APP_CMD;
+			cmd.flags = MMC_RSP_R1;
+			cmd.arg = 0;
+			err = sdhi_boot_request(base, &cmd);
+			if (err)
+				break;
+
+			cmd.opcode = SD_APP_OP_COND;
+			cmd.flags = MMC_RSP_R3;
+			cmd.arg = (VOLTAGES & 0xff8000);
+			if (sd_v2)
+				cmd.arg |= OCR_HCS;
+			cmd.arg |= OCR_FASTBOOT;
+			err = sdhi_boot_request(base, &cmd);
+			if (err)
+				break;
+
+			msleep(1);
+		} while((!(cmd.resp[0] & OCR_BUSY)) && --timeout);
+
+		if (!err && timeout) {
+			if (!sd_v2)
+				sd_v1_0 = true;
+			high_capacity = (cmd.resp[0] & OCR_HCS) == OCR_HCS;
+		}
+	}
+
+	/* CMD1 - Get OCR (MMC) */
+	if (!sd_v2 && !sd_v1_0) {
+		int timeout = 1000;
+		struct mmc_command cmd;
+
+		do {
+			cmd.opcode = MMC_SEND_OP_COND;
+			cmd.arg = VOLTAGES | OCR_HCS;
+			cmd.flags = MMC_RSP_R3;
+			err = sdhi_boot_request(base, &cmd);
+			if (err)
+				return err;
+
+			msleep(1);
+		} while((!(cmd.resp[0] & OCR_BUSY)) && --timeout);
+
+		if (!timeout)
+			return -EAGAIN;
+
+		high_capacity = (cmd.resp[0] & OCR_HCS) == OCR_HCS;
+	}
+
+	/* CMD2 - Get CID */
+	{
+		struct mmc_command cmd;
+		cmd.opcode = MMC_ALL_SEND_CID;
+		cmd.arg = 0;
+		cmd.flags = MMC_RSP_R2;
+		err = sdhi_boot_request(base, &cmd);
+		if (err)
+			return err;
+	}
+
+	/* CMD3
+	 * MMC: Set the relative address
+	 * SD:  Get the relative address
+	 * Also puts the card into the standby state
+	 */
+	{
+		struct mmc_command cmd;
+		cmd.opcode = MMC_SET_RELATIVE_ADDR;
+		cmd.arg = 0;
+		cmd.flags = MMC_RSP_R1;
+		err = sdhi_boot_request(base, &cmd);
+		if (err)
+			return err;
+		cid = cmd.resp[0] >> 16;
+	}
+
+	/* CMD9 - Get CSD */
+	{
+		struct mmc_command cmd;
+		cmd.opcode = MMC_SEND_CSD;
+		cmd.arg = cid << 16;
+		cmd.flags = MMC_RSP_R2;
+		err = sdhi_boot_request(base, &cmd);
+		if (err)
+			return err;
+	}
+
+	/* CMD7 - Select the card */
+	{
+		struct mmc_command cmd;
+		cmd.opcode = MMC_SELECT_CARD;
+		//cmd.arg = rca << 16;
+		cmd.arg = cid << 16;
+		//cmd.flags = MMC_RSP_R1B;
+		cmd.flags = MMC_RSP_R1;
+		err = sdhi_boot_request(base, &cmd);
+		if (err)
+			return err;
+	}
+
+	/* CMD16 - Set the block size */
+	{
+		struct mmc_command cmd;
+		cmd.opcode = MMC_SET_BLOCKLEN;
+		cmd.arg = TMIO_BBS;
+		cmd.flags = MMC_RSP_R1;
+		err = sdhi_boot_request(base, &cmd);
+		if (err)
+			return err;
+	}
+
+	return high_capacity;
+}
diff --git a/arch/arm/boot/compressed/sdhi-shmobile.h b/arch/arm/boot/compressed/sdhi-shmobile.h
new file mode 100644
index 000000000000..92eaa09f985e
--- /dev/null
+++ b/arch/arm/boot/compressed/sdhi-shmobile.h
@@ -0,0 +1,11 @@
+#ifndef SDHI_MOBILE_H
+#define SDHI_MOBILE_H
+
+#include <linux/compiler.h>
+
+int sdhi_boot_do_read(void __iomem *base, int high_capacity,
+		      unsigned long offset, unsigned short count,
+		      unsigned short *buf);
+int sdhi_boot_init(void __iomem *base);
+
+#endif
diff --git a/arch/arm/boot/compressed/string.c b/arch/arm/boot/compressed/string.c
new file mode 100644
index 000000000000..36e53ef9200f
--- /dev/null
+++ b/arch/arm/boot/compressed/string.c
@@ -0,0 +1,127 @@
+/*
+ * arch/arm/boot/compressed/string.c
+ *
+ * Small subset of simple string routines
+ */
+
+#include <linux/string.h>
+
+void *memcpy(void *__dest, __const void *__src, size_t __n)
+{
+	int i = 0;
+	unsigned char *d = (unsigned char *)__dest, *s = (unsigned char *)__src;
+
+	for (i = __n >> 3; i > 0; i--) {
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+	}
+
+	if (__n & 1 << 2) {
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+	}
+
+	if (__n & 1 << 1) {
+		*d++ = *s++;
+		*d++ = *s++;
+	}
+
+	if (__n & 1)
+		*d++ = *s++;
+
+	return __dest;
+}
+
+void *memmove(void *__dest, __const void *__src, size_t count)
+{
+	unsigned char *d = __dest;
+	const unsigned char *s = __src;
+
+	if (__dest == __src)
+		return __dest;
+
+	if (__dest < __src)
+		return memcpy(__dest, __src, count);
+
+	while (count--)
+		d[count] = s[count];
+	return __dest;
+}
+
+size_t strlen(const char *s)
+{
+	const char *sc = s;
+
+	while (*sc != '\0')
+		sc++;
+	return sc - s;
+}
+
+int memcmp(const void *cs, const void *ct, size_t count)
+{
+	const unsigned char *su1 = cs, *su2 = ct, *end = su1 + count;
+	int res = 0;
+
+	while (su1 < end) {
+		res = *su1++ - *su2++;
+		if (res)
+			break;
+	}
+	return res;
+}
+
+int strcmp(const char *cs, const char *ct)
+{
+	unsigned char c1, c2;
+	int res = 0;
+
+	do {
+		c1 = *cs++;
+		c2 = *ct++;
+		res = c1 - c2;
+		if (res)
+			break;
+	} while (c1);
+	return res;
+}
+
+void *memchr(const void *s, int c, size_t count)
+{
+	const unsigned char *p = s;
+
+	while (count--)
+		if ((unsigned char)c == *p++)
+			return (void *)(p - 1);
+	return NULL;
+}
+
+char *strchr(const char *s, int c)
+{
+	while (*s != (char)c)
+		if (*s++ == '\0')
+			return NULL;
+	return (char *)s;
+}
+
+#undef memset
+
+void *memset(void *s, int c, size_t count)
+{
+	char *xs = s;
+	while (count--)
+		*xs++ = c;
+	return s;
+}
+
+void __memzero(void *s, size_t count)
+{
+	memset(s, 0, count);
+}
diff --git a/arch/arm/boot/compressed/vmlinux.lds.in b/arch/arm/boot/compressed/vmlinux.lds.in
index ea80abe78844..4919f2ac8b89 100644
--- a/arch/arm/boot/compressed/vmlinux.lds.in
+++ b/arch/arm/boot/compressed/vmlinux.lds.in
@@ -33,20 +33,28 @@ SECTIONS
     *(.text.*)
     *(.fixup)
     *(.gnu.warning)
+    *(.glue_7t)
+    *(.glue_7)
+  }
+  .rodata : {
     *(.rodata)
     *(.rodata.*)
-    *(.glue_7)
-    *(.glue_7t)
+  }
+  .piggydata : {
     *(.piggydata)
-    . = ALIGN(4);
   }
 
+  . = ALIGN(4);
   _etext = .;
 
+  .got.plt		: { *(.got.plt) }
   _got_start = .;
   .got			: { *(.got) }
   _got_end = .;
-  .got.plt		: { *(.got.plt) }
+
+  /* ensure the zImage file size is always a multiple of 64 bits */
+  /* (without a dummy byte, ld just ignores the empty section) */
+  .pad			: { BYTE(0); . = ALIGN(8); }
   _edata = .;
 
   . = BSS_START;