Merge branch 'origin' into devel-stable

Conflicts:
	arch/arm/mach-mx2/devices.c
	arch/arm/mach-mx2/devices.h
	sound/soc/pxa/pxa-ssp.c

412 files changed, 20633 insertions, 9432 deletions

diff --git a/fs/9p/fid.c b/fs/9p/fid.c
index 14d944204571..08b2eb157048 100644
--- a/fs/9p/fid.c
+++ b/fs/9p/fid.c
@@ -151,7 +151,7 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry)
 			if (access == V9FS_ACCESS_SINGLE)
 				return ERR_PTR(-EPERM);
 
-			if (v9fs_extended(v9ses))
+			if (v9fs_proto_dotu(v9ses))
 				uname = NULL;
 			else
 				uname = v9ses->uname;
diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c
index 7d6c2139891d..6c7f6a251115 100644
--- a/fs/9p/v9fs.c
+++ b/fs/9p/v9fs.c
@@ -241,7 +241,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses,
 	list_add(&v9ses->slist, &v9fs_sessionlist);
 	spin_unlock(&v9fs_sessionlist_lock);
 
-	v9ses->flags = V9FS_EXTENDED | V9FS_ACCESS_USER;
+	v9ses->flags = V9FS_PROTO_2000U | V9FS_ACCESS_USER;
 	strcpy(v9ses->uname, V9FS_DEFUSER);
 	strcpy(v9ses->aname, V9FS_DEFANAME);
 	v9ses->uid = ~0;
@@ -262,13 +262,13 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses,
 		goto error;
 	}
 
-	if (!v9ses->clnt->dotu)
-		v9ses->flags &= ~V9FS_EXTENDED;
+	if (!p9_is_proto_dotu(v9ses->clnt))
+		v9ses->flags &= ~V9FS_PROTO_2000U;
 
 	v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ;
 
 	/* for legacy mode, fall back to V9FS_ACCESS_ANY */
-	if (!v9fs_extended(v9ses) &&
+	if (!v9fs_proto_dotu(v9ses) &&
 		((v9ses->flags&V9FS_ACCESS_MASK) == V9FS_ACCESS_USER)) {
 
 		v9ses->flags &= ~V9FS_ACCESS_MASK;
diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h
index 019f4ccb70c1..79000bf62491 100644
--- a/fs/9p/v9fs.h
+++ b/fs/9p/v9fs.h
@@ -23,7 +23,8 @@
 
 /**
  * enum p9_session_flags - option flags for each 9P session
- * @V9FS_EXTENDED: whether or not to use 9P2000.u extensions
+ * @V9FS_PROTO_2000U: whether or not to use 9P2000.u extensions
+ * @V9FS_PROTO_2010L: whether or not to use 9P2010.l extensions
  * @V9FS_ACCESS_SINGLE: only the mounting user can access the hierarchy
  * @V9FS_ACCESS_USER: a new attach will be issued for every user (default)
  * @V9FS_ACCESS_ANY: use a single attach for all users
@@ -32,11 +33,12 @@
  * Session flags reflect options selected by users at mount time
  */
 enum p9_session_flags {
-	V9FS_EXTENDED		= 0x01,
-	V9FS_ACCESS_SINGLE	= 0x02,
-	V9FS_ACCESS_USER	= 0x04,
-	V9FS_ACCESS_ANY		= 0x06,
-	V9FS_ACCESS_MASK	= 0x06,
+	V9FS_PROTO_2000U	= 0x01,
+	V9FS_PROTO_2010L	= 0x02,
+	V9FS_ACCESS_SINGLE	= 0x04,
+	V9FS_ACCESS_USER	= 0x08,
+	V9FS_ACCESS_ANY		= 0x0C,
+	V9FS_ACCESS_MASK	= 0x0C,
 };
 
 /* possible values of ->cache */
@@ -121,7 +123,12 @@ static inline struct v9fs_session_info *v9fs_inode2v9ses(struct inode *inode)
 	return (inode->i_sb->s_fs_info);
 }
 
-static inline int v9fs_extended(struct v9fs_session_info *v9ses)
+static inline int v9fs_proto_dotu(struct v9fs_session_info *v9ses)
 {
-	return v9ses->flags & V9FS_EXTENDED;
+	return v9ses->flags & V9FS_PROTO_2000U;
+}
+
+static inline int v9fs_proto_dotl(struct v9fs_session_info *v9ses)
+{
+	return v9ses->flags & V9FS_PROTO_2010L;
 }
diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c
index 15cce53bf61e..6580aa449541 100644
--- a/fs/9p/vfs_dir.c
+++ b/fs/9p/vfs_dir.c
@@ -135,7 +135,7 @@ static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir)
 		while (rdir->head < rdir->tail) {
 			err = p9stat_read(rdir->buf + rdir->head,
 						buflen - rdir->head, &st,
-						fid->clnt->dotu);
+						fid->clnt->proto_version);
 			if (err) {
 				P9_DPRINTK(P9_DEBUG_VFS, "returned %d\n", err);
 				err = -EIO;
diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index 74a0461a9ac0..36122683fae8 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -61,7 +61,7 @@ int v9fs_file_open(struct inode *inode, struct file *file)
 
 	P9_DPRINTK(P9_DEBUG_VFS, "inode: %p file: %p \n", inode, file);
 	v9ses = v9fs_inode2v9ses(inode);
-	omode = v9fs_uflags2omode(file->f_flags, v9fs_extended(v9ses));
+	omode = v9fs_uflags2omode(file->f_flags, v9fs_proto_dotu(v9ses));
 	fid = file->private_data;
 	if (!fid) {
 		fid = v9fs_fid_clone(file->f_path.dentry);
@@ -77,7 +77,7 @@ int v9fs_file_open(struct inode *inode, struct file *file)
 			i_size_write(inode, 0);
 			inode->i_blocks = 0;
 		}
-		if ((file->f_flags & O_APPEND) && (!v9fs_extended(v9ses)))
+		if ((file->f_flags & O_APPEND) && (!v9fs_proto_dotu(v9ses)))
 			generic_file_llseek(file, 0, SEEK_END);
 	}
 
diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c
index a407fa3388c0..5fe45d692c9f 100644
--- a/fs/9p/vfs_inode.c
+++ b/fs/9p/vfs_inode.c
@@ -60,7 +60,7 @@ static int unixmode2p9mode(struct v9fs_session_info *v9ses, int mode)
 	res = mode & 0777;
 	if (S_ISDIR(mode))
 		res |= P9_DMDIR;
-	if (v9fs_extended(v9ses)) {
+	if (v9fs_proto_dotu(v9ses)) {
 		if (S_ISLNK(mode))
 			res |= P9_DMSYMLINK;
 		if (v9ses->nodev == 0) {
@@ -102,21 +102,21 @@ static int p9mode2unixmode(struct v9fs_session_info *v9ses, int mode)
 
 	if ((mode & P9_DMDIR) == P9_DMDIR)
 		res |= S_IFDIR;
-	else if ((mode & P9_DMSYMLINK) && (v9fs_extended(v9ses)))
+	else if ((mode & P9_DMSYMLINK) && (v9fs_proto_dotu(v9ses)))
 		res |= S_IFLNK;
-	else if ((mode & P9_DMSOCKET) && (v9fs_extended(v9ses))
+	else if ((mode & P9_DMSOCKET) && (v9fs_proto_dotu(v9ses))
 		 && (v9ses->nodev == 0))
 		res |= S_IFSOCK;
-	else if ((mode & P9_DMNAMEDPIPE) && (v9fs_extended(v9ses))
+	else if ((mode & P9_DMNAMEDPIPE) && (v9fs_proto_dotu(v9ses))
 		 && (v9ses->nodev == 0))
 		res |= S_IFIFO;
-	else if ((mode & P9_DMDEVICE) && (v9fs_extended(v9ses))
+	else if ((mode & P9_DMDEVICE) && (v9fs_proto_dotu(v9ses))
 		 && (v9ses->nodev == 0))
 		res |= S_IFBLK;
 	else
 		res |= S_IFREG;
 
-	if (v9fs_extended(v9ses)) {
+	if (v9fs_proto_dotu(v9ses)) {
 		if ((mode & P9_DMSETUID) == P9_DMSETUID)
 			res |= S_ISUID;
 
@@ -265,7 +265,7 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode)
 	case S_IFBLK:
 	case S_IFCHR:
 	case S_IFSOCK:
-		if (!v9fs_extended(v9ses)) {
+		if (!v9fs_proto_dotu(v9ses)) {
 			P9_DPRINTK(P9_DEBUG_ERROR,
 				   "special files without extended mode\n");
 			err = -EINVAL;
@@ -278,7 +278,7 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode)
 		inode->i_fop = &v9fs_file_operations;
 		break;
 	case S_IFLNK:
-		if (!v9fs_extended(v9ses)) {
+		if (!v9fs_proto_dotu(v9ses)) {
 			P9_DPRINTK(P9_DEBUG_ERROR,
 				   "extended modes used w/o 9P2000.u\n");
 			err = -EINVAL;
@@ -288,7 +288,7 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode)
 		break;
 	case S_IFDIR:
 		inc_nlink(inode);
-		if (v9fs_extended(v9ses))
+		if (v9fs_proto_dotu(v9ses))
 			inode->i_op = &v9fs_dir_inode_operations_ext;
 		else
 			inode->i_op = &v9fs_dir_inode_operations;
@@ -575,7 +575,8 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 		flags = O_RDWR;
 
 	fid = v9fs_create(v9ses, dir, dentry, NULL, perm,
-				v9fs_uflags2omode(flags, v9fs_extended(v9ses)));
+				v9fs_uflags2omode(flags,
+						v9fs_proto_dotu(v9ses)));
 	if (IS_ERR(fid)) {
 		err = PTR_ERR(fid);
 		fid = NULL;
@@ -858,7 +859,7 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr)
 	if (iattr->ia_valid & ATTR_SIZE)
 		wstat.length = iattr->ia_size;
 
-	if (v9fs_extended(v9ses)) {
+	if (v9fs_proto_dotu(v9ses)) {
 		if (iattr->ia_valid & ATTR_UID)
 			wstat.n_uid = iattr->ia_uid;
 
@@ -886,6 +887,8 @@ v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode,
 	struct super_block *sb)
 {
 	char ext[32];
+	char tag_name[14];
+	unsigned int i_nlink;
 	struct v9fs_session_info *v9ses = sb->s_fs_info;
 
 	inode->i_nlink = 1;
@@ -897,11 +900,26 @@ v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode,
 	inode->i_uid = v9ses->dfltuid;
 	inode->i_gid = v9ses->dfltgid;
 
-	if (v9fs_extended(v9ses)) {
+	if (v9fs_proto_dotu(v9ses)) {
 		inode->i_uid = stat->n_uid;
 		inode->i_gid = stat->n_gid;
 	}
-
+	if ((S_ISREG(inode->i_mode)) || (S_ISDIR(inode->i_mode))) {
+		if (v9fs_proto_dotu(v9ses) && (stat->extension[0] != '\0')) {
+			/*
+			 * Hadlink support got added later to
+			 * to the .u extension. So there can be
+			 * server out there that doesn't support
+			 * this even with .u extension. So check
+			 * for non NULL stat->extension
+			 */
+			strncpy(ext, stat->extension, sizeof(ext));
+			/* HARDLINKCOUNT %u */
+			sscanf(ext, "%13s %u", tag_name, &i_nlink);
+			if (!strncmp(tag_name, "HARDLINKCOUNT", 13))
+				inode->i_nlink = i_nlink;
+		}
+	}
 	inode->i_mode = p9mode2unixmode(v9ses, stat->mode);
 	if ((S_ISBLK(inode->i_mode)) || (S_ISCHR(inode->i_mode))) {
 		char type = 0;
@@ -976,7 +994,7 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen)
 	if (IS_ERR(fid))
 		return PTR_ERR(fid);
 
-	if (!v9fs_extended(v9ses))
+	if (!v9fs_proto_dotu(v9ses))
 		return -EBADF;
 
 	st = p9_client_stat(fid);
@@ -1066,7 +1084,7 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry,
 	struct p9_fid *fid;
 
 	v9ses = v9fs_inode2v9ses(dir);
-	if (!v9fs_extended(v9ses)) {
+	if (!v9fs_proto_dotu(v9ses)) {
 		P9_DPRINTK(P9_DEBUG_ERROR, "not extended\n");
 		return -EPERM;
 	}
diff --git a/fs/Kconfig b/fs/Kconfig
index 64d44efad7a5..7405f071be67 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -177,6 +177,7 @@ source "fs/efs/Kconfig"
 source "fs/jffs2/Kconfig"
 # UBIFS File system configuration
 source "fs/ubifs/Kconfig"
+source "fs/logfs/Kconfig"
 source "fs/cramfs/Kconfig"
 source "fs/squashfs/Kconfig"
 source "fs/freevxfs/Kconfig"
diff --git a/fs/Makefile b/fs/Makefile
index af6d04700d9c..c3633aa46911 100644
--- a/fs/Makefile
+++ b/fs/Makefile
@@ -99,6 +99,7 @@ obj-$(CONFIG_NTFS_FS)		+= ntfs/
 obj-$(CONFIG_UFS_FS)		+= ufs/
 obj-$(CONFIG_EFS_FS)		+= efs/
 obj-$(CONFIG_JFFS2_FS)		+= jffs2/
+obj-$(CONFIG_LOGFS)		+= logfs/
 obj-$(CONFIG_UBIFS_FS)		+= ubifs/
 obj-$(CONFIG_AFFS_FS)		+= affs/
 obj-$(CONFIG_ROMFS_FS)		+= romfs/
diff --git a/fs/adfs/adfs.h b/fs/adfs/adfs.h
index 9cc18775b832..2ff622f6f547 100644
--- a/fs/adfs/adfs.h
+++ b/fs/adfs/adfs.h
@@ -121,7 +121,7 @@ struct adfs_discmap {
 
 /* Inode stuff */
 struct inode *adfs_iget(struct super_block *sb, struct object_info *obj);
-int adfs_write_inode(struct inode *inode,int unused);
+int adfs_write_inode(struct inode *inode, struct writeback_control *wbc);
 int adfs_notify_change(struct dentry *dentry, struct iattr *attr);
 
 /* map.c */
diff --git a/fs/adfs/inode.c b/fs/adfs/inode.c
index 3f57ce4bee5d..0f5e30978135 100644
--- a/fs/adfs/inode.c
+++ b/fs/adfs/inode.c
@@ -9,6 +9,7 @@
  */
 #include <linux/smp_lock.h>
 #include <linux/buffer_head.h>
+#include <linux/writeback.h>
 #include "adfs.h"
 
 /*
@@ -360,7 +361,7 @@ out:
  * The adfs-specific inode data has already been updated by
  * adfs_notify_change()
  */
-int adfs_write_inode(struct inode *inode, int wait)
+int adfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct super_block *sb = inode->i_sb;
 	struct object_info obj;
@@ -375,7 +376,7 @@ int adfs_write_inode(struct inode *inode, int wait)
 	obj.attr	= ADFS_I(inode)->attr;
 	obj.size	= inode->i_size;
 
-	ret = adfs_dir_update(sb, &obj, wait);
+	ret = adfs_dir_update(sb, &obj, wbc->sync_mode == WB_SYNC_ALL);
 	unlock_kernel();
 	return ret;
 }
diff --git a/fs/affs/affs.h b/fs/affs/affs.h
index 0e40caaba456..861dae68ac12 100644
--- a/fs/affs/affs.h
+++ b/fs/affs/affs.h
@@ -175,7 +175,8 @@ extern void			 affs_delete_inode(struct inode *inode);
 extern void			 affs_clear_inode(struct inode *inode);
 extern struct inode		*affs_iget(struct super_block *sb,
 					unsigned long ino);
-extern int			 affs_write_inode(struct inode *inode, int);
+extern int			 affs_write_inode(struct inode *inode,
+					struct writeback_control *wbc);
 extern int			 affs_add_entry(struct inode *dir, struct inode *inode, struct dentry *dentry, s32 type);
 
 /* file.c */
diff --git a/fs/affs/inode.c b/fs/affs/inode.c
index 3c4ec7d864c4..c9744d771d98 100644
--- a/fs/affs/inode.c
+++ b/fs/affs/inode.c
@@ -166,7 +166,7 @@ bad_inode:
 }
 
 int
-affs_write_inode(struct inode *inode, int unused)
+affs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct super_block	*sb = inode->i_sb;
 	struct buffer_head	*bh;
diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index 6ece2a13bf71..c54dad4e6063 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -733,7 +733,6 @@ extern int afs_write_end(struct file *file, struct address_space *mapping,
 			struct page *page, void *fsdata);
 extern int afs_writepage(struct page *, struct writeback_control *);
 extern int afs_writepages(struct address_space *, struct writeback_control *);
-extern int afs_write_inode(struct inode *, int);
 extern void afs_pages_written_back(struct afs_vnode *, struct afs_call *);
 extern ssize_t afs_file_write(struct kiocb *, const struct iovec *,
 			      unsigned long, loff_t);
diff --git a/fs/afs/super.c b/fs/afs/super.c
index e1ea1c240b6a..14f6431598ad 100644
--- a/fs/afs/super.c
+++ b/fs/afs/super.c
@@ -48,7 +48,6 @@ struct file_system_type afs_fs_type = {
 static const struct super_operations afs_super_ops = {
 	.statfs		= afs_statfs,
 	.alloc_inode	= afs_alloc_inode,
-	.write_inode	= afs_write_inode,
 	.destroy_inode	= afs_destroy_inode,
 	.clear_inode	= afs_clear_inode,
 	.put_super	= afs_put_super,
diff --git a/fs/afs/write.c b/fs/afs/write.c
index 5e15a21dbf9f..3bed54a294d4 100644
--- a/fs/afs/write.c
+++ b/fs/afs/write.c
@@ -585,27 +585,6 @@ int afs_writepages(struct address_space *mapping,
 }
 
 /*
- * write an inode back
- */
-int afs_write_inode(struct inode *inode, int sync)
-{
-	struct afs_vnode *vnode = AFS_FS_I(inode);
-	int ret;
-
-	_enter("{%x:%u},", vnode->fid.vid, vnode->fid.vnode);
-
-	ret = 0;
-	if (sync) {
-		ret = filemap_fdatawait(inode->i_mapping);
-		if (ret < 0)
-			__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
-	}
-
-	_leave(" = %d", ret);
-	return ret;
-}
-
-/*
  * completion of write to server
  */
 void afs_pages_written_back(struct afs_vnode *vnode, struct afs_call *call)
diff --git a/fs/attr.c b/fs/attr.c
index 96d394bdaddf..0815e93bb487 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -12,7 +12,6 @@
 #include <linux/capability.h>
 #include <linux/fsnotify.h>
 #include <linux/fcntl.h>
-#include <linux/quotaops.h>
 #include <linux/security.h>
 
 /* Taken over from the old code... */
@@ -82,7 +81,7 @@ int inode_newsize_ok(const struct inode *inode, loff_t offset)
 	if (inode->i_size < offset) {
 		unsigned long limit;
 
-		limit = current->signal->rlim[RLIMIT_FSIZE].rlim_cur;
+		limit = rlimit(RLIMIT_FSIZE);
 		if (limit != RLIM_INFINITY && offset > limit)
 			goto out_sig;
 		if (offset > inode->i_sb->s_maxbytes)
@@ -212,14 +211,8 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
 		error = inode->i_op->setattr(dentry, attr);
 	} else {
 		error = inode_change_ok(inode, attr);
-		if (!error) {
-			if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
-			    (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid))
-				error = vfs_dq_transfer(inode, attr) ?
-					-EDQUOT : 0;
-			if (!error)
-				error = inode_setattr(inode, attr);
-		}
+		if (!error)
+			error = inode_setattr(inode, attr);
 	}
 
 	if (ia_valid & ATTR_SIZE)
diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h
index 0118d67221b2..3d283abf67d7 100644
--- a/fs/autofs4/autofs_i.h
+++ b/fs/autofs4/autofs_i.h
@@ -60,11 +60,6 @@ do {							\
 		current->pid, __func__, ##args);	\
 } while (0)
 
-struct rehash_entry {
-	struct task_struct *task;
-	struct list_head list;
-};
-
 /* Unified info structure.  This is pointed to by both the dentry and
    inode structures.  Each file in the filesystem has an instance of this
    structure.  It holds a reference to the dentry, so dentries are never
@@ -81,7 +76,6 @@ struct autofs_info {
 
 	struct list_head active;
 	int active_count;
-	struct list_head rehash_list;
 
 	struct list_head expiring;
 
@@ -104,7 +98,6 @@ struct autofs_info {
 #define AUTOFS_INF_EXPIRING	(1<<0) /* dentry is in the process of expiring */
 #define AUTOFS_INF_MOUNTPOINT	(1<<1) /* mountpoint status for direct expire */
 #define AUTOFS_INF_PENDING	(1<<2) /* dentry pending mount */
-#define AUTOFS_INF_REHASH	(1<<3) /* dentry in transit to ->lookup() */
 
 struct autofs_wait_queue {
 	wait_queue_head_t queue;
diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c
index 00bf8fcb245f..c8a80dffb455 100644
--- a/fs/autofs4/dev-ioctl.c
+++ b/fs/autofs4/dev-ioctl.c
@@ -544,10 +544,9 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp,
 			goto out;
 		devid = new_encode_dev(path.mnt->mnt_sb->s_dev);
 		err = 0;
-		if (path.dentry->d_inode &&
-		    path.mnt->mnt_root == path.dentry) {
+		if (path.mnt->mnt_root == path.dentry) {
 			err = 1;
-			magic = path.dentry->d_inode->i_sb->s_magic;
+			magic = path.mnt->mnt_sb->s_magic;
 		}
 	} else {
 		dev_t dev = sbi->sb->s_dev;
@@ -560,10 +559,8 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp,
 
 		err = have_submounts(path.dentry);
 
-		if (path.mnt->mnt_mountpoint != path.mnt->mnt_root) {
-			if (follow_down(&path))
-				magic = path.mnt->mnt_sb->s_magic;
-		}
+		if (follow_down(&path))
+			magic = path.mnt->mnt_sb->s_magic;
 	}
 
 	param->ismountpoint.out.devid = devid;
diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c
index 74bc9aa6df31..a796c9417fb1 100644
--- a/fs/autofs4/expire.c
+++ b/fs/autofs4/expire.c
@@ -279,7 +279,6 @@ struct dentry *autofs4_expire_direct(struct super_block *sb,
 			root->d_mounted--;
 		}
 		ino->flags |= AUTOFS_INF_EXPIRING;
-		autofs4_add_expiring(root);
 		init_completion(&ino->expire_complete);
 		spin_unlock(&sbi->fs_lock);
 		return root;
@@ -407,7 +406,6 @@ found:
 		expired, (int)expired->d_name.len, expired->d_name.name);
 	ino = autofs4_dentry_ino(expired);
 	ino->flags |= AUTOFS_INF_EXPIRING;
-	autofs4_add_expiring(expired);
 	init_completion(&ino->expire_complete);
 	spin_unlock(&sbi->fs_lock);
 	spin_lock(&dcache_lock);
@@ -435,7 +433,7 @@ int autofs4_expire_wait(struct dentry *dentry)
 
 		DPRINTK("expire done status=%d", status);
 
-		if (d_unhashed(dentry) && IS_DEADDIR(dentry->d_inode))
+		if (d_unhashed(dentry))
 			return -EAGAIN;
 
 		return status;
@@ -475,7 +473,6 @@ int autofs4_expire_run(struct super_block *sb,
 	spin_lock(&sbi->fs_lock);
 	ino = autofs4_dentry_ino(dentry);
 	ino->flags &= ~AUTOFS_INF_EXPIRING;
-	autofs4_del_expiring(dentry);
 	complete_all(&ino->expire_complete);
 	spin_unlock(&sbi->fs_lock);
 
@@ -506,7 +503,6 @@ int autofs4_do_expire_multi(struct super_block *sb, struct vfsmount *mnt,
 			ino->flags &= ~AUTOFS_INF_MOUNTPOINT;
 		}
 		ino->flags &= ~AUTOFS_INF_EXPIRING;
-		autofs4_del_expiring(dentry);
 		complete_all(&ino->expire_complete);
 		spin_unlock(&sbi->fs_lock);
 		dput(dentry);
diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c
index d0a3de247458..821b2b955dac 100644
--- a/fs/autofs4/inode.c
+++ b/fs/autofs4/inode.c
@@ -49,7 +49,6 @@ struct autofs_info *autofs4_init_ino(struct autofs_info *ino,
 		ino->dentry = NULL;
 		ino->size = 0;
 		INIT_LIST_HEAD(&ino->active);
-		INIT_LIST_HEAD(&ino->rehash_list);
 		ino->active_count = 0;
 		INIT_LIST_HEAD(&ino->expiring);
 		atomic_set(&ino->count, 0);
@@ -97,63 +96,6 @@ void autofs4_free_ino(struct autofs_info *ino)
 	kfree(ino);
 }
 
-/*
- * Deal with the infamous "Busy inodes after umount ..." message.
- *
- * Clean up the dentry tree. This happens with autofs if the user
- * space program goes away due to a SIGKILL, SIGSEGV etc.
- */
-static void autofs4_force_release(struct autofs_sb_info *sbi)
-{
-	struct dentry *this_parent = sbi->sb->s_root;
-	struct list_head *next;
-
-	if (!sbi->sb->s_root)
-		return;
-
-	spin_lock(&dcache_lock);
-repeat:
-	next = this_parent->d_subdirs.next;
-resume:
-	while (next != &this_parent->d_subdirs) {
-		struct dentry *dentry = list_entry(next, struct dentry, d_u.d_child);
-
-		/* Negative dentry - don`t care */
-		if (!simple_positive(dentry)) {
-			next = next->next;
-			continue;
-		}
-
-		if (!list_empty(&dentry->d_subdirs)) {
-			this_parent = dentry;
-			goto repeat;
-		}
-
-		next = next->next;
-		spin_unlock(&dcache_lock);
-
-		DPRINTK("dentry %p %.*s",
-			dentry, (int)dentry->d_name.len, dentry->d_name.name);
-
-		dput(dentry);
-		spin_lock(&dcache_lock);
-	}
-
-	if (this_parent != sbi->sb->s_root) {
-		struct dentry *dentry = this_parent;
-
-		next = this_parent->d_u.d_child.next;
-		this_parent = this_parent->d_parent;
-		spin_unlock(&dcache_lock);
-		DPRINTK("parent dentry %p %.*s",
-			dentry, (int)dentry->d_name.len, dentry->d_name.name);
-		dput(dentry);
-		spin_lock(&dcache_lock);
-		goto resume;
-	}
-	spin_unlock(&dcache_lock);
-}
-
 void autofs4_kill_sb(struct super_block *sb)
 {
 	struct autofs_sb_info *sbi = autofs4_sbi(sb);
@@ -170,15 +112,12 @@ void autofs4_kill_sb(struct super_block *sb)
 	/* Free wait queues, close pipe */
 	autofs4_catatonic_mode(sbi);
 
-	/* Clean up and release dangling references */
-	autofs4_force_release(sbi);
-
 	sb->s_fs_info = NULL;
 	kfree(sbi);
 
 out_kill_sb:
 	DPRINTK("shutting down");
-	kill_anon_super(sb);
+	kill_litter_super(sb);
 }
 
 static int autofs4_show_options(struct seq_file *m, struct vfsmount *mnt)
diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c
index 30cc9ddf4b70..a015b49891df 100644
--- a/fs/autofs4/root.c
+++ b/fs/autofs4/root.c
@@ -104,99 +104,6 @@ static void autofs4_del_active(struct dentry *dentry)
 	return;
 }
 
-static void autofs4_add_rehash_entry(struct autofs_info *ino,
-				     struct rehash_entry *entry)
-{
-	entry->task = current;
-	INIT_LIST_HEAD(&entry->list);
-	list_add(&entry->list, &ino->rehash_list);
-	return;
-}
-
-static void autofs4_remove_rehash_entry(struct autofs_info *ino)
-{
-	struct list_head *head = &ino->rehash_list;
-	struct rehash_entry *entry;
-	list_for_each_entry(entry, head, list) {
-		if (entry->task == current) {
-			list_del(&entry->list);
-			kfree(entry);
-			break;
-		}
-	}
-	return;
-}
-
-static void autofs4_remove_rehash_entrys(struct autofs_info *ino)
-{
-	struct autofs_sb_info *sbi = ino->sbi;
-	struct rehash_entry *entry, *next;
-	struct list_head *head;
-
-	spin_lock(&sbi->fs_lock);
-	spin_lock(&sbi->lookup_lock);
-	if (!(ino->flags & AUTOFS_INF_REHASH)) {
-		spin_unlock(&sbi->lookup_lock);
-		spin_unlock(&sbi->fs_lock);
-		return;
-	}
-	ino->flags &= ~AUTOFS_INF_REHASH;
-	head = &ino->rehash_list;
-	list_for_each_entry_safe(entry, next, head, list) {
-		list_del(&entry->list);
-		kfree(entry);
-	}
-	spin_unlock(&sbi->lookup_lock);
-	spin_unlock(&sbi->fs_lock);
-	dput(ino->dentry);
-
-	return;
-}
-
-static void autofs4_revalidate_drop(struct dentry *dentry,
-				    struct rehash_entry *entry)
-{
-	struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb);
-	struct autofs_info *ino = autofs4_dentry_ino(dentry);
-	/*
-	 * Add to the active list so we can pick this up in
-	 * ->lookup(). Also add an entry to a rehash list so
-	 * we know when there are no dentrys in flight so we
-	 * know when we can rehash the dentry.
-	 */
-	spin_lock(&sbi->lookup_lock);
-	if (list_empty(&ino->active))
-		list_add(&ino->active, &sbi->active_list);
-	autofs4_add_rehash_entry(ino, entry);
-	spin_unlock(&sbi->lookup_lock);
-	if (!(ino->flags & AUTOFS_INF_REHASH)) {
-		ino->flags |= AUTOFS_INF_REHASH;
-		dget(dentry);
-		spin_lock(&dentry->d_lock);
-		__d_drop(dentry);
-		spin_unlock(&dentry->d_lock);
-	}
-	return;
-}
-
-static void autofs4_revalidate_rehash(struct dentry *dentry)
-{
-	struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb);
-	struct autofs_info *ino = autofs4_dentry_ino(dentry);
-	if (ino->flags & AUTOFS_INF_REHASH) {
-		spin_lock(&sbi->lookup_lock);
-		autofs4_remove_rehash_entry(ino);
-		if (list_empty(&ino->rehash_list)) {
-			spin_unlock(&sbi->lookup_lock);
-			ino->flags &= ~AUTOFS_INF_REHASH;
-			d_rehash(dentry);
-			dput(ino->dentry);
-		} else
-			spin_unlock(&sbi->lookup_lock);
-	}
-	return;
-}
-
 static unsigned int autofs4_need_mount(unsigned int flags)
 {
 	unsigned int res = 0;
@@ -236,7 +143,7 @@ out:
 	return dcache_dir_open(inode, file);
 }
 
-static int try_to_fill_dentry(struct dentry *dentry)
+static int try_to_fill_dentry(struct dentry *dentry, int flags)
 {
 	struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb);
 	struct autofs_info *ino = autofs4_dentry_ino(dentry);
@@ -249,17 +156,55 @@ static int try_to_fill_dentry(struct dentry *dentry)
 	 * Wait for a pending mount, triggering one if there
 	 * isn't one already
 	 */
-	DPRINTK("waiting for mount name=%.*s",
-		 dentry->d_name.len, dentry->d_name.name);
+	if (dentry->d_inode == NULL) {
+		DPRINTK("waiting for mount name=%.*s",
+			 dentry->d_name.len, dentry->d_name.name);
 
-	status = autofs4_wait(sbi, dentry, NFY_MOUNT);
+		status = autofs4_wait(sbi, dentry, NFY_MOUNT);
 
-	DPRINTK("mount done status=%d", status);
+		DPRINTK("mount done status=%d", status);
 
-	/* Update expiry counter */
-	ino->last_used = jiffies;
+		/* Turn this into a real negative dentry? */
+		if (status == -ENOENT) {
+			spin_lock(&sbi->fs_lock);
+			ino->flags &= ~AUTOFS_INF_PENDING;
+			spin_unlock(&sbi->fs_lock);
+			return status;
+		} else if (status) {
+			/* Return a negative dentry, but leave it "pending" */
+			return status;
+		}
+	/* Trigger mount for path component or follow link */
+	} else if (ino->flags & AUTOFS_INF_PENDING ||
+			autofs4_need_mount(flags) ||
+			current->link_count) {
+		DPRINTK("waiting for mount name=%.*s",
+			dentry->d_name.len, dentry->d_name.name);
 
-	return status;
+		spin_lock(&sbi->fs_lock);
+		ino->flags |= AUTOFS_INF_PENDING;
+		spin_unlock(&sbi->fs_lock);
+		status = autofs4_wait(sbi, dentry, NFY_MOUNT);
+
+		DPRINTK("mount done status=%d", status);
+
+		if (status) {
+			spin_lock(&sbi->fs_lock);
+			ino->flags &= ~AUTOFS_INF_PENDING;
+			spin_unlock(&sbi->fs_lock);
+			return status;
+		}
+	}
+
+	/* Initialize expiry counter after successful mount */
+	if (ino)
+		ino->last_used = jiffies;
+
+	spin_lock(&sbi->fs_lock);
+	ino->flags &= ~AUTOFS_INF_PENDING;
+	spin_unlock(&sbi->fs_lock);
+
+	return 0;
 }
 
 /* For autofs direct mounts the follow link triggers the mount */
@@ -313,16 +258,10 @@ static void *autofs4_follow_link(struct dentry *dentry, struct nameidata *nd)
 	 */
 	if (ino->flags & AUTOFS_INF_PENDING ||
 	    (!d_mountpoint(dentry) && list_empty(&dentry->d_subdirs))) {
-		ino->flags |= AUTOFS_INF_PENDING;
 		spin_unlock(&dcache_lock);
 		spin_unlock(&sbi->fs_lock);
 
-		status = try_to_fill_dentry(dentry);
-
-		spin_lock(&sbi->fs_lock);
-		ino->flags &= ~AUTOFS_INF_PENDING;
-		spin_unlock(&sbi->fs_lock);
-
+		status = try_to_fill_dentry(dentry, 0);
 		if (status)
 			goto out_error;
 
@@ -361,47 +300,18 @@ static int autofs4_revalidate(struct dentry *dentry, struct nameidata *nd)
 {
 	struct inode *dir = dentry->d_parent->d_inode;
 	struct autofs_sb_info *sbi = autofs4_sbi(dir->i_sb);
-	struct autofs_info *ino = autofs4_dentry_ino(dentry);
-	struct rehash_entry *entry;
+	int oz_mode = autofs4_oz_mode(sbi);
 	int flags = nd ? nd->flags : 0;
-	unsigned int mutex_aquired;
+	int status = 1;
 
-	DPRINTK("name = %.*s oz_mode = %d",
-		dentry->d_name.len, dentry->d_name.name, oz_mode);
-
-	/* Daemon never causes a mount to trigger */
-	if (autofs4_oz_mode(sbi))
-		return 1;
-
-	entry = kmalloc(sizeof(struct rehash_entry), GFP_KERNEL);
-	if (!entry)
-		return -ENOMEM;
-
-	mutex_aquired = mutex_trylock(&dir->i_mutex);
-
-	spin_lock(&sbi->fs_lock);
-	spin_lock(&dcache_lock);
 	/* Pending dentry */
+	spin_lock(&sbi->fs_lock);
 	if (autofs4_ispending(dentry)) {
-		int status;
-
-		/*
-		 * We can only unhash and send this to ->lookup() if
-		 * the directory mutex is held over d_revalidate() and
-		 * ->lookup(). This prevents the VFS from incorrectly
-		 * seeing the dentry as non-existent.
-		 */
-		ino->flags |= AUTOFS_INF_PENDING;
-		if (!mutex_aquired) {
-			autofs4_revalidate_drop(dentry, entry);
-			spin_unlock(&dcache_lock);
-			spin_unlock(&sbi->fs_lock);
-			return 0;
-		}
-		spin_unlock(&dcache_lock);
+		/* The daemon never causes a mount to trigger */
 		spin_unlock(&sbi->fs_lock);
-		mutex_unlock(&dir->i_mutex);
-		kfree(entry);
+
+		if (oz_mode)
+			return 1;
 
 		/*
 		 * If the directory has gone away due to an expire
@@ -415,82 +325,45 @@ static int autofs4_revalidate(struct dentry *dentry, struct nameidata *nd)
 		 * A zero status is success otherwise we have a
 		 * negative error code.
 		 */
-		status = try_to_fill_dentry(dentry);
-
-		spin_lock(&sbi->fs_lock);
-		ino->flags &= ~AUTOFS_INF_PENDING;
-		spin_unlock(&sbi->fs_lock);
-
+		status = try_to_fill_dentry(dentry, flags);
 		if (status == 0)
 			return 1;
 
 		return status;
 	}
+	spin_unlock(&sbi->fs_lock);
+
+	/* Negative dentry.. invalidate if "old" */
+	if (dentry->d_inode == NULL)
+		return 0;
 
 	/* Check for a non-mountpoint directory with no contents */
+	spin_lock(&dcache_lock);
 	if (S_ISDIR(dentry->d_inode->i_mode) &&
 	    !d_mountpoint(dentry) && list_empty(&dentry->d_subdirs)) {
 		DPRINTK("dentry=%p %.*s, emptydir",
 			 dentry, dentry->d_name.len, dentry->d_name.name);
+		spin_unlock(&dcache_lock);
 
-		if (autofs4_need_mount(flags) || current->link_count) {
-			int status;
-
-			/*
-			 * We can only unhash and send this to ->lookup() if
-			 * the directory mutex is held over d_revalidate() and
-			 * ->lookup(). This prevents the VFS from incorrectly
-			 * seeing the dentry as non-existent.
-			 */
-			ino->flags |= AUTOFS_INF_PENDING;
-			if (!mutex_aquired) {
-				autofs4_revalidate_drop(dentry, entry);
-				spin_unlock(&dcache_lock);
-				spin_unlock(&sbi->fs_lock);
-				return 0;
-			}
-			spin_unlock(&dcache_lock);
-			spin_unlock(&sbi->fs_lock);
-			mutex_unlock(&dir->i_mutex);
-			kfree(entry);
-
-			/*
-			 * A zero status is success otherwise we have a
-			 * negative error code.
-			 */
-			status = try_to_fill_dentry(dentry);
-
-			spin_lock(&sbi->fs_lock);
-			ino->flags &= ~AUTOFS_INF_PENDING;
-			spin_unlock(&sbi->fs_lock);
+		/* The daemon never causes a mount to trigger */
+		if (oz_mode)
+			return 1;
 
-			if (status == 0)
-				return 1;
+		/*
+		 * A zero status is success otherwise we have a
+		 * negative error code.
+		 */
+		status = try_to_fill_dentry(dentry, flags);
+		if (status == 0)
+			return 1;
 
-			return status;
-		}
+		return status;
 	}
 	spin_unlock(&dcache_lock);
-	spin_unlock(&sbi->fs_lock);
-
-	if (mutex_aquired)
-		mutex_unlock(&dir->i_mutex);
-
-	kfree(entry);
 
 	return 1;
 }
 
-static void autofs4_free_rehash_entrys(struct autofs_info *inf)
-{
-	struct list_head *head = &inf->rehash_list;
-	struct rehash_entry *entry, *next;
-	list_for_each_entry_safe(entry, next, head, list) {
-		list_del(&entry->list);
-		kfree(entry);
-	}
-}
-
 void autofs4_dentry_release(struct dentry *de)
 {
 	struct autofs_info *inf;
@@ -509,8 +382,6 @@ void autofs4_dentry_release(struct dentry *de)
 				list_del(&inf->active);
 			if (!list_empty(&inf->expiring))
 				list_del(&inf->expiring);
-			if (!list_empty(&inf->rehash_list))
-				autofs4_free_rehash_entrys(inf);
 			spin_unlock(&sbi->lookup_lock);
 		}
 
@@ -543,7 +414,6 @@ static struct dentry *autofs4_lookup_active(struct dentry *dentry)
 	const unsigned char *str = name->name;
 	struct list_head *p, *head;
 
-restart:
 	spin_lock(&dcache_lock);
 	spin_lock(&sbi->lookup_lock);
 	head = &sbi->active_list;
@@ -561,19 +431,6 @@ restart:
 		if (atomic_read(&active->d_count) == 0)
 			goto next;
 
-		if (active->d_inode && IS_DEADDIR(active->d_inode)) {
-			if (!list_empty(&ino->rehash_list)) {
-				dget(active);
-				spin_unlock(&active->d_lock);
-				spin_unlock(&sbi->lookup_lock);
-				spin_unlock(&dcache_lock);
-				autofs4_remove_rehash_entrys(ino);
-				dput(active);
-				goto restart;
-			}
-			goto next;
-		}
-
 		qstr = &active->d_name;
 
 		if (active->d_name.hash != hash)
@@ -586,11 +443,13 @@ restart:
 		if (memcmp(qstr->name, str, len))
 			goto next;
 
-		dget(active);
-		spin_unlock(&active->d_lock);
-		spin_unlock(&sbi->lookup_lock);
-		spin_unlock(&dcache_lock);
-		return active;
+		if (d_unhashed(active)) {
+			dget(active);
+			spin_unlock(&active->d_lock);
+			spin_unlock(&sbi->lookup_lock);
+			spin_unlock(&dcache_lock);
+			return active;
+		}
 next:
 		spin_unlock(&active->d_lock);
 	}
@@ -639,11 +498,13 @@ static struct dentry *autofs4_lookup_expiring(struct dentry *dentry)
 		if (memcmp(qstr->name, str, len))
 			goto next;
 
-		dget(expiring);
-		spin_unlock(&expiring->d_lock);
-		spin_unlock(&sbi->lookup_lock);
-		spin_unlock(&dcache_lock);
-		return expiring;
+		if (d_unhashed(expiring)) {
+			dget(expiring);
+			spin_unlock(&expiring->d_lock);
+			spin_unlock(&sbi->lookup_lock);
+			spin_unlock(&dcache_lock);
+			return expiring;
+		}
 next:
 		spin_unlock(&expiring->d_lock);
 	}
@@ -653,48 +514,6 @@ next:
 	return NULL;
 }
 
-static struct autofs_info *init_new_dentry(struct autofs_sb_info *sbi,
-					   struct dentry *dentry, int oz_mode)
-{
-	struct autofs_info *ino;
-
-	/*
-	 * Mark the dentry incomplete but don't hash it. We do this
-	 * to serialize our inode creation operations (symlink and
-	 * mkdir) which prevents deadlock during the callback to
-	 * the daemon. Subsequent user space lookups for the same
-	 * dentry are placed on the wait queue while the daemon
-	 * itself is allowed passage unresticted so the create
-	 * operation itself can then hash the dentry. Finally,
-	 * we check for the hashed dentry and return the newly
-	 * hashed dentry.
-	 */
-	dentry->d_op = &autofs4_root_dentry_operations;
-
-	/*
-	 * And we need to ensure that the same dentry is used for
-	 * all following lookup calls until it is hashed so that
-	 * the dentry flags are persistent throughout the request.
-	 */
-	ino = autofs4_init_ino(NULL, sbi, 0555);
-	if (!ino)
-		return ERR_PTR(-ENOMEM);
-
-	dentry->d_fsdata = ino;
-	ino->dentry = dentry;
-
-	/*
-	 * Only set the mount pending flag for new dentrys not created
-	 * by the daemon.
-	 */
-	if (!oz_mode)
-		ino->flags |= AUTOFS_INF_PENDING;
-
-	d_instantiate(dentry, NULL);
-
-	return ino;
-}
-
 /* Lookups in the root directory */
 static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
 {
@@ -702,7 +521,6 @@ static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, s
 	struct autofs_info *ino;
 	struct dentry *expiring, *active;
 	int oz_mode;
-	int status = 0;
 
 	DPRINTK("name = %.*s",
 		dentry->d_name.len, dentry->d_name.name);
@@ -717,26 +535,44 @@ static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, s
 	DPRINTK("pid = %u, pgrp = %u, catatonic = %d, oz_mode = %d",
 		 current->pid, task_pgrp_nr(current), sbi->catatonic, oz_mode);
 
-	spin_lock(&sbi->fs_lock);
 	active = autofs4_lookup_active(dentry);
 	if (active) {
 		dentry = active;
 		ino = autofs4_dentry_ino(dentry);
-		/* If this came from revalidate, rehash it */
-		autofs4_revalidate_rehash(dentry);
-		spin_unlock(&sbi->fs_lock);
 	} else {
-		spin_unlock(&sbi->fs_lock);
-		ino = init_new_dentry(sbi, dentry, oz_mode);
-		if (IS_ERR(ino))
-			return (struct dentry *) ino;
-	}
+		/*
+		 * Mark the dentry incomplete but don't hash it. We do this
+		 * to serialize our inode creation operations (symlink and
+		 * mkdir) which prevents deadlock during the callback to
+		 * the daemon. Subsequent user space lookups for the same
+		 * dentry are placed on the wait queue while the daemon
+		 * itself is allowed passage unresticted so the create
+		 * operation itself can then hash the dentry. Finally,
+		 * we check for the hashed dentry and return the newly
+		 * hashed dentry.
+		 */
+		dentry->d_op = &autofs4_root_dentry_operations;
+
+		/*
+		 * And we need to ensure that the same dentry is used for
+		 * all following lookup calls until it is hashed so that
+		 * the dentry flags are persistent throughout the request.
+		 */
+		ino = autofs4_init_ino(NULL, sbi, 0555);
+		if (!ino)
+			return ERR_PTR(-ENOMEM);
 
-	autofs4_add_active(dentry);
+		dentry->d_fsdata = ino;
+		ino->dentry = dentry;
+
+		autofs4_add_active(dentry);
+
+		d_instantiate(dentry, NULL);
+	}
 
 	if (!oz_mode) {
-		expiring = autofs4_lookup_expiring(dentry);
 		mutex_unlock(&dir->i_mutex);
+		expiring = autofs4_lookup_expiring(dentry);
 		if (expiring) {
 			/*
 			 * If we are racing with expire the request might not
@@ -744,22 +580,23 @@ static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, s
 			 * so it must have been successful, so just wait for it.
 			 */
 			autofs4_expire_wait(expiring);
+			autofs4_del_expiring(expiring);
 			dput(expiring);
 		}
-		status = try_to_fill_dentry(dentry);
-		mutex_lock(&dir->i_mutex);
+
 		spin_lock(&sbi->fs_lock);
-		ino->flags &= ~AUTOFS_INF_PENDING;
+		ino->flags |= AUTOFS_INF_PENDING;
 		spin_unlock(&sbi->fs_lock);
+		if (dentry->d_op && dentry->d_op->d_revalidate)
+			(dentry->d_op->d_revalidate)(dentry, nd);
+		mutex_lock(&dir->i_mutex);
 	}
 
-	autofs4_del_active(dentry);
-
 	/*
-	 * If we had a mount fail, check if we had to handle
+	 * If we are still pending, check if we had to handle
 	 * a signal. If so we can force a restart..
 	 */
-	if (status) {
+	if (ino->flags & AUTOFS_INF_PENDING) {
 		/* See if we were interrupted */
 		if (signal_pending(current)) {
 			sigset_t *sigset = &current->pending.signal;
@@ -771,46 +608,43 @@ static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, s
 			    return ERR_PTR(-ERESTARTNOINTR);
 			}
 		}
-	}
-
-	/*
-	 * User space can (and has done in the past) remove and re-create
-	 * this directory during the callback. This can leave us with an
-	 * unhashed dentry, but a successful mount!  So we need to
-	 * perform another cached lookup in case the dentry now exists.
-	 */
-	if (!oz_mode && !have_submounts(dentry)) {
-		struct dentry *new;
-		new = d_lookup(dentry->d_parent, &dentry->d_name);
-		if (new) {
-			if (active)
-				dput(active);
-			return new;
-		} else {
-			if (!status)
-				status = -ENOENT;
+		if (!oz_mode) {
+			spin_lock(&sbi->fs_lock);
+			ino->flags &= ~AUTOFS_INF_PENDING;
+			spin_unlock(&sbi->fs_lock);
 		}
 	}
 
 	/*
-	 * If we had a mount failure, return status to user space.
-	 * If the mount succeeded and we used a dentry from the active queue
-	 * return it.
+	 * If this dentry is unhashed, then we shouldn't honour this
+	 * lookup.  Returning ENOENT here doesn't do the right thing
+	 * for all system calls, but it should be OK for the operations
+	 * we permit from an autofs.
 	 */
-	if (status) {
-		dentry = ERR_PTR(status);
-		if (active)
-			dput(active);
-		return dentry;
-	} else {
+	if (!oz_mode && d_unhashed(dentry)) {
 		/*
-		 * Valid successful mount, return active dentry or NULL
-		 * for a new dentry.
+		 * A user space application can (and has done in the past)
+		 * remove and re-create this directory during the callback.
+		 * This can leave us with an unhashed dentry, but a
+		 * successful mount!  So we need to perform another
+		 * cached lookup in case the dentry now exists.
 		 */
+		struct dentry *parent = dentry->d_parent;
+		struct dentry *new = d_lookup(parent, &dentry->d_name);
+		if (new != NULL)
+			dentry = new;
+		else
+			dentry = ERR_PTR(-ENOENT);
+
 		if (active)
-			return active;
+			dput(active);
+
+		return dentry;
 	}
 
+	if (active)
+		return active;
+
 	return NULL;
 }
 
@@ -834,6 +668,8 @@ static int autofs4_dir_symlink(struct inode *dir,
 	if (!ino)
 		return -ENOMEM;
 
+	autofs4_del_active(dentry);
+
 	ino->size = strlen(symname);
 	cp = kmalloc(ino->size + 1, GFP_KERNEL);
 	if (!cp) {
@@ -910,6 +746,7 @@ static int autofs4_dir_unlink(struct inode *dir, struct dentry *dentry)
 	dir->i_mtime = CURRENT_TIME;
 
 	spin_lock(&dcache_lock);
+	autofs4_add_expiring(dentry);
 	spin_lock(&dentry->d_lock);
 	__d_drop(dentry);
 	spin_unlock(&dentry->d_lock);
@@ -935,6 +772,7 @@ static int autofs4_dir_rmdir(struct inode *dir, struct dentry *dentry)
 		spin_unlock(&dcache_lock);
 		return -ENOTEMPTY;
 	}
+	autofs4_add_expiring(dentry);
 	spin_lock(&dentry->d_lock);
 	__d_drop(dentry);
 	spin_unlock(&dentry->d_lock);
@@ -972,6 +810,8 @@ static int autofs4_dir_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	if (!ino)
 		return -ENOMEM;
 
+	autofs4_del_active(dentry);
+
 	inode = autofs4_get_inode(dir->i_sb, ino);
 	if (!inode) {
 		if (!dentry->d_fsdata)
diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c
index 8f3d9fd89604..f22a7d3dc362 100644
--- a/fs/bfs/inode.c
+++ b/fs/bfs/inode.c
@@ -15,6 +15,7 @@
 #include <linux/smp_lock.h>
 #include <linux/buffer_head.h>
 #include <linux/vfs.h>
+#include <linux/writeback.h>
 #include <asm/uaccess.h>
 #include "bfs.h"
 
@@ -98,7 +99,7 @@ error:
 	return ERR_PTR(-EIO);
 }
 
-static int bfs_write_inode(struct inode *inode, int wait)
+static int bfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct bfs_sb_info *info = BFS_SB(inode->i_sb);
 	unsigned int ino = (u16)inode->i_ino;
@@ -147,7 +148,7 @@ static int bfs_write_inode(struct inode *inode, int wait)
 	di->i_eoffset = cpu_to_le32(i_sblock * BFS_BSIZE + inode->i_size - 1);
 
 	mark_buffer_dirty(bh);
-	if (wait) {
+	if (wbc->sync_mode == WB_SYNC_ALL) {
 		sync_dirty_buffer(bh);
 		if (buffer_req(bh) && !buffer_uptodate(bh))
 			err = -EIO;
diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c
index fdd397099172..15d80bb35d6f 100644
--- a/fs/binfmt_aout.c
+++ b/fs/binfmt_aout.c
@@ -24,6 +24,7 @@
 #include <linux/binfmts.h>
 #include <linux/personality.h>
 #include <linux/init.h>
+#include <linux/coredump.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
@@ -60,26 +61,6 @@ static int set_brk(unsigned long start, unsigned long end)
 }
 
 /*
- * These are the only things you should do on a core-file: use only these
- * macros to write out all the necessary info.
- */
-
-static int dump_write(struct file *file, const void *addr, int nr)
-{
-	return file->f_op->write(file, addr, nr, &file->f_pos) == nr;
-}
-
-#define DUMP_WRITE(addr, nr)	\
-	if (!dump_write(file, (void *)(addr), (nr))) \
-		goto end_coredump;
-
-#define DUMP_SEEK(offset) \
-if (file->f_op->llseek) { \
-	if (file->f_op->llseek(file,(offset),0) != (offset)) \
- 		goto end_coredump; \
-} else file->f_pos = (offset)
-
-/*
  * Routine writes a core dump image in the current directory.
  * Currently only a stub-function.
  *
@@ -130,26 +111,31 @@ static int aout_core_dump(struct coredump_params *cprm)
 
 	set_fs(KERNEL_DS);
 /* struct user */
-	DUMP_WRITE(&dump,sizeof(dump));
+	if (!dump_write(file, &dump, sizeof(dump)))
+		goto end_coredump;
 /* Now dump all of the user data.  Include malloced stuff as well */
-	DUMP_SEEK(PAGE_SIZE);
+	if (!dump_seek(cprm->file, PAGE_SIZE - sizeof(dump)))
+		goto end_coredump;
 /* now we start writing out the user space info */
 	set_fs(USER_DS);
 /* Dump the data area */
 	if (dump.u_dsize != 0) {
 		dump_start = START_DATA(dump);
 		dump_size = dump.u_dsize << PAGE_SHIFT;
-		DUMP_WRITE(dump_start,dump_size);
+		if (!dump_write(file, dump_start, dump_size))
+			goto end_coredump;
 	}
 /* Now prepare to dump the stack area */
 	if (dump.u_ssize != 0) {
 		dump_start = START_STACK(dump);
 		dump_size = dump.u_ssize << PAGE_SHIFT;
-		DUMP_WRITE(dump_start,dump_size);
+		if (!dump_write(file, dump_start, dump_size))
+			goto end_coredump;
 	}
 /* Finally dump the task struct.  Not be used by gdb, but could be useful */
 	set_fs(KERNEL_DS);
-	DUMP_WRITE(current,sizeof(*current));
+	if (!dump_write(file, current, sizeof(*current)))
+		goto end_coredump;
 end_coredump:
 	set_fs(fs);
 	return has_dumped;
@@ -247,7 +233,7 @@ static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs)
 	 * size limits imposed on them by creating programs with large
 	 * arrays in the data or bss.
 	 */
-	rlim = current->signal->rlim[RLIMIT_DATA].rlim_cur;
+	rlim = rlimit(RLIMIT_DATA);
 	if (rlim >= RLIM_INFINITY)
 		rlim = ~0;
 	if (ex.a_data + ex.a_bss > rlim)
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index fd5b2ea5d299..535e763ab1a6 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -31,6 +31,7 @@
 #include <linux/random.h>
 #include <linux/elf.h>
 #include <linux/utsname.h>
+#include <linux/coredump.h>
 #include <asm/uaccess.h>
 #include <asm/param.h>
 #include <asm/page.h>
@@ -1085,36 +1086,6 @@ out:
  * Modelled on fs/exec.c:aout_core_dump()
  * Jeremy Fitzhardinge <jeremy@sw.oz.au>
  */
-/*
- * These are the only things you should do on a core-file: use only these
- * functions to write out all the necessary info.
- */
-static int dump_write(struct file *file, const void *addr, int nr)
-{
-	return file->f_op->write(file, addr, nr, &file->f_pos) == nr;
-}
-
-static int dump_seek(struct file *file, loff_t off)
-{
-	if (file->f_op->llseek && file->f_op->llseek != no_llseek) {
-		if (file->f_op->llseek(file, off, SEEK_CUR) < 0)
-			return 0;
-	} else {
-		char *buf = (char *)get_zeroed_page(GFP_KERNEL);
-		if (!buf)
-			return 0;
-		while (off > 0) {
-			unsigned long n = off;
-			if (n > PAGE_SIZE)
-				n = PAGE_SIZE;
-			if (!dump_write(file, buf, n))
-				return 0;
-			off -= n;
-		}
-		free_page((unsigned long)buf);
-	}
-	return 1;
-}
 
 /*
  * Decide what to dump of a segment, part, all or none.
@@ -1249,11 +1220,6 @@ static int writenote(struct memelfnote *men, struct file *file,
 }
 #undef DUMP_WRITE
 
-#define DUMP_WRITE(addr, nr)				\
-	if ((size += (nr)) > cprm->limit ||		\
-	    !dump_write(cprm->file, (addr), (nr)))	\
-		goto end_coredump;
-
 static void fill_elf_header(struct elfhdr *elf, int segs,
 			    u16 machine, u32 flags, u8 osabi)
 {
@@ -1872,6 +1838,34 @@ static struct vm_area_struct *next_vma(struct vm_area_struct *this_vma,
 	return gate_vma;
 }
 
+static void fill_extnum_info(struct elfhdr *elf, struct elf_shdr *shdr4extnum,
+			     elf_addr_t e_shoff, int segs)
+{
+	elf->e_shoff = e_shoff;
+	elf->e_shentsize = sizeof(*shdr4extnum);
+	elf->e_shnum = 1;
+	elf->e_shstrndx = SHN_UNDEF;
+
+	memset(shdr4extnum, 0, sizeof(*shdr4extnum));
+
+	shdr4extnum->sh_type = SHT_NULL;
+	shdr4extnum->sh_size = elf->e_shnum;
+	shdr4extnum->sh_link = elf->e_shstrndx;
+	shdr4extnum->sh_info = segs;
+}
+
+static size_t elf_core_vma_data_size(struct vm_area_struct *gate_vma,
+				     unsigned long mm_flags)
+{
+	struct vm_area_struct *vma;
+	size_t size = 0;
+
+	for (vma = first_vma(current, gate_vma); vma != NULL;
+	     vma = next_vma(vma, gate_vma))
+		size += vma_dump_size(vma, mm_flags);
+	return size;
+}
+
 /*
  * Actual dumper
  *
@@ -1888,8 +1882,11 @@ static int elf_core_dump(struct coredump_params *cprm)
 	struct vm_area_struct *vma, *gate_vma;
 	struct elfhdr *elf = NULL;
 	loff_t offset = 0, dataoff, foffset;
-	unsigned long mm_flags;
 	struct elf_note_info info;
+	struct elf_phdr *phdr4note = NULL;
+	struct elf_shdr *shdr4extnum = NULL;
+	Elf_Half e_phnum;
+	elf_addr_t e_shoff;
 
 	/*
 	 * We no longer stop all VM operations.
@@ -1912,20 +1909,25 @@ static int elf_core_dump(struct coredump_params *cprm)
 	 * Please check DEFAULT_MAX_MAP_COUNT definition when you modify here.
 	 */
 	segs = current->mm->map_count;
-#ifdef ELF_CORE_EXTRA_PHDRS
-	segs += ELF_CORE_EXTRA_PHDRS;
-#endif
+	segs += elf_core_extra_phdrs();
 
 	gate_vma = get_gate_vma(current);
 	if (gate_vma != NULL)
 		segs++;
 
+	/* for notes section */
+	segs++;
+
+	/* If segs > PN_XNUM(0xffff), then e_phnum overflows. To avoid
+	 * this, kernel supports extended numbering. Have a look at
+	 * include/linux/elf.h for further information. */
+	e_phnum = segs > PN_XNUM ? PN_XNUM : segs;
+
 	/*
 	 * Collect all the non-memory information about the process for the
 	 * notes.  This also sets up the file header.
 	 */
-	if (!fill_note_info(elf, segs + 1, /* including notes section */
-			    &info, cprm->signr, cprm->regs))
+	if (!fill_note_info(elf, e_phnum, &info, cprm->signr, cprm->regs))
 		goto cleanup;
 
 	has_dumped = 1;
@@ -1934,31 +1936,47 @@ static int elf_core_dump(struct coredump_params *cprm)
 	fs = get_fs();
 	set_fs(KERNEL_DS);
 
-	DUMP_WRITE(elf, sizeof(*elf));
 	offset += sizeof(*elf);				/* Elf header */
-	offset += (segs + 1) * sizeof(struct elf_phdr); /* Program headers */
+	offset += segs * sizeof(struct elf_phdr);	/* Program headers */
 	foffset = offset;
 
 	/* Write notes phdr entry */
 	{
-		struct elf_phdr phdr;
 		size_t sz = get_note_info_size(&info);
 
 		sz += elf_coredump_extra_notes_size();
 
-		fill_elf_note_phdr(&phdr, sz, offset);
+		phdr4note = kmalloc(sizeof(*phdr4note), GFP_KERNEL);
+		if (!phdr4note)
+			goto end_coredump;
+
+		fill_elf_note_phdr(phdr4note, sz, offset);
 		offset += sz;
-		DUMP_WRITE(&phdr, sizeof(phdr));
 	}
 
 	dataoff = offset = roundup(offset, ELF_EXEC_PAGESIZE);
 
-	/*
-	 * We must use the same mm->flags while dumping core to avoid
-	 * inconsistency between the program headers and bodies, otherwise an
-	 * unusable core file can be generated.
-	 */
-	mm_flags = current->mm->flags;
+	offset += elf_core_vma_data_size(gate_vma, cprm->mm_flags);
+	offset += elf_core_extra_data_size();
+	e_shoff = offset;
+
+	if (e_phnum == PN_XNUM) {
+		shdr4extnum = kmalloc(sizeof(*shdr4extnum), GFP_KERNEL);
+		if (!shdr4extnum)
+			goto end_coredump;
+		fill_extnum_info(elf, shdr4extnum, e_shoff, segs);
+	}
+
+	offset = dataoff;
+
+	size += sizeof(*elf);
+	if (size > cprm->limit || !dump_write(cprm->file, elf, sizeof(*elf)))
+		goto end_coredump;
+
+	size += sizeof(*phdr4note);
+	if (size > cprm->limit
+	    || !dump_write(cprm->file, phdr4note, sizeof(*phdr4note)))
+		goto end_coredump;
 
 	/* Write program headers for segments dump */
 	for (vma = first_vma(current, gate_vma); vma != NULL;
@@ -1969,7 +1987,7 @@ static int elf_core_dump(struct coredump_params *cprm)
 		phdr.p_offset = offset;
 		phdr.p_vaddr = vma->vm_start;
 		phdr.p_paddr = 0;
-		phdr.p_filesz = vma_dump_size(vma, mm_flags);
+		phdr.p_filesz = vma_dump_size(vma, cprm->mm_flags);
 		phdr.p_memsz = vma->vm_end - vma->vm_start;
 		offset += phdr.p_filesz;
 		phdr.p_flags = vma->vm_flags & VM_READ ? PF_R : 0;
@@ -1979,12 +1997,14 @@ static int elf_core_dump(struct coredump_params *cprm)
 			phdr.p_flags |= PF_X;
 		phdr.p_align = ELF_EXEC_PAGESIZE;
 
-		DUMP_WRITE(&phdr, sizeof(phdr));
+		size += sizeof(phdr);
+		if (size > cprm->limit
+		    || !dump_write(cprm->file, &phdr, sizeof(phdr)))
+			goto end_coredump;
 	}
 
-#ifdef ELF_CORE_WRITE_EXTRA_PHDRS
-	ELF_CORE_WRITE_EXTRA_PHDRS;
-#endif
+	if (!elf_core_write_extra_phdrs(cprm->file, offset, &size, cprm->limit))
+		goto end_coredump;
 
  	/* write out the notes section */
 	if (!write_note_info(&info, cprm->file, &foffset))
@@ -2002,7 +2022,7 @@ static int elf_core_dump(struct coredump_params *cprm)
 		unsigned long addr;
 		unsigned long end;
 
-		end = vma->vm_start + vma_dump_size(vma, mm_flags);
+		end = vma->vm_start + vma_dump_size(vma, cprm->mm_flags);
 
 		for (addr = vma->vm_start; addr < end; addr += PAGE_SIZE) {
 			struct page *page;
@@ -2023,15 +2043,24 @@ static int elf_core_dump(struct coredump_params *cprm)
 		}
 	}
 
-#ifdef ELF_CORE_WRITE_EXTRA_DATA
-	ELF_CORE_WRITE_EXTRA_DATA;
-#endif
+	if (!elf_core_write_extra_data(cprm->file, &size, cprm->limit))
+		goto end_coredump;
+
+	if (e_phnum == PN_XNUM) {
+		size += sizeof(*shdr4extnum);
+		if (size > cprm->limit
+		    || !dump_write(cprm->file, shdr4extnum,
+				   sizeof(*shdr4extnum)))
+			goto end_coredump;
+	}
 
 end_coredump:
 	set_fs(fs);
 
 cleanup:
 	free_note_info(&info);
+	kfree(shdr4extnum);
+	kfree(phdr4note);
 	kfree(elf);
 out:
 	return has_dumped;
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index 18d77297ccc8..6d6a16c5e9bb 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -34,6 +34,7 @@
 #include <linux/elf.h>
 #include <linux/elf-fdpic.h>
 #include <linux/elfcore.h>
+#include <linux/coredump.h>
 
 #include <asm/uaccess.h>
 #include <asm/param.h>
@@ -1216,26 +1217,6 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params,
 #ifdef CONFIG_ELF_CORE
 
 /*
- * These are the only things you should do on a core-file: use only these
- * functions to write out all the necessary info.
- */
-static int dump_write(struct file *file, const void *addr, int nr)
-{
-	return file->f_op->write(file, addr, nr, &file->f_pos) == nr;
-}
-
-static int dump_seek(struct file *file, loff_t off)
-{
-	if (file->f_op->llseek) {
-		if (file->f_op->llseek(file, off, SEEK_SET) != off)
-			return 0;
-	} else {
-		file->f_pos = off;
-	}
-	return 1;
-}
-
-/*
  * Decide whether a segment is worth dumping; default is yes to be
  * sure (missing info is worse than too much; etc).
  * Personally I'd include everything, and use the coredump limit...
@@ -1313,35 +1294,35 @@ static int notesize(struct memelfnote *en)
 
 /* #define DEBUG */
 
-#define DUMP_WRITE(addr, nr)	\
-	do { if (!dump_write(file, (addr), (nr))) return 0; } while(0)
-#define DUMP_SEEK(off)	\
-	do { if (!dump_seek(file, (off))) return 0; } while(0)
+#define DUMP_WRITE(addr, nr, foffset)	\
+	do { if (!dump_write(file, (addr), (nr))) return 0; *foffset += (nr); } while(0)
 
-static int writenote(struct memelfnote *men, struct file *file)
+static int alignfile(struct file *file, loff_t *foffset)
 {
-	struct elf_note en;
+	static const char buf[4] = { 0, };
+	DUMP_WRITE(buf, roundup(*foffset, 4) - *foffset, foffset);
+	return 1;
+}
 
+static int writenote(struct memelfnote *men, struct file *file,
+			loff_t *foffset)
+{
+	struct elf_note en;
 	en.n_namesz = strlen(men->name) + 1;
 	en.n_descsz = men->datasz;
 	en.n_type = men->type;
 
-	DUMP_WRITE(&en, sizeof(en));
-	DUMP_WRITE(men->name, en.n_namesz);
-	/* XXX - cast from long long to long to avoid need for libgcc.a */
-	DUMP_SEEK(roundup((unsigned long)file->f_pos, 4));	/* XXX */
-	DUMP_WRITE(men->data, men->datasz);
-	DUMP_SEEK(roundup((unsigned long)file->f_pos, 4));	/* XXX */
+	DUMP_WRITE(&en, sizeof(en), foffset);
+	DUMP_WRITE(men->name, en.n_namesz, foffset);
+	if (!alignfile(file, foffset))
+		return 0;
+	DUMP_WRITE(men->data, men->datasz, foffset);
+	if (!alignfile(file, foffset))
+		return 0;
 
 	return 1;
 }
 #undef DUMP_WRITE
-#undef DUMP_SEEK
-
-#define DUMP_WRITE(addr, nr)				\
-	if ((size += (nr)) > cprm->limit ||		\
-	    !dump_write(cprm->file, (addr), (nr)))	\
-		goto end_coredump;
 
 static inline void fill_elf_fdpic_header(struct elfhdr *elf, int segs)
 {
@@ -1524,6 +1505,22 @@ static int elf_dump_thread_status(long signr, struct elf_thread_status *t)
 	return sz;
 }
 
+static void fill_extnum_info(struct elfhdr *elf, struct elf_shdr *shdr4extnum,
+			     elf_addr_t e_shoff, int segs)
+{
+	elf->e_shoff = e_shoff;
+	elf->e_shentsize = sizeof(*shdr4extnum);
+	elf->e_shnum = 1;
+	elf->e_shstrndx = SHN_UNDEF;
+
+	memset(shdr4extnum, 0, sizeof(*shdr4extnum));
+
+	shdr4extnum->sh_type = SHT_NULL;
+	shdr4extnum->sh_size = elf->e_shnum;
+	shdr4extnum->sh_link = elf->e_shstrndx;
+	shdr4extnum->sh_info = segs;
+}
+
 /*
  * dump the segments for an MMU process
  */
@@ -1552,7 +1549,7 @@ static int elf_fdpic_dump_segments(struct file *file, size_t *size,
 					err = -EIO;
 				kunmap(page);
 				page_cache_release(page);
-			} else if (!dump_seek(file, file->f_pos + PAGE_SIZE))
+			} else if (!dump_seek(file, PAGE_SIZE))
 				err = -EFBIG;
 			if (err)
 				goto out;
@@ -1588,6 +1585,17 @@ static int elf_fdpic_dump_segments(struct file *file, size_t *size,
 }
 #endif
 
+static size_t elf_core_vma_data_size(unsigned long mm_flags)
+{
+	struct vm_area_struct *vma;
+	size_t size = 0;
+
+	for (vma = current->mm->mmap; vma; vma->vm_next)
+		if (maydump(vma, mm_flags))
+			size += vma->vm_end - vma->vm_start;
+	return size;
+}
+
 /*
  * Actual dumper
  *
@@ -1605,7 +1613,7 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 	int i;
 	struct vm_area_struct *vma;
 	struct elfhdr *elf = NULL;
-	loff_t offset = 0, dataoff;
+	loff_t offset = 0, dataoff, foffset;
 	int numnote;
 	struct memelfnote *notes = NULL;
 	struct elf_prstatus *prstatus = NULL;	/* NT_PRSTATUS */
@@ -1618,7 +1626,10 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 #endif
 	int thread_status_size = 0;
 	elf_addr_t *auxv;
-	unsigned long mm_flags;
+	struct elf_phdr *phdr4note = NULL;
+	struct elf_shdr *shdr4extnum = NULL;
+	Elf_Half e_phnum;
+	elf_addr_t e_shoff;
 
 	/*
 	 * We no longer stop all VM operations.
@@ -1683,12 +1694,18 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 	elf_core_copy_regs(&prstatus->pr_reg, cprm->regs);
 
 	segs = current->mm->map_count;
-#ifdef ELF_CORE_EXTRA_PHDRS
-	segs += ELF_CORE_EXTRA_PHDRS;
-#endif
+	segs += elf_core_extra_phdrs();
+
+	/* for notes section */
+	segs++;
+
+	/* If segs > PN_XNUM(0xffff), then e_phnum overflows. To avoid
+	 * this, kernel supports extended numbering. Have a look at
+	 * include/linux/elf.h for further information. */
+	e_phnum = segs > PN_XNUM ? PN_XNUM : segs;
 
 	/* Set up header */
-	fill_elf_fdpic_header(elf, segs + 1);	/* including notes section */
+	fill_elf_fdpic_header(elf, e_phnum);
 
 	has_dumped = 1;
 	current->flags |= PF_DUMPCORE;
@@ -1727,13 +1744,12 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 	fs = get_fs();
 	set_fs(KERNEL_DS);
 
-	DUMP_WRITE(elf, sizeof(*elf));
 	offset += sizeof(*elf);				/* Elf header */
-	offset += (segs+1) * sizeof(struct elf_phdr);	/* Program headers */
+	offset += segs * sizeof(struct elf_phdr);	/* Program headers */
+	foffset = offset;
 
 	/* Write notes phdr entry */
 	{
-		struct elf_phdr phdr;
 		int sz = 0;
 
 		for (i = 0; i < numnote; i++)
@@ -1741,20 +1757,38 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 
 		sz += thread_status_size;
 
-		fill_elf_note_phdr(&phdr, sz, offset);
+		phdr4note = kmalloc(sizeof(*phdr4note), GFP_KERNEL);
+		if (!phdr4note)
+			goto end_coredump;
+
+		fill_elf_note_phdr(phdr4note, sz, offset);
 		offset += sz;
-		DUMP_WRITE(&phdr, sizeof(phdr));
 	}
 
 	/* Page-align dumped data */
 	dataoff = offset = roundup(offset, ELF_EXEC_PAGESIZE);
 
-	/*
-	 * We must use the same mm->flags while dumping core to avoid
-	 * inconsistency between the program headers and bodies, otherwise an
-	 * unusable core file can be generated.
-	 */
-	mm_flags = current->mm->flags;
+	offset += elf_core_vma_data_size(cprm->mm_flags);
+	offset += elf_core_extra_data_size();
+	e_shoff = offset;
+
+	if (e_phnum == PN_XNUM) {
+		shdr4extnum = kmalloc(sizeof(*shdr4extnum), GFP_KERNEL);
+		if (!shdr4extnum)
+			goto end_coredump;
+		fill_extnum_info(elf, shdr4extnum, e_shoff, segs);
+	}
+
+	offset = dataoff;
+
+	size += sizeof(*elf);
+	if (size > cprm->limit || !dump_write(cprm->file, elf, sizeof(*elf)))
+		goto end_coredump;
+
+	size += sizeof(*phdr4note);
+	if (size > cprm->limit
+	    || !dump_write(cprm->file, phdr4note, sizeof(*phdr4note)))
+		goto end_coredump;
 
 	/* write program headers for segments dump */
 	for (vma = current->mm->mmap; vma; vma = vma->vm_next) {
@@ -1767,7 +1801,7 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 		phdr.p_offset = offset;
 		phdr.p_vaddr = vma->vm_start;
 		phdr.p_paddr = 0;
-		phdr.p_filesz = maydump(vma, mm_flags) ? sz : 0;
+		phdr.p_filesz = maydump(vma, cprm->mm_flags) ? sz : 0;
 		phdr.p_memsz = sz;
 		offset += phdr.p_filesz;
 		phdr.p_flags = vma->vm_flags & VM_READ ? PF_R : 0;
@@ -1777,16 +1811,18 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 			phdr.p_flags |= PF_X;
 		phdr.p_align = ELF_EXEC_PAGESIZE;
 
-		DUMP_WRITE(&phdr, sizeof(phdr));
+		size += sizeof(phdr);
+		if (size > cprm->limit
+		    || !dump_write(cprm->file, &phdr, sizeof(phdr)))
+			goto end_coredump;
 	}
 
-#ifdef ELF_CORE_WRITE_EXTRA_PHDRS
-	ELF_CORE_WRITE_EXTRA_PHDRS;
-#endif
+	if (!elf_core_write_extra_phdrs(cprm->file, offset, &size, cprm->limit))
+		goto end_coredump;
 
  	/* write out the notes section */
 	for (i = 0; i < numnote; i++)
-		if (!writenote(notes + i, cprm->file))
+		if (!writenote(notes + i, cprm->file, &foffset))
 			goto end_coredump;
 
 	/* write out the thread status notes section */
@@ -1795,20 +1831,27 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 				list_entry(t, struct elf_thread_status, list);
 
 		for (i = 0; i < tmp->num_notes; i++)
-			if (!writenote(&tmp->notes[i], cprm->file))
+			if (!writenote(&tmp->notes[i], cprm->file, &foffset))
 				goto end_coredump;
 	}
 
-	if (!dump_seek(cprm->file, dataoff))
+	if (!dump_seek(cprm->file, dataoff - foffset))
 		goto end_coredump;
 
 	if (elf_fdpic_dump_segments(cprm->file, &size, &cprm->limit,
-				    mm_flags) < 0)
+				    cprm->mm_flags) < 0)
 		goto end_coredump;
 
-#ifdef ELF_CORE_WRITE_EXTRA_DATA
-	ELF_CORE_WRITE_EXTRA_DATA;
-#endif
+	if (!elf_core_write_extra_data(cprm->file, &size, cprm->limit))
+		goto end_coredump;
+
+	if (e_phnum == PN_XNUM) {
+		size += sizeof(*shdr4extnum);
+		if (size > cprm->limit
+		    || !dump_write(cprm->file, shdr4extnum,
+				   sizeof(*shdr4extnum)))
+			goto end_coredump;
+	}
 
 	if (cprm->file->f_pos != offset) {
 		/* Sanity check */
@@ -1826,7 +1869,7 @@ cleanup:
 		list_del(tmp);
 		kfree(list_entry(tmp, struct elf_thread_status, list));
 	}
-
+	kfree(phdr4note);
 	kfree(elf);
 	kfree(prstatus);
 	kfree(psinfo);
diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c
index 42c6b4a54445..e0e769bdca59 100644
--- a/fs/binfmt_flat.c
+++ b/fs/binfmt_flat.c
@@ -501,7 +501,7 @@ static int load_flat_file(struct linux_binprm * bprm,
 	 * size limits imposed on them by creating programs with large
 	 * arrays in the data or bss.
 	 */
-	rlim = current->signal->rlim[RLIMIT_DATA].rlim_cur;
+	rlim = rlimit(RLIMIT_DATA);
 	if (rlim >= RLIM_INFINITY)
 		rlim = ~0;
 	if (data_len + bss_len > rlim) {
diff --git a/fs/bio.c b/fs/bio.c
index 88094afc29ea..dc17afd672e3 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -507,10 +507,8 @@ int bio_get_nr_vecs(struct block_device *bdev)
 	int nr_pages;
 
 	nr_pages = ((queue_max_sectors(q) << 9) + PAGE_SIZE - 1) >> PAGE_SHIFT;
-	if (nr_pages > queue_max_phys_segments(q))
-		nr_pages = queue_max_phys_segments(q);
-	if (nr_pages > queue_max_hw_segments(q))
-		nr_pages = queue_max_hw_segments(q);
+	if (nr_pages > queue_max_segments(q))
+		nr_pages = queue_max_segments(q);
 
 	return nr_pages;
 }
@@ -575,8 +573,7 @@ static int __bio_add_page(struct request_queue *q, struct bio *bio, struct page
 	 * make this too complex.
 	 */
 
-	while (bio->bi_phys_segments >= queue_max_phys_segments(q)
-	       || bio->bi_phys_segments >= queue_max_hw_segments(q)) {
+	while (bio->bi_phys_segments >= queue_max_segments(q)) {
 
 		if (retried_segments)
 			return 0;
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 2aa8ec6a0981..8b5cfdd4bfc1 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2326,7 +2326,7 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf);
 int btrfs_readpage(struct file *file, struct page *page);
 void btrfs_delete_inode(struct inode *inode);
 void btrfs_put_inode(struct inode *inode);
-int btrfs_write_inode(struct inode *inode, int wait);
+int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc);
 void btrfs_dirty_inode(struct inode *inode);
 struct inode *btrfs_alloc_inode(struct super_block *sb);
 void btrfs_destroy_inode(struct inode *inode);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 4deb280f8969..c41db6d45ab6 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -3968,7 +3968,7 @@ err:
 	return ret;
 }
 
-int btrfs_write_inode(struct inode *inode, int wait)
+int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_trans_handle *trans;
@@ -3977,7 +3977,7 @@ int btrfs_write_inode(struct inode *inode, int wait)
 	if (root->fs_info->btree_inode == inode)
 		return 0;
 
-	if (wait) {
+	if (wbc->sync_mode == WB_SYNC_ALL) {
 		trans = btrfs_join_transaction(root, 1);
 		btrfs_set_trans_block_group(trans, inode);
 		ret = btrfs_commit_transaction(trans, root);
diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c
index 14ac4806e291..eeb4986ea7db 100644
--- a/fs/cachefiles/namei.c
+++ b/fs/cachefiles/namei.c
@@ -348,7 +348,17 @@ int cachefiles_delete_object(struct cachefiles_cache *cache,
 	dir = dget_parent(object->dentry);
 
 	mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_PARENT);
-	ret = cachefiles_bury_object(cache, dir, object->dentry);
+
+	/* we need to check that our parent is _still_ our parent - it may have
+	 * been renamed */
+	if (dir == object->dentry->d_parent) {
+		ret = cachefiles_bury_object(cache, dir, object->dentry);
+	} else {
+		/* it got moved, presumably by cachefilesd culling it, so it's
+		 * no longer in the key path and we can ignore it */
+		mutex_unlock(&dir->d_inode->i_mutex);
+		ret = 0;
+	}
 
 	dput(dir);
 	_leave(" = %d", ret);
diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES
index 49503d2edc7e..bc0025cdd1c9 100644
--- a/fs/cifs/CHANGES
+++ b/fs/cifs/CHANGES
@@ -1,6 +1,7 @@
 Version 1.62
 ------------
-Add sockopt=TCP_NODELAY mount option.
+Add sockopt=TCP_NODELAY mount option. EA (xattr) routines hardened
+to more strictly handle corrupt frames.
 
 Version 1.61
 ------------
diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h
index ed751bb657db..a1c817eb291a 100644
--- a/fs/cifs/cifsglob.h
+++ b/fs/cifs/cifsglob.h
@@ -205,7 +205,7 @@ struct cifsUidInfo {
 struct cifsSesInfo {
 	struct list_head smb_ses_list;
 	struct list_head tcon_list;
-	struct semaphore sesSem;
+	struct mutex session_mutex;
 #if 0
 	struct cifsUidInfo *uidInfo;	/* pointer to user info */
 #endif
diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h
index 3877737f96a6..14d036d8db11 100644
--- a/fs/cifs/cifspdu.h
+++ b/fs/cifs/cifspdu.h
@@ -415,10 +415,10 @@ struct smb_hdr {
 	__u8 WordCount;
 } __attribute__((packed));
 /* given a pointer to an smb_hdr retrieve the value of byte count */
-#define BCC(smb_var) (*(__u16 *)((char *)smb_var + sizeof(struct smb_hdr) + (2 * smb_var->WordCount)))
-#define BCC_LE(smb_var) (*(__le16 *)((char *)smb_var + sizeof(struct smb_hdr) + (2 * smb_var->WordCount)))
+#define BCC(smb_var) (*(__u16 *)((char *)(smb_var) + sizeof(struct smb_hdr) + (2 * (smb_var)->WordCount)))
+#define BCC_LE(smb_var) (*(__le16 *)((char *)(smb_var) + sizeof(struct smb_hdr) + (2 * (smb_var)->WordCount)))
 /* given a pointer to an smb_hdr retrieve the pointer to the byte area */
-#define pByteArea(smb_var) ((unsigned char *)smb_var + sizeof(struct smb_hdr) + (2 * smb_var->WordCount) + 2)
+#define pByteArea(smb_var) ((unsigned char *)(smb_var) + sizeof(struct smb_hdr) + (2 * (smb_var)->WordCount) + 2)
 
 /*
  * Computer Name Length (since Netbios name was length 16 with last byte 0x20)
diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h
index 5646727e33f5..88e2bc44ac58 100644
--- a/fs/cifs/cifsproto.h
+++ b/fs/cifs/cifsproto.h
@@ -363,13 +363,10 @@ extern int CIFSSMBNotify(const int xid, struct cifsTconInfo *tcon,
 			__u32 filter, struct file *file, int multishot,
 			const struct nls_table *nls_codepage);
 extern ssize_t CIFSSMBQAllEAs(const int xid, struct cifsTconInfo *tcon,
-			const unsigned char *searchName, char *EAData,
+			const unsigned char *searchName,
+			const unsigned char *ea_name, char *EAData,
 			size_t bufsize, const struct nls_table *nls_codepage,
 			int remap_special_chars);
-extern ssize_t CIFSSMBQueryEA(const int xid, struct cifsTconInfo *tcon,
-		const unsigned char *searchName, const unsigned char *ea_name,
-		unsigned char *ea_value, size_t buf_size,
-		const struct nls_table *nls_codepage, int remap_special_chars);
 extern int CIFSSMBSetEA(const int xid, struct cifsTconInfo *tcon,
 		const char *fileName, const char *ea_name,
 		const void *ea_value, const __u16 ea_value_len,
diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c
index 941441d3e386..9d17df3e0768 100644
--- a/fs/cifs/cifssmb.c
+++ b/fs/cifs/cifssmb.c
@@ -170,19 +170,19 @@ cifs_reconnect_tcon(struct cifsTconInfo *tcon, int smb_command)
 	 * need to prevent multiple threads trying to simultaneously
 	 * reconnect the same SMB session
 	 */
-	down(&ses->sesSem);
+	mutex_lock(&ses->session_mutex);
 	if (ses->need_reconnect)
 		rc = cifs_setup_session(0, ses, nls_codepage);
 
 	/* do we need to reconnect tcon? */
 	if (rc || !tcon->need_reconnect) {
-		up(&ses->sesSem);
+		mutex_unlock(&ses->session_mutex);
 		goto out;
 	}
 
 	mark_open_files_invalid(tcon);
 	rc = CIFSTCon(0, ses, tcon->treeName, tcon, nls_codepage);
-	up(&ses->sesSem);
+	mutex_unlock(&ses->session_mutex);
 	cFYI(1, ("reconnect tcon rc = %d", rc));
 
 	if (rc)
@@ -700,13 +700,13 @@ CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses)
 	if (!ses || !ses->server)
 		return -EIO;
 
-	down(&ses->sesSem);
+	mutex_lock(&ses->session_mutex);
 	if (ses->need_reconnect)
 		goto session_already_dead; /* no need to send SMBlogoff if uid
 					      already closed due to reconnect */
 	rc = small_smb_init(SMB_COM_LOGOFF_ANDX, 2, NULL, (void **)&pSMB);
 	if (rc) {
-		up(&ses->sesSem);
+		mutex_unlock(&ses->session_mutex);
 		return rc;
 	}
 
@@ -721,7 +721,7 @@ CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses)
 	pSMB->AndXCommand = 0xFF;
 	rc = SendReceiveNoRsp(xid, ses, (struct smb_hdr *) pSMB, 0);
 session_already_dead:
-	up(&ses->sesSem);
+	mutex_unlock(&ses->session_mutex);
 
 	/* if session dead then we do not need to do ulogoff,
 		since server closed smb session, no sense reporting
@@ -5269,22 +5269,34 @@ int CIFSSMBNotify(const int xid, struct cifsTconInfo *tcon,
 	cifs_buf_release(pSMB);
 	return rc;
 }
+
 #ifdef CONFIG_CIFS_XATTR
+/*
+ * Do a path-based QUERY_ALL_EAS call and parse the result. This is a common
+ * function used by listxattr and getxattr type calls. When ea_name is set,
+ * it looks for that attribute name and stuffs that value into the EAData
+ * buffer. When ea_name is NULL, it stuffs a list of attribute names into the
+ * buffer. In both cases, the return value is either the length of the
+ * resulting data or a negative error code. If EAData is a NULL pointer then
+ * the data isn't copied to it, but the length is returned.
+ */
 ssize_t
 CIFSSMBQAllEAs(const int xid, struct cifsTconInfo *tcon,
-		 const unsigned char *searchName,
-		 char *EAData, size_t buf_size,
-		 const struct nls_table *nls_codepage, int remap)
+		const unsigned char *searchName, const unsigned char *ea_name,
+		char *EAData, size_t buf_size,
+		const struct nls_table *nls_codepage, int remap)
 {
 		/* BB assumes one setup word */
 	TRANSACTION2_QPI_REQ *pSMB = NULL;
 	TRANSACTION2_QPI_RSP *pSMBr = NULL;
 	int rc = 0;
 	int bytes_returned;
-	int name_len;
+	int list_len;
+	struct fealist *ea_response_data;
 	struct fea *temp_fea;
 	char *temp_ptr;
-	__u16 params, byte_count;
+	char *end_of_smb;
+	__u16 params, byte_count, data_offset;
 
 	cFYI(1, ("In Query All EAs path %s", searchName));
 QAllEAsRetry:
@@ -5294,22 +5306,22 @@ QAllEAsRetry:
 		return rc;
 
 	if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) {
-		name_len =
+		list_len =
 		    cifsConvertToUCS((__le16 *) pSMB->FileName, searchName,
 				     PATH_MAX, nls_codepage, remap);
-		name_len++;	/* trailing null */
-		name_len *= 2;
+		list_len++;	/* trailing null */
+		list_len *= 2;
 	} else {	/* BB improve the check for buffer overruns BB */
-		name_len = strnlen(searchName, PATH_MAX);
-		name_len++;	/* trailing null */
-		strncpy(pSMB->FileName, searchName, name_len);
+		list_len = strnlen(searchName, PATH_MAX);
+		list_len++;	/* trailing null */
+		strncpy(pSMB->FileName, searchName, list_len);
 	}
 
-	params = 2 /* level */ + 4 /* reserved */ + name_len /* includes NUL */;
+	params = 2 /* level */ + 4 /* reserved */ + list_len /* includes NUL */;
 	pSMB->TotalDataCount = 0;
 	pSMB->MaxParameterCount = cpu_to_le16(2);
 	/* BB find exact max SMB PDU from sess structure BB */
-	pSMB->MaxDataCount = cpu_to_le16(4000);
+	pSMB->MaxDataCount = cpu_to_le16(CIFSMaxBufSize);
 	pSMB->MaxSetupCount = 0;
 	pSMB->Reserved = 0;
 	pSMB->Flags = 0;
@@ -5334,237 +5346,117 @@ QAllEAsRetry:
 			 (struct smb_hdr *) pSMBr, &bytes_returned, 0);
 	if (rc) {
 		cFYI(1, ("Send error in QueryAllEAs = %d", rc));
-	} else {		/* decode response */
-		rc = validate_t2((struct smb_t2_rsp *)pSMBr);
+		goto QAllEAsOut;
+	}
 
-		/* BB also check enough total bytes returned */
-		/* BB we need to improve the validity checking
-		of these trans2 responses */
-		if (rc || (pSMBr->ByteCount < 4))
-			rc = -EIO;	/* bad smb */
-	   /* else if (pFindData){
-			memcpy((char *) pFindData,
-			       (char *) &pSMBr->hdr.Protocol +
-			       data_offset, kl);
-		}*/ else {
-			/* check that length of list is not more than bcc */
-			/* check that each entry does not go beyond length
-			   of list */
-			/* check that each element of each entry does not
-			   go beyond end of list */
-			__u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset);
-			struct fealist *ea_response_data;
-			rc = 0;
-			/* validate_trans2_offsets() */
-			/* BB check if start of smb + data_offset > &bcc+ bcc */
-			ea_response_data = (struct fealist *)
-				(((char *) &pSMBr->hdr.Protocol) +
-				data_offset);
-			name_len = le32_to_cpu(ea_response_data->list_len);
-			cFYI(1, ("ea length %d", name_len));
-			if (name_len <= 8) {
-			/* returned EA size zeroed at top of function */
-				cFYI(1, ("empty EA list returned from server"));
-			} else {
-				/* account for ea list len */
-				name_len -= 4;
-				temp_fea = ea_response_data->list;
-				temp_ptr = (char *)temp_fea;
-				while (name_len > 0) {
-					__u16 value_len;
-					name_len -= 4;
-					temp_ptr += 4;
-					rc += temp_fea->name_len;
-				/* account for prefix user. and trailing null */
-					rc = rc + 5 + 1;
-					if (rc < (int)buf_size) {
-						memcpy(EAData, "user.", 5);
-						EAData += 5;
-						memcpy(EAData, temp_ptr,
-						       temp_fea->name_len);
-						EAData += temp_fea->name_len;
-						/* null terminate name */
-						*EAData = 0;
-						EAData = EAData + 1;
-					} else if (buf_size == 0) {
-						/* skip copy - calc size only */
-					} else {
-						/* stop before overrun buffer */
-						rc = -ERANGE;
-						break;
-					}
-					name_len -= temp_fea->name_len;
-					temp_ptr += temp_fea->name_len;
-					/* account for trailing null */
-					name_len--;
-					temp_ptr++;
-					value_len =
-					      le16_to_cpu(temp_fea->value_len);
-					name_len -= value_len;
-					temp_ptr += value_len;
-					/* BB check that temp_ptr is still
-					      within the SMB BB*/
-
-					/* no trailing null to account for
-					   in value len */
-					/* go on to next EA */
-					temp_fea = (struct fea *)temp_ptr;
-				}
-			}
-		}
+
+	/* BB also check enough total bytes returned */
+	/* BB we need to improve the validity checking
+	of these trans2 responses */
+
+	rc = validate_t2((struct smb_t2_rsp *)pSMBr);
+	if (rc || (pSMBr->ByteCount < 4)) {
+		rc = -EIO;	/* bad smb */
+		goto QAllEAsOut;
 	}
-	cifs_buf_release(pSMB);
-	if (rc == -EAGAIN)
-		goto QAllEAsRetry;
 
-	return (ssize_t)rc;
-}
+	/* check that length of list is not more than bcc */
+	/* check that each entry does not go beyond length
+	   of list */
+	/* check that each element of each entry does not
+	   go beyond end of list */
+	/* validate_trans2_offsets() */
+	/* BB check if start of smb + data_offset > &bcc+ bcc */
 
-ssize_t CIFSSMBQueryEA(const int xid, struct cifsTconInfo *tcon,
-		const unsigned char *searchName, const unsigned char *ea_name,
-		unsigned char *ea_value, size_t buf_size,
-		const struct nls_table *nls_codepage, int remap)
-{
-	TRANSACTION2_QPI_REQ *pSMB = NULL;
-	TRANSACTION2_QPI_RSP *pSMBr = NULL;
-	int rc = 0;
-	int bytes_returned;
-	int name_len;
-	struct fea *temp_fea;
-	char *temp_ptr;
-	__u16 params, byte_count;
+	data_offset = le16_to_cpu(pSMBr->t2.DataOffset);
+	ea_response_data = (struct fealist *)
+				(((char *) &pSMBr->hdr.Protocol) + data_offset);
 
-	cFYI(1, ("In Query EA path %s", searchName));
-QEARetry:
-	rc = smb_init(SMB_COM_TRANSACTION2, 15, tcon, (void **) &pSMB,
-		      (void **) &pSMBr);
-	if (rc)
-		return rc;
+	list_len = le32_to_cpu(ea_response_data->list_len);
+	cFYI(1, ("ea length %d", list_len));
+	if (list_len <= 8) {
+		cFYI(1, ("empty EA list returned from server"));
+		goto QAllEAsOut;
+	}
 
-	if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) {
-		name_len =
-		    cifsConvertToUCS((__le16 *) pSMB->FileName, searchName,
-				     PATH_MAX, nls_codepage, remap);
-		name_len++;	/* trailing null */
-		name_len *= 2;
-	} else {	/* BB improve the check for buffer overruns BB */
-		name_len = strnlen(searchName, PATH_MAX);
-		name_len++;	/* trailing null */
-		strncpy(pSMB->FileName, searchName, name_len);
+	/* make sure list_len doesn't go past end of SMB */
+	end_of_smb = (char *)pByteArea(&pSMBr->hdr) + BCC(&pSMBr->hdr);
+	if ((char *)ea_response_data + list_len > end_of_smb) {
+		cFYI(1, ("EA list appears to go beyond SMB"));
+		rc = -EIO;
+		goto QAllEAsOut;
 	}
 
-	params = 2 /* level */ + 4 /* reserved */ + name_len /* includes NUL */;
-	pSMB->TotalDataCount = 0;
-	pSMB->MaxParameterCount = cpu_to_le16(2);
-	/* BB find exact max SMB PDU from sess structure BB */
-	pSMB->MaxDataCount = cpu_to_le16(4000);
-	pSMB->MaxSetupCount = 0;
-	pSMB->Reserved = 0;
-	pSMB->Flags = 0;
-	pSMB->Timeout = 0;
-	pSMB->Reserved2 = 0;
-	pSMB->ParameterOffset = cpu_to_le16(offsetof(
-		struct smb_com_transaction2_qpi_req, InformationLevel) - 4);
-	pSMB->DataCount = 0;
-	pSMB->DataOffset = 0;
-	pSMB->SetupCount = 1;
-	pSMB->Reserved3 = 0;
-	pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_PATH_INFORMATION);
-	byte_count = params + 1 /* pad */ ;
-	pSMB->TotalParameterCount = cpu_to_le16(params);
-	pSMB->ParameterCount = pSMB->TotalParameterCount;
-	pSMB->InformationLevel = cpu_to_le16(SMB_INFO_QUERY_ALL_EAS);
-	pSMB->Reserved4 = 0;
-	pSMB->hdr.smb_buf_length += byte_count;
-	pSMB->ByteCount = cpu_to_le16(byte_count);
+	/* account for ea list len */
+	list_len -= 4;
+	temp_fea = ea_response_data->list;
+	temp_ptr = (char *)temp_fea;
+	while (list_len > 0) {
+		unsigned int name_len;
+		__u16 value_len;
+
+		list_len -= 4;
+		temp_ptr += 4;
+		/* make sure we can read name_len and value_len */
+		if (list_len < 0) {
+			cFYI(1, ("EA entry goes beyond length of list"));
+			rc = -EIO;
+			goto QAllEAsOut;
+		}
 
-	rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB,
-			 (struct smb_hdr *) pSMBr, &bytes_returned, 0);
-	if (rc) {
-		cFYI(1, ("Send error in Query EA = %d", rc));
-	} else {		/* decode response */
-		rc = validate_t2((struct smb_t2_rsp *)pSMBr);
+		name_len = temp_fea->name_len;
+		value_len = le16_to_cpu(temp_fea->value_len);
+		list_len -= name_len + 1 + value_len;
+		if (list_len < 0) {
+			cFYI(1, ("EA entry goes beyond length of list"));
+			rc = -EIO;
+			goto QAllEAsOut;
+		}
 
-		/* BB also check enough total bytes returned */
-		/* BB we need to improve the validity checking
-		of these trans2 responses */
-		if (rc || (pSMBr->ByteCount < 4))
-			rc = -EIO;	/* bad smb */
-	   /* else if (pFindData){
-			memcpy((char *) pFindData,
-			       (char *) &pSMBr->hdr.Protocol +
-			       data_offset, kl);
-		}*/ else {
-			/* check that length of list is not more than bcc */
-			/* check that each entry does not go beyond length
-			   of list */
-			/* check that each element of each entry does not
-			   go beyond end of list */
-			__u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset);
-			struct fealist *ea_response_data;
-			rc = -ENODATA;
-			/* validate_trans2_offsets() */
-			/* BB check if start of smb + data_offset > &bcc+ bcc*/
-			ea_response_data = (struct fealist *)
-				(((char *) &pSMBr->hdr.Protocol) +
-				data_offset);
-			name_len = le32_to_cpu(ea_response_data->list_len);
-			cFYI(1, ("ea length %d", name_len));
-			if (name_len <= 8) {
-			/* returned EA size zeroed at top of function */
-				cFYI(1, ("empty EA list returned from server"));
-			} else {
-				/* account for ea list len */
-				name_len -= 4;
-				temp_fea = ea_response_data->list;
-				temp_ptr = (char *)temp_fea;
-				/* loop through checking if we have a matching
-				name and then return the associated value */
-				while (name_len > 0) {
-					__u16 value_len;
-					name_len -= 4;
-					temp_ptr += 4;
-					value_len =
-					      le16_to_cpu(temp_fea->value_len);
-				/* BB validate that value_len falls within SMB,
-				even though maximum for name_len is 255 */
-					if (memcmp(temp_fea->name, ea_name,
-						  temp_fea->name_len) == 0) {
-						/* found a match */
-						rc = value_len;
-				/* account for prefix user. and trailing null */
-						if (rc <= (int)buf_size) {
-							memcpy(ea_value,
-								temp_fea->name+temp_fea->name_len+1,
-								rc);
-							/* ea values, unlike ea
-							   names, are not null
-							   terminated */
-						} else if (buf_size == 0) {
-						/* skip copy - calc size only */
-						} else {
-						/* stop before overrun buffer */
-							rc = -ERANGE;
-						}
-						break;
-					}
-					name_len -= temp_fea->name_len;
-					temp_ptr += temp_fea->name_len;
-					/* account for trailing null */
-					name_len--;
-					temp_ptr++;
-					name_len -= value_len;
-					temp_ptr += value_len;
-					/* No trailing null to account for in
-					   value_len.  Go on to next EA */
-					temp_fea = (struct fea *)temp_ptr;
+		if (ea_name) {
+			if (strncmp(ea_name, temp_ptr, name_len) == 0) {
+				temp_ptr += name_len + 1;
+				rc = value_len;
+				if (buf_size == 0)
+					goto QAllEAsOut;
+				if ((size_t)value_len > buf_size) {
+					rc = -ERANGE;
+					goto QAllEAsOut;
 				}
+				memcpy(EAData, temp_ptr, value_len);
+				goto QAllEAsOut;
+			}
+		} else {
+			/* account for prefix user. and trailing null */
+			rc += (5 + 1 + name_len);
+			if (rc < (int) buf_size) {
+				memcpy(EAData, "user.", 5);
+				EAData += 5;
+				memcpy(EAData, temp_ptr, name_len);
+				EAData += name_len;
+				/* null terminate name */
+				*EAData = 0;
+				++EAData;
+			} else if (buf_size == 0) {
+				/* skip copy - calc size only */
+			} else {
+				/* stop before overrun buffer */
+				rc = -ERANGE;
+				break;
 			}
 		}
+		temp_ptr += name_len + 1 + value_len;
+		temp_fea = (struct fea *)temp_ptr;
 	}
+
+	/* didn't find the named attribute */
+	if (ea_name)
+		rc = -ENODATA;
+
+QAllEAsOut:
 	cifs_buf_release(pSMB);
 	if (rc == -EAGAIN)
-		goto QEARetry;
+		goto QAllEAsRetry;
 
 	return (ssize_t)rc;
 }
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 2e9e09ca0e30..45eb6cba793f 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -2388,13 +2388,13 @@ try_mount_again:
 		 */
 		cifs_put_tcp_session(srvTcp);
 
-		down(&pSesInfo->sesSem);
+		mutex_lock(&pSesInfo->session_mutex);
 		if (pSesInfo->need_reconnect) {
 			cFYI(1, ("Session needs reconnect"));
 			rc = cifs_setup_session(xid, pSesInfo,
 						cifs_sb->local_nls);
 		}
-		up(&pSesInfo->sesSem);
+		mutex_unlock(&pSesInfo->session_mutex);
 	} else if (!rc) {
 		cFYI(1, ("Existing smb sess not found"));
 		pSesInfo = sesInfoAlloc();
@@ -2437,12 +2437,12 @@ try_mount_again:
 		}
 		pSesInfo->linux_uid = volume_info->linux_uid;
 		pSesInfo->overrideSecFlg = volume_info->secFlg;
-		down(&pSesInfo->sesSem);
+		mutex_lock(&pSesInfo->session_mutex);
 
 		/* BB FIXME need to pass vol->secFlgs BB */
 		rc = cifs_setup_session(xid, pSesInfo,
 					cifs_sb->local_nls);
-		up(&pSesInfo->sesSem);
+		mutex_unlock(&pSesInfo->session_mutex);
 	}
 
 	/* search for existing tcon to this server share */
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index 057e1dae12ab..3d8f8a96f5a3 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -2289,9 +2289,9 @@ cifs_oplock_break(struct slow_work *work)
 	if (inode && S_ISREG(inode->i_mode)) {
 #ifdef CONFIG_CIFS_EXPERIMENTAL
 		if (cinode->clientCanCacheAll == 0)
-			break_lease(inode, FMODE_READ);
+			break_lease(inode, O_RDONLY);
 		else if (cinode->clientCanCacheRead == 0)
-			break_lease(inode, FMODE_WRITE);
+			break_lease(inode, O_WRONLY);
 #endif
 		rc = filemap_fdatawrite(inode->i_mapping);
 		if (cinode->clientCanCacheRead == 0) {
diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c
index e3fda978f481..8bdbc818164c 100644
--- a/fs/cifs/inode.c
+++ b/fs/cifs/inode.c
@@ -111,6 +111,7 @@ cifs_fattr_to_inode(struct inode *inode, struct cifs_fattr *fattr)
 
 	cifs_i->delete_pending = fattr->cf_flags & CIFS_FATTR_DELETE_PENDING;
 
+	cifs_i->server_eof = fattr->cf_eof;
 	/*
 	 * Can't safely change the file size here if the client is writing to
 	 * it due to potential races.
@@ -366,7 +367,7 @@ static int cifs_sfu_mode(struct cifs_fattr *fattr, const unsigned char *path,
 	char ea_value[4];
 	__u32 mode;
 
-	rc = CIFSSMBQueryEA(xid, cifs_sb->tcon, path, "SETFILEBITS",
+	rc = CIFSSMBQAllEAs(xid, cifs_sb->tcon, path, "SETFILEBITS",
 			    ea_value, 4 /* size of buf */, cifs_sb->local_nls,
 			    cifs_sb->mnt_cifs_flags &
 				CIFS_MOUNT_MAP_SPECIAL_CHR);
diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c
index d27d4ec6579b..d1474996a812 100644
--- a/fs/cifs/misc.c
+++ b/fs/cifs/misc.c
@@ -79,7 +79,7 @@ sesInfoAlloc(void)
 		++ret_buf->ses_count;
 		INIT_LIST_HEAD(&ret_buf->smb_ses_list);
 		INIT_LIST_HEAD(&ret_buf->tcon_list);
-		init_MUTEX(&ret_buf->sesSem);
+		mutex_init(&ret_buf->session_mutex);
 	}
 	return ret_buf;
 }
diff --git a/fs/cifs/xattr.c b/fs/cifs/xattr.c
index a75afa3dd9e1..3e2ef0de1209 100644
--- a/fs/cifs/xattr.c
+++ b/fs/cifs/xattr.c
@@ -244,7 +244,7 @@ ssize_t cifs_getxattr(struct dentry *direntry, const char *ea_name,
 			/* revalidate/getattr then populate from inode */
 		} /* BB add else when above is implemented */
 		ea_name += 5; /* skip past user. prefix */
-		rc = CIFSSMBQueryEA(xid, pTcon, full_path, ea_name, ea_value,
+		rc = CIFSSMBQAllEAs(xid, pTcon, full_path, ea_name, ea_value,
 			buf_size, cifs_sb->local_nls,
 			cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR);
 	} else if (strncmp(ea_name, CIFS_XATTR_OS2_PREFIX, 4) == 0) {
@@ -252,7 +252,7 @@ ssize_t cifs_getxattr(struct dentry *direntry, const char *ea_name,
 			goto get_ea_exit;
 
 		ea_name += 4; /* skip past os2. prefix */
-		rc = CIFSSMBQueryEA(xid, pTcon, full_path, ea_name, ea_value,
+		rc = CIFSSMBQAllEAs(xid, pTcon, full_path, ea_name, ea_value,
 			buf_size, cifs_sb->local_nls,
 			cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR);
 	} else if (strncmp(ea_name, POSIX_ACL_XATTR_ACCESS,
@@ -364,8 +364,8 @@ ssize_t cifs_listxattr(struct dentry *direntry, char *data, size_t buf_size)
 	/* if proc/fs/cifs/streamstoxattr is set then
 		search server for EAs or streams to
 		returns as xattrs */
-	rc = CIFSSMBQAllEAs(xid, pTcon, full_path, data, buf_size,
-				cifs_sb->local_nls,
+	rc = CIFSSMBQAllEAs(xid, pTcon, full_path, NULL, data,
+				buf_size, cifs_sb->local_nls,
 				cifs_sb->mnt_cifs_flags &
 					CIFS_MOUNT_MAP_SPECIAL_CHR);
 
diff --git a/fs/compat_binfmt_elf.c b/fs/compat_binfmt_elf.c
index 0adced2f296f..112e45a17e99 100644
--- a/fs/compat_binfmt_elf.c
+++ b/fs/compat_binfmt_elf.c
@@ -28,10 +28,12 @@
 
 #undef	elfhdr
 #undef	elf_phdr
+#undef	elf_shdr
 #undef	elf_note
 #undef	elf_addr_t
 #define elfhdr		elf32_hdr
 #define elf_phdr	elf32_phdr
+#define elf_shdr	elf32_shdr
 #define elf_note	elf32_note
 #define elf_addr_t	Elf32_Addr
 
diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c
index 0ca9ec4a79c3..6d55b61bfa79 100644
--- a/fs/compat_ioctl.c
+++ b/fs/compat_ioctl.c
@@ -545,7 +545,7 @@ static int mt_ioctl_trans(unsigned int fd, unsigned int cmd, void __user *argp)
 		kcmd = MTIOCPOS;
 		karg = &pos;
 		break;
-	case MTIOCGET32:
+	default:	/* MTIOCGET32 */
 		kcmd = MTIOCGET;
 		karg = &get;
 		break;
@@ -663,7 +663,7 @@ static int raw_ioctl(unsigned fd, unsigned cmd,
 
         switch (cmd) {
         case RAW_SETBIND:
-        case RAW_GETBIND: {
+	default: {	/* RAW_GETBIND */
                 struct raw_config_request req;
                 mm_segment_t oldfs = get_fs();
 
diff --git a/fs/dcache.c b/fs/dcache.c
index 953173a293a9..f1358e5c3a59 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -257,6 +257,7 @@ kill_it:
 	if (dentry)
 		goto repeat;
 }
+EXPORT_SYMBOL(dput);
 
 /**
  * d_invalidate - invalidate a dentry
@@ -314,6 +315,7 @@ int d_invalidate(struct dentry * dentry)
 	spin_unlock(&dcache_lock);
 	return 0;
 }
+EXPORT_SYMBOL(d_invalidate);
 
 /* This should be called _only_ with dcache_lock held */
 
@@ -328,6 +330,7 @@ struct dentry * dget_locked(struct dentry *dentry)
 {
 	return __dget_locked(dentry);
 }
+EXPORT_SYMBOL(dget_locked);
 
 /**
  * d_find_alias - grab a hashed alias of inode
@@ -384,6 +387,7 @@ struct dentry * d_find_alias(struct inode *inode)
 	}
 	return de;
 }
+EXPORT_SYMBOL(d_find_alias);
 
 /*
  *	Try to kill dentries associated with this inode.
@@ -408,6 +412,7 @@ restart:
 	}
 	spin_unlock(&dcache_lock);
 }
+EXPORT_SYMBOL(d_prune_aliases);
 
 /*
  * Throw away a dentry - free the inode, dput the parent.  This requires that
@@ -610,6 +615,7 @@ void shrink_dcache_sb(struct super_block * sb)
 {
 	__shrink_dcache_sb(sb, NULL, 0);
 }
+EXPORT_SYMBOL(shrink_dcache_sb);
 
 /*
  * destroy a single subtree of dentries for unmount
@@ -792,6 +798,7 @@ positive:
 	spin_unlock(&dcache_lock);
 	return 1;
 }
+EXPORT_SYMBOL(have_submounts);
 
 /*
  * Search the dentry child list for the specified parent,
@@ -876,6 +883,7 @@ void shrink_dcache_parent(struct dentry * parent)
 	while ((found = select_parent(parent)) != 0)
 		__shrink_dcache_sb(sb, &found, 0);
 }
+EXPORT_SYMBOL(shrink_dcache_parent);
 
 /*
  * Scan `nr' dentries and return the number which remain.
@@ -968,6 +976,7 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
 
 	return dentry;
 }
+EXPORT_SYMBOL(d_alloc);
 
 struct dentry *d_alloc_name(struct dentry *parent, const char *name)
 {
@@ -1012,6 +1021,7 @@ void d_instantiate(struct dentry *entry, struct inode * inode)
 	spin_unlock(&dcache_lock);
 	security_d_instantiate(entry, inode);
 }
+EXPORT_SYMBOL(d_instantiate);
 
 /**
  * d_instantiate_unique - instantiate a non-aliased dentry
@@ -1108,6 +1118,7 @@ struct dentry * d_alloc_root(struct inode * root_inode)
 	}
 	return res;
 }
+EXPORT_SYMBOL(d_alloc_root);
 
 static inline struct hlist_head *d_hash(struct dentry *parent,
 					unsigned long hash)
@@ -1211,7 +1222,6 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry)
 			BUG_ON(!(new->d_flags & DCACHE_DISCONNECTED));
 			spin_unlock(&dcache_lock);
 			security_d_instantiate(new, inode);
-			d_rehash(dentry);
 			d_move(new, dentry);
 			iput(inode);
 		} else {
@@ -1225,6 +1235,7 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry)
 		d_add(dentry, inode);
 	return new;
 }
+EXPORT_SYMBOL(d_splice_alias);
 
 /**
  * d_add_ci - lookup or allocate new dentry with case-exact name
@@ -1314,6 +1325,7 @@ err_out:
 	iput(inode);
 	return ERR_PTR(error);
 }
+EXPORT_SYMBOL(d_add_ci);
 
 /**
  * d_lookup - search for a dentry
@@ -1357,6 +1369,7 @@ struct dentry * d_lookup(struct dentry * parent, struct qstr * name)
 	} while (read_seqretry(&rename_lock, seq));
 	return dentry;
 }
+EXPORT_SYMBOL(d_lookup);
 
 struct dentry * __d_lookup(struct dentry * parent, struct qstr * name)
 {
@@ -1483,6 +1496,7 @@ int d_validate(struct dentry *dentry, struct dentry *dparent)
 out:
 	return 0;
 }
+EXPORT_SYMBOL(d_validate);
 
 /*
  * When a file is deleted, we have two options:
@@ -1528,6 +1542,7 @@ void d_delete(struct dentry * dentry)
 
 	fsnotify_nameremove(dentry, isdir);
 }
+EXPORT_SYMBOL(d_delete);
 
 static void __d_rehash(struct dentry * entry, struct hlist_head *list)
 {
@@ -1556,6 +1571,7 @@ void d_rehash(struct dentry * entry)
 	spin_unlock(&entry->d_lock);
 	spin_unlock(&dcache_lock);
 }
+EXPORT_SYMBOL(d_rehash);
 
 /*
  * When switching names, the actual string doesn't strictly have to
@@ -1702,6 +1718,7 @@ void d_move(struct dentry * dentry, struct dentry * target)
 	d_move_locked(dentry, target);
 	spin_unlock(&dcache_lock);
 }
+EXPORT_SYMBOL(d_move);
 
 /**
  * d_ancestor - search for an ancestor
@@ -1868,6 +1885,7 @@ shouldnt_be_hashed:
 	spin_unlock(&dcache_lock);
 	BUG();
 }
+EXPORT_SYMBOL_GPL(d_materialise_unique);
 
 static int prepend(char **buffer, int *buflen, const char *str, int namelen)
 {
@@ -2005,6 +2023,7 @@ char *d_path(const struct path *path, char *buf, int buflen)
 	path_put(&root);
 	return res;
 }
+EXPORT_SYMBOL(d_path);
 
 /*
  * Helper function for dentry_operations.d_dname() members
@@ -2171,6 +2190,30 @@ int is_subdir(struct dentry *new_dentry, struct dentry *old_dentry)
 	return result;
 }
 
+int path_is_under(struct path *path1, struct path *path2)
+{
+	struct vfsmount *mnt = path1->mnt;
+	struct dentry *dentry = path1->dentry;
+	int res;
+	spin_lock(&vfsmount_lock);
+	if (mnt != path2->mnt) {
+		for (;;) {
+			if (mnt->mnt_parent == mnt) {
+				spin_unlock(&vfsmount_lock);
+				return 0;
+			}
+			if (mnt->mnt_parent == path2->mnt)
+				break;
+			mnt = mnt->mnt_parent;
+		}
+		dentry = mnt->mnt_mountpoint;
+	}
+	res = is_subdir(dentry, path2->dentry);
+	spin_unlock(&vfsmount_lock);
+	return res;
+}
+EXPORT_SYMBOL(path_is_under);
+
 void d_genocide(struct dentry *root)
 {
 	struct dentry *this_parent = root;
@@ -2228,6 +2271,7 @@ ino_t find_inode_number(struct dentry *dir, struct qstr *name)
 	}
 	return ino;
 }
+EXPORT_SYMBOL(find_inode_number);
 
 static __initdata unsigned long dhash_entries;
 static int __init set_dhash_entries(char *str)
@@ -2297,6 +2341,7 @@ static void __init dcache_init(void)
 
 /* SLAB cache for __getname() consumers */
 struct kmem_cache *names_cachep __read_mostly;
+EXPORT_SYMBOL(names_cachep);
 
 EXPORT_SYMBOL(d_genocide);
 
@@ -2326,26 +2371,3 @@ void __init vfs_caches_init(unsigned long mempages)
 	bdev_cache_init();
 	chrdev_init();
 }
-
-EXPORT_SYMBOL(d_alloc);
-EXPORT_SYMBOL(d_alloc_root);
-EXPORT_SYMBOL(d_delete);
-EXPORT_SYMBOL(d_find_alias);
-EXPORT_SYMBOL(d_instantiate);
-EXPORT_SYMBOL(d_invalidate);
-EXPORT_SYMBOL(d_lookup);
-EXPORT_SYMBOL(d_move);
-EXPORT_SYMBOL_GPL(d_materialise_unique);
-EXPORT_SYMBOL(d_path);
-EXPORT_SYMBOL(d_prune_aliases);
-EXPORT_SYMBOL(d_rehash);
-EXPORT_SYMBOL(d_splice_alias);
-EXPORT_SYMBOL(d_add_ci);
-EXPORT_SYMBOL(d_validate);
-EXPORT_SYMBOL(dget_locked);
-EXPORT_SYMBOL(dput);
-EXPORT_SYMBOL(find_inode_number);
-EXPORT_SYMBOL(have_submounts);
-EXPORT_SYMBOL(names_cachep);
-EXPORT_SYMBOL(shrink_dcache_parent);
-EXPORT_SYMBOL(shrink_dcache_sb);
diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index 274ac865bae8..049d6c36da09 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -496,7 +496,7 @@ struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
 	}
 	d_move(old_dentry, dentry);
 	fsnotify_move(old_dir->d_inode, new_dir->d_inode, old_name,
-		old_dentry->d_name.name, S_ISDIR(old_dentry->d_inode->i_mode),
+		S_ISDIR(old_dentry->d_inode->i_mode),
 		NULL, old_dentry);
 	fsnotify_oldname_free(old_name);
 	unlock_rename(new_dir, old_dir);
diff --git a/fs/dlm/ast.c b/fs/dlm/ast.c
index dc2ad6008b2d..4314f0d48d85 100644
--- a/fs/dlm/ast.c
+++ b/fs/dlm/ast.c
@@ -2,7 +2,7 @@
 *******************************************************************************
 **
 **  Copyright (C) Sistina Software, Inc.  1997-2003  All rights reserved.
-**  Copyright (C) 2004-2008 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2004-2010 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -33,10 +33,10 @@ void dlm_del_ast(struct dlm_lkb *lkb)
 	spin_unlock(&ast_queue_lock);
 }
 
-void dlm_add_ast(struct dlm_lkb *lkb, int type, int bastmode)
+void dlm_add_ast(struct dlm_lkb *lkb, int type, int mode)
 {
 	if (lkb->lkb_flags & DLM_IFL_USER) {
-		dlm_user_add_ast(lkb, type, bastmode);
+		dlm_user_add_ast(lkb, type, mode);
 		return;
 	}
 
@@ -44,10 +44,21 @@ void dlm_add_ast(struct dlm_lkb *lkb, int type, int bastmode)
 	if (!(lkb->lkb_ast_type & (AST_COMP | AST_BAST))) {
 		kref_get(&lkb->lkb_ref);
 		list_add_tail(&lkb->lkb_astqueue, &ast_queue);
+		lkb->lkb_ast_first = type;
 	}
+
+	/* sanity check, this should not happen */
+
+	if ((type == AST_COMP) && (lkb->lkb_ast_type & AST_COMP))
+		log_print("repeat cast %d castmode %d lock %x %s",
+			  mode, lkb->lkb_castmode,
+			  lkb->lkb_id, lkb->lkb_resource->res_name);
+
 	lkb->lkb_ast_type |= type;
-	if (bastmode)
-		lkb->lkb_bastmode = bastmode;
+	if (type == AST_BAST)
+		lkb->lkb_bastmode = mode;
+	else
+		lkb->lkb_castmode = mode;
 	spin_unlock(&ast_queue_lock);
 
 	set_bit(WAKE_ASTS, &astd_wakeflags);
@@ -59,9 +70,9 @@ static void process_asts(void)
 	struct dlm_ls *ls = NULL;
 	struct dlm_rsb *r = NULL;
 	struct dlm_lkb *lkb;
-	void (*cast) (void *astparam);
-	void (*bast) (void *astparam, int mode);
-	int type = 0, bastmode;
+	void (*castfn) (void *astparam);
+	void (*bastfn) (void *astparam, int mode);
+	int type, first, bastmode, castmode, do_bast, do_cast, last_castmode;
 
 repeat:
 	spin_lock(&ast_queue_lock);
@@ -75,17 +86,48 @@ repeat:
 		list_del(&lkb->lkb_astqueue);
 		type = lkb->lkb_ast_type;
 		lkb->lkb_ast_type = 0;
+		first = lkb->lkb_ast_first;
+		lkb->lkb_ast_first = 0;
 		bastmode = lkb->lkb_bastmode;
-
+		castmode = lkb->lkb_castmode;
+		castfn = lkb->lkb_astfn;
+		bastfn = lkb->lkb_bastfn;
 		spin_unlock(&ast_queue_lock);
-		cast = lkb->lkb_astfn;
-		bast = lkb->lkb_bastfn;
-
-		if ((type & AST_COMP) && cast)
-			cast(lkb->lkb_astparam);
 
-		if ((type & AST_BAST) && bast)
-			bast(lkb->lkb_astparam, bastmode);
+		do_cast = (type & AST_COMP) && castfn;
+		do_bast = (type & AST_BAST) && bastfn;
+
+		/* Skip a bast if its blocking mode is compatible with the
+		   granted mode of the preceding cast. */
+
+		if (do_bast) {
+			if (first == AST_COMP)
+				last_castmode = castmode;
+			else
+				last_castmode = lkb->lkb_castmode_done;
+			if (dlm_modes_compat(bastmode, last_castmode))
+				do_bast = 0;
+		}
+
+		if (first == AST_COMP) {
+			if (do_cast)
+				castfn(lkb->lkb_astparam);
+			if (do_bast)
+				bastfn(lkb->lkb_astparam, bastmode);
+		} else if (first == AST_BAST) {
+			if (do_bast)
+				bastfn(lkb->lkb_astparam, bastmode);
+			if (do_cast)
+				castfn(lkb->lkb_astparam);
+		} else {
+			log_error(ls, "bad ast_first %d ast_type %d",
+				  first, type);
+		}
+
+		if (do_cast)
+			lkb->lkb_castmode_done = castmode;
+		if (do_bast)
+			lkb->lkb_bastmode_done = bastmode;
 
 		/* this removes the reference added by dlm_add_ast
 		   and may result in the lkb being freed */
diff --git a/fs/dlm/ast.h b/fs/dlm/ast.h
index 1b5fc5f428fd..bcb1aaba519d 100644
--- a/fs/dlm/ast.h
+++ b/fs/dlm/ast.h
@@ -1,7 +1,7 @@
 /******************************************************************************
 *******************************************************************************
 **
-**  Copyright (C) 2005-2008 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2005-2010 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -13,7 +13,7 @@
 #ifndef __ASTD_DOT_H__
 #define __ASTD_DOT_H__
 
-void dlm_add_ast(struct dlm_lkb *lkb, int type, int bastmode);
+void dlm_add_ast(struct dlm_lkb *lkb, int type, int mode);
 void dlm_del_ast(struct dlm_lkb *lkb);
 
 void dlm_astd_wake(void);
diff --git a/fs/dlm/debug_fs.c b/fs/dlm/debug_fs.c
index 375a2359b3bf..29d6139c35fc 100644
--- a/fs/dlm/debug_fs.c
+++ b/fs/dlm/debug_fs.c
@@ -256,7 +256,7 @@ static int print_format3_lock(struct seq_file *s, struct dlm_lkb *lkb,
 			lkb->lkb_status,
 			lkb->lkb_grmode,
 			lkb->lkb_rqmode,
-			lkb->lkb_highbast,
+			lkb->lkb_bastmode,
 			rsb_lookup,
 			lkb->lkb_wait_type,
 			lkb->lkb_lvbseq,
diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h
index 826d3dc6e0ab..f632b58cd222 100644
--- a/fs/dlm/dlm_internal.h
+++ b/fs/dlm/dlm_internal.h
@@ -2,7 +2,7 @@
 *******************************************************************************
 **
 **  Copyright (C) Sistina Software, Inc.  1997-2003  All rights reserved.
-**  Copyright (C) 2004-2008 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2004-2010 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -232,11 +232,17 @@ struct dlm_lkb {
 	int8_t			lkb_status;     /* granted, waiting, convert */
 	int8_t			lkb_rqmode;	/* requested lock mode */
 	int8_t			lkb_grmode;	/* granted lock mode */
-	int8_t			lkb_bastmode;	/* requested mode */
 	int8_t			lkb_highbast;	/* highest mode bast sent for */
+
 	int8_t			lkb_wait_type;	/* type of reply waiting for */
 	int8_t			lkb_wait_count;
 	int8_t			lkb_ast_type;	/* type of ast queued for */
+	int8_t			lkb_ast_first;	/* type of first ast queued */
+
+	int8_t			lkb_bastmode;	/* req mode of queued bast */
+	int8_t			lkb_castmode;	/* gr mode of queued cast */
+	int8_t			lkb_bastmode_done; /* last delivered bastmode */
+	int8_t			lkb_castmode_done; /* last delivered castmode */
 
 	struct list_head	lkb_idtbl_list;	/* lockspace lkbtbl */
 	struct list_head	lkb_statequeue;	/* rsb g/c/w list */
diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c
index 9c0c1db1e105..46ffd3eeaaf7 100644
--- a/fs/dlm/lock.c
+++ b/fs/dlm/lock.c
@@ -1,7 +1,7 @@
 /******************************************************************************
 *******************************************************************************
 **
-**  Copyright (C) 2005-2008 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2005-2010 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -307,7 +307,7 @@ static void queue_cast(struct dlm_rsb *r, struct dlm_lkb *lkb, int rv)
 	lkb->lkb_lksb->sb_status = rv;
 	lkb->lkb_lksb->sb_flags = lkb->lkb_sbflags;
 
-	dlm_add_ast(lkb, AST_COMP, 0);
+	dlm_add_ast(lkb, AST_COMP, lkb->lkb_grmode);
 }
 
 static inline void queue_cast_overlap(struct dlm_rsb *r, struct dlm_lkb *lkb)
@@ -320,10 +320,12 @@ static void queue_bast(struct dlm_rsb *r, struct dlm_lkb *lkb, int rqmode)
 {
 	lkb->lkb_time_bast = ktime_get();
 
-	if (is_master_copy(lkb))
+	if (is_master_copy(lkb)) {
+		lkb->lkb_bastmode = rqmode; /* printed by debugfs */
 		send_bast(r, lkb, rqmode);
-	else
+	} else {
 		dlm_add_ast(lkb, AST_BAST, rqmode);
+	}
 }
 
 /*
@@ -2280,20 +2282,30 @@ static int do_request(struct dlm_rsb *r, struct dlm_lkb *lkb)
 	if (can_be_queued(lkb)) {
 		error = -EINPROGRESS;
 		add_lkb(r, lkb, DLM_LKSTS_WAITING);
-		send_blocking_asts(r, lkb);
 		add_timeout(lkb);
 		goto out;
 	}
 
 	error = -EAGAIN;
-	if (force_blocking_asts(lkb))
-		send_blocking_asts_all(r, lkb);
 	queue_cast(r, lkb, -EAGAIN);
-
  out:
 	return error;
 }
 
+static void do_request_effects(struct dlm_rsb *r, struct dlm_lkb *lkb,
+			       int error)
+{
+	switch (error) {
+	case -EAGAIN:
+		if (force_blocking_asts(lkb))
+			send_blocking_asts_all(r, lkb);
+		break;
+	case -EINPROGRESS:
+		send_blocking_asts(r, lkb);
+		break;
+	}
+}
+
 static int do_convert(struct dlm_rsb *r, struct dlm_lkb *lkb)
 {
 	int error = 0;
@@ -2304,7 +2316,6 @@ static int do_convert(struct dlm_rsb *r, struct dlm_lkb *lkb)
 	if (can_be_granted(r, lkb, 1, &deadlk)) {
 		grant_lock(r, lkb);
 		queue_cast(r, lkb, 0);
-		grant_pending_locks(r);
 		goto out;
 	}
 
@@ -2334,7 +2345,6 @@ static int do_convert(struct dlm_rsb *r, struct dlm_lkb *lkb)
 		if (_can_be_granted(r, lkb, 1)) {
 			grant_lock(r, lkb);
 			queue_cast(r, lkb, 0);
-			grant_pending_locks(r);
 			goto out;
 		}
 		/* else fall through and move to convert queue */
@@ -2344,28 +2354,47 @@ static int do_convert(struct dlm_rsb *r, struct dlm_lkb *lkb)
 		error = -EINPROGRESS;
 		del_lkb(r, lkb);
 		add_lkb(r, lkb, DLM_LKSTS_CONVERT);
-		send_blocking_asts(r, lkb);
 		add_timeout(lkb);
 		goto out;
 	}
 
 	error = -EAGAIN;
-	if (force_blocking_asts(lkb))
-		send_blocking_asts_all(r, lkb);
 	queue_cast(r, lkb, -EAGAIN);
-
  out:
 	return error;
 }
 
+static void do_convert_effects(struct dlm_rsb *r, struct dlm_lkb *lkb,
+			       int error)
+{
+	switch (error) {
+	case 0:
+		grant_pending_locks(r);
+		/* grant_pending_locks also sends basts */
+		break;
+	case -EAGAIN:
+		if (force_blocking_asts(lkb))
+			send_blocking_asts_all(r, lkb);
+		break;
+	case -EINPROGRESS:
+		send_blocking_asts(r, lkb);
+		break;
+	}
+}
+
 static int do_unlock(struct dlm_rsb *r, struct dlm_lkb *lkb)
 {
 	remove_lock(r, lkb);
 	queue_cast(r, lkb, -DLM_EUNLOCK);
-	grant_pending_locks(r);
 	return -DLM_EUNLOCK;
 }
 
+static void do_unlock_effects(struct dlm_rsb *r, struct dlm_lkb *lkb,
+			      int error)
+{
+	grant_pending_locks(r);
+}
+
 /* returns: 0 did nothing, -DLM_ECANCEL canceled lock */
  
 static int do_cancel(struct dlm_rsb *r, struct dlm_lkb *lkb)
@@ -2375,12 +2404,18 @@ static int do_cancel(struct dlm_rsb *r, struct dlm_lkb *lkb)
 	error = revert_lock(r, lkb);
 	if (error) {
 		queue_cast(r, lkb, -DLM_ECANCEL);
-		grant_pending_locks(r);
 		return -DLM_ECANCEL;
 	}
 	return 0;
 }
 
+static void do_cancel_effects(struct dlm_rsb *r, struct dlm_lkb *lkb,
+			      int error)
+{
+	if (error)
+		grant_pending_locks(r);
+}
+
 /*
  * Four stage 3 varieties:
  * _request_lock(), _convert_lock(), _unlock_lock(), _cancel_lock()
@@ -2402,11 +2437,15 @@ static int _request_lock(struct dlm_rsb *r, struct dlm_lkb *lkb)
 		goto out;
 	}
 
-	if (is_remote(r))
+	if (is_remote(r)) {
 		/* receive_request() calls do_request() on remote node */
 		error = send_request(r, lkb);
-	else
+	} else {
 		error = do_request(r, lkb);
+		/* for remote locks the request_reply is sent
+		   between do_request and do_request_effects */
+		do_request_effects(r, lkb, error);
+	}
  out:
 	return error;
 }
@@ -2417,11 +2456,15 @@ static int _convert_lock(struct dlm_rsb *r, struct dlm_lkb *lkb)
 {
 	int error;
 
-	if (is_remote(r))
+	if (is_remote(r)) {
 		/* receive_convert() calls do_convert() on remote node */
 		error = send_convert(r, lkb);
-	else
+	} else {
 		error = do_convert(r, lkb);
+		/* for remote locks the convert_reply is sent
+		   between do_convert and do_convert_effects */
+		do_convert_effects(r, lkb, error);
+	}
 
 	return error;
 }
@@ -2432,11 +2475,15 @@ static int _unlock_lock(struct dlm_rsb *r, struct dlm_lkb *lkb)
 {
 	int error;
 
-	if (is_remote(r))
+	if (is_remote(r)) {
 		/* receive_unlock() calls do_unlock() on remote node */
 		error = send_unlock(r, lkb);
-	else
+	} else {
 		error = do_unlock(r, lkb);
+		/* for remote locks the unlock_reply is sent
+		   between do_unlock and do_unlock_effects */
+		do_unlock_effects(r, lkb, error);
+	}
 
 	return error;
 }
@@ -2447,11 +2494,15 @@ static int _cancel_lock(struct dlm_rsb *r, struct dlm_lkb *lkb)
 {
 	int error;
 
-	if (is_remote(r))
+	if (is_remote(r)) {
 		/* receive_cancel() calls do_cancel() on remote node */
 		error = send_cancel(r, lkb);
-	else
+	} else {
 		error = do_cancel(r, lkb);
+		/* for remote locks the cancel_reply is sent
+		   between do_cancel and do_cancel_effects */
+		do_cancel_effects(r, lkb, error);
+	}
 
 	return error;
 }
@@ -3191,6 +3242,7 @@ static void receive_request(struct dlm_ls *ls, struct dlm_message *ms)
 	attach_lkb(r, lkb);
 	error = do_request(r, lkb);
 	send_request_reply(r, lkb, error);
+	do_request_effects(r, lkb, error);
 
 	unlock_rsb(r);
 	put_rsb(r);
@@ -3226,15 +3278,19 @@ static void receive_convert(struct dlm_ls *ls, struct dlm_message *ms)
 		goto out;
 
 	receive_flags(lkb, ms);
+
 	error = receive_convert_args(ls, lkb, ms);
-	if (error)
-		goto out_reply;
+	if (error) {
+		send_convert_reply(r, lkb, error);
+		goto out;
+	}
+
 	reply = !down_conversion(lkb);
 
 	error = do_convert(r, lkb);
- out_reply:
 	if (reply)
 		send_convert_reply(r, lkb, error);
+	do_convert_effects(r, lkb, error);
  out:
 	unlock_rsb(r);
 	put_rsb(r);
@@ -3266,13 +3322,16 @@ static void receive_unlock(struct dlm_ls *ls, struct dlm_message *ms)
 		goto out;
 
 	receive_flags(lkb, ms);
+
 	error = receive_unlock_args(ls, lkb, ms);
-	if (error)
-		goto out_reply;
+	if (error) {
+		send_unlock_reply(r, lkb, error);
+		goto out;
+	}
 
 	error = do_unlock(r, lkb);
- out_reply:
 	send_unlock_reply(r, lkb, error);
+	do_unlock_effects(r, lkb, error);
  out:
 	unlock_rsb(r);
 	put_rsb(r);
@@ -3307,6 +3366,7 @@ static void receive_cancel(struct dlm_ls *ls, struct dlm_message *ms)
 
 	error = do_cancel(r, lkb);
 	send_cancel_reply(r, lkb, error);
+	do_cancel_effects(r, lkb, error);
  out:
 	unlock_rsb(r);
 	put_rsb(r);
diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c
index c010ecfc0d29..26a8bd40400a 100644
--- a/fs/dlm/lockspace.c
+++ b/fs/dlm/lockspace.c
@@ -191,6 +191,18 @@ static int do_uevent(struct dlm_ls *ls, int in)
 	return error;
 }
 
+static int dlm_uevent(struct kset *kset, struct kobject *kobj,
+		      struct kobj_uevent_env *env)
+{
+	struct dlm_ls *ls = container_of(kobj, struct dlm_ls, ls_kobj);
+
+	add_uevent_var(env, "LOCKSPACE=%s", ls->ls_name);
+	return 0;
+}
+
+static struct kset_uevent_ops dlm_uevent_ops = {
+	.uevent = dlm_uevent,
+};
 
 int __init dlm_lockspace_init(void)
 {
@@ -199,7 +211,7 @@ int __init dlm_lockspace_init(void)
 	INIT_LIST_HEAD(&lslist);
 	spin_lock_init(&lslist_lock);
 
-	dlm_kset = kset_create_and_add("dlm", NULL, kernel_kobj);
+	dlm_kset = kset_create_and_add("dlm", &dlm_uevent_ops, kernel_kobj);
 	if (!dlm_kset) {
 		printk(KERN_WARNING "%s: can not create kset\n", __func__);
 		return -ENOMEM;
diff --git a/fs/dlm/user.c b/fs/dlm/user.c
index e73a4bb572aa..a4bfd31ac45b 100644
--- a/fs/dlm/user.c
+++ b/fs/dlm/user.c
@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2006-2009 Red Hat, Inc.  All rights reserved.
+ * Copyright (C) 2006-2010 Red Hat, Inc.  All rights reserved.
  *
  * This copyrighted material is made available to anyone wishing to use,
  * modify, copy, or redistribute it subject to the terms and conditions
@@ -173,7 +173,7 @@ static int lkb_is_endoflife(struct dlm_lkb *lkb, int sb_status, int type)
 /* we could possibly check if the cancel of an orphan has resulted in the lkb
    being removed and then remove that lkb from the orphans list and free it */
 
-void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int bastmode)
+void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int mode)
 {
 	struct dlm_ls *ls;
 	struct dlm_user_args *ua;
@@ -206,8 +206,10 @@ void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int bastmode)
 
 	ast_type = lkb->lkb_ast_type;
 	lkb->lkb_ast_type |= type;
-	if (bastmode)
-		lkb->lkb_bastmode = bastmode;
+	if (type == AST_BAST)
+		lkb->lkb_bastmode = mode;
+	else
+		lkb->lkb_castmode = mode;
 
 	if (!ast_type) {
 		kref_get(&lkb->lkb_ref);
diff --git a/fs/dlm/user.h b/fs/dlm/user.h
index 1c9686492286..f196091dd7ff 100644
--- a/fs/dlm/user.h
+++ b/fs/dlm/user.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2006-2008 Red Hat, Inc.  All rights reserved.
+ * Copyright (C) 2006-2010 Red Hat, Inc.  All rights reserved.
  *
  * This copyrighted material is made available to anyone wishing to use,
  * modify, copy, or redistribute it subject to the terms and conditions
@@ -9,7 +9,7 @@
 #ifndef __USER_DOT_H__
 #define __USER_DOT_H__
 
-void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int bastmode);
+void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int mode);
 int dlm_user_init(void);
 void dlm_user_exit(void);
 int dlm_device_deregister(struct dlm_ls *ls);
diff --git a/fs/exec.c b/fs/exec.c
index e95c692ef0e4..49cdaa19e5b9 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -195,7 +195,7 @@ static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos,
 		 *    to work from.
 		 */
 		rlim = current->signal->rlim;
-		if (size > rlim[RLIMIT_STACK].rlim_cur / 4) {
+		if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4) {
 			put_page(page);
 			return NULL;
 		}
@@ -246,6 +246,7 @@ static int __bprm_mm_init(struct linux_binprm *bprm)
 	vma->vm_start = vma->vm_end - PAGE_SIZE;
 	vma->vm_flags = VM_STACK_FLAGS;
 	vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
+	INIT_LIST_HEAD(&vma->anon_vma_chain);
 	err = insert_vm_struct(mm, vma);
 	if (err)
 		goto err;
@@ -516,7 +517,8 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift)
 	/*
 	 * cover the whole range: [new_start, old_end)
 	 */
-	vma_adjust(vma, new_start, old_end, vma->vm_pgoff, NULL);
+	if (vma_adjust(vma, new_start, old_end, vma->vm_pgoff, NULL))
+		return -ENOMEM;
 
 	/*
 	 * move the page tables downwards, on failure we rely on
@@ -547,15 +549,13 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift)
 	tlb_finish_mmu(tlb, new_end, old_end);
 
 	/*
-	 * shrink the vma to just the new range.
+	 * Shrink the vma to just the new range.  Always succeeds.
 	 */
 	vma_adjust(vma, new_start, new_end, vma->vm_pgoff, NULL);
 
 	return 0;
 }
 
-#define EXTRA_STACK_VM_PAGES	20	/* random */
-
 /*
  * Finalizes the stack vm_area_struct. The flags and permissions are updated,
  * the stack is optionally relocated, and some extra space is added.
@@ -577,7 +577,7 @@ int setup_arg_pages(struct linux_binprm *bprm,
 
 #ifdef CONFIG_STACK_GROWSUP
 	/* Limit stack size to 1GB */
-	stack_base = current->signal->rlim[RLIMIT_STACK].rlim_max;
+	stack_base = rlimit_max(RLIMIT_STACK);
 	if (stack_base > (1 << 30))
 		stack_base = 1 << 30;
 
@@ -630,14 +630,13 @@ int setup_arg_pages(struct linux_binprm *bprm,
 			goto out_unlock;
 	}
 
-	stack_expand = EXTRA_STACK_VM_PAGES * PAGE_SIZE;
+	stack_expand = 131072UL; /* randomly 32*4k (or 2*64k) pages */
 	stack_size = vma->vm_end - vma->vm_start;
 	/*
 	 * Align this down to a page boundary as expand_stack
 	 * will align it up.
 	 */
 	rlim_stack = rlimit(RLIMIT_STACK) & PAGE_MASK;
-	rlim_stack = min(rlim_stack, stack_size);
 #ifdef CONFIG_STACK_GROWSUP
 	if (stack_size + stack_expand > rlim_stack)
 		stack_base = vma->vm_start + rlim_stack;
@@ -719,6 +718,7 @@ static int exec_mmap(struct mm_struct *mm)
 	/* Notify parent that we're no longer interested in the old VM */
 	tsk = current;
 	old_mm = current->mm;
+	sync_mm_rss(tsk, old_mm);
 	mm_release(tsk, old_mm);
 
 	if (old_mm) {
@@ -1533,7 +1533,7 @@ static int format_corename(char *corename, long signr)
 			/* core limit size */
 			case 'c':
 				rc = snprintf(out_ptr, out_end - out_ptr,
-					      "%lu", current->signal->rlim[RLIMIT_CORE].rlim_cur);
+					      "%lu", rlimit(RLIMIT_CORE));
 				if (rc > out_end - out_ptr)
 					goto out;
 				out_ptr += rc;
@@ -1561,12 +1561,13 @@ out:
 	return ispipe;
 }
 
-static int zap_process(struct task_struct *start)
+static int zap_process(struct task_struct *start, int exit_code)
 {
 	struct task_struct *t;
 	int nr = 0;
 
 	start->signal->flags = SIGNAL_GROUP_EXIT;
+	start->signal->group_exit_code = exit_code;
 	start->signal->group_stop_count = 0;
 
 	t = start;
@@ -1591,8 +1592,7 @@ static inline int zap_threads(struct task_struct *tsk, struct mm_struct *mm,
 	spin_lock_irq(&tsk->sighand->siglock);
 	if (!signal_group_exit(tsk->signal)) {
 		mm->core_state = core_state;
-		tsk->signal->group_exit_code = exit_code;
-		nr = zap_process(tsk);
+		nr = zap_process(tsk, exit_code);
 	}
 	spin_unlock_irq(&tsk->sighand->siglock);
 	if (unlikely(nr < 0))
@@ -1641,7 +1641,7 @@ static inline int zap_threads(struct task_struct *tsk, struct mm_struct *mm,
 			if (p->mm) {
 				if (unlikely(p->mm == mm)) {
 					lock_task_sighand(p, &flags);
-					nr += zap_process(p);
+					nr += zap_process(p, exit_code);
 					unlock_task_sighand(p, &flags);
 				}
 				break;
@@ -1748,14 +1748,19 @@ void set_dumpable(struct mm_struct *mm, int value)
 	}
 }
 
-int get_dumpable(struct mm_struct *mm)
+static int __get_dumpable(unsigned long mm_flags)
 {
 	int ret;
 
-	ret = mm->flags & 0x3;
+	ret = mm_flags & MMF_DUMPABLE_MASK;
 	return (ret >= 2) ? 2 : ret;
 }
 
+int get_dumpable(struct mm_struct *mm)
+{
+	return __get_dumpable(mm->flags);
+}
+
 static void wait_for_dump_helpers(struct file *file)
 {
 	struct pipe_inode_info *pipe;
@@ -1798,7 +1803,13 @@ void do_coredump(long signr, int exit_code, struct pt_regs *regs)
 	struct coredump_params cprm = {
 		.signr = signr,
 		.regs = regs,
-		.limit = current->signal->rlim[RLIMIT_CORE].rlim_cur,
+		.limit = rlimit(RLIMIT_CORE),
+		/*
+		 * We must use the same mm->flags while dumping core to avoid
+		 * inconsistency of bit flags, since this flag is not protected
+		 * by any locks.
+		 */
+		.mm_flags = mm->flags,
 	};
 
 	audit_core_dumps(signr);
@@ -1817,7 +1828,7 @@ void do_coredump(long signr, int exit_code, struct pt_regs *regs)
 	/*
 	 * If another thread got here first, or we are not dumpable, bail out.
 	 */
-	if (mm->core_state || !get_dumpable(mm)) {
+	if (mm->core_state || !__get_dumpable(cprm.mm_flags)) {
 		up_write(&mm->mmap_sem);
 		put_cred(cred);
 		goto fail;
@@ -1828,7 +1839,8 @@ void do_coredump(long signr, int exit_code, struct pt_regs *regs)
 	 *	process nor do we know its entire history. We only know it
 	 *	was tainted so we dump it as root in mode 2.
 	 */
-	if (get_dumpable(mm) == 2) {	/* Setuid core dump mode */
+	if (__get_dumpable(cprm.mm_flags) == 2) {
+		/* Setuid core dump mode */
 		flag = O_EXCL;		/* Stop rewrite attacks */
 		cred->fsuid = 0;	/* Dump root private */
 	}
@@ -1924,8 +1936,9 @@ void do_coredump(long signr, int exit_code, struct pt_regs *regs)
 	/*
 	 * Dont allow local users get cute and trick others to coredump
 	 * into their pre-created files:
+	 * Note, this is not relevant for pipes
 	 */
-	if (inode->i_uid != current_fsuid())
+	if (!ispipe && (inode->i_uid != current_fsuid()))
 		goto close_fail;
 	if (!cprm.file->f_op)
 		goto close_fail;
diff --git a/fs/exofs/common.h b/fs/exofs/common.h
index b1b178e61718..f0d520312d8b 100644
--- a/fs/exofs/common.h
+++ b/fs/exofs/common.h
@@ -55,6 +55,8 @@
 /* exofs Application specific page/attribute */
 # define EXOFS_APAGE_FS_DATA	(OSD_APAGE_APP_DEFINED_FIRST + 3)
 # define EXOFS_ATTR_INODE_DATA	1
+# define EXOFS_ATTR_INODE_FILE_LAYOUT	2
+# define EXOFS_ATTR_INODE_DIR_LAYOUT	3
 
 /*
  * The maximum number of files we can have is limited by the size of the
@@ -206,4 +208,41 @@ enum {
 	(((name_len) + offsetof(struct exofs_dir_entry, name)  + \
 	  EXOFS_DIR_ROUND) & ~EXOFS_DIR_ROUND)
 
+/*
+ * The on-disk (optional) layout structure.
+ * sits in an EXOFS_ATTR_INODE_FILE_LAYOUT or EXOFS_ATTR_INODE_DIR_LAYOUT
+ * attribute, attached to any inode, usually to a directory.
+ */
+
+enum exofs_inode_layout_gen_functions {
+	LAYOUT_MOVING_WINDOW = 0,
+	LAYOUT_IMPLICT = 1,
+};
+
+struct exofs_on_disk_inode_layout {
+	__le16 gen_func; /* One of enum exofs_inode_layout_gen_functions */
+	__le16 pad;
+	union {
+		/* gen_func == LAYOUT_MOVING_WINDOW (default) */
+		struct exofs_layout_sliding_window {
+			__le32 num_devices; /* first n devices in global-table*/
+		} sliding_window __packed;
+
+		/* gen_func == LAYOUT_IMPLICT */
+		struct exofs_layout_implict_list {
+			struct exofs_dt_data_map data_map;
+			/* Variable array of size data_map.cb_num_comps. These
+			 * are device indexes of the devices in the global table
+			 */
+			__le32 dev_indexes[];
+		} implict __packed;
+	};
+} __packed;
+
+static inline size_t exofs_on_disk_inode_layout_size(unsigned max_devs)
+{
+	return sizeof(struct exofs_on_disk_inode_layout) +
+		max_devs * sizeof(__le32);
+}
+
 #endif /*ifndef __EXOFS_COM_H__*/
diff --git a/fs/exofs/exofs.h b/fs/exofs/exofs.h
index c35fd4623986..8442e353309f 100644
--- a/fs/exofs/exofs.h
+++ b/fs/exofs/exofs.h
@@ -55,12 +55,28 @@
 /* u64 has problems with printk this will cast it to unsigned long long */
 #define _LLU(x) (unsigned long long)(x)
 
+struct exofs_layout {
+	osd_id		s_pid;			/* partition ID of file system*/
+
+	/* Our way of looking at the data_map */
+	unsigned stripe_unit;
+	unsigned mirrors_p1;
+
+	unsigned group_width;
+	u64	 group_depth;
+	unsigned group_count;
+
+	enum exofs_inode_layout_gen_functions lay_func;
+
+	unsigned	s_numdevs;		/* Num of devices in array    */
+	struct osd_dev	*s_ods[0];		/* Variable length            */
+};
+
 /*
  * our extension to the in-memory superblock
  */
 struct exofs_sb_info {
 	struct exofs_fscb s_fscb;		/* Written often, pre-allocate*/
-	osd_id		s_pid;			/* partition ID of file system*/
 	int		s_timeout;		/* timeout for OSD operations */
 	uint64_t	s_nextid;		/* highest object ID used     */
 	uint32_t	s_numfiles;		/* number of files on fs      */
@@ -69,22 +85,27 @@ struct exofs_sb_info {
 	atomic_t	s_curr_pending;		/* number of pending commands */
 	uint8_t		s_cred[OSD_CAP_LEN];	/* credential for the fscb    */
 
-	struct pnfs_osd_data_map data_map;	/* Default raid to use        */
-	unsigned	s_numdevs;		/* Num of devices in array    */
-	struct osd_dev	*s_ods[1];		/* Variable length, minimum 1 */
+	struct pnfs_osd_data_map data_map;	/* Default raid to use
+						 * FIXME: Needed ?
+						 */
+/*	struct exofs_layout	dir_layout;*/	/* Default dir layout */
+	struct exofs_layout	layout;		/* Default files layout,
+						 * contains the variable osd_dev
+						 * array. Keep last */
+	struct osd_dev	*_min_one_dev[1];	/* Place holder for one dev   */
 };
 
 /*
  * our extension to the in-memory inode
  */
 struct exofs_i_info {
+	struct inode   vfs_inode;          /* normal in-memory inode          */
+	wait_queue_head_t i_wq;            /* wait queue for inode            */
 	unsigned long  i_flags;            /* various atomic flags            */
 	uint32_t       i_data[EXOFS_IDATA];/*short symlink names and device #s*/
 	uint32_t       i_dir_start_lookup; /* which page to start lookup      */
-	wait_queue_head_t i_wq;            /* wait queue for inode            */
 	uint64_t       i_commit_size;      /* the object's written length     */
 	uint8_t        i_cred[OSD_CAP_LEN];/* all-powerful credential         */
-	struct inode   vfs_inode;          /* normal in-memory inode          */
 };
 
 static inline osd_id exofs_oi_objno(struct exofs_i_info *oi)
@@ -101,7 +122,7 @@ struct exofs_io_state {
 	void			*private;
 	exofs_io_done_fn	done;
 
-	struct exofs_sb_info	*sbi;
+	struct exofs_layout	*layout;
 	struct osd_obj_id	obj;
 	u8			*cred;
 
@@ -109,7 +130,11 @@ struct exofs_io_state {
 	loff_t			offset;
 	unsigned long		length;
 	void			*kern_buff;
-	struct bio		*bio;
+
+	struct page		**pages;
+	unsigned		nr_pages;
+	unsigned		pgbase;
+	unsigned		pages_consumed;
 
 	/* Attributes */
 	unsigned		in_attr_len;
@@ -122,6 +147,9 @@ struct exofs_io_state {
 	struct exofs_per_dev_state {
 		struct osd_request *or;
 		struct bio *bio;
+		loff_t offset;
+		unsigned length;
+		unsigned dev;
 	} per_dev[];
 };
 
@@ -175,6 +203,12 @@ static inline struct exofs_i_info *exofs_i(struct inode *inode)
 }
 
 /*
+ * Given a layout, object_number and stripe_index return the associated global
+ * dev_index
+ */
+unsigned exofs_layout_od_id(struct exofs_layout *layout,
+			    osd_id obj_no, unsigned layout_index);
+/*
  * Maximum count of links to a file
  */
 #define EXOFS_LINK_MAX           32000
@@ -189,7 +223,8 @@ void exofs_make_credential(u8 cred_a[OSD_CAP_LEN],
 int exofs_read_kern(struct osd_dev *od, u8 *cred, struct osd_obj_id *obj,
 		    u64 offset, void *p, unsigned length);
 
-int  exofs_get_io_state(struct exofs_sb_info *sbi, struct exofs_io_state** ios);
+int  exofs_get_io_state(struct exofs_layout *layout,
+			struct exofs_io_state **ios);
 void exofs_put_io_state(struct exofs_io_state *ios);
 
 int exofs_check_io(struct exofs_io_state *ios, u64 *resid);
@@ -226,7 +261,7 @@ int exofs_write_begin(struct file *file, struct address_space *mapping,
 		struct page **pagep, void **fsdata);
 extern struct inode *exofs_iget(struct super_block *, unsigned long);
 struct inode *exofs_new_inode(struct inode *, int);
-extern int exofs_write_inode(struct inode *, int);
+extern int exofs_write_inode(struct inode *, struct writeback_control *wbc);
 extern void exofs_delete_inode(struct inode *);
 
 /* dir.c:                */
diff --git a/fs/exofs/inode.c b/fs/exofs/inode.c
index 2afbcebeda71..a17e4b733e35 100644
--- a/fs/exofs/inode.c
+++ b/fs/exofs/inode.c
@@ -41,16 +41,18 @@
 
 enum { BIO_MAX_PAGES_KMALLOC =
 		(PAGE_SIZE - sizeof(struct bio)) / sizeof(struct bio_vec),
+	MAX_PAGES_KMALLOC =
+		PAGE_SIZE / sizeof(struct page *),
 };
 
 struct page_collect {
 	struct exofs_sb_info *sbi;
-	struct request_queue *req_q;
 	struct inode *inode;
 	unsigned expected_pages;
 	struct exofs_io_state *ios;
 
-	struct bio *bio;
+	struct page **pages;
+	unsigned alloc_pages;
 	unsigned nr_pages;
 	unsigned long length;
 	loff_t pg_first; /* keep 64bit also in 32-arches */
@@ -62,15 +64,12 @@ static void _pcol_init(struct page_collect *pcol, unsigned expected_pages,
 	struct exofs_sb_info *sbi = inode->i_sb->s_fs_info;
 
 	pcol->sbi = sbi;
-	/* Create master bios on first Q, later on cloning, each clone will be
-	 * allocated on it's destination Q
-	 */
-	pcol->req_q = osd_request_queue(sbi->s_ods[0]);
 	pcol->inode = inode;
 	pcol->expected_pages = expected_pages;
 
 	pcol->ios = NULL;
-	pcol->bio = NULL;
+	pcol->pages = NULL;
+	pcol->alloc_pages = 0;
 	pcol->nr_pages = 0;
 	pcol->length = 0;
 	pcol->pg_first = -1;
@@ -80,7 +79,8 @@ static void _pcol_reset(struct page_collect *pcol)
 {
 	pcol->expected_pages -= min(pcol->nr_pages, pcol->expected_pages);
 
-	pcol->bio = NULL;
+	pcol->pages = NULL;
+	pcol->alloc_pages = 0;
 	pcol->nr_pages = 0;
 	pcol->length = 0;
 	pcol->pg_first = -1;
@@ -90,38 +90,43 @@ static void _pcol_reset(struct page_collect *pcol)
 	 * it might not end here. don't be left with nothing
 	 */
 	if (!pcol->expected_pages)
-		pcol->expected_pages = BIO_MAX_PAGES_KMALLOC;
+		pcol->expected_pages = MAX_PAGES_KMALLOC;
 }
 
 static int pcol_try_alloc(struct page_collect *pcol)
 {
-	int pages = min_t(unsigned, pcol->expected_pages,
-			  BIO_MAX_PAGES_KMALLOC);
+	unsigned pages = min_t(unsigned, pcol->expected_pages,
+			  MAX_PAGES_KMALLOC);
 
 	if (!pcol->ios) { /* First time allocate io_state */
-		int ret = exofs_get_io_state(pcol->sbi, &pcol->ios);
+		int ret = exofs_get_io_state(&pcol->sbi->layout, &pcol->ios);
 
 		if (ret)
 			return ret;
 	}
 
+	/* TODO: easily support bio chaining */
+	pages =  min_t(unsigned, pages,
+		       pcol->sbi->layout.group_width * BIO_MAX_PAGES_KMALLOC);
+
 	for (; pages; pages >>= 1) {
-		pcol->bio = bio_kmalloc(GFP_KERNEL, pages);
-		if (likely(pcol->bio))
+		pcol->pages = kmalloc(pages * sizeof(struct page *),
+				      GFP_KERNEL);
+		if (likely(pcol->pages)) {
+			pcol->alloc_pages = pages;
 			return 0;
+		}
 	}
 
-	EXOFS_ERR("Failed to bio_kmalloc expected_pages=%u\n",
+	EXOFS_ERR("Failed to kmalloc expected_pages=%u\n",
 		  pcol->expected_pages);
 	return -ENOMEM;
 }
 
 static void pcol_free(struct page_collect *pcol)
 {
-	if (pcol->bio) {
-		bio_put(pcol->bio);
-		pcol->bio = NULL;
-	}
+	kfree(pcol->pages);
+	pcol->pages = NULL;
 
 	if (pcol->ios) {
 		exofs_put_io_state(pcol->ios);
@@ -132,11 +137,10 @@ static void pcol_free(struct page_collect *pcol)
 static int pcol_add_page(struct page_collect *pcol, struct page *page,
 			 unsigned len)
 {
-	int added_len = bio_add_pc_page(pcol->req_q, pcol->bio, page, len, 0);
-	if (unlikely(len != added_len))
+	if (unlikely(pcol->nr_pages >= pcol->alloc_pages))
 		return -ENOMEM;
 
-	++pcol->nr_pages;
+	pcol->pages[pcol->nr_pages++] = page;
 	pcol->length += len;
 	return 0;
 }
@@ -181,7 +185,6 @@ static void update_write_page(struct page *page, int ret)
  */
 static int __readpages_done(struct page_collect *pcol, bool do_unlock)
 {
-	struct bio_vec *bvec;
 	int i;
 	u64 resid;
 	u64 good_bytes;
@@ -193,13 +196,13 @@ static int __readpages_done(struct page_collect *pcol, bool do_unlock)
 	else
 		good_bytes = pcol->length - resid;
 
-	EXOFS_DBGMSG("readpages_done(0x%lx) good_bytes=0x%llx"
+	EXOFS_DBGMSG2("readpages_done(0x%lx) good_bytes=0x%llx"
 		     " length=0x%lx nr_pages=%u\n",
 		     pcol->inode->i_ino, _LLU(good_bytes), pcol->length,
 		     pcol->nr_pages);
 
-	__bio_for_each_segment(bvec, pcol->bio, i, 0) {
-		struct page *page = bvec->bv_page;
+	for (i = 0; i < pcol->nr_pages; i++) {
+		struct page *page = pcol->pages[i];
 		struct inode *inode = page->mapping->host;
 		int page_stat;
 
@@ -218,11 +221,11 @@ static int __readpages_done(struct page_collect *pcol, bool do_unlock)
 		ret = update_read_page(page, page_stat);
 		if (do_unlock)
 			unlock_page(page);
-		length += bvec->bv_len;
+		length += PAGE_SIZE;
 	}
 
 	pcol_free(pcol);
-	EXOFS_DBGMSG("readpages_done END\n");
+	EXOFS_DBGMSG2("readpages_done END\n");
 	return ret;
 }
 
@@ -238,11 +241,10 @@ static void readpages_done(struct exofs_io_state *ios, void *p)
 
 static void _unlock_pcol_pages(struct page_collect *pcol, int ret, int rw)
 {
-	struct bio_vec *bvec;
 	int i;
 
-	__bio_for_each_segment(bvec, pcol->bio, i, 0) {
-		struct page *page = bvec->bv_page;
+	for (i = 0; i < pcol->nr_pages; i++) {
+		struct page *page = pcol->pages[i];
 
 		if (rw == READ)
 			update_read_page(page, ret);
@@ -260,13 +262,14 @@ static int read_exec(struct page_collect *pcol, bool is_sync)
 	struct page_collect *pcol_copy = NULL;
 	int ret;
 
-	if (!pcol->bio)
+	if (!pcol->pages)
 		return 0;
 
 	/* see comment in _readpage() about sync reads */
 	WARN_ON(is_sync && (pcol->nr_pages != 1));
 
-	ios->bio = pcol->bio;
+	ios->pages = pcol->pages;
+	ios->nr_pages = pcol->nr_pages;
 	ios->length = pcol->length;
 	ios->offset = pcol->pg_first << PAGE_CACHE_SHIFT;
 
@@ -290,7 +293,7 @@ static int read_exec(struct page_collect *pcol, bool is_sync)
 
 	atomic_inc(&pcol->sbi->s_curr_pending);
 
-	EXOFS_DBGMSG("read_exec obj=0x%llx start=0x%llx length=0x%lx\n",
+	EXOFS_DBGMSG2("read_exec obj=0x%llx start=0x%llx length=0x%lx\n",
 		  ios->obj.id, _LLU(ios->offset), pcol->length);
 
 	/* pages ownership was passed to pcol_copy */
@@ -366,7 +369,7 @@ try_again:
 		goto try_again;
 	}
 
-	if (!pcol->bio) {
+	if (!pcol->pages) {
 		ret = pcol_try_alloc(pcol);
 		if (unlikely(ret))
 			goto fail;
@@ -448,7 +451,6 @@ static int exofs_readpage(struct file *file, struct page *page)
 static void writepages_done(struct exofs_io_state *ios, void *p)
 {
 	struct page_collect *pcol = p;
-	struct bio_vec *bvec;
 	int i;
 	u64 resid;
 	u64  good_bytes;
@@ -462,13 +464,13 @@ static void writepages_done(struct exofs_io_state *ios, void *p)
 	else
 		good_bytes = pcol->length - resid;
 
-	EXOFS_DBGMSG("writepages_done(0x%lx) good_bytes=0x%llx"
+	EXOFS_DBGMSG2("writepages_done(0x%lx) good_bytes=0x%llx"
 		     " length=0x%lx nr_pages=%u\n",
 		     pcol->inode->i_ino, _LLU(good_bytes), pcol->length,
 		     pcol->nr_pages);
 
-	__bio_for_each_segment(bvec, pcol->bio, i, 0) {
-		struct page *page = bvec->bv_page;
+	for (i = 0; i < pcol->nr_pages; i++) {
+		struct page *page = pcol->pages[i];
 		struct inode *inode = page->mapping->host;
 		int page_stat;
 
@@ -485,12 +487,12 @@ static void writepages_done(struct exofs_io_state *ios, void *p)
 		EXOFS_DBGMSG2("    writepages_done(0x%lx, 0x%lx) status=%d\n",
 			     inode->i_ino, page->index, page_stat);
 
-		length += bvec->bv_len;
+		length += PAGE_SIZE;
 	}
 
 	pcol_free(pcol);
 	kfree(pcol);
-	EXOFS_DBGMSG("writepages_done END\n");
+	EXOFS_DBGMSG2("writepages_done END\n");
 }
 
 static int write_exec(struct page_collect *pcol)
@@ -500,7 +502,7 @@ static int write_exec(struct page_collect *pcol)
 	struct page_collect *pcol_copy = NULL;
 	int ret;
 
-	if (!pcol->bio)
+	if (!pcol->pages)
 		return 0;
 
 	pcol_copy = kmalloc(sizeof(*pcol_copy), GFP_KERNEL);
@@ -512,9 +514,8 @@ static int write_exec(struct page_collect *pcol)
 
 	*pcol_copy = *pcol;
 
-	pcol_copy->bio->bi_rw |= (1 << BIO_RW); /* FIXME: bio_set_dir() */
-
-	ios->bio = pcol_copy->bio;
+	ios->pages = pcol_copy->pages;
+	ios->nr_pages = pcol_copy->nr_pages;
 	ios->offset = pcol_copy->pg_first << PAGE_CACHE_SHIFT;
 	ios->length = pcol_copy->length;
 	ios->done = writepages_done;
@@ -527,7 +528,7 @@ static int write_exec(struct page_collect *pcol)
 	}
 
 	atomic_inc(&pcol->sbi->s_curr_pending);
-	EXOFS_DBGMSG("write_exec(0x%lx, 0x%llx) start=0x%llx length=0x%lx\n",
+	EXOFS_DBGMSG2("write_exec(0x%lx, 0x%llx) start=0x%llx length=0x%lx\n",
 		  pcol->inode->i_ino, pcol->pg_first, _LLU(ios->offset),
 		  pcol->length);
 	/* pages ownership was passed to pcol_copy */
@@ -605,7 +606,7 @@ try_again:
 		goto try_again;
 	}
 
-	if (!pcol->bio) {
+	if (!pcol->pages) {
 		ret = pcol_try_alloc(pcol);
 		if (unlikely(ret))
 			goto fail;
@@ -616,7 +617,7 @@ try_again:
 
 	ret = pcol_add_page(pcol, page, len);
 	if (unlikely(ret)) {
-		EXOFS_DBGMSG("Failed pcol_add_page "
+		EXOFS_DBGMSG2("Failed pcol_add_page "
 			     "nr_pages=%u total_length=0x%lx\n",
 			     pcol->nr_pages, pcol->length);
 
@@ -663,7 +664,7 @@ static int exofs_writepages(struct address_space *mapping,
 	if (expected_pages < 32L)
 		expected_pages = 32L;
 
-	EXOFS_DBGMSG("inode(0x%lx) wbc->start=0x%llx wbc->end=0x%llx "
+	EXOFS_DBGMSG2("inode(0x%lx) wbc->start=0x%llx wbc->end=0x%llx "
 		     "nrpages=%lu start=0x%lx end=0x%lx expected_pages=%ld\n",
 		     mapping->host->i_ino, wbc->range_start, wbc->range_end,
 		     mapping->nrpages, start, end, expected_pages);
@@ -859,20 +860,33 @@ int exofs_setattr(struct dentry *dentry, struct iattr *iattr)
 	return error;
 }
 
+static const struct osd_attr g_attr_inode_file_layout = ATTR_DEF(
+	EXOFS_APAGE_FS_DATA,
+	EXOFS_ATTR_INODE_FILE_LAYOUT,
+	0);
+static const struct osd_attr g_attr_inode_dir_layout = ATTR_DEF(
+	EXOFS_APAGE_FS_DATA,
+	EXOFS_ATTR_INODE_DIR_LAYOUT,
+	0);
+
 /*
- * Read an inode from the OSD, and return it as is.  We also return the size
- * attribute in the 'obj_size' argument.
+ * Read the Linux inode info from the OSD, and return it as is. In exofs the
+ * inode info is in an application specific page/attribute of the osd-object.
  */
 static int exofs_get_inode(struct super_block *sb, struct exofs_i_info *oi,
-		    struct exofs_fcb *inode, uint64_t *obj_size)
+		    struct exofs_fcb *inode)
 {
 	struct exofs_sb_info *sbi = sb->s_fs_info;
-	struct osd_attr attrs[2];
+	struct osd_attr attrs[] = {
+		[0] = g_attr_inode_data,
+		[1] = g_attr_inode_file_layout,
+		[2] = g_attr_inode_dir_layout,
+	};
 	struct exofs_io_state *ios;
+	struct exofs_on_disk_inode_layout *layout;
 	int ret;
 
-	*obj_size = ~0;
-	ret = exofs_get_io_state(sbi, &ios);
+	ret = exofs_get_io_state(&sbi->layout, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: exofs_get_io_state failed.\n", __func__);
 		return ret;
@@ -882,14 +896,25 @@ static int exofs_get_inode(struct super_block *sb, struct exofs_i_info *oi,
 	exofs_make_credential(oi->i_cred, &ios->obj);
 	ios->cred = oi->i_cred;
 
-	attrs[0] = g_attr_inode_data;
-	attrs[1] = g_attr_logical_length;
+	attrs[1].len = exofs_on_disk_inode_layout_size(sbi->layout.s_numdevs);
+	attrs[2].len = exofs_on_disk_inode_layout_size(sbi->layout.s_numdevs);
+
 	ios->in_attr = attrs;
 	ios->in_attr_len = ARRAY_SIZE(attrs);
 
 	ret = exofs_sbi_read(ios);
-	if (ret)
+	if (unlikely(ret)) {
+		EXOFS_ERR("object(0x%llx) corrupted, return empty file=>%d\n",
+			  _LLU(ios->obj.id), ret);
+		memset(inode, 0, sizeof(*inode));
+		inode->i_mode = 0040000 | (0777 & ~022);
+		/* If object is lost on target we might as well enable it's
+		 * delete.
+		 */
+		if ((ret == -ENOENT) || (ret == -EINVAL))
+			ret = 0;
 		goto out;
+	}
 
 	ret = extract_attr_from_ios(ios, &attrs[0]);
 	if (ret) {
@@ -901,11 +926,33 @@ static int exofs_get_inode(struct super_block *sb, struct exofs_i_info *oi,
 
 	ret = extract_attr_from_ios(ios, &attrs[1]);
 	if (ret) {
-		EXOFS_ERR("%s: extract_attr of logical_length failed\n",
-			  __func__);
+		EXOFS_ERR("%s: extract_attr of inode_data failed\n", __func__);
+		goto out;
+	}
+	if (attrs[1].len) {
+		layout = attrs[1].val_ptr;
+		if (layout->gen_func != cpu_to_le16(LAYOUT_MOVING_WINDOW)) {
+			EXOFS_ERR("%s: unsupported files layout %d\n",
+				__func__, layout->gen_func);
+			ret = -ENOTSUPP;
+			goto out;
+		}
+	}
+
+	ret = extract_attr_from_ios(ios, &attrs[2]);
+	if (ret) {
+		EXOFS_ERR("%s: extract_attr of inode_data failed\n", __func__);
 		goto out;
 	}
-	*obj_size = get_unaligned_be64(attrs[1].val_ptr);
+	if (attrs[2].len) {
+		layout = attrs[2].val_ptr;
+		if (layout->gen_func != cpu_to_le16(LAYOUT_MOVING_WINDOW)) {
+			EXOFS_ERR("%s: unsupported meta-data layout %d\n",
+				__func__, layout->gen_func);
+			ret = -ENOTSUPP;
+			goto out;
+		}
+	}
 
 out:
 	exofs_put_io_state(ios);
@@ -925,7 +972,6 @@ struct inode *exofs_iget(struct super_block *sb, unsigned long ino)
 	struct exofs_i_info *oi;
 	struct exofs_fcb fcb;
 	struct inode *inode;
-	uint64_t obj_size;
 	int ret;
 
 	inode = iget_locked(sb, ino);
@@ -937,7 +983,7 @@ struct inode *exofs_iget(struct super_block *sb, unsigned long ino)
 	__oi_init(oi);
 
 	/* read the inode from the osd */
-	ret = exofs_get_inode(sb, oi, &fcb, &obj_size);
+	ret = exofs_get_inode(sb, oi, &fcb);
 	if (ret)
 		goto bad_inode;
 
@@ -958,13 +1004,6 @@ struct inode *exofs_iget(struct super_block *sb, unsigned long ino)
 	inode->i_blkbits = EXOFS_BLKSHIFT;
 	inode->i_generation = le32_to_cpu(fcb.i_generation);
 
-	if ((inode->i_size != obj_size) &&
-		(!exofs_inode_is_fast_symlink(inode))) {
-		EXOFS_ERR("WARNING: Size of inode=%llu != object=%llu\n",
-			  inode->i_size, _LLU(obj_size));
-		/* FIXME: call exofs_inode_recovery() */
-	}
-
 	oi->i_dir_start_lookup = 0;
 
 	if ((inode->i_nlink == 0) && (inode->i_mode == 0)) {
@@ -1043,7 +1082,7 @@ static void create_done(struct exofs_io_state *ios, void *p)
 
 	if (unlikely(ret)) {
 		EXOFS_ERR("object=0x%llx creation faild in pid=0x%llx",
-			  _LLU(exofs_oi_objno(oi)), _LLU(sbi->s_pid));
+			  _LLU(exofs_oi_objno(oi)), _LLU(sbi->layout.s_pid));
 		/*TODO: When FS is corrupted creation can fail, object already
 		 * exist. Get rid of this asynchronous creation, if exist
 		 * increment the obj counter and try the next object. Until we
@@ -1104,7 +1143,7 @@ struct inode *exofs_new_inode(struct inode *dir, int mode)
 
 	mark_inode_dirty(inode);
 
-	ret = exofs_get_io_state(sbi, &ios);
+	ret = exofs_get_io_state(&sbi->layout, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("exofs_new_inode: exofs_get_io_state failed\n");
 		return ERR_PTR(ret);
@@ -1170,8 +1209,10 @@ static int exofs_update_inode(struct inode *inode, int do_sync)
 	int ret;
 
 	args = kzalloc(sizeof(*args), GFP_KERNEL);
-	if (!args)
+	if (!args) {
+		EXOFS_DBGMSG("Faild kzalloc of args\n");
 		return -ENOMEM;
+	}
 
 	fcb = &args->fcb;
 
@@ -1200,7 +1241,7 @@ static int exofs_update_inode(struct inode *inode, int do_sync)
 	} else
 		memcpy(fcb->i_data, oi->i_data, sizeof(fcb->i_data));
 
-	ret = exofs_get_io_state(sbi, &ios);
+	ret = exofs_get_io_state(&sbi->layout, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: exofs_get_io_state failed.\n", __func__);
 		goto free_args;
@@ -1234,13 +1275,14 @@ static int exofs_update_inode(struct inode *inode, int do_sync)
 free_args:
 	kfree(args);
 out:
-	EXOFS_DBGMSG("ret=>%d\n", ret);
+	EXOFS_DBGMSG("(0x%lx) do_sync=%d ret=>%d\n",
+		     inode->i_ino, do_sync, ret);
 	return ret;
 }
 
-int exofs_write_inode(struct inode *inode, int wait)
+int exofs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
-	return exofs_update_inode(inode, wait);
+	return exofs_update_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
 }
 
 /*
@@ -1283,7 +1325,7 @@ void exofs_delete_inode(struct inode *inode)
 
 	clear_inode(inode);
 
-	ret = exofs_get_io_state(sbi, &ios);
+	ret = exofs_get_io_state(&sbi->layout, &ios);
 	if (unlikely(ret)) {
 		EXOFS_ERR("%s: exofs_get_io_state failed\n", __func__);
 		return;
diff --git a/fs/exofs/ios.c b/fs/exofs/ios.c
index 5bad01fa1f9f..5293bc411d17 100644
--- a/fs/exofs/ios.c
+++ b/fs/exofs/ios.c
@@ -23,9 +23,13 @@
  */
 
 #include <scsi/scsi_device.h>
+#include <asm/div64.h>
 
 #include "exofs.h"
 
+#define EXOFS_DBGMSG2(M...) do {} while (0)
+/* #define EXOFS_DBGMSG2 EXOFS_DBGMSG */
+
 void exofs_make_credential(u8 cred_a[OSD_CAP_LEN], const struct osd_obj_id *obj)
 {
 	osd_sec_init_nosec_doall_caps(cred_a, obj, false, true);
@@ -64,21 +68,24 @@ out:
 	return ret;
 }
 
-int exofs_get_io_state(struct exofs_sb_info *sbi, struct exofs_io_state** pios)
+int exofs_get_io_state(struct exofs_layout *layout,
+		       struct exofs_io_state **pios)
 {
 	struct exofs_io_state *ios;
 
 	/*TODO: Maybe use kmem_cach per sbi of size
-	 * exofs_io_state_size(sbi->s_numdevs)
+	 * exofs_io_state_size(layout->s_numdevs)
 	 */
-	ios = kzalloc(exofs_io_state_size(sbi->s_numdevs), GFP_KERNEL);
+	ios = kzalloc(exofs_io_state_size(layout->s_numdevs), GFP_KERNEL);
 	if (unlikely(!ios)) {
+		EXOFS_DBGMSG("Faild kzalloc bytes=%d\n",
+			     exofs_io_state_size(layout->s_numdevs));
 		*pios = NULL;
 		return -ENOMEM;
 	}
 
-	ios->sbi = sbi;
-	ios->obj.partition = sbi->s_pid;
+	ios->layout = layout;
+	ios->obj.partition = layout->s_pid;
 	*pios = ios;
 	return 0;
 }
@@ -101,6 +108,29 @@ void exofs_put_io_state(struct exofs_io_state *ios)
 	}
 }
 
+unsigned exofs_layout_od_id(struct exofs_layout *layout,
+			    osd_id obj_no, unsigned layout_index)
+{
+/*	switch (layout->lay_func) {
+	case LAYOUT_MOVING_WINDOW:
+	{*/
+		unsigned dev_mod = obj_no;
+
+		return (layout_index + dev_mod * layout->mirrors_p1) %
+							      layout->s_numdevs;
+/*	}
+	case LAYOUT_FUNC_IMPLICT:
+		return layout->devs[layout_index];
+	}*/
+}
+
+static inline struct osd_dev *exofs_ios_od(struct exofs_io_state *ios,
+					   unsigned layout_index)
+{
+	return ios->layout->s_ods[
+		exofs_layout_od_id(ios->layout, ios->obj.id, layout_index)];
+}
+
 static void _sync_done(struct exofs_io_state *ios, void *p)
 {
 	struct completion *waiting = p;
@@ -168,6 +198,21 @@ static int exofs_io_execute(struct exofs_io_state *ios)
 	return ret;
 }
 
+static void _clear_bio(struct bio *bio)
+{
+	struct bio_vec *bv;
+	unsigned i;
+
+	__bio_for_each_segment(bv, bio, i, 0) {
+		unsigned this_count = bv->bv_len;
+
+		if (likely(PAGE_SIZE == this_count))
+			clear_highpage(bv->bv_page);
+		else
+			zero_user(bv->bv_page, bv->bv_offset, this_count);
+	}
+}
+
 int exofs_check_io(struct exofs_io_state *ios, u64 *resid)
 {
 	enum osd_err_priority acumulated_osd_err = 0;
@@ -176,16 +221,25 @@ int exofs_check_io(struct exofs_io_state *ios, u64 *resid)
 
 	for (i = 0; i < ios->numdevs; i++) {
 		struct osd_sense_info osi;
-		int ret = osd_req_decode_sense(ios->per_dev[i].or, &osi);
+		struct osd_request *or = ios->per_dev[i].or;
+		int ret;
+
+		if (unlikely(!or))
+			continue;
 
+		ret = osd_req_decode_sense(or, &osi);
 		if (likely(!ret))
 			continue;
 
-		if (unlikely(ret == -EFAULT)) {
-			EXOFS_DBGMSG("%s: EFAULT Need page clear\n", __func__);
-			/*FIXME: All the pages in this device range should:
-			 *	clear_highpage(page);
-			 */
+		if (OSD_ERR_PRI_CLEAR_PAGES == osi.osd_err_pri) {
+			/* start read offset passed endof file */
+			_clear_bio(ios->per_dev[i].bio);
+			EXOFS_DBGMSG("start read offset passed end of file "
+				"offset=0x%llx, length=0x%llx\n",
+				_LLU(ios->per_dev[i].offset),
+				_LLU(ios->per_dev[i].length));
+
+			continue; /* we recovered */
 		}
 
 		if (osi.osd_err_pri >= acumulated_osd_err) {
@@ -205,14 +259,259 @@ int exofs_check_io(struct exofs_io_state *ios, u64 *resid)
 	return acumulated_lin_err;
 }
 
+/*
+ * L - logical offset into the file
+ *
+ * U - The number of bytes in a stripe within a group
+ *
+ *	U = stripe_unit * group_width
+ *
+ * T - The number of bytes striped within a group of component objects
+ *     (before advancing to the next group)
+ *
+ *	T = stripe_unit * group_width * group_depth
+ *
+ * S - The number of bytes striped across all component objects
+ *     before the pattern repeats
+ *
+ *	S = stripe_unit * group_width * group_depth * group_count
+ *
+ * M - The "major" (i.e., across all components) stripe number
+ *
+ *	M = L / S
+ *
+ * G - Counts the groups from the beginning of the major stripe
+ *
+ *	G = (L - (M * S)) / T	[or (L % S) / T]
+ *
+ * H - The byte offset within the group
+ *
+ *	H = (L - (M * S)) % T	[or (L % S) % T]
+ *
+ * N - The "minor" (i.e., across the group) stripe number
+ *
+ *	N = H / U
+ *
+ * C - The component index coresponding to L
+ *
+ *	C = (H - (N * U)) / stripe_unit + G * group_width
+ *	[or (L % U) / stripe_unit + G * group_width]
+ *
+ * O - The component offset coresponding to L
+ *
+ *	O = L % stripe_unit + N * stripe_unit + M * group_depth * stripe_unit
+ */
+struct _striping_info {
+	u64 obj_offset;
+	u64 group_length;
+	u64 total_group_length;
+	u64 Major;
+	unsigned dev;
+	unsigned unit_off;
+};
+
+static void _calc_stripe_info(struct exofs_io_state *ios, u64 file_offset,
+			      struct _striping_info *si)
+{
+	u32	stripe_unit = ios->layout->stripe_unit;
+	u32	group_width = ios->layout->group_width;
+	u64	group_depth = ios->layout->group_depth;
+
+	u32	U = stripe_unit * group_width;
+	u64	T = U * group_depth;
+	u64	S = T * ios->layout->group_count;
+	u64	M = div64_u64(file_offset, S);
+
+	/*
+	G = (L - (M * S)) / T
+	H = (L - (M * S)) % T
+	*/
+	u64	LmodS = file_offset - M * S;
+	u32	G = div64_u64(LmodS, T);
+	u64	H = LmodS - G * T;
+
+	u32	N = div_u64(H, U);
+
+	/* "H - (N * U)" is just "H % U" so it's bound to u32 */
+	si->dev = (u32)(H - (N * U)) / stripe_unit + G * group_width;
+	si->dev *= ios->layout->mirrors_p1;
+
+	div_u64_rem(file_offset, stripe_unit, &si->unit_off);
+
+	si->obj_offset = si->unit_off + (N * stripe_unit) +
+				  (M * group_depth * stripe_unit);
+
+	si->group_length = T - H;
+	si->total_group_length = T;
+	si->Major = M;
+}
+
+static int _add_stripe_unit(struct exofs_io_state *ios,  unsigned *cur_pg,
+		unsigned pgbase, struct exofs_per_dev_state *per_dev,
+		int cur_len)
+{
+	unsigned pg = *cur_pg;
+	struct request_queue *q =
+			osd_request_queue(exofs_ios_od(ios, per_dev->dev));
+
+	per_dev->length += cur_len;
+
+	if (per_dev->bio == NULL) {
+		unsigned pages_in_stripe = ios->layout->group_width *
+					(ios->layout->stripe_unit / PAGE_SIZE);
+		unsigned bio_size = (ios->nr_pages + pages_in_stripe) /
+						ios->layout->group_width;
+
+		per_dev->bio = bio_kmalloc(GFP_KERNEL, bio_size);
+		if (unlikely(!per_dev->bio)) {
+			EXOFS_DBGMSG("Faild to allocate BIO size=%u\n",
+				     bio_size);
+			return -ENOMEM;
+		}
+	}
+
+	while (cur_len > 0) {
+		unsigned pglen = min_t(unsigned, PAGE_SIZE - pgbase, cur_len);
+		unsigned added_len;
+
+		BUG_ON(ios->nr_pages <= pg);
+		cur_len -= pglen;
+
+		added_len = bio_add_pc_page(q, per_dev->bio, ios->pages[pg],
+					    pglen, pgbase);
+		if (unlikely(pglen != added_len))
+			return -ENOMEM;
+		pgbase = 0;
+		++pg;
+	}
+	BUG_ON(cur_len);
+
+	*cur_pg = pg;
+	return 0;
+}
+
+static int _prepare_one_group(struct exofs_io_state *ios, u64 length,
+			      struct _striping_info *si, unsigned first_comp)
+{
+	unsigned stripe_unit = ios->layout->stripe_unit;
+	unsigned mirrors_p1 = ios->layout->mirrors_p1;
+	unsigned devs_in_group = ios->layout->group_width * mirrors_p1;
+	unsigned dev = si->dev;
+	unsigned first_dev = dev - (dev % devs_in_group);
+	unsigned comp = first_comp + (dev - first_dev);
+	unsigned max_comp = ios->numdevs ? ios->numdevs - mirrors_p1 : 0;
+	unsigned cur_pg = ios->pages_consumed;
+	int ret = 0;
+
+	while (length) {
+		struct exofs_per_dev_state *per_dev = &ios->per_dev[comp];
+		unsigned cur_len, page_off = 0;
+
+		if (!per_dev->length) {
+			per_dev->dev = dev;
+			if (dev < si->dev) {
+				per_dev->offset = si->obj_offset + stripe_unit -
+								   si->unit_off;
+				cur_len = stripe_unit;
+			} else if (dev == si->dev) {
+				per_dev->offset = si->obj_offset;
+				cur_len = stripe_unit - si->unit_off;
+				page_off = si->unit_off & ~PAGE_MASK;
+				BUG_ON(page_off && (page_off != ios->pgbase));
+			} else { /* dev > si->dev */
+				per_dev->offset = si->obj_offset - si->unit_off;
+				cur_len = stripe_unit;
+			}
+
+			if (max_comp < comp)
+				max_comp = comp;
+
+			dev += mirrors_p1;
+			dev = (dev % devs_in_group) + first_dev;
+		} else {
+			cur_len = stripe_unit;
+		}
+		if (cur_len >= length)
+			cur_len = length;
+
+		ret = _add_stripe_unit(ios, &cur_pg, page_off , per_dev,
+				       cur_len);
+		if (unlikely(ret))
+			goto out;
+
+		comp += mirrors_p1;
+		comp = (comp % devs_in_group) + first_comp;
+
+		length -= cur_len;
+	}
+out:
+	ios->numdevs = max_comp + mirrors_p1;
+	ios->pages_consumed = cur_pg;
+	return ret;
+}
+
+static int _prepare_for_striping(struct exofs_io_state *ios)
+{
+	u64 length = ios->length;
+	struct _striping_info si;
+	unsigned devs_in_group = ios->layout->group_width *
+				 ios->layout->mirrors_p1;
+	unsigned first_comp = 0;
+	int ret = 0;
+
+	_calc_stripe_info(ios, ios->offset, &si);
+
+	if (!ios->pages) {
+		if (ios->kern_buff) {
+			struct exofs_per_dev_state *per_dev = &ios->per_dev[0];
+
+			per_dev->offset = si.obj_offset;
+			per_dev->dev = si.dev;
+
+			/* no cross device without page array */
+			BUG_ON((ios->layout->group_width > 1) &&
+			       (si.unit_off + ios->length >
+				ios->layout->stripe_unit));
+		}
+		ios->numdevs = ios->layout->mirrors_p1;
+		return 0;
+	}
+
+	while (length) {
+		if (length < si.group_length)
+			si.group_length = length;
+
+		ret = _prepare_one_group(ios, si.group_length, &si, first_comp);
+		if (unlikely(ret))
+			goto out;
+
+		length -= si.group_length;
+
+		si.group_length = si.total_group_length;
+		si.unit_off = 0;
+		++si.Major;
+		si.obj_offset = si.Major * ios->layout->stripe_unit *
+						ios->layout->group_depth;
+
+		si.dev = (si.dev - (si.dev % devs_in_group)) + devs_in_group;
+		si.dev %= ios->layout->s_numdevs;
+
+		first_comp += devs_in_group;
+		first_comp %= ios->layout->s_numdevs;
+	}
+
+out:
+	return ret;
+}
+
 int exofs_sbi_create(struct exofs_io_state *ios)
 {
 	int i, ret;
 
-	for (i = 0; i < ios->sbi->s_numdevs; i++) {
+	for (i = 0; i < ios->layout->s_numdevs; i++) {
 		struct osd_request *or;
 
-		or = osd_start_request(ios->sbi->s_ods[i], GFP_KERNEL);
+		or = osd_start_request(exofs_ios_od(ios, i), GFP_KERNEL);
 		if (unlikely(!or)) {
 			EXOFS_ERR("%s: osd_start_request failed\n", __func__);
 			ret = -ENOMEM;
@@ -233,10 +532,10 @@ int exofs_sbi_remove(struct exofs_io_state *ios)
 {
 	int i, ret;
 
-	for (i = 0; i < ios->sbi->s_numdevs; i++) {
+	for (i = 0; i < ios->layout->s_numdevs; i++) {
 		struct osd_request *or;
 
-		or = osd_start_request(ios->sbi->s_ods[i], GFP_KERNEL);
+		or = osd_start_request(exofs_ios_od(ios, i), GFP_KERNEL);
 		if (unlikely(!or)) {
 			EXOFS_ERR("%s: osd_start_request failed\n", __func__);
 			ret = -ENOMEM;
@@ -253,51 +552,74 @@ out:
 	return ret;
 }
 
-int exofs_sbi_write(struct exofs_io_state *ios)
+static int _sbi_write_mirror(struct exofs_io_state *ios, int cur_comp)
 {
-	int i, ret;
+	struct exofs_per_dev_state *master_dev = &ios->per_dev[cur_comp];
+	unsigned dev = ios->per_dev[cur_comp].dev;
+	unsigned last_comp = cur_comp + ios->layout->mirrors_p1;
+	int ret = 0;
 
-	for (i = 0; i < ios->sbi->s_numdevs; i++) {
+	if (ios->pages && !master_dev->length)
+		return 0; /* Just an empty slot */
+
+	for (; cur_comp < last_comp; ++cur_comp, ++dev) {
+		struct exofs_per_dev_state *per_dev = &ios->per_dev[cur_comp];
 		struct osd_request *or;
 
-		or = osd_start_request(ios->sbi->s_ods[i], GFP_KERNEL);
+		or = osd_start_request(exofs_ios_od(ios, dev), GFP_KERNEL);
 		if (unlikely(!or)) {
 			EXOFS_ERR("%s: osd_start_request failed\n", __func__);
 			ret = -ENOMEM;
 			goto out;
 		}
-		ios->per_dev[i].or = or;
-		ios->numdevs++;
+		per_dev->or = or;
+		per_dev->offset = master_dev->offset;
 
-		if (ios->bio) {
+		if (ios->pages) {
 			struct bio *bio;
 
-			if (i != 0) {
+			if (per_dev != master_dev) {
 				bio = bio_kmalloc(GFP_KERNEL,
-						  ios->bio->bi_max_vecs);
+						  master_dev->bio->bi_max_vecs);
 				if (unlikely(!bio)) {
+					EXOFS_DBGMSG(
+					      "Faild to allocate BIO size=%u\n",
+					      master_dev->bio->bi_max_vecs);
 					ret = -ENOMEM;
 					goto out;
 				}
 
-				__bio_clone(bio, ios->bio);
+				__bio_clone(bio, master_dev->bio);
 				bio->bi_bdev = NULL;
 				bio->bi_next = NULL;
-				ios->per_dev[i].bio =  bio;
+				per_dev->length = master_dev->length;
+				per_dev->bio =  bio;
+				per_dev->dev = dev;
 			} else {
-				bio = ios->bio;
+				bio = master_dev->bio;
+				/* FIXME: bio_set_dir() */
+				bio->bi_rw |= (1 << BIO_RW);
 			}
 
-			osd_req_write(or, &ios->obj, ios->offset, bio,
-				      ios->length);
-/*			EXOFS_DBGMSG("write sync=%d\n", sync);*/
+			osd_req_write(or, &ios->obj, per_dev->offset, bio,
+				      per_dev->length);
+			EXOFS_DBGMSG("write(0x%llx) offset=0x%llx "
+				      "length=0x%llx dev=%d\n",
+				     _LLU(ios->obj.id), _LLU(per_dev->offset),
+				     _LLU(per_dev->length), dev);
 		} else if (ios->kern_buff) {
-			osd_req_write_kern(or, &ios->obj, ios->offset,
+			ret = osd_req_write_kern(or, &ios->obj, per_dev->offset,
 					   ios->kern_buff, ios->length);
-/*			EXOFS_DBGMSG("write_kern sync=%d\n", sync);*/
+			if (unlikely(ret))
+				goto out;
+			EXOFS_DBGMSG2("write_kern(0x%llx) offset=0x%llx "
+				      "length=0x%llx dev=%d\n",
+				     _LLU(ios->obj.id), _LLU(per_dev->offset),
+				     _LLU(ios->length), dev);
 		} else {
 			osd_req_set_attributes(or, &ios->obj);
-/*			EXOFS_DBGMSG("set_attributes sync=%d\n", sync);*/
+			EXOFS_DBGMSG2("obj(0x%llx) set_attributes=%d dev=%d\n",
+				     _LLU(ios->obj.id), ios->out_attr_len, dev);
 		}
 
 		if (ios->out_attr)
@@ -308,54 +630,93 @@ int exofs_sbi_write(struct exofs_io_state *ios)
 			osd_req_add_get_attr_list(or, ios->in_attr,
 						  ios->in_attr_len);
 	}
-	ret = exofs_io_execute(ios);
 
 out:
 	return ret;
 }
 
-int exofs_sbi_read(struct exofs_io_state *ios)
+int exofs_sbi_write(struct exofs_io_state *ios)
 {
-	int i, ret;
+	int i;
+	int ret;
 
-	for (i = 0; i < 1; i++) {
-		struct osd_request *or;
-		unsigned first_dev = (unsigned)ios->obj.id;
+	ret = _prepare_for_striping(ios);
+	if (unlikely(ret))
+		return ret;
 
-		first_dev %= ios->sbi->s_numdevs;
-		or = osd_start_request(ios->sbi->s_ods[first_dev], GFP_KERNEL);
-		if (unlikely(!or)) {
-			EXOFS_ERR("%s: osd_start_request failed\n", __func__);
-			ret = -ENOMEM;
-			goto out;
-		}
-		ios->per_dev[i].or = or;
-		ios->numdevs++;
+	for (i = 0; i < ios->numdevs; i += ios->layout->mirrors_p1) {
+		ret = _sbi_write_mirror(ios, i);
+		if (unlikely(ret))
+			return ret;
+	}
 
-		if (ios->bio) {
-			osd_req_read(or, &ios->obj, ios->offset, ios->bio,
-				     ios->length);
-/*			EXOFS_DBGMSG("read sync=%d\n", sync);*/
-		} else if (ios->kern_buff) {
-			osd_req_read_kern(or, &ios->obj, ios->offset,
-					   ios->kern_buff, ios->length);
-/*			EXOFS_DBGMSG("read_kern sync=%d\n", sync);*/
-		} else {
-			osd_req_get_attributes(or, &ios->obj);
-/*			EXOFS_DBGMSG("get_attributes sync=%d\n", sync);*/
-		}
+	ret = exofs_io_execute(ios);
+	return ret;
+}
 
-		if (ios->out_attr)
-			osd_req_add_set_attr_list(or, ios->out_attr,
-						  ios->out_attr_len);
+static int _sbi_read_mirror(struct exofs_io_state *ios, unsigned cur_comp)
+{
+	struct osd_request *or;
+	struct exofs_per_dev_state *per_dev = &ios->per_dev[cur_comp];
+	unsigned first_dev = (unsigned)ios->obj.id;
 
-		if (ios->in_attr)
-			osd_req_add_get_attr_list(or, ios->in_attr,
-						  ios->in_attr_len);
+	if (ios->pages && !per_dev->length)
+		return 0; /* Just an empty slot */
+
+	first_dev = per_dev->dev + first_dev % ios->layout->mirrors_p1;
+	or = osd_start_request(exofs_ios_od(ios, first_dev), GFP_KERNEL);
+	if (unlikely(!or)) {
+		EXOFS_ERR("%s: osd_start_request failed\n", __func__);
+		return -ENOMEM;
 	}
-	ret = exofs_io_execute(ios);
+	per_dev->or = or;
+
+	if (ios->pages) {
+		osd_req_read(or, &ios->obj, per_dev->offset,
+				per_dev->bio, per_dev->length);
+		EXOFS_DBGMSG("read(0x%llx) offset=0x%llx length=0x%llx"
+			     " dev=%d\n", _LLU(ios->obj.id),
+			     _LLU(per_dev->offset), _LLU(per_dev->length),
+			     first_dev);
+	} else if (ios->kern_buff) {
+		int ret = osd_req_read_kern(or, &ios->obj, per_dev->offset,
+					    ios->kern_buff, ios->length);
+		EXOFS_DBGMSG2("read_kern(0x%llx) offset=0x%llx "
+			      "length=0x%llx dev=%d ret=>%d\n",
+			      _LLU(ios->obj.id), _LLU(per_dev->offset),
+			      _LLU(ios->length), first_dev, ret);
+		if (unlikely(ret))
+			return ret;
+	} else {
+		osd_req_get_attributes(or, &ios->obj);
+		EXOFS_DBGMSG2("obj(0x%llx) get_attributes=%d dev=%d\n",
+			      _LLU(ios->obj.id), ios->in_attr_len, first_dev);
+	}
+	if (ios->out_attr)
+		osd_req_add_set_attr_list(or, ios->out_attr, ios->out_attr_len);
 
-out:
+	if (ios->in_attr)
+		osd_req_add_get_attr_list(or, ios->in_attr, ios->in_attr_len);
+
+	return 0;
+}
+
+int exofs_sbi_read(struct exofs_io_state *ios)
+{
+	int i;
+	int ret;
+
+	ret = _prepare_for_striping(ios);
+	if (unlikely(ret))
+		return ret;
+
+	for (i = 0; i < ios->numdevs; i += ios->layout->mirrors_p1) {
+		ret = _sbi_read_mirror(ios, i);
+		if (unlikely(ret))
+			return ret;
+	}
+
+	ret = exofs_io_execute(ios);
 	return ret;
 }
 
@@ -380,42 +741,82 @@ int extract_attr_from_ios(struct exofs_io_state *ios, struct osd_attr *attr)
 	return -EIO;
 }
 
+static int _truncate_mirrors(struct exofs_io_state *ios, unsigned cur_comp,
+			     struct osd_attr *attr)
+{
+	int last_comp = cur_comp + ios->layout->mirrors_p1;
+
+	for (; cur_comp < last_comp; ++cur_comp) {
+		struct exofs_per_dev_state *per_dev = &ios->per_dev[cur_comp];
+		struct osd_request *or;
+
+		or = osd_start_request(exofs_ios_od(ios, cur_comp), GFP_KERNEL);
+		if (unlikely(!or)) {
+			EXOFS_ERR("%s: osd_start_request failed\n", __func__);
+			return -ENOMEM;
+		}
+		per_dev->or = or;
+
+		osd_req_set_attributes(or, &ios->obj);
+		osd_req_add_set_attr_list(or, attr, 1);
+	}
+
+	return 0;
+}
+
 int exofs_oi_truncate(struct exofs_i_info *oi, u64 size)
 {
 	struct exofs_sb_info *sbi = oi->vfs_inode.i_sb->s_fs_info;
 	struct exofs_io_state *ios;
-	struct osd_attr attr;
-	__be64 newsize;
+	struct exofs_trunc_attr {
+		struct osd_attr attr;
+		__be64 newsize;
+	} *size_attrs;
+	struct _striping_info si;
 	int i, ret;
 
-	if (exofs_get_io_state(sbi, &ios))
-		return -ENOMEM;
+	ret = exofs_get_io_state(&sbi->layout, &ios);
+	if (unlikely(ret))
+		return ret;
+
+	size_attrs = kcalloc(ios->layout->group_width, sizeof(*size_attrs),
+			     GFP_KERNEL);
+	if (unlikely(!size_attrs)) {
+		ret = -ENOMEM;
+		goto out;
+	}
 
 	ios->obj.id = exofs_oi_objno(oi);
 	ios->cred = oi->i_cred;
 
-	newsize = cpu_to_be64(size);
-	attr = g_attr_logical_length;
-	attr.val_ptr = &newsize;
+	ios->numdevs = ios->layout->s_numdevs;
+	_calc_stripe_info(ios, size, &si);
 
-	for (i = 0; i < sbi->s_numdevs; i++) {
-		struct osd_request *or;
+	for (i = 0; i < ios->layout->group_width; ++i) {
+		struct exofs_trunc_attr *size_attr = &size_attrs[i];
+		u64 obj_size;
 
-		or = osd_start_request(sbi->s_ods[i], GFP_KERNEL);
-		if (unlikely(!or)) {
-			EXOFS_ERR("%s: osd_start_request failed\n", __func__);
-			ret = -ENOMEM;
-			goto out;
-		}
-		ios->per_dev[i].or = or;
-		ios->numdevs++;
+		if (i < si.dev)
+			obj_size = si.obj_offset +
+					ios->layout->stripe_unit - si.unit_off;
+		else if (i == si.dev)
+			obj_size = si.obj_offset;
+		else /* i > si.dev */
+			obj_size = si.obj_offset - si.unit_off;
 
-		osd_req_set_attributes(or, &ios->obj);
-		osd_req_add_set_attr_list(or, &attr, 1);
+		size_attr->newsize = cpu_to_be64(obj_size);
+		size_attr->attr = g_attr_logical_length;
+		size_attr->attr.val_ptr = &size_attr->newsize;
+
+		ret = _truncate_mirrors(ios, i * ios->layout->mirrors_p1,
+					&size_attr->attr);
+		if (unlikely(ret))
+			goto out;
 	}
 	ret = exofs_io_execute(ios);
 
 out:
+	kfree(size_attrs);
 	exofs_put_io_state(ios);
 	return ret;
 }
diff --git a/fs/exofs/super.c b/fs/exofs/super.c
index a1d1e77b12eb..6cf5e4e84d61 100644
--- a/fs/exofs/super.c
+++ b/fs/exofs/super.c
@@ -210,7 +210,7 @@ int exofs_sync_fs(struct super_block *sb, int wait)
 	sbi = sb->s_fs_info;
 	fscb = &sbi->s_fscb;
 
-	ret = exofs_get_io_state(sbi, &ios);
+	ret = exofs_get_io_state(&sbi->layout, &ios);
 	if (ret)
 		goto out;
 
@@ -264,12 +264,12 @@ static void _exofs_print_device(const char *msg, const char *dev_path,
 
 void exofs_free_sbi(struct exofs_sb_info *sbi)
 {
-	while (sbi->s_numdevs) {
-		int i = --sbi->s_numdevs;
-		struct osd_dev *od = sbi->s_ods[i];
+	while (sbi->layout.s_numdevs) {
+		int i = --sbi->layout.s_numdevs;
+		struct osd_dev *od = sbi->layout.s_ods[i];
 
 		if (od) {
-			sbi->s_ods[i] = NULL;
+			sbi->layout.s_ods[i] = NULL;
 			osduld_put_device(od);
 		}
 	}
@@ -298,7 +298,8 @@ static void exofs_put_super(struct super_block *sb)
 				  msecs_to_jiffies(100));
 	}
 
-	_exofs_print_device("Unmounting", NULL, sbi->s_ods[0], sbi->s_pid);
+	_exofs_print_device("Unmounting", NULL, sbi->layout.s_ods[0],
+			    sbi->layout.s_pid);
 
 	exofs_free_sbi(sbi);
 	sb->s_fs_info = NULL;
@@ -307,6 +308,8 @@ static void exofs_put_super(struct super_block *sb)
 static int _read_and_match_data_map(struct exofs_sb_info *sbi, unsigned numdevs,
 				    struct exofs_device_table *dt)
 {
+	u64 stripe_length;
+
 	sbi->data_map.odm_num_comps   =
 				le32_to_cpu(dt->dt_data_map.cb_num_comps);
 	sbi->data_map.odm_stripe_unit =
@@ -320,14 +323,63 @@ static int _read_and_match_data_map(struct exofs_sb_info *sbi, unsigned numdevs,
 	sbi->data_map.odm_raid_algorithm  =
 				le32_to_cpu(dt->dt_data_map.cb_raid_algorithm);
 
-/* FIXME: Hard coded mirror only for now. if not so do not mount */
-	if ((sbi->data_map.odm_num_comps != numdevs) ||
-	    (sbi->data_map.odm_stripe_unit != EXOFS_BLKSIZE) ||
-	    (sbi->data_map.odm_raid_algorithm != PNFS_OSD_RAID_0) ||
-	    (sbi->data_map.odm_mirror_cnt != (numdevs - 1)))
+/* FIXME: Only raid0 for now. if not so, do not mount */
+	if (sbi->data_map.odm_num_comps != numdevs) {
+		EXOFS_ERR("odm_num_comps(%u) != numdevs(%u)\n",
+			  sbi->data_map.odm_num_comps, numdevs);
 		return -EINVAL;
-	else
-		return 0;
+	}
+	if (sbi->data_map.odm_raid_algorithm != PNFS_OSD_RAID_0) {
+		EXOFS_ERR("Only RAID_0 for now\n");
+		return -EINVAL;
+	}
+	if (0 != (numdevs % (sbi->data_map.odm_mirror_cnt + 1))) {
+		EXOFS_ERR("Data Map wrong, numdevs=%d mirrors=%d\n",
+			  numdevs, sbi->data_map.odm_mirror_cnt);
+		return -EINVAL;
+	}
+
+	if (0 != (sbi->data_map.odm_stripe_unit & ~PAGE_MASK)) {
+		EXOFS_ERR("Stripe Unit(0x%llx)"
+			  " must be Multples of PAGE_SIZE(0x%lx)\n",
+			  _LLU(sbi->data_map.odm_stripe_unit), PAGE_SIZE);
+		return -EINVAL;
+	}
+
+	sbi->layout.stripe_unit = sbi->data_map.odm_stripe_unit;
+	sbi->layout.mirrors_p1 = sbi->data_map.odm_mirror_cnt + 1;
+
+	if (sbi->data_map.odm_group_width) {
+		sbi->layout.group_width = sbi->data_map.odm_group_width;
+		sbi->layout.group_depth = sbi->data_map.odm_group_depth;
+		if (!sbi->layout.group_depth) {
+			EXOFS_ERR("group_depth == 0 && group_width != 0\n");
+			return -EINVAL;
+		}
+		sbi->layout.group_count = sbi->data_map.odm_num_comps /
+						sbi->layout.mirrors_p1 /
+						sbi->data_map.odm_group_width;
+	} else {
+		if (sbi->data_map.odm_group_depth) {
+			printk(KERN_NOTICE "Warning: group_depth ignored "
+				"group_width == 0 && group_depth == %d\n",
+				sbi->data_map.odm_group_depth);
+			sbi->data_map.odm_group_depth = 0;
+		}
+		sbi->layout.group_width = sbi->data_map.odm_num_comps /
+							sbi->layout.mirrors_p1;
+		sbi->layout.group_depth = -1;
+		sbi->layout.group_count = 1;
+	}
+
+	stripe_length = (u64)sbi->layout.group_width * sbi->layout.stripe_unit;
+	if (stripe_length >= (1ULL << 32)) {
+		EXOFS_ERR("Total Stripe length(0x%llx)"
+			  " >= 32bit is not supported\n", _LLU(stripe_length));
+		return -EINVAL;
+	}
+
+	return 0;
 }
 
 /* @odi is valid only as long as @fscb_dev is valid */
@@ -361,7 +413,7 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info **psbi,
 {
 	struct exofs_sb_info *sbi = *psbi;
 	struct osd_dev *fscb_od;
-	struct osd_obj_id obj = {.partition = sbi->s_pid,
+	struct osd_obj_id obj = {.partition = sbi->layout.s_pid,
 				 .id = EXOFS_DEVTABLE_ID};
 	struct exofs_device_table *dt;
 	unsigned table_bytes = table_count * sizeof(dt->dt_dev_table[0]) +
@@ -376,9 +428,9 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info **psbi,
 		return -ENOMEM;
 	}
 
-	fscb_od = sbi->s_ods[0];
-	sbi->s_ods[0] = NULL;
-	sbi->s_numdevs = 0;
+	fscb_od = sbi->layout.s_ods[0];
+	sbi->layout.s_ods[0] = NULL;
+	sbi->layout.s_numdevs = 0;
 	ret = exofs_read_kern(fscb_od, sbi->s_cred, &obj, 0, dt, table_bytes);
 	if (unlikely(ret)) {
 		EXOFS_ERR("ERROR: reading device table\n");
@@ -397,14 +449,15 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info **psbi,
 		goto out;
 
 	if (likely(numdevs > 1)) {
-		unsigned size = numdevs * sizeof(sbi->s_ods[0]);
+		unsigned size = numdevs * sizeof(sbi->layout.s_ods[0]);
 
 		sbi = krealloc(sbi, sizeof(*sbi) + size, GFP_KERNEL);
 		if (unlikely(!sbi)) {
 			ret = -ENOMEM;
 			goto out;
 		}
-		memset(&sbi->s_ods[1], 0, size - sizeof(sbi->s_ods[0]));
+		memset(&sbi->layout.s_ods[1], 0,
+		       size - sizeof(sbi->layout.s_ods[0]));
 		*psbi = sbi;
 	}
 
@@ -427,8 +480,8 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info **psbi,
 		 * line. We always keep them in device-table order.
 		 */
 		if (fscb_od && osduld_device_same(fscb_od, &odi)) {
-			sbi->s_ods[i] = fscb_od;
-			++sbi->s_numdevs;
+			sbi->layout.s_ods[i] = fscb_od;
+			++sbi->layout.s_numdevs;
 			fscb_od = NULL;
 			continue;
 		}
@@ -441,8 +494,8 @@ static int exofs_read_lookup_dev_table(struct exofs_sb_info **psbi,
 			goto out;
 		}
 
-		sbi->s_ods[i] = od;
-		++sbi->s_numdevs;
+		sbi->layout.s_ods[i] = od;
+		++sbi->layout.s_numdevs;
 
 		/* Read the fscb of the other devices to make sure the FS
 		 * partition is there.
@@ -499,9 +552,15 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_sbi;
 	}
 
-	sbi->s_ods[0] = od;
-	sbi->s_numdevs = 1;
-	sbi->s_pid = opts->pid;
+	/* Default layout in case we do not have a device-table */
+	sbi->layout.stripe_unit = PAGE_SIZE;
+	sbi->layout.mirrors_p1 = 1;
+	sbi->layout.group_width = 1;
+	sbi->layout.group_depth = -1;
+	sbi->layout.group_count = 1;
+	sbi->layout.s_ods[0] = od;
+	sbi->layout.s_numdevs = 1;
+	sbi->layout.s_pid = opts->pid;
 	sbi->s_timeout = opts->timeout;
 
 	/* fill in some other data by hand */
@@ -514,7 +573,7 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent)
 	sb->s_bdev = NULL;
 	sb->s_dev = 0;
 
-	obj.partition = sbi->s_pid;
+	obj.partition = sbi->layout.s_pid;
 	obj.id = EXOFS_SUPER_ID;
 	exofs_make_credential(sbi->s_cred, &obj);
 
@@ -578,13 +637,13 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_sbi;
 	}
 
-	_exofs_print_device("Mounting", opts->dev_name, sbi->s_ods[0],
-			    sbi->s_pid);
+	_exofs_print_device("Mounting", opts->dev_name, sbi->layout.s_ods[0],
+			    sbi->layout.s_pid);
 	return 0;
 
 free_sbi:
 	EXOFS_ERR("Unable to mount exofs on %s pid=0x%llx err=%d\n",
-		  opts->dev_name, sbi->s_pid, ret);
+		  opts->dev_name, sbi->layout.s_pid, ret);
 	exofs_free_sbi(sbi);
 	return ret;
 }
@@ -627,7 +686,7 @@ static int exofs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	uint8_t cred_a[OSD_CAP_LEN];
 	int ret;
 
-	ret = exofs_get_io_state(sbi, &ios);
+	ret = exofs_get_io_state(&sbi->layout, &ios);
 	if (ret) {
 		EXOFS_DBGMSG("exofs_get_io_state failed.\n");
 		return ret;
diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c
index 7f8d2e5a7ea6..1d081f0cfec2 100644
--- a/fs/ext2/balloc.c
+++ b/fs/ext2/balloc.c
@@ -570,7 +570,7 @@ do_more:
 error_return:
 	brelse(bitmap_bh);
 	release_blocks(sb, freed);
-	vfs_dq_free_block(inode, freed);
+	dquot_free_block(inode, freed);
 }
 
 /**
@@ -1236,6 +1236,7 @@ ext2_fsblk_t ext2_new_blocks(struct inode *inode, ext2_fsblk_t goal,
 	unsigned short windowsz = 0;
 	unsigned long ngroups;
 	unsigned long num = *count;
+	int ret;
 
 	*errp = -ENOSPC;
 	sb = inode->i_sb;
@@ -1247,8 +1248,9 @@ ext2_fsblk_t ext2_new_blocks(struct inode *inode, ext2_fsblk_t goal,
 	/*
 	 * Check quota for allocation of this block.
 	 */
-	if (vfs_dq_alloc_block(inode, num)) {
-		*errp = -EDQUOT;
+	ret = dquot_alloc_block(inode, num);
+	if (ret) {
+		*errp = ret;
 		return 0;
 	}
 
@@ -1409,7 +1411,7 @@ allocated:
 
 	*errp = 0;
 	brelse(bitmap_bh);
-	vfs_dq_free_block(inode, *count-num);
+	dquot_free_block(inode, *count-num);
 	*count = num;
 	return ret_block;
 
@@ -1420,7 +1422,7 @@ out:
 	 * Undo the block allocation
 	 */
 	if (!performed_allocation)
-		vfs_dq_free_block(inode, *count);
+		dquot_free_block(inode, *count);
 	brelse(bitmap_bh);
 	return 0;
 }
diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h
index 061914add3cf..0b038e47ad2f 100644
--- a/fs/ext2/ext2.h
+++ b/fs/ext2/ext2.h
@@ -118,7 +118,7 @@ extern unsigned long ext2_count_free (struct buffer_head *, unsigned);
 
 /* inode.c */
 extern struct inode *ext2_iget (struct super_block *, unsigned long);
-extern int ext2_write_inode (struct inode *, int);
+extern int ext2_write_inode (struct inode *, struct writeback_control *);
 extern void ext2_delete_inode (struct inode *);
 extern int ext2_sync_inode (struct inode *);
 extern int ext2_get_block(struct inode *, sector_t, struct buffer_head *, int);
diff --git a/fs/ext2/file.c b/fs/ext2/file.c
index 586e3589d4c2..5d198d0697fb 100644
--- a/fs/ext2/file.c
+++ b/fs/ext2/file.c
@@ -20,6 +20,7 @@
 
 #include <linux/time.h>
 #include <linux/pagemap.h>
+#include <linux/quotaops.h>
 #include "ext2.h"
 #include "xattr.h"
 #include "acl.h"
@@ -70,7 +71,7 @@ const struct file_operations ext2_file_operations = {
 	.compat_ioctl	= ext2_compat_ioctl,
 #endif
 	.mmap		= generic_file_mmap,
-	.open		= generic_file_open,
+	.open		= dquot_file_open,
 	.release	= ext2_release_file,
 	.fsync		= ext2_fsync,
 	.splice_read	= generic_file_splice_read,
@@ -87,7 +88,7 @@ const struct file_operations ext2_xip_file_operations = {
 	.compat_ioctl	= ext2_compat_ioctl,
 #endif
 	.mmap		= xip_file_mmap,
-	.open		= generic_file_open,
+	.open		= dquot_file_open,
 	.release	= ext2_release_file,
 	.fsync		= ext2_fsync,
 };
diff --git a/fs/ext2/ialloc.c b/fs/ext2/ialloc.c
index 15387c9c17d8..ad7d572ee8dc 100644
--- a/fs/ext2/ialloc.c
+++ b/fs/ext2/ialloc.c
@@ -121,8 +121,8 @@ void ext2_free_inode (struct inode * inode)
 	if (!is_bad_inode(inode)) {
 		/* Quota is already initialized in iput() */
 		ext2_xattr_delete_inode(inode);
-		vfs_dq_free_inode(inode);
-		vfs_dq_drop(inode);
+		dquot_free_inode(inode);
+		dquot_drop(inode);
 	}
 
 	es = EXT2_SB(sb)->s_es;
@@ -586,10 +586,10 @@ got:
 		goto fail_drop;
 	}
 
-	if (vfs_dq_alloc_inode(inode)) {
-		err = -EDQUOT;
+	dquot_initialize(inode);
+	err = dquot_alloc_inode(inode);
+	if (err)
 		goto fail_drop;
-	}
 
 	err = ext2_init_acl(inode, dir);
 	if (err)
@@ -605,10 +605,10 @@ got:
 	return inode;
 
 fail_free_drop:
-	vfs_dq_free_inode(inode);
+	dquot_free_inode(inode);
 
 fail_drop:
-	vfs_dq_drop(inode);
+	dquot_drop(inode);
 	inode->i_flags |= S_NOQUOTA;
 	inode->i_nlink = 0;
 	unlock_new_inode(inode);
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index 71b032c65a02..fc13cc119aad 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -41,6 +41,8 @@ MODULE_AUTHOR("Remy Card and others");
 MODULE_DESCRIPTION("Second Extended Filesystem");
 MODULE_LICENSE("GPL");
 
+static int __ext2_write_inode(struct inode *inode, int do_sync);
+
 /*
  * Test whether an inode is a fast symlink.
  */
@@ -58,13 +60,15 @@ static inline int ext2_inode_is_fast_symlink(struct inode *inode)
  */
 void ext2_delete_inode (struct inode * inode)
 {
+	if (!is_bad_inode(inode))
+		dquot_initialize(inode);
 	truncate_inode_pages(&inode->i_data, 0);
 
 	if (is_bad_inode(inode))
 		goto no_delete;
 	EXT2_I(inode)->i_dtime	= get_seconds();
 	mark_inode_dirty(inode);
-	ext2_write_inode(inode, inode_needs_sync(inode));
+	__ext2_write_inode(inode, inode_needs_sync(inode));
 
 	inode->i_size = 0;
 	if (inode->i_blocks)
@@ -1335,7 +1339,7 @@ bad_inode:
 	return ERR_PTR(ret);
 }
 
-int ext2_write_inode(struct inode *inode, int do_sync)
+static int __ext2_write_inode(struct inode *inode, int do_sync)
 {
 	struct ext2_inode_info *ei = EXT2_I(inode);
 	struct super_block *sb = inode->i_sb;
@@ -1440,6 +1444,11 @@ int ext2_write_inode(struct inode *inode, int do_sync)
 	return err;
 }
 
+int ext2_write_inode(struct inode *inode, struct writeback_control *wbc)
+{
+	return __ext2_write_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
+}
+
 int ext2_sync_inode(struct inode *inode)
 {
 	struct writeback_control wbc = {
@@ -1457,9 +1466,12 @@ int ext2_setattr(struct dentry *dentry, struct iattr *iattr)
 	error = inode_change_ok(inode, iattr);
 	if (error)
 		return error;
+
+	if (iattr->ia_valid & ATTR_SIZE)
+		dquot_initialize(inode);
 	if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) ||
 	    (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) {
-		error = vfs_dq_transfer(inode, iattr) ? -EDQUOT : 0;
+		error = dquot_transfer(inode, iattr);
 		if (error)
 			return error;
 	}
diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c
index dd7175ce5606..71efb0e9a3f2 100644
--- a/fs/ext2/namei.c
+++ b/fs/ext2/namei.c
@@ -31,6 +31,7 @@
  */
 
 #include <linux/pagemap.h>
+#include <linux/quotaops.h>
 #include "ext2.h"
 #include "xattr.h"
 #include "acl.h"
@@ -99,24 +100,27 @@ struct dentry *ext2_get_parent(struct dentry *child)
  */
 static int ext2_create (struct inode * dir, struct dentry * dentry, int mode, struct nameidata *nd)
 {
-	struct inode * inode = ext2_new_inode (dir, mode);
-	int err = PTR_ERR(inode);
-	if (!IS_ERR(inode)) {
-		inode->i_op = &ext2_file_inode_operations;
-		if (ext2_use_xip(inode->i_sb)) {
-			inode->i_mapping->a_ops = &ext2_aops_xip;
-			inode->i_fop = &ext2_xip_file_operations;
-		} else if (test_opt(inode->i_sb, NOBH)) {
-			inode->i_mapping->a_ops = &ext2_nobh_aops;
-			inode->i_fop = &ext2_file_operations;
-		} else {
-			inode->i_mapping->a_ops = &ext2_aops;
-			inode->i_fop = &ext2_file_operations;
-		}
-		mark_inode_dirty(inode);
-		err = ext2_add_nondir(dentry, inode);
+	struct inode *inode;
+
+	dquot_initialize(dir);
+
+	inode = ext2_new_inode(dir, mode);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+
+	inode->i_op = &ext2_file_inode_operations;
+	if (ext2_use_xip(inode->i_sb)) {
+		inode->i_mapping->a_ops = &ext2_aops_xip;
+		inode->i_fop = &ext2_xip_file_operations;
+	} else if (test_opt(inode->i_sb, NOBH)) {
+		inode->i_mapping->a_ops = &ext2_nobh_aops;
+		inode->i_fop = &ext2_file_operations;
+	} else {
+		inode->i_mapping->a_ops = &ext2_aops;
+		inode->i_fop = &ext2_file_operations;
 	}
-	return err;
+	mark_inode_dirty(inode);
+	return ext2_add_nondir(dentry, inode);
 }
 
 static int ext2_mknod (struct inode * dir, struct dentry *dentry, int mode, dev_t rdev)
@@ -127,6 +131,8 @@ static int ext2_mknod (struct inode * dir, struct dentry *dentry, int mode, dev_
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
 
+	dquot_initialize(dir);
+
 	inode = ext2_new_inode (dir, mode);
 	err = PTR_ERR(inode);
 	if (!IS_ERR(inode)) {
@@ -151,6 +157,8 @@ static int ext2_symlink (struct inode * dir, struct dentry * dentry,
 	if (l > sb->s_blocksize)
 		goto out;
 
+	dquot_initialize(dir);
+
 	inode = ext2_new_inode (dir, S_IFLNK | S_IRWXUGO);
 	err = PTR_ERR(inode);
 	if (IS_ERR(inode))
@@ -194,6 +202,8 @@ static int ext2_link (struct dentry * old_dentry, struct inode * dir,
 	if (inode->i_nlink >= EXT2_LINK_MAX)
 		return -EMLINK;
 
+	dquot_initialize(dir);
+
 	inode->i_ctime = CURRENT_TIME_SEC;
 	inode_inc_link_count(inode);
 	atomic_inc(&inode->i_count);
@@ -216,6 +226,8 @@ static int ext2_mkdir(struct inode * dir, struct dentry * dentry, int mode)
 	if (dir->i_nlink >= EXT2_LINK_MAX)
 		goto out;
 
+	dquot_initialize(dir);
+
 	inode_inc_link_count(dir);
 
 	inode = ext2_new_inode (dir, S_IFDIR | mode);
@@ -262,6 +274,8 @@ static int ext2_unlink(struct inode * dir, struct dentry *dentry)
 	struct page * page;
 	int err = -ENOENT;
 
+	dquot_initialize(dir);
+
 	de = ext2_find_entry (dir, &dentry->d_name, &page);
 	if (!de)
 		goto out;
@@ -304,6 +318,9 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry,
 	struct ext2_dir_entry_2 * old_de;
 	int err = -ENOENT;
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	old_de = ext2_find_entry (old_dir, &old_dentry->d_name, &old_page);
 	if (!old_de)
 		goto out;
diff --git a/fs/ext2/super.c b/fs/ext2/super.c
index f9cb54a585ce..42e4a303b675 100644
--- a/fs/ext2/super.c
+++ b/fs/ext2/super.c
@@ -194,6 +194,8 @@ static void destroy_inodecache(void)
 static void ext2_clear_inode(struct inode *inode)
 {
 	struct ext2_block_alloc_info *rsv = EXT2_I(inode)->i_block_alloc_info;
+
+	dquot_drop(inode);
 	ext2_discard_reservation(inode);
 	EXT2_I(inode)->i_block_alloc_info = NULL;
 	if (unlikely(rsv))
diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c
index 904f00642f84..e44dc92609be 100644
--- a/fs/ext2/xattr.c
+++ b/fs/ext2/xattr.c
@@ -644,8 +644,8 @@ ext2_xattr_set2(struct inode *inode, struct buffer_head *old_bh,
 				   the inode.  */
 				ea_bdebug(new_bh, "reusing block");
 
-				error = -EDQUOT;
-				if (vfs_dq_alloc_block(inode, 1)) {
+				error = dquot_alloc_block(inode, 1);
+				if (error) {
 					unlock_buffer(new_bh);
 					goto cleanup;
 				}
@@ -702,7 +702,7 @@ ext2_xattr_set2(struct inode *inode, struct buffer_head *old_bh,
 		 * as if nothing happened and cleanup the unused block */
 		if (error && error != -ENOSPC) {
 			if (new_bh && new_bh != old_bh)
-				vfs_dq_free_block(inode, 1);
+				dquot_free_block(inode, 1);
 			goto cleanup;
 		}
 	} else
@@ -734,7 +734,7 @@ ext2_xattr_set2(struct inode *inode, struct buffer_head *old_bh,
 			le32_add_cpu(&HDR(old_bh)->h_refcount, -1);
 			if (ce)
 				mb_cache_entry_release(ce);
-			vfs_dq_free_block(inode, 1);
+			dquot_free_block(inode, 1);
 			mark_buffer_dirty(old_bh);
 			ea_bdebug(old_bh, "refcount now=%d",
 				le32_to_cpu(HDR(old_bh)->h_refcount));
@@ -797,7 +797,7 @@ ext2_xattr_delete_inode(struct inode *inode)
 		mark_buffer_dirty(bh);
 		if (IS_SYNC(inode))
 			sync_dirty_buffer(bh);
-		vfs_dq_free_block(inode, 1);
+		dquot_free_block(inode, 1);
 	}
 	EXT2_I(inode)->i_file_acl = 0;
 
diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c
index 27967f92e820..161da2d3f890 100644
--- a/fs/ext3/balloc.c
+++ b/fs/ext3/balloc.c
@@ -676,7 +676,7 @@ void ext3_free_blocks(handle_t *handle, struct inode *inode,
 	}
 	ext3_free_blocks_sb(handle, sb, block, count, &dquot_freed_blocks);
 	if (dquot_freed_blocks)
-		vfs_dq_free_block(inode, dquot_freed_blocks);
+		dquot_free_block(inode, dquot_freed_blocks);
 	return;
 }
 
@@ -1502,8 +1502,9 @@ ext3_fsblk_t ext3_new_blocks(handle_t *handle, struct inode *inode,
 	/*
 	 * Check quota for allocation of this block.
 	 */
-	if (vfs_dq_alloc_block(inode, num)) {
-		*errp = -EDQUOT;
+	err = dquot_alloc_block(inode, num);
+	if (err) {
+		*errp = err;
 		return 0;
 	}
 
@@ -1713,7 +1714,7 @@ allocated:
 
 	*errp = 0;
 	brelse(bitmap_bh);
-	vfs_dq_free_block(inode, *count-num);
+	dquot_free_block(inode, *count-num);
 	*count = num;
 	return ret_block;
 
@@ -1728,7 +1729,7 @@ out:
 	 * Undo the block allocation
 	 */
 	if (!performed_allocation)
-		vfs_dq_free_block(inode, *count);
+		dquot_free_block(inode, *count);
 	brelse(bitmap_bh);
 	return 0;
 }
diff --git a/fs/ext3/file.c b/fs/ext3/file.c
index 388bbdfa0b4e..f55df0e61cbd 100644
--- a/fs/ext3/file.c
+++ b/fs/ext3/file.c
@@ -21,6 +21,7 @@
 #include <linux/time.h>
 #include <linux/fs.h>
 #include <linux/jbd.h>
+#include <linux/quotaops.h>
 #include <linux/ext3_fs.h>
 #include <linux/ext3_jbd.h>
 #include "xattr.h"
@@ -33,9 +34,9 @@
  */
 static int ext3_release_file (struct inode * inode, struct file * filp)
 {
-	if (EXT3_I(inode)->i_state & EXT3_STATE_FLUSH_ON_CLOSE) {
+	if (ext3_test_inode_state(inode, EXT3_STATE_FLUSH_ON_CLOSE)) {
 		filemap_flush(inode->i_mapping);
-		EXT3_I(inode)->i_state &= ~EXT3_STATE_FLUSH_ON_CLOSE;
+		ext3_clear_inode_state(inode, EXT3_STATE_FLUSH_ON_CLOSE);
 	}
 	/* if we are the last writer on the inode, drop the block reservation */
 	if ((filp->f_mode & FMODE_WRITE) &&
@@ -62,7 +63,7 @@ const struct file_operations ext3_file_operations = {
 	.compat_ioctl	= ext3_compat_ioctl,
 #endif
 	.mmap		= generic_file_mmap,
-	.open		= generic_file_open,
+	.open		= dquot_file_open,
 	.release	= ext3_release_file,
 	.fsync		= ext3_sync_file,
 	.splice_read	= generic_file_splice_read,
diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c
index b39991285136..ef9008b885b5 100644
--- a/fs/ext3/ialloc.c
+++ b/fs/ext3/ialloc.c
@@ -123,10 +123,10 @@ void ext3_free_inode (handle_t *handle, struct inode * inode)
 	 * Note: we must free any quota before locking the superblock,
 	 * as writing the quota to disk may need the lock as well.
 	 */
-	vfs_dq_init(inode);
+	dquot_initialize(inode);
 	ext3_xattr_delete_inode(handle, inode);
-	vfs_dq_free_inode(inode);
-	vfs_dq_drop(inode);
+	dquot_free_inode(inode);
+	dquot_drop(inode);
 
 	is_directory = S_ISDIR(inode->i_mode);
 
@@ -588,10 +588,10 @@ got:
 		sizeof(struct ext3_inode) - EXT3_GOOD_OLD_INODE_SIZE : 0;
 
 	ret = inode;
-	if (vfs_dq_alloc_inode(inode)) {
-		err = -EDQUOT;
+	dquot_initialize(inode);
+	err = dquot_alloc_inode(inode);
+	if (err)
 		goto fail_drop;
-	}
 
 	err = ext3_init_acl(handle, inode, dir);
 	if (err)
@@ -619,10 +619,10 @@ really_out:
 	return ret;
 
 fail_free_drop:
-	vfs_dq_free_inode(inode);
+	dquot_free_inode(inode);
 
 fail_drop:
-	vfs_dq_drop(inode);
+	dquot_drop(inode);
 	inode->i_flags |= S_NOQUOTA;
 	inode->i_nlink = 0;
 	unlock_new_inode(inode);
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 455e6e6e5cb9..7f920b7263a4 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -196,6 +196,9 @@ void ext3_delete_inode (struct inode * inode)
 {
 	handle_t *handle;
 
+	if (!is_bad_inode(inode))
+		dquot_initialize(inode);
+
 	truncate_inode_pages(&inode->i_data, 0);
 
 	if (is_bad_inode(inode))
@@ -1378,7 +1381,7 @@ static int ext3_journalled_write_end(struct file *file,
 	 */
 	if (pos + len > inode->i_size && ext3_can_truncate(inode))
 		ext3_orphan_add(handle, inode);
-	EXT3_I(inode)->i_state |= EXT3_STATE_JDATA;
+	ext3_set_inode_state(inode, EXT3_STATE_JDATA);
 	if (inode->i_size > EXT3_I(inode)->i_disksize) {
 		EXT3_I(inode)->i_disksize = inode->i_size;
 		ret2 = ext3_mark_inode_dirty(handle, inode);
@@ -1417,7 +1420,7 @@ static sector_t ext3_bmap(struct address_space *mapping, sector_t block)
 	journal_t *journal;
 	int err;
 
-	if (EXT3_I(inode)->i_state & EXT3_STATE_JDATA) {
+	if (ext3_test_inode_state(inode, EXT3_STATE_JDATA)) {
 		/*
 		 * This is a REALLY heavyweight approach, but the use of
 		 * bmap on dirty files is expected to be extremely rare:
@@ -1436,7 +1439,7 @@ static sector_t ext3_bmap(struct address_space *mapping, sector_t block)
 		 * everything they get.
 		 */
 
-		EXT3_I(inode)->i_state &= ~EXT3_STATE_JDATA;
+		ext3_clear_inode_state(inode, EXT3_STATE_JDATA);
 		journal = EXT3_JOURNAL(inode);
 		journal_lock_updates(journal);
 		err = journal_flush(journal);
@@ -1528,6 +1531,7 @@ static int ext3_ordered_writepage(struct page *page,
 	int err;
 
 	J_ASSERT(PageLocked(page));
+	WARN_ON_ONCE(IS_RDONLY(inode));
 
 	/*
 	 * We give up here if we're reentered, because it might be for a
@@ -1600,6 +1604,9 @@ static int ext3_writeback_writepage(struct page *page,
 	int ret = 0;
 	int err;
 
+	J_ASSERT(PageLocked(page));
+	WARN_ON_ONCE(IS_RDONLY(inode));
+
 	if (ext3_journal_current_handle())
 		goto out_fail;
 
@@ -1642,6 +1649,9 @@ static int ext3_journalled_writepage(struct page *page,
 	int ret = 0;
 	int err;
 
+	J_ASSERT(PageLocked(page));
+	WARN_ON_ONCE(IS_RDONLY(inode));
+
 	if (ext3_journal_current_handle())
 		goto no_write;
 
@@ -1670,7 +1680,7 @@ static int ext3_journalled_writepage(struct page *page,
 				PAGE_CACHE_SIZE, NULL, write_end_fn);
 		if (ret == 0)
 			ret = err;
-		EXT3_I(inode)->i_state |= EXT3_STATE_JDATA;
+		ext3_set_inode_state(inode, EXT3_STATE_JDATA);
 		unlock_page(page);
 	} else {
 		/*
@@ -1785,8 +1795,9 @@ retry:
 		handle = ext3_journal_start(inode, 2);
 		if (IS_ERR(handle)) {
 			/* This is really bad luck. We've written the data
-			 * but cannot extend i_size. Bail out and pretend
-			 * the write failed... */
+			 * but cannot extend i_size. Truncate allocated blocks
+			 * and pretend the write failed... */
+			ext3_truncate(inode);
 			ret = PTR_ERR(handle);
 			goto out;
 		}
@@ -2402,7 +2413,7 @@ void ext3_truncate(struct inode *inode)
 		goto out_notrans;
 
 	if (inode->i_size == 0 && ext3_should_writeback_data(inode))
-		ei->i_state |= EXT3_STATE_FLUSH_ON_CLOSE;
+		ext3_set_inode_state(inode, EXT3_STATE_FLUSH_ON_CLOSE);
 
 	/*
 	 * We have to lock the EOF page here, because lock_page() nests
@@ -2721,7 +2732,7 @@ int ext3_get_inode_loc(struct inode *inode, struct ext3_iloc *iloc)
 {
 	/* We have all inode data except xattrs in memory here. */
 	return __ext3_get_inode_loc(inode, iloc,
-		!(EXT3_I(inode)->i_state & EXT3_STATE_XATTR));
+		!ext3_test_inode_state(inode, EXT3_STATE_XATTR));
 }
 
 void ext3_set_inode_flags(struct inode *inode)
@@ -2893,7 +2904,7 @@ struct inode *ext3_iget(struct super_block *sb, unsigned long ino)
 					EXT3_GOOD_OLD_INODE_SIZE +
 					ei->i_extra_isize;
 			if (*magic == cpu_to_le32(EXT3_XATTR_MAGIC))
-				 ei->i_state |= EXT3_STATE_XATTR;
+				 ext3_set_inode_state(inode, EXT3_STATE_XATTR);
 		}
 	} else
 		ei->i_extra_isize = 0;
@@ -2955,7 +2966,7 @@ again:
 
 	/* For fields not not tracking in the in-memory inode,
 	 * initialise them to zero for new inodes. */
-	if (ei->i_state & EXT3_STATE_NEW)
+	if (ext3_test_inode_state(inode, EXT3_STATE_NEW))
 		memset(raw_inode, 0, EXT3_SB(inode->i_sb)->s_inode_size);
 
 	ext3_get_inode_flags(ei);
@@ -3052,7 +3063,7 @@ again:
 	rc = ext3_journal_dirty_metadata(handle, bh);
 	if (!err)
 		err = rc;
-	ei->i_state &= ~EXT3_STATE_NEW;
+	ext3_clear_inode_state(inode, EXT3_STATE_NEW);
 
 	atomic_set(&ei->i_sync_tid, handle->h_transaction->t_tid);
 out_brelse:
@@ -3096,7 +3107,7 @@ out_brelse:
  * `stuff()' is running, and the new i_size will be lost.  Plus the inode
  * will no longer be on the superblock's dirty inode list.
  */
-int ext3_write_inode(struct inode *inode, int wait)
+int ext3_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	if (current->flags & PF_MEMALLOC)
 		return 0;
@@ -3107,7 +3118,7 @@ int ext3_write_inode(struct inode *inode, int wait)
 		return -EIO;
 	}
 
-	if (!wait)
+	if (wbc->sync_mode != WB_SYNC_ALL)
 		return 0;
 
 	return ext3_force_commit(inode->i_sb);
@@ -3140,6 +3151,8 @@ int ext3_setattr(struct dentry *dentry, struct iattr *attr)
 	if (error)
 		return error;
 
+	if (ia_valid & ATTR_SIZE)
+		dquot_initialize(inode);
 	if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
 		(ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) {
 		handle_t *handle;
@@ -3152,7 +3165,7 @@ int ext3_setattr(struct dentry *dentry, struct iattr *attr)
 			error = PTR_ERR(handle);
 			goto err_out;
 		}
-		error = vfs_dq_transfer(inode, attr) ? -EDQUOT : 0;
+		error = dquot_transfer(inode, attr);
 		if (error) {
 			ext3_journal_stop(handle);
 			return error;
@@ -3237,7 +3250,7 @@ static int ext3_writepage_trans_blocks(struct inode *inode)
 		ret = 2 * (bpp + indirects) + 2;
 
 #ifdef CONFIG_QUOTA
-	/* We know that structure was already allocated during vfs_dq_init so
+	/* We know that structure was already allocated during dquot_initialize so
 	 * we will be updating only the data blocks + inodes */
 	ret += EXT3_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
 #endif
@@ -3328,7 +3341,7 @@ int ext3_mark_inode_dirty(handle_t *handle, struct inode *inode)
  * i_size has been changed by generic_commit_write() and we thus need
  * to include the updated inode in the current transaction.
  *
- * Also, vfs_dq_alloc_space() will always dirty the inode when blocks
+ * Also, dquot_alloc_space() will always dirty the inode when blocks
  * are allocated to the file.
  *
  * If the inode is marked synchronous, we don't honour that here - doing
diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c
index 7b0e44f7d66f..ee184084ca42 100644
--- a/fs/ext3/namei.c
+++ b/fs/ext3/namei.c
@@ -1696,6 +1696,8 @@ static int ext3_create (struct inode * dir, struct dentry * dentry, int mode,
 	struct inode * inode;
 	int err, retries = 0;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 +
@@ -1730,6 +1732,8 @@ static int ext3_mknod (struct inode * dir, struct dentry *dentry,
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 +
@@ -1766,6 +1770,8 @@ static int ext3_mkdir(struct inode * dir, struct dentry * dentry, int mode)
 	if (dir->i_nlink >= EXT3_LINK_MAX)
 		return -EMLINK;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 +
@@ -2060,7 +2066,9 @@ static int ext3_rmdir (struct inode * dir, struct dentry *dentry)
 
 	/* Initialize quotas before so that eventual writes go in
 	 * separate transaction */
-	vfs_dq_init(dentry->d_inode);
+	dquot_initialize(dir);
+	dquot_initialize(dentry->d_inode);
+
 	handle = ext3_journal_start(dir, EXT3_DELETE_TRANS_BLOCKS(dir->i_sb));
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
@@ -2119,7 +2127,9 @@ static int ext3_unlink(struct inode * dir, struct dentry *dentry)
 
 	/* Initialize quotas before so that eventual writes go
 	 * in separate transaction */
-	vfs_dq_init(dentry->d_inode);
+	dquot_initialize(dir);
+	dquot_initialize(dentry->d_inode);
+
 	handle = ext3_journal_start(dir, EXT3_DELETE_TRANS_BLOCKS(dir->i_sb));
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
@@ -2174,6 +2184,8 @@ static int ext3_symlink (struct inode * dir,
 	if (l > dir->i_sb->s_blocksize)
 		return -ENAMETOOLONG;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT3_INDEX_EXTRA_TRANS_BLOCKS + 5 +
@@ -2228,6 +2240,9 @@ static int ext3_link (struct dentry * old_dentry,
 
 	if (inode->i_nlink >= EXT3_LINK_MAX)
 		return -EMLINK;
+
+	dquot_initialize(dir);
+
 	/*
 	 * Return -ENOENT if we've raced with unlink and i_nlink is 0.  Doing
 	 * otherwise has the potential to corrupt the orphan inode list.
@@ -2278,12 +2293,15 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry,
 	struct ext3_dir_entry_2 * old_de, * new_de;
 	int retval, flush_file = 0;
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	old_bh = new_bh = dir_bh = NULL;
 
 	/* Initialize quotas before so that eventual writes go
 	 * in separate transaction */
 	if (new_dentry->d_inode)
-		vfs_dq_init(new_dentry->d_inode);
+		dquot_initialize(new_dentry->d_inode);
 	handle = ext3_journal_start(old_dir, 2 *
 					EXT3_DATA_TRANS_BLOCKS(old_dir->i_sb) +
 					EXT3_INDEX_EXTRA_TRANS_BLOCKS + 2);
diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index afa2b569da10..e844accbf55d 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -181,7 +181,7 @@ static void ext3_handle_error(struct super_block *sb)
 	if (!test_opt (sb, ERRORS_CONT)) {
 		journal_t *journal = EXT3_SB(sb)->s_journal;
 
-		EXT3_SB(sb)->s_mount_opt |= EXT3_MOUNT_ABORT;
+		set_opt(EXT3_SB(sb)->s_mount_opt, ABORT);
 		if (journal)
 			journal_abort(journal, -EIO);
 	}
@@ -296,7 +296,7 @@ void ext3_abort (struct super_block * sb, const char * function,
 		"error: remounting filesystem read-only");
 	EXT3_SB(sb)->s_mount_state |= EXT3_ERROR_FS;
 	sb->s_flags |= MS_RDONLY;
-	EXT3_SB(sb)->s_mount_opt |= EXT3_MOUNT_ABORT;
+	set_opt(EXT3_SB(sb)->s_mount_opt, ABORT);
 	if (EXT3_SB(sb)->s_journal)
 		journal_abort(EXT3_SB(sb)->s_journal, -EIO);
 }
@@ -528,6 +528,8 @@ static void destroy_inodecache(void)
 static void ext3_clear_inode(struct inode *inode)
 {
 	struct ext3_block_alloc_info *rsv = EXT3_I(inode)->i_block_alloc_info;
+
+	dquot_drop(inode);
 	ext3_discard_reservation(inode);
 	EXT3_I(inode)->i_block_alloc_info = NULL;
 	if (unlikely(rsv))
@@ -562,10 +564,10 @@ static inline void ext3_show_quota_options(struct seq_file *seq, struct super_bl
 	if (sbi->s_qf_names[GRPQUOTA])
 		seq_printf(seq, ",grpjquota=%s", sbi->s_qf_names[GRPQUOTA]);
 
-	if (sbi->s_mount_opt & EXT3_MOUNT_USRQUOTA)
+	if (test_opt(sb, USRQUOTA))
 		seq_puts(seq, ",usrquota");
 
-	if (sbi->s_mount_opt & EXT3_MOUNT_GRPQUOTA)
+	if (test_opt(sb, GRPQUOTA))
 		seq_puts(seq, ",grpquota");
 #endif
 }
@@ -656,8 +658,7 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	if (test_opt(sb, NOBH))
 		seq_puts(seq, ",nobh");
 
-	seq_printf(seq, ",data=%s", data_mode_string(sbi->s_mount_opt &
-						     EXT3_MOUNT_DATA_FLAGS));
+	seq_printf(seq, ",data=%s", data_mode_string(test_opt(sb, DATA_FLAGS)));
 	if (test_opt(sb, DATA_ERR_ABORT))
 		seq_puts(seq, ",data_err=abort");
 
@@ -751,13 +752,6 @@ static ssize_t ext3_quota_write(struct super_block *sb, int type,
 				const char *data, size_t len, loff_t off);
 
 static const struct dquot_operations ext3_quota_operations = {
-	.initialize	= dquot_initialize,
-	.drop		= dquot_drop,
-	.alloc_space	= dquot_alloc_space,
-	.alloc_inode	= dquot_alloc_inode,
-	.free_space	= dquot_free_space,
-	.free_inode	= dquot_free_inode,
-	.transfer	= dquot_transfer,
 	.write_dquot	= ext3_write_dquot,
 	.acquire_dquot	= ext3_acquire_dquot,
 	.release_dquot	= ext3_release_dquot,
@@ -896,6 +890,63 @@ static ext3_fsblk_t get_sb_block(void **data, struct super_block *sb)
 	return sb_block;
 }
 
+#ifdef CONFIG_QUOTA
+static int set_qf_name(struct super_block *sb, int qtype, substring_t *args)
+{
+	struct ext3_sb_info *sbi = EXT3_SB(sb);
+	char *qname;
+
+	if (sb_any_quota_loaded(sb) &&
+		!sbi->s_qf_names[qtype]) {
+		ext3_msg(sb, KERN_ERR,
+			"Cannot change journaled "
+			"quota options when quota turned on");
+		return 0;
+	}
+	qname = match_strdup(args);
+	if (!qname) {
+		ext3_msg(sb, KERN_ERR,
+			"Not enough memory for storing quotafile name");
+		return 0;
+	}
+	if (sbi->s_qf_names[qtype] &&
+		strcmp(sbi->s_qf_names[qtype], qname)) {
+		ext3_msg(sb, KERN_ERR,
+			"%s quota file already specified", QTYPE2NAME(qtype));
+		kfree(qname);
+		return 0;
+	}
+	sbi->s_qf_names[qtype] = qname;
+	if (strchr(sbi->s_qf_names[qtype], '/')) {
+		ext3_msg(sb, KERN_ERR,
+			"quotafile must be on filesystem root");
+		kfree(sbi->s_qf_names[qtype]);
+		sbi->s_qf_names[qtype] = NULL;
+		return 0;
+	}
+	set_opt(sbi->s_mount_opt, QUOTA);
+	return 1;
+}
+
+static int clear_qf_name(struct super_block *sb, int qtype) {
+
+	struct ext3_sb_info *sbi = EXT3_SB(sb);
+
+	if (sb_any_quota_loaded(sb) &&
+		sbi->s_qf_names[qtype]) {
+		ext3_msg(sb, KERN_ERR, "Cannot change journaled quota options"
+			" when quota turned on");
+		return 0;
+	}
+	/*
+	 * The space will be released later when all options are confirmed
+	 * to be correct
+	 */
+	sbi->s_qf_names[qtype] = NULL;
+	return 1;
+}
+#endif
+
 static int parse_options (char *options, struct super_block *sb,
 			  unsigned int *inum, unsigned long *journal_devnum,
 			  ext3_fsblk_t *n_blocks_count, int is_remount)
@@ -906,8 +957,7 @@ static int parse_options (char *options, struct super_block *sb,
 	int data_opt = 0;
 	int option;
 #ifdef CONFIG_QUOTA
-	int qtype, qfmt;
-	char *qname;
+	int qfmt;
 #endif
 
 	if (!options)
@@ -1065,20 +1115,19 @@ static int parse_options (char *options, struct super_block *sb,
 			data_opt = EXT3_MOUNT_WRITEBACK_DATA;
 		datacheck:
 			if (is_remount) {
-				if ((sbi->s_mount_opt & EXT3_MOUNT_DATA_FLAGS)
-						== data_opt)
+				if (test_opt(sb, DATA_FLAGS) == data_opt)
 					break;
 				ext3_msg(sb, KERN_ERR,
 					"error: cannot change "
 					"data mode on remount. The filesystem "
 					"is mounted in data=%s mode and you "
 					"try to remount it in data=%s mode.",
-					data_mode_string(sbi->s_mount_opt &
-							EXT3_MOUNT_DATA_FLAGS),
+					data_mode_string(test_opt(sb,
+							DATA_FLAGS)),
 					data_mode_string(data_opt));
 				return 0;
 			} else {
-				sbi->s_mount_opt &= ~EXT3_MOUNT_DATA_FLAGS;
+				clear_opt(sbi->s_mount_opt, DATA_FLAGS);
 				sbi->s_mount_opt |= data_opt;
 			}
 			break;
@@ -1090,62 +1139,20 @@ static int parse_options (char *options, struct super_block *sb,
 			break;
 #ifdef CONFIG_QUOTA
 		case Opt_usrjquota:
-			qtype = USRQUOTA;
-			goto set_qf_name;
-		case Opt_grpjquota:
-			qtype = GRPQUOTA;
-set_qf_name:
-			if (sb_any_quota_loaded(sb) &&
-			    !sbi->s_qf_names[qtype]) {
-				ext3_msg(sb, KERN_ERR,
-					"error: cannot change journaled "
-					"quota options when quota turned on.");
-				return 0;
-			}
-			qname = match_strdup(&args[0]);
-			if (!qname) {
-				ext3_msg(sb, KERN_ERR,
-					"error: not enough memory for "
-					"storing quotafile name.");
+			if (!set_qf_name(sb, USRQUOTA, &args[0]))
 				return 0;
-			}
-			if (sbi->s_qf_names[qtype] &&
-			    strcmp(sbi->s_qf_names[qtype], qname)) {
-				ext3_msg(sb, KERN_ERR,
-					"error: %s quota file already "
-					"specified.", QTYPE2NAME(qtype));
-				kfree(qname);
-				return 0;
-			}
-			sbi->s_qf_names[qtype] = qname;
-			if (strchr(sbi->s_qf_names[qtype], '/')) {
-				ext3_msg(sb, KERN_ERR,
-					"error: quotafile must be on "
-					"filesystem root.");
-				kfree(sbi->s_qf_names[qtype]);
-				sbi->s_qf_names[qtype] = NULL;
+			break;
+		case Opt_grpjquota:
+			if (!set_qf_name(sb, GRPQUOTA, &args[0]))
 				return 0;
-			}
-			set_opt(sbi->s_mount_opt, QUOTA);
 			break;
 		case Opt_offusrjquota:
-			qtype = USRQUOTA;
-			goto clear_qf_name;
+			if (!clear_qf_name(sb, USRQUOTA))
+				return 0;
+			break;
 		case Opt_offgrpjquota:
-			qtype = GRPQUOTA;
-clear_qf_name:
-			if (sb_any_quota_loaded(sb) &&
-			    sbi->s_qf_names[qtype]) {
-				ext3_msg(sb, KERN_ERR, "error: cannot change "
-					"journaled quota options when "
-					"quota turned on.");
+			if (!clear_qf_name(sb, GRPQUOTA))
 				return 0;
-			}
-			/*
-			 * The space will be released later when all options
-			 * are confirmed to be correct
-			 */
-			sbi->s_qf_names[qtype] = NULL;
 			break;
 		case Opt_jqfmt_vfsold:
 			qfmt = QFMT_VFS_OLD;
@@ -1244,18 +1251,12 @@ set_qf_format:
 	}
 #ifdef CONFIG_QUOTA
 	if (sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) {
-		if ((sbi->s_mount_opt & EXT3_MOUNT_USRQUOTA) &&
-		     sbi->s_qf_names[USRQUOTA])
+		if (test_opt(sb, USRQUOTA) && sbi->s_qf_names[USRQUOTA])
 			clear_opt(sbi->s_mount_opt, USRQUOTA);
-
-		if ((sbi->s_mount_opt & EXT3_MOUNT_GRPQUOTA) &&
-		     sbi->s_qf_names[GRPQUOTA])
+		if (test_opt(sb, GRPQUOTA) && sbi->s_qf_names[GRPQUOTA])
 			clear_opt(sbi->s_mount_opt, GRPQUOTA);
 
-		if ((sbi->s_qf_names[USRQUOTA] &&
-				(sbi->s_mount_opt & EXT3_MOUNT_GRPQUOTA)) ||
-		    (sbi->s_qf_names[GRPQUOTA] &&
-				(sbi->s_mount_opt & EXT3_MOUNT_USRQUOTA))) {
+		if (test_opt(sb, GRPQUOTA) || test_opt(sb, USRQUOTA)) {
 			ext3_msg(sb, KERN_ERR, "error: old and new quota "
 					"format mixing.");
 			return 0;
@@ -1478,7 +1479,7 @@ static void ext3_orphan_cleanup (struct super_block * sb,
 		}
 
 		list_add(&EXT3_I(inode)->i_orphan, &EXT3_SB(sb)->s_orphan);
-		vfs_dq_init(inode);
+		dquot_initialize(inode);
 		if (inode->i_nlink) {
 			printk(KERN_DEBUG
 				"%s: truncating inode %lu to %Ld bytes\n",
@@ -1671,11 +1672,11 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent)
 		set_opt(sbi->s_mount_opt, POSIX_ACL);
 #endif
 	if ((def_mount_opts & EXT3_DEFM_JMODE) == EXT3_DEFM_JMODE_DATA)
-		sbi->s_mount_opt |= EXT3_MOUNT_JOURNAL_DATA;
+		set_opt(sbi->s_mount_opt, JOURNAL_DATA);
 	else if ((def_mount_opts & EXT3_DEFM_JMODE) == EXT3_DEFM_JMODE_ORDERED)
-		sbi->s_mount_opt |= EXT3_MOUNT_ORDERED_DATA;
+		set_opt(sbi->s_mount_opt, ORDERED_DATA);
 	else if ((def_mount_opts & EXT3_DEFM_JMODE) == EXT3_DEFM_JMODE_WBACK)
-		sbi->s_mount_opt |= EXT3_MOUNT_WRITEBACK_DATA;
+		set_opt(sbi->s_mount_opt, WRITEBACK_DATA);
 
 	if (le16_to_cpu(sbi->s_es->s_errors) == EXT3_ERRORS_PANIC)
 		set_opt(sbi->s_mount_opt, ERRORS_PANIC);
@@ -1694,7 +1695,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent)
 		goto failed_mount;
 
 	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
-		((sbi->s_mount_opt & EXT3_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+		(test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
 
 	if (le32_to_cpu(es->s_rev_level) == EXT3_GOOD_OLD_REV &&
 	    (EXT3_HAS_COMPAT_FEATURE(sb, ~0U) ||
@@ -2561,11 +2562,11 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data)
 		goto restore_opts;
 	}
 
-	if (sbi->s_mount_opt & EXT3_MOUNT_ABORT)
+	if (test_opt(sb, ABORT))
 		ext3_abort(sb, __func__, "Abort forced by user");
 
 	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
-		((sbi->s_mount_opt & EXT3_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+		(test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
 
 	es = sbi->s_es;
 
@@ -2573,7 +2574,7 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data)
 
 	if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY) ||
 		n_blocks_count > le32_to_cpu(es->s_blocks_count)) {
-		if (sbi->s_mount_opt & EXT3_MOUNT_ABORT) {
+		if (test_opt(sb, ABORT)) {
 			err = -EROFS;
 			goto restore_opts;
 		}
@@ -2734,7 +2735,7 @@ static int ext3_statfs (struct dentry * dentry, struct kstatfs * buf)
  * Process 1                         Process 2
  * ext3_create()                     quota_sync()
  *   journal_start()                   write_dquot()
- *   vfs_dq_init()                       down(dqio_mutex)
+ *   dquot_initialize()                       down(dqio_mutex)
  *     down(dqio_mutex)                    journal_start()
  *
  */
@@ -2942,9 +2943,7 @@ static ssize_t ext3_quota_write(struct super_block *sb, int type,
 	sector_t blk = off >> EXT3_BLOCK_SIZE_BITS(sb);
 	int err = 0;
 	int offset = off & (sb->s_blocksize - 1);
-	int tocopy;
 	int journal_quota = EXT3_SB(sb)->s_qf_names[type] != NULL;
-	size_t towrite = len;
 	struct buffer_head *bh;
 	handle_t *handle = journal_current_handle();
 
@@ -2955,53 +2954,54 @@ static ssize_t ext3_quota_write(struct super_block *sb, int type,
 			(unsigned long long)off, (unsigned long long)len);
 		return -EIO;
 	}
+
+	/*
+	 * Since we account only one data block in transaction credits,
+	 * then it is impossible to cross a block boundary.
+	 */
+	if (sb->s_blocksize - offset < len) {
+		ext3_msg(sb, KERN_WARNING, "Quota write (off=%llu, len=%llu)"
+			" cancelled because not block aligned",
+			(unsigned long long)off, (unsigned long long)len);
+		return -EIO;
+	}
 	mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA);
-	while (towrite > 0) {
-		tocopy = sb->s_blocksize - offset < towrite ?
-				sb->s_blocksize - offset : towrite;
-		bh = ext3_bread(handle, inode, blk, 1, &err);
-		if (!bh)
+	bh = ext3_bread(handle, inode, blk, 1, &err);
+	if (!bh)
+		goto out;
+	if (journal_quota) {
+		err = ext3_journal_get_write_access(handle, bh);
+		if (err) {
+			brelse(bh);
 			goto out;
-		if (journal_quota) {
-			err = ext3_journal_get_write_access(handle, bh);
-			if (err) {
-				brelse(bh);
-				goto out;
-			}
-		}
-		lock_buffer(bh);
-		memcpy(bh->b_data+offset, data, tocopy);
-		flush_dcache_page(bh->b_page);
-		unlock_buffer(bh);
-		if (journal_quota)
-			err = ext3_journal_dirty_metadata(handle, bh);
-		else {
-			/* Always do at least ordered writes for quotas */
-			err = ext3_journal_dirty_data(handle, bh);
-			mark_buffer_dirty(bh);
 		}
-		brelse(bh);
-		if (err)
-			goto out;
-		offset = 0;
-		towrite -= tocopy;
-		data += tocopy;
-		blk++;
 	}
+	lock_buffer(bh);
+	memcpy(bh->b_data+offset, data, len);
+	flush_dcache_page(bh->b_page);
+	unlock_buffer(bh);
+	if (journal_quota)
+		err = ext3_journal_dirty_metadata(handle, bh);
+	else {
+		/* Always do at least ordered writes for quotas */
+		err = ext3_journal_dirty_data(handle, bh);
+		mark_buffer_dirty(bh);
+	}
+	brelse(bh);
 out:
-	if (len == towrite) {
+	if (err) {
 		mutex_unlock(&inode->i_mutex);
 		return err;
 	}
-	if (inode->i_size < off+len-towrite) {
-		i_size_write(inode, off+len-towrite);
+	if (inode->i_size < off + len) {
+		i_size_write(inode, off + len);
 		EXT3_I(inode)->i_disksize = inode->i_size;
 	}
 	inode->i_version++;
 	inode->i_mtime = inode->i_ctime = CURRENT_TIME;
 	ext3_mark_inode_dirty(handle, inode);
 	mutex_unlock(&inode->i_mutex);
-	return len - towrite;
+	return len;
 }
 
 #endif
diff --git a/fs/ext3/xattr.c b/fs/ext3/xattr.c
index 66895ccf76c7..534a94c3a933 100644
--- a/fs/ext3/xattr.c
+++ b/fs/ext3/xattr.c
@@ -274,7 +274,7 @@ ext3_xattr_ibody_get(struct inode *inode, int name_index, const char *name,
 	void *end;
 	int error;
 
-	if (!(EXT3_I(inode)->i_state & EXT3_STATE_XATTR))
+	if (!ext3_test_inode_state(inode, EXT3_STATE_XATTR))
 		return -ENODATA;
 	error = ext3_get_inode_loc(inode, &iloc);
 	if (error)
@@ -403,7 +403,7 @@ ext3_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size)
 	void *end;
 	int error;
 
-	if (!(EXT3_I(inode)->i_state & EXT3_STATE_XATTR))
+	if (!ext3_test_inode_state(inode, EXT3_STATE_XATTR))
 		return 0;
 	error = ext3_get_inode_loc(inode, &iloc);
 	if (error)
@@ -500,7 +500,7 @@ ext3_xattr_release_block(handle_t *handle, struct inode *inode,
 		error = ext3_journal_dirty_metadata(handle, bh);
 		if (IS_SYNC(inode))
 			handle->h_sync = 1;
-		vfs_dq_free_block(inode, 1);
+		dquot_free_block(inode, 1);
 		ea_bdebug(bh, "refcount now=%d; releasing",
 			  le32_to_cpu(BHDR(bh)->h_refcount));
 		if (ce)
@@ -775,8 +775,8 @@ inserted:
 			else {
 				/* The old block is released after updating
 				   the inode. */
-				error = -EDQUOT;
-				if (vfs_dq_alloc_block(inode, 1))
+				error = dquot_alloc_block(inode, 1);
+				if (error)
 					goto cleanup;
 				error = ext3_journal_get_write_access(handle,
 								      new_bh);
@@ -850,7 +850,7 @@ cleanup:
 	return error;
 
 cleanup_dquot:
-	vfs_dq_free_block(inode, 1);
+	dquot_free_block(inode, 1);
 	goto cleanup;
 
 bad_block:
@@ -882,7 +882,7 @@ ext3_xattr_ibody_find(struct inode *inode, struct ext3_xattr_info *i,
 	is->s.base = is->s.first = IFIRST(header);
 	is->s.here = is->s.first;
 	is->s.end = (void *)raw_inode + EXT3_SB(inode->i_sb)->s_inode_size;
-	if (EXT3_I(inode)->i_state & EXT3_STATE_XATTR) {
+	if (ext3_test_inode_state(inode, EXT3_STATE_XATTR)) {
 		error = ext3_xattr_check_names(IFIRST(header), is->s.end);
 		if (error)
 			return error;
@@ -914,10 +914,10 @@ ext3_xattr_ibody_set(handle_t *handle, struct inode *inode,
 	header = IHDR(inode, ext3_raw_inode(&is->iloc));
 	if (!IS_LAST_ENTRY(s->first)) {
 		header->h_magic = cpu_to_le32(EXT3_XATTR_MAGIC);
-		EXT3_I(inode)->i_state |= EXT3_STATE_XATTR;
+		ext3_set_inode_state(inode, EXT3_STATE_XATTR);
 	} else {
 		header->h_magic = cpu_to_le32(0);
-		EXT3_I(inode)->i_state &= ~EXT3_STATE_XATTR;
+		ext3_clear_inode_state(inode, EXT3_STATE_XATTR);
 	}
 	return 0;
 }
@@ -967,10 +967,10 @@ ext3_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
 	if (error)
 		goto cleanup;
 
-	if (EXT3_I(inode)->i_state & EXT3_STATE_NEW) {
+	if (ext3_test_inode_state(inode, EXT3_STATE_NEW)) {
 		struct ext3_inode *raw_inode = ext3_raw_inode(&is.iloc);
 		memset(raw_inode, 0, EXT3_SB(inode->i_sb)->s_inode_size);
-		EXT3_I(inode)->i_state &= ~EXT3_STATE_NEW;
+		ext3_clear_inode_state(inode, EXT3_STATE_NEW);
 	}
 
 	error = ext3_xattr_ibody_find(inode, &i, &is);
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 22bc7435d913..d2f37a5516c7 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -97,8 +97,8 @@ unsigned ext4_init_block_bitmap(struct super_block *sb, struct buffer_head *bh,
 		/* If checksum is bad mark all blocks used to prevent allocation
 		 * essentially implementing a per-group read-only flag. */
 		if (!ext4_group_desc_csum_verify(sbi, block_group, gdp)) {
-			ext4_error(sb, __func__,
-				  "Checksum bad for group %u", block_group);
+			ext4_error(sb, "Checksum bad for group %u",
+					block_group);
 			ext4_free_blks_set(sb, gdp, 0);
 			ext4_free_inodes_set(sb, gdp, 0);
 			ext4_itable_unused_set(sb, gdp, 0);
@@ -130,8 +130,7 @@ unsigned ext4_init_block_bitmap(struct super_block *sb, struct buffer_head *bh,
 		 * to make sure we calculate the right free blocks
 		 */
 		group_blocks = ext4_blocks_count(sbi->s_es) -
-			le32_to_cpu(sbi->s_es->s_first_data_block) -
-			(EXT4_BLOCKS_PER_GROUP(sb) * (ngroups - 1));
+			ext4_group_first_block_no(sb, ngroups - 1);
 	} else {
 		group_blocks = EXT4_BLOCKS_PER_GROUP(sb);
 	}
@@ -189,9 +188,6 @@ unsigned ext4_init_block_bitmap(struct super_block *sb, struct buffer_head *bh,
  * when a file system is mounted (see ext4_fill_super).
  */
 
-
-#define in_range(b, first, len)	((b) >= (first) && (b) <= (first) + (len) - 1)
-
 /**
  * ext4_get_group_desc() -- load group descriptor from disk
  * @sb:			super block
@@ -210,10 +206,8 @@ struct ext4_group_desc * ext4_get_group_desc(struct super_block *sb,
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
 
 	if (block_group >= ngroups) {
-		ext4_error(sb, "ext4_get_group_desc",
-			   "block_group >= groups_count - "
-			   "block_group = %u, groups_count = %u",
-			   block_group, ngroups);
+		ext4_error(sb, "block_group >= groups_count - block_group = %u,"
+			   " groups_count = %u", block_group, ngroups);
 
 		return NULL;
 	}
@@ -221,8 +215,7 @@ struct ext4_group_desc * ext4_get_group_desc(struct super_block *sb,
 	group_desc = block_group >> EXT4_DESC_PER_BLOCK_BITS(sb);
 	offset = block_group & (EXT4_DESC_PER_BLOCK(sb) - 1);
 	if (!sbi->s_group_desc[group_desc]) {
-		ext4_error(sb, "ext4_get_group_desc",
-			   "Group descriptor not loaded - "
+		ext4_error(sb, "Group descriptor not loaded - "
 			   "block_group = %u, group_desc = %u, desc = %u",
 			   block_group, group_desc, offset);
 		return NULL;
@@ -282,9 +275,7 @@ static int ext4_valid_block_bitmap(struct super_block *sb,
 		return 1;
 
 err_out:
-	ext4_error(sb, __func__,
-			"Invalid block bitmap - "
-			"block_group = %d, block = %llu",
+	ext4_error(sb, "Invalid block bitmap - block_group = %d, block = %llu",
 			block_group, bitmap_blk);
 	return 0;
 }
@@ -311,8 +302,7 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group)
 	bitmap_blk = ext4_block_bitmap(sb, desc);
 	bh = sb_getblk(sb, bitmap_blk);
 	if (unlikely(!bh)) {
-		ext4_error(sb, __func__,
-			    "Cannot read block bitmap - "
+		ext4_error(sb, "Cannot read block bitmap - "
 			    "block_group = %u, block_bitmap = %llu",
 			    block_group, bitmap_blk);
 		return NULL;
@@ -354,8 +344,7 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group)
 	set_bitmap_uptodate(bh);
 	if (bh_submit_read(bh) < 0) {
 		put_bh(bh);
-		ext4_error(sb, __func__,
-			    "Cannot read block bitmap - "
+		ext4_error(sb, "Cannot read block bitmap - "
 			    "block_group = %u, block_bitmap = %llu",
 			    block_group, bitmap_blk);
 		return NULL;
@@ -419,8 +408,7 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb,
 	    in_range(block, ext4_inode_table(sb, desc), sbi->s_itb_per_group) ||
 	    in_range(block + count - 1, ext4_inode_table(sb, desc),
 		     sbi->s_itb_per_group)) {
-		ext4_error(sb, __func__,
-			   "Adding blocks in system zones - "
+		ext4_error(sb, "Adding blocks in system zones - "
 			   "Block = %llu, count = %lu",
 			   block, count);
 		goto error_return;
@@ -453,8 +441,7 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb,
 		BUFFER_TRACE(bitmap_bh, "clear bit");
 		if (!ext4_clear_bit_atomic(ext4_group_lock_ptr(sb, block_group),
 						bit + i, bitmap_bh->b_data)) {
-			ext4_error(sb, __func__,
-				   "bit already cleared for block %llu",
+			ext4_error(sb, "bit already cleared for block %llu",
 				   (ext4_fsblk_t)(block + i));
 			BUFFER_TRACE(bitmap_bh, "bit already cleared");
 		} else {
diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c
index a60ab9aad57d..983f0e127493 100644
--- a/fs/ext4/block_validity.c
+++ b/fs/ext4/block_validity.c
@@ -205,14 +205,14 @@ void ext4_release_system_zone(struct super_block *sb)
 		entry = rb_entry(n, struct ext4_system_zone, node);
 		kmem_cache_free(ext4_system_zone_cachep, entry);
 		if (!parent)
-			EXT4_SB(sb)->system_blks.rb_node = NULL;
+			EXT4_SB(sb)->system_blks = RB_ROOT;
 		else if (parent->rb_left == n)
 			parent->rb_left = NULL;
 		else if (parent->rb_right == n)
 			parent->rb_right = NULL;
 		n = parent;
 	}
-	EXT4_SB(sb)->system_blks.rb_node = NULL;
+	EXT4_SB(sb)->system_blks = RB_ROOT;
 }
 
 /*
diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c
index 9dc93168e262..86cb6d86a048 100644
--- a/fs/ext4/dir.c
+++ b/fs/ext4/dir.c
@@ -83,10 +83,12 @@ int ext4_check_dir_entry(const char *function, struct inode *dir,
 		error_msg = "inode out of bounds";
 
 	if (error_msg != NULL)
-		ext4_error(dir->i_sb, function,
-			"bad entry in directory #%lu: %s - "
-			"offset=%u, inode=%u, rec_len=%d, name_len=%d",
-			dir->i_ino, error_msg, offset,
+		__ext4_error(dir->i_sb, function,
+			"bad entry in directory #%lu: %s - block=%llu"
+			"offset=%u(%u), inode=%u, rec_len=%d, name_len=%d",
+			dir->i_ino, error_msg, 
+			(unsigned long long) bh->b_blocknr,     
+			(unsigned) (offset%bh->b_size), offset,
 			le32_to_cpu(de->inode),
 			rlen, de->name_len);
 	return error_msg == NULL ? 1 : 0;
@@ -150,7 +152,7 @@ static int ext4_readdir(struct file *filp,
 		 */
 		if (!bh) {
 			if (!dir_has_error) {
-				ext4_error(sb, __func__, "directory #%lu "
+				ext4_error(sb, "directory #%lu "
 					   "contains a hole at offset %Lu",
 					   inode->i_ino,
 					   (unsigned long long) filp->f_pos);
@@ -303,7 +305,7 @@ static void free_rb_tree_fname(struct rb_root *root)
 			kfree(old);
 		}
 		if (!parent)
-			root->rb_node = NULL;
+			*root = RB_ROOT;
 		else if (parent->rb_left == n)
 			parent->rb_left = NULL;
 		else if (parent->rb_right == n)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 874d169a193e..bf938cf7c5f0 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -53,6 +53,12 @@
 #define ext4_debug(f, a...)	do {} while (0)
 #endif
 
+#define EXT4_ERROR_INODE(inode, fmt, a...) \
+	ext4_error_inode(__func__, (inode), (fmt), ## a);
+
+#define EXT4_ERROR_FILE(file, fmt, a...)	\
+	ext4_error_file(__func__, (file), (fmt), ## a);
+
 /* data type for block offset of block group */
 typedef int ext4_grpblk_t;
 
@@ -133,14 +139,14 @@ struct mpage_da_data {
 	int pages_written;
 	int retval;
 };
-#define	DIO_AIO_UNWRITTEN	0x1
+#define	EXT4_IO_UNWRITTEN	0x1
 typedef struct ext4_io_end {
 	struct list_head	list;		/* per-file finished AIO list */
 	struct inode		*inode;		/* file being written to */
 	unsigned int		flag;		/* unwritten or not */
-	int			error;		/* I/O error code */
-	ext4_lblk_t		offset;		/* offset in the file */
-	size_t			size;		/* size of the extent */
+	struct page		*page;		/* page struct for buffer write */
+	loff_t			offset;		/* offset in the file */
+	ssize_t			size;		/* size of the extent */
 	struct work_struct	work;		/* data work queue */
 } ext4_io_end_t;
 
@@ -284,10 +290,12 @@ struct flex_groups {
 #define EXT4_TOPDIR_FL			0x00020000 /* Top of directory hierarchies*/
 #define EXT4_HUGE_FILE_FL               0x00040000 /* Set to each huge file */
 #define EXT4_EXTENTS_FL			0x00080000 /* Inode uses extents */
+#define EXT4_EA_INODE_FL	        0x00200000 /* Inode used for large EA */
+#define EXT4_EOFBLOCKS_FL		0x00400000 /* Blocks allocated beyond EOF */
 #define EXT4_RESERVED_FL		0x80000000 /* reserved for ext4 lib */
 
-#define EXT4_FL_USER_VISIBLE		0x000BDFFF /* User visible flags */
-#define EXT4_FL_USER_MODIFIABLE		0x000B80FF /* User modifiable flags */
+#define EXT4_FL_USER_VISIBLE		0x004BDFFF /* User visible flags */
+#define EXT4_FL_USER_MODIFIABLE		0x004B80FF /* User modifiable flags */
 
 /* Flags that should be inherited by new inodes from their parent. */
 #define EXT4_FL_INHERITED (EXT4_SECRM_FL | EXT4_UNRM_FL | EXT4_COMPR_FL |\
@@ -313,17 +321,6 @@ static inline __u32 ext4_mask_flags(umode_t mode, __u32 flags)
 		return flags & EXT4_OTHER_FLMASK;
 }
 
-/*
- * Inode dynamic state flags
- */
-#define EXT4_STATE_JDATA		0x00000001 /* journaled data exists */
-#define EXT4_STATE_NEW			0x00000002 /* inode is newly created */
-#define EXT4_STATE_XATTR		0x00000004 /* has in-inode xattrs */
-#define EXT4_STATE_NO_EXPAND		0x00000008 /* No space for expansion */
-#define EXT4_STATE_DA_ALLOC_CLOSE	0x00000010 /* Alloc DA blks on close */
-#define EXT4_STATE_EXT_MIGRATE		0x00000020 /* Inode is migrating */
-#define EXT4_STATE_DIO_UNWRITTEN	0x00000040 /* need convert on dio done*/
-
 /* Used to pass group descriptor data when online resize is done */
 struct ext4_new_group_input {
 	__u32 group;		/* Group number for this data */
@@ -364,19 +361,20 @@ struct ext4_new_group_data {
 	/* caller is from the direct IO path, request to creation of an
 	unitialized extents if not allocated, split the uninitialized
 	extent if blocks has been preallocated already*/
-#define EXT4_GET_BLOCKS_DIO			0x0008
+#define EXT4_GET_BLOCKS_PRE_IO			0x0008
 #define EXT4_GET_BLOCKS_CONVERT			0x0010
-#define EXT4_GET_BLOCKS_DIO_CREATE_EXT		(EXT4_GET_BLOCKS_DIO|\
+#define EXT4_GET_BLOCKS_IO_CREATE_EXT		(EXT4_GET_BLOCKS_PRE_IO|\
+					 EXT4_GET_BLOCKS_CREATE_UNINIT_EXT)
+	/* Convert extent to initialized after IO complete */
+#define EXT4_GET_BLOCKS_IO_CONVERT_EXT		(EXT4_GET_BLOCKS_CONVERT|\
 					 EXT4_GET_BLOCKS_CREATE_UNINIT_EXT)
-	/* Convert extent to initialized after direct IO complete */
-#define EXT4_GET_BLOCKS_DIO_CONVERT_EXT		(EXT4_GET_BLOCKS_CONVERT|\
-					 EXT4_GET_BLOCKS_DIO_CREATE_EXT)
 
 /*
  * Flags used by ext4_free_blocks
  */
 #define EXT4_FREE_BLOCKS_METADATA	0x0001
 #define EXT4_FREE_BLOCKS_FORGET		0x0002
+#define EXT4_FREE_BLOCKS_VALIDATED	0x0004
 
 /*
  * ioctl commands
@@ -630,7 +628,7 @@ struct ext4_inode_info {
 	 * near to their parent directory's inode.
 	 */
 	ext4_group_t	i_block_group;
-	__u32	i_state;		/* Dynamic state flags for ext4 */
+	unsigned long	i_state_flags;		/* Dynamic state flags */
 
 	ext4_lblk_t		i_dir_start_lookup;
 #ifdef CONFIG_EXT4_FS_XATTR
@@ -708,8 +706,9 @@ struct ext4_inode_info {
 	qsize_t i_reserved_quota;
 #endif
 
-	/* completed async DIOs that might need unwritten extents handling */
-	struct list_head i_aio_dio_complete_list;
+	/* completed IOs that might need unwritten extents handling */
+	struct list_head i_completed_io_list;
+	spinlock_t i_completed_io_lock;
 	/* current io_end structure for async DIO write*/
 	ext4_io_end_t *cur_aio_dio;
 
@@ -760,6 +759,7 @@ struct ext4_inode_info {
 #define EXT4_MOUNT_QUOTA		0x80000 /* Some quota option set */
 #define EXT4_MOUNT_USRQUOTA		0x100000 /* "old" user quota */
 #define EXT4_MOUNT_GRPQUOTA		0x200000 /* "old" group quota */
+#define EXT4_MOUNT_DIOREAD_NOLOCK	0x400000 /* Enable support for dio read nolocking */
 #define EXT4_MOUNT_JOURNAL_CHECKSUM	0x800000 /* Journal checksums */
 #define EXT4_MOUNT_JOURNAL_ASYNC_COMMIT	0x1000000 /* Journal Async Commit */
 #define EXT4_MOUNT_I_VERSION            0x2000000 /* i_version support */
@@ -1014,7 +1014,7 @@ struct ext4_sb_info {
 	atomic_t s_lock_busy;
 
 	/* locality groups */
-	struct ext4_locality_group *s_locality_groups;
+	struct ext4_locality_group __percpu *s_locality_groups;
 
 	/* for write statistics */
 	unsigned long s_sectors_written_start;
@@ -1050,6 +1050,34 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino)
 		(ino >= EXT4_FIRST_INO(sb) &&
 		 ino <= le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count));
 }
+
+/*
+ * Inode dynamic state flags
+ */
+enum {
+	EXT4_STATE_JDATA,		/* journaled data exists */
+	EXT4_STATE_NEW,			/* inode is newly created */
+	EXT4_STATE_XATTR,		/* has in-inode xattrs */
+	EXT4_STATE_NO_EXPAND,		/* No space for expansion */
+	EXT4_STATE_DA_ALLOC_CLOSE,	/* Alloc DA blks on close */
+	EXT4_STATE_EXT_MIGRATE,		/* Inode is migrating */
+	EXT4_STATE_DIO_UNWRITTEN,	/* need convert on dio done*/
+};
+
+static inline int ext4_test_inode_state(struct inode *inode, int bit)
+{
+	return test_bit(bit, &EXT4_I(inode)->i_state_flags);
+}
+
+static inline void ext4_set_inode_state(struct inode *inode, int bit)
+{
+	set_bit(bit, &EXT4_I(inode)->i_state_flags);
+}
+
+static inline void ext4_clear_inode_state(struct inode *inode, int bit)
+{
+	clear_bit(bit, &EXT4_I(inode)->i_state_flags);
+}
 #else
 /* Assume that user mode programs are passing in an ext4fs superblock, not
  * a kernel struct super_block.  This will allow us to call the feature-test
@@ -1126,6 +1154,8 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino)
 #define EXT4_FEATURE_INCOMPAT_64BIT		0x0080
 #define EXT4_FEATURE_INCOMPAT_MMP               0x0100
 #define EXT4_FEATURE_INCOMPAT_FLEX_BG		0x0200
+#define EXT4_FEATURE_INCOMPAT_EA_INODE		0x0400 /* EA in inode */
+#define EXT4_FEATURE_INCOMPAT_DIRDATA		0x1000 /* data in dirent */
 
 #define EXT4_FEATURE_COMPAT_SUPP	EXT2_FEATURE_COMPAT_EXT_ATTR
 #define EXT4_FEATURE_INCOMPAT_SUPP	(EXT4_FEATURE_INCOMPAT_FILETYPE| \
@@ -1416,7 +1446,7 @@ int ext4_get_block(struct inode *inode, sector_t iblock,
 				struct buffer_head *bh_result, int create);
 
 extern struct inode *ext4_iget(struct super_block *, unsigned long);
-extern int  ext4_write_inode(struct inode *, int);
+extern int  ext4_write_inode(struct inode *, struct writeback_control *);
 extern int  ext4_setattr(struct dentry *, struct iattr *);
 extern int  ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
 				struct kstat *stat);
@@ -1439,7 +1469,7 @@ extern int ext4_block_truncate_page(handle_t *handle,
 		struct address_space *mapping, loff_t from);
 extern int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf);
 extern qsize_t *ext4_get_reserved_space(struct inode *inode);
-extern int flush_aio_dio_completed_IO(struct inode *inode);
+extern int flush_completed_IO(struct inode *inode);
 extern void ext4_da_update_reserve_space(struct inode *inode,
 					int used, int quota_claim);
 /* ioctl.c */
@@ -1465,13 +1495,20 @@ extern int ext4_group_extend(struct super_block *sb,
 				ext4_fsblk_t n_blocks_count);
 
 /* super.c */
-extern void ext4_error(struct super_block *, const char *, const char *, ...)
+extern void __ext4_error(struct super_block *, const char *, const char *, ...)
+	__attribute__ ((format (printf, 3, 4)));
+#define ext4_error(sb, message...)	__ext4_error(sb, __func__, ## message)
+extern void ext4_error_inode(const char *, struct inode *, const char *, ...)
+	__attribute__ ((format (printf, 3, 4)));
+extern void ext4_error_file(const char *, struct file *, const char *, ...)
 	__attribute__ ((format (printf, 3, 4)));
 extern void __ext4_std_error(struct super_block *, const char *, int);
 extern void ext4_abort(struct super_block *, const char *, const char *, ...)
 	__attribute__ ((format (printf, 3, 4)));
-extern void ext4_warning(struct super_block *, const char *, const char *, ...)
+extern void __ext4_warning(struct super_block *, const char *,
+			  const char *, ...)
 	__attribute__ ((format (printf, 3, 4)));
+#define ext4_warning(sb, message...)	__ext4_warning(sb, __func__, ## message)
 extern void ext4_msg(struct super_block *, const char *, const char *, ...)
 	__attribute__ ((format (printf, 3, 4)));
 extern void ext4_grp_locked_error(struct super_block *, ext4_group_t,
@@ -1744,7 +1781,7 @@ extern void ext4_ext_release(struct super_block *);
 extern long ext4_fallocate(struct inode *inode, int mode, loff_t offset,
 			  loff_t len);
 extern int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
-			  loff_t len);
+			  ssize_t len);
 extern int ext4_get_blocks(handle_t *handle, struct inode *inode,
 			   sector_t block, unsigned int max_blocks,
 			   struct buffer_head *bh, int flags);
@@ -1756,6 +1793,15 @@ extern int ext4_move_extents(struct file *o_filp, struct file *d_filp,
 			     __u64 len, __u64 *moved_len);
 
 
+/* BH_Uninit flag: blocks are allocated but uninitialized on disk */
+enum ext4_state_bits {
+	BH_Uninit	/* blocks are allocated but uninitialized on disk */
+	  = BH_JBDPrivateStart,
+};
+
+BUFFER_FNS(Uninit, uninit)
+TAS_BUFFER_FNS(Uninit, uninit)
+
 /*
  * Add new method to test wether block and inode bitmaps are properly
  * initialized. With uninit_bg reading the block from disk is not enough
@@ -1773,6 +1819,8 @@ static inline void set_bitmap_uptodate(struct buffer_head *bh)
 	set_bit(BH_BITMAP_UPTODATE, &(bh)->b_state);
 }
 
+#define in_range(b, first, len)	((b) >= (first) && (b) <= (first) + (len) - 1)
+
 #endif	/* __KERNEL__ */
 
 #endif	/* _EXT4_H */
diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index b57e5c711b6d..53d2764d71ca 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -125,14 +125,14 @@ int __ext4_handle_dirty_metadata(const char *where, handle_t *handle,
 			ext4_journal_abort_handle(where, __func__, bh,
 						  handle, err);
 	} else {
-		if (inode && bh)
+		if (inode)
 			mark_buffer_dirty_inode(bh, inode);
 		else
 			mark_buffer_dirty(bh);
 		if (inode && inode_needs_sync(inode)) {
 			sync_dirty_buffer(bh);
 			if (buffer_req(bh) && !buffer_uptodate(bh)) {
-				ext4_error(inode->i_sb, __func__,
+				ext4_error(inode->i_sb,
 					   "IO error syncing inode, "
 					   "inode=%lu, block=%llu",
 					   inode->i_ino,
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 05eca817d704..b79ad5126468 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -304,4 +304,28 @@ static inline int ext4_should_writeback_data(struct inode *inode)
 	return 0;
 }
 
+/*
+ * This function controls whether or not we should try to go down the
+ * dioread_nolock code paths, which makes it safe to avoid taking
+ * i_mutex for direct I/O reads.  This only works for extent-based
+ * files, and it doesn't work for nobh or if data journaling is
+ * enabled, since the dioread_nolock code uses b_private to pass
+ * information back to the I/O completion handler, and this conflicts
+ * with the jbd's use of b_private.
+ */
+static inline int ext4_should_dioread_nolock(struct inode *inode)
+{
+	if (!test_opt(inode->i_sb, DIOREAD_NOLOCK))
+		return 0;
+	if (test_opt(inode->i_sb, NOBH))
+		return 0;
+	if (!S_ISREG(inode->i_mode))
+		return 0;
+	if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+		return 0;
+	if (ext4_should_journal_data(inode))
+		return 0;
+	return 1;
+}
+
 #endif	/* _EXT4_JBD2_H */
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 765a4826b118..94c8ee81f5e1 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -195,8 +195,7 @@ static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
 		if (S_ISREG(inode->i_mode))
 			block_group++;
 	}
-	bg_start = (block_group * EXT4_BLOCKS_PER_GROUP(inode->i_sb)) +
-		le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_first_data_block);
+	bg_start = ext4_group_first_block_no(inode->i_sb, block_group);
 	last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1;
 
 	/*
@@ -440,7 +439,7 @@ static int __ext4_ext_check(const char *function, struct inode *inode,
 	return 0;
 
 corrupted:
-	ext4_error(inode->i_sb, function,
+	__ext4_error(inode->i_sb, function,
 			"bad header/extent in inode #%lu: %s - magic %x, "
 			"entries %u, max %u(%u), depth %u(%u)",
 			inode->i_ino, error_msg, le16_to_cpu(eh->eh_magic),
@@ -703,7 +702,12 @@ ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
 		}
 		eh = ext_block_hdr(bh);
 		ppos++;
-		BUG_ON(ppos > depth);
+		if (unlikely(ppos > depth)) {
+			put_bh(bh);
+			EXT4_ERROR_INODE(inode,
+					 "ppos %d > depth %d", ppos, depth);
+			goto err;
+		}
 		path[ppos].p_bh = bh;
 		path[ppos].p_hdr = eh;
 		i--;
@@ -749,7 +753,12 @@ int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
 	if (err)
 		return err;
 
-	BUG_ON(logical == le32_to_cpu(curp->p_idx->ei_block));
+	if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) {
+		EXT4_ERROR_INODE(inode,
+				 "logical %d == ei_block %d!",
+				 logical, le32_to_cpu(curp->p_idx->ei_block));
+		return -EIO;
+	}
 	len = EXT_MAX_INDEX(curp->p_hdr) - curp->p_idx;
 	if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
 		/* insert after */
@@ -779,9 +788,17 @@ int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
 	ext4_idx_store_pblock(ix, ptr);
 	le16_add_cpu(&curp->p_hdr->eh_entries, 1);
 
-	BUG_ON(le16_to_cpu(curp->p_hdr->eh_entries)
-			     > le16_to_cpu(curp->p_hdr->eh_max));
-	BUG_ON(ix > EXT_LAST_INDEX(curp->p_hdr));
+	if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries)
+			     > le16_to_cpu(curp->p_hdr->eh_max))) {
+		EXT4_ERROR_INODE(inode,
+				 "logical %d == ei_block %d!",
+				 logical, le32_to_cpu(curp->p_idx->ei_block));
+		return -EIO;
+	}
+	if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) {
+		EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!");
+		return -EIO;
+	}
 
 	err = ext4_ext_dirty(handle, inode, curp);
 	ext4_std_error(inode->i_sb, err);
@@ -819,7 +836,10 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
 
 	/* if current leaf will be split, then we should use
 	 * border from split point */
-	BUG_ON(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr));
+	if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) {
+		EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!");
+		return -EIO;
+	}
 	if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
 		border = path[depth].p_ext[1].ee_block;
 		ext_debug("leaf will be split."
@@ -860,7 +880,11 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
 
 	/* initialize new leaf */
 	newblock = ablocks[--a];
-	BUG_ON(newblock == 0);
+	if (unlikely(newblock == 0)) {
+		EXT4_ERROR_INODE(inode, "newblock == 0!");
+		err = -EIO;
+		goto cleanup;
+	}
 	bh = sb_getblk(inode->i_sb, newblock);
 	if (!bh) {
 		err = -EIO;
@@ -880,7 +904,14 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
 	ex = EXT_FIRST_EXTENT(neh);
 
 	/* move remainder of path[depth] to the new leaf */
-	BUG_ON(path[depth].p_hdr->eh_entries != path[depth].p_hdr->eh_max);
+	if (unlikely(path[depth].p_hdr->eh_entries !=
+		     path[depth].p_hdr->eh_max)) {
+		EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!",
+				 path[depth].p_hdr->eh_entries,
+				 path[depth].p_hdr->eh_max);
+		err = -EIO;
+		goto cleanup;
+	}
 	/* start copy from next extent */
 	/* TODO: we could do it by single memmove */
 	m = 0;
@@ -927,7 +958,11 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
 
 	/* create intermediate indexes */
 	k = depth - at - 1;
-	BUG_ON(k < 0);
+	if (unlikely(k < 0)) {
+		EXT4_ERROR_INODE(inode, "k %d < 0!", k);
+		err = -EIO;
+		goto cleanup;
+	}
 	if (k)
 		ext_debug("create %d intermediate indices\n", k);
 	/* insert new index into current index block */
@@ -964,8 +999,14 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
 
 		ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
 				EXT_MAX_INDEX(path[i].p_hdr));
-		BUG_ON(EXT_MAX_INDEX(path[i].p_hdr) !=
-				EXT_LAST_INDEX(path[i].p_hdr));
+		if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) !=
+					EXT_LAST_INDEX(path[i].p_hdr))) {
+			EXT4_ERROR_INODE(inode,
+					 "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!",
+					 le32_to_cpu(path[i].p_ext->ee_block));
+			err = -EIO;
+			goto cleanup;
+		}
 		while (path[i].p_idx <= EXT_MAX_INDEX(path[i].p_hdr)) {
 			ext_debug("%d: move %d:%llu in new index %llu\n", i,
 					le32_to_cpu(path[i].p_idx->ei_block),
@@ -1203,7 +1244,10 @@ ext4_ext_search_left(struct inode *inode, struct ext4_ext_path *path,
 	struct ext4_extent *ex;
 	int depth, ee_len;
 
-	BUG_ON(path == NULL);
+	if (unlikely(path == NULL)) {
+		EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
+		return -EIO;
+	}
 	depth = path->p_depth;
 	*phys = 0;
 
@@ -1217,15 +1261,33 @@ ext4_ext_search_left(struct inode *inode, struct ext4_ext_path *path,
 	ex = path[depth].p_ext;
 	ee_len = ext4_ext_get_actual_len(ex);
 	if (*logical < le32_to_cpu(ex->ee_block)) {
-		BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
+		if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
+			EXT4_ERROR_INODE(inode,
+					 "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!",
+					 *logical, le32_to_cpu(ex->ee_block));
+			return -EIO;
+		}
 		while (--depth >= 0) {
 			ix = path[depth].p_idx;
-			BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
+			if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
+				EXT4_ERROR_INODE(inode,
+				  "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!",
+				  ix != NULL ? ix->ei_block : 0,
+				  EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ?
+				    EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block : 0,
+				  depth);
+				return -EIO;
+			}
 		}
 		return 0;
 	}
 
-	BUG_ON(*logical < (le32_to_cpu(ex->ee_block) + ee_len));
+	if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
+		EXT4_ERROR_INODE(inode,
+				 "logical %d < ee_block %d + ee_len %d!",
+				 *logical, le32_to_cpu(ex->ee_block), ee_len);
+		return -EIO;
+	}
 
 	*logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
 	*phys = ext_pblock(ex) + ee_len - 1;
@@ -1251,7 +1313,10 @@ ext4_ext_search_right(struct inode *inode, struct ext4_ext_path *path,
 	int depth;	/* Note, NOT eh_depth; depth from top of tree */
 	int ee_len;
 
-	BUG_ON(path == NULL);
+	if (unlikely(path == NULL)) {
+		EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
+		return -EIO;
+	}
 	depth = path->p_depth;
 	*phys = 0;
 
@@ -1265,17 +1330,32 @@ ext4_ext_search_right(struct inode *inode, struct ext4_ext_path *path,
 	ex = path[depth].p_ext;
 	ee_len = ext4_ext_get_actual_len(ex);
 	if (*logical < le32_to_cpu(ex->ee_block)) {
-		BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
+		if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
+			EXT4_ERROR_INODE(inode,
+					 "first_extent(path[%d].p_hdr) != ex",
+					 depth);
+			return -EIO;
+		}
 		while (--depth >= 0) {
 			ix = path[depth].p_idx;
-			BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
+			if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
+				EXT4_ERROR_INODE(inode,
+						 "ix != EXT_FIRST_INDEX *logical %d!",
+						 *logical);
+				return -EIO;
+			}
 		}
 		*logical = le32_to_cpu(ex->ee_block);
 		*phys = ext_pblock(ex);
 		return 0;
 	}
 
-	BUG_ON(*logical < (le32_to_cpu(ex->ee_block) + ee_len));
+	if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
+		EXT4_ERROR_INODE(inode,
+				 "logical %d < ee_block %d + ee_len %d!",
+				 *logical, le32_to_cpu(ex->ee_block), ee_len);
+		return -EIO;
+	}
 
 	if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
 		/* next allocated block in this leaf */
@@ -1414,8 +1494,12 @@ static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
 
 	eh = path[depth].p_hdr;
 	ex = path[depth].p_ext;
-	BUG_ON(ex == NULL);
-	BUG_ON(eh == NULL);
+
+	if (unlikely(ex == NULL || eh == NULL)) {
+		EXT4_ERROR_INODE(inode,
+				 "ex %p == NULL or eh %p == NULL", ex, eh);
+		return -EIO;
+	}
 
 	if (depth == 0) {
 		/* there is no tree at all */
@@ -1538,8 +1622,9 @@ int ext4_ext_try_to_merge(struct inode *inode,
 		merge_done = 1;
 		WARN_ON(eh->eh_entries == 0);
 		if (!eh->eh_entries)
-			ext4_error(inode->i_sb, "ext4_ext_try_to_merge",
-			   "inode#%lu, eh->eh_entries = 0!", inode->i_ino);
+			ext4_error(inode->i_sb,
+				   "inode#%lu, eh->eh_entries = 0!",
+				   inode->i_ino);
 	}
 
 	return merge_done;
@@ -1612,13 +1697,19 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
 	ext4_lblk_t next;
 	unsigned uninitialized = 0;
 
-	BUG_ON(ext4_ext_get_actual_len(newext) == 0);
+	if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
+		EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
+		return -EIO;
+	}
 	depth = ext_depth(inode);
 	ex = path[depth].p_ext;
-	BUG_ON(path[depth].p_hdr == NULL);
+	if (unlikely(path[depth].p_hdr == NULL)) {
+		EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
+		return -EIO;
+	}
 
 	/* try to insert block into found extent and return */
-	if (ex && (flag != EXT4_GET_BLOCKS_DIO_CREATE_EXT)
+	if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)
 		&& ext4_can_extents_be_merged(inode, ex, newext)) {
 		ext_debug("append [%d]%d block to %d:[%d]%d (from %llu)\n",
 				ext4_ext_is_uninitialized(newext),
@@ -1739,7 +1830,7 @@ has_space:
 
 merge:
 	/* try to merge extents to the right */
-	if (flag != EXT4_GET_BLOCKS_DIO_CREATE_EXT)
+	if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
 		ext4_ext_try_to_merge(inode, path, nearex);
 
 	/* try to merge extents to the left */
@@ -1787,7 +1878,11 @@ int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
 		}
 
 		depth = ext_depth(inode);
-		BUG_ON(path[depth].p_hdr == NULL);
+		if (unlikely(path[depth].p_hdr == NULL)) {
+			EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
+			err = -EIO;
+			break;
+		}
 		ex = path[depth].p_ext;
 		next = ext4_ext_next_allocated_block(path);
 
@@ -1838,7 +1933,11 @@ int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
 			cbex.ec_type = EXT4_EXT_CACHE_EXTENT;
 		}
 
-		BUG_ON(cbex.ec_len == 0);
+		if (unlikely(cbex.ec_len == 0)) {
+			EXT4_ERROR_INODE(inode, "cbex.ec_len == 0");
+			err = -EIO;
+			break;
+		}
 		err = func(inode, path, &cbex, ex, cbdata);
 		ext4_ext_drop_refs(path);
 
@@ -1952,7 +2051,7 @@ ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
 
 	BUG_ON(cex->ec_type != EXT4_EXT_CACHE_GAP &&
 			cex->ec_type != EXT4_EXT_CACHE_EXTENT);
-	if (block >= cex->ec_block && block < cex->ec_block + cex->ec_len) {
+	if (in_range(block, cex->ec_block, cex->ec_len)) {
 		ex->ee_block = cpu_to_le32(cex->ec_block);
 		ext4_ext_store_pblock(ex, cex->ec_start);
 		ex->ee_len = cpu_to_le16(cex->ec_len);
@@ -1981,7 +2080,10 @@ static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
 	/* free index block */
 	path--;
 	leaf = idx_pblock(path->p_idx);
-	BUG_ON(path->p_hdr->eh_entries == 0);
+	if (unlikely(path->p_hdr->eh_entries == 0)) {
+		EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
+		return -EIO;
+	}
 	err = ext4_ext_get_access(handle, inode, path);
 	if (err)
 		return err;
@@ -2119,8 +2221,10 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 	if (!path[depth].p_hdr)
 		path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
 	eh = path[depth].p_hdr;
-	BUG_ON(eh == NULL);
-
+	if (unlikely(path[depth].p_hdr == NULL)) {
+		EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
+		return -EIO;
+	}
 	/* find where to start removing */
 	ex = EXT_LAST_EXTENT(eh);
 
@@ -2983,7 +3087,7 @@ fix_extent_len:
 	ext4_ext_dirty(handle, inode, path + depth);
 	return err;
 }
-static int ext4_convert_unwritten_extents_dio(handle_t *handle,
+static int ext4_convert_unwritten_extents_endio(handle_t *handle,
 					      struct inode *inode,
 					      struct ext4_ext_path *path)
 {
@@ -3063,8 +3167,8 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
 		  flags, allocated);
 	ext4_ext_show_leaf(inode, path);
 
-	/* DIO get_block() before submit the IO, split the extent */
-	if (flags == EXT4_GET_BLOCKS_DIO_CREATE_EXT) {
+	/* get_block() before submit the IO, split the extent */
+	if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
 		ret = ext4_split_unwritten_extents(handle,
 						inode, path, iblock,
 						max_blocks, flags);
@@ -3074,14 +3178,16 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
 		 * completed
 		 */
 		if (io)
-			io->flag = DIO_AIO_UNWRITTEN;
+			io->flag = EXT4_IO_UNWRITTEN;
 		else
-			EXT4_I(inode)->i_state |= EXT4_STATE_DIO_UNWRITTEN;
+			ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
+		if (ext4_should_dioread_nolock(inode))
+			set_buffer_uninit(bh_result);
 		goto out;
 	}
-	/* async DIO end_io complete, convert the filled extent to written */
-	if (flags == EXT4_GET_BLOCKS_DIO_CONVERT_EXT) {
-		ret = ext4_convert_unwritten_extents_dio(handle, inode,
+	/* IO end_io complete, convert the filled extent to written */
+	if ((flags & EXT4_GET_BLOCKS_CONVERT)) {
+		ret = ext4_convert_unwritten_extents_endio(handle, inode,
 							path);
 		if (ret >= 0)
 			ext4_update_inode_fsync_trans(handle, inode, 1);
@@ -3185,7 +3291,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 {
 	struct ext4_ext_path *path = NULL;
 	struct ext4_extent_header *eh;
-	struct ext4_extent newex, *ex;
+	struct ext4_extent newex, *ex, *last_ex;
 	ext4_fsblk_t newblock;
 	int err = 0, depth, ret, cache_type;
 	unsigned int allocated = 0;
@@ -3237,10 +3343,10 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 	 * this situation is possible, though, _during_ tree modification;
 	 * this is why assert can't be put in ext4_ext_find_extent()
 	 */
-	if (path[depth].p_ext == NULL && depth != 0) {
-		ext4_error(inode->i_sb, __func__, "bad extent address "
-			   "inode: %lu, iblock: %d, depth: %d",
-			   inode->i_ino, iblock, depth);
+	if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
+		EXT4_ERROR_INODE(inode, "bad extent address "
+				 "iblock: %d, depth: %d pblock %lld",
+				 iblock, depth, path[depth].p_block);
 		err = -EIO;
 		goto out2;
 	}
@@ -3258,7 +3364,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 		 */
 		ee_len = ext4_ext_get_actual_len(ex);
 		/* if found extent covers block, simply return it */
-		if (iblock >= ee_block && iblock < ee_block + ee_len) {
+		if (in_range(iblock, ee_block, ee_len)) {
 			newblock = iblock - ee_block + ee_start;
 			/* number of remaining blocks in the extent */
 			allocated = ee_len - (iblock - ee_block);
@@ -3350,21 +3456,35 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 	if (flags & EXT4_GET_BLOCKS_UNINIT_EXT){
 		ext4_ext_mark_uninitialized(&newex);
 		/*
-		 * io_end structure was created for every async
-		 * direct IO write to the middle of the file.
-		 * To avoid unecessary convertion for every aio dio rewrite
-		 * to the mid of file, here we flag the IO that is really
-		 * need the convertion.
+		 * io_end structure was created for every IO write to an
+		 * uninitialized extent. To avoid unecessary conversion,
+		 * here we flag the IO that really needs the conversion.
 		 * For non asycn direct IO case, flag the inode state
 		 * that we need to perform convertion when IO is done.
 		 */
-		if (flags == EXT4_GET_BLOCKS_DIO_CREATE_EXT) {
+		if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
 			if (io)
-				io->flag = DIO_AIO_UNWRITTEN;
+				io->flag = EXT4_IO_UNWRITTEN;
 			else
-				EXT4_I(inode)->i_state |=
-					EXT4_STATE_DIO_UNWRITTEN;;
+				ext4_set_inode_state(inode,
+						     EXT4_STATE_DIO_UNWRITTEN);
+		}
+		if (ext4_should_dioread_nolock(inode))
+			set_buffer_uninit(bh_result);
+	}
+
+	if (unlikely(EXT4_I(inode)->i_flags & EXT4_EOFBLOCKS_FL)) {
+		if (unlikely(!eh->eh_entries)) {
+			EXT4_ERROR_INODE(inode,
+					 "eh->eh_entries == 0 ee_block %d",
+					 ex->ee_block);
+			err = -EIO;
+			goto out2;
 		}
+		last_ex = EXT_LAST_EXTENT(eh);
+		if (iblock + ar.len > le32_to_cpu(last_ex->ee_block)
+		    + ext4_ext_get_actual_len(last_ex))
+			EXT4_I(inode)->i_flags &= ~EXT4_EOFBLOCKS_FL;
 	}
 	err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
 	if (err) {
@@ -3499,6 +3619,13 @@ static void ext4_falloc_update_inode(struct inode *inode,
 			i_size_write(inode, new_size);
 		if (new_size > EXT4_I(inode)->i_disksize)
 			ext4_update_i_disksize(inode, new_size);
+	} else {
+		/*
+		 * Mark that we allocate beyond EOF so the subsequent truncate
+		 * can proceed even if the new size is the same as i_size.
+		 */
+		if (new_size > i_size_read(inode))
+			EXT4_I(inode)->i_flags |= EXT4_EOFBLOCKS_FL;
 	}
 
 }
@@ -3603,7 +3730,7 @@ retry:
  * Returns 0 on success.
  */
 int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
-				    loff_t len)
+				    ssize_t len)
 {
 	handle_t *handle;
 	ext4_lblk_t block;
@@ -3635,7 +3762,7 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
 		map_bh.b_state = 0;
 		ret = ext4_get_blocks(handle, inode, block,
 				      max_blocks, &map_bh,
-				      EXT4_GET_BLOCKS_DIO_CONVERT_EXT);
+				      EXT4_GET_BLOCKS_IO_CONVERT_EXT);
 		if (ret <= 0) {
 			WARN_ON(ret <= 0);
 			printk(KERN_ERR "%s: ext4_ext_get_blocks "
@@ -3739,7 +3866,7 @@ static int ext4_xattr_fiemap(struct inode *inode,
 	int error = 0;
 
 	/* in-inode? */
-	if (EXT4_I(inode)->i_state & EXT4_STATE_XATTR) {
+	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
 		struct ext4_iloc iloc;
 		int offset;	/* offset of xattr in inode */
 
@@ -3767,7 +3894,6 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len)
 {
 	ext4_lblk_t start_blk;
-	ext4_lblk_t len_blks;
 	int error = 0;
 
 	/* fallback to generic here if not in extents fmt */
@@ -3781,8 +3907,14 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
 		error = ext4_xattr_fiemap(inode, fieinfo);
 	} else {
+		ext4_lblk_t len_blks;
+		__u64 last_blk;
+
 		start_blk = start >> inode->i_sb->s_blocksize_bits;
-		len_blks = len >> inode->i_sb->s_blocksize_bits;
+		last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
+		if (last_blk >= EXT_MAX_BLOCK)
+			last_blk = EXT_MAX_BLOCK-1;
+		len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
 
 		/*
 		 * Walk the extent tree gathering extent information.
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 9630583cef28..d0776e410f34 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -23,6 +23,7 @@
 #include <linux/jbd2.h>
 #include <linux/mount.h>
 #include <linux/path.h>
+#include <linux/quotaops.h>
 #include "ext4.h"
 #include "ext4_jbd2.h"
 #include "xattr.h"
@@ -35,9 +36,9 @@
  */
 static int ext4_release_file(struct inode *inode, struct file *filp)
 {
-	if (EXT4_I(inode)->i_state & EXT4_STATE_DA_ALLOC_CLOSE) {
+	if (ext4_test_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE)) {
 		ext4_alloc_da_blocks(inode);
-		EXT4_I(inode)->i_state &= ~EXT4_STATE_DA_ALLOC_CLOSE;
+		ext4_clear_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE);
 	}
 	/* if we are the last writer on the inode, drop the block reservation */
 	if ((filp->f_mode & FMODE_WRITE) &&
@@ -116,18 +117,16 @@ static int ext4_file_open(struct inode * inode, struct file * filp)
 		 * devices or filesystem images.
 		 */
 		memset(buf, 0, sizeof(buf));
-		path.mnt = mnt->mnt_parent;
-		path.dentry = mnt->mnt_mountpoint;
-		path_get(&path);
+		path.mnt = mnt;
+		path.dentry = mnt->mnt_root;
 		cp = d_path(&path, buf, sizeof(buf));
-		path_put(&path);
 		if (!IS_ERR(cp)) {
 			memcpy(sbi->s_es->s_last_mounted, cp,
 			       sizeof(sbi->s_es->s_last_mounted));
 			sb->s_dirt = 1;
 		}
 	}
-	return generic_file_open(inode, filp);
+	return dquot_file_open(inode, filp);
 }
 
 const struct file_operations ext4_file_operations = {
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index 98bd140aad01..0d0c3239c1cd 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -63,7 +63,7 @@ int ext4_sync_file(struct file *file, struct dentry *dentry, int datasync)
 	if (inode->i_sb->s_flags & MS_RDONLY)
 		return 0;
 
-	ret = flush_aio_dio_completed_IO(inode);
+	ret = flush_completed_IO(inode);
 	if (ret < 0)
 		return ret;
 	
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index f3624ead4f6c..361c0b9962a8 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -76,8 +76,7 @@ unsigned ext4_init_inode_bitmap(struct super_block *sb, struct buffer_head *bh,
 	/* If checksum is bad mark all blocks and inodes use to prevent
 	 * allocation, essentially implementing a per-group read-only flag. */
 	if (!ext4_group_desc_csum_verify(sbi, block_group, gdp)) {
-		ext4_error(sb, __func__, "Checksum bad for group %u",
-			   block_group);
+		ext4_error(sb, "Checksum bad for group %u", block_group);
 		ext4_free_blks_set(sb, gdp, 0);
 		ext4_free_inodes_set(sb, gdp, 0);
 		ext4_itable_unused_set(sb, gdp, 0);
@@ -111,8 +110,7 @@ ext4_read_inode_bitmap(struct super_block *sb, ext4_group_t block_group)
 	bitmap_blk = ext4_inode_bitmap(sb, desc);
 	bh = sb_getblk(sb, bitmap_blk);
 	if (unlikely(!bh)) {
-		ext4_error(sb, __func__,
-			    "Cannot read inode bitmap - "
+		ext4_error(sb, "Cannot read inode bitmap - "
 			    "block_group = %u, inode_bitmap = %llu",
 			    block_group, bitmap_blk);
 		return NULL;
@@ -153,8 +151,7 @@ ext4_read_inode_bitmap(struct super_block *sb, ext4_group_t block_group)
 	set_bitmap_uptodate(bh);
 	if (bh_submit_read(bh) < 0) {
 		put_bh(bh);
-		ext4_error(sb, __func__,
-			    "Cannot read inode bitmap - "
+		ext4_error(sb, "Cannot read inode bitmap - "
 			    "block_group = %u, inode_bitmap = %llu",
 			    block_group, bitmap_blk);
 		return NULL;
@@ -217,10 +214,10 @@ void ext4_free_inode(handle_t *handle, struct inode *inode)
 	 * Note: we must free any quota before locking the superblock,
 	 * as writing the quota to disk may need the lock as well.
 	 */
-	vfs_dq_init(inode);
+	dquot_initialize(inode);
 	ext4_xattr_delete_inode(handle, inode);
-	vfs_dq_free_inode(inode);
-	vfs_dq_drop(inode);
+	dquot_free_inode(inode);
+	dquot_drop(inode);
 
 	is_directory = S_ISDIR(inode->i_mode);
 
@@ -229,8 +226,7 @@ void ext4_free_inode(handle_t *handle, struct inode *inode)
 
 	es = EXT4_SB(sb)->s_es;
 	if (ino < EXT4_FIRST_INO(sb) || ino > le32_to_cpu(es->s_inodes_count)) {
-		ext4_error(sb, "ext4_free_inode",
-			   "reserved or nonexistent inode %lu", ino);
+		ext4_error(sb, "reserved or nonexistent inode %lu", ino);
 		goto error_return;
 	}
 	block_group = (ino - 1) / EXT4_INODES_PER_GROUP(sb);
@@ -248,8 +244,7 @@ void ext4_free_inode(handle_t *handle, struct inode *inode)
 	cleared = ext4_clear_bit_atomic(ext4_group_lock_ptr(sb, block_group),
 					bit, bitmap_bh->b_data);
 	if (!cleared)
-		ext4_error(sb, "ext4_free_inode",
-			   "bit already cleared for inode %lu", ino);
+		ext4_error(sb, "bit already cleared for inode %lu", ino);
 	else {
 		gdp = ext4_get_group_desc(sb, block_group, &bh2);
 
@@ -736,8 +731,7 @@ static int ext4_claim_inode(struct super_block *sb,
 	if ((group == 0 && ino < EXT4_FIRST_INO(sb)) ||
 			ino > EXT4_INODES_PER_GROUP(sb)) {
 		ext4_unlock_group(sb, group);
-		ext4_error(sb, __func__,
-			   "reserved inode or inode > inodes count - "
+		ext4_error(sb, "reserved inode or inode > inodes count - "
 			   "block_group = %u, inode=%lu", group,
 			   ino + group * EXT4_INODES_PER_GROUP(sb));
 		return 1;
@@ -904,7 +898,7 @@ repeat_in_this_group:
 				BUFFER_TRACE(inode_bitmap_bh,
 					"call ext4_handle_dirty_metadata");
 				err = ext4_handle_dirty_metadata(handle,
-								 inode,
+								 NULL,
 							inode_bitmap_bh);
 				if (err)
 					goto fail;
@@ -1029,15 +1023,16 @@ got:
 	inode->i_generation = sbi->s_next_generation++;
 	spin_unlock(&sbi->s_next_gen_lock);
 
-	ei->i_state = EXT4_STATE_NEW;
+	ei->i_state_flags = 0;
+	ext4_set_inode_state(inode, EXT4_STATE_NEW);
 
 	ei->i_extra_isize = EXT4_SB(sb)->s_want_extra_isize;
 
 	ret = inode;
-	if (vfs_dq_alloc_inode(inode)) {
-		err = -EDQUOT;
+	dquot_initialize(inode);
+	err = dquot_alloc_inode(inode);
+	if (err)
 		goto fail_drop;
-	}
 
 	err = ext4_init_acl(handle, inode, dir);
 	if (err)
@@ -1074,10 +1069,10 @@ really_out:
 	return ret;
 
 fail_free_drop:
-	vfs_dq_free_inode(inode);
+	dquot_free_inode(inode);
 
 fail_drop:
-	vfs_dq_drop(inode);
+	dquot_drop(inode);
 	inode->i_flags |= S_NOQUOTA;
 	inode->i_nlink = 0;
 	unlock_new_inode(inode);
@@ -1098,8 +1093,7 @@ struct inode *ext4_orphan_get(struct super_block *sb, unsigned long ino)
 
 	/* Error cases - e2fsck has already cleaned up for us */
 	if (ino > max_ino) {
-		ext4_warning(sb, __func__,
-			     "bad orphan ino %lu!  e2fsck was run?", ino);
+		ext4_warning(sb, "bad orphan ino %lu!  e2fsck was run?", ino);
 		goto error;
 	}
 
@@ -1107,8 +1101,7 @@ struct inode *ext4_orphan_get(struct super_block *sb, unsigned long ino)
 	bit = (ino - 1) % EXT4_INODES_PER_GROUP(sb);
 	bitmap_bh = ext4_read_inode_bitmap(sb, block_group);
 	if (!bitmap_bh) {
-		ext4_warning(sb, __func__,
-			     "inode bitmap error for orphan %lu", ino);
+		ext4_warning(sb, "inode bitmap error for orphan %lu", ino);
 		goto error;
 	}
 
@@ -1140,8 +1133,7 @@ iget_failed:
 	err = PTR_ERR(inode);
 	inode = NULL;
 bad_orphan:
-	ext4_warning(sb, __func__,
-		     "bad orphan inode %lu!  e2fsck was run?", ino);
+	ext4_warning(sb, "bad orphan inode %lu!  e2fsck was run?", ino);
 	printk(KERN_NOTICE "ext4_test_bit(bit=%d, block=%llu) = %d\n",
 	       bit, (unsigned long long)bitmap_bh->b_blocknr,
 	       ext4_test_bit(bit, bitmap_bh->b_data));
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index e11952404e02..986120f30066 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -38,6 +38,7 @@
 #include <linux/uio.h>
 #include <linux/bio.h>
 #include <linux/workqueue.h>
+#include <linux/kernel.h>
 
 #include "ext4_jbd2.h"
 #include "xattr.h"
@@ -170,6 +171,9 @@ void ext4_delete_inode(struct inode *inode)
 	handle_t *handle;
 	int err;
 
+	if (!is_bad_inode(inode))
+		dquot_initialize(inode);
+
 	if (ext4_should_order_data(inode))
 		ext4_begin_ordered_truncate(inode, 0);
 	truncate_inode_pages(&inode->i_data, 0);
@@ -194,7 +198,7 @@ void ext4_delete_inode(struct inode *inode)
 	inode->i_size = 0;
 	err = ext4_mark_inode_dirty(handle, inode);
 	if (err) {
-		ext4_warning(inode->i_sb, __func__,
+		ext4_warning(inode->i_sb,
 			     "couldn't mark inode dirty (err %d)", err);
 		goto stop_handle;
 	}
@@ -212,7 +216,7 @@ void ext4_delete_inode(struct inode *inode)
 		if (err > 0)
 			err = ext4_journal_restart(handle, 3);
 		if (err != 0) {
-			ext4_warning(inode->i_sb, __func__,
+			ext4_warning(inode->i_sb,
 				     "couldn't extend journal (err %d)", err);
 		stop_handle:
 			ext4_journal_stop(handle);
@@ -323,8 +327,7 @@ static int ext4_block_to_path(struct inode *inode,
 		offsets[n++] = i_block & (ptrs - 1);
 		final = ptrs;
 	} else {
-		ext4_warning(inode->i_sb, "ext4_block_to_path",
-			     "block %lu > max in inode %lu",
+		ext4_warning(inode->i_sb, "block %lu > max in inode %lu",
 			     i_block + direct_blocks +
 			     indirect_blocks + double_blocks, inode->i_ino);
 	}
@@ -344,7 +347,7 @@ static int __ext4_check_blockref(const char *function, struct inode *inode,
 		if (blk &&
 		    unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb),
 						    blk, 1))) {
-			ext4_error(inode->i_sb, function,
+			__ext4_error(inode->i_sb, function,
 				   "invalid block reference %u "
 				   "in inode #%lu", blk, inode->i_ino);
 			return -EIO;
@@ -607,7 +610,14 @@ static int ext4_alloc_blocks(handle_t *handle, struct inode *inode,
 		if (*err)
 			goto failed_out;
 
-		BUG_ON(current_block + count > EXT4_MAX_BLOCK_FILE_PHYS);
+		if (unlikely(current_block + count > EXT4_MAX_BLOCK_FILE_PHYS)) {
+			EXT4_ERROR_INODE(inode,
+					 "current_block %llu + count %lu > %d!",
+					 current_block, count,
+					 EXT4_MAX_BLOCK_FILE_PHYS);
+			*err = -EIO;
+			goto failed_out;
+		}
 
 		target -= count;
 		/* allocate blocks for indirect blocks */
@@ -643,7 +653,14 @@ static int ext4_alloc_blocks(handle_t *handle, struct inode *inode,
 		ar.flags = EXT4_MB_HINT_DATA;
 
 	current_block = ext4_mb_new_blocks(handle, &ar, err);
-	BUG_ON(current_block + ar.len > EXT4_MAX_BLOCK_FILE_PHYS);
+	if (unlikely(current_block + ar.len > EXT4_MAX_BLOCK_FILE_PHYS)) {
+		EXT4_ERROR_INODE(inode,
+				 "current_block %llu + ar.len %d > %d!",
+				 current_block, ar.len,
+				 EXT4_MAX_BLOCK_FILE_PHYS);
+		*err = -EIO;
+		goto failed_out;
+	}
 
 	if (*err && (target == blks)) {
 		/*
@@ -1061,6 +1078,7 @@ void ext4_da_update_reserve_space(struct inode *inode,
 	int mdb_free = 0, allocated_meta_blocks = 0;
 
 	spin_lock(&ei->i_block_reservation_lock);
+	trace_ext4_da_update_reserve_space(inode, used);
 	if (unlikely(used > ei->i_reserved_data_blocks)) {
 		ext4_msg(inode->i_sb, KERN_NOTICE, "%s: ino %lu, used %d "
 			 "with only %d reserved data blocks\n",
@@ -1093,9 +1111,9 @@ void ext4_da_update_reserve_space(struct inode *inode,
 
 	/* Update quota subsystem */
 	if (quota_claim) {
-		vfs_dq_claim_block(inode, used);
+		dquot_claim_block(inode, used);
 		if (mdb_free)
-			vfs_dq_release_reservation_block(inode, mdb_free);
+			dquot_release_reservation_block(inode, mdb_free);
 	} else {
 		/*
 		 * We did fallocate with an offset that is already delayed
@@ -1106,8 +1124,8 @@ void ext4_da_update_reserve_space(struct inode *inode,
 		 * that
 		 */
 		if (allocated_meta_blocks)
-			vfs_dq_claim_block(inode, allocated_meta_blocks);
-		vfs_dq_release_reservation_block(inode, mdb_free + used);
+			dquot_claim_block(inode, allocated_meta_blocks);
+		dquot_release_reservation_block(inode, mdb_free + used);
 	}
 
 	/*
@@ -1124,7 +1142,7 @@ static int check_block_validity(struct inode *inode, const char *msg,
 				sector_t logical, sector_t phys, int len)
 {
 	if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), phys, len)) {
-		ext4_error(inode->i_sb, msg,
+		__ext4_error(inode->i_sb, msg,
 			   "inode #%lu logical block %llu mapped to %llu "
 			   "(size %d)", inode->i_ino,
 			   (unsigned long long) logical,
@@ -1306,7 +1324,7 @@ int ext4_get_blocks(handle_t *handle, struct inode *inode, sector_t block,
 			 * i_data's format changing.  Force the migrate
 			 * to fail by clearing migrate flags
 			 */
-			EXT4_I(inode)->i_state &= ~EXT4_STATE_EXT_MIGRATE;
+			ext4_clear_inode_state(inode, EXT4_STATE_EXT_MIGRATE);
 		}
 
 		/*
@@ -1534,6 +1552,8 @@ static void ext4_truncate_failed_write(struct inode *inode)
 	ext4_truncate(inode);
 }
 
+static int ext4_get_block_write(struct inode *inode, sector_t iblock,
+		   struct buffer_head *bh_result, int create);
 static int ext4_write_begin(struct file *file, struct address_space *mapping,
 			    loff_t pos, unsigned len, unsigned flags,
 			    struct page **pagep, void **fsdata)
@@ -1575,8 +1595,12 @@ retry:
 	}
 	*pagep = page;
 
-	ret = block_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
-				ext4_get_block);
+	if (ext4_should_dioread_nolock(inode))
+		ret = block_write_begin(file, mapping, pos, len, flags, pagep,
+				fsdata, ext4_get_block_write);
+	else
+		ret = block_write_begin(file, mapping, pos, len, flags, pagep,
+				fsdata, ext4_get_block);
 
 	if (!ret && ext4_should_journal_data(inode)) {
 		ret = walk_page_buffers(handle, page_buffers(page),
@@ -1793,7 +1817,7 @@ static int ext4_journalled_write_end(struct file *file,
 	new_i_size = pos + copied;
 	if (new_i_size > inode->i_size)
 		i_size_write(inode, pos+copied);
-	EXT4_I(inode)->i_state |= EXT4_STATE_JDATA;
+	ext4_set_inode_state(inode, EXT4_STATE_JDATA);
 	if (new_i_size > EXT4_I(inode)->i_disksize) {
 		ext4_update_i_disksize(inode, new_i_size);
 		ret2 = ext4_mark_inode_dirty(handle, inode);
@@ -1836,6 +1860,7 @@ static int ext4_da_reserve_space(struct inode *inode, sector_t lblock)
 	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 	struct ext4_inode_info *ei = EXT4_I(inode);
 	unsigned long md_needed, md_reserved;
+	int ret;
 
 	/*
 	 * recalculate the amount of metadata blocks to reserve
@@ -1846,6 +1871,7 @@ repeat:
 	spin_lock(&ei->i_block_reservation_lock);
 	md_reserved = ei->i_reserved_meta_blocks;
 	md_needed = ext4_calc_metadata_amount(inode, lblock);
+	trace_ext4_da_reserve_space(inode, md_needed);
 	spin_unlock(&ei->i_block_reservation_lock);
 
 	/*
@@ -1853,11 +1879,12 @@ repeat:
 	 * later. Real quota accounting is done at pages writeout
 	 * time.
 	 */
-	if (vfs_dq_reserve_block(inode, md_needed + 1))
-		return -EDQUOT;
+	ret = dquot_reserve_block(inode, md_needed + 1);
+	if (ret)
+		return ret;
 
 	if (ext4_claim_free_blocks(sbi, md_needed + 1)) {
-		vfs_dq_release_reservation_block(inode, md_needed + 1);
+		dquot_release_reservation_block(inode, md_needed + 1);
 		if (ext4_should_retry_alloc(inode->i_sb, &retries)) {
 			yield();
 			goto repeat;
@@ -1914,7 +1941,7 @@ static void ext4_da_release_space(struct inode *inode, int to_free)
 
 	spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
 
-	vfs_dq_release_reservation_block(inode, to_free);
+	dquot_release_reservation_block(inode, to_free);
 }
 
 static void ext4_da_page_release_reservation(struct page *page,
@@ -2091,6 +2118,8 @@ static void mpage_put_bnr_to_bhs(struct mpage_da_data *mpd, sector_t logical,
 				} else if (buffer_mapped(bh))
 					BUG_ON(bh->b_blocknr != pblock);
 
+				if (buffer_uninit(exbh))
+					set_buffer_uninit(bh);
 				cur_logical++;
 				pblock++;
 			} while ((bh = bh->b_this_page) != head);
@@ -2133,17 +2162,16 @@ static void ext4_da_block_invalidatepages(struct mpage_da_data *mpd,
 			break;
 		for (i = 0; i < nr_pages; i++) {
 			struct page *page = pvec.pages[i];
-			index = page->index;
-			if (index > end)
+			if (page->index > end)
 				break;
-			index++;
-
 			BUG_ON(!PageLocked(page));
 			BUG_ON(PageWriteback(page));
 			block_invalidatepage(page, 0);
 			ClearPageUptodate(page);
 			unlock_page(page);
 		}
+		index = pvec.pages[nr_pages - 1]->index + 1;
+		pagevec_release(&pvec);
 	}
 	return;
 }
@@ -2220,6 +2248,8 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
 	 */
 	new.b_state = 0;
 	get_blocks_flags = EXT4_GET_BLOCKS_CREATE;
+	if (ext4_should_dioread_nolock(mpd->inode))
+		get_blocks_flags |= EXT4_GET_BLOCKS_IO_CREATE_EXT;
 	if (mpd->b_state & (1 << BH_Delay))
 		get_blocks_flags |= EXT4_GET_BLOCKS_DELALLOC_RESERVE;
 
@@ -2630,11 +2660,14 @@ static int __ext4_journalled_writepage(struct page *page,
 		ret = err;
 
 	walk_page_buffers(handle, page_bufs, 0, len, NULL, bput_one);
-	EXT4_I(inode)->i_state |= EXT4_STATE_JDATA;
+	ext4_set_inode_state(inode, EXT4_STATE_JDATA);
 out:
 	return ret;
 }
 
+static int ext4_set_bh_endio(struct buffer_head *bh, struct inode *inode);
+static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate);
+
 /*
  * Note that we don't need to start a transaction unless we're journaling data
  * because we should have holes filled from ext4_page_mkwrite(). We even don't
@@ -2682,7 +2715,7 @@ static int ext4_writepage(struct page *page,
 	int ret = 0;
 	loff_t size;
 	unsigned int len;
-	struct buffer_head *page_bufs;
+	struct buffer_head *page_bufs = NULL;
 	struct inode *inode = page->mapping->host;
 
 	trace_ext4_writepage(inode, page);
@@ -2758,7 +2791,11 @@ static int ext4_writepage(struct page *page,
 
 	if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode))
 		ret = nobh_writepage(page, noalloc_get_block_write, wbc);
-	else
+	else if (page_bufs && buffer_uninit(page_bufs)) {
+		ext4_set_bh_endio(page_bufs, inode);
+		ret = block_write_full_page_endio(page, noalloc_get_block_write,
+					    wbc, ext4_end_io_buffer_write);
+	} else
 		ret = block_write_full_page(page, noalloc_get_block_write,
 					    wbc);
 
@@ -3301,7 +3338,8 @@ static sector_t ext4_bmap(struct address_space *mapping, sector_t block)
 		filemap_write_and_wait(mapping);
 	}
 
-	if (EXT4_JOURNAL(inode) && EXT4_I(inode)->i_state & EXT4_STATE_JDATA) {
+	if (EXT4_JOURNAL(inode) &&
+	    ext4_test_inode_state(inode, EXT4_STATE_JDATA)) {
 		/*
 		 * This is a REALLY heavyweight approach, but the use of
 		 * bmap on dirty files is expected to be extremely rare:
@@ -3320,7 +3358,7 @@ static sector_t ext4_bmap(struct address_space *mapping, sector_t block)
 		 * everything they get.
 		 */
 
-		EXT4_I(inode)->i_state &= ~EXT4_STATE_JDATA;
+		ext4_clear_inode_state(inode, EXT4_STATE_JDATA);
 		journal = EXT4_JOURNAL(inode);
 		jbd2_journal_lock_updates(journal);
 		err = jbd2_journal_flush(journal);
@@ -3345,11 +3383,45 @@ ext4_readpages(struct file *file, struct address_space *mapping,
 	return mpage_readpages(mapping, pages, nr_pages, ext4_get_block);
 }
 
+static void ext4_free_io_end(ext4_io_end_t *io)
+{
+	BUG_ON(!io);
+	if (io->page)
+		put_page(io->page);
+	iput(io->inode);
+	kfree(io);
+}
+
+static void ext4_invalidatepage_free_endio(struct page *page, unsigned long offset)
+{
+	struct buffer_head *head, *bh;
+	unsigned int curr_off = 0;
+
+	if (!page_has_buffers(page))
+		return;
+	head = bh = page_buffers(page);
+	do {
+		if (offset <= curr_off && test_clear_buffer_uninit(bh)
+					&& bh->b_private) {
+			ext4_free_io_end(bh->b_private);
+			bh->b_private = NULL;
+			bh->b_end_io = NULL;
+		}
+		curr_off = curr_off + bh->b_size;
+		bh = bh->b_this_page;
+	} while (bh != head);
+}
+
 static void ext4_invalidatepage(struct page *page, unsigned long offset)
 {
 	journal_t *journal = EXT4_JOURNAL(page->mapping->host);
 
 	/*
+	 * free any io_end structure allocated for buffers to be discarded
+	 */
+	if (ext4_should_dioread_nolock(page->mapping->host))
+		ext4_invalidatepage_free_endio(page, offset);
+	/*
 	 * If it's a full truncate we just forget about the pending dirtying
 	 */
 	if (offset == 0)
@@ -3420,7 +3492,14 @@ static ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb,
 	}
 
 retry:
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
+	if (rw == READ && ext4_should_dioread_nolock(inode))
+		ret = blockdev_direct_IO_no_locking(rw, iocb, inode,
+				 inode->i_sb->s_bdev, iov,
+				 offset, nr_segs,
+				 ext4_get_block, NULL);
+	else
+		ret = blockdev_direct_IO(rw, iocb, inode,
+				 inode->i_sb->s_bdev, iov,
 				 offset, nr_segs,
 				 ext4_get_block, NULL);
 	if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
@@ -3436,6 +3515,9 @@ retry:
 			 * but cannot extend i_size. Bail out and pretend
 			 * the write failed... */
 			ret = PTR_ERR(handle);
+			if (inode->i_nlink)
+				ext4_orphan_del(NULL, inode);
+
 			goto out;
 		}
 		if (inode->i_nlink)
@@ -3463,75 +3545,63 @@ out:
 	return ret;
 }
 
-static int ext4_get_block_dio_write(struct inode *inode, sector_t iblock,
+static int ext4_get_block_write(struct inode *inode, sector_t iblock,
 		   struct buffer_head *bh_result, int create)
 {
-	handle_t *handle = NULL;
+	handle_t *handle = ext4_journal_current_handle();
 	int ret = 0;
 	unsigned max_blocks = bh_result->b_size >> inode->i_blkbits;
 	int dio_credits;
+	int started = 0;
 
-	ext4_debug("ext4_get_block_dio_write: inode %lu, create flag %d\n",
+	ext4_debug("ext4_get_block_write: inode %lu, create flag %d\n",
 		   inode->i_ino, create);
 	/*
-	 * DIO VFS code passes create = 0 flag for write to
-	 * the middle of file. It does this to avoid block
-	 * allocation for holes, to prevent expose stale data
-	 * out when there is parallel buffered read (which does
-	 * not hold the i_mutex lock) while direct IO write has
-	 * not completed. DIO request on holes finally falls back
-	 * to buffered IO for this reason.
-	 *
-	 * For ext4 extent based file, since we support fallocate,
-	 * new allocated extent as uninitialized, for holes, we
-	 * could fallocate blocks for holes, thus parallel
-	 * buffered IO read will zero out the page when read on
-	 * a hole while parallel DIO write to the hole has not completed.
-	 *
-	 * when we come here, we know it's a direct IO write to
-	 * to the middle of file (<i_size)
-	 * so it's safe to override the create flag from VFS.
+	 * ext4_get_block in prepare for a DIO write or buffer write.
+	 * We allocate an uinitialized extent if blocks haven't been allocated.
+	 * The extent will be converted to initialized after IO complete.
 	 */
-	create = EXT4_GET_BLOCKS_DIO_CREATE_EXT;
+	create = EXT4_GET_BLOCKS_IO_CREATE_EXT;
 
-	if (max_blocks > DIO_MAX_BLOCKS)
-		max_blocks = DIO_MAX_BLOCKS;
-	dio_credits = ext4_chunk_trans_blocks(inode, max_blocks);
-	handle = ext4_journal_start(inode, dio_credits);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		goto out;
+	if (!handle) {
+		if (max_blocks > DIO_MAX_BLOCKS)
+			max_blocks = DIO_MAX_BLOCKS;
+		dio_credits = ext4_chunk_trans_blocks(inode, max_blocks);
+		handle = ext4_journal_start(inode, dio_credits);
+		if (IS_ERR(handle)) {
+			ret = PTR_ERR(handle);
+			goto out;
+		}
+		started = 1;
 	}
+
 	ret = ext4_get_blocks(handle, inode, iblock, max_blocks, bh_result,
 			      create);
 	if (ret > 0) {
 		bh_result->b_size = (ret << inode->i_blkbits);
 		ret = 0;
 	}
-	ext4_journal_stop(handle);
+	if (started)
+		ext4_journal_stop(handle);
 out:
 	return ret;
 }
 
-static void ext4_free_io_end(ext4_io_end_t *io)
-{
-	BUG_ON(!io);
-	iput(io->inode);
-	kfree(io);
-}
-static void dump_aio_dio_list(struct inode * inode)
+static void dump_completed_IO(struct inode * inode)
 {
 #ifdef	EXT4_DEBUG
 	struct list_head *cur, *before, *after;
 	ext4_io_end_t *io, *io0, *io1;
+	unsigned long flags;
 
-	if (list_empty(&EXT4_I(inode)->i_aio_dio_complete_list)){
-		ext4_debug("inode %lu aio dio list is empty\n", inode->i_ino);
+	if (list_empty(&EXT4_I(inode)->i_completed_io_list)){
+		ext4_debug("inode %lu completed_io list is empty\n", inode->i_ino);
 		return;
 	}
 
-	ext4_debug("Dump inode %lu aio_dio_completed_IO list \n", inode->i_ino);
-	list_for_each_entry(io, &EXT4_I(inode)->i_aio_dio_complete_list, list){
+	ext4_debug("Dump inode %lu completed_io list \n", inode->i_ino);
+	spin_lock_irqsave(&EXT4_I(inode)->i_completed_io_lock, flags);
+	list_for_each_entry(io, &EXT4_I(inode)->i_completed_io_list, list){
 		cur = &io->list;
 		before = cur->prev;
 		io0 = container_of(before, ext4_io_end_t, list);
@@ -3541,32 +3611,31 @@ static void dump_aio_dio_list(struct inode * inode)
 		ext4_debug("io 0x%p from inode %lu,prev 0x%p,next 0x%p\n",
 			    io, inode->i_ino, io0, io1);
 	}
+	spin_unlock_irqrestore(&EXT4_I(inode)->i_completed_io_lock, flags);
 #endif
 }
 
 /*
  * check a range of space and convert unwritten extents to written.
  */
-static int ext4_end_aio_dio_nolock(ext4_io_end_t *io)
+static int ext4_end_io_nolock(ext4_io_end_t *io)
 {
 	struct inode *inode = io->inode;
 	loff_t offset = io->offset;
-	size_t size = io->size;
+	ssize_t size = io->size;
 	int ret = 0;
 
-	ext4_debug("end_aio_dio_onlock: io 0x%p from inode %lu,list->next 0x%p,"
+	ext4_debug("ext4_end_io_nolock: io 0x%p from inode %lu,list->next 0x%p,"
 		   "list->prev 0x%p\n",
 	           io, inode->i_ino, io->list.next, io->list.prev);
 
 	if (list_empty(&io->list))
 		return ret;
 
-	if (io->flag != DIO_AIO_UNWRITTEN)
+	if (io->flag != EXT4_IO_UNWRITTEN)
 		return ret;
 
-	if (offset + size <= i_size_read(inode))
-		ret = ext4_convert_unwritten_extents(inode, offset, size);
-
+	ret = ext4_convert_unwritten_extents(inode, offset, size);
 	if (ret < 0) {
 		printk(KERN_EMERG "%s: failed to convert unwritten"
 			"extents to written extents, error is %d"
@@ -3579,50 +3648,64 @@ static int ext4_end_aio_dio_nolock(ext4_io_end_t *io)
 	io->flag = 0;
 	return ret;
 }
+
 /*
  * work on completed aio dio IO, to convert unwritten extents to extents
  */
-static void ext4_end_aio_dio_work(struct work_struct *work)
+static void ext4_end_io_work(struct work_struct *work)
 {
-	ext4_io_end_t *io  = container_of(work, ext4_io_end_t, work);
-	struct inode *inode = io->inode;
-	int ret = 0;
+	ext4_io_end_t		*io = container_of(work, ext4_io_end_t, work);
+	struct inode		*inode = io->inode;
+	struct ext4_inode_info	*ei = EXT4_I(inode);
+	unsigned long		flags;
+	int			ret;
 
 	mutex_lock(&inode->i_mutex);
-	ret = ext4_end_aio_dio_nolock(io);
-	if (ret >= 0) {
-		if (!list_empty(&io->list))
-			list_del_init(&io->list);
-		ext4_free_io_end(io);
+	ret = ext4_end_io_nolock(io);
+	if (ret < 0) {
+		mutex_unlock(&inode->i_mutex);
+		return;
 	}
+
+	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
+	if (!list_empty(&io->list))
+		list_del_init(&io->list);
+	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
 	mutex_unlock(&inode->i_mutex);
+	ext4_free_io_end(io);
 }
+
 /*
  * This function is called from ext4_sync_file().
  *
- * When AIO DIO IO is completed, the work to convert unwritten
- * extents to written is queued on workqueue but may not get immediately
+ * When IO is completed, the work to convert unwritten extents to
+ * written is queued on workqueue but may not get immediately
  * scheduled. When fsync is called, we need to ensure the
  * conversion is complete before fsync returns.
- * The inode keeps track of a list of completed AIO from DIO path
- * that might needs to do the conversion. This function walks through
- * the list and convert the related unwritten extents to written.
+ * The inode keeps track of a list of pending/completed IO that
+ * might needs to do the conversion. This function walks through
+ * the list and convert the related unwritten extents for completed IO
+ * to written.
+ * The function return the number of pending IOs on success.
  */
-int flush_aio_dio_completed_IO(struct inode *inode)
+int flush_completed_IO(struct inode *inode)
 {
 	ext4_io_end_t *io;
+	struct ext4_inode_info *ei = EXT4_I(inode);
+	unsigned long flags;
 	int ret = 0;
 	int ret2 = 0;
 
-	if (list_empty(&EXT4_I(inode)->i_aio_dio_complete_list))
+	if (list_empty(&ei->i_completed_io_list))
 		return ret;
 
-	dump_aio_dio_list(inode);
-	while (!list_empty(&EXT4_I(inode)->i_aio_dio_complete_list)){
-		io = list_entry(EXT4_I(inode)->i_aio_dio_complete_list.next,
+	dump_completed_IO(inode);
+	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
+	while (!list_empty(&ei->i_completed_io_list)){
+		io = list_entry(ei->i_completed_io_list.next,
 				ext4_io_end_t, list);
 		/*
-		 * Calling ext4_end_aio_dio_nolock() to convert completed
+		 * Calling ext4_end_io_nolock() to convert completed
 		 * IO to written.
 		 *
 		 * When ext4_sync_file() is called, run_queue() may already
@@ -3635,20 +3718,23 @@ int flush_aio_dio_completed_IO(struct inode *inode)
 		 * avoid double converting from both fsync and background work
 		 * queue work.
 		 */
-		ret = ext4_end_aio_dio_nolock(io);
+		spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
+		ret = ext4_end_io_nolock(io);
+		spin_lock_irqsave(&ei->i_completed_io_lock, flags);
 		if (ret < 0)
 			ret2 = ret;
 		else
 			list_del_init(&io->list);
 	}
+	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
 	return (ret2 < 0) ? ret2 : 0;
 }
 
-static ext4_io_end_t *ext4_init_io_end (struct inode *inode)
+static ext4_io_end_t *ext4_init_io_end (struct inode *inode, gfp_t flags)
 {
 	ext4_io_end_t *io = NULL;
 
-	io = kmalloc(sizeof(*io), GFP_NOFS);
+	io = kmalloc(sizeof(*io), flags);
 
 	if (io) {
 		igrab(inode);
@@ -3656,8 +3742,8 @@ static ext4_io_end_t *ext4_init_io_end (struct inode *inode)
 		io->flag = 0;
 		io->offset = 0;
 		io->size = 0;
-		io->error = 0;
-		INIT_WORK(&io->work, ext4_end_aio_dio_work);
+		io->page = NULL;
+		INIT_WORK(&io->work, ext4_end_io_work);
 		INIT_LIST_HEAD(&io->list);
 	}
 
@@ -3669,6 +3755,8 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 {
         ext4_io_end_t *io_end = iocb->private;
 	struct workqueue_struct *wq;
+	unsigned long flags;
+	struct ext4_inode_info *ei;
 
 	/* if not async direct IO or dio with 0 bytes write, just return */
 	if (!io_end || !size)
@@ -3680,7 +3768,7 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 		  size);
 
 	/* if not aio dio with unwritten extents, just free io and return */
-	if (io_end->flag != DIO_AIO_UNWRITTEN){
+	if (io_end->flag != EXT4_IO_UNWRITTEN){
 		ext4_free_io_end(io_end);
 		iocb->private = NULL;
 		return;
@@ -3688,16 +3776,85 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 
 	io_end->offset = offset;
 	io_end->size = size;
+	io_end->flag = EXT4_IO_UNWRITTEN;
 	wq = EXT4_SB(io_end->inode->i_sb)->dio_unwritten_wq;
 
 	/* queue the work to convert unwritten extents to written */
 	queue_work(wq, &io_end->work);
 
 	/* Add the io_end to per-inode completed aio dio list*/
-	list_add_tail(&io_end->list,
-		 &EXT4_I(io_end->inode)->i_aio_dio_complete_list);
+	ei = EXT4_I(io_end->inode);
+	spin_lock_irqsave(&ei->i_completed_io_lock, flags);
+	list_add_tail(&io_end->list, &ei->i_completed_io_list);
+	spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
 	iocb->private = NULL;
 }
+
+static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate)
+{
+	ext4_io_end_t *io_end = bh->b_private;
+	struct workqueue_struct *wq;
+	struct inode *inode;
+	unsigned long flags;
+
+	if (!test_clear_buffer_uninit(bh) || !io_end)
+		goto out;
+
+	if (!(io_end->inode->i_sb->s_flags & MS_ACTIVE)) {
+		printk("sb umounted, discard end_io request for inode %lu\n",
+			io_end->inode->i_ino);
+		ext4_free_io_end(io_end);
+		goto out;
+	}
+
+	io_end->flag = EXT4_IO_UNWRITTEN;
+	inode = io_end->inode;
+
+	/* Add the io_end to per-inode completed io list*/
+	spin_lock_irqsave(&EXT4_I(inode)->i_completed_io_lock, flags);
+	list_add_tail(&io_end->list, &EXT4_I(inode)->i_completed_io_list);
+	spin_unlock_irqrestore(&EXT4_I(inode)->i_completed_io_lock, flags);
+
+	wq = EXT4_SB(inode->i_sb)->dio_unwritten_wq;
+	/* queue the work to convert unwritten extents to written */
+	queue_work(wq, &io_end->work);
+out:
+	bh->b_private = NULL;
+	bh->b_end_io = NULL;
+	clear_buffer_uninit(bh);
+	end_buffer_async_write(bh, uptodate);
+}
+
+static int ext4_set_bh_endio(struct buffer_head *bh, struct inode *inode)
+{
+	ext4_io_end_t *io_end;
+	struct page *page = bh->b_page;
+	loff_t offset = (sector_t)page->index << PAGE_CACHE_SHIFT;
+	size_t size = bh->b_size;
+
+retry:
+	io_end = ext4_init_io_end(inode, GFP_ATOMIC);
+	if (!io_end) {
+		if (printk_ratelimit())
+			printk(KERN_WARNING "%s: allocation fail\n", __func__);
+		schedule();
+		goto retry;
+	}
+	io_end->offset = offset;
+	io_end->size = size;
+	/*
+	 * We need to hold a reference to the page to make sure it
+	 * doesn't get evicted before ext4_end_io_work() has a chance
+	 * to convert the extent from written to unwritten.
+	 */
+	io_end->page = page;
+	get_page(io_end->page);
+
+	bh->b_private = io_end;
+	bh->b_end_io = ext4_end_io_buffer_write;
+	return 0;
+}
+
 /*
  * For ext4 extent files, ext4 will do direct-io write to holes,
  * preallocated extents, and those write extend the file, no need to
@@ -3751,7 +3908,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 		iocb->private = NULL;
 		EXT4_I(inode)->cur_aio_dio = NULL;
 		if (!is_sync_kiocb(iocb)) {
-			iocb->private = ext4_init_io_end(inode);
+			iocb->private = ext4_init_io_end(inode, GFP_NOFS);
 			if (!iocb->private)
 				return -ENOMEM;
 			/*
@@ -3767,7 +3924,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 		ret = blockdev_direct_IO(rw, iocb, inode,
 					 inode->i_sb->s_bdev, iov,
 					 offset, nr_segs,
-					 ext4_get_block_dio_write,
+					 ext4_get_block_write,
 					 ext4_end_io_dio);
 		if (iocb->private)
 			EXT4_I(inode)->cur_aio_dio = NULL;
@@ -3788,8 +3945,8 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 		if (ret != -EIOCBQUEUED && ret <= 0 && iocb->private) {
 			ext4_free_io_end(iocb->private);
 			iocb->private = NULL;
-		} else if (ret > 0 && (EXT4_I(inode)->i_state &
-				       EXT4_STATE_DIO_UNWRITTEN)) {
+		} else if (ret > 0 && ext4_test_inode_state(inode,
+						EXT4_STATE_DIO_UNWRITTEN)) {
 			int err;
 			/*
 			 * for non AIO case, since the IO is already
@@ -3799,7 +3956,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 							     offset, ret);
 			if (err < 0)
 				ret = err;
-			EXT4_I(inode)->i_state &= ~EXT4_STATE_DIO_UNWRITTEN;
+			ext4_clear_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
 		}
 		return ret;
 	}
@@ -4130,18 +4287,27 @@ no_top:
  * We release `count' blocks on disk, but (last - first) may be greater
  * than `count' because there can be holes in there.
  */
-static void ext4_clear_blocks(handle_t *handle, struct inode *inode,
-			      struct buffer_head *bh,
-			      ext4_fsblk_t block_to_free,
-			      unsigned long count, __le32 *first,
-			      __le32 *last)
+static int ext4_clear_blocks(handle_t *handle, struct inode *inode,
+			     struct buffer_head *bh,
+			     ext4_fsblk_t block_to_free,
+			     unsigned long count, __le32 *first,
+			     __le32 *last)
 {
 	__le32 *p;
-	int	flags = EXT4_FREE_BLOCKS_FORGET;
+	int	flags = EXT4_FREE_BLOCKS_FORGET | EXT4_FREE_BLOCKS_VALIDATED;
 
 	if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
 		flags |= EXT4_FREE_BLOCKS_METADATA;
 
+	if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), block_to_free,
+				   count)) {
+		ext4_error(inode->i_sb, "inode #%lu: "
+			   "attempt to clear blocks %llu len %lu, invalid",
+			   inode->i_ino, (unsigned long long) block_to_free,
+			   count);
+		return 1;
+	}
+
 	if (try_to_extend_transaction(handle, inode)) {
 		if (bh) {
 			BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
@@ -4160,6 +4326,7 @@ static void ext4_clear_blocks(handle_t *handle, struct inode *inode,
 		*p = 0;
 
 	ext4_free_blocks(handle, inode, 0, block_to_free, count, flags);
+	return 0;
 }
 
 /**
@@ -4215,9 +4382,10 @@ static void ext4_free_data(handle_t *handle, struct inode *inode,
 			} else if (nr == block_to_free + count) {
 				count++;
 			} else {
-				ext4_clear_blocks(handle, inode, this_bh,
-						  block_to_free,
-						  count, block_to_free_p, p);
+				if (ext4_clear_blocks(handle, inode, this_bh,
+						      block_to_free, count,
+						      block_to_free_p, p))
+					break;
 				block_to_free = nr;
 				block_to_free_p = p;
 				count = 1;
@@ -4241,7 +4409,7 @@ static void ext4_free_data(handle_t *handle, struct inode *inode,
 		if ((EXT4_JOURNAL(inode) == NULL) || bh2jh(this_bh))
 			ext4_handle_dirty_metadata(handle, inode, this_bh);
 		else
-			ext4_error(inode->i_sb, __func__,
+			ext4_error(inode->i_sb,
 				   "circular indirect block detected, "
 				   "inode=%lu, block=%llu",
 				   inode->i_ino,
@@ -4281,6 +4449,16 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode,
 			if (!nr)
 				continue;		/* A hole */
 
+			if (!ext4_data_block_valid(EXT4_SB(inode->i_sb),
+						   nr, 1)) {
+				ext4_error(inode->i_sb,
+					   "indirect mapped block in inode "
+					   "#%lu invalid (level %d, blk #%lu)",
+					   inode->i_ino, depth,
+					   (unsigned long) nr);
+				break;
+			}
+
 			/* Go read the buffer for the next level down */
 			bh = sb_bread(inode->i_sb, nr);
 
@@ -4289,7 +4467,7 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode,
 			 * (should be rare).
 			 */
 			if (!bh) {
-				ext4_error(inode->i_sb, "ext4_free_branches",
+				ext4_error(inode->i_sb,
 					   "Read failure, inode=%lu, block=%llu",
 					   inode->i_ino, nr);
 				continue;
@@ -4433,8 +4611,10 @@ void ext4_truncate(struct inode *inode)
 	if (!ext4_can_truncate(inode))
 		return;
 
+	EXT4_I(inode)->i_flags &= ~EXT4_EOFBLOCKS_FL;
+
 	if (inode->i_size == 0 && !test_opt(inode->i_sb, NO_AUTO_DA_ALLOC))
-		ei->i_state |= EXT4_STATE_DA_ALLOC_CLOSE;
+		ext4_set_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE);
 
 	if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) {
 		ext4_ext_truncate(inode);
@@ -4604,9 +4784,8 @@ static int __ext4_get_inode_loc(struct inode *inode,
 
 	bh = sb_getblk(sb, block);
 	if (!bh) {
-		ext4_error(sb, "ext4_get_inode_loc", "unable to read "
-			   "inode block - inode=%lu, block=%llu",
-			   inode->i_ino, block);
+		ext4_error(sb, "unable to read inode block - "
+			   "inode=%lu, block=%llu", inode->i_ino, block);
 		return -EIO;
 	}
 	if (!buffer_uptodate(bh)) {
@@ -4704,9 +4883,8 @@ make_io:
 		submit_bh(READ_META, bh);
 		wait_on_buffer(bh);
 		if (!buffer_uptodate(bh)) {
-			ext4_error(sb, __func__,
-				   "unable to read inode block - inode=%lu, "
-				   "block=%llu", inode->i_ino, block);
+			ext4_error(sb, "unable to read inode block - inode=%lu,"
+				   " block=%llu", inode->i_ino, block);
 			brelse(bh);
 			return -EIO;
 		}
@@ -4720,7 +4898,7 @@ int ext4_get_inode_loc(struct inode *inode, struct ext4_iloc *iloc)
 {
 	/* We have all inode data except xattrs in memory here. */
 	return __ext4_get_inode_loc(inode, iloc,
-		!(EXT4_I(inode)->i_state & EXT4_STATE_XATTR));
+		!ext4_test_inode_state(inode, EXT4_STATE_XATTR));
 }
 
 void ext4_set_inode_flags(struct inode *inode)
@@ -4814,7 +4992,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	}
 	inode->i_nlink = le16_to_cpu(raw_inode->i_links_count);
 
-	ei->i_state = 0;
+	ei->i_state_flags = 0;
 	ei->i_dir_start_lookup = 0;
 	ei->i_dtime = le32_to_cpu(raw_inode->i_dtime);
 	/* We now have enough fields to check if the inode was active or not.
@@ -4897,7 +5075,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 					EXT4_GOOD_OLD_INODE_SIZE +
 					ei->i_extra_isize;
 			if (*magic == cpu_to_le32(EXT4_XATTR_MAGIC))
-				ei->i_state |= EXT4_STATE_XATTR;
+				ext4_set_inode_state(inode, EXT4_STATE_XATTR);
 		}
 	} else
 		ei->i_extra_isize = 0;
@@ -4917,8 +5095,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	ret = 0;
 	if (ei->i_file_acl &&
 	    !ext4_data_block_valid(EXT4_SB(sb), ei->i_file_acl, 1)) {
-		ext4_error(sb, __func__,
-			   "bad extended attribute block %llu in inode #%lu",
+		ext4_error(sb, "bad extended attribute block %llu inode #%lu",
 			   ei->i_file_acl, inode->i_ino);
 		ret = -EIO;
 		goto bad_inode;
@@ -4964,8 +5141,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 			   new_decode_dev(le32_to_cpu(raw_inode->i_block[1])));
 	} else {
 		ret = -EIO;
-		ext4_error(inode->i_sb, __func__,
-			   "bogus i_mode (%o) for inode=%lu",
+		ext4_error(inode->i_sb, "bogus i_mode (%o) for inode=%lu",
 			   inode->i_mode, inode->i_ino);
 		goto bad_inode;
 	}
@@ -5037,7 +5213,7 @@ static int ext4_do_update_inode(handle_t *handle,
 
 	/* For fields not not tracking in the in-memory inode,
 	 * initialise them to zero for new inodes. */
-	if (ei->i_state & EXT4_STATE_NEW)
+	if (ext4_test_inode_state(inode, EXT4_STATE_NEW))
 		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
 
 	ext4_get_inode_flags(ei);
@@ -5101,7 +5277,7 @@ static int ext4_do_update_inode(handle_t *handle,
 					EXT4_FEATURE_RO_COMPAT_LARGE_FILE);
 			sb->s_dirt = 1;
 			ext4_handle_sync(handle);
-			err = ext4_handle_dirty_metadata(handle, inode,
+			err = ext4_handle_dirty_metadata(handle, NULL,
 					EXT4_SB(sb)->s_sbh);
 		}
 	}
@@ -5130,10 +5306,10 @@ static int ext4_do_update_inode(handle_t *handle,
 	}
 
 	BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata");
-	rc = ext4_handle_dirty_metadata(handle, inode, bh);
+	rc = ext4_handle_dirty_metadata(handle, NULL, bh);
 	if (!err)
 		err = rc;
-	ei->i_state &= ~EXT4_STATE_NEW;
+	ext4_clear_inode_state(inode, EXT4_STATE_NEW);
 
 	ext4_update_inode_fsync_trans(handle, inode, 0);
 out_brelse:
@@ -5177,7 +5353,7 @@ out_brelse:
  * `stuff()' is running, and the new i_size will be lost.  Plus the inode
  * will no longer be on the superblock's dirty inode list.
  */
-int ext4_write_inode(struct inode *inode, int wait)
+int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	int err;
 
@@ -5191,7 +5367,7 @@ int ext4_write_inode(struct inode *inode, int wait)
 			return -EIO;
 		}
 
-		if (!wait)
+		if (wbc->sync_mode != WB_SYNC_ALL)
 			return 0;
 
 		err = ext4_force_commit(inode->i_sb);
@@ -5201,13 +5377,11 @@ int ext4_write_inode(struct inode *inode, int wait)
 		err = ext4_get_inode_loc(inode, &iloc);
 		if (err)
 			return err;
-		if (wait)
+		if (wbc->sync_mode == WB_SYNC_ALL)
 			sync_dirty_buffer(iloc.bh);
 		if (buffer_req(iloc.bh) && !buffer_uptodate(iloc.bh)) {
-			ext4_error(inode->i_sb, __func__,
-				   "IO error syncing inode, "
-				   "inode=%lu, block=%llu",
-				   inode->i_ino,
+			ext4_error(inode->i_sb, "IO error syncing inode, "
+				   "inode=%lu, block=%llu", inode->i_ino,
 				   (unsigned long long)iloc.bh->b_blocknr);
 			err = -EIO;
 		}
@@ -5249,6 +5423,8 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	if (error)
 		return error;
 
+	if (ia_valid & ATTR_SIZE)
+		dquot_initialize(inode);
 	if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
 		(ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) {
 		handle_t *handle;
@@ -5261,7 +5437,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 			error = PTR_ERR(handle);
 			goto err_out;
 		}
-		error = vfs_dq_transfer(inode, attr) ? -EDQUOT : 0;
+		error = dquot_transfer(inode, attr);
 		if (error) {
 			ext4_journal_stop(handle);
 			return error;
@@ -5288,7 +5464,9 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (S_ISREG(inode->i_mode) &&
-	    attr->ia_valid & ATTR_SIZE && attr->ia_size < inode->i_size) {
+	    attr->ia_valid & ATTR_SIZE &&
+	    (attr->ia_size < inode->i_size ||
+	     (EXT4_I(inode)->i_flags & EXT4_EOFBLOCKS_FL))) {
 		handle_t *handle;
 
 		handle = ext4_journal_start(inode, 3);
@@ -5319,6 +5497,9 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 				goto err_out;
 			}
 		}
+		/* ext4_truncate will clear the flag */
+		if ((EXT4_I(inode)->i_flags & EXT4_EOFBLOCKS_FL))
+			ext4_truncate(inode);
 	}
 
 	rc = inode_setattr(inode, attr);
@@ -5557,8 +5738,8 @@ static int ext4_expand_extra_isize(struct inode *inode,
 	entry = IFIRST(header);
 
 	/* No extended attributes present */
-	if (!(EXT4_I(inode)->i_state & EXT4_STATE_XATTR) ||
-		header->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC)) {
+	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR) ||
+	    header->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC)) {
 		memset((void *)raw_inode + EXT4_GOOD_OLD_INODE_SIZE, 0,
 			new_extra_isize);
 		EXT4_I(inode)->i_extra_isize = new_extra_isize;
@@ -5602,7 +5783,7 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode)
 	err = ext4_reserve_inode_write(handle, inode, &iloc);
 	if (ext4_handle_valid(handle) &&
 	    EXT4_I(inode)->i_extra_isize < sbi->s_want_extra_isize &&
-	    !(EXT4_I(inode)->i_state & EXT4_STATE_NO_EXPAND)) {
+	    !ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND)) {
 		/*
 		 * We need extra buffer credits since we may write into EA block
 		 * with this same handle. If journal_extend fails, then it will
@@ -5616,10 +5797,11 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode)
 						      sbi->s_want_extra_isize,
 						      iloc, handle);
 			if (ret) {
-				EXT4_I(inode)->i_state |= EXT4_STATE_NO_EXPAND;
+				ext4_set_inode_state(inode,
+						     EXT4_STATE_NO_EXPAND);
 				if (mnt_count !=
 					le16_to_cpu(sbi->s_es->s_mnt_count)) {
-					ext4_warning(inode->i_sb, __func__,
+					ext4_warning(inode->i_sb,
 					"Unable to expand inode %lu. Delete"
 					" some EAs or run e2fsck.",
 					inode->i_ino);
@@ -5641,7 +5823,7 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode)
  * i_size has been changed by generic_commit_write() and we thus need
  * to include the updated inode in the current transaction.
  *
- * Also, vfs_dq_alloc_block() will always dirty the inode when blocks
+ * Also, dquot_alloc_block() will always dirty the inode when blocks
  * are allocated to the file.
  *
  * If the inode is marked synchronous, we don't honour that here - doing
@@ -5683,7 +5865,7 @@ static int ext4_pin_inode(handle_t *handle, struct inode *inode)
 			err = jbd2_journal_get_write_access(handle, iloc.bh);
 			if (!err)
 				err = ext4_handle_dirty_metadata(handle,
-								 inode,
+								 NULL,
 								 iloc.bh);
 			brelse(iloc.bh);
 		}
diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index b63d193126db..016d0249294f 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -92,6 +92,15 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 			flags &= ~EXT4_EXTENTS_FL;
 		}
 
+		if (flags & EXT4_EOFBLOCKS_FL) {
+			/* we don't support adding EOFBLOCKS flag */
+			if (!(oldflags & EXT4_EOFBLOCKS_FL)) {
+				err = -EOPNOTSUPP;
+				goto flags_out;
+			}
+		} else if (oldflags & EXT4_EOFBLOCKS_FL)
+			ext4_truncate(inode);
+
 		handle = ext4_journal_start(inode, 1);
 		if (IS_ERR(handle)) {
 			err = PTR_ERR(handle);
@@ -249,7 +258,8 @@ setversion_out:
 		if (me.moved_len > 0)
 			file_remove_suid(donor_filp);
 
-		if (copy_to_user((struct move_extent *)arg, &me, sizeof(me)))
+		if (copy_to_user((struct move_extent __user *)arg, 
+				 &me, sizeof(me)))
 			err = -EFAULT;
 mext_out:
 		fput(donor_filp);
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index d34afad3e137..506713a2ebd8 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -441,10 +441,9 @@ static void mb_free_blocks_double(struct inode *inode, struct ext4_buddy *e4b,
 	for (i = 0; i < count; i++) {
 		if (!mb_test_bit(first + i, e4b->bd_info->bb_bitmap)) {
 			ext4_fsblk_t blocknr;
-			blocknr = e4b->bd_group * EXT4_BLOCKS_PER_GROUP(sb);
+
+			blocknr = ext4_group_first_block_no(sb, e4b->bd_group);
 			blocknr += first + i;
-			blocknr +=
-			    le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block);
 			ext4_grp_locked_error(sb, e4b->bd_group,
 				   __func__, "double-free of inode"
 				   " %lu's block %llu(bit %u in group %u)",
@@ -1255,10 +1254,9 @@ static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b,
 
 		if (!mb_test_bit(block, EXT4_MB_BITMAP(e4b))) {
 			ext4_fsblk_t blocknr;
-			blocknr = e4b->bd_group * EXT4_BLOCKS_PER_GROUP(sb);
+
+			blocknr = ext4_group_first_block_no(sb, e4b->bd_group);
 			blocknr += block;
-			blocknr +=
-			    le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block);
 			ext4_grp_locked_error(sb, e4b->bd_group,
 				   __func__, "double-free of inode"
 				   " %lu's block %llu(bit %u in group %u)",
@@ -1631,7 +1629,6 @@ int ext4_mb_find_by_goal(struct ext4_allocation_context *ac,
 	int max;
 	int err;
 	struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb);
-	struct ext4_super_block *es = sbi->s_es;
 	struct ext4_free_extent ex;
 
 	if (!(ac->ac_flags & EXT4_MB_HINT_TRY_GOAL))
@@ -1648,8 +1645,8 @@ int ext4_mb_find_by_goal(struct ext4_allocation_context *ac,
 	if (max >= ac->ac_g_ex.fe_len && ac->ac_g_ex.fe_len == sbi->s_stripe) {
 		ext4_fsblk_t start;
 
-		start = (e4b->bd_group * EXT4_BLOCKS_PER_GROUP(ac->ac_sb)) +
-			ex.fe_start + le32_to_cpu(es->s_first_data_block);
+		start = ext4_group_first_block_no(ac->ac_sb, e4b->bd_group) +
+			ex.fe_start;
 		/* use do_div to get remainder (would be 64-bit modulo) */
 		if (do_div(start, sbi->s_stripe) == 0) {
 			ac->ac_found++;
@@ -1803,8 +1800,8 @@ void ext4_mb_scan_aligned(struct ext4_allocation_context *ac,
 	BUG_ON(sbi->s_stripe == 0);
 
 	/* find first stripe-aligned block in group */
-	first_group_block = e4b->bd_group * EXT4_BLOCKS_PER_GROUP(sb)
-		+ le32_to_cpu(sbi->s_es->s_first_data_block);
+	first_group_block = ext4_group_first_block_no(sb, e4b->bd_group);
+
 	a = first_group_block + sbi->s_stripe - 1;
 	do_div(a, sbi->s_stripe);
 	i = (a * sbi->s_stripe) - first_group_block;
@@ -2256,7 +2253,7 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
 
 	INIT_LIST_HEAD(&meta_group_info[i]->bb_prealloc_list);
 	init_rwsem(&meta_group_info[i]->alloc_sem);
-	meta_group_info[i]->bb_free_root.rb_node = NULL;
+	meta_group_info[i]->bb_free_root = RB_ROOT;
 
 #ifdef DOUBLE_CHECK
 	{
@@ -2560,12 +2557,9 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn)
 		ext4_unlock_group(sb, entry->group);
 		if (test_opt(sb, DISCARD)) {
 			ext4_fsblk_t discard_block;
-			struct ext4_super_block *es = EXT4_SB(sb)->s_es;
 
-			discard_block = (ext4_fsblk_t)entry->group *
-						EXT4_BLOCKS_PER_GROUP(sb)
-					+ entry->start_blk
-					+ le32_to_cpu(es->s_first_data_block);
+			discard_block = entry->start_blk +
+				ext4_group_first_block_no(sb, entry->group);
 			trace_ext4_discard_blocks(sb,
 					(unsigned long long)discard_block,
 					entry->count);
@@ -2703,14 +2697,11 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac,
 	if (err)
 		goto out_err;
 
-	block = ac->ac_b_ex.fe_group * EXT4_BLOCKS_PER_GROUP(sb)
-		+ ac->ac_b_ex.fe_start
-		+ le32_to_cpu(es->s_first_data_block);
+	block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex);
 
 	len = ac->ac_b_ex.fe_len;
 	if (!ext4_data_block_valid(sbi, block, len)) {
-		ext4_error(sb, __func__,
-			   "Allocating blocks %llu-%llu which overlap "
+		ext4_error(sb, "Allocating blocks %llu-%llu which overlap "
 			   "fs metadata\n", block, block+len);
 		/* File system mounted not to panic on error
 		 * Fix the bitmap and repeat the block allocation
@@ -3161,9 +3152,7 @@ ext4_mb_use_preallocated(struct ext4_allocation_context *ac)
 		/* The max size of hash table is PREALLOC_TB_SIZE */
 		order = PREALLOC_TB_SIZE - 1;
 
-	goal_block = ac->ac_g_ex.fe_group * EXT4_BLOCKS_PER_GROUP(ac->ac_sb) +
-		     ac->ac_g_ex.fe_start +
-		     le32_to_cpu(EXT4_SB(ac->ac_sb)->s_es->s_first_data_block);
+	goal_block = ext4_grp_offs_to_block(ac->ac_sb, &ac->ac_g_ex);
 	/*
 	 * search for the prealloc space that is having
 	 * minimal distance from the goal block.
@@ -3526,8 +3515,7 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
 		if (bit >= end)
 			break;
 		next = mb_find_next_bit(bitmap_bh->b_data, end, bit);
-		start = group * EXT4_BLOCKS_PER_GROUP(sb) + bit +
-				le32_to_cpu(sbi->s_es->s_first_data_block);
+		start = ext4_group_first_block_no(sb, group) + bit;
 		mb_debug(1, "    free preallocated %u/%u in group %u\n",
 				(unsigned) start, (unsigned) next - bit,
 				(unsigned) group);
@@ -3623,15 +3611,13 @@ ext4_mb_discard_group_preallocations(struct super_block *sb,
 
 	bitmap_bh = ext4_read_block_bitmap(sb, group);
 	if (bitmap_bh == NULL) {
-		ext4_error(sb, __func__, "Error in reading block "
-				"bitmap for %u", group);
+		ext4_error(sb, "Error reading block bitmap for %u", group);
 		return 0;
 	}
 
 	err = ext4_mb_load_buddy(sb, group, &e4b);
 	if (err) {
-		ext4_error(sb, __func__, "Error in loading buddy "
-				"information for %u", group);
+		ext4_error(sb, "Error loading buddy information for %u", group);
 		put_bh(bitmap_bh);
 		return 0;
 	}
@@ -3804,15 +3790,15 @@ repeat:
 
 		err = ext4_mb_load_buddy(sb, group, &e4b);
 		if (err) {
-			ext4_error(sb, __func__, "Error in loading buddy "
-					"information for %u", group);
+			ext4_error(sb, "Error loading buddy information for %u",
+					group);
 			continue;
 		}
 
 		bitmap_bh = ext4_read_block_bitmap(sb, group);
 		if (bitmap_bh == NULL) {
-			ext4_error(sb, __func__, "Error in reading block "
-					"bitmap for %u", group);
+			ext4_error(sb, "Error reading block bitmap for %u",
+					group);
 			ext4_mb_release_desc(&e4b);
 			continue;
 		}
@@ -3938,7 +3924,7 @@ static void ext4_mb_group_or_file(struct ext4_allocation_context *ac)
 
 	/* don't use group allocation for large files */
 	size = max(size, isize);
-	if (size >= sbi->s_mb_stream_request) {
+	if (size > sbi->s_mb_stream_request) {
 		ac->ac_flags |= EXT4_MB_STREAM_ALLOC;
 		return;
 	}
@@ -4077,8 +4063,8 @@ ext4_mb_discard_lg_preallocations(struct super_block *sb,
 
 		ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, NULL);
 		if (ext4_mb_load_buddy(sb, group, &e4b)) {
-			ext4_error(sb, __func__, "Error in loading buddy "
-					"information for %u", group);
+			ext4_error(sb, "Error loading buddy information for %u",
+					group);
 			continue;
 		}
 		ext4_lock_group(sb, group);
@@ -4254,7 +4240,7 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle,
 			return 0;
 		}
 		reserv_blks = ar->len;
-		while (ar->len && vfs_dq_alloc_block(ar->inode, ar->len)) {
+		while (ar->len && dquot_alloc_block(ar->inode, ar->len)) {
 			ar->flags |= EXT4_MB_HINT_NOPREALLOC;
 			ar->len--;
 		}
@@ -4331,7 +4317,7 @@ out2:
 	kmem_cache_free(ext4_ac_cachep, ac);
 out1:
 	if (inquota && ar->len < inquota)
-		vfs_dq_free_block(ar->inode, inquota - ar->len);
+		dquot_free_block(ar->inode, inquota - ar->len);
 out3:
 	if (!ar->len) {
 		if (!EXT4_I(ar->inode)->i_delalloc_reserved_flag)
@@ -4476,10 +4462,10 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode,
 
 	sbi = EXT4_SB(sb);
 	es = EXT4_SB(sb)->s_es;
-	if (!ext4_data_block_valid(sbi, block, count)) {
-		ext4_error(sb, __func__,
-			    "Freeing blocks not in datazone - "
-			    "block = %llu, count = %lu", block, count);
+	if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) &&
+	    !ext4_data_block_valid(sbi, block, count)) {
+		ext4_error(sb, "Freeing blocks not in datazone - "
+			   "block = %llu, count = %lu", block, count);
 		goto error_return;
 	}
 
@@ -4547,8 +4533,7 @@ do_more:
 	    in_range(block + count - 1, ext4_inode_table(sb, gdp),
 		      EXT4_SB(sb)->s_itb_per_group)) {
 
-		ext4_error(sb, __func__,
-			   "Freeing blocks in system zone - "
+		ext4_error(sb, "Freeing blocks in system zone - "
 			   "Block = %llu, count = %lu", block, count);
 		/* err = 0. ext4_std_error should be a no op */
 		goto error_return;
@@ -4646,7 +4631,7 @@ do_more:
 	sb->s_dirt = 1;
 error_return:
 	if (freed)
-		vfs_dq_free_block(inode, freed);
+		dquot_free_block(inode, freed);
 	brelse(bitmap_bh);
 	ext4_std_error(sb, err);
 	if (ac)
diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h
index 436521cae456..b619322c76f0 100644
--- a/fs/ext4/mballoc.h
+++ b/fs/ext4/mballoc.h
@@ -220,16 +220,9 @@ struct ext4_buddy {
 #define EXT4_MB_BITMAP(e4b)	((e4b)->bd_bitmap)
 #define EXT4_MB_BUDDY(e4b)	((e4b)->bd_buddy)
 
-#define in_range(b, first, len)	((b) >= (first) && (b) <= (first) + (len) - 1)
-
 static inline ext4_fsblk_t ext4_grp_offs_to_block(struct super_block *sb,
 					struct ext4_free_extent *fex)
 {
-	ext4_fsblk_t block;
-
-	block = (ext4_fsblk_t) fex->fe_group * EXT4_BLOCKS_PER_GROUP(sb)
-			+ fex->fe_start
-			+ le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block);
-	return block;
+	return ext4_group_first_block_no(sb, fex->fe_group) + fex->fe_start;
 }
 #endif
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index 81415814b00b..8b87bd0eac95 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -365,12 +365,12 @@ static int ext4_ext_swap_inode_data(handle_t *handle, struct inode *inode,
 	 * happened after we started the migrate. We need to
 	 * fail the migrate
 	 */
-	if (!(EXT4_I(inode)->i_state & EXT4_STATE_EXT_MIGRATE)) {
+	if (!ext4_test_inode_state(inode, EXT4_STATE_EXT_MIGRATE)) {
 		retval = -EAGAIN;
 		up_write(&EXT4_I(inode)->i_data_sem);
 		goto err_out;
 	} else
-		EXT4_I(inode)->i_state &= ~EXT4_STATE_EXT_MIGRATE;
+		ext4_clear_inode_state(inode, EXT4_STATE_EXT_MIGRATE);
 	/*
 	 * We have the extent map build with the tmp inode.
 	 * Now copy the i_data across
@@ -503,14 +503,10 @@ int ext4_ext_migrate(struct inode *inode)
 	}
 	i_size_write(tmp_inode, i_size_read(inode));
 	/*
-	 * We don't want the inode to be reclaimed
-	 * if we got interrupted in between. We have
-	 * this tmp inode carrying reference to the
-	 * data blocks of the original file. We set
-	 * the i_nlink to zero at the last stage after
-	 * switching the original file to extent format
+	 * Set the i_nlink to zero so it will be deleted later
+	 * when we drop inode reference.
 	 */
-	tmp_inode->i_nlink = 1;
+	tmp_inode->i_nlink = 0;
 
 	ext4_ext_tree_init(handle, tmp_inode);
 	ext4_orphan_add(handle, tmp_inode);
@@ -533,10 +529,20 @@ int ext4_ext_migrate(struct inode *inode)
 	 * allocation.
 	 */
 	down_read((&EXT4_I(inode)->i_data_sem));
-	EXT4_I(inode)->i_state |= EXT4_STATE_EXT_MIGRATE;
+	ext4_set_inode_state(inode, EXT4_STATE_EXT_MIGRATE);
 	up_read((&EXT4_I(inode)->i_data_sem));
 
 	handle = ext4_journal_start(inode, 1);
+	if (IS_ERR(handle)) {
+		/*
+		 * It is impossible to update on-disk structures without
+		 * a handle, so just rollback in-core changes and live other
+		 * work to orphan_list_cleanup()
+		 */
+		ext4_orphan_del(NULL, tmp_inode);
+		retval = PTR_ERR(handle);
+		goto out;
+	}
 
 	ei = EXT4_I(inode);
 	i_data = ei->i_data;
@@ -618,15 +624,8 @@ err_out:
 
 	/* Reset the extent details */
 	ext4_ext_tree_init(handle, tmp_inode);
-
-	/*
-	 * Set the i_nlink to zero so that
-	 * generic_drop_inode really deletes the
-	 * inode
-	 */
-	tmp_inode->i_nlink = 0;
-
 	ext4_journal_stop(handle);
+out:
 	unlock_new_inode(tmp_inode);
 	iput(tmp_inode);
 
diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index 82c415be87a4..aa5fe28d180f 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -152,12 +152,12 @@ mext_check_null_inode(struct inode *inode1, struct inode *inode2,
 	int ret = 0;
 
 	if (inode1 == NULL) {
-		ext4_error(inode2->i_sb, function,
+		__ext4_error(inode2->i_sb, function,
 			"Both inodes should not be NULL: "
 			"inode1 NULL inode2 %lu", inode2->i_ino);
 		ret = -EIO;
 	} else if (inode2 == NULL) {
-		ext4_error(inode1->i_sb, function,
+		__ext4_error(inode1->i_sb, function,
 			"Both inodes should not be NULL: "
 			"inode1 %lu inode2 NULL", inode1->i_ino);
 		ret = -EIO;
@@ -252,6 +252,7 @@ mext_insert_across_blocks(handle_t *handle, struct inode *orig_inode,
 		}
 
 		o_start->ee_len = start_ext->ee_len;
+		eblock = le32_to_cpu(start_ext->ee_block);
 		new_flag = 1;
 
 	} else if (start_ext->ee_len && new_ext->ee_len &&
@@ -262,6 +263,7 @@ mext_insert_across_blocks(handle_t *handle, struct inode *orig_inode,
 		 * orig  |------------------------------|
 		 */
 		o_start->ee_len = start_ext->ee_len;
+		eblock = le32_to_cpu(start_ext->ee_block);
 		new_flag = 1;
 
 	} else if (!start_ext->ee_len && new_ext->ee_len &&
@@ -475,7 +477,6 @@ mext_leaf_block(handle_t *handle, struct inode *orig_inode,
 	struct ext4_extent *oext, *o_start, *o_end, *prev_ext;
 	struct ext4_extent new_ext, start_ext, end_ext;
 	ext4_lblk_t new_ext_end;
-	ext4_fsblk_t new_phys_end;
 	int oext_alen, new_ext_alen, end_ext_alen;
 	int depth = ext_depth(orig_inode);
 	int ret;
@@ -489,7 +490,6 @@ mext_leaf_block(handle_t *handle, struct inode *orig_inode,
 	new_ext.ee_len = dext->ee_len;
 	new_ext_alen = ext4_ext_get_actual_len(&new_ext);
 	new_ext_end = le32_to_cpu(new_ext.ee_block) + new_ext_alen - 1;
-	new_phys_end = ext_pblock(&new_ext) + new_ext_alen - 1;
 
 	/*
 	 * Case: original extent is first
@@ -502,6 +502,7 @@ mext_leaf_block(handle_t *handle, struct inode *orig_inode,
 		le32_to_cpu(oext->ee_block) + oext_alen) {
 		start_ext.ee_len = cpu_to_le16(le32_to_cpu(new_ext.ee_block) -
 					       le32_to_cpu(oext->ee_block));
+		start_ext.ee_block = oext->ee_block;
 		copy_extent_status(oext, &start_ext);
 	} else if (oext > EXT_FIRST_EXTENT(orig_path[depth].p_hdr)) {
 		prev_ext = oext - 1;
@@ -515,6 +516,7 @@ mext_leaf_block(handle_t *handle, struct inode *orig_inode,
 			start_ext.ee_len = cpu_to_le16(
 				ext4_ext_get_actual_len(prev_ext) +
 				new_ext_alen);
+			start_ext.ee_block = oext->ee_block;
 			copy_extent_status(prev_ext, &start_ext);
 			new_ext.ee_len = 0;
 		}
@@ -526,7 +528,7 @@ mext_leaf_block(handle_t *handle, struct inode *orig_inode,
 	 * new_ext       |-------|
 	 */
 	if (le32_to_cpu(oext->ee_block) + oext_alen - 1 < new_ext_end) {
-		ext4_error(orig_inode->i_sb, __func__,
+		ext4_error(orig_inode->i_sb,
 			"new_ext_end(%u) should be less than or equal to "
 			"oext->ee_block(%u) + oext_alen(%d) - 1",
 			new_ext_end, le32_to_cpu(oext->ee_block),
@@ -689,12 +691,12 @@ mext_replace_branches(handle_t *handle, struct inode *orig_inode,
 	while (1) {
 		/* The extent for donor must be found. */
 		if (!dext) {
-			ext4_error(donor_inode->i_sb, __func__,
+			ext4_error(donor_inode->i_sb,
 				   "The extent for donor must be found");
 			*err = -EIO;
 			goto out;
 		} else if (donor_off != le32_to_cpu(tmp_dext.ee_block)) {
-			ext4_error(donor_inode->i_sb, __func__,
+			ext4_error(donor_inode->i_sb,
 				"Donor offset(%u) and the first block of donor "
 				"extent(%u) should be equal",
 				donor_off,
@@ -928,7 +930,7 @@ out2:
 }
 
 /**
- * mext_check_argumants - Check whether move extent can be done
+ * mext_check_arguments - Check whether move extent can be done
  *
  * @orig_inode:		original inode
  * @donor_inode:	donor inode
@@ -949,14 +951,6 @@ mext_check_arguments(struct inode *orig_inode,
 	unsigned int blkbits = orig_inode->i_blkbits;
 	unsigned int blocksize = 1 << blkbits;
 
-	/* Regular file check */
-	if (!S_ISREG(orig_inode->i_mode) || !S_ISREG(donor_inode->i_mode)) {
-		ext4_debug("ext4 move extent: The argument files should be "
-			"regular file [ino:orig %lu, donor %lu]\n",
-			orig_inode->i_ino, donor_inode->i_ino);
-		return -EINVAL;
-	}
-
 	if (donor_inode->i_mode & (S_ISUID|S_ISGID)) {
 		ext4_debug("ext4 move extent: suid or sgid is set"
 			   " to donor file [ino:orig %lu, donor %lu]\n",
@@ -1204,6 +1198,14 @@ ext4_move_extents(struct file *o_filp, struct file *d_filp,
 		return -EINVAL;
 	}
 
+	/* Regular file check */
+	if (!S_ISREG(orig_inode->i_mode) || !S_ISREG(donor_inode->i_mode)) {
+		ext4_debug("ext4 move extent: The argument files should be "
+			"regular file [ino:orig %lu, donor %lu]\n",
+			orig_inode->i_ino, donor_inode->i_ino);
+		return -EINVAL;
+	}
+
 	/* Protect orig and donor inodes against a truncate */
 	ret1 = mext_inode_double_lock(orig_inode, donor_inode);
 	if (ret1 < 0)
@@ -1351,7 +1353,7 @@ ext4_move_extents(struct file *o_filp, struct file *d_filp,
 			if (ret1 < 0)
 				break;
 			if (*moved_len > len) {
-				ext4_error(orig_inode->i_sb, __func__,
+				ext4_error(orig_inode->i_sb,
 					"We replaced blocks too much! "
 					"sum of replaced: %llu requested: %llu",
 					*moved_len, len);
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 17a17e10dd60..0c070fabd108 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -383,8 +383,7 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 	if (root->info.hash_version != DX_HASH_TEA &&
 	    root->info.hash_version != DX_HASH_HALF_MD4 &&
 	    root->info.hash_version != DX_HASH_LEGACY) {
-		ext4_warning(dir->i_sb, __func__,
-			     "Unrecognised inode hash code %d",
+		ext4_warning(dir->i_sb, "Unrecognised inode hash code %d",
 			     root->info.hash_version);
 		brelse(bh);
 		*err = ERR_BAD_DX_DIR;
@@ -399,8 +398,7 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 	hash = hinfo->hash;
 
 	if (root->info.unused_flags & 1) {
-		ext4_warning(dir->i_sb, __func__,
-			     "Unimplemented inode hash flags: %#06x",
+		ext4_warning(dir->i_sb, "Unimplemented inode hash flags: %#06x",
 			     root->info.unused_flags);
 		brelse(bh);
 		*err = ERR_BAD_DX_DIR;
@@ -408,8 +406,7 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 	}
 
 	if ((indirect = root->info.indirect_levels) > 1) {
-		ext4_warning(dir->i_sb, __func__,
-			     "Unimplemented inode hash depth: %#06x",
+		ext4_warning(dir->i_sb, "Unimplemented inode hash depth: %#06x",
 			     root->info.indirect_levels);
 		brelse(bh);
 		*err = ERR_BAD_DX_DIR;
@@ -421,8 +418,7 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 
 	if (dx_get_limit(entries) != dx_root_limit(dir,
 						   root->info.info_length)) {
-		ext4_warning(dir->i_sb, __func__,
-			     "dx entry: limit != root limit");
+		ext4_warning(dir->i_sb, "dx entry: limit != root limit");
 		brelse(bh);
 		*err = ERR_BAD_DX_DIR;
 		goto fail;
@@ -433,7 +429,7 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 	{
 		count = dx_get_count(entries);
 		if (!count || count > dx_get_limit(entries)) {
-			ext4_warning(dir->i_sb, __func__,
+			ext4_warning(dir->i_sb,
 				     "dx entry: no count or count > limit");
 			brelse(bh);
 			*err = ERR_BAD_DX_DIR;
@@ -478,7 +474,7 @@ dx_probe(const struct qstr *d_name, struct inode *dir,
 			goto fail2;
 		at = entries = ((struct dx_node *) bh->b_data)->entries;
 		if (dx_get_limit(entries) != dx_node_limit (dir)) {
-			ext4_warning(dir->i_sb, __func__,
+			ext4_warning(dir->i_sb,
 				     "dx entry: limit != node limit");
 			brelse(bh);
 			*err = ERR_BAD_DX_DIR;
@@ -494,7 +490,7 @@ fail2:
 	}
 fail:
 	if (*err == ERR_BAD_DX_DIR)
-		ext4_warning(dir->i_sb, __func__,
+		ext4_warning(dir->i_sb,
 			     "Corrupt dir inode %ld, running e2fsck is "
 			     "recommended.", dir->i_ino);
 	return NULL;
@@ -947,9 +943,8 @@ restart:
 		wait_on_buffer(bh);
 		if (!buffer_uptodate(bh)) {
 			/* read error, skip block & hope for the best */
-			ext4_error(sb, __func__, "reading directory #%lu "
-				   "offset %lu", dir->i_ino,
-				   (unsigned long)block);
+			ext4_error(sb, "reading directory #%lu offset %lu",
+				   dir->i_ino, (unsigned long)block);
 			brelse(bh);
 			goto next;
 		}
@@ -1041,7 +1036,7 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q
 		retval = ext4_htree_next_block(dir, hash, frame,
 					       frames, NULL);
 		if (retval < 0) {
-			ext4_warning(sb, __func__,
+			ext4_warning(sb,
 			     "error reading index page in directory #%lu",
 			     dir->i_ino);
 			*err = retval;
@@ -1071,14 +1066,13 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, stru
 		__u32 ino = le32_to_cpu(de->inode);
 		brelse(bh);
 		if (!ext4_valid_inum(dir->i_sb, ino)) {
-			ext4_error(dir->i_sb, "ext4_lookup",
-				   "bad inode number: %u", ino);
+			ext4_error(dir->i_sb, "bad inode number: %u", ino);
 			return ERR_PTR(-EIO);
 		}
 		inode = ext4_iget(dir->i_sb, ino);
 		if (unlikely(IS_ERR(inode))) {
 			if (PTR_ERR(inode) == -ESTALE) {
-				ext4_error(dir->i_sb, __func__,
+				ext4_error(dir->i_sb,
 						"deleted inode referenced: %u",
 						ino);
 				return ERR_PTR(-EIO);
@@ -1110,7 +1104,7 @@ struct dentry *ext4_get_parent(struct dentry *child)
 	brelse(bh);
 
 	if (!ext4_valid_inum(child->d_inode->i_sb, ino)) {
-		ext4_error(child->d_inode->i_sb, "ext4_get_parent",
+		ext4_error(child->d_inode->i_sb,
 			   "bad inode number: %u", ino);
 		return ERR_PTR(-EIO);
 	}
@@ -1410,7 +1404,7 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry,
 	de = (struct ext4_dir_entry_2 *)((char *)fde +
 		ext4_rec_len_from_disk(fde->rec_len, blocksize));
 	if ((char *) de >= (((char *) root) + blocksize)) {
-		ext4_error(dir->i_sb, __func__,
+		ext4_error(dir->i_sb,
 			   "invalid rec_len for '..' in inode %lu",
 			   dir->i_ino);
 		brelse(bh);
@@ -1575,8 +1569,7 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry,
 
 		if (levels && (dx_get_count(frames->entries) ==
 			       dx_get_limit(frames->entries))) {
-			ext4_warning(sb, __func__,
-				     "Directory index full!");
+			ext4_warning(sb, "Directory index full!");
 			err = -ENOSPC;
 			goto cleanup;
 		}
@@ -1766,6 +1759,8 @@ static int ext4_create(struct inode *dir, struct dentry *dentry, int mode,
 	struct inode *inode;
 	int err, retries = 0;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 +
@@ -1800,6 +1795,8 @@ static int ext4_mknod(struct inode *dir, struct dentry *dentry,
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 +
@@ -1837,6 +1834,8 @@ static int ext4_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	if (EXT4_DIR_LINK_MAX(dir))
 		return -EMLINK;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 +
@@ -1916,11 +1915,11 @@ static int empty_dir(struct inode *inode)
 	if (inode->i_size < EXT4_DIR_REC_LEN(1) + EXT4_DIR_REC_LEN(2) ||
 	    !(bh = ext4_bread(NULL, inode, 0, 0, &err))) {
 		if (err)
-			ext4_error(inode->i_sb, __func__,
+			ext4_error(inode->i_sb,
 				   "error %d reading directory #%lu offset 0",
 				   err, inode->i_ino);
 		else
-			ext4_warning(inode->i_sb, __func__,
+			ext4_warning(inode->i_sb,
 				     "bad directory (dir #%lu) - no data block",
 				     inode->i_ino);
 		return 1;
@@ -1931,7 +1930,7 @@ static int empty_dir(struct inode *inode)
 			!le32_to_cpu(de1->inode) ||
 			strcmp(".", de->name) ||
 			strcmp("..", de1->name)) {
-		ext4_warning(inode->i_sb, "empty_dir",
+		ext4_warning(inode->i_sb,
 			     "bad directory (dir #%lu) - no `.' or `..'",
 			     inode->i_ino);
 		brelse(bh);
@@ -1949,7 +1948,7 @@ static int empty_dir(struct inode *inode)
 				offset >> EXT4_BLOCK_SIZE_BITS(sb), 0, &err);
 			if (!bh) {
 				if (err)
-					ext4_error(sb, __func__,
+					ext4_error(sb,
 						   "error %d reading directory"
 						   " #%lu offset %u",
 						   err, inode->i_ino, offset);
@@ -2020,11 +2019,18 @@ int ext4_orphan_add(handle_t *handle, struct inode *inode)
 	err = ext4_reserve_inode_write(handle, inode, &iloc);
 	if (err)
 		goto out_unlock;
+	/*
+	 * Due to previous errors inode may be already a part of on-disk
+	 * orphan list. If so skip on-disk list modification.
+	 */
+	if (NEXT_ORPHAN(inode) && NEXT_ORPHAN(inode) <=
+		(le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count)))
+			goto mem_insert;
 
 	/* Insert this inode at the head of the on-disk orphan list... */
 	NEXT_ORPHAN(inode) = le32_to_cpu(EXT4_SB(sb)->s_es->s_last_orphan);
 	EXT4_SB(sb)->s_es->s_last_orphan = cpu_to_le32(inode->i_ino);
-	err = ext4_handle_dirty_metadata(handle, inode, EXT4_SB(sb)->s_sbh);
+	err = ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
 	rc = ext4_mark_iloc_dirty(handle, inode, &iloc);
 	if (!err)
 		err = rc;
@@ -2037,6 +2043,7 @@ int ext4_orphan_add(handle_t *handle, struct inode *inode)
 	 *
 	 * This is safe: on error we're going to ignore the orphan list
 	 * anyway on the next recovery. */
+mem_insert:
 	if (!err)
 		list_add(&EXT4_I(inode)->i_orphan, &EXT4_SB(sb)->s_orphan);
 
@@ -2096,7 +2103,7 @@ int ext4_orphan_del(handle_t *handle, struct inode *inode)
 		if (err)
 			goto out_brelse;
 		sbi->s_es->s_last_orphan = cpu_to_le32(ino_next);
-		err = ext4_handle_dirty_metadata(handle, inode, sbi->s_sbh);
+		err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh);
 	} else {
 		struct ext4_iloc iloc2;
 		struct inode *i_prev =
@@ -2136,7 +2143,9 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry)
 
 	/* Initialize quotas before so that eventual writes go in
 	 * separate transaction */
-	vfs_dq_init(dentry->d_inode);
+	dquot_initialize(dir);
+	dquot_initialize(dentry->d_inode);
+
 	handle = ext4_journal_start(dir, EXT4_DELETE_TRANS_BLOCKS(dir->i_sb));
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
@@ -2163,7 +2172,7 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry)
 	if (retval)
 		goto end_rmdir;
 	if (!EXT4_DIR_LINK_EMPTY(inode))
-		ext4_warning(inode->i_sb, "ext4_rmdir",
+		ext4_warning(inode->i_sb,
 			     "empty directory has too many links (%d)",
 			     inode->i_nlink);
 	inode->i_version++;
@@ -2195,7 +2204,9 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry)
 
 	/* Initialize quotas before so that eventual writes go
 	 * in separate transaction */
-	vfs_dq_init(dentry->d_inode);
+	dquot_initialize(dir);
+	dquot_initialize(dentry->d_inode);
+
 	handle = ext4_journal_start(dir, EXT4_DELETE_TRANS_BLOCKS(dir->i_sb));
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
@@ -2215,7 +2226,7 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry)
 		goto end_unlink;
 
 	if (!inode->i_nlink) {
-		ext4_warning(inode->i_sb, "ext4_unlink",
+		ext4_warning(inode->i_sb,
 			     "Deleting nonexistent file (%lu), %d",
 			     inode->i_ino, inode->i_nlink);
 		inode->i_nlink = 1;
@@ -2250,6 +2261,8 @@ static int ext4_symlink(struct inode *dir,
 	if (l > dir->i_sb->s_blocksize)
 		return -ENAMETOOLONG;
 
+	dquot_initialize(dir);
+
 retry:
 	handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
 					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 5 +
@@ -2308,6 +2321,8 @@ static int ext4_link(struct dentry *old_dentry,
 	if (inode->i_nlink >= EXT4_LINK_MAX)
 		return -EMLINK;
 
+	dquot_initialize(dir);
+
 	/*
 	 * Return -ENOENT if we've raced with unlink and i_nlink is 0.  Doing
 	 * otherwise has the potential to corrupt the orphan inode list.
@@ -2358,12 +2373,15 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
 	struct ext4_dir_entry_2 *old_de, *new_de;
 	int retval, force_da_alloc = 0;
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	old_bh = new_bh = dir_bh = NULL;
 
 	/* Initialize quotas before so that eventual writes go
 	 * in separate transaction */
 	if (new_dentry->d_inode)
-		vfs_dq_init(new_dentry->d_inode);
+		dquot_initialize(new_dentry->d_inode);
 	handle = ext4_journal_start(old_dir, 2 *
 					EXT4_DATA_TRANS_BLOCKS(old_dir->i_sb) +
 					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 2);
@@ -2462,7 +2480,7 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
 		}
 	}
 	if (retval) {
-		ext4_warning(old_dir->i_sb, "ext4_rename",
+		ext4_warning(old_dir->i_sb,
 				"Deleting old file (%lu), %d, error=%d",
 				old_dir->i_ino, old_dir->i_nlink, retval);
 	}
diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
index 3b2c5541d8a6..5692c48754a0 100644
--- a/fs/ext4/resize.c
+++ b/fs/ext4/resize.c
@@ -48,65 +48,54 @@ static int verify_group_input(struct super_block *sb,
 
 	ext4_get_group_no_and_offset(sb, start, NULL, &offset);
 	if (group != sbi->s_groups_count)
-		ext4_warning(sb, __func__,
-			     "Cannot add at group %u (only %u groups)",
+		ext4_warning(sb, "Cannot add at group %u (only %u groups)",
 			     input->group, sbi->s_groups_count);
 	else if (offset != 0)
-			ext4_warning(sb, __func__, "Last group not full");
+			ext4_warning(sb, "Last group not full");
 	else if (input->reserved_blocks > input->blocks_count / 5)
-		ext4_warning(sb, __func__, "Reserved blocks too high (%u)",
+		ext4_warning(sb, "Reserved blocks too high (%u)",
 			     input->reserved_blocks);
 	else if (free_blocks_count < 0)
-		ext4_warning(sb, __func__, "Bad blocks count %u",
+		ext4_warning(sb, "Bad blocks count %u",
 			     input->blocks_count);
 	else if (!(bh = sb_bread(sb, end - 1)))
-		ext4_warning(sb, __func__,
-			     "Cannot read last block (%llu)",
+		ext4_warning(sb, "Cannot read last block (%llu)",
 			     end - 1);
 	else if (outside(input->block_bitmap, start, end))
-		ext4_warning(sb, __func__,
-			     "Block bitmap not in group (block %llu)",
+		ext4_warning(sb, "Block bitmap not in group (block %llu)",
 			     (unsigned long long)input->block_bitmap);
 	else if (outside(input->inode_bitmap, start, end))
-		ext4_warning(sb, __func__,
-			     "Inode bitmap not in group (block %llu)",
+		ext4_warning(sb, "Inode bitmap not in group (block %llu)",
 			     (unsigned long long)input->inode_bitmap);
 	else if (outside(input->inode_table, start, end) ||
 		 outside(itend - 1, start, end))
-		ext4_warning(sb, __func__,
-			     "Inode table not in group (blocks %llu-%llu)",
+		ext4_warning(sb, "Inode table not in group (blocks %llu-%llu)",
 			     (unsigned long long)input->inode_table, itend - 1);
 	else if (input->inode_bitmap == input->block_bitmap)
-		ext4_warning(sb, __func__,
-			     "Block bitmap same as inode bitmap (%llu)",
+		ext4_warning(sb, "Block bitmap same as inode bitmap (%llu)",
 			     (unsigned long long)input->block_bitmap);
 	else if (inside(input->block_bitmap, input->inode_table, itend))
-		ext4_warning(sb, __func__,
-			     "Block bitmap (%llu) in inode table (%llu-%llu)",
+		ext4_warning(sb, "Block bitmap (%llu) in inode table "
+			     "(%llu-%llu)",
 			     (unsigned long long)input->block_bitmap,
 			     (unsigned long long)input->inode_table, itend - 1);
 	else if (inside(input->inode_bitmap, input->inode_table, itend))
-		ext4_warning(sb, __func__,
-			     "Inode bitmap (%llu) in inode table (%llu-%llu)",
+		ext4_warning(sb, "Inode bitmap (%llu) in inode table "
+			     "(%llu-%llu)",
 			     (unsigned long long)input->inode_bitmap,
 			     (unsigned long long)input->inode_table, itend - 1);
 	else if (inside(input->block_bitmap, start, metaend))
-		ext4_warning(sb, __func__,
-			     "Block bitmap (%llu) in GDT table"
-			     " (%llu-%llu)",
+		ext4_warning(sb, "Block bitmap (%llu) in GDT table (%llu-%llu)",
 			     (unsigned long long)input->block_bitmap,
 			     start, metaend - 1);
 	else if (inside(input->inode_bitmap, start, metaend))
-		ext4_warning(sb, __func__,
-			     "Inode bitmap (%llu) in GDT table"
-			     " (%llu-%llu)",
+		ext4_warning(sb, "Inode bitmap (%llu) in GDT table (%llu-%llu)",
 			     (unsigned long long)input->inode_bitmap,
 			     start, metaend - 1);
 	else if (inside(input->inode_table, start, metaend) ||
 		 inside(itend - 1, start, metaend))
-		ext4_warning(sb, __func__,
-			     "Inode table (%llu-%llu) overlaps"
-			     "GDT table (%llu-%llu)",
+		ext4_warning(sb, "Inode table (%llu-%llu) overlaps GDT table "
+			     "(%llu-%llu)",
 			     (unsigned long long)input->inode_table,
 			     itend - 1, start, metaend - 1);
 	else
@@ -364,8 +353,7 @@ static int verify_reserved_gdb(struct super_block *sb,
 	while ((grp = ext4_list_backups(sb, &three, &five, &seven)) < end) {
 		if (le32_to_cpu(*p++) !=
 		    grp * EXT4_BLOCKS_PER_GROUP(sb) + blk){
-			ext4_warning(sb, __func__,
-				     "reserved GDT %llu"
+			ext4_warning(sb, "reserved GDT %llu"
 				     " missing grp %d (%llu)",
 				     blk, grp,
 				     grp *
@@ -420,8 +408,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
          */
 	if (EXT4_SB(sb)->s_sbh->b_blocknr !=
 	    le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block)) {
-		ext4_warning(sb, __func__,
-			"won't resize using backup superblock at %llu",
+		ext4_warning(sb, "won't resize using backup superblock at %llu",
 			(unsigned long long)EXT4_SB(sb)->s_sbh->b_blocknr);
 		return -EPERM;
 	}
@@ -444,8 +431,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 
 	data = (__le32 *)dind->b_data;
 	if (le32_to_cpu(data[gdb_num % EXT4_ADDR_PER_BLOCK(sb)]) != gdblock) {
-		ext4_warning(sb, __func__,
-			     "new group %u GDT block %llu not reserved",
+		ext4_warning(sb, "new group %u GDT block %llu not reserved",
 			     input->group, gdblock);
 		err = -EINVAL;
 		goto exit_dind;
@@ -468,7 +454,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 			GFP_NOFS);
 	if (!n_group_desc) {
 		err = -ENOMEM;
-		ext4_warning(sb, __func__,
+		ext4_warning(sb,
 			      "not enough memory for %lu groups", gdb_num + 1);
 		goto exit_inode;
 	}
@@ -567,8 +553,7 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode,
 	/* Get each reserved primary GDT block and verify it holds backups */
 	for (res = 0; res < reserved_gdb; res++, blk++) {
 		if (le32_to_cpu(*data) != blk) {
-			ext4_warning(sb, __func__,
-				     "reserved block %llu"
+			ext4_warning(sb, "reserved block %llu"
 				     " not at offset %ld",
 				     blk,
 				     (long)(data - (__le32 *)dind->b_data));
@@ -713,8 +698,7 @@ static void update_backups(struct super_block *sb,
 	 */
 exit_err:
 	if (err) {
-		ext4_warning(sb, __func__,
-			     "can't update backup for group %u (err %d), "
+		ext4_warning(sb, "can't update backup for group %u (err %d), "
 			     "forcing fsck on next reboot", group, err);
 		sbi->s_mount_state &= ~EXT4_VALID_FS;
 		sbi->s_es->s_state &= cpu_to_le16(~EXT4_VALID_FS);
@@ -753,20 +737,19 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input)
 
 	if (gdb_off == 0 && !EXT4_HAS_RO_COMPAT_FEATURE(sb,
 					EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER)) {
-		ext4_warning(sb, __func__,
-			     "Can't resize non-sparse filesystem further");
+		ext4_warning(sb, "Can't resize non-sparse filesystem further");
 		return -EPERM;
 	}
 
 	if (ext4_blocks_count(es) + input->blocks_count <
 	    ext4_blocks_count(es)) {
-		ext4_warning(sb, __func__, "blocks_count overflow");
+		ext4_warning(sb, "blocks_count overflow");
 		return -EINVAL;
 	}
 
 	if (le32_to_cpu(es->s_inodes_count) + EXT4_INODES_PER_GROUP(sb) <
 	    le32_to_cpu(es->s_inodes_count)) {
-		ext4_warning(sb, __func__, "inodes_count overflow");
+		ext4_warning(sb, "inodes_count overflow");
 		return -EINVAL;
 	}
 
@@ -774,14 +757,13 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input)
 		if (!EXT4_HAS_COMPAT_FEATURE(sb,
 					     EXT4_FEATURE_COMPAT_RESIZE_INODE)
 		    || !le16_to_cpu(es->s_reserved_gdt_blocks)) {
-			ext4_warning(sb, __func__,
+			ext4_warning(sb,
 				     "No reserved GDT blocks, can't resize");
 			return -EPERM;
 		}
 		inode = ext4_iget(sb, EXT4_RESIZE_INO);
 		if (IS_ERR(inode)) {
-			ext4_warning(sb, __func__,
-				     "Error opening resize inode");
+			ext4_warning(sb, "Error opening resize inode");
 			return PTR_ERR(inode);
 		}
 	}
@@ -810,8 +792,7 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input)
 
 	mutex_lock(&sbi->s_resize_lock);
 	if (input->group != sbi->s_groups_count) {
-		ext4_warning(sb, __func__,
-			     "multiple resizers run on filesystem!");
+		ext4_warning(sb, "multiple resizers run on filesystem!");
 		err = -EBUSY;
 		goto exit_journal;
 	}
@@ -997,13 +978,12 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 			" too large to resize to %llu blocks safely\n",
 			sb->s_id, n_blocks_count);
 		if (sizeof(sector_t) < 8)
-			ext4_warning(sb, __func__, "CONFIG_LBDAF not enabled");
+			ext4_warning(sb, "CONFIG_LBDAF not enabled");
 		return -EINVAL;
 	}
 
 	if (n_blocks_count < o_blocks_count) {
-		ext4_warning(sb, __func__,
-			     "can't shrink FS - resize aborted");
+		ext4_warning(sb, "can't shrink FS - resize aborted");
 		return -EBUSY;
 	}
 
@@ -1011,15 +991,14 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 	ext4_get_group_no_and_offset(sb, o_blocks_count, &group, &last);
 
 	if (last == 0) {
-		ext4_warning(sb, __func__,
-			     "need to use ext2online to resize further");
+		ext4_warning(sb, "need to use ext2online to resize further");
 		return -EPERM;
 	}
 
 	add = EXT4_BLOCKS_PER_GROUP(sb) - last;
 
 	if (o_blocks_count + add < o_blocks_count) {
-		ext4_warning(sb, __func__, "blocks_count overflow");
+		ext4_warning(sb, "blocks_count overflow");
 		return -EINVAL;
 	}
 
@@ -1027,16 +1006,13 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 		add = n_blocks_count - o_blocks_count;
 
 	if (o_blocks_count + add < n_blocks_count)
-		ext4_warning(sb, __func__,
-			     "will only finish group (%llu"
-			     " blocks, %u new)",
+		ext4_warning(sb, "will only finish group (%llu blocks, %u new)",
 			     o_blocks_count + add, add);
 
 	/* See if the device is actually as big as what was requested */
 	bh = sb_bread(sb, o_blocks_count + add - 1);
 	if (!bh) {
-		ext4_warning(sb, __func__,
-			     "can't read last block, resize aborted");
+		ext4_warning(sb, "can't read last block, resize aborted");
 		return -ENOSPC;
 	}
 	brelse(bh);
@@ -1047,14 +1023,13 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 	handle = ext4_journal_start_sb(sb, 3);
 	if (IS_ERR(handle)) {
 		err = PTR_ERR(handle);
-		ext4_warning(sb, __func__, "error %d on journal start", err);
+		ext4_warning(sb, "error %d on journal start", err);
 		goto exit_put;
 	}
 
 	mutex_lock(&EXT4_SB(sb)->s_resize_lock);
 	if (o_blocks_count != ext4_blocks_count(es)) {
-		ext4_warning(sb, __func__,
-			     "multiple resizers run on filesystem!");
+		ext4_warning(sb, "multiple resizers run on filesystem!");
 		mutex_unlock(&EXT4_SB(sb)->s_resize_lock);
 		ext4_journal_stop(handle);
 		err = -EBUSY;
@@ -1063,8 +1038,7 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es,
 
 	if ((err = ext4_journal_get_write_access(handle,
 						 EXT4_SB(sb)->s_sbh))) {
-		ext4_warning(sb, __func__,
-			     "error %d on journal write access", err);
+		ext4_warning(sb, "error %d on journal write access", err);
 		mutex_unlock(&EXT4_SB(sb)->s_resize_lock);
 		ext4_journal_stop(handle);
 		goto exit_put;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 735c20d5fd56..2b83b96cb2eb 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -333,7 +333,7 @@ static void ext4_handle_error(struct super_block *sb)
 			sb->s_id);
 }
 
-void ext4_error(struct super_block *sb, const char *function,
+void __ext4_error(struct super_block *sb, const char *function,
 		const char *fmt, ...)
 {
 	va_list args;
@@ -347,6 +347,42 @@ void ext4_error(struct super_block *sb, const char *function,
 	ext4_handle_error(sb);
 }
 
+void ext4_error_inode(const char *function, struct inode *inode,
+		      const char *fmt, ...)
+{
+	va_list args;
+
+	va_start(args, fmt);
+	printk(KERN_CRIT "EXT4-fs error (device %s): %s: inode #%lu: (comm %s) ",
+	       inode->i_sb->s_id, function, inode->i_ino, current->comm);
+	vprintk(fmt, args);
+	printk("\n");
+	va_end(args);
+
+	ext4_handle_error(inode->i_sb);
+}
+
+void ext4_error_file(const char *function, struct file *file,
+		     const char *fmt, ...)
+{
+	va_list args;
+	struct inode *inode = file->f_dentry->d_inode;
+	char pathname[80], *path;
+
+	va_start(args, fmt);
+	path = d_path(&(file->f_path), pathname, sizeof(pathname));
+	if (!path)
+		path = "(unknown)";
+	printk(KERN_CRIT
+	       "EXT4-fs error (device %s): %s: inode #%lu (comm %s path %s): ",
+	       inode->i_sb->s_id, function, inode->i_ino, current->comm, path);
+	vprintk(fmt, args);
+	printk("\n");
+	va_end(args);
+
+	ext4_handle_error(inode->i_sb);
+}
+
 static const char *ext4_decode_error(struct super_block *sb, int errno,
 				     char nbuf[16])
 {
@@ -450,7 +486,7 @@ void ext4_msg (struct super_block * sb, const char *prefix,
 	va_end(args);
 }
 
-void ext4_warning(struct super_block *sb, const char *function,
+void __ext4_warning(struct super_block *sb, const char *function,
 		  const char *fmt, ...)
 {
 	va_list args;
@@ -507,7 +543,7 @@ void ext4_update_dynamic_rev(struct super_block *sb)
 	if (le32_to_cpu(es->s_rev_level) > EXT4_GOOD_OLD_REV)
 		return;
 
-	ext4_warning(sb, __func__,
+	ext4_warning(sb,
 		     "updating to rev %d because of new feature flag, "
 		     "running e2fsck is recommended",
 		     EXT4_DYNAMIC_REV);
@@ -708,7 +744,8 @@ static struct inode *ext4_alloc_inode(struct super_block *sb)
 #ifdef CONFIG_QUOTA
 	ei->i_reserved_quota = 0;
 #endif
-	INIT_LIST_HEAD(&ei->i_aio_dio_complete_list);
+	INIT_LIST_HEAD(&ei->i_completed_io_list);
+	spin_lock_init(&ei->i_completed_io_lock);
 	ei->cur_aio_dio = NULL;
 	ei->i_sync_tid = 0;
 	ei->i_datasync_tid = 0;
@@ -761,6 +798,7 @@ static void destroy_inodecache(void)
 
 static void ext4_clear_inode(struct inode *inode)
 {
+	dquot_drop(inode);
 	ext4_discard_preallocations(inode);
 	if (EXT4_JOURNAL(inode))
 		jbd2_journal_release_jbd_inode(EXT4_SB(inode->i_sb)->s_journal,
@@ -796,10 +834,10 @@ static inline void ext4_show_quota_options(struct seq_file *seq,
 	if (sbi->s_qf_names[GRPQUOTA])
 		seq_printf(seq, ",grpjquota=%s", sbi->s_qf_names[GRPQUOTA]);
 
-	if (sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA)
+	if (test_opt(sb, USRQUOTA))
 		seq_puts(seq, ",usrquota");
 
-	if (sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA)
+	if (test_opt(sb, GRPQUOTA))
 		seq_puts(seq, ",grpquota");
 #endif
 }
@@ -926,6 +964,9 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	if (test_opt(sb, NOLOAD))
 		seq_puts(seq, ",norecovery");
 
+	if (test_opt(sb, DIOREAD_NOLOCK))
+		seq_puts(seq, ",dioread_nolock");
+
 	ext4_show_quota_options(seq, sb);
 
 	return 0;
@@ -1012,19 +1053,9 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type,
 				const char *data, size_t len, loff_t off);
 
 static const struct dquot_operations ext4_quota_operations = {
-	.initialize	= dquot_initialize,
-	.drop		= dquot_drop,
-	.alloc_space	= dquot_alloc_space,
-	.reserve_space	= dquot_reserve_space,
-	.claim_space	= dquot_claim_space,
-	.release_rsv	= dquot_release_reserved_space,
 #ifdef CONFIG_QUOTA
 	.get_reserved_space = ext4_get_reserved_space,
 #endif
-	.alloc_inode	= dquot_alloc_inode,
-	.free_space	= dquot_free_space,
-	.free_inode	= dquot_free_inode,
-	.transfer	= dquot_transfer,
 	.write_dquot	= ext4_write_dquot,
 	.acquire_dquot	= ext4_acquire_dquot,
 	.release_dquot	= ext4_release_dquot,
@@ -1109,6 +1140,7 @@ enum {
 	Opt_stripe, Opt_delalloc, Opt_nodelalloc,
 	Opt_block_validity, Opt_noblock_validity,
 	Opt_inode_readahead_blks, Opt_journal_ioprio,
+	Opt_dioread_nolock, Opt_dioread_lock,
 	Opt_discard, Opt_nodiscard,
 };
 
@@ -1176,6 +1208,8 @@ static const match_table_t tokens = {
 	{Opt_auto_da_alloc, "auto_da_alloc=%u"},
 	{Opt_auto_da_alloc, "auto_da_alloc"},
 	{Opt_noauto_da_alloc, "noauto_da_alloc"},
+	{Opt_dioread_nolock, "dioread_nolock"},
+	{Opt_dioread_lock, "dioread_lock"},
 	{Opt_discard, "discard"},
 	{Opt_nodiscard, "nodiscard"},
 	{Opt_err, NULL},
@@ -1205,6 +1239,66 @@ static ext4_fsblk_t get_sb_block(void **data)
 }
 
 #define DEFAULT_JOURNAL_IOPRIO (IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, 3))
+static char deprecated_msg[] = "Mount option \"%s\" will be removed by %s\n"
+	"Contact linux-ext4@vger.kernel.org if you think we should keep it.\n";
+
+#ifdef CONFIG_QUOTA
+static int set_qf_name(struct super_block *sb, int qtype, substring_t *args)
+{
+	struct ext4_sb_info *sbi = EXT4_SB(sb);
+	char *qname;
+
+	if (sb_any_quota_loaded(sb) &&
+		!sbi->s_qf_names[qtype]) {
+		ext4_msg(sb, KERN_ERR,
+			"Cannot change journaled "
+			"quota options when quota turned on");
+		return 0;
+	}
+	qname = match_strdup(args);
+	if (!qname) {
+		ext4_msg(sb, KERN_ERR,
+			"Not enough memory for storing quotafile name");
+		return 0;
+	}
+	if (sbi->s_qf_names[qtype] &&
+		strcmp(sbi->s_qf_names[qtype], qname)) {
+		ext4_msg(sb, KERN_ERR,
+			"%s quota file already specified", QTYPE2NAME(qtype));
+		kfree(qname);
+		return 0;
+	}
+	sbi->s_qf_names[qtype] = qname;
+	if (strchr(sbi->s_qf_names[qtype], '/')) {
+		ext4_msg(sb, KERN_ERR,
+			"quotafile must be on filesystem root");
+		kfree(sbi->s_qf_names[qtype]);
+		sbi->s_qf_names[qtype] = NULL;
+		return 0;
+	}
+	set_opt(sbi->s_mount_opt, QUOTA);
+	return 1;
+}
+
+static int clear_qf_name(struct super_block *sb, int qtype)
+{
+
+	struct ext4_sb_info *sbi = EXT4_SB(sb);
+
+	if (sb_any_quota_loaded(sb) &&
+		sbi->s_qf_names[qtype]) {
+		ext4_msg(sb, KERN_ERR, "Cannot change journaled quota options"
+			" when quota turned on");
+		return 0;
+	}
+	/*
+	 * The space will be released later when all options are confirmed
+	 * to be correct
+	 */
+	sbi->s_qf_names[qtype] = NULL;
+	return 1;
+}
+#endif
 
 static int parse_options(char *options, struct super_block *sb,
 			 unsigned long *journal_devnum,
@@ -1217,8 +1311,7 @@ static int parse_options(char *options, struct super_block *sb,
 	int data_opt = 0;
 	int option;
 #ifdef CONFIG_QUOTA
-	int qtype, qfmt;
-	char *qname;
+	int qfmt;
 #endif
 
 	if (!options)
@@ -1229,19 +1322,31 @@ static int parse_options(char *options, struct super_block *sb,
 		if (!*p)
 			continue;
 
+		/*
+		 * Initialize args struct so we know whether arg was
+		 * found; some options take optional arguments.
+		 */
+		args[0].to = args[0].from = 0;
 		token = match_token(p, tokens, args);
 		switch (token) {
 		case Opt_bsd_df:
+			ext4_msg(sb, KERN_WARNING, deprecated_msg, p, "2.6.38");
 			clear_opt(sbi->s_mount_opt, MINIX_DF);
 			break;
 		case Opt_minix_df:
+			ext4_msg(sb, KERN_WARNING, deprecated_msg, p, "2.6.38");
 			set_opt(sbi->s_mount_opt, MINIX_DF);
+
 			break;
 		case Opt_grpid:
+			ext4_msg(sb, KERN_WARNING, deprecated_msg, p, "2.6.38");
 			set_opt(sbi->s_mount_opt, GRPID);
+
 			break;
 		case Opt_nogrpid:
+			ext4_msg(sb, KERN_WARNING, deprecated_msg, p, "2.6.38");
 			clear_opt(sbi->s_mount_opt, GRPID);
+
 			break;
 		case Opt_resuid:
 			if (match_int(&args[0], &option))
@@ -1378,14 +1483,13 @@ static int parse_options(char *options, struct super_block *sb,
 			data_opt = EXT4_MOUNT_WRITEBACK_DATA;
 		datacheck:
 			if (is_remount) {
-				if ((sbi->s_mount_opt & EXT4_MOUNT_DATA_FLAGS)
-						!= data_opt) {
+				if (test_opt(sb, DATA_FLAGS) != data_opt) {
 					ext4_msg(sb, KERN_ERR,
 						"Cannot change data mode on remount");
 					return 0;
 				}
 			} else {
-				sbi->s_mount_opt &= ~EXT4_MOUNT_DATA_FLAGS;
+				clear_opt(sbi->s_mount_opt, DATA_FLAGS);
 				sbi->s_mount_opt |= data_opt;
 			}
 			break;
@@ -1397,63 +1501,22 @@ static int parse_options(char *options, struct super_block *sb,
 			break;
 #ifdef CONFIG_QUOTA
 		case Opt_usrjquota:
-			qtype = USRQUOTA;
-			goto set_qf_name;
-		case Opt_grpjquota:
-			qtype = GRPQUOTA;
-set_qf_name:
-			if (sb_any_quota_loaded(sb) &&
-			    !sbi->s_qf_names[qtype]) {
-				ext4_msg(sb, KERN_ERR,
-				       "Cannot change journaled "
-				       "quota options when quota turned on");
+			if (!set_qf_name(sb, USRQUOTA, &args[0]))
 				return 0;
-			}
-			qname = match_strdup(&args[0]);
-			if (!qname) {
-				ext4_msg(sb, KERN_ERR,
-					"Not enough memory for "
-					"storing quotafile name");
-				return 0;
-			}
-			if (sbi->s_qf_names[qtype] &&
-			    strcmp(sbi->s_qf_names[qtype], qname)) {
-				ext4_msg(sb, KERN_ERR,
-					"%s quota file already "
-					"specified", QTYPE2NAME(qtype));
-				kfree(qname);
-				return 0;
-			}
-			sbi->s_qf_names[qtype] = qname;
-			if (strchr(sbi->s_qf_names[qtype], '/')) {
-				ext4_msg(sb, KERN_ERR,
-					"quotafile must be on "
-					"filesystem root");
-				kfree(sbi->s_qf_names[qtype]);
-				sbi->s_qf_names[qtype] = NULL;
+			break;
+		case Opt_grpjquota:
+			if (!set_qf_name(sb, GRPQUOTA, &args[0]))
 				return 0;
-			}
-			set_opt(sbi->s_mount_opt, QUOTA);
 			break;
 		case Opt_offusrjquota:
-			qtype = USRQUOTA;
-			goto clear_qf_name;
+			if (!clear_qf_name(sb, USRQUOTA))
+				return 0;
+			break;
 		case Opt_offgrpjquota:
-			qtype = GRPQUOTA;
-clear_qf_name:
-			if (sb_any_quota_loaded(sb) &&
-			    sbi->s_qf_names[qtype]) {
-				ext4_msg(sb, KERN_ERR, "Cannot change "
-					"journaled quota options when "
-					"quota turned on");
+			if (!clear_qf_name(sb, GRPQUOTA))
 				return 0;
-			}
-			/*
-			 * The space will be released later when all options
-			 * are confirmed to be correct
-			 */
-			sbi->s_qf_names[qtype] = NULL;
 			break;
+
 		case Opt_jqfmt_vfsold:
 			qfmt = QFMT_VFS_OLD;
 			goto set_qf_format;
@@ -1518,10 +1581,11 @@ set_qf_format:
 			clear_opt(sbi->s_mount_opt, BARRIER);
 			break;
 		case Opt_barrier:
-			if (match_int(&args[0], &option)) {
-				set_opt(sbi->s_mount_opt, BARRIER);
-				break;
-			}
+			if (args[0].from) {
+				if (match_int(&args[0], &option))
+					return 0;
+			} else
+				option = 1;	/* No argument, default to 1 */
 			if (option)
 				set_opt(sbi->s_mount_opt, BARRIER);
 			else
@@ -1594,10 +1658,11 @@ set_qf_format:
 			set_opt(sbi->s_mount_opt,NO_AUTO_DA_ALLOC);
 			break;
 		case Opt_auto_da_alloc:
-			if (match_int(&args[0], &option)) {
-				clear_opt(sbi->s_mount_opt, NO_AUTO_DA_ALLOC);
-				break;
-			}
+			if (args[0].from) {
+				if (match_int(&args[0], &option))
+					return 0;
+			} else
+				option = 1;	/* No argument, default to 1 */
 			if (option)
 				clear_opt(sbi->s_mount_opt, NO_AUTO_DA_ALLOC);
 			else
@@ -1609,6 +1674,12 @@ set_qf_format:
 		case Opt_nodiscard:
 			clear_opt(sbi->s_mount_opt, DISCARD);
 			break;
+		case Opt_dioread_nolock:
+			set_opt(sbi->s_mount_opt, DIOREAD_NOLOCK);
+			break;
+		case Opt_dioread_lock:
+			clear_opt(sbi->s_mount_opt, DIOREAD_NOLOCK);
+			break;
 		default:
 			ext4_msg(sb, KERN_ERR,
 			       "Unrecognized mount option \"%s\" "
@@ -1618,18 +1689,13 @@ set_qf_format:
 	}
 #ifdef CONFIG_QUOTA
 	if (sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) {
-		if ((sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA) &&
-		     sbi->s_qf_names[USRQUOTA])
+		if (test_opt(sb, USRQUOTA) && sbi->s_qf_names[USRQUOTA])
 			clear_opt(sbi->s_mount_opt, USRQUOTA);
 
-		if ((sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA) &&
-		     sbi->s_qf_names[GRPQUOTA])
+		if (test_opt(sb, GRPQUOTA) && sbi->s_qf_names[GRPQUOTA])
 			clear_opt(sbi->s_mount_opt, GRPQUOTA);
 
-		if ((sbi->s_qf_names[USRQUOTA] &&
-				(sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA)) ||
-		    (sbi->s_qf_names[GRPQUOTA] &&
-				(sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA))) {
+		if (test_opt(sb, GRPQUOTA) || test_opt(sb, USRQUOTA)) {
 			ext4_msg(sb, KERN_ERR, "old and new quota "
 					"format mixing");
 			return 0;
@@ -1939,7 +2005,7 @@ static void ext4_orphan_cleanup(struct super_block *sb,
 		}
 
 		list_add(&EXT4_I(inode)->i_orphan, &EXT4_SB(sb)->s_orphan);
-		vfs_dq_init(inode);
+		dquot_initialize(inode);
 		if (inode->i_nlink) {
 			ext4_msg(sb, KERN_DEBUG,
 				"%s: truncating inode %lu to %lld bytes",
@@ -2432,8 +2498,11 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	def_mount_opts = le32_to_cpu(es->s_default_mount_opts);
 	if (def_mount_opts & EXT4_DEFM_DEBUG)
 		set_opt(sbi->s_mount_opt, DEBUG);
-	if (def_mount_opts & EXT4_DEFM_BSDGROUPS)
+	if (def_mount_opts & EXT4_DEFM_BSDGROUPS) {
+		ext4_msg(sb, KERN_WARNING, deprecated_msg, "bsdgroups",
+			"2.6.38");
 		set_opt(sbi->s_mount_opt, GRPID);
+	}
 	if (def_mount_opts & EXT4_DEFM_UID16)
 		set_opt(sbi->s_mount_opt, NO_UID32);
 #ifdef CONFIG_EXT4_FS_XATTR
@@ -2445,11 +2514,11 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 		set_opt(sbi->s_mount_opt, POSIX_ACL);
 #endif
 	if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_DATA)
-		sbi->s_mount_opt |= EXT4_MOUNT_JOURNAL_DATA;
+		set_opt(sbi->s_mount_opt, JOURNAL_DATA);
 	else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_ORDERED)
-		sbi->s_mount_opt |= EXT4_MOUNT_ORDERED_DATA;
+		set_opt(sbi->s_mount_opt, ORDERED_DATA);
 	else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_WBACK)
-		sbi->s_mount_opt |= EXT4_MOUNT_WRITEBACK_DATA;
+		set_opt(sbi->s_mount_opt, WRITEBACK_DATA);
 
 	if (le16_to_cpu(sbi->s_es->s_errors) == EXT4_ERRORS_PANIC)
 		set_opt(sbi->s_mount_opt, ERRORS_PANIC);
@@ -2477,7 +2546,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 		goto failed_mount;
 
 	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
-		((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+		(test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
 
 	if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV &&
 	    (EXT4_HAS_COMPAT_FEATURE(sb, ~0U) ||
@@ -2766,7 +2835,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	      EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER)) {
 		ext4_msg(sb, KERN_ERR, "required journal recovery "
 		       "suppressed and not mounted read-only");
-		goto failed_mount4;
+		goto failed_mount_wq;
 	} else {
 		clear_opt(sbi->s_mount_opt, DATA_FLAGS);
 		set_opt(sbi->s_mount_opt, WRITEBACK_DATA);
@@ -2779,7 +2848,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	    !jbd2_journal_set_features(EXT4_SB(sb)->s_journal, 0, 0,
 				       JBD2_FEATURE_INCOMPAT_64BIT)) {
 		ext4_msg(sb, KERN_ERR, "Failed to set 64-bit journal feature");
-		goto failed_mount4;
+		goto failed_mount_wq;
 	}
 
 	if (test_opt(sb, JOURNAL_ASYNC_COMMIT)) {
@@ -2818,7 +2887,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 		    (sbi->s_journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE)) {
 			ext4_msg(sb, KERN_ERR, "Journal does not support "
 			       "requested data journaling mode");
-			goto failed_mount4;
+			goto failed_mount_wq;
 		}
 	default:
 		break;
@@ -2826,13 +2895,17 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	set_task_ioprio(sbi->s_journal->j_task, journal_ioprio);
 
 no_journal:
-
 	if (test_opt(sb, NOBH)) {
 		if (!(test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)) {
 			ext4_msg(sb, KERN_WARNING, "Ignoring nobh option - "
 				"its supported only with writeback mode");
 			clear_opt(sbi->s_mount_opt, NOBH);
 		}
+		if (test_opt(sb, DIOREAD_NOLOCK)) {
+			ext4_msg(sb, KERN_WARNING, "dioread_nolock option is "
+				"not supported with nobh mode");
+			goto failed_mount_wq;
+		}
 	}
 	EXT4_SB(sb)->dio_unwritten_wq = create_workqueue("ext4-dio-unwritten");
 	if (!EXT4_SB(sb)->dio_unwritten_wq) {
@@ -2897,6 +2970,18 @@ no_journal:
 			 "requested data journaling mode");
 		clear_opt(sbi->s_mount_opt, DELALLOC);
 	}
+	if (test_opt(sb, DIOREAD_NOLOCK)) {
+		if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) {
+			ext4_msg(sb, KERN_WARNING, "Ignoring dioread_nolock "
+				"option - requested data journaling mode");
+			clear_opt(sbi->s_mount_opt, DIOREAD_NOLOCK);
+		}
+		if (sb->s_blocksize < PAGE_SIZE) {
+			ext4_msg(sb, KERN_WARNING, "Ignoring dioread_nolock "
+				"option - block size is too small");
+			clear_opt(sbi->s_mount_opt, DIOREAD_NOLOCK);
+		}
+	}
 
 	err = ext4_setup_system_zone(sb);
 	if (err) {
@@ -3360,10 +3445,9 @@ static void ext4_clear_journal_err(struct super_block *sb,
 		char nbuf[16];
 
 		errstr = ext4_decode_error(sb, j_errno, nbuf);
-		ext4_warning(sb, __func__, "Filesystem error recorded "
+		ext4_warning(sb, "Filesystem error recorded "
 			     "from previous mount: %s", errstr);
-		ext4_warning(sb, __func__, "Marking fs in need of "
-			     "filesystem check.");
+		ext4_warning(sb, "Marking fs in need of filesystem check.");
 
 		EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
 		es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
@@ -3514,7 +3598,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 		ext4_abort(sb, __func__, "Abort forced by user");
 
 	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
-		((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+		(test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
 
 	es = sbi->s_es;
 
@@ -3708,7 +3792,7 @@ static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf)
  * Process 1                         Process 2
  * ext4_create()                     quota_sync()
  *   jbd2_journal_start()                  write_dquot()
- *   vfs_dq_init()                         down(dqio_mutex)
+ *   dquot_initialize()                         down(dqio_mutex)
  *     down(dqio_mutex)                    jbd2_journal_start()
  *
  */
@@ -3917,9 +4001,7 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type,
 	ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb);
 	int err = 0;
 	int offset = off & (sb->s_blocksize - 1);
-	int tocopy;
 	int journal_quota = EXT4_SB(sb)->s_qf_names[type] != NULL;
-	size_t towrite = len;
 	struct buffer_head *bh;
 	handle_t *handle = journal_current_handle();
 
@@ -3929,52 +4011,53 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type,
 			(unsigned long long)off, (unsigned long long)len);
 		return -EIO;
 	}
+	/*
+	 * Since we account only one data block in transaction credits,
+	 * then it is impossible to cross a block boundary.
+	 */
+	if (sb->s_blocksize - offset < len) {
+		ext4_msg(sb, KERN_WARNING, "Quota write (off=%llu, len=%llu)"
+			" cancelled because not block aligned",
+			(unsigned long long)off, (unsigned long long)len);
+		return -EIO;
+	}
+
 	mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA);
-	while (towrite > 0) {
-		tocopy = sb->s_blocksize - offset < towrite ?
-				sb->s_blocksize - offset : towrite;
-		bh = ext4_bread(handle, inode, blk, 1, &err);
-		if (!bh)
+	bh = ext4_bread(handle, inode, blk, 1, &err);
+	if (!bh)
+		goto out;
+	if (journal_quota) {
+		err = ext4_journal_get_write_access(handle, bh);
+		if (err) {
+			brelse(bh);
 			goto out;
-		if (journal_quota) {
-			err = ext4_journal_get_write_access(handle, bh);
-			if (err) {
-				brelse(bh);
-				goto out;
-			}
 		}
-		lock_buffer(bh);
-		memcpy(bh->b_data+offset, data, tocopy);
-		flush_dcache_page(bh->b_page);
-		unlock_buffer(bh);
-		if (journal_quota)
-			err = ext4_handle_dirty_metadata(handle, NULL, bh);
-		else {
-			/* Always do at least ordered writes for quotas */
-			err = ext4_jbd2_file_inode(handle, inode);
-			mark_buffer_dirty(bh);
-		}
-		brelse(bh);
-		if (err)
-			goto out;
-		offset = 0;
-		towrite -= tocopy;
-		data += tocopy;
-		blk++;
 	}
+	lock_buffer(bh);
+	memcpy(bh->b_data+offset, data, len);
+	flush_dcache_page(bh->b_page);
+	unlock_buffer(bh);
+	if (journal_quota)
+		err = ext4_handle_dirty_metadata(handle, NULL, bh);
+	else {
+		/* Always do at least ordered writes for quotas */
+		err = ext4_jbd2_file_inode(handle, inode);
+		mark_buffer_dirty(bh);
+	}
+	brelse(bh);
 out:
-	if (len == towrite) {
+	if (err) {
 		mutex_unlock(&inode->i_mutex);
 		return err;
 	}
-	if (inode->i_size < off+len-towrite) {
-		i_size_write(inode, off+len-towrite);
+	if (inode->i_size < off + len) {
+		i_size_write(inode, off + len);
 		EXT4_I(inode)->i_disksize = inode->i_size;
 	}
 	inode->i_mtime = inode->i_ctime = CURRENT_TIME;
 	ext4_mark_inode_dirty(handle, inode);
 	mutex_unlock(&inode->i_mutex);
-	return len - towrite;
+	return len;
 }
 
 #endif
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index f3a2f7ed45aa..b4c5aa8489d8 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -227,7 +227,8 @@ ext4_xattr_block_get(struct inode *inode, int name_index, const char *name,
 	ea_bdebug(bh, "b_count=%d, refcount=%d",
 		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
 	if (ext4_xattr_check_block(bh)) {
-bad_block:	ext4_error(inode->i_sb, __func__,
+bad_block:
+		ext4_error(inode->i_sb,
 			   "inode %lu: bad block %llu", inode->i_ino,
 			   EXT4_I(inode)->i_file_acl);
 		error = -EIO;
@@ -267,7 +268,7 @@ ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name,
 	void *end;
 	int error;
 
-	if (!(EXT4_I(inode)->i_state & EXT4_STATE_XATTR))
+	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
 		return -ENODATA;
 	error = ext4_get_inode_loc(inode, &iloc);
 	if (error)
@@ -371,7 +372,7 @@ ext4_xattr_block_list(struct dentry *dentry, char *buffer, size_t buffer_size)
 	ea_bdebug(bh, "b_count=%d, refcount=%d",
 		atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount));
 	if (ext4_xattr_check_block(bh)) {
-		ext4_error(inode->i_sb, __func__,
+		ext4_error(inode->i_sb,
 			   "inode %lu: bad block %llu", inode->i_ino,
 			   EXT4_I(inode)->i_file_acl);
 		error = -EIO;
@@ -396,7 +397,7 @@ ext4_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size)
 	void *end;
 	int error;
 
-	if (!(EXT4_I(inode)->i_state & EXT4_STATE_XATTR))
+	if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR))
 		return 0;
 	error = ext4_get_inode_loc(inode, &iloc);
 	if (error)
@@ -494,7 +495,7 @@ ext4_xattr_release_block(handle_t *handle, struct inode *inode,
 		error = ext4_handle_dirty_metadata(handle, inode, bh);
 		if (IS_SYNC(inode))
 			ext4_handle_sync(handle);
-		vfs_dq_free_block(inode, 1);
+		dquot_free_block(inode, 1);
 		ea_bdebug(bh, "refcount now=%d; releasing",
 			  le32_to_cpu(BHDR(bh)->h_refcount));
 		if (ce)
@@ -665,9 +666,8 @@ ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
 			atomic_read(&(bs->bh->b_count)),
 			le32_to_cpu(BHDR(bs->bh)->h_refcount));
 		if (ext4_xattr_check_block(bs->bh)) {
-			ext4_error(sb, __func__,
-				"inode %lu: bad block %llu", inode->i_ino,
-				EXT4_I(inode)->i_file_acl);
+			ext4_error(sb, "inode %lu: bad block %llu",
+				   inode->i_ino, EXT4_I(inode)->i_file_acl);
 			error = -EIO;
 			goto cleanup;
 		}
@@ -787,8 +787,8 @@ inserted:
 			else {
 				/* The old block is released after updating
 				   the inode. */
-				error = -EDQUOT;
-				if (vfs_dq_alloc_block(inode, 1))
+				error = dquot_alloc_block(inode, 1);
+				if (error)
 					goto cleanup;
 				error = ext4_journal_get_write_access(handle,
 								      new_bh);
@@ -876,13 +876,12 @@ cleanup:
 	return error;
 
 cleanup_dquot:
-	vfs_dq_free_block(inode, 1);
+	dquot_free_block(inode, 1);
 	goto cleanup;
 
 bad_block:
-	ext4_error(inode->i_sb, __func__,
-		   "inode %lu: bad block %llu", inode->i_ino,
-		   EXT4_I(inode)->i_file_acl);
+	ext4_error(inode->i_sb, "inode %lu: bad block %llu",
+		   inode->i_ino, EXT4_I(inode)->i_file_acl);
 	goto cleanup;
 
 #undef header
@@ -908,7 +907,7 @@ ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i,
 	is->s.base = is->s.first = IFIRST(header);
 	is->s.here = is->s.first;
 	is->s.end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
-	if (EXT4_I(inode)->i_state & EXT4_STATE_XATTR) {
+	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
 		error = ext4_xattr_check_names(IFIRST(header), is->s.end);
 		if (error)
 			return error;
@@ -940,10 +939,10 @@ ext4_xattr_ibody_set(handle_t *handle, struct inode *inode,
 	header = IHDR(inode, ext4_raw_inode(&is->iloc));
 	if (!IS_LAST_ENTRY(s->first)) {
 		header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC);
-		EXT4_I(inode)->i_state |= EXT4_STATE_XATTR;
+		ext4_set_inode_state(inode, EXT4_STATE_XATTR);
 	} else {
 		header->h_magic = cpu_to_le32(0);
-		EXT4_I(inode)->i_state &= ~EXT4_STATE_XATTR;
+		ext4_clear_inode_state(inode, EXT4_STATE_XATTR);
 	}
 	return 0;
 }
@@ -986,8 +985,8 @@ ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
 	if (strlen(name) > 255)
 		return -ERANGE;
 	down_write(&EXT4_I(inode)->xattr_sem);
-	no_expand = EXT4_I(inode)->i_state & EXT4_STATE_NO_EXPAND;
-	EXT4_I(inode)->i_state |= EXT4_STATE_NO_EXPAND;
+	no_expand = ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND);
+	ext4_set_inode_state(inode, EXT4_STATE_NO_EXPAND);
 
 	error = ext4_get_inode_loc(inode, &is.iloc);
 	if (error)
@@ -997,10 +996,10 @@ ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
 	if (error)
 		goto cleanup;
 
-	if (EXT4_I(inode)->i_state & EXT4_STATE_NEW) {
+	if (ext4_test_inode_state(inode, EXT4_STATE_NEW)) {
 		struct ext4_inode *raw_inode = ext4_raw_inode(&is.iloc);
 		memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size);
-		EXT4_I(inode)->i_state &= ~EXT4_STATE_NEW;
+		ext4_clear_inode_state(inode, EXT4_STATE_NEW);
 	}
 
 	error = ext4_xattr_ibody_find(inode, &i, &is);
@@ -1052,7 +1051,7 @@ ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
 		ext4_xattr_update_super_block(handle, inode->i_sb);
 		inode->i_ctime = ext4_current_time(inode);
 		if (!value)
-			EXT4_I(inode)->i_state &= ~EXT4_STATE_NO_EXPAND;
+			ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
 		error = ext4_mark_iloc_dirty(handle, inode, &is.iloc);
 		/*
 		 * The bh is consumed by ext4_mark_iloc_dirty, even with
@@ -1067,7 +1066,7 @@ cleanup:
 	brelse(is.iloc.bh);
 	brelse(bs.bh);
 	if (no_expand == 0)
-		EXT4_I(inode)->i_state &= ~EXT4_STATE_NO_EXPAND;
+		ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND);
 	up_write(&EXT4_I(inode)->xattr_sem);
 	return error;
 }
@@ -1195,9 +1194,8 @@ retry:
 		if (!bh)
 			goto cleanup;
 		if (ext4_xattr_check_block(bh)) {
-			ext4_error(inode->i_sb, __func__,
-				"inode %lu: bad block %llu", inode->i_ino,
-				EXT4_I(inode)->i_file_acl);
+			ext4_error(inode->i_sb, "inode %lu: bad block %llu",
+				   inode->i_ino, EXT4_I(inode)->i_file_acl);
 			error = -EIO;
 			goto cleanup;
 		}
@@ -1302,6 +1300,8 @@ retry:
 
 		/* Remove the chosen entry from the inode */
 		error = ext4_xattr_ibody_set(handle, inode, &i, is);
+		if (error)
+			goto cleanup;
 
 		entry = IFIRST(header);
 		if (entry_size + EXT4_XATTR_SIZE(size) >= new_extra_isize)
@@ -1372,16 +1372,14 @@ ext4_xattr_delete_inode(handle_t *handle, struct inode *inode)
 		goto cleanup;
 	bh = sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl);
 	if (!bh) {
-		ext4_error(inode->i_sb, __func__,
-			"inode %lu: block %llu read error", inode->i_ino,
-			EXT4_I(inode)->i_file_acl);
+		ext4_error(inode->i_sb, "inode %lu: block %llu read error",
+			   inode->i_ino, EXT4_I(inode)->i_file_acl);
 		goto cleanup;
 	}
 	if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) ||
 	    BHDR(bh)->h_blocks != cpu_to_le32(1)) {
-		ext4_error(inode->i_sb, __func__,
-			"inode %lu: bad block %llu", inode->i_ino,
-			EXT4_I(inode)->i_file_acl);
+		ext4_error(inode->i_sb, "inode %lu: bad block %llu",
+			   inode->i_ino, EXT4_I(inode)->i_file_acl);
 		goto cleanup;
 	}
 	ext4_xattr_release_block(handle, inode, bh);
@@ -1506,7 +1504,7 @@ again:
 		}
 		bh = sb_bread(inode->i_sb, ce->e_block);
 		if (!bh) {
-			ext4_error(inode->i_sb, __func__,
+			ext4_error(inode->i_sb,
 				"inode %lu: block %lu read error",
 				inode->i_ino, (unsigned long) ce->e_block);
 		} else if (le32_to_cpu(BHDR(bh)->h_refcount) >=
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index 14da530b05ca..fbeecdc194dc 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -577,7 +577,7 @@ static inline loff_t fat_i_pos_read(struct msdos_sb_info *sbi,
 	return i_pos;
 }
 
-static int fat_write_inode(struct inode *inode, int wait)
+static int __fat_write_inode(struct inode *inode, int wait)
 {
 	struct super_block *sb = inode->i_sb;
 	struct msdos_sb_info *sbi = MSDOS_SB(sb);
@@ -634,9 +634,14 @@ retry:
 	return err;
 }
 
+static int fat_write_inode(struct inode *inode, struct writeback_control *wbc)
+{
+	return __fat_write_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
+}
+
 int fat_sync_inode(struct inode *inode)
 {
-	return fat_write_inode(inode, 1);
+	return __fat_write_inode(inode, 1);
 }
 
 EXPORT_SYMBOL_GPL(fat_sync_inode);
diff --git a/fs/fcntl.c b/fs/fcntl.c
index 97e01dc0d95f..452d02f9075e 100644
--- a/fs/fcntl.c
+++ b/fs/fcntl.c
@@ -344,7 +344,7 @@ static long do_fcntl(int fd, unsigned int cmd, unsigned long arg,
 	switch (cmd) {
 	case F_DUPFD:
 	case F_DUPFD_CLOEXEC:
-		if (arg >= current->signal->rlim[RLIMIT_NOFILE].rlim_cur)
+		if (arg >= rlimit(RLIMIT_NOFILE))
 			break;
 		err = alloc_fd(arg, cmd == F_DUPFD_CLOEXEC ? O_CLOEXEC : 0);
 		if (err >= 0) {
diff --git a/fs/file.c b/fs/file.c
index 87e129030ab1..34bb7f71d994 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -257,7 +257,7 @@ int expand_files(struct files_struct *files, int nr)
 	 * N.B. For clone tasks sharing a files structure, this test
 	 * will limit the total number of files that can be opened.
 	 */
-	if (nr >= current->signal->rlim[RLIMIT_NOFILE].rlim_cur)
+	if (nr >= rlimit(RLIMIT_NOFILE))
 		return -EMFILE;
 
 	/* Do we need to expand? */
@@ -478,7 +478,7 @@ repeat:
 	error = fd;
 #if 1
 	/* Sanity check */
-	if (rcu_dereference(fdt->fd[fd]) != NULL) {
+	if (rcu_dereference_raw(fdt->fd[fd]) != NULL) {
 		printk(KERN_WARNING "alloc_fd: slot %d not NULL!\n", fd);
 		rcu_assign_pointer(fdt->fd[fd], NULL);
 	}
diff --git a/fs/file_table.c b/fs/file_table.c
index b98404b54383..32d12b78bac8 100644
--- a/fs/file_table.c
+++ b/fs/file_table.c
@@ -393,7 +393,9 @@ retry:
 			continue;
 		if (!(f->f_mode & FMODE_WRITE))
 			continue;
+		spin_lock(&f->f_lock);
 		f->f_mode &= ~FMODE_WRITE;
+		spin_unlock(&f->f_lock);
 		if (file_check_writeable(f) != 0)
 			continue;
 		file_release_write(f);
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 1a7c42c64ff4..76fc4d594acb 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -381,10 +381,10 @@ static void queue_io(struct bdi_writeback *wb, unsigned long *older_than_this)
 	move_expired_inodes(&wb->b_dirty, &wb->b_io, older_than_this);
 }
 
-static int write_inode(struct inode *inode, int sync)
+static int write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	if (inode->i_sb->s_op->write_inode && !is_bad_inode(inode))
-		return inode->i_sb->s_op->write_inode(inode, sync);
+		return inode->i_sb->s_op->write_inode(inode, wbc);
 	return 0;
 }
 
@@ -421,7 +421,6 @@ static int
 writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct address_space *mapping = inode->i_mapping;
-	int wait = wbc->sync_mode == WB_SYNC_ALL;
 	unsigned dirty;
 	int ret;
 
@@ -439,7 +438,7 @@ writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
 		 * We'll have another go at writing back this inode when we
 		 * completed a full scan of b_io.
 		 */
-		if (!wait) {
+		if (wbc->sync_mode != WB_SYNC_ALL) {
 			requeue_io(inode);
 			return 0;
 		}
@@ -461,15 +460,20 @@ writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
 
 	ret = do_writepages(mapping, wbc);
 
-	/* Don't write the inode if only I_DIRTY_PAGES was set */
-	if (dirty & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) {
-		int err = write_inode(inode, wait);
+	/*
+	 * Make sure to wait on the data before writing out the metadata.
+	 * This is important for filesystems that modify metadata on data
+	 * I/O completion.
+	 */
+	if (wbc->sync_mode == WB_SYNC_ALL) {
+		int err = filemap_fdatawait(mapping);
 		if (ret == 0)
 			ret = err;
 	}
 
-	if (wait) {
-		int err = filemap_fdatawait(mapping);
+	/* Don't write the inode if only I_DIRTY_PAGES was set */
+	if (dirty & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) {
+		int err = write_inode(inode, wbc);
 		if (ret == 0)
 			ret = err;
 	}
diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
index 51d9e33d634f..eb7e9423691f 100644
--- a/fs/fuse/dev.c
+++ b/fs/fuse/dev.c
@@ -865,13 +865,10 @@ static int fuse_notify_inval_inode(struct fuse_conn *fc, unsigned int size,
 
 	down_read(&fc->killsb);
 	err = -ENOENT;
-	if (!fc->sb)
-		goto err_unlock;
-
-	err = fuse_reverse_inval_inode(fc->sb, outarg.ino,
-				       outarg.off, outarg.len);
-
-err_unlock:
+	if (fc->sb) {
+		err = fuse_reverse_inval_inode(fc->sb, outarg.ino,
+					       outarg.off, outarg.len);
+	}
 	up_read(&fc->killsb);
 	return err;
 
@@ -884,10 +881,15 @@ static int fuse_notify_inval_entry(struct fuse_conn *fc, unsigned int size,
 				   struct fuse_copy_state *cs)
 {
 	struct fuse_notify_inval_entry_out outarg;
-	int err = -EINVAL;
-	char buf[FUSE_NAME_MAX+1];
+	int err = -ENOMEM;
+	char *buf;
 	struct qstr name;
 
+	buf = kzalloc(FUSE_NAME_MAX + 1, GFP_KERNEL);
+	if (!buf)
+		goto err;
+
+	err = -EINVAL;
 	if (size < sizeof(outarg))
 		goto err;
 
@@ -910,16 +912,14 @@ static int fuse_notify_inval_entry(struct fuse_conn *fc, unsigned int size,
 
 	down_read(&fc->killsb);
 	err = -ENOENT;
-	if (!fc->sb)
-		goto err_unlock;
-
-	err = fuse_reverse_inval_entry(fc->sb, outarg.parent, &name);
-
-err_unlock:
+	if (fc->sb)
+		err = fuse_reverse_inval_entry(fc->sb, outarg.parent, &name);
 	up_read(&fc->killsb);
+	kfree(buf);
 	return err;
 
 err:
+	kfree(buf);
 	fuse_copy_finish(cs);
 	return err;
 }
diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 7b8da9415267..0c1d0b82dcf1 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -1061,8 +1061,8 @@ out:
 
 int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 {
-	struct inode *aspace = page->mapping->host;
-	struct gfs2_sbd *sdp = aspace->i_sb->s_fs_info;
+	struct address_space *mapping = page->mapping;
+	struct gfs2_sbd *sdp = gfs2_mapping2sbd(mapping);
 	struct buffer_head *bh, *head;
 	struct gfs2_bufdata *bd;
 
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index f42663325931..454d4b4eb36b 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -19,7 +19,6 @@
 #include <linux/list.h>
 #include <linux/wait.h>
 #include <linux/module.h>
-#include <linux/rwsem.h>
 #include <asm/uaccess.h>
 #include <linux/seq_file.h>
 #include <linux/debugfs.h>
@@ -60,7 +59,6 @@ static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl);
 #define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { __dump_glock(NULL, gl); BUG(); } } while(0)
 static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int target);
 
-static DECLARE_RWSEM(gfs2_umount_flush_sem);
 static struct dentry *gfs2_root;
 static struct workqueue_struct *glock_workqueue;
 struct workqueue_struct *gfs2_delete_workqueue;
@@ -154,12 +152,14 @@ static unsigned int gl_hash(const struct gfs2_sbd *sdp,
 static void glock_free(struct gfs2_glock *gl)
 {
 	struct gfs2_sbd *sdp = gl->gl_sbd;
-	struct inode *aspace = gl->gl_aspace;
+	struct address_space *mapping = gfs2_glock2aspace(gl);
+	struct kmem_cache *cachep = gfs2_glock_cachep;
 
-	if (aspace)
-		gfs2_aspace_put(aspace);
+	GLOCK_BUG_ON(gl, mapping && mapping->nrpages);
 	trace_gfs2_glock_put(gl);
-	sdp->sd_lockstruct.ls_ops->lm_put_lock(gfs2_glock_cachep, gl);
+	if (mapping)
+		cachep = gfs2_glock_aspace_cachep;
+	sdp->sd_lockstruct.ls_ops->lm_put_lock(cachep, gl);
 }
 
 /**
@@ -712,7 +712,6 @@ static void glock_work_func(struct work_struct *work)
 		finish_xmote(gl, gl->gl_reply);
 		drop_ref = 1;
 	}
-	down_read(&gfs2_umount_flush_sem);
 	spin_lock(&gl->gl_spin);
 	if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
 	    gl->gl_state != LM_ST_UNLOCKED &&
@@ -725,7 +724,6 @@ static void glock_work_func(struct work_struct *work)
 	}
 	run_queue(gl, 0);
 	spin_unlock(&gl->gl_spin);
-	up_read(&gfs2_umount_flush_sem);
 	if (!delay ||
 	    queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
 		gfs2_glock_put(gl);
@@ -750,10 +748,11 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 		   const struct gfs2_glock_operations *glops, int create,
 		   struct gfs2_glock **glp)
 {
+	struct super_block *s = sdp->sd_vfs;
 	struct lm_lockname name = { .ln_number = number, .ln_type = glops->go_type };
 	struct gfs2_glock *gl, *tmp;
 	unsigned int hash = gl_hash(sdp, &name);
-	int error;
+	struct address_space *mapping;
 
 	read_lock(gl_lock_addr(hash));
 	gl = search_bucket(hash, sdp, &name);
@@ -765,7 +764,10 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 	if (!create)
 		return -ENOENT;
 
-	gl = kmem_cache_alloc(gfs2_glock_cachep, GFP_KERNEL);
+	if (glops->go_flags & GLOF_ASPACE)
+		gl = kmem_cache_alloc(gfs2_glock_aspace_cachep, GFP_KERNEL);
+	else
+		gl = kmem_cache_alloc(gfs2_glock_cachep, GFP_KERNEL);
 	if (!gl)
 		return -ENOMEM;
 
@@ -784,18 +786,18 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 	gl->gl_tchange = jiffies;
 	gl->gl_object = NULL;
 	gl->gl_sbd = sdp;
-	gl->gl_aspace = NULL;
 	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
 	INIT_WORK(&gl->gl_delete, delete_work_func);
 
-	/* If this glock protects actual on-disk data or metadata blocks,
-	   create a VFS inode to manage the pages/buffers holding them. */
-	if (glops == &gfs2_inode_glops || glops == &gfs2_rgrp_glops) {
-		gl->gl_aspace = gfs2_aspace_get(sdp);
-		if (!gl->gl_aspace) {
-			error = -ENOMEM;
-			goto fail;
-		}
+	mapping = gfs2_glock2aspace(gl);
+	if (mapping) {
+                mapping->a_ops = &gfs2_meta_aops;
+		mapping->host = s->s_bdev->bd_inode;
+		mapping->flags = 0;
+		mapping_set_gfp_mask(mapping, GFP_NOFS);
+		mapping->assoc_mapping = NULL;
+		mapping->backing_dev_info = s->s_bdi;
+		mapping->writeback_index = 0;
 	}
 
 	write_lock(gl_lock_addr(hash));
@@ -812,10 +814,6 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 	*glp = gl;
 
 	return 0;
-
-fail:
-	kmem_cache_free(gfs2_glock_cachep, gl);
-	return error;
 }
 
 /**
@@ -1510,35 +1508,10 @@ void gfs2_glock_thaw(struct gfs2_sbd *sdp)
 
 void gfs2_gl_hash_clear(struct gfs2_sbd *sdp)
 {
-	unsigned long t;
 	unsigned int x;
-	int cont;
 
-	t = jiffies;
-
-	for (;;) {
-		cont = 0;
-		for (x = 0; x < GFS2_GL_HASH_SIZE; x++) {
-			if (examine_bucket(clear_glock, sdp, x))
-				cont = 1;
-		}
-
-		if (!cont)
-			break;
-
-		if (time_after_eq(jiffies,
-				  t + gfs2_tune_get(sdp, gt_stall_secs) * HZ)) {
-			fs_warn(sdp, "Unmount seems to be stalled. "
-				     "Dumping lock state...\n");
-			gfs2_dump_lockstate(sdp);
-			t = jiffies;
-		}
-
-		down_write(&gfs2_umount_flush_sem);
-		invalidate_inodes(sdp->sd_vfs);
-		up_write(&gfs2_umount_flush_sem);
-		msleep(10);
-	}
+	for (x = 0; x < GFS2_GL_HASH_SIZE; x++)
+		examine_bucket(clear_glock, sdp, x);
 	flush_workqueue(glock_workqueue);
 	wait_event(sdp->sd_glock_wait, atomic_read(&sdp->sd_glock_disposal) == 0);
 	gfs2_dump_lockstate(sdp);
@@ -1685,7 +1658,7 @@ static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl)
 	dtime *= 1000000/HZ; /* demote time in uSec */
 	if (!test_bit(GLF_DEMOTE, &gl->gl_flags))
 		dtime = 0;
-	gfs2_print_dbg(seq, "G:  s:%s n:%u/%llu f:%s t:%s d:%s/%llu a:%d r:%d\n",
+	gfs2_print_dbg(seq, "G:  s:%s n:%u/%llx f:%s t:%s d:%s/%llu a:%d r:%d\n",
 		  state2str(gl->gl_state),
 		  gl->gl_name.ln_type,
 		  (unsigned long long)gl->gl_name.ln_number,
diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h
index c0262faf4725..2bda1911b156 100644
--- a/fs/gfs2/glock.h
+++ b/fs/gfs2/glock.h
@@ -180,6 +180,13 @@ static inline int gfs2_glock_is_held_shrd(struct gfs2_glock *gl)
 	return gl->gl_state == LM_ST_SHARED;
 }
 
+static inline struct address_space *gfs2_glock2aspace(struct gfs2_glock *gl)
+{
+	if (gl->gl_ops->go_flags & GLOF_ASPACE)
+		return (struct address_space *)(gl + 1);
+	return NULL;
+}
+
 int gfs2_glock_get(struct gfs2_sbd *sdp,
 		   u64 number, const struct gfs2_glock_operations *glops,
 		   int create, struct gfs2_glock **glp);
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 78554acc0605..38e3749d476c 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -87,7 +87,7 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 
 static void rgrp_go_sync(struct gfs2_glock *gl)
 {
-	struct address_space *metamapping = gl->gl_aspace->i_mapping;
+	struct address_space *metamapping = gfs2_glock2aspace(gl);
 	int error;
 
 	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
@@ -113,7 +113,7 @@ static void rgrp_go_sync(struct gfs2_glock *gl)
 
 static void rgrp_go_inval(struct gfs2_glock *gl, int flags)
 {
-	struct address_space *mapping = gl->gl_aspace->i_mapping;
+	struct address_space *mapping = gfs2_glock2aspace(gl);
 
 	BUG_ON(!(flags & DIO_METADATA));
 	gfs2_assert_withdraw(gl->gl_sbd, !atomic_read(&gl->gl_ail_count));
@@ -134,7 +134,7 @@ static void rgrp_go_inval(struct gfs2_glock *gl, int flags)
 static void inode_go_sync(struct gfs2_glock *gl)
 {
 	struct gfs2_inode *ip = gl->gl_object;
-	struct address_space *metamapping = gl->gl_aspace->i_mapping;
+	struct address_space *metamapping = gfs2_glock2aspace(gl);
 	int error;
 
 	if (ip && !S_ISREG(ip->i_inode.i_mode))
@@ -183,7 +183,7 @@ static void inode_go_inval(struct gfs2_glock *gl, int flags)
 	gfs2_assert_withdraw(gl->gl_sbd, !atomic_read(&gl->gl_ail_count));
 
 	if (flags & DIO_METADATA) {
-		struct address_space *mapping = gl->gl_aspace->i_mapping;
+		struct address_space *mapping = gfs2_glock2aspace(gl);
 		truncate_inode_pages(mapping, 0);
 		if (ip) {
 			set_bit(GIF_INVALID, &ip->i_flags);
@@ -282,7 +282,8 @@ static int inode_go_dump(struct seq_file *seq, const struct gfs2_glock *gl)
 
 static int rgrp_go_demote_ok(const struct gfs2_glock *gl)
 {
-	return !gl->gl_aspace->i_mapping->nrpages;
+	const struct address_space *mapping = (const struct address_space *)(gl + 1);
+	return !mapping->nrpages;
 }
 
 /**
@@ -387,8 +388,7 @@ static void iopen_go_callback(struct gfs2_glock *gl)
 	struct gfs2_inode *ip = (struct gfs2_inode *)gl->gl_object;
 
 	if (gl->gl_demote_state == LM_ST_UNLOCKED &&
-	    gl->gl_state == LM_ST_SHARED &&
-	    ip && test_bit(GIF_USER, &ip->i_flags)) {
+	    gl->gl_state == LM_ST_SHARED && ip) {
 		gfs2_glock_hold(gl);
 		if (queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0)
 			gfs2_glock_put_nolock(gl);
@@ -407,6 +407,7 @@ const struct gfs2_glock_operations gfs2_inode_glops = {
 	.go_dump = inode_go_dump,
 	.go_type = LM_TYPE_INODE,
 	.go_min_hold_time = HZ / 5,
+	.go_flags = GLOF_ASPACE,
 };
 
 const struct gfs2_glock_operations gfs2_rgrp_glops = {
@@ -418,6 +419,7 @@ const struct gfs2_glock_operations gfs2_rgrp_glops = {
 	.go_dump = gfs2_rgrp_dump,
 	.go_type = LM_TYPE_RGRP,
 	.go_min_hold_time = HZ / 5,
+	.go_flags = GLOF_ASPACE,
 };
 
 const struct gfs2_glock_operations gfs2_trans_glops = {
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index bc0ad158e6b4..b8025e51cabf 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -162,6 +162,8 @@ struct gfs2_glock_operations {
 	void (*go_callback) (struct gfs2_glock *gl);
 	const int go_type;
 	const unsigned long go_min_hold_time;
+	const unsigned long go_flags;
+#define GLOF_ASPACE 1
 };
 
 enum {
@@ -225,7 +227,6 @@ struct gfs2_glock {
 
 	struct gfs2_sbd *gl_sbd;
 
-	struct inode *gl_aspace;
 	struct list_head gl_ail_list;
 	atomic_t gl_ail_count;
 	struct delayed_work gl_work;
@@ -258,7 +259,6 @@ enum {
 	GIF_INVALID		= 0,
 	GIF_QD_LOCKED		= 1,
 	GIF_SW_PAGED		= 3,
-	GIF_USER                = 4, /* user inode, not metadata addr space */
 };
 
 
@@ -451,7 +451,6 @@ struct gfs2_tune {
 	unsigned int gt_quota_quantum; /* Secs between syncs to quota file */
 	unsigned int gt_new_files_jdata;
 	unsigned int gt_max_readahead; /* Max bytes to read-ahead from disk */
-	unsigned int gt_stall_secs; /* Detects trouble! */
 	unsigned int gt_complain_secs;
 	unsigned int gt_statfs_quantum;
 	unsigned int gt_statfs_slow;
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 6e220f4eee7d..b1bf2694fb2b 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -45,7 +45,7 @@ static int iget_test(struct inode *inode, void *opaque)
 	struct gfs2_inode *ip = GFS2_I(inode);
 	u64 *no_addr = opaque;
 
-	if (ip->i_no_addr == *no_addr && test_bit(GIF_USER, &ip->i_flags))
+	if (ip->i_no_addr == *no_addr)
 		return 1;
 
 	return 0;
@@ -58,7 +58,6 @@ static int iget_set(struct inode *inode, void *opaque)
 
 	inode->i_ino = (unsigned long)*no_addr;
 	ip->i_no_addr = *no_addr;
-	set_bit(GIF_USER, &ip->i_flags);
 	return 0;
 }
 
@@ -84,7 +83,7 @@ static int iget_skip_test(struct inode *inode, void *opaque)
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_skip_data *data = opaque;
 
-	if (ip->i_no_addr == data->no_addr && test_bit(GIF_USER, &ip->i_flags)){
+	if (ip->i_no_addr == data->no_addr) {
 		if (inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE)){
 			data->skipped = 1;
 			return 0;
@@ -103,7 +102,6 @@ static int iget_skip_set(struct inode *inode, void *opaque)
 		return 1;
 	inode->i_ino = (unsigned long)(data->no_addr);
 	ip->i_no_addr = data->no_addr;
-	set_bit(GIF_USER, &ip->i_flags);
 	return 0;
 }
 
diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c
index 0e5e0e7022e5..569b46240f61 100644
--- a/fs/gfs2/lock_dlm.c
+++ b/fs/gfs2/lock_dlm.c
@@ -30,7 +30,10 @@ static void gdlm_ast(void *arg)
 
 	switch (gl->gl_lksb.sb_status) {
 	case -DLM_EUNLOCK: /* Unlocked, so glock can be freed */
-		kmem_cache_free(gfs2_glock_cachep, gl);
+		if (gl->gl_ops->go_flags & GLOF_ASPACE)
+			kmem_cache_free(gfs2_glock_aspace_cachep, gl);
+		else
+			kmem_cache_free(gfs2_glock_cachep, gl);
 		if (atomic_dec_and_test(&sdp->sd_glock_disposal))
 			wake_up(&sdp->sd_glock_wait);
 		return;
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index de97632ba32f..adc260fbea90 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -528,9 +528,9 @@ static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 		gfs2_pin(sdp, bd->bd_bh);
 		tr->tr_num_databuf_new++;
 		sdp->sd_log_num_databuf++;
-		list_add(&le->le_list, &sdp->sd_log_le_databuf);
+		list_add_tail(&le->le_list, &sdp->sd_log_le_databuf);
 	} else {
-		list_add(&le->le_list, &sdp->sd_log_le_ordered);
+		list_add_tail(&le->le_list, &sdp->sd_log_le_ordered);
 	}
 out:
 	gfs2_log_unlock(sdp);
diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c
index 5b31f7741a8f..a88fadc704bb 100644
--- a/fs/gfs2/main.c
+++ b/fs/gfs2/main.c
@@ -52,6 +52,22 @@ static void gfs2_init_glock_once(void *foo)
 	atomic_set(&gl->gl_ail_count, 0);
 }
 
+static void gfs2_init_gl_aspace_once(void *foo)
+{
+	struct gfs2_glock *gl = foo;
+	struct address_space *mapping = (struct address_space *)(gl + 1);
+
+	gfs2_init_glock_once(gl);
+	memset(mapping, 0, sizeof(*mapping));
+	INIT_RADIX_TREE(&mapping->page_tree, GFP_ATOMIC);
+	spin_lock_init(&mapping->tree_lock);
+	spin_lock_init(&mapping->i_mmap_lock);
+	INIT_LIST_HEAD(&mapping->private_list);
+	spin_lock_init(&mapping->private_lock);
+	INIT_RAW_PRIO_TREE_ROOT(&mapping->i_mmap);
+	INIT_LIST_HEAD(&mapping->i_mmap_nonlinear);
+}
+
 /**
  * init_gfs2_fs - Register GFS2 as a filesystem
  *
@@ -78,6 +94,14 @@ static int __init init_gfs2_fs(void)
 	if (!gfs2_glock_cachep)
 		goto fail;
 
+	gfs2_glock_aspace_cachep = kmem_cache_create("gfs2_glock (aspace)",
+					sizeof(struct gfs2_glock) +
+					sizeof(struct address_space),
+					0, 0, gfs2_init_gl_aspace_once);
+
+	if (!gfs2_glock_aspace_cachep)
+		goto fail;
+
 	gfs2_inode_cachep = kmem_cache_create("gfs2_inode",
 					      sizeof(struct gfs2_inode),
 					      0,  SLAB_RECLAIM_ACCOUNT|
@@ -144,6 +168,9 @@ fail:
 	if (gfs2_inode_cachep)
 		kmem_cache_destroy(gfs2_inode_cachep);
 
+	if (gfs2_glock_aspace_cachep)
+		kmem_cache_destroy(gfs2_glock_aspace_cachep);
+
 	if (gfs2_glock_cachep)
 		kmem_cache_destroy(gfs2_glock_cachep);
 
@@ -169,6 +196,7 @@ static void __exit exit_gfs2_fs(void)
 	kmem_cache_destroy(gfs2_rgrpd_cachep);
 	kmem_cache_destroy(gfs2_bufdata_cachep);
 	kmem_cache_destroy(gfs2_inode_cachep);
+	kmem_cache_destroy(gfs2_glock_aspace_cachep);
 	kmem_cache_destroy(gfs2_glock_cachep);
 
 	gfs2_sys_uninit();
diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c
index 6f68a5f18eb8..0bb12c80937a 100644
--- a/fs/gfs2/meta_io.c
+++ b/fs/gfs2/meta_io.c
@@ -93,49 +93,13 @@ static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wb
 	return err;
 }
 
-static const struct address_space_operations aspace_aops = {
+const struct address_space_operations gfs2_meta_aops = {
 	.writepage = gfs2_aspace_writepage,
 	.releasepage = gfs2_releasepage,
 	.sync_page = block_sync_page,
 };
 
 /**
- * gfs2_aspace_get - Create and initialize a struct inode structure
- * @sdp: the filesystem the aspace is in
- *
- * Right now a struct inode is just a struct inode.  Maybe Linux
- * will supply a more lightweight address space construct (that works)
- * in the future.
- *
- * Make sure pages/buffers in this aspace aren't in high memory.
- *
- * Returns: the aspace
- */
-
-struct inode *gfs2_aspace_get(struct gfs2_sbd *sdp)
-{
-	struct inode *aspace;
-	struct gfs2_inode *ip;
-
-	aspace = new_inode(sdp->sd_vfs);
-	if (aspace) {
-		mapping_set_gfp_mask(aspace->i_mapping, GFP_NOFS);
-		aspace->i_mapping->a_ops = &aspace_aops;
-		aspace->i_size = MAX_LFS_FILESIZE;
-		ip = GFS2_I(aspace);
-		clear_bit(GIF_USER, &ip->i_flags);
-		insert_inode_hash(aspace);
-	}
-	return aspace;
-}
-
-void gfs2_aspace_put(struct inode *aspace)
-{
-	remove_inode_hash(aspace);
-	iput(aspace);
-}
-
-/**
  * gfs2_meta_sync - Sync all buffers associated with a glock
  * @gl: The glock
  *
@@ -143,7 +107,7 @@ void gfs2_aspace_put(struct inode *aspace)
 
 void gfs2_meta_sync(struct gfs2_glock *gl)
 {
-	struct address_space *mapping = gl->gl_aspace->i_mapping;
+	struct address_space *mapping = gfs2_glock2aspace(gl);
 	int error;
 
 	filemap_fdatawrite(mapping);
@@ -164,7 +128,7 @@ void gfs2_meta_sync(struct gfs2_glock *gl)
 
 struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, int create)
 {
-	struct address_space *mapping = gl->gl_aspace->i_mapping;
+	struct address_space *mapping = gfs2_glock2aspace(gl);
 	struct gfs2_sbd *sdp = gl->gl_sbd;
 	struct page *page;
 	struct buffer_head *bh;
@@ -344,8 +308,10 @@ void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh,
 
 void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int meta)
 {
-	struct gfs2_sbd *sdp = GFS2_SB(bh->b_page->mapping->host);
+	struct address_space *mapping = bh->b_page->mapping;
+	struct gfs2_sbd *sdp = gfs2_mapping2sbd(mapping);
 	struct gfs2_bufdata *bd = bh->b_private;
+
 	if (test_clear_buffer_pinned(bh)) {
 		list_del_init(&bd->bd_le.le_list);
 		if (meta) {
diff --git a/fs/gfs2/meta_io.h b/fs/gfs2/meta_io.h
index de270c2f9b63..6a1d9ba16411 100644
--- a/fs/gfs2/meta_io.h
+++ b/fs/gfs2/meta_io.h
@@ -37,8 +37,16 @@ static inline void gfs2_buffer_copy_tail(struct buffer_head *to_bh,
 	       0, from_head - to_head);
 }
 
-struct inode *gfs2_aspace_get(struct gfs2_sbd *sdp);
-void gfs2_aspace_put(struct inode *aspace);
+extern const struct address_space_operations gfs2_meta_aops;
+
+static inline struct gfs2_sbd *gfs2_mapping2sbd(struct address_space *mapping)
+{
+	struct inode *inode = mapping->host;
+	if (mapping->a_ops == &gfs2_meta_aops)
+		return (((struct gfs2_glock *)mapping) - 1)->gl_sbd;
+	else
+		return inode->i_sb->s_fs_info;
+}
 
 void gfs2_meta_sync(struct gfs2_glock *gl);
 
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index a86ed6381566..a054b526dc08 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -65,7 +65,6 @@ static void gfs2_tune_init(struct gfs2_tune *gt)
 	gt->gt_quota_scale_den = 1;
 	gt->gt_new_files_jdata = 0;
 	gt->gt_max_readahead = 1 << 18;
-	gt->gt_stall_secs = 600;
 	gt->gt_complain_secs = 10;
 }
 
@@ -1241,10 +1240,9 @@ fail_sb:
 fail_locking:
 	init_locking(sdp, &mount_gh, UNDO);
 fail_lm:
+	invalidate_inodes(sb);
 	gfs2_gl_hash_clear(sdp);
 	gfs2_lm_unmount(sdp);
-	while (invalidate_inodes(sb))
-		yield();
 fail_sys:
 	gfs2_sys_fs_del(sdp);
 fail:
diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c
index 84350e1be66d..4e64352d49de 100644
--- a/fs/gfs2/ops_inode.c
+++ b/fs/gfs2/ops_inode.c
@@ -976,122 +976,62 @@ out:
 }
 
 /**
- * gfs2_readlinki - return the contents of a symlink
- * @ip: the symlink's inode
- * @buf: a pointer to the buffer to be filled
- * @len: a pointer to the length of @buf
+ * gfs2_follow_link - Follow a symbolic link
+ * @dentry: The dentry of the link
+ * @nd: Data that we pass to vfs_follow_link()
  *
- * If @buf is too small, a piece of memory is kmalloc()ed and needs
- * to be freed by the caller.
+ * This can handle symlinks of any size.
  *
- * Returns: errno
+ * Returns: 0 on success or error code
  */
 
-static int gfs2_readlinki(struct gfs2_inode *ip, char **buf, unsigned int *len)
+static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd)
 {
+	struct gfs2_inode *ip = GFS2_I(dentry->d_inode);
 	struct gfs2_holder i_gh;
 	struct buffer_head *dibh;
 	unsigned int x;
+	char *buf;
 	int error;
 
 	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &i_gh);
 	error = gfs2_glock_nq(&i_gh);
 	if (error) {
 		gfs2_holder_uninit(&i_gh);
-		return error;
+		nd_set_link(nd, ERR_PTR(error));
+		return NULL;
 	}
 
 	if (!ip->i_disksize) {
 		gfs2_consist_inode(ip);
-		error = -EIO;
+		buf = ERR_PTR(-EIO);
 		goto out;
 	}
 
 	error = gfs2_meta_inode_buffer(ip, &dibh);
-	if (error)
+	if (error) {
+		buf = ERR_PTR(error);
 		goto out;
-
-	x = ip->i_disksize + 1;
-	if (x > *len) {
-		*buf = kmalloc(x, GFP_NOFS);
-		if (!*buf) {
-			error = -ENOMEM;
-			goto out_brelse;
-		}
 	}
 
-	memcpy(*buf, dibh->b_data + sizeof(struct gfs2_dinode), x);
-	*len = x;
-
-out_brelse:
+	x = ip->i_disksize + 1;
+	buf = kmalloc(x, GFP_NOFS);
+	if (!buf)
+		buf = ERR_PTR(-ENOMEM);
+	else
+		memcpy(buf, dibh->b_data + sizeof(struct gfs2_dinode), x);
 	brelse(dibh);
 out:
 	gfs2_glock_dq_uninit(&i_gh);
-	return error;
-}
-
-/**
- * gfs2_readlink - Read the value of a symlink
- * @dentry: the symlink
- * @buf: the buffer to read the symlink data into
- * @size: the size of the buffer
- *
- * Returns: errno
- */
-
-static int gfs2_readlink(struct dentry *dentry, char __user *user_buf,
-			 int user_size)
-{
-	struct gfs2_inode *ip = GFS2_I(dentry->d_inode);
-	char array[GFS2_FAST_NAME_SIZE], *buf = array;
-	unsigned int len = GFS2_FAST_NAME_SIZE;
-	int error;
-
-	error = gfs2_readlinki(ip, &buf, &len);
-	if (error)
-		return error;
-
-	if (user_size > len - 1)
-		user_size = len - 1;
-
-	if (copy_to_user(user_buf, buf, user_size))
-		error = -EFAULT;
-	else
-		error = user_size;
-
-	if (buf != array)
-		kfree(buf);
-
-	return error;
+	nd_set_link(nd, buf);
+	return NULL;
 }
 
-/**
- * gfs2_follow_link - Follow a symbolic link
- * @dentry: The dentry of the link
- * @nd: Data that we pass to vfs_follow_link()
- *
- * This can handle symlinks of any size. It is optimised for symlinks
- * under GFS2_FAST_NAME_SIZE.
- *
- * Returns: 0 on success or error code
- */
-
-static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd)
+static void gfs2_put_link(struct dentry *dentry, struct nameidata *nd, void *p)
 {
-	struct gfs2_inode *ip = GFS2_I(dentry->d_inode);
-	char array[GFS2_FAST_NAME_SIZE], *buf = array;
-	unsigned int len = GFS2_FAST_NAME_SIZE;
-	int error;
-
-	error = gfs2_readlinki(ip, &buf, &len);
-	if (!error) {
-		error = vfs_follow_link(nd, buf);
-		if (buf != array)
-			kfree(buf);
-	} else
-		path_put(&nd->path);
-
-	return ERR_PTR(error);
+	char *s = nd_get_link(nd);
+	if (!IS_ERR(s))
+		kfree(s);
 }
 
 /**
@@ -1426,8 +1366,9 @@ const struct inode_operations gfs2_dir_iops = {
 };
 
 const struct inode_operations gfs2_symlink_iops = {
-	.readlink = gfs2_readlink,
+	.readlink = generic_readlink,
 	.follow_link = gfs2_follow_link,
+	.put_link = gfs2_put_link,
 	.permission = gfs2_permission,
 	.setattr = gfs2_setattr,
 	.getattr = gfs2_getattr,
diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index e3bf6eab8750..6dbcbad6ab17 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -1083,7 +1083,7 @@ void gfs2_quota_change(struct gfs2_inode *ip, s64 change,
 	}
 }
 
-int gfs2_quota_sync(struct super_block *sb, int type)
+int gfs2_quota_sync(struct super_block *sb, int type, int wait)
 {
 	struct gfs2_sbd *sdp = sb->s_fs_info;
 	struct gfs2_quota_data **qda;
@@ -1127,6 +1127,11 @@ int gfs2_quota_sync(struct super_block *sb, int type)
 	return error;
 }
 
+static int gfs2_quota_sync_timeo(struct super_block *sb, int type)
+{
+	return gfs2_quota_sync(sb, type, 0);
+}
+
 int gfs2_quota_refresh(struct gfs2_sbd *sdp, int user, u32 id)
 {
 	struct gfs2_quota_data *qd;
@@ -1382,7 +1387,7 @@ int gfs2_quotad(void *data)
 					   &tune->gt_statfs_quantum);
 
 		/* Update quota file */
-		quotad_check_timeo(sdp, "sync", gfs2_quota_sync, t,
+		quotad_check_timeo(sdp, "sync", gfs2_quota_sync_timeo, t,
 				   &quotad_timeo, &tune->gt_quota_quantum);
 
 		/* Check for & recover partially truncated inodes */
diff --git a/fs/gfs2/quota.h b/fs/gfs2/quota.h
index e271fa07ad02..195f60c8bd14 100644
--- a/fs/gfs2/quota.h
+++ b/fs/gfs2/quota.h
@@ -25,7 +25,7 @@ extern int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid);
 extern void gfs2_quota_change(struct gfs2_inode *ip, s64 change,
 			      u32 uid, u32 gid);
 
-extern int gfs2_quota_sync(struct super_block *sb, int type);
+extern int gfs2_quota_sync(struct super_block *sb, int type, int wait);
 extern int gfs2_quota_refresh(struct gfs2_sbd *sdp, int user, u32 id);
 
 extern int gfs2_quota_init(struct gfs2_sbd *sdp);
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index b9dd3da22c0a..50aac606b990 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -22,6 +22,7 @@
 #include <linux/crc32.h>
 #include <linux/time.h>
 #include <linux/wait.h>
+#include <linux/writeback.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -711,7 +712,7 @@ void gfs2_unfreeze_fs(struct gfs2_sbd *sdp)
  * Returns: errno
  */
 
-static int gfs2_write_inode(struct inode *inode, int sync)
+static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
@@ -722,8 +723,7 @@ static int gfs2_write_inode(struct inode *inode, int sync)
 	int ret = 0;
 
 	/* Check this is a "normal" inode, etc */
-	if (!test_bit(GIF_USER, &ip->i_flags) ||
-	    (current->flags & PF_MEMALLOC))
+	if (current->flags & PF_MEMALLOC)
 		return 0;
 	ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
 	if (ret)
@@ -746,7 +746,7 @@ static int gfs2_write_inode(struct inode *inode, int sync)
 do_unlock:
 	gfs2_glock_dq_uninit(&gh);
 do_flush:
-	if (sync != 0)
+	if (wbc->sync_mode == WB_SYNC_ALL)
 		gfs2_log_flush(GFS2_SB(inode), ip->i_gl);
 	return ret;
 }
@@ -764,7 +764,7 @@ static int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
 	int error;
 
 	flush_workqueue(gfs2_delete_workqueue);
-	gfs2_quota_sync(sdp->sd_vfs, 0);
+	gfs2_quota_sync(sdp->sd_vfs, 0, 1);
 	gfs2_statfs_sync(sdp->sd_vfs, 0);
 
 	error = gfs2_glock_nq_init(sdp->sd_trans_gl, LM_ST_SHARED, GL_NOCACHE,
@@ -860,6 +860,7 @@ restart:
 	gfs2_clear_rgrpd(sdp);
 	gfs2_jindex_free(sdp);
 	/*  Take apart glock structures and buffer lists  */
+	invalidate_inodes(sdp->sd_vfs);
 	gfs2_gl_hash_clear(sdp);
 	/*  Unmount the locking protocol  */
 	gfs2_lm_unmount(sdp);
@@ -1194,7 +1195,7 @@ static void gfs2_drop_inode(struct inode *inode)
 {
 	struct gfs2_inode *ip = GFS2_I(inode);
 
-	if (test_bit(GIF_USER, &ip->i_flags) && inode->i_nlink) {
+	if (inode->i_nlink) {
 		struct gfs2_glock *gl = ip->i_iopen_gh.gh_gl;
 		if (gl && test_bit(GLF_DEMOTE, &gl->gl_flags))
 			clear_nlink(inode);
@@ -1212,18 +1213,12 @@ static void gfs2_clear_inode(struct inode *inode)
 {
 	struct gfs2_inode *ip = GFS2_I(inode);
 
-	/* This tells us its a "real" inode and not one which only
-	 * serves to contain an address space (see rgrp.c, meta_io.c)
-	 * which therefore doesn't have its own glocks.
-	 */
-	if (test_bit(GIF_USER, &ip->i_flags)) {
-		ip->i_gl->gl_object = NULL;
-		gfs2_glock_put(ip->i_gl);
-		ip->i_gl = NULL;
-		if (ip->i_iopen_gh.gh_gl) {
-			ip->i_iopen_gh.gh_gl->gl_object = NULL;
-			gfs2_glock_dq_uninit(&ip->i_iopen_gh);
-		}
+	ip->i_gl->gl_object = NULL;
+	gfs2_glock_put(ip->i_gl);
+	ip->i_gl = NULL;
+	if (ip->i_iopen_gh.gh_gl) {
+		ip->i_iopen_gh.gh_gl->gl_object = NULL;
+		gfs2_glock_dq_uninit(&ip->i_iopen_gh);
 	}
 }
 
@@ -1358,9 +1353,6 @@ static void gfs2_delete_inode(struct inode *inode)
 	struct gfs2_holder gh;
 	int error;
 
-	if (!test_bit(GIF_USER, &ip->i_flags))
-		goto out;
-
 	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
 	if (unlikely(error)) {
 		gfs2_glock_dq_uninit(&ip->i_iopen_gh);
diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c
index 0dc34621f6a6..b5f1a46133c8 100644
--- a/fs/gfs2/sys.c
+++ b/fs/gfs2/sys.c
@@ -167,7 +167,7 @@ static ssize_t quota_sync_store(struct gfs2_sbd *sdp, const char *buf,
 	if (simple_strtol(buf, NULL, 0) != 1)
 		return -EINVAL;
 
-	gfs2_quota_sync(sdp->sd_vfs, 0);
+	gfs2_quota_sync(sdp->sd_vfs, 0, 1);
 	return len;
 }
 
@@ -478,7 +478,6 @@ TUNE_ATTR(complain_secs, 0);
 TUNE_ATTR(statfs_slow, 0);
 TUNE_ATTR(new_files_jdata, 0);
 TUNE_ATTR(quota_simul_sync, 1);
-TUNE_ATTR(stall_secs, 1);
 TUNE_ATTR(statfs_quantum, 1);
 TUNE_ATTR_3(quota_scale, quota_scale_show, quota_scale_store);
 
@@ -491,7 +490,6 @@ static struct attribute *tune_attrs[] = {
 	&tune_attr_complain_secs.attr,
 	&tune_attr_statfs_slow.attr,
 	&tune_attr_quota_simul_sync.attr,
-	&tune_attr_stall_secs.attr,
 	&tune_attr_statfs_quantum.attr,
 	&tune_attr_quota_scale.attr,
 	&tune_attr_new_files_jdata.attr,
diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index f6a7efa34eb9..226f2bfbf16a 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -21,6 +21,7 @@
 #include "util.h"
 
 struct kmem_cache *gfs2_glock_cachep __read_mostly;
+struct kmem_cache *gfs2_glock_aspace_cachep __read_mostly;
 struct kmem_cache *gfs2_inode_cachep __read_mostly;
 struct kmem_cache *gfs2_bufdata_cachep __read_mostly;
 struct kmem_cache *gfs2_rgrpd_cachep __read_mostly;
diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h
index 33e96b0ce9ab..b432e04600de 100644
--- a/fs/gfs2/util.h
+++ b/fs/gfs2/util.h
@@ -145,6 +145,7 @@ gfs2_io_error_bh_i((sdp), (bh), __func__, __FILE__, __LINE__);
 
 
 extern struct kmem_cache *gfs2_glock_cachep;
+extern struct kmem_cache *gfs2_glock_aspace_cachep;
 extern struct kmem_cache *gfs2_inode_cachep;
 extern struct kmem_cache *gfs2_bufdata_cachep;
 extern struct kmem_cache *gfs2_rgrpd_cachep;
diff --git a/fs/hfs/hfs_fs.h b/fs/hfs/hfs_fs.h
index 052387e11671..fe35e3b626c4 100644
--- a/fs/hfs/hfs_fs.h
+++ b/fs/hfs/hfs_fs.h
@@ -188,7 +188,7 @@ extern const struct address_space_operations hfs_btree_aops;
 
 extern struct inode *hfs_new_inode(struct inode *, struct qstr *, int);
 extern void hfs_inode_write_fork(struct inode *, struct hfs_extent *, __be32 *, __be32 *);
-extern int hfs_write_inode(struct inode *, int);
+extern int hfs_write_inode(struct inode *, struct writeback_control *);
 extern int hfs_inode_setattr(struct dentry *, struct iattr *);
 extern void hfs_inode_read_fork(struct inode *inode, struct hfs_extent *ext,
 			__be32 log_size, __be32 phys_size, u32 clump_size);
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index a1cbff2b4d99..14f5cb1b9fdc 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -381,7 +381,7 @@ void hfs_inode_write_fork(struct inode *inode, struct hfs_extent *ext,
 					 HFS_SB(inode->i_sb)->alloc_blksz);
 }
 
-int hfs_write_inode(struct inode *inode, int unused)
+int hfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct inode *main_inode = inode;
 	struct hfs_find_data fd;
diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c
index 43022f3d5148..74b473a8ef92 100644
--- a/fs/hfsplus/super.c
+++ b/fs/hfsplus/super.c
@@ -87,7 +87,8 @@ bad_inode:
 	return ERR_PTR(err);
 }
 
-static int hfsplus_write_inode(struct inode *inode, int unused)
+static int hfsplus_write_inode(struct inode *inode,
+		struct writeback_control *wbc)
 {
 	struct hfsplus_vh *vhdr;
 	int ret = 0;
diff --git a/fs/hpfs/anode.c b/fs/hpfs/anode.c
index 1aa88c4e0964..6a2f04bf3df0 100644
--- a/fs/hpfs/anode.c
+++ b/fs/hpfs/anode.c
@@ -353,7 +353,7 @@ int hpfs_ea_read(struct super_block *s, secno a, int ano, unsigned pos,
 }
 
 int hpfs_ea_write(struct super_block *s, secno a, int ano, unsigned pos,
-	     unsigned len, char *buf)
+	     unsigned len, const char *buf)
 {
 	struct buffer_head *bh;
 	char *data;
diff --git a/fs/hpfs/dentry.c b/fs/hpfs/dentry.c
index 940d6d150bee..67d9d36b3d5f 100644
--- a/fs/hpfs/dentry.c
+++ b/fs/hpfs/dentry.c
@@ -20,8 +20,8 @@ static int hpfs_hash_dentry(struct dentry *dentry, struct qstr *qstr)
 
 	if (l == 1) if (qstr->name[0]=='.') goto x;
 	if (l == 2) if (qstr->name[0]=='.' || qstr->name[1]=='.') goto x;
-	hpfs_adjust_length((char *)qstr->name, &l);
-	/*if (hpfs_chk_name((char *)qstr->name,&l))*/
+	hpfs_adjust_length(qstr->name, &l);
+	/*if (hpfs_chk_name(qstr->name,&l))*/
 		/*return -ENAMETOOLONG;*/
 		/*return -ENOENT;*/
 	x:
@@ -38,14 +38,16 @@ static int hpfs_compare_dentry(struct dentry *dentry, struct qstr *a, struct qst
 {
 	unsigned al=a->len;
 	unsigned bl=b->len;
-	hpfs_adjust_length((char *)a->name, &al);
-	/*hpfs_adjust_length((char *)b->name, &bl);*/
+	hpfs_adjust_length(a->name, &al);
+	/*hpfs_adjust_length(b->name, &bl);*/
 	/* 'a' is the qstr of an already existing dentry, so the name
 	 * must be valid. 'b' must be validated first.
 	 */
 
-	if (hpfs_chk_name((char *)b->name, &bl)) return 1;
-	if (hpfs_compare_names(dentry->d_sb, (char *)a->name, al, (char *)b->name, bl, 0)) return 1;
+	if (hpfs_chk_name(b->name, &bl))
+		return 1;
+	if (hpfs_compare_names(dentry->d_sb, a->name, al, b->name, bl, 0))
+		return 1;
 	return 0;
 }
 
diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c
index 8865c94f55f6..26e3964a4b8c 100644
--- a/fs/hpfs/dir.c
+++ b/fs/hpfs/dir.c
@@ -59,7 +59,7 @@ static int hpfs_readdir(struct file *filp, void *dirent, filldir_t filldir)
 	struct hpfs_dirent *de;
 	int lc;
 	long old_pos;
-	char *tempname;
+	unsigned char *tempname;
 	int c1, c2 = 0;
 	int ret = 0;
 
@@ -158,11 +158,11 @@ static int hpfs_readdir(struct file *filp, void *dirent, filldir_t filldir)
 		tempname = hpfs_translate_name(inode->i_sb, de->name, de->namelen, lc, de->not_8x3);
 		if (filldir(dirent, tempname, de->namelen, old_pos, de->fnode, DT_UNKNOWN) < 0) {
 			filp->f_pos = old_pos;
-			if (tempname != (char *)de->name) kfree(tempname);
+			if (tempname != de->name) kfree(tempname);
 			hpfs_brelse4(&qbh);
 			goto out;
 		}
-		if (tempname != (char *)de->name) kfree(tempname);
+		if (tempname != de->name) kfree(tempname);
 		hpfs_brelse4(&qbh);
 	}
 out:
@@ -187,7 +187,7 @@ out:
 
 struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
 {
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	unsigned len = dentry->d_name.len;
 	struct quad_buffer_head qbh;
 	struct hpfs_dirent *de;
@@ -197,7 +197,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name
 	struct hpfs_inode_info *hpfs_result;
 
 	lock_kernel();
-	if ((err = hpfs_chk_name((char *)name, &len))) {
+	if ((err = hpfs_chk_name(name, &len))) {
 		if (err == -ENAMETOOLONG) {
 			unlock_kernel();
 			return ERR_PTR(-ENAMETOOLONG);
@@ -209,7 +209,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name
 	 * '.' and '..' will never be passed here.
 	 */
 
-	de = map_dirent(dir, hpfs_i(dir)->i_dno, (char *) name, len, NULL, &qbh);
+	de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, NULL, &qbh);
 
 	/*
 	 * This is not really a bailout, just means file not found.
@@ -250,7 +250,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name
 	hpfs_result = hpfs_i(result);
 	if (!de->directory) hpfs_result->i_parent_dir = dir->i_ino;
 
-	hpfs_decide_conv(result, (char *)name, len);
+	hpfs_decide_conv(result, name, len);
 
 	if (de->has_acl || de->has_xtd_perm) if (!(dir->i_sb->s_flags & MS_RDONLY)) {
 		hpfs_error(result->i_sb, "ACLs or XPERM found. This is probably HPFS386. This driver doesn't support it now. Send me some info on these structures");
diff --git a/fs/hpfs/dnode.c b/fs/hpfs/dnode.c
index fe83c2b7d2d8..9b2ffadfc8c4 100644
--- a/fs/hpfs/dnode.c
+++ b/fs/hpfs/dnode.c
@@ -158,7 +158,8 @@ static void set_last_pointer(struct super_block *s, struct dnode *d, dnode_secno
 
 /* Add an entry to dnode and don't care if it grows over 2048 bytes */
 
-struct hpfs_dirent *hpfs_add_de(struct super_block *s, struct dnode *d, unsigned char *name,
+struct hpfs_dirent *hpfs_add_de(struct super_block *s, struct dnode *d,
+				const unsigned char *name,
 				unsigned namelen, secno down_ptr)
 {
 	struct hpfs_dirent *de;
@@ -223,7 +224,7 @@ static void fix_up_ptrs(struct super_block *s, struct dnode *d)
 /* Add an entry to dnode and do dnode splitting if required */
 
 static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno,
-			     unsigned char *name, unsigned namelen,
+			     const unsigned char *name, unsigned namelen,
 			     struct hpfs_dirent *new_de, dnode_secno down_ptr)
 {
 	struct quad_buffer_head qbh, qbh1, qbh2;
@@ -231,7 +232,7 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno,
 	dnode_secno adno, rdno;
 	struct hpfs_dirent *de;
 	struct hpfs_dirent nde;
-	char *nname;
+	unsigned char *nname;
 	int h;
 	int pos;
 	struct buffer_head *bh;
@@ -305,7 +306,9 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno,
 		pos++;
 	}
 	copy_de(new_de = &nde, de);
-	memcpy(name = nname, de->name, namelen = de->namelen);
+	memcpy(nname, de->name, de->namelen);
+	name = nname;
+	namelen = de->namelen;
 	for_all_poss(i, hpfs_pos_subst, ((loff_t)dno << 4) | pos, 4);
 	down_ptr = adno;
 	set_last_pointer(i->i_sb, ad, de->down ? de_down_pointer(de) : 0);
@@ -368,7 +371,8 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno,
  * I hope, now it's finally bug-free.
  */
 
-int hpfs_add_dirent(struct inode *i, unsigned char *name, unsigned namelen,
+int hpfs_add_dirent(struct inode *i,
+		    const unsigned char *name, unsigned namelen,
 		    struct hpfs_dirent *new_de, int cdepth)
 {
 	struct hpfs_inode_info *hpfs_inode = hpfs_i(i);
@@ -897,7 +901,8 @@ struct hpfs_dirent *map_pos_dirent(struct inode *inode, loff_t *posp,
 
 /* Find a dirent in tree */
 
-struct hpfs_dirent *map_dirent(struct inode *inode, dnode_secno dno, char *name, unsigned len,
+struct hpfs_dirent *map_dirent(struct inode *inode, dnode_secno dno,
+			       const unsigned char *name, unsigned len,
 			       dnode_secno *dd, struct quad_buffer_head *qbh)
 {
 	struct dnode *dnode;
@@ -988,8 +993,8 @@ void hpfs_remove_dtree(struct super_block *s, dnode_secno dno)
 struct hpfs_dirent *map_fnode_dirent(struct super_block *s, fnode_secno fno,
 				     struct fnode *f, struct quad_buffer_head *qbh)
 {
-	char *name1;
-	char *name2;
+	unsigned char *name1;
+	unsigned char *name2;
 	int name1len, name2len;
 	struct dnode *d;
 	dnode_secno dno, downd;
diff --git a/fs/hpfs/ea.c b/fs/hpfs/ea.c
index 547a8384571f..45e53d972b42 100644
--- a/fs/hpfs/ea.c
+++ b/fs/hpfs/ea.c
@@ -62,8 +62,8 @@ static char *get_indirect_ea(struct super_block *s, int ano, secno a, int size)
 	return ret;
 }
 
-static void set_indirect_ea(struct super_block *s, int ano, secno a, char *data,
-			    int size)
+static void set_indirect_ea(struct super_block *s, int ano, secno a,
+			    const char *data, int size)
 {
 	hpfs_ea_write(s, a, ano, 0, size, data);
 }
@@ -186,7 +186,8 @@ char *hpfs_get_ea(struct super_block *s, struct fnode *fnode, char *key, int *si
  * This driver can't change sizes of eas ('cause I just don't need it).
  */
 
-void hpfs_set_ea(struct inode *inode, struct fnode *fnode, char *key, char *data, int size)
+void hpfs_set_ea(struct inode *inode, struct fnode *fnode, const char *key,
+		 const char *data, int size)
 {
 	fnode_secno fno = inode->i_ino;
 	struct super_block *s = inode->i_sb;
diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h
index 701ca54c0867..97bf738cd5d6 100644
--- a/fs/hpfs/hpfs_fn.h
+++ b/fs/hpfs/hpfs_fn.h
@@ -215,7 +215,7 @@ secno hpfs_bplus_lookup(struct super_block *, struct inode *, struct bplus_heade
 secno hpfs_add_sector_to_btree(struct super_block *, secno, int, unsigned);
 void hpfs_remove_btree(struct super_block *, struct bplus_header *);
 int hpfs_ea_read(struct super_block *, secno, int, unsigned, unsigned, char *);
-int hpfs_ea_write(struct super_block *, secno, int, unsigned, unsigned, char *);
+int hpfs_ea_write(struct super_block *, secno, int, unsigned, unsigned, const char *);
 void hpfs_ea_remove(struct super_block *, secno, int, unsigned);
 void hpfs_truncate_btree(struct super_block *, secno, int, unsigned);
 void hpfs_remove_fnode(struct super_block *, fnode_secno fno);
@@ -244,13 +244,17 @@ extern const struct file_operations hpfs_dir_ops;
 
 void hpfs_add_pos(struct inode *, loff_t *);
 void hpfs_del_pos(struct inode *, loff_t *);
-struct hpfs_dirent *hpfs_add_de(struct super_block *, struct dnode *, unsigned char *, unsigned, secno);
-int hpfs_add_dirent(struct inode *, unsigned char *, unsigned, struct hpfs_dirent *, int);
+struct hpfs_dirent *hpfs_add_de(struct super_block *, struct dnode *,
+				const unsigned char *, unsigned, secno);
+int hpfs_add_dirent(struct inode *, const unsigned char *, unsigned,
+		    struct hpfs_dirent *, int);
 int hpfs_remove_dirent(struct inode *, dnode_secno, struct hpfs_dirent *, struct quad_buffer_head *, int);
 void hpfs_count_dnodes(struct super_block *, dnode_secno, int *, int *, int *);
 dnode_secno hpfs_de_as_down_as_possible(struct super_block *, dnode_secno dno);
 struct hpfs_dirent *map_pos_dirent(struct inode *, loff_t *, struct quad_buffer_head *);
-struct hpfs_dirent *map_dirent(struct inode *, dnode_secno, char *, unsigned, dnode_secno *, struct quad_buffer_head *);
+struct hpfs_dirent *map_dirent(struct inode *, dnode_secno,
+			       const unsigned char *, unsigned, dnode_secno *,
+			       struct quad_buffer_head *);
 void hpfs_remove_dtree(struct super_block *, dnode_secno);
 struct hpfs_dirent *map_fnode_dirent(struct super_block *, fnode_secno, struct fnode *, struct quad_buffer_head *);
 
@@ -259,7 +263,8 @@ struct hpfs_dirent *map_fnode_dirent(struct super_block *, fnode_secno, struct f
 void hpfs_ea_ext_remove(struct super_block *, secno, int, unsigned);
 int hpfs_read_ea(struct super_block *, struct fnode *, char *, char *, int);
 char *hpfs_get_ea(struct super_block *, struct fnode *, char *, int *);
-void hpfs_set_ea(struct inode *, struct fnode *, char *, char *, int);
+void hpfs_set_ea(struct inode *, struct fnode *, const char *,
+		 const char *, int);
 
 /* file.c */
 
@@ -282,7 +287,7 @@ void hpfs_delete_inode(struct inode *);
 
 unsigned *hpfs_map_dnode_bitmap(struct super_block *, struct quad_buffer_head *);
 unsigned *hpfs_map_bitmap(struct super_block *, unsigned, struct quad_buffer_head *, char *);
-char *hpfs_load_code_page(struct super_block *, secno);
+unsigned char *hpfs_load_code_page(struct super_block *, secno);
 secno *hpfs_load_bitmap_directory(struct super_block *, secno bmp);
 struct fnode *hpfs_map_fnode(struct super_block *s, ino_t, struct buffer_head **);
 struct anode *hpfs_map_anode(struct super_block *s, anode_secno, struct buffer_head **);
@@ -292,12 +297,13 @@ dnode_secno hpfs_fnode_dno(struct super_block *s, ino_t ino);
 /* name.c */
 
 unsigned char hpfs_upcase(unsigned char *, unsigned char);
-int hpfs_chk_name(unsigned char *, unsigned *);
-char *hpfs_translate_name(struct super_block *, unsigned char *, unsigned, int, int);
-int hpfs_compare_names(struct super_block *, unsigned char *, unsigned, unsigned char *, unsigned, int);
-int hpfs_is_name_long(unsigned char *, unsigned);
-void hpfs_adjust_length(unsigned char *, unsigned *);
-void hpfs_decide_conv(struct inode *, unsigned char *, unsigned);
+int hpfs_chk_name(const unsigned char *, unsigned *);
+unsigned char *hpfs_translate_name(struct super_block *, unsigned char *, unsigned, int, int);
+int hpfs_compare_names(struct super_block *, const unsigned char *, unsigned,
+		       const unsigned char *, unsigned, int);
+int hpfs_is_name_long(const unsigned char *, unsigned);
+void hpfs_adjust_length(const unsigned char *, unsigned *);
+void hpfs_decide_conv(struct inode *, const unsigned char *, unsigned);
 
 /* namei.c */
 
diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c
index fe703ae46bc7..ff90affb94e1 100644
--- a/fs/hpfs/inode.c
+++ b/fs/hpfs/inode.c
@@ -46,7 +46,7 @@ void hpfs_read_inode(struct inode *i)
 	struct fnode *fnode;
 	struct super_block *sb = i->i_sb;
 	struct hpfs_inode_info *hpfs_inode = hpfs_i(i);
-	unsigned char *ea;
+	void *ea;
 	int ea_size;
 
 	if (!(fnode = hpfs_map_fnode(sb, i->i_ino, &bh))) {
@@ -112,7 +112,7 @@ void hpfs_read_inode(struct inode *i)
 		}
 	}
 	if (fnode->dirflag) {
-		unsigned n_dnodes, n_subdirs;
+		int n_dnodes, n_subdirs;
 		i->i_mode |= S_IFDIR;
 		i->i_op = &hpfs_dir_iops;
 		i->i_fop = &hpfs_dir_ops;
diff --git a/fs/hpfs/map.c b/fs/hpfs/map.c
index c4724589b2eb..840d033ecee8 100644
--- a/fs/hpfs/map.c
+++ b/fs/hpfs/map.c
@@ -35,7 +35,7 @@ unsigned int *hpfs_map_bitmap(struct super_block *s, unsigned bmp_block,
  * lowercasing table
  */
 
-char *hpfs_load_code_page(struct super_block *s, secno cps)
+unsigned char *hpfs_load_code_page(struct super_block *s, secno cps)
 {
 	struct buffer_head *bh;
 	secno cpds;
@@ -71,7 +71,7 @@ char *hpfs_load_code_page(struct super_block *s, secno cps)
 		brelse(bh);
 		return NULL;
 	}
-	ptr = (char *)cpd + cpd->offs[cpi] + 6;
+	ptr = (unsigned char *)cpd + cpd->offs[cpi] + 6;
 	if (!(cp_table = kmalloc(256, GFP_KERNEL))) {
 		printk("HPFS: out of memory for code page table\n");
 		brelse(bh);
@@ -217,7 +217,7 @@ struct dnode *hpfs_map_dnode(struct super_block *s, unsigned secno,
 	if ((dnode = hpfs_map_4sectors(s, secno, qbh, DNODE_RD_AHEAD)))
 		if (hpfs_sb(s)->sb_chk) {
 			unsigned p, pp = 0;
-			unsigned char *d = (char *)dnode;
+			unsigned char *d = (unsigned char *)dnode;
 			int b = 0;
 			if (dnode->magic != DNODE_MAGIC) {
 				hpfs_error(s, "bad magic on dnode %08x", secno);
diff --git a/fs/hpfs/name.c b/fs/hpfs/name.c
index 1f4a964384eb..f24736d7a439 100644
--- a/fs/hpfs/name.c
+++ b/fs/hpfs/name.c
@@ -8,16 +8,16 @@
 
 #include "hpfs_fn.h"
 
-static char *text_postfix[]={
+static const char *text_postfix[]={
 ".ASM", ".BAS", ".BAT", ".C", ".CC", ".CFG", ".CMD", ".CON", ".CPP", ".DEF",
 ".DOC", ".DPR", ".ERX", ".H", ".HPP", ".HTM", ".HTML", ".JAVA", ".LOG", ".PAS",
 ".RC", ".TEX", ".TXT", ".Y", ""};
 
-static char *text_prefix[]={
+static const char *text_prefix[]={
 "AUTOEXEC.", "CHANGES", "COPYING", "CONFIG.", "CREDITS", "FAQ", "FILE_ID.DIZ",
 "MAKEFILE", "READ.ME", "README", "TERMCAP", ""};
 
-void hpfs_decide_conv(struct inode *inode, unsigned char *name, unsigned len)
+void hpfs_decide_conv(struct inode *inode, const unsigned char *name, unsigned len)
 {
 	struct hpfs_inode_info *hpfs_inode = hpfs_i(inode);
 	int i;
@@ -71,7 +71,7 @@ static inline unsigned char locase(unsigned char *dir, unsigned char a)
 	return dir[a];
 }
 
-int hpfs_chk_name(unsigned char *name, unsigned *len)
+int hpfs_chk_name(const unsigned char *name, unsigned *len)
 {
 	int i;
 	if (*len > 254) return -ENAMETOOLONG;
@@ -83,10 +83,10 @@ int hpfs_chk_name(unsigned char *name, unsigned *len)
 	return 0;
 }
 
-char *hpfs_translate_name(struct super_block *s, unsigned char *from,
+unsigned char *hpfs_translate_name(struct super_block *s, unsigned char *from,
 			  unsigned len, int lc, int lng)
 {
-	char *to;
+	unsigned char *to;
 	int i;
 	if (hpfs_sb(s)->sb_chk >= 2) if (hpfs_is_name_long(from, len) != lng) {
 		printk("HPFS: Long name flag mismatch - name ");
@@ -103,8 +103,9 @@ char *hpfs_translate_name(struct super_block *s, unsigned char *from,
 	return to;
 }
 
-int hpfs_compare_names(struct super_block *s, unsigned char *n1, unsigned l1,
-		       unsigned char *n2, unsigned l2, int last)
+int hpfs_compare_names(struct super_block *s,
+		       const unsigned char *n1, unsigned l1,
+		       const unsigned char *n2, unsigned l2, int last)
 {
 	unsigned l = l1 < l2 ? l1 : l2;
 	unsigned i;
@@ -120,7 +121,7 @@ int hpfs_compare_names(struct super_block *s, unsigned char *n1, unsigned l1,
 	return 0;
 }
 
-int hpfs_is_name_long(unsigned char *name, unsigned len)
+int hpfs_is_name_long(const unsigned char *name, unsigned len)
 {
 	int i,j;
 	for (i = 0; i < len && name[i] != '.'; i++)
@@ -134,7 +135,7 @@ int hpfs_is_name_long(unsigned char *name, unsigned len)
 
 /* OS/2 clears dots and spaces at the end of file name, so we have to */
 
-void hpfs_adjust_length(unsigned char *name, unsigned *len)
+void hpfs_adjust_length(const unsigned char *name, unsigned *len)
 {
 	if (!*len) return;
 	if (*len == 1 && name[0] == '.') return;
diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c
index 82b9c4ba9ed0..11c2b4080f65 100644
--- a/fs/hpfs/namei.c
+++ b/fs/hpfs/namei.c
@@ -11,7 +11,7 @@
 
 static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 {
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	unsigned len = dentry->d_name.len;
 	struct quad_buffer_head qbh0;
 	struct buffer_head *bh;
@@ -24,7 +24,7 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	int r;
 	struct hpfs_dirent dee;
 	int err;
-	if ((err = hpfs_chk_name((char *)name, &len))) return err==-ENOENT ? -EINVAL : err;
+	if ((err = hpfs_chk_name(name, &len))) return err==-ENOENT ? -EINVAL : err;
 	lock_kernel();
 	err = -ENOSPC;
 	fnode = hpfs_alloc_fnode(dir->i_sb, hpfs_i(dir)->i_dno, &fno, &bh);
@@ -62,7 +62,7 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 		result->i_mode &= ~0222;
 
 	mutex_lock(&hpfs_i(dir)->i_mutex);
-	r = hpfs_add_dirent(dir, (char *)name, len, &dee, 0);
+	r = hpfs_add_dirent(dir, name, len, &dee, 0);
 	if (r == 1)
 		goto bail3;
 	if (r == -1) {
@@ -121,7 +121,7 @@ bail:
 
 static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struct nameidata *nd)
 {
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	unsigned len = dentry->d_name.len;
 	struct inode *result = NULL;
 	struct buffer_head *bh;
@@ -130,7 +130,7 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc
 	int r;
 	struct hpfs_dirent dee;
 	int err;
-	if ((err = hpfs_chk_name((char *)name, &len)))
+	if ((err = hpfs_chk_name(name, &len)))
 		return err==-ENOENT ? -EINVAL : err;
 	lock_kernel();
 	err = -ENOSPC;
@@ -155,7 +155,7 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc
 	result->i_op = &hpfs_file_iops;
 	result->i_fop = &hpfs_file_ops;
 	result->i_nlink = 1;
-	hpfs_decide_conv(result, (char *)name, len);
+	hpfs_decide_conv(result, name, len);
 	hpfs_i(result)->i_parent_dir = dir->i_ino;
 	result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date);
 	result->i_ctime.tv_nsec = 0;
@@ -170,7 +170,7 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc
 	hpfs_i(result)->mmu_private = 0;
 
 	mutex_lock(&hpfs_i(dir)->i_mutex);
-	r = hpfs_add_dirent(dir, (char *)name, len, &dee, 0);
+	r = hpfs_add_dirent(dir, name, len, &dee, 0);
 	if (r == 1)
 		goto bail2;
 	if (r == -1) {
@@ -211,7 +211,7 @@ bail:
 
 static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev)
 {
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	unsigned len = dentry->d_name.len;
 	struct buffer_head *bh;
 	struct fnode *fnode;
@@ -220,7 +220,7 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t
 	struct hpfs_dirent dee;
 	struct inode *result = NULL;
 	int err;
-	if ((err = hpfs_chk_name((char *)name, &len))) return err==-ENOENT ? -EINVAL : err;
+	if ((err = hpfs_chk_name(name, &len))) return err==-ENOENT ? -EINVAL : err;
 	if (hpfs_sb(dir->i_sb)->sb_eas < 2) return -EPERM;
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
@@ -256,7 +256,7 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t
 	init_special_inode(result, mode, rdev);
 
 	mutex_lock(&hpfs_i(dir)->i_mutex);
-	r = hpfs_add_dirent(dir, (char *)name, len, &dee, 0);
+	r = hpfs_add_dirent(dir, name, len, &dee, 0);
 	if (r == 1)
 		goto bail2;
 	if (r == -1) {
@@ -289,7 +289,7 @@ bail:
 
 static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *symlink)
 {
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	unsigned len = dentry->d_name.len;
 	struct buffer_head *bh;
 	struct fnode *fnode;
@@ -298,7 +298,7 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy
 	struct hpfs_dirent dee;
 	struct inode *result;
 	int err;
-	if ((err = hpfs_chk_name((char *)name, &len))) return err==-ENOENT ? -EINVAL : err;
+	if ((err = hpfs_chk_name(name, &len))) return err==-ENOENT ? -EINVAL : err;
 	lock_kernel();
 	if (hpfs_sb(dir->i_sb)->sb_eas < 2) {
 		unlock_kernel();
@@ -335,7 +335,7 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy
 	result->i_data.a_ops = &hpfs_symlink_aops;
 
 	mutex_lock(&hpfs_i(dir)->i_mutex);
-	r = hpfs_add_dirent(dir, (char *)name, len, &dee, 0);
+	r = hpfs_add_dirent(dir, name, len, &dee, 0);
 	if (r == 1)
 		goto bail2;
 	if (r == -1) {
@@ -345,7 +345,7 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy
 	fnode->len = len;
 	memcpy(fnode->name, name, len > 15 ? 15 : len);
 	fnode->up = dir->i_ino;
-	hpfs_set_ea(result, fnode, "SYMLINK", (char *)symlink, strlen(symlink));
+	hpfs_set_ea(result, fnode, "SYMLINK", symlink, strlen(symlink));
 	mark_buffer_dirty(bh);
 	brelse(bh);
 
@@ -369,7 +369,7 @@ bail:
 
 static int hpfs_unlink(struct inode *dir, struct dentry *dentry)
 {
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	unsigned len = dentry->d_name.len;
 	struct quad_buffer_head qbh;
 	struct hpfs_dirent *de;
@@ -381,12 +381,12 @@ static int hpfs_unlink(struct inode *dir, struct dentry *dentry)
 	int err;
 
 	lock_kernel();
-	hpfs_adjust_length((char *)name, &len);
+	hpfs_adjust_length(name, &len);
 again:
 	mutex_lock(&hpfs_i(inode)->i_parent_mutex);
 	mutex_lock(&hpfs_i(dir)->i_mutex);
 	err = -ENOENT;
-	de = map_dirent(dir, hpfs_i(dir)->i_dno, (char *)name, len, &dno, &qbh);
+	de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, &dno, &qbh);
 	if (!de)
 		goto out;
 
@@ -413,22 +413,25 @@ again:
 
 		mutex_unlock(&hpfs_i(dir)->i_mutex);
 		mutex_unlock(&hpfs_i(inode)->i_parent_mutex);
-		d_drop(dentry);
-		spin_lock(&dentry->d_lock);
-		if (atomic_read(&dentry->d_count) > 1 ||
-		    generic_permission(inode, MAY_WRITE, NULL) ||
+		dentry_unhash(dentry);
+		if (!d_unhashed(dentry)) {
+			dput(dentry);
+			unlock_kernel();
+			return -ENOSPC;
+		}
+		if (generic_permission(inode, MAY_WRITE, NULL) ||
 		    !S_ISREG(inode->i_mode) ||
 		    get_write_access(inode)) {
-			spin_unlock(&dentry->d_lock);
 			d_rehash(dentry);
+			dput(dentry);
 		} else {
 			struct iattr newattrs;
-			spin_unlock(&dentry->d_lock);
 			/*printk("HPFS: truncating file before delete.\n");*/
 			newattrs.ia_size = 0;
 			newattrs.ia_valid = ATTR_SIZE | ATTR_CTIME;
 			err = notify_change(dentry, &newattrs);
 			put_write_access(inode);
+			dput(dentry);
 			if (!err)
 				goto again;
 		}
@@ -451,7 +454,7 @@ out:
 
 static int hpfs_rmdir(struct inode *dir, struct dentry *dentry)
 {
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	unsigned len = dentry->d_name.len;
 	struct quad_buffer_head qbh;
 	struct hpfs_dirent *de;
@@ -462,12 +465,12 @@ static int hpfs_rmdir(struct inode *dir, struct dentry *dentry)
 	int err;
 	int r;
 
-	hpfs_adjust_length((char *)name, &len);
+	hpfs_adjust_length(name, &len);
 	lock_kernel();
 	mutex_lock(&hpfs_i(inode)->i_parent_mutex);
 	mutex_lock(&hpfs_i(dir)->i_mutex);
 	err = -ENOENT;
-	de = map_dirent(dir, hpfs_i(dir)->i_dno, (char *)name, len, &dno, &qbh);
+	de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, &dno, &qbh);
 	if (!de)
 		goto out;
 
@@ -546,10 +549,10 @@ const struct address_space_operations hpfs_symlink_aops = {
 static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		struct inode *new_dir, struct dentry *new_dentry)
 {
-	char *old_name = (char *)old_dentry->d_name.name;
-	int old_len = old_dentry->d_name.len;
-	char *new_name = (char *)new_dentry->d_name.name;
-	int new_len = new_dentry->d_name.len;
+	const unsigned char *old_name = old_dentry->d_name.name;
+	unsigned old_len = old_dentry->d_name.len;
+	const unsigned char *new_name = new_dentry->d_name.name;
+	unsigned new_len = new_dentry->d_name.len;
 	struct inode *i = old_dentry->d_inode;
 	struct inode *new_inode = new_dentry->d_inode;
 	struct quad_buffer_head qbh, qbh1;
@@ -560,9 +563,9 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	struct buffer_head *bh;
 	struct fnode *fnode;
 	int err;
-	if ((err = hpfs_chk_name((char *)new_name, &new_len))) return err;
+	if ((err = hpfs_chk_name(new_name, &new_len))) return err;
 	err = 0;
-	hpfs_adjust_length((char *)old_name, &old_len);
+	hpfs_adjust_length(old_name, &old_len);
 
 	lock_kernel();
 	/* order doesn't matter, due to VFS exclusion */
@@ -579,7 +582,7 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		goto end1;
 	}
 
-	if (!(dep = map_dirent(old_dir, hpfs_i(old_dir)->i_dno, (char *)old_name, old_len, &dno, &qbh))) {
+	if (!(dep = map_dirent(old_dir, hpfs_i(old_dir)->i_dno, old_name, old_len, &dno, &qbh))) {
 		hpfs_error(i->i_sb, "lookup succeeded but map dirent failed");
 		err = -ENOENT;
 		goto end1;
@@ -590,7 +593,7 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (new_inode) {
 		int r;
 		if ((r = hpfs_remove_dirent(old_dir, dno, dep, &qbh, 1)) != 2) {
-			if ((nde = map_dirent(new_dir, hpfs_i(new_dir)->i_dno, (char *)new_name, new_len, NULL, &qbh1))) {
+			if ((nde = map_dirent(new_dir, hpfs_i(new_dir)->i_dno, new_name, new_len, NULL, &qbh1))) {
 				clear_nlink(new_inode);
 				copy_de(nde, &de);
 				memcpy(nde->name, new_name, new_len);
@@ -618,7 +621,7 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	}
 	
 	if (new_dir == old_dir)
-		if (!(dep = map_dirent(old_dir, hpfs_i(old_dir)->i_dno, (char *)old_name, old_len, &dno, &qbh))) {
+		if (!(dep = map_dirent(old_dir, hpfs_i(old_dir)->i_dno, old_name, old_len, &dno, &qbh))) {
 			hpfs_unlock_creation(i->i_sb);
 			hpfs_error(i->i_sb, "lookup succeeded but map dirent failed at #2");
 			err = -ENOENT;
@@ -648,7 +651,7 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		brelse(bh);
 	}
 	hpfs_i(i)->i_conv = hpfs_sb(i->i_sb)->sb_conv;
-	hpfs_decide_conv(i, (char *)new_name, new_len);
+	hpfs_decide_conv(i, new_name, new_len);
 end1:
 	if (old_dir != new_dir)
 		mutex_unlock(&hpfs_i(new_dir)->i_mutex);
diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c
index 7239efc690d8..2e4dfa8593da 100644
--- a/fs/hppfs/hppfs.c
+++ b/fs/hppfs/hppfs.c
@@ -718,7 +718,7 @@ static int hppfs_fill_super(struct super_block *sb, void *d, int silent)
 	struct vfsmount *proc_mnt;
 	int err = -ENOENT;
 
-	proc_mnt = do_kern_mount("proc", 0, "proc", NULL);
+	proc_mnt = mntget(current->nsproxy->pid_ns->proc_mnt);
 	if (IS_ERR(proc_mnt))
 		goto out;
 
diff --git a/fs/inode.c b/fs/inode.c
index 03dfeb2e3928..407bf392e20a 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -8,7 +8,6 @@
 #include <linux/mm.h>
 #include <linux/dcache.h>
 #include <linux/init.h>
-#include <linux/quotaops.h>
 #include <linux/slab.h>
 #include <linux/writeback.h>
 #include <linux/module.h>
@@ -314,7 +313,6 @@ void clear_inode(struct inode *inode)
 	BUG_ON(!(inode->i_state & I_FREEING));
 	BUG_ON(inode->i_state & I_CLEAR);
 	inode_sync_wait(inode);
-	vfs_dq_drop(inode);
 	if (inode->i_sb->s_op->clear_inode)
 		inode->i_sb->s_op->clear_inode(inode);
 	if (S_ISBLK(inode->i_mode) && inode->i_bdev)
@@ -1211,8 +1209,6 @@ void generic_delete_inode(struct inode *inode)
 
 	if (op->delete_inode) {
 		void (*delete)(struct inode *) = op->delete_inode;
-		if (!is_bad_inode(inode))
-			vfs_dq_init(inode);
 		/* Filesystems implementing their own
 		 * s_op->delete_inode are required to call
 		 * truncate_inode_pages and clear_inode()
diff --git a/fs/internal.h b/fs/internal.h
index e96a1667d749..8a03a5447bdf 100644
--- a/fs/internal.h
+++ b/fs/internal.h
@@ -70,6 +70,8 @@ extern struct vfsmount *copy_tree(struct vfsmount *, struct dentry *, int);
 
 extern void __init mnt_init(void);
 
+extern spinlock_t vfsmount_lock;
+
 /*
  * fs_struct.c
  */
diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c
index 4bd882548c45..2c90e3ef625f 100644
--- a/fs/jbd/commit.c
+++ b/fs/jbd/commit.c
@@ -862,12 +862,12 @@ restart_loop:
 		/* A buffer which has been freed while still being
 		 * journaled by a previous transaction may end up still
 		 * being dirty here, but we want to avoid writing back
-		 * that buffer in the future now that the last use has
-		 * been committed.  That's not only a performance gain,
-		 * it also stops aliasing problems if the buffer is left
-		 * behind for writeback and gets reallocated for another
+		 * that buffer in the future after the "add to orphan"
+		 * operation been committed,  That's not only a performance
+		 * gain, it also stops aliasing problems if the buffer is
+		 * left behind for writeback and gets reallocated for another
 		 * use in a different page. */
-		if (buffer_freed(bh)) {
+		if (buffer_freed(bh) && !jh->b_next_transaction) {
 			clear_buffer_freed(bh);
 			clear_buffer_jbddirty(bh);
 		}
diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c
index 006f9ad838a2..99e9fea11077 100644
--- a/fs/jbd/transaction.c
+++ b/fs/jbd/transaction.c
@@ -1864,6 +1864,21 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 	if (!jh)
 		goto zap_buffer_no_jh;
 
+	/*
+	 * We cannot remove the buffer from checkpoint lists until the
+	 * transaction adding inode to orphan list (let's call it T)
+	 * is committed.  Otherwise if the transaction changing the
+	 * buffer would be cleaned from the journal before T is
+	 * committed, a crash will cause that the correct contents of
+	 * the buffer will be lost.  On the other hand we have to
+	 * clear the buffer dirty bit at latest at the moment when the
+	 * transaction marking the buffer as freed in the filesystem
+	 * structures is committed because from that moment on the
+	 * buffer can be reallocated and used by a different page.
+	 * Since the block hasn't been freed yet but the inode has
+	 * already been added to orphan list, it is safe for us to add
+	 * the buffer to BJ_Forget list of the newest transaction.
+	 */
 	transaction = jh->b_transaction;
 	if (transaction == NULL) {
 		/* First case: not on any transaction.  If it
@@ -1929,16 +1944,15 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 			goto zap_buffer;
 		}
 		/*
-		 * If it is committing, we simply cannot touch it.  We
-		 * can remove it's next_transaction pointer from the
-		 * running transaction if that is set, but nothing
-		 * else. */
+		 * The buffer is committing, we simply cannot touch
+		 * it. So we just set j_next_transaction to the
+		 * running transaction (if there is one) and mark
+		 * buffer as freed so that commit code knows it should
+		 * clear dirty bits when it is done with the buffer.
+		 */
 		set_buffer_freed(bh);
-		if (jh->b_next_transaction) {
-			J_ASSERT(jh->b_next_transaction ==
-					journal->j_running_transaction);
-			jh->b_next_transaction = NULL;
-		}
+		if (journal->j_running_transaction && buffer_jbddirty(bh))
+			jh->b_next_transaction = journal->j_running_transaction;
 		journal_put_journal_head(jh);
 		spin_unlock(&journal->j_list_lock);
 		jbd_unlock_bh_state(bh);
@@ -2120,7 +2134,7 @@ void journal_file_buffer(struct journal_head *jh,
  */
 void __journal_refile_buffer(struct journal_head *jh)
 {
-	int was_dirty;
+	int was_dirty, jlist;
 	struct buffer_head *bh = jh2bh(jh);
 
 	J_ASSERT_JH(jh, jbd_is_locked_bh_state(bh));
@@ -2142,8 +2156,13 @@ void __journal_refile_buffer(struct journal_head *jh)
 	__journal_temp_unlink_buffer(jh);
 	jh->b_transaction = jh->b_next_transaction;
 	jh->b_next_transaction = NULL;
-	__journal_file_buffer(jh, jh->b_transaction,
-				jh->b_modified ? BJ_Metadata : BJ_Reserved);
+	if (buffer_freed(bh))
+		jlist = BJ_Forget;
+	else if (jh->b_modified)
+		jlist = BJ_Metadata;
+	else
+		jlist = BJ_Reserved;
+	__journal_file_buffer(jh, jh->b_transaction, jlist);
 	J_ASSERT_JH(jh, jh->b_transaction->t_state == T_RUNNING);
 
 	if (was_dirty)
diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 886849370950..30beb11ef928 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -507,6 +507,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
 	if (blocknr < journal->j_tail)
 		freed = freed + journal->j_last - journal->j_first;
 
+	trace_jbd2_cleanup_journal_tail(journal, first_tid, blocknr, freed);
 	jbd_debug(1,
 		  "Cleaning journal tail from %d to %d (offset %lu), "
 		  "freeing %lu\n",
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 1bc74b6f26d2..671da7fb7ffd 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -883,8 +883,7 @@ restart_loop:
 		spin_unlock(&journal->j_list_lock);
 		bh = jh2bh(jh);
 		jbd_lock_bh_state(bh);
-		J_ASSERT_JH(jh,	jh->b_transaction == commit_transaction ||
-			jh->b_transaction == journal->j_running_transaction);
+		J_ASSERT_JH(jh,	jh->b_transaction == commit_transaction);
 
 		/*
 		 * If there is undo-protected committed data against
@@ -930,12 +929,12 @@ restart_loop:
 		/* A buffer which has been freed while still being
 		 * journaled by a previous transaction may end up still
 		 * being dirty here, but we want to avoid writing back
-		 * that buffer in the future now that the last use has
-		 * been committed.  That's not only a performance gain,
-		 * it also stops aliasing problems if the buffer is left
-		 * behind for writeback and gets reallocated for another
+		 * that buffer in the future after the "add to orphan"
+		 * operation been committed,  That's not only a performance
+		 * gain, it also stops aliasing problems if the buffer is
+		 * left behind for writeback and gets reallocated for another
 		 * use in a different page. */
-		if (buffer_freed(bh)) {
+		if (buffer_freed(bh) && !jh->b_next_transaction) {
 			clear_buffer_freed(bh);
 			clear_buffer_jbddirty(bh);
 		}
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index ac0d027595d0..c03d4dce4d76 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -39,6 +39,8 @@
 #include <linux/seq_file.h>
 #include <linux/math64.h>
 #include <linux/hash.h>
+#include <linux/log2.h>
+#include <linux/vmalloc.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/jbd2.h>
@@ -93,6 +95,7 @@ EXPORT_SYMBOL(jbd2_journal_begin_ordered_truncate);
 
 static int journal_convert_superblock_v1(journal_t *, journal_superblock_t *);
 static void __journal_abort_soft (journal_t *journal, int errno);
+static int jbd2_journal_create_slab(size_t slab_size);
 
 /*
  * Helper function used to manage commit timeouts
@@ -1248,6 +1251,13 @@ int jbd2_journal_load(journal_t *journal)
 		}
 	}
 
+	/*
+	 * Create a slab for this blocksize
+	 */
+	err = jbd2_journal_create_slab(be32_to_cpu(sb->s_blocksize));
+	if (err)
+		return err;
+
 	/* Let the recovery code check whether it needs to recover any
 	 * data from the journal. */
 	if (jbd2_journal_recover(journal))
@@ -1807,6 +1817,127 @@ size_t journal_tag_bytes(journal_t *journal)
 }
 
 /*
+ * JBD memory management
+ *
+ * These functions are used to allocate block-sized chunks of memory
+ * used for making copies of buffer_head data.  Very often it will be
+ * page-sized chunks of data, but sometimes it will be in
+ * sub-page-size chunks.  (For example, 16k pages on Power systems
+ * with a 4k block file system.)  For blocks smaller than a page, we
+ * use a SLAB allocator.  There are slab caches for each block size,
+ * which are allocated at mount time, if necessary, and we only free
+ * (all of) the slab caches when/if the jbd2 module is unloaded.  For
+ * this reason we don't need to a mutex to protect access to
+ * jbd2_slab[] allocating or releasing memory; only in
+ * jbd2_journal_create_slab().
+ */
+#define JBD2_MAX_SLABS 8
+static struct kmem_cache *jbd2_slab[JBD2_MAX_SLABS];
+static DECLARE_MUTEX(jbd2_slab_create_sem);
+
+static const char *jbd2_slab_names[JBD2_MAX_SLABS] = {
+	"jbd2_1k", "jbd2_2k", "jbd2_4k", "jbd2_8k",
+	"jbd2_16k", "jbd2_32k", "jbd2_64k", "jbd2_128k"
+};
+
+
+static void jbd2_journal_destroy_slabs(void)
+{
+	int i;
+
+	for (i = 0; i < JBD2_MAX_SLABS; i++) {
+		if (jbd2_slab[i])
+			kmem_cache_destroy(jbd2_slab[i]);
+		jbd2_slab[i] = NULL;
+	}
+}
+
+static int jbd2_journal_create_slab(size_t size)
+{
+	int i = order_base_2(size) - 10;
+	size_t slab_size;
+
+	if (size == PAGE_SIZE)
+		return 0;
+
+	if (i >= JBD2_MAX_SLABS)
+		return -EINVAL;
+
+	if (unlikely(i < 0))
+		i = 0;
+	down(&jbd2_slab_create_sem);
+	if (jbd2_slab[i]) {
+		up(&jbd2_slab_create_sem);
+		return 0;	/* Already created */
+	}
+
+	slab_size = 1 << (i+10);
+	jbd2_slab[i] = kmem_cache_create(jbd2_slab_names[i], slab_size,
+					 slab_size, 0, NULL);
+	up(&jbd2_slab_create_sem);
+	if (!jbd2_slab[i]) {
+		printk(KERN_EMERG "JBD2: no memory for jbd2_slab cache\n");
+		return -ENOMEM;
+	}
+	return 0;
+}
+
+static struct kmem_cache *get_slab(size_t size)
+{
+	int i = order_base_2(size) - 10;
+
+	BUG_ON(i >= JBD2_MAX_SLABS);
+	if (unlikely(i < 0))
+		i = 0;
+	BUG_ON(jbd2_slab[i] == 0);
+	return jbd2_slab[i];
+}
+
+void *jbd2_alloc(size_t size, gfp_t flags)
+{
+	void *ptr;
+
+	BUG_ON(size & (size-1)); /* Must be a power of 2 */
+
+	flags |= __GFP_REPEAT;
+	if (size == PAGE_SIZE)
+		ptr = (void *)__get_free_pages(flags, 0);
+	else if (size > PAGE_SIZE) {
+		int order = get_order(size);
+
+		if (order < 3)
+			ptr = (void *)__get_free_pages(flags, order);
+		else
+			ptr = vmalloc(size);
+	} else
+		ptr = kmem_cache_alloc(get_slab(size), flags);
+
+	/* Check alignment; SLUB has gotten this wrong in the past,
+	 * and this can lead to user data corruption! */
+	BUG_ON(((unsigned long) ptr) & (size-1));
+
+	return ptr;
+}
+
+void jbd2_free(void *ptr, size_t size)
+{
+	if (size == PAGE_SIZE) {
+		free_pages((unsigned long)ptr, 0);
+		return;
+	}
+	if (size > PAGE_SIZE) {
+		int order = get_order(size);
+
+		if (order < 3)
+			free_pages((unsigned long)ptr, order);
+		else
+			vfree(ptr);
+		return;
+	}
+	kmem_cache_free(get_slab(size), ptr);
+};
+
+/*
  * Journal_head storage management
  */
 static struct kmem_cache *jbd2_journal_head_cache;
@@ -2204,6 +2335,7 @@ static void jbd2_journal_destroy_caches(void)
 	jbd2_journal_destroy_revoke_caches();
 	jbd2_journal_destroy_jbd2_journal_head_cache();
 	jbd2_journal_destroy_handle_cache();
+	jbd2_journal_destroy_slabs();
 }
 
 static int __init journal_init(void)
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index a0512700542f..bfc70f57900f 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -1727,6 +1727,21 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 	if (!jh)
 		goto zap_buffer_no_jh;
 
+	/*
+	 * We cannot remove the buffer from checkpoint lists until the
+	 * transaction adding inode to orphan list (let's call it T)
+	 * is committed.  Otherwise if the transaction changing the
+	 * buffer would be cleaned from the journal before T is
+	 * committed, a crash will cause that the correct contents of
+	 * the buffer will be lost.  On the other hand we have to
+	 * clear the buffer dirty bit at latest at the moment when the
+	 * transaction marking the buffer as freed in the filesystem
+	 * structures is committed because from that moment on the
+	 * buffer can be reallocated and used by a different page.
+	 * Since the block hasn't been freed yet but the inode has
+	 * already been added to orphan list, it is safe for us to add
+	 * the buffer to BJ_Forget list of the newest transaction.
+	 */
 	transaction = jh->b_transaction;
 	if (transaction == NULL) {
 		/* First case: not on any transaction.  If it
@@ -1783,16 +1798,15 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
 	} else if (transaction == journal->j_committing_transaction) {
 		JBUFFER_TRACE(jh, "on committing transaction");
 		/*
-		 * If it is committing, we simply cannot touch it.  We
-		 * can remove it's next_transaction pointer from the
-		 * running transaction if that is set, but nothing
-		 * else. */
+		 * The buffer is committing, we simply cannot touch
+		 * it. So we just set j_next_transaction to the
+		 * running transaction (if there is one) and mark
+		 * buffer as freed so that commit code knows it should
+		 * clear dirty bits when it is done with the buffer.
+		 */
 		set_buffer_freed(bh);
-		if (jh->b_next_transaction) {
-			J_ASSERT(jh->b_next_transaction ==
-					journal->j_running_transaction);
-			jh->b_next_transaction = NULL;
-		}
+		if (journal->j_running_transaction && buffer_jbddirty(bh))
+			jh->b_next_transaction = journal->j_running_transaction;
 		jbd2_journal_put_journal_head(jh);
 		spin_unlock(&journal->j_list_lock);
 		jbd_unlock_bh_state(bh);
@@ -1969,7 +1983,7 @@ void jbd2_journal_file_buffer(struct journal_head *jh,
  */
 void __jbd2_journal_refile_buffer(struct journal_head *jh)
 {
-	int was_dirty;
+	int was_dirty, jlist;
 	struct buffer_head *bh = jh2bh(jh);
 
 	J_ASSERT_JH(jh, jbd_is_locked_bh_state(bh));
@@ -1991,8 +2005,13 @@ void __jbd2_journal_refile_buffer(struct journal_head *jh)
 	__jbd2_journal_temp_unlink_buffer(jh);
 	jh->b_transaction = jh->b_next_transaction;
 	jh->b_next_transaction = NULL;
-	__jbd2_journal_file_buffer(jh, jh->b_transaction,
-				jh->b_modified ? BJ_Metadata : BJ_Reserved);
+	if (buffer_freed(bh))
+		jlist = BJ_Forget;
+	else if (jh->b_modified)
+		jlist = BJ_Metadata;
+	else
+		jlist = BJ_Reserved;
+	__jbd2_journal_file_buffer(jh, jh->b_transaction, jlist);
 	J_ASSERT_JH(jh, jh->b_transaction->t_state == T_RUNNING);
 
 	if (was_dirty)
diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c
index d66477c34306..213169780b6c 100644
--- a/fs/jfs/acl.c
+++ b/fs/jfs/acl.c
@@ -20,7 +20,6 @@
 
 #include <linux/sched.h>
 #include <linux/fs.h>
-#include <linux/quotaops.h>
 #include <linux/posix_acl_xattr.h>
 #include "jfs_incore.h"
 #include "jfs_txnmgr.h"
@@ -174,7 +173,7 @@ cleanup:
 	return rc;
 }
 
-static int jfs_acl_chmod(struct inode *inode)
+int jfs_acl_chmod(struct inode *inode)
 {
 	struct posix_acl *acl, *clone;
 	int rc;
@@ -205,26 +204,3 @@ static int jfs_acl_chmod(struct inode *inode)
 	posix_acl_release(clone);
 	return rc;
 }
-
-int jfs_setattr(struct dentry *dentry, struct iattr *iattr)
-{
-	struct inode *inode = dentry->d_inode;
-	int rc;
-
-	rc = inode_change_ok(inode, iattr);
-	if (rc)
-		return rc;
-
-	if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) ||
-	    (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) {
-		if (vfs_dq_transfer(inode, iattr))
-			return -EDQUOT;
-	}
-
-	rc = inode_setattr(inode, iattr);
-
-	if (!rc && (iattr->ia_valid & ATTR_MODE))
-		rc = jfs_acl_chmod(inode);
-
-	return rc;
-}
diff --git a/fs/jfs/file.c b/fs/jfs/file.c
index 2b70fa78e4a7..14ba982b3f24 100644
--- a/fs/jfs/file.c
+++ b/fs/jfs/file.c
@@ -18,6 +18,7 @@
  */
 
 #include <linux/fs.h>
+#include <linux/quotaops.h>
 #include "jfs_incore.h"
 #include "jfs_inode.h"
 #include "jfs_dmap.h"
@@ -47,7 +48,7 @@ static int jfs_open(struct inode *inode, struct file *file)
 {
 	int rc;
 
-	if ((rc = generic_file_open(inode, file)))
+	if ((rc = dquot_file_open(inode, file)))
 		return rc;
 
 	/*
@@ -88,14 +89,40 @@ static int jfs_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
+int jfs_setattr(struct dentry *dentry, struct iattr *iattr)
+{
+	struct inode *inode = dentry->d_inode;
+	int rc;
+
+	rc = inode_change_ok(inode, iattr);
+	if (rc)
+		return rc;
+
+	if (iattr->ia_valid & ATTR_SIZE)
+		dquot_initialize(inode);
+	if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) ||
+	    (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) {
+		rc = dquot_transfer(inode, iattr);
+		if (rc)
+			return rc;
+	}
+
+	rc = inode_setattr(inode, iattr);
+
+	if (!rc && (iattr->ia_valid & ATTR_MODE))
+		rc = jfs_acl_chmod(inode);
+
+	return rc;
+}
+
 const struct inode_operations jfs_file_inode_operations = {
 	.truncate	= jfs_truncate,
 	.setxattr	= jfs_setxattr,
 	.getxattr	= jfs_getxattr,
 	.listxattr	= jfs_listxattr,
 	.removexattr	= jfs_removexattr,
-#ifdef CONFIG_JFS_POSIX_ACL
 	.setattr	= jfs_setattr,
+#ifdef CONFIG_JFS_POSIX_ACL
 	.check_acl	= jfs_check_acl,
 #endif
 };
diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c
index b2ae190a77ba..9dd126276c9f 100644
--- a/fs/jfs/inode.c
+++ b/fs/jfs/inode.c
@@ -22,6 +22,7 @@
 #include <linux/buffer_head.h>
 #include <linux/pagemap.h>
 #include <linux/quotaops.h>
+#include <linux/writeback.h>
 #include "jfs_incore.h"
 #include "jfs_inode.h"
 #include "jfs_filsys.h"
@@ -120,8 +121,10 @@ int jfs_commit_inode(struct inode *inode, int wait)
 	return rc;
 }
 
-int jfs_write_inode(struct inode *inode, int wait)
+int jfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
+	int wait = wbc->sync_mode == WB_SYNC_ALL;
+
 	if (test_cflag(COMMIT_Nolink, inode))
 		return 0;
 	/*
@@ -146,6 +149,9 @@ void jfs_delete_inode(struct inode *inode)
 {
 	jfs_info("In jfs_delete_inode, inode = 0x%p", inode);
 
+	if (!is_bad_inode(inode))
+		dquot_initialize(inode);
+
 	if (!is_bad_inode(inode) &&
 	    (JFS_IP(inode)->fileset == FILESYSTEM_I)) {
 		truncate_inode_pages(&inode->i_data, 0);
@@ -158,9 +164,9 @@ void jfs_delete_inode(struct inode *inode)
 		/*
 		 * Free the inode from the quota allocation.
 		 */
-		vfs_dq_init(inode);
-		vfs_dq_free_inode(inode);
-		vfs_dq_drop(inode);
+		dquot_initialize(inode);
+		dquot_free_inode(inode);
+		dquot_drop(inode);
 	}
 
 	clear_inode(inode);
diff --git a/fs/jfs/jfs_acl.h b/fs/jfs/jfs_acl.h
index b07bd417ef85..54e07559878d 100644
--- a/fs/jfs/jfs_acl.h
+++ b/fs/jfs/jfs_acl.h
@@ -22,7 +22,7 @@
 
 int jfs_check_acl(struct inode *, int);
 int jfs_init_acl(tid_t, struct inode *, struct inode *);
-int jfs_setattr(struct dentry *, struct iattr *);
+int jfs_acl_chmod(struct inode *inode);
 
 #else
 
@@ -32,5 +32,10 @@ static inline int jfs_init_acl(tid_t tid, struct inode *inode,
 	return 0;
 }
 
+static inline int jfs_acl_chmod(struct inode *inode)
+{
+	return 0;
+}
+
 #endif
 #endif		/* _H_JFS_ACL */
diff --git a/fs/jfs/jfs_dtree.c b/fs/jfs/jfs_dtree.c
index 925871e9887b..0e4623be70ce 100644
--- a/fs/jfs/jfs_dtree.c
+++ b/fs/jfs/jfs_dtree.c
@@ -381,10 +381,10 @@ static u32 add_index(tid_t tid, struct inode *ip, s64 bn, int slot)
 		 * It's time to move the inline table to an external
 		 * page and begin to build the xtree
 		 */
-		if (vfs_dq_alloc_block(ip, sbi->nbperpage))
+		if (dquot_alloc_block(ip, sbi->nbperpage))
 			goto clean_up;
 		if (dbAlloc(ip, 0, sbi->nbperpage, &xaddr)) {
-			vfs_dq_free_block(ip, sbi->nbperpage);
+			dquot_free_block(ip, sbi->nbperpage);
 			goto clean_up;
 		}
 
@@ -408,7 +408,7 @@ static u32 add_index(tid_t tid, struct inode *ip, s64 bn, int slot)
 			memcpy(&jfs_ip->i_dirtable, temp_table,
 			       sizeof (temp_table));
 			dbFree(ip, xaddr, sbi->nbperpage);
-			vfs_dq_free_block(ip, sbi->nbperpage);
+			dquot_free_block(ip, sbi->nbperpage);
 			goto clean_up;
 		}
 		ip->i_size = PSIZE;
@@ -1027,10 +1027,9 @@ static int dtSplitUp(tid_t tid,
 			n = xlen;
 
 		/* Allocate blocks to quota. */
-		if (vfs_dq_alloc_block(ip, n)) {
-			rc = -EDQUOT;
+		rc = dquot_alloc_block(ip, n);
+		if (rc)
 			goto extendOut;
-		}
 		quota_allocation += n;
 
 		if ((rc = dbReAlloc(sbi->ipbmap, xaddr, (s64) xlen,
@@ -1308,7 +1307,7 @@ static int dtSplitUp(tid_t tid,
 
 	/* Rollback quota allocation */
 	if (rc && quota_allocation)
-		vfs_dq_free_block(ip, quota_allocation);
+		dquot_free_block(ip, quota_allocation);
 
       dtSplitUp_Exit:
 
@@ -1369,9 +1368,10 @@ static int dtSplitPage(tid_t tid, struct inode *ip, struct dtsplit * split,
 		return -EIO;
 
 	/* Allocate blocks to quota. */
-	if (vfs_dq_alloc_block(ip, lengthPXD(pxd))) {
+	rc = dquot_alloc_block(ip, lengthPXD(pxd));
+	if (rc) {
 		release_metapage(rmp);
-		return -EDQUOT;
+		return rc;
 	}
 
 	jfs_info("dtSplitPage: ip:0x%p smp:0x%p rmp:0x%p", ip, smp, rmp);
@@ -1892,6 +1892,7 @@ static int dtSplitRoot(tid_t tid,
 	struct dt_lock *dtlck;
 	struct tlock *tlck;
 	struct lv *lv;
+	int rc;
 
 	/* get split root page */
 	smp = split->mp;
@@ -1916,9 +1917,10 @@ static int dtSplitRoot(tid_t tid,
 	rp = rmp->data;
 
 	/* Allocate blocks to quota. */
-	if (vfs_dq_alloc_block(ip, lengthPXD(pxd))) {
+	rc = dquot_alloc_block(ip, lengthPXD(pxd));
+	if (rc) {
 		release_metapage(rmp);
-		return -EDQUOT;
+		return rc;
 	}
 
 	BT_MARK_DIRTY(rmp, ip);
@@ -2287,7 +2289,7 @@ static int dtDeleteUp(tid_t tid, struct inode *ip,
 	xlen = lengthPXD(&fp->header.self);
 
 	/* Free quota allocation. */
-	vfs_dq_free_block(ip, xlen);
+	dquot_free_block(ip, xlen);
 
 	/* free/invalidate its buffer page */
 	discard_metapage(fmp);
@@ -2363,7 +2365,7 @@ static int dtDeleteUp(tid_t tid, struct inode *ip,
 				xlen = lengthPXD(&p->header.self);
 
 				/* Free quota allocation */
-				vfs_dq_free_block(ip, xlen);
+				dquot_free_block(ip, xlen);
 
 				/* free/invalidate its buffer page */
 				discard_metapage(mp);
diff --git a/fs/jfs/jfs_extent.c b/fs/jfs/jfs_extent.c
index 41d6045dbeb0..5d3bbd10f8db 100644
--- a/fs/jfs/jfs_extent.c
+++ b/fs/jfs/jfs_extent.c
@@ -141,10 +141,11 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, bool abnr)
 	}
 
 	/* Allocate blocks to quota. */
-	if (vfs_dq_alloc_block(ip, nxlen)) {
+	rc = dquot_alloc_block(ip, nxlen);
+	if (rc) {
 		dbFree(ip, nxaddr, (s64) nxlen);
 		mutex_unlock(&JFS_IP(ip)->commit_mutex);
-		return -EDQUOT;
+		return rc;
 	}
 
 	/* determine the value of the extent flag */
@@ -164,7 +165,7 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, bool abnr)
 	 */
 	if (rc) {
 		dbFree(ip, nxaddr, nxlen);
-		vfs_dq_free_block(ip, nxlen);
+		dquot_free_block(ip, nxlen);
 		mutex_unlock(&JFS_IP(ip)->commit_mutex);
 		return (rc);
 	}
@@ -256,10 +257,11 @@ int extRealloc(struct inode *ip, s64 nxlen, xad_t * xp, bool abnr)
 		goto exit;
 
 	/* Allocat blocks to quota. */
-	if (vfs_dq_alloc_block(ip, nxlen)) {
+	rc = dquot_alloc_block(ip, nxlen);
+	if (rc) {
 		dbFree(ip, nxaddr, (s64) nxlen);
 		mutex_unlock(&JFS_IP(ip)->commit_mutex);
-		return -EDQUOT;
+		return rc;
 	}
 
 	delta = nxlen - xlen;
@@ -297,7 +299,7 @@ int extRealloc(struct inode *ip, s64 nxlen, xad_t * xp, bool abnr)
 		/* extend the extent */
 		if ((rc = xtExtend(0, ip, xoff + xlen, (int) nextend, 0))) {
 			dbFree(ip, xaddr + xlen, delta);
-			vfs_dq_free_block(ip, nxlen);
+			dquot_free_block(ip, nxlen);
 			goto exit;
 		}
 	} else {
@@ -308,7 +310,7 @@ int extRealloc(struct inode *ip, s64 nxlen, xad_t * xp, bool abnr)
 		 */
 		if ((rc = xtTailgate(0, ip, xoff, (int) ntail, nxaddr, 0))) {
 			dbFree(ip, nxaddr, nxlen);
-			vfs_dq_free_block(ip, nxlen);
+			dquot_free_block(ip, nxlen);
 			goto exit;
 		}
 	}
diff --git a/fs/jfs/jfs_inode.c b/fs/jfs/jfs_inode.c
index dc0e02159ac9..829921b67765 100644
--- a/fs/jfs/jfs_inode.c
+++ b/fs/jfs/jfs_inode.c
@@ -116,10 +116,10 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
 	/*
 	 * Allocate inode to quota.
 	 */
-	if (vfs_dq_alloc_inode(inode)) {
-		rc = -EDQUOT;
+	dquot_initialize(inode);
+	rc = dquot_alloc_inode(inode);
+	if (rc)
 		goto fail_drop;
-	}
 
 	inode->i_mode = mode;
 	/* inherit flags from parent */
@@ -162,7 +162,7 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
 	return inode;
 
 fail_drop:
-	vfs_dq_drop(inode);
+	dquot_drop(inode);
 	inode->i_flags |= S_NOQUOTA;
 fail_unlock:
 	inode->i_nlink = 0;
diff --git a/fs/jfs/jfs_inode.h b/fs/jfs/jfs_inode.h
index 1eff7db34d63..79e2c79661df 100644
--- a/fs/jfs/jfs_inode.h
+++ b/fs/jfs/jfs_inode.h
@@ -26,7 +26,7 @@ extern long jfs_ioctl(struct file *, unsigned int, unsigned long);
 extern long jfs_compat_ioctl(struct file *, unsigned int, unsigned long);
 extern struct inode *jfs_iget(struct super_block *, unsigned long);
 extern int jfs_commit_inode(struct inode *, int);
-extern int jfs_write_inode(struct inode*, int);
+extern int jfs_write_inode(struct inode *, struct writeback_control *);
 extern void jfs_delete_inode(struct inode *);
 extern void jfs_dirty_inode(struct inode *);
 extern void jfs_truncate(struct inode *);
@@ -40,6 +40,7 @@ extern struct dentry *jfs_fh_to_parent(struct super_block *sb, struct fid *fid,
 	int fh_len, int fh_type);
 extern void jfs_set_inode_flags(struct inode *);
 extern int jfs_get_block(struct inode *, sector_t, struct buffer_head *, int);
+extern int jfs_setattr(struct dentry *, struct iattr *);
 
 extern const struct address_space_operations jfs_aops;
 extern const struct inode_operations jfs_dir_inode_operations;
diff --git a/fs/jfs/jfs_xtree.c b/fs/jfs/jfs_xtree.c
index d654a6458648..6c50871e6220 100644
--- a/fs/jfs/jfs_xtree.c
+++ b/fs/jfs/jfs_xtree.c
@@ -585,10 +585,10 @@ int xtInsert(tid_t tid,		/* transaction id */
 			hint = addressXAD(xad) + lengthXAD(xad) - 1;
 		} else
 			hint = 0;
-		if ((rc = vfs_dq_alloc_block(ip, xlen)))
+		if ((rc = dquot_alloc_block(ip, xlen)))
 			goto out;
 		if ((rc = dbAlloc(ip, hint, (s64) xlen, &xaddr))) {
-			vfs_dq_free_block(ip, xlen);
+			dquot_free_block(ip, xlen);
 			goto out;
 		}
 	}
@@ -617,7 +617,7 @@ int xtInsert(tid_t tid,		/* transaction id */
 			/* undo data extent allocation */
 			if (*xaddrp == 0) {
 				dbFree(ip, xaddr, (s64) xlen);
-				vfs_dq_free_block(ip, xlen);
+				dquot_free_block(ip, xlen);
 			}
 			return rc;
 		}
@@ -985,10 +985,9 @@ xtSplitPage(tid_t tid, struct inode *ip,
 	rbn = addressPXD(pxd);
 
 	/* Allocate blocks to quota. */
-	if (vfs_dq_alloc_block(ip, lengthPXD(pxd))) {
-		rc = -EDQUOT;
+	rc = dquot_alloc_block(ip, lengthPXD(pxd));
+	if (rc)
 		goto clean_up;
-	}
 
 	quota_allocation += lengthPXD(pxd);
 
@@ -1195,7 +1194,7 @@ xtSplitPage(tid_t tid, struct inode *ip,
 
 	/* Rollback quota allocation. */
 	if (quota_allocation)
-		vfs_dq_free_block(ip, quota_allocation);
+		dquot_free_block(ip, quota_allocation);
 
 	return (rc);
 }
@@ -1235,6 +1234,7 @@ xtSplitRoot(tid_t tid,
 	struct pxdlist *pxdlist;
 	struct tlock *tlck;
 	struct xtlock *xtlck;
+	int rc;
 
 	sp = &JFS_IP(ip)->i_xtroot;
 
@@ -1252,9 +1252,10 @@ xtSplitRoot(tid_t tid,
 		return -EIO;
 
 	/* Allocate blocks to quota. */
-	if (vfs_dq_alloc_block(ip, lengthPXD(pxd))) {
+	rc = dquot_alloc_block(ip, lengthPXD(pxd));
+	if (rc) {
 		release_metapage(rmp);
-		return -EDQUOT;
+		return rc;
 	}
 
 	jfs_info("xtSplitRoot: ip:0x%p rmp:0x%p", ip, rmp);
@@ -3680,7 +3681,7 @@ s64 xtTruncate(tid_t tid, struct inode *ip, s64 newsize, int flag)
 		ip->i_size = newsize;
 
 	/* update quota allocation to reflect freed blocks */
-	vfs_dq_free_block(ip, nfreed);
+	dquot_free_block(ip, nfreed);
 
 	/*
 	 * free tlock of invalidated pages
diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index c79a4270f083..4a3e9f39c21d 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -85,6 +85,8 @@ static int jfs_create(struct inode *dip, struct dentry *dentry, int mode,
 
 	jfs_info("jfs_create: dip:0x%p name:%s", dip, dentry->d_name.name);
 
+	dquot_initialize(dip);
+
 	/*
 	 * search parent directory for entry/freespace
 	 * (dtSearch() returns parent directory page pinned)
@@ -215,6 +217,8 @@ static int jfs_mkdir(struct inode *dip, struct dentry *dentry, int mode)
 
 	jfs_info("jfs_mkdir: dip:0x%p name:%s", dip, dentry->d_name.name);
 
+	dquot_initialize(dip);
+
 	/* link count overflow on parent directory ? */
 	if (dip->i_nlink == JFS_LINK_MAX) {
 		rc = -EMLINK;
@@ -356,7 +360,8 @@ static int jfs_rmdir(struct inode *dip, struct dentry *dentry)
 	jfs_info("jfs_rmdir: dip:0x%p name:%s", dip, dentry->d_name.name);
 
 	/* Init inode for quota operations. */
-	vfs_dq_init(ip);
+	dquot_initialize(dip);
+	dquot_initialize(ip);
 
 	/* directory must be empty to be removed */
 	if (!dtEmpty(ip)) {
@@ -483,7 +488,8 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry)
 	jfs_info("jfs_unlink: dip:0x%p name:%s", dip, dentry->d_name.name);
 
 	/* Init inode for quota operations. */
-	vfs_dq_init(ip);
+	dquot_initialize(dip);
+	dquot_initialize(ip);
 
 	if ((rc = get_UCSname(&dname, dentry)))
 		goto out;
@@ -805,6 +811,8 @@ static int jfs_link(struct dentry *old_dentry,
 	if (ip->i_nlink == 0)
 		return -ENOENT;
 
+	dquot_initialize(dir);
+
 	tid = txBegin(ip->i_sb, 0);
 
 	mutex_lock_nested(&JFS_IP(dir)->commit_mutex, COMMIT_MUTEX_PARENT);
@@ -896,6 +904,8 @@ static int jfs_symlink(struct inode *dip, struct dentry *dentry,
 
 	jfs_info("jfs_symlink: dip:0x%p name:%s", dip, name);
 
+	dquot_initialize(dip);
+
 	ssize = strlen(name) + 1;
 
 	/*
@@ -1087,6 +1097,9 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	jfs_info("jfs_rename: %s %s", old_dentry->d_name.name,
 		 new_dentry->d_name.name);
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	old_ip = old_dentry->d_inode;
 	new_ip = new_dentry->d_inode;
 
@@ -1136,7 +1149,7 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	} else if (new_ip) {
 		IWRITE_LOCK(new_ip, RDWRLOCK_NORMAL);
 		/* Init inode for quota operations. */
-		vfs_dq_init(new_ip);
+		dquot_initialize(new_ip);
 	}
 
 	/*
@@ -1360,6 +1373,8 @@ static int jfs_mknod(struct inode *dir, struct dentry *dentry,
 
 	jfs_info("jfs_mknod: %s", dentry->d_name.name);
 
+	dquot_initialize(dir);
+
 	if ((rc = get_UCSname(&dname, dentry)))
 		goto out;
 
@@ -1541,8 +1556,8 @@ const struct inode_operations jfs_dir_inode_operations = {
 	.getxattr	= jfs_getxattr,
 	.listxattr	= jfs_listxattr,
 	.removexattr	= jfs_removexattr,
-#ifdef CONFIG_JFS_POSIX_ACL
 	.setattr	= jfs_setattr,
+#ifdef CONFIG_JFS_POSIX_ACL
 	.check_acl	= jfs_check_acl,
 #endif
 };
diff --git a/fs/jfs/super.c b/fs/jfs/super.c
index d929a822a74e..266699deb1c6 100644
--- a/fs/jfs/super.c
+++ b/fs/jfs/super.c
@@ -131,6 +131,11 @@ static void jfs_destroy_inode(struct inode *inode)
 	kmem_cache_free(jfs_inode_cachep, ji);
 }
 
+static void jfs_clear_inode(struct inode *inode)
+{
+	dquot_drop(inode);
+}
+
 static int jfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 {
 	struct jfs_sb_info *sbi = JFS_SBI(dentry->d_sb);
@@ -745,6 +750,7 @@ static const struct super_operations jfs_super_operations = {
 	.dirty_inode	= jfs_dirty_inode,
 	.write_inode	= jfs_write_inode,
 	.delete_inode	= jfs_delete_inode,
+	.clear_inode	= jfs_clear_inode,
 	.put_super	= jfs_put_super,
 	.sync_fs	= jfs_sync_fs,
 	.freeze_fs	= jfs_freeze,
diff --git a/fs/jfs/xattr.c b/fs/jfs/xattr.c
index fad364548bc9..1f594ab21895 100644
--- a/fs/jfs/xattr.c
+++ b/fs/jfs/xattr.c
@@ -260,14 +260,14 @@ static int ea_write(struct inode *ip, struct jfs_ea_list *ealist, int size,
 	nblocks = (size + (sb->s_blocksize - 1)) >> sb->s_blocksize_bits;
 
 	/* Allocate new blocks to quota. */
-	if (vfs_dq_alloc_block(ip, nblocks)) {
-		return -EDQUOT;
-	}
+	rc = dquot_alloc_block(ip, nblocks);
+	if (rc)
+		return rc;
 
 	rc = dbAlloc(ip, INOHINT(ip), nblocks, &blkno);
 	if (rc) {
 		/*Rollback quota allocation. */
-		vfs_dq_free_block(ip, nblocks);
+		dquot_free_block(ip, nblocks);
 		return rc;
 	}
 
@@ -332,7 +332,7 @@ static int ea_write(struct inode *ip, struct jfs_ea_list *ealist, int size,
 
       failed:
 	/* Rollback quota allocation. */
-	vfs_dq_free_block(ip, nblocks);
+	dquot_free_block(ip, nblocks);
 
 	dbFree(ip, blkno, nblocks);
 	return rc;
@@ -538,7 +538,8 @@ static int ea_get(struct inode *inode, struct ea_buffer *ea_buf, int min_size)
 
 	if (blocks_needed > current_blocks) {
 		/* Allocate new blocks to quota. */
-		if (vfs_dq_alloc_block(inode, blocks_needed))
+		rc = dquot_alloc_block(inode, blocks_needed);
+		if (rc)
 			return -EDQUOT;
 
 		quota_allocation = blocks_needed;
@@ -602,7 +603,7 @@ static int ea_get(struct inode *inode, struct ea_buffer *ea_buf, int min_size)
       clean_up:
 	/* Rollback quota allocation */
 	if (quota_allocation)
-		vfs_dq_free_block(inode, quota_allocation);
+		dquot_free_block(inode, quota_allocation);
 
 	return (rc);
 }
@@ -677,7 +678,7 @@ static int ea_put(tid_t tid, struct inode *inode, struct ea_buffer *ea_buf,
 
 	/* If old blocks exist, they must be removed from quota allocation. */
 	if (old_blocks)
-		vfs_dq_free_block(inode, old_blocks);
+		dquot_free_block(inode, old_blocks);
 
 	inode->i_ctime = CURRENT_TIME;
 
diff --git a/fs/libfs.c b/fs/libfs.c
index 6e8d17e1dc4c..9e50bcf55857 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -338,28 +338,14 @@ int simple_readpage(struct file *file, struct page *page)
 	return 0;
 }
 
-int simple_prepare_write(struct file *file, struct page *page,
-			unsigned from, unsigned to)
-{
-	if (!PageUptodate(page)) {
-		if (to - from != PAGE_CACHE_SIZE)
-			zero_user_segments(page,
-				0, from,
-				to, PAGE_CACHE_SIZE);
-	}
-	return 0;
-}
-
 int simple_write_begin(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned flags,
 			struct page **pagep, void **fsdata)
 {
 	struct page *page;
 	pgoff_t index;
-	unsigned from;
 
 	index = pos >> PAGE_CACHE_SHIFT;
-	from = pos & (PAGE_CACHE_SIZE - 1);
 
 	page = grab_cache_page_write_begin(mapping, index, flags);
 	if (!page)
@@ -367,43 +353,59 @@ int simple_write_begin(struct file *file, struct address_space *mapping,
 
 	*pagep = page;
 
-	return simple_prepare_write(file, page, from, from+len);
-}
-
-static int simple_commit_write(struct file *file, struct page *page,
-			       unsigned from, unsigned to)
-{
-	struct inode *inode = page->mapping->host;
-	loff_t pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
+	if (!PageUptodate(page) && (len != PAGE_CACHE_SIZE)) {
+		unsigned from = pos & (PAGE_CACHE_SIZE - 1);
 
-	if (!PageUptodate(page))
-		SetPageUptodate(page);
-	/*
-	 * No need to use i_size_read() here, the i_size
-	 * cannot change under us because we hold the i_mutex.
-	 */
-	if (pos > inode->i_size)
-		i_size_write(inode, pos);
-	set_page_dirty(page);
+		zero_user_segments(page, 0, from, from + len, PAGE_CACHE_SIZE);
+	}
 	return 0;
 }
 
+/**
+ * simple_write_end - .write_end helper for non-block-device FSes
+ * @available: See .write_end of address_space_operations
+ * @file: 		"
+ * @mapping: 		"
+ * @pos: 		"
+ * @len: 		"
+ * @copied: 		"
+ * @page: 		"
+ * @fsdata: 		"
+ *
+ * simple_write_end does the minimum needed for updating a page after writing is
+ * done. It has the same API signature as the .write_end of
+ * address_space_operations vector. So it can just be set onto .write_end for
+ * FSes that don't need any other processing. i_mutex is assumed to be held.
+ * Block based filesystems should use generic_write_end().
+ * NOTE: Even though i_size might get updated by this function, mark_inode_dirty
+ * is not called, so a filesystem that actually does store data in .write_inode
+ * should extend on what's done here with a call to mark_inode_dirty() in the
+ * case that i_size has changed.
+ */
 int simple_write_end(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned copied,
 			struct page *page, void *fsdata)
 {
-	unsigned from = pos & (PAGE_CACHE_SIZE - 1);
+	struct inode *inode = page->mapping->host;
+	loff_t last_pos = pos + copied;
 
 	/* zero the stale part of the page if we did a short copy */
 	if (copied < len) {
-		void *kaddr = kmap_atomic(page, KM_USER0);
-		memset(kaddr + from + copied, 0, len - copied);
-		flush_dcache_page(page);
-		kunmap_atomic(kaddr, KM_USER0);
+		unsigned from = pos & (PAGE_CACHE_SIZE - 1);
+
+		zero_user(page, from + copied, len - copied);
 	}
 
-	simple_commit_write(file, page, from, from+copied);
+	if (!PageUptodate(page))
+		SetPageUptodate(page);
+	/*
+	 * No need to use i_size_read() here, the i_size
+	 * cannot change under us because we hold the i_mutex.
+	 */
+	if (last_pos > inode->i_size)
+		i_size_write(inode, last_pos);
 
+	set_page_dirty(page);
 	unlock_page(page);
 	page_cache_release(page);
 
@@ -853,7 +855,6 @@ EXPORT_SYMBOL(simple_getattr);
 EXPORT_SYMBOL(simple_link);
 EXPORT_SYMBOL(simple_lookup);
 EXPORT_SYMBOL(simple_pin_fs);
-EXPORT_UNUSED_SYMBOL(simple_prepare_write);
 EXPORT_SYMBOL(simple_readpage);
 EXPORT_SYMBOL(simple_release_fs);
 EXPORT_SYMBOL(simple_rename);
diff --git a/fs/lockd/host.c b/fs/lockd/host.c
index 4600c2037b8b..bb464d12104c 100644
--- a/fs/lockd/host.c
+++ b/fs/lockd/host.c
@@ -479,8 +479,8 @@ again:	mutex_lock(&nlm_host_mutex);
 			}
 		}
 	}
-
 	mutex_unlock(&nlm_host_mutex);
+	nsm_release(nsm);
 }
 
 /*
diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c
index f956651d0f65..fefa4df3f005 100644
--- a/fs/lockd/mon.c
+++ b/fs/lockd/mon.c
@@ -349,9 +349,9 @@ retry:
  * nsm_reboot_lookup - match NLMPROC_SM_NOTIFY arguments to an nsm_handle
  * @info: pointer to NLMPROC_SM_NOTIFY arguments
  *
- * Returns a matching nsm_handle if found in the nsm cache; the returned
- * nsm_handle's reference count is bumped and sm_monitored is cleared.
- * Otherwise returns NULL if some error occurred.
+ * Returns a matching nsm_handle if found in the nsm cache. The returned
+ * nsm_handle's reference count is bumped. Otherwise returns NULL if some
+ * error occurred.
  */
 struct nsm_handle *nsm_reboot_lookup(const struct nlm_reboot *info)
 {
@@ -370,12 +370,6 @@ struct nsm_handle *nsm_reboot_lookup(const struct nlm_reboot *info)
 	atomic_inc(&cached->sm_count);
 	spin_unlock(&nsm_lock);
 
-	/*
-	 * During subsequent lock activity, force a fresh
-	 * notification to be set up for this host.
-	 */
-	cached->sm_monitored = 0;
-
 	dprintk("lockd: host %s (%s) rebooted, cnt %d\n",
 			cached->sm_name, cached->sm_addrbuf,
 			atomic_read(&cached->sm_count));
diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c
index e50cfa3d9654..7d150517ddf0 100644
--- a/fs/lockd/svc.c
+++ b/fs/lockd/svc.c
@@ -243,11 +243,9 @@ static int make_socks(struct svc_serv *serv)
 	if (err < 0)
 		goto out_err;
 
-#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
 	err = create_lockd_family(serv, PF_INET6);
 	if (err < 0 && err != -EAFNOSUPPORT)
 		goto out_err;
-#endif	/* CONFIG_IPV6 || CONFIG_IPV6_MODULE */
 
 	warned = 0;
 	return 0;
diff --git a/fs/locks.c b/fs/locks.c
index a8794f233bc9..ae9ded026b7c 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -1182,8 +1182,9 @@ int __break_lease(struct inode *inode, unsigned int mode)
 	struct file_lock *fl;
 	unsigned long break_time;
 	int i_have_this_lease = 0;
+	int want_write = (mode & O_ACCMODE) != O_RDONLY;
 
-	new_fl = lease_alloc(NULL, mode & FMODE_WRITE ? F_WRLCK : F_RDLCK);
+	new_fl = lease_alloc(NULL, want_write ? F_WRLCK : F_RDLCK);
 
 	lock_kernel();
 
@@ -1197,7 +1198,7 @@ int __break_lease(struct inode *inode, unsigned int mode)
 		if (fl->fl_owner == current->files)
 			i_have_this_lease = 1;
 
-	if (mode & FMODE_WRITE) {
+	if (want_write) {
 		/* If we want write access, we have to revoke any lease. */
 		future = F_UNLCK | F_INPROGRESS;
 	} else if (flock->fl_type & F_INPROGRESS) {
diff --git a/fs/logfs/Kconfig b/fs/logfs/Kconfig
new file mode 100644
index 000000000000..daf9a9b32dd3
--- /dev/null
+++ b/fs/logfs/Kconfig
@@ -0,0 +1,17 @@
+config LOGFS
+	tristate "LogFS file system (EXPERIMENTAL)"
+	depends on (MTD || BLOCK) && EXPERIMENTAL
+	select ZLIB_INFLATE
+	select ZLIB_DEFLATE
+	select CRC32
+	select BTREE
+	help
+	  Flash filesystem aimed to scale efficiently to large devices.
+	  In comparison to JFFS2 it offers significantly faster mount
+	  times and potentially less RAM usage, although the latter has
+	  not been measured yet.
+
+	  In its current state it is still very experimental and should
+	  not be used for other than testing purposes.
+
+	  If unsure, say N.
diff --git a/fs/logfs/Makefile b/fs/logfs/Makefile
new file mode 100644
index 000000000000..4820027787ee
--- /dev/null
+++ b/fs/logfs/Makefile
@@ -0,0 +1,13 @@
+obj-$(CONFIG_LOGFS)	+= logfs.o
+
+logfs-y	+= compr.o
+logfs-y	+= dir.o
+logfs-y	+= file.o
+logfs-y	+= gc.o
+logfs-y	+= inode.o
+logfs-y	+= journal.o
+logfs-y	+= readwrite.o
+logfs-y	+= segment.o
+logfs-y	+= super.o
+logfs-$(CONFIG_BLOCK)	+= dev_bdev.o
+logfs-$(CONFIG_MTD)	+= dev_mtd.o
diff --git a/fs/logfs/compr.c b/fs/logfs/compr.c
new file mode 100644
index 000000000000..44bbfd249abc
--- /dev/null
+++ b/fs/logfs/compr.c
@@ -0,0 +1,95 @@
+/*
+ * fs/logfs/compr.c	- compression routines
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+#include <linux/vmalloc.h>
+#include <linux/zlib.h>
+
+#define COMPR_LEVEL 3
+
+static DEFINE_MUTEX(compr_mutex);
+static struct z_stream_s stream;
+
+int logfs_compress(void *in, void *out, size_t inlen, size_t outlen)
+{
+	int err, ret;
+
+	ret = -EIO;
+	mutex_lock(&compr_mutex);
+	err = zlib_deflateInit(&stream, COMPR_LEVEL);
+	if (err != Z_OK)
+		goto error;
+
+	stream.next_in = in;
+	stream.avail_in = inlen;
+	stream.total_in = 0;
+	stream.next_out = out;
+	stream.avail_out = outlen;
+	stream.total_out = 0;
+
+	err = zlib_deflate(&stream, Z_FINISH);
+	if (err != Z_STREAM_END)
+		goto error;
+
+	err = zlib_deflateEnd(&stream);
+	if (err != Z_OK)
+		goto error;
+
+	if (stream.total_out >= stream.total_in)
+		goto error;
+
+	ret = stream.total_out;
+error:
+	mutex_unlock(&compr_mutex);
+	return ret;
+}
+
+int logfs_uncompress(void *in, void *out, size_t inlen, size_t outlen)
+{
+	int err, ret;
+
+	ret = -EIO;
+	mutex_lock(&compr_mutex);
+	err = zlib_inflateInit(&stream);
+	if (err != Z_OK)
+		goto error;
+
+	stream.next_in = in;
+	stream.avail_in = inlen;
+	stream.total_in = 0;
+	stream.next_out = out;
+	stream.avail_out = outlen;
+	stream.total_out = 0;
+
+	err = zlib_inflate(&stream, Z_FINISH);
+	if (err != Z_STREAM_END)
+		goto error;
+
+	err = zlib_inflateEnd(&stream);
+	if (err != Z_OK)
+		goto error;
+
+	ret = 0;
+error:
+	mutex_unlock(&compr_mutex);
+	return ret;
+}
+
+int __init logfs_compr_init(void)
+{
+	size_t size = max(zlib_deflate_workspacesize(),
+			zlib_inflate_workspacesize());
+	stream.workspace = vmalloc(size);
+	if (!stream.workspace)
+		return -ENOMEM;
+	return 0;
+}
+
+void logfs_compr_exit(void)
+{
+	vfree(stream.workspace);
+}
diff --git a/fs/logfs/dev_bdev.c b/fs/logfs/dev_bdev.c
new file mode 100644
index 000000000000..9718c22f186d
--- /dev/null
+++ b/fs/logfs/dev_bdev.c
@@ -0,0 +1,327 @@
+/*
+ * fs/logfs/dev_bdev.c	- Device access methods for block devices
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+#include <linux/bio.h>
+#include <linux/blkdev.h>
+#include <linux/buffer_head.h>
+
+#define PAGE_OFS(ofs) ((ofs) & (PAGE_SIZE-1))
+
+static void request_complete(struct bio *bio, int err)
+{
+	complete((struct completion *)bio->bi_private);
+}
+
+static int sync_request(struct page *page, struct block_device *bdev, int rw)
+{
+	struct bio bio;
+	struct bio_vec bio_vec;
+	struct completion complete;
+
+	bio_init(&bio);
+	bio.bi_io_vec = &bio_vec;
+	bio_vec.bv_page = page;
+	bio_vec.bv_len = PAGE_SIZE;
+	bio_vec.bv_offset = 0;
+	bio.bi_vcnt = 1;
+	bio.bi_idx = 0;
+	bio.bi_size = PAGE_SIZE;
+	bio.bi_bdev = bdev;
+	bio.bi_sector = page->index * (PAGE_SIZE >> 9);
+	init_completion(&complete);
+	bio.bi_private = &complete;
+	bio.bi_end_io = request_complete;
+
+	submit_bio(rw, &bio);
+	generic_unplug_device(bdev_get_queue(bdev));
+	wait_for_completion(&complete);
+	return test_bit(BIO_UPTODATE, &bio.bi_flags) ? 0 : -EIO;
+}
+
+static int bdev_readpage(void *_sb, struct page *page)
+{
+	struct super_block *sb = _sb;
+	struct block_device *bdev = logfs_super(sb)->s_bdev;
+	int err;
+
+	err = sync_request(page, bdev, READ);
+	if (err) {
+		ClearPageUptodate(page);
+		SetPageError(page);
+	} else {
+		SetPageUptodate(page);
+		ClearPageError(page);
+	}
+	unlock_page(page);
+	return err;
+}
+
+static DECLARE_WAIT_QUEUE_HEAD(wq);
+
+static void writeseg_end_io(struct bio *bio, int err)
+{
+	const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
+	struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1;
+	struct super_block *sb = bio->bi_private;
+	struct logfs_super *super = logfs_super(sb);
+	struct page *page;
+
+	BUG_ON(!uptodate); /* FIXME: Retry io or write elsewhere */
+	BUG_ON(err);
+	BUG_ON(bio->bi_vcnt == 0);
+	do {
+		page = bvec->bv_page;
+		if (--bvec >= bio->bi_io_vec)
+			prefetchw(&bvec->bv_page->flags);
+
+		end_page_writeback(page);
+	} while (bvec >= bio->bi_io_vec);
+	bio_put(bio);
+	if (atomic_dec_and_test(&super->s_pending_writes))
+		wake_up(&wq);
+}
+
+static int __bdev_writeseg(struct super_block *sb, u64 ofs, pgoff_t index,
+		size_t nr_pages)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	struct bio *bio;
+	struct page *page;
+	struct request_queue *q = bdev_get_queue(sb->s_bdev);
+	unsigned int max_pages = queue_max_hw_sectors(q) >> (PAGE_SHIFT - 9);
+	int i;
+
+	bio = bio_alloc(GFP_NOFS, max_pages);
+	BUG_ON(!bio); /* FIXME: handle this */
+
+	for (i = 0; i < nr_pages; i++) {
+		if (i >= max_pages) {
+			/* Block layer cannot split bios :( */
+			bio->bi_vcnt = i;
+			bio->bi_idx = 0;
+			bio->bi_size = i * PAGE_SIZE;
+			bio->bi_bdev = super->s_bdev;
+			bio->bi_sector = ofs >> 9;
+			bio->bi_private = sb;
+			bio->bi_end_io = writeseg_end_io;
+			atomic_inc(&super->s_pending_writes);
+			submit_bio(WRITE, bio);
+
+			ofs += i * PAGE_SIZE;
+			index += i;
+			nr_pages -= i;
+			i = 0;
+
+			bio = bio_alloc(GFP_NOFS, max_pages);
+			BUG_ON(!bio);
+		}
+		page = find_lock_page(mapping, index + i);
+		BUG_ON(!page);
+		bio->bi_io_vec[i].bv_page = page;
+		bio->bi_io_vec[i].bv_len = PAGE_SIZE;
+		bio->bi_io_vec[i].bv_offset = 0;
+
+		BUG_ON(PageWriteback(page));
+		set_page_writeback(page);
+		unlock_page(page);
+	}
+	bio->bi_vcnt = nr_pages;
+	bio->bi_idx = 0;
+	bio->bi_size = nr_pages * PAGE_SIZE;
+	bio->bi_bdev = super->s_bdev;
+	bio->bi_sector = ofs >> 9;
+	bio->bi_private = sb;
+	bio->bi_end_io = writeseg_end_io;
+	atomic_inc(&super->s_pending_writes);
+	submit_bio(WRITE, bio);
+	return 0;
+}
+
+static void bdev_writeseg(struct super_block *sb, u64 ofs, size_t len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int head;
+
+	BUG_ON(super->s_flags & LOGFS_SB_FLAG_RO);
+
+	if (len == 0) {
+		/* This can happen when the object fit perfectly into a
+		 * segment, the segment gets written per sync and subsequently
+		 * closed.
+		 */
+		return;
+	}
+	head = ofs & (PAGE_SIZE - 1);
+	if (head) {
+		ofs -= head;
+		len += head;
+	}
+	len = PAGE_ALIGN(len);
+	__bdev_writeseg(sb, ofs, ofs >> PAGE_SHIFT, len >> PAGE_SHIFT);
+	generic_unplug_device(bdev_get_queue(logfs_super(sb)->s_bdev));
+}
+
+
+static void erase_end_io(struct bio *bio, int err) 
+{ 
+	const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); 
+	struct super_block *sb = bio->bi_private; 
+	struct logfs_super *super = logfs_super(sb); 
+
+	BUG_ON(!uptodate); /* FIXME: Retry io or write elsewhere */ 
+	BUG_ON(err); 
+	BUG_ON(bio->bi_vcnt == 0); 
+	bio_put(bio); 
+	if (atomic_dec_and_test(&super->s_pending_writes))
+		wake_up(&wq); 
+} 
+
+static int do_erase(struct super_block *sb, u64 ofs, pgoff_t index,
+		size_t nr_pages)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct bio *bio;
+	struct request_queue *q = bdev_get_queue(sb->s_bdev);
+	unsigned int max_pages = queue_max_hw_sectors(q) >> (PAGE_SHIFT - 9);
+	int i;
+
+	bio = bio_alloc(GFP_NOFS, max_pages);
+	BUG_ON(!bio); /* FIXME: handle this */
+
+	for (i = 0; i < nr_pages; i++) {
+		if (i >= max_pages) {
+			/* Block layer cannot split bios :( */
+			bio->bi_vcnt = i;
+			bio->bi_idx = 0;
+			bio->bi_size = i * PAGE_SIZE;
+			bio->bi_bdev = super->s_bdev;
+			bio->bi_sector = ofs >> 9;
+			bio->bi_private = sb;
+			bio->bi_end_io = erase_end_io;
+			atomic_inc(&super->s_pending_writes);
+			submit_bio(WRITE, bio);
+
+			ofs += i * PAGE_SIZE;
+			index += i;
+			nr_pages -= i;
+			i = 0;
+
+			bio = bio_alloc(GFP_NOFS, max_pages);
+			BUG_ON(!bio);
+		}
+		bio->bi_io_vec[i].bv_page = super->s_erase_page;
+		bio->bi_io_vec[i].bv_len = PAGE_SIZE;
+		bio->bi_io_vec[i].bv_offset = 0;
+	}
+	bio->bi_vcnt = nr_pages;
+	bio->bi_idx = 0;
+	bio->bi_size = nr_pages * PAGE_SIZE;
+	bio->bi_bdev = super->s_bdev;
+	bio->bi_sector = ofs >> 9;
+	bio->bi_private = sb;
+	bio->bi_end_io = erase_end_io;
+	atomic_inc(&super->s_pending_writes);
+	submit_bio(WRITE, bio);
+	return 0;
+}
+
+static int bdev_erase(struct super_block *sb, loff_t to, size_t len,
+		int ensure_write)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	BUG_ON(to & (PAGE_SIZE - 1));
+	BUG_ON(len & (PAGE_SIZE - 1));
+
+	if (super->s_flags & LOGFS_SB_FLAG_RO)
+		return -EROFS;
+
+	if (ensure_write) {
+		/*
+		 * Object store doesn't care whether erases happen or not.
+		 * But for the journal they are required.  Otherwise a scan
+		 * can find an old commit entry and assume it is the current
+		 * one, travelling back in time.
+		 */
+		do_erase(sb, to, to >> PAGE_SHIFT, len >> PAGE_SHIFT);
+	}
+
+	return 0;
+}
+
+static void bdev_sync(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	wait_event(wq, atomic_read(&super->s_pending_writes) == 0);
+}
+
+static struct page *bdev_find_first_sb(struct super_block *sb, u64 *ofs)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	filler_t *filler = bdev_readpage;
+
+	*ofs = 0;
+	return read_cache_page(mapping, 0, filler, sb);
+}
+
+static struct page *bdev_find_last_sb(struct super_block *sb, u64 *ofs)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	filler_t *filler = bdev_readpage;
+	u64 pos = (super->s_bdev->bd_inode->i_size & ~0xfffULL) - 0x1000;
+	pgoff_t index = pos >> PAGE_SHIFT;
+
+	*ofs = pos;
+	return read_cache_page(mapping, index, filler, sb);
+}
+
+static int bdev_write_sb(struct super_block *sb, struct page *page)
+{
+	struct block_device *bdev = logfs_super(sb)->s_bdev;
+
+	/* Nothing special to do for block devices. */
+	return sync_request(page, bdev, WRITE);
+}
+
+static void bdev_put_device(struct super_block *sb)
+{
+	close_bdev_exclusive(logfs_super(sb)->s_bdev, FMODE_READ|FMODE_WRITE);
+}
+
+static const struct logfs_device_ops bd_devops = {
+	.find_first_sb	= bdev_find_first_sb,
+	.find_last_sb	= bdev_find_last_sb,
+	.write_sb	= bdev_write_sb,
+	.readpage	= bdev_readpage,
+	.writeseg	= bdev_writeseg,
+	.erase		= bdev_erase,
+	.sync		= bdev_sync,
+	.put_device	= bdev_put_device,
+};
+
+int logfs_get_sb_bdev(struct file_system_type *type, int flags,
+		const char *devname, struct vfsmount *mnt)
+{
+	struct block_device *bdev;
+
+	bdev = open_bdev_exclusive(devname, FMODE_READ|FMODE_WRITE, type);
+	if (IS_ERR(bdev))
+		return PTR_ERR(bdev);
+
+	if (MAJOR(bdev->bd_dev) == MTD_BLOCK_MAJOR) {
+		int mtdnr = MINOR(bdev->bd_dev);
+		close_bdev_exclusive(bdev, FMODE_READ|FMODE_WRITE);
+		return logfs_get_sb_mtd(type, flags, mtdnr, mnt);
+	}
+
+	return logfs_get_sb_device(type, flags, NULL, bdev, &bd_devops, mnt);
+}
diff --git a/fs/logfs/dev_mtd.c b/fs/logfs/dev_mtd.c
new file mode 100644
index 000000000000..cafb6ef2e05b
--- /dev/null
+++ b/fs/logfs/dev_mtd.c
@@ -0,0 +1,254 @@
+/*
+ * fs/logfs/dev_mtd.c	- Device access methods for MTD
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+#include <linux/completion.h>
+#include <linux/mount.h>
+#include <linux/sched.h>
+
+#define PAGE_OFS(ofs) ((ofs) & (PAGE_SIZE-1))
+
+static int mtd_read(struct super_block *sb, loff_t ofs, size_t len, void *buf)
+{
+	struct mtd_info *mtd = logfs_super(sb)->s_mtd;
+	size_t retlen;
+	int ret;
+
+	ret = mtd->read(mtd, ofs, len, &retlen, buf);
+	BUG_ON(ret == -EINVAL);
+	if (ret)
+		return ret;
+
+	/* Not sure if we should loop instead. */
+	if (retlen != len)
+		return -EIO;
+
+	return 0;
+}
+
+static int mtd_write(struct super_block *sb, loff_t ofs, size_t len, void *buf)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct mtd_info *mtd = super->s_mtd;
+	size_t retlen;
+	loff_t page_start, page_end;
+	int ret;
+
+	if (super->s_flags & LOGFS_SB_FLAG_RO)
+		return -EROFS;
+
+	BUG_ON((ofs >= mtd->size) || (len > mtd->size - ofs));
+	BUG_ON(ofs != (ofs >> super->s_writeshift) << super->s_writeshift);
+	BUG_ON(len > PAGE_CACHE_SIZE);
+	page_start = ofs & PAGE_CACHE_MASK;
+	page_end = PAGE_CACHE_ALIGN(ofs + len) - 1;
+	ret = mtd->write(mtd, ofs, len, &retlen, buf);
+	if (ret || (retlen != len))
+		return -EIO;
+
+	return 0;
+}
+
+/*
+ * For as long as I can remember (since about 2001) mtd->erase has been an
+ * asynchronous interface lacking the first driver to actually use the
+ * asynchronous properties.  So just to prevent the first implementor of such
+ * a thing from breaking logfs in 2350, we do the usual pointless dance to
+ * declare a completion variable and wait for completion before returning
+ * from mtd_erase().  What an excercise in futility!
+ */
+static void logfs_erase_callback(struct erase_info *ei)
+{
+	complete((struct completion *)ei->priv);
+}
+
+static int mtd_erase_mapping(struct super_block *sb, loff_t ofs, size_t len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	struct page *page;
+	pgoff_t index = ofs >> PAGE_SHIFT;
+
+	for (index = ofs >> PAGE_SHIFT; index < (ofs + len) >> PAGE_SHIFT; index++) {
+		page = find_get_page(mapping, index);
+		if (!page)
+			continue;
+		memset(page_address(page), 0xFF, PAGE_SIZE);
+		page_cache_release(page);
+	}
+	return 0;
+}
+
+static int mtd_erase(struct super_block *sb, loff_t ofs, size_t len,
+		int ensure_write)
+{
+	struct mtd_info *mtd = logfs_super(sb)->s_mtd;
+	struct erase_info ei;
+	DECLARE_COMPLETION_ONSTACK(complete);
+	int ret;
+
+	BUG_ON(len % mtd->erasesize);
+	if (logfs_super(sb)->s_flags & LOGFS_SB_FLAG_RO)
+		return -EROFS;
+
+	memset(&ei, 0, sizeof(ei));
+	ei.mtd = mtd;
+	ei.addr = ofs;
+	ei.len = len;
+	ei.callback = logfs_erase_callback;
+	ei.priv = (long)&complete;
+	ret = mtd->erase(mtd, &ei);
+	if (ret)
+		return -EIO;
+
+	wait_for_completion(&complete);
+	if (ei.state != MTD_ERASE_DONE)
+		return -EIO;
+	return mtd_erase_mapping(sb, ofs, len);
+}
+
+static void mtd_sync(struct super_block *sb)
+{
+	struct mtd_info *mtd = logfs_super(sb)->s_mtd;
+
+	if (mtd->sync)
+		mtd->sync(mtd);
+}
+
+static int mtd_readpage(void *_sb, struct page *page)
+{
+	struct super_block *sb = _sb;
+	int err;
+
+	err = mtd_read(sb, page->index << PAGE_SHIFT, PAGE_SIZE,
+			page_address(page));
+	if (err == -EUCLEAN) {
+		err = 0;
+		/* FIXME: force GC this segment */
+	}
+	if (err) {
+		ClearPageUptodate(page);
+		SetPageError(page);
+	} else {
+		SetPageUptodate(page);
+		ClearPageError(page);
+	}
+	unlock_page(page);
+	return err;
+}
+
+static struct page *mtd_find_first_sb(struct super_block *sb, u64 *ofs)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	filler_t *filler = mtd_readpage;
+	struct mtd_info *mtd = super->s_mtd;
+
+	if (!mtd->block_isbad)
+		return NULL;
+
+	*ofs = 0;
+	while (mtd->block_isbad(mtd, *ofs)) {
+		*ofs += mtd->erasesize;
+		if (*ofs >= mtd->size)
+			return NULL;
+	}
+	BUG_ON(*ofs & ~PAGE_MASK);
+	return read_cache_page(mapping, *ofs >> PAGE_SHIFT, filler, sb);
+}
+
+static struct page *mtd_find_last_sb(struct super_block *sb, u64 *ofs)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	filler_t *filler = mtd_readpage;
+	struct mtd_info *mtd = super->s_mtd;
+
+	if (!mtd->block_isbad)
+		return NULL;
+
+	*ofs = mtd->size - mtd->erasesize;
+	while (mtd->block_isbad(mtd, *ofs)) {
+		*ofs -= mtd->erasesize;
+		if (*ofs <= 0)
+			return NULL;
+	}
+	*ofs = *ofs + mtd->erasesize - 0x1000;
+	BUG_ON(*ofs & ~PAGE_MASK);
+	return read_cache_page(mapping, *ofs >> PAGE_SHIFT, filler, sb);
+}
+
+static int __mtd_writeseg(struct super_block *sb, u64 ofs, pgoff_t index,
+		size_t nr_pages)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	struct page *page;
+	int i, err;
+
+	for (i = 0; i < nr_pages; i++) {
+		page = find_lock_page(mapping, index + i);
+		BUG_ON(!page);
+
+		err = mtd_write(sb, page->index << PAGE_SHIFT, PAGE_SIZE,
+				page_address(page));
+		unlock_page(page);
+		page_cache_release(page);
+		if (err)
+			return err;
+	}
+	return 0;
+}
+
+static void mtd_writeseg(struct super_block *sb, u64 ofs, size_t len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int head;
+
+	if (super->s_flags & LOGFS_SB_FLAG_RO)
+		return;
+
+	if (len == 0) {
+		/* This can happen when the object fit perfectly into a
+		 * segment, the segment gets written per sync and subsequently
+		 * closed.
+		 */
+		return;
+	}
+	head = ofs & (PAGE_SIZE - 1);
+	if (head) {
+		ofs -= head;
+		len += head;
+	}
+	len = PAGE_ALIGN(len);
+	__mtd_writeseg(sb, ofs, ofs >> PAGE_SHIFT, len >> PAGE_SHIFT);
+}
+
+static void mtd_put_device(struct super_block *sb)
+{
+	put_mtd_device(logfs_super(sb)->s_mtd);
+}
+
+static const struct logfs_device_ops mtd_devops = {
+	.find_first_sb	= mtd_find_first_sb,
+	.find_last_sb	= mtd_find_last_sb,
+	.readpage	= mtd_readpage,
+	.writeseg	= mtd_writeseg,
+	.erase		= mtd_erase,
+	.sync		= mtd_sync,
+	.put_device	= mtd_put_device,
+};
+
+int logfs_get_sb_mtd(struct file_system_type *type, int flags,
+		int mtdnr, struct vfsmount *mnt)
+{
+	struct mtd_info *mtd;
+	const struct logfs_device_ops *devops = &mtd_devops;
+
+	mtd = get_mtd_device(NULL, mtdnr);
+	return logfs_get_sb_device(type, flags, mtd, NULL, devops, mnt);
+}
diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c
new file mode 100644
index 000000000000..56a8bfbb0120
--- /dev/null
+++ b/fs/logfs/dir.c
@@ -0,0 +1,827 @@
+/*
+ * fs/logfs/dir.c	- directory-related code
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+
+
+/*
+ * Atomic dir operations
+ *
+ * Directory operations are by default not atomic.  Dentries and Inodes are
+ * created/removed/altered in seperate operations.  Therefore we need to do
+ * a small amount of journaling.
+ *
+ * Create, link, mkdir, mknod and symlink all share the same function to do
+ * the work: __logfs_create.  This function works in two atomic steps:
+ * 1. allocate inode (remember in journal)
+ * 2. allocate dentry (clear journal)
+ *
+ * As we can only get interrupted between the two, when the inode we just
+ * created is simply stored in the anchor.  On next mount, if we were
+ * interrupted, we delete the inode.  From a users point of view the
+ * operation never happened.
+ *
+ * Unlink and rmdir also share the same function: unlink.  Again, this
+ * function works in two atomic steps
+ * 1. remove dentry (remember inode in journal)
+ * 2. unlink inode (clear journal)
+ *
+ * And again, on the next mount, if we were interrupted, we delete the inode.
+ * From a users point of view the operation succeeded.
+ *
+ * Rename is the real pain to deal with, harder than all the other methods
+ * combined.  Depending on the circumstances we can run into three cases.
+ * A "target rename" where the target dentry already existed, a "local
+ * rename" where both parent directories are identical or a "cross-directory
+ * rename" in the remaining case.
+ *
+ * Local rename is atomic, as the old dentry is simply rewritten with a new
+ * name.
+ *
+ * Cross-directory rename works in two steps, similar to __logfs_create and
+ * logfs_unlink:
+ * 1. Write new dentry (remember old dentry in journal)
+ * 2. Remove old dentry (clear journal)
+ *
+ * Here we remember a dentry instead of an inode.  On next mount, if we were
+ * interrupted, we delete the dentry.  From a users point of view, the
+ * operation succeeded.
+ *
+ * Target rename works in three atomic steps:
+ * 1. Attach old inode to new dentry (remember old dentry and new inode)
+ * 2. Remove old dentry (still remember the new inode)
+ * 3. Remove victim inode
+ *
+ * Here we remember both an inode an a dentry.  If we get interrupted
+ * between steps 1 and 2, we delete both the dentry and the inode.  If
+ * we get interrupted between steps 2 and 3, we delete just the inode.
+ * In either case, the remaining objects are deleted on next mount.  From
+ * a users point of view, the operation succeeded.
+ */
+
+static int write_dir(struct inode *dir, struct logfs_disk_dentry *dd,
+		loff_t pos)
+{
+	return logfs_inode_write(dir, dd, sizeof(*dd), pos, WF_LOCK, NULL);
+}
+
+static int write_inode(struct inode *inode)
+{
+	return __logfs_write_inode(inode, WF_LOCK);
+}
+
+static s64 dir_seek_data(struct inode *inode, s64 pos)
+{
+	s64 new_pos = logfs_seek_data(inode, pos);
+
+	return max(pos, new_pos - 1);
+}
+
+static int beyond_eof(struct inode *inode, loff_t bix)
+{
+	loff_t pos = bix << inode->i_sb->s_blocksize_bits;
+	return pos >= i_size_read(inode);
+}
+
+/*
+ * Prime value was chosen to be roughly 256 + 26.  r5 hash uses 11,
+ * so short names (len <= 9) don't even occupy the complete 32bit name
+ * space.  A prime >256 ensures short names quickly spread the 32bit
+ * name space.  Add about 26 for the estimated amount of information
+ * of each character and pick a prime nearby, preferrably a bit-sparse
+ * one.
+ */
+static u32 hash_32(const char *s, int len, u32 seed)
+{
+	u32 hash = seed;
+	int i;
+
+	for (i = 0; i < len; i++)
+		hash = hash * 293 + s[i];
+	return hash;
+}
+
+/*
+ * We have to satisfy several conflicting requirements here.  Small
+ * directories should stay fairly compact and not require too many
+ * indirect blocks.  The number of possible locations for a given hash
+ * should be small to make lookup() fast.  And we should try hard not
+ * to overflow the 32bit name space or nfs and 32bit host systems will
+ * be unhappy.
+ *
+ * So we use the following scheme.  First we reduce the hash to 0..15
+ * and try a direct block.  If that is occupied we reduce the hash to
+ * 16..255 and try an indirect block.  Same for 2x and 3x indirect
+ * blocks.  Lastly we reduce the hash to 0x800_0000 .. 0xffff_ffff,
+ * but use buckets containing eight entries instead of a single one.
+ *
+ * Using 16 entries should allow for a reasonable amount of hash
+ * collisions, so the 32bit name space can be packed fairly tight
+ * before overflowing.  Oh and currently we don't overflow but return
+ * and error.
+ *
+ * How likely are collisions?  Doing the appropriate math is beyond me
+ * and the Bronstein textbook.  But running a test program to brute
+ * force collisions for a couple of days showed that on average the
+ * first collision occurs after 598M entries, with 290M being the
+ * smallest result.  Obviously 21 entries could already cause a
+ * collision if all entries are carefully chosen.
+ */
+static pgoff_t hash_index(u32 hash, int round)
+{
+	u32 i0_blocks = I0_BLOCKS;
+	u32 i1_blocks = I1_BLOCKS;
+	u32 i2_blocks = I2_BLOCKS;
+	u32 i3_blocks = I3_BLOCKS;
+
+	switch (round) {
+	case 0:
+		return hash % i0_blocks;
+	case 1:
+		return i0_blocks + hash % (i1_blocks - i0_blocks);
+	case 2:
+		return i1_blocks + hash % (i2_blocks - i1_blocks);
+	case 3:
+		return i2_blocks + hash % (i3_blocks - i2_blocks);
+	case 4 ... 19:
+		return i3_blocks + 16 * (hash % (((1<<31) - i3_blocks) / 16))
+			+ round - 4;
+	}
+	BUG();
+}
+
+static struct page *logfs_get_dd_page(struct inode *dir, struct dentry *dentry)
+{
+	struct qstr *name = &dentry->d_name;
+	struct page *page;
+	struct logfs_disk_dentry *dd;
+	u32 hash = hash_32(name->name, name->len, 0);
+	pgoff_t index;
+	int round;
+
+	if (name->len > LOGFS_MAX_NAMELEN)
+		return ERR_PTR(-ENAMETOOLONG);
+
+	for (round = 0; round < 20; round++) {
+		index = hash_index(hash, round);
+
+		if (beyond_eof(dir, index))
+			return NULL;
+		if (!logfs_exist_block(dir, index))
+			continue;
+		page = read_cache_page(dir->i_mapping, index,
+				(filler_t *)logfs_readpage, NULL);
+		if (IS_ERR(page))
+			return page;
+		dd = kmap_atomic(page, KM_USER0);
+		BUG_ON(dd->namelen == 0);
+
+		if (name->len != be16_to_cpu(dd->namelen) ||
+				memcmp(name->name, dd->name, name->len)) {
+			kunmap_atomic(dd, KM_USER0);
+			page_cache_release(page);
+			continue;
+		}
+
+		kunmap_atomic(dd, KM_USER0);
+		return page;
+	}
+	return NULL;
+}
+
+static int logfs_remove_inode(struct inode *inode)
+{
+	int ret;
+
+	inode->i_nlink--;
+	ret = write_inode(inode);
+	LOGFS_BUG_ON(ret, inode->i_sb);
+	return ret;
+}
+
+static void abort_transaction(struct inode *inode, struct logfs_transaction *ta)
+{
+	if (logfs_inode(inode)->li_block)
+		logfs_inode(inode)->li_block->ta = NULL;
+	kfree(ta);
+}
+
+static int logfs_unlink(struct inode *dir, struct dentry *dentry)
+{
+	struct logfs_super *super = logfs_super(dir->i_sb);
+	struct inode *inode = dentry->d_inode;
+	struct logfs_transaction *ta;
+	struct page *page;
+	pgoff_t index;
+	int ret;
+
+	ta = kzalloc(sizeof(*ta), GFP_KERNEL);
+	if (!ta)
+		return -ENOMEM;
+
+	ta->state = UNLINK_1;
+	ta->ino = inode->i_ino;
+
+	inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME;
+
+	page = logfs_get_dd_page(dir, dentry);
+	if (!page) {
+		kfree(ta);
+		return -ENOENT;
+	}
+	if (IS_ERR(page)) {
+		kfree(ta);
+		return PTR_ERR(page);
+	}
+	index = page->index;
+	page_cache_release(page);
+
+	mutex_lock(&super->s_dirop_mutex);
+	logfs_add_transaction(dir, ta);
+
+	ret = logfs_delete(dir, index, NULL);
+	if (!ret)
+		ret = write_inode(dir);
+
+	if (ret) {
+		abort_transaction(dir, ta);
+		printk(KERN_ERR"LOGFS: unable to delete inode\n");
+		goto out;
+	}
+
+	ta->state = UNLINK_2;
+	logfs_add_transaction(inode, ta);
+	ret = logfs_remove_inode(inode);
+out:
+	mutex_unlock(&super->s_dirop_mutex);
+	return ret;
+}
+
+static inline int logfs_empty_dir(struct inode *dir)
+{
+	u64 data;
+
+	data = logfs_seek_data(dir, 0) << dir->i_sb->s_blocksize_bits;
+	return data >= i_size_read(dir);
+}
+
+static int logfs_rmdir(struct inode *dir, struct dentry *dentry)
+{
+	struct inode *inode = dentry->d_inode;
+
+	if (!logfs_empty_dir(inode))
+		return -ENOTEMPTY;
+
+	return logfs_unlink(dir, dentry);
+}
+
+/* FIXME: readdir currently has it's own dir_walk code.  I don't see a good
+ * way to combine the two copies */
+#define IMPLICIT_NODES 2
+static int __logfs_readdir(struct file *file, void *buf, filldir_t filldir)
+{
+	struct inode *dir = file->f_dentry->d_inode;
+	loff_t pos = file->f_pos - IMPLICIT_NODES;
+	struct page *page;
+	struct logfs_disk_dentry *dd;
+	int full;
+
+	BUG_ON(pos < 0);
+	for (;; pos++) {
+		if (beyond_eof(dir, pos))
+			break;
+		if (!logfs_exist_block(dir, pos)) {
+			/* deleted dentry */
+			pos = dir_seek_data(dir, pos);
+			continue;
+		}
+		page = read_cache_page(dir->i_mapping, pos,
+				(filler_t *)logfs_readpage, NULL);
+		if (IS_ERR(page))
+			return PTR_ERR(page);
+		dd = kmap_atomic(page, KM_USER0);
+		BUG_ON(dd->namelen == 0);
+
+		full = filldir(buf, (char *)dd->name, be16_to_cpu(dd->namelen),
+				pos, be64_to_cpu(dd->ino), dd->type);
+		kunmap_atomic(dd, KM_USER0);
+		page_cache_release(page);
+		if (full)
+			break;
+	}
+
+	file->f_pos = pos + IMPLICIT_NODES;
+	return 0;
+}
+
+static int logfs_readdir(struct file *file, void *buf, filldir_t filldir)
+{
+	struct inode *inode = file->f_dentry->d_inode;
+	ino_t pino = parent_ino(file->f_dentry);
+	int err;
+
+	if (file->f_pos < 0)
+		return -EINVAL;
+
+	if (file->f_pos == 0) {
+		if (filldir(buf, ".", 1, 1, inode->i_ino, DT_DIR) < 0)
+			return 0;
+		file->f_pos++;
+	}
+	if (file->f_pos == 1) {
+		if (filldir(buf, "..", 2, 2, pino, DT_DIR) < 0)
+			return 0;
+		file->f_pos++;
+	}
+
+	err = __logfs_readdir(file, buf, filldir);
+	return err;
+}
+
+static void logfs_set_name(struct logfs_disk_dentry *dd, struct qstr *name)
+{
+	dd->namelen = cpu_to_be16(name->len);
+	memcpy(dd->name, name->name, name->len);
+}
+
+static struct dentry *logfs_lookup(struct inode *dir, struct dentry *dentry,
+		struct nameidata *nd)
+{
+	struct page *page;
+	struct logfs_disk_dentry *dd;
+	pgoff_t index;
+	u64 ino = 0;
+	struct inode *inode;
+
+	page = logfs_get_dd_page(dir, dentry);
+	if (IS_ERR(page))
+		return ERR_CAST(page);
+	if (!page) {
+		d_add(dentry, NULL);
+		return NULL;
+	}
+	index = page->index;
+	dd = kmap_atomic(page, KM_USER0);
+	ino = be64_to_cpu(dd->ino);
+	kunmap_atomic(dd, KM_USER0);
+	page_cache_release(page);
+
+	inode = logfs_iget(dir->i_sb, ino);
+	if (IS_ERR(inode)) {
+		printk(KERN_ERR"LogFS: Cannot read inode #%llx for dentry (%lx, %lx)n",
+				ino, dir->i_ino, index);
+		return ERR_CAST(inode);
+	}
+	return d_splice_alias(inode, dentry);
+}
+
+static void grow_dir(struct inode *dir, loff_t index)
+{
+	index = (index + 1) << dir->i_sb->s_blocksize_bits;
+	if (i_size_read(dir) < index)
+		i_size_write(dir, index);
+}
+
+static int logfs_write_dir(struct inode *dir, struct dentry *dentry,
+		struct inode *inode)
+{
+	struct page *page;
+	struct logfs_disk_dentry *dd;
+	u32 hash = hash_32(dentry->d_name.name, dentry->d_name.len, 0);
+	pgoff_t index;
+	int round, err;
+
+	for (round = 0; round < 20; round++) {
+		index = hash_index(hash, round);
+
+		if (logfs_exist_block(dir, index))
+			continue;
+		page = find_or_create_page(dir->i_mapping, index, GFP_KERNEL);
+		if (!page)
+			return -ENOMEM;
+
+		dd = kmap_atomic(page, KM_USER0);
+		memset(dd, 0, sizeof(*dd));
+		dd->ino = cpu_to_be64(inode->i_ino);
+		dd->type = logfs_type(inode);
+		logfs_set_name(dd, &dentry->d_name);
+		kunmap_atomic(dd, KM_USER0);
+
+		err = logfs_write_buf(dir, page, WF_LOCK);
+		unlock_page(page);
+		page_cache_release(page);
+		if (!err)
+			grow_dir(dir, index);
+		return err;
+	}
+	/* FIXME: Is there a better return value?  In most cases neither
+	 * the filesystem nor the directory are full.  But we have had
+	 * too many collisions for this particular hash and no fallback.
+	 */
+	return -ENOSPC;
+}
+
+static int __logfs_create(struct inode *dir, struct dentry *dentry,
+		struct inode *inode, const char *dest, long destlen)
+{
+	struct logfs_super *super = logfs_super(dir->i_sb);
+	struct logfs_inode *li = logfs_inode(inode);
+	struct logfs_transaction *ta;
+	int ret;
+
+	ta = kzalloc(sizeof(*ta), GFP_KERNEL);
+	if (!ta)
+		return -ENOMEM;
+
+	ta->state = CREATE_1;
+	ta->ino = inode->i_ino;
+	mutex_lock(&super->s_dirop_mutex);
+	logfs_add_transaction(inode, ta);
+
+	if (dest) {
+		/* symlink */
+		ret = logfs_inode_write(inode, dest, destlen, 0, WF_LOCK, NULL);
+		if (!ret)
+			ret = write_inode(inode);
+	} else {
+		/* creat/mkdir/mknod */
+		ret = write_inode(inode);
+	}
+	if (ret) {
+		abort_transaction(inode, ta);
+		li->li_flags |= LOGFS_IF_STILLBORN;
+		/* FIXME: truncate symlink */
+		inode->i_nlink--;
+		iput(inode);
+		goto out;
+	}
+
+	ta->state = CREATE_2;
+	logfs_add_transaction(dir, ta);
+	ret = logfs_write_dir(dir, dentry, inode);
+	/* sync directory */
+	if (!ret)
+		ret = write_inode(dir);
+
+	if (ret) {
+		logfs_del_transaction(dir, ta);
+		ta->state = CREATE_2;
+		logfs_add_transaction(inode, ta);
+		logfs_remove_inode(inode);
+		iput(inode);
+		goto out;
+	}
+	d_instantiate(dentry, inode);
+out:
+	mutex_unlock(&super->s_dirop_mutex);
+	return ret;
+}
+
+static int logfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
+{
+	struct inode *inode;
+
+	/*
+	 * FIXME: why do we have to fill in S_IFDIR, while the mode is
+	 * correct for mknod, creat, etc.?  Smells like the vfs *should*
+	 * do it for us but for some reason fails to do so.
+	 */
+	inode = logfs_new_inode(dir, S_IFDIR | mode);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+
+	inode->i_op = &logfs_dir_iops;
+	inode->i_fop = &logfs_dir_fops;
+
+	return __logfs_create(dir, dentry, inode, NULL, 0);
+}
+
+static int logfs_create(struct inode *dir, struct dentry *dentry, int mode,
+		struct nameidata *nd)
+{
+	struct inode *inode;
+
+	inode = logfs_new_inode(dir, mode);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+
+	inode->i_op = &logfs_reg_iops;
+	inode->i_fop = &logfs_reg_fops;
+	inode->i_mapping->a_ops = &logfs_reg_aops;
+
+	return __logfs_create(dir, dentry, inode, NULL, 0);
+}
+
+static int logfs_mknod(struct inode *dir, struct dentry *dentry, int mode,
+		dev_t rdev)
+{
+	struct inode *inode;
+
+	if (dentry->d_name.len > LOGFS_MAX_NAMELEN)
+		return -ENAMETOOLONG;
+
+	inode = logfs_new_inode(dir, mode);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+
+	init_special_inode(inode, mode, rdev);
+
+	return __logfs_create(dir, dentry, inode, NULL, 0);
+}
+
+static int logfs_symlink(struct inode *dir, struct dentry *dentry,
+		const char *target)
+{
+	struct inode *inode;
+	size_t destlen = strlen(target) + 1;
+
+	if (destlen > dir->i_sb->s_blocksize)
+		return -ENAMETOOLONG;
+
+	inode = logfs_new_inode(dir, S_IFLNK | 0777);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+
+	inode->i_op = &logfs_symlink_iops;
+	inode->i_mapping->a_ops = &logfs_reg_aops;
+
+	return __logfs_create(dir, dentry, inode, target, destlen);
+}
+
+static int logfs_permission(struct inode *inode, int mask)
+{
+	return generic_permission(inode, mask, NULL);
+}
+
+static int logfs_link(struct dentry *old_dentry, struct inode *dir,
+		struct dentry *dentry)
+{
+	struct inode *inode = old_dentry->d_inode;
+
+	if (inode->i_nlink >= LOGFS_LINK_MAX)
+		return -EMLINK;
+
+	inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME;
+	atomic_inc(&inode->i_count);
+	inode->i_nlink++;
+	mark_inode_dirty_sync(inode);
+
+	return __logfs_create(dir, dentry, inode, NULL, 0);
+}
+
+static int logfs_get_dd(struct inode *dir, struct dentry *dentry,
+		struct logfs_disk_dentry *dd, loff_t *pos)
+{
+	struct page *page;
+	void *map;
+
+	page = logfs_get_dd_page(dir, dentry);
+	if (IS_ERR(page))
+		return PTR_ERR(page);
+	*pos = page->index;
+	map = kmap_atomic(page, KM_USER0);
+	memcpy(dd, map, sizeof(*dd));
+	kunmap_atomic(map, KM_USER0);
+	page_cache_release(page);
+	return 0;
+}
+
+static int logfs_delete_dd(struct inode *dir, loff_t pos)
+{
+	/*
+	 * Getting called with pos somewhere beyond eof is either a goofup
+	 * within this file or means someone maliciously edited the
+	 * (crc-protected) journal.
+	 */
+	BUG_ON(beyond_eof(dir, pos));
+	dir->i_ctime = dir->i_mtime = CURRENT_TIME;
+	log_dir(" Delete dentry (%lx, %llx)\n", dir->i_ino, pos);
+	return logfs_delete(dir, pos, NULL);
+}
+
+/*
+ * Cross-directory rename, target does not exist.  Just a little nasty.
+ * Create a new dentry in the target dir, then remove the old dentry,
+ * all the while taking care to remember our operation in the journal.
+ */
+static int logfs_rename_cross(struct inode *old_dir, struct dentry *old_dentry,
+			      struct inode *new_dir, struct dentry *new_dentry)
+{
+	struct logfs_super *super = logfs_super(old_dir->i_sb);
+	struct logfs_disk_dentry dd;
+	struct logfs_transaction *ta;
+	loff_t pos;
+	int err;
+
+	/* 1. locate source dd */
+	err = logfs_get_dd(old_dir, old_dentry, &dd, &pos);
+	if (err)
+		return err;
+
+	ta = kzalloc(sizeof(*ta), GFP_KERNEL);
+	if (!ta)
+		return -ENOMEM;
+
+	ta->state = CROSS_RENAME_1;
+	ta->dir = old_dir->i_ino;
+	ta->pos = pos;
+
+	/* 2. write target dd */
+	mutex_lock(&super->s_dirop_mutex);
+	logfs_add_transaction(new_dir, ta);
+	err = logfs_write_dir(new_dir, new_dentry, old_dentry->d_inode);
+	if (!err)
+		err = write_inode(new_dir);
+
+	if (err) {
+		super->s_rename_dir = 0;
+		super->s_rename_pos = 0;
+		abort_transaction(new_dir, ta);
+		goto out;
+	}
+
+	/* 3. remove source dd */
+	ta->state = CROSS_RENAME_2;
+	logfs_add_transaction(old_dir, ta);
+	err = logfs_delete_dd(old_dir, pos);
+	if (!err)
+		err = write_inode(old_dir);
+	LOGFS_BUG_ON(err, old_dir->i_sb);
+out:
+	mutex_unlock(&super->s_dirop_mutex);
+	return err;
+}
+
+static int logfs_replace_inode(struct inode *dir, struct dentry *dentry,
+		struct logfs_disk_dentry *dd, struct inode *inode)
+{
+	loff_t pos;
+	int err;
+
+	err = logfs_get_dd(dir, dentry, dd, &pos);
+	if (err)
+		return err;
+	dd->ino = cpu_to_be64(inode->i_ino);
+	dd->type = logfs_type(inode);
+
+	err = write_dir(dir, dd, pos);
+	if (err)
+		return err;
+	log_dir("Replace dentry (%lx, %llx) %s -> %llx\n", dir->i_ino, pos,
+			dd->name, be64_to_cpu(dd->ino));
+	return write_inode(dir);
+}
+
+/* Target dentry exists - the worst case.  We need to attach the source
+ * inode to the target dentry, then remove the orphaned target inode and
+ * source dentry.
+ */
+static int logfs_rename_target(struct inode *old_dir, struct dentry *old_dentry,
+			       struct inode *new_dir, struct dentry *new_dentry)
+{
+	struct logfs_super *super = logfs_super(old_dir->i_sb);
+	struct inode *old_inode = old_dentry->d_inode;
+	struct inode *new_inode = new_dentry->d_inode;
+	int isdir = S_ISDIR(old_inode->i_mode);
+	struct logfs_disk_dentry dd;
+	struct logfs_transaction *ta;
+	loff_t pos;
+	int err;
+
+	BUG_ON(isdir != S_ISDIR(new_inode->i_mode));
+	if (isdir) {
+		if (!logfs_empty_dir(new_inode))
+			return -ENOTEMPTY;
+	}
+
+	/* 1. locate source dd */
+	err = logfs_get_dd(old_dir, old_dentry, &dd, &pos);
+	if (err)
+		return err;
+
+	ta = kzalloc(sizeof(*ta), GFP_KERNEL);
+	if (!ta)
+		return -ENOMEM;
+
+	ta->state = TARGET_RENAME_1;
+	ta->dir = old_dir->i_ino;
+	ta->pos = pos;
+	ta->ino = new_inode->i_ino;
+
+	/* 2. attach source inode to target dd */
+	mutex_lock(&super->s_dirop_mutex);
+	logfs_add_transaction(new_dir, ta);
+	err = logfs_replace_inode(new_dir, new_dentry, &dd, old_inode);
+	if (err) {
+		super->s_rename_dir = 0;
+		super->s_rename_pos = 0;
+		super->s_victim_ino = 0;
+		abort_transaction(new_dir, ta);
+		goto out;
+	}
+
+	/* 3. remove source dd */
+	ta->state = TARGET_RENAME_2;
+	logfs_add_transaction(old_dir, ta);
+	err = logfs_delete_dd(old_dir, pos);
+	if (!err)
+		err = write_inode(old_dir);
+	LOGFS_BUG_ON(err, old_dir->i_sb);
+
+	/* 4. remove target inode */
+	ta->state = TARGET_RENAME_3;
+	logfs_add_transaction(new_inode, ta);
+	err = logfs_remove_inode(new_inode);
+
+out:
+	mutex_unlock(&super->s_dirop_mutex);
+	return err;
+}
+
+static int logfs_rename(struct inode *old_dir, struct dentry *old_dentry,
+			struct inode *new_dir, struct dentry *new_dentry)
+{
+	if (new_dentry->d_inode)
+		return logfs_rename_target(old_dir, old_dentry,
+					   new_dir, new_dentry);
+	return logfs_rename_cross(old_dir, old_dentry, new_dir, new_dentry);
+}
+
+/* No locking done here, as this is called before .get_sb() returns. */
+int logfs_replay_journal(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *inode;
+	u64 ino, pos;
+	int err;
+
+	if (super->s_victim_ino) {
+		/* delete victim inode */
+		ino = super->s_victim_ino;
+		printk(KERN_INFO"LogFS: delete unmapped inode #%llx\n", ino);
+		inode = logfs_iget(sb, ino);
+		if (IS_ERR(inode))
+			goto fail;
+
+		LOGFS_BUG_ON(i_size_read(inode) > 0, sb);
+		super->s_victim_ino = 0;
+		err = logfs_remove_inode(inode);
+		iput(inode);
+		if (err) {
+			super->s_victim_ino = ino;
+			goto fail;
+		}
+	}
+	if (super->s_rename_dir) {
+		/* delete old dd from rename */
+		ino = super->s_rename_dir;
+		pos = super->s_rename_pos;
+		printk(KERN_INFO"LogFS: delete unbacked dentry (%llx, %llx)\n",
+				ino, pos);
+		inode = logfs_iget(sb, ino);
+		if (IS_ERR(inode))
+			goto fail;
+
+		super->s_rename_dir = 0;
+		super->s_rename_pos = 0;
+		err = logfs_delete_dd(inode, pos);
+		iput(inode);
+		if (err) {
+			super->s_rename_dir = ino;
+			super->s_rename_pos = pos;
+			goto fail;
+		}
+	}
+	return 0;
+fail:
+	LOGFS_BUG(sb);
+	return -EIO;
+}
+
+const struct inode_operations logfs_symlink_iops = {
+	.readlink	= generic_readlink,
+	.follow_link	= page_follow_link_light,
+};
+
+const struct inode_operations logfs_dir_iops = {
+	.create		= logfs_create,
+	.link		= logfs_link,
+	.lookup		= logfs_lookup,
+	.mkdir		= logfs_mkdir,
+	.mknod		= logfs_mknod,
+	.rename		= logfs_rename,
+	.rmdir		= logfs_rmdir,
+	.permission	= logfs_permission,
+	.symlink	= logfs_symlink,
+	.unlink		= logfs_unlink,
+};
+const struct file_operations logfs_dir_fops = {
+	.fsync		= logfs_fsync,
+	.ioctl		= logfs_ioctl,
+	.readdir	= logfs_readdir,
+	.read		= generic_read_dir,
+};
diff --git a/fs/logfs/file.c b/fs/logfs/file.c
new file mode 100644
index 000000000000..370f367a933e
--- /dev/null
+++ b/fs/logfs/file.c
@@ -0,0 +1,263 @@
+/*
+ * fs/logfs/file.c	- prepare_write, commit_write and friends
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+#include <linux/sched.h>
+#include <linux/writeback.h>
+
+static int logfs_write_begin(struct file *file, struct address_space *mapping,
+		loff_t pos, unsigned len, unsigned flags,
+		struct page **pagep, void **fsdata)
+{
+	struct inode *inode = mapping->host;
+	struct page *page;
+	pgoff_t index = pos >> PAGE_CACHE_SHIFT;
+
+	page = grab_cache_page_write_begin(mapping, index, flags);
+	if (!page)
+		return -ENOMEM;
+	*pagep = page;
+
+	if ((len == PAGE_CACHE_SIZE) || PageUptodate(page))
+		return 0;
+	if ((pos & PAGE_CACHE_MASK) >= i_size_read(inode)) {
+		unsigned start = pos & (PAGE_CACHE_SIZE - 1);
+		unsigned end = start + len;
+
+		/* Reading beyond i_size is simple: memset to zero */
+		zero_user_segments(page, 0, start, end, PAGE_CACHE_SIZE);
+		return 0;
+	}
+	return logfs_readpage_nolock(page);
+}
+
+static int logfs_write_end(struct file *file, struct address_space *mapping,
+		loff_t pos, unsigned len, unsigned copied, struct page *page,
+		void *fsdata)
+{
+	struct inode *inode = mapping->host;
+	pgoff_t index = page->index;
+	unsigned start = pos & (PAGE_CACHE_SIZE - 1);
+	unsigned end = start + copied;
+	int ret = 0;
+
+	BUG_ON(PAGE_CACHE_SIZE != inode->i_sb->s_blocksize);
+	BUG_ON(page->index > I3_BLOCKS);
+
+	if (copied < len) {
+		/*
+		 * Short write of a non-initialized paged.  Just tell userspace
+		 * to retry the entire page.
+		 */
+		if (!PageUptodate(page)) {
+			copied = 0;
+			goto out;
+		}
+	}
+	if (copied == 0)
+		goto out; /* FIXME: do we need to update inode? */
+
+	if (i_size_read(inode) < (index << PAGE_CACHE_SHIFT) + end) {
+		i_size_write(inode, (index << PAGE_CACHE_SHIFT) + end);
+		mark_inode_dirty_sync(inode);
+	}
+
+	SetPageUptodate(page);
+	if (!PageDirty(page)) {
+		if (!get_page_reserve(inode, page))
+			__set_page_dirty_nobuffers(page);
+		else
+			ret = logfs_write_buf(inode, page, WF_LOCK);
+	}
+out:
+	unlock_page(page);
+	page_cache_release(page);
+	return ret ? ret : copied;
+}
+
+int logfs_readpage(struct file *file, struct page *page)
+{
+	int ret;
+
+	ret = logfs_readpage_nolock(page);
+	unlock_page(page);
+	return ret;
+}
+
+/* Clear the page's dirty flag in the radix tree. */
+/* TODO: mucking with PageWriteback is silly.  Add a generic function to clear
+ * the dirty bit from the radix tree for filesystems that don't have to wait
+ * for page writeback to finish (i.e. any compressing filesystem).
+ */
+static void clear_radix_tree_dirty(struct page *page)
+{
+	BUG_ON(PagePrivate(page) || page->private);
+	set_page_writeback(page);
+	end_page_writeback(page);
+}
+
+static int __logfs_writepage(struct page *page)
+{
+	struct inode *inode = page->mapping->host;
+	int err;
+
+	err = logfs_write_buf(inode, page, WF_LOCK);
+	if (err)
+		set_page_dirty(page);
+	else
+		clear_radix_tree_dirty(page);
+	unlock_page(page);
+	return err;
+}
+
+static int logfs_writepage(struct page *page, struct writeback_control *wbc)
+{
+	struct inode *inode = page->mapping->host;
+	loff_t i_size = i_size_read(inode);
+	pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
+	unsigned offset;
+	u64 bix;
+	level_t level;
+
+	log_file("logfs_writepage(%lx, %lx, %p)\n", inode->i_ino, page->index,
+			page);
+
+	logfs_unpack_index(page->index, &bix, &level);
+
+	/* Indirect blocks are never truncated */
+	if (level != 0)
+		return __logfs_writepage(page);
+
+	/*
+	 * TODO: everything below is a near-verbatim copy of nobh_writepage().
+	 * The relevant bits should be factored out after logfs is merged.
+	 */
+
+	/* Is the page fully inside i_size? */
+	if (bix < end_index)
+		return __logfs_writepage(page);
+
+	 /* Is the page fully outside i_size? (truncate in progress) */
+	offset = i_size & (PAGE_CACHE_SIZE-1);
+	if (bix > end_index || offset == 0) {
+		unlock_page(page);
+		return 0; /* don't care */
+	}
+
+	/*
+	 * The page straddles i_size.  It must be zeroed out on each and every
+	 * writepage invokation because it may be mmapped.  "A file is mapped
+	 * in multiples of the page size.  For a file that is not a multiple of
+	 * the  page size, the remaining memory is zeroed when mapped, and
+	 * writes to that region are not written out to the file."
+	 */
+	zero_user_segment(page, offset, PAGE_CACHE_SIZE);
+	return __logfs_writepage(page);
+}
+
+static void logfs_invalidatepage(struct page *page, unsigned long offset)
+{
+	move_page_to_btree(page);
+	BUG_ON(PagePrivate(page) || page->private);
+}
+
+static int logfs_releasepage(struct page *page, gfp_t only_xfs_uses_this)
+{
+	return 0; /* None of these are easy to release */
+}
+
+
+int logfs_ioctl(struct inode *inode, struct file *file, unsigned int cmd,
+		unsigned long arg)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	unsigned int oldflags, flags;
+	int err;
+
+	switch (cmd) {
+	case FS_IOC_GETFLAGS:
+		flags = li->li_flags & LOGFS_FL_USER_VISIBLE;
+		return put_user(flags, (int __user *)arg);
+	case FS_IOC_SETFLAGS:
+		if (IS_RDONLY(inode))
+			return -EROFS;
+
+		if (!is_owner_or_cap(inode))
+			return -EACCES;
+
+		err = get_user(flags, (int __user *)arg);
+		if (err)
+			return err;
+
+		mutex_lock(&inode->i_mutex);
+		oldflags = li->li_flags;
+		flags &= LOGFS_FL_USER_MODIFIABLE;
+		flags |= oldflags & ~LOGFS_FL_USER_MODIFIABLE;
+		li->li_flags = flags;
+		mutex_unlock(&inode->i_mutex);
+
+		inode->i_ctime = CURRENT_TIME;
+		mark_inode_dirty_sync(inode);
+		return 0;
+
+	default:
+		return -ENOTTY;
+	}
+}
+
+int logfs_fsync(struct file *file, struct dentry *dentry, int datasync)
+{
+	struct super_block *sb = dentry->d_inode->i_sb;
+	struct logfs_super *super = logfs_super(sb);
+
+	/* FIXME: write anchor */
+	super->s_devops->sync(sb);
+	return 0;
+}
+
+static int logfs_setattr(struct dentry *dentry, struct iattr *attr)
+{
+	struct inode *inode = dentry->d_inode;
+	int err = 0;
+
+	if (attr->ia_valid & ATTR_SIZE)
+		err = logfs_truncate(inode, attr->ia_size);
+	attr->ia_valid &= ~ATTR_SIZE;
+
+	if (!err)
+		err = inode_change_ok(inode, attr);
+	if (!err)
+		err = inode_setattr(inode, attr);
+	return err;
+}
+
+const struct inode_operations logfs_reg_iops = {
+	.setattr	= logfs_setattr,
+};
+
+const struct file_operations logfs_reg_fops = {
+	.aio_read	= generic_file_aio_read,
+	.aio_write	= generic_file_aio_write,
+	.fsync		= logfs_fsync,
+	.ioctl		= logfs_ioctl,
+	.llseek		= generic_file_llseek,
+	.mmap		= generic_file_readonly_mmap,
+	.open		= generic_file_open,
+	.read		= do_sync_read,
+	.write		= do_sync_write,
+};
+
+const struct address_space_operations logfs_reg_aops = {
+	.invalidatepage	= logfs_invalidatepage,
+	.readpage	= logfs_readpage,
+	.releasepage	= logfs_releasepage,
+	.set_page_dirty	= __set_page_dirty_nobuffers,
+	.writepage	= logfs_writepage,
+	.writepages	= generic_writepages,
+	.write_begin	= logfs_write_begin,
+	.write_end	= logfs_write_end,
+};
diff --git a/fs/logfs/gc.c b/fs/logfs/gc.c
new file mode 100644
index 000000000000..92949f95a901
--- /dev/null
+++ b/fs/logfs/gc.c
@@ -0,0 +1,730 @@
+/*
+ * fs/logfs/gc.c	- garbage collection code
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+#include <linux/sched.h>
+
+/*
+ * Wear leveling needs to kick in when the difference between low erase
+ * counts and high erase counts gets too big.  A good value for "too big"
+ * may be somewhat below 10% of maximum erase count for the device.
+ * Why not 397, to pick a nice round number with no specific meaning? :)
+ *
+ * WL_RATELIMIT is the minimum time between two wear level events.  A huge
+ * number of segments may fulfil the requirements for wear leveling at the
+ * same time.  If that happens we don't want to cause a latency from hell,
+ * but just gently pick one segment every so often and minimize overhead.
+ */
+#define WL_DELTA 397
+#define WL_RATELIMIT 100
+#define MAX_OBJ_ALIASES	2600
+#define SCAN_RATIO 512	/* number of scanned segments per gc'd segment */
+#define LIST_SIZE 64	/* base size of candidate lists */
+#define SCAN_ROUNDS 128	/* maximum number of complete medium scans */
+#define SCAN_ROUNDS_HIGH 4 /* maximum number of higher-level scans */
+
+static int no_free_segments(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	return super->s_free_list.count;
+}
+
+/* journal has distance -1, top-most ifile layer distance 0 */
+static u8 root_distance(struct super_block *sb, gc_level_t __gc_level)
+{
+	struct logfs_super *super = logfs_super(sb);
+	u8 gc_level = (__force u8)__gc_level;
+
+	switch (gc_level) {
+	case 0: /* fall through */
+	case 1: /* fall through */
+	case 2: /* fall through */
+	case 3:
+		/* file data or indirect blocks */
+		return super->s_ifile_levels + super->s_iblock_levels - gc_level;
+	case 6: /* fall through */
+	case 7: /* fall through */
+	case 8: /* fall through */
+	case 9:
+		/* inode file data or indirect blocks */
+		return super->s_ifile_levels - (gc_level - 6);
+	default:
+		printk(KERN_ERR"LOGFS: segment of unknown level %x found\n",
+				gc_level);
+		WARN_ON(1);
+		return super->s_ifile_levels + super->s_iblock_levels;
+	}
+}
+
+static int segment_is_reserved(struct super_block *sb, u32 segno)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_area *area;
+	void *reserved;
+	int i;
+
+	/* Some segments are reserved.  Just pretend they were all valid */
+	reserved = btree_lookup32(&super->s_reserved_segments, segno);
+	if (reserved)
+		return 1;
+
+	/* Currently open segments */
+	for_each_area(i) {
+		area = super->s_area[i];
+		if (area->a_is_open && area->a_segno == segno)
+			return 1;
+	}
+
+	return 0;
+}
+
+static void logfs_mark_segment_bad(struct super_block *sb, u32 segno)
+{
+	BUG();
+}
+
+/*
+ * Returns the bytes consumed by valid objects in this segment.  Object headers
+ * are counted, the segment header is not.
+ */
+static u32 logfs_valid_bytes(struct super_block *sb, u32 segno, u32 *ec,
+		gc_level_t *gc_level)
+{
+	struct logfs_segment_entry se;
+	u32 ec_level;
+
+	logfs_get_segment_entry(sb, segno, &se);
+	if (se.ec_level == cpu_to_be32(BADSEG) ||
+			se.valid == cpu_to_be32(RESERVED))
+		return RESERVED;
+
+	ec_level = be32_to_cpu(se.ec_level);
+	*ec = ec_level >> 4;
+	*gc_level = GC_LEVEL(ec_level & 0xf);
+	return be32_to_cpu(se.valid);
+}
+
+static void logfs_cleanse_block(struct super_block *sb, u64 ofs, u64 ino,
+		u64 bix, gc_level_t gc_level)
+{
+	struct inode *inode;
+	int err, cookie;
+
+	inode = logfs_safe_iget(sb, ino, &cookie);
+	err = logfs_rewrite_block(inode, bix, ofs, gc_level, 0);
+	BUG_ON(err);
+	logfs_safe_iput(inode, cookie);
+}
+
+static u32 logfs_gc_segment(struct super_block *sb, u32 segno, u8 dist)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_segment_header sh;
+	struct logfs_object_header oh;
+	u64 ofs, ino, bix;
+	u32 seg_ofs, logical_segno, cleaned = 0;
+	int err, len, valid;
+	gc_level_t gc_level;
+
+	LOGFS_BUG_ON(segment_is_reserved(sb, segno), sb);
+
+	btree_insert32(&super->s_reserved_segments, segno, (void *)1, GFP_NOFS);
+	err = wbuf_read(sb, dev_ofs(sb, segno, 0), sizeof(sh), &sh);
+	BUG_ON(err);
+	gc_level = GC_LEVEL(sh.level);
+	logical_segno = be32_to_cpu(sh.segno);
+	if (sh.crc != logfs_crc32(&sh, sizeof(sh), 4)) {
+		logfs_mark_segment_bad(sb, segno);
+		cleaned = -1;
+		goto out;
+	}
+
+	for (seg_ofs = LOGFS_SEGMENT_HEADERSIZE;
+			seg_ofs + sizeof(oh) < super->s_segsize; ) {
+		ofs = dev_ofs(sb, logical_segno, seg_ofs);
+		err = wbuf_read(sb, dev_ofs(sb, segno, seg_ofs), sizeof(oh),
+				&oh);
+		BUG_ON(err);
+
+		if (!memchr_inv(&oh, 0xff, sizeof(oh)))
+			break;
+
+		if (oh.crc != logfs_crc32(&oh, sizeof(oh) - 4, 4)) {
+			logfs_mark_segment_bad(sb, segno);
+			cleaned = super->s_segsize - 1;
+			goto out;
+		}
+
+		ino = be64_to_cpu(oh.ino);
+		bix = be64_to_cpu(oh.bix);
+		len = sizeof(oh) + be16_to_cpu(oh.len);
+		valid = logfs_is_valid_block(sb, ofs, ino, bix, gc_level);
+		if (valid == 1) {
+			logfs_cleanse_block(sb, ofs, ino, bix, gc_level);
+			cleaned += len;
+		} else if (valid == 2) {
+			/* Will be invalid upon journal commit */
+			cleaned += len;
+		}
+		seg_ofs += len;
+	}
+out:
+	btree_remove32(&super->s_reserved_segments, segno);
+	return cleaned;
+}
+
+static struct gc_candidate *add_list(struct gc_candidate *cand,
+		struct candidate_list *list)
+{
+	struct rb_node **p = &list->rb_tree.rb_node;
+	struct rb_node *parent = NULL;
+	struct gc_candidate *cur;
+	int comp;
+
+	cand->list = list;
+	while (*p) {
+		parent = *p;
+		cur = rb_entry(parent, struct gc_candidate, rb_node);
+
+		if (list->sort_by_ec)
+			comp = cand->erase_count < cur->erase_count;
+		else
+			comp = cand->valid < cur->valid;
+
+		if (comp)
+			p = &parent->rb_left;
+		else
+			p = &parent->rb_right;
+	}
+	rb_link_node(&cand->rb_node, parent, p);
+	rb_insert_color(&cand->rb_node, &list->rb_tree);
+
+	if (list->count <= list->maxcount) {
+		list->count++;
+		return NULL;
+	}
+	cand = rb_entry(rb_last(&list->rb_tree), struct gc_candidate, rb_node);
+	rb_erase(&cand->rb_node, &list->rb_tree);
+	cand->list = NULL;
+	return cand;
+}
+
+static void remove_from_list(struct gc_candidate *cand)
+{
+	struct candidate_list *list = cand->list;
+
+	rb_erase(&cand->rb_node, &list->rb_tree);
+	list->count--;
+}
+
+static void free_candidate(struct super_block *sb, struct gc_candidate *cand)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	btree_remove32(&super->s_cand_tree, cand->segno);
+	kfree(cand);
+}
+
+u32 get_best_cand(struct super_block *sb, struct candidate_list *list, u32 *ec)
+{
+	struct gc_candidate *cand;
+	u32 segno;
+
+	BUG_ON(list->count == 0);
+
+	cand = rb_entry(rb_first(&list->rb_tree), struct gc_candidate, rb_node);
+	remove_from_list(cand);
+	segno = cand->segno;
+	if (ec)
+		*ec = cand->erase_count;
+	free_candidate(sb, cand);
+	return segno;
+}
+
+/*
+ * We have several lists to manage segments with.  The reserve_list is used to
+ * deal with bad blocks.  We try to keep the best (lowest ec) segments on this
+ * list.
+ * The free_list contains free segments for normal usage.  It usually gets the
+ * second pick after the reserve_list.  But when the free_list is running short
+ * it is more important to keep the free_list full than to keep a reserve.
+ *
+ * Segments that are not free are put onto a per-level low_list.  If we have
+ * to run garbage collection, we pick a candidate from there.  All segments on
+ * those lists should have at least some free space so GC will make progress.
+ *
+ * And last we have the ec_list, which is used to pick segments for wear
+ * leveling.
+ *
+ * If all appropriate lists are full, we simply free the candidate and forget
+ * about that segment for a while.  We have better candidates for each purpose.
+ */
+static void __add_candidate(struct super_block *sb, struct gc_candidate *cand)
+{
+	struct logfs_super *super = logfs_super(sb);
+	u32 full = super->s_segsize - LOGFS_SEGMENT_RESERVE;
+
+	if (cand->valid == 0) {
+		/* 100% free segments */
+		log_gc_noisy("add reserve segment %x (ec %x) at %llx\n",
+				cand->segno, cand->erase_count,
+				dev_ofs(sb, cand->segno, 0));
+		cand = add_list(cand, &super->s_reserve_list);
+		if (cand) {
+			log_gc_noisy("add free segment %x (ec %x) at %llx\n",
+					cand->segno, cand->erase_count,
+					dev_ofs(sb, cand->segno, 0));
+			cand = add_list(cand, &super->s_free_list);
+		}
+	} else {
+		/* good candidates for Garbage Collection */
+		if (cand->valid < full)
+			cand = add_list(cand, &super->s_low_list[cand->dist]);
+		/* good candidates for wear leveling,
+		 * segments that were recently written get ignored */
+		if (cand)
+			cand = add_list(cand, &super->s_ec_list);
+	}
+	if (cand)
+		free_candidate(sb, cand);
+}
+
+static int add_candidate(struct super_block *sb, u32 segno, u32 valid, u32 ec,
+		u8 dist)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct gc_candidate *cand;
+
+	cand = kmalloc(sizeof(*cand), GFP_NOFS);
+	if (!cand)
+		return -ENOMEM;
+
+	cand->segno = segno;
+	cand->valid = valid;
+	cand->erase_count = ec;
+	cand->dist = dist;
+
+	btree_insert32(&super->s_cand_tree, segno, cand, GFP_NOFS);
+	__add_candidate(sb, cand);
+	return 0;
+}
+
+static void remove_segment_from_lists(struct super_block *sb, u32 segno)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct gc_candidate *cand;
+
+	cand = btree_lookup32(&super->s_cand_tree, segno);
+	if (cand) {
+		remove_from_list(cand);
+		free_candidate(sb, cand);
+	}
+}
+
+static void scan_segment(struct super_block *sb, u32 segno)
+{
+	u32 valid, ec = 0;
+	gc_level_t gc_level = 0;
+	u8 dist;
+
+	if (segment_is_reserved(sb, segno))
+		return;
+
+	remove_segment_from_lists(sb, segno);
+	valid = logfs_valid_bytes(sb, segno, &ec, &gc_level);
+	if (valid == RESERVED)
+		return;
+
+	dist = root_distance(sb, gc_level);
+	add_candidate(sb, segno, valid, ec, dist);
+}
+
+static struct gc_candidate *first_in_list(struct candidate_list *list)
+{
+	if (list->count == 0)
+		return NULL;
+	return rb_entry(rb_first(&list->rb_tree), struct gc_candidate, rb_node);
+}
+
+/*
+ * Find the best segment for garbage collection.  Main criterion is
+ * the segment requiring the least effort to clean.  Secondary
+ * criterion is to GC on the lowest level available.
+ *
+ * So we search the least effort segment on the lowest level first,
+ * then move up and pick another segment iff is requires significantly
+ * less effort.  Hence the LOGFS_MAX_OBJECTSIZE in the comparison.
+ */
+static struct gc_candidate *get_candidate(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int i, max_dist;
+	struct gc_candidate *cand = NULL, *this;
+
+	max_dist = min(no_free_segments(sb), LOGFS_NO_AREAS);
+
+	for (i = max_dist; i >= 0; i--) {
+		this = first_in_list(&super->s_low_list[i]);
+		if (!this)
+			continue;
+		if (!cand)
+			cand = this;
+		if (this->valid + LOGFS_MAX_OBJECTSIZE <= cand->valid)
+			cand = this;
+	}
+	return cand;
+}
+
+static int __logfs_gc_once(struct super_block *sb, struct gc_candidate *cand)
+{
+	struct logfs_super *super = logfs_super(sb);
+	gc_level_t gc_level;
+	u32 cleaned, valid, segno, ec;
+	u8 dist;
+
+	if (!cand) {
+		log_gc("GC attempted, but no candidate found\n");
+		return 0;
+	}
+
+	segno = cand->segno;
+	dist = cand->dist;
+	valid = logfs_valid_bytes(sb, segno, &ec, &gc_level);
+	free_candidate(sb, cand);
+	log_gc("GC segment #%02x at %llx, %x required, %x free, %x valid, %llx free\n",
+			segno, (u64)segno << super->s_segshift,
+			dist, no_free_segments(sb), valid,
+			super->s_free_bytes);
+	cleaned = logfs_gc_segment(sb, segno, dist);
+	log_gc("GC segment #%02x complete - now %x valid\n", segno,
+			valid - cleaned);
+	BUG_ON(cleaned != valid);
+	return 1;
+}
+
+static int logfs_gc_once(struct super_block *sb)
+{
+	struct gc_candidate *cand;
+
+	cand = get_candidate(sb);
+	if (cand)
+		remove_from_list(cand);
+	return __logfs_gc_once(sb, cand);
+}
+
+/* returns 1 if a wrap occurs, 0 otherwise */
+static int logfs_scan_some(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	u32 segno;
+	int i, ret = 0;
+
+	segno = super->s_sweeper;
+	for (i = SCAN_RATIO; i > 0; i--) {
+		segno++;
+		if (segno >= super->s_no_segs) {
+			segno = 0;
+			ret = 1;
+			/* Break out of the loop.  We want to read a single
+			 * block from the segment size on next invocation if
+			 * SCAN_RATIO is set to match block size
+			 */
+			break;
+		}
+
+		scan_segment(sb, segno);
+	}
+	super->s_sweeper = segno;
+	return ret;
+}
+
+/*
+ * In principle, this function should loop forever, looking for GC candidates
+ * and moving data.  LogFS is designed in such a way that this loop is
+ * guaranteed to terminate.
+ *
+ * Limiting the loop to some iterations serves purely to catch cases when
+ * these guarantees have failed.  An actual endless loop is an obvious bug
+ * and should be reported as such.
+ */
+static void __logfs_gc_pass(struct super_block *sb, int target)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_block *block;
+	int round, progress, last_progress = 0;
+
+	if (no_free_segments(sb) >= target &&
+			super->s_no_object_aliases < MAX_OBJ_ALIASES)
+		return;
+
+	log_gc("__logfs_gc_pass(%x)\n", target);
+	for (round = 0; round < SCAN_ROUNDS; ) {
+		if (no_free_segments(sb) >= target)
+			goto write_alias;
+
+		/* Sync in-memory state with on-medium state in case they
+		 * diverged */
+		logfs_write_anchor(sb);
+		round += logfs_scan_some(sb);
+		if (no_free_segments(sb) >= target)
+			goto write_alias;
+		progress = logfs_gc_once(sb);
+		if (progress)
+			last_progress = round;
+		else if (round - last_progress > 2)
+			break;
+		continue;
+
+		/*
+		 * The goto logic is nasty, I just don't know a better way to
+		 * code it.  GC is supposed to ensure two things:
+		 * 1. Enough free segments are available.
+		 * 2. The number of aliases is bounded.
+		 * When 1. is achieved, we take a look at 2. and write back
+		 * some alias-containing blocks, if necessary.  However, after
+		 * each such write we need to go back to 1., as writes can
+		 * consume free segments.
+		 */
+write_alias:
+		if (super->s_no_object_aliases < MAX_OBJ_ALIASES)
+			return;
+		if (list_empty(&super->s_object_alias)) {
+			/* All aliases are still in btree */
+			return;
+		}
+		log_gc("Write back one alias\n");
+		block = list_entry(super->s_object_alias.next,
+				struct logfs_block, alias_list);
+		block->ops->write_block(block);
+		/*
+		 * To round off the nasty goto logic, we reset round here.  It
+		 * is a safety-net for GC not making any progress and limited
+		 * to something reasonably small.  If incremented it for every
+		 * single alias, the loop could terminate rather quickly.
+		 */
+		round = 0;
+	}
+	LOGFS_BUG(sb);
+}
+
+static int wl_ratelimit(struct super_block *sb, u64 *next_event)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	if (*next_event < super->s_gec) {
+		*next_event = super->s_gec + WL_RATELIMIT;
+		return 0;
+	}
+	return 1;
+}
+
+static void logfs_wl_pass(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct gc_candidate *wl_cand, *free_cand;
+
+	if (wl_ratelimit(sb, &super->s_wl_gec_ostore))
+		return;
+
+	wl_cand = first_in_list(&super->s_ec_list);
+	if (!wl_cand)
+		return;
+	free_cand = first_in_list(&super->s_free_list);
+	if (!free_cand)
+		return;
+
+	if (wl_cand->erase_count < free_cand->erase_count + WL_DELTA) {
+		remove_from_list(wl_cand);
+		__logfs_gc_once(sb, wl_cand);
+	}
+}
+
+/*
+ * The journal needs wear leveling as well.  But moving the journal is an
+ * expensive operation so we try to avoid it as much as possible.  And if we
+ * have to do it, we move the whole journal, not individual segments.
+ *
+ * Ratelimiting is not strictly necessary here, it mainly serves to avoid the
+ * calculations.  First we check whether moving the journal would be a
+ * significant improvement.  That means that a) the current journal segments
+ * have more wear than the future journal segments and b) the current journal
+ * segments have more wear than normal ostore segments.
+ * Rationale for b) is that we don't have to move the journal if it is aging
+ * less than the ostore, even if the reserve segments age even less (they are
+ * excluded from wear leveling, after all).
+ * Next we check that the superblocks have less wear than the journal.  Since
+ * moving the journal requires writing the superblocks, we have to protect the
+ * superblocks even more than the journal.
+ *
+ * Also we double the acceptable wear difference, compared to ostore wear
+ * leveling.  Journal data is read and rewritten rapidly, comparatively.  So
+ * soft errors have much less time to accumulate and we allow the journal to
+ * be a bit worse than the ostore.
+ */
+static void logfs_journal_wl_pass(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct gc_candidate *cand;
+	u32 min_journal_ec = -1, max_reserve_ec = 0;
+	int i;
+
+	if (wl_ratelimit(sb, &super->s_wl_gec_journal))
+		return;
+
+	if (super->s_reserve_list.count < super->s_no_journal_segs) {
+		/* Reserve is not full enough to move complete journal */
+		return;
+	}
+
+	journal_for_each(i)
+		if (super->s_journal_seg[i])
+			min_journal_ec = min(min_journal_ec,
+					super->s_journal_ec[i]);
+	cand = rb_entry(rb_first(&super->s_free_list.rb_tree),
+			struct gc_candidate, rb_node);
+	max_reserve_ec = cand->erase_count;
+	for (i = 0; i < 2; i++) {
+		struct logfs_segment_entry se;
+		u32 segno = seg_no(sb, super->s_sb_ofs[i]);
+		u32 ec;
+
+		logfs_get_segment_entry(sb, segno, &se);
+		ec = be32_to_cpu(se.ec_level) >> 4;
+		max_reserve_ec = max(max_reserve_ec, ec);
+	}
+
+	if (min_journal_ec > max_reserve_ec + 2 * WL_DELTA) {
+		do_logfs_journal_wl_pass(sb);
+	}
+}
+
+void logfs_gc_pass(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	//BUG_ON(mutex_trylock(&logfs_super(sb)->s_w_mutex));
+	/* Write journal before free space is getting saturated with dirty
+	 * objects.
+	 */
+	if (super->s_dirty_used_bytes + super->s_dirty_free_bytes
+			+ LOGFS_MAX_OBJECTSIZE >= super->s_free_bytes)
+		logfs_write_anchor(sb);
+	__logfs_gc_pass(sb, super->s_total_levels);
+	logfs_wl_pass(sb);
+	logfs_journal_wl_pass(sb);
+}
+
+static int check_area(struct super_block *sb, int i)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_area *area = super->s_area[i];
+	struct logfs_object_header oh;
+	u32 segno = area->a_segno;
+	u32 ofs = area->a_used_bytes;
+	__be32 crc;
+	int err;
+
+	if (!area->a_is_open)
+		return 0;
+
+	for (ofs = area->a_used_bytes;
+	     ofs <= super->s_segsize - sizeof(oh);
+	     ofs += (u32)be16_to_cpu(oh.len) + sizeof(oh)) {
+		err = wbuf_read(sb, dev_ofs(sb, segno, ofs), sizeof(oh), &oh);
+		if (err)
+			return err;
+
+		if (!memchr_inv(&oh, 0xff, sizeof(oh)))
+			break;
+
+		crc = logfs_crc32(&oh, sizeof(oh) - 4, 4);
+		if (crc != oh.crc) {
+			printk(KERN_INFO "interrupted header at %llx\n",
+					dev_ofs(sb, segno, ofs));
+			return 0;
+		}
+	}
+	if (ofs != area->a_used_bytes) {
+		printk(KERN_INFO "%x bytes unaccounted data found at %llx\n",
+				ofs - area->a_used_bytes,
+				dev_ofs(sb, segno, area->a_used_bytes));
+		area->a_used_bytes = ofs;
+	}
+	return 0;
+}
+
+int logfs_check_areas(struct super_block *sb)
+{
+	int i, err;
+
+	for_each_area(i) {
+		err = check_area(sb, i);
+		if (err)
+			return err;
+	}
+	return 0;
+}
+
+static void logfs_init_candlist(struct candidate_list *list, int maxcount,
+		int sort_by_ec)
+{
+	list->count = 0;
+	list->maxcount = maxcount;
+	list->sort_by_ec = sort_by_ec;
+	list->rb_tree = RB_ROOT;
+}
+
+int logfs_init_gc(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int i;
+
+	btree_init_mempool32(&super->s_cand_tree, super->s_btree_pool);
+	logfs_init_candlist(&super->s_free_list, LIST_SIZE + SCAN_RATIO, 1);
+	logfs_init_candlist(&super->s_reserve_list,
+			super->s_bad_seg_reserve, 1);
+	for_each_area(i)
+		logfs_init_candlist(&super->s_low_list[i], LIST_SIZE, 0);
+	logfs_init_candlist(&super->s_ec_list, LIST_SIZE, 1);
+	return 0;
+}
+
+static void logfs_cleanup_list(struct super_block *sb,
+		struct candidate_list *list)
+{
+	struct gc_candidate *cand;
+
+	while (list->count) {
+		cand = rb_entry(list->rb_tree.rb_node, struct gc_candidate,
+				rb_node);
+		remove_from_list(cand);
+		free_candidate(sb, cand);
+	}
+	BUG_ON(list->rb_tree.rb_node);
+}
+
+void logfs_cleanup_gc(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int i;
+
+	if (!super->s_free_list.count)
+		return;
+
+	/*
+	 * FIXME: The btree may still contain a single empty node.  So we
+	 * call the grim visitor to clean up that mess.  Btree code should
+	 * do it for us, really.
+	 */
+	btree_grim_visitor32(&super->s_cand_tree, 0, NULL);
+	logfs_cleanup_list(sb, &super->s_free_list);
+	logfs_cleanup_list(sb, &super->s_reserve_list);
+	for_each_area(i)
+		logfs_cleanup_list(sb, &super->s_low_list[i]);
+	logfs_cleanup_list(sb, &super->s_ec_list);
+}
diff --git a/fs/logfs/inode.c b/fs/logfs/inode.c
new file mode 100644
index 000000000000..33ec1aeaeec4
--- /dev/null
+++ b/fs/logfs/inode.c
@@ -0,0 +1,417 @@
+/*
+ * fs/logfs/inode.c	- inode handling code
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+#include <linux/writeback.h>
+#include <linux/backing-dev.h>
+
+/*
+ * How soon to reuse old inode numbers?  LogFS doesn't store deleted inodes
+ * on the medium.  It therefore also lacks a method to store the previous
+ * generation number for deleted inodes.  Instead a single generation number
+ * is stored which will be used for new inodes.  Being just a 32bit counter,
+ * this can obvious wrap relatively quickly.  So we only reuse inodes if we
+ * know that a fair number of inodes can be created before we have to increment
+ * the generation again - effectively adding some bits to the counter.
+ * But being too aggressive here means we keep a very large and very sparse
+ * inode file, wasting space on indirect blocks.
+ * So what is a good value?  Beats me.  64k seems moderately bad on both
+ * fronts, so let's use that for now...
+ *
+ * NFS sucks, as everyone already knows.
+ */
+#define INOS_PER_WRAP (0x10000)
+
+/*
+ * Logfs' requirement to read inodes for garbage collection makes life a bit
+ * harder.  GC may have to read inodes that are in I_FREEING state, when they
+ * are being written out - and waiting for GC to make progress, naturally.
+ *
+ * So we cannot just call iget() or some variant of it, but first have to check
+ * wether the inode in question might be in I_FREEING state.  Therefore we
+ * maintain our own per-sb list of "almost deleted" inodes and check against
+ * that list first.  Normally this should be at most 1-2 entries long.
+ *
+ * Also, inodes have logfs-specific reference counting on top of what the vfs
+ * does.  When .destroy_inode is called, normally the reference count will drop
+ * to zero and the inode gets deleted.  But if GC accessed the inode, its
+ * refcount will remain nonzero and final deletion will have to wait.
+ *
+ * As a result we have two sets of functions to get/put inodes:
+ * logfs_safe_iget/logfs_safe_iput	- safe to call from GC context
+ * logfs_iget/iput			- normal version
+ */
+static struct kmem_cache *logfs_inode_cache;
+
+static DEFINE_SPINLOCK(logfs_inode_lock);
+
+static void logfs_inode_setops(struct inode *inode)
+{
+	switch (inode->i_mode & S_IFMT) {
+	case S_IFDIR:
+		inode->i_op = &logfs_dir_iops;
+		inode->i_fop = &logfs_dir_fops;
+		inode->i_mapping->a_ops = &logfs_reg_aops;
+		break;
+	case S_IFREG:
+		inode->i_op = &logfs_reg_iops;
+		inode->i_fop = &logfs_reg_fops;
+		inode->i_mapping->a_ops = &logfs_reg_aops;
+		break;
+	case S_IFLNK:
+		inode->i_op = &logfs_symlink_iops;
+		inode->i_mapping->a_ops = &logfs_reg_aops;
+		break;
+	case S_IFSOCK:	/* fall through */
+	case S_IFBLK:	/* fall through */
+	case S_IFCHR:	/* fall through */
+	case S_IFIFO:
+		init_special_inode(inode, inode->i_mode, inode->i_rdev);
+		break;
+	default:
+		BUG();
+	}
+}
+
+static struct inode *__logfs_iget(struct super_block *sb, ino_t ino)
+{
+	struct inode *inode = iget_locked(sb, ino);
+	int err;
+
+	if (!inode)
+		return ERR_PTR(-ENOMEM);
+	if (!(inode->i_state & I_NEW))
+		return inode;
+
+	err = logfs_read_inode(inode);
+	if (err || inode->i_nlink == 0) {
+		/* inode->i_nlink == 0 can be true when called from
+		 * block validator */
+		/* set i_nlink to 0 to prevent caching */
+		inode->i_nlink = 0;
+		logfs_inode(inode)->li_flags |= LOGFS_IF_ZOMBIE;
+		iget_failed(inode);
+		if (!err)
+			err = -ENOENT;
+		return ERR_PTR(err);
+	}
+
+	logfs_inode_setops(inode);
+	unlock_new_inode(inode);
+	return inode;
+}
+
+struct inode *logfs_iget(struct super_block *sb, ino_t ino)
+{
+	BUG_ON(ino == LOGFS_INO_MASTER);
+	BUG_ON(ino == LOGFS_INO_SEGFILE);
+	return __logfs_iget(sb, ino);
+}
+
+/*
+ * is_cached is set to 1 if we hand out a cached inode, 0 otherwise.
+ * this allows logfs_iput to do the right thing later
+ */
+struct inode *logfs_safe_iget(struct super_block *sb, ino_t ino, int *is_cached)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_inode *li;
+
+	if (ino == LOGFS_INO_MASTER)
+		return super->s_master_inode;
+	if (ino == LOGFS_INO_SEGFILE)
+		return super->s_segfile_inode;
+
+	spin_lock(&logfs_inode_lock);
+	list_for_each_entry(li, &super->s_freeing_list, li_freeing_list)
+		if (li->vfs_inode.i_ino == ino) {
+			li->li_refcount++;
+			spin_unlock(&logfs_inode_lock);
+			*is_cached = 1;
+			return &li->vfs_inode;
+		}
+	spin_unlock(&logfs_inode_lock);
+
+	*is_cached = 0;
+	return __logfs_iget(sb, ino);
+}
+
+static void __logfs_destroy_inode(struct inode *inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	BUG_ON(li->li_block);
+	list_del(&li->li_freeing_list);
+	kmem_cache_free(logfs_inode_cache, li);
+}
+
+static void logfs_destroy_inode(struct inode *inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	BUG_ON(list_empty(&li->li_freeing_list));
+	spin_lock(&logfs_inode_lock);
+	li->li_refcount--;
+	if (li->li_refcount == 0)
+		__logfs_destroy_inode(inode);
+	spin_unlock(&logfs_inode_lock);
+}
+
+void logfs_safe_iput(struct inode *inode, int is_cached)
+{
+	if (inode->i_ino == LOGFS_INO_MASTER)
+		return;
+	if (inode->i_ino == LOGFS_INO_SEGFILE)
+		return;
+
+	if (is_cached) {
+		logfs_destroy_inode(inode);
+		return;
+	}
+
+	iput(inode);
+}
+
+static void logfs_init_inode(struct super_block *sb, struct inode *inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	int i;
+
+	li->li_flags	= 0;
+	li->li_height	= 0;
+	li->li_used_bytes = 0;
+	li->li_block	= NULL;
+	inode->i_uid	= 0;
+	inode->i_gid	= 0;
+	inode->i_size	= 0;
+	inode->i_blocks	= 0;
+	inode->i_ctime	= CURRENT_TIME;
+	inode->i_mtime	= CURRENT_TIME;
+	inode->i_nlink	= 1;
+	INIT_LIST_HEAD(&li->li_freeing_list);
+
+	for (i = 0; i < LOGFS_EMBEDDED_FIELDS; i++)
+		li->li_data[i] = 0;
+
+	return;
+}
+
+static struct inode *logfs_alloc_inode(struct super_block *sb)
+{
+	struct logfs_inode *li;
+
+	li = kmem_cache_alloc(logfs_inode_cache, GFP_NOFS);
+	if (!li)
+		return NULL;
+	logfs_init_inode(sb, &li->vfs_inode);
+	return &li->vfs_inode;
+}
+
+/*
+ * In logfs inodes are written to an inode file.  The inode file, like any
+ * other file, is managed with a inode.  The inode file's inode, aka master
+ * inode, requires special handling in several respects.  First, it cannot be
+ * written to the inode file, so it is stored in the journal instead.
+ *
+ * Secondly, this inode cannot be written back and destroyed before all other
+ * inodes have been written.  The ordering is important.  Linux' VFS is happily
+ * unaware of the ordering constraint and would ordinarily destroy the master
+ * inode at umount time while other inodes are still in use and dirty.  Not
+ * good.
+ *
+ * So logfs makes sure the master inode is not written until all other inodes
+ * have been destroyed.  Sadly, this method has another side-effect.  The VFS
+ * will notice one remaining inode and print a frightening warning message.
+ * Worse, it is impossible to judge whether such a warning was caused by the
+ * master inode or any other inodes have leaked as well.
+ *
+ * Our attempt of solving this is with logfs_new_meta_inode() below.  Its
+ * purpose is to create a new inode that will not trigger the warning if such
+ * an inode is still in use.  An ugly hack, no doubt.  Suggections for
+ * improvement are welcome.
+ */
+struct inode *logfs_new_meta_inode(struct super_block *sb, u64 ino)
+{
+	struct inode *inode;
+
+	inode = logfs_alloc_inode(sb);
+	if (!inode)
+		return ERR_PTR(-ENOMEM);
+
+	inode->i_mode = S_IFREG;
+	inode->i_ino = ino;
+	inode->i_sb = sb;
+
+	/* This is a blatant copy of alloc_inode code.  We'd need alloc_inode
+	 * to be nonstatic, alas. */
+	{
+		struct address_space * const mapping = &inode->i_data;
+
+		mapping->a_ops = &logfs_reg_aops;
+		mapping->host = inode;
+		mapping->flags = 0;
+		mapping_set_gfp_mask(mapping, GFP_NOFS);
+		mapping->assoc_mapping = NULL;
+		mapping->backing_dev_info = &default_backing_dev_info;
+		inode->i_mapping = mapping;
+		inode->i_nlink = 1;
+	}
+
+	return inode;
+}
+
+struct inode *logfs_read_meta_inode(struct super_block *sb, u64 ino)
+{
+	struct inode *inode;
+	int err;
+
+	inode = logfs_new_meta_inode(sb, ino);
+	if (IS_ERR(inode))
+		return inode;
+
+	err = logfs_read_inode(inode);
+	if (err) {
+		destroy_meta_inode(inode);
+		return ERR_PTR(err);
+	}
+	logfs_inode_setops(inode);
+	return inode;
+}
+
+static int logfs_write_inode(struct inode *inode, struct writeback_control *wbc)
+{
+	int ret;
+	long flags = WF_LOCK;
+
+	/* Can only happen if creat() failed.  Safe to skip. */
+	if (logfs_inode(inode)->li_flags & LOGFS_IF_STILLBORN)
+		return 0;
+
+	ret = __logfs_write_inode(inode, flags);
+	LOGFS_BUG_ON(ret, inode->i_sb);
+	return ret;
+}
+
+void destroy_meta_inode(struct inode *inode)
+{
+	if (inode) {
+		if (inode->i_data.nrpages)
+			truncate_inode_pages(&inode->i_data, 0);
+		logfs_clear_inode(inode);
+		kmem_cache_free(logfs_inode_cache, logfs_inode(inode));
+	}
+}
+
+/* called with inode_lock held */
+static void logfs_drop_inode(struct inode *inode)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+	struct logfs_inode *li = logfs_inode(inode);
+
+	spin_lock(&logfs_inode_lock);
+	list_move(&li->li_freeing_list, &super->s_freeing_list);
+	spin_unlock(&logfs_inode_lock);
+	generic_drop_inode(inode);
+}
+
+static void logfs_set_ino_generation(struct super_block *sb,
+		struct inode *inode)
+{
+	struct logfs_super *super = logfs_super(sb);
+	u64 ino;
+
+	mutex_lock(&super->s_journal_mutex);
+	ino = logfs_seek_hole(super->s_master_inode, super->s_last_ino);
+	super->s_last_ino = ino;
+	super->s_inos_till_wrap--;
+	if (super->s_inos_till_wrap < 0) {
+		super->s_last_ino = LOGFS_RESERVED_INOS;
+		super->s_generation++;
+		super->s_inos_till_wrap = INOS_PER_WRAP;
+	}
+	inode->i_ino = ino;
+	inode->i_generation = super->s_generation;
+	mutex_unlock(&super->s_journal_mutex);
+}
+
+struct inode *logfs_new_inode(struct inode *dir, int mode)
+{
+	struct super_block *sb = dir->i_sb;
+	struct inode *inode;
+
+	inode = new_inode(sb);
+	if (!inode)
+		return ERR_PTR(-ENOMEM);
+
+	logfs_init_inode(sb, inode);
+
+	/* inherit parent flags */
+	logfs_inode(inode)->li_flags |=
+		logfs_inode(dir)->li_flags & LOGFS_FL_INHERITED;
+
+	inode->i_mode = mode;
+	logfs_set_ino_generation(sb, inode);
+
+	inode->i_uid = current_fsuid();
+	inode->i_gid = current_fsgid();
+	if (dir->i_mode & S_ISGID) {
+		inode->i_gid = dir->i_gid;
+		if (S_ISDIR(mode))
+			inode->i_mode |= S_ISGID;
+	}
+
+	logfs_inode_setops(inode);
+	insert_inode_hash(inode);
+
+	return inode;
+}
+
+static void logfs_init_once(void *_li)
+{
+	struct logfs_inode *li = _li;
+	int i;
+
+	li->li_flags = 0;
+	li->li_used_bytes = 0;
+	li->li_refcount = 1;
+	for (i = 0; i < LOGFS_EMBEDDED_FIELDS; i++)
+		li->li_data[i] = 0;
+	inode_init_once(&li->vfs_inode);
+}
+
+static int logfs_sync_fs(struct super_block *sb, int wait)
+{
+	/* FIXME: write anchor */
+	logfs_super(sb)->s_devops->sync(sb);
+	return 0;
+}
+
+const struct super_operations logfs_super_operations = {
+	.alloc_inode	= logfs_alloc_inode,
+	.clear_inode	= logfs_clear_inode,
+	.delete_inode	= logfs_delete_inode,
+	.destroy_inode	= logfs_destroy_inode,
+	.drop_inode	= logfs_drop_inode,
+	.write_inode	= logfs_write_inode,
+	.statfs		= logfs_statfs,
+	.sync_fs	= logfs_sync_fs,
+};
+
+int logfs_init_inode_cache(void)
+{
+	logfs_inode_cache = kmem_cache_create("logfs_inode_cache",
+			sizeof(struct logfs_inode), 0, SLAB_RECLAIM_ACCOUNT,
+			logfs_init_once);
+	if (!logfs_inode_cache)
+		return -ENOMEM;
+	return 0;
+}
+
+void logfs_destroy_inode_cache(void)
+{
+	kmem_cache_destroy(logfs_inode_cache);
+}
diff --git a/fs/logfs/journal.c b/fs/logfs/journal.c
new file mode 100644
index 000000000000..6ad30a4c9052
--- /dev/null
+++ b/fs/logfs/journal.c
@@ -0,0 +1,883 @@
+/*
+ * fs/logfs/journal.c	- journal handling code
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ */
+#include "logfs.h"
+
+static void logfs_calc_free(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	u64 reserve, no_segs = super->s_no_segs;
+	s64 free;
+	int i;
+
+	/* superblock segments */
+	no_segs -= 2;
+	super->s_no_journal_segs = 0;
+	/* journal */
+	journal_for_each(i)
+		if (super->s_journal_seg[i]) {
+			no_segs--;
+			super->s_no_journal_segs++;
+		}
+
+	/* open segments plus one extra per level for GC */
+	no_segs -= 2 * super->s_total_levels;
+
+	free = no_segs * (super->s_segsize - LOGFS_SEGMENT_RESERVE);
+	free -= super->s_used_bytes;
+	/* just a bit extra */
+	free -= super->s_total_levels * 4096;
+
+	/* Bad blocks are 'paid' for with speed reserve - the filesystem
+	 * simply gets slower as bad blocks accumulate.  Until the bad blocks
+	 * exceed the speed reserve - then the filesystem gets smaller.
+	 */
+	reserve = super->s_bad_segments + super->s_bad_seg_reserve;
+	reserve *= super->s_segsize - LOGFS_SEGMENT_RESERVE;
+	reserve = max(reserve, super->s_speed_reserve);
+	free -= reserve;
+	if (free < 0)
+		free = 0;
+
+	super->s_free_bytes = free;
+}
+
+static void reserve_sb_and_journal(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct btree_head32 *head = &super->s_reserved_segments;
+	int i, err;
+
+	err = btree_insert32(head, seg_no(sb, super->s_sb_ofs[0]), (void *)1,
+			GFP_KERNEL);
+	BUG_ON(err);
+
+	err = btree_insert32(head, seg_no(sb, super->s_sb_ofs[1]), (void *)1,
+			GFP_KERNEL);
+	BUG_ON(err);
+
+	journal_for_each(i) {
+		if (!super->s_journal_seg[i])
+			continue;
+		err = btree_insert32(head, super->s_journal_seg[i], (void *)1,
+				GFP_KERNEL);
+		BUG_ON(err);
+	}
+}
+
+static void read_dynsb(struct super_block *sb,
+		struct logfs_je_dynsb *dynsb)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	super->s_gec		= be64_to_cpu(dynsb->ds_gec);
+	super->s_sweeper	= be64_to_cpu(dynsb->ds_sweeper);
+	super->s_victim_ino	= be64_to_cpu(dynsb->ds_victim_ino);
+	super->s_rename_dir	= be64_to_cpu(dynsb->ds_rename_dir);
+	super->s_rename_pos	= be64_to_cpu(dynsb->ds_rename_pos);
+	super->s_used_bytes	= be64_to_cpu(dynsb->ds_used_bytes);
+	super->s_generation	= be32_to_cpu(dynsb->ds_generation);
+}
+
+static void read_anchor(struct super_block *sb,
+		struct logfs_je_anchor *da)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *inode = super->s_master_inode;
+	struct logfs_inode *li = logfs_inode(inode);
+	int i;
+
+	super->s_last_ino = be64_to_cpu(da->da_last_ino);
+	li->li_flags	= 0;
+	li->li_height	= da->da_height;
+	i_size_write(inode, be64_to_cpu(da->da_size));
+	li->li_used_bytes = be64_to_cpu(da->da_used_bytes);
+
+	for (i = 0; i < LOGFS_EMBEDDED_FIELDS; i++)
+		li->li_data[i] = be64_to_cpu(da->da_data[i]);
+}
+
+static void read_erasecount(struct super_block *sb,
+		struct logfs_je_journal_ec *ec)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int i;
+
+	journal_for_each(i)
+		super->s_journal_ec[i] = be32_to_cpu(ec->ec[i]);
+}
+
+static int read_area(struct super_block *sb, struct logfs_je_area *a)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_area *area = super->s_area[a->gc_level];
+	u64 ofs;
+	u32 writemask = ~(super->s_writesize - 1);
+
+	if (a->gc_level >= LOGFS_NO_AREAS)
+		return -EIO;
+	if (a->vim != VIM_DEFAULT)
+		return -EIO; /* TODO: close area and continue */
+
+	area->a_used_bytes = be32_to_cpu(a->used_bytes);
+	area->a_written_bytes = area->a_used_bytes & writemask;
+	area->a_segno = be32_to_cpu(a->segno);
+	if (area->a_segno)
+		area->a_is_open = 1;
+
+	ofs = dev_ofs(sb, area->a_segno, area->a_written_bytes);
+	if (super->s_writesize > 1)
+		logfs_buf_recover(area, ofs, a + 1, super->s_writesize);
+	else
+		logfs_buf_recover(area, ofs, NULL, 0);
+	return 0;
+}
+
+static void *unpack(void *from, void *to)
+{
+	struct logfs_journal_header *jh = from;
+	void *data = from + sizeof(struct logfs_journal_header);
+	int err;
+	size_t inlen, outlen;
+
+	inlen = be16_to_cpu(jh->h_len);
+	outlen = be16_to_cpu(jh->h_datalen);
+
+	if (jh->h_compr == COMPR_NONE)
+		memcpy(to, data, inlen);
+	else {
+		err = logfs_uncompress(data, to, inlen, outlen);
+		BUG_ON(err);
+	}
+	return to;
+}
+
+static int __read_je_header(struct super_block *sb, u64 ofs,
+		struct logfs_journal_header *jh)
+{
+	struct logfs_super *super = logfs_super(sb);
+	size_t bufsize = max_t(size_t, sb->s_blocksize, super->s_writesize)
+		+ MAX_JOURNAL_HEADER;
+	u16 type, len, datalen;
+	int err;
+
+	/* read header only */
+	err = wbuf_read(sb, ofs, sizeof(*jh), jh);
+	if (err)
+		return err;
+	type = be16_to_cpu(jh->h_type);
+	len = be16_to_cpu(jh->h_len);
+	datalen = be16_to_cpu(jh->h_datalen);
+	if (len > sb->s_blocksize)
+		return -EIO;
+	if ((type < JE_FIRST) || (type > JE_LAST))
+		return -EIO;
+	if (datalen > bufsize)
+		return -EIO;
+	return 0;
+}
+
+static int __read_je_payload(struct super_block *sb, u64 ofs,
+		struct logfs_journal_header *jh)
+{
+	u16 len;
+	int err;
+
+	len = be16_to_cpu(jh->h_len);
+	err = wbuf_read(sb, ofs + sizeof(*jh), len, jh + 1);
+	if (err)
+		return err;
+	if (jh->h_crc != logfs_crc32(jh, len + sizeof(*jh), 4)) {
+		/* Old code was confused.  It forgot about the header length
+		 * and stopped calculating the crc 16 bytes before the end
+		 * of data - ick!
+		 * FIXME: Remove this hack once the old code is fixed.
+		 */
+		if (jh->h_crc == logfs_crc32(jh, len, 4))
+			WARN_ON_ONCE(1);
+		else
+			return -EIO;
+	}
+	return 0;
+}
+
+/*
+ * jh needs to be large enough to hold the complete entry, not just the header
+ */
+static int __read_je(struct super_block *sb, u64 ofs,
+		struct logfs_journal_header *jh)
+{
+	int err;
+
+	err = __read_je_header(sb, ofs, jh);
+	if (err)
+		return err;
+	return __read_je_payload(sb, ofs, jh);
+}
+
+static int read_je(struct super_block *sb, u64 ofs)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_journal_header *jh = super->s_compressed_je;
+	void *scratch = super->s_je;
+	u16 type, datalen;
+	int err;
+
+	err = __read_je(sb, ofs, jh);
+	if (err)
+		return err;
+	type = be16_to_cpu(jh->h_type);
+	datalen = be16_to_cpu(jh->h_datalen);
+
+	switch (type) {
+	case JE_DYNSB:
+		read_dynsb(sb, unpack(jh, scratch));
+		break;
+	case JE_ANCHOR:
+		read_anchor(sb, unpack(jh, scratch));
+		break;
+	case JE_ERASECOUNT:
+		read_erasecount(sb, unpack(jh, scratch));
+		break;
+	case JE_AREA:
+		read_area(sb, unpack(jh, scratch));
+		break;
+	case JE_OBJ_ALIAS:
+		err = logfs_load_object_aliases(sb, unpack(jh, scratch),
+				datalen);
+		break;
+	default:
+		WARN_ON_ONCE(1);
+		return -EIO;
+	}
+	return err;
+}
+
+static int logfs_read_segment(struct super_block *sb, u32 segno)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_journal_header *jh = super->s_compressed_je;
+	u64 ofs, seg_ofs = dev_ofs(sb, segno, 0);
+	u32 h_ofs, last_ofs = 0;
+	u16 len, datalen, last_len = 0;
+	int i, err;
+
+	/* search for most recent commit */
+	for (h_ofs = 0; h_ofs < super->s_segsize; h_ofs += sizeof(*jh)) {
+		ofs = seg_ofs + h_ofs;
+		err = __read_je_header(sb, ofs, jh);
+		if (err)
+			continue;
+		if (jh->h_type != cpu_to_be16(JE_COMMIT))
+			continue;
+		err = __read_je_payload(sb, ofs, jh);
+		if (err)
+			continue;
+		len = be16_to_cpu(jh->h_len);
+		datalen = be16_to_cpu(jh->h_datalen);
+		if ((datalen > sizeof(super->s_je_array)) ||
+				(datalen % sizeof(__be64)))
+			continue;
+		last_ofs = h_ofs;
+		last_len = datalen;
+		h_ofs += ALIGN(len, sizeof(*jh)) - sizeof(*jh);
+	}
+	/* read commit */
+	if (last_ofs == 0)
+		return -ENOENT;
+	ofs = seg_ofs + last_ofs;
+	log_journal("Read commit from %llx\n", ofs);
+	err = __read_je(sb, ofs, jh);
+	BUG_ON(err); /* We should have caught it in the scan loop already */
+	if (err)
+		return err;
+	/* uncompress */
+	unpack(jh, super->s_je_array);
+	super->s_no_je = last_len / sizeof(__be64);
+	/* iterate over array */
+	for (i = 0; i < super->s_no_je; i++) {
+		err = read_je(sb, be64_to_cpu(super->s_je_array[i]));
+		if (err)
+			return err;
+	}
+	super->s_journal_area->a_segno = segno;
+	return 0;
+}
+
+static u64 read_gec(struct super_block *sb, u32 segno)
+{
+	struct logfs_segment_header sh;
+	__be32 crc;
+	int err;
+
+	if (!segno)
+		return 0;
+	err = wbuf_read(sb, dev_ofs(sb, segno, 0), sizeof(sh), &sh);
+	if (err)
+		return 0;
+	crc = logfs_crc32(&sh, sizeof(sh), 4);
+	if (crc != sh.crc) {
+		WARN_ON(sh.gec != cpu_to_be64(0xffffffffffffffffull));
+		/* Most likely it was just erased */
+		return 0;
+	}
+	return be64_to_cpu(sh.gec);
+}
+
+static int logfs_read_journal(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	u64 gec[LOGFS_JOURNAL_SEGS], max;
+	u32 segno;
+	int i, max_i;
+
+	max = 0;
+	max_i = -1;
+	journal_for_each(i) {
+		segno = super->s_journal_seg[i];
+		gec[i] = read_gec(sb, super->s_journal_seg[i]);
+		if (gec[i] > max) {
+			max = gec[i];
+			max_i = i;
+		}
+	}
+	if (max_i == -1)
+		return -EIO;
+	/* FIXME: Try older segments in case of error */
+	return logfs_read_segment(sb, super->s_journal_seg[max_i]);
+}
+
+/*
+ * First search the current segment (outer loop), then pick the next segment
+ * in the array, skipping any zero entries (inner loop).
+ */
+static void journal_get_free_segment(struct logfs_area *area)
+{
+	struct logfs_super *super = logfs_super(area->a_sb);
+	int i;
+
+	journal_for_each(i) {
+		if (area->a_segno != super->s_journal_seg[i])
+			continue;
+
+		do {
+			i++;
+			if (i == LOGFS_JOURNAL_SEGS)
+				i = 0;
+		} while (!super->s_journal_seg[i]);
+
+		area->a_segno = super->s_journal_seg[i];
+		area->a_erase_count = ++(super->s_journal_ec[i]);
+		log_journal("Journal now at %x (ec %x)\n", area->a_segno,
+				area->a_erase_count);
+		return;
+	}
+	BUG();
+}
+
+static void journal_get_erase_count(struct logfs_area *area)
+{
+	/* erase count is stored globally and incremented in
+	 * journal_get_free_segment() - nothing to do here */
+}
+
+static int journal_erase_segment(struct logfs_area *area)
+{
+	struct super_block *sb = area->a_sb;
+	struct logfs_segment_header sh;
+	u64 ofs;
+	int err;
+
+	err = logfs_erase_segment(sb, area->a_segno, 1);
+	if (err)
+		return err;
+
+	sh.pad = 0;
+	sh.type = SEG_JOURNAL;
+	sh.level = 0;
+	sh.segno = cpu_to_be32(area->a_segno);
+	sh.ec = cpu_to_be32(area->a_erase_count);
+	sh.gec = cpu_to_be64(logfs_super(sb)->s_gec);
+	sh.crc = logfs_crc32(&sh, sizeof(sh), 4);
+
+	/* This causes a bug in segment.c.  Not yet. */
+	//logfs_set_segment_erased(sb, area->a_segno, area->a_erase_count, 0);
+
+	ofs = dev_ofs(sb, area->a_segno, 0);
+	area->a_used_bytes = ALIGN(sizeof(sh), 16);
+	logfs_buf_write(area, ofs, &sh, sizeof(sh));
+	return 0;
+}
+
+static size_t __logfs_write_header(struct logfs_super *super,
+		struct logfs_journal_header *jh, size_t len, size_t datalen,
+		u16 type, u8 compr)
+{
+	jh->h_len	= cpu_to_be16(len);
+	jh->h_type	= cpu_to_be16(type);
+	jh->h_datalen	= cpu_to_be16(datalen);
+	jh->h_compr	= compr;
+	jh->h_pad[0]	= 'H';
+	jh->h_pad[1]	= 'E';
+	jh->h_pad[2]	= 'A';
+	jh->h_pad[3]	= 'D';
+	jh->h_pad[4]	= 'R';
+	jh->h_crc	= logfs_crc32(jh, len + sizeof(*jh), 4);
+	return ALIGN(len, 16) + sizeof(*jh);
+}
+
+static size_t logfs_write_header(struct logfs_super *super,
+		struct logfs_journal_header *jh, size_t datalen, u16 type)
+{
+	size_t len = datalen;
+
+	return __logfs_write_header(super, jh, len, datalen, type, COMPR_NONE);
+}
+
+static inline size_t logfs_journal_erasecount_size(struct logfs_super *super)
+{
+	return LOGFS_JOURNAL_SEGS * sizeof(__be32);
+}
+
+static void *logfs_write_erasecount(struct super_block *sb, void *_ec,
+		u16 *type, size_t *len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_je_journal_ec *ec = _ec;
+	int i;
+
+	journal_for_each(i)
+		ec->ec[i] = cpu_to_be32(super->s_journal_ec[i]);
+	*type = JE_ERASECOUNT;
+	*len = logfs_journal_erasecount_size(super);
+	return ec;
+}
+
+static void account_shadow(void *_shadow, unsigned long _sb, u64 ignore,
+		size_t ignore2)
+{
+	struct logfs_shadow *shadow = _shadow;
+	struct super_block *sb = (void *)_sb;
+	struct logfs_super *super = logfs_super(sb);
+
+	/* consume new space */
+	super->s_free_bytes	  -= shadow->new_len;
+	super->s_used_bytes	  += shadow->new_len;
+	super->s_dirty_used_bytes -= shadow->new_len;
+
+	/* free up old space */
+	super->s_free_bytes	  += shadow->old_len;
+	super->s_used_bytes	  -= shadow->old_len;
+	super->s_dirty_free_bytes -= shadow->old_len;
+
+	logfs_set_segment_used(sb, shadow->old_ofs, -shadow->old_len);
+	logfs_set_segment_used(sb, shadow->new_ofs, shadow->new_len);
+
+	log_journal("account_shadow(%llx, %llx, %x) %llx->%llx %x->%x\n",
+			shadow->ino, shadow->bix, shadow->gc_level,
+			shadow->old_ofs, shadow->new_ofs,
+			shadow->old_len, shadow->new_len);
+	mempool_free(shadow, super->s_shadow_pool);
+}
+
+static void account_shadows(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *inode = super->s_master_inode;
+	struct logfs_inode *li = logfs_inode(inode);
+	struct shadow_tree *tree = &super->s_shadow_tree;
+
+	btree_grim_visitor64(&tree->new, (unsigned long)sb, account_shadow);
+	btree_grim_visitor64(&tree->old, (unsigned long)sb, account_shadow);
+
+	if (li->li_block) {
+		/*
+		 * We never actually use the structure, when attached to the
+		 * master inode.  But it is easier to always free it here than
+		 * to have checks in several places elsewhere when allocating
+		 * it.
+		 */
+		li->li_block->ops->free_block(sb, li->li_block);
+	}
+	BUG_ON((s64)li->li_used_bytes < 0);
+}
+
+static void *__logfs_write_anchor(struct super_block *sb, void *_da,
+		u16 *type, size_t *len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_je_anchor *da = _da;
+	struct inode *inode = super->s_master_inode;
+	struct logfs_inode *li = logfs_inode(inode);
+	int i;
+
+	da->da_height	= li->li_height;
+	da->da_last_ino = cpu_to_be64(super->s_last_ino);
+	da->da_size	= cpu_to_be64(i_size_read(inode));
+	da->da_used_bytes = cpu_to_be64(li->li_used_bytes);
+	for (i = 0; i < LOGFS_EMBEDDED_FIELDS; i++)
+		da->da_data[i] = cpu_to_be64(li->li_data[i]);
+	*type = JE_ANCHOR;
+	*len = sizeof(*da);
+	return da;
+}
+
+static void *logfs_write_dynsb(struct super_block *sb, void *_dynsb,
+		u16 *type, size_t *len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_je_dynsb *dynsb = _dynsb;
+
+	dynsb->ds_gec		= cpu_to_be64(super->s_gec);
+	dynsb->ds_sweeper	= cpu_to_be64(super->s_sweeper);
+	dynsb->ds_victim_ino	= cpu_to_be64(super->s_victim_ino);
+	dynsb->ds_rename_dir	= cpu_to_be64(super->s_rename_dir);
+	dynsb->ds_rename_pos	= cpu_to_be64(super->s_rename_pos);
+	dynsb->ds_used_bytes	= cpu_to_be64(super->s_used_bytes);
+	dynsb->ds_generation	= cpu_to_be32(super->s_generation);
+	*type = JE_DYNSB;
+	*len = sizeof(*dynsb);
+	return dynsb;
+}
+
+static void write_wbuf(struct super_block *sb, struct logfs_area *area,
+		void *wbuf)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	u64 ofs;
+	pgoff_t index;
+	int page_ofs;
+	struct page *page;
+
+	ofs = dev_ofs(sb, area->a_segno,
+			area->a_used_bytes & ~(super->s_writesize - 1));
+	index = ofs >> PAGE_SHIFT;
+	page_ofs = ofs & (PAGE_SIZE - 1);
+
+	page = find_lock_page(mapping, index);
+	BUG_ON(!page);
+	memcpy(wbuf, page_address(page) + page_ofs, super->s_writesize);
+	unlock_page(page);
+}
+
+static void *logfs_write_area(struct super_block *sb, void *_a,
+		u16 *type, size_t *len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_area *area = super->s_area[super->s_sum_index];
+	struct logfs_je_area *a = _a;
+
+	a->vim = VIM_DEFAULT;
+	a->gc_level = super->s_sum_index;
+	a->used_bytes = cpu_to_be32(area->a_used_bytes);
+	a->segno = cpu_to_be32(area->a_segno);
+	if (super->s_writesize > 1)
+		write_wbuf(sb, area, a + 1);
+
+	*type = JE_AREA;
+	*len = sizeof(*a) + super->s_writesize;
+	return a;
+}
+
+static void *logfs_write_commit(struct super_block *sb, void *h,
+		u16 *type, size_t *len)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	*type = JE_COMMIT;
+	*len = super->s_no_je * sizeof(__be64);
+	return super->s_je_array;
+}
+
+static size_t __logfs_write_je(struct super_block *sb, void *buf, u16 type,
+		size_t len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	void *header = super->s_compressed_je;
+	void *data = header + sizeof(struct logfs_journal_header);
+	ssize_t compr_len, pad_len;
+	u8 compr = COMPR_ZLIB;
+
+	if (len == 0)
+		return logfs_write_header(super, header, 0, type);
+
+	compr_len = logfs_compress(buf, data, len, sb->s_blocksize);
+	if (compr_len < 0 || type == JE_ANCHOR) {
+		BUG_ON(len > sb->s_blocksize);
+		memcpy(data, buf, len);
+		compr_len = len;
+		compr = COMPR_NONE;
+	}
+
+	pad_len = ALIGN(compr_len, 16);
+	memset(data + compr_len, 0, pad_len - compr_len);
+
+	return __logfs_write_header(super, header, compr_len, len, type, compr);
+}
+
+static s64 logfs_get_free_bytes(struct logfs_area *area, size_t *bytes,
+		int must_pad)
+{
+	u32 writesize = logfs_super(area->a_sb)->s_writesize;
+	s32 ofs;
+	int ret;
+
+	ret = logfs_open_area(area, *bytes);
+	if (ret)
+		return -EAGAIN;
+
+	ofs = area->a_used_bytes;
+	area->a_used_bytes += *bytes;
+
+	if (must_pad) {
+		area->a_used_bytes = ALIGN(area->a_used_bytes, writesize);
+		*bytes = area->a_used_bytes - ofs;
+	}
+
+	return dev_ofs(area->a_sb, area->a_segno, ofs);
+}
+
+static int logfs_write_je_buf(struct super_block *sb, void *buf, u16 type,
+		size_t buf_len)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_area *area = super->s_journal_area;
+	struct logfs_journal_header *jh = super->s_compressed_je;
+	size_t len;
+	int must_pad = 0;
+	s64 ofs;
+
+	len = __logfs_write_je(sb, buf, type, buf_len);
+	if (jh->h_type == cpu_to_be16(JE_COMMIT))
+		must_pad = 1;
+
+	ofs = logfs_get_free_bytes(area, &len, must_pad);
+	if (ofs < 0)
+		return ofs;
+	logfs_buf_write(area, ofs, super->s_compressed_je, len);
+	super->s_je_array[super->s_no_je++] = cpu_to_be64(ofs);
+	return 0;
+}
+
+static int logfs_write_je(struct super_block *sb,
+		void* (*write)(struct super_block *sb, void *scratch,
+			u16 *type, size_t *len))
+{
+	void *buf;
+	size_t len;
+	u16 type;
+
+	buf = write(sb, logfs_super(sb)->s_je, &type, &len);
+	return logfs_write_je_buf(sb, buf, type, len);
+}
+
+int write_alias_journal(struct super_block *sb, u64 ino, u64 bix,
+		level_t level, int child_no, __be64 val)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_obj_alias *oa = super->s_je;
+	int err = 0, fill = super->s_je_fill;
+
+	log_aliases("logfs_write_obj_aliases #%x(%llx, %llx, %x, %x) %llx\n",
+			fill, ino, bix, level, child_no, be64_to_cpu(val));
+	oa[fill].ino = cpu_to_be64(ino);
+	oa[fill].bix = cpu_to_be64(bix);
+	oa[fill].val = val;
+	oa[fill].level = (__force u8)level;
+	oa[fill].child_no = cpu_to_be16(child_no);
+	fill++;
+	if (fill >= sb->s_blocksize / sizeof(*oa)) {
+		err = logfs_write_je_buf(sb, oa, JE_OBJ_ALIAS, sb->s_blocksize);
+		fill = 0;
+	}
+
+	super->s_je_fill = fill;
+	return err;
+}
+
+static int logfs_write_obj_aliases(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int err;
+
+	log_journal("logfs_write_obj_aliases: %d aliases to write\n",
+			super->s_no_object_aliases);
+	super->s_je_fill = 0;
+	err = logfs_write_obj_aliases_pagecache(sb);
+	if (err)
+		return err;
+
+	if (super->s_je_fill)
+		err = logfs_write_je_buf(sb, super->s_je, JE_OBJ_ALIAS,
+				super->s_je_fill
+				* sizeof(struct logfs_obj_alias));
+	return err;
+}
+
+/*
+ * Write all journal entries.  The goto logic ensures that all journal entries
+ * are written whenever a new segment is used.  It is ugly and potentially a
+ * bit wasteful, but robustness is more important.  With this we can *always*
+ * erase all journal segments except the one containing the most recent commit.
+ */
+void logfs_write_anchor(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_area *area = super->s_journal_area;
+	int i, err;
+
+	if (!(super->s_flags & LOGFS_SB_FLAG_DIRTY))
+		return;
+	super->s_flags &= ~LOGFS_SB_FLAG_DIRTY;
+
+	BUG_ON(super->s_flags & LOGFS_SB_FLAG_SHUTDOWN);
+	mutex_lock(&super->s_journal_mutex);
+
+	/* Do this first or suffer corruption */
+	logfs_sync_segments(sb);
+	account_shadows(sb);
+
+again:
+	super->s_no_je = 0;
+	for_each_area(i) {
+		if (!super->s_area[i]->a_is_open)
+			continue;
+		super->s_sum_index = i;
+		err = logfs_write_je(sb, logfs_write_area);
+		if (err)
+			goto again;
+	}
+	err = logfs_write_obj_aliases(sb);
+	if (err)
+		goto again;
+	err = logfs_write_je(sb, logfs_write_erasecount);
+	if (err)
+		goto again;
+	err = logfs_write_je(sb, __logfs_write_anchor);
+	if (err)
+		goto again;
+	err = logfs_write_je(sb, logfs_write_dynsb);
+	if (err)
+		goto again;
+	/*
+	 * Order is imperative.  First we sync all writes, including the
+	 * non-committed journal writes.  Then we write the final commit and
+	 * sync the current journal segment.
+	 * There is a theoretical bug here.  Syncing the journal segment will
+	 * write a number of journal entries and the final commit.  All these
+	 * are written in a single operation.  If the device layer writes the
+	 * data back-to-front, the commit will precede the other journal
+	 * entries, leaving a race window.
+	 * Two fixes are possible.  Preferred is to fix the device layer to
+	 * ensure writes happen front-to-back.  Alternatively we can insert
+	 * another logfs_sync_area() super->s_devops->sync() combo before
+	 * writing the commit.
+	 */
+	/*
+	 * On another subject, super->s_devops->sync is usually not necessary.
+	 * Unless called from sys_sync or friends, a barrier would suffice.
+	 */
+	super->s_devops->sync(sb);
+	err = logfs_write_je(sb, logfs_write_commit);
+	if (err)
+		goto again;
+	log_journal("Write commit to %llx\n",
+			be64_to_cpu(super->s_je_array[super->s_no_je - 1]));
+	logfs_sync_area(area);
+	BUG_ON(area->a_used_bytes != area->a_written_bytes);
+	super->s_devops->sync(sb);
+
+	mutex_unlock(&super->s_journal_mutex);
+	return;
+}
+
+void do_logfs_journal_wl_pass(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_area *area = super->s_journal_area;
+	u32 segno, ec;
+	int i, err;
+
+	log_journal("Journal requires wear-leveling.\n");
+	/* Drop old segments */
+	journal_for_each(i)
+		if (super->s_journal_seg[i]) {
+			logfs_set_segment_unreserved(sb,
+					super->s_journal_seg[i],
+					super->s_journal_ec[i]);
+			super->s_journal_seg[i] = 0;
+			super->s_journal_ec[i] = 0;
+		}
+	/* Get new segments */
+	for (i = 0; i < super->s_no_journal_segs; i++) {
+		segno = get_best_cand(sb, &super->s_reserve_list, &ec);
+		super->s_journal_seg[i] = segno;
+		super->s_journal_ec[i] = ec;
+		logfs_set_segment_reserved(sb, segno);
+	}
+	/* Manually move journal_area */
+	area->a_segno = super->s_journal_seg[0];
+	area->a_is_open = 0;
+	area->a_used_bytes = 0;
+	/* Write journal */
+	logfs_write_anchor(sb);
+	/* Write superblocks */
+	err = logfs_write_sb(sb);
+	BUG_ON(err);
+}
+
+static const struct logfs_area_ops journal_area_ops = {
+	.get_free_segment	= journal_get_free_segment,
+	.get_erase_count	= journal_get_erase_count,
+	.erase_segment		= journal_erase_segment,
+};
+
+int logfs_init_journal(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	size_t bufsize = max_t(size_t, sb->s_blocksize, super->s_writesize)
+		+ MAX_JOURNAL_HEADER;
+	int ret = -ENOMEM;
+
+	mutex_init(&super->s_journal_mutex);
+	btree_init_mempool32(&super->s_reserved_segments, super->s_btree_pool);
+
+	super->s_je = kzalloc(bufsize, GFP_KERNEL);
+	if (!super->s_je)
+		return ret;
+
+	super->s_compressed_je = kzalloc(bufsize, GFP_KERNEL);
+	if (!super->s_compressed_je)
+		return ret;
+
+	super->s_master_inode = logfs_new_meta_inode(sb, LOGFS_INO_MASTER);
+	if (IS_ERR(super->s_master_inode))
+		return PTR_ERR(super->s_master_inode);
+
+	ret = logfs_read_journal(sb);
+	if (ret)
+		return -EIO;
+
+	reserve_sb_and_journal(sb);
+	logfs_calc_free(sb);
+
+	super->s_journal_area->a_ops = &journal_area_ops;
+	return 0;
+}
+
+void logfs_cleanup_journal(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	btree_grim_visitor32(&super->s_reserved_segments, 0, NULL);
+	destroy_meta_inode(super->s_master_inode);
+	super->s_master_inode = NULL;
+
+	kfree(super->s_compressed_je);
+	kfree(super->s_je);
+}
diff --git a/fs/logfs/logfs.h b/fs/logfs/logfs.h
new file mode 100644
index 000000000000..129779431373
--- /dev/null
+++ b/fs/logfs/logfs.h
@@ -0,0 +1,724 @@
+/*
+ * fs/logfs/logfs.h
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ *
+ * Private header for logfs.
+ */
+#ifndef FS_LOGFS_LOGFS_H
+#define FS_LOGFS_LOGFS_H
+
+#undef __CHECK_ENDIAN__
+#define __CHECK_ENDIAN__
+
+#include <linux/btree.h>
+#include <linux/crc32.h>
+#include <linux/fs.h>
+#include <linux/kernel.h>
+#include <linux/mempool.h>
+#include <linux/pagemap.h>
+#include <linux/mtd/mtd.h>
+#include "logfs_abi.h"
+
+#define LOGFS_DEBUG_SUPER	(0x0001)
+#define LOGFS_DEBUG_SEGMENT	(0x0002)
+#define LOGFS_DEBUG_JOURNAL	(0x0004)
+#define LOGFS_DEBUG_DIR		(0x0008)
+#define LOGFS_DEBUG_FILE	(0x0010)
+#define LOGFS_DEBUG_INODE	(0x0020)
+#define LOGFS_DEBUG_READWRITE	(0x0040)
+#define LOGFS_DEBUG_GC		(0x0080)
+#define LOGFS_DEBUG_GC_NOISY	(0x0100)
+#define LOGFS_DEBUG_ALIASES	(0x0200)
+#define LOGFS_DEBUG_BLOCKMOVE	(0x0400)
+#define LOGFS_DEBUG_ALL		(0xffffffff)
+
+#define LOGFS_DEBUG		(0x01)
+/*
+ * To enable specific log messages, simply define LOGFS_DEBUG to match any
+ * or all of the above.
+ */
+#ifndef LOGFS_DEBUG
+#define LOGFS_DEBUG		(0)
+#endif
+
+#define log_cond(cond, fmt, arg...) do {	\
+	if (cond)				\
+		printk(KERN_DEBUG fmt, ##arg);	\
+} while (0)
+
+#define log_super(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_SUPER, fmt, ##arg)
+#define log_segment(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_SEGMENT, fmt, ##arg)
+#define log_journal(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_JOURNAL, fmt, ##arg)
+#define log_dir(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_DIR, fmt, ##arg)
+#define log_file(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_FILE, fmt, ##arg)
+#define log_inode(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_INODE, fmt, ##arg)
+#define log_readwrite(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_READWRITE, fmt, ##arg)
+#define log_gc(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_GC, fmt, ##arg)
+#define log_gc_noisy(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_GC_NOISY, fmt, ##arg)
+#define log_aliases(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_ALIASES, fmt, ##arg)
+#define log_blockmove(fmt, arg...) \
+	log_cond(LOGFS_DEBUG & LOGFS_DEBUG_BLOCKMOVE, fmt, ##arg)
+
+#define PG_pre_locked		PG_owner_priv_1
+#define PagePreLocked(page)	test_bit(PG_pre_locked, &(page)->flags)
+#define SetPagePreLocked(page)	set_bit(PG_pre_locked, &(page)->flags)
+#define ClearPagePreLocked(page) clear_bit(PG_pre_locked, &(page)->flags)
+
+/* FIXME: This should really be somewhere in the 64bit area. */
+#define LOGFS_LINK_MAX		(1<<30)
+
+/* Read-only filesystem */
+#define LOGFS_SB_FLAG_RO	0x0001
+#define LOGFS_SB_FLAG_DIRTY	0x0002
+#define LOGFS_SB_FLAG_OBJ_ALIAS	0x0004
+#define LOGFS_SB_FLAG_SHUTDOWN	0x0008
+
+/* Write Control Flags */
+#define WF_LOCK			0x01 /* take write lock */
+#define WF_WRITE		0x02 /* write block */
+#define WF_DELETE		0x04 /* delete old block */
+
+typedef u8 __bitwise level_t;
+typedef u8 __bitwise gc_level_t;
+
+#define LEVEL(level) ((__force level_t)(level))
+#define GC_LEVEL(gc_level) ((__force gc_level_t)(gc_level))
+
+#define SUBLEVEL(level) ( (void)((level) == LEVEL(1)),	\
+		(__force level_t)((__force u8)(level) - 1) )
+
+/**
+ * struct logfs_area - area management information
+ *
+ * @a_sb:			the superblock this area belongs to
+ * @a_is_open:			1 if the area is currently open, else 0
+ * @a_segno:			segment number of area
+ * @a_written_bytes:		number of bytes already written back
+ * @a_used_bytes:		number of used bytes
+ * @a_ops:			area operations (either journal or ostore)
+ * @a_erase_count:		erase count
+ * @a_level:			GC level
+ */
+struct logfs_area { /* a segment open for writing */
+	struct super_block *a_sb;
+	int	a_is_open;
+	u32	a_segno;
+	u32	a_written_bytes;
+	u32	a_used_bytes;
+	const struct logfs_area_ops *a_ops;
+	u32	a_erase_count;
+	gc_level_t a_level;
+};
+
+/**
+ * struct logfs_area_ops - area operations
+ *
+ * @get_free_segment:		fill area->ofs with the offset of a free segment
+ * @get_erase_count:		fill area->erase_count (needs area->ofs)
+ * @erase_segment:		erase and setup segment
+ */
+struct logfs_area_ops {
+	void	(*get_free_segment)(struct logfs_area *area);
+	void	(*get_erase_count)(struct logfs_area *area);
+	int	(*erase_segment)(struct logfs_area *area);
+};
+
+/**
+ * struct logfs_device_ops - device access operations
+ *
+ * @readpage:			read one page (mm page)
+ * @writeseg:			write one segment.  may be a partial segment
+ * @erase:			erase one segment
+ * @read:			read from the device
+ * @erase:			erase part of the device
+ */
+struct logfs_device_ops {
+	struct page *(*find_first_sb)(struct super_block *sb, u64 *ofs);
+	struct page *(*find_last_sb)(struct super_block *sb, u64 *ofs);
+	int (*write_sb)(struct super_block *sb, struct page *page);
+	int (*readpage)(void *_sb, struct page *page);
+	void (*writeseg)(struct super_block *sb, u64 ofs, size_t len);
+	int (*erase)(struct super_block *sb, loff_t ofs, size_t len,
+			int ensure_write);
+	void (*sync)(struct super_block *sb);
+	void (*put_device)(struct super_block *sb);
+};
+
+/**
+ * struct candidate_list - list of similar candidates
+ */
+struct candidate_list {
+	struct rb_root rb_tree;
+	int count;
+	int maxcount;
+	int sort_by_ec;
+};
+
+/**
+ * struct gc_candidate - "candidate" segment to be garbage collected next
+ *
+ * @list:			list (either free of low)
+ * @segno:			segment number
+ * @valid:			number of valid bytes
+ * @erase_count:		erase count of segment
+ * @dist:			distance from tree root
+ *
+ * Candidates can be on two lists.  The free list contains electees rather
+ * than candidates - segments that no longer contain any valid data.  The
+ * low list contains candidates to be picked for GC.  It should be kept
+ * short.  It is not required to always pick a perfect candidate.  In the
+ * worst case GC will have to move more data than absolutely necessary.
+ */
+struct gc_candidate {
+	struct rb_node rb_node;
+	struct candidate_list *list;
+	u32	segno;
+	u32	valid;
+	u32	erase_count;
+	u8	dist;
+};
+
+/**
+ * struct logfs_journal_entry - temporary structure used during journal scan
+ *
+ * @used:
+ * @version:			normalized version
+ * @len:			length
+ * @offset:			offset
+ */
+struct logfs_journal_entry {
+	int used;
+	s16 version;
+	u16 len;
+	u16 datalen;
+	u64 offset;
+};
+
+enum transaction_state {
+	CREATE_1 = 1,
+	CREATE_2,
+	UNLINK_1,
+	UNLINK_2,
+	CROSS_RENAME_1,
+	CROSS_RENAME_2,
+	TARGET_RENAME_1,
+	TARGET_RENAME_2,
+	TARGET_RENAME_3
+};
+
+/**
+ * struct logfs_transaction - essential fields to support atomic dirops
+ *
+ * @ino:			target inode
+ * @dir:			inode of directory containing dentry
+ * @pos:			pos of dentry in directory
+ */
+struct logfs_transaction {
+	enum transaction_state state;
+	u64	 ino;
+	u64	 dir;
+	u64	 pos;
+};
+
+/**
+ * struct logfs_shadow - old block in the shadow of a not-yet-committed new one
+ * @old_ofs:			offset of old block on medium
+ * @new_ofs:			offset of new block on medium
+ * @ino:			inode number
+ * @bix:			block index
+ * @old_len:			size of old block, including header
+ * @new_len:			size of new block, including header
+ * @level:			block level
+ */
+struct logfs_shadow {
+	u64 old_ofs;
+	u64 new_ofs;
+	u64 ino;
+	u64 bix;
+	int old_len;
+	int new_len;
+	gc_level_t gc_level;
+};
+
+/**
+ * struct shadow_tree
+ * @new:			shadows where old_ofs==0, indexed by new_ofs
+ * @old:			shadows where old_ofs!=0, indexed by old_ofs
+ */
+struct shadow_tree {
+	struct btree_head64 new;
+	struct btree_head64 old;
+};
+
+struct object_alias_item {
+	struct list_head list;
+	__be64 val;
+	int child_no;
+};
+
+/**
+ * struct logfs_block - contains any block state
+ * @type:			indirect block or inode
+ * @full:			number of fully populated children
+ * @partial:			number of partially populated children
+ *
+ * Most blocks are directly represented by page cache pages.  But when a block
+ * becomes dirty, is part of a transaction, contains aliases or is otherwise
+ * special, a struct logfs_block is allocated to track the additional state.
+ * Inodes are very similar to indirect blocks, so they can also get one of
+ * these structures added when appropriate.
+ */
+#define BLOCK_INDIRECT	1	/* Indirect block */
+#define BLOCK_INODE	2	/* Inode */
+struct logfs_block_ops;
+struct logfs_block {
+	struct list_head alias_list;
+	struct list_head item_list;
+	struct super_block *sb;
+	u64 ino;
+	u64 bix;
+	level_t level;
+	struct page *page;
+	struct inode *inode;
+	struct logfs_transaction *ta;
+	unsigned long alias_map[LOGFS_BLOCK_FACTOR / BITS_PER_LONG];
+	struct logfs_block_ops *ops;
+	int full;
+	int partial;
+	int reserved_bytes;
+};
+
+typedef int write_alias_t(struct super_block *sb, u64 ino, u64 bix,
+		level_t level, int child_no, __be64 val);
+struct logfs_block_ops {
+	void	(*write_block)(struct logfs_block *block);
+	gc_level_t	(*block_level)(struct logfs_block *block);
+	void	(*free_block)(struct super_block *sb, struct logfs_block*block);
+	int	(*write_alias)(struct super_block *sb,
+			struct logfs_block *block,
+			write_alias_t *write_one_alias);
+};
+
+struct logfs_super {
+	struct mtd_info *s_mtd;			/* underlying device */
+	struct block_device *s_bdev;		/* underlying device */
+	const struct logfs_device_ops *s_devops;/* device access */
+	struct inode	*s_master_inode;	/* inode file */
+	struct inode	*s_segfile_inode;	/* segment file */
+	struct inode *s_mapping_inode;		/* device mapping */
+	atomic_t s_pending_writes;		/* outstanting bios */
+	long	 s_flags;
+	mempool_t *s_btree_pool;		/* for btree nodes */
+	mempool_t *s_alias_pool;		/* aliases in segment.c */
+	u64	 s_feature_incompat;
+	u64	 s_feature_ro_compat;
+	u64	 s_feature_compat;
+	u64	 s_feature_flags;
+	u64	 s_sb_ofs[2];
+	struct page *s_erase_page;		/* for dev_bdev.c */
+	/* alias.c fields */
+	struct btree_head32 s_segment_alias;	/* remapped segments */
+	int	 s_no_object_aliases;
+	struct list_head s_object_alias;	/* remapped objects */
+	struct btree_head128 s_object_alias_tree; /* remapped objects */
+	struct mutex s_object_alias_mutex;
+	/* dir.c fields */
+	struct mutex s_dirop_mutex;		/* for creat/unlink/rename */
+	u64	 s_victim_ino;			/* used for atomic dir-ops */
+	u64	 s_rename_dir;			/* source directory ino */
+	u64	 s_rename_pos;			/* position of source dd */
+	/* gc.c fields */
+	long	 s_segsize;			/* size of a segment */
+	int	 s_segshift;			/* log2 of segment size */
+	long	 s_segmask;			/* 1 << s_segshift - 1 */
+	long	 s_no_segs;			/* segments on device */
+	long	 s_no_journal_segs;		/* segments used for journal */
+	long	 s_no_blocks;			/* blocks per segment */
+	long	 s_writesize;			/* minimum write size */
+	int	 s_writeshift;			/* log2 of write size */
+	u64	 s_size;			/* filesystem size */
+	struct logfs_area *s_area[LOGFS_NO_AREAS];	/* open segment array */
+	u64	 s_gec;				/* global erase count */
+	u64	 s_wl_gec_ostore;		/* time of last wl event */
+	u64	 s_wl_gec_journal;		/* time of last wl event */
+	u64	 s_sweeper;			/* current sweeper pos */
+	u8	 s_ifile_levels;		/* max level of ifile */
+	u8	 s_iblock_levels;		/* max level of regular files */
+	u8	 s_data_levels;			/* # of segments to leaf block*/
+	u8	 s_total_levels;		/* sum of above three */
+	struct btree_head32 s_cand_tree;	/* all candidates */
+	struct candidate_list s_free_list;	/* 100% free segments */
+	struct candidate_list s_reserve_list;	/* Bad segment reserve */
+	struct candidate_list s_low_list[LOGFS_NO_AREAS];/* good candidates */
+	struct candidate_list s_ec_list;	/* wear level candidates */
+	struct btree_head32 s_reserved_segments;/* sb, journal, bad, etc. */
+	/* inode.c fields */
+	u64	 s_last_ino;			/* highest ino used */
+	long	 s_inos_till_wrap;
+	u32	 s_generation;			/* i_generation for new files */
+	struct list_head s_freeing_list;	/* inodes being freed */
+	/* journal.c fields */
+	struct mutex s_journal_mutex;
+	void	*s_je;				/* journal entry to compress */
+	void	*s_compressed_je;		/* block to write to journal */
+	u32	 s_journal_seg[LOGFS_JOURNAL_SEGS]; /* journal segments */
+	u32	 s_journal_ec[LOGFS_JOURNAL_SEGS]; /* journal erasecounts */
+	u64	 s_last_version;
+	struct logfs_area *s_journal_area;	/* open journal segment */
+	__be64	s_je_array[64];
+	int	s_no_je;
+
+	int	 s_sum_index;			/* for the 12 summaries */
+	struct shadow_tree s_shadow_tree;
+	int	 s_je_fill;			/* index of current je */
+	/* readwrite.c fields */
+	struct mutex s_write_mutex;
+	int	 s_lock_count;
+	mempool_t *s_block_pool;		/* struct logfs_block pool */
+	mempool_t *s_shadow_pool;		/* struct logfs_shadow pool */
+	/*
+	 * Space accounting:
+	 * - s_used_bytes specifies space used to store valid data objects.
+	 * - s_dirty_used_bytes is space used to store non-committed data
+	 *   objects.  Those objects have already been written themselves,
+	 *   but they don't become valid until all indirect blocks up to the
+	 *   journal have been written as well.
+	 * - s_dirty_free_bytes is space used to store the old copy of a
+	 *   replaced object, as long as the replacement is non-committed.
+	 *   In other words, it is the amount of space freed when all dirty
+	 *   blocks are written back.
+	 * - s_free_bytes is the amount of free space available for any
+	 *   purpose.
+	 * - s_root_reserve is the amount of free space available only to
+	 *   the root user.  Non-privileged users can no longer write once
+	 *   this watermark has been reached.
+	 * - s_speed_reserve is space which remains unused to speed up
+	 *   garbage collection performance.
+	 * - s_dirty_pages is the space reserved for currently dirty pages.
+	 *   It is a pessimistic estimate, so some/most will get freed on
+	 *   page writeback.
+	 *
+	 * s_used_bytes + s_free_bytes + s_speed_reserve = total usable size
+	 */
+	u64	 s_free_bytes;
+	u64	 s_used_bytes;
+	u64	 s_dirty_free_bytes;
+	u64	 s_dirty_used_bytes;
+	u64	 s_root_reserve;
+	u64	 s_speed_reserve;
+	u64	 s_dirty_pages;
+	/* Bad block handling:
+	 * - s_bad_seg_reserve is a number of segments usually kept
+	 *   free.  When encountering bad blocks, the affected segment's data
+	 *   is _temporarily_ moved to a reserved segment.
+	 * - s_bad_segments is the number of known bad segments.
+	 */
+	u32	 s_bad_seg_reserve;
+	u32	 s_bad_segments;
+};
+
+/**
+ * struct logfs_inode - in-memory inode
+ *
+ * @vfs_inode:			struct inode
+ * @li_data:			data pointers
+ * @li_used_bytes:		number of used bytes
+ * @li_freeing_list:		used to track inodes currently being freed
+ * @li_flags:			inode flags
+ * @li_refcount:		number of internal (GC-induced) references
+ */
+struct logfs_inode {
+	struct inode vfs_inode;
+	u64	li_data[LOGFS_EMBEDDED_FIELDS];
+	u64	li_used_bytes;
+	struct list_head li_freeing_list;
+	struct logfs_block *li_block;
+	u32	li_flags;
+	u8	li_height;
+	int	li_refcount;
+};
+
+#define journal_for_each(__i) for (__i = 0; __i < LOGFS_JOURNAL_SEGS; __i++)
+#define for_each_area(__i) for (__i = 0; __i < LOGFS_NO_AREAS; __i++)
+#define for_each_area_down(__i) for (__i = LOGFS_NO_AREAS - 1; __i >= 0; __i--)
+
+/* compr.c */
+int logfs_compress(void *in, void *out, size_t inlen, size_t outlen);
+int logfs_uncompress(void *in, void *out, size_t inlen, size_t outlen);
+int __init logfs_compr_init(void);
+void logfs_compr_exit(void);
+
+/* dev_bdev.c */
+#ifdef CONFIG_BLOCK
+int logfs_get_sb_bdev(struct file_system_type *type, int flags,
+		const char *devname, struct vfsmount *mnt);
+#else
+static inline int logfs_get_sb_bdev(struct file_system_type *type, int flags,
+		const char *devname, struct vfsmount *mnt)
+{
+	return -ENODEV;
+}
+#endif
+
+/* dev_mtd.c */
+#ifdef CONFIG_MTD
+int logfs_get_sb_mtd(struct file_system_type *type, int flags,
+		int mtdnr, struct vfsmount *mnt);
+#else
+static inline int logfs_get_sb_mtd(struct file_system_type *type, int flags,
+		int mtdnr, struct vfsmount *mnt)
+{
+	return -ENODEV;
+}
+#endif
+
+/* dir.c */
+extern const struct inode_operations logfs_symlink_iops;
+extern const struct inode_operations logfs_dir_iops;
+extern const struct file_operations logfs_dir_fops;
+int logfs_replay_journal(struct super_block *sb);
+
+/* file.c */
+extern const struct inode_operations logfs_reg_iops;
+extern const struct file_operations logfs_reg_fops;
+extern const struct address_space_operations logfs_reg_aops;
+int logfs_readpage(struct file *file, struct page *page);
+int logfs_ioctl(struct inode *inode, struct file *file, unsigned int cmd,
+		unsigned long arg);
+int logfs_fsync(struct file *file, struct dentry *dentry, int datasync);
+
+/* gc.c */
+u32 get_best_cand(struct super_block *sb, struct candidate_list *list, u32 *ec);
+void logfs_gc_pass(struct super_block *sb);
+int logfs_check_areas(struct super_block *sb);
+int logfs_init_gc(struct super_block *sb);
+void logfs_cleanup_gc(struct super_block *sb);
+
+/* inode.c */
+extern const struct super_operations logfs_super_operations;
+struct inode *logfs_iget(struct super_block *sb, ino_t ino);
+struct inode *logfs_safe_iget(struct super_block *sb, ino_t ino, int *cookie);
+void logfs_safe_iput(struct inode *inode, int cookie);
+struct inode *logfs_new_inode(struct inode *dir, int mode);
+struct inode *logfs_new_meta_inode(struct super_block *sb, u64 ino);
+struct inode *logfs_read_meta_inode(struct super_block *sb, u64 ino);
+int logfs_init_inode_cache(void);
+void logfs_destroy_inode_cache(void);
+void destroy_meta_inode(struct inode *inode);
+void logfs_set_blocks(struct inode *inode, u64 no);
+/* these logically belong into inode.c but actually reside in readwrite.c */
+int logfs_read_inode(struct inode *inode);
+int __logfs_write_inode(struct inode *inode, long flags);
+void logfs_delete_inode(struct inode *inode);
+void logfs_clear_inode(struct inode *inode);
+
+/* journal.c */
+void logfs_write_anchor(struct super_block *sb);
+int logfs_init_journal(struct super_block *sb);
+void logfs_cleanup_journal(struct super_block *sb);
+int write_alias_journal(struct super_block *sb, u64 ino, u64 bix,
+		level_t level, int child_no, __be64 val);
+void do_logfs_journal_wl_pass(struct super_block *sb);
+
+/* readwrite.c */
+pgoff_t logfs_pack_index(u64 bix, level_t level);
+void logfs_unpack_index(pgoff_t index, u64 *bix, level_t *level);
+int logfs_inode_write(struct inode *inode, const void *buf, size_t count,
+		loff_t bix, long flags, struct shadow_tree *shadow_tree);
+int logfs_readpage_nolock(struct page *page);
+int logfs_write_buf(struct inode *inode, struct page *page, long flags);
+int logfs_delete(struct inode *inode, pgoff_t index,
+		struct shadow_tree *shadow_tree);
+int logfs_rewrite_block(struct inode *inode, u64 bix, u64 ofs,
+		gc_level_t gc_level, long flags);
+int logfs_is_valid_block(struct super_block *sb, u64 ofs, u64 ino, u64 bix,
+		gc_level_t gc_level);
+int logfs_truncate(struct inode *inode, u64 size);
+u64 logfs_seek_hole(struct inode *inode, u64 bix);
+u64 logfs_seek_data(struct inode *inode, u64 bix);
+int logfs_open_segfile(struct super_block *sb);
+int logfs_init_rw(struct super_block *sb);
+void logfs_cleanup_rw(struct super_block *sb);
+void logfs_add_transaction(struct inode *inode, struct logfs_transaction *ta);
+void logfs_del_transaction(struct inode *inode, struct logfs_transaction *ta);
+void logfs_write_block(struct logfs_block *block, long flags);
+int logfs_write_obj_aliases_pagecache(struct super_block *sb);
+void logfs_get_segment_entry(struct super_block *sb, u32 segno,
+		struct logfs_segment_entry *se);
+void logfs_set_segment_used(struct super_block *sb, u64 ofs, int increment);
+void logfs_set_segment_erased(struct super_block *sb, u32 segno, u32 ec,
+		gc_level_t gc_level);
+void logfs_set_segment_reserved(struct super_block *sb, u32 segno);
+void logfs_set_segment_unreserved(struct super_block *sb, u32 segno, u32 ec);
+struct logfs_block *__alloc_block(struct super_block *sb,
+		u64 ino, u64 bix, level_t level);
+void __free_block(struct super_block *sb, struct logfs_block *block);
+void btree_write_block(struct logfs_block *block);
+void initialize_block_counters(struct page *page, struct logfs_block *block,
+		__be64 *array, int page_is_empty);
+int logfs_exist_block(struct inode *inode, u64 bix);
+int get_page_reserve(struct inode *inode, struct page *page);
+extern struct logfs_block_ops indirect_block_ops;
+
+/* segment.c */
+int logfs_erase_segment(struct super_block *sb, u32 ofs, int ensure_erase);
+int wbuf_read(struct super_block *sb, u64 ofs, size_t len, void *buf);
+int logfs_segment_read(struct inode *inode, struct page *page, u64 ofs, u64 bix,
+		level_t level);
+int logfs_segment_write(struct inode *inode, struct page *page,
+		struct logfs_shadow *shadow);
+int logfs_segment_delete(struct inode *inode, struct logfs_shadow *shadow);
+int logfs_load_object_aliases(struct super_block *sb,
+		struct logfs_obj_alias *oa, int count);
+void move_page_to_btree(struct page *page);
+int logfs_init_mapping(struct super_block *sb);
+void logfs_sync_area(struct logfs_area *area);
+void logfs_sync_segments(struct super_block *sb);
+
+/* area handling */
+int logfs_init_areas(struct super_block *sb);
+void logfs_cleanup_areas(struct super_block *sb);
+int logfs_open_area(struct logfs_area *area, size_t bytes);
+void __logfs_buf_write(struct logfs_area *area, u64 ofs, void *buf, size_t len,
+		int use_filler);
+
+static inline void logfs_buf_write(struct logfs_area *area, u64 ofs,
+		void *buf, size_t len)
+{
+	__logfs_buf_write(area, ofs, buf, len, 0);
+}
+
+static inline void logfs_buf_recover(struct logfs_area *area, u64 ofs,
+		void *buf, size_t len)
+{
+	__logfs_buf_write(area, ofs, buf, len, 1);
+}
+
+/* super.c */
+struct page *emergency_read_begin(struct address_space *mapping, pgoff_t index);
+void emergency_read_end(struct page *page);
+void logfs_crash_dump(struct super_block *sb);
+void *memchr_inv(const void *s, int c, size_t n);
+int logfs_statfs(struct dentry *dentry, struct kstatfs *stats);
+int logfs_get_sb_device(struct file_system_type *type, int flags,
+		struct mtd_info *mtd, struct block_device *bdev,
+		const struct logfs_device_ops *devops, struct vfsmount *mnt);
+int logfs_check_ds(struct logfs_disk_super *ds);
+int logfs_write_sb(struct super_block *sb);
+
+static inline struct logfs_super *logfs_super(struct super_block *sb)
+{
+	return sb->s_fs_info;
+}
+
+static inline struct logfs_inode *logfs_inode(struct inode *inode)
+{
+	return container_of(inode, struct logfs_inode, vfs_inode);
+}
+
+static inline void logfs_set_ro(struct super_block *sb)
+{
+	logfs_super(sb)->s_flags |= LOGFS_SB_FLAG_RO;
+}
+
+#define LOGFS_BUG(sb) do {					\
+	struct super_block *__sb = sb;				\
+	logfs_crash_dump(__sb);					\
+	logfs_super(__sb)->s_flags |= LOGFS_SB_FLAG_RO;		\
+	BUG();							\
+} while (0)
+
+#define LOGFS_BUG_ON(condition, sb) \
+	do { if (unlikely(condition)) LOGFS_BUG((sb)); } while (0)
+
+static inline __be32 logfs_crc32(void *data, size_t len, size_t skip)
+{
+	return cpu_to_be32(crc32(~0, data+skip, len-skip));
+}
+
+static inline u8 logfs_type(struct inode *inode)
+{
+	return (inode->i_mode >> 12) & 15;
+}
+
+static inline pgoff_t logfs_index(struct super_block *sb, u64 pos)
+{
+	return pos >> sb->s_blocksize_bits;
+}
+
+static inline u64 dev_ofs(struct super_block *sb, u32 segno, u32 ofs)
+{
+	return ((u64)segno << logfs_super(sb)->s_segshift) + ofs;
+}
+
+static inline u32 seg_no(struct super_block *sb, u64 ofs)
+{
+	return ofs >> logfs_super(sb)->s_segshift;
+}
+
+static inline u32 seg_ofs(struct super_block *sb, u64 ofs)
+{
+	return ofs & logfs_super(sb)->s_segmask;
+}
+
+static inline u64 seg_align(struct super_block *sb, u64 ofs)
+{
+	return ofs & ~logfs_super(sb)->s_segmask;
+}
+
+static inline struct logfs_block *logfs_block(struct page *page)
+{
+	return (void *)page->private;
+}
+
+static inline level_t shrink_level(gc_level_t __level)
+{
+	u8 level = (__force u8)__level;
+
+	if (level >= LOGFS_MAX_LEVELS)
+		level -= LOGFS_MAX_LEVELS;
+	return (__force level_t)level;
+}
+
+static inline gc_level_t expand_level(u64 ino, level_t __level)
+{
+	u8 level = (__force u8)__level;
+
+	if (ino == LOGFS_INO_MASTER) {
+		/* ifile has seperate areas */
+		level += LOGFS_MAX_LEVELS;
+	}
+	return (__force gc_level_t)level;
+}
+
+static inline int logfs_block_shift(struct super_block *sb, level_t level)
+{
+	level = shrink_level((__force gc_level_t)level);
+	return (__force int)level * (sb->s_blocksize_bits - 3);
+}
+
+static inline u64 logfs_block_mask(struct super_block *sb, level_t level)
+{
+	return ~0ull << logfs_block_shift(sb, level);
+}
+
+static inline struct logfs_area *get_area(struct super_block *sb,
+		gc_level_t gc_level)
+{
+	return logfs_super(sb)->s_area[(__force u8)gc_level];
+}
+
+#endif
diff --git a/fs/logfs/logfs_abi.h b/fs/logfs/logfs_abi.h
new file mode 100644
index 000000000000..f674725663fe
--- /dev/null
+++ b/fs/logfs/logfs_abi.h
@@ -0,0 +1,629 @@
+/*
+ * fs/logfs/logfs_abi.h
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ *
+ * Public header for logfs.
+ */
+#ifndef FS_LOGFS_LOGFS_ABI_H
+#define FS_LOGFS_LOGFS_ABI_H
+
+/* For out-of-kernel compiles */
+#ifndef BUILD_BUG_ON
+#define BUILD_BUG_ON(condition) /**/
+#endif
+
+#define SIZE_CHECK(type, size)					\
+static inline void check_##type(void)				\
+{								\
+	BUILD_BUG_ON(sizeof(struct type) != (size));		\
+}
+
+/*
+ * Throughout the logfs code, we're constantly dealing with blocks at
+ * various positions or offsets.  To remove confusion, we stricly
+ * distinguish between a "position" - the logical position within a
+ * file and an "offset" - the physical location within the device.
+ *
+ * Any usage of the term offset for a logical location or position for
+ * a physical one is a bug and should get fixed.
+ */
+
+/*
+ * Block are allocated in one of several segments depending on their
+ * level.  The following levels are used:
+ *  0	- regular data block
+ *  1	- i1 indirect blocks
+ *  2	- i2 indirect blocks
+ *  3	- i3 indirect blocks
+ *  4	- i4 indirect blocks
+ *  5	- i5 indirect blocks
+ *  6	- ifile data blocks
+ *  7	- ifile i1 indirect blocks
+ *  8	- ifile i2 indirect blocks
+ *  9	- ifile i3 indirect blocks
+ * 10	- ifile i4 indirect blocks
+ * 11	- ifile i5 indirect blocks
+ * Potential levels to be used in the future:
+ * 12	- gc recycled blocks, long-lived data
+ * 13	- replacement blocks, short-lived data
+ *
+ * Levels 1-11 are necessary for robust gc operations and help seperate
+ * short-lived metadata from longer-lived file data.  In the future,
+ * file data should get seperated into several segments based on simple
+ * heuristics.  Old data recycled during gc operation is expected to be
+ * long-lived.  New data is of uncertain life expectancy.  New data
+ * used to replace older blocks in existing files is expected to be
+ * short-lived.
+ */
+
+
+/* Magic numbers.  64bit for superblock, 32bit for statfs f_type */
+#define LOGFS_MAGIC		0x7a3a8e5cb9d5bf67ull
+#define LOGFS_MAGIC_U32		0xc97e8168u
+
+/*
+ * Various blocksize related macros.  Blocksize is currently fixed at 4KiB.
+ * Sooner or later that should become configurable and the macros replaced
+ * by something superblock-dependent.  Pointers in indirect blocks are and
+ * will remain 64bit.
+ *
+ * LOGFS_BLOCKSIZE	- self-explaining
+ * LOGFS_BLOCK_FACTOR	- number of pointers per indirect block
+ * LOGFS_BLOCK_BITS	- log2 of LOGFS_BLOCK_FACTOR, used for shifts
+ */
+#define LOGFS_BLOCKSIZE		(4096ull)
+#define LOGFS_BLOCK_FACTOR	(LOGFS_BLOCKSIZE / sizeof(u64))
+#define LOGFS_BLOCK_BITS	(9)
+
+/*
+ * Number of blocks at various levels of indirection.  There are 16 direct
+ * block pointers plus a single indirect pointer.
+ */
+#define I0_BLOCKS		(16)
+#define I1_BLOCKS		LOGFS_BLOCK_FACTOR
+#define I2_BLOCKS		(LOGFS_BLOCK_FACTOR * I1_BLOCKS)
+#define I3_BLOCKS		(LOGFS_BLOCK_FACTOR * I2_BLOCKS)
+#define I4_BLOCKS		(LOGFS_BLOCK_FACTOR * I3_BLOCKS)
+#define I5_BLOCKS		(LOGFS_BLOCK_FACTOR * I4_BLOCKS)
+
+#define INDIRECT_INDEX		I0_BLOCKS
+#define LOGFS_EMBEDDED_FIELDS	(I0_BLOCKS + 1)
+
+/*
+ * Sizes at which files require another level of indirection.  Files smaller
+ * than LOGFS_EMBEDDED_SIZE can be completely stored in the inode itself,
+ * similar like ext2 fast symlinks.
+ *
+ * Data at a position smaller than LOGFS_I0_SIZE is accessed through the
+ * direct pointers, else through the 1x indirect pointer and so forth.
+ */
+#define LOGFS_EMBEDDED_SIZE	(LOGFS_EMBEDDED_FIELDS * sizeof(u64))
+#define LOGFS_I0_SIZE		(I0_BLOCKS * LOGFS_BLOCKSIZE)
+#define LOGFS_I1_SIZE		(I1_BLOCKS * LOGFS_BLOCKSIZE)
+#define LOGFS_I2_SIZE		(I2_BLOCKS * LOGFS_BLOCKSIZE)
+#define LOGFS_I3_SIZE		(I3_BLOCKS * LOGFS_BLOCKSIZE)
+#define LOGFS_I4_SIZE		(I4_BLOCKS * LOGFS_BLOCKSIZE)
+#define LOGFS_I5_SIZE		(I5_BLOCKS * LOGFS_BLOCKSIZE)
+
+/*
+ * Each indirect block pointer must have this flag set, if all block pointers
+ * behind it are set, i.e. there is no hole hidden in the shadow of this
+ * indirect block pointer.
+ */
+#define LOGFS_FULLY_POPULATED (1ULL << 63)
+#define pure_ofs(ofs) (ofs & ~LOGFS_FULLY_POPULATED)
+
+/*
+ * LogFS needs to seperate data into levels.  Each level is defined as the
+ * maximal possible distance from the master inode (inode of the inode file).
+ * Data blocks reside on level 0, 1x indirect block on level 1, etc.
+ * Inodes reside on level 6, indirect blocks for the inode file on levels 7-11.
+ * This effort is necessary to guarantee garbage collection to always make
+ * progress.
+ *
+ * LOGFS_MAX_INDIRECT is the maximal indirection through indirect blocks,
+ * LOGFS_MAX_LEVELS is one more for the actual data level of a file.  It is
+ * the maximal number of levels for one file.
+ * LOGFS_NO_AREAS is twice that, as the inode file and regular files are
+ * effectively stacked on top of each other.
+ */
+#define LOGFS_MAX_INDIRECT	(5)
+#define LOGFS_MAX_LEVELS	(LOGFS_MAX_INDIRECT + 1)
+#define LOGFS_NO_AREAS		(2 * LOGFS_MAX_LEVELS)
+
+/* Maximum size of filenames */
+#define LOGFS_MAX_NAMELEN	(255)
+
+/* Number of segments in the primary journal. */
+#define LOGFS_JOURNAL_SEGS	(16)
+
+/* Maximum number of free/erased/etc. segments in journal entries */
+#define MAX_CACHED_SEGS		(64)
+
+
+/*
+ * LOGFS_OBJECT_HEADERSIZE is the size of a single header in the object store,
+ * LOGFS_MAX_OBJECTSIZE the size of the largest possible object, including
+ * its header,
+ * LOGFS_SEGMENT_RESERVE is the amount of space reserved for each segment for
+ * its segment header and the padded space at the end when no further objects
+ * fit.
+ */
+#define LOGFS_OBJECT_HEADERSIZE	(0x1c)
+#define LOGFS_SEGMENT_HEADERSIZE (0x18)
+#define LOGFS_MAX_OBJECTSIZE	(LOGFS_OBJECT_HEADERSIZE + LOGFS_BLOCKSIZE)
+#define LOGFS_SEGMENT_RESERVE	\
+	(LOGFS_SEGMENT_HEADERSIZE + LOGFS_MAX_OBJECTSIZE - 1)
+
+/*
+ * Segment types:
+ * SEG_SUPER	- Data or indirect block
+ * SEG_JOURNAL	- Inode
+ * SEG_OSTORE	- Dentry
+ */
+enum {
+	SEG_SUPER	= 0x01,
+	SEG_JOURNAL	= 0x02,
+	SEG_OSTORE	= 0x03,
+};
+
+/**
+ * struct logfs_segment_header - per-segment header in the ostore
+ *
+ * @crc:			crc32 of header (there is no data)
+ * @pad:			unused, must be 0
+ * @type:			segment type, see above
+ * @level:			GC level for all objects in this segment
+ * @segno:			segment number
+ * @ec:				erase count for this segment
+ * @gec:			global erase count at time of writing
+ */
+struct logfs_segment_header {
+	__be32	crc;
+	__be16	pad;
+	__u8	type;
+	__u8	level;
+	__be32	segno;
+	__be32	ec;
+	__be64	gec;
+};
+
+SIZE_CHECK(logfs_segment_header, LOGFS_SEGMENT_HEADERSIZE);
+
+#define LOGFS_FEATURES_INCOMPAT		(0ull)
+#define LOGFS_FEATURES_RO_COMPAT	(0ull)
+#define LOGFS_FEATURES_COMPAT		(0ull)
+
+/**
+ * struct logfs_disk_super - on-medium superblock
+ *
+ * @ds_magic:			magic number, must equal LOGFS_MAGIC
+ * @ds_crc:			crc32 of structure starting with the next field
+ * @ds_ifile_levels:		maximum number of levels for ifile
+ * @ds_iblock_levels:		maximum number of levels for regular files
+ * @ds_data_levels:		number of seperate levels for data
+ * @pad0:			reserved, must be 0
+ * @ds_feature_incompat:	incompatible filesystem features
+ * @ds_feature_ro_compat:	read-only compatible filesystem features
+ * @ds_feature_compat:		compatible filesystem features
+ * @ds_flags:			flags
+ * @ds_segment_shift:		log2 of segment size
+ * @ds_block_shift:		log2 of block size
+ * @ds_write_shift:		log2 of write size
+ * @pad1:			reserved, must be 0
+ * @ds_journal_seg:		segments used by primary journal
+ * @ds_root_reserve:		bytes reserved for the superuser
+ * @ds_speed_reserve:		bytes reserved to speed up GC
+ * @ds_bad_seg_reserve:		number of segments reserved to handle bad blocks
+ * @pad2:			reserved, must be 0
+ * @pad3:			reserved, must be 0
+ *
+ * Contains only read-only fields.  Read-write fields like the amount of used
+ * space is tracked in the dynamic superblock, which is stored in the journal.
+ */
+struct logfs_disk_super {
+	struct logfs_segment_header ds_sh;
+	__be64	ds_magic;
+
+	__be32	ds_crc;
+	__u8	ds_ifile_levels;
+	__u8	ds_iblock_levels;
+	__u8	ds_data_levels;
+	__u8	ds_segment_shift;
+	__u8	ds_block_shift;
+	__u8	ds_write_shift;
+	__u8	pad0[6];
+
+	__be64	ds_filesystem_size;
+	__be32	ds_segment_size;
+	__be32  ds_bad_seg_reserve;
+
+	__be64	ds_feature_incompat;
+	__be64	ds_feature_ro_compat;
+
+	__be64	ds_feature_compat;
+	__be64	ds_feature_flags;
+
+	__be64	ds_root_reserve;
+	__be64  ds_speed_reserve;
+
+	__be32	ds_journal_seg[LOGFS_JOURNAL_SEGS];
+
+	__be64	ds_super_ofs[2];
+	__be64	pad3[8];
+};
+
+SIZE_CHECK(logfs_disk_super, 256);
+
+/*
+ * Object types:
+ * OBJ_BLOCK	- Data or indirect block
+ * OBJ_INODE	- Inode
+ * OBJ_DENTRY	- Dentry
+ */
+enum {
+	OBJ_BLOCK	= 0x04,
+	OBJ_INODE	= 0x05,
+	OBJ_DENTRY	= 0x06,
+};
+
+/**
+ * struct logfs_object_header - per-object header in the ostore
+ *
+ * @crc:			crc32 of header, excluding data_crc
+ * @len:			length of data
+ * @type:			object type, see above
+ * @compr:			compression type
+ * @ino:			inode number
+ * @bix:			block index
+ * @data_crc:			crc32 of payload
+ */
+struct logfs_object_header {
+	__be32	crc;
+	__be16	len;
+	__u8	type;
+	__u8	compr;
+	__be64	ino;
+	__be64	bix;
+	__be32	data_crc;
+} __attribute__((packed));
+
+SIZE_CHECK(logfs_object_header, LOGFS_OBJECT_HEADERSIZE);
+
+/*
+ * Reserved inode numbers:
+ * LOGFS_INO_MASTER	- master inode (for inode file)
+ * LOGFS_INO_ROOT	- root directory
+ * LOGFS_INO_SEGFILE	- per-segment used bytes and erase count
+ */
+enum {
+	LOGFS_INO_MAPPING	= 0x00,
+	LOGFS_INO_MASTER	= 0x01,
+	LOGFS_INO_ROOT		= 0x02,
+	LOGFS_INO_SEGFILE	= 0x03,
+	LOGFS_RESERVED_INOS	= 0x10,
+};
+
+/*
+ * Inode flags.  High bits should never be written to the medium.  They are
+ * reserved for in-memory usage.
+ * Low bits should either remain in sync with the corresponding FS_*_FL or
+ * reuse slots that obviously don't make sense for logfs.
+ *
+ * LOGFS_IF_DIRTY	Inode must be written back
+ * LOGFS_IF_ZOMBIE	Inode has been deleted
+ * LOGFS_IF_STILLBORN	-ENOSPC happened when creating inode
+ */
+#define LOGFS_IF_COMPRESSED	0x00000004 /* == FS_COMPR_FL */
+#define LOGFS_IF_DIRTY		0x20000000
+#define LOGFS_IF_ZOMBIE		0x40000000
+#define LOGFS_IF_STILLBORN	0x80000000
+
+/* Flags available to chattr */
+#define LOGFS_FL_USER_VISIBLE	(LOGFS_IF_COMPRESSED)
+#define LOGFS_FL_USER_MODIFIABLE (LOGFS_IF_COMPRESSED)
+/* Flags inherited from parent directory on file/directory creation */
+#define LOGFS_FL_INHERITED	(LOGFS_IF_COMPRESSED)
+
+/**
+ * struct logfs_disk_inode - on-medium inode
+ *
+ * @di_mode:			file mode
+ * @di_pad:			reserved, must be 0
+ * @di_flags:			inode flags, see above
+ * @di_uid:			user id
+ * @di_gid:			group id
+ * @di_ctime:			change time
+ * @di_mtime:			modify time
+ * @di_refcount:		reference count (aka nlink or link count)
+ * @di_generation:		inode generation, for nfs
+ * @di_used_bytes:		number of bytes used
+ * @di_size:			file size
+ * @di_data:			data pointers
+ */
+struct logfs_disk_inode {
+	__be16	di_mode;
+	__u8	di_height;
+	__u8	di_pad;
+	__be32	di_flags;
+	__be32	di_uid;
+	__be32	di_gid;
+
+	__be64	di_ctime;
+	__be64	di_mtime;
+
+	__be64	di_atime;
+	__be32	di_refcount;
+	__be32	di_generation;
+
+	__be64	di_used_bytes;
+	__be64	di_size;
+
+	__be64	di_data[LOGFS_EMBEDDED_FIELDS];
+};
+
+SIZE_CHECK(logfs_disk_inode, 200);
+
+#define INODE_POINTER_OFS \
+	(offsetof(struct logfs_disk_inode, di_data) / sizeof(__be64))
+#define INODE_USED_OFS \
+	(offsetof(struct logfs_disk_inode, di_used_bytes) / sizeof(__be64))
+#define INODE_SIZE_OFS \
+	(offsetof(struct logfs_disk_inode, di_size) / sizeof(__be64))
+#define INODE_HEIGHT_OFS	(0)
+
+/**
+ * struct logfs_disk_dentry - on-medium dentry structure
+ *
+ * @ino:			inode number
+ * @namelen:			length of file name
+ * @type:			file type, identical to bits 12..15 of mode
+ * @name:			file name
+ */
+/* FIXME: add 6 bytes of padding to remove the __packed */
+struct logfs_disk_dentry {
+	__be64	ino;
+	__be16	namelen;
+	__u8	type;
+	__u8	name[LOGFS_MAX_NAMELEN];
+} __attribute__((packed));
+
+SIZE_CHECK(logfs_disk_dentry, 266);
+
+#define RESERVED		0xffffffff
+#define BADSEG			0xffffffff
+/**
+ * struct logfs_segment_entry - segment file entry
+ *
+ * @ec_level:			erase count and level
+ * @valid:			number of valid bytes
+ *
+ * Segment file contains one entry for every segment.  ec_level contains the
+ * erasecount in the upper 28 bits and the level in the lower 4 bits.  An
+ * ec_level of BADSEG (-1) identifies bad segments.  valid contains the number
+ * of valid bytes or RESERVED (-1 again) if the segment is used for either the
+ * superblock or the journal, or when the segment is bad.
+ */
+struct logfs_segment_entry {
+	__be32	ec_level;
+	__be32	valid;
+};
+
+SIZE_CHECK(logfs_segment_entry, 8);
+
+/**
+ * struct logfs_journal_header - header for journal entries (JEs)
+ *
+ * @h_crc:			crc32 of journal entry
+ * @h_len:			length of compressed journal entry,
+ *				not including header
+ * @h_datalen:			length of uncompressed data
+ * @h_type:			JE type
+ * @h_compr:			compression type
+ * @h_pad:			reserved
+ */
+struct logfs_journal_header {
+	__be32	h_crc;
+	__be16	h_len;
+	__be16	h_datalen;
+	__be16	h_type;
+	__u8	h_compr;
+	__u8	h_pad[5];
+};
+
+SIZE_CHECK(logfs_journal_header, 16);
+
+/*
+ * Life expectency of data.
+ * VIM_DEFAULT		- default vim
+ * VIM_SEGFILE		- for segment file only - very short-living
+ * VIM_GC		- GC'd data - likely long-living
+ */
+enum logfs_vim {
+	VIM_DEFAULT	= 0,
+	VIM_SEGFILE	= 1,
+};
+
+/**
+ * struct logfs_je_area - wbuf header
+ *
+ * @segno:			segment number of area
+ * @used_bytes:			number of bytes already used
+ * @gc_level:			GC level
+ * @vim:			life expectancy of data
+ *
+ * "Areas" are segments currently being used for writing.  There is at least
+ * one area per GC level.  Several may be used to seperate long-living from
+ * short-living data.  If an area with unknown vim is encountered, it can
+ * simply be closed.
+ * The write buffer immediately follow this header.
+ */
+struct logfs_je_area {
+	__be32	segno;
+	__be32	used_bytes;
+	__u8	gc_level;
+	__u8	vim;
+} __attribute__((packed));
+
+SIZE_CHECK(logfs_je_area, 10);
+
+#define MAX_JOURNAL_HEADER \
+	(sizeof(struct logfs_journal_header) + sizeof(struct logfs_je_area))
+
+/**
+ * struct logfs_je_dynsb - dynamic superblock
+ *
+ * @ds_gec:			global erase count
+ * @ds_sweeper:			current position of GC "sweeper"
+ * @ds_rename_dir:		source directory ino (see dir.c documentation)
+ * @ds_rename_pos:		position of source dd (see dir.c documentation)
+ * @ds_victim_ino:		victims of incomplete dir operation (see dir.c)
+ * @ds_victim_ino:		parent inode of victim (see dir.c)
+ * @ds_used_bytes:		number of used bytes
+ */
+struct logfs_je_dynsb {
+	__be64	ds_gec;
+	__be64	ds_sweeper;
+
+	__be64	ds_rename_dir;
+	__be64	ds_rename_pos;
+
+	__be64	ds_victim_ino;
+	__be64	ds_victim_parent; /* XXX */
+
+	__be64	ds_used_bytes;
+	__be32	ds_generation;
+	__be32	pad;
+};
+
+SIZE_CHECK(logfs_je_dynsb, 64);
+
+/**
+ * struct logfs_je_anchor - anchor of filesystem tree, aka master inode
+ *
+ * @da_size:			size of inode file
+ * @da_last_ino:		last created inode
+ * @da_used_bytes:		number of bytes used
+ * @da_data:			data pointers
+ */
+struct logfs_je_anchor {
+	__be64	da_size;
+	__be64	da_last_ino;
+
+	__be64	da_used_bytes;
+	u8	da_height;
+	u8	pad[7];
+
+	__be64	da_data[LOGFS_EMBEDDED_FIELDS];
+};
+
+SIZE_CHECK(logfs_je_anchor, 168);
+
+/**
+ * struct logfs_je_spillout - spillout entry (from 1st to 2nd journal)
+ *
+ * @so_segment:			segments used for 2nd journal
+ *
+ * Length of the array is given by h_len field in the header.
+ */
+struct logfs_je_spillout {
+	__be64	so_segment[0];
+};
+
+SIZE_CHECK(logfs_je_spillout, 0);
+
+/**
+ * struct logfs_je_journal_ec - erase counts for all journal segments
+ *
+ * @ec:				erase count
+ *
+ * Length of the array is given by h_len field in the header.
+ */
+struct logfs_je_journal_ec {
+	__be32	ec[0];
+};
+
+SIZE_CHECK(logfs_je_journal_ec, 0);
+
+/**
+ * struct logfs_je_free_segments - list of free segmetns with erase count
+ */
+struct logfs_je_free_segments {
+	__be32	segno;
+	__be32	ec;
+};
+
+SIZE_CHECK(logfs_je_free_segments, 8);
+
+/**
+ * struct logfs_seg_alias - list of segment aliases
+ */
+struct logfs_seg_alias {
+	__be32	old_segno;
+	__be32	new_segno;
+};
+
+SIZE_CHECK(logfs_seg_alias, 8);
+
+/**
+ * struct logfs_obj_alias - list of object aliases
+ */
+struct logfs_obj_alias {
+	__be64	ino;
+	__be64	bix;
+	__be64	val;
+	u8	level;
+	u8	pad[5];
+	__be16	child_no;
+};
+
+SIZE_CHECK(logfs_obj_alias, 32);
+
+/**
+ * Compression types.
+ *
+ * COMPR_NONE	- uncompressed
+ * COMPR_ZLIB	- compressed with zlib
+ */
+enum {
+	COMPR_NONE	= 0,
+	COMPR_ZLIB	= 1,
+};
+
+/*
+ * Journal entries come in groups of 16.  First group contains unique
+ * entries, next groups contain one entry per level
+ *
+ * JE_FIRST	- smallest possible journal entry number
+ *
+ * JEG_BASE	- base group, containing unique entries
+ * JE_COMMIT	- commit entry, validates all previous entries
+ * JE_DYNSB	- dynamic superblock, anything that ought to be in the
+ *		  superblock but cannot because it is read-write data
+ * JE_ANCHOR	- anchor aka master inode aka inode file's inode
+ * JE_ERASECOUNT  erasecounts for all journal segments
+ * JE_SPILLOUT	- unused
+ * JE_SEG_ALIAS	- aliases segments
+ * JE_AREA	- area description
+ *
+ * JE_LAST	- largest possible journal entry number
+ */
+enum {
+	JE_FIRST	= 0x01,
+
+	JEG_BASE	= 0x00,
+	JE_COMMIT	= 0x02,
+	JE_DYNSB	= 0x03,
+	JE_ANCHOR	= 0x04,
+	JE_ERASECOUNT	= 0x05,
+	JE_SPILLOUT	= 0x06,
+	JE_OBJ_ALIAS	= 0x0d,
+	JE_AREA		= 0x0e,
+
+	JE_LAST		= 0x0e,
+};
+
+#endif
diff --git a/fs/logfs/readwrite.c b/fs/logfs/readwrite.c
new file mode 100644
index 000000000000..7a23b3e7c0a7
--- /dev/null
+++ b/fs/logfs/readwrite.c
@@ -0,0 +1,2246 @@
+/*
+ * fs/logfs/readwrite.c
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ *
+ *
+ * Actually contains five sets of very similar functions:
+ * read		read blocks from a file
+ * seek_hole	find next hole
+ * seek_data	find next data block
+ * valid	check whether a block still belongs to a file
+ * write	write blocks to a file
+ * delete	delete a block (for directories and ifile)
+ * rewrite	move existing blocks of a file to a new location (gc helper)
+ * truncate	truncate a file
+ */
+#include "logfs.h"
+#include <linux/sched.h>
+
+static u64 adjust_bix(u64 bix, level_t level)
+{
+	switch (level) {
+	case 0:
+		return bix;
+	case LEVEL(1):
+		return max_t(u64, bix, I0_BLOCKS);
+	case LEVEL(2):
+		return max_t(u64, bix, I1_BLOCKS);
+	case LEVEL(3):
+		return max_t(u64, bix, I2_BLOCKS);
+	case LEVEL(4):
+		return max_t(u64, bix, I3_BLOCKS);
+	case LEVEL(5):
+		return max_t(u64, bix, I4_BLOCKS);
+	default:
+		WARN_ON(1);
+		return bix;
+	}
+}
+
+static inline u64 maxbix(u8 height)
+{
+	return 1ULL << (LOGFS_BLOCK_BITS * height);
+}
+
+/**
+ * The inode address space is cut in two halves.  Lower half belongs to data
+ * pages, upper half to indirect blocks.  If the high bit (INDIRECT_BIT) is
+ * set, the actual block index (bix) and level can be derived from the page
+ * index.
+ *
+ * The lowest three bits of the block index are set to 0 after packing and
+ * unpacking.  Since the lowest n bits (9 for 4KiB blocksize) are ignored
+ * anyway this is harmless.
+ */
+#define ARCH_SHIFT	(BITS_PER_LONG - 32)
+#define INDIRECT_BIT	(0x80000000UL << ARCH_SHIFT)
+#define LEVEL_SHIFT	(28 + ARCH_SHIFT)
+static inline pgoff_t first_indirect_block(void)
+{
+	return INDIRECT_BIT | (1ULL << LEVEL_SHIFT);
+}
+
+pgoff_t logfs_pack_index(u64 bix, level_t level)
+{
+	pgoff_t index;
+
+	BUG_ON(bix >= INDIRECT_BIT);
+	if (level == 0)
+		return bix;
+
+	index  = INDIRECT_BIT;
+	index |= (__force long)level << LEVEL_SHIFT;
+	index |= bix >> ((__force u8)level * LOGFS_BLOCK_BITS);
+	return index;
+}
+
+void logfs_unpack_index(pgoff_t index, u64 *bix, level_t *level)
+{
+	u8 __level;
+
+	if (!(index & INDIRECT_BIT)) {
+		*bix = index;
+		*level = 0;
+		return;
+	}
+
+	__level = (index & ~INDIRECT_BIT) >> LEVEL_SHIFT;
+	*level = LEVEL(__level);
+	*bix = (index << (__level * LOGFS_BLOCK_BITS)) & ~INDIRECT_BIT;
+	*bix = adjust_bix(*bix, *level);
+	return;
+}
+#undef ARCH_SHIFT
+#undef INDIRECT_BIT
+#undef LEVEL_SHIFT
+
+/*
+ * Time is stored as nanoseconds since the epoch.
+ */
+static struct timespec be64_to_timespec(__be64 betime)
+{
+	return ns_to_timespec(be64_to_cpu(betime));
+}
+
+static __be64 timespec_to_be64(struct timespec tsp)
+{
+	return cpu_to_be64((u64)tsp.tv_sec * NSEC_PER_SEC + tsp.tv_nsec);
+}
+
+static void logfs_disk_to_inode(struct logfs_disk_inode *di, struct inode*inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	int i;
+
+	inode->i_mode	= be16_to_cpu(di->di_mode);
+	li->li_height	= di->di_height;
+	li->li_flags	= be32_to_cpu(di->di_flags);
+	inode->i_uid	= be32_to_cpu(di->di_uid);
+	inode->i_gid	= be32_to_cpu(di->di_gid);
+	inode->i_size	= be64_to_cpu(di->di_size);
+	logfs_set_blocks(inode, be64_to_cpu(di->di_used_bytes));
+	inode->i_atime	= be64_to_timespec(di->di_atime);
+	inode->i_ctime	= be64_to_timespec(di->di_ctime);
+	inode->i_mtime	= be64_to_timespec(di->di_mtime);
+	inode->i_nlink	= be32_to_cpu(di->di_refcount);
+	inode->i_generation = be32_to_cpu(di->di_generation);
+
+	switch (inode->i_mode & S_IFMT) {
+	case S_IFSOCK:	/* fall through */
+	case S_IFBLK:	/* fall through */
+	case S_IFCHR:	/* fall through */
+	case S_IFIFO:
+		inode->i_rdev = be64_to_cpu(di->di_data[0]);
+		break;
+	case S_IFDIR:	/* fall through */
+	case S_IFREG:	/* fall through */
+	case S_IFLNK:
+		for (i = 0; i < LOGFS_EMBEDDED_FIELDS; i++)
+			li->li_data[i] = be64_to_cpu(di->di_data[i]);
+		break;
+	default:
+		BUG();
+	}
+}
+
+static void logfs_inode_to_disk(struct inode *inode, struct logfs_disk_inode*di)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	int i;
+
+	di->di_mode	= cpu_to_be16(inode->i_mode);
+	di->di_height	= li->li_height;
+	di->di_pad	= 0;
+	di->di_flags	= cpu_to_be32(li->li_flags);
+	di->di_uid	= cpu_to_be32(inode->i_uid);
+	di->di_gid	= cpu_to_be32(inode->i_gid);
+	di->di_size	= cpu_to_be64(i_size_read(inode));
+	di->di_used_bytes = cpu_to_be64(li->li_used_bytes);
+	di->di_atime	= timespec_to_be64(inode->i_atime);
+	di->di_ctime	= timespec_to_be64(inode->i_ctime);
+	di->di_mtime	= timespec_to_be64(inode->i_mtime);
+	di->di_refcount	= cpu_to_be32(inode->i_nlink);
+	di->di_generation = cpu_to_be32(inode->i_generation);
+
+	switch (inode->i_mode & S_IFMT) {
+	case S_IFSOCK:	/* fall through */
+	case S_IFBLK:	/* fall through */
+	case S_IFCHR:	/* fall through */
+	case S_IFIFO:
+		di->di_data[0] = cpu_to_be64(inode->i_rdev);
+		break;
+	case S_IFDIR:	/* fall through */
+	case S_IFREG:	/* fall through */
+	case S_IFLNK:
+		for (i = 0; i < LOGFS_EMBEDDED_FIELDS; i++)
+			di->di_data[i] = cpu_to_be64(li->li_data[i]);
+		break;
+	default:
+		BUG();
+	}
+}
+
+static void __logfs_set_blocks(struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+	struct logfs_inode *li = logfs_inode(inode);
+
+	inode->i_blocks = ULONG_MAX;
+	if (li->li_used_bytes >> sb->s_blocksize_bits < ULONG_MAX)
+		inode->i_blocks = ALIGN(li->li_used_bytes, 512) >> 9;
+}
+
+void logfs_set_blocks(struct inode *inode, u64 bytes)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	li->li_used_bytes = bytes;
+	__logfs_set_blocks(inode);
+}
+
+static void prelock_page(struct super_block *sb, struct page *page, int lock)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	BUG_ON(!PageLocked(page));
+	if (lock) {
+		BUG_ON(PagePreLocked(page));
+		SetPagePreLocked(page);
+	} else {
+		/* We are in GC path. */
+		if (PagePreLocked(page))
+			super->s_lock_count++;
+		else
+			SetPagePreLocked(page);
+	}
+}
+
+static void preunlock_page(struct super_block *sb, struct page *page, int lock)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	BUG_ON(!PageLocked(page));
+	if (lock)
+		ClearPagePreLocked(page);
+	else {
+		/* We are in GC path. */
+		BUG_ON(!PagePreLocked(page));
+		if (super->s_lock_count)
+			super->s_lock_count--;
+		else
+			ClearPagePreLocked(page);
+	}
+}
+
+/*
+ * Logfs is prone to an AB-BA deadlock where one task tries to acquire
+ * s_write_mutex with a locked page and GC tries to get that page while holding
+ * s_write_mutex.
+ * To solve this issue logfs will ignore the page lock iff the page in question
+ * is waiting for s_write_mutex.  We annotate this fact by setting PG_pre_locked
+ * in addition to PG_locked.
+ */
+static void logfs_get_wblocks(struct super_block *sb, struct page *page,
+		int lock)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	if (page)
+		prelock_page(sb, page, lock);
+
+	if (lock) {
+		mutex_lock(&super->s_write_mutex);
+		logfs_gc_pass(sb);
+		/* FIXME: We also have to check for shadowed space
+		 * and mempool fill grade */
+	}
+}
+
+static void logfs_put_wblocks(struct super_block *sb, struct page *page,
+		int lock)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	if (page)
+		preunlock_page(sb, page, lock);
+	/* Order matters - we must clear PG_pre_locked before releasing
+	 * s_write_mutex or we could race against another task. */
+	if (lock)
+		mutex_unlock(&super->s_write_mutex);
+}
+
+static struct page *logfs_get_read_page(struct inode *inode, u64 bix,
+		level_t level)
+{
+	return find_or_create_page(inode->i_mapping,
+			logfs_pack_index(bix, level), GFP_NOFS);
+}
+
+static void logfs_put_read_page(struct page *page)
+{
+	unlock_page(page);
+	page_cache_release(page);
+}
+
+static void logfs_lock_write_page(struct page *page)
+{
+	int loop = 0;
+
+	while (unlikely(!trylock_page(page))) {
+		if (loop++ > 0x1000) {
+			/* Has been observed once so far... */
+			printk(KERN_ERR "stack at %p\n", &loop);
+			BUG();
+		}
+		if (PagePreLocked(page)) {
+			/* Holder of page lock is waiting for us, it
+			 * is safe to use this page. */
+			break;
+		}
+		/* Some other process has this page locked and has
+		 * nothing to do with us.  Wait for it to finish.
+		 */
+		schedule();
+	}
+	BUG_ON(!PageLocked(page));
+}
+
+static struct page *logfs_get_write_page(struct inode *inode, u64 bix,
+		level_t level)
+{
+	struct address_space *mapping = inode->i_mapping;
+	pgoff_t index = logfs_pack_index(bix, level);
+	struct page *page;
+	int err;
+
+repeat:
+	page = find_get_page(mapping, index);
+	if (!page) {
+		page = __page_cache_alloc(GFP_NOFS);
+		if (!page)
+			return NULL;
+		err = add_to_page_cache_lru(page, mapping, index, GFP_NOFS);
+		if (unlikely(err)) {
+			page_cache_release(page);
+			if (err == -EEXIST)
+				goto repeat;
+			return NULL;
+		}
+	} else logfs_lock_write_page(page);
+	BUG_ON(!PageLocked(page));
+	return page;
+}
+
+static void logfs_unlock_write_page(struct page *page)
+{
+	if (!PagePreLocked(page))
+		unlock_page(page);
+}
+
+static void logfs_put_write_page(struct page *page)
+{
+	logfs_unlock_write_page(page);
+	page_cache_release(page);
+}
+
+static struct page *logfs_get_page(struct inode *inode, u64 bix, level_t level,
+		int rw)
+{
+	if (rw == READ)
+		return logfs_get_read_page(inode, bix, level);
+	else
+		return logfs_get_write_page(inode, bix, level);
+}
+
+static void logfs_put_page(struct page *page, int rw)
+{
+	if (rw == READ)
+		logfs_put_read_page(page);
+	else
+		logfs_put_write_page(page);
+}
+
+static unsigned long __get_bits(u64 val, int skip, int no)
+{
+	u64 ret = val;
+
+	ret >>= skip * no;
+	ret <<= 64 - no;
+	ret >>= 64 - no;
+	return ret;
+}
+
+static unsigned long get_bits(u64 val, level_t skip)
+{
+	return __get_bits(val, (__force int)skip, LOGFS_BLOCK_BITS);
+}
+
+static inline void init_shadow_tree(struct super_block *sb,
+		struct shadow_tree *tree)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	btree_init_mempool64(&tree->new, super->s_btree_pool);
+	btree_init_mempool64(&tree->old, super->s_btree_pool);
+}
+
+static void indirect_write_block(struct logfs_block *block)
+{
+	struct page *page;
+	struct inode *inode;
+	int ret;
+
+	page = block->page;
+	inode = page->mapping->host;
+	logfs_lock_write_page(page);
+	ret = logfs_write_buf(inode, page, 0);
+	logfs_unlock_write_page(page);
+	/*
+	 * This needs some rework.  Unless you want your filesystem to run
+	 * completely synchronously (you don't), the filesystem will always
+	 * report writes as 'successful' before the actual work has been
+	 * done.  The actual work gets done here and this is where any errors
+	 * will show up.  And there isn't much we can do about it, really.
+	 *
+	 * Some attempts to fix the errors (move from bad blocks, retry io,...)
+	 * have already been done, so anything left should be either a broken
+	 * device or a bug somewhere in logfs itself.  Being relatively new,
+	 * the odds currently favor a bug, so for now the line below isn't
+	 * entirely tasteles.
+	 */
+	BUG_ON(ret);
+}
+
+static void inode_write_block(struct logfs_block *block)
+{
+	struct inode *inode;
+	int ret;
+
+	inode = block->inode;
+	if (inode->i_ino == LOGFS_INO_MASTER)
+		logfs_write_anchor(inode->i_sb);
+	else {
+		ret = __logfs_write_inode(inode, 0);
+		/* see indirect_write_block comment */
+		BUG_ON(ret);
+	}
+}
+
+static gc_level_t inode_block_level(struct logfs_block *block)
+{
+	BUG_ON(block->inode->i_ino == LOGFS_INO_MASTER);
+	return GC_LEVEL(LOGFS_MAX_LEVELS);
+}
+
+static gc_level_t indirect_block_level(struct logfs_block *block)
+{
+	struct page *page;
+	struct inode *inode;
+	u64 bix;
+	level_t level;
+
+	page = block->page;
+	inode = page->mapping->host;
+	logfs_unpack_index(page->index, &bix, &level);
+	return expand_level(inode->i_ino, level);
+}
+
+/*
+ * This silences a false, yet annoying gcc warning.  I hate it when my editor
+ * jumps into bitops.h each time I recompile this file.
+ * TODO: Complain to gcc folks about this and upgrade compiler.
+ */
+static unsigned long fnb(const unsigned long *addr,
+		unsigned long size, unsigned long offset)
+{
+	return find_next_bit(addr, size, offset);
+}
+
+static __be64 inode_val0(struct inode *inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	u64 val;
+
+	/*
+	 * Explicit shifting generates good code, but must match the format
+	 * of the structure.  Add some paranoia just in case.
+	 */
+	BUILD_BUG_ON(offsetof(struct logfs_disk_inode, di_mode) != 0);
+	BUILD_BUG_ON(offsetof(struct logfs_disk_inode, di_height) != 2);
+	BUILD_BUG_ON(offsetof(struct logfs_disk_inode, di_flags) != 4);
+
+	val =	(u64)inode->i_mode << 48 |
+		(u64)li->li_height << 40 |
+		(u64)li->li_flags;
+	return cpu_to_be64(val);
+}
+
+static int inode_write_alias(struct super_block *sb,
+		struct logfs_block *block, write_alias_t *write_one_alias)
+{
+	struct inode *inode = block->inode;
+	struct logfs_inode *li = logfs_inode(inode);
+	unsigned long pos;
+	u64 ino , bix;
+	__be64 val;
+	level_t level;
+	int err;
+
+	for (pos = 0; ; pos++) {
+		pos = fnb(block->alias_map, LOGFS_BLOCK_FACTOR, pos);
+		if (pos >= LOGFS_EMBEDDED_FIELDS + INODE_POINTER_OFS)
+			return 0;
+
+		switch (pos) {
+		case INODE_HEIGHT_OFS:
+			val = inode_val0(inode);
+			break;
+		case INODE_USED_OFS:
+			val = cpu_to_be64(li->li_used_bytes);;
+			break;
+		case INODE_SIZE_OFS:
+			val = cpu_to_be64(i_size_read(inode));
+			break;
+		case INODE_POINTER_OFS ... INODE_POINTER_OFS + LOGFS_EMBEDDED_FIELDS - 1:
+			val = cpu_to_be64(li->li_data[pos - INODE_POINTER_OFS]);
+			break;
+		default:
+			BUG();
+		}
+
+		ino = LOGFS_INO_MASTER;
+		bix = inode->i_ino;
+		level = LEVEL(0);
+		err = write_one_alias(sb, ino, bix, level, pos, val);
+		if (err)
+			return err;
+	}
+}
+
+static int indirect_write_alias(struct super_block *sb,
+		struct logfs_block *block, write_alias_t *write_one_alias)
+{
+	unsigned long pos;
+	struct page *page = block->page;
+	u64 ino , bix;
+	__be64 *child, val;
+	level_t level;
+	int err;
+
+	for (pos = 0; ; pos++) {
+		pos = fnb(block->alias_map, LOGFS_BLOCK_FACTOR, pos);
+		if (pos >= LOGFS_BLOCK_FACTOR)
+			return 0;
+
+		ino = page->mapping->host->i_ino;
+		logfs_unpack_index(page->index, &bix, &level);
+		child = kmap_atomic(page, KM_USER0);
+		val = child[pos];
+		kunmap_atomic(child, KM_USER0);
+		err = write_one_alias(sb, ino, bix, level, pos, val);
+		if (err)
+			return err;
+	}
+}
+
+int logfs_write_obj_aliases_pagecache(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_block *block;
+	int err;
+
+	list_for_each_entry(block, &super->s_object_alias, alias_list) {
+		err = block->ops->write_alias(sb, block, write_alias_journal);
+		if (err)
+			return err;
+	}
+	return 0;
+}
+
+void __free_block(struct super_block *sb, struct logfs_block *block)
+{
+	BUG_ON(!list_empty(&block->item_list));
+	list_del(&block->alias_list);
+	mempool_free(block, logfs_super(sb)->s_block_pool);
+}
+
+static void inode_free_block(struct super_block *sb, struct logfs_block *block)
+{
+	struct inode *inode = block->inode;
+
+	logfs_inode(inode)->li_block = NULL;
+	__free_block(sb, block);
+}
+
+static void indirect_free_block(struct super_block *sb,
+		struct logfs_block *block)
+{
+	ClearPagePrivate(block->page);
+	block->page->private = 0;
+	__free_block(sb, block);
+}
+
+
+static struct logfs_block_ops inode_block_ops = {
+	.write_block = inode_write_block,
+	.block_level = inode_block_level,
+	.free_block = inode_free_block,
+	.write_alias = inode_write_alias,
+};
+
+struct logfs_block_ops indirect_block_ops = {
+	.write_block = indirect_write_block,
+	.block_level = indirect_block_level,
+	.free_block = indirect_free_block,
+	.write_alias = indirect_write_alias,
+};
+
+struct logfs_block *__alloc_block(struct super_block *sb,
+		u64 ino, u64 bix, level_t level)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_block *block;
+
+	block = mempool_alloc(super->s_block_pool, GFP_NOFS);
+	memset(block, 0, sizeof(*block));
+	INIT_LIST_HEAD(&block->alias_list);
+	INIT_LIST_HEAD(&block->item_list);
+	block->sb = sb;
+	block->ino = ino;
+	block->bix = bix;
+	block->level = level;
+	return block;
+}
+
+static void alloc_inode_block(struct inode *inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	struct logfs_block *block;
+
+	if (li->li_block)
+		return;
+
+	block = __alloc_block(inode->i_sb, LOGFS_INO_MASTER, inode->i_ino, 0);
+	block->inode = inode;
+	li->li_block = block;
+	block->ops = &inode_block_ops;
+}
+
+void initialize_block_counters(struct page *page, struct logfs_block *block,
+		__be64 *array, int page_is_empty)
+{
+	u64 ptr;
+	int i, start;
+
+	block->partial = 0;
+	block->full = 0;
+	start = 0;
+	if (page->index < first_indirect_block()) {
+		/* Counters are pointless on level 0 */
+		return;
+	}
+	if (page->index == first_indirect_block()) {
+		/* Skip unused pointers */
+		start = I0_BLOCKS;
+		block->full = I0_BLOCKS;
+	}
+	if (!page_is_empty) {
+		for (i = start; i < LOGFS_BLOCK_FACTOR; i++) {
+			ptr = be64_to_cpu(array[i]);
+			if (ptr)
+				block->partial++;
+			if (ptr & LOGFS_FULLY_POPULATED)
+				block->full++;
+		}
+	}
+}
+
+static void alloc_data_block(struct inode *inode, struct page *page)
+{
+	struct logfs_block *block;
+	u64 bix;
+	level_t level;
+
+	if (PagePrivate(page))
+		return;
+
+	logfs_unpack_index(page->index, &bix, &level);
+	block = __alloc_block(inode->i_sb, inode->i_ino, bix, level);
+	block->page = page;
+	SetPagePrivate(page);
+	page->private = (unsigned long)block;
+	block->ops = &indirect_block_ops;
+}
+
+static void alloc_indirect_block(struct inode *inode, struct page *page,
+		int page_is_empty)
+{
+	struct logfs_block *block;
+	__be64 *array;
+
+	if (PagePrivate(page))
+		return;
+
+	alloc_data_block(inode, page);
+
+	block = logfs_block(page);
+	array = kmap_atomic(page, KM_USER0);
+	initialize_block_counters(page, block, array, page_is_empty);
+	kunmap_atomic(array, KM_USER0);
+}
+
+static void block_set_pointer(struct page *page, int index, u64 ptr)
+{
+	struct logfs_block *block = logfs_block(page);
+	__be64 *array;
+	u64 oldptr;
+
+	BUG_ON(!block);
+	array = kmap_atomic(page, KM_USER0);
+	oldptr = be64_to_cpu(array[index]);
+	array[index] = cpu_to_be64(ptr);
+	kunmap_atomic(array, KM_USER0);
+	SetPageUptodate(page);
+
+	block->full += !!(ptr & LOGFS_FULLY_POPULATED)
+		- !!(oldptr & LOGFS_FULLY_POPULATED);
+	block->partial += !!ptr - !!oldptr;
+}
+
+static u64 block_get_pointer(struct page *page, int index)
+{
+	__be64 *block;
+	u64 ptr;
+
+	block = kmap_atomic(page, KM_USER0);
+	ptr = be64_to_cpu(block[index]);
+	kunmap_atomic(block, KM_USER0);
+	return ptr;
+}
+
+static int logfs_read_empty(struct page *page)
+{
+	zero_user_segment(page, 0, PAGE_CACHE_SIZE);
+	return 0;
+}
+
+static int logfs_read_direct(struct inode *inode, struct page *page)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	pgoff_t index = page->index;
+	u64 block;
+
+	block = li->li_data[index];
+	if (!block)
+		return logfs_read_empty(page);
+
+	return logfs_segment_read(inode, page, block, index, 0);
+}
+
+static int logfs_read_loop(struct inode *inode, struct page *page,
+		int rw_context)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	u64 bix, bofs = li->li_data[INDIRECT_INDEX];
+	level_t level, target_level;
+	int ret;
+	struct page *ipage;
+
+	logfs_unpack_index(page->index, &bix, &target_level);
+	if (!bofs)
+		return logfs_read_empty(page);
+
+	if (bix >= maxbix(li->li_height))
+		return logfs_read_empty(page);
+
+	for (level = LEVEL(li->li_height);
+			(__force u8)level > (__force u8)target_level;
+			level = SUBLEVEL(level)){
+		ipage = logfs_get_page(inode, bix, level, rw_context);
+		if (!ipage)
+			return -ENOMEM;
+
+		ret = logfs_segment_read(inode, ipage, bofs, bix, level);
+		if (ret) {
+			logfs_put_read_page(ipage);
+			return ret;
+		}
+
+		bofs = block_get_pointer(ipage, get_bits(bix, SUBLEVEL(level)));
+		logfs_put_page(ipage, rw_context);
+		if (!bofs)
+			return logfs_read_empty(page);
+	}
+
+	return logfs_segment_read(inode, page, bofs, bix, 0);
+}
+
+static int logfs_read_block(struct inode *inode, struct page *page,
+		int rw_context)
+{
+	pgoff_t index = page->index;
+
+	if (index < I0_BLOCKS)
+		return logfs_read_direct(inode, page);
+	return logfs_read_loop(inode, page, rw_context);
+}
+
+static int logfs_exist_loop(struct inode *inode, u64 bix)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	u64 bofs = li->li_data[INDIRECT_INDEX];
+	level_t level;
+	int ret;
+	struct page *ipage;
+
+	if (!bofs)
+		return 0;
+	if (bix >= maxbix(li->li_height))
+		return 0;
+
+	for (level = LEVEL(li->li_height); level != 0; level = SUBLEVEL(level)) {
+		ipage = logfs_get_read_page(inode, bix, level);
+		if (!ipage)
+			return -ENOMEM;
+
+		ret = logfs_segment_read(inode, ipage, bofs, bix, level);
+		if (ret) {
+			logfs_put_read_page(ipage);
+			return ret;
+		}
+
+		bofs = block_get_pointer(ipage, get_bits(bix, SUBLEVEL(level)));
+		logfs_put_read_page(ipage);
+		if (!bofs)
+			return 0;
+	}
+
+	return 1;
+}
+
+int logfs_exist_block(struct inode *inode, u64 bix)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	if (bix < I0_BLOCKS)
+		return !!li->li_data[bix];
+	return logfs_exist_loop(inode, bix);
+}
+
+static u64 seek_holedata_direct(struct inode *inode, u64 bix, int data)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	for (; bix < I0_BLOCKS; bix++)
+		if (data ^ (li->li_data[bix] == 0))
+			return bix;
+	return I0_BLOCKS;
+}
+
+static u64 seek_holedata_loop(struct inode *inode, u64 bix, int data)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	__be64 *rblock;
+	u64 increment, bofs = li->li_data[INDIRECT_INDEX];
+	level_t level;
+	int ret, slot;
+	struct page *page;
+
+	BUG_ON(!bofs);
+
+	for (level = LEVEL(li->li_height); level != 0; level = SUBLEVEL(level)) {
+		increment = 1 << (LOGFS_BLOCK_BITS * ((__force u8)level-1));
+		page = logfs_get_read_page(inode, bix, level);
+		if (!page)
+			return bix;
+
+		ret = logfs_segment_read(inode, page, bofs, bix, level);
+		if (ret) {
+			logfs_put_read_page(page);
+			return bix;
+		}
+
+		slot = get_bits(bix, SUBLEVEL(level));
+		rblock = kmap_atomic(page, KM_USER0);
+		while (slot < LOGFS_BLOCK_FACTOR) {
+			if (data && (rblock[slot] != 0))
+				break;
+			if (!data && !(be64_to_cpu(rblock[slot]) & LOGFS_FULLY_POPULATED))
+				break;
+			slot++;
+			bix += increment;
+			bix &= ~(increment - 1);
+		}
+		if (slot >= LOGFS_BLOCK_FACTOR) {
+			kunmap_atomic(rblock, KM_USER0);
+			logfs_put_read_page(page);
+			return bix;
+		}
+		bofs = be64_to_cpu(rblock[slot]);
+		kunmap_atomic(rblock, KM_USER0);
+		logfs_put_read_page(page);
+		if (!bofs) {
+			BUG_ON(data);
+			return bix;
+		}
+	}
+	return bix;
+}
+
+/**
+ * logfs_seek_hole - find next hole starting at a given block index
+ * @inode:		inode to search in
+ * @bix:		block index to start searching
+ *
+ * Returns next hole.  If the file doesn't contain any further holes, the
+ * block address next to eof is returned instead.
+ */
+u64 logfs_seek_hole(struct inode *inode, u64 bix)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	if (bix < I0_BLOCKS) {
+		bix = seek_holedata_direct(inode, bix, 0);
+		if (bix < I0_BLOCKS)
+			return bix;
+	}
+
+	if (!li->li_data[INDIRECT_INDEX])
+		return bix;
+	else if (li->li_data[INDIRECT_INDEX] & LOGFS_FULLY_POPULATED)
+		bix = maxbix(li->li_height);
+	else {
+		bix = seek_holedata_loop(inode, bix, 0);
+		if (bix < maxbix(li->li_height))
+			return bix;
+		/* Should not happen anymore.  But if some port writes semi-
+		 * corrupt images (as this one used to) we might run into it.
+		 */
+		WARN_ON_ONCE(bix == maxbix(li->li_height));
+	}
+
+	return bix;
+}
+
+static u64 __logfs_seek_data(struct inode *inode, u64 bix)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	if (bix < I0_BLOCKS) {
+		bix = seek_holedata_direct(inode, bix, 1);
+		if (bix < I0_BLOCKS)
+			return bix;
+	}
+
+	if (bix < maxbix(li->li_height)) {
+		if (!li->li_data[INDIRECT_INDEX])
+			bix = maxbix(li->li_height);
+		else
+			return seek_holedata_loop(inode, bix, 1);
+	}
+
+	return bix;
+}
+
+/**
+ * logfs_seek_data - find next data block after a given block index
+ * @inode:		inode to search in
+ * @bix:		block index to start searching
+ *
+ * Returns next data block.  If the file doesn't contain any further data
+ * blocks, the last block in the file is returned instead.
+ */
+u64 logfs_seek_data(struct inode *inode, u64 bix)
+{
+	struct super_block *sb = inode->i_sb;
+	u64 ret, end;
+
+	ret = __logfs_seek_data(inode, bix);
+	end = i_size_read(inode) >> sb->s_blocksize_bits;
+	if (ret >= end)
+		ret = max(bix, end);
+	return ret;
+}
+
+static int logfs_is_valid_direct(struct logfs_inode *li, u64 bix, u64 ofs)
+{
+	return pure_ofs(li->li_data[bix]) == ofs;
+}
+
+static int __logfs_is_valid_loop(struct inode *inode, u64 bix,
+		u64 ofs, u64 bofs)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	level_t level;
+	int ret;
+	struct page *page;
+
+	for (level = LEVEL(li->li_height); level != 0; level = SUBLEVEL(level)){
+		page = logfs_get_write_page(inode, bix, level);
+		BUG_ON(!page);
+
+		ret = logfs_segment_read(inode, page, bofs, bix, level);
+		if (ret) {
+			logfs_put_write_page(page);
+			return 0;
+		}
+
+		bofs = block_get_pointer(page, get_bits(bix, SUBLEVEL(level)));
+		logfs_put_write_page(page);
+		if (!bofs)
+			return 0;
+
+		if (pure_ofs(bofs) == ofs)
+			return 1;
+	}
+	return 0;
+}
+
+static int logfs_is_valid_loop(struct inode *inode, u64 bix, u64 ofs)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	u64 bofs = li->li_data[INDIRECT_INDEX];
+
+	if (!bofs)
+		return 0;
+
+	if (bix >= maxbix(li->li_height))
+		return 0;
+
+	if (pure_ofs(bofs) == ofs)
+		return 1;
+
+	return __logfs_is_valid_loop(inode, bix, ofs, bofs);
+}
+
+static int __logfs_is_valid_block(struct inode *inode, u64 bix, u64 ofs)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	if ((inode->i_nlink == 0) && atomic_read(&inode->i_count) == 1)
+		return 0;
+
+	if (bix < I0_BLOCKS)
+		return logfs_is_valid_direct(li, bix, ofs);
+	return logfs_is_valid_loop(inode, bix, ofs);
+}
+
+/**
+ * logfs_is_valid_block - check whether this block is still valid
+ *
+ * @sb	- superblock
+ * @ofs	- block physical offset
+ * @ino	- block inode number
+ * @bix	- block index
+ * @level - block level
+ *
+ * Returns 0 if the block is invalid, 1 if it is valid and 2 if it will
+ * become invalid once the journal is written.
+ */
+int logfs_is_valid_block(struct super_block *sb, u64 ofs, u64 ino, u64 bix,
+		gc_level_t gc_level)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *inode;
+	int ret, cookie;
+
+	/* Umount closes a segment with free blocks remaining.  Those
+	 * blocks are by definition invalid. */
+	if (ino == -1)
+		return 0;
+
+	LOGFS_BUG_ON((u64)(u_long)ino != ino, sb);
+
+	inode = logfs_safe_iget(sb, ino, &cookie);
+	if (IS_ERR(inode))
+		goto invalid;
+
+	ret = __logfs_is_valid_block(inode, bix, ofs);
+	logfs_safe_iput(inode, cookie);
+	if (ret)
+		return ret;
+
+invalid:
+	/* Block is nominally invalid, but may still sit in the shadow tree,
+	 * waiting for a journal commit.
+	 */
+	if (btree_lookup64(&super->s_shadow_tree.old, ofs))
+		return 2;
+	return 0;
+}
+
+int logfs_readpage_nolock(struct page *page)
+{
+	struct inode *inode = page->mapping->host;
+	int ret = -EIO;
+
+	ret = logfs_read_block(inode, page, READ);
+
+	if (ret) {
+		ClearPageUptodate(page);
+		SetPageError(page);
+	} else {
+		SetPageUptodate(page);
+		ClearPageError(page);
+	}
+	flush_dcache_page(page);
+
+	return ret;
+}
+
+static int logfs_reserve_bytes(struct inode *inode, int bytes)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+	u64 available = super->s_free_bytes + super->s_dirty_free_bytes
+			- super->s_dirty_used_bytes - super->s_dirty_pages;
+
+	if (!bytes)
+		return 0;
+
+	if (available < bytes)
+		return -ENOSPC;
+
+	if (available < bytes + super->s_root_reserve &&
+			!capable(CAP_SYS_RESOURCE))
+		return -ENOSPC;
+
+	return 0;
+}
+
+int get_page_reserve(struct inode *inode, struct page *page)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+	int ret;
+
+	if (logfs_block(page) && logfs_block(page)->reserved_bytes)
+		return 0;
+
+	logfs_get_wblocks(inode->i_sb, page, WF_LOCK);
+	ret = logfs_reserve_bytes(inode, 6 * LOGFS_MAX_OBJECTSIZE);
+	if (!ret) {
+		alloc_data_block(inode, page);
+		logfs_block(page)->reserved_bytes += 6 * LOGFS_MAX_OBJECTSIZE;
+		super->s_dirty_pages += 6 * LOGFS_MAX_OBJECTSIZE;
+	}
+	logfs_put_wblocks(inode->i_sb, page, WF_LOCK);
+	return ret;
+}
+
+/*
+ * We are protected by write lock.  Push victims up to superblock level
+ * and release transaction when appropriate.
+ */
+/* FIXME: This is currently called from the wrong spots. */
+static void logfs_handle_transaction(struct inode *inode,
+		struct logfs_transaction *ta)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+
+	if (!ta)
+		return;
+	logfs_inode(inode)->li_block->ta = NULL;
+
+	if (inode->i_ino != LOGFS_INO_MASTER) {
+		BUG(); /* FIXME: Yes, this needs more thought */
+		/* just remember the transaction until inode is written */
+		//BUG_ON(logfs_inode(inode)->li_transaction);
+		//logfs_inode(inode)->li_transaction = ta;
+		return;
+	}
+
+	switch (ta->state) {
+	case CREATE_1: /* fall through */
+	case UNLINK_1:
+		BUG_ON(super->s_victim_ino);
+		super->s_victim_ino = ta->ino;
+		break;
+	case CREATE_2: /* fall through */
+	case UNLINK_2:
+		BUG_ON(super->s_victim_ino != ta->ino);
+		super->s_victim_ino = 0;
+		/* transaction ends here - free it */
+		kfree(ta);
+		break;
+	case CROSS_RENAME_1:
+		BUG_ON(super->s_rename_dir);
+		BUG_ON(super->s_rename_pos);
+		super->s_rename_dir = ta->dir;
+		super->s_rename_pos = ta->pos;
+		break;
+	case CROSS_RENAME_2:
+		BUG_ON(super->s_rename_dir != ta->dir);
+		BUG_ON(super->s_rename_pos != ta->pos);
+		super->s_rename_dir = 0;
+		super->s_rename_pos = 0;
+		kfree(ta);
+		break;
+	case TARGET_RENAME_1:
+		BUG_ON(super->s_rename_dir);
+		BUG_ON(super->s_rename_pos);
+		BUG_ON(super->s_victim_ino);
+		super->s_rename_dir = ta->dir;
+		super->s_rename_pos = ta->pos;
+		super->s_victim_ino = ta->ino;
+		break;
+	case TARGET_RENAME_2:
+		BUG_ON(super->s_rename_dir != ta->dir);
+		BUG_ON(super->s_rename_pos != ta->pos);
+		BUG_ON(super->s_victim_ino != ta->ino);
+		super->s_rename_dir = 0;
+		super->s_rename_pos = 0;
+		break;
+	case TARGET_RENAME_3:
+		BUG_ON(super->s_rename_dir);
+		BUG_ON(super->s_rename_pos);
+		BUG_ON(super->s_victim_ino != ta->ino);
+		super->s_victim_ino = 0;
+		kfree(ta);
+		break;
+	default:
+		BUG();
+	}
+}
+
+/*
+ * Not strictly a reservation, but rather a check that we still have enough
+ * space to satisfy the write.
+ */
+static int logfs_reserve_blocks(struct inode *inode, int blocks)
+{
+	return logfs_reserve_bytes(inode, blocks * LOGFS_MAX_OBJECTSIZE);
+}
+
+struct write_control {
+	u64 ofs;
+	long flags;
+};
+
+static struct logfs_shadow *alloc_shadow(struct inode *inode, u64 bix,
+		level_t level, u64 old_ofs)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+	struct logfs_shadow *shadow;
+
+	shadow = mempool_alloc(super->s_shadow_pool, GFP_NOFS);
+	memset(shadow, 0, sizeof(*shadow));
+	shadow->ino = inode->i_ino;
+	shadow->bix = bix;
+	shadow->gc_level = expand_level(inode->i_ino, level);
+	shadow->old_ofs = old_ofs & ~LOGFS_FULLY_POPULATED;
+	return shadow;
+}
+
+static void free_shadow(struct inode *inode, struct logfs_shadow *shadow)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+
+	mempool_free(shadow, super->s_shadow_pool);
+}
+
+/**
+ * fill_shadow_tree - Propagate shadow tree changes due to a write
+ * @inode:	Inode owning the page
+ * @page:	Struct page that was written
+ * @shadow:	Shadow for the current write
+ *
+ * Writes in logfs can result in two semi-valid objects.  The old object
+ * is still valid as long as it can be reached by following pointers on
+ * the medium.  Only when writes propagate all the way up to the journal
+ * has the new object safely replaced the old one.
+ *
+ * To handle this problem, a struct logfs_shadow is used to represent
+ * every single write.  It is attached to the indirect block, which is
+ * marked dirty.  When the indirect block is written, its shadows are
+ * handed up to the next indirect block (or inode).  Untimately they
+ * will reach the master inode and be freed upon journal commit.
+ *
+ * This function handles a single step in the propagation.  It adds the
+ * shadow for the current write to the tree, along with any shadows in
+ * the page's tree, in case it was an indirect block.  If a page is
+ * written, the inode parameter is left NULL, if an inode is written,
+ * the page parameter is left NULL.
+ */
+static void fill_shadow_tree(struct inode *inode, struct page *page,
+		struct logfs_shadow *shadow)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+	struct logfs_block *block = logfs_block(page);
+	struct shadow_tree *tree = &super->s_shadow_tree;
+
+	if (PagePrivate(page)) {
+		if (block->alias_map)
+			super->s_no_object_aliases -= bitmap_weight(
+					block->alias_map, LOGFS_BLOCK_FACTOR);
+		logfs_handle_transaction(inode, block->ta);
+		block->ops->free_block(inode->i_sb, block);
+	}
+	if (shadow) {
+		if (shadow->old_ofs)
+			btree_insert64(&tree->old, shadow->old_ofs, shadow,
+					GFP_NOFS);
+		else
+			btree_insert64(&tree->new, shadow->new_ofs, shadow,
+					GFP_NOFS);
+
+		super->s_dirty_used_bytes += shadow->new_len;
+		super->s_dirty_free_bytes += shadow->old_len;
+	}
+}
+
+static void logfs_set_alias(struct super_block *sb, struct logfs_block *block,
+		long child_no)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	if (block->inode && block->inode->i_ino == LOGFS_INO_MASTER) {
+		/* Aliases in the master inode are pointless. */
+		return;
+	}
+
+	if (!test_bit(child_no, block->alias_map)) {
+		set_bit(child_no, block->alias_map);
+		super->s_no_object_aliases++;
+	}
+	list_move_tail(&block->alias_list, &super->s_object_alias);
+}
+
+/*
+ * Object aliases can and often do change the size and occupied space of a
+ * file.  So not only do we have to change the pointers, we also have to
+ * change inode->i_size and li->li_used_bytes.  Which is done by setting
+ * another two object aliases for the inode itself.
+ */
+static void set_iused(struct inode *inode, struct logfs_shadow *shadow)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	if (shadow->new_len == shadow->old_len)
+		return;
+
+	alloc_inode_block(inode);
+	li->li_used_bytes += shadow->new_len - shadow->old_len;
+	__logfs_set_blocks(inode);
+	logfs_set_alias(inode->i_sb, li->li_block, INODE_USED_OFS);
+	logfs_set_alias(inode->i_sb, li->li_block, INODE_SIZE_OFS);
+}
+
+static int logfs_write_i0(struct inode *inode, struct page *page,
+		struct write_control *wc)
+{
+	struct logfs_shadow *shadow;
+	u64 bix;
+	level_t level;
+	int full, err = 0;
+
+	logfs_unpack_index(page->index, &bix, &level);
+	if (wc->ofs == 0)
+		if (logfs_reserve_blocks(inode, 1))
+			return -ENOSPC;
+
+	shadow = alloc_shadow(inode, bix, level, wc->ofs);
+	if (wc->flags & WF_WRITE)
+		err = logfs_segment_write(inode, page, shadow);
+	if (wc->flags & WF_DELETE)
+		logfs_segment_delete(inode, shadow);
+	if (err) {
+		free_shadow(inode, shadow);
+		return err;
+	}
+
+	set_iused(inode, shadow);
+	full = 1;
+	if (level != 0) {
+		alloc_indirect_block(inode, page, 0);
+		full = logfs_block(page)->full == LOGFS_BLOCK_FACTOR;
+	}
+	fill_shadow_tree(inode, page, shadow);
+	wc->ofs = shadow->new_ofs;
+	if (wc->ofs && full)
+		wc->ofs |= LOGFS_FULLY_POPULATED;
+	return 0;
+}
+
+static int logfs_write_direct(struct inode *inode, struct page *page,
+		long flags)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	struct write_control wc = {
+		.ofs = li->li_data[page->index],
+		.flags = flags,
+	};
+	int err;
+
+	alloc_inode_block(inode);
+
+	err = logfs_write_i0(inode, page, &wc);
+	if (err)
+		return err;
+
+	li->li_data[page->index] = wc.ofs;
+	logfs_set_alias(inode->i_sb, li->li_block,
+			page->index + INODE_POINTER_OFS);
+	return 0;
+}
+
+static int ptr_change(u64 ofs, struct page *page)
+{
+	struct logfs_block *block = logfs_block(page);
+	int empty0, empty1, full0, full1;
+
+	empty0 = ofs == 0;
+	empty1 = block->partial == 0;
+	if (empty0 != empty1)
+		return 1;
+
+	/* The !! is necessary to shrink result to int */
+	full0 = !!(ofs & LOGFS_FULLY_POPULATED);
+	full1 = block->full == LOGFS_BLOCK_FACTOR;
+	if (full0 != full1)
+		return 1;
+	return 0;
+}
+
+static int __logfs_write_rec(struct inode *inode, struct page *page,
+		struct write_control *this_wc,
+		pgoff_t bix, level_t target_level, level_t level)
+{
+	int ret, page_empty = 0;
+	int child_no = get_bits(bix, SUBLEVEL(level));
+	struct page *ipage;
+	struct write_control child_wc = {
+		.flags = this_wc->flags,
+	};
+
+	ipage = logfs_get_write_page(inode, bix, level);
+	if (!ipage)
+		return -ENOMEM;
+
+	if (this_wc->ofs) {
+		ret = logfs_segment_read(inode, ipage, this_wc->ofs, bix, level);
+		if (ret)
+			goto out;
+	} else if (!PageUptodate(ipage)) {
+		page_empty = 1;
+		logfs_read_empty(ipage);
+	}
+
+	child_wc.ofs = block_get_pointer(ipage, child_no);
+
+	if ((__force u8)level-1 > (__force u8)target_level)
+		ret = __logfs_write_rec(inode, page, &child_wc, bix,
+				target_level, SUBLEVEL(level));
+	else
+		ret = logfs_write_i0(inode, page, &child_wc);
+
+	if (ret)
+		goto out;
+
+	alloc_indirect_block(inode, ipage, page_empty);
+	block_set_pointer(ipage, child_no, child_wc.ofs);
+	/* FIXME: first condition seems superfluous */
+	if (child_wc.ofs || logfs_block(ipage)->partial)
+		this_wc->flags |= WF_WRITE;
+	/* the condition on this_wc->ofs ensures that we won't consume extra
+	 * space for indirect blocks in the future, which we cannot reserve */
+	if (!this_wc->ofs || ptr_change(this_wc->ofs, ipage))
+		ret = logfs_write_i0(inode, ipage, this_wc);
+	else
+		logfs_set_alias(inode->i_sb, logfs_block(ipage), child_no);
+out:
+	logfs_put_write_page(ipage);
+	return ret;
+}
+
+static int logfs_write_rec(struct inode *inode, struct page *page,
+		pgoff_t bix, level_t target_level, long flags)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	struct write_control wc = {
+		.ofs = li->li_data[INDIRECT_INDEX],
+		.flags = flags,
+	};
+	int ret;
+
+	alloc_inode_block(inode);
+
+	if (li->li_height > (__force u8)target_level)
+		ret = __logfs_write_rec(inode, page, &wc, bix, target_level,
+				LEVEL(li->li_height));
+	else
+		ret = logfs_write_i0(inode, page, &wc);
+	if (ret)
+		return ret;
+
+	if (li->li_data[INDIRECT_INDEX] != wc.ofs) {
+		li->li_data[INDIRECT_INDEX] = wc.ofs;
+		logfs_set_alias(inode->i_sb, li->li_block,
+				INDIRECT_INDEX + INODE_POINTER_OFS);
+	}
+	return ret;
+}
+
+void logfs_add_transaction(struct inode *inode, struct logfs_transaction *ta)
+{
+	alloc_inode_block(inode);
+	logfs_inode(inode)->li_block->ta = ta;
+}
+
+void logfs_del_transaction(struct inode *inode, struct logfs_transaction *ta)
+{
+	struct logfs_block *block = logfs_inode(inode)->li_block;
+
+	if (block && block->ta)
+		block->ta = NULL;
+}
+
+static int grow_inode(struct inode *inode, u64 bix, level_t level)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	u8 height = (__force u8)level;
+	struct page *page;
+	struct write_control wc = {
+		.flags = WF_WRITE,
+	};
+	int err;
+
+	BUG_ON(height > 5 || li->li_height > 5);
+	while (height > li->li_height || bix >= maxbix(li->li_height)) {
+		page = logfs_get_write_page(inode, I0_BLOCKS + 1,
+				LEVEL(li->li_height + 1));
+		if (!page)
+			return -ENOMEM;
+		logfs_read_empty(page);
+		alloc_indirect_block(inode, page, 1);
+		block_set_pointer(page, 0, li->li_data[INDIRECT_INDEX]);
+		err = logfs_write_i0(inode, page, &wc);
+		logfs_put_write_page(page);
+		if (err)
+			return err;
+		li->li_data[INDIRECT_INDEX] = wc.ofs;
+		wc.ofs = 0;
+		li->li_height++;
+		logfs_set_alias(inode->i_sb, li->li_block, INODE_HEIGHT_OFS);
+	}
+	return 0;
+}
+
+static int __logfs_write_buf(struct inode *inode, struct page *page, long flags)
+{
+	struct logfs_super *super = logfs_super(inode->i_sb);
+	pgoff_t index = page->index;
+	u64 bix;
+	level_t level;
+	int err;
+
+	flags |= WF_WRITE | WF_DELETE;
+	inode->i_ctime = inode->i_mtime = CURRENT_TIME;
+
+	logfs_unpack_index(index, &bix, &level);
+	if (logfs_block(page) && logfs_block(page)->reserved_bytes)
+		super->s_dirty_pages -= logfs_block(page)->reserved_bytes;
+
+	if (index < I0_BLOCKS)
+		return logfs_write_direct(inode, page, flags);
+
+	bix = adjust_bix(bix, level);
+	err = grow_inode(inode, bix, level);
+	if (err)
+		return err;
+	return logfs_write_rec(inode, page, bix, level, flags);
+}
+
+int logfs_write_buf(struct inode *inode, struct page *page, long flags)
+{
+	struct super_block *sb = inode->i_sb;
+	int ret;
+
+	logfs_get_wblocks(sb, page, flags & WF_LOCK);
+	ret = __logfs_write_buf(inode, page, flags);
+	logfs_put_wblocks(sb, page, flags & WF_LOCK);
+	return ret;
+}
+
+static int __logfs_delete(struct inode *inode, struct page *page)
+{
+	long flags = WF_DELETE;
+
+	inode->i_ctime = inode->i_mtime = CURRENT_TIME;
+
+	if (page->index < I0_BLOCKS)
+		return logfs_write_direct(inode, page, flags);
+	return logfs_write_rec(inode, page, page->index, 0, flags);
+}
+
+int logfs_delete(struct inode *inode, pgoff_t index,
+		struct shadow_tree *shadow_tree)
+{
+	struct super_block *sb = inode->i_sb;
+	struct page *page;
+	int ret;
+
+	page = logfs_get_read_page(inode, index, 0);
+	if (!page)
+		return -ENOMEM;
+
+	logfs_get_wblocks(sb, page, 1);
+	ret = __logfs_delete(inode, page);
+	logfs_put_wblocks(sb, page, 1);
+
+	logfs_put_read_page(page);
+
+	return ret;
+}
+
+/* Rewrite cannot mark the inode dirty but has to write it immediatly. */
+int logfs_rewrite_block(struct inode *inode, u64 bix, u64 ofs,
+		gc_level_t gc_level, long flags)
+{
+	level_t level = shrink_level(gc_level);
+	struct page *page;
+	int err;
+
+	page = logfs_get_write_page(inode, bix, level);
+	if (!page)
+		return -ENOMEM;
+
+	err = logfs_segment_read(inode, page, ofs, bix, level);
+	if (!err) {
+		if (level != 0)
+			alloc_indirect_block(inode, page, 0);
+		err = logfs_write_buf(inode, page, flags);
+	}
+	logfs_put_write_page(page);
+	return err;
+}
+
+static int truncate_data_block(struct inode *inode, struct page *page,
+		u64 ofs, struct logfs_shadow *shadow, u64 size)
+{
+	loff_t pageofs = page->index << inode->i_sb->s_blocksize_bits;
+	u64 bix;
+	level_t level;
+	int err;
+
+	/* Does truncation happen within this page? */
+	if (size <= pageofs || size - pageofs >= PAGE_SIZE)
+		return 0;
+
+	logfs_unpack_index(page->index, &bix, &level);
+	BUG_ON(level != 0);
+
+	err = logfs_segment_read(inode, page, ofs, bix, level);
+	if (err)
+		return err;
+
+	zero_user_segment(page, size - pageofs, PAGE_CACHE_SIZE);
+	return logfs_segment_write(inode, page, shadow);
+}
+
+static int logfs_truncate_i0(struct inode *inode, struct page *page,
+		struct write_control *wc, u64 size)
+{
+	struct logfs_shadow *shadow;
+	u64 bix;
+	level_t level;
+	int err = 0;
+
+	logfs_unpack_index(page->index, &bix, &level);
+	BUG_ON(level != 0);
+	shadow = alloc_shadow(inode, bix, level, wc->ofs);
+
+	err = truncate_data_block(inode, page, wc->ofs, shadow, size);
+	if (err) {
+		free_shadow(inode, shadow);
+		return err;
+	}
+
+	logfs_segment_delete(inode, shadow);
+	set_iused(inode, shadow);
+	fill_shadow_tree(inode, page, shadow);
+	wc->ofs = shadow->new_ofs;
+	return 0;
+}
+
+static int logfs_truncate_direct(struct inode *inode, u64 size)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	struct write_control wc;
+	struct page *page;
+	int e;
+	int err;
+
+	alloc_inode_block(inode);
+
+	for (e = I0_BLOCKS - 1; e >= 0; e--) {
+		if (size > (e+1) * LOGFS_BLOCKSIZE)
+			break;
+
+		wc.ofs = li->li_data[e];
+		if (!wc.ofs)
+			continue;
+
+		page = logfs_get_write_page(inode, e, 0);
+		if (!page)
+			return -ENOMEM;
+		err = logfs_segment_read(inode, page, wc.ofs, e, 0);
+		if (err) {
+			logfs_put_write_page(page);
+			return err;
+		}
+		err = logfs_truncate_i0(inode, page, &wc, size);
+		logfs_put_write_page(page);
+		if (err)
+			return err;
+
+		li->li_data[e] = wc.ofs;
+	}
+	return 0;
+}
+
+/* FIXME: these need to become per-sb once we support different blocksizes */
+static u64 __logfs_step[] = {
+	1,
+	I1_BLOCKS,
+	I2_BLOCKS,
+	I3_BLOCKS,
+};
+
+static u64 __logfs_start_index[] = {
+	I0_BLOCKS,
+	I1_BLOCKS,
+	I2_BLOCKS,
+	I3_BLOCKS
+};
+
+static inline u64 logfs_step(level_t level)
+{
+	return __logfs_step[(__force u8)level];
+}
+
+static inline u64 logfs_factor(u8 level)
+{
+	return __logfs_step[level] * LOGFS_BLOCKSIZE;
+}
+
+static inline u64 logfs_start_index(level_t level)
+{
+	return __logfs_start_index[(__force u8)level];
+}
+
+static void logfs_unpack_raw_index(pgoff_t index, u64 *bix, level_t *level)
+{
+	logfs_unpack_index(index, bix, level);
+	if (*bix <= logfs_start_index(SUBLEVEL(*level)))
+		*bix = 0;
+}
+
+static int __logfs_truncate_rec(struct inode *inode, struct page *ipage,
+		struct write_control *this_wc, u64 size)
+{
+	int truncate_happened = 0;
+	int e, err = 0;
+	u64 bix, child_bix, next_bix;
+	level_t level;
+	struct page *page;
+	struct write_control child_wc = { /* FIXME: flags */ };
+
+	logfs_unpack_raw_index(ipage->index, &bix, &level);
+	err = logfs_segment_read(inode, ipage, this_wc->ofs, bix, level);
+	if (err)
+		return err;
+
+	for (e = LOGFS_BLOCK_FACTOR - 1; e >= 0; e--) {
+		child_bix = bix + e * logfs_step(SUBLEVEL(level));
+		next_bix = child_bix + logfs_step(SUBLEVEL(level));
+		if (size > next_bix * LOGFS_BLOCKSIZE)
+			break;
+
+		child_wc.ofs = pure_ofs(block_get_pointer(ipage, e));
+		if (!child_wc.ofs)
+			continue;
+
+		page = logfs_get_write_page(inode, child_bix, SUBLEVEL(level));
+		if (!page)
+			return -ENOMEM;
+
+		if ((__force u8)level > 1)
+			err = __logfs_truncate_rec(inode, page, &child_wc, size);
+		else
+			err = logfs_truncate_i0(inode, page, &child_wc, size);
+		logfs_put_write_page(page);
+		if (err)
+			return err;
+
+		truncate_happened = 1;
+		alloc_indirect_block(inode, ipage, 0);
+		block_set_pointer(ipage, e, child_wc.ofs);
+	}
+
+	if (!truncate_happened) {
+		printk("ineffectual truncate (%lx, %lx, %llx)\n", inode->i_ino, ipage->index, size);
+		return 0;
+	}
+
+	this_wc->flags = WF_DELETE;
+	if (logfs_block(ipage)->partial)
+		this_wc->flags |= WF_WRITE;
+
+	return logfs_write_i0(inode, ipage, this_wc);
+}
+
+static int logfs_truncate_rec(struct inode *inode, u64 size)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	struct write_control wc = {
+		.ofs = li->li_data[INDIRECT_INDEX],
+	};
+	struct page *page;
+	int err;
+
+	alloc_inode_block(inode);
+
+	if (!wc.ofs)
+		return 0;
+
+	page = logfs_get_write_page(inode, 0, LEVEL(li->li_height));
+	if (!page)
+		return -ENOMEM;
+
+	err = __logfs_truncate_rec(inode, page, &wc, size);
+	logfs_put_write_page(page);
+	if (err)
+		return err;
+
+	if (li->li_data[INDIRECT_INDEX] != wc.ofs)
+		li->li_data[INDIRECT_INDEX] = wc.ofs;
+	return 0;
+}
+
+static int __logfs_truncate(struct inode *inode, u64 size)
+{
+	int ret;
+
+	if (size >= logfs_factor(logfs_inode(inode)->li_height))
+		return 0;
+
+	ret = logfs_truncate_rec(inode, size);
+	if (ret)
+		return ret;
+
+	return logfs_truncate_direct(inode, size);
+}
+
+int logfs_truncate(struct inode *inode, u64 size)
+{
+	struct super_block *sb = inode->i_sb;
+	int err;
+
+	logfs_get_wblocks(sb, NULL, 1);
+	err = __logfs_truncate(inode, size);
+	if (!err)
+		err = __logfs_write_inode(inode, 0);
+	logfs_put_wblocks(sb, NULL, 1);
+
+	if (!err)
+		err = vmtruncate(inode, size);
+
+	/* I don't trust error recovery yet. */
+	WARN_ON(err);
+	return err;
+}
+
+static void move_page_to_inode(struct inode *inode, struct page *page)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	struct logfs_block *block = logfs_block(page);
+
+	if (!block)
+		return;
+
+	log_blockmove("move_page_to_inode(%llx, %llx, %x)\n",
+			block->ino, block->bix, block->level);
+	BUG_ON(li->li_block);
+	block->ops = &inode_block_ops;
+	block->inode = inode;
+	li->li_block = block;
+
+	block->page = NULL;
+	page->private = 0;
+	ClearPagePrivate(page);
+}
+
+static void move_inode_to_page(struct page *page, struct inode *inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+	struct logfs_block *block = li->li_block;
+
+	if (!block)
+		return;
+
+	log_blockmove("move_inode_to_page(%llx, %llx, %x)\n",
+			block->ino, block->bix, block->level);
+	BUG_ON(PagePrivate(page));
+	block->ops = &indirect_block_ops;
+	block->page = page;
+	page->private = (unsigned long)block;
+	SetPagePrivate(page);
+
+	block->inode = NULL;
+	li->li_block = NULL;
+}
+
+int logfs_read_inode(struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *master_inode = super->s_master_inode;
+	struct page *page;
+	struct logfs_disk_inode *di;
+	u64 ino = inode->i_ino;
+
+	if (ino << sb->s_blocksize_bits > i_size_read(master_inode))
+		return -ENODATA;
+	if (!logfs_exist_block(master_inode, ino))
+		return -ENODATA;
+
+	page = read_cache_page(master_inode->i_mapping, ino,
+			(filler_t *)logfs_readpage, NULL);
+	if (IS_ERR(page))
+		return PTR_ERR(page);
+
+	di = kmap_atomic(page, KM_USER0);
+	logfs_disk_to_inode(di, inode);
+	kunmap_atomic(di, KM_USER0);
+	move_page_to_inode(inode, page);
+	page_cache_release(page);
+	return 0;
+}
+
+/* Caller must logfs_put_write_page(page); */
+static struct page *inode_to_page(struct inode *inode)
+{
+	struct inode *master_inode = logfs_super(inode->i_sb)->s_master_inode;
+	struct logfs_disk_inode *di;
+	struct page *page;
+
+	BUG_ON(inode->i_ino == LOGFS_INO_MASTER);
+
+	page = logfs_get_write_page(master_inode, inode->i_ino, 0);
+	if (!page)
+		return NULL;
+
+	di = kmap_atomic(page, KM_USER0);
+	logfs_inode_to_disk(inode, di);
+	kunmap_atomic(di, KM_USER0);
+	move_inode_to_page(page, inode);
+	return page;
+}
+
+/* Cheaper version of write_inode.  All changes are concealed in
+ * aliases, which are moved back.  No write to the medium happens.
+ */
+void logfs_clear_inode(struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+	struct logfs_inode *li = logfs_inode(inode);
+	struct logfs_block *block = li->li_block;
+	struct page *page;
+
+	/* Only deleted files may be dirty at this point */
+	BUG_ON(inode->i_state & I_DIRTY && inode->i_nlink);
+	if (!block)
+		return;
+	if ((logfs_super(sb)->s_flags & LOGFS_SB_FLAG_SHUTDOWN)) {
+		block->ops->free_block(inode->i_sb, block);
+		return;
+	}
+
+	BUG_ON(inode->i_ino < LOGFS_RESERVED_INOS);
+	page = inode_to_page(inode);
+	BUG_ON(!page); /* FIXME: Use emergency page */
+	logfs_put_write_page(page);
+}
+
+static int do_write_inode(struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+	struct inode *master_inode = logfs_super(sb)->s_master_inode;
+	loff_t size = (inode->i_ino + 1) << inode->i_sb->s_blocksize_bits;
+	struct page *page;
+	int err;
+
+	BUG_ON(inode->i_ino == LOGFS_INO_MASTER);
+	/* FIXME: lock inode */
+
+	if (i_size_read(master_inode) < size)
+		i_size_write(master_inode, size);
+
+	/* TODO: Tell vfs this inode is clean now */
+
+	page = inode_to_page(inode);
+	if (!page)
+		return -ENOMEM;
+
+	/* FIXME: transaction is part of logfs_block now.  Is that enough? */
+	err = logfs_write_buf(master_inode, page, 0);
+	logfs_put_write_page(page);
+	return err;
+}
+
+static void logfs_mod_segment_entry(struct super_block *sb, u32 segno,
+		int write,
+		void (*change_se)(struct logfs_segment_entry *, long),
+		long arg)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *inode;
+	struct page *page;
+	struct logfs_segment_entry *se;
+	pgoff_t page_no;
+	int child_no;
+
+	page_no = segno >> (sb->s_blocksize_bits - 3);
+	child_no = segno & ((sb->s_blocksize >> 3) - 1);
+
+	inode = super->s_segfile_inode;
+	page = logfs_get_write_page(inode, page_no, 0);
+	BUG_ON(!page); /* FIXME: We need some reserve page for this case */
+	if (!PageUptodate(page))
+		logfs_read_block(inode, page, WRITE);
+
+	if (write)
+		alloc_indirect_block(inode, page, 0);
+	se = kmap_atomic(page, KM_USER0);
+	change_se(se + child_no, arg);
+	if (write) {
+		logfs_set_alias(sb, logfs_block(page), child_no);
+		BUG_ON((int)be32_to_cpu(se[child_no].valid) > super->s_segsize);
+	}
+	kunmap_atomic(se, KM_USER0);
+
+	logfs_put_write_page(page);
+}
+
+static void __get_segment_entry(struct logfs_segment_entry *se, long _target)
+{
+	struct logfs_segment_entry *target = (void *)_target;
+
+	*target = *se;
+}
+
+void logfs_get_segment_entry(struct super_block *sb, u32 segno,
+		struct logfs_segment_entry *se)
+{
+	logfs_mod_segment_entry(sb, segno, 0, __get_segment_entry, (long)se);
+}
+
+static void __set_segment_used(struct logfs_segment_entry *se, long increment)
+{
+	u32 valid;
+
+	valid = be32_to_cpu(se->valid);
+	valid += increment;
+	se->valid = cpu_to_be32(valid);
+}
+
+void logfs_set_segment_used(struct super_block *sb, u64 ofs, int increment)
+{
+	struct logfs_super *super = logfs_super(sb);
+	u32 segno = ofs >> super->s_segshift;
+
+	if (!increment)
+		return;
+
+	logfs_mod_segment_entry(sb, segno, 1, __set_segment_used, increment);
+}
+
+static void __set_segment_erased(struct logfs_segment_entry *se, long ec_level)
+{
+	se->ec_level = cpu_to_be32(ec_level);
+}
+
+void logfs_set_segment_erased(struct super_block *sb, u32 segno, u32 ec,
+		gc_level_t gc_level)
+{
+	u32 ec_level = ec << 4 | (__force u8)gc_level;
+
+	logfs_mod_segment_entry(sb, segno, 1, __set_segment_erased, ec_level);
+}
+
+static void __set_segment_reserved(struct logfs_segment_entry *se, long ignore)
+{
+	se->valid = cpu_to_be32(RESERVED);
+}
+
+void logfs_set_segment_reserved(struct super_block *sb, u32 segno)
+{
+	logfs_mod_segment_entry(sb, segno, 1, __set_segment_reserved, 0);
+}
+
+static void __set_segment_unreserved(struct logfs_segment_entry *se,
+		long ec_level)
+{
+	se->valid = 0;
+	se->ec_level = cpu_to_be32(ec_level);
+}
+
+void logfs_set_segment_unreserved(struct super_block *sb, u32 segno, u32 ec)
+{
+	u32 ec_level = ec << 4;
+
+	logfs_mod_segment_entry(sb, segno, 1, __set_segment_unreserved,
+			ec_level);
+}
+
+int __logfs_write_inode(struct inode *inode, long flags)
+{
+	struct super_block *sb = inode->i_sb;
+	int ret;
+
+	logfs_get_wblocks(sb, NULL, flags & WF_LOCK);
+	ret = do_write_inode(inode);
+	logfs_put_wblocks(sb, NULL, flags & WF_LOCK);
+	return ret;
+}
+
+static int do_delete_inode(struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+	struct inode *master_inode = logfs_super(sb)->s_master_inode;
+	struct page *page;
+	int ret;
+
+	page = logfs_get_write_page(master_inode, inode->i_ino, 0);
+	if (!page)
+		return -ENOMEM;
+
+	move_inode_to_page(page, inode);
+
+	logfs_get_wblocks(sb, page, 1);
+	ret = __logfs_delete(master_inode, page);
+	logfs_put_wblocks(sb, page, 1);
+
+	logfs_put_write_page(page);
+	return ret;
+}
+
+/*
+ * ZOMBIE inodes have already been deleted before and should remain dead,
+ * if it weren't for valid checking.  No need to kill them again here.
+ */
+void logfs_delete_inode(struct inode *inode)
+{
+	struct logfs_inode *li = logfs_inode(inode);
+
+	if (!(li->li_flags & LOGFS_IF_ZOMBIE)) {
+		li->li_flags |= LOGFS_IF_ZOMBIE;
+		if (i_size_read(inode) > 0)
+			logfs_truncate(inode, 0);
+		do_delete_inode(inode);
+	}
+	truncate_inode_pages(&inode->i_data, 0);
+	clear_inode(inode);
+}
+
+void btree_write_block(struct logfs_block *block)
+{
+	struct inode *inode;
+	struct page *page;
+	int err, cookie;
+
+	inode = logfs_safe_iget(block->sb, block->ino, &cookie);
+	page = logfs_get_write_page(inode, block->bix, block->level);
+
+	err = logfs_readpage_nolock(page);
+	BUG_ON(err);
+	BUG_ON(!PagePrivate(page));
+	BUG_ON(logfs_block(page) != block);
+	err = __logfs_write_buf(inode, page, 0);
+	BUG_ON(err);
+	BUG_ON(PagePrivate(page) || page->private);
+
+	logfs_put_write_page(page);
+	logfs_safe_iput(inode, cookie);
+}
+
+/**
+ * logfs_inode_write - write inode or dentry objects
+ *
+ * @inode:		parent inode (ifile or directory)
+ * @buf:		object to write (inode or dentry)
+ * @n:			object size
+ * @_pos:		object number (file position in blocks/objects)
+ * @flags:		write flags
+ * @lock:		0 if write lock is already taken, 1 otherwise
+ * @shadow_tree:	shadow below this inode
+ *
+ * FIXME: All caller of this put a 200-300 byte variable on the stack,
+ * only to call here and do a memcpy from that stack variable.  A good
+ * example of wasted performance and stack space.
+ */
+int logfs_inode_write(struct inode *inode, const void *buf, size_t count,
+		loff_t bix, long flags, struct shadow_tree *shadow_tree)
+{
+	loff_t pos = bix << inode->i_sb->s_blocksize_bits;
+	int err;
+	struct page *page;
+	void *pagebuf;
+
+	BUG_ON(pos & (LOGFS_BLOCKSIZE-1));
+	BUG_ON(count > LOGFS_BLOCKSIZE);
+	page = logfs_get_write_page(inode, bix, 0);
+	if (!page)
+		return -ENOMEM;
+
+	pagebuf = kmap_atomic(page, KM_USER0);
+	memcpy(pagebuf, buf, count);
+	flush_dcache_page(page);
+	kunmap_atomic(pagebuf, KM_USER0);
+
+	if (i_size_read(inode) < pos + LOGFS_BLOCKSIZE)
+		i_size_write(inode, pos + LOGFS_BLOCKSIZE);
+
+	err = logfs_write_buf(inode, page, flags);
+	logfs_put_write_page(page);
+	return err;
+}
+
+int logfs_open_segfile(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *inode;
+
+	inode = logfs_read_meta_inode(sb, LOGFS_INO_SEGFILE);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+	super->s_segfile_inode = inode;
+	return 0;
+}
+
+int logfs_init_rw(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int min_fill = 3 * super->s_no_blocks;
+
+	INIT_LIST_HEAD(&super->s_object_alias);
+	mutex_init(&super->s_write_mutex);
+	super->s_block_pool = mempool_create_kmalloc_pool(min_fill,
+			sizeof(struct logfs_block));
+	super->s_shadow_pool = mempool_create_kmalloc_pool(min_fill,
+			sizeof(struct logfs_shadow));
+	return 0;
+}
+
+void logfs_cleanup_rw(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	destroy_meta_inode(super->s_segfile_inode);
+	if (super->s_block_pool)
+		mempool_destroy(super->s_block_pool);
+	if (super->s_shadow_pool)
+		mempool_destroy(super->s_shadow_pool);
+}
diff --git a/fs/logfs/segment.c b/fs/logfs/segment.c
new file mode 100644
index 000000000000..1a14f9910d55
--- /dev/null
+++ b/fs/logfs/segment.c
@@ -0,0 +1,927 @@
+/*
+ * fs/logfs/segment.c	- Handling the Object Store
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ *
+ * Object store or ostore makes up the complete device with exception of
+ * the superblock and journal areas.  Apart from its own metadata it stores
+ * three kinds of objects: inodes, dentries and blocks, both data and indirect.
+ */
+#include "logfs.h"
+
+static int logfs_mark_segment_bad(struct super_block *sb, u32 segno)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct btree_head32 *head = &super->s_reserved_segments;
+	int err;
+
+	err = btree_insert32(head, segno, (void *)1, GFP_NOFS);
+	if (err)
+		return err;
+	logfs_super(sb)->s_bad_segments++;
+	/* FIXME: write to journal */
+	return 0;
+}
+
+int logfs_erase_segment(struct super_block *sb, u32 segno, int ensure_erase)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	super->s_gec++;
+
+	return super->s_devops->erase(sb, (u64)segno << super->s_segshift,
+			super->s_segsize, ensure_erase);
+}
+
+static s64 logfs_get_free_bytes(struct logfs_area *area, size_t bytes)
+{
+	s32 ofs;
+
+	logfs_open_area(area, bytes);
+
+	ofs = area->a_used_bytes;
+	area->a_used_bytes += bytes;
+	BUG_ON(area->a_used_bytes >= logfs_super(area->a_sb)->s_segsize);
+
+	return dev_ofs(area->a_sb, area->a_segno, ofs);
+}
+
+static struct page *get_mapping_page(struct super_block *sb, pgoff_t index,
+		int use_filler)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	filler_t *filler = super->s_devops->readpage;
+	struct page *page;
+
+	BUG_ON(mapping_gfp_mask(mapping) & __GFP_FS);
+	if (use_filler)
+		page = read_cache_page(mapping, index, filler, sb);
+	else {
+		page = find_or_create_page(mapping, index, GFP_NOFS);
+		unlock_page(page);
+	}
+	return page;
+}
+
+void __logfs_buf_write(struct logfs_area *area, u64 ofs, void *buf, size_t len,
+		int use_filler)
+{
+	pgoff_t index = ofs >> PAGE_SHIFT;
+	struct page *page;
+	long offset = ofs & (PAGE_SIZE-1);
+	long copylen;
+
+	/* Only logfs_wbuf_recover may use len==0 */
+	BUG_ON(!len && !use_filler);
+	do {
+		copylen = min((ulong)len, PAGE_SIZE - offset);
+
+		page = get_mapping_page(area->a_sb, index, use_filler);
+		SetPageUptodate(page);
+		BUG_ON(!page); /* FIXME: reserve a pool */
+		memcpy(page_address(page) + offset, buf, copylen);
+		SetPagePrivate(page);
+		page_cache_release(page);
+
+		buf += copylen;
+		len -= copylen;
+		offset = 0;
+		index++;
+	} while (len);
+}
+
+/*
+ * bdev_writeseg will write full pages.  Memset the tail to prevent data leaks.
+ */
+static void pad_wbuf(struct logfs_area *area, int final)
+{
+	struct super_block *sb = area->a_sb;
+	struct logfs_super *super = logfs_super(sb);
+	struct page *page;
+	u64 ofs = dev_ofs(sb, area->a_segno, area->a_used_bytes);
+	pgoff_t index = ofs >> PAGE_SHIFT;
+	long offset = ofs & (PAGE_SIZE-1);
+	u32 len = PAGE_SIZE - offset;
+
+	if (len == PAGE_SIZE) {
+		/* The math in this function can surely use some love */
+		len = 0;
+	}
+	if (len) {
+		BUG_ON(area->a_used_bytes >= super->s_segsize);
+
+		page = get_mapping_page(area->a_sb, index, 0);
+		BUG_ON(!page); /* FIXME: reserve a pool */
+		memset(page_address(page) + offset, 0xff, len);
+		SetPagePrivate(page);
+		page_cache_release(page);
+	}
+
+	if (!final)
+		return;
+
+	area->a_used_bytes += len;
+	for ( ; area->a_used_bytes < super->s_segsize;
+			area->a_used_bytes += PAGE_SIZE) {
+		/* Memset another page */
+		index++;
+		page = get_mapping_page(area->a_sb, index, 0);
+		BUG_ON(!page); /* FIXME: reserve a pool */
+		memset(page_address(page), 0xff, PAGE_SIZE);
+		SetPagePrivate(page);
+		page_cache_release(page);
+	}
+}
+
+/*
+ * We have to be careful with the alias tree.  Since lookup is done by bix,
+ * it needs to be normalized, so 14, 15, 16, etc. all match when dealing with
+ * indirect blocks.  So always use it through accessor functions.
+ */
+static void *alias_tree_lookup(struct super_block *sb, u64 ino, u64 bix,
+		level_t level)
+{
+	struct btree_head128 *head = &logfs_super(sb)->s_object_alias_tree;
+	pgoff_t index = logfs_pack_index(bix, level);
+
+	return btree_lookup128(head, ino, index);
+}
+
+static int alias_tree_insert(struct super_block *sb, u64 ino, u64 bix,
+		level_t level, void *val)
+{
+	struct btree_head128 *head = &logfs_super(sb)->s_object_alias_tree;
+	pgoff_t index = logfs_pack_index(bix, level);
+
+	return btree_insert128(head, ino, index, val, GFP_NOFS);
+}
+
+static int btree_write_alias(struct super_block *sb, struct logfs_block *block,
+		write_alias_t *write_one_alias)
+{
+	struct object_alias_item *item;
+	int err;
+
+	list_for_each_entry(item, &block->item_list, list) {
+		err = write_alias_journal(sb, block->ino, block->bix,
+				block->level, item->child_no, item->val);
+		if (err)
+			return err;
+	}
+	return 0;
+}
+
+static gc_level_t btree_block_level(struct logfs_block *block)
+{
+	return expand_level(block->ino, block->level);
+}
+
+static struct logfs_block_ops btree_block_ops = {
+	.write_block	= btree_write_block,
+	.block_level	= btree_block_level,
+	.free_block	= __free_block,
+	.write_alias	= btree_write_alias,
+};
+
+int logfs_load_object_aliases(struct super_block *sb,
+		struct logfs_obj_alias *oa, int count)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_block *block;
+	struct object_alias_item *item;
+	u64 ino, bix;
+	level_t level;
+	int i, err;
+
+	super->s_flags |= LOGFS_SB_FLAG_OBJ_ALIAS;
+	count /= sizeof(*oa);
+	for (i = 0; i < count; i++) {
+		item = mempool_alloc(super->s_alias_pool, GFP_NOFS);
+		if (!item)
+			return -ENOMEM;
+		memset(item, 0, sizeof(*item));
+
+		super->s_no_object_aliases++;
+		item->val = oa[i].val;
+		item->child_no = be16_to_cpu(oa[i].child_no);
+
+		ino = be64_to_cpu(oa[i].ino);
+		bix = be64_to_cpu(oa[i].bix);
+		level = LEVEL(oa[i].level);
+
+		log_aliases("logfs_load_object_aliases(%llx, %llx, %x, %x) %llx\n",
+				ino, bix, level, item->child_no,
+				be64_to_cpu(item->val));
+		block = alias_tree_lookup(sb, ino, bix, level);
+		if (!block) {
+			block = __alloc_block(sb, ino, bix, level);
+			block->ops = &btree_block_ops;
+			err = alias_tree_insert(sb, ino, bix, level, block);
+			BUG_ON(err); /* mempool empty */
+		}
+		if (test_and_set_bit(item->child_no, block->alias_map)) {
+			printk(KERN_ERR"LogFS: Alias collision detected\n");
+			return -EIO;
+		}
+		list_move_tail(&block->alias_list, &super->s_object_alias);
+		list_add(&item->list, &block->item_list);
+	}
+	return 0;
+}
+
+static void kill_alias(void *_block, unsigned long ignore0,
+		u64 ignore1, u64 ignore2, size_t ignore3)
+{
+	struct logfs_block *block = _block;
+	struct super_block *sb = block->sb;
+	struct logfs_super *super = logfs_super(sb);
+	struct object_alias_item *item;
+
+	while (!list_empty(&block->item_list)) {
+		item = list_entry(block->item_list.next, typeof(*item), list);
+		list_del(&item->list);
+		mempool_free(item, super->s_alias_pool);
+	}
+	block->ops->free_block(sb, block);
+}
+
+static int obj_type(struct inode *inode, level_t level)
+{
+	if (level == 0) {
+		if (S_ISDIR(inode->i_mode))
+			return OBJ_DENTRY;
+		if (inode->i_ino == LOGFS_INO_MASTER)
+			return OBJ_INODE;
+	}
+	return OBJ_BLOCK;
+}
+
+static int obj_len(struct super_block *sb, int obj_type)
+{
+	switch (obj_type) {
+	case OBJ_DENTRY:
+		return sizeof(struct logfs_disk_dentry);
+	case OBJ_INODE:
+		return sizeof(struct logfs_disk_inode);
+	case OBJ_BLOCK:
+		return sb->s_blocksize;
+	default:
+		BUG();
+	}
+}
+
+static int __logfs_segment_write(struct inode *inode, void *buf,
+		struct logfs_shadow *shadow, int type, int len, int compr)
+{
+	struct logfs_area *area;
+	struct super_block *sb = inode->i_sb;
+	s64 ofs;
+	struct logfs_object_header h;
+	int acc_len;
+
+	if (shadow->gc_level == 0)
+		acc_len = len;
+	else
+		acc_len = obj_len(sb, type);
+
+	area = get_area(sb, shadow->gc_level);
+	ofs = logfs_get_free_bytes(area, len + LOGFS_OBJECT_HEADERSIZE);
+	LOGFS_BUG_ON(ofs <= 0, sb);
+	/*
+	 * Order is important.  logfs_get_free_bytes(), by modifying the
+	 * segment file, may modify the content of the very page we're about
+	 * to write now.  Which is fine, as long as the calculated crc and
+	 * written data still match.  So do the modifications _before_
+	 * calculating the crc.
+	 */
+
+	h.len	= cpu_to_be16(len);
+	h.type	= type;
+	h.compr	= compr;
+	h.ino	= cpu_to_be64(inode->i_ino);
+	h.bix	= cpu_to_be64(shadow->bix);
+	h.crc	= logfs_crc32(&h, sizeof(h) - 4, 4);
+	h.data_crc = logfs_crc32(buf, len, 0);
+
+	logfs_buf_write(area, ofs, &h, sizeof(h));
+	logfs_buf_write(area, ofs + LOGFS_OBJECT_HEADERSIZE, buf, len);
+
+	shadow->new_ofs = ofs;
+	shadow->new_len = acc_len + LOGFS_OBJECT_HEADERSIZE;
+
+	return 0;
+}
+
+static s64 logfs_segment_write_compress(struct inode *inode, void *buf,
+		struct logfs_shadow *shadow, int type, int len)
+{
+	struct super_block *sb = inode->i_sb;
+	void *compressor_buf = logfs_super(sb)->s_compressed_je;
+	ssize_t compr_len;
+	int ret;
+
+	mutex_lock(&logfs_super(sb)->s_journal_mutex);
+	compr_len = logfs_compress(buf, compressor_buf, len, len);
+
+	if (compr_len >= 0) {
+		ret = __logfs_segment_write(inode, compressor_buf, shadow,
+				type, compr_len, COMPR_ZLIB);
+	} else {
+		ret = __logfs_segment_write(inode, buf, shadow, type, len,
+				COMPR_NONE);
+	}
+	mutex_unlock(&logfs_super(sb)->s_journal_mutex);
+	return ret;
+}
+
+/**
+ * logfs_segment_write - write data block to object store
+ * @inode:		inode containing data
+ *
+ * Returns an errno or zero.
+ */
+int logfs_segment_write(struct inode *inode, struct page *page,
+		struct logfs_shadow *shadow)
+{
+	struct super_block *sb = inode->i_sb;
+	struct logfs_super *super = logfs_super(sb);
+	int do_compress, type, len;
+	int ret;
+	void *buf;
+
+	super->s_flags |= LOGFS_SB_FLAG_DIRTY;
+	BUG_ON(super->s_flags & LOGFS_SB_FLAG_SHUTDOWN);
+	do_compress = logfs_inode(inode)->li_flags & LOGFS_IF_COMPRESSED;
+	if (shadow->gc_level != 0) {
+		/* temporarily disable compression for indirect blocks */
+		do_compress = 0;
+	}
+
+	type = obj_type(inode, shrink_level(shadow->gc_level));
+	len = obj_len(sb, type);
+	buf = kmap(page);
+	if (do_compress)
+		ret = logfs_segment_write_compress(inode, buf, shadow, type,
+				len);
+	else
+		ret = __logfs_segment_write(inode, buf, shadow, type, len,
+				COMPR_NONE);
+	kunmap(page);
+
+	log_segment("logfs_segment_write(%llx, %llx, %x) %llx->%llx %x->%x\n",
+			shadow->ino, shadow->bix, shadow->gc_level,
+			shadow->old_ofs, shadow->new_ofs,
+			shadow->old_len, shadow->new_len);
+	/* this BUG_ON did catch a locking bug.  useful */
+	BUG_ON(!(shadow->new_ofs & (super->s_segsize - 1)));
+	return ret;
+}
+
+int wbuf_read(struct super_block *sb, u64 ofs, size_t len, void *buf)
+{
+	pgoff_t index = ofs >> PAGE_SHIFT;
+	struct page *page;
+	long offset = ofs & (PAGE_SIZE-1);
+	long copylen;
+
+	while (len) {
+		copylen = min((ulong)len, PAGE_SIZE - offset);
+
+		page = get_mapping_page(sb, index, 1);
+		if (IS_ERR(page))
+			return PTR_ERR(page);
+		memcpy(buf, page_address(page) + offset, copylen);
+		page_cache_release(page);
+
+		buf += copylen;
+		len -= copylen;
+		offset = 0;
+		index++;
+	}
+	return 0;
+}
+
+/*
+ * The "position" of indirect blocks is ambiguous.  It can be the position
+ * of any data block somewhere behind this indirect block.  So we need to
+ * normalize the positions through logfs_block_mask() before comparing.
+ */
+static int check_pos(struct super_block *sb, u64 pos1, u64 pos2, level_t level)
+{
+	return	(pos1 & logfs_block_mask(sb, level)) !=
+		(pos2 & logfs_block_mask(sb, level));
+}
+
+#if 0
+static int read_seg_header(struct super_block *sb, u64 ofs,
+		struct logfs_segment_header *sh)
+{
+	__be32 crc;
+	int err;
+
+	err = wbuf_read(sb, ofs, sizeof(*sh), sh);
+	if (err)
+		return err;
+	crc = logfs_crc32(sh, sizeof(*sh), 4);
+	if (crc != sh->crc) {
+		printk(KERN_ERR"LOGFS: header crc error at %llx: expected %x, "
+				"got %x\n", ofs, be32_to_cpu(sh->crc),
+				be32_to_cpu(crc));
+		return -EIO;
+	}
+	return 0;
+}
+#endif
+
+static int read_obj_header(struct super_block *sb, u64 ofs,
+		struct logfs_object_header *oh)
+{
+	__be32 crc;
+	int err;
+
+	err = wbuf_read(sb, ofs, sizeof(*oh), oh);
+	if (err)
+		return err;
+	crc = logfs_crc32(oh, sizeof(*oh) - 4, 4);
+	if (crc != oh->crc) {
+		printk(KERN_ERR"LOGFS: header crc error at %llx: expected %x, "
+				"got %x\n", ofs, be32_to_cpu(oh->crc),
+				be32_to_cpu(crc));
+		return -EIO;
+	}
+	return 0;
+}
+
+static void move_btree_to_page(struct inode *inode, struct page *page,
+		__be64 *data)
+{
+	struct super_block *sb = inode->i_sb;
+	struct logfs_super *super = logfs_super(sb);
+	struct btree_head128 *head = &super->s_object_alias_tree;
+	struct logfs_block *block;
+	struct object_alias_item *item, *next;
+
+	if (!(super->s_flags & LOGFS_SB_FLAG_OBJ_ALIAS))
+		return;
+
+	block = btree_remove128(head, inode->i_ino, page->index);
+	if (!block)
+		return;
+
+	log_blockmove("move_btree_to_page(%llx, %llx, %x)\n",
+			block->ino, block->bix, block->level);
+	list_for_each_entry_safe(item, next, &block->item_list, list) {
+		data[item->child_no] = item->val;
+		list_del(&item->list);
+		mempool_free(item, super->s_alias_pool);
+	}
+	block->page = page;
+	SetPagePrivate(page);
+	page->private = (unsigned long)block;
+	block->ops = &indirect_block_ops;
+	initialize_block_counters(page, block, data, 0);
+}
+
+/*
+ * This silences a false, yet annoying gcc warning.  I hate it when my editor
+ * jumps into bitops.h each time I recompile this file.
+ * TODO: Complain to gcc folks about this and upgrade compiler.
+ */
+static unsigned long fnb(const unsigned long *addr,
+		unsigned long size, unsigned long offset)
+{
+	return find_next_bit(addr, size, offset);
+}
+
+void move_page_to_btree(struct page *page)
+{
+	struct logfs_block *block = logfs_block(page);
+	struct super_block *sb = block->sb;
+	struct logfs_super *super = logfs_super(sb);
+	struct object_alias_item *item;
+	unsigned long pos;
+	__be64 *child;
+	int err;
+
+	if (super->s_flags & LOGFS_SB_FLAG_SHUTDOWN) {
+		block->ops->free_block(sb, block);
+		return;
+	}
+	log_blockmove("move_page_to_btree(%llx, %llx, %x)\n",
+			block->ino, block->bix, block->level);
+	super->s_flags |= LOGFS_SB_FLAG_OBJ_ALIAS;
+
+	for (pos = 0; ; pos++) {
+		pos = fnb(block->alias_map, LOGFS_BLOCK_FACTOR, pos);
+		if (pos >= LOGFS_BLOCK_FACTOR)
+			break;
+
+		item = mempool_alloc(super->s_alias_pool, GFP_NOFS);
+		BUG_ON(!item); /* mempool empty */
+		memset(item, 0, sizeof(*item));
+
+		child = kmap_atomic(page, KM_USER0);
+		item->val = child[pos];
+		kunmap_atomic(child, KM_USER0);
+		item->child_no = pos;
+		list_add(&item->list, &block->item_list);
+	}
+	block->page = NULL;
+	ClearPagePrivate(page);
+	page->private = 0;
+	block->ops = &btree_block_ops;
+	err = alias_tree_insert(block->sb, block->ino, block->bix, block->level,
+			block);
+	BUG_ON(err); /* mempool empty */
+	ClearPageUptodate(page);
+}
+
+static int __logfs_segment_read(struct inode *inode, void *buf,
+		u64 ofs, u64 bix, level_t level)
+{
+	struct super_block *sb = inode->i_sb;
+	void *compressor_buf = logfs_super(sb)->s_compressed_je;
+	struct logfs_object_header oh;
+	__be32 crc;
+	u16 len;
+	int err, block_len;
+
+	block_len = obj_len(sb, obj_type(inode, level));
+	err = read_obj_header(sb, ofs, &oh);
+	if (err)
+		goto out_err;
+
+	err = -EIO;
+	if (be64_to_cpu(oh.ino) != inode->i_ino
+			|| check_pos(sb, be64_to_cpu(oh.bix), bix, level)) {
+		printk(KERN_ERR"LOGFS: (ino, bix) don't match at %llx: "
+				"expected (%lx, %llx), got (%llx, %llx)\n",
+				ofs, inode->i_ino, bix,
+				be64_to_cpu(oh.ino), be64_to_cpu(oh.bix));
+		goto out_err;
+	}
+
+	len = be16_to_cpu(oh.len);
+
+	switch (oh.compr) {
+	case COMPR_NONE:
+		err = wbuf_read(sb, ofs + LOGFS_OBJECT_HEADERSIZE, len, buf);
+		if (err)
+			goto out_err;
+		crc = logfs_crc32(buf, len, 0);
+		if (crc != oh.data_crc) {
+			printk(KERN_ERR"LOGFS: uncompressed data crc error at "
+					"%llx: expected %x, got %x\n", ofs,
+					be32_to_cpu(oh.data_crc),
+					be32_to_cpu(crc));
+			goto out_err;
+		}
+		break;
+	case COMPR_ZLIB:
+		mutex_lock(&logfs_super(sb)->s_journal_mutex);
+		err = wbuf_read(sb, ofs + LOGFS_OBJECT_HEADERSIZE, len,
+				compressor_buf);
+		if (err) {
+			mutex_unlock(&logfs_super(sb)->s_journal_mutex);
+			goto out_err;
+		}
+		crc = logfs_crc32(compressor_buf, len, 0);
+		if (crc != oh.data_crc) {
+			printk(KERN_ERR"LOGFS: compressed data crc error at "
+					"%llx: expected %x, got %x\n", ofs,
+					be32_to_cpu(oh.data_crc),
+					be32_to_cpu(crc));
+			mutex_unlock(&logfs_super(sb)->s_journal_mutex);
+			goto out_err;
+		}
+		err = logfs_uncompress(compressor_buf, buf, len, block_len);
+		mutex_unlock(&logfs_super(sb)->s_journal_mutex);
+		if (err) {
+			printk(KERN_ERR"LOGFS: uncompress error at %llx\n", ofs);
+			goto out_err;
+		}
+		break;
+	default:
+		LOGFS_BUG(sb);
+		err = -EIO;
+		goto out_err;
+	}
+	return 0;
+
+out_err:
+	logfs_set_ro(sb);
+	printk(KERN_ERR"LOGFS: device is read-only now\n");
+	LOGFS_BUG(sb);
+	return err;
+}
+
+/**
+ * logfs_segment_read - read data block from object store
+ * @inode:		inode containing data
+ * @buf:		data buffer
+ * @ofs:		physical data offset
+ * @bix:		block index
+ * @level:		block level
+ *
+ * Returns 0 on success or a negative errno.
+ */
+int logfs_segment_read(struct inode *inode, struct page *page,
+		u64 ofs, u64 bix, level_t level)
+{
+	int err;
+	void *buf;
+
+	if (PageUptodate(page))
+		return 0;
+
+	ofs &= ~LOGFS_FULLY_POPULATED;
+
+	buf = kmap(page);
+	err = __logfs_segment_read(inode, buf, ofs, bix, level);
+	if (!err) {
+		move_btree_to_page(inode, page, buf);
+		SetPageUptodate(page);
+	}
+	kunmap(page);
+	log_segment("logfs_segment_read(%lx, %llx, %x) %llx (%d)\n",
+			inode->i_ino, bix, level, ofs, err);
+	return err;
+}
+
+int logfs_segment_delete(struct inode *inode, struct logfs_shadow *shadow)
+{
+	struct super_block *sb = inode->i_sb;
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_object_header h;
+	u16 len;
+	int err;
+
+	super->s_flags |= LOGFS_SB_FLAG_DIRTY;
+	BUG_ON(super->s_flags & LOGFS_SB_FLAG_SHUTDOWN);
+	BUG_ON(shadow->old_ofs & LOGFS_FULLY_POPULATED);
+	if (!shadow->old_ofs)
+		return 0;
+
+	log_segment("logfs_segment_delete(%llx, %llx, %x) %llx->%llx %x->%x\n",
+			shadow->ino, shadow->bix, shadow->gc_level,
+			shadow->old_ofs, shadow->new_ofs,
+			shadow->old_len, shadow->new_len);
+	err = read_obj_header(sb, shadow->old_ofs, &h);
+	LOGFS_BUG_ON(err, sb);
+	LOGFS_BUG_ON(be64_to_cpu(h.ino) != inode->i_ino, sb);
+	LOGFS_BUG_ON(check_pos(sb, shadow->bix, be64_to_cpu(h.bix),
+				shrink_level(shadow->gc_level)), sb);
+
+	if (shadow->gc_level == 0)
+		len = be16_to_cpu(h.len);
+	else
+		len = obj_len(sb, h.type);
+	shadow->old_len = len + sizeof(h);
+	return 0;
+}
+
+static void freeseg(struct super_block *sb, u32 segno)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping = super->s_mapping_inode->i_mapping;
+	struct page *page;
+	u64 ofs, start, end;
+
+	start = dev_ofs(sb, segno, 0);
+	end = dev_ofs(sb, segno + 1, 0);
+	for (ofs = start; ofs < end; ofs += PAGE_SIZE) {
+		page = find_get_page(mapping, ofs >> PAGE_SHIFT);
+		if (!page)
+			continue;
+		ClearPagePrivate(page);
+		page_cache_release(page);
+	}
+}
+
+int logfs_open_area(struct logfs_area *area, size_t bytes)
+{
+	struct super_block *sb = area->a_sb;
+	struct logfs_super *super = logfs_super(sb);
+	int err, closed = 0;
+
+	if (area->a_is_open && area->a_used_bytes + bytes <= super->s_segsize)
+		return 0;
+
+	if (area->a_is_open) {
+		u64 ofs = dev_ofs(sb, area->a_segno, area->a_written_bytes);
+		u32 len = super->s_segsize - area->a_written_bytes;
+
+		log_gc("logfs_close_area(%x)\n", area->a_segno);
+		pad_wbuf(area, 1);
+		super->s_devops->writeseg(area->a_sb, ofs, len);
+		freeseg(sb, area->a_segno);
+		closed = 1;
+	}
+
+	area->a_used_bytes = 0;
+	area->a_written_bytes = 0;
+again:
+	area->a_ops->get_free_segment(area);
+	area->a_ops->get_erase_count(area);
+
+	log_gc("logfs_open_area(%x, %x)\n", area->a_segno, area->a_level);
+	err = area->a_ops->erase_segment(area);
+	if (err) {
+		printk(KERN_WARNING "LogFS: Error erasing segment %x\n",
+				area->a_segno);
+		logfs_mark_segment_bad(sb, area->a_segno);
+		goto again;
+	}
+	area->a_is_open = 1;
+	return closed;
+}
+
+void logfs_sync_area(struct logfs_area *area)
+{
+	struct super_block *sb = area->a_sb;
+	struct logfs_super *super = logfs_super(sb);
+	u64 ofs = dev_ofs(sb, area->a_segno, area->a_written_bytes);
+	u32 len = (area->a_used_bytes - area->a_written_bytes);
+
+	if (super->s_writesize)
+		len &= ~(super->s_writesize - 1);
+	if (len == 0)
+		return;
+	pad_wbuf(area, 0);
+	super->s_devops->writeseg(sb, ofs, len);
+	area->a_written_bytes += len;
+}
+
+void logfs_sync_segments(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int i;
+
+	for_each_area(i)
+		logfs_sync_area(super->s_area[i]);
+}
+
+/*
+ * Pick a free segment to be used for this area.  Effectively takes a
+ * candidate from the free list (not really a candidate anymore).
+ */
+static void ostore_get_free_segment(struct logfs_area *area)
+{
+	struct super_block *sb = area->a_sb;
+	struct logfs_super *super = logfs_super(sb);
+
+	if (super->s_free_list.count == 0) {
+		printk(KERN_ERR"LOGFS: ran out of free segments\n");
+		LOGFS_BUG(sb);
+	}
+
+	area->a_segno = get_best_cand(sb, &super->s_free_list, NULL);
+}
+
+static void ostore_get_erase_count(struct logfs_area *area)
+{
+	struct logfs_segment_entry se;
+	u32 ec_level;
+
+	logfs_get_segment_entry(area->a_sb, area->a_segno, &se);
+	BUG_ON(se.ec_level == cpu_to_be32(BADSEG) ||
+			se.valid == cpu_to_be32(RESERVED));
+
+	ec_level = be32_to_cpu(se.ec_level);
+	area->a_erase_count = (ec_level >> 4) + 1;
+}
+
+static int ostore_erase_segment(struct logfs_area *area)
+{
+	struct super_block *sb = area->a_sb;
+	struct logfs_segment_header sh;
+	u64 ofs;
+	int err;
+
+	err = logfs_erase_segment(sb, area->a_segno, 0);
+	if (err)
+		return err;
+
+	sh.pad = 0;
+	sh.type = SEG_OSTORE;
+	sh.level = (__force u8)area->a_level;
+	sh.segno = cpu_to_be32(area->a_segno);
+	sh.ec = cpu_to_be32(area->a_erase_count);
+	sh.gec = cpu_to_be64(logfs_super(sb)->s_gec);
+	sh.crc = logfs_crc32(&sh, sizeof(sh), 4);
+
+	logfs_set_segment_erased(sb, area->a_segno, area->a_erase_count,
+			area->a_level);
+
+	ofs = dev_ofs(sb, area->a_segno, 0);
+	area->a_used_bytes = sizeof(sh);
+	logfs_buf_write(area, ofs, &sh, sizeof(sh));
+	return 0;
+}
+
+static const struct logfs_area_ops ostore_area_ops = {
+	.get_free_segment	= ostore_get_free_segment,
+	.get_erase_count	= ostore_get_erase_count,
+	.erase_segment		= ostore_erase_segment,
+};
+
+static void free_area(struct logfs_area *area)
+{
+	if (area)
+		freeseg(area->a_sb, area->a_segno);
+	kfree(area);
+}
+
+static struct logfs_area *alloc_area(struct super_block *sb)
+{
+	struct logfs_area *area;
+
+	area = kzalloc(sizeof(*area), GFP_KERNEL);
+	if (!area)
+		return NULL;
+
+	area->a_sb = sb;
+	return area;
+}
+
+static void map_invalidatepage(struct page *page, unsigned long l)
+{
+	BUG();
+}
+
+static int map_releasepage(struct page *page, gfp_t g)
+{
+	/* Don't release these pages */
+	return 0;
+}
+
+static const struct address_space_operations mapping_aops = {
+	.invalidatepage = map_invalidatepage,
+	.releasepage	= map_releasepage,
+	.set_page_dirty = __set_page_dirty_nobuffers,
+};
+
+int logfs_init_mapping(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct address_space *mapping;
+	struct inode *inode;
+
+	inode = logfs_new_meta_inode(sb, LOGFS_INO_MAPPING);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+	super->s_mapping_inode = inode;
+	mapping = inode->i_mapping;
+	mapping->a_ops = &mapping_aops;
+	/* Would it be possible to use __GFP_HIGHMEM as well? */
+	mapping_set_gfp_mask(mapping, GFP_NOFS);
+	return 0;
+}
+
+int logfs_init_areas(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int i = -1;
+
+	super->s_alias_pool = mempool_create_kmalloc_pool(600,
+			sizeof(struct object_alias_item));
+	if (!super->s_alias_pool)
+		return -ENOMEM;
+
+	super->s_journal_area = alloc_area(sb);
+	if (!super->s_journal_area)
+		goto err;
+
+	for_each_area(i) {
+		super->s_area[i] = alloc_area(sb);
+		if (!super->s_area[i])
+			goto err;
+		super->s_area[i]->a_level = GC_LEVEL(i);
+		super->s_area[i]->a_ops = &ostore_area_ops;
+	}
+	btree_init_mempool128(&super->s_object_alias_tree,
+			super->s_btree_pool);
+	return 0;
+
+err:
+	for (i--; i >= 0; i--)
+		free_area(super->s_area[i]);
+	free_area(super->s_journal_area);
+	mempool_destroy(super->s_alias_pool);
+	return -ENOMEM;
+}
+
+void logfs_cleanup_areas(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int i;
+
+	btree_grim_visitor128(&super->s_object_alias_tree, 0, kill_alias);
+	for_each_area(i)
+		free_area(super->s_area[i]);
+	free_area(super->s_journal_area);
+	destroy_meta_inode(super->s_mapping_inode);
+}
diff --git a/fs/logfs/super.c b/fs/logfs/super.c
new file mode 100644
index 000000000000..c66beab78dee
--- /dev/null
+++ b/fs/logfs/super.c
@@ -0,0 +1,650 @@
+/*
+ * fs/logfs/super.c
+ *
+ * As should be obvious for Linux kernel code, license is GPLv2
+ *
+ * Copyright (c) 2005-2008 Joern Engel <joern@logfs.org>
+ *
+ * Generally contains mount/umount code and also serves as a dump area for
+ * any functions that don't fit elsewhere and neither justify a file of their
+ * own.
+ */
+#include "logfs.h"
+#include <linux/bio.h>
+#include <linux/mtd/mtd.h>
+#include <linux/statfs.h>
+#include <linux/buffer_head.h>
+
+static DEFINE_MUTEX(emergency_mutex);
+static struct page *emergency_page;
+
+struct page *emergency_read_begin(struct address_space *mapping, pgoff_t index)
+{
+	filler_t *filler = (filler_t *)mapping->a_ops->readpage;
+	struct page *page;
+	int err;
+
+	page = read_cache_page(mapping, index, filler, NULL);
+	if (page)
+		return page;
+
+	/* No more pages available, switch to emergency page */
+	printk(KERN_INFO"Logfs: Using emergency page\n");
+	mutex_lock(&emergency_mutex);
+	err = filler(NULL, emergency_page);
+	if (err) {
+		mutex_unlock(&emergency_mutex);
+		printk(KERN_EMERG"Logfs: Error reading emergency page\n");
+		return ERR_PTR(err);
+	}
+	return emergency_page;
+}
+
+void emergency_read_end(struct page *page)
+{
+	if (page == emergency_page)
+		mutex_unlock(&emergency_mutex);
+	else
+		page_cache_release(page);
+}
+
+static void dump_segfile(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_segment_entry se;
+	u32 segno;
+
+	for (segno = 0; segno < super->s_no_segs; segno++) {
+		logfs_get_segment_entry(sb, segno, &se);
+		printk("%3x: %6x %8x", segno, be32_to_cpu(se.ec_level),
+				be32_to_cpu(se.valid));
+		if (++segno < super->s_no_segs) {
+			logfs_get_segment_entry(sb, segno, &se);
+			printk(" %6x %8x", be32_to_cpu(se.ec_level),
+					be32_to_cpu(se.valid));
+		}
+		if (++segno < super->s_no_segs) {
+			logfs_get_segment_entry(sb, segno, &se);
+			printk(" %6x %8x", be32_to_cpu(se.ec_level),
+					be32_to_cpu(se.valid));
+		}
+		if (++segno < super->s_no_segs) {
+			logfs_get_segment_entry(sb, segno, &se);
+			printk(" %6x %8x", be32_to_cpu(se.ec_level),
+					be32_to_cpu(se.valid));
+		}
+		printk("\n");
+	}
+}
+
+/*
+ * logfs_crash_dump - dump debug information to device
+ *
+ * The LogFS superblock only occupies part of a segment.  This function will
+ * write as much debug information as it can gather into the spare space.
+ */
+void logfs_crash_dump(struct super_block *sb)
+{
+	dump_segfile(sb);
+}
+
+/*
+ * TODO: move to lib/string.c
+ */
+/**
+ * memchr_inv - Find a character in an area of memory.
+ * @s: The memory area
+ * @c: The byte to search for
+ * @n: The size of the area.
+ *
+ * returns the address of the first character other than @c, or %NULL
+ * if the whole buffer contains just @c.
+ */
+void *memchr_inv(const void *s, int c, size_t n)
+{
+	const unsigned char *p = s;
+	while (n-- != 0)
+		if ((unsigned char)c != *p++)
+			return (void *)(p - 1);
+
+	return NULL;
+}
+
+/*
+ * FIXME: There should be a reserve for root, similar to ext2.
+ */
+int logfs_statfs(struct dentry *dentry, struct kstatfs *stats)
+{
+	struct super_block *sb = dentry->d_sb;
+	struct logfs_super *super = logfs_super(sb);
+
+	stats->f_type		= LOGFS_MAGIC_U32;
+	stats->f_bsize		= sb->s_blocksize;
+	stats->f_blocks		= super->s_size >> LOGFS_BLOCK_BITS >> 3;
+	stats->f_bfree		= super->s_free_bytes >> sb->s_blocksize_bits;
+	stats->f_bavail		= super->s_free_bytes >> sb->s_blocksize_bits;
+	stats->f_files		= 0;
+	stats->f_ffree		= 0;
+	stats->f_namelen	= LOGFS_MAX_NAMELEN;
+	return 0;
+}
+
+static int logfs_sb_set(struct super_block *sb, void *_super)
+{
+	struct logfs_super *super = _super;
+
+	sb->s_fs_info = super;
+	sb->s_mtd = super->s_mtd;
+	sb->s_bdev = super->s_bdev;
+	return 0;
+}
+
+static int logfs_sb_test(struct super_block *sb, void *_super)
+{
+	struct logfs_super *super = _super;
+	struct mtd_info *mtd = super->s_mtd;
+
+	if (mtd && sb->s_mtd == mtd)
+		return 1;
+	if (super->s_bdev && sb->s_bdev == super->s_bdev)
+		return 1;
+	return 0;
+}
+
+static void set_segment_header(struct logfs_segment_header *sh, u8 type,
+		u8 level, u32 segno, u32 ec)
+{
+	sh->pad = 0;
+	sh->type = type;
+	sh->level = level;
+	sh->segno = cpu_to_be32(segno);
+	sh->ec = cpu_to_be32(ec);
+	sh->gec = cpu_to_be64(segno);
+	sh->crc = logfs_crc32(sh, LOGFS_SEGMENT_HEADERSIZE, 4);
+}
+
+static void logfs_write_ds(struct super_block *sb, struct logfs_disk_super *ds,
+		u32 segno, u32 ec)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_segment_header *sh = &ds->ds_sh;
+	int i;
+
+	memset(ds, 0, sizeof(*ds));
+	set_segment_header(sh, SEG_SUPER, 0, segno, ec);
+
+	ds->ds_ifile_levels	= super->s_ifile_levels;
+	ds->ds_iblock_levels	= super->s_iblock_levels;
+	ds->ds_data_levels	= super->s_data_levels; /* XXX: Remove */
+	ds->ds_segment_shift	= super->s_segshift;
+	ds->ds_block_shift	= sb->s_blocksize_bits;
+	ds->ds_write_shift	= super->s_writeshift;
+	ds->ds_filesystem_size	= cpu_to_be64(super->s_size);
+	ds->ds_segment_size	= cpu_to_be32(super->s_segsize);
+	ds->ds_bad_seg_reserve	= cpu_to_be32(super->s_bad_seg_reserve);
+	ds->ds_feature_incompat	= cpu_to_be64(super->s_feature_incompat);
+	ds->ds_feature_ro_compat= cpu_to_be64(super->s_feature_ro_compat);
+	ds->ds_feature_compat	= cpu_to_be64(super->s_feature_compat);
+	ds->ds_feature_flags	= cpu_to_be64(super->s_feature_flags);
+	ds->ds_root_reserve	= cpu_to_be64(super->s_root_reserve);
+	ds->ds_speed_reserve	= cpu_to_be64(super->s_speed_reserve);
+	journal_for_each(i)
+		ds->ds_journal_seg[i] = cpu_to_be32(super->s_journal_seg[i]);
+	ds->ds_magic		= cpu_to_be64(LOGFS_MAGIC);
+	ds->ds_crc = logfs_crc32(ds, sizeof(*ds),
+			LOGFS_SEGMENT_HEADERSIZE + 12);
+}
+
+static int write_one_sb(struct super_block *sb,
+		struct page *(*find_sb)(struct super_block *sb, u64 *ofs))
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_disk_super *ds;
+	struct logfs_segment_entry se;
+	struct page *page;
+	u64 ofs;
+	u32 ec, segno;
+	int err;
+
+	page = find_sb(sb, &ofs);
+	if (!page)
+		return -EIO;
+	ds = page_address(page);
+	segno = seg_no(sb, ofs);
+	logfs_get_segment_entry(sb, segno, &se);
+	ec = be32_to_cpu(se.ec_level) >> 4;
+	ec++;
+	logfs_set_segment_erased(sb, segno, ec, 0);
+	logfs_write_ds(sb, ds, segno, ec);
+	err = super->s_devops->write_sb(sb, page);
+	page_cache_release(page);
+	return err;
+}
+
+int logfs_write_sb(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int err;
+
+	/* First superblock */
+	err = write_one_sb(sb, super->s_devops->find_first_sb);
+	if (err)
+		return err;
+
+	/* Last superblock */
+	err = write_one_sb(sb, super->s_devops->find_last_sb);
+	if (err)
+		return err;
+	return 0;
+}
+
+static int ds_cmp(const void *ds0, const void *ds1)
+{
+	size_t len = sizeof(struct logfs_disk_super);
+
+	/* We know the segment headers differ, so ignore them */
+	len -= LOGFS_SEGMENT_HEADERSIZE;
+	ds0 += LOGFS_SEGMENT_HEADERSIZE;
+	ds1 += LOGFS_SEGMENT_HEADERSIZE;
+	return memcmp(ds0, ds1, len);
+}
+
+static int logfs_recover_sb(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct logfs_disk_super _ds0, *ds0 = &_ds0;
+	struct logfs_disk_super _ds1, *ds1 = &_ds1;
+	int err, valid0, valid1;
+
+	/* read first superblock */
+	err = wbuf_read(sb, super->s_sb_ofs[0], sizeof(*ds0), ds0);
+	if (err)
+		return err;
+	/* read last superblock */
+	err = wbuf_read(sb, super->s_sb_ofs[1], sizeof(*ds1), ds1);
+	if (err)
+		return err;
+	valid0 = logfs_check_ds(ds0) == 0;
+	valid1 = logfs_check_ds(ds1) == 0;
+
+	if (!valid0 && valid1) {
+		printk(KERN_INFO"First superblock is invalid - fixing.\n");
+		return write_one_sb(sb, super->s_devops->find_first_sb);
+	}
+	if (valid0 && !valid1) {
+		printk(KERN_INFO"Last superblock is invalid - fixing.\n");
+		return write_one_sb(sb, super->s_devops->find_last_sb);
+	}
+	if (valid0 && valid1 && ds_cmp(ds0, ds1)) {
+		printk(KERN_INFO"Superblocks don't match - fixing.\n");
+		return write_one_sb(sb, super->s_devops->find_last_sb);
+	}
+	/* If neither is valid now, something's wrong.  Didn't we properly
+	 * check them before?!? */
+	BUG_ON(!valid0 && !valid1);
+	return 0;
+}
+
+static int logfs_make_writeable(struct super_block *sb)
+{
+	int err;
+
+	/* Repair any broken superblock copies */
+	err = logfs_recover_sb(sb);
+	if (err)
+		return err;
+
+	/* Check areas for trailing unaccounted data */
+	err = logfs_check_areas(sb);
+	if (err)
+		return err;
+
+	err = logfs_open_segfile(sb);
+	if (err)
+		return err;
+
+	/* Do one GC pass before any data gets dirtied */
+	logfs_gc_pass(sb);
+
+	/* after all initializations are done, replay the journal
+	 * for rw-mounts, if necessary */
+	err = logfs_replay_journal(sb);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+static int logfs_get_sb_final(struct super_block *sb, struct vfsmount *mnt)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct inode *rootdir;
+	int err;
+
+	/* root dir */
+	rootdir = logfs_iget(sb, LOGFS_INO_ROOT);
+	if (IS_ERR(rootdir))
+		goto fail;
+
+	sb->s_root = d_alloc_root(rootdir);
+	if (!sb->s_root)
+		goto fail;
+
+	super->s_erase_page = alloc_pages(GFP_KERNEL, 0);
+	if (!super->s_erase_page)
+		goto fail2;
+	memset(page_address(super->s_erase_page), 0xFF, PAGE_SIZE);
+
+	/* FIXME: check for read-only mounts */
+	err = logfs_make_writeable(sb);
+	if (err)
+		goto fail3;
+
+	log_super("LogFS: Finished mounting\n");
+	simple_set_mnt(mnt, sb);
+	return 0;
+
+fail3:
+	__free_page(super->s_erase_page);
+fail2:
+	iput(rootdir);
+fail:
+	iput(logfs_super(sb)->s_master_inode);
+	return -EIO;
+}
+
+int logfs_check_ds(struct logfs_disk_super *ds)
+{
+	struct logfs_segment_header *sh = &ds->ds_sh;
+
+	if (ds->ds_magic != cpu_to_be64(LOGFS_MAGIC))
+		return -EINVAL;
+	if (sh->crc != logfs_crc32(sh, LOGFS_SEGMENT_HEADERSIZE, 4))
+		return -EINVAL;
+	if (ds->ds_crc != logfs_crc32(ds, sizeof(*ds),
+				LOGFS_SEGMENT_HEADERSIZE + 12))
+		return -EINVAL;
+	return 0;
+}
+
+static struct page *find_super_block(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct page *first, *last;
+
+	first = super->s_devops->find_first_sb(sb, &super->s_sb_ofs[0]);
+	if (!first || IS_ERR(first))
+		return NULL;
+	last = super->s_devops->find_last_sb(sb, &super->s_sb_ofs[1]);
+	if (!last || IS_ERR(first)) {
+		page_cache_release(first);
+		return NULL;
+	}
+
+	if (!logfs_check_ds(page_address(first))) {
+		page_cache_release(last);
+		return first;
+	}
+
+	/* First one didn't work, try the second superblock */
+	if (!logfs_check_ds(page_address(last))) {
+		page_cache_release(first);
+		return last;
+	}
+
+	/* Neither worked, sorry folks */
+	page_cache_release(first);
+	page_cache_release(last);
+	return NULL;
+}
+
+static int __logfs_read_sb(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+	struct page *page;
+	struct logfs_disk_super *ds;
+	int i;
+
+	page = find_super_block(sb);
+	if (!page)
+		return -EIO;
+
+	ds = page_address(page);
+	super->s_size = be64_to_cpu(ds->ds_filesystem_size);
+	super->s_root_reserve = be64_to_cpu(ds->ds_root_reserve);
+	super->s_speed_reserve = be64_to_cpu(ds->ds_speed_reserve);
+	super->s_bad_seg_reserve = be32_to_cpu(ds->ds_bad_seg_reserve);
+	super->s_segsize = 1 << ds->ds_segment_shift;
+	super->s_segmask = (1 << ds->ds_segment_shift) - 1;
+	super->s_segshift = ds->ds_segment_shift;
+	sb->s_blocksize = 1 << ds->ds_block_shift;
+	sb->s_blocksize_bits = ds->ds_block_shift;
+	super->s_writesize = 1 << ds->ds_write_shift;
+	super->s_writeshift = ds->ds_write_shift;
+	super->s_no_segs = super->s_size >> super->s_segshift;
+	super->s_no_blocks = super->s_segsize >> sb->s_blocksize_bits;
+	super->s_feature_incompat = be64_to_cpu(ds->ds_feature_incompat);
+	super->s_feature_ro_compat = be64_to_cpu(ds->ds_feature_ro_compat);
+	super->s_feature_compat = be64_to_cpu(ds->ds_feature_compat);
+	super->s_feature_flags = be64_to_cpu(ds->ds_feature_flags);
+
+	journal_for_each(i)
+		super->s_journal_seg[i] = be32_to_cpu(ds->ds_journal_seg[i]);
+
+	super->s_ifile_levels = ds->ds_ifile_levels;
+	super->s_iblock_levels = ds->ds_iblock_levels;
+	super->s_data_levels = ds->ds_data_levels;
+	super->s_total_levels = super->s_ifile_levels + super->s_iblock_levels
+		+ super->s_data_levels;
+	page_cache_release(page);
+	return 0;
+}
+
+static int logfs_read_sb(struct super_block *sb, int read_only)
+{
+	struct logfs_super *super = logfs_super(sb);
+	int ret;
+
+	super->s_btree_pool = mempool_create(32, btree_alloc, btree_free, NULL);
+	if (!super->s_btree_pool)
+		return -ENOMEM;
+
+	btree_init_mempool64(&super->s_shadow_tree.new, super->s_btree_pool);
+	btree_init_mempool64(&super->s_shadow_tree.old, super->s_btree_pool);
+
+	ret = logfs_init_mapping(sb);
+	if (ret)
+		return ret;
+
+	ret = __logfs_read_sb(sb);
+	if (ret)
+		return ret;
+
+	if (super->s_feature_incompat & ~LOGFS_FEATURES_INCOMPAT)
+		return -EIO;
+	if ((super->s_feature_ro_compat & ~LOGFS_FEATURES_RO_COMPAT) &&
+			!read_only)
+		return -EIO;
+
+	mutex_init(&super->s_dirop_mutex);
+	mutex_init(&super->s_object_alias_mutex);
+	INIT_LIST_HEAD(&super->s_freeing_list);
+
+	ret = logfs_init_rw(sb);
+	if (ret)
+		return ret;
+
+	ret = logfs_init_areas(sb);
+	if (ret)
+		return ret;
+
+	ret = logfs_init_gc(sb);
+	if (ret)
+		return ret;
+
+	ret = logfs_init_journal(sb);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static void logfs_kill_sb(struct super_block *sb)
+{
+	struct logfs_super *super = logfs_super(sb);
+
+	log_super("LogFS: Start unmounting\n");
+	/* Alias entries slow down mount, so evict as many as possible */
+	sync_filesystem(sb);
+	logfs_write_anchor(sb);
+
+	/*
+	 * From this point on alias entries are simply dropped - and any
+	 * writes to the object store are considered bugs.
+	 */
+	super->s_flags |= LOGFS_SB_FLAG_SHUTDOWN;
+	log_super("LogFS: Now in shutdown\n");
+	generic_shutdown_super(sb);
+
+	BUG_ON(super->s_dirty_used_bytes || super->s_dirty_free_bytes);
+
+	logfs_cleanup_gc(sb);
+	logfs_cleanup_journal(sb);
+	logfs_cleanup_areas(sb);
+	logfs_cleanup_rw(sb);
+	if (super->s_erase_page)
+		__free_page(super->s_erase_page);
+	super->s_devops->put_device(sb);
+	mempool_destroy(super->s_btree_pool);
+	mempool_destroy(super->s_alias_pool);
+	kfree(super);
+	log_super("LogFS: Finished unmounting\n");
+}
+
+int logfs_get_sb_device(struct file_system_type *type, int flags,
+		struct mtd_info *mtd, struct block_device *bdev,
+		const struct logfs_device_ops *devops, struct vfsmount *mnt)
+{
+	struct logfs_super *super;
+	struct super_block *sb;
+	int err = -ENOMEM;
+	static int mount_count;
+
+	log_super("LogFS: Start mount %x\n", mount_count++);
+	super = kzalloc(sizeof(*super), GFP_KERNEL);
+	if (!super)
+		goto err0;
+
+	super->s_mtd	= mtd;
+	super->s_bdev	= bdev;
+	err = -EINVAL;
+	sb = sget(type, logfs_sb_test, logfs_sb_set, super);
+	if (IS_ERR(sb))
+		goto err0;
+
+	if (sb->s_root) {
+		/* Device is already in use */
+		err = 0;
+		simple_set_mnt(mnt, sb);
+		goto err0;
+	}
+
+	super->s_devops = devops;
+
+	/*
+	 * sb->s_maxbytes is limited to 8TB.  On 32bit systems, the page cache
+	 * only covers 16TB and the upper 8TB are used for indirect blocks.
+	 * On 64bit system we could bump up the limit, but that would make
+	 * the filesystem incompatible with 32bit systems.
+	 */
+	sb->s_maxbytes	= (1ull << 43) - 1;
+	sb->s_op	= &logfs_super_operations;
+	sb->s_flags	= flags | MS_NOATIME;
+
+	err = logfs_read_sb(sb, sb->s_flags & MS_RDONLY);
+	if (err)
+		goto err1;
+
+	sb->s_flags |= MS_ACTIVE;
+	err = logfs_get_sb_final(sb, mnt);
+	if (err)
+		goto err1;
+	return 0;
+
+err1:
+	up_write(&sb->s_umount);
+	deactivate_super(sb);
+	return err;
+err0:
+	kfree(super);
+	//devops->put_device(sb);
+	return err;
+}
+
+static int logfs_get_sb(struct file_system_type *type, int flags,
+		const char *devname, void *data, struct vfsmount *mnt)
+{
+	ulong mtdnr;
+
+	if (!devname)
+		return logfs_get_sb_bdev(type, flags, devname, mnt);
+	if (strncmp(devname, "mtd", 3))
+		return logfs_get_sb_bdev(type, flags, devname, mnt);
+
+	{
+		char *garbage;
+		mtdnr = simple_strtoul(devname+3, &garbage, 0);
+		if (*garbage)
+			return -EINVAL;
+	}
+
+	return logfs_get_sb_mtd(type, flags, mtdnr, mnt);
+}
+
+static struct file_system_type logfs_fs_type = {
+	.owner		= THIS_MODULE,
+	.name		= "logfs",
+	.get_sb		= logfs_get_sb,
+	.kill_sb	= logfs_kill_sb,
+	.fs_flags	= FS_REQUIRES_DEV,
+
+};
+
+static int __init logfs_init(void)
+{
+	int ret;
+
+	emergency_page = alloc_pages(GFP_KERNEL, 0);
+	if (!emergency_page)
+		return -ENOMEM;
+
+	ret = logfs_compr_init();
+	if (ret)
+		goto out1;
+
+	ret = logfs_init_inode_cache();
+	if (ret)
+		goto out2;
+
+	return register_filesystem(&logfs_fs_type);
+out2:
+	logfs_compr_exit();
+out1:
+	__free_pages(emergency_page, 0);
+	return ret;
+}
+
+static void __exit logfs_exit(void)
+{
+	unregister_filesystem(&logfs_fs_type);
+	logfs_destroy_inode_cache();
+	logfs_compr_exit();
+	__free_pages(emergency_page, 0);
+}
+
+module_init(logfs_init);
+module_exit(logfs_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Joern Engel <joern@logfs.org>");
+MODULE_DESCRIPTION("scalable flash filesystem");
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index 74ea82d72164..756f8c93780c 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -17,8 +17,10 @@
 #include <linux/init.h>
 #include <linux/highuid.h>
 #include <linux/vfs.h>
+#include <linux/writeback.h>
 
-static int minix_write_inode(struct inode * inode, int wait);
+static int minix_write_inode(struct inode *inode,
+		struct writeback_control *wbc);
 static int minix_statfs(struct dentry *dentry, struct kstatfs *buf);
 static int minix_remount (struct super_block * sb, int * flags, char * data);
 
@@ -552,7 +554,7 @@ static struct buffer_head * V2_minix_update_inode(struct inode * inode)
 	return bh;
 }
 
-static int minix_write_inode(struct inode *inode, int wait)
+static int minix_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	int err = 0;
 	struct buffer_head *bh;
@@ -563,7 +565,7 @@ static int minix_write_inode(struct inode *inode, int wait)
 		bh = V2_minix_update_inode(inode);
 	if (!bh)
 		return -EIO;
-	if (wait && buffer_dirty(bh)) {
+	if (wbc->sync_mode == WB_SYNC_ALL && buffer_dirty(bh)) {
 		sync_dirty_buffer(bh);
 		if (buffer_req(bh) && !buffer_uptodate(bh)) {
 			printk("IO error syncing minix inode [%s:%08lx]\n",
diff --git a/fs/namei.c b/fs/namei.c
index d62fdc875f22..48e60a187325 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -19,7 +19,6 @@
 #include <linux/slab.h>
 #include <linux/fs.h>
 #include <linux/namei.h>
-#include <linux/quotaops.h>
 #include <linux/pagemap.h>
 #include <linux/fsnotify.h>
 #include <linux/personality.h>
@@ -498,8 +497,6 @@ static int link_path_walk(const char *, struct nameidata *);
 
 static __always_inline int __vfs_follow_link(struct nameidata *nd, const char *link)
 {
-	int res = 0;
-	char *name;
 	if (IS_ERR(link))
 		goto fail;
 
@@ -510,22 +507,7 @@ static __always_inline int __vfs_follow_link(struct nameidata *nd, const char *l
 		path_get(&nd->root);
 	}
 
-	res = link_path_walk(link, nd);
-	if (nd->depth || res || nd->last_type!=LAST_NORM)
-		return res;
-	/*
-	 * If it is an iterative symlinks resolution in open_namei() we
-	 * have to copy the last component. And all that crap because of
-	 * bloody create() on broken symlinks. Furrfu...
-	 */
-	name = __getname();
-	if (unlikely(!name)) {
-		path_put(&nd->path);
-		return -ENOMEM;
-	}
-	strcpy(name, nd->last.name);
-	nd->last.name = name;
-	return 0;
+	return link_path_walk(link, nd);
 fail:
 	path_put(&nd->path);
 	return PTR_ERR(link);
@@ -547,10 +529,10 @@ static inline void path_to_nameidata(struct path *path, struct nameidata *nd)
 	nd->path.dentry = path->dentry;
 }
 
-static __always_inline int __do_follow_link(struct path *path, struct nameidata *nd)
+static __always_inline int
+__do_follow_link(struct path *path, struct nameidata *nd, void **p)
 {
 	int error;
-	void *cookie;
 	struct dentry *dentry = path->dentry;
 
 	touch_atime(path->mnt, dentry);
@@ -562,9 +544,9 @@ static __always_inline int __do_follow_link(struct path *path, struct nameidata
 	}
 	mntget(path->mnt);
 	nd->last_type = LAST_BIND;
-	cookie = dentry->d_inode->i_op->follow_link(dentry, nd);
-	error = PTR_ERR(cookie);
-	if (!IS_ERR(cookie)) {
+	*p = dentry->d_inode->i_op->follow_link(dentry, nd);
+	error = PTR_ERR(*p);
+	if (!IS_ERR(*p)) {
 		char *s = nd_get_link(nd);
 		error = 0;
 		if (s)
@@ -574,8 +556,6 @@ static __always_inline int __do_follow_link(struct path *path, struct nameidata
 			if (error)
 				path_put(&nd->path);
 		}
-		if (dentry->d_inode->i_op->put_link)
-			dentry->d_inode->i_op->put_link(dentry, nd, cookie);
 	}
 	return error;
 }
@@ -589,6 +569,7 @@ static __always_inline int __do_follow_link(struct path *path, struct nameidata
  */
 static inline int do_follow_link(struct path *path, struct nameidata *nd)
 {
+	void *cookie;
 	int err = -ELOOP;
 	if (current->link_count >= MAX_NESTED_LINKS)
 		goto loop;
@@ -602,7 +583,9 @@ static inline int do_follow_link(struct path *path, struct nameidata *nd)
 	current->link_count++;
 	current->total_link_count++;
 	nd->depth++;
-	err = __do_follow_link(path, nd);
+	err = __do_follow_link(path, nd, &cookie);
+	if (!IS_ERR(cookie) && path->dentry->d_inode->i_op->put_link)
+		path->dentry->d_inode->i_op->put_link(path->dentry, nd, cookie);
 	path_put(path);
 	current->link_count--;
 	nd->depth--;
@@ -689,33 +672,20 @@ static __always_inline void follow_dotdot(struct nameidata *nd)
 	set_root(nd);
 
 	while(1) {
-		struct vfsmount *parent;
 		struct dentry *old = nd->path.dentry;
 
 		if (nd->path.dentry == nd->root.dentry &&
 		    nd->path.mnt == nd->root.mnt) {
 			break;
 		}
-		spin_lock(&dcache_lock);
 		if (nd->path.dentry != nd->path.mnt->mnt_root) {
-			nd->path.dentry = dget(nd->path.dentry->d_parent);
-			spin_unlock(&dcache_lock);
+			/* rare case of legitimate dget_parent()... */
+			nd->path.dentry = dget_parent(nd->path.dentry);
 			dput(old);
 			break;
 		}
-		spin_unlock(&dcache_lock);
-		spin_lock(&vfsmount_lock);
-		parent = nd->path.mnt->mnt_parent;
-		if (parent == nd->path.mnt) {
-			spin_unlock(&vfsmount_lock);
+		if (!follow_up(&nd->path))
 			break;
-		}
-		mntget(parent);
-		nd->path.dentry = dget(nd->path.mnt->mnt_mountpoint);
-		spin_unlock(&vfsmount_lock);
-		dput(old);
-		mntput(nd->path.mnt);
-		nd->path.mnt = parent;
 	}
 	follow_mount(&nd->path);
 }
@@ -823,6 +793,17 @@ fail:
 }
 
 /*
+ * This is a temporary kludge to deal with "automount" symlinks; proper
+ * solution is to trigger them on follow_mount(), so that do_lookup()
+ * would DTRT.  To be killed before 2.6.34-final.
+ */
+static inline int follow_on_final(struct inode *inode, unsigned lookup_flags)
+{
+	return inode && unlikely(inode->i_op->follow_link) &&
+		((lookup_flags & LOOKUP_FOLLOW) || S_ISDIR(inode->i_mode));
+}
+
+/*
  * Name resolution.
  * This is the basic name resolution function, turning a pathname into
  * the final dentry. We expect 'base' to be positive and a directory.
@@ -942,8 +923,7 @@ last_component:
 		if (err)
 			break;
 		inode = next.dentry->d_inode;
-		if ((lookup_flags & LOOKUP_FOLLOW)
-		    && inode && inode->i_op->follow_link) {
+		if (follow_on_final(inode, lookup_flags)) {
 			err = do_follow_link(&next, nd);
 			if (err)
 				goto return_err;
@@ -1337,7 +1317,7 @@ static int may_delete(struct inode *dir,struct dentry *victim,int isdir)
 		return -ENOENT;
 
 	BUG_ON(victim->d_parent->d_inode != dir);
-	audit_inode_child(victim->d_name.name, victim, dir);
+	audit_inode_child(victim, dir);
 
 	error = inode_permission(dir, MAY_WRITE | MAY_EXEC);
 	if (error)
@@ -1378,22 +1358,6 @@ static inline int may_create(struct inode *dir, struct dentry *child)
 	return inode_permission(dir, MAY_WRITE | MAY_EXEC);
 }
 
-/* 
- * O_DIRECTORY translates into forcing a directory lookup.
- */
-static inline int lookup_flags(unsigned int f)
-{
-	unsigned long retval = LOOKUP_FOLLOW;
-
-	if (f & O_NOFOLLOW)
-		retval &= ~LOOKUP_FOLLOW;
-	
-	if (f & O_DIRECTORY)
-		retval |= LOOKUP_DIRECTORY;
-
-	return retval;
-}
-
 /*
  * p1 and p2 should be directories on the same fs.
  */
@@ -1451,7 +1415,6 @@ int vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	error = security_inode_create(dir, dentry, mode);
 	if (error)
 		return error;
-	vfs_dq_init(dir);
 	error = dir->i_op->create(dir, dentry, mode, nd);
 	if (!error)
 		fsnotify_create(dir, dentry);
@@ -1493,7 +1456,7 @@ int may_open(struct path *path, int acc_mode, int flag)
 	 * An append-only file must be opened in append mode for writing.
 	 */
 	if (IS_APPEND(inode)) {
-		if  ((flag & FMODE_WRITE) && !(flag & O_APPEND))
+		if  ((flag & O_ACCMODE) != O_RDONLY && !(flag & O_APPEND))
 			return -EPERM;
 		if (flag & O_TRUNC)
 			return -EPERM;
@@ -1537,7 +1500,7 @@ static int handle_truncate(struct path *path)
  * what get passed to sys_open().
  */
 static int __open_namei_create(struct nameidata *nd, struct path *path,
-				int flag, int mode)
+				int open_flag, int mode)
 {
 	int error;
 	struct dentry *dir = nd->path.dentry;
@@ -1555,7 +1518,7 @@ out_unlock:
 	if (error)
 		return error;
 	/* Don't check for write permission, don't truncate */
-	return may_open(&nd->path, 0, flag & ~O_TRUNC);
+	return may_open(&nd->path, 0, open_flag & ~O_TRUNC);
 }
 
 /*
@@ -1593,129 +1556,132 @@ static int open_will_truncate(int flag, struct inode *inode)
 	return (flag & O_TRUNC);
 }
 
-/*
- * Note that the low bits of the passed in "open_flag"
- * are not the same as in the local variable "flag". See
- * open_to_namei_flags() for more details.
- */
-struct file *do_filp_open(int dfd, const char *pathname,
-		int open_flag, int mode, int acc_mode)
+static struct file *finish_open(struct nameidata *nd,
+				int open_flag, int acc_mode)
 {
 	struct file *filp;
-	struct nameidata nd;
-	int error;
-	struct path path;
-	struct dentry *dir;
-	int count = 0;
 	int will_truncate;
-	int flag = open_to_namei_flags(open_flag);
-	int force_reval = 0;
+	int error;
 
+	will_truncate = open_will_truncate(open_flag, nd->path.dentry->d_inode);
+	if (will_truncate) {
+		error = mnt_want_write(nd->path.mnt);
+		if (error)
+			goto exit;
+	}
+	error = may_open(&nd->path, acc_mode, open_flag);
+	if (error) {
+		if (will_truncate)
+			mnt_drop_write(nd->path.mnt);
+		goto exit;
+	}
+	filp = nameidata_to_filp(nd);
+	if (!IS_ERR(filp)) {
+		error = ima_file_check(filp, acc_mode);
+		if (error) {
+			fput(filp);
+			filp = ERR_PTR(error);
+		}
+	}
+	if (!IS_ERR(filp)) {
+		if (will_truncate) {
+			error = handle_truncate(&nd->path);
+			if (error) {
+				fput(filp);
+				filp = ERR_PTR(error);
+			}
+		}
+	}
 	/*
-	 * O_SYNC is implemented as __O_SYNC|O_DSYNC.  As many places only
-	 * check for O_DSYNC if the need any syncing at all we enforce it's
-	 * always set instead of having to deal with possibly weird behaviour
-	 * for malicious applications setting only __O_SYNC.
+	 * It is now safe to drop the mnt write
+	 * because the filp has had a write taken
+	 * on its behalf.
 	 */
-	if (open_flag & __O_SYNC)
-		open_flag |= O_DSYNC;
-
-	if (!acc_mode)
-		acc_mode = MAY_OPEN | ACC_MODE(open_flag);
+	if (will_truncate)
+		mnt_drop_write(nd->path.mnt);
+	return filp;
 
-	/* O_TRUNC implies we need access checks for write permissions */
-	if (flag & O_TRUNC)
-		acc_mode |= MAY_WRITE;
+exit:
+	if (!IS_ERR(nd->intent.open.file))
+		release_open_intent(nd);
+	path_put(&nd->path);
+	return ERR_PTR(error);
+}
 
-	/* Allow the LSM permission hook to distinguish append 
-	   access from general write access. */
-	if (flag & O_APPEND)
-		acc_mode |= MAY_APPEND;
+static struct file *do_last(struct nameidata *nd, struct path *path,
+			    int open_flag, int acc_mode,
+			    int mode, const char *pathname,
+			    int *want_dir)
+{
+	struct dentry *dir = nd->path.dentry;
+	struct file *filp;
+	int error = -EISDIR;
 
-	/*
-	 * The simplest case - just a plain lookup.
-	 */
-	if (!(flag & O_CREAT)) {
-		filp = get_empty_filp();
-
-		if (filp == NULL)
-			return ERR_PTR(-ENFILE);
-		nd.intent.open.file = filp;
-		filp->f_flags = open_flag;
-		nd.intent.open.flags = flag;
-		nd.intent.open.create_mode = 0;
-		error = do_path_lookup(dfd, pathname,
-					lookup_flags(flag)|LOOKUP_OPEN, &nd);
-		if (IS_ERR(nd.intent.open.file)) {
-			if (error == 0) {
-				error = PTR_ERR(nd.intent.open.file);
-				path_put(&nd.path);
+	switch (nd->last_type) {
+	case LAST_DOTDOT:
+		follow_dotdot(nd);
+		dir = nd->path.dentry;
+		if (nd->path.mnt->mnt_sb->s_type->fs_flags & FS_REVAL_DOT) {
+			if (!dir->d_op->d_revalidate(dir, nd)) {
+				error = -ESTALE;
+				goto exit;
 			}
-		} else if (error)
-			release_open_intent(&nd);
-		if (error)
-			return ERR_PTR(error);
+		}
+		/* fallthrough */
+	case LAST_DOT:
+	case LAST_ROOT:
+		if (open_flag & O_CREAT)
+			goto exit;
+		/* fallthrough */
+	case LAST_BIND:
+		audit_inode(pathname, dir);
 		goto ok;
 	}
 
-	/*
-	 * Create - we need to know the parent.
-	 */
-reval:
-	error = path_init(dfd, pathname, LOOKUP_PARENT, &nd);
-	if (error)
-		return ERR_PTR(error);
-	if (force_reval)
-		nd.flags |= LOOKUP_REVAL;
-	error = path_walk(pathname, &nd);
-	if (error) {
-		if (nd.root.mnt)
-			path_put(&nd.root);
-		return ERR_PTR(error);
+	/* trailing slashes? */
+	if (nd->last.name[nd->last.len]) {
+		if (open_flag & O_CREAT)
+			goto exit;
+		*want_dir = 1;
 	}
-	if (unlikely(!audit_dummy_context()))
-		audit_inode(pathname, nd.path.dentry);
 
-	/*
-	 * We have the parent and last component. First of all, check
-	 * that we are not asked to creat(2) an obvious directory - that
-	 * will not do.
-	 */
-	error = -EISDIR;
-	if (nd.last_type != LAST_NORM || nd.last.name[nd.last.len])
-		goto exit_parent;
+	/* just plain open? */
+	if (!(open_flag & O_CREAT)) {
+		error = do_lookup(nd, &nd->last, path);
+		if (error)
+			goto exit;
+		error = -ENOENT;
+		if (!path->dentry->d_inode)
+			goto exit_dput;
+		if (path->dentry->d_inode->i_op->follow_link)
+			return NULL;
+		error = -ENOTDIR;
+		if (*want_dir && !path->dentry->d_inode->i_op->lookup)
+			goto exit_dput;
+		path_to_nameidata(path, nd);
+		audit_inode(pathname, nd->path.dentry);
+		goto ok;
+	}
 
-	error = -ENFILE;
-	filp = get_empty_filp();
-	if (filp == NULL)
-		goto exit_parent;
-	nd.intent.open.file = filp;
-	filp->f_flags = open_flag;
-	nd.intent.open.flags = flag;
-	nd.intent.open.create_mode = mode;
-	dir = nd.path.dentry;
-	nd.flags &= ~LOOKUP_PARENT;
-	nd.flags |= LOOKUP_CREATE | LOOKUP_OPEN;
-	if (flag & O_EXCL)
-		nd.flags |= LOOKUP_EXCL;
+	/* OK, it's O_CREAT */
 	mutex_lock(&dir->d_inode->i_mutex);
-	path.dentry = lookup_hash(&nd);
-	path.mnt = nd.path.mnt;
 
-do_last:
-	error = PTR_ERR(path.dentry);
-	if (IS_ERR(path.dentry)) {
+	path->dentry = lookup_hash(nd);
+	path->mnt = nd->path.mnt;
+
+	error = PTR_ERR(path->dentry);
+	if (IS_ERR(path->dentry)) {
 		mutex_unlock(&dir->d_inode->i_mutex);
 		goto exit;
 	}
 
-	if (IS_ERR(nd.intent.open.file)) {
-		error = PTR_ERR(nd.intent.open.file);
+	if (IS_ERR(nd->intent.open.file)) {
+		error = PTR_ERR(nd->intent.open.file);
 		goto exit_mutex_unlock;
 	}
 
 	/* Negative dentry, just create the file */
-	if (!path.dentry->d_inode) {
+	if (!path->dentry->d_inode) {
 		/*
 		 * This write is needed to ensure that a
 		 * ro->rw transition does not occur between
@@ -1723,18 +1689,16 @@ do_last:
 		 * a permanent write count is taken through
 		 * the 'struct file' in nameidata_to_filp().
 		 */
-		error = mnt_want_write(nd.path.mnt);
+		error = mnt_want_write(nd->path.mnt);
 		if (error)
 			goto exit_mutex_unlock;
-		error = __open_namei_create(&nd, &path, flag, mode);
+		error = __open_namei_create(nd, path, open_flag, mode);
 		if (error) {
-			mnt_drop_write(nd.path.mnt);
+			mnt_drop_write(nd->path.mnt);
 			goto exit;
 		}
-		filp = nameidata_to_filp(&nd);
-		mnt_drop_write(nd.path.mnt);
-		if (nd.root.mnt)
-			path_put(&nd.root);
+		filp = nameidata_to_filp(nd);
+		mnt_drop_write(nd->path.mnt);
 		if (!IS_ERR(filp)) {
 			error = ima_file_check(filp, acc_mode);
 			if (error) {
@@ -1749,150 +1713,181 @@ do_last:
 	 * It already exists.
 	 */
 	mutex_unlock(&dir->d_inode->i_mutex);
-	audit_inode(pathname, path.dentry);
+	audit_inode(pathname, path->dentry);
 
 	error = -EEXIST;
-	if (flag & O_EXCL)
+	if (open_flag & O_EXCL)
 		goto exit_dput;
 
-	if (__follow_mount(&path)) {
+	if (__follow_mount(path)) {
 		error = -ELOOP;
-		if (flag & O_NOFOLLOW)
+		if (open_flag & O_NOFOLLOW)
 			goto exit_dput;
 	}
 
 	error = -ENOENT;
-	if (!path.dentry->d_inode)
+	if (!path->dentry->d_inode)
 		goto exit_dput;
-	if (path.dentry->d_inode->i_op->follow_link)
-		goto do_link;
 
-	path_to_nameidata(&path, &nd);
+	if (path->dentry->d_inode->i_op->follow_link)
+		return NULL;
+
+	path_to_nameidata(path, nd);
 	error = -EISDIR;
-	if (S_ISDIR(path.dentry->d_inode->i_mode))
+	if (S_ISDIR(path->dentry->d_inode->i_mode))
 		goto exit;
 ok:
+	filp = finish_open(nd, open_flag, acc_mode);
+	return filp;
+
+exit_mutex_unlock:
+	mutex_unlock(&dir->d_inode->i_mutex);
+exit_dput:
+	path_put_conditional(path, nd);
+exit:
+	if (!IS_ERR(nd->intent.open.file))
+		release_open_intent(nd);
+	path_put(&nd->path);
+	return ERR_PTR(error);
+}
+
+/*
+ * Note that the low bits of the passed in "open_flag"
+ * are not the same as in the local variable "flag". See
+ * open_to_namei_flags() for more details.
+ */
+struct file *do_filp_open(int dfd, const char *pathname,
+		int open_flag, int mode, int acc_mode)
+{
+	struct file *filp;
+	struct nameidata nd;
+	int error;
+	struct path path;
+	int count = 0;
+	int flag = open_to_namei_flags(open_flag);
+	int force_reval = 0;
+	int want_dir = open_flag & O_DIRECTORY;
+
+	if (!(open_flag & O_CREAT))
+		mode = 0;
+
 	/*
-	 * Consider:
-	 * 1. may_open() truncates a file
-	 * 2. a rw->ro mount transition occurs
-	 * 3. nameidata_to_filp() fails due to
-	 *    the ro mount.
-	 * That would be inconsistent, and should
-	 * be avoided. Taking this mnt write here
-	 * ensures that (2) can not occur.
+	 * O_SYNC is implemented as __O_SYNC|O_DSYNC.  As many places only
+	 * check for O_DSYNC if the need any syncing at all we enforce it's
+	 * always set instead of having to deal with possibly weird behaviour
+	 * for malicious applications setting only __O_SYNC.
 	 */
-	will_truncate = open_will_truncate(flag, nd.path.dentry->d_inode);
-	if (will_truncate) {
-		error = mnt_want_write(nd.path.mnt);
-		if (error)
-			goto exit;
-	}
-	error = may_open(&nd.path, acc_mode, flag);
+	if (open_flag & __O_SYNC)
+		open_flag |= O_DSYNC;
+
+	if (!acc_mode)
+		acc_mode = MAY_OPEN | ACC_MODE(open_flag);
+
+	/* O_TRUNC implies we need access checks for write permissions */
+	if (open_flag & O_TRUNC)
+		acc_mode |= MAY_WRITE;
+
+	/* Allow the LSM permission hook to distinguish append 
+	   access from general write access. */
+	if (open_flag & O_APPEND)
+		acc_mode |= MAY_APPEND;
+
+	/* find the parent */
+reval:
+	error = path_init(dfd, pathname, LOOKUP_PARENT, &nd);
+	if (error)
+		return ERR_PTR(error);
+	if (force_reval)
+		nd.flags |= LOOKUP_REVAL;
+
+	current->total_link_count = 0;
+	error = link_path_walk(pathname, &nd);
 	if (error) {
-		if (will_truncate)
-			mnt_drop_write(nd.path.mnt);
-		goto exit;
-	}
-	filp = nameidata_to_filp(&nd);
-	if (!IS_ERR(filp)) {
-		error = ima_file_check(filp, acc_mode);
-		if (error) {
-			fput(filp);
-			filp = ERR_PTR(error);
-		}
+		filp = ERR_PTR(error);
+		goto out;
 	}
-	if (!IS_ERR(filp)) {
-		if (acc_mode & MAY_WRITE)
-			vfs_dq_init(nd.path.dentry->d_inode);
+	if (unlikely(!audit_dummy_context()) && (open_flag & O_CREAT))
+		audit_inode(pathname, nd.path.dentry);
 
-		if (will_truncate) {
-			error = handle_truncate(&nd.path);
-			if (error) {
-				fput(filp);
-				filp = ERR_PTR(error);
-			}
-		}
-	}
 	/*
-	 * It is now safe to drop the mnt write
-	 * because the filp has had a write taken
-	 * on its behalf.
+	 * We have the parent and last component.
 	 */
-	if (will_truncate)
-		mnt_drop_write(nd.path.mnt);
+
+	error = -ENFILE;
+	filp = get_empty_filp();
+	if (filp == NULL)
+		goto exit_parent;
+	nd.intent.open.file = filp;
+	filp->f_flags = open_flag;
+	nd.intent.open.flags = flag;
+	nd.intent.open.create_mode = mode;
+	nd.flags &= ~LOOKUP_PARENT;
+	nd.flags |= LOOKUP_OPEN;
+	if (open_flag & O_CREAT) {
+		nd.flags |= LOOKUP_CREATE;
+		if (open_flag & O_EXCL)
+			nd.flags |= LOOKUP_EXCL;
+	}
+	filp = do_last(&nd, &path, open_flag, acc_mode, mode, pathname, &want_dir);
+	while (unlikely(!filp)) { /* trailing symlink */
+		struct path holder;
+		struct inode *inode = path.dentry->d_inode;
+		void *cookie;
+		error = -ELOOP;
+		/* S_ISDIR part is a temporary automount kludge */
+		if ((open_flag & O_NOFOLLOW) && !S_ISDIR(inode->i_mode))
+			goto exit_dput;
+		if (count++ == 32)
+			goto exit_dput;
+		/*
+		 * This is subtle. Instead of calling do_follow_link() we do
+		 * the thing by hands. The reason is that this way we have zero
+		 * link_count and path_walk() (called from ->follow_link)
+		 * honoring LOOKUP_PARENT.  After that we have the parent and
+		 * last component, i.e. we are in the same situation as after
+		 * the first path_walk().  Well, almost - if the last component
+		 * is normal we get its copy 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);
+		if (error)
+			goto exit_dput;
+		error = __do_follow_link(&path, &nd, &cookie);
+		if (unlikely(error)) {
+			/* nd.path had been dropped */
+			if (!IS_ERR(cookie) && inode->i_op->put_link)
+				inode->i_op->put_link(path.dentry, &nd, cookie);
+			path_put(&path);
+			release_open_intent(&nd);
+			filp = ERR_PTR(error);
+			goto out;
+		}
+		holder = path;
+		nd.flags &= ~LOOKUP_PARENT;
+		filp = do_last(&nd, &path, open_flag, acc_mode, mode, pathname, &want_dir);
+		if (inode->i_op->put_link)
+			inode->i_op->put_link(holder.dentry, &nd, cookie);
+		path_put(&holder);
+	}
+out:
 	if (nd.root.mnt)
 		path_put(&nd.root);
+	if (filp == ERR_PTR(-ESTALE) && !force_reval) {
+		force_reval = 1;
+		goto reval;
+	}
 	return filp;
 
-exit_mutex_unlock:
-	mutex_unlock(&dir->d_inode->i_mutex);
 exit_dput:
 	path_put_conditional(&path, &nd);
-exit:
 	if (!IS_ERR(nd.intent.open.file))
 		release_open_intent(&nd);
 exit_parent:
-	if (nd.root.mnt)
-		path_put(&nd.root);
 	path_put(&nd.path);
-	return ERR_PTR(error);
-
-do_link:
-	error = -ELOOP;
-	if (flag & O_NOFOLLOW)
-		goto exit_dput;
-	/*
-	 * This is subtle. Instead of calling do_follow_link() we do the
-	 * thing by hands. The reason is that this way we have zero link_count
-	 * and path_walk() (called from ->follow_link) honoring LOOKUP_PARENT.
-	 * After that we have the parent and last component, i.e.
-	 * we are in the same situation as after the first path_walk().
-	 * Well, almost - if the last component is normal we get its copy
-	 * 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);
-	if (error)
-		goto exit_dput;
-	error = __do_follow_link(&path, &nd);
-	path_put(&path);
-	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);
-		if (nd.root.mnt)
-			path_put(&nd.root);
-		if (error == -ESTALE && !force_reval) {
-			force_reval = 1;
-			goto reval;
-		}
-		return ERR_PTR(error);
-	}
-	nd.flags &= ~LOOKUP_PARENT;
-	if (nd.last_type == LAST_BIND)
-		goto ok;
-	error = -EISDIR;
-	if (nd.last_type != LAST_NORM)
-		goto exit;
-	if (nd.last.name[nd.last.len]) {
-		__putname(nd.last.name);
-		goto exit;
-	}
-	error = -ELOOP;
-	if (count++==32) {
-		__putname(nd.last.name);
-		goto exit;
-	}
-	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);
-	goto do_last;
+	filp = ERR_PTR(error);
+	goto out;
 }
 
 /**
@@ -1986,7 +1981,6 @@ int vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev)
 	if (error)
 		return error;
 
-	vfs_dq_init(dir);
 	error = dir->i_op->mknod(dir, dentry, mode, dev);
 	if (!error)
 		fsnotify_create(dir, dentry);
@@ -2085,7 +2079,6 @@ int vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	if (error)
 		return error;
 
-	vfs_dq_init(dir);
 	error = dir->i_op->mkdir(dir, dentry, mode);
 	if (!error)
 		fsnotify_mkdir(dir, dentry);
@@ -2171,8 +2164,6 @@ int vfs_rmdir(struct inode *dir, struct dentry *dentry)
 	if (!dir->i_op->rmdir)
 		return -EPERM;
 
-	vfs_dq_init(dir);
-
 	mutex_lock(&dentry->d_inode->i_mutex);
 	dentry_unhash(dentry);
 	if (d_mountpoint(dentry))
@@ -2258,15 +2249,16 @@ int vfs_unlink(struct inode *dir, struct dentry *dentry)
 	if (!dir->i_op->unlink)
 		return -EPERM;
 
-	vfs_dq_init(dir);
-
 	mutex_lock(&dentry->d_inode->i_mutex);
 	if (d_mountpoint(dentry))
 		error = -EBUSY;
 	else {
 		error = security_inode_unlink(dir, dentry);
-		if (!error)
+		if (!error) {
 			error = dir->i_op->unlink(dir, dentry);
+			if (!error)
+				dentry->d_inode->i_flags |= S_DEAD;
+		}
 	}
 	mutex_unlock(&dentry->d_inode->i_mutex);
 
@@ -2369,7 +2361,6 @@ int vfs_symlink(struct inode *dir, struct dentry *dentry, const char *oldname)
 	if (error)
 		return error;
 
-	vfs_dq_init(dir);
 	error = dir->i_op->symlink(dir, dentry, oldname);
 	if (!error)
 		fsnotify_create(dir, dentry);
@@ -2453,7 +2444,6 @@ int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_de
 		return error;
 
 	mutex_lock(&inode->i_mutex);
-	vfs_dq_init(dir);
 	error = dir->i_op->link(old_dentry, dir, new_dentry);
 	mutex_unlock(&inode->i_mutex);
 	if (!error)
@@ -2619,6 +2609,8 @@ static int vfs_rename_other(struct inode *old_dir, struct dentry *old_dentry,
 	else
 		error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
 	if (!error) {
+		if (target)
+			target->i_flags |= S_DEAD;
 		if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
 			d_move(old_dentry, new_dentry);
 	}
@@ -2652,20 +2644,15 @@ int vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (!old_dir->i_op->rename)
 		return -EPERM;
 
-	vfs_dq_init(old_dir);
-	vfs_dq_init(new_dir);
-
 	old_name = fsnotify_oldname_init(old_dentry->d_name.name);
 
 	if (is_dir)
 		error = vfs_rename_dir(old_dir,old_dentry,new_dir,new_dentry);
 	else
 		error = vfs_rename_other(old_dir,old_dentry,new_dir,new_dentry);
-	if (!error) {
-		const char *new_name = old_dentry->d_name.name;
-		fsnotify_move(old_dir, new_dir, old_name, new_name, is_dir,
+	if (!error)
+		fsnotify_move(old_dir, new_dir, old_name, is_dir,
 			      new_dentry->d_inode, old_dentry);
-	}
 	fsnotify_oldname_free(old_name);
 
 	return error;
diff --git a/fs/namespace.c b/fs/namespace.c
index c768f733c8d6..8174c8ab5c70 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -573,7 +573,7 @@ static struct vfsmount *clone_mnt(struct vfsmount *old, struct dentry *root,
 			mnt->mnt_master = old;
 			CLEAR_MNT_SHARED(mnt);
 		} else if (!(flag & CL_PRIVATE)) {
-			if ((flag & CL_PROPAGATION) || IS_MNT_SHARED(old))
+			if ((flag & CL_MAKE_SHARED) || IS_MNT_SHARED(old))
 				list_add(&mnt->mnt_share, &old->mnt_share);
 			if (IS_MNT_SLAVE(old))
 				list_add(&mnt->mnt_slave, &old->mnt_slave);
@@ -737,6 +737,21 @@ static void m_stop(struct seq_file *m, void *v)
 	up_read(&namespace_sem);
 }
 
+int mnt_had_events(struct proc_mounts *p)
+{
+	struct mnt_namespace *ns = p->ns;
+	int res = 0;
+
+	spin_lock(&vfsmount_lock);
+	if (p->event != ns->event) {
+		p->event = ns->event;
+		res = 1;
+	}
+	spin_unlock(&vfsmount_lock);
+
+	return res;
+}
+
 struct proc_fs_info {
 	int flag;
 	const char *str;
@@ -1121,8 +1136,15 @@ SYSCALL_DEFINE2(umount, char __user *, name, int, flags)
 {
 	struct path path;
 	int retval;
+	int lookup_flags = 0;
 
-	retval = user_path(name, &path);
+	if (flags & ~(MNT_FORCE | MNT_DETACH | MNT_EXPIRE | UMOUNT_NOFOLLOW))
+		return -EINVAL;
+
+	if (!(flags & UMOUNT_NOFOLLOW))
+		lookup_flags |= LOOKUP_FOLLOW;
+
+	retval = user_path_at(AT_FDCWD, name, lookup_flags, &path);
 	if (retval)
 		goto out;
 	retval = -EINVAL;
@@ -1246,6 +1268,21 @@ void drop_collected_mounts(struct vfsmount *mnt)
 	release_mounts(&umount_list);
 }
 
+int iterate_mounts(int (*f)(struct vfsmount *, void *), void *arg,
+		   struct vfsmount *root)
+{
+	struct vfsmount *mnt;
+	int res = f(root, arg);
+	if (res)
+		return res;
+	list_for_each_entry(mnt, &root->mnt_list, mnt_list) {
+		res = f(mnt, arg);
+		if (res)
+			return res;
+	}
+	return 0;
+}
+
 static void cleanup_group_ids(struct vfsmount *mnt, struct vfsmount *end)
 {
 	struct vfsmount *p;
@@ -1538,7 +1575,7 @@ static int do_remount(struct path *path, int flags, int mnt_flags,
 		err = do_remount_sb(sb, flags, data, 0);
 	if (!err) {
 		spin_lock(&vfsmount_lock);
-		mnt_flags |= path->mnt->mnt_flags & MNT_PNODE_MASK;
+		mnt_flags |= path->mnt->mnt_flags & MNT_PROPAGATION_MASK;
 		path->mnt->mnt_flags = mnt_flags;
 		spin_unlock(&vfsmount_lock);
 	}
@@ -1671,7 +1708,7 @@ int do_add_mount(struct vfsmount *newmnt, struct path *path,
 {
 	int err;
 
-	mnt_flags &= ~(MNT_SHARED | MNT_WRITE_HOLD);
+	mnt_flags &= ~(MNT_SHARED | MNT_WRITE_HOLD | MNT_INTERNAL);
 
 	down_write(&namespace_sem);
 	/* Something was mounted here while we slept */
@@ -2314,17 +2351,13 @@ void __init mnt_init(void)
 
 void put_mnt_ns(struct mnt_namespace *ns)
 {
-	struct vfsmount *root;
 	LIST_HEAD(umount_list);
 
-	if (!atomic_dec_and_lock(&ns->count, &vfsmount_lock))
+	if (!atomic_dec_and_test(&ns->count))
 		return;
-	root = ns->root;
-	ns->root = NULL;
-	spin_unlock(&vfsmount_lock);
 	down_write(&namespace_sem);
 	spin_lock(&vfsmount_lock);
-	umount_tree(root, 0, &umount_list);
+	umount_tree(ns->root, 0, &umount_list);
 	spin_unlock(&vfsmount_lock);
 	up_write(&namespace_sem);
 	release_mounts(&umount_list);
diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig
index 59e5673b4597..a43d07e7b924 100644
--- a/fs/nfs/Kconfig
+++ b/fs/nfs/Kconfig
@@ -95,8 +95,7 @@ config ROOT_NFS
 	  Most people say N here.
 
 config NFS_FSCACHE
-	bool "Provide NFS client caching support (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
+	bool "Provide NFS client caching support"
 	depends on NFS_FS=m && FSCACHE || NFS_FS=y && FSCACHE=y
 	help
 	  Say Y here if you want NFS data to be cached locally on disc through
diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index 73ab220354df..36dfdae95123 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -118,7 +118,6 @@ nfs4_callback_up(struct svc_serv *serv)
 	dprintk("NFS: Callback listener port = %u (af %u)\n",
 			nfs_callback_tcpport, PF_INET);
 
-#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
 	ret = svc_create_xprt(serv, "tcp", PF_INET6,
 				nfs_callback_set_tcpport, SVC_SOCK_ANONYMOUS);
 	if (ret > 0) {
@@ -129,7 +128,6 @@ nfs4_callback_up(struct svc_serv *serv)
 		ret = 0;
 	else
 		goto out_err;
-#endif	/* defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) */
 
 	return svc_prepare_thread(serv, &serv->sv_pools[0]);
 
diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h
index d4036be0b589..85a7cfd1b8dd 100644
--- a/fs/nfs/callback.h
+++ b/fs/nfs/callback.h
@@ -119,6 +119,14 @@ struct cb_recallanyargs {
 };
 
 extern unsigned nfs4_callback_recallany(struct cb_recallanyargs *args, void *dummy);
+
+struct cb_recallslotargs {
+	struct sockaddr	*crsa_addr;
+	uint32_t	crsa_target_max_slots;
+};
+extern unsigned nfs4_callback_recallslot(struct cb_recallslotargs *args,
+					  void *dummy);
+
 #endif /* CONFIG_NFS_V4_1 */
 
 extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res);
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index defa9b4c470e..84761b5bb8e2 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -143,44 +143,49 @@ int nfs41_validate_delegation_stateid(struct nfs_delegation *delegation, const n
  * Return success if the sequenceID is one more than what we last saw on
  * this slot, accounting for wraparound.  Increments the slot's sequence.
  *
- * We don't yet implement a duplicate request cache, so at this time
- * we will log replays, and process them as if we had not seen them before,
- * but we don't bump the sequence in the slot.  Not too worried about it,
+ * We don't yet implement a duplicate request cache, instead we set the
+ * back channel ca_maxresponsesize_cached to zero. This is OK for now
  * since we only currently implement idempotent callbacks anyway.
  *
  * We have a single slot backchannel at this time, so we don't bother
  * checking the used_slots bit array on the table.  The lower layer guarantees
  * a single outstanding callback request at a time.
  */
-static int
-validate_seqid(struct nfs4_slot_table *tbl, u32 slotid, u32 seqid)
+static __be32
+validate_seqid(struct nfs4_slot_table *tbl, struct cb_sequenceargs * args)
 {
 	struct nfs4_slot *slot;
 
 	dprintk("%s enter. slotid %d seqid %d\n",
-		__func__, slotid, seqid);
+		__func__, args->csa_slotid, args->csa_sequenceid);
 
-	if (slotid > NFS41_BC_MAX_CALLBACKS)
+	if (args->csa_slotid > NFS41_BC_MAX_CALLBACKS)
 		return htonl(NFS4ERR_BADSLOT);
 
-	slot = tbl->slots + slotid;
+	slot = tbl->slots + args->csa_slotid;
 	dprintk("%s slot table seqid: %d\n", __func__, slot->seq_nr);
 
 	/* Normal */
-	if (likely(seqid == slot->seq_nr + 1)) {
+	if (likely(args->csa_sequenceid == slot->seq_nr + 1)) {
 		slot->seq_nr++;
 		return htonl(NFS4_OK);
 	}
 
 	/* Replay */
-	if (seqid == slot->seq_nr) {
-		dprintk("%s seqid %d is a replay - no DRC available\n",
-			__func__, seqid);
-		return htonl(NFS4_OK);
+	if (args->csa_sequenceid == slot->seq_nr) {
+		dprintk("%s seqid %d is a replay\n",
+			__func__, args->csa_sequenceid);
+		/* Signal process_op to set this error on next op */
+		if (args->csa_cachethis == 0)
+			return htonl(NFS4ERR_RETRY_UNCACHED_REP);
+
+		/* The ca_maxresponsesize_cached is 0 with no DRC */
+		else if (args->csa_cachethis == 1)
+			return htonl(NFS4ERR_REP_TOO_BIG_TO_CACHE);
 	}
 
 	/* Wraparound */
-	if (seqid == 1 && (slot->seq_nr + 1) == 0) {
+	if (args->csa_sequenceid == 1 && (slot->seq_nr + 1) == 0) {
 		slot->seq_nr = 1;
 		return htonl(NFS4_OK);
 	}
@@ -225,27 +230,87 @@ validate_seqid(struct nfs4_slot_table *tbl, u32 slotid, u32 seqid)
 	return NULL;
 }
 
-/* FIXME: referring calls should be processed */
-unsigned nfs4_callback_sequence(struct cb_sequenceargs *args,
+/*
+ * For each referring call triple, check the session's slot table for
+ * a match.  If the slot is in use and the sequence numbers match, the
+ * client is still waiting for a response to the original request.
+ */
+static bool referring_call_exists(struct nfs_client *clp,
+				  uint32_t nrclists,
+				  struct referring_call_list *rclists)
+{
+	bool status = 0;
+	int i, j;
+	struct nfs4_session *session;
+	struct nfs4_slot_table *tbl;
+	struct referring_call_list *rclist;
+	struct referring_call *ref;
+
+	/*
+	 * XXX When client trunking is implemented, this becomes
+	 * a session lookup from within the loop
+	 */
+	session = clp->cl_session;
+	tbl = &session->fc_slot_table;
+
+	for (i = 0; i < nrclists; i++) {
+		rclist = &rclists[i];
+		if (memcmp(session->sess_id.data,
+			   rclist->rcl_sessionid.data,
+			   NFS4_MAX_SESSIONID_LEN) != 0)
+			continue;
+
+		for (j = 0; j < rclist->rcl_nrefcalls; j++) {
+			ref = &rclist->rcl_refcalls[j];
+
+			dprintk("%s: sessionid %x:%x:%x:%x sequenceid %u "
+				"slotid %u\n", __func__,
+				((u32 *)&rclist->rcl_sessionid.data)[0],
+				((u32 *)&rclist->rcl_sessionid.data)[1],
+				((u32 *)&rclist->rcl_sessionid.data)[2],
+				((u32 *)&rclist->rcl_sessionid.data)[3],
+				ref->rc_sequenceid, ref->rc_slotid);
+
+			spin_lock(&tbl->slot_tbl_lock);
+			status = (test_bit(ref->rc_slotid, tbl->used_slots) &&
+				  tbl->slots[ref->rc_slotid].seq_nr ==
+					ref->rc_sequenceid);
+			spin_unlock(&tbl->slot_tbl_lock);
+			if (status)
+				goto out;
+		}
+	}
+
+out:
+	return status;
+}
+
+__be32 nfs4_callback_sequence(struct cb_sequenceargs *args,
 				struct cb_sequenceres *res)
 {
 	struct nfs_client *clp;
-	int i, status;
-
-	for (i = 0; i < args->csa_nrclists; i++)
-		kfree(args->csa_rclists[i].rcl_refcalls);
-	kfree(args->csa_rclists);
+	int i;
+	__be32 status;
 
 	status = htonl(NFS4ERR_BADSESSION);
 	clp = find_client_with_session(args->csa_addr, 4, &args->csa_sessionid);
 	if (clp == NULL)
 		goto out;
 
-	status = validate_seqid(&clp->cl_session->bc_slot_table,
-				args->csa_slotid, args->csa_sequenceid);
+	status = validate_seqid(&clp->cl_session->bc_slot_table, args);
 	if (status)
 		goto out_putclient;
 
+	/*
+	 * Check for pending referring calls.  If a match is found, a
+	 * related callback was received before the response to the original
+	 * call.
+	 */
+	if (referring_call_exists(clp, args->csa_nrclists, args->csa_rclists)) {
+		status = htonl(NFS4ERR_DELAY);
+		goto out_putclient;
+	}
+
 	memcpy(&res->csr_sessionid, &args->csa_sessionid,
 	       sizeof(res->csr_sessionid));
 	res->csr_sequenceid = args->csa_sequenceid;
@@ -256,15 +321,23 @@ unsigned nfs4_callback_sequence(struct cb_sequenceargs *args,
 out_putclient:
 	nfs_put_client(clp);
 out:
-	dprintk("%s: exit with status = %d\n", __func__, ntohl(status));
-	res->csr_status = status;
-	return res->csr_status;
+	for (i = 0; i < args->csa_nrclists; i++)
+		kfree(args->csa_rclists[i].rcl_refcalls);
+	kfree(args->csa_rclists);
+
+	if (status == htonl(NFS4ERR_RETRY_UNCACHED_REP))
+		res->csr_status = 0;
+	else
+		res->csr_status = status;
+	dprintk("%s: exit with status = %d res->csr_status %d\n", __func__,
+		ntohl(status), ntohl(res->csr_status));
+	return status;
 }
 
-unsigned nfs4_callback_recallany(struct cb_recallanyargs *args, void *dummy)
+__be32 nfs4_callback_recallany(struct cb_recallanyargs *args, void *dummy)
 {
 	struct nfs_client *clp;
-	int status;
+	__be32 status;
 	fmode_t flags = 0;
 
 	status = htonl(NFS4ERR_OP_NOT_IN_SESSION);
@@ -289,4 +362,40 @@ out:
 	dprintk("%s: exit with status = %d\n", __func__, ntohl(status));
 	return status;
 }
+
+/* Reduce the fore channel's max_slots to the target value */
+__be32 nfs4_callback_recallslot(struct cb_recallslotargs *args, void *dummy)
+{
+	struct nfs_client *clp;
+	struct nfs4_slot_table *fc_tbl;
+	__be32 status;
+
+	status = htonl(NFS4ERR_OP_NOT_IN_SESSION);
+	clp = nfs_find_client(args->crsa_addr, 4);
+	if (clp == NULL)
+		goto out;
+
+	dprintk("NFS: CB_RECALL_SLOT request from %s target max slots %d\n",
+		rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR),
+		args->crsa_target_max_slots);
+
+	fc_tbl = &clp->cl_session->fc_slot_table;
+
+	status = htonl(NFS4ERR_BAD_HIGH_SLOT);
+	if (args->crsa_target_max_slots > fc_tbl->max_slots ||
+	    args->crsa_target_max_slots < 1)
+		goto out_putclient;
+
+	status = htonl(NFS4_OK);
+	if (args->crsa_target_max_slots == fc_tbl->max_slots)
+		goto out_putclient;
+
+	fc_tbl->target_max_slots = args->crsa_target_max_slots;
+	nfs41_handle_recall_slot(clp);
+out_putclient:
+	nfs_put_client(clp);	/* balance nfs_find_client */
+out:
+	dprintk("%s: exit with status = %d\n", __func__, ntohl(status));
+	return status;
+}
 #endif /* CONFIG_NFS_V4_1 */
diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c
index 8e1a2511c8be..db30c0b398b5 100644
--- a/fs/nfs/callback_xdr.c
+++ b/fs/nfs/callback_xdr.c
@@ -24,10 +24,14 @@
 #define CB_OP_SEQUENCE_RES_MAXSZ	(CB_OP_HDR_RES_MAXSZ + \
 					4 + 1 + 3)
 #define CB_OP_RECALLANY_RES_MAXSZ	(CB_OP_HDR_RES_MAXSZ)
+#define CB_OP_RECALLSLOT_RES_MAXSZ	(CB_OP_HDR_RES_MAXSZ)
 #endif /* CONFIG_NFS_V4_1 */
 
 #define NFSDBG_FACILITY NFSDBG_CALLBACK
 
+/* Internal error code */
+#define NFS4ERR_RESOURCE_HDR	11050
+
 typedef __be32 (*callback_process_op_t)(void *, void *);
 typedef __be32 (*callback_decode_arg_t)(struct svc_rqst *, struct xdr_stream *, void *);
 typedef __be32 (*callback_encode_res_t)(struct svc_rqst *, struct xdr_stream *, void *);
@@ -173,7 +177,7 @@ static __be32 decode_op_hdr(struct xdr_stream *xdr, unsigned int *op)
 	__be32 *p;
 	p = read_buf(xdr, 4);
 	if (unlikely(p == NULL))
-		return htonl(NFS4ERR_RESOURCE);
+		return htonl(NFS4ERR_RESOURCE_HDR);
 	*op = ntohl(*p);
 	return 0;
 }
@@ -215,10 +219,10 @@ out:
 
 #if defined(CONFIG_NFS_V4_1)
 
-static unsigned decode_sessionid(struct xdr_stream *xdr,
+static __be32 decode_sessionid(struct xdr_stream *xdr,
 				 struct nfs4_sessionid *sid)
 {
-	uint32_t *p;
+	__be32 *p;
 	int len = NFS4_MAX_SESSIONID_LEN;
 
 	p = read_buf(xdr, len);
@@ -229,12 +233,12 @@ static unsigned decode_sessionid(struct xdr_stream *xdr,
 	return 0;
 }
 
-static unsigned decode_rc_list(struct xdr_stream *xdr,
+static __be32 decode_rc_list(struct xdr_stream *xdr,
 			       struct referring_call_list *rc_list)
 {
-	uint32_t *p;
+	__be32 *p;
 	int i;
-	unsigned status;
+	__be32 status;
 
 	status = decode_sessionid(xdr, &rc_list->rcl_sessionid);
 	if (status)
@@ -267,13 +271,13 @@ out:
 	return status;
 }
 
-static unsigned decode_cb_sequence_args(struct svc_rqst *rqstp,
+static __be32 decode_cb_sequence_args(struct svc_rqst *rqstp,
 					struct xdr_stream *xdr,
 					struct cb_sequenceargs *args)
 {
-	uint32_t *p;
+	__be32 *p;
 	int i;
-	unsigned status;
+	__be32 status;
 
 	status = decode_sessionid(xdr, &args->csa_sessionid);
 	if (status)
@@ -327,11 +331,11 @@ out_free:
 	goto out;
 }
 
-static unsigned decode_recallany_args(struct svc_rqst *rqstp,
+static __be32 decode_recallany_args(struct svc_rqst *rqstp,
 				      struct xdr_stream *xdr,
 				      struct cb_recallanyargs *args)
 {
-	uint32_t *p;
+	__be32 *p;
 
 	args->craa_addr = svc_addr(rqstp);
 	p = read_buf(xdr, 4);
@@ -346,6 +350,20 @@ static unsigned decode_recallany_args(struct svc_rqst *rqstp,
 	return 0;
 }
 
+static __be32 decode_recallslot_args(struct svc_rqst *rqstp,
+					struct xdr_stream *xdr,
+					struct cb_recallslotargs *args)
+{
+	__be32 *p;
+
+	args->crsa_addr = svc_addr(rqstp);
+	p = read_buf(xdr, 4);
+	if (unlikely(p == NULL))
+		return htonl(NFS4ERR_BADXDR);
+	args->crsa_target_max_slots = ntohl(*p++);
+	return 0;
+}
+
 #endif /* CONFIG_NFS_V4_1 */
 
 static __be32 encode_string(struct xdr_stream *xdr, unsigned int len, const char *str)
@@ -465,7 +483,7 @@ static __be32 encode_op_hdr(struct xdr_stream *xdr, uint32_t op, __be32 res)
 	
 	p = xdr_reserve_space(xdr, 8);
 	if (unlikely(p == NULL))
-		return htonl(NFS4ERR_RESOURCE);
+		return htonl(NFS4ERR_RESOURCE_HDR);
 	*p++ = htonl(op);
 	*p = res;
 	return 0;
@@ -499,10 +517,10 @@ out:
 
 #if defined(CONFIG_NFS_V4_1)
 
-static unsigned encode_sessionid(struct xdr_stream *xdr,
+static __be32 encode_sessionid(struct xdr_stream *xdr,
 				 const struct nfs4_sessionid *sid)
 {
-	uint32_t *p;
+	__be32 *p;
 	int len = NFS4_MAX_SESSIONID_LEN;
 
 	p = xdr_reserve_space(xdr, len);
@@ -513,11 +531,11 @@ static unsigned encode_sessionid(struct xdr_stream *xdr,
 	return 0;
 }
 
-static unsigned encode_cb_sequence_res(struct svc_rqst *rqstp,
+static __be32 encode_cb_sequence_res(struct svc_rqst *rqstp,
 				       struct xdr_stream *xdr,
 				       const struct cb_sequenceres *res)
 {
-	uint32_t *p;
+	__be32 *p;
 	unsigned status = res->csr_status;
 
 	if (unlikely(status != 0))
@@ -554,6 +572,7 @@ preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op)
 	case OP_CB_RECALL:
 	case OP_CB_SEQUENCE:
 	case OP_CB_RECALL_ANY:
+	case OP_CB_RECALL_SLOT:
 		*op = &callback_ops[op_nr];
 		break;
 
@@ -562,7 +581,6 @@ preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op)
 	case OP_CB_NOTIFY:
 	case OP_CB_PUSH_DELEG:
 	case OP_CB_RECALLABLE_OBJ_AVAIL:
-	case OP_CB_RECALL_SLOT:
 	case OP_CB_WANTS_CANCELLED:
 	case OP_CB_NOTIFY_LOCK:
 		return htonl(NFS4ERR_NOTSUPP);
@@ -602,20 +620,18 @@ preprocess_nfs4_op(unsigned int op_nr, struct callback_op **op)
 static __be32 process_op(uint32_t minorversion, int nop,
 		struct svc_rqst *rqstp,
 		struct xdr_stream *xdr_in, void *argp,
-		struct xdr_stream *xdr_out, void *resp)
+		struct xdr_stream *xdr_out, void *resp, int* drc_status)
 {
 	struct callback_op *op = &callback_ops[0];
-	unsigned int op_nr = OP_CB_ILLEGAL;
+	unsigned int op_nr;
 	__be32 status;
 	long maxlen;
 	__be32 res;
 
 	dprintk("%s: start\n", __func__);
 	status = decode_op_hdr(xdr_in, &op_nr);
-	if (unlikely(status)) {
-		status = htonl(NFS4ERR_OP_ILLEGAL);
-		goto out;
-	}
+	if (unlikely(status))
+		return status;
 
 	dprintk("%s: minorversion=%d nop=%d op_nr=%u\n",
 		__func__, minorversion, nop, op_nr);
@@ -624,19 +640,32 @@ static __be32 process_op(uint32_t minorversion, int nop,
 				preprocess_nfs4_op(op_nr, &op);
 	if (status == htonl(NFS4ERR_OP_ILLEGAL))
 		op_nr = OP_CB_ILLEGAL;
-out:
+	if (status)
+		goto encode_hdr;
+
+	if (*drc_status) {
+		status = *drc_status;
+		goto encode_hdr;
+	}
+
 	maxlen = xdr_out->end - xdr_out->p;
 	if (maxlen > 0 && maxlen < PAGE_SIZE) {
-		if (likely(status == 0 && op->decode_args != NULL))
-			status = op->decode_args(rqstp, xdr_in, argp);
-		if (likely(status == 0 && op->process_op != NULL))
+		status = op->decode_args(rqstp, xdr_in, argp);
+		if (likely(status == 0))
 			status = op->process_op(argp, resp);
 	} else
 		status = htonl(NFS4ERR_RESOURCE);
 
+	/* Only set by OP_CB_SEQUENCE processing */
+	if (status == htonl(NFS4ERR_RETRY_UNCACHED_REP)) {
+		*drc_status = status;
+		status = 0;
+	}
+
+encode_hdr:
 	res = encode_op_hdr(xdr_out, op_nr, status);
-	if (status == 0)
-		status = res;
+	if (unlikely(res))
+		return res;
 	if (op->encode_res != NULL && status == 0)
 		status = op->encode_res(rqstp, xdr_out, resp);
 	dprintk("%s: done, status = %d\n", __func__, ntohl(status));
@@ -652,7 +681,7 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r
 	struct cb_compound_hdr_res hdr_res = { NULL };
 	struct xdr_stream xdr_in, xdr_out;
 	__be32 *p;
-	__be32 status;
+	__be32 status, drc_status = 0;
 	unsigned int nops = 0;
 
 	dprintk("%s: start\n", __func__);
@@ -672,11 +701,18 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r
 		return rpc_system_err;
 
 	while (status == 0 && nops != hdr_arg.nops) {
-		status = process_op(hdr_arg.minorversion, nops,
-				    rqstp, &xdr_in, argp, &xdr_out, resp);
+		status = process_op(hdr_arg.minorversion, nops, rqstp,
+				    &xdr_in, argp, &xdr_out, resp, &drc_status);
 		nops++;
 	}
 
+	/* Buffer overflow in decode_ops_hdr or encode_ops_hdr. Return
+	* resource error in cb_compound status without returning op */
+	if (unlikely(status == htonl(NFS4ERR_RESOURCE_HDR))) {
+		status = htonl(NFS4ERR_RESOURCE);
+		nops--;
+	}
+
 	*hdr_res.status = status;
 	*hdr_res.nops = htonl(nops);
 	dprintk("%s: done, status = %u\n", __func__, ntohl(status));
@@ -713,6 +749,11 @@ static struct callback_op callback_ops[] = {
 		.decode_args = (callback_decode_arg_t)decode_recallany_args,
 		.res_maxsize = CB_OP_RECALLANY_RES_MAXSZ,
 	},
+	[OP_CB_RECALL_SLOT] = {
+		.process_op = (callback_process_op_t)nfs4_callback_recallslot,
+		.decode_args = (callback_decode_arg_t)decode_recallslot_args,
+		.res_maxsize = CB_OP_RECALLSLOT_RES_MAXSZ,
+	},
 #endif /* CONFIG_NFS_V4_1 */
 };
 
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index ee77713ce68b..2274f1737336 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -164,30 +164,7 @@ error_0:
 	return ERR_PTR(err);
 }
 
-static void nfs4_shutdown_client(struct nfs_client *clp)
-{
-#ifdef CONFIG_NFS_V4
-	if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state))
-		nfs4_kill_renewd(clp);
-	BUG_ON(!RB_EMPTY_ROOT(&clp->cl_state_owners));
-	if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state))
-		nfs_idmap_delete(clp);
-
-	rpc_destroy_wait_queue(&clp->cl_rpcwaitq);
-#endif
-}
-
-/*
- * Destroy the NFS4 callback service
- */
-static void nfs4_destroy_callback(struct nfs_client *clp)
-{
 #ifdef CONFIG_NFS_V4
-	if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state))
-		nfs_callback_down(clp->cl_minorversion);
-#endif /* CONFIG_NFS_V4 */
-}
-
 /*
  * Clears/puts all minor version specific parts from an nfs_client struct
  * reverting it to minorversion 0.
@@ -202,9 +179,33 @@ static void nfs4_clear_client_minor_version(struct nfs_client *clp)
 
 	clp->cl_call_sync = _nfs4_call_sync;
 #endif /* CONFIG_NFS_V4_1 */
+}
 
+/*
+ * Destroy the NFS4 callback service
+ */
+static void nfs4_destroy_callback(struct nfs_client *clp)
+{
+	if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state))
+		nfs_callback_down(clp->cl_minorversion);
+}
+
+static void nfs4_shutdown_client(struct nfs_client *clp)
+{
+	if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state))
+		nfs4_kill_renewd(clp);
+	nfs4_clear_client_minor_version(clp);
 	nfs4_destroy_callback(clp);
+	if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state))
+		nfs_idmap_delete(clp);
+
+	rpc_destroy_wait_queue(&clp->cl_rpcwaitq);
 }
+#else
+static void nfs4_shutdown_client(struct nfs_client *clp)
+{
+}
+#endif /* CONFIG_NFS_V4 */
 
 /*
  * Destroy a shared client record
@@ -213,7 +214,6 @@ static void nfs_free_client(struct nfs_client *clp)
 {
 	dprintk("--> nfs_free_client(%u)\n", clp->rpc_ops->version);
 
-	nfs4_clear_client_minor_version(clp);
 	nfs4_shutdown_client(clp);
 
 	nfs_fscache_release_client_cookie(clp);
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 3c7f03b669fb..a1f6b4438fb1 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -560,7 +560,7 @@ static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir)
 	desc->entry = &my_entry;
 
 	nfs_block_sillyrename(dentry);
-	res = nfs_revalidate_mapping_nolock(inode, filp->f_mapping);
+	res = nfs_revalidate_mapping(inode, filp->f_mapping);
 	if (res < 0)
 		goto out;
 
diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c
index 95e1ca765d47..3f0cd4dfddaf 100644
--- a/fs/nfs/dns_resolve.c
+++ b/fs/nfs/dns_resolve.c
@@ -36,6 +36,19 @@ struct nfs_dns_ent {
 };
 
 
+static void nfs_dns_ent_update(struct cache_head *cnew,
+		struct cache_head *ckey)
+{
+	struct nfs_dns_ent *new;
+	struct nfs_dns_ent *key;
+
+	new = container_of(cnew, struct nfs_dns_ent, h);
+	key = container_of(ckey, struct nfs_dns_ent, h);
+
+	memcpy(&new->addr, &key->addr, key->addrlen);
+	new->addrlen = key->addrlen;
+}
+
 static void nfs_dns_ent_init(struct cache_head *cnew,
 		struct cache_head *ckey)
 {
@@ -49,8 +62,7 @@ static void nfs_dns_ent_init(struct cache_head *cnew,
 	new->hostname = kstrndup(key->hostname, key->namelen, GFP_KERNEL);
 	if (new->hostname) {
 		new->namelen = key->namelen;
-		memcpy(&new->addr, &key->addr, key->addrlen);
-		new->addrlen = key->addrlen;
+		nfs_dns_ent_update(cnew, ckey);
 	} else {
 		new->namelen = 0;
 		new->addrlen = 0;
@@ -234,7 +246,7 @@ static struct cache_detail nfs_dns_resolve = {
 	.cache_show = nfs_dns_show,
 	.match = nfs_dns_match,
 	.init = nfs_dns_ent_init,
-	.update = nfs_dns_ent_init,
+	.update = nfs_dns_ent_update,
 	.alloc = nfs_dns_ent_alloc,
 };
 
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 63f2071d6445..ae8d02294e46 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -123,11 +123,11 @@ nfs_file_open(struct inode *inode, struct file *filp)
 			filp->f_path.dentry->d_parent->d_name.name,
 			filp->f_path.dentry->d_name.name);
 
+	nfs_inc_stats(inode, NFSIOS_VFSOPEN);
 	res = nfs_check_flags(filp->f_flags);
 	if (res)
 		return res;
 
-	nfs_inc_stats(inode, NFSIOS_VFSOPEN);
 	res = nfs_open(inode, filp);
 	return res;
 }
@@ -237,9 +237,9 @@ nfs_file_flush(struct file *file, fl_owner_t id)
 			dentry->d_parent->d_name.name,
 			dentry->d_name.name);
 
+	nfs_inc_stats(inode, NFSIOS_VFSFLUSH);
 	if ((file->f_mode & FMODE_WRITE) == 0)
 		return 0;
-	nfs_inc_stats(inode, NFSIOS_VFSFLUSH);
 
 	/* Flush writes to the server and return any errors */
 	return nfs_do_fsync(ctx, inode);
@@ -262,9 +262,11 @@ nfs_file_read(struct kiocb *iocb, const struct iovec *iov,
 		(unsigned long) count, (unsigned long) pos);
 
 	result = nfs_revalidate_mapping(inode, iocb->ki_filp->f_mapping);
-	nfs_add_stats(inode, NFSIOS_NORMALREADBYTES, count);
-	if (!result)
+	if (!result) {
 		result = generic_file_aio_read(iocb, iov, nr_segs, pos);
+		if (result > 0)
+			nfs_add_stats(inode, NFSIOS_NORMALREADBYTES, result);
+	}
 	return result;
 }
 
@@ -282,8 +284,11 @@ nfs_file_splice_read(struct file *filp, loff_t *ppos,
 		(unsigned long) count, (unsigned long long) *ppos);
 
 	res = nfs_revalidate_mapping(inode, filp->f_mapping);
-	if (!res)
+	if (!res) {
 		res = generic_file_splice_read(filp, ppos, pipe, count, flags);
+		if (res > 0)
+			nfs_add_stats(inode, NFSIOS_NORMALREADBYTES, res);
+	}
 	return res;
 }
 
@@ -596,6 +601,7 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov,
 {
 	struct dentry * dentry = iocb->ki_filp->f_path.dentry;
 	struct inode * inode = dentry->d_inode;
+	unsigned long written = 0;
 	ssize_t result;
 	size_t count = iov_length(iov, nr_segs);
 
@@ -622,14 +628,18 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov,
 	if (!count)
 		goto out;
 
-	nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, count);
 	result = generic_file_aio_write(iocb, iov, nr_segs, pos);
+	if (result > 0)
+		written = result;
+
 	/* Return error values for O_DSYNC and IS_SYNC() */
 	if (result >= 0 && nfs_need_sync_write(iocb->ki_filp, inode)) {
 		int err = nfs_do_fsync(nfs_file_open_context(iocb->ki_filp), inode);
 		if (err < 0)
 			result = err;
 	}
+	if (result > 0)
+		nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, written);
 out:
 	return result;
 
@@ -644,6 +654,7 @@ static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe,
 {
 	struct dentry *dentry = filp->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
+	unsigned long written = 0;
 	ssize_t ret;
 
 	dprintk("NFS splice_write(%s/%s, %lu@%llu)\n",
@@ -654,14 +665,17 @@ static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe,
 	 * The combination of splice and an O_APPEND destination is disallowed.
 	 */
 
-	nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, count);
-
 	ret = generic_file_splice_write(pipe, filp, ppos, count, flags);
+	if (ret > 0)
+		written = ret;
+
 	if (ret >= 0 && nfs_need_sync_write(filp, inode)) {
 		int err = nfs_do_fsync(nfs_file_open_context(filp), inode);
 		if (err < 0)
 			ret = err;
 	}
+	if (ret > 0)
+		nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, written);
 	return ret;
 }
 
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index f141bde7756a..657201acda84 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -97,22 +97,6 @@ u64 nfs_compat_user_ino64(u64 fileid)
 	return ino;
 }
 
-int nfs_write_inode(struct inode *inode, int sync)
-{
-	int ret;
-
-	if (sync) {
-		ret = filemap_fdatawait(inode->i_mapping);
-		if (ret == 0)
-			ret = nfs_commit_inode(inode, FLUSH_SYNC);
-	} else
-		ret = nfs_commit_inode(inode, 0);
-	if (ret >= 0)
-		return 0;
-	__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
-	return ret;
-}
-
 void nfs_clear_inode(struct inode *inode)
 {
 	/*
@@ -130,16 +114,12 @@ void nfs_clear_inode(struct inode *inode)
  */
 int nfs_sync_mapping(struct address_space *mapping)
 {
-	int ret;
+	int ret = 0;
 
-	if (mapping->nrpages == 0)
-		return 0;
-	unmap_mapping_range(mapping, 0, 0, 0);
-	ret = filemap_write_and_wait(mapping);
-	if (ret != 0)
-		goto out;
-	ret = nfs_wb_all(mapping->host);
-out:
+	if (mapping->nrpages != 0) {
+		unmap_mapping_range(mapping, 0, 0, 0);
+		ret = nfs_wb_all(mapping->host);
+	}
 	return ret;
 }
 
@@ -511,17 +491,11 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 	int need_atime = NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATIME;
 	int err;
 
-	/*
-	 * Flush out writes to the server in order to update c/mtime.
-	 *
-	 * Hold the i_mutex to suspend application writes temporarily;
-	 * this prevents long-running writing applications from blocking
-	 * nfs_wb_nocommit.
-	 */
+	/* Flush out writes to the server in order to update c/mtime.  */
 	if (S_ISREG(inode->i_mode)) {
-		mutex_lock(&inode->i_mutex);
-		nfs_wb_nocommit(inode);
-		mutex_unlock(&inode->i_mutex);
+		err = filemap_write_and_wait(inode->i_mapping);
+		if (err)
+			goto out;
 	}
 
 	/*
@@ -545,6 +519,7 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 		generic_fillattr(inode, stat);
 		stat->ino = nfs_compat_user_ino64(NFS_FILEID(inode));
 	}
+out:
 	return err;
 }
 
@@ -574,14 +549,14 @@ void nfs_close_context(struct nfs_open_context *ctx, int is_sync)
 	nfs_revalidate_inode(server, inode);
 }
 
-static struct nfs_open_context *alloc_nfs_open_context(struct vfsmount *mnt, struct dentry *dentry, struct rpc_cred *cred)
+static struct nfs_open_context *alloc_nfs_open_context(struct path *path, struct rpc_cred *cred)
 {
 	struct nfs_open_context *ctx;
 
 	ctx = kmalloc(sizeof(*ctx), GFP_KERNEL);
 	if (ctx != NULL) {
-		ctx->path.dentry = dget(dentry);
-		ctx->path.mnt = mntget(mnt);
+		ctx->path = *path;
+		path_get(&ctx->path);
 		ctx->cred = get_rpccred(cred);
 		ctx->state = NULL;
 		ctx->lockowner = current->files;
@@ -620,11 +595,6 @@ void put_nfs_open_context(struct nfs_open_context *ctx)
 	__put_nfs_open_context(ctx, 0);
 }
 
-static void put_nfs_open_context_sync(struct nfs_open_context *ctx)
-{
-	__put_nfs_open_context(ctx, 1);
-}
-
 /*
  * Ensure that mmap has a recent RPC credential for use when writing out
  * shared pages
@@ -671,7 +641,7 @@ static void nfs_file_clear_open_context(struct file *filp)
 		spin_lock(&inode->i_lock);
 		list_move_tail(&ctx->list, &NFS_I(inode)->open_files);
 		spin_unlock(&inode->i_lock);
-		put_nfs_open_context_sync(ctx);
+		__put_nfs_open_context(ctx, filp->f_flags & O_DIRECT ? 0 : 1);
 	}
 }
 
@@ -686,7 +656,7 @@ int nfs_open(struct inode *inode, struct file *filp)
 	cred = rpc_lookup_cred();
 	if (IS_ERR(cred))
 		return PTR_ERR(cred);
-	ctx = alloc_nfs_open_context(filp->f_path.mnt, filp->f_path.dentry, cred);
+	ctx = alloc_nfs_open_context(&filp->f_path, cred);
 	put_rpccred(cred);
 	if (ctx == NULL)
 		return -ENOMEM;
@@ -779,7 +749,7 @@ int nfs_revalidate_inode(struct nfs_server *server, struct inode *inode)
 	return __nfs_revalidate_inode(server, inode);
 }
 
-static int nfs_invalidate_mapping_nolock(struct inode *inode, struct address_space *mapping)
+static int nfs_invalidate_mapping(struct inode *inode, struct address_space *mapping)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 	
@@ -800,49 +770,10 @@ static int nfs_invalidate_mapping_nolock(struct inode *inode, struct address_spa
 	return 0;
 }
 
-static int nfs_invalidate_mapping(struct inode *inode, struct address_space *mapping)
-{
-	int ret = 0;
-
-	mutex_lock(&inode->i_mutex);
-	if (NFS_I(inode)->cache_validity & NFS_INO_INVALID_DATA) {
-		ret = nfs_sync_mapping(mapping);
-		if (ret == 0)
-			ret = nfs_invalidate_mapping_nolock(inode, mapping);
-	}
-	mutex_unlock(&inode->i_mutex);
-	return ret;
-}
-
-/**
- * nfs_revalidate_mapping_nolock - Revalidate the pagecache
- * @inode - pointer to host inode
- * @mapping - pointer to mapping
- */
-int nfs_revalidate_mapping_nolock(struct inode *inode, struct address_space *mapping)
-{
-	struct nfs_inode *nfsi = NFS_I(inode);
-	int ret = 0;
-
-	if ((nfsi->cache_validity & NFS_INO_REVAL_PAGECACHE)
-			|| nfs_attribute_timeout(inode) || NFS_STALE(inode)) {
-		ret = __nfs_revalidate_inode(NFS_SERVER(inode), inode);
-		if (ret < 0)
-			goto out;
-	}
-	if (nfsi->cache_validity & NFS_INO_INVALID_DATA)
-		ret = nfs_invalidate_mapping_nolock(inode, mapping);
-out:
-	return ret;
-}
-
 /**
  * nfs_revalidate_mapping - Revalidate the pagecache
  * @inode - pointer to host inode
  * @mapping - pointer to mapping
- *
- * This version of the function will take the inode->i_mutex and attempt to
- * flush out all dirty data if it needs to invalidate the page cache.
  */
 int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping)
 {
@@ -1420,6 +1351,7 @@ static void init_once(void *foo)
 	INIT_LIST_HEAD(&nfsi->access_cache_inode_lru);
 	INIT_RADIX_TREE(&nfsi->nfs_page_tree, GFP_ATOMIC);
 	nfsi->npages = 0;
+	nfsi->ncommit = 0;
 	atomic_set(&nfsi->silly_count, 1);
 	INIT_HLIST_HEAD(&nfsi->silly_list);
 	init_waitqueue_head(&nfsi->waitqueue);
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 29e464d23b32..11f82f03c5de 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -211,7 +211,7 @@ extern int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask);
 extern struct workqueue_struct *nfsiod_workqueue;
 extern struct inode *nfs_alloc_inode(struct super_block *sb);
 extern void nfs_destroy_inode(struct inode *);
-extern int nfs_write_inode(struct inode *,int);
+extern int nfs_write_inode(struct inode *, struct writeback_control *);
 extern void nfs_clear_inode(struct inode *);
 #ifdef CONFIG_NFS_V4
 extern void nfs4_clear_inode(struct inode *);
diff --git a/fs/nfs/iostat.h b/fs/nfs/iostat.h
index 46d779abafd3..1d8d5c813b01 100644
--- a/fs/nfs/iostat.h
+++ b/fs/nfs/iostat.h
@@ -57,12 +57,12 @@ static inline void nfs_add_fscache_stats(struct inode *inode,
 }
 #endif
 
-static inline struct nfs_iostats *nfs_alloc_iostats(void)
+static inline struct nfs_iostats __percpu *nfs_alloc_iostats(void)
 {
 	return alloc_percpu(struct nfs_iostats);
 }
 
-static inline void nfs_free_iostats(struct nfs_iostats *stats)
+static inline void nfs_free_iostats(struct nfs_iostats __percpu *stats)
 {
 	if (stats != NULL)
 		free_percpu(stats);
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 3f8881d1a050..24992f0a29f2 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -22,14 +22,14 @@
 
 #define NFSDBG_FACILITY		NFSDBG_PROC
 
-/* A wrapper to handle the EJUKEBOX error message */
+/* A wrapper to handle the EJUKEBOX and EKEYEXPIRED error messages */
 static int
 nfs3_rpc_wrapper(struct rpc_clnt *clnt, struct rpc_message *msg, int flags)
 {
 	int res;
 	do {
 		res = rpc_call_sync(clnt, msg, flags);
-		if (res != -EJUKEBOX)
+		if (res != -EJUKEBOX && res != -EKEYEXPIRED)
 			break;
 		schedule_timeout_killable(NFS_JUKEBOX_RETRY_TIME);
 		res = -ERESTARTSYS;
@@ -42,9 +42,10 @@ nfs3_rpc_wrapper(struct rpc_clnt *clnt, struct rpc_message *msg, int flags)
 static int
 nfs3_async_handle_jukebox(struct rpc_task *task, struct inode *inode)
 {
-	if (task->tk_status != -EJUKEBOX)
+	if (task->tk_status != -EJUKEBOX && task->tk_status != -EKEYEXPIRED)
 		return 0;
-	nfs_inc_stats(inode, NFSIOS_DELAY);
+	if (task->tk_status == -EJUKEBOX)
+		nfs_inc_stats(inode, NFSIOS_DELAY);
 	task->tk_status = 0;
 	rpc_restart_call(task);
 	rpc_delay(task, NFS_JUKEBOX_RETRY_TIME);
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index 0c6fda33d66e..a187200a7aac 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -46,6 +46,7 @@ enum nfs4_client_state {
 	NFS4CLNT_DELEGRETURN,
 	NFS4CLNT_SESSION_RESET,
 	NFS4CLNT_SESSION_DRAINING,
+	NFS4CLNT_RECALL_SLOT,
 };
 
 /*
@@ -280,6 +281,7 @@ extern void nfs4_schedule_state_manager(struct nfs_client *);
 extern int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *state);
 extern int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *state);
 extern void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags);
+extern void nfs41_handle_recall_slot(struct nfs_client *clp);
 extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp);
 extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl);
 extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 375f0fae2c6a..eda74c42d552 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -281,6 +281,7 @@ static int nfs4_handle_exception(const struct nfs_server *server, int errorcode,
 			}
 		case -NFS4ERR_GRACE:
 		case -NFS4ERR_DELAY:
+		case -EKEYEXPIRED:
 			ret = nfs4_delay(server->client, &exception->timeout);
 			if (ret != 0)
 				break;
@@ -418,7 +419,8 @@ static void nfs41_sequence_done(struct nfs_client *clp,
 			clp->cl_last_renewal = timestamp;
 		spin_unlock(&clp->cl_lock);
 		/* Check sequence flags */
-		nfs41_handle_sequence_flag_errors(clp, res->sr_status_flags);
+		if (atomic_read(&clp->cl_count) > 1)
+			nfs41_handle_sequence_flag_errors(clp, res->sr_status_flags);
 	}
 out:
 	/* The session may be reset by one of the error handlers. */
@@ -724,8 +726,8 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
 	p->o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid);
 	if (p->o_arg.seqid == NULL)
 		goto err_free;
-	p->path.mnt = mntget(path->mnt);
-	p->path.dentry = dget(path->dentry);
+	path_get(path);
+	p->path = *path;
 	p->dir = parent;
 	p->owner = sp;
 	atomic_inc(&sp->so_count);
@@ -1163,7 +1165,7 @@ static int nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state
 	int err;
 	do {
 		err = _nfs4_do_open_reclaim(ctx, state);
-		if (err != -NFS4ERR_DELAY)
+		if (err != -NFS4ERR_DELAY && err != -EKEYEXPIRED)
 			break;
 		nfs4_handle_exception(server, err, &exception);
 	} while (exception.retry);
@@ -1582,6 +1584,7 @@ static int nfs4_do_open_expired(struct nfs_open_context *ctx, struct nfs4_state
 			goto out;
 		case -NFS4ERR_GRACE:
 		case -NFS4ERR_DELAY:
+		case -EKEYEXPIRED:
 			nfs4_handle_exception(server, err, &exception);
 			err = 0;
 		}
@@ -1944,8 +1947,8 @@ int nfs4_do_close(struct path *path, struct nfs4_state *state, int wait)
 	calldata->res.seqid = calldata->arg.seqid;
 	calldata->res.server = server;
 	calldata->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE;
-	calldata->path.mnt = mntget(path->mnt);
-	calldata->path.dentry = dget(path->dentry);
+	path_get(path);
+	calldata->path = *path;
 
 	msg.rpc_argp = &calldata->arg,
 	msg.rpc_resp = &calldata->res,
@@ -3145,10 +3148,19 @@ static void nfs4_proc_commit_setup(struct nfs_write_data *data, struct rpc_messa
  * nfs4_proc_async_renew(): This is not one of the nfs_rpc_ops; it is a special
  * standalone procedure for queueing an asynchronous RENEW.
  */
+static void nfs4_renew_release(void *data)
+{
+	struct nfs_client *clp = data;
+
+	if (atomic_read(&clp->cl_count) > 1)
+		nfs4_schedule_state_renewal(clp);
+	nfs_put_client(clp);
+}
+
 static void nfs4_renew_done(struct rpc_task *task, void *data)
 {
-	struct nfs_client *clp = (struct nfs_client *)task->tk_msg.rpc_argp;
-	unsigned long timestamp = (unsigned long)data;
+	struct nfs_client *clp = data;
+	unsigned long timestamp = task->tk_start;
 
 	if (task->tk_status < 0) {
 		/* Unless we're shutting down, schedule state recovery! */
@@ -3164,6 +3176,7 @@ static void nfs4_renew_done(struct rpc_task *task, void *data)
 
 static const struct rpc_call_ops nfs4_renew_ops = {
 	.rpc_call_done = nfs4_renew_done,
+	.rpc_release = nfs4_renew_release,
 };
 
 int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred)
@@ -3174,8 +3187,10 @@ int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred)
 		.rpc_cred	= cred,
 	};
 
+	if (!atomic_inc_not_zero(&clp->cl_count))
+		return -EIO;
 	return rpc_call_async(clp->cl_rpcclient, &msg, RPC_TASK_SOFT,
-			&nfs4_renew_ops, (void *)jiffies);
+			&nfs4_renew_ops, clp);
 }
 
 int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred)
@@ -3452,6 +3467,7 @@ _nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server,
 			if (server)
 				nfs_inc_server_stats(server, NFSIOS_DELAY);
 		case -NFS4ERR_GRACE:
+		case -EKEYEXPIRED:
 			rpc_delay(task, NFS4_POLL_RETRY_MAX);
 			task->tk_status = 0;
 			return -EAGAIN;
@@ -3564,6 +3580,7 @@ int nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred)
 			case -NFS4ERR_RESOURCE:
 				/* The IBM lawyers misread another document! */
 			case -NFS4ERR_DELAY:
+			case -EKEYEXPIRED:
 				err = nfs4_delay(clp->cl_rpcclient, &timeout);
 		}
 	} while (err == 0);
@@ -4179,7 +4196,7 @@ static int nfs4_lock_reclaim(struct nfs4_state *state, struct file_lock *request
 		if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0)
 			return 0;
 		err = _nfs4_do_setlk(state, F_SETLK, request, NFS_LOCK_RECLAIM);
-		if (err != -NFS4ERR_DELAY)
+		if (err != -NFS4ERR_DELAY && err != -EKEYEXPIRED)
 			break;
 		nfs4_handle_exception(server, err, &exception);
 	} while (exception.retry);
@@ -4204,6 +4221,7 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request
 			goto out;
 		case -NFS4ERR_GRACE:
 		case -NFS4ERR_DELAY:
+		case -EKEYEXPIRED:
 			nfs4_handle_exception(server, err, &exception);
 			err = 0;
 		}
@@ -4355,6 +4373,7 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl)
 				err = 0;
 				goto out;
 			case -NFS4ERR_DELAY:
+			case -EKEYEXPIRED:
 				break;
 		}
 		err = nfs4_handle_exception(server, err, &exception);
@@ -4500,7 +4519,7 @@ int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred)
 
 		status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
 
-		if (status != NFS4ERR_CLID_INUSE)
+		if (status != -NFS4ERR_CLID_INUSE)
 			break;
 
 		if (signalled())
@@ -4554,6 +4573,7 @@ static void nfs4_get_lease_time_done(struct rpc_task *task, void *calldata)
 	switch (task->tk_status) {
 	case -NFS4ERR_DELAY:
 	case -NFS4ERR_GRACE:
+	case -EKEYEXPIRED:
 		dprintk("%s Retry: tk_status %d\n", __func__, task->tk_status);
 		rpc_delay(task, NFS4_POLL_RETRY_MIN);
 		task->tk_status = 0;
@@ -4611,26 +4631,32 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo)
 /*
  * Reset a slot table
  */
-static int nfs4_reset_slot_table(struct nfs4_slot_table *tbl, int max_slots,
-		int old_max_slots, int ivalue)
+static int nfs4_reset_slot_table(struct nfs4_slot_table *tbl, u32 max_reqs,
+				 int ivalue)
 {
+	struct nfs4_slot *new = NULL;
 	int i;
 	int ret = 0;
 
-	dprintk("--> %s: max_reqs=%u, tbl %p\n", __func__, max_slots, tbl);
+	dprintk("--> %s: max_reqs=%u, tbl->max_slots %d\n", __func__,
+		max_reqs, tbl->max_slots);
 
-	/*
-	 * Until we have dynamic slot table adjustment, insist
-	 * upon the same slot table size
-	 */
-	if (max_slots != old_max_slots) {
-		dprintk("%s reset slot table does't match old\n",
-			__func__);
-		ret = -EINVAL; /*XXX NFS4ERR_REQ_TOO_BIG ? */
-		goto out;
+	/* Does the newly negotiated max_reqs match the existing slot table? */
+	if (max_reqs != tbl->max_slots) {
+		ret = -ENOMEM;
+		new = kmalloc(max_reqs * sizeof(struct nfs4_slot),
+			      GFP_KERNEL);
+		if (!new)
+			goto out;
+		ret = 0;
+		kfree(tbl->slots);
 	}
 	spin_lock(&tbl->slot_tbl_lock);
-	for (i = 0; i < max_slots; ++i)
+	if (new) {
+		tbl->slots = new;
+		tbl->max_slots = max_reqs;
+	}
+	for (i = 0; i < tbl->max_slots; ++i)
 		tbl->slots[i].seq_nr = ivalue;
 	spin_unlock(&tbl->slot_tbl_lock);
 	dprintk("%s: tbl=%p slots=%p max_slots=%d\n", __func__,
@@ -4648,16 +4674,12 @@ static int nfs4_reset_slot_tables(struct nfs4_session *session)
 	int status;
 
 	status = nfs4_reset_slot_table(&session->fc_slot_table,
-			session->fc_attrs.max_reqs,
-			session->fc_slot_table.max_slots,
-			1);
+			session->fc_attrs.max_reqs, 1);
 	if (status)
 		return status;
 
 	status = nfs4_reset_slot_table(&session->bc_slot_table,
-			session->bc_attrs.max_reqs,
-			session->bc_slot_table.max_slots,
-			0);
+			session->bc_attrs.max_reqs, 0);
 	return status;
 }
 
@@ -4798,16 +4820,14 @@ static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args)
 	args->fc_attrs.headerpadsz = 0;
 	args->fc_attrs.max_rqst_sz = mxrqst_sz;
 	args->fc_attrs.max_resp_sz = mxresp_sz;
-	args->fc_attrs.max_resp_sz_cached = mxresp_sz;
 	args->fc_attrs.max_ops = NFS4_MAX_OPS;
 	args->fc_attrs.max_reqs = session->clp->cl_rpcclient->cl_xprt->max_reqs;
 
 	dprintk("%s: Fore Channel : max_rqst_sz=%u max_resp_sz=%u "
-		"max_resp_sz_cached=%u max_ops=%u max_reqs=%u\n",
+		"max_ops=%u max_reqs=%u\n",
 		__func__,
 		args->fc_attrs.max_rqst_sz, args->fc_attrs.max_resp_sz,
-		args->fc_attrs.max_resp_sz_cached, args->fc_attrs.max_ops,
-		args->fc_attrs.max_reqs);
+		args->fc_attrs.max_ops, args->fc_attrs.max_reqs);
 
 	/* Back channel attributes */
 	args->bc_attrs.headerpadsz = 0;
@@ -5016,7 +5036,16 @@ static int nfs4_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred)
 				       &res, args.sa_cache_this, 1);
 }
 
-void nfs41_sequence_call_done(struct rpc_task *task, void *data)
+static void nfs41_sequence_release(void *data)
+{
+	struct nfs_client *clp = (struct nfs_client *)data;
+
+	if (atomic_read(&clp->cl_count) > 1)
+		nfs4_schedule_state_renewal(clp);
+	nfs_put_client(clp);
+}
+
+static void nfs41_sequence_call_done(struct rpc_task *task, void *data)
 {
 	struct nfs_client *clp = (struct nfs_client *)data;
 
@@ -5024,6 +5053,8 @@ void nfs41_sequence_call_done(struct rpc_task *task, void *data)
 
 	if (task->tk_status < 0) {
 		dprintk("%s ERROR %d\n", __func__, task->tk_status);
+		if (atomic_read(&clp->cl_count) == 1)
+			goto out;
 
 		if (_nfs4_async_handle_error(task, NULL, clp, NULL)
 								== -EAGAIN) {
@@ -5032,7 +5063,7 @@ void nfs41_sequence_call_done(struct rpc_task *task, void *data)
 		}
 	}
 	dprintk("%s rpc_cred %p\n", __func__, task->tk_msg.rpc_cred);
-
+out:
 	kfree(task->tk_msg.rpc_argp);
 	kfree(task->tk_msg.rpc_resp);
 
@@ -5057,6 +5088,7 @@ static void nfs41_sequence_prepare(struct rpc_task *task, void *data)
 static const struct rpc_call_ops nfs41_sequence_ops = {
 	.rpc_call_done = nfs41_sequence_call_done,
 	.rpc_call_prepare = nfs41_sequence_prepare,
+	.rpc_release = nfs41_sequence_release,
 };
 
 static int nfs41_proc_async_sequence(struct nfs_client *clp,
@@ -5069,12 +5101,13 @@ static int nfs41_proc_async_sequence(struct nfs_client *clp,
 		.rpc_cred = cred,
 	};
 
+	if (!atomic_inc_not_zero(&clp->cl_count))
+		return -EIO;
 	args = kzalloc(sizeof(*args), GFP_KERNEL);
-	if (!args)
-		return -ENOMEM;
 	res = kzalloc(sizeof(*res), GFP_KERNEL);
-	if (!res) {
+	if (!args || !res) {
 		kfree(args);
+		nfs_put_client(clp);
 		return -ENOMEM;
 	}
 	res->sr_slotid = NFS4_MAX_SLOT_TABLE;
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index 0156c01c212c..d87f10327b72 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -36,11 +36,6 @@
  * as an rpc_task, not a real kernel thread, so it always runs in rpciod's
  * context.  There is one renewd per nfs_server.
  *
- * TODO: If the send queue gets backlogged (e.g., if the server goes down),
- * we will keep filling the queue with periodic RENEW requests.  We need a
- * mechanism for ensuring that if renewd successfully sends off a request,
- * then it only wakes up when the request is finished.  Maybe use the
- * child task framework of the RPC layer?
  */
 
 #include <linux/mm.h>
@@ -63,7 +58,7 @@ nfs4_renew_state(struct work_struct *work)
 	struct nfs_client *clp =
 		container_of(work, struct nfs_client, cl_renewd.work);
 	struct rpc_cred *cred;
-	long lease, timeout;
+	long lease;
 	unsigned long last, now;
 
 	ops = nfs4_state_renewal_ops[clp->cl_minorversion];
@@ -75,7 +70,6 @@ nfs4_renew_state(struct work_struct *work)
 	lease = clp->cl_lease_time;
 	last = clp->cl_last_renewal;
 	now = jiffies;
-	timeout = (2 * lease) / 3 + (long)last - (long)now;
 	/* Are we close to a lease timeout? */
 	if (time_after(now, last + lease/3)) {
 		cred = ops->get_state_renewal_cred_locked(clp);
@@ -90,19 +84,15 @@ nfs4_renew_state(struct work_struct *work)
 			/* Queue an asynchronous RENEW. */
 			ops->sched_state_renewal(clp, cred);
 			put_rpccred(cred);
+			goto out_exp;
 		}
-		timeout = (2 * lease) / 3;
-		spin_lock(&clp->cl_lock);
-	} else
+	} else {
 		dprintk("%s: failed to call renewd. Reason: lease not expired \n",
 				__func__);
-	if (timeout < 5 * HZ)    /* safeguard */
-		timeout = 5 * HZ;
-	dprintk("%s: requeueing work. Lease period = %ld\n",
-			__func__, (timeout + HZ - 1) / HZ);
-	cancel_delayed_work(&clp->cl_renewd);
-	schedule_delayed_work(&clp->cl_renewd, timeout);
-	spin_unlock(&clp->cl_lock);
+		spin_unlock(&clp->cl_lock);
+	}
+	nfs4_schedule_state_renewal(clp);
+out_exp:
 	nfs_expire_unreferenced_delegations(clp);
 out:
 	dprintk("%s: done\n", __func__);
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index c1e2733f4fa4..6c5ed51f105e 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1249,26 +1249,65 @@ static int nfs4_reclaim_lease(struct nfs_client *clp)
 }
 
 #ifdef CONFIG_NFS_V4_1
+void nfs41_handle_recall_slot(struct nfs_client *clp)
+{
+	set_bit(NFS4CLNT_RECALL_SLOT, &clp->cl_state);
+	nfs4_schedule_state_recovery(clp);
+}
+
+static void nfs4_reset_all_state(struct nfs_client *clp)
+{
+	if (test_and_set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) == 0) {
+		clp->cl_boot_time = CURRENT_TIME;
+		nfs4_state_start_reclaim_nograce(clp);
+		nfs4_schedule_state_recovery(clp);
+	}
+}
+
+static void nfs41_handle_server_reboot(struct nfs_client *clp)
+{
+	if (test_and_set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) == 0) {
+		nfs4_state_start_reclaim_reboot(clp);
+		nfs4_schedule_state_recovery(clp);
+	}
+}
+
+static void nfs41_handle_state_revoked(struct nfs_client *clp)
+{
+	/* Temporary */
+	nfs4_reset_all_state(clp);
+}
+
+static void nfs41_handle_recallable_state_revoked(struct nfs_client *clp)
+{
+	/* This will need to handle layouts too */
+	nfs_expire_all_delegations(clp);
+}
+
+static void nfs41_handle_cb_path_down(struct nfs_client *clp)
+{
+	nfs_expire_all_delegations(clp);
+	if (test_and_set_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state) == 0)
+		nfs4_schedule_state_recovery(clp);
+}
+
 void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags)
 {
 	if (!flags)
 		return;
-	else if (flags & SEQ4_STATUS_RESTART_RECLAIM_NEEDED) {
-		set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
-		nfs4_state_start_reclaim_reboot(clp);
-		nfs4_schedule_state_recovery(clp);
-	} else if (flags & (SEQ4_STATUS_EXPIRED_ALL_STATE_REVOKED |
+	else if (flags & SEQ4_STATUS_RESTART_RECLAIM_NEEDED)
+		nfs41_handle_server_reboot(clp);
+	else if (flags & (SEQ4_STATUS_EXPIRED_ALL_STATE_REVOKED |
 			    SEQ4_STATUS_EXPIRED_SOME_STATE_REVOKED |
 			    SEQ4_STATUS_ADMIN_STATE_REVOKED |
-			    SEQ4_STATUS_RECALLABLE_STATE_REVOKED |
-			    SEQ4_STATUS_LEASE_MOVED)) {
-		set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
-		nfs4_state_start_reclaim_nograce(clp);
-		nfs4_schedule_state_recovery(clp);
-	} else if (flags & (SEQ4_STATUS_CB_PATH_DOWN |
+			    SEQ4_STATUS_LEASE_MOVED))
+		nfs41_handle_state_revoked(clp);
+	else if (flags & SEQ4_STATUS_RECALLABLE_STATE_REVOKED)
+		nfs41_handle_recallable_state_revoked(clp);
+	else if (flags & (SEQ4_STATUS_CB_PATH_DOWN |
 			    SEQ4_STATUS_BACKCHANNEL_FAULT |
 			    SEQ4_STATUS_CB_PATH_DOWN_SESSION))
-		nfs_expire_all_delegations(clp);
+		nfs41_handle_cb_path_down(clp);
 }
 
 static int nfs4_reset_session(struct nfs_client *clp)
@@ -1285,23 +1324,52 @@ static int nfs4_reset_session(struct nfs_client *clp)
 
 	memset(clp->cl_session->sess_id.data, 0, NFS4_MAX_SESSIONID_LEN);
 	status = nfs4_proc_create_session(clp);
-	if (status)
+	if (status) {
 		status = nfs4_recovery_handle_error(clp, status);
+		goto out;
+	}
+	/* create_session negotiated new slot table */
+	clear_bit(NFS4CLNT_RECALL_SLOT, &clp->cl_state);
 
-out:
-	/*
-	 * Let the state manager reestablish state
-	 */
-	if (!test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) &&
-	    status == 0)
+	 /* Let the state manager reestablish state */
+	if (!test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
 		nfs41_setup_state_renewal(clp);
-
+out:
 	return status;
 }
 
+static int nfs4_recall_slot(struct nfs_client *clp)
+{
+	struct nfs4_slot_table *fc_tbl = &clp->cl_session->fc_slot_table;
+	struct nfs4_channel_attrs *fc_attrs = &clp->cl_session->fc_attrs;
+	struct nfs4_slot *new, *old;
+	int i;
+
+	nfs4_begin_drain_session(clp);
+	new = kmalloc(fc_tbl->target_max_slots * sizeof(struct nfs4_slot),
+		      GFP_KERNEL);
+        if (!new)
+		return -ENOMEM;
+
+	spin_lock(&fc_tbl->slot_tbl_lock);
+	for (i = 0; i < fc_tbl->target_max_slots; i++)
+		new[i].seq_nr = fc_tbl->slots[i].seq_nr;
+	old = fc_tbl->slots;
+	fc_tbl->slots = new;
+	fc_tbl->max_slots = fc_tbl->target_max_slots;
+	fc_tbl->target_max_slots = 0;
+	fc_attrs->max_reqs = fc_tbl->max_slots;
+	spin_unlock(&fc_tbl->slot_tbl_lock);
+
+	kfree(old);
+	nfs4_end_drain_session(clp);
+	return 0;
+}
+
 #else /* CONFIG_NFS_V4_1 */
 static int nfs4_reset_session(struct nfs_client *clp) { return 0; }
 static int nfs4_end_drain_session(struct nfs_client *clp) { return 0; }
+static int nfs4_recall_slot(struct nfs_client *clp) { return 0; }
 #endif /* CONFIG_NFS_V4_1 */
 
 /* Set NFS4CLNT_LEASE_EXPIRED for all v4.0 errors and for recoverable errors
@@ -1314,6 +1382,7 @@ static void nfs4_set_lease_expired(struct nfs_client *clp, int status)
 		case -NFS4ERR_DELAY:
 		case -NFS4ERR_CLID_INUSE:
 		case -EAGAIN:
+		case -EKEYEXPIRED:
 			break;
 
 		case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery
@@ -1397,6 +1466,15 @@ static void nfs4_state_manager(struct nfs_client *clp)
 			nfs_client_return_marked_delegations(clp);
 			continue;
 		}
+		/* Recall session slots */
+		if (test_and_clear_bit(NFS4CLNT_RECALL_SLOT, &clp->cl_state)
+		   && nfs4_has_session(clp)) {
+			status = nfs4_recall_slot(clp);
+			if (status < 0)
+				goto out_error;
+			continue;
+		}
+
 
 		nfs4_clear_state_manager_bit(clp);
 		/* Did we race with an attempt to give us more work? */
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 5cd5184b56db..4d338be492cb 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -1578,6 +1578,14 @@ static void encode_create_session(struct xdr_stream *xdr,
 	char machine_name[NFS4_MAX_MACHINE_NAME_LEN];
 	uint32_t len;
 	struct nfs_client *clp = args->client;
+	u32 max_resp_sz_cached;
+
+	/*
+	 * Assumes OPEN is the biggest non-idempotent compound.
+	 * 2 is the verifier.
+	 */
+	max_resp_sz_cached = (NFS4_dec_open_sz + RPC_REPHDRSIZE +
+			      RPC_MAX_AUTH_SIZE + 2) * XDR_UNIT;
 
 	len = scnprintf(machine_name, sizeof(machine_name), "%s",
 			clp->cl_ipaddr);
@@ -1592,7 +1600,7 @@ static void encode_create_session(struct xdr_stream *xdr,
 	*p++ = cpu_to_be32(args->fc_attrs.headerpadsz);	/* header padding size */
 	*p++ = cpu_to_be32(args->fc_attrs.max_rqst_sz);	/* max req size */
 	*p++ = cpu_to_be32(args->fc_attrs.max_resp_sz);	/* max resp size */
-	*p++ = cpu_to_be32(args->fc_attrs.max_resp_sz_cached);	/* Max resp sz cached */
+	*p++ = cpu_to_be32(max_resp_sz_cached);		/* Max resp sz cached */
 	*p++ = cpu_to_be32(args->fc_attrs.max_ops);	/* max operations */
 	*p++ = cpu_to_be32(args->fc_attrs.max_reqs);	/* max requests */
 	*p++ = cpu_to_be32(0);				/* rdmachannel_attrs */
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index ef583854d8d0..c752d944fe9e 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -47,6 +47,39 @@
 #define NFSDBG_FACILITY		NFSDBG_PROC
 
 /*
+ * wrapper to handle the -EKEYEXPIRED error message. This should generally
+ * only happen if using krb5 auth and a user's TGT expires. NFSv2 doesn't
+ * support the NFSERR_JUKEBOX error code, but we handle this situation in the
+ * same way that we handle that error with NFSv3.
+ */
+static int
+nfs_rpc_wrapper(struct rpc_clnt *clnt, struct rpc_message *msg, int flags)
+{
+	int res;
+	do {
+		res = rpc_call_sync(clnt, msg, flags);
+		if (res != -EKEYEXPIRED)
+			break;
+		schedule_timeout_killable(NFS_JUKEBOX_RETRY_TIME);
+		res = -ERESTARTSYS;
+	} while (!fatal_signal_pending(current));
+	return res;
+}
+
+#define rpc_call_sync(clnt, msg, flags)	nfs_rpc_wrapper(clnt, msg, flags)
+
+static int
+nfs_async_handle_expired_key(struct rpc_task *task)
+{
+	if (task->tk_status != -EKEYEXPIRED)
+		return 0;
+	task->tk_status = 0;
+	rpc_restart_call(task);
+	rpc_delay(task, NFS_JUKEBOX_RETRY_TIME);
+	return 1;
+}
+
+/*
  * Bare-bones access to getattr: this is for nfs_read_super.
  */
 static int
@@ -307,6 +340,8 @@ nfs_proc_unlink_setup(struct rpc_message *msg, struct inode *dir)
 
 static int nfs_proc_unlink_done(struct rpc_task *task, struct inode *dir)
 {
+	if (nfs_async_handle_expired_key(task))
+		return 0;
 	nfs_mark_for_revalidate(dir);
 	return 1;
 }
@@ -560,6 +595,9 @@ nfs_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle,
 
 static int nfs_read_done(struct rpc_task *task, struct nfs_read_data *data)
 {
+	if (nfs_async_handle_expired_key(task))
+		return -EAGAIN;
+
 	nfs_invalidate_atime(data->inode);
 	if (task->tk_status >= 0) {
 		nfs_refresh_inode(data->inode, data->res.fattr);
@@ -579,6 +617,9 @@ static void nfs_proc_read_setup(struct nfs_read_data *data, struct rpc_message *
 
 static int nfs_write_done(struct rpc_task *task, struct nfs_write_data *data)
 {
+	if (nfs_async_handle_expired_key(task))
+		return -EAGAIN;
+
 	if (task->tk_status >= 0)
 		nfs_post_op_update_inode_force_wcc(data->inode, data->res.fattr);
 	return 0;
diff --git a/fs/nfs/symlink.c b/fs/nfs/symlink.c
index 412738dbfbc7..2ea9e5c27e55 100644
--- a/fs/nfs/symlink.c
+++ b/fs/nfs/symlink.c
@@ -50,7 +50,7 @@ static void *nfs_follow_link(struct dentry *dentry, struct nameidata *nd)
 	struct page *page;
 	void *err;
 
-	err = ERR_PTR(nfs_revalidate_mapping_nolock(inode, inode->i_mapping));
+	err = ERR_PTR(nfs_revalidate_mapping(inode, inode->i_mapping));
 	if (err)
 		goto read_failed;
 	page = read_cache_page(&inode->i_data, 0,
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index d63d964a0392..53ff70e23993 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -438,6 +438,7 @@ nfs_mark_request_commit(struct nfs_page *req)
 	radix_tree_tag_set(&nfsi->nfs_page_tree,
 			req->wb_index,
 			NFS_PAGE_TAG_COMMIT);
+	nfsi->ncommit++;
 	spin_unlock(&inode->i_lock);
 	inc_zone_page_state(req->wb_page, NR_UNSTABLE_NFS);
 	inc_bdi_stat(req->wb_page->mapping->backing_dev_info, BDI_RECLAIMABLE);
@@ -501,57 +502,6 @@ int nfs_reschedule_unstable_write(struct nfs_page *req)
 }
 #endif
 
-/*
- * Wait for a request to complete.
- *
- * Interruptible by fatal signals only.
- */
-static int nfs_wait_on_requests_locked(struct inode *inode, pgoff_t idx_start, unsigned int npages)
-{
-	struct nfs_inode *nfsi = NFS_I(inode);
-	struct nfs_page *req;
-	pgoff_t idx_end, next;
-	unsigned int		res = 0;
-	int			error;
-
-	if (npages == 0)
-		idx_end = ~0;
-	else
-		idx_end = idx_start + npages - 1;
-
-	next = idx_start;
-	while (radix_tree_gang_lookup_tag(&nfsi->nfs_page_tree, (void **)&req, next, 1, NFS_PAGE_TAG_LOCKED)) {
-		if (req->wb_index > idx_end)
-			break;
-
-		next = req->wb_index + 1;
-		BUG_ON(!NFS_WBACK_BUSY(req));
-
-		kref_get(&req->wb_kref);
-		spin_unlock(&inode->i_lock);
-		error = nfs_wait_on_request(req);
-		nfs_release_request(req);
-		spin_lock(&inode->i_lock);
-		if (error < 0)
-			return error;
-		res++;
-	}
-	return res;
-}
-
-static void nfs_cancel_commit_list(struct list_head *head)
-{
-	struct nfs_page *req;
-
-	while(!list_empty(head)) {
-		req = nfs_list_entry(head->next);
-		nfs_list_remove_request(req);
-		nfs_clear_request_commit(req);
-		nfs_inode_remove_request(req);
-		nfs_unlock_request(req);
-	}
-}
-
 #if defined(CONFIG_NFS_V3) || defined(CONFIG_NFS_V4)
 static int
 nfs_need_commit(struct nfs_inode *nfsi)
@@ -573,11 +523,17 @@ static int
 nfs_scan_commit(struct inode *inode, struct list_head *dst, pgoff_t idx_start, unsigned int npages)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
+	int ret;
 
 	if (!nfs_need_commit(nfsi))
 		return 0;
 
-	return nfs_scan_list(nfsi, dst, idx_start, npages, NFS_PAGE_TAG_COMMIT);
+	ret = nfs_scan_list(nfsi, dst, idx_start, npages, NFS_PAGE_TAG_COMMIT);
+	if (ret > 0)
+		nfsi->ncommit -= ret;
+	if (nfs_need_commit(NFS_I(inode)))
+		__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
+	return ret;
 }
 #else
 static inline int nfs_need_commit(struct nfs_inode *nfsi)
@@ -642,9 +598,10 @@ static struct nfs_page *nfs_try_to_update_request(struct inode *inode,
 		spin_lock(&inode->i_lock);
 	}
 
-	if (nfs_clear_request_commit(req))
-		radix_tree_tag_clear(&NFS_I(inode)->nfs_page_tree,
-				req->wb_index, NFS_PAGE_TAG_COMMIT);
+	if (nfs_clear_request_commit(req) &&
+			radix_tree_tag_clear(&NFS_I(inode)->nfs_page_tree,
+				req->wb_index, NFS_PAGE_TAG_COMMIT) != NULL)
+		NFS_I(inode)->ncommit--;
 
 	/* Okay, the request matches. Update the region */
 	if (offset < req->wb_offset) {
@@ -1391,7 +1348,7 @@ static const struct rpc_call_ops nfs_commit_ops = {
 	.rpc_release = nfs_commit_release,
 };
 
-int nfs_commit_inode(struct inode *inode, int how)
+static int nfs_commit_inode(struct inode *inode, int how)
 {
 	LIST_HEAD(head);
 	int res;
@@ -1406,92 +1363,51 @@ int nfs_commit_inode(struct inode *inode, int how)
 	}
 	return res;
 }
-#else
-static inline int nfs_commit_list(struct inode *inode, struct list_head *head, int how)
-{
-	return 0;
-}
-#endif
 
-long nfs_sync_mapping_wait(struct address_space *mapping, struct writeback_control *wbc, int how)
+static int nfs_commit_unstable_pages(struct inode *inode, struct writeback_control *wbc)
 {
-	struct inode *inode = mapping->host;
-	pgoff_t idx_start, idx_end;
-	unsigned int npages = 0;
-	LIST_HEAD(head);
-	int nocommit = how & FLUSH_NOCOMMIT;
-	long pages, ret;
-
-	/* FIXME */
-	if (wbc->range_cyclic)
-		idx_start = 0;
-	else {
-		idx_start = wbc->range_start >> PAGE_CACHE_SHIFT;
-		idx_end = wbc->range_end >> PAGE_CACHE_SHIFT;
-		if (idx_end > idx_start) {
-			pgoff_t l_npages = 1 + idx_end - idx_start;
-			npages = l_npages;
-			if (sizeof(npages) != sizeof(l_npages) &&
-					(pgoff_t)npages != l_npages)
-				npages = 0;
+	struct nfs_inode *nfsi = NFS_I(inode);
+	int flags = FLUSH_SYNC;
+	int ret = 0;
+
+	/* Don't commit yet if this is a non-blocking flush and there are
+	 * lots of outstanding writes for this mapping.
+	 */
+	if (wbc->sync_mode == WB_SYNC_NONE &&
+	    nfsi->ncommit <= (nfsi->npages >> 1))
+		goto out_mark_dirty;
+
+	if (wbc->nonblocking || wbc->for_background)
+		flags = 0;
+	ret = nfs_commit_inode(inode, flags);
+	if (ret >= 0) {
+		if (wbc->sync_mode == WB_SYNC_NONE) {
+			if (ret < wbc->nr_to_write)
+				wbc->nr_to_write -= ret;
+			else
+				wbc->nr_to_write = 0;
 		}
+		return 0;
 	}
-	how &= ~FLUSH_NOCOMMIT;
-	spin_lock(&inode->i_lock);
-	do {
-		ret = nfs_wait_on_requests_locked(inode, idx_start, npages);
-		if (ret != 0)
-			continue;
-		if (nocommit)
-			break;
-		pages = nfs_scan_commit(inode, &head, idx_start, npages);
-		if (pages == 0)
-			break;
-		if (how & FLUSH_INVALIDATE) {
-			spin_unlock(&inode->i_lock);
-			nfs_cancel_commit_list(&head);
-			ret = pages;
-			spin_lock(&inode->i_lock);
-			continue;
-		}
-		pages += nfs_scan_commit(inode, &head, 0, 0);
-		spin_unlock(&inode->i_lock);
-		ret = nfs_commit_list(inode, &head, how);
-		spin_lock(&inode->i_lock);
-
-	} while (ret >= 0);
-	spin_unlock(&inode->i_lock);
+out_mark_dirty:
+	__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
 	return ret;
 }
-
-static int __nfs_write_mapping(struct address_space *mapping, struct writeback_control *wbc, int how)
+#else
+static int nfs_commit_inode(struct inode *inode, int how)
 {
-	int ret;
-
-	ret = nfs_writepages(mapping, wbc);
-	if (ret < 0)
-		goto out;
-	ret = nfs_sync_mapping_wait(mapping, wbc, how);
-	if (ret < 0)
-		goto out;
 	return 0;
-out:
-	__mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
-	return ret;
 }
 
-/* Two pass sync: first using WB_SYNC_NONE, then WB_SYNC_ALL */
-static int nfs_write_mapping(struct address_space *mapping, int how)
+static int nfs_commit_unstable_pages(struct inode *inode, struct writeback_control *wbc)
 {
-	struct writeback_control wbc = {
-		.bdi = mapping->backing_dev_info,
-		.sync_mode = WB_SYNC_ALL,
-		.nr_to_write = LONG_MAX,
-		.range_start = 0,
-		.range_end = LLONG_MAX,
-	};
+	return 0;
+}
+#endif
 
-	return __nfs_write_mapping(mapping, &wbc, how);
+int nfs_write_inode(struct inode *inode, struct writeback_control *wbc)
+{
+	return nfs_commit_unstable_pages(inode, wbc);
 }
 
 /*
@@ -1499,37 +1415,26 @@ static int nfs_write_mapping(struct address_space *mapping, int how)
  */
 int nfs_wb_all(struct inode *inode)
 {
-	return nfs_write_mapping(inode->i_mapping, 0);
-}
+	struct writeback_control wbc = {
+		.sync_mode = WB_SYNC_ALL,
+		.nr_to_write = LONG_MAX,
+		.range_start = 0,
+		.range_end = LLONG_MAX,
+	};
 
-int nfs_wb_nocommit(struct inode *inode)
-{
-	return nfs_write_mapping(inode->i_mapping, FLUSH_NOCOMMIT);
+	return sync_inode(inode, &wbc);
 }
 
 int nfs_wb_page_cancel(struct inode *inode, struct page *page)
 {
 	struct nfs_page *req;
-	loff_t range_start = page_offset(page);
-	loff_t range_end = range_start + (loff_t)(PAGE_CACHE_SIZE - 1);
-	struct writeback_control wbc = {
-		.bdi = page->mapping->backing_dev_info,
-		.sync_mode = WB_SYNC_ALL,
-		.nr_to_write = LONG_MAX,
-		.range_start = range_start,
-		.range_end = range_end,
-	};
 	int ret = 0;
 
 	BUG_ON(!PageLocked(page));
 	for (;;) {
 		req = nfs_page_find_request(page);
 		if (req == NULL)
-			goto out;
-		if (test_bit(PG_CLEAN, &req->wb_flags)) {
-			nfs_release_request(req);
 			break;
-		}
 		if (nfs_lock_request_dontget(req)) {
 			nfs_inode_remove_request(req);
 			/*
@@ -1543,54 +1448,54 @@ int nfs_wb_page_cancel(struct inode *inode, struct page *page)
 		ret = nfs_wait_on_request(req);
 		nfs_release_request(req);
 		if (ret < 0)
-			goto out;
+			break;
 	}
-	if (!PagePrivate(page))
-		return 0;
-	ret = nfs_sync_mapping_wait(page->mapping, &wbc, FLUSH_INVALIDATE);
-out:
 	return ret;
 }
 
-static int nfs_wb_page_priority(struct inode *inode, struct page *page,
-				int how)
+/*
+ * Write back all requests on one page - we do this before reading it.
+ */
+int nfs_wb_page(struct inode *inode, struct page *page)
 {
 	loff_t range_start = page_offset(page);
 	loff_t range_end = range_start + (loff_t)(PAGE_CACHE_SIZE - 1);
 	struct writeback_control wbc = {
-		.bdi = page->mapping->backing_dev_info,
 		.sync_mode = WB_SYNC_ALL,
-		.nr_to_write = LONG_MAX,
+		.nr_to_write = 0,
 		.range_start = range_start,
 		.range_end = range_end,
 	};
+	struct nfs_page *req;
+	int need_commit;
 	int ret;
 
-	do {
+	while(PagePrivate(page)) {
 		if (clear_page_dirty_for_io(page)) {
 			ret = nfs_writepage_locked(page, &wbc);
 			if (ret < 0)
 				goto out_error;
-		} else if (!PagePrivate(page))
+		}
+		req = nfs_find_and_lock_request(page);
+		if (!req)
 			break;
-		ret = nfs_sync_mapping_wait(page->mapping, &wbc, how);
-		if (ret < 0)
+		if (IS_ERR(req)) {
+			ret = PTR_ERR(req);
 			goto out_error;
-	} while (PagePrivate(page));
+		}
+		need_commit = test_bit(PG_CLEAN, &req->wb_flags);
+		nfs_clear_page_tag_locked(req);
+		if (need_commit) {
+			ret = nfs_commit_inode(inode, FLUSH_SYNC);
+			if (ret < 0)
+				goto out_error;
+		}
+	}
 	return 0;
 out_error:
-	__mark_inode_dirty(inode, I_DIRTY_PAGES);
 	return ret;
 }
 
-/*
- * Write back all requests on one page - we do this before reading it.
- */
-int nfs_wb_page(struct inode *inode, struct page* page)
-{
-	return nfs_wb_page_priority(inode, page, FLUSH_STABLE);
-}
-
 #ifdef CONFIG_MIGRATION
 int nfs_migrate_page(struct address_space *mapping, struct page *newpage,
 		struct page *page)
diff --git a/fs/nfsctl.c b/fs/nfsctl.c
index d3854d94b7cf..bf9cbd242ddd 100644
--- a/fs/nfsctl.c
+++ b/fs/nfsctl.c
@@ -36,10 +36,9 @@ static struct file *do_open(char *name, int flags)
 		return ERR_PTR(error);
 
 	if (flags == O_RDWR)
-		error = may_open(&nd.path, MAY_READ|MAY_WRITE,
-					   FMODE_READ|FMODE_WRITE);
+		error = may_open(&nd.path, MAY_READ|MAY_WRITE, flags);
 	else
-		error = may_open(&nd.path, MAY_WRITE, FMODE_WRITE);
+		error = may_open(&nd.path, MAY_WRITE, flags);
 
 	if (!error)
 		return dentry_open(nd.path.dentry, nd.path.mnt, flags,
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index c6eed2a3b093..4bc22c763de7 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -525,6 +525,8 @@ static struct rpc_cred *callback_cred;
 
 int set_callback_cred(void)
 {
+	if (callback_cred)
+		return 0;
 	callback_cred = rpc_lookup_machine_cred();
 	if (!callback_cred)
 		return -ENOMEM;
@@ -542,7 +544,8 @@ void do_probe_callback(struct nfs4_client *clp)
 	};
 	int status;
 
-	status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_SOFT,
+	status = rpc_call_async(cb->cb_client, &msg,
+				RPC_TASK_SOFT | RPC_TASK_SOFTCONN,
 				&nfsd4_cb_probe_ops, (void *)clp);
 	if (status) {
 		warn_no_callback_path(clp, status);
diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index 5a754f7b71ed..98fb98e330b4 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -119,9 +119,7 @@ out_no_tfm:
 static void
 nfsd4_sync_rec_dir(void)
 {
-	mutex_lock(&rec_dir.dentry->d_inode->i_mutex);
-	nfsd_sync_dir(rec_dir.dentry);
-	mutex_unlock(&rec_dir.dentry->d_inode->i_mutex);
+	vfs_fsync(NULL, rec_dir.dentry, 0);
 }
 
 int
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index f19ed866c95f..c97fddbd17db 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -1998,7 +1998,9 @@ nfs4_file_downgrade(struct file *filp, unsigned int share_access)
 {
 	if (share_access & NFS4_SHARE_ACCESS_WRITE) {
 		drop_file_write_access(filp);
+		spin_lock(&filp->f_lock);
 		filp->f_mode = (filp->f_mode | FMODE_READ) & ~FMODE_WRITE;
+		spin_unlock(&filp->f_lock);
 	}
 }
 
@@ -2480,8 +2482,10 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
 	}
 	memcpy(&open->op_stateid, &stp->st_stateid, sizeof(stateid_t));
 
-	if (nfsd4_has_session(&resp->cstate))
+	if (nfsd4_has_session(&resp->cstate)) {
 		open->op_stateowner->so_confirmed = 1;
+		nfsd4_create_clid_dir(open->op_stateowner->so_client);
+	}
 
 	/*
 	* Attempt to hand out a delegation. No error return, because the
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index a8587e90fd5a..78c7e24e5129 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -1434,7 +1434,7 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp)
 		}
 		op->opnum = ntohl(*argp->p++);
 
-		if (op->opnum >= OP_ACCESS && op->opnum < ops->nops)
+		if (op->opnum >= FIRST_NFS4_OP && op->opnum <= LAST_NFS4_OP)
 			op->status = ops->decoders[op->opnum](argp, &op->u);
 		else {
 			op->opnum = OP_ILLEGAL;
@@ -2121,9 +2121,15 @@ out_acl:
 		 * and this is the root of a cross-mounted filesystem.
 		 */
 		if (ignore_crossmnt == 0 &&
-		    exp->ex_path.mnt->mnt_root->d_inode == dentry->d_inode) {
-			err = vfs_getattr(exp->ex_path.mnt->mnt_parent,
-				exp->ex_path.mnt->mnt_mountpoint, &stat);
+		    dentry == exp->ex_path.mnt->mnt_root) {
+			struct path path = exp->ex_path;
+			path_get(&path);
+			while (follow_up(&path)) {
+				if (path.dentry != path.mnt->mnt_root)
+					break;
+			}
+			err = vfs_getattr(path.mnt, path.dentry, &stat);
+			path_put(&path);
 			if (err)
 				goto out_nfserr;
 		}
diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index 2604c3e70ea5..0f0e77f2012f 100644
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -988,6 +988,7 @@ static ssize_t __write_ports_delfd(char *buf)
 static ssize_t __write_ports_addxprt(char *buf)
 {
 	char transport[16];
+	struct svc_xprt *xprt;
 	int port, err;
 
 	if (sscanf(buf, "%15s %4u", transport, &port) != 2)
@@ -1002,13 +1003,24 @@ static ssize_t __write_ports_addxprt(char *buf)
 
 	err = svc_create_xprt(nfsd_serv, transport,
 				PF_INET, port, SVC_SOCK_ANONYMOUS);
-	if (err < 0) {
-		/* Give a reasonable perror msg for bad transport string */
-		if (err == -ENOENT)
-			err = -EPROTONOSUPPORT;
-		return err;
-	}
+	if (err < 0)
+		goto out_err;
+
+	err = svc_create_xprt(nfsd_serv, transport,
+				PF_INET6, port, SVC_SOCK_ANONYMOUS);
+	if (err < 0 && err != -EAFNOSUPPORT)
+		goto out_close;
 	return 0;
+out_close:
+	xprt = svc_find_xprt(nfsd_serv, transport, PF_INET, port);
+	if (xprt != NULL) {
+		svc_close_xprt(xprt);
+		svc_xprt_put(xprt);
+	}
+out_err:
+	/* Decrease the count, but don't shut down the service */
+	nfsd_serv->sv_nrthreads--;
+	return err;
 }
 
 /*
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 97d79eff6b7f..a11b0e8678ee 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -20,13 +20,14 @@
 #include <linux/fcntl.h>
 #include <linux/namei.h>
 #include <linux/delay.h>
-#include <linux/quotaops.h>
 #include <linux/fsnotify.h>
 #include <linux/posix_acl_xattr.h>
 #include <linux/xattr.h>
 #include <linux/jhash.h>
 #include <linux/ima.h>
 #include <asm/uaccess.h>
+#include <linux/exportfs.h>
+#include <linux/writeback.h>
 
 #ifdef CONFIG_NFSD_V3
 #include "xdr3.h"
@@ -271,6 +272,32 @@ out:
 	return err;
 }
 
+/*
+ * Commit metadata changes to stable storage.
+ */
+static int
+commit_metadata(struct svc_fh *fhp)
+{
+	struct inode *inode = fhp->fh_dentry->d_inode;
+	const struct export_operations *export_ops = inode->i_sb->s_export_op;
+	int error = 0;
+
+	if (!EX_ISSYNC(fhp->fh_export))
+		return 0;
+
+	if (export_ops->commit_metadata) {
+		error = export_ops->commit_metadata(inode);
+	} else {
+		struct writeback_control wbc = {
+			.sync_mode = WB_SYNC_ALL,
+			.nr_to_write = 0, /* metadata only */
+		};
+
+		error = sync_inode(inode, &wbc);
+	}
+
+	return error;
+}
 
 /*
  * Set various file attributes.
@@ -361,7 +388,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
 		 * If we are changing the size of the file, then
 		 * we need to break all leases.
 		 */
-		host_err = break_lease(inode, FMODE_WRITE | O_NONBLOCK);
+		host_err = break_lease(inode, O_WRONLY | O_NONBLOCK);
 		if (host_err == -EWOULDBLOCK)
 			host_err = -ETIMEDOUT;
 		if (host_err) /* ENOMEM or EWOULDBLOCK */
@@ -377,7 +404,6 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
 			put_write_access(inode);
 			goto out_nfserr;
 		}
-		vfs_dq_init(inode);
 	}
 
 	/* sanitize the mode change */
@@ -734,7 +760,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	 * Check to see if there are any leases on this file.
 	 * This may block while leases are broken.
 	 */
-	host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? FMODE_WRITE : 0));
+	host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? O_WRONLY : 0));
 	if (host_err == -EWOULDBLOCK)
 		host_err = -ETIMEDOUT;
 	if (host_err) /* NOMEM or WOULDBLOCK */
@@ -745,14 +771,13 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 			flags = O_RDWR|O_LARGEFILE;
 		else
 			flags = O_WRONLY|O_LARGEFILE;
-
-		vfs_dq_init(inode);
 	}
 	*filp = dentry_open(dget(dentry), mntget(fhp->fh_export->ex_path.mnt),
 			    flags, current_cred());
 	if (IS_ERR(*filp))
 		host_err = PTR_ERR(*filp);
-	host_err = ima_file_check(*filp, access);
+	else
+		host_err = ima_file_check(*filp, access);
 out_nfserr:
 	err = nfserrno(host_err);
 out:
@@ -770,43 +795,6 @@ nfsd_close(struct file *filp)
 }
 
 /*
- * Sync a file
- * As this calls fsync (not fdatasync) there is no need for a write_inode
- * after it.
- */
-static inline int nfsd_dosync(struct file *filp, struct dentry *dp,
-			      const struct file_operations *fop)
-{
-	struct inode *inode = dp->d_inode;
-	int (*fsync) (struct file *, struct dentry *, int);
-	int err;
-
-	err = filemap_write_and_wait(inode->i_mapping);
-	if (err == 0 && fop && (fsync = fop->fsync))
-		err = fsync(filp, dp, 0);
-	return err;
-}
-
-static int
-nfsd_sync(struct file *filp)
-{
-        int err;
-	struct inode *inode = filp->f_path.dentry->d_inode;
-	dprintk("nfsd: sync file %s\n", filp->f_path.dentry->d_name.name);
-	mutex_lock(&inode->i_mutex);
-	err=nfsd_dosync(filp, filp->f_path.dentry, filp->f_op);
-	mutex_unlock(&inode->i_mutex);
-
-	return err;
-}
-
-int
-nfsd_sync_dir(struct dentry *dp)
-{
-	return nfsd_dosync(NULL, dp, dp->d_inode->i_fop);
-}
-
-/*
  * Obtain the readahead parameters for the file
  * specified by (dev, ino).
  */
@@ -1009,7 +997,7 @@ static int wait_for_concurrent_writes(struct file *file)
 
 	if (inode->i_state & I_DIRTY) {
 		dprintk("nfsd: write sync %d\n", task_pid_nr(current));
-		err = nfsd_sync(file);
+		err = vfs_fsync(file, file->f_path.dentry, 0);
 	}
 	last_ino = inode->i_ino;
 	last_dev = inode->i_sb->s_dev;
@@ -1157,8 +1145,9 @@ out:
 #ifdef CONFIG_NFSD_V3
 /*
  * Commit all pending writes to stable storage.
- * Strictly speaking, we could sync just the indicated file region here,
- * but there's currently no way we can ask the VFS to do so.
+ *
+ * Note: we only guarantee that data that lies within the range specified
+ * by the 'offset' and 'count' parameters will be synced.
  *
  * Unfortunately we cannot lock the file to make sure we return full WCC
  * data to the client, as locking happens lower down in the filesystem.
@@ -1168,23 +1157,32 @@ nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp,
                loff_t offset, unsigned long count)
 {
 	struct file	*file;
-	__be32		err;
+	loff_t		end = LLONG_MAX;
+	__be32		err = nfserr_inval;
 
-	if ((u64)count > ~(u64)offset)
-		return nfserr_inval;
+	if (offset < 0)
+		goto out;
+	if (count != 0) {
+		end = offset + (loff_t)count - 1;
+		if (end < offset)
+			goto out;
+	}
 
 	err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_WRITE, &file);
 	if (err)
-		return err;
+		goto out;
 	if (EX_ISSYNC(fhp->fh_export)) {
-		if (file->f_op && file->f_op->fsync) {
-			err = nfserrno(nfsd_sync(file));
-		} else {
+		int err2 = vfs_fsync_range(file, file->f_path.dentry,
+				offset, end, 0);
+
+		if (err2 != -EINVAL)
+			err = nfserrno(err2);
+		else
 			err = nfserr_notsupp;
-		}
 	}
 
 	nfsd_close(file);
+out:
 	return err;
 }
 #endif /* CONFIG_NFSD_V3 */
@@ -1337,12 +1335,14 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		goto out_nfserr;
 	}
 
-	if (EX_ISSYNC(fhp->fh_export)) {
-		err = nfserrno(nfsd_sync_dir(dentry));
-		write_inode_now(dchild->d_inode, 1);
-	}
+	err = nfsd_create_setattr(rqstp, resfhp, iap);
 
-	err2 = nfsd_create_setattr(rqstp, resfhp, iap);
+	/*
+	 * nfsd_setattr already committed the child.  Transactional filesystems
+	 * had a chance to commit changes for both parent and child
+	 * simultaneously making the following commit_metadata a noop.
+	 */
+	err2 = nfserrno(commit_metadata(fhp));
 	if (err2)
 		err = err2;
 	mnt_drop_write(fhp->fh_export->ex_path.mnt);
@@ -1374,7 +1374,6 @@ nfsd_create_v3(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	struct dentry	*dentry, *dchild = NULL;
 	struct inode	*dirp;
 	__be32		err;
-	__be32		err2;
 	int		host_err;
 	__u32		v_mtime=0, v_atime=0;
 
@@ -1469,11 +1468,6 @@ nfsd_create_v3(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	if (created)
 		*created = 1;
 
-	if (EX_ISSYNC(fhp->fh_export)) {
-		err = nfserrno(nfsd_sync_dir(dentry));
-		/* setattr will sync the child (or not) */
-	}
-
 	nfsd_check_ignore_resizing(iap);
 
 	if (createmode == NFS3_CREATE_EXCLUSIVE) {
@@ -1488,9 +1482,13 @@ nfsd_create_v3(struct svc_rqst *rqstp, struct svc_fh *fhp,
 	}
 
  set_attr:
-	err2 = nfsd_create_setattr(rqstp, resfhp, iap);
-	if (err2)
-		err = err2;
+	err = nfsd_create_setattr(rqstp, resfhp, iap);
+
+	/*
+	 * nfsd_setattr already committed the child (and possibly also the parent).
+	 */
+	if (!err)
+		err = nfserrno(commit_metadata(fhp));
 
 	mnt_drop_write(fhp->fh_export->ex_path.mnt);
 	/*
@@ -1605,12 +1603,9 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		}
 	} else
 		host_err = vfs_symlink(dentry->d_inode, dnew, path);
-
-	if (!host_err) {
-		if (EX_ISSYNC(fhp->fh_export))
-			host_err = nfsd_sync_dir(dentry);
-	}
 	err = nfserrno(host_err);
+	if (!err)
+		err = nfserrno(commit_metadata(fhp));
 	fh_unlock(fhp);
 
 	mnt_drop_write(fhp->fh_export->ex_path.mnt);
@@ -1672,11 +1667,9 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 	}
 	host_err = vfs_link(dold, dirp, dnew);
 	if (!host_err) {
-		if (EX_ISSYNC(ffhp->fh_export)) {
-			err = nfserrno(nfsd_sync_dir(ddir));
-			write_inode_now(dest, 1);
-		}
-		err = 0;
+		err = nfserrno(commit_metadata(ffhp));
+		if (!err)
+			err = nfserrno(commit_metadata(tfhp));
 	} else {
 		if (host_err == -EXDEV && rqstp->rq_vers == 2)
 			err = nfserr_acces;
@@ -1772,10 +1765,10 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
 		goto out_dput_new;
 
 	host_err = vfs_rename(fdir, odentry, tdir, ndentry);
-	if (!host_err && EX_ISSYNC(tfhp->fh_export)) {
-		host_err = nfsd_sync_dir(tdentry);
+	if (!host_err) {
+		host_err = commit_metadata(tfhp);
 		if (!host_err)
-			host_err = nfsd_sync_dir(fdentry);
+			host_err = commit_metadata(ffhp);
 	}
 
 	mnt_drop_write(ffhp->fh_export->ex_path.mnt);
@@ -1856,12 +1849,9 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 
 	dput(rdentry);
 
-	if (host_err)
-		goto out_drop;
-	if (EX_ISSYNC(fhp->fh_export))
-		host_err = nfsd_sync_dir(dentry);
+	if (!host_err)
+		host_err = commit_metadata(fhp);
 
-out_drop:
 	mnt_drop_write(fhp->fh_export->ex_path.mnt);
 out_nfserr:
 	err = nfserrno(host_err);
diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c
index 187dd07ba86c..9d1e5de91afb 100644
--- a/fs/nilfs2/dat.c
+++ b/fs/nilfs2/dat.c
@@ -388,8 +388,7 @@ int nilfs_dat_translate(struct inode *dat, __u64 vblocknr, sector_t *blocknrp)
 		ret = -ENOENT;
 		goto out;
 	}
-	if (blocknrp != NULL)
-		*blocknrp = blocknr;
+	*blocknrp = blocknr;
 
  out:
 	kunmap_atomic(kaddr, KM_USER0);
diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c
index 76d803e060a9..0092840492ee 100644
--- a/fs/nilfs2/dir.c
+++ b/fs/nilfs2/dir.c
@@ -224,7 +224,7 @@ fail:
  * len <= NILFS_NAME_LEN and de != NULL are guaranteed by caller.
  */
 static int
-nilfs_match(int len, const char * const name, struct nilfs_dir_entry *de)
+nilfs_match(int len, const unsigned char *name, struct nilfs_dir_entry *de)
 {
 	if (len != de->name_len)
 		return 0;
@@ -349,11 +349,11 @@ done:
  * Entry is guaranteed to be valid.
  */
 struct nilfs_dir_entry *
-nilfs_find_entry(struct inode *dir, struct dentry *dentry,
+nilfs_find_entry(struct inode *dir, const struct qstr *qstr,
 		 struct page **res_page)
 {
-	const char *name = dentry->d_name.name;
-	int namelen = dentry->d_name.len;
+	const unsigned char *name = qstr->name;
+	int namelen = qstr->len;
 	unsigned reclen = NILFS_DIR_REC_LEN(namelen);
 	unsigned long start, n;
 	unsigned long npages = dir_pages(dir);
@@ -424,13 +424,13 @@ struct nilfs_dir_entry *nilfs_dotdot(struct inode *dir, struct page **p)
 	return de;
 }
 
-ino_t nilfs_inode_by_name(struct inode *dir, struct dentry *dentry)
+ino_t nilfs_inode_by_name(struct inode *dir, const struct qstr *qstr)
 {
 	ino_t res = 0;
 	struct nilfs_dir_entry *de;
 	struct page *page;
 
-	de = nilfs_find_entry(dir, dentry, &page);
+	de = nilfs_find_entry(dir, qstr, &page);
 	if (de) {
 		res = le64_to_cpu(de->inode);
 		kunmap(page);
@@ -465,7 +465,7 @@ void nilfs_set_link(struct inode *dir, struct nilfs_dir_entry *de,
 int nilfs_add_link(struct dentry *dentry, struct inode *inode)
 {
 	struct inode *dir = dentry->d_parent->d_inode;
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	int namelen = dentry->d_name.len;
 	unsigned chunk_size = nilfs_chunk_size(dir);
 	unsigned reclen = NILFS_DIR_REC_LEN(namelen);
diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c
index d6b2b83de363..313d0a21da48 100644
--- a/fs/nilfs2/ioctl.c
+++ b/fs/nilfs2/ioctl.c
@@ -26,6 +26,7 @@
 #include <linux/capability.h>	/* capable() */
 #include <linux/uaccess.h>	/* copy_from_user(), copy_to_user() */
 #include <linux/vmalloc.h>
+#include <linux/mount.h>	/* mnt_want_write(), mnt_drop_write() */
 #include <linux/nilfs2_fs.h>
 #include "nilfs.h"
 #include "segment.h"
@@ -107,20 +108,28 @@ static int nilfs_ioctl_change_cpmode(struct inode *inode, struct file *filp,
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
+
+	ret = mnt_want_write(filp->f_path.mnt);
+	if (ret)
+		return ret;
+
+	ret = -EFAULT;
 	if (copy_from_user(&cpmode, argp, sizeof(cpmode)))
-		return -EFAULT;
+		goto out;
 
 	mutex_lock(&nilfs->ns_mount_mutex);
+
 	nilfs_transaction_begin(inode->i_sb, &ti, 0);
 	ret = nilfs_cpfile_change_cpmode(
 		cpfile, cpmode.cm_cno, cpmode.cm_mode);
-	if (unlikely(ret < 0)) {
+	if (unlikely(ret < 0))
 		nilfs_transaction_abort(inode->i_sb);
-		mutex_unlock(&nilfs->ns_mount_mutex);
-		return ret;
-	}
-	nilfs_transaction_commit(inode->i_sb); /* never fails */
+	else
+		nilfs_transaction_commit(inode->i_sb); /* never fails */
+
 	mutex_unlock(&nilfs->ns_mount_mutex);
+out:
+	mnt_drop_write(filp->f_path.mnt);
 	return ret;
 }
 
@@ -135,16 +144,23 @@ nilfs_ioctl_delete_checkpoint(struct inode *inode, struct file *filp,
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
+
+	ret = mnt_want_write(filp->f_path.mnt);
+	if (ret)
+		return ret;
+
+	ret = -EFAULT;
 	if (copy_from_user(&cno, argp, sizeof(cno)))
-		return -EFAULT;
+		goto out;
 
 	nilfs_transaction_begin(inode->i_sb, &ti, 0);
 	ret = nilfs_cpfile_delete_checkpoint(cpfile, cno);
-	if (unlikely(ret < 0)) {
+	if (unlikely(ret < 0))
 		nilfs_transaction_abort(inode->i_sb);
-		return ret;
-	}
-	nilfs_transaction_commit(inode->i_sb); /* never fails */
+	else
+		nilfs_transaction_commit(inode->i_sb); /* never fails */
+out:
+	mnt_drop_write(filp->f_path.mnt);
 	return ret;
 }
 
@@ -496,12 +512,19 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp,
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
+	ret = mnt_want_write(filp->f_path.mnt);
+	if (ret)
+		return ret;
+
+	ret = -EFAULT;
 	if (copy_from_user(argv, argp, sizeof(argv)))
-		return -EFAULT;
+		goto out;
 
+	ret = -EINVAL;
 	nsegs = argv[4].v_nmembs;
 	if (argv[4].v_size != argsz[4])
-		return -EINVAL;
+		goto out;
+
 	/*
 	 * argv[4] points to segment numbers this ioctl cleans.  We
 	 * use kmalloc() for its buffer because memory used for the
@@ -509,9 +532,10 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp,
 	 */
 	kbufs[4] = memdup_user((void __user *)(unsigned long)argv[4].v_base,
 			       nsegs * sizeof(__u64));
-	if (IS_ERR(kbufs[4]))
-		return PTR_ERR(kbufs[4]);
-
+	if (IS_ERR(kbufs[4])) {
+		ret = PTR_ERR(kbufs[4]);
+		goto out;
+	}
 	nilfs = NILFS_SB(inode->i_sb)->s_nilfs;
 
 	for (n = 0; n < 4; n++) {
@@ -563,10 +587,12 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp,
 		nilfs_remove_all_gcinode(nilfs);
 	clear_nilfs_gc_running(nilfs);
 
- out_free:
+out_free:
 	while (--n >= 0)
 		vfree(kbufs[n]);
 	kfree(kbufs[4]);
+out:
+	mnt_drop_write(filp->f_path.mnt);
 	return ret;
 }
 
@@ -575,13 +601,17 @@ static int nilfs_ioctl_sync(struct inode *inode, struct file *filp,
 {
 	__u64 cno;
 	int ret;
+	struct the_nilfs *nilfs;
 
 	ret = nilfs_construct_segment(inode->i_sb);
 	if (ret < 0)
 		return ret;
 
 	if (argp != NULL) {
-		cno = NILFS_SB(inode->i_sb)->s_nilfs->ns_cno - 1;
+		nilfs = NILFS_SB(inode->i_sb)->s_nilfs;
+		down_read(&nilfs->ns_segctor_sem);
+		cno = nilfs->ns_cno - 1;
+		up_read(&nilfs->ns_segctor_sem);
 		if (copy_to_user(argp, &cno, sizeof(cno)))
 			return -EFAULT;
 	}
diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c
index 07ba838ef089..ad6ed2cf19b4 100644
--- a/fs/nilfs2/namei.c
+++ b/fs/nilfs2/namei.c
@@ -67,7 +67,7 @@ nilfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
 	if (dentry->d_name.len > NILFS_NAME_LEN)
 		return ERR_PTR(-ENAMETOOLONG);
 
-	ino = nilfs_inode_by_name(dir, dentry);
+	ino = nilfs_inode_by_name(dir, &dentry->d_name);
 	inode = NULL;
 	if (ino) {
 		inode = nilfs_iget(dir->i_sb, ino);
@@ -81,10 +81,7 @@ struct dentry *nilfs_get_parent(struct dentry *child)
 {
 	unsigned long ino;
 	struct inode *inode;
-	struct dentry dotdot;
-
-	dotdot.d_name.name = "..";
-	dotdot.d_name.len = 2;
+	struct qstr dotdot = {.name = "..", .len = 2};
 
 	ino = nilfs_inode_by_name(child->d_inode, &dotdot);
 	if (!ino)
@@ -296,7 +293,7 @@ static int nilfs_do_unlink(struct inode *dir, struct dentry *dentry)
 	int err;
 
 	err = -ENOENT;
-	de = nilfs_find_entry(dir, dentry, &page);
+	de = nilfs_find_entry(dir, &dentry->d_name, &page);
 	if (!de)
 		goto out;
 
@@ -389,7 +386,7 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		return err;
 
 	err = -ENOENT;
-	old_de = nilfs_find_entry(old_dir, old_dentry, &old_page);
+	old_de = nilfs_find_entry(old_dir, &old_dentry->d_name, &old_page);
 	if (!old_de)
 		goto out;
 
@@ -409,7 +406,7 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 			goto out_dir;
 
 		err = -ENOENT;
-		new_de = nilfs_find_entry(new_dir, new_dentry, &new_page);
+		new_de = nilfs_find_entry(new_dir, &new_dentry->d_name, &new_page);
 		if (!new_de)
 			goto out_dir;
 		inc_nlink(old_inode);
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index 4da6f67e9a91..8723e5bfd071 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -217,10 +217,10 @@ static inline int nilfs_init_acl(struct inode *inode, struct inode *dir)
 
 /* dir.c */
 extern int nilfs_add_link(struct dentry *, struct inode *);
-extern ino_t nilfs_inode_by_name(struct inode *, struct dentry *);
+extern ino_t nilfs_inode_by_name(struct inode *, const struct qstr *);
 extern int nilfs_make_empty(struct inode *, struct inode *);
 extern struct nilfs_dir_entry *
-nilfs_find_entry(struct inode *, struct dentry *, struct page **);
+nilfs_find_entry(struct inode *, const struct qstr *, struct page **);
 extern int nilfs_delete_entry(struct nilfs_dir_entry *, struct page *);
 extern int nilfs_empty_dir(struct inode *);
 extern struct nilfs_dir_entry *nilfs_dotdot(struct inode *, struct page **);
diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c
index c9c96c7825dc..017bedc761a0 100644
--- a/fs/nilfs2/recovery.c
+++ b/fs/nilfs2/recovery.c
@@ -39,7 +39,6 @@ enum {
 	NILFS_SEG_FAIL_IO,
 	NILFS_SEG_FAIL_MAGIC,
 	NILFS_SEG_FAIL_SEQ,
-	NILFS_SEG_FAIL_CHECKSUM_SEGSUM,
 	NILFS_SEG_FAIL_CHECKSUM_SUPER_ROOT,
 	NILFS_SEG_FAIL_CHECKSUM_FULL,
 	NILFS_SEG_FAIL_CONSISTENCY,
@@ -71,10 +70,6 @@ static int nilfs_warn_segment_error(int err)
 		printk(KERN_WARNING
 		       "NILFS warning: Sequence number mismatch\n");
 		break;
-	case NILFS_SEG_FAIL_CHECKSUM_SEGSUM:
-		printk(KERN_WARNING
-		       "NILFS warning: Checksum error in segment summary\n");
-		break;
 	case NILFS_SEG_FAIL_CHECKSUM_SUPER_ROOT:
 		printk(KERN_WARNING
 		       "NILFS warning: Checksum error in super root\n");
@@ -206,19 +201,15 @@ int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block,
  * @pseg_start: start disk block number of partial segment
  * @seg_seq: sequence number requested
  * @ssi: pointer to nilfs_segsum_info struct to store information
- * @full_check: full check flag
- *              (0: only checks segment summary CRC, 1: data CRC)
  */
 static int
 load_segment_summary(struct nilfs_sb_info *sbi, sector_t pseg_start,
-		     u64 seg_seq, struct nilfs_segsum_info *ssi,
-		     int full_check)
+		     u64 seg_seq, struct nilfs_segsum_info *ssi)
 {
 	struct buffer_head *bh_sum;
 	struct nilfs_segment_summary *sum;
-	unsigned long offset, nblock;
-	u64 check_bytes;
-	u32 crc, crc_sum;
+	unsigned long nblock;
+	u32 crc;
 	int ret = NILFS_SEG_FAIL_IO;
 
 	bh_sum = sb_bread(sbi->s_super, pseg_start);
@@ -237,34 +228,24 @@ load_segment_summary(struct nilfs_sb_info *sbi, sector_t pseg_start,
 		ret = NILFS_SEG_FAIL_SEQ;
 		goto failed;
 	}
-	if (full_check) {
-		offset = sizeof(sum->ss_datasum);
-		check_bytes =
-			((u64)ssi->nblocks << sbi->s_super->s_blocksize_bits);
-		nblock = ssi->nblocks;
-		crc_sum = le32_to_cpu(sum->ss_datasum);
-		ret = NILFS_SEG_FAIL_CHECKSUM_FULL;
-	} else { /* only checks segment summary */
-		offset = sizeof(sum->ss_datasum) + sizeof(sum->ss_sumsum);
-		check_bytes = ssi->sumbytes;
-		nblock = ssi->nsumblk;
-		crc_sum = le32_to_cpu(sum->ss_sumsum);
-		ret = NILFS_SEG_FAIL_CHECKSUM_SEGSUM;
-	}
 
+	nblock = ssi->nblocks;
 	if (unlikely(nblock == 0 ||
 		     nblock > sbi->s_nilfs->ns_blocks_per_segment)) {
 		/* This limits the number of blocks read in the CRC check */
 		ret = NILFS_SEG_FAIL_CONSISTENCY;
 		goto failed;
 	}
-	if (calc_crc_cont(sbi, bh_sum, &crc, offset, check_bytes,
+	if (calc_crc_cont(sbi, bh_sum, &crc, sizeof(sum->ss_datasum),
+			  ((u64)nblock << sbi->s_super->s_blocksize_bits),
 			  pseg_start, nblock)) {
 		ret = NILFS_SEG_FAIL_IO;
 		goto failed;
 	}
-	if (crc == crc_sum)
+	if (crc == le32_to_cpu(sum->ss_datasum))
 		ret = 0;
+	else
+		ret = NILFS_SEG_FAIL_CHECKSUM_FULL;
  failed:
 	brelse(bh_sum);
  out:
@@ -598,7 +579,7 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs,
 
 	while (segnum != ri->ri_segnum || pseg_start <= ri->ri_pseg_start) {
 
-		ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi, 1);
+		ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi);
 		if (ret) {
 			if (ret == NILFS_SEG_FAIL_IO) {
 				err = -EIO;
@@ -821,7 +802,7 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi,
 
 	for (;;) {
 		/* Load segment summary */
-		ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi, 1);
+		ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi);
 		if (ret) {
 			if (ret == NILFS_SEG_FAIL_IO)
 				goto failed;
diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c
index 645c78656aa0..ab56fe44e377 100644
--- a/fs/nilfs2/segbuf.c
+++ b/fs/nilfs2/segbuf.c
@@ -40,6 +40,11 @@ struct nilfs_write_info {
 };
 
 
+static int nilfs_segbuf_write(struct nilfs_segment_buffer *segbuf,
+			      struct the_nilfs *nilfs);
+static int nilfs_segbuf_wait(struct nilfs_segment_buffer *segbuf);
+
+
 static struct kmem_cache *nilfs_segbuf_cachep;
 
 static void nilfs_segbuf_init_once(void *obj)
@@ -302,6 +307,19 @@ void nilfs_truncate_logs(struct list_head *logs,
 	}
 }
 
+int nilfs_write_logs(struct list_head *logs, struct the_nilfs *nilfs)
+{
+	struct nilfs_segment_buffer *segbuf;
+	int ret = 0;
+
+	list_for_each_entry(segbuf, logs, sb_list) {
+		ret = nilfs_segbuf_write(segbuf, nilfs);
+		if (ret)
+			break;
+	}
+	return ret;
+}
+
 int nilfs_wait_on_logs(struct list_head *logs)
 {
 	struct nilfs_segment_buffer *segbuf;
diff --git a/fs/nilfs2/segbuf.h b/fs/nilfs2/segbuf.h
index 6af1630fb401..94dfd3517bc0 100644
--- a/fs/nilfs2/segbuf.h
+++ b/fs/nilfs2/segbuf.h
@@ -166,13 +166,10 @@ nilfs_segbuf_add_file_buffer(struct nilfs_segment_buffer *segbuf,
 	segbuf->sb_sum.nfileblk++;
 }
 
-int nilfs_segbuf_write(struct nilfs_segment_buffer *segbuf,
-		       struct the_nilfs *nilfs);
-int nilfs_segbuf_wait(struct nilfs_segment_buffer *segbuf);
-
 void nilfs_clear_logs(struct list_head *logs);
 void nilfs_truncate_logs(struct list_head *logs,
 			 struct nilfs_segment_buffer *last);
+int nilfs_write_logs(struct list_head *logs, struct the_nilfs *nilfs);
 int nilfs_wait_on_logs(struct list_head *logs);
 
 static inline void nilfs_destroy_logs(struct list_head *logs)
diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index 105b508b47a8..ada2f1b947a3 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -1764,14 +1764,9 @@ static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci,
 static int nilfs_segctor_write(struct nilfs_sc_info *sci,
 			       struct the_nilfs *nilfs)
 {
-	struct nilfs_segment_buffer *segbuf;
-	int ret = 0;
+	int ret;
 
-	list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) {
-		ret = nilfs_segbuf_write(segbuf, nilfs);
-		if (ret)
-			break;
-	}
+	ret = nilfs_write_logs(&sci->sc_segbufs, nilfs);
 	list_splice_tail_init(&sci->sc_segbufs, &sci->sc_write_logs);
 	return ret;
 }
@@ -1937,8 +1932,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci)
 {
 	struct nilfs_segment_buffer *segbuf;
 	struct page *bd_page = NULL, *fs_page = NULL;
-	struct nilfs_sb_info *sbi = sci->sc_sbi;
-	struct the_nilfs *nilfs = sbi->s_nilfs;
+	struct the_nilfs *nilfs = sci->sc_sbi->s_nilfs;
 	int update_sr = (sci->sc_super_root != NULL);
 
 	list_for_each_entry(segbuf, &sci->sc_write_logs, sb_list) {
@@ -2020,7 +2014,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci)
 	if (update_sr) {
 		nilfs_set_last_segment(nilfs, segbuf->sb_pseg_start,
 				       segbuf->sb_sum.seg_seq, nilfs->ns_cno++);
-		sbi->s_super->s_dirt = 1;
+		set_nilfs_sb_dirty(nilfs);
 
 		clear_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags);
 		clear_bit(NILFS_SC_DIRTY, &sci->sc_flags);
@@ -2425,43 +2419,43 @@ int nilfs_construct_dsync_segment(struct super_block *sb, struct inode *inode,
 	return err;
 }
 
-struct nilfs_segctor_req {
-	int mode;
-	__u32 seq_accepted;
-	int sc_err;  /* construction failure */
-	int sb_err;  /* super block writeback failure */
-};
-
 #define FLUSH_FILE_BIT	(0x1) /* data file only */
 #define FLUSH_DAT_BIT	(1 << NILFS_DAT_INO) /* DAT only */
 
-static void nilfs_segctor_accept(struct nilfs_sc_info *sci,
-				 struct nilfs_segctor_req *req)
+/**
+ * nilfs_segctor_accept - record accepted sequence count of log-write requests
+ * @sci: segment constructor object
+ */
+static void nilfs_segctor_accept(struct nilfs_sc_info *sci)
 {
-	req->sc_err = req->sb_err = 0;
 	spin_lock(&sci->sc_state_lock);
-	req->seq_accepted = sci->sc_seq_request;
+	sci->sc_seq_accepted = sci->sc_seq_request;
 	spin_unlock(&sci->sc_state_lock);
 
 	if (sci->sc_timer)
 		del_timer_sync(sci->sc_timer);
 }
 
-static void nilfs_segctor_notify(struct nilfs_sc_info *sci,
-				 struct nilfs_segctor_req *req)
+/**
+ * nilfs_segctor_notify - notify the result of request to caller threads
+ * @sci: segment constructor object
+ * @mode: mode of log forming
+ * @err: error code to be notified
+ */
+static void nilfs_segctor_notify(struct nilfs_sc_info *sci, int mode, int err)
 {
 	/* Clear requests (even when the construction failed) */
 	spin_lock(&sci->sc_state_lock);
 
-	if (req->mode == SC_LSEG_SR) {
+	if (mode == SC_LSEG_SR) {
 		sci->sc_state &= ~NILFS_SEGCTOR_COMMIT;
-		sci->sc_seq_done = req->seq_accepted;
-		nilfs_segctor_wakeup(sci, req->sc_err ? : req->sb_err);
+		sci->sc_seq_done = sci->sc_seq_accepted;
+		nilfs_segctor_wakeup(sci, err);
 		sci->sc_flush_request = 0;
 	} else {
-		if (req->mode == SC_FLUSH_FILE)
+		if (mode == SC_FLUSH_FILE)
 			sci->sc_flush_request &= ~FLUSH_FILE_BIT;
-		else if (req->mode == SC_FLUSH_DAT)
+		else if (mode == SC_FLUSH_DAT)
 			sci->sc_flush_request &= ~FLUSH_DAT_BIT;
 
 		/* re-enable timer if checkpoint creation was not done */
@@ -2472,30 +2466,37 @@ static void nilfs_segctor_notify(struct nilfs_sc_info *sci,
 	spin_unlock(&sci->sc_state_lock);
 }
 
-static int nilfs_segctor_construct(struct nilfs_sc_info *sci,
-				   struct nilfs_segctor_req *req)
+/**
+ * nilfs_segctor_construct - form logs and write them to disk
+ * @sci: segment constructor object
+ * @mode: mode of log forming
+ */
+static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode)
 {
 	struct nilfs_sb_info *sbi = sci->sc_sbi;
 	struct the_nilfs *nilfs = sbi->s_nilfs;
 	int err = 0;
 
+	nilfs_segctor_accept(sci);
+
 	if (nilfs_discontinued(nilfs))
-		req->mode = SC_LSEG_SR;
-	if (!nilfs_segctor_confirm(sci)) {
-		err = nilfs_segctor_do_construct(sci, req->mode);
-		req->sc_err = err;
-	}
+		mode = SC_LSEG_SR;
+	if (!nilfs_segctor_confirm(sci))
+		err = nilfs_segctor_do_construct(sci, mode);
+
 	if (likely(!err)) {
-		if (req->mode != SC_FLUSH_DAT)
+		if (mode != SC_FLUSH_DAT)
 			atomic_set(&nilfs->ns_ndirtyblks, 0);
 		if (test_bit(NILFS_SC_SUPER_ROOT, &sci->sc_flags) &&
 		    nilfs_discontinued(nilfs)) {
 			down_write(&nilfs->ns_sem);
-			req->sb_err = nilfs_commit_super(sbi,
-					nilfs_altsb_need_update(nilfs));
+			err = nilfs_commit_super(
+				sbi, nilfs_altsb_need_update(nilfs));
 			up_write(&nilfs->ns_sem);
 		}
 	}
+
+	nilfs_segctor_notify(sci, mode, err);
 	return err;
 }
 
@@ -2526,7 +2527,6 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv,
 	struct nilfs_sc_info *sci = NILFS_SC(sbi);
 	struct the_nilfs *nilfs = sbi->s_nilfs;
 	struct nilfs_transaction_info ti;
-	struct nilfs_segctor_req req = { .mode = SC_LSEG_SR };
 	int err;
 
 	if (unlikely(!sci))
@@ -2547,10 +2547,8 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv,
 	list_splice_tail_init(&nilfs->ns_gc_inodes, &sci->sc_gc_inodes);
 
 	for (;;) {
-		nilfs_segctor_accept(sci, &req);
-		err = nilfs_segctor_construct(sci, &req);
+		err = nilfs_segctor_construct(sci, SC_LSEG_SR);
 		nilfs_remove_written_gcinodes(nilfs, &sci->sc_gc_inodes);
-		nilfs_segctor_notify(sci, &req);
 
 		if (likely(!err))
 			break;
@@ -2560,6 +2558,16 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv,
 		set_current_state(TASK_INTERRUPTIBLE);
 		schedule_timeout(sci->sc_interval);
 	}
+	if (nilfs_test_opt(sbi, DISCARD)) {
+		int ret = nilfs_discard_segments(nilfs, sci->sc_freesegs,
+						 sci->sc_nfreesegs);
+		if (ret) {
+			printk(KERN_WARNING
+			       "NILFS warning: error %d on discard request, "
+			       "turning discards off for the device\n", ret);
+			nilfs_clear_opt(sbi, DISCARD);
+		}
+	}
 
  out_unlock:
 	sci->sc_freesegs = NULL;
@@ -2573,13 +2581,9 @@ static void nilfs_segctor_thread_construct(struct nilfs_sc_info *sci, int mode)
 {
 	struct nilfs_sb_info *sbi = sci->sc_sbi;
 	struct nilfs_transaction_info ti;
-	struct nilfs_segctor_req req = { .mode = mode };
 
 	nilfs_transaction_lock(sbi, &ti, 0);
-
-	nilfs_segctor_accept(sci, &req);
-	nilfs_segctor_construct(sci, &req);
-	nilfs_segctor_notify(sci, &req);
+	nilfs_segctor_construct(sci, mode);
 
 	/*
 	 * Unclosed segment should be retried.  We do this using sc_timer.
@@ -2635,6 +2639,7 @@ static int nilfs_segctor_flush_mode(struct nilfs_sc_info *sci)
 static int nilfs_segctor_thread(void *arg)
 {
 	struct nilfs_sc_info *sci = (struct nilfs_sc_info *)arg;
+	struct the_nilfs *nilfs = sci->sc_sbi->s_nilfs;
 	struct timer_list timer;
 	int timeout = 0;
 
@@ -2680,7 +2685,6 @@ static int nilfs_segctor_thread(void *arg)
 	} else {
 		DEFINE_WAIT(wait);
 		int should_sleep = 1;
-		struct the_nilfs *nilfs;
 
 		prepare_to_wait(&sci->sc_wait_daemon, &wait,
 				TASK_INTERRUPTIBLE);
@@ -2701,8 +2705,8 @@ static int nilfs_segctor_thread(void *arg)
 		finish_wait(&sci->sc_wait_daemon, &wait);
 		timeout = ((sci->sc_state & NILFS_SEGCTOR_COMMIT) &&
 			   time_after_eq(jiffies, sci->sc_timer->expires));
-		nilfs = sci->sc_sbi->s_nilfs;
-		if (sci->sc_super->s_dirt && nilfs_sb_need_update(nilfs))
+
+		if (nilfs_sb_dirty(nilfs) && nilfs_sb_need_update(nilfs))
 			set_nilfs_discontinued(nilfs);
 	}
 	goto loop;
@@ -2797,12 +2801,9 @@ static void nilfs_segctor_write_out(struct nilfs_sc_info *sci)
 	do {
 		struct nilfs_sb_info *sbi = sci->sc_sbi;
 		struct nilfs_transaction_info ti;
-		struct nilfs_segctor_req req = { .mode = SC_LSEG_SR };
 
 		nilfs_transaction_lock(sbi, &ti, 0);
-		nilfs_segctor_accept(sci, &req);
-		ret = nilfs_segctor_construct(sci, &req);
-		nilfs_segctor_notify(sci, &req);
+		ret = nilfs_segctor_construct(sci, SC_LSEG_SR);
 		nilfs_transaction_unlock(sbi);
 
 	} while (ret && retrycount-- > 0);
@@ -2865,8 +2866,15 @@ int nilfs_attach_segment_constructor(struct nilfs_sb_info *sbi)
 	struct the_nilfs *nilfs = sbi->s_nilfs;
 	int err;
 
-	/* Each field of nilfs_segctor is cleared through the initialization
-	   of super-block info */
+	if (NILFS_SC(sbi)) {
+		/*
+		 * This happens if the filesystem was remounted
+		 * read/write after nilfs_error degenerated it into a
+		 * read-only mount.
+		 */
+		nilfs_detach_segment_constructor(sbi);
+	}
+
 	sbi->s_sc_info = nilfs_segctor_new(sbi);
 	if (!sbi->s_sc_info)
 		return -ENOMEM;
diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h
index 3d3ab2f9864c..3155e0c7f415 100644
--- a/fs/nilfs2/segment.h
+++ b/fs/nilfs2/segment.h
@@ -116,6 +116,7 @@ struct nilfs_segsum_pointer {
  * @sc_wait_daemon: Daemon wait queue
  * @sc_wait_task: Start/end wait queue to control segctord task
  * @sc_seq_request: Request counter
+ * @sc_seq_accept: Accepted request count
  * @sc_seq_done: Completion counter
  * @sc_sync: Request of explicit sync operation
  * @sc_interval: Timeout value of background construction
@@ -169,6 +170,7 @@ struct nilfs_sc_info {
 	wait_queue_head_t	sc_wait_task;
 
 	__u32			sc_seq_request;
+	__u32			sc_seq_accepted;
 	__u32			sc_seq_done;
 
 	int			sc_sync;
diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c
index 8173faee31e6..92579cc4c935 100644
--- a/fs/nilfs2/super.c
+++ b/fs/nilfs2/super.c
@@ -96,9 +96,6 @@ void nilfs_error(struct super_block *sb, const char *function,
 	if (!(sb->s_flags & MS_RDONLY)) {
 		struct the_nilfs *nilfs = sbi->s_nilfs;
 
-		if (!nilfs_test_opt(sbi, ERRORS_CONT))
-			nilfs_detach_segment_constructor(sbi);
-
 		down_write(&nilfs->ns_sem);
 		if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) {
 			nilfs->ns_mount_state |= NILFS_ERROR_FS;
@@ -301,7 +298,7 @@ int nilfs_commit_super(struct nilfs_sb_info *sbi, int dupsb)
 		memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);
 		nilfs->ns_sbwtime[1] = t;
 	}
-	sbi->s_super->s_dirt = 0;
+	clear_nilfs_sb_dirty(nilfs);
 	return nilfs_sync_super(sbi, dupsb);
 }
 
@@ -345,7 +342,7 @@ static int nilfs_sync_fs(struct super_block *sb, int wait)
 		err = nilfs_construct_segment(sb);
 
 	down_write(&nilfs->ns_sem);
-	if (sb->s_dirt)
+	if (nilfs_sb_dirty(nilfs))
 		nilfs_commit_super(sbi, 1);
 	up_write(&nilfs->ns_sem);
 
@@ -481,6 +478,8 @@ static int nilfs_show_options(struct seq_file *seq, struct vfsmount *vfs)
 		seq_printf(seq, ",order=strict");
 	if (nilfs_test_opt(sbi, NORECOVERY))
 		seq_printf(seq, ",norecovery");
+	if (nilfs_test_opt(sbi, DISCARD))
+		seq_printf(seq, ",discard");
 
 	return 0;
 }
@@ -550,7 +549,7 @@ static const struct export_operations nilfs_export_ops = {
 enum {
 	Opt_err_cont, Opt_err_panic, Opt_err_ro,
 	Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery,
-	Opt_err,
+	Opt_discard, Opt_err,
 };
 
 static match_table_t tokens = {
@@ -561,6 +560,7 @@ static match_table_t tokens = {
 	{Opt_snapshot, "cp=%u"},
 	{Opt_order, "order=%s"},
 	{Opt_norecovery, "norecovery"},
+	{Opt_discard, "discard"},
 	{Opt_err, NULL}
 };
 
@@ -614,6 +614,9 @@ static int parse_options(char *options, struct super_block *sb)
 		case Opt_norecovery:
 			nilfs_set_opt(sbi, NORECOVERY);
 			break;
+		case Opt_discard:
+			nilfs_set_opt(sbi, DISCARD);
+			break;
 		default:
 			printk(KERN_ERR
 			       "NILFS: Unrecognized mount option \"%s\"\n", p);
diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c
index 6241e1722efc..92733d5651d2 100644
--- a/fs/nilfs2/the_nilfs.c
+++ b/fs/nilfs2/the_nilfs.c
@@ -646,6 +646,44 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data)
 	goto out;
 }
 
+int nilfs_discard_segments(struct the_nilfs *nilfs, __u64 *segnump,
+			    size_t nsegs)
+{
+	sector_t seg_start, seg_end;
+	sector_t start = 0, nblocks = 0;
+	unsigned int sects_per_block;
+	__u64 *sn;
+	int ret = 0;
+
+	sects_per_block = (1 << nilfs->ns_blocksize_bits) /
+		bdev_logical_block_size(nilfs->ns_bdev);
+	for (sn = segnump; sn < segnump + nsegs; sn++) {
+		nilfs_get_segment_range(nilfs, *sn, &seg_start, &seg_end);
+
+		if (!nblocks) {
+			start = seg_start;
+			nblocks = seg_end - seg_start + 1;
+		} else if (start + nblocks == seg_start) {
+			nblocks += seg_end - seg_start + 1;
+		} else {
+			ret = blkdev_issue_discard(nilfs->ns_bdev,
+						   start * sects_per_block,
+						   nblocks * sects_per_block,
+						   GFP_NOFS,
+						   DISCARD_FL_BARRIER);
+			if (ret < 0)
+				return ret;
+			nblocks = 0;
+		}
+	}
+	if (nblocks)
+		ret = blkdev_issue_discard(nilfs->ns_bdev,
+					   start * sects_per_block,
+					   nblocks * sects_per_block,
+					   GFP_NOFS, DISCARD_FL_BARRIER);
+	return ret;
+}
+
 int nilfs_count_free_blocks(struct the_nilfs *nilfs, sector_t *nblocks)
 {
 	struct inode *dat = nilfs_dat_inode(nilfs);
diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h
index 589786e33464..e9795f1724d7 100644
--- a/fs/nilfs2/the_nilfs.h
+++ b/fs/nilfs2/the_nilfs.h
@@ -38,6 +38,7 @@ enum {
 				   the latest checkpoint was loaded */
 	THE_NILFS_DISCONTINUED,	/* 'next' pointer chain has broken */
 	THE_NILFS_GC_RUNNING,	/* gc process is running */
+	THE_NILFS_SB_DIRTY,	/* super block is dirty */
 };
 
 /**
@@ -197,6 +198,7 @@ THE_NILFS_FNS(INIT, init)
 THE_NILFS_FNS(LOADED, loaded)
 THE_NILFS_FNS(DISCONTINUED, discontinued)
 THE_NILFS_FNS(GC_RUNNING, gc_running)
+THE_NILFS_FNS(SB_DIRTY, sb_dirty)
 
 /* Minimum interval of periodical update of superblocks (in seconds) */
 #define NILFS_SB_FREQ		10
@@ -221,6 +223,7 @@ struct the_nilfs *find_or_create_nilfs(struct block_device *);
 void put_nilfs(struct the_nilfs *);
 int init_nilfs(struct the_nilfs *, struct nilfs_sb_info *, char *);
 int load_nilfs(struct the_nilfs *, struct nilfs_sb_info *);
+int nilfs_discard_segments(struct the_nilfs *, __u64 *, size_t);
 int nilfs_count_free_blocks(struct the_nilfs *, sector_t *);
 struct nilfs_sb_info *nilfs_find_sbinfo(struct the_nilfs *, int, __u64);
 int nilfs_checkpoint_is_mounted(struct the_nilfs *, __u64, int);
diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c
index a94e8bd8eb1f..472cdf29ef82 100644
--- a/fs/notify/inotify/inotify_user.c
+++ b/fs/notify/inotify/inotify_user.c
@@ -29,14 +29,12 @@
 #include <linux/init.h> /* module_init */
 #include <linux/inotify.h>
 #include <linux/kernel.h> /* roundup() */
-#include <linux/magic.h> /* superblock magic number */
-#include <linux/mount.h> /* mntget */
 #include <linux/namei.h> /* LOOKUP_FOLLOW */
-#include <linux/path.h> /* struct path */
 #include <linux/sched.h> /* struct user */
 #include <linux/slab.h> /* struct kmem_cache */
 #include <linux/syscalls.h>
 #include <linux/types.h>
+#include <linux/anon_inodes.h>
 #include <linux/uaccess.h>
 #include <linux/poll.h>
 #include <linux/wait.h>
@@ -45,8 +43,6 @@
 
 #include <asm/ioctls.h>
 
-static struct vfsmount *inotify_mnt __read_mostly;
-
 /* these are configurable via /proc/sys/fs/inotify/ */
 static int inotify_max_user_instances __read_mostly;
 static int inotify_max_queued_events __read_mostly;
@@ -645,9 +641,7 @@ SYSCALL_DEFINE1(inotify_init1, int, flags)
 {
 	struct fsnotify_group *group;
 	struct user_struct *user;
-	struct file *filp;
-	struct path path;
-	int fd, ret;
+	int ret;
 
 	/* Check the IN_* constants for consistency.  */
 	BUILD_BUG_ON(IN_CLOEXEC != O_CLOEXEC);
@@ -656,10 +650,6 @@ SYSCALL_DEFINE1(inotify_init1, int, flags)
 	if (flags & ~(IN_CLOEXEC | IN_NONBLOCK))
 		return -EINVAL;
 
-	fd = get_unused_fd_flags(flags & O_CLOEXEC);
-	if (fd < 0)
-		return fd;
-
 	user = get_current_user();
 	if (unlikely(atomic_read(&user->inotify_devs) >=
 			inotify_max_user_instances)) {
@@ -676,27 +666,14 @@ SYSCALL_DEFINE1(inotify_init1, int, flags)
 
 	atomic_inc(&user->inotify_devs);
 
-	path.mnt = inotify_mnt;
-	path.dentry = inotify_mnt->mnt_root;
-	path_get(&path);
-	filp = alloc_file(&path, FMODE_READ, &inotify_fops);
-	if (!filp)
-		goto Enfile;
+	ret = anon_inode_getfd("inotify", &inotify_fops, group,
+				  O_RDONLY | flags);
+	if (ret >= 0)
+		return ret;
 
-	filp->f_flags = O_RDONLY | (flags & O_NONBLOCK);
-	filp->private_data = group;
-
-	fd_install(fd, filp);
-
-	return fd;
-
-Enfile:
-	ret = -ENFILE;
-	path_put(&path);
 	atomic_dec(&user->inotify_devs);
 out_free_uid:
 	free_uid(user);
-	put_unused_fd(fd);
 	return ret;
 }
 
@@ -783,20 +760,6 @@ out:
 	return ret;
 }
 
-static int
-inotify_get_sb(struct file_system_type *fs_type, int flags,
-	       const char *dev_name, void *data, struct vfsmount *mnt)
-{
-	return get_sb_pseudo(fs_type, "inotify", NULL,
-			INOTIFYFS_SUPER_MAGIC, mnt);
-}
-
-static struct file_system_type inotify_fs_type = {
-    .name	= "inotifyfs",
-    .get_sb	= inotify_get_sb,
-    .kill_sb	= kill_anon_super,
-};
-
 /*
  * inotify_user_setup - Our initialization function.  Note that we cannnot return
  * error because we have compiled-in VFS hooks.  So an (unlikely) failure here
@@ -804,16 +767,6 @@ static struct file_system_type inotify_fs_type = {
  */
 static int __init inotify_user_setup(void)
 {
-	int ret;
-
-	ret = register_filesystem(&inotify_fs_type);
-	if (unlikely(ret))
-		panic("inotify: register_filesystem returned %d!\n", ret);
-
-	inotify_mnt = kern_mount(&inotify_fs_type);
-	if (IS_ERR(inotify_mnt))
-		panic("inotify: kern_mount ret %ld!\n", PTR_ERR(inotify_mnt));
-
 	inotify_inode_mark_cachep = KMEM_CACHE(inotify_inode_mark_entry, SLAB_PANIC);
 	event_priv_cachep = KMEM_CACHE(inotify_event_private_data, SLAB_PANIC);
 
diff --git a/fs/ntfs/dir.c b/fs/ntfs/dir.c
index 5a9e34475e37..9173e82a45d1 100644
--- a/fs/ntfs/dir.c
+++ b/fs/ntfs/dir.c
@@ -1545,7 +1545,7 @@ static int ntfs_dir_fsync(struct file *filp, struct dentry *dentry,
  		write_inode_now(bmp_vi, !datasync);
 		iput(bmp_vi);
 	}
-	ret = ntfs_write_inode(vi, 1);
+	ret = __ntfs_write_inode(vi, 1);
 	write_inode_now(vi, !datasync);
 	err = sync_blockdev(vi->i_sb->s_bdev);
 	if (unlikely(err && !ret))
diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c
index 43179ddd336f..b681c71d7069 100644
--- a/fs/ntfs/file.c
+++ b/fs/ntfs/file.c
@@ -2182,7 +2182,7 @@ static int ntfs_file_fsync(struct file *filp, struct dentry *dentry,
 	ntfs_debug("Entering for inode 0x%lx.", vi->i_ino);
 	BUG_ON(S_ISDIR(vi->i_mode));
 	if (!datasync || !NInoNonResident(NTFS_I(vi)))
-		ret = ntfs_write_inode(vi, 1);
+		ret = __ntfs_write_inode(vi, 1);
 	write_inode_now(vi, !datasync);
 	/*
 	 * NOTE: If we were to use mapping->private_list (see ext2 and
diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c
index dc2505abb6d7..4b57fb1eac2a 100644
--- a/fs/ntfs/inode.c
+++ b/fs/ntfs/inode.c
@@ -2957,7 +2957,7 @@ out:
  *
  * Return 0 on success and -errno on error.
  */
-int ntfs_write_inode(struct inode *vi, int sync)
+int __ntfs_write_inode(struct inode *vi, int sync)
 {
 	sle64 nt;
 	ntfs_inode *ni = NTFS_I(vi);
diff --git a/fs/ntfs/inode.h b/fs/ntfs/inode.h
index 117eaf8032a3..9a113544605d 100644
--- a/fs/ntfs/inode.h
+++ b/fs/ntfs/inode.h
@@ -307,12 +307,12 @@ extern void ntfs_truncate_vfs(struct inode *vi);
 
 extern int ntfs_setattr(struct dentry *dentry, struct iattr *attr);
 
-extern int ntfs_write_inode(struct inode *vi, int sync);
+extern int __ntfs_write_inode(struct inode *vi, int sync);
 
 static inline void ntfs_commit_inode(struct inode *vi)
 {
 	if (!is_bad_inode(vi))
-		ntfs_write_inode(vi, 1);
+		__ntfs_write_inode(vi, 1);
 	return;
 }
 
diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c
index 80b04770e8e9..1cf39dfaee7a 100644
--- a/fs/ntfs/super.c
+++ b/fs/ntfs/super.c
@@ -39,6 +39,7 @@
 #include "dir.h"
 #include "debug.h"
 #include "index.h"
+#include "inode.h"
 #include "aops.h"
 #include "layout.h"
 #include "malloc.h"
@@ -2662,6 +2663,13 @@ static int ntfs_statfs(struct dentry *dentry, struct kstatfs *sfs)
 	return 0;
 }
 
+#ifdef NTFS_RW
+static int ntfs_write_inode(struct inode *vi, struct writeback_control *wbc)
+{
+	return __ntfs_write_inode(vi, wbc->sync_mode == WB_SYNC_ALL);
+}
+#endif
+
 /**
  * The complete super operations.
  */
diff --git a/fs/ocfs2/Makefile b/fs/ocfs2/Makefile
index 600d2d2ade11..791c0886c060 100644
--- a/fs/ocfs2/Makefile
+++ b/fs/ocfs2/Makefile
@@ -46,6 +46,7 @@ ocfs2_stackglue-objs := stackglue.o
 ocfs2_stack_o2cb-objs := stack_o2cb.o
 ocfs2_stack_user-objs := stack_user.o
 
+obj-$(CONFIG_OCFS2_FS) += dlmfs/
 # cluster/ is always needed when OCFS2_FS for masklog support
 obj-$(CONFIG_OCFS2_FS) += cluster/
 obj-$(CONFIG_OCFS2_FS_O2CB) += dlm/
diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index d17bdc718f74..9f8bd913c51e 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -1050,7 +1050,8 @@ static int ocfs2_create_new_meta_bhs(handle_t *handle,
 			strcpy(eb->h_signature, OCFS2_EXTENT_BLOCK_SIGNATURE);
 			eb->h_blkno = cpu_to_le64(first_blkno);
 			eb->h_fs_generation = cpu_to_le32(osb->fs_generation);
-			eb->h_suballoc_slot = cpu_to_le16(osb->slot_num);
+			eb->h_suballoc_slot =
+				cpu_to_le16(meta_ac->ac_alloc_slot);
 			eb->h_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 			eb->h_list.l_count =
 				cpu_to_le16(ocfs2_extent_recs_per_eb(osb->sb));
@@ -5712,7 +5713,7 @@ int ocfs2_remove_btree_range(struct inode *inode,
 		goto out;
 	}
 
-	vfs_dq_free_space_nodirty(inode,
+	dquot_free_space_nodirty(inode,
 				  ocfs2_clusters_to_bytes(inode->i_sb, len));
 
 	ret = ocfs2_remove_extent(handle, et, cpos, len, meta_ac, dealloc);
@@ -6037,7 +6038,7 @@ static void ocfs2_truncate_log_worker(struct work_struct *work)
 	if (status < 0)
 		mlog_errno(status);
 	else
-		ocfs2_init_inode_steal_slot(osb);
+		ocfs2_init_steal_slots(osb);
 
 	mlog_exit(status);
 }
@@ -6935,7 +6936,7 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb,
 		goto bail;
 	}
 
-	vfs_dq_free_space_nodirty(inode,
+	dquot_free_space_nodirty(inode,
 			ocfs2_clusters_to_bytes(osb->sb, clusters_to_del));
 	spin_lock(&OCFS2_I(inode)->ip_lock);
 	OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters) -
@@ -7300,11 +7301,10 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 		unsigned int page_end;
 		u64 phys;
 
-		if (vfs_dq_alloc_space_nodirty(inode,
-				       ocfs2_clusters_to_bytes(osb->sb, 1))) {
-			ret = -EDQUOT;
+		ret = dquot_alloc_space_nodirty(inode,
+				       ocfs2_clusters_to_bytes(osb->sb, 1));
+		if (ret)
 			goto out_commit;
-		}
 		did_quota = 1;
 
 		ret = ocfs2_claim_clusters(osb, handle, data_ac, 1, &bit_off,
@@ -7380,7 +7380,7 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 
 out_commit:
 	if (ret < 0 && did_quota)
-		vfs_dq_free_space_nodirty(inode,
+		dquot_free_space_nodirty(inode,
 					  ocfs2_clusters_to_bytes(osb->sb, 1));
 
 	ocfs2_commit_trans(osb, handle);
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 7e9df11260f4..21441ddb5506 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -577,8 +577,9 @@ static int ocfs2_direct_IO_get_blocks(struct inode *inode, sector_t iblock,
 		goto bail;
 	}
 
-	/* We should already CoW the refcounted extent. */
-	BUG_ON(ext_flags & OCFS2_EXT_REFCOUNTED);
+	/* We should already CoW the refcounted extent in case of create. */
+	BUG_ON(create && (ext_flags & OCFS2_EXT_REFCOUNTED));
+
 	/*
 	 * get_more_blocks() expects us to describe a hole by clearing
 	 * the mapped bit on bh_result().
@@ -1763,10 +1764,11 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 
 	wc->w_handle = handle;
 
-	if (clusters_to_alloc && vfs_dq_alloc_space_nodirty(inode,
-			ocfs2_clusters_to_bytes(osb->sb, clusters_to_alloc))) {
-		ret = -EDQUOT;
-		goto out_commit;
+	if (clusters_to_alloc) {
+		ret = dquot_alloc_space_nodirty(inode,
+			ocfs2_clusters_to_bytes(osb->sb, clusters_to_alloc));
+		if (ret)
+			goto out_commit;
 	}
 	/*
 	 * We don't want this to fail in ocfs2_write_end(), so do it
@@ -1809,7 +1811,7 @@ success:
 	return 0;
 out_quota:
 	if (clusters_to_alloc)
-		vfs_dq_free_space(inode,
+		dquot_free_space(inode,
 			  ocfs2_clusters_to_bytes(osb->sb, clusters_to_alloc));
 out_commit:
 	ocfs2_commit_trans(osb, handle);
diff --git a/fs/ocfs2/cluster/masklog.c b/fs/ocfs2/cluster/masklog.c
index 1cd2934de615..b39da877b12f 100644
--- a/fs/ocfs2/cluster/masklog.c
+++ b/fs/ocfs2/cluster/masklog.c
@@ -112,6 +112,7 @@ static struct mlog_attribute mlog_attrs[MLOG_MAX_BITS] = {
 	define_mask(XATTR),
 	define_mask(QUOTA),
 	define_mask(REFCOUNT),
+	define_mask(BASTS),
 	define_mask(ERROR),
 	define_mask(NOTICE),
 	define_mask(KTHREAD),
diff --git a/fs/ocfs2/cluster/masklog.h b/fs/ocfs2/cluster/masklog.h
index 9b4d11726cf2..3dfddbec32f2 100644
--- a/fs/ocfs2/cluster/masklog.h
+++ b/fs/ocfs2/cluster/masklog.h
@@ -114,6 +114,7 @@
 #define ML_XATTR	0x0000000020000000ULL /* ocfs2 extended attributes */
 #define ML_QUOTA	0x0000000040000000ULL /* ocfs2 quota operations */
 #define ML_REFCOUNT	0x0000000080000000ULL /* refcount tree operations */
+#define ML_BASTS	0x0000001000000000ULL /* dlmglue asts and basts */
 /* bits that are infrequently given and frequently matched in the high word */
 #define ML_ERROR	0x0000000100000000ULL /* sent to KERN_ERR */
 #define ML_NOTICE	0x0000000200000000ULL /* setn to KERN_NOTICE */
@@ -194,9 +195,9 @@ extern struct mlog_bits mlog_and_bits, mlog_not_bits;
  * previous token if args expands to nothing.
  */
 #define __mlog_printk(level, fmt, args...)				\
-	printk(level "(%u,%lu):%s:%d " fmt, task_pid_nr(current),	\
-	       __mlog_cpu_guess, __PRETTY_FUNCTION__, __LINE__ ,	\
-	       ##args)
+	printk(level "(%s,%u,%lu):%s:%d " fmt, current->comm,		\
+	       task_pid_nr(current), __mlog_cpu_guess,			\
+	       __PRETTY_FUNCTION__, __LINE__ , ##args)
 
 #define mlog(mask, fmt, args...) do {					\
 	u64 __m = MLOG_MASK_PREFIX | (mask);				\
diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c
index 28c3ec238796..efd77d071c80 100644
--- a/fs/ocfs2/dir.c
+++ b/fs/ocfs2/dir.c
@@ -2439,7 +2439,7 @@ static int ocfs2_dx_dir_attach_index(struct ocfs2_super *osb,
 	dx_root = (struct ocfs2_dx_root_block *)dx_root_bh->b_data;
 	memset(dx_root, 0, osb->sb->s_blocksize);
 	strcpy(dx_root->dr_signature, OCFS2_DX_ROOT_SIGNATURE);
-	dx_root->dr_suballoc_slot = cpu_to_le16(osb->slot_num);
+	dx_root->dr_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
 	dx_root->dr_suballoc_bit = cpu_to_le16(dr_suballoc_bit);
 	dx_root->dr_fs_generation = cpu_to_le32(osb->fs_generation);
 	dx_root->dr_blkno = cpu_to_le64(dr_blkno);
@@ -2964,12 +2964,10 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 		goto out;
 	}
 
-	if (vfs_dq_alloc_space_nodirty(dir,
-				ocfs2_clusters_to_bytes(osb->sb,
-							alloc + dx_alloc))) {
-		ret = -EDQUOT;
+	ret = dquot_alloc_space_nodirty(dir,
+		ocfs2_clusters_to_bytes(osb->sb, alloc + dx_alloc));
+	if (ret)
 		goto out_commit;
-	}
 	did_quota = 1;
 
 	if (ocfs2_supports_indexed_dirs(osb) && !dx_inline) {
@@ -3178,7 +3176,7 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 
 out_commit:
 	if (ret < 0 && did_quota)
-		vfs_dq_free_space_nodirty(dir, bytes_allocated);
+		dquot_free_space_nodirty(dir, bytes_allocated);
 
 	ocfs2_commit_trans(osb, handle);
 
@@ -3221,11 +3219,10 @@ static int ocfs2_do_extend_dir(struct super_block *sb,
 	if (extend) {
 		u32 offset = OCFS2_I(dir)->ip_clusters;
 
-		if (vfs_dq_alloc_space_nodirty(dir,
-					ocfs2_clusters_to_bytes(sb, 1))) {
-			status = -EDQUOT;
+		status = dquot_alloc_space_nodirty(dir,
+					ocfs2_clusters_to_bytes(sb, 1));
+		if (status)
 			goto bail;
-		}
 		did_quota = 1;
 
 		status = ocfs2_add_inode_data(OCFS2_SB(sb), dir, &offset,
@@ -3254,7 +3251,7 @@ static int ocfs2_do_extend_dir(struct super_block *sb,
 	status = 0;
 bail:
 	if (did_quota && status < 0)
-		vfs_dq_free_space_nodirty(dir, ocfs2_clusters_to_bytes(sb, 1));
+		dquot_free_space_nodirty(dir, ocfs2_clusters_to_bytes(sb, 1));
 	mlog_exit(status);
 	return status;
 }
@@ -3889,11 +3886,10 @@ static int ocfs2_dx_dir_rebalance(struct ocfs2_super *osb, struct inode *dir,
 		goto out;
 	}
 
-	if (vfs_dq_alloc_space_nodirty(dir,
-				       ocfs2_clusters_to_bytes(dir->i_sb, 1))) {
-		ret = -EDQUOT;
+	ret = dquot_alloc_space_nodirty(dir,
+				       ocfs2_clusters_to_bytes(dir->i_sb, 1));
+	if (ret)
 		goto out_commit;
-	}
 	did_quota = 1;
 
 	ret = ocfs2_journal_access_dl(handle, INODE_CACHE(dir), dx_leaf_bh,
@@ -3983,7 +3979,7 @@ static int ocfs2_dx_dir_rebalance(struct ocfs2_super *osb, struct inode *dir,
 
 out_commit:
 	if (ret < 0 && did_quota)
-		vfs_dq_free_space_nodirty(dir,
+		dquot_free_space_nodirty(dir,
 				ocfs2_clusters_to_bytes(dir->i_sb, 1));
 
 	ocfs2_commit_trans(osb, handle);
@@ -4165,11 +4161,10 @@ static int ocfs2_expand_inline_dx_root(struct inode *dir,
 		goto out;
 	}
 
-	if (vfs_dq_alloc_space_nodirty(dir,
-				       ocfs2_clusters_to_bytes(osb->sb, 1))) {
-		ret = -EDQUOT;
+	ret = dquot_alloc_space_nodirty(dir,
+				       ocfs2_clusters_to_bytes(osb->sb, 1));
+	if (ret)
 		goto out_commit;
-	}
 	did_quota = 1;
 
 	/*
@@ -4229,7 +4224,7 @@ static int ocfs2_expand_inline_dx_root(struct inode *dir,
 
 out_commit:
 	if (ret < 0 && did_quota)
-		vfs_dq_free_space_nodirty(dir,
+		dquot_free_space_nodirty(dir,
 					  ocfs2_clusters_to_bytes(dir->i_sb, 1));
 
 	ocfs2_commit_trans(osb, handle);
diff --git a/fs/ocfs2/dlm/Makefile b/fs/ocfs2/dlm/Makefile
index 190361375700..dcebf0d920fa 100644
--- a/fs/ocfs2/dlm/Makefile
+++ b/fs/ocfs2/dlm/Makefile
@@ -1,8 +1,7 @@
 EXTRA_CFLAGS += -Ifs/ocfs2
 
-obj-$(CONFIG_OCFS2_FS_O2CB) += ocfs2_dlm.o ocfs2_dlmfs.o
+obj-$(CONFIG_OCFS2_FS_O2CB) += ocfs2_dlm.o
 
 ocfs2_dlm-objs := dlmdomain.o dlmdebug.o dlmthread.o dlmrecovery.o \
 	dlmmaster.o dlmast.o dlmconvert.o dlmlock.o dlmunlock.o dlmver.o
 
-ocfs2_dlmfs-objs := userdlm.o dlmfs.o dlmfsver.o
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 344bcf90cbf4..b4f99de2caf3 100644
--- a/fs/ocfs2/dlm/dlmrecovery.c
+++ b/fs/ocfs2/dlm/dlmrecovery.c
@@ -310,7 +310,7 @@ static int dlm_recovery_thread(void *data)
 	mlog(0, "dlm thread running for %s...\n", dlm->name);
 
 	while (!kthread_should_stop()) {
-		if (dlm_joined(dlm)) {
+		if (dlm_domain_fully_joined(dlm)) {
 			status = dlm_do_recovery(dlm);
 			if (status == -EAGAIN) {
 				/* do not sleep, recheck immediately. */
diff --git a/fs/ocfs2/dlmfs/Makefile b/fs/ocfs2/dlmfs/Makefile
new file mode 100644
index 000000000000..df69b4856d0d
--- /dev/null
+++ b/fs/ocfs2/dlmfs/Makefile
@@ -0,0 +1,5 @@
+EXTRA_CFLAGS += -Ifs/ocfs2
+
+obj-$(CONFIG_OCFS2_FS) += ocfs2_dlmfs.o
+
+ocfs2_dlmfs-objs := userdlm.o dlmfs.o dlmfsver.o
diff --git a/fs/ocfs2/dlm/dlmfs.c b/fs/ocfs2/dlmfs/dlmfs.c
index 02bf17808bdc..1b0de157a08c 100644
--- a/fs/ocfs2/dlm/dlmfs.c
+++ b/fs/ocfs2/dlmfs/dlmfs.c
@@ -43,24 +43,17 @@
 #include <linux/init.h>
 #include <linux/string.h>
 #include <linux/backing-dev.h>
+#include <linux/poll.h>
 
 #include <asm/uaccess.h>
 
-
-#include "cluster/nodemanager.h"
-#include "cluster/heartbeat.h"
-#include "cluster/tcp.h"
-
-#include "dlmapi.h"
-
+#include "stackglue.h"
 #include "userdlm.h"
-
 #include "dlmfsver.h"
 
 #define MLOG_MASK_PREFIX ML_DLMFS
 #include "cluster/masklog.h"
 
-#include "ocfs2_lockingver.h"
 
 static const struct super_operations dlmfs_ops;
 static const struct file_operations dlmfs_file_operations;
@@ -71,15 +64,46 @@ static struct kmem_cache *dlmfs_inode_cache;
 
 struct workqueue_struct *user_dlm_worker;
 
+
+
 /*
- * This is the userdlmfs locking protocol version.
+ * These are the ABI capabilities of dlmfs.
+ *
+ * Over time, dlmfs has added some features that were not part of the
+ * initial ABI.  Unfortunately, some of these features are not detectable
+ * via standard usage.  For example, Linux's default poll always returns
+ * POLLIN, so there is no way for a caller of poll(2) to know when dlmfs
+ * added poll support.  Instead, we provide this list of new capabilities.
+ *
+ * Capabilities is a read-only attribute.  We do it as a module parameter
+ * so we can discover it whether dlmfs is built in, loaded, or even not
+ * loaded.
  *
- * See fs/ocfs2/dlmglue.c for more details on locking versions.
+ * The ABI features are local to this machine's dlmfs mount.  This is
+ * distinct from the locking protocol, which is concerned with inter-node
+ * interaction.
+ *
+ * Capabilities:
+ * - bast	: POLLIN against the file descriptor of a held lock
+ *		  signifies a bast fired on the lock.
  */
-static const struct dlm_protocol_version user_locking_protocol = {
-	.pv_major = OCFS2_LOCKING_PROTOCOL_MAJOR,
-	.pv_minor = OCFS2_LOCKING_PROTOCOL_MINOR,
-};
+#define DLMFS_CAPABILITIES "bast stackglue"
+extern int param_set_dlmfs_capabilities(const char *val,
+					struct kernel_param *kp)
+{
+	printk(KERN_ERR "%s: readonly parameter\n", kp->name);
+	return -EINVAL;
+}
+static int param_get_dlmfs_capabilities(char *buffer,
+					struct kernel_param *kp)
+{
+	return strlcpy(buffer, DLMFS_CAPABILITIES,
+		       strlen(DLMFS_CAPABILITIES) + 1);
+}
+module_param_call(capabilities, param_set_dlmfs_capabilities,
+		  param_get_dlmfs_capabilities, NULL, 0444);
+MODULE_PARM_DESC(capabilities, DLMFS_CAPABILITIES);
+
 
 /*
  * decodes a set of open flags into a valid lock level and a set of flags.
@@ -179,13 +203,46 @@ static int dlmfs_file_release(struct inode *inode,
 	return 0;
 }
 
+/*
+ * We do ->setattr() just to override size changes.  Our size is the size
+ * of the LVB and nothing else.
+ */
+static int dlmfs_file_setattr(struct dentry *dentry, struct iattr *attr)
+{
+	int error;
+	struct inode *inode = dentry->d_inode;
+
+	attr->ia_valid &= ~ATTR_SIZE;
+	error = inode_change_ok(inode, attr);
+	if (!error)
+		error = inode_setattr(inode, attr);
+
+	return error;
+}
+
+static unsigned int dlmfs_file_poll(struct file *file, poll_table *wait)
+{
+	int event = 0;
+	struct inode *inode = file->f_path.dentry->d_inode;
+	struct dlmfs_inode_private *ip = DLMFS_I(inode);
+
+	poll_wait(file, &ip->ip_lockres.l_event, wait);
+
+	spin_lock(&ip->ip_lockres.l_lock);
+	if (ip->ip_lockres.l_flags & USER_LOCK_BLOCKED)
+		event = POLLIN | POLLRDNORM;
+	spin_unlock(&ip->ip_lockres.l_lock);
+
+	return event;
+}
+
 static ssize_t dlmfs_file_read(struct file *filp,
 			       char __user *buf,
 			       size_t count,
 			       loff_t *ppos)
 {
 	int bytes_left;
-	ssize_t readlen;
+	ssize_t readlen, got;
 	char *lvb_buf;
 	struct inode *inode = filp->f_path.dentry->d_inode;
 
@@ -211,9 +268,13 @@ static ssize_t dlmfs_file_read(struct file *filp,
 	if (!lvb_buf)
 		return -ENOMEM;
 
-	user_dlm_read_lvb(inode, lvb_buf, readlen);
-	bytes_left = __copy_to_user(buf, lvb_buf, readlen);
-	readlen -= bytes_left;
+	got = user_dlm_read_lvb(inode, lvb_buf, readlen);
+	if (got) {
+		BUG_ON(got != readlen);
+		bytes_left = __copy_to_user(buf, lvb_buf, readlen);
+		readlen -= bytes_left;
+	} else
+		readlen = 0;
 
 	kfree(lvb_buf);
 
@@ -272,7 +333,7 @@ static void dlmfs_init_once(void *foo)
 	struct dlmfs_inode_private *ip =
 		(struct dlmfs_inode_private *) foo;
 
-	ip->ip_dlm = NULL;
+	ip->ip_conn = NULL;
 	ip->ip_parent = NULL;
 
 	inode_init_once(&ip->ip_vfs_inode);
@@ -314,14 +375,14 @@ static void dlmfs_clear_inode(struct inode *inode)
 		goto clear_fields;
 	}
 
-	mlog(0, "we're a directory, ip->ip_dlm = 0x%p\n", ip->ip_dlm);
+	mlog(0, "we're a directory, ip->ip_conn = 0x%p\n", ip->ip_conn);
 	/* we must be a directory. If required, lets unregister the
 	 * dlm context now. */
-	if (ip->ip_dlm)
-		user_dlm_unregister_context(ip->ip_dlm);
+	if (ip->ip_conn)
+		user_dlm_unregister(ip->ip_conn);
 clear_fields:
 	ip->ip_parent = NULL;
-	ip->ip_dlm = NULL;
+	ip->ip_conn = NULL;
 }
 
 static struct backing_dev_info dlmfs_backing_dev_info = {
@@ -371,7 +432,7 @@ static struct inode *dlmfs_get_inode(struct inode *parent,
 	inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
 
 	ip = DLMFS_I(inode);
-	ip->ip_dlm = DLMFS_I(parent)->ip_dlm;
+	ip->ip_conn = DLMFS_I(parent)->ip_conn;
 
 	switch (mode & S_IFMT) {
 	default:
@@ -425,13 +486,12 @@ static int dlmfs_mkdir(struct inode * dir,
 	struct inode *inode = NULL;
 	struct qstr *domain = &dentry->d_name;
 	struct dlmfs_inode_private *ip;
-	struct dlm_ctxt *dlm;
-	struct dlm_protocol_version proto = user_locking_protocol;
+	struct ocfs2_cluster_connection *conn;
 
 	mlog(0, "mkdir %.*s\n", domain->len, domain->name);
 
 	/* verify that we have a proper domain */
-	if (domain->len >= O2NM_MAX_NAME_LEN) {
+	if (domain->len >= GROUP_NAME_MAX) {
 		status = -EINVAL;
 		mlog(ML_ERROR, "invalid domain name for directory.\n");
 		goto bail;
@@ -446,14 +506,14 @@ static int dlmfs_mkdir(struct inode * dir,
 
 	ip = DLMFS_I(inode);
 
-	dlm = user_dlm_register_context(domain, &proto);
-	if (IS_ERR(dlm)) {
-		status = PTR_ERR(dlm);
+	conn = user_dlm_register(domain);
+	if (IS_ERR(conn)) {
+		status = PTR_ERR(conn);
 		mlog(ML_ERROR, "Error %d could not register domain \"%.*s\"\n",
 		     status, domain->len, domain->name);
 		goto bail;
 	}
-	ip->ip_dlm = dlm;
+	ip->ip_conn = conn;
 
 	inc_nlink(dir);
 	d_instantiate(dentry, inode);
@@ -549,6 +609,7 @@ static int dlmfs_fill_super(struct super_block * sb,
 static const struct file_operations dlmfs_file_operations = {
 	.open		= dlmfs_file_open,
 	.release	= dlmfs_file_release,
+	.poll		= dlmfs_file_poll,
 	.read		= dlmfs_file_read,
 	.write		= dlmfs_file_write,
 };
@@ -576,6 +637,7 @@ static const struct super_operations dlmfs_ops = {
 
 static const struct inode_operations dlmfs_file_inode_operations = {
 	.getattr	= simple_getattr,
+	.setattr	= dlmfs_file_setattr,
 };
 
 static int dlmfs_get_sb(struct file_system_type *fs_type,
@@ -620,6 +682,7 @@ static int __init init_dlmfs_fs(void)
 	}
 	cleanup_worker = 1;
 
+	user_dlm_set_locking_protocol();
 	status = register_filesystem(&dlmfs_fs_type);
 bail:
 	if (status) {
diff --git a/fs/ocfs2/dlm/dlmfsver.c b/fs/ocfs2/dlmfs/dlmfsver.c
index a733b3321f83..a733b3321f83 100644
--- a/fs/ocfs2/dlm/dlmfsver.c
+++ b/fs/ocfs2/dlmfs/dlmfsver.c
diff --git a/fs/ocfs2/dlm/dlmfsver.h b/fs/ocfs2/dlmfs/dlmfsver.h
index f35eadbed25c..f35eadbed25c 100644
--- a/fs/ocfs2/dlm/dlmfsver.h
+++ b/fs/ocfs2/dlmfs/dlmfsver.h
diff --git a/fs/ocfs2/dlm/userdlm.c b/fs/ocfs2/dlmfs/userdlm.c
index 4cb1d3dae250..0499e3fb7bdb 100644
--- a/fs/ocfs2/dlm/userdlm.c
+++ b/fs/ocfs2/dlmfs/userdlm.c
@@ -34,18 +34,19 @@
 #include <linux/types.h>
 #include <linux/crc32.h>
 
-
-#include "cluster/nodemanager.h"
-#include "cluster/heartbeat.h"
-#include "cluster/tcp.h"
-
-#include "dlmapi.h"
-
+#include "ocfs2_lockingver.h"
+#include "stackglue.h"
 #include "userdlm.h"
 
 #define MLOG_MASK_PREFIX ML_DLMFS
 #include "cluster/masklog.h"
 
+
+static inline struct user_lock_res *user_lksb_to_lock_res(struct ocfs2_dlm_lksb *lksb)
+{
+	return container_of(lksb, struct user_lock_res, l_lksb);
+}
+
 static inline int user_check_wait_flag(struct user_lock_res *lockres,
 				       int flag)
 {
@@ -73,15 +74,15 @@ static inline void user_wait_on_blocked_lock(struct user_lock_res *lockres)
 }
 
 /* I heart container_of... */
-static inline struct dlm_ctxt *
-dlm_ctxt_from_user_lockres(struct user_lock_res *lockres)
+static inline struct ocfs2_cluster_connection *
+cluster_connection_from_user_lockres(struct user_lock_res *lockres)
 {
 	struct dlmfs_inode_private *ip;
 
 	ip = container_of(lockres,
 			  struct dlmfs_inode_private,
 			  ip_lockres);
-	return ip->ip_dlm;
+	return ip->ip_conn;
 }
 
 static struct inode *
@@ -103,9 +104,9 @@ static inline void user_recover_from_dlm_error(struct user_lock_res *lockres)
 }
 
 #define user_log_dlm_error(_func, _stat, _lockres) do {			\
-	mlog(ML_ERROR, "Dlm error \"%s\" while calling %s on "		\
-		"resource %.*s: %s\n", dlm_errname(_stat), _func,	\
-		_lockres->l_namelen, _lockres->l_name, dlm_errmsg(_stat)); \
+	mlog(ML_ERROR, "Dlm error %d while calling %s on "		\
+		"resource %.*s\n", _stat, _func,			\
+		_lockres->l_namelen, _lockres->l_name); 		\
 } while (0)
 
 /* WARNING: This function lives in a world where the only three lock
@@ -113,34 +114,35 @@ static inline void user_recover_from_dlm_error(struct user_lock_res *lockres)
  * lock types are added. */
 static inline int user_highest_compat_lock_level(int level)
 {
-	int new_level = LKM_EXMODE;
+	int new_level = DLM_LOCK_EX;
 
-	if (level == LKM_EXMODE)
-		new_level = LKM_NLMODE;
-	else if (level == LKM_PRMODE)
-		new_level = LKM_PRMODE;
+	if (level == DLM_LOCK_EX)
+		new_level = DLM_LOCK_NL;
+	else if (level == DLM_LOCK_PR)
+		new_level = DLM_LOCK_PR;
 	return new_level;
 }
 
-static void user_ast(void *opaque)
+static void user_ast(struct ocfs2_dlm_lksb *lksb)
 {
-	struct user_lock_res *lockres = opaque;
-	struct dlm_lockstatus *lksb;
+	struct user_lock_res *lockres = user_lksb_to_lock_res(lksb);
+	int status;
 
-	mlog(0, "AST fired for lockres %.*s\n", lockres->l_namelen,
-	     lockres->l_name);
+	mlog(ML_BASTS, "AST fired for lockres %.*s, level %d => %d\n",
+	     lockres->l_namelen, lockres->l_name, lockres->l_level,
+	     lockres->l_requested);
 
 	spin_lock(&lockres->l_lock);
 
-	lksb = &(lockres->l_lksb);
-	if (lksb->status != DLM_NORMAL) {
+	status = ocfs2_dlm_lock_status(&lockres->l_lksb);
+	if (status) {
 		mlog(ML_ERROR, "lksb status value of %u on lockres %.*s\n",
-		     lksb->status, lockres->l_namelen, lockres->l_name);
+		     status, lockres->l_namelen, lockres->l_name);
 		spin_unlock(&lockres->l_lock);
 		return;
 	}
 
-	mlog_bug_on_msg(lockres->l_requested == LKM_IVMODE,
+	mlog_bug_on_msg(lockres->l_requested == DLM_LOCK_IV,
 			"Lockres %.*s, requested ivmode. flags 0x%x\n",
 			lockres->l_namelen, lockres->l_name, lockres->l_flags);
 
@@ -148,13 +150,13 @@ static void user_ast(void *opaque)
 	if (lockres->l_requested < lockres->l_level) {
 		if (lockres->l_requested <=
 		    user_highest_compat_lock_level(lockres->l_blocking)) {
-			lockres->l_blocking = LKM_NLMODE;
+			lockres->l_blocking = DLM_LOCK_NL;
 			lockres->l_flags &= ~USER_LOCK_BLOCKED;
 		}
 	}
 
 	lockres->l_level = lockres->l_requested;
-	lockres->l_requested = LKM_IVMODE;
+	lockres->l_requested = DLM_LOCK_IV;
 	lockres->l_flags |= USER_LOCK_ATTACHED;
 	lockres->l_flags &= ~USER_LOCK_BUSY;
 
@@ -193,11 +195,11 @@ static void __user_dlm_cond_queue_lockres(struct user_lock_res *lockres)
 		return;
 
 	switch (lockres->l_blocking) {
-	case LKM_EXMODE:
+	case DLM_LOCK_EX:
 		if (!lockres->l_ex_holders && !lockres->l_ro_holders)
 			queue = 1;
 		break;
-	case LKM_PRMODE:
+	case DLM_LOCK_PR:
 		if (!lockres->l_ex_holders)
 			queue = 1;
 		break;
@@ -209,12 +211,12 @@ static void __user_dlm_cond_queue_lockres(struct user_lock_res *lockres)
 		__user_dlm_queue_lockres(lockres);
 }
 
-static void user_bast(void *opaque, int level)
+static void user_bast(struct ocfs2_dlm_lksb *lksb, int level)
 {
-	struct user_lock_res *lockres = opaque;
+	struct user_lock_res *lockres = user_lksb_to_lock_res(lksb);
 
-	mlog(0, "Blocking AST fired for lockres %.*s. Blocking level %d\n",
-	     lockres->l_namelen, lockres->l_name, level);
+	mlog(ML_BASTS, "BAST fired for lockres %.*s, blocking %d, level %d\n",
+	     lockres->l_namelen, lockres->l_name, level, lockres->l_level);
 
 	spin_lock(&lockres->l_lock);
 	lockres->l_flags |= USER_LOCK_BLOCKED;
@@ -227,15 +229,15 @@ static void user_bast(void *opaque, int level)
 	wake_up(&lockres->l_event);
 }
 
-static void user_unlock_ast(void *opaque, enum dlm_status status)
+static void user_unlock_ast(struct ocfs2_dlm_lksb *lksb, int status)
 {
-	struct user_lock_res *lockres = opaque;
+	struct user_lock_res *lockres = user_lksb_to_lock_res(lksb);
 
-	mlog(0, "UNLOCK AST called on lock %.*s\n", lockres->l_namelen,
-	     lockres->l_name);
+	mlog(ML_BASTS, "UNLOCK AST fired for lockres %.*s, flags 0x%x\n",
+	     lockres->l_namelen, lockres->l_name, lockres->l_flags);
 
-	if (status != DLM_NORMAL && status != DLM_CANCELGRANT)
-		mlog(ML_ERROR, "Dlm returns status %d\n", status);
+	if (status)
+		mlog(ML_ERROR, "dlm returns status %d\n", status);
 
 	spin_lock(&lockres->l_lock);
 	/* The teardown flag gets set early during the unlock process,
@@ -243,7 +245,7 @@ static void user_unlock_ast(void *opaque, enum dlm_status status)
 	 * for a concurrent cancel. */
 	if (lockres->l_flags & USER_LOCK_IN_TEARDOWN
 	    && !(lockres->l_flags & USER_LOCK_IN_CANCEL)) {
-		lockres->l_level = LKM_IVMODE;
+		lockres->l_level = DLM_LOCK_IV;
 	} else if (status == DLM_CANCELGRANT) {
 		/* We tried to cancel a convert request, but it was
 		 * already granted. Don't clear the busy flag - the
@@ -254,7 +256,7 @@ static void user_unlock_ast(void *opaque, enum dlm_status status)
 	} else {
 		BUG_ON(!(lockres->l_flags & USER_LOCK_IN_CANCEL));
 		/* Cancel succeeded, we want to re-queue */
-		lockres->l_requested = LKM_IVMODE; /* cancel an
+		lockres->l_requested = DLM_LOCK_IV; /* cancel an
 						    * upconvert
 						    * request. */
 		lockres->l_flags &= ~USER_LOCK_IN_CANCEL;
@@ -271,6 +273,21 @@ out_noclear:
 	wake_up(&lockres->l_event);
 }
 
+/*
+ * This is the userdlmfs locking protocol version.
+ *
+ * See fs/ocfs2/dlmglue.c for more details on locking versions.
+ */
+static struct ocfs2_locking_protocol user_dlm_lproto = {
+	.lp_max_version = {
+		.pv_major = OCFS2_LOCKING_PROTOCOL_MAJOR,
+		.pv_minor = OCFS2_LOCKING_PROTOCOL_MINOR,
+	},
+	.lp_lock_ast		= user_ast,
+	.lp_blocking_ast	= user_bast,
+	.lp_unlock_ast		= user_unlock_ast,
+};
+
 static inline void user_dlm_drop_inode_ref(struct user_lock_res *lockres)
 {
 	struct inode *inode;
@@ -283,10 +300,10 @@ static void user_dlm_unblock_lock(struct work_struct *work)
 	int new_level, status;
 	struct user_lock_res *lockres =
 		container_of(work, struct user_lock_res, l_work);
-	struct dlm_ctxt *dlm = dlm_ctxt_from_user_lockres(lockres);
+	struct ocfs2_cluster_connection *conn =
+		cluster_connection_from_user_lockres(lockres);
 
-	mlog(0, "processing lockres %.*s\n", lockres->l_namelen,
-	     lockres->l_name);
+	mlog(0, "lockres %.*s\n", lockres->l_namelen, lockres->l_name);
 
 	spin_lock(&lockres->l_lock);
 
@@ -304,17 +321,23 @@ static void user_dlm_unblock_lock(struct work_struct *work)
 	 * flag, and finally we might get another bast which re-queues
 	 * us before our ast for the downconvert is called. */
 	if (!(lockres->l_flags & USER_LOCK_BLOCKED)) {
+		mlog(ML_BASTS, "lockres %.*s USER_LOCK_BLOCKED\n",
+		     lockres->l_namelen, lockres->l_name);
 		spin_unlock(&lockres->l_lock);
 		goto drop_ref;
 	}
 
 	if (lockres->l_flags & USER_LOCK_IN_TEARDOWN) {
+		mlog(ML_BASTS, "lockres %.*s USER_LOCK_IN_TEARDOWN\n",
+		     lockres->l_namelen, lockres->l_name);
 		spin_unlock(&lockres->l_lock);
 		goto drop_ref;
 	}
 
 	if (lockres->l_flags & USER_LOCK_BUSY) {
 		if (lockres->l_flags & USER_LOCK_IN_CANCEL) {
+			mlog(ML_BASTS, "lockres %.*s USER_LOCK_IN_CANCEL\n",
+			     lockres->l_namelen, lockres->l_name);
 			spin_unlock(&lockres->l_lock);
 			goto drop_ref;
 		}
@@ -322,32 +345,31 @@ static void user_dlm_unblock_lock(struct work_struct *work)
 		lockres->l_flags |= USER_LOCK_IN_CANCEL;
 		spin_unlock(&lockres->l_lock);
 
-		status = dlmunlock(dlm,
-				   &lockres->l_lksb,
-				   LKM_CANCEL,
-				   user_unlock_ast,
-				   lockres);
-		if (status != DLM_NORMAL)
-			user_log_dlm_error("dlmunlock", status, lockres);
+		status = ocfs2_dlm_unlock(conn, &lockres->l_lksb,
+					  DLM_LKF_CANCEL);
+		if (status)
+			user_log_dlm_error("ocfs2_dlm_unlock", status, lockres);
 		goto drop_ref;
 	}
 
 	/* If there are still incompat holders, we can exit safely
 	 * without worrying about re-queueing this lock as that will
 	 * happen on the last call to user_cluster_unlock. */
-	if ((lockres->l_blocking == LKM_EXMODE)
+	if ((lockres->l_blocking == DLM_LOCK_EX)
 	    && (lockres->l_ex_holders || lockres->l_ro_holders)) {
 		spin_unlock(&lockres->l_lock);
-		mlog(0, "can't downconvert for ex: ro = %u, ex = %u\n",
-			lockres->l_ro_holders, lockres->l_ex_holders);
+		mlog(ML_BASTS, "lockres %.*s, EX/PR Holders %u,%u\n",
+		     lockres->l_namelen, lockres->l_name,
+		     lockres->l_ex_holders, lockres->l_ro_holders);
 		goto drop_ref;
 	}
 
-	if ((lockres->l_blocking == LKM_PRMODE)
+	if ((lockres->l_blocking == DLM_LOCK_PR)
 	    && lockres->l_ex_holders) {
 		spin_unlock(&lockres->l_lock);
-		mlog(0, "can't downconvert for pr: ex = %u\n",
-			lockres->l_ex_holders);
+		mlog(ML_BASTS, "lockres %.*s, EX Holders %u\n",
+		     lockres->l_namelen, lockres->l_name,
+		     lockres->l_ex_holders);
 		goto drop_ref;
 	}
 
@@ -355,22 +377,17 @@ static void user_dlm_unblock_lock(struct work_struct *work)
 	new_level = user_highest_compat_lock_level(lockres->l_blocking);
 	lockres->l_requested = new_level;
 	lockres->l_flags |= USER_LOCK_BUSY;
-	mlog(0, "Downconvert lock from %d to %d\n",
-		lockres->l_level, new_level);
+	mlog(ML_BASTS, "lockres %.*s, downconvert %d => %d\n",
+	     lockres->l_namelen, lockres->l_name, lockres->l_level, new_level);
 	spin_unlock(&lockres->l_lock);
 
 	/* need lock downconvert request now... */
-	status = dlmlock(dlm,
-			 new_level,
-			 &lockres->l_lksb,
-			 LKM_CONVERT|LKM_VALBLK,
-			 lockres->l_name,
-			 lockres->l_namelen,
-			 user_ast,
-			 lockres,
-			 user_bast);
-	if (status != DLM_NORMAL) {
-		user_log_dlm_error("dlmlock", status, lockres);
+	status = ocfs2_dlm_lock(conn, new_level, &lockres->l_lksb,
+				DLM_LKF_CONVERT|DLM_LKF_VALBLK,
+				lockres->l_name,
+				lockres->l_namelen);
+	if (status) {
+		user_log_dlm_error("ocfs2_dlm_lock", status, lockres);
 		user_recover_from_dlm_error(lockres);
 	}
 
@@ -382,10 +399,10 @@ static inline void user_dlm_inc_holders(struct user_lock_res *lockres,
 					int level)
 {
 	switch(level) {
-	case LKM_EXMODE:
+	case DLM_LOCK_EX:
 		lockres->l_ex_holders++;
 		break;
-	case LKM_PRMODE:
+	case DLM_LOCK_PR:
 		lockres->l_ro_holders++;
 		break;
 	default:
@@ -410,20 +427,19 @@ int user_dlm_cluster_lock(struct user_lock_res *lockres,
 			  int lkm_flags)
 {
 	int status, local_flags;
-	struct dlm_ctxt *dlm = dlm_ctxt_from_user_lockres(lockres);
+	struct ocfs2_cluster_connection *conn =
+		cluster_connection_from_user_lockres(lockres);
 
-	if (level != LKM_EXMODE &&
-	    level != LKM_PRMODE) {
+	if (level != DLM_LOCK_EX &&
+	    level != DLM_LOCK_PR) {
 		mlog(ML_ERROR, "lockres %.*s: invalid request!\n",
 		     lockres->l_namelen, lockres->l_name);
 		status = -EINVAL;
 		goto bail;
 	}
 
-	mlog(0, "lockres %.*s: asking for %s lock, passed flags = 0x%x\n",
-	     lockres->l_namelen, lockres->l_name,
-	     (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE",
-	     lkm_flags);
+	mlog(ML_BASTS, "lockres %.*s, level %d, flags = 0x%x\n",
+	     lockres->l_namelen, lockres->l_name, level, lkm_flags);
 
 again:
 	if (signal_pending(current)) {
@@ -457,35 +473,26 @@ again:
 	}
 
 	if (level > lockres->l_level) {
-		local_flags = lkm_flags | LKM_VALBLK;
-		if (lockres->l_level != LKM_IVMODE)
-			local_flags |= LKM_CONVERT;
+		local_flags = lkm_flags | DLM_LKF_VALBLK;
+		if (lockres->l_level != DLM_LOCK_IV)
+			local_flags |= DLM_LKF_CONVERT;
 
 		lockres->l_requested = level;
 		lockres->l_flags |= USER_LOCK_BUSY;
 		spin_unlock(&lockres->l_lock);
 
-		BUG_ON(level == LKM_IVMODE);
-		BUG_ON(level == LKM_NLMODE);
+		BUG_ON(level == DLM_LOCK_IV);
+		BUG_ON(level == DLM_LOCK_NL);
 
 		/* call dlm_lock to upgrade lock now */
-		status = dlmlock(dlm,
-				 level,
-				 &lockres->l_lksb,
-				 local_flags,
-				 lockres->l_name,
-				 lockres->l_namelen,
-				 user_ast,
-				 lockres,
-				 user_bast);
-		if (status != DLM_NORMAL) {
-			if ((lkm_flags & LKM_NOQUEUE) &&
-			    (status == DLM_NOTQUEUED))
-				status = -EAGAIN;
-			else {
-				user_log_dlm_error("dlmlock", status, lockres);
-				status = -EINVAL;
-			}
+		status = ocfs2_dlm_lock(conn, level, &lockres->l_lksb,
+					local_flags, lockres->l_name,
+					lockres->l_namelen);
+		if (status) {
+			if ((lkm_flags & DLM_LKF_NOQUEUE) &&
+			    (status != -EAGAIN))
+				user_log_dlm_error("ocfs2_dlm_lock",
+						   status, lockres);
 			user_recover_from_dlm_error(lockres);
 			goto bail;
 		}
@@ -506,11 +513,11 @@ static inline void user_dlm_dec_holders(struct user_lock_res *lockres,
 					int level)
 {
 	switch(level) {
-	case LKM_EXMODE:
+	case DLM_LOCK_EX:
 		BUG_ON(!lockres->l_ex_holders);
 		lockres->l_ex_holders--;
 		break;
-	case LKM_PRMODE:
+	case DLM_LOCK_PR:
 		BUG_ON(!lockres->l_ro_holders);
 		lockres->l_ro_holders--;
 		break;
@@ -522,8 +529,8 @@ static inline void user_dlm_dec_holders(struct user_lock_res *lockres,
 void user_dlm_cluster_unlock(struct user_lock_res *lockres,
 			     int level)
 {
-	if (level != LKM_EXMODE &&
-	    level != LKM_PRMODE) {
+	if (level != DLM_LOCK_EX &&
+	    level != DLM_LOCK_PR) {
 		mlog(ML_ERROR, "lockres %.*s: invalid request!\n",
 		     lockres->l_namelen, lockres->l_name);
 		return;
@@ -540,33 +547,40 @@ void user_dlm_write_lvb(struct inode *inode,
 			unsigned int len)
 {
 	struct user_lock_res *lockres = &DLMFS_I(inode)->ip_lockres;
-	char *lvb = lockres->l_lksb.lvb;
+	char *lvb;
 
 	BUG_ON(len > DLM_LVB_LEN);
 
 	spin_lock(&lockres->l_lock);
 
-	BUG_ON(lockres->l_level < LKM_EXMODE);
+	BUG_ON(lockres->l_level < DLM_LOCK_EX);
+	lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
 	memcpy(lvb, val, len);
 
 	spin_unlock(&lockres->l_lock);
 }
 
-void user_dlm_read_lvb(struct inode *inode,
-		       char *val,
-		       unsigned int len)
+ssize_t user_dlm_read_lvb(struct inode *inode,
+			  char *val,
+			  unsigned int len)
 {
 	struct user_lock_res *lockres = &DLMFS_I(inode)->ip_lockres;
-	char *lvb = lockres->l_lksb.lvb;
+	char *lvb;
+	ssize_t ret = len;
 
 	BUG_ON(len > DLM_LVB_LEN);
 
 	spin_lock(&lockres->l_lock);
 
-	BUG_ON(lockres->l_level < LKM_PRMODE);
-	memcpy(val, lvb, len);
+	BUG_ON(lockres->l_level < DLM_LOCK_PR);
+	if (ocfs2_dlm_lvb_valid(&lockres->l_lksb)) {
+		lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
+		memcpy(val, lvb, len);
+	} else
+		ret = 0;
 
 	spin_unlock(&lockres->l_lock);
+	return ret;
 }
 
 void user_dlm_lock_res_init(struct user_lock_res *lockres,
@@ -576,9 +590,9 @@ void user_dlm_lock_res_init(struct user_lock_res *lockres,
 
 	spin_lock_init(&lockres->l_lock);
 	init_waitqueue_head(&lockres->l_event);
-	lockres->l_level = LKM_IVMODE;
-	lockres->l_requested = LKM_IVMODE;
-	lockres->l_blocking = LKM_IVMODE;
+	lockres->l_level = DLM_LOCK_IV;
+	lockres->l_requested = DLM_LOCK_IV;
+	lockres->l_blocking = DLM_LOCK_IV;
 
 	/* should have been checked before getting here. */
 	BUG_ON(dentry->d_name.len >= USER_DLM_LOCK_ID_MAX_LEN);
@@ -592,9 +606,10 @@ void user_dlm_lock_res_init(struct user_lock_res *lockres,
 int user_dlm_destroy_lock(struct user_lock_res *lockres)
 {
 	int status = -EBUSY;
-	struct dlm_ctxt *dlm = dlm_ctxt_from_user_lockres(lockres);
+	struct ocfs2_cluster_connection *conn =
+		cluster_connection_from_user_lockres(lockres);
 
-	mlog(0, "asked to destroy %.*s\n", lockres->l_namelen, lockres->l_name);
+	mlog(ML_BASTS, "lockres %.*s\n", lockres->l_namelen, lockres->l_name);
 
 	spin_lock(&lockres->l_lock);
 	if (lockres->l_flags & USER_LOCK_IN_TEARDOWN) {
@@ -627,14 +642,9 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres)
 	lockres->l_flags |= USER_LOCK_BUSY;
 	spin_unlock(&lockres->l_lock);
 
-	status = dlmunlock(dlm,
-			   &lockres->l_lksb,
-			   LKM_VALBLK,
-			   user_unlock_ast,
-			   lockres);
-	if (status != DLM_NORMAL) {
-		user_log_dlm_error("dlmunlock", status, lockres);
-		status = -EINVAL;
+	status = ocfs2_dlm_unlock(conn, &lockres->l_lksb, DLM_LKF_VALBLK);
+	if (status) {
+		user_log_dlm_error("ocfs2_dlm_unlock", status, lockres);
 		goto bail;
 	}
 
@@ -645,32 +655,34 @@ bail:
 	return status;
 }
 
-struct dlm_ctxt *user_dlm_register_context(struct qstr *name,
-					   struct dlm_protocol_version *proto)
+static void user_dlm_recovery_handler_noop(int node_num,
+					   void *recovery_data)
 {
-	struct dlm_ctxt *dlm;
-	u32 dlm_key;
-	char *domain;
-
-	domain = kmalloc(name->len + 1, GFP_NOFS);
-	if (!domain) {
-		mlog_errno(-ENOMEM);
-		return ERR_PTR(-ENOMEM);
-	}
+	/* We ignore recovery events */
+	return;
+}
 
-	dlm_key = crc32_le(0, name->name, name->len);
+void user_dlm_set_locking_protocol(void)
+{
+	ocfs2_stack_glue_set_max_proto_version(&user_dlm_lproto.lp_max_version);
+}
 
-	snprintf(domain, name->len + 1, "%.*s", name->len, name->name);
+struct ocfs2_cluster_connection *user_dlm_register(struct qstr *name)
+{
+	int rc;
+	struct ocfs2_cluster_connection *conn;
 
-	dlm = dlm_register_domain(domain, dlm_key, proto);
-	if (IS_ERR(dlm))
-		mlog_errno(PTR_ERR(dlm));
+	rc = ocfs2_cluster_connect_agnostic(name->name, name->len,
+					    &user_dlm_lproto,
+					    user_dlm_recovery_handler_noop,
+					    NULL, &conn);
+	if (rc)
+		mlog_errno(rc);
 
-	kfree(domain);
-	return dlm;
+	return rc ? ERR_PTR(rc) : conn;
 }
 
-void user_dlm_unregister_context(struct dlm_ctxt *dlm)
+void user_dlm_unregister(struct ocfs2_cluster_connection *conn)
 {
-	dlm_unregister_domain(dlm);
+	ocfs2_cluster_disconnect(conn, 0);
 }
diff --git a/fs/ocfs2/dlm/userdlm.h b/fs/ocfs2/dlmfs/userdlm.h
index 0c3cc03c61fa..3b42d79531d7 100644
--- a/fs/ocfs2/dlm/userdlm.h
+++ b/fs/ocfs2/dlmfs/userdlm.h
@@ -57,7 +57,7 @@ struct user_lock_res {
 	int                      l_level;
 	unsigned int             l_ro_holders;
 	unsigned int             l_ex_holders;
-	struct dlm_lockstatus    l_lksb;
+	struct ocfs2_dlm_lksb    l_lksb;
 
 	int                      l_requested;
 	int                      l_blocking;
@@ -80,15 +80,15 @@ void user_dlm_cluster_unlock(struct user_lock_res *lockres,
 void user_dlm_write_lvb(struct inode *inode,
 			const char *val,
 			unsigned int len);
-void user_dlm_read_lvb(struct inode *inode,
-		       char *val,
-		       unsigned int len);
-struct dlm_ctxt *user_dlm_register_context(struct qstr *name,
-					   struct dlm_protocol_version *proto);
-void user_dlm_unregister_context(struct dlm_ctxt *dlm);
+ssize_t user_dlm_read_lvb(struct inode *inode,
+			  char *val,
+			  unsigned int len);
+struct ocfs2_cluster_connection *user_dlm_register(struct qstr *name);
+void user_dlm_unregister(struct ocfs2_cluster_connection *conn);
+void user_dlm_set_locking_protocol(void);
 
 struct dlmfs_inode_private {
-	struct dlm_ctxt             *ip_dlm;
+	struct ocfs2_cluster_connection	*ip_conn;
 
 	struct user_lock_res ip_lockres; /* unused for directories. */
 	struct inode         *ip_parent;
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index e044019cb3b1..8298608d4165 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -297,6 +297,11 @@ static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres)
 		lockres->l_type == OCFS2_LOCK_TYPE_OPEN;
 }
 
+static inline struct ocfs2_lock_res *ocfs2_lksb_to_lock_res(struct ocfs2_dlm_lksb *lksb)
+{
+	return container_of(lksb, struct ocfs2_lock_res, l_lksb);
+}
+
 static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres)
 {
 	BUG_ON(!ocfs2_is_inode_lock(lockres));
@@ -927,6 +932,10 @@ static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres,
 		lockres->l_blocking = level;
 	}
 
+	mlog(ML_BASTS, "lockres %s, block %d, level %d, l_block %d, dwn %d\n",
+	     lockres->l_name, level, lockres->l_level, lockres->l_blocking,
+	     needs_downconvert);
+
 	if (needs_downconvert)
 		lockres_or_flags(lockres, OCFS2_LOCK_BLOCKED);
 
@@ -1040,18 +1049,17 @@ static unsigned int lockres_set_pending(struct ocfs2_lock_res *lockres)
 	return lockres->l_pending_gen;
 }
 
-
-static void ocfs2_blocking_ast(void *opaque, int level)
+static void ocfs2_blocking_ast(struct ocfs2_dlm_lksb *lksb, int level)
 {
-	struct ocfs2_lock_res *lockres = opaque;
+	struct ocfs2_lock_res *lockres = ocfs2_lksb_to_lock_res(lksb);
 	struct ocfs2_super *osb = ocfs2_get_lockres_osb(lockres);
 	int needs_downconvert;
 	unsigned long flags;
 
 	BUG_ON(level <= DLM_LOCK_NL);
 
-	mlog(0, "BAST fired for lockres %s, blocking %d, level %d type %s\n",
-	     lockres->l_name, level, lockres->l_level,
+	mlog(ML_BASTS, "BAST fired for lockres %s, blocking %d, level %d, "
+	     "type %s\n", lockres->l_name, level, lockres->l_level,
 	     ocfs2_lock_type_string(lockres->l_type));
 
 	/*
@@ -1072,9 +1080,9 @@ static void ocfs2_blocking_ast(void *opaque, int level)
 	ocfs2_wake_downconvert_thread(osb);
 }
 
-static void ocfs2_locking_ast(void *opaque)
+static void ocfs2_locking_ast(struct ocfs2_dlm_lksb *lksb)
 {
-	struct ocfs2_lock_res *lockres = opaque;
+	struct ocfs2_lock_res *lockres = ocfs2_lksb_to_lock_res(lksb);
 	struct ocfs2_super *osb = ocfs2_get_lockres_osb(lockres);
 	unsigned long flags;
 	int status;
@@ -1095,6 +1103,10 @@ static void ocfs2_locking_ast(void *opaque)
 		return;
 	}
 
+	mlog(ML_BASTS, "AST fired for lockres %s, action %d, unlock %d, "
+	     "level %d => %d\n", lockres->l_name, lockres->l_action,
+	     lockres->l_unlock_action, lockres->l_level, lockres->l_requested);
+
 	switch(lockres->l_action) {
 	case OCFS2_AST_ATTACH:
 		ocfs2_generic_handle_attach_action(lockres);
@@ -1107,8 +1119,8 @@ static void ocfs2_locking_ast(void *opaque)
 		ocfs2_generic_handle_downconvert_action(lockres);
 		break;
 	default:
-		mlog(ML_ERROR, "lockres %s: ast fired with invalid action: %u "
-		     "lockres flags = 0x%lx, unlock action: %u\n",
+		mlog(ML_ERROR, "lockres %s: AST fired with invalid action: %u, "
+		     "flags 0x%lx, unlock: %u\n",
 		     lockres->l_name, lockres->l_action, lockres->l_flags,
 		     lockres->l_unlock_action);
 		BUG();
@@ -1134,6 +1146,88 @@ out:
 	spin_unlock_irqrestore(&lockres->l_lock, flags);
 }
 
+static void ocfs2_unlock_ast(struct ocfs2_dlm_lksb *lksb, int error)
+{
+	struct ocfs2_lock_res *lockres = ocfs2_lksb_to_lock_res(lksb);
+	unsigned long flags;
+
+	mlog_entry_void();
+
+	mlog(ML_BASTS, "UNLOCK AST fired for lockres %s, action = %d\n",
+	     lockres->l_name, lockres->l_unlock_action);
+
+	spin_lock_irqsave(&lockres->l_lock, flags);
+	if (error) {
+		mlog(ML_ERROR, "Dlm passes error %d for lock %s, "
+		     "unlock_action %d\n", error, lockres->l_name,
+		     lockres->l_unlock_action);
+		spin_unlock_irqrestore(&lockres->l_lock, flags);
+		mlog_exit_void();
+		return;
+	}
+
+	switch(lockres->l_unlock_action) {
+	case OCFS2_UNLOCK_CANCEL_CONVERT:
+		mlog(0, "Cancel convert success for %s\n", lockres->l_name);
+		lockres->l_action = OCFS2_AST_INVALID;
+		/* Downconvert thread may have requeued this lock, we
+		 * need to wake it. */
+		if (lockres->l_flags & OCFS2_LOCK_BLOCKED)
+			ocfs2_wake_downconvert_thread(ocfs2_get_lockres_osb(lockres));
+		break;
+	case OCFS2_UNLOCK_DROP_LOCK:
+		lockres->l_level = DLM_LOCK_IV;
+		break;
+	default:
+		BUG();
+	}
+
+	lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
+	lockres->l_unlock_action = OCFS2_UNLOCK_INVALID;
+	wake_up(&lockres->l_event);
+	spin_unlock_irqrestore(&lockres->l_lock, flags);
+
+	mlog_exit_void();
+}
+
+/*
+ * This is the filesystem locking protocol.  It provides the lock handling
+ * hooks for the underlying DLM.  It has a maximum version number.
+ * The version number allows interoperability with systems running at
+ * the same major number and an equal or smaller minor number.
+ *
+ * Whenever the filesystem does new things with locks (adds or removes a
+ * lock, orders them differently, does different things underneath a lock),
+ * the version must be changed.  The protocol is negotiated when joining
+ * the dlm domain.  A node may join the domain if its major version is
+ * identical to all other nodes and its minor version is greater than
+ * or equal to all other nodes.  When its minor version is greater than
+ * the other nodes, it will run at the minor version specified by the
+ * other nodes.
+ *
+ * If a locking change is made that will not be compatible with older
+ * versions, the major number must be increased and the minor version set
+ * to zero.  If a change merely adds a behavior that can be disabled when
+ * speaking to older versions, the minor version must be increased.  If a
+ * change adds a fully backwards compatible change (eg, LVB changes that
+ * are just ignored by older versions), the version does not need to be
+ * updated.
+ */
+static struct ocfs2_locking_protocol lproto = {
+	.lp_max_version = {
+		.pv_major = OCFS2_LOCKING_PROTOCOL_MAJOR,
+		.pv_minor = OCFS2_LOCKING_PROTOCOL_MINOR,
+	},
+	.lp_lock_ast		= ocfs2_locking_ast,
+	.lp_blocking_ast	= ocfs2_blocking_ast,
+	.lp_unlock_ast		= ocfs2_unlock_ast,
+};
+
+void ocfs2_set_locking_protocol(void)
+{
+	ocfs2_stack_glue_set_max_proto_version(&lproto.lp_max_version);
+}
+
 static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres,
 						int convert)
 {
@@ -1189,8 +1283,7 @@ static int ocfs2_lock_create(struct ocfs2_super *osb,
 			     &lockres->l_lksb,
 			     dlm_flags,
 			     lockres->l_name,
-			     OCFS2_LOCK_ID_MAX_LEN - 1,
-			     lockres);
+			     OCFS2_LOCK_ID_MAX_LEN - 1);
 	lockres_clear_pending(lockres, gen, osb);
 	if (ret) {
 		ocfs2_log_dlm_error("ocfs2_dlm_lock", ret, lockres);
@@ -1412,7 +1505,7 @@ again:
 		BUG_ON(level == DLM_LOCK_IV);
 		BUG_ON(level == DLM_LOCK_NL);
 
-		mlog(0, "lock %s, convert from %d to level = %d\n",
+		mlog(ML_BASTS, "lockres %s, convert from %d to %d\n",
 		     lockres->l_name, lockres->l_level, level);
 
 		/* call dlm_lock to upgrade lock now */
@@ -1421,8 +1514,7 @@ again:
 				     &lockres->l_lksb,
 				     lkm_flags,
 				     lockres->l_name,
-				     OCFS2_LOCK_ID_MAX_LEN - 1,
-				     lockres);
+				     OCFS2_LOCK_ID_MAX_LEN - 1);
 		lockres_clear_pending(lockres, gen, osb);
 		if (ret) {
 			if (!(lkm_flags & DLM_LKF_NOQUEUE) ||
@@ -1859,8 +1951,7 @@ int ocfs2_file_lock(struct file *file, int ex, int trylock)
 	spin_unlock_irqrestore(&lockres->l_lock, flags);
 
 	ret = ocfs2_dlm_lock(osb->cconn, level, &lockres->l_lksb, lkm_flags,
-			     lockres->l_name, OCFS2_LOCK_ID_MAX_LEN - 1,
-			     lockres);
+			     lockres->l_name, OCFS2_LOCK_ID_MAX_LEN - 1);
 	if (ret) {
 		if (!trylock || (ret != -EAGAIN)) {
 			ocfs2_log_dlm_error("ocfs2_dlm_lock", ret, lockres);
@@ -2989,7 +3080,7 @@ int ocfs2_dlm_init(struct ocfs2_super *osb)
 	status = ocfs2_cluster_connect(osb->osb_cluster_stack,
 				       osb->uuid_str,
 				       strlen(osb->uuid_str),
-				       ocfs2_do_node_down, osb,
+				       &lproto, ocfs2_do_node_down, osb,
 				       &conn);
 	if (status) {
 		mlog_errno(status);
@@ -3056,50 +3147,6 @@ void ocfs2_dlm_shutdown(struct ocfs2_super *osb,
 	mlog_exit_void();
 }
 
-static void ocfs2_unlock_ast(void *opaque, int error)
-{
-	struct ocfs2_lock_res *lockres = opaque;
-	unsigned long flags;
-
-	mlog_entry_void();
-
-	mlog(0, "UNLOCK AST called on lock %s, action = %d\n", lockres->l_name,
-	     lockres->l_unlock_action);
-
-	spin_lock_irqsave(&lockres->l_lock, flags);
-	if (error) {
-		mlog(ML_ERROR, "Dlm passes error %d for lock %s, "
-		     "unlock_action %d\n", error, lockres->l_name,
-		     lockres->l_unlock_action);
-		spin_unlock_irqrestore(&lockres->l_lock, flags);
-		mlog_exit_void();
-		return;
-	}
-
-	switch(lockres->l_unlock_action) {
-	case OCFS2_UNLOCK_CANCEL_CONVERT:
-		mlog(0, "Cancel convert success for %s\n", lockres->l_name);
-		lockres->l_action = OCFS2_AST_INVALID;
-		/* Downconvert thread may have requeued this lock, we
-		 * need to wake it. */
-		if (lockres->l_flags & OCFS2_LOCK_BLOCKED)
-			ocfs2_wake_downconvert_thread(ocfs2_get_lockres_osb(lockres));
-		break;
-	case OCFS2_UNLOCK_DROP_LOCK:
-		lockres->l_level = DLM_LOCK_IV;
-		break;
-	default:
-		BUG();
-	}
-
-	lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
-	lockres->l_unlock_action = OCFS2_UNLOCK_INVALID;
-	wake_up(&lockres->l_event);
-	spin_unlock_irqrestore(&lockres->l_lock, flags);
-
-	mlog_exit_void();
-}
-
 static int ocfs2_drop_lock(struct ocfs2_super *osb,
 			   struct ocfs2_lock_res *lockres)
 {
@@ -3167,8 +3214,7 @@ static int ocfs2_drop_lock(struct ocfs2_super *osb,
 
 	mlog(0, "lock %s\n", lockres->l_name);
 
-	ret = ocfs2_dlm_unlock(osb->cconn, &lockres->l_lksb, lkm_flags,
-			       lockres);
+	ret = ocfs2_dlm_unlock(osb->cconn, &lockres->l_lksb, lkm_flags);
 	if (ret) {
 		ocfs2_log_dlm_error("ocfs2_dlm_unlock", ret, lockres);
 		mlog(ML_ERROR, "lockres flags: %lu\n", lockres->l_flags);
@@ -3276,13 +3322,20 @@ static unsigned int ocfs2_prepare_downconvert(struct ocfs2_lock_res *lockres,
 	BUG_ON(lockres->l_blocking <= DLM_LOCK_NL);
 
 	if (lockres->l_level <= new_level) {
-		mlog(ML_ERROR, "lockres->l_level (%d) <= new_level (%d)\n",
-		     lockres->l_level, new_level);
+		mlog(ML_ERROR, "lockres %s, lvl %d <= %d, blcklst %d, mask %d, "
+		     "type %d, flags 0x%lx, hold %d %d, act %d %d, req %d, "
+		     "block %d, pgen %d\n", lockres->l_name, lockres->l_level,
+		     new_level, list_empty(&lockres->l_blocked_list),
+		     list_empty(&lockres->l_mask_waiters), lockres->l_type,
+		     lockres->l_flags, lockres->l_ro_holders,
+		     lockres->l_ex_holders, lockres->l_action,
+		     lockres->l_unlock_action, lockres->l_requested,
+		     lockres->l_blocking, lockres->l_pending_gen);
 		BUG();
 	}
 
-	mlog(0, "lock %s, new_level = %d, l_blocking = %d\n",
-	     lockres->l_name, new_level, lockres->l_blocking);
+	mlog(ML_BASTS, "lockres %s, level %d => %d, blocking %d\n",
+	     lockres->l_name, lockres->l_level, new_level, lockres->l_blocking);
 
 	lockres->l_action = OCFS2_AST_DOWNCONVERT;
 	lockres->l_requested = new_level;
@@ -3301,6 +3354,9 @@ static int ocfs2_downconvert_lock(struct ocfs2_super *osb,
 
 	mlog_entry_void();
 
+	mlog(ML_BASTS, "lockres %s, level %d => %d\n", lockres->l_name,
+	     lockres->l_level, new_level);
+
 	if (lvb)
 		dlm_flags |= DLM_LKF_VALBLK;
 
@@ -3309,8 +3365,7 @@ static int ocfs2_downconvert_lock(struct ocfs2_super *osb,
 			     &lockres->l_lksb,
 			     dlm_flags,
 			     lockres->l_name,
-			     OCFS2_LOCK_ID_MAX_LEN - 1,
-			     lockres);
+			     OCFS2_LOCK_ID_MAX_LEN - 1);
 	lockres_clear_pending(lockres, generation, osb);
 	if (ret) {
 		ocfs2_log_dlm_error("ocfs2_dlm_lock", ret, lockres);
@@ -3331,14 +3386,12 @@ static int ocfs2_prepare_cancel_convert(struct ocfs2_super *osb,
 	assert_spin_locked(&lockres->l_lock);
 
 	mlog_entry_void();
-	mlog(0, "lock %s\n", lockres->l_name);
 
 	if (lockres->l_unlock_action == OCFS2_UNLOCK_CANCEL_CONVERT) {
 		/* If we're already trying to cancel a lock conversion
 		 * then just drop the spinlock and allow the caller to
 		 * requeue this lock. */
-
-		mlog(0, "Lockres %s, skip convert\n", lockres->l_name);
+		mlog(ML_BASTS, "lockres %s, skip convert\n", lockres->l_name);
 		return 0;
 	}
 
@@ -3353,6 +3406,8 @@ static int ocfs2_prepare_cancel_convert(struct ocfs2_super *osb,
 			"lock %s, invalid flags: 0x%lx\n",
 			lockres->l_name, lockres->l_flags);
 
+	mlog(ML_BASTS, "lockres %s\n", lockres->l_name);
+
 	return 1;
 }
 
@@ -3362,16 +3417,15 @@ static int ocfs2_cancel_convert(struct ocfs2_super *osb,
 	int ret;
 
 	mlog_entry_void();
-	mlog(0, "lock %s\n", lockres->l_name);
 
 	ret = ocfs2_dlm_unlock(osb->cconn, &lockres->l_lksb,
-			       DLM_LKF_CANCEL, lockres);
+			       DLM_LKF_CANCEL);
 	if (ret) {
 		ocfs2_log_dlm_error("ocfs2_dlm_unlock", ret, lockres);
 		ocfs2_recover_from_dlm_error(lockres, 0);
 	}
 
-	mlog(0, "lock %s return from ocfs2_dlm_unlock\n", lockres->l_name);
+	mlog(ML_BASTS, "lockres %s\n", lockres->l_name);
 
 	mlog_exit(ret);
 	return ret;
@@ -3428,8 +3482,11 @@ recheck:
 		 * at the same time they set OCFS2_DLM_BUSY.  They must
 		 * clear OCFS2_DLM_PENDING after dlm_lock() returns.
 		 */
-		if (lockres->l_flags & OCFS2_LOCK_PENDING)
+		if (lockres->l_flags & OCFS2_LOCK_PENDING) {
+			mlog(ML_BASTS, "lockres %s, ReQ: Pending\n",
+			     lockres->l_name);
 			goto leave_requeue;
+		}
 
 		ctl->requeue = 1;
 		ret = ocfs2_prepare_cancel_convert(osb, lockres);
@@ -3461,6 +3518,7 @@ recheck:
 	 */
 	if (lockres->l_level == DLM_LOCK_NL) {
 		BUG_ON(lockres->l_ex_holders || lockres->l_ro_holders);
+		mlog(ML_BASTS, "lockres %s, Aborting dc\n", lockres->l_name);
 		lockres->l_blocking = DLM_LOCK_NL;
 		lockres_clear_flags(lockres, OCFS2_LOCK_BLOCKED);
 		spin_unlock_irqrestore(&lockres->l_lock, flags);
@@ -3470,28 +3528,41 @@ recheck:
 	/* if we're blocking an exclusive and we have *any* holders,
 	 * then requeue. */
 	if ((lockres->l_blocking == DLM_LOCK_EX)
-	    && (lockres->l_ex_holders || lockres->l_ro_holders))
+	    && (lockres->l_ex_holders || lockres->l_ro_holders)) {
+		mlog(ML_BASTS, "lockres %s, ReQ: EX/PR Holders %u,%u\n",
+		     lockres->l_name, lockres->l_ex_holders,
+		     lockres->l_ro_holders);
 		goto leave_requeue;
+	}
 
 	/* If it's a PR we're blocking, then only
 	 * requeue if we've got any EX holders */
 	if (lockres->l_blocking == DLM_LOCK_PR &&
-	    lockres->l_ex_holders)
+	    lockres->l_ex_holders) {
+		mlog(ML_BASTS, "lockres %s, ReQ: EX Holders %u\n",
+		     lockres->l_name, lockres->l_ex_holders);
 		goto leave_requeue;
+	}
 
 	/*
 	 * Can we get a lock in this state if the holder counts are
 	 * zero? The meta data unblock code used to check this.
 	 */
 	if ((lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH)
-	    && (lockres->l_flags & OCFS2_LOCK_REFRESHING))
+	    && (lockres->l_flags & OCFS2_LOCK_REFRESHING)) {
+		mlog(ML_BASTS, "lockres %s, ReQ: Lock Refreshing\n",
+		     lockres->l_name);
 		goto leave_requeue;
+	}
 
 	new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking);
 
 	if (lockres->l_ops->check_downconvert
-	    && !lockres->l_ops->check_downconvert(lockres, new_level))
+	    && !lockres->l_ops->check_downconvert(lockres, new_level)) {
+		mlog(ML_BASTS, "lockres %s, ReQ: Checkpointing\n",
+		     lockres->l_name);
 		goto leave_requeue;
+	}
 
 	/* If we get here, then we know that there are no more
 	 * incompatible holders (and anyone asking for an incompatible
@@ -3509,13 +3580,19 @@ recheck:
 
 	ctl->unblock_action = lockres->l_ops->downconvert_worker(lockres, blocking);
 
-	if (ctl->unblock_action == UNBLOCK_STOP_POST)
+	if (ctl->unblock_action == UNBLOCK_STOP_POST) {
+		mlog(ML_BASTS, "lockres %s, UNBLOCK_STOP_POST\n",
+		     lockres->l_name);
 		goto leave;
+	}
 
 	spin_lock_irqsave(&lockres->l_lock, flags);
 	if ((blocking != lockres->l_blocking) || (level != lockres->l_level)) {
 		/* If this changed underneath us, then we can't drop
 		 * it just yet. */
+		mlog(ML_BASTS, "lockres %s, block=%d:%d, level=%d:%d, "
+		     "Recheck\n", lockres->l_name, blocking,
+		     lockres->l_blocking, level, lockres->l_level);
 		goto recheck;
 	}
 
@@ -3910,45 +3987,6 @@ void ocfs2_refcount_unlock(struct ocfs2_refcount_tree *ref_tree, int ex)
 		ocfs2_cluster_unlock(osb, lockres, level);
 }
 
-/*
- * This is the filesystem locking protocol.  It provides the lock handling
- * hooks for the underlying DLM.  It has a maximum version number.
- * The version number allows interoperability with systems running at
- * the same major number and an equal or smaller minor number.
- *
- * Whenever the filesystem does new things with locks (adds or removes a
- * lock, orders them differently, does different things underneath a lock),
- * the version must be changed.  The protocol is negotiated when joining
- * the dlm domain.  A node may join the domain if its major version is
- * identical to all other nodes and its minor version is greater than
- * or equal to all other nodes.  When its minor version is greater than
- * the other nodes, it will run at the minor version specified by the
- * other nodes.
- *
- * If a locking change is made that will not be compatible with older
- * versions, the major number must be increased and the minor version set
- * to zero.  If a change merely adds a behavior that can be disabled when
- * speaking to older versions, the minor version must be increased.  If a
- * change adds a fully backwards compatible change (eg, LVB changes that
- * are just ignored by older versions), the version does not need to be
- * updated.
- */
-static struct ocfs2_locking_protocol lproto = {
-	.lp_max_version = {
-		.pv_major = OCFS2_LOCKING_PROTOCOL_MAJOR,
-		.pv_minor = OCFS2_LOCKING_PROTOCOL_MINOR,
-	},
-	.lp_lock_ast		= ocfs2_locking_ast,
-	.lp_blocking_ast	= ocfs2_blocking_ast,
-	.lp_unlock_ast		= ocfs2_unlock_ast,
-};
-
-void ocfs2_set_locking_protocol(void)
-{
-	ocfs2_stack_glue_set_locking_protocol(&lproto);
-}
-
-
 static void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
 				       struct ocfs2_lock_res *lockres)
 {
@@ -3965,7 +4003,7 @@ static void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
 	BUG_ON(!lockres);
 	BUG_ON(!lockres->l_ops);
 
-	mlog(0, "lockres %s blocked.\n", lockres->l_name);
+	mlog(ML_BASTS, "lockres %s blocked\n", lockres->l_name);
 
 	/* Detect whether a lock has been marked as going away while
 	 * the downconvert thread was processing other things. A lock can
@@ -3988,7 +4026,7 @@ unqueue:
 	} else
 		ocfs2_schedule_blocked_lock(osb, lockres);
 
-	mlog(0, "lockres %s, requeue = %s.\n", lockres->l_name,
+	mlog(ML_BASTS, "lockres %s, requeue = %s.\n", lockres->l_name,
 	     ctl.requeue ? "yes" : "no");
 	spin_unlock_irqrestore(&lockres->l_lock, flags);
 
@@ -4010,7 +4048,7 @@ static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb,
 		/* Do not schedule a lock for downconvert when it's on
 		 * the way to destruction - any nodes wanting access
 		 * to the resource will get it soon. */
-		mlog(0, "Lockres %s won't be scheduled: flags 0x%lx\n",
+		mlog(ML_BASTS, "lockres %s won't be scheduled: flags 0x%lx\n",
 		     lockres->l_name, lockres->l_flags);
 		return;
 	}
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 558ce0312421..17947dc8341e 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -107,6 +107,9 @@ static int ocfs2_file_open(struct inode *inode, struct file *file)
 	mlog_entry("(0x%p, 0x%p, '%.*s')\n", inode, file,
 		   file->f_path.dentry->d_name.len, file->f_path.dentry->d_name.name);
 
+	if (file->f_mode & FMODE_WRITE)
+		dquot_initialize(inode);
+
 	spin_lock(&oi->ip_lock);
 
 	/* Check that the inode hasn't been wiped from disk by another
@@ -629,11 +632,10 @@ restart_all:
 	}
 
 restarted_transaction:
-	if (vfs_dq_alloc_space_nodirty(inode, ocfs2_clusters_to_bytes(osb->sb,
-	    clusters_to_add))) {
-		status = -EDQUOT;
+	status = dquot_alloc_space_nodirty(inode,
+			ocfs2_clusters_to_bytes(osb->sb, clusters_to_add));
+	if (status)
 		goto leave;
-	}
 	did_quota = 1;
 
 	/* reserve a write to the file entry early on - that we if we
@@ -674,7 +676,7 @@ restarted_transaction:
 	clusters_to_add -= (OCFS2_I(inode)->ip_clusters - prev_clusters);
 	spin_unlock(&OCFS2_I(inode)->ip_lock);
 	/* Release unused quota reservation */
-	vfs_dq_free_space(inode,
+	dquot_free_space(inode,
 			ocfs2_clusters_to_bytes(osb->sb, clusters_to_add));
 	did_quota = 0;
 
@@ -710,7 +712,7 @@ restarted_transaction:
 
 leave:
 	if (status < 0 && did_quota)
-		vfs_dq_free_space(inode,
+		dquot_free_space(inode,
 			ocfs2_clusters_to_bytes(osb->sb, clusters_to_add));
 	if (handle) {
 		ocfs2_commit_trans(osb, handle);
@@ -978,6 +980,8 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 
 	size_change = S_ISREG(inode->i_mode) && attr->ia_valid & ATTR_SIZE;
 	if (size_change) {
+		dquot_initialize(inode);
+
 		status = ocfs2_rw_lock(inode, 1);
 		if (status < 0) {
 			mlog_errno(status);
@@ -993,10 +997,9 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (size_change && attr->ia_size != i_size_read(inode)) {
-		if (attr->ia_size > sb->s_maxbytes) {
-			status = -EFBIG;
+		status = inode_newsize_ok(inode, attr->ia_size);
+		if (status)
 			goto bail_unlock;
-		}
 
 		if (i_size_read(inode) > attr->ia_size) {
 			if (ocfs2_should_order_data(inode)) {
@@ -1021,7 +1024,7 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 		/*
 		 * Gather pointers to quota structures so that allocation /
 		 * freeing of quota structures happens here and not inside
-		 * vfs_dq_transfer() where we have problems with lock ordering
+		 * dquot_transfer() where we have problems with lock ordering
 		 */
 		if (attr->ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid
 		    && OCFS2_HAS_RO_COMPAT_FEATURE(sb,
@@ -1054,7 +1057,7 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 			mlog_errno(status);
 			goto bail_unlock;
 		}
-		status = vfs_dq_transfer(inode, attr) ? -EDQUOT : 0;
+		status = dquot_transfer(inode, attr);
 		if (status < 0)
 			goto bail_commit;
 	} else {
@@ -1836,6 +1839,8 @@ static int ocfs2_prepare_inode_for_write(struct dentry *dentry,
 							       &meta_level);
 			if (has_refcount)
 				*has_refcount = 1;
+			if (direct_io)
+				*direct_io = 0;
 		}
 
 		if (ret < 0) {
@@ -1859,10 +1864,6 @@ static int ocfs2_prepare_inode_for_write(struct dentry *dentry,
 			break;
 		}
 
-		if (has_refcount && *has_refcount == 1) {
-			*direct_io = 0;
-			break;
-		}
 		/*
 		 * Allowing concurrent direct writes means
 		 * i_size changes wouldn't be synchronized, so
@@ -2043,7 +2044,7 @@ out_dio:
 	 * async dio is going to do it in the future or an end_io after an
 	 * error has already done it.
 	 */
-	if (ret == -EIOCBQUEUED || !ocfs2_iocb_is_rw_locked(iocb)) {
+	if ((ret == -EIOCBQUEUED) || (!ocfs2_iocb_is_rw_locked(iocb))) {
 		rw_level = -1;
 		have_alloc_sem = 0;
 	}
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index 88459bdd1ff3..278a223aae14 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -665,7 +665,7 @@ static int ocfs2_remove_inode(struct inode *inode,
 	}
 
 	ocfs2_remove_from_cache(INODE_CACHE(inode), di_bh);
-	vfs_dq_free_inode(inode);
+	dquot_free_inode(inode);
 
 	status = ocfs2_free_dinode(handle, inode_alloc_inode,
 				   inode_alloc_bh, di);
@@ -971,6 +971,8 @@ void ocfs2_delete_inode(struct inode *inode)
 		goto bail;
 	}
 
+	dquot_initialize(inode);
+
 	if (!ocfs2_inode_is_valid_to_delete(inode)) {
 		/* It's probably not necessary to truncate_inode_pages
 		 * here but we do it for safety anyway (it will most
@@ -1087,6 +1089,8 @@ void ocfs2_clear_inode(struct inode *inode)
 	mlog_bug_on_msg(OCFS2_SB(inode->i_sb) == NULL,
 			"Inode=%lu\n", inode->i_ino);
 
+	dquot_drop(inode);
+
 	/* To preven remote deletes we hold open lock before, now it
 	 * is time to unlock PR and EX open locks. */
 	ocfs2_open_unlock(inode);
diff --git a/fs/ocfs2/ioctl.h b/fs/ocfs2/ioctl.h
index cf9a5ee30fef..0cd5323bd3f0 100644
--- a/fs/ocfs2/ioctl.h
+++ b/fs/ocfs2/ioctl.h
@@ -7,10 +7,10 @@
  *
  */
 
-#ifndef OCFS2_IOCTL_H
-#define OCFS2_IOCTL_H
+#ifndef OCFS2_IOCTL_PROTO_H
+#define OCFS2_IOCTL_PROTO_H
 
 long ocfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg);
 long ocfs2_compat_ioctl(struct file *file, unsigned cmd, unsigned long arg);
 
-#endif /* OCFS2_IOCTL_H */
+#endif /* OCFS2_IOCTL_PROTO_H */
diff --git a/fs/ocfs2/localalloc.c b/fs/ocfs2/localalloc.c
index ac10f83edb95..ca992d91f511 100644
--- a/fs/ocfs2/localalloc.c
+++ b/fs/ocfs2/localalloc.c
@@ -476,7 +476,7 @@ out_mutex:
 
 out:
 	if (!status)
-		ocfs2_init_inode_steal_slot(osb);
+		ocfs2_init_steal_slots(osb);
 	mlog_exit(status);
 	return status;
 }
diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index 50fb26a6a5f5..d9cd4e373a53 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -212,7 +212,7 @@ static struct inode *ocfs2_get_init_inode(struct inode *dir, int mode)
 	} else
 		inode->i_gid = current_fsgid();
 	inode->i_mode = mode;
-	vfs_dq_init(inode);
+	dquot_initialize(inode);
 	return inode;
 }
 
@@ -244,6 +244,8 @@ static int ocfs2_mknod(struct inode *dir,
 		   (unsigned long)dev, dentry->d_name.len,
 		   dentry->d_name.name);
 
+	dquot_initialize(dir);
+
 	/* get our super block */
 	osb = OCFS2_SB(dir->i_sb);
 
@@ -348,13 +350,9 @@ static int ocfs2_mknod(struct inode *dir,
 		goto leave;
 	}
 
-	/* We don't use standard VFS wrapper because we don't want vfs_dq_init
-	 * to be called. */
-	if (sb_any_quota_active(osb->sb) &&
-	    osb->sb->dq_op->alloc_inode(inode, 1) == NO_QUOTA) {
-		status = -EDQUOT;
+	status = dquot_alloc_inode(inode);
+	if (status)
 		goto leave;
-	}
 	did_quota_inode = 1;
 
 	mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry,
@@ -431,7 +429,7 @@ static int ocfs2_mknod(struct inode *dir,
 	status = 0;
 leave:
 	if (status < 0 && did_quota_inode)
-		vfs_dq_free_inode(inode);
+		dquot_free_inode(inode);
 	if (handle)
 		ocfs2_commit_trans(osb, handle);
 
@@ -636,6 +634,8 @@ static int ocfs2_link(struct dentry *old_dentry,
 	if (S_ISDIR(inode->i_mode))
 		return -EPERM;
 
+	dquot_initialize(dir);
+
 	err = ocfs2_inode_lock_nested(dir, &parent_fe_bh, 1, OI_LS_PARENT);
 	if (err < 0) {
 		if (err != -ENOENT)
@@ -791,6 +791,8 @@ static int ocfs2_unlink(struct inode *dir,
 	mlog_entry("(0x%p, 0x%p, '%.*s')\n", dir, dentry,
 		   dentry->d_name.len, dentry->d_name.name);
 
+	dquot_initialize(dir);
+
 	BUG_ON(dentry->d_parent->d_inode != dir);
 
 	mlog(0, "ino = %llu\n", (unsigned long long)OCFS2_I(inode)->ip_blkno);
@@ -1051,6 +1053,9 @@ static int ocfs2_rename(struct inode *old_dir,
 		   old_dentry->d_name.len, old_dentry->d_name.name,
 		   new_dentry->d_name.len, new_dentry->d_name.name);
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	osb = OCFS2_SB(old_dir->i_sb);
 
 	if (new_inode) {
@@ -1599,6 +1604,8 @@ static int ocfs2_symlink(struct inode *dir,
 	mlog_entry("(0x%p, 0x%p, symname='%s' actual='%.*s')\n", dir,
 		   dentry, symname, dentry->d_name.len, dentry->d_name.name);
 
+	dquot_initialize(dir);
+
 	sb = dir->i_sb;
 	osb = OCFS2_SB(sb);
 
@@ -1688,13 +1695,9 @@ static int ocfs2_symlink(struct inode *dir,
 		goto bail;
 	}
 
-	/* We don't use standard VFS wrapper because we don't want vfs_dq_init
-	 * to be called. */
-	if (sb_any_quota_active(osb->sb) &&
-	    osb->sb->dq_op->alloc_inode(inode, 1) == NO_QUOTA) {
-		status = -EDQUOT;
+	status = dquot_alloc_inode(inode);
+	if (status)
 		goto bail;
-	}
 	did_quota_inode = 1;
 
 	mlog_entry("(0x%p, 0x%p, %d, '%.*s')\n", dir, dentry,
@@ -1716,11 +1719,10 @@ static int ocfs2_symlink(struct inode *dir,
 		u32 offset = 0;
 
 		inode->i_op = &ocfs2_symlink_inode_operations;
-		if (vfs_dq_alloc_space_nodirty(inode,
-		    ocfs2_clusters_to_bytes(osb->sb, 1))) {
-			status = -EDQUOT;
+		status = dquot_alloc_space_nodirty(inode,
+		    ocfs2_clusters_to_bytes(osb->sb, 1));
+		if (status)
 			goto bail;
-		}
 		did_quota = 1;
 		status = ocfs2_add_inode_data(osb, inode, &offset, 1, 0,
 					      new_fe_bh,
@@ -1788,10 +1790,10 @@ static int ocfs2_symlink(struct inode *dir,
 	d_instantiate(dentry, inode);
 bail:
 	if (status < 0 && did_quota)
-		vfs_dq_free_space_nodirty(inode,
+		dquot_free_space_nodirty(inode,
 					ocfs2_clusters_to_bytes(osb->sb, 1));
 	if (status < 0 && did_quota_inode)
-		vfs_dq_free_inode(inode);
+		dquot_free_inode(inode);
 	if (handle)
 		ocfs2_commit_trans(osb, handle);
 
@@ -2099,13 +2101,9 @@ int ocfs2_create_inode_in_orphan(struct inode *dir,
 		goto leave;
 	}
 
-	/* We don't use standard VFS wrapper because we don't want vfs_dq_init
-	 * to be called. */
-	if (sb_any_quota_active(osb->sb) &&
-	    osb->sb->dq_op->alloc_inode(inode, 1) == NO_QUOTA) {
-		status = -EDQUOT;
+	status = dquot_alloc_inode(inode);
+	if (status)
 		goto leave;
-	}
 	did_quota_inode = 1;
 
 	inode->i_nlink = 0;
@@ -2140,7 +2138,7 @@ int ocfs2_create_inode_in_orphan(struct inode *dir,
 	insert_inode_hash(inode);
 leave:
 	if (status < 0 && did_quota_inode)
-		vfs_dq_free_inode(inode);
+		dquot_free_inode(inode);
 	if (handle)
 		ocfs2_commit_trans(osb, handle);
 
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 740f448041e2..1238b491db90 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -42,6 +42,7 @@
 
 #include "ocfs2_fs.h"
 #include "ocfs2_lockid.h"
+#include "ocfs2_ioctl.h"
 
 /* For struct ocfs2_blockcheck_stats */
 #include "blockcheck.h"
@@ -159,7 +160,7 @@ struct ocfs2_lock_res {
 	int                      l_level;
 	unsigned int             l_ro_holders;
 	unsigned int             l_ex_holders;
-	union ocfs2_dlm_lksb     l_lksb;
+	struct ocfs2_dlm_lksb    l_lksb;
 
 	/* used from AST/BAST funcs. */
 	enum ocfs2_ast_action    l_action;
@@ -305,7 +306,9 @@ struct ocfs2_super
 	u32 s_next_generation;
 	unsigned long osb_flags;
 	s16 s_inode_steal_slot;
+	s16 s_meta_steal_slot;
 	atomic_t s_num_inodes_stolen;
+	atomic_t s_num_meta_stolen;
 
 	unsigned long s_mount_opt;
 	unsigned int s_atime_quantum;
@@ -760,33 +763,6 @@ static inline unsigned int ocfs2_megabytes_to_clusters(struct super_block *sb,
 	return megs << (20 - OCFS2_SB(sb)->s_clustersize_bits);
 }
 
-static inline void ocfs2_init_inode_steal_slot(struct ocfs2_super *osb)
-{
-	spin_lock(&osb->osb_lock);
-	osb->s_inode_steal_slot = OCFS2_INVALID_SLOT;
-	spin_unlock(&osb->osb_lock);
-	atomic_set(&osb->s_num_inodes_stolen, 0);
-}
-
-static inline void ocfs2_set_inode_steal_slot(struct ocfs2_super *osb,
-					      s16 slot)
-{
-	spin_lock(&osb->osb_lock);
-	osb->s_inode_steal_slot = slot;
-	spin_unlock(&osb->osb_lock);
-}
-
-static inline s16 ocfs2_get_inode_steal_slot(struct ocfs2_super *osb)
-{
-	s16 slot;
-
-	spin_lock(&osb->osb_lock);
-	slot = osb->s_inode_steal_slot;
-	spin_unlock(&osb->osb_lock);
-
-	return slot;
-}
-
 #define ocfs2_set_bit ext2_set_bit
 #define ocfs2_clear_bit ext2_clear_bit
 #define ocfs2_test_bit ext2_test_bit
diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h
index 7638a38c32bc..bb37218a7978 100644
--- a/fs/ocfs2/ocfs2_fs.h
+++ b/fs/ocfs2/ocfs2_fs.h
@@ -254,63 +254,6 @@
 						 * refcount tree */
 
 /*
- * ioctl commands
- */
-#define OCFS2_IOC_GETFLAGS	_IOR('f', 1, long)
-#define OCFS2_IOC_SETFLAGS	_IOW('f', 2, long)
-#define OCFS2_IOC32_GETFLAGS	_IOR('f', 1, int)
-#define OCFS2_IOC32_SETFLAGS	_IOW('f', 2, int)
-
-/*
- * Space reservation / allocation / free ioctls and argument structure
- * are designed to be compatible with XFS.
- *
- * ALLOCSP* and FREESP* are not and will never be supported, but are
- * included here for completeness.
- */
-struct ocfs2_space_resv {
-	__s16		l_type;
-	__s16		l_whence;
-	__s64		l_start;
-	__s64		l_len;		/* len == 0 means until end of file */
-	__s32		l_sysid;
-	__u32		l_pid;
-	__s32		l_pad[4];	/* reserve area			    */
-};
-
-#define OCFS2_IOC_ALLOCSP		_IOW ('X', 10, struct ocfs2_space_resv)
-#define OCFS2_IOC_FREESP		_IOW ('X', 11, struct ocfs2_space_resv)
-#define OCFS2_IOC_RESVSP		_IOW ('X', 40, struct ocfs2_space_resv)
-#define OCFS2_IOC_UNRESVSP	_IOW ('X', 41, struct ocfs2_space_resv)
-#define OCFS2_IOC_ALLOCSP64	_IOW ('X', 36, struct ocfs2_space_resv)
-#define OCFS2_IOC_FREESP64	_IOW ('X', 37, struct ocfs2_space_resv)
-#define OCFS2_IOC_RESVSP64	_IOW ('X', 42, struct ocfs2_space_resv)
-#define OCFS2_IOC_UNRESVSP64	_IOW ('X', 43, struct ocfs2_space_resv)
-
-/* Used to pass group descriptor data when online resize is done */
-struct ocfs2_new_group_input {
-	__u64 group;		/* Group descriptor's blkno. */
-	__u32 clusters;		/* Total number of clusters in this group */
-	__u32 frees;		/* Total free clusters in this group */
-	__u16 chain;		/* Chain for this group */
-	__u16 reserved1;
-	__u32 reserved2;
-};
-
-#define OCFS2_IOC_GROUP_EXTEND	_IOW('o', 1, int)
-#define OCFS2_IOC_GROUP_ADD	_IOW('o', 2,struct ocfs2_new_group_input)
-#define OCFS2_IOC_GROUP_ADD64	_IOW('o', 3,struct ocfs2_new_group_input)
-
-/* Used to pass 2 file names to reflink. */
-struct reflink_arguments {
-	__u64 old_path;
-	__u64 new_path;
-	__u64 preserve;
-};
-#define OCFS2_IOC_REFLINK	_IOW('o', 4, struct reflink_arguments)
-
-
-/*
  * Journal Flags (ocfs2_dinode.id1.journal1.i_flags)
  */
 #define OCFS2_JOURNAL_DIRTY_FL	(0x00000001)	/* Journal needs recovery */
diff --git a/fs/ocfs2/ocfs2_ioctl.h b/fs/ocfs2/ocfs2_ioctl.h
new file mode 100644
index 000000000000..2d3420af1a83
--- /dev/null
+++ b/fs/ocfs2/ocfs2_ioctl.h
@@ -0,0 +1,79 @@
+/* -*- mode: c; c-basic-offset: 8; -*-
+ * vim: noexpandtab sw=8 ts=8 sts=0:
+ *
+ * ocfs2_ioctl.h
+ *
+ * Defines OCFS2 ioctls.
+ *
+ * Copyright (C) 2010 Oracle.  All rights reserved.
+ *
+ * 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.
+ *
+ * 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 OCFS2_IOCTL_H
+#define OCFS2_IOCTL_H
+
+/*
+ * ioctl commands
+ */
+#define OCFS2_IOC_GETFLAGS	_IOR('f', 1, long)
+#define OCFS2_IOC_SETFLAGS	_IOW('f', 2, long)
+#define OCFS2_IOC32_GETFLAGS	_IOR('f', 1, int)
+#define OCFS2_IOC32_SETFLAGS	_IOW('f', 2, int)
+
+/*
+ * Space reservation / allocation / free ioctls and argument structure
+ * are designed to be compatible with XFS.
+ *
+ * ALLOCSP* and FREESP* are not and will never be supported, but are
+ * included here for completeness.
+ */
+struct ocfs2_space_resv {
+	__s16		l_type;
+	__s16		l_whence;
+	__s64		l_start;
+	__s64		l_len;		/* len == 0 means until end of file */
+	__s32		l_sysid;
+	__u32		l_pid;
+	__s32		l_pad[4];	/* reserve area			    */
+};
+
+#define OCFS2_IOC_ALLOCSP		_IOW ('X', 10, struct ocfs2_space_resv)
+#define OCFS2_IOC_FREESP		_IOW ('X', 11, struct ocfs2_space_resv)
+#define OCFS2_IOC_RESVSP		_IOW ('X', 40, struct ocfs2_space_resv)
+#define OCFS2_IOC_UNRESVSP	_IOW ('X', 41, struct ocfs2_space_resv)
+#define OCFS2_IOC_ALLOCSP64	_IOW ('X', 36, struct ocfs2_space_resv)
+#define OCFS2_IOC_FREESP64	_IOW ('X', 37, struct ocfs2_space_resv)
+#define OCFS2_IOC_RESVSP64	_IOW ('X', 42, struct ocfs2_space_resv)
+#define OCFS2_IOC_UNRESVSP64	_IOW ('X', 43, struct ocfs2_space_resv)
+
+/* Used to pass group descriptor data when online resize is done */
+struct ocfs2_new_group_input {
+	__u64 group;		/* Group descriptor's blkno. */
+	__u32 clusters;		/* Total number of clusters in this group */
+	__u32 frees;		/* Total free clusters in this group */
+	__u16 chain;		/* Chain for this group */
+	__u16 reserved1;
+	__u32 reserved2;
+};
+
+#define OCFS2_IOC_GROUP_EXTEND	_IOW('o', 1, int)
+#define OCFS2_IOC_GROUP_ADD	_IOW('o', 2,struct ocfs2_new_group_input)
+#define OCFS2_IOC_GROUP_ADD64	_IOW('o', 3,struct ocfs2_new_group_input)
+
+/* Used to pass 2 file names to reflink. */
+struct reflink_arguments {
+	__u64 old_path;
+	__u64 new_path;
+	__u64 preserve;
+};
+#define OCFS2_IOC_REFLINK	_IOW('o', 4, struct reflink_arguments)
+
+#endif /* OCFS2_IOCTL_H */
diff --git a/fs/ocfs2/ocfs2_lockingver.h b/fs/ocfs2/ocfs2_lockingver.h
index 82d5eeac0fff..2e45c8d2ea7e 100644
--- a/fs/ocfs2/ocfs2_lockingver.h
+++ b/fs/ocfs2/ocfs2_lockingver.h
@@ -23,6 +23,8 @@
 /*
  * The protocol version for ocfs2 cluster locking.  See dlmglue.c for
  * more details.
+ *
+ * 1.0 - Initial locking version from ocfs2 1.4.
  */
 #define OCFS2_LOCKING_PROTOCOL_MAJOR 1
 #define OCFS2_LOCKING_PROTOCOL_MINOR 0
diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c
index b437dc0c4cad..355f41d1d520 100644
--- a/fs/ocfs2/quota_global.c
+++ b/fs/ocfs2/quota_global.c
@@ -851,13 +851,6 @@ static void ocfs2_destroy_dquot(struct dquot *dquot)
 }
 
 const struct dquot_operations ocfs2_quota_operations = {
-	.initialize	= dquot_initialize,
-	.drop		= dquot_drop,
-	.alloc_space	= dquot_alloc_space,
-	.alloc_inode	= dquot_alloc_inode,
-	.free_space	= dquot_free_space,
-	.free_inode	= dquot_free_inode,
-	.transfer	= dquot_transfer,
 	.write_dquot	= ocfs2_write_dquot,
 	.acquire_dquot	= ocfs2_acquire_dquot,
 	.release_dquot	= ocfs2_release_dquot,
diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c
index 21f9e71223ca..a6467f3d262e 100644
--- a/fs/ocfs2/quota_local.c
+++ b/fs/ocfs2/quota_local.c
@@ -457,7 +457,7 @@ static int ocfs2_recover_local_quota_file(struct inode *lqinode,
 			break;
 		}
 		dchunk = (struct ocfs2_local_disk_chunk *)hbh->b_data;
-		for_each_bit(bit, rchunk->rc_bitmap, ol_chunk_entries(sb)) {
+		for_each_set_bit(bit, rchunk->rc_bitmap, ol_chunk_entries(sb)) {
 			qbh = NULL;
 			status = ocfs2_read_quota_block(lqinode,
 						ol_dqblk_block(sb, chunk, bit),
diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index 8ae65c9c020c..9e96921dffda 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -626,7 +626,7 @@ static int ocfs2_create_refcount_tree(struct inode *inode,
 	rb = (struct ocfs2_refcount_block *)new_bh->b_data;
 	memset(rb, 0, inode->i_sb->s_blocksize);
 	strcpy((void *)rb, OCFS2_REFCOUNT_BLOCK_SIGNATURE);
-	rb->rf_suballoc_slot = cpu_to_le16(osb->slot_num);
+	rb->rf_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
 	rb->rf_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 	rb->rf_fs_generation = cpu_to_le32(osb->fs_generation);
 	rb->rf_blkno = cpu_to_le64(first_blkno);
@@ -1330,7 +1330,7 @@ static int ocfs2_expand_inline_ref_root(handle_t *handle,
 	memcpy(new_bh->b_data, ref_root_bh->b_data, sb->s_blocksize);
 
 	new_rb = (struct ocfs2_refcount_block *)new_bh->b_data;
-	new_rb->rf_suballoc_slot = cpu_to_le16(OCFS2_SB(sb)->slot_num);
+	new_rb->rf_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
 	new_rb->rf_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 	new_rb->rf_blkno = cpu_to_le64(blkno);
 	new_rb->rf_cpos = cpu_to_le32(0);
@@ -1576,7 +1576,7 @@ static int ocfs2_new_leaf_refcount_block(handle_t *handle,
 	new_rb = (struct ocfs2_refcount_block *)new_bh->b_data;
 	memset(new_rb, 0, sb->s_blocksize);
 	strcpy((void *)new_rb, OCFS2_REFCOUNT_BLOCK_SIGNATURE);
-	new_rb->rf_suballoc_slot = cpu_to_le16(OCFS2_SB(sb)->slot_num);
+	new_rb->rf_suballoc_slot = cpu_to_le16(meta_ac->ac_alloc_slot);
 	new_rb->rf_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 	new_rb->rf_fs_generation = cpu_to_le32(OCFS2_SB(sb)->fs_generation);
 	new_rb->rf_blkno = cpu_to_le64(blkno);
@@ -4390,7 +4390,7 @@ static int ocfs2_vfs_reflink(struct dentry *old_dentry, struct inode *dir,
 	}
 
 	mutex_lock(&inode->i_mutex);
-	vfs_dq_init(dir);
+	dquot_initialize(dir);
 	error = ocfs2_reflink(old_dentry, dir, new_dentry, preserve);
 	mutex_unlock(&inode->i_mutex);
 	if (!error)
diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c
index 3038c92af493..7020e1253ffa 100644
--- a/fs/ocfs2/stack_o2cb.c
+++ b/fs/ocfs2/stack_o2cb.c
@@ -161,24 +161,23 @@ static int dlm_status_to_errno(enum dlm_status status)
 
 static void o2dlm_lock_ast_wrapper(void *astarg)
 {
-	BUG_ON(o2cb_stack.sp_proto == NULL);
+	struct ocfs2_dlm_lksb *lksb = astarg;
 
-	o2cb_stack.sp_proto->lp_lock_ast(astarg);
+	lksb->lksb_conn->cc_proto->lp_lock_ast(lksb);
 }
 
 static void o2dlm_blocking_ast_wrapper(void *astarg, int level)
 {
-	BUG_ON(o2cb_stack.sp_proto == NULL);
+	struct ocfs2_dlm_lksb *lksb = astarg;
 
-	o2cb_stack.sp_proto->lp_blocking_ast(astarg, level);
+	lksb->lksb_conn->cc_proto->lp_blocking_ast(lksb, level);
 }
 
 static void o2dlm_unlock_ast_wrapper(void *astarg, enum dlm_status status)
 {
+	struct ocfs2_dlm_lksb *lksb = astarg;
 	int error = dlm_status_to_errno(status);
 
-	BUG_ON(o2cb_stack.sp_proto == NULL);
-
 	/*
 	 * In o2dlm, you can get both the lock_ast() for the lock being
 	 * granted and the unlock_ast() for the CANCEL failing.  A
@@ -193,16 +192,15 @@ static void o2dlm_unlock_ast_wrapper(void *astarg, enum dlm_status status)
 	if (status == DLM_CANCELGRANT)
 		return;
 
-	o2cb_stack.sp_proto->lp_unlock_ast(astarg, error);
+	lksb->lksb_conn->cc_proto->lp_unlock_ast(lksb, error);
 }
 
 static int o2cb_dlm_lock(struct ocfs2_cluster_connection *conn,
 			 int mode,
-			 union ocfs2_dlm_lksb *lksb,
+			 struct ocfs2_dlm_lksb *lksb,
 			 u32 flags,
 			 void *name,
-			 unsigned int namelen,
-			 void *astarg)
+			 unsigned int namelen)
 {
 	enum dlm_status status;
 	int o2dlm_mode = mode_to_o2dlm(mode);
@@ -211,28 +209,27 @@ static int o2cb_dlm_lock(struct ocfs2_cluster_connection *conn,
 
 	status = dlmlock(conn->cc_lockspace, o2dlm_mode, &lksb->lksb_o2dlm,
 			 o2dlm_flags, name, namelen,
-			 o2dlm_lock_ast_wrapper, astarg,
+			 o2dlm_lock_ast_wrapper, lksb,
 			 o2dlm_blocking_ast_wrapper);
 	ret = dlm_status_to_errno(status);
 	return ret;
 }
 
 static int o2cb_dlm_unlock(struct ocfs2_cluster_connection *conn,
-			   union ocfs2_dlm_lksb *lksb,
-			   u32 flags,
-			   void *astarg)
+			   struct ocfs2_dlm_lksb *lksb,
+			   u32 flags)
 {
 	enum dlm_status status;
 	int o2dlm_flags = flags_to_o2dlm(flags);
 	int ret;
 
 	status = dlmunlock(conn->cc_lockspace, &lksb->lksb_o2dlm,
-			   o2dlm_flags, o2dlm_unlock_ast_wrapper, astarg);
+			   o2dlm_flags, o2dlm_unlock_ast_wrapper, lksb);
 	ret = dlm_status_to_errno(status);
 	return ret;
 }
 
-static int o2cb_dlm_lock_status(union ocfs2_dlm_lksb *lksb)
+static int o2cb_dlm_lock_status(struct ocfs2_dlm_lksb *lksb)
 {
 	return dlm_status_to_errno(lksb->lksb_o2dlm.status);
 }
@@ -242,17 +239,17 @@ static int o2cb_dlm_lock_status(union ocfs2_dlm_lksb *lksb)
  * contents, it will zero out the LVB.  Thus the caller can always trust
  * the contents.
  */
-static int o2cb_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb)
+static int o2cb_dlm_lvb_valid(struct ocfs2_dlm_lksb *lksb)
 {
 	return 1;
 }
 
-static void *o2cb_dlm_lvb(union ocfs2_dlm_lksb *lksb)
+static void *o2cb_dlm_lvb(struct ocfs2_dlm_lksb *lksb)
 {
 	return (void *)(lksb->lksb_o2dlm.lvb);
 }
 
-static void o2cb_dump_lksb(union ocfs2_dlm_lksb *lksb)
+static void o2cb_dump_lksb(struct ocfs2_dlm_lksb *lksb)
 {
 	dlm_print_one_lock(lksb->lksb_o2dlm.lockid);
 }
@@ -280,7 +277,7 @@ static int o2cb_cluster_connect(struct ocfs2_cluster_connection *conn)
 	struct dlm_protocol_version fs_version;
 
 	BUG_ON(conn == NULL);
-	BUG_ON(o2cb_stack.sp_proto == NULL);
+	BUG_ON(conn->cc_proto == NULL);
 
 	/* for now we only have one cluster/node, make sure we see it
 	 * in the heartbeat universe */
diff --git a/fs/ocfs2/stack_user.c b/fs/ocfs2/stack_user.c
index da78a2a334fd..5ae8812b2864 100644
--- a/fs/ocfs2/stack_user.c
+++ b/fs/ocfs2/stack_user.c
@@ -25,7 +25,6 @@
 #include <linux/reboot.h>
 #include <asm/uaccess.h>
 
-#include "ocfs2.h"  /* For struct ocfs2_lock_res */
 #include "stackglue.h"
 
 #include <linux/dlm_plock.h>
@@ -63,8 +62,8 @@
  * negotiated by the client.  The client negotiates based on the maximum
  * version advertised in /sys/fs/ocfs2/max_locking_protocol.  The major
  * number from the "SETV" message must match
- * ocfs2_user_plugin.sp_proto->lp_max_version.pv_major, and the minor number
- * must be less than or equal to ...->lp_max_version.pv_minor.
+ * ocfs2_user_plugin.sp_max_proto.pv_major, and the minor number
+ * must be less than or equal to ...sp_max_version.pv_minor.
  *
  * Once this information has been set, mounts will be allowed.  From this
  * point on, the "DOWN" message can be sent for node down notification.
@@ -401,7 +400,7 @@ static int ocfs2_control_do_setversion_msg(struct file *file,
 	char *ptr = NULL;
 	struct ocfs2_control_private *p = file->private_data;
 	struct ocfs2_protocol_version *max =
-		&ocfs2_user_plugin.sp_proto->lp_max_version;
+		&ocfs2_user_plugin.sp_max_proto;
 
 	if (ocfs2_control_get_handshake_state(file) !=
 	    OCFS2_CONTROL_HANDSHAKE_PROTOCOL)
@@ -664,18 +663,10 @@ static void ocfs2_control_exit(void)
 		       -rc);
 }
 
-static struct dlm_lksb *fsdlm_astarg_to_lksb(void *astarg)
-{
-	struct ocfs2_lock_res *res = astarg;
-	return &res->l_lksb.lksb_fsdlm;
-}
-
 static void fsdlm_lock_ast_wrapper(void *astarg)
 {
-	struct dlm_lksb *lksb = fsdlm_astarg_to_lksb(astarg);
-	int status = lksb->sb_status;
-
-	BUG_ON(ocfs2_user_plugin.sp_proto == NULL);
+	struct ocfs2_dlm_lksb *lksb = astarg;
+	int status = lksb->lksb_fsdlm.sb_status;
 
 	/*
 	 * For now we're punting on the issue of other non-standard errors
@@ -688,25 +679,24 @@ static void fsdlm_lock_ast_wrapper(void *astarg)
 	 */
 
 	if (status == -DLM_EUNLOCK || status == -DLM_ECANCEL)
-		ocfs2_user_plugin.sp_proto->lp_unlock_ast(astarg, 0);
+		lksb->lksb_conn->cc_proto->lp_unlock_ast(lksb, 0);
 	else
-		ocfs2_user_plugin.sp_proto->lp_lock_ast(astarg);
+		lksb->lksb_conn->cc_proto->lp_lock_ast(lksb);
 }
 
 static void fsdlm_blocking_ast_wrapper(void *astarg, int level)
 {
-	BUG_ON(ocfs2_user_plugin.sp_proto == NULL);
+	struct ocfs2_dlm_lksb *lksb = astarg;
 
-	ocfs2_user_plugin.sp_proto->lp_blocking_ast(astarg, level);
+	lksb->lksb_conn->cc_proto->lp_blocking_ast(lksb, level);
 }
 
 static int user_dlm_lock(struct ocfs2_cluster_connection *conn,
 			 int mode,
-			 union ocfs2_dlm_lksb *lksb,
+			 struct ocfs2_dlm_lksb *lksb,
 			 u32 flags,
 			 void *name,
-			 unsigned int namelen,
-			 void *astarg)
+			 unsigned int namelen)
 {
 	int ret;
 
@@ -716,36 +706,35 @@ static int user_dlm_lock(struct ocfs2_cluster_connection *conn,
 
 	ret = dlm_lock(conn->cc_lockspace, mode, &lksb->lksb_fsdlm,
 		       flags|DLM_LKF_NODLCKWT, name, namelen, 0,
-		       fsdlm_lock_ast_wrapper, astarg,
+		       fsdlm_lock_ast_wrapper, lksb,
 		       fsdlm_blocking_ast_wrapper);
 	return ret;
 }
 
 static int user_dlm_unlock(struct ocfs2_cluster_connection *conn,
-			   union ocfs2_dlm_lksb *lksb,
-			   u32 flags,
-			   void *astarg)
+			   struct ocfs2_dlm_lksb *lksb,
+			   u32 flags)
 {
 	int ret;
 
 	ret = dlm_unlock(conn->cc_lockspace, lksb->lksb_fsdlm.sb_lkid,
-			 flags, &lksb->lksb_fsdlm, astarg);
+			 flags, &lksb->lksb_fsdlm, lksb);
 	return ret;
 }
 
-static int user_dlm_lock_status(union ocfs2_dlm_lksb *lksb)
+static int user_dlm_lock_status(struct ocfs2_dlm_lksb *lksb)
 {
 	return lksb->lksb_fsdlm.sb_status;
 }
 
-static int user_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb)
+static int user_dlm_lvb_valid(struct ocfs2_dlm_lksb *lksb)
 {
 	int invalid = lksb->lksb_fsdlm.sb_flags & DLM_SBF_VALNOTVALID;
 
 	return !invalid;
 }
 
-static void *user_dlm_lvb(union ocfs2_dlm_lksb *lksb)
+static void *user_dlm_lvb(struct ocfs2_dlm_lksb *lksb)
 {
 	if (!lksb->lksb_fsdlm.sb_lvbptr)
 		lksb->lksb_fsdlm.sb_lvbptr = (char *)lksb +
@@ -753,7 +742,7 @@ static void *user_dlm_lvb(union ocfs2_dlm_lksb *lksb)
 	return (void *)(lksb->lksb_fsdlm.sb_lvbptr);
 }
 
-static void user_dlm_dump_lksb(union ocfs2_dlm_lksb *lksb)
+static void user_dlm_dump_lksb(struct ocfs2_dlm_lksb *lksb)
 {
 }
 
diff --git a/fs/ocfs2/stackglue.c b/fs/ocfs2/stackglue.c
index f3df0baa9a48..39abf89697ed 100644
--- a/fs/ocfs2/stackglue.c
+++ b/fs/ocfs2/stackglue.c
@@ -36,7 +36,7 @@
 #define OCFS2_STACK_PLUGIN_USER		"user"
 #define OCFS2_MAX_HB_CTL_PATH		256
 
-static struct ocfs2_locking_protocol *lproto;
+static struct ocfs2_protocol_version locking_max_version;
 static DEFINE_SPINLOCK(ocfs2_stack_lock);
 static LIST_HEAD(ocfs2_stack_list);
 static char cluster_stack_name[OCFS2_STACK_LABEL_LEN + 1];
@@ -176,7 +176,7 @@ int ocfs2_stack_glue_register(struct ocfs2_stack_plugin *plugin)
 	spin_lock(&ocfs2_stack_lock);
 	if (!ocfs2_stack_lookup(plugin->sp_name)) {
 		plugin->sp_count = 0;
-		plugin->sp_proto = lproto;
+		plugin->sp_max_proto = locking_max_version;
 		list_add(&plugin->sp_list, &ocfs2_stack_list);
 		printk(KERN_INFO "ocfs2: Registered cluster interface %s\n",
 		       plugin->sp_name);
@@ -213,77 +213,76 @@ void ocfs2_stack_glue_unregister(struct ocfs2_stack_plugin *plugin)
 }
 EXPORT_SYMBOL_GPL(ocfs2_stack_glue_unregister);
 
-void ocfs2_stack_glue_set_locking_protocol(struct ocfs2_locking_protocol *proto)
+void ocfs2_stack_glue_set_max_proto_version(struct ocfs2_protocol_version *max_proto)
 {
 	struct ocfs2_stack_plugin *p;
 
-	BUG_ON(proto == NULL);
-
 	spin_lock(&ocfs2_stack_lock);
-	BUG_ON(active_stack != NULL);
+	if (memcmp(max_proto, &locking_max_version,
+		   sizeof(struct ocfs2_protocol_version))) {
+		BUG_ON(locking_max_version.pv_major != 0);
 
-	lproto = proto;
-	list_for_each_entry(p, &ocfs2_stack_list, sp_list) {
-		p->sp_proto = lproto;
+		locking_max_version = *max_proto;
+		list_for_each_entry(p, &ocfs2_stack_list, sp_list) {
+			p->sp_max_proto = locking_max_version;
+		}
 	}
-
 	spin_unlock(&ocfs2_stack_lock);
 }
-EXPORT_SYMBOL_GPL(ocfs2_stack_glue_set_locking_protocol);
+EXPORT_SYMBOL_GPL(ocfs2_stack_glue_set_max_proto_version);
 
 
 /*
- * The ocfs2_dlm_lock() and ocfs2_dlm_unlock() functions take
- * "struct ocfs2_lock_res *astarg" instead of "void *astarg" because the
- * underlying stack plugins need to pilfer the lksb off of the lock_res.
- * If some other structure needs to be passed as an astarg, the plugins
- * will need to be given a different avenue to the lksb.
+ * The ocfs2_dlm_lock() and ocfs2_dlm_unlock() functions take no argument
+ * for the ast and bast functions.  They will pass the lksb to the ast
+ * and bast.  The caller can wrap the lksb with their own structure to
+ * get more information.
  */
 int ocfs2_dlm_lock(struct ocfs2_cluster_connection *conn,
 		   int mode,
-		   union ocfs2_dlm_lksb *lksb,
+		   struct ocfs2_dlm_lksb *lksb,
 		   u32 flags,
 		   void *name,
-		   unsigned int namelen,
-		   struct ocfs2_lock_res *astarg)
+		   unsigned int namelen)
 {
-	BUG_ON(lproto == NULL);
-
+	if (!lksb->lksb_conn)
+		lksb->lksb_conn = conn;
+	else
+		BUG_ON(lksb->lksb_conn != conn);
 	return active_stack->sp_ops->dlm_lock(conn, mode, lksb, flags,
-					      name, namelen, astarg);
+					      name, namelen);
 }
 EXPORT_SYMBOL_GPL(ocfs2_dlm_lock);
 
 int ocfs2_dlm_unlock(struct ocfs2_cluster_connection *conn,
-		     union ocfs2_dlm_lksb *lksb,
-		     u32 flags,
-		     struct ocfs2_lock_res *astarg)
+		     struct ocfs2_dlm_lksb *lksb,
+		     u32 flags)
 {
-	BUG_ON(lproto == NULL);
+	BUG_ON(lksb->lksb_conn == NULL);
 
-	return active_stack->sp_ops->dlm_unlock(conn, lksb, flags, astarg);
+	return active_stack->sp_ops->dlm_unlock(conn, lksb, flags);
 }
 EXPORT_SYMBOL_GPL(ocfs2_dlm_unlock);
 
-int ocfs2_dlm_lock_status(union ocfs2_dlm_lksb *lksb)
+int ocfs2_dlm_lock_status(struct ocfs2_dlm_lksb *lksb)
 {
 	return active_stack->sp_ops->lock_status(lksb);
 }
 EXPORT_SYMBOL_GPL(ocfs2_dlm_lock_status);
 
-int ocfs2_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb)
+int ocfs2_dlm_lvb_valid(struct ocfs2_dlm_lksb *lksb)
 {
 	return active_stack->sp_ops->lvb_valid(lksb);
 }
 EXPORT_SYMBOL_GPL(ocfs2_dlm_lvb_valid);
 
-void *ocfs2_dlm_lvb(union ocfs2_dlm_lksb *lksb)
+void *ocfs2_dlm_lvb(struct ocfs2_dlm_lksb *lksb)
 {
 	return active_stack->sp_ops->lock_lvb(lksb);
 }
 EXPORT_SYMBOL_GPL(ocfs2_dlm_lvb);
 
-void ocfs2_dlm_dump_lksb(union ocfs2_dlm_lksb *lksb)
+void ocfs2_dlm_dump_lksb(struct ocfs2_dlm_lksb *lksb)
 {
 	active_stack->sp_ops->dump_lksb(lksb);
 }
@@ -312,6 +311,7 @@ EXPORT_SYMBOL_GPL(ocfs2_plock);
 int ocfs2_cluster_connect(const char *stack_name,
 			  const char *group,
 			  int grouplen,
+			  struct ocfs2_locking_protocol *lproto,
 			  void (*recovery_handler)(int node_num,
 						   void *recovery_data),
 			  void *recovery_data,
@@ -329,6 +329,12 @@ int ocfs2_cluster_connect(const char *stack_name,
 		goto out;
 	}
 
+	if (memcmp(&lproto->lp_max_version, &locking_max_version,
+		   sizeof(struct ocfs2_protocol_version))) {
+		rc = -EINVAL;
+		goto out;
+	}
+
 	new_conn = kzalloc(sizeof(struct ocfs2_cluster_connection),
 			   GFP_KERNEL);
 	if (!new_conn) {
@@ -341,6 +347,7 @@ int ocfs2_cluster_connect(const char *stack_name,
 	new_conn->cc_recovery_handler = recovery_handler;
 	new_conn->cc_recovery_data = recovery_data;
 
+	new_conn->cc_proto = lproto;
 	/* Start the new connection at our maximum compatibility level */
 	new_conn->cc_version = lproto->lp_max_version;
 
@@ -366,6 +373,24 @@ out:
 }
 EXPORT_SYMBOL_GPL(ocfs2_cluster_connect);
 
+/* The caller will ensure all nodes have the same cluster stack */
+int ocfs2_cluster_connect_agnostic(const char *group,
+				   int grouplen,
+				   struct ocfs2_locking_protocol *lproto,
+				   void (*recovery_handler)(int node_num,
+							    void *recovery_data),
+				   void *recovery_data,
+				   struct ocfs2_cluster_connection **conn)
+{
+	char *stack_name = NULL;
+
+	if (cluster_stack_name[0])
+		stack_name = cluster_stack_name;
+	return ocfs2_cluster_connect(stack_name, group, grouplen, lproto,
+				     recovery_handler, recovery_data, conn);
+}
+EXPORT_SYMBOL_GPL(ocfs2_cluster_connect_agnostic);
+
 /* If hangup_pending is 0, the stack driver will be dropped */
 int ocfs2_cluster_disconnect(struct ocfs2_cluster_connection *conn,
 			     int hangup_pending)
@@ -453,10 +478,10 @@ static ssize_t ocfs2_max_locking_protocol_show(struct kobject *kobj,
 	ssize_t ret = 0;
 
 	spin_lock(&ocfs2_stack_lock);
-	if (lproto)
+	if (locking_max_version.pv_major)
 		ret = snprintf(buf, PAGE_SIZE, "%u.%u\n",
-			       lproto->lp_max_version.pv_major,
-			       lproto->lp_max_version.pv_minor);
+			       locking_max_version.pv_major,
+			       locking_max_version.pv_minor);
 	spin_unlock(&ocfs2_stack_lock);
 
 	return ret;
@@ -685,7 +710,10 @@ static int __init ocfs2_stack_glue_init(void)
 
 static void __exit ocfs2_stack_glue_exit(void)
 {
-	lproto = NULL;
+	memset(&locking_max_version, 0,
+	       sizeof(struct ocfs2_protocol_version));
+	locking_max_version.pv_major = 0;
+	locking_max_version.pv_minor = 0;
 	ocfs2_sysfs_exit();
 	if (ocfs2_table_header)
 		unregister_sysctl_table(ocfs2_table_header);
diff --git a/fs/ocfs2/stackglue.h b/fs/ocfs2/stackglue.h
index 03a44d60eac9..8ce7398ae1d2 100644
--- a/fs/ocfs2/stackglue.h
+++ b/fs/ocfs2/stackglue.h
@@ -56,17 +56,6 @@ struct ocfs2_protocol_version {
 };
 
 /*
- * The ocfs2_locking_protocol defines the handlers called on ocfs2's behalf.
- */
-struct ocfs2_locking_protocol {
-	struct ocfs2_protocol_version lp_max_version;
-	void (*lp_lock_ast)(void *astarg);
-	void (*lp_blocking_ast)(void *astarg, int level);
-	void (*lp_unlock_ast)(void *astarg, int error);
-};
-
-
-/*
  * The dlm_lockstatus struct includes lvb space, but the dlm_lksb struct only
  * has a pointer to separately allocated lvb space.  This struct exists only to
  * include in the lksb union to make space for a combined dlm_lksb and lvb.
@@ -81,12 +70,27 @@ struct fsdlm_lksb_plus_lvb {
  * size of the union is known.  Lock status structures are embedded in
  * ocfs2 inodes.
  */
-union ocfs2_dlm_lksb {
-	struct dlm_lockstatus lksb_o2dlm;
-	struct dlm_lksb lksb_fsdlm;
-	struct fsdlm_lksb_plus_lvb padding;
+struct ocfs2_cluster_connection;
+struct ocfs2_dlm_lksb {
+	 union {
+		 struct dlm_lockstatus lksb_o2dlm;
+		 struct dlm_lksb lksb_fsdlm;
+		 struct fsdlm_lksb_plus_lvb padding;
+	 };
+	 struct ocfs2_cluster_connection *lksb_conn;
+};
+
+/*
+ * The ocfs2_locking_protocol defines the handlers called on ocfs2's behalf.
+ */
+struct ocfs2_locking_protocol {
+	struct ocfs2_protocol_version lp_max_version;
+	void (*lp_lock_ast)(struct ocfs2_dlm_lksb *lksb);
+	void (*lp_blocking_ast)(struct ocfs2_dlm_lksb *lksb, int level);
+	void (*lp_unlock_ast)(struct ocfs2_dlm_lksb *lksb, int error);
 };
 
+
 /*
  * A cluster connection.  Mostly opaque to ocfs2, the connection holds
  * state for the underlying stack.  ocfs2 does use cc_version to determine
@@ -96,6 +100,7 @@ struct ocfs2_cluster_connection {
 	char cc_name[GROUP_NAME_MAX];
 	int cc_namelen;
 	struct ocfs2_protocol_version cc_version;
+	struct ocfs2_locking_protocol *cc_proto;
 	void (*cc_recovery_handler)(int node_num, void *recovery_data);
 	void *cc_recovery_data;
 	void *cc_lockspace;
@@ -155,27 +160,29 @@ struct ocfs2_stack_operations {
 	 *
 	 * ast and bast functions are not part of the call because the
 	 * stack will likely want to wrap ast and bast calls before passing
-	 * them to stack->sp_proto.
+	 * them to stack->sp_proto.  There is no astarg.  The lksb will
+	 * be passed back to the ast and bast functions.  The caller can
+	 * use this to find their object.
 	 */
 	int (*dlm_lock)(struct ocfs2_cluster_connection *conn,
 			int mode,
-			union ocfs2_dlm_lksb *lksb,
+			struct ocfs2_dlm_lksb *lksb,
 			u32 flags,
 			void *name,
-			unsigned int namelen,
-			void *astarg);
+			unsigned int namelen);
 
 	/*
 	 * Call the underlying dlm unlock function.  The ->dlm_unlock()
 	 * function should convert the flags as appropriate.
 	 *
 	 * The unlock ast is not passed, as the stack will want to wrap
-	 * it before calling stack->sp_proto->lp_unlock_ast().
+	 * it before calling stack->sp_proto->lp_unlock_ast().  There is
+	 * no astarg.  The lksb will be passed back to the unlock ast
+	 * function.  The caller can use this to find their object.
 	 */
 	int (*dlm_unlock)(struct ocfs2_cluster_connection *conn,
-			  union ocfs2_dlm_lksb *lksb,
-			  u32 flags,
-			  void *astarg);
+			  struct ocfs2_dlm_lksb *lksb,
+			  u32 flags);
 
 	/*
 	 * Return the status of the current lock status block.  The fs
@@ -183,17 +190,17 @@ struct ocfs2_stack_operations {
 	 * callback pulls out the stack-specific lksb, converts the status
 	 * to a proper errno, and returns it.
 	 */
-	int (*lock_status)(union ocfs2_dlm_lksb *lksb);
+	int (*lock_status)(struct ocfs2_dlm_lksb *lksb);
 
 	/*
 	 * Return non-zero if the LVB is valid.
 	 */
-	int (*lvb_valid)(union ocfs2_dlm_lksb *lksb);
+	int (*lvb_valid)(struct ocfs2_dlm_lksb *lksb);
 
 	/*
 	 * Pull the lvb pointer off of the stack-specific lksb.
 	 */
-	void *(*lock_lvb)(union ocfs2_dlm_lksb *lksb);
+	void *(*lock_lvb)(struct ocfs2_dlm_lksb *lksb);
 
 	/*
 	 * Cluster-aware posix locks
@@ -210,7 +217,7 @@ struct ocfs2_stack_operations {
 	 * This is an optoinal debugging hook.  If provided, the
 	 * stack can dump debugging information about this lock.
 	 */
-	void (*dump_lksb)(union ocfs2_dlm_lksb *lksb);
+	void (*dump_lksb)(struct ocfs2_dlm_lksb *lksb);
 };
 
 /*
@@ -226,7 +233,7 @@ struct ocfs2_stack_plugin {
 	/* These are managed by the stackglue code. */
 	struct list_head sp_list;
 	unsigned int sp_count;
-	struct ocfs2_locking_protocol *sp_proto;
+	struct ocfs2_protocol_version sp_max_proto;
 };
 
 
@@ -234,10 +241,22 @@ struct ocfs2_stack_plugin {
 int ocfs2_cluster_connect(const char *stack_name,
 			  const char *group,
 			  int grouplen,
+			  struct ocfs2_locking_protocol *lproto,
 			  void (*recovery_handler)(int node_num,
 						   void *recovery_data),
 			  void *recovery_data,
 			  struct ocfs2_cluster_connection **conn);
+/*
+ * Used by callers that don't store their stack name.  They must ensure
+ * all nodes have the same stack.
+ */
+int ocfs2_cluster_connect_agnostic(const char *group,
+				   int grouplen,
+				   struct ocfs2_locking_protocol *lproto,
+				   void (*recovery_handler)(int node_num,
+							    void *recovery_data),
+				   void *recovery_data,
+				   struct ocfs2_cluster_connection **conn);
 int ocfs2_cluster_disconnect(struct ocfs2_cluster_connection *conn,
 			     int hangup_pending);
 void ocfs2_cluster_hangup(const char *group, int grouplen);
@@ -246,26 +265,24 @@ int ocfs2_cluster_this_node(unsigned int *node);
 struct ocfs2_lock_res;
 int ocfs2_dlm_lock(struct ocfs2_cluster_connection *conn,
 		   int mode,
-		   union ocfs2_dlm_lksb *lksb,
+		   struct ocfs2_dlm_lksb *lksb,
 		   u32 flags,
 		   void *name,
-		   unsigned int namelen,
-		   struct ocfs2_lock_res *astarg);
+		   unsigned int namelen);
 int ocfs2_dlm_unlock(struct ocfs2_cluster_connection *conn,
-		     union ocfs2_dlm_lksb *lksb,
-		     u32 flags,
-		     struct ocfs2_lock_res *astarg);
+		     struct ocfs2_dlm_lksb *lksb,
+		     u32 flags);
 
-int ocfs2_dlm_lock_status(union ocfs2_dlm_lksb *lksb);
-int ocfs2_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb);
-void *ocfs2_dlm_lvb(union ocfs2_dlm_lksb *lksb);
-void ocfs2_dlm_dump_lksb(union ocfs2_dlm_lksb *lksb);
+int ocfs2_dlm_lock_status(struct ocfs2_dlm_lksb *lksb);
+int ocfs2_dlm_lvb_valid(struct ocfs2_dlm_lksb *lksb);
+void *ocfs2_dlm_lvb(struct ocfs2_dlm_lksb *lksb);
+void ocfs2_dlm_dump_lksb(struct ocfs2_dlm_lksb *lksb);
 
 int ocfs2_stack_supports_plocks(void);
 int ocfs2_plock(struct ocfs2_cluster_connection *conn, u64 ino,
 		struct file *file, int cmd, struct file_lock *fl);
 
-void ocfs2_stack_glue_set_locking_protocol(struct ocfs2_locking_protocol *proto);
+void ocfs2_stack_glue_set_max_proto_version(struct ocfs2_protocol_version *max_proto);
 
 
 /* Used by stack plugins */
diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c
index c30b644d9572..c3c60bc3e072 100644
--- a/fs/ocfs2/suballoc.c
+++ b/fs/ocfs2/suballoc.c
@@ -51,7 +51,7 @@
 #define ALLOC_NEW_GROUP			0x1
 #define ALLOC_GROUPS_FROM_GLOBAL	0x2
 
-#define OCFS2_MAX_INODES_TO_STEAL	1024
+#define OCFS2_MAX_TO_STEAL		1024
 
 static inline void ocfs2_debug_bg(struct ocfs2_group_desc *bg);
 static inline void ocfs2_debug_suballoc_inode(struct ocfs2_dinode *fe);
@@ -637,12 +637,113 @@ bail:
 	return status;
 }
 
+static void ocfs2_init_inode_steal_slot(struct ocfs2_super *osb)
+{
+	spin_lock(&osb->osb_lock);
+	osb->s_inode_steal_slot = OCFS2_INVALID_SLOT;
+	spin_unlock(&osb->osb_lock);
+	atomic_set(&osb->s_num_inodes_stolen, 0);
+}
+
+static void ocfs2_init_meta_steal_slot(struct ocfs2_super *osb)
+{
+	spin_lock(&osb->osb_lock);
+	osb->s_meta_steal_slot = OCFS2_INVALID_SLOT;
+	spin_unlock(&osb->osb_lock);
+	atomic_set(&osb->s_num_meta_stolen, 0);
+}
+
+void ocfs2_init_steal_slots(struct ocfs2_super *osb)
+{
+	ocfs2_init_inode_steal_slot(osb);
+	ocfs2_init_meta_steal_slot(osb);
+}
+
+static void __ocfs2_set_steal_slot(struct ocfs2_super *osb, int slot, int type)
+{
+	spin_lock(&osb->osb_lock);
+	if (type == INODE_ALLOC_SYSTEM_INODE)
+		osb->s_inode_steal_slot = slot;
+	else if (type == EXTENT_ALLOC_SYSTEM_INODE)
+		osb->s_meta_steal_slot = slot;
+	spin_unlock(&osb->osb_lock);
+}
+
+static int __ocfs2_get_steal_slot(struct ocfs2_super *osb, int type)
+{
+	int slot = OCFS2_INVALID_SLOT;
+
+	spin_lock(&osb->osb_lock);
+	if (type == INODE_ALLOC_SYSTEM_INODE)
+		slot = osb->s_inode_steal_slot;
+	else if (type == EXTENT_ALLOC_SYSTEM_INODE)
+		slot = osb->s_meta_steal_slot;
+	spin_unlock(&osb->osb_lock);
+
+	return slot;
+}
+
+static int ocfs2_get_inode_steal_slot(struct ocfs2_super *osb)
+{
+	return __ocfs2_get_steal_slot(osb, INODE_ALLOC_SYSTEM_INODE);
+}
+
+static int ocfs2_get_meta_steal_slot(struct ocfs2_super *osb)
+{
+	return __ocfs2_get_steal_slot(osb, EXTENT_ALLOC_SYSTEM_INODE);
+}
+
+static int ocfs2_steal_resource(struct ocfs2_super *osb,
+				struct ocfs2_alloc_context *ac,
+				int type)
+{
+	int i, status = -ENOSPC;
+	int slot = __ocfs2_get_steal_slot(osb, type);
+
+	/* Start to steal resource from the first slot after ours. */
+	if (slot == OCFS2_INVALID_SLOT)
+		slot = osb->slot_num + 1;
+
+	for (i = 0; i < osb->max_slots; i++, slot++) {
+		if (slot == osb->max_slots)
+			slot = 0;
+
+		if (slot == osb->slot_num)
+			continue;
+
+		status = ocfs2_reserve_suballoc_bits(osb, ac,
+						     type,
+						     (u32)slot, NULL,
+						     NOT_ALLOC_NEW_GROUP);
+		if (status >= 0) {
+			__ocfs2_set_steal_slot(osb, slot, type);
+			break;
+		}
+
+		ocfs2_free_ac_resource(ac);
+	}
+
+	return status;
+}
+
+static int ocfs2_steal_inode(struct ocfs2_super *osb,
+			     struct ocfs2_alloc_context *ac)
+{
+	return ocfs2_steal_resource(osb, ac, INODE_ALLOC_SYSTEM_INODE);
+}
+
+static int ocfs2_steal_meta(struct ocfs2_super *osb,
+			    struct ocfs2_alloc_context *ac)
+{
+	return ocfs2_steal_resource(osb, ac, EXTENT_ALLOC_SYSTEM_INODE);
+}
+
 int ocfs2_reserve_new_metadata_blocks(struct ocfs2_super *osb,
 				      int blocks,
 				      struct ocfs2_alloc_context **ac)
 {
 	int status;
-	u32 slot;
+	int slot = ocfs2_get_meta_steal_slot(osb);
 
 	*ac = kzalloc(sizeof(struct ocfs2_alloc_context), GFP_KERNEL);
 	if (!(*ac)) {
@@ -653,12 +754,34 @@ int ocfs2_reserve_new_metadata_blocks(struct ocfs2_super *osb,
 
 	(*ac)->ac_bits_wanted = blocks;
 	(*ac)->ac_which = OCFS2_AC_USE_META;
-	slot = osb->slot_num;
 	(*ac)->ac_group_search = ocfs2_block_group_search;
 
+	if (slot != OCFS2_INVALID_SLOT &&
+		atomic_read(&osb->s_num_meta_stolen) < OCFS2_MAX_TO_STEAL)
+		goto extent_steal;
+
+	atomic_set(&osb->s_num_meta_stolen, 0);
 	status = ocfs2_reserve_suballoc_bits(osb, (*ac),
 					     EXTENT_ALLOC_SYSTEM_INODE,
-					     slot, NULL, ALLOC_NEW_GROUP);
+					     (u32)osb->slot_num, NULL,
+					     ALLOC_NEW_GROUP);
+
+
+	if (status >= 0) {
+		status = 0;
+		if (slot != OCFS2_INVALID_SLOT)
+			ocfs2_init_meta_steal_slot(osb);
+		goto bail;
+	} else if (status < 0 && status != -ENOSPC) {
+		mlog_errno(status);
+		goto bail;
+	}
+
+	ocfs2_free_ac_resource(*ac);
+
+extent_steal:
+	status = ocfs2_steal_meta(osb, *ac);
+	atomic_inc(&osb->s_num_meta_stolen);
 	if (status < 0) {
 		if (status != -ENOSPC)
 			mlog_errno(status);
@@ -685,43 +808,11 @@ int ocfs2_reserve_new_metadata(struct ocfs2_super *osb,
 					ac);
 }
 
-static int ocfs2_steal_inode_from_other_nodes(struct ocfs2_super *osb,
-					      struct ocfs2_alloc_context *ac)
-{
-	int i, status = -ENOSPC;
-	s16 slot = ocfs2_get_inode_steal_slot(osb);
-
-	/* Start to steal inodes from the first slot after ours. */
-	if (slot == OCFS2_INVALID_SLOT)
-		slot = osb->slot_num + 1;
-
-	for (i = 0; i < osb->max_slots; i++, slot++) {
-		if (slot == osb->max_slots)
-			slot = 0;
-
-		if (slot == osb->slot_num)
-			continue;
-
-		status = ocfs2_reserve_suballoc_bits(osb, ac,
-						     INODE_ALLOC_SYSTEM_INODE,
-						     slot, NULL,
-						     NOT_ALLOC_NEW_GROUP);
-		if (status >= 0) {
-			ocfs2_set_inode_steal_slot(osb, slot);
-			break;
-		}
-
-		ocfs2_free_ac_resource(ac);
-	}
-
-	return status;
-}
-
 int ocfs2_reserve_new_inode(struct ocfs2_super *osb,
 			    struct ocfs2_alloc_context **ac)
 {
 	int status;
-	s16 slot = ocfs2_get_inode_steal_slot(osb);
+	int slot = ocfs2_get_inode_steal_slot(osb);
 	u64 alloc_group;
 
 	*ac = kzalloc(sizeof(struct ocfs2_alloc_context), GFP_KERNEL);
@@ -754,14 +845,14 @@ int ocfs2_reserve_new_inode(struct ocfs2_super *osb,
 	 * need to check our slots to see whether there is some space for us.
 	 */
 	if (slot != OCFS2_INVALID_SLOT &&
-	    atomic_read(&osb->s_num_inodes_stolen) < OCFS2_MAX_INODES_TO_STEAL)
+	    atomic_read(&osb->s_num_inodes_stolen) < OCFS2_MAX_TO_STEAL)
 		goto inode_steal;
 
 	atomic_set(&osb->s_num_inodes_stolen, 0);
 	alloc_group = osb->osb_inode_alloc_group;
 	status = ocfs2_reserve_suballoc_bits(osb, *ac,
 					     INODE_ALLOC_SYSTEM_INODE,
-					     osb->slot_num,
+					     (u32)osb->slot_num,
 					     &alloc_group,
 					     ALLOC_NEW_GROUP |
 					     ALLOC_GROUPS_FROM_GLOBAL);
@@ -789,7 +880,7 @@ int ocfs2_reserve_new_inode(struct ocfs2_super *osb,
 	ocfs2_free_ac_resource(*ac);
 
 inode_steal:
-	status = ocfs2_steal_inode_from_other_nodes(osb, *ac);
+	status = ocfs2_steal_inode(osb, *ac);
 	atomic_inc(&osb->s_num_inodes_stolen);
 	if (status < 0) {
 		if (status != -ENOSPC)
diff --git a/fs/ocfs2/suballoc.h b/fs/ocfs2/suballoc.h
index 8c9a78a43164..fa60723c43e8 100644
--- a/fs/ocfs2/suballoc.h
+++ b/fs/ocfs2/suballoc.h
@@ -56,6 +56,7 @@ struct ocfs2_alloc_context {
 				 is the same as ~0 - unlimited */
 };
 
+void ocfs2_init_steal_slots(struct ocfs2_super *osb);
 void ocfs2_free_alloc_context(struct ocfs2_alloc_context *ac);
 static inline int ocfs2_alloc_context_bits_left(struct ocfs2_alloc_context *ac)
 {
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 755cd49a5ef3..dee03197a494 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -69,6 +69,7 @@
 #include "xattr.h"
 #include "quota.h"
 #include "refcounttree.h"
+#include "suballoc.h"
 
 #include "buffer_head_io.h"
 
@@ -301,9 +302,12 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len)
 
 	spin_lock(&osb->osb_lock);
 	out += snprintf(buf + out, len - out,
-			"%10s => Slot: %d  NumStolen: %d\n", "Steal",
+			"%10s => InodeSlot: %d  StolenInodes: %d, "
+			"MetaSlot: %d  StolenMeta: %d\n", "Steal",
 			osb->s_inode_steal_slot,
-			atomic_read(&osb->s_num_inodes_stolen));
+			atomic_read(&osb->s_num_inodes_stolen),
+			osb->s_meta_steal_slot,
+			atomic_read(&osb->s_num_meta_stolen));
 	spin_unlock(&osb->osb_lock);
 
 	out += snprintf(buf + out, len - out, "OrphanScan => ");
@@ -1997,7 +2001,7 @@ static int ocfs2_initialize_super(struct super_block *sb,
 	osb->blocked_lock_count = 0;
 	spin_lock_init(&osb->osb_lock);
 	spin_lock_init(&osb->osb_xattr_lock);
-	ocfs2_init_inode_steal_slot(osb);
+	ocfs2_init_steal_slots(osb);
 
 	atomic_set(&osb->alloc_stats.moves, 0);
 	atomic_set(&osb->alloc_stats.local_data, 0);
diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c
index 8fc6fb071c6d..d1b0d386f6d1 100644
--- a/fs/ocfs2/xattr.c
+++ b/fs/ocfs2/xattr.c
@@ -116,10 +116,11 @@ static struct xattr_handler *ocfs2_xattr_handler_map[OCFS2_XATTR_MAX] = {
 };
 
 struct ocfs2_xattr_info {
-	int name_index;
-	const char *name;
-	const void *value;
-	size_t value_len;
+	int		xi_name_index;
+	const char	*xi_name;
+	int		xi_name_len;
+	const void	*xi_value;
+	size_t		xi_value_len;
 };
 
 struct ocfs2_xattr_search {
@@ -137,6 +138,115 @@ struct ocfs2_xattr_search {
 	int not_found;
 };
 
+/* Operations on struct ocfs2_xa_entry */
+struct ocfs2_xa_loc;
+struct ocfs2_xa_loc_operations {
+	/*
+	 * Journal functions
+	 */
+	int (*xlo_journal_access)(handle_t *handle, struct ocfs2_xa_loc *loc,
+				  int type);
+	void (*xlo_journal_dirty)(handle_t *handle, struct ocfs2_xa_loc *loc);
+
+	/*
+	 * Return a pointer to the appropriate buffer in loc->xl_storage
+	 * at the given offset from loc->xl_header.
+	 */
+	void *(*xlo_offset_pointer)(struct ocfs2_xa_loc *loc, int offset);
+
+	/* Can we reuse the existing entry for the new value? */
+	int (*xlo_can_reuse)(struct ocfs2_xa_loc *loc,
+			     struct ocfs2_xattr_info *xi);
+
+	/* How much space is needed for the new value? */
+	int (*xlo_check_space)(struct ocfs2_xa_loc *loc,
+			       struct ocfs2_xattr_info *xi);
+
+	/*
+	 * Return the offset of the first name+value pair.  This is
+	 * the start of our downward-filling free space.
+	 */
+	int (*xlo_get_free_start)(struct ocfs2_xa_loc *loc);
+
+	/*
+	 * Remove the name+value at this location.  Do whatever is
+	 * appropriate with the remaining name+value pairs.
+	 */
+	void (*xlo_wipe_namevalue)(struct ocfs2_xa_loc *loc);
+
+	/* Fill xl_entry with a new entry */
+	void (*xlo_add_entry)(struct ocfs2_xa_loc *loc, u32 name_hash);
+
+	/* Add name+value storage to an entry */
+	void (*xlo_add_namevalue)(struct ocfs2_xa_loc *loc, int size);
+
+	/*
+	 * Initialize the value buf's access and bh fields for this entry.
+	 * ocfs2_xa_fill_value_buf() will handle the xv pointer.
+	 */
+	void (*xlo_fill_value_buf)(struct ocfs2_xa_loc *loc,
+				   struct ocfs2_xattr_value_buf *vb);
+};
+
+/*
+ * Describes an xattr entry location.  This is a memory structure
+ * tracking the on-disk structure.
+ */
+struct ocfs2_xa_loc {
+	/* This xattr belongs to this inode */
+	struct inode *xl_inode;
+
+	/* The ocfs2_xattr_header inside the on-disk storage. Not NULL. */
+	struct ocfs2_xattr_header *xl_header;
+
+	/* Bytes from xl_header to the end of the storage */
+	int xl_size;
+
+	/*
+	 * The ocfs2_xattr_entry this location describes.  If this is
+	 * NULL, this location describes the on-disk structure where it
+	 * would have been.
+	 */
+	struct ocfs2_xattr_entry *xl_entry;
+
+	/*
+	 * Internal housekeeping
+	 */
+
+	/* Buffer(s) containing this entry */
+	void *xl_storage;
+
+	/* Operations on the storage backing this location */
+	const struct ocfs2_xa_loc_operations *xl_ops;
+};
+
+/*
+ * Convenience functions to calculate how much space is needed for a
+ * given name+value pair
+ */
+static int namevalue_size(int name_len, uint64_t value_len)
+{
+	if (value_len > OCFS2_XATTR_INLINE_SIZE)
+		return OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
+	else
+		return OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_SIZE(value_len);
+}
+
+static int namevalue_size_xi(struct ocfs2_xattr_info *xi)
+{
+	return namevalue_size(xi->xi_name_len, xi->xi_value_len);
+}
+
+static int namevalue_size_xe(struct ocfs2_xattr_entry *xe)
+{
+	u64 value_len = le64_to_cpu(xe->xe_value_size);
+
+	BUG_ON((value_len > OCFS2_XATTR_INLINE_SIZE) &&
+	       ocfs2_xattr_is_local(xe));
+	return namevalue_size(xe->xe_name_len, value_len);
+}
+
+
 static int ocfs2_xattr_bucket_get_name_value(struct super_block *sb,
 					     struct ocfs2_xattr_header *xh,
 					     int index,
@@ -212,14 +322,6 @@ static inline u16 ocfs2_blocks_per_xattr_bucket(struct super_block *sb)
 	return OCFS2_XATTR_BUCKET_SIZE / (1 << sb->s_blocksize_bits);
 }
 
-static inline u16 ocfs2_xattr_max_xe_in_bucket(struct super_block *sb)
-{
-	u16 len = sb->s_blocksize -
-		 offsetof(struct ocfs2_xattr_header, xh_entries);
-
-	return len / sizeof(struct ocfs2_xattr_entry);
-}
-
 #define bucket_blkno(_b) ((_b)->bu_bhs[0]->b_blocknr)
 #define bucket_block(_b, _n) ((_b)->bu_bhs[(_n)]->b_data)
 #define bucket_xh(_b) ((struct ocfs2_xattr_header *)bucket_block((_b), 0))
@@ -463,35 +565,22 @@ static u32 ocfs2_xattr_name_hash(struct inode *inode,
 	return hash;
 }
 
-/*
- * ocfs2_xattr_hash_entry()
- *
- * Compute the hash of an extended attribute.
- */
-static void ocfs2_xattr_hash_entry(struct inode *inode,
-				   struct ocfs2_xattr_header *header,
-				   struct ocfs2_xattr_entry *entry)
+static int ocfs2_xattr_entry_real_size(int name_len, size_t value_len)
 {
-	u32 hash = 0;
-	char *name = (char *)header + le16_to_cpu(entry->xe_name_offset);
-
-	hash = ocfs2_xattr_name_hash(inode, name, entry->xe_name_len);
-	entry->xe_name_hash = cpu_to_le32(hash);
-
-	return;
+	return namevalue_size(name_len, value_len) +
+		sizeof(struct ocfs2_xattr_entry);
 }
 
-static int ocfs2_xattr_entry_real_size(int name_len, size_t value_len)
+static int ocfs2_xi_entry_usage(struct ocfs2_xattr_info *xi)
 {
-	int size = 0;
-
-	if (value_len <= OCFS2_XATTR_INLINE_SIZE)
-		size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_SIZE(value_len);
-	else
-		size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
-	size += sizeof(struct ocfs2_xattr_entry);
+	return namevalue_size_xi(xi) +
+		sizeof(struct ocfs2_xattr_entry);
+}
 
-	return size;
+static int ocfs2_xe_entry_usage(struct ocfs2_xattr_entry *xe)
+{
+	return namevalue_size_xe(xe) +
+		sizeof(struct ocfs2_xattr_entry);
 }
 
 int ocfs2_calc_security_init(struct inode *dir,
@@ -1308,452 +1397,897 @@ out:
 	return ret;
 }
 
-static int ocfs2_xattr_cleanup(struct inode *inode,
-			       handle_t *handle,
-			       struct ocfs2_xattr_info *xi,
-			       struct ocfs2_xattr_search *xs,
-			       struct ocfs2_xattr_value_buf *vb,
-			       size_t offs)
+static int ocfs2_xa_check_space_helper(int needed_space, int free_start,
+				       int num_entries)
 {
-	int ret = 0;
-	size_t name_len = strlen(xi->name);
-	void *val = xs->base + offs;
-	size_t size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
+	int free_space;
 
-	ret = vb->vb_access(handle, INODE_CACHE(inode), vb->vb_bh,
-			    OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
-	/* Decrease xattr count */
-	le16_add_cpu(&xs->header->xh_count, -1);
-	/* Remove the xattr entry and tree root which has already be set*/
-	memset((void *)xs->here, 0, sizeof(struct ocfs2_xattr_entry));
-	memset(val, 0, size);
+	if (!needed_space)
+		return 0;
 
-	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
-	if (ret < 0)
-		mlog_errno(ret);
-out:
-	return ret;
+	free_space = free_start -
+		sizeof(struct ocfs2_xattr_header) -
+		(num_entries * sizeof(struct ocfs2_xattr_entry)) -
+		OCFS2_XATTR_HEADER_GAP;
+	if (free_space < 0)
+		return -EIO;
+	if (free_space < needed_space)
+		return -ENOSPC;
+
+	return 0;
 }
 
-static int ocfs2_xattr_update_entry(struct inode *inode,
-				    handle_t *handle,
-				    struct ocfs2_xattr_info *xi,
-				    struct ocfs2_xattr_search *xs,
-				    struct ocfs2_xattr_value_buf *vb,
-				    size_t offs)
+static int ocfs2_xa_journal_access(handle_t *handle, struct ocfs2_xa_loc *loc,
+				   int type)
 {
-	int ret;
+	return loc->xl_ops->xlo_journal_access(handle, loc, type);
+}
 
-	ret = vb->vb_access(handle, INODE_CACHE(inode), vb->vb_bh,
-			    OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
+static void ocfs2_xa_journal_dirty(handle_t *handle, struct ocfs2_xa_loc *loc)
+{
+	loc->xl_ops->xlo_journal_dirty(handle, loc);
+}
 
-	xs->here->xe_name_offset = cpu_to_le16(offs);
-	xs->here->xe_value_size = cpu_to_le64(xi->value_len);
-	if (xi->value_len <= OCFS2_XATTR_INLINE_SIZE)
-		ocfs2_xattr_set_local(xs->here, 1);
-	else
-		ocfs2_xattr_set_local(xs->here, 0);
-	ocfs2_xattr_hash_entry(inode, xs->header, xs->here);
+/* Give a pointer into the storage for the given offset */
+static void *ocfs2_xa_offset_pointer(struct ocfs2_xa_loc *loc, int offset)
+{
+	BUG_ON(offset >= loc->xl_size);
+	return loc->xl_ops->xlo_offset_pointer(loc, offset);
+}
 
-	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
-	if (ret < 0)
-		mlog_errno(ret);
-out:
-	return ret;
+/*
+ * Wipe the name+value pair and allow the storage to reclaim it.  This
+ * must be followed by either removal of the entry or a call to
+ * ocfs2_xa_add_namevalue().
+ */
+static void ocfs2_xa_wipe_namevalue(struct ocfs2_xa_loc *loc)
+{
+	loc->xl_ops->xlo_wipe_namevalue(loc);
 }
 
 /*
- * ocfs2_xattr_set_value_outside()
- *
- * Set large size value in B tree.
+ * Find lowest offset to a name+value pair.  This is the start of our
+ * downward-growing free space.
  */
-static int ocfs2_xattr_set_value_outside(struct inode *inode,
-					 struct ocfs2_xattr_info *xi,
-					 struct ocfs2_xattr_search *xs,
-					 struct ocfs2_xattr_set_ctxt *ctxt,
-					 struct ocfs2_xattr_value_buf *vb,
-					 size_t offs)
+static int ocfs2_xa_get_free_start(struct ocfs2_xa_loc *loc)
 {
-	size_t name_len = strlen(xi->name);
-	void *val = xs->base + offs;
-	struct ocfs2_xattr_value_root *xv = NULL;
-	size_t size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
-	int ret = 0;
+	return loc->xl_ops->xlo_get_free_start(loc);
+}
 
-	memset(val, 0, size);
-	memcpy(val, xi->name, name_len);
-	xv = (struct ocfs2_xattr_value_root *)
-		(val + OCFS2_XATTR_SIZE(name_len));
-	xv->xr_clusters = 0;
-	xv->xr_last_eb_blk = 0;
-	xv->xr_list.l_tree_depth = 0;
-	xv->xr_list.l_count = cpu_to_le16(1);
-	xv->xr_list.l_next_free_rec = 0;
-	vb->vb_xv = xv;
-
-	ret = ocfs2_xattr_value_truncate(inode, vb, xi->value_len, ctxt);
-	if (ret < 0) {
-		mlog_errno(ret);
-		return ret;
+/* Can we reuse loc->xl_entry for xi? */
+static int ocfs2_xa_can_reuse_entry(struct ocfs2_xa_loc *loc,
+				    struct ocfs2_xattr_info *xi)
+{
+	return loc->xl_ops->xlo_can_reuse(loc, xi);
+}
+
+/* How much free space is needed to set the new value */
+static int ocfs2_xa_check_space(struct ocfs2_xa_loc *loc,
+				struct ocfs2_xattr_info *xi)
+{
+	return loc->xl_ops->xlo_check_space(loc, xi);
+}
+
+static void ocfs2_xa_add_entry(struct ocfs2_xa_loc *loc, u32 name_hash)
+{
+	loc->xl_ops->xlo_add_entry(loc, name_hash);
+	loc->xl_entry->xe_name_hash = cpu_to_le32(name_hash);
+	/*
+	 * We can't leave the new entry's xe_name_offset at zero or
+	 * add_namevalue() will go nuts.  We set it to the size of our
+	 * storage so that it can never be less than any other entry.
+	 */
+	loc->xl_entry->xe_name_offset = cpu_to_le16(loc->xl_size);
+}
+
+static void ocfs2_xa_add_namevalue(struct ocfs2_xa_loc *loc,
+				   struct ocfs2_xattr_info *xi)
+{
+	int size = namevalue_size_xi(xi);
+	int nameval_offset;
+	char *nameval_buf;
+
+	loc->xl_ops->xlo_add_namevalue(loc, size);
+	loc->xl_entry->xe_value_size = cpu_to_le64(xi->xi_value_len);
+	loc->xl_entry->xe_name_len = xi->xi_name_len;
+	ocfs2_xattr_set_type(loc->xl_entry, xi->xi_name_index);
+	ocfs2_xattr_set_local(loc->xl_entry,
+			      xi->xi_value_len <= OCFS2_XATTR_INLINE_SIZE);
+
+	nameval_offset = le16_to_cpu(loc->xl_entry->xe_name_offset);
+	nameval_buf = ocfs2_xa_offset_pointer(loc, nameval_offset);
+	memset(nameval_buf, 0, size);
+	memcpy(nameval_buf, xi->xi_name, xi->xi_name_len);
+}
+
+static void ocfs2_xa_fill_value_buf(struct ocfs2_xa_loc *loc,
+				    struct ocfs2_xattr_value_buf *vb)
+{
+	int nameval_offset = le16_to_cpu(loc->xl_entry->xe_name_offset);
+	int name_size = OCFS2_XATTR_SIZE(loc->xl_entry->xe_name_len);
+
+	/* Value bufs are for value trees */
+	BUG_ON(ocfs2_xattr_is_local(loc->xl_entry));
+	BUG_ON(namevalue_size_xe(loc->xl_entry) !=
+	       (name_size + OCFS2_XATTR_ROOT_SIZE));
+
+	loc->xl_ops->xlo_fill_value_buf(loc, vb);
+	vb->vb_xv =
+		(struct ocfs2_xattr_value_root *)ocfs2_xa_offset_pointer(loc,
+							nameval_offset +
+							name_size);
+}
+
+static int ocfs2_xa_block_journal_access(handle_t *handle,
+					 struct ocfs2_xa_loc *loc, int type)
+{
+	struct buffer_head *bh = loc->xl_storage;
+	ocfs2_journal_access_func access;
+
+	if (loc->xl_size == (bh->b_size -
+			     offsetof(struct ocfs2_xattr_block,
+				      xb_attrs.xb_header)))
+		access = ocfs2_journal_access_xb;
+	else
+		access = ocfs2_journal_access_di;
+	return access(handle, INODE_CACHE(loc->xl_inode), bh, type);
+}
+
+static void ocfs2_xa_block_journal_dirty(handle_t *handle,
+					 struct ocfs2_xa_loc *loc)
+{
+	struct buffer_head *bh = loc->xl_storage;
+
+	ocfs2_journal_dirty(handle, bh);
+}
+
+static void *ocfs2_xa_block_offset_pointer(struct ocfs2_xa_loc *loc,
+					   int offset)
+{
+	return (char *)loc->xl_header + offset;
+}
+
+static int ocfs2_xa_block_can_reuse(struct ocfs2_xa_loc *loc,
+				    struct ocfs2_xattr_info *xi)
+{
+	/*
+	 * Block storage is strict.  If the sizes aren't exact, we will
+	 * remove the old one and reinsert the new.
+	 */
+	return namevalue_size_xe(loc->xl_entry) ==
+		namevalue_size_xi(xi);
+}
+
+static int ocfs2_xa_block_get_free_start(struct ocfs2_xa_loc *loc)
+{
+	struct ocfs2_xattr_header *xh = loc->xl_header;
+	int i, count = le16_to_cpu(xh->xh_count);
+	int offset, free_start = loc->xl_size;
+
+	for (i = 0; i < count; i++) {
+		offset = le16_to_cpu(xh->xh_entries[i].xe_name_offset);
+		if (offset < free_start)
+			free_start = offset;
 	}
-	ret = ocfs2_xattr_update_entry(inode, ctxt->handle, xi, xs, vb, offs);
-	if (ret < 0) {
-		mlog_errno(ret);
-		return ret;
+
+	return free_start;
+}
+
+static int ocfs2_xa_block_check_space(struct ocfs2_xa_loc *loc,
+				      struct ocfs2_xattr_info *xi)
+{
+	int count = le16_to_cpu(loc->xl_header->xh_count);
+	int free_start = ocfs2_xa_get_free_start(loc);
+	int needed_space = ocfs2_xi_entry_usage(xi);
+
+	/*
+	 * Block storage will reclaim the original entry before inserting
+	 * the new value, so we only need the difference.  If the new
+	 * entry is smaller than the old one, we don't need anything.
+	 */
+	if (loc->xl_entry) {
+		/* Don't need space if we're reusing! */
+		if (ocfs2_xa_can_reuse_entry(loc, xi))
+			needed_space = 0;
+		else
+			needed_space -= ocfs2_xe_entry_usage(loc->xl_entry);
 	}
-	ret = __ocfs2_xattr_set_value_outside(inode, ctxt->handle, vb,
-					      xi->value, xi->value_len);
-	if (ret < 0)
-		mlog_errno(ret);
+	if (needed_space < 0)
+		needed_space = 0;
+	return ocfs2_xa_check_space_helper(needed_space, free_start, count);
+}
 
-	return ret;
+/*
+ * Block storage for xattrs keeps the name+value pairs compacted.  When
+ * we remove one, we have to shift any that preceded it towards the end.
+ */
+static void ocfs2_xa_block_wipe_namevalue(struct ocfs2_xa_loc *loc)
+{
+	int i, offset;
+	int namevalue_offset, first_namevalue_offset, namevalue_size;
+	struct ocfs2_xattr_entry *entry = loc->xl_entry;
+	struct ocfs2_xattr_header *xh = loc->xl_header;
+	int count = le16_to_cpu(xh->xh_count);
+
+	namevalue_offset = le16_to_cpu(entry->xe_name_offset);
+	namevalue_size = namevalue_size_xe(entry);
+	first_namevalue_offset = ocfs2_xa_get_free_start(loc);
+
+	/* Shift the name+value pairs */
+	memmove((char *)xh + first_namevalue_offset + namevalue_size,
+		(char *)xh + first_namevalue_offset,
+		namevalue_offset - first_namevalue_offset);
+	memset((char *)xh + first_namevalue_offset, 0, namevalue_size);
+
+	/* Now tell xh->xh_entries about it */
+	for (i = 0; i < count; i++) {
+		offset = le16_to_cpu(xh->xh_entries[i].xe_name_offset);
+		if (offset < namevalue_offset)
+			le16_add_cpu(&xh->xh_entries[i].xe_name_offset,
+				     namevalue_size);
+	}
+
+	/*
+	 * Note that we don't update xh_free_start or xh_name_value_len
+	 * because they're not used in block-stored xattrs.
+	 */
+}
+
+static void ocfs2_xa_block_add_entry(struct ocfs2_xa_loc *loc, u32 name_hash)
+{
+	int count = le16_to_cpu(loc->xl_header->xh_count);
+	loc->xl_entry = &(loc->xl_header->xh_entries[count]);
+	le16_add_cpu(&loc->xl_header->xh_count, 1);
+	memset(loc->xl_entry, 0, sizeof(struct ocfs2_xattr_entry));
+}
+
+static void ocfs2_xa_block_add_namevalue(struct ocfs2_xa_loc *loc, int size)
+{
+	int free_start = ocfs2_xa_get_free_start(loc);
+
+	loc->xl_entry->xe_name_offset = cpu_to_le16(free_start - size);
+}
+
+static void ocfs2_xa_block_fill_value_buf(struct ocfs2_xa_loc *loc,
+					  struct ocfs2_xattr_value_buf *vb)
+{
+	struct buffer_head *bh = loc->xl_storage;
+
+	if (loc->xl_size == (bh->b_size -
+			     offsetof(struct ocfs2_xattr_block,
+				      xb_attrs.xb_header)))
+		vb->vb_access = ocfs2_journal_access_xb;
+	else
+		vb->vb_access = ocfs2_journal_access_di;
+	vb->vb_bh = bh;
 }
 
 /*
- * ocfs2_xattr_set_entry_local()
- *
- * Set, replace or remove extended attribute in local.
+ * Operations for xattrs stored in blocks.  This includes inline inode
+ * storage and unindexed ocfs2_xattr_blocks.
  */
-static void ocfs2_xattr_set_entry_local(struct inode *inode,
-					struct ocfs2_xattr_info *xi,
-					struct ocfs2_xattr_search *xs,
-					struct ocfs2_xattr_entry *last,
-					size_t min_offs)
+static const struct ocfs2_xa_loc_operations ocfs2_xa_block_loc_ops = {
+	.xlo_journal_access	= ocfs2_xa_block_journal_access,
+	.xlo_journal_dirty	= ocfs2_xa_block_journal_dirty,
+	.xlo_offset_pointer	= ocfs2_xa_block_offset_pointer,
+	.xlo_check_space	= ocfs2_xa_block_check_space,
+	.xlo_can_reuse		= ocfs2_xa_block_can_reuse,
+	.xlo_get_free_start	= ocfs2_xa_block_get_free_start,
+	.xlo_wipe_namevalue	= ocfs2_xa_block_wipe_namevalue,
+	.xlo_add_entry		= ocfs2_xa_block_add_entry,
+	.xlo_add_namevalue	= ocfs2_xa_block_add_namevalue,
+	.xlo_fill_value_buf	= ocfs2_xa_block_fill_value_buf,
+};
+
+static int ocfs2_xa_bucket_journal_access(handle_t *handle,
+					  struct ocfs2_xa_loc *loc, int type)
 {
-	size_t name_len = strlen(xi->name);
-	int i;
+	struct ocfs2_xattr_bucket *bucket = loc->xl_storage;
 
-	if (xi->value && xs->not_found) {
-		/* Insert the new xattr entry. */
-		le16_add_cpu(&xs->header->xh_count, 1);
-		ocfs2_xattr_set_type(last, xi->name_index);
-		ocfs2_xattr_set_local(last, 1);
-		last->xe_name_len = name_len;
-	} else {
-		void *first_val;
-		void *val;
-		size_t offs, size;
-
-		first_val = xs->base + min_offs;
-		offs = le16_to_cpu(xs->here->xe_name_offset);
-		val = xs->base + offs;
-
-		if (le64_to_cpu(xs->here->xe_value_size) >
-		    OCFS2_XATTR_INLINE_SIZE)
-			size = OCFS2_XATTR_SIZE(name_len) +
-				OCFS2_XATTR_ROOT_SIZE;
+	return ocfs2_xattr_bucket_journal_access(handle, bucket, type);
+}
+
+static void ocfs2_xa_bucket_journal_dirty(handle_t *handle,
+					  struct ocfs2_xa_loc *loc)
+{
+	struct ocfs2_xattr_bucket *bucket = loc->xl_storage;
+
+	ocfs2_xattr_bucket_journal_dirty(handle, bucket);
+}
+
+static void *ocfs2_xa_bucket_offset_pointer(struct ocfs2_xa_loc *loc,
+					    int offset)
+{
+	struct ocfs2_xattr_bucket *bucket = loc->xl_storage;
+	int block, block_offset;
+
+	/* The header is at the front of the bucket */
+	block = offset >> loc->xl_inode->i_sb->s_blocksize_bits;
+	block_offset = offset % loc->xl_inode->i_sb->s_blocksize;
+
+	return bucket_block(bucket, block) + block_offset;
+}
+
+static int ocfs2_xa_bucket_can_reuse(struct ocfs2_xa_loc *loc,
+				     struct ocfs2_xattr_info *xi)
+{
+	return namevalue_size_xe(loc->xl_entry) >=
+		namevalue_size_xi(xi);
+}
+
+static int ocfs2_xa_bucket_get_free_start(struct ocfs2_xa_loc *loc)
+{
+	struct ocfs2_xattr_bucket *bucket = loc->xl_storage;
+	return le16_to_cpu(bucket_xh(bucket)->xh_free_start);
+}
+
+static int ocfs2_bucket_align_free_start(struct super_block *sb,
+					 int free_start, int size)
+{
+	/*
+	 * We need to make sure that the name+value pair fits within
+	 * one block.
+	 */
+	if (((free_start - size) >> sb->s_blocksize_bits) !=
+	    ((free_start - 1) >> sb->s_blocksize_bits))
+		free_start -= free_start % sb->s_blocksize;
+
+	return free_start;
+}
+
+static int ocfs2_xa_bucket_check_space(struct ocfs2_xa_loc *loc,
+				       struct ocfs2_xattr_info *xi)
+{
+	int rc;
+	int count = le16_to_cpu(loc->xl_header->xh_count);
+	int free_start = ocfs2_xa_get_free_start(loc);
+	int needed_space = ocfs2_xi_entry_usage(xi);
+	int size = namevalue_size_xi(xi);
+	struct super_block *sb = loc->xl_inode->i_sb;
+
+	/*
+	 * Bucket storage does not reclaim name+value pairs it cannot
+	 * reuse.  They live as holes until the bucket fills, and then
+	 * the bucket is defragmented.  However, the bucket can reclaim
+	 * the ocfs2_xattr_entry.
+	 */
+	if (loc->xl_entry) {
+		/* Don't need space if we're reusing! */
+		if (ocfs2_xa_can_reuse_entry(loc, xi))
+			needed_space = 0;
 		else
-			size = OCFS2_XATTR_SIZE(name_len) +
-			OCFS2_XATTR_SIZE(le64_to_cpu(xs->here->xe_value_size));
-
-		if (xi->value && size == OCFS2_XATTR_SIZE(name_len) +
-				OCFS2_XATTR_SIZE(xi->value_len)) {
-			/* The old and the new value have the
-			   same size. Just replace the value. */
-			ocfs2_xattr_set_local(xs->here, 1);
-			xs->here->xe_value_size = cpu_to_le64(xi->value_len);
-			/* Clear value bytes. */
-			memset(val + OCFS2_XATTR_SIZE(name_len),
-			       0,
-			       OCFS2_XATTR_SIZE(xi->value_len));
-			memcpy(val + OCFS2_XATTR_SIZE(name_len),
-			       xi->value,
-			       xi->value_len);
-			return;
-		}
-		/* Remove the old name+value. */
-		memmove(first_val + size, first_val, val - first_val);
-		memset(first_val, 0, size);
-		xs->here->xe_name_hash = 0;
-		xs->here->xe_name_offset = 0;
-		ocfs2_xattr_set_local(xs->here, 1);
-		xs->here->xe_value_size = 0;
-
-		min_offs += size;
-
-		/* Adjust all value offsets. */
-		last = xs->header->xh_entries;
-		for (i = 0 ; i < le16_to_cpu(xs->header->xh_count); i++) {
-			size_t o = le16_to_cpu(last->xe_name_offset);
-
-			if (o < offs)
-				last->xe_name_offset = cpu_to_le16(o + size);
-			last += 1;
-		}
+			needed_space -= sizeof(struct ocfs2_xattr_entry);
+	}
+	BUG_ON(needed_space < 0);
 
-		if (!xi->value) {
-			/* Remove the old entry. */
-			last -= 1;
-			memmove(xs->here, xs->here + 1,
-				(void *)last - (void *)xs->here);
-			memset(last, 0, sizeof(struct ocfs2_xattr_entry));
-			le16_add_cpu(&xs->header->xh_count, -1);
-		}
+	if (free_start < size) {
+		if (needed_space)
+			return -ENOSPC;
+	} else {
+		/*
+		 * First we check if it would fit in the first place.
+		 * Below, we align the free start to a block.  This may
+		 * slide us below the minimum gap.  By checking unaligned
+		 * first, we avoid that error.
+		 */
+		rc = ocfs2_xa_check_space_helper(needed_space, free_start,
+						 count);
+		if (rc)
+			return rc;
+		free_start = ocfs2_bucket_align_free_start(sb, free_start,
+							   size);
 	}
-	if (xi->value) {
-		/* Insert the new name+value. */
-		size_t size = OCFS2_XATTR_SIZE(name_len) +
-				OCFS2_XATTR_SIZE(xi->value_len);
-		void *val = xs->base + min_offs - size;
+	return ocfs2_xa_check_space_helper(needed_space, free_start, count);
+}
+
+static void ocfs2_xa_bucket_wipe_namevalue(struct ocfs2_xa_loc *loc)
+{
+	le16_add_cpu(&loc->xl_header->xh_name_value_len,
+		     -namevalue_size_xe(loc->xl_entry));
+}
 
-		xs->here->xe_name_offset = cpu_to_le16(min_offs - size);
-		memset(val, 0, size);
-		memcpy(val, xi->name, name_len);
-		memcpy(val + OCFS2_XATTR_SIZE(name_len),
-		       xi->value,
-		       xi->value_len);
-		xs->here->xe_value_size = cpu_to_le64(xi->value_len);
-		ocfs2_xattr_set_local(xs->here, 1);
-		ocfs2_xattr_hash_entry(inode, xs->header, xs->here);
+static void ocfs2_xa_bucket_add_entry(struct ocfs2_xa_loc *loc, u32 name_hash)
+{
+	struct ocfs2_xattr_header *xh = loc->xl_header;
+	int count = le16_to_cpu(xh->xh_count);
+	int low = 0, high = count - 1, tmp;
+	struct ocfs2_xattr_entry *tmp_xe;
+
+	/*
+	 * We keep buckets sorted by name_hash, so we need to find
+	 * our insert place.
+	 */
+	while (low <= high && count) {
+		tmp = (low + high) / 2;
+		tmp_xe = &xh->xh_entries[tmp];
+
+		if (name_hash > le32_to_cpu(tmp_xe->xe_name_hash))
+			low = tmp + 1;
+		else if (name_hash < le32_to_cpu(tmp_xe->xe_name_hash))
+			high = tmp - 1;
+		else {
+			low = tmp;
+			break;
+		}
 	}
 
-	return;
+	if (low != count)
+		memmove(&xh->xh_entries[low + 1],
+			&xh->xh_entries[low],
+			((count - low) * sizeof(struct ocfs2_xattr_entry)));
+
+	le16_add_cpu(&xh->xh_count, 1);
+	loc->xl_entry = &xh->xh_entries[low];
+	memset(loc->xl_entry, 0, sizeof(struct ocfs2_xattr_entry));
+}
+
+static void ocfs2_xa_bucket_add_namevalue(struct ocfs2_xa_loc *loc, int size)
+{
+	int free_start = ocfs2_xa_get_free_start(loc);
+	struct ocfs2_xattr_header *xh = loc->xl_header;
+	struct super_block *sb = loc->xl_inode->i_sb;
+	int nameval_offset;
+
+	free_start = ocfs2_bucket_align_free_start(sb, free_start, size);
+	nameval_offset = free_start - size;
+	loc->xl_entry->xe_name_offset = cpu_to_le16(nameval_offset);
+	xh->xh_free_start = cpu_to_le16(nameval_offset);
+	le16_add_cpu(&xh->xh_name_value_len, size);
+
+}
+
+static void ocfs2_xa_bucket_fill_value_buf(struct ocfs2_xa_loc *loc,
+					   struct ocfs2_xattr_value_buf *vb)
+{
+	struct ocfs2_xattr_bucket *bucket = loc->xl_storage;
+	struct super_block *sb = loc->xl_inode->i_sb;
+	int nameval_offset = le16_to_cpu(loc->xl_entry->xe_name_offset);
+	int size = namevalue_size_xe(loc->xl_entry);
+	int block_offset = nameval_offset >> sb->s_blocksize_bits;
+
+	/* Values are not allowed to straddle block boundaries */
+	BUG_ON(block_offset !=
+	       ((nameval_offset + size - 1) >> sb->s_blocksize_bits));
+	/* We expect the bucket to be filled in */
+	BUG_ON(!bucket->bu_bhs[block_offset]);
+
+	vb->vb_access = ocfs2_journal_access;
+	vb->vb_bh = bucket->bu_bhs[block_offset];
+}
+
+/* Operations for xattrs stored in buckets. */
+static const struct ocfs2_xa_loc_operations ocfs2_xa_bucket_loc_ops = {
+	.xlo_journal_access	= ocfs2_xa_bucket_journal_access,
+	.xlo_journal_dirty	= ocfs2_xa_bucket_journal_dirty,
+	.xlo_offset_pointer	= ocfs2_xa_bucket_offset_pointer,
+	.xlo_check_space	= ocfs2_xa_bucket_check_space,
+	.xlo_can_reuse		= ocfs2_xa_bucket_can_reuse,
+	.xlo_get_free_start	= ocfs2_xa_bucket_get_free_start,
+	.xlo_wipe_namevalue	= ocfs2_xa_bucket_wipe_namevalue,
+	.xlo_add_entry		= ocfs2_xa_bucket_add_entry,
+	.xlo_add_namevalue	= ocfs2_xa_bucket_add_namevalue,
+	.xlo_fill_value_buf	= ocfs2_xa_bucket_fill_value_buf,
+};
+
+static unsigned int ocfs2_xa_value_clusters(struct ocfs2_xa_loc *loc)
+{
+	struct ocfs2_xattr_value_buf vb;
+
+	if (ocfs2_xattr_is_local(loc->xl_entry))
+		return 0;
+
+	ocfs2_xa_fill_value_buf(loc, &vb);
+	return le32_to_cpu(vb.vb_xv->xr_clusters);
+}
+
+static int ocfs2_xa_value_truncate(struct ocfs2_xa_loc *loc, u64 bytes,
+				   struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int trunc_rc, access_rc;
+	struct ocfs2_xattr_value_buf vb;
+
+	ocfs2_xa_fill_value_buf(loc, &vb);
+	trunc_rc = ocfs2_xattr_value_truncate(loc->xl_inode, &vb, bytes,
+					      ctxt);
+
+	/*
+	 * The caller of ocfs2_xa_value_truncate() has already called
+	 * ocfs2_xa_journal_access on the loc.  However, The truncate code
+	 * calls ocfs2_extend_trans().  This may commit the previous
+	 * transaction and open a new one.  If this is a bucket, truncate
+	 * could leave only vb->vb_bh set up for journaling.  Meanwhile,
+	 * the caller is expecting to dirty the entire bucket.  So we must
+	 * reset the journal work.  We do this even if truncate has failed,
+	 * as it could have failed after committing the extend.
+	 */
+	access_rc = ocfs2_xa_journal_access(ctxt->handle, loc,
+					    OCFS2_JOURNAL_ACCESS_WRITE);
+
+	/* Errors in truncate take precedence */
+	return trunc_rc ? trunc_rc : access_rc;
+}
+
+static void ocfs2_xa_remove_entry(struct ocfs2_xa_loc *loc)
+{
+	int index, count;
+	struct ocfs2_xattr_header *xh = loc->xl_header;
+	struct ocfs2_xattr_entry *entry = loc->xl_entry;
+
+	ocfs2_xa_wipe_namevalue(loc);
+	loc->xl_entry = NULL;
+
+	le16_add_cpu(&xh->xh_count, -1);
+	count = le16_to_cpu(xh->xh_count);
+
+	/*
+	 * Only zero out the entry if there are more remaining.  This is
+	 * important for an empty bucket, as it keeps track of the
+	 * bucket's hash value.  It doesn't hurt empty block storage.
+	 */
+	if (count) {
+		index = ((char *)entry - (char *)&xh->xh_entries) /
+			sizeof(struct ocfs2_xattr_entry);
+		memmove(&xh->xh_entries[index], &xh->xh_entries[index + 1],
+			(count - index) * sizeof(struct ocfs2_xattr_entry));
+		memset(&xh->xh_entries[count], 0,
+		       sizeof(struct ocfs2_xattr_entry));
+	}
 }
 
 /*
- * ocfs2_xattr_set_entry()
+ * If we have a problem adjusting the size of an external value during
+ * ocfs2_xa_prepare_entry() or ocfs2_xa_remove(), we may have an xattr
+ * in an intermediate state.  For example, the value may be partially
+ * truncated.
+ *
+ * If the value tree hasn't changed, the extend/truncate went nowhere.
+ * We have nothing to do.  The caller can treat it as a straight error.
  *
- * Set extended attribute entry into inode or block.
+ * If the value tree got partially truncated, we now have a corrupted
+ * extended attribute.  We're going to wipe its entry and leak the
+ * clusters.  Better to leak some storage than leave a corrupt entry.
  *
- * If extended attribute value size > OCFS2_XATTR_INLINE_SIZE,
- * We first insert tree root(ocfs2_xattr_value_root) with set_entry_local(),
- * then set value in B tree with set_value_outside().
+ * If the value tree grew, it obviously didn't grow enough for the
+ * new entry.  We're not going to try and reclaim those clusters either.
+ * If there was already an external value there (orig_clusters != 0),
+ * the new clusters are attached safely and we can just leave the old
+ * value in place.  If there was no external value there, we remove
+ * the entry.
+ *
+ * This way, the xattr block we store in the journal will be consistent.
+ * If the size change broke because of the journal, no changes will hit
+ * disk anyway.
  */
-static int ocfs2_xattr_set_entry(struct inode *inode,
-				 struct ocfs2_xattr_info *xi,
-				 struct ocfs2_xattr_search *xs,
-				 struct ocfs2_xattr_set_ctxt *ctxt,
-				 int flag)
-{
-	struct ocfs2_xattr_entry *last;
-	struct ocfs2_inode_info *oi = OCFS2_I(inode);
-	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
-	size_t min_offs = xs->end - xs->base, name_len = strlen(xi->name);
-	size_t size_l = 0;
-	handle_t *handle = ctxt->handle;
-	int free, i, ret;
-	struct ocfs2_xattr_info xi_l = {
-		.name_index = xi->name_index,
-		.name = xi->name,
-		.value = xi->value,
-		.value_len = xi->value_len,
-	};
-	struct ocfs2_xattr_value_buf vb = {
-		.vb_bh = xs->xattr_bh,
-		.vb_access = ocfs2_journal_access_di,
-	};
+static void ocfs2_xa_cleanup_value_truncate(struct ocfs2_xa_loc *loc,
+					    const char *what,
+					    unsigned int orig_clusters)
+{
+	unsigned int new_clusters = ocfs2_xa_value_clusters(loc);
+	char *nameval_buf = ocfs2_xa_offset_pointer(loc,
+				le16_to_cpu(loc->xl_entry->xe_name_offset));
+
+	if (new_clusters < orig_clusters) {
+		mlog(ML_ERROR,
+		     "Partial truncate while %s xattr %.*s.  Leaking "
+		     "%u clusters and removing the entry\n",
+		     what, loc->xl_entry->xe_name_len, nameval_buf,
+		     orig_clusters - new_clusters);
+		ocfs2_xa_remove_entry(loc);
+	} else if (!orig_clusters) {
+		mlog(ML_ERROR,
+		     "Unable to allocate an external value for xattr "
+		     "%.*s safely.  Leaking %u clusters and removing the "
+		     "entry\n",
+		     loc->xl_entry->xe_name_len, nameval_buf,
+		     new_clusters - orig_clusters);
+		ocfs2_xa_remove_entry(loc);
+	} else if (new_clusters > orig_clusters)
+		mlog(ML_ERROR,
+		     "Unable to grow xattr %.*s safely.  %u new clusters "
+		     "have been added, but the value will not be "
+		     "modified\n",
+		     loc->xl_entry->xe_name_len, nameval_buf,
+		     new_clusters - orig_clusters);
+}
+
+static int ocfs2_xa_remove(struct ocfs2_xa_loc *loc,
+			   struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int rc = 0;
+	unsigned int orig_clusters;
+
+	if (!ocfs2_xattr_is_local(loc->xl_entry)) {
+		orig_clusters = ocfs2_xa_value_clusters(loc);
+		rc = ocfs2_xa_value_truncate(loc, 0, ctxt);
+		if (rc) {
+			mlog_errno(rc);
+			/*
+			 * Since this is remove, we can return 0 if
+			 * ocfs2_xa_cleanup_value_truncate() is going to
+			 * wipe the entry anyway.  So we check the
+			 * cluster count as well.
+			 */
+			if (orig_clusters != ocfs2_xa_value_clusters(loc))
+				rc = 0;
+			ocfs2_xa_cleanup_value_truncate(loc, "removing",
+							orig_clusters);
+			if (rc)
+				goto out;
+		}
+	}
 
-	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
-		BUG_ON(xs->xattr_bh == xs->inode_bh);
-		vb.vb_access = ocfs2_journal_access_xb;
-	} else
-		BUG_ON(xs->xattr_bh != xs->inode_bh);
+	ocfs2_xa_remove_entry(loc);
 
-	/* Compute min_offs, last and free space. */
-	last = xs->header->xh_entries;
+out:
+	return rc;
+}
 
-	for (i = 0 ; i < le16_to_cpu(xs->header->xh_count); i++) {
-		size_t offs = le16_to_cpu(last->xe_name_offset);
-		if (offs < min_offs)
-			min_offs = offs;
-		last += 1;
-	}
+static void ocfs2_xa_install_value_root(struct ocfs2_xa_loc *loc)
+{
+	int name_size = OCFS2_XATTR_SIZE(loc->xl_entry->xe_name_len);
+	char *nameval_buf;
 
-	free = min_offs - ((void *)last - xs->base) - OCFS2_XATTR_HEADER_GAP;
-	if (free < 0)
-		return -EIO;
+	nameval_buf = ocfs2_xa_offset_pointer(loc,
+				le16_to_cpu(loc->xl_entry->xe_name_offset));
+	memcpy(nameval_buf + name_size, &def_xv, OCFS2_XATTR_ROOT_SIZE);
+}
 
-	if (!xs->not_found) {
-		size_t size = 0;
-		if (ocfs2_xattr_is_local(xs->here))
-			size = OCFS2_XATTR_SIZE(name_len) +
-			OCFS2_XATTR_SIZE(le64_to_cpu(xs->here->xe_value_size));
-		else
-			size = OCFS2_XATTR_SIZE(name_len) +
-				OCFS2_XATTR_ROOT_SIZE;
-		free += (size + sizeof(struct ocfs2_xattr_entry));
-	}
-	/* Check free space in inode or block */
-	if (xi->value && xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
-		if (free < sizeof(struct ocfs2_xattr_entry) +
-			   OCFS2_XATTR_SIZE(name_len) +
-			   OCFS2_XATTR_ROOT_SIZE) {
-			ret = -ENOSPC;
-			goto out;
+/*
+ * Take an existing entry and make it ready for the new value.  This
+ * won't allocate space, but it may free space.  It should be ready for
+ * ocfs2_xa_prepare_entry() to finish the work.
+ */
+static int ocfs2_xa_reuse_entry(struct ocfs2_xa_loc *loc,
+				struct ocfs2_xattr_info *xi,
+				struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int rc = 0;
+	int name_size = OCFS2_XATTR_SIZE(xi->xi_name_len);
+	unsigned int orig_clusters;
+	char *nameval_buf;
+	int xe_local = ocfs2_xattr_is_local(loc->xl_entry);
+	int xi_local = xi->xi_value_len <= OCFS2_XATTR_INLINE_SIZE;
+
+	BUG_ON(OCFS2_XATTR_SIZE(loc->xl_entry->xe_name_len) !=
+	       name_size);
+
+	nameval_buf = ocfs2_xa_offset_pointer(loc,
+				le16_to_cpu(loc->xl_entry->xe_name_offset));
+	if (xe_local) {
+		memset(nameval_buf + name_size, 0,
+		       namevalue_size_xe(loc->xl_entry) - name_size);
+		if (!xi_local)
+			ocfs2_xa_install_value_root(loc);
+	} else {
+		orig_clusters = ocfs2_xa_value_clusters(loc);
+		if (xi_local) {
+			rc = ocfs2_xa_value_truncate(loc, 0, ctxt);
+			if (rc < 0)
+				mlog_errno(rc);
+			else
+				memset(nameval_buf + name_size, 0,
+				       namevalue_size_xe(loc->xl_entry) -
+				       name_size);
+		} else if (le64_to_cpu(loc->xl_entry->xe_value_size) >
+			   xi->xi_value_len) {
+			rc = ocfs2_xa_value_truncate(loc, xi->xi_value_len,
+						     ctxt);
+			if (rc < 0)
+				mlog_errno(rc);
 		}
-		size_l = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
-		xi_l.value = (void *)&def_xv;
-		xi_l.value_len = OCFS2_XATTR_ROOT_SIZE;
-	} else if (xi->value) {
-		if (free < sizeof(struct ocfs2_xattr_entry) +
-			   OCFS2_XATTR_SIZE(name_len) +
-			   OCFS2_XATTR_SIZE(xi->value_len)) {
-			ret = -ENOSPC;
+
+		if (rc) {
+			ocfs2_xa_cleanup_value_truncate(loc, "reusing",
+							orig_clusters);
 			goto out;
 		}
 	}
 
-	if (!xs->not_found) {
-		/* For existing extended attribute */
-		size_t size = OCFS2_XATTR_SIZE(name_len) +
-			OCFS2_XATTR_SIZE(le64_to_cpu(xs->here->xe_value_size));
-		size_t offs = le16_to_cpu(xs->here->xe_name_offset);
-		void *val = xs->base + offs;
+	loc->xl_entry->xe_value_size = cpu_to_le64(xi->xi_value_len);
+	ocfs2_xattr_set_local(loc->xl_entry, xi_local);
 
-		if (ocfs2_xattr_is_local(xs->here) && size == size_l) {
-			/* Replace existing local xattr with tree root */
-			ret = ocfs2_xattr_set_value_outside(inode, xi, xs,
-							    ctxt, &vb, offs);
-			if (ret < 0)
-				mlog_errno(ret);
-			goto out;
-		} else if (!ocfs2_xattr_is_local(xs->here)) {
-			/* For existing xattr which has value outside */
-			vb.vb_xv = (struct ocfs2_xattr_value_root *)
-				(val + OCFS2_XATTR_SIZE(name_len));
+out:
+	return rc;
+}
 
-			if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
-				/*
-				 * If new value need set outside also,
-				 * first truncate old value to new value,
-				 * then set new value with set_value_outside().
-				 */
-				ret = ocfs2_xattr_value_truncate(inode,
-								 &vb,
-								 xi->value_len,
-								 ctxt);
-				if (ret < 0) {
-					mlog_errno(ret);
-					goto out;
-				}
+/*
+ * Prepares loc->xl_entry to receive the new xattr.  This includes
+ * properly setting up the name+value pair region.  If loc->xl_entry
+ * already exists, it will take care of modifying it appropriately.
+ *
+ * Note that this modifies the data.  You did journal_access already,
+ * right?
+ */
+static int ocfs2_xa_prepare_entry(struct ocfs2_xa_loc *loc,
+				  struct ocfs2_xattr_info *xi,
+				  u32 name_hash,
+				  struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int rc = 0;
+	unsigned int orig_clusters;
+	__le64 orig_value_size = 0;
 
-				ret = ocfs2_xattr_update_entry(inode,
-							       handle,
-							       xi,
-							       xs,
-							       &vb,
-							       offs);
-				if (ret < 0) {
-					mlog_errno(ret);
-					goto out;
-				}
+	rc = ocfs2_xa_check_space(loc, xi);
+	if (rc)
+		goto out;
 
-				ret = __ocfs2_xattr_set_value_outside(inode,
-								handle,
-								&vb,
-								xi->value,
-								xi->value_len);
-				if (ret < 0)
-					mlog_errno(ret);
+	if (loc->xl_entry) {
+		if (ocfs2_xa_can_reuse_entry(loc, xi)) {
+			orig_value_size = loc->xl_entry->xe_value_size;
+			rc = ocfs2_xa_reuse_entry(loc, xi, ctxt);
+			if (rc)
+				goto out;
+			goto alloc_value;
+		}
+
+		if (!ocfs2_xattr_is_local(loc->xl_entry)) {
+			orig_clusters = ocfs2_xa_value_clusters(loc);
+			rc = ocfs2_xa_value_truncate(loc, 0, ctxt);
+			if (rc) {
+				mlog_errno(rc);
+				ocfs2_xa_cleanup_value_truncate(loc,
+								"overwriting",
+								orig_clusters);
 				goto out;
-			} else {
-				/*
-				 * If new value need set in local,
-				 * just trucate old value to zero.
-				 */
-				 ret = ocfs2_xattr_value_truncate(inode,
-								  &vb,
-								  0,
-								  ctxt);
-				if (ret < 0)
-					mlog_errno(ret);
 			}
 		}
+		ocfs2_xa_wipe_namevalue(loc);
+	} else
+		ocfs2_xa_add_entry(loc, name_hash);
+
+	/*
+	 * If we get here, we have a blank entry.  Fill it.  We grow our
+	 * name+value pair back from the end.
+	 */
+	ocfs2_xa_add_namevalue(loc, xi);
+	if (xi->xi_value_len > OCFS2_XATTR_INLINE_SIZE)
+		ocfs2_xa_install_value_root(loc);
+
+alloc_value:
+	if (xi->xi_value_len > OCFS2_XATTR_INLINE_SIZE) {
+		orig_clusters = ocfs2_xa_value_clusters(loc);
+		rc = ocfs2_xa_value_truncate(loc, xi->xi_value_len, ctxt);
+		if (rc < 0) {
+			/*
+			 * If we tried to grow an existing external value,
+			 * ocfs2_xa_cleanuP-value_truncate() is going to
+			 * let it stand.  We have to restore its original
+			 * value size.
+			 */
+			loc->xl_entry->xe_value_size = orig_value_size;
+			ocfs2_xa_cleanup_value_truncate(loc, "growing",
+							orig_clusters);
+			mlog_errno(rc);
+		}
 	}
 
-	ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), xs->inode_bh,
+out:
+	return rc;
+}
+
+/*
+ * Store the value portion of the name+value pair.  This will skip
+ * values that are stored externally.  Their tree roots were set up
+ * by ocfs2_xa_prepare_entry().
+ */
+static int ocfs2_xa_store_value(struct ocfs2_xa_loc *loc,
+				struct ocfs2_xattr_info *xi,
+				struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int rc = 0;
+	int nameval_offset = le16_to_cpu(loc->xl_entry->xe_name_offset);
+	int name_size = OCFS2_XATTR_SIZE(xi->xi_name_len);
+	char *nameval_buf;
+	struct ocfs2_xattr_value_buf vb;
+
+	nameval_buf = ocfs2_xa_offset_pointer(loc, nameval_offset);
+	if (xi->xi_value_len > OCFS2_XATTR_INLINE_SIZE) {
+		ocfs2_xa_fill_value_buf(loc, &vb);
+		rc = __ocfs2_xattr_set_value_outside(loc->xl_inode,
+						     ctxt->handle, &vb,
+						     xi->xi_value,
+						     xi->xi_value_len);
+	} else
+		memcpy(nameval_buf + name_size, xi->xi_value, xi->xi_value_len);
+
+	return rc;
+}
+
+static int ocfs2_xa_set(struct ocfs2_xa_loc *loc,
+			struct ocfs2_xattr_info *xi,
+			struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int ret;
+	u32 name_hash = ocfs2_xattr_name_hash(loc->xl_inode, xi->xi_name,
+					      xi->xi_name_len);
+
+	ret = ocfs2_xa_journal_access(ctxt->handle, loc,
 				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
-		ret = vb.vb_access(handle, INODE_CACHE(inode), vb.vb_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
-		if (ret) {
-			mlog_errno(ret);
-			goto out;
-		}
-	}
-
 	/*
-	 * Set value in local, include set tree root in local.
-	 * This is the first step for value size >INLINE_SIZE.
+	 * From here on out, everything is going to modify the buffer a
+	 * little.  Errors are going to leave the xattr header in a
+	 * sane state.  Thus, even with errors we dirty the sucker.
 	 */
-	ocfs2_xattr_set_entry_local(inode, &xi_l, xs, last, min_offs);
 
-	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
-		ret = ocfs2_journal_dirty(handle, xs->xattr_bh);
-		if (ret < 0) {
-			mlog_errno(ret);
-			goto out;
-		}
+	/* Don't worry, we are never called with !xi_value and !xl_entry */
+	if (!xi->xi_value) {
+		ret = ocfs2_xa_remove(loc, ctxt);
+		goto out_dirty;
 	}
 
-	if (!(oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL) &&
-	    (flag & OCFS2_INLINE_XATTR_FL)) {
-		struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-		unsigned int xattrsize = osb->s_xattr_inline_size;
-
-		/*
-		 * Adjust extent record count or inline data size
-		 * to reserve space for extended attribute.
-		 */
-		if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
-			struct ocfs2_inline_data *idata = &di->id2.i_data;
-			le16_add_cpu(&idata->id_count, -xattrsize);
-		} else if (!(ocfs2_inode_is_fast_symlink(inode))) {
-			struct ocfs2_extent_list *el = &di->id2.i_list;
-			le16_add_cpu(&el->l_count, -(xattrsize /
-					sizeof(struct ocfs2_extent_rec)));
-		}
-		di->i_xattr_inline_size = cpu_to_le16(xattrsize);
+	ret = ocfs2_xa_prepare_entry(loc, xi, name_hash, ctxt);
+	if (ret) {
+		if (ret != -ENOSPC)
+			mlog_errno(ret);
+		goto out_dirty;
 	}
-	/* Update xattr flag */
-	spin_lock(&oi->ip_lock);
-	oi->ip_dyn_features |= flag;
-	di->i_dyn_features = cpu_to_le16(oi->ip_dyn_features);
-	spin_unlock(&oi->ip_lock);
 
-	ret = ocfs2_journal_dirty(handle, xs->inode_bh);
-	if (ret < 0)
+	ret = ocfs2_xa_store_value(loc, xi, ctxt);
+	if (ret)
 		mlog_errno(ret);
 
-	if (!ret && xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
-		/*
-		 * Set value outside in B tree.
-		 * This is the second step for value size > INLINE_SIZE.
-		 */
-		size_t offs = le16_to_cpu(xs->here->xe_name_offset);
-		ret = ocfs2_xattr_set_value_outside(inode, xi, xs, ctxt,
-						    &vb, offs);
-		if (ret < 0) {
-			int ret2;
+out_dirty:
+	ocfs2_xa_journal_dirty(ctxt->handle, loc);
 
-			mlog_errno(ret);
-			/*
-			 * If set value outside failed, we have to clean
-			 * the junk tree root we have already set in local.
-			 */
-			ret2 = ocfs2_xattr_cleanup(inode, ctxt->handle,
-						   xi, xs, &vb, offs);
-			if (ret2 < 0)
-				mlog_errno(ret2);
-		}
-	}
 out:
 	return ret;
 }
 
+static void ocfs2_init_dinode_xa_loc(struct ocfs2_xa_loc *loc,
+				     struct inode *inode,
+				     struct buffer_head *bh,
+				     struct ocfs2_xattr_entry *entry)
+{
+	struct ocfs2_dinode *di = (struct ocfs2_dinode *)bh->b_data;
+
+	BUG_ON(!(OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_XATTR_FL));
+
+	loc->xl_inode = inode;
+	loc->xl_ops = &ocfs2_xa_block_loc_ops;
+	loc->xl_storage = bh;
+	loc->xl_entry = entry;
+	loc->xl_size = le16_to_cpu(di->i_xattr_inline_size);
+	loc->xl_header =
+		(struct ocfs2_xattr_header *)(bh->b_data + bh->b_size -
+					      loc->xl_size);
+}
+
+static void ocfs2_init_xattr_block_xa_loc(struct ocfs2_xa_loc *loc,
+					  struct inode *inode,
+					  struct buffer_head *bh,
+					  struct ocfs2_xattr_entry *entry)
+{
+	struct ocfs2_xattr_block *xb =
+		(struct ocfs2_xattr_block *)bh->b_data;
+
+	BUG_ON(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED);
+
+	loc->xl_inode = inode;
+	loc->xl_ops = &ocfs2_xa_block_loc_ops;
+	loc->xl_storage = bh;
+	loc->xl_header = &(xb->xb_attrs.xb_header);
+	loc->xl_entry = entry;
+	loc->xl_size = bh->b_size - offsetof(struct ocfs2_xattr_block,
+					     xb_attrs.xb_header);
+}
+
+static void ocfs2_init_xattr_bucket_xa_loc(struct ocfs2_xa_loc *loc,
+					   struct ocfs2_xattr_bucket *bucket,
+					   struct ocfs2_xattr_entry *entry)
+{
+	loc->xl_inode = bucket->bu_inode;
+	loc->xl_ops = &ocfs2_xa_bucket_loc_ops;
+	loc->xl_storage = bucket;
+	loc->xl_header = bucket_xh(bucket);
+	loc->xl_entry = entry;
+	loc->xl_size = OCFS2_XATTR_BUCKET_SIZE;
+}
+
 /*
  * In xattr remove, if it is stored outside and refcounted, we may have
  * the chance to split the refcount tree. So need the allocators.
@@ -2149,6 +2683,55 @@ static int ocfs2_xattr_ibody_find(struct inode *inode,
 	return 0;
 }
 
+static int ocfs2_xattr_ibody_init(struct inode *inode,
+				  struct buffer_head *di_bh,
+				  struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int ret;
+	struct ocfs2_inode_info *oi = OCFS2_I(inode);
+	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	unsigned int xattrsize = osb->s_xattr_inline_size;
+
+	if (!ocfs2_xattr_has_space_inline(inode, di)) {
+		ret = -ENOSPC;
+		goto out;
+	}
+
+	ret = ocfs2_journal_access_di(ctxt->handle, INODE_CACHE(inode), di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
+
+	/*
+	 * Adjust extent record count or inline data size
+	 * to reserve space for extended attribute.
+	 */
+	if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
+		struct ocfs2_inline_data *idata = &di->id2.i_data;
+		le16_add_cpu(&idata->id_count, -xattrsize);
+	} else if (!(ocfs2_inode_is_fast_symlink(inode))) {
+		struct ocfs2_extent_list *el = &di->id2.i_list;
+		le16_add_cpu(&el->l_count, -(xattrsize /
+					     sizeof(struct ocfs2_extent_rec)));
+	}
+	di->i_xattr_inline_size = cpu_to_le16(xattrsize);
+
+	spin_lock(&oi->ip_lock);
+	oi->ip_dyn_features |= OCFS2_INLINE_XATTR_FL|OCFS2_HAS_XATTR_FL;
+	di->i_dyn_features = cpu_to_le16(oi->ip_dyn_features);
+	spin_unlock(&oi->ip_lock);
+
+	ret = ocfs2_journal_dirty(ctxt->handle, di_bh);
+	if (ret < 0)
+		mlog_errno(ret);
+
+out:
+	return ret;
+}
+
 /*
  * ocfs2_xattr_ibody_set()
  *
@@ -2160,9 +2743,10 @@ static int ocfs2_xattr_ibody_set(struct inode *inode,
 				 struct ocfs2_xattr_search *xs,
 				 struct ocfs2_xattr_set_ctxt *ctxt)
 {
+	int ret;
 	struct ocfs2_inode_info *oi = OCFS2_I(inode);
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
-	int ret;
+	struct ocfs2_xa_loc loc;
 
 	if (inode->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE)
 		return -ENOSPC;
@@ -2175,8 +2759,25 @@ static int ocfs2_xattr_ibody_set(struct inode *inode,
 		}
 	}
 
-	ret = ocfs2_xattr_set_entry(inode, xi, xs, ctxt,
-				(OCFS2_INLINE_XATTR_FL | OCFS2_HAS_XATTR_FL));
+	if (!(oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL)) {
+		ret = ocfs2_xattr_ibody_init(inode, xs->inode_bh, ctxt);
+		if (ret) {
+			if (ret != -ENOSPC)
+				mlog_errno(ret);
+			goto out;
+		}
+	}
+
+	ocfs2_init_dinode_xa_loc(&loc, inode, xs->inode_bh,
+				 xs->not_found ? NULL : xs->here);
+	ret = ocfs2_xa_set(&loc, xi, ctxt);
+	if (ret) {
+		if (ret != -ENOSPC)
+			mlog_errno(ret);
+		goto out;
+	}
+	xs->here = loc.xl_entry;
+
 out:
 	up_write(&oi->ip_alloc_sem);
 
@@ -2236,12 +2837,11 @@ cleanup:
 	return ret;
 }
 
-static int ocfs2_create_xattr_block(handle_t *handle,
-				    struct inode *inode,
+static int ocfs2_create_xattr_block(struct inode *inode,
 				    struct buffer_head *inode_bh,
-				    struct ocfs2_alloc_context *meta_ac,
-				    struct buffer_head **ret_bh,
-				    int indexed)
+				    struct ocfs2_xattr_set_ctxt *ctxt,
+				    int indexed,
+				    struct buffer_head **ret_bh)
 {
 	int ret;
 	u16 suballoc_bit_start;
@@ -2252,14 +2852,14 @@ static int ocfs2_create_xattr_block(handle_t *handle,
 	struct buffer_head *new_bh = NULL;
 	struct ocfs2_xattr_block *xblk;
 
-	ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), inode_bh,
-				      OCFS2_JOURNAL_ACCESS_CREATE);
+	ret = ocfs2_journal_access_di(ctxt->handle, INODE_CACHE(inode),
+				      inode_bh, OCFS2_JOURNAL_ACCESS_CREATE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto end;
 	}
 
-	ret = ocfs2_claim_metadata(osb, handle, meta_ac, 1,
+	ret = ocfs2_claim_metadata(osb, ctxt->handle, ctxt->meta_ac, 1,
 				   &suballoc_bit_start, &num_got,
 				   &first_blkno);
 	if (ret < 0) {
@@ -2270,7 +2870,7 @@ static int ocfs2_create_xattr_block(handle_t *handle,
 	new_bh = sb_getblk(inode->i_sb, first_blkno);
 	ocfs2_set_new_buffer_uptodate(INODE_CACHE(inode), new_bh);
 
-	ret = ocfs2_journal_access_xb(handle, INODE_CACHE(inode),
+	ret = ocfs2_journal_access_xb(ctxt->handle, INODE_CACHE(inode),
 				      new_bh,
 				      OCFS2_JOURNAL_ACCESS_CREATE);
 	if (ret < 0) {
@@ -2282,11 +2882,10 @@ static int ocfs2_create_xattr_block(handle_t *handle,
 	xblk = (struct ocfs2_xattr_block *)new_bh->b_data;
 	memset(xblk, 0, inode->i_sb->s_blocksize);
 	strcpy((void *)xblk, OCFS2_XATTR_BLOCK_SIGNATURE);
-	xblk->xb_suballoc_slot = cpu_to_le16(osb->slot_num);
+	xblk->xb_suballoc_slot = cpu_to_le16(ctxt->meta_ac->ac_alloc_slot);
 	xblk->xb_suballoc_bit = cpu_to_le16(suballoc_bit_start);
 	xblk->xb_fs_generation = cpu_to_le32(osb->fs_generation);
 	xblk->xb_blkno = cpu_to_le64(first_blkno);
-
 	if (indexed) {
 		struct ocfs2_xattr_tree_root *xr = &xblk->xb_attrs.xb_root;
 		xr->xt_clusters = cpu_to_le32(1);
@@ -2297,14 +2896,17 @@ static int ocfs2_create_xattr_block(handle_t *handle,
 		xr->xt_list.l_next_free_rec = cpu_to_le16(1);
 		xblk->xb_flags = cpu_to_le16(OCFS2_XATTR_INDEXED);
 	}
+	ocfs2_journal_dirty(ctxt->handle, new_bh);
 
-	ret = ocfs2_journal_dirty(handle, new_bh);
-	if (ret < 0) {
-		mlog_errno(ret);
-		goto end;
-	}
+	/* Add it to the inode */
 	di->i_xattr_loc = cpu_to_le64(first_blkno);
-	ocfs2_journal_dirty(handle, inode_bh);
+
+	spin_lock(&OCFS2_I(inode)->ip_lock);
+	OCFS2_I(inode)->ip_dyn_features |= OCFS2_HAS_XATTR_FL;
+	di->i_dyn_features = cpu_to_le16(OCFS2_I(inode)->ip_dyn_features);
+	spin_unlock(&OCFS2_I(inode)->ip_lock);
+
+	ocfs2_journal_dirty(ctxt->handle, inode_bh);
 
 	*ret_bh = new_bh;
 	new_bh = NULL;
@@ -2326,13 +2928,13 @@ static int ocfs2_xattr_block_set(struct inode *inode,
 				 struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	struct buffer_head *new_bh = NULL;
-	handle_t *handle = ctxt->handle;
 	struct ocfs2_xattr_block *xblk = NULL;
 	int ret;
+	struct ocfs2_xa_loc loc;
 
 	if (!xs->xattr_bh) {
-		ret = ocfs2_create_xattr_block(handle, inode, xs->inode_bh,
-					       ctxt->meta_ac, &new_bh, 0);
+		ret = ocfs2_create_xattr_block(inode, xs->inode_bh, ctxt,
+					       0, &new_bh);
 		if (ret) {
 			mlog_errno(ret);
 			goto end;
@@ -2348,21 +2950,25 @@ static int ocfs2_xattr_block_set(struct inode *inode,
 		xblk = (struct ocfs2_xattr_block *)xs->xattr_bh->b_data;
 
 	if (!(le16_to_cpu(xblk->xb_flags) & OCFS2_XATTR_INDEXED)) {
-		/* Set extended attribute into external block */
-		ret = ocfs2_xattr_set_entry(inode, xi, xs, ctxt,
-					    OCFS2_HAS_XATTR_FL);
-		if (!ret || ret != -ENOSPC)
-			goto end;
+		ocfs2_init_xattr_block_xa_loc(&loc, inode, xs->xattr_bh,
+					      xs->not_found ? NULL : xs->here);
 
-		ret = ocfs2_xattr_create_index_block(inode, xs, ctxt);
-		if (ret)
+		ret = ocfs2_xa_set(&loc, xi, ctxt);
+		if (!ret)
+			xs->here = loc.xl_entry;
+		else if (ret != -ENOSPC)
 			goto end;
+		else {
+			ret = ocfs2_xattr_create_index_block(inode, xs, ctxt);
+			if (ret)
+				goto end;
+		}
 	}
 
-	ret = ocfs2_xattr_set_entry_index_block(inode, xi, xs, ctxt);
+	if (le16_to_cpu(xblk->xb_flags) & OCFS2_XATTR_INDEXED)
+		ret = ocfs2_xattr_set_entry_index_block(inode, xi, xs, ctxt);
 
 end:
-
 	return ret;
 }
 
@@ -2371,7 +2977,6 @@ static int ocfs2_xattr_can_be_in_inode(struct inode *inode,
 				       struct ocfs2_xattr_info *xi,
 				       struct ocfs2_xattr_search *xs)
 {
-	u64 value_size;
 	struct ocfs2_xattr_entry *last;
 	int free, i;
 	size_t min_offs = xs->end - xs->base;
@@ -2394,13 +2999,7 @@ static int ocfs2_xattr_can_be_in_inode(struct inode *inode,
 
 	BUG_ON(!xs->not_found);
 
-	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE)
-		value_size = OCFS2_XATTR_ROOT_SIZE;
-	else
-		value_size = OCFS2_XATTR_SIZE(xi->value_len);
-
-	if (free >= sizeof(struct ocfs2_xattr_entry) +
-		   OCFS2_XATTR_SIZE(strlen(xi->name)) + value_size)
+	if (free >= (sizeof(struct ocfs2_xattr_entry) + namevalue_size_xi(xi)))
 		return 1;
 
 	return 0;
@@ -2424,7 +3023,7 @@ static int ocfs2_calc_xattr_set_need(struct inode *inode,
 	char *base = NULL;
 	int name_offset, name_len = 0;
 	u32 new_clusters = ocfs2_clusters_for_bytes(inode->i_sb,
-						    xi->value_len);
+						    xi->xi_value_len);
 	u64 value_size;
 
 	/*
@@ -2432,14 +3031,14 @@ static int ocfs2_calc_xattr_set_need(struct inode *inode,
 	 * No matter whether we replace an old one or add a new one,
 	 * we need this for writing.
 	 */
-	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE)
+	if (xi->xi_value_len > OCFS2_XATTR_INLINE_SIZE)
 		credits += new_clusters *
 			   ocfs2_clusters_to_blocks(inode->i_sb, 1);
 
 	if (xis->not_found && xbs->not_found) {
 		credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
 
-		if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
+		if (xi->xi_value_len > OCFS2_XATTR_INLINE_SIZE) {
 			clusters_add += new_clusters;
 			credits += ocfs2_calc_extend_credits(inode->i_sb,
 							&def_xv.xv.xr_list,
@@ -2484,7 +3083,7 @@ static int ocfs2_calc_xattr_set_need(struct inode *inode,
 	 * The credits for removing the value tree will be extended
 	 * by ocfs2_remove_extent itself.
 	 */
-	if (!xi->value) {
+	if (!xi->xi_value) {
 		if (!ocfs2_xattr_is_local(xe))
 			credits += ocfs2_remove_extent_credits(inode->i_sb);
 
@@ -2514,7 +3113,7 @@ static int ocfs2_calc_xattr_set_need(struct inode *inode,
 		}
 	}
 
-	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
+	if (xi->xi_value_len > OCFS2_XATTR_INLINE_SIZE) {
 		/* the new values will be stored outside. */
 		u32 old_clusters = 0;
 
@@ -2547,9 +3146,10 @@ static int ocfs2_calc_xattr_set_need(struct inode *inode,
 		 * value, we don't need any allocation, otherwise we have
 		 * to guess metadata allocation.
 		 */
-		if ((ocfs2_xattr_is_local(xe) && value_size >= xi->value_len) ||
+		if ((ocfs2_xattr_is_local(xe) &&
+		     (value_size >= xi->xi_value_len)) ||
 		    (!ocfs2_xattr_is_local(xe) &&
-		     OCFS2_XATTR_ROOT_SIZE >= xi->value_len))
+		     OCFS2_XATTR_ROOT_SIZE >= xi->xi_value_len))
 			goto out;
 	}
 
@@ -2639,7 +3239,7 @@ static int ocfs2_init_xattr_set_ctxt(struct inode *inode,
 
 	meta_add += extra_meta;
 	mlog(0, "Set xattr %s, reserve meta blocks = %d, clusters = %d, "
-	     "credits = %d\n", xi->name, meta_add, clusters_add, *credits);
+	     "credits = %d\n", xi->xi_name, meta_add, clusters_add, *credits);
 
 	if (meta_add) {
 		ret = ocfs2_reserve_new_metadata_blocks(osb, meta_add,
@@ -2679,7 +3279,7 @@ static int __ocfs2_xattr_set_handle(struct inode *inode,
 {
 	int ret = 0, credits, old_found;
 
-	if (!xi->value) {
+	if (!xi->xi_value) {
 		/* Remove existing extended attribute */
 		if (!xis->not_found)
 			ret = ocfs2_xattr_ibody_set(inode, xi, xis, ctxt);
@@ -2693,8 +3293,8 @@ static int __ocfs2_xattr_set_handle(struct inode *inode,
 			 * If succeed and that extended attribute existing in
 			 * external block, then we will remove it.
 			 */
-			xi->value = NULL;
-			xi->value_len = 0;
+			xi->xi_value = NULL;
+			xi->xi_value_len = 0;
 
 			old_found = xis->not_found;
 			xis->not_found = -ENODATA;
@@ -2722,8 +3322,8 @@ static int __ocfs2_xattr_set_handle(struct inode *inode,
 		} else if (ret == -ENOSPC) {
 			if (di->i_xattr_loc && !xbs->xattr_bh) {
 				ret = ocfs2_xattr_block_find(inode,
-							     xi->name_index,
-							     xi->name, xbs);
+							     xi->xi_name_index,
+							     xi->xi_name, xbs);
 				if (ret)
 					goto out;
 
@@ -2762,8 +3362,8 @@ static int __ocfs2_xattr_set_handle(struct inode *inode,
 				 * If succeed and that extended attribute
 				 * existing in inode, we will remove it.
 				 */
-				xi->value = NULL;
-				xi->value_len = 0;
+				xi->xi_value = NULL;
+				xi->xi_value_len = 0;
 				xbs->not_found = -ENODATA;
 				ret = ocfs2_calc_xattr_set_need(inode,
 								di,
@@ -2829,10 +3429,11 @@ int ocfs2_xattr_set_handle(handle_t *handle,
 	int ret;
 
 	struct ocfs2_xattr_info xi = {
-		.name_index = name_index,
-		.name = name,
-		.value = value,
-		.value_len = value_len,
+		.xi_name_index = name_index,
+		.xi_name = name,
+		.xi_name_len = strlen(name),
+		.xi_value = value,
+		.xi_value_len = value_len,
 	};
 
 	struct ocfs2_xattr_search xis = {
@@ -2912,10 +3513,11 @@ int ocfs2_xattr_set(struct inode *inode,
 	struct ocfs2_refcount_tree *ref_tree = NULL;
 
 	struct ocfs2_xattr_info xi = {
-		.name_index = name_index,
-		.name = name,
-		.value = value,
-		.value_len = value_len,
+		.xi_name_index = name_index,
+		.xi_name = name,
+		.xi_name_len = strlen(name),
+		.xi_value = value,
+		.xi_value_len = value_len,
 	};
 
 	struct ocfs2_xattr_search xis = {
@@ -3759,7 +4361,7 @@ static int ocfs2_defrag_xattr_bucket(struct inode *inode,
 				     struct ocfs2_xattr_bucket *bucket)
 {
 	int ret, i;
-	size_t end, offset, len, value_len;
+	size_t end, offset, len;
 	struct ocfs2_xattr_header *xh;
 	char *entries, *buf, *bucket_buf = NULL;
 	u64 blkno = bucket_blkno(bucket);
@@ -3813,12 +4415,7 @@ static int ocfs2_defrag_xattr_bucket(struct inode *inode,
 	end = OCFS2_XATTR_BUCKET_SIZE;
 	for (i = 0; i < le16_to_cpu(xh->xh_count); i++, xe++) {
 		offset = le16_to_cpu(xe->xe_name_offset);
-		if (ocfs2_xattr_is_local(xe))
-			value_len = OCFS2_XATTR_SIZE(
-					le64_to_cpu(xe->xe_value_size));
-		else
-			value_len = OCFS2_XATTR_ROOT_SIZE;
-		len = OCFS2_XATTR_SIZE(xe->xe_name_len) + value_len;
+		len = namevalue_size_xe(xe);
 
 		/*
 		 * We must make sure that the name/value pair
@@ -4007,7 +4604,7 @@ static int ocfs2_divide_xattr_bucket(struct inode *inode,
 				    int new_bucket_head)
 {
 	int ret, i;
-	int count, start, len, name_value_len = 0, xe_len, name_offset = 0;
+	int count, start, len, name_value_len = 0, name_offset = 0;
 	struct ocfs2_xattr_bucket *s_bucket = NULL, *t_bucket = NULL;
 	struct ocfs2_xattr_header *xh;
 	struct ocfs2_xattr_entry *xe;
@@ -4098,13 +4695,7 @@ static int ocfs2_divide_xattr_bucket(struct inode *inode,
 	name_value_len = 0;
 	for (i = 0; i < start; i++) {
 		xe = &xh->xh_entries[i];
-		xe_len = OCFS2_XATTR_SIZE(xe->xe_name_len);
-		if (ocfs2_xattr_is_local(xe))
-			xe_len +=
-			   OCFS2_XATTR_SIZE(le64_to_cpu(xe->xe_value_size));
-		else
-			xe_len += OCFS2_XATTR_ROOT_SIZE;
-		name_value_len += xe_len;
+		name_value_len += namevalue_size_xe(xe);
 		if (le16_to_cpu(xe->xe_name_offset) < name_offset)
 			name_offset = le16_to_cpu(xe->xe_name_offset);
 	}
@@ -4134,12 +4725,6 @@ static int ocfs2_divide_xattr_bucket(struct inode *inode,
 	xh->xh_free_start = cpu_to_le16(OCFS2_XATTR_BUCKET_SIZE);
 	for (i = 0; i < le16_to_cpu(xh->xh_count); i++) {
 		xe = &xh->xh_entries[i];
-		xe_len = OCFS2_XATTR_SIZE(xe->xe_name_len);
-		if (ocfs2_xattr_is_local(xe))
-			xe_len +=
-			   OCFS2_XATTR_SIZE(le64_to_cpu(xe->xe_value_size));
-		else
-			xe_len += OCFS2_XATTR_ROOT_SIZE;
 		if (le16_to_cpu(xe->xe_name_offset) <
 		    le16_to_cpu(xh->xh_free_start))
 			xh->xh_free_start = xe->xe_name_offset;
@@ -4751,195 +5336,6 @@ static inline char *ocfs2_xattr_bucket_get_val(struct inode *inode,
 }
 
 /*
- * Handle the normal xattr set, including replace, delete and new.
- *
- * Note: "local" indicates the real data's locality. So we can't
- * just its bucket locality by its length.
- */
-static void ocfs2_xattr_set_entry_normal(struct inode *inode,
-					 struct ocfs2_xattr_info *xi,
-					 struct ocfs2_xattr_search *xs,
-					 u32 name_hash,
-					 int local)
-{
-	struct ocfs2_xattr_entry *last, *xe;
-	int name_len = strlen(xi->name);
-	struct ocfs2_xattr_header *xh = xs->header;
-	u16 count = le16_to_cpu(xh->xh_count), start;
-	size_t blocksize = inode->i_sb->s_blocksize;
-	char *val;
-	size_t offs, size, new_size;
-
-	last = &xh->xh_entries[count];
-	if (!xs->not_found) {
-		xe = xs->here;
-		offs = le16_to_cpu(xe->xe_name_offset);
-		if (ocfs2_xattr_is_local(xe))
-			size = OCFS2_XATTR_SIZE(name_len) +
-			OCFS2_XATTR_SIZE(le64_to_cpu(xe->xe_value_size));
-		else
-			size = OCFS2_XATTR_SIZE(name_len) +
-			OCFS2_XATTR_SIZE(OCFS2_XATTR_ROOT_SIZE);
-
-		/*
-		 * If the new value will be stored outside, xi->value has been
-		 * initalized as an empty ocfs2_xattr_value_root, and the same
-		 * goes with xi->value_len, so we can set new_size safely here.
-		 * See ocfs2_xattr_set_in_bucket.
-		 */
-		new_size = OCFS2_XATTR_SIZE(name_len) +
-			   OCFS2_XATTR_SIZE(xi->value_len);
-
-		le16_add_cpu(&xh->xh_name_value_len, -size);
-		if (xi->value) {
-			if (new_size > size)
-				goto set_new_name_value;
-
-			/* Now replace the old value with new one. */
-			if (local)
-				xe->xe_value_size = cpu_to_le64(xi->value_len);
-			else
-				xe->xe_value_size = 0;
-
-			val = ocfs2_xattr_bucket_get_val(inode,
-							 xs->bucket, offs);
-			memset(val + OCFS2_XATTR_SIZE(name_len), 0,
-			       size - OCFS2_XATTR_SIZE(name_len));
-			if (OCFS2_XATTR_SIZE(xi->value_len) > 0)
-				memcpy(val + OCFS2_XATTR_SIZE(name_len),
-				       xi->value, xi->value_len);
-
-			le16_add_cpu(&xh->xh_name_value_len, new_size);
-			ocfs2_xattr_set_local(xe, local);
-			return;
-		} else {
-			/*
-			 * Remove the old entry if there is more than one.
-			 * We don't remove the last entry so that we can
-			 * use it to indicate the hash value of the empty
-			 * bucket.
-			 */
-			last -= 1;
-			le16_add_cpu(&xh->xh_count, -1);
-			if (xh->xh_count) {
-				memmove(xe, xe + 1,
-					(void *)last - (void *)xe);
-				memset(last, 0,
-				       sizeof(struct ocfs2_xattr_entry));
-			} else
-				xh->xh_free_start =
-					cpu_to_le16(OCFS2_XATTR_BUCKET_SIZE);
-
-			return;
-		}
-	} else {
-		/* find a new entry for insert. */
-		int low = 0, high = count - 1, tmp;
-		struct ocfs2_xattr_entry *tmp_xe;
-
-		while (low <= high && count) {
-			tmp = (low + high) / 2;
-			tmp_xe = &xh->xh_entries[tmp];
-
-			if (name_hash > le32_to_cpu(tmp_xe->xe_name_hash))
-				low = tmp + 1;
-			else if (name_hash <
-				 le32_to_cpu(tmp_xe->xe_name_hash))
-				high = tmp - 1;
-			else {
-				low = tmp;
-				break;
-			}
-		}
-
-		xe = &xh->xh_entries[low];
-		if (low != count)
-			memmove(xe + 1, xe, (void *)last - (void *)xe);
-
-		le16_add_cpu(&xh->xh_count, 1);
-		memset(xe, 0, sizeof(struct ocfs2_xattr_entry));
-		xe->xe_name_hash = cpu_to_le32(name_hash);
-		xe->xe_name_len = name_len;
-		ocfs2_xattr_set_type(xe, xi->name_index);
-	}
-
-set_new_name_value:
-	/* Insert the new name+value. */
-	size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_SIZE(xi->value_len);
-
-	/*
-	 * We must make sure that the name/value pair
-	 * exists in the same block.
-	 */
-	offs = le16_to_cpu(xh->xh_free_start);
-	start = offs - size;
-
-	if (start >> inode->i_sb->s_blocksize_bits !=
-	    (offs - 1) >> inode->i_sb->s_blocksize_bits) {
-		offs = offs - offs % blocksize;
-		xh->xh_free_start = cpu_to_le16(offs);
-	}
-
-	val = ocfs2_xattr_bucket_get_val(inode, xs->bucket, offs - size);
-	xe->xe_name_offset = cpu_to_le16(offs - size);
-
-	memset(val, 0, size);
-	memcpy(val, xi->name, name_len);
-	memcpy(val + OCFS2_XATTR_SIZE(name_len), xi->value, xi->value_len);
-
-	xe->xe_value_size = cpu_to_le64(xi->value_len);
-	ocfs2_xattr_set_local(xe, local);
-	xs->here = xe;
-	le16_add_cpu(&xh->xh_free_start, -size);
-	le16_add_cpu(&xh->xh_name_value_len, size);
-
-	return;
-}
-
-/*
- * Set the xattr entry in the specified bucket.
- * The bucket is indicated by xs->bucket and it should have the enough
- * space for the xattr insertion.
- */
-static int ocfs2_xattr_set_entry_in_bucket(struct inode *inode,
-					   handle_t *handle,
-					   struct ocfs2_xattr_info *xi,
-					   struct ocfs2_xattr_search *xs,
-					   u32 name_hash,
-					   int local)
-{
-	int ret;
-	u64 blkno;
-
-	mlog(0, "Set xattr entry len = %lu index = %d in bucket %llu\n",
-	     (unsigned long)xi->value_len, xi->name_index,
-	     (unsigned long long)bucket_blkno(xs->bucket));
-
-	if (!xs->bucket->bu_bhs[1]) {
-		blkno = bucket_blkno(xs->bucket);
-		ocfs2_xattr_bucket_relse(xs->bucket);
-		ret = ocfs2_read_xattr_bucket(xs->bucket, blkno);
-		if (ret) {
-			mlog_errno(ret);
-			goto out;
-		}
-	}
-
-	ret = ocfs2_xattr_bucket_journal_access(handle, xs->bucket,
-						OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret < 0) {
-		mlog_errno(ret);
-		goto out;
-	}
-
-	ocfs2_xattr_set_entry_normal(inode, xi, xs, name_hash, local);
-	ocfs2_xattr_bucket_journal_dirty(handle, xs->bucket);
-
-out:
-	return ret;
-}
-
-/*
  * Truncate the specified xe_off entry in xattr bucket.
  * bucket is indicated by header_bh and len is the new length.
  * Both the ocfs2_xattr_value_root and the entry will be updated here.
@@ -5009,66 +5405,6 @@ out:
 	return ret;
 }
 
-static int ocfs2_xattr_bucket_value_truncate_xs(struct inode *inode,
-					struct ocfs2_xattr_search *xs,
-					int len,
-					struct ocfs2_xattr_set_ctxt *ctxt)
-{
-	int ret, offset;
-	struct ocfs2_xattr_entry *xe = xs->here;
-	struct ocfs2_xattr_header *xh = (struct ocfs2_xattr_header *)xs->base;
-
-	BUG_ON(!xs->bucket->bu_bhs[0] || !xe || ocfs2_xattr_is_local(xe));
-
-	offset = xe - xh->xh_entries;
-	ret = ocfs2_xattr_bucket_value_truncate(inode, xs->bucket,
-						offset, len, ctxt);
-	if (ret)
-		mlog_errno(ret);
-
-	return ret;
-}
-
-static int ocfs2_xattr_bucket_set_value_outside(struct inode *inode,
-						handle_t *handle,
-						struct ocfs2_xattr_search *xs,
-						char *val,
-						int value_len)
-{
-	int ret, offset, block_off;
-	struct ocfs2_xattr_value_root *xv;
-	struct ocfs2_xattr_entry *xe = xs->here;
-	struct ocfs2_xattr_header *xh = bucket_xh(xs->bucket);
-	void *base;
-	struct ocfs2_xattr_value_buf vb = {
-		.vb_access = ocfs2_journal_access,
-	};
-
-	BUG_ON(!xs->base || !xe || ocfs2_xattr_is_local(xe));
-
-	ret = ocfs2_xattr_bucket_get_name_value(inode->i_sb, xh,
-						xe - xh->xh_entries,
-						&block_off,
-						&offset);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
-
-	base = bucket_block(xs->bucket, block_off);
-	xv = (struct ocfs2_xattr_value_root *)(base + offset +
-		 OCFS2_XATTR_SIZE(xe->xe_name_len));
-
-	vb.vb_xv = xv;
-	vb.vb_bh = xs->bucket->bu_bhs[block_off];
-	ret = __ocfs2_xattr_set_value_outside(inode, handle,
-					      &vb, val, value_len);
-	if (ret)
-		mlog_errno(ret);
-out:
-	return ret;
-}
-
 static int ocfs2_rm_xattr_cluster(struct inode *inode,
 				  struct buffer_head *root_bh,
 				  u64 blkno,
@@ -5167,128 +5503,6 @@ out:
 	return ret;
 }
 
-static void ocfs2_xattr_bucket_remove_xs(struct inode *inode,
-					 handle_t *handle,
-					 struct ocfs2_xattr_search *xs)
-{
-	struct ocfs2_xattr_header *xh = bucket_xh(xs->bucket);
-	struct ocfs2_xattr_entry *last = &xh->xh_entries[
-						le16_to_cpu(xh->xh_count) - 1];
-	int ret = 0;
-
-	ret = ocfs2_xattr_bucket_journal_access(handle, xs->bucket,
-						OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret) {
-		mlog_errno(ret);
-		return;
-	}
-
-	/* Remove the old entry. */
-	memmove(xs->here, xs->here + 1,
-		(void *)last - (void *)xs->here);
-	memset(last, 0, sizeof(struct ocfs2_xattr_entry));
-	le16_add_cpu(&xh->xh_count, -1);
-
-	ocfs2_xattr_bucket_journal_dirty(handle, xs->bucket);
-}
-
-/*
- * Set the xattr name/value in the bucket specified in xs.
- *
- * As the new value in xi may be stored in the bucket or in an outside cluster,
- * we divide the whole process into 3 steps:
- * 1. insert name/value in the bucket(ocfs2_xattr_set_entry_in_bucket)
- * 2. truncate of the outside cluster(ocfs2_xattr_bucket_value_truncate_xs)
- * 3. Set the value to the outside cluster(ocfs2_xattr_bucket_set_value_outside)
- * 4. If the clusters for the new outside value can't be allocated, we need
- *    to free the xattr we allocated in set.
- */
-static int ocfs2_xattr_set_in_bucket(struct inode *inode,
-				     struct ocfs2_xattr_info *xi,
-				     struct ocfs2_xattr_search *xs,
-				     struct ocfs2_xattr_set_ctxt *ctxt)
-{
-	int ret, local = 1;
-	size_t value_len;
-	char *val = (char *)xi->value;
-	struct ocfs2_xattr_entry *xe = xs->here;
-	u32 name_hash = ocfs2_xattr_name_hash(inode, xi->name,
-					      strlen(xi->name));
-
-	if (!xs->not_found && !ocfs2_xattr_is_local(xe)) {
-		/*
-		 * We need to truncate the xattr storage first.
-		 *
-		 * If both the old and new value are stored to
-		 * outside block, we only need to truncate
-		 * the storage and then set the value outside.
-		 *
-		 * If the new value should be stored within block,
-		 * we should free all the outside block first and
-		 * the modification to the xattr block will be done
-		 * by following steps.
-		 */
-		if (xi->value_len > OCFS2_XATTR_INLINE_SIZE)
-			value_len = xi->value_len;
-		else
-			value_len = 0;
-
-		ret = ocfs2_xattr_bucket_value_truncate_xs(inode, xs,
-							   value_len,
-							   ctxt);
-		if (ret)
-			goto out;
-
-		if (value_len)
-			goto set_value_outside;
-	}
-
-	value_len = xi->value_len;
-	/* So we have to handle the inside block change now. */
-	if (value_len > OCFS2_XATTR_INLINE_SIZE) {
-		/*
-		 * If the new value will be stored outside of block,
-		 * initalize a new empty value root and insert it first.
-		 */
-		local = 0;
-		xi->value = &def_xv;
-		xi->value_len = OCFS2_XATTR_ROOT_SIZE;
-	}
-
-	ret = ocfs2_xattr_set_entry_in_bucket(inode, ctxt->handle, xi, xs,
-					      name_hash, local);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
-
-	if (value_len <= OCFS2_XATTR_INLINE_SIZE)
-		goto out;
-
-	/* allocate the space now for the outside block storage. */
-	ret = ocfs2_xattr_bucket_value_truncate_xs(inode, xs,
-						   value_len, ctxt);
-	if (ret) {
-		mlog_errno(ret);
-
-		if (xs->not_found) {
-			/*
-			 * We can't allocate enough clusters for outside
-			 * storage and we have allocated xattr already,
-			 * so need to remove it.
-			 */
-			ocfs2_xattr_bucket_remove_xs(inode, ctxt->handle, xs);
-		}
-		goto out;
-	}
-
-set_value_outside:
-	ret = ocfs2_xattr_bucket_set_value_outside(inode, ctxt->handle,
-						   xs, val, value_len);
-out:
-	return ret;
-}
-
 /*
  * check whether the xattr bucket is filled up with the same hash value.
  * If we want to insert the xattr with the same hash, return -ENOSPC.
@@ -5317,156 +5531,116 @@ static int ocfs2_check_xattr_bucket_collision(struct inode *inode,
 	return 0;
 }
 
-static int ocfs2_xattr_set_entry_index_block(struct inode *inode,
-					     struct ocfs2_xattr_info *xi,
-					     struct ocfs2_xattr_search *xs,
-					     struct ocfs2_xattr_set_ctxt *ctxt)
+/*
+ * Try to set the entry in the current bucket.  If we fail, the caller
+ * will handle getting us another bucket.
+ */
+static int ocfs2_xattr_set_entry_bucket(struct inode *inode,
+					struct ocfs2_xattr_info *xi,
+					struct ocfs2_xattr_search *xs,
+					struct ocfs2_xattr_set_ctxt *ctxt)
 {
-	struct ocfs2_xattr_header *xh;
-	struct ocfs2_xattr_entry *xe;
-	u16 count, header_size, xh_free_start;
-	int free, max_free, need, old;
-	size_t value_size = 0, name_len = strlen(xi->name);
-	size_t blocksize = inode->i_sb->s_blocksize;
-	int ret, allocation = 0;
-
-	mlog_entry("Set xattr %s in xattr index block\n", xi->name);
-
-try_again:
-	xh = xs->header;
-	count = le16_to_cpu(xh->xh_count);
-	xh_free_start = le16_to_cpu(xh->xh_free_start);
-	header_size = sizeof(struct ocfs2_xattr_header) +
-			count * sizeof(struct ocfs2_xattr_entry);
-	max_free = OCFS2_XATTR_BUCKET_SIZE - header_size -
-		le16_to_cpu(xh->xh_name_value_len) - OCFS2_XATTR_HEADER_GAP;
-
-	mlog_bug_on_msg(header_size > blocksize, "bucket %llu has header size "
-			"of %u which exceed block size\n",
-			(unsigned long long)bucket_blkno(xs->bucket),
-			header_size);
+	int ret;
+	struct ocfs2_xa_loc loc;
 
-	if (xi->value && xi->value_len > OCFS2_XATTR_INLINE_SIZE)
-		value_size = OCFS2_XATTR_ROOT_SIZE;
-	else if (xi->value)
-		value_size = OCFS2_XATTR_SIZE(xi->value_len);
+	mlog_entry("Set xattr %s in xattr bucket\n", xi->xi_name);
 
-	if (xs->not_found)
-		need = sizeof(struct ocfs2_xattr_entry) +
-			OCFS2_XATTR_SIZE(name_len) + value_size;
-	else {
-		need = value_size + OCFS2_XATTR_SIZE(name_len);
+	ocfs2_init_xattr_bucket_xa_loc(&loc, xs->bucket,
+				       xs->not_found ? NULL : xs->here);
+	ret = ocfs2_xa_set(&loc, xi, ctxt);
+	if (!ret) {
+		xs->here = loc.xl_entry;
+		goto out;
+	}
+	if (ret != -ENOSPC) {
+		mlog_errno(ret);
+		goto out;
+	}
 
-		/*
-		 * We only replace the old value if the new length is smaller
-		 * than the old one. Otherwise we will allocate new space in the
-		 * bucket to store it.
-		 */
-		xe = xs->here;
-		if (ocfs2_xattr_is_local(xe))
-			old = OCFS2_XATTR_SIZE(le64_to_cpu(xe->xe_value_size));
-		else
-			old = OCFS2_XATTR_SIZE(OCFS2_XATTR_ROOT_SIZE);
+	/* Ok, we need space.  Let's try defragmenting the bucket. */
+	ret = ocfs2_defrag_xattr_bucket(inode, ctxt->handle,
+					xs->bucket);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
 
-		if (old >= value_size)
-			need = 0;
+	ret = ocfs2_xa_set(&loc, xi, ctxt);
+	if (!ret) {
+		xs->here = loc.xl_entry;
+		goto out;
 	}
+	if (ret != -ENOSPC)
+		mlog_errno(ret);
 
-	free = xh_free_start - header_size - OCFS2_XATTR_HEADER_GAP;
-	/*
-	 * We need to make sure the new name/value pair
-	 * can exist in the same block.
-	 */
-	if (xh_free_start % blocksize < need)
-		free -= xh_free_start % blocksize;
-
-	mlog(0, "xs->not_found = %d, in xattr bucket %llu: free = %d, "
-	     "need = %d, max_free = %d, xh_free_start = %u, xh_name_value_len ="
-	     " %u\n", xs->not_found,
-	     (unsigned long long)bucket_blkno(xs->bucket),
-	     free, need, max_free, le16_to_cpu(xh->xh_free_start),
-	     le16_to_cpu(xh->xh_name_value_len));
-
-	if (free < need ||
-	    (xs->not_found &&
-	     count == ocfs2_xattr_max_xe_in_bucket(inode->i_sb))) {
-		if (need <= max_free &&
-		    count < ocfs2_xattr_max_xe_in_bucket(inode->i_sb)) {
-			/*
-			 * We can create the space by defragment. Since only the
-			 * name/value will be moved, the xe shouldn't be changed
-			 * in xs.
-			 */
-			ret = ocfs2_defrag_xattr_bucket(inode, ctxt->handle,
-							xs->bucket);
-			if (ret) {
-				mlog_errno(ret);
-				goto out;
-			}
 
-			xh_free_start = le16_to_cpu(xh->xh_free_start);
-			free = xh_free_start - header_size
-				- OCFS2_XATTR_HEADER_GAP;
-			if (xh_free_start % blocksize < need)
-				free -= xh_free_start % blocksize;
+out:
+	mlog_exit(ret);
+	return ret;
+}
 
-			if (free >= need)
-				goto xattr_set;
+static int ocfs2_xattr_set_entry_index_block(struct inode *inode,
+					     struct ocfs2_xattr_info *xi,
+					     struct ocfs2_xattr_search *xs,
+					     struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int ret;
 
-			mlog(0, "Can't get enough space for xattr insert by "
-			     "defragment. Need %u bytes, but we have %d, so "
-			     "allocate new bucket for it.\n", need, free);
-		}
+	mlog_entry("Set xattr %s in xattr index block\n", xi->xi_name);
 
-		/*
-		 * We have to add new buckets or clusters and one
-		 * allocation should leave us enough space for insert.
-		 */
-		BUG_ON(allocation);
+	ret = ocfs2_xattr_set_entry_bucket(inode, xi, xs, ctxt);
+	if (!ret)
+		goto out;
+	if (ret != -ENOSPC) {
+		mlog_errno(ret);
+		goto out;
+	}
 
-		/*
-		 * We do not allow for overlapping ranges between buckets. And
-		 * the maximum number of collisions we will allow for then is
-		 * one bucket's worth, so check it here whether we need to
-		 * add a new bucket for the insert.
-		 */
-		ret = ocfs2_check_xattr_bucket_collision(inode,
-							 xs->bucket,
-							 xi->name);
-		if (ret) {
-			mlog_errno(ret);
-			goto out;
-		}
+	/* Ack, need more space.  Let's try to get another bucket! */
 
-		ret = ocfs2_add_new_xattr_bucket(inode,
-						 xs->xattr_bh,
+	/*
+	 * We do not allow for overlapping ranges between buckets. And
+	 * the maximum number of collisions we will allow for then is
+	 * one bucket's worth, so check it here whether we need to
+	 * add a new bucket for the insert.
+	 */
+	ret = ocfs2_check_xattr_bucket_collision(inode,
 						 xs->bucket,
-						 ctxt);
-		if (ret) {
-			mlog_errno(ret);
-			goto out;
-		}
+						 xi->xi_name);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
 
-		/*
-		 * ocfs2_add_new_xattr_bucket() will have updated
-		 * xs->bucket if it moved, but it will not have updated
-		 * any of the other search fields.  Thus, we drop it and
-		 * re-search.  Everything should be cached, so it'll be
-		 * quick.
-		 */
-		ocfs2_xattr_bucket_relse(xs->bucket);
-		ret = ocfs2_xattr_index_block_find(inode, xs->xattr_bh,
-						   xi->name_index,
-						   xi->name, xs);
-		if (ret && ret != -ENODATA)
-			goto out;
-		xs->not_found = ret;
-		allocation = 1;
-		goto try_again;
+	ret = ocfs2_add_new_xattr_bucket(inode,
+					 xs->xattr_bh,
+					 xs->bucket,
+					 ctxt);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
 	}
 
-xattr_set:
-	ret = ocfs2_xattr_set_in_bucket(inode, xi, xs, ctxt);
+	/*
+	 * ocfs2_add_new_xattr_bucket() will have updated
+	 * xs->bucket if it moved, but it will not have updated
+	 * any of the other search fields.  Thus, we drop it and
+	 * re-search.  Everything should be cached, so it'll be
+	 * quick.
+	 */
+	ocfs2_xattr_bucket_relse(xs->bucket);
+	ret = ocfs2_xattr_index_block_find(inode, xs->xattr_bh,
+					   xi->xi_name_index,
+					   xi->xi_name, xs);
+	if (ret && ret != -ENODATA)
+		goto out;
+	xs->not_found = ret;
+
+	/* Ok, we have a new bucket, let's try again */
+	ret = ocfs2_xattr_set_entry_bucket(inode, xi, xs, ctxt);
+	if (ret && (ret != -ENOSPC))
+		mlog_errno(ret);
+
 out:
 	mlog_exit(ret);
 	return ret;
@@ -5678,7 +5852,7 @@ static int ocfs2_prepare_refcount_xattr(struct inode *inode,
 	 * refcount tree, and make the original extent become 3. So we will need
 	 * 2 * cluster more extent recs at most.
 	 */
-	if (!xi->value || xi->value_len <= OCFS2_XATTR_INLINE_SIZE) {
+	if (!xi->xi_value || xi->xi_value_len <= OCFS2_XATTR_INLINE_SIZE) {
 
 		ret = ocfs2_refcounted_xattr_delete_need(inode,
 							 &(*ref_tree)->rf_ci,
@@ -6354,9 +6528,11 @@ static int ocfs2_create_empty_xattr_block(struct inode *inode,
 					  int indexed)
 {
 	int ret;
-	handle_t *handle;
 	struct ocfs2_alloc_context *meta_ac;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct ocfs2_xattr_set_ctxt ctxt = {
+		.meta_ac = meta_ac,
+	};
 
 	ret = ocfs2_reserve_new_metadata_blocks(osb, 1, &meta_ac);
 	if (ret < 0) {
@@ -6364,21 +6540,21 @@ static int ocfs2_create_empty_xattr_block(struct inode *inode,
 		return ret;
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_XATTR_BLOCK_CREATE_CREDITS);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
+	ctxt.handle = ocfs2_start_trans(osb, OCFS2_XATTR_BLOCK_CREATE_CREDITS);
+	if (IS_ERR(ctxt.handle)) {
+		ret = PTR_ERR(ctxt.handle);
 		mlog_errno(ret);
 		goto out;
 	}
 
 	mlog(0, "create new xattr block for inode %llu, index = %d\n",
 	     (unsigned long long)fe_bh->b_blocknr, indexed);
-	ret = ocfs2_create_xattr_block(handle, inode, fe_bh,
-				       meta_ac, ret_bh, indexed);
+	ret = ocfs2_create_xattr_block(inode, fe_bh, &ctxt, indexed,
+				       ret_bh);
 	if (ret)
 		mlog_errno(ret);
 
-	ocfs2_commit_trans(osb, handle);
+	ocfs2_commit_trans(osb, ctxt.handle);
 out:
 	ocfs2_free_alloc_context(meta_ac);
 	return ret;
diff --git a/fs/omfs/inode.c b/fs/omfs/inode.c
index f3b7c1541f3a..75d9b5ba1d45 100644
--- a/fs/omfs/inode.c
+++ b/fs/omfs/inode.c
@@ -11,6 +11,7 @@
 #include <linux/parser.h>
 #include <linux/buffer_head.h>
 #include <linux/vmalloc.h>
+#include <linux/writeback.h>
 #include <linux/crc-itu-t.h>
 #include "omfs.h"
 
@@ -89,7 +90,7 @@ static void omfs_update_checksums(struct omfs_inode *oi)
 	oi->i_head.h_check_xor = xor;
 }
 
-static int omfs_write_inode(struct inode *inode, int wait)
+static int __omfs_write_inode(struct inode *inode, int wait)
 {
 	struct omfs_inode *oi;
 	struct omfs_sb_info *sbi = OMFS_SB(inode->i_sb);
@@ -162,9 +163,14 @@ out:
 	return ret;
 }
 
+static int omfs_write_inode(struct inode *inode, struct writeback_control *wbc)
+{
+	return __omfs_write_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
+}
+
 int omfs_sync_inode(struct inode *inode)
 {
-	return omfs_write_inode(inode, 1);
+	return __omfs_write_inode(inode, 1);
 }
 
 /*
diff --git a/fs/open.c b/fs/open.c
index 040cef72bc00..e17f54454b50 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -8,7 +8,6 @@
 #include <linux/mm.h>
 #include <linux/file.h>
 #include <linux/fdtable.h>
-#include <linux/quotaops.h>
 #include <linux/fsnotify.h>
 #include <linux/module.h>
 #include <linux/slab.h>
@@ -271,17 +270,15 @@ static long do_sys_truncate(const char __user *pathname, loff_t length)
 	 * Make sure that there are no leases.  get_write_access() protects
 	 * against the truncate racing with a lease-granting setlease().
 	 */
-	error = break_lease(inode, FMODE_WRITE);
+	error = break_lease(inode, O_WRONLY);
 	if (error)
 		goto put_write_and_out;
 
 	error = locks_verify_truncate(inode, NULL, length);
 	if (!error)
 		error = security_path_truncate(&path, length, 0);
-	if (!error) {
-		vfs_dq_init(inode);
+	if (!error)
 		error = do_truncate(path.dentry, length, 0, NULL);
-	}
 
 put_write_and_out:
 	put_write_access(inode);
diff --git a/fs/partitions/check.c b/fs/partitions/check.c
index 64bc8998ac9a..e8865c11777f 100644
--- a/fs/partitions/check.c
+++ b/fs/partitions/check.c
@@ -412,9 +412,10 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno,
 	pdev = part_to_dev(p);
 
 	p->start_sect = start;
-	p->alignment_offset = queue_sector_alignment_offset(disk->queue, start);
-	p->discard_alignment = queue_sector_discard_alignment(disk->queue,
-							      start);
+	p->alignment_offset =
+		queue_limit_alignment_offset(&disk->queue->limits, start);
+	p->discard_alignment =
+		queue_limit_discard_alignment(&disk->queue->limits, start);
 	p->nr_sects = len;
 	p->partno = partno;
 	p->policy = get_disk_ro(disk);
diff --git a/fs/pnode.c b/fs/pnode.c
index 8d5f392ec3d3..5cc564a83149 100644
--- a/fs/pnode.c
+++ b/fs/pnode.c
@@ -86,7 +86,7 @@ static int do_make_slave(struct vfsmount *mnt)
 
 	/*
 	 * slave 'mnt' to a peer mount that has the
-	 * same root dentry. If none is available than
+	 * same root dentry. If none is available then
 	 * slave it to anything that is available.
 	 */
 	while ((peer_mnt = next_peer(peer_mnt)) != mnt &&
@@ -147,6 +147,11 @@ void change_mnt_propagation(struct vfsmount *mnt, int type)
  * get the next mount in the propagation tree.
  * @m: the mount seen last
  * @origin: the original mount from where the tree walk initiated
+ *
+ * Note that peer groups form contiguous segments of slave lists.
+ * We rely on that in get_source() to be able to find out if
+ * vfsmount found while iterating with propagation_next() is
+ * a peer of one we'd found earlier.
  */
 static struct vfsmount *propagation_next(struct vfsmount *m,
 					 struct vfsmount *origin)
@@ -186,10 +191,6 @@ static struct vfsmount *get_source(struct vfsmount *dest,
 {
 	struct vfsmount *p_last_src = NULL;
 	struct vfsmount *p_last_dest = NULL;
-	*type = CL_PROPAGATION;
-
-	if (IS_MNT_SHARED(dest))
-		*type |= CL_MAKE_SHARED;
 
 	while (last_dest != dest->mnt_master) {
 		p_last_dest = last_dest;
@@ -202,13 +203,18 @@ static struct vfsmount *get_source(struct vfsmount *dest,
 		do {
 			p_last_dest = next_peer(p_last_dest);
 		} while (IS_MNT_NEW(p_last_dest));
+		/* is that a peer of the earlier? */
+		if (dest == p_last_dest) {
+			*type = CL_MAKE_SHARED;
+			return p_last_src;
+		}
 	}
-
-	if (dest != p_last_dest) {
-		*type |= CL_SLAVE;
-		return last_src;
-	} else
-		return p_last_src;
+	/* slave of the earlier, then */
+	*type = CL_SLAVE;
+	/* beginning of peer group among the slaves? */
+	if (IS_MNT_SHARED(dest))
+		*type |= CL_MAKE_SHARED;
+	return last_src;
 }
 
 /*
diff --git a/fs/pnode.h b/fs/pnode.h
index 958665d662af..1ea4ae1efcd3 100644
--- a/fs/pnode.h
+++ b/fs/pnode.h
@@ -21,12 +21,11 @@
 #define CL_SLAVE     		0x02
 #define CL_COPY_ALL 		0x04
 #define CL_MAKE_SHARED 		0x08
-#define CL_PROPAGATION 		0x10
-#define CL_PRIVATE 		0x20
+#define CL_PRIVATE 		0x10
 
 static inline void set_mnt_shared(struct vfsmount *mnt)
 {
-	mnt->mnt_flags &= ~MNT_PNODE_MASK;
+	mnt->mnt_flags &= ~MNT_SHARED_MASK;
 	mnt->mnt_flags |= MNT_SHARED;
 }
 
diff --git a/fs/proc/array.c b/fs/proc/array.c
index 13b5d0708175..aa8637b81028 100644
--- a/fs/proc/array.c
+++ b/fs/proc/array.c
@@ -270,8 +270,10 @@ static inline void task_sig(struct seq_file *m, struct task_struct *p)
 		blocked = p->blocked;
 		collect_sigign_sigcatch(p, &ignored, &caught);
 		num_threads = atomic_read(&p->signal->count);
+		rcu_read_lock();  /* FIXME: is this correct? */
 		qsize = atomic_read(&__task_cred(p)->user->sigpending);
-		qlim = p->signal->rlim[RLIMIT_SIGPENDING].rlim_cur;
+		rcu_read_unlock();
+		qlim = task_rlimit(p, RLIMIT_SIGPENDING);
 		unlock_task_sighand(p, &flags);
 	}
 
@@ -418,7 +420,7 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns,
 		cutime = sig->cutime;
 		cstime = sig->cstime;
 		cgtime = sig->cgtime;
-		rsslim = sig->rlim[RLIMIT_RSS].rlim_cur;
+		rsslim = ACCESS_ONCE(sig->rlim[RLIMIT_RSS].rlim_cur);
 
 		/* add up live thread stats at the group level */
 		if (whole) {
diff --git a/fs/proc/base.c b/fs/proc/base.c
index e42bbd843ed1..a7310841c831 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -647,17 +647,11 @@ static int mounts_release(struct inode *inode, struct file *file)
 static unsigned mounts_poll(struct file *file, poll_table *wait)
 {
 	struct proc_mounts *p = file->private_data;
-	struct mnt_namespace *ns = p->ns;
 	unsigned res = POLLIN | POLLRDNORM;
 
-	poll_wait(file, &ns->poll, wait);
-
-	spin_lock(&vfsmount_lock);
-	if (p->event != ns->event) {
-		p->event = ns->event;
+	poll_wait(file, &p->ns->poll, wait);
+	if (mnt_had_events(p))
 		res |= POLLERR | POLLPRI;
-	}
-	spin_unlock(&vfsmount_lock);
 
 	return res;
 }
@@ -1095,8 +1089,12 @@ static ssize_t proc_loginuid_write(struct file * file, const char __user * buf,
 	if (!capable(CAP_AUDIT_CONTROL))
 		return -EPERM;
 
-	if (current != pid_task(proc_pid(inode), PIDTYPE_PID))
+	rcu_read_lock();
+	if (current != pid_task(proc_pid(inode), PIDTYPE_PID)) {
+		rcu_read_unlock();
 		return -EPERM;
+	}
+	rcu_read_unlock();
 
 	if (count >= PAGE_SIZE)
 		count = PAGE_SIZE - 1;
@@ -2369,16 +2367,30 @@ static void *proc_self_follow_link(struct dentry *dentry, struct nameidata *nd)
 {
 	struct pid_namespace *ns = dentry->d_sb->s_fs_info;
 	pid_t tgid = task_tgid_nr_ns(current, ns);
-	char tmp[PROC_NUMBUF];
-	if (!tgid)
-		return ERR_PTR(-ENOENT);
-	sprintf(tmp, "%d", task_tgid_nr_ns(current, ns));
-	return ERR_PTR(vfs_follow_link(nd,tmp));
+	char *name = ERR_PTR(-ENOENT);
+	if (tgid) {
+		name = __getname();
+		if (!name)
+			name = ERR_PTR(-ENOMEM);
+		else
+			sprintf(name, "%d", tgid);
+	}
+	nd_set_link(nd, name);
+	return NULL;
+}
+
+static void proc_self_put_link(struct dentry *dentry, struct nameidata *nd,
+				void *cookie)
+{
+	char *s = nd_get_link(nd);
+	if (!IS_ERR(s))
+		__putname(s);
 }
 
 static const struct inode_operations proc_self_inode_operations = {
 	.readlink	= proc_self_readlink,
 	.follow_link	= proc_self_follow_link,
+	.put_link	= proc_self_put_link,
 };
 
 /*
diff --git a/fs/proc/generic.c b/fs/proc/generic.c
index 480cb1065eec..08f4d71dacd7 100644
--- a/fs/proc/generic.c
+++ b/fs/proc/generic.c
@@ -291,19 +291,17 @@ static const struct inode_operations proc_file_inode_operations = {
  * returns the struct proc_dir_entry for "/proc/tty/driver", and
  * returns "serial" in residual.
  */
-static int xlate_proc_name(const char *name,
-			   struct proc_dir_entry **ret, const char **residual)
+static int __xlate_proc_name(const char *name, struct proc_dir_entry **ret,
+			     const char **residual)
 {
 	const char     		*cp = name, *next;
 	struct proc_dir_entry	*de;
 	int			len;
-	int 			rtn = 0;
 
 	de = *ret;
 	if (!de)
 		de = &proc_root;
 
-	spin_lock(&proc_subdir_lock);
 	while (1) {
 		next = strchr(cp, '/');
 		if (!next)
@@ -315,16 +313,25 @@ static int xlate_proc_name(const char *name,
 				break;
 		}
 		if (!de) {
-			rtn = -ENOENT;
-			goto out;
+			WARN(1, "name '%s'\n", name);
+			return -ENOENT;
 		}
 		cp += len + 1;
 	}
 	*residual = cp;
 	*ret = de;
-out:
+	return 0;
+}
+
+static int xlate_proc_name(const char *name, struct proc_dir_entry **ret,
+			   const char **residual)
+{
+	int rv;
+
+	spin_lock(&proc_subdir_lock);
+	rv = __xlate_proc_name(name, ret, residual);
 	spin_unlock(&proc_subdir_lock);
-	return rtn;
+	return rv;
 }
 
 static DEFINE_IDA(proc_inum_ida);
@@ -662,6 +669,7 @@ struct proc_dir_entry *proc_symlink(const char *name,
 	}
 	return ent;
 }
+EXPORT_SYMBOL(proc_symlink);
 
 struct proc_dir_entry *proc_mkdir_mode(const char *name, mode_t mode,
 		struct proc_dir_entry *parent)
@@ -700,6 +708,7 @@ struct proc_dir_entry *proc_mkdir(const char *name,
 {
 	return proc_mkdir_mode(name, S_IRUGO | S_IXUGO, parent);
 }
+EXPORT_SYMBOL(proc_mkdir);
 
 struct proc_dir_entry *create_proc_entry(const char *name, mode_t mode,
 					 struct proc_dir_entry *parent)
@@ -728,6 +737,7 @@ struct proc_dir_entry *create_proc_entry(const char *name, mode_t mode,
 	}
 	return ent;
 }
+EXPORT_SYMBOL(create_proc_entry);
 
 struct proc_dir_entry *proc_create_data(const char *name, mode_t mode,
 					struct proc_dir_entry *parent,
@@ -762,6 +772,7 @@ out_free:
 out:
 	return NULL;
 }
+EXPORT_SYMBOL(proc_create_data);
 
 static void free_proc_entry(struct proc_dir_entry *de)
 {
@@ -793,11 +804,13 @@ void remove_proc_entry(const char *name, struct proc_dir_entry *parent)
 	const char *fn = name;
 	int len;
 
-	if (xlate_proc_name(name, &parent, &fn) != 0)
+	spin_lock(&proc_subdir_lock);
+	if (__xlate_proc_name(name, &parent, &fn) != 0) {
+		spin_unlock(&proc_subdir_lock);
 		return;
+	}
 	len = strlen(fn);
 
-	spin_lock(&proc_subdir_lock);
 	for (p = &parent->subdir; *p; p=&(*p)->next ) {
 		if (proc_match(len, fn, *p)) {
 			de = *p;
@@ -807,8 +820,10 @@ void remove_proc_entry(const char *name, struct proc_dir_entry *parent)
 		}
 	}
 	spin_unlock(&proc_subdir_lock);
-	if (!de)
+	if (!de) {
+		WARN(1, "name '%s'\n", name);
 		return;
+	}
 
 	spin_lock(&de->pde_unload_lock);
 	/*
@@ -853,3 +868,4 @@ continue_removing:
 			de->parent->name, de->name, de->subdir->name);
 	pde_put(de);
 }
+EXPORT_SYMBOL(remove_proc_entry);
diff --git a/fs/proc/kmsg.c b/fs/proc/kmsg.c
index 7ca78346d3f0..cfe90a48a6e8 100644
--- a/fs/proc/kmsg.c
+++ b/fs/proc/kmsg.c
@@ -12,37 +12,37 @@
 #include <linux/poll.h>
 #include <linux/proc_fs.h>
 #include <linux/fs.h>
+#include <linux/syslog.h>
 
 #include <asm/uaccess.h>
 #include <asm/io.h>
 
 extern wait_queue_head_t log_wait;
 
-extern int do_syslog(int type, char __user *bug, int count);
-
 static int kmsg_open(struct inode * inode, struct file * file)
 {
-	return do_syslog(1,NULL,0);
+	return do_syslog(SYSLOG_ACTION_OPEN, NULL, 0, SYSLOG_FROM_FILE);
 }
 
 static int kmsg_release(struct inode * inode, struct file * file)
 {
-	(void) do_syslog(0,NULL,0);
+	(void) do_syslog(SYSLOG_ACTION_CLOSE, NULL, 0, SYSLOG_FROM_FILE);
 	return 0;
 }
 
 static ssize_t kmsg_read(struct file *file, char __user *buf,
 			 size_t count, loff_t *ppos)
 {
-	if ((file->f_flags & O_NONBLOCK) && !do_syslog(9, NULL, 0))
+	if ((file->f_flags & O_NONBLOCK) &&
+	    !do_syslog(SYSLOG_ACTION_SIZE_UNREAD, NULL, 0, SYSLOG_FROM_FILE))
 		return -EAGAIN;
-	return do_syslog(2, buf, count);
+	return do_syslog(SYSLOG_ACTION_READ, buf, count, SYSLOG_FROM_FILE);
 }
 
 static unsigned int kmsg_poll(struct file *file, poll_table *wait)
 {
 	poll_wait(file, &log_wait, wait);
-	if (do_syslog(9, NULL, 0))
+	if (do_syslog(SYSLOG_ACTION_SIZE_UNREAD, NULL, 0, SYSLOG_FROM_FILE))
 		return POLLIN | POLLRDNORM;
 	return 0;
 }
diff --git a/fs/proc/proc_devtree.c b/fs/proc/proc_devtree.c
index 123257bb356b..f8650dce74fb 100644
--- a/fs/proc/proc_devtree.c
+++ b/fs/proc/proc_devtree.c
@@ -10,16 +10,19 @@
 #include <linux/seq_file.h>
 #include <linux/stat.h>
 #include <linux/string.h>
+#include <linux/of.h>
+#include <linux/module.h>
 #include <asm/prom.h>
 #include <asm/uaccess.h>
 #include "internal.h"
 
-#ifndef HAVE_ARCH_DEVTREE_FIXUPS
 static inline void set_node_proc_entry(struct device_node *np,
 				       struct proc_dir_entry *de)
 {
-}
+#ifdef HAVE_ARCH_DEVTREE_FIXUPS
+	np->pde = de;
 #endif
+}
 
 static struct proc_dir_entry *proc_device_tree;
 
diff --git a/fs/proc/root.c b/fs/proc/root.c
index b080b791d9e3..757c069f2a65 100644
--- a/fs/proc/root.c
+++ b/fs/proc/root.c
@@ -220,9 +220,3 @@ void pid_ns_release_proc(struct pid_namespace *ns)
 {
 	mntput(ns->proc_mnt);
 }
-
-EXPORT_SYMBOL(proc_symlink);
-EXPORT_SYMBOL(proc_mkdir);
-EXPORT_SYMBOL(create_proc_entry);
-EXPORT_SYMBOL(proc_create_data);
-EXPORT_SYMBOL(remove_proc_entry);
diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index f277c4a111cb..183f8ff5f400 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -16,7 +16,7 @@
 
 void task_mem(struct seq_file *m, struct mm_struct *mm)
 {
-	unsigned long data, text, lib;
+	unsigned long data, text, lib, swap;
 	unsigned long hiwater_vm, total_vm, hiwater_rss, total_rss;
 
 	/*
@@ -36,6 +36,7 @@ void task_mem(struct seq_file *m, struct mm_struct *mm)
 	data = mm->total_vm - mm->shared_vm - mm->stack_vm;
 	text = (PAGE_ALIGN(mm->end_code) - (mm->start_code & PAGE_MASK)) >> 10;
 	lib = (mm->exec_vm << (PAGE_SHIFT-10)) - text;
+	swap = get_mm_counter(mm, MM_SWAPENTS);
 	seq_printf(m,
 		"VmPeak:\t%8lu kB\n"
 		"VmSize:\t%8lu kB\n"
@@ -46,7 +47,8 @@ void task_mem(struct seq_file *m, struct mm_struct *mm)
 		"VmStk:\t%8lu kB\n"
 		"VmExe:\t%8lu kB\n"
 		"VmLib:\t%8lu kB\n"
-		"VmPTE:\t%8lu kB\n",
+		"VmPTE:\t%8lu kB\n"
+		"VmSwap:\t%8lu kB\n",
 		hiwater_vm << (PAGE_SHIFT-10),
 		(total_vm - mm->reserved_vm) << (PAGE_SHIFT-10),
 		mm->locked_vm << (PAGE_SHIFT-10),
@@ -54,7 +56,8 @@ void task_mem(struct seq_file *m, struct mm_struct *mm)
 		total_rss << (PAGE_SHIFT-10),
 		data << (PAGE_SHIFT-10),
 		mm->stack_vm << (PAGE_SHIFT-10), text, lib,
-		(PTRS_PER_PTE*sizeof(pte_t)*mm->nr_ptes) >> 10);
+		(PTRS_PER_PTE*sizeof(pte_t)*mm->nr_ptes) >> 10,
+		swap << (PAGE_SHIFT-10));
 }
 
 unsigned long task_vsize(struct mm_struct *mm)
@@ -65,11 +68,11 @@ unsigned long task_vsize(struct mm_struct *mm)
 int task_statm(struct mm_struct *mm, int *shared, int *text,
 	       int *data, int *resident)
 {
-	*shared = get_mm_counter(mm, file_rss);
+	*shared = get_mm_counter(mm, MM_FILEPAGES);
 	*text = (PAGE_ALIGN(mm->end_code) - (mm->start_code & PAGE_MASK))
 								>> PAGE_SHIFT;
 	*data = mm->total_vm - mm->shared_vm;
-	*resident = *shared + get_mm_counter(mm, anon_rss);
+	*resident = *shared + get_mm_counter(mm, MM_ANONPAGES);
 	return mm->total_vm;
 }
 
diff --git a/fs/quota/Kconfig b/fs/quota/Kconfig
index efc02ebb8c70..dad7fb247ddc 100644
--- a/fs/quota/Kconfig
+++ b/fs/quota/Kconfig
@@ -59,3 +59,8 @@ config QUOTACTL
 	bool
 	depends on XFS_QUOTA || QUOTA
 	default y
+
+config QUOTACTL_COMPAT
+	bool
+	depends on QUOTACTL && COMPAT_FOR_U64_ALIGNMENT
+	default y
diff --git a/fs/quota/Makefile b/fs/quota/Makefile
index 68d4f6dc0578..5f9e9e276af0 100644
--- a/fs/quota/Makefile
+++ b/fs/quota/Makefile
@@ -3,3 +3,5 @@ obj-$(CONFIG_QFMT_V1)		+= quota_v1.o
 obj-$(CONFIG_QFMT_V2)		+= quota_v2.o
 obj-$(CONFIG_QUOTA_TREE)	+= quota_tree.o
 obj-$(CONFIG_QUOTACTL)		+= quota.o
+obj-$(CONFIG_QUOTACTL_COMPAT)	+= compat.o
+obj-$(CONFIG_QUOTA_NETLINK_INTERFACE)	+= netlink.o
diff --git a/fs/quota/compat.c b/fs/quota/compat.c
new file mode 100644
index 000000000000..fb1892fe3e56
--- /dev/null
+++ b/fs/quota/compat.c
@@ -0,0 +1,118 @@
+
+#include <linux/syscalls.h>
+#include <linux/compat.h>
+#include <linux/quotaops.h>
+
+/*
+ * This code works only for 32 bit quota tools over 64 bit OS (x86_64, ia64)
+ * and is necessary due to alignment problems.
+ */
+struct compat_if_dqblk {
+	compat_u64 dqb_bhardlimit;
+	compat_u64 dqb_bsoftlimit;
+	compat_u64 dqb_curspace;
+	compat_u64 dqb_ihardlimit;
+	compat_u64 dqb_isoftlimit;
+	compat_u64 dqb_curinodes;
+	compat_u64 dqb_btime;
+	compat_u64 dqb_itime;
+	compat_uint_t dqb_valid;
+};
+
+/* XFS structures */
+struct compat_fs_qfilestat {
+	compat_u64 dqb_bhardlimit;
+	compat_u64 qfs_nblks;
+	compat_uint_t qfs_nextents;
+};
+
+struct compat_fs_quota_stat {
+	__s8		qs_version;
+	__u16		qs_flags;
+	__s8		qs_pad;
+	struct compat_fs_qfilestat	qs_uquota;
+	struct compat_fs_qfilestat	qs_gquota;
+	compat_uint_t	qs_incoredqs;
+	compat_int_t	qs_btimelimit;
+	compat_int_t	qs_itimelimit;
+	compat_int_t	qs_rtbtimelimit;
+	__u16		qs_bwarnlimit;
+	__u16		qs_iwarnlimit;
+};
+
+asmlinkage long sys32_quotactl(unsigned int cmd, const char __user *special,
+						qid_t id, void __user *addr)
+{
+	unsigned int cmds;
+	struct if_dqblk __user *dqblk;
+	struct compat_if_dqblk __user *compat_dqblk;
+	struct fs_quota_stat __user *fsqstat;
+	struct compat_fs_quota_stat __user *compat_fsqstat;
+	compat_uint_t data;
+	u16 xdata;
+	long ret;
+
+	cmds = cmd >> SUBCMDSHIFT;
+
+	switch (cmds) {
+	case Q_GETQUOTA:
+		dqblk = compat_alloc_user_space(sizeof(struct if_dqblk));
+		compat_dqblk = addr;
+		ret = sys_quotactl(cmd, special, id, dqblk);
+		if (ret)
+			break;
+		if (copy_in_user(compat_dqblk, dqblk, sizeof(*compat_dqblk)) ||
+			get_user(data, &dqblk->dqb_valid) ||
+			put_user(data, &compat_dqblk->dqb_valid))
+			ret = -EFAULT;
+		break;
+	case Q_SETQUOTA:
+		dqblk = compat_alloc_user_space(sizeof(struct if_dqblk));
+		compat_dqblk = addr;
+		ret = -EFAULT;
+		if (copy_in_user(dqblk, compat_dqblk, sizeof(*compat_dqblk)) ||
+			get_user(data, &compat_dqblk->dqb_valid) ||
+			put_user(data, &dqblk->dqb_valid))
+			break;
+		ret = sys_quotactl(cmd, special, id, dqblk);
+		break;
+	case Q_XGETQSTAT:
+		fsqstat = compat_alloc_user_space(sizeof(struct fs_quota_stat));
+		compat_fsqstat = addr;
+		ret = sys_quotactl(cmd, special, id, fsqstat);
+		if (ret)
+			break;
+		ret = -EFAULT;
+		/* Copying qs_version, qs_flags, qs_pad */
+		if (copy_in_user(compat_fsqstat, fsqstat,
+			offsetof(struct compat_fs_quota_stat, qs_uquota)))
+			break;
+		/* Copying qs_uquota */
+		if (copy_in_user(&compat_fsqstat->qs_uquota,
+			&fsqstat->qs_uquota,
+			sizeof(compat_fsqstat->qs_uquota)) ||
+			get_user(data, &fsqstat->qs_uquota.qfs_nextents) ||
+			put_user(data, &compat_fsqstat->qs_uquota.qfs_nextents))
+			break;
+		/* Copying qs_gquota */
+		if (copy_in_user(&compat_fsqstat->qs_gquota,
+			&fsqstat->qs_gquota,
+			sizeof(compat_fsqstat->qs_gquota)) ||
+			get_user(data, &fsqstat->qs_gquota.qfs_nextents) ||
+			put_user(data, &compat_fsqstat->qs_gquota.qfs_nextents))
+			break;
+		/* Copying the rest */
+		if (copy_in_user(&compat_fsqstat->qs_incoredqs,
+			&fsqstat->qs_incoredqs,
+			sizeof(struct compat_fs_quota_stat) -
+			offsetof(struct compat_fs_quota_stat, qs_incoredqs)) ||
+			get_user(xdata, &fsqstat->qs_iwarnlimit) ||
+			put_user(xdata, &compat_fsqstat->qs_iwarnlimit))
+			break;
+		ret = 0;
+		break;
+	default:
+		ret = sys_quotactl(cmd, special, id, addr);
+	}
+	return ret;
+}
diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
index 3fc62b097bed..e0b870f4749f 100644
--- a/fs/quota/dquot.c
+++ b/fs/quota/dquot.c
@@ -100,9 +100,13 @@
  *
  * Any operation working on dquots via inode pointers must hold dqptr_sem.  If
  * operation is just reading pointers from inode (or not using them at all) the
- * read lock is enough. If pointers are altered function must hold write lock
- * (these locking rules also apply for S_NOQUOTA flag in the inode - note that
- * for altering the flag i_mutex is also needed).
+ * read lock is enough. If pointers are altered function must hold write lock.
+ * Special care needs to be taken about S_NOQUOTA inode flag (marking that
+ * inode is a quota file). Functions adding pointers from inode to dquots have
+ * to check this flag under dqptr_sem and then (if S_NOQUOTA is not set) they
+ * have to do all pointer modifications before dropping dqptr_sem. This makes
+ * sure they cannot race with quotaon which first sets S_NOQUOTA flag and
+ * then drops all pointers to dquots from an inode.
  *
  * Each dquot has its dq_lock mutex. Locked dquots might not be referenced
  * from inodes (dquot_alloc_space() and such don't check the dq_lock).
@@ -225,6 +229,9 @@ static struct hlist_head *dquot_hash;
 struct dqstats dqstats;
 EXPORT_SYMBOL(dqstats);
 
+static qsize_t inode_get_rsv_space(struct inode *inode);
+static void __dquot_initialize(struct inode *inode, int type);
+
 static inline unsigned int
 hashfn(const struct super_block *sb, unsigned int id, int type)
 {
@@ -564,7 +571,7 @@ out:
 }
 EXPORT_SYMBOL(dquot_scan_active);
 
-int vfs_quota_sync(struct super_block *sb, int type)
+int vfs_quota_sync(struct super_block *sb, int type, int wait)
 {
 	struct list_head *dirty;
 	struct dquot *dquot;
@@ -609,6 +616,33 @@ int vfs_quota_sync(struct super_block *sb, int type)
 	spin_unlock(&dq_list_lock);
 	mutex_unlock(&dqopt->dqonoff_mutex);
 
+	if (!wait || (sb_dqopt(sb)->flags & DQUOT_QUOTA_SYS_FILE))
+		return 0;
+
+	/* This is not very clever (and fast) but currently I don't know about
+	 * any other simple way of getting quota data to disk and we must get
+	 * them there for userspace to be visible... */
+	if (sb->s_op->sync_fs)
+		sb->s_op->sync_fs(sb, 1);
+	sync_blockdev(sb->s_bdev);
+
+	/*
+	 * Now when everything is written we can discard the pagecache so
+	 * that userspace sees the changes.
+	 */
+	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		if (type != -1 && cnt != type)
+			continue;
+		if (!sb_has_quota_active(sb, cnt))
+			continue;
+		mutex_lock_nested(&sb_dqopt(sb)->files[cnt]->i_mutex,
+				  I_MUTEX_QUOTA);
+		truncate_inode_pages(&sb_dqopt(sb)->files[cnt]->i_data, 0);
+		mutex_unlock(&sb_dqopt(sb)->files[cnt]->i_mutex);
+	}
+	mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
+
 	return 0;
 }
 EXPORT_SYMBOL(vfs_quota_sync);
@@ -840,11 +874,14 @@ static int dqinit_needed(struct inode *inode, int type)
 static void add_dquot_ref(struct super_block *sb, int type)
 {
 	struct inode *inode, *old_inode = NULL;
+	int reserved = 0;
 
 	spin_lock(&inode_lock);
 	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
 		if (inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE|I_NEW))
 			continue;
+		if (unlikely(inode_get_rsv_space(inode) > 0))
+			reserved = 1;
 		if (!atomic_read(&inode->i_writecount))
 			continue;
 		if (!dqinit_needed(inode, type))
@@ -854,7 +891,7 @@ static void add_dquot_ref(struct super_block *sb, int type)
 		spin_unlock(&inode_lock);
 
 		iput(old_inode);
-		sb->dq_op->initialize(inode, type);
+		__dquot_initialize(inode, type);
 		/* We hold a reference to 'inode' so it couldn't have been
 		 * removed from s_inodes list while we dropped the inode_lock.
 		 * We cannot iput the inode now as we can be holding the last
@@ -865,6 +902,12 @@ static void add_dquot_ref(struct super_block *sb, int type)
 	}
 	spin_unlock(&inode_lock);
 	iput(old_inode);
+
+	if (reserved) {
+		printk(KERN_WARNING "VFS (%s): Writes happened before quota"
+			" was turned on thus quota information is probably "
+			"inconsistent. Please run quotacheck(8).\n", sb->s_id);
+	}
 }
 
 /*
@@ -978,10 +1021,12 @@ static inline void dquot_resv_space(struct dquot *dquot, qsize_t number)
 /*
  * Claim reserved quota space
  */
-static void dquot_claim_reserved_space(struct dquot *dquot,
-						qsize_t number)
+static void dquot_claim_reserved_space(struct dquot *dquot, qsize_t number)
 {
-	WARN_ON(dquot->dq_dqb.dqb_rsvspace < number);
+	if (dquot->dq_dqb.dqb_rsvspace < number) {
+		WARN_ON_ONCE(1);
+		number = dquot->dq_dqb.dqb_rsvspace;
+	}
 	dquot->dq_dqb.dqb_curspace += number;
 	dquot->dq_dqb.dqb_rsvspace -= number;
 }
@@ -989,7 +1034,12 @@ static void dquot_claim_reserved_space(struct dquot *dquot,
 static inline
 void dquot_free_reserved_space(struct dquot *dquot, qsize_t number)
 {
-	dquot->dq_dqb.dqb_rsvspace -= number;
+	if (dquot->dq_dqb.dqb_rsvspace >= number)
+		dquot->dq_dqb.dqb_rsvspace -= number;
+	else {
+		WARN_ON_ONCE(1);
+		dquot->dq_dqb.dqb_rsvspace = 0;
+	}
 }
 
 static void dquot_decr_inodes(struct dquot *dquot, qsize_t number)
@@ -1131,13 +1181,13 @@ static int check_idq(struct dquot *dquot, qsize_t inodes, char *warntype)
 	*warntype = QUOTA_NL_NOWARN;
 	if (!sb_has_quota_limits_enabled(dquot->dq_sb, dquot->dq_type) ||
 	    test_bit(DQ_FAKE_B, &dquot->dq_flags))
-		return QUOTA_OK;
+		return 0;
 
 	if (dquot->dq_dqb.dqb_ihardlimit &&
 	    newinodes > dquot->dq_dqb.dqb_ihardlimit &&
             !ignore_hardlimit(dquot)) {
 		*warntype = QUOTA_NL_IHARDWARN;
-		return NO_QUOTA;
+		return -EDQUOT;
 	}
 
 	if (dquot->dq_dqb.dqb_isoftlimit &&
@@ -1146,7 +1196,7 @@ static int check_idq(struct dquot *dquot, qsize_t inodes, char *warntype)
 	    get_seconds() >= dquot->dq_dqb.dqb_itime &&
             !ignore_hardlimit(dquot)) {
 		*warntype = QUOTA_NL_ISOFTLONGWARN;
-		return NO_QUOTA;
+		return -EDQUOT;
 	}
 
 	if (dquot->dq_dqb.dqb_isoftlimit &&
@@ -1157,7 +1207,7 @@ static int check_idq(struct dquot *dquot, qsize_t inodes, char *warntype)
 		    sb_dqopt(dquot->dq_sb)->info[dquot->dq_type].dqi_igrace;
 	}
 
-	return QUOTA_OK;
+	return 0;
 }
 
 /* needs dq_data_lock */
@@ -1169,7 +1219,7 @@ static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *war
 	*warntype = QUOTA_NL_NOWARN;
 	if (!sb_has_quota_limits_enabled(sb, dquot->dq_type) ||
 	    test_bit(DQ_FAKE_B, &dquot->dq_flags))
-		return QUOTA_OK;
+		return 0;
 
 	tspace = dquot->dq_dqb.dqb_curspace + dquot->dq_dqb.dqb_rsvspace
 		+ space;
@@ -1179,7 +1229,7 @@ static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *war
             !ignore_hardlimit(dquot)) {
 		if (!prealloc)
 			*warntype = QUOTA_NL_BHARDWARN;
-		return NO_QUOTA;
+		return -EDQUOT;
 	}
 
 	if (dquot->dq_dqb.dqb_bsoftlimit &&
@@ -1189,7 +1239,7 @@ static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *war
             !ignore_hardlimit(dquot)) {
 		if (!prealloc)
 			*warntype = QUOTA_NL_BSOFTLONGWARN;
-		return NO_QUOTA;
+		return -EDQUOT;
 	}
 
 	if (dquot->dq_dqb.dqb_bsoftlimit &&
@@ -1205,10 +1255,10 @@ static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *war
 			 * We don't allow preallocation to exceed softlimit so exceeding will
 			 * be always printed
 			 */
-			return NO_QUOTA;
+			return -EDQUOT;
 	}
 
-	return QUOTA_OK;
+	return 0;
 }
 
 static int info_idq_free(struct dquot *dquot, qsize_t inodes)
@@ -1242,25 +1292,32 @@ static int info_bdq_free(struct dquot *dquot, qsize_t space)
 		return QUOTA_NL_BHARDBELOW;
 	return QUOTA_NL_NOWARN;
 }
+
 /*
- *	Initialize quota pointers in inode
- *	We do things in a bit complicated way but by that we avoid calling
- *	dqget() and thus filesystem callbacks under dqptr_sem.
+ * Initialize quota pointers in inode
+ *
+ * We do things in a bit complicated way but by that we avoid calling
+ * dqget() and thus filesystem callbacks under dqptr_sem.
+ *
+ * It is better to call this function outside of any transaction as it
+ * might need a lot of space in journal for dquot structure allocation.
  */
-int dquot_initialize(struct inode *inode, int type)
+static void __dquot_initialize(struct inode *inode, int type)
 {
 	unsigned int id = 0;
-	int cnt, ret = 0;
-	struct dquot *got[MAXQUOTAS] = { NULL, NULL };
+	int cnt;
+	struct dquot *got[MAXQUOTAS];
 	struct super_block *sb = inode->i_sb;
+	qsize_t rsv;
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (IS_NOQUOTA(inode))
-		return 0;
+	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode))
+		return;
 
 	/* First get references to structures we might need. */
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		got[cnt] = NULL;
 		if (type != -1 && cnt != type)
 			continue;
 		switch (cnt) {
@@ -1275,7 +1332,6 @@ int dquot_initialize(struct inode *inode, int type)
 	}
 
 	down_write(&sb_dqopt(sb)->dqptr_sem);
-	/* Having dqptr_sem we know NOQUOTA flags can't be altered... */
 	if (IS_NOQUOTA(inode))
 		goto out_err;
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
@@ -1287,20 +1343,31 @@ int dquot_initialize(struct inode *inode, int type)
 		if (!inode->i_dquot[cnt]) {
 			inode->i_dquot[cnt] = got[cnt];
 			got[cnt] = NULL;
+			/*
+			 * Make quota reservation system happy if someone
+			 * did a write before quota was turned on
+			 */
+			rsv = inode_get_rsv_space(inode);
+			if (unlikely(rsv))
+				dquot_resv_space(inode->i_dquot[cnt], rsv);
 		}
 	}
 out_err:
 	up_write(&sb_dqopt(sb)->dqptr_sem);
 	/* Drop unused references */
 	dqput_all(got);
-	return ret;
+}
+
+void dquot_initialize(struct inode *inode)
+{
+	__dquot_initialize(inode, -1);
 }
 EXPORT_SYMBOL(dquot_initialize);
 
 /*
  * 	Release all quotas referenced by inode
  */
-int dquot_drop(struct inode *inode)
+static void __dquot_drop(struct inode *inode)
 {
 	int cnt;
 	struct dquot *put[MAXQUOTAS];
@@ -1312,32 +1379,31 @@ int dquot_drop(struct inode *inode)
 	}
 	up_write(&sb_dqopt(inode->i_sb)->dqptr_sem);
 	dqput_all(put);
-	return 0;
 }
-EXPORT_SYMBOL(dquot_drop);
 
-/* Wrapper to remove references to quota structures from inode */
-void vfs_dq_drop(struct inode *inode)
-{
-	/* Here we can get arbitrary inode from clear_inode() so we have
-	 * to be careful. OTOH we don't need locking as quota operations
-	 * are allowed to change only at mount time */
-	if (!IS_NOQUOTA(inode) && inode->i_sb && inode->i_sb->dq_op
-	    && inode->i_sb->dq_op->drop) {
-		int cnt;
-		/* Test before calling to rule out calls from proc and such
-                 * where we are not allowed to block. Note that this is
-		 * actually reliable test even without the lock - the caller
-		 * must assure that nobody can come after the DQUOT_DROP and
-		 * add quota pointers back anyway */
-		for (cnt = 0; cnt < MAXQUOTAS; cnt++)
-			if (inode->i_dquot[cnt])
-				break;
-		if (cnt < MAXQUOTAS)
-			inode->i_sb->dq_op->drop(inode);
-	}
-}
-EXPORT_SYMBOL(vfs_dq_drop);
+void dquot_drop(struct inode *inode)
+{
+	int cnt;
+
+	if (IS_NOQUOTA(inode))
+		return;
+
+	/*
+	 * Test before calling to rule out calls from proc and such
+	 * where we are not allowed to block. Note that this is
+	 * actually reliable test even without the lock - the caller
+	 * must assure that nobody can come after the DQUOT_DROP and
+	 * add quota pointers back anyway.
+	 */
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		if (inode->i_dquot[cnt])
+			break;
+	}
+
+	if (cnt < MAXQUOTAS)
+		__dquot_drop(inode);
+}
+EXPORT_SYMBOL(dquot_drop);
 
 /*
  * inode_reserved_space is managed internally by quota, and protected by
@@ -1351,28 +1417,30 @@ static qsize_t *inode_reserved_space(struct inode * inode)
 	return inode->i_sb->dq_op->get_reserved_space(inode);
 }
 
-static void inode_add_rsv_space(struct inode *inode, qsize_t number)
+void inode_add_rsv_space(struct inode *inode, qsize_t number)
 {
 	spin_lock(&inode->i_lock);
 	*inode_reserved_space(inode) += number;
 	spin_unlock(&inode->i_lock);
 }
+EXPORT_SYMBOL(inode_add_rsv_space);
 
-
-static void inode_claim_rsv_space(struct inode *inode, qsize_t number)
+void inode_claim_rsv_space(struct inode *inode, qsize_t number)
 {
 	spin_lock(&inode->i_lock);
 	*inode_reserved_space(inode) -= number;
 	__inode_add_bytes(inode, number);
 	spin_unlock(&inode->i_lock);
 }
+EXPORT_SYMBOL(inode_claim_rsv_space);
 
-static void inode_sub_rsv_space(struct inode *inode, qsize_t number)
+void inode_sub_rsv_space(struct inode *inode, qsize_t number)
 {
 	spin_lock(&inode->i_lock);
 	*inode_reserved_space(inode) -= number;
 	spin_unlock(&inode->i_lock);
 }
+EXPORT_SYMBOL(inode_sub_rsv_space);
 
 static qsize_t inode_get_rsv_space(struct inode *inode)
 {
@@ -1404,38 +1472,34 @@ static void inode_decr_space(struct inode *inode, qsize_t number, int reserve)
 }
 
 /*
- * Following four functions update i_blocks+i_bytes fields and
- * quota information (together with appropriate checks)
- * NOTE: We absolutely rely on the fact that caller dirties
- * the inode (usually macros in quotaops.h care about this) and
- * holds a handle for the current transaction so that dquot write and
- * inode write go into the same transaction.
+ * This functions updates i_blocks+i_bytes fields and quota information
+ * (together with appropriate checks).
+ *
+ * NOTE: We absolutely rely on the fact that caller dirties the inode
+ * (usually helpers in quotaops.h care about this) and holds a handle for
+ * the current transaction so that dquot write and inode write go into the
+ * same transaction.
  */
 
 /*
  * This operation can block, but only after everything is updated
  */
 int __dquot_alloc_space(struct inode *inode, qsize_t number,
-			int warn, int reserve)
+		int warn, int reserve)
 {
-	int cnt, ret = QUOTA_OK;
+	int cnt, ret = 0;
 	char warntype[MAXQUOTAS];
 
 	/*
 	 * First test before acquiring mutex - solves deadlocks when we
 	 * re-enter the quota code and are already holding the mutex
 	 */
-	if (IS_NOQUOTA(inode)) {
+	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) {
 		inode_incr_space(inode, number, reserve);
 		goto out;
 	}
 
 	down_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	if (IS_NOQUOTA(inode)) {
-		inode_incr_space(inode, number, reserve);
-		goto out_unlock;
-	}
-
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
 		warntype[cnt] = QUOTA_NL_NOWARN;
 
@@ -1443,9 +1507,9 @@ int __dquot_alloc_space(struct inode *inode, qsize_t number,
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (!inode->i_dquot[cnt])
 			continue;
-		if (check_bdq(inode->i_dquot[cnt], number, warn, warntype+cnt)
-		    == NO_QUOTA) {
-			ret = NO_QUOTA;
+		ret = check_bdq(inode->i_dquot[cnt], number, !warn,
+				warntype+cnt);
+		if (ret) {
 			spin_unlock(&dq_data_lock);
 			goto out_flush_warn;
 		}
@@ -1466,61 +1530,45 @@ int __dquot_alloc_space(struct inode *inode, qsize_t number,
 	mark_all_dquot_dirty(inode->i_dquot);
 out_flush_warn:
 	flush_warnings(inode->i_dquot, warntype);
-out_unlock:
 	up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
 out:
 	return ret;
 }
-
-int dquot_alloc_space(struct inode *inode, qsize_t number, int warn)
-{
-	return __dquot_alloc_space(inode, number, warn, 0);
-}
-EXPORT_SYMBOL(dquot_alloc_space);
-
-int dquot_reserve_space(struct inode *inode, qsize_t number, int warn)
-{
-	return __dquot_alloc_space(inode, number, warn, 1);
-}
-EXPORT_SYMBOL(dquot_reserve_space);
+EXPORT_SYMBOL(__dquot_alloc_space);
 
 /*
  * This operation can block, but only after everything is updated
  */
-int dquot_alloc_inode(const struct inode *inode, qsize_t number)
+int dquot_alloc_inode(const struct inode *inode)
 {
-	int cnt, ret = NO_QUOTA;
+	int cnt, ret = 0;
 	char warntype[MAXQUOTAS];
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (IS_NOQUOTA(inode))
-		return QUOTA_OK;
+	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode))
+		return 0;
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
 		warntype[cnt] = QUOTA_NL_NOWARN;
 	down_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	if (IS_NOQUOTA(inode)) {
-		up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-		return QUOTA_OK;
-	}
 	spin_lock(&dq_data_lock);
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (!inode->i_dquot[cnt])
 			continue;
-		if (check_idq(inode->i_dquot[cnt], number, warntype+cnt)
-		    == NO_QUOTA)
+		ret = check_idq(inode->i_dquot[cnt], 1, warntype + cnt);
+		if (ret)
 			goto warn_put_all;
 	}
 
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (!inode->i_dquot[cnt])
 			continue;
-		dquot_incr_inodes(inode->i_dquot[cnt], number);
+		dquot_incr_inodes(inode->i_dquot[cnt], 1);
 	}
-	ret = QUOTA_OK;
+
 warn_put_all:
 	spin_unlock(&dq_data_lock);
-	if (ret == QUOTA_OK)
+	if (ret == 0)
 		mark_all_dquot_dirty(inode->i_dquot);
 	flush_warnings(inode->i_dquot, warntype);
 	up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
@@ -1528,23 +1576,19 @@ warn_put_all:
 }
 EXPORT_SYMBOL(dquot_alloc_inode);
 
-int dquot_claim_space(struct inode *inode, qsize_t number)
+/*
+ * Convert in-memory reserved quotas to real consumed quotas
+ */
+int dquot_claim_space_nodirty(struct inode *inode, qsize_t number)
 {
 	int cnt;
-	int ret = QUOTA_OK;
 
-	if (IS_NOQUOTA(inode)) {
+	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) {
 		inode_claim_rsv_space(inode, number);
-		goto out;
+		return 0;
 	}
 
 	down_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	if (IS_NOQUOTA(inode))	{
-		up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-		inode_claim_rsv_space(inode, number);
-		goto out;
-	}
-
 	spin_lock(&dq_data_lock);
 	/* Claim reserved quotas to allocated quotas */
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
@@ -1557,33 +1601,26 @@ int dquot_claim_space(struct inode *inode, qsize_t number)
 	spin_unlock(&dq_data_lock);
 	mark_all_dquot_dirty(inode->i_dquot);
 	up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-out:
-	return ret;
+	return 0;
 }
-EXPORT_SYMBOL(dquot_claim_space);
+EXPORT_SYMBOL(dquot_claim_space_nodirty);
 
 /*
  * This operation can block, but only after everything is updated
  */
-int __dquot_free_space(struct inode *inode, qsize_t number, int reserve)
+void __dquot_free_space(struct inode *inode, qsize_t number, int reserve)
 {
 	unsigned int cnt;
 	char warntype[MAXQUOTAS];
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (IS_NOQUOTA(inode)) {
-out_sub:
+	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) {
 		inode_decr_space(inode, number, reserve);
-		return QUOTA_OK;
+		return;
 	}
 
 	down_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	/* Now recheck reliably when holding dqptr_sem */
-	if (IS_NOQUOTA(inode)) {
-		up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-		goto out_sub;
-	}
 	spin_lock(&dq_data_lock);
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (!inode->i_dquot[cnt])
@@ -1603,56 +1640,34 @@ out_sub:
 out_unlock:
 	flush_warnings(inode->i_dquot, warntype);
 	up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	return QUOTA_OK;
-}
-
-int dquot_free_space(struct inode *inode, qsize_t number)
-{
-	return  __dquot_free_space(inode, number, 0);
 }
-EXPORT_SYMBOL(dquot_free_space);
-
-/*
- * Release reserved quota space
- */
-void dquot_release_reserved_space(struct inode *inode, qsize_t number)
-{
-	__dquot_free_space(inode, number, 1);
-
-}
-EXPORT_SYMBOL(dquot_release_reserved_space);
+EXPORT_SYMBOL(__dquot_free_space);
 
 /*
  * This operation can block, but only after everything is updated
  */
-int dquot_free_inode(const struct inode *inode, qsize_t number)
+void dquot_free_inode(const struct inode *inode)
 {
 	unsigned int cnt;
 	char warntype[MAXQUOTAS];
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
-	if (IS_NOQUOTA(inode))
-		return QUOTA_OK;
+	if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode))
+		return;
 
 	down_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	/* Now recheck reliably when holding dqptr_sem */
-	if (IS_NOQUOTA(inode)) {
-		up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-		return QUOTA_OK;
-	}
 	spin_lock(&dq_data_lock);
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (!inode->i_dquot[cnt])
 			continue;
-		warntype[cnt] = info_idq_free(inode->i_dquot[cnt], number);
-		dquot_decr_inodes(inode->i_dquot[cnt], number);
+		warntype[cnt] = info_idq_free(inode->i_dquot[cnt], 1);
+		dquot_decr_inodes(inode->i_dquot[cnt], 1);
 	}
 	spin_unlock(&dq_data_lock);
 	mark_all_dquot_dirty(inode->i_dquot);
 	flush_warnings(inode->i_dquot, warntype);
 	up_read(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	return QUOTA_OK;
 }
 EXPORT_SYMBOL(dquot_free_inode);
 
@@ -1662,37 +1677,31 @@ EXPORT_SYMBOL(dquot_free_inode);
  * This operation can block, but only after everything is updated
  * A transaction must be started when entering this function.
  */
-int dquot_transfer(struct inode *inode, struct iattr *iattr)
+static int __dquot_transfer(struct inode *inode, qid_t *chid, unsigned long mask)
 {
 	qsize_t space, cur_space;
 	qsize_t rsv_space = 0;
 	struct dquot *transfer_from[MAXQUOTAS];
 	struct dquot *transfer_to[MAXQUOTAS];
-	int cnt, ret = QUOTA_OK;
-	int chuid = iattr->ia_valid & ATTR_UID && inode->i_uid != iattr->ia_uid,
-	    chgid = iattr->ia_valid & ATTR_GID && inode->i_gid != iattr->ia_gid;
+	int cnt, ret = 0;
 	char warntype_to[MAXQUOTAS];
 	char warntype_from_inodes[MAXQUOTAS], warntype_from_space[MAXQUOTAS];
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
 	if (IS_NOQUOTA(inode))
-		return QUOTA_OK;
+		return 0;
 	/* Initialize the arrays */
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		transfer_from[cnt] = NULL;
 		transfer_to[cnt] = NULL;
 		warntype_to[cnt] = QUOTA_NL_NOWARN;
 	}
-	if (chuid)
-		transfer_to[USRQUOTA] = dqget(inode->i_sb, iattr->ia_uid,
-					      USRQUOTA);
-	if (chgid)
-		transfer_to[GRPQUOTA] = dqget(inode->i_sb, iattr->ia_gid,
-					      GRPQUOTA);
-
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		if (mask & (1 << cnt))
+			transfer_to[cnt] = dqget(inode->i_sb, chid[cnt], cnt);
+	}
 	down_write(&sb_dqopt(inode->i_sb)->dqptr_sem);
-	/* Now recheck reliably when holding dqptr_sem */
 	if (IS_NOQUOTA(inode)) {	/* File without quota accounting? */
 		up_write(&sb_dqopt(inode->i_sb)->dqptr_sem);
 		goto put_all;
@@ -1706,9 +1715,11 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr)
 		if (!transfer_to[cnt])
 			continue;
 		transfer_from[cnt] = inode->i_dquot[cnt];
-		if (check_idq(transfer_to[cnt], 1, warntype_to + cnt) ==
-		    NO_QUOTA || check_bdq(transfer_to[cnt], space, 0,
-		    warntype_to + cnt) == NO_QUOTA)
+		ret = check_idq(transfer_to[cnt], 1, warntype_to + cnt);
+		if (ret)
+			goto over_quota;
+		ret = check_bdq(transfer_to[cnt], space, 0, warntype_to + cnt);
+		if (ret)
 			goto over_quota;
 	}
 
@@ -1762,22 +1773,32 @@ over_quota:
 	/* Clear dquot pointers we don't want to dqput() */
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
 		transfer_from[cnt] = NULL;
-	ret = NO_QUOTA;
 	goto warn_put_all;
 }
-EXPORT_SYMBOL(dquot_transfer);
 
-/* Wrapper for transferring ownership of an inode */
-int vfs_dq_transfer(struct inode *inode, struct iattr *iattr)
+/* Wrapper for transferring ownership of an inode for uid/gid only
+ * Called from FSXXX_setattr()
+ */
+int dquot_transfer(struct inode *inode, struct iattr *iattr)
 {
+	qid_t chid[MAXQUOTAS];
+	unsigned long mask = 0;
+
+	if (iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) {
+		mask |= 1 << USRQUOTA;
+		chid[USRQUOTA] = iattr->ia_uid;
+	}
+	if (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid) {
+		mask |= 1 << GRPQUOTA;
+		chid[GRPQUOTA] = iattr->ia_gid;
+	}
 	if (sb_any_quota_active(inode->i_sb) && !IS_NOQUOTA(inode)) {
-		vfs_dq_init(inode);
-		if (inode->i_sb->dq_op->transfer(inode, iattr) == NO_QUOTA)
-			return 1;
+		dquot_initialize(inode);
+		return __dquot_transfer(inode, chid, mask);
 	}
 	return 0;
 }
-EXPORT_SYMBOL(vfs_dq_transfer);
+EXPORT_SYMBOL(dquot_transfer);
 
 /*
  * Write info of quota file to disk
@@ -1798,13 +1819,6 @@ EXPORT_SYMBOL(dquot_commit_info);
  * Definitions of diskquota operations.
  */
 const struct dquot_operations dquot_operations = {
-	.initialize	= dquot_initialize,
-	.drop		= dquot_drop,
-	.alloc_space	= dquot_alloc_space,
-	.alloc_inode	= dquot_alloc_inode,
-	.free_space	= dquot_free_space,
-	.free_inode	= dquot_free_inode,
-	.transfer	= dquot_transfer,
 	.write_dquot	= dquot_commit,
 	.acquire_dquot	= dquot_acquire,
 	.release_dquot	= dquot_release,
@@ -1815,6 +1829,20 @@ const struct dquot_operations dquot_operations = {
 };
 
 /*
+ * Generic helper for ->open on filesystems supporting disk quotas.
+ */
+int dquot_file_open(struct inode *inode, struct file *file)
+{
+	int error;
+
+	error = generic_file_open(inode, file);
+	if (!error && (file->f_mode & FMODE_WRITE))
+		dquot_initialize(inode);
+	return error;
+}
+EXPORT_SYMBOL(dquot_file_open);
+
+/*
  * Turn quota off on a device. type == -1 ==> quotaoff for all types (umount)
  */
 int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags)
@@ -1993,11 +2021,13 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id,
 	}
 
 	if (!(dqopt->flags & DQUOT_QUOTA_SYS_FILE)) {
-		/* As we bypass the pagecache we must now flush the inode so
-		 * that we see all the changes from userspace... */
-		write_inode_now(inode, 1);
-		/* And now flush the block cache so that kernel sees the
-		 * changes */
+		/* As we bypass the pagecache we must now flush all the
+		 * dirty data and invalidate caches so that kernel sees
+		 * changes from userspace. It is not enough to just flush
+		 * the quota file since if blocksize < pagesize, invalidation
+		 * of the cache could fail because of other unrelated dirty
+		 * data */
+		sync_filesystem(sb);
 		invalidate_bdev(sb->s_bdev);
 	}
 	mutex_lock(&dqopt->dqonoff_mutex);
@@ -2010,14 +2040,16 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id,
 		/* We don't want quota and atime on quota files (deadlocks
 		 * possible) Also nobody should write to the file - we use
 		 * special IO operations which ignore the immutable bit. */
-		down_write(&dqopt->dqptr_sem);
 		mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA);
 		oldflags = inode->i_flags & (S_NOATIME | S_IMMUTABLE |
 					     S_NOQUOTA);
 		inode->i_flags |= S_NOQUOTA | S_NOATIME | S_IMMUTABLE;
 		mutex_unlock(&inode->i_mutex);
-		up_write(&dqopt->dqptr_sem);
-		sb->dq_op->drop(inode);
+		/*
+		 * When S_NOQUOTA is set, remove dquot references as no more
+		 * references can be added
+		 */
+		__dquot_drop(inode);
 	}
 
 	error = -EIO;
@@ -2053,14 +2085,12 @@ out_file_init:
 	iput(inode);
 out_lock:
 	if (oldflags != -1) {
-		down_write(&dqopt->dqptr_sem);
 		mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA);
 		/* Set the flags back (in the case of accidental quotaon()
 		 * on a wrong file we don't want to mess up the flags) */
 		inode->i_flags &= ~(S_NOATIME | S_NOQUOTA | S_IMMUTABLE);
 		inode->i_flags |= oldflags;
 		mutex_unlock(&inode->i_mutex);
-		up_write(&dqopt->dqptr_sem);
 	}
 	mutex_unlock(&dqopt->dqonoff_mutex);
 out_fmt:
diff --git a/fs/quota/netlink.c b/fs/quota/netlink.c
new file mode 100644
index 000000000000..2663ed90fb03
--- /dev/null
+++ b/fs/quota/netlink.c
@@ -0,0 +1,95 @@
+
+#include <linux/cred.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/quotaops.h>
+#include <linux/sched.h>
+#include <net/netlink.h>
+#include <net/genetlink.h>
+
+/* Netlink family structure for quota */
+static struct genl_family quota_genl_family = {
+	.id = GENL_ID_GENERATE,
+	.hdrsize = 0,
+	.name = "VFS_DQUOT",
+	.version = 1,
+	.maxattr = QUOTA_NL_A_MAX,
+};
+
+/**
+ * quota_send_warning - Send warning to userspace about exceeded quota
+ * @type: The quota type: USRQQUOTA, GRPQUOTA,...
+ * @id: The user or group id of the quota that was exceeded
+ * @dev: The device on which the fs is mounted (sb->s_dev)
+ * @warntype: The type of the warning: QUOTA_NL_...
+ *
+ * This can be used by filesystems (including those which don't use
+ * dquot) to send a message to userspace relating to quota limits.
+ *
+ */
+
+void quota_send_warning(short type, unsigned int id, dev_t dev,
+			const char warntype)
+{
+	static atomic_t seq;
+	struct sk_buff *skb;
+	void *msg_head;
+	int ret;
+	int msg_size = 4 * nla_total_size(sizeof(u32)) +
+		       2 * nla_total_size(sizeof(u64));
+
+	/* We have to allocate using GFP_NOFS as we are called from a
+	 * filesystem performing write and thus further recursion into
+	 * the fs to free some data could cause deadlocks. */
+	skb = genlmsg_new(msg_size, GFP_NOFS);
+	if (!skb) {
+		printk(KERN_ERR
+		  "VFS: Not enough memory to send quota warning.\n");
+		return;
+	}
+	msg_head = genlmsg_put(skb, 0, atomic_add_return(1, &seq),
+			&quota_genl_family, 0, QUOTA_NL_C_WARNING);
+	if (!msg_head) {
+		printk(KERN_ERR
+		  "VFS: Cannot store netlink header in quota warning.\n");
+		goto err_out;
+	}
+	ret = nla_put_u32(skb, QUOTA_NL_A_QTYPE, type);
+	if (ret)
+		goto attr_err_out;
+	ret = nla_put_u64(skb, QUOTA_NL_A_EXCESS_ID, id);
+	if (ret)
+		goto attr_err_out;
+	ret = nla_put_u32(skb, QUOTA_NL_A_WARNING, warntype);
+	if (ret)
+		goto attr_err_out;
+	ret = nla_put_u32(skb, QUOTA_NL_A_DEV_MAJOR, MAJOR(dev));
+	if (ret)
+		goto attr_err_out;
+	ret = nla_put_u32(skb, QUOTA_NL_A_DEV_MINOR, MINOR(dev));
+	if (ret)
+		goto attr_err_out;
+	ret = nla_put_u64(skb, QUOTA_NL_A_CAUSED_ID, current_uid());
+	if (ret)
+		goto attr_err_out;
+	genlmsg_end(skb, msg_head);
+
+	genlmsg_multicast(skb, 0, quota_genl_family.id, GFP_NOFS);
+	return;
+attr_err_out:
+	printk(KERN_ERR "VFS: Not enough space to compose quota message!\n");
+err_out:
+	kfree_skb(skb);
+}
+EXPORT_SYMBOL(quota_send_warning);
+
+static int __init quota_init(void)
+{
+	if (genl_register_family(&quota_genl_family) != 0)
+		printk(KERN_ERR
+		       "VFS: Failed to create quota netlink interface.\n");
+	return 0;
+};
+
+module_init(quota_init);
diff --git a/fs/quota/quota.c b/fs/quota/quota.c
index ee91e2756950..95388f9b7356 100644
--- a/fs/quota/quota.c
+++ b/fs/quota/quota.c
@@ -10,7 +10,6 @@
 #include <linux/slab.h>
 #include <asm/current.h>
 #include <asm/uaccess.h>
-#include <linux/compat.h>
 #include <linux/kernel.h>
 #include <linux/security.h>
 #include <linux/syscalls.h>
@@ -18,220 +17,205 @@
 #include <linux/capability.h>
 #include <linux/quotaops.h>
 #include <linux/types.h>
-#include <net/netlink.h>
-#include <net/genetlink.h>
+#include <linux/writeback.h>
 
-/* Check validity of generic quotactl commands */
-static int generic_quotactl_valid(struct super_block *sb, int type, int cmd,
-				  qid_t id)
+static int check_quotactl_permission(struct super_block *sb, int type, int cmd,
+				     qid_t id)
 {
-	if (type >= MAXQUOTAS)
-		return -EINVAL;
-	if (!sb && cmd != Q_SYNC)
-		return -ENODEV;
-	/* Is operation supported? */
-	if (sb && !sb->s_qcop)
-		return -ENOSYS;
-
 	switch (cmd) {
-		case Q_GETFMT:
-			break;
-		case Q_QUOTAON:
-			if (!sb->s_qcop->quota_on)
-				return -ENOSYS;
-			break;
-		case Q_QUOTAOFF:
-			if (!sb->s_qcop->quota_off)
-				return -ENOSYS;
-			break;
-		case Q_SETINFO:
-			if (!sb->s_qcop->set_info)
-				return -ENOSYS;
-			break;
-		case Q_GETINFO:
-			if (!sb->s_qcop->get_info)
-				return -ENOSYS;
-			break;
-		case Q_SETQUOTA:
-			if (!sb->s_qcop->set_dqblk)
-				return -ENOSYS;
-			break;
-		case Q_GETQUOTA:
-			if (!sb->s_qcop->get_dqblk)
-				return -ENOSYS;
-			break;
-		case Q_SYNC:
-			if (sb && !sb->s_qcop->quota_sync)
-				return -ENOSYS;
+	/* these commands do not require any special privilegues */
+	case Q_GETFMT:
+	case Q_SYNC:
+	case Q_GETINFO:
+	case Q_XGETQSTAT:
+	case Q_XQUOTASYNC:
+		break;
+	/* allow to query information for dquots we "own" */
+	case Q_GETQUOTA:
+	case Q_XGETQUOTA:
+		if ((type == USRQUOTA && current_euid() == id) ||
+		    (type == GRPQUOTA && in_egroup_p(id)))
 			break;
-		default:
-			return -EINVAL;
+		/*FALLTHROUGH*/
+	default:
+		if (!capable(CAP_SYS_ADMIN))
+			return -EPERM;
 	}
 
-	/* Is quota turned on for commands which need it? */
-	switch (cmd) {
-		case Q_GETFMT:
-		case Q_GETINFO:
-		case Q_SETINFO:
-		case Q_SETQUOTA:
-		case Q_GETQUOTA:
-			/* This is just an informative test so we are satisfied
-			 * without the lock */
-			if (!sb_has_quota_active(sb, type))
-				return -ESRCH;
-	}
+	return security_quotactl(cmd, type, id, sb);
+}
 
-	/* Check privileges */
-	if (cmd == Q_GETQUOTA) {
-		if (((type == USRQUOTA && current_euid() != id) ||
-		     (type == GRPQUOTA && !in_egroup_p(id))) &&
-		    !capable(CAP_SYS_ADMIN))
-			return -EPERM;
+static int quota_sync_all(int type)
+{
+	struct super_block *sb;
+	int ret;
+
+	if (type >= MAXQUOTAS)
+		return -EINVAL;
+	ret = security_quotactl(Q_SYNC, type, 0, NULL);
+	if (ret)
+		return ret;
+
+	spin_lock(&sb_lock);
+restart:
+	list_for_each_entry(sb, &super_blocks, s_list) {
+		if (!sb->s_qcop || !sb->s_qcop->quota_sync)
+			continue;
+
+		sb->s_count++;
+		spin_unlock(&sb_lock);
+		down_read(&sb->s_umount);
+		if (sb->s_root)
+			sb->s_qcop->quota_sync(sb, type, 1);
+		up_read(&sb->s_umount);
+		spin_lock(&sb_lock);
+		if (__put_super_and_need_restart(sb))
+			goto restart;
 	}
-	else if (cmd != Q_GETFMT && cmd != Q_SYNC && cmd != Q_GETINFO)
-		if (!capable(CAP_SYS_ADMIN))
-			return -EPERM;
+	spin_unlock(&sb_lock);
 
 	return 0;
 }
 
-/* Check validity of XFS Quota Manager commands */
-static int xqm_quotactl_valid(struct super_block *sb, int type, int cmd,
-			      qid_t id)
+static int quota_quotaon(struct super_block *sb, int type, int cmd, qid_t id,
+		         void __user *addr)
 {
-	if (type >= XQM_MAXQUOTAS)
-		return -EINVAL;
-	if (!sb)
-		return -ENODEV;
-	if (!sb->s_qcop)
-		return -ENOSYS;
+	char *pathname;
+	int ret = -ENOSYS;
+
+	pathname = getname(addr);
+	if (IS_ERR(pathname))
+		return PTR_ERR(pathname);
+	if (sb->s_qcop->quota_on)
+		ret = sb->s_qcop->quota_on(sb, type, id, pathname, 0);
+	putname(pathname);
+	return ret;
+}
 
-	switch (cmd) {
-		case Q_XQUOTAON:
-		case Q_XQUOTAOFF:
-		case Q_XQUOTARM:
-			if (!sb->s_qcop->set_xstate)
-				return -ENOSYS;
-			break;
-		case Q_XGETQSTAT:
-			if (!sb->s_qcop->get_xstate)
-				return -ENOSYS;
-			break;
-		case Q_XSETQLIM:
-			if (!sb->s_qcop->set_xquota)
-				return -ENOSYS;
-			break;
-		case Q_XGETQUOTA:
-			if (!sb->s_qcop->get_xquota)
-				return -ENOSYS;
-			break;
-		case Q_XQUOTASYNC:
-			if (!sb->s_qcop->quota_sync)
-				return -ENOSYS;
-			break;
-		default:
-			return -EINVAL;
-	}
+static int quota_getfmt(struct super_block *sb, int type, void __user *addr)
+{
+	__u32 fmt;
 
-	/* Check privileges */
-	if (cmd == Q_XGETQUOTA) {
-		if (((type == XQM_USRQUOTA && current_euid() != id) ||
-		     (type == XQM_GRPQUOTA && !in_egroup_p(id))) &&
-		     !capable(CAP_SYS_ADMIN))
-			return -EPERM;
-	} else if (cmd != Q_XGETQSTAT && cmd != Q_XQUOTASYNC) {
-		if (!capable(CAP_SYS_ADMIN))
-			return -EPERM;
+	down_read(&sb_dqopt(sb)->dqptr_sem);
+	if (!sb_has_quota_active(sb, type)) {
+		up_read(&sb_dqopt(sb)->dqptr_sem);
+		return -ESRCH;
 	}
+	fmt = sb_dqopt(sb)->info[type].dqi_format->qf_fmt_id;
+	up_read(&sb_dqopt(sb)->dqptr_sem);
+	if (copy_to_user(addr, &fmt, sizeof(fmt)))
+		return -EFAULT;
+	return 0;
+}
 
+static int quota_getinfo(struct super_block *sb, int type, void __user *addr)
+{
+	struct if_dqinfo info;
+	int ret;
+
+	if (!sb_has_quota_active(sb, type))
+		return -ESRCH;
+	if (!sb->s_qcop->get_info)
+		return -ENOSYS;
+	ret = sb->s_qcop->get_info(sb, type, &info);
+	if (!ret && copy_to_user(addr, &info, sizeof(info)))
+		return -EFAULT;
+	return ret;
+}
+
+static int quota_setinfo(struct super_block *sb, int type, void __user *addr)
+{
+	struct if_dqinfo info;
+
+	if (copy_from_user(&info, addr, sizeof(info)))
+		return -EFAULT;
+	if (!sb_has_quota_active(sb, type))
+		return -ESRCH;
+	if (!sb->s_qcop->set_info)
+		return -ENOSYS;
+	return sb->s_qcop->set_info(sb, type, &info);
+}
+
+static int quota_getquota(struct super_block *sb, int type, qid_t id,
+			  void __user *addr)
+{
+	struct if_dqblk idq;
+	int ret;
+
+	if (!sb_has_quota_active(sb, type))
+		return -ESRCH;
+	if (!sb->s_qcop->get_dqblk)
+		return -ENOSYS;
+	ret = sb->s_qcop->get_dqblk(sb, type, id, &idq);
+	if (ret)
+		return ret;
+	if (copy_to_user(addr, &idq, sizeof(idq)))
+		return -EFAULT;
 	return 0;
 }
 
-static int check_quotactl_valid(struct super_block *sb, int type, int cmd,
-				qid_t id)
+static int quota_setquota(struct super_block *sb, int type, qid_t id,
+			  void __user *addr)
 {
-	int error;
-
-	if (XQM_COMMAND(cmd))
-		error = xqm_quotactl_valid(sb, type, cmd, id);
-	else
-		error = generic_quotactl_valid(sb, type, cmd, id);
-	if (!error)
-		error = security_quotactl(cmd, type, id, sb);
-	return error;
+	struct if_dqblk idq;
+
+	if (copy_from_user(&idq, addr, sizeof(idq)))
+		return -EFAULT;
+	if (!sb_has_quota_active(sb, type))
+		return -ESRCH;
+	if (!sb->s_qcop->set_dqblk)
+		return -ENOSYS;
+	return sb->s_qcop->set_dqblk(sb, type, id, &idq);
 }
 
-#ifdef CONFIG_QUOTA
-void sync_quota_sb(struct super_block *sb, int type)
+static int quota_setxstate(struct super_block *sb, int cmd, void __user *addr)
 {
-	int cnt;
+	__u32 flags;
 
-	if (!sb->s_qcop->quota_sync)
-		return;
+	if (copy_from_user(&flags, addr, sizeof(flags)))
+		return -EFAULT;
+	if (!sb->s_qcop->set_xstate)
+		return -ENOSYS;
+	return sb->s_qcop->set_xstate(sb, flags, cmd);
+}
 
-	sb->s_qcop->quota_sync(sb, type);
+static int quota_getxstate(struct super_block *sb, void __user *addr)
+{
+	struct fs_quota_stat fqs;
+	int ret;
 
-	if (sb_dqopt(sb)->flags & DQUOT_QUOTA_SYS_FILE)
-		return;
-	/* This is not very clever (and fast) but currently I don't know about
-	 * any other simple way of getting quota data to disk and we must get
-	 * them there for userspace to be visible... */
-	if (sb->s_op->sync_fs)
-		sb->s_op->sync_fs(sb, 1);
-	sync_blockdev(sb->s_bdev);
+	if (!sb->s_qcop->get_xstate)
+		return -ENOSYS;
+	ret = sb->s_qcop->get_xstate(sb, &fqs);
+	if (!ret && copy_to_user(addr, &fqs, sizeof(fqs)))
+		return -EFAULT;
+	return ret;
+}
 
-	/*
-	 * Now when everything is written we can discard the pagecache so
-	 * that userspace sees the changes.
-	 */
-	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
-	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
-		if (type != -1 && cnt != type)
-			continue;
-		if (!sb_has_quota_active(sb, cnt))
-			continue;
-		mutex_lock_nested(&sb_dqopt(sb)->files[cnt]->i_mutex,
-				  I_MUTEX_QUOTA);
-		truncate_inode_pages(&sb_dqopt(sb)->files[cnt]->i_data, 0);
-		mutex_unlock(&sb_dqopt(sb)->files[cnt]->i_mutex);
-	}
-	mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
+static int quota_setxquota(struct super_block *sb, int type, qid_t id,
+			   void __user *addr)
+{
+	struct fs_disk_quota fdq;
+
+	if (copy_from_user(&fdq, addr, sizeof(fdq)))
+		return -EFAULT;
+	if (!sb->s_qcop->set_xquota)
+		return -ENOSYS;
+	return sb->s_qcop->set_xquota(sb, type, id, &fdq);
 }
-#endif
 
-static void sync_dquots(int type)
+static int quota_getxquota(struct super_block *sb, int type, qid_t id,
+			   void __user *addr)
 {
-	struct super_block *sb;
-	int cnt;
+	struct fs_disk_quota fdq;
+	int ret;
 
-	spin_lock(&sb_lock);
-restart:
-	list_for_each_entry(sb, &super_blocks, s_list) {
-		/* This test just improves performance so it needn't be
-		 * reliable... */
-		for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
-			if (type != -1 && type != cnt)
-				continue;
-			if (!sb_has_quota_active(sb, cnt))
-				continue;
-			if (!info_dirty(&sb_dqopt(sb)->info[cnt]) &&
-			   list_empty(&sb_dqopt(sb)->info[cnt].dqi_dirty_list))
-				continue;
-			break;
-		}
-		if (cnt == MAXQUOTAS)
-			continue;
-		sb->s_count++;
-		spin_unlock(&sb_lock);
-		down_read(&sb->s_umount);
-		if (sb->s_root)
-			sync_quota_sb(sb, type);
-		up_read(&sb->s_umount);
-		spin_lock(&sb_lock);
-		if (__put_super_and_need_restart(sb))
-			goto restart;
-	}
-	spin_unlock(&sb_lock);
+	if (!sb->s_qcop->get_xquota)
+		return -ENOSYS;
+	ret = sb->s_qcop->get_xquota(sb, type, id, &fdq);
+	if (!ret && copy_to_user(addr, &fdq, sizeof(fdq)))
+		return -EFAULT;
+	return ret;
 }
 
 /* Copy parameters and call proper function */
@@ -240,117 +224,55 @@ static int do_quotactl(struct super_block *sb, int type, int cmd, qid_t id,
 {
 	int ret;
 
+	if (type >= (XQM_COMMAND(cmd) ? XQM_MAXQUOTAS : MAXQUOTAS))
+		return -EINVAL;
+	if (!sb->s_qcop)
+		return -ENOSYS;
+
+	ret = check_quotactl_permission(sb, type, cmd, id);
+	if (ret < 0)
+		return ret;
+
 	switch (cmd) {
-		case Q_QUOTAON: {
-			char *pathname;
-
-			pathname = getname(addr);
-			if (IS_ERR(pathname))
-				return PTR_ERR(pathname);
-			ret = sb->s_qcop->quota_on(sb, type, id, pathname, 0);
-			putname(pathname);
-			return ret;
-		}
-		case Q_QUOTAOFF:
-			return sb->s_qcop->quota_off(sb, type, 0);
-
-		case Q_GETFMT: {
-			__u32 fmt;
-
-			down_read(&sb_dqopt(sb)->dqptr_sem);
-			if (!sb_has_quota_active(sb, type)) {
-				up_read(&sb_dqopt(sb)->dqptr_sem);
-				return -ESRCH;
-			}
-			fmt = sb_dqopt(sb)->info[type].dqi_format->qf_fmt_id;
-			up_read(&sb_dqopt(sb)->dqptr_sem);
-			if (copy_to_user(addr, &fmt, sizeof(fmt)))
-				return -EFAULT;
-			return 0;
-		}
-		case Q_GETINFO: {
-			struct if_dqinfo info;
-
-			ret = sb->s_qcop->get_info(sb, type, &info);
-			if (ret)
-				return ret;
-			if (copy_to_user(addr, &info, sizeof(info)))
-				return -EFAULT;
-			return 0;
-		}
-		case Q_SETINFO: {
-			struct if_dqinfo info;
-
-			if (copy_from_user(&info, addr, sizeof(info)))
-				return -EFAULT;
-			return sb->s_qcop->set_info(sb, type, &info);
-		}
-		case Q_GETQUOTA: {
-			struct if_dqblk idq;
-
-			ret = sb->s_qcop->get_dqblk(sb, type, id, &idq);
-			if (ret)
-				return ret;
-			if (copy_to_user(addr, &idq, sizeof(idq)))
-				return -EFAULT;
-			return 0;
-		}
-		case Q_SETQUOTA: {
-			struct if_dqblk idq;
-
-			if (copy_from_user(&idq, addr, sizeof(idq)))
-				return -EFAULT;
-			return sb->s_qcop->set_dqblk(sb, type, id, &idq);
-		}
-		case Q_SYNC:
-			if (sb)
-				sync_quota_sb(sb, type);
-			else
-				sync_dquots(type);
-			return 0;
-
-		case Q_XQUOTAON:
-		case Q_XQUOTAOFF:
-		case Q_XQUOTARM: {
-			__u32 flags;
-
-			if (copy_from_user(&flags, addr, sizeof(flags)))
-				return -EFAULT;
-			return sb->s_qcop->set_xstate(sb, flags, cmd);
-		}
-		case Q_XGETQSTAT: {
-			struct fs_quota_stat fqs;
-		
-			if ((ret = sb->s_qcop->get_xstate(sb, &fqs)))
-				return ret;
-			if (copy_to_user(addr, &fqs, sizeof(fqs)))
-				return -EFAULT;
-			return 0;
-		}
-		case Q_XSETQLIM: {
-			struct fs_disk_quota fdq;
-
-			if (copy_from_user(&fdq, addr, sizeof(fdq)))
-				return -EFAULT;
-		       return sb->s_qcop->set_xquota(sb, type, id, &fdq);
-		}
-		case Q_XGETQUOTA: {
-			struct fs_disk_quota fdq;
-
-			ret = sb->s_qcop->get_xquota(sb, type, id, &fdq);
-			if (ret)
-				return ret;
-			if (copy_to_user(addr, &fdq, sizeof(fdq)))
-				return -EFAULT;
-			return 0;
-		}
-		case Q_XQUOTASYNC:
-			return sb->s_qcop->quota_sync(sb, type);
-		/* We never reach here unless validity check is broken */
-		default:
-			BUG();
+	case Q_QUOTAON:
+		return quota_quotaon(sb, type, cmd, id, addr);
+	case Q_QUOTAOFF:
+		if (!sb->s_qcop->quota_off)
+			return -ENOSYS;
+		return sb->s_qcop->quota_off(sb, type, 0);
+	case Q_GETFMT:
+		return quota_getfmt(sb, type, addr);
+	case Q_GETINFO:
+		return quota_getinfo(sb, type, addr);
+	case Q_SETINFO:
+		return quota_setinfo(sb, type, addr);
+	case Q_GETQUOTA:
+		return quota_getquota(sb, type, id, addr);
+	case Q_SETQUOTA:
+		return quota_setquota(sb, type, id, addr);
+	case Q_SYNC:
+		if (!sb->s_qcop->quota_sync)
+			return -ENOSYS;
+		return sb->s_qcop->quota_sync(sb, type, 1);
+	case Q_XQUOTAON:
+	case Q_XQUOTAOFF:
+	case Q_XQUOTARM:
+		return quota_setxstate(sb, cmd, addr);
+	case Q_XGETQSTAT:
+		return quota_getxstate(sb, addr);
+	case Q_XSETQLIM:
+		return quota_setxquota(sb, type, id, addr);
+	case Q_XGETQUOTA:
+		return quota_getxquota(sb, type, id, addr);
+	case Q_XQUOTASYNC:
+		/* caller already holds s_umount */
+		if (sb->s_flags & MS_RDONLY)
+			return -EROFS;
+		writeback_inodes_sb(sb);
+		return 0;
+	default:
+		return -EINVAL;
 	}
-	return 0;
 }
 
 /*
@@ -397,224 +319,23 @@ SYSCALL_DEFINE4(quotactl, unsigned int, cmd, const char __user *, special,
 	cmds = cmd >> SUBCMDSHIFT;
 	type = cmd & SUBCMDMASK;
 
-	if (cmds != Q_SYNC || special) {
-		sb = quotactl_block(special);
-		if (IS_ERR(sb))
-			return PTR_ERR(sb);
+	/*
+	 * As a special case Q_SYNC can be called without a specific device.
+	 * It will iterate all superblocks that have quota enabled and call
+	 * the sync action on each of them.
+	 */
+	if (!special) {
+		if (cmds == Q_SYNC)
+			return quota_sync_all(type);
+		return -ENODEV;
 	}
 
-	ret = check_quotactl_valid(sb, type, cmds, id);
-	if (ret >= 0)
-		ret = do_quotactl(sb, type, cmds, id, addr);
-	if (sb)
-		drop_super(sb);
+	sb = quotactl_block(special);
+	if (IS_ERR(sb))
+		return PTR_ERR(sb);
 
-	return ret;
-}
-
-#if defined(CONFIG_COMPAT_FOR_U64_ALIGNMENT)
-/*
- * This code works only for 32 bit quota tools over 64 bit OS (x86_64, ia64)
- * and is necessary due to alignment problems.
- */
-struct compat_if_dqblk {
-	compat_u64 dqb_bhardlimit;
-	compat_u64 dqb_bsoftlimit;
-	compat_u64 dqb_curspace;
-	compat_u64 dqb_ihardlimit;
-	compat_u64 dqb_isoftlimit;
-	compat_u64 dqb_curinodes;
-	compat_u64 dqb_btime;
-	compat_u64 dqb_itime;
-	compat_uint_t dqb_valid;
-};
-
-/* XFS structures */
-struct compat_fs_qfilestat {
-	compat_u64 dqb_bhardlimit;
-	compat_u64 qfs_nblks;
-	compat_uint_t qfs_nextents;
-};
-
-struct compat_fs_quota_stat {
-	__s8		qs_version;
-	__u16		qs_flags;
-	__s8		qs_pad;
-	struct compat_fs_qfilestat	qs_uquota;
-	struct compat_fs_qfilestat	qs_gquota;
-	compat_uint_t	qs_incoredqs;
-	compat_int_t	qs_btimelimit;
-	compat_int_t	qs_itimelimit;
-	compat_int_t	qs_rtbtimelimit;
-	__u16		qs_bwarnlimit;
-	__u16		qs_iwarnlimit;
-};
-
-asmlinkage long sys32_quotactl(unsigned int cmd, const char __user *special,
-						qid_t id, void __user *addr)
-{
-	unsigned int cmds;
-	struct if_dqblk __user *dqblk;
-	struct compat_if_dqblk __user *compat_dqblk;
-	struct fs_quota_stat __user *fsqstat;
-	struct compat_fs_quota_stat __user *compat_fsqstat;
-	compat_uint_t data;
-	u16 xdata;
-	long ret;
+	ret = do_quotactl(sb, type, cmds, id, addr);
 
-	cmds = cmd >> SUBCMDSHIFT;
-
-	switch (cmds) {
-	case Q_GETQUOTA:
-		dqblk = compat_alloc_user_space(sizeof(struct if_dqblk));
-		compat_dqblk = addr;
-		ret = sys_quotactl(cmd, special, id, dqblk);
-		if (ret)
-			break;
-		if (copy_in_user(compat_dqblk, dqblk, sizeof(*compat_dqblk)) ||
-			get_user(data, &dqblk->dqb_valid) ||
-			put_user(data, &compat_dqblk->dqb_valid))
-			ret = -EFAULT;
-		break;
-	case Q_SETQUOTA:
-		dqblk = compat_alloc_user_space(sizeof(struct if_dqblk));
-		compat_dqblk = addr;
-		ret = -EFAULT;
-		if (copy_in_user(dqblk, compat_dqblk, sizeof(*compat_dqblk)) ||
-			get_user(data, &compat_dqblk->dqb_valid) ||
-			put_user(data, &dqblk->dqb_valid))
-			break;
-		ret = sys_quotactl(cmd, special, id, dqblk);
-		break;
-	case Q_XGETQSTAT:
-		fsqstat = compat_alloc_user_space(sizeof(struct fs_quota_stat));
-		compat_fsqstat = addr;
-		ret = sys_quotactl(cmd, special, id, fsqstat);
-		if (ret)
-			break;
-		ret = -EFAULT;
-		/* Copying qs_version, qs_flags, qs_pad */
-		if (copy_in_user(compat_fsqstat, fsqstat,
-			offsetof(struct compat_fs_quota_stat, qs_uquota)))
-			break;
-		/* Copying qs_uquota */
-		if (copy_in_user(&compat_fsqstat->qs_uquota,
-			&fsqstat->qs_uquota,
-			sizeof(compat_fsqstat->qs_uquota)) ||
-			get_user(data, &fsqstat->qs_uquota.qfs_nextents) ||
-			put_user(data, &compat_fsqstat->qs_uquota.qfs_nextents))
-			break;
-		/* Copying qs_gquota */
-		if (copy_in_user(&compat_fsqstat->qs_gquota,
-			&fsqstat->qs_gquota,
-			sizeof(compat_fsqstat->qs_gquota)) ||
-			get_user(data, &fsqstat->qs_gquota.qfs_nextents) ||
-			put_user(data, &compat_fsqstat->qs_gquota.qfs_nextents))
-			break;
-		/* Copying the rest */
-		if (copy_in_user(&compat_fsqstat->qs_incoredqs,
-			&fsqstat->qs_incoredqs,
-			sizeof(struct compat_fs_quota_stat) -
-			offsetof(struct compat_fs_quota_stat, qs_incoredqs)) ||
-			get_user(xdata, &fsqstat->qs_iwarnlimit) ||
-			put_user(xdata, &compat_fsqstat->qs_iwarnlimit))
-			break;
-		ret = 0;
-		break;
-	default:
-		ret = sys_quotactl(cmd, special, id, addr);
-	}
+	drop_super(sb);
 	return ret;
 }
-#endif
-
-
-#ifdef CONFIG_QUOTA_NETLINK_INTERFACE
-
-/* Netlink family structure for quota */
-static struct genl_family quota_genl_family = {
-	.id = GENL_ID_GENERATE,
-	.hdrsize = 0,
-	.name = "VFS_DQUOT",
-	.version = 1,
-	.maxattr = QUOTA_NL_A_MAX,
-};
-
-/**
- * quota_send_warning - Send warning to userspace about exceeded quota
- * @type: The quota type: USRQQUOTA, GRPQUOTA,...
- * @id: The user or group id of the quota that was exceeded
- * @dev: The device on which the fs is mounted (sb->s_dev)
- * @warntype: The type of the warning: QUOTA_NL_...
- *
- * This can be used by filesystems (including those which don't use
- * dquot) to send a message to userspace relating to quota limits.
- *
- */
-
-void quota_send_warning(short type, unsigned int id, dev_t dev,
-			const char warntype)
-{
-	static atomic_t seq;
-	struct sk_buff *skb;
-	void *msg_head;
-	int ret;
-	int msg_size = 4 * nla_total_size(sizeof(u32)) +
-		       2 * nla_total_size(sizeof(u64));
-
-	/* We have to allocate using GFP_NOFS as we are called from a
-	 * filesystem performing write and thus further recursion into
-	 * the fs to free some data could cause deadlocks. */
-	skb = genlmsg_new(msg_size, GFP_NOFS);
-	if (!skb) {
-		printk(KERN_ERR
-		  "VFS: Not enough memory to send quota warning.\n");
-		return;
-	}
-	msg_head = genlmsg_put(skb, 0, atomic_add_return(1, &seq),
-			&quota_genl_family, 0, QUOTA_NL_C_WARNING);
-	if (!msg_head) {
-		printk(KERN_ERR
-		  "VFS: Cannot store netlink header in quota warning.\n");
-		goto err_out;
-	}
-	ret = nla_put_u32(skb, QUOTA_NL_A_QTYPE, type);
-	if (ret)
-		goto attr_err_out;
-	ret = nla_put_u64(skb, QUOTA_NL_A_EXCESS_ID, id);
-	if (ret)
-		goto attr_err_out;
-	ret = nla_put_u32(skb, QUOTA_NL_A_WARNING, warntype);
-	if (ret)
-		goto attr_err_out;
-	ret = nla_put_u32(skb, QUOTA_NL_A_DEV_MAJOR, MAJOR(dev));
-	if (ret)
-		goto attr_err_out;
-	ret = nla_put_u32(skb, QUOTA_NL_A_DEV_MINOR, MINOR(dev));
-	if (ret)
-		goto attr_err_out;
-	ret = nla_put_u64(skb, QUOTA_NL_A_CAUSED_ID, current_uid());
-	if (ret)
-		goto attr_err_out;
-	genlmsg_end(skb, msg_head);
-
-	genlmsg_multicast(skb, 0, quota_genl_family.id, GFP_NOFS);
-	return;
-attr_err_out:
-	printk(KERN_ERR "VFS: Not enough space to compose quota message!\n");
-err_out:
-	kfree_skb(skb);
-}
-EXPORT_SYMBOL(quota_send_warning);
-
-static int __init quota_init(void)
-{
-	if (genl_register_family(&quota_genl_family) != 0)
-		printk(KERN_ERR
-		       "VFS: Failed to create quota netlink interface.\n");
-	return 0;
-};
-
-module_init(quota_init);
-#endif
-
diff --git a/fs/reiserfs/bitmap.c b/fs/reiserfs/bitmap.c
index 65c872761177..dc014f7def05 100644
--- a/fs/reiserfs/bitmap.c
+++ b/fs/reiserfs/bitmap.c
@@ -425,7 +425,7 @@ static void _reiserfs_free_block(struct reiserfs_transaction_handle *th,
 
 	journal_mark_dirty(th, s, sbh);
 	if (for_unformatted)
-		vfs_dq_free_block_nodirty(inode, 1);
+		dquot_free_block_nodirty(inode, 1);
 }
 
 void reiserfs_free_block(struct reiserfs_transaction_handle *th,
@@ -1049,7 +1049,7 @@ static inline int blocknrs_and_prealloc_arrays_from_search_start
 			       amount_needed, hint->inode->i_uid);
 #endif
 		quota_ret =
-		    vfs_dq_alloc_block_nodirty(hint->inode, amount_needed);
+		    dquot_alloc_block_nodirty(hint->inode, amount_needed);
 		if (quota_ret)	/* Quota exceeded? */
 			return QUOTA_EXCEEDED;
 		if (hint->preallocate && hint->prealloc_size) {
@@ -1058,7 +1058,7 @@ static inline int blocknrs_and_prealloc_arrays_from_search_start
 				       "reiserquota: allocating (prealloc) %d blocks id=%u",
 				       hint->prealloc_size, hint->inode->i_uid);
 #endif
-			quota_ret = vfs_dq_prealloc_block_nodirty(hint->inode,
+			quota_ret = dquot_prealloc_block_nodirty(hint->inode,
 							 hint->prealloc_size);
 			if (quota_ret)
 				hint->preallocate = hint->prealloc_size = 0;
@@ -1092,7 +1092,7 @@ static inline int blocknrs_and_prealloc_arrays_from_search_start
 					       hint->inode->i_uid);
 #endif
 				/* Free not allocated blocks */
-				vfs_dq_free_block_nodirty(hint->inode,
+				dquot_free_block_nodirty(hint->inode,
 					amount_needed + hint->prealloc_size -
 					nr_allocated);
 			}
@@ -1125,7 +1125,7 @@ static inline int blocknrs_and_prealloc_arrays_from_search_start
 			       REISERFS_I(hint->inode)->i_prealloc_count,
 			       hint->inode->i_uid);
 #endif
-		vfs_dq_free_block_nodirty(hint->inode, amount_needed +
+		dquot_free_block_nodirty(hint->inode, amount_needed +
 					 hint->prealloc_size - nr_allocated -
 					 REISERFS_I(hint->inode)->
 					 i_prealloc_count);
diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c
index da2dba082e2d..1d9c12714c5c 100644
--- a/fs/reiserfs/file.c
+++ b/fs/reiserfs/file.c
@@ -289,7 +289,7 @@ const struct file_operations reiserfs_file_operations = {
 	.compat_ioctl = reiserfs_compat_ioctl,
 #endif
 	.mmap = reiserfs_file_mmap,
-	.open = generic_file_open,
+	.open = dquot_file_open,
 	.release = reiserfs_file_release,
 	.fsync = reiserfs_sync_file,
 	.aio_read = generic_file_aio_read,
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 2df0f5c7c60b..d1da94b82d8f 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -34,6 +34,9 @@ void reiserfs_delete_inode(struct inode *inode)
 	int depth;
 	int err;
 
+	if (!is_bad_inode(inode))
+		dquot_initialize(inode);
+
 	truncate_inode_pages(&inode->i_data, 0);
 
 	depth = reiserfs_write_lock_once(inode->i_sb);
@@ -54,7 +57,7 @@ void reiserfs_delete_inode(struct inode *inode)
 		 * after delete_object so that quota updates go into the same transaction as
 		 * stat data deletion */
 		if (!err) 
-			vfs_dq_free_inode(inode);
+			dquot_free_inode(inode);
 
 		if (journal_end(&th, inode->i_sb, jbegin_count))
 			goto out;
@@ -1615,7 +1618,7 @@ int reiserfs_encode_fh(struct dentry *dentry, __u32 * data, int *lenp,
 ** to properly mark inodes for datasync and such, but only actually
 ** does something when called for a synchronous update.
 */
-int reiserfs_write_inode(struct inode *inode, int do_sync)
+int reiserfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct reiserfs_transaction_handle th;
 	int jbegin_count = 1;
@@ -1627,7 +1630,7 @@ int reiserfs_write_inode(struct inode *inode, int do_sync)
 	 ** inode needs to reach disk for safety, and they can safely be
 	 ** ignored because the altered inode has already been logged.
 	 */
-	if (do_sync && !(current->flags & PF_MEMALLOC)) {
+	if (wbc->sync_mode == WB_SYNC_ALL && !(current->flags & PF_MEMALLOC)) {
 		reiserfs_write_lock(inode->i_sb);
 		if (!journal_begin(&th, inode->i_sb, jbegin_count)) {
 			reiserfs_update_sd(&th, inode);
@@ -1765,10 +1768,10 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th,
 
 	BUG_ON(!th->t_trans_id);
 
-	if (vfs_dq_alloc_inode(inode)) {
-		err = -EDQUOT;
+	dquot_initialize(inode);
+	err = dquot_alloc_inode(inode);
+	if (err)
 		goto out_end_trans;
-	}
 	if (!dir->i_nlink) {
 		err = -EPERM;
 		goto out_bad_inode;
@@ -1959,12 +1962,12 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th,
 	INODE_PKEY(inode)->k_objectid = 0;
 
 	/* Quota change must be inside a transaction for journaling */
-	vfs_dq_free_inode(inode);
+	dquot_free_inode(inode);
 
       out_end_trans:
 	journal_end(th, th->t_super, th->t_blocks_allocated);
 	/* Drop can be outside and it needs more credits so it's better to have it outside */
-	vfs_dq_drop(inode);
+	dquot_drop(inode);
 	inode->i_flags |= S_NOQUOTA;
 	make_bad_inode(inode);
 
@@ -3073,6 +3076,8 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr)
 
 	depth = reiserfs_write_lock_once(inode->i_sb);
 	if (attr->ia_valid & ATTR_SIZE) {
+		dquot_initialize(inode);
+
 		/* version 2 items will be caught by the s_maxbytes check
 		 ** done for us in vmtruncate
 		 */
@@ -3134,8 +3139,7 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr)
 						  jbegin_count);
 				if (error)
 					goto out;
-				error =
-				    vfs_dq_transfer(inode, attr) ? -EDQUOT : 0;
+				error = dquot_transfer(inode, attr);
 				if (error) {
 					journal_end(&th, inode->i_sb,
 						    jbegin_count);
diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c
index 9d4dcf0b07cb..96e4cbbfaa18 100644
--- a/fs/reiserfs/namei.c
+++ b/fs/reiserfs/namei.c
@@ -546,7 +546,7 @@ static int reiserfs_add_entry(struct reiserfs_transaction_handle *th,
 */
 static int drop_new_inode(struct inode *inode)
 {
-	vfs_dq_drop(inode);
+	dquot_drop(inode);
 	make_bad_inode(inode);
 	inode->i_flags |= S_NOQUOTA;
 	iput(inode);
@@ -554,7 +554,7 @@ static int drop_new_inode(struct inode *inode)
 }
 
 /* utility function that does setup for reiserfs_new_inode.
-** vfs_dq_init needs lots of credits so it's better to have it
+** dquot_initialize needs lots of credits so it's better to have it
 ** outside of a transaction, so we had to pull some bits of
 ** reiserfs_new_inode out into this func.
 */
@@ -577,7 +577,7 @@ static int new_inode_init(struct inode *inode, struct inode *dir, int mode)
 	} else {
 		inode->i_gid = current_fsgid();
 	}
-	vfs_dq_init(inode);
+	dquot_initialize(inode);
 	return 0;
 }
 
@@ -594,6 +594,8 @@ static int reiserfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	struct reiserfs_transaction_handle th;
 	struct reiserfs_security_handle security;
 
+	dquot_initialize(dir);
+
 	if (!(inode = new_inode(dir->i_sb))) {
 		return -ENOMEM;
 	}
@@ -666,6 +668,8 @@ static int reiserfs_mknod(struct inode *dir, struct dentry *dentry, int mode,
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
 
+	dquot_initialize(dir);
+
 	if (!(inode = new_inode(dir->i_sb))) {
 		return -ENOMEM;
 	}
@@ -739,6 +743,8 @@ static int reiserfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	    2 * (REISERFS_QUOTA_INIT_BLOCKS(dir->i_sb) +
 		 REISERFS_QUOTA_TRANS_BLOCKS(dir->i_sb));
 
+	dquot_initialize(dir);
+
 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
 	/* set flag that new packing locality created and new blocks for the content     * of that directory are not displaced yet */
 	REISERFS_I(dir)->new_packing_locality = 1;
@@ -842,6 +848,8 @@ static int reiserfs_rmdir(struct inode *dir, struct dentry *dentry)
 	    JOURNAL_PER_BALANCE_CNT * 2 + 2 +
 	    4 * REISERFS_QUOTA_TRANS_BLOCKS(dir->i_sb);
 
+	dquot_initialize(dir);
+
 	reiserfs_write_lock(dir->i_sb);
 	retval = journal_begin(&th, dir->i_sb, jbegin_count);
 	if (retval)
@@ -923,6 +931,8 @@ static int reiserfs_unlink(struct inode *dir, struct dentry *dentry)
 	unsigned long savelink;
 	int depth;
 
+	dquot_initialize(dir);
+
 	inode = dentry->d_inode;
 
 	/* in this transaction we can be doing at max two balancings and update
@@ -1024,6 +1034,8 @@ static int reiserfs_symlink(struct inode *parent_dir,
 	    2 * (REISERFS_QUOTA_INIT_BLOCKS(parent_dir->i_sb) +
 		 REISERFS_QUOTA_TRANS_BLOCKS(parent_dir->i_sb));
 
+	dquot_initialize(parent_dir);
+
 	if (!(inode = new_inode(parent_dir->i_sb))) {
 		return -ENOMEM;
 	}
@@ -1111,6 +1123,8 @@ static int reiserfs_link(struct dentry *old_dentry, struct inode *dir,
 	    JOURNAL_PER_BALANCE_CNT * 3 +
 	    2 * REISERFS_QUOTA_TRANS_BLOCKS(dir->i_sb);
 
+	dquot_initialize(dir);
+
 	reiserfs_write_lock(dir->i_sb);
 	if (inode->i_nlink >= REISERFS_LINK_MAX) {
 		//FIXME: sd_nlink is 32 bit for new files
@@ -1235,6 +1249,9 @@ static int reiserfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	    JOURNAL_PER_BALANCE_CNT * 3 + 5 +
 	    4 * REISERFS_QUOTA_TRANS_BLOCKS(old_dir->i_sb);
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	old_inode = old_dentry->d_inode;
 	new_dentry_inode = new_dentry->d_inode;
 
diff --git a/fs/reiserfs/stree.c b/fs/reiserfs/stree.c
index 5fa7118f04e1..313d39d639eb 100644
--- a/fs/reiserfs/stree.c
+++ b/fs/reiserfs/stree.c
@@ -1299,7 +1299,7 @@ int reiserfs_delete_item(struct reiserfs_transaction_handle *th,
 		       "reiserquota delete_item(): freeing %u, id=%u type=%c",
 		       quota_cut_bytes, inode->i_uid, head2type(&s_ih));
 #endif
-	vfs_dq_free_space_nodirty(inode, quota_cut_bytes);
+	dquot_free_space_nodirty(inode, quota_cut_bytes);
 
 	/* Return deleted body length */
 	return ret_value;
@@ -1383,7 +1383,7 @@ void reiserfs_delete_solid_item(struct reiserfs_transaction_handle *th,
 					       quota_cut_bytes, inode->i_uid,
 					       key2type(key));
 #endif
-				vfs_dq_free_space_nodirty(inode,
+				dquot_free_space_nodirty(inode,
 							 quota_cut_bytes);
 			}
 			break;
@@ -1733,7 +1733,7 @@ int reiserfs_cut_from_item(struct reiserfs_transaction_handle *th,
 		       "reiserquota cut_from_item(): freeing %u id=%u type=%c",
 		       quota_cut_bytes, inode->i_uid, '?');
 #endif
-	vfs_dq_free_space_nodirty(inode, quota_cut_bytes);
+	dquot_free_space_nodirty(inode, quota_cut_bytes);
 	return ret_value;
 }
 
@@ -1968,9 +1968,10 @@ int reiserfs_paste_into_item(struct reiserfs_transaction_handle *th, struct tree
 		       key2type(&(key->on_disk_key)));
 #endif
 
-	if (vfs_dq_alloc_space_nodirty(inode, pasted_size)) {
+	retval = dquot_alloc_space_nodirty(inode, pasted_size);
+	if (retval) {
 		pathrelse(search_path);
-		return -EDQUOT;
+		return retval;
 	}
 	init_tb_struct(th, &s_paste_balance, th->t_super, search_path,
 		       pasted_size);
@@ -2024,7 +2025,7 @@ int reiserfs_paste_into_item(struct reiserfs_transaction_handle *th, struct tree
 		       pasted_size, inode->i_uid,
 		       key2type(&(key->on_disk_key)));
 #endif
-	vfs_dq_free_space_nodirty(inode, pasted_size);
+	dquot_free_space_nodirty(inode, pasted_size);
 	return retval;
 }
 
@@ -2062,9 +2063,10 @@ int reiserfs_insert_item(struct reiserfs_transaction_handle *th,
 #endif
 		/* We can't dirty inode here. It would be immediately written but
 		 * appropriate stat item isn't inserted yet... */
-		if (vfs_dq_alloc_space_nodirty(inode, quota_bytes)) {
+		retval = dquot_alloc_space_nodirty(inode, quota_bytes);
+		if (retval) {
 			pathrelse(path);
-			return -EDQUOT;
+			return retval;
 		}
 	}
 	init_tb_struct(th, &s_ins_balance, th->t_super, path,
@@ -2113,6 +2115,6 @@ int reiserfs_insert_item(struct reiserfs_transaction_handle *th,
 		       quota_bytes, inode->i_uid, head2type(ih));
 #endif
 	if (inode)
-		vfs_dq_free_space_nodirty(inode, quota_bytes);
+		dquot_free_space_nodirty(inode, quota_bytes);
 	return retval;
 }
diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c
index b4a7dd03bdb9..04bf5d791bda 100644
--- a/fs/reiserfs/super.c
+++ b/fs/reiserfs/super.c
@@ -246,7 +246,7 @@ static int finish_unfinished(struct super_block *s)
 			retval = remove_save_link_only(s, &save_link_key, 0);
 			continue;
 		}
-		vfs_dq_init(inode);
+		dquot_initialize(inode);
 
 		if (truncate && S_ISDIR(inode->i_mode)) {
 			/* We got a truncate request for a dir which is impossible.
@@ -578,6 +578,11 @@ out:
 	reiserfs_write_unlock_once(inode->i_sb, lock_depth);
 }
 
+static void reiserfs_clear_inode(struct inode *inode)
+{
+	dquot_drop(inode);
+}
+
 #ifdef CONFIG_QUOTA
 static ssize_t reiserfs_quota_write(struct super_block *, int, const char *,
 				    size_t, loff_t);
@@ -590,6 +595,7 @@ static const struct super_operations reiserfs_sops = {
 	.destroy_inode = reiserfs_destroy_inode,
 	.write_inode = reiserfs_write_inode,
 	.dirty_inode = reiserfs_dirty_inode,
+	.clear_inode = reiserfs_clear_inode,
 	.delete_inode = reiserfs_delete_inode,
 	.put_super = reiserfs_put_super,
 	.write_super = reiserfs_write_super,
@@ -616,13 +622,6 @@ static int reiserfs_write_info(struct super_block *, int);
 static int reiserfs_quota_on(struct super_block *, int, int, char *, int);
 
 static const struct dquot_operations reiserfs_quota_operations = {
-	.initialize = dquot_initialize,
-	.drop = dquot_drop,
-	.alloc_space = dquot_alloc_space,
-	.alloc_inode = dquot_alloc_inode,
-	.free_space = dquot_free_space,
-	.free_inode = dquot_free_inode,
-	.transfer = dquot_transfer,
 	.write_dquot = reiserfs_write_dquot,
 	.acquire_dquot = reiserfs_acquire_dquot,
 	.release_dquot = reiserfs_release_dquot,
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index 81f09fab8ae4..37d034ca7d99 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -61,7 +61,6 @@
 static int xattr_create(struct inode *dir, struct dentry *dentry, int mode)
 {
 	BUG_ON(!mutex_is_locked(&dir->i_mutex));
-	vfs_dq_init(dir);
 	return dir->i_op->create(dir, dentry, mode, NULL);
 }
 #endif
@@ -69,7 +68,6 @@ static int xattr_create(struct inode *dir, struct dentry *dentry, int mode)
 static int xattr_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 {
 	BUG_ON(!mutex_is_locked(&dir->i_mutex));
-	vfs_dq_init(dir);
 	return dir->i_op->mkdir(dir, dentry, mode);
 }
 
@@ -81,7 +79,6 @@ static int xattr_unlink(struct inode *dir, struct dentry *dentry)
 {
 	int error;
 	BUG_ON(!mutex_is_locked(&dir->i_mutex));
-	vfs_dq_init(dir);
 
 	reiserfs_mutex_lock_nested_safe(&dentry->d_inode->i_mutex,
 					I_MUTEX_CHILD, dir->i_sb);
@@ -97,7 +94,6 @@ static int xattr_rmdir(struct inode *dir, struct dentry *dentry)
 {
 	int error;
 	BUG_ON(!mutex_is_locked(&dir->i_mutex));
-	vfs_dq_init(dir);
 
 	reiserfs_mutex_lock_nested_safe(&dentry->d_inode->i_mutex,
 					I_MUTEX_CHILD, dir->i_sb);
diff --git a/fs/select.c b/fs/select.c
index fd38ce2e32e3..73715e90030f 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -821,7 +821,7 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds,
  	struct poll_list *walk = head;
  	unsigned long todo = nfds;
 
-	if (nfds > current->signal->rlim[RLIMIT_NOFILE].rlim_cur)
+	if (nfds > rlimit(RLIMIT_NOFILE))
 		return -EINVAL;
 
 	len = min_t(unsigned int, nfds, N_STACK_PPS);
diff --git a/fs/seq_file.c b/fs/seq_file.c
index eae7d9dbf3ff..e1f437be6c3c 100644
--- a/fs/seq_file.c
+++ b/fs/seq_file.c
@@ -674,7 +674,6 @@ struct list_head *seq_list_start(struct list_head *head, loff_t pos)
 
 	return NULL;
 }
-
 EXPORT_SYMBOL(seq_list_start);
 
 struct list_head *seq_list_start_head(struct list_head *head, loff_t pos)
@@ -684,7 +683,6 @@ struct list_head *seq_list_start_head(struct list_head *head, loff_t pos)
 
 	return seq_list_start(head, pos - 1);
 }
-
 EXPORT_SYMBOL(seq_list_start_head);
 
 struct list_head *seq_list_next(void *v, struct list_head *head, loff_t *ppos)
@@ -695,5 +693,131 @@ struct list_head *seq_list_next(void *v, struct list_head *head, loff_t *ppos)
 	++*ppos;
 	return lh == head ? NULL : lh;
 }
-
 EXPORT_SYMBOL(seq_list_next);
+
+/**
+ * seq_hlist_start - start an iteration of a hlist
+ * @head: the head of the hlist
+ * @pos:  the start position of the sequence
+ *
+ * Called at seq_file->op->start().
+ */
+struct hlist_node *seq_hlist_start(struct hlist_head *head, loff_t pos)
+{
+	struct hlist_node *node;
+
+	hlist_for_each(node, head)
+		if (pos-- == 0)
+			return node;
+	return NULL;
+}
+EXPORT_SYMBOL(seq_hlist_start);
+
+/**
+ * seq_hlist_start_head - start an iteration of a hlist
+ * @head: the head of the hlist
+ * @pos:  the start position of the sequence
+ *
+ * Called at seq_file->op->start(). Call this function if you want to
+ * print a header at the top of the output.
+ */
+struct hlist_node *seq_hlist_start_head(struct hlist_head *head, loff_t pos)
+{
+	if (!pos)
+		return SEQ_START_TOKEN;
+
+	return seq_hlist_start(head, pos - 1);
+}
+EXPORT_SYMBOL(seq_hlist_start_head);
+
+/**
+ * seq_hlist_next - move to the next position of the hlist
+ * @v:    the current iterator
+ * @head: the head of the hlist
+ * @ppos: the current position
+ *
+ * Called at seq_file->op->next().
+ */
+struct hlist_node *seq_hlist_next(void *v, struct hlist_head *head,
+				  loff_t *ppos)
+{
+	struct hlist_node *node = v;
+
+	++*ppos;
+	if (v == SEQ_START_TOKEN)
+		return head->first;
+	else
+		return node->next;
+}
+EXPORT_SYMBOL(seq_hlist_next);
+
+/**
+ * seq_hlist_start_rcu - start an iteration of a hlist protected by RCU
+ * @head: the head of the hlist
+ * @pos:  the start position of the sequence
+ *
+ * Called at seq_file->op->start().
+ *
+ * This list-traversal primitive may safely run concurrently with
+ * the _rcu list-mutation primitives such as hlist_add_head_rcu()
+ * as long as the traversal is guarded by rcu_read_lock().
+ */
+struct hlist_node *seq_hlist_start_rcu(struct hlist_head *head,
+				       loff_t pos)
+{
+	struct hlist_node *node;
+
+	__hlist_for_each_rcu(node, head)
+		if (pos-- == 0)
+			return node;
+	return NULL;
+}
+EXPORT_SYMBOL(seq_hlist_start_rcu);
+
+/**
+ * seq_hlist_start_head_rcu - start an iteration of a hlist protected by RCU
+ * @head: the head of the hlist
+ * @pos:  the start position of the sequence
+ *
+ * Called at seq_file->op->start(). Call this function if you want to
+ * print a header at the top of the output.
+ *
+ * This list-traversal primitive may safely run concurrently with
+ * the _rcu list-mutation primitives such as hlist_add_head_rcu()
+ * as long as the traversal is guarded by rcu_read_lock().
+ */
+struct hlist_node *seq_hlist_start_head_rcu(struct hlist_head *head,
+					    loff_t pos)
+{
+	if (!pos)
+		return SEQ_START_TOKEN;
+
+	return seq_hlist_start_rcu(head, pos - 1);
+}
+EXPORT_SYMBOL(seq_hlist_start_head_rcu);
+
+/**
+ * seq_hlist_next_rcu - move to the next position of the hlist protected by RCU
+ * @v:    the current iterator
+ * @head: the head of the hlist
+ * @ppos: the current position
+ *
+ * Called at seq_file->op->next().
+ *
+ * This list-traversal primitive may safely run concurrently with
+ * the _rcu list-mutation primitives such as hlist_add_head_rcu()
+ * as long as the traversal is guarded by rcu_read_lock().
+ */
+struct hlist_node *seq_hlist_next_rcu(void *v,
+				      struct hlist_head *head,
+				      loff_t *ppos)
+{
+	struct hlist_node *node = v;
+
+	++*ppos;
+	if (v == SEQ_START_TOKEN)
+		return rcu_dereference(head->first);
+	else
+		return rcu_dereference(node->next);
+}
+EXPORT_SYMBOL(seq_hlist_next_rcu);
diff --git a/fs/squashfs/Makefile b/fs/squashfs/Makefile
index 70e3244fa30f..df8a19ef870d 100644
--- a/fs/squashfs/Makefile
+++ b/fs/squashfs/Makefile
@@ -4,4 +4,4 @@
 
 obj-$(CONFIG_SQUASHFS) += squashfs.o
 squashfs-y += block.o cache.o dir.o export.o file.o fragment.o id.o inode.o
-squashfs-y += namei.o super.o symlink.o
+squashfs-y += namei.o super.o symlink.o zlib_wrapper.o decompressor.o
diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c
index 2a7960310349..1cb0d81b164b 100644
--- a/fs/squashfs/block.c
+++ b/fs/squashfs/block.c
@@ -29,15 +29,14 @@
 #include <linux/fs.h>
 #include <linux/vfs.h>
 #include <linux/slab.h>
-#include <linux/mutex.h>
 #include <linux/string.h>
 #include <linux/buffer_head.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
 #include "squashfs_fs_i.h"
 #include "squashfs.h"
+#include "decompressor.h"
 
 /*
  * Read the metadata block length, this is stored in the first two
@@ -153,72 +152,10 @@ int squashfs_read_data(struct super_block *sb, void **buffer, u64 index,
 	}
 
 	if (compressed) {
-		int zlib_err = 0, zlib_init = 0;
-
-		/*
-		 * Uncompress block.
-		 */
-
-		mutex_lock(&msblk->read_data_mutex);
-
-		msblk->stream.avail_out = 0;
-		msblk->stream.avail_in = 0;
-
-		bytes = length;
-		do {
-			if (msblk->stream.avail_in == 0 && k < b) {
-				avail = min(bytes, msblk->devblksize - offset);
-				bytes -= avail;
-				wait_on_buffer(bh[k]);
-				if (!buffer_uptodate(bh[k]))
-					goto release_mutex;
-
-				if (avail == 0) {
-					offset = 0;
-					put_bh(bh[k++]);
-					continue;
-				}
-
-				msblk->stream.next_in = bh[k]->b_data + offset;
-				msblk->stream.avail_in = avail;
-				offset = 0;
-			}
-
-			if (msblk->stream.avail_out == 0 && page < pages) {
-				msblk->stream.next_out = buffer[page++];
-				msblk->stream.avail_out = PAGE_CACHE_SIZE;
-			}
-
-			if (!zlib_init) {
-				zlib_err = zlib_inflateInit(&msblk->stream);
-				if (zlib_err != Z_OK) {
-					ERROR("zlib_inflateInit returned"
-						" unexpected result 0x%x,"
-						" srclength %d\n", zlib_err,
-						srclength);
-					goto release_mutex;
-				}
-				zlib_init = 1;
-			}
-
-			zlib_err = zlib_inflate(&msblk->stream, Z_SYNC_FLUSH);
-
-			if (msblk->stream.avail_in == 0 && k < b)
-				put_bh(bh[k++]);
-		} while (zlib_err == Z_OK);
-
-		if (zlib_err != Z_STREAM_END) {
-			ERROR("zlib_inflate error, data probably corrupt\n");
-			goto release_mutex;
-		}
-
-		zlib_err = zlib_inflateEnd(&msblk->stream);
-		if (zlib_err != Z_OK) {
-			ERROR("zlib_inflate error, data probably corrupt\n");
-			goto release_mutex;
-		}
-		length = msblk->stream.total_out;
-		mutex_unlock(&msblk->read_data_mutex);
+		length = squashfs_decompress(msblk, buffer, bh, b, offset,
+			 length, srclength, pages);
+		if (length < 0)
+			goto read_failure;
 	} else {
 		/*
 		 * Block is uncompressed.
@@ -255,9 +192,6 @@ int squashfs_read_data(struct super_block *sb, void **buffer, u64 index,
 	kfree(bh);
 	return length;
 
-release_mutex:
-	mutex_unlock(&msblk->read_data_mutex);
-
 block_release:
 	for (; k < b; k++)
 		put_bh(bh[k]);
diff --git a/fs/squashfs/cache.c b/fs/squashfs/cache.c
index 40c98fa6b5d6..57314bee9059 100644
--- a/fs/squashfs/cache.c
+++ b/fs/squashfs/cache.c
@@ -51,7 +51,6 @@
 #include <linux/sched.h>
 #include <linux/spinlock.h>
 #include <linux/wait.h>
-#include <linux/zlib.h>
 #include <linux/pagemap.h>
 
 #include "squashfs_fs.h"
diff --git a/fs/squashfs/decompressor.c b/fs/squashfs/decompressor.c
new file mode 100644
index 000000000000..157478da6ac9
--- /dev/null
+++ b/fs/squashfs/decompressor.c
@@ -0,0 +1,68 @@
+/*
+ * Squashfs - a compressed read only filesystem for Linux
+ *
+ * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009
+ * Phillip Lougher <phillip@lougher.demon.co.uk>
+ *
+ * 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, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
+ * decompressor.c
+ */
+
+#include <linux/types.h>
+#include <linux/mutex.h>
+#include <linux/buffer_head.h>
+
+#include "squashfs_fs.h"
+#include "squashfs_fs_sb.h"
+#include "squashfs_fs_i.h"
+#include "decompressor.h"
+#include "squashfs.h"
+
+/*
+ * This file (and decompressor.h) implements a decompressor framework for
+ * Squashfs, allowing multiple decompressors to be easily supported
+ */
+
+static const struct squashfs_decompressor squashfs_lzma_unsupported_comp_ops = {
+	NULL, NULL, NULL, LZMA_COMPRESSION, "lzma", 0
+};
+
+static const struct squashfs_decompressor squashfs_lzo_unsupported_comp_ops = {
+	NULL, NULL, NULL, LZO_COMPRESSION, "lzo", 0
+};
+
+static const struct squashfs_decompressor squashfs_unknown_comp_ops = {
+	NULL, NULL, NULL, 0, "unknown", 0
+};
+
+static const struct squashfs_decompressor *decompressor[] = {
+	&squashfs_zlib_comp_ops,
+	&squashfs_lzma_unsupported_comp_ops,
+	&squashfs_lzo_unsupported_comp_ops,
+	&squashfs_unknown_comp_ops
+};
+
+
+const struct squashfs_decompressor *squashfs_lookup_decompressor(int id)
+{
+	int i;
+
+	for (i = 0; decompressor[i]->id; i++)
+		if (id == decompressor[i]->id)
+			break;
+
+	return decompressor[i];
+}
diff --git a/fs/squashfs/decompressor.h b/fs/squashfs/decompressor.h
new file mode 100644
index 000000000000..7425f80783f6
--- /dev/null
+++ b/fs/squashfs/decompressor.h
@@ -0,0 +1,55 @@
+#ifndef DECOMPRESSOR_H
+#define DECOMPRESSOR_H
+/*
+ * Squashfs - a compressed read only filesystem for Linux
+ *
+ * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009
+ * Phillip Lougher <phillip@lougher.demon.co.uk>
+ *
+ * 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, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
+ * decompressor.h
+ */
+
+struct squashfs_decompressor {
+	void	*(*init)(struct squashfs_sb_info *);
+	void	(*free)(void *);
+	int	(*decompress)(struct squashfs_sb_info *, void **,
+		struct buffer_head **, int, int, int, int, int);
+	int	id;
+	char	*name;
+	int	supported;
+};
+
+static inline void *squashfs_decompressor_init(struct squashfs_sb_info *msblk)
+{
+	return msblk->decompressor->init(msblk);
+}
+
+static inline void squashfs_decompressor_free(struct squashfs_sb_info *msblk,
+	void *s)
+{
+	if (msblk->decompressor)
+		msblk->decompressor->free(s);
+}
+
+static inline int squashfs_decompress(struct squashfs_sb_info *msblk,
+	void **buffer, struct buffer_head **bh, int b, int offset, int length,
+	int srclength, int pages)
+{
+	return msblk->decompressor->decompress(msblk, buffer, bh, b, offset,
+		length, srclength, pages);
+}
+#endif
diff --git a/fs/squashfs/dir.c b/fs/squashfs/dir.c
index 566b0eaed868..12b933ac6585 100644
--- a/fs/squashfs/dir.c
+++ b/fs/squashfs/dir.c
@@ -30,7 +30,6 @@
 #include <linux/fs.h>
 #include <linux/vfs.h>
 #include <linux/slab.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
diff --git a/fs/squashfs/export.c b/fs/squashfs/export.c
index 2b1b8fe5e037..7f93d5a9ee05 100644
--- a/fs/squashfs/export.c
+++ b/fs/squashfs/export.c
@@ -39,7 +39,6 @@
 #include <linux/vfs.h>
 #include <linux/dcache.h>
 #include <linux/exportfs.h>
-#include <linux/zlib.h>
 #include <linux/slab.h>
 
 #include "squashfs_fs.h"
diff --git a/fs/squashfs/file.c b/fs/squashfs/file.c
index 717767d831df..a25c5060bdcb 100644
--- a/fs/squashfs/file.c
+++ b/fs/squashfs/file.c
@@ -47,7 +47,6 @@
 #include <linux/string.h>
 #include <linux/pagemap.h>
 #include <linux/mutex.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
diff --git a/fs/squashfs/fragment.c b/fs/squashfs/fragment.c
index b5a2c15bbbc7..7c90bbd6879d 100644
--- a/fs/squashfs/fragment.c
+++ b/fs/squashfs/fragment.c
@@ -36,7 +36,6 @@
 #include <linux/fs.h>
 #include <linux/vfs.h>
 #include <linux/slab.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
diff --git a/fs/squashfs/id.c b/fs/squashfs/id.c
index 3795b837ba28..b7f64bcd2b70 100644
--- a/fs/squashfs/id.c
+++ b/fs/squashfs/id.c
@@ -34,7 +34,6 @@
 #include <linux/fs.h>
 #include <linux/vfs.h>
 #include <linux/slab.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
diff --git a/fs/squashfs/inode.c b/fs/squashfs/inode.c
index 9101dbde39ec..49daaf669e41 100644
--- a/fs/squashfs/inode.c
+++ b/fs/squashfs/inode.c
@@ -40,7 +40,6 @@
 
 #include <linux/fs.h>
 #include <linux/vfs.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
diff --git a/fs/squashfs/namei.c b/fs/squashfs/namei.c
index 9e398653b22b..5266bd8ad932 100644
--- a/fs/squashfs/namei.c
+++ b/fs/squashfs/namei.c
@@ -57,7 +57,6 @@
 #include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/dcache.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
diff --git a/fs/squashfs/squashfs.h b/fs/squashfs/squashfs.h
index 0e9feb6adf7e..fe2587af5512 100644
--- a/fs/squashfs/squashfs.h
+++ b/fs/squashfs/squashfs.h
@@ -51,6 +51,9 @@ extern struct squashfs_cache_entry *squashfs_get_datablock(struct super_block *,
 				u64, int);
 extern int squashfs_read_table(struct super_block *, void *, u64, int);
 
+/* decompressor.c */
+extern const struct squashfs_decompressor *squashfs_lookup_decompressor(int);
+
 /* export.c */
 extern __le64 *squashfs_read_inode_lookup_table(struct super_block *, u64,
 				unsigned int);
@@ -71,7 +74,7 @@ extern struct inode *squashfs_iget(struct super_block *, long long,
 extern int squashfs_read_inode(struct inode *, long long);
 
 /*
- * Inodes and files operations
+ * Inodes, files and decompressor operations
  */
 
 /* dir.c */
@@ -88,3 +91,6 @@ extern const struct inode_operations squashfs_dir_inode_ops;
 
 /* symlink.c */
 extern const struct address_space_operations squashfs_symlink_aops;
+
+/* zlib_wrapper.c */
+extern const struct squashfs_decompressor squashfs_zlib_comp_ops;
diff --git a/fs/squashfs/squashfs_fs.h b/fs/squashfs/squashfs_fs.h
index 283daafc568e..79024245ea00 100644
--- a/fs/squashfs/squashfs_fs.h
+++ b/fs/squashfs/squashfs_fs.h
@@ -183,8 +183,6 @@
 #define SQUASHFS_MAX_FILE_SIZE		(1LL << \
 					(SQUASHFS_MAX_FILE_SIZE_LOG - 2))
 
-#define SQUASHFS_MARKER_BYTE		0xff
-
 /* meta index cache */
 #define SQUASHFS_META_INDEXES	(SQUASHFS_METADATA_SIZE / sizeof(unsigned int))
 #define SQUASHFS_META_ENTRIES	127
@@ -211,7 +209,9 @@ struct meta_index {
 /*
  * definitions for structures on disk
  */
-#define ZLIB_COMPRESSION	 1
+#define ZLIB_COMPRESSION	1
+#define LZMA_COMPRESSION	2
+#define LZO_COMPRESSION		3
 
 struct squashfs_super_block {
 	__le32			s_magic;
diff --git a/fs/squashfs/squashfs_fs_sb.h b/fs/squashfs/squashfs_fs_sb.h
index c8c65614dd1c..2e77dc547e25 100644
--- a/fs/squashfs/squashfs_fs_sb.h
+++ b/fs/squashfs/squashfs_fs_sb.h
@@ -52,25 +52,25 @@ struct squashfs_cache_entry {
 };
 
 struct squashfs_sb_info {
-	int			devblksize;
-	int			devblksize_log2;
-	struct squashfs_cache	*block_cache;
-	struct squashfs_cache	*fragment_cache;
-	struct squashfs_cache	*read_page;
-	int			next_meta_index;
-	__le64			*id_table;
-	__le64			*fragment_index;
-	unsigned int		*fragment_index_2;
-	struct mutex		read_data_mutex;
-	struct mutex		meta_index_mutex;
-	struct meta_index	*meta_index;
-	z_stream		stream;
-	__le64			*inode_lookup_table;
-	u64			inode_table;
-	u64			directory_table;
-	unsigned int		block_size;
-	unsigned short		block_log;
-	long long		bytes_used;
-	unsigned int		inodes;
+	const struct squashfs_decompressor	*decompressor;
+	int					devblksize;
+	int					devblksize_log2;
+	struct squashfs_cache			*block_cache;
+	struct squashfs_cache			*fragment_cache;
+	struct squashfs_cache			*read_page;
+	int					next_meta_index;
+	__le64					*id_table;
+	__le64					*fragment_index;
+	struct mutex				read_data_mutex;
+	struct mutex				meta_index_mutex;
+	struct meta_index			*meta_index;
+	void					*stream;
+	__le64					*inode_lookup_table;
+	u64					inode_table;
+	u64					directory_table;
+	unsigned int				block_size;
+	unsigned short				block_log;
+	long long				bytes_used;
+	unsigned int				inodes;
 };
 #endif
diff --git a/fs/squashfs/super.c b/fs/squashfs/super.c
index 6c197ef53add..3550aec2f655 100644
--- a/fs/squashfs/super.c
+++ b/fs/squashfs/super.c
@@ -35,34 +35,41 @@
 #include <linux/pagemap.h>
 #include <linux/init.h>
 #include <linux/module.h>
-#include <linux/zlib.h>
 #include <linux/magic.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
 #include "squashfs_fs_i.h"
 #include "squashfs.h"
+#include "decompressor.h"
 
 static struct file_system_type squashfs_fs_type;
 static const struct super_operations squashfs_super_ops;
 
-static int supported_squashfs_filesystem(short major, short minor, short comp)
+static const struct squashfs_decompressor *supported_squashfs_filesystem(short
+	major, short minor, short id)
 {
+	const struct squashfs_decompressor *decompressor;
+
 	if (major < SQUASHFS_MAJOR) {
 		ERROR("Major/Minor mismatch, older Squashfs %d.%d "
 			"filesystems are unsupported\n", major, minor);
-		return -EINVAL;
+		return NULL;
 	} else if (major > SQUASHFS_MAJOR || minor > SQUASHFS_MINOR) {
 		ERROR("Major/Minor mismatch, trying to mount newer "
 			"%d.%d filesystem\n", major, minor);
 		ERROR("Please update your kernel\n");
-		return -EINVAL;
+		return NULL;
 	}
 
-	if (comp != ZLIB_COMPRESSION)
-		return -EINVAL;
+	decompressor = squashfs_lookup_decompressor(id);
+	if (!decompressor->supported) {
+		ERROR("Filesystem uses \"%s\" compression. This is not "
+			"supported\n", decompressor->name);
+		return NULL;
+	}
 
-	return 0;
+	return decompressor;
 }
 
 
@@ -87,13 +94,6 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent)
 	}
 	msblk = sb->s_fs_info;
 
-	msblk->stream.workspace = kmalloc(zlib_inflate_workspacesize(),
-		GFP_KERNEL);
-	if (msblk->stream.workspace == NULL) {
-		ERROR("Failed to allocate zlib workspace\n");
-		goto failure;
-	}
-
 	sblk = kzalloc(sizeof(*sblk), GFP_KERNEL);
 	if (sblk == NULL) {
 		ERROR("Failed to allocate squashfs_super_block\n");
@@ -120,25 +120,25 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent)
 		goto failed_mount;
 	}
 
+	err = -EINVAL;
+
 	/* Check it is a SQUASHFS superblock */
 	sb->s_magic = le32_to_cpu(sblk->s_magic);
 	if (sb->s_magic != SQUASHFS_MAGIC) {
 		if (!silent)
 			ERROR("Can't find a SQUASHFS superblock on %s\n",
 						bdevname(sb->s_bdev, b));
-		err = -EINVAL;
 		goto failed_mount;
 	}
 
-	/* Check the MAJOR & MINOR versions and compression type */
-	err = supported_squashfs_filesystem(le16_to_cpu(sblk->s_major),
+	/* Check the MAJOR & MINOR versions and lookup compression type */
+	msblk->decompressor = supported_squashfs_filesystem(
+			le16_to_cpu(sblk->s_major),
 			le16_to_cpu(sblk->s_minor),
 			le16_to_cpu(sblk->compression));
-	if (err < 0)
+	if (msblk->decompressor == NULL)
 		goto failed_mount;
 
-	err = -EINVAL;
-
 	/*
 	 * Check if there's xattrs in the filesystem.  These are not
 	 * supported in this version, so warn that they will be ignored.
@@ -205,6 +205,10 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent)
 
 	err = -ENOMEM;
 
+	msblk->stream = squashfs_decompressor_init(msblk);
+	if (msblk->stream == NULL)
+		goto failed_mount;
+
 	msblk->block_cache = squashfs_cache_init("metadata",
 			SQUASHFS_CACHED_BLKS, SQUASHFS_METADATA_SIZE);
 	if (msblk->block_cache == NULL)
@@ -292,17 +296,16 @@ failed_mount:
 	squashfs_cache_delete(msblk->block_cache);
 	squashfs_cache_delete(msblk->fragment_cache);
 	squashfs_cache_delete(msblk->read_page);
+	squashfs_decompressor_free(msblk, msblk->stream);
 	kfree(msblk->inode_lookup_table);
 	kfree(msblk->fragment_index);
 	kfree(msblk->id_table);
-	kfree(msblk->stream.workspace);
 	kfree(sb->s_fs_info);
 	sb->s_fs_info = NULL;
 	kfree(sblk);
 	return err;
 
 failure:
-	kfree(msblk->stream.workspace);
 	kfree(sb->s_fs_info);
 	sb->s_fs_info = NULL;
 	return -ENOMEM;
@@ -346,10 +349,10 @@ static void squashfs_put_super(struct super_block *sb)
 		squashfs_cache_delete(sbi->block_cache);
 		squashfs_cache_delete(sbi->fragment_cache);
 		squashfs_cache_delete(sbi->read_page);
+		squashfs_decompressor_free(sbi, sbi->stream);
 		kfree(sbi->id_table);
 		kfree(sbi->fragment_index);
 		kfree(sbi->meta_index);
-		kfree(sbi->stream.workspace);
 		kfree(sb->s_fs_info);
 		sb->s_fs_info = NULL;
 	}
diff --git a/fs/squashfs/symlink.c b/fs/squashfs/symlink.c
index 83d87880aac8..e80be2022a7f 100644
--- a/fs/squashfs/symlink.c
+++ b/fs/squashfs/symlink.c
@@ -36,7 +36,6 @@
 #include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/pagemap.h>
-#include <linux/zlib.h>
 
 #include "squashfs_fs.h"
 #include "squashfs_fs_sb.h"
diff --git a/fs/squashfs/zlib_wrapper.c b/fs/squashfs/zlib_wrapper.c
new file mode 100644
index 000000000000..4dd70e04333b
--- /dev/null
+++ b/fs/squashfs/zlib_wrapper.c
@@ -0,0 +1,150 @@
+/*
+ * Squashfs - a compressed read only filesystem for Linux
+ *
+ * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009
+ * Phillip Lougher <phillip@lougher.demon.co.uk>
+ *
+ * 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, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
+ * zlib_wrapper.c
+ */
+
+
+#include <linux/mutex.h>
+#include <linux/buffer_head.h>
+#include <linux/zlib.h>
+
+#include "squashfs_fs.h"
+#include "squashfs_fs_sb.h"
+#include "squashfs_fs_i.h"
+#include "squashfs.h"
+#include "decompressor.h"
+
+static void *zlib_init(struct squashfs_sb_info *dummy)
+{
+	z_stream *stream = kmalloc(sizeof(z_stream), GFP_KERNEL);
+	if (stream == NULL)
+		goto failed;
+	stream->workspace = kmalloc(zlib_inflate_workspacesize(),
+		GFP_KERNEL);
+	if (stream->workspace == NULL)
+		goto failed;
+
+	return stream;
+
+failed:
+	ERROR("Failed to allocate zlib workspace\n");
+	kfree(stream);
+	return NULL;
+}
+
+
+static void zlib_free(void *strm)
+{
+	z_stream *stream = strm;
+
+	if (stream)
+		kfree(stream->workspace);
+	kfree(stream);
+}
+
+
+static int zlib_uncompress(struct squashfs_sb_info *msblk, void **buffer,
+	struct buffer_head **bh, int b, int offset, int length, int srclength,
+	int pages)
+{
+	int zlib_err = 0, zlib_init = 0;
+	int avail, bytes, k = 0, page = 0;
+	z_stream *stream = msblk->stream;
+
+	mutex_lock(&msblk->read_data_mutex);
+
+	stream->avail_out = 0;
+	stream->avail_in = 0;
+
+	bytes = length;
+	do {
+		if (stream->avail_in == 0 && k < b) {
+			avail = min(bytes, msblk->devblksize - offset);
+			bytes -= avail;
+			wait_on_buffer(bh[k]);
+			if (!buffer_uptodate(bh[k]))
+				goto release_mutex;
+
+			if (avail == 0) {
+				offset = 0;
+				put_bh(bh[k++]);
+				continue;
+			}
+
+			stream->next_in = bh[k]->b_data + offset;
+			stream->avail_in = avail;
+			offset = 0;
+		}
+
+		if (stream->avail_out == 0 && page < pages) {
+			stream->next_out = buffer[page++];
+			stream->avail_out = PAGE_CACHE_SIZE;
+		}
+
+		if (!zlib_init) {
+			zlib_err = zlib_inflateInit(stream);
+			if (zlib_err != Z_OK) {
+				ERROR("zlib_inflateInit returned unexpected "
+					"result 0x%x, srclength %d\n",
+					zlib_err, srclength);
+				goto release_mutex;
+			}
+			zlib_init = 1;
+		}
+
+		zlib_err = zlib_inflate(stream, Z_SYNC_FLUSH);
+
+		if (stream->avail_in == 0 && k < b)
+			put_bh(bh[k++]);
+	} while (zlib_err == Z_OK);
+
+	if (zlib_err != Z_STREAM_END) {
+		ERROR("zlib_inflate error, data probably corrupt\n");
+		goto release_mutex;
+	}
+
+	zlib_err = zlib_inflateEnd(stream);
+	if (zlib_err != Z_OK) {
+		ERROR("zlib_inflate error, data probably corrupt\n");
+		goto release_mutex;
+	}
+
+	mutex_unlock(&msblk->read_data_mutex);
+	return stream->total_out;
+
+release_mutex:
+	mutex_unlock(&msblk->read_data_mutex);
+
+	for (; k < b; k++)
+		put_bh(bh[k]);
+
+	return -EIO;
+}
+
+const struct squashfs_decompressor squashfs_zlib_comp_ops = {
+	.init = zlib_init,
+	.free = zlib_free,
+	.decompress = zlib_uncompress,
+	.id = ZLIB_COMPRESSION,
+	.name = "zlib",
+	.supported = 1
+};
+
diff --git a/fs/super.c b/fs/super.c
index aff046b0fe78..f35ac6022109 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -568,7 +568,7 @@ out:
 int do_remount_sb(struct super_block *sb, int flags, void *data, int force)
 {
 	int retval;
-	int remount_rw;
+	int remount_rw, remount_ro;
 
 	if (sb->s_frozen != SB_UNFROZEN)
 		return -EBUSY;
@@ -583,9 +583,12 @@ int do_remount_sb(struct super_block *sb, int flags, void *data, int force)
 	shrink_dcache_sb(sb);
 	sync_filesystem(sb);
 
+	remount_ro = (flags & MS_RDONLY) && !(sb->s_flags & MS_RDONLY);
+	remount_rw = !(flags & MS_RDONLY) && (sb->s_flags & MS_RDONLY);
+
 	/* If we are remounting RDONLY and current sb is read/write,
 	   make sure there are no rw files opened */
-	if ((flags & MS_RDONLY) && !(sb->s_flags & MS_RDONLY)) {
+	if (remount_ro) {
 		if (force)
 			mark_files_ro(sb);
 		else if (!fs_may_remount_ro(sb))
@@ -594,7 +597,6 @@ int do_remount_sb(struct super_block *sb, int flags, void *data, int force)
 		if (retval < 0 && retval != -ENOSYS)
 			return -EBUSY;
 	}
-	remount_rw = !(flags & MS_RDONLY) && (sb->s_flags & MS_RDONLY);
 
 	if (sb->s_op->remount_fs) {
 		retval = sb->s_op->remount_fs(sb, &flags, data);
@@ -604,6 +606,16 @@ int do_remount_sb(struct super_block *sb, int flags, void *data, int force)
 	sb->s_flags = (sb->s_flags & ~MS_RMT_MASK) | (flags & MS_RMT_MASK);
 	if (remount_rw)
 		vfs_dq_quota_on_remount(sb);
+	/*
+	 * Some filesystems modify their metadata via some other path than the
+	 * bdev buffer cache (eg. use a private mapping, or directories in
+	 * pagecache, etc). Also file data modifications go via their own
+	 * mappings. So If we try to mount readonly then copy the filesystem
+	 * from bdev, we could get stale data, so invalidate it to give a best
+	 * effort at coherency.
+	 */
+	if (remount_ro && sb->s_bdev)
+		invalidate_bdev(sb->s_bdev);
 	return 0;
 }
 
@@ -925,6 +937,9 @@ vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void
 	if (!mnt)
 		goto out;
 
+	if (flags & MS_KERNMOUNT)
+		mnt->mnt_flags = MNT_INTERNAL;
+
 	if (data && !(type->fs_flags & FS_BINARY_MOUNTDATA)) {
 		secdata = alloc_secdata();
 		if (!secdata)
diff --git a/fs/sync.c b/fs/sync.c
index 418727a2a239..f557d71cb097 100644
--- a/fs/sync.c
+++ b/fs/sync.c
@@ -34,14 +34,14 @@ static int __sync_filesystem(struct super_block *sb, int wait)
 	if (!sb->s_bdi)
 		return 0;
 
-	/* Avoid doing twice syncing and cache pruning for quota sync */
-	if (!wait) {
-		writeout_quota_sb(sb, -1);
-		writeback_inodes_sb(sb);
-	} else {
-		sync_quota_sb(sb, -1);
+	if (sb->s_qcop && sb->s_qcop->quota_sync)
+		sb->s_qcop->quota_sync(sb, -1, wait);
+
+	if (wait)
 		sync_inodes_sb(sb);
-	}
+	else
+		writeback_inodes_sb(sb);
+
 	if (sb->s_op->sync_fs)
 		sb->s_op->sync_fs(sb, wait);
 	return __sync_blockdev(sb->s_bdev, wait);
diff --git a/fs/sysv/inode.c b/fs/sysv/inode.c
index 9824743832a7..4573734d723d 100644
--- a/fs/sysv/inode.c
+++ b/fs/sysv/inode.c
@@ -26,6 +26,7 @@
 #include <linux/init.h>
 #include <linux/buffer_head.h>
 #include <linux/vfs.h>
+#include <linux/writeback.h>
 #include <linux/namei.h>
 #include <asm/byteorder.h>
 #include "sysv.h"
@@ -246,7 +247,7 @@ bad_inode:
 	return ERR_PTR(-EIO);
 }
 
-int sysv_write_inode(struct inode *inode, int wait)
+static int __sysv_write_inode(struct inode *inode, int wait)
 {
 	struct super_block * sb = inode->i_sb;
 	struct sysv_sb_info * sbi = SYSV_SB(sb);
@@ -296,9 +297,14 @@ int sysv_write_inode(struct inode *inode, int wait)
 	return 0;
 }
 
+int sysv_write_inode(struct inode *inode, struct writeback_control *wbc)
+{
+	return __sysv_write_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
+}
+
 int sysv_sync_inode(struct inode *inode)
 {
-	return sysv_write_inode(inode, 1);
+	return __sysv_write_inode(inode, 1);
 }
 
 static void sysv_delete_inode(struct inode *inode)
diff --git a/fs/sysv/sysv.h b/fs/sysv/sysv.h
index 53786eb5cf60..94cb9b4d76c2 100644
--- a/fs/sysv/sysv.h
+++ b/fs/sysv/sysv.h
@@ -142,7 +142,7 @@ extern int __sysv_write_begin(struct file *file, struct address_space *mapping,
 
 /* inode.c */
 extern struct inode *sysv_iget(struct super_block *, unsigned int);
-extern int sysv_write_inode(struct inode *, int);
+extern int sysv_write_inode(struct inode *, struct writeback_control *wbc);
 extern int sysv_sync_inode(struct inode *);
 extern void sysv_set_inode(struct inode *, dev_t);
 extern int sysv_getattr(struct vfsmount *, struct dentry *, struct kstat *);
diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c
index 552fb0111fff..401e503d44a1 100644
--- a/fs/ubifs/dir.c
+++ b/fs/ubifs/dir.c
@@ -1120,7 +1120,7 @@ static int ubifs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	if (release)
 		ubifs_release_budget(c, &ino_req);
 	if (IS_SYNC(old_inode))
-		err = old_inode->i_sb->s_op->write_inode(old_inode, 1);
+		err = old_inode->i_sb->s_op->write_inode(old_inode, NULL);
 	return err;
 
 out_cancel:
diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 16a6444330ec..e26c02ab6cd5 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -1011,7 +1011,7 @@ static int ubifs_writepage(struct page *page, struct writeback_control *wbc)
 	/* Is the page fully inside @i_size? */
 	if (page->index < end_index) {
 		if (page->index >= synced_i_size >> PAGE_CACHE_SHIFT) {
-			err = inode->i_sb->s_op->write_inode(inode, 1);
+			err = inode->i_sb->s_op->write_inode(inode, NULL);
 			if (err)
 				goto out_unlock;
 			/*
@@ -1039,7 +1039,7 @@ static int ubifs_writepage(struct page *page, struct writeback_control *wbc)
 	kunmap_atomic(kaddr, KM_USER0);
 
 	if (i_size > synced_i_size) {
-		err = inode->i_sb->s_op->write_inode(inode, 1);
+		err = inode->i_sb->s_op->write_inode(inode, NULL);
 		if (err)
 			goto out_unlock;
 	}
@@ -1242,7 +1242,7 @@ static int do_setattr(struct ubifs_info *c, struct inode *inode,
 	if (release)
 		ubifs_release_budget(c, &req);
 	if (IS_SYNC(inode))
-		err = inode->i_sb->s_op->write_inode(inode, 1);
+		err = inode->i_sb->s_op->write_inode(inode, NULL);
 	return err;
 
 out:
@@ -1316,7 +1316,7 @@ int ubifs_fsync(struct file *file, struct dentry *dentry, int datasync)
 	 * the inode unless this is a 'datasync()' call.
 	 */
 	if (!datasync || (inode->i_state & I_DIRTY_DATASYNC)) {
-		err = inode->i_sb->s_op->write_inode(inode, 1);
+		err = inode->i_sb->s_op->write_inode(inode, NULL);
 		if (err)
 			return err;
 	}
diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
index 43f9d19a6f33..4d2f2157dd3f 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -283,7 +283,7 @@ static void ubifs_destroy_inode(struct inode *inode)
 /*
  * Note, Linux write-back code calls this without 'i_mutex'.
  */
-static int ubifs_write_inode(struct inode *inode, int wait)
+static int ubifs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	int err = 0;
 	struct ubifs_info *c = inode->i_sb->s_fs_info;
diff --git a/fs/udf/balloc.c b/fs/udf/balloc.c
index 82372e332f08..ccc3ad7242d4 100644
--- a/fs/udf/balloc.c
+++ b/fs/udf/balloc.c
@@ -208,7 +208,7 @@ static void udf_bitmap_free_blocks(struct super_block *sb,
 					((char *)bh->b_data)[(bit + i) >> 3]);
 			} else {
 				if (inode)
-					vfs_dq_free_block(inode, 1);
+					dquot_free_block(inode, 1);
 				udf_add_free_space(sb, sbi->s_partition, 1);
 			}
 		}
@@ -260,11 +260,11 @@ static int udf_bitmap_prealloc_blocks(struct super_block *sb,
 		while (bit < (sb->s_blocksize << 3) && block_count > 0) {
 			if (!udf_test_bit(bit, bh->b_data))
 				goto out;
-			else if (vfs_dq_prealloc_block(inode, 1))
+			else if (dquot_prealloc_block(inode, 1))
 				goto out;
 			else if (!udf_clear_bit(bit, bh->b_data)) {
 				udf_debug("bit already cleared for block %d\n", bit);
-				vfs_dq_free_block(inode, 1);
+				dquot_free_block(inode, 1);
 				goto out;
 			}
 			block_count--;
@@ -390,10 +390,14 @@ got_block:
 	/*
 	 * Check quota for allocation of this block.
 	 */
-	if (inode && vfs_dq_alloc_block(inode, 1)) {
-		mutex_unlock(&sbi->s_alloc_mutex);
-		*err = -EDQUOT;
-		return 0;
+	if (inode) {
+		int ret = dquot_alloc_block(inode, 1);
+
+		if (ret) {
+			mutex_unlock(&sbi->s_alloc_mutex);
+			*err = ret;
+			return 0;
+		}
 	}
 
 	newblock = bit + (block_group << (sb->s_blocksize_bits + 3)) -
@@ -449,7 +453,7 @@ static void udf_table_free_blocks(struct super_block *sb,
 	/* We do this up front - There are some error conditions that
 	   could occure, but.. oh well */
 	if (inode)
-		vfs_dq_free_block(inode, count);
+		dquot_free_block(inode, count);
 	udf_add_free_space(sb, sbi->s_partition, count);
 
 	start = bloc->logicalBlockNum + offset;
@@ -547,7 +551,7 @@ static void udf_table_free_blocks(struct super_block *sb,
 		}
 
 		if (epos.offset + (2 * adsize) > sb->s_blocksize) {
-			char *sptr, *dptr;
+			unsigned char *sptr, *dptr;
 			int loffset;
 
 			brelse(oepos.bh);
@@ -694,7 +698,7 @@ static int udf_table_prealloc_blocks(struct super_block *sb,
 		epos.offset -= adsize;
 
 		alloc_count = (elen >> sb->s_blocksize_bits);
-		if (inode && vfs_dq_prealloc_block(inode,
+		if (inode && dquot_prealloc_block(inode,
 			alloc_count > block_count ? block_count : alloc_count))
 			alloc_count = 0;
 		else if (alloc_count > block_count) {
@@ -797,12 +801,13 @@ static int udf_table_new_block(struct super_block *sb,
 	newblock = goal_eloc.logicalBlockNum;
 	goal_eloc.logicalBlockNum++;
 	goal_elen -= sb->s_blocksize;
-
-	if (inode && vfs_dq_alloc_block(inode, 1)) {
-		brelse(goal_epos.bh);
-		mutex_unlock(&sbi->s_alloc_mutex);
-		*err = -EDQUOT;
-		return 0;
+	if (inode) {
+		*err = dquot_alloc_block(inode, 1);
+		if (*err) {
+			brelse(goal_epos.bh);
+			mutex_unlock(&sbi->s_alloc_mutex);
+			return 0;
+		}
 	}
 
 	if (goal_elen)
diff --git a/fs/udf/dir.c b/fs/udf/dir.c
index 61d9a76a3a69..f0f2a436251e 100644
--- a/fs/udf/dir.c
+++ b/fs/udf/dir.c
@@ -45,8 +45,8 @@ static int do_udf_readdir(struct inode *dir, struct file *filp,
 	int block, iblock;
 	loff_t nf_pos = (filp->f_pos - 1) << 2;
 	int flen;
-	char *fname = NULL;
-	char *nameptr;
+	unsigned char *fname = NULL;
+	unsigned char *nameptr;
 	uint16_t liu;
 	uint8_t lfi;
 	loff_t size = udf_ext0_offset(dir) + dir->i_size;
diff --git a/fs/udf/file.c b/fs/udf/file.c
index f311d509b6a3..1eb06774ed90 100644
--- a/fs/udf/file.c
+++ b/fs/udf/file.c
@@ -34,6 +34,7 @@
 #include <linux/errno.h>
 #include <linux/smp_lock.h>
 #include <linux/pagemap.h>
+#include <linux/quotaops.h>
 #include <linux/buffer_head.h>
 #include <linux/aio.h>
 
@@ -207,7 +208,7 @@ const struct file_operations udf_file_operations = {
 	.read			= do_sync_read,
 	.aio_read		= generic_file_aio_read,
 	.ioctl			= udf_ioctl,
-	.open			= generic_file_open,
+	.open			= dquot_file_open,
 	.mmap			= generic_file_mmap,
 	.write			= do_sync_write,
 	.aio_write		= udf_file_aio_write,
@@ -217,6 +218,29 @@ const struct file_operations udf_file_operations = {
 	.llseek			= generic_file_llseek,
 };
 
+static int udf_setattr(struct dentry *dentry, struct iattr *iattr)
+{
+	struct inode *inode = dentry->d_inode;
+	int error;
+
+	error = inode_change_ok(inode, iattr);
+	if (error)
+		return error;
+
+	if (iattr->ia_valid & ATTR_SIZE)
+		dquot_initialize(inode);
+
+	if ((iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) ||
+            (iattr->ia_valid & ATTR_GID && iattr->ia_gid != inode->i_gid)) {
+		error = dquot_transfer(inode, iattr);
+		if (error)
+			return error;
+	}
+
+	return inode_setattr(inode, iattr);
+}
+
 const struct inode_operations udf_file_inode_operations = {
-	.truncate = udf_truncate,
+	.truncate		= udf_truncate,
+	.setattr		= udf_setattr,
 };
diff --git a/fs/udf/ialloc.c b/fs/udf/ialloc.c
index c10fa39f97e2..fb68c9cd0c3e 100644
--- a/fs/udf/ialloc.c
+++ b/fs/udf/ialloc.c
@@ -36,8 +36,8 @@ void udf_free_inode(struct inode *inode)
 	 * Note: we must free any quota before locking the superblock,
 	 * as writing the quota to disk may need the lock as well.
 	 */
-	vfs_dq_free_inode(inode);
-	vfs_dq_drop(inode);
+	dquot_free_inode(inode);
+	dquot_drop(inode);
 
 	clear_inode(inode);
 
@@ -61,7 +61,7 @@ struct inode *udf_new_inode(struct inode *dir, int mode, int *err)
 	struct super_block *sb = dir->i_sb;
 	struct udf_sb_info *sbi = UDF_SB(sb);
 	struct inode *inode;
-	int block;
+	int block, ret;
 	uint32_t start = UDF_I(dir)->i_location.logicalBlockNum;
 	struct udf_inode_info *iinfo;
 	struct udf_inode_info *dinfo = UDF_I(dir);
@@ -153,12 +153,14 @@ struct inode *udf_new_inode(struct inode *dir, int mode, int *err)
 	insert_inode_hash(inode);
 	mark_inode_dirty(inode);
 
-	if (vfs_dq_alloc_inode(inode)) {
-		vfs_dq_drop(inode);
+	dquot_initialize(inode);
+	ret = dquot_alloc_inode(inode);
+	if (ret) {
+		dquot_drop(inode);
 		inode->i_flags |= S_NOQUOTA;
 		inode->i_nlink = 0;
 		iput(inode);
-		*err = -EDQUOT;
+		*err = ret;
 		return NULL;
 	}
 
diff --git a/fs/udf/inode.c b/fs/udf/inode.c
index f90231eb2916..b57ab0402d89 100644
--- a/fs/udf/inode.c
+++ b/fs/udf/inode.c
@@ -36,6 +36,7 @@
 #include <linux/pagemap.h>
 #include <linux/buffer_head.h>
 #include <linux/writeback.h>
+#include <linux/quotaops.h>
 #include <linux/slab.h>
 #include <linux/crc-itu-t.h>
 
@@ -70,6 +71,9 @@ static int udf_get_block(struct inode *, sector_t, struct buffer_head *, int);
 
 void udf_delete_inode(struct inode *inode)
 {
+	if (!is_bad_inode(inode))
+		dquot_initialize(inode);
+
 	truncate_inode_pages(&inode->i_data, 0);
 
 	if (is_bad_inode(inode))
@@ -108,6 +112,8 @@ void udf_clear_inode(struct inode *inode)
 			(unsigned long long)inode->i_size,
 			(unsigned long long)iinfo->i_lenExtents);
 	}
+
+	dquot_drop(inode);
 	kfree(iinfo->i_ext.i_data);
 	iinfo->i_ext.i_data = NULL;
 }
@@ -1373,12 +1379,12 @@ static mode_t udf_convert_permissions(struct fileEntry *fe)
 	return mode;
 }
 
-int udf_write_inode(struct inode *inode, int sync)
+int udf_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	int ret;
 
 	lock_kernel();
-	ret = udf_update_inode(inode, sync);
+	ret = udf_update_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
 	unlock_kernel();
 
 	return ret;
@@ -1672,7 +1678,7 @@ int8_t udf_add_aext(struct inode *inode, struct extent_position *epos,
 		return -1;
 
 	if (epos->offset + (2 * adsize) > inode->i_sb->s_blocksize) {
-		char *sptr, *dptr;
+		unsigned char *sptr, *dptr;
 		struct buffer_head *nbh;
 		int err, loffset;
 		struct kernel_lb_addr obloc = epos->block;
diff --git a/fs/udf/namei.c b/fs/udf/namei.c
index cd2115060fdc..db423ab078b1 100644
--- a/fs/udf/namei.c
+++ b/fs/udf/namei.c
@@ -34,8 +34,8 @@
 #include <linux/crc-itu-t.h>
 #include <linux/exportfs.h>
 
-static inline int udf_match(int len1, const char *name1, int len2,
-			    const char *name2)
+static inline int udf_match(int len1, const unsigned char *name1, int len2,
+			    const unsigned char *name2)
 {
 	if (len1 != len2)
 		return 0;
@@ -142,15 +142,15 @@ int udf_write_fi(struct inode *inode, struct fileIdentDesc *cfi,
 }
 
 static struct fileIdentDesc *udf_find_entry(struct inode *dir,
-					    struct qstr *child,
+					    const struct qstr *child,
 					    struct udf_fileident_bh *fibh,
 					    struct fileIdentDesc *cfi)
 {
 	struct fileIdentDesc *fi = NULL;
 	loff_t f_pos;
 	int block, flen;
-	char *fname = NULL;
-	char *nameptr;
+	unsigned char *fname = NULL;
+	unsigned char *nameptr;
 	uint8_t lfi;
 	uint16_t liu;
 	loff_t size;
@@ -308,7 +308,7 @@ static struct fileIdentDesc *udf_add_entry(struct inode *dir,
 {
 	struct super_block *sb = dir->i_sb;
 	struct fileIdentDesc *fi = NULL;
-	char *name = NULL;
+	unsigned char *name = NULL;
 	int namelen;
 	loff_t f_pos;
 	loff_t size = udf_ext0_offset(dir) + dir->i_size;
@@ -563,6 +563,8 @@ static int udf_create(struct inode *dir, struct dentry *dentry, int mode,
 	int err;
 	struct udf_inode_info *iinfo;
 
+	dquot_initialize(dir);
+
 	lock_kernel();
 	inode = udf_new_inode(dir, mode, &err);
 	if (!inode) {
@@ -616,6 +618,8 @@ static int udf_mknod(struct inode *dir, struct dentry *dentry, int mode,
 	if (!old_valid_dev(rdev))
 		return -EINVAL;
 
+	dquot_initialize(dir);
+
 	lock_kernel();
 	err = -EIO;
 	inode = udf_new_inode(dir, mode, &err);
@@ -662,6 +666,8 @@ static int udf_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	struct udf_inode_info *dinfo = UDF_I(dir);
 	struct udf_inode_info *iinfo;
 
+	dquot_initialize(dir);
+
 	lock_kernel();
 	err = -EMLINK;
 	if (dir->i_nlink >= (256 << sizeof(dir->i_nlink)) - 1)
@@ -799,6 +805,8 @@ static int udf_rmdir(struct inode *dir, struct dentry *dentry)
 	struct fileIdentDesc *fi, cfi;
 	struct kernel_lb_addr tloc;
 
+	dquot_initialize(dir);
+
 	retval = -ENOENT;
 	lock_kernel();
 	fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi);
@@ -845,6 +853,8 @@ static int udf_unlink(struct inode *dir, struct dentry *dentry)
 	struct fileIdentDesc cfi;
 	struct kernel_lb_addr tloc;
 
+	dquot_initialize(dir);
+
 	retval = -ENOENT;
 	lock_kernel();
 	fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi);
@@ -885,20 +895,22 @@ static int udf_symlink(struct inode *dir, struct dentry *dentry,
 {
 	struct inode *inode;
 	struct pathComponent *pc;
-	char *compstart;
+	const char *compstart;
 	struct udf_fileident_bh fibh;
 	struct extent_position epos = {};
 	int eoffset, elen = 0;
 	struct fileIdentDesc *fi;
 	struct fileIdentDesc cfi;
-	char *ea;
+	uint8_t *ea;
 	int err;
 	int block;
-	char *name = NULL;
+	unsigned char *name = NULL;
 	int namelen;
 	struct buffer_head *bh;
 	struct udf_inode_info *iinfo;
 
+	dquot_initialize(dir);
+
 	lock_kernel();
 	inode = udf_new_inode(dir, S_IFLNK, &err);
 	if (!inode)
@@ -970,7 +982,7 @@ static int udf_symlink(struct inode *dir, struct dentry *dentry,
 
 		pc = (struct pathComponent *)(ea + elen);
 
-		compstart = (char *)symname;
+		compstart = symname;
 
 		do {
 			symname++;
@@ -1069,6 +1081,8 @@ static int udf_link(struct dentry *old_dentry, struct inode *dir,
 	int err;
 	struct buffer_head *bh;
 
+	dquot_initialize(dir);
+
 	lock_kernel();
 	if (inode->i_nlink >= (256 << sizeof(inode->i_nlink)) - 1) {
 		unlock_kernel();
@@ -1131,6 +1145,9 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry,
 	struct kernel_lb_addr tloc;
 	struct udf_inode_info *old_iinfo = UDF_I(old_inode);
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	lock_kernel();
 	ofi = udf_find_entry(old_dir, &old_dentry->d_name, &ofibh, &ocfi);
 	if (ofi) {
diff --git a/fs/udf/symlink.c b/fs/udf/symlink.c
index c3265e1385d4..852e91845688 100644
--- a/fs/udf/symlink.c
+++ b/fs/udf/symlink.c
@@ -32,12 +32,12 @@
 #include <linux/buffer_head.h>
 #include "udf_i.h"
 
-static void udf_pc_to_char(struct super_block *sb, char *from, int fromlen,
-			   char *to)
+static void udf_pc_to_char(struct super_block *sb, unsigned char *from,
+			   int fromlen, unsigned char *to)
 {
 	struct pathComponent *pc;
 	int elen = 0;
-	char *p = to;
+	unsigned char *p = to;
 
 	while (elen < fromlen) {
 		pc = (struct pathComponent *)(from + elen);
@@ -75,9 +75,9 @@ static int udf_symlink_filler(struct file *file, struct page *page)
 {
 	struct inode *inode = page->mapping->host;
 	struct buffer_head *bh = NULL;
-	char *symlink;
+	unsigned char *symlink;
 	int err = -EIO;
-	char *p = kmap(page);
+	unsigned char *p = kmap(page);
 	struct udf_inode_info *iinfo;
 
 	lock_kernel();
diff --git a/fs/udf/udfdecl.h b/fs/udf/udfdecl.h
index 8d46f4294ee7..4223ac855da9 100644
--- a/fs/udf/udfdecl.h
+++ b/fs/udf/udfdecl.h
@@ -142,7 +142,7 @@ extern void udf_truncate(struct inode *);
 extern void udf_read_inode(struct inode *);
 extern void udf_delete_inode(struct inode *);
 extern void udf_clear_inode(struct inode *);
-extern int udf_write_inode(struct inode *, int);
+extern int udf_write_inode(struct inode *, struct writeback_control *wbc);
 extern long udf_block_map(struct inode *, sector_t);
 extern int udf_extend_file(struct inode *, struct extent_position *,
 			   struct kernel_long_ad *, sector_t);
diff --git a/fs/ufs/balloc.c b/fs/ufs/balloc.c
index 54c16ec95dff..5cfa4d85ccf2 100644
--- a/fs/ufs/balloc.c
+++ b/fs/ufs/balloc.c
@@ -85,7 +85,7 @@ void ufs_free_fragments(struct inode *inode, u64 fragment, unsigned count)
 				   "bit already cleared for fragment %u", i);
 	}
 	
-	vfs_dq_free_block(inode, count);
+	dquot_free_block(inode, count);
 
 	
 	fs32_add(sb, &ucg->cg_cs.cs_nffree, count);
@@ -195,7 +195,7 @@ do_more:
 		ubh_setblock(UCPI_UBH(ucpi), ucpi->c_freeoff, blkno);
 		if ((UFS_SB(sb)->s_flags & UFS_CG_MASK) == UFS_CG_44BSD)
 			ufs_clusteracct (sb, ucpi, blkno, 1);
-		vfs_dq_free_block(inode, uspi->s_fpb);
+		dquot_free_block(inode, uspi->s_fpb);
 
 		fs32_add(sb, &ucg->cg_cs.cs_nbfree, 1);
 		uspi->cs_total.cs_nbfree++;
@@ -511,6 +511,7 @@ static u64 ufs_add_fragments(struct inode *inode, u64 fragment,
 	struct ufs_cg_private_info * ucpi;
 	struct ufs_cylinder_group * ucg;
 	unsigned cgno, fragno, fragoff, count, fragsize, i;
+	int ret;
 	
 	UFSD("ENTER, fragment %llu, oldcount %u, newcount %u\n",
 	     (unsigned long long)fragment, oldcount, newcount);
@@ -556,8 +557,9 @@ static u64 ufs_add_fragments(struct inode *inode, u64 fragment,
 		fs32_add(sb, &ucg->cg_frsum[fragsize - count], 1);
 	for (i = oldcount; i < newcount; i++)
 		ubh_clrbit (UCPI_UBH(ucpi), ucpi->c_freeoff, fragno + i);
-	if (vfs_dq_alloc_block(inode, count)) {
-		*err = -EDQUOT;
+	ret = dquot_alloc_block(inode, count);
+	if (ret) {
+		*err = ret;
 		return 0;
 	}
 
@@ -596,6 +598,7 @@ static u64 ufs_alloc_fragments(struct inode *inode, unsigned cgno,
 	struct ufs_cylinder_group * ucg;
 	unsigned oldcg, i, j, k, allocsize;
 	u64 result;
+	int ret;
 	
 	UFSD("ENTER, ino %lu, cgno %u, goal %llu, count %u\n",
 	     inode->i_ino, cgno, (unsigned long long)goal, count);
@@ -664,7 +667,7 @@ cg_found:
 		for (i = count; i < uspi->s_fpb; i++)
 			ubh_setbit (UCPI_UBH(ucpi), ucpi->c_freeoff, goal + i);
 		i = uspi->s_fpb - count;
-		vfs_dq_free_block(inode, i);
+		dquot_free_block(inode, i);
 
 		fs32_add(sb, &ucg->cg_cs.cs_nffree, i);
 		uspi->cs_total.cs_nffree += i;
@@ -676,8 +679,9 @@ cg_found:
 	result = ufs_bitmap_search (sb, ucpi, goal, allocsize);
 	if (result == INVBLOCK)
 		return 0;
-	if (vfs_dq_alloc_block(inode, count)) {
-		*err = -EDQUOT;
+	ret = dquot_alloc_block(inode, count);
+	if (ret) {
+		*err = ret;
 		return 0;
 	}
 	for (i = 0; i < count; i++)
@@ -714,6 +718,7 @@ static u64 ufs_alloccg_block(struct inode *inode,
 	struct ufs_super_block_first * usb1;
 	struct ufs_cylinder_group * ucg;
 	u64 result, blkno;
+	int ret;
 
 	UFSD("ENTER, goal %llu\n", (unsigned long long)goal);
 
@@ -747,8 +752,9 @@ gotit:
 	ubh_clrblock (UCPI_UBH(ucpi), ucpi->c_freeoff, blkno);
 	if ((UFS_SB(sb)->s_flags & UFS_CG_MASK) == UFS_CG_44BSD)
 		ufs_clusteracct (sb, ucpi, blkno, -1);
-	if (vfs_dq_alloc_block(inode, uspi->s_fpb)) {
-		*err = -EDQUOT;
+	ret = dquot_alloc_block(inode, uspi->s_fpb);
+	if (ret) {
+		*err = ret;
 		return INVBLOCK;
 	}
 
diff --git a/fs/ufs/dir.c b/fs/ufs/dir.c
index 22af68f8b682..317a0d444f6b 100644
--- a/fs/ufs/dir.c
+++ b/fs/ufs/dir.c
@@ -31,7 +31,7 @@
  * len <= UFS_MAXNAMLEN and de != NULL are guaranteed by caller.
  */
 static inline int ufs_match(struct super_block *sb, int len,
-		const char * const name, struct ufs_dir_entry * de)
+		const unsigned char *name, struct ufs_dir_entry *de)
 {
 	if (len != ufs_get_de_namlen(sb, de))
 		return 0;
@@ -70,7 +70,7 @@ static inline unsigned long ufs_dir_pages(struct inode *inode)
 	return (inode->i_size+PAGE_CACHE_SIZE-1)>>PAGE_CACHE_SHIFT;
 }
 
-ino_t ufs_inode_by_name(struct inode *dir, struct qstr *qstr)
+ino_t ufs_inode_by_name(struct inode *dir, const struct qstr *qstr)
 {
 	ino_t res = 0;
 	struct ufs_dir_entry *de;
@@ -249,11 +249,11 @@ struct ufs_dir_entry *ufs_dotdot(struct inode *dir, struct page **p)
  * (as a parameter - res_dir). Page is returned mapped and unlocked.
  * Entry is guaranteed to be valid.
  */
-struct ufs_dir_entry *ufs_find_entry(struct inode *dir, struct qstr *qstr,
+struct ufs_dir_entry *ufs_find_entry(struct inode *dir, const struct qstr *qstr,
 				     struct page **res_page)
 {
 	struct super_block *sb = dir->i_sb;
-	const char *name = qstr->name;
+	const unsigned char *name = qstr->name;
 	int namelen = qstr->len;
 	unsigned reclen = UFS_DIR_REC_LEN(namelen);
 	unsigned long start, n;
@@ -313,7 +313,7 @@ found:
 int ufs_add_link(struct dentry *dentry, struct inode *inode)
 {
 	struct inode *dir = dentry->d_parent->d_inode;
-	const char *name = dentry->d_name.name;
+	const unsigned char *name = dentry->d_name.name;
 	int namelen = dentry->d_name.len;
 	struct super_block *sb = dir->i_sb;
 	unsigned reclen = UFS_DIR_REC_LEN(namelen);
diff --git a/fs/ufs/file.c b/fs/ufs/file.c
index 73655c61240a..a8962cecde5b 100644
--- a/fs/ufs/file.c
+++ b/fs/ufs/file.c
@@ -24,6 +24,7 @@
  */
 
 #include <linux/fs.h>
+#include <linux/quotaops.h>
 
 #include "ufs_fs.h"
 #include "ufs.h"
@@ -40,7 +41,7 @@ const struct file_operations ufs_file_operations = {
 	.write		= do_sync_write,
 	.aio_write	= generic_file_aio_write,
 	.mmap		= generic_file_mmap,
-	.open           = generic_file_open,
+	.open           = dquot_file_open,
 	.fsync		= simple_fsync,
 	.splice_read	= generic_file_splice_read,
 };
diff --git a/fs/ufs/ialloc.c b/fs/ufs/ialloc.c
index 3527c00fef0d..230ecf608026 100644
--- a/fs/ufs/ialloc.c
+++ b/fs/ufs/ialloc.c
@@ -95,8 +95,8 @@ void ufs_free_inode (struct inode * inode)
 
 	is_directory = S_ISDIR(inode->i_mode);
 
-	vfs_dq_free_inode(inode);
-	vfs_dq_drop(inode);
+	dquot_free_inode(inode);
+	dquot_drop(inode);
 
 	clear_inode (inode);
 
@@ -355,9 +355,10 @@ cg_found:
 
 	unlock_super (sb);
 
-	if (vfs_dq_alloc_inode(inode)) {
-		vfs_dq_drop(inode);
-		err = -EDQUOT;
+	dquot_initialize(inode);
+	err = dquot_alloc_inode(inode);
+	if (err) {
+		dquot_drop(inode);
 		goto fail_without_unlock;
 	}
 
diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c
index 7cf33379fd46..80b68c3702d1 100644
--- a/fs/ufs/inode.c
+++ b/fs/ufs/inode.c
@@ -36,6 +36,8 @@
 #include <linux/mm.h>
 #include <linux/smp_lock.h>
 #include <linux/buffer_head.h>
+#include <linux/writeback.h>
+#include <linux/quotaops.h>
 
 #include "ufs_fs.h"
 #include "ufs.h"
@@ -890,11 +892,11 @@ static int ufs_update_inode(struct inode * inode, int do_sync)
 	return 0;
 }
 
-int ufs_write_inode (struct inode * inode, int wait)
+int ufs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	int ret;
 	lock_kernel();
-	ret = ufs_update_inode (inode, wait);
+	ret = ufs_update_inode(inode, wbc->sync_mode == WB_SYNC_ALL);
 	unlock_kernel();
 	return ret;
 }
@@ -908,6 +910,9 @@ void ufs_delete_inode (struct inode * inode)
 {
 	loff_t old_i_size;
 
+	if (!is_bad_inode(inode))
+		dquot_initialize(inode);
+
 	truncate_inode_pages(&inode->i_data, 0);
 	if (is_bad_inode(inode))
 		goto no_delete;
diff --git a/fs/ufs/namei.c b/fs/ufs/namei.c
index 4c26d9e8bc94..118556243e7a 100644
--- a/fs/ufs/namei.c
+++ b/fs/ufs/namei.c
@@ -30,6 +30,7 @@
 #include <linux/time.h>
 #include <linux/fs.h>
 #include <linux/smp_lock.h>
+#include <linux/quotaops.h>
 
 #include "ufs_fs.h"
 #include "ufs.h"
@@ -84,6 +85,9 @@ static int ufs_create (struct inode * dir, struct dentry * dentry, int mode,
 	int err;
 
 	UFSD("BEGIN\n");
+
+	dquot_initialize(dir);
+
 	inode = ufs_new_inode(dir, mode);
 	err = PTR_ERR(inode);
 
@@ -107,6 +111,9 @@ static int ufs_mknod (struct inode * dir, struct dentry *dentry, int mode, dev_t
 
 	if (!old_valid_dev(rdev))
 		return -EINVAL;
+
+	dquot_initialize(dir);
+
 	inode = ufs_new_inode(dir, mode);
 	err = PTR_ERR(inode);
 	if (!IS_ERR(inode)) {
@@ -131,6 +138,8 @@ static int ufs_symlink (struct inode * dir, struct dentry * dentry,
 	if (l > sb->s_blocksize)
 		goto out_notlocked;
 
+	dquot_initialize(dir);
+
 	lock_kernel();
 	inode = ufs_new_inode(dir, S_IFLNK | S_IRWXUGO);
 	err = PTR_ERR(inode);
@@ -176,6 +185,8 @@ static int ufs_link (struct dentry * old_dentry, struct inode * dir,
 		return -EMLINK;
 	}
 
+	dquot_initialize(dir);
+
 	inode->i_ctime = CURRENT_TIME_SEC;
 	inode_inc_link_count(inode);
 	atomic_inc(&inode->i_count);
@@ -193,6 +204,8 @@ static int ufs_mkdir(struct inode * dir, struct dentry * dentry, int mode)
 	if (dir->i_nlink >= UFS_LINK_MAX)
 		goto out;
 
+	dquot_initialize(dir);
+
 	lock_kernel();
 	inode_inc_link_count(dir);
 
@@ -237,6 +250,8 @@ static int ufs_unlink(struct inode *dir, struct dentry *dentry)
 	struct page *page;
 	int err = -ENOENT;
 
+	dquot_initialize(dir);
+
 	de = ufs_find_entry(dir, &dentry->d_name, &page);
 	if (!de)
 		goto out;
@@ -281,6 +296,9 @@ static int ufs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	struct ufs_dir_entry *old_de;
 	int err = -ENOENT;
 
+	dquot_initialize(old_dir);
+	dquot_initialize(new_dir);
+
 	old_de = ufs_find_entry(old_dir, &old_dentry->d_name, &old_page);
 	if (!old_de)
 		goto out;
diff --git a/fs/ufs/super.c b/fs/ufs/super.c
index 143c20bfb04b..66b63a751615 100644
--- a/fs/ufs/super.c
+++ b/fs/ufs/super.c
@@ -1432,6 +1432,11 @@ static void destroy_inodecache(void)
 	kmem_cache_destroy(ufs_inode_cachep);
 }
 
+static void ufs_clear_inode(struct inode *inode)
+{
+	dquot_drop(inode);
+}
+
 #ifdef CONFIG_QUOTA
 static ssize_t ufs_quota_read(struct super_block *, int, char *,size_t, loff_t);
 static ssize_t ufs_quota_write(struct super_block *, int, const char *, size_t, loff_t);
@@ -1442,6 +1447,7 @@ static const struct super_operations ufs_super_ops = {
 	.destroy_inode	= ufs_destroy_inode,
 	.write_inode	= ufs_write_inode,
 	.delete_inode	= ufs_delete_inode,
+	.clear_inode	= ufs_clear_inode,
 	.put_super	= ufs_put_super,
 	.write_super	= ufs_write_super,
 	.sync_fs	= ufs_sync_fs,
diff --git a/fs/ufs/truncate.c b/fs/ufs/truncate.c
index 41dd431ce228..d3b6270cb377 100644
--- a/fs/ufs/truncate.c
+++ b/fs/ufs/truncate.c
@@ -44,6 +44,7 @@
 #include <linux/buffer_head.h>
 #include <linux/blkdev.h>
 #include <linux/sched.h>
+#include <linux/quotaops.h>
 
 #include "ufs_fs.h"
 #include "ufs.h"
@@ -517,9 +518,18 @@ static int ufs_setattr(struct dentry *dentry, struct iattr *attr)
 	if (error)
 		return error;
 
+	if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
+	    (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) {
+		error = dquot_transfer(inode, attr);
+		if (error)
+			return error;
+	}
 	if (ia_valid & ATTR_SIZE &&
 	    attr->ia_size != i_size_read(inode)) {
 		loff_t old_i_size = inode->i_size;
+
+		dquot_initialize(inode);
+
 		error = vmtruncate(inode, attr->ia_size);
 		if (error)
 			return error;
diff --git a/fs/ufs/ufs.h b/fs/ufs/ufs.h
index 0b4c39bc0d9e..43f9f5d5670e 100644
--- a/fs/ufs/ufs.h
+++ b/fs/ufs/ufs.h
@@ -86,9 +86,9 @@ extern void ufs_put_cylinder (struct super_block *, unsigned);
 /* dir.c */
 extern const struct inode_operations ufs_dir_inode_operations;
 extern int ufs_add_link (struct dentry *, struct inode *);
-extern ino_t ufs_inode_by_name(struct inode *, struct qstr *);
+extern ino_t ufs_inode_by_name(struct inode *, const struct qstr *);
 extern int ufs_make_empty(struct inode *, struct inode *);
-extern struct ufs_dir_entry *ufs_find_entry(struct inode *, struct qstr *, struct page **);
+extern struct ufs_dir_entry *ufs_find_entry(struct inode *, const struct qstr *, struct page **);
 extern int ufs_delete_entry(struct inode *, struct ufs_dir_entry *, struct page *);
 extern int ufs_empty_dir (struct inode *);
 extern struct ufs_dir_entry *ufs_dotdot(struct inode *, struct page **);
@@ -106,7 +106,7 @@ extern struct inode * ufs_new_inode (struct inode *, int);
 
 /* inode.c */
 extern struct inode *ufs_iget(struct super_block *, unsigned long);
-extern int ufs_write_inode (struct inode *, int);
+extern int ufs_write_inode (struct inode *, struct writeback_control *);
 extern int ufs_sync_inode (struct inode *);
 extern void ufs_delete_inode (struct inode *);
 extern struct buffer_head * ufs_bread (struct inode *, unsigned, int, int *);
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 56641fe52a23..b4769e40e8bc 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -16,7 +16,7 @@
 # Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 #
 
-EXTRA_CFLAGS +=	 -I$(src) -I$(src)/linux-2.6 -funsigned-char
+EXTRA_CFLAGS +=	 -I$(src) -I$(src)/linux-2.6
 
 XFS_LINUX := linux-2.6
 
@@ -105,7 +105,6 @@ xfs-y				+= $(addprefix $(XFS_LINUX)/, \
 				   xfs_globals.o \
 				   xfs_ioctl.o \
 				   xfs_iops.o \
-				   xfs_lrw.o \
 				   xfs_super.o \
 				   xfs_sync.o \
 				   xfs_xattr.o)
diff --git a/fs/xfs/linux-2.6/kmem.c b/fs/xfs/linux-2.6/kmem.c
index 2d3f90afe5f1..bc7405585def 100644
--- a/fs/xfs/linux-2.6/kmem.c
+++ b/fs/xfs/linux-2.6/kmem.c
@@ -16,7 +16,6 @@
  * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
  */
 #include <linux/mm.h>
-#include <linux/vmalloc.h>
 #include <linux/highmem.h>
 #include <linux/swap.h>
 #include <linux/blkdev.h>
@@ -24,8 +23,25 @@
 #include "time.h"
 #include "kmem.h"
 
-#define MAX_VMALLOCS	6
-#define MAX_SLAB_SIZE	0x20000
+/*
+ * Greedy allocation.  May fail and may return vmalloced memory.
+ *
+ * Must be freed using kmem_free_large.
+ */
+void *
+kmem_zalloc_greedy(size_t *size, size_t minsize, size_t maxsize)
+{
+	void		*ptr;
+	size_t		kmsize = maxsize;
+
+	while (!(ptr = kmem_zalloc_large(kmsize))) {
+		if ((kmsize >>= 1) <= minsize)
+			kmsize = minsize;
+	}
+	if (ptr)
+		*size = kmsize;
+	return ptr;
+}
 
 void *
 kmem_alloc(size_t size, unsigned int __nocast flags)
@@ -34,19 +50,8 @@ kmem_alloc(size_t size, unsigned int __nocast flags)
 	gfp_t	lflags = kmem_flags_convert(flags);
 	void	*ptr;
 
-#ifdef DEBUG
-	if (unlikely(!(flags & KM_LARGE) && (size > PAGE_SIZE))) {
-		printk(KERN_WARNING "Large %s attempt, size=%ld\n",
-			__func__, (long)size);
-		dump_stack();
-	}
-#endif
-
 	do {
-		if (size < MAX_SLAB_SIZE || retries > MAX_VMALLOCS)
-			ptr = kmalloc(size, lflags);
-		else
-			ptr = __vmalloc(size, lflags, PAGE_KERNEL);
+		ptr = kmalloc(size, lflags);
 		if (ptr || (flags & (KM_MAYFAIL|KM_NOSLEEP)))
 			return ptr;
 		if (!(++retries % 100))
@@ -68,27 +73,6 @@ kmem_zalloc(size_t size, unsigned int __nocast flags)
 	return ptr;
 }
 
-void *
-kmem_zalloc_greedy(size_t *size, size_t minsize, size_t maxsize,
-		   unsigned int __nocast flags)
-{
-	void		*ptr;
-	size_t		kmsize = maxsize;
-	unsigned int	kmflags = (flags & ~KM_SLEEP) | KM_NOSLEEP;
-
-	while (!(ptr = kmem_zalloc(kmsize, kmflags))) {
-		if ((kmsize <= minsize) && (flags & KM_NOSLEEP))
-			break;
-		if ((kmsize >>= 1) <= minsize) {
-			kmsize = minsize;
-			kmflags = flags;
-		}
-	}
-	if (ptr)
-		*size = kmsize;
-	return ptr;
-}
-
 void
 kmem_free(const void *ptr)
 {
diff --git a/fs/xfs/linux-2.6/kmem.h b/fs/xfs/linux-2.6/kmem.h
index 179cbd630f69..f7c8f7a9ea6d 100644
--- a/fs/xfs/linux-2.6/kmem.h
+++ b/fs/xfs/linux-2.6/kmem.h
@@ -21,6 +21,7 @@
 #include <linux/slab.h>
 #include <linux/sched.h>
 #include <linux/mm.h>
+#include <linux/vmalloc.h>
 
 /*
  * General memory allocation interfaces
@@ -30,7 +31,6 @@
 #define KM_NOSLEEP	0x0002u
 #define KM_NOFS		0x0004u
 #define KM_MAYFAIL	0x0008u
-#define KM_LARGE	0x0010u
 
 /*
  * We use a special process flag to avoid recursive callbacks into
@@ -42,7 +42,7 @@ kmem_flags_convert(unsigned int __nocast flags)
 {
 	gfp_t	lflags;
 
-	BUG_ON(flags & ~(KM_SLEEP|KM_NOSLEEP|KM_NOFS|KM_MAYFAIL|KM_LARGE));
+	BUG_ON(flags & ~(KM_SLEEP|KM_NOSLEEP|KM_NOFS|KM_MAYFAIL));
 
 	if (flags & KM_NOSLEEP) {
 		lflags = GFP_ATOMIC | __GFP_NOWARN;
@@ -56,10 +56,25 @@ kmem_flags_convert(unsigned int __nocast flags)
 
 extern void *kmem_alloc(size_t, unsigned int __nocast);
 extern void *kmem_zalloc(size_t, unsigned int __nocast);
-extern void *kmem_zalloc_greedy(size_t *, size_t, size_t, unsigned int __nocast);
 extern void *kmem_realloc(const void *, size_t, size_t, unsigned int __nocast);
 extern void  kmem_free(const void *);
 
+static inline void *kmem_zalloc_large(size_t size)
+{
+	void *ptr;
+
+	ptr = vmalloc(size);
+	if (ptr)
+		memset(ptr, 0, size);
+	return ptr;
+}
+static inline void kmem_free_large(void *ptr)
+{
+	vfree(ptr);
+}
+
+extern void *kmem_zalloc_greedy(size_t *, size_t, size_t);
+
 /*
  * Zone interfaces
  */
diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c
index 883ca5ab8af5..bf85bbe4a9ae 100644
--- a/fs/xfs/linux-2.6/xfs_acl.c
+++ b/fs/xfs/linux-2.6/xfs_acl.c
@@ -106,7 +106,7 @@ xfs_get_acl(struct inode *inode, int type)
 	struct posix_acl *acl;
 	struct xfs_acl *xfs_acl;
 	int len = sizeof(struct xfs_acl);
-	char *ea_name;
+	unsigned char *ea_name;
 	int error;
 
 	acl = get_cached_acl(inode, type);
@@ -133,7 +133,8 @@ xfs_get_acl(struct inode *inode, int type)
 	if (!xfs_acl)
 		return ERR_PTR(-ENOMEM);
 
-	error = -xfs_attr_get(ip, ea_name, (char *)xfs_acl, &len, ATTR_ROOT);
+	error = -xfs_attr_get(ip, ea_name, (unsigned char *)xfs_acl,
+							&len, ATTR_ROOT);
 	if (error) {
 		/*
 		 * If the attribute doesn't exist make sure we have a negative
@@ -162,7 +163,7 @@ STATIC int
 xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 {
 	struct xfs_inode *ip = XFS_I(inode);
-	char *ea_name;
+	unsigned char *ea_name;
 	int error;
 
 	if (S_ISLNK(inode->i_mode))
@@ -194,7 +195,7 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 			(sizeof(struct xfs_acl_entry) *
 			 (XFS_ACL_MAX_ENTRIES - acl->a_count));
 
-		error = -xfs_attr_set(ip, ea_name, (char *)xfs_acl,
+		error = -xfs_attr_set(ip, ea_name, (unsigned char *)xfs_acl,
 				len, ATTR_ROOT);
 
 		kfree(xfs_acl);
@@ -262,7 +263,7 @@ xfs_set_mode(struct inode *inode, mode_t mode)
 }
 
 static int
-xfs_acl_exists(struct inode *inode, char *name)
+xfs_acl_exists(struct inode *inode, unsigned char *name)
 {
 	int len = sizeof(struct xfs_acl);
 
diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index 66abe36c1213..9083357f9e44 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -39,6 +39,7 @@
 #include "xfs_iomap.h"
 #include "xfs_vnodeops.h"
 #include "xfs_trace.h"
+#include "xfs_bmap.h"
 #include <linux/mpage.h>
 #include <linux/pagevec.h>
 #include <linux/writeback.h>
@@ -163,14 +164,17 @@ xfs_ioend_new_eof(
 }
 
 /*
- * Update on-disk file size now that data has been written to disk.
- * The current in-memory file size is i_size.  If a write is beyond
- * eof i_new_size will be the intended file size until i_size is
- * updated.  If this write does not extend all the way to the valid
- * file size then restrict this update to the end of the write.
+ * Update on-disk file size now that data has been written to disk.  The
+ * current in-memory file size is i_size.  If a write is beyond eof i_new_size
+ * will be the intended file size until i_size is updated.  If this write does
+ * not extend all the way to the valid file size then restrict this update to
+ * the end of the write.
+ *
+ * This function does not block as blocking on the inode lock in IO completion
+ * can lead to IO completion order dependency deadlocks.. If it can't get the
+ * inode ilock it will return EAGAIN. Callers must handle this.
  */
-
-STATIC void
+STATIC int
 xfs_setfilesize(
 	xfs_ioend_t		*ioend)
 {
@@ -181,16 +185,40 @@ xfs_setfilesize(
 	ASSERT(ioend->io_type != IOMAP_READ);
 
 	if (unlikely(ioend->io_error))
-		return;
+		return 0;
+
+	if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL))
+		return EAGAIN;
 
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	isize = xfs_ioend_new_eof(ioend);
 	if (isize) {
 		ip->i_d.di_size = isize;
-		xfs_mark_inode_dirty_sync(ip);
+		xfs_mark_inode_dirty(ip);
 	}
 
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	return 0;
+}
+
+/*
+ * Schedule IO completion handling on a xfsdatad if this was
+ * the final hold on this ioend. If we are asked to wait,
+ * flush the workqueue.
+ */
+STATIC void
+xfs_finish_ioend(
+	xfs_ioend_t	*ioend,
+	int		wait)
+{
+	if (atomic_dec_and_test(&ioend->io_remaining)) {
+		struct workqueue_struct *wq;
+
+		wq = (ioend->io_type == IOMAP_UNWRITTEN) ?
+			xfsconvertd_workqueue : xfsdatad_workqueue;
+		queue_work(wq, &ioend->io_work);
+		if (wait)
+			flush_workqueue(wq);
+	}
 }
 
 /*
@@ -198,11 +226,11 @@ xfs_setfilesize(
  */
 STATIC void
 xfs_end_io(
-	struct work_struct	*work)
+	struct work_struct *work)
 {
-	xfs_ioend_t		*ioend =
-		container_of(work, xfs_ioend_t, io_work);
-	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
+	xfs_ioend_t	*ioend = container_of(work, xfs_ioend_t, io_work);
+	struct xfs_inode *ip = XFS_I(ioend->io_inode);
+	int		error = 0;
 
 	/*
 	 * For unwritten extents we need to issue transactions to convert a
@@ -210,7 +238,6 @@ xfs_end_io(
 	 */
 	if (ioend->io_type == IOMAP_UNWRITTEN &&
 	    likely(!ioend->io_error && !XFS_FORCED_SHUTDOWN(ip->i_mount))) {
-		int error;
 
 		error = xfs_iomap_write_unwritten(ip, ioend->io_offset,
 						 ioend->io_size);
@@ -222,30 +249,23 @@ xfs_end_io(
 	 * We might have to update the on-disk file size after extending
 	 * writes.
 	 */
-	if (ioend->io_type != IOMAP_READ)
-		xfs_setfilesize(ioend);
-	xfs_destroy_ioend(ioend);
-}
-
-/*
- * Schedule IO completion handling on a xfsdatad if this was
- * the final hold on this ioend. If we are asked to wait,
- * flush the workqueue.
- */
-STATIC void
-xfs_finish_ioend(
-	xfs_ioend_t	*ioend,
-	int		wait)
-{
-	if (atomic_dec_and_test(&ioend->io_remaining)) {
-		struct workqueue_struct *wq;
-
-		wq = (ioend->io_type == IOMAP_UNWRITTEN) ?
-			xfsconvertd_workqueue : xfsdatad_workqueue;
-		queue_work(wq, &ioend->io_work);
-		if (wait)
-			flush_workqueue(wq);
+	if (ioend->io_type != IOMAP_READ) {
+		error = xfs_setfilesize(ioend);
+		ASSERT(!error || error == EAGAIN);
 	}
+
+	/*
+	 * If we didn't complete processing of the ioend, requeue it to the
+	 * tail of the workqueue for another attempt later. Otherwise destroy
+	 * it.
+	 */
+	if (error == EAGAIN) {
+		atomic_inc(&ioend->io_remaining);
+		xfs_finish_ioend(ioend, 0);
+		/* ensure we don't spin on blocked ioends */
+		delay(1);
+	} else
+		xfs_destroy_ioend(ioend);
 }
 
 /*
@@ -341,7 +361,7 @@ xfs_submit_ioend_bio(
 	 * but don't update the inode size until I/O completion.
 	 */
 	if (xfs_ioend_new_eof(ioend))
-		xfs_mark_inode_dirty_sync(XFS_I(ioend->io_inode));
+		xfs_mark_inode_dirty(XFS_I(ioend->io_inode));
 
 	submit_bio(wbc->sync_mode == WB_SYNC_ALL ?
 		   WRITE_SYNC_PLUG : WRITE, bio);
@@ -874,6 +894,118 @@ xfs_cluster_write(
 	}
 }
 
+STATIC void
+xfs_vm_invalidatepage(
+	struct page		*page,
+	unsigned long		offset)
+{
+	trace_xfs_invalidatepage(page->mapping->host, page, offset);
+	block_invalidatepage(page, offset);
+}
+
+/*
+ * If the page has delalloc buffers on it, we need to punch them out before we
+ * invalidate the page. If we don't, we leave a stale delalloc mapping on the
+ * inode that can trip a BUG() in xfs_get_blocks() later on if a direct IO read
+ * is done on that same region - the delalloc extent is returned when none is
+ * supposed to be there.
+ *
+ * We prevent this by truncating away the delalloc regions on the page before
+ * invalidating it. Because they are delalloc, we can do this without needing a
+ * transaction. Indeed - if we get ENOSPC errors, we have to be able to do this
+ * truncation without a transaction as there is no space left for block
+ * reservation (typically why we see a ENOSPC in writeback).
+ *
+ * This is not a performance critical path, so for now just do the punching a
+ * buffer head at a time.
+ */
+STATIC void
+xfs_aops_discard_page(
+	struct page		*page)
+{
+	struct inode		*inode = page->mapping->host;
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct buffer_head	*bh, *head;
+	loff_t			offset = page_offset(page);
+	ssize_t			len = 1 << inode->i_blkbits;
+
+	if (!xfs_is_delayed_page(page, IOMAP_DELAY))
+		goto out_invalidate;
+
+	xfs_fs_cmn_err(CE_ALERT, ip->i_mount,
+		"page discard on page %p, inode 0x%llx, offset %llu.",
+			page, ip->i_ino, offset);
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	bh = head = page_buffers(page);
+	do {
+		int		done;
+		xfs_fileoff_t	offset_fsb;
+		xfs_bmbt_irec_t	imap;
+		int		nimaps = 1;
+		int		error;
+		xfs_fsblock_t	firstblock;
+		xfs_bmap_free_t flist;
+
+		if (!buffer_delay(bh))
+			goto next_buffer;
+
+		offset_fsb = XFS_B_TO_FSBT(ip->i_mount, offset);
+
+		/*
+		 * Map the range first and check that it is a delalloc extent
+		 * before trying to unmap the range. Otherwise we will be
+		 * trying to remove a real extent (which requires a
+		 * transaction) or a hole, which is probably a bad idea...
+		 */
+		error = xfs_bmapi(NULL, ip, offset_fsb, 1,
+				XFS_BMAPI_ENTIRE,  NULL, 0, &imap,
+				&nimaps, NULL, NULL);
+
+		if (error) {
+			/* something screwed, just bail */
+			xfs_fs_cmn_err(CE_ALERT, ip->i_mount,
+			"page discard failed delalloc mapping lookup.");
+			break;
+		}
+		if (!nimaps) {
+			/* nothing there */
+			goto next_buffer;
+		}
+		if (imap.br_startblock != DELAYSTARTBLOCK) {
+			/* been converted, ignore */
+			goto next_buffer;
+		}
+		WARN_ON(imap.br_blockcount == 0);
+
+		/*
+		 * Note: while we initialise the firstblock/flist pair, they
+		 * should never be used because blocks should never be
+		 * allocated or freed for a delalloc extent and hence we need
+		 * don't cancel or finish them after the xfs_bunmapi() call.
+		 */
+		xfs_bmap_init(&flist, &firstblock);
+		error = xfs_bunmapi(NULL, ip, offset_fsb, 1, 0, 1, &firstblock,
+					&flist, NULL, &done);
+
+		ASSERT(!flist.xbf_count && !flist.xbf_first);
+		if (error) {
+			/* something screwed, just bail */
+			xfs_fs_cmn_err(CE_ALERT, ip->i_mount,
+			"page discard unable to remove delalloc mapping.");
+			break;
+		}
+next_buffer:
+		offset += len;
+
+	} while ((bh = bh->b_this_page) != head);
+
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+out_invalidate:
+	xfs_vm_invalidatepage(page, 0);
+	return;
+}
+
 /*
  * Calling this without startio set means we are being asked to make a dirty
  * page ready for freeing it's buffers.  When called with startio set then
@@ -1125,7 +1257,7 @@ error:
 	 */
 	if (err != -EAGAIN) {
 		if (!unmapped)
-			block_invalidatepage(page, 0);
+			xfs_aops_discard_page(page);
 		ClearPageUptodate(page);
 	}
 	return err;
@@ -1535,15 +1667,6 @@ xfs_vm_readpages(
 	return mpage_readpages(mapping, pages, nr_pages, xfs_get_blocks);
 }
 
-STATIC void
-xfs_vm_invalidatepage(
-	struct page		*page,
-	unsigned long		offset)
-{
-	trace_xfs_invalidatepage(page->mapping->host, page, offset);
-	block_invalidatepage(page, offset);
-}
-
 const struct address_space_operations xfs_address_space_operations = {
 	.readpage		= xfs_vm_readpage,
 	.readpages		= xfs_vm_readpages,
diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c
index 77b8be81c769..6f76ba85f193 100644
--- a/fs/xfs/linux-2.6/xfs_buf.c
+++ b/fs/xfs/linux-2.6/xfs_buf.c
@@ -33,6 +33,7 @@
 #include <linux/migrate.h>
 #include <linux/backing-dev.h>
 #include <linux/freezer.h>
+#include <linux/list_sort.h>
 
 #include "xfs_sb.h"
 #include "xfs_inum.h"
@@ -76,6 +77,27 @@ struct workqueue_struct *xfsconvertd_workqueue;
 #define xfs_buf_deallocate(bp) \
 	kmem_zone_free(xfs_buf_zone, (bp));
 
+static inline int
+xfs_buf_is_vmapped(
+	struct xfs_buf	*bp)
+{
+	/*
+	 * Return true if the buffer is vmapped.
+	 *
+	 * The XBF_MAPPED flag is set if the buffer should be mapped, but the
+	 * code is clever enough to know it doesn't have to map a single page,
+	 * so the check has to be both for XBF_MAPPED and bp->b_page_count > 1.
+	 */
+	return (bp->b_flags & XBF_MAPPED) && bp->b_page_count > 1;
+}
+
+static inline int
+xfs_buf_vmap_len(
+	struct xfs_buf	*bp)
+{
+	return (bp->b_page_count * PAGE_SIZE) - bp->b_offset;
+}
+
 /*
  *	Page Region interfaces.
  *
@@ -314,7 +336,7 @@ xfs_buf_free(
 	if (bp->b_flags & (_XBF_PAGE_CACHE|_XBF_PAGES)) {
 		uint		i;
 
-		if ((bp->b_flags & XBF_MAPPED) && (bp->b_page_count > 1))
+		if (xfs_buf_is_vmapped(bp))
 			free_address(bp->b_addr - bp->b_offset);
 
 		for (i = 0; i < bp->b_page_count; i++) {
@@ -1051,22 +1073,30 @@ xfs_buf_ioerror(
 }
 
 int
-xfs_bawrite(
-	void			*mp,
+xfs_bwrite(
+	struct xfs_mount	*mp,
 	struct xfs_buf		*bp)
 {
-	trace_xfs_buf_bawrite(bp, _RET_IP_);
+	int			iowait = (bp->b_flags & XBF_ASYNC) == 0;
+	int			error = 0;
 
-	ASSERT(bp->b_bn != XFS_BUF_DADDR_NULL);
+	bp->b_strat = xfs_bdstrat_cb;
+	bp->b_mount = mp;
+	bp->b_flags |= XBF_WRITE;
+	if (!iowait)
+		bp->b_flags |= _XBF_RUN_QUEUES;
 
 	xfs_buf_delwri_dequeue(bp);
+	xfs_buf_iostrategy(bp);
 
-	bp->b_flags &= ~(XBF_READ | XBF_DELWRI | XBF_READ_AHEAD);
-	bp->b_flags |= (XBF_WRITE | XBF_ASYNC | _XBF_RUN_QUEUES);
+	if (iowait) {
+		error = xfs_buf_iowait(bp);
+		if (error)
+			xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR);
+		xfs_buf_relse(bp);
+	}
 
-	bp->b_mount = mp;
-	bp->b_strat = xfs_bdstrat_cb;
-	return xfs_bdstrat_cb(bp);
+	return error;
 }
 
 void
@@ -1085,6 +1115,126 @@ xfs_bdwrite(
 	xfs_buf_delwri_queue(bp, 1);
 }
 
+/*
+ * Called when we want to stop a buffer from getting written or read.
+ * We attach the EIO error, muck with its flags, and call biodone
+ * so that the proper iodone callbacks get called.
+ */
+STATIC int
+xfs_bioerror(
+	xfs_buf_t *bp)
+{
+#ifdef XFSERRORDEBUG
+	ASSERT(XFS_BUF_ISREAD(bp) || bp->b_iodone);
+#endif
+
+	/*
+	 * No need to wait until the buffer is unpinned, we aren't flushing it.
+	 */
+	XFS_BUF_ERROR(bp, EIO);
+
+	/*
+	 * We're calling biodone, so delete XBF_DONE flag.
+	 */
+	XFS_BUF_UNREAD(bp);
+	XFS_BUF_UNDELAYWRITE(bp);
+	XFS_BUF_UNDONE(bp);
+	XFS_BUF_STALE(bp);
+
+	XFS_BUF_CLR_BDSTRAT_FUNC(bp);
+	xfs_biodone(bp);
+
+	return EIO;
+}
+
+/*
+ * Same as xfs_bioerror, except that we are releasing the buffer
+ * here ourselves, and avoiding the biodone call.
+ * This is meant for userdata errors; metadata bufs come with
+ * iodone functions attached, so that we can track down errors.
+ */
+STATIC int
+xfs_bioerror_relse(
+	struct xfs_buf	*bp)
+{
+	int64_t		fl = XFS_BUF_BFLAGS(bp);
+	/*
+	 * No need to wait until the buffer is unpinned.
+	 * We aren't flushing it.
+	 *
+	 * chunkhold expects B_DONE to be set, whether
+	 * we actually finish the I/O or not. We don't want to
+	 * change that interface.
+	 */
+	XFS_BUF_UNREAD(bp);
+	XFS_BUF_UNDELAYWRITE(bp);
+	XFS_BUF_DONE(bp);
+	XFS_BUF_STALE(bp);
+	XFS_BUF_CLR_IODONE_FUNC(bp);
+	XFS_BUF_CLR_BDSTRAT_FUNC(bp);
+	if (!(fl & XBF_ASYNC)) {
+		/*
+		 * Mark b_error and B_ERROR _both_.
+		 * Lot's of chunkcache code assumes that.
+		 * There's no reason to mark error for
+		 * ASYNC buffers.
+		 */
+		XFS_BUF_ERROR(bp, EIO);
+		XFS_BUF_FINISH_IOWAIT(bp);
+	} else {
+		xfs_buf_relse(bp);
+	}
+
+	return EIO;
+}
+
+
+/*
+ * All xfs metadata buffers except log state machine buffers
+ * get this attached as their b_bdstrat callback function.
+ * This is so that we can catch a buffer
+ * after prematurely unpinning it to forcibly shutdown the filesystem.
+ */
+int
+xfs_bdstrat_cb(
+	struct xfs_buf	*bp)
+{
+	if (XFS_FORCED_SHUTDOWN(bp->b_mount)) {
+		trace_xfs_bdstrat_shut(bp, _RET_IP_);
+		/*
+		 * Metadata write that didn't get logged but
+		 * written delayed anyway. These aren't associated
+		 * with a transaction, and can be ignored.
+		 */
+		if (!bp->b_iodone && !XFS_BUF_ISREAD(bp))
+			return xfs_bioerror_relse(bp);
+		else
+			return xfs_bioerror(bp);
+	}
+
+	xfs_buf_iorequest(bp);
+	return 0;
+}
+
+/*
+ * Wrapper around bdstrat so that we can stop data from going to disk in case
+ * we are shutting down the filesystem.  Typically user data goes thru this
+ * path; one of the exceptions is the superblock.
+ */
+void
+xfsbdstrat(
+	struct xfs_mount	*mp,
+	struct xfs_buf		*bp)
+{
+	if (XFS_FORCED_SHUTDOWN(mp)) {
+		trace_xfs_bdstrat_shut(bp, _RET_IP_);
+		xfs_bioerror_relse(bp);
+		return;
+	}
+
+	xfs_buf_iorequest(bp);
+}
+
 STATIC void
 _xfs_buf_ioend(
 	xfs_buf_t		*bp,
@@ -1107,6 +1257,9 @@ xfs_buf_bio_end_io(
 
 	xfs_buf_ioerror(bp, -error);
 
+	if (!error && xfs_buf_is_vmapped(bp) && (bp->b_flags & XBF_READ))
+		invalidate_kernel_vmap_range(bp->b_addr, xfs_buf_vmap_len(bp));
+
 	do {
 		struct page	*page = bvec->bv_page;
 
@@ -1216,6 +1369,10 @@ next_chunk:
 
 submit_io:
 	if (likely(bio->bi_size)) {
+		if (xfs_buf_is_vmapped(bp)) {
+			flush_kernel_vmap_range(bp->b_addr,
+						xfs_buf_vmap_len(bp));
+		}
 		submit_bio(rw, bio);
 		if (size)
 			goto next_chunk;
@@ -1296,7 +1453,7 @@ xfs_buf_iomove(
 	xfs_buf_t		*bp,	/* buffer to process		*/
 	size_t			boff,	/* starting buffer offset	*/
 	size_t			bsize,	/* length to copy		*/
-	caddr_t			data,	/* data address			*/
+	void			*data,	/* data address			*/
 	xfs_buf_rw_t		mode)	/* read/write/zero flag		*/
 {
 	size_t			bend, cpoff, csize;
@@ -1378,8 +1535,8 @@ xfs_alloc_bufhash(
 
 	btp->bt_hashshift = external ? 3 : 8;	/* 8 or 256 buckets */
 	btp->bt_hashmask = (1 << btp->bt_hashshift) - 1;
-	btp->bt_hash = kmem_zalloc((1 << btp->bt_hashshift) *
-					sizeof(xfs_bufhash_t), KM_SLEEP | KM_LARGE);
+	btp->bt_hash = kmem_zalloc_large((1 << btp->bt_hashshift) *
+					 sizeof(xfs_bufhash_t));
 	for (i = 0; i < (1 << btp->bt_hashshift); i++) {
 		spin_lock_init(&btp->bt_hash[i].bh_lock);
 		INIT_LIST_HEAD(&btp->bt_hash[i].bh_list);
@@ -1390,7 +1547,7 @@ STATIC void
 xfs_free_bufhash(
 	xfs_buftarg_t		*btp)
 {
-	kmem_free(btp->bt_hash);
+	kmem_free_large(btp->bt_hash);
 	btp->bt_hash = NULL;
 }
 
@@ -1595,6 +1752,11 @@ xfs_buf_delwri_queue(
 		list_del(&bp->b_list);
 	}
 
+	if (list_empty(dwq)) {
+		/* start xfsbufd as it is about to have something to do */
+		wake_up_process(bp->b_target->bt_task);
+	}
+
 	bp->b_flags |= _XBF_DELWRI_Q;
 	list_add_tail(&bp->b_list, dwq);
 	bp->b_queuetime = jiffies;
@@ -1626,6 +1788,35 @@ xfs_buf_delwri_dequeue(
 	trace_xfs_buf_delwri_dequeue(bp, _RET_IP_);
 }
 
+/*
+ * If a delwri buffer needs to be pushed before it has aged out, then promote
+ * it to the head of the delwri queue so that it will be flushed on the next
+ * xfsbufd run. We do this by resetting the queuetime of the buffer to be older
+ * than the age currently needed to flush the buffer. Hence the next time the
+ * xfsbufd sees it is guaranteed to be considered old enough to flush.
+ */
+void
+xfs_buf_delwri_promote(
+	struct xfs_buf	*bp)
+{
+	struct xfs_buftarg *btp = bp->b_target;
+	long		age = xfs_buf_age_centisecs * msecs_to_jiffies(10) + 1;
+
+	ASSERT(bp->b_flags & XBF_DELWRI);
+	ASSERT(bp->b_flags & _XBF_DELWRI_Q);
+
+	/*
+	 * Check the buffer age before locking the delayed write queue as we
+	 * don't need to promote buffers that are already past the flush age.
+	 */
+	if (bp->b_queuetime < jiffies - age)
+		return;
+	bp->b_queuetime = jiffies - age;
+	spin_lock(&btp->bt_delwrite_lock);
+	list_move(&bp->b_list, &btp->bt_delwrite_queue);
+	spin_unlock(&btp->bt_delwrite_lock);
+}
+
 STATIC void
 xfs_buf_runall_queues(
 	struct workqueue_struct	*queue)
@@ -1644,6 +1835,8 @@ xfsbufd_wakeup(
 	list_for_each_entry(btp, &xfs_buftarg_list, bt_list) {
 		if (test_bit(XBT_FORCE_SLEEP, &btp->bt_flags))
 			continue;
+		if (list_empty(&btp->bt_delwrite_queue))
+			continue;
 		set_bit(XBT_FORCE_FLUSH, &btp->bt_flags);
 		wake_up_process(btp->bt_task);
 	}
@@ -1694,20 +1887,53 @@ xfs_buf_delwri_split(
 
 }
 
+/*
+ * Compare function is more complex than it needs to be because
+ * the return value is only 32 bits and we are doing comparisons
+ * on 64 bit values
+ */
+static int
+xfs_buf_cmp(
+	void		*priv,
+	struct list_head *a,
+	struct list_head *b)
+{
+	struct xfs_buf	*ap = container_of(a, struct xfs_buf, b_list);
+	struct xfs_buf	*bp = container_of(b, struct xfs_buf, b_list);
+	xfs_daddr_t		diff;
+
+	diff = ap->b_bn - bp->b_bn;
+	if (diff < 0)
+		return -1;
+	if (diff > 0)
+		return 1;
+	return 0;
+}
+
+void
+xfs_buf_delwri_sort(
+	xfs_buftarg_t	*target,
+	struct list_head *list)
+{
+	list_sort(NULL, list, xfs_buf_cmp);
+}
+
 STATIC int
 xfsbufd(
 	void		*data)
 {
-	struct list_head tmp;
-	xfs_buftarg_t	*target = (xfs_buftarg_t *)data;
-	int		count;
-	xfs_buf_t	*bp;
+	xfs_buftarg_t   *target = (xfs_buftarg_t *)data;
 
 	current->flags |= PF_MEMALLOC;
 
 	set_freezable();
 
 	do {
+		long	age = xfs_buf_age_centisecs * msecs_to_jiffies(10);
+		long	tout = xfs_buf_timer_centisecs * msecs_to_jiffies(10);
+		int	count = 0;
+		struct list_head tmp;
+
 		if (unlikely(freezing(current))) {
 			set_bit(XBT_FORCE_SLEEP, &target->bt_flags);
 			refrigerator();
@@ -1715,17 +1941,16 @@ xfsbufd(
 			clear_bit(XBT_FORCE_SLEEP, &target->bt_flags);
 		}
 
-		schedule_timeout_interruptible(
-			xfs_buf_timer_centisecs * msecs_to_jiffies(10));
+		/* sleep for a long time if there is nothing to do. */
+		if (list_empty(&target->bt_delwrite_queue))
+			tout = MAX_SCHEDULE_TIMEOUT;
+		schedule_timeout_interruptible(tout);
 
-		xfs_buf_delwri_split(target, &tmp,
-				xfs_buf_age_centisecs * msecs_to_jiffies(10));
-
-		count = 0;
+		xfs_buf_delwri_split(target, &tmp, age);
+		list_sort(NULL, &tmp, xfs_buf_cmp);
 		while (!list_empty(&tmp)) {
-			bp = list_entry(tmp.next, xfs_buf_t, b_list);
-			ASSERT(target == bp->b_target);
-
+			struct xfs_buf *bp;
+			bp = list_first_entry(&tmp, struct xfs_buf, b_list);
 			list_del_init(&bp->b_list);
 			xfs_buf_iostrategy(bp);
 			count++;
@@ -1751,42 +1976,45 @@ xfs_flush_buftarg(
 	xfs_buftarg_t	*target,
 	int		wait)
 {
-	struct list_head tmp;
-	xfs_buf_t	*bp, *n;
+	xfs_buf_t	*bp;
 	int		pincount = 0;
+	LIST_HEAD(tmp_list);
+	LIST_HEAD(wait_list);
 
 	xfs_buf_runall_queues(xfsconvertd_workqueue);
 	xfs_buf_runall_queues(xfsdatad_workqueue);
 	xfs_buf_runall_queues(xfslogd_workqueue);
 
 	set_bit(XBT_FORCE_FLUSH, &target->bt_flags);
-	pincount = xfs_buf_delwri_split(target, &tmp, 0);
+	pincount = xfs_buf_delwri_split(target, &tmp_list, 0);
 
 	/*
-	 * Dropped the delayed write list lock, now walk the temporary list
+	 * Dropped the delayed write list lock, now walk the temporary list.
+	 * All I/O is issued async and then if we need to wait for completion
+	 * we do that after issuing all the IO.
 	 */
-	list_for_each_entry_safe(bp, n, &tmp, b_list) {
+	list_sort(NULL, &tmp_list, xfs_buf_cmp);
+	while (!list_empty(&tmp_list)) {
+		bp = list_first_entry(&tmp_list, struct xfs_buf, b_list);
 		ASSERT(target == bp->b_target);
-		if (wait)
+		list_del_init(&bp->b_list);
+		if (wait) {
 			bp->b_flags &= ~XBF_ASYNC;
-		else
-			list_del_init(&bp->b_list);
-
+			list_add(&bp->b_list, &wait_list);
+		}
 		xfs_buf_iostrategy(bp);
 	}
 
-	if (wait)
+	if (wait) {
+		/* Expedite and wait for IO to complete. */
 		blk_run_address_space(target->bt_mapping);
+		while (!list_empty(&wait_list)) {
+			bp = list_first_entry(&wait_list, struct xfs_buf, b_list);
 
-	/*
-	 * Remaining list items must be flushed before returning
-	 */
-	while (!list_empty(&tmp)) {
-		bp = list_entry(tmp.next, xfs_buf_t, b_list);
-
-		list_del_init(&bp->b_list);
-		xfs_iowait(bp);
-		xfs_buf_relse(bp);
+			list_del_init(&bp->b_list);
+			xfs_iowait(bp);
+			xfs_buf_relse(bp);
+		}
 	}
 
 	return pincount;
diff --git a/fs/xfs/linux-2.6/xfs_buf.h b/fs/xfs/linux-2.6/xfs_buf.h
index a34c7b54822d..386e7361e50e 100644
--- a/fs/xfs/linux-2.6/xfs_buf.h
+++ b/fs/xfs/linux-2.6/xfs_buf.h
@@ -232,13 +232,17 @@ extern void xfs_buf_lock(xfs_buf_t *);
 extern void xfs_buf_unlock(xfs_buf_t *);
 
 /* Buffer Read and Write Routines */
-extern int xfs_bawrite(void *mp, xfs_buf_t *bp);
+extern int xfs_bwrite(struct xfs_mount *mp, struct xfs_buf *bp);
 extern void xfs_bdwrite(void *mp, xfs_buf_t *bp);
+
+extern void xfsbdstrat(struct xfs_mount *, struct xfs_buf *);
+extern int xfs_bdstrat_cb(struct xfs_buf *);
+
 extern void xfs_buf_ioend(xfs_buf_t *,	int);
 extern void xfs_buf_ioerror(xfs_buf_t *, int);
 extern int xfs_buf_iorequest(xfs_buf_t *);
 extern int xfs_buf_iowait(xfs_buf_t *);
-extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, xfs_caddr_t,
+extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, void *,
 				xfs_buf_rw_t);
 
 static inline int xfs_buf_iostrategy(xfs_buf_t *bp)
@@ -261,6 +265,7 @@ extern int xfs_buf_ispin(xfs_buf_t *);
 
 /* Delayed Write Buffer Routines */
 extern void xfs_buf_delwri_dequeue(xfs_buf_t *);
+extern void xfs_buf_delwri_promote(xfs_buf_t *);
 
 /* Buffer Daemon Setup Routines */
 extern int xfs_buf_init(void);
@@ -270,33 +275,19 @@ extern void xfs_buf_terminate(void);
 	({ char __b[BDEVNAME_SIZE]; bdevname((target)->bt_bdev, __b); __b; })
 
 
-#define XFS_B_ASYNC		XBF_ASYNC
-#define XFS_B_DELWRI		XBF_DELWRI
-#define XFS_B_READ		XBF_READ
-#define XFS_B_WRITE		XBF_WRITE
-#define XFS_B_STALE		XBF_STALE
-
-#define XFS_BUF_TRYLOCK		XBF_TRYLOCK
-#define XFS_INCORE_TRYLOCK	XBF_TRYLOCK
-#define XFS_BUF_LOCK		XBF_LOCK
-#define XFS_BUF_MAPPED		XBF_MAPPED
-
-#define BUF_BUSY		XBF_DONT_BLOCK
-
 #define XFS_BUF_BFLAGS(bp)	((bp)->b_flags)
 #define XFS_BUF_ZEROFLAGS(bp)	((bp)->b_flags &= \
 		~(XBF_READ|XBF_WRITE|XBF_ASYNC|XBF_DELWRI|XBF_ORDERED))
 
-#define XFS_BUF_STALE(bp)	((bp)->b_flags |= XFS_B_STALE)
-#define XFS_BUF_UNSTALE(bp)	((bp)->b_flags &= ~XFS_B_STALE)
-#define XFS_BUF_ISSTALE(bp)	((bp)->b_flags & XFS_B_STALE)
+#define XFS_BUF_STALE(bp)	((bp)->b_flags |= XBF_STALE)
+#define XFS_BUF_UNSTALE(bp)	((bp)->b_flags &= ~XBF_STALE)
+#define XFS_BUF_ISSTALE(bp)	((bp)->b_flags & XBF_STALE)
 #define XFS_BUF_SUPER_STALE(bp)	do {				\
 					XFS_BUF_STALE(bp);	\
 					xfs_buf_delwri_dequeue(bp);	\
 					XFS_BUF_DONE(bp);	\
 				} while (0)
 
-#define XFS_BUF_MANAGE		XBF_FS_MANAGED
 #define XFS_BUF_UNMANAGE(bp)	((bp)->b_flags &= ~XBF_FS_MANAGED)
 
 #define XFS_BUF_DELAYWRITE(bp)		((bp)->b_flags |= XBF_DELWRI)
@@ -385,31 +376,11 @@ static inline void xfs_buf_relse(xfs_buf_t *bp)
 
 #define xfs_biomove(bp, off, len, data, rw) \
 	    xfs_buf_iomove((bp), (off), (len), (data), \
-		((rw) == XFS_B_WRITE) ? XBRW_WRITE : XBRW_READ)
+		((rw) == XBF_WRITE) ? XBRW_WRITE : XBRW_READ)
 
 #define xfs_biozero(bp, off, len) \
 	    xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)
 
-
-static inline int XFS_bwrite(xfs_buf_t *bp)
-{
-	int	iowait = (bp->b_flags & XBF_ASYNC) == 0;
-	int	error = 0;
-
-	if (!iowait)
-		bp->b_flags |= _XBF_RUN_QUEUES;
-
-	xfs_buf_delwri_dequeue(bp);
-	xfs_buf_iostrategy(bp);
-	if (iowait) {
-		error = xfs_buf_iowait(bp);
-		xfs_buf_relse(bp);
-	}
-	return error;
-}
-
-#define XFS_bdstrat(bp) xfs_buf_iorequest(bp)
-
 #define xfs_iowait(bp)	xfs_buf_iowait(bp)
 
 #define xfs_baread(target, rablkno, ralen)  \
@@ -424,6 +395,7 @@ extern void xfs_free_buftarg(struct xfs_mount *, struct xfs_buftarg *);
 extern void xfs_wait_buftarg(xfs_buftarg_t *);
 extern int xfs_setsize_buftarg(xfs_buftarg_t *, unsigned int, unsigned int);
 extern int xfs_flush_buftarg(xfs_buftarg_t *, int);
+
 #ifdef CONFIG_KDB_MODULES
 extern struct list_head *xfs_get_buftarg_list(void);
 #endif
diff --git a/fs/xfs/linux-2.6/xfs_export.c b/fs/xfs/linux-2.6/xfs_export.c
index 87b8cbd23d4b..846b75aeb2ab 100644
--- a/fs/xfs/linux-2.6/xfs_export.c
+++ b/fs/xfs/linux-2.6/xfs_export.c
@@ -29,6 +29,7 @@
 #include "xfs_vnodeops.h"
 #include "xfs_bmap_btree.h"
 #include "xfs_inode.h"
+#include "xfs_inode_item.h"
 
 /*
  * Note that we only accept fileids which are long enough rather than allow
@@ -215,9 +216,28 @@ xfs_fs_get_parent(
 	return d_obtain_alias(VFS_I(cip));
 }
 
+STATIC int
+xfs_fs_nfs_commit_metadata(
+	struct inode		*inode)
+{
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
+	int			error = 0;
+
+	xfs_ilock(ip, XFS_ILOCK_SHARED);
+	if (xfs_ipincount(ip)) {
+		error = _xfs_log_force_lsn(mp, ip->i_itemp->ili_last_lsn,
+				XFS_LOG_SYNC, NULL);
+	}
+	xfs_iunlock(ip, XFS_ILOCK_SHARED);
+
+	return error;
+}
+
 const struct export_operations xfs_export_operations = {
 	.encode_fh		= xfs_fs_encode_fh,
 	.fh_to_dentry		= xfs_fs_fh_to_dentry,
 	.fh_to_parent		= xfs_fs_fh_to_parent,
 	.get_parent		= xfs_fs_get_parent,
+	.commit_metadata	= xfs_fs_nfs_commit_metadata,
 };
diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c
index e4caeb28ce2e..42dd3bcfba6b 100644
--- a/fs/xfs/linux-2.6/xfs_file.c
+++ b/fs/xfs/linux-2.6/xfs_file.c
@@ -16,6 +16,7 @@
  * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
  */
 #include "xfs.h"
+#include "xfs_fs.h"
 #include "xfs_bit.h"
 #include "xfs_log.h"
 #include "xfs_inum.h"
@@ -34,52 +35,279 @@
 #include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
+#include "xfs_inode_item.h"
+#include "xfs_bmap.h"
 #include "xfs_error.h"
 #include "xfs_rw.h"
 #include "xfs_vnodeops.h"
 #include "xfs_da_btree.h"
 #include "xfs_ioctl.h"
+#include "xfs_trace.h"
 
 #include <linux/dcache.h>
 
 static const struct vm_operations_struct xfs_file_vm_ops;
 
-STATIC ssize_t
-xfs_file_aio_read(
-	struct kiocb		*iocb,
-	const struct iovec	*iov,
-	unsigned long		nr_segs,
-	loff_t			pos)
+/*
+ *	xfs_iozero
+ *
+ *	xfs_iozero clears the specified range of buffer supplied,
+ *	and marks all the affected blocks as valid and modified.  If
+ *	an affected block is not allocated, it will be allocated.  If
+ *	an affected block is not completely overwritten, and is not
+ *	valid before the operation, it will be read from disk before
+ *	being partially zeroed.
+ */
+STATIC int
+xfs_iozero(
+	struct xfs_inode	*ip,	/* inode			*/
+	loff_t			pos,	/* offset in file		*/
+	size_t			count)	/* size of data to zero		*/
 {
-	struct file		*file = iocb->ki_filp;
-	int			ioflags = 0;
+	struct page		*page;
+	struct address_space	*mapping;
+	int			status;
 
-	BUG_ON(iocb->ki_pos != pos);
-	if (unlikely(file->f_flags & O_DIRECT))
-		ioflags |= IO_ISDIRECT;
-	if (file->f_mode & FMODE_NOCMTIME)
-		ioflags |= IO_INVIS;
-	return xfs_read(XFS_I(file->f_path.dentry->d_inode), iocb, iov,
-				nr_segs, &iocb->ki_pos, ioflags);
+	mapping = VFS_I(ip)->i_mapping;
+	do {
+		unsigned offset, bytes;
+		void *fsdata;
+
+		offset = (pos & (PAGE_CACHE_SIZE -1)); /* Within page */
+		bytes = PAGE_CACHE_SIZE - offset;
+		if (bytes > count)
+			bytes = count;
+
+		status = pagecache_write_begin(NULL, mapping, pos, bytes,
+					AOP_FLAG_UNINTERRUPTIBLE,
+					&page, &fsdata);
+		if (status)
+			break;
+
+		zero_user(page, offset, bytes);
+
+		status = pagecache_write_end(NULL, mapping, pos, bytes, bytes,
+					page, fsdata);
+		WARN_ON(status <= 0); /* can't return less than zero! */
+		pos += bytes;
+		count -= bytes;
+		status = 0;
+	} while (count);
+
+	return (-status);
+}
+
+STATIC int
+xfs_file_fsync(
+	struct file		*file,
+	struct dentry		*dentry,
+	int			datasync)
+{
+	struct xfs_inode	*ip = XFS_I(dentry->d_inode);
+	struct xfs_trans	*tp;
+	int			error = 0;
+	int			log_flushed = 0;
+
+	xfs_itrace_entry(ip);
+
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
+		return -XFS_ERROR(EIO);
+
+	xfs_iflags_clear(ip, XFS_ITRUNCATED);
+
+	/*
+	 * We always need to make sure that the required inode state is safe on
+	 * disk.  The inode might be clean but we still might need to force the
+	 * log because of committed transactions that haven't hit the disk yet.
+	 * Likewise, there could be unflushed non-transactional changes to the
+	 * inode core that have to go to disk and this requires us to issue
+	 * a synchronous transaction to capture these changes correctly.
+	 *
+	 * This code relies on the assumption that if the i_update_core field
+	 * of the inode is clear and the inode is unpinned then it is clean
+	 * and no action is required.
+	 */
+	xfs_ilock(ip, XFS_ILOCK_SHARED);
+
+	/*
+	 * First check if the VFS inode is marked dirty.  All the dirtying
+	 * of non-transactional updates no goes through mark_inode_dirty*,
+	 * which allows us to distinguish beteeen pure timestamp updates
+	 * and i_size updates which need to be caught for fdatasync.
+	 * After that also theck for the dirty state in the XFS inode, which
+	 * might gets cleared when the inode gets written out via the AIL
+	 * or xfs_iflush_cluster.
+	 */
+	if (((dentry->d_inode->i_state & I_DIRTY_DATASYNC) ||
+	    ((dentry->d_inode->i_state & I_DIRTY_SYNC) && !datasync)) &&
+	    ip->i_update_core) {
+		/*
+		 * Kick off a transaction to log the inode core to get the
+		 * updates.  The sync transaction will also force the log.
+		 */
+		xfs_iunlock(ip, XFS_ILOCK_SHARED);
+		tp = xfs_trans_alloc(ip->i_mount, XFS_TRANS_FSYNC_TS);
+		error = xfs_trans_reserve(tp, 0,
+				XFS_FSYNC_TS_LOG_RES(ip->i_mount), 0, 0, 0);
+		if (error) {
+			xfs_trans_cancel(tp, 0);
+			return -error;
+		}
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+
+		/*
+		 * Note - it's possible that we might have pushed ourselves out
+		 * of the way during trans_reserve which would flush the inode.
+		 * But there's no guarantee that the inode buffer has actually
+		 * gone out yet (it's delwri).	Plus the buffer could be pinned
+		 * anyway if it's part of an inode in another recent
+		 * transaction.	 So we play it safe and fire off the
+		 * transaction anyway.
+		 */
+		xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+		xfs_trans_ihold(tp, ip);
+		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+		xfs_trans_set_sync(tp);
+		error = _xfs_trans_commit(tp, 0, &log_flushed);
+
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	} else {
+		/*
+		 * Timestamps/size haven't changed since last inode flush or
+		 * inode transaction commit.  That means either nothing got
+		 * written or a transaction committed which caught the updates.
+		 * If the latter happened and the transaction hasn't hit the
+		 * disk yet, the inode will be still be pinned.  If it is,
+		 * force the log.
+		 */
+		if (xfs_ipincount(ip)) {
+			error = _xfs_log_force_lsn(ip->i_mount,
+					ip->i_itemp->ili_last_lsn,
+					XFS_LOG_SYNC, &log_flushed);
+		}
+		xfs_iunlock(ip, XFS_ILOCK_SHARED);
+	}
+
+	if (ip->i_mount->m_flags & XFS_MOUNT_BARRIER) {
+		/*
+		 * If the log write didn't issue an ordered tag we need
+		 * to flush the disk cache for the data device now.
+		 */
+		if (!log_flushed)
+			xfs_blkdev_issue_flush(ip->i_mount->m_ddev_targp);
+
+		/*
+		 * If this inode is on the RT dev we need to flush that
+		 * cache as well.
+		 */
+		if (XFS_IS_REALTIME_INODE(ip))
+			xfs_blkdev_issue_flush(ip->i_mount->m_rtdev_targp);
+	}
+
+	return -error;
 }
 
 STATIC ssize_t
-xfs_file_aio_write(
+xfs_file_aio_read(
 	struct kiocb		*iocb,
-	const struct iovec	*iov,
+	const struct iovec	*iovp,
 	unsigned long		nr_segs,
 	loff_t			pos)
 {
 	struct file		*file = iocb->ki_filp;
+	struct inode		*inode = file->f_mapping->host;
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
+	size_t			size = 0;
+	ssize_t			ret = 0;
 	int			ioflags = 0;
+	xfs_fsize_t		n;
+	unsigned long		seg;
+
+	XFS_STATS_INC(xs_read_calls);
 
 	BUG_ON(iocb->ki_pos != pos);
+
 	if (unlikely(file->f_flags & O_DIRECT))
 		ioflags |= IO_ISDIRECT;
 	if (file->f_mode & FMODE_NOCMTIME)
 		ioflags |= IO_INVIS;
-	return xfs_write(XFS_I(file->f_mapping->host), iocb, iov, nr_segs,
-				&iocb->ki_pos, ioflags);
+
+	/* START copy & waste from filemap.c */
+	for (seg = 0; seg < nr_segs; seg++) {
+		const struct iovec *iv = &iovp[seg];
+
+		/*
+		 * If any segment has a negative length, or the cumulative
+		 * length ever wraps negative then return -EINVAL.
+		 */
+		size += iv->iov_len;
+		if (unlikely((ssize_t)(size|iv->iov_len) < 0))
+			return XFS_ERROR(-EINVAL);
+	}
+	/* END copy & waste from filemap.c */
+
+	if (unlikely(ioflags & IO_ISDIRECT)) {
+		xfs_buftarg_t	*target =
+			XFS_IS_REALTIME_INODE(ip) ?
+				mp->m_rtdev_targp : mp->m_ddev_targp;
+		if ((iocb->ki_pos & target->bt_smask) ||
+		    (size & target->bt_smask)) {
+			if (iocb->ki_pos == ip->i_size)
+				return 0;
+			return -XFS_ERROR(EINVAL);
+		}
+	}
+
+	n = XFS_MAXIOFFSET(mp) - iocb->ki_pos;
+	if (n <= 0 || size == 0)
+		return 0;
+
+	if (n < size)
+		size = n;
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return -EIO;
+
+	if (unlikely(ioflags & IO_ISDIRECT))
+		mutex_lock(&inode->i_mutex);
+	xfs_ilock(ip, XFS_IOLOCK_SHARED);
+
+	if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && !(ioflags & IO_INVIS)) {
+		int dmflags = FILP_DELAY_FLAG(file) | DM_SEM_FLAG_RD(ioflags);
+		int iolock = XFS_IOLOCK_SHARED;
+
+		ret = -XFS_SEND_DATA(mp, DM_EVENT_READ, ip, iocb->ki_pos, size,
+					dmflags, &iolock);
+		if (ret) {
+			xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+			if (unlikely(ioflags & IO_ISDIRECT))
+				mutex_unlock(&inode->i_mutex);
+			return ret;
+		}
+	}
+
+	if (unlikely(ioflags & IO_ISDIRECT)) {
+		if (inode->i_mapping->nrpages) {
+			ret = -xfs_flushinval_pages(ip,
+					(iocb->ki_pos & PAGE_CACHE_MASK),
+					-1, FI_REMAPF_LOCKED);
+		}
+		mutex_unlock(&inode->i_mutex);
+		if (ret) {
+			xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+			return ret;
+		}
+	}
+
+	trace_xfs_file_read(ip, size, iocb->ki_pos, ioflags);
+
+	ret = generic_file_aio_read(iocb, iovp, nr_segs, iocb->ki_pos);
+	if (ret > 0)
+		XFS_STATS_ADD(xs_read_bytes, ret);
+
+	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+	return ret;
 }
 
 STATIC ssize_t
@@ -87,16 +315,44 @@ xfs_file_splice_read(
 	struct file		*infilp,
 	loff_t			*ppos,
 	struct pipe_inode_info	*pipe,
-	size_t			len,
+	size_t			count,
 	unsigned int		flags)
 {
+	struct xfs_inode	*ip = XFS_I(infilp->f_mapping->host);
+	struct xfs_mount	*mp = ip->i_mount;
 	int			ioflags = 0;
+	ssize_t			ret;
+
+	XFS_STATS_INC(xs_read_calls);
 
 	if (infilp->f_mode & FMODE_NOCMTIME)
 		ioflags |= IO_INVIS;
 
-	return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode),
-				   infilp, ppos, pipe, len, flags, ioflags);
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
+		return -EIO;
+
+	xfs_ilock(ip, XFS_IOLOCK_SHARED);
+
+	if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && !(ioflags & IO_INVIS)) {
+		int iolock = XFS_IOLOCK_SHARED;
+		int error;
+
+		error = XFS_SEND_DATA(mp, DM_EVENT_READ, ip, *ppos, count,
+					FILP_DELAY_FLAG(infilp), &iolock);
+		if (error) {
+			xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+			return -error;
+		}
+	}
+
+	trace_xfs_file_splice_read(ip, count, *ppos, ioflags);
+
+	ret = generic_file_splice_read(infilp, ppos, pipe, count, flags);
+	if (ret > 0)
+		XFS_STATS_ADD(xs_read_bytes, ret);
+
+	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+	return ret;
 }
 
 STATIC ssize_t
@@ -104,16 +360,538 @@ xfs_file_splice_write(
 	struct pipe_inode_info	*pipe,
 	struct file		*outfilp,
 	loff_t			*ppos,
-	size_t			len,
+	size_t			count,
 	unsigned int		flags)
 {
+	struct inode		*inode = outfilp->f_mapping->host;
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
+	xfs_fsize_t		isize, new_size;
 	int			ioflags = 0;
+	ssize_t			ret;
+
+	XFS_STATS_INC(xs_write_calls);
 
 	if (outfilp->f_mode & FMODE_NOCMTIME)
 		ioflags |= IO_INVIS;
 
-	return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode),
-				    pipe, outfilp, ppos, len, flags, ioflags);
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
+		return -EIO;
+
+	xfs_ilock(ip, XFS_IOLOCK_EXCL);
+
+	if (DM_EVENT_ENABLED(ip, DM_EVENT_WRITE) && !(ioflags & IO_INVIS)) {
+		int iolock = XFS_IOLOCK_EXCL;
+		int error;
+
+		error = XFS_SEND_DATA(mp, DM_EVENT_WRITE, ip, *ppos, count,
+					FILP_DELAY_FLAG(outfilp), &iolock);
+		if (error) {
+			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
+			return -error;
+		}
+	}
+
+	new_size = *ppos + count;
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	if (new_size > ip->i_size)
+		ip->i_new_size = new_size;
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+
+	trace_xfs_file_splice_write(ip, count, *ppos, ioflags);
+
+	ret = generic_file_splice_write(pipe, outfilp, ppos, count, flags);
+	if (ret > 0)
+		XFS_STATS_ADD(xs_write_bytes, ret);
+
+	isize = i_size_read(inode);
+	if (unlikely(ret < 0 && ret != -EFAULT && *ppos > isize))
+		*ppos = isize;
+
+	if (*ppos > ip->i_size) {
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+		if (*ppos > ip->i_size)
+			ip->i_size = *ppos;
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	}
+
+	if (ip->i_new_size) {
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+		ip->i_new_size = 0;
+		if (ip->i_d.di_size > ip->i_size)
+			ip->i_d.di_size = ip->i_size;
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	}
+	xfs_iunlock(ip, XFS_IOLOCK_EXCL);
+	return ret;
+}
+
+/*
+ * This routine is called to handle zeroing any space in the last
+ * block of the file that is beyond the EOF.  We do this since the
+ * size is being increased without writing anything to that block
+ * and we don't want anyone to read the garbage on the disk.
+ */
+STATIC int				/* error (positive) */
+xfs_zero_last_block(
+	xfs_inode_t	*ip,
+	xfs_fsize_t	offset,
+	xfs_fsize_t	isize)
+{
+	xfs_fileoff_t	last_fsb;
+	xfs_mount_t	*mp = ip->i_mount;
+	int		nimaps;
+	int		zero_offset;
+	int		zero_len;
+	int		error = 0;
+	xfs_bmbt_irec_t	imap;
+
+	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
+
+	zero_offset = XFS_B_FSB_OFFSET(mp, isize);
+	if (zero_offset == 0) {
+		/*
+		 * There are no extra bytes in the last block on disk to
+		 * zero, so return.
+		 */
+		return 0;
+	}
+
+	last_fsb = XFS_B_TO_FSBT(mp, isize);
+	nimaps = 1;
+	error = xfs_bmapi(NULL, ip, last_fsb, 1, 0, NULL, 0, &imap,
+			  &nimaps, NULL, NULL);
+	if (error) {
+		return error;
+	}
+	ASSERT(nimaps > 0);
+	/*
+	 * If the block underlying isize is just a hole, then there
+	 * is nothing to zero.
+	 */
+	if (imap.br_startblock == HOLESTARTBLOCK) {
+		return 0;
+	}
+	/*
+	 * Zero the part of the last block beyond the EOF, and write it
+	 * out sync.  We need to drop the ilock while we do this so we
+	 * don't deadlock when the buffer cache calls back to us.
+	 */
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+
+	zero_len = mp->m_sb.sb_blocksize - zero_offset;
+	if (isize + zero_len > offset)
+		zero_len = offset - isize;
+	error = xfs_iozero(ip, isize, zero_len);
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	ASSERT(error >= 0);
+	return error;
+}
+
+/*
+ * Zero any on disk space between the current EOF and the new,
+ * larger EOF.  This handles the normal case of zeroing the remainder
+ * of the last block in the file and the unusual case of zeroing blocks
+ * out beyond the size of the file.  This second case only happens
+ * with fixed size extents and when the system crashes before the inode
+ * size was updated but after blocks were allocated.  If fill is set,
+ * then any holes in the range are filled and zeroed.  If not, the holes
+ * are left alone as holes.
+ */
+
+int					/* error (positive) */
+xfs_zero_eof(
+	xfs_inode_t	*ip,
+	xfs_off_t	offset,		/* starting I/O offset */
+	xfs_fsize_t	isize)		/* current inode size */
+{
+	xfs_mount_t	*mp = ip->i_mount;
+	xfs_fileoff_t	start_zero_fsb;
+	xfs_fileoff_t	end_zero_fsb;
+	xfs_fileoff_t	zero_count_fsb;
+	xfs_fileoff_t	last_fsb;
+	xfs_fileoff_t	zero_off;
+	xfs_fsize_t	zero_len;
+	int		nimaps;
+	int		error = 0;
+	xfs_bmbt_irec_t	imap;
+
+	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
+	ASSERT(offset > isize);
+
+	/*
+	 * First handle zeroing the block on which isize resides.
+	 * We only zero a part of that block so it is handled specially.
+	 */
+	error = xfs_zero_last_block(ip, offset, isize);
+	if (error) {
+		ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
+		return error;
+	}
+
+	/*
+	 * Calculate the range between the new size and the old
+	 * where blocks needing to be zeroed may exist.  To get the
+	 * block where the last byte in the file currently resides,
+	 * we need to subtract one from the size and truncate back
+	 * to a block boundary.  We subtract 1 in case the size is
+	 * exactly on a block boundary.
+	 */
+	last_fsb = isize ? XFS_B_TO_FSBT(mp, isize - 1) : (xfs_fileoff_t)-1;
+	start_zero_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)isize);
+	end_zero_fsb = XFS_B_TO_FSBT(mp, offset - 1);
+	ASSERT((xfs_sfiloff_t)last_fsb < (xfs_sfiloff_t)start_zero_fsb);
+	if (last_fsb == end_zero_fsb) {
+		/*
+		 * The size was only incremented on its last block.
+		 * We took care of that above, so just return.
+		 */
+		return 0;
+	}
+
+	ASSERT(start_zero_fsb <= end_zero_fsb);
+	while (start_zero_fsb <= end_zero_fsb) {
+		nimaps = 1;
+		zero_count_fsb = end_zero_fsb - start_zero_fsb + 1;
+		error = xfs_bmapi(NULL, ip, start_zero_fsb, zero_count_fsb,
+				  0, NULL, 0, &imap, &nimaps, NULL, NULL);
+		if (error) {
+			ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
+			return error;
+		}
+		ASSERT(nimaps > 0);
+
+		if (imap.br_state == XFS_EXT_UNWRITTEN ||
+		    imap.br_startblock == HOLESTARTBLOCK) {
+			/*
+			 * This loop handles initializing pages that were
+			 * partially initialized by the code below this
+			 * loop. It basically zeroes the part of the page
+			 * that sits on a hole and sets the page as P_HOLE
+			 * and calls remapf if it is a mapped file.
+			 */
+			start_zero_fsb = imap.br_startoff + imap.br_blockcount;
+			ASSERT(start_zero_fsb <= (end_zero_fsb + 1));
+			continue;
+		}
+
+		/*
+		 * There are blocks we need to zero.
+		 * Drop the inode lock while we're doing the I/O.
+		 * We'll still have the iolock to protect us.
+		 */
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+
+		zero_off = XFS_FSB_TO_B(mp, start_zero_fsb);
+		zero_len = XFS_FSB_TO_B(mp, imap.br_blockcount);
+
+		if ((zero_off + zero_len) > offset)
+			zero_len = offset - zero_off;
+
+		error = xfs_iozero(ip, zero_off, zero_len);
+		if (error) {
+			goto out_lock;
+		}
+
+		start_zero_fsb = imap.br_startoff + imap.br_blockcount;
+		ASSERT(start_zero_fsb <= (end_zero_fsb + 1));
+
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+	}
+
+	return 0;
+
+out_lock:
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	ASSERT(error >= 0);
+	return error;
+}
+
+STATIC ssize_t
+xfs_file_aio_write(
+	struct kiocb		*iocb,
+	const struct iovec	*iovp,
+	unsigned long		nr_segs,
+	loff_t			pos)
+{
+	struct file		*file = iocb->ki_filp;
+	struct address_space	*mapping = file->f_mapping;
+	struct inode		*inode = mapping->host;
+	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
+	ssize_t			ret = 0, error = 0;
+	int			ioflags = 0;
+	xfs_fsize_t		isize, new_size;
+	int			iolock;
+	int			eventsent = 0;
+	size_t			ocount = 0, count;
+	int			need_i_mutex;
+
+	XFS_STATS_INC(xs_write_calls);
+
+	BUG_ON(iocb->ki_pos != pos);
+
+	if (unlikely(file->f_flags & O_DIRECT))
+		ioflags |= IO_ISDIRECT;
+	if (file->f_mode & FMODE_NOCMTIME)
+		ioflags |= IO_INVIS;
+
+	error = generic_segment_checks(iovp, &nr_segs, &ocount, VERIFY_READ);
+	if (error)
+		return error;
+
+	count = ocount;
+	if (count == 0)
+		return 0;
+
+	xfs_wait_for_freeze(mp, SB_FREEZE_WRITE);
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return -EIO;
+
+relock:
+	if (ioflags & IO_ISDIRECT) {
+		iolock = XFS_IOLOCK_SHARED;
+		need_i_mutex = 0;
+	} else {
+		iolock = XFS_IOLOCK_EXCL;
+		need_i_mutex = 1;
+		mutex_lock(&inode->i_mutex);
+	}
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL|iolock);
+
+start:
+	error = -generic_write_checks(file, &pos, &count,
+					S_ISBLK(inode->i_mode));
+	if (error) {
+		xfs_iunlock(ip, XFS_ILOCK_EXCL|iolock);
+		goto out_unlock_mutex;
+	}
+
+	if ((DM_EVENT_ENABLED(ip, DM_EVENT_WRITE) &&
+	    !(ioflags & IO_INVIS) && !eventsent)) {
+		int		dmflags = FILP_DELAY_FLAG(file);
+
+		if (need_i_mutex)
+			dmflags |= DM_FLAGS_IMUX;
+
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+		error = XFS_SEND_DATA(ip->i_mount, DM_EVENT_WRITE, ip,
+				      pos, count, dmflags, &iolock);
+		if (error) {
+			goto out_unlock_internal;
+		}
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+		eventsent = 1;
+
+		/*
+		 * The iolock was dropped and reacquired in XFS_SEND_DATA
+		 * so we have to recheck the size when appending.
+		 * We will only "goto start;" once, since having sent the
+		 * event prevents another call to XFS_SEND_DATA, which is
+		 * what allows the size to change in the first place.
+		 */
+		if ((file->f_flags & O_APPEND) && pos != ip->i_size)
+			goto start;
+	}
+
+	if (ioflags & IO_ISDIRECT) {
+		xfs_buftarg_t	*target =
+			XFS_IS_REALTIME_INODE(ip) ?
+				mp->m_rtdev_targp : mp->m_ddev_targp;
+
+		if ((pos & target->bt_smask) || (count & target->bt_smask)) {
+			xfs_iunlock(ip, XFS_ILOCK_EXCL|iolock);
+			return XFS_ERROR(-EINVAL);
+		}
+
+		if (!need_i_mutex && (mapping->nrpages || pos > ip->i_size)) {
+			xfs_iunlock(ip, XFS_ILOCK_EXCL|iolock);
+			iolock = XFS_IOLOCK_EXCL;
+			need_i_mutex = 1;
+			mutex_lock(&inode->i_mutex);
+			xfs_ilock(ip, XFS_ILOCK_EXCL|iolock);
+			goto start;
+		}
+	}
+
+	new_size = pos + count;
+	if (new_size > ip->i_size)
+		ip->i_new_size = new_size;
+
+	if (likely(!(ioflags & IO_INVIS)))
+		file_update_time(file);
+
+	/*
+	 * If the offset is beyond the size of the file, we have a couple
+	 * of things to do. First, if there is already space allocated
+	 * we need to either create holes or zero the disk or ...
+	 *
+	 * If there is a page where the previous size lands, we need
+	 * to zero it out up to the new size.
+	 */
+
+	if (pos > ip->i_size) {
+		error = xfs_zero_eof(ip, pos, ip->i_size);
+		if (error) {
+			xfs_iunlock(ip, XFS_ILOCK_EXCL);
+			goto out_unlock_internal;
+		}
+	}
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+
+	/*
+	 * If we're writing the file then make sure to clear the
+	 * setuid and setgid bits if the process is not being run
+	 * by root.  This keeps people from modifying setuid and
+	 * setgid binaries.
+	 */
+	error = -file_remove_suid(file);
+	if (unlikely(error))
+		goto out_unlock_internal;
+
+	/* We can write back this queue in page reclaim */
+	current->backing_dev_info = mapping->backing_dev_info;
+
+	if ((ioflags & IO_ISDIRECT)) {
+		if (mapping->nrpages) {
+			WARN_ON(need_i_mutex == 0);
+			error = xfs_flushinval_pages(ip,
+					(pos & PAGE_CACHE_MASK),
+					-1, FI_REMAPF_LOCKED);
+			if (error)
+				goto out_unlock_internal;
+		}
+
+		if (need_i_mutex) {
+			/* demote the lock now the cached pages are gone */
+			xfs_ilock_demote(ip, XFS_IOLOCK_EXCL);
+			mutex_unlock(&inode->i_mutex);
+
+			iolock = XFS_IOLOCK_SHARED;
+			need_i_mutex = 0;
+		}
+
+		trace_xfs_file_direct_write(ip, count, iocb->ki_pos, ioflags);
+		ret = generic_file_direct_write(iocb, iovp,
+				&nr_segs, pos, &iocb->ki_pos, count, ocount);
+
+		/*
+		 * direct-io write to a hole: fall through to buffered I/O
+		 * for completing the rest of the request.
+		 */
+		if (ret >= 0 && ret != count) {
+			XFS_STATS_ADD(xs_write_bytes, ret);
+
+			pos += ret;
+			count -= ret;
+
+			ioflags &= ~IO_ISDIRECT;
+			xfs_iunlock(ip, iolock);
+			goto relock;
+		}
+	} else {
+		int enospc = 0;
+		ssize_t ret2 = 0;
+
+write_retry:
+		trace_xfs_file_buffered_write(ip, count, iocb->ki_pos, ioflags);
+		ret2 = generic_file_buffered_write(iocb, iovp, nr_segs,
+				pos, &iocb->ki_pos, count, ret);
+		/*
+		 * if we just got an ENOSPC, flush the inode now we
+		 * aren't holding any page locks and retry *once*
+		 */
+		if (ret2 == -ENOSPC && !enospc) {
+			error = xfs_flush_pages(ip, 0, -1, 0, FI_NONE);
+			if (error)
+				goto out_unlock_internal;
+			enospc = 1;
+			goto write_retry;
+		}
+		ret = ret2;
+	}
+
+	current->backing_dev_info = NULL;
+
+	isize = i_size_read(inode);
+	if (unlikely(ret < 0 && ret != -EFAULT && iocb->ki_pos > isize))
+		iocb->ki_pos = isize;
+
+	if (iocb->ki_pos > ip->i_size) {
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+		if (iocb->ki_pos > ip->i_size)
+			ip->i_size = iocb->ki_pos;
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	}
+
+	if (ret == -ENOSPC &&
+	    DM_EVENT_ENABLED(ip, DM_EVENT_NOSPACE) && !(ioflags & IO_INVIS)) {
+		xfs_iunlock(ip, iolock);
+		if (need_i_mutex)
+			mutex_unlock(&inode->i_mutex);
+		error = XFS_SEND_NAMESP(ip->i_mount, DM_EVENT_NOSPACE, ip,
+				DM_RIGHT_NULL, ip, DM_RIGHT_NULL, NULL, NULL,
+				0, 0, 0); /* Delay flag intentionally  unused */
+		if (need_i_mutex)
+			mutex_lock(&inode->i_mutex);
+		xfs_ilock(ip, iolock);
+		if (error)
+			goto out_unlock_internal;
+		goto start;
+	}
+
+	error = -ret;
+	if (ret <= 0)
+		goto out_unlock_internal;
+
+	XFS_STATS_ADD(xs_write_bytes, ret);
+
+	/* Handle various SYNC-type writes */
+	if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) {
+		loff_t end = pos + ret - 1;
+		int error2;
+
+		xfs_iunlock(ip, iolock);
+		if (need_i_mutex)
+			mutex_unlock(&inode->i_mutex);
+
+		error2 = filemap_write_and_wait_range(mapping, pos, end);
+		if (!error)
+			error = error2;
+		if (need_i_mutex)
+			mutex_lock(&inode->i_mutex);
+		xfs_ilock(ip, iolock);
+
+		error2 = -xfs_file_fsync(file, file->f_path.dentry,
+					 (file->f_flags & __O_SYNC) ? 0 : 1);
+		if (!error)
+			error = error2;
+	}
+
+ out_unlock_internal:
+	if (ip->i_new_size) {
+		xfs_ilock(ip, XFS_ILOCK_EXCL);
+		ip->i_new_size = 0;
+		/*
+		 * If this was a direct or synchronous I/O that failed (such
+		 * as ENOSPC) then part of the I/O may have been written to
+		 * disk before the error occured.  In this case the on-disk
+		 * file size may have been adjusted beyond the in-memory file
+		 * size and now needs to be truncated back.
+		 */
+		if (ip->i_d.di_size > ip->i_size)
+			ip->i_d.di_size = ip->i_size;
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	}
+	xfs_iunlock(ip, iolock);
+ out_unlock_mutex:
+	if (need_i_mutex)
+		mutex_unlock(&inode->i_mutex);
+	return -error;
 }
 
 STATIC int
@@ -160,28 +938,6 @@ xfs_file_release(
 	return -xfs_release(XFS_I(inode));
 }
 
-/*
- * We ignore the datasync flag here because a datasync is effectively
- * identical to an fsync. That is, datasync implies that we need to write
- * only the metadata needed to be able to access the data that is written
- * if we crash after the call completes. Hence if we are writing beyond
- * EOF we have to log the inode size change as well, which makes it a
- * full fsync. If we don't write beyond EOF, the inode core will be
- * clean in memory and so we don't need to log the inode, just like
- * fsync.
- */
-STATIC int
-xfs_file_fsync(
-	struct file		*file,
-	struct dentry		*dentry,
-	int			datasync)
-{
-	struct xfs_inode	*ip = XFS_I(dentry->d_inode);
-
-	xfs_iflags_clear(ip, XFS_ITRUNCATED);
-	return -xfs_fsync(ip);
-}
-
 STATIC int
 xfs_file_readdir(
 	struct file	*filp,
@@ -203,9 +959,9 @@ xfs_file_readdir(
 	 *
 	 * Try to give it an estimate that's good enough, maybe at some
 	 * point we can change the ->readdir prototype to include the
-	 * buffer size.
+	 * buffer size.  For now we use the current glibc buffer size.
 	 */
-	bufsize = (size_t)min_t(loff_t, PAGE_SIZE, ip->i_d.di_size);
+	bufsize = (size_t)min_t(loff_t, 32768, ip->i_d.di_size);
 
 	error = xfs_readdir(ip, dirent, bufsize,
 				(xfs_off_t *)&filp->f_pos, filldir);
diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.c b/fs/xfs/linux-2.6/xfs_fs_subr.c
index 7501b85fd860..b6918d76bc7b 100644
--- a/fs/xfs/linux-2.6/xfs_fs_subr.c
+++ b/fs/xfs/linux-2.6/xfs_fs_subr.c
@@ -79,7 +79,7 @@ xfs_flush_pages(
 		xfs_iflags_clear(ip, XFS_ITRUNCATED);
 		ret = -filemap_fdatawrite(mapping);
 	}
-	if (flags & XFS_B_ASYNC)
+	if (flags & XBF_ASYNC)
 		return ret;
 	ret2 = xfs_wait_on_pages(ip, first, last);
 	if (!ret)
diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c
index a034cf624437..4ea1ee18aded 100644
--- a/fs/xfs/linux-2.6/xfs_ioctl.c
+++ b/fs/xfs/linux-2.6/xfs_ioctl.c
@@ -447,12 +447,12 @@ xfs_attrlist_by_handle(
 int
 xfs_attrmulti_attr_get(
 	struct inode		*inode,
-	char			*name,
-	char			__user *ubuf,
+	unsigned char		*name,
+	unsigned char		__user *ubuf,
 	__uint32_t		*len,
 	__uint32_t		flags)
 {
-	char			*kbuf;
+	unsigned char		*kbuf;
 	int			error = EFAULT;
 
 	if (*len > XATTR_SIZE_MAX)
@@ -476,12 +476,12 @@ xfs_attrmulti_attr_get(
 int
 xfs_attrmulti_attr_set(
 	struct inode		*inode,
-	char			*name,
-	const char		__user *ubuf,
+	unsigned char		*name,
+	const unsigned char	__user *ubuf,
 	__uint32_t		len,
 	__uint32_t		flags)
 {
-	char			*kbuf;
+	unsigned char		*kbuf;
 	int			error = EFAULT;
 
 	if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
@@ -501,7 +501,7 @@ xfs_attrmulti_attr_set(
 int
 xfs_attrmulti_attr_remove(
 	struct inode		*inode,
-	char			*name,
+	unsigned char		*name,
 	__uint32_t		flags)
 {
 	if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
@@ -519,7 +519,7 @@ xfs_attrmulti_by_handle(
 	xfs_fsop_attrmulti_handlereq_t am_hreq;
 	struct dentry		*dentry;
 	unsigned int		i, size;
-	char			*attr_name;
+	unsigned char		*attr_name;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -XFS_ERROR(EPERM);
@@ -547,7 +547,7 @@ xfs_attrmulti_by_handle(
 
 	error = 0;
 	for (i = 0; i < am_hreq.opcount; i++) {
-		ops[i].am_error = strncpy_from_user(attr_name,
+		ops[i].am_error = strncpy_from_user((char *)attr_name,
 				ops[i].am_attrname, MAXNAMELEN);
 		if (ops[i].am_error == 0 || ops[i].am_error == MAXNAMELEN)
 			error = -ERANGE;
@@ -1431,6 +1431,9 @@ xfs_file_ioctl(
 		if (!capable(CAP_SYS_ADMIN))
 			return -EPERM;
 
+		if (mp->m_flags & XFS_MOUNT_RDONLY)
+			return -XFS_ERROR(EROFS);
+
 		if (copy_from_user(&inout, arg, sizeof(inout)))
 			return -XFS_ERROR(EFAULT);
 
diff --git a/fs/xfs/linux-2.6/xfs_ioctl.h b/fs/xfs/linux-2.6/xfs_ioctl.h
index 7bd7c6afc1eb..d56173b34a2a 100644
--- a/fs/xfs/linux-2.6/xfs_ioctl.h
+++ b/fs/xfs/linux-2.6/xfs_ioctl.h
@@ -45,23 +45,23 @@ xfs_readlink_by_handle(
 extern int
 xfs_attrmulti_attr_get(
 	struct inode		*inode,
-	char			*name,
-	char			__user *ubuf,
+	unsigned char		*name,
+	unsigned char		__user *ubuf,
 	__uint32_t		*len,
 	__uint32_t		flags);
 
 extern int
-	xfs_attrmulti_attr_set(
+xfs_attrmulti_attr_set(
 	struct inode		*inode,
-	char			*name,
-	const char		__user *ubuf,
+	unsigned char		*name,
+	const unsigned char	__user *ubuf,
 	__uint32_t		len,
 	__uint32_t		flags);
 
 extern int
 xfs_attrmulti_attr_remove(
 	struct inode		*inode,
-	char			*name,
+	unsigned char		*name,
 	__uint32_t		flags);
 
 extern struct dentry *
diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c
index be1527b1670c..0bf6d61f0528 100644
--- a/fs/xfs/linux-2.6/xfs_ioctl32.c
+++ b/fs/xfs/linux-2.6/xfs_ioctl32.c
@@ -411,7 +411,7 @@ xfs_compat_attrmulti_by_handle(
 	compat_xfs_fsop_attrmulti_handlereq_t	am_hreq;
 	struct dentry				*dentry;
 	unsigned int				i, size;
-	char					*attr_name;
+	unsigned char				*attr_name;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -XFS_ERROR(EPERM);
@@ -440,7 +440,7 @@ xfs_compat_attrmulti_by_handle(
 
 	error = 0;
 	for (i = 0; i < am_hreq.opcount; i++) {
-		ops[i].am_error = strncpy_from_user(attr_name,
+		ops[i].am_error = strncpy_from_user((char *)attr_name,
 				compat_ptr(ops[i].am_attrname),
 				MAXNAMELEN);
 		if (ops[i].am_error == 0 || ops[i].am_error == MAXNAMELEN)
diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c
index 225946012d0b..61a99608731e 100644
--- a/fs/xfs/linux-2.6/xfs_iops.c
+++ b/fs/xfs/linux-2.6/xfs_iops.c
@@ -91,6 +91,16 @@ xfs_mark_inode_dirty_sync(
 		mark_inode_dirty_sync(inode);
 }
 
+void
+xfs_mark_inode_dirty(
+	xfs_inode_t	*ip)
+{
+	struct inode	*inode = VFS_I(ip);
+
+	if (!(inode->i_state & (I_WILL_FREE|I_FREEING|I_CLEAR)))
+		mark_inode_dirty(inode);
+}
+
 /*
  * Change the requested timestamp in the given inode.
  * We don't lock across timestamp updates, and we don't log them but
@@ -140,10 +150,10 @@ xfs_init_security(
 	struct xfs_inode *ip = XFS_I(inode);
 	size_t		length;
 	void		*value;
-	char		*name;
+	unsigned char	*name;
 	int		error;
 
-	error = security_inode_init_security(inode, dir, &name,
+	error = security_inode_init_security(inode, dir, (char **)&name,
 					     &value, &length);
 	if (error) {
 		if (error == -EOPNOTSUPP)
diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h
index 5af0c81ca1ae..facfb323a706 100644
--- a/fs/xfs/linux-2.6/xfs_linux.h
+++ b/fs/xfs/linux-2.6/xfs_linux.h
@@ -88,7 +88,6 @@
 #include <xfs_super.h>
 #include <xfs_globals.h>
 #include <xfs_fs_subr.h>
-#include <xfs_lrw.h>
 #include <xfs_buf.h>
 
 /*
diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c
deleted file mode 100644
index 0d32457abef1..000000000000
--- a/fs/xfs/linux-2.6/xfs_lrw.c
+++ /dev/null
@@ -1,852 +0,0 @@
-/*
- * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc.
- * All Rights Reserved.
- *
- * 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.
- *
- * This program is distributed in the hope that it would 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 the Free Software Foundation,
- * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
- */
-#include "xfs.h"
-#include "xfs_fs.h"
-#include "xfs_bit.h"
-#include "xfs_log.h"
-#include "xfs_inum.h"
-#include "xfs_trans.h"
-#include "xfs_sb.h"
-#include "xfs_ag.h"
-#include "xfs_dir2.h"
-#include "xfs_alloc.h"
-#include "xfs_dmapi.h"
-#include "xfs_quota.h"
-#include "xfs_mount.h"
-#include "xfs_bmap_btree.h"
-#include "xfs_alloc_btree.h"
-#include "xfs_ialloc_btree.h"
-#include "xfs_dir2_sf.h"
-#include "xfs_attr_sf.h"
-#include "xfs_dinode.h"
-#include "xfs_inode.h"
-#include "xfs_bmap.h"
-#include "xfs_btree.h"
-#include "xfs_ialloc.h"
-#include "xfs_rtalloc.h"
-#include "xfs_error.h"
-#include "xfs_itable.h"
-#include "xfs_rw.h"
-#include "xfs_attr.h"
-#include "xfs_inode_item.h"
-#include "xfs_buf_item.h"
-#include "xfs_utils.h"
-#include "xfs_iomap.h"
-#include "xfs_vnodeops.h"
-#include "xfs_trace.h"
-
-#include <linux/capability.h>
-#include <linux/writeback.h>
-
-
-/*
- *	xfs_iozero
- *
- *	xfs_iozero clears the specified range of buffer supplied,
- *	and marks all the affected blocks as valid and modified.  If
- *	an affected block is not allocated, it will be allocated.  If
- *	an affected block is not completely overwritten, and is not
- *	valid before the operation, it will be read from disk before
- *	being partially zeroed.
- */
-STATIC int
-xfs_iozero(
-	struct xfs_inode	*ip,	/* inode			*/
-	loff_t			pos,	/* offset in file		*/
-	size_t			count)	/* size of data to zero		*/
-{
-	struct page		*page;
-	struct address_space	*mapping;
-	int			status;
-
-	mapping = VFS_I(ip)->i_mapping;
-	do {
-		unsigned offset, bytes;
-		void *fsdata;
-
-		offset = (pos & (PAGE_CACHE_SIZE -1)); /* Within page */
-		bytes = PAGE_CACHE_SIZE - offset;
-		if (bytes > count)
-			bytes = count;
-
-		status = pagecache_write_begin(NULL, mapping, pos, bytes,
-					AOP_FLAG_UNINTERRUPTIBLE,
-					&page, &fsdata);
-		if (status)
-			break;
-
-		zero_user(page, offset, bytes);
-
-		status = pagecache_write_end(NULL, mapping, pos, bytes, bytes,
-					page, fsdata);
-		WARN_ON(status <= 0); /* can't return less than zero! */
-		pos += bytes;
-		count -= bytes;
-		status = 0;
-	} while (count);
-
-	return (-status);
-}
-
-ssize_t			/* bytes read, or (-)  error */
-xfs_read(
-	xfs_inode_t		*ip,
-	struct kiocb		*iocb,
-	const struct iovec	*iovp,
-	unsigned int		segs,
-	loff_t			*offset,
-	int			ioflags)
-{
-	struct file		*file = iocb->ki_filp;
-	struct inode		*inode = file->f_mapping->host;
-	xfs_mount_t		*mp = ip->i_mount;
-	size_t			size = 0;
-	ssize_t			ret = 0;
-	xfs_fsize_t		n;
-	unsigned long		seg;
-
-
-	XFS_STATS_INC(xs_read_calls);
-
-	/* START copy & waste from filemap.c */
-	for (seg = 0; seg < segs; seg++) {
-		const struct iovec *iv = &iovp[seg];
-
-		/*
-		 * If any segment has a negative length, or the cumulative
-		 * length ever wraps negative then return -EINVAL.
-		 */
-		size += iv->iov_len;
-		if (unlikely((ssize_t)(size|iv->iov_len) < 0))
-			return XFS_ERROR(-EINVAL);
-	}
-	/* END copy & waste from filemap.c */
-
-	if (unlikely(ioflags & IO_ISDIRECT)) {
-		xfs_buftarg_t	*target =
-			XFS_IS_REALTIME_INODE(ip) ?
-				mp->m_rtdev_targp : mp->m_ddev_targp;
-		if ((*offset & target->bt_smask) ||
-		    (size & target->bt_smask)) {
-			if (*offset == ip->i_size) {
-				return (0);
-			}
-			return -XFS_ERROR(EINVAL);
-		}
-	}
-
-	n = XFS_MAXIOFFSET(mp) - *offset;
-	if ((n <= 0) || (size == 0))
-		return 0;
-
-	if (n < size)
-		size = n;
-
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return -EIO;
-
-	if (unlikely(ioflags & IO_ISDIRECT))
-		mutex_lock(&inode->i_mutex);
-	xfs_ilock(ip, XFS_IOLOCK_SHARED);
-
-	if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && !(ioflags & IO_INVIS)) {
-		int dmflags = FILP_DELAY_FLAG(file) | DM_SEM_FLAG_RD(ioflags);
-		int iolock = XFS_IOLOCK_SHARED;
-
-		ret = -XFS_SEND_DATA(mp, DM_EVENT_READ, ip, *offset, size,
-					dmflags, &iolock);
-		if (ret) {
-			xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-			if (unlikely(ioflags & IO_ISDIRECT))
-				mutex_unlock(&inode->i_mutex);
-			return ret;
-		}
-	}
-
-	if (unlikely(ioflags & IO_ISDIRECT)) {
-		if (inode->i_mapping->nrpages)
-			ret = -xfs_flushinval_pages(ip, (*offset & PAGE_CACHE_MASK),
-						    -1, FI_REMAPF_LOCKED);
-		mutex_unlock(&inode->i_mutex);
-		if (ret) {
-			xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-			return ret;
-		}
-	}
-
-	trace_xfs_file_read(ip, size, *offset, ioflags);
-
-	iocb->ki_pos = *offset;
-	ret = generic_file_aio_read(iocb, iovp, segs, *offset);
-	if (ret > 0)
-		XFS_STATS_ADD(xs_read_bytes, ret);
-
-	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-	return ret;
-}
-
-ssize_t
-xfs_splice_read(
-	xfs_inode_t		*ip,
-	struct file		*infilp,
-	loff_t			*ppos,
-	struct pipe_inode_info	*pipe,
-	size_t			count,
-	int			flags,
-	int			ioflags)
-{
-	xfs_mount_t		*mp = ip->i_mount;
-	ssize_t			ret;
-
-	XFS_STATS_INC(xs_read_calls);
-	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
-		return -EIO;
-
-	xfs_ilock(ip, XFS_IOLOCK_SHARED);
-
-	if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && !(ioflags & IO_INVIS)) {
-		int iolock = XFS_IOLOCK_SHARED;
-		int error;
-
-		error = XFS_SEND_DATA(mp, DM_EVENT_READ, ip, *ppos, count,
-					FILP_DELAY_FLAG(infilp), &iolock);
-		if (error) {
-			xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-			return -error;
-		}
-	}
-
-	trace_xfs_file_splice_read(ip, count, *ppos, ioflags);
-
-	ret = generic_file_splice_read(infilp, ppos, pipe, count, flags);
-	if (ret > 0)
-		XFS_STATS_ADD(xs_read_bytes, ret);
-
-	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-	return ret;
-}
-
-ssize_t
-xfs_splice_write(
-	xfs_inode_t		*ip,
-	struct pipe_inode_info	*pipe,
-	struct file		*outfilp,
-	loff_t			*ppos,
-	size_t			count,
-	int			flags,
-	int			ioflags)
-{
-	xfs_mount_t		*mp = ip->i_mount;
-	ssize_t			ret;
-	struct inode		*inode = outfilp->f_mapping->host;
-	xfs_fsize_t		isize, new_size;
-
-	XFS_STATS_INC(xs_write_calls);
-	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
-		return -EIO;
-
-	xfs_ilock(ip, XFS_IOLOCK_EXCL);
-
-	if (DM_EVENT_ENABLED(ip, DM_EVENT_WRITE) && !(ioflags & IO_INVIS)) {
-		int iolock = XFS_IOLOCK_EXCL;
-		int error;
-
-		error = XFS_SEND_DATA(mp, DM_EVENT_WRITE, ip, *ppos, count,
-					FILP_DELAY_FLAG(outfilp), &iolock);
-		if (error) {
-			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
-			return -error;
-		}
-	}
-
-	new_size = *ppos + count;
-
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	if (new_size > ip->i_size)
-		ip->i_new_size = new_size;
-	xfs_iunlock(ip, XFS_ILOCK_EXCL);
-
-	trace_xfs_file_splice_write(ip, count, *ppos, ioflags);
-
-	ret = generic_file_splice_write(pipe, outfilp, ppos, count, flags);
-	if (ret > 0)
-		XFS_STATS_ADD(xs_write_bytes, ret);
-
-	isize = i_size_read(inode);
-	if (unlikely(ret < 0 && ret != -EFAULT && *ppos > isize))
-		*ppos = isize;
-
-	if (*ppos > ip->i_size) {
-		xfs_ilock(ip, XFS_ILOCK_EXCL);
-		if (*ppos > ip->i_size)
-			ip->i_size = *ppos;
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-	}
-
-	if (ip->i_new_size) {
-		xfs_ilock(ip, XFS_ILOCK_EXCL);
-		ip->i_new_size = 0;
-		if (ip->i_d.di_size > ip->i_size)
-			ip->i_d.di_size = ip->i_size;
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-	}
-	xfs_iunlock(ip, XFS_IOLOCK_EXCL);
-	return ret;
-}
-
-/*
- * This routine is called to handle zeroing any space in the last
- * block of the file that is beyond the EOF.  We do this since the
- * size is being increased without writing anything to that block
- * and we don't want anyone to read the garbage on the disk.
- */
-STATIC int				/* error (positive) */
-xfs_zero_last_block(
-	xfs_inode_t	*ip,
-	xfs_fsize_t	offset,
-	xfs_fsize_t	isize)
-{
-	xfs_fileoff_t	last_fsb;
-	xfs_mount_t	*mp = ip->i_mount;
-	int		nimaps;
-	int		zero_offset;
-	int		zero_len;
-	int		error = 0;
-	xfs_bmbt_irec_t	imap;
-
-	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
-
-	zero_offset = XFS_B_FSB_OFFSET(mp, isize);
-	if (zero_offset == 0) {
-		/*
-		 * There are no extra bytes in the last block on disk to
-		 * zero, so return.
-		 */
-		return 0;
-	}
-
-	last_fsb = XFS_B_TO_FSBT(mp, isize);
-	nimaps = 1;
-	error = xfs_bmapi(NULL, ip, last_fsb, 1, 0, NULL, 0, &imap,
-			  &nimaps, NULL, NULL);
-	if (error) {
-		return error;
-	}
-	ASSERT(nimaps > 0);
-	/*
-	 * If the block underlying isize is just a hole, then there
-	 * is nothing to zero.
-	 */
-	if (imap.br_startblock == HOLESTARTBLOCK) {
-		return 0;
-	}
-	/*
-	 * Zero the part of the last block beyond the EOF, and write it
-	 * out sync.  We need to drop the ilock while we do this so we
-	 * don't deadlock when the buffer cache calls back to us.
-	 */
-	xfs_iunlock(ip, XFS_ILOCK_EXCL);
-
-	zero_len = mp->m_sb.sb_blocksize - zero_offset;
-	if (isize + zero_len > offset)
-		zero_len = offset - isize;
-	error = xfs_iozero(ip, isize, zero_len);
-
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	ASSERT(error >= 0);
-	return error;
-}
-
-/*
- * Zero any on disk space between the current EOF and the new,
- * larger EOF.  This handles the normal case of zeroing the remainder
- * of the last block in the file and the unusual case of zeroing blocks
- * out beyond the size of the file.  This second case only happens
- * with fixed size extents and when the system crashes before the inode
- * size was updated but after blocks were allocated.  If fill is set,
- * then any holes in the range are filled and zeroed.  If not, the holes
- * are left alone as holes.
- */
-
-int					/* error (positive) */
-xfs_zero_eof(
-	xfs_inode_t	*ip,
-	xfs_off_t	offset,		/* starting I/O offset */
-	xfs_fsize_t	isize)		/* current inode size */
-{
-	xfs_mount_t	*mp = ip->i_mount;
-	xfs_fileoff_t	start_zero_fsb;
-	xfs_fileoff_t	end_zero_fsb;
-	xfs_fileoff_t	zero_count_fsb;
-	xfs_fileoff_t	last_fsb;
-	xfs_fileoff_t	zero_off;
-	xfs_fsize_t	zero_len;
-	int		nimaps;
-	int		error = 0;
-	xfs_bmbt_irec_t	imap;
-
-	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
-	ASSERT(offset > isize);
-
-	/*
-	 * First handle zeroing the block on which isize resides.
-	 * We only zero a part of that block so it is handled specially.
-	 */
-	error = xfs_zero_last_block(ip, offset, isize);
-	if (error) {
-		ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
-		return error;
-	}
-
-	/*
-	 * Calculate the range between the new size and the old
-	 * where blocks needing to be zeroed may exist.  To get the
-	 * block where the last byte in the file currently resides,
-	 * we need to subtract one from the size and truncate back
-	 * to a block boundary.  We subtract 1 in case the size is
-	 * exactly on a block boundary.
-	 */
-	last_fsb = isize ? XFS_B_TO_FSBT(mp, isize - 1) : (xfs_fileoff_t)-1;
-	start_zero_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)isize);
-	end_zero_fsb = XFS_B_TO_FSBT(mp, offset - 1);
-	ASSERT((xfs_sfiloff_t)last_fsb < (xfs_sfiloff_t)start_zero_fsb);
-	if (last_fsb == end_zero_fsb) {
-		/*
-		 * The size was only incremented on its last block.
-		 * We took care of that above, so just return.
-		 */
-		return 0;
-	}
-
-	ASSERT(start_zero_fsb <= end_zero_fsb);
-	while (start_zero_fsb <= end_zero_fsb) {
-		nimaps = 1;
-		zero_count_fsb = end_zero_fsb - start_zero_fsb + 1;
-		error = xfs_bmapi(NULL, ip, start_zero_fsb, zero_count_fsb,
-				  0, NULL, 0, &imap, &nimaps, NULL, NULL);
-		if (error) {
-			ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
-			return error;
-		}
-		ASSERT(nimaps > 0);
-
-		if (imap.br_state == XFS_EXT_UNWRITTEN ||
-		    imap.br_startblock == HOLESTARTBLOCK) {
-			/*
-			 * This loop handles initializing pages that were
-			 * partially initialized by the code below this
-			 * loop. It basically zeroes the part of the page
-			 * that sits on a hole and sets the page as P_HOLE
-			 * and calls remapf if it is a mapped file.
-			 */
-			start_zero_fsb = imap.br_startoff + imap.br_blockcount;
-			ASSERT(start_zero_fsb <= (end_zero_fsb + 1));
-			continue;
-		}
-
-		/*
-		 * There are blocks we need to zero.
-		 * Drop the inode lock while we're doing the I/O.
-		 * We'll still have the iolock to protect us.
-		 */
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-
-		zero_off = XFS_FSB_TO_B(mp, start_zero_fsb);
-		zero_len = XFS_FSB_TO_B(mp, imap.br_blockcount);
-
-		if ((zero_off + zero_len) > offset)
-			zero_len = offset - zero_off;
-
-		error = xfs_iozero(ip, zero_off, zero_len);
-		if (error) {
-			goto out_lock;
-		}
-
-		start_zero_fsb = imap.br_startoff + imap.br_blockcount;
-		ASSERT(start_zero_fsb <= (end_zero_fsb + 1));
-
-		xfs_ilock(ip, XFS_ILOCK_EXCL);
-	}
-
-	return 0;
-
-out_lock:
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	ASSERT(error >= 0);
-	return error;
-}
-
-ssize_t				/* bytes written, or (-) error */
-xfs_write(
-	struct xfs_inode	*xip,
-	struct kiocb		*iocb,
-	const struct iovec	*iovp,
-	unsigned int		nsegs,
-	loff_t			*offset,
-	int			ioflags)
-{
-	struct file		*file = iocb->ki_filp;
-	struct address_space	*mapping = file->f_mapping;
-	struct inode		*inode = mapping->host;
-	unsigned long		segs = nsegs;
-	xfs_mount_t		*mp;
-	ssize_t			ret = 0, error = 0;
-	xfs_fsize_t		isize, new_size;
-	int			iolock;
-	int			eventsent = 0;
-	size_t			ocount = 0, count;
-	loff_t			pos;
-	int			need_i_mutex;
-
-	XFS_STATS_INC(xs_write_calls);
-
-	error = generic_segment_checks(iovp, &segs, &ocount, VERIFY_READ);
-	if (error)
-		return error;
-
-	count = ocount;
-	pos = *offset;
-
-	if (count == 0)
-		return 0;
-
-	mp = xip->i_mount;
-
-	xfs_wait_for_freeze(mp, SB_FREEZE_WRITE);
-
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return -EIO;
-
-relock:
-	if (ioflags & IO_ISDIRECT) {
-		iolock = XFS_IOLOCK_SHARED;
-		need_i_mutex = 0;
-	} else {
-		iolock = XFS_IOLOCK_EXCL;
-		need_i_mutex = 1;
-		mutex_lock(&inode->i_mutex);
-	}
-
-	xfs_ilock(xip, XFS_ILOCK_EXCL|iolock);
-
-start:
-	error = -generic_write_checks(file, &pos, &count,
-					S_ISBLK(inode->i_mode));
-	if (error) {
-		xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
-		goto out_unlock_mutex;
-	}
-
-	if ((DM_EVENT_ENABLED(xip, DM_EVENT_WRITE) &&
-	    !(ioflags & IO_INVIS) && !eventsent)) {
-		int		dmflags = FILP_DELAY_FLAG(file);
-
-		if (need_i_mutex)
-			dmflags |= DM_FLAGS_IMUX;
-
-		xfs_iunlock(xip, XFS_ILOCK_EXCL);
-		error = XFS_SEND_DATA(xip->i_mount, DM_EVENT_WRITE, xip,
-				      pos, count, dmflags, &iolock);
-		if (error) {
-			goto out_unlock_internal;
-		}
-		xfs_ilock(xip, XFS_ILOCK_EXCL);
-		eventsent = 1;
-
-		/*
-		 * The iolock was dropped and reacquired in XFS_SEND_DATA
-		 * so we have to recheck the size when appending.
-		 * We will only "goto start;" once, since having sent the
-		 * event prevents another call to XFS_SEND_DATA, which is
-		 * what allows the size to change in the first place.
-		 */
-		if ((file->f_flags & O_APPEND) && pos != xip->i_size)
-			goto start;
-	}
-
-	if (ioflags & IO_ISDIRECT) {
-		xfs_buftarg_t	*target =
-			XFS_IS_REALTIME_INODE(xip) ?
-				mp->m_rtdev_targp : mp->m_ddev_targp;
-
-		if ((pos & target->bt_smask) || (count & target->bt_smask)) {
-			xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
-			return XFS_ERROR(-EINVAL);
-		}
-
-		if (!need_i_mutex && (mapping->nrpages || pos > xip->i_size)) {
-			xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
-			iolock = XFS_IOLOCK_EXCL;
-			need_i_mutex = 1;
-			mutex_lock(&inode->i_mutex);
-			xfs_ilock(xip, XFS_ILOCK_EXCL|iolock);
-			goto start;
-		}
-	}
-
-	new_size = pos + count;
-	if (new_size > xip->i_size)
-		xip->i_new_size = new_size;
-
-	if (likely(!(ioflags & IO_INVIS)))
-		file_update_time(file);
-
-	/*
-	 * If the offset is beyond the size of the file, we have a couple
-	 * of things to do. First, if there is already space allocated
-	 * we need to either create holes or zero the disk or ...
-	 *
-	 * If there is a page where the previous size lands, we need
-	 * to zero it out up to the new size.
-	 */
-
-	if (pos > xip->i_size) {
-		error = xfs_zero_eof(xip, pos, xip->i_size);
-		if (error) {
-			xfs_iunlock(xip, XFS_ILOCK_EXCL);
-			goto out_unlock_internal;
-		}
-	}
-	xfs_iunlock(xip, XFS_ILOCK_EXCL);
-
-	/*
-	 * If we're writing the file then make sure to clear the
-	 * setuid and setgid bits if the process is not being run
-	 * by root.  This keeps people from modifying setuid and
-	 * setgid binaries.
-	 */
-
-	if (((xip->i_d.di_mode & S_ISUID) ||
-	    ((xip->i_d.di_mode & (S_ISGID | S_IXGRP)) ==
-		(S_ISGID | S_IXGRP))) &&
-	     !capable(CAP_FSETID)) {
-		error = xfs_write_clear_setuid(xip);
-		if (likely(!error))
-			error = -file_remove_suid(file);
-		if (unlikely(error)) {
-			goto out_unlock_internal;
-		}
-	}
-
-	/* We can write back this queue in page reclaim */
-	current->backing_dev_info = mapping->backing_dev_info;
-
-	if ((ioflags & IO_ISDIRECT)) {
-		if (mapping->nrpages) {
-			WARN_ON(need_i_mutex == 0);
-			error = xfs_flushinval_pages(xip,
-					(pos & PAGE_CACHE_MASK),
-					-1, FI_REMAPF_LOCKED);
-			if (error)
-				goto out_unlock_internal;
-		}
-
-		if (need_i_mutex) {
-			/* demote the lock now the cached pages are gone */
-			xfs_ilock_demote(xip, XFS_IOLOCK_EXCL);
-			mutex_unlock(&inode->i_mutex);
-
-			iolock = XFS_IOLOCK_SHARED;
-			need_i_mutex = 0;
-		}
-
-		trace_xfs_file_direct_write(xip, count, *offset, ioflags);
-		ret = generic_file_direct_write(iocb, iovp,
-				&segs, pos, offset, count, ocount);
-
-		/*
-		 * direct-io write to a hole: fall through to buffered I/O
-		 * for completing the rest of the request.
-		 */
-		if (ret >= 0 && ret != count) {
-			XFS_STATS_ADD(xs_write_bytes, ret);
-
-			pos += ret;
-			count -= ret;
-
-			ioflags &= ~IO_ISDIRECT;
-			xfs_iunlock(xip, iolock);
-			goto relock;
-		}
-	} else {
-		int enospc = 0;
-		ssize_t ret2 = 0;
-
-write_retry:
-		trace_xfs_file_buffered_write(xip, count, *offset, ioflags);
-		ret2 = generic_file_buffered_write(iocb, iovp, segs,
-				pos, offset, count, ret);
-		/*
-		 * if we just got an ENOSPC, flush the inode now we
-		 * aren't holding any page locks and retry *once*
-		 */
-		if (ret2 == -ENOSPC && !enospc) {
-			error = xfs_flush_pages(xip, 0, -1, 0, FI_NONE);
-			if (error)
-				goto out_unlock_internal;
-			enospc = 1;
-			goto write_retry;
-		}
-		ret = ret2;
-	}
-
-	current->backing_dev_info = NULL;
-
-	isize = i_size_read(inode);
-	if (unlikely(ret < 0 && ret != -EFAULT && *offset > isize))
-		*offset = isize;
-
-	if (*offset > xip->i_size) {
-		xfs_ilock(xip, XFS_ILOCK_EXCL);
-		if (*offset > xip->i_size)
-			xip->i_size = *offset;
-		xfs_iunlock(xip, XFS_ILOCK_EXCL);
-	}
-
-	if (ret == -ENOSPC &&
-	    DM_EVENT_ENABLED(xip, DM_EVENT_NOSPACE) && !(ioflags & IO_INVIS)) {
-		xfs_iunlock(xip, iolock);
-		if (need_i_mutex)
-			mutex_unlock(&inode->i_mutex);
-		error = XFS_SEND_NAMESP(xip->i_mount, DM_EVENT_NOSPACE, xip,
-				DM_RIGHT_NULL, xip, DM_RIGHT_NULL, NULL, NULL,
-				0, 0, 0); /* Delay flag intentionally  unused */
-		if (need_i_mutex)
-			mutex_lock(&inode->i_mutex);
-		xfs_ilock(xip, iolock);
-		if (error)
-			goto out_unlock_internal;
-		goto start;
-	}
-
-	error = -ret;
-	if (ret <= 0)
-		goto out_unlock_internal;
-
-	XFS_STATS_ADD(xs_write_bytes, ret);
-
-	/* Handle various SYNC-type writes */
-	if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) {
-		loff_t end = pos + ret - 1;
-		int error2;
-
-		xfs_iunlock(xip, iolock);
-		if (need_i_mutex)
-			mutex_unlock(&inode->i_mutex);
-
-		error2 = filemap_write_and_wait_range(mapping, pos, end);
-		if (!error)
-			error = error2;
-		if (need_i_mutex)
-			mutex_lock(&inode->i_mutex);
-		xfs_ilock(xip, iolock);
-
-		error2 = xfs_fsync(xip);
-		if (!error)
-			error = error2;
-	}
-
- out_unlock_internal:
-	if (xip->i_new_size) {
-		xfs_ilock(xip, XFS_ILOCK_EXCL);
-		xip->i_new_size = 0;
-		/*
-		 * If this was a direct or synchronous I/O that failed (such
-		 * as ENOSPC) then part of the I/O may have been written to
-		 * disk before the error occured.  In this case the on-disk
-		 * file size may have been adjusted beyond the in-memory file
-		 * size and now needs to be truncated back.
-		 */
-		if (xip->i_d.di_size > xip->i_size)
-			xip->i_d.di_size = xip->i_size;
-		xfs_iunlock(xip, XFS_ILOCK_EXCL);
-	}
-	xfs_iunlock(xip, iolock);
- out_unlock_mutex:
-	if (need_i_mutex)
-		mutex_unlock(&inode->i_mutex);
-	return -error;
-}
-
-/*
- * All xfs metadata buffers except log state machine buffers
- * get this attached as their b_bdstrat callback function.
- * This is so that we can catch a buffer
- * after prematurely unpinning it to forcibly shutdown the filesystem.
- */
-int
-xfs_bdstrat_cb(struct xfs_buf *bp)
-{
-	if (XFS_FORCED_SHUTDOWN(bp->b_mount)) {
-		trace_xfs_bdstrat_shut(bp, _RET_IP_);
-		/*
-		 * Metadata write that didn't get logged but
-		 * written delayed anyway. These aren't associated
-		 * with a transaction, and can be ignored.
-		 */
-		if (XFS_BUF_IODONE_FUNC(bp) == NULL &&
-		    (XFS_BUF_ISREAD(bp)) == 0)
-			return (xfs_bioerror_relse(bp));
-		else
-			return (xfs_bioerror(bp));
-	}
-
-	xfs_buf_iorequest(bp);
-	return 0;
-}
-
-/*
- * Wrapper around bdstrat so that we can stop data from going to disk in case
- * we are shutting down the filesystem.  Typically user data goes thru this
- * path; one of the exceptions is the superblock.
- */
-void
-xfsbdstrat(
-	struct xfs_mount	*mp,
-	struct xfs_buf		*bp)
-{
-	ASSERT(mp);
-	if (!XFS_FORCED_SHUTDOWN(mp)) {
-		xfs_buf_iorequest(bp);
-		return;
-	}
-
-	trace_xfs_bdstrat_shut(bp, _RET_IP_);
-	xfs_bioerror_relse(bp);
-}
-
-/*
- * If the underlying (data/log/rt) device is readonly, there are some
- * operations that cannot proceed.
- */
-int
-xfs_dev_is_read_only(
-	xfs_mount_t		*mp,
-	char			*message)
-{
-	if (xfs_readonly_buftarg(mp->m_ddev_targp) ||
-	    xfs_readonly_buftarg(mp->m_logdev_targp) ||
-	    (mp->m_rtdev_targp && xfs_readonly_buftarg(mp->m_rtdev_targp))) {
-		cmn_err(CE_NOTE,
-			"XFS: %s required on read-only device.", message);
-		cmn_err(CE_NOTE,
-			"XFS: write access unavailable, cannot proceed.");
-		return EROFS;
-	}
-	return 0;
-}
diff --git a/fs/xfs/linux-2.6/xfs_lrw.h b/fs/xfs/linux-2.6/xfs_lrw.h
deleted file mode 100644
index d1f7789c7ffb..000000000000
--- a/fs/xfs/linux-2.6/xfs_lrw.h
+++ /dev/null
@@ -1,32 +0,0 @@
-/*
- * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc.
- * All Rights Reserved.
- *
- * 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.
- *
- * This program is distributed in the hope that it would 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 the Free Software Foundation,
- * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
- */
-#ifndef __XFS_LRW_H__
-#define __XFS_LRW_H__
-
-struct xfs_mount;
-struct xfs_inode;
-struct xfs_buf;
-
-/* errors from xfsbdstrat() must be extracted from the buffer */
-extern void xfsbdstrat(struct xfs_mount *, struct xfs_buf *);
-extern int xfs_bdstrat_cb(struct xfs_buf *);
-extern int xfs_dev_is_read_only(struct xfs_mount *, char *);
-
-extern int xfs_zero_eof(struct xfs_inode *, xfs_off_t, xfs_fsize_t);
-
-#endif	/* __XFS_LRW_H__ */
diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c
index 3d4a0c84d634..1947514ce1ad 100644
--- a/fs/xfs/linux-2.6/xfs_quotaops.c
+++ b/fs/xfs/linux-2.6/xfs_quotaops.c
@@ -44,20 +44,6 @@ xfs_quota_type(int type)
 }
 
 STATIC int
-xfs_fs_quota_sync(
-	struct super_block	*sb,
-	int			type)
-{
-	struct xfs_mount	*mp = XFS_M(sb);
-
-	if (sb->s_flags & MS_RDONLY)
-		return -EROFS;
-	if (!XFS_IS_QUOTA_RUNNING(mp))
-		return -ENOSYS;
-	return -xfs_sync_data(mp, 0);
-}
-
-STATIC int
 xfs_fs_get_xstate(
 	struct super_block	*sb,
 	struct fs_quota_stat	*fqs)
@@ -82,8 +68,6 @@ xfs_fs_set_xstate(
 		return -EROFS;
 	if (op != Q_XQUOTARM && !XFS_IS_QUOTA_RUNNING(mp))
 		return -ENOSYS;
-	if (!capable(CAP_SYS_ADMIN))
-		return -EPERM;
 
 	if (uflags & XFS_QUOTA_UDQ_ACCT)
 		flags |= XFS_UQUOTA_ACCT;
@@ -144,14 +128,11 @@ xfs_fs_set_xquota(
 		return -ENOSYS;
 	if (!XFS_IS_QUOTA_ON(mp))
 		return -ESRCH;
-	if (!capable(CAP_SYS_ADMIN))
-		return -EPERM;
 
 	return -xfs_qm_scall_setqlim(mp, id, xfs_quota_type(type), fdq);
 }
 
 const struct quotactl_ops xfs_quotactl_operations = {
-	.quota_sync		= xfs_fs_quota_sync,
 	.get_xstate		= xfs_fs_get_xstate,
 	.set_xstate		= xfs_fs_set_xstate,
 	.get_xquota		= xfs_fs_get_xquota,
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index 77414db10dc2..71345a370d9f 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -877,12 +877,11 @@ xfsaild(
 {
 	struct xfs_ail	*ailp = data;
 	xfs_lsn_t	last_pushed_lsn = 0;
-	long		tout = 0;
+	long		tout = 0; /* milliseconds */
 
 	while (!kthread_should_stop()) {
-		if (tout)
-			schedule_timeout_interruptible(msecs_to_jiffies(tout));
-		tout = 1000;
+		schedule_timeout_interruptible(tout ?
+				msecs_to_jiffies(tout) : MAX_SCHEDULE_TIMEOUT);
 
 		/* swsusp */
 		try_to_freeze();
@@ -1022,59 +1021,108 @@ xfs_fs_dirty_inode(
 	XFS_I(inode)->i_update_core = 1;
 }
 
-/*
- * Attempt to flush the inode, this will actually fail
- * if the inode is pinned, but we dirty the inode again
- * at the point when it is unpinned after a log write,
- * since this is when the inode itself becomes flushable.
- */
+STATIC int
+xfs_log_inode(
+	struct xfs_inode	*ip)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_trans	*tp;
+	int			error;
+
+	xfs_iunlock(ip, XFS_ILOCK_SHARED);
+	tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
+	error = xfs_trans_reserve(tp, 0, XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
+
+	if (error) {
+		xfs_trans_cancel(tp, 0);
+		/* we need to return with the lock hold shared */
+		xfs_ilock(ip, XFS_ILOCK_SHARED);
+		return error;
+	}
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+
+	/*
+	 * Note - it's possible that we might have pushed ourselves out of the
+	 * way during trans_reserve which would flush the inode.  But there's
+	 * no guarantee that the inode buffer has actually gone out yet (it's
+	 * delwri).  Plus the buffer could be pinned anyway if it's part of
+	 * an inode in another recent transaction.  So we play it safe and
+	 * fire off the transaction anyway.
+	 */
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_ihold(tp, ip);
+	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+	xfs_trans_set_sync(tp);
+	error = xfs_trans_commit(tp, 0);
+	xfs_ilock_demote(ip, XFS_ILOCK_EXCL);
+
+	return error;
+}
+
 STATIC int
 xfs_fs_write_inode(
 	struct inode		*inode,
-	int			sync)
+	struct writeback_control *wbc)
 {
 	struct xfs_inode	*ip = XFS_I(inode);
 	struct xfs_mount	*mp = ip->i_mount;
-	int			error = 0;
+	int			error = EAGAIN;
 
 	xfs_itrace_entry(ip);
 
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return XFS_ERROR(EIO);
 
-	if (sync) {
-		error = xfs_wait_on_pages(ip, 0, -1);
-		if (error)
+	if (wbc->sync_mode == WB_SYNC_ALL) {
+		/*
+		 * Make sure the inode has hit stable storage.  By using the
+		 * log and the fsync transactions we reduce the IOs we have
+		 * to do here from two (log and inode) to just the log.
+		 *
+		 * Note: We still need to do a delwri write of the inode after
+		 * this to flush it to the backing buffer so that bulkstat
+		 * works properly if this is the first time the inode has been
+		 * written.  Because we hold the ilock atomically over the
+		 * transaction commit and the inode flush we are guaranteed
+		 * that the inode is not pinned when it returns. If the flush
+		 * lock is already held, then the inode has already been
+		 * flushed once and we don't need to flush it again.  Hence
+		 * the code will only flush the inode if it isn't already
+		 * being flushed.
+		 */
+		xfs_ilock(ip, XFS_ILOCK_SHARED);
+		if (ip->i_update_core) {
+			error = xfs_log_inode(ip);
+			if (error)
+				goto out_unlock;
+		}
+	} else {
+		/*
+		 * We make this non-blocking if the inode is contended, return
+		 * EAGAIN to indicate to the caller that they did not succeed.
+		 * This prevents the flush path from blocking on inodes inside
+		 * another operation right now, they get caught later by xfs_sync.
+		 */
+		if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED))
 			goto out;
 	}
 
-	/*
-	 * Bypass inodes which have already been cleaned by
-	 * the inode flush clustering code inside xfs_iflush
-	 */
-	if (xfs_inode_clean(ip))
-		goto out;
+	if (xfs_ipincount(ip) || !xfs_iflock_nowait(ip))
+		goto out_unlock;
 
 	/*
-	 * We make this non-blocking if the inode is contended, return
-	 * EAGAIN to indicate to the caller that they did not succeed.
-	 * This prevents the flush path from blocking on inodes inside
-	 * another operation right now, they get caught later by xfs_sync.
+	 * Now we have the flush lock and the inode is not pinned, we can check
+	 * if the inode is really clean as we know that there are no pending
+	 * transaction completions, it is not waiting on the delayed write
+	 * queue and there is no IO in progress.
 	 */
-	if (sync) {
-		xfs_ilock(ip, XFS_ILOCK_SHARED);
-		xfs_iflock(ip);
-
-		error = xfs_iflush(ip, XFS_IFLUSH_SYNC);
-	} else {
-		error = EAGAIN;
-		if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED))
-			goto out;
-		if (xfs_ipincount(ip) || !xfs_iflock_nowait(ip))
-			goto out_unlock;
-
-		error = xfs_iflush(ip, XFS_IFLUSH_ASYNC_NOBLOCK);
+	if (xfs_inode_clean(ip)) {
+		xfs_ifunlock(ip);
+		error = 0;
+		goto out_unlock;
 	}
+	error = xfs_iflush(ip, 0);
 
  out_unlock:
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
@@ -1257,6 +1305,29 @@ xfs_fs_statfs(
 	return 0;
 }
 
+STATIC void
+xfs_save_resvblks(struct xfs_mount *mp)
+{
+	__uint64_t resblks = 0;
+
+	mp->m_resblks_save = mp->m_resblks;
+	xfs_reserve_blocks(mp, &resblks, NULL);
+}
+
+STATIC void
+xfs_restore_resvblks(struct xfs_mount *mp)
+{
+	__uint64_t resblks;
+
+	if (mp->m_resblks_save) {
+		resblks = mp->m_resblks_save;
+		mp->m_resblks_save = 0;
+	} else
+		resblks = xfs_default_resblks(mp);
+
+	xfs_reserve_blocks(mp, &resblks, NULL);
+}
+
 STATIC int
 xfs_fs_remount(
 	struct super_block	*sb,
@@ -1336,11 +1407,27 @@ xfs_fs_remount(
 			}
 			mp->m_update_flags = 0;
 		}
+
+		/*
+		 * Fill out the reserve pool if it is empty. Use the stashed
+		 * value if it is non-zero, otherwise go with the default.
+		 */
+		xfs_restore_resvblks(mp);
 	}
 
 	/* rw -> ro */
 	if (!(mp->m_flags & XFS_MOUNT_RDONLY) && (*flags & MS_RDONLY)) {
+		/*
+		 * After we have synced the data but before we sync the
+		 * metadata, we need to free up the reserve block pool so that
+		 * the used block count in the superblock on disk is correct at
+		 * the end of the remount. Stash the current reserve pool size
+		 * so that if we get remounted rw, we can return it to the same
+		 * size.
+		 */
+
 		xfs_quiesce_data(mp);
+		xfs_save_resvblks(mp);
 		xfs_quiesce_attr(mp);
 		mp->m_flags |= XFS_MOUNT_RDONLY;
 	}
@@ -1359,11 +1446,22 @@ xfs_fs_freeze(
 {
 	struct xfs_mount	*mp = XFS_M(sb);
 
+	xfs_save_resvblks(mp);
 	xfs_quiesce_attr(mp);
 	return -xfs_fs_log_dummy(mp);
 }
 
 STATIC int
+xfs_fs_unfreeze(
+	struct super_block	*sb)
+{
+	struct xfs_mount	*mp = XFS_M(sb);
+
+	xfs_restore_resvblks(mp);
+	return 0;
+}
+
+STATIC int
 xfs_fs_show_options(
 	struct seq_file		*m,
 	struct vfsmount		*mnt)
@@ -1585,6 +1683,7 @@ static const struct super_operations xfs_super_operations = {
 	.put_super		= xfs_fs_put_super,
 	.sync_fs		= xfs_fs_sync_fs,
 	.freeze_fs		= xfs_fs_freeze,
+	.unfreeze_fs		= xfs_fs_unfreeze,
 	.statfs			= xfs_fs_statfs,
 	.remount_fs		= xfs_fs_remount,
 	.show_options		= xfs_fs_show_options,
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index 1f5e4bb5e970..05cd85317f6f 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -90,14 +90,13 @@ xfs_inode_ag_lookup(
 STATIC int
 xfs_inode_ag_walk(
 	struct xfs_mount	*mp,
-	xfs_agnumber_t		ag,
+	struct xfs_perag	*pag,
 	int			(*execute)(struct xfs_inode *ip,
 					   struct xfs_perag *pag, int flags),
 	int			flags,
 	int			tag,
 	int			exclusive)
 {
-	struct xfs_perag	*pag = &mp->m_perag[ag];
 	uint32_t		first_index;
 	int			last_error = 0;
 	int			skipped;
@@ -141,8 +140,6 @@ restart:
 		delay(1);
 		goto restart;
 	}
-
-	xfs_put_perag(mp, pag);
 	return last_error;
 }
 
@@ -160,10 +157,16 @@ xfs_inode_ag_iterator(
 	xfs_agnumber_t		ag;
 
 	for (ag = 0; ag < mp->m_sb.sb_agcount; ag++) {
-		if (!mp->m_perag[ag].pag_ici_init)
+		struct xfs_perag	*pag;
+
+		pag = xfs_perag_get(mp, ag);
+		if (!pag->pag_ici_init) {
+			xfs_perag_put(pag);
 			continue;
-		error = xfs_inode_ag_walk(mp, ag, execute, flags, tag,
+		}
+		error = xfs_inode_ag_walk(mp, pag, execute, flags, tag,
 						exclusive);
+		xfs_perag_put(pag);
 		if (error) {
 			last_error = error;
 			if (error == EFSCORRUPTED)
@@ -231,7 +234,7 @@ xfs_sync_inode_data(
 	}
 
 	error = xfs_flush_pages(ip, 0, -1, (flags & SYNC_WAIT) ?
-				0 : XFS_B_ASYNC, FI_NONE);
+				0 : XBF_ASYNC, FI_NONE);
 	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
 
  out_wait:
@@ -267,8 +270,7 @@ xfs_sync_inode_attr(
 		goto out_unlock;
 	}
 
-	error = xfs_iflush(ip, (flags & SYNC_WAIT) ?
-			   XFS_IFLUSH_SYNC : XFS_IFLUSH_DELWRI);
+	error = xfs_iflush(ip, flags);
 
  out_unlock:
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
@@ -293,10 +295,7 @@ xfs_sync_data(
 	if (error)
 		return XFS_ERROR(error);
 
-	xfs_log_force(mp, 0,
-		      (flags & SYNC_WAIT) ?
-		       XFS_LOG_FORCE | XFS_LOG_SYNC :
-		       XFS_LOG_FORCE);
+	xfs_log_force(mp, (flags & SYNC_WAIT) ? XFS_LOG_SYNC : 0);
 	return 0;
 }
 
@@ -322,10 +321,6 @@ xfs_commit_dummy_trans(
 	struct xfs_inode	*ip = mp->m_rootip;
 	struct xfs_trans	*tp;
 	int			error;
-	int			log_flags = XFS_LOG_FORCE;
-
-	if (flags & SYNC_WAIT)
-		log_flags |= XFS_LOG_SYNC;
 
 	/*
 	 * Put a dummy transaction in the log to tell recovery
@@ -347,11 +342,11 @@ xfs_commit_dummy_trans(
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
 
 	/* the log force ensures this transaction is pushed to disk */
-	xfs_log_force(mp, 0, log_flags);
+	xfs_log_force(mp, (flags & SYNC_WAIT) ? XFS_LOG_SYNC : 0);
 	return error;
 }
 
-int
+STATIC int
 xfs_sync_fsdata(
 	struct xfs_mount	*mp,
 	int			flags)
@@ -367,7 +362,7 @@ xfs_sync_fsdata(
 	if (flags & SYNC_TRYLOCK) {
 		ASSERT(!(flags & SYNC_WAIT));
 
-		bp = xfs_getsb(mp, XFS_BUF_TRYLOCK);
+		bp = xfs_getsb(mp, XBF_TRYLOCK);
 		if (!bp)
 			goto out;
 
@@ -387,7 +382,7 @@ xfs_sync_fsdata(
 		 * become pinned in between there and here.
 		 */
 		if (XFS_BUF_ISPINNED(bp))
-			xfs_log_force(mp, 0, XFS_LOG_FORCE);
+			xfs_log_force(mp, 0);
 	}
 
 
@@ -448,9 +443,6 @@ xfs_quiesce_data(
 	xfs_sync_data(mp, SYNC_WAIT);
 	xfs_qm_sync(mp, SYNC_WAIT);
 
-	/* drop inode references pinned by filestreams */
-	xfs_filestream_flush(mp);
-
 	/* write superblock and hoover up shutdown errors */
 	error = xfs_sync_fsdata(mp, SYNC_WAIT);
 
@@ -467,16 +459,18 @@ xfs_quiesce_fs(
 {
 	int	count = 0, pincount;
 
+	xfs_reclaim_inodes(mp, 0);
 	xfs_flush_buftarg(mp->m_ddev_targp, 0);
-	xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
 
 	/*
 	 * This loop must run at least twice.  The first instance of the loop
 	 * will flush most meta data but that will generate more meta data
 	 * (typically directory updates).  Which then must be flushed and
-	 * logged before we can write the unmount record.
+	 * logged before we can write the unmount record. We also so sync
+	 * reclaim of inodes to catch any that the above delwri flush skipped.
 	 */
 	do {
+		xfs_reclaim_inodes(mp, SYNC_WAIT);
 		xfs_sync_attr(mp, SYNC_WAIT);
 		pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1);
 		if (!pincount) {
@@ -575,7 +569,7 @@ xfs_flush_inodes(
 	igrab(inode);
 	xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inodes_work, &completion);
 	wait_for_completion(&completion);
-	xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC);
+	xfs_log_force(ip->i_mount, XFS_LOG_SYNC);
 }
 
 /*
@@ -591,8 +585,8 @@ xfs_sync_worker(
 	int		error;
 
 	if (!(mp->m_flags & XFS_MOUNT_RDONLY)) {
-		xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
-		xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
+		xfs_log_force(mp, 0);
+		xfs_reclaim_inodes(mp, 0);
 		/* dgc: errors ignored here */
 		error = xfs_qm_sync(mp, SYNC_TRYLOCK);
 		error = xfs_sync_fsdata(mp, SYNC_TRYLOCK);
@@ -613,7 +607,8 @@ xfssyncd(
 	set_freezable();
 	timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10);
 	for (;;) {
-		timeleft = schedule_timeout_interruptible(timeleft);
+		if (list_empty(&mp->m_sync_list))
+			timeleft = schedule_timeout_interruptible(timeleft);
 		/* swsusp */
 		try_to_freeze();
 		if (kthread_should_stop() && list_empty(&mp->m_sync_list))
@@ -633,8 +628,7 @@ xfssyncd(
 			list_add_tail(&mp->m_sync_work.w_list,
 					&mp->m_sync_list);
 		}
-		list_for_each_entry_safe(work, n, &mp->m_sync_list, w_list)
-			list_move(&work->w_list, &tmp);
+		list_splice_init(&mp->m_sync_list, &tmp);
 		spin_unlock(&mp->m_sync_lock);
 
 		list_for_each_entry_safe(work, n, &tmp, w_list) {
@@ -690,16 +684,17 @@ void
 xfs_inode_set_reclaim_tag(
 	xfs_inode_t	*ip)
 {
-	xfs_mount_t	*mp = ip->i_mount;
-	xfs_perag_t	*pag = xfs_get_perag(mp, ip->i_ino);
+	struct xfs_mount *mp = ip->i_mount;
+	struct xfs_perag *pag;
 
-	read_lock(&pag->pag_ici_lock);
+	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
+	write_lock(&pag->pag_ici_lock);
 	spin_lock(&ip->i_flags_lock);
 	__xfs_inode_set_reclaim_tag(pag, ip);
 	__xfs_iflags_set(ip, XFS_IRECLAIMABLE);
 	spin_unlock(&ip->i_flags_lock);
-	read_unlock(&pag->pag_ici_lock);
-	xfs_put_perag(mp, pag);
+	write_unlock(&pag->pag_ici_lock);
+	xfs_perag_put(pag);
 }
 
 void
@@ -712,12 +707,64 @@ __xfs_inode_clear_reclaim_tag(
 			XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG);
 }
 
+/*
+ * Inodes in different states need to be treated differently, and the return
+ * value of xfs_iflush is not sufficient to get this right. The following table
+ * lists the inode states and the reclaim actions necessary for non-blocking
+ * reclaim:
+ *
+ *
+ *	inode state	     iflush ret		required action
+ *      ---------------      ----------         ---------------
+ *	bad			-		reclaim
+ *	shutdown		EIO		unpin and reclaim
+ *	clean, unpinned		0		reclaim
+ *	stale, unpinned		0		reclaim
+ *	clean, pinned(*)	0		requeue
+ *	stale, pinned		EAGAIN		requeue
+ *	dirty, delwri ok	0		requeue
+ *	dirty, delwri blocked	EAGAIN		requeue
+ *	dirty, sync flush	0		reclaim
+ *
+ * (*) dgc: I don't think the clean, pinned state is possible but it gets
+ * handled anyway given the order of checks implemented.
+ *
+ * As can be seen from the table, the return value of xfs_iflush() is not
+ * sufficient to correctly decide the reclaim action here. The checks in
+ * xfs_iflush() might look like duplicates, but they are not.
+ *
+ * Also, because we get the flush lock first, we know that any inode that has
+ * been flushed delwri has had the flush completed by the time we check that
+ * the inode is clean. The clean inode check needs to be done before flushing
+ * the inode delwri otherwise we would loop forever requeuing clean inodes as
+ * we cannot tell apart a successful delwri flush and a clean inode from the
+ * return value of xfs_iflush().
+ *
+ * Note that because the inode is flushed delayed write by background
+ * writeback, the flush lock may already be held here and waiting on it can
+ * result in very long latencies. Hence for sync reclaims, where we wait on the
+ * flush lock, the caller should push out delayed write inodes first before
+ * trying to reclaim them to minimise the amount of time spent waiting. For
+ * background relaim, we just requeue the inode for the next pass.
+ *
+ * Hence the order of actions after gaining the locks should be:
+ *	bad		=> reclaim
+ *	shutdown	=> unpin and reclaim
+ *	pinned, delwri	=> requeue
+ *	pinned, sync	=> unpin
+ *	stale		=> reclaim
+ *	clean		=> reclaim
+ *	dirty, delwri	=> flush and requeue
+ *	dirty, sync	=> flush, wait and reclaim
+ */
 STATIC int
 xfs_reclaim_inode(
 	struct xfs_inode	*ip,
 	struct xfs_perag	*pag,
 	int			sync_mode)
 {
+	int	error = 0;
+
 	/*
 	 * The radix tree lock here protects a thread in xfs_iget from racing
 	 * with us starting reclaim on the inode.  Once we have the
@@ -735,33 +782,70 @@ xfs_reclaim_inode(
 	spin_unlock(&ip->i_flags_lock);
 	write_unlock(&pag->pag_ici_lock);
 
-	/*
-	 * If the inode is still dirty, then flush it out.  If the inode
-	 * is not in the AIL, then it will be OK to flush it delwri as
-	 * long as xfs_iflush() does not keep any references to the inode.
-	 * We leave that decision up to xfs_iflush() since it has the
-	 * knowledge of whether it's OK to simply do a delwri flush of
-	 * the inode or whether we need to wait until the inode is
-	 * pulled from the AIL.
-	 * We get the flush lock regardless, though, just to make sure
-	 * we don't free it while it is being flushed.
-	 */
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_iflock(ip);
+	if (!xfs_iflock_nowait(ip)) {
+		if (!(sync_mode & SYNC_WAIT))
+			goto out;
+		xfs_iflock(ip);
+	}
+
+	if (is_bad_inode(VFS_I(ip)))
+		goto reclaim;
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+		xfs_iunpin_wait(ip);
+		goto reclaim;
+	}
+	if (xfs_ipincount(ip)) {
+		if (!(sync_mode & SYNC_WAIT)) {
+			xfs_ifunlock(ip);
+			goto out;
+		}
+		xfs_iunpin_wait(ip);
+	}
+	if (xfs_iflags_test(ip, XFS_ISTALE))
+		goto reclaim;
+	if (xfs_inode_clean(ip))
+		goto reclaim;
+
+	/* Now we have an inode that needs flushing */
+	error = xfs_iflush(ip, sync_mode);
+	if (sync_mode & SYNC_WAIT) {
+		xfs_iflock(ip);
+		goto reclaim;
+	}
 
 	/*
-	 * In the case of a forced shutdown we rely on xfs_iflush() to
-	 * wait for the inode to be unpinned before returning an error.
+	 * When we have to flush an inode but don't have SYNC_WAIT set, we
+	 * flush the inode out using a delwri buffer and wait for the next
+	 * call into reclaim to find it in a clean state instead of waiting for
+	 * it now. We also don't return errors here - if the error is transient
+	 * then the next reclaim pass will flush the inode, and if the error
+	 * is permanent then the next sync reclaim will relcaim the inode and
+	 * pass on the error.
 	 */
-	if (!is_bad_inode(VFS_I(ip)) && xfs_iflush(ip, sync_mode) == 0) {
-		/* synchronize with xfs_iflush_done */
-		xfs_iflock(ip);
-		xfs_ifunlock(ip);
+	if (error && !XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+		xfs_fs_cmn_err(CE_WARN, ip->i_mount,
+			"inode 0x%llx background reclaim flush failed with %d",
+			(long long)ip->i_ino, error);
 	}
+out:
+	xfs_iflags_clear(ip, XFS_IRECLAIM);
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	/*
+	 * We could return EAGAIN here to make reclaim rescan the inode tree in
+	 * a short while. However, this just burns CPU time scanning the tree
+	 * waiting for IO to complete and xfssyncd never goes back to the idle
+	 * state. Instead, return 0 to let the next scheduled background reclaim
+	 * attempt to reclaim the inode again.
+	 */
+	return 0;
 
+reclaim:
+	xfs_ifunlock(ip);
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
 	xfs_ireclaim(ip);
-	return 0;
+	return error;
+
 }
 
 int
diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h
index ea932b43335d..d480c346cabb 100644
--- a/fs/xfs/linux-2.6/xfs_sync.h
+++ b/fs/xfs/linux-2.6/xfs_sync.h
@@ -37,7 +37,6 @@ void xfs_syncd_stop(struct xfs_mount *mp);
 
 int xfs_sync_attr(struct xfs_mount *mp, int flags);
 int xfs_sync_data(struct xfs_mount *mp, int flags);
-int xfs_sync_fsdata(struct xfs_mount *mp, int flags);
 
 int xfs_quiesce_data(struct xfs_mount *mp);
 void xfs_quiesce_attr(struct xfs_mount *mp);
diff --git a/fs/xfs/linux-2.6/xfs_trace.c b/fs/xfs/linux-2.6/xfs_trace.c
index 856eb3c8d605..5a107601e969 100644
--- a/fs/xfs/linux-2.6/xfs_trace.c
+++ b/fs/xfs/linux-2.6/xfs_trace.c
@@ -52,22 +52,6 @@
 #include "quota/xfs_dquot.h"
 
 /*
- * Format fsblock number into a static buffer & return it.
- */
-STATIC char *xfs_fmtfsblock(xfs_fsblock_t bno)
-{
-	static char rval[50];
-
-	if (bno == NULLFSBLOCK)
-		sprintf(rval, "NULLFSBLOCK");
-	else if (isnullstartblock(bno))
-		sprintf(rval, "NULLSTARTBLOCK(%lld)", startblockval(bno));
-	else
-		sprintf(rval, "%lld", (xfs_dfsbno_t)bno);
-	return rval;
-}
-
-/*
  * We include this last to have the helpers above available for the trace
  * event implementations.
  */
diff --git a/fs/xfs/linux-2.6/xfs_trace.h b/fs/xfs/linux-2.6/xfs_trace.h
index c22a608321a3..fcaa62f0799e 100644
--- a/fs/xfs/linux-2.6/xfs_trace.h
+++ b/fs/xfs/linux-2.6/xfs_trace.h
@@ -78,6 +78,33 @@ DECLARE_EVENT_CLASS(xfs_attr_list_class,
 	)
 )
 
+#define DEFINE_PERAG_REF_EVENT(name) \
+TRACE_EVENT(name, \
+	TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, int refcount, \
+		 unsigned long caller_ip), \
+	TP_ARGS(mp, agno, refcount, caller_ip), \
+	TP_STRUCT__entry( \
+		__field(dev_t, dev) \
+		__field(xfs_agnumber_t, agno) \
+		__field(int, refcount) \
+		__field(unsigned long, caller_ip) \
+	), \
+	TP_fast_assign( \
+		__entry->dev = mp->m_super->s_dev; \
+		__entry->agno = agno; \
+		__entry->refcount = refcount; \
+		__entry->caller_ip = caller_ip; \
+	), \
+	TP_printk("dev %d:%d agno %u refcount %d caller %pf", \
+		  MAJOR(__entry->dev), MINOR(__entry->dev), \
+		  __entry->agno, \
+		  __entry->refcount, \
+		  (char *)__entry->caller_ip) \
+);
+
+DEFINE_PERAG_REF_EVENT(xfs_perag_get)
+DEFINE_PERAG_REF_EVENT(xfs_perag_put)
+
 #define DEFINE_ATTR_LIST_EVENT(name) \
 DEFINE_EVENT(xfs_attr_list_class, name, \
 	TP_PROTO(struct xfs_attr_list_context *ctx), \
@@ -170,13 +197,13 @@ TRACE_EVENT(xfs_iext_insert,
 		__entry->caller_ip = caller_ip;
 	),
 	TP_printk("dev %d:%d ino 0x%llx state %s idx %ld "
-		  "offset %lld block %s count %lld flag %d caller %pf",
+		  "offset %lld block %lld count %lld flag %d caller %pf",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
 		  __entry->ino,
 		  __print_flags(__entry->bmap_state, "|", XFS_BMAP_EXT_FLAGS),
 		  (long)__entry->idx,
 		  __entry->startoff,
-		  xfs_fmtfsblock(__entry->startblock),
+		  (__int64_t)__entry->startblock,
 		  __entry->blockcount,
 		  __entry->state,
 		  (char *)__entry->caller_ip)
@@ -214,13 +241,13 @@ DECLARE_EVENT_CLASS(xfs_bmap_class,
 		__entry->caller_ip = caller_ip;
 	),
 	TP_printk("dev %d:%d ino 0x%llx state %s idx %ld "
-		  "offset %lld block %s count %lld flag %d caller %pf",
+		  "offset %lld block %lld count %lld flag %d caller %pf",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
 		  __entry->ino,
 		  __print_flags(__entry->bmap_state, "|", XFS_BMAP_EXT_FLAGS),
 		  (long)__entry->idx,
 		  __entry->startoff,
-		  xfs_fmtfsblock(__entry->startblock),
+		  (__int64_t)__entry->startblock,
 		  __entry->blockcount,
 		  __entry->state,
 		  (char *)__entry->caller_ip)
@@ -456,6 +483,7 @@ DEFINE_BUF_ITEM_EVENT(xfs_buf_item_unlock);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_unlock_stale);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_committed);
 DEFINE_BUF_ITEM_EVENT(xfs_buf_item_push);
+DEFINE_BUF_ITEM_EVENT(xfs_buf_item_pushbuf);
 DEFINE_BUF_ITEM_EVENT(xfs_trans_get_buf);
 DEFINE_BUF_ITEM_EVENT(xfs_trans_get_buf_recur);
 DEFINE_BUF_ITEM_EVENT(xfs_trans_getsb);
@@ -565,7 +593,7 @@ DECLARE_EVENT_CLASS(xfs_dquot_class,
 	TP_ARGS(dqp),
 	TP_STRUCT__entry(
 		__field(dev_t, dev)
-		__field(__be32, id)
+		__field(u32, id)
 		__field(unsigned, flags)
 		__field(unsigned, nrefs)
 		__field(unsigned long long, res_bcount)
@@ -578,7 +606,7 @@ DECLARE_EVENT_CLASS(xfs_dquot_class,
 	), \
 	TP_fast_assign(
 		__entry->dev = dqp->q_mount->m_super->s_dev;
-		__entry->id = dqp->q_core.d_id;
+		__entry->id = be32_to_cpu(dqp->q_core.d_id);
 		__entry->flags = dqp->dq_flags;
 		__entry->nrefs = dqp->q_nrefs;
 		__entry->res_bcount = dqp->q_res_bcount;
@@ -594,10 +622,10 @@ DECLARE_EVENT_CLASS(xfs_dquot_class,
 			be64_to_cpu(dqp->q_core.d_ino_softlimit);
 	),
 	TP_printk("dev %d:%d id 0x%x flags %s nrefs %u res_bc 0x%llx "
-		  "bcnt 0x%llx [hard 0x%llx | soft 0x%llx] "
-		  "icnt 0x%llx [hard 0x%llx | soft 0x%llx]",
+		  "bcnt 0x%llx bhardlimit 0x%llx bsoftlimit 0x%llx "
+		  "icnt 0x%llx ihardlimit 0x%llx isoftlimit 0x%llx]",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
-		  be32_to_cpu(__entry->id),
+		  __entry->id,
 		  __print_flags(__entry->flags, "|", XFS_DQ_FLAGS),
 		  __entry->nrefs,
 		  __entry->res_bcount,
@@ -853,7 +881,7 @@ TRACE_EVENT(name, \
 	), \
 	TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " \
 		  "offset 0x%llx count %zd flags %s " \
-		  "startoff 0x%llx startblock %s blockcount 0x%llx", \
+		  "startoff 0x%llx startblock %lld blockcount 0x%llx", \
 		  MAJOR(__entry->dev), MINOR(__entry->dev), \
 		  __entry->ino, \
 		  __entry->size, \
@@ -862,7 +890,7 @@ TRACE_EVENT(name, \
 		  __entry->count, \
 		  __print_flags(__entry->flags, "|", BMAPI_FLAGS), \
 		  __entry->startoff, \
-		  xfs_fmtfsblock(__entry->startblock), \
+		  (__int64_t)__entry->startblock, \
 		  __entry->blockcount) \
 )
 DEFINE_IOMAP_EVENT(xfs_iomap_enter);
@@ -1414,6 +1442,59 @@ TRACE_EVENT(xfs_dir2_leafn_moveents,
 		  __entry->count)
 );
 
+#define XFS_SWAPEXT_INODES \
+	{ 0,	"target" }, \
+	{ 1,	"temp" }
+
+#define XFS_INODE_FORMAT_STR \
+	{ 0,	"invalid" }, \
+	{ 1,	"local" }, \
+	{ 2,	"extent" }, \
+	{ 3,	"btree" }
+
+DECLARE_EVENT_CLASS(xfs_swap_extent_class,
+	TP_PROTO(struct xfs_inode *ip, int which),
+	TP_ARGS(ip, which),
+	TP_STRUCT__entry(
+		__field(dev_t, dev)
+		__field(int, which)
+		__field(xfs_ino_t, ino)
+		__field(int, format)
+		__field(int, nex)
+		__field(int, max_nex)
+		__field(int, broot_size)
+		__field(int, fork_off)
+	),
+	TP_fast_assign(
+		__entry->dev = VFS_I(ip)->i_sb->s_dev;
+		__entry->which = which;
+		__entry->ino = ip->i_ino;
+		__entry->format = ip->i_d.di_format;
+		__entry->nex = ip->i_d.di_nextents;
+		__entry->max_nex = ip->i_df.if_ext_max;
+		__entry->broot_size = ip->i_df.if_broot_bytes;
+		__entry->fork_off = XFS_IFORK_BOFF(ip);
+	),
+	TP_printk("dev %d:%d ino 0x%llx (%s), %s format, num_extents %d, "
+		  "Max in-fork extents %d, broot size %d, fork offset %d",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->ino,
+		  __print_symbolic(__entry->which, XFS_SWAPEXT_INODES),
+		  __print_symbolic(__entry->format, XFS_INODE_FORMAT_STR),
+		  __entry->nex,
+		  __entry->max_nex,
+		  __entry->broot_size,
+		  __entry->fork_off)
+)
+
+#define DEFINE_SWAPEXT_EVENT(name) \
+DEFINE_EVENT(xfs_swap_extent_class, name, \
+	TP_PROTO(struct xfs_inode *ip, int which), \
+	TP_ARGS(ip, which))
+
+DEFINE_SWAPEXT_EVENT(xfs_swap_extent_before);
+DEFINE_SWAPEXT_EVENT(xfs_swap_extent_after);
+
 #endif /* _TRACE_XFS_H */
 
 #undef TRACE_INCLUDE_PATH
diff --git a/fs/xfs/linux-2.6/xfs_xattr.c b/fs/xfs/linux-2.6/xfs_xattr.c
index 0b1878857fc3..fa01b9daba6b 100644
--- a/fs/xfs/linux-2.6/xfs_xattr.c
+++ b/fs/xfs/linux-2.6/xfs_xattr.c
@@ -45,7 +45,7 @@ xfs_xattr_get(struct dentry *dentry, const char *name,
 		value = NULL;
 	}
 
-	error = -xfs_attr_get(ip, name, value, &asize, xflags);
+	error = -xfs_attr_get(ip, (unsigned char *)name, value, &asize, xflags);
 	if (error)
 		return error;
 	return asize;
@@ -67,8 +67,9 @@ xfs_xattr_set(struct dentry *dentry, const char *name, const void *value,
 		xflags |= ATTR_REPLACE;
 
 	if (!value)
-		return -xfs_attr_remove(ip, name, xflags);
-	return -xfs_attr_set(ip, name, (void *)value, size, xflags);
+		return -xfs_attr_remove(ip, (unsigned char *)name, xflags);
+	return -xfs_attr_set(ip, (unsigned char *)name,
+				(void *)value, size, xflags);
 }
 
 static struct xattr_handler xfs_xattr_user_handler = {
@@ -124,8 +125,13 @@ static const char *xfs_xattr_prefix(int flags)
 }
 
 static int
-xfs_xattr_put_listent(struct xfs_attr_list_context *context, int flags,
-		char *name, int namelen, int valuelen, char *value)
+xfs_xattr_put_listent(
+	struct xfs_attr_list_context *context,
+	int		flags,
+	unsigned char	*name,
+	int		namelen,
+	int		valuelen,
+	unsigned char	*value)
 {
 	unsigned int prefix_len = xfs_xattr_prefix_len(flags);
 	char *offset;
@@ -148,7 +154,7 @@ xfs_xattr_put_listent(struct xfs_attr_list_context *context, int flags,
 	offset = (char *)context->alist + context->count;
 	strncpy(offset, xfs_xattr_prefix(flags), prefix_len);
 	offset += prefix_len;
-	strncpy(offset, name, namelen);			/* real name */
+	strncpy(offset, (char *)name, namelen);			/* real name */
 	offset += namelen;
 	*offset = '\0';
 	context->count += prefix_len + namelen + 1;
@@ -156,8 +162,13 @@ xfs_xattr_put_listent(struct xfs_attr_list_context *context, int flags,
 }
 
 static int
-xfs_xattr_put_listent_sizes(struct xfs_attr_list_context *context, int flags,
-		char *name, int namelen, int valuelen, char *value)
+xfs_xattr_put_listent_sizes(
+	struct xfs_attr_list_context *context,
+	int		flags,
+	unsigned char	*name,
+	int		namelen,
+	int		valuelen,
+	unsigned char	*value)
 {
 	context->count += xfs_xattr_prefix_len(flags) + namelen + 1;
 	return 0;
diff --git a/fs/xfs/quota/xfs_dquot.c b/fs/xfs/quota/xfs_dquot.c
index d7c7eea09fc2..5f79dd78626b 100644
--- a/fs/xfs/quota/xfs_dquot.c
+++ b/fs/xfs/quota/xfs_dquot.c
@@ -1187,7 +1187,7 @@ xfs_qm_dqflush(
 	 * block, nada.
 	 */
 	if (!XFS_DQ_IS_DIRTY(dqp) ||
-	    (!(flags & XFS_QMOPT_SYNC) && atomic_read(&dqp->q_pincount) > 0)) {
+	    (!(flags & SYNC_WAIT) && atomic_read(&dqp->q_pincount) > 0)) {
 		xfs_dqfunlock(dqp);
 		return 0;
 	}
@@ -1248,23 +1248,20 @@ xfs_qm_dqflush(
 	 */
 	if (XFS_BUF_ISPINNED(bp)) {
 		trace_xfs_dqflush_force(dqp);
-		xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
+		xfs_log_force(mp, 0);
 	}
 
-	if (flags & XFS_QMOPT_DELWRI) {
-		xfs_bdwrite(mp, bp);
-	} else if (flags & XFS_QMOPT_ASYNC) {
-		error = xfs_bawrite(mp, bp);
-	} else {
+	if (flags & SYNC_WAIT)
 		error = xfs_bwrite(mp, bp);
-	}
+	else
+		xfs_bdwrite(mp, bp);
 
 	trace_xfs_dqflush_done(dqp);
 
 	/*
 	 * dqp is still locked, but caller is free to unlock it now.
 	 */
-	return (error);
+	return error;
 
 }
 
@@ -1445,7 +1442,7 @@ xfs_qm_dqpurge(
 		 * We don't care about getting disk errors here. We need
 		 * to purge this dquot anyway, so we go ahead regardless.
 		 */
-		error = xfs_qm_dqflush(dqp, XFS_QMOPT_SYNC);
+		error = xfs_qm_dqflush(dqp, SYNC_WAIT);
 		if (error)
 			xfs_fs_cmn_err(CE_WARN, mp,
 				"xfs_qm_dqpurge: dquot %p flush failed", dqp);
@@ -1529,25 +1526,17 @@ xfs_qm_dqflock_pushbuf_wait(
 	 * the flush lock when the I/O completes.
 	 */
 	bp = xfs_incore(dqp->q_mount->m_ddev_targp, dqp->q_blkno,
-		    XFS_QI_DQCHUNKLEN(dqp->q_mount),
-		    XFS_INCORE_TRYLOCK);
-	if (bp != NULL) {
-		if (XFS_BUF_ISDELAYWRITE(bp)) {
-			int	error;
-			if (XFS_BUF_ISPINNED(bp)) {
-				xfs_log_force(dqp->q_mount,
-					      (xfs_lsn_t)0,
-					      XFS_LOG_FORCE);
-			}
-			error = xfs_bawrite(dqp->q_mount, bp);
-			if (error)
-				xfs_fs_cmn_err(CE_WARN, dqp->q_mount,
-					"xfs_qm_dqflock_pushbuf_wait: "
-					"pushbuf error %d on dqp %p, bp %p",
-					error, dqp, bp);
-		} else {
-			xfs_buf_relse(bp);
-		}
+		    XFS_QI_DQCHUNKLEN(dqp->q_mount), XBF_TRYLOCK);
+	if (!bp)
+		goto out_lock;
+
+	if (XFS_BUF_ISDELAYWRITE(bp)) {
+		if (XFS_BUF_ISPINNED(bp))
+			xfs_log_force(dqp->q_mount, 0);
+		xfs_buf_delwri_promote(bp);
+		wake_up_process(bp->b_target->bt_task);
 	}
+	xfs_buf_relse(bp);
+out_lock:
 	xfs_dqflock(dqp);
 }
diff --git a/fs/xfs/quota/xfs_dquot_item.c b/fs/xfs/quota/xfs_dquot_item.c
index d0d4a9a0bbd7..4e4ee9a57194 100644
--- a/fs/xfs/quota/xfs_dquot_item.c
+++ b/fs/xfs/quota/xfs_dquot_item.c
@@ -74,11 +74,11 @@ xfs_qm_dquot_logitem_format(
 
 	logvec->i_addr = (xfs_caddr_t)&logitem->qli_format;
 	logvec->i_len  = sizeof(xfs_dq_logformat_t);
-	XLOG_VEC_SET_TYPE(logvec, XLOG_REG_TYPE_QFORMAT);
+	logvec->i_type = XLOG_REG_TYPE_QFORMAT;
 	logvec++;
 	logvec->i_addr = (xfs_caddr_t)&logitem->qli_dquot->q_core;
 	logvec->i_len  = sizeof(xfs_disk_dquot_t);
-	XLOG_VEC_SET_TYPE(logvec, XLOG_REG_TYPE_DQUOT);
+	logvec->i_type = XLOG_REG_TYPE_DQUOT;
 
 	ASSERT(2 == logitem->qli_item.li_desc->lid_size);
 	logitem->qli_format.qlf_size = 2;
@@ -153,7 +153,7 @@ xfs_qm_dquot_logitem_push(
 	 * lock without sleeping, then there must not have been
 	 * anyone in the process of flushing the dquot.
 	 */
-	error = xfs_qm_dqflush(dqp, XFS_QMOPT_DELWRI);
+	error = xfs_qm_dqflush(dqp, 0);
 	if (error)
 		xfs_fs_cmn_err(CE_WARN, dqp->q_mount,
 			"xfs_qm_dquot_logitem_push: push error %d on dqp %p",
@@ -190,7 +190,7 @@ xfs_qm_dqunpin_wait(
 	/*
 	 * Give the log a push so we don't wait here too long.
 	 */
-	xfs_log_force(dqp->q_mount, (xfs_lsn_t)0, XFS_LOG_FORCE);
+	xfs_log_force(dqp->q_mount, 0);
 	wait_event(dqp->q_pinwait, (atomic_read(&dqp->q_pincount) == 0));
 }
 
@@ -212,68 +212,31 @@ xfs_qm_dquot_logitem_pushbuf(
 	xfs_dquot_t	*dqp;
 	xfs_mount_t	*mp;
 	xfs_buf_t	*bp;
-	uint		dopush;
 
 	dqp = qip->qli_dquot;
 	ASSERT(XFS_DQ_IS_LOCKED(dqp));
 
 	/*
-	 * The qli_pushbuf_flag keeps others from
-	 * trying to duplicate our effort.
-	 */
-	ASSERT(qip->qli_pushbuf_flag != 0);
-	ASSERT(qip->qli_push_owner == current_pid());
-
-	/*
 	 * If flushlock isn't locked anymore, chances are that the
 	 * inode flush completed and the inode was taken off the AIL.
 	 * So, just get out.
 	 */
 	if (completion_done(&dqp->q_flush)  ||
 	    ((qip->qli_item.li_flags & XFS_LI_IN_AIL) == 0)) {
-		qip->qli_pushbuf_flag = 0;
 		xfs_dqunlock(dqp);
 		return;
 	}
 	mp = dqp->q_mount;
 	bp = xfs_incore(mp->m_ddev_targp, qip->qli_format.qlf_blkno,
-		    XFS_QI_DQCHUNKLEN(mp),
-		    XFS_INCORE_TRYLOCK);
-	if (bp != NULL) {
-		if (XFS_BUF_ISDELAYWRITE(bp)) {
-			dopush = ((qip->qli_item.li_flags & XFS_LI_IN_AIL) &&
-				  !completion_done(&dqp->q_flush));
-			qip->qli_pushbuf_flag = 0;
-			xfs_dqunlock(dqp);
-
-			if (XFS_BUF_ISPINNED(bp)) {
-				xfs_log_force(mp, (xfs_lsn_t)0,
-					      XFS_LOG_FORCE);
-			}
-			if (dopush) {
-				int	error;
-#ifdef XFSRACEDEBUG
-				delay_for_intr();
-				delay(300);
-#endif
-				error = xfs_bawrite(mp, bp);
-				if (error)
-					xfs_fs_cmn_err(CE_WARN, mp,
-	"xfs_qm_dquot_logitem_pushbuf: pushbuf error %d on qip %p, bp %p",
-							error, qip, bp);
-			} else {
-				xfs_buf_relse(bp);
-			}
-		} else {
-			qip->qli_pushbuf_flag = 0;
-			xfs_dqunlock(dqp);
-			xfs_buf_relse(bp);
-		}
+		    XFS_QI_DQCHUNKLEN(mp), XBF_TRYLOCK);
+	xfs_dqunlock(dqp);
+	if (!bp)
 		return;
-	}
+	if (XFS_BUF_ISDELAYWRITE(bp))
+		xfs_buf_delwri_promote(bp);
+	xfs_buf_relse(bp);
+	return;
 
-	qip->qli_pushbuf_flag = 0;
-	xfs_dqunlock(dqp);
 }
 
 /*
@@ -291,50 +254,24 @@ xfs_qm_dquot_logitem_trylock(
 	xfs_dq_logitem_t	*qip)
 {
 	xfs_dquot_t		*dqp;
-	uint			retval;
 
 	dqp = qip->qli_dquot;
 	if (atomic_read(&dqp->q_pincount) > 0)
-		return (XFS_ITEM_PINNED);
+		return XFS_ITEM_PINNED;
 
 	if (! xfs_qm_dqlock_nowait(dqp))
-		return (XFS_ITEM_LOCKED);
+		return XFS_ITEM_LOCKED;
 
-	retval = XFS_ITEM_SUCCESS;
 	if (!xfs_dqflock_nowait(dqp)) {
 		/*
-		 * The dquot is already being flushed.	It may have been
-		 * flushed delayed write, however, and we don't want to
-		 * get stuck waiting for that to complete.  So, we want to check
-		 * to see if we can lock the dquot's buffer without sleeping.
-		 * If we can and it is marked for delayed write, then we
-		 * hold it and send it out from the push routine.  We don't
-		 * want to do that now since we might sleep in the device
-		 * strategy routine.  We also don't want to grab the buffer lock
-		 * here because we'd like not to call into the buffer cache
-		 * while holding the AIL lock.
-		 * Make sure to only return PUSHBUF if we set pushbuf_flag
-		 * ourselves.  If someone else is doing it then we don't
-		 * want to go to the push routine and duplicate their efforts.
+		 * dquot has already been flushed to the backing buffer,
+		 * leave it locked, pushbuf routine will unlock it.
 		 */
-		if (qip->qli_pushbuf_flag == 0) {
-			qip->qli_pushbuf_flag = 1;
-			ASSERT(qip->qli_format.qlf_blkno == dqp->q_blkno);
-#ifdef DEBUG
-			qip->qli_push_owner = current_pid();
-#endif
-			/*
-			 * The dquot is left locked.
-			 */
-			retval = XFS_ITEM_PUSHBUF;
-		} else {
-			retval = XFS_ITEM_FLUSHING;
-			xfs_dqunlock_nonotify(dqp);
-		}
+		return XFS_ITEM_PUSHBUF;
 	}
 
 	ASSERT(qip->qli_item.li_flags & XFS_LI_IN_AIL);
-	return (retval);
+	return XFS_ITEM_SUCCESS;
 }
 
 
@@ -467,7 +404,7 @@ xfs_qm_qoff_logitem_format(xfs_qoff_logitem_t	*qf,
 
 	log_vector->i_addr = (xfs_caddr_t)&(qf->qql_format);
 	log_vector->i_len = sizeof(xfs_qoff_logitem_t);
-	XLOG_VEC_SET_TYPE(log_vector, XLOG_REG_TYPE_QUOTAOFF);
+	log_vector->i_type = XLOG_REG_TYPE_QUOTAOFF;
 	qf->qql_format.qf_size = 1;
 }
 
diff --git a/fs/xfs/quota/xfs_dquot_item.h b/fs/xfs/quota/xfs_dquot_item.h
index 5a632531f843..5acae2ada70b 100644
--- a/fs/xfs/quota/xfs_dquot_item.h
+++ b/fs/xfs/quota/xfs_dquot_item.h
@@ -27,10 +27,6 @@ typedef struct xfs_dq_logitem {
 	xfs_log_item_t		 qli_item;	   /* common portion */
 	struct xfs_dquot	*qli_dquot;	   /* dquot ptr */
 	xfs_lsn_t		 qli_flush_lsn;	   /* lsn at last flush */
-	unsigned short		 qli_pushbuf_flag; /* 1 bit used in push_ail */
-#ifdef DEBUG
-	uint64_t		 qli_push_owner;
-#endif
 	xfs_dq_logformat_t	 qli_format;	   /* logged structure */
 } xfs_dq_logitem_t;
 
diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c
index 9e627a8b5b0e..417e61e3d9dd 100644
--- a/fs/xfs/quota/xfs_qm.c
+++ b/fs/xfs/quota/xfs_qm.c
@@ -118,9 +118,14 @@ xfs_Gqm_init(void)
 	 */
 	udqhash = kmem_zalloc_greedy(&hsize,
 				     XFS_QM_HASHSIZE_LOW * sizeof(xfs_dqhash_t),
-				     XFS_QM_HASHSIZE_HIGH * sizeof(xfs_dqhash_t),
-				     KM_SLEEP | KM_MAYFAIL | KM_LARGE);
-	gdqhash = kmem_zalloc(hsize, KM_SLEEP | KM_LARGE);
+				     XFS_QM_HASHSIZE_HIGH * sizeof(xfs_dqhash_t));
+	if (!udqhash)
+		goto out;
+
+	gdqhash = kmem_zalloc_large(hsize);
+	if (!gdqhash)
+		goto out_free_udqhash;
+
 	hsize /= sizeof(xfs_dqhash_t);
 	ndquot = hsize << 8;
 
@@ -170,6 +175,11 @@ xfs_Gqm_init(void)
 	mutex_init(&qcheck_lock);
 #endif
 	return xqm;
+
+ out_free_udqhash:
+	kmem_free_large(udqhash);
+ out:
+	return NULL;
 }
 
 /*
@@ -189,8 +199,8 @@ xfs_qm_destroy(
 		xfs_qm_list_destroy(&(xqm->qm_usr_dqhtable[i]));
 		xfs_qm_list_destroy(&(xqm->qm_grp_dqhtable[i]));
 	}
-	kmem_free(xqm->qm_usr_dqhtable);
-	kmem_free(xqm->qm_grp_dqhtable);
+	kmem_free_large(xqm->qm_usr_dqhtable);
+	kmem_free_large(xqm->qm_grp_dqhtable);
 	xqm->qm_usr_dqhtable = NULL;
 	xqm->qm_grp_dqhtable = NULL;
 	xqm->qm_dqhashmask = 0;
@@ -219,8 +229,12 @@ xfs_qm_hold_quotafs_ref(
 	 */
 	mutex_lock(&xfs_Gqm_lock);
 
-	if (xfs_Gqm == NULL)
+	if (!xfs_Gqm) {
 		xfs_Gqm = xfs_Gqm_init();
+		if (!xfs_Gqm)
+			return ENOMEM;
+	}
+
 	/*
 	 * We can keep a list of all filesystems with quotas mounted for
 	 * debugging and statistical purposes, but ...
@@ -436,7 +450,7 @@ xfs_qm_unmount_quotas(
 STATIC int
 xfs_qm_dqflush_all(
 	xfs_mount_t	*mp,
-	int		flags)
+	int		sync_mode)
 {
 	int		recl;
 	xfs_dquot_t	*dqp;
@@ -472,7 +486,7 @@ again:
 		 * across a disk write.
 		 */
 		xfs_qm_mplist_unlock(mp);
-		error = xfs_qm_dqflush(dqp, flags);
+		error = xfs_qm_dqflush(dqp, sync_mode);
 		xfs_dqunlock(dqp);
 		if (error)
 			return error;
@@ -912,13 +926,11 @@ xfs_qm_sync(
 {
 	int		recl, restarts;
 	xfs_dquot_t	*dqp;
-	uint		flush_flags;
 	int		error;
 
 	if (!XFS_IS_QUOTA_RUNNING(mp) || !XFS_IS_QUOTA_ON(mp))
 		return 0;
 
-	flush_flags = (flags & SYNC_WAIT) ? XFS_QMOPT_SYNC : XFS_QMOPT_DELWRI;
 	restarts = 0;
 
   again:
@@ -978,7 +990,7 @@ xfs_qm_sync(
 		 * across a disk write
 		 */
 		xfs_qm_mplist_unlock(mp);
-		error = xfs_qm_dqflush(dqp, flush_flags);
+		error = xfs_qm_dqflush(dqp, flags);
 		xfs_dqunlock(dqp);
 		if (error && XFS_FORCED_SHUTDOWN(mp))
 			return 0;	/* Need to prevent umount failure */
@@ -1782,7 +1794,7 @@ xfs_qm_quotacheck(
 	 * successfully.
 	 */
 	if (!error)
-		error = xfs_qm_dqflush_all(mp, XFS_QMOPT_DELWRI);
+		error = xfs_qm_dqflush_all(mp, 0);
 
 	/*
 	 * We can get this error if we couldn't do a dquot allocation inside
@@ -2004,7 +2016,7 @@ xfs_qm_shake_freelist(
 			 * We flush it delayed write, so don't bother
 			 * releasing the mplock.
 			 */
-			error = xfs_qm_dqflush(dqp, XFS_QMOPT_DELWRI);
+			error = xfs_qm_dqflush(dqp, 0);
 			if (error) {
 				xfs_fs_cmn_err(CE_WARN, dqp->q_mount,
 			"xfs_qm_dqflush_all: dquot %p flush failed", dqp);
@@ -2187,7 +2199,7 @@ xfs_qm_dqreclaim_one(void)
 			 * We flush it delayed write, so don't bother
 			 * releasing the freelist lock.
 			 */
-			error = xfs_qm_dqflush(dqp, XFS_QMOPT_DELWRI);
+			error = xfs_qm_dqflush(dqp, 0);
 			if (error) {
 				xfs_fs_cmn_err(CE_WARN, dqp->q_mount,
 			"xfs_qm_dqreclaim: dquot %p flush failed", dqp);
diff --git a/fs/xfs/quota/xfs_qm_bhv.c b/fs/xfs/quota/xfs_qm_bhv.c
index a5346630dfae..97b410c12794 100644
--- a/fs/xfs/quota/xfs_qm_bhv.c
+++ b/fs/xfs/quota/xfs_qm_bhv.c
@@ -59,7 +59,7 @@ xfs_fill_statvfs_from_dquot(
 		be64_to_cpu(dp->d_blk_hardlimit);
 	if (limit && statp->f_blocks > limit) {
 		statp->f_blocks = limit;
-		statp->f_bfree =
+		statp->f_bfree = statp->f_bavail =
 			(statp->f_blocks > be64_to_cpu(dp->d_bcount)) ?
 			 (statp->f_blocks - be64_to_cpu(dp->d_bcount)) : 0;
 	}
diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c
index 873e07e29074..5d0ee8d492db 100644
--- a/fs/xfs/quota/xfs_qm_syscalls.c
+++ b/fs/xfs/quota/xfs_qm_syscalls.c
@@ -1192,9 +1192,9 @@ xfs_qm_internalqcheck(
 	if (! XFS_IS_QUOTA_ON(mp))
 		return XFS_ERROR(ESRCH);
 
-	xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC);
+	xfs_log_force(mp, XFS_LOG_SYNC);
 	XFS_bflush(mp->m_ddev_targp);
-	xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC);
+	xfs_log_force(mp, XFS_LOG_SYNC);
 	XFS_bflush(mp->m_ddev_targp);
 
 	mutex_lock(&qcheck_lock);
diff --git a/fs/xfs/quota/xfs_trans_dquot.c b/fs/xfs/quota/xfs_trans_dquot.c
index 97ac9640be98..c3ab75cb1d9a 100644
--- a/fs/xfs/quota/xfs_trans_dquot.c
+++ b/fs/xfs/quota/xfs_trans_dquot.c
@@ -589,12 +589,18 @@ xfs_trans_unreserve_and_mod_dquots(
 	}
 }
 
-STATIC int
-xfs_quota_error(uint flags)
+STATIC void
+xfs_quota_warn(
+	struct xfs_mount	*mp,
+	struct xfs_dquot	*dqp,
+	int			type)
 {
-	if (flags & XFS_QMOPT_ENOSPC)
-		return ENOSPC;
-	return EDQUOT;
+	/* no warnings for project quotas - we just return ENOSPC later */
+	if (dqp->dq_flags & XFS_DQ_PROJ)
+		return;
+	quota_send_warning((dqp->dq_flags & XFS_DQ_USER) ? USRQUOTA : GRPQUOTA,
+			   be32_to_cpu(dqp->q_core.d_id), mp->m_super->s_dev,
+			   type);
 }
 
 /*
@@ -612,7 +618,6 @@ xfs_trans_dqresv(
 	long		ninos,
 	uint		flags)
 {
-	int		error;
 	xfs_qcnt_t	hardlimit;
 	xfs_qcnt_t	softlimit;
 	time_t		timer;
@@ -649,7 +654,6 @@ xfs_trans_dqresv(
 		warnlimit = XFS_QI_RTBWARNLIMIT(dqp->q_mount);
 		resbcountp = &dqp->q_res_rtbcount;
 	}
-	error = 0;
 
 	if ((flags & XFS_QMOPT_FORCE_RES) == 0 &&
 	    dqp->q_core.d_id &&
@@ -667,18 +671,20 @@ xfs_trans_dqresv(
 			 * nblks.
 			 */
 			if (hardlimit > 0ULL &&
-			     (hardlimit <= nblks + *resbcountp)) {
-				error = xfs_quota_error(flags);
+			    hardlimit <= nblks + *resbcountp) {
+				xfs_quota_warn(mp, dqp, QUOTA_NL_BHARDWARN);
 				goto error_return;
 			}
-
 			if (softlimit > 0ULL &&
-			     (softlimit <= nblks + *resbcountp)) {
+			    softlimit <= nblks + *resbcountp) {
 				if ((timer != 0 && get_seconds() > timer) ||
 				    (warns != 0 && warns >= warnlimit)) {
-					error = xfs_quota_error(flags);
+					xfs_quota_warn(mp, dqp,
+						       QUOTA_NL_BSOFTLONGWARN);
 					goto error_return;
 				}
+
+				xfs_quota_warn(mp, dqp, QUOTA_NL_BSOFTWARN);
 			}
 		}
 		if (ninos > 0) {
@@ -692,15 +698,19 @@ xfs_trans_dqresv(
 			softlimit = be64_to_cpu(dqp->q_core.d_ino_softlimit);
 			if (!softlimit)
 				softlimit = q->qi_isoftlimit;
+
 			if (hardlimit > 0ULL && count >= hardlimit) {
-				error = xfs_quota_error(flags);
+				xfs_quota_warn(mp, dqp, QUOTA_NL_IHARDWARN);
 				goto error_return;
-			} else if (softlimit > 0ULL && count >= softlimit) {
-				if ((timer != 0 && get_seconds() > timer) ||
+			}
+			if (softlimit > 0ULL && count >= softlimit) {
+				if  ((timer != 0 && get_seconds() > timer) ||
 				     (warns != 0 && warns >= warnlimit)) {
-					error = xfs_quota_error(flags);
+					xfs_quota_warn(mp, dqp,
+						       QUOTA_NL_ISOFTLONGWARN);
 					goto error_return;
 				}
+				xfs_quota_warn(mp, dqp, QUOTA_NL_ISOFTWARN);
 			}
 		}
 	}
@@ -736,9 +746,14 @@ xfs_trans_dqresv(
 	ASSERT(dqp->q_res_rtbcount >= be64_to_cpu(dqp->q_core.d_rtbcount));
 	ASSERT(dqp->q_res_icount >= be64_to_cpu(dqp->q_core.d_icount));
 
+	xfs_dqunlock(dqp);
+	return 0;
+
 error_return:
 	xfs_dqunlock(dqp);
-	return error;
+	if (flags & XFS_QMOPT_ENOSPC)
+		return ENOSPC;
+	return EDQUOT;
 }
 
 
diff --git a/fs/xfs/xfs_acl.h b/fs/xfs/xfs_acl.h
index 00fd357c3e46..d13eeba2c8f8 100644
--- a/fs/xfs/xfs_acl.h
+++ b/fs/xfs/xfs_acl.h
@@ -36,8 +36,8 @@ struct xfs_acl {
 };
 
 /* On-disk XFS extended attribute names */
-#define SGI_ACL_FILE		"SGI_ACL_FILE"
-#define SGI_ACL_DEFAULT		"SGI_ACL_DEFAULT"
+#define SGI_ACL_FILE		(unsigned char *)"SGI_ACL_FILE"
+#define SGI_ACL_DEFAULT		(unsigned char *)"SGI_ACL_DEFAULT"
 #define SGI_ACL_FILE_SIZE	(sizeof(SGI_ACL_FILE)-1)
 #define SGI_ACL_DEFAULT_SIZE	(sizeof(SGI_ACL_DEFAULT)-1)
 
diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h
index 6702bd865811..b1a5a1ff88ea 100644
--- a/fs/xfs/xfs_ag.h
+++ b/fs/xfs/xfs_ag.h
@@ -187,17 +187,13 @@ typedef struct xfs_perag_busy {
 /*
  * Per-ag incore structure, copies of information in agf and agi,
  * to improve the performance of allocation group selection.
- *
- * pick sizes which fit in allocation buckets well
  */
-#if (BITS_PER_LONG == 32)
-#define XFS_PAGB_NUM_SLOTS	84
-#elif (BITS_PER_LONG == 64)
 #define XFS_PAGB_NUM_SLOTS	128
-#endif
 
-typedef struct xfs_perag
-{
+typedef struct xfs_perag {
+	struct xfs_mount *pag_mount;	/* owner filesystem */
+	xfs_agnumber_t	pag_agno;	/* AG this structure belongs to */
+	atomic_t	pag_ref;	/* perag reference count */
 	char		pagf_init;	/* this agf's entry is initialized */
 	char		pagi_init;	/* this agi's entry is initialized */
 	char		pagf_metadata;	/* the agf is preferred to be metadata */
@@ -210,8 +206,6 @@ typedef struct xfs_perag
 	__uint32_t	pagf_btreeblks;	/* # of blocks held in AGF btrees */
 	xfs_agino_t	pagi_freecount;	/* number of free inodes */
 	xfs_agino_t	pagi_count;	/* number of allocated inodes */
-	int		pagb_count;	/* pagb slots in use */
-	xfs_perag_busy_t *pagb_list;	/* unstable blocks */
 
 	/*
 	 * Inode allocation search lookup optimisation.
@@ -230,6 +224,8 @@ typedef struct xfs_perag
 	rwlock_t	pag_ici_lock;	/* incore inode lock */
 	struct radix_tree_root pag_ici_root;	/* incore inode cache root */
 #endif
+	int		pagb_count;	/* pagb slots in use */
+	xfs_perag_busy_t pagb_list[XFS_PAGB_NUM_SLOTS];	/* unstable blocks */
 } xfs_perag_t;
 
 /*
diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 275b1f4f9430..94cddbfb2560 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -1662,11 +1662,13 @@ xfs_free_ag_extent(
 		xfs_agf_t	*agf;
 		xfs_perag_t	*pag;		/* per allocation group data */
 
+		pag = xfs_perag_get(mp, agno);
+		pag->pagf_freeblks += len;
+		xfs_perag_put(pag);
+
 		agf = XFS_BUF_TO_AGF(agbp);
-		pag = &mp->m_perag[agno];
 		be32_add_cpu(&agf->agf_freeblks, len);
 		xfs_trans_agblocks_delta(tp, len);
-		pag->pagf_freeblks += len;
 		XFS_WANT_CORRUPTED_GOTO(
 			be32_to_cpu(agf->agf_freeblks) <=
 			be32_to_cpu(agf->agf_length),
@@ -1969,10 +1971,12 @@ xfs_alloc_get_freelist(
 	xfs_trans_brelse(tp, agflbp);
 	if (be32_to_cpu(agf->agf_flfirst) == XFS_AGFL_SIZE(mp))
 		agf->agf_flfirst = 0;
-	pag = &mp->m_perag[be32_to_cpu(agf->agf_seqno)];
+
+	pag = xfs_perag_get(mp, be32_to_cpu(agf->agf_seqno));
 	be32_add_cpu(&agf->agf_flcount, -1);
 	xfs_trans_agflist_delta(tp, -1);
 	pag->pagf_flcount--;
+	xfs_perag_put(pag);
 
 	logflags = XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT;
 	if (btreeblk) {
@@ -2078,7 +2082,8 @@ xfs_alloc_put_freelist(
 	be32_add_cpu(&agf->agf_fllast, 1);
 	if (be32_to_cpu(agf->agf_fllast) == XFS_AGFL_SIZE(mp))
 		agf->agf_fllast = 0;
-	pag = &mp->m_perag[be32_to_cpu(agf->agf_seqno)];
+
+	pag = xfs_perag_get(mp, be32_to_cpu(agf->agf_seqno));
 	be32_add_cpu(&agf->agf_flcount, 1);
 	xfs_trans_agflist_delta(tp, 1);
 	pag->pagf_flcount++;
@@ -2089,6 +2094,7 @@ xfs_alloc_put_freelist(
 		pag->pagf_btreeblks--;
 		logflags |= XFS_AGF_BTREEBLKS;
 	}
+	xfs_perag_put(pag);
 
 	xfs_alloc_log_agf(tp, agbp, logflags);
 
@@ -2152,7 +2158,6 @@ xfs_read_agf(
 		xfs_trans_brelse(tp, *bpp);
 		return XFS_ERROR(EFSCORRUPTED);
 	}
-
 	XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_AGF, XFS_AGF_REF);
 	return 0;
 }
@@ -2175,7 +2180,7 @@ xfs_alloc_read_agf(
 	ASSERT(agno != NULLAGNUMBER);
 
 	error = xfs_read_agf(mp, tp, agno,
-			(flags & XFS_ALLOC_FLAG_TRYLOCK) ? XFS_BUF_TRYLOCK : 0,
+			(flags & XFS_ALLOC_FLAG_TRYLOCK) ? XBF_TRYLOCK : 0,
 			bpp);
 	if (error)
 		return error;
@@ -2184,7 +2189,7 @@ xfs_alloc_read_agf(
 	ASSERT(!XFS_BUF_GETERROR(*bpp));
 
 	agf = XFS_BUF_TO_AGF(*bpp);
-	pag = &mp->m_perag[agno];
+	pag = xfs_perag_get(mp, agno);
 	if (!pag->pagf_init) {
 		pag->pagf_freeblks = be32_to_cpu(agf->agf_freeblks);
 		pag->pagf_btreeblks = be32_to_cpu(agf->agf_btreeblks);
@@ -2195,8 +2200,8 @@ xfs_alloc_read_agf(
 		pag->pagf_levels[XFS_BTNUM_CNTi] =
 			be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNTi]);
 		spin_lock_init(&pag->pagb_lock);
-		pag->pagb_list = kmem_zalloc(XFS_PAGB_NUM_SLOTS *
-					sizeof(xfs_perag_busy_t), KM_SLEEP);
+		pag->pagb_count = 0;
+		memset(pag->pagb_list, 0, sizeof(pag->pagb_list));
 		pag->pagf_init = 1;
 	}
 #ifdef DEBUG
@@ -2211,6 +2216,7 @@ xfs_alloc_read_agf(
 		       be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNTi]));
 	}
 #endif
+	xfs_perag_put(pag);
 	return 0;
 }
 
@@ -2270,8 +2276,7 @@ xfs_alloc_vextent(
 		 * These three force us into a single a.g.
 		 */
 		args->agno = XFS_FSB_TO_AGNO(mp, args->fsbno);
-		down_read(&mp->m_peraglock);
-		args->pag = &mp->m_perag[args->agno];
+		args->pag = xfs_perag_get(mp, args->agno);
 		args->minleft = 0;
 		error = xfs_alloc_fix_freelist(args, 0);
 		args->minleft = minleft;
@@ -2280,14 +2285,12 @@ xfs_alloc_vextent(
 			goto error0;
 		}
 		if (!args->agbp) {
-			up_read(&mp->m_peraglock);
 			trace_xfs_alloc_vextent_noagbp(args);
 			break;
 		}
 		args->agbno = XFS_FSB_TO_AGBNO(mp, args->fsbno);
 		if ((error = xfs_alloc_ag_vextent(args)))
 			goto error0;
-		up_read(&mp->m_peraglock);
 		break;
 	case XFS_ALLOCTYPE_START_BNO:
 		/*
@@ -2339,9 +2342,8 @@ xfs_alloc_vextent(
 		 * Loop over allocation groups twice; first time with
 		 * trylock set, second time without.
 		 */
-		down_read(&mp->m_peraglock);
 		for (;;) {
-			args->pag = &mp->m_perag[args->agno];
+			args->pag = xfs_perag_get(mp, args->agno);
 			if (no_min) args->minleft = 0;
 			error = xfs_alloc_fix_freelist(args, flags);
 			args->minleft = minleft;
@@ -2400,8 +2402,8 @@ xfs_alloc_vextent(
 					}
 				}
 			}
+			xfs_perag_put(args->pag);
 		}
-		up_read(&mp->m_peraglock);
 		if (bump_rotor || (type == XFS_ALLOCTYPE_ANY_AG)) {
 			if (args->agno == sagno)
 				mp->m_agfrotor = (mp->m_agfrotor + 1) %
@@ -2427,9 +2429,10 @@ xfs_alloc_vextent(
 			args->len);
 #endif
 	}
+	xfs_perag_put(args->pag);
 	return 0;
 error0:
-	up_read(&mp->m_peraglock);
+	xfs_perag_put(args->pag);
 	return error;
 }
 
@@ -2454,8 +2457,7 @@ xfs_free_extent(
 	args.agno = XFS_FSB_TO_AGNO(args.mp, bno);
 	ASSERT(args.agno < args.mp->m_sb.sb_agcount);
 	args.agbno = XFS_FSB_TO_AGBNO(args.mp, bno);
-	down_read(&args.mp->m_peraglock);
-	args.pag = &args.mp->m_perag[args.agno];
+	args.pag = xfs_perag_get(args.mp, args.agno);
 	if ((error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING)))
 		goto error0;
 #ifdef DEBUG
@@ -2465,7 +2467,7 @@ xfs_free_extent(
 #endif
 	error = xfs_free_ag_extent(tp, args.agbp, args.agno, args.agbno, len, 0);
 error0:
-	up_read(&args.mp->m_peraglock);
+	xfs_perag_put(args.pag);
 	return error;
 }
 
@@ -2486,15 +2488,15 @@ xfs_alloc_mark_busy(xfs_trans_t *tp,
 		    xfs_agblock_t bno,
 		    xfs_extlen_t len)
 {
-	xfs_mount_t		*mp;
 	xfs_perag_busy_t	*bsy;
+	struct xfs_perag	*pag;
 	int			n;
 
-	mp = tp->t_mountp;
-	spin_lock(&mp->m_perag[agno].pagb_lock);
+	pag = xfs_perag_get(tp->t_mountp, agno);
+	spin_lock(&pag->pagb_lock);
 
 	/* search pagb_list for an open slot */
-	for (bsy = mp->m_perag[agno].pagb_list, n = 0;
+	for (bsy = pag->pagb_list, n = 0;
 	     n < XFS_PAGB_NUM_SLOTS;
 	     bsy++, n++) {
 		if (bsy->busy_tp == NULL) {
@@ -2502,11 +2504,11 @@ xfs_alloc_mark_busy(xfs_trans_t *tp,
 		}
 	}
 
-	trace_xfs_alloc_busy(mp, agno, bno, len, n);
+	trace_xfs_alloc_busy(tp->t_mountp, agno, bno, len, n);
 
 	if (n < XFS_PAGB_NUM_SLOTS) {
-		bsy = &mp->m_perag[agno].pagb_list[n];
-		mp->m_perag[agno].pagb_count++;
+		bsy = &pag->pagb_list[n];
+		pag->pagb_count++;
 		bsy->busy_start = bno;
 		bsy->busy_length = len;
 		bsy->busy_tp = tp;
@@ -2521,7 +2523,8 @@ xfs_alloc_mark_busy(xfs_trans_t *tp,
 		xfs_trans_set_sync(tp);
 	}
 
-	spin_unlock(&mp->m_perag[agno].pagb_lock);
+	spin_unlock(&pag->pagb_lock);
+	xfs_perag_put(pag);
 }
 
 void
@@ -2529,24 +2532,23 @@ xfs_alloc_clear_busy(xfs_trans_t *tp,
 		     xfs_agnumber_t agno,
 		     int idx)
 {
-	xfs_mount_t		*mp;
+	struct xfs_perag	*pag;
 	xfs_perag_busy_t	*list;
 
-	mp = tp->t_mountp;
-
-	spin_lock(&mp->m_perag[agno].pagb_lock);
-	list = mp->m_perag[agno].pagb_list;
-
 	ASSERT(idx < XFS_PAGB_NUM_SLOTS);
+	pag = xfs_perag_get(tp->t_mountp, agno);
+	spin_lock(&pag->pagb_lock);
+	list = pag->pagb_list;
 
-	trace_xfs_alloc_unbusy(mp, agno, idx, list[idx].busy_tp == tp);
+	trace_xfs_alloc_unbusy(tp->t_mountp, agno, idx, list[idx].busy_tp == tp);
 
 	if (list[idx].busy_tp == tp) {
 		list[idx].busy_tp = NULL;
-		mp->m_perag[agno].pagb_count--;
+		pag->pagb_count--;
 	}
 
-	spin_unlock(&mp->m_perag[agno].pagb_lock);
+	spin_unlock(&pag->pagb_lock);
+	xfs_perag_put(pag);
 }
 
 
@@ -2560,17 +2562,15 @@ xfs_alloc_search_busy(xfs_trans_t *tp,
 		    xfs_agblock_t bno,
 		    xfs_extlen_t len)
 {
-	xfs_mount_t		*mp;
+	struct xfs_perag	*pag;
 	xfs_perag_busy_t	*bsy;
 	xfs_agblock_t		uend, bend;
 	xfs_lsn_t		lsn = 0;
 	int			cnt;
 
-	mp = tp->t_mountp;
-
-	spin_lock(&mp->m_perag[agno].pagb_lock);
-
-	uend = bno + len - 1;
+	pag = xfs_perag_get(tp->t_mountp, agno);
+	spin_lock(&pag->pagb_lock);
+	cnt = pag->pagb_count;
 
 	/*
 	 * search pagb_list for this slot, skipping open slots. We have to
@@ -2578,8 +2578,9 @@ xfs_alloc_search_busy(xfs_trans_t *tp,
 	 * we have to get the most recent LSN for the log force to push out
 	 * all the transactions that span the range.
 	 */
-	for (cnt = 0; cnt < mp->m_perag[agno].pagb_count; cnt++) {
-		bsy = &mp->m_perag[agno].pagb_list[cnt];
+	uend = bno + len - 1;
+	for (cnt = 0; cnt < pag->pagb_count; cnt++) {
+		bsy = &pag->pagb_list[cnt];
 		if (!bsy->busy_tp)
 			continue;
 
@@ -2591,7 +2592,8 @@ xfs_alloc_search_busy(xfs_trans_t *tp,
 		if (XFS_LSN_CMP(bsy->busy_tp->t_commit_lsn, lsn) > 0)
 			lsn = bsy->busy_tp->t_commit_lsn;
 	}
-	spin_unlock(&mp->m_perag[agno].pagb_lock);
+	spin_unlock(&pag->pagb_lock);
+	xfs_perag_put(pag);
 	trace_xfs_alloc_busysearch(tp->t_mountp, agno, bno, len, lsn);
 
 	/*
@@ -2599,5 +2601,5 @@ xfs_alloc_search_busy(xfs_trans_t *tp,
 	 * transaction that freed the block
 	 */
 	if (lsn)
-		xfs_log_force(mp, lsn, XFS_LOG_FORCE|XFS_LOG_SYNC);
+		xfs_log_force_lsn(tp->t_mountp, lsn, XFS_LOG_SYNC);
 }
diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c
index adbd9141aea1..b726e10d2c1c 100644
--- a/fs/xfs/xfs_alloc_btree.c
+++ b/fs/xfs/xfs_alloc_btree.c
@@ -61,12 +61,14 @@ xfs_allocbt_set_root(
 	struct xfs_agf		*agf = XFS_BUF_TO_AGF(agbp);
 	xfs_agnumber_t		seqno = be32_to_cpu(agf->agf_seqno);
 	int			btnum = cur->bc_btnum;
+	struct xfs_perag	*pag = xfs_perag_get(cur->bc_mp, seqno);
 
 	ASSERT(ptr->s != 0);
 
 	agf->agf_roots[btnum] = ptr->s;
 	be32_add_cpu(&agf->agf_levels[btnum], inc);
-	cur->bc_mp->m_perag[seqno].pagf_levels[btnum] += inc;
+	pag->pagf_levels[btnum] += inc;
+	xfs_perag_put(pag);
 
 	xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_ROOTS | XFS_AGF_LEVELS);
 }
@@ -150,6 +152,7 @@ xfs_allocbt_update_lastrec(
 {
 	struct xfs_agf		*agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp);
 	xfs_agnumber_t		seqno = be32_to_cpu(agf->agf_seqno);
+	struct xfs_perag	*pag;
 	__be32			len;
 	int			numrecs;
 
@@ -193,7 +196,9 @@ xfs_allocbt_update_lastrec(
 	}
 
 	agf->agf_longest = len;
-	cur->bc_mp->m_perag[seqno].pagf_longest = be32_to_cpu(len);
+	pag = xfs_perag_get(cur->bc_mp, seqno);
+	pag->pagf_longest = be32_to_cpu(len);
+	xfs_perag_put(pag);
 	xfs_alloc_log_agf(cur->bc_tp, cur->bc_private.a.agbp, XFS_AGF_LONGEST);
 }
 
diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c
index e953b6cfb2a8..b9c196a53c42 100644
--- a/fs/xfs/xfs_attr.c
+++ b/fs/xfs/xfs_attr.c
@@ -93,12 +93,12 @@ STATIC int xfs_attr_rmtval_remove(xfs_da_args_t *args);
 STATIC int
 xfs_attr_name_to_xname(
 	struct xfs_name	*xname,
-	const char	*aname)
+	const unsigned char *aname)
 {
 	if (!aname)
 		return EINVAL;
 	xname->name = aname;
-	xname->len = strlen(aname);
+	xname->len = strlen((char *)aname);
 	if (xname->len >= MAXNAMELEN)
 		return EFAULT;		/* match IRIX behaviour */
 
@@ -124,7 +124,7 @@ STATIC int
 xfs_attr_get_int(
 	struct xfs_inode	*ip,
 	struct xfs_name		*name,
-	char			*value,
+	unsigned char		*value,
 	int			*valuelenp,
 	int			flags)
 {
@@ -171,8 +171,8 @@ xfs_attr_get_int(
 int
 xfs_attr_get(
 	xfs_inode_t	*ip,
-	const char	*name,
-	char		*value,
+	const unsigned char *name,
+	unsigned char	*value,
 	int		*valuelenp,
 	int		flags)
 {
@@ -197,7 +197,7 @@ xfs_attr_get(
 /*
  * Calculate how many blocks we need for the new attribute,
  */
-int
+STATIC int
 xfs_attr_calc_size(
 	struct xfs_inode 	*ip,
 	int			namelen,
@@ -235,8 +235,12 @@ xfs_attr_calc_size(
 }
 
 STATIC int
-xfs_attr_set_int(xfs_inode_t *dp, struct xfs_name *name,
-		char *value, int valuelen, int flags)
+xfs_attr_set_int(
+	struct xfs_inode *dp,
+	struct xfs_name	*name,
+	unsigned char	*value,
+	int		valuelen,
+	int		flags)
 {
 	xfs_da_args_t	args;
 	xfs_fsblock_t	firstblock;
@@ -452,8 +456,8 @@ out:
 int
 xfs_attr_set(
 	xfs_inode_t	*dp,
-	const char	*name,
-	char		*value,
+	const unsigned char *name,
+	unsigned char	*value,
 	int		valuelen,
 	int		flags)
 {
@@ -600,7 +604,7 @@ out:
 int
 xfs_attr_remove(
 	xfs_inode_t	*dp,
-	const char	*name,
+	const unsigned char *name,
 	int		flags)
 {
 	int		error;
@@ -669,9 +673,13 @@ xfs_attr_list_int(xfs_attr_list_context_t *context)
  */
 /*ARGSUSED*/
 STATIC int
-xfs_attr_put_listent(xfs_attr_list_context_t *context, int flags,
-		     char *name, int namelen,
-		     int valuelen, char *value)
+xfs_attr_put_listent(
+	xfs_attr_list_context_t *context,
+	int		flags,
+	unsigned char	*name,
+	int		namelen,
+	int		valuelen,
+	unsigned char	*value)
 {
 	struct attrlist *alist = (struct attrlist *)context->alist;
 	attrlist_ent_t *aep;
@@ -1980,7 +1988,7 @@ xfs_attr_rmtval_get(xfs_da_args_t *args)
 	xfs_bmbt_irec_t map[ATTR_RMTVALUE_MAPSIZE];
 	xfs_mount_t *mp;
 	xfs_daddr_t dblkno;
-	xfs_caddr_t dst;
+	void *dst;
 	xfs_buf_t *bp;
 	int nmap, error, tmp, valuelen, blkcnt, i;
 	xfs_dablk_t lblkno;
@@ -2007,15 +2015,14 @@ xfs_attr_rmtval_get(xfs_da_args_t *args)
 			dblkno = XFS_FSB_TO_DADDR(mp, map[i].br_startblock);
 			blkcnt = XFS_FSB_TO_BB(mp, map[i].br_blockcount);
 			error = xfs_read_buf(mp, mp->m_ddev_targp, dblkno,
-					     blkcnt,
-					     XFS_BUF_LOCK | XBF_DONT_BLOCK,
+					     blkcnt, XBF_LOCK | XBF_DONT_BLOCK,
 					     &bp);
 			if (error)
 				return(error);
 
 			tmp = (valuelen < XFS_BUF_SIZE(bp))
 				? valuelen : XFS_BUF_SIZE(bp);
-			xfs_biomove(bp, 0, tmp, dst, XFS_B_READ);
+			xfs_biomove(bp, 0, tmp, dst, XBF_READ);
 			xfs_buf_relse(bp);
 			dst += tmp;
 			valuelen -= tmp;
@@ -2039,7 +2046,7 @@ xfs_attr_rmtval_set(xfs_da_args_t *args)
 	xfs_inode_t *dp;
 	xfs_bmbt_irec_t map;
 	xfs_daddr_t dblkno;
-	xfs_caddr_t src;
+	void *src;
 	xfs_buf_t *bp;
 	xfs_dablk_t lblkno;
 	int blkcnt, valuelen, nmap, error, tmp, committed;
@@ -2141,13 +2148,13 @@ xfs_attr_rmtval_set(xfs_da_args_t *args)
 		blkcnt = XFS_FSB_TO_BB(mp, map.br_blockcount);
 
 		bp = xfs_buf_get(mp->m_ddev_targp, dblkno, blkcnt,
-				 XFS_BUF_LOCK | XBF_DONT_BLOCK);
+				 XBF_LOCK | XBF_DONT_BLOCK);
 		ASSERT(bp);
 		ASSERT(!XFS_BUF_GETERROR(bp));
 
 		tmp = (valuelen < XFS_BUF_SIZE(bp)) ? valuelen :
 							XFS_BUF_SIZE(bp);
-		xfs_biomove(bp, 0, tmp, src, XFS_B_WRITE);
+		xfs_biomove(bp, 0, tmp, src, XBF_WRITE);
 		if (tmp < XFS_BUF_SIZE(bp))
 			xfs_biozero(bp, tmp, XFS_BUF_SIZE(bp) - tmp);
 		if ((error = xfs_bwrite(mp, bp))) {/* GROT: NOTE: synchronous write */
@@ -2208,8 +2215,7 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args)
 		/*
 		 * If the "remote" value is in the cache, remove it.
 		 */
-		bp = xfs_incore(mp->m_ddev_targp, dblkno, blkcnt,
-				XFS_INCORE_TRYLOCK);
+		bp = xfs_incore(mp->m_ddev_targp, dblkno, blkcnt, XBF_TRYLOCK);
 		if (bp) {
 			XFS_BUF_STALE(bp);
 			XFS_BUF_UNDELAYWRITE(bp);
diff --git a/fs/xfs/xfs_attr.h b/fs/xfs/xfs_attr.h
index 59b410ce69a1..e920d68ef509 100644
--- a/fs/xfs/xfs_attr.h
+++ b/fs/xfs/xfs_attr.h
@@ -113,7 +113,7 @@ typedef struct attrlist_cursor_kern {
 
 
 typedef int (*put_listent_func_t)(struct xfs_attr_list_context *, int,
-				      char *, int, int, char *);
+			      unsigned char *, int, int, unsigned char *);
 
 typedef struct xfs_attr_list_context {
 	struct xfs_inode		*dp;		/* inode */
@@ -139,7 +139,6 @@ typedef struct xfs_attr_list_context {
 /*
  * Overall external interface routines.
  */
-int xfs_attr_calc_size(struct xfs_inode *, int, int, int *);
 int xfs_attr_inactive(struct xfs_inode *dp);
 int xfs_attr_rmtval_get(struct xfs_da_args *args);
 int xfs_attr_list_int(struct xfs_attr_list_context *);
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index baf41b5af756..a90ce74fc256 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -521,11 +521,11 @@ xfs_attr_shortform_to_leaf(xfs_da_args_t *args)
 
 	sfe = &sf->list[0];
 	for (i = 0; i < sf->hdr.count; i++) {
-		nargs.name = (char *)sfe->nameval;
+		nargs.name = sfe->nameval;
 		nargs.namelen = sfe->namelen;
-		nargs.value = (char *)&sfe->nameval[nargs.namelen];
+		nargs.value = &sfe->nameval[nargs.namelen];
 		nargs.valuelen = sfe->valuelen;
-		nargs.hashval = xfs_da_hashname((char *)sfe->nameval,
+		nargs.hashval = xfs_da_hashname(sfe->nameval,
 						sfe->namelen);
 		nargs.flags = XFS_ATTR_NSP_ONDISK_TO_ARGS(sfe->flags);
 		error = xfs_attr_leaf_lookup_int(bp, &nargs); /* set a->index */
@@ -612,10 +612,10 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context)
 		for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
 			error = context->put_listent(context,
 					   sfe->flags,
-					   (char *)sfe->nameval,
+					   sfe->nameval,
 					   (int)sfe->namelen,
 					   (int)sfe->valuelen,
-					   (char*)&sfe->nameval[sfe->namelen]);
+					   &sfe->nameval[sfe->namelen]);
 
 			/*
 			 * Either search callback finished early or
@@ -659,8 +659,8 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context)
 		}
 
 		sbp->entno = i;
-		sbp->hash = xfs_da_hashname((char *)sfe->nameval, sfe->namelen);
-		sbp->name = (char *)sfe->nameval;
+		sbp->hash = xfs_da_hashname(sfe->nameval, sfe->namelen);
+		sbp->name = sfe->nameval;
 		sbp->namelen = sfe->namelen;
 		/* These are bytes, and both on-disk, don't endian-flip */
 		sbp->valuelen = sfe->valuelen;
@@ -818,9 +818,9 @@ xfs_attr_leaf_to_shortform(xfs_dabuf_t *bp, xfs_da_args_t *args, int forkoff)
 			continue;
 		ASSERT(entry->flags & XFS_ATTR_LOCAL);
 		name_loc = xfs_attr_leaf_name_local(leaf, i);
-		nargs.name = (char *)name_loc->nameval;
+		nargs.name = name_loc->nameval;
 		nargs.namelen = name_loc->namelen;
-		nargs.value = (char *)&name_loc->nameval[nargs.namelen];
+		nargs.value = &name_loc->nameval[nargs.namelen];
 		nargs.valuelen = be16_to_cpu(name_loc->valuelen);
 		nargs.hashval = be32_to_cpu(entry->hashval);
 		nargs.flags = XFS_ATTR_NSP_ONDISK_TO_ARGS(entry->flags);
@@ -2370,10 +2370,10 @@ xfs_attr_leaf_list_int(xfs_dabuf_t *bp, xfs_attr_list_context_t *context)
 
 			retval = context->put_listent(context,
 						entry->flags,
-						(char *)name_loc->nameval,
+						name_loc->nameval,
 						(int)name_loc->namelen,
 						be16_to_cpu(name_loc->valuelen),
-						(char *)&name_loc->nameval[name_loc->namelen]);
+						&name_loc->nameval[name_loc->namelen]);
 			if (retval)
 				return retval;
 		} else {
@@ -2397,15 +2397,15 @@ xfs_attr_leaf_list_int(xfs_dabuf_t *bp, xfs_attr_list_context_t *context)
 					return retval;
 				retval = context->put_listent(context,
 						entry->flags,
-						(char *)name_rmt->name,
+						name_rmt->name,
 						(int)name_rmt->namelen,
 						valuelen,
-						(char*)args.value);
+						args.value);
 				kmem_free(args.value);
 			} else {
 				retval = context->put_listent(context,
 						entry->flags,
-						(char *)name_rmt->name,
+						name_rmt->name,
 						(int)name_rmt->namelen,
 						valuelen,
 						NULL);
@@ -2950,7 +2950,7 @@ xfs_attr_leaf_freextent(xfs_trans_t **trans, xfs_inode_t *dp,
 						map.br_blockcount);
 			bp = xfs_trans_get_buf(*trans,
 					dp->i_mount->m_ddev_targp,
-					dblkno, dblkcnt, XFS_BUF_LOCK);
+					dblkno, dblkcnt, XBF_LOCK);
 			xfs_trans_binval(*trans, bp);
 			/*
 			 * Roll to next transaction.
diff --git a/fs/xfs/xfs_attr_sf.h b/fs/xfs/xfs_attr_sf.h
index 76ab7b0cbb3a..919756e3ba53 100644
--- a/fs/xfs/xfs_attr_sf.h
+++ b/fs/xfs/xfs_attr_sf.h
@@ -52,7 +52,7 @@ typedef struct xfs_attr_sf_sort {
 	__uint8_t	valuelen;	/* length of value */
 	__uint8_t	flags;		/* flags bits (see xfs_attr_leaf.h) */
 	xfs_dahash_t	hash;		/* this entry's hash value */
-	char		*name;		/* name value, pointer into buffer */
+	unsigned char	*name;		/* name value, pointer into buffer */
 } xfs_attr_sf_sort_t;
 
 #define XFS_ATTR_SF_ENTSIZE_BYNAME(nlen,vlen)	/* space name/value uses */ \
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index 98251cdc52aa..5c11e4d17010 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -2550,22 +2550,134 @@ xfs_bmap_rtalloc(
 }
 
 STATIC int
+xfs_bmap_btalloc_nullfb(
+	struct xfs_bmalloca	*ap,
+	struct xfs_alloc_arg	*args,
+	xfs_extlen_t		*blen)
+{
+	struct xfs_mount	*mp = ap->ip->i_mount;
+	struct xfs_perag	*pag;
+	xfs_agnumber_t		ag, startag;
+	int			notinit = 0;
+	int			error;
+
+	if (ap->userdata && xfs_inode_is_filestream(ap->ip))
+		args->type = XFS_ALLOCTYPE_NEAR_BNO;
+	else
+		args->type = XFS_ALLOCTYPE_START_BNO;
+	args->total = ap->total;
+
+	/*
+	 * Search for an allocation group with a single extent large enough
+	 * for the request.  If one isn't found, then adjust the minimum
+	 * allocation size to the largest space found.
+	 */
+	startag = ag = XFS_FSB_TO_AGNO(mp, args->fsbno);
+	if (startag == NULLAGNUMBER)
+		startag = ag = 0;
+
+	pag = xfs_perag_get(mp, ag);
+	while (*blen < ap->alen) {
+		if (!pag->pagf_init) {
+			error = xfs_alloc_pagf_init(mp, args->tp, ag,
+						    XFS_ALLOC_FLAG_TRYLOCK);
+			if (error) {
+				xfs_perag_put(pag);
+				return error;
+			}
+		}
+
+		/*
+		 * See xfs_alloc_fix_freelist...
+		 */
+		if (pag->pagf_init) {
+			xfs_extlen_t	longest;
+			longest = xfs_alloc_longest_free_extent(mp, pag);
+			if (*blen < longest)
+				*blen = longest;
+		} else
+			notinit = 1;
+
+		if (xfs_inode_is_filestream(ap->ip)) {
+			if (*blen >= ap->alen)
+				break;
+
+			if (ap->userdata) {
+				/*
+				 * If startag is an invalid AG, we've
+				 * come here once before and
+				 * xfs_filestream_new_ag picked the
+				 * best currently available.
+				 *
+				 * Don't continue looping, since we
+				 * could loop forever.
+				 */
+				if (startag == NULLAGNUMBER)
+					break;
+
+				error = xfs_filestream_new_ag(ap, &ag);
+				xfs_perag_put(pag);
+				if (error)
+					return error;
+
+				/* loop again to set 'blen'*/
+				startag = NULLAGNUMBER;
+				pag = xfs_perag_get(mp, ag);
+				continue;
+			}
+		}
+		if (++ag == mp->m_sb.sb_agcount)
+			ag = 0;
+		if (ag == startag)
+			break;
+		xfs_perag_put(pag);
+		pag = xfs_perag_get(mp, ag);
+	}
+	xfs_perag_put(pag);
+
+	/*
+	 * Since the above loop did a BUF_TRYLOCK, it is
+	 * possible that there is space for this request.
+	 */
+	if (notinit || *blen < ap->minlen)
+		args->minlen = ap->minlen;
+	/*
+	 * If the best seen length is less than the request
+	 * length, use the best as the minimum.
+	 */
+	else if (*blen < ap->alen)
+		args->minlen = *blen;
+	/*
+	 * Otherwise we've seen an extent as big as alen,
+	 * use that as the minimum.
+	 */
+	else
+		args->minlen = ap->alen;
+
+	/*
+	 * set the failure fallback case to look in the selected
+	 * AG as the stream may have moved.
+	 */
+	if (xfs_inode_is_filestream(ap->ip))
+		ap->rval = args->fsbno = XFS_AGB_TO_FSB(mp, ag, 0);
+
+	return 0;
+}
+
+STATIC int
 xfs_bmap_btalloc(
 	xfs_bmalloca_t	*ap)		/* bmap alloc argument struct */
 {
 	xfs_mount_t	*mp;		/* mount point structure */
 	xfs_alloctype_t	atype = 0;	/* type for allocation routines */
 	xfs_extlen_t	align;		/* minimum allocation alignment */
-	xfs_agnumber_t	ag;
 	xfs_agnumber_t	fb_agno;	/* ag number of ap->firstblock */
-	xfs_agnumber_t	startag;
+	xfs_agnumber_t	ag;
 	xfs_alloc_arg_t	args;
 	xfs_extlen_t	blen;
 	xfs_extlen_t	nextminlen = 0;
-	xfs_perag_t	*pag;
 	int		nullfb;		/* true if ap->firstblock isn't set */
 	int		isaligned;
-	int		notinit;
 	int		tryagain;
 	int		error;
 
@@ -2612,102 +2724,9 @@ xfs_bmap_btalloc(
 	args.firstblock = ap->firstblock;
 	blen = 0;
 	if (nullfb) {
-		if (ap->userdata && xfs_inode_is_filestream(ap->ip))
-			args.type = XFS_ALLOCTYPE_NEAR_BNO;
-		else
-			args.type = XFS_ALLOCTYPE_START_BNO;
-		args.total = ap->total;
-
-		/*
-		 * Search for an allocation group with a single extent
-		 * large enough for the request.
-		 *
-		 * If one isn't found, then adjust the minimum allocation
-		 * size to the largest space found.
-		 */
-		startag = ag = XFS_FSB_TO_AGNO(mp, args.fsbno);
-		if (startag == NULLAGNUMBER)
-			startag = ag = 0;
-		notinit = 0;
-		down_read(&mp->m_peraglock);
-		while (blen < ap->alen) {
-			pag = &mp->m_perag[ag];
-			if (!pag->pagf_init &&
-			    (error = xfs_alloc_pagf_init(mp, args.tp,
-				    ag, XFS_ALLOC_FLAG_TRYLOCK))) {
-				up_read(&mp->m_peraglock);
-				return error;
-			}
-			/*
-			 * See xfs_alloc_fix_freelist...
-			 */
-			if (pag->pagf_init) {
-				xfs_extlen_t	longest;
-				longest = xfs_alloc_longest_free_extent(mp, pag);
-				if (blen < longest)
-					blen = longest;
-			} else
-				notinit = 1;
-
-			if (xfs_inode_is_filestream(ap->ip)) {
-				if (blen >= ap->alen)
-					break;
-
-				if (ap->userdata) {
-					/*
-					 * If startag is an invalid AG, we've
-					 * come here once before and
-					 * xfs_filestream_new_ag picked the
-					 * best currently available.
-					 *
-					 * Don't continue looping, since we
-					 * could loop forever.
-					 */
-					if (startag == NULLAGNUMBER)
-						break;
-
-					error = xfs_filestream_new_ag(ap, &ag);
-					if (error) {
-						up_read(&mp->m_peraglock);
-						return error;
-					}
-
-					/* loop again to set 'blen'*/
-					startag = NULLAGNUMBER;
-					continue;
-				}
-			}
-			if (++ag == mp->m_sb.sb_agcount)
-				ag = 0;
-			if (ag == startag)
-				break;
-		}
-		up_read(&mp->m_peraglock);
-		/*
-		 * Since the above loop did a BUF_TRYLOCK, it is
-		 * possible that there is space for this request.
-		 */
-		if (notinit || blen < ap->minlen)
-			args.minlen = ap->minlen;
-		/*
-		 * If the best seen length is less than the request
-		 * length, use the best as the minimum.
-		 */
-		else if (blen < ap->alen)
-			args.minlen = blen;
-		/*
-		 * Otherwise we've seen an extent as big as alen,
-		 * use that as the minimum.
-		 */
-		else
-			args.minlen = ap->alen;
-
-		/*
-		 * set the failure fallback case to look in the selected
-		 * AG as the stream may have moved.
-		 */
-		if (xfs_inode_is_filestream(ap->ip))
-			ap->rval = args.fsbno = XFS_AGB_TO_FSB(mp, ag, 0);
+		error = xfs_bmap_btalloc_nullfb(ap, &args, &blen);
+		if (error)
+			return error;
 	} else if (ap->low) {
 		if (xfs_inode_is_filestream(ap->ip))
 			args.type = XFS_ALLOCTYPE_FIRST_AG;
@@ -4470,7 +4489,7 @@ xfs_bmapi(
 	xfs_fsblock_t	abno;		/* allocated block number */
 	xfs_extlen_t	alen;		/* allocated extent length */
 	xfs_fileoff_t	aoff;		/* allocated file offset */
-	xfs_bmalloca_t	bma;		/* args for xfs_bmap_alloc */
+	xfs_bmalloca_t	bma = { 0 };	/* args for xfs_bmap_alloc */
 	xfs_btree_cur_t	*cur;		/* bmap btree cursor */
 	xfs_fileoff_t	end;		/* end of mapped file region */
 	int		eof;		/* we've hit the end of extents */
diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c
index 38751d5fac6f..416e47e54b83 100644
--- a/fs/xfs/xfs_bmap_btree.c
+++ b/fs/xfs/xfs_bmap_btree.c
@@ -334,7 +334,7 @@ xfs_bmbt_disk_set_allf(
 /*
  * Set all the fields in a bmap extent record from the uncompressed form.
  */
-void
+STATIC void
 xfs_bmbt_disk_set_all(
 	xfs_bmbt_rec_t	*r,
 	xfs_bmbt_irec_t *s)
diff --git a/fs/xfs/xfs_bmap_btree.h b/fs/xfs/xfs_bmap_btree.h
index cf07ca7c22e7..0e66c4ea0f85 100644
--- a/fs/xfs/xfs_bmap_btree.h
+++ b/fs/xfs/xfs_bmap_btree.h
@@ -223,7 +223,6 @@ extern void xfs_bmbt_set_startblock(xfs_bmbt_rec_host_t *r, xfs_fsblock_t v);
 extern void xfs_bmbt_set_startoff(xfs_bmbt_rec_host_t *r, xfs_fileoff_t v);
 extern void xfs_bmbt_set_state(xfs_bmbt_rec_host_t *r, xfs_exntst_t v);
 
-extern void xfs_bmbt_disk_set_all(xfs_bmbt_rec_t *r, xfs_bmbt_irec_t *s);
 extern void xfs_bmbt_disk_set_allf(xfs_bmbt_rec_t *r, xfs_fileoff_t o,
 			xfs_fsblock_t b, xfs_filblks_t c, xfs_exntst_t v);
 
diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c
index 36a0992dd669..96be4b0f2496 100644
--- a/fs/xfs/xfs_btree.c
+++ b/fs/xfs/xfs_btree.c
@@ -977,7 +977,7 @@ xfs_btree_get_buf_block(
 	xfs_daddr_t		d;
 
 	/* need to sort out how callers deal with failures first */
-	ASSERT(!(flags & XFS_BUF_TRYLOCK));
+	ASSERT(!(flags & XBF_TRYLOCK));
 
 	d = xfs_btree_ptr_to_daddr(cur, ptr);
 	*bpp = xfs_trans_get_buf(cur->bc_tp, mp->m_ddev_targp, d,
@@ -1008,7 +1008,7 @@ xfs_btree_read_buf_block(
 	int			error;
 
 	/* need to sort out how callers deal with failures first */
-	ASSERT(!(flags & XFS_BUF_TRYLOCK));
+	ASSERT(!(flags & XBF_TRYLOCK));
 
 	d = xfs_btree_ptr_to_daddr(cur, ptr);
 	error = xfs_trans_read_buf(mp, cur->bc_tp, mp->m_ddev_targp, d,
diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index a30f7e9eb2b9..f3c49e69eab9 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -250,7 +250,7 @@ xfs_buf_item_format(
 		       ((bip->bli_format.blf_map_size - 1) * sizeof(uint)));
 	vecp->i_addr = (xfs_caddr_t)&bip->bli_format;
 	vecp->i_len = base_size;
-	XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_BFORMAT);
+	vecp->i_type = XLOG_REG_TYPE_BFORMAT;
 	vecp++;
 	nvecs = 1;
 
@@ -297,14 +297,14 @@ xfs_buf_item_format(
 			buffer_offset = first_bit * XFS_BLI_CHUNK;
 			vecp->i_addr = xfs_buf_offset(bp, buffer_offset);
 			vecp->i_len = nbits * XFS_BLI_CHUNK;
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_BCHUNK);
+			vecp->i_type = XLOG_REG_TYPE_BCHUNK;
 			nvecs++;
 			break;
 		} else if (next_bit != last_bit + 1) {
 			buffer_offset = first_bit * XFS_BLI_CHUNK;
 			vecp->i_addr = xfs_buf_offset(bp, buffer_offset);
 			vecp->i_len = nbits * XFS_BLI_CHUNK;
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_BCHUNK);
+			vecp->i_type = XLOG_REG_TYPE_BCHUNK;
 			nvecs++;
 			vecp++;
 			first_bit = next_bit;
@@ -316,7 +316,7 @@ xfs_buf_item_format(
 			buffer_offset = first_bit * XFS_BLI_CHUNK;
 			vecp->i_addr = xfs_buf_offset(bp, buffer_offset);
 			vecp->i_len = nbits * XFS_BLI_CHUNK;
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_BCHUNK);
+			vecp->i_type = XLOG_REG_TYPE_BCHUNK;
 /* You would think we need to bump the nvecs here too, but we do not
  * this number is used by recovery, and it gets confused by the boundary
  * split here
@@ -467,8 +467,10 @@ xfs_buf_item_unpin_remove(
 /*
  * This is called to attempt to lock the buffer associated with this
  * buf log item.  Don't sleep on the buffer lock.  If we can't get
- * the lock right away, return 0.  If we can get the lock, pull the
- * buffer from the free list, mark it busy, and return 1.
+ * the lock right away, return 0.  If we can get the lock, take a
+ * reference to the buffer. If this is a delayed write buffer that
+ * needs AIL help to be written back, invoke the pushbuf routine
+ * rather than the normal success path.
  */
 STATIC uint
 xfs_buf_item_trylock(
@@ -477,24 +479,18 @@ xfs_buf_item_trylock(
 	xfs_buf_t	*bp;
 
 	bp = bip->bli_buf;
-
-	if (XFS_BUF_ISPINNED(bp)) {
+	if (XFS_BUF_ISPINNED(bp))
 		return XFS_ITEM_PINNED;
-	}
-
-	if (!XFS_BUF_CPSEMA(bp)) {
+	if (!XFS_BUF_CPSEMA(bp))
 		return XFS_ITEM_LOCKED;
-	}
 
-	/*
-	 * Remove the buffer from the free list.  Only do this
-	 * if it's on the free list.  Private buffers like the
-	 * superblock buffer are not.
-	 */
+	/* take a reference to the buffer.  */
 	XFS_BUF_HOLD(bp);
 
 	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
 	trace_xfs_buf_item_trylock(bip);
+	if (XFS_BUF_ISDELAYWRITE(bp))
+		return XFS_ITEM_PUSHBUF;
 	return XFS_ITEM_SUCCESS;
 }
 
@@ -626,11 +622,9 @@ xfs_buf_item_committed(
 }
 
 /*
- * This is called to asynchronously write the buffer associated with this
- * buf log item out to disk. The buffer will already have been locked by
- * a successful call to xfs_buf_item_trylock().  If the buffer still has
- * B_DELWRI set, then get it going out to disk with a call to bawrite().
- * If not, then just release the buffer.
+ * The buffer is locked, but is not a delayed write buffer. This happens
+ * if we race with IO completion and hence we don't want to try to write it
+ * again. Just release the buffer.
  */
 STATIC void
 xfs_buf_item_push(
@@ -642,17 +636,29 @@ xfs_buf_item_push(
 	trace_xfs_buf_item_push(bip);
 
 	bp = bip->bli_buf;
+	ASSERT(!XFS_BUF_ISDELAYWRITE(bp));
+	xfs_buf_relse(bp);
+}
 
-	if (XFS_BUF_ISDELAYWRITE(bp)) {
-		int	error;
-		error = xfs_bawrite(bip->bli_item.li_mountp, bp);
-		if (error)
-			xfs_fs_cmn_err(CE_WARN, bip->bli_item.li_mountp,
-			"xfs_buf_item_push: pushbuf error %d on bip %p, bp %p",
-					error, bip, bp);
-	} else {
-		xfs_buf_relse(bp);
-	}
+/*
+ * The buffer is locked and is a delayed write buffer. Promote the buffer
+ * in the delayed write queue as the caller knows that they must invoke
+ * the xfsbufd to get this buffer written. We have to unlock the buffer
+ * to allow the xfsbufd to write it, too.
+ */
+STATIC void
+xfs_buf_item_pushbuf(
+	xfs_buf_log_item_t	*bip)
+{
+	xfs_buf_t	*bp;
+
+	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
+	trace_xfs_buf_item_pushbuf(bip);
+
+	bp = bip->bli_buf;
+	ASSERT(XFS_BUF_ISDELAYWRITE(bp));
+	xfs_buf_delwri_promote(bp);
+	xfs_buf_relse(bp);
 }
 
 /* ARGSUSED */
@@ -677,7 +683,7 @@ static struct xfs_item_ops xfs_buf_item_ops = {
 	.iop_committed	= (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t))
 					xfs_buf_item_committed,
 	.iop_push	= (void(*)(xfs_log_item_t*))xfs_buf_item_push,
-	.iop_pushbuf	= NULL,
+	.iop_pushbuf	= (void(*)(xfs_log_item_t*))xfs_buf_item_pushbuf,
 	.iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t))
 					xfs_buf_item_committing
 };
diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c
index c0c8869115b1..0ca556b4bf31 100644
--- a/fs/xfs/xfs_da_btree.c
+++ b/fs/xfs/xfs_da_btree.c
@@ -1534,8 +1534,8 @@ xfs_da_hashname(const __uint8_t *name, int namelen)
 enum xfs_dacmp
 xfs_da_compname(
 	struct xfs_da_args *args,
-	const char 	*name,
-	int 		len)
+	const unsigned char *name,
+	int		len)
 {
 	return (args->namelen == len && memcmp(args->name, name, len) == 0) ?
 					XFS_CMP_EXACT : XFS_CMP_DIFFERENT;
diff --git a/fs/xfs/xfs_da_btree.h b/fs/xfs/xfs_da_btree.h
index 30cd08f56a3a..fe9f5a8c1d2a 100644
--- a/fs/xfs/xfs_da_btree.h
+++ b/fs/xfs/xfs_da_btree.h
@@ -209,7 +209,8 @@ typedef struct xfs_da_state {
  */
 struct xfs_nameops {
 	xfs_dahash_t	(*hashname)(struct xfs_name *);
-	enum xfs_dacmp	(*compname)(struct xfs_da_args *, const char *, int);
+	enum xfs_dacmp	(*compname)(struct xfs_da_args *,
+					const unsigned char *, int);
 };
 
 
@@ -260,7 +261,7 @@ int	xfs_da_shrink_inode(xfs_da_args_t *args, xfs_dablk_t dead_blkno,
 
 uint xfs_da_hashname(const __uint8_t *name_string, int name_length);
 enum xfs_dacmp xfs_da_compname(struct xfs_da_args *args,
-				const char *name, int len);
+				const unsigned char *name, int len);
 
 
 xfs_da_state_t *xfs_da_state_alloc(void);
diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c
index 84ca1cf16a1e..cd27c9d6c71f 100644
--- a/fs/xfs/xfs_dfrag.c
+++ b/fs/xfs/xfs_dfrag.c
@@ -45,15 +45,21 @@
 #include "xfs_vnodeops.h"
 #include "xfs_trace.h"
 
+
+static int xfs_swap_extents(
+	xfs_inode_t	*ip,	/* target inode */
+	xfs_inode_t	*tip,	/* tmp inode */
+	xfs_swapext_t	*sxp);
+
 /*
- * Syssgi interface for swapext
+ * ioctl interface for swapext
  */
 int
 xfs_swapext(
 	xfs_swapext_t	*sxp)
 {
 	xfs_inode_t     *ip, *tip;
-	struct file	*file, *target_file;
+	struct file	*file, *tmp_file;
 	int		error = 0;
 
 	/* Pull information for the target fd */
@@ -68,46 +74,46 @@ xfs_swapext(
 		goto out_put_file;
 	}
 
-	target_file = fget((int)sxp->sx_fdtmp);
-	if (!target_file) {
+	tmp_file = fget((int)sxp->sx_fdtmp);
+	if (!tmp_file) {
 		error = XFS_ERROR(EINVAL);
 		goto out_put_file;
 	}
 
-	if (!(target_file->f_mode & FMODE_WRITE) ||
-	    (target_file->f_flags & O_APPEND)) {
+	if (!(tmp_file->f_mode & FMODE_WRITE) ||
+	    (tmp_file->f_flags & O_APPEND)) {
 		error = XFS_ERROR(EBADF);
-		goto out_put_target_file;
+		goto out_put_tmp_file;
 	}
 
 	if (IS_SWAPFILE(file->f_path.dentry->d_inode) ||
-	    IS_SWAPFILE(target_file->f_path.dentry->d_inode)) {
+	    IS_SWAPFILE(tmp_file->f_path.dentry->d_inode)) {
 		error = XFS_ERROR(EINVAL);
-		goto out_put_target_file;
+		goto out_put_tmp_file;
 	}
 
 	ip = XFS_I(file->f_path.dentry->d_inode);
-	tip = XFS_I(target_file->f_path.dentry->d_inode);
+	tip = XFS_I(tmp_file->f_path.dentry->d_inode);
 
 	if (ip->i_mount != tip->i_mount) {
 		error = XFS_ERROR(EINVAL);
-		goto out_put_target_file;
+		goto out_put_tmp_file;
 	}
 
 	if (ip->i_ino == tip->i_ino) {
 		error = XFS_ERROR(EINVAL);
-		goto out_put_target_file;
+		goto out_put_tmp_file;
 	}
 
 	if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
 		error = XFS_ERROR(EIO);
-		goto out_put_target_file;
+		goto out_put_tmp_file;
 	}
 
 	error = xfs_swap_extents(ip, tip, sxp);
 
- out_put_target_file:
-	fput(target_file);
+ out_put_tmp_file:
+	fput(tmp_file);
  out_put_file:
 	fput(file);
  out:
@@ -186,7 +192,7 @@ xfs_swap_extents_check_format(
 	return 0;
 }
 
-int
+static int
 xfs_swap_extents(
 	xfs_inode_t	*ip,	/* target inode */
 	xfs_inode_t	*tip,	/* tmp inode */
@@ -254,6 +260,9 @@ xfs_swap_extents(
 		goto out_unlock;
 	}
 
+	trace_xfs_swap_extent_before(ip, 0);
+	trace_xfs_swap_extent_before(tip, 1);
+
 	/* check inode formats now that data is flushed */
 	error = xfs_swap_extents_check_format(ip, tip);
 	if (error) {
@@ -421,6 +430,8 @@ xfs_swap_extents(
 
 	error = xfs_trans_commit(tp, XFS_TRANS_SWAPEXT);
 
+	trace_xfs_swap_extent_after(ip, 0);
+	trace_xfs_swap_extent_after(tip, 1);
 out:
 	kmem_free(tempifp);
 	return error;
diff --git a/fs/xfs/xfs_dfrag.h b/fs/xfs/xfs_dfrag.h
index 4f55a6306558..20bdd935c121 100644
--- a/fs/xfs/xfs_dfrag.h
+++ b/fs/xfs/xfs_dfrag.h
@@ -48,9 +48,6 @@ typedef struct xfs_swapext
  */
 int	xfs_swapext(struct xfs_swapext *sx);
 
-int	xfs_swap_extents(struct xfs_inode *ip, struct xfs_inode *tip,
-		struct xfs_swapext *sxp);
-
 #endif	/* __KERNEL__ */
 
 #endif	/* __XFS_DFRAG_H__ */
diff --git a/fs/xfs/xfs_dir2.c b/fs/xfs/xfs_dir2.c
index 93634a7e90e9..42520f041265 100644
--- a/fs/xfs/xfs_dir2.c
+++ b/fs/xfs/xfs_dir2.c
@@ -44,7 +44,7 @@
 #include "xfs_vnodeops.h"
 #include "xfs_trace.h"
 
-struct xfs_name xfs_name_dotdot = {"..", 2};
+struct xfs_name xfs_name_dotdot = { (unsigned char *)"..", 2};
 
 /*
  * ASCII case-insensitive (ie. A-Z) support for directories that was
@@ -66,8 +66,8 @@ xfs_ascii_ci_hashname(
 STATIC enum xfs_dacmp
 xfs_ascii_ci_compname(
 	struct xfs_da_args *args,
-	const char	*name,
-	int 		len)
+	const unsigned char *name,
+	int		len)
 {
 	enum xfs_dacmp	result;
 	int		i;
@@ -247,7 +247,7 @@ xfs_dir_createname(
 int
 xfs_dir_cilookup_result(
 	struct xfs_da_args *args,
-	const char	*name,
+	const unsigned char *name,
 	int		len)
 {
 	if (args->cmpresult == XFS_CMP_DIFFERENT)
diff --git a/fs/xfs/xfs_dir2.h b/fs/xfs/xfs_dir2.h
index 1d9ef96f33aa..74a3b1057685 100644
--- a/fs/xfs/xfs_dir2.h
+++ b/fs/xfs/xfs_dir2.h
@@ -100,7 +100,7 @@ extern int xfs_dir2_isleaf(struct xfs_trans *tp, struct xfs_inode *dp,
 extern int xfs_dir2_shrink_inode(struct xfs_da_args *args, xfs_dir2_db_t db,
 				struct xfs_dabuf *bp);
 
-extern int xfs_dir_cilookup_result(struct xfs_da_args *args, const char *name,
-				int len);
+extern int xfs_dir_cilookup_result(struct xfs_da_args *args,
+				const unsigned char *name, int len);
 
 #endif	/* __XFS_DIR2_H__ */
diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c
index ddc4ecc7807f..779a267b0a84 100644
--- a/fs/xfs/xfs_dir2_block.c
+++ b/fs/xfs/xfs_dir2_block.c
@@ -57,8 +57,8 @@ static xfs_dahash_t xfs_dir_hash_dot, xfs_dir_hash_dotdot;
 void
 xfs_dir_startup(void)
 {
-	xfs_dir_hash_dot = xfs_da_hashname(".", 1);
-	xfs_dir_hash_dotdot = xfs_da_hashname("..", 2);
+	xfs_dir_hash_dot = xfs_da_hashname((unsigned char *)".", 1);
+	xfs_dir_hash_dotdot = xfs_da_hashname((unsigned char *)"..", 2);
 }
 
 /*
@@ -513,8 +513,9 @@ xfs_dir2_block_getdents(
 		/*
 		 * If it didn't fit, set the final offset to here & return.
 		 */
-		if (filldir(dirent, dep->name, dep->namelen, cook & 0x7fffffff,
-			    be64_to_cpu(dep->inumber), DT_UNKNOWN)) {
+		if (filldir(dirent, (char *)dep->name, dep->namelen,
+			    cook & 0x7fffffff, be64_to_cpu(dep->inumber),
+			    DT_UNKNOWN)) {
 			*offset = cook & 0x7fffffff;
 			xfs_da_brelse(NULL, bp);
 			return 0;
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index 29f484c11b3a..e2d89854ec9e 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -1081,7 +1081,7 @@ xfs_dir2_leaf_getdents(
 		dep = (xfs_dir2_data_entry_t *)ptr;
 		length = xfs_dir2_data_entsize(dep->namelen);
 
-		if (filldir(dirent, dep->name, dep->namelen,
+		if (filldir(dirent, (char *)dep->name, dep->namelen,
 			    xfs_dir2_byte_to_dataptr(mp, curoff) & 0x7fffffff,
 			    be64_to_cpu(dep->inumber), DT_UNKNOWN))
 			break;
diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c
index ce6e355199b5..78fc4d9ae756 100644
--- a/fs/xfs/xfs_dir2_node.c
+++ b/fs/xfs/xfs_dir2_node.c
@@ -65,7 +65,7 @@ static int xfs_dir2_node_addname_int(xfs_da_args_t *args,
 /*
  * Log entries from a freespace block.
  */
-void
+STATIC void
 xfs_dir2_free_log_bests(
 	xfs_trans_t		*tp,		/* transaction pointer */
 	xfs_dabuf_t		*bp,		/* freespace buffer */
diff --git a/fs/xfs/xfs_dir2_node.h b/fs/xfs/xfs_dir2_node.h
index dde72db3d695..82dfe7147195 100644
--- a/fs/xfs/xfs_dir2_node.h
+++ b/fs/xfs/xfs_dir2_node.h
@@ -75,8 +75,6 @@ xfs_dir2_db_to_fdindex(struct xfs_mount *mp, xfs_dir2_db_t db)
 	return ((db) % XFS_DIR2_MAX_FREE_BESTS(mp));
 }
 
-extern void xfs_dir2_free_log_bests(struct xfs_trans *tp, struct xfs_dabuf *bp,
-				    int first, int last);
 extern int xfs_dir2_leaf_to_node(struct xfs_da_args *args,
 				 struct xfs_dabuf *lbp);
 extern xfs_dahash_t xfs_dir2_leafn_lasthash(struct xfs_dabuf *bp, int *count);
diff --git a/fs/xfs/xfs_dir2_sf.c b/fs/xfs/xfs_dir2_sf.c
index 9d4f17a69676..c1a5945d463a 100644
--- a/fs/xfs/xfs_dir2_sf.c
+++ b/fs/xfs/xfs_dir2_sf.c
@@ -782,7 +782,7 @@ xfs_dir2_sf_getdents(
 		}
 
 		ino = xfs_dir2_sf_get_inumber(sfp, xfs_dir2_sf_inumberp(sfep));
-		if (filldir(dirent, sfep->name, sfep->namelen,
+		if (filldir(dirent, (char *)sfep->name, sfep->namelen,
 			    off & 0x7fffffff, ino, DT_UNKNOWN)) {
 			*offset = off & 0x7fffffff;
 			return 0;
diff --git a/fs/xfs/xfs_extfree_item.c b/fs/xfs/xfs_extfree_item.c
index 05a4bdd4be39..6f35ed1b39b9 100644
--- a/fs/xfs/xfs_extfree_item.c
+++ b/fs/xfs/xfs_extfree_item.c
@@ -82,7 +82,7 @@ xfs_efi_item_format(xfs_efi_log_item_t	*efip,
 
 	log_vector->i_addr = (xfs_caddr_t)&(efip->efi_format);
 	log_vector->i_len = size;
-	XLOG_VEC_SET_TYPE(log_vector, XLOG_REG_TYPE_EFI_FORMAT);
+	log_vector->i_type = XLOG_REG_TYPE_EFI_FORMAT;
 	ASSERT(size >= sizeof(xfs_efi_log_format_t));
 }
 
@@ -406,7 +406,7 @@ xfs_efd_item_format(xfs_efd_log_item_t	*efdp,
 
 	log_vector->i_addr = (xfs_caddr_t)&(efdp->efd_format);
 	log_vector->i_len = size;
-	XLOG_VEC_SET_TYPE(log_vector, XLOG_REG_TYPE_EFD_FORMAT);
+	log_vector->i_type = XLOG_REG_TYPE_EFD_FORMAT;
 	ASSERT(size >= sizeof(xfs_efd_log_format_t));
 }
 
diff --git a/fs/xfs/xfs_filestream.c b/fs/xfs/xfs_filestream.c
index a631e1451abb..390850ee6603 100644
--- a/fs/xfs/xfs_filestream.c
+++ b/fs/xfs/xfs_filestream.c
@@ -140,6 +140,7 @@ _xfs_filestream_pick_ag(
 	int		flags,
 	xfs_extlen_t	minlen)
 {
+	int		streams, max_streams;
 	int		err, trylock, nscan;
 	xfs_extlen_t	longest, free, minfree, maxfree = 0;
 	xfs_agnumber_t	ag, max_ag = NULLAGNUMBER;
@@ -155,15 +156,15 @@ _xfs_filestream_pick_ag(
 	trylock = XFS_ALLOC_FLAG_TRYLOCK;
 
 	for (nscan = 0; 1; nscan++) {
-
-		TRACE_AG_SCAN(mp, ag, xfs_filestream_peek_ag(mp, ag));
-
-		pag = mp->m_perag + ag;
+		pag = xfs_perag_get(mp, ag);
+		TRACE_AG_SCAN(mp, ag, atomic_read(&pag->pagf_fstrms));
 
 		if (!pag->pagf_init) {
 			err = xfs_alloc_pagf_init(mp, NULL, ag, trylock);
-			if (err && !trylock)
+			if (err && !trylock) {
+				xfs_perag_put(pag);
 				return err;
+			}
 		}
 
 		/* Might fail sometimes during the 1st pass with trylock set. */
@@ -173,6 +174,7 @@ _xfs_filestream_pick_ag(
 		/* Keep track of the AG with the most free blocks. */
 		if (pag->pagf_freeblks > maxfree) {
 			maxfree = pag->pagf_freeblks;
+			max_streams = atomic_read(&pag->pagf_fstrms);
 			max_ag = ag;
 		}
 
@@ -195,6 +197,8 @@ _xfs_filestream_pick_ag(
 
 			/* Break out, retaining the reference on the AG. */
 			free = pag->pagf_freeblks;
+			streams = atomic_read(&pag->pagf_fstrms);
+			xfs_perag_put(pag);
 			*agp = ag;
 			break;
 		}
@@ -202,6 +206,7 @@ _xfs_filestream_pick_ag(
 		/* Drop the reference on this AG, it's not usable. */
 		xfs_filestream_put_ag(mp, ag);
 next_ag:
+		xfs_perag_put(pag);
 		/* Move to the next AG, wrapping to AG 0 if necessary. */
 		if (++ag >= mp->m_sb.sb_agcount)
 			ag = 0;
@@ -229,6 +234,7 @@ next_ag:
 		if (max_ag != NULLAGNUMBER) {
 			xfs_filestream_get_ag(mp, max_ag);
 			TRACE_AG_PICK1(mp, max_ag, maxfree);
+			streams = max_streams;
 			free = maxfree;
 			*agp = max_ag;
 			break;
@@ -240,16 +246,14 @@ next_ag:
 		return 0;
 	}
 
-	TRACE_AG_PICK2(mp, startag, *agp, xfs_filestream_peek_ag(mp, *agp),
-			free, nscan, flags);
+	TRACE_AG_PICK2(mp, startag, *agp, streams, free, nscan, flags);
 
 	return 0;
 }
 
 /*
  * Set the allocation group number for a file or a directory, updating inode
- * references and per-AG references as appropriate.  Must be called with the
- * m_peraglock held in read mode.
+ * references and per-AG references as appropriate.
  */
 static int
 _xfs_filestream_update_ag(
@@ -451,20 +455,6 @@ xfs_filestream_unmount(
 }
 
 /*
- * If the mount point's m_perag array is going to be reallocated, all
- * outstanding cache entries must be flushed to avoid accessing reference count
- * addresses that have been freed.  The call to xfs_filestream_flush() must be
- * made inside the block that holds the m_peraglock in write mode to do the
- * reallocation.
- */
-void
-xfs_filestream_flush(
-	xfs_mount_t	*mp)
-{
-	xfs_mru_cache_flush(mp->m_filestream);
-}
-
-/*
  * Return the AG of the filestream the file or directory belongs to, or
  * NULLAGNUMBER otherwise.
  */
@@ -526,7 +516,6 @@ xfs_filestream_associate(
 
 	mp = pip->i_mount;
 	cache = mp->m_filestream;
-	down_read(&mp->m_peraglock);
 
 	/*
 	 * We have a problem, Houston.
@@ -543,10 +532,8 @@ xfs_filestream_associate(
 	 *
 	 * So, if we can't get the iolock without sleeping then just give up
 	 */
-	if (!xfs_ilock_nowait(pip, XFS_IOLOCK_EXCL)) {
-		up_read(&mp->m_peraglock);
+	if (!xfs_ilock_nowait(pip, XFS_IOLOCK_EXCL))
 		return 1;
-	}
 
 	/* If the parent directory is already in the cache, use its AG. */
 	item = xfs_mru_cache_lookup(cache, pip->i_ino);
@@ -601,7 +588,6 @@ exit_did_pick:
 
 exit:
 	xfs_iunlock(pip, XFS_IOLOCK_EXCL);
-	up_read(&mp->m_peraglock);
 	return -err;
 }
 
diff --git a/fs/xfs/xfs_filestream.h b/fs/xfs/xfs_filestream.h
index 4aba67c5f64f..260f757bbc5d 100644
--- a/fs/xfs/xfs_filestream.h
+++ b/fs/xfs/xfs_filestream.h
@@ -79,12 +79,21 @@ extern ktrace_t *xfs_filestreams_trace_buf;
  * the cache that reference per-ag array elements that have since been
  * reallocated.
  */
+/*
+ * xfs_filestream_peek_ag is only used in tracing code
+ */
 static inline int
 xfs_filestream_peek_ag(
 	xfs_mount_t	*mp,
 	xfs_agnumber_t	agno)
 {
-	return atomic_read(&mp->m_perag[agno].pagf_fstrms);
+	struct xfs_perag *pag;
+	int		ret;
+
+	pag = xfs_perag_get(mp, agno);
+	ret = atomic_read(&pag->pagf_fstrms);
+	xfs_perag_put(pag);
+	return ret;
 }
 
 static inline int
@@ -92,7 +101,13 @@ xfs_filestream_get_ag(
 	xfs_mount_t	*mp,
 	xfs_agnumber_t	agno)
 {
-	return atomic_inc_return(&mp->m_perag[agno].pagf_fstrms);
+	struct xfs_perag *pag;
+	int		ret;
+
+	pag = xfs_perag_get(mp, agno);
+	ret = atomic_inc_return(&pag->pagf_fstrms);
+	xfs_perag_put(pag);
+	return ret;
 }
 
 static inline int
@@ -100,7 +115,13 @@ xfs_filestream_put_ag(
 	xfs_mount_t	*mp,
 	xfs_agnumber_t	agno)
 {
-	return atomic_dec_return(&mp->m_perag[agno].pagf_fstrms);
+	struct xfs_perag *pag;
+	int		ret;
+
+	pag = xfs_perag_get(mp, agno);
+	ret = atomic_dec_return(&pag->pagf_fstrms);
+	xfs_perag_put(pag);
+	return ret;
 }
 
 /* allocation selection flags */
@@ -114,7 +135,6 @@ int xfs_filestream_init(void);
 void xfs_filestream_uninit(void);
 int xfs_filestream_mount(struct xfs_mount *mp);
 void xfs_filestream_unmount(struct xfs_mount *mp);
-void xfs_filestream_flush(struct xfs_mount *mp);
 xfs_agnumber_t xfs_filestream_lookup_ag(struct xfs_inode *ip);
 int xfs_filestream_associate(struct xfs_inode *dip, struct xfs_inode *ip);
 void xfs_filestream_deassociate(struct xfs_inode *ip);
diff --git a/fs/xfs/xfs_fs.h b/fs/xfs/xfs_fs.h
index f52ac276277e..7cf7220e7d5f 100644
--- a/fs/xfs/xfs_fs.h
+++ b/fs/xfs/xfs_fs.h
@@ -292,7 +292,8 @@ typedef struct xfs_bstat {
 	__s32		bs_extents;	/* number of extents		*/
 	__u32		bs_gen;		/* generation count		*/
 	__u16		bs_projid;	/* project id			*/
-	unsigned char	bs_pad[14];	/* pad space, unused		*/
+	__u16		bs_forkoff;	/* inode fork offset in bytes	*/
+	unsigned char	bs_pad[12];	/* pad space, unused		*/
 	__u32		bs_dmevmask;	/* DMIG event mask		*/
 	__u16		bs_dmstate;	/* DMIG state info		*/
 	__u16		bs_aextents;	/* attribute number of extents	*/
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index a13919a6a364..37a6f62c57b6 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -167,27 +167,14 @@ xfs_growfs_data_private(
 	}
 	new = nb - mp->m_sb.sb_dblocks;
 	oagcount = mp->m_sb.sb_agcount;
-	if (nagcount > oagcount) {
-		void *new_perag, *old_perag;
-
-		xfs_filestream_flush(mp);
-
-		new_perag = kmem_zalloc(sizeof(xfs_perag_t) * nagcount,
-					KM_MAYFAIL);
-		if (!new_perag)
-			return XFS_ERROR(ENOMEM);
-
-		down_write(&mp->m_peraglock);
-		memcpy(new_perag, mp->m_perag, sizeof(xfs_perag_t) * oagcount);
-		old_perag = mp->m_perag;
-		mp->m_perag = new_perag;
-
-		mp->m_flags |= XFS_MOUNT_32BITINODES;
-		nagimax = xfs_initialize_perag(mp, nagcount);
-		up_write(&mp->m_peraglock);
 
-		kmem_free(old_perag);
+	/* allocate the new per-ag structures */
+	if (nagcount > oagcount) {
+		error = xfs_initialize_perag(mp, nagcount, &nagimax);
+		if (error)
+			return error;
 	}
+
 	tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFS);
 	tp->t_flags |= XFS_TRANS_RESERVE;
 	if ((error = xfs_trans_reserve(tp, XFS_GROWFS_SPACE_RES(mp),
@@ -196,6 +183,11 @@ xfs_growfs_data_private(
 		return error;
 	}
 
+	/*
+	 * Write new AG headers to disk. Non-transactional, but written
+	 * synchronously so they are completed prior to the growfs transaction
+	 * being logged.
+	 */
 	nfree = 0;
 	for (agno = nagcount - 1; agno >= oagcount; agno--, new -= agsize) {
 		/*
@@ -359,6 +351,12 @@ xfs_growfs_data_private(
 			goto error0;
 		}
 	}
+
+	/*
+	 * Update changed superblock fields transactionally. These are not
+	 * seen by the rest of the world until the transaction commit applies
+	 * them atomically to the superblock.
+	 */
 	if (nagcount > oagcount)
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_AGCOUNT, nagcount - oagcount);
 	if (nb > mp->m_sb.sb_dblocks)
@@ -369,9 +367,9 @@ xfs_growfs_data_private(
 	if (dpct)
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_IMAXPCT, dpct);
 	error = xfs_trans_commit(tp, 0);
-	if (error) {
+	if (error)
 		return error;
-	}
+
 	/* New allocation groups fully initialized, so update mount struct */
 	if (nagimax)
 		mp->m_maxagi = nagimax;
@@ -381,6 +379,8 @@ xfs_growfs_data_private(
 		mp->m_maxicount = icount << mp->m_sb.sb_inopblog;
 	} else
 		mp->m_maxicount = 0;
+
+	/* update secondary superblocks. */
 	for (agno = 1; agno < nagcount; agno++) {
 		error = xfs_read_buf(mp, mp->m_ddev_targp,
 				  XFS_AGB_TO_DADDR(mp, agno, XFS_SB_BLOCK(mp)),
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c
index cb907ba69c4c..9d884c127bb9 100644
--- a/fs/xfs/xfs_ialloc.c
+++ b/fs/xfs/xfs_ialloc.c
@@ -205,7 +205,7 @@ xfs_ialloc_inode_init(
 		d = XFS_AGB_TO_DADDR(mp, agno, agbno + (j * blks_per_cluster));
 		fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
 					 mp->m_bsize * blks_per_cluster,
-					 XFS_BUF_LOCK);
+					 XBF_LOCK);
 		ASSERT(fbuf);
 		ASSERT(!XFS_BUF_GETERROR(fbuf));
 
@@ -253,6 +253,7 @@ xfs_ialloc_ag_alloc(
 	xfs_agino_t	thisino;	/* current inode number, for loop */
 	int		isaligned = 0;	/* inode allocation at stripe unit */
 					/* boundary */
+	struct xfs_perag *pag;
 
 	args.tp = tp;
 	args.mp = tp->t_mountp;
@@ -382,9 +383,9 @@ xfs_ialloc_ag_alloc(
 	newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0);
 	be32_add_cpu(&agi->agi_count, newlen);
 	be32_add_cpu(&agi->agi_freecount, newlen);
-	down_read(&args.mp->m_peraglock);
-	args.mp->m_perag[agno].pagi_freecount += newlen;
-	up_read(&args.mp->m_peraglock);
+	pag = xfs_perag_get(args.mp, agno);
+	pag->pagi_freecount += newlen;
+	xfs_perag_put(pag);
 	agi->agi_newino = cpu_to_be32(newino);
 
 	/*
@@ -486,9 +487,8 @@ xfs_ialloc_ag_select(
 	 */
 	agno = pagno;
 	flags = XFS_ALLOC_FLAG_TRYLOCK;
-	down_read(&mp->m_peraglock);
 	for (;;) {
-		pag = &mp->m_perag[agno];
+		pag = xfs_perag_get(mp, agno);
 		if (!pag->pagi_init) {
 			if (xfs_ialloc_read_agi(mp, tp, agno, &agbp)) {
 				agbp = NULL;
@@ -527,7 +527,7 @@ xfs_ialloc_ag_select(
 					agbp = NULL;
 					goto nextag;
 				}
-				up_read(&mp->m_peraglock);
+				xfs_perag_put(pag);
 				return agbp;
 			}
 		}
@@ -535,22 +535,19 @@ unlock_nextag:
 		if (agbp)
 			xfs_trans_brelse(tp, agbp);
 nextag:
+		xfs_perag_put(pag);
 		/*
 		 * No point in iterating over the rest, if we're shutting
 		 * down.
 		 */
-		if (XFS_FORCED_SHUTDOWN(mp)) {
-			up_read(&mp->m_peraglock);
+		if (XFS_FORCED_SHUTDOWN(mp))
 			return NULL;
-		}
 		agno++;
 		if (agno >= agcount)
 			agno = 0;
 		if (agno == pagno) {
-			if (flags == 0) {
-				up_read(&mp->m_peraglock);
+			if (flags == 0)
 				return NULL;
-			}
 			flags = 0;
 		}
 	}
@@ -672,6 +669,7 @@ xfs_dialloc(
 	xfs_agnumber_t	tagno;		/* testing allocation group number */
 	xfs_btree_cur_t	*tcur;		/* temp cursor */
 	xfs_inobt_rec_incore_t trec;	/* temp inode allocation record */
+	struct xfs_perag *pag;
 
 
 	if (*IO_agbp == NULL) {
@@ -771,13 +769,13 @@ nextag:
 			*inop = NULLFSINO;
 			return noroom ? ENOSPC : 0;
 		}
-		down_read(&mp->m_peraglock);
-		if (mp->m_perag[tagno].pagi_inodeok == 0) {
-			up_read(&mp->m_peraglock);
+		pag = xfs_perag_get(mp, tagno);
+		if (pag->pagi_inodeok == 0) {
+			xfs_perag_put(pag);
 			goto nextag;
 		}
 		error = xfs_ialloc_read_agi(mp, tp, tagno, &agbp);
-		up_read(&mp->m_peraglock);
+		xfs_perag_put(pag);
 		if (error)
 			goto nextag;
 		agi = XFS_BUF_TO_AGI(agbp);
@@ -790,6 +788,7 @@ nextag:
 	 */
 	agno = tagno;
 	*IO_agbp = NULL;
+	pag = xfs_perag_get(mp, agno);
 
  restart_pagno:
 	cur = xfs_inobt_init_cursor(mp, tp, agbp, be32_to_cpu(agi->agi_seqno));
@@ -808,7 +807,6 @@ nextag:
 	 * If in the same AG as the parent, try to get near the parent.
 	 */
 	if (pagno == agno) {
-		xfs_perag_t	*pag = &mp->m_perag[agno];
 		int		doneleft;	/* done, to the left */
 		int		doneright;	/* done, to the right */
 		int		searchdistance = 10;
@@ -1006,9 +1004,7 @@ alloc_inode:
 		goto error0;
 	be32_add_cpu(&agi->agi_freecount, -1);
 	xfs_ialloc_log_agi(tp, agbp, XFS_AGI_FREECOUNT);
-	down_read(&mp->m_peraglock);
-	mp->m_perag[tagno].pagi_freecount--;
-	up_read(&mp->m_peraglock);
+	pag->pagi_freecount--;
 
 	error = xfs_check_agi_freecount(cur, agi);
 	if (error)
@@ -1016,12 +1012,14 @@ alloc_inode:
 
 	xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
 	xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -1);
+	xfs_perag_put(pag);
 	*inop = ino;
 	return 0;
 error1:
 	xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR);
 error0:
 	xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
+	xfs_perag_put(pag);
 	return error;
 }
 
@@ -1052,6 +1050,7 @@ xfs_difree(
 	xfs_mount_t	*mp;	/* mount structure for filesystem */
 	int		off;	/* offset of inode in inode chunk */
 	xfs_inobt_rec_incore_t rec;	/* btree record */
+	struct xfs_perag *pag;
 
 	mp = tp->t_mountp;
 
@@ -1088,9 +1087,7 @@ xfs_difree(
 	/*
 	 * Get the allocation group header.
 	 */
-	down_read(&mp->m_peraglock);
 	error = xfs_ialloc_read_agi(mp, tp, agno, &agbp);
-	up_read(&mp->m_peraglock);
 	if (error) {
 		cmn_err(CE_WARN,
 			"xfs_difree: xfs_ialloc_read_agi() returned an error %d on %s.  Returning error.",
@@ -1157,9 +1154,9 @@ xfs_difree(
 		be32_add_cpu(&agi->agi_count, -ilen);
 		be32_add_cpu(&agi->agi_freecount, -(ilen - 1));
 		xfs_ialloc_log_agi(tp, agbp, XFS_AGI_COUNT | XFS_AGI_FREECOUNT);
-		down_read(&mp->m_peraglock);
-		mp->m_perag[agno].pagi_freecount -= ilen - 1;
-		up_read(&mp->m_peraglock);
+		pag = xfs_perag_get(mp, agno);
+		pag->pagi_freecount -= ilen - 1;
+		xfs_perag_put(pag);
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_ICOUNT, -ilen);
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -(ilen - 1));
 
@@ -1188,9 +1185,9 @@ xfs_difree(
 		 */
 		be32_add_cpu(&agi->agi_freecount, 1);
 		xfs_ialloc_log_agi(tp, agbp, XFS_AGI_FREECOUNT);
-		down_read(&mp->m_peraglock);
-		mp->m_perag[agno].pagi_freecount++;
-		up_read(&mp->m_peraglock);
+		pag = xfs_perag_get(mp, agno);
+		pag->pagi_freecount++;
+		xfs_perag_put(pag);
 		xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, 1);
 	}
 
@@ -1312,9 +1309,7 @@ xfs_imap(
 		xfs_buf_t	*agbp;	/* agi buffer */
 		int		i;	/* temp state */
 
-		down_read(&mp->m_peraglock);
 		error = xfs_ialloc_read_agi(mp, tp, agno, &agbp);
-		up_read(&mp->m_peraglock);
 		if (error) {
 			xfs_fs_cmn_err(CE_ALERT, mp, "xfs_imap: "
 					"xfs_ialloc_read_agi() returned "
@@ -1379,7 +1374,6 @@ xfs_imap(
 			XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks));
 		return XFS_ERROR(EINVAL);
 	}
-
 	return 0;
 }
 
@@ -1523,8 +1517,7 @@ xfs_ialloc_read_agi(
 		return error;
 
 	agi = XFS_BUF_TO_AGI(*bpp);
-	pag = &mp->m_perag[agno];
-
+	pag = xfs_perag_get(mp, agno);
 	if (!pag->pagi_init) {
 		pag->pagi_freecount = be32_to_cpu(agi->agi_freecount);
 		pag->pagi_count = be32_to_cpu(agi->agi_count);
@@ -1537,6 +1530,7 @@ xfs_ialloc_read_agi(
 	 */
 	ASSERT(pag->pagi_freecount == be32_to_cpu(agi->agi_freecount) ||
 		XFS_FORCED_SHUTDOWN(mp));
+	xfs_perag_put(pag);
 	return 0;
 }
 
diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c
index 155e798f30a1..6845db90818f 100644
--- a/fs/xfs/xfs_iget.c
+++ b/fs/xfs/xfs_iget.c
@@ -190,13 +190,12 @@ xfs_iget_cache_hit(
 		trace_xfs_iget_reclaim(ip);
 
 		/*
-		 * We need to set XFS_INEW atomically with clearing the
-		 * reclaimable tag so that we do have an indicator of the
-		 * inode still being initialized.
+		 * We need to set XFS_IRECLAIM to prevent xfs_reclaim_inode
+		 * from stomping over us while we recycle the inode.  We can't
+		 * clear the radix tree reclaimable tag yet as it requires
+		 * pag_ici_lock to be held exclusive.
 		 */
-		ip->i_flags |= XFS_INEW;
-		ip->i_flags &= ~XFS_IRECLAIMABLE;
-		__xfs_inode_clear_reclaim_tag(mp, pag, ip);
+		ip->i_flags |= XFS_IRECLAIM;
 
 		spin_unlock(&ip->i_flags_lock);
 		read_unlock(&pag->pag_ici_lock);
@@ -216,7 +215,15 @@ xfs_iget_cache_hit(
 			trace_xfs_iget_reclaim(ip);
 			goto out_error;
 		}
+
+		write_lock(&pag->pag_ici_lock);
+		spin_lock(&ip->i_flags_lock);
+		ip->i_flags &= ~(XFS_IRECLAIMABLE | XFS_IRECLAIM);
+		ip->i_flags |= XFS_INEW;
+		__xfs_inode_clear_reclaim_tag(mp, pag, ip);
 		inode->i_state = I_NEW;
+		spin_unlock(&ip->i_flags_lock);
+		write_unlock(&pag->pag_ici_lock);
 	} else {
 		/* If the VFS inode is being torn down, pause and try again. */
 		if (!igrab(inode)) {
@@ -374,7 +381,7 @@ xfs_iget(
 		return EINVAL;
 
 	/* get the perag structure and ensure that it's inode capable */
-	pag = xfs_get_perag(mp, ino);
+	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ino));
 	if (!pag->pagi_inodeok)
 		return EINVAL;
 	ASSERT(pag->pag_ici_init);
@@ -398,7 +405,7 @@ again:
 		if (error)
 			goto out_error_or_again;
 	}
-	xfs_put_perag(mp, pag);
+	xfs_perag_put(pag);
 
 	*ipp = ip;
 
@@ -417,7 +424,7 @@ out_error_or_again:
 		delay(1);
 		goto again;
 	}
-	xfs_put_perag(mp, pag);
+	xfs_perag_put(pag);
 	return error;
 }
 
@@ -488,12 +495,12 @@ xfs_ireclaim(
 	 * added to the tree assert that it's been there before to catch
 	 * problems with the inode life time early on.
 	 */
-	pag = xfs_get_perag(mp, ip->i_ino);
+	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
 	write_lock(&pag->pag_ici_lock);
 	if (!radix_tree_delete(&pag->pag_ici_root, agino))
 		ASSERT(0);
 	write_unlock(&pag->pag_ici_lock);
-	xfs_put_perag(mp, pag);
+	xfs_perag_put(pag);
 
 	/*
 	 * Here we do an (almost) spurious inode lock in order to coordinate
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index ef77fd88c8e3..0ffd56447045 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -151,7 +151,7 @@ xfs_imap_to_bp(
 				"an error %d on %s.  Returning error.",
 				error, mp->m_fsname);
 		} else {
-			ASSERT(buf_flags & XFS_BUF_TRYLOCK);
+			ASSERT(buf_flags & XBF_TRYLOCK);
 		}
 		return error;
 	}
@@ -239,7 +239,7 @@ xfs_inotobp(
 	if (error)
 		return error;
 
-	error = xfs_imap_to_bp(mp, tp, &imap, &bp, XFS_BUF_LOCK, imap_flags);
+	error = xfs_imap_to_bp(mp, tp, &imap, &bp, XBF_LOCK, imap_flags);
 	if (error)
 		return error;
 
@@ -285,7 +285,7 @@ xfs_itobp(
 		return error;
 
 	if (!bp) {
-		ASSERT(buf_flags & XFS_BUF_TRYLOCK);
+		ASSERT(buf_flags & XBF_TRYLOCK);
 		ASSERT(tp == NULL);
 		*bpp = NULL;
 		return EAGAIN;
@@ -807,7 +807,7 @@ xfs_iread(
 	 * Get pointers to the on-disk inode and the buffer containing it.
 	 */
 	error = xfs_imap_to_bp(mp, tp, &ip->i_imap, &bp,
-			       XFS_BUF_LOCK, iget_flags);
+			       XBF_LOCK, iget_flags);
 	if (error)
 		return error;
 	dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset);
@@ -1751,7 +1751,7 @@ xfs_iunlink(
 		 * Here we put the head pointer into our next pointer,
 		 * and then we fall through to point the head at us.
 		 */
-		error = xfs_itobp(mp, tp, ip, &dip, &ibp, XFS_BUF_LOCK);
+		error = xfs_itobp(mp, tp, ip, &dip, &ibp, XBF_LOCK);
 		if (error)
 			return error;
 
@@ -1833,7 +1833,7 @@ xfs_iunlink_remove(
 		 * of dealing with the buffer when there is no need to
 		 * change it.
 		 */
-		error = xfs_itobp(mp, tp, ip, &dip, &ibp, XFS_BUF_LOCK);
+		error = xfs_itobp(mp, tp, ip, &dip, &ibp, XBF_LOCK);
 		if (error) {
 			cmn_err(CE_WARN,
 				"xfs_iunlink_remove: xfs_itobp()  returned an error %d on %s.  Returning error.",
@@ -1895,7 +1895,7 @@ xfs_iunlink_remove(
 		 * Now last_ibp points to the buffer previous to us on
 		 * the unlinked list.  Pull us from the list.
 		 */
-		error = xfs_itobp(mp, tp, ip, &dip, &ibp, XFS_BUF_LOCK);
+		error = xfs_itobp(mp, tp, ip, &dip, &ibp, XBF_LOCK);
 		if (error) {
 			cmn_err(CE_WARN,
 				"xfs_iunlink_remove: xfs_itobp()  returned an error %d on %s.  Returning error.",
@@ -1946,8 +1946,9 @@ xfs_ifree_cluster(
 	xfs_inode_t		*ip, **ip_found;
 	xfs_inode_log_item_t	*iip;
 	xfs_log_item_t		*lip;
-	xfs_perag_t		*pag = xfs_get_perag(mp, inum);
+	struct xfs_perag	*pag;
 
+	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, inum));
 	if (mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(mp)) {
 		blks_per_cluster = 1;
 		ninodes = mp->m_sb.sb_inopblock;
@@ -2039,7 +2040,7 @@ xfs_ifree_cluster(
 
 		bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, blkno, 
 					mp->m_bsize * blks_per_cluster,
-					XFS_BUF_LOCK);
+					XBF_LOCK);
 
 		pre_flushed = 0;
 		lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
@@ -2088,7 +2089,7 @@ xfs_ifree_cluster(
 	}
 
 	kmem_free(ip_found);
-	xfs_put_perag(mp, pag);
+	xfs_perag_put(pag);
 }
 
 /*
@@ -2150,7 +2151,7 @@ xfs_ifree(
 
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 
-	error = xfs_itobp(ip->i_mount, tp, ip, &dip, &ibp, XFS_BUF_LOCK);
+	error = xfs_itobp(ip->i_mount, tp, ip, &dip, &ibp, XBF_LOCK);
 	if (error)
 		return error;
 
@@ -2438,72 +2439,31 @@ xfs_idestroy_fork(
 }
 
 /*
- * Increment the pin count of the given buffer.
- * This value is protected by ipinlock spinlock in the mount structure.
+ * This is called to unpin an inode.  The caller must have the inode locked
+ * in at least shared mode so that the buffer cannot be subsequently pinned
+ * once someone is waiting for it to be unpinned.
  */
-void
-xfs_ipin(
-	xfs_inode_t	*ip)
-{
-	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
-
-	atomic_inc(&ip->i_pincount);
-}
-
-/*
- * Decrement the pin count of the given inode, and wake up
- * anyone in xfs_iwait_unpin() if the count goes to 0.  The
- * inode must have been previously pinned with a call to xfs_ipin().
- */
-void
-xfs_iunpin(
-	xfs_inode_t	*ip)
-{
-	ASSERT(atomic_read(&ip->i_pincount) > 0);
-
-	if (atomic_dec_and_test(&ip->i_pincount))
-		wake_up(&ip->i_ipin_wait);
-}
-
-/*
- * This is called to unpin an inode. It can be directed to wait or to return
- * immediately without waiting for the inode to be unpinned.  The caller must
- * have the inode locked in at least shared mode so that the buffer cannot be
- * subsequently pinned once someone is waiting for it to be unpinned.
- */
-STATIC void
-__xfs_iunpin_wait(
-	xfs_inode_t	*ip,
-	int		wait)
+static void
+xfs_iunpin_nowait(
+	struct xfs_inode	*ip)
 {
-	xfs_inode_log_item_t	*iip = ip->i_itemp;
-
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED));
-	if (atomic_read(&ip->i_pincount) == 0)
-		return;
 
 	/* Give the log a push to start the unpinning I/O */
-	xfs_log_force(ip->i_mount, (iip && iip->ili_last_lsn) ?
-				iip->ili_last_lsn : 0, XFS_LOG_FORCE);
-	if (wait)
-		wait_event(ip->i_ipin_wait, (atomic_read(&ip->i_pincount) == 0));
-}
+	xfs_log_force_lsn(ip->i_mount, ip->i_itemp->ili_last_lsn, 0);
 
-static inline void
-xfs_iunpin_wait(
-	xfs_inode_t	*ip)
-{
-	__xfs_iunpin_wait(ip, 1);
 }
 
-static inline void
-xfs_iunpin_nowait(
-	xfs_inode_t	*ip)
+void
+xfs_iunpin_wait(
+	struct xfs_inode	*ip)
 {
-	__xfs_iunpin_wait(ip, 0);
+	if (xfs_ipincount(ip)) {
+		xfs_iunpin_nowait(ip);
+		wait_event(ip->i_ipin_wait, (xfs_ipincount(ip) == 0));
+	}
 }
 
-
 /*
  * xfs_iextents_copy()
  *
@@ -2675,7 +2635,7 @@ xfs_iflush_cluster(
 	xfs_buf_t	*bp)
 {
 	xfs_mount_t		*mp = ip->i_mount;
-	xfs_perag_t		*pag = xfs_get_perag(mp, ip->i_ino);
+	struct xfs_perag	*pag;
 	unsigned long		first_index, mask;
 	unsigned long		inodes_per_cluster;
 	int			ilist_size;
@@ -2686,6 +2646,7 @@ xfs_iflush_cluster(
 	int			bufwasdelwri;
 	int			i;
 
+	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
 	ASSERT(pag->pagi_inodeok);
 	ASSERT(pag->pag_ici_init);
 
@@ -2693,7 +2654,7 @@ xfs_iflush_cluster(
 	ilist_size = inodes_per_cluster * sizeof(xfs_inode_t *);
 	ilist = kmem_alloc(ilist_size, KM_MAYFAIL|KM_NOFS);
 	if (!ilist)
-		return 0;
+		goto out_put;
 
 	mask = ~(((XFS_INODE_CLUSTER_SIZE(mp) >> mp->m_sb.sb_inodelog)) - 1);
 	first_index = XFS_INO_TO_AGINO(mp, ip->i_ino) & mask;
@@ -2762,6 +2723,8 @@ xfs_iflush_cluster(
 out_free:
 	read_unlock(&pag->pag_ici_lock);
 	kmem_free(ilist);
+out_put:
+	xfs_perag_put(pag);
 	return 0;
 
 
@@ -2805,6 +2768,7 @@ cluster_corrupt_out:
 	 */
 	xfs_iflush_abort(iq);
 	kmem_free(ilist);
+	xfs_perag_put(pag);
 	return XFS_ERROR(EFSCORRUPTED);
 }
 
@@ -2827,8 +2791,6 @@ xfs_iflush(
 	xfs_dinode_t		*dip;
 	xfs_mount_t		*mp;
 	int			error;
-	int			noblock = (flags == XFS_IFLUSH_ASYNC_NOBLOCK);
-	enum { INT_DELWRI = (1 << 0), INT_ASYNC = (1 << 1) };
 
 	XFS_STATS_INC(xs_iflush_count);
 
@@ -2841,15 +2803,6 @@ xfs_iflush(
 	mp = ip->i_mount;
 
 	/*
-	 * If the inode isn't dirty, then just release the inode flush lock and
-	 * do nothing.
-	 */
-	if (xfs_inode_clean(ip)) {
-		xfs_ifunlock(ip);
-		return 0;
-	}
-
-	/*
 	 * We can't flush the inode until it is unpinned, so wait for it if we
 	 * are allowed to block.  We know noone new can pin it, because we are
 	 * holding the inode lock shared and you need to hold it exclusively to
@@ -2860,7 +2813,7 @@ xfs_iflush(
 	 * in the same cluster are dirty, they will probably write the inode
 	 * out for us if they occur after the log force completes.
 	 */
-	if (noblock && xfs_ipincount(ip)) {
+	if (!(flags & SYNC_WAIT) && xfs_ipincount(ip)) {
 		xfs_iunpin_nowait(ip);
 		xfs_ifunlock(ip);
 		return EAGAIN;
@@ -2894,60 +2847,10 @@ xfs_iflush(
 	}
 
 	/*
-	 * Decide how buffer will be flushed out.  This is done before
-	 * the call to xfs_iflush_int because this field is zeroed by it.
-	 */
-	if (iip != NULL && iip->ili_format.ilf_fields != 0) {
-		/*
-		 * Flush out the inode buffer according to the directions
-		 * of the caller.  In the cases where the caller has given
-		 * us a choice choose the non-delwri case.  This is because
-		 * the inode is in the AIL and we need to get it out soon.
-		 */
-		switch (flags) {
-		case XFS_IFLUSH_SYNC:
-		case XFS_IFLUSH_DELWRI_ELSE_SYNC:
-			flags = 0;
-			break;
-		case XFS_IFLUSH_ASYNC_NOBLOCK:
-		case XFS_IFLUSH_ASYNC:
-		case XFS_IFLUSH_DELWRI_ELSE_ASYNC:
-			flags = INT_ASYNC;
-			break;
-		case XFS_IFLUSH_DELWRI:
-			flags = INT_DELWRI;
-			break;
-		default:
-			ASSERT(0);
-			flags = 0;
-			break;
-		}
-	} else {
-		switch (flags) {
-		case XFS_IFLUSH_DELWRI_ELSE_SYNC:
-		case XFS_IFLUSH_DELWRI_ELSE_ASYNC:
-		case XFS_IFLUSH_DELWRI:
-			flags = INT_DELWRI;
-			break;
-		case XFS_IFLUSH_ASYNC_NOBLOCK:
-		case XFS_IFLUSH_ASYNC:
-			flags = INT_ASYNC;
-			break;
-		case XFS_IFLUSH_SYNC:
-			flags = 0;
-			break;
-		default:
-			ASSERT(0);
-			flags = 0;
-			break;
-		}
-	}
-
-	/*
 	 * Get the buffer containing the on-disk inode.
 	 */
 	error = xfs_itobp(mp, NULL, ip, &dip, &bp,
-				noblock ? XFS_BUF_TRYLOCK : XFS_BUF_LOCK);
+				(flags & SYNC_WAIT) ? XBF_LOCK : XBF_TRYLOCK);
 	if (error || !bp) {
 		xfs_ifunlock(ip);
 		return error;
@@ -2965,7 +2868,7 @@ xfs_iflush(
 	 * get stuck waiting in the write for too long.
 	 */
 	if (XFS_BUF_ISPINNED(bp))
-		xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
+		xfs_log_force(mp, 0);
 
 	/*
 	 * inode clustering:
@@ -2975,13 +2878,10 @@ xfs_iflush(
 	if (error)
 		goto cluster_corrupt_out;
 
-	if (flags & INT_DELWRI) {
-		xfs_bdwrite(mp, bp);
-	} else if (flags & INT_ASYNC) {
-		error = xfs_bawrite(mp, bp);
-	} else {
+	if (flags & SYNC_WAIT)
 		error = xfs_bwrite(mp, bp);
-	}
+	else
+		xfs_bdwrite(mp, bp);
 	return error;
 
 corrupt_out:
@@ -3016,16 +2916,6 @@ xfs_iflush_int(
 	iip = ip->i_itemp;
 	mp = ip->i_mount;
 
-
-	/*
-	 * If the inode isn't dirty, then just release the inode
-	 * flush lock and do nothing.
-	 */
-	if (xfs_inode_clean(ip)) {
-		xfs_ifunlock(ip);
-		return 0;
-	}
-
 	/* set *dip = inode's place in the buffer */
 	dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset);
 
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index ec1f28c4fc4f..9965e40a4615 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -420,16 +420,6 @@ static inline void xfs_ifunlock(xfs_inode_t *ip)
 #define XFS_ILOCK_DEP(flags)	(((flags) & XFS_ILOCK_DEP_MASK) >> XFS_ILOCK_SHIFT)
 
 /*
- * Flags for xfs_iflush()
- */
-#define	XFS_IFLUSH_DELWRI_ELSE_SYNC	1
-#define	XFS_IFLUSH_DELWRI_ELSE_ASYNC	2
-#define	XFS_IFLUSH_SYNC			3
-#define	XFS_IFLUSH_ASYNC		4
-#define	XFS_IFLUSH_DELWRI		5
-#define	XFS_IFLUSH_ASYNC_NOBLOCK	6
-
-/*
  * Flags for xfs_itruncate_start().
  */
 #define	XFS_ITRUNC_DEFINITE	0x1
@@ -481,14 +471,14 @@ int		xfs_itruncate_finish(struct xfs_trans **, xfs_inode_t *,
 int		xfs_iunlink(struct xfs_trans *, xfs_inode_t *);
 
 void		xfs_iext_realloc(xfs_inode_t *, int, int);
-void		xfs_ipin(xfs_inode_t *);
-void		xfs_iunpin(xfs_inode_t *);
+void		xfs_iunpin_wait(xfs_inode_t *);
 int		xfs_iflush(xfs_inode_t *, uint);
 void		xfs_ichgtime(xfs_inode_t *, int);
 void		xfs_lock_inodes(xfs_inode_t **, int, uint);
 void		xfs_lock_two_inodes(xfs_inode_t *, xfs_inode_t *, uint);
 
 void		xfs_synchronize_times(xfs_inode_t *);
+void		xfs_mark_inode_dirty(xfs_inode_t *);
 void		xfs_mark_inode_dirty_sync(xfs_inode_t *);
 
 #define IHOLD(ip) \
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index f38855d21ea5..7bfea8540159 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -228,7 +228,7 @@ xfs_inode_item_format(
 
 	vecp->i_addr = (xfs_caddr_t)&iip->ili_format;
 	vecp->i_len  = sizeof(xfs_inode_log_format_t);
-	XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IFORMAT);
+	vecp->i_type = XLOG_REG_TYPE_IFORMAT;
 	vecp++;
 	nvecs	     = 1;
 
@@ -279,7 +279,7 @@ xfs_inode_item_format(
 
 	vecp->i_addr = (xfs_caddr_t)&ip->i_d;
 	vecp->i_len  = sizeof(struct xfs_icdinode);
-	XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_ICORE);
+	vecp->i_type = XLOG_REG_TYPE_ICORE;
 	vecp++;
 	nvecs++;
 	iip->ili_format.ilf_fields |= XFS_ILOG_CORE;
@@ -336,7 +336,7 @@ xfs_inode_item_format(
 				vecp->i_addr =
 					(char *)(ip->i_df.if_u1.if_extents);
 				vecp->i_len = ip->i_df.if_bytes;
-				XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IEXT);
+				vecp->i_type = XLOG_REG_TYPE_IEXT;
 			} else
 #endif
 			{
@@ -355,7 +355,7 @@ xfs_inode_item_format(
 				vecp->i_addr = (xfs_caddr_t)ext_buffer;
 				vecp->i_len = xfs_iextents_copy(ip, ext_buffer,
 						XFS_DATA_FORK);
-				XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IEXT);
+				vecp->i_type = XLOG_REG_TYPE_IEXT;
 			}
 			ASSERT(vecp->i_len <= ip->i_df.if_bytes);
 			iip->ili_format.ilf_dsize = vecp->i_len;
@@ -373,7 +373,7 @@ xfs_inode_item_format(
 			ASSERT(ip->i_df.if_broot != NULL);
 			vecp->i_addr = (xfs_caddr_t)ip->i_df.if_broot;
 			vecp->i_len = ip->i_df.if_broot_bytes;
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IBROOT);
+			vecp->i_type = XLOG_REG_TYPE_IBROOT;
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_dsize = ip->i_df.if_broot_bytes;
@@ -399,7 +399,7 @@ xfs_inode_item_format(
 			ASSERT((ip->i_df.if_real_bytes == 0) ||
 			       (ip->i_df.if_real_bytes == data_bytes));
 			vecp->i_len = (int)data_bytes;
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_ILOCAL);
+			vecp->i_type = XLOG_REG_TYPE_ILOCAL;
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_dsize = (unsigned)data_bytes;
@@ -477,7 +477,7 @@ xfs_inode_item_format(
 			vecp->i_len = xfs_iextents_copy(ip, ext_buffer,
 					XFS_ATTR_FORK);
 #endif
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IATTR_EXT);
+			vecp->i_type = XLOG_REG_TYPE_IATTR_EXT;
 			iip->ili_format.ilf_asize = vecp->i_len;
 			vecp++;
 			nvecs++;
@@ -492,7 +492,7 @@ xfs_inode_item_format(
 			ASSERT(ip->i_afp->if_broot != NULL);
 			vecp->i_addr = (xfs_caddr_t)ip->i_afp->if_broot;
 			vecp->i_len = ip->i_afp->if_broot_bytes;
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IATTR_BROOT);
+			vecp->i_type = XLOG_REG_TYPE_IATTR_BROOT;
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_asize = ip->i_afp->if_broot_bytes;
@@ -516,7 +516,7 @@ xfs_inode_item_format(
 			ASSERT((ip->i_afp->if_real_bytes == 0) ||
 			       (ip->i_afp->if_real_bytes == data_bytes));
 			vecp->i_len = (int)data_bytes;
-			XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IATTR_LOCAL);
+			vecp->i_type = XLOG_REG_TYPE_IATTR_LOCAL;
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_asize = (unsigned)data_bytes;
@@ -535,23 +535,23 @@ xfs_inode_item_format(
 
 /*
  * This is called to pin the inode associated with the inode log
- * item in memory so it cannot be written out.  Do this by calling
- * xfs_ipin() to bump the pin count in the inode while holding the
- * inode pin lock.
+ * item in memory so it cannot be written out.
  */
 STATIC void
 xfs_inode_item_pin(
 	xfs_inode_log_item_t	*iip)
 {
 	ASSERT(xfs_isilocked(iip->ili_inode, XFS_ILOCK_EXCL));
-	xfs_ipin(iip->ili_inode);
+
+	atomic_inc(&iip->ili_inode->i_pincount);
 }
 
 
 /*
  * This is called to unpin the inode associated with the inode log
  * item which was previously pinned with a call to xfs_inode_item_pin().
- * Just call xfs_iunpin() on the inode to do this.
+ *
+ * Also wake up anyone in xfs_iunpin_wait() if the count goes to 0.
  */
 /* ARGSUSED */
 STATIC void
@@ -559,7 +559,11 @@ xfs_inode_item_unpin(
 	xfs_inode_log_item_t	*iip,
 	int			stale)
 {
-	xfs_iunpin(iip->ili_inode);
+	struct xfs_inode	*ip = iip->ili_inode;
+
+	ASSERT(atomic_read(&ip->i_pincount) > 0);
+	if (atomic_dec_and_test(&ip->i_pincount))
+		wake_up(&ip->i_ipin_wait);
 }
 
 /* ARGSUSED */
@@ -568,7 +572,7 @@ xfs_inode_item_unpin_remove(
 	xfs_inode_log_item_t	*iip,
 	xfs_trans_t		*tp)
 {
-	xfs_iunpin(iip->ili_inode);
+	xfs_inode_item_unpin(iip, 0);
 }
 
 /*
@@ -602,33 +606,20 @@ xfs_inode_item_trylock(
 
 	if (!xfs_iflock_nowait(ip)) {
 		/*
-		 * If someone else isn't already trying to push the inode
-		 * buffer, we get to do it.
+		 * inode has already been flushed to the backing buffer,
+		 * leave it locked in shared mode, pushbuf routine will
+		 * unlock it.
 		 */
-		if (iip->ili_pushbuf_flag == 0) {
-			iip->ili_pushbuf_flag = 1;
-#ifdef DEBUG
-			iip->ili_push_owner = current_pid();
-#endif
-			/*
-			 * Inode is left locked in shared mode.
-			 * Pushbuf routine gets to unlock it.
-			 */
-			return XFS_ITEM_PUSHBUF;
-		} else {
-			/*
-			 * We hold the AIL lock, so we must specify the
-			 * NONOTIFY flag so that we won't double trip.
-			 */
-			xfs_iunlock(ip, XFS_ILOCK_SHARED|XFS_IUNLOCK_NONOTIFY);
-			return XFS_ITEM_FLUSHING;
-		}
-		/* NOTREACHED */
+		return XFS_ITEM_PUSHBUF;
 	}
 
 	/* Stale items should force out the iclog */
 	if (ip->i_flags & XFS_ISTALE) {
 		xfs_ifunlock(ip);
+		/*
+		 * we hold the AIL lock - notify the unlock routine of this
+		 * so it doesn't try to get the lock again.
+		 */
 		xfs_iunlock(ip, XFS_ILOCK_SHARED|XFS_IUNLOCK_NONOTIFY);
 		return XFS_ITEM_PINNED;
 	}
@@ -746,11 +737,8 @@ xfs_inode_item_committed(
  * This gets called by xfs_trans_push_ail(), when IOP_TRYLOCK
  * failed to get the inode flush lock but did get the inode locked SHARED.
  * Here we're trying to see if the inode buffer is incore, and if so whether it's
- * marked delayed write. If that's the case, we'll initiate a bawrite on that
- * buffer to expedite the process.
- *
- * We aren't holding the AIL lock (or the flush lock) when this gets called,
- * so it is inherently race-y.
+ * marked delayed write. If that's the case, we'll promote it and that will
+ * allow the caller to write the buffer by triggering the xfsbufd to run.
  */
 STATIC void
 xfs_inode_item_pushbuf(
@@ -759,82 +747,30 @@ xfs_inode_item_pushbuf(
 	xfs_inode_t	*ip;
 	xfs_mount_t	*mp;
 	xfs_buf_t	*bp;
-	uint		dopush;
 
 	ip = iip->ili_inode;
-
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED));
 
 	/*
-	 * The ili_pushbuf_flag keeps others from
-	 * trying to duplicate our effort.
-	 */
-	ASSERT(iip->ili_pushbuf_flag != 0);
-	ASSERT(iip->ili_push_owner == current_pid());
-
-	/*
 	 * If a flush is not in progress anymore, chances are that the
 	 * inode was taken off the AIL. So, just get out.
 	 */
 	if (completion_done(&ip->i_flush) ||
 	    ((iip->ili_item.li_flags & XFS_LI_IN_AIL) == 0)) {
-		iip->ili_pushbuf_flag = 0;
 		xfs_iunlock(ip, XFS_ILOCK_SHARED);
 		return;
 	}
 
 	mp = ip->i_mount;
 	bp = xfs_incore(mp->m_ddev_targp, iip->ili_format.ilf_blkno,
-		    iip->ili_format.ilf_len, XFS_INCORE_TRYLOCK);
-
-	if (bp != NULL) {
-		if (XFS_BUF_ISDELAYWRITE(bp)) {
-			/*
-			 * We were racing with iflush because we don't hold
-			 * the AIL lock or the flush lock. However, at this point,
-			 * we have the buffer, and we know that it's dirty.
-			 * So, it's possible that iflush raced with us, and
-			 * this item is already taken off the AIL.
-			 * If not, we can flush it async.
-			 */
-			dopush = ((iip->ili_item.li_flags & XFS_LI_IN_AIL) &&
-				  !completion_done(&ip->i_flush));
-			iip->ili_pushbuf_flag = 0;
-			xfs_iunlock(ip, XFS_ILOCK_SHARED);
-
-			trace_xfs_inode_item_push(bp, _RET_IP_);
+		    iip->ili_format.ilf_len, XBF_TRYLOCK);
 
-			if (XFS_BUF_ISPINNED(bp)) {
-				xfs_log_force(mp, (xfs_lsn_t)0,
-					      XFS_LOG_FORCE);
-			}
-			if (dopush) {
-				int	error;
-				error = xfs_bawrite(mp, bp);
-				if (error)
-					xfs_fs_cmn_err(CE_WARN, mp,
-		"xfs_inode_item_pushbuf: pushbuf error %d on iip %p, bp %p",
-							error, iip, bp);
-			} else {
-				xfs_buf_relse(bp);
-			}
-		} else {
-			iip->ili_pushbuf_flag = 0;
-			xfs_iunlock(ip, XFS_ILOCK_SHARED);
-			xfs_buf_relse(bp);
-		}
-		return;
-	}
-	/*
-	 * We have to be careful about resetting pushbuf flag too early (above).
-	 * Even though in theory we can do it as soon as we have the buflock,
-	 * we don't want others to be doing work needlessly. They'll come to
-	 * this function thinking that pushing the buffer is their
-	 * responsibility only to find that the buffer is still locked by
-	 * another doing the same thing
-	 */
-	iip->ili_pushbuf_flag = 0;
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
+	if (!bp)
+		return;
+	if (XFS_BUF_ISDELAYWRITE(bp))
+		xfs_buf_delwri_promote(bp);
+	xfs_buf_relse(bp);
 	return;
 }
 
@@ -867,10 +803,14 @@ xfs_inode_item_push(
 	       iip->ili_format.ilf_fields != 0);
 
 	/*
-	 * Write out the inode.  The completion routine ('iflush_done') will
-	 * pull it from the AIL, mark it clean, unlock the flush lock.
+	 * Push the inode to it's backing buffer. This will not remove the
+	 * inode from the AIL - a further push will be required to trigger a
+	 * buffer push. However, this allows all the dirty inodes to be pushed
+	 * to the buffer before it is pushed to disk. THe buffer IO completion
+	 * will pull th einode from the AIL, mark it clean and unlock the flush
+	 * lock.
 	 */
-	(void) xfs_iflush(ip, XFS_IFLUSH_ASYNC);
+	(void) xfs_iflush(ip, 0);
 	xfs_iunlock(ip, XFS_ILOCK_SHARED);
 
 	return;
@@ -934,7 +874,6 @@ xfs_inode_item_init(
 	/*
 	   We have zeroed memory. No need ...
 	   iip->ili_extents_buf = NULL;
-	   iip->ili_pushbuf_flag = 0;
 	 */
 
 	iip->ili_format.ilf_type = XFS_LI_INODE;
diff --git a/fs/xfs/xfs_inode_item.h b/fs/xfs/xfs_inode_item.h
index cc8df1ac7783..9a467958ecdd 100644
--- a/fs/xfs/xfs_inode_item.h
+++ b/fs/xfs/xfs_inode_item.h
@@ -144,12 +144,6 @@ typedef struct xfs_inode_log_item {
 						      data exts */
 	struct xfs_bmbt_rec	*ili_aextents_buf; /* array of logged
 						      attr exts */
-	unsigned int            ili_pushbuf_flag;  /* one bit used in push_ail */
-
-#ifdef DEBUG
-	uint64_t                ili_push_owner;    /* one who sets pushbuf_flag
-						      above gets to push the buf */
-#endif
 #ifdef XFS_TRANS_DEBUG
 	int			ili_root_size;
 	char			*ili_orig_root;
diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c
index 62efab2f3839..b1b801e4a28e 100644
--- a/fs/xfs/xfs_itable.c
+++ b/fs/xfs/xfs_itable.c
@@ -106,6 +106,7 @@ xfs_bulkstat_one_iget(
 	buf->bs_dmevmask = dic->di_dmevmask;
 	buf->bs_dmstate = dic->di_dmstate;
 	buf->bs_aextents = dic->di_anextents;
+	buf->bs_forkoff = XFS_IFORK_BOFF(ip);
 
 	switch (dic->di_format) {
 	case XFS_DINODE_FMT_DEV:
@@ -176,6 +177,7 @@ xfs_bulkstat_one_dinode(
 	buf->bs_dmevmask = be32_to_cpu(dic->di_dmevmask);
 	buf->bs_dmstate = be16_to_cpu(dic->di_dmstate);
 	buf->bs_aextents = be16_to_cpu(dic->di_anextents);
+	buf->bs_forkoff = XFS_DFORK_BOFF(dic);
 
 	switch (dic->di_format) {
 	case XFS_DINODE_FMT_DEV:
@@ -408,8 +410,10 @@ xfs_bulkstat(
 		(XFS_INODE_CLUSTER_SIZE(mp) >> mp->m_sb.sb_inodelog);
 	nimask = ~(nicluster - 1);
 	nbcluster = nicluster >> mp->m_sb.sb_inopblog;
-	irbuf = kmem_zalloc_greedy(&irbsize, PAGE_SIZE, PAGE_SIZE * 4,
-				   KM_SLEEP | KM_MAYFAIL | KM_LARGE);
+	irbuf = kmem_zalloc_greedy(&irbsize, PAGE_SIZE, PAGE_SIZE * 4);
+	if (!irbuf)
+		return ENOMEM;
+
 	nirbuf = irbsize / sizeof(*irbuf);
 
 	/*
@@ -420,9 +424,7 @@ xfs_bulkstat(
 	while (XFS_BULKSTAT_UBLEFT(ubleft) && agno < mp->m_sb.sb_agcount) {
 		cond_resched();
 		bp = NULL;
-		down_read(&mp->m_peraglock);
 		error = xfs_ialloc_read_agi(mp, NULL, agno, &agbp);
-		up_read(&mp->m_peraglock);
 		if (error) {
 			/*
 			 * Skip this allocation group and go to the next one.
@@ -729,7 +731,7 @@ xfs_bulkstat(
 	/*
 	 * Done, we're either out of filesystem or space to put the data.
 	 */
-	kmem_free(irbuf);
+	kmem_free_large(irbuf);
 	*ubcountp = ubelem;
 	/*
 	 * Found some inodes, return them now and return the error next time.
@@ -849,9 +851,7 @@ xfs_inumbers(
 	agbp = NULL;
 	while (left > 0 && agno < mp->m_sb.sb_agcount) {
 		if (agbp == NULL) {
-			down_read(&mp->m_peraglock);
 			error = xfs_ialloc_read_agi(mp, NULL, agno, &agbp);
-			up_read(&mp->m_peraglock);
 			if (error) {
 				/*
 				 * If we can't read the AGI of this ag,
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 600b5b06aaeb..e8fba92d7cd9 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -50,7 +50,6 @@ kmem_zone_t	*xfs_log_ticket_zone;
 	  (off) += (bytes);}
 
 /* Local miscellaneous function prototypes */
-STATIC int	 xlog_bdstrat_cb(struct xfs_buf *);
 STATIC int	 xlog_commit_record(xfs_mount_t *mp, xlog_ticket_t *ticket,
 				    xlog_in_core_t **, xfs_lsn_t *);
 STATIC xlog_t *  xlog_alloc_log(xfs_mount_t	*mp,
@@ -61,7 +60,7 @@ STATIC int	 xlog_space_left(xlog_t *log, int cycle, int bytes);
 STATIC int	 xlog_sync(xlog_t *log, xlog_in_core_t *iclog);
 STATIC void	 xlog_dealloc_log(xlog_t *log);
 STATIC int	 xlog_write(xfs_mount_t *mp, xfs_log_iovec_t region[],
-			    int nentries, xfs_log_ticket_t tic,
+			    int nentries, struct xlog_ticket *tic,
 			    xfs_lsn_t *start_lsn,
 			    xlog_in_core_t **commit_iclog,
 			    uint flags);
@@ -80,11 +79,6 @@ STATIC int  xlog_state_release_iclog(xlog_t		*log,
 STATIC void xlog_state_switch_iclogs(xlog_t		*log,
 				     xlog_in_core_t *iclog,
 				     int		eventual_size);
-STATIC int  xlog_state_sync(xlog_t			*log,
-			    xfs_lsn_t 			lsn,
-			    uint			flags,
-			    int				*log_flushed);
-STATIC int  xlog_state_sync_all(xlog_t *log, uint flags, int *log_flushed);
 STATIC void xlog_state_want_sync(xlog_t	*log, xlog_in_core_t *iclog);
 
 /* local functions to manipulate grant head */
@@ -249,14 +243,14 @@ xlog_tic_add_region(xlog_ticket_t *tic, uint len, uint type)
  * out when the next write occurs.
  */
 xfs_lsn_t
-xfs_log_done(xfs_mount_t	*mp,
-	     xfs_log_ticket_t	xtic,
-	     void		**iclog,
-	     uint		flags)
+xfs_log_done(
+	struct xfs_mount	*mp,
+	struct xlog_ticket	*ticket,
+	struct xlog_in_core	**iclog,
+	uint			flags)
 {
-	xlog_t		*log    = mp->m_log;
-	xlog_ticket_t	*ticket = (xfs_log_ticket_t) xtic;
-	xfs_lsn_t	lsn	= 0;
+	struct log		*log = mp->m_log;
+	xfs_lsn_t		lsn = 0;
 
 	if (XLOG_FORCED_SHUTDOWN(log) ||
 	    /*
@@ -264,8 +258,7 @@ xfs_log_done(xfs_mount_t	*mp,
 	     * If we get an error, just continue and give back the log ticket.
 	     */
 	    (((ticket->t_flags & XLOG_TIC_INITED) == 0) &&
-	     (xlog_commit_record(mp, ticket,
-				 (xlog_in_core_t **)iclog, &lsn)))) {
+	     (xlog_commit_record(mp, ticket, iclog, &lsn)))) {
 		lsn = (xfs_lsn_t) -1;
 		if (ticket->t_flags & XLOG_TIC_PERM_RESERV) {
 			flags |= XFS_LOG_REL_PERM_RESERV;
@@ -295,67 +288,8 @@ xfs_log_done(xfs_mount_t	*mp,
 	}
 
 	return lsn;
-}	/* xfs_log_done */
-
-
-/*
- * Force the in-core log to disk.  If flags == XFS_LOG_SYNC,
- *	the force is done synchronously.
- *
- * Asynchronous forces are implemented by setting the WANT_SYNC
- * bit in the appropriate in-core log and then returning.
- *
- * Synchronous forces are implemented with a signal variable. All callers
- * to force a given lsn to disk will wait on a the sv attached to the
- * specific in-core log.  When given in-core log finally completes its
- * write to disk, that thread will wake up all threads waiting on the
- * sv.
- */
-int
-_xfs_log_force(
-	xfs_mount_t	*mp,
-	xfs_lsn_t	lsn,
-	uint		flags,
-	int		*log_flushed)
-{
-	xlog_t		*log = mp->m_log;
-	int		dummy;
-
-	if (!log_flushed)
-		log_flushed = &dummy;
-
-	ASSERT(flags & XFS_LOG_FORCE);
-
-	XFS_STATS_INC(xs_log_force);
-
-	if (log->l_flags & XLOG_IO_ERROR)
-		return XFS_ERROR(EIO);
-	if (lsn == 0)
-		return xlog_state_sync_all(log, flags, log_flushed);
-	else
-		return xlog_state_sync(log, lsn, flags, log_flushed);
-}	/* _xfs_log_force */
-
-/*
- * Wrapper for _xfs_log_force(), to be used when caller doesn't care
- * about errors or whether the log was flushed or not. This is the normal
- * interface to use when trying to unpin items or move the log forward.
- */
-void
-xfs_log_force(
-	xfs_mount_t	*mp,
-	xfs_lsn_t	lsn,
-	uint		flags)
-{
-	int	error;
-	error = _xfs_log_force(mp, lsn, flags, NULL);
-	if (error) {
-		xfs_fs_cmn_err(CE_WARN, mp, "xfs_log_force: "
-			"error %d returned.", error);
-	}
 }
 
-
 /*
  * Attaches a new iclog I/O completion callback routine during
  * transaction commit.  If the log is in error state, a non-zero
@@ -363,11 +297,11 @@ xfs_log_force(
  * executing the callback at an appropriate time.
  */
 int
-xfs_log_notify(xfs_mount_t	  *mp,		/* mount of partition */
-	       void		  *iclog_hndl,	/* iclog to hang callback off */
-	       xfs_log_callback_t *cb)
+xfs_log_notify(
+	struct xfs_mount	*mp,
+	struct xlog_in_core	*iclog,
+	xfs_log_callback_t	*cb)
 {
-	xlog_in_core_t	  *iclog = (xlog_in_core_t *)iclog_hndl;
 	int	abortflg;
 
 	spin_lock(&iclog->ic_callback_lock);
@@ -381,16 +315,14 @@ xfs_log_notify(xfs_mount_t	  *mp,		/* mount of partition */
 	}
 	spin_unlock(&iclog->ic_callback_lock);
 	return abortflg;
-}	/* xfs_log_notify */
+}
 
 int
-xfs_log_release_iclog(xfs_mount_t *mp,
-		      void	  *iclog_hndl)
+xfs_log_release_iclog(
+	struct xfs_mount	*mp,
+	struct xlog_in_core	*iclog)
 {
-	xlog_t *log = mp->m_log;
-	xlog_in_core_t	  *iclog = (xlog_in_core_t *)iclog_hndl;
-
-	if (xlog_state_release_iclog(log, iclog)) {
+	if (xlog_state_release_iclog(mp->m_log, iclog)) {
 		xfs_force_shutdown(mp, SHUTDOWN_LOG_IO_ERROR);
 		return EIO;
 	}
@@ -409,17 +341,18 @@ xfs_log_release_iclog(xfs_mount_t *mp,
  * reservation, we prevent over allocation problems.
  */
 int
-xfs_log_reserve(xfs_mount_t	 *mp,
-		int		 unit_bytes,
-		int		 cnt,
-		xfs_log_ticket_t *ticket,
-		__uint8_t	 client,
-		uint		 flags,
-		uint		 t_type)
+xfs_log_reserve(
+	struct xfs_mount	*mp,
+	int		 	unit_bytes,
+	int		 	cnt,
+	struct xlog_ticket	**ticket,
+	__uint8_t	 	client,
+	uint		 	flags,
+	uint		 	t_type)
 {
-	xlog_t		*log = mp->m_log;
-	xlog_ticket_t	*internal_ticket;
-	int		retval = 0;
+	struct log		*log = mp->m_log;
+	struct xlog_ticket	*internal_ticket;
+	int			retval = 0;
 
 	ASSERT(client == XFS_TRANSACTION || client == XFS_LOG);
 	ASSERT((flags & XFS_LOG_NOSLEEP) == 0);
@@ -432,7 +365,7 @@ xfs_log_reserve(xfs_mount_t	 *mp,
 
 	if (*ticket != NULL) {
 		ASSERT(flags & XFS_LOG_PERM_RESERV);
-		internal_ticket = (xlog_ticket_t *)*ticket;
+		internal_ticket = *ticket;
 
 		trace_xfs_log_reserve(log, internal_ticket);
 
@@ -584,7 +517,7 @@ xfs_log_unmount_write(xfs_mount_t *mp)
 	xlog_in_core_t	 *first_iclog;
 #endif
 	xfs_log_iovec_t  reg[1];
-	xfs_log_ticket_t tic = NULL;
+	xlog_ticket_t	*tic = NULL;
 	xfs_lsn_t	 lsn;
 	int		 error;
 
@@ -602,7 +535,7 @@ xfs_log_unmount_write(xfs_mount_t *mp)
 	if (mp->m_flags & XFS_MOUNT_RDONLY)
 		return 0;
 
-	error = _xfs_log_force(mp, 0, XFS_LOG_FORCE|XFS_LOG_SYNC, NULL);
+	error = _xfs_log_force(mp, XFS_LOG_SYNC, NULL);
 	ASSERT(error || !(XLOG_FORCED_SHUTDOWN(log)));
 
 #ifdef DEBUG
@@ -618,7 +551,7 @@ xfs_log_unmount_write(xfs_mount_t *mp)
 	if (! (XLOG_FORCED_SHUTDOWN(log))) {
 		reg[0].i_addr = (void*)&magic;
 		reg[0].i_len  = sizeof(magic);
-		XLOG_VEC_SET_TYPE(&reg[0], XLOG_REG_TYPE_UNMOUNT);
+		reg[0].i_type = XLOG_REG_TYPE_UNMOUNT;
 
 		error = xfs_log_reserve(mp, 600, 1, &tic,
 					XFS_LOG, 0, XLOG_UNMOUNT_REC_TYPE);
@@ -721,24 +654,24 @@ xfs_log_unmount(xfs_mount_t *mp)
  * transaction occur with one call to xfs_log_write().
  */
 int
-xfs_log_write(xfs_mount_t *	mp,
-	      xfs_log_iovec_t	reg[],
-	      int		nentries,
-	      xfs_log_ticket_t	tic,
-	      xfs_lsn_t		*start_lsn)
+xfs_log_write(
+	struct xfs_mount	*mp,
+	struct xfs_log_iovec	reg[],
+	int			nentries,
+	struct xlog_ticket	*tic,
+	xfs_lsn_t		*start_lsn)
 {
-	int	error;
-	xlog_t *log = mp->m_log;
+	struct log		*log = mp->m_log;
+	int			error;
 
 	if (XLOG_FORCED_SHUTDOWN(log))
 		return XFS_ERROR(EIO);
 
-	if ((error = xlog_write(mp, reg, nentries, tic, start_lsn, NULL, 0))) {
+	error = xlog_write(mp, reg, nentries, tic, start_lsn, NULL, 0);
+	if (error)
 		xfs_force_shutdown(mp, SHUTDOWN_LOG_IO_ERROR);
-	}
 	return error;
-}	/* xfs_log_write */
-
+}
 
 void
 xfs_log_move_tail(xfs_mount_t	*mp,
@@ -988,35 +921,6 @@ xlog_iodone(xfs_buf_t *bp)
 }	/* xlog_iodone */
 
 /*
- * The bdstrat callback function for log bufs. This gives us a central
- * place to trap bufs in case we get hit by a log I/O error and need to
- * shutdown. Actually, in practice, even when we didn't get a log error,
- * we transition the iclogs to IOERROR state *after* flushing all existing
- * iclogs to disk. This is because we don't want anymore new transactions to be
- * started or completed afterwards.
- */
-STATIC int
-xlog_bdstrat_cb(struct xfs_buf *bp)
-{
-	xlog_in_core_t *iclog;
-
-	iclog = XFS_BUF_FSPRIVATE(bp, xlog_in_core_t *);
-
-	if ((iclog->ic_state & XLOG_STATE_IOERROR) == 0) {
-	  /* note for irix bstrat will need  struct bdevsw passed
-	   * Fix the following macro if the code ever is merged
-	   */
-	    XFS_bdstrat(bp);
-		return 0;
-	}
-
-	XFS_BUF_ERROR(bp, EIO);
-	XFS_BUF_STALE(bp);
-	xfs_biodone(bp);
-	return XFS_ERROR(EIO);
-}
-
-/*
  * Return size of each in-core log record buffer.
  *
  * All machines get 8 x 32kB buffers by default, unless tuned otherwise.
@@ -1158,7 +1062,6 @@ xlog_alloc_log(xfs_mount_t	*mp,
 	if (!bp)
 		goto out_free_log;
 	XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone);
-	XFS_BUF_SET_BDSTRAT_FUNC(bp, xlog_bdstrat_cb);
 	XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1);
 	ASSERT(XFS_BUF_ISBUSY(bp));
 	ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
@@ -1196,7 +1099,6 @@ xlog_alloc_log(xfs_mount_t	*mp,
 		if (!XFS_BUF_CPSEMA(bp))
 			ASSERT(0);
 		XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone);
-		XFS_BUF_SET_BDSTRAT_FUNC(bp, xlog_bdstrat_cb);
 		XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1);
 		iclog->ic_bp = bp;
 		iclog->ic_data = bp->b_addr;
@@ -1268,7 +1170,7 @@ xlog_commit_record(xfs_mount_t  *mp,
 
 	reg[0].i_addr = NULL;
 	reg[0].i_len = 0;
-	XLOG_VEC_SET_TYPE(&reg[0], XLOG_REG_TYPE_COMMIT);
+	reg[0].i_type = XLOG_REG_TYPE_COMMIT;
 
 	ASSERT_ALWAYS(iclog);
 	if ((error = xlog_write(mp, reg, 1, ticket, commitlsnp,
@@ -1343,6 +1245,37 @@ xlog_grant_push_ail(xfs_mount_t	*mp,
 	    xfs_trans_ail_push(log->l_ailp, threshold_lsn);
 }	/* xlog_grant_push_ail */
 
+/*
+ * The bdstrat callback function for log bufs. This gives us a central
+ * place to trap bufs in case we get hit by a log I/O error and need to
+ * shutdown. Actually, in practice, even when we didn't get a log error,
+ * we transition the iclogs to IOERROR state *after* flushing all existing
+ * iclogs to disk. This is because we don't want anymore new transactions to be
+ * started or completed afterwards.
+ */
+STATIC int
+xlog_bdstrat(
+	struct xfs_buf		*bp)
+{
+	struct xlog_in_core	*iclog;
+
+	iclog = XFS_BUF_FSPRIVATE(bp, xlog_in_core_t *);
+	if (iclog->ic_state & XLOG_STATE_IOERROR) {
+		XFS_BUF_ERROR(bp, EIO);
+		XFS_BUF_STALE(bp);
+		xfs_biodone(bp);
+		/*
+		 * It would seem logical to return EIO here, but we rely on
+		 * the log state machine to propagate I/O errors instead of
+		 * doing it here.
+		 */
+		return 0;
+	}
+
+	bp->b_flags |= _XBF_RUN_QUEUES;
+	xfs_buf_iorequest(bp);
+	return 0;
+}
 
 /*
  * Flush out the in-core log (iclog) to the on-disk log in an asynchronous 
@@ -1462,7 +1395,7 @@ xlog_sync(xlog_t		*log,
 	 */
 	XFS_BUF_WRITE(bp);
 
-	if ((error = XFS_bwrite(bp))) {
+	if ((error = xlog_bdstrat(bp))) {
 		xfs_ioerror_alert("xlog_sync", log->l_mp, bp,
 				  XFS_BUF_ADDR(bp));
 		return error;
@@ -1502,7 +1435,7 @@ xlog_sync(xlog_t		*log,
 		/* account for internal log which doesn't start at block #0 */
 		XFS_BUF_SET_ADDR(bp, XFS_BUF_ADDR(bp) + log->l_logBBstart);
 		XFS_BUF_WRITE(bp);
-		if ((error = XFS_bwrite(bp))) {
+		if ((error = xlog_bdstrat(bp))) {
 			xfs_ioerror_alert("xlog_sync (split)", log->l_mp,
 					  bp, XFS_BUF_ADDR(bp));
 			return error;
@@ -1707,16 +1640,16 @@ xlog_print_tic_res(xfs_mount_t *mp, xlog_ticket_t *ticket)
  *	bytes have been written out.
  */
 STATIC int
-xlog_write(xfs_mount_t *	mp,
-	   xfs_log_iovec_t	reg[],
-	   int			nentries,
-	   xfs_log_ticket_t	tic,
-	   xfs_lsn_t		*start_lsn,
-	   xlog_in_core_t	**commit_iclog,
-	   uint			flags)
+xlog_write(
+	struct xfs_mount	*mp,
+	struct xfs_log_iovec	reg[],
+	int			nentries,
+	struct xlog_ticket	*ticket,
+	xfs_lsn_t		*start_lsn,
+	struct xlog_in_core	**commit_iclog,
+	uint			flags)
 {
     xlog_t	     *log = mp->m_log;
-    xlog_ticket_t    *ticket = (xlog_ticket_t *)tic;
     xlog_in_core_t   *iclog = NULL;  /* ptr to current in-core log */
     xlog_op_header_t *logop_head;    /* ptr to log operation header */
     __psint_t	     ptr;	     /* copy address into data region */
@@ -1830,7 +1763,7 @@ xlog_write(xfs_mount_t *	mp,
 	    default:
 		xfs_fs_cmn_err(CE_WARN, mp,
 		    "Bad XFS transaction clientid 0x%x in ticket 0x%p",
-		    logop_head->oh_clientid, tic);
+		    logop_head->oh_clientid, ticket);
 		return XFS_ERROR(EIO);
 	    }
 
@@ -2854,7 +2787,6 @@ xlog_state_switch_iclogs(xlog_t		*log,
 	log->l_iclog = iclog->ic_next;
 }	/* xlog_state_switch_iclogs */
 
-
 /*
  * Write out all data in the in-core log as of this exact moment in time.
  *
@@ -2882,11 +2814,17 @@ xlog_state_switch_iclogs(xlog_t		*log,
  *	   b) when we return from flushing out this iclog, it is still
  *		not in the active nor dirty state.
  */
-STATIC int
-xlog_state_sync_all(xlog_t *log, uint flags, int *log_flushed)
+int
+_xfs_log_force(
+	struct xfs_mount	*mp,
+	uint			flags,
+	int			*log_flushed)
 {
-	xlog_in_core_t	*iclog;
-	xfs_lsn_t	lsn;
+	struct log		*log = mp->m_log;
+	struct xlog_in_core	*iclog;
+	xfs_lsn_t		lsn;
+
+	XFS_STATS_INC(xs_log_force);
 
 	spin_lock(&log->l_icloglock);
 
@@ -2932,7 +2870,9 @@ xlog_state_sync_all(xlog_t *log, uint flags, int *log_flushed)
 
 				if (xlog_state_release_iclog(log, iclog))
 					return XFS_ERROR(EIO);
-				*log_flushed = 1;
+
+				if (log_flushed)
+					*log_flushed = 1;
 				spin_lock(&log->l_icloglock);
 				if (be64_to_cpu(iclog->ic_header.h_lsn) == lsn &&
 				    iclog->ic_state != XLOG_STATE_DIRTY)
@@ -2976,19 +2916,37 @@ maybe_sleep:
 		 */
 		if (iclog->ic_state & XLOG_STATE_IOERROR)
 			return XFS_ERROR(EIO);
-		*log_flushed = 1;
-
+		if (log_flushed)
+			*log_flushed = 1;
 	} else {
 
 no_sleep:
 		spin_unlock(&log->l_icloglock);
 	}
 	return 0;
-}	/* xlog_state_sync_all */
+}
 
+/*
+ * Wrapper for _xfs_log_force(), to be used when caller doesn't care
+ * about errors or whether the log was flushed or not. This is the normal
+ * interface to use when trying to unpin items or move the log forward.
+ */
+void
+xfs_log_force(
+	xfs_mount_t	*mp,
+	uint		flags)
+{
+	int	error;
+
+	error = _xfs_log_force(mp, flags, NULL);
+	if (error) {
+		xfs_fs_cmn_err(CE_WARN, mp, "xfs_log_force: "
+			"error %d returned.", error);
+	}
+}
 
 /*
- * Used by code which implements synchronous log forces.
+ * Force the in-core log to disk for a specific LSN.
  *
  * Find in-core log with lsn.
  *	If it is in the DIRTY state, just return.
@@ -2996,109 +2954,142 @@ no_sleep:
  *		state and go to sleep or return.
  *	If it is in any other state, go to sleep or return.
  *
- * If filesystem activity goes to zero, the iclog will get flushed only by
- * bdflush().
+ * Synchronous forces are implemented with a signal variable. All callers
+ * to force a given lsn to disk will wait on a the sv attached to the
+ * specific in-core log.  When given in-core log finally completes its
+ * write to disk, that thread will wake up all threads waiting on the
+ * sv.
  */
-STATIC int
-xlog_state_sync(xlog_t	  *log,
-		xfs_lsn_t lsn,
-		uint	  flags,
-		int	  *log_flushed)
+int
+_xfs_log_force_lsn(
+	struct xfs_mount	*mp,
+	xfs_lsn_t		lsn,
+	uint			flags,
+	int			*log_flushed)
 {
-    xlog_in_core_t	*iclog;
-    int			already_slept = 0;
-
-try_again:
-    spin_lock(&log->l_icloglock);
-    iclog = log->l_iclog;
+	struct log		*log = mp->m_log;
+	struct xlog_in_core	*iclog;
+	int			already_slept = 0;
 
-    if (iclog->ic_state & XLOG_STATE_IOERROR) {
-	    spin_unlock(&log->l_icloglock);
-	    return XFS_ERROR(EIO);
-    }
+	ASSERT(lsn != 0);
 
-    do {
-	if (be64_to_cpu(iclog->ic_header.h_lsn) != lsn) {
-		iclog = iclog->ic_next;
-		continue;
-	}
+	XFS_STATS_INC(xs_log_force);
 
-	if (iclog->ic_state == XLOG_STATE_DIRTY) {
+try_again:
+	spin_lock(&log->l_icloglock);
+	iclog = log->l_iclog;
+	if (iclog->ic_state & XLOG_STATE_IOERROR) {
 		spin_unlock(&log->l_icloglock);
-		return 0;
+		return XFS_ERROR(EIO);
 	}
 
-	if (iclog->ic_state == XLOG_STATE_ACTIVE) {
-		/*
-		 * We sleep here if we haven't already slept (e.g.
-		 * this is the first time we've looked at the correct
-		 * iclog buf) and the buffer before us is going to
-		 * be sync'ed. The reason for this is that if we
-		 * are doing sync transactions here, by waiting for
-		 * the previous I/O to complete, we can allow a few
-		 * more transactions into this iclog before we close
-		 * it down.
-		 *
-		 * Otherwise, we mark the buffer WANT_SYNC, and bump
-		 * up the refcnt so we can release the log (which drops
-		 * the ref count).  The state switch keeps new transaction
-		 * commits from using this buffer.  When the current commits
-		 * finish writing into the buffer, the refcount will drop to
-		 * zero and the buffer will go out then.
-		 */
-		if (!already_slept &&
-		    (iclog->ic_prev->ic_state & (XLOG_STATE_WANT_SYNC |
-						 XLOG_STATE_SYNCING))) {
-			ASSERT(!(iclog->ic_state & XLOG_STATE_IOERROR));
-			XFS_STATS_INC(xs_log_force_sleep);
-			sv_wait(&iclog->ic_prev->ic_write_wait, PSWP,
-				&log->l_icloglock, s);
-			*log_flushed = 1;
-			already_slept = 1;
-			goto try_again;
-		} else {
+	do {
+		if (be64_to_cpu(iclog->ic_header.h_lsn) != lsn) {
+			iclog = iclog->ic_next;
+			continue;
+		}
+
+		if (iclog->ic_state == XLOG_STATE_DIRTY) {
+			spin_unlock(&log->l_icloglock);
+			return 0;
+		}
+
+		if (iclog->ic_state == XLOG_STATE_ACTIVE) {
+			/*
+			 * We sleep here if we haven't already slept (e.g.
+			 * this is the first time we've looked at the correct
+			 * iclog buf) and the buffer before us is going to
+			 * be sync'ed. The reason for this is that if we
+			 * are doing sync transactions here, by waiting for
+			 * the previous I/O to complete, we can allow a few
+			 * more transactions into this iclog before we close
+			 * it down.
+			 *
+			 * Otherwise, we mark the buffer WANT_SYNC, and bump
+			 * up the refcnt so we can release the log (which
+			 * drops the ref count).  The state switch keeps new
+			 * transaction commits from using this buffer.  When
+			 * the current commits finish writing into the buffer,
+			 * the refcount will drop to zero and the buffer will
+			 * go out then.
+			 */
+			if (!already_slept &&
+			    (iclog->ic_prev->ic_state &
+			     (XLOG_STATE_WANT_SYNC | XLOG_STATE_SYNCING))) {
+				ASSERT(!(iclog->ic_state & XLOG_STATE_IOERROR));
+
+				XFS_STATS_INC(xs_log_force_sleep);
+
+				sv_wait(&iclog->ic_prev->ic_write_wait,
+					PSWP, &log->l_icloglock, s);
+				if (log_flushed)
+					*log_flushed = 1;
+				already_slept = 1;
+				goto try_again;
+			}
 			atomic_inc(&iclog->ic_refcnt);
 			xlog_state_switch_iclogs(log, iclog, 0);
 			spin_unlock(&log->l_icloglock);
 			if (xlog_state_release_iclog(log, iclog))
 				return XFS_ERROR(EIO);
-			*log_flushed = 1;
+			if (log_flushed)
+				*log_flushed = 1;
 			spin_lock(&log->l_icloglock);
 		}
-	}
 
-	if ((flags & XFS_LOG_SYNC) && /* sleep */
-	    !(iclog->ic_state & (XLOG_STATE_ACTIVE | XLOG_STATE_DIRTY))) {
+		if ((flags & XFS_LOG_SYNC) && /* sleep */
+		    !(iclog->ic_state &
+		      (XLOG_STATE_ACTIVE | XLOG_STATE_DIRTY))) {
+			/*
+			 * Don't wait on completion if we know that we've
+			 * gotten a log write error.
+			 */
+			if (iclog->ic_state & XLOG_STATE_IOERROR) {
+				spin_unlock(&log->l_icloglock);
+				return XFS_ERROR(EIO);
+			}
+			XFS_STATS_INC(xs_log_force_sleep);
+			sv_wait(&iclog->ic_force_wait, PSWP, &log->l_icloglock, s);
+			/*
+			 * No need to grab the log lock here since we're
+			 * only deciding whether or not to return EIO
+			 * and the memory read should be atomic.
+			 */
+			if (iclog->ic_state & XLOG_STATE_IOERROR)
+				return XFS_ERROR(EIO);
 
-		/*
-		 * Don't wait on completion if we know that we've
-		 * gotten a log write error.
-		 */
-		if (iclog->ic_state & XLOG_STATE_IOERROR) {
+			if (log_flushed)
+				*log_flushed = 1;
+		} else {		/* just return */
 			spin_unlock(&log->l_icloglock);
-			return XFS_ERROR(EIO);
 		}
-		XFS_STATS_INC(xs_log_force_sleep);
-		sv_wait(&iclog->ic_force_wait, PSWP, &log->l_icloglock, s);
-		/*
-		 * No need to grab the log lock here since we're
-		 * only deciding whether or not to return EIO
-		 * and the memory read should be atomic.
-		 */
-		if (iclog->ic_state & XLOG_STATE_IOERROR)
-			return XFS_ERROR(EIO);
-		*log_flushed = 1;
-	} else {		/* just return */
-		spin_unlock(&log->l_icloglock);
-	}
-	return 0;
 
-    } while (iclog != log->l_iclog);
+		return 0;
+	} while (iclog != log->l_iclog);
 
-    spin_unlock(&log->l_icloglock);
-    return 0;
-}	/* xlog_state_sync */
+	spin_unlock(&log->l_icloglock);
+	return 0;
+}
+
+/*
+ * Wrapper for _xfs_log_force_lsn(), to be used when caller doesn't care
+ * about errors or whether the log was flushed or not. This is the normal
+ * interface to use when trying to unpin items or move the log forward.
+ */
+void
+xfs_log_force_lsn(
+	xfs_mount_t	*mp,
+	xfs_lsn_t	lsn,
+	uint		flags)
+{
+	int	error;
 
+	error = _xfs_log_force_lsn(mp, lsn, flags, NULL);
+	if (error) {
+		xfs_fs_cmn_err(CE_WARN, mp, "xfs_log_force: "
+			"error %d returned.", error);
+	}
+}
 
 /*
  * Called when we want to mark the current iclog as being ready to sync to
@@ -3463,7 +3454,6 @@ xfs_log_force_umount(
 	xlog_ticket_t	*tic;
 	xlog_t		*log;
 	int		retval;
-	int		dummy;
 
 	log = mp->m_log;
 
@@ -3537,13 +3527,14 @@ xfs_log_force_umount(
 	}
 	spin_unlock(&log->l_grant_lock);
 
-	if (! (log->l_iclog->ic_state & XLOG_STATE_IOERROR)) {
+	if (!(log->l_iclog->ic_state & XLOG_STATE_IOERROR)) {
 		ASSERT(!logerror);
 		/*
 		 * Force the incore logs to disk before shutting the
 		 * log down completely.
 		 */
-		xlog_state_sync_all(log, XFS_LOG_FORCE|XFS_LOG_SYNC, &dummy);
+		_xfs_log_force(mp, XFS_LOG_SYNC, NULL);
+
 		spin_lock(&log->l_icloglock);
 		retval = xlog_state_ioerror(log);
 		spin_unlock(&log->l_icloglock);
diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h
index d0c9baa50b1a..97a24c7795a4 100644
--- a/fs/xfs/xfs_log.h
+++ b/fs/xfs/xfs_log.h
@@ -70,14 +70,8 @@ static inline xfs_lsn_t	_lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2)
  * Flags to xfs_log_force()
  *
  *	XFS_LOG_SYNC:	Synchronous force in-core log to disk
- *	XFS_LOG_FORCE:	Start in-core log write now.
- *	XFS_LOG_URGE:	Start write within some window of time.
- *
- * Note: Either XFS_LOG_FORCE or XFS_LOG_URGE must be set.
  */
 #define XFS_LOG_SYNC		0x1
-#define XFS_LOG_FORCE		0x2
-#define XFS_LOG_URGE		0x4
 
 #endif	/* __KERNEL__ */
 
@@ -110,16 +104,12 @@ static inline xfs_lsn_t	_lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2)
 #define XLOG_REG_TYPE_TRANSHDR		19
 #define XLOG_REG_TYPE_MAX		19
 
-#define XLOG_VEC_SET_TYPE(vecp, t) ((vecp)->i_type = (t))
-
 typedef struct xfs_log_iovec {
-	xfs_caddr_t		i_addr;		/* beginning address of region */
+	xfs_caddr_t	i_addr;		/* beginning address of region */
 	int		i_len;		/* length in bytes of region */
 	uint		i_type;		/* type of region */
 } xfs_log_iovec_t;
 
-typedef void* xfs_log_ticket_t;
-
 /*
  * Structure used to pass callback function and the function's argument
  * to the log manager.
@@ -134,18 +124,25 @@ typedef struct xfs_log_callback {
 #ifdef __KERNEL__
 /* Log manager interfaces */
 struct xfs_mount;
+struct xlog_in_core;
 struct xlog_ticket;
+
 xfs_lsn_t xfs_log_done(struct xfs_mount *mp,
-		       xfs_log_ticket_t ticket,
-		       void		**iclog,
+		       struct xlog_ticket *ticket,
+		       struct xlog_in_core **iclog,
 		       uint		flags);
 int	  _xfs_log_force(struct xfs_mount *mp,
-			 xfs_lsn_t	lsn,
 			 uint		flags,
 			 int		*log_forced);
 void	  xfs_log_force(struct xfs_mount	*mp,
-			xfs_lsn_t		lsn,
 			uint			flags);
+int	  _xfs_log_force_lsn(struct xfs_mount *mp,
+			     xfs_lsn_t		lsn,
+			     uint		flags,
+			     int		*log_forced);
+void	  xfs_log_force_lsn(struct xfs_mount	*mp,
+			    xfs_lsn_t		lsn,
+			    uint		flags);
 int	  xfs_log_mount(struct xfs_mount	*mp,
 			struct xfs_buftarg	*log_target,
 			xfs_daddr_t		start_block,
@@ -154,21 +151,21 @@ int	  xfs_log_mount_finish(struct xfs_mount *mp);
 void	  xfs_log_move_tail(struct xfs_mount	*mp,
 			    xfs_lsn_t		tail_lsn);
 int	  xfs_log_notify(struct xfs_mount	*mp,
-			 void			*iclog,
+			 struct xlog_in_core	*iclog,
 			 xfs_log_callback_t	*callback_entry);
 int	  xfs_log_release_iclog(struct xfs_mount *mp,
-			 void			 *iclog_hndl);
+			 struct xlog_in_core	 *iclog);
 int	  xfs_log_reserve(struct xfs_mount *mp,
 			  int		   length,
 			  int		   count,
-			  xfs_log_ticket_t *ticket,
+			  struct xlog_ticket **ticket,
 			  __uint8_t	   clientid,
 			  uint		   flags,
 			  uint		   t_type);
 int	  xfs_log_write(struct xfs_mount *mp,
 			xfs_log_iovec_t  region[],
 			int		 nentries,
-			xfs_log_ticket_t ticket,
+			struct xlog_ticket *ticket,
 			xfs_lsn_t	 *start_lsn);
 int	  xfs_log_unmount_write(struct xfs_mount *mp);
 void      xfs_log_unmount(struct xfs_mount *mp);
diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h
index d55662db7077..fd02a18facd5 100644
--- a/fs/xfs/xfs_log_priv.h
+++ b/fs/xfs/xfs_log_priv.h
@@ -443,14 +443,9 @@ typedef struct log {
 
 /* common routines */
 extern xfs_lsn_t xlog_assign_tail_lsn(struct xfs_mount *mp);
-extern int	 xlog_find_tail(xlog_t	*log,
-				xfs_daddr_t *head_blk,
-				xfs_daddr_t *tail_blk);
 extern int	 xlog_recover(xlog_t *log);
 extern int	 xlog_recover_finish(xlog_t *log);
 extern void	 xlog_pack_data(xlog_t *log, xlog_in_core_t *iclog, int);
-extern struct xfs_buf *xlog_get_bp(xlog_t *, int);
-extern void	 xlog_put_bp(struct xfs_buf *);
 
 extern kmem_zone_t	*xfs_log_ticket_zone;
 
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 69ac2e5ef20c..22e6efdc17ea 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -50,8 +50,6 @@
 
 STATIC int	xlog_find_zeroed(xlog_t *, xfs_daddr_t *);
 STATIC int	xlog_clear_stale_blocks(xlog_t *, xfs_lsn_t);
-STATIC void	xlog_recover_insert_item_backq(xlog_recover_item_t **q,
-					       xlog_recover_item_t *item);
 #if defined(DEBUG)
 STATIC void	xlog_recover_check_summary(xlog_t *);
 #else
@@ -68,7 +66,7 @@ STATIC void	xlog_recover_check_summary(xlog_t *);
 	((bbs + (log)->l_sectbb_mask + 1) & ~(log)->l_sectbb_mask) : (bbs) )
 #define XLOG_SECTOR_ROUNDDOWN_BLKNO(log, bno)	((bno) & ~(log)->l_sectbb_mask)
 
-xfs_buf_t *
+STATIC xfs_buf_t *
 xlog_get_bp(
 	xlog_t		*log,
 	int		nbblks)
@@ -88,7 +86,7 @@ xlog_get_bp(
 	return xfs_buf_get_noaddr(BBTOB(nbblks), log->l_mp->m_logdev_targp);
 }
 
-void
+STATIC void
 xlog_put_bp(
 	xfs_buf_t	*bp)
 {
@@ -805,7 +803,7 @@ xlog_find_head(
  * We could speed up search by using current head_blk buffer, but it is not
  * available.
  */
-int
+STATIC int
 xlog_find_tail(
 	xlog_t			*log,
 	xfs_daddr_t		*head_blk,
@@ -1367,36 +1365,45 @@ xlog_clear_stale_blocks(
 
 STATIC xlog_recover_t *
 xlog_recover_find_tid(
-	xlog_recover_t		*q,
+	struct hlist_head	*head,
 	xlog_tid_t		tid)
 {
-	xlog_recover_t		*p = q;
+	xlog_recover_t		*trans;
+	struct hlist_node	*n;
 
-	while (p != NULL) {
-		if (p->r_log_tid == tid)
-		    break;
-		p = p->r_next;
+	hlist_for_each_entry(trans, n, head, r_list) {
+		if (trans->r_log_tid == tid)
+			return trans;
 	}
-	return p;
+	return NULL;
 }
 
 STATIC void
-xlog_recover_put_hashq(
-	xlog_recover_t		**q,
-	xlog_recover_t		*trans)
+xlog_recover_new_tid(
+	struct hlist_head	*head,
+	xlog_tid_t		tid,
+	xfs_lsn_t		lsn)
 {
-	trans->r_next = *q;
-	*q = trans;
+	xlog_recover_t		*trans;
+
+	trans = kmem_zalloc(sizeof(xlog_recover_t), KM_SLEEP);
+	trans->r_log_tid   = tid;
+	trans->r_lsn	   = lsn;
+	INIT_LIST_HEAD(&trans->r_itemq);
+
+	INIT_HLIST_NODE(&trans->r_list);
+	hlist_add_head(&trans->r_list, head);
 }
 
 STATIC void
 xlog_recover_add_item(
-	xlog_recover_item_t	**itemq)
+	struct list_head	*head)
 {
 	xlog_recover_item_t	*item;
 
 	item = kmem_zalloc(sizeof(xlog_recover_item_t), KM_SLEEP);
-	xlog_recover_insert_item_backq(itemq, item);
+	INIT_LIST_HEAD(&item->ri_list);
+	list_add_tail(&item->ri_list, head);
 }
 
 STATIC int
@@ -1409,8 +1416,7 @@ xlog_recover_add_to_cont_trans(
 	xfs_caddr_t		ptr, old_ptr;
 	int			old_len;
 
-	item = trans->r_itemq;
-	if (item == NULL) {
+	if (list_empty(&trans->r_itemq)) {
 		/* finish copying rest of trans header */
 		xlog_recover_add_item(&trans->r_itemq);
 		ptr = (xfs_caddr_t) &trans->r_theader +
@@ -1418,7 +1424,8 @@ xlog_recover_add_to_cont_trans(
 		memcpy(ptr, dp, len); /* d, s, l */
 		return 0;
 	}
-	item = item->ri_prev;
+	/* take the tail entry */
+	item = list_entry(trans->r_itemq.prev, xlog_recover_item_t, ri_list);
 
 	old_ptr = item->ri_buf[item->ri_cnt-1].i_addr;
 	old_len = item->ri_buf[item->ri_cnt-1].i_len;
@@ -1455,8 +1462,7 @@ xlog_recover_add_to_trans(
 
 	if (!len)
 		return 0;
-	item = trans->r_itemq;
-	if (item == NULL) {
+	if (list_empty(&trans->r_itemq)) {
 		/* we need to catch log corruptions here */
 		if (*(uint *)dp != XFS_TRANS_HEADER_MAGIC) {
 			xlog_warn("XFS: xlog_recover_add_to_trans: "
@@ -1474,12 +1480,15 @@ xlog_recover_add_to_trans(
 	memcpy(ptr, dp, len);
 	in_f = (xfs_inode_log_format_t *)ptr;
 
-	if (item->ri_prev->ri_total != 0 &&
-	     item->ri_prev->ri_total == item->ri_prev->ri_cnt) {
+	/* take the tail entry */
+	item = list_entry(trans->r_itemq.prev, xlog_recover_item_t, ri_list);
+	if (item->ri_total != 0 &&
+	     item->ri_total == item->ri_cnt) {
+		/* tail item is in use, get a new one */
 		xlog_recover_add_item(&trans->r_itemq);
+		item = list_entry(trans->r_itemq.prev,
+					xlog_recover_item_t, ri_list);
 	}
-	item = trans->r_itemq;
-	item = item->ri_prev;
 
 	if (item->ri_total == 0) {		/* first region to be added */
 		if (in_f->ilf_size == 0 ||
@@ -1504,96 +1513,29 @@ xlog_recover_add_to_trans(
 	return 0;
 }
 
-STATIC void
-xlog_recover_new_tid(
-	xlog_recover_t		**q,
-	xlog_tid_t		tid,
-	xfs_lsn_t		lsn)
-{
-	xlog_recover_t		*trans;
-
-	trans = kmem_zalloc(sizeof(xlog_recover_t), KM_SLEEP);
-	trans->r_log_tid   = tid;
-	trans->r_lsn	   = lsn;
-	xlog_recover_put_hashq(q, trans);
-}
-
-STATIC int
-xlog_recover_unlink_tid(
-	xlog_recover_t		**q,
-	xlog_recover_t		*trans)
-{
-	xlog_recover_t		*tp;
-	int			found = 0;
-
-	ASSERT(trans != NULL);
-	if (trans == *q) {
-		*q = (*q)->r_next;
-	} else {
-		tp = *q;
-		while (tp) {
-			if (tp->r_next == trans) {
-				found = 1;
-				break;
-			}
-			tp = tp->r_next;
-		}
-		if (!found) {
-			xlog_warn(
-			     "XFS: xlog_recover_unlink_tid: trans not found");
-			ASSERT(0);
-			return XFS_ERROR(EIO);
-		}
-		tp->r_next = tp->r_next->r_next;
-	}
-	return 0;
-}
-
-STATIC void
-xlog_recover_insert_item_backq(
-	xlog_recover_item_t	**q,
-	xlog_recover_item_t	*item)
-{
-	if (*q == NULL) {
-		item->ri_prev = item->ri_next = item;
-		*q = item;
-	} else {
-		item->ri_next		= *q;
-		item->ri_prev		= (*q)->ri_prev;
-		(*q)->ri_prev		= item;
-		item->ri_prev->ri_next	= item;
-	}
-}
-
-STATIC void
-xlog_recover_insert_item_frontq(
-	xlog_recover_item_t	**q,
-	xlog_recover_item_t	*item)
-{
-	xlog_recover_insert_item_backq(q, item);
-	*q = item;
-}
-
+/*
+ * Sort the log items in the transaction. Cancelled buffers need
+ * to be put first so they are processed before any items that might
+ * modify the buffers. If they are cancelled, then the modifications
+ * don't need to be replayed.
+ */
 STATIC int
 xlog_recover_reorder_trans(
 	xlog_recover_t		*trans)
 {
-	xlog_recover_item_t	*first_item, *itemq, *itemq_next;
-	xfs_buf_log_format_t	*buf_f;
-	ushort			flags = 0;
+	xlog_recover_item_t	*item, *n;
+	LIST_HEAD(sort_list);
 
-	first_item = itemq = trans->r_itemq;
-	trans->r_itemq = NULL;
-	do {
-		itemq_next = itemq->ri_next;
-		buf_f = (xfs_buf_log_format_t *)itemq->ri_buf[0].i_addr;
+	list_splice_init(&trans->r_itemq, &sort_list);
+	list_for_each_entry_safe(item, n, &sort_list, ri_list) {
+		xfs_buf_log_format_t	*buf_f;
 
-		switch (ITEM_TYPE(itemq)) {
+		buf_f = (xfs_buf_log_format_t *)item->ri_buf[0].i_addr;
+
+		switch (ITEM_TYPE(item)) {
 		case XFS_LI_BUF:
-			flags = buf_f->blf_flags;
-			if (!(flags & XFS_BLI_CANCEL)) {
-				xlog_recover_insert_item_frontq(&trans->r_itemq,
-								itemq);
+			if (!(buf_f->blf_flags & XFS_BLI_CANCEL)) {
+				list_move(&item->ri_list, &trans->r_itemq);
 				break;
 			}
 		case XFS_LI_INODE:
@@ -1601,7 +1543,7 @@ xlog_recover_reorder_trans(
 		case XFS_LI_QUOTAOFF:
 		case XFS_LI_EFD:
 		case XFS_LI_EFI:
-			xlog_recover_insert_item_backq(&trans->r_itemq, itemq);
+			list_move_tail(&item->ri_list, &trans->r_itemq);
 			break;
 		default:
 			xlog_warn(
@@ -1609,8 +1551,8 @@ xlog_recover_reorder_trans(
 			ASSERT(0);
 			return XFS_ERROR(EIO);
 		}
-		itemq = itemq_next;
-	} while (first_item != itemq);
+	}
+	ASSERT(list_empty(&sort_list));
 	return 0;
 }
 
@@ -2242,9 +2184,9 @@ xlog_recover_do_buffer_trans(
 	}
 
 	mp = log->l_mp;
-	buf_flags = XFS_BUF_LOCK;
+	buf_flags = XBF_LOCK;
 	if (!(flags & XFS_BLI_INODE_BUF))
-		buf_flags |= XFS_BUF_MAPPED;
+		buf_flags |= XBF_MAPPED;
 
 	bp = xfs_buf_read(mp->m_ddev_targp, blkno, len, buf_flags);
 	if (XFS_BUF_ISERROR(bp)) {
@@ -2346,7 +2288,7 @@ xlog_recover_do_inode_trans(
 	}
 
 	bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len,
-			  XFS_BUF_LOCK);
+			  XBF_LOCK);
 	if (XFS_BUF_ISERROR(bp)) {
 		xfs_ioerror_alert("xlog_recover_do..(read#2)", mp,
 				  bp, in_f->ilf_blkno);
@@ -2814,14 +2756,13 @@ xlog_recover_do_trans(
 	int			pass)
 {
 	int			error = 0;
-	xlog_recover_item_t	*item, *first_item;
+	xlog_recover_item_t	*item;
 
 	error = xlog_recover_reorder_trans(trans);
 	if (error)
 		return error;
 
-	first_item = item = trans->r_itemq;
-	do {
+	list_for_each_entry(item, &trans->r_itemq, ri_list) {
 		switch (ITEM_TYPE(item)) {
 		case XFS_LI_BUF:
 			error = xlog_recover_do_buffer_trans(log, item, pass);
@@ -2854,8 +2795,7 @@ xlog_recover_do_trans(
 
 		if (error)
 			return error;
-		item = item->ri_next;
-	} while (first_item != item);
+	}
 
 	return 0;
 }
@@ -2869,21 +2809,18 @@ STATIC void
 xlog_recover_free_trans(
 	xlog_recover_t		*trans)
 {
-	xlog_recover_item_t	*first_item, *item, *free_item;
+	xlog_recover_item_t	*item, *n;
 	int			i;
 
-	item = first_item = trans->r_itemq;
-	do {
-		free_item = item;
-		item = item->ri_next;
-		 /* Free the regions in the item. */
-		for (i = 0; i < free_item->ri_cnt; i++) {
-			kmem_free(free_item->ri_buf[i].i_addr);
-		}
+	list_for_each_entry_safe(item, n, &trans->r_itemq, ri_list) {
+		/* Free the regions in the item. */
+		list_del(&item->ri_list);
+		for (i = 0; i < item->ri_cnt; i++)
+			kmem_free(item->ri_buf[i].i_addr);
 		/* Free the item itself */
-		kmem_free(free_item->ri_buf);
-		kmem_free(free_item);
-	} while (first_item != item);
+		kmem_free(item->ri_buf);
+		kmem_free(item);
+	}
 	/* Free the transaction recover structure */
 	kmem_free(trans);
 }
@@ -2891,14 +2828,12 @@ xlog_recover_free_trans(
 STATIC int
 xlog_recover_commit_trans(
 	xlog_t			*log,
-	xlog_recover_t		**q,
 	xlog_recover_t		*trans,
 	int			pass)
 {
 	int			error;
 
-	if ((error = xlog_recover_unlink_tid(q, trans)))
-		return error;
+	hlist_del(&trans->r_list);
 	if ((error = xlog_recover_do_trans(log, trans, pass)))
 		return error;
 	xlog_recover_free_trans(trans);			/* no error */
@@ -2926,7 +2861,7 @@ xlog_recover_unmount_trans(
 STATIC int
 xlog_recover_process_data(
 	xlog_t			*log,
-	xlog_recover_t		*rhash[],
+	struct hlist_head	rhash[],
 	xlog_rec_header_t	*rhead,
 	xfs_caddr_t		dp,
 	int			pass)
@@ -2960,7 +2895,7 @@ xlog_recover_process_data(
 		}
 		tid = be32_to_cpu(ohead->oh_tid);
 		hash = XLOG_RHASH(tid);
-		trans = xlog_recover_find_tid(rhash[hash], tid);
+		trans = xlog_recover_find_tid(&rhash[hash], tid);
 		if (trans == NULL) {		   /* not found; add new tid */
 			if (ohead->oh_flags & XLOG_START_TRANS)
 				xlog_recover_new_tid(&rhash[hash], tid,
@@ -2978,7 +2913,7 @@ xlog_recover_process_data(
 			switch (flags) {
 			case XLOG_COMMIT_TRANS:
 				error = xlog_recover_commit_trans(log,
-						&rhash[hash], trans, pass);
+								trans, pass);
 				break;
 			case XLOG_UNMOUNT_TRANS:
 				error = xlog_recover_unmount_trans(trans);
@@ -3211,7 +3146,7 @@ xlog_recover_process_one_iunlink(
 	/*
 	 * Get the on disk inode to find the next inode in the bucket.
 	 */
-	error = xfs_itobp(mp, NULL, ip, &dip, &ibp, XFS_BUF_LOCK);
+	error = xfs_itobp(mp, NULL, ip, &dip, &ibp, XBF_LOCK);
 	if (error)
 		goto fail_iput;
 
@@ -3517,7 +3452,7 @@ xlog_do_recovery_pass(
 	int			error = 0, h_size;
 	int			bblks, split_bblks;
 	int			hblks, split_hblks, wrapped_hblks;
-	xlog_recover_t		*rhash[XLOG_RHASH_SIZE];
+	struct hlist_head	rhash[XLOG_RHASH_SIZE];
 
 	ASSERT(head_blk != tail_blk);
 
@@ -3978,8 +3913,7 @@ xlog_recover_finish(
 		 * case the unlink transactions would have problems
 		 * pushing the EFIs out of the way.
 		 */
-		xfs_log_force(log->l_mp, (xfs_lsn_t)0,
-			      (XFS_LOG_FORCE | XFS_LOG_SYNC));
+		xfs_log_force(log->l_mp, XFS_LOG_SYNC);
 
 		xlog_recover_process_iunlinks(log);
 
diff --git a/fs/xfs/xfs_log_recover.h b/fs/xfs/xfs_log_recover.h
index b22545555301..75d749207258 100644
--- a/fs/xfs/xfs_log_recover.h
+++ b/fs/xfs/xfs_log_recover.h
@@ -35,22 +35,21 @@
  * item headers are in ri_buf[0].  Additional buffers follow.
  */
 typedef struct xlog_recover_item {
-	struct xlog_recover_item *ri_next;
-	struct xlog_recover_item *ri_prev;
-	int			 ri_type;
-	int			 ri_cnt;	/* count of regions found */
-	int			 ri_total;	/* total regions */
-	xfs_log_iovec_t		 *ri_buf;	/* ptr to regions buffer */
+	struct list_head	ri_list;
+	int			ri_type;
+	int			ri_cnt;	/* count of regions found */
+	int			ri_total;	/* total regions */
+	xfs_log_iovec_t		*ri_buf;	/* ptr to regions buffer */
 } xlog_recover_item_t;
 
 struct xlog_tid;
 typedef struct xlog_recover {
-	struct xlog_recover *r_next;
-	xlog_tid_t	    r_log_tid;		/* log's transaction id */
-	xfs_trans_header_t  r_theader;		/* trans header for partial */
-	int		    r_state;		/* not needed */
-	xfs_lsn_t	    r_lsn;		/* xact lsn */
-	xlog_recover_item_t *r_itemq;		/* q for items */
+	struct hlist_node	r_list;
+	xlog_tid_t		r_log_tid;	/* log's transaction id */
+	xfs_trans_header_t	r_theader;	/* trans header for partial */
+	int			r_state;	/* not needed */
+	xfs_lsn_t		r_lsn;		/* xact lsn */
+	struct list_head	r_itemq;	/* q for items */
 } xlog_recover_t;
 
 #define ITEM_TYPE(i)	(*(ushort *)(i)->ri_buf[0].i_addr)
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index eb403b40e120..e79b56b4bca6 100644
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -201,6 +201,38 @@ xfs_uuid_unmount(
 
 
 /*
+ * Reference counting access wrappers to the perag structures.
+ */
+struct xfs_perag *
+xfs_perag_get(struct xfs_mount *mp, xfs_agnumber_t agno)
+{
+	struct xfs_perag	*pag;
+	int			ref = 0;
+
+	spin_lock(&mp->m_perag_lock);
+	pag = radix_tree_lookup(&mp->m_perag_tree, agno);
+	if (pag) {
+		ASSERT(atomic_read(&pag->pag_ref) >= 0);
+		/* catch leaks in the positive direction during testing */
+		ASSERT(atomic_read(&pag->pag_ref) < 1000);
+		ref = atomic_inc_return(&pag->pag_ref);
+	}
+	spin_unlock(&mp->m_perag_lock);
+	trace_xfs_perag_get(mp, agno, ref, _RET_IP_);
+	return pag;
+}
+
+void
+xfs_perag_put(struct xfs_perag *pag)
+{
+	int	ref;
+
+	ASSERT(atomic_read(&pag->pag_ref) > 0);
+	ref = atomic_dec_return(&pag->pag_ref);
+	trace_xfs_perag_put(pag->pag_mount, pag->pag_agno, ref, _RET_IP_);
+}
+
+/*
  * Free up the resources associated with a mount structure.  Assume that
  * the structure was initially zeroed, so we can tell which fields got
  * initialized.
@@ -209,13 +241,16 @@ STATIC void
 xfs_free_perag(
 	xfs_mount_t	*mp)
 {
-	if (mp->m_perag) {
-		int	agno;
+	xfs_agnumber_t	agno;
+	struct xfs_perag *pag;
 
-		for (agno = 0; agno < mp->m_maxagi; agno++)
-			if (mp->m_perag[agno].pagb_list)
-				kmem_free(mp->m_perag[agno].pagb_list);
-		kmem_free(mp->m_perag);
+	for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) {
+		spin_lock(&mp->m_perag_lock);
+		pag = radix_tree_delete(&mp->m_perag_tree, agno);
+		ASSERT(pag);
+		ASSERT(atomic_read(&pag->pag_ref) == 0);
+		spin_unlock(&mp->m_perag_lock);
+		kmem_free(pag);
 	}
 }
 
@@ -389,22 +424,57 @@ xfs_initialize_perag_icache(
 	}
 }
 
-xfs_agnumber_t
+int
 xfs_initialize_perag(
 	xfs_mount_t	*mp,
-	xfs_agnumber_t	agcount)
+	xfs_agnumber_t	agcount,
+	xfs_agnumber_t	*maxagi)
 {
 	xfs_agnumber_t	index, max_metadata;
+	xfs_agnumber_t	first_initialised = 0;
 	xfs_perag_t	*pag;
 	xfs_agino_t	agino;
 	xfs_ino_t	ino;
 	xfs_sb_t	*sbp = &mp->m_sb;
 	xfs_ino_t	max_inum = XFS_MAXINUMBER_32;
+	int		error = -ENOMEM;
 
 	/* Check to see if the filesystem can overflow 32 bit inodes */
 	agino = XFS_OFFBNO_TO_AGINO(mp, sbp->sb_agblocks - 1, 0);
 	ino = XFS_AGINO_TO_INO(mp, agcount - 1, agino);
 
+	/*
+	 * Walk the current per-ag tree so we don't try to initialise AGs
+	 * that already exist (growfs case). Allocate and insert all the
+	 * AGs we don't find ready for initialisation.
+	 */
+	for (index = 0; index < agcount; index++) {
+		pag = xfs_perag_get(mp, index);
+		if (pag) {
+			xfs_perag_put(pag);
+			continue;
+		}
+		if (!first_initialised)
+			first_initialised = index;
+		pag = kmem_zalloc(sizeof(*pag), KM_MAYFAIL);
+		if (!pag)
+			goto out_unwind;
+		if (radix_tree_preload(GFP_NOFS))
+			goto out_unwind;
+		spin_lock(&mp->m_perag_lock);
+		if (radix_tree_insert(&mp->m_perag_tree, index, pag)) {
+			BUG();
+			spin_unlock(&mp->m_perag_lock);
+			radix_tree_preload_end();
+			error = -EEXIST;
+			goto out_unwind;
+		}
+		pag->pag_agno = index;
+		pag->pag_mount = mp;
+		spin_unlock(&mp->m_perag_lock);
+		radix_tree_preload_end();
+	}
+
 	/* Clear the mount flag if no inode can overflow 32 bits
 	 * on this filesystem, or if specifically requested..
 	 */
@@ -438,21 +508,33 @@ xfs_initialize_perag(
 			}
 
 			/* This ag is preferred for inodes */
-			pag = &mp->m_perag[index];
+			pag = xfs_perag_get(mp, index);
 			pag->pagi_inodeok = 1;
 			if (index < max_metadata)
 				pag->pagf_metadata = 1;
 			xfs_initialize_perag_icache(pag);
+			xfs_perag_put(pag);
 		}
 	} else {
 		/* Setup default behavior for smaller filesystems */
 		for (index = 0; index < agcount; index++) {
-			pag = &mp->m_perag[index];
+			pag = xfs_perag_get(mp, index);
 			pag->pagi_inodeok = 1;
 			xfs_initialize_perag_icache(pag);
+			xfs_perag_put(pag);
 		}
 	}
-	return index;
+	if (maxagi)
+		*maxagi = index;
+	return 0;
+
+out_unwind:
+	kmem_free(pag);
+	for (; index > first_initialised; index--) {
+		pag = radix_tree_delete(&mp->m_perag_tree, index);
+		kmem_free(pag);
+	}
+	return error;
 }
 
 void
@@ -583,7 +665,7 @@ xfs_readsb(xfs_mount_t *mp, int flags)
 	 * access to the superblock.
 	 */
 	sector_size = xfs_getsize_buftarg(mp->m_ddev_targp);
-	extra_flags = XFS_BUF_LOCK | XFS_BUF_MANAGE | XFS_BUF_MAPPED;
+	extra_flags = XBF_LOCK | XBF_FS_MANAGED | XBF_MAPPED;
 
 	bp = xfs_buf_read(mp->m_ddev_targp, XFS_SB_DADDR, BTOBB(sector_size),
 			  extra_flags);
@@ -731,12 +813,13 @@ xfs_initialize_perag_data(xfs_mount_t *mp, xfs_agnumber_t agcount)
 		error = xfs_ialloc_pagi_init(mp, NULL, index);
 		if (error)
 			return error;
-		pag = &mp->m_perag[index];
+		pag = xfs_perag_get(mp, index);
 		ifree += pag->pagi_freecount;
 		ialloc += pag->pagi_count;
 		bfree += pag->pagf_freeblks;
 		bfreelst += pag->pagf_flcount;
 		btree += pag->pagf_btreeblks;
+		xfs_perag_put(pag);
 	}
 	/*
 	 * Overwrite incore superblock counters with just-read data
@@ -1008,6 +1091,24 @@ xfs_mount_reset_sbqflags(
 	return xfs_trans_commit(tp, 0);
 }
 
+__uint64_t
+xfs_default_resblks(xfs_mount_t *mp)
+{
+	__uint64_t resblks;
+
+	/*
+	 * We default to 5% or 8192 fsbs of space reserved, whichever is
+	 * smaller.  This is intended to cover concurrent allocation
+	 * transactions when we initially hit enospc. These each require a 4
+	 * block reservation. Hence by default we cover roughly 2000 concurrent
+	 * allocation reservations.
+	 */
+	resblks = mp->m_sb.sb_dblocks;
+	do_div(resblks, 20);
+	resblks = min_t(__uint64_t, resblks, 8192);
+	return resblks;
+}
+
 /*
  * This function does the following on an initial mount of a file system:
  *	- reads the superblock from disk and init the mount struct
@@ -1152,13 +1253,13 @@ xfs_mountfs(
 	/*
 	 * Allocate and initialize the per-ag data.
 	 */
-	init_rwsem(&mp->m_peraglock);
-	mp->m_perag = kmem_zalloc(sbp->sb_agcount * sizeof(xfs_perag_t),
-				  KM_MAYFAIL);
-	if (!mp->m_perag)
+	spin_lock_init(&mp->m_perag_lock);
+	INIT_RADIX_TREE(&mp->m_perag_tree, GFP_NOFS);
+	error = xfs_initialize_perag(mp, sbp->sb_agcount, &mp->m_maxagi);
+	if (error) {
+		cmn_err(CE_WARN, "XFS: Failed per-ag init: %d", error);
 		goto out_remove_uuid;
-
-	mp->m_maxagi = xfs_initialize_perag(mp, sbp->sb_agcount);
+	}
 
 	if (!sbp->sb_logblocks) {
 		cmn_err(CE_WARN, "XFS: no log defined");
@@ -1319,17 +1420,16 @@ xfs_mountfs(
 	 * attr, unwritten extent conversion at ENOSPC, etc. Data allocations
 	 * are not allowed to use this reserved space.
 	 *
-	 * We default to 5% or 1024 fsbs of space reserved, whichever is smaller.
 	 * This may drive us straight to ENOSPC on mount, but that implies
 	 * we were already there on the last unmount. Warn if this occurs.
 	 */
-	resblks = mp->m_sb.sb_dblocks;
-	do_div(resblks, 20);
-	resblks = min_t(__uint64_t, resblks, 1024);
-	error = xfs_reserve_blocks(mp, &resblks, NULL);
-	if (error)
-		cmn_err(CE_WARN, "XFS: Unable to allocate reserve blocks. "
-				"Continuing without a reserve pool.");
+	if (!(mp->m_flags & XFS_MOUNT_RDONLY)) {
+		resblks = xfs_default_resblks(mp);
+		error = xfs_reserve_blocks(mp, &resblks, NULL);
+		if (error)
+			cmn_err(CE_WARN, "XFS: Unable to allocate reserve "
+				"blocks. Continuing without a reserve pool.");
+	}
 
 	return 0;
 
@@ -1372,8 +1472,19 @@ xfs_unmountfs(
 	 * push out the iclog we will never get that unlocked. hence we
 	 * need to force the log first.
 	 */
-	xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC);
-	xfs_reclaim_inodes(mp, XFS_IFLUSH_ASYNC);
+	xfs_log_force(mp, XFS_LOG_SYNC);
+
+	/*
+	 * Do a delwri reclaim pass first so that as many dirty inodes are
+	 * queued up for IO as possible. Then flush the buffers before making
+	 * a synchronous path to catch all the remaining inodes are reclaimed.
+	 * This makes the reclaim process as quick as possible by avoiding
+	 * synchronous writeout and blocking on inodes already in the delwri
+	 * state as much as possible.
+	 */
+	xfs_reclaim_inodes(mp, 0);
+	XFS_bflush(mp->m_ddev_targp);
+	xfs_reclaim_inodes(mp, SYNC_WAIT);
 
 	xfs_qm_unmount(mp);
 
@@ -1382,7 +1493,7 @@ xfs_unmountfs(
 	 * that nothing is pinned.  This is important because bflush()
 	 * will skip pinned buffers.
 	 */
-	xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC);
+	xfs_log_force(mp, XFS_LOG_SYNC);
 
 	xfs_binval(mp->m_ddev_targp);
 	if (mp->m_rtdev_targp) {
@@ -1548,15 +1659,14 @@ xfs_mod_sb(xfs_trans_t *tp, __int64_t fields)
 	xfs_sb_to_disk(XFS_BUF_TO_SBP(bp), &mp->m_sb, fields);
 
 	/* find modified range */
+	f = (xfs_sb_field_t)xfs_highbit64((__uint64_t)fields);
+	ASSERT((1LL << f) & XFS_SB_MOD_BITS);
+	last = xfs_sb_info[f + 1].offset - 1;
 
 	f = (xfs_sb_field_t)xfs_lowbit64((__uint64_t)fields);
 	ASSERT((1LL << f) & XFS_SB_MOD_BITS);
 	first = xfs_sb_info[f].offset;
 
-	f = (xfs_sb_field_t)xfs_highbit64((__uint64_t)fields);
-	ASSERT((1LL << f) & XFS_SB_MOD_BITS);
-	last = xfs_sb_info[f + 1].offset - 1;
-
 	xfs_trans_log_buf(tp, bp, first, last);
 }
 
@@ -1620,26 +1730,30 @@ xfs_mod_incore_sb_unlocked(
 				lcounter += rem;
 			}
 		} else {				/* Taking blocks away */
-
 			lcounter += delta;
+			if (lcounter >= 0) {
+				mp->m_sb.sb_fdblocks = lcounter +
+							XFS_ALLOC_SET_ASIDE(mp);
+				return 0;
+			}
 
-		/*
-		 * If were out of blocks, use any available reserved blocks if
-		 * were allowed to.
-		 */
+			/*
+			 * We are out of blocks, use any available reserved
+			 * blocks if were allowed to.
+			 */
+			if (!rsvd)
+				return XFS_ERROR(ENOSPC);
 
-			if (lcounter < 0) {
-				if (rsvd) {
-					lcounter = (long long)mp->m_resblks_avail + delta;
-					if (lcounter < 0) {
-						return XFS_ERROR(ENOSPC);
-					}
-					mp->m_resblks_avail = lcounter;
-					return 0;
-				} else {	/* not reserved */
-					return XFS_ERROR(ENOSPC);
-				}
+			lcounter = (long long)mp->m_resblks_avail + delta;
+			if (lcounter >= 0) {
+				mp->m_resblks_avail = lcounter;
+				return 0;
 			}
+			printk_once(KERN_WARNING
+				"Filesystem \"%s\": reserve blocks depleted! "
+				"Consider increasing reserve pool size.",
+				mp->m_fsname);
+			return XFS_ERROR(ENOSPC);
 		}
 
 		mp->m_sb.sb_fdblocks = lcounter + XFS_ALLOC_SET_ASIDE(mp);
@@ -1887,7 +2001,7 @@ xfs_getsb(
 
 	ASSERT(mp->m_sb_bp != NULL);
 	bp = mp->m_sb_bp;
-	if (flags & XFS_BUF_TRYLOCK) {
+	if (flags & XBF_TRYLOCK) {
 		if (!XFS_BUF_CPSEMA(bp)) {
 			return NULL;
 		}
@@ -1947,6 +2061,26 @@ xfs_mount_log_sb(
 	return error;
 }
 
+/*
+ * If the underlying (data/log/rt) device is readonly, there are some
+ * operations that cannot proceed.
+ */
+int
+xfs_dev_is_read_only(
+	struct xfs_mount	*mp,
+	char			*message)
+{
+	if (xfs_readonly_buftarg(mp->m_ddev_targp) ||
+	    xfs_readonly_buftarg(mp->m_logdev_targp) ||
+	    (mp->m_rtdev_targp && xfs_readonly_buftarg(mp->m_rtdev_targp))) {
+		cmn_err(CE_NOTE,
+			"XFS: %s required on read-only device.", message);
+		cmn_err(CE_NOTE,
+			"XFS: write access unavailable, cannot proceed.");
+		return EROFS;
+	}
+	return 0;
+}
 
 #ifdef HAVE_PERCPU_SB
 /*
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 1df7e4502967..4fa0bc7b983e 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -78,7 +78,8 @@ typedef int	(*xfs_send_destroy_t)(struct xfs_inode *, dm_right_t);
 typedef int	(*xfs_send_namesp_t)(dm_eventtype_t, struct xfs_mount *,
 			struct xfs_inode *, dm_right_t,
 			struct xfs_inode *, dm_right_t,
-			const char *, const char *, mode_t, int, int);
+			const unsigned char *, const unsigned char *,
+			mode_t, int, int);
 typedef int	(*xfs_send_mount_t)(struct xfs_mount *, dm_right_t,
 			char *, char *);
 typedef void	(*xfs_send_unmount_t)(struct xfs_mount *, struct xfs_inode *,
@@ -207,8 +208,8 @@ typedef struct xfs_mount {
 	uint			m_ag_maxlevels;	/* XFS_AG_MAXLEVELS */
 	uint			m_bm_maxlevels[2]; /* XFS_BM_MAXLEVELS */
 	uint			m_in_maxlevels;	/* max inobt btree levels. */
-	struct xfs_perag	*m_perag;	/* per-ag accounting info */
-	struct rw_semaphore	m_peraglock;	/* lock for m_perag (pointer) */
+	struct radix_tree_root	m_perag_tree;	/* per-ag accounting info */
+	spinlock_t		m_perag_lock;	/* lock for m_perag_tree */
 	struct mutex		m_growlock;	/* growfs mutex */
 	int			m_fixedfsid[2];	/* unchanged for life of FS */
 	uint			m_dmevmask;	/* DMI events for this FS */
@@ -224,6 +225,7 @@ typedef struct xfs_mount {
 	__uint64_t		m_maxioffset;	/* maximum inode offset */
 	__uint64_t		m_resblks;	/* total reserved blocks */
 	__uint64_t		m_resblks_avail;/* available reserved blocks */
+	__uint64_t		m_resblks_save;	/* reserved blks @ remount,ro */
 	int			m_dalign;	/* stripe unit */
 	int			m_swidth;	/* stripe width */
 	int			m_sinoalign;	/* stripe unit inode alignment */
@@ -243,7 +245,7 @@ typedef struct xfs_mount {
 	struct xfs_qmops	*m_qm_ops;	/* vector of XQM ops */
 	atomic_t		m_active_trans;	/* number trans frozen */
 #ifdef HAVE_PERCPU_SB
-	xfs_icsb_cnts_t		*m_sb_cnts;	/* per-cpu superblock counters */
+	xfs_icsb_cnts_t __percpu *m_sb_cnts;	/* per-cpu superblock counters */
 	unsigned long		m_icsb_counters; /* disabled per-cpu counters */
 	struct notifier_block	m_icsb_notifier; /* hotplug cpu notifier */
 	struct mutex		m_icsb_mutex;	/* balancer sync lock */
@@ -384,19 +386,10 @@ xfs_daddr_to_agbno(struct xfs_mount *mp, xfs_daddr_t d)
 }
 
 /*
- * perag get/put wrappers for eventual ref counting
+ * perag get/put wrappers for ref counting
  */
-static inline xfs_perag_t *
-xfs_get_perag(struct xfs_mount *mp, xfs_ino_t ino)
-{
-	return &mp->m_perag[XFS_INO_TO_AGNO(mp, ino)];
-}
-
-static inline void
-xfs_put_perag(struct xfs_mount *mp, xfs_perag_t *pag)
-{
-	/* nothing to see here, move along */
-}
+struct xfs_perag *xfs_perag_get(struct xfs_mount *mp, xfs_agnumber_t agno);
+void	xfs_perag_put(struct xfs_perag *pag);
 
 /*
  * Per-cpu superblock locking functions
@@ -428,6 +421,7 @@ typedef struct xfs_mod_sb {
 } xfs_mod_sb_t;
 
 extern int	xfs_log_sbcount(xfs_mount_t *, uint);
+extern __uint64_t xfs_default_resblks(xfs_mount_t *mp);
 extern int	xfs_mountfs(xfs_mount_t *mp);
 
 extern void	xfs_unmountfs(xfs_mount_t *);
@@ -442,6 +436,8 @@ extern void	xfs_freesb(xfs_mount_t *);
 extern int	xfs_fs_writable(xfs_mount_t *);
 extern int	xfs_sb_validate_fsb_count(struct xfs_sb *, __uint64_t);
 
+extern int	xfs_dev_is_read_only(struct xfs_mount *, char *);
+
 extern int	xfs_dmops_get(struct xfs_mount *);
 extern void	xfs_dmops_put(struct xfs_mount *);
 
@@ -450,7 +446,8 @@ extern struct xfs_dmops xfs_dmcore_xfs;
 #endif	/* __KERNEL__ */
 
 extern void	xfs_mod_sb(struct xfs_trans *, __int64_t);
-extern xfs_agnumber_t	xfs_initialize_perag(struct xfs_mount *, xfs_agnumber_t);
+extern int	xfs_initialize_perag(struct xfs_mount *, xfs_agnumber_t,
+					xfs_agnumber_t *);
 extern void	xfs_sb_from_disk(struct xfs_sb *, struct xfs_dsb *);
 extern void	xfs_sb_to_disk(struct xfs_dsb *, struct xfs_sb *, __int64_t);
 
diff --git a/fs/xfs/xfs_mru_cache.c b/fs/xfs/xfs_mru_cache.c
index 4b0613d99faa..45ce15dc5b2b 100644
--- a/fs/xfs/xfs_mru_cache.c
+++ b/fs/xfs/xfs_mru_cache.c
@@ -398,7 +398,7 @@ exit:
  * guaranteed that all the free functions for all the elements have finished
  * executing and the reaper is not running.
  */
-void
+static void
 xfs_mru_cache_flush(
 	xfs_mru_cache_t		*mru)
 {
diff --git a/fs/xfs/xfs_mru_cache.h b/fs/xfs/xfs_mru_cache.h
index 5d439f34b0c9..36dd3ec8b4eb 100644
--- a/fs/xfs/xfs_mru_cache.h
+++ b/fs/xfs/xfs_mru_cache.h
@@ -42,7 +42,6 @@ void xfs_mru_cache_uninit(void);
 int xfs_mru_cache_create(struct xfs_mru_cache **mrup, unsigned int lifetime_ms,
 			     unsigned int grp_count,
 			     xfs_mru_cache_free_func_t free_func);
-void xfs_mru_cache_flush(xfs_mru_cache_t *mru);
 void xfs_mru_cache_destroy(struct xfs_mru_cache *mru);
 int xfs_mru_cache_insert(struct xfs_mru_cache *mru, unsigned long key,
 				void *value);
diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h
index 91bfd60f4c74..fdcab3f81dde 100644
--- a/fs/xfs/xfs_quota.h
+++ b/fs/xfs/xfs_quota.h
@@ -223,16 +223,9 @@ typedef struct xfs_qoff_logformat {
 #define XFS_QMOPT_RES_INOS	0x0800000
 
 /*
- * flags for dqflush and dqflush_all.
- */
-#define XFS_QMOPT_SYNC		0x1000000
-#define XFS_QMOPT_ASYNC		0x2000000
-#define XFS_QMOPT_DELWRI	0x4000000
-
-/*
  * flags for dqalloc.
  */
-#define XFS_QMOPT_INHERIT	0x8000000
+#define XFS_QMOPT_INHERIT	0x1000000
 
 /*
  * flags to xfs_trans_mod_dquot.
diff --git a/fs/xfs/xfs_rw.c b/fs/xfs/xfs_rw.c
index 5aa07caea5f1..e336742a58a4 100644
--- a/fs/xfs/xfs_rw.c
+++ b/fs/xfs/xfs_rw.c
@@ -47,48 +47,6 @@
 #include "xfs_trace.h"
 
 /*
- * This is a subroutine for xfs_write() and other writers (xfs_ioctl)
- * which clears the setuid and setgid bits when a file is written.
- */
-int
-xfs_write_clear_setuid(
-	xfs_inode_t	*ip)
-{
-	xfs_mount_t	*mp;
-	xfs_trans_t	*tp;
-	int		error;
-
-	mp = ip->i_mount;
-	tp = xfs_trans_alloc(mp, XFS_TRANS_WRITEID);
-	if ((error = xfs_trans_reserve(tp, 0,
-				      XFS_WRITEID_LOG_RES(mp),
-				      0, 0, 0))) {
-		xfs_trans_cancel(tp, 0);
-		return error;
-	}
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
-	xfs_trans_ihold(tp, ip);
-	ip->i_d.di_mode &= ~S_ISUID;
-
-	/*
-	 * Note that we don't have to worry about mandatory
-	 * file locking being disabled here because we only
-	 * clear the S_ISGID bit if the Group execute bit is
-	 * on, but if it was on then mandatory locking wouldn't
-	 * have been enabled.
-	 */
-	if (ip->i_d.di_mode & S_IXGRP) {
-		ip->i_d.di_mode &= ~S_ISGID;
-	}
-	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-	xfs_trans_set_sync(tp);
-	error = xfs_trans_commit(tp, 0);
-	xfs_iunlock(ip, XFS_ILOCK_EXCL);
-	return 0;
-}
-
-/*
  * Force a shutdown of the filesystem instantly while keeping
  * the filesystem consistent. We don't do an unmount here; just shutdown
  * the shop, make sure that absolutely nothing persistent happens to
@@ -153,88 +111,6 @@ xfs_do_force_shutdown(
 	}
 }
 
-
-/*
- * Called when we want to stop a buffer from getting written or read.
- * We attach the EIO error, muck with its flags, and call biodone
- * so that the proper iodone callbacks get called.
- */
-int
-xfs_bioerror(
-	xfs_buf_t *bp)
-{
-
-#ifdef XFSERRORDEBUG
-	ASSERT(XFS_BUF_ISREAD(bp) || bp->b_iodone);
-#endif
-
-	/*
-	 * No need to wait until the buffer is unpinned.
-	 * We aren't flushing it.
-	 */
-	XFS_BUF_ERROR(bp, EIO);
-	/*
-	 * We're calling biodone, so delete B_DONE flag. Either way
-	 * we have to call the iodone callback, and calling biodone
-	 * probably is the best way since it takes care of
-	 * GRIO as well.
-	 */
-	XFS_BUF_UNREAD(bp);
-	XFS_BUF_UNDELAYWRITE(bp);
-	XFS_BUF_UNDONE(bp);
-	XFS_BUF_STALE(bp);
-
-	XFS_BUF_CLR_BDSTRAT_FUNC(bp);
-	xfs_biodone(bp);
-
-	return (EIO);
-}
-
-/*
- * Same as xfs_bioerror, except that we are releasing the buffer
- * here ourselves, and avoiding the biodone call.
- * This is meant for userdata errors; metadata bufs come with
- * iodone functions attached, so that we can track down errors.
- */
-int
-xfs_bioerror_relse(
-	xfs_buf_t *bp)
-{
-	int64_t fl;
-
-	ASSERT(XFS_BUF_IODONE_FUNC(bp) != xfs_buf_iodone_callbacks);
-	ASSERT(XFS_BUF_IODONE_FUNC(bp) != xlog_iodone);
-
-	fl = XFS_BUF_BFLAGS(bp);
-	/*
-	 * No need to wait until the buffer is unpinned.
-	 * We aren't flushing it.
-	 *
-	 * chunkhold expects B_DONE to be set, whether
-	 * we actually finish the I/O or not. We don't want to
-	 * change that interface.
-	 */
-	XFS_BUF_UNREAD(bp);
-	XFS_BUF_UNDELAYWRITE(bp);
-	XFS_BUF_DONE(bp);
-	XFS_BUF_STALE(bp);
-	XFS_BUF_CLR_IODONE_FUNC(bp);
-	XFS_BUF_CLR_BDSTRAT_FUNC(bp);
-	if (!(fl & XFS_B_ASYNC)) {
-		/*
-		 * Mark b_error and B_ERROR _both_.
-		 * Lot's of chunkcache code assumes that.
-		 * There's no reason to mark error for
-		 * ASYNC buffers.
-		 */
-		XFS_BUF_ERROR(bp, EIO);
-		XFS_BUF_FINISH_IOWAIT(bp);
-	} else {
-		xfs_buf_relse(bp);
-	}
-	return (EIO);
-}
-
 /*
  * Prints out an ALERT message about I/O error.
  */
@@ -306,37 +182,6 @@ xfs_read_buf(
 }
 
 /*
- * Wrapper around bwrite() so that we can trap
- * write errors, and act accordingly.
- */
-int
-xfs_bwrite(
-	struct xfs_mount *mp,
-	struct xfs_buf	 *bp)
-{
-	int	error;
-
-	/*
-	 * XXXsup how does this work for quotas.
-	 */
-	XFS_BUF_SET_BDSTRAT_FUNC(bp, xfs_bdstrat_cb);
-	bp->b_mount = mp;
-	XFS_BUF_WRITE(bp);
-
-	if ((error = XFS_bwrite(bp))) {
-		ASSERT(mp);
-		/*
-		 * Cannot put a buftrace here since if the buffer is not
-		 * B_HOLD then we will brelse() the buffer before returning
-		 * from bwrite and we could be tracing a buffer that has
-		 * been reused.
-		 */
-		xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR);
-	}
-	return (error);
-}
-
-/*
  * helper function to extract extent size hint from inode
  */
 xfs_extlen_t
diff --git a/fs/xfs/xfs_rw.h b/fs/xfs/xfs_rw.h
index 571f2174435c..11c41ec6ed75 100644
--- a/fs/xfs/xfs_rw.h
+++ b/fs/xfs/xfs_rw.h
@@ -39,10 +39,6 @@ xfs_fsb_to_db(struct xfs_inode *ip, xfs_fsblock_t fsb)
 /*
  * Prototypes for functions in xfs_rw.c.
  */
-extern int xfs_write_clear_setuid(struct xfs_inode *ip);
-extern int xfs_bwrite(struct xfs_mount *mp, struct xfs_buf *bp);
-extern int xfs_bioerror(struct xfs_buf *bp);
-extern int xfs_bioerror_relse(struct xfs_buf *bp);
 extern int xfs_read_buf(struct xfs_mount *mp, xfs_buftarg_t *btp,
 			xfs_daddr_t blkno, int len, uint flags,
 			struct xfs_buf **bpp);
diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index 237badcbac3b..f73e358bae8d 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -796,7 +796,7 @@ _xfs_trans_commit(
 	int			sync;
 #define	XFS_TRANS_LOGVEC_COUNT	16
 	xfs_log_iovec_t		log_vector_fast[XFS_TRANS_LOGVEC_COUNT];
-	void			*commit_iclog;
+	struct xlog_in_core	*commit_iclog;
 	int			shutdown;
 
 	commit_lsn = -1;
@@ -981,9 +981,8 @@ shut_us_down:
 	 */
 	if (sync) {
 		if (!error) {
-			error = _xfs_log_force(mp, commit_lsn,
-				      XFS_LOG_FORCE | XFS_LOG_SYNC,
-				      log_flushed);
+			error = _xfs_log_force_lsn(mp, commit_lsn,
+				      XFS_LOG_SYNC, log_flushed);
 		}
 		XFS_STATS_INC(xs_trans_sync);
 	} else {
@@ -1121,7 +1120,7 @@ xfs_trans_fill_vecs(
 	tp->t_header.th_num_items = nitems;
 	log_vector->i_addr = (xfs_caddr_t)&tp->t_header;
 	log_vector->i_len = sizeof(xfs_trans_header_t);
-	XLOG_VEC_SET_TYPE(log_vector, XLOG_REG_TYPE_TRANSHDR);
+	log_vector->i_type = XLOG_REG_TYPE_TRANSHDR;
 }
 
 
diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index ca64f33c63a3..79c8bab9dfff 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -861,8 +861,7 @@ typedef struct xfs_item_ops {
 #define	XFS_ITEM_SUCCESS	0
 #define	XFS_ITEM_PINNED		1
 #define	XFS_ITEM_LOCKED		2
-#define	XFS_ITEM_FLUSHING	3
-#define XFS_ITEM_PUSHBUF	4
+#define XFS_ITEM_PUSHBUF	3
 
 /*
  * This structure is used to maintain a list of block ranges that have been
@@ -911,7 +910,7 @@ typedef struct xfs_trans {
 	unsigned int		t_blk_res_used;	/* # of resvd blocks used */
 	unsigned int		t_rtx_res;	/* # of rt extents resvd */
 	unsigned int		t_rtx_res_used;	/* # of resvd rt extents used */
-	xfs_log_ticket_t	t_ticket;	/* log mgr ticket */
+	struct xlog_ticket	*t_ticket;	/* log mgr ticket */
 	xfs_lsn_t		t_lsn;		/* log seq num of start of
 						 * transaction. */
 	xfs_lsn_t		t_commit_lsn;	/* log seq num of end of
diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c
index 2ffc570679be..e799824f7245 100644
--- a/fs/xfs/xfs_trans_ail.c
+++ b/fs/xfs/xfs_trans_ail.c
@@ -237,14 +237,15 @@ out:
 }
 
 /*
- * Function that does the work of pushing on the AIL
+ * xfsaild_push does the work of pushing on the AIL.  Returning a timeout of
+ * zero indicates that the caller should sleep until woken.
  */
 long
 xfsaild_push(
 	struct xfs_ail	*ailp,
 	xfs_lsn_t	*last_lsn)
 {
-	long		tout = 1000; /* milliseconds */
+	long		tout = 0;
 	xfs_lsn_t	last_pushed_lsn = *last_lsn;
 	xfs_lsn_t	target =  ailp->xa_target;
 	xfs_lsn_t	lsn;
@@ -252,6 +253,7 @@ xfsaild_push(
 	int		flush_log, count, stuck;
 	xfs_mount_t	*mp = ailp->xa_mount;
 	struct xfs_ail_cursor	*cur = &ailp->xa_cursors;
+	int		push_xfsbufd = 0;
 
 	spin_lock(&ailp->xa_lock);
 	xfs_trans_ail_cursor_init(ailp, cur);
@@ -262,7 +264,7 @@ xfsaild_push(
 		 */
 		xfs_trans_ail_cursor_done(ailp, cur);
 		spin_unlock(&ailp->xa_lock);
-		last_pushed_lsn = 0;
+		*last_lsn = 0;
 		return tout;
 	}
 
@@ -279,7 +281,6 @@ xfsaild_push(
 	 * prevents use from spinning when we can't do anything or there is
 	 * lots of contention on the AIL lists.
 	 */
-	tout = 10;
 	lsn = lip->li_lsn;
 	flush_log = stuck = count = 0;
 	while ((XFS_LSN_CMP(lip->li_lsn, target) < 0)) {
@@ -308,6 +309,7 @@ xfsaild_push(
 			XFS_STATS_INC(xs_push_ail_pushbuf);
 			IOP_PUSHBUF(lip);
 			last_pushed_lsn = lsn;
+			push_xfsbufd = 1;
 			break;
 
 		case XFS_ITEM_PINNED:
@@ -322,12 +324,6 @@ xfsaild_push(
 			stuck++;
 			break;
 
-		case XFS_ITEM_FLUSHING:
-			XFS_STATS_INC(xs_push_ail_flushing);
-			last_pushed_lsn = lsn;
-			stuck++;
-			break;
-
 		default:
 			ASSERT(0);
 			break;
@@ -371,19 +367,24 @@ xfsaild_push(
 		 * move forward in the AIL.
 		 */
 		XFS_STATS_INC(xs_push_ail_flush);
-		xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
+		xfs_log_force(mp, 0);
+	}
+
+	if (push_xfsbufd) {
+		/* we've got delayed write buffers to flush */
+		wake_up_process(mp->m_ddev_targp->bt_task);
 	}
 
 	if (!count) {
 		/* We're past our target or empty, so idle */
-		tout = 1000;
+		last_pushed_lsn = 0;
 	} else if (XFS_LSN_CMP(lsn, target) >= 0) {
 		/*
 		 * We reached the target so wait a bit longer for I/O to
 		 * complete and remove pushed items from the AIL before we
 		 * start the next scan from the start of the AIL.
 		 */
-		tout += 20;
+		tout = 50;
 		last_pushed_lsn = 0;
 	} else if ((stuck * 100) / count > 90) {
 		/*
@@ -395,11 +396,14 @@ xfsaild_push(
 		 * Backoff a bit more to allow some I/O to complete before
 		 * continuing from where we were.
 		 */
-		tout += 10;
+		tout = 20;
+	} else {
+		/* more to do, but wait a short while before continuing */
+		tout = 10;
 	}
 	*last_lsn = last_pushed_lsn;
 	return tout;
-}	/* xfsaild_push */
+}
 
 
 /*
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index 49130628d5ef..fb586360d1c9 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -46,6 +46,65 @@ STATIC xfs_buf_t *xfs_trans_buf_item_match(xfs_trans_t *, xfs_buftarg_t *,
 STATIC xfs_buf_t *xfs_trans_buf_item_match_all(xfs_trans_t *, xfs_buftarg_t *,
 		xfs_daddr_t, int);
 
+/*
+ * Add the locked buffer to the transaction.
+ *
+ * The buffer must be locked, and it cannot be associated with any
+ * transaction.
+ *
+ * If the buffer does not yet have a buf log item associated with it,
+ * then allocate one for it.  Then add the buf item to the transaction.
+ */
+STATIC void
+_xfs_trans_bjoin(
+	struct xfs_trans	*tp,
+	struct xfs_buf		*bp,
+	int			reset_recur)
+{
+	struct xfs_buf_log_item	*bip;
+
+	ASSERT(XFS_BUF_ISBUSY(bp));
+	ASSERT(XFS_BUF_FSPRIVATE2(bp, void *) == NULL);
+
+	/*
+	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
+	 * it doesn't have one yet, then allocate one and initialize it.
+	 * The checks to see if one is there are in xfs_buf_item_init().
+	 */
+	xfs_buf_item_init(bp, tp->t_mountp);
+	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
+	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
+	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
+	if (reset_recur)
+		bip->bli_recur = 0;
+
+	/*
+	 * Take a reference for this transaction on the buf item.
+	 */
+	atomic_inc(&bip->bli_refcount);
+
+	/*
+	 * Get a log_item_desc to point at the new item.
+	 */
+	(void) xfs_trans_add_item(tp, (xfs_log_item_t *)bip);
+
+	/*
+	 * Initialize b_fsprivate2 so we can find it with incore_match()
+	 * in xfs_trans_get_buf() and friends above.
+	 */
+	XFS_BUF_SET_FSPRIVATE2(bp, tp);
+
+}
+
+void
+xfs_trans_bjoin(
+	struct xfs_trans	*tp,
+	struct xfs_buf		*bp)
+{
+	_xfs_trans_bjoin(tp, bp, 0);
+	trace_xfs_trans_bjoin(bp->b_fspriv);
+}
 
 /*
  * Get and lock the buffer for the caller if it is not already
@@ -75,13 +134,14 @@ xfs_trans_get_buf(xfs_trans_t	*tp,
 	xfs_buf_log_item_t	*bip;
 
 	if (flags == 0)
-		flags = XFS_BUF_LOCK | XFS_BUF_MAPPED;
+		flags = XBF_LOCK | XBF_MAPPED;
 
 	/*
 	 * Default to a normal get_buf() call if the tp is NULL.
 	 */
 	if (tp == NULL)
-		return xfs_buf_get(target_dev, blkno, len, flags | BUF_BUSY);
+		return xfs_buf_get(target_dev, blkno, len,
+				   flags | XBF_DONT_BLOCK);
 
 	/*
 	 * If we find the buffer in the cache with this transaction
@@ -117,54 +177,22 @@ xfs_trans_get_buf(xfs_trans_t	*tp,
 	}
 
 	/*
-	 * We always specify the BUF_BUSY flag within a transaction so
-	 * that get_buf does not try to push out a delayed write buffer
+	 * We always specify the XBF_DONT_BLOCK flag within a transaction
+	 * so that get_buf does not try to push out a delayed write buffer
 	 * which might cause another transaction to take place (if the
 	 * buffer was delayed alloc).  Such recursive transactions can
 	 * easily deadlock with our current transaction as well as cause
 	 * us to run out of stack space.
 	 */
-	bp = xfs_buf_get(target_dev, blkno, len, flags | BUF_BUSY);
+	bp = xfs_buf_get(target_dev, blkno, len, flags | XBF_DONT_BLOCK);
 	if (bp == NULL) {
 		return NULL;
 	}
 
 	ASSERT(!XFS_BUF_GETERROR(bp));
 
-	/*
-	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
-	 * it doesn't have one yet, then allocate one and initialize it.
-	 * The checks to see if one is there are in xfs_buf_item_init().
-	 */
-	xfs_buf_item_init(bp, tp->t_mountp);
-
-	/*
-	 * Set the recursion count for the buffer within this transaction
-	 * to 0.
-	 */
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
-	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
-	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
-	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
-	bip->bli_recur = 0;
-
-	/*
-	 * Take a reference for this transaction on the buf item.
-	 */
-	atomic_inc(&bip->bli_refcount);
-
-	/*
-	 * Get a log_item_desc to point at the new item.
-	 */
-	(void) xfs_trans_add_item(tp, (xfs_log_item_t*)bip);
-
-	/*
-	 * Initialize b_fsprivate2 so we can find it with incore_match()
-	 * above.
-	 */
-	XFS_BUF_SET_FSPRIVATE2(bp, tp);
-
-	trace_xfs_trans_get_buf(bip);
+	_xfs_trans_bjoin(tp, bp, 1);
+	trace_xfs_trans_get_buf(bp->b_fspriv);
 	return (bp);
 }
 
@@ -209,44 +237,11 @@ xfs_trans_getsb(xfs_trans_t	*tp,
 	}
 
 	bp = xfs_getsb(mp, flags);
-	if (bp == NULL) {
+	if (bp == NULL)
 		return NULL;
-	}
-
-	/*
-	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
-	 * it doesn't have one yet, then allocate one and initialize it.
-	 * The checks to see if one is there are in xfs_buf_item_init().
-	 */
-	xfs_buf_item_init(bp, mp);
-
-	/*
-	 * Set the recursion count for the buffer within this transaction
-	 * to 0.
-	 */
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
-	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
-	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
-	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
-	bip->bli_recur = 0;
-
-	/*
-	 * Take a reference for this transaction on the buf item.
-	 */
-	atomic_inc(&bip->bli_refcount);
-
-	/*
-	 * Get a log_item_desc to point at the new item.
-	 */
-	(void) xfs_trans_add_item(tp, (xfs_log_item_t*)bip);
-
-	/*
-	 * Initialize b_fsprivate2 so we can find it with incore_match()
-	 * above.
-	 */
-	XFS_BUF_SET_FSPRIVATE2(bp, tp);
 
-	trace_xfs_trans_getsb(bip);
+	_xfs_trans_bjoin(tp, bp, 1);
+	trace_xfs_trans_getsb(bp->b_fspriv);
 	return (bp);
 }
 
@@ -290,15 +285,15 @@ xfs_trans_read_buf(
 	int			error;
 
 	if (flags == 0)
-		flags = XFS_BUF_LOCK | XFS_BUF_MAPPED;
+		flags = XBF_LOCK | XBF_MAPPED;
 
 	/*
 	 * Default to a normal get_buf() call if the tp is NULL.
 	 */
 	if (tp == NULL) {
-		bp = xfs_buf_read(target, blkno, len, flags | BUF_BUSY);
+		bp = xfs_buf_read(target, blkno, len, flags | XBF_DONT_BLOCK);
 		if (!bp)
-			return (flags & XFS_BUF_TRYLOCK) ?
+			return (flags & XBF_TRYLOCK) ?
 					EAGAIN : XFS_ERROR(ENOMEM);
 
 		if (XFS_BUF_GETERROR(bp) != 0) {
@@ -385,14 +380,14 @@ xfs_trans_read_buf(
 	}
 
 	/*
-	 * We always specify the BUF_BUSY flag within a transaction so
-	 * that get_buf does not try to push out a delayed write buffer
+	 * We always specify the XBF_DONT_BLOCK flag within a transaction
+	 * so that get_buf does not try to push out a delayed write buffer
 	 * which might cause another transaction to take place (if the
 	 * buffer was delayed alloc).  Such recursive transactions can
 	 * easily deadlock with our current transaction as well as cause
 	 * us to run out of stack space.
 	 */
-	bp = xfs_buf_read(target, blkno, len, flags | BUF_BUSY);
+	bp = xfs_buf_read(target, blkno, len, flags | XBF_DONT_BLOCK);
 	if (bp == NULL) {
 		*bpp = NULL;
 		return 0;
@@ -424,40 +419,9 @@ xfs_trans_read_buf(
 	if (XFS_FORCED_SHUTDOWN(mp))
 		goto shutdown_abort;
 
-	/*
-	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
-	 * it doesn't have one yet, then allocate one and initialize it.
-	 * The checks to see if one is there are in xfs_buf_item_init().
-	 */
-	xfs_buf_item_init(bp, tp->t_mountp);
+	_xfs_trans_bjoin(tp, bp, 1);
+	trace_xfs_trans_read_buf(bp->b_fspriv);
 
-	/*
-	 * Set the recursion count for the buffer within this transaction
-	 * to 0.
-	 */
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
-	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
-	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
-	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
-	bip->bli_recur = 0;
-
-	/*
-	 * Take a reference for this transaction on the buf item.
-	 */
-	atomic_inc(&bip->bli_refcount);
-
-	/*
-	 * Get a log_item_desc to point at the new item.
-	 */
-	(void) xfs_trans_add_item(tp, (xfs_log_item_t*)bip);
-
-	/*
-	 * Initialize b_fsprivate2 so we can find it with incore_match()
-	 * above.
-	 */
-	XFS_BUF_SET_FSPRIVATE2(bp, tp);
-
-	trace_xfs_trans_read_buf(bip);
 	*bpp = bp;
 	return 0;
 
@@ -472,8 +436,8 @@ shutdown_abort:
 	if (XFS_BUF_ISSTALE(bp) && XFS_BUF_ISDELAYWRITE(bp))
 		cmn_err(CE_NOTE, "about to pop assert, bp == 0x%p", bp);
 #endif
-	ASSERT((XFS_BUF_BFLAGS(bp) & (XFS_B_STALE|XFS_B_DELWRI)) !=
-						(XFS_B_STALE|XFS_B_DELWRI));
+	ASSERT((XFS_BUF_BFLAGS(bp) & (XBF_STALE|XBF_DELWRI)) !=
+				     (XBF_STALE|XBF_DELWRI));
 
 	trace_xfs_trans_read_buf_shut(bp, _RET_IP_);
 	xfs_buf_relse(bp);
@@ -622,53 +586,6 @@ xfs_trans_brelse(xfs_trans_t	*tp,
 }
 
 /*
- * Add the locked buffer to the transaction.
- * The buffer must be locked, and it cannot be associated with any
- * transaction.
- *
- * If the buffer does not yet have a buf log item associated with it,
- * then allocate one for it.  Then add the buf item to the transaction.
- */
-void
-xfs_trans_bjoin(xfs_trans_t	*tp,
-		xfs_buf_t	*bp)
-{
-	xfs_buf_log_item_t	*bip;
-
-	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, void *) == NULL);
-
-	/*
-	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
-	 * it doesn't have one yet, then allocate one and initialize it.
-	 * The checks to see if one is there are in xfs_buf_item_init().
-	 */
-	xfs_buf_item_init(bp, tp->t_mountp);
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
-	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
-	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
-	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
-
-	/*
-	 * Take a reference for this transaction on the buf item.
-	 */
-	atomic_inc(&bip->bli_refcount);
-
-	/*
-	 * Get a log_item_desc to point at the new item.
-	 */
-	(void) xfs_trans_add_item(tp, (xfs_log_item_t *)bip);
-
-	/*
-	 * Initialize b_fsprivate2 so we can find it with incore_match()
-	 * in xfs_trans_get_buf() and friends above.
-	 */
-	XFS_BUF_SET_FSPRIVATE2(bp, tp);
-
-	trace_xfs_trans_bjoin(bip);
-}
-
-/*
  * Mark the buffer as not needing to be unlocked when the buf item's
  * IOP_UNLOCK() routine is called.  The buffer must already be locked
  * and associated with the given transaction.
diff --git a/fs/xfs/xfs_types.h b/fs/xfs/xfs_types.h
index d725428c9df6..b09904555d07 100644
--- a/fs/xfs/xfs_types.h
+++ b/fs/xfs/xfs_types.h
@@ -151,8 +151,8 @@ typedef enum {
 } xfs_btnum_t;
 
 struct xfs_name {
-	const char	*name;
-	int		len;
+	const unsigned char	*name;
+	int			len;
 };
 
 #endif	/* __XFS_TYPES_H__ */
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index 6f268756bf36..9d376be0ea38 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -256,7 +256,7 @@ xfs_setattr(
 		    iattr->ia_size > ip->i_d.di_size) {
 			code = xfs_flush_pages(ip,
 					ip->i_d.di_size, iattr->ia_size,
-					XFS_B_ASYNC, FI_NONE);
+					XBF_ASYNC, FI_NONE);
 		}
 
 		/* wait for all I/O to complete */
@@ -584,116 +584,6 @@ xfs_readlink(
 }
 
 /*
- * xfs_fsync
- *
- * This is called to sync the inode and its data out to disk.  We need to hold
- * the I/O lock while flushing the data, and the inode lock while flushing the
- * inode.  The inode lock CANNOT be held while flushing the data, so acquire
- * after we're done with that.
- */
-int
-xfs_fsync(
-	xfs_inode_t	*ip)
-{
-	xfs_trans_t	*tp;
-	int		error = 0;
-	int		log_flushed = 0, changed = 1;
-
-	xfs_itrace_entry(ip);
-
-	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
-		return XFS_ERROR(EIO);
-
-	/*
-	 * We always need to make sure that the required inode state is safe on
-	 * disk.  The inode might be clean but we still might need to force the
-	 * log because of committed transactions that haven't hit the disk yet.
-	 * Likewise, there could be unflushed non-transactional changes to the
-	 * inode core that have to go to disk and this requires us to issue
-	 * a synchronous transaction to capture these changes correctly.
-	 *
-	 * This code relies on the assumption that if the update_* fields
-	 * of the inode are clear and the inode is unpinned then it is clean
-	 * and no action is required.
-	 */
-	xfs_ilock(ip, XFS_ILOCK_SHARED);
-
-	if (!ip->i_update_core) {
-		/*
-		 * Timestamps/size haven't changed since last inode flush or
-		 * inode transaction commit.  That means either nothing got
-		 * written or a transaction committed which caught the updates.
-		 * If the latter happened and the transaction hasn't hit the
-		 * disk yet, the inode will be still be pinned.  If it is,
-		 * force the log.
-		 */
-
-		xfs_iunlock(ip, XFS_ILOCK_SHARED);
-
-		if (xfs_ipincount(ip)) {
-			error = _xfs_log_force(ip->i_mount, (xfs_lsn_t)0,
-				      XFS_LOG_FORCE | XFS_LOG_SYNC,
-				      &log_flushed);
-		} else {
-			/*
-			 * If the inode is not pinned and nothing has changed
-			 * we don't need to flush the cache.
-			 */
-			changed = 0;
-		}
-	} else	{
-		/*
-		 * Kick off a transaction to log the inode core to get the
-		 * updates.  The sync transaction will also force the log.
-		 */
-		xfs_iunlock(ip, XFS_ILOCK_SHARED);
-		tp = xfs_trans_alloc(ip->i_mount, XFS_TRANS_FSYNC_TS);
-		error = xfs_trans_reserve(tp, 0,
-				XFS_FSYNC_TS_LOG_RES(ip->i_mount), 0, 0, 0);
-		if (error) {
-			xfs_trans_cancel(tp, 0);
-			return error;
-		}
-		xfs_ilock(ip, XFS_ILOCK_EXCL);
-
-		/*
-		 * Note - it's possible that we might have pushed ourselves out
-		 * of the way during trans_reserve which would flush the inode.
-		 * But there's no guarantee that the inode buffer has actually
-		 * gone out yet (it's delwri).	Plus the buffer could be pinned
-		 * anyway if it's part of an inode in another recent
-		 * transaction.	 So we play it safe and fire off the
-		 * transaction anyway.
-		 */
-		xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
-		xfs_trans_ihold(tp, ip);
-		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-		xfs_trans_set_sync(tp);
-		error = _xfs_trans_commit(tp, 0, &log_flushed);
-
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-	}
-
-	if ((ip->i_mount->m_flags & XFS_MOUNT_BARRIER) && changed) {
-		/*
-		 * If the log write didn't issue an ordered tag we need
-		 * to flush the disk cache for the data device now.
-		 */
-		if (!log_flushed)
-			xfs_blkdev_issue_flush(ip->i_mount->m_ddev_targp);
-
-		/*
-		 * If this inode is on the RT dev we need to flush that
-		 * cache as well.
-		 */
-		if (XFS_IS_REALTIME_INODE(ip))
-			xfs_blkdev_issue_flush(ip->i_mount->m_rtdev_targp);
-	}
-
-	return error;
-}
-
-/*
  * Flags for xfs_free_eofblocks
  */
 #define XFS_FREE_EOF_TRYLOCK	(1<<0)
@@ -1096,7 +986,7 @@ xfs_release(
 		 */
 		truncated = xfs_iflags_test_and_clear(ip, XFS_ITRUNCATED);
 		if (truncated && VN_DIRTY(VFS_I(ip)) && ip->i_delayed_blks > 0)
-			xfs_flush_pages(ip, 0, -1, XFS_B_ASYNC, FI_NONE);
+			xfs_flush_pages(ip, 0, -1, XBF_ASYNC, FI_NONE);
 	}
 
 	if (ip->i_d.di_nlink != 0) {
@@ -2199,7 +2089,8 @@ xfs_symlink(
 	if (DM_EVENT_ENABLED(dp, DM_EVENT_SYMLINK)) {
 		error = XFS_SEND_NAMESP(mp, DM_EVENT_SYMLINK, dp,
 					DM_RIGHT_NULL, NULL, DM_RIGHT_NULL,
-					link_name->name, target_path, 0, 0, 0);
+					link_name->name,
+					(unsigned char *)target_path, 0, 0, 0);
 		if (error)
 			return error;
 	}
@@ -2395,7 +2286,8 @@ std_return:
 					dp, DM_RIGHT_NULL,
 					error ? NULL : ip,
 					DM_RIGHT_NULL, link_name->name,
-					target_path, 0, error, 0);
+					(unsigned char *)target_path,
+					0, error, 0);
 	}
 
 	if (!error)
diff --git a/fs/xfs/xfs_vnodeops.h b/fs/xfs/xfs_vnodeops.h
index 167a467403a5..d8dfa8d0dadd 100644
--- a/fs/xfs/xfs_vnodeops.h
+++ b/fs/xfs/xfs_vnodeops.h
@@ -21,7 +21,6 @@ int xfs_setattr(struct xfs_inode *ip, struct iattr *vap, int flags);
 #define XFS_ATTR_NOACL		0x08	/* Don't call xfs_acl_chmod */
 
 int xfs_readlink(struct xfs_inode *ip, char *link);
-int xfs_fsync(struct xfs_inode *ip);
 int xfs_release(struct xfs_inode *ip);
 int xfs_inactive(struct xfs_inode *ip);
 int xfs_lookup(struct xfs_inode *dp, struct xfs_name *name,
@@ -43,25 +42,13 @@ int xfs_change_file_space(struct xfs_inode *ip, int cmd,
 int xfs_rename(struct xfs_inode *src_dp, struct xfs_name *src_name,
 		struct xfs_inode *src_ip, struct xfs_inode *target_dp,
 		struct xfs_name *target_name, struct xfs_inode *target_ip);
-int xfs_attr_get(struct xfs_inode *ip, const char *name, char *value,
-		int *valuelenp, int flags);
-int xfs_attr_set(struct xfs_inode *dp, const char *name, char *value,
-		int valuelen, int flags);
-int xfs_attr_remove(struct xfs_inode *dp, const char *name, int flags);
+int xfs_attr_get(struct xfs_inode *ip, const unsigned char *name,
+		unsigned char *value, int *valuelenp, int flags);
+int xfs_attr_set(struct xfs_inode *dp, const unsigned char *name,
+		unsigned char *value, int valuelen, int flags);
+int xfs_attr_remove(struct xfs_inode *dp, const unsigned char *name, int flags);
 int xfs_attr_list(struct xfs_inode *dp, char *buffer, int bufsize,
 		int flags, struct attrlist_cursor_kern *cursor);
-ssize_t xfs_read(struct xfs_inode *ip, struct kiocb *iocb,
-		const struct iovec *iovp, unsigned int segs,
-		loff_t *offset, int ioflags);
-ssize_t xfs_splice_read(struct xfs_inode *ip, struct file *infilp,
-		loff_t *ppos, struct pipe_inode_info *pipe, size_t count,
-		int flags, int ioflags);
-ssize_t xfs_splice_write(struct xfs_inode *ip,
-		struct pipe_inode_info *pipe, struct file *outfilp,
-		loff_t *ppos, size_t count, int flags, int ioflags);
-ssize_t xfs_write(struct xfs_inode *xip, struct kiocb *iocb,
-		const struct iovec *iovp, unsigned int nsegs,
-		loff_t *offset, int ioflags);
 int xfs_bmap(struct xfs_inode *ip, xfs_off_t offset, ssize_t count,
 		int flags, struct xfs_iomap *iomapp, int *niomaps);
 void xfs_tosspages(struct xfs_inode *inode, xfs_off_t first,
@@ -72,4 +59,6 @@ int xfs_flush_pages(struct xfs_inode *ip, xfs_off_t first,
 		xfs_off_t last, uint64_t flags, int fiopt);
 int xfs_wait_on_pages(struct xfs_inode *ip, xfs_off_t first, xfs_off_t last);
 
+int xfs_zero_eof(struct xfs_inode *, xfs_off_t, xfs_fsize_t);
+
 #endif /* _XFS_VNODEOPS_H */