Merge branch 'mw-3.1-jul25' of git://oss.oracle.com/git/smushran/linux-2.6 into ocfs2-fixes

389 files changed, 9983 insertions, 10005 deletions

diff --git a/fs/9p/acl.c b/fs/9p/acl.c
index 535ab6eccb1..e98f56d3787 100644
--- a/fs/9p/acl.c
+++ b/fs/9p/acl.c
@@ -96,12 +96,12 @@ static struct posix_acl *v9fs_get_cached_acl(struct inode *inode, int type)
 	return acl;
 }
 
-int v9fs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int v9fs_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	struct v9fs_session_info *v9ses;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	v9ses = v9fs_inode2v9ses(inode);
diff --git a/fs/9p/acl.h b/fs/9p/acl.h
index 7ef3ac9f6d9..59e18c2e8c7 100644
--- a/fs/9p/acl.h
+++ b/fs/9p/acl.h
@@ -16,7 +16,7 @@
 
 #ifdef CONFIG_9P_FS_POSIX_ACL
 extern int v9fs_get_acl(struct inode *, struct p9_fid *);
-extern int v9fs_check_acl(struct inode *inode, int mask, unsigned int flags);
+extern int v9fs_check_acl(struct inode *inode, int mask);
 extern int v9fs_acl_chmod(struct dentry *);
 extern int v9fs_set_create_acl(struct dentry *,
 			       struct posix_acl *, struct posix_acl *);
diff --git a/fs/9p/cache.c b/fs/9p/cache.c
index 5b335c5086a..945aa5f02f9 100644
--- a/fs/9p/cache.c
+++ b/fs/9p/cache.c
@@ -108,11 +108,10 @@ static uint16_t v9fs_cache_inode_get_key(const void *cookie_netfs_data,
 					 void *buffer, uint16_t bufmax)
 {
 	const struct v9fs_inode *v9inode = cookie_netfs_data;
-	memcpy(buffer, &v9inode->fscache_key->path,
-	       sizeof(v9inode->fscache_key->path));
+	memcpy(buffer, &v9inode->qid.path, sizeof(v9inode->qid.path));
 	P9_DPRINTK(P9_DEBUG_FSC, "inode %p get key %llu", &v9inode->vfs_inode,
-		   v9inode->fscache_key->path);
-	return sizeof(v9inode->fscache_key->path);
+		   v9inode->qid.path);
+	return sizeof(v9inode->qid.path);
 }
 
 static void v9fs_cache_inode_get_attr(const void *cookie_netfs_data,
@@ -129,11 +128,10 @@ static uint16_t v9fs_cache_inode_get_aux(const void *cookie_netfs_data,
 					 void *buffer, uint16_t buflen)
 {
 	const struct v9fs_inode *v9inode = cookie_netfs_data;
-	memcpy(buffer, &v9inode->fscache_key->version,
-	       sizeof(v9inode->fscache_key->version));
+	memcpy(buffer, &v9inode->qid.version, sizeof(v9inode->qid.version));
 	P9_DPRINTK(P9_DEBUG_FSC, "inode %p get aux %u", &v9inode->vfs_inode,
-		   v9inode->fscache_key->version);
-	return sizeof(v9inode->fscache_key->version);
+		   v9inode->qid.version);
+	return sizeof(v9inode->qid.version);
 }
 
 static enum
@@ -143,11 +141,11 @@ fscache_checkaux v9fs_cache_inode_check_aux(void *cookie_netfs_data,
 {
 	const struct v9fs_inode *v9inode = cookie_netfs_data;
 
-	if (buflen != sizeof(v9inode->fscache_key->version))
+	if (buflen != sizeof(v9inode->qid.version))
 		return FSCACHE_CHECKAUX_OBSOLETE;
 
-	if (memcmp(buffer, &v9inode->fscache_key->version,
-		   sizeof(v9inode->fscache_key->version)))
+	if (memcmp(buffer, &v9inode->qid.version,
+		   sizeof(v9inode->qid.version)))
 		return FSCACHE_CHECKAUX_OBSOLETE;
 
 	return FSCACHE_CHECKAUX_OKAY;
diff --git a/fs/9p/cache.h b/fs/9p/cache.h
index 049507a5b01..40cc54ced5d 100644
--- a/fs/9p/cache.h
+++ b/fs/9p/cache.h
@@ -93,15 +93,6 @@ static inline void v9fs_uncache_page(struct inode *inode, struct page *page)
 	BUG_ON(PageFsCache(page));
 }
 
-static inline void v9fs_fscache_set_key(struct inode *inode,
-					struct p9_qid *qid)
-{
-	struct v9fs_inode *v9inode = V9FS_I(inode);
-	spin_lock(&v9inode->fscache_lock);
-	v9inode->fscache_key = qid;
-	spin_unlock(&v9inode->fscache_lock);
-}
-
 static inline void v9fs_fscache_wait_on_page_write(struct inode *inode,
 						   struct page *page)
 {
diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c
index c82b017f51f..ef966188611 100644
--- a/fs/9p/v9fs.c
+++ b/fs/9p/v9fs.c
@@ -78,6 +78,25 @@ static const match_table_t tokens = {
 	{Opt_err, NULL}
 };
 
+/* Interpret mount options for cache mode */
+static int get_cache_mode(char *s)
+{
+	int version = -EINVAL;
+
+	if (!strcmp(s, "loose")) {
+		version = CACHE_LOOSE;
+		P9_DPRINTK(P9_DEBUG_9P, "Cache mode: loose\n");
+	} else if (!strcmp(s, "fscache")) {
+		version = CACHE_FSCACHE;
+		P9_DPRINTK(P9_DEBUG_9P, "Cache mode: fscache\n");
+	} else if (!strcmp(s, "none")) {
+		version = CACHE_NONE;
+		P9_DPRINTK(P9_DEBUG_9P, "Cache mode: none\n");
+	} else
+		printk(KERN_INFO "9p: Unknown Cache mode %s.\n", s);
+	return version;
+}
+
 /**
  * v9fs_parse_options - parse mount options into session structure
  * @v9ses: existing v9fs session information
@@ -97,7 +116,7 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts)
 	/* setup defaults */
 	v9ses->afid = ~0;
 	v9ses->debug = 0;
-	v9ses->cache = 0;
+	v9ses->cache = CACHE_NONE;
 #ifdef CONFIG_9P_FSCACHE
 	v9ses->cachetag = NULL;
 #endif
@@ -171,13 +190,13 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts)
 				  "problem allocating copy of cache arg\n");
 				goto free_and_return;
 			}
+			ret = get_cache_mode(s);
+			if (ret == -EINVAL) {
+				kfree(s);
+				goto free_and_return;
+			}
 
-			if (strcmp(s, "loose") == 0)
-				v9ses->cache = CACHE_LOOSE;
-			else if (strcmp(s, "fscache") == 0)
-				v9ses->cache = CACHE_FSCACHE;
-			else
-				v9ses->cache = CACHE_NONE;
+			v9ses->cache = ret;
 			kfree(s);
 			break;
 
@@ -200,9 +219,15 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts)
 			} else {
 				v9ses->flags |= V9FS_ACCESS_SINGLE;
 				v9ses->uid = simple_strtoul(s, &e, 10);
-				if (*e != '\0')
-					v9ses->uid = ~0;
+				if (*e != '\0') {
+					ret = -EINVAL;
+					printk(KERN_INFO "9p: Unknown access "
+							"argument %s.\n", s);
+					kfree(s);
+					goto free_and_return;
+				}
 			}
+
 			kfree(s);
 			break;
 
@@ -487,8 +512,8 @@ static void v9fs_inode_init_once(void *foo)
 	struct v9fs_inode *v9inode = (struct v9fs_inode *)foo;
 #ifdef CONFIG_9P_FSCACHE
 	v9inode->fscache = NULL;
-	v9inode->fscache_key = NULL;
 #endif
+	memset(&v9inode->qid, 0, sizeof(v9inode->qid));
 	inode_init_once(&v9inode->vfs_inode);
 }
 
diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h
index e5ebedfc5ed..e78956cbd70 100644
--- a/fs/9p/v9fs.h
+++ b/fs/9p/v9fs.h
@@ -125,8 +125,8 @@ struct v9fs_inode {
 #ifdef CONFIG_9P_FSCACHE
 	spinlock_t fscache_lock;
 	struct fscache_cookie *fscache;
-	struct p9_qid *fscache_key;
 #endif
+	struct p9_qid qid;
 	unsigned int cache_validity;
 	struct p9_fid *writeback_fid;
 	struct mutex v_mutex;
@@ -153,13 +153,13 @@ extern void v9fs_vfs_put_link(struct dentry *dentry, struct nameidata *nd,
 			void *p);
 extern struct inode *v9fs_inode_from_fid(struct v9fs_session_info *v9ses,
 					 struct p9_fid *fid,
-					 struct super_block *sb);
+					 struct super_block *sb, int new);
 extern const struct inode_operations v9fs_dir_inode_operations_dotl;
 extern const struct inode_operations v9fs_file_inode_operations_dotl;
 extern const struct inode_operations v9fs_symlink_inode_operations_dotl;
 extern struct inode *v9fs_inode_from_fid_dotl(struct v9fs_session_info *v9ses,
 					      struct p9_fid *fid,
-					      struct super_block *sb);
+					      struct super_block *sb, int new);
 
 /* other default globals */
 #define V9FS_PORT	564
@@ -201,8 +201,27 @@ v9fs_get_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid,
 			struct super_block *sb)
 {
 	if (v9fs_proto_dotl(v9ses))
-		return v9fs_inode_from_fid_dotl(v9ses, fid, sb);
+		return v9fs_inode_from_fid_dotl(v9ses, fid, sb, 0);
 	else
-		return v9fs_inode_from_fid(v9ses, fid, sb);
+		return v9fs_inode_from_fid(v9ses, fid, sb, 0);
 }
+
+/**
+ * v9fs_get_new_inode_from_fid - Helper routine to populate an inode by
+ * issuing a attribute request
+ * @v9ses: session information
+ * @fid: fid to issue attribute request for
+ * @sb: superblock on which to create inode
+ *
+ */
+static inline struct inode *
+v9fs_get_new_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid,
+			    struct super_block *sb)
+{
+	if (v9fs_proto_dotl(v9ses))
+		return v9fs_inode_from_fid_dotl(v9ses, fid, sb, 1);
+	else
+		return v9fs_inode_from_fid(v9ses, fid, sb, 1);
+}
+
 #endif
diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h
index 4014160903a..46ce357ca1a 100644
--- a/fs/9p/v9fs_vfs.h
+++ b/fs/9p/v9fs_vfs.h
@@ -70,7 +70,8 @@ ssize_t v9fs_file_readn(struct file *, char *, char __user *, u32, u64);
 ssize_t v9fs_fid_readn(struct p9_fid *, char *, char __user *, u32, u64);
 void v9fs_blank_wstat(struct p9_wstat *wstat);
 int v9fs_vfs_setattr_dotl(struct dentry *, struct iattr *);
-int v9fs_file_fsync_dotl(struct file *filp, int datasync);
+int v9fs_file_fsync_dotl(struct file *filp, loff_t start, loff_t end,
+			 int datasync);
 ssize_t v9fs_file_write_internal(struct inode *, struct p9_fid *,
 				 const char __user *, size_t, loff_t *, int);
 int v9fs_refresh_inode(struct p9_fid *fid, struct inode *inode);
diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index ffed55817f0..3c173fcc2c5 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -519,32 +519,50 @@ out:
 }
 
 
-static int v9fs_file_fsync(struct file *filp, int datasync)
+static int v9fs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			   int datasync)
 {
 	struct p9_fid *fid;
+	struct inode *inode = filp->f_mapping->host;
 	struct p9_wstat wstat;
 	int retval;
 
+	retval = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (retval)
+		return retval;
+
+	mutex_lock(&inode->i_mutex);
 	P9_DPRINTK(P9_DEBUG_VFS, "filp %p datasync %x\n", filp, datasync);
 
 	fid = filp->private_data;
 	v9fs_blank_wstat(&wstat);
 
 	retval = p9_client_wstat(fid, &wstat);
+	mutex_unlock(&inode->i_mutex);
+
 	return retval;
 }
 
-int v9fs_file_fsync_dotl(struct file *filp, int datasync)
+int v9fs_file_fsync_dotl(struct file *filp, loff_t start, loff_t end,
+			 int datasync)
 {
 	struct p9_fid *fid;
+	struct inode *inode = filp->f_mapping->host;
 	int retval;
 
+	retval = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (retval)
+		return retval;
+
+	mutex_lock(&inode->i_mutex);
 	P9_DPRINTK(P9_DEBUG_VFS, "v9fs_file_fsync_dotl: filp %p datasync %x\n",
 			filp, datasync);
 
 	fid = filp->private_data;
 
 	retval = p9_client_fsync(fid, datasync);
+	mutex_unlock(&inode->i_mutex);
+
 	return retval;
 }
 
diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c
index 7f6c6770319..8bb5507e822 100644
--- a/fs/9p/vfs_inode.c
+++ b/fs/9p/vfs_inode.c
@@ -216,7 +216,6 @@ struct inode *v9fs_alloc_inode(struct super_block *sb)
 		return NULL;
 #ifdef CONFIG_9P_FSCACHE
 	v9inode->fscache = NULL;
-	v9inode->fscache_key = NULL;
 	spin_lock_init(&v9inode->fscache_lock);
 #endif
 	v9inode->writeback_fid = NULL;
@@ -433,17 +432,60 @@ void v9fs_evict_inode(struct inode *inode)
 	}
 }
 
+static int v9fs_test_inode(struct inode *inode, void *data)
+{
+	int umode;
+	struct v9fs_inode *v9inode = V9FS_I(inode);
+	struct p9_wstat *st = (struct p9_wstat *)data;
+	struct v9fs_session_info *v9ses = v9fs_inode2v9ses(inode);
+
+	umode = p9mode2unixmode(v9ses, st->mode);
+	/* don't match inode of different type */
+	if ((inode->i_mode & S_IFMT) != (umode & S_IFMT))
+		return 0;
+
+	/* compare qid details */
+	if (memcmp(&v9inode->qid.version,
+		   &st->qid.version, sizeof(v9inode->qid.version)))
+		return 0;
+
+	if (v9inode->qid.type != st->qid.type)
+		return 0;
+	return 1;
+}
+
+static int v9fs_test_new_inode(struct inode *inode, void *data)
+{
+	return 0;
+}
+
+static int v9fs_set_inode(struct inode *inode,  void *data)
+{
+	struct v9fs_inode *v9inode = V9FS_I(inode);
+	struct p9_wstat *st = (struct p9_wstat *)data;
+
+	memcpy(&v9inode->qid, &st->qid, sizeof(st->qid));
+	return 0;
+}
+
 static struct inode *v9fs_qid_iget(struct super_block *sb,
 				   struct p9_qid *qid,
-				   struct p9_wstat *st)
+				   struct p9_wstat *st,
+				   int new)
 {
 	int retval, umode;
 	unsigned long i_ino;
 	struct inode *inode;
 	struct v9fs_session_info *v9ses = sb->s_fs_info;
+	int (*test)(struct inode *, void *);
+
+	if (new)
+		test = v9fs_test_new_inode;
+	else
+		test = v9fs_test_inode;
 
 	i_ino = v9fs_qid2ino(qid);
-	inode = iget_locked(sb, i_ino);
+	inode = iget5_locked(sb, i_ino, test, v9fs_set_inode, st);
 	if (!inode)
 		return ERR_PTR(-ENOMEM);
 	if (!(inode->i_state & I_NEW))
@@ -453,6 +495,7 @@ static struct inode *v9fs_qid_iget(struct super_block *sb,
 	 * FIXME!! we may need support for stale inodes
 	 * later.
 	 */
+	inode->i_ino = i_ino;
 	umode = p9mode2unixmode(v9ses, st->mode);
 	retval = v9fs_init_inode(v9ses, inode, umode);
 	if (retval)
@@ -460,7 +503,6 @@ static struct inode *v9fs_qid_iget(struct super_block *sb,
 
 	v9fs_stat2inode(st, inode, sb);
 #ifdef CONFIG_9P_FSCACHE
-	v9fs_fscache_set_key(inode, &st->qid);
 	v9fs_cache_inode_get_cookie(inode);
 #endif
 	unlock_new_inode(inode);
@@ -474,7 +516,7 @@ error:
 
 struct inode *
 v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid,
-		    struct super_block *sb)
+		    struct super_block *sb, int new)
 {
 	struct p9_wstat *st;
 	struct inode *inode = NULL;
@@ -483,7 +525,7 @@ v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid,
 	if (IS_ERR(st))
 		return ERR_CAST(st);
 
-	inode = v9fs_qid_iget(sb, &st->qid, st);
+	inode = v9fs_qid_iget(sb, &st->qid, st, new);
 	p9stat_free(st);
 	kfree(st);
 	return inode;
@@ -492,38 +534,50 @@ v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid,
 /**
  * v9fs_remove - helper function to remove files and directories
  * @dir: directory inode that is being deleted
- * @file:  dentry that is being deleted
+ * @dentry:  dentry that is being deleted
  * @rmdir: removing a directory
  *
  */
 
-static int v9fs_remove(struct inode *dir, struct dentry *file, int rmdir)
+static int v9fs_remove(struct inode *dir, struct dentry *dentry, int flags)
 {
-	int retval;
-	struct p9_fid *v9fid;
-	struct inode *file_inode;
-
-	P9_DPRINTK(P9_DEBUG_VFS, "inode: %p dentry: %p rmdir: %d\n", dir, file,
-		rmdir);
+	struct inode *inode;
+	int retval = -EOPNOTSUPP;
+	struct p9_fid *v9fid, *dfid;
+	struct v9fs_session_info *v9ses;
 
-	file_inode = file->d_inode;
-	v9fid = v9fs_fid_clone(file);
-	if (IS_ERR(v9fid))
-		return PTR_ERR(v9fid);
+	P9_DPRINTK(P9_DEBUG_VFS, "inode: %p dentry: %p rmdir: %x\n",
+		   dir, dentry, flags);
 
-	retval = p9_client_remove(v9fid);
+	v9ses = v9fs_inode2v9ses(dir);
+	inode = dentry->d_inode;
+	dfid = v9fs_fid_lookup(dentry->d_parent);
+	if (IS_ERR(dfid)) {
+		retval = PTR_ERR(dfid);
+		P9_DPRINTK(P9_DEBUG_VFS, "fid lookup failed %d\n", retval);
+		return retval;
+	}
+	if (v9fs_proto_dotl(v9ses))
+		retval = p9_client_unlinkat(dfid, dentry->d_name.name, flags);
+	if (retval == -EOPNOTSUPP) {
+		/* Try the one based on path */
+		v9fid = v9fs_fid_clone(dentry);
+		if (IS_ERR(v9fid))
+			return PTR_ERR(v9fid);
+		retval = p9_client_remove(v9fid);
+	}
 	if (!retval) {
 		/*
 		 * directories on unlink should have zero
 		 * link count
 		 */
-		if (rmdir) {
-			clear_nlink(file_inode);
+		if (flags & AT_REMOVEDIR) {
+			clear_nlink(inode);
 			drop_nlink(dir);
 		} else
-			drop_nlink(file_inode);
+			drop_nlink(inode);
 
-		v9fs_invalidate_inode_attr(file_inode);
+		v9fs_invalidate_inode_attr(inode);
 		v9fs_invalidate_inode_attr(dir);
 	}
 	return retval;
@@ -585,7 +639,7 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir,
 	}
 
 	/* instantiate inode and assign the unopened fid to the dentry */
-	inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb);
+	inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err);
@@ -633,8 +687,8 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	fid = NULL;
 	v9ses = v9fs_inode2v9ses(dir);
 	perm = unixmode2p9mode(v9ses, mode);
-	if (nd && nd->flags & LOOKUP_OPEN)
-		flags = nd->intent.open.flags - 1;
+	if (nd)
+		flags = nd->intent.open.flags;
 	else
 		flags = O_RDWR;
 
@@ -649,7 +703,7 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 
 	v9fs_invalidate_inode_attr(dir);
 	/* if we are opening a file, assign the open fid to the file */
-	if (nd && nd->flags & LOOKUP_OPEN) {
+	if (nd) {
 		v9inode = V9FS_I(dentry->d_inode);
 		mutex_lock(&v9inode->v_mutex);
 		if (v9ses->cache && !v9inode->writeback_fid &&
@@ -814,7 +868,7 @@ int v9fs_vfs_unlink(struct inode *i, struct dentry *d)
 
 int v9fs_vfs_rmdir(struct inode *i, struct dentry *d)
 {
-	return v9fs_remove(i, d, 1);
+	return v9fs_remove(i, d, AT_REMOVEDIR);
 }
 
 /**
@@ -862,9 +916,12 @@ v9fs_vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 
 	down_write(&v9ses->rename_sem);
 	if (v9fs_proto_dotl(v9ses)) {
-		retval = p9_client_rename(oldfid, newdirfid,
-					(char *) new_dentry->d_name.name);
-		if (retval != -ENOSYS)
+		retval = p9_client_renameat(olddirfid, old_dentry->d_name.name,
+					    newdirfid, new_dentry->d_name.name);
+		if (retval == -EOPNOTSUPP)
+			retval = p9_client_rename(oldfid, newdirfid,
+						  new_dentry->d_name.name);
+		if (retval != -EOPNOTSUPP)
 			goto clunk_newdir;
 	}
 	if (old_dentry->d_parent != new_dentry->d_parent) {
@@ -889,11 +946,6 @@ clunk_newdir:
 				clear_nlink(new_inode);
 			else
 				drop_nlink(new_inode);
-			/*
-			 * Work around vfs rename rehash bug with
-			 * FS_RENAME_DOES_D_MOVE
-			 */
-			v9fs_invalidate_inode_attr(new_inode);
 		}
 		if (S_ISDIR(old_inode->i_mode)) {
 			if (!new_inode)
diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c
index 691c78f58be..276f4a69ecd 100644
--- a/fs/9p/vfs_inode_dotl.c
+++ b/fs/9p/vfs_inode_dotl.c
@@ -86,18 +86,63 @@ static struct dentry *v9fs_dentry_from_dir_inode(struct inode *inode)
 	return dentry;
 }
 
+static int v9fs_test_inode_dotl(struct inode *inode, void *data)
+{
+	struct v9fs_inode *v9inode = V9FS_I(inode);
+	struct p9_stat_dotl *st = (struct p9_stat_dotl *)data;
+
+	/* don't match inode of different type */
+	if ((inode->i_mode & S_IFMT) != (st->st_mode & S_IFMT))
+		return 0;
+
+	if (inode->i_generation != st->st_gen)
+		return 0;
+
+	/* compare qid details */
+	if (memcmp(&v9inode->qid.version,
+		   &st->qid.version, sizeof(v9inode->qid.version)))
+		return 0;
+
+	if (v9inode->qid.type != st->qid.type)
+		return 0;
+	return 1;
+}
+
+/* Always get a new inode */
+static int v9fs_test_new_inode_dotl(struct inode *inode, void *data)
+{
+	return 0;
+}
+
+static int v9fs_set_inode_dotl(struct inode *inode,  void *data)
+{
+	struct v9fs_inode *v9inode = V9FS_I(inode);
+	struct p9_stat_dotl *st = (struct p9_stat_dotl *)data;
+
+	memcpy(&v9inode->qid, &st->qid, sizeof(st->qid));
+	inode->i_generation = st->st_gen;
+	return 0;
+}
+
 static struct inode *v9fs_qid_iget_dotl(struct super_block *sb,
 					struct p9_qid *qid,
 					struct p9_fid *fid,
-					struct p9_stat_dotl *st)
+					struct p9_stat_dotl *st,
+					int new)
 {
 	int retval;
 	unsigned long i_ino;
 	struct inode *inode;
 	struct v9fs_session_info *v9ses = sb->s_fs_info;
+	int (*test)(struct inode *, void *);
+
+	if (new)
+		test = v9fs_test_new_inode_dotl;
+	else
+		test = v9fs_test_inode_dotl;
 
 	i_ino = v9fs_qid2ino(qid);
-	inode = iget_locked(sb, i_ino);
+	inode = iget5_locked(sb, i_ino, test, v9fs_set_inode_dotl, st);
 	if (!inode)
 		return ERR_PTR(-ENOMEM);
 	if (!(inode->i_state & I_NEW))
@@ -107,13 +152,13 @@ static struct inode *v9fs_qid_iget_dotl(struct super_block *sb,
 	 * FIXME!! we may need support for stale inodes
 	 * later.
 	 */
+	inode->i_ino = i_ino;
 	retval = v9fs_init_inode(v9ses, inode, st->st_mode);
 	if (retval)
 		goto error;
 
 	v9fs_stat2inode_dotl(st, inode);
 #ifdef CONFIG_9P_FSCACHE
-	v9fs_fscache_set_key(inode, &st->qid);
 	v9fs_cache_inode_get_cookie(inode);
 #endif
 	retval = v9fs_get_acl(inode, fid);
@@ -131,16 +176,16 @@ error:
 
 struct inode *
 v9fs_inode_from_fid_dotl(struct v9fs_session_info *v9ses, struct p9_fid *fid,
-			 struct super_block *sb)
+			 struct super_block *sb, int new)
 {
 	struct p9_stat_dotl *st;
 	struct inode *inode = NULL;
 
-	st = p9_client_getattr_dotl(fid, P9_STATS_BASIC);
+	st = p9_client_getattr_dotl(fid, P9_STATS_BASIC | P9_STATS_GEN);
 	if (IS_ERR(st))
 		return ERR_CAST(st);
 
-	inode = v9fs_qid_iget_dotl(sb, &st->qid, fid, st);
+	inode = v9fs_qid_iget_dotl(sb, &st->qid, fid, st, new);
 	kfree(st);
 	return inode;
 }
@@ -173,8 +218,8 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode,
 	struct posix_acl *pacl = NULL, *dacl = NULL;
 
 	v9ses = v9fs_inode2v9ses(dir);
-	if (nd && nd->flags & LOOKUP_OPEN)
-		flags = nd->intent.open.flags - 1;
+	if (nd)
+		flags = nd->intent.open.flags;
 	else {
 		/*
 		 * create call without LOOKUP_OPEN is due
@@ -230,7 +275,7 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode,
 		fid = NULL;
 		goto error;
 	}
-	inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb);
+	inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err);
@@ -350,7 +395,7 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir,
 			goto error;
 		}
 
-		inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb);
+		inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb);
 		if (IS_ERR(inode)) {
 			err = PTR_ERR(inode);
 			P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n",
@@ -547,7 +592,7 @@ v9fs_stat2inode_dotl(struct p9_stat_dotl *stat, struct inode *inode)
 			inode->i_blocks = stat->st_blocks;
 	}
 	if (stat->st_result_mask & P9_STATS_GEN)
-			inode->i_generation = stat->st_gen;
+		inode->i_generation = stat->st_gen;
 
 	/* Currently we don't support P9_STATS_BTIME and P9_STATS_DATA_VERSION
 	 * because the inode structure does not have fields for them.
@@ -603,7 +648,7 @@ v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry,
 		}
 
 		/* instantiate inode and assign the unopened fid to dentry */
-		inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb);
+		inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb);
 		if (IS_ERR(inode)) {
 			err = PTR_ERR(inode);
 			P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n",
@@ -756,7 +801,7 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int omode,
 			goto error;
 		}
 
-		inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb);
+		inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb);
 		if (IS_ERR(inode)) {
 			err = PTR_ERR(inode);
 			P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n",
diff --git a/fs/affs/affs.h b/fs/affs/affs.h
index 0e95f73a702..c2b9c79eb64 100644
--- a/fs/affs/affs.h
+++ b/fs/affs/affs.h
@@ -182,7 +182,7 @@ extern int			 affs_add_entry(struct inode *dir, struct inode *inode, struct dent
 
 void		affs_free_prealloc(struct inode *inode);
 extern void	affs_truncate(struct inode *);
-int		affs_file_fsync(struct file *, int);
+int		affs_file_fsync(struct file *, loff_t, loff_t, int);
 
 /* dir.c */
 
diff --git a/fs/affs/file.c b/fs/affs/file.c
index acf321b70fc..2f4c935cb32 100644
--- a/fs/affs/file.c
+++ b/fs/affs/file.c
@@ -923,14 +923,20 @@ affs_truncate(struct inode *inode)
 	affs_free_prealloc(inode);
 }
 
-int affs_file_fsync(struct file *filp, int datasync)
+int affs_file_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int ret, err;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	ret = write_inode_now(inode, 0);
 	err = sync_blockdev(inode->i_sb->s_bdev);
 	if (!ret)
 		ret = err;
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
diff --git a/fs/afs/afs_vl.h b/fs/afs/afs_vl.h
index 8bbefe009ed..800f607ffaf 100644
--- a/fs/afs/afs_vl.h
+++ b/fs/afs/afs_vl.h
@@ -49,7 +49,7 @@ enum AFSVL_Errors {
 	AFSVL_BADVOLOPER 	= 363542,	/* Bad volume operation code */
 	AFSVL_BADRELLOCKTYPE 	= 363543,	/* Bad release lock type */
 	AFSVL_RERELEASE 	= 363544,	/* Status report: last release was aborted */
-	AFSVL_BADSERVERFLAG 	= 363545,	/* Invalid replication site server °ag */
+	AFSVL_BADSERVERFLAG 	= 363545,	/* Invalid replication site server flag */
 	AFSVL_PERM 		= 363546,	/* No permission access */
 	AFSVL_NOMEM 		= 363547,	/* malloc/realloc failed to alloc enough memory */
 };
diff --git a/fs/afs/dir.c b/fs/afs/dir.c
index 20c106f2492..1b0b1955001 100644
--- a/fs/afs/dir.c
+++ b/fs/afs/dir.c
@@ -584,11 +584,11 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry,
 
 success:
 	d_add(dentry, inode);
-	_leave(" = 0 { vn=%u u=%u } -> { ino=%lu v=%llu }",
+	_leave(" = 0 { vn=%u u=%u } -> { ino=%lu v=%u }",
 	       fid.vnode,
 	       fid.unique,
 	       dentry->d_inode->i_ino,
-	       (unsigned long long)dentry->d_inode->i_version);
+	       dentry->d_inode->i_generation);
 
 	return NULL;
 }
@@ -671,10 +671,10 @@ static int afs_d_revalidate(struct dentry *dentry, struct nameidata *nd)
 		 * been deleted and replaced, and the original vnode ID has
 		 * been reused */
 		if (fid.unique != vnode->fid.unique) {
-			_debug("%s: file deleted (uq %u -> %u I:%llu)",
+			_debug("%s: file deleted (uq %u -> %u I:%u)",
 			       dentry->d_name.name, fid.unique,
 			       vnode->fid.unique,
-			       (unsigned long long)dentry->d_inode->i_version);
+			       dentry->d_inode->i_generation);
 			spin_lock(&vnode->lock);
 			set_bit(AFS_VNODE_DELETED, &vnode->flags);
 			spin_unlock(&vnode->lock);
diff --git a/fs/afs/fsclient.c b/fs/afs/fsclient.c
index 4bd0218473a..346e3289abd 100644
--- a/fs/afs/fsclient.c
+++ b/fs/afs/fsclient.c
@@ -89,7 +89,7 @@ static void xdr_decode_AFSFetchStatus(const __be32 **_bp,
 			i_size_write(&vnode->vfs_inode, size);
 			vnode->vfs_inode.i_uid = status->owner;
 			vnode->vfs_inode.i_gid = status->group;
-			vnode->vfs_inode.i_version = vnode->fid.unique;
+			vnode->vfs_inode.i_generation = vnode->fid.unique;
 			vnode->vfs_inode.i_nlink = status->nlink;
 
 			mode = vnode->vfs_inode.i_mode;
@@ -102,6 +102,7 @@ static void xdr_decode_AFSFetchStatus(const __be32 **_bp,
 		vnode->vfs_inode.i_ctime.tv_sec	= status->mtime_server;
 		vnode->vfs_inode.i_mtime	= vnode->vfs_inode.i_ctime;
 		vnode->vfs_inode.i_atime	= vnode->vfs_inode.i_ctime;
+		vnode->vfs_inode.i_version	= data_version;
 	}
 
 	expected_version = status->data_version;
diff --git a/fs/afs/inode.c b/fs/afs/inode.c
index db66c520147..0fdab6e03d8 100644
--- a/fs/afs/inode.c
+++ b/fs/afs/inode.c
@@ -75,7 +75,8 @@ static int afs_inode_map_status(struct afs_vnode *vnode, struct key *key)
 	inode->i_ctime.tv_nsec	= 0;
 	inode->i_atime		= inode->i_mtime = inode->i_ctime;
 	inode->i_blocks		= 0;
-	inode->i_version	= vnode->fid.unique;
+	inode->i_generation	= vnode->fid.unique;
+	inode->i_version	= vnode->status.data_version;
 	inode->i_mapping->a_ops	= &afs_fs_aops;
 
 	/* check to see whether a symbolic link is really a mountpoint */
@@ -100,7 +101,7 @@ static int afs_iget5_test(struct inode *inode, void *opaque)
 	struct afs_iget_data *data = opaque;
 
 	return inode->i_ino == data->fid.vnode &&
-		inode->i_version == data->fid.unique;
+		inode->i_generation == data->fid.unique;
 }
 
 /*
@@ -122,7 +123,7 @@ static int afs_iget5_set(struct inode *inode, void *opaque)
 	struct afs_vnode *vnode = AFS_FS_I(inode);
 
 	inode->i_ino = data->fid.vnode;
-	inode->i_version = data->fid.unique;
+	inode->i_generation = data->fid.unique;
 	vnode->fid = data->fid;
 	vnode->volume = data->volume;
 
@@ -380,8 +381,7 @@ int afs_getattr(struct vfsmount *mnt, struct dentry *dentry,
 
 	inode = dentry->d_inode;
 
-	_enter("{ ino=%lu v=%llu }", inode->i_ino,
-		(unsigned long long)inode->i_version);
+	_enter("{ ino=%lu v=%u }", inode->i_ino, inode->i_generation);
 
 	generic_fillattr(inode, stat);
 	return 0;
diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index 5a9b6843bac..d2b0888126d 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -627,7 +627,7 @@ extern void afs_clear_permits(struct afs_vnode *);
 extern void afs_cache_permit(struct afs_vnode *, struct key *, long);
 extern void afs_zap_permits(struct rcu_head *);
 extern struct key *afs_request_key(struct afs_cell *);
-extern int afs_permission(struct inode *, int, unsigned int);
+extern int afs_permission(struct inode *, int);
 
 /*
  * server.c
@@ -750,7 +750,7 @@ 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);
 extern int afs_writeback_all(struct afs_vnode *);
-extern int afs_fsync(struct file *, int);
+extern int afs_fsync(struct file *, loff_t, loff_t, int);
 
 
 /*****************************************************************************/
diff --git a/fs/afs/security.c b/fs/afs/security.c
index f44b9d35537..8d010422dc8 100644
--- a/fs/afs/security.c
+++ b/fs/afs/security.c
@@ -285,14 +285,14 @@ static int afs_check_permit(struct afs_vnode *vnode, struct key *key,
  * - AFS ACLs are attached to directories only, and a file is controlled by its
  *   parent directory's ACL
  */
-int afs_permission(struct inode *inode, int mask, unsigned int flags)
+int afs_permission(struct inode *inode, int mask)
 {
 	struct afs_vnode *vnode = AFS_FS_I(inode);
 	afs_access_t uninitialized_var(access);
 	struct key *key;
 	int ret;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	_enter("{{%x:%u},%lx},%x,",
@@ -350,7 +350,7 @@ int afs_permission(struct inode *inode, int mask, unsigned int flags)
 	}
 
 	key_put(key);
-	ret = generic_permission(inode, mask, flags, NULL);
+	ret = generic_permission(inode, mask);
 	_leave(" = %d", ret);
 	return ret;
 
diff --git a/fs/afs/super.c b/fs/afs/super.c
index fb240e8766d..356dcf0929e 100644
--- a/fs/afs/super.c
+++ b/fs/afs/super.c
@@ -31,8 +31,8 @@
 static void afs_i_init_once(void *foo);
 static struct dentry *afs_mount(struct file_system_type *fs_type,
 		      int flags, const char *dev_name, void *data);
+static void afs_kill_super(struct super_block *sb);
 static struct inode *afs_alloc_inode(struct super_block *sb);
-static void afs_put_super(struct super_block *sb);
 static void afs_destroy_inode(struct inode *inode);
 static int afs_statfs(struct dentry *dentry, struct kstatfs *buf);
 
@@ -40,7 +40,7 @@ struct file_system_type afs_fs_type = {
 	.owner		= THIS_MODULE,
 	.name		= "afs",
 	.mount		= afs_mount,
-	.kill_sb	= kill_anon_super,
+	.kill_sb	= afs_kill_super,
 	.fs_flags	= 0,
 };
 
@@ -50,7 +50,6 @@ static const struct super_operations afs_super_ops = {
 	.drop_inode	= afs_drop_inode,
 	.destroy_inode	= afs_destroy_inode,
 	.evict_inode	= afs_evict_inode,
-	.put_super	= afs_put_super,
 	.show_options	= generic_show_options,
 };
 
@@ -282,19 +281,25 @@ static int afs_parse_device_name(struct afs_mount_params *params,
  */
 static int afs_test_super(struct super_block *sb, void *data)
 {
-	struct afs_mount_params *params = data;
+	struct afs_super_info *as1 = data;
 	struct afs_super_info *as = sb->s_fs_info;
 
-	return as->volume == params->volume;
+	return as->volume == as1->volume;
+}
+
+static int afs_set_super(struct super_block *sb, void *data)
+{
+	sb->s_fs_info = data;
+	return set_anon_super(sb, NULL);
 }
 
 /*
  * fill in the superblock
  */
-static int afs_fill_super(struct super_block *sb, void *data)
+static int afs_fill_super(struct super_block *sb,
+			  struct afs_mount_params *params)
 {
-	struct afs_mount_params *params = data;
-	struct afs_super_info *as = NULL;
+	struct afs_super_info *as = sb->s_fs_info;
 	struct afs_fid fid;
 	struct dentry *root = NULL;
 	struct inode *inode = NULL;
@@ -302,23 +307,13 @@ static int afs_fill_super(struct super_block *sb, void *data)
 
 	_enter("");
 
-	/* allocate a superblock info record */
-	as = kzalloc(sizeof(struct afs_super_info), GFP_KERNEL);
-	if (!as) {
-		_leave(" = -ENOMEM");
-		return -ENOMEM;
-	}
-
-	afs_get_volume(params->volume);
-	as->volume = params->volume;
-
 	/* fill in the superblock */
 	sb->s_blocksize		= PAGE_CACHE_SIZE;
 	sb->s_blocksize_bits	= PAGE_CACHE_SHIFT;
 	sb->s_magic		= AFS_FS_MAGIC;
 	sb->s_op		= &afs_super_ops;
-	sb->s_fs_info		= as;
 	sb->s_bdi		= &as->volume->bdi;
+	strlcpy(sb->s_id, as->volume->vlocation->vldb.name, sizeof(sb->s_id));
 
 	/* allocate the root inode and dentry */
 	fid.vid		= as->volume->vid;
@@ -326,7 +321,7 @@ static int afs_fill_super(struct super_block *sb, void *data)
 	fid.unique	= 1;
 	inode = afs_iget(sb, params->key, &fid, NULL, NULL);
 	if (IS_ERR(inode))
-		goto error_inode;
+		return PTR_ERR(inode);
 
 	if (params->autocell)
 		set_bit(AFS_VNODE_AUTOCELL, &AFS_FS_I(inode)->flags);
@@ -342,16 +337,8 @@ static int afs_fill_super(struct super_block *sb, void *data)
 	_leave(" = 0");
 	return 0;
 
-error_inode:
-	ret = PTR_ERR(inode);
-	inode = NULL;
 error:
 	iput(inode);
-	afs_put_volume(as->volume);
-	kfree(as);
-
-	sb->s_fs_info = NULL;
-
 	_leave(" = %d", ret);
 	return ret;
 }
@@ -367,6 +354,7 @@ static struct dentry *afs_mount(struct file_system_type *fs_type,
 	struct afs_volume *vol;
 	struct key *key;
 	char *new_opts = kstrdup(options, GFP_KERNEL);
+	struct afs_super_info *as;
 	int ret;
 
 	_enter(",,%s,%p", dev_name, options);
@@ -399,12 +387,22 @@ static struct dentry *afs_mount(struct file_system_type *fs_type,
 		ret = PTR_ERR(vol);
 		goto error;
 	}
-	params.volume = vol;
+
+	/* allocate a superblock info record */
+	as = kzalloc(sizeof(struct afs_super_info), GFP_KERNEL);
+	if (!as) {
+		ret = -ENOMEM;
+		afs_put_volume(vol);
+		goto error;
+	}
+	as->volume = vol;
 
 	/* allocate a deviceless superblock */
-	sb = sget(fs_type, afs_test_super, set_anon_super, &params);
+	sb = sget(fs_type, afs_test_super, afs_set_super, as);
 	if (IS_ERR(sb)) {
 		ret = PTR_ERR(sb);
+		afs_put_volume(vol);
+		kfree(as);
 		goto error;
 	}
 
@@ -422,16 +420,16 @@ static struct dentry *afs_mount(struct file_system_type *fs_type,
 	} else {
 		_debug("reuse");
 		ASSERTCMP(sb->s_flags, &, MS_ACTIVE);
+		afs_put_volume(vol);
+		kfree(as);
 	}
 
-	afs_put_volume(params.volume);
 	afs_put_cell(params.cell);
 	kfree(new_opts);
 	_leave(" = 0 [%p]", sb);
 	return dget(sb->s_root);
 
 error:
-	afs_put_volume(params.volume);
 	afs_put_cell(params.cell);
 	key_put(params.key);
 	kfree(new_opts);
@@ -439,18 +437,12 @@ error:
 	return ERR_PTR(ret);
 }
 
-/*
- * finish the unmounting process on the superblock
- */
-static void afs_put_super(struct super_block *sb)
+static void afs_kill_super(struct super_block *sb)
 {
 	struct afs_super_info *as = sb->s_fs_info;
-
-	_enter("");
-
+	kill_anon_super(sb);
 	afs_put_volume(as->volume);
-
-	_leave("");
+	kfree(as);
 }
 
 /*
diff --git a/fs/afs/write.c b/fs/afs/write.c
index 789b3afb342..9aa52d93c73 100644
--- a/fs/afs/write.c
+++ b/fs/afs/write.c
@@ -84,23 +84,21 @@ void afs_put_writeback(struct afs_writeback *wb)
  * partly or wholly fill a page that's under preparation for writing
  */
 static int afs_fill_page(struct afs_vnode *vnode, struct key *key,
-			 loff_t pos, unsigned len, struct page *page)
+			 loff_t pos, struct page *page)
 {
 	loff_t i_size;
-	unsigned eof;
 	int ret;
+	int len;
 
-	_enter(",,%llu,%u", (unsigned long long)pos, len);
-
-	ASSERTCMP(len, <=, PAGE_CACHE_SIZE);
+	_enter(",,%llu", (unsigned long long)pos);
 
 	i_size = i_size_read(&vnode->vfs_inode);
-	if (pos + len > i_size)
-		eof = i_size;
+	if (pos + PAGE_CACHE_SIZE > i_size)
+		len = i_size - pos;
 	else
-		eof = PAGE_CACHE_SIZE;
+		len = PAGE_CACHE_SIZE;
 
-	ret = afs_vnode_fetch_data(vnode, key, 0, eof, page);
+	ret = afs_vnode_fetch_data(vnode, key, pos, len, page);
 	if (ret < 0) {
 		if (ret == -ENOENT) {
 			_debug("got NOENT from server"
@@ -153,9 +151,8 @@ int afs_write_begin(struct file *file, struct address_space *mapping,
 	*pagep = page;
 	/* page won't leak in error case: it eventually gets cleaned off LRU */
 
-	if (!PageUptodate(page)) {
-		_debug("not up to date");
-		ret = afs_fill_page(vnode, key, pos, len, page);
+	if (!PageUptodate(page) && len != PAGE_CACHE_SIZE) {
+		ret = afs_fill_page(vnode, key, index << PAGE_CACHE_SHIFT, page);
 		if (ret < 0) {
 			kfree(candidate);
 			_leave(" = %d [prep]", ret);
@@ -684,9 +681,10 @@ int afs_writeback_all(struct afs_vnode *vnode)
  * - the return status from this call provides a reliable indication of
  *   whether any write errors occurred for this process.
  */
-int afs_fsync(struct file *file, int datasync)
+int afs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct dentry *dentry = file->f_path.dentry;
+	struct inode *inode = file->f_mapping->host;
 	struct afs_writeback *wb, *xwb;
 	struct afs_vnode *vnode = AFS_FS_I(dentry->d_inode);
 	int ret;
@@ -695,12 +693,19 @@ int afs_fsync(struct file *file, int datasync)
 	       vnode->fid.vid, vnode->fid.vnode, dentry->d_name.name,
 	       datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	/* use a writeback record as a marker in the queue - when this reaches
 	 * the front of the queue, all the outstanding writes are either
 	 * completed or rejected */
 	wb = kzalloc(sizeof(*wb), GFP_KERNEL);
-	if (!wb)
-		return -ENOMEM;
+	if (!wb) {
+		ret = -ENOMEM;
+		goto out;
+	}
 	wb->vnode = vnode;
 	wb->first = 0;
 	wb->last = -1;
@@ -723,7 +728,7 @@ int afs_fsync(struct file *file, int datasync)
 	if (ret < 0) {
 		afs_put_writeback(wb);
 		_leave(" = %d [wb]", ret);
-		return ret;
+		goto out;
 	}
 
 	/* wait for the preceding writes to actually complete */
@@ -732,6 +737,8 @@ int afs_fsync(struct file *file, int datasync)
 				       vnode->writebacks.next == &wb->link);
 	afs_put_writeback(wb);
 	_leave(" = %d", ret);
+out:
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/attr.c b/fs/attr.c
index caf2aa521e2..538e27959d3 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -232,17 +232,11 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
 	if (error)
 		return error;
 
-	if (ia_valid & ATTR_SIZE)
-		down_write(&dentry->d_inode->i_alloc_sem);
-
 	if (inode->i_op->setattr)
 		error = inode->i_op->setattr(dentry, attr);
 	else
 		error = simple_setattr(dentry, attr);
 
-	if (ia_valid & ATTR_SIZE)
-		up_write(&dentry->d_inode->i_alloc_sem);
-
 	if (!error)
 		fsnotify_change(dentry, ia_valid);
 
diff --git a/fs/bad_inode.c b/fs/bad_inode.c
index 9ad2369d9e3..9205cf25f1c 100644
--- a/fs/bad_inode.c
+++ b/fs/bad_inode.c
@@ -87,7 +87,8 @@ static int bad_file_release(struct inode *inode, struct file *filp)
 	return -EIO;
 }
 
-static int bad_file_fsync(struct file *file, int datasync)
+static int bad_file_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync)
 {
 	return -EIO;
 }
@@ -229,11 +230,8 @@ static int bad_inode_readlink(struct dentry *dentry, char __user *buffer,
 	return -EIO;
 }
 
-static int bad_inode_permission(struct inode *inode, int mask, unsigned int flags)
+static int bad_inode_permission(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
 	return -EIO;
 }
 
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 303983fabfd..dd0fdfc56d3 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -668,8 +668,7 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 			 * mm->dumpable = 0 regardless of the interpreter's
 			 * permissions.
 			 */
-			if (file_permission(interpreter, MAY_READ) < 0)
-				bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+			would_dump(bprm, interpreter);
 
 			retval = kernel_read(interpreter, 0, bprm->buf,
 					     BINPRM_BUF_SIZE);
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index 63039ed9576..30745f459fa 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -245,8 +245,7 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm,
 			 * mm->dumpable = 0 regardless of the interpreter's
 			 * permissions.
 			 */
-			if (file_permission(interpreter, MAY_READ) < 0)
-				bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+			would_dump(bprm, interpreter);
 
 			retval = kernel_read(interpreter, 0, bprm->buf,
 					     BINPRM_BUF_SIZE);
@@ -1864,6 +1863,7 @@ cleanup:
 	kfree(psinfo);
 	kfree(notes);
 	kfree(fpu);
+	kfree(shdr4extnum);
 #ifdef ELF_CORE_COPY_XFPREGS
 	kfree(xfpu);
 #endif
diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c
index 1befe2ec818..ba1a1ae4a18 100644
--- a/fs/binfmt_misc.c
+++ b/fs/binfmt_misc.c
@@ -149,8 +149,7 @@ static int load_misc_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 
 		/* if the binary is not readable than enforce mm->dumpable=0
 		   regardless of the interpreter's permissions */
-		if (file_permission(bprm->file, MAY_READ))
-			bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+		would_dump(bprm, bprm->file);
 
 		allow_write_access(bprm->file);
 		bprm->file = NULL;
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 1f2b1997833..9fb0b15331d 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -355,25 +355,30 @@ static loff_t block_llseek(struct file *file, loff_t offset, int origin)
 	mutex_lock(&bd_inode->i_mutex);
 	size = i_size_read(bd_inode);
 
+	retval = -EINVAL;
 	switch (origin) {
-		case 2:
+		case SEEK_END:
 			offset += size;
 			break;
-		case 1:
+		case SEEK_CUR:
 			offset += file->f_pos;
+		case SEEK_SET:
+			break;
+		default:
+			goto out;
 	}
-	retval = -EINVAL;
 	if (offset >= 0 && offset <= size) {
 		if (offset != file->f_pos) {
 			file->f_pos = offset;
 		}
 		retval = offset;
 	}
+out:
 	mutex_unlock(&bd_inode->i_mutex);
 	return retval;
 }
 	
-int blkdev_fsync(struct file *filp, int datasync)
+int blkdev_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *bd_inode = filp->f_mapping->host;
 	struct block_device *bdev = I_BDEV(bd_inode);
@@ -384,14 +389,10 @@ int blkdev_fsync(struct file *filp, int datasync)
 	 * i_mutex and doing so causes performance issues with concurrent
 	 * O_SYNC writers to a block device.
 	 */
-	mutex_unlock(&bd_inode->i_mutex);
-
 	error = blkdev_issue_flush(bdev, GFP_KERNEL, NULL);
 	if (error == -EOPNOTSUPP)
 		error = 0;
 
-	mutex_lock(&bd_inode->i_mutex);
-
 	return error;
 }
 EXPORT_SYMBOL(blkdev_fsync);
@@ -762,7 +763,19 @@ static struct block_device *bd_start_claiming(struct block_device *bdev,
 	if (!disk)
 		return ERR_PTR(-ENXIO);
 
-	whole = bdget_disk(disk, 0);
+	/*
+	 * Normally, @bdev should equal what's returned from bdget_disk()
+	 * if partno is 0; however, some drivers (floppy) use multiple
+	 * bdev's for the same physical device and @bdev may be one of the
+	 * aliases.  Keep @bdev if partno is 0.  This means claimer
+	 * tracking is broken for those devices but it has always been that
+	 * way.
+	 */
+	if (partno)
+		whole = bdget_disk(disk, 0);
+	else
+		whole = bdgrab(bdev);
+
 	module_put(disk->fops->owner);
 	put_disk(disk);
 	if (!whole)
@@ -1272,8 +1285,8 @@ int blkdev_get(struct block_device *bdev, fmode_t mode, void *holder)
 		 * individual writeable reference is too fragile given the
 		 * way @mode is used in blkdev_get/put().
 		 */
-		if ((disk->flags & GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE) &&
-		    !res && (mode & FMODE_WRITE) && !bdev->bd_write_holder) {
+		if (!res && (mode & FMODE_WRITE) && !bdev->bd_write_holder &&
+		    (disk->flags & GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE)) {
 			bdev->bd_write_holder = true;
 			disk_block_events(disk);
 		}
diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c
index f66fc995973..9f62ab2a728 100644
--- a/fs/btrfs/acl.c
+++ b/fs/btrfs/acl.c
@@ -195,14 +195,13 @@ out:
 	return ret;
 }
 
-int btrfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int btrfs_check_acl(struct inode *inode, int mask)
 {
 	int error = -EAGAIN;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			error = -ECHILD;
-
 	} else {
 		struct posix_acl *acl;
 		acl = btrfs_get_acl(inode, ACL_TYPE_ACCESS);
diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index 93b1aa93201..52d7eca8c7b 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -121,9 +121,6 @@ struct btrfs_inode {
 	 */
 	u64 index_cnt;
 
-	/* the start of block group preferred for allocations. */
-	u64 block_group;
-
 	/* the fsync log has some corner cases that mean we have to check
 	 * directories to see if any unlinks have been done before
 	 * the directory was logged.  See tree-log.c for all the
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index b0e18d986e0..2e667868e0d 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -43,8 +43,6 @@ struct btrfs_path *btrfs_alloc_path(void)
 {
 	struct btrfs_path *path;
 	path = kmem_cache_zalloc(btrfs_path_cachep, GFP_NOFS);
-	if (path)
-		path->reada = 1;
 	return path;
 }
 
@@ -1224,11 +1222,13 @@ static void reada_for_search(struct btrfs_root *root,
 	u64 search;
 	u64 target;
 	u64 nread = 0;
+	u64 gen;
 	int direction = path->reada;
 	struct extent_buffer *eb;
 	u32 nr;
 	u32 blocksize;
 	u32 nscan = 0;
+	bool map = true;
 
 	if (level != 1)
 		return;
@@ -1250,7 +1250,19 @@ static void reada_for_search(struct btrfs_root *root,
 
 	nritems = btrfs_header_nritems(node);
 	nr = slot;
+	if (node->map_token || path->skip_locking)
+		map = false;
+
 	while (1) {
+		if (map && !node->map_token) {
+			unsigned long offset = btrfs_node_key_ptr_offset(nr);
+			map_private_extent_buffer(node, offset,
+						  sizeof(struct btrfs_key_ptr),
+						  &node->map_token,
+						  &node->kaddr,
+						  &node->map_start,
+						  &node->map_len, KM_USER1);
+		}
 		if (direction < 0) {
 			if (nr == 0)
 				break;
@@ -1268,14 +1280,23 @@ static void reada_for_search(struct btrfs_root *root,
 		search = btrfs_node_blockptr(node, nr);
 		if ((search <= target && target - search <= 65536) ||
 		    (search > target && search - target <= 65536)) {
-			readahead_tree_block(root, search, blocksize,
-				     btrfs_node_ptr_generation(node, nr));
+			gen = btrfs_node_ptr_generation(node, nr);
+			if (map && node->map_token) {
+				unmap_extent_buffer(node, node->map_token,
+						    KM_USER1);
+				node->map_token = NULL;
+			}
+			readahead_tree_block(root, search, blocksize, gen);
 			nread += blocksize;
 		}
 		nscan++;
 		if ((nread > 65536 || nscan > 32))
 			break;
 	}
+	if (map && node->map_token) {
+		unmap_extent_buffer(node, node->map_token, KM_USER1);
+		node->map_token = NULL;
+	}
 }
 
 /*
@@ -1648,9 +1669,6 @@ again:
 		}
 cow_done:
 		BUG_ON(!cow && ins_len);
-		if (level != btrfs_header_level(b))
-			WARN_ON(1);
-		level = btrfs_header_level(b);
 
 		p->nodes[level] = b;
 		if (!p->skip_locking)
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 6c093fa98f6..82be74efbb2 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -19,7 +19,6 @@
 #ifndef __BTRFS_CTREE__
 #define __BTRFS_CTREE__
 
-#include <linux/version.h>
 #include <linux/mm.h>
 #include <linux/highmem.h>
 #include <linux/fs.h>
@@ -930,7 +929,6 @@ struct btrfs_fs_info {
 	 * is required instead of the faster short fsync log commits
 	 */
 	u64 last_trans_log_full_commit;
-	u64 open_ioctl_trans;
 	unsigned long mount_opt:20;
 	unsigned long compress_type:4;
 	u64 max_inline;
@@ -947,7 +945,6 @@ struct btrfs_fs_info {
 	struct super_block *sb;
 	struct inode *btree_inode;
 	struct backing_dev_info bdi;
-	struct mutex trans_mutex;
 	struct mutex tree_log_mutex;
 	struct mutex transaction_kthread_mutex;
 	struct mutex cleaner_mutex;
@@ -968,6 +965,13 @@ struct btrfs_fs_info {
 	struct rw_semaphore subvol_sem;
 	struct srcu_struct subvol_srcu;
 
+	spinlock_t trans_lock;
+	/*
+	 * the reloc mutex goes with the trans lock, it is taken
+	 * during commit to protect us from the relocation code
+	 */
+	struct mutex reloc_mutex;
+
 	struct list_head trans_list;
 	struct list_head hashers;
 	struct list_head dead_roots;
@@ -980,6 +984,7 @@ struct btrfs_fs_info {
 	atomic_t async_submit_draining;
 	atomic_t nr_async_bios;
 	atomic_t async_delalloc_pages;
+	atomic_t open_ioctl_trans;
 
 	/*
 	 * this is used by the balancing code to wait for all the pending
@@ -1044,6 +1049,7 @@ struct btrfs_fs_info {
 	int closing;
 	int log_root_recovering;
 	int enospc_unlink;
+	int trans_no_join;
 
 	u64 total_pinned;
 
@@ -1065,7 +1071,6 @@ struct btrfs_fs_info {
 	struct reloc_control *reloc_ctl;
 
 	spinlock_t delalloc_lock;
-	spinlock_t new_trans_lock;
 	u64 delalloc_bytes;
 
 	/* data_alloc_cluster is only used in ssd mode */
@@ -1172,6 +1177,14 @@ struct btrfs_root {
 	u32 type;
 
 	u64 highest_objectid;
+
+	/* btrfs_record_root_in_trans is a multi-step process,
+	 * and it can race with the balancing code.   But the
+	 * race is very small, and only the first time the root
+	 * is added to each transaction.  So in_trans_setup
+	 * is used to tell us when more checks are required
+	 */
+	unsigned long in_trans_setup;
 	int ref_cows;
 	int track_dirty;
 	int in_radix;
@@ -1181,7 +1194,6 @@ struct btrfs_root {
 	struct btrfs_key defrag_max;
 	int defrag_running;
 	char *name;
-	int in_sysfs;
 
 	/* the dirty list is only used by non-reference counted roots */
 	struct list_head dirty_list;
@@ -1207,7 +1219,7 @@ struct btrfs_root {
 	 * right now this just gets used so that a root has its own devid
 	 * for stat.  It may be used for more later
 	 */
-	struct super_block anon_super;
+	dev_t anon_dev;
 };
 
 struct btrfs_ioctl_defrag_range_args {
@@ -1323,6 +1335,11 @@ struct btrfs_ioctl_defrag_range_args {
  */
 #define BTRFS_STRING_ITEM_KEY	253
 
+/*
+ * Flags for mount options.
+ *
+ * Note: don't forget to add new options to btrfs_show_options()
+ */
 #define BTRFS_MOUNT_NODATASUM		(1 << 0)
 #define BTRFS_MOUNT_NODATACOW		(1 << 1)
 #define BTRFS_MOUNT_NOBARRIER		(1 << 2)
@@ -1340,6 +1357,7 @@ struct btrfs_ioctl_defrag_range_args {
 #define BTRFS_MOUNT_USER_SUBVOL_RM_ALLOWED (1 << 14)
 #define BTRFS_MOUNT_ENOSPC_DEBUG	 (1 << 15)
 #define BTRFS_MOUNT_AUTO_DEFRAG		(1 << 16)
+#define BTRFS_MOUNT_INODE_MAP_CACHE	(1 << 17)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
@@ -2238,6 +2256,9 @@ int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv,
 void btrfs_block_rsv_release(struct btrfs_root *root,
 			     struct btrfs_block_rsv *block_rsv,
 			     u64 num_bytes);
+int btrfs_truncate_reserve_metadata(struct btrfs_trans_handle *trans,
+				    struct btrfs_root *root,
+				    struct btrfs_block_rsv *rsv);
 int btrfs_set_block_group_ro(struct btrfs_root *root,
 			     struct btrfs_block_group_cache *cache);
 int btrfs_set_block_group_rw(struct btrfs_root *root,
@@ -2350,6 +2371,15 @@ int btrfs_drop_subtree(struct btrfs_trans_handle *trans,
 			struct btrfs_root *root,
 			struct extent_buffer *node,
 			struct extent_buffer *parent);
+static inline int btrfs_fs_closing(struct btrfs_fs_info *fs_info)
+{
+	/*
+	 * Get synced with close_ctree()
+	 */
+	smp_mb();
+	return fs_info->closing;
+}
+
 /* root-item.c */
 int btrfs_find_root_ref(struct btrfs_root *tree_root,
 			struct btrfs_path *path,
@@ -2480,6 +2510,9 @@ int btrfs_csum_truncate(struct btrfs_trans_handle *trans,
 int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 			     struct list_head *list, int search_commit);
 /* inode.c */
+struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *page,
+					   size_t pg_offset, u64 start, u64 len,
+					   int create);
 
 /* RHEL and EL kernels have a patch that renames PG_checked to FsMisc */
 #if defined(ClearPageFsMisc) && !defined(ClearPageChecked)
@@ -2512,8 +2545,7 @@ int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end,
 int btrfs_writepages(struct address_space *mapping,
 		     struct writeback_control *wbc);
 int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *new_root,
-			     u64 new_dirid, u64 alloc_hint);
+			     struct btrfs_root *new_root, u64 new_dirid);
 int btrfs_merge_bio_hook(struct page *page, unsigned long offset,
 			 size_t size, struct bio *bio, unsigned long bio_flags);
 
@@ -2573,7 +2605,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
 			   struct inode *inode);
 int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info);
-int btrfs_sync_file(struct file *file, int datasync);
+int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync);
 int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 			    int skip_pinned);
 extern const struct file_operations btrfs_file_operations;
@@ -2613,7 +2645,7 @@ do {								\
 
 /* acl.c */
 #ifdef CONFIG_BTRFS_FS_POSIX_ACL
-int btrfs_check_acl(struct inode *inode, int mask, unsigned int flags);
+int btrfs_check_acl(struct inode *inode, int mask);
 #else
 #define btrfs_check_acl NULL
 #endif
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index 01e29503a54..98c68e658a9 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -82,19 +82,16 @@ static inline struct btrfs_delayed_root *btrfs_get_delayed_root(
 	return root->fs_info->delayed_root;
 }
 
-static struct btrfs_delayed_node *btrfs_get_or_create_delayed_node(
-							struct inode *inode)
+static struct btrfs_delayed_node *btrfs_get_delayed_node(struct inode *inode)
 {
-	struct btrfs_delayed_node *node;
 	struct btrfs_inode *btrfs_inode = BTRFS_I(inode);
 	struct btrfs_root *root = btrfs_inode->root;
 	u64 ino = btrfs_ino(inode);
-	int ret;
+	struct btrfs_delayed_node *node;
 
-again:
 	node = ACCESS_ONCE(btrfs_inode->delayed_node);
 	if (node) {
-		atomic_inc(&node->refs);	/* can be accessed */
+		atomic_inc(&node->refs);
 		return node;
 	}
 
@@ -102,8 +99,10 @@ again:
 	node = radix_tree_lookup(&root->delayed_nodes_tree, ino);
 	if (node) {
 		if (btrfs_inode->delayed_node) {
+			atomic_inc(&node->refs);	/* can be accessed */
+			BUG_ON(btrfs_inode->delayed_node != node);
 			spin_unlock(&root->inode_lock);
-			goto again;
+			return node;
 		}
 		btrfs_inode->delayed_node = node;
 		atomic_inc(&node->refs);	/* can be accessed */
@@ -113,6 +112,23 @@ again:
 	}
 	spin_unlock(&root->inode_lock);
 
+	return NULL;
+}
+
+static struct btrfs_delayed_node *btrfs_get_or_create_delayed_node(
+							struct inode *inode)
+{
+	struct btrfs_delayed_node *node;
+	struct btrfs_inode *btrfs_inode = BTRFS_I(inode);
+	struct btrfs_root *root = btrfs_inode->root;
+	u64 ino = btrfs_ino(inode);
+	int ret;
+
+again:
+	node = btrfs_get_delayed_node(inode);
+	if (node)
+		return node;
+
 	node = kmem_cache_alloc(delayed_node_cache, GFP_NOFS);
 	if (!node)
 		return ERR_PTR(-ENOMEM);
@@ -297,7 +313,6 @@ struct btrfs_delayed_item *btrfs_alloc_delayed_item(u32 data_len)
 		item->data_len = data_len;
 		item->ins_or_del = 0;
 		item->bytes_reserved = 0;
-		item->block_rsv = NULL;
 		item->delayed_node = NULL;
 		atomic_set(&item->refs, 1);
 	}
@@ -549,19 +564,6 @@ struct btrfs_delayed_item *__btrfs_next_delayed_item(
 	return next;
 }
 
-static inline struct btrfs_delayed_node *btrfs_get_delayed_node(
-							struct inode *inode)
-{
-	struct btrfs_inode *btrfs_inode = BTRFS_I(inode);
-	struct btrfs_delayed_node *delayed_node;
-
-	delayed_node = btrfs_inode->delayed_node;
-	if (delayed_node)
-		atomic_inc(&delayed_node->refs);
-
-	return delayed_node;
-}
-
 static inline struct btrfs_root *btrfs_get_fs_root(struct btrfs_root *root,
 						   u64 root_id)
 {
@@ -593,10 +595,8 @@ static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans,
 
 	num_bytes = btrfs_calc_trans_metadata_size(root, 1);
 	ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes);
-	if (!ret) {
+	if (!ret)
 		item->bytes_reserved = num_bytes;
-		item->block_rsv = dst_rsv;
-	}
 
 	return ret;
 }
@@ -604,10 +604,13 @@ static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans,
 static void btrfs_delayed_item_release_metadata(struct btrfs_root *root,
 						struct btrfs_delayed_item *item)
 {
+	struct btrfs_block_rsv *rsv;
+
 	if (!item->bytes_reserved)
 		return;
 
-	btrfs_block_rsv_release(root, item->block_rsv,
+	rsv = &root->fs_info->global_block_rsv;
+	btrfs_block_rsv_release(root, rsv,
 				item->bytes_reserved);
 }
 
@@ -678,6 +681,7 @@ static int btrfs_batch_insert_items(struct btrfs_trans_handle *trans,
 	INIT_LIST_HEAD(&head);
 
 	next = item;
+	nitems = 0;
 
 	/*
 	 * count the number of the continuous items that we can insert in batch
@@ -1013,6 +1017,7 @@ int btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
 	struct btrfs_delayed_root *delayed_root;
 	struct btrfs_delayed_node *curr_node, *prev_node;
 	struct btrfs_path *path;
+	struct btrfs_block_rsv *block_rsv;
 	int ret = 0;
 
 	path = btrfs_alloc_path();
@@ -1020,6 +1025,9 @@ int btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 	path->leave_spinning = 1;
 
+	block_rsv = trans->block_rsv;
+	trans->block_rsv = &root->fs_info->global_block_rsv;
+
 	delayed_root = btrfs_get_delayed_root(root);
 
 	curr_node = btrfs_first_delayed_node(delayed_root);
@@ -1044,6 +1052,7 @@ int btrfs_run_delayed_items(struct btrfs_trans_handle *trans,
 	}
 
 	btrfs_free_path(path);
+	trans->block_rsv = block_rsv;
 	return ret;
 }
 
@@ -1051,6 +1060,7 @@ static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
 					      struct btrfs_delayed_node *node)
 {
 	struct btrfs_path *path;
+	struct btrfs_block_rsv *block_rsv;
 	int ret;
 
 	path = btrfs_alloc_path();
@@ -1058,6 +1068,9 @@ static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 	path->leave_spinning = 1;
 
+	block_rsv = trans->block_rsv;
+	trans->block_rsv = &node->root->fs_info->global_block_rsv;
+
 	ret = btrfs_insert_delayed_items(trans, path, node->root, node);
 	if (!ret)
 		ret = btrfs_delete_delayed_items(trans, path, node->root, node);
@@ -1065,6 +1078,7 @@ static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans,
 		ret = btrfs_update_delayed_inode(trans, node->root, path, node);
 	btrfs_free_path(path);
 
+	trans->block_rsv = block_rsv;
 	return ret;
 }
 
@@ -1115,6 +1129,7 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work)
 	struct btrfs_path *path;
 	struct btrfs_delayed_node *delayed_node = NULL;
 	struct btrfs_root *root;
+	struct btrfs_block_rsv *block_rsv;
 	unsigned long nr = 0;
 	int need_requeue = 0;
 	int ret;
@@ -1129,10 +1144,13 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work)
 	delayed_node = async_node->delayed_node;
 	root = delayed_node->root;
 
-	trans = btrfs_join_transaction(root, 0);
+	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans))
 		goto free_path;
 
+	block_rsv = trans->block_rsv;
+	trans->block_rsv = &root->fs_info->global_block_rsv;
+
 	ret = btrfs_insert_delayed_items(trans, path, root, delayed_node);
 	if (!ret)
 		ret = btrfs_delete_delayed_items(trans, path, root,
@@ -1175,6 +1193,7 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work)
 
 	nr = trans->blocks_used;
 
+	trans->block_rsv = block_rsv;
 	btrfs_end_transaction_dmeta(trans, root);
 	__btrfs_btree_balance_dirty(root, nr);
 free_path:
@@ -1221,6 +1240,13 @@ again:
 	return 0;
 }
 
+void btrfs_assert_delayed_root_empty(struct btrfs_root *root)
+{
+	struct btrfs_delayed_root *delayed_root;
+	delayed_root = btrfs_get_delayed_root(root);
+	WARN_ON(btrfs_first_delayed_node(delayed_root));
+}
+
 void btrfs_balance_delayed_items(struct btrfs_root *root)
 {
 	struct btrfs_delayed_root *delayed_root;
@@ -1381,8 +1407,7 @@ end:
 
 int btrfs_inode_delayed_dir_index_count(struct inode *inode)
 {
-	struct btrfs_delayed_node *delayed_node = BTRFS_I(inode)->delayed_node;
-	int ret = 0;
+	struct btrfs_delayed_node *delayed_node = btrfs_get_delayed_node(inode);
 
 	if (!delayed_node)
 		return -ENOENT;
@@ -1392,11 +1417,14 @@ int btrfs_inode_delayed_dir_index_count(struct inode *inode)
 	 * a new directory index is added into the delayed node and index_cnt
 	 * is updated now. So we needn't lock the delayed node.
 	 */
-	if (!delayed_node->index_cnt)
+	if (!delayed_node->index_cnt) {
+		btrfs_release_delayed_node(delayed_node);
 		return -EINVAL;
+	}
 
 	BTRFS_I(inode)->index_cnt = delayed_node->index_cnt;
-	return ret;
+	btrfs_release_delayed_node(delayed_node);
+	return 0;
 }
 
 void btrfs_get_delayed_items(struct inode *inode, struct list_head *ins_list,
@@ -1572,8 +1600,7 @@ static void fill_stack_inode_item(struct btrfs_trans_handle *trans,
 	btrfs_set_stack_inode_transid(inode_item, trans->transid);
 	btrfs_set_stack_inode_rdev(inode_item, inode->i_rdev);
 	btrfs_set_stack_inode_flags(inode_item, BTRFS_I(inode)->flags);
-	btrfs_set_stack_inode_block_group(inode_item,
-					  BTRFS_I(inode)->block_group);
+	btrfs_set_stack_inode_block_group(inode_item, 0);
 
 	btrfs_set_stack_timespec_sec(btrfs_inode_atime(inode_item),
 				     inode->i_atime.tv_sec);
@@ -1591,11 +1618,62 @@ static void fill_stack_inode_item(struct btrfs_trans_handle *trans,
 				      inode->i_ctime.tv_nsec);
 }
 
+int btrfs_fill_inode(struct inode *inode, u32 *rdev)
+{
+	struct btrfs_delayed_node *delayed_node;
+	struct btrfs_inode_item *inode_item;
+	struct btrfs_timespec *tspec;
+
+	delayed_node = btrfs_get_delayed_node(inode);
+	if (!delayed_node)
+		return -ENOENT;
+
+	mutex_lock(&delayed_node->mutex);
+	if (!delayed_node->inode_dirty) {
+		mutex_unlock(&delayed_node->mutex);
+		btrfs_release_delayed_node(delayed_node);
+		return -ENOENT;
+	}
+
+	inode_item = &delayed_node->inode_item;
+
+	inode->i_uid = btrfs_stack_inode_uid(inode_item);
+	inode->i_gid = btrfs_stack_inode_gid(inode_item);
+	btrfs_i_size_write(inode, btrfs_stack_inode_size(inode_item));
+	inode->i_mode = btrfs_stack_inode_mode(inode_item);
+	inode->i_nlink = btrfs_stack_inode_nlink(inode_item);
+	inode_set_bytes(inode, btrfs_stack_inode_nbytes(inode_item));
+	BTRFS_I(inode)->generation = btrfs_stack_inode_generation(inode_item);
+	BTRFS_I(inode)->sequence = btrfs_stack_inode_sequence(inode_item);
+	inode->i_rdev = 0;
+	*rdev = btrfs_stack_inode_rdev(inode_item);
+	BTRFS_I(inode)->flags = btrfs_stack_inode_flags(inode_item);
+
+	tspec = btrfs_inode_atime(inode_item);
+	inode->i_atime.tv_sec = btrfs_stack_timespec_sec(tspec);
+	inode->i_atime.tv_nsec = btrfs_stack_timespec_nsec(tspec);
+
+	tspec = btrfs_inode_mtime(inode_item);
+	inode->i_mtime.tv_sec = btrfs_stack_timespec_sec(tspec);
+	inode->i_mtime.tv_nsec = btrfs_stack_timespec_nsec(tspec);
+
+	tspec = btrfs_inode_ctime(inode_item);
+	inode->i_ctime.tv_sec = btrfs_stack_timespec_sec(tspec);
+	inode->i_ctime.tv_nsec = btrfs_stack_timespec_nsec(tspec);
+
+	inode->i_generation = BTRFS_I(inode)->generation;
+	BTRFS_I(inode)->index_cnt = (u64)-1;
+
+	mutex_unlock(&delayed_node->mutex);
+	btrfs_release_delayed_node(delayed_node);
+	return 0;
+}
+
 int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root, struct inode *inode)
 {
 	struct btrfs_delayed_node *delayed_node;
-	int ret;
+	int ret = 0;
 
 	delayed_node = btrfs_get_or_create_delayed_node(inode);
 	if (IS_ERR(delayed_node))
diff --git a/fs/btrfs/delayed-inode.h b/fs/btrfs/delayed-inode.h
index eb7d240aa64..8d27af4bd8b 100644
--- a/fs/btrfs/delayed-inode.h
+++ b/fs/btrfs/delayed-inode.h
@@ -75,7 +75,6 @@ struct btrfs_delayed_item {
 	struct list_head tree_list;	/* used for batch insert/delete items */
 	struct list_head readdir_list;	/* used for readdir items */
 	u64 bytes_reserved;
-	struct btrfs_block_rsv *block_rsv;
 	struct btrfs_delayed_node *delayed_node;
 	atomic_t refs;
 	int ins_or_del;
@@ -120,6 +119,7 @@ void btrfs_kill_delayed_inode_items(struct inode *inode);
 
 int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root, struct inode *inode);
+int btrfs_fill_inode(struct inode *inode, u32 *rdev);
 
 /* Used for drop dead root */
 void btrfs_kill_all_delayed_nodes(struct btrfs_root *root);
@@ -138,4 +138,8 @@ int btrfs_readdir_delayed_dir_index(struct file *filp, void *dirent,
 /* for init */
 int __init btrfs_delayed_inode_init(void);
 void btrfs_delayed_inode_exit(void);
+
+/* for debugging */
+void btrfs_assert_delayed_root_empty(struct btrfs_root *root);
+
 #endif
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 98b6a71decb..b231ae13b26 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1044,7 +1044,6 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	root->last_trans = 0;
 	root->highest_objectid = 0;
 	root->name = NULL;
-	root->in_sysfs = 0;
 	root->inode_tree = RB_ROOT;
 	INIT_RADIX_TREE(&root->delayed_nodes_tree, GFP_ATOMIC);
 	root->block_rsv = NULL;
@@ -1078,12 +1077,7 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	init_completion(&root->kobj_unregister);
 	root->defrag_running = 0;
 	root->root_key.objectid = objectid;
-	root->anon_super.s_root = NULL;
-	root->anon_super.s_dev = 0;
-	INIT_LIST_HEAD(&root->anon_super.s_list);
-	INIT_LIST_HEAD(&root->anon_super.s_instances);
-	init_rwsem(&root->anon_super.s_umount);
-
+	root->anon_dev = 0;
 	return 0;
 }
 
@@ -1300,19 +1294,21 @@ again:
 		return root;
 
 	root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS);
-	if (!root->free_ino_ctl)
-		goto fail;
 	root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned),
 					GFP_NOFS);
-	if (!root->free_ino_pinned)
+	if (!root->free_ino_pinned || !root->free_ino_ctl) {
+		ret = -ENOMEM;
 		goto fail;
+	}
 
 	btrfs_init_free_ino_ctl(root);
 	mutex_init(&root->fs_commit_mutex);
 	spin_lock_init(&root->cache_lock);
 	init_waitqueue_head(&root->cache_wait);
 
-	set_anon_super(&root->anon_super, NULL);
+	ret = get_anon_bdev(&root->anon_dev);
+	if (ret)
+		goto fail;
 
 	if (btrfs_root_refs(&root->root_item) == 0) {
 		ret = -ENOENT;
@@ -1505,24 +1501,24 @@ static int transaction_kthread(void *arg)
 		vfs_check_frozen(root->fs_info->sb, SB_FREEZE_WRITE);
 		mutex_lock(&root->fs_info->transaction_kthread_mutex);
 
-		spin_lock(&root->fs_info->new_trans_lock);
+		spin_lock(&root->fs_info->trans_lock);
 		cur = root->fs_info->running_transaction;
 		if (!cur) {
-			spin_unlock(&root->fs_info->new_trans_lock);
+			spin_unlock(&root->fs_info->trans_lock);
 			goto sleep;
 		}
 
 		now = get_seconds();
 		if (!cur->blocked &&
 		    (now < cur->start_time || now - cur->start_time < 30)) {
-			spin_unlock(&root->fs_info->new_trans_lock);
+			spin_unlock(&root->fs_info->trans_lock);
 			delay = HZ * 5;
 			goto sleep;
 		}
 		transid = cur->transid;
-		spin_unlock(&root->fs_info->new_trans_lock);
+		spin_unlock(&root->fs_info->trans_lock);
 
-		trans = btrfs_join_transaction(root, 1);
+		trans = btrfs_join_transaction(root);
 		BUG_ON(IS_ERR(trans));
 		if (transid == trans->transid) {
 			ret = btrfs_commit_transaction(trans, root);
@@ -1613,11 +1609,12 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	INIT_LIST_HEAD(&fs_info->ordered_operations);
 	INIT_LIST_HEAD(&fs_info->caching_block_groups);
 	spin_lock_init(&fs_info->delalloc_lock);
-	spin_lock_init(&fs_info->new_trans_lock);
+	spin_lock_init(&fs_info->trans_lock);
 	spin_lock_init(&fs_info->ref_cache_lock);
 	spin_lock_init(&fs_info->fs_roots_radix_lock);
 	spin_lock_init(&fs_info->delayed_iput_lock);
 	spin_lock_init(&fs_info->defrag_inodes_lock);
+	mutex_init(&fs_info->reloc_mutex);
 
 	init_completion(&fs_info->kobj_unregister);
 	fs_info->tree_root = tree_root;
@@ -1645,6 +1642,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	fs_info->max_inline = 8192 * 1024;
 	fs_info->metadata_ratio = 0;
 	fs_info->defrag_inodes = RB_ROOT;
+	fs_info->trans_no_join = 0;
 
 	fs_info->thread_pool_size = min_t(unsigned long,
 					  num_online_cpus() + 2, 8);
@@ -1667,8 +1665,6 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	init_waitqueue_head(&fs_info->scrub_pause_wait);
 	init_rwsem(&fs_info->scrub_super_lock);
 	fs_info->scrub_workers_refcnt = 0;
-	btrfs_init_workers(&fs_info->scrub_workers, "scrub",
-			   fs_info->thread_pool_size, &fs_info->generic_worker);
 
 	sb->s_blocksize = 4096;
 	sb->s_blocksize_bits = blksize_bits(4096);
@@ -1709,7 +1705,6 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	fs_info->do_barriers = 1;
 
 
-	mutex_init(&fs_info->trans_mutex);
 	mutex_init(&fs_info->ordered_operations_mutex);
 	mutex_init(&fs_info->tree_log_mutex);
 	mutex_init(&fs_info->chunk_mutex);
@@ -2393,10 +2388,8 @@ static void free_fs_root(struct btrfs_root *root)
 {
 	iput(root->cache_inode);
 	WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
-	if (root->anon_super.s_dev) {
-		down_write(&root->anon_super.s_umount);
-		kill_anon_super(&root->anon_super);
-	}
+	if (root->anon_dev)
+		free_anon_bdev(root->anon_dev);
 	free_extent_buffer(root->node);
 	free_extent_buffer(root->commit_root);
 	kfree(root->free_ino_ctl);
@@ -2479,13 +2472,13 @@ int btrfs_commit_super(struct btrfs_root *root)
 	down_write(&root->fs_info->cleanup_work_sem);
 	up_write(&root->fs_info->cleanup_work_sem);
 
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 	ret = btrfs_commit_transaction(trans, root);
 	BUG_ON(ret);
 	/* run commit again to drop the original snapshot */
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 	btrfs_commit_transaction(trans, root);
@@ -2911,9 +2904,8 @@ static int btrfs_destroy_delalloc_inodes(struct btrfs_root *root)
 
 	INIT_LIST_HEAD(&splice);
 
-	list_splice_init(&root->fs_info->delalloc_inodes, &splice);
-
 	spin_lock(&root->fs_info->delalloc_lock);
+	list_splice_init(&root->fs_info->delalloc_inodes, &splice);
 
 	while (!list_empty(&splice)) {
 		btrfs_inode = list_entry(splice.next, struct btrfs_inode,
@@ -3024,10 +3016,13 @@ static int btrfs_cleanup_transaction(struct btrfs_root *root)
 
 	WARN_ON(1);
 
-	mutex_lock(&root->fs_info->trans_mutex);
 	mutex_lock(&root->fs_info->transaction_kthread_mutex);
 
+	spin_lock(&root->fs_info->trans_lock);
 	list_splice_init(&root->fs_info->trans_list, &list);
+	root->fs_info->trans_no_join = 1;
+	spin_unlock(&root->fs_info->trans_lock);
+
 	while (!list_empty(&list)) {
 		t = list_entry(list.next, struct btrfs_transaction, list);
 		if (!t)
@@ -3052,23 +3047,18 @@ static int btrfs_cleanup_transaction(struct btrfs_root *root)
 		t->blocked = 0;
 		if (waitqueue_active(&root->fs_info->transaction_wait))
 			wake_up(&root->fs_info->transaction_wait);
-		mutex_unlock(&root->fs_info->trans_mutex);
 
-		mutex_lock(&root->fs_info->trans_mutex);
 		t->commit_done = 1;
 		if (waitqueue_active(&t->commit_wait))
 			wake_up(&t->commit_wait);
-		mutex_unlock(&root->fs_info->trans_mutex);
-
-		mutex_lock(&root->fs_info->trans_mutex);
 
 		btrfs_destroy_pending_snapshots(t);
 
 		btrfs_destroy_delalloc_inodes(root);
 
-		spin_lock(&root->fs_info->new_trans_lock);
+		spin_lock(&root->fs_info->trans_lock);
 		root->fs_info->running_transaction = NULL;
-		spin_unlock(&root->fs_info->new_trans_lock);
+		spin_unlock(&root->fs_info->trans_lock);
 
 		btrfs_destroy_marked_extents(root, &t->dirty_pages,
 					     EXTENT_DIRTY);
@@ -3082,8 +3072,10 @@ static int btrfs_cleanup_transaction(struct btrfs_root *root)
 		kmem_cache_free(btrfs_transaction_cachep, t);
 	}
 
+	spin_lock(&root->fs_info->trans_lock);
+	root->fs_info->trans_no_join = 0;
+	spin_unlock(&root->fs_info->trans_lock);
 	mutex_unlock(&root->fs_info->transaction_kthread_mutex);
-	mutex_unlock(&root->fs_info->trans_mutex);
 
 	return 0;
 }
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 169bd62ce77..71cd456fdb6 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -348,7 +348,7 @@ static int caching_kthread(void *data)
 	 */
 	path->skip_locking = 1;
 	path->search_commit_root = 1;
-	path->reada = 2;
+	path->reada = 1;
 
 	key.objectid = last;
 	key.offset = 0;
@@ -366,8 +366,7 @@ again:
 	nritems = btrfs_header_nritems(leaf);
 
 	while (1) {
-		smp_mb();
-		if (fs_info->closing > 1) {
+		if (btrfs_fs_closing(fs_info) > 1) {
 			last = (u64)-1;
 			break;
 		}
@@ -379,15 +378,18 @@ again:
 			if (ret)
 				break;
 
-			caching_ctl->progress = last;
-			btrfs_release_path(path);
-			up_read(&fs_info->extent_commit_sem);
-			mutex_unlock(&caching_ctl->mutex);
-			if (btrfs_transaction_in_commit(fs_info))
-				schedule_timeout(1);
-			else
+			if (need_resched() ||
+			    btrfs_next_leaf(extent_root, path)) {
+				caching_ctl->progress = last;
+				btrfs_release_path(path);
+				up_read(&fs_info->extent_commit_sem);
+				mutex_unlock(&caching_ctl->mutex);
 				cond_resched();
-			goto again;
+				goto again;
+			}
+			leaf = path->nodes[0];
+			nritems = btrfs_header_nritems(leaf);
+			continue;
 		}
 
 		if (key.objectid < block_group->key.objectid) {
@@ -3065,7 +3067,7 @@ again:
 			spin_unlock(&data_sinfo->lock);
 alloc:
 			alloc_target = btrfs_get_alloc_profile(root, 1);
-			trans = btrfs_join_transaction(root, 1);
+			trans = btrfs_join_transaction(root);
 			if (IS_ERR(trans))
 				return PTR_ERR(trans);
 
@@ -3087,13 +3089,21 @@ alloc:
 			}
 			goto again;
 		}
+
+		/*
+		 * If we have less pinned bytes than we want to allocate then
+		 * don't bother committing the transaction, it won't help us.
+		 */
+		if (data_sinfo->bytes_pinned < bytes)
+			committed = 1;
 		spin_unlock(&data_sinfo->lock);
 
 		/* commit the current transaction and try again */
 commit_trans:
-		if (!committed && !root->fs_info->open_ioctl_trans) {
+		if (!committed &&
+		    !atomic_read(&root->fs_info->open_ioctl_trans)) {
 			committed = 1;
-			trans = btrfs_join_transaction(root, 1);
+			trans = btrfs_join_transaction(root);
 			if (IS_ERR(trans))
 				return PTR_ERR(trans);
 			ret = btrfs_commit_transaction(trans, root);
@@ -3304,10 +3314,6 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans,
 	if (reserved == 0)
 		return 0;
 
-	/* nothing to shrink - nothing to reclaim */
-	if (root->fs_info->delalloc_bytes == 0)
-		return 0;
-
 	max_reclaim = min(reserved, to_reclaim);
 
 	while (loops < 1024) {
@@ -3472,7 +3478,7 @@ again:
 		goto out;
 
 	ret = -ENOSPC;
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans))
 		goto out;
 	ret = btrfs_commit_transaction(trans, root);
@@ -3699,7 +3705,7 @@ int btrfs_block_rsv_check(struct btrfs_trans_handle *trans,
 		if (trans)
 			return -EAGAIN;
 
-		trans = btrfs_join_transaction(root, 1);
+		trans = btrfs_join_transaction(root);
 		BUG_ON(IS_ERR(trans));
 		ret = btrfs_commit_transaction(trans, root);
 		return 0;
@@ -3837,6 +3843,37 @@ static void release_global_block_rsv(struct btrfs_fs_info *fs_info)
 	WARN_ON(fs_info->chunk_block_rsv.reserved > 0);
 }
 
+int btrfs_truncate_reserve_metadata(struct btrfs_trans_handle *trans,
+				    struct btrfs_root *root,
+				    struct btrfs_block_rsv *rsv)
+{
+	struct btrfs_block_rsv *trans_rsv = &root->fs_info->trans_block_rsv;
+	u64 num_bytes;
+	int ret;
+
+	/*
+	 * Truncate should be freeing data, but give us 2 items just in case it
+	 * needs to use some space.  We may want to be smarter about this in the
+	 * future.
+	 */
+	num_bytes = btrfs_calc_trans_metadata_size(root, 2);
+
+	/* We already have enough bytes, just return */
+	if (rsv->reserved >= num_bytes)
+		return 0;
+
+	num_bytes -= rsv->reserved;
+
+	/*
+	 * You should have reserved enough space before hand to do this, so this
+	 * should not fail.
+	 */
+	ret = block_rsv_migrate_bytes(trans_rsv, rsv, num_bytes);
+	BUG_ON(ret);
+
+	return 0;
+}
+
 int btrfs_trans_reserve_metadata(struct btrfs_trans_handle *trans,
 				 struct btrfs_root *root,
 				 int num_items)
@@ -3877,23 +3914,18 @@ int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans,
 	struct btrfs_block_rsv *dst_rsv = root->orphan_block_rsv;
 
 	/*
-	 * one for deleting orphan item, one for updating inode and
-	 * two for calling btrfs_truncate_inode_items.
-	 *
-	 * btrfs_truncate_inode_items is a delete operation, it frees
-	 * more space than it uses in most cases. So two units of
-	 * metadata space should be enough for calling it many times.
-	 * If all of the metadata space is used, we can commit
-	 * transaction and use space it freed.
+	 * We need to hold space in order to delete our orphan item once we've
+	 * added it, so this takes the reservation so we can release it later
+	 * when we are truly done with the orphan item.
 	 */
-	u64 num_bytes = btrfs_calc_trans_metadata_size(root, 4);
+	u64 num_bytes = btrfs_calc_trans_metadata_size(root, 1);
 	return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes);
 }
 
 void btrfs_orphan_release_metadata(struct inode *inode)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	u64 num_bytes = btrfs_calc_trans_metadata_size(root, 4);
+	u64 num_bytes = btrfs_calc_trans_metadata_size(root, 1);
 	btrfs_block_rsv_release(root, root->orphan_block_rsv, num_bytes);
 }
 
@@ -4810,7 +4842,7 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
 				     u64 num_bytes, u64 empty_size,
 				     u64 search_start, u64 search_end,
 				     u64 hint_byte, struct btrfs_key *ins,
-				     int data)
+				     u64 data)
 {
 	int ret = 0;
 	struct btrfs_root *root = orig_root->fs_info->extent_root;
@@ -4837,7 +4869,7 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
 
 	space_info = __find_space_info(root->fs_info, data);
 	if (!space_info) {
-		printk(KERN_ERR "No space info for %d\n", data);
+		printk(KERN_ERR "No space info for %llu\n", data);
 		return -ENOSPC;
 	}
 
@@ -4987,6 +5019,15 @@ have_block_group:
 		if (unlikely(block_group->ro))
 			goto loop;
 
+		spin_lock(&block_group->free_space_ctl->tree_lock);
+		if (cached &&
+		    block_group->free_space_ctl->free_space <
+		    num_bytes + empty_size) {
+			spin_unlock(&block_group->free_space_ctl->tree_lock);
+			goto loop;
+		}
+		spin_unlock(&block_group->free_space_ctl->tree_lock);
+
 		/*
 		 * Ok we want to try and use the cluster allocator, so lets look
 		 * there, unless we are on LOOP_NO_EMPTY_SIZE, since we will
@@ -5150,6 +5191,7 @@ checks:
 			btrfs_add_free_space(block_group, offset,
 					     search_start - offset);
 		BUG_ON(offset > search_start);
+		btrfs_put_block_group(block_group);
 		break;
 loop:
 		failed_cluster_refill = false;
@@ -5172,9 +5214,7 @@ loop:
 	 * LOOP_NO_EMPTY_SIZE, set empty_size and empty_cluster to 0 and try
 	 *			again
 	 */
-	if (!ins->objectid && loop < LOOP_NO_EMPTY_SIZE &&
-	    (found_uncached_bg || empty_size || empty_cluster ||
-	     allowed_chunk_alloc)) {
+	if (!ins->objectid && loop < LOOP_NO_EMPTY_SIZE) {
 		index = 0;
 		if (loop == LOOP_FIND_IDEAL && found_uncached_bg) {
 			found_uncached_bg = false;
@@ -5214,42 +5254,39 @@ loop:
 			goto search;
 		}
 
-		if (loop < LOOP_CACHING_WAIT) {
-			loop++;
-			goto search;
-		}
+		loop++;
 
 		if (loop == LOOP_ALLOC_CHUNK) {
-			empty_size = 0;
-			empty_cluster = 0;
-		}
+		       if (allowed_chunk_alloc) {
+				ret = do_chunk_alloc(trans, root, num_bytes +
+						     2 * 1024 * 1024, data,
+						     CHUNK_ALLOC_LIMITED);
+				allowed_chunk_alloc = 0;
+				if (ret == 1)
+					done_chunk_alloc = 1;
+			} else if (!done_chunk_alloc &&
+				   space_info->force_alloc ==
+				   CHUNK_ALLOC_NO_FORCE) {
+				space_info->force_alloc = CHUNK_ALLOC_LIMITED;
+			}
 
-		if (allowed_chunk_alloc) {
-			ret = do_chunk_alloc(trans, root, num_bytes +
-					     2 * 1024 * 1024, data,
-					     CHUNK_ALLOC_LIMITED);
-			allowed_chunk_alloc = 0;
-			done_chunk_alloc = 1;
-		} else if (!done_chunk_alloc &&
-			   space_info->force_alloc == CHUNK_ALLOC_NO_FORCE) {
-			space_info->force_alloc = CHUNK_ALLOC_LIMITED;
+		       /*
+			* We didn't allocate a chunk, go ahead and drop the
+			* empty size and loop again.
+			*/
+		       if (!done_chunk_alloc)
+			       loop = LOOP_NO_EMPTY_SIZE;
 		}
 
-		if (loop < LOOP_NO_EMPTY_SIZE) {
-			loop++;
-			goto search;
+		if (loop == LOOP_NO_EMPTY_SIZE) {
+			empty_size = 0;
+			empty_cluster = 0;
 		}
-		ret = -ENOSPC;
+
+		goto search;
 	} else if (!ins->objectid) {
 		ret = -ENOSPC;
-	}
-
-	/* we found what we needed */
-	if (ins->objectid) {
-		if (!(data & BTRFS_BLOCK_GROUP_DATA))
-			trans->block_group = block_group->key.objectid;
-
-		btrfs_put_block_group(block_group);
+	} else if (ins->objectid) {
 		ret = 0;
 	}
 
@@ -6526,7 +6563,7 @@ int btrfs_set_block_group_ro(struct btrfs_root *root,
 
 	BUG_ON(cache->ro);
 
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	BUG_ON(IS_ERR(trans));
 
 	alloc_flags = update_block_group_flags(root, cache->flags);
@@ -6882,6 +6919,7 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+	path->reada = 1;
 
 	cache_gen = btrfs_super_cache_generation(&root->fs_info->super_copy);
 	if (cache_gen != 0 &&
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index c5d9fbb92bc..7055d11c1ef 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -1476,7 +1476,7 @@ u64 count_range_bits(struct extent_io_tree *tree,
 			if (total_bytes >= max_bytes)
 				break;
 			if (!found) {
-				*start = state->start;
+				*start = max(cur_start, state->start);
 				found = 1;
 			}
 			last = state->end;
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index 4e8445a4757..a11a92ee2d3 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -126,9 +126,9 @@ struct extent_buffer {
 	unsigned long map_len;
 	struct page *first_page;
 	unsigned long bflags;
-	atomic_t refs;
 	struct list_head leak_list;
 	struct rcu_head rcu_head;
+	atomic_t refs;
 
 	/* the spinlock is used to protect most operations */
 	spinlock_t lock;
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index c6a22d783c3..59cbdb120ad 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -129,7 +129,7 @@ int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
 	if (!btrfs_test_opt(root, AUTO_DEFRAG))
 		return 0;
 
-	if (root->fs_info->closing)
+	if (btrfs_fs_closing(root->fs_info))
 		return 0;
 
 	if (BTRFS_I(inode)->in_defrag)
@@ -144,7 +144,7 @@ int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
 	if (!defrag)
 		return -ENOMEM;
 
-	defrag->ino = inode->i_ino;
+	defrag->ino = btrfs_ino(inode);
 	defrag->transid = transid;
 	defrag->root = root->root_key.objectid;
 
@@ -229,7 +229,7 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
 		first_ino = defrag->ino + 1;
 		rb_erase(&defrag->rb_node, &fs_info->defrag_inodes);
 
-		if (fs_info->closing)
+		if (btrfs_fs_closing(fs_info))
 			goto next_free;
 
 		spin_unlock(&fs_info->defrag_inodes_lock);
@@ -1452,7 +1452,7 @@ int btrfs_release_file(struct inode *inode, struct file *filp)
  * important optimization for directories because holding the mutex prevents
  * new operations on the dir while we write to disk.
  */
-int btrfs_sync_file(struct file *file, int datasync)
+int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct dentry *dentry = file->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
@@ -1462,9 +1462,13 @@ int btrfs_sync_file(struct file *file, int datasync)
 
 	trace_btrfs_sync_file(file, datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	/* we wait first, since the writeback may change the inode */
 	root->log_batch++;
-	/* the VFS called filemap_fdatawrite for us */
 	btrfs_wait_ordered_range(inode, 0, (u64)-1);
 	root->log_batch++;
 
@@ -1472,22 +1476,23 @@ int btrfs_sync_file(struct file *file, int datasync)
 	 * check the transaction that last modified this inode
 	 * and see if its already been committed
 	 */
-	if (!BTRFS_I(inode)->last_trans)
+	if (!BTRFS_I(inode)->last_trans) {
+		mutex_unlock(&inode->i_mutex);
 		goto out;
+	}
 
 	/*
 	 * if the last transaction that changed this file was before
 	 * the current transaction, we can bail out now without any
 	 * syncing
 	 */
-	mutex_lock(&root->fs_info->trans_mutex);
+	smp_mb();
 	if (BTRFS_I(inode)->last_trans <=
 	    root->fs_info->last_trans_committed) {
 		BTRFS_I(inode)->last_trans = 0;
-		mutex_unlock(&root->fs_info->trans_mutex);
+		mutex_unlock(&inode->i_mutex);
 		goto out;
 	}
-	mutex_unlock(&root->fs_info->trans_mutex);
 
 	/*
 	 * ok we haven't committed the transaction yet, lets do a commit
@@ -1498,12 +1503,15 @@ int btrfs_sync_file(struct file *file, int datasync)
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
+		mutex_unlock(&inode->i_mutex);
 		goto out;
 	}
 
 	ret = btrfs_log_dentry_safe(trans, root, dentry);
-	if (ret < 0)
+	if (ret < 0) {
+		mutex_unlock(&inode->i_mutex);
 		goto out;
+	}
 
 	/* we've logged all the items and now have a consistent
 	 * version of the file in the log.  It is possible that
@@ -1515,7 +1523,7 @@ int btrfs_sync_file(struct file *file, int datasync)
 	 * file again, but that will end up using the synchronization
 	 * inside btrfs_sync_log to keep things safe.
 	 */
-	mutex_unlock(&dentry->d_inode->i_mutex);
+	mutex_unlock(&inode->i_mutex);
 
 	if (ret != BTRFS_NO_LOG_SYNC) {
 		if (ret > 0) {
@@ -1530,7 +1538,6 @@ int btrfs_sync_file(struct file *file, int datasync)
 	} else {
 		ret = btrfs_end_transaction(trans, root);
 	}
-	mutex_lock(&dentry->d_inode->i_mutex);
 out:
 	return ret > 0 ? -EIO : ret;
 }
@@ -1666,8 +1673,154 @@ out:
 	return ret;
 }
 
+static int find_desired_extent(struct inode *inode, loff_t *offset, int origin)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct extent_map *em;
+	struct extent_state *cached_state = NULL;
+	u64 lockstart = *offset;
+	u64 lockend = i_size_read(inode);
+	u64 start = *offset;
+	u64 orig_start = *offset;
+	u64 len = i_size_read(inode);
+	u64 last_end = 0;
+	int ret = 0;
+
+	lockend = max_t(u64, root->sectorsize, lockend);
+	if (lockend <= lockstart)
+		lockend = lockstart + root->sectorsize;
+
+	len = lockend - lockstart + 1;
+
+	len = max_t(u64, len, root->sectorsize);
+	if (inode->i_size == 0)
+		return -ENXIO;
+
+	lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend, 0,
+			 &cached_state, GFP_NOFS);
+
+	/*
+	 * Delalloc is such a pain.  If we have a hole and we have pending
+	 * delalloc for a portion of the hole we will get back a hole that
+	 * exists for the entire range since it hasn't been actually written
+	 * yet.  So to take care of this case we need to look for an extent just
+	 * before the position we want in case there is outstanding delalloc
+	 * going on here.
+	 */
+	if (origin == SEEK_HOLE && start != 0) {
+		if (start <= root->sectorsize)
+			em = btrfs_get_extent_fiemap(inode, NULL, 0, 0,
+						     root->sectorsize, 0);
+		else
+			em = btrfs_get_extent_fiemap(inode, NULL, 0,
+						     start - root->sectorsize,
+						     root->sectorsize, 0);
+		if (IS_ERR(em)) {
+			ret = -ENXIO;
+			goto out;
+		}
+		last_end = em->start + em->len;
+		if (em->block_start == EXTENT_MAP_DELALLOC)
+			last_end = min_t(u64, last_end, inode->i_size);
+		free_extent_map(em);
+	}
+
+	while (1) {
+		em = btrfs_get_extent_fiemap(inode, NULL, 0, start, len, 0);
+		if (IS_ERR(em)) {
+			ret = -ENXIO;
+			break;
+		}
+
+		if (em->block_start == EXTENT_MAP_HOLE) {
+			if (test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
+				if (last_end <= orig_start) {
+					free_extent_map(em);
+					ret = -ENXIO;
+					break;
+				}
+			}
+
+			if (origin == SEEK_HOLE) {
+				*offset = start;
+				free_extent_map(em);
+				break;
+			}
+		} else {
+			if (origin == SEEK_DATA) {
+				if (em->block_start == EXTENT_MAP_DELALLOC) {
+					if (start >= inode->i_size) {
+						free_extent_map(em);
+						ret = -ENXIO;
+						break;
+					}
+				}
+
+				*offset = start;
+				free_extent_map(em);
+				break;
+			}
+		}
+
+		start = em->start + em->len;
+		last_end = em->start + em->len;
+
+		if (em->block_start == EXTENT_MAP_DELALLOC)
+			last_end = min_t(u64, last_end, inode->i_size);
+
+		if (test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
+			free_extent_map(em);
+			ret = -ENXIO;
+			break;
+		}
+		free_extent_map(em);
+		cond_resched();
+	}
+	if (!ret)
+		*offset = min(*offset, inode->i_size);
+out:
+	unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend,
+			     &cached_state, GFP_NOFS);
+	return ret;
+}
+
+static loff_t btrfs_file_llseek(struct file *file, loff_t offset, int origin)
+{
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	mutex_lock(&inode->i_mutex);
+	switch (origin) {
+	case SEEK_END:
+	case SEEK_CUR:
+		offset = generic_file_llseek_unlocked(file, offset, origin);
+		goto out;
+	case SEEK_DATA:
+	case SEEK_HOLE:
+		ret = find_desired_extent(inode, &offset, origin);
+		if (ret) {
+			mutex_unlock(&inode->i_mutex);
+			return ret;
+		}
+	}
+
+	if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET))
+		return -EINVAL;
+	if (offset > inode->i_sb->s_maxbytes)
+		return -EINVAL;
+
+	/* Special lock needed here? */
+	if (offset != file->f_pos) {
+		file->f_pos = offset;
+		file->f_version = 0;
+	}
+out:
+	mutex_unlock(&inode->i_mutex);
+	return offset;
+}
+
 const struct file_operations btrfs_file_operations = {
-	.llseek		= generic_file_llseek,
+	.llseek		= btrfs_file_llseek,
 	.read		= do_sync_read,
 	.write		= do_sync_write,
 	.aio_read       = generic_file_aio_read,
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 70d45795d75..bf0d61567f3 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -98,7 +98,7 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root,
 		return inode;
 
 	spin_lock(&block_group->lock);
-	if (!root->fs_info->closing) {
+	if (!btrfs_fs_closing(root->fs_info)) {
 		block_group->inode = igrab(inode);
 		block_group->iref = 1;
 	}
@@ -250,7 +250,7 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 	pgoff_t index = 0;
 	unsigned long first_page_offset;
 	int num_checksums;
-	int ret = 0, ret2;
+	int ret = 0;
 
 	INIT_LIST_HEAD(&bitmaps);
 
@@ -402,7 +402,14 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 				spin_lock(&ctl->tree_lock);
 				ret = link_free_space(ctl, e);
 				spin_unlock(&ctl->tree_lock);
-				BUG_ON(ret);
+				if (ret) {
+					printk(KERN_ERR "Duplicate entries in "
+					       "free space cache, dumping\n");
+					kunmap(page);
+					unlock_page(page);
+					page_cache_release(page);
+					goto free_cache;
+				}
 			} else {
 				e->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS);
 				if (!e->bitmap) {
@@ -414,10 +421,18 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 					goto free_cache;
 				}
 				spin_lock(&ctl->tree_lock);
-				ret2 = link_free_space(ctl, e);
+				ret = link_free_space(ctl, e);
 				ctl->total_bitmaps++;
 				ctl->op->recalc_thresholds(ctl);
 				spin_unlock(&ctl->tree_lock);
+				if (ret) {
+					printk(KERN_ERR "Duplicate entries in "
+					       "free space cache, dumping\n");
+					kunmap(page);
+					unlock_page(page);
+					page_cache_release(page);
+					goto free_cache;
+				}
 				list_add_tail(&e->list, &bitmaps);
 			}
 
@@ -478,8 +493,7 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info,
 	 * If we're unmounting then just return, since this does a search on the
 	 * normal root and not the commit root and we could deadlock.
 	 */
-	smp_mb();
-	if (fs_info->closing)
+	if (btrfs_fs_closing(fs_info))
 		return 0;
 
 	/*
@@ -575,10 +589,25 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
 
 	num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >>
 		PAGE_CACHE_SHIFT;
+
+	/* Since the first page has all of our checksums and our generation we
+	 * need to calculate the offset into the page that we can start writing
+	 * our entries.
+	 */
+	first_page_offset = (sizeof(u32) * num_pages) + sizeof(u64);
+
 	filemap_write_and_wait(inode->i_mapping);
 	btrfs_wait_ordered_range(inode, inode->i_size &
 				 ~(root->sectorsize - 1), (u64)-1);
 
+	/* make sure we don't overflow that first page */
+	if (first_page_offset + sizeof(struct btrfs_free_space_entry) >= PAGE_CACHE_SIZE) {
+		/* this is really the same as running out of space, where we also return 0 */
+		printk(KERN_CRIT "Btrfs: free space cache was too big for the crc page\n");
+		ret = 0;
+		goto out_update;
+	}
+
 	/* We need a checksum per page. */
 	crc = checksums = kzalloc(sizeof(u32) * num_pages, GFP_NOFS);
 	if (!crc)
@@ -590,12 +619,6 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
 		return -1;
 	}
 
-	/* Since the first page has all of our checksums and our generation we
-	 * need to calculate the offset into the page that we can start writing
-	 * our entries.
-	 */
-	first_page_offset = (sizeof(u32) * num_pages) + sizeof(u64);
-
 	/* Get the cluster for this block_group if it exists */
 	if (block_group && !list_empty(&block_group->cluster_list))
 		cluster = list_entry(block_group->cluster_list.next,
@@ -857,12 +880,14 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
 	ret = 1;
 
 out_free:
+	kfree(checksums);
+	kfree(pages);
+
+out_update:
 	if (ret != 1) {
 		invalidate_inode_pages2_range(inode->i_mapping, 0, index);
 		BTRFS_I(inode)->generation = 0;
 	}
-	kfree(checksums);
-	kfree(pages);
 	btrfs_update_inode(trans, root, inode);
 	return ret;
 }
@@ -963,10 +988,16 @@ static int tree_insert_offset(struct rb_root *root, u64 offset,
 			 * logically.
 			 */
 			if (bitmap) {
-				WARN_ON(info->bitmap);
+				if (info->bitmap) {
+					WARN_ON_ONCE(1);
+					return -EEXIST;
+				}
 				p = &(*p)->rb_right;
 			} else {
-				WARN_ON(!info->bitmap);
+				if (!info->bitmap) {
+					WARN_ON_ONCE(1);
+					return -EEXIST;
+				}
 				p = &(*p)->rb_left;
 			}
 		}
@@ -1386,6 +1417,23 @@ again:
 	return 0;
 }
 
+static u64 add_bytes_to_bitmap(struct btrfs_free_space_ctl *ctl,
+			       struct btrfs_free_space *info, u64 offset,
+			       u64 bytes)
+{
+	u64 bytes_to_set = 0;
+	u64 end;
+
+	end = info->offset + (u64)(BITS_PER_BITMAP * ctl->unit);
+
+	bytes_to_set = min(end - offset, bytes);
+
+	bitmap_set_bits(ctl, info, offset, bytes_to_set);
+
+	return bytes_to_set;
+
+}
+
 static bool use_bitmap(struct btrfs_free_space_ctl *ctl,
 		      struct btrfs_free_space *info)
 {
@@ -1422,12 +1470,18 @@ static bool use_bitmap(struct btrfs_free_space_ctl *ctl,
 	return true;
 }
 
+static struct btrfs_free_space_op free_space_op = {
+	.recalc_thresholds	= recalculate_thresholds,
+	.use_bitmap		= use_bitmap,
+};
+
 static int insert_into_bitmap(struct btrfs_free_space_ctl *ctl,
 			      struct btrfs_free_space *info)
 {
 	struct btrfs_free_space *bitmap_info;
+	struct btrfs_block_group_cache *block_group = NULL;
 	int added = 0;
-	u64 bytes, offset, end;
+	u64 bytes, offset, bytes_added;
 	int ret;
 
 	bytes = info->bytes;
@@ -1436,7 +1490,49 @@ static int insert_into_bitmap(struct btrfs_free_space_ctl *ctl,
 	if (!ctl->op->use_bitmap(ctl, info))
 		return 0;
 
+	if (ctl->op == &free_space_op)
+		block_group = ctl->private;
 again:
+	/*
+	 * Since we link bitmaps right into the cluster we need to see if we
+	 * have a cluster here, and if so and it has our bitmap we need to add
+	 * the free space to that bitmap.
+	 */
+	if (block_group && !list_empty(&block_group->cluster_list)) {
+		struct btrfs_free_cluster *cluster;
+		struct rb_node *node;
+		struct btrfs_free_space *entry;
+
+		cluster = list_entry(block_group->cluster_list.next,
+				     struct btrfs_free_cluster,
+				     block_group_list);
+		spin_lock(&cluster->lock);
+		node = rb_first(&cluster->root);
+		if (!node) {
+			spin_unlock(&cluster->lock);
+			goto no_cluster_bitmap;
+		}
+
+		entry = rb_entry(node, struct btrfs_free_space, offset_index);
+		if (!entry->bitmap) {
+			spin_unlock(&cluster->lock);
+			goto no_cluster_bitmap;
+		}
+
+		if (entry->offset == offset_to_bitmap(ctl, offset)) {
+			bytes_added = add_bytes_to_bitmap(ctl, entry,
+							  offset, bytes);
+			bytes -= bytes_added;
+			offset += bytes_added;
+		}
+		spin_unlock(&cluster->lock);
+		if (!bytes) {
+			ret = 1;
+			goto out;
+		}
+	}
+
+no_cluster_bitmap:
 	bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset),
 					 1, 0);
 	if (!bitmap_info) {
@@ -1444,19 +1540,10 @@ again:
 		goto new_bitmap;
 	}
 
-	end = bitmap_info->offset + (u64)(BITS_PER_BITMAP * ctl->unit);
-
-	if (offset >= bitmap_info->offset && offset + bytes > end) {
-		bitmap_set_bits(ctl, bitmap_info, offset, end - offset);
-		bytes -= end - offset;
-		offset = end;
-		added = 0;
-	} else if (offset >= bitmap_info->offset && offset + bytes <= end) {
-		bitmap_set_bits(ctl, bitmap_info, offset, bytes);
-		bytes = 0;
-	} else {
-		BUG();
-	}
+	bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes);
+	bytes -= bytes_added;
+	offset += bytes_added;
+	added = 0;
 
 	if (!bytes) {
 		ret = 1;
@@ -1735,11 +1822,6 @@ void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
 	       "\n", count);
 }
 
-static struct btrfs_free_space_op free_space_op = {
-	.recalc_thresholds	= recalculate_thresholds,
-	.use_bitmap		= use_bitmap,
-};
-
 void btrfs_init_free_space_ctl(struct btrfs_block_group_cache *block_group)
 {
 	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
@@ -1811,9 +1893,12 @@ void __btrfs_remove_free_space_cache_locked(struct btrfs_free_space_ctl *ctl)
 
 	while ((node = rb_last(&ctl->free_space_offset)) != NULL) {
 		info = rb_entry(node, struct btrfs_free_space, offset_index);
-		unlink_free_space(ctl, info);
-		kfree(info->bitmap);
-		kmem_cache_free(btrfs_free_space_cachep, info);
+		if (!info->bitmap) {
+			unlink_free_space(ctl, info);
+			kmem_cache_free(btrfs_free_space_cachep, info);
+		} else {
+			free_bitmap(ctl, info);
+		}
 		if (need_resched()) {
 			spin_unlock(&ctl->tree_lock);
 			cond_resched();
@@ -2111,9 +2196,11 @@ again:
 /*
  * This searches the block group for just extents to fill the cluster with.
  */
-static int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
-				   struct btrfs_free_cluster *cluster,
-				   u64 offset, u64 bytes, u64 min_bytes)
+static noinline int
+setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
+			struct btrfs_free_cluster *cluster,
+			struct list_head *bitmaps, u64 offset, u64 bytes,
+			u64 min_bytes)
 {
 	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
 	struct btrfs_free_space *first = NULL;
@@ -2135,6 +2222,8 @@ static int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
 	 * extent entry.
 	 */
 	while (entry->bitmap) {
+		if (list_empty(&entry->list))
+			list_add_tail(&entry->list, bitmaps);
 		node = rb_next(&entry->offset_index);
 		if (!node)
 			return -ENOSPC;
@@ -2154,8 +2243,12 @@ static int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
 			return -ENOSPC;
 		entry = rb_entry(node, struct btrfs_free_space, offset_index);
 
-		if (entry->bitmap)
+		if (entry->bitmap) {
+			if (list_empty(&entry->list))
+				list_add_tail(&entry->list, bitmaps);
 			continue;
+		}
+
 		/*
 		 * we haven't filled the empty size and the window is
 		 * very large.  reset and try again
@@ -2207,9 +2300,11 @@ static int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group,
  * This specifically looks for bitmaps that may work in the cluster, we assume
  * that we have already failed to find extents that will work.
  */
-static int setup_cluster_bitmap(struct btrfs_block_group_cache *block_group,
-				struct btrfs_free_cluster *cluster,
-				u64 offset, u64 bytes, u64 min_bytes)
+static noinline int
+setup_cluster_bitmap(struct btrfs_block_group_cache *block_group,
+		     struct btrfs_free_cluster *cluster,
+		     struct list_head *bitmaps, u64 offset, u64 bytes,
+		     u64 min_bytes)
 {
 	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
 	struct btrfs_free_space *entry;
@@ -2219,10 +2314,39 @@ static int setup_cluster_bitmap(struct btrfs_block_group_cache *block_group,
 	if (ctl->total_bitmaps == 0)
 		return -ENOSPC;
 
+	/*
+	 * First check our cached list of bitmaps and see if there is an entry
+	 * here that will work.
+	 */
+	list_for_each_entry(entry, bitmaps, list) {
+		if (entry->bytes < min_bytes)
+			continue;
+		ret = btrfs_bitmap_cluster(block_group, entry, cluster, offset,
+					   bytes, min_bytes);
+		if (!ret)
+			return 0;
+	}
+
+	/*
+	 * If we do have entries on our list and we are here then we didn't find
+	 * anything, so go ahead and get the next entry after the last entry in
+	 * this list and start the search from there.
+	 */
+	if (!list_empty(bitmaps)) {
+		entry = list_entry(bitmaps->prev, struct btrfs_free_space,
+				   list);
+		node = rb_next(&entry->offset_index);
+		if (!node)
+			return -ENOSPC;
+		entry = rb_entry(node, struct btrfs_free_space, offset_index);
+		goto search;
+	}
+
 	entry = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), 0, 1);
 	if (!entry)
 		return -ENOSPC;
 
+search:
 	node = &entry->offset_index;
 	do {
 		entry = rb_entry(node, struct btrfs_free_space, offset_index);
@@ -2253,6 +2377,8 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
 			     u64 offset, u64 bytes, u64 empty_size)
 {
 	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
+	struct list_head bitmaps;
+	struct btrfs_free_space *entry, *tmp;
 	u64 min_bytes;
 	int ret;
 
@@ -2291,11 +2417,16 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans,
 		goto out;
 	}
 
-	ret = setup_cluster_no_bitmap(block_group, cluster, offset, bytes,
-				      min_bytes);
+	INIT_LIST_HEAD(&bitmaps);
+	ret = setup_cluster_no_bitmap(block_group, cluster, &bitmaps, offset,
+				      bytes, min_bytes);
 	if (ret)
-		ret = setup_cluster_bitmap(block_group, cluster, offset,
-					   bytes, min_bytes);
+		ret = setup_cluster_bitmap(block_group, cluster, &bitmaps,
+					   offset, bytes, min_bytes);
+
+	/* Clear our temporary list */
+	list_for_each_entry_safe(entry, tmp, &bitmaps, list)
+		list_del_init(&entry->list);
 
 	if (!ret) {
 		atomic_inc(&block_group->count);
@@ -2481,7 +2612,7 @@ struct inode *lookup_free_ino_inode(struct btrfs_root *root,
 		return inode;
 
 	spin_lock(&root->cache_lock);
-	if (!root->fs_info->closing)
+	if (!btrfs_fs_closing(root->fs_info))
 		root->cache_inode = igrab(inode);
 	spin_unlock(&root->cache_lock);
 
@@ -2504,12 +2635,14 @@ int load_free_ino_cache(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
 	int ret = 0;
 	u64 root_gen = btrfs_root_generation(&root->root_item);
 
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return 0;
+
 	/*
 	 * If we're unmounting then just return, since this does a search on the
 	 * normal root and not the commit root and we could deadlock.
 	 */
-	smp_mb();
-	if (fs_info->closing)
+	if (btrfs_fs_closing(fs_info))
 		return 0;
 
 	path = btrfs_alloc_path();
@@ -2543,6 +2676,9 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root,
 	struct inode *inode;
 	int ret;
 
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return 0;
+
 	inode = lookup_free_ino_inode(root, path);
 	if (IS_ERR(inode))
 		return 0;
diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c
index 3262cd17a12..b4087e0fa87 100644
--- a/fs/btrfs/inode-map.c
+++ b/fs/btrfs/inode-map.c
@@ -38,6 +38,9 @@ static int caching_kthread(void *data)
 	int slot;
 	int ret;
 
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return 0;
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -59,8 +62,7 @@ again:
 		goto out;
 
 	while (1) {
-		smp_mb();
-		if (fs_info->closing)
+		if (btrfs_fs_closing(fs_info))
 			goto out;
 
 		leaf = path->nodes[0];
@@ -141,6 +143,9 @@ static void start_caching(struct btrfs_root *root)
 	int ret;
 	u64 objectid;
 
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return;
+
 	spin_lock(&root->cache_lock);
 	if (root->cached != BTRFS_CACHE_NO) {
 		spin_unlock(&root->cache_lock);
@@ -178,6 +183,9 @@ static void start_caching(struct btrfs_root *root)
 
 int btrfs_find_free_ino(struct btrfs_root *root, u64 *objectid)
 {
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return btrfs_find_free_objectid(root, objectid);
+
 again:
 	*objectid = btrfs_find_ino_for_alloc(root);
 
@@ -201,6 +209,10 @@ void btrfs_return_ino(struct btrfs_root *root, u64 objectid)
 {
 	struct btrfs_free_space_ctl *ctl = root->free_ino_ctl;
 	struct btrfs_free_space_ctl *pinned = root->free_ino_pinned;
+
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return;
+
 again:
 	if (root->cached == BTRFS_CACHE_FINISHED) {
 		__btrfs_add_free_space(ctl, objectid, 1);
@@ -250,6 +262,9 @@ void btrfs_unpin_free_ino(struct btrfs_root *root)
 	struct rb_node *n;
 	u64 count;
 
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return;
+
 	while (1) {
 		n = rb_first(rbroot);
 		if (!n)
@@ -388,9 +403,24 @@ int btrfs_save_ino_cache(struct btrfs_root *root,
 	int prealloc;
 	bool retry = false;
 
+	/* only fs tree and subvol/snap needs ino cache */
+	if (root->root_key.objectid != BTRFS_FS_TREE_OBJECTID &&
+	    (root->root_key.objectid < BTRFS_FIRST_FREE_OBJECTID ||
+	     root->root_key.objectid > BTRFS_LAST_FREE_OBJECTID))
+		return 0;
+
+	/* Don't save inode cache if we are deleting this root */
+	if (btrfs_root_refs(&root->root_item) == 0 &&
+	    root != root->fs_info->tree_root)
+		return 0;
+
+	if (!btrfs_test_opt(root, INODE_MAP_CACHE))
+		return 0;
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+
 again:
 	inode = lookup_free_ino_inode(root, path);
 	if (IS_ERR(inode) && PTR_ERR(inode) != -ENOENT) {
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 39a9d5750ef..2548a04a023 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -138,7 +138,6 @@ static noinline int insert_inline_extent(struct btrfs_trans_handle *trans,
 		return -ENOMEM;
 
 	path->leave_spinning = 1;
-	btrfs_set_trans_block_group(trans, inode);
 
 	key.objectid = btrfs_ino(inode);
 	key.offset = start;
@@ -426,9 +425,8 @@ again:
 		}
 	}
 	if (start == 0) {
-		trans = btrfs_join_transaction(root, 1);
+		trans = btrfs_join_transaction(root);
 		BUG_ON(IS_ERR(trans));
-		btrfs_set_trans_block_group(trans, inode);
 		trans->block_rsv = &root->fs_info->delalloc_block_rsv;
 
 		/* lets try to make an inline extent */
@@ -623,8 +621,9 @@ retry:
 			    async_extent->start + async_extent->ram_size - 1,
 			    GFP_NOFS);
 
-		trans = btrfs_join_transaction(root, 1);
+		trans = btrfs_join_transaction(root);
 		BUG_ON(IS_ERR(trans));
+		trans->block_rsv = &root->fs_info->delalloc_block_rsv;
 		ret = btrfs_reserve_extent(trans, root,
 					   async_extent->compressed_size,
 					   async_extent->compressed_size,
@@ -793,9 +792,8 @@ static noinline int cow_file_range(struct inode *inode,
 	int ret = 0;
 
 	BUG_ON(is_free_space_inode(root, inode));
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	BUG_ON(IS_ERR(trans));
-	btrfs_set_trans_block_group(trans, inode);
 	trans->block_rsv = &root->fs_info->delalloc_block_rsv;
 
 	num_bytes = (end - start + blocksize) & ~(blocksize - 1);
@@ -1077,10 +1075,12 @@ static noinline int run_delalloc_nocow(struct inode *inode,
 	nolock = is_free_space_inode(root, inode);
 
 	if (nolock)
-		trans = btrfs_join_transaction_nolock(root, 1);
+		trans = btrfs_join_transaction_nolock(root);
 	else
-		trans = btrfs_join_transaction(root, 1);
+		trans = btrfs_join_transaction(root);
+
 	BUG_ON(IS_ERR(trans));
+	trans->block_rsv = &root->fs_info->delalloc_block_rsv;
 
 	cow_start = (u64)-1;
 	cur_offset = start;
@@ -1519,8 +1519,6 @@ static noinline int add_pending_csums(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_ordered_sum *sum;
 
-	btrfs_set_trans_block_group(trans, inode);
-
 	list_for_each_entry(sum, list, list) {
 		btrfs_csum_file_blocks(trans,
 		       BTRFS_I(inode)->root->fs_info->csum_root, sum);
@@ -1735,11 +1733,10 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
 		ret = btrfs_ordered_update_i_size(inode, 0, ordered_extent);
 		if (!ret) {
 			if (nolock)
-				trans = btrfs_join_transaction_nolock(root, 1);
+				trans = btrfs_join_transaction_nolock(root);
 			else
-				trans = btrfs_join_transaction(root, 1);
+				trans = btrfs_join_transaction(root);
 			BUG_ON(IS_ERR(trans));
-			btrfs_set_trans_block_group(trans, inode);
 			trans->block_rsv = &root->fs_info->delalloc_block_rsv;
 			ret = btrfs_update_inode(trans, root, inode);
 			BUG_ON(ret);
@@ -1752,11 +1749,10 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
 			 0, &cached_state, GFP_NOFS);
 
 	if (nolock)
-		trans = btrfs_join_transaction_nolock(root, 1);
+		trans = btrfs_join_transaction_nolock(root);
 	else
-		trans = btrfs_join_transaction(root, 1);
+		trans = btrfs_join_transaction(root);
 	BUG_ON(IS_ERR(trans));
-	btrfs_set_trans_block_group(trans, inode);
 	trans->block_rsv = &root->fs_info->delalloc_block_rsv;
 
 	if (test_bit(BTRFS_ORDERED_COMPRESSED, &ordered_extent->flags))
@@ -1990,7 +1986,7 @@ static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 	}
 
 	if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)
-		return 0;
+		goto good;
 
 	if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID &&
 	    test_range_bit(io_tree, start, end, EXTENT_NODATASUM, 1, NULL)) {
@@ -2431,7 +2427,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
 					(u64)-1);
 
 	if (root->orphan_block_rsv || root->orphan_item_inserted) {
-		trans = btrfs_join_transaction(root, 1);
+		trans = btrfs_join_transaction(root);
 		if (!IS_ERR(trans))
 			btrfs_end_transaction(trans, root);
 	}
@@ -2511,12 +2507,17 @@ static void btrfs_read_locked_inode(struct inode *inode)
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct btrfs_key location;
 	int maybe_acls;
-	u64 alloc_group_block;
 	u32 rdev;
 	int ret;
+	bool filled = false;
+
+	ret = btrfs_fill_inode(inode, &rdev);
+	if (!ret)
+		filled = true;
 
 	path = btrfs_alloc_path();
 	BUG_ON(!path);
+	path->leave_spinning = 1;
 	memcpy(&location, &BTRFS_I(inode)->location, sizeof(location));
 
 	ret = btrfs_lookup_inode(NULL, root, path, &location, 0);
@@ -2524,8 +2525,18 @@ static void btrfs_read_locked_inode(struct inode *inode)
 		goto make_bad;
 
 	leaf = path->nodes[0];
+
+	if (filled)
+		goto cache_acl;
+
 	inode_item = btrfs_item_ptr(leaf, path->slots[0],
 				    struct btrfs_inode_item);
+	if (!leaf->map_token)
+		map_private_extent_buffer(leaf, (unsigned long)inode_item,
+					  sizeof(struct btrfs_inode_item),
+					  &leaf->map_token, &leaf->kaddr,
+					  &leaf->map_start, &leaf->map_len,
+					  KM_USER1);
 
 	inode->i_mode = btrfs_inode_mode(leaf, inode_item);
 	inode->i_nlink = btrfs_inode_nlink(leaf, inode_item);
@@ -2554,9 +2565,7 @@ static void btrfs_read_locked_inode(struct inode *inode)
 
 	BTRFS_I(inode)->index_cnt = (u64)-1;
 	BTRFS_I(inode)->flags = btrfs_inode_flags(leaf, inode_item);
-
-	alloc_group_block = btrfs_inode_block_group(leaf, inode_item);
-
+cache_acl:
 	/*
 	 * try to precache a NULL acl entry for files that don't have
 	 * any xattrs or acls
@@ -2566,10 +2575,12 @@ static void btrfs_read_locked_inode(struct inode *inode)
 	if (!maybe_acls)
 		cache_no_acl(inode);
 
-	BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0,
-						alloc_group_block, 0);
+	if (leaf->map_token) {
+		unmap_extent_buffer(leaf, leaf->map_token, KM_USER1);
+		leaf->map_token = NULL;
+	}
+
 	btrfs_free_path(path);
-	inode_item = NULL;
 
 	switch (inode->i_mode & S_IFMT) {
 	case S_IFREG:
@@ -2647,7 +2658,7 @@ static void fill_inode_item(struct btrfs_trans_handle *trans,
 	btrfs_set_inode_transid(leaf, item, trans->transid);
 	btrfs_set_inode_rdev(leaf, item, inode->i_rdev);
 	btrfs_set_inode_flags(leaf, item, BTRFS_I(inode)->flags);
-	btrfs_set_inode_block_group(leaf, item, BTRFS_I(inode)->block_group);
+	btrfs_set_inode_block_group(leaf, item, 0);
 
 	if (leaf->map_token) {
 		unmap_extent_buffer(leaf, leaf->map_token, KM_USER1);
@@ -2667,12 +2678,14 @@ noinline int btrfs_update_inode(struct btrfs_trans_handle *trans,
 	int ret;
 
 	/*
-	 * If root is tree root, it means this inode is used to
-	 * store free space information. And these inodes are updated
-	 * when committing the transaction, so they needn't delaye to
-	 * be updated, or deadlock will occured.
+	 * If the inode is a free space inode, we can deadlock during commit
+	 * if we put it into the delayed code.
+	 *
+	 * The data relocation inode should also be directly updated
+	 * without delay
 	 */
-	if (!is_free_space_inode(root, inode)) {
+	if (!is_free_space_inode(root, inode)
+	    && root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID) {
 		ret = btrfs_delayed_update_inode(trans, root, inode);
 		if (!ret)
 			btrfs_set_inode_last_trans(trans, inode);
@@ -3004,8 +3017,6 @@ static int btrfs_unlink(struct inode *dir, struct dentry *dentry)
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 
-	btrfs_set_trans_block_group(trans, dir);
-
 	btrfs_record_unlink_dir(trans, dir, dentry->d_inode, 0);
 
 	ret = btrfs_unlink_inode(trans, root, dir, dentry->d_inode,
@@ -3075,6 +3086,7 @@ int btrfs_unlink_subvol(struct btrfs_trans_handle *trans,
 	ret = btrfs_update_inode(trans, root, dir);
 	BUG_ON(ret);
 
+	btrfs_free_path(path);
 	return 0;
 }
 
@@ -3094,8 +3106,6 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry)
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 
-	btrfs_set_trans_block_group(trans, dir);
-
 	if (unlikely(btrfs_ino(inode) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) {
 		err = btrfs_unlink_subvol(trans, root, dir,
 					  BTRFS_I(inode)->location.objectid,
@@ -3514,7 +3524,6 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 				err = PTR_ERR(trans);
 				break;
 			}
-			btrfs_set_trans_block_group(trans, inode);
 
 			err = btrfs_drop_extents(trans, inode, cur_offset,
 						 cur_offset + hole_size,
@@ -3648,9 +3657,8 @@ void btrfs_evict_inode(struct inode *inode)
 	btrfs_i_size_write(inode, 0);
 
 	while (1) {
-		trans = btrfs_start_transaction(root, 0);
+		trans = btrfs_join_transaction(root);
 		BUG_ON(IS_ERR(trans));
-		btrfs_set_trans_block_group(trans, inode);
 		trans->block_rsv = root->orphan_block_rsv;
 
 		ret = btrfs_block_rsv_check(trans, root,
@@ -4071,13 +4079,7 @@ static int btrfs_dentry_delete(const struct dentry *dentry)
 static struct dentry *btrfs_lookup(struct inode *dir, struct dentry *dentry,
 				   struct nameidata *nd)
 {
-	struct inode *inode;
-
-	inode = btrfs_lookup_dentry(dir, dentry);
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-
-	return d_splice_alias(inode, dentry);
+	return d_splice_alias(btrfs_lookup_dentry(dir, dentry), dentry);
 }
 
 unsigned char btrfs_filetype_table[] = {
@@ -4133,7 +4135,8 @@ static int btrfs_real_readdir(struct file *filp, void *dirent,
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
-	path->reada = 2;
+
+	path->reada = 1;
 
 	if (key_type == BTRFS_DIR_INDEX_KEY) {
 		INIT_LIST_HEAD(&ins_list);
@@ -4268,18 +4271,16 @@ int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 	if (BTRFS_I(inode)->dummy_inode)
 		return 0;
 
-	smp_mb();
-	if (root->fs_info->closing && is_free_space_inode(root, inode))
+	if (btrfs_fs_closing(root->fs_info) && is_free_space_inode(root, inode))
 		nolock = true;
 
 	if (wbc->sync_mode == WB_SYNC_ALL) {
 		if (nolock)
-			trans = btrfs_join_transaction_nolock(root, 1);
+			trans = btrfs_join_transaction_nolock(root);
 		else
-			trans = btrfs_join_transaction(root, 1);
+			trans = btrfs_join_transaction(root);
 		if (IS_ERR(trans))
 			return PTR_ERR(trans);
-		btrfs_set_trans_block_group(trans, inode);
 		if (nolock)
 			ret = btrfs_end_transaction_nolock(trans, root);
 		else
@@ -4303,9 +4304,8 @@ void btrfs_dirty_inode(struct inode *inode, int flags)
 	if (BTRFS_I(inode)->dummy_inode)
 		return;
 
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	BUG_ON(IS_ERR(trans));
-	btrfs_set_trans_block_group(trans, inode);
 
 	ret = btrfs_update_inode(trans, root, inode);
 	if (ret && ret == -ENOSPC) {
@@ -4319,7 +4319,6 @@ void btrfs_dirty_inode(struct inode *inode, int flags)
 				       PTR_ERR(trans));
 			return;
 		}
-		btrfs_set_trans_block_group(trans, inode);
 
 		ret = btrfs_update_inode(trans, root, inode);
 		if (ret) {
@@ -4418,8 +4417,8 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root,
 				     struct inode *dir,
 				     const char *name, int name_len,
-				     u64 ref_objectid, u64 objectid,
-				     u64 alloc_hint, int mode, u64 *index)
+				     u64 ref_objectid, u64 objectid, int mode,
+				     u64 *index)
 {
 	struct inode *inode;
 	struct btrfs_inode_item *inode_item;
@@ -4472,8 +4471,6 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 		owner = 0;
 	else
 		owner = 1;
-	BTRFS_I(inode)->block_group =
-			btrfs_find_block_group(root, 0, alloc_hint, owner);
 
 	key[0].objectid = objectid;
 	btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY);
@@ -4527,6 +4524,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
 	inode_tree_add(inode);
 
 	trace_btrfs_inode_new(inode);
+	btrfs_set_inode_last_trans(trans, inode);
 
 	return inode;
 fail:
@@ -4629,15 +4627,13 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry,
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 
-	btrfs_set_trans_block_group(trans, dir);
-
 	err = btrfs_find_free_ino(root, &objectid);
 	if (err)
 		goto out_unlock;
 
 	inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
 				dentry->d_name.len, btrfs_ino(dir), objectid,
-				BTRFS_I(dir)->block_group, mode, &index);
+				mode, &index);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		goto out_unlock;
@@ -4649,7 +4645,6 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry,
 		goto out_unlock;
 	}
 
-	btrfs_set_trans_block_group(trans, inode);
 	err = btrfs_add_nondir(trans, dir, dentry, inode, 0, index);
 	if (err)
 		drop_inode = 1;
@@ -4658,8 +4653,6 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry,
 		init_special_inode(inode, inode->i_mode, rdev);
 		btrfs_update_inode(trans, root, inode);
 	}
-	btrfs_update_inode_block_group(trans, inode);
-	btrfs_update_inode_block_group(trans, dir);
 out_unlock:
 	nr = trans->blocks_used;
 	btrfs_end_transaction_throttle(trans, root);
@@ -4692,15 +4685,13 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry,
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 
-	btrfs_set_trans_block_group(trans, dir);
-
 	err = btrfs_find_free_ino(root, &objectid);
 	if (err)
 		goto out_unlock;
 
 	inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
 				dentry->d_name.len, btrfs_ino(dir), objectid,
-				BTRFS_I(dir)->block_group, mode, &index);
+				mode, &index);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		goto out_unlock;
@@ -4712,7 +4703,6 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry,
 		goto out_unlock;
 	}
 
-	btrfs_set_trans_block_group(trans, inode);
 	err = btrfs_add_nondir(trans, dir, dentry, inode, 0, index);
 	if (err)
 		drop_inode = 1;
@@ -4723,8 +4713,6 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry,
 		inode->i_op = &btrfs_file_inode_operations;
 		BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
 	}
-	btrfs_update_inode_block_group(trans, inode);
-	btrfs_update_inode_block_group(trans, dir);
 out_unlock:
 	nr = trans->blocks_used;
 	btrfs_end_transaction_throttle(trans, root);
@@ -4771,8 +4759,6 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 
 	btrfs_inc_nlink(inode);
 	inode->i_ctime = CURRENT_TIME;
-
-	btrfs_set_trans_block_group(trans, dir);
 	ihold(inode);
 
 	err = btrfs_add_nondir(trans, dir, dentry, inode, 1, index);
@@ -4780,12 +4766,10 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 	if (err) {
 		drop_inode = 1;
 	} else {
-		struct dentry *parent = dget_parent(dentry);
-		btrfs_update_inode_block_group(trans, dir);
+		struct dentry *parent = dentry->d_parent;
 		err = btrfs_update_inode(trans, root, inode);
 		BUG_ON(err);
 		btrfs_log_new_name(trans, inode, NULL, parent);
-		dput(parent);
 	}
 
 	nr = trans->blocks_used;
@@ -4818,7 +4802,6 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	trans = btrfs_start_transaction(root, 5);
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
-	btrfs_set_trans_block_group(trans, dir);
 
 	err = btrfs_find_free_ino(root, &objectid);
 	if (err)
@@ -4826,8 +4809,7 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 
 	inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
 				dentry->d_name.len, btrfs_ino(dir), objectid,
-				BTRFS_I(dir)->block_group, S_IFDIR | mode,
-				&index);
+				S_IFDIR | mode, &index);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		goto out_fail;
@@ -4841,7 +4823,6 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 
 	inode->i_op = &btrfs_dir_inode_operations;
 	inode->i_fop = &btrfs_dir_file_operations;
-	btrfs_set_trans_block_group(trans, inode);
 
 	btrfs_i_size_write(inode, 0);
 	err = btrfs_update_inode(trans, root, inode);
@@ -4855,8 +4836,6 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 
 	d_instantiate(dentry, inode);
 	drop_on_err = 0;
-	btrfs_update_inode_block_group(trans, inode);
-	btrfs_update_inode_block_group(trans, dir);
 
 out_fail:
 	nr = trans->blocks_used;
@@ -4989,7 +4968,15 @@ again:
 
 	if (!path) {
 		path = btrfs_alloc_path();
-		BUG_ON(!path);
+		if (!path) {
+			err = -ENOMEM;
+			goto out;
+		}
+		/*
+		 * Chances are we'll be called again, so go ahead and do
+		 * readahead
+		 */
+		path->reada = 1;
 	}
 
 	ret = btrfs_lookup_file_extent(trans, root, path,
@@ -5130,8 +5117,10 @@ again:
 				kunmap(page);
 				free_extent_map(em);
 				em = NULL;
+
 				btrfs_release_path(path);
-				trans = btrfs_join_transaction(root, 1);
+				trans = btrfs_join_transaction(root);
+
 				if (IS_ERR(trans))
 					return ERR_CAST(trans);
 				goto again;
@@ -5375,7 +5364,7 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode,
 		btrfs_drop_extent_cache(inode, start, start + len - 1, 0);
 	}
 
-	trans = btrfs_join_transaction(root, 0);
+	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans))
 		return ERR_CAST(trans);
 
@@ -5611,7 +5600,7 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 		 * to make sure the current transaction stays open
 		 * while we look for nocow cross refs
 		 */
-		trans = btrfs_join_transaction(root, 0);
+		trans = btrfs_join_transaction(root);
 		if (IS_ERR(trans))
 			goto must_cow;
 
@@ -5750,7 +5739,7 @@ again:
 
 	BUG_ON(!ordered);
 
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans)) {
 		err = -ENOMEM;
 		goto out;
@@ -6500,6 +6489,7 @@ out:
 static int btrfs_truncate(struct inode *inode)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct btrfs_block_rsv *rsv;
 	int ret;
 	int err = 0;
 	struct btrfs_trans_handle *trans;
@@ -6513,28 +6503,80 @@ static int btrfs_truncate(struct inode *inode)
 	btrfs_wait_ordered_range(inode, inode->i_size & (~mask), (u64)-1);
 	btrfs_ordered_update_i_size(inode, inode->i_size, NULL);
 
-	trans = btrfs_start_transaction(root, 5);
-	if (IS_ERR(trans))
-		return PTR_ERR(trans);
+	/*
+	 * Yes ladies and gentelment, this is indeed ugly.  The fact is we have
+	 * 3 things going on here
+	 *
+	 * 1) We need to reserve space for our orphan item and the space to
+	 * delete our orphan item.  Lord knows we don't want to have a dangling
+	 * orphan item because we didn't reserve space to remove it.
+	 *
+	 * 2) We need to reserve space to update our inode.
+	 *
+	 * 3) We need to have something to cache all the space that is going to
+	 * be free'd up by the truncate operation, but also have some slack
+	 * space reserved in case it uses space during the truncate (thank you
+	 * very much snapshotting).
+	 *
+	 * And we need these to all be seperate.  The fact is we can use alot of
+	 * space doing the truncate, and we have no earthly idea how much space
+	 * we will use, so we need the truncate reservation to be seperate so it
+	 * doesn't end up using space reserved for updating the inode or
+	 * removing the orphan item.  We also need to be able to stop the
+	 * transaction and start a new one, which means we need to be able to
+	 * update the inode several times, and we have no idea of knowing how
+	 * many times that will be, so we can't just reserve 1 item for the
+	 * entirety of the opration, so that has to be done seperately as well.
+	 * Then there is the orphan item, which does indeed need to be held on
+	 * to for the whole operation, and we need nobody to touch this reserved
+	 * space except the orphan code.
+	 *
+	 * So that leaves us with
+	 *
+	 * 1) root->orphan_block_rsv - for the orphan deletion.
+	 * 2) rsv - for the truncate reservation, which we will steal from the
+	 * transaction reservation.
+	 * 3) fs_info->trans_block_rsv - this will have 1 items worth left for
+	 * updating the inode.
+	 */
+	rsv = btrfs_alloc_block_rsv(root);
+	if (!rsv)
+		return -ENOMEM;
+	btrfs_add_durable_block_rsv(root->fs_info, rsv);
 
-	btrfs_set_trans_block_group(trans, inode);
+	trans = btrfs_start_transaction(root, 4);
+	if (IS_ERR(trans)) {
+		err = PTR_ERR(trans);
+		goto out;
+	}
+
+	/*
+	 * Reserve space for the truncate process.  Truncate should be adding
+	 * space, but if there are snapshots it may end up using space.
+	 */
+	ret = btrfs_truncate_reserve_metadata(trans, root, rsv);
+	BUG_ON(ret);
 
 	ret = btrfs_orphan_add(trans, inode);
 	if (ret) {
 		btrfs_end_transaction(trans, root);
-		return ret;
+		goto out;
 	}
 
 	nr = trans->blocks_used;
 	btrfs_end_transaction(trans, root);
 	btrfs_btree_balance_dirty(root, nr);
 
-	/* Now start a transaction for the truncate */
-	trans = btrfs_start_transaction(root, 0);
-	if (IS_ERR(trans))
-		return PTR_ERR(trans);
-	btrfs_set_trans_block_group(trans, inode);
-	trans->block_rsv = root->orphan_block_rsv;
+	/*
+	 * Ok so we've already migrated our bytes over for the truncate, so here
+	 * just reserve the one slot we need for updating the inode.
+	 */
+	trans = btrfs_start_transaction(root, 1);
+	if (IS_ERR(trans)) {
+		err = PTR_ERR(trans);
+		goto out;
+	}
+	trans->block_rsv = rsv;
 
 	/*
 	 * setattr is responsible for setting the ordered_data_close flag,
@@ -6558,24 +6600,17 @@ static int btrfs_truncate(struct inode *inode)
 
 	while (1) {
 		if (!trans) {
-			trans = btrfs_start_transaction(root, 0);
-			if (IS_ERR(trans))
-				return PTR_ERR(trans);
-			btrfs_set_trans_block_group(trans, inode);
-			trans->block_rsv = root->orphan_block_rsv;
-		}
+			trans = btrfs_start_transaction(root, 3);
+			if (IS_ERR(trans)) {
+				err = PTR_ERR(trans);
+				goto out;
+			}
 
-		ret = btrfs_block_rsv_check(trans, root,
-					    root->orphan_block_rsv, 0, 5);
-		if (ret == -EAGAIN) {
-			ret = btrfs_commit_transaction(trans, root);
-			if (ret)
-				return ret;
-			trans = NULL;
-			continue;
-		} else if (ret) {
-			err = ret;
-			break;
+			ret = btrfs_truncate_reserve_metadata(trans, root,
+							      rsv);
+			BUG_ON(ret);
+
+			trans->block_rsv = rsv;
 		}
 
 		ret = btrfs_truncate_inode_items(trans, root, inode,
@@ -6586,6 +6621,7 @@ static int btrfs_truncate(struct inode *inode)
 			break;
 		}
 
+		trans->block_rsv = &root->fs_info->trans_block_rsv;
 		ret = btrfs_update_inode(trans, root, inode);
 		if (ret) {
 			err = ret;
@@ -6599,6 +6635,7 @@ static int btrfs_truncate(struct inode *inode)
 	}
 
 	if (ret == 0 && inode->i_nlink > 0) {
+		trans->block_rsv = root->orphan_block_rsv;
 		ret = btrfs_orphan_del(trans, inode);
 		if (ret)
 			err = ret;
@@ -6610,15 +6647,20 @@ static int btrfs_truncate(struct inode *inode)
 		ret = btrfs_orphan_del(NULL, inode);
 	}
 
+	trans->block_rsv = &root->fs_info->trans_block_rsv;
 	ret = btrfs_update_inode(trans, root, inode);
 	if (ret && !err)
 		err = ret;
 
 	nr = trans->blocks_used;
 	ret = btrfs_end_transaction_throttle(trans, root);
+	btrfs_btree_balance_dirty(root, nr);
+
+out:
+	btrfs_free_block_rsv(root, rsv);
+
 	if (ret && !err)
 		err = ret;
-	btrfs_btree_balance_dirty(root, nr);
 
 	return err;
 }
@@ -6627,15 +6669,14 @@ static int btrfs_truncate(struct inode *inode)
  * create a new subvolume directory/inode (helper for the ioctl).
  */
 int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
-			     struct btrfs_root *new_root,
-			     u64 new_dirid, u64 alloc_hint)
+			     struct btrfs_root *new_root, u64 new_dirid)
 {
 	struct inode *inode;
 	int err;
 	u64 index = 0;
 
 	inode = btrfs_new_inode(trans, new_root, NULL, "..", 2, new_dirid,
-				new_dirid, alloc_hint, S_IFDIR | 0700, &index);
+				new_dirid, S_IFDIR | 0700, &index);
 	if (IS_ERR(inode))
 		return PTR_ERR(inode);
 	inode->i_op = &btrfs_dir_inode_operations;
@@ -6748,21 +6789,6 @@ void btrfs_destroy_inode(struct inode *inode)
 		spin_unlock(&root->fs_info->ordered_extent_lock);
 	}
 
-	if (root == root->fs_info->tree_root) {
-		struct btrfs_block_group_cache *block_group;
-
-		block_group = btrfs_lookup_block_group(root->fs_info,
-						BTRFS_I(inode)->block_group);
-		if (block_group && block_group->inode == inode) {
-			spin_lock(&block_group->lock);
-			block_group->inode = NULL;
-			spin_unlock(&block_group->lock);
-			btrfs_put_block_group(block_group);
-		} else if (block_group) {
-			btrfs_put_block_group(block_group);
-		}
-	}
-
 	spin_lock(&root->orphan_lock);
 	if (!list_empty(&BTRFS_I(inode)->i_orphan)) {
 		printk(KERN_INFO "BTRFS: inode %llu still on the orphan list\n",
@@ -6867,7 +6893,7 @@ static int btrfs_getattr(struct vfsmount *mnt,
 {
 	struct inode *inode = dentry->d_inode;
 	generic_fillattr(inode, stat);
-	stat->dev = BTRFS_I(inode)->root->anon_super.s_dev;
+	stat->dev = BTRFS_I(inode)->root->anon_dev;
 	stat->blksize = PAGE_CACHE_SIZE;
 	stat->blocks = (inode_get_bytes(inode) +
 			BTRFS_I(inode)->delalloc_bytes) >> 9;
@@ -6948,8 +6974,6 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
                 goto out_notrans;
         }
 
-	btrfs_set_trans_block_group(trans, new_dir);
-
 	if (dest != root)
 		btrfs_record_root_in_trans(trans, dest);
 
@@ -7037,9 +7061,8 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	BUG_ON(ret);
 
 	if (old_ino != BTRFS_FIRST_FREE_OBJECTID) {
-		struct dentry *parent = dget_parent(new_dentry);
+		struct dentry *parent = new_dentry->d_parent;
 		btrfs_log_new_name(trans, old_inode, old_dir, parent);
-		dput(parent);
 		btrfs_end_log_trans(root);
 	}
 out_fail:
@@ -7131,16 +7154,13 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 
-	btrfs_set_trans_block_group(trans, dir);
-
 	err = btrfs_find_free_ino(root, &objectid);
 	if (err)
 		goto out_unlock;
 
 	inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
 				dentry->d_name.len, btrfs_ino(dir), objectid,
-				BTRFS_I(dir)->block_group, S_IFLNK|S_IRWXUGO,
-				&index);
+				S_IFLNK|S_IRWXUGO, &index);
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		goto out_unlock;
@@ -7152,7 +7172,6 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 		goto out_unlock;
 	}
 
-	btrfs_set_trans_block_group(trans, inode);
 	err = btrfs_add_nondir(trans, dir, dentry, inode, 0, index);
 	if (err)
 		drop_inode = 1;
@@ -7163,8 +7182,6 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 		inode->i_op = &btrfs_file_inode_operations;
 		BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
 	}
-	btrfs_update_inode_block_group(trans, inode);
-	btrfs_update_inode_block_group(trans, dir);
 	if (drop_inode)
 		goto out_unlock;
 
@@ -7306,7 +7323,7 @@ static int btrfs_set_page_dirty(struct page *page)
 	return __set_page_dirty_nobuffers(page);
 }
 
-static int btrfs_permission(struct inode *inode, int mask, unsigned int flags)
+static int btrfs_permission(struct inode *inode, int mask)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 
@@ -7314,7 +7331,7 @@ static int btrfs_permission(struct inode *inode, int mask, unsigned int flags)
 		return -EROFS;
 	if ((BTRFS_I(inode)->flags & BTRFS_INODE_READONLY) && (mask & MAY_WRITE))
 		return -EACCES;
-	return generic_permission(inode, mask, flags, btrfs_check_acl);
+	return generic_permission(inode, mask);
 }
 
 static const struct inode_operations btrfs_dir_inode_operations = {
@@ -7334,10 +7351,12 @@ static const struct inode_operations btrfs_dir_inode_operations = {
 	.listxattr	= btrfs_listxattr,
 	.removexattr	= btrfs_removexattr,
 	.permission	= btrfs_permission,
+	.check_acl	= btrfs_check_acl,
 };
 static const struct inode_operations btrfs_dir_ro_inode_operations = {
 	.lookup		= btrfs_lookup,
 	.permission	= btrfs_permission,
+	.check_acl	= btrfs_check_acl,
 };
 
 static const struct file_operations btrfs_dir_file_operations = {
@@ -7406,6 +7425,7 @@ static const struct inode_operations btrfs_file_inode_operations = {
 	.removexattr	= btrfs_removexattr,
 	.permission	= btrfs_permission,
 	.fiemap		= btrfs_fiemap,
+	.check_acl	= btrfs_check_acl,
 };
 static const struct inode_operations btrfs_special_inode_operations = {
 	.getattr	= btrfs_getattr,
@@ -7415,6 +7435,7 @@ static const struct inode_operations btrfs_special_inode_operations = {
 	.getxattr	= btrfs_getxattr,
 	.listxattr	= btrfs_listxattr,
 	.removexattr	= btrfs_removexattr,
+	.check_acl	= btrfs_check_acl,
 };
 static const struct inode_operations btrfs_symlink_inode_operations = {
 	.readlink	= generic_readlink,
@@ -7426,6 +7447,7 @@ static const struct inode_operations btrfs_symlink_inode_operations = {
 	.getxattr	= btrfs_getxattr,
 	.listxattr	= btrfs_listxattr,
 	.removexattr	= btrfs_removexattr,
+	.check_acl	= btrfs_check_acl,
 };
 
 const struct dentry_operations btrfs_dentry_operations = {
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 85e818ce00c..622543309eb 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -243,7 +243,7 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
 	}
 
-	trans = btrfs_join_transaction(root, 1);
+	trans = btrfs_join_transaction(root);
 	BUG_ON(IS_ERR(trans));
 
 	ret = btrfs_update_inode(trans, root, inode);
@@ -323,7 +323,7 @@ static noinline int create_subvol(struct btrfs_root *root,
 	struct btrfs_inode_item *inode_item;
 	struct extent_buffer *leaf;
 	struct btrfs_root *new_root;
-	struct dentry *parent = dget_parent(dentry);
+	struct dentry *parent = dentry->d_parent;
 	struct inode *dir;
 	int ret;
 	int err;
@@ -332,10 +332,8 @@ static noinline int create_subvol(struct btrfs_root *root,
 	u64 index = 0;
 
 	ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
-	if (ret) {
-		dput(parent);
+	if (ret)
 		return ret;
-	}
 
 	dir = parent->d_inode;
 
@@ -346,10 +344,8 @@ static noinline int create_subvol(struct btrfs_root *root,
 	 * 2 - dir items
 	 */
 	trans = btrfs_start_transaction(root, 6);
-	if (IS_ERR(trans)) {
-		dput(parent);
+	if (IS_ERR(trans))
 		return PTR_ERR(trans);
-	}
 
 	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
 				      0, objectid, NULL, 0, 0, 0);
@@ -414,8 +410,7 @@ static noinline int create_subvol(struct btrfs_root *root,
 
 	btrfs_record_root_in_trans(trans, new_root);
 
-	ret = btrfs_create_subvol_root(trans, new_root, new_dirid,
-				       BTRFS_I(dir)->block_group);
+	ret = btrfs_create_subvol_root(trans, new_root, new_dirid);
 	/*
 	 * insert the directory item
 	 */
@@ -440,7 +435,6 @@ static noinline int create_subvol(struct btrfs_root *root,
 
 	d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry));
 fail:
-	dput(parent);
 	if (async_transid) {
 		*async_transid = trans->transid;
 		err = btrfs_commit_transaction_async(trans, root, 1);
@@ -457,7 +451,6 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
 			   bool readonly)
 {
 	struct inode *inode;
-	struct dentry *parent;
 	struct btrfs_pending_snapshot *pending_snapshot;
 	struct btrfs_trans_handle *trans;
 	int ret;
@@ -483,8 +476,10 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
 	ret = btrfs_snap_reserve_metadata(trans, pending_snapshot);
 	BUG_ON(ret);
 
+	spin_lock(&root->fs_info->trans_lock);
 	list_add(&pending_snapshot->list,
 		 &trans->transaction->pending_snapshots);
+	spin_unlock(&root->fs_info->trans_lock);
 	if (async_transid) {
 		*async_transid = trans->transid;
 		ret = btrfs_commit_transaction_async(trans,
@@ -503,9 +498,7 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
 	if (ret)
 		goto fail;
 
-	parent = dget_parent(dentry);
-	inode = btrfs_lookup_dentry(parent->d_inode, dentry);
-	dput(parent);
+	inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry);
 	if (IS_ERR(inode)) {
 		ret = PTR_ERR(inode);
 		goto fail;
@@ -707,16 +700,17 @@ static int find_new_extents(struct btrfs_root *root,
 	struct btrfs_file_extent_item *extent;
 	int type;
 	int ret;
+	u64 ino = btrfs_ino(inode);
 
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
 
-	min_key.objectid = inode->i_ino;
+	min_key.objectid = ino;
 	min_key.type = BTRFS_EXTENT_DATA_KEY;
 	min_key.offset = *off;
 
-	max_key.objectid = inode->i_ino;
+	max_key.objectid = ino;
 	max_key.type = (u8)-1;
 	max_key.offset = (u64)-1;
 
@@ -727,7 +721,7 @@ static int find_new_extents(struct btrfs_root *root,
 					   path, 0, newer_than);
 		if (ret != 0)
 			goto none;
-		if (min_key.objectid != inode->i_ino)
+		if (min_key.objectid != ino)
 			goto none;
 		if (min_key.type != BTRFS_EXTENT_DATA_KEY)
 			goto none;
@@ -2054,29 +2048,34 @@ static long btrfs_ioctl_rm_dev(struct btrfs_root *root, void __user *arg)
 
 static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
 {
-	struct btrfs_ioctl_fs_info_args fi_args;
+	struct btrfs_ioctl_fs_info_args *fi_args;
 	struct btrfs_device *device;
 	struct btrfs_device *next;
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
+	int ret = 0;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	fi_args.num_devices = fs_devices->num_devices;
-	fi_args.max_id = 0;
-	memcpy(&fi_args.fsid, root->fs_info->fsid, sizeof(fi_args.fsid));
+	fi_args = kzalloc(sizeof(*fi_args), GFP_KERNEL);
+	if (!fi_args)
+		return -ENOMEM;
+
+	fi_args->num_devices = fs_devices->num_devices;
+	memcpy(&fi_args->fsid, root->fs_info->fsid, sizeof(fi_args->fsid));
 
 	mutex_lock(&fs_devices->device_list_mutex);
 	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
-		if (device->devid > fi_args.max_id)
-			fi_args.max_id = device->devid;
+		if (device->devid > fi_args->max_id)
+			fi_args->max_id = device->devid;
 	}
 	mutex_unlock(&fs_devices->device_list_mutex);
 
-	if (copy_to_user(arg, &fi_args, sizeof(fi_args)))
-		return -EFAULT;
+	if (copy_to_user(arg, fi_args, sizeof(*fi_args)))
+		ret = -EFAULT;
 
-	return 0;
+	kfree(fi_args);
+	return ret;
 }
 
 static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
@@ -2489,12 +2488,10 @@ static long btrfs_ioctl_trans_start(struct file *file)
 	if (ret)
 		goto out;
 
-	mutex_lock(&root->fs_info->trans_mutex);
-	root->fs_info->open_ioctl_trans++;
-	mutex_unlock(&root->fs_info->trans_mutex);
+	atomic_inc(&root->fs_info->open_ioctl_trans);
 
 	ret = -ENOMEM;
-	trans = btrfs_start_ioctl_transaction(root, 0);
+	trans = btrfs_start_ioctl_transaction(root);
 	if (IS_ERR(trans))
 		goto out_drop;
 
@@ -2502,9 +2499,7 @@ static long btrfs_ioctl_trans_start(struct file *file)
 	return 0;
 
 out_drop:
-	mutex_lock(&root->fs_info->trans_mutex);
-	root->fs_info->open_ioctl_trans--;
-	mutex_unlock(&root->fs_info->trans_mutex);
+	atomic_dec(&root->fs_info->open_ioctl_trans);
 	mnt_drop_write(file->f_path.mnt);
 out:
 	return ret;
@@ -2738,9 +2733,7 @@ long btrfs_ioctl_trans_end(struct file *file)
 
 	btrfs_end_transaction(trans, root);
 
-	mutex_lock(&root->fs_info->trans_mutex);
-	root->fs_info->open_ioctl_trans--;
-	mutex_unlock(&root->fs_info->trans_mutex);
+	atomic_dec(&root->fs_info->open_ioctl_trans);
 
 	mnt_drop_write(file->f_path.mnt);
 	return 0;
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index ca38eca70af..5e0a3dc79a4 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -677,6 +677,8 @@ struct backref_node *build_backref_tree(struct reloc_control *rc,
 		err = -ENOMEM;
 		goto out;
 	}
+	path1->reada = 1;
+	path2->reada = 2;
 
 	node = alloc_backref_node(cache);
 	if (!node) {
@@ -1366,7 +1368,7 @@ int btrfs_update_reloc_root(struct btrfs_trans_handle *trans,
 	int ret;
 
 	if (!root->reloc_root)
-		return 0;
+		goto out;
 
 	reloc_root = root->reloc_root;
 	root_item = &reloc_root->root_item;
@@ -1388,6 +1390,8 @@ int btrfs_update_reloc_root(struct btrfs_trans_handle *trans,
 	ret = btrfs_update_root(trans, root->fs_info->tree_root,
 				&reloc_root->root_key, root_item);
 	BUG_ON(ret);
+
+out:
 	return 0;
 }
 
@@ -1999,6 +2003,7 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+	path->reada = 1;
 
 	reloc_root = root->reloc_root;
 	root_item = &reloc_root->root_item;
@@ -2139,10 +2144,11 @@ int prepare_to_merge(struct reloc_control *rc, int err)
 	u64 num_bytes = 0;
 	int ret;
 
-	mutex_lock(&root->fs_info->trans_mutex);
+	mutex_lock(&root->fs_info->reloc_mutex);
 	rc->merging_rsv_size += root->nodesize * (BTRFS_MAX_LEVEL - 1) * 2;
 	rc->merging_rsv_size += rc->nodes_relocated * 2;
-	mutex_unlock(&root->fs_info->trans_mutex);
+	mutex_unlock(&root->fs_info->reloc_mutex);
+
 again:
 	if (!err) {
 		num_bytes = rc->merging_rsv_size;
@@ -2152,7 +2158,7 @@ again:
 			err = ret;
 	}
 
-	trans = btrfs_join_transaction(rc->extent_root, 1);
+	trans = btrfs_join_transaction(rc->extent_root);
 	if (IS_ERR(trans)) {
 		if (!err)
 			btrfs_block_rsv_release(rc->extent_root,
@@ -2211,9 +2217,16 @@ int merge_reloc_roots(struct reloc_control *rc)
 	int ret;
 again:
 	root = rc->extent_root;
-	mutex_lock(&root->fs_info->trans_mutex);
+
+	/*
+	 * this serializes us with btrfs_record_root_in_transaction,
+	 * we have to make sure nobody is in the middle of
+	 * adding their roots to the list while we are
+	 * doing this splice
+	 */
+	mutex_lock(&root->fs_info->reloc_mutex);
 	list_splice_init(&rc->reloc_roots, &reloc_roots);
-	mutex_unlock(&root->fs_info->trans_mutex);
+	mutex_unlock(&root->fs_info->reloc_mutex);
 
 	while (!list_empty(&reloc_roots)) {
 		found = 1;
@@ -3236,7 +3249,7 @@ truncate:
 		goto out;
 	}
 
-	trans = btrfs_join_transaction(root, 0);
+	trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans)) {
 		btrfs_free_path(path);
 		ret = PTR_ERR(trans);
@@ -3300,6 +3313,7 @@ static int find_data_references(struct reloc_control *rc,
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+	path->reada = 1;
 
 	root = read_fs_root(rc->extent_root->fs_info, ref_root);
 	if (IS_ERR(root)) {
@@ -3586,17 +3600,19 @@ next:
 static void set_reloc_control(struct reloc_control *rc)
 {
 	struct btrfs_fs_info *fs_info = rc->extent_root->fs_info;
-	mutex_lock(&fs_info->trans_mutex);
+
+	mutex_lock(&fs_info->reloc_mutex);
 	fs_info->reloc_ctl = rc;
-	mutex_unlock(&fs_info->trans_mutex);
+	mutex_unlock(&fs_info->reloc_mutex);
 }
 
 static void unset_reloc_control(struct reloc_control *rc)
 {
 	struct btrfs_fs_info *fs_info = rc->extent_root->fs_info;
-	mutex_lock(&fs_info->trans_mutex);
+
+	mutex_lock(&fs_info->reloc_mutex);
 	fs_info->reloc_ctl = NULL;
-	mutex_unlock(&fs_info->trans_mutex);
+	mutex_unlock(&fs_info->reloc_mutex);
 }
 
 static int check_extent_flags(u64 flags)
@@ -3645,7 +3661,7 @@ int prepare_to_relocate(struct reloc_control *rc)
 	rc->create_reloc_tree = 1;
 	set_reloc_control(rc);
 
-	trans = btrfs_join_transaction(rc->extent_root, 1);
+	trans = btrfs_join_transaction(rc->extent_root);
 	BUG_ON(IS_ERR(trans));
 	btrfs_commit_transaction(trans, rc->extent_root);
 	return 0;
@@ -3668,6 +3684,7 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+	path->reada = 1;
 
 	ret = prepare_to_relocate(rc);
 	if (ret) {
@@ -3834,7 +3851,7 @@ restart:
 	btrfs_block_rsv_release(rc->extent_root, rc->block_rsv, (u64)-1);
 
 	/* get rid of pinned extents */
-	trans = btrfs_join_transaction(rc->extent_root, 1);
+	trans = btrfs_join_transaction(rc->extent_root);
 	if (IS_ERR(trans))
 		err = PTR_ERR(trans);
 	else
@@ -4093,6 +4110,7 @@ int btrfs_recover_relocation(struct btrfs_root *root)
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
+	path->reada = -1;
 
 	key.objectid = BTRFS_TREE_RELOC_OBJECTID;
 	key.type = BTRFS_ROOT_ITEM_KEY;
@@ -4159,7 +4177,7 @@ int btrfs_recover_relocation(struct btrfs_root *root)
 
 	set_reloc_control(rc);
 
-	trans = btrfs_join_transaction(rc->extent_root, 1);
+	trans = btrfs_join_transaction(rc->extent_root);
 	if (IS_ERR(trans)) {
 		unset_reloc_control(rc);
 		err = PTR_ERR(trans);
@@ -4193,7 +4211,7 @@ int btrfs_recover_relocation(struct btrfs_root *root)
 
 	unset_reloc_control(rc);
 
-	trans = btrfs_join_transaction(rc->extent_root, 1);
+	trans = btrfs_join_transaction(rc->extent_root);
 	if (IS_ERR(trans))
 		err = PTR_ERR(trans);
 	else
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index 6dfed0c27ac..a8d03d5efb5 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -16,13 +16,7 @@
  * Boston, MA 021110-1307, USA.
  */
 
-#include <linux/sched.h>
-#include <linux/pagemap.h>
-#include <linux/writeback.h>
 #include <linux/blkdev.h>
-#include <linux/rbtree.h>
-#include <linux/slab.h>
-#include <linux/workqueue.h>
 #include "ctree.h"
 #include "volumes.h"
 #include "disk-io.h"
@@ -117,33 +111,37 @@ static void scrub_free_csums(struct scrub_dev *sdev)
 	}
 }
 
+static void scrub_free_bio(struct bio *bio)
+{
+	int i;
+	struct page *last_page = NULL;
+
+	if (!bio)
+		return;
+
+	for (i = 0; i < bio->bi_vcnt; ++i) {
+		if (bio->bi_io_vec[i].bv_page == last_page)
+			continue;
+		last_page = bio->bi_io_vec[i].bv_page;
+		__free_page(last_page);
+	}
+	bio_put(bio);
+}
+
 static noinline_for_stack void scrub_free_dev(struct scrub_dev *sdev)
 {
 	int i;
-	int j;
-	struct page *last_page;
 
 	if (!sdev)
 		return;
 
 	for (i = 0; i < SCRUB_BIOS_PER_DEV; ++i) {
 		struct scrub_bio *sbio = sdev->bios[i];
-		struct bio *bio;
 
 		if (!sbio)
 			break;
 
-		bio = sbio->bio;
-		if (bio) {
-			last_page = NULL;
-			for (j = 0; j < bio->bi_vcnt; ++j) {
-				if (bio->bi_io_vec[j].bv_page == last_page)
-					continue;
-				last_page = bio->bi_io_vec[j].bv_page;
-				__free_page(last_page);
-			}
-			bio_put(bio);
-		}
+		scrub_free_bio(sbio->bio);
 		kfree(sbio);
 	}
 
@@ -156,8 +154,6 @@ struct scrub_dev *scrub_setup_dev(struct btrfs_device *dev)
 {
 	struct scrub_dev *sdev;
 	int		i;
-	int		j;
-	int		ret;
 	struct btrfs_fs_info *fs_info = dev->dev_root->fs_info;
 
 	sdev = kzalloc(sizeof(*sdev), GFP_NOFS);
@@ -165,7 +161,6 @@ struct scrub_dev *scrub_setup_dev(struct btrfs_device *dev)
 		goto nomem;
 	sdev->dev = dev;
 	for (i = 0; i < SCRUB_BIOS_PER_DEV; ++i) {
-		struct bio *bio;
 		struct scrub_bio *sbio;
 
 		sbio = kzalloc(sizeof(*sbio), GFP_NOFS);
@@ -173,32 +168,10 @@ struct scrub_dev *scrub_setup_dev(struct btrfs_device *dev)
 			goto nomem;
 		sdev->bios[i] = sbio;
 
-		bio = bio_kmalloc(GFP_NOFS, SCRUB_PAGES_PER_BIO);
-		if (!bio)
-			goto nomem;
-
 		sbio->index = i;
 		sbio->sdev = sdev;
-		sbio->bio = bio;
 		sbio->count = 0;
 		sbio->work.func = scrub_checksum;
-		bio->bi_private = sdev->bios[i];
-		bio->bi_end_io = scrub_bio_end_io;
-		bio->bi_sector = 0;
-		bio->bi_bdev = dev->bdev;
-		bio->bi_size = 0;
-
-		for (j = 0; j < SCRUB_PAGES_PER_BIO; ++j) {
-			struct page *page;
-			page = alloc_page(GFP_NOFS);
-			if (!page)
-				goto nomem;
-
-			ret = bio_add_page(bio, page, PAGE_SIZE, 0);
-			if (!ret)
-				goto nomem;
-		}
-		WARN_ON(bio->bi_vcnt != SCRUB_PAGES_PER_BIO);
 
 		if (i != SCRUB_BIOS_PER_DEV-1)
 			sdev->bios[i]->next_free = i + 1;
@@ -369,9 +342,6 @@ static int scrub_fixup_io(int rw, struct block_device *bdev, sector_t sector,
 	int ret;
 	DECLARE_COMPLETION_ONSTACK(complete);
 
-	/* we are going to wait on this IO */
-	rw |= REQ_SYNC;
-
 	bio = bio_alloc(GFP_NOFS, 1);
 	bio->bi_bdev = bdev;
 	bio->bi_sector = sector;
@@ -380,6 +350,7 @@ static int scrub_fixup_io(int rw, struct block_device *bdev, sector_t sector,
 	bio->bi_private = &complete;
 	submit_bio(rw, bio);
 
+	/* this will also unplug the queue */
 	wait_for_completion(&complete);
 
 	ret = !test_bit(BIO_UPTODATE, &bio->bi_flags);
@@ -394,6 +365,7 @@ static void scrub_bio_end_io(struct bio *bio, int err)
 	struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info;
 
 	sbio->err = err;
+	sbio->bio = bio;
 
 	btrfs_queue_worker(&fs_info->scrub_workers, &sbio->work);
 }
@@ -453,6 +425,8 @@ static void scrub_checksum(struct btrfs_work *work)
 	}
 
 out:
+	scrub_free_bio(sbio->bio);
+	sbio->bio = NULL;
 	spin_lock(&sdev->list_lock);
 	sbio->next_free = sdev->first_free;
 	sdev->first_free = sbio->index;
@@ -583,25 +557,50 @@ static int scrub_checksum_super(struct scrub_bio *sbio, void *buffer)
 static int scrub_submit(struct scrub_dev *sdev)
 {
 	struct scrub_bio *sbio;
+	struct bio *bio;
+	int i;
 
 	if (sdev->curr == -1)
 		return 0;
 
 	sbio = sdev->bios[sdev->curr];
 
-	sbio->bio->bi_sector = sbio->physical >> 9;
-	sbio->bio->bi_size = sbio->count * PAGE_SIZE;
-	sbio->bio->bi_next = NULL;
-	sbio->bio->bi_flags |= 1 << BIO_UPTODATE;
-	sbio->bio->bi_comp_cpu = -1;
-	sbio->bio->bi_bdev = sdev->dev->bdev;
+	bio = bio_alloc(GFP_NOFS, sbio->count);
+	if (!bio)
+		goto nomem;
+
+	bio->bi_private = sbio;
+	bio->bi_end_io = scrub_bio_end_io;
+	bio->bi_bdev = sdev->dev->bdev;
+	bio->bi_sector = sbio->physical >> 9;
+
+	for (i = 0; i < sbio->count; ++i) {
+		struct page *page;
+		int ret;
+
+		page = alloc_page(GFP_NOFS);
+		if (!page)
+			goto nomem;
+
+		ret = bio_add_page(bio, page, PAGE_SIZE, 0);
+		if (!ret) {
+			__free_page(page);
+			goto nomem;
+		}
+	}
+
 	sbio->err = 0;
 	sdev->curr = -1;
 	atomic_inc(&sdev->in_flight);
 
-	submit_bio(0, sbio->bio);
+	submit_bio(READ, bio);
 
 	return 0;
+
+nomem:
+	scrub_free_bio(bio);
+
+	return -ENOMEM;
 }
 
 static int scrub_page(struct scrub_dev *sdev, u64 logical, u64 len,
@@ -633,7 +632,11 @@ again:
 		sbio->logical = logical;
 	} else if (sbio->physical + sbio->count * PAGE_SIZE != physical ||
 		   sbio->logical + sbio->count * PAGE_SIZE != logical) {
-		scrub_submit(sdev);
+		int ret;
+
+		ret = scrub_submit(sdev);
+		if (ret)
+			return ret;
 		goto again;
 	}
 	sbio->spag[sbio->count].flags = flags;
@@ -645,8 +648,13 @@ again:
 		memcpy(sbio->spag[sbio->count].csum, csum, sdev->csum_size);
 	}
 	++sbio->count;
-	if (sbio->count == SCRUB_PAGES_PER_BIO || force)
-		scrub_submit(sdev);
+	if (sbio->count == SCRUB_PAGES_PER_BIO || force) {
+		int ret;
+
+		ret = scrub_submit(sdev);
+		if (ret)
+			return ret;
+	}
 
 	return 0;
 }
@@ -727,6 +735,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_dev *sdev,
 	struct btrfs_root *root = fs_info->extent_root;
 	struct btrfs_root *csum_root = fs_info->csum_root;
 	struct btrfs_extent_item *extent;
+	struct blk_plug plug;
 	u64 flags;
 	int ret;
 	int slot;
@@ -789,18 +798,12 @@ static noinline_for_stack int scrub_stripe(struct scrub_dev *sdev,
 
 		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 		if (ret < 0)
-			goto out;
-
-		l = path->nodes[0];
-		slot = path->slots[0];
-		btrfs_item_key_to_cpu(l, &key, slot);
-		if (key.objectid != logical) {
-			ret = btrfs_previous_item(root, path, 0,
-						  BTRFS_EXTENT_ITEM_KEY);
-			if (ret < 0)
-				goto out;
-		}
+			goto out_noplug;
 
+		/*
+		 * we might miss half an extent here, but that doesn't matter,
+		 * as it's only the prefetch
+		 */
 		while (1) {
 			l = path->nodes[0];
 			slot = path->slots[0];
@@ -809,7 +812,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_dev *sdev,
 				if (ret == 0)
 					continue;
 				if (ret < 0)
-					goto out;
+					goto out_noplug;
 
 				break;
 			}
@@ -831,6 +834,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_dev *sdev,
 	 * the scrub. This might currently (crc32) end up to be about 1MB
 	 */
 	start_stripe = 0;
+	blk_start_plug(&plug);
 again:
 	logical = base + offset + start_stripe * increment;
 	for (i = start_stripe; i < nstripes; ++i) {
@@ -890,15 +894,20 @@ again:
 		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 		if (ret < 0)
 			goto out;
-
-		l = path->nodes[0];
-		slot = path->slots[0];
-		btrfs_item_key_to_cpu(l, &key, slot);
-		if (key.objectid != logical) {
+		if (ret > 0) {
 			ret = btrfs_previous_item(root, path, 0,
 						  BTRFS_EXTENT_ITEM_KEY);
 			if (ret < 0)
 				goto out;
+			if (ret > 0) {
+				/* there's no smaller item, so stick with the
+				 * larger one */
+				btrfs_release_path(path);
+				ret = btrfs_search_slot(NULL, root, &key,
+							path, 0, 0);
+				if (ret < 0)
+					goto out;
+			}
 		}
 
 		while (1) {
@@ -972,6 +981,8 @@ next:
 	scrub_submit(sdev);
 
 out:
+	blk_finish_plug(&plug);
+out_noplug:
 	btrfs_free_path(path);
 	return ret < 0 ? ret : 0;
 }
@@ -1047,8 +1058,15 @@ int scrub_enumerate_chunks(struct scrub_dev *sdev, u64 start, u64 end)
 	while (1) {
 		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
 		if (ret < 0)
-			goto out;
-		ret = 0;
+			break;
+		if (ret > 0) {
+			if (path->slots[0] >=
+			    btrfs_header_nritems(path->nodes[0])) {
+				ret = btrfs_next_leaf(root, path);
+				if (ret)
+					break;
+			}
+		}
 
 		l = path->nodes[0];
 		slot = path->slots[0];
@@ -1058,7 +1076,7 @@ int scrub_enumerate_chunks(struct scrub_dev *sdev, u64 start, u64 end)
 		if (found_key.objectid != sdev->dev->devid)
 			break;
 
-		if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY)
+		if (btrfs_key_type(&found_key) != BTRFS_DEV_EXTENT_KEY)
 			break;
 
 		if (found_key.offset >= end)
@@ -1087,7 +1105,7 @@ int scrub_enumerate_chunks(struct scrub_dev *sdev, u64 start, u64 end)
 		cache = btrfs_lookup_block_group(fs_info, chunk_offset);
 		if (!cache) {
 			ret = -ENOENT;
-			goto out;
+			break;
 		}
 		ret = scrub_chunk(sdev, chunk_tree, chunk_objectid,
 				  chunk_offset, length);
@@ -1099,9 +1117,13 @@ int scrub_enumerate_chunks(struct scrub_dev *sdev, u64 start, u64 end)
 		btrfs_release_path(path);
 	}
 
-out:
 	btrfs_free_path(path);
-	return ret;
+
+	/*
+	 * ret can still be 1 from search_slot or next_leaf,
+	 * that's not an error
+	 */
+	return ret < 0 ? ret : 0;
 }
 
 static noinline_for_stack int scrub_supers(struct scrub_dev *sdev)
@@ -1138,8 +1160,12 @@ static noinline_for_stack int scrub_workers_get(struct btrfs_root *root)
 	struct btrfs_fs_info *fs_info = root->fs_info;
 
 	mutex_lock(&fs_info->scrub_lock);
-	if (fs_info->scrub_workers_refcnt == 0)
+	if (fs_info->scrub_workers_refcnt == 0) {
+		btrfs_init_workers(&fs_info->scrub_workers, "scrub",
+			   fs_info->thread_pool_size, &fs_info->generic_worker);
+		fs_info->scrub_workers.idle_thresh = 4;
 		btrfs_start_workers(&fs_info->scrub_workers, 1);
+	}
 	++fs_info->scrub_workers_refcnt;
 	mutex_unlock(&fs_info->scrub_lock);
 
@@ -1166,7 +1192,7 @@ int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end,
 	int ret;
 	struct btrfs_device *dev;
 
-	if (root->fs_info->closing)
+	if (btrfs_fs_closing(root->fs_info))
 		return -EINVAL;
 
 	/*
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 9b2e7e5bc3e..15634d4648d 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -161,7 +161,8 @@ enum {
 	Opt_compress_type, Opt_compress_force, Opt_compress_force_type,
 	Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard,
 	Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed,
-	Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, Opt_err,
+	Opt_enospc_debug, Opt_subvolrootid, Opt_defrag,
+	Opt_inode_cache, Opt_err,
 };
 
 static match_table_t tokens = {
@@ -193,6 +194,7 @@ static match_table_t tokens = {
 	{Opt_enospc_debug, "enospc_debug"},
 	{Opt_subvolrootid, "subvolrootid=%d"},
 	{Opt_defrag, "autodefrag"},
+	{Opt_inode_cache, "inode_cache"},
 	{Opt_err, NULL},
 };
 
@@ -361,6 +363,10 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			printk(KERN_INFO "btrfs: enabling disk space caching\n");
 			btrfs_set_opt(info->mount_opt, SPACE_CACHE);
 			break;
+		case Opt_inode_cache:
+			printk(KERN_INFO "btrfs: enabling inode map caching\n");
+			btrfs_set_opt(info->mount_opt, INODE_MAP_CACHE);
+			break;
 		case Opt_clear_cache:
 			printk(KERN_INFO "btrfs: force clearing of disk cache\n");
 			btrfs_set_opt(info->mount_opt, CLEAR_CACHE);
@@ -717,6 +723,12 @@ static int btrfs_show_options(struct seq_file *seq, struct vfsmount *vfs)
 		seq_puts(seq, ",clear_cache");
 	if (btrfs_test_opt(root, USER_SUBVOL_RM_ALLOWED))
 		seq_puts(seq, ",user_subvol_rm_allowed");
+	if (btrfs_test_opt(root, ENOSPC_DEBUG))
+		seq_puts(seq, ",enospc_debug");
+	if (btrfs_test_opt(root, AUTO_DEFRAG))
+		seq_puts(seq, ",autodefrag");
+	if (btrfs_test_opt(root, INODE_MAP_CACHE))
+		seq_puts(seq, ",inode_cache");
 	return 0;
 }
 
@@ -819,7 +831,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
 	} else {
 		char b[BDEVNAME_SIZE];
 
-		s->s_flags = flags;
+		s->s_flags = flags | MS_NOSEC;
 		strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
 		error = btrfs_fill_super(s, fs_devices, data,
 					 flags & MS_SILENT ? 1 : 0);
diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c
index c3c223ae669..daac9ae6d73 100644
--- a/fs/btrfs/sysfs.c
+++ b/fs/btrfs/sysfs.c
@@ -28,152 +28,6 @@
 #include "disk-io.h"
 #include "transaction.h"
 
-static ssize_t root_blocks_used_show(struct btrfs_root *root, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_root_used(&root->root_item));
-}
-
-static ssize_t root_block_limit_show(struct btrfs_root *root, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_root_limit(&root->root_item));
-}
-
-static ssize_t super_blocks_used_show(struct btrfs_fs_info *fs, char *buf)
-{
-
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_super_bytes_used(&fs->super_copy));
-}
-
-static ssize_t super_total_blocks_show(struct btrfs_fs_info *fs, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_super_total_bytes(&fs->super_copy));
-}
-
-static ssize_t super_blocksize_show(struct btrfs_fs_info *fs, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%llu\n",
-		(unsigned long long)btrfs_super_sectorsize(&fs->super_copy));
-}
-
-/* this is for root attrs (subvols/snapshots) */
-struct btrfs_root_attr {
-	struct attribute attr;
-	ssize_t (*show)(struct btrfs_root *, char *);
-	ssize_t (*store)(struct btrfs_root *, const char *, size_t);
-};
-
-#define ROOT_ATTR(name, mode, show, store) \
-static struct btrfs_root_attr btrfs_root_attr_##name = __ATTR(name, mode, \
-							      show, store)
-
-ROOT_ATTR(blocks_used,	0444,	root_blocks_used_show,	NULL);
-ROOT_ATTR(block_limit,	0644,	root_block_limit_show,	NULL);
-
-static struct attribute *btrfs_root_attrs[] = {
-	&btrfs_root_attr_blocks_used.attr,
-	&btrfs_root_attr_block_limit.attr,
-	NULL,
-};
-
-/* this is for super attrs (actual full fs) */
-struct btrfs_super_attr {
-	struct attribute attr;
-	ssize_t (*show)(struct btrfs_fs_info *, char *);
-	ssize_t (*store)(struct btrfs_fs_info *, const char *, size_t);
-};
-
-#define SUPER_ATTR(name, mode, show, store) \
-static struct btrfs_super_attr btrfs_super_attr_##name = __ATTR(name, mode, \
-								show, store)
-
-SUPER_ATTR(blocks_used,		0444,	super_blocks_used_show,		NULL);
-SUPER_ATTR(total_blocks,	0444,	super_total_blocks_show,	NULL);
-SUPER_ATTR(blocksize,		0444,	super_blocksize_show,		NULL);
-
-static struct attribute *btrfs_super_attrs[] = {
-	&btrfs_super_attr_blocks_used.attr,
-	&btrfs_super_attr_total_blocks.attr,
-	&btrfs_super_attr_blocksize.attr,
-	NULL,
-};
-
-static ssize_t btrfs_super_attr_show(struct kobject *kobj,
-				    struct attribute *attr, char *buf)
-{
-	struct btrfs_fs_info *fs = container_of(kobj, struct btrfs_fs_info,
-						super_kobj);
-	struct btrfs_super_attr *a = container_of(attr,
-						  struct btrfs_super_attr,
-						  attr);
-
-	return a->show ? a->show(fs, buf) : 0;
-}
-
-static ssize_t btrfs_super_attr_store(struct kobject *kobj,
-				     struct attribute *attr,
-				     const char *buf, size_t len)
-{
-	struct btrfs_fs_info *fs = container_of(kobj, struct btrfs_fs_info,
-						super_kobj);
-	struct btrfs_super_attr *a = container_of(attr,
-						  struct btrfs_super_attr,
-						  attr);
-
-	return a->store ? a->store(fs, buf, len) : 0;
-}
-
-static ssize_t btrfs_root_attr_show(struct kobject *kobj,
-				    struct attribute *attr, char *buf)
-{
-	struct btrfs_root *root = container_of(kobj, struct btrfs_root,
-						root_kobj);
-	struct btrfs_root_attr *a = container_of(attr,
-						 struct btrfs_root_attr,
-						 attr);
-
-	return a->show ? a->show(root, buf) : 0;
-}
-
-static ssize_t btrfs_root_attr_store(struct kobject *kobj,
-				     struct attribute *attr,
-				     const char *buf, size_t len)
-{
-	struct btrfs_root *root = container_of(kobj, struct btrfs_root,
-						root_kobj);
-	struct btrfs_root_attr *a = container_of(attr,
-						 struct btrfs_root_attr,
-						 attr);
-	return a->store ? a->store(root, buf, len) : 0;
-}
-
-static void btrfs_super_release(struct kobject *kobj)
-{
-	struct btrfs_fs_info *fs = container_of(kobj, struct btrfs_fs_info,
-						super_kobj);
-	complete(&fs->kobj_unregister);
-}
-
-static void btrfs_root_release(struct kobject *kobj)
-{
-	struct btrfs_root *root = container_of(kobj, struct btrfs_root,
-						root_kobj);
-	complete(&root->kobj_unregister);
-}
-
-static const struct sysfs_ops btrfs_super_attr_ops = {
-	.show	= btrfs_super_attr_show,
-	.store	= btrfs_super_attr_store,
-};
-
-static const struct sysfs_ops btrfs_root_attr_ops = {
-	.show	= btrfs_root_attr_show,
-	.store	= btrfs_root_attr_store,
-};
-
 /* /sys/fs/btrfs/ entry */
 static struct kset *btrfs_kset;
 
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index dc80f715692..51dcec86757 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -35,6 +35,7 @@ static noinline void put_transaction(struct btrfs_transaction *transaction)
 {
 	WARN_ON(atomic_read(&transaction->use_count) == 0);
 	if (atomic_dec_and_test(&transaction->use_count)) {
+		BUG_ON(!list_empty(&transaction->list));
 		memset(transaction, 0, sizeof(*transaction));
 		kmem_cache_free(btrfs_transaction_cachep, transaction);
 	}
@@ -49,46 +50,72 @@ static noinline void switch_commit_root(struct btrfs_root *root)
 /*
  * either allocate a new transaction or hop into the existing one
  */
-static noinline int join_transaction(struct btrfs_root *root)
+static noinline int join_transaction(struct btrfs_root *root, int nofail)
 {
 	struct btrfs_transaction *cur_trans;
+
+	spin_lock(&root->fs_info->trans_lock);
+	if (root->fs_info->trans_no_join) {
+		if (!nofail) {
+			spin_unlock(&root->fs_info->trans_lock);
+			return -EBUSY;
+		}
+	}
+
 	cur_trans = root->fs_info->running_transaction;
-	if (!cur_trans) {
-		cur_trans = kmem_cache_alloc(btrfs_transaction_cachep,
-					     GFP_NOFS);
-		if (!cur_trans)
-			return -ENOMEM;
-		root->fs_info->generation++;
-		atomic_set(&cur_trans->num_writers, 1);
-		cur_trans->num_joined = 0;
-		cur_trans->transid = root->fs_info->generation;
-		init_waitqueue_head(&cur_trans->writer_wait);
-		init_waitqueue_head(&cur_trans->commit_wait);
-		cur_trans->in_commit = 0;
-		cur_trans->blocked = 0;
-		atomic_set(&cur_trans->use_count, 1);
-		cur_trans->commit_done = 0;
-		cur_trans->start_time = get_seconds();
-
-		cur_trans->delayed_refs.root = RB_ROOT;
-		cur_trans->delayed_refs.num_entries = 0;
-		cur_trans->delayed_refs.num_heads_ready = 0;
-		cur_trans->delayed_refs.num_heads = 0;
-		cur_trans->delayed_refs.flushing = 0;
-		cur_trans->delayed_refs.run_delayed_start = 0;
-		spin_lock_init(&cur_trans->delayed_refs.lock);
-
-		INIT_LIST_HEAD(&cur_trans->pending_snapshots);
-		list_add_tail(&cur_trans->list, &root->fs_info->trans_list);
-		extent_io_tree_init(&cur_trans->dirty_pages,
-				     root->fs_info->btree_inode->i_mapping);
-		spin_lock(&root->fs_info->new_trans_lock);
-		root->fs_info->running_transaction = cur_trans;
-		spin_unlock(&root->fs_info->new_trans_lock);
-	} else {
+	if (cur_trans) {
+		atomic_inc(&cur_trans->use_count);
+		atomic_inc(&cur_trans->num_writers);
+		cur_trans->num_joined++;
+		spin_unlock(&root->fs_info->trans_lock);
+		return 0;
+	}
+	spin_unlock(&root->fs_info->trans_lock);
+
+	cur_trans = kmem_cache_alloc(btrfs_transaction_cachep, GFP_NOFS);
+	if (!cur_trans)
+		return -ENOMEM;
+	spin_lock(&root->fs_info->trans_lock);
+	if (root->fs_info->running_transaction) {
+		kmem_cache_free(btrfs_transaction_cachep, cur_trans);
+		cur_trans = root->fs_info->running_transaction;
+		atomic_inc(&cur_trans->use_count);
 		atomic_inc(&cur_trans->num_writers);
 		cur_trans->num_joined++;
+		spin_unlock(&root->fs_info->trans_lock);
+		return 0;
 	}
+	atomic_set(&cur_trans->num_writers, 1);
+	cur_trans->num_joined = 0;
+	init_waitqueue_head(&cur_trans->writer_wait);
+	init_waitqueue_head(&cur_trans->commit_wait);
+	cur_trans->in_commit = 0;
+	cur_trans->blocked = 0;
+	/*
+	 * One for this trans handle, one so it will live on until we
+	 * commit the transaction.
+	 */
+	atomic_set(&cur_trans->use_count, 2);
+	cur_trans->commit_done = 0;
+	cur_trans->start_time = get_seconds();
+
+	cur_trans->delayed_refs.root = RB_ROOT;
+	cur_trans->delayed_refs.num_entries = 0;
+	cur_trans->delayed_refs.num_heads_ready = 0;
+	cur_trans->delayed_refs.num_heads = 0;
+	cur_trans->delayed_refs.flushing = 0;
+	cur_trans->delayed_refs.run_delayed_start = 0;
+	spin_lock_init(&cur_trans->commit_lock);
+	spin_lock_init(&cur_trans->delayed_refs.lock);
+
+	INIT_LIST_HEAD(&cur_trans->pending_snapshots);
+	list_add_tail(&cur_trans->list, &root->fs_info->trans_list);
+	extent_io_tree_init(&cur_trans->dirty_pages,
+			     root->fs_info->btree_inode->i_mapping);
+	root->fs_info->generation++;
+	cur_trans->transid = root->fs_info->generation;
+	root->fs_info->running_transaction = cur_trans;
+	spin_unlock(&root->fs_info->trans_lock);
 
 	return 0;
 }
@@ -99,36 +126,82 @@ static noinline int join_transaction(struct btrfs_root *root)
  * to make sure the old root from before we joined the transaction is deleted
  * when the transaction commits
  */
-static noinline int record_root_in_trans(struct btrfs_trans_handle *trans,
-					 struct btrfs_root *root)
+static int record_root_in_trans(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *root)
 {
 	if (root->ref_cows && root->last_trans < trans->transid) {
 		WARN_ON(root == root->fs_info->extent_root);
 		WARN_ON(root->commit_root != root->node);
 
+		/*
+		 * see below for in_trans_setup usage rules
+		 * we have the reloc mutex held now, so there
+		 * is only one writer in this function
+		 */
+		root->in_trans_setup = 1;
+
+		/* make sure readers find in_trans_setup before
+		 * they find our root->last_trans update
+		 */
+		smp_wmb();
+
+		spin_lock(&root->fs_info->fs_roots_radix_lock);
+		if (root->last_trans == trans->transid) {
+			spin_unlock(&root->fs_info->fs_roots_radix_lock);
+			return 0;
+		}
 		radix_tree_tag_set(&root->fs_info->fs_roots_radix,
 			   (unsigned long)root->root_key.objectid,
 			   BTRFS_ROOT_TRANS_TAG);
+		spin_unlock(&root->fs_info->fs_roots_radix_lock);
 		root->last_trans = trans->transid;
+
+		/* this is pretty tricky.  We don't want to
+		 * take the relocation lock in btrfs_record_root_in_trans
+		 * unless we're really doing the first setup for this root in
+		 * this transaction.
+		 *
+		 * Normally we'd use root->last_trans as a flag to decide
+		 * if we want to take the expensive mutex.
+		 *
+		 * But, we have to set root->last_trans before we
+		 * init the relocation root, otherwise, we trip over warnings
+		 * in ctree.c.  The solution used here is to flag ourselves
+		 * with root->in_trans_setup.  When this is 1, we're still
+		 * fixing up the reloc trees and everyone must wait.
+		 *
+		 * When this is zero, they can trust root->last_trans and fly
+		 * through btrfs_record_root_in_trans without having to take the
+		 * lock.  smp_wmb() makes sure that all the writes above are
+		 * done before we pop in the zero below
+		 */
 		btrfs_init_reloc_root(trans, root);
+		smp_wmb();
+		root->in_trans_setup = 0;
 	}
 	return 0;
 }
 
+
 int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root)
 {
 	if (!root->ref_cows)
 		return 0;
 
-	mutex_lock(&root->fs_info->trans_mutex);
-	if (root->last_trans == trans->transid) {
-		mutex_unlock(&root->fs_info->trans_mutex);
+	/*
+	 * see record_root_in_trans for comments about in_trans_setup usage
+	 * and barriers
+	 */
+	smp_rmb();
+	if (root->last_trans == trans->transid &&
+	    !root->in_trans_setup)
 		return 0;
-	}
 
+	mutex_lock(&root->fs_info->reloc_mutex);
 	record_root_in_trans(trans, root);
-	mutex_unlock(&root->fs_info->trans_mutex);
+	mutex_unlock(&root->fs_info->reloc_mutex);
+
 	return 0;
 }
 
@@ -140,21 +213,23 @@ static void wait_current_trans(struct btrfs_root *root)
 {
 	struct btrfs_transaction *cur_trans;
 
+	spin_lock(&root->fs_info->trans_lock);
 	cur_trans = root->fs_info->running_transaction;
 	if (cur_trans && cur_trans->blocked) {
 		DEFINE_WAIT(wait);
 		atomic_inc(&cur_trans->use_count);
+		spin_unlock(&root->fs_info->trans_lock);
 		while (1) {
 			prepare_to_wait(&root->fs_info->transaction_wait, &wait,
 					TASK_UNINTERRUPTIBLE);
 			if (!cur_trans->blocked)
 				break;
-			mutex_unlock(&root->fs_info->trans_mutex);
 			schedule();
-			mutex_lock(&root->fs_info->trans_mutex);
 		}
 		finish_wait(&root->fs_info->transaction_wait, &wait);
 		put_transaction(cur_trans);
+	} else {
+		spin_unlock(&root->fs_info->trans_lock);
 	}
 }
 
@@ -167,10 +242,16 @@ enum btrfs_trans_type {
 
 static int may_wait_transaction(struct btrfs_root *root, int type)
 {
-	if (!root->fs_info->log_root_recovering &&
-	    ((type == TRANS_START && !root->fs_info->open_ioctl_trans) ||
-	     type == TRANS_USERSPACE))
+	if (root->fs_info->log_root_recovering)
+		return 0;
+
+	if (type == TRANS_USERSPACE)
 		return 1;
+
+	if (type == TRANS_START &&
+	    !atomic_read(&root->fs_info->open_ioctl_trans))
+		return 1;
+
 	return 0;
 }
 
@@ -184,36 +265,44 @@ static struct btrfs_trans_handle *start_transaction(struct btrfs_root *root,
 
 	if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR)
 		return ERR_PTR(-EROFS);
+
+	if (current->journal_info) {
+		WARN_ON(type != TRANS_JOIN && type != TRANS_JOIN_NOLOCK);
+		h = current->journal_info;
+		h->use_count++;
+		h->orig_rsv = h->block_rsv;
+		h->block_rsv = NULL;
+		goto got_it;
+	}
 again:
 	h = kmem_cache_alloc(btrfs_trans_handle_cachep, GFP_NOFS);
 	if (!h)
 		return ERR_PTR(-ENOMEM);
 
-	if (type != TRANS_JOIN_NOLOCK)
-		mutex_lock(&root->fs_info->trans_mutex);
 	if (may_wait_transaction(root, type))
 		wait_current_trans(root);
 
-	ret = join_transaction(root);
+	do {
+		ret = join_transaction(root, type == TRANS_JOIN_NOLOCK);
+		if (ret == -EBUSY)
+			wait_current_trans(root);
+	} while (ret == -EBUSY);
+
 	if (ret < 0) {
 		kmem_cache_free(btrfs_trans_handle_cachep, h);
-		if (type != TRANS_JOIN_NOLOCK)
-			mutex_unlock(&root->fs_info->trans_mutex);
 		return ERR_PTR(ret);
 	}
 
 	cur_trans = root->fs_info->running_transaction;
-	atomic_inc(&cur_trans->use_count);
-	if (type != TRANS_JOIN_NOLOCK)
-		mutex_unlock(&root->fs_info->trans_mutex);
 
 	h->transid = cur_trans->transid;
 	h->transaction = cur_trans;
 	h->blocks_used = 0;
-	h->block_group = 0;
 	h->bytes_reserved = 0;
 	h->delayed_ref_updates = 0;
+	h->use_count = 1;
 	h->block_rsv = NULL;
+	h->orig_rsv = NULL;
 
 	smp_mb();
 	if (cur_trans->blocked && may_wait_transaction(root, type)) {
@@ -241,11 +330,8 @@ again:
 		}
 	}
 
-	if (type != TRANS_JOIN_NOLOCK)
-		mutex_lock(&root->fs_info->trans_mutex);
-	record_root_in_trans(h, root);
-	if (type != TRANS_JOIN_NOLOCK)
-		mutex_unlock(&root->fs_info->trans_mutex);
+got_it:
+	btrfs_record_root_in_trans(h, root);
 
 	if (!current->journal_info && type != TRANS_USERSPACE)
 		current->journal_info = h;
@@ -257,22 +343,19 @@ struct btrfs_trans_handle *btrfs_start_transaction(struct btrfs_root *root,
 {
 	return start_transaction(root, num_items, TRANS_START);
 }
-struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root,
-						   int num_blocks)
+struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root)
 {
 	return start_transaction(root, 0, TRANS_JOIN);
 }
 
-struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root,
-							  int num_blocks)
+struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root)
 {
 	return start_transaction(root, 0, TRANS_JOIN_NOLOCK);
 }
 
-struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *r,
-							 int num_blocks)
+struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *root)
 {
-	return start_transaction(r, 0, TRANS_USERSPACE);
+	return start_transaction(root, 0, TRANS_USERSPACE);
 }
 
 /* wait for a transaction commit to be fully complete */
@@ -280,17 +363,13 @@ static noinline int wait_for_commit(struct btrfs_root *root,
 				    struct btrfs_transaction *commit)
 {
 	DEFINE_WAIT(wait);
-	mutex_lock(&root->fs_info->trans_mutex);
 	while (!commit->commit_done) {
 		prepare_to_wait(&commit->commit_wait, &wait,
 				TASK_UNINTERRUPTIBLE);
 		if (commit->commit_done)
 			break;
-		mutex_unlock(&root->fs_info->trans_mutex);
 		schedule();
-		mutex_lock(&root->fs_info->trans_mutex);
 	}
-	mutex_unlock(&root->fs_info->trans_mutex);
 	finish_wait(&commit->commit_wait, &wait);
 	return 0;
 }
@@ -300,59 +379,56 @@ int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid)
 	struct btrfs_transaction *cur_trans = NULL, *t;
 	int ret;
 
-	mutex_lock(&root->fs_info->trans_mutex);
-
 	ret = 0;
 	if (transid) {
 		if (transid <= root->fs_info->last_trans_committed)
-			goto out_unlock;
+			goto out;
 
 		/* find specified transaction */
+		spin_lock(&root->fs_info->trans_lock);
 		list_for_each_entry(t, &root->fs_info->trans_list, list) {
 			if (t->transid == transid) {
 				cur_trans = t;
+				atomic_inc(&cur_trans->use_count);
 				break;
 			}
 			if (t->transid > transid)
 				break;
 		}
+		spin_unlock(&root->fs_info->trans_lock);
 		ret = -EINVAL;
 		if (!cur_trans)
-			goto out_unlock;  /* bad transid */
+			goto out;  /* bad transid */
 	} else {
 		/* find newest transaction that is committing | committed */
+		spin_lock(&root->fs_info->trans_lock);
 		list_for_each_entry_reverse(t, &root->fs_info->trans_list,
 					    list) {
 			if (t->in_commit) {
 				if (t->commit_done)
-					goto out_unlock;
+					break;
 				cur_trans = t;
+				atomic_inc(&cur_trans->use_count);
 				break;
 			}
 		}
+		spin_unlock(&root->fs_info->trans_lock);
 		if (!cur_trans)
-			goto out_unlock;  /* nothing committing|committed */
+			goto out;  /* nothing committing|committed */
 	}
 
-	atomic_inc(&cur_trans->use_count);
-	mutex_unlock(&root->fs_info->trans_mutex);
-
 	wait_for_commit(root, cur_trans);
 
-	mutex_lock(&root->fs_info->trans_mutex);
 	put_transaction(cur_trans);
 	ret = 0;
-out_unlock:
-	mutex_unlock(&root->fs_info->trans_mutex);
+out:
 	return ret;
 }
 
 void btrfs_throttle(struct btrfs_root *root)
 {
-	mutex_lock(&root->fs_info->trans_mutex);
-	if (!root->fs_info->open_ioctl_trans)
+	if (!atomic_read(&root->fs_info->open_ioctl_trans))
 		wait_current_trans(root);
-	mutex_unlock(&root->fs_info->trans_mutex);
 }
 
 static int should_end_transaction(struct btrfs_trans_handle *trans,
@@ -370,6 +446,7 @@ int btrfs_should_end_transaction(struct btrfs_trans_handle *trans,
 	struct btrfs_transaction *cur_trans = trans->transaction;
 	int updates;
 
+	smp_mb();
 	if (cur_trans->blocked || cur_trans->delayed_refs.flushing)
 		return 1;
 
@@ -388,6 +465,11 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 	struct btrfs_fs_info *info = root->fs_info;
 	int count = 0;
 
+	if (--trans->use_count) {
+		trans->block_rsv = trans->orig_rsv;
+		return 0;
+	}
+
 	while (count < 4) {
 		unsigned long cur = trans->delayed_ref_updates;
 		trans->delayed_ref_updates = 0;
@@ -410,9 +492,11 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 
 	btrfs_trans_release_metadata(trans, root);
 
-	if (lock && !root->fs_info->open_ioctl_trans &&
-	    should_end_transaction(trans, root))
+	if (lock && !atomic_read(&root->fs_info->open_ioctl_trans) &&
+	    should_end_transaction(trans, root)) {
 		trans->transaction->blocked = 1;
+		smp_wmb();
+	}
 
 	if (lock && cur_trans->blocked && !cur_trans->in_commit) {
 		if (throttle)
@@ -703,9 +787,9 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans,
  */
 int btrfs_add_dead_root(struct btrfs_root *root)
 {
-	mutex_lock(&root->fs_info->trans_mutex);
+	spin_lock(&root->fs_info->trans_lock);
 	list_add(&root->root_list, &root->fs_info->dead_roots);
-	mutex_unlock(&root->fs_info->trans_mutex);
+	spin_unlock(&root->fs_info->trans_lock);
 	return 0;
 }
 
@@ -721,6 +805,7 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans,
 	int ret;
 	int err = 0;
 
+	spin_lock(&fs_info->fs_roots_radix_lock);
 	while (1) {
 		ret = radix_tree_gang_lookup_tag(&fs_info->fs_roots_radix,
 						 (void **)gang, 0,
@@ -733,6 +818,7 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans,
 			radix_tree_tag_clear(&fs_info->fs_roots_radix,
 					(unsigned long)root->root_key.objectid,
 					BTRFS_ROOT_TRANS_TAG);
+			spin_unlock(&fs_info->fs_roots_radix_lock);
 
 			btrfs_free_log(trans, root);
 			btrfs_update_reloc_root(trans, root);
@@ -753,10 +839,12 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans,
 			err = btrfs_update_root(trans, fs_info->tree_root,
 						&root->root_key,
 						&root->root_item);
+			spin_lock(&fs_info->fs_roots_radix_lock);
 			if (err)
 				break;
 		}
 	}
+	spin_unlock(&fs_info->fs_roots_radix_lock);
 	return err;
 }
 
@@ -786,7 +874,7 @@ int btrfs_defrag_root(struct btrfs_root *root, int cacheonly)
 		btrfs_btree_balance_dirty(info->tree_root, nr);
 		cond_resched();
 
-		if (root->fs_info->closing || ret != -EAGAIN)
+		if (btrfs_fs_closing(root->fs_info) || ret != -EAGAIN)
 			break;
 	}
 	root->defrag_running = 0;
@@ -869,6 +957,15 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	ret = btrfs_update_inode(trans, parent_root, parent_inode);
 	BUG_ON(ret);
 
+	/*
+	 * pull in the delayed directory update
+	 * and the delayed inode item
+	 * otherwise we corrupt the FS during
+	 * snapshot
+	 */
+	ret = btrfs_run_delayed_items(trans, root);
+	BUG_ON(ret);
+
 	record_root_in_trans(trans, root);
 	btrfs_set_root_last_snapshot(&root->root_item, trans->transid);
 	memcpy(new_root_item, &root->root_item, sizeof(*new_root_item));
@@ -930,14 +1027,6 @@ static noinline int create_pending_snapshots(struct btrfs_trans_handle *trans,
 	int ret;
 
 	list_for_each_entry(pending, head, list) {
-		/*
-		 * We must deal with the delayed items before creating
-		 * snapshots, or we will create a snapthot with inconsistent
-		 * information.
-		*/
-		ret = btrfs_run_delayed_items(trans, fs_info->fs_root);
-		BUG_ON(ret);
-
 		ret = create_pending_snapshot(trans, fs_info, pending);
 		BUG_ON(ret);
 	}
@@ -967,20 +1056,20 @@ static void update_super_roots(struct btrfs_root *root)
 int btrfs_transaction_in_commit(struct btrfs_fs_info *info)
 {
 	int ret = 0;
-	spin_lock(&info->new_trans_lock);
+	spin_lock(&info->trans_lock);
 	if (info->running_transaction)
 		ret = info->running_transaction->in_commit;
-	spin_unlock(&info->new_trans_lock);
+	spin_unlock(&info->trans_lock);
 	return ret;
 }
 
 int btrfs_transaction_blocked(struct btrfs_fs_info *info)
 {
 	int ret = 0;
-	spin_lock(&info->new_trans_lock);
+	spin_lock(&info->trans_lock);
 	if (info->running_transaction)
 		ret = info->running_transaction->blocked;
-	spin_unlock(&info->new_trans_lock);
+	spin_unlock(&info->trans_lock);
 	return ret;
 }
 
@@ -1004,9 +1093,7 @@ static void wait_current_trans_commit_start(struct btrfs_root *root,
 				    &wait);
 			break;
 		}
-		mutex_unlock(&root->fs_info->trans_mutex);
 		schedule();
-		mutex_lock(&root->fs_info->trans_mutex);
 		finish_wait(&root->fs_info->transaction_blocked_wait, &wait);
 	}
 }
@@ -1032,9 +1119,7 @@ static void wait_current_trans_commit_start_and_unblock(struct btrfs_root *root,
 				    &wait);
 			break;
 		}
-		mutex_unlock(&root->fs_info->trans_mutex);
 		schedule();
-		mutex_lock(&root->fs_info->trans_mutex);
 		finish_wait(&root->fs_info->transaction_wait,
 			    &wait);
 	}
@@ -1072,7 +1157,7 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans,
 
 	INIT_DELAYED_WORK(&ac->work, do_async_commit);
 	ac->root = root;
-	ac->newtrans = btrfs_join_transaction(root, 0);
+	ac->newtrans = btrfs_join_transaction(root);
 	if (IS_ERR(ac->newtrans)) {
 		int err = PTR_ERR(ac->newtrans);
 		kfree(ac);
@@ -1080,23 +1165,22 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans,
 	}
 
 	/* take transaction reference */
-	mutex_lock(&root->fs_info->trans_mutex);
 	cur_trans = trans->transaction;
 	atomic_inc(&cur_trans->use_count);
-	mutex_unlock(&root->fs_info->trans_mutex);
 
 	btrfs_end_transaction(trans, root);
 	schedule_delayed_work(&ac->work, 0);
 
 	/* wait for transaction to start and unblock */
-	mutex_lock(&root->fs_info->trans_mutex);
 	if (wait_for_unblock)
 		wait_current_trans_commit_start_and_unblock(root, cur_trans);
 	else
 		wait_current_trans_commit_start(root, cur_trans);
-	put_transaction(cur_trans);
-	mutex_unlock(&root->fs_info->trans_mutex);
 
+	if (current->journal_info == trans)
+		current->journal_info = NULL;
+
+	put_transaction(cur_trans);
 	return 0;
 }
 
@@ -1139,38 +1223,41 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	ret = btrfs_run_delayed_refs(trans, root, 0);
 	BUG_ON(ret);
 
-	mutex_lock(&root->fs_info->trans_mutex);
+	spin_lock(&cur_trans->commit_lock);
 	if (cur_trans->in_commit) {
+		spin_unlock(&cur_trans->commit_lock);
 		atomic_inc(&cur_trans->use_count);
-		mutex_unlock(&root->fs_info->trans_mutex);
 		btrfs_end_transaction(trans, root);
 
 		ret = wait_for_commit(root, cur_trans);
 		BUG_ON(ret);
 
-		mutex_lock(&root->fs_info->trans_mutex);
 		put_transaction(cur_trans);
-		mutex_unlock(&root->fs_info->trans_mutex);
 
 		return 0;
 	}
 
 	trans->transaction->in_commit = 1;
 	trans->transaction->blocked = 1;
+	spin_unlock(&cur_trans->commit_lock);
 	wake_up(&root->fs_info->transaction_blocked_wait);
 
+	spin_lock(&root->fs_info->trans_lock);
 	if (cur_trans->list.prev != &root->fs_info->trans_list) {
 		prev_trans = list_entry(cur_trans->list.prev,
 					struct btrfs_transaction, list);
 		if (!prev_trans->commit_done) {
 			atomic_inc(&prev_trans->use_count);
-			mutex_unlock(&root->fs_info->trans_mutex);
+			spin_unlock(&root->fs_info->trans_lock);
 
 			wait_for_commit(root, prev_trans);
 
-			mutex_lock(&root->fs_info->trans_mutex);
 			put_transaction(prev_trans);
+		} else {
+			spin_unlock(&root->fs_info->trans_lock);
 		}
+	} else {
+		spin_unlock(&root->fs_info->trans_lock);
 	}
 
 	if (now < cur_trans->start_time || now - cur_trans->start_time < 1)
@@ -1178,12 +1265,12 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	do {
 		int snap_pending = 0;
+
 		joined = cur_trans->num_joined;
 		if (!list_empty(&trans->transaction->pending_snapshots))
 			snap_pending = 1;
 
 		WARN_ON(cur_trans != trans->transaction);
-		mutex_unlock(&root->fs_info->trans_mutex);
 
 		if (flush_on_commit || snap_pending) {
 			btrfs_start_delalloc_inodes(root, 1);
@@ -1206,26 +1293,48 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 		prepare_to_wait(&cur_trans->writer_wait, &wait,
 				TASK_UNINTERRUPTIBLE);
 
-		smp_mb();
 		if (atomic_read(&cur_trans->num_writers) > 1)
 			schedule_timeout(MAX_SCHEDULE_TIMEOUT);
 		else if (should_grow)
 			schedule_timeout(1);
 
-		mutex_lock(&root->fs_info->trans_mutex);
 		finish_wait(&cur_trans->writer_wait, &wait);
 	} while (atomic_read(&cur_trans->num_writers) > 1 ||
 		 (should_grow && cur_trans->num_joined != joined));
 
-	ret = create_pending_snapshots(trans, root->fs_info);
-	BUG_ON(ret);
+	/*
+	 * Ok now we need to make sure to block out any other joins while we
+	 * commit the transaction.  We could have started a join before setting
+	 * no_join so make sure to wait for num_writers to == 1 again.
+	 */
+	spin_lock(&root->fs_info->trans_lock);
+	root->fs_info->trans_no_join = 1;
+	spin_unlock(&root->fs_info->trans_lock);
+	wait_event(cur_trans->writer_wait,
+		   atomic_read(&cur_trans->num_writers) == 1);
+
+	/*
+	 * the reloc mutex makes sure that we stop
+	 * the balancing code from coming in and moving
+	 * extents around in the middle of the commit
+	 */
+	mutex_lock(&root->fs_info->reloc_mutex);
 
 	ret = btrfs_run_delayed_items(trans, root);
 	BUG_ON(ret);
 
+	ret = create_pending_snapshots(trans, root->fs_info);
+	BUG_ON(ret);
+
 	ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1);
 	BUG_ON(ret);
 
+	/*
+	 * make sure none of the code above managed to slip in a
+	 * delayed item
+	 */
+	btrfs_assert_delayed_root_empty(root);
+
 	WARN_ON(cur_trans != trans->transaction);
 
 	btrfs_scrub_pause(root);
@@ -1258,9 +1367,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	btrfs_prepare_extent_commit(trans, root);
 
 	cur_trans = root->fs_info->running_transaction;
-	spin_lock(&root->fs_info->new_trans_lock);
-	root->fs_info->running_transaction = NULL;
-	spin_unlock(&root->fs_info->new_trans_lock);
 
 	btrfs_set_root_node(&root->fs_info->tree_root->root_item,
 			    root->fs_info->tree_root->node);
@@ -1281,10 +1387,14 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	       sizeof(root->fs_info->super_copy));
 
 	trans->transaction->blocked = 0;
+	spin_lock(&root->fs_info->trans_lock);
+	root->fs_info->running_transaction = NULL;
+	root->fs_info->trans_no_join = 0;
+	spin_unlock(&root->fs_info->trans_lock);
+	mutex_unlock(&root->fs_info->reloc_mutex);
 
 	wake_up(&root->fs_info->transaction_wait);
 
-	mutex_unlock(&root->fs_info->trans_mutex);
 	ret = btrfs_write_and_wait_transaction(trans, root);
 	BUG_ON(ret);
 	write_ctree_super(trans, root, 0);
@@ -1297,22 +1407,21 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	btrfs_finish_extent_commit(trans, root);
 
-	mutex_lock(&root->fs_info->trans_mutex);
-
 	cur_trans->commit_done = 1;
 
 	root->fs_info->last_trans_committed = cur_trans->transid;
 
 	wake_up(&cur_trans->commit_wait);
 
+	spin_lock(&root->fs_info->trans_lock);
 	list_del_init(&cur_trans->list);
+	spin_unlock(&root->fs_info->trans_lock);
+
 	put_transaction(cur_trans);
 	put_transaction(cur_trans);
 
 	trace_btrfs_transaction_commit(root);
 
-	mutex_unlock(&root->fs_info->trans_mutex);
-
 	btrfs_scrub_continue(root);
 
 	if (current->journal_info == trans)
@@ -1334,9 +1443,9 @@ int btrfs_clean_old_snapshots(struct btrfs_root *root)
 	LIST_HEAD(list);
 	struct btrfs_fs_info *fs_info = root->fs_info;
 
-	mutex_lock(&fs_info->trans_mutex);
+	spin_lock(&fs_info->trans_lock);
 	list_splice_init(&fs_info->dead_roots, &list);
-	mutex_unlock(&fs_info->trans_mutex);
+	spin_unlock(&fs_info->trans_lock);
 
 	while (!list_empty(&list)) {
 		root = list_entry(list.next, struct btrfs_root, root_list);
diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h
index 804c88639e5..02564e6230a 100644
--- a/fs/btrfs/transaction.h
+++ b/fs/btrfs/transaction.h
@@ -28,10 +28,12 @@ struct btrfs_transaction {
 	 * transaction can end
 	 */
 	atomic_t num_writers;
+	atomic_t use_count;
 
 	unsigned long num_joined;
+
+	spinlock_t commit_lock;
 	int in_commit;
-	atomic_t use_count;
 	int commit_done;
 	int blocked;
 	struct list_head list;
@@ -45,13 +47,14 @@ struct btrfs_transaction {
 
 struct btrfs_trans_handle {
 	u64 transid;
-	u64 block_group;
 	u64 bytes_reserved;
+	unsigned long use_count;
 	unsigned long blocks_reserved;
 	unsigned long blocks_used;
 	unsigned long delayed_ref_updates;
 	struct btrfs_transaction *transaction;
 	struct btrfs_block_rsv *block_rsv;
+	struct btrfs_block_rsv *orig_rsv;
 };
 
 struct btrfs_pending_snapshot {
@@ -66,19 +69,6 @@ struct btrfs_pending_snapshot {
 	struct list_head list;
 };
 
-static inline void btrfs_set_trans_block_group(struct btrfs_trans_handle *trans,
-					       struct inode *inode)
-{
-	trans->block_group = BTRFS_I(inode)->block_group;
-}
-
-static inline void btrfs_update_inode_block_group(
-					  struct btrfs_trans_handle *trans,
-					  struct inode *inode)
-{
-	BTRFS_I(inode)->block_group = trans->block_group;
-}
-
 static inline void btrfs_set_inode_last_trans(struct btrfs_trans_handle *trans,
 					      struct inode *inode)
 {
@@ -92,12 +82,9 @@ int btrfs_end_transaction_nolock(struct btrfs_trans_handle *trans,
 				 struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_start_transaction(struct btrfs_root *root,
 						   int num_items);
-struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root,
-						  int num_blocks);
-struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root,
-							  int num_blocks);
-struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *r,
-							 int num_blocks);
+struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root);
+struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root);
+struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *root);
 int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid);
 int btrfs_write_and_wait_transaction(struct btrfs_trans_handle *trans,
 				     struct btrfs_root *root);
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 592396c6dc4..4ce8a9f41d1 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -3177,7 +3177,7 @@ again:
 		tmp_key.offset = (u64)-1;
 
 		wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key);
-		BUG_ON(!wc.replay_dest);
+		BUG_ON(IS_ERR_OR_NULL(wc.replay_dest));
 
 		wc.replay_dest->log_root = log;
 		btrfs_record_root_in_trans(trans, wc.replay_dest);
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index c48214ef5c0..19450bc5363 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -504,7 +504,7 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 		BUG_ON(!new_device);
 		memcpy(new_device, device, sizeof(*new_device));
 		new_device->name = kstrdup(device->name, GFP_NOFS);
-		BUG_ON(!new_device->name);
+		BUG_ON(device->name && !new_device->name);
 		new_device->bdev = NULL;
 		new_device->writeable = 0;
 		new_device->in_fs_metadata = 0;
@@ -689,12 +689,8 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 	transid = btrfs_super_generation(disk_super);
 	if (disk_super->label[0])
 		printk(KERN_INFO "device label %s ", disk_super->label);
-	else {
-		/* FIXME, make a readl uuid parser */
-		printk(KERN_INFO "device fsid %llx-%llx ",
-		       *(unsigned long long *)disk_super->fsid,
-		       *(unsigned long long *)(disk_super->fsid + 8));
-	}
+	else
+		printk(KERN_INFO "device fsid %pU ", disk_super->fsid);
 	printk(KERN_CONT "devid %llu transid %llu %s\n",
 	       (unsigned long long)devid, (unsigned long long)transid, path);
 	ret = device_list_add(path, disk_super, devid, fs_devices_ret);
@@ -2102,7 +2098,8 @@ int btrfs_balance(struct btrfs_root *dev_root)
 					   chunk_root->root_key.objectid,
 					   found_key.objectid,
 					   found_key.offset);
-		BUG_ON(ret && ret != -ENOSPC);
+		if (ret && ret != -ENOSPC)
+			goto error;
 		key.offset = found_key.offset - 1;
 	}
 	ret = 0;
diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c
index f3107e4b4d5..5366fe452ab 100644
--- a/fs/btrfs/xattr.c
+++ b/fs/btrfs/xattr.c
@@ -158,8 +158,6 @@ int __btrfs_setxattr(struct btrfs_trans_handle *trans,
 	if (IS_ERR(trans))
 		return PTR_ERR(trans);
 
-	btrfs_set_trans_block_group(trans, inode);
-
 	ret = do_setxattr(trans, inode, name, value, size, flags);
 	if (ret)
 		goto out;
diff --git a/fs/buffer.c b/fs/buffer.c
index 49c9aada037..1a80b048ade 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -1902,10 +1902,8 @@ int __block_write_begin(struct page *page, loff_t pos, unsigned len,
 		if (!buffer_uptodate(*wait_bh))
 			err = -EIO;
 	}
-	if (unlikely(err)) {
+	if (unlikely(err))
 		page_zero_new_buffers(page, from, to);
-		ClearPageUptodate(page);
-	}
 	return err;
 }
 EXPORT_SYMBOL(__block_write_begin);
diff --git a/fs/cachefiles/bind.c b/fs/cachefiles/bind.c
index a2603e7c0bb..622f4696e48 100644
--- a/fs/cachefiles/bind.c
+++ b/fs/cachefiles/bind.c
@@ -129,8 +129,6 @@ static int cachefiles_daemon_add_cache(struct cachefiles_cache *cache)
 	    !root->d_inode->i_op->mkdir ||
 	    !root->d_inode->i_op->setxattr ||
 	    !root->d_inode->i_op->getxattr ||
-	    !root->d_sb ||
-	    !root->d_sb->s_op ||
 	    !root->d_sb->s_op->statfs ||
 	    !root->d_sb->s_op->sync_fs)
 		goto error_unsupported;
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 33da49dc3cc..5a3953db811 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -453,7 +453,7 @@ static int ceph_writepage(struct page *page, struct writeback_control *wbc)
 	int err;
 	struct inode *inode = page->mapping->host;
 	BUG_ON(!inode);
-	igrab(inode);
+	ihold(inode);
 	err = writepage_nounlock(page, wbc);
 	unlock_page(page);
 	iput(inode);
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index 1f72b00447c..8d74ad7ba55 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -1811,7 +1811,7 @@ out:
 	spin_unlock(&ci->i_unsafe_lock);
 }
 
-int ceph_fsync(struct file *file, int datasync)
+int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ceph_inode_info *ci = ceph_inode(inode);
@@ -1822,9 +1822,10 @@ int ceph_fsync(struct file *file, int datasync)
 	dout("fsync %p%s\n", inode, datasync ? " datasync" : "");
 	sync_write_wait(inode);
 
-	ret = filemap_write_and_wait(inode->i_mapping);
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
 	if (ret < 0)
 		return ret;
+	mutex_lock(&inode->i_mutex);
 
 	dirty = try_flush_caps(inode, NULL, &flush_tid);
 	dout("fsync dirty caps are %s\n", ceph_cap_string(dirty));
@@ -1841,6 +1842,7 @@ int ceph_fsync(struct file *file, int datasync)
 	}
 
 	dout("fsync %p%s done\n", inode, datasync ? " datasync" : "");
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
@@ -2940,14 +2942,12 @@ void ceph_flush_dirty_caps(struct ceph_mds_client *mdsc)
 	while (!list_empty(&mdsc->cap_dirty)) {
 		ci = list_first_entry(&mdsc->cap_dirty, struct ceph_inode_info,
 				      i_dirty_item);
-		inode = igrab(&ci->vfs_inode);
+		inode = &ci->vfs_inode;
+		ihold(inode);
 		dout("flush_dirty_caps %p\n", inode);
 		spin_unlock(&mdsc->cap_dirty_lock);
-		if (inode) {
-			ceph_check_caps(ci, CHECK_CAPS_NODELAY|CHECK_CAPS_FLUSH,
-					NULL);
-			iput(inode);
-		}
+		ceph_check_caps(ci, CHECK_CAPS_NODELAY|CHECK_CAPS_FLUSH, NULL);
+		iput(inode);
 		spin_lock(&mdsc->cap_dirty_lock);
 	}
 	spin_unlock(&mdsc->cap_dirty_lock);
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 33729e822bb..1065ac77984 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -252,7 +252,7 @@ static int ceph_readdir(struct file *filp, void *dirent, filldir_t filldir)
 		off = 1;
 	}
 	if (filp->f_pos == 1) {
-		ino_t ino = filp->f_dentry->d_parent->d_inode->i_ino;
+		ino_t ino = parent_ino(filp->f_dentry);
 		dout("readdir off 1 -> '..'\n");
 		if (filldir(dirent, "..", 2, ceph_make_fpos(0, 1),
 			    ceph_translate_ino(inode->i_sb, ino),
@@ -308,7 +308,8 @@ more:
 		req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
 		if (IS_ERR(req))
 			return PTR_ERR(req);
-		req->r_inode = igrab(inode);
+		req->r_inode = inode;
+		ihold(inode);
 		req->r_dentry = dget(filp->f_dentry);
 		/* hints to request -> mds selection code */
 		req->r_direct_mode = USE_AUTH_MDS;
@@ -445,14 +446,19 @@ static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int origin)
 	loff_t retval;
 
 	mutex_lock(&inode->i_mutex);
+	retval = -EINVAL;
 	switch (origin) {
 	case SEEK_END:
 		offset += inode->i_size + 2;   /* FIXME */
 		break;
 	case SEEK_CUR:
 		offset += file->f_pos;
+	case SEEK_SET:
+		break;
+	default:
+		goto out;
 	}
-	retval = -EINVAL;
+
 	if (offset >= 0 && offset <= inode->i_sb->s_maxbytes) {
 		if (offset != file->f_pos) {
 			file->f_pos = offset;
@@ -476,6 +482,7 @@ static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int origin)
 		if (offset > old_offset)
 			fi->dir_release_count--;
 	}
+out:
 	mutex_unlock(&inode->i_mutex);
 	return retval;
 }
@@ -565,7 +572,6 @@ static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry,
 	/* open (but not create!) intent? */
 	if (nd &&
 	    (nd->flags & LOOKUP_OPEN) &&
-	    (nd->flags & LOOKUP_CONTINUE) == 0 && /* only open last component */
 	    !(nd->intent.open.flags & O_CREAT)) {
 		int mode = nd->intent.open.create_mode & ~current->fs->umask;
 		return ceph_lookup_open(dir, dentry, nd, mode, 1);
@@ -787,10 +793,12 @@ static int ceph_link(struct dentry *old_dentry, struct inode *dir,
 	req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
 	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
 	err = ceph_mdsc_do_request(mdsc, dir, req);
-	if (err)
+	if (err) {
 		d_drop(dentry);
-	else if (!req->r_reply_info.head->is_dentry)
-		d_instantiate(dentry, igrab(old_dentry->d_inode));
+	} else if (!req->r_reply_info.head->is_dentry) {
+		ihold(old_dentry->d_inode);
+		d_instantiate(dentry, old_dentry->d_inode);
+	}
 	ceph_mdsc_put_request(req);
 	return err;
 }
@@ -1110,7 +1118,8 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size,
  * an fsync() on a dir will wait for any uncommitted directory
  * operations to commit.
  */
-static int ceph_dir_fsync(struct file *file, int datasync)
+static int ceph_dir_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync)
 {
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct ceph_inode_info *ci = ceph_inode(inode);
@@ -1120,6 +1129,11 @@ static int ceph_dir_fsync(struct file *file, int datasync)
 	int ret = 0;
 
 	dout("dir_fsync %p\n", inode);
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	spin_lock(&ci->i_unsafe_lock);
 	if (list_empty(head))
 		goto out;
@@ -1153,6 +1167,8 @@ static int ceph_dir_fsync(struct file *file, int datasync)
 	} while (req->r_tid < last_tid);
 out:
 	spin_unlock(&ci->i_unsafe_lock);
+	mutex_unlock(&inode->i_mutex);
+
 	return ret;
 }
 
diff --git a/fs/ceph/export.c b/fs/ceph/export.c
index a610d3d6748..f67b687550d 100644
--- a/fs/ceph/export.c
+++ b/fs/ceph/export.c
@@ -109,7 +109,7 @@ static struct dentry *__fh_to_dentry(struct super_block *sb,
 		err = ceph_mdsc_do_request(mdsc, NULL, req);
 		inode = req->r_target_inode;
 		if (inode)
-			igrab(inode);
+			ihold(inode);
 		ceph_mdsc_put_request(req);
 		if (!inode)
 			return ERR_PTR(-ESTALE);
@@ -167,7 +167,7 @@ static struct dentry *__cfh_to_dentry(struct super_block *sb,
 		err = ceph_mdsc_do_request(mdsc, NULL, req);
 		inode = req->r_target_inode;
 		if (inode)
-			igrab(inode);
+			ihold(inode);
 		ceph_mdsc_put_request(req);
 		if (!inode)
 			return ERR_PTR(err ? err : -ESTALE);
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 203252d88d9..0d0eae05598 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -191,7 +191,8 @@ int ceph_open(struct inode *inode, struct file *file)
 		err = PTR_ERR(req);
 		goto out;
 	}
-	req->r_inode = igrab(inode);
+	req->r_inode = inode;
+	ihold(inode);
 	req->r_num_caps = 1;
 	err = ceph_mdsc_do_request(mdsc, parent_inode, req);
 	if (!err)
@@ -225,7 +226,7 @@ struct dentry *ceph_lookup_open(struct inode *dir, struct dentry *dentry,
 	struct inode *parent_inode = get_dentry_parent_inode(file->f_dentry);
 	struct ceph_mds_request *req;
 	int err;
-	int flags = nd->intent.open.flags - 1;  /* silly vfs! */
+	int flags = nd->intent.open.flags;
 
 	dout("ceph_lookup_open dentry %p '%.*s' flags %d mode 0%o\n",
 	     dentry, dentry->d_name.len, dentry->d_name.name, flags, mode);
@@ -282,14 +283,13 @@ int ceph_release(struct inode *inode, struct file *file)
 static int striped_read(struct inode *inode,
 			u64 off, u64 len,
 			struct page **pages, int num_pages,
-			int *checkeof, bool align_to_pages,
+			int *checkeof, bool o_direct,
 			unsigned long buf_align)
 {
 	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	u64 pos, this_len;
 	int io_align, page_align;
-	int page_off = off & ~PAGE_CACHE_MASK; /* first byte's offset in page */
 	int left, pages_left;
 	int read;
 	struct page **page_pos;
@@ -307,7 +307,7 @@ static int striped_read(struct inode *inode,
 	io_align = off & ~PAGE_MASK;
 
 more:
-	if (align_to_pages)
+	if (o_direct)
 		page_align = (pos - io_align + buf_align) & ~PAGE_MASK;
 	else
 		page_align = pos & ~PAGE_MASK;
@@ -317,20 +317,19 @@ more:
 				  ci->i_truncate_seq,
 				  ci->i_truncate_size,
 				  page_pos, pages_left, page_align);
-	hit_stripe = this_len < left;
-	was_short = ret >= 0 && ret < this_len;
 	if (ret == -ENOENT)
 		ret = 0;
+	hit_stripe = this_len < left;
+	was_short = ret >= 0 && ret < this_len;
 	dout("striped_read %llu~%u (read %u) got %d%s%s\n", pos, left, read,
 	     ret, hit_stripe ? " HITSTRIPE" : "", was_short ? " SHORT" : "");
 
 	if (ret > 0) {
-		int didpages =
-			((pos & ~PAGE_CACHE_MASK) + ret) >> PAGE_CACHE_SHIFT;
+		int didpages = (page_align + ret) >> PAGE_CACHE_SHIFT;
 
 		if (read < pos - off) {
 			dout(" zero gap %llu to %llu\n", off + read, pos);
-			ceph_zero_page_vector_range(page_off + read,
+			ceph_zero_page_vector_range(page_align + read,
 						    pos - off - read, pages);
 		}
 		pos += ret;
@@ -345,20 +344,22 @@ more:
 	}
 
 	if (was_short) {
-		/* was original extent fully inside i_size? */
-		if (pos + left <= inode->i_size) {
-			dout("zero tail\n");
-			ceph_zero_page_vector_range(page_off + read, len - read,
+		/* did we bounce off eof? */
+		if (pos + left > inode->i_size)
+			*checkeof = 1;
+
+		/* zero trailing bytes (inside i_size) */
+		if (left > 0 && pos < inode->i_size) {
+			if (pos + left > inode->i_size)
+				left = inode->i_size - pos;
+
+			dout("zero tail %d\n", left);
+			ceph_zero_page_vector_range(page_align + read, left,
 						    pages);
-			read = len;
-			goto out;
+			read += left;
 		}
-
-		/* check i_size */
-		*checkeof = 1;
 	}
 
-out:
 	if (ret >= 0)
 		ret = read;
 	dout("striped_read returns %d\n", ret);
@@ -475,9 +476,6 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
 	else
 		pos = *offset;
 
-	io_align = pos & ~PAGE_MASK;
-	buf_align = (unsigned long)data & ~PAGE_MASK;
-
 	ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + left);
 	if (ret < 0)
 		return ret;
@@ -501,6 +499,8 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
 	 * boundary.  this isn't atomic, unfortunately.  :(
 	 */
 more:
+	io_align = pos & ~PAGE_MASK;
+	buf_align = (unsigned long)data & ~PAGE_MASK;
 	len = left;
 	if (file->f_flags & O_DIRECT) {
 		/* write from beginning of first page, regardless of
@@ -590,6 +590,7 @@ out:
 		pos += len;
 		written += len;
 		left -= len;
+		data += written;
 		if (left)
 			goto more;
 
@@ -658,7 +659,7 @@ out:
 
 		/* hit EOF or hole? */
 		if (statret == 0 && *ppos < inode->i_size) {
-			dout("aio_read sync_read hit hole, reading more\n");
+			dout("aio_read sync_read hit hole, ppos %lld < size %lld, reading more\n", *ppos, inode->i_size);
 			read += ret;
 			base += ret;
 			len -= ret;
@@ -767,13 +768,16 @@ static loff_t ceph_llseek(struct file *file, loff_t offset, int origin)
 
 	mutex_lock(&inode->i_mutex);
 	__ceph_do_pending_vmtruncate(inode);
-	switch (origin) {
-	case SEEK_END:
+	if (origin != SEEK_CUR || origin != SEEK_SET) {
 		ret = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE);
 		if (ret < 0) {
 			offset = ret;
 			goto out;
 		}
+	}
+
+	switch (origin) {
+	case SEEK_END:
 		offset += inode->i_size;
 		break;
 	case SEEK_CUR:
@@ -789,6 +793,19 @@ static loff_t ceph_llseek(struct file *file, loff_t offset, int origin)
 		}
 		offset += file->f_pos;
 		break;
+	case SEEK_DATA:
+		if (offset >= inode->i_size) {
+			ret = -ENXIO;
+			goto out;
+		}
+		break;
+	case SEEK_HOLE:
+		if (offset >= inode->i_size) {
+			ret = -ENXIO;
+			goto out;
+		}
+		offset = inode->i_size;
+		break;
 	}
 
 	if (offset < 0 || offset > inode->i_sb->s_maxbytes) {
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 70b6a4839c3..dfb2831d8d8 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -1101,10 +1101,10 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req,
 				goto done;
 			}
 			req->r_dentry = dn;  /* may have spliced */
-			igrab(in);
+			ihold(in);
 		} else if (ceph_ino(in) == vino.ino &&
 			   ceph_snap(in) == vino.snap) {
-			igrab(in);
+			ihold(in);
 		} else {
 			dout(" %p links to %p %llx.%llx, not %llx.%llx\n",
 			     dn, in, ceph_ino(in), ceph_snap(in),
@@ -1144,7 +1144,7 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req,
 			goto done;
 		}
 		req->r_dentry = dn;  /* may have spliced */
-		igrab(in);
+		ihold(in);
 		rinfo->head->is_dentry = 1;  /* fool notrace handlers */
 	}
 
@@ -1328,7 +1328,7 @@ void ceph_queue_writeback(struct inode *inode)
 	if (queue_work(ceph_inode_to_client(inode)->wb_wq,
 		       &ceph_inode(inode)->i_wb_work)) {
 		dout("ceph_queue_writeback %p\n", inode);
-		igrab(inode);
+		ihold(inode);
 	} else {
 		dout("ceph_queue_writeback %p failed\n", inode);
 	}
@@ -1353,7 +1353,7 @@ void ceph_queue_invalidate(struct inode *inode)
 	if (queue_work(ceph_inode_to_client(inode)->pg_inv_wq,
 		       &ceph_inode(inode)->i_pg_inv_work)) {
 		dout("ceph_queue_invalidate %p\n", inode);
-		igrab(inode);
+		ihold(inode);
 	} else {
 		dout("ceph_queue_invalidate %p failed\n", inode);
 	}
@@ -1477,7 +1477,7 @@ void ceph_queue_vmtruncate(struct inode *inode)
 	if (queue_work(ceph_sb_to_client(inode->i_sb)->trunc_wq,
 		       &ci->i_vmtruncate_work)) {
 		dout("ceph_queue_vmtruncate %p\n", inode);
-		igrab(inode);
+		ihold(inode);
 	} else {
 		dout("ceph_queue_vmtruncate %p failed, pending=%d\n",
 		     inode, ci->i_truncate_pending);
@@ -1738,7 +1738,8 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr)
 		__mark_inode_dirty(inode, inode_dirty_flags);
 
 	if (mask) {
-		req->r_inode = igrab(inode);
+		req->r_inode = inode;
+		ihold(inode);
 		req->r_inode_drop = release;
 		req->r_args.setattr.mask = cpu_to_le32(mask);
 		req->r_num_caps = 1;
@@ -1779,7 +1780,8 @@ int ceph_do_getattr(struct inode *inode, int mask)
 	req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_GETATTR, USE_ANY_MDS);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
-	req->r_inode = igrab(inode);
+	req->r_inode = inode;
+	ihold(inode);
 	req->r_num_caps = 1;
 	req->r_args.getattr.mask = cpu_to_le32(mask);
 	err = ceph_mdsc_do_request(mdsc, NULL, req);
@@ -1793,17 +1795,17 @@ int ceph_do_getattr(struct inode *inode, int mask)
  * Check inode permissions.  We verify we have a valid value for
  * the AUTH cap, then call the generic handler.
  */
-int ceph_permission(struct inode *inode, int mask, unsigned int flags)
+int ceph_permission(struct inode *inode, int mask)
 {
 	int err;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED);
 
 	if (!err)
-		err = generic_permission(inode, mask, flags, NULL);
+		err = generic_permission(inode, mask);
 	return err;
 }
 
diff --git a/fs/ceph/ioctl.c b/fs/ceph/ioctl.c
index 8888c9ba68d..ef0b5f48e13 100644
--- a/fs/ceph/ioctl.c
+++ b/fs/ceph/ioctl.c
@@ -73,7 +73,8 @@ static long ceph_ioctl_set_layout(struct file *file, void __user *arg)
 				       USE_AUTH_MDS);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
-	req->r_inode = igrab(inode);
+	req->r_inode = inode;
+	ihold(inode);
 	req->r_inode_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_FILE_EXCL;
 
 	req->r_args.setlayout.layout.fl_stripe_unit =
@@ -135,7 +136,8 @@ static long ceph_ioctl_set_layout_policy (struct file *file, void __user *arg)
 
 	if (IS_ERR(req))
 		return PTR_ERR(req);
-	req->r_inode = igrab(inode);
+	req->r_inode = inode;
+	ihold(inode);
 
 	req->r_args.setlayout.layout.fl_stripe_unit =
 			cpu_to_le32(l.stripe_unit);
diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c
index 476b329867d..80576d05d68 100644
--- a/fs/ceph/locks.c
+++ b/fs/ceph/locks.c
@@ -23,7 +23,8 @@ static int ceph_lock_message(u8 lock_type, u16 operation, struct file *file,
 	req = ceph_mdsc_create_request(mdsc, operation, USE_AUTH_MDS);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
-	req->r_inode = igrab(inode);
+	req->r_inode = inode;
+	ihold(inode);
 
 	/* mds requires start and length rather than start and end */
 	if (LLONG_MAX == fl->fl_end)
@@ -32,11 +33,10 @@ static int ceph_lock_message(u8 lock_type, u16 operation, struct file *file,
 		length = fl->fl_end - fl->fl_start + 1;
 
 	dout("ceph_lock_message: rule: %d, op: %d, pid: %llu, start: %llu, "
-	     "length: %llu, wait: %d, type`: %d", (int)lock_type,
+	     "length: %llu, wait: %d, type: %d", (int)lock_type,
 	     (int)operation, (u64)fl->fl_pid, fl->fl_start,
 	     length, wait, fl->fl_type);
 
-
 	req->r_args.filelock_change.rule = lock_type;
 	req->r_args.filelock_change.type = cmd;
 	req->r_args.filelock_change.pid = cpu_to_le64((u64)fl->fl_pid);
@@ -70,7 +70,7 @@ static int ceph_lock_message(u8 lock_type, u16 operation, struct file *file,
 	}
 	ceph_mdsc_put_request(req);
 	dout("ceph_lock_message: rule: %d, op: %d, pid: %llu, start: %llu, "
-	     "length: %llu, wait: %d, type`: %d, err code %d", (int)lock_type,
+	     "length: %llu, wait: %d, type: %d, err code %d", (int)lock_type,
 	     (int)operation, (u64)fl->fl_pid, fl->fl_start,
 	     length, wait, fl->fl_type, err);
 	return err;
@@ -109,16 +109,20 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
 			dout("mds locked, locking locally");
 			err = posix_lock_file(file, fl, NULL);
 			if (err && (CEPH_MDS_OP_SETFILELOCK == op)) {
-				/* undo! This should only happen if the kernel detects
-				 * local deadlock. */
+				/* undo! This should only happen if
+				 * the kernel detects local
+				 * deadlock. */
 				ceph_lock_message(CEPH_LOCK_FCNTL, op, file,
 						  CEPH_LOCK_UNLOCK, 0, fl);
-				dout("got %d on posix_lock_file, undid lock", err);
+				dout("got %d on posix_lock_file, undid lock",
+				     err);
 			}
 		}
 
-	} else {
-		dout("mds returned error code %d", err);
+	} else if (err == -ERESTARTSYS) {
+		dout("undoing lock\n");
+		ceph_lock_message(CEPH_LOCK_FCNTL, op, file,
+				  CEPH_LOCK_UNLOCK, 0, fl);
 	}
 	return err;
 }
@@ -155,8 +159,11 @@ int ceph_flock(struct file *file, int cmd, struct file_lock *fl)
 					  file, CEPH_LOCK_UNLOCK, 0, fl);
 			dout("got %d on flock_lock_file_wait, undid lock", err);
 		}
-	} else {
-		dout("mds error code %d", err);
+	} else if (err == -ERESTARTSYS) {
+		dout("undoing lock\n");
+		ceph_lock_message(CEPH_LOCK_FLOCK,
+				  CEPH_MDS_OP_SETFILELOCK,
+				  file, CEPH_LOCK_UNLOCK, 0, fl);
 	}
 	return err;
 }
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 79743d146be..0c1d9175652 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -1438,12 +1438,15 @@ char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base,
 	struct dentry *temp;
 	char *path;
 	int len, pos;
+	unsigned seq;
 
 	if (dentry == NULL)
 		return ERR_PTR(-EINVAL);
 
 retry:
 	len = 0;
+	seq = read_seqbegin(&rename_lock);
+	rcu_read_lock();
 	for (temp = dentry; !IS_ROOT(temp);) {
 		struct inode *inode = temp->d_inode;
 		if (inode && ceph_snap(inode) == CEPH_SNAPDIR)
@@ -1455,10 +1458,12 @@ retry:
 			len += 1 + temp->d_name.len;
 		temp = temp->d_parent;
 		if (temp == NULL) {
+			rcu_read_unlock();
 			pr_err("build_path corrupt dentry %p\n", dentry);
 			return ERR_PTR(-EINVAL);
 		}
 	}
+	rcu_read_unlock();
 	if (len)
 		len--;  /* no leading '/' */
 
@@ -1467,9 +1472,12 @@ retry:
 		return ERR_PTR(-ENOMEM);
 	pos = len;
 	path[pos] = 0;	/* trailing null */
+	rcu_read_lock();
 	for (temp = dentry; !IS_ROOT(temp) && pos != 0; ) {
-		struct inode *inode = temp->d_inode;
+		struct inode *inode;
 
+		spin_lock(&temp->d_lock);
+		inode = temp->d_inode;
 		if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
 			dout("build_path path+%d: %p SNAPDIR\n",
 			     pos, temp);
@@ -1478,21 +1486,26 @@ retry:
 			break;
 		} else {
 			pos -= temp->d_name.len;
-			if (pos < 0)
+			if (pos < 0) {
+				spin_unlock(&temp->d_lock);
 				break;
+			}
 			strncpy(path + pos, temp->d_name.name,
 				temp->d_name.len);
 		}
+		spin_unlock(&temp->d_lock);
 		if (pos)
 			path[--pos] = '/';
 		temp = temp->d_parent;
 		if (temp == NULL) {
+			rcu_read_unlock();
 			pr_err("build_path corrupt dentry\n");
 			kfree(path);
 			return ERR_PTR(-EINVAL);
 		}
 	}
-	if (pos != 0) {
+	rcu_read_unlock();
+	if (pos != 0 || read_seqretry(&rename_lock, seq)) {
 		pr_err("build_path did not end path lookup where "
 		       "expected, namelen is %d, pos is %d\n", len, pos);
 		/* presumably this is only possible if racing with a
diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c
index 24067d68a55..54b14de2e72 100644
--- a/fs/ceph/snap.c
+++ b/fs/ceph/snap.c
@@ -722,7 +722,7 @@ static void flush_snaps(struct ceph_mds_client *mdsc)
 		ci = list_first_entry(&mdsc->snap_flush_list,
 				struct ceph_inode_info, i_snap_flush_item);
 		inode = &ci->vfs_inode;
-		igrab(inode);
+		ihold(inode);
 		spin_unlock(&mdsc->snap_flush_lock);
 		spin_lock(&inode->i_lock);
 		__ceph_flush_snaps(ci, &session, 0);
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index f5cabefa98d..30446b144e3 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -692,7 +692,7 @@ extern void ceph_queue_invalidate(struct inode *inode);
 extern void ceph_queue_writeback(struct inode *inode);
 
 extern int ceph_do_getattr(struct inode *inode, int mask);
-extern int ceph_permission(struct inode *inode, int mask, unsigned int flags);
+extern int ceph_permission(struct inode *inode, int mask);
 extern int ceph_setattr(struct dentry *dentry, struct iattr *attr);
 extern int ceph_getattr(struct vfsmount *mnt, struct dentry *dentry,
 			struct kstat *stat);
@@ -728,7 +728,8 @@ extern void ceph_put_cap(struct ceph_mds_client *mdsc,
 
 extern void ceph_queue_caps_release(struct inode *inode);
 extern int ceph_write_inode(struct inode *inode, struct writeback_control *wbc);
-extern int ceph_fsync(struct file *file, int datasync);
+extern int ceph_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync);
 extern void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc,
 				    struct ceph_mds_session *session);
 extern struct ceph_cap *ceph_get_cap_for_mds(struct ceph_inode_info *ci,
diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c
index f2b62869618..f42d730f1b6 100644
--- a/fs/ceph/xattr.c
+++ b/fs/ceph/xattr.c
@@ -665,7 +665,8 @@ static int ceph_sync_setxattr(struct dentry *dentry, const char *name,
 		err = PTR_ERR(req);
 		goto out;
 	}
-	req->r_inode = igrab(inode);
+	req->r_inode = inode;
+	ihold(inode);
 	req->r_inode_drop = CEPH_CAP_XATTR_SHARED;
 	req->r_num_caps = 1;
 	req->r_args.setxattr.flags = cpu_to_le32(flags);
@@ -795,7 +796,8 @@ static int ceph_send_removexattr(struct dentry *dentry, const char *name)
 				       USE_AUTH_MDS);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
-	req->r_inode = igrab(inode);
+	req->r_inode = inode;
+	ihold(inode);
 	req->r_inode_drop = CEPH_CAP_XATTR_SHARED;
 	req->r_num_caps = 1;
 	req->r_path2 = kstrdup(name, GFP_NOFS);
diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig
index 1cd4c3a1862..f66cc162515 100644
--- a/fs/cifs/Kconfig
+++ b/fs/cifs/Kconfig
@@ -7,6 +7,7 @@ config CIFS
 	select CRYPTO_MD5
 	select CRYPTO_HMAC
 	select CRYPTO_ARC4
+	select CRYPTO_ECB
 	select CRYPTO_DES
 	help
 	  This is the client VFS module for the Common Internet File System
@@ -148,13 +149,13 @@ config CIFS_FSCACHE
 
 config CIFS_ACL
 	  bool "Provide CIFS ACL support (EXPERIMENTAL)"
-	  depends on EXPERIMENTAL && CIFS_XATTR
+	  depends on EXPERIMENTAL && CIFS_XATTR && KEYS
 	  help
 	    Allows to fetch CIFS/NTFS ACL from the server.  The DACL blob
 	    is handed over to the application/caller.
 
 config CIFS_NFSD_EXPORT
 	  bool "Allow nfsd to export CIFS file system (EXPERIMENTAL)"
-	  depends on CIFS && EXPERIMENTAL
+	  depends on CIFS && EXPERIMENTAL && BROKEN
 	  help
 	   Allows NFS server to export a CIFS mounted share (nfsd over cifs)
diff --git a/fs/cifs/cache.c b/fs/cifs/cache.c
index dd8584d35a1..545509c3313 100644
--- a/fs/cifs/cache.c
+++ b/fs/cifs/cache.c
@@ -92,7 +92,7 @@ static uint16_t cifs_server_get_key(const void *cookie_netfs_data,
 		break;
 
 	default:
-		cERROR(1, "CIFS: Unknown network family '%d'", sa->sa_family);
+		cERROR(1, "Unknown network family '%d'", sa->sa_family);
 		key_len = 0;
 		break;
 	}
@@ -152,7 +152,7 @@ static uint16_t cifs_super_get_key(const void *cookie_netfs_data, void *buffer,
 
 	sharename = extract_sharename(tcon->treeName);
 	if (IS_ERR(sharename)) {
-		cFYI(1, "CIFS: couldn't extract sharename\n");
+		cFYI(1, "%s: couldn't extract sharename\n", __func__);
 		sharename = NULL;
 		return 0;
 	}
@@ -302,7 +302,7 @@ static void cifs_fscache_inode_now_uncached(void *cookie_netfs_data)
 	pagevec_init(&pvec, 0);
 	first = 0;
 
-	cFYI(1, "cifs inode 0x%p now uncached", cifsi);
+	cFYI(1, "%s: cifs inode 0x%p now uncached", __func__, cifsi);
 
 	for (;;) {
 		nr_pages = pagevec_lookup(&pvec,
diff --git a/fs/cifs/cifs_fs_sb.h b/fs/cifs/cifs_fs_sb.h
index ffb1459dc6e..7260e11e21f 100644
--- a/fs/cifs/cifs_fs_sb.h
+++ b/fs/cifs/cifs_fs_sb.h
@@ -42,6 +42,7 @@
 #define CIFS_MOUNT_MULTIUSER	0x20000 /* multiuser mount */
 #define CIFS_MOUNT_STRICT_IO	0x40000 /* strict cache mode */
 #define CIFS_MOUNT_RWPIDFORWARD	0x80000 /* use pid forwarding for rw */
+#define CIFS_MOUNT_POSIXACL	0x100000 /* mirror of MS_POSIXACL in mnt_cifs_flags */
 
 struct cifs_sb_info {
 	struct rb_root tlink_tree;
diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c
index dfbd9f1f373..5a0ee7f2af0 100644
--- a/fs/cifs/cifsencrypt.c
+++ b/fs/cifs/cifsencrypt.c
@@ -184,7 +184,7 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu,
 	if (cifs_pdu == NULL || server == NULL)
 		return -EINVAL;
 
-	if (cifs_pdu->Command == SMB_COM_NEGOTIATE)
+	if (!server->session_estab)
 		return 0;
 
 	if (cifs_pdu->Command == SMB_COM_LOCKING_ANDX) {
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index 989442dcfb4..86551747096 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -35,6 +35,7 @@
 #include <linux/delay.h>
 #include <linux/kthread.h>
 #include <linux/freezer.h>
+#include <linux/namei.h>
 #include <net/ipv6.h>
 #include "cifsfs.h"
 #include "cifspdu.h"
@@ -104,8 +105,7 @@ cifs_sb_deactive(struct super_block *sb)
 }
 
 static int
-cifs_read_super(struct super_block *sb, struct smb_vol *volume_info,
-		const char *devname, int silent)
+cifs_read_super(struct super_block *sb)
 {
 	struct inode *inode;
 	struct cifs_sb_info *cifs_sb;
@@ -113,22 +113,16 @@ cifs_read_super(struct super_block *sb, struct smb_vol *volume_info,
 
 	cifs_sb = CIFS_SB(sb);
 
-	spin_lock_init(&cifs_sb->tlink_tree_lock);
-	cifs_sb->tlink_tree = RB_ROOT;
+	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIXACL)
+		sb->s_flags |= MS_POSIXACL;
 
-	rc = bdi_setup_and_register(&cifs_sb->bdi, "cifs", BDI_CAP_MAP_COPY);
-	if (rc)
-		return rc;
-
-	cifs_sb->bdi.ra_pages = default_backing_dev_info.ra_pages;
-
-	rc = cifs_mount(sb, cifs_sb, volume_info, devname);
+	if (cifs_sb_master_tcon(cifs_sb)->ses->capabilities & CAP_LARGE_FILES)
+		sb->s_maxbytes = MAX_LFS_FILESIZE;
+	else
+		sb->s_maxbytes = MAX_NON_LFS;
 
-	if (rc) {
-		if (!silent)
-			cERROR(1, "cifs_mount failed w/return code = %d", rc);
-		goto out_mount_failed;
-	}
+	/* BB FIXME fix time_gran to be larger for LANMAN sessions */
+	sb->s_time_gran = 100;
 
 	sb->s_magic = CIFS_MAGIC_NUMBER;
 	sb->s_op = &cifs_super_ops;
@@ -170,37 +164,14 @@ out_no_root:
 	if (inode)
 		iput(inode);
 
-	cifs_umount(sb, cifs_sb);
-
-out_mount_failed:
-	bdi_destroy(&cifs_sb->bdi);
 	return rc;
 }
 
-static void
-cifs_put_super(struct super_block *sb)
+static void cifs_kill_sb(struct super_block *sb)
 {
-	int rc = 0;
-	struct cifs_sb_info *cifs_sb;
-
-	cFYI(1, "In cifs_put_super");
-	cifs_sb = CIFS_SB(sb);
-	if (cifs_sb == NULL) {
-		cFYI(1, "Empty cifs superblock info passed to unmount");
-		return;
-	}
-
-	rc = cifs_umount(sb, cifs_sb);
-	if (rc)
-		cERROR(1, "cifs_umount failed with return code %d", rc);
-	if (cifs_sb->mountdata) {
-		kfree(cifs_sb->mountdata);
-		cifs_sb->mountdata = NULL;
-	}
-
-	unload_nls(cifs_sb->local_nls);
-	bdi_destroy(&cifs_sb->bdi);
-	kfree(cifs_sb);
+	struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
+	kill_anon_super(sb);
+	cifs_umount(cifs_sb);
 }
 
 static int
@@ -253,13 +224,10 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	return 0;
 }
 
-static int cifs_permission(struct inode *inode, int mask, unsigned int flags)
+static int cifs_permission(struct inode *inode, int mask)
 {
 	struct cifs_sb_info *cifs_sb;
 
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
 	cifs_sb = CIFS_SB(inode->i_sb);
 
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM) {
@@ -271,7 +239,7 @@ static int cifs_permission(struct inode *inode, int mask, unsigned int flags)
 		on the client (above and beyond ACL on servers) for
 		servers which do not support setting and viewing mode bits,
 		so allowing client to check permissions is useful */
-		return generic_permission(inode, mask, flags, NULL);
+		return generic_permission(inode, mask);
 }
 
 static struct kmem_cache *cifs_inode_cachep;
@@ -352,6 +320,37 @@ cifs_show_address(struct seq_file *s, struct TCP_Server_Info *server)
 	}
 }
 
+static void
+cifs_show_security(struct seq_file *s, struct TCP_Server_Info *server)
+{
+	seq_printf(s, ",sec=");
+
+	switch (server->secType) {
+	case LANMAN:
+		seq_printf(s, "lanman");
+		break;
+	case NTLMv2:
+		seq_printf(s, "ntlmv2");
+		break;
+	case NTLM:
+		seq_printf(s, "ntlm");
+		break;
+	case Kerberos:
+		seq_printf(s, "krb5");
+		break;
+	case RawNTLMSSP:
+		seq_printf(s, "ntlmssp");
+		break;
+	default:
+		/* shouldn't ever happen */
+		seq_printf(s, "unknown");
+		break;
+	}
+
+	if (server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))
+		seq_printf(s, "i");
+}
+
 /*
  * cifs_show_options() is for displaying mount options in /proc/mounts.
  * Not all settable options are displayed but most of the important
@@ -365,6 +364,8 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m)
 	struct sockaddr *srcaddr;
 	srcaddr = (struct sockaddr *)&tcon->ses->server->srcaddr;
 
+	cifs_show_security(s, tcon->ses->server);
+
 	seq_printf(s, ",unc=%s", tcon->treeName);
 
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER)
@@ -518,7 +519,6 @@ static int cifs_drop_inode(struct inode *inode)
 }
 
 static const struct super_operations cifs_super_ops = {
-	.put_super = cifs_put_super,
 	.statfs = cifs_statfs,
 	.alloc_inode = cifs_alloc_inode,
 	.destroy_inode = cifs_destroy_inode,
@@ -543,91 +543,55 @@ static const struct super_operations cifs_super_ops = {
 static struct dentry *
 cifs_get_root(struct smb_vol *vol, struct super_block *sb)
 {
-	int xid, rc;
-	struct inode *inode;
-	struct qstr name;
-	struct dentry *dparent = NULL, *dchild = NULL, *alias;
+	struct dentry *dentry;
 	struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
-	unsigned int i, full_len, len;
-	char *full_path = NULL, *pstart;
+	char *full_path = NULL;
+	char *s, *p;
 	char sep;
+	int xid;
 
 	full_path = cifs_build_path_to_root(vol, cifs_sb,
 					    cifs_sb_master_tcon(cifs_sb));
 	if (full_path == NULL)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	cFYI(1, "Get root dentry for %s", full_path);
 
 	xid = GetXid();
 	sep = CIFS_DIR_SEP(cifs_sb);
-	dparent = dget(sb->s_root);
-	full_len = strlen(full_path);
-	full_path[full_len] = sep;
-	pstart = full_path + 1;
-
-	for (i = 1, len = 0; i <= full_len; i++) {
-		if (full_path[i] != sep || !len) {
-			len++;
-			continue;
-		}
-
-		full_path[i] = 0;
-		cFYI(1, "get dentry for %s", pstart);
-
-		name.name = pstart;
-		name.len = len;
-		name.hash = full_name_hash(pstart, len);
-		dchild = d_lookup(dparent, &name);
-		if (dchild == NULL) {
-			cFYI(1, "not exists");
-			dchild = d_alloc(dparent, &name);
-			if (dchild == NULL) {
-				dput(dparent);
-				dparent = NULL;
-				goto out;
-			}
-		}
-
-		cFYI(1, "get inode");
-		if (dchild->d_inode == NULL) {
-			cFYI(1, "not exists");
-			inode = NULL;
-			if (cifs_sb_master_tcon(CIFS_SB(sb))->unix_ext)
-				rc = cifs_get_inode_info_unix(&inode, full_path,
-							      sb, xid);
-			else
-				rc = cifs_get_inode_info(&inode, full_path,
-							 NULL, sb, xid, NULL);
-			if (rc) {
-				dput(dchild);
-				dput(dparent);
-				dparent = NULL;
-				goto out;
-			}
-			alias = d_materialise_unique(dchild, inode);
-			if (alias != NULL) {
-				dput(dchild);
-				if (IS_ERR(alias)) {
-					dput(dparent);
-					dparent = NULL;
-					goto out;
-				}
-				dchild = alias;
-			}
-		}
-		cFYI(1, "parent %p, child %p", dparent, dchild);
-
-		dput(dparent);
-		dparent = dchild;
-		len = 0;
-		pstart = full_path + i + 1;
-		full_path[i] = sep;
-	}
-out:
+	dentry = dget(sb->s_root);
+	p = s = full_path;
+
+	do {
+		struct inode *dir = dentry->d_inode;
+		struct dentry *child;
+
+		/* skip separators */
+		while (*s == sep)
+			s++;
+		if (!*s)
+			break;
+		p = s++;
+		/* next separator */
+		while (*s && *s != sep)
+			s++;
+
+		mutex_lock(&dir->i_mutex);
+		child = lookup_one_len(p, dentry, s - p);
+		mutex_unlock(&dir->i_mutex);
+		dput(dentry);
+		dentry = child;
+	} while (!IS_ERR(dentry));
 	_FreeXid(xid);
 	kfree(full_path);
-	return dparent;
+	return dentry;
+}
+
+static int cifs_set_super(struct super_block *sb, void *data)
+{
+	struct cifs_mnt_data *mnt_data = data;
+	sb->s_fs_info = mnt_data->cifs_sb;
+	return set_anon_super(sb, NULL);
 }
 
 static struct dentry *
@@ -643,82 +607,80 @@ cifs_do_mount(struct file_system_type *fs_type,
 
 	cFYI(1, "Devname: %s flags: %d ", dev_name, flags);
 
-	rc = cifs_setup_volume_info(&volume_info, (char *)data, dev_name);
-	if (rc)
-		return ERR_PTR(rc);
+	volume_info = cifs_get_volume_info((char *)data, dev_name);
+	if (IS_ERR(volume_info))
+		return ERR_CAST(volume_info);
 
 	cifs_sb = kzalloc(sizeof(struct cifs_sb_info), GFP_KERNEL);
 	if (cifs_sb == NULL) {
 		root = ERR_PTR(-ENOMEM);
-		goto out;
+		goto out_nls;
+	}
+
+	cifs_sb->mountdata = kstrndup(data, PAGE_SIZE, GFP_KERNEL);
+	if (cifs_sb->mountdata == NULL) {
+		root = ERR_PTR(-ENOMEM);
+		goto out_cifs_sb;
 	}
 
 	cifs_setup_cifs_sb(volume_info, cifs_sb);
 
+	rc = cifs_mount(cifs_sb, volume_info);
+	if (rc) {
+		if (!(flags & MS_SILENT))
+			cERROR(1, "cifs_mount failed w/return code = %d", rc);
+		root = ERR_PTR(rc);
+		goto out_mountdata;
+	}
+
 	mnt_data.vol = volume_info;
 	mnt_data.cifs_sb = cifs_sb;
 	mnt_data.flags = flags;
 
-	sb = sget(fs_type, cifs_match_super, set_anon_super, &mnt_data);
+	sb = sget(fs_type, cifs_match_super, cifs_set_super, &mnt_data);
 	if (IS_ERR(sb)) {
 		root = ERR_CAST(sb);
-		goto out_cifs_sb;
+		cifs_umount(cifs_sb);
+		goto out;
 	}
 
-	if (sb->s_fs_info) {
+	if (sb->s_root) {
 		cFYI(1, "Use existing superblock");
-		goto out_shared;
-	}
-
-	/*
-	 * Copy mount params for use in submounts. Better to do
-	 * the copy here and deal with the error before cleanup gets
-	 * complicated post-mount.
-	 */
-	cifs_sb->mountdata = kstrndup(data, PAGE_SIZE, GFP_KERNEL);
-	if (cifs_sb->mountdata == NULL) {
-		root = ERR_PTR(-ENOMEM);
-		goto out_super;
-	}
-
-	sb->s_flags = flags;
-	/* BB should we make this contingent on mount parm? */
-	sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
-	sb->s_fs_info = cifs_sb;
+		cifs_umount(cifs_sb);
+	} else {
+		sb->s_flags = flags;
+		/* BB should we make this contingent on mount parm? */
+		sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
+
+		rc = cifs_read_super(sb);
+		if (rc) {
+			root = ERR_PTR(rc);
+			goto out_super;
+		}
 
-	rc = cifs_read_super(sb, volume_info, dev_name,
-			     flags & MS_SILENT ? 1 : 0);
-	if (rc) {
-		root = ERR_PTR(rc);
-		goto out_super;
+		sb->s_flags |= MS_ACTIVE;
 	}
 
-	sb->s_flags |= MS_ACTIVE;
-
 	root = cifs_get_root(volume_info, sb);
-	if (root == NULL)
+	if (IS_ERR(root))
 		goto out_super;
 
 	cFYI(1, "dentry root is: %p", root);
 	goto out;
 
-out_shared:
-	root = cifs_get_root(volume_info, sb);
-	if (root)
-		cFYI(1, "dentry root is: %p", root);
-	goto out;
-
 out_super:
-	kfree(cifs_sb->mountdata);
 	deactivate_locked_super(sb);
+out:
+	cifs_cleanup_volume_info(volume_info);
+	return root;
 
+out_mountdata:
+	kfree(cifs_sb->mountdata);
 out_cifs_sb:
-	unload_nls(cifs_sb->local_nls);
 	kfree(cifs_sb);
-
-out:
-	cifs_cleanup_volume_info(&volume_info);
-	return root;
+out_nls:
+	unload_nls(volume_info->local_nls);
+	goto out;
 }
 
 static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
@@ -742,8 +704,11 @@ static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
 
 static loff_t cifs_llseek(struct file *file, loff_t offset, int origin)
 {
-	/* origin == SEEK_END => we must revalidate the cached file length */
-	if (origin == SEEK_END) {
+	/*
+	 * origin == SEEK_END || SEEK_DATA || SEEK_HOLE => we must revalidate
+	 * the cached file length
+	 */
+	if (origin != SEEK_SET || origin != SEEK_CUR) {
 		int rc;
 		struct inode *inode = file->f_path.dentry->d_inode;
 
@@ -807,7 +772,7 @@ struct file_system_type cifs_fs_type = {
 	.owner = THIS_MODULE,
 	.name = "cifs",
 	.mount = cifs_do_mount,
-	.kill_sb = kill_anon_super,
+	.kill_sb = cifs_kill_sb,
 	/*  .fs_flags */
 };
 const struct inode_operations cifs_dir_inode_ops = {
diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h
index 64313f778eb..fbd050c8d52 100644
--- a/fs/cifs/cifsfs.h
+++ b/fs/cifs/cifsfs.h
@@ -91,8 +91,8 @@ extern ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov,
 extern ssize_t cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov,
 				  unsigned long nr_segs, loff_t pos);
 extern int cifs_lock(struct file *, int, struct file_lock *);
-extern int cifs_fsync(struct file *, int);
-extern int cifs_strict_fsync(struct file *, int);
+extern int cifs_fsync(struct file *, loff_t, loff_t, int);
+extern int cifs_strict_fsync(struct file *, loff_t, loff_t, int);
 extern int cifs_flush(struct file *, fl_owner_t id);
 extern int cifs_file_mmap(struct file * , struct vm_area_struct *);
 extern int cifs_file_strict_mmap(struct file * , struct vm_area_struct *);
@@ -129,5 +129,5 @@ extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg);
 extern const struct export_operations cifs_export_ops;
 #endif /* CIFS_NFSD_EXPORT */
 
-#define CIFS_VERSION   "1.72"
+#define CIFS_VERSION   "1.74"
 #endif				/* _CIFSFS_H */
diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h
index 953f84413c7..8df28e925e5 100644
--- a/fs/cifs/cifsproto.h
+++ b/fs/cifs/cifsproto.h
@@ -154,12 +154,11 @@ extern int set_cifs_acl(struct cifs_ntsd *, __u32, struct inode *,
 extern void cifs_setup_cifs_sb(struct smb_vol *pvolume_info,
 			       struct cifs_sb_info *cifs_sb);
 extern int cifs_match_super(struct super_block *, void *);
-extern void cifs_cleanup_volume_info(struct smb_vol **pvolume_info);
-extern int cifs_setup_volume_info(struct smb_vol **pvolume_info,
-				  char *mount_data, const char *devname);
-extern int cifs_mount(struct super_block *, struct cifs_sb_info *,
-		      struct smb_vol *, const char *);
-extern int cifs_umount(struct super_block *, struct cifs_sb_info *);
+extern void cifs_cleanup_volume_info(struct smb_vol *pvolume_info);
+extern struct smb_vol *cifs_get_volume_info(char *mount_data,
+					    const char *devname);
+extern int cifs_mount(struct cifs_sb_info *, struct smb_vol *);
+extern void cifs_umount(struct cifs_sb_info *);
 extern void cifs_dfs_release_automount_timer(void);
 void cifs_proc_init(void);
 void cifs_proc_clean(void);
@@ -218,7 +217,8 @@ extern int get_dfs_path(int xid, struct cifs_ses *pSesInfo,
 			struct dfs_info3_param **preferrals,
 			int remap);
 extern void reset_cifs_unix_caps(int xid, struct cifs_tcon *tcon,
-				 struct super_block *sb, struct smb_vol *vol);
+				 struct cifs_sb_info *cifs_sb,
+				 struct smb_vol *vol);
 extern int CIFSSMBQFSInfo(const int xid, struct cifs_tcon *tcon,
 			struct kstatfs *FSData);
 extern int SMBOldQFSInfo(const int xid, struct cifs_tcon *tcon,
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 6d88b82537c..e66297bad41 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -65,6 +65,8 @@ static int ip_connect(struct TCP_Server_Info *server);
 static int generic_ip_connect(struct TCP_Server_Info *server);
 static void tlink_rb_insert(struct rb_root *root, struct tcon_link *new_tlink);
 static void cifs_prune_tlinks(struct work_struct *work);
+static int cifs_setup_volume_info(struct smb_vol *volume_info, char *mount_data,
+					const char *devname);
 
 /*
  * cifs tcp session reconnection
@@ -152,7 +154,7 @@ cifs_reconnect(struct TCP_Server_Info *server)
 		mid_entry->callback(mid_entry);
 	}
 
-	while (server->tcpStatus == CifsNeedReconnect) {
+	do {
 		try_to_freeze();
 
 		/* we should try only the port we connected to before */
@@ -167,7 +169,7 @@ cifs_reconnect(struct TCP_Server_Info *server)
 				server->tcpStatus = CifsNeedNegotiate;
 			spin_unlock(&GlobalMid_Lock);
 		}
-	}
+	} while (server->tcpStatus == CifsNeedReconnect);
 
 	return rc;
 }
@@ -318,9 +320,10 @@ requeue_echo:
 }
 
 static int
-cifs_demultiplex_thread(struct TCP_Server_Info *server)
+cifs_demultiplex_thread(void *p)
 {
 	int length;
+	struct TCP_Server_Info *server = p;
 	unsigned int pdu_length, total_read;
 	struct smb_hdr *smb_buffer = NULL;
 	struct smb_hdr *bigbuf = NULL;
@@ -784,7 +787,7 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			 struct smb_vol *vol)
 {
 	char *value, *data, *end;
-	char *mountdata_copy, *options;
+	char *mountdata_copy = NULL, *options;
 	unsigned int  temp_len, i, j;
 	char separator[2];
 	short int override_uid = -1;
@@ -1391,7 +1394,7 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 				"/proc/fs/cifs/LookupCacheEnabled to 0\n");
 		} else if (strnicmp(data, "fsc", 3) == 0) {
 #ifndef CONFIG_CIFS_FSCACHE
-			cERROR(1, "FS-Cache support needs CONFIG_CIFS_FSCACHE"
+			cERROR(1, "FS-Cache support needs CONFIG_CIFS_FSCACHE "
 				  "kernel config option set");
 			goto cifs_parse_mount_err;
 #endif
@@ -1789,7 +1792,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info)
 	 * this will succeed. No need for try_module_get().
 	 */
 	__module_get(THIS_MODULE);
-	tcp_ses->tsk = kthread_run((void *)(void *)cifs_demultiplex_thread,
+	tcp_ses->tsk = kthread_run(cifs_demultiplex_thread,
 				  tcp_ses, "cifsd");
 	if (IS_ERR(tcp_ses->tsk)) {
 		rc = PTR_ERR(tcp_ses->tsk);
@@ -1976,7 +1979,7 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info)
 		warned_on_ntlm = true;
 		cERROR(1, "default security mechanism requested.  The default "
 			"security mechanism will be upgraded from ntlm to "
-			"ntlmv2 in kernel release 2.6.41");
+			"ntlmv2 in kernel release 3.1");
 	}
 	ses->overrideSecFlg = volume_info->secFlg;
 
@@ -2149,7 +2152,10 @@ cifs_put_tlink(struct tcon_link *tlink)
 }
 
 static inline struct tcon_link *
-cifs_sb_master_tlink(struct cifs_sb_info *cifs_sb);
+cifs_sb_master_tlink(struct cifs_sb_info *cifs_sb)
+{
+	return cifs_sb->master_tlink;
+}
 
 static int
 compare_mount_options(struct super_block *sb, struct cifs_mnt_data *mnt_data)
@@ -2237,8 +2243,8 @@ cifs_match_super(struct super_block *sb, void *data)
 
 	rc = compare_mount_options(sb, mnt_data);
 out:
-	cifs_put_tlink(tlink);
 	spin_unlock(&cifs_tcp_ses_lock);
+	cifs_put_tlink(tlink);
 	return rc;
 }
 
@@ -2471,14 +2477,6 @@ generic_ip_connect(struct TCP_Server_Info *server)
 	if (rc < 0)
 		return rc;
 
-	rc = socket->ops->connect(socket, saddr, slen, 0);
-	if (rc < 0) {
-		cFYI(1, "Error %d connecting to server", rc);
-		sock_release(socket);
-		server->ssocket = NULL;
-		return rc;
-	}
-
 	/*
 	 * Eventually check for other socket options to change from
 	 * the default. sock_setsockopt not used because it expects
@@ -2507,6 +2505,14 @@ generic_ip_connect(struct TCP_Server_Info *server)
 		 socket->sk->sk_sndbuf,
 		 socket->sk->sk_rcvbuf, socket->sk->sk_rcvtimeo);
 
+	rc = socket->ops->connect(socket, saddr, slen, 0);
+	if (rc < 0) {
+		cFYI(1, "Error %d connecting to server", rc);
+		sock_release(socket);
+		server->ssocket = NULL;
+		return rc;
+	}
+
 	if (sport == htons(RFC1001_PORT))
 		rc = ip_rfc1001_connect(server);
 
@@ -2543,7 +2549,7 @@ ip_connect(struct TCP_Server_Info *server)
 }
 
 void reset_cifs_unix_caps(int xid, struct cifs_tcon *tcon,
-			  struct super_block *sb, struct smb_vol *vol_info)
+			  struct cifs_sb_info *cifs_sb, struct smb_vol *vol_info)
 {
 	/* if we are reconnecting then should we check to see if
 	 * any requested capabilities changed locally e.g. via
@@ -2597,22 +2603,23 @@ void reset_cifs_unix_caps(int xid, struct cifs_tcon *tcon,
 			cap &= ~CIFS_UNIX_POSIX_ACL_CAP;
 		else if (CIFS_UNIX_POSIX_ACL_CAP & cap) {
 			cFYI(1, "negotiated posix acl support");
-			if (sb)
-				sb->s_flags |= MS_POSIXACL;
+			if (cifs_sb)
+				cifs_sb->mnt_cifs_flags |=
+					CIFS_MOUNT_POSIXACL;
 		}
 
 		if (vol_info && vol_info->posix_paths == 0)
 			cap &= ~CIFS_UNIX_POSIX_PATHNAMES_CAP;
 		else if (cap & CIFS_UNIX_POSIX_PATHNAMES_CAP) {
 			cFYI(1, "negotiate posix pathnames");
-			if (sb)
-				CIFS_SB(sb)->mnt_cifs_flags |=
+			if (cifs_sb)
+				cifs_sb->mnt_cifs_flags |=
 					CIFS_MOUNT_POSIX_PATHS;
 		}
 
-		if (sb && (CIFS_SB(sb)->rsize > 127 * 1024)) {
+		if (cifs_sb && (cifs_sb->rsize > 127 * 1024)) {
 			if ((cap & CIFS_UNIX_LARGE_READ_CAP) == 0) {
-				CIFS_SB(sb)->rsize = 127 * 1024;
+				cifs_sb->rsize = 127 * 1024;
 				cFYI(DBG2, "larger reads not supported by srv");
 			}
 		}
@@ -2659,6 +2666,9 @@ void cifs_setup_cifs_sb(struct smb_vol *pvolume_info,
 {
 	INIT_DELAYED_WORK(&cifs_sb->prune_tlinks, cifs_prune_tlinks);
 
+	spin_lock_init(&cifs_sb->tlink_tree_lock);
+	cifs_sb->tlink_tree = RB_ROOT;
+
 	if (pvolume_info->rsize > CIFSMaxBufSize) {
 		cERROR(1, "rsize %d too large, using MaxBufSize",
 			pvolume_info->rsize);
@@ -2747,21 +2757,21 @@ void cifs_setup_cifs_sb(struct smb_vol *pvolume_info,
 
 /*
  * When the server supports very large writes via POSIX extensions, we can
- * allow up to 2^24 - PAGE_CACHE_SIZE.
+ * allow up to 2^24-1, minus the size of a WRITE_AND_X header, not including
+ * the RFC1001 length.
  *
  * Note that this might make for "interesting" allocation problems during
- * writeback however (as we have to allocate an array of pointers for the
- * pages). A 16M write means ~32kb page array with PAGE_CACHE_SIZE == 4096.
+ * writeback however as we have to allocate an array of pointers for the
+ * pages. A 16M write means ~32kb page array with PAGE_CACHE_SIZE == 4096.
  */
-#define CIFS_MAX_WSIZE ((1<<24) - PAGE_CACHE_SIZE)
+#define CIFS_MAX_WSIZE ((1<<24) - 1 - sizeof(WRITE_REQ) + 4)
 
 /*
- * When the server doesn't allow large posix writes, default to a wsize of
- * 128k - PAGE_CACHE_SIZE -- one page less than the largest frame size
- * described in RFC1001. This allows space for the header without going over
- * that by default.
+ * When the server doesn't allow large posix writes, only allow a wsize of
+ * 128k minus the size of the WRITE_AND_X header. That allows for a write up
+ * to the maximum size described by RFC1002.
  */
-#define CIFS_MAX_RFC1001_WSIZE (128 * 1024 - PAGE_CACHE_SIZE)
+#define CIFS_MAX_RFC1002_WSIZE (128 * 1024 - sizeof(WRITE_REQ) + 4)
 
 /*
  * The default wsize is 1M. find_get_pages seems to return a maximum of 256
@@ -2780,11 +2790,18 @@ cifs_negotiate_wsize(struct cifs_tcon *tcon, struct smb_vol *pvolume_info)
 
 	/* can server support 24-bit write sizes? (via UNIX extensions) */
 	if (!tcon->unix_ext || !(unix_cap & CIFS_UNIX_LARGE_WRITE_CAP))
-		wsize = min_t(unsigned int, wsize, CIFS_MAX_RFC1001_WSIZE);
+		wsize = min_t(unsigned int, wsize, CIFS_MAX_RFC1002_WSIZE);
 
-	/* no CAP_LARGE_WRITE_X? Limit it to 16 bits */
-	if (!(server->capabilities & CAP_LARGE_WRITE_X))
-		wsize = min_t(unsigned int, wsize, USHRT_MAX);
+	/*
+	 * no CAP_LARGE_WRITE_X or is signing enabled without CAP_UNIX set?
+	 * Limit it to max buffer offered by the server, minus the size of the
+	 * WRITEX header, not including the 4 byte RFC1001 length.
+	 */
+	if (!(server->capabilities & CAP_LARGE_WRITE_X) ||
+	    (!(server->capabilities & CAP_UNIX) &&
+	     (server->sec_mode & (SECMODE_SIGN_ENABLED|SECMODE_SIGN_REQUIRED))))
+		wsize = min_t(unsigned int, wsize,
+				server->maxBuf - sizeof(WRITE_REQ) + 4);
 
 	/* hard limit of CIFS_MAX_WSIZE */
 	wsize = min_t(unsigned int, wsize, CIFS_MAX_WSIZE);
@@ -2816,15 +2833,9 @@ is_path_accessible(int xid, struct cifs_tcon *tcon,
 	return rc;
 }
 
-void
-cifs_cleanup_volume_info(struct smb_vol **pvolume_info)
+static void
+cleanup_volume_info_contents(struct smb_vol *volume_info)
 {
-	struct smb_vol *volume_info;
-
-	if (!pvolume_info || !*pvolume_info)
-		return;
-
-	volume_info = *pvolume_info;
 	kfree(volume_info->username);
 	kzfree(volume_info->password);
 	kfree(volume_info->UNC);
@@ -2832,28 +2843,44 @@ cifs_cleanup_volume_info(struct smb_vol **pvolume_info)
 	kfree(volume_info->domainname);
 	kfree(volume_info->iocharset);
 	kfree(volume_info->prepath);
+}
+
+void
+cifs_cleanup_volume_info(struct smb_vol *volume_info)
+{
+	if (!volume_info)
+		return;
+	cleanup_volume_info_contents(volume_info);
 	kfree(volume_info);
-	*pvolume_info = NULL;
-	return;
 }
 
+
 #ifdef CONFIG_CIFS_DFS_UPCALL
 /* build_path_to_root returns full path to root when
  * we do not have an exiting connection (tcon) */
 static char *
-build_unc_path_to_root(const struct smb_vol *volume_info,
+build_unc_path_to_root(const struct smb_vol *vol,
 		const struct cifs_sb_info *cifs_sb)
 {
-	char *full_path;
+	char *full_path, *pos;
+	unsigned int pplen = vol->prepath ? strlen(vol->prepath) : 0;
+	unsigned int unc_len = strnlen(vol->UNC, MAX_TREE_SIZE + 1);
 
-	int unc_len = strnlen(volume_info->UNC, MAX_TREE_SIZE + 1);
-	full_path = kmalloc(unc_len + 1, GFP_KERNEL);
+	full_path = kmalloc(unc_len + pplen + 1, GFP_KERNEL);
 	if (full_path == NULL)
 		return ERR_PTR(-ENOMEM);
 
-	strncpy(full_path, volume_info->UNC, unc_len);
-	full_path[unc_len] = 0; /* add trailing null */
+	strncpy(full_path, vol->UNC, unc_len);
+	pos = full_path + unc_len;
+
+	if (pplen) {
+		strncpy(pos, vol->prepath, pplen);
+		pos += pplen;
+	}
+
+	*pos = '\0'; /* add trailing null */
 	convert_delimiter(full_path, CIFS_DIR_SEP(cifs_sb));
+	cFYI(1, "%s: full_path=%s", __func__, full_path);
 	return full_path;
 }
 
@@ -2896,15 +2923,18 @@ expand_dfs_referral(int xid, struct cifs_ses *pSesInfo,
 						   &fake_devname);
 
 		free_dfs_info_array(referrals, num_referrals);
-		kfree(fake_devname);
-
-		if (cifs_sb->mountdata != NULL)
-			kfree(cifs_sb->mountdata);
 
 		if (IS_ERR(mdata)) {
 			rc = PTR_ERR(mdata);
 			mdata = NULL;
+		} else {
+			cleanup_volume_info_contents(volume_info);
+			memset(volume_info, '\0', sizeof(*volume_info));
+			rc = cifs_setup_volume_info(volume_info, mdata,
+							fake_devname);
 		}
+		kfree(fake_devname);
+		kfree(cifs_sb->mountdata);
 		cifs_sb->mountdata = mdata;
 	}
 	kfree(full_path);
@@ -2912,29 +2942,20 @@ expand_dfs_referral(int xid, struct cifs_ses *pSesInfo,
 }
 #endif
 
-int cifs_setup_volume_info(struct smb_vol **pvolume_info, char *mount_data,
-			   const char *devname)
+static int
+cifs_setup_volume_info(struct smb_vol *volume_info, char *mount_data,
+			const char *devname)
 {
-	struct smb_vol *volume_info;
 	int rc = 0;
 
-	*pvolume_info = NULL;
-
-	volume_info = kzalloc(sizeof(struct smb_vol), GFP_KERNEL);
-	if (!volume_info) {
-		rc = -ENOMEM;
-		goto out;
-	}
-
-	if (cifs_parse_mount_options(mount_data, devname,
-				     volume_info)) {
-		rc = -EINVAL;
-		goto out;
-	}
+	if (cifs_parse_mount_options(mount_data, devname, volume_info))
+		return -EINVAL;
 
 	if (volume_info->nullauth) {
 		cFYI(1, "null user");
-		volume_info->username = "";
+		volume_info->username = kzalloc(1, GFP_KERNEL);
+		if (volume_info->username == NULL)
+			return -ENOMEM;
 	} else if (volume_info->username) {
 		/* BB fixme parse for domain name here */
 		cFYI(1, "Username: %s", volume_info->username);
@@ -2942,8 +2963,7 @@ int cifs_setup_volume_info(struct smb_vol **pvolume_info, char *mount_data,
 		cifserror("No username specified");
 	/* In userspace mount helper we can get user name from alternate
 	   locations such as env variables and files on disk */
-		rc = -EINVAL;
-		goto out;
+		return -EINVAL;
 	}
 
 	/* this is needed for ASCII cp to Unicode converts */
@@ -2955,21 +2975,34 @@ int cifs_setup_volume_info(struct smb_vol **pvolume_info, char *mount_data,
 		if (volume_info->local_nls == NULL) {
 			cERROR(1, "CIFS mount error: iocharset %s not found",
 				 volume_info->iocharset);
-			rc = -ELIBACC;
-			goto out;
+			return -ELIBACC;
 		}
 	}
 
-	*pvolume_info = volume_info;
-	return rc;
-out:
-	cifs_cleanup_volume_info(&volume_info);
 	return rc;
 }
 
+struct smb_vol *
+cifs_get_volume_info(char *mount_data, const char *devname)
+{
+	int rc;
+	struct smb_vol *volume_info;
+
+	volume_info = kzalloc(sizeof(struct smb_vol), GFP_KERNEL);
+	if (!volume_info)
+		return ERR_PTR(-ENOMEM);
+
+	rc = cifs_setup_volume_info(volume_info, mount_data, devname);
+	if (rc) {
+		cifs_cleanup_volume_info(volume_info);
+		volume_info = ERR_PTR(rc);
+	}
+
+	return volume_info;
+}
+
 int
-cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
-	   struct smb_vol *volume_info, const char *devname)
+cifs_mount(struct cifs_sb_info *cifs_sb, struct smb_vol *volume_info)
 {
 	int rc = 0;
 	int xid;
@@ -2980,6 +3013,15 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
 	struct tcon_link *tlink;
 #ifdef CONFIG_CIFS_DFS_UPCALL
 	int referral_walks_count = 0;
+#endif
+
+	rc = bdi_setup_and_register(&cifs_sb->bdi, "cifs", BDI_CAP_MAP_COPY);
+	if (rc)
+		return rc;
+
+	cifs_sb->bdi.ra_pages = default_backing_dev_info.ra_pages;
+
+#ifdef CONFIG_CIFS_DFS_UPCALL
 try_mount_again:
 	/* cleanup activities if we're chasing a referral */
 	if (referral_walks_count) {
@@ -2988,7 +3030,6 @@ try_mount_again:
 		else if (pSesInfo)
 			cifs_put_smb_ses(pSesInfo);
 
-		cifs_cleanup_volume_info(&volume_info);
 		FreeXid(xid);
 	}
 #endif
@@ -3004,6 +3045,7 @@ try_mount_again:
 	srvTcp = cifs_get_tcp_session(volume_info);
 	if (IS_ERR(srvTcp)) {
 		rc = PTR_ERR(srvTcp);
+		bdi_destroy(&cifs_sb->bdi);
 		goto out;
 	}
 
@@ -3015,14 +3057,6 @@ try_mount_again:
 		goto mount_fail_check;
 	}
 
-	if (pSesInfo->capabilities & CAP_LARGE_FILES)
-		sb->s_maxbytes = MAX_LFS_FILESIZE;
-	else
-		sb->s_maxbytes = MAX_NON_LFS;
-
-	/* BB FIXME fix time_gran to be larger for LANMAN sessions */
-	sb->s_time_gran = 100;
-
 	/* search for existing tcon to this server share */
 	tcon = cifs_get_tcon(pSesInfo, volume_info);
 	if (IS_ERR(tcon)) {
@@ -3035,7 +3069,7 @@ try_mount_again:
 	if (tcon->ses->capabilities & CAP_UNIX) {
 		/* reset of caps checks mount to see if unix extensions
 		   disabled for just this mount */
-		reset_cifs_unix_caps(xid, tcon, sb, volume_info);
+		reset_cifs_unix_caps(xid, tcon, cifs_sb, volume_info);
 		if ((tcon->ses->server->tcpStatus == CifsNeedReconnect) &&
 		    (le64_to_cpu(tcon->fsUnixInfo.Capability) &
 		     CIFS_UNIX_TRANSPORT_ENCRYPTION_MANDATORY_CAP)) {
@@ -3158,6 +3192,7 @@ mount_fail_check:
 			cifs_put_smb_ses(pSesInfo);
 		else
 			cifs_put_tcp_session(srvTcp);
+		bdi_destroy(&cifs_sb->bdi);
 		goto out;
 	}
 
@@ -3171,6 +3206,10 @@ out:
 	return rc;
 }
 
+/*
+ * Issue a TREE_CONNECT request. Note that for IPC$ shares, that the tcon
+ * pointer may be NULL.
+ */
 int
 CIFSTCon(unsigned int xid, struct cifs_ses *ses,
 	 const char *tree, struct cifs_tcon *tcon,
@@ -3205,7 +3244,7 @@ CIFSTCon(unsigned int xid, struct cifs_ses *ses,
 	pSMB->AndXCommand = 0xFF;
 	pSMB->Flags = cpu_to_le16(TCON_EXTENDED_SECINFO);
 	bcc_ptr = &pSMB->Password[0];
-	if ((ses->server->sec_mode) & SECMODE_USER) {
+	if (!tcon || (ses->server->sec_mode & SECMODE_USER)) {
 		pSMB->PasswordLength = cpu_to_le16(1);	/* minimum */
 		*bcc_ptr = 0; /* password is null byte */
 		bcc_ptr++;              /* skip password */
@@ -3328,8 +3367,8 @@ CIFSTCon(unsigned int xid, struct cifs_ses *ses,
 	return rc;
 }
 
-int
-cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb)
+void
+cifs_umount(struct cifs_sb_info *cifs_sb)
 {
 	struct rb_root *root = &cifs_sb->tlink_tree;
 	struct rb_node *node;
@@ -3350,7 +3389,10 @@ cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb)
 	}
 	spin_unlock(&cifs_sb->tlink_tree_lock);
 
-	return 0;
+	bdi_destroy(&cifs_sb->bdi);
+	kfree(cifs_sb->mountdata);
+	unload_nls(cifs_sb->local_nls);
+	kfree(cifs_sb);
 }
 
 int cifs_negotiate_protocol(unsigned int xid, struct cifs_ses *ses)
@@ -3371,7 +3413,7 @@ int cifs_negotiate_protocol(unsigned int xid, struct cifs_ses *ses)
 	}
 	if (rc == 0) {
 		spin_lock(&GlobalMid_Lock);
-		if (server->tcpStatus != CifsExiting)
+		if (server->tcpStatus == CifsNeedNegotiate)
 			server->tcpStatus = CifsGood;
 		else
 			rc = -EHOSTDOWN;
@@ -3444,7 +3486,7 @@ cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid)
 		goto out;
 	}
 
-	snprintf(username, MAX_USERNAME_SIZE, "krb50x%x", fsuid);
+	snprintf(username, sizeof(username), "krb50x%x", fsuid);
 	vol_info->username = username;
 	vol_info->local_nls = cifs_sb->local_nls;
 	vol_info->linux_uid = fsuid;
@@ -3484,12 +3526,6 @@ out:
 	return tcon;
 }
 
-static inline struct tcon_link *
-cifs_sb_master_tlink(struct cifs_sb_info *cifs_sb)
-{
-	return cifs_sb->master_tlink;
-}
-
 struct cifs_tcon *
 cifs_sb_master_tcon(struct cifs_sb_info *cifs_sb)
 {
diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c
index 81914df47ef..14d602f178c 100644
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -55,6 +55,7 @@ build_path_from_dentry(struct dentry *direntry)
 	char dirsep;
 	struct cifs_sb_info *cifs_sb = CIFS_SB(direntry->d_sb);
 	struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb);
+	unsigned seq;
 
 	if (direntry == NULL)
 		return NULL;  /* not much we can do if dentry is freed and
@@ -68,22 +69,29 @@ build_path_from_dentry(struct dentry *direntry)
 		dfsplen = 0;
 cifs_bp_rename_retry:
 	namelen = dfsplen;
+	seq = read_seqbegin(&rename_lock);
+	rcu_read_lock();
 	for (temp = direntry; !IS_ROOT(temp);) {
 		namelen += (1 + temp->d_name.len);
 		temp = temp->d_parent;
 		if (temp == NULL) {
 			cERROR(1, "corrupt dentry");
+			rcu_read_unlock();
 			return NULL;
 		}
 	}
+	rcu_read_unlock();
 
 	full_path = kmalloc(namelen+1, GFP_KERNEL);
 	if (full_path == NULL)
 		return full_path;
 	full_path[namelen] = 0;	/* trailing null */
+	rcu_read_lock();
 	for (temp = direntry; !IS_ROOT(temp);) {
+		spin_lock(&temp->d_lock);
 		namelen -= 1 + temp->d_name.len;
 		if (namelen < 0) {
+			spin_unlock(&temp->d_lock);
 			break;
 		} else {
 			full_path[namelen] = dirsep;
@@ -91,14 +99,17 @@ cifs_bp_rename_retry:
 				temp->d_name.len);
 			cFYI(0, "name: %s", full_path + namelen);
 		}
+		spin_unlock(&temp->d_lock);
 		temp = temp->d_parent;
 		if (temp == NULL) {
 			cERROR(1, "corrupt dentry");
+			rcu_read_unlock();
 			kfree(full_path);
 			return NULL;
 		}
 	}
-	if (namelen != dfsplen) {
+	rcu_read_unlock();
+	if (namelen != dfsplen || read_seqretry(&rename_lock, seq)) {
 		cERROR(1, "did not end path lookup where expected namelen is %d",
 			namelen);
 		/* presumably this is only possible if racing with a rename
@@ -168,7 +179,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode,
 	if (oplockEnabled)
 		oplock = REQ_OPLOCK;
 
-	if (nd && (nd->flags & LOOKUP_OPEN))
+	if (nd)
 		oflags = nd->intent.open.file->f_flags;
 	else
 		oflags = O_RDONLY | O_CREAT;
@@ -203,7 +214,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode,
 		   which should be rare for path not covered on files) */
 	}
 
-	if (nd && (nd->flags & LOOKUP_OPEN)) {
+	if (nd) {
 		/* if the file is going to stay open, then we
 		   need to set the desired access properly */
 		desiredAccess = 0;
@@ -317,7 +328,7 @@ cifs_create_set_dentry:
 	else
 		cFYI(1, "Create worked, get_inode_info failed rc = %d", rc);
 
-	if (newinode && nd && (nd->flags & LOOKUP_OPEN)) {
+	if (newinode && nd) {
 		struct cifsFileInfo *pfile_info;
 		struct file *filp;
 
@@ -557,7 +568,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry,
 	 * reduction in network traffic in the other paths.
 	 */
 	if (pTcon->unix_ext) {
-		if (nd && !(nd->flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY)) &&
+		if (nd && !(nd->flags & LOOKUP_DIRECTORY) &&
 		     (nd->flags & LOOKUP_OPEN) && !pTcon->broken_posix_open &&
 		     (nd->intent.open.file->f_flags & O_CREAT)) {
 			rc = cifs_posix_open(full_path, &newInode,
@@ -652,10 +663,8 @@ cifs_d_revalidate(struct dentry *direntry, struct nameidata *nd)
 	 * case sensitive name which is specified by user if this is
 	 * for creation.
 	 */
-	if (!(nd->flags & (LOOKUP_CONTINUE | LOOKUP_PARENT))) {
-		if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
-			return 0;
-	}
+	if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
+		return 0;
 
 	if (time_after(jiffies, direntry->d_time + HZ) || !lookupCacheEnabled)
 		return 0;
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index bb71471a4d9..378acdafa35 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -1401,7 +1401,8 @@ static int cifs_write_end(struct file *file, struct address_space *mapping,
 	return rc;
 }
 
-int cifs_strict_fsync(struct file *file, int datasync)
+int cifs_strict_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync)
 {
 	int xid;
 	int rc = 0;
@@ -1410,6 +1411,11 @@ int cifs_strict_fsync(struct file *file, int datasync)
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb);
 
+	rc = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (rc)
+		return rc;
+	mutex_lock(&inode->i_mutex);
+
 	xid = GetXid();
 
 	cFYI(1, "Sync file - name: %s datasync: 0x%x",
@@ -1428,16 +1434,23 @@ int cifs_strict_fsync(struct file *file, int datasync)
 		rc = CIFSSMBFlush(xid, tcon, smbfile->netfid);
 
 	FreeXid(xid);
+	mutex_unlock(&inode->i_mutex);
 	return rc;
 }
 
-int cifs_fsync(struct file *file, int datasync)
+int cifs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	int xid;
 	int rc = 0;
 	struct cifs_tcon *tcon;
 	struct cifsFileInfo *smbfile = file->private_data;
 	struct cifs_sb_info *cifs_sb = CIFS_SB(file->f_path.dentry->d_sb);
+	struct inode *inode = file->f_mapping->host;
+
+	rc = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (rc)
+		return rc;
+	mutex_lock(&inode->i_mutex);
 
 	xid = GetXid();
 
@@ -1449,6 +1462,7 @@ int cifs_fsync(struct file *file, int datasync)
 		rc = CIFSSMBFlush(xid, tcon, smbfile->netfid);
 
 	FreeXid(xid);
+	mutex_unlock(&inode->i_mutex);
 	return rc;
 }
 
@@ -1737,7 +1751,7 @@ cifs_iovec_read(struct file *file, const struct iovec *iov,
 			io_parms.pid = pid;
 			io_parms.tcon = pTcon;
 			io_parms.offset = *poffset;
-			io_parms.length = len;
+			io_parms.length = cur_len;
 			rc = CIFSSMBRead(xid, &io_parms, &bytes_read,
 					 &read_data, &buf_type);
 			pSMBr = (struct smb_com_read_rsp *)read_data;
diff --git a/fs/cifs/fscache.c b/fs/cifs/fscache.c
index d368a47ba5e..42e5363b410 100644
--- a/fs/cifs/fscache.c
+++ b/fs/cifs/fscache.c
@@ -28,14 +28,14 @@ void cifs_fscache_get_client_cookie(struct TCP_Server_Info *server)
 	server->fscache =
 		fscache_acquire_cookie(cifs_fscache_netfs.primary_index,
 				&cifs_fscache_server_index_def, server);
-	cFYI(1, "CIFS: get client cookie (0x%p/0x%p)", server,
-				server->fscache);
+	cFYI(1, "%s: (0x%p/0x%p)", __func__, server,
+			server->fscache);
 }
 
 void cifs_fscache_release_client_cookie(struct TCP_Server_Info *server)
 {
-	cFYI(1, "CIFS: release client cookie (0x%p/0x%p)", server,
-				server->fscache);
+	cFYI(1, "%s: (0x%p/0x%p)", __func__, server,
+			server->fscache);
 	fscache_relinquish_cookie(server->fscache, 0);
 	server->fscache = NULL;
 }
@@ -47,13 +47,13 @@ void cifs_fscache_get_super_cookie(struct cifs_tcon *tcon)
 	tcon->fscache =
 		fscache_acquire_cookie(server->fscache,
 				&cifs_fscache_super_index_def, tcon);
-	cFYI(1, "CIFS: get superblock cookie (0x%p/0x%p)",
-				server->fscache, tcon->fscache);
+	cFYI(1, "%s: (0x%p/0x%p)", __func__, server->fscache,
+			tcon->fscache);
 }
 
 void cifs_fscache_release_super_cookie(struct cifs_tcon *tcon)
 {
-	cFYI(1, "CIFS: releasing superblock cookie (0x%p)", tcon->fscache);
+	cFYI(1, "%s: (0x%p)", __func__, tcon->fscache);
 	fscache_relinquish_cookie(tcon->fscache, 0);
 	tcon->fscache = NULL;
 }
@@ -70,8 +70,8 @@ static void cifs_fscache_enable_inode_cookie(struct inode *inode)
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_FSCACHE) {
 		cifsi->fscache = fscache_acquire_cookie(tcon->fscache,
 				&cifs_fscache_inode_object_def, cifsi);
-		cFYI(1, "CIFS: got FH cookie (0x%p/0x%p)", tcon->fscache,
-				cifsi->fscache);
+		cFYI(1, "%s: got FH cookie (0x%p/0x%p)", __func__,
+				tcon->fscache, cifsi->fscache);
 	}
 }
 
@@ -80,8 +80,7 @@ void cifs_fscache_release_inode_cookie(struct inode *inode)
 	struct cifsInodeInfo *cifsi = CIFS_I(inode);
 
 	if (cifsi->fscache) {
-		cFYI(1, "CIFS releasing inode cookie (0x%p)",
-				cifsi->fscache);
+		cFYI(1, "%s: (0x%p)", __func__, cifsi->fscache);
 		fscache_relinquish_cookie(cifsi->fscache, 0);
 		cifsi->fscache = NULL;
 	}
@@ -92,8 +91,8 @@ static void cifs_fscache_disable_inode_cookie(struct inode *inode)
 	struct cifsInodeInfo *cifsi = CIFS_I(inode);
 
 	if (cifsi->fscache) {
-		cFYI(1, "CIFS disabling inode cookie (0x%p)",
-				cifsi->fscache);
+		cFYI(1, "%s: (0x%p)", __func__, cifsi->fscache);
+		fscache_uncache_all_inode_pages(cifsi->fscache, inode);
 		fscache_relinquish_cookie(cifsi->fscache, 1);
 		cifsi->fscache = NULL;
 	}
@@ -121,8 +120,8 @@ void cifs_fscache_reset_inode_cookie(struct inode *inode)
 					cifs_sb_master_tcon(cifs_sb)->fscache,
 					&cifs_fscache_inode_object_def,
 					cifsi);
-		cFYI(1, "CIFS: new cookie 0x%p oldcookie 0x%p",
-				cifsi->fscache, old);
+		cFYI(1, "%s: new cookie 0x%p oldcookie 0x%p",
+				__func__, cifsi->fscache, old);
 	}
 }
 
@@ -132,8 +131,8 @@ int cifs_fscache_release_page(struct page *page, gfp_t gfp)
 		struct inode *inode = page->mapping->host;
 		struct cifsInodeInfo *cifsi = CIFS_I(inode);
 
-		cFYI(1, "CIFS: fscache release page (0x%p/0x%p)",
-				page, cifsi->fscache);
+		cFYI(1, "%s: (0x%p/0x%p)", __func__, page,
+				cifsi->fscache);
 		if (!fscache_maybe_release_page(cifsi->fscache, page, gfp))
 			return 0;
 	}
@@ -144,8 +143,7 @@ int cifs_fscache_release_page(struct page *page, gfp_t gfp)
 static void cifs_readpage_from_fscache_complete(struct page *page, void *ctx,
 						int error)
 {
-	cFYI(1, "CFS: readpage_from_fscache_complete (0x%p/%d)",
-			page, error);
+	cFYI(1, "%s: (0x%p/%d)", __func__, page, error);
 	if (!error)
 		SetPageUptodate(page);
 	unlock_page(page);
@@ -158,7 +156,7 @@ int __cifs_readpage_from_fscache(struct inode *inode, struct page *page)
 {
 	int ret;
 
-	cFYI(1, "CIFS: readpage_from_fscache(fsc:%p, p:%p, i:0x%p",
+	cFYI(1, "%s: (fsc:%p, p:%p, i:0x%p", __func__,
 			CIFS_I(inode)->fscache, page, inode);
 	ret = fscache_read_or_alloc_page(CIFS_I(inode)->fscache, page,
 					 cifs_readpage_from_fscache_complete,
@@ -167,11 +165,11 @@ int __cifs_readpage_from_fscache(struct inode *inode, struct page *page)
 	switch (ret) {
 
 	case 0: /* page found in fscache, read submitted */
-		cFYI(1, "CIFS: readpage_from_fscache: submitted");
+		cFYI(1, "%s: submitted", __func__);
 		return ret;
 	case -ENOBUFS:	/* page won't be cached */
 	case -ENODATA:	/* page not in cache */
-		cFYI(1, "CIFS: readpage_from_fscache %d", ret);
+		cFYI(1, "%s: %d", __func__, ret);
 		return 1;
 
 	default:
@@ -190,7 +188,7 @@ int __cifs_readpages_from_fscache(struct inode *inode,
 {
 	int ret;
 
-	cFYI(1, "CIFS: __cifs_readpages_from_fscache (0x%p/%u/0x%p)",
+	cFYI(1, "%s: (0x%p/%u/0x%p)", __func__,
 			CIFS_I(inode)->fscache, *nr_pages, inode);
 	ret = fscache_read_or_alloc_pages(CIFS_I(inode)->fscache, mapping,
 					  pages, nr_pages,
@@ -199,12 +197,12 @@ int __cifs_readpages_from_fscache(struct inode *inode,
 					  mapping_gfp_mask(mapping));
 	switch (ret) {
 	case 0:	/* read submitted to the cache for all pages */
-		cFYI(1, "CIFS: readpages_from_fscache: submitted");
+		cFYI(1, "%s: submitted", __func__);
 		return ret;
 
 	case -ENOBUFS:	/* some pages are not cached and can't be */
 	case -ENODATA:	/* some pages are not cached */
-		cFYI(1, "CIFS: readpages_from_fscache: no page");
+		cFYI(1, "%s: no page", __func__);
 		return 1;
 
 	default:
@@ -218,7 +216,7 @@ void __cifs_readpage_to_fscache(struct inode *inode, struct page *page)
 {
 	int ret;
 
-	cFYI(1, "CIFS: readpage_to_fscache(fsc: %p, p: %p, i: %p",
+	cFYI(1, "%s: (fsc: %p, p: %p, i: %p)", __func__,
 			CIFS_I(inode)->fscache, page, inode);
 	ret = fscache_write_page(CIFS_I(inode)->fscache, page, GFP_KERNEL);
 	if (ret != 0)
@@ -230,7 +228,7 @@ void __cifs_fscache_invalidate_page(struct page *page, struct inode *inode)
 	struct cifsInodeInfo *cifsi = CIFS_I(inode);
 	struct fscache_cookie *cookie = cifsi->fscache;
 
-	cFYI(1, "CIFS: fscache invalidatepage (0x%p/0x%p)", page, cookie);
+	cFYI(1, "%s: (0x%p/0x%p)", __func__, page, cookie);
 	fscache_wait_on_page_write(cookie, page);
 	fscache_uncache_page(cookie, page);
 }
diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c
index 6751e745bbc..965a3af186a 100644
--- a/fs/cifs/readdir.c
+++ b/fs/cifs/readdir.c
@@ -796,7 +796,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
 		file->f_pos++;
 	case 1:
 		if (filldir(direntry, "..", 2, file->f_pos,
-		     file->f_path.dentry->d_parent->d_inode->i_ino, DT_DIR) < 0) {
+		     parent_ino(file->f_path.dentry), DT_DIR) < 0) {
 			cERROR(1, "Filldir for parent dir failed");
 			rc = -ENOMEM;
 			break;
diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c
index 3892ab817a3..d3e619692ee 100644
--- a/fs/cifs/sess.c
+++ b/fs/cifs/sess.c
@@ -428,8 +428,7 @@ static void build_ntlmssp_negotiate_blob(unsigned char *pbuffer,
 			(SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) {
 		flags |= NTLMSSP_NEGOTIATE_SIGN;
 		if (!ses->server->session_estab)
-			flags |= NTLMSSP_NEGOTIATE_KEY_XCH |
-				NTLMSSP_NEGOTIATE_EXTENDED_SEC;
+			flags |= NTLMSSP_NEGOTIATE_KEY_XCH;
 	}
 
 	sec_blob->NegotiateFlags = cpu_to_le32(flags);
@@ -465,10 +464,11 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer,
 		NTLMSSP_NEGOTIATE_128 | NTLMSSP_NEGOTIATE_UNICODE |
 		NTLMSSP_NEGOTIATE_NTLM | NTLMSSP_NEGOTIATE_EXTENDED_SEC;
 	if (ses->server->sec_mode &
-	   (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED))
+	   (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) {
 		flags |= NTLMSSP_NEGOTIATE_SIGN;
-	if (ses->server->sec_mode & SECMODE_SIGN_REQUIRED)
-		flags |= NTLMSSP_NEGOTIATE_ALWAYS_SIGN;
+		if (!ses->server->session_estab)
+			flags |= NTLMSSP_NEGOTIATE_KEY_XCH;
+	}
 
 	tmp = pbuffer + sizeof(AUTHENTICATE_MESSAGE);
 	sec_blob->NegotiateFlags = cpu_to_le32(flags);
diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c
index 1525d5e662b..1c5b770c314 100644
--- a/fs/cifs/smbencrypt.c
+++ b/fs/cifs/smbencrypt.c
@@ -90,12 +90,10 @@ smbhash(unsigned char *out, const unsigned char *in, unsigned char *key)
 	sg_init_one(&sgout, out, 8);
 
 	rc = crypto_blkcipher_encrypt(&desc, &sgout, &sgin, 8);
-	if (rc) {
+	if (rc)
 		cERROR(1, "could not encrypt crypt key rc: %d\n", rc);
-		crypto_free_blkcipher(tfm_des);
-		goto smbhash_err;
-	}
 
+	crypto_free_blkcipher(tfm_des);
 smbhash_err:
 	return rc;
 }
diff --git a/fs/coda/coda_int.h b/fs/coda/coda_int.h
index 6b443ff43a1..b7143cf783a 100644
--- a/fs/coda/coda_int.h
+++ b/fs/coda/coda_int.h
@@ -11,7 +11,7 @@ extern int coda_fake_statfs;
 
 void coda_destroy_inodecache(void);
 int coda_init_inodecache(void);
-int coda_fsync(struct file *coda_file, int datasync);
+int coda_fsync(struct file *coda_file, loff_t start, loff_t end, int datasync);
 void coda_sysctl_init(void);
 void coda_sysctl_clean(void);
 
diff --git a/fs/coda/coda_linux.h b/fs/coda/coda_linux.h
index 9b0c5323890..44e17e9c21a 100644
--- a/fs/coda/coda_linux.h
+++ b/fs/coda/coda_linux.h
@@ -39,7 +39,7 @@ extern const struct file_operations coda_ioctl_operations;
 /* operations shared over more than one file */
 int coda_open(struct inode *i, struct file *f);
 int coda_release(struct inode *i, struct file *f);
-int coda_permission(struct inode *inode, int mask, unsigned int flags);
+int coda_permission(struct inode *inode, int mask);
 int coda_revalidate_inode(struct dentry *);
 int coda_getattr(struct vfsmount *, struct dentry *, struct kstat *);
 int coda_setattr(struct dentry *, struct iattr *);
diff --git a/fs/coda/dir.c b/fs/coda/dir.c
index 2b8dae4d121..0239433f50c 100644
--- a/fs/coda/dir.c
+++ b/fs/coda/dir.c
@@ -132,11 +132,11 @@ exit:
 }
 
 
-int coda_permission(struct inode *inode, int mask, unsigned int flags)
+int coda_permission(struct inode *inode, int mask)
 {
 	int error;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
@@ -449,8 +449,7 @@ static int coda_venus_readdir(struct file *coda_file, void *buf,
 	struct file *host_file;
 	struct dentry *de;
 	struct venus_dirent *vdir;
-	unsigned long vdir_size =
-	    (unsigned long)(&((struct venus_dirent *)0)->d_name);
+	unsigned long vdir_size = offsetof(struct venus_dirent, d_name);
 	unsigned int type;
 	struct qstr name;
 	ino_t ino;
@@ -474,7 +473,7 @@ static int coda_venus_readdir(struct file *coda_file, void *buf,
 		coda_file->f_pos++;
 	}
 	if (coda_file->f_pos == 1) {
-		ret = filldir(buf, "..", 2, 1, de->d_parent->d_inode->i_ino, DT_DIR);
+		ret = filldir(buf, "..", 2, 1, parent_ino(de), DT_DIR);
 		if (ret < 0)
 			goto out;
 		result++;
diff --git a/fs/coda/file.c b/fs/coda/file.c
index 0433057be33..8edd404e641 100644
--- a/fs/coda/file.c
+++ b/fs/coda/file.c
@@ -199,7 +199,7 @@ int coda_release(struct inode *coda_inode, struct file *coda_file)
 	return 0;
 }
 
-int coda_fsync(struct file *coda_file, int datasync)
+int coda_fsync(struct file *coda_file, loff_t start, loff_t end, int datasync)
 {
 	struct file *host_file;
 	struct inode *coda_inode = coda_file->f_path.dentry->d_inode;
@@ -210,6 +210,11 @@ int coda_fsync(struct file *coda_file, int datasync)
 	      S_ISLNK(coda_inode->i_mode)))
 		return -EINVAL;
 
+	err = filemap_write_and_wait_range(coda_inode->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&coda_inode->i_mutex);
+
 	cfi = CODA_FTOC(coda_file);
 	BUG_ON(!cfi || cfi->cfi_magic != CODA_MAGIC);
 	host_file = cfi->cfi_container;
@@ -217,6 +222,7 @@ int coda_fsync(struct file *coda_file, int datasync)
 	err = vfs_fsync(host_file, datasync);
 	if (!err && !datasync)
 		err = venus_fsync(coda_inode->i_sb, coda_i2f(coda_inode));
+	mutex_unlock(&coda_inode->i_mutex);
 
 	return err;
 }
diff --git a/fs/coda/pioctl.c b/fs/coda/pioctl.c
index 6cbb3afb36d..ee0981f1375 100644
--- a/fs/coda/pioctl.c
+++ b/fs/coda/pioctl.c
@@ -24,7 +24,7 @@
 #include "coda_linux.h"
 
 /* pioctl ops */
-static int coda_ioctl_permission(struct inode *inode, int mask, unsigned int flags);
+static int coda_ioctl_permission(struct inode *inode, int mask);
 static long coda_pioctl(struct file *filp, unsigned int cmd,
 			unsigned long user_data);
 
@@ -41,10 +41,8 @@ const struct file_operations coda_ioctl_operations = {
 };
 
 /* the coda pioctl inode ops */
-static int coda_ioctl_permission(struct inode *inode, int mask, unsigned int flags)
+static int coda_ioctl_permission(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
 	return (mask & MAY_EXEC) ? -EACCES : 0;
 }
 
diff --git a/fs/cramfs/inode.c b/fs/cramfs/inode.c
index e141939080f..739fb59bcdc 100644
--- a/fs/cramfs/inode.c
+++ b/fs/cramfs/inode.c
@@ -37,7 +37,7 @@ static DEFINE_MUTEX(read_mutex);
 /* These macros may change in future, to provide better st_ino semantics. */
 #define OFFSET(x)	((x)->i_ino)
 
-static unsigned long cramino(struct cramfs_inode *cino, unsigned int offset)
+static unsigned long cramino(const struct cramfs_inode *cino, unsigned int offset)
 {
 	if (!cino->offset)
 		return offset + 1;
@@ -61,7 +61,7 @@ static unsigned long cramino(struct cramfs_inode *cino, unsigned int offset)
 }
 
 static struct inode *get_cramfs_inode(struct super_block *sb,
-	struct cramfs_inode *cramfs_inode, unsigned int offset)
+	const struct cramfs_inode *cramfs_inode, unsigned int offset)
 {
 	struct inode *inode;
 	static struct timespec zerotime;
@@ -317,7 +317,7 @@ static int cramfs_fill_super(struct super_block *sb, void *data, int silent)
 	/* Set it all up.. */
 	sb->s_op = &cramfs_ops;
 	root = get_cramfs_inode(sb, &super.root, 0);
-	if (!root)
+	if (IS_ERR(root))
 		goto out;
 	sb->s_root = d_alloc_root(root);
 	if (!sb->s_root) {
@@ -423,6 +423,7 @@ static int cramfs_readdir(struct file *filp, void *dirent, filldir_t filldir)
 static struct dentry * cramfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
 {
 	unsigned int offset = 0;
+	struct inode *inode = NULL;
 	int sorted;
 
 	mutex_lock(&read_mutex);
@@ -449,8 +450,8 @@ static struct dentry * cramfs_lookup(struct inode *dir, struct dentry *dentry, s
 
 		for (;;) {
 			if (!namelen) {
-				mutex_unlock(&read_mutex);
-				return ERR_PTR(-EIO);
+				inode = ERR_PTR(-EIO);
+				goto out;
 			}
 			if (name[namelen-1])
 				break;
@@ -462,17 +463,18 @@ static struct dentry * cramfs_lookup(struct inode *dir, struct dentry *dentry, s
 		if (retval > 0)
 			continue;
 		if (!retval) {
-			struct cramfs_inode entry = *de;
-			mutex_unlock(&read_mutex);
-			d_add(dentry, get_cramfs_inode(dir->i_sb, &entry, dir_off));
-			return NULL;
+			inode = get_cramfs_inode(dir->i_sb, de, dir_off);
+			break;
 		}
 		/* else (retval < 0) */
 		if (sorted)
 			break;
 	}
+out:
 	mutex_unlock(&read_mutex);
-	d_add(dentry, NULL);
+	if (IS_ERR(inode))
+		return ERR_CAST(inode);
+	d_add(dentry, inode);
 	return NULL;
 }
 
diff --git a/fs/dcache.c b/fs/dcache.c
index 37f72ee5bf7..be18598c7fd 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -344,6 +344,24 @@ void d_drop(struct dentry *dentry)
 EXPORT_SYMBOL(d_drop);
 
 /*
+ * d_clear_need_lookup - drop a dentry from cache and clear the need lookup flag
+ * @dentry: dentry to drop
+ *
+ * This is called when we do a lookup on a placeholder dentry that needed to be
+ * looked up.  The dentry should have been hashed in order for it to be found by
+ * the lookup code, but now needs to be unhashed while we do the actual lookup
+ * and clear the DCACHE_NEED_LOOKUP flag.
+ */
+void d_clear_need_lookup(struct dentry *dentry)
+{
+	spin_lock(&dentry->d_lock);
+	__d_drop(dentry);
+	dentry->d_flags &= ~DCACHE_NEED_LOOKUP;
+	spin_unlock(&dentry->d_lock);
+}
+EXPORT_SYMBOL(d_clear_need_lookup);
+
+/*
  * Finish off a dentry we've decided to kill.
  * dentry->d_lock must be held, returns with it unlocked.
  * If ref is non-zero, then decrement the refcount too.
@@ -432,8 +450,13 @@ repeat:
  	if (d_unhashed(dentry))
 		goto kill_it;
 
-	/* Otherwise leave it cached and ensure it's on the LRU */
-	dentry->d_flags |= DCACHE_REFERENCED;
+	/*
+	 * If this dentry needs lookup, don't set the referenced flag so that it
+	 * is more likely to be cleaned up by the dcache shrinker in case of
+	 * memory pressure.
+	 */
+	if (!d_need_lookup(dentry))
+		dentry->d_flags |= DCACHE_REFERENCED;
 	dentry_lru_add(dentry);
 
 	dentry->d_count--;
@@ -526,10 +549,6 @@ repeat:
 	 */
 	rcu_read_lock();
 	ret = dentry->d_parent;
-	if (!ret) {
-		rcu_read_unlock();
-		goto out;
-	}
 	spin_lock(&ret->d_lock);
 	if (unlikely(ret != dentry->d_parent)) {
 		spin_unlock(&ret->d_lock);
@@ -540,7 +559,6 @@ repeat:
 	BUG_ON(!ret->d_count);
 	ret->d_count++;
 	spin_unlock(&ret->d_lock);
-out:
 	return ret;
 }
 EXPORT_SYMBOL(dget_parent);
@@ -720,13 +738,11 @@ static void shrink_dentry_list(struct list_head *list)
  *
  * If flags contains DCACHE_REFERENCED reference dentries will not be pruned.
  */
-static void __shrink_dcache_sb(struct super_block *sb, int *count, int flags)
+static void __shrink_dcache_sb(struct super_block *sb, int count, int flags)
 {
-	/* called from prune_dcache() and shrink_dcache_parent() */
 	struct dentry *dentry;
 	LIST_HEAD(referenced);
 	LIST_HEAD(tmp);
-	int cnt = *count;
 
 relock:
 	spin_lock(&dcache_lru_lock);
@@ -754,7 +770,7 @@ relock:
 		} else {
 			list_move_tail(&dentry->d_lru, &tmp);
 			spin_unlock(&dentry->d_lock);
-			if (!--cnt)
+			if (!--count)
 				break;
 		}
 		cond_resched_lock(&dcache_lru_lock);
@@ -764,83 +780,22 @@ relock:
 	spin_unlock(&dcache_lru_lock);
 
 	shrink_dentry_list(&tmp);
-
-	*count = cnt;
 }
 
 /**
- * prune_dcache - shrink the dcache
- * @count: number of entries to try to free
+ * prune_dcache_sb - shrink the dcache
+ * @nr_to_scan: number of entries to try to free
  *
- * Shrink the dcache. This is done when we need more memory, or simply when we
- * need to unmount something (at which point we need to unuse all dentries).
+ * Attempt to shrink the superblock dcache LRU by @nr_to_scan entries. This is
+ * done when we need more memory an called from the superblock shrinker
+ * function.
  *
- * This function may fail to free any resources if all the dentries are in use.
+ * This function may fail to free any resources if all the dentries are in
+ * use.
  */
-static void prune_dcache(int count)
+void prune_dcache_sb(struct super_block *sb, int nr_to_scan)
 {
-	struct super_block *sb, *p = NULL;
-	int w_count;
-	int unused = dentry_stat.nr_unused;
-	int prune_ratio;
-	int pruned;
-
-	if (unused == 0 || count == 0)
-		return;
-	if (count >= unused)
-		prune_ratio = 1;
-	else
-		prune_ratio = unused / count;
-	spin_lock(&sb_lock);
-	list_for_each_entry(sb, &super_blocks, s_list) {
-		if (list_empty(&sb->s_instances))
-			continue;
-		if (sb->s_nr_dentry_unused == 0)
-			continue;
-		sb->s_count++;
-		/* Now, we reclaim unused dentrins with fairness.
-		 * We reclaim them same percentage from each superblock.
-		 * We calculate number of dentries to scan on this sb
-		 * as follows, but the implementation is arranged to avoid
-		 * overflows:
-		 * number of dentries to scan on this sb =
-		 * count * (number of dentries on this sb /
-		 * number of dentries in the machine)
-		 */
-		spin_unlock(&sb_lock);
-		if (prune_ratio != 1)
-			w_count = (sb->s_nr_dentry_unused / prune_ratio) + 1;
-		else
-			w_count = sb->s_nr_dentry_unused;
-		pruned = w_count;
-		/*
-		 * We need to be sure this filesystem isn't being unmounted,
-		 * otherwise we could race with generic_shutdown_super(), and
-		 * end up holding a reference to an inode while the filesystem
-		 * is unmounted.  So we try to get s_umount, and make sure
-		 * s_root isn't NULL.
-		 */
-		if (down_read_trylock(&sb->s_umount)) {
-			if ((sb->s_root != NULL) &&
-			    (!list_empty(&sb->s_dentry_lru))) {
-				__shrink_dcache_sb(sb, &w_count,
-						DCACHE_REFERENCED);
-				pruned -= w_count;
-			}
-			up_read(&sb->s_umount);
-		}
-		spin_lock(&sb_lock);
-		if (p)
-			__put_super(p);
-		count -= pruned;
-		p = sb;
-		/* more work left to do? */
-		if (count <= 0)
-			break;
-	}
-	if (p)
-		__put_super(p);
-	spin_unlock(&sb_lock);
+	__shrink_dcache_sb(sb, nr_to_scan, DCACHE_REFERENCED);
 }
 
 /**
@@ -1215,45 +1170,13 @@ void shrink_dcache_parent(struct dentry * parent)
 	int found;
 
 	while ((found = select_parent(parent)) != 0)
-		__shrink_dcache_sb(sb, &found, 0);
+		__shrink_dcache_sb(sb, found, 0);
 }
 EXPORT_SYMBOL(shrink_dcache_parent);
 
-/*
- * Scan `sc->nr_slab_to_reclaim' dentries and return the number which remain.
- *
- * We need to avoid reentering the filesystem if the caller is performing a
- * GFP_NOFS allocation attempt.  One example deadlock is:
- *
- * ext2_new_block->getblk->GFP->shrink_dcache_memory->prune_dcache->
- * prune_one_dentry->dput->dentry_iput->iput->inode->i_sb->s_op->put_inode->
- * ext2_discard_prealloc->ext2_free_blocks->lock_super->DEADLOCK.
- *
- * In this case we return -1 to tell the caller that we baled.
- */
-static int shrink_dcache_memory(struct shrinker *shrink,
-				struct shrink_control *sc)
-{
-	int nr = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
-
-	if (nr) {
-		if (!(gfp_mask & __GFP_FS))
-			return -1;
-		prune_dcache(nr);
-	}
-
-	return (dentry_stat.nr_unused / 100) * sysctl_vfs_cache_pressure;
-}
-
-static struct shrinker dcache_shrinker = {
-	.shrink = shrink_dcache_memory,
-	.seeks = DEFAULT_SEEKS,
-};
-
 /**
- * d_alloc	-	allocate a dcache entry
- * @parent: parent of entry to allocate
+ * __d_alloc	-	allocate a dcache entry
+ * @sb: filesystem it will belong to
  * @name: qstr of the name
  *
  * Allocates a dentry. It returns %NULL if there is insufficient memory
@@ -1261,7 +1184,7 @@ static struct shrinker dcache_shrinker = {
  * copied and the copy passed in may be reused after this call.
  */
  
-struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
+struct dentry *__d_alloc(struct super_block *sb, const struct qstr *name)
 {
 	struct dentry *dentry;
 	char *dname;
@@ -1291,8 +1214,8 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
 	spin_lock_init(&dentry->d_lock);
 	seqcount_init(&dentry->d_seq);
 	dentry->d_inode = NULL;
-	dentry->d_parent = NULL;
-	dentry->d_sb = NULL;
+	dentry->d_parent = dentry;
+	dentry->d_sb = sb;
 	dentry->d_op = NULL;
 	dentry->d_fsdata = NULL;
 	INIT_HLIST_BL_NODE(&dentry->d_hash);
@@ -1300,36 +1223,47 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
 	INIT_LIST_HEAD(&dentry->d_subdirs);
 	INIT_LIST_HEAD(&dentry->d_alias);
 	INIT_LIST_HEAD(&dentry->d_u.d_child);
-
-	if (parent) {
-		spin_lock(&parent->d_lock);
-		/*
-		 * don't need child lock because it is not subject
-		 * to concurrency here
-		 */
-		__dget_dlock(parent);
-		dentry->d_parent = parent;
-		dentry->d_sb = parent->d_sb;
-		d_set_d_op(dentry, dentry->d_sb->s_d_op);
-		list_add(&dentry->d_u.d_child, &parent->d_subdirs);
-		spin_unlock(&parent->d_lock);
-	}
+	d_set_d_op(dentry, dentry->d_sb->s_d_op);
 
 	this_cpu_inc(nr_dentry);
 
 	return dentry;
 }
+
+/**
+ * d_alloc	-	allocate a dcache entry
+ * @parent: parent of entry to allocate
+ * @name: qstr of the name
+ *
+ * Allocates a dentry. It returns %NULL if there is insufficient memory
+ * available. On a success the dentry is returned. The name passed in is
+ * copied and the copy passed in may be reused after this call.
+ */
+struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
+{
+	struct dentry *dentry = __d_alloc(parent->d_sb, name);
+	if (!dentry)
+		return NULL;
+
+	spin_lock(&parent->d_lock);
+	/*
+	 * don't need child lock because it is not subject
+	 * to concurrency here
+	 */
+	__dget_dlock(parent);
+	dentry->d_parent = parent;
+	list_add(&dentry->d_u.d_child, &parent->d_subdirs);
+	spin_unlock(&parent->d_lock);
+
+	return dentry;
+}
 EXPORT_SYMBOL(d_alloc);
 
 struct dentry *d_alloc_pseudo(struct super_block *sb, const struct qstr *name)
 {
-	struct dentry *dentry = d_alloc(NULL, name);
-	if (dentry) {
-		dentry->d_sb = sb;
-		d_set_d_op(dentry, dentry->d_sb->s_d_op);
-		dentry->d_parent = dentry;
+	struct dentry *dentry = __d_alloc(sb, name);
+	if (dentry)
 		dentry->d_flags |= DCACHE_DISCONNECTED;
-	}
 	return dentry;
 }
 EXPORT_SYMBOL(d_alloc_pseudo);
@@ -1499,13 +1433,9 @@ struct dentry * d_alloc_root(struct inode * root_inode)
 	if (root_inode) {
 		static const struct qstr name = { .name = "/", .len = 1 };
 
-		res = d_alloc(NULL, &name);
-		if (res) {
-			res->d_sb = root_inode->i_sb;
-			d_set_d_op(res, res->d_sb->s_d_op);
-			res->d_parent = res;
+		res = __d_alloc(root_inode->i_sb, &name);
+		if (res)
 			d_instantiate(res, root_inode);
-		}
 	}
 	return res;
 }
@@ -1566,13 +1496,11 @@ struct dentry *d_obtain_alias(struct inode *inode)
 	if (res)
 		goto out_iput;
 
-	tmp = d_alloc(NULL, &anonstring);
+	tmp = __d_alloc(inode->i_sb, &anonstring);
 	if (!tmp) {
 		res = ERR_PTR(-ENOMEM);
 		goto out_iput;
 	}
-	tmp->d_parent = tmp; /* make sure dput doesn't croak */
-
 
 	spin_lock(&inode->i_lock);
 	res = __d_find_any_alias(inode);
@@ -1584,8 +1512,6 @@ struct dentry *d_obtain_alias(struct inode *inode)
 
 	/* attach a disconnected dentry */
 	spin_lock(&tmp->d_lock);
-	tmp->d_sb = inode->i_sb;
-	d_set_d_op(tmp, tmp->d_sb->s_d_op);
 	tmp->d_inode = inode;
 	tmp->d_flags |= DCACHE_DISCONNECTED;
 	list_add(&tmp->d_alias, &inode->i_dentry);
@@ -1626,6 +1552,9 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry)
 {
 	struct dentry *new = NULL;
 
+	if (IS_ERR(inode))
+		return ERR_CAST(inode);
+
 	if (inode && S_ISDIR(inode->i_mode)) {
 		spin_lock(&inode->i_lock);
 		new = __d_find_alias(inode, 1);
@@ -1708,29 +1637,22 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
 	}
 
 	/*
-	 * Negative dentry: instantiate it unless the inode is a directory and
-	 * already has a dentry.
+	 * We are going to instantiate this dentry, unhash it and clear the
+	 * lookup flag so we can do that.
 	 */
-	spin_lock(&inode->i_lock);
-	if (!S_ISDIR(inode->i_mode) || list_empty(&inode->i_dentry)) {
-		__d_instantiate(found, inode);
-		spin_unlock(&inode->i_lock);
-		security_d_instantiate(found, inode);
-		return found;
-	}
+	if (unlikely(d_need_lookup(found)))
+		d_clear_need_lookup(found);
 
 	/*
-	 * In case a directory already has a (disconnected) entry grab a
-	 * reference to it, move it in place and use it.
+	 * Negative dentry: instantiate it unless the inode is a directory and
+	 * already has a dentry.
 	 */
-	new = list_entry(inode->i_dentry.next, struct dentry, d_alias);
-	__dget(new);
-	spin_unlock(&inode->i_lock);
-	security_d_instantiate(found, inode);
-	d_move(new, found);
-	iput(inode);
-	dput(found);
-	return new;
+	new = d_splice_alias(inode, found);
+	if (new) {
+		dput(found);
+		found = new;
+	}
+	return found;
 
 err_out:
 	iput(inode);
@@ -1813,8 +1735,6 @@ seqretry:
 		tname = dentry->d_name.name;
 		i = dentry->d_inode;
 		prefetch(tname);
-		if (i)
-			prefetch(i);
 		/*
 		 * This seqcount check is required to ensure name and
 		 * len are loaded atomically, so as not to walk off the
@@ -2213,14 +2133,15 @@ static void dentry_unlock_parents_for_move(struct dentry *dentry,
  * The hash value has to match the hash queue that the dentry is on..
  */
 /*
- * d_move - move a dentry
+ * __d_move - move a dentry
  * @dentry: entry to move
  * @target: new dentry
  *
  * Update the dcache to reflect the move of a file name. Negative
- * dcache entries should not be moved in this way.
+ * dcache entries should not be moved in this way.  Caller hold
+ * rename_lock.
  */
-void d_move(struct dentry * dentry, struct dentry * target)
+static void __d_move(struct dentry * dentry, struct dentry * target)
 {
 	if (!dentry->d_inode)
 		printk(KERN_WARNING "VFS: moving negative dcache entry\n");
@@ -2228,8 +2149,6 @@ void d_move(struct dentry * dentry, struct dentry * target)
 	BUG_ON(d_ancestor(dentry, target));
 	BUG_ON(d_ancestor(target, dentry));
 
-	write_seqlock(&rename_lock);
-
 	dentry_lock_for_move(dentry, target);
 
 	write_seqcount_begin(&dentry->d_seq);
@@ -2275,6 +2194,20 @@ void d_move(struct dentry * dentry, struct dentry * target)
 	spin_unlock(&target->d_lock);
 	fsnotify_d_move(dentry);
 	spin_unlock(&dentry->d_lock);
+}
+
+/*
+ * d_move - move a dentry
+ * @dentry: entry to move
+ * @target: new dentry
+ *
+ * Update the dcache to reflect the move of a file name. Negative
+ * dcache entries should not be moved in this way.
+ */
+void d_move(struct dentry *dentry, struct dentry *target)
+{
+	write_seqlock(&rename_lock);
+	__d_move(dentry, target);
 	write_sequnlock(&rename_lock);
 }
 EXPORT_SYMBOL(d_move);
@@ -2302,7 +2235,7 @@ struct dentry *d_ancestor(struct dentry *p1, struct dentry *p2)
  * This helper attempts to cope with remotely renamed directories
  *
  * It assumes that the caller is already holding
- * dentry->d_parent->d_inode->i_mutex and the inode->i_lock
+ * dentry->d_parent->d_inode->i_mutex, inode->i_lock and rename_lock
  *
  * Note: If ever the locking in lock_rename() changes, then please
  * remember to update this too...
@@ -2317,11 +2250,6 @@ static struct dentry *__d_unalias(struct inode *inode,
 	if (alias->d_parent == dentry->d_parent)
 		goto out_unalias;
 
-	/* Check for loops */
-	ret = ERR_PTR(-ELOOP);
-	if (d_ancestor(alias, dentry))
-		goto out_err;
-
 	/* See lock_rename() */
 	ret = ERR_PTR(-EBUSY);
 	if (!mutex_trylock(&dentry->d_sb->s_vfs_rename_mutex))
@@ -2331,7 +2259,7 @@ static struct dentry *__d_unalias(struct inode *inode,
 		goto out_err;
 	m2 = &alias->d_parent->d_inode->i_mutex;
 out_unalias:
-	d_move(alias, dentry);
+	__d_move(alias, dentry);
 	ret = alias;
 out_err:
 	spin_unlock(&inode->i_lock);
@@ -2416,15 +2344,24 @@ struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode)
 		alias = __d_find_alias(inode, 0);
 		if (alias) {
 			actual = alias;
-			/* Is this an anonymous mountpoint that we could splice
-			 * into our tree? */
-			if (IS_ROOT(alias)) {
+			write_seqlock(&rename_lock);
+
+			if (d_ancestor(alias, dentry)) {
+				/* Check for loops */
+				actual = ERR_PTR(-ELOOP);
+			} else if (IS_ROOT(alias)) {
+				/* Is this an anonymous mountpoint that we
+				 * could splice into our tree? */
 				__d_materialise_dentry(dentry, alias);
+				write_sequnlock(&rename_lock);
 				__d_drop(alias);
 				goto found;
+			} else {
+				/* Nope, but we must(!) avoid directory
+				 * aliasing */
+				actual = __d_unalias(inode, dentry, alias);
 			}
-			/* Nope, but we must(!) avoid directory aliasing */
-			actual = __d_unalias(inode, dentry, alias);
+			write_sequnlock(&rename_lock);
 			if (IS_ERR(actual))
 				dput(alias);
 			goto out_nolock;
@@ -3030,8 +2967,6 @@ static void __init dcache_init(void)
 	 */
 	dentry_cache = KMEM_CACHE(dentry,
 		SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|SLAB_MEM_SPREAD);
-	
-	register_shrinker(&dcache_shrinker);
 
 	/* Hash may have been set up in dcache_init_early */
 	if (!hashdist)
diff --git a/fs/dcookies.c b/fs/dcookies.c
index a21cabdbd87..dda0dc702d1 100644
--- a/fs/dcookies.c
+++ b/fs/dcookies.c
@@ -178,6 +178,8 @@ SYSCALL_DEFINE(lookup_dcookie)(u64 cookie64, char __user * buf, size_t len)
 	/* FIXME: (deleted) ? */
 	path = d_path(&dcs->path, kbuf, PAGE_SIZE);
 
+	mutex_unlock(&dcookie_mutex);
+
 	if (IS_ERR(path)) {
 		err = PTR_ERR(path);
 		goto out_free;
@@ -194,6 +196,7 @@ SYSCALL_DEFINE(lookup_dcookie)(u64 cookie64, char __user * buf, size_t len)
 
 out_free:
 	kfree(kbuf);
+	return err;
 out:
 	mutex_unlock(&dcookie_mutex);
 	return err;
diff --git a/fs/direct-io.c b/fs/direct-io.c
index ac5f164170e..01d2d9ef609 100644
--- a/fs/direct-io.c
+++ b/fs/direct-io.c
@@ -135,6 +135,50 @@ struct dio {
 	struct page *pages[DIO_PAGES];	/* page buffer */
 };
 
+static void __inode_dio_wait(struct inode *inode)
+{
+	wait_queue_head_t *wq = bit_waitqueue(&inode->i_state, __I_DIO_WAKEUP);
+	DEFINE_WAIT_BIT(q, &inode->i_state, __I_DIO_WAKEUP);
+
+	do {
+		prepare_to_wait(wq, &q.wait, TASK_UNINTERRUPTIBLE);
+		if (atomic_read(&inode->i_dio_count))
+			schedule();
+	} while (atomic_read(&inode->i_dio_count));
+	finish_wait(wq, &q.wait);
+}
+
+/**
+ * inode_dio_wait - wait for outstanding DIO requests to finish
+ * @inode: inode to wait for
+ *
+ * Waits for all pending direct I/O requests to finish so that we can
+ * proceed with a truncate or equivalent operation.
+ *
+ * Must be called under a lock that serializes taking new references
+ * to i_dio_count, usually by inode->i_mutex.
+ */
+void inode_dio_wait(struct inode *inode)
+{
+	if (atomic_read(&inode->i_dio_count))
+		__inode_dio_wait(inode);
+}
+EXPORT_SYMBOL_GPL(inode_dio_wait);
+
+/*
+ * inode_dio_done - signal finish of a direct I/O requests
+ * @inode: inode the direct I/O happens on
+ *
+ * This is called once we've finished processing a direct I/O request,
+ * and is used to wake up callers waiting for direct I/O to be quiesced.
+ */
+void inode_dio_done(struct inode *inode)
+{
+	if (atomic_dec_and_test(&inode->i_dio_count))
+		wake_up_bit(&inode->i_state, __I_DIO_WAKEUP);
+}
+EXPORT_SYMBOL_GPL(inode_dio_done);
+
 /*
  * How many pages are in the queue?
  */
@@ -249,14 +293,12 @@ static ssize_t dio_complete(struct dio *dio, loff_t offset, ssize_t ret, bool is
 	if (dio->end_io && dio->result) {
 		dio->end_io(dio->iocb, offset, transferred,
 			    dio->map_bh.b_private, ret, is_async);
-	} else if (is_async) {
-		aio_complete(dio->iocb, ret, 0);
+	} else {
+		if (is_async)
+			aio_complete(dio->iocb, ret, 0);
+		inode_dio_done(dio->inode);
 	}
 
-	if (dio->flags & DIO_LOCKING)
-		/* lockdep: non-owner release */
-		up_read_non_owner(&dio->inode->i_alloc_sem);
-
 	return ret;
 }
 
@@ -980,9 +1022,6 @@ out:
 	return ret;
 }
 
-/*
- * Releases both i_mutex and i_alloc_sem
- */
 static ssize_t
 direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode, 
 	const struct iovec *iov, loff_t offset, unsigned long nr_segs, 
@@ -1146,15 +1185,16 @@ direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode,
  *    For writes this function is called under i_mutex and returns with
  *    i_mutex held, for reads, i_mutex is not held on entry, but it is
  *    taken and dropped again before returning.
- *    For reads and writes i_alloc_sem is taken in shared mode and released
- *    on I/O completion (which may happen asynchronously after returning to
- *    the caller).
- *
  *  - if the flags value does NOT contain DIO_LOCKING we don't use any
  *    internal locking but rather rely on the filesystem to synchronize
  *    direct I/O reads/writes versus each other and truncate.
- *    For reads and writes both i_mutex and i_alloc_sem are not held on
- *    entry and are never taken.
+ *
+ * To help with locking against truncate we incremented the i_dio_count
+ * counter before starting direct I/O, and decrement it once we are done.
+ * Truncate can wait for it to reach zero to provide exclusion.  It is
+ * expected that filesystem provide exclusion between new direct I/O
+ * and truncates.  For DIO_LOCKING filesystems this is done by i_mutex,
+ * but other filesystems need to take care of this on their own.
  */
 ssize_t
 __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
@@ -1200,6 +1240,10 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 		}
 	}
 
+	/* watch out for a 0 len io from a tricksy fs */
+	if (rw == READ && end == offset)
+		return 0;
+
 	dio = kmalloc(sizeof(*dio), GFP_KERNEL);
 	retval = -ENOMEM;
 	if (!dio)
@@ -1213,8 +1257,7 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 
 	dio->flags = flags;
 	if (dio->flags & DIO_LOCKING) {
-		/* watch out for a 0 len io from a tricksy fs */
-		if (rw == READ && end > offset) {
+		if (rw == READ) {
 			struct address_space *mapping =
 					iocb->ki_filp->f_mapping;
 
@@ -1229,15 +1272,14 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 				goto out;
 			}
 		}
-
-		/*
-		 * Will be released at I/O completion, possibly in a
-		 * different thread.
-		 */
-		down_read_non_owner(&inode->i_alloc_sem);
 	}
 
 	/*
+	 * Will be decremented at I/O completion time.
+	 */
+	atomic_inc(&inode->i_dio_count);
+
+	/*
 	 * For file extending writes updating i_size before data
 	 * writeouts complete can expose uninitialized blocks. So
 	 * even for AIO, we need to wait for i/o to complete before
diff --git a/fs/dlm/ast.c b/fs/dlm/ast.c
index abc49f29245..90e5997262e 100644
--- a/fs/dlm/ast.c
+++ b/fs/dlm/ast.c
@@ -14,17 +14,9 @@
 #include "dlm_internal.h"
 #include "lock.h"
 #include "user.h"
-#include "ast.h"
-
-#define WAKE_ASTS  0
-
-static uint64_t			ast_seq_count;
-static struct list_head		ast_queue;
-static spinlock_t		ast_queue_lock;
-static struct task_struct *	astd_task;
-static unsigned long		astd_wakeflags;
-static struct mutex		astd_running;
 
+static uint64_t			dlm_cb_seq;
+static spinlock_t		dlm_cb_seq_spin;
 
 static void dlm_dump_lkb_callbacks(struct dlm_lkb *lkb)
 {
@@ -57,21 +49,13 @@ static void dlm_dump_lkb_callbacks(struct dlm_lkb *lkb)
 	}
 }
 
-void dlm_del_ast(struct dlm_lkb *lkb)
-{
-	spin_lock(&ast_queue_lock);
-	if (!list_empty(&lkb->lkb_astqueue))
-		list_del_init(&lkb->lkb_astqueue);
-	spin_unlock(&ast_queue_lock);
-}
-
 int dlm_add_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode,
 			 int status, uint32_t sbflags, uint64_t seq)
 {
 	struct dlm_ls *ls = lkb->lkb_resource->res_ls;
 	uint64_t prev_seq;
 	int prev_mode;
-	int i;
+	int i, rv;
 
 	for (i = 0; i < DLM_CALLBACKS_SIZE; i++) {
 		if (lkb->lkb_callbacks[i].seq)
@@ -100,7 +84,8 @@ int dlm_add_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode,
 					  mode,
 					  (unsigned long long)prev_seq,
 					  prev_mode);
-				return 0;
+				rv = 0;
+				goto out;
 			}
 		}
 
@@ -109,6 +94,7 @@ int dlm_add_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode,
 		lkb->lkb_callbacks[i].mode = mode;
 		lkb->lkb_callbacks[i].sb_status = status;
 		lkb->lkb_callbacks[i].sb_flags = (sbflags & 0x000000FF);
+		rv = 0;
 		break;
 	}
 
@@ -117,21 +103,24 @@ int dlm_add_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode,
 			  lkb->lkb_id, (unsigned long long)seq,
 			  flags, mode, status, sbflags);
 		dlm_dump_lkb_callbacks(lkb);
-		return -1;
+		rv = -1;
+		goto out;
 	}
-
-	return 0;
+ out:
+	return rv;
 }
 
 int dlm_rem_lkb_callback(struct dlm_ls *ls, struct dlm_lkb *lkb,
 			 struct dlm_callback *cb, int *resid)
 {
-	int i;
+	int i, rv;
 
 	*resid = 0;
 
-	if (!lkb->lkb_callbacks[0].seq)
-		return -ENOENT;
+	if (!lkb->lkb_callbacks[0].seq) {
+		rv = -ENOENT;
+		goto out;
+	}
 
 	/* oldest undelivered cb is callbacks[0] */
 
@@ -163,7 +152,8 @@ int dlm_rem_lkb_callback(struct dlm_ls *ls, struct dlm_lkb *lkb,
 				  cb->mode,
 				  (unsigned long long)lkb->lkb_last_cast.seq,
 				  lkb->lkb_last_cast.mode);
-			return 0;
+			rv = 0;
+			goto out;
 		}
 	}
 
@@ -176,171 +166,150 @@ int dlm_rem_lkb_callback(struct dlm_ls *ls, struct dlm_lkb *lkb,
 		memcpy(&lkb->lkb_last_bast, cb, sizeof(struct dlm_callback));
 		lkb->lkb_last_bast_time = ktime_get();
 	}
-
-	return 0;
+	rv = 0;
+ out:
+	return rv;
 }
 
-void dlm_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode, int status,
-		 uint32_t sbflags)
+void dlm_add_cb(struct dlm_lkb *lkb, uint32_t flags, int mode, int status,
+		uint32_t sbflags)
 {
-	uint64_t seq;
+	struct dlm_ls *ls = lkb->lkb_resource->res_ls;
+	uint64_t new_seq, prev_seq;
 	int rv;
 
-	spin_lock(&ast_queue_lock);
-
-	seq = ++ast_seq_count;
+	spin_lock(&dlm_cb_seq_spin);
+	new_seq = ++dlm_cb_seq;
+	spin_unlock(&dlm_cb_seq_spin);
 
 	if (lkb->lkb_flags & DLM_IFL_USER) {
-		spin_unlock(&ast_queue_lock);
-		dlm_user_add_ast(lkb, flags, mode, status, sbflags, seq);
+		dlm_user_add_ast(lkb, flags, mode, status, sbflags, new_seq);
 		return;
 	}
 
-	rv = dlm_add_lkb_callback(lkb, flags, mode, status, sbflags, seq);
-	if (rv < 0) {
-		spin_unlock(&ast_queue_lock);
-		return;
-	}
+	mutex_lock(&lkb->lkb_cb_mutex);
+	prev_seq = lkb->lkb_callbacks[0].seq;
 
-	if (list_empty(&lkb->lkb_astqueue)) {
+	rv = dlm_add_lkb_callback(lkb, flags, mode, status, sbflags, new_seq);
+	if (rv < 0)
+		goto out;
+
+	if (!prev_seq) {
 		kref_get(&lkb->lkb_ref);
-		list_add_tail(&lkb->lkb_astqueue, &ast_queue);
-	}
-	spin_unlock(&ast_queue_lock);
 
-	set_bit(WAKE_ASTS, &astd_wakeflags);
-	wake_up_process(astd_task);
+		if (test_bit(LSFL_CB_DELAY, &ls->ls_flags)) {
+			mutex_lock(&ls->ls_cb_mutex);
+			list_add(&lkb->lkb_cb_list, &ls->ls_cb_delay);
+			mutex_unlock(&ls->ls_cb_mutex);
+		} else {
+			queue_work(ls->ls_callback_wq, &lkb->lkb_cb_work);
+		}
+	}
+ out:
+	mutex_unlock(&lkb->lkb_cb_mutex);
 }
 
-static void process_asts(void)
+void dlm_callback_work(struct work_struct *work)
 {
-	struct dlm_ls *ls = NULL;
-	struct dlm_rsb *r = NULL;
-	struct dlm_lkb *lkb;
+	struct dlm_lkb *lkb = container_of(work, struct dlm_lkb, lkb_cb_work);
+	struct dlm_ls *ls = lkb->lkb_resource->res_ls;
 	void (*castfn) (void *astparam);
 	void (*bastfn) (void *astparam, int mode);
 	struct dlm_callback callbacks[DLM_CALLBACKS_SIZE];
 	int i, rv, resid;
 
-repeat:
-	spin_lock(&ast_queue_lock);
-	list_for_each_entry(lkb, &ast_queue, lkb_astqueue) {
-		r = lkb->lkb_resource;
-		ls = r->res_ls;
+	memset(&callbacks, 0, sizeof(callbacks));
 
-		if (dlm_locking_stopped(ls))
-			continue;
-
-		/* we remove from astqueue list and remove everything in
-		   lkb_callbacks before releasing the spinlock so empty
-		   lkb_astqueue is always consistent with empty lkb_callbacks */
-
-		list_del_init(&lkb->lkb_astqueue);
-
-		castfn = lkb->lkb_astfn;
-		bastfn = lkb->lkb_bastfn;
+	mutex_lock(&lkb->lkb_cb_mutex);
+	if (!lkb->lkb_callbacks[0].seq) {
+		/* no callback work exists, shouldn't happen */
+		log_error(ls, "dlm_callback_work %x no work", lkb->lkb_id);
+		dlm_print_lkb(lkb);
+		dlm_dump_lkb_callbacks(lkb);
+	}
 
-		memset(&callbacks, 0, sizeof(callbacks));
+	for (i = 0; i < DLM_CALLBACKS_SIZE; i++) {
+		rv = dlm_rem_lkb_callback(ls, lkb, &callbacks[i], &resid);
+		if (rv < 0)
+			break;
+	}
 
-		for (i = 0; i < DLM_CALLBACKS_SIZE; i++) {
-			rv = dlm_rem_lkb_callback(ls, lkb, &callbacks[i], &resid);
-			if (rv < 0)
-				break;
-		}
-		spin_unlock(&ast_queue_lock);
+	if (resid) {
+		/* cbs remain, loop should have removed all, shouldn't happen */
+		log_error(ls, "dlm_callback_work %x resid %d", lkb->lkb_id,
+			  resid);
+		dlm_print_lkb(lkb);
+		dlm_dump_lkb_callbacks(lkb);
+	}
+	mutex_unlock(&lkb->lkb_cb_mutex);
 
-		if (resid) {
-			/* shouldn't happen, for loop should have removed all */
-			log_error(ls, "callback resid %d lkb %x",
-				  resid, lkb->lkb_id);
-		}
+	castfn = lkb->lkb_astfn;
+	bastfn = lkb->lkb_bastfn;
 
-		for (i = 0; i < DLM_CALLBACKS_SIZE; i++) {
-			if (!callbacks[i].seq)
-				break;
-			if (callbacks[i].flags & DLM_CB_SKIP) {
-				continue;
-			} else if (callbacks[i].flags & DLM_CB_BAST) {
-				bastfn(lkb->lkb_astparam, callbacks[i].mode);
-			} else if (callbacks[i].flags & DLM_CB_CAST) {
-				lkb->lkb_lksb->sb_status = callbacks[i].sb_status;
-				lkb->lkb_lksb->sb_flags = callbacks[i].sb_flags;
-				castfn(lkb->lkb_astparam);
-			}
+	for (i = 0; i < DLM_CALLBACKS_SIZE; i++) {
+		if (!callbacks[i].seq)
+			break;
+		if (callbacks[i].flags & DLM_CB_SKIP) {
+			continue;
+		} else if (callbacks[i].flags & DLM_CB_BAST) {
+			bastfn(lkb->lkb_astparam, callbacks[i].mode);
+		} else if (callbacks[i].flags & DLM_CB_CAST) {
+			lkb->lkb_lksb->sb_status = callbacks[i].sb_status;
+			lkb->lkb_lksb->sb_flags = callbacks[i].sb_flags;
+			castfn(lkb->lkb_astparam);
 		}
-
-		/* removes ref for ast_queue, may cause lkb to be freed */
-		dlm_put_lkb(lkb);
-
-		cond_resched();
-		goto repeat;
 	}
-	spin_unlock(&ast_queue_lock);
-}
-
-static inline int no_asts(void)
-{
-	int ret;
 
-	spin_lock(&ast_queue_lock);
-	ret = list_empty(&ast_queue);
-	spin_unlock(&ast_queue_lock);
-	return ret;
+	/* undo kref_get from dlm_add_callback, may cause lkb to be freed */
+	dlm_put_lkb(lkb);
 }
 
-static int dlm_astd(void *data)
+int dlm_callback_start(struct dlm_ls *ls)
 {
-	while (!kthread_should_stop()) {
-		set_current_state(TASK_INTERRUPTIBLE);
-		if (!test_bit(WAKE_ASTS, &astd_wakeflags))
-			schedule();
-		set_current_state(TASK_RUNNING);
-
-		mutex_lock(&astd_running);
-		if (test_and_clear_bit(WAKE_ASTS, &astd_wakeflags))
-			process_asts();
-		mutex_unlock(&astd_running);
+	ls->ls_callback_wq = alloc_workqueue("dlm_callback",
+					     WQ_UNBOUND |
+					     WQ_MEM_RECLAIM |
+					     WQ_NON_REENTRANT,
+					     0);
+	if (!ls->ls_callback_wq) {
+		log_print("can't start dlm_callback workqueue");
+		return -ENOMEM;
 	}
 	return 0;
 }
 
-void dlm_astd_wake(void)
+void dlm_callback_stop(struct dlm_ls *ls)
 {
-	if (!no_asts()) {
-		set_bit(WAKE_ASTS, &astd_wakeflags);
-		wake_up_process(astd_task);
-	}
+	if (ls->ls_callback_wq)
+		destroy_workqueue(ls->ls_callback_wq);
 }
 
-int dlm_astd_start(void)
+void dlm_callback_suspend(struct dlm_ls *ls)
 {
-	struct task_struct *p;
-	int error = 0;
-
-	INIT_LIST_HEAD(&ast_queue);
-	spin_lock_init(&ast_queue_lock);
-	mutex_init(&astd_running);
-
-	p = kthread_run(dlm_astd, NULL, "dlm_astd");
-	if (IS_ERR(p))
-		error = PTR_ERR(p);
-	else
-		astd_task = p;
-	return error;
-}
+	set_bit(LSFL_CB_DELAY, &ls->ls_flags);
 
-void dlm_astd_stop(void)
-{
-	kthread_stop(astd_task);
+	if (ls->ls_callback_wq)
+		flush_workqueue(ls->ls_callback_wq);
 }
 
-void dlm_astd_suspend(void)
+void dlm_callback_resume(struct dlm_ls *ls)
 {
-	mutex_lock(&astd_running);
-}
+	struct dlm_lkb *lkb, *safe;
+	int count = 0;
 
-void dlm_astd_resume(void)
-{
-	mutex_unlock(&astd_running);
+	clear_bit(LSFL_CB_DELAY, &ls->ls_flags);
+
+	if (!ls->ls_callback_wq)
+		return;
+
+	mutex_lock(&ls->ls_cb_mutex);
+	list_for_each_entry_safe(lkb, safe, &ls->ls_cb_delay, lkb_cb_list) {
+		list_del_init(&lkb->lkb_cb_list);
+		queue_work(ls->ls_callback_wq, &lkb->lkb_cb_work);
+		count++;
+	}
+	mutex_unlock(&ls->ls_cb_mutex);
+
+	log_debug(ls, "dlm_callback_resume %d", count);
 }
 
diff --git a/fs/dlm/ast.h b/fs/dlm/ast.h
index 8aa89c9b561..757b551c682 100644
--- a/fs/dlm/ast.h
+++ b/fs/dlm/ast.h
@@ -18,14 +18,15 @@ int dlm_add_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode,
                          int status, uint32_t sbflags, uint64_t seq);
 int dlm_rem_lkb_callback(struct dlm_ls *ls, struct dlm_lkb *lkb,
                          struct dlm_callback *cb, int *resid);
-void dlm_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode, int status,
-		 uint32_t sbflags);
+void dlm_add_cb(struct dlm_lkb *lkb, uint32_t flags, int mode, int status,
+                uint32_t sbflags);
 
-void dlm_astd_wake(void);
-int dlm_astd_start(void);
-void dlm_astd_stop(void);
-void dlm_astd_suspend(void);
-void dlm_astd_resume(void);
+void dlm_callback_work(struct work_struct *work);
+int dlm_callback_start(struct dlm_ls *ls);
+void dlm_callback_stop(struct dlm_ls *ls);
+void dlm_callback_suspend(struct dlm_ls *ls);
+void dlm_callback_resume(struct dlm_ls *ls);
 
 #endif
 
+
diff --git a/fs/dlm/config.c b/fs/dlm/config.c
index 9b026ea8baa..6cf72fcc0d0 100644
--- a/fs/dlm/config.c
+++ b/fs/dlm/config.c
@@ -28,7 +28,8 @@
  * /config/dlm/<cluster>/spaces/<space>/nodes/<node>/weight
  * /config/dlm/<cluster>/comms/<comm>/nodeid
  * /config/dlm/<cluster>/comms/<comm>/local
- * /config/dlm/<cluster>/comms/<comm>/addr
+ * /config/dlm/<cluster>/comms/<comm>/addr      (write only)
+ * /config/dlm/<cluster>/comms/<comm>/addr_list (read only)
  * The <cluster> level is useless, but I haven't figured out how to avoid it.
  */
 
@@ -80,6 +81,7 @@ static ssize_t comm_local_write(struct dlm_comm *cm, const char *buf,
 				size_t len);
 static ssize_t comm_addr_write(struct dlm_comm *cm, const char *buf,
 				size_t len);
+static ssize_t comm_addr_list_read(struct dlm_comm *cm, char *buf);
 static ssize_t node_nodeid_read(struct dlm_node *nd, char *buf);
 static ssize_t node_nodeid_write(struct dlm_node *nd, const char *buf,
 				size_t len);
@@ -92,7 +94,6 @@ struct dlm_cluster {
 	unsigned int cl_tcp_port;
 	unsigned int cl_buffer_size;
 	unsigned int cl_rsbtbl_size;
-	unsigned int cl_lkbtbl_size;
 	unsigned int cl_dirtbl_size;
 	unsigned int cl_recover_timer;
 	unsigned int cl_toss_secs;
@@ -101,13 +102,13 @@ struct dlm_cluster {
 	unsigned int cl_protocol;
 	unsigned int cl_timewarn_cs;
 	unsigned int cl_waitwarn_us;
+	unsigned int cl_new_rsb_count;
 };
 
 enum {
 	CLUSTER_ATTR_TCP_PORT = 0,
 	CLUSTER_ATTR_BUFFER_SIZE,
 	CLUSTER_ATTR_RSBTBL_SIZE,
-	CLUSTER_ATTR_LKBTBL_SIZE,
 	CLUSTER_ATTR_DIRTBL_SIZE,
 	CLUSTER_ATTR_RECOVER_TIMER,
 	CLUSTER_ATTR_TOSS_SECS,
@@ -116,6 +117,7 @@ enum {
 	CLUSTER_ATTR_PROTOCOL,
 	CLUSTER_ATTR_TIMEWARN_CS,
 	CLUSTER_ATTR_WAITWARN_US,
+	CLUSTER_ATTR_NEW_RSB_COUNT,
 };
 
 struct cluster_attribute {
@@ -160,7 +162,6 @@ __CONFIGFS_ATTR(name, 0644, name##_read, name##_write)
 CLUSTER_ATTR(tcp_port, 1);
 CLUSTER_ATTR(buffer_size, 1);
 CLUSTER_ATTR(rsbtbl_size, 1);
-CLUSTER_ATTR(lkbtbl_size, 1);
 CLUSTER_ATTR(dirtbl_size, 1);
 CLUSTER_ATTR(recover_timer, 1);
 CLUSTER_ATTR(toss_secs, 1);
@@ -169,12 +170,12 @@ CLUSTER_ATTR(log_debug, 0);
 CLUSTER_ATTR(protocol, 0);
 CLUSTER_ATTR(timewarn_cs, 1);
 CLUSTER_ATTR(waitwarn_us, 0);
+CLUSTER_ATTR(new_rsb_count, 0);
 
 static struct configfs_attribute *cluster_attrs[] = {
 	[CLUSTER_ATTR_TCP_PORT] = &cluster_attr_tcp_port.attr,
 	[CLUSTER_ATTR_BUFFER_SIZE] = &cluster_attr_buffer_size.attr,
 	[CLUSTER_ATTR_RSBTBL_SIZE] = &cluster_attr_rsbtbl_size.attr,
-	[CLUSTER_ATTR_LKBTBL_SIZE] = &cluster_attr_lkbtbl_size.attr,
 	[CLUSTER_ATTR_DIRTBL_SIZE] = &cluster_attr_dirtbl_size.attr,
 	[CLUSTER_ATTR_RECOVER_TIMER] = &cluster_attr_recover_timer.attr,
 	[CLUSTER_ATTR_TOSS_SECS] = &cluster_attr_toss_secs.attr,
@@ -183,6 +184,7 @@ static struct configfs_attribute *cluster_attrs[] = {
 	[CLUSTER_ATTR_PROTOCOL] = &cluster_attr_protocol.attr,
 	[CLUSTER_ATTR_TIMEWARN_CS] = &cluster_attr_timewarn_cs.attr,
 	[CLUSTER_ATTR_WAITWARN_US] = &cluster_attr_waitwarn_us.attr,
+	[CLUSTER_ATTR_NEW_RSB_COUNT] = &cluster_attr_new_rsb_count.attr,
 	NULL,
 };
 
@@ -190,6 +192,7 @@ enum {
 	COMM_ATTR_NODEID = 0,
 	COMM_ATTR_LOCAL,
 	COMM_ATTR_ADDR,
+	COMM_ATTR_ADDR_LIST,
 };
 
 struct comm_attribute {
@@ -217,14 +220,22 @@ static struct comm_attribute comm_attr_local = {
 static struct comm_attribute comm_attr_addr = {
 	.attr   = { .ca_owner = THIS_MODULE,
                     .ca_name = "addr",
-                    .ca_mode = S_IRUGO | S_IWUSR },
+                    .ca_mode = S_IWUSR },
 	.store  = comm_addr_write,
 };
 
+static struct comm_attribute comm_attr_addr_list = {
+	.attr   = { .ca_owner = THIS_MODULE,
+                    .ca_name = "addr_list",
+                    .ca_mode = S_IRUGO },
+	.show   = comm_addr_list_read,
+};
+
 static struct configfs_attribute *comm_attrs[] = {
 	[COMM_ATTR_NODEID] = &comm_attr_nodeid.attr,
 	[COMM_ATTR_LOCAL] = &comm_attr_local.attr,
 	[COMM_ATTR_ADDR] = &comm_attr_addr.attr,
+	[COMM_ATTR_ADDR_LIST] = &comm_attr_addr_list.attr,
 	NULL,
 };
 
@@ -435,7 +446,6 @@ static struct config_group *make_cluster(struct config_group *g,
 	cl->cl_tcp_port = dlm_config.ci_tcp_port;
 	cl->cl_buffer_size = dlm_config.ci_buffer_size;
 	cl->cl_rsbtbl_size = dlm_config.ci_rsbtbl_size;
-	cl->cl_lkbtbl_size = dlm_config.ci_lkbtbl_size;
 	cl->cl_dirtbl_size = dlm_config.ci_dirtbl_size;
 	cl->cl_recover_timer = dlm_config.ci_recover_timer;
 	cl->cl_toss_secs = dlm_config.ci_toss_secs;
@@ -444,6 +454,7 @@ static struct config_group *make_cluster(struct config_group *g,
 	cl->cl_protocol = dlm_config.ci_protocol;
 	cl->cl_timewarn_cs = dlm_config.ci_timewarn_cs;
 	cl->cl_waitwarn_us = dlm_config.ci_waitwarn_us;
+	cl->cl_new_rsb_count = dlm_config.ci_new_rsb_count;
 
 	space_list = &sps->ss_group;
 	comm_list = &cms->cs_group;
@@ -720,6 +731,50 @@ static ssize_t comm_addr_write(struct dlm_comm *cm, const char *buf, size_t len)
 	return len;
 }
 
+static ssize_t comm_addr_list_read(struct dlm_comm *cm, char *buf)
+{
+	ssize_t s;
+	ssize_t allowance;
+	int i;
+	struct sockaddr_storage *addr;
+	struct sockaddr_in *addr_in;
+	struct sockaddr_in6 *addr_in6;
+	
+	/* Taken from ip6_addr_string() defined in lib/vsprintf.c */
+	char buf0[sizeof("AF_INET6	xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:255.255.255.255\n")];
+	
+
+	/* Derived from SIMPLE_ATTR_SIZE of fs/configfs/file.c */
+	allowance = 4096;
+	buf[0] = '\0';
+
+	for (i = 0; i < cm->addr_count; i++) {
+		addr = cm->addr[i];
+
+		switch(addr->ss_family) {
+		case AF_INET:
+			addr_in = (struct sockaddr_in *)addr;
+			s = sprintf(buf0, "AF_INET	%pI4\n", &addr_in->sin_addr.s_addr);
+			break;
+		case AF_INET6:
+			addr_in6 = (struct sockaddr_in6 *)addr;
+			s = sprintf(buf0, "AF_INET6	%pI6\n", &addr_in6->sin6_addr);
+			break;
+		default:
+			s = sprintf(buf0, "%s\n", "<UNKNOWN>");
+			break;
+		}
+		allowance -= s;
+		if (allowance >= 0)
+			strcat(buf, buf0);
+		else {
+			allowance += s;
+			break;
+		}
+	}
+	return 4096 - allowance;
+}
+
 static ssize_t show_node(struct config_item *i, struct configfs_attribute *a,
 			 char *buf)
 {
@@ -983,7 +1038,6 @@ int dlm_our_addr(struct sockaddr_storage *addr, int num)
 #define DEFAULT_TCP_PORT       21064
 #define DEFAULT_BUFFER_SIZE     4096
 #define DEFAULT_RSBTBL_SIZE     1024
-#define DEFAULT_LKBTBL_SIZE     1024
 #define DEFAULT_DIRTBL_SIZE     1024
 #define DEFAULT_RECOVER_TIMER      5
 #define DEFAULT_TOSS_SECS         10
@@ -992,12 +1046,12 @@ int dlm_our_addr(struct sockaddr_storage *addr, int num)
 #define DEFAULT_PROTOCOL           0
 #define DEFAULT_TIMEWARN_CS      500 /* 5 sec = 500 centiseconds */
 #define DEFAULT_WAITWARN_US	   0
+#define DEFAULT_NEW_RSB_COUNT    128
 
 struct dlm_config_info dlm_config = {
 	.ci_tcp_port = DEFAULT_TCP_PORT,
 	.ci_buffer_size = DEFAULT_BUFFER_SIZE,
 	.ci_rsbtbl_size = DEFAULT_RSBTBL_SIZE,
-	.ci_lkbtbl_size = DEFAULT_LKBTBL_SIZE,
 	.ci_dirtbl_size = DEFAULT_DIRTBL_SIZE,
 	.ci_recover_timer = DEFAULT_RECOVER_TIMER,
 	.ci_toss_secs = DEFAULT_TOSS_SECS,
@@ -1005,6 +1059,7 @@ struct dlm_config_info dlm_config = {
 	.ci_log_debug = DEFAULT_LOG_DEBUG,
 	.ci_protocol = DEFAULT_PROTOCOL,
 	.ci_timewarn_cs = DEFAULT_TIMEWARN_CS,
-	.ci_waitwarn_us = DEFAULT_WAITWARN_US
+	.ci_waitwarn_us = DEFAULT_WAITWARN_US,
+	.ci_new_rsb_count = DEFAULT_NEW_RSB_COUNT
 };
 
diff --git a/fs/dlm/config.h b/fs/dlm/config.h
index dd0ce24d5a8..3099d0dd26c 100644
--- a/fs/dlm/config.h
+++ b/fs/dlm/config.h
@@ -20,7 +20,6 @@ struct dlm_config_info {
 	int ci_tcp_port;
 	int ci_buffer_size;
 	int ci_rsbtbl_size;
-	int ci_lkbtbl_size;
 	int ci_dirtbl_size;
 	int ci_recover_timer;
 	int ci_toss_secs;
@@ -29,6 +28,7 @@ struct dlm_config_info {
 	int ci_protocol;
 	int ci_timewarn_cs;
 	int ci_waitwarn_us;
+	int ci_new_rsb_count;
 };
 
 extern struct dlm_config_info dlm_config;
diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h
index 0262451eb9c..fe2860c0244 100644
--- a/fs/dlm/dlm_internal.h
+++ b/fs/dlm/dlm_internal.h
@@ -37,6 +37,7 @@
 #include <linux/jhash.h>
 #include <linux/miscdevice.h>
 #include <linux/mutex.h>
+#include <linux/idr.h>
 #include <asm/uaccess.h>
 
 #include <linux/dlm.h>
@@ -52,7 +53,6 @@ struct dlm_ls;
 struct dlm_lkb;
 struct dlm_rsb;
 struct dlm_member;
-struct dlm_lkbtable;
 struct dlm_rsbtable;
 struct dlm_dirtable;
 struct dlm_direntry;
@@ -108,11 +108,6 @@ struct dlm_rsbtable {
 	spinlock_t		lock;
 };
 
-struct dlm_lkbtable {
-	struct list_head	list;
-	rwlock_t		lock;
-	uint16_t		counter;
-};
 
 /*
  * Lockspace member (per node in a ls)
@@ -248,17 +243,18 @@ struct dlm_lkb {
 	int8_t			lkb_wait_count;
 	int			lkb_wait_nodeid; /* for debugging */
 
-	struct list_head	lkb_idtbl_list;	/* lockspace lkbtbl */
 	struct list_head	lkb_statequeue;	/* rsb g/c/w list */
 	struct list_head	lkb_rsb_lookup;	/* waiting for rsb lookup */
 	struct list_head	lkb_wait_reply;	/* waiting for remote reply */
-	struct list_head	lkb_astqueue;	/* need ast to be sent */
 	struct list_head	lkb_ownqueue;	/* list of locks for a process */
 	struct list_head	lkb_time_list;
 	ktime_t			lkb_timestamp;
 	ktime_t			lkb_wait_time;
 	unsigned long		lkb_timeout_cs;
 
+	struct mutex		lkb_cb_mutex;
+	struct work_struct	lkb_cb_work;
+	struct list_head	lkb_cb_list; /* for ls_cb_delay or proc->asts */
 	struct dlm_callback	lkb_callbacks[DLM_CALLBACKS_SIZE];
 	struct dlm_callback	lkb_last_cast;
 	struct dlm_callback	lkb_last_bast;
@@ -299,7 +295,7 @@ struct dlm_rsb {
 	int			res_recover_locks_count;
 
 	char			*res_lvbptr;
-	char			res_name[1];
+	char			res_name[DLM_RESNAME_MAXLEN+1];
 };
 
 /* find_rsb() flags */
@@ -465,12 +461,12 @@ struct dlm_ls {
 	unsigned long		ls_scan_time;
 	struct kobject		ls_kobj;
 
+	struct idr		ls_lkbidr;
+	spinlock_t		ls_lkbidr_spin;
+
 	struct dlm_rsbtable	*ls_rsbtbl;
 	uint32_t		ls_rsbtbl_size;
 
-	struct dlm_lkbtable	*ls_lkbtbl;
-	uint32_t		ls_lkbtbl_size;
-
 	struct dlm_dirtable	*ls_dirtbl;
 	uint32_t		ls_dirtbl_size;
 
@@ -483,6 +479,10 @@ struct dlm_ls {
 	struct mutex		ls_timeout_mutex;
 	struct list_head	ls_timeout;
 
+	spinlock_t		ls_new_rsb_spin;
+	int			ls_new_rsb_count;
+	struct list_head	ls_new_rsb;	/* new rsb structs */
+
 	struct list_head	ls_nodes;	/* current nodes in ls */
 	struct list_head	ls_nodes_gone;	/* dead node list, recovery */
 	int			ls_num_nodes;	/* number of nodes in ls */
@@ -506,8 +506,12 @@ struct dlm_ls {
 
 	struct miscdevice       ls_device;
 
+	struct workqueue_struct	*ls_callback_wq;
+
 	/* recovery related */
 
+	struct mutex		ls_cb_mutex;
+	struct list_head	ls_cb_delay; /* save for queue_work later */
 	struct timer_list	ls_timer;
 	struct task_struct	*ls_recoverd_task;
 	struct mutex		ls_recoverd_active;
@@ -544,6 +548,7 @@ struct dlm_ls {
 #define LSFL_RCOM_WAIT		4
 #define LSFL_UEVENT_WAIT	5
 #define LSFL_TIMEWARN		6
+#define LSFL_CB_DELAY		7
 
 /* much of this is just saving user space pointers associated with the
    lock that we pass back to the user lib with an ast */
diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c
index f71d0b5abd9..83b5e32514e 100644
--- a/fs/dlm/lock.c
+++ b/fs/dlm/lock.c
@@ -305,7 +305,7 @@ static void queue_cast(struct dlm_rsb *r, struct dlm_lkb *lkb, int rv)
 		rv = -EDEADLK;
 	}
 
-	dlm_add_ast(lkb, DLM_CB_CAST, lkb->lkb_grmode, rv, lkb->lkb_sbflags);
+	dlm_add_cb(lkb, DLM_CB_CAST, lkb->lkb_grmode, rv, lkb->lkb_sbflags);
 }
 
 static inline void queue_cast_overlap(struct dlm_rsb *r, struct dlm_lkb *lkb)
@@ -319,7 +319,7 @@ static void queue_bast(struct dlm_rsb *r, struct dlm_lkb *lkb, int rqmode)
 	if (is_master_copy(lkb)) {
 		send_bast(r, lkb, rqmode);
 	} else {
-		dlm_add_ast(lkb, DLM_CB_BAST, rqmode, 0, 0);
+		dlm_add_cb(lkb, DLM_CB_BAST, rqmode, 0, 0);
 	}
 }
 
@@ -327,19 +327,68 @@ static void queue_bast(struct dlm_rsb *r, struct dlm_lkb *lkb, int rqmode)
  * Basic operations on rsb's and lkb's
  */
 
-static struct dlm_rsb *create_rsb(struct dlm_ls *ls, char *name, int len)
+static int pre_rsb_struct(struct dlm_ls *ls)
+{
+	struct dlm_rsb *r1, *r2;
+	int count = 0;
+
+	spin_lock(&ls->ls_new_rsb_spin);
+	if (ls->ls_new_rsb_count > dlm_config.ci_new_rsb_count / 2) {
+		spin_unlock(&ls->ls_new_rsb_spin);
+		return 0;
+	}
+	spin_unlock(&ls->ls_new_rsb_spin);
+
+	r1 = dlm_allocate_rsb(ls);
+	r2 = dlm_allocate_rsb(ls);
+
+	spin_lock(&ls->ls_new_rsb_spin);
+	if (r1) {
+		list_add(&r1->res_hashchain, &ls->ls_new_rsb);
+		ls->ls_new_rsb_count++;
+	}
+	if (r2) {
+		list_add(&r2->res_hashchain, &ls->ls_new_rsb);
+		ls->ls_new_rsb_count++;
+	}
+	count = ls->ls_new_rsb_count;
+	spin_unlock(&ls->ls_new_rsb_spin);
+
+	if (!count)
+		return -ENOMEM;
+	return 0;
+}
+
+/* If ls->ls_new_rsb is empty, return -EAGAIN, so the caller can
+   unlock any spinlocks, go back and call pre_rsb_struct again.
+   Otherwise, take an rsb off the list and return it. */
+
+static int get_rsb_struct(struct dlm_ls *ls, char *name, int len,
+			  struct dlm_rsb **r_ret)
 {
 	struct dlm_rsb *r;
+	int count;
 
-	r = dlm_allocate_rsb(ls, len);
-	if (!r)
-		return NULL;
+	spin_lock(&ls->ls_new_rsb_spin);
+	if (list_empty(&ls->ls_new_rsb)) {
+		count = ls->ls_new_rsb_count;
+		spin_unlock(&ls->ls_new_rsb_spin);
+		log_debug(ls, "find_rsb retry %d %d %s",
+			  count, dlm_config.ci_new_rsb_count, name);
+		return -EAGAIN;
+	}
+
+	r = list_first_entry(&ls->ls_new_rsb, struct dlm_rsb, res_hashchain);
+	list_del(&r->res_hashchain);
+	ls->ls_new_rsb_count--;
+	spin_unlock(&ls->ls_new_rsb_spin);
 
 	r->res_ls = ls;
 	r->res_length = len;
 	memcpy(r->res_name, name, len);
 	mutex_init(&r->res_mutex);
 
+	INIT_LIST_HEAD(&r->res_hashchain);
 	INIT_LIST_HEAD(&r->res_lookup);
 	INIT_LIST_HEAD(&r->res_grantqueue);
 	INIT_LIST_HEAD(&r->res_convertqueue);
@@ -347,7 +396,8 @@ static struct dlm_rsb *create_rsb(struct dlm_ls *ls, char *name, int len)
 	INIT_LIST_HEAD(&r->res_root_list);
 	INIT_LIST_HEAD(&r->res_recover_list);
 
-	return r;
+	*r_ret = r;
+	return 0;
 }
 
 static int search_rsb_list(struct list_head *head, char *name, int len,
@@ -405,16 +455,6 @@ static int _search_rsb(struct dlm_ls *ls, char *name, int len, int b,
 	return error;
 }
 
-static int search_rsb(struct dlm_ls *ls, char *name, int len, int b,
-		      unsigned int flags, struct dlm_rsb **r_ret)
-{
-	int error;
-	spin_lock(&ls->ls_rsbtbl[b].lock);
-	error = _search_rsb(ls, name, len, b, flags, r_ret);
-	spin_unlock(&ls->ls_rsbtbl[b].lock);
-	return error;
-}
-
 /*
  * Find rsb in rsbtbl and potentially create/add one
  *
@@ -432,35 +472,48 @@ static int search_rsb(struct dlm_ls *ls, char *name, int len, int b,
 static int find_rsb(struct dlm_ls *ls, char *name, int namelen,
 		    unsigned int flags, struct dlm_rsb **r_ret)
 {
-	struct dlm_rsb *r = NULL, *tmp;
+	struct dlm_rsb *r = NULL;
 	uint32_t hash, bucket;
-	int error = -EINVAL;
+	int error;
 
-	if (namelen > DLM_RESNAME_MAXLEN)
+	if (namelen > DLM_RESNAME_MAXLEN) {
+		error = -EINVAL;
 		goto out;
+	}
 
 	if (dlm_no_directory(ls))
 		flags |= R_CREATE;
 
-	error = 0;
 	hash = jhash(name, namelen, 0);
 	bucket = hash & (ls->ls_rsbtbl_size - 1);
 
-	error = search_rsb(ls, name, namelen, bucket, flags, &r);
+ retry:
+	if (flags & R_CREATE) {
+		error = pre_rsb_struct(ls);
+		if (error < 0)
+			goto out;
+	}
+
+	spin_lock(&ls->ls_rsbtbl[bucket].lock);
+
+	error = _search_rsb(ls, name, namelen, bucket, flags, &r);
 	if (!error)
-		goto out;
+		goto out_unlock;
 
 	if (error == -EBADR && !(flags & R_CREATE))
-		goto out;
+		goto out_unlock;
 
 	/* the rsb was found but wasn't a master copy */
 	if (error == -ENOTBLK)
-		goto out;
+		goto out_unlock;
 
-	error = -ENOMEM;
-	r = create_rsb(ls, name, namelen);
-	if (!r)
-		goto out;
+	error = get_rsb_struct(ls, name, namelen, &r);
+	if (error == -EAGAIN) {
+		spin_unlock(&ls->ls_rsbtbl[bucket].lock);
+		goto retry;
+	}
+	if (error)
+		goto out_unlock;
 
 	r->res_hash = hash;
 	r->res_bucket = bucket;
@@ -474,18 +527,10 @@ static int find_rsb(struct dlm_ls *ls, char *name, int namelen,
 			nodeid = 0;
 		r->res_nodeid = nodeid;
 	}
-
-	spin_lock(&ls->ls_rsbtbl[bucket].lock);
-	error = _search_rsb(ls, name, namelen, bucket, 0, &tmp);
-	if (!error) {
-		spin_unlock(&ls->ls_rsbtbl[bucket].lock);
-		dlm_free_rsb(r);
-		r = tmp;
-		goto out;
-	}
 	list_add(&r->res_hashchain, &ls->ls_rsbtbl[bucket].list);
-	spin_unlock(&ls->ls_rsbtbl[bucket].lock);
 	error = 0;
+ out_unlock:
+	spin_unlock(&ls->ls_rsbtbl[bucket].lock);
  out:
 	*r_ret = r;
 	return error;
@@ -580,9 +625,8 @@ static void detach_lkb(struct dlm_lkb *lkb)
 
 static int create_lkb(struct dlm_ls *ls, struct dlm_lkb **lkb_ret)
 {
-	struct dlm_lkb *lkb, *tmp;
-	uint32_t lkid = 0;
-	uint16_t bucket;
+	struct dlm_lkb *lkb;
+	int rv, id;
 
 	lkb = dlm_allocate_lkb(ls);
 	if (!lkb)
@@ -594,60 +638,42 @@ static int create_lkb(struct dlm_ls *ls, struct dlm_lkb **lkb_ret)
 	INIT_LIST_HEAD(&lkb->lkb_ownqueue);
 	INIT_LIST_HEAD(&lkb->lkb_rsb_lookup);
 	INIT_LIST_HEAD(&lkb->lkb_time_list);
-	INIT_LIST_HEAD(&lkb->lkb_astqueue);
+	INIT_LIST_HEAD(&lkb->lkb_cb_list);
+	mutex_init(&lkb->lkb_cb_mutex);
+	INIT_WORK(&lkb->lkb_cb_work, dlm_callback_work);
 
-	get_random_bytes(&bucket, sizeof(bucket));
-	bucket &= (ls->ls_lkbtbl_size - 1);
-
-	write_lock(&ls->ls_lkbtbl[bucket].lock);
+ retry:
+	rv = idr_pre_get(&ls->ls_lkbidr, GFP_NOFS);
+	if (!rv)
+		return -ENOMEM;
 
-	/* counter can roll over so we must verify lkid is not in use */
+	spin_lock(&ls->ls_lkbidr_spin);
+	rv = idr_get_new_above(&ls->ls_lkbidr, lkb, 1, &id);
+	if (!rv)
+		lkb->lkb_id = id;
+	spin_unlock(&ls->ls_lkbidr_spin);
 
-	while (lkid == 0) {
-		lkid = (bucket << 16) | ls->ls_lkbtbl[bucket].counter++;
+	if (rv == -EAGAIN)
+		goto retry;
 
-		list_for_each_entry(tmp, &ls->ls_lkbtbl[bucket].list,
-				    lkb_idtbl_list) {
-			if (tmp->lkb_id != lkid)
-				continue;
-			lkid = 0;
-			break;
-		}
+	if (rv < 0) {
+		log_error(ls, "create_lkb idr error %d", rv);
+		return rv;
 	}
 
-	lkb->lkb_id = lkid;
-	list_add(&lkb->lkb_idtbl_list, &ls->ls_lkbtbl[bucket].list);
-	write_unlock(&ls->ls_lkbtbl[bucket].lock);
-
 	*lkb_ret = lkb;
 	return 0;
 }
 
-static struct dlm_lkb *__find_lkb(struct dlm_ls *ls, uint32_t lkid)
-{
-	struct dlm_lkb *lkb;
-	uint16_t bucket = (lkid >> 16);
-
-	list_for_each_entry(lkb, &ls->ls_lkbtbl[bucket].list, lkb_idtbl_list) {
-		if (lkb->lkb_id == lkid)
-			return lkb;
-	}
-	return NULL;
-}
-
 static int find_lkb(struct dlm_ls *ls, uint32_t lkid, struct dlm_lkb **lkb_ret)
 {
 	struct dlm_lkb *lkb;
-	uint16_t bucket = (lkid >> 16);
-
-	if (bucket >= ls->ls_lkbtbl_size)
-		return -EBADSLT;
 
-	read_lock(&ls->ls_lkbtbl[bucket].lock);
-	lkb = __find_lkb(ls, lkid);
+	spin_lock(&ls->ls_lkbidr_spin);
+	lkb = idr_find(&ls->ls_lkbidr, lkid);
 	if (lkb)
 		kref_get(&lkb->lkb_ref);
-	read_unlock(&ls->ls_lkbtbl[bucket].lock);
+	spin_unlock(&ls->ls_lkbidr_spin);
 
 	*lkb_ret = lkb;
 	return lkb ? 0 : -ENOENT;
@@ -668,12 +694,12 @@ static void kill_lkb(struct kref *kref)
 
 static int __put_lkb(struct dlm_ls *ls, struct dlm_lkb *lkb)
 {
-	uint16_t bucket = (lkb->lkb_id >> 16);
+	uint32_t lkid = lkb->lkb_id;
 
-	write_lock(&ls->ls_lkbtbl[bucket].lock);
+	spin_lock(&ls->ls_lkbidr_spin);
 	if (kref_put(&lkb->lkb_ref, kill_lkb)) {
-		list_del(&lkb->lkb_idtbl_list);
-		write_unlock(&ls->ls_lkbtbl[bucket].lock);
+		idr_remove(&ls->ls_lkbidr, lkid);
+		spin_unlock(&ls->ls_lkbidr_spin);
 
 		detach_lkb(lkb);
 
@@ -683,7 +709,7 @@ static int __put_lkb(struct dlm_ls *ls, struct dlm_lkb *lkb)
 		dlm_free_lkb(lkb);
 		return 1;
 	} else {
-		write_unlock(&ls->ls_lkbtbl[bucket].lock);
+		spin_unlock(&ls->ls_lkbidr_spin);
 		return 0;
 	}
 }
@@ -849,9 +875,7 @@ void dlm_scan_waiters(struct dlm_ls *ls)
 
 		if (!num_nodes) {
 			num_nodes = ls->ls_num_nodes;
-			warned = kmalloc(GFP_KERNEL, num_nodes * sizeof(int));
-			if (warned)
-				memset(warned, 0, num_nodes * sizeof(int));
+			warned = kzalloc(num_nodes * sizeof(int), GFP_KERNEL);
 		}
 		if (!warned)
 			continue;
@@ -863,9 +887,7 @@ void dlm_scan_waiters(struct dlm_ls *ls)
 			  dlm_config.ci_waitwarn_us, lkb->lkb_wait_nodeid);
 	}
 	mutex_unlock(&ls->ls_waiters_mutex);
-
-	if (warned)
-		kfree(warned);
+	kfree(warned);
 
 	if (debug_expired)
 		log_debug(ls, "scan_waiters %u warn %u over %d us max %lld us",
@@ -2401,9 +2423,6 @@ static int do_convert(struct dlm_rsb *r, struct dlm_lkb *lkb)
 
 	if (deadlk) {
 		/* it's left on the granted queue */
-		log_debug(r->res_ls, "deadlock %x node %d sts%d g%d r%d %s",
-			  lkb->lkb_id, lkb->lkb_nodeid, lkb->lkb_status,
-			  lkb->lkb_grmode, lkb->lkb_rqmode, r->res_name);
 		revert_lock(r, lkb);
 		queue_cast(r, lkb, -EDEADLK);
 		error = -EDEADLK;
@@ -3993,8 +4012,6 @@ static void _receive_message(struct dlm_ls *ls, struct dlm_message *ms)
 	default:
 		log_error(ls, "unknown message type %d", ms->m_type);
 	}
-
-	dlm_astd_wake();
 }
 
 /* If the lockspace is in recovery mode (locking stopped), then normal
@@ -4133,7 +4150,7 @@ void dlm_recover_waiters_pre(struct dlm_ls *ls)
 	struct dlm_message *ms_stub;
 	int wait_type, stub_unlock_result, stub_cancel_result;
 
-	ms_stub = kmalloc(GFP_KERNEL, sizeof(struct dlm_message));
+	ms_stub = kmalloc(sizeof(struct dlm_message), GFP_KERNEL);
 	if (!ms_stub) {
 		log_error(ls, "dlm_recover_waiters_pre no mem");
 		return;
@@ -4809,7 +4826,7 @@ int dlm_user_unlock(struct dlm_ls *ls, struct dlm_user_args *ua_tmp,
 		goto out_put;
 
 	spin_lock(&ua->proc->locks_spin);
-	/* dlm_user_add_ast() may have already taken lkb off the proc list */
+	/* dlm_user_add_cb() may have already taken lkb off the proc list */
 	if (!list_empty(&lkb->lkb_ownqueue))
 		list_move(&lkb->lkb_ownqueue, &ua->proc->unlocking);
 	spin_unlock(&ua->proc->locks_spin);
@@ -4946,7 +4963,7 @@ static int unlock_proc_lock(struct dlm_ls *ls, struct dlm_lkb *lkb)
 
 /* We have to release clear_proc_locks mutex before calling unlock_proc_lock()
    (which does lock_rsb) due to deadlock with receiving a message that does
-   lock_rsb followed by dlm_user_add_ast() */
+   lock_rsb followed by dlm_user_add_cb() */
 
 static struct dlm_lkb *del_proc_lock(struct dlm_ls *ls,
 				     struct dlm_user_proc *proc)
@@ -4969,7 +4986,7 @@ static struct dlm_lkb *del_proc_lock(struct dlm_ls *ls,
 	return lkb;
 }
 
-/* The ls_clear_proc_locks mutex protects against dlm_user_add_asts() which
+/* The ls_clear_proc_locks mutex protects against dlm_user_add_cb() which
    1) references lkb->ua which we free here and 2) adds lkbs to proc->asts,
    which we clear here. */
 
@@ -5011,10 +5028,10 @@ void dlm_clear_proc_locks(struct dlm_ls *ls, struct dlm_user_proc *proc)
 		dlm_put_lkb(lkb);
 	}
 
-	list_for_each_entry_safe(lkb, safe, &proc->asts, lkb_astqueue) {
+	list_for_each_entry_safe(lkb, safe, &proc->asts, lkb_cb_list) {
 		memset(&lkb->lkb_callbacks, 0,
 		       sizeof(struct dlm_callback) * DLM_CALLBACKS_SIZE);
-		list_del_init(&lkb->lkb_astqueue);
+		list_del_init(&lkb->lkb_cb_list);
 		dlm_put_lkb(lkb);
 	}
 
@@ -5053,10 +5070,10 @@ static void purge_proc_locks(struct dlm_ls *ls, struct dlm_user_proc *proc)
 	spin_unlock(&proc->locks_spin);
 
 	spin_lock(&proc->asts_spin);
-	list_for_each_entry_safe(lkb, safe, &proc->asts, lkb_astqueue) {
+	list_for_each_entry_safe(lkb, safe, &proc->asts, lkb_cb_list) {
 		memset(&lkb->lkb_callbacks, 0,
 		       sizeof(struct dlm_callback) * DLM_CALLBACKS_SIZE);
-		list_del_init(&lkb->lkb_astqueue);
+		list_del_init(&lkb->lkb_cb_list);
 		dlm_put_lkb(lkb);
 	}
 	spin_unlock(&proc->asts_spin);
diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c
index 14cbf409975..a1d8f1af144 100644
--- a/fs/dlm/lockspace.c
+++ b/fs/dlm/lockspace.c
@@ -15,7 +15,6 @@
 #include "lockspace.h"
 #include "member.h"
 #include "recoverd.h"
-#include "ast.h"
 #include "dir.h"
 #include "lowcomms.h"
 #include "config.h"
@@ -24,6 +23,7 @@
 #include "recover.h"
 #include "requestqueue.h"
 #include "user.h"
+#include "ast.h"
 
 static int			ls_count;
 static struct mutex		ls_lock;
@@ -359,17 +359,10 @@ static int threads_start(void)
 {
 	int error;
 
-	/* Thread which process lock requests for all lockspace's */
-	error = dlm_astd_start();
-	if (error) {
-		log_print("cannot start dlm_astd thread %d", error);
-		goto fail;
-	}
-
 	error = dlm_scand_start();
 	if (error) {
 		log_print("cannot start dlm_scand thread %d", error);
-		goto astd_fail;
+		goto fail;
 	}
 
 	/* Thread for sending/receiving messages for all lockspace's */
@@ -383,8 +376,6 @@ static int threads_start(void)
 
  scand_fail:
 	dlm_scand_stop();
- astd_fail:
-	dlm_astd_stop();
  fail:
 	return error;
 }
@@ -393,7 +384,6 @@ static void threads_stop(void)
 {
 	dlm_scand_stop();
 	dlm_lowcomms_stop();
-	dlm_astd_stop();
 }
 
 static int new_lockspace(const char *name, int namelen, void **lockspace,
@@ -463,7 +453,7 @@ static int new_lockspace(const char *name, int namelen, void **lockspace,
 	size = dlm_config.ci_rsbtbl_size;
 	ls->ls_rsbtbl_size = size;
 
-	ls->ls_rsbtbl = kmalloc(sizeof(struct dlm_rsbtable) * size, GFP_NOFS);
+	ls->ls_rsbtbl = vmalloc(sizeof(struct dlm_rsbtable) * size);
 	if (!ls->ls_rsbtbl)
 		goto out_lsfree;
 	for (i = 0; i < size; i++) {
@@ -472,22 +462,13 @@ static int new_lockspace(const char *name, int namelen, void **lockspace,
 		spin_lock_init(&ls->ls_rsbtbl[i].lock);
 	}
 
-	size = dlm_config.ci_lkbtbl_size;
-	ls->ls_lkbtbl_size = size;
-
-	ls->ls_lkbtbl = kmalloc(sizeof(struct dlm_lkbtable) * size, GFP_NOFS);
-	if (!ls->ls_lkbtbl)
-		goto out_rsbfree;
-	for (i = 0; i < size; i++) {
-		INIT_LIST_HEAD(&ls->ls_lkbtbl[i].list);
-		rwlock_init(&ls->ls_lkbtbl[i].lock);
-		ls->ls_lkbtbl[i].counter = 1;
-	}
+	idr_init(&ls->ls_lkbidr);
+	spin_lock_init(&ls->ls_lkbidr_spin);
 
 	size = dlm_config.ci_dirtbl_size;
 	ls->ls_dirtbl_size = size;
 
-	ls->ls_dirtbl = kmalloc(sizeof(struct dlm_dirtable) * size, GFP_NOFS);
+	ls->ls_dirtbl = vmalloc(sizeof(struct dlm_dirtable) * size);
 	if (!ls->ls_dirtbl)
 		goto out_lkbfree;
 	for (i = 0; i < size; i++) {
@@ -502,6 +483,9 @@ static int new_lockspace(const char *name, int namelen, void **lockspace,
 	INIT_LIST_HEAD(&ls->ls_timeout);
 	mutex_init(&ls->ls_timeout_mutex);
 
+	INIT_LIST_HEAD(&ls->ls_new_rsb);
+	spin_lock_init(&ls->ls_new_rsb_spin);
+
 	INIT_LIST_HEAD(&ls->ls_nodes);
 	INIT_LIST_HEAD(&ls->ls_nodes_gone);
 	ls->ls_num_nodes = 0;
@@ -520,6 +504,9 @@ static int new_lockspace(const char *name, int namelen, void **lockspace,
 	init_completion(&ls->ls_members_done);
 	ls->ls_members_result = -1;
 
+	mutex_init(&ls->ls_cb_mutex);
+	INIT_LIST_HEAD(&ls->ls_cb_delay);
+
 	ls->ls_recoverd_task = NULL;
 	mutex_init(&ls->ls_recoverd_active);
 	spin_lock_init(&ls->ls_recover_lock);
@@ -553,18 +540,26 @@ static int new_lockspace(const char *name, int namelen, void **lockspace,
 	list_add(&ls->ls_list, &lslist);
 	spin_unlock(&lslist_lock);
 
+	if (flags & DLM_LSFL_FS) {
+		error = dlm_callback_start(ls);
+		if (error) {
+			log_error(ls, "can't start dlm_callback %d", error);
+			goto out_delist;
+		}
+	}
+
 	/* needs to find ls in lslist */
 	error = dlm_recoverd_start(ls);
 	if (error) {
 		log_error(ls, "can't start dlm_recoverd %d", error);
-		goto out_delist;
+		goto out_callback;
 	}
 
 	ls->ls_kobj.kset = dlm_kset;
 	error = kobject_init_and_add(&ls->ls_kobj, &dlm_ktype, NULL,
 				     "%s", ls->ls_name);
 	if (error)
-		goto out_stop;
+		goto out_recoverd;
 	kobject_uevent(&ls->ls_kobj, KOBJ_ADD);
 
 	/* let kobject handle freeing of ls if there's an error */
@@ -578,7 +573,7 @@ static int new_lockspace(const char *name, int namelen, void **lockspace,
 
 	error = do_uevent(ls, 1);
 	if (error)
-		goto out_stop;
+		goto out_recoverd;
 
 	wait_for_completion(&ls->ls_members_done);
 	error = ls->ls_members_result;
@@ -595,19 +590,20 @@ static int new_lockspace(const char *name, int namelen, void **lockspace,
 	do_uevent(ls, 0);
 	dlm_clear_members(ls);
 	kfree(ls->ls_node_array);
- out_stop:
+ out_recoverd:
 	dlm_recoverd_stop(ls);
+ out_callback:
+	dlm_callback_stop(ls);
  out_delist:
 	spin_lock(&lslist_lock);
 	list_del(&ls->ls_list);
 	spin_unlock(&lslist_lock);
 	kfree(ls->ls_recover_buf);
  out_dirfree:
-	kfree(ls->ls_dirtbl);
+	vfree(ls->ls_dirtbl);
  out_lkbfree:
-	kfree(ls->ls_lkbtbl);
- out_rsbfree:
-	kfree(ls->ls_rsbtbl);
+	idr_destroy(&ls->ls_lkbidr);
+	vfree(ls->ls_rsbtbl);
  out_lsfree:
 	if (do_unreg)
 		kobject_put(&ls->ls_kobj);
@@ -641,50 +637,64 @@ int dlm_new_lockspace(const char *name, int namelen, void **lockspace,
 	return error;
 }
 
-/* Return 1 if the lockspace still has active remote locks,
- *        2 if the lockspace still has active local locks.
- */
-static int lockspace_busy(struct dlm_ls *ls)
-{
-	int i, lkb_found = 0;
-	struct dlm_lkb *lkb;
-
-	/* NOTE: We check the lockidtbl here rather than the resource table.
-	   This is because there may be LKBs queued as ASTs that have been
-	   unlinked from their RSBs and are pending deletion once the AST has
-	   been delivered */
-
-	for (i = 0; i < ls->ls_lkbtbl_size; i++) {
-		read_lock(&ls->ls_lkbtbl[i].lock);
-		if (!list_empty(&ls->ls_lkbtbl[i].list)) {
-			lkb_found = 1;
-			list_for_each_entry(lkb, &ls->ls_lkbtbl[i].list,
-					    lkb_idtbl_list) {
-				if (!lkb->lkb_nodeid) {
-					read_unlock(&ls->ls_lkbtbl[i].lock);
-					return 2;
-				}
-			}
-		}
-		read_unlock(&ls->ls_lkbtbl[i].lock);
+static int lkb_idr_is_local(int id, void *p, void *data)
+{
+	struct dlm_lkb *lkb = p;
+
+	if (!lkb->lkb_nodeid)
+		return 1;
+	return 0;
+}
+
+static int lkb_idr_is_any(int id, void *p, void *data)
+{
+	return 1;
+}
+
+static int lkb_idr_free(int id, void *p, void *data)
+{
+	struct dlm_lkb *lkb = p;
+
+	if (lkb->lkb_lvbptr && lkb->lkb_flags & DLM_IFL_MSTCPY)
+		dlm_free_lvb(lkb->lkb_lvbptr);
+
+	dlm_free_lkb(lkb);
+	return 0;
+}
+
+/* NOTE: We check the lkbidr here rather than the resource table.
+   This is because there may be LKBs queued as ASTs that have been unlinked
+   from their RSBs and are pending deletion once the AST has been delivered */
+
+static int lockspace_busy(struct dlm_ls *ls, int force)
+{
+	int rv;
+
+	spin_lock(&ls->ls_lkbidr_spin);
+	if (force == 0) {
+		rv = idr_for_each(&ls->ls_lkbidr, lkb_idr_is_any, ls);
+	} else if (force == 1) {
+		rv = idr_for_each(&ls->ls_lkbidr, lkb_idr_is_local, ls);
+	} else {
+		rv = 0;
 	}
-	return lkb_found;
+	spin_unlock(&ls->ls_lkbidr_spin);
+	return rv;
 }
 
 static int release_lockspace(struct dlm_ls *ls, int force)
 {
-	struct dlm_lkb *lkb;
 	struct dlm_rsb *rsb;
 	struct list_head *head;
 	int i, busy, rv;
 
-	busy = lockspace_busy(ls);
+	busy = lockspace_busy(ls, force);
 
 	spin_lock(&lslist_lock);
 	if (ls->ls_create_count == 1) {
-		if (busy > force)
+		if (busy) {
 			rv = -EBUSY;
-		else {
+		} else {
 			/* remove_lockspace takes ls off lslist */
 			ls->ls_create_count = 0;
 			rv = 0;
@@ -708,12 +718,12 @@ static int release_lockspace(struct dlm_ls *ls, int force)
 
 	dlm_recoverd_stop(ls);
 
+	dlm_callback_stop(ls);
+
 	remove_lockspace(ls);
 
 	dlm_delete_debug_file(ls);
 
-	dlm_astd_suspend();
-
 	kfree(ls->ls_recover_buf);
 
 	/*
@@ -721,31 +731,15 @@ static int release_lockspace(struct dlm_ls *ls, int force)
 	 */
 
 	dlm_dir_clear(ls);
-	kfree(ls->ls_dirtbl);
+	vfree(ls->ls_dirtbl);
 
 	/*
-	 * Free all lkb's on lkbtbl[] lists.
+	 * Free all lkb's in idr
 	 */
 
-	for (i = 0; i < ls->ls_lkbtbl_size; i++) {
-		head = &ls->ls_lkbtbl[i].list;
-		while (!list_empty(head)) {
-			lkb = list_entry(head->next, struct dlm_lkb,
-					 lkb_idtbl_list);
-
-			list_del(&lkb->lkb_idtbl_list);
-
-			dlm_del_ast(lkb);
-
-			if (lkb->lkb_lvbptr && lkb->lkb_flags & DLM_IFL_MSTCPY)
-				dlm_free_lvb(lkb->lkb_lvbptr);
-
-			dlm_free_lkb(lkb);
-		}
-	}
-	dlm_astd_resume();
-
-	kfree(ls->ls_lkbtbl);
+	idr_for_each(&ls->ls_lkbidr, lkb_idr_free, ls);
+	idr_remove_all(&ls->ls_lkbidr);
+	idr_destroy(&ls->ls_lkbidr);
 
 	/*
 	 * Free all rsb's on rsbtbl[] lists
@@ -770,7 +764,14 @@ static int release_lockspace(struct dlm_ls *ls, int force)
 		}
 	}
 
-	kfree(ls->ls_rsbtbl);
+	vfree(ls->ls_rsbtbl);
+
+	while (!list_empty(&ls->ls_new_rsb)) {
+		rsb = list_first_entry(&ls->ls_new_rsb, struct dlm_rsb,
+				       res_hashchain);
+		list_del(&rsb->res_hashchain);
+		dlm_free_rsb(rsb);
+	}
 
 	/*
 	 * Free structures on any other lists
diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index 5e2c71f05e4..990626e7da8 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -512,12 +512,10 @@ static void process_sctp_notification(struct connection *con,
 			}
 			make_sockaddr(&prim.ssp_addr, 0, &addr_len);
 			if (dlm_addr_to_nodeid(&prim.ssp_addr, &nodeid)) {
-				int i;
 				unsigned char *b=(unsigned char *)&prim.ssp_addr;
 				log_print("reject connect from unknown addr");
-				for (i=0; i<sizeof(struct sockaddr_storage);i++)
-					printk("%02x ", b[i]);
-				printk("\n");
+				print_hex_dump_bytes("ss: ", DUMP_PREFIX_NONE, 
+						     b, sizeof(struct sockaddr_storage));
 				sctp_send_shutdown(prim.ssp_assoc_id);
 				return;
 			}
@@ -748,7 +746,10 @@ static int tcp_accept_from_sock(struct connection *con)
 	/* Get the new node's NODEID */
 	make_sockaddr(&peeraddr, 0, &len);
 	if (dlm_addr_to_nodeid(&peeraddr, &nodeid)) {
+		unsigned char *b=(unsigned char *)&peeraddr;
 		log_print("connect from non cluster node");
+		print_hex_dump_bytes("ss: ", DUMP_PREFIX_NONE, 
+				     b, sizeof(struct sockaddr_storage));
 		sock_release(newsock);
 		mutex_unlock(&con->sock_mutex);
 		return -1;
diff --git a/fs/dlm/memory.c b/fs/dlm/memory.c
index 8e0d00db004..da64df7576e 100644
--- a/fs/dlm/memory.c
+++ b/fs/dlm/memory.c
@@ -16,6 +16,7 @@
 #include "memory.h"
 
 static struct kmem_cache *lkb_cache;
+static struct kmem_cache *rsb_cache;
 
 
 int __init dlm_memory_init(void)
@@ -26,6 +27,14 @@ int __init dlm_memory_init(void)
 				__alignof__(struct dlm_lkb), 0, NULL);
 	if (!lkb_cache)
 		ret = -ENOMEM;
+
+	rsb_cache = kmem_cache_create("dlm_rsb", sizeof(struct dlm_rsb),
+				__alignof__(struct dlm_rsb), 0, NULL);
+	if (!rsb_cache) {
+		kmem_cache_destroy(lkb_cache);
+		ret = -ENOMEM;
+	}
+
 	return ret;
 }
 
@@ -33,6 +42,8 @@ void dlm_memory_exit(void)
 {
 	if (lkb_cache)
 		kmem_cache_destroy(lkb_cache);
+	if (rsb_cache)
+		kmem_cache_destroy(rsb_cache);
 }
 
 char *dlm_allocate_lvb(struct dlm_ls *ls)
@@ -48,16 +59,11 @@ void dlm_free_lvb(char *p)
 	kfree(p);
 }
 
-/* FIXME: have some minimal space built-in to rsb for the name and
-   kmalloc a separate name if needed, like dentries are done */
-
-struct dlm_rsb *dlm_allocate_rsb(struct dlm_ls *ls, int namelen)
+struct dlm_rsb *dlm_allocate_rsb(struct dlm_ls *ls)
 {
 	struct dlm_rsb *r;
 
-	DLM_ASSERT(namelen <= DLM_RESNAME_MAXLEN,);
-
-	r = kzalloc(sizeof(*r) + namelen, GFP_NOFS);
+	r = kmem_cache_zalloc(rsb_cache, GFP_NOFS);
 	return r;
 }
 
@@ -65,7 +71,7 @@ void dlm_free_rsb(struct dlm_rsb *r)
 {
 	if (r->res_lvbptr)
 		dlm_free_lvb(r->res_lvbptr);
-	kfree(r);
+	kmem_cache_free(rsb_cache, r);
 }
 
 struct dlm_lkb *dlm_allocate_lkb(struct dlm_ls *ls)
diff --git a/fs/dlm/memory.h b/fs/dlm/memory.h
index 485fb29143b..177c11cbb0a 100644
--- a/fs/dlm/memory.h
+++ b/fs/dlm/memory.h
@@ -16,7 +16,7 @@
 
 int dlm_memory_init(void);
 void dlm_memory_exit(void);
-struct dlm_rsb *dlm_allocate_rsb(struct dlm_ls *ls, int namelen);
+struct dlm_rsb *dlm_allocate_rsb(struct dlm_ls *ls);
 void dlm_free_rsb(struct dlm_rsb *r);
 struct dlm_lkb *dlm_allocate_lkb(struct dlm_ls *ls);
 void dlm_free_lkb(struct dlm_lkb *l);
diff --git a/fs/dlm/recoverd.c b/fs/dlm/recoverd.c
index fd677c8c3d3..774da3cf92c 100644
--- a/fs/dlm/recoverd.c
+++ b/fs/dlm/recoverd.c
@@ -58,13 +58,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 
 	mutex_lock(&ls->ls_recoverd_active);
 
-	/*
-	 * Suspending and resuming dlm_astd ensures that no lkb's from this ls
-	 * will be processed by dlm_astd during recovery.
-	 */
-
-	dlm_astd_suspend();
-	dlm_astd_resume();
+	dlm_callback_suspend(ls);
 
 	/*
 	 * Free non-master tossed rsb's.  Master rsb's are kept on toss
@@ -202,6 +196,8 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 
 	dlm_adjust_timeouts(ls);
 
+	dlm_callback_resume(ls);
+
 	error = enable_locking(ls, rv->seq);
 	if (error) {
 		log_debug(ls, "enable_locking failed %d", error);
@@ -222,8 +218,6 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 
 	dlm_grant_after_purge(ls);
 
-	dlm_astd_wake();
-
 	log_debug(ls, "recover %llx done: %u ms",
 		  (unsigned long long)rv->seq,
 		  jiffies_to_msecs(jiffies - start));
diff --git a/fs/dlm/user.c b/fs/dlm/user.c
index e96bf3e9be8..d8ea6075640 100644
--- a/fs/dlm/user.c
+++ b/fs/dlm/user.c
@@ -213,9 +213,9 @@ void dlm_user_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode,
 		goto out;
 	}
 
-	if (list_empty(&lkb->lkb_astqueue)) {
+	if (list_empty(&lkb->lkb_cb_list)) {
 		kref_get(&lkb->lkb_ref);
-		list_add_tail(&lkb->lkb_astqueue, &proc->asts);
+		list_add_tail(&lkb->lkb_cb_list, &proc->asts);
 		wake_up_interruptible(&proc->wait);
 	}
 	spin_unlock(&proc->asts_spin);
@@ -832,24 +832,24 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count,
 	}
 
 	/* if we empty lkb_callbacks, we don't want to unlock the spinlock
-	   without removing lkb_astqueue; so empty lkb_astqueue is always
+	   without removing lkb_cb_list; so empty lkb_cb_list is always
 	   consistent with empty lkb_callbacks */
 
-	lkb = list_entry(proc->asts.next, struct dlm_lkb, lkb_astqueue);
+	lkb = list_entry(proc->asts.next, struct dlm_lkb, lkb_cb_list);
 
 	rv = dlm_rem_lkb_callback(lkb->lkb_resource->res_ls, lkb, &cb, &resid);
 	if (rv < 0) {
 		/* this shouldn't happen; lkb should have been removed from
 		   list when resid was zero */
 		log_print("dlm_rem_lkb_callback empty %x", lkb->lkb_id);
-		list_del_init(&lkb->lkb_astqueue);
+		list_del_init(&lkb->lkb_cb_list);
 		spin_unlock(&proc->asts_spin);
 		/* removes ref for proc->asts, may cause lkb to be freed */
 		dlm_put_lkb(lkb);
 		goto try_another;
 	}
 	if (!resid)
-		list_del_init(&lkb->lkb_astqueue);
+		list_del_init(&lkb->lkb_cb_list);
 	spin_unlock(&proc->asts_spin);
 
 	if (cb.flags & DLM_CB_SKIP) {
diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c
index 4ec9eb00a24..c6ac98cf9ba 100644
--- a/fs/ecryptfs/file.c
+++ b/fs/ecryptfs/file.c
@@ -270,14 +270,15 @@ static int ecryptfs_release(struct inode *inode, struct file *file)
 }
 
 static int
-ecryptfs_fsync(struct file *file, int datasync)
+ecryptfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	int rc = 0;
 
-	rc = generic_file_fsync(file, datasync);
+	rc = generic_file_fsync(file, start, end, datasync);
 	if (rc)
 		goto out;
-	rc = vfs_fsync(ecryptfs_file_to_lower(file), datasync);
+	rc = vfs_fsync_range(ecryptfs_file_to_lower(file), start, end,
+			     datasync);
 out:
 	return rc;
 }
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 7349ade17de..340c657a108 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -147,7 +147,6 @@ static int ecryptfs_interpose(struct dentry *lower_dentry,
  * @lower_dir_inode: inode of the parent in the lower fs of the new file
  * @dentry: New file's dentry
  * @mode: The mode of the new file
- * @nd: nameidata of ecryptfs' parent's dentry & vfsmount
  *
  * Creates the file in the lower file system.
  *
@@ -155,31 +154,10 @@ static int ecryptfs_interpose(struct dentry *lower_dentry,
  */
 static int
 ecryptfs_create_underlying_file(struct inode *lower_dir_inode,
-				struct dentry *dentry, int mode,
-				struct nameidata *nd)
+				struct dentry *dentry, int mode)
 {
 	struct dentry *lower_dentry = ecryptfs_dentry_to_lower(dentry);
-	struct vfsmount *lower_mnt = ecryptfs_dentry_to_lower_mnt(dentry);
-	struct dentry *dentry_save;
-	struct vfsmount *vfsmount_save;
-	unsigned int flags_save;
-	int rc;
-
-	if (nd) {
-		dentry_save = nd->path.dentry;
-		vfsmount_save = nd->path.mnt;
-		flags_save = nd->flags;
-		nd->path.dentry = lower_dentry;
-		nd->path.mnt = lower_mnt;
-		nd->flags &= ~LOOKUP_OPEN;
-	}
-	rc = vfs_create(lower_dir_inode, lower_dentry, mode, nd);
-	if (nd) {
-		nd->path.dentry = dentry_save;
-		nd->path.mnt = vfsmount_save;
-		nd->flags = flags_save;
-	}
-	return rc;
+	return vfs_create(lower_dir_inode, lower_dentry, mode, NULL);
 }
 
 /**
@@ -197,8 +175,7 @@ ecryptfs_create_underlying_file(struct inode *lower_dir_inode,
  */
 static int
 ecryptfs_do_create(struct inode *directory_inode,
-		   struct dentry *ecryptfs_dentry, int mode,
-		   struct nameidata *nd)
+		   struct dentry *ecryptfs_dentry, int mode)
 {
 	int rc;
 	struct dentry *lower_dentry;
@@ -213,7 +190,7 @@ ecryptfs_do_create(struct inode *directory_inode,
 		goto out;
 	}
 	rc = ecryptfs_create_underlying_file(lower_dir_dentry->d_inode,
-					     ecryptfs_dentry, mode, nd);
+					     ecryptfs_dentry, mode);
 	if (rc) {
 		printk(KERN_ERR "%s: Failure to create dentry in lower fs; "
 		       "rc = [%d]\n", __func__, rc);
@@ -294,7 +271,7 @@ ecryptfs_create(struct inode *directory_inode, struct dentry *ecryptfs_dentry,
 	int rc;
 
 	/* ecryptfs_do_create() calls ecryptfs_interpose() */
-	rc = ecryptfs_do_create(directory_inode, ecryptfs_dentry, mode, nd);
+	rc = ecryptfs_do_create(directory_inode, ecryptfs_dentry, mode);
 	if (unlikely(rc)) {
 		ecryptfs_printk(KERN_WARNING, "Failed to create file in"
 				"lower filesystem\n");
@@ -942,10 +919,8 @@ int ecryptfs_truncate(struct dentry *dentry, loff_t new_length)
 }
 
 static int
-ecryptfs_permission(struct inode *inode, int mask, unsigned int flags)
+ecryptfs_permission(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
 	return inode_permission(ecryptfs_inode_to_lower(inode), mask);
 }
 
diff --git a/fs/efs/namei.c b/fs/efs/namei.c
index 1511bf9e5f8..832b10ded82 100644
--- a/fs/efs/namei.c
+++ b/fs/efs/namei.c
@@ -60,14 +60,11 @@ static efs_ino_t efs_find_entry(struct inode *inode, const char *name, int len)
 
 struct dentry *efs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) {
 	efs_ino_t inodenum;
-	struct inode * inode = NULL;
+	struct inode *inode = NULL;
 
 	inodenum = efs_find_entry(dir, dentry->d_name.name, dentry->d_name.len);
-	if (inodenum) {
+	if (inodenum)
 		inode = efs_iget(dir->i_sb, inodenum);
-		if (IS_ERR(inode))
-			return ERR_CAST(inode);
-	}
 
 	return d_splice_alias(inode, dentry);
 }
diff --git a/fs/exec.c b/fs/exec.c
index ea5f748906a..842d5700c15 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -963,9 +963,18 @@ static int de_thread(struct task_struct *tsk)
 		leader->group_leader = tsk;
 
 		tsk->exit_signal = SIGCHLD;
+		leader->exit_signal = -1;
 
 		BUG_ON(leader->exit_state != EXIT_ZOMBIE);
 		leader->exit_state = EXIT_DEAD;
+
+		/*
+		 * We are going to release_task()->ptrace_unlink() silently,
+		 * the tracer can sleep in do_wait(). EXIT_DEAD guarantees
+		 * the tracer wont't block again waiting for this thread.
+		 */
+		if (unlikely(leader->ptrace))
+			__wake_up_parent(leader, leader->parent);
 		write_unlock_irq(&tasklist_lock);
 
 		release_task(leader);
@@ -1093,6 +1102,7 @@ int flush_old_exec(struct linux_binprm * bprm)
 
 	bprm->mm = NULL;		/* We're using it now */
 
+	set_fs(USER_DS);
 	current->flags &= ~(PF_RANDOMIZE | PF_KTHREAD);
 	flush_thread();
 	current->personality &= ~bprm->per_clear;
@@ -1104,6 +1114,13 @@ out:
 }
 EXPORT_SYMBOL(flush_old_exec);
 
+void would_dump(struct linux_binprm *bprm, struct file *file)
+{
+	if (inode_permission(file->f_path.dentry->d_inode, MAY_READ) < 0)
+		bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+}
+EXPORT_SYMBOL(would_dump);
+
 void setup_new_exec(struct linux_binprm * bprm)
 {
 	int i, ch;
@@ -1143,9 +1160,10 @@ void setup_new_exec(struct linux_binprm * bprm)
 	if (bprm->cred->uid != current_euid() ||
 	    bprm->cred->gid != current_egid()) {
 		current->pdeath_signal = 0;
-	} else if (file_permission(bprm->file, MAY_READ) ||
-		   bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP) {
-		set_dumpable(current->mm, suid_dumpable);
+	} else {
+		would_dump(bprm, bprm->file);
+		if (bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP)
+			set_dumpable(current->mm, suid_dumpable);
 	}
 
 	/*
@@ -1224,7 +1242,12 @@ int check_unsafe_exec(struct linux_binprm *bprm)
 	unsigned n_fs;
 	int res = 0;
 
-	bprm->unsafe = tracehook_unsafe_exec(p);
+	if (p->ptrace) {
+		if (p->ptrace & PT_PTRACE_CAP)
+			bprm->unsafe |= LSM_UNSAFE_PTRACE_CAP;
+		else
+			bprm->unsafe |= LSM_UNSAFE_PTRACE;
+	}
 
 	n_fs = 1;
 	spin_lock(&p->fs->lock);
@@ -1352,19 +1375,21 @@ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs)
 	unsigned int depth = bprm->recursion_depth;
 	int try,retval;
 	struct linux_binfmt *fmt;
+	pid_t old_pid;
 
 	retval = security_bprm_check(bprm);
 	if (retval)
 		return retval;
 
-	/* kernel module loader fixup */
-	/* so we don't try to load run modprobe in kernel space. */
-	set_fs(USER_DS);
-
 	retval = audit_bprm(bprm);
 	if (retval)
 		return retval;
 
+	/* Need to fetch pid before load_binary changes it */
+	rcu_read_lock();
+	old_pid = task_pid_nr_ns(current, task_active_pid_ns(current->parent));
+	rcu_read_unlock();
+
 	retval = -ENOENT;
 	for (try=0; try<2; try++) {
 		read_lock(&binfmt_lock);
@@ -1384,7 +1409,8 @@ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs)
 			bprm->recursion_depth = depth;
 			if (retval >= 0) {
 				if (depth == 0)
-					tracehook_report_exec(fmt, bprm, regs);
+					ptrace_event(PTRACE_EVENT_EXEC,
+							old_pid);
 				put_binfmt(fmt);
 				allow_write_access(bprm->file);
 				if (bprm->file)
@@ -1772,7 +1798,7 @@ static int zap_process(struct task_struct *start, int exit_code)
 
 	t = start;
 	do {
-		task_clear_group_stop_pending(t);
+		task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
 		if (t != current && t->mm) {
 			sigaddset(&t->pending.signal, SIGKILL);
 			signal_wake_up(t, 1);
@@ -1999,7 +2025,7 @@ static void wait_for_dump_helpers(struct file *file)
  * is a special value that we use to trap recursive
  * core dumps
  */
-static int umh_pipe_setup(struct subprocess_info *info)
+static int umh_pipe_setup(struct subprocess_info *info, struct cred *new)
 {
 	struct file *rp, *wp;
 	struct fdtable *fdt;
diff --git a/fs/exofs/file.c b/fs/exofs/file.c
index 45ca323d836..491c6c078e7 100644
--- a/fs/exofs/file.c
+++ b/fs/exofs/file.c
@@ -42,11 +42,19 @@ static int exofs_release_file(struct inode *inode, struct file *filp)
  *   Note, in exofs all metadata is written as part of inode, regardless.
  *   The writeout is synchronous
  */
-static int exofs_file_fsync(struct file *filp, int datasync)
+static int exofs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			    int datasync)
 {
+	struct inode *inode = filp->f_mapping->host;
 	int ret;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+
+	mutex_lock(&inode->i_mutex);
 	ret = sync_inode_metadata(filp->f_mapping->host, 1);
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/exofs/namei.c b/fs/exofs/namei.c
index 4d70db110cf..b54c43775f1 100644
--- a/fs/exofs/namei.c
+++ b/fs/exofs/namei.c
@@ -55,12 +55,7 @@ static struct dentry *exofs_lookup(struct inode *dir, struct dentry *dentry,
 		return ERR_PTR(-ENAMETOOLONG);
 
 	ino = exofs_inode_by_name(dir, dentry);
-	inode = NULL;
-	if (ino) {
-		inode = exofs_iget(dir->i_sb, ino);
-		if (IS_ERR(inode))
-			return ERR_CAST(inode);
-	}
+	inode = ino ? exofs_iget(dir->i_sb, ino) : NULL;
 	return d_splice_alias(inode, dentry);
 }
 
diff --git a/fs/exofs/super.c b/fs/exofs/super.c
index 06065bd37fc..c57beddcc21 100644
--- a/fs/exofs/super.c
+++ b/fs/exofs/super.c
@@ -913,7 +913,7 @@ struct dentry *exofs_get_parent(struct dentry *child)
 	unsigned long ino = exofs_parent_ino(child);
 
 	if (!ino)
-		return NULL;
+		return ERR_PTR(-ESTALE);
 
 	return d_obtain_alias(exofs_iget(child->d_inode->i_sb, ino));
 }
diff --git a/fs/ext2/acl.c b/fs/ext2/acl.c
index abea5a17c76..bfe651f9ae1 100644
--- a/fs/ext2/acl.c
+++ b/fs/ext2/acl.c
@@ -232,11 +232,11 @@ ext2_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 }
 
 int
-ext2_check_acl(struct inode *inode, int mask, unsigned int flags)
+ext2_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/ext2/acl.h b/fs/ext2/acl.h
index c939b7b1209..3ff6cbb9ac4 100644
--- a/fs/ext2/acl.h
+++ b/fs/ext2/acl.h
@@ -54,7 +54,7 @@ static inline int ext2_acl_count(size_t size)
 #ifdef CONFIG_EXT2_FS_POSIX_ACL
 
 /* acl.c */
-extern int ext2_check_acl (struct inode *, int, unsigned int);
+extern int ext2_check_acl (struct inode *, int);
 extern int ext2_acl_chmod (struct inode *);
 extern int ext2_init_acl (struct inode *, struct inode *);
 
diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h
index 645be9e7ee4..af9fc89b1b2 100644
--- a/fs/ext2/ext2.h
+++ b/fs/ext2/ext2.h
@@ -150,7 +150,8 @@ extern void ext2_write_super (struct super_block *);
 extern const struct file_operations ext2_dir_operations;
 
 /* file.c */
-extern int ext2_fsync(struct file *file, int datasync);
+extern int ext2_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync);
 extern const struct inode_operations ext2_file_inode_operations;
 extern const struct file_operations ext2_file_operations;
 extern const struct file_operations ext2_xip_file_operations;
diff --git a/fs/ext2/file.c b/fs/ext2/file.c
index 49eec9456c5..82e06321de3 100644
--- a/fs/ext2/file.c
+++ b/fs/ext2/file.c
@@ -40,13 +40,13 @@ static int ext2_release_file (struct inode * inode, struct file * filp)
 	return 0;
 }
 
-int ext2_fsync(struct file *file, int datasync)
+int ext2_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	int ret;
 	struct super_block *sb = file->f_mapping->host->i_sb;
 	struct address_space *mapping = sb->s_bdev->bd_inode->i_mapping;
 
-	ret = generic_file_fsync(file, datasync);
+	ret = generic_file_fsync(file, start, end, datasync);
 	if (ret == -EIO || test_and_clear_bit(AS_EIO, &mapping->flags)) {
 		/* We don't really know where the IO error happened... */
 		ext2_error(sb, __func__,
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index 788e09a07f7..a8a58f63f07 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -843,8 +843,8 @@ ext2_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
 	struct inode *inode = mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev,
-				iov, offset, nr_segs, ext2_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 ext2_get_block);
 	if (ret < 0 && (rw & WRITE))
 		ext2_write_failed(mapping, offset + iov_length(iov, nr_segs));
 	return ret;
@@ -1184,6 +1184,8 @@ static int ext2_setsize(struct inode *inode, loff_t newsize)
 	if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
 		return -EPERM;
 
+	inode_dio_wait(inode);
+
 	if (mapping_is_xip(inode->i_mapping))
 		error = xip_truncate_page(inode->i_mapping, newsize);
 	else if (test_opt(inode->i_sb, NOBH))
diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c
index ed5c5d496ee..d60b7099e2d 100644
--- a/fs/ext2/namei.c
+++ b/fs/ext2/namei.c
@@ -67,15 +67,11 @@ static struct dentry *ext2_lookup(struct inode * dir, struct dentry *dentry, str
 	inode = NULL;
 	if (ino) {
 		inode = ext2_iget(dir->i_sb, ino);
-		if (IS_ERR(inode)) {
-			if (PTR_ERR(inode) == -ESTALE) {
-				ext2_error(dir->i_sb, __func__,
-						"deleted inode referenced: %lu",
-						(unsigned long) ino);
-				return ERR_PTR(-EIO);
-			} else {
-				return ERR_CAST(inode);
-			}
+		if (inode == ERR_PTR(-ESTALE)) {
+			ext2_error(dir->i_sb, __func__,
+					"deleted inode referenced: %lu",
+					(unsigned long) ino);
+			return ERR_PTR(-EIO);
 		}
 	}
 	return d_splice_alias(inode, dentry);
diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c
index 9d021c0d472..edfeb293d4c 100644
--- a/fs/ext3/acl.c
+++ b/fs/ext3/acl.c
@@ -240,11 +240,11 @@ ext3_set_acl(handle_t *handle, struct inode *inode, int type,
 }
 
 int
-ext3_check_acl(struct inode *inode, int mask, unsigned int flags)
+ext3_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/ext3/acl.h b/fs/ext3/acl.h
index 5faf8048e90..597334626de 100644
--- a/fs/ext3/acl.h
+++ b/fs/ext3/acl.h
@@ -54,7 +54,7 @@ static inline int ext3_acl_count(size_t size)
 #ifdef CONFIG_EXT3_FS_POSIX_ACL
 
 /* acl.c */
-extern int ext3_check_acl (struct inode *, int, unsigned int);
+extern int ext3_check_acl (struct inode *, int);
 extern int ext3_acl_chmod (struct inode *);
 extern int ext3_init_acl (handle_t *, struct inode *, struct inode *);
 
diff --git a/fs/ext3/fsync.c b/fs/ext3/fsync.c
index 09b13bb34c9..0bcf63adb80 100644
--- a/fs/ext3/fsync.c
+++ b/fs/ext3/fsync.c
@@ -43,7 +43,7 @@
  * inode to disk.
  */
 
-int ext3_sync_file(struct file *file, int datasync)
+int ext3_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ext3_inode_info *ei = EXT3_I(inode);
@@ -54,6 +54,17 @@ int ext3_sync_file(struct file *file, int datasync)
 	if (inode->i_sb->s_flags & MS_RDONLY)
 		return 0;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+
+	/*
+	 * Taking the mutex here just to keep consistent with how fsync was
+	 * called previously, however it looks like we don't need to take
+	 * i_mutex at all.
+	 */
+	mutex_lock(&inode->i_mutex);
+
 	J_ASSERT(ext3_journal_current_handle() == NULL);
 
 	/*
@@ -70,8 +81,10 @@ int ext3_sync_file(struct file *file, int datasync)
 	 *  (they were dirtied by commit).  But that's OK - the blocks are
 	 *  safe in-journal, which is all fsync() needs to ensure.
 	 */
-	if (ext3_should_journal_data(inode))
+	if (ext3_should_journal_data(inode)) {
+		mutex_unlock(&inode->i_mutex);
 		return ext3_force_commit(inode->i_sb);
+	}
 
 	if (datasync)
 		commit_tid = atomic_read(&ei->i_datasync_tid);
@@ -91,5 +104,6 @@ int ext3_sync_file(struct file *file, int datasync)
 	 */
 	if (needs_barrier)
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 3451d23c3ba..2978a2a17a5 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -1816,9 +1816,8 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb,
 	}
 
 retry:
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				 offset, nr_segs,
-				 ext3_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 ext3_get_block);
 	/*
 	 * In case of error extending write may have instantiated a few
 	 * blocks outside i_size. Trim these off again.
@@ -3216,6 +3215,9 @@ int ext3_setattr(struct dentry *dentry, struct iattr *attr)
 		ext3_journal_stop(handle);
 	}
 
+	if (attr->ia_valid & ATTR_SIZE)
+		inode_dio_wait(inode);
+
 	if (S_ISREG(inode->i_mode) &&
 	    attr->ia_valid & ATTR_SIZE && attr->ia_size < inode->i_size) {
 		handle_t *handle;
diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c
index 34b6d9bfc48..c095cf5640c 100644
--- a/fs/ext3/namei.c
+++ b/fs/ext3/namei.c
@@ -1038,15 +1038,11 @@ static struct dentry *ext3_lookup(struct inode * dir, struct dentry *dentry, str
 			return ERR_PTR(-EIO);
 		}
 		inode = ext3_iget(dir->i_sb, ino);
-		if (IS_ERR(inode)) {
-			if (PTR_ERR(inode) == -ESTALE) {
-				ext3_error(dir->i_sb, __func__,
-						"deleted inode referenced: %lu",
-						ino);
-				return ERR_PTR(-EIO);
-			} else {
-				return ERR_CAST(inode);
-			}
+		if (inode == ERR_PTR(-ESTALE)) {
+			ext3_error(dir->i_sb, __func__,
+					"deleted inode referenced: %lu",
+					ino);
+			return ERR_PTR(-EIO);
 		}
 	}
 	return d_splice_alias(inode, dentry);
diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index aad153ef6b7..b57ea2f9126 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -1718,6 +1718,8 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent)
 	sbi->s_resuid = le16_to_cpu(es->s_def_resuid);
 	sbi->s_resgid = le16_to_cpu(es->s_def_resgid);
 
+	/* enable barriers by default */
+	set_opt(sbi->s_mount_opt, BARRIER);
 	set_opt(sbi->s_mount_opt, RESERVATION);
 
 	if (!parse_options ((char *) data, sb, &journal_inum, &journal_devnum,
diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c
index 21eacd7b7d7..60d900fcc3d 100644
--- a/fs/ext4/acl.c
+++ b/fs/ext4/acl.c
@@ -238,11 +238,11 @@ ext4_set_acl(handle_t *handle, struct inode *inode, int type,
 }
 
 int
-ext4_check_acl(struct inode *inode, int mask, unsigned int flags)
+ext4_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/ext4/acl.h b/fs/ext4/acl.h
index dec821168fd..9d843d5deac 100644
--- a/fs/ext4/acl.h
+++ b/fs/ext4/acl.h
@@ -54,7 +54,7 @@ static inline int ext4_acl_count(size_t size)
 #ifdef CONFIG_EXT4_FS_POSIX_ACL
 
 /* acl.c */
-extern int ext4_check_acl(struct inode *, int, unsigned int);
+extern int ext4_check_acl(struct inode *, int);
 extern int ext4_acl_chmod(struct inode *);
 extern int ext4_init_acl(handle_t *, struct inode *, struct inode *);
 
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 1921392cd70..fa44df87971 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1758,7 +1758,7 @@ extern int ext4_htree_store_dirent(struct file *dir_file, __u32 hash,
 extern void ext4_htree_free_dir_info(struct dir_private_info *p);
 
 /* fsync.c */
-extern int ext4_sync_file(struct file *, int);
+extern int ext4_sync_file(struct file *, loff_t, loff_t, int);
 extern int ext4_flush_completed_IO(struct inode *);
 
 /* hash.c */
diff --git a/fs/ext4/ext4_extents.h b/fs/ext4/ext4_extents.h
index 2e29abb30f7..095c36f3b61 100644
--- a/fs/ext4/ext4_extents.h
+++ b/fs/ext4/ext4_extents.h
@@ -125,7 +125,7 @@ struct ext4_ext_path {
  * positive retcode - signal for ext4_ext_walk_space(), see below
  * callback must return valid extent (passed or newly created)
  */
-typedef int (*ext_prepare_callback)(struct inode *, struct ext4_ext_path *,
+typedef int (*ext_prepare_callback)(struct inode *, ext4_lblk_t,
 					struct ext4_ext_cache *,
 					struct ext4_extent *, void *);
 
@@ -133,8 +133,11 @@ typedef int (*ext_prepare_callback)(struct inode *, struct ext4_ext_path *,
 #define EXT_BREAK      1
 #define EXT_REPEAT     2
 
-/* Maximum logical block in a file; ext4_extent's ee_block is __le32 */
-#define EXT_MAX_BLOCK	0xffffffff
+/*
+ * Maximum number of logical blocks in a file; ext4_extent's ee_block is
+ * __le32.
+ */
+#define EXT_MAX_BLOCKS	0xffffffff
 
 /*
  * EXT_INIT_MAX_LEN is the maximum number of blocks we can have in an
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 5199bac7fc6..f815cc81e7a 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -1408,7 +1408,7 @@ got_index:
 
 /*
  * ext4_ext_next_allocated_block:
- * returns allocated block in subsequent extent or EXT_MAX_BLOCK.
+ * returns allocated block in subsequent extent or EXT_MAX_BLOCKS.
  * NOTE: it considers block number from index entry as
  * allocated block. Thus, index entries have to be consistent
  * with leaves.
@@ -1422,7 +1422,7 @@ ext4_ext_next_allocated_block(struct ext4_ext_path *path)
 	depth = path->p_depth;
 
 	if (depth == 0 && path->p_ext == NULL)
-		return EXT_MAX_BLOCK;
+		return EXT_MAX_BLOCKS;
 
 	while (depth >= 0) {
 		if (depth == path->p_depth) {
@@ -1439,12 +1439,12 @@ ext4_ext_next_allocated_block(struct ext4_ext_path *path)
 		depth--;
 	}
 
-	return EXT_MAX_BLOCK;
+	return EXT_MAX_BLOCKS;
 }
 
 /*
  * ext4_ext_next_leaf_block:
- * returns first allocated block from next leaf or EXT_MAX_BLOCK
+ * returns first allocated block from next leaf or EXT_MAX_BLOCKS
  */
 static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode,
 					struct ext4_ext_path *path)
@@ -1456,7 +1456,7 @@ static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode,
 
 	/* zero-tree has no leaf blocks at all */
 	if (depth == 0)
-		return EXT_MAX_BLOCK;
+		return EXT_MAX_BLOCKS;
 
 	/* go to index block */
 	depth--;
@@ -1469,7 +1469,7 @@ static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode,
 		depth--;
 	}
 
-	return EXT_MAX_BLOCK;
+	return EXT_MAX_BLOCKS;
 }
 
 /*
@@ -1677,13 +1677,13 @@ static unsigned int ext4_ext_check_overlap(struct inode *inode,
 	 */
 	if (b2 < b1) {
 		b2 = ext4_ext_next_allocated_block(path);
-		if (b2 == EXT_MAX_BLOCK)
+		if (b2 == EXT_MAX_BLOCKS)
 			goto out;
 	}
 
 	/* check for wrap through zero on extent logical start block*/
 	if (b1 + len1 < b1) {
-		len1 = EXT_MAX_BLOCK - b1;
+		len1 = EXT_MAX_BLOCKS - b1;
 		newext->ee_len = cpu_to_le16(len1);
 		ret = 1;
 	}
@@ -1767,7 +1767,7 @@ repeat:
 	fex = EXT_LAST_EXTENT(eh);
 	next = ext4_ext_next_leaf_block(inode, path);
 	if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)
-	    && next != EXT_MAX_BLOCK) {
+	    && next != EXT_MAX_BLOCKS) {
 		ext_debug("next leaf block - %d\n", next);
 		BUG_ON(npath != NULL);
 		npath = ext4_ext_find_extent(inode, next, NULL);
@@ -1887,7 +1887,7 @@ static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
 	BUG_ON(func == NULL);
 	BUG_ON(inode == NULL);
 
-	while (block < last && block != EXT_MAX_BLOCK) {
+	while (block < last && block != EXT_MAX_BLOCKS) {
 		num = last - block;
 		/* find extent for this block */
 		down_read(&EXT4_I(inode)->i_data_sem);
@@ -1958,7 +1958,7 @@ static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
 			err = -EIO;
 			break;
 		}
-		err = func(inode, path, &cbex, ex, cbdata);
+		err = func(inode, next, &cbex, ex, cbdata);
 		ext4_ext_drop_refs(path);
 
 		if (err < 0)
@@ -2020,7 +2020,7 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
 	if (ex == NULL) {
 		/* there is no extent yet, so gap is [0;-] */
 		lblock = 0;
-		len = EXT_MAX_BLOCK;
+		len = EXT_MAX_BLOCKS;
 		ext_debug("cache gap(whole file):");
 	} else if (block < le32_to_cpu(ex->ee_block)) {
 		lblock = block;
@@ -2350,7 +2350,7 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 			 * never happen because at least one of the end points
 			 * needs to be on the edge of the extent.
 			 */
-			if (end == EXT_MAX_BLOCK) {
+			if (end == EXT_MAX_BLOCKS - 1) {
 				ext_debug("  bad truncate %u:%u\n",
 						start, end);
 				block = 0;
@@ -2398,7 +2398,7 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 			 * If this is a truncate, this condition
 			 * should never happen
 			 */
-			if (end == EXT_MAX_BLOCK) {
+			if (end == EXT_MAX_BLOCKS - 1) {
 				ext_debug("  bad truncate %u:%u\n",
 					start, end);
 				err = -EIO;
@@ -2478,7 +2478,7 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 		 * we need to remove it from the leaf
 		 */
 		if (num == 0) {
-			if (end != EXT_MAX_BLOCK) {
+			if (end != EXT_MAX_BLOCKS - 1) {
 				/*
 				 * For hole punching, we need to scoot all the
 				 * extents up when an extent is removed so that
@@ -3699,7 +3699,7 @@ void ext4_ext_truncate(struct inode *inode)
 
 	last_block = (inode->i_size + sb->s_blocksize - 1)
 			>> EXT4_BLOCK_SIZE_BITS(sb);
-	err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCK);
+	err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1);
 
 	/* In a multi-transaction truncate, we only make the final
 	 * transaction synchronous.
@@ -3914,14 +3914,13 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
 /*
  * Callback function called for each extent to gather FIEMAP information.
  */
-static int ext4_ext_fiemap_cb(struct inode *inode, struct ext4_ext_path *path,
+static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next,
 		       struct ext4_ext_cache *newex, struct ext4_extent *ex,
 		       void *data)
 {
 	__u64	logical;
 	__u64	physical;
 	__u64	length;
-	loff_t	size;
 	__u32	flags = 0;
 	int		ret = 0;
 	struct fiemap_extent_info *fieinfo = data;
@@ -4103,8 +4102,7 @@ found_delayed_extent:
 	if (ex && ext4_ext_is_uninitialized(ex))
 		flags |= FIEMAP_EXTENT_UNWRITTEN;
 
-	size = i_size_read(inode);
-	if (logical + length >= size)
+	if (next == EXT_MAX_BLOCKS)
 		flags |= FIEMAP_EXTENT_LAST;
 
 	ret = fiemap_fill_next_extent(fieinfo, logical, physical,
@@ -4347,8 +4345,8 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 
 		start_blk = start >> 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;
+		if (last_blk >= EXT_MAX_BLOCKS)
+			last_blk = EXT_MAX_BLOCKS-1;
 		len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
 
 		/*
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 2c097232200..ce766f974b1 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -236,6 +236,27 @@ loff_t ext4_llseek(struct file *file, loff_t offset, int origin)
 		}
 		offset += file->f_pos;
 		break;
+	case SEEK_DATA:
+		/*
+		 * In the generic case the entire file is data, so as long as
+		 * offset isn't at the end of the file then the offset is data.
+		 */
+		if (offset >= inode->i_size) {
+			mutex_unlock(&inode->i_mutex);
+			return -ENXIO;
+		}
+		break;
+	case SEEK_HOLE:
+		/*
+		 * There is a virtual hole at the end of the file, so as long as
+		 * offset isn't i_size or larger, return i_size.
+		 */
+		if (offset >= inode->i_size) {
+			mutex_unlock(&inode->i_mutex);
+			return -ENXIO;
+		}
+		offset = inode->i_size;
+		break;
 	}
 
 	if (offset < 0 || offset > maxbytes) {
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index ce66d2fe826..da3bed3e0c2 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -151,6 +151,32 @@ static int ext4_sync_parent(struct inode *inode)
 	return ret;
 }
 
+/**
+ * __sync_file - generic_file_fsync without the locking and filemap_write
+ * @inode:	inode to sync
+ * @datasync:	only sync essential metadata if true
+ *
+ * This is just generic_file_fsync without the locking.  This is needed for
+ * nojournal mode to make sure this inodes data/metadata makes it to disk
+ * properly.  The i_mutex should be held already.
+ */
+static int __sync_inode(struct inode *inode, int datasync)
+{
+	int err;
+	int ret;
+
+	ret = sync_mapping_buffers(inode->i_mapping);
+	if (!(inode->i_state & I_DIRTY))
+		return ret;
+	if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
+		return ret;
+
+	err = sync_inode_metadata(inode, 1);
+	if (ret == 0)
+		ret = err;
+	return ret;
+}
+
 /*
  * akpm: A new design for ext4_sync_file().
  *
@@ -165,7 +191,7 @@ static int ext4_sync_parent(struct inode *inode)
  * i_mutex lock is held when entering and exiting this function
  */
 
-int ext4_sync_file(struct file *file, int datasync)
+int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ext4_inode_info *ei = EXT4_I(inode);
@@ -178,15 +204,20 @@ int ext4_sync_file(struct file *file, int datasync)
 
 	trace_ext4_sync_file_enter(file, datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	if (inode->i_sb->s_flags & MS_RDONLY)
-		return 0;
+		goto out;
 
 	ret = ext4_flush_completed_IO(inode);
 	if (ret < 0)
 		goto out;
 
 	if (!journal) {
-		ret = generic_file_fsync(file, datasync);
+		ret = __sync_inode(inode, datasync);
 		if (!ret && !list_empty(&inode->i_dentry))
 			ret = ext4_sync_parent(inode);
 		goto out;
@@ -220,6 +251,7 @@ int ext4_sync_file(struct file *file, int datasync)
 	if (needs_barrier)
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
  out:
+	mutex_unlock(&inode->i_mutex);
 	trace_ext4_sync_file_exit(inode, ret);
 	return ret;
 }
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index a5763e3505b..678cde834f1 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2634,7 +2634,7 @@ static int ext4_writepage(struct page *page,
 	struct buffer_head *page_bufs = NULL;
 	struct inode *inode = page->mapping->host;
 
-	trace_ext4_writepage(inode, page);
+	trace_ext4_writepage(page);
 	size = i_size_read(inode);
 	if (page->index == size >> PAGE_CACHE_SHIFT)
 		len = size & ~PAGE_CACHE_MASK;
@@ -3501,10 +3501,8 @@ retry:
 				 offset, nr_segs,
 				 ext4_get_block, NULL, NULL, 0);
 	else {
-		ret = blockdev_direct_IO(rw, iocb, inode,
-				 inode->i_sb->s_bdev, iov,
-				 offset, nr_segs,
-				 ext4_get_block, NULL);
+		ret = blockdev_direct_IO(rw, iocb, inode, iov,
+				 offset, nr_segs, ext4_get_block);
 
 		if (unlikely((rw & WRITE) && ret < 0)) {
 			loff_t isize = i_size_read(inode);
@@ -3575,6 +3573,7 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 			    ssize_t size, void *private, int ret,
 			    bool is_async)
 {
+	struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode;
         ext4_io_end_t *io_end = iocb->private;
 	struct workqueue_struct *wq;
 	unsigned long flags;
@@ -3596,6 +3595,7 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 out:
 		if (is_async)
 			aio_complete(iocb, ret, 0);
+		inode_dio_done(inode);
 		return;
 	}
 
@@ -3616,6 +3616,9 @@ out:
 	/* queue the work to convert unwritten extents to written */
 	queue_work(wq, &io_end->work);
 	iocb->private = NULL;
+
+	/* XXX: probably should move into the real I/O completion handler */
+	inode_dio_done(inode);
 }
 
 static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate)
@@ -3748,11 +3751,13 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 			EXT4_I(inode)->cur_aio_dio = iocb->private;
 		}
 
-		ret = blockdev_direct_IO(rw, iocb, inode,
+		ret = __blockdev_direct_IO(rw, iocb, inode,
 					 inode->i_sb->s_bdev, iov,
 					 offset, nr_segs,
 					 ext4_get_block_write,
-					 ext4_end_io_dio);
+					 ext4_end_io_dio,
+					 NULL,
+					 DIO_LOCKING | DIO_SKIP_HOLES);
 		if (iocb->private)
 			EXT4_I(inode)->cur_aio_dio = NULL;
 		/*
@@ -5351,6 +5356,8 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
+		inode_dio_wait(inode);
+
 		if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
 			struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 
@@ -5843,80 +5850,84 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	struct page *page = vmf->page;
 	loff_t size;
 	unsigned long len;
-	int ret = -EINVAL;
-	void *fsdata;
+	int ret;
 	struct file *file = vma->vm_file;
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct address_space *mapping = inode->i_mapping;
+	handle_t *handle;
+	get_block_t *get_block;
+	int retries = 0;
 
 	/*
-	 * Get i_alloc_sem to stop truncates messing with the inode. We cannot
-	 * get i_mutex because we are already holding mmap_sem.
+	 * This check is racy but catches the common case. We rely on
+	 * __block_page_mkwrite() to do a reliable check.
 	 */
-	down_read(&inode->i_alloc_sem);
-	size = i_size_read(inode);
-	if (page->mapping != mapping || size <= page_offset(page)
-	    || !PageUptodate(page)) {
-		/* page got truncated from under us? */
-		goto out_unlock;
+	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
+	/* Delalloc case is easy... */
+	if (test_opt(inode->i_sb, DELALLOC) &&
+	    !ext4_should_journal_data(inode) &&
+	    !ext4_nonda_switch(inode->i_sb)) {
+		do {
+			ret = __block_page_mkwrite(vma, vmf,
+						   ext4_da_get_block_prep);
+		} while (ret == -ENOSPC &&
+		       ext4_should_retry_alloc(inode->i_sb, &retries));
+		goto out_ret;
 	}
-	ret = 0;
 
 	lock_page(page);
-	wait_on_page_writeback(page);
-	if (PageMappedToDisk(page)) {
-		up_read(&inode->i_alloc_sem);
-		return VM_FAULT_LOCKED;
+	size = i_size_read(inode);
+	/* Page got truncated from under us? */
+	if (page->mapping != mapping || page_offset(page) > size) {
+		unlock_page(page);
+		ret = VM_FAULT_NOPAGE;
+		goto out;
 	}
 
 	if (page->index == size >> PAGE_CACHE_SHIFT)
 		len = size & ~PAGE_CACHE_MASK;
 	else
 		len = PAGE_CACHE_SIZE;
-
 	/*
-	 * return if we have all the buffers mapped. This avoid
-	 * the need to call write_begin/write_end which does a
-	 * journal_start/journal_stop which can block and take
-	 * long time
+	 * Return if we have all the buffers mapped. This avoids the need to do
+	 * journal_start/journal_stop which can block and take a long time
 	 */
 	if (page_has_buffers(page)) {
 		if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL,
 					ext4_bh_unmapped)) {
-			up_read(&inode->i_alloc_sem);
-			return VM_FAULT_LOCKED;
+			/* Wait so that we don't change page under IO */
+			wait_on_page_writeback(page);
+			ret = VM_FAULT_LOCKED;
+			goto out;
 		}
 	}
 	unlock_page(page);
-	/*
-	 * OK, we need to fill the hole... Do write_begin write_end
-	 * to do block allocation/reservation.We are not holding
-	 * inode.i__mutex here. That allow * parallel write_begin,
-	 * write_end call. lock_page prevent this from happening
-	 * on the same page though
-	 */
-	ret = mapping->a_ops->write_begin(file, mapping, page_offset(page),
-			len, AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata);
-	if (ret < 0)
-		goto out_unlock;
-	ret = mapping->a_ops->write_end(file, mapping, page_offset(page),
-			len, len, page, fsdata);
-	if (ret < 0)
-		goto out_unlock;
-	ret = 0;
-
-	/*
-	 * write_begin/end might have created a dirty page and someone
-	 * could wander in and start the IO.  Make sure that hasn't
-	 * happened.
-	 */
-	lock_page(page);
-	wait_on_page_writeback(page);
-	up_read(&inode->i_alloc_sem);
-	return VM_FAULT_LOCKED;
-out_unlock:
-	if (ret)
+	/* OK, we need to fill the hole... */
+	if (ext4_should_dioread_nolock(inode))
+		get_block = ext4_get_block_write;
+	else
+		get_block = ext4_get_block;
+retry_alloc:
+	handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode));
+	if (IS_ERR(handle)) {
 		ret = VM_FAULT_SIGBUS;
-	up_read(&inode->i_alloc_sem);
+		goto out;
+	}
+	ret = __block_page_mkwrite(vma, vmf, get_block);
+	if (!ret && ext4_should_journal_data(inode)) {
+		if (walk_page_buffers(handle, page_buffers(page), 0,
+			  PAGE_CACHE_SIZE, NULL, do_journal_get_write_access)) {
+			unlock_page(page);
+			ret = VM_FAULT_SIGBUS;
+			goto out;
+		}
+		ext4_set_inode_state(inode, EXT4_STATE_JDATA);
+	}
+	ext4_journal_stop(handle);
+	if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
+		goto retry_alloc;
+out_ret:
+	ret = block_page_mkwrite_return(ret);
+out:
 	return ret;
 }
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 859f2ae8864..6ed859d5685 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -3578,8 +3578,8 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
 		free += next - bit;
 
 		trace_ext4_mballoc_discard(sb, NULL, group, bit, next - bit);
-		trace_ext4_mb_release_inode_pa(sb, pa->pa_inode, pa,
-					       grp_blk_start + bit, next - bit);
+		trace_ext4_mb_release_inode_pa(pa, grp_blk_start + bit,
+					       next - bit);
 		mb_free_blocks(pa->pa_inode, e4b, bit, next - bit);
 		bit = next + 1;
 	}
@@ -3608,7 +3608,7 @@ ext4_mb_release_group_pa(struct ext4_buddy *e4b,
 	ext4_group_t group;
 	ext4_grpblk_t bit;
 
-	trace_ext4_mb_release_group_pa(sb, pa);
+	trace_ext4_mb_release_group_pa(pa);
 	BUG_ON(pa->pa_deleted == 0);
 	ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit);
 	BUG_ON(group != e4b->bd_group && pa->pa_len != 0);
@@ -4448,7 +4448,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
  * @inode:		inode
  * @block:		start physical block to free
  * @count:		number of blocks to count
- * @metadata: 		Are these metadata blocks
+ * @flags:		flags used by ext4_free_blocks
  */
 void ext4_free_blocks(handle_t *handle, struct inode *inode,
 		      struct buffer_head *bh, ext4_fsblk_t block,
diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index 2b8304bf3c5..f57455a1b1b 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -1002,12 +1002,12 @@ mext_check_arguments(struct inode *orig_inode,
 		return -EINVAL;
 	}
 
-	if ((orig_start > EXT_MAX_BLOCK) ||
-	    (donor_start > EXT_MAX_BLOCK) ||
-	    (*len > EXT_MAX_BLOCK) ||
-	    (orig_start + *len > EXT_MAX_BLOCK))  {
+	if ((orig_start >= EXT_MAX_BLOCKS) ||
+	    (donor_start >= EXT_MAX_BLOCKS) ||
+	    (*len > EXT_MAX_BLOCKS) ||
+	    (orig_start + *len >= EXT_MAX_BLOCKS))  {
 		ext4_debug("ext4 move extent: Can't handle over [%u] blocks "
-			"[ino:orig %lu, donor %lu]\n", EXT_MAX_BLOCK,
+			"[ino:orig %lu, donor %lu]\n", EXT_MAX_BLOCKS,
 			orig_inode->i_ino, donor_inode->i_ino);
 		return -EINVAL;
 	}
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index b754b7721f5..707d605bf76 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -1037,15 +1037,11 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, stru
 			return ERR_PTR(-EIO);
 		}
 		inode = ext4_iget(dir->i_sb, ino);
-		if (IS_ERR(inode)) {
-			if (PTR_ERR(inode) == -ESTALE) {
-				EXT4_ERROR_INODE(dir,
-						 "deleted inode referenced: %u",
-						 ino);
-				return ERR_PTR(-EIO);
-			} else {
-				return ERR_CAST(inode);
-			}
+		if (inode == ERR_PTR(-ESTALE)) {
+			EXT4_ERROR_INODE(dir,
+					 "deleted inode referenced: %u",
+					 ino);
+			return ERR_PTR(-EIO);
 		}
 	}
 	return d_splice_alias(inode, dentry);
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index cc5c157aa11..9ea71aa864b 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -2243,6 +2243,12 @@ static void ext4_orphan_cleanup(struct super_block *sb,
  * in the vfs.  ext4 inode has 48 bits of i_block in fsblock units,
  * so that won't be a limiting factor.
  *
+ * However there is other limiting factor. We do store extents in the form
+ * of starting block and length, hence the resulting length of the extent
+ * covering maximum file size must fit into on-disk format containers as
+ * well. Given that length is always by 1 unit bigger than max unit (because
+ * we count 0 as well) we have to lower the s_maxbytes by one fs block.
+ *
  * Note, this does *not* consider any metadata overhead for vfs i_blocks.
  */
 static loff_t ext4_max_size(int blkbits, int has_huge_files)
@@ -2264,10 +2270,13 @@ static loff_t ext4_max_size(int blkbits, int has_huge_files)
 		upper_limit <<= blkbits;
 	}
 
-	/* 32-bit extent-start container, ee_block */
-	res = 1LL << 32;
+	/*
+	 * 32-bit extent-start container, ee_block. We lower the maxbytes
+	 * by one fs block, so ee_len can cover the extent of maximum file
+	 * size
+	 */
+	res = (1LL << 32) - 1;
 	res <<= blkbits;
-	res -= 1;
 
 	/* Sanity check against vm- & vfs- imposed limits */
 	if (res > upper_limit)
diff --git a/fs/fat/fat.h b/fs/fat/fat.h
index 8276cc282de..a5d3853822e 100644
--- a/fs/fat/fat.h
+++ b/fs/fat/fat.h
@@ -109,6 +109,7 @@ struct msdos_inode_info {
 	int i_attrs;		/* unused attribute bits */
 	loff_t i_pos;		/* on-disk position of directory entry or 0 */
 	struct hlist_node i_fat_hash;	/* hash by i_location */
+	struct rw_semaphore truncate_lock; /* protect bmap against truncate */
 	struct inode vfs_inode;
 };
 
@@ -309,7 +310,8 @@ extern int fat_setattr(struct dentry * dentry, struct iattr * attr);
 extern void fat_truncate_blocks(struct inode *inode, loff_t offset);
 extern int fat_getattr(struct vfsmount *mnt, struct dentry *dentry,
 		       struct kstat *stat);
-extern int fat_file_fsync(struct file *file, int datasync);
+extern int fat_file_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync);
 
 /* fat/inode.c */
 extern void fat_attach(struct inode *inode, loff_t i_pos);
diff --git a/fs/fat/file.c b/fs/fat/file.c
index 7257752b6d5..c118acf16e4 100644
--- a/fs/fat/file.c
+++ b/fs/fat/file.c
@@ -102,7 +102,7 @@ static int fat_ioctl_set_attributes(struct file *file, u32 __user *user_attr)
 		if (attr & ATTR_SYS)
 			inode->i_flags |= S_IMMUTABLE;
 		else
-			inode->i_flags &= S_IMMUTABLE;
+			inode->i_flags &= ~S_IMMUTABLE;
 	}
 
 	fat_save_attrs(inode, attr);
@@ -149,12 +149,12 @@ static int fat_file_release(struct inode *inode, struct file *filp)
 	return 0;
 }
 
-int fat_file_fsync(struct file *filp, int datasync)
+int fat_file_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int res, err;
 
-	res = generic_file_fsync(filp, datasync);
+	res = generic_file_fsync(filp, start, end, datasync);
 	err = sync_mapping_buffers(MSDOS_SB(inode->i_sb)->fat_inode->i_mapping);
 
 	return res ? res : err;
@@ -397,6 +397,8 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr)
 	 * sequence.
 	 */
 	if (attr->ia_valid & ATTR_SIZE) {
+		inode_dio_wait(inode);
+
 		if (attr->ia_size > inode->i_size) {
 			error = fat_cont_expand(inode, attr->ia_size);
 			if (error || attr->ia_valid == ATTR_SIZE)
@@ -429,8 +431,10 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
+		down_write(&MSDOS_I(inode)->truncate_lock);
 		truncate_setsize(inode, attr->ia_size);
 		fat_truncate_blocks(inode, attr->ia_size);
+		up_write(&MSDOS_I(inode)->truncate_lock);
 	}
 
 	setattr_copy(inode, attr);
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index cb8d8391ac0..5942fec22c6 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -211,8 +211,8 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb,
 	 * FAT need to use the DIO_LOCKING for avoiding the race
 	 * condition of fat_get_block() and ->truncate().
 	 */
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev,
-				 iov, offset, nr_segs, fat_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 fat_get_block);
 	if (ret < 0 && (rw & WRITE))
 		fat_write_failed(mapping, offset + iov_length(iov, nr_segs));
 
@@ -224,9 +224,9 @@ static sector_t _fat_bmap(struct address_space *mapping, sector_t block)
 	sector_t blocknr;
 
 	/* fat_get_cluster() assumes the requested blocknr isn't truncated. */
-	down_read(&mapping->host->i_alloc_sem);
+	down_read(&MSDOS_I(mapping->host)->truncate_lock);
 	blocknr = generic_block_bmap(mapping, block, fat_get_block);
-	up_read(&mapping->host->i_alloc_sem);
+	up_read(&MSDOS_I(mapping->host)->truncate_lock);
 
 	return blocknr;
 }
@@ -510,6 +510,8 @@ static struct inode *fat_alloc_inode(struct super_block *sb)
 	ei = kmem_cache_alloc(fat_inode_cachep, GFP_NOFS);
 	if (!ei)
 		return NULL;
+
+	init_rwsem(&ei->truncate_lock);
 	return &ei->vfs_inode;
 }
 
diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c
index 3b222dafd15..66e83b84545 100644
--- a/fs/fat/namei_msdos.c
+++ b/fs/fat/namei_msdos.c
@@ -209,29 +209,20 @@ static struct dentry *msdos_lookup(struct inode *dir, struct dentry *dentry,
 	int err;
 
 	lock_super(sb);
-
 	err = msdos_find(dir, dentry->d_name.name, dentry->d_name.len, &sinfo);
-	if (err) {
-		if (err == -ENOENT) {
-			inode = NULL;
-			goto out;
-		}
-		goto error;
-	}
-
-	inode = fat_build_inode(sb, sinfo.de, sinfo.i_pos);
-	brelse(sinfo.bh);
-	if (IS_ERR(inode)) {
-		err = PTR_ERR(inode);
-		goto error;
+	switch (err) {
+	case -ENOENT:
+		inode = NULL;
+		break;
+	case 0:
+		inode = fat_build_inode(sb, sinfo.de, sinfo.i_pos);
+		brelse(sinfo.bh);
+		break;
+	default:
+		inode = ERR_PTR(err);
 	}
-out:
 	unlock_super(sb);
 	return d_splice_alias(inode, dentry);
-
-error:
-	unlock_super(sb);
-	return ERR_PTR(err);
 }
 
 /***** Creates a directory entry (name is already formatted). */
diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c
index 20b4ea53fdc..bb3f29c3557 100644
--- a/fs/fat/namei_vfat.c
+++ b/fs/fat/namei_vfat.c
@@ -82,10 +82,8 @@ static int vfat_revalidate_ci(struct dentry *dentry, struct nameidata *nd)
 	 * case sensitive name which is specified by user if this is
 	 * for creation.
 	 */
-	if (!(nd->flags & (LOOKUP_CONTINUE | LOOKUP_PARENT))) {
-		if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
-			return 0;
-	}
+	if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
+		return 0;
 
 	return vfat_revalidate_shortname(dentry);
 }
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 0f015a0468d..b8c507ca42f 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -461,32 +461,6 @@ writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
 }
 
 /*
- * For background writeback the caller does not have the sb pinned
- * before calling writeback. So make sure that we do pin it, so it doesn't
- * go away while we are writing inodes from it.
- */
-static bool pin_sb_for_writeback(struct super_block *sb)
-{
-	spin_lock(&sb_lock);
-	if (list_empty(&sb->s_instances)) {
-		spin_unlock(&sb_lock);
-		return false;
-	}
-
-	sb->s_count++;
-	spin_unlock(&sb_lock);
-
-	if (down_read_trylock(&sb->s_umount)) {
-		if (sb->s_root)
-			return true;
-		up_read(&sb->s_umount);
-	}
-
-	put_super(sb);
-	return false;
-}
-
-/*
  * Write a portion of b_io inodes which belong to @sb.
  *
  * If @only_this_sb is true, then find and write all such
@@ -585,7 +559,7 @@ void writeback_inodes_wb(struct bdi_writeback *wb,
 		struct inode *inode = wb_inode(wb->b_io.prev);
 		struct super_block *sb = inode->i_sb;
 
-		if (!pin_sb_for_writeback(sb)) {
+		if (!grab_super_passive(sb)) {
 			requeue_io(inode);
 			continue;
 		}
diff --git a/fs/fscache/page.c b/fs/fscache/page.c
index a2a5d19ece6..3f7a59bfa7a 100644
--- a/fs/fscache/page.c
+++ b/fs/fscache/page.c
@@ -954,3 +954,43 @@ void fscache_mark_pages_cached(struct fscache_retrieval *op,
 	pagevec_reinit(pagevec);
 }
 EXPORT_SYMBOL(fscache_mark_pages_cached);
+
+/*
+ * Uncache all the pages in an inode that are marked PG_fscache, assuming them
+ * to be associated with the given cookie.
+ */
+void __fscache_uncache_all_inode_pages(struct fscache_cookie *cookie,
+				       struct inode *inode)
+{
+	struct address_space *mapping = inode->i_mapping;
+	struct pagevec pvec;
+	pgoff_t next;
+	int i;
+
+	_enter("%p,%p", cookie, inode);
+
+	if (!mapping || mapping->nrpages == 0) {
+		_leave(" [no pages]");
+		return;
+	}
+
+	pagevec_init(&pvec, 0);
+	next = 0;
+	do {
+		if (!pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE))
+			break;
+		for (i = 0; i < pagevec_count(&pvec); i++) {
+			struct page *page = pvec.pages[i];
+			next = page->index;
+			if (PageFsCache(page)) {
+				__fscache_wait_on_page_write(cookie, page);
+				__fscache_uncache_page(cookie, page);
+			}
+		}
+		pagevec_release(&pvec);
+		cond_resched();
+	} while (++next);
+
+	_leave("");
+}
+EXPORT_SYMBOL(__fscache_uncache_all_inode_pages);
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index d5016071459..9f63e493a9b 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -382,7 +382,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode,
 	struct fuse_entry_out outentry;
 	struct fuse_file *ff;
 	struct file *file;
-	int flags = nd->intent.open.flags - 1;
+	int flags = nd->intent.open.flags;
 
 	if (fc->no_create)
 		return -ENOSYS;
@@ -576,7 +576,7 @@ static int fuse_mknod(struct inode *dir, struct dentry *entry, int mode,
 static int fuse_create(struct inode *dir, struct dentry *entry, int mode,
 		       struct nameidata *nd)
 {
-	if (nd && (nd->flags & LOOKUP_OPEN)) {
+	if (nd) {
 		int err = fuse_create_open(dir, entry, mode, nd);
 		if (err != -ENOSYS)
 			return err;
@@ -971,9 +971,9 @@ static int fuse_access(struct inode *inode, int mask)
 	return err;
 }
 
-static int fuse_perm_getattr(struct inode *inode, int flags)
+static int fuse_perm_getattr(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	return fuse_do_getattr(inode, NULL, NULL);
@@ -992,7 +992,7 @@ static int fuse_perm_getattr(struct inode *inode, int flags)
  * access request is sent.  Execute permission is still checked
  * locally based on file mode.
  */
-static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
+static int fuse_permission(struct inode *inode, int mask)
 {
 	struct fuse_conn *fc = get_fuse_conn(inode);
 	bool refreshed = false;
@@ -1011,23 +1011,22 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 		if (fi->i_time < get_jiffies_64()) {
 			refreshed = true;
 
-			err = fuse_perm_getattr(inode, flags);
+			err = fuse_perm_getattr(inode, mask);
 			if (err)
 				return err;
 		}
 	}
 
 	if (fc->flags & FUSE_DEFAULT_PERMISSIONS) {
-		err = generic_permission(inode, mask, flags, NULL);
+		err = generic_permission(inode, mask);
 
 		/* If permission is denied, try to refresh file
 		   attributes.  This is also needed, because the root
 		   node will at first have no permissions */
 		if (err == -EACCES && !refreshed) {
-			err = fuse_perm_getattr(inode, flags);
+			err = fuse_perm_getattr(inode, mask);
 			if (!err)
-				err = generic_permission(inode, mask,
-							flags, NULL);
+				err = generic_permission(inode, mask);
 		}
 
 		/* Note: the opposite of the above test does not
@@ -1035,7 +1034,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 		   noticed immediately, only after the attribute
 		   timeout has expired */
 	} else if (mask & (MAY_ACCESS | MAY_CHDIR)) {
-		if (flags & IPERM_FLAG_RCU)
+		if (mask & MAY_NOT_BLOCK)
 			return -ECHILD;
 
 		err = fuse_access(inode, mask);
@@ -1044,7 +1043,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 			if (refreshed)
 				return -EACCES;
 
-			err = fuse_perm_getattr(inode, flags);
+			err = fuse_perm_getattr(inode, mask);
 			if (!err && !(inode->i_mode & S_IXUGO))
 				return -EACCES;
 		}
@@ -1177,9 +1176,10 @@ static int fuse_dir_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static int fuse_dir_fsync(struct file *file, int datasync)
+static int fuse_dir_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync)
 {
-	return fuse_fsync_common(file, datasync, 1);
+	return fuse_fsync_common(file, start, end, datasync, 1);
 }
 
 static bool update_mtime(unsigned ivalid)
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index 82a66466a24..7bb685cdd00 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -400,7 +400,8 @@ static void fuse_sync_writes(struct inode *inode)
 	fuse_release_nowrite(inode);
 }
 
-int fuse_fsync_common(struct file *file, int datasync, int isdir)
+int fuse_fsync_common(struct file *file, loff_t start, loff_t end,
+		      int datasync, int isdir)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct fuse_conn *fc = get_fuse_conn(inode);
@@ -412,9 +413,15 @@ int fuse_fsync_common(struct file *file, int datasync, int isdir)
 	if (is_bad_inode(inode))
 		return -EIO;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
 	if ((!isdir && fc->no_fsync) || (isdir && fc->no_fsyncdir))
 		return 0;
 
+	mutex_lock(&inode->i_mutex);
+
 	/*
 	 * Start writeback against all dirty pages of the inode, then
 	 * wait for all outstanding writes, before sending the FSYNC
@@ -422,13 +429,15 @@ int fuse_fsync_common(struct file *file, int datasync, int isdir)
 	 */
 	err = write_inode_now(inode, 0);
 	if (err)
-		return err;
+		goto out;
 
 	fuse_sync_writes(inode);
 
 	req = fuse_get_req(fc);
-	if (IS_ERR(req))
-		return PTR_ERR(req);
+	if (IS_ERR(req)) {
+		err = PTR_ERR(req);
+		goto out;
+	}
 
 	memset(&inarg, 0, sizeof(inarg));
 	inarg.fh = ff->fh;
@@ -448,12 +457,15 @@ int fuse_fsync_common(struct file *file, int datasync, int isdir)
 			fc->no_fsync = 1;
 		err = 0;
 	}
+out:
+	mutex_unlock(&inode->i_mutex);
 	return err;
 }
 
-static int fuse_fsync(struct file *file, int datasync)
+static int fuse_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync)
 {
-	return fuse_fsync_common(file, datasync, 0);
+	return fuse_fsync_common(file, start, end, datasync, 0);
 }
 
 void fuse_read_fill(struct fuse_req *req, struct file *file, loff_t pos,
@@ -1600,15 +1612,32 @@ static loff_t fuse_file_llseek(struct file *file, loff_t offset, int origin)
 	struct inode *inode = file->f_path.dentry->d_inode;
 
 	mutex_lock(&inode->i_mutex);
-	switch (origin) {
-	case SEEK_END:
+	if (origin != SEEK_CUR || origin != SEEK_SET) {
 		retval = fuse_update_attributes(inode, NULL, file, NULL);
 		if (retval)
 			goto exit;
+	}
+
+	switch (origin) {
+	case SEEK_END:
 		offset += i_size_read(inode);
 		break;
 	case SEEK_CUR:
 		offset += file->f_pos;
+		break;
+	case SEEK_DATA:
+		if (offset >= i_size_read(inode)) {
+			retval = -ENXIO;
+			goto exit;
+		}
+		break;
+	case SEEK_HOLE:
+		if (offset >= i_size_read(inode)) {
+			retval = -ENXIO;
+			goto exit;
+		}
+		offset = i_size_read(inode);
+		break;
 	}
 	retval = -EINVAL;
 	if (offset >= 0 && offset <= inode->i_sb->s_maxbytes) {
diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
index b788becada7..c6aa2d4b851 100644
--- a/fs/fuse/fuse_i.h
+++ b/fs/fuse/fuse_i.h
@@ -589,7 +589,8 @@ void fuse_release_common(struct file *file, int opcode);
 /**
  * Send FSYNC or FSYNCDIR request
  */
-int fuse_fsync_common(struct file *file, int datasync, int isdir);
+int fuse_fsync_common(struct file *file, loff_t start, loff_t end,
+		      int datasync, int isdir);
 
 /**
  * Notify poll wakeup
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index cc6ec4b2f0f..38f84cd48b6 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -921,6 +921,8 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent)
 	if (sb->s_flags & MS_MANDLOCK)
 		goto err;
 
+	sb->s_flags &= ~MS_NOSEC;
+
 	if (!parse_fuse_opt((char *) data, &d, is_bdev))
 		goto err;
 
diff --git a/fs/generic_acl.c b/fs/generic_acl.c
index 8f26d1a5891..70e90b4974c 100644
--- a/fs/generic_acl.c
+++ b/fs/generic_acl.c
@@ -190,9 +190,9 @@ generic_acl_chmod(struct inode *inode)
 }
 
 int
-generic_check_acl(struct inode *inode, int mask, unsigned int flags)
+generic_check_acl(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 	} else {
diff --git a/fs/gfs2/acl.c b/fs/gfs2/acl.c
index cbc07155b1a..8ef1079f166 100644
--- a/fs/gfs2/acl.c
+++ b/fs/gfs2/acl.c
@@ -75,12 +75,12 @@ static struct posix_acl *gfs2_acl_get(struct gfs2_inode *ip, int type)
  * Returns: errno
  */
 
-int gfs2_check_acl(struct inode *inode, int mask, unsigned int flags)
+int gfs2_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	int error;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/gfs2/acl.h b/fs/gfs2/acl.h
index a93907c8159..b522b0cb39e 100644
--- a/fs/gfs2/acl.h
+++ b/fs/gfs2/acl.h
@@ -16,7 +16,7 @@
 #define GFS2_POSIX_ACL_DEFAULT		"posix_acl_default"
 #define GFS2_ACL_MAX_ENTRIES		25
 
-extern int gfs2_check_acl(struct inode *inode, int mask, unsigned int);
+extern int gfs2_check_acl(struct inode *inode, int mask);
 extern int gfs2_acl_create(struct gfs2_inode *dip, struct inode *inode);
 extern int gfs2_acl_chmod(struct gfs2_inode *ip, struct iattr *attr);
 extern const struct xattr_handler gfs2_xattr_system_handler;
diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 802ac5eeba2..f9fbbe96c22 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -1069,6 +1069,7 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 		return 0;
 
 	gfs2_log_lock(sdp);
+	spin_lock(&sdp->sd_ail_lock);
 	head = bh = page_buffers(page);
 	do {
 		if (atomic_read(&bh->b_count))
@@ -1080,6 +1081,7 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 			goto not_possible;
 		bh = bh->b_this_page;
 	} while(bh != head);
+	spin_unlock(&sdp->sd_ail_lock);
 	gfs2_log_unlock(sdp);
 
 	head = bh = page_buffers(page);
@@ -1112,6 +1114,7 @@ not_possible: /* Should never happen */
 	WARN_ON(buffer_dirty(bh));
 	WARN_ON(buffer_pinned(bh));
 cannot_release:
+	spin_unlock(&sdp->sd_ail_lock);
 	gfs2_log_unlock(sdp);
 	return 0;
 }
diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index e65493a8ac0..7878c473ae6 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -854,11 +854,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh,
 			blen++;
 		else {
 			if (bstart) {
-				if (metadata)
-					__gfs2_free_meta(ip, bstart, blen);
-				else
-					__gfs2_free_data(ip, bstart, blen);
-
+				__gfs2_free_blocks(ip, bstart, blen, metadata);
 				btotal += blen;
 			}
 
@@ -870,11 +866,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh,
 		gfs2_add_inode_blocks(&ip->i_inode, -1);
 	}
 	if (bstart) {
-		if (metadata)
-			__gfs2_free_meta(ip, bstart, blen);
-		else
-			__gfs2_free_data(ip, bstart, blen);
-
+		__gfs2_free_blocks(ip, bstart, blen, metadata);
 		btotal += blen;
 	}
 
@@ -1224,6 +1216,8 @@ int gfs2_setattr_size(struct inode *inode, u64 newsize)
 	if (ret)
 		return ret;
 
+	inode_dio_wait(inode);
+
 	oldsize = inode->i_size;
 	if (newsize >= oldsize)
 		return do_grow(inode, newsize);
diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c
index 091ee477953..1cc2f8ec52a 100644
--- a/fs/gfs2/dir.c
+++ b/fs/gfs2/dir.c
@@ -339,6 +339,67 @@ fail:
 	return (copied) ? copied : error;
 }
 
+/**
+ * gfs2_dir_get_hash_table - Get pointer to the dir hash table
+ * @ip: The inode in question
+ *
+ * Returns: The hash table or an error
+ */
+
+static __be64 *gfs2_dir_get_hash_table(struct gfs2_inode *ip)
+{
+	struct inode *inode = &ip->i_inode;
+	int ret;
+	u32 hsize;
+	__be64 *hc;
+
+	BUG_ON(!(ip->i_diskflags & GFS2_DIF_EXHASH));
+
+	hc = ip->i_hash_cache;
+	if (hc)
+		return hc;
+
+	hsize = 1 << ip->i_depth;
+	hsize *= sizeof(__be64);
+	if (hsize != i_size_read(&ip->i_inode)) {
+		gfs2_consist_inode(ip);
+		return ERR_PTR(-EIO);
+	}
+
+	hc = kmalloc(hsize, GFP_NOFS);
+	ret = -ENOMEM;
+	if (hc == NULL)
+		return ERR_PTR(-ENOMEM);
+
+	ret = gfs2_dir_read_data(ip, (char *)hc, 0, hsize, 1);
+	if (ret < 0) {
+		kfree(hc);
+		return ERR_PTR(ret);
+	}
+
+	spin_lock(&inode->i_lock);
+	if (ip->i_hash_cache)
+		kfree(hc);
+	else
+		ip->i_hash_cache = hc;
+	spin_unlock(&inode->i_lock);
+
+	return ip->i_hash_cache;
+}
+
+/**
+ * gfs2_dir_hash_inval - Invalidate dir hash
+ * @ip: The directory inode
+ *
+ * Must be called with an exclusive glock, or during glock invalidation.
+ */
+void gfs2_dir_hash_inval(struct gfs2_inode *ip)
+{
+	__be64 *hc = ip->i_hash_cache;
+	ip->i_hash_cache = NULL;
+	kfree(hc);
+}
+
 static inline int gfs2_dirent_sentinel(const struct gfs2_dirent *dent)
 {
 	return dent->de_inum.no_addr == 0 || dent->de_inum.no_formal_ino == 0;
@@ -686,17 +747,12 @@ static int get_leaf(struct gfs2_inode *dip, u64 leaf_no,
 static int get_leaf_nr(struct gfs2_inode *dip, u32 index,
 		       u64 *leaf_out)
 {
-	__be64 leaf_no;
-	int error;
-
-	error = gfs2_dir_read_data(dip, (char *)&leaf_no,
-				    index * sizeof(__be64),
-				    sizeof(__be64), 0);
-	if (error != sizeof(u64))
-		return (error < 0) ? error : -EIO;
-
-	*leaf_out = be64_to_cpu(leaf_no);
+	__be64 *hash;
 
+	hash = gfs2_dir_get_hash_table(dip);
+	if (IS_ERR(hash))
+		return PTR_ERR(hash);
+	*leaf_out = be64_to_cpu(*(hash + index));
 	return 0;
 }
 
@@ -966,6 +1022,8 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name)
 	for (x = 0; x < half_len; x++)
 		lp[x] = cpu_to_be64(bn);
 
+	gfs2_dir_hash_inval(dip);
+
 	error = gfs2_dir_write_data(dip, (char *)lp, start * sizeof(u64),
 				    half_len * sizeof(u64));
 	if (error != half_len * sizeof(u64)) {
@@ -1052,70 +1110,54 @@ fail_brelse:
 
 static int dir_double_exhash(struct gfs2_inode *dip)
 {
-	struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
 	struct buffer_head *dibh;
 	u32 hsize;
-	u64 *buf;
-	u64 *from, *to;
-	u64 block;
-	u64 disksize = i_size_read(&dip->i_inode);
+	u32 hsize_bytes;
+	__be64 *hc;
+	__be64 *hc2, *h;
 	int x;
 	int error = 0;
 
 	hsize = 1 << dip->i_depth;
-	if (hsize * sizeof(u64) != disksize) {
-		gfs2_consist_inode(dip);
-		return -EIO;
-	}
+	hsize_bytes = hsize * sizeof(__be64);
 
-	/*  Allocate both the "from" and "to" buffers in one big chunk  */
+	hc = gfs2_dir_get_hash_table(dip);
+	if (IS_ERR(hc))
+		return PTR_ERR(hc);
 
-	buf = kcalloc(3, sdp->sd_hash_bsize, GFP_NOFS);
-	if (!buf)
+	h = hc2 = kmalloc(hsize_bytes * 2, GFP_NOFS);
+	if (!hc2)
 		return -ENOMEM;
 
-	for (block = disksize >> sdp->sd_hash_bsize_shift; block--;) {
-		error = gfs2_dir_read_data(dip, (char *)buf,
-					    block * sdp->sd_hash_bsize,
-					    sdp->sd_hash_bsize, 1);
-		if (error != sdp->sd_hash_bsize) {
-			if (error >= 0)
-				error = -EIO;
-			goto fail;
-		}
-
-		from = buf;
-		to = (u64 *)((char *)buf + sdp->sd_hash_bsize);
-
-		for (x = sdp->sd_hash_ptrs; x--; from++) {
-			*to++ = *from;	/*  No endianess worries  */
-			*to++ = *from;
-		}
+	error = gfs2_meta_inode_buffer(dip, &dibh);
+	if (error)
+		goto out_kfree;
 
-		error = gfs2_dir_write_data(dip,
-					     (char *)buf + sdp->sd_hash_bsize,
-					     block * sdp->sd_sb.sb_bsize,
-					     sdp->sd_sb.sb_bsize);
-		if (error != sdp->sd_sb.sb_bsize) {
-			if (error >= 0)
-				error = -EIO;
-			goto fail;
-		}
+	for (x = 0; x < hsize; x++) {
+		*h++ = *hc;
+		*h++ = *hc;
+		hc++;
 	}
 
-	kfree(buf);
-
-	error = gfs2_meta_inode_buffer(dip, &dibh);
-	if (!gfs2_assert_withdraw(sdp, !error)) {
-		dip->i_depth++;
-		gfs2_dinode_out(dip, dibh->b_data);
-		brelse(dibh);
-	}
+	error = gfs2_dir_write_data(dip, (char *)hc2, 0, hsize_bytes * 2);
+	if (error != (hsize_bytes * 2))
+		goto fail;
 
-	return error;
+	gfs2_dir_hash_inval(dip);
+	dip->i_hash_cache = hc2;
+	dip->i_depth++;
+	gfs2_dinode_out(dip, dibh->b_data);
+	brelse(dibh);
+	return 0;
 
 fail:
-	kfree(buf);
+	/* Replace original hash table & size */
+	gfs2_dir_write_data(dip, (char *)hc, 0, hsize_bytes);
+	i_size_write(&dip->i_inode, hsize_bytes);
+	gfs2_dinode_out(dip, dibh->b_data);
+	brelse(dibh);
+out_kfree:
+	kfree(hc2);
 	return error;
 }
 
@@ -1348,6 +1390,7 @@ out:
 	return error;
 }
 
+
 /**
  * dir_e_read - Reads the entries from a directory into a filldir buffer
  * @dip: dinode pointer
@@ -1362,9 +1405,7 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque,
 		      filldir_t filldir)
 {
 	struct gfs2_inode *dip = GFS2_I(inode);
-	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	u32 hsize, len = 0;
-	u32 ht_offset, lp_offset, ht_offset_cur = -1;
 	u32 hash, index;
 	__be64 *lp;
 	int copied = 0;
@@ -1372,37 +1413,17 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque,
 	unsigned depth = 0;
 
 	hsize = 1 << dip->i_depth;
-	if (hsize * sizeof(u64) != i_size_read(inode)) {
-		gfs2_consist_inode(dip);
-		return -EIO;
-	}
-
 	hash = gfs2_dir_offset2hash(*offset);
 	index = hash >> (32 - dip->i_depth);
 
-	lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS);
-	if (!lp)
-		return -ENOMEM;
+	lp = gfs2_dir_get_hash_table(dip);
+	if (IS_ERR(lp))
+		return PTR_ERR(lp);
 
 	while (index < hsize) {
-		lp_offset = index & (sdp->sd_hash_ptrs - 1);
-		ht_offset = index - lp_offset;
-
-		if (ht_offset_cur != ht_offset) {
-			error = gfs2_dir_read_data(dip, (char *)lp,
-						ht_offset * sizeof(__be64),
-						sdp->sd_hash_bsize, 1);
-			if (error != sdp->sd_hash_bsize) {
-				if (error >= 0)
-					error = -EIO;
-				goto out;
-			}
-			ht_offset_cur = ht_offset;
-		}
-
 		error = gfs2_dir_read_leaf(inode, offset, opaque, filldir,
 					   &copied, &depth,
-					   be64_to_cpu(lp[lp_offset]));
+					   be64_to_cpu(lp[index]));
 		if (error)
 			break;
 
@@ -1410,8 +1431,6 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque,
 		index = (index & ~(len - 1)) + len;
 	}
 
-out:
-	kfree(lp);
 	if (error > 0)
 		error = 0;
 	return error;
@@ -1914,43 +1933,22 @@ out:
 
 int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip)
 {
-	struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
 	struct buffer_head *bh;
 	struct gfs2_leaf *leaf;
 	u32 hsize, len;
-	u32 ht_offset, lp_offset, ht_offset_cur = -1;
 	u32 index = 0, next_index;
 	__be64 *lp;
 	u64 leaf_no;
 	int error = 0, last;
 
 	hsize = 1 << dip->i_depth;
-	if (hsize * sizeof(u64) != i_size_read(&dip->i_inode)) {
-		gfs2_consist_inode(dip);
-		return -EIO;
-	}
 
-	lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS);
-	if (!lp)
-		return -ENOMEM;
+	lp = gfs2_dir_get_hash_table(dip);
+	if (IS_ERR(lp))
+		return PTR_ERR(lp);
 
 	while (index < hsize) {
-		lp_offset = index & (sdp->sd_hash_ptrs - 1);
-		ht_offset = index - lp_offset;
-
-		if (ht_offset_cur != ht_offset) {
-			error = gfs2_dir_read_data(dip, (char *)lp,
-						ht_offset * sizeof(__be64),
-						sdp->sd_hash_bsize, 1);
-			if (error != sdp->sd_hash_bsize) {
-				if (error >= 0)
-					error = -EIO;
-				goto out;
-			}
-			ht_offset_cur = ht_offset;
-		}
-
-		leaf_no = be64_to_cpu(lp[lp_offset]);
+		leaf_no = be64_to_cpu(lp[index]);
 		if (leaf_no) {
 			error = get_leaf(dip, leaf_no, &bh);
 			if (error)
@@ -1976,7 +1974,6 @@ int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip)
 	}
 
 out:
-	kfree(lp);
 
 	return error;
 }
diff --git a/fs/gfs2/dir.h b/fs/gfs2/dir.h
index e686af11bec..ff5772fbf02 100644
--- a/fs/gfs2/dir.h
+++ b/fs/gfs2/dir.h
@@ -35,6 +35,7 @@ extern int gfs2_diradd_alloc_required(struct inode *dir,
 				      const struct qstr *filename);
 extern int gfs2_dir_get_new_buffer(struct gfs2_inode *ip, u64 block,
 				   struct buffer_head **bhp);
+extern void gfs2_dir_hash_inval(struct gfs2_inode *ip);
 
 static inline u32 gfs2_disk_hash(const char *data, int len)
 {
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index a9f5cbe45cd..edeb9e80290 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -174,7 +174,9 @@ void gfs2_set_inode_flags(struct inode *inode)
 	struct gfs2_inode *ip = GFS2_I(inode);
 	unsigned int flags = inode->i_flags;
 
-	flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);
+	flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC|S_NOSEC);
+	if ((ip->i_eattr == 0) && !is_sxid(inode->i_mode))
+		inode->i_flags |= S_NOSEC;
 	if (ip->i_diskflags & GFS2_DIF_IMMUTABLE)
 		flags |= S_IMMUTABLE;
 	if (ip->i_diskflags & GFS2_DIF_APPENDONLY)
@@ -243,7 +245,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask)
 	    !capable(CAP_LINUX_IMMUTABLE))
 		goto out;
 	if (!IS_IMMUTABLE(inode)) {
-		error = gfs2_permission(inode, MAY_WRITE, 0);
+		error = gfs2_permission(inode, MAY_WRITE);
 		if (error)
 			goto out;
 	}
@@ -544,7 +546,9 @@ static int gfs2_close(struct inode *inode, struct file *file)
 
 /**
  * gfs2_fsync - sync the dirty data for a file (across the cluster)
- * @file: the file that points to the dentry (we ignore this)
+ * @file: the file that points to the dentry
+ * @start: the start position in the file to sync
+ * @end: the end position in the file to sync
  * @datasync: set if we can ignore timestamp changes
  *
  * The VFS will flush data for us. We only need to worry
@@ -553,23 +557,32 @@ static int gfs2_close(struct inode *inode, struct file *file)
  * Returns: errno
  */
 
-static int gfs2_fsync(struct file *file, int datasync)
+static int gfs2_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	int sync_state = inode->i_state & (I_DIRTY_SYNC|I_DIRTY_DATASYNC);
 	struct gfs2_inode *ip = GFS2_I(inode);
 	int ret;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	if (datasync)
 		sync_state &= ~I_DIRTY_SYNC;
 
 	if (sync_state) {
 		ret = sync_inode_metadata(inode, 1);
-		if (ret)
+		if (ret) {
+			mutex_unlock(&inode->i_mutex);
 			return ret;
+		}
 		gfs2_ail_flush(ip->i_gl);
 	}
 
+	mutex_unlock(&inode->i_mutex);
 	return 0;
 }
 
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 2792a790e50..88e8a23d002 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -409,6 +409,10 @@ static void state_change(struct gfs2_glock *gl, unsigned int new_state)
 	if (held1 && held2 && list_empty(&gl->gl_holders))
 		clear_bit(GLF_QUEUED, &gl->gl_flags);
 
+	if (new_state != gl->gl_target)
+		/* shorten our minimum hold time */
+		gl->gl_hold_time = max(gl->gl_hold_time - GL_GLOCK_HOLD_DECR,
+				       GL_GLOCK_MIN_HOLD);
 	gl->gl_state = new_state;
 	gl->gl_tchange = jiffies;
 }
@@ -663,20 +667,30 @@ static void glock_work_func(struct work_struct *work)
 		drop_ref = 1;
 	}
 	spin_lock(&gl->gl_spin);
-	if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
+	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
 	    gl->gl_state != LM_ST_UNLOCKED &&
 	    gl->gl_demote_state != LM_ST_EXCLUSIVE) {
 		unsigned long holdtime, now = jiffies;
-		holdtime = gl->gl_tchange + gl->gl_ops->go_min_hold_time;
+
+		holdtime = gl->gl_tchange + gl->gl_hold_time;
 		if (time_before(now, holdtime))
 			delay = holdtime - now;
-		set_bit(delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE, &gl->gl_flags);
+
+		if (!delay) {
+			clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags);
+			set_bit(GLF_DEMOTE, &gl->gl_flags);
+		}
 	}
 	run_queue(gl, 0);
 	spin_unlock(&gl->gl_spin);
-	if (!delay ||
-	    queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
+	if (!delay)
 		gfs2_glock_put(gl);
+	else {
+		if (gl->gl_name.ln_type != LM_TYPE_INODE)
+			delay = 0;
+		if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
+			gfs2_glock_put(gl);
+	}
 	if (drop_ref)
 		gfs2_glock_put(gl);
 }
@@ -738,6 +752,7 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 	gl->gl_tchange = jiffies;
 	gl->gl_object = NULL;
 	gl->gl_sbd = sdp;
+	gl->gl_hold_time = GL_GLOCK_DFT_HOLD;
 	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
 	INIT_WORK(&gl->gl_delete, delete_work_func);
 
@@ -850,8 +865,15 @@ static int gfs2_glock_demote_wait(void *word)
 
 static void wait_on_holder(struct gfs2_holder *gh)
 {
+	unsigned long time1 = jiffies;
+
 	might_sleep();
 	wait_on_bit(&gh->gh_iflags, HIF_WAIT, gfs2_glock_holder_wait, TASK_UNINTERRUPTIBLE);
+	if (time_after(jiffies, time1 + HZ)) /* have we waited > a second? */
+		/* Lengthen the minimum hold time. */
+		gh->gh_gl->gl_hold_time = min(gh->gh_gl->gl_hold_time +
+					      GL_GLOCK_HOLD_INCR,
+					      GL_GLOCK_MAX_HOLD);
 }
 
 static void wait_on_demote(struct gfs2_glock *gl)
@@ -1088,8 +1110,9 @@ void gfs2_glock_dq(struct gfs2_holder *gh)
 
 	gfs2_glock_hold(gl);
 	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
-	    !test_bit(GLF_DEMOTE, &gl->gl_flags))
-		delay = gl->gl_ops->go_min_hold_time;
+	    !test_bit(GLF_DEMOTE, &gl->gl_flags) &&
+	    gl->gl_name.ln_type == LM_TYPE_INODE)
+		delay = gl->gl_hold_time;
 	if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
 		gfs2_glock_put(gl);
 }
@@ -1268,12 +1291,13 @@ void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state)
 	unsigned long now = jiffies;
 
 	gfs2_glock_hold(gl);
-	holdtime = gl->gl_tchange + gl->gl_ops->go_min_hold_time;
-	if (test_bit(GLF_QUEUED, &gl->gl_flags)) {
+	holdtime = gl->gl_tchange + gl->gl_hold_time;
+	if (test_bit(GLF_QUEUED, &gl->gl_flags) &&
+	    gl->gl_name.ln_type == LM_TYPE_INODE) {
 		if (time_before(now, holdtime))
 			delay = holdtime - now;
 		if (test_bit(GLF_REPLY_PENDING, &gl->gl_flags))
-			delay = gl->gl_ops->go_min_hold_time;
+			delay = gl->gl_hold_time;
 	}
 
 	spin_lock(&gl->gl_spin);
@@ -1662,7 +1686,7 @@ 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/%llx f:%s t:%s d:%s/%llu a:%d v:%d r:%d\n",
+	gfs2_print_dbg(seq, "G:  s:%s n:%u/%llx f:%s t:%s d:%s/%llu a:%d v:%d r:%d m:%ld\n",
 		  state2str(gl->gl_state),
 		  gl->gl_name.ln_type,
 		  (unsigned long long)gl->gl_name.ln_number,
@@ -1671,7 +1695,7 @@ static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl)
 		  state2str(gl->gl_demote_state), dtime,
 		  atomic_read(&gl->gl_ail_count),
 		  atomic_read(&gl->gl_revokes),
-		  atomic_read(&gl->gl_ref));
+		  atomic_read(&gl->gl_ref), gl->gl_hold_time);
 
 	list_for_each_entry(gh, &gl->gl_holders, gh_list) {
 		error = dump_holder(seq, gh);
diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h
index 6b2f757b928..66707118af2 100644
--- a/fs/gfs2/glock.h
+++ b/fs/gfs2/glock.h
@@ -113,6 +113,12 @@ enum {
 
 #define GLR_TRYFAILED		13
 
+#define GL_GLOCK_MAX_HOLD        (long)(HZ / 5)
+#define GL_GLOCK_DFT_HOLD        (long)(HZ / 5)
+#define GL_GLOCK_MIN_HOLD        (long)(10)
+#define GL_GLOCK_HOLD_INCR       (long)(HZ / 20)
+#define GL_GLOCK_HOLD_DECR       (long)(HZ / 40)
+
 struct lm_lockops {
 	const char *lm_proto_name;
 	int (*lm_mount) (struct gfs2_sbd *sdp, const char *fsname);
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 8ef70f46473..da21ecaafcc 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -26,6 +26,7 @@
 #include "rgrp.h"
 #include "util.h"
 #include "trans.h"
+#include "dir.h"
 
 /**
  * __gfs2_ail_flush - remove all buffers for a given lock from the AIL
@@ -47,10 +48,10 @@ static void __gfs2_ail_flush(struct gfs2_glock *gl)
 				bd_ail_gl_list);
 		bh = bd->bd_bh;
 		gfs2_remove_from_ail(bd);
-		spin_unlock(&sdp->sd_ail_lock);
-
 		bd->bd_bh = NULL;
 		bh->b_private = NULL;
+		spin_unlock(&sdp->sd_ail_lock);
+
 		bd->bd_blkno = bh->b_blocknr;
 		gfs2_log_lock(sdp);
 		gfs2_assert_withdraw(sdp, !buffer_busy(bh));
@@ -218,11 +219,14 @@ static void inode_go_inval(struct gfs2_glock *gl, int flags)
 		if (ip) {
 			set_bit(GIF_INVALID, &ip->i_flags);
 			forget_all_cached_acls(&ip->i_inode);
+			gfs2_dir_hash_inval(ip);
 		}
 	}
 
-	if (ip == GFS2_I(gl->gl_sbd->sd_rindex))
+	if (ip == GFS2_I(gl->gl_sbd->sd_rindex)) {
+		gfs2_log_flush(gl->gl_sbd, NULL);
 		gl->gl_sbd->sd_rindex_uptodate = 0;
+	}
 	if (ip && S_ISREG(ip->i_inode.i_mode))
 		truncate_inode_pages(ip->i_inode.i_mapping, 0);
 }
@@ -314,6 +318,8 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf)
 	ip->i_generation = be64_to_cpu(str->di_generation);
 
 	ip->i_diskflags = be32_to_cpu(str->di_flags);
+	ip->i_eattr = be64_to_cpu(str->di_eattr);
+	/* i_diskflags and i_eattr must be set before gfs2_set_inode_flags() */
 	gfs2_set_inode_flags(&ip->i_inode);
 	height = be16_to_cpu(str->di_height);
 	if (unlikely(height > GFS2_MAX_META_HEIGHT))
@@ -326,7 +332,6 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf)
 	ip->i_depth = (u8)depth;
 	ip->i_entries = be32_to_cpu(str->di_entries);
 
-	ip->i_eattr = be64_to_cpu(str->di_eattr);
 	if (S_ISREG(ip->i_inode.i_mode))
 		gfs2_set_aops(&ip->i_inode);
 
@@ -547,7 +552,6 @@ const struct gfs2_glock_operations gfs2_inode_glops = {
 	.go_lock = inode_go_lock,
 	.go_dump = inode_go_dump,
 	.go_type = LM_TYPE_INODE,
-	.go_min_hold_time = HZ / 5,
 	.go_flags = GLOF_ASPACE,
 };
 
@@ -558,7 +562,6 @@ const struct gfs2_glock_operations gfs2_rgrp_glops = {
 	.go_unlock = rgrp_go_unlock,
 	.go_dump = gfs2_rgrp_dump,
 	.go_type = LM_TYPE_RGRP,
-	.go_min_hold_time = HZ / 5,
 	.go_flags = GLOF_ASPACE,
 };
 
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 0a064e91ac7..892ac37de8a 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -17,6 +17,7 @@
 #include <linux/buffer_head.h>
 #include <linux/rcupdate.h>
 #include <linux/rculist_bl.h>
+#include <linux/completion.h>
 
 #define DIO_WAIT	0x00000010
 #define DIO_METADATA	0x00000020
@@ -162,7 +163,6 @@ struct gfs2_glock_operations {
 	int (*go_dump)(struct seq_file *seq, const struct gfs2_glock *gl);
 	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
 };
@@ -220,6 +220,7 @@ struct gfs2_glock {
 
 	unsigned int gl_hash;
 	unsigned long gl_demote_time; /* time of first demote request */
+	long gl_hold_time;
 	struct list_head gl_holders;
 
 	const struct gfs2_glock_operations *gl_ops;
@@ -284,6 +285,7 @@ struct gfs2_inode {
 	u64 i_goal;	/* goal block for allocations */
 	struct rw_semaphore i_rw_mutex;
 	struct list_head i_trunc_list;
+	__be64 *i_hash_cache;
 	u32 i_entries;
 	u32 i_diskflags;
 	u8 i_height;
@@ -546,6 +548,7 @@ struct gfs2_sbd {
 	struct gfs2_glock *sd_trans_gl;
 	wait_queue_head_t sd_glock_wait;
 	atomic_t sd_glock_disposal;
+	struct completion sd_locking_init;
 
 	/* Inode Stuff */
 
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 03e0c529063..0fb51a96eff 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -307,7 +307,7 @@ struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name,
 	}
 
 	if (!is_root) {
-		error = gfs2_permission(dir, MAY_EXEC, 0);
+		error = gfs2_permission(dir, MAY_EXEC);
 		if (error)
 			goto out;
 	}
@@ -337,7 +337,7 @@ static int create_ok(struct gfs2_inode *dip, const struct qstr *name,
 {
 	int error;
 
-	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0);
+	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC);
 	if (error)
 		return error;
 
@@ -792,13 +792,8 @@ static int gfs2_create(struct inode *dir, struct dentry *dentry,
 static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry,
 				  struct nameidata *nd)
 {
-	struct inode *inode = NULL;
-
-	inode = gfs2_lookupi(dir, &dentry->d_name, 0);
-	if (inode && IS_ERR(inode))
-		return ERR_CAST(inode);
-
-	if (inode) {
+	struct inode *inode = gfs2_lookupi(dir, &dentry->d_name, 0);
+	if (inode && !IS_ERR(inode)) {
 		struct gfs2_glock *gl = GFS2_I(inode)->i_gl;
 		struct gfs2_holder gh;
 		int error;
@@ -808,11 +803,8 @@ static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry,
 			return ERR_PTR(error);
 		}
 		gfs2_glock_dq_uninit(&gh);
-		return d_splice_alias(inode, dentry);
 	}
-	d_add(dentry, inode);
-
-	return NULL;
+	return d_splice_alias(inode, dentry);
 }
 
 /**
@@ -857,7 +849,7 @@ static int gfs2_link(struct dentry *old_dentry, struct inode *dir,
 	if (inode->i_nlink == 0)
 		goto out_gunlock;
 
-	error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC, 0);
+	error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC);
 	if (error)
 		goto out_gunlock;
 
@@ -990,7 +982,7 @@ static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name,
 	if (IS_APPEND(&dip->i_inode))
 		return -EPERM;
 
-	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0);
+	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC);
 	if (error)
 		return error;
 
@@ -1336,7 +1328,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry,
 			}
 		}
 	} else {
-		error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC, 0);
+		error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC);
 		if (error)
 			goto out_gunlock;
 
@@ -1371,7 +1363,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry,
 	/* Check out the dir to be renamed */
 
 	if (dir_rename) {
-		error = gfs2_permission(odentry->d_inode, MAY_WRITE, 0);
+		error = gfs2_permission(odentry->d_inode, MAY_WRITE);
 		if (error)
 			goto out_gunlock;
 	}
@@ -1543,7 +1535,7 @@ static void gfs2_put_link(struct dentry *dentry, struct nameidata *nd, void *p)
  * Returns: errno
  */
 
-int gfs2_permission(struct inode *inode, int mask, unsigned int flags)
+int gfs2_permission(struct inode *inode, int mask)
 {
 	struct gfs2_inode *ip;
 	struct gfs2_holder i_gh;
@@ -1553,7 +1545,7 @@ int gfs2_permission(struct inode *inode, int mask, unsigned int flags)
 
 	ip = GFS2_I(inode);
 	if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) {
-		if (flags & IPERM_FLAG_RCU)
+		if (mask & MAY_NOT_BLOCK)
 			return -ECHILD;
 		error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh);
 		if (error)
@@ -1564,7 +1556,7 @@ int gfs2_permission(struct inode *inode, int mask, unsigned int flags)
 	if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode))
 		error = -EACCES;
 	else
-		error = generic_permission(inode, mask, flags, gfs2_check_acl);
+		error = generic_permission(inode, mask);
 	if (unlock)
 		gfs2_glock_dq_uninit(&i_gh);
 
@@ -1854,6 +1846,7 @@ const struct inode_operations gfs2_file_iops = {
 	.listxattr = gfs2_listxattr,
 	.removexattr = gfs2_removexattr,
 	.fiemap = gfs2_fiemap,
+	.check_acl = gfs2_check_acl,
 };
 
 const struct inode_operations gfs2_dir_iops = {
@@ -1874,6 +1867,7 @@ const struct inode_operations gfs2_dir_iops = {
 	.listxattr = gfs2_listxattr,
 	.removexattr = gfs2_removexattr,
 	.fiemap = gfs2_fiemap,
+	.check_acl = gfs2_check_acl,
 };
 
 const struct inode_operations gfs2_symlink_iops = {
@@ -1888,5 +1882,6 @@ const struct inode_operations gfs2_symlink_iops = {
 	.listxattr = gfs2_listxattr,
 	.removexattr = gfs2_removexattr,
 	.fiemap = gfs2_fiemap,
+	.check_acl = gfs2_check_acl,
 };
 
diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h
index 31606076f70..8d90e0c0767 100644
--- a/fs/gfs2/inode.h
+++ b/fs/gfs2/inode.h
@@ -108,7 +108,7 @@ extern int gfs2_inode_refresh(struct gfs2_inode *ip);
 
 extern struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name,
 				  int is_root);
-extern int gfs2_permission(struct inode *inode, int mask, unsigned int flags);
+extern int gfs2_permission(struct inode *inode, int mask);
 extern int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr);
 extern struct inode *gfs2_lookup_simple(struct inode *dip, const char *name);
 extern void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf);
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 903115f2bb3..85c62923ee2 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -903,6 +903,7 @@ void gfs2_meta_syncfs(struct gfs2_sbd *sdp)
 		if (gfs2_ail1_empty(sdp))
 			break;
 	}
+	gfs2_log_flush(sdp, NULL);
 }
 
 static inline int gfs2_jrnl_flush_reqd(struct gfs2_sbd *sdp)
diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c
index c2b34cd2abe..29e1ace7953 100644
--- a/fs/gfs2/main.c
+++ b/fs/gfs2/main.c
@@ -41,6 +41,7 @@ static void gfs2_init_inode_once(void *foo)
 	init_rwsem(&ip->i_rw_mutex);
 	INIT_LIST_HEAD(&ip->i_trunc_list);
 	ip->i_alloc = NULL;
+	ip->i_hash_cache = NULL;
 }
 
 static void gfs2_init_glock_once(void *foo)
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 8ac9ae189b5..516516e0c2a 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -72,6 +72,7 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb)
 
 	init_waitqueue_head(&sdp->sd_glock_wait);
 	atomic_set(&sdp->sd_glock_disposal, 0);
+	init_completion(&sdp->sd_locking_init);
 	spin_lock_init(&sdp->sd_statfs_spin);
 
 	spin_lock_init(&sdp->sd_rindex_spin);
@@ -1017,11 +1018,13 @@ hostdata_error:
 		fsname++;
 	if (lm->lm_mount == NULL) {
 		fs_info(sdp, "Now mounting FS...\n");
+		complete(&sdp->sd_locking_init);
 		return 0;
 	}
 	ret = lm->lm_mount(sdp, fsname);
 	if (ret == 0)
 		fs_info(sdp, "Joined cluster. Now mounting FS...\n");
+	complete(&sdp->sd_locking_init);
 	return ret;
 }
 
@@ -1091,6 +1094,7 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent
 	if (sdp->sd_args.ar_nobarrier)
 		set_bit(SDF_NOBARRIERS, &sdp->sd_flags);
 
+	sb->s_flags |= MS_NOSEC;
 	sb->s_magic = GFS2_MAGIC;
 	sb->s_op = &gfs2_super_ops;
 	sb->s_d_op = &gfs2_dops;
diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index 9b780df3fd5..7f8af1eb02d 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -1607,14 +1607,15 @@ rgrp_error:
 }
 
 /**
- * gfs2_free_data - free a contiguous run of data block(s)
+ * __gfs2_free_blocks - free a contiguous run of block(s)
  * @ip: the inode these blocks are being freed from
  * @bstart: first block of a run of contiguous blocks
  * @blen: the length of the block run
+ * @meta: 1 if the blocks represent metadata
  *
  */
 
-void __gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen)
+void __gfs2_free_blocks(struct gfs2_inode *ip, u64 bstart, u32 blen, int meta)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 	struct gfs2_rgrpd *rgd;
@@ -1631,54 +1632,11 @@ void __gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen)
 	gfs2_trans_add_rg(rgd);
 
 	/* Directories keep their data in the metadata address space */
-	if (ip->i_depth)
+	if (meta || ip->i_depth)
 		gfs2_meta_wipe(ip, bstart, blen);
 }
 
 /**
- * gfs2_free_data - free a contiguous run of data block(s)
- * @ip: the inode these blocks are being freed from
- * @bstart: first block of a run of contiguous blocks
- * @blen: the length of the block run
- *
- */
-
-void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen)
-{
-	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
-
-	__gfs2_free_data(ip, bstart, blen);
-	gfs2_statfs_change(sdp, 0, +blen, 0);
-	gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
-}
-
-/**
- * gfs2_free_meta - free a contiguous run of data block(s)
- * @ip: the inode these blocks are being freed from
- * @bstart: first block of a run of contiguous blocks
- * @blen: the length of the block run
- *
- */
-
-void __gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen)
-{
-	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
-	struct gfs2_rgrpd *rgd;
-
-	rgd = rgblk_free(sdp, bstart, blen, GFS2_BLKST_FREE);
-	if (!rgd)
-		return;
-	trace_gfs2_block_alloc(ip, bstart, blen, GFS2_BLKST_FREE);
-	rgd->rd_free += blen;
-
-	gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1);
-	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
-
-	gfs2_trans_add_rg(rgd);
-	gfs2_meta_wipe(ip, bstart, blen);
-}
-
-/**
  * gfs2_free_meta - free a contiguous run of data block(s)
  * @ip: the inode these blocks are being freed from
  * @bstart: first block of a run of contiguous blocks
@@ -1690,7 +1648,7 @@ void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 
-	__gfs2_free_meta(ip, bstart, blen);
+	__gfs2_free_blocks(ip, bstart, blen, 1);
 	gfs2_statfs_change(sdp, 0, +blen, 0);
 	gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid);
 }
diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h
index a80e3034ac4..d253f9a8c70 100644
--- a/fs/gfs2/rgrp.h
+++ b/fs/gfs2/rgrp.h
@@ -52,9 +52,7 @@ extern int gfs2_ri_update(struct gfs2_inode *ip);
 extern int gfs2_alloc_block(struct gfs2_inode *ip, u64 *bn, unsigned int *n);
 extern int gfs2_alloc_di(struct gfs2_inode *ip, u64 *bn, u64 *generation);
 
-extern void __gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen);
-extern void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen);
-extern void __gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen);
+extern void __gfs2_free_blocks(struct gfs2_inode *ip, u64 bstart, u32 blen, int meta);
 extern void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen);
 extern void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip);
 extern void gfs2_unlink_di(struct inode *inode);
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index ed540e7018b..b7beadd9ba4 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -757,13 +757,17 @@ static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc)
 	struct timespec atime;
 	struct gfs2_dinode *di;
 	int ret = -EAGAIN;
+	int unlock_required = 0;
 
 	/* Skip timestamp update, if this is from a memalloc */
 	if (current->flags & PF_MEMALLOC)
 		goto do_flush;
-	ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
-	if (ret)
-		goto do_flush;
+	if (!gfs2_glock_is_locked_by_me(ip->i_gl)) {
+		ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
+		if (ret)
+			goto do_flush;
+		unlock_required = 1;
+	}
 	ret = gfs2_trans_begin(sdp, RES_DINODE, 0);
 	if (ret)
 		goto do_unlock;
@@ -780,7 +784,8 @@ static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc)
 	}
 	gfs2_trans_end(sdp);
 do_unlock:
-	gfs2_glock_dq_uninit(&gh);
+	if (unlock_required)
+		gfs2_glock_dq_uninit(&gh);
 do_flush:
 	if (wbc->sync_mode == WB_SYNC_ALL)
 		gfs2_log_flush(GFS2_SB(inode), ip->i_gl);
@@ -1427,7 +1432,20 @@ out:
 	return error;
 }
 
-/*
+/**
+ * gfs2_evict_inode - Remove an inode from cache
+ * @inode: The inode to evict
+ *
+ * There are three cases to consider:
+ * 1. i_nlink == 0, we are final opener (and must deallocate)
+ * 2. i_nlink == 0, we are not the final opener (and cannot deallocate)
+ * 3. i_nlink > 0
+ *
+ * If the fs is read only, then we have to treat all cases as per #3
+ * since we are unable to do any deallocation. The inode will be
+ * deallocated by the next read/write node to attempt an allocation
+ * in the same resource group
+ *
  * We have to (at the moment) hold the inodes main lock to cover
  * the gap between unlocking the shared lock on the iopen lock and
  * taking the exclusive lock. I'd rather do a shared -> exclusive
@@ -1470,6 +1488,8 @@ static void gfs2_evict_inode(struct inode *inode)
 	if (error)
 		goto out_truncate;
 
+	/* Case 1 starts here */
+
 	if (S_ISDIR(inode->i_mode) &&
 	    (ip->i_diskflags & GFS2_DIF_EXHASH)) {
 		error = gfs2_dir_exhash_dealloc(ip);
@@ -1493,13 +1513,16 @@ static void gfs2_evict_inode(struct inode *inode)
 	goto out_unlock;
 
 out_truncate:
+	/* Case 2 starts here */
 	error = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks);
 	if (error)
 		goto out_unlock;
-	gfs2_final_release_pages(ip);
+	/* Needs to be done before glock release & also in a transaction */
+	truncate_inode_pages(&inode->i_data, 0);
 	gfs2_trans_end(sdp);
 
 out_unlock:
+	/* Error path for case 1 */
 	if (test_bit(HIF_HOLDER, &ip->i_iopen_gh.gh_iflags))
 		gfs2_glock_dq(&ip->i_iopen_gh);
 	gfs2_holder_uninit(&ip->i_iopen_gh);
@@ -1507,9 +1530,10 @@ out_unlock:
 	if (error && error != GLR_TRYFAILED && error != -EROFS)
 		fs_warn(sdp, "gfs2_evict_inode: %d\n", error);
 out:
+	/* Case 3 starts here */
 	truncate_inode_pages(&inode->i_data, 0);
 	end_writeback(inode);
-
+	gfs2_dir_hash_inval(ip);
 	ip->i_gl->gl_object = NULL;
 	gfs2_glock_add_to_lru(ip->i_gl);
 	gfs2_glock_put(ip->i_gl);
diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c
index e20eab37bc8..443cabcfcd2 100644
--- a/fs/gfs2/sys.c
+++ b/fs/gfs2/sys.c
@@ -338,6 +338,9 @@ static ssize_t lkfirst_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
 	rv = sscanf(buf, "%u", &first);
 	if (rv != 1 || first > 1)
 		return -EINVAL;
+	rv = wait_for_completion_killable(&sdp->sd_locking_init);
+	if (rv)
+		return rv;
 	spin_lock(&sdp->sd_jindex_spin);
 	rv = -EBUSY;
 	if (test_bit(SDF_NOJOURNALID, &sdp->sd_flags) == 0)
@@ -414,7 +417,9 @@ static ssize_t jid_store(struct gfs2_sbd *sdp, const char *buf, size_t len)
 	rv = sscanf(buf, "%d", &jid);
 	if (rv != 1)
 		return -EINVAL;
-
+	rv = wait_for_completion_killable(&sdp->sd_locking_init);
+	if (rv)
+		return rv;
 	spin_lock(&sdp->sd_jindex_spin);
 	rv = -EINVAL;
 	if (sdp->sd_lockstruct.ls_ops->lm_mount == NULL)
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index fff16c968e6..96a1b625fc7 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -123,8 +123,8 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_path.dentry->d_inode->i_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs, hfs_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 hfs_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
@@ -615,6 +615,8 @@ int hfs_inode_setattr(struct dentry *dentry, struct iattr * attr)
 
 	if ((attr->ia_valid & ATTR_SIZE) &&
 	    attr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		error = vmtruncate(inode, attr->ia_size);
 		if (error)
 			return error;
@@ -625,12 +627,18 @@ int hfs_inode_setattr(struct dentry *dentry, struct iattr * attr)
 	return 0;
 }
 
-static int hfs_file_fsync(struct file *filp, int datasync)
+static int hfs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			  int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	struct super_block * sb;
 	int ret, err;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	/* sync the inode to buffers */
 	ret = write_inode_now(inode, 0);
 
@@ -647,6 +655,7 @@ static int hfs_file_fsync(struct file *filp, int datasync)
 	err = sync_blockdev(sb->s_bdev);
 	if (!ret)
 		ret = err;
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/hfsplus/brec.c b/fs/hfsplus/brec.c
index 2312de34bd4..2a734cfccc9 100644
--- a/fs/hfsplus/brec.c
+++ b/fs/hfsplus/brec.c
@@ -43,6 +43,10 @@ u16 hfs_brec_keylen(struct hfs_bnode *node, u16 rec)
 			node->tree->node_size - (rec + 1) * 2);
 		if (!recoff)
 			return 0;
+		if (recoff > node->tree->node_size - 2) {
+			printk(KERN_ERR "hfs: recoff %d too large\n", recoff);
+			return 0;
+		}
 
 		retval = hfs_bnode_read_u16(node, recoff) + 2;
 		if (retval > node->tree->max_key_len + 2) {
diff --git a/fs/hfsplus/catalog.c b/fs/hfsplus/catalog.c
index b4ba1b31933..4dfbfec357e 100644
--- a/fs/hfsplus/catalog.c
+++ b/fs/hfsplus/catalog.c
@@ -212,7 +212,9 @@ int hfsplus_create_cat(u32 cnid, struct inode *dir,
 
 	dprint(DBG_CAT_MOD, "create_cat: %s,%u(%d)\n",
 		str->name, cnid, inode->i_nlink);
-	hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	err = hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	if (err)
+		return err;
 
 	hfsplus_cat_build_key(sb, fd.search_key, cnid, NULL);
 	entry_size = hfsplus_fill_cat_thread(sb, &entry,
@@ -269,7 +271,9 @@ int hfsplus_delete_cat(u32 cnid, struct inode *dir, struct qstr *str)
 
 	dprint(DBG_CAT_MOD, "delete_cat: %s,%u\n",
 		str ? str->name : NULL, cnid);
-	hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	err = hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	if (err)
+		return err;
 
 	if (!str) {
 		int len;
@@ -347,12 +351,14 @@ int hfsplus_rename_cat(u32 cnid,
 	struct hfs_find_data src_fd, dst_fd;
 	hfsplus_cat_entry entry;
 	int entry_size, type;
-	int err = 0;
+	int err;
 
 	dprint(DBG_CAT_MOD, "rename_cat: %u - %lu,%s - %lu,%s\n",
 		cnid, src_dir->i_ino, src_name->name,
 		dst_dir->i_ino, dst_name->name);
-	hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &src_fd);
+	err = hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &src_fd);
+	if (err)
+		return err;
 	dst_fd = src_fd;
 
 	/* find the old dir entry and read the data */
diff --git a/fs/hfsplus/dir.c b/fs/hfsplus/dir.c
index 4df5059c25d..25b2443a004 100644
--- a/fs/hfsplus/dir.c
+++ b/fs/hfsplus/dir.c
@@ -38,7 +38,9 @@ static struct dentry *hfsplus_lookup(struct inode *dir, struct dentry *dentry,
 	sb = dir->i_sb;
 
 	dentry->d_fsdata = NULL;
-	hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	err = hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	if (err)
+		return ERR_PTR(err);
 	hfsplus_cat_build_key(sb, fd.search_key, dir->i_ino, &dentry->d_name);
 again:
 	err = hfs_brec_read(&fd, &entry, sizeof(entry));
@@ -132,7 +134,9 @@ static int hfsplus_readdir(struct file *filp, void *dirent, filldir_t filldir)
 	if (filp->f_pos >= inode->i_size)
 		return 0;
 
-	hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	err = hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	if (err)
+		return err;
 	hfsplus_cat_build_key(sb, fd.search_key, inode->i_ino, NULL);
 	err = hfs_brec_find(&fd);
 	if (err)
diff --git a/fs/hfsplus/extents.c b/fs/hfsplus/extents.c
index b1991a2a08e..5849e3ef35c 100644
--- a/fs/hfsplus/extents.c
+++ b/fs/hfsplus/extents.c
@@ -119,22 +119,31 @@ static void __hfsplus_ext_write_extent(struct inode *inode,
 	set_bit(HFSPLUS_I_EXT_DIRTY, &hip->flags);
 }
 
-static void hfsplus_ext_write_extent_locked(struct inode *inode)
+static int hfsplus_ext_write_extent_locked(struct inode *inode)
 {
+	int res;
+
 	if (HFSPLUS_I(inode)->extent_state & HFSPLUS_EXT_DIRTY) {
 		struct hfs_find_data fd;
 
-		hfs_find_init(HFSPLUS_SB(inode->i_sb)->ext_tree, &fd);
+		res = hfs_find_init(HFSPLUS_SB(inode->i_sb)->ext_tree, &fd);
+		if (res)
+			return res;
 		__hfsplus_ext_write_extent(inode, &fd);
 		hfs_find_exit(&fd);
 	}
+	return 0;
 }
 
-void hfsplus_ext_write_extent(struct inode *inode)
+int hfsplus_ext_write_extent(struct inode *inode)
 {
+	int res;
+
 	mutex_lock(&HFSPLUS_I(inode)->extents_lock);
-	hfsplus_ext_write_extent_locked(inode);
+	res = hfsplus_ext_write_extent_locked(inode);
 	mutex_unlock(&HFSPLUS_I(inode)->extents_lock);
+
+	return res;
 }
 
 static inline int __hfsplus_ext_read_extent(struct hfs_find_data *fd,
@@ -194,9 +203,11 @@ static int hfsplus_ext_read_extent(struct inode *inode, u32 block)
 	    block < hip->cached_start + hip->cached_blocks)
 		return 0;
 
-	hfs_find_init(HFSPLUS_SB(inode->i_sb)->ext_tree, &fd);
-	res = __hfsplus_ext_cache_extent(&fd, inode, block);
-	hfs_find_exit(&fd);
+	res = hfs_find_init(HFSPLUS_SB(inode->i_sb)->ext_tree, &fd);
+	if (!res) {
+		res = __hfsplus_ext_cache_extent(&fd, inode, block);
+		hfs_find_exit(&fd);
+	}
 	return res;
 }
 
@@ -209,6 +220,7 @@ int hfsplus_get_block(struct inode *inode, sector_t iblock,
 	struct hfsplus_inode_info *hip = HFSPLUS_I(inode);
 	int res = -EIO;
 	u32 ablock, dblock, mask;
+	sector_t sector;
 	int was_dirty = 0;
 	int shift;
 
@@ -255,10 +267,12 @@ int hfsplus_get_block(struct inode *inode, sector_t iblock,
 done:
 	dprint(DBG_EXTENT, "get_block(%lu): %llu - %u\n",
 		inode->i_ino, (long long)iblock, dblock);
+
 	mask = (1 << sbi->fs_shift) - 1;
-	map_bh(bh_result, sb,
-		(dblock << sbi->fs_shift) + sbi->blockoffset +
-			(iblock & mask));
+	sector = ((sector_t)dblock << sbi->fs_shift) +
+		  sbi->blockoffset + (iblock & mask);
+	map_bh(bh_result, sb, sector);
+
 	if (create) {
 		set_buffer_new(bh_result);
 		hip->phys_size += sb->s_blocksize;
@@ -371,7 +385,9 @@ int hfsplus_free_fork(struct super_block *sb, u32 cnid,
 	if (total_blocks == blocks)
 		return 0;
 
-	hfs_find_init(HFSPLUS_SB(sb)->ext_tree, &fd);
+	res = hfs_find_init(HFSPLUS_SB(sb)->ext_tree, &fd);
+	if (res)
+		return res;
 	do {
 		res = __hfsplus_ext_read_extent(&fd, ext_entry, cnid,
 						total_blocks, type);
@@ -469,7 +485,9 @@ out:
 
 insert_extent:
 	dprint(DBG_EXTENT, "insert new extent\n");
-	hfsplus_ext_write_extent_locked(inode);
+	res = hfsplus_ext_write_extent_locked(inode);
+	if (res)
+		goto out;
 
 	memset(hip->cached_extents, 0, sizeof(hfsplus_extent_rec));
 	hip->cached_extents[0].start_block = cpu_to_be32(start);
@@ -500,7 +518,6 @@ void hfsplus_file_truncate(struct inode *inode)
 		struct page *page;
 		void *fsdata;
 		u32 size = inode->i_size;
-		int res;
 
 		res = pagecache_write_begin(NULL, mapping, size, 0,
 						AOP_FLAG_UNINTERRUPTIBLE,
@@ -523,7 +540,12 @@ void hfsplus_file_truncate(struct inode *inode)
 		goto out;
 
 	mutex_lock(&hip->extents_lock);
-	hfs_find_init(HFSPLUS_SB(sb)->ext_tree, &fd);
+	res = hfs_find_init(HFSPLUS_SB(sb)->ext_tree, &fd);
+	if (res) {
+		mutex_unlock(&hip->extents_lock);
+		/* XXX: We lack error handling of hfsplus_file_truncate() */
+		return;
+	}
 	while (1) {
 		if (alloc_cnt == hip->first_blocks) {
 			hfsplus_free_extents(sb, hip->first_extents,
diff --git a/fs/hfsplus/hfsplus_fs.h b/fs/hfsplus/hfsplus_fs.h
index d6857523336..d7674d051f5 100644
--- a/fs/hfsplus/hfsplus_fs.h
+++ b/fs/hfsplus/hfsplus_fs.h
@@ -13,6 +13,7 @@
 #include <linux/fs.h>
 #include <linux/mutex.h>
 #include <linux/buffer_head.h>
+#include <linux/blkdev.h>
 #include "hfsplus_raw.h"
 
 #define DBG_BNODE_REFS	0x00000001
@@ -110,7 +111,9 @@ struct hfsplus_vh;
 struct hfs_btree;
 
 struct hfsplus_sb_info {
+	void *s_vhdr_buf;
 	struct hfsplus_vh *s_vhdr;
+	void *s_backup_vhdr_buf;
 	struct hfsplus_vh *s_backup_vhdr;
 	struct hfs_btree *ext_tree;
 	struct hfs_btree *cat_tree;
@@ -258,6 +261,15 @@ struct hfsplus_readdir_data {
 	struct hfsplus_cat_key key;
 };
 
+/*
+ * Find minimum acceptible I/O size for an hfsplus sb.
+ */
+static inline unsigned short hfsplus_min_io_size(struct super_block *sb)
+{
+	return max_t(unsigned short, bdev_logical_block_size(sb->s_bdev),
+		     HFSPLUS_SECTOR_SIZE);
+}
+
 #define hfs_btree_open hfsplus_btree_open
 #define hfs_btree_close hfsplus_btree_close
 #define hfs_btree_write hfsplus_btree_write
@@ -374,7 +386,7 @@ extern const struct file_operations hfsplus_dir_operations;
 
 /* extents.c */
 int hfsplus_ext_cmp_key(const hfsplus_btree_key *, const hfsplus_btree_key *);
-void hfsplus_ext_write_extent(struct inode *);
+int hfsplus_ext_write_extent(struct inode *);
 int hfsplus_get_block(struct inode *, sector_t, struct buffer_head *, int);
 int hfsplus_free_fork(struct super_block *, u32,
 		struct hfsplus_fork_raw *, int);
@@ -392,7 +404,8 @@ int hfsplus_cat_read_inode(struct inode *, struct hfs_find_data *);
 int hfsplus_cat_write_inode(struct inode *);
 struct inode *hfsplus_new_inode(struct super_block *, int);
 void hfsplus_delete_inode(struct inode *);
-int hfsplus_file_fsync(struct file *file, int datasync);
+int hfsplus_file_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync);
 
 /* ioctl.c */
 long hfsplus_ioctl(struct file *filp, unsigned int cmd, unsigned long arg);
@@ -436,8 +449,8 @@ int hfsplus_compare_dentry(const struct dentry *parent,
 /* wrapper.c */
 int hfsplus_read_wrapper(struct super_block *);
 int hfs_part_find(struct super_block *, sector_t *, sector_t *);
-int hfsplus_submit_bio(struct block_device *bdev, sector_t sector,
-		void *data, int rw);
+int hfsplus_submit_bio(struct super_block *sb, sector_t sector,
+		void *buf, void **data, int rw);
 
 /* time macros */
 #define __hfsp_mt2ut(t)		(be32_to_cpu(t) - 2082844800U)
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index b248a6cfcad..4cc1e3a36ec 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -119,8 +119,8 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_path.dentry->d_inode->i_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs, hfsplus_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 hfsplus_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
@@ -195,11 +195,13 @@ static struct dentry *hfsplus_file_lookup(struct inode *dir,
 	hip->flags = 0;
 	set_bit(HFSPLUS_I_RSRC, &hip->flags);
 
-	hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
-	err = hfsplus_find_cat(sb, dir->i_ino, &fd);
-	if (!err)
-		err = hfsplus_cat_read_inode(inode, &fd);
-	hfs_find_exit(&fd);
+	err = hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd);
+	if (!err) {
+		err = hfsplus_find_cat(sb, dir->i_ino, &fd);
+		if (!err)
+			err = hfsplus_cat_read_inode(inode, &fd);
+		hfs_find_exit(&fd);
+	}
 	if (err) {
 		iput(inode);
 		return ERR_PTR(err);
@@ -296,6 +298,8 @@ static int hfsplus_setattr(struct dentry *dentry, struct iattr *attr)
 
 	if ((attr->ia_valid & ATTR_SIZE) &&
 	    attr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		error = vmtruncate(inode, attr->ia_size);
 		if (error)
 			return error;
@@ -306,13 +310,19 @@ static int hfsplus_setattr(struct dentry *dentry, struct iattr *attr)
 	return 0;
 }
 
-int hfsplus_file_fsync(struct file *file, int datasync)
+int hfsplus_file_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct hfsplus_inode_info *hip = HFSPLUS_I(inode);
 	struct hfsplus_sb_info *sbi = HFSPLUS_SB(inode->i_sb);
 	int error = 0, error2;
 
+	error = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (error)
+		return error;
+	mutex_lock(&inode->i_mutex);
+
 	/*
 	 * Sync inode metadata into the catalog and extent trees.
 	 */
@@ -340,6 +350,8 @@ int hfsplus_file_fsync(struct file *file, int datasync)
 	if (!test_bit(HFSPLUS_SB_NOBARRIER, &sbi->flags))
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
 
+	mutex_unlock(&inode->i_mutex);
+
 	return error;
 }
 
diff --git a/fs/hfsplus/part_tbl.c b/fs/hfsplus/part_tbl.c
index 40ad88c12c6..eb355d81e27 100644
--- a/fs/hfsplus/part_tbl.c
+++ b/fs/hfsplus/part_tbl.c
@@ -88,11 +88,12 @@ static int hfs_parse_old_pmap(struct super_block *sb, struct old_pmap *pm,
 	return -ENOENT;
 }
 
-static int hfs_parse_new_pmap(struct super_block *sb, struct new_pmap *pm,
-		sector_t *part_start, sector_t *part_size)
+static int hfs_parse_new_pmap(struct super_block *sb, void *buf,
+		struct new_pmap *pm, sector_t *part_start, sector_t *part_size)
 {
 	struct hfsplus_sb_info *sbi = HFSPLUS_SB(sb);
 	int size = be32_to_cpu(pm->pmMapBlkCnt);
+	int buf_size = hfsplus_min_io_size(sb);
 	int res;
 	int i = 0;
 
@@ -107,11 +108,14 @@ static int hfs_parse_new_pmap(struct super_block *sb, struct new_pmap *pm,
 		if (++i >= size)
 			return -ENOENT;
 
-		res = hfsplus_submit_bio(sb->s_bdev,
-					 *part_start + HFS_PMAP_BLK + i,
-					 pm, READ);
-		if (res)
-			return res;
+		pm = (struct new_pmap *)((u8 *)pm + HFSPLUS_SECTOR_SIZE);
+		if ((u8 *)pm - (u8 *)buf >= buf_size) {
+			res = hfsplus_submit_bio(sb,
+						 *part_start + HFS_PMAP_BLK + i,
+						 buf, (void **)&pm, READ);
+			if (res)
+				return res;
+		}
 	} while (pm->pmSig == cpu_to_be16(HFS_NEW_PMAP_MAGIC));
 
 	return -ENOENT;
@@ -124,15 +128,15 @@ static int hfs_parse_new_pmap(struct super_block *sb, struct new_pmap *pm,
 int hfs_part_find(struct super_block *sb,
 		sector_t *part_start, sector_t *part_size)
 {
-	void *data;
+	void *buf, *data;
 	int res;
 
-	data = kmalloc(HFSPLUS_SECTOR_SIZE, GFP_KERNEL);
-	if (!data)
+	buf = kmalloc(hfsplus_min_io_size(sb), GFP_KERNEL);
+	if (!buf)
 		return -ENOMEM;
 
-	res = hfsplus_submit_bio(sb->s_bdev, *part_start + HFS_PMAP_BLK,
-				 data, READ);
+	res = hfsplus_submit_bio(sb, *part_start + HFS_PMAP_BLK,
+				 buf, &data, READ);
 	if (res)
 		goto out;
 
@@ -141,13 +145,13 @@ int hfs_part_find(struct super_block *sb,
 		res = hfs_parse_old_pmap(sb, data, part_start, part_size);
 		break;
 	case HFS_NEW_PMAP_MAGIC:
-		res = hfs_parse_new_pmap(sb, data, part_start, part_size);
+		res = hfs_parse_new_pmap(sb, buf, data, part_start, part_size);
 		break;
 	default:
 		res = -ENOENT;
 		break;
 	}
 out:
-	kfree(data);
+	kfree(buf);
 	return res;
 }
diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c
index b49b55584c8..c106ca22e81 100644
--- a/fs/hfsplus/super.c
+++ b/fs/hfsplus/super.c
@@ -73,11 +73,13 @@ struct inode *hfsplus_iget(struct super_block *sb, unsigned long ino)
 
 	if (inode->i_ino >= HFSPLUS_FIRSTUSER_CNID ||
 	    inode->i_ino == HFSPLUS_ROOT_CNID) {
-		hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &fd);
-		err = hfsplus_find_cat(inode->i_sb, inode->i_ino, &fd);
-		if (!err)
-			err = hfsplus_cat_read_inode(inode, &fd);
-		hfs_find_exit(&fd);
+		err = hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &fd);
+		if (!err) {
+			err = hfsplus_find_cat(inode->i_sb, inode->i_ino, &fd);
+			if (!err)
+				err = hfsplus_cat_read_inode(inode, &fd);
+			hfs_find_exit(&fd);
+		}
 	} else {
 		err = hfsplus_system_read_inode(inode);
 	}
@@ -133,9 +135,13 @@ static int hfsplus_system_write_inode(struct inode *inode)
 static int hfsplus_write_inode(struct inode *inode,
 		struct writeback_control *wbc)
 {
+	int err;
+
 	dprint(DBG_INODE, "hfsplus_write_inode: %lu\n", inode->i_ino);
 
-	hfsplus_ext_write_extent(inode);
+	err = hfsplus_ext_write_extent(inode);
+	if (err)
+		return err;
 
 	if (inode->i_ino >= HFSPLUS_FIRSTUSER_CNID ||
 	    inode->i_ino == HFSPLUS_ROOT_CNID)
@@ -197,17 +203,17 @@ int hfsplus_sync_fs(struct super_block *sb, int wait)
 		write_backup = 1;
 	}
 
-	error2 = hfsplus_submit_bio(sb->s_bdev,
+	error2 = hfsplus_submit_bio(sb,
 				   sbi->part_start + HFSPLUS_VOLHEAD_SECTOR,
-				   sbi->s_vhdr, WRITE_SYNC);
+				   sbi->s_vhdr_buf, NULL, WRITE_SYNC);
 	if (!error)
 		error = error2;
 	if (!write_backup)
 		goto out;
 
-	error2 = hfsplus_submit_bio(sb->s_bdev,
+	error2 = hfsplus_submit_bio(sb,
 				  sbi->part_start + sbi->sect_count - 2,
-				  sbi->s_backup_vhdr, WRITE_SYNC);
+				  sbi->s_backup_vhdr_buf, NULL, WRITE_SYNC);
 	if (!error)
 		error2 = error;
 out:
@@ -251,8 +257,8 @@ static void hfsplus_put_super(struct super_block *sb)
 	hfs_btree_close(sbi->ext_tree);
 	iput(sbi->alloc_file);
 	iput(sbi->hidden_dir);
-	kfree(sbi->s_vhdr);
-	kfree(sbi->s_backup_vhdr);
+	kfree(sbi->s_vhdr_buf);
+	kfree(sbi->s_backup_vhdr_buf);
 	unload_nls(sbi->nls);
 	kfree(sb->s_fs_info);
 	sb->s_fs_info = NULL;
@@ -393,6 +399,13 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent)
 	if (!sbi->rsrc_clump_blocks)
 		sbi->rsrc_clump_blocks = 1;
 
+	err = generic_check_addressable(sbi->alloc_blksz_shift,
+					sbi->total_blocks);
+	if (err) {
+		printk(KERN_ERR "hfs: filesystem size too large.\n");
+		goto out_free_vhdr;
+	}
+
 	/* Set up operations so we can load metadata */
 	sb->s_op = &hfsplus_sops;
 	sb->s_maxbytes = MAX_LFS_FILESIZE;
@@ -417,6 +430,8 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent)
 		sb->s_flags |= MS_RDONLY;
 	}
 
+	err = -EINVAL;
+
 	/* Load metadata objects (B*Trees) */
 	sbi->ext_tree = hfs_btree_open(sb, HFSPLUS_EXT_CNID);
 	if (!sbi->ext_tree) {
@@ -447,7 +462,9 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent)
 
 	str.len = sizeof(HFSP_HIDDENDIR_NAME) - 1;
 	str.name = HFSP_HIDDENDIR_NAME;
-	hfs_find_init(sbi->cat_tree, &fd);
+	err = hfs_find_init(sbi->cat_tree, &fd);
+	if (err)
+		goto out_put_root;
 	hfsplus_cat_build_key(sb, fd.search_key, HFSPLUS_ROOT_CNID, &str);
 	if (!hfs_brec_read(&fd, &entry, sizeof(entry))) {
 		hfs_find_exit(&fd);
@@ -500,7 +517,7 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent)
 out_put_hidden_dir:
 	iput(sbi->hidden_dir);
 out_put_root:
-	iput(sbi->alloc_file);
+	iput(root);
 out_put_alloc_file:
 	iput(sbi->alloc_file);
 out_close_cat_tree:
diff --git a/fs/hfsplus/unicode.c b/fs/hfsplus/unicode.c
index a3f0bfcc881..a32998f29f0 100644
--- a/fs/hfsplus/unicode.c
+++ b/fs/hfsplus/unicode.c
@@ -142,7 +142,11 @@ int hfsplus_uni2asc(struct super_block *sb,
 		/* search for single decomposed char */
 		if (likely(compose))
 			ce1 = hfsplus_compose_lookup(hfsplus_compose_table, c0);
-		if (ce1 && (cc = ce1[0])) {
+		if (ce1)
+			cc = ce1[0];
+		else
+			cc = 0;
+		if (cc) {
 			/* start of a possibly decomposed Hangul char */
 			if (cc != 0xffff)
 				goto done;
@@ -209,7 +213,8 @@ int hfsplus_uni2asc(struct super_block *sb,
 				i++;
 				ce2 = ce1;
 			}
-			if ((cc = ce2[0])) {
+			cc = ce2[0];
+			if (cc) {
 				ip += i;
 				ustrlen -= i;
 				goto done;
@@ -301,7 +306,11 @@ int hfsplus_asc2uni(struct super_block *sb, struct hfsplus_unistr *ustr,
 	while (outlen < HFSPLUS_MAX_STRLEN && len > 0) {
 		size = asc2unichar(sb, astr, len, &c);
 
-		if (decompose && (dstr = decompose_unichar(c, &dsize))) {
+		if (decompose)
+			dstr = decompose_unichar(c, &dsize);
+		else
+			dstr = NULL;
+		if (dstr) {
 			if (outlen + dsize > HFSPLUS_MAX_STRLEN)
 				break;
 			do {
@@ -346,15 +355,23 @@ int hfsplus_hash_dentry(const struct dentry *dentry, const struct inode *inode,
 		astr += size;
 		len -= size;
 
-		if (decompose && (dstr = decompose_unichar(c, &dsize))) {
+		if (decompose)
+			dstr = decompose_unichar(c, &dsize);
+		else
+			dstr = NULL;
+		if (dstr) {
 			do {
 				c2 = *dstr++;
-				if (!casefold || (c2 = case_fold(c2)))
+				if (casefold)
+					c2 = case_fold(c2);
+				if (!casefold || c2)
 					hash = partial_name_hash(c2, hash);
 			} while (--dsize > 0);
 		} else {
 			c2 = c;
-			if (!casefold || (c2 = case_fold(c2)))
+			if (casefold)
+				c2 = case_fold(c2);
+			if (!casefold || c2)
 				hash = partial_name_hash(c2, hash);
 		}
 	}
@@ -422,12 +439,14 @@ int hfsplus_compare_dentry(const struct dentry *parent,
 		c1 = *dstr1;
 		c2 = *dstr2;
 		if (casefold) {
-			if  (!(c1 = case_fold(c1))) {
+			c1 = case_fold(c1);
+			if (!c1) {
 				dstr1++;
 				dsize1--;
 				continue;
 			}
-			if (!(c2 = case_fold(c2))) {
+			c2 = case_fold(c2);
+			if (!c2) {
 				dstr2++;
 				dsize2--;
 				continue;
diff --git a/fs/hfsplus/wrapper.c b/fs/hfsplus/wrapper.c
index 3031d81f5f0..10e515a0d45 100644
--- a/fs/hfsplus/wrapper.c
+++ b/fs/hfsplus/wrapper.c
@@ -31,31 +31,77 @@ static void hfsplus_end_io_sync(struct bio *bio, int err)
 	complete(bio->bi_private);
 }
 
-int hfsplus_submit_bio(struct block_device *bdev, sector_t sector,
-		void *data, int rw)
+/*
+ * hfsplus_submit_bio - Perfrom block I/O
+ * @sb: super block of volume for I/O
+ * @sector: block to read or write, for blocks of HFSPLUS_SECTOR_SIZE bytes
+ * @buf: buffer for I/O
+ * @data: output pointer for location of requested data
+ * @rw: direction of I/O
+ *
+ * The unit of I/O is hfsplus_min_io_size(sb), which may be bigger than
+ * HFSPLUS_SECTOR_SIZE, and @buf must be sized accordingly. On reads
+ * @data will return a pointer to the start of the requested sector,
+ * which may not be the same location as @buf.
+ *
+ * If @sector is not aligned to the bdev logical block size it will
+ * be rounded down. For writes this means that @buf should contain data
+ * that starts at the rounded-down address. As long as the data was
+ * read using hfsplus_submit_bio() and the same buffer is used things
+ * will work correctly.
+ */
+int hfsplus_submit_bio(struct super_block *sb, sector_t sector,
+		void *buf, void **data, int rw)
 {
 	DECLARE_COMPLETION_ONSTACK(wait);
 	struct bio *bio;
+	int ret = 0;
+	unsigned int io_size;
+	loff_t start;
+	int offset;
+
+	/*
+	 * Align sector to hardware sector size and find offset. We
+	 * assume that io_size is a power of two, which _should_
+	 * be true.
+	 */
+	io_size = hfsplus_min_io_size(sb);
+	start = (loff_t)sector << HFSPLUS_SECTOR_SHIFT;
+	offset = start & (io_size - 1);
+	sector &= ~((io_size >> HFSPLUS_SECTOR_SHIFT) - 1);
 
 	bio = bio_alloc(GFP_NOIO, 1);
 	bio->bi_sector = sector;
-	bio->bi_bdev = bdev;
+	bio->bi_bdev = sb->s_bdev;
 	bio->bi_end_io = hfsplus_end_io_sync;
 	bio->bi_private = &wait;
 
-	/*
-	 * We always submit one sector at a time, so bio_add_page must not fail.
-	 */
-	if (bio_add_page(bio, virt_to_page(data), HFSPLUS_SECTOR_SIZE,
-			 offset_in_page(data)) != HFSPLUS_SECTOR_SIZE)
-		BUG();
+	if (!(rw & WRITE) && data)
+		*data = (u8 *)buf + offset;
+
+	while (io_size > 0) {
+		unsigned int page_offset = offset_in_page(buf);
+		unsigned int len = min_t(unsigned int, PAGE_SIZE - page_offset,
+					 io_size);
+
+		ret = bio_add_page(bio, virt_to_page(buf), len, page_offset);
+		if (ret != len) {
+			ret = -EIO;
+			goto out;
+		}
+		io_size -= len;
+		buf = (u8 *)buf + len;
+	}
 
 	submit_bio(rw, bio);
 	wait_for_completion(&wait);
 
 	if (!bio_flagged(bio, BIO_UPTODATE))
-		return -EIO;
-	return 0;
+		ret = -EIO;
+
+out:
+	bio_put(bio);
+	return ret < 0 ? ret : 0;
 }
 
 static int hfsplus_read_mdb(void *bufptr, struct hfsplus_wd *wd)
@@ -138,23 +184,19 @@ int hfsplus_read_wrapper(struct super_block *sb)
 
 	if (hfsplus_get_last_session(sb, &part_start, &part_size))
 		goto out;
-	if ((u64)part_start + part_size > 0x100000000ULL) {
-		pr_err("hfs: volumes larger than 2TB are not supported yet\n");
-		goto out;
-	}
 
 	error = -ENOMEM;
-	sbi->s_vhdr = kmalloc(HFSPLUS_SECTOR_SIZE, GFP_KERNEL);
-	if (!sbi->s_vhdr)
+	sbi->s_vhdr_buf = kmalloc(hfsplus_min_io_size(sb), GFP_KERNEL);
+	if (!sbi->s_vhdr_buf)
 		goto out;
-	sbi->s_backup_vhdr = kmalloc(HFSPLUS_SECTOR_SIZE, GFP_KERNEL);
-	if (!sbi->s_backup_vhdr)
+	sbi->s_backup_vhdr_buf = kmalloc(hfsplus_min_io_size(sb), GFP_KERNEL);
+	if (!sbi->s_backup_vhdr_buf)
 		goto out_free_vhdr;
 
 reread:
-	error = hfsplus_submit_bio(sb->s_bdev,
-				   part_start + HFSPLUS_VOLHEAD_SECTOR,
-				   sbi->s_vhdr, READ);
+	error = hfsplus_submit_bio(sb, part_start + HFSPLUS_VOLHEAD_SECTOR,
+				   sbi->s_vhdr_buf, (void **)&sbi->s_vhdr,
+				   READ);
 	if (error)
 		goto out_free_backup_vhdr;
 
@@ -169,8 +211,9 @@ reread:
 		if (!hfsplus_read_mdb(sbi->s_vhdr, &wd))
 			goto out_free_backup_vhdr;
 		wd.ablk_size >>= HFSPLUS_SECTOR_SHIFT;
-		part_start += wd.ablk_start + wd.embed_start * wd.ablk_size;
-		part_size = wd.embed_count * wd.ablk_size;
+		part_start += (sector_t)wd.ablk_start +
+			       (sector_t)wd.embed_start * wd.ablk_size;
+		part_size = (sector_t)wd.embed_count * wd.ablk_size;
 		goto reread;
 	default:
 		/*
@@ -183,9 +226,9 @@ reread:
 		goto reread;
 	}
 
-	error = hfsplus_submit_bio(sb->s_bdev,
-				   part_start + part_size - 2,
-				   sbi->s_backup_vhdr, READ);
+	error = hfsplus_submit_bio(sb, part_start + part_size - 2,
+				   sbi->s_backup_vhdr_buf,
+				   (void **)&sbi->s_backup_vhdr, READ);
 	if (error)
 		goto out_free_backup_vhdr;
 
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index 2638c834ed2..0d22afdd461 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -362,9 +362,20 @@ retry:
 	return 0;
 }
 
-int hostfs_fsync(struct file *file, int datasync)
+int hostfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
-	return fsync_file(HOSTFS_I(file->f_mapping->host)->fd, datasync);
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+
+	mutex_lock(&inode->i_mutex);
+	ret = fsync_file(HOSTFS_I(inode)->fd, datasync);
+	mutex_unlock(&inode->i_mutex);
+
+	return ret;
 }
 
 static const struct file_operations hostfs_file_fops = {
@@ -748,12 +759,12 @@ int hostfs_rename(struct inode *from_ino, struct dentry *from,
 	return err;
 }
 
-int hostfs_permission(struct inode *ino, int desired, unsigned int flags)
+int hostfs_permission(struct inode *ino, int desired)
 {
 	char *name;
 	int r = 0, w = 0, x = 0, err;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (desired & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	if (desired & MAY_READ) r = 1;
@@ -770,7 +781,7 @@ int hostfs_permission(struct inode *ino, int desired, unsigned int flags)
 		err = access_file(name, r, w, x);
 	__putname(name);
 	if (!err)
-		err = generic_permission(ino, desired, flags, NULL);
+		err = generic_permission(ino, desired);
 	return err;
 }
 
diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c
index f46ae025bfb..96a8ed91ced 100644
--- a/fs/hpfs/dir.c
+++ b/fs/hpfs/dir.c
@@ -29,6 +29,10 @@ static loff_t hpfs_dir_lseek(struct file *filp, loff_t off, int whence)
 	struct hpfs_inode_info *hpfs_inode = hpfs_i(i);
 	struct super_block *s = i->i_sb;
 
+	/* Somebody else will have to figure out what to do here */
+	if (whence == SEEK_DATA || whence == SEEK_HOLE)
+		return -EINVAL;
+
 	hpfs_lock(s);
 
 	/*printk("dir lseek\n");*/
diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c
index 89c500ee521..89d2a5803ae 100644
--- a/fs/hpfs/file.c
+++ b/fs/hpfs/file.c
@@ -18,9 +18,14 @@ static int hpfs_file_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-int hpfs_file_fsync(struct file *file, int datasync)
+int hpfs_file_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	ret = filemap_write_and_wait_range(file->f_mapping, start, end);
+	if (ret)
+		return ret;
 	return sync_blockdev(inode->i_sb->s_bdev);
 }
 
diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h
index dd552f862c8..331b5e234ef 100644
--- a/fs/hpfs/hpfs_fn.h
+++ b/fs/hpfs/hpfs_fn.h
@@ -258,7 +258,7 @@ void hpfs_set_ea(struct inode *, struct fnode *, const char *,
 
 /* file.c */
 
-int hpfs_file_fsync(struct file *, int);
+int hpfs_file_fsync(struct file *, loff_t, loff_t, int);
 extern const struct file_operations hpfs_file_ops;
 extern const struct inode_operations hpfs_file_iops;
 extern const struct address_space_operations hpfs_aops;
diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c
index acf95dab2aa..2df69e2f07c 100644
--- a/fs/hpfs/namei.c
+++ b/fs/hpfs/namei.c
@@ -398,7 +398,7 @@ again:
 			hpfs_unlock(dir->i_sb);
 			return -ENOSPC;
 		}
-		if (generic_permission(inode, MAY_WRITE, 0, NULL) ||
+		if (generic_permission(inode, MAY_WRITE) ||
 		    !S_ISREG(inode->i_mode) ||
 		    get_write_access(inode)) {
 			d_rehash(dentry);
diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c
index 87ed48e0343..8635be5ffd9 100644
--- a/fs/hppfs/hppfs.c
+++ b/fs/hppfs/hppfs.c
@@ -139,7 +139,8 @@ static int file_removed(struct dentry *dentry, const char *file)
 static struct dentry *hppfs_lookup(struct inode *ino, struct dentry *dentry,
 				   struct nameidata *nd)
 {
-	struct dentry *proc_dentry, *new, *parent;
+	struct dentry *proc_dentry, *parent;
+	struct qstr *name = &dentry->d_name;
 	struct inode *inode;
 	int err, deleted;
 
@@ -149,23 +150,9 @@ static struct dentry *hppfs_lookup(struct inode *ino, struct dentry *dentry,
 	else if (deleted)
 		return ERR_PTR(-ENOENT);
 
-	err = -ENOMEM;
 	parent = HPPFS_I(ino)->proc_dentry;
 	mutex_lock(&parent->d_inode->i_mutex);
-	proc_dentry = d_lookup(parent, &dentry->d_name);
-	if (proc_dentry == NULL) {
-		proc_dentry = d_alloc(parent, &dentry->d_name);
-		if (proc_dentry == NULL) {
-			mutex_unlock(&parent->d_inode->i_mutex);
-			goto out;
-		}
-		new = (*parent->d_inode->i_op->lookup)(parent->d_inode,
-						       proc_dentry, NULL);
-		if (new) {
-			dput(proc_dentry);
-			proc_dentry = new;
-		}
-	}
+	proc_dentry = lookup_one_len(name->name, parent, name->len);
 	mutex_unlock(&parent->d_inode->i_mutex);
 
 	if (IS_ERR(proc_dentry))
@@ -174,13 +161,11 @@ static struct dentry *hppfs_lookup(struct inode *ino, struct dentry *dentry,
 	err = -ENOMEM;
 	inode = get_inode(ino->i_sb, proc_dentry);
 	if (!inode)
-		goto out_dput;
+		goto out;
 
  	d_add(dentry, inode);
 	return NULL;
 
- out_dput:
-	dput(proc_dentry);
  out:
 	return ERR_PTR(err);
 }
@@ -588,9 +573,10 @@ static int hppfs_readdir(struct file *file, void *ent, filldir_t filldir)
 	return err;
 }
 
-static int hppfs_fsync(struct file *file, int datasync)
+static int hppfs_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync)
 {
-	return 0;
+	return filemap_write_and_wait_range(file->f_mapping, start, end);
 }
 
 static const struct file_operations hppfs_dir_fops = {
@@ -690,8 +676,10 @@ static struct inode *get_inode(struct super_block *sb, struct dentry *dentry)
 	struct inode *proc_ino = dentry->d_inode;
 	struct inode *inode = new_inode(sb);
 
-	if (!inode)
+	if (!inode) {
+		dput(dentry);
 		return ERR_PTR(-ENOMEM);
+	}
 
 	if (S_ISDIR(dentry->d_inode->i_mode)) {
 		inode->i_op = &hppfs_dir_iops;
@@ -704,7 +692,7 @@ static struct inode *get_inode(struct super_block *sb, struct dentry *dentry)
 		inode->i_fop = &hppfs_file_fops;
 	}
 
-	HPPFS_I(inode)->proc_dentry = dget(dentry);
+	HPPFS_I(inode)->proc_dentry = dentry;
 
 	inode->i_uid = proc_ino->i_uid;
 	inode->i_gid = proc_ino->i_gid;
@@ -737,7 +725,7 @@ static int hppfs_fill_super(struct super_block *sb, void *d, int silent)
 	sb->s_fs_info = proc_mnt;
 
 	err = -ENOMEM;
-	root_inode = get_inode(sb, proc_mnt->mnt_sb->s_root);
+	root_inode = get_inode(sb, dget(proc_mnt->mnt_sb->s_root));
 	if (!root_inode)
 		goto out_mntput;
 
diff --git a/fs/inode.c b/fs/inode.c
index 0f7e88a7803..96c77b81167 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -33,8 +33,8 @@
  *
  * inode->i_lock protects:
  *   inode->i_state, inode->i_hash, __iget()
- * inode_lru_lock protects:
- *   inode_lru, inode->i_lru
+ * inode->i_sb->s_inode_lru_lock protects:
+ *   inode->i_sb->s_inode_lru, inode->i_lru
  * inode_sb_list_lock protects:
  *   sb->s_inodes, inode->i_sb_list
  * inode_wb_list_lock protects:
@@ -46,7 +46,7 @@
  *
  * inode_sb_list_lock
  *   inode->i_lock
- *     inode_lru_lock
+ *     inode->i_sb->s_inode_lru_lock
  *
  * inode_wb_list_lock
  *   inode->i_lock
@@ -64,24 +64,10 @@ static unsigned int i_hash_shift __read_mostly;
 static struct hlist_head *inode_hashtable __read_mostly;
 static __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_hash_lock);
 
-static LIST_HEAD(inode_lru);
-static DEFINE_SPINLOCK(inode_lru_lock);
-
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_sb_list_lock);
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_wb_list_lock);
 
 /*
- * iprune_sem provides exclusion between the icache shrinking and the
- * umount path.
- *
- * We don't actually need it to protect anything in the umount path,
- * but only need to cycle through it to make sure any inode that
- * prune_icache took off the LRU list has been fully torn down by the
- * time we are past evict_inodes.
- */
-static DECLARE_RWSEM(iprune_sem);
-
-/*
  * Empty aops. Can be used for the cases where the user does not
  * define any of the address_space operations.
  */
@@ -95,6 +81,7 @@ EXPORT_SYMBOL(empty_aops);
 struct inodes_stat_t inodes_stat;
 
 static DEFINE_PER_CPU(unsigned int, nr_inodes);
+static DEFINE_PER_CPU(unsigned int, nr_unused);
 
 static struct kmem_cache *inode_cachep __read_mostly;
 
@@ -109,7 +96,11 @@ static int get_nr_inodes(void)
 
 static inline int get_nr_inodes_unused(void)
 {
-	return inodes_stat.nr_unused;
+	int i;
+	int sum = 0;
+	for_each_possible_cpu(i)
+		sum += per_cpu(nr_unused, i);
+	return sum < 0 ? 0 : sum;
 }
 
 int get_nr_dirty_inodes(void)
@@ -127,6 +118,7 @@ int proc_nr_inodes(ctl_table *table, int write,
 		   void __user *buffer, size_t *lenp, loff_t *ppos)
 {
 	inodes_stat.nr_inodes = get_nr_inodes();
+	inodes_stat.nr_unused = get_nr_inodes_unused();
 	return proc_dointvec(table, write, buffer, lenp, ppos);
 }
 #endif
@@ -176,8 +168,7 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
 	mutex_init(&inode->i_mutex);
 	lockdep_set_class(&inode->i_mutex, &sb->s_type->i_mutex_key);
 
-	init_rwsem(&inode->i_alloc_sem);
-	lockdep_set_class(&inode->i_alloc_sem, &sb->s_type->i_alloc_sem_key);
+	atomic_set(&inode->i_dio_count, 0);
 
 	mapping->a_ops = &empty_aops;
 	mapping->host = inode;
@@ -337,22 +328,24 @@ EXPORT_SYMBOL(ihold);
 
 static void inode_lru_list_add(struct inode *inode)
 {
-	spin_lock(&inode_lru_lock);
+	spin_lock(&inode->i_sb->s_inode_lru_lock);
 	if (list_empty(&inode->i_lru)) {
-		list_add(&inode->i_lru, &inode_lru);
-		inodes_stat.nr_unused++;
+		list_add(&inode->i_lru, &inode->i_sb->s_inode_lru);
+		inode->i_sb->s_nr_inodes_unused++;
+		this_cpu_inc(nr_unused);
 	}
-	spin_unlock(&inode_lru_lock);
+	spin_unlock(&inode->i_sb->s_inode_lru_lock);
 }
 
 static void inode_lru_list_del(struct inode *inode)
 {
-	spin_lock(&inode_lru_lock);
+	spin_lock(&inode->i_sb->s_inode_lru_lock);
 	if (!list_empty(&inode->i_lru)) {
 		list_del_init(&inode->i_lru);
-		inodes_stat.nr_unused--;
+		inode->i_sb->s_nr_inodes_unused--;
+		this_cpu_dec(nr_unused);
 	}
-	spin_unlock(&inode_lru_lock);
+	spin_unlock(&inode->i_sb->s_inode_lru_lock);
 }
 
 /**
@@ -423,7 +416,14 @@ EXPORT_SYMBOL(remove_inode_hash);
 void end_writeback(struct inode *inode)
 {
 	might_sleep();
+	/*
+	 * We have to cycle tree_lock here because reclaim can be still in the
+	 * process of removing the last page (in __delete_from_page_cache())
+	 * and we must not free mapping under it.
+	 */
+	spin_lock_irq(&inode->i_data.tree_lock);
 	BUG_ON(inode->i_data.nrpages);
+	spin_unlock_irq(&inode->i_data.tree_lock);
 	BUG_ON(!list_empty(&inode->i_data.private_list));
 	BUG_ON(!(inode->i_state & I_FREEING));
 	BUG_ON(inode->i_state & I_CLEAR);
@@ -530,14 +530,6 @@ void evict_inodes(struct super_block *sb)
 	spin_unlock(&inode_sb_list_lock);
 
 	dispose_list(&dispose);
-
-	/*
-	 * Cycle through iprune_sem to make sure any inode that prune_icache
-	 * moved off the list before we took the lock has been fully torn
-	 * down.
-	 */
-	down_write(&iprune_sem);
-	up_write(&iprune_sem);
 }
 
 /**
@@ -600,8 +592,10 @@ static int can_unuse(struct inode *inode)
 }
 
 /*
- * Scan `goal' inodes on the unused list for freeable ones. They are moved to a
- * temporary list and then are freed outside inode_lru_lock by dispose_list().
+ * Walk the superblock inode LRU for freeable inodes and attempt to free them.
+ * This is called from the superblock shrinker function with a number of inodes
+ * to trim from the LRU. Inodes to be freed are moved to a temporary list and
+ * then are freed outside inode_lock by dispose_list().
  *
  * Any inodes which are pinned purely because of attached pagecache have their
  * pagecache removed.  If the inode has metadata buffers attached to
@@ -615,29 +609,28 @@ static int can_unuse(struct inode *inode)
  * LRU does not have strict ordering. Hence we don't want to reclaim inodes
  * with this flag set because they are the inodes that are out of order.
  */
-static void prune_icache(int nr_to_scan)
+void prune_icache_sb(struct super_block *sb, int nr_to_scan)
 {
 	LIST_HEAD(freeable);
 	int nr_scanned;
 	unsigned long reap = 0;
 
-	down_read(&iprune_sem);
-	spin_lock(&inode_lru_lock);
-	for (nr_scanned = 0; nr_scanned < nr_to_scan; nr_scanned++) {
+	spin_lock(&sb->s_inode_lru_lock);
+	for (nr_scanned = nr_to_scan; nr_scanned >= 0; nr_scanned--) {
 		struct inode *inode;
 
-		if (list_empty(&inode_lru))
+		if (list_empty(&sb->s_inode_lru))
 			break;
 
-		inode = list_entry(inode_lru.prev, struct inode, i_lru);
+		inode = list_entry(sb->s_inode_lru.prev, struct inode, i_lru);
 
 		/*
-		 * we are inverting the inode_lru_lock/inode->i_lock here,
+		 * we are inverting the sb->s_inode_lru_lock/inode->i_lock here,
 		 * so use a trylock. If we fail to get the lock, just move the
 		 * inode to the back of the list so we don't spin on it.
 		 */
 		if (!spin_trylock(&inode->i_lock)) {
-			list_move(&inode->i_lru, &inode_lru);
+			list_move(&inode->i_lru, &sb->s_inode_lru);
 			continue;
 		}
 
@@ -649,28 +642,29 @@ static void prune_icache(int nr_to_scan)
 		    (inode->i_state & ~I_REFERENCED)) {
 			list_del_init(&inode->i_lru);
 			spin_unlock(&inode->i_lock);
-			inodes_stat.nr_unused--;
+			sb->s_nr_inodes_unused--;
+			this_cpu_dec(nr_unused);
 			continue;
 		}
 
 		/* recently referenced inodes get one more pass */
 		if (inode->i_state & I_REFERENCED) {
 			inode->i_state &= ~I_REFERENCED;
-			list_move(&inode->i_lru, &inode_lru);
+			list_move(&inode->i_lru, &sb->s_inode_lru);
 			spin_unlock(&inode->i_lock);
 			continue;
 		}
 		if (inode_has_buffers(inode) || inode->i_data.nrpages) {
 			__iget(inode);
 			spin_unlock(&inode->i_lock);
-			spin_unlock(&inode_lru_lock);
+			spin_unlock(&sb->s_inode_lru_lock);
 			if (remove_inode_buffers(inode))
 				reap += invalidate_mapping_pages(&inode->i_data,
 								0, -1);
 			iput(inode);
-			spin_lock(&inode_lru_lock);
+			spin_lock(&sb->s_inode_lru_lock);
 
-			if (inode != list_entry(inode_lru.next,
+			if (inode != list_entry(sb->s_inode_lru.next,
 						struct inode, i_lru))
 				continue;	/* wrong inode or list_empty */
 			/* avoid lock inversions with trylock */
@@ -686,51 +680,18 @@ static void prune_icache(int nr_to_scan)
 		spin_unlock(&inode->i_lock);
 
 		list_move(&inode->i_lru, &freeable);
-		inodes_stat.nr_unused--;
+		sb->s_nr_inodes_unused--;
+		this_cpu_dec(nr_unused);
 	}
 	if (current_is_kswapd())
 		__count_vm_events(KSWAPD_INODESTEAL, reap);
 	else
 		__count_vm_events(PGINODESTEAL, reap);
-	spin_unlock(&inode_lru_lock);
+	spin_unlock(&sb->s_inode_lru_lock);
 
 	dispose_list(&freeable);
-	up_read(&iprune_sem);
 }
 
-/*
- * shrink_icache_memory() will attempt to reclaim some unused inodes.  Here,
- * "unused" means that no dentries are referring to the inodes: the files are
- * not open and the dcache references to those inodes have already been
- * reclaimed.
- *
- * This function is passed the number of inodes to scan, and it returns the
- * total number of remaining possibly-reclaimable inodes.
- */
-static int shrink_icache_memory(struct shrinker *shrink,
-				struct shrink_control *sc)
-{
-	int nr = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
-
-	if (nr) {
-		/*
-		 * Nasty deadlock avoidance.  We may hold various FS locks,
-		 * and we don't want to recurse into the FS that called us
-		 * in clear_inode() and friends..
-		 */
-		if (!(gfp_mask & __GFP_FS))
-			return -1;
-		prune_icache(nr);
-	}
-	return (get_nr_inodes_unused() / 100) * sysctl_vfs_cache_pressure;
-}
-
-static struct shrinker icache_shrinker = {
-	.shrink = shrink_icache_memory,
-	.seeks = DEFAULT_SEEKS,
-};
-
 static void __wait_on_freeing_inode(struct inode *inode);
 /*
  * Called with the inode lock held.
@@ -1324,7 +1285,7 @@ static void iput_final(struct inode *inode)
 
 	WARN_ON(inode->i_state & I_NEW);
 
-	if (op && op->drop_inode)
+	if (op->drop_inode)
 		drop = op->drop_inode(inode);
 	else
 		drop = generic_drop_inode(inode);
@@ -1610,7 +1571,6 @@ void __init inode_init(void)
 					 (SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|
 					 SLAB_MEM_SPREAD),
 					 init_once);
-	register_shrinker(&icache_shrinker);
 
 	/* Hash may have been set up in inode_init_early */
 	if (!hashdist)
diff --git a/fs/internal.h b/fs/internal.h
index b29c46e4e32..fe327c20af8 100644
--- a/fs/internal.h
+++ b/fs/internal.h
@@ -97,6 +97,7 @@ extern struct file *get_empty_filp(void);
  * super.c
  */
 extern int do_remount_sb(struct super_block *, int, void *, int);
+extern bool grab_super_passive(struct super_block *sb);
 extern void __put_super(struct super_block *sb);
 extern void put_super(struct super_block *sb);
 extern struct dentry *mount_fs(struct file_system_type *,
@@ -135,3 +136,8 @@ extern void inode_wb_list_del(struct inode *inode);
 extern int get_nr_dirty_inodes(void);
 extern void evict_inodes(struct super_block *);
 extern int invalidate_inodes(struct super_block *, bool);
+
+/*
+ * dcache.c
+ */
+extern struct dentry *__d_alloc(struct super_block *, const struct qstr *);
diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c
index 0542b6eedf8..f20437c068a 100644
--- a/fs/isofs/dir.c
+++ b/fs/isofs/dir.c
@@ -254,19 +254,16 @@ static int isofs_readdir(struct file *filp,
 	char *tmpname;
 	struct iso_directory_record *tmpde;
 	struct inode *inode = filp->f_path.dentry->d_inode;
-	struct isofs_sb_info *sbi = ISOFS_SB(inode->i_sb);
 
 	tmpname = (char *)__get_free_page(GFP_KERNEL);
 	if (tmpname == NULL)
 		return -ENOMEM;
 
-	mutex_lock(&sbi->s_mutex);
 	tmpde = (struct iso_directory_record *) (tmpname+1024);
 
 	result = do_isofs_readdir(inode, filp, dirent, filldir, tmpname, tmpde);
 
 	free_page((unsigned long) tmpname);
-	mutex_unlock(&sbi->s_mutex);
 	return result;
 }
 
diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c
index 3db5ba4568f..a5d03672d04 100644
--- a/fs/isofs/inode.c
+++ b/fs/isofs/inode.c
@@ -863,7 +863,6 @@ root_found:
 	sbi->s_utf8 = opt.utf8;
 	sbi->s_nocompress = opt.nocompress;
 	sbi->s_overriderockperm = opt.overriderockperm;
-	mutex_init(&sbi->s_mutex);
 	/*
 	 * It would be incredibly stupid to allow people to mark every file
 	 * on the disk as suid, so we merely allow them to set the default
@@ -974,7 +973,7 @@ out_no_inode:
 out_no_read:
 	printk(KERN_WARNING "%s: bread failed, dev=%s, iso_blknum=%d, block=%d\n",
 		__func__, s->s_id, iso_blknum, block);
-	goto out_freesbi;
+	goto out_freebh;
 out_bad_zone_size:
 	printk(KERN_WARNING "ISOFS: Bad logical zone size %ld\n",
 		sbi->s_log_zone_size);
@@ -989,6 +988,7 @@ out_unknown_format:
 
 out_freebh:
 	brelse(bh);
+	brelse(pri_bh);
 out_freesbi:
 	kfree(opt.iocharset);
 	kfree(sbi);
diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h
index 2882dc089f8..7d33de84f52 100644
--- a/fs/isofs/isofs.h
+++ b/fs/isofs/isofs.h
@@ -55,7 +55,6 @@ struct isofs_sb_info {
 	gid_t s_gid;
 	uid_t s_uid;
 	struct nls_table *s_nls_iocharset; /* Native language support table */
-	struct mutex s_mutex; /* replaces BKL, please remove if possible */
 };
 
 #define ISOFS_INVALID_MODE ((mode_t) -1)
diff --git a/fs/isofs/namei.c b/fs/isofs/namei.c
index 4fb3e8074fd..1e2946f2a69 100644
--- a/fs/isofs/namei.c
+++ b/fs/isofs/namei.c
@@ -168,7 +168,6 @@ struct dentry *isofs_lookup(struct inode *dir, struct dentry *dentry, struct nam
 	int found;
 	unsigned long uninitialized_var(block);
 	unsigned long uninitialized_var(offset);
-	struct isofs_sb_info *sbi = ISOFS_SB(dir->i_sb);
 	struct inode *inode;
 	struct page *page;
 
@@ -176,21 +175,13 @@ struct dentry *isofs_lookup(struct inode *dir, struct dentry *dentry, struct nam
 	if (!page)
 		return ERR_PTR(-ENOMEM);
 
-	mutex_lock(&sbi->s_mutex);
 	found = isofs_find_entry(dir, dentry,
 				&block, &offset,
 				page_address(page),
 				1024 + page_address(page));
 	__free_page(page);
 
-	inode = NULL;
-	if (found) {
-		inode = isofs_iget(dir->i_sb, block, offset);
-		if (IS_ERR(inode)) {
-			mutex_unlock(&sbi->s_mutex);
-			return ERR_CAST(inode);
-		}
-	}
-	mutex_unlock(&sbi->s_mutex);
+	inode = found ? isofs_iget(dir->i_sb, block, offset) : NULL;
+
 	return d_splice_alias(inode, dentry);
 }
diff --git a/fs/isofs/rock.c b/fs/isofs/rock.c
index f9cd04db6ea..1fbc7de88f5 100644
--- a/fs/isofs/rock.c
+++ b/fs/isofs/rock.c
@@ -678,7 +678,6 @@ static int rock_ridge_symlink_readpage(struct file *file, struct page *page)
 
 	init_rock_state(&rs, inode);
 	block = ei->i_iget5_block;
-	mutex_lock(&sbi->s_mutex);
 	bh = sb_bread(inode->i_sb, block);
 	if (!bh)
 		goto out_noread;
@@ -748,7 +747,6 @@ repeat:
 		goto fail;
 	brelse(bh);
 	*rpnt = '\0';
-	mutex_unlock(&sbi->s_mutex);
 	SetPageUptodate(page);
 	kunmap(page);
 	unlock_page(page);
@@ -765,7 +763,6 @@ out_bad_span:
 	printk("symlink spans iso9660 blocks\n");
 fail:
 	brelse(bh);
-	mutex_unlock(&sbi->s_mutex);
 error:
 	SetPageError(page);
 	kunmap(page);
diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 6a79fd0a1a3..2c62c5aae82 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -97,10 +97,14 @@ static int __try_to_free_cp_buf(struct journal_head *jh)
 
 	if (jh->b_jlist == BJ_None && !buffer_locked(bh) &&
 	    !buffer_dirty(bh) && !buffer_write_io_error(bh)) {
+		/*
+		 * Get our reference so that bh cannot be freed before
+		 * we unlock it
+		 */
+		get_bh(bh);
 		JBUFFER_TRACE(jh, "remove from checkpoint list");
 		ret = __jbd2_journal_remove_checkpoint(jh) + 1;
 		jbd_unlock_bh_state(bh);
-		jbd2_journal_remove_journal_head(bh);
 		BUFFER_TRACE(bh, "release");
 		__brelse(bh);
 	} else {
@@ -223,8 +227,8 @@ restart:
 			spin_lock(&journal->j_list_lock);
 			goto restart;
 		}
+		get_bh(bh);
 		if (buffer_locked(bh)) {
-			atomic_inc(&bh->b_count);
 			spin_unlock(&journal->j_list_lock);
 			jbd_unlock_bh_state(bh);
 			wait_on_buffer(bh);
@@ -243,7 +247,6 @@ restart:
 		 */
 		released = __jbd2_journal_remove_checkpoint(jh);
 		jbd_unlock_bh_state(bh);
-		jbd2_journal_remove_journal_head(bh);
 		__brelse(bh);
 	}
 
@@ -284,7 +287,7 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh,
 	int ret = 0;
 
 	if (buffer_locked(bh)) {
-		atomic_inc(&bh->b_count);
+		get_bh(bh);
 		spin_unlock(&journal->j_list_lock);
 		jbd_unlock_bh_state(bh);
 		wait_on_buffer(bh);
@@ -316,12 +319,12 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh,
 		ret = 1;
 		if (unlikely(buffer_write_io_error(bh)))
 			ret = -EIO;
+		get_bh(bh);
 		J_ASSERT_JH(jh, !buffer_jbddirty(bh));
 		BUFFER_TRACE(bh, "remove from checkpoint");
 		__jbd2_journal_remove_checkpoint(jh);
 		spin_unlock(&journal->j_list_lock);
 		jbd_unlock_bh_state(bh);
-		jbd2_journal_remove_journal_head(bh);
 		__brelse(bh);
 	} else {
 		/*
@@ -554,7 +557,8 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
 /*
  * journal_clean_one_cp_list
  *
- * Find all the written-back checkpoint buffers in the given list and release them.
+ * Find all the written-back checkpoint buffers in the given list and
+ * release them.
  *
  * Called with the journal locked.
  * Called with j_list_lock held.
@@ -663,8 +667,8 @@ out:
  * checkpoint lists.
  *
  * The function returns 1 if it frees the transaction, 0 otherwise.
+ * The function can free jh and bh.
  *
- * This function is called with the journal locked.
  * This function is called with j_list_lock held.
  * This function is called with jbd_lock_bh_state(jh2bh(jh))
  */
@@ -684,13 +688,14 @@ int __jbd2_journal_remove_checkpoint(struct journal_head *jh)
 	}
 	journal = transaction->t_journal;
 
+	JBUFFER_TRACE(jh, "removing from transaction");
 	__buffer_unlink(jh);
 	jh->b_cp_transaction = NULL;
+	jbd2_journal_put_journal_head(jh);
 
 	if (transaction->t_checkpoint_list != NULL ||
 	    transaction->t_checkpoint_io_list != NULL)
 		goto out;
-	JBUFFER_TRACE(jh, "transaction has no more buffers");
 
 	/*
 	 * There is one special case to worry about: if we have just pulled the
@@ -701,10 +706,8 @@ int __jbd2_journal_remove_checkpoint(struct journal_head *jh)
 	 * The locking here around t_state is a bit sleazy.
 	 * See the comment at the end of jbd2_journal_commit_transaction().
 	 */
-	if (transaction->t_state != T_FINISHED) {
-		JBUFFER_TRACE(jh, "belongs to running/committing transaction");
+	if (transaction->t_state != T_FINISHED)
 		goto out;
-	}
 
 	/* OK, that was the last buffer for the transaction: we can now
 	   safely remove this transaction from the log */
@@ -723,7 +726,6 @@ int __jbd2_journal_remove_checkpoint(struct journal_head *jh)
 	wake_up(&journal->j_wait_logspace);
 	ret = 1;
 out:
-	JBUFFER_TRACE(jh, "exit");
 	return ret;
 }
 
@@ -742,6 +744,8 @@ void __jbd2_journal_insert_checkpoint(struct journal_head *jh,
 	J_ASSERT_JH(jh, buffer_dirty(jh2bh(jh)) || buffer_jbddirty(jh2bh(jh)));
 	J_ASSERT_JH(jh, jh->b_cp_transaction == NULL);
 
+	/* Get reference for checkpointing transaction */
+	jbd2_journal_grab_journal_head(jh2bh(jh));
 	jh->b_cp_transaction = transaction;
 
 	if (!transaction->t_checkpoint_list) {
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 7f21cf3aaf9..eef6979821a 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -848,10 +848,16 @@ restart_loop:
 	while (commit_transaction->t_forget) {
 		transaction_t *cp_transaction;
 		struct buffer_head *bh;
+		int try_to_free = 0;
 
 		jh = commit_transaction->t_forget;
 		spin_unlock(&journal->j_list_lock);
 		bh = jh2bh(jh);
+		/*
+		 * Get a reference so that bh cannot be freed before we are
+		 * done with it.
+		 */
+		get_bh(bh);
 		jbd_lock_bh_state(bh);
 		J_ASSERT_JH(jh,	jh->b_transaction == commit_transaction);
 
@@ -914,28 +920,27 @@ restart_loop:
 			__jbd2_journal_insert_checkpoint(jh, commit_transaction);
 			if (is_journal_aborted(journal))
 				clear_buffer_jbddirty(bh);
-			JBUFFER_TRACE(jh, "refile for checkpoint writeback");
-			__jbd2_journal_refile_buffer(jh);
-			jbd_unlock_bh_state(bh);
 		} else {
 			J_ASSERT_BH(bh, !buffer_dirty(bh));
-			/* The buffer on BJ_Forget list and not jbddirty means
+			/*
+			 * The buffer on BJ_Forget list and not jbddirty means
 			 * it has been freed by this transaction and hence it
 			 * could not have been reallocated until this
 			 * transaction has committed. *BUT* it could be
 			 * reallocated once we have written all the data to
 			 * disk and before we process the buffer on BJ_Forget
-			 * list. */
-			JBUFFER_TRACE(jh, "refile or unfile freed buffer");
-			__jbd2_journal_refile_buffer(jh);
-			if (!jh->b_transaction) {
-				jbd_unlock_bh_state(bh);
-				 /* needs a brelse */
-				jbd2_journal_remove_journal_head(bh);
-				release_buffer_page(bh);
-			} else
-				jbd_unlock_bh_state(bh);
+			 * list.
+			 */
+			if (!jh->b_next_transaction)
+				try_to_free = 1;
 		}
+		JBUFFER_TRACE(jh, "refile or unfile buffer");
+		__jbd2_journal_refile_buffer(jh);
+		jbd_unlock_bh_state(bh);
+		if (try_to_free)
+			release_buffer_page(bh);	/* Drops bh reference */
+		else
+			__brelse(bh);
 		cond_resched_lock(&journal->j_list_lock);
 	}
 	spin_unlock(&journal->j_list_lock);
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 9a782699030..0dfa5b598e6 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -2078,10 +2078,9 @@ static void journal_free_journal_head(struct journal_head *jh)
  * When a buffer has its BH_JBD bit set it is immune from being released by
  * core kernel code, mainly via ->b_count.
  *
- * A journal_head may be detached from its buffer_head when the journal_head's
- * b_transaction, b_cp_transaction and b_next_transaction pointers are NULL.
- * Various places in JBD call jbd2_journal_remove_journal_head() to indicate that the
- * journal_head can be dropped if needed.
+ * A journal_head is detached from its buffer_head when the journal_head's
+ * b_jcount reaches zero. Running transaction (b_transaction) and checkpoint
+ * transaction (b_cp_transaction) hold their references to b_jcount.
  *
  * Various places in the kernel want to attach a journal_head to a buffer_head
  * _before_ attaching the journal_head to a transaction.  To protect the
@@ -2094,17 +2093,16 @@ static void journal_free_journal_head(struct journal_head *jh)
  *	(Attach a journal_head if needed.  Increments b_jcount)
  *	struct journal_head *jh = jbd2_journal_add_journal_head(bh);
  *	...
+ *      (Get another reference for transaction)
+ *	jbd2_journal_grab_journal_head(bh);
  *	jh->b_transaction = xxx;
+ *	(Put original reference)
  *	jbd2_journal_put_journal_head(jh);
- *
- * Now, the journal_head's b_jcount is zero, but it is safe from being released
- * because it has a non-zero b_transaction.
  */
 
 /*
  * Give a buffer_head a journal_head.
  *
- * Doesn't need the journal lock.
  * May sleep.
  */
 struct journal_head *jbd2_journal_add_journal_head(struct buffer_head *bh)
@@ -2168,61 +2166,29 @@ static void __journal_remove_journal_head(struct buffer_head *bh)
 	struct journal_head *jh = bh2jh(bh);
 
 	J_ASSERT_JH(jh, jh->b_jcount >= 0);
-
-	get_bh(bh);
-	if (jh->b_jcount == 0) {
-		if (jh->b_transaction == NULL &&
-				jh->b_next_transaction == NULL &&
-				jh->b_cp_transaction == NULL) {
-			J_ASSERT_JH(jh, jh->b_jlist == BJ_None);
-			J_ASSERT_BH(bh, buffer_jbd(bh));
-			J_ASSERT_BH(bh, jh2bh(jh) == bh);
-			BUFFER_TRACE(bh, "remove journal_head");
-			if (jh->b_frozen_data) {
-				printk(KERN_WARNING "%s: freeing "
-						"b_frozen_data\n",
-						__func__);
-				jbd2_free(jh->b_frozen_data, bh->b_size);
-			}
-			if (jh->b_committed_data) {
-				printk(KERN_WARNING "%s: freeing "
-						"b_committed_data\n",
-						__func__);
-				jbd2_free(jh->b_committed_data, bh->b_size);
-			}
-			bh->b_private = NULL;
-			jh->b_bh = NULL;	/* debug, really */
-			clear_buffer_jbd(bh);
-			__brelse(bh);
-			journal_free_journal_head(jh);
-		} else {
-			BUFFER_TRACE(bh, "journal_head was locked");
-		}
+	J_ASSERT_JH(jh, jh->b_transaction == NULL);
+	J_ASSERT_JH(jh, jh->b_next_transaction == NULL);
+	J_ASSERT_JH(jh, jh->b_cp_transaction == NULL);
+	J_ASSERT_JH(jh, jh->b_jlist == BJ_None);
+	J_ASSERT_BH(bh, buffer_jbd(bh));
+	J_ASSERT_BH(bh, jh2bh(jh) == bh);
+	BUFFER_TRACE(bh, "remove journal_head");
+	if (jh->b_frozen_data) {
+		printk(KERN_WARNING "%s: freeing b_frozen_data\n", __func__);
+		jbd2_free(jh->b_frozen_data, bh->b_size);
 	}
+	if (jh->b_committed_data) {
+		printk(KERN_WARNING "%s: freeing b_committed_data\n", __func__);
+		jbd2_free(jh->b_committed_data, bh->b_size);
+	}
+	bh->b_private = NULL;
+	jh->b_bh = NULL;	/* debug, really */
+	clear_buffer_jbd(bh);
+	journal_free_journal_head(jh);
 }
 
 /*
- * jbd2_journal_remove_journal_head(): if the buffer isn't attached to a transaction
- * and has a zero b_jcount then remove and release its journal_head.   If we did
- * see that the buffer is not used by any transaction we also "logically"
- * decrement ->b_count.
- *
- * We in fact take an additional increment on ->b_count as a convenience,
- * because the caller usually wants to do additional things with the bh
- * after calling here.
- * The caller of jbd2_journal_remove_journal_head() *must* run __brelse(bh) at some
- * time.  Once the caller has run __brelse(), the buffer is eligible for
- * reaping by try_to_free_buffers().
- */
-void jbd2_journal_remove_journal_head(struct buffer_head *bh)
-{
-	jbd_lock_bh_journal_head(bh);
-	__journal_remove_journal_head(bh);
-	jbd_unlock_bh_journal_head(bh);
-}
-
-/*
- * Drop a reference on the passed journal_head.  If it fell to zero then try to
+ * Drop a reference on the passed journal_head.  If it fell to zero then
  * release the journal_head from the buffer_head.
  */
 void jbd2_journal_put_journal_head(struct journal_head *jh)
@@ -2232,11 +2198,12 @@ void jbd2_journal_put_journal_head(struct journal_head *jh)
 	jbd_lock_bh_journal_head(bh);
 	J_ASSERT_JH(jh, jh->b_jcount > 0);
 	--jh->b_jcount;
-	if (!jh->b_jcount && !jh->b_transaction) {
+	if (!jh->b_jcount) {
 		__journal_remove_journal_head(bh);
+		jbd_unlock_bh_journal_head(bh);
 		__brelse(bh);
-	}
-	jbd_unlock_bh_journal_head(bh);
+	} else
+		jbd_unlock_bh_journal_head(bh);
 }
 
 /*
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index 3eec82d32fd..2d7109414cd 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -30,6 +30,7 @@
 #include <linux/module.h>
 
 static void __jbd2_journal_temp_unlink_buffer(struct journal_head *jh);
+static void __jbd2_journal_unfile_buffer(struct journal_head *jh);
 
 /*
  * jbd2_get_transaction: obtain a new transaction_t object.
@@ -764,7 +765,6 @@ repeat:
 	if (!jh->b_transaction) {
 		JBUFFER_TRACE(jh, "no transaction");
 		J_ASSERT_JH(jh, !jh->b_next_transaction);
-		jh->b_transaction = transaction;
 		JBUFFER_TRACE(jh, "file as BJ_Reserved");
 		spin_lock(&journal->j_list_lock);
 		__jbd2_journal_file_buffer(jh, transaction, BJ_Reserved);
@@ -814,7 +814,6 @@ out:
  * int jbd2_journal_get_write_access() - notify intent to modify a buffer for metadata (not data) update.
  * @handle: transaction to add buffer modifications to
  * @bh:     bh to be used for metadata writes
- * @credits: variable that will receive credits for the buffer
  *
  * Returns an error code or 0 on success.
  *
@@ -896,8 +895,6 @@ int jbd2_journal_get_create_access(handle_t *handle, struct buffer_head *bh)
 		 * committed and so it's safe to clear the dirty bit.
 		 */
 		clear_buffer_dirty(jh2bh(jh));
-		jh->b_transaction = transaction;
-
 		/* first access by this transaction */
 		jh->b_modified = 0;
 
@@ -932,7 +929,6 @@ out:
  *     non-rewindable consequences
  * @handle: transaction
  * @bh: buffer to undo
- * @credits: store the number of taken credits here (if not NULL)
  *
  * Sometimes there is a need to distinguish between metadata which has
  * been committed to disk and that which has not.  The ext3fs code uses
@@ -1232,8 +1228,6 @@ int jbd2_journal_forget (handle_t *handle, struct buffer_head *bh)
 			__jbd2_journal_file_buffer(jh, transaction, BJ_Forget);
 		} else {
 			__jbd2_journal_unfile_buffer(jh);
-			jbd2_journal_remove_journal_head(bh);
-			__brelse(bh);
 			if (!buffer_jbd(bh)) {
 				spin_unlock(&journal->j_list_lock);
 				jbd_unlock_bh_state(bh);
@@ -1556,19 +1550,32 @@ void __jbd2_journal_temp_unlink_buffer(struct journal_head *jh)
 		mark_buffer_dirty(bh);	/* Expose it to the VM */
 }
 
-void __jbd2_journal_unfile_buffer(struct journal_head *jh)
+/*
+ * Remove buffer from all transactions.
+ *
+ * Called with bh_state lock and j_list_lock
+ *
+ * jh and bh may be already freed when this function returns.
+ */
+static void __jbd2_journal_unfile_buffer(struct journal_head *jh)
 {
 	__jbd2_journal_temp_unlink_buffer(jh);
 	jh->b_transaction = NULL;
+	jbd2_journal_put_journal_head(jh);
 }
 
 void jbd2_journal_unfile_buffer(journal_t *journal, struct journal_head *jh)
 {
-	jbd_lock_bh_state(jh2bh(jh));
+	struct buffer_head *bh = jh2bh(jh);
+
+	/* Get reference so that buffer cannot be freed before we unlock it */
+	get_bh(bh);
+	jbd_lock_bh_state(bh);
 	spin_lock(&journal->j_list_lock);
 	__jbd2_journal_unfile_buffer(jh);
 	spin_unlock(&journal->j_list_lock);
-	jbd_unlock_bh_state(jh2bh(jh));
+	jbd_unlock_bh_state(bh);
+	__brelse(bh);
 }
 
 /*
@@ -1595,8 +1602,6 @@ __journal_try_to_free_buffer(journal_t *journal, struct buffer_head *bh)
 		if (jh->b_jlist == BJ_None) {
 			JBUFFER_TRACE(jh, "remove from checkpoint list");
 			__jbd2_journal_remove_checkpoint(jh);
-			jbd2_journal_remove_journal_head(bh);
-			__brelse(bh);
 		}
 	}
 	spin_unlock(&journal->j_list_lock);
@@ -1659,7 +1664,6 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal,
 		/*
 		 * We take our own ref against the journal_head here to avoid
 		 * having to add tons of locking around each instance of
-		 * jbd2_journal_remove_journal_head() and
 		 * jbd2_journal_put_journal_head().
 		 */
 		jh = jbd2_journal_grab_journal_head(bh);
@@ -1697,10 +1701,9 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction)
 	int may_free = 1;
 	struct buffer_head *bh = jh2bh(jh);
 
-	__jbd2_journal_unfile_buffer(jh);
-
 	if (jh->b_cp_transaction) {
 		JBUFFER_TRACE(jh, "on running+cp transaction");
+		__jbd2_journal_temp_unlink_buffer(jh);
 		/*
 		 * We don't want to write the buffer anymore, clear the
 		 * bit so that we don't confuse checks in
@@ -1711,8 +1714,7 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction)
 		may_free = 0;
 	} else {
 		JBUFFER_TRACE(jh, "on running transaction");
-		jbd2_journal_remove_journal_head(bh);
-		__brelse(bh);
+		__jbd2_journal_unfile_buffer(jh);
 	}
 	return may_free;
 }
@@ -1990,6 +1992,8 @@ void __jbd2_journal_file_buffer(struct journal_head *jh,
 
 	if (jh->b_transaction)
 		__jbd2_journal_temp_unlink_buffer(jh);
+	else
+		jbd2_journal_grab_journal_head(bh);
 	jh->b_transaction = transaction;
 
 	switch (jlist) {
@@ -2041,9 +2045,10 @@ void jbd2_journal_file_buffer(struct journal_head *jh,
  * already started to be used by a subsequent transaction, refile the
  * buffer on that transaction's metadata list.
  *
- * Called under journal->j_list_lock
- *
+ * Called under j_list_lock
  * Called under jbd_lock_bh_state(jh2bh(jh))
+ *
+ * jh and bh may be already free when this function returns
  */
 void __jbd2_journal_refile_buffer(struct journal_head *jh)
 {
@@ -2067,6 +2072,11 @@ void __jbd2_journal_refile_buffer(struct journal_head *jh)
 
 	was_dirty = test_clear_buffer_jbddirty(bh);
 	__jbd2_journal_temp_unlink_buffer(jh);
+	/*
+	 * We set b_transaction here because b_next_transaction will inherit
+	 * our jh reference and thus __jbd2_journal_file_buffer() must not
+	 * take a new one.
+	 */
 	jh->b_transaction = jh->b_next_transaction;
 	jh->b_next_transaction = NULL;
 	if (buffer_freed(bh))
@@ -2083,30 +2093,21 @@ void __jbd2_journal_refile_buffer(struct journal_head *jh)
 }
 
 /*
- * For the unlocked version of this call, also make sure that any
- * hanging journal_head is cleaned up if necessary.
- *
- * __jbd2_journal_refile_buffer is usually called as part of a single locked
- * operation on a buffer_head, in which the caller is probably going to
- * be hooking the journal_head onto other lists.  In that case it is up
- * to the caller to remove the journal_head if necessary.  For the
- * unlocked jbd2_journal_refile_buffer call, the caller isn't going to be
- * doing anything else to the buffer so we need to do the cleanup
- * ourselves to avoid a jh leak.
- *
- * *** The journal_head may be freed by this call! ***
+ * __jbd2_journal_refile_buffer() with necessary locking added. We take our
+ * bh reference so that we can safely unlock bh.
+ *
+ * The jh and bh may be freed by this call.
  */
 void jbd2_journal_refile_buffer(journal_t *journal, struct journal_head *jh)
 {
 	struct buffer_head *bh = jh2bh(jh);
 
+	/* Get reference so that buffer cannot be freed before we unlock it */
+	get_bh(bh);
 	jbd_lock_bh_state(bh);
 	spin_lock(&journal->j_list_lock);
-
 	__jbd2_journal_refile_buffer(jh);
 	jbd_unlock_bh_state(bh);
-	jbd2_journal_remove_journal_head(bh);
-
 	spin_unlock(&journal->j_list_lock);
 	__brelse(bh);
 }
diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c
index 828a0e1ea43..3675b3cdee8 100644
--- a/fs/jffs2/acl.c
+++ b/fs/jffs2/acl.c
@@ -259,12 +259,12 @@ static int jffs2_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 	return rc;
 }
 
-int jffs2_check_acl(struct inode *inode, int mask, unsigned int flags)
+int jffs2_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	int rc;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	acl = jffs2_get_acl(inode, ACL_TYPE_ACCESS);
diff --git a/fs/jffs2/acl.h b/fs/jffs2/acl.h
index 3119f59253d..5e42de8d954 100644
--- a/fs/jffs2/acl.h
+++ b/fs/jffs2/acl.h
@@ -26,7 +26,7 @@ struct jffs2_acl_header {
 
 #ifdef CONFIG_JFFS2_FS_POSIX_ACL
 
-extern int jffs2_check_acl(struct inode *, int, unsigned int);
+extern int jffs2_check_acl(struct inode *, int);
 extern int jffs2_acl_chmod(struct inode *);
 extern int jffs2_init_acl_pre(struct inode *, struct inode *, int *);
 extern int jffs2_init_acl_post(struct inode *);
diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c
index 4bca6a2e5c0..5f243cd63af 100644
--- a/fs/jffs2/dir.c
+++ b/fs/jffs2/dir.c
@@ -102,10 +102,8 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target,
 	mutex_unlock(&dir_f->sem);
 	if (ino) {
 		inode = jffs2_iget(dir_i->i_sb, ino);
-		if (IS_ERR(inode)) {
+		if (IS_ERR(inode))
 			printk(KERN_WARNING "iget() failed for ino #%u\n", ino);
-			return ERR_CAST(inode);
-		}
 	}
 
 	return d_splice_alias(inode, target);
@@ -822,7 +820,10 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry,
 
 	if (victim_f) {
 		/* There was a victim. Kill it off nicely */
-		drop_nlink(new_dentry->d_inode);
+		if (S_ISDIR(new_dentry->d_inode->i_mode))
+			clear_nlink(new_dentry->d_inode);
+		else
+			drop_nlink(new_dentry->d_inode);
 		/* Don't oops if the victim was a dirent pointing to an
 		   inode which didn't exist. */
 		if (victim_f->inocache) {
diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index 1c0a08d711a..3989f7e09f7 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -27,13 +27,20 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 			struct page **pagep, void **fsdata);
 static int jffs2_readpage (struct file *filp, struct page *pg);
 
-int jffs2_fsync(struct file *filp, int datasync)
+int jffs2_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb);
+	int ret;
+
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
 
+	mutex_lock(&inode->i_mutex);
 	/* Trigger GC to flush any pending writes for this inode */
 	jffs2_flush_wbuf_gc(c, inode->i_ino);
+	mutex_unlock(&inode->i_mutex);
 
 	return 0;
 }
diff --git a/fs/jffs2/os-linux.h b/fs/jffs2/os-linux.h
index 65c6c43ca48..9c252835e8e 100644
--- a/fs/jffs2/os-linux.h
+++ b/fs/jffs2/os-linux.h
@@ -158,7 +158,7 @@ extern const struct inode_operations jffs2_dir_inode_operations;
 extern const struct file_operations jffs2_file_operations;
 extern const struct inode_operations jffs2_file_inode_operations;
 extern const struct address_space_operations jffs2_file_address_operations;
-int jffs2_fsync(struct file *, int);
+int jffs2_fsync(struct file *, loff_t, loff_t, int);
 int jffs2_do_readpage_unlock (struct inode *inode, struct page *pg);
 
 /* ioctl.c */
diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c
index e5de9422fa3..8a0a0666d5a 100644
--- a/fs/jfs/acl.c
+++ b/fs/jfs/acl.c
@@ -114,11 +114,11 @@ out:
 	return rc;
 }
 
-int jfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int jfs_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	acl = jfs_get_acl(inode, ACL_TYPE_ACCESS);
diff --git a/fs/jfs/file.c b/fs/jfs/file.c
index c5ce6c1d1ff..7527855b5cc 100644
--- a/fs/jfs/file.c
+++ b/fs/jfs/file.c
@@ -28,19 +28,26 @@
 #include "jfs_acl.h"
 #include "jfs_debug.h"
 
-int jfs_fsync(struct file *file, int datasync)
+int jfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	int rc = 0;
 
+	rc = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (rc)
+		return rc;
+
+	mutex_lock(&inode->i_mutex);
 	if (!(inode->i_state & I_DIRTY) ||
 	    (datasync && !(inode->i_state & I_DIRTY_DATASYNC))) {
 		/* Make sure committed changes hit the disk */
 		jfs_flush_journal(JFS_SBI(inode->i_sb)->log, 1);
+		mutex_unlock(&inode->i_mutex);
 		return rc;
 	}
 
 	rc |= jfs_commit_inode(inode, 1);
+	mutex_unlock(&inode->i_mutex);
 
 	return rc ? -EIO : 0;
 }
@@ -66,9 +73,9 @@ static int jfs_open(struct inode *inode, struct file *file)
 		struct jfs_inode_info *ji = JFS_IP(inode);
 		spin_lock_irq(&ji->ag_lock);
 		if (ji->active_ag == -1) {
-			ji->active_ag = ji->agno;
-			atomic_inc(
-			    &JFS_SBI(inode->i_sb)->bmap->db_active[ji->agno]);
+			struct jfs_sb_info *jfs_sb = JFS_SBI(inode->i_sb);
+			ji->active_ag = BLKTOAG(addressPXD(&ji->ixpxd), jfs_sb);
+			atomic_inc( &jfs_sb->bmap->db_active[ji->active_ag]);
 		}
 		spin_unlock_irq(&ji->ag_lock);
 	}
@@ -110,6 +117,8 @@ int jfs_setattr(struct dentry *dentry, struct iattr *iattr)
 
 	if ((iattr->ia_valid & ATTR_SIZE) &&
 	    iattr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		rc = vmtruncate(inode, iattr->ia_size);
 		if (rc)
 			return rc;
diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c
index 109655904bb..77b69b27f82 100644
--- a/fs/jfs/inode.c
+++ b/fs/jfs/inode.c
@@ -329,8 +329,8 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				offset, nr_segs, jfs_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 jfs_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
diff --git a/fs/jfs/jfs_acl.h b/fs/jfs/jfs_acl.h
index f9285c4900f..54e07559878 100644
--- a/fs/jfs/jfs_acl.h
+++ b/fs/jfs/jfs_acl.h
@@ -20,7 +20,7 @@
 
 #ifdef CONFIG_JFS_POSIX_ACL
 
-int jfs_check_acl(struct inode *, int, unsigned int flags);
+int jfs_check_acl(struct inode *, int);
 int jfs_init_acl(tid_t, struct inode *, struct inode *);
 int jfs_acl_chmod(struct inode *inode);
 
diff --git a/fs/jfs/jfs_imap.c b/fs/jfs/jfs_imap.c
index ed53a474016..b78b2f978f0 100644
--- a/fs/jfs/jfs_imap.c
+++ b/fs/jfs/jfs_imap.c
@@ -397,7 +397,7 @@ int diRead(struct inode *ip)
 	release_metapage(mp);
 
 	/* set the ag for the inode */
-	JFS_IP(ip)->agno = BLKTOAG(agstart, sbi);
+	JFS_IP(ip)->agstart = agstart;
 	JFS_IP(ip)->active_ag = -1;
 
 	return (rc);
@@ -901,7 +901,7 @@ int diFree(struct inode *ip)
 
 	/* get the allocation group for this ino.
 	 */
-	agno = JFS_IP(ip)->agno;
+	agno = BLKTOAG(JFS_IP(ip)->agstart, JFS_SBI(ip->i_sb));
 
 	/* Lock the AG specific inode map information
 	 */
@@ -1315,12 +1315,11 @@ int diFree(struct inode *ip)
 static inline void
 diInitInode(struct inode *ip, int iagno, int ino, int extno, struct iag * iagp)
 {
-	struct jfs_sb_info *sbi = JFS_SBI(ip->i_sb);
 	struct jfs_inode_info *jfs_ip = JFS_IP(ip);
 
 	ip->i_ino = (iagno << L2INOSPERIAG) + ino;
 	jfs_ip->ixpxd = iagp->inoext[extno];
-	jfs_ip->agno = BLKTOAG(le64_to_cpu(iagp->agstart), sbi);
+	jfs_ip->agstart = le64_to_cpu(iagp->agstart);
 	jfs_ip->active_ag = -1;
 }
 
@@ -1379,7 +1378,7 @@ int diAlloc(struct inode *pip, bool dir, struct inode *ip)
 	 */
 
 	/* get the ag number of this iag */
-	agno = JFS_IP(pip)->agno;
+	agno = BLKTOAG(JFS_IP(pip)->agstart, JFS_SBI(pip->i_sb));
 
 	if (atomic_read(&JFS_SBI(pip->i_sb)->bmap->db_active[agno])) {
 		/*
@@ -2921,10 +2920,9 @@ int diExtendFS(struct inode *ipimap, struct inode *ipbmap)
 			continue;
 		}
 
-		/* agstart that computes to the same ag is treated as same; */
 		agstart = le64_to_cpu(iagp->agstart);
-		/* iagp->agstart = agstart & ~(mp->db_agsize - 1); */
 		n = agstart >> mp->db_agl2size;
+		iagp->agstart = cpu_to_le64((s64)n << mp->db_agl2size);
 
 		/* compute backed inodes */
 		numinos = (EXTSPERIAG - le32_to_cpu(iagp->nfreeexts))
diff --git a/fs/jfs/jfs_incore.h b/fs/jfs/jfs_incore.h
index 1439f119ec8..584a4a1a6e8 100644
--- a/fs/jfs/jfs_incore.h
+++ b/fs/jfs/jfs_incore.h
@@ -50,8 +50,9 @@ struct jfs_inode_info {
 	short	btindex;	/* btpage entry index*/
 	struct inode *ipimap;	/* inode map			*/
 	unsigned long cflag;	/* commit flags		*/
+	u64	agstart;	/* agstart of the containing IAG */
 	u16	bxflag;		/* xflag of pseudo buffer?	*/
-	unchar	agno;		/* ag number			*/
+	unchar	pad;
 	signed char active_ag;	/* ag currently allocating from	*/
 	lid_t	blid;		/* lid of pseudo buffer?	*/
 	lid_t	atlhead;	/* anonymous tlock list head	*/
diff --git a/fs/jfs/jfs_inode.h b/fs/jfs/jfs_inode.h
index ec2fb8b945f..9271cfe4a14 100644
--- a/fs/jfs/jfs_inode.h
+++ b/fs/jfs/jfs_inode.h
@@ -21,7 +21,7 @@
 struct fid;
 
 extern struct inode *ialloc(struct inode *, umode_t);
-extern int jfs_fsync(struct file *, int);
+extern int jfs_fsync(struct file *, loff_t, loff_t, int);
 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);
diff --git a/fs/jfs/jfs_logmgr.c b/fs/jfs/jfs_logmgr.c
index 278e3fb40b7..583636f745e 100644
--- a/fs/jfs/jfs_logmgr.c
+++ b/fs/jfs/jfs_logmgr.c
@@ -1123,7 +1123,7 @@ int lmLogOpen(struct super_block *sb)
 	bdev = blkdev_get_by_dev(sbi->logdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL,
 				 log);
 	if (IS_ERR(bdev)) {
-		rc = -PTR_ERR(bdev);
+		rc = PTR_ERR(bdev);
 		goto free;
 	}
 
diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index eaaf2b511e8..03787ef6a11 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -1456,34 +1456,23 @@ static struct dentry *jfs_lookup(struct inode *dip, struct dentry *dentry, struc
 	ino_t inum;
 	struct inode *ip;
 	struct component_name key;
-	const char *name = dentry->d_name.name;
-	int len = dentry->d_name.len;
 	int rc;
 
-	jfs_info("jfs_lookup: name = %s", name);
-
-	if ((name[0] == '.') && (len == 1))
-		inum = dip->i_ino;
-	else if (strcmp(name, "..") == 0)
-		inum = PARENT(dip);
-	else {
-		if ((rc = get_UCSname(&key, dentry)))
-			return ERR_PTR(rc);
-		rc = dtSearch(dip, &key, &inum, &btstack, JFS_LOOKUP);
-		free_UCSname(&key);
-		if (rc == -ENOENT) {
-			d_add(dentry, NULL);
-			return NULL;
-		} else if (rc) {
-			jfs_err("jfs_lookup: dtSearch returned %d", rc);
-			return ERR_PTR(rc);
-		}
-	}
-
-	ip = jfs_iget(dip->i_sb, inum);
-	if (IS_ERR(ip)) {
-		jfs_err("jfs_lookup: iget failed on inum %d", (uint) inum);
-		return ERR_CAST(ip);
+	jfs_info("jfs_lookup: name = %s", dentry->d_name.name);
+
+	if ((rc = get_UCSname(&key, dentry)))
+		return ERR_PTR(rc);
+	rc = dtSearch(dip, &key, &inum, &btstack, JFS_LOOKUP);
+	free_UCSname(&key);
+	if (rc == -ENOENT) {
+		ip = NULL;
+	} else if (rc) {
+		jfs_err("jfs_lookup: dtSearch returned %d", rc);
+		ip = ERR_PTR(rc);
+	} else {
+		ip = jfs_iget(dip->i_sb, inum);
+		if (IS_ERR(ip))
+			jfs_err("jfs_lookup: iget failed on inum %d", (uint)inum);
 	}
 
 	return d_splice_alias(ip, dentry);
@@ -1597,8 +1586,6 @@ out:
 
 static int jfs_ci_revalidate(struct dentry *dentry, struct nameidata *nd)
 {
-	if (nd && nd->flags & LOOKUP_RCU)
-		return -ECHILD;
 	/*
 	 * This is not negative dentry. Always valid.
 	 *
@@ -1624,10 +1611,8 @@ static int jfs_ci_revalidate(struct dentry *dentry, struct nameidata *nd)
 	 * case sensitive name which is specified by user if this is
 	 * for creation.
 	 */
-	if (!(nd->flags & (LOOKUP_CONTINUE | LOOKUP_PARENT))) {
-		if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
-			return 0;
-	}
+	if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
+		return 0;
 	return 1;
 }
 
diff --git a/fs/jfs/resize.c b/fs/jfs/resize.c
index 8ea5efb5a34..8d0c1c7c082 100644
--- a/fs/jfs/resize.c
+++ b/fs/jfs/resize.c
@@ -80,7 +80,7 @@ int jfs_extendfs(struct super_block *sb, s64 newLVSize, int newLogSize)
 	int log_formatted = 0;
 	struct inode *iplist[1];
 	struct jfs_superblock *j_sb, *j_sb2;
-	uint old_agsize;
+	s64 old_agsize;
 	int agsizechanged = 0;
 	struct buffer_head *bh, *bh2;
 
diff --git a/fs/libfs.c b/fs/libfs.c
index c88eab55aec..c18e9a1235b 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -16,6 +16,8 @@
 
 #include <asm/uaccess.h>
 
+#include "internal.h"
+
 static inline int simple_positive(struct dentry *dentry)
 {
 	return dentry->d_inode && !d_unhashed(dentry);
@@ -246,13 +248,11 @@ struct dentry *mount_pseudo(struct file_system_type *fs_type, char *name,
 	root->i_ino = 1;
 	root->i_mode = S_IFDIR | S_IRUSR | S_IWUSR;
 	root->i_atime = root->i_mtime = root->i_ctime = CURRENT_TIME;
-	dentry = d_alloc(NULL, &d_name);
+	dentry = __d_alloc(s, &d_name);
 	if (!dentry) {
 		iput(root);
 		goto Enomem;
 	}
-	dentry->d_sb = s;
-	dentry->d_parent = dentry;
 	d_instantiate(dentry, root);
 	s->s_root = dentry;
 	s->s_d_op = dops;
@@ -328,8 +328,10 @@ int simple_rename(struct inode *old_dir, struct dentry *old_dentry,
 
 	if (new_dentry->d_inode) {
 		simple_unlink(new_dir, new_dentry);
-		if (they_are_dirs)
+		if (they_are_dirs) {
+			drop_nlink(new_dentry->d_inode);
 			drop_nlink(old_dir);
+		}
 	} else if (they_are_dirs) {
 		drop_nlink(old_dir);
 		inc_nlink(new_dir);
@@ -822,7 +824,7 @@ ssize_t simple_attr_write(struct file *file, const char __user *buf,
 		goto out;
 
 	attr->set_buf[size] = '\0';
-	val = simple_strtol(attr->set_buf, NULL, 0);
+	val = simple_strtoll(attr->set_buf, NULL, 0);
 	ret = attr->set(attr->data, val);
 	if (ret == 0)
 		ret = len; /* on success, claim we got the whole input */
@@ -905,21 +907,29 @@ EXPORT_SYMBOL_GPL(generic_fh_to_parent);
  * filesystems which track all non-inode metadata in the buffers list
  * hanging off the address_space structure.
  */
-int generic_file_fsync(struct file *file, int datasync)
+int generic_file_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	int err;
 	int ret;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	ret = sync_mapping_buffers(inode->i_mapping);
 	if (!(inode->i_state & I_DIRTY))
-		return ret;
+		goto out;
 	if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
-		return ret;
+		goto out;
 
 	err = sync_inode_metadata(inode, 1);
 	if (ret == 0)
 		ret = err;
+out:
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 EXPORT_SYMBOL(generic_file_fsync);
@@ -956,7 +966,7 @@ EXPORT_SYMBOL(generic_check_addressable);
 /*
  * No-op implementation of ->fsync for in-memory filesystems.
  */
-int noop_fsync(struct file *file, int datasync)
+int noop_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	return 0;
 }
diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
index adb45ec9038..e374050a911 100644
--- a/fs/lockd/clntproc.c
+++ b/fs/lockd/clntproc.c
@@ -708,7 +708,13 @@ static void nlmclnt_unlock_callback(struct rpc_task *task, void *data)
 
 	if (task->tk_status < 0) {
 		dprintk("lockd: unlock failed (err = %d)\n", -task->tk_status);
-		goto retry_rebind;
+		switch (task->tk_status) {
+		case -EACCES:
+		case -EIO:
+			goto die;
+		default:
+			goto retry_rebind;
+		}
 	}
 	if (status == NLM_LCK_DENIED_GRACE_PERIOD) {
 		rpc_delay(task, NLMCLNT_GRACE_WAIT);
diff --git a/fs/locks.c b/fs/locks.c
index 0a4f50dfadf..b286539d547 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -160,10 +160,28 @@ EXPORT_SYMBOL_GPL(unlock_flocks);
 
 static struct kmem_cache *filelock_cache __read_mostly;
 
+static void locks_init_lock_always(struct file_lock *fl)
+{
+	fl->fl_next = NULL;
+	fl->fl_fasync = NULL;
+	fl->fl_owner = NULL;
+	fl->fl_pid = 0;
+	fl->fl_nspid = NULL;
+	fl->fl_file = NULL;
+	fl->fl_flags = 0;
+	fl->fl_type = 0;
+	fl->fl_start = fl->fl_end = 0;
+}
+
 /* Allocate an empty lock structure. */
 struct file_lock *locks_alloc_lock(void)
 {
-	return kmem_cache_alloc(filelock_cache, GFP_KERNEL);
+	struct file_lock *fl = kmem_cache_alloc(filelock_cache, GFP_KERNEL);
+
+	if (fl)
+		locks_init_lock_always(fl);
+
+	return fl;
 }
 EXPORT_SYMBOL_GPL(locks_alloc_lock);
 
@@ -200,17 +218,9 @@ void locks_init_lock(struct file_lock *fl)
 	INIT_LIST_HEAD(&fl->fl_link);
 	INIT_LIST_HEAD(&fl->fl_block);
 	init_waitqueue_head(&fl->fl_wait);
-	fl->fl_next = NULL;
-	fl->fl_fasync = NULL;
-	fl->fl_owner = NULL;
-	fl->fl_pid = 0;
-	fl->fl_nspid = NULL;
-	fl->fl_file = NULL;
-	fl->fl_flags = 0;
-	fl->fl_type = 0;
-	fl->fl_start = fl->fl_end = 0;
 	fl->fl_ops = NULL;
 	fl->fl_lmops = NULL;
+	locks_init_lock_always(fl);
 }
 
 EXPORT_SYMBOL(locks_init_lock);
diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c
index 9ed89d1663f..b3ff3d89416 100644
--- a/fs/logfs/dir.c
+++ b/fs/logfs/dir.c
@@ -371,11 +371,9 @@ static struct dentry *logfs_lookup(struct inode *dir, struct dentry *dentry,
 	page_cache_release(page);
 
 	inode = logfs_iget(dir->i_sb, ino);
-	if (IS_ERR(inode)) {
+	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);
 }
 
@@ -555,13 +553,6 @@ static int logfs_symlink(struct inode *dir, struct dentry *dentry,
 	return __logfs_create(dir, dentry, inode, target, destlen);
 }
 
-static int logfs_permission(struct inode *inode, int mask, unsigned int flags)
-{
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-	return generic_permission(inode, mask, flags, NULL);
-}
-
 static int logfs_link(struct dentry *old_dentry, struct inode *dir,
 		struct dentry *dentry)
 {
@@ -820,7 +811,6 @@ const struct inode_operations logfs_dir_iops = {
 	.mknod		= logfs_mknod,
 	.rename		= logfs_rename,
 	.rmdir		= logfs_rmdir,
-	.permission	= logfs_permission,
 	.symlink	= logfs_symlink,
 	.unlink		= logfs_unlink,
 };
diff --git a/fs/logfs/file.c b/fs/logfs/file.c
index c2ad7028def..b548c87a86f 100644
--- a/fs/logfs/file.c
+++ b/fs/logfs/file.c
@@ -219,11 +219,20 @@ long logfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	}
 }
 
-int logfs_fsync(struct file *file, int datasync)
+int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct super_block *sb = file->f_mapping->host->i_sb;
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
 
+	mutex_lock(&inode->i_mutex);
 	logfs_write_anchor(sb);
+	mutex_unlock(&inode->i_mutex);
+
 	return 0;
 }
 
diff --git a/fs/logfs/logfs.h b/fs/logfs/logfs.h
index 57afd4a6fab..f22d108bfa5 100644
--- a/fs/logfs/logfs.h
+++ b/fs/logfs/logfs.h
@@ -506,7 +506,7 @@ 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);
 long logfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
-int logfs_fsync(struct file *file, int datasync);
+int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync);
 
 /* gc.c */
 u32 get_best_cand(struct super_block *sb, struct candidate_list *list, u32 *ec);
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index adcdc0a4e18..e7d23e25bf1 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -596,8 +596,7 @@ static int minix_write_inode(struct inode *inode, struct writeback_control *wbc)
 
 int minix_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 {
-	struct inode *dir = dentry->d_parent->d_inode;
-	struct super_block *sb = dir->i_sb;
+	struct super_block *sb = dentry->d_sb;
 	generic_fillattr(dentry->d_inode, stat);
 	if (INODE_VERSION(dentry->d_inode) == MINIX_V1)
 		stat->blocks = (BLOCK_SIZE / 512) * V1_minix_blocks(stat->size, sb);
diff --git a/fs/namei.c b/fs/namei.c
index e2e4e8d032e..b7fad009bbf 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -176,12 +176,12 @@ EXPORT_SYMBOL(putname);
 /*
  * This does basic POSIX ACL permission checking
  */
-static int acl_permission_check(struct inode *inode, int mask, unsigned int flags,
-		int (*check_acl)(struct inode *inode, int mask, unsigned int flags))
+static int acl_permission_check(struct inode *inode, int mask)
 {
+	int (*check_acl)(struct inode *inode, int mask);
 	unsigned int mode = inode->i_mode;
 
-	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
+	mask &= MAY_READ | MAY_WRITE | MAY_EXEC | MAY_NOT_BLOCK;
 
 	if (current_user_ns() != inode_userns(inode))
 		goto other_perms;
@@ -189,8 +189,9 @@ static int acl_permission_check(struct inode *inode, int mask, unsigned int flag
 	if (current_fsuid() == inode->i_uid)
 		mode >>= 6;
 	else {
+		check_acl = inode->i_op->check_acl;
 		if (IS_POSIXACL(inode) && (mode & S_IRWXG) && check_acl) {
-			int error = check_acl(inode, mask, flags);
+			int error = check_acl(inode, mask);
 			if (error != -EAGAIN)
 				return error;
 		}
@@ -203,7 +204,7 @@ other_perms:
 	/*
 	 * If the DACs are ok we don't need any capability check.
 	 */
-	if ((mask & ~mode) == 0)
+	if ((mask & ~mode & (MAY_READ | MAY_WRITE | MAY_EXEC)) == 0)
 		return 0;
 	return -EACCES;
 }
@@ -212,8 +213,6 @@ other_perms:
  * generic_permission -  check for access rights on a Posix-like filesystem
  * @inode:	inode to check access rights for
  * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
- * @check_acl:	optional callback to check for Posix ACLs
- * @flags:	IPERM_FLAG_ flags.
  *
  * Used to check for read/write/execute permissions on a file.
  * We use "fsuid" for this, letting us set arbitrary permissions
@@ -224,23 +223,32 @@ other_perms:
  * request cannot be satisfied (eg. requires blocking or too much complexity).
  * It would then be called again in ref-walk mode.
  */
-int generic_permission(struct inode *inode, int mask, unsigned int flags,
-	int (*check_acl)(struct inode *inode, int mask, unsigned int flags))
+int generic_permission(struct inode *inode, int mask)
 {
 	int ret;
 
 	/*
 	 * Do the basic POSIX ACL permission checks.
 	 */
-	ret = acl_permission_check(inode, mask, flags, check_acl);
+	ret = acl_permission_check(inode, mask);
 	if (ret != -EACCES)
 		return ret;
 
+	if (S_ISDIR(inode->i_mode)) {
+		/* DACs are overridable for directories */
+		if (ns_capable(inode_userns(inode), CAP_DAC_OVERRIDE))
+			return 0;
+		if (!(mask & MAY_WRITE))
+			if (ns_capable(inode_userns(inode), CAP_DAC_READ_SEARCH))
+				return 0;
+		return -EACCES;
+	}
 	/*
 	 * Read/write DACs are always overridable.
-	 * Executable DACs are overridable if at least one exec bit is set.
+	 * Executable DACs are overridable when there is
+	 * at least one exec bit set.
 	 */
-	if (!(mask & MAY_EXEC) || execute_ok(inode))
+	if (!(mask & MAY_EXEC) || (inode->i_mode & S_IXUGO))
 		if (ns_capable(inode_userns(inode), CAP_DAC_OVERRIDE))
 			return 0;
 
@@ -248,7 +256,7 @@ int generic_permission(struct inode *inode, int mask, unsigned int flags,
 	 * Searching includes executable on directories, else just read.
 	 */
 	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
-	if (mask == MAY_READ || (S_ISDIR(inode->i_mode) && !(mask & MAY_WRITE)))
+	if (mask == MAY_READ)
 		if (ns_capable(inode_userns(inode), CAP_DAC_READ_SEARCH))
 			return 0;
 
@@ -287,10 +295,9 @@ int inode_permission(struct inode *inode, int mask)
 	}
 
 	if (inode->i_op->permission)
-		retval = inode->i_op->permission(inode, mask, 0);
+		retval = inode->i_op->permission(inode, mask);
 	else
-		retval = generic_permission(inode, mask, 0,
-				inode->i_op->check_acl);
+		retval = generic_permission(inode, mask);
 
 	if (retval)
 		return retval;
@@ -303,69 +310,6 @@ int inode_permission(struct inode *inode, int mask)
 }
 
 /**
- * file_permission  -  check for additional access rights to a given file
- * @file:	file to check access rights for
- * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
- *
- * Used to check for read/write/execute permissions on an already opened
- * file.
- *
- * Note:
- *	Do not use this function in new code.  All access checks should
- *	be done using inode_permission().
- */
-int file_permission(struct file *file, int mask)
-{
-	return inode_permission(file->f_path.dentry->d_inode, mask);
-}
-
-/*
- * get_write_access() gets write permission for a file.
- * put_write_access() releases this write permission.
- * This is used for regular files.
- * We cannot support write (and maybe mmap read-write shared) accesses and
- * MAP_DENYWRITE mmappings simultaneously. The i_writecount field of an inode
- * can have the following values:
- * 0: no writers, no VM_DENYWRITE mappings
- * < 0: (-i_writecount) vm_area_structs with VM_DENYWRITE set exist
- * > 0: (i_writecount) users are writing to the file.
- *
- * Normally we operate on that counter with atomic_{inc,dec} and it's safe
- * except for the cases where we don't hold i_writecount yet. Then we need to
- * use {get,deny}_write_access() - these functions check the sign and refuse
- * to do the change if sign is wrong. Exclusion between them is provided by
- * the inode->i_lock spinlock.
- */
-
-int get_write_access(struct inode * inode)
-{
-	spin_lock(&inode->i_lock);
-	if (atomic_read(&inode->i_writecount) < 0) {
-		spin_unlock(&inode->i_lock);
-		return -ETXTBSY;
-	}
-	atomic_inc(&inode->i_writecount);
-	spin_unlock(&inode->i_lock);
-
-	return 0;
-}
-
-int deny_write_access(struct file * file)
-{
-	struct inode *inode = file->f_path.dentry->d_inode;
-
-	spin_lock(&inode->i_lock);
-	if (atomic_read(&inode->i_writecount) > 0) {
-		spin_unlock(&inode->i_lock);
-		return -ETXTBSY;
-	}
-	atomic_dec(&inode->i_writecount);
-	spin_unlock(&inode->i_lock);
-
-	return 0;
-}
-
-/**
  * path_get - get a reference to a path
  * @path: path to get the reference to
  *
@@ -432,6 +376,8 @@ static int unlazy_walk(struct nameidata *nd, struct dentry *dentry)
 			goto err_parent;
 		BUG_ON(nd->inode != parent->d_inode);
 	} else {
+		if (dentry->d_parent != parent)
+			goto err_parent;
 		spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED);
 		if (!__d_rcu_to_refcount(dentry, nd->seq))
 			goto err_child;
@@ -489,28 +435,6 @@ static inline int d_revalidate(struct dentry *dentry, struct nameidata *nd)
 	return dentry->d_op->d_revalidate(dentry, nd);
 }
 
-static struct dentry *
-do_revalidate(struct dentry *dentry, struct nameidata *nd)
-{
-	int status = d_revalidate(dentry, nd);
-	if (unlikely(status <= 0)) {
-		/*
-		 * The dentry failed validation.
-		 * If d_revalidate returned 0 attempt to invalidate
-		 * the dentry otherwise d_revalidate is asking us
-		 * to return a fail status.
-		 */
-		if (status < 0) {
-			dput(dentry);
-			dentry = ERR_PTR(status);
-		} else if (!d_invalidate(dentry)) {
-			dput(dentry);
-			dentry = NULL;
-		}
-	}
-	return dentry;
-}
-
 /**
  * complete_walk - successful completion of path walk
  * @nd:  pointer nameidata
@@ -565,40 +489,6 @@ static int complete_walk(struct nameidata *nd)
 	return status;
 }
 
-/*
- * Short-cut version of permission(), for calling on directories
- * during pathname resolution.  Combines parts of permission()
- * and generic_permission(), and tests ONLY for MAY_EXEC permission.
- *
- * If appropriate, check DAC only.  If not appropriate, or
- * short-cut DAC fails, then call ->permission() to do more
- * complete permission check.
- */
-static inline int exec_permission(struct inode *inode, unsigned int flags)
-{
-	int ret;
-	struct user_namespace *ns = inode_userns(inode);
-
-	if (inode->i_op->permission) {
-		ret = inode->i_op->permission(inode, MAY_EXEC, flags);
-	} else {
-		ret = acl_permission_check(inode, MAY_EXEC, flags,
-				inode->i_op->check_acl);
-	}
-	if (likely(!ret))
-		goto ok;
-	if (ret == -ECHILD)
-		return ret;
-
-	if (ns_capable(ns, CAP_DAC_OVERRIDE) ||
-			ns_capable(ns, CAP_DAC_READ_SEARCH))
-		goto ok;
-
-	return ret;
-ok:
-	return security_inode_exec_permission(inode, flags);
-}
-
 static __always_inline void set_root(struct nameidata *nd)
 {
 	if (!nd->root.mnt)
@@ -773,7 +663,7 @@ static int follow_automount(struct path *path, unsigned flags,
 	/* We don't want to mount if someone supplied AT_NO_AUTOMOUNT
 	 * and this is the terminal part of the path.
 	 */
-	if ((flags & LOOKUP_NO_AUTOMOUNT) && !(flags & LOOKUP_CONTINUE))
+	if ((flags & LOOKUP_NO_AUTOMOUNT) && !(flags & LOOKUP_PARENT))
 		return -EISDIR; /* we actually want to stop here */
 
 	/* We want to mount if someone is trying to open/create a file of any
@@ -785,7 +675,7 @@ static int follow_automount(struct path *path, unsigned flags,
 	 * appended a '/' to the name.
 	 */
 	if (!(flags & LOOKUP_FOLLOW) &&
-	    !(flags & (LOOKUP_CONTINUE | LOOKUP_DIRECTORY |
+	    !(flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY |
 		       LOOKUP_OPEN | LOOKUP_CREATE)))
 		return -EISDIR;
 
@@ -804,7 +694,7 @@ static int follow_automount(struct path *path, unsigned flags,
 		 * the path being looked up; if it wasn't then the remainder of
 		 * the path is inaccessible and we should say so.
 		 */
-		if (PTR_ERR(mnt) == -EISDIR && (flags & LOOKUP_CONTINUE))
+		if (PTR_ERR(mnt) == -EISDIR && (flags & LOOKUP_PARENT))
 			return -EREMOTE;
 		return PTR_ERR(mnt);
 	}
@@ -812,6 +702,11 @@ static int follow_automount(struct path *path, unsigned flags,
 	if (!mnt) /* mount collision */
 		return 0;
 
+	if (!*need_mntput) {
+		/* lock_mount() may release path->mnt on error */
+		mntget(path->mnt);
+		*need_mntput = true;
+	}
 	err = finish_automount(mnt, path);
 
 	switch (err) {
@@ -819,12 +714,9 @@ static int follow_automount(struct path *path, unsigned flags,
 		/* Someone else made a mount here whilst we were busy */
 		return 0;
 	case 0:
-		dput(path->dentry);
-		if (*need_mntput)
-			mntput(path->mnt);
+		path_put(path);
 		path->mnt = mnt;
 		path->dentry = dget(mnt->mnt_root);
-		*need_mntput = true;
 		return 0;
 	default:
 		return err;
@@ -844,9 +736,10 @@ static int follow_automount(struct path *path, unsigned flags,
  */
 static int follow_managed(struct path *path, unsigned flags)
 {
+	struct vfsmount *mnt = path->mnt; /* held by caller, must be left alone */
 	unsigned managed;
 	bool need_mntput = false;
-	int ret;
+	int ret = 0;
 
 	/* Given that we're not holding a lock here, we retain the value in a
 	 * local variable for each dentry as we look at it so that we don't see
@@ -861,7 +754,7 @@ static int follow_managed(struct path *path, unsigned flags)
 			BUG_ON(!path->dentry->d_op->d_manage);
 			ret = path->dentry->d_op->d_manage(path->dentry, false);
 			if (ret < 0)
-				return ret == -EISDIR ? 0 : ret;
+				break;
 		}
 
 		/* Transit to a mounted filesystem. */
@@ -887,14 +780,19 @@ static int follow_managed(struct path *path, unsigned flags)
 		if (managed & DCACHE_NEED_AUTOMOUNT) {
 			ret = follow_automount(path, flags, &need_mntput);
 			if (ret < 0)
-				return ret == -EISDIR ? 0 : ret;
+				break;
 			continue;
 		}
 
 		/* We didn't change the current path point */
 		break;
 	}
-	return 0;
+
+	if (need_mntput && path->mnt == mnt)
+		mntput(path->mnt);
+	if (ret == -EISDIR)
+		ret = 0;
+	return ret;
 }
 
 int follow_down_one(struct path *path)
@@ -931,7 +829,6 @@ static bool __follow_mount_rcu(struct nameidata *nd, struct path *path,
 		 * Don't forget we might have a non-mountpoint managed dentry
 		 * that wants to block transit.
 		 */
-		*inode = path->dentry->d_inode;
 		if (unlikely(managed_dentry_might_block(path->dentry)))
 			return false;
 
@@ -944,6 +841,12 @@ static bool __follow_mount_rcu(struct nameidata *nd, struct path *path,
 		path->mnt = mounted;
 		path->dentry = mounted->mnt_root;
 		nd->seq = read_seqcount_begin(&path->dentry->d_seq);
+		/*
+		 * Update the inode too. We don't need to re-check the
+		 * dentry sequence number here after this d_inode read,
+		 * because a mount-point is always pinned.
+		 */
+		*inode = path->dentry->d_inode;
 	}
 	return true;
 }
@@ -1003,9 +906,6 @@ failed:
  * Follow down to the covering mount currently visible to userspace.  At each
  * point, the filesystem owning that dentry may be queried as to whether the
  * caller is permitted to proceed or not.
- *
- * Care must be taken as namespace_sem may be held (indicated by mounting_here
- * being true).
  */
 int follow_down(struct path *path)
 {
@@ -1121,6 +1021,30 @@ static struct dentry *d_alloc_and_lookup(struct dentry *parent,
 }
 
 /*
+ * We already have a dentry, but require a lookup to be performed on the parent
+ * directory to fill in d_inode. Returns the new dentry, or ERR_PTR on error.
+ * parent->d_inode->i_mutex must be held. d_lookup must have verified that no
+ * child exists while under i_mutex.
+ */
+static struct dentry *d_inode_lookup(struct dentry *parent, struct dentry *dentry,
+				     struct nameidata *nd)
+{
+	struct inode *inode = parent->d_inode;
+	struct dentry *old;
+
+	/* Don't create child dentry for a dead directory. */
+	if (unlikely(IS_DEADDIR(inode)))
+		return ERR_PTR(-ENOENT);
+
+	old = inode->i_op->lookup(inode, dentry, nd);
+	if (unlikely(old)) {
+		dput(dentry);
+		dentry = old;
+	}
+	return dentry;
+}
+
+/*
  *  It's more convoluted than I'd like it to be, but... it's still fairly
  *  small and for now I'd prefer to have fast path as straight as possible.
  *  It _is_ time-critical.
@@ -1159,6 +1083,8 @@ static int do_lookup(struct nameidata *nd, struct qstr *name,
 				goto unlazy;
 			}
 		}
+		if (unlikely(d_need_lookup(dentry)))
+			goto unlazy;
 		path->mnt = mnt;
 		path->dentry = dentry;
 		if (unlikely(!__follow_mount_rcu(nd, path, inode)))
@@ -1173,6 +1099,10 @@ unlazy:
 		dentry = __d_lookup(parent, name);
 	}
 
+	if (dentry && unlikely(d_need_lookup(dentry))) {
+		dput(dentry);
+		dentry = NULL;
+	}
 retry:
 	if (unlikely(!dentry)) {
 		struct inode *dir = parent->d_inode;
@@ -1189,6 +1119,15 @@ retry:
 			/* known good */
 			need_reval = 0;
 			status = 1;
+		} else if (unlikely(d_need_lookup(dentry))) {
+			dentry = d_inode_lookup(parent, dentry, nd);
+			if (IS_ERR(dentry)) {
+				mutex_unlock(&dir->i_mutex);
+				return PTR_ERR(dentry);
+			}
+			/* known good */
+			need_reval = 0;
+			status = 1;
 		}
 		mutex_unlock(&dir->i_mutex);
 	}
@@ -1221,13 +1160,13 @@ retry:
 static inline int may_lookup(struct nameidata *nd)
 {
 	if (nd->flags & LOOKUP_RCU) {
-		int err = exec_permission(nd->inode, IPERM_FLAG_RCU);
+		int err = inode_permission(nd->inode, MAY_EXEC|MAY_NOT_BLOCK);
 		if (err != -ECHILD)
 			return err;
 		if (unlazy_walk(nd, NULL))
 			return -ECHILD;
 	}
-	return exec_permission(nd->inode, 0);
+	return inode_permission(nd->inode, MAY_EXEC);
 }
 
 static inline int handle_dots(struct nameidata *nd, int type)
@@ -1341,7 +1280,6 @@ static int link_path_walk(const char *name, struct nameidata *nd)
 {
 	struct path next;
 	int err;
-	unsigned int lookup_flags = nd->flags;
 	
 	while (*name=='/')
 		name++;
@@ -1355,8 +1293,6 @@ static int link_path_walk(const char *name, struct nameidata *nd)
 		unsigned int c;
 		int type;
 
-		nd->flags |= LOOKUP_CONTINUE;
-
 		err = may_lookup(nd);
  		if (err)
 			break;
@@ -1418,8 +1354,6 @@ static int link_path_walk(const char *name, struct nameidata *nd)
 		/* here ends the main loop */
 
 last_component:
-		/* Clear LOOKUP_CONTINUE iff it was previously unset */
-		nd->flags &= lookup_flags | ~LOOKUP_CONTINUE;
 		nd->last = this;
 		nd->last_type = type;
 		return 0;
@@ -1502,7 +1436,7 @@ static int path_init(int dfd, const char *name, unsigned int flags,
 			if (!S_ISDIR(dentry->d_inode->i_mode))
 				goto fput_fail;
 
-			retval = file_permission(file, MAY_EXEC);
+			retval = inode_permission(dentry->d_inode, MAY_EXEC);
 			if (retval)
 				goto fput_fail;
 		}
@@ -1640,16 +1574,22 @@ int kern_path(const char *name, unsigned int flags, struct path *path)
  * @mnt: pointer to vfs mount of the base directory
  * @name: pointer to file name
  * @flags: lookup flags
- * @nd: pointer to nameidata
+ * @path: pointer to struct path to fill
  */
 int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt,
 		    const char *name, unsigned int flags,
-		    struct nameidata *nd)
+		    struct path *path)
 {
-	nd->root.dentry = dentry;
-	nd->root.mnt = mnt;
+	struct nameidata nd;
+	int err;
+	nd.root.dentry = dentry;
+	nd.root.mnt = mnt;
+	BUG_ON(flags & LOOKUP_PARENT);
 	/* the first argument of do_path_lookup() is ignored with LOOKUP_ROOT */
-	return do_path_lookup(AT_FDCWD, name, flags | LOOKUP_ROOT, nd);
+	err = do_path_lookup(AT_FDCWD, name, flags | LOOKUP_ROOT, &nd);
+	if (!err)
+		*path = nd.path;
+	return err;
 }
 
 static struct dentry *__lookup_hash(struct qstr *name,
@@ -1659,7 +1599,7 @@ static struct dentry *__lookup_hash(struct qstr *name,
 	struct dentry *dentry;
 	int err;
 
-	err = exec_permission(inode, 0);
+	err = inode_permission(inode, MAY_EXEC);
 	if (err)
 		return ERR_PTR(err);
 
@@ -1670,8 +1610,34 @@ static struct dentry *__lookup_hash(struct qstr *name,
 	 */
 	dentry = d_lookup(base, name);
 
-	if (dentry && (dentry->d_flags & DCACHE_OP_REVALIDATE))
-		dentry = do_revalidate(dentry, nd);
+	if (dentry && d_need_lookup(dentry)) {
+		/*
+		 * __lookup_hash is called with the parent dir's i_mutex already
+		 * held, so we are good to go here.
+		 */
+		dentry = d_inode_lookup(base, dentry, nd);
+		if (IS_ERR(dentry))
+			return dentry;
+	}
+
+	if (dentry && (dentry->d_flags & DCACHE_OP_REVALIDATE)) {
+		int status = d_revalidate(dentry, nd);
+		if (unlikely(status <= 0)) {
+			/*
+			 * The dentry failed validation.
+			 * If d_revalidate returned 0 attempt to invalidate
+			 * the dentry otherwise d_revalidate is asking us
+			 * to return a fail status.
+			 */
+			if (status < 0) {
+				dput(dentry);
+				return ERR_PTR(status);
+			} else if (!d_invalidate(dentry)) {
+				dput(dentry);
+				dentry = NULL;
+			}
+		}
+	}
 
 	if (!dentry)
 		dentry = d_alloc_and_lookup(base, name, nd);
@@ -1999,27 +1965,10 @@ static int handle_truncate(struct file *filp)
 	return error;
 }
 
-/*
- * Note that while the flag value (low two bits) for sys_open means:
- *	00 - read-only
- *	01 - write-only
- *	10 - read-write
- *	11 - special
- * it is changed into
- *	00 - no permissions needed
- *	01 - read-permission
- *	10 - write-permission
- *	11 - read-write
- * for the internal routines (ie open_namei()/follow_link() etc)
- * This is more logical, and also allows the 00 "no perm needed"
- * to be used for symlinks (where the permissions are checked
- * later).
- *
-*/
 static inline int open_to_namei_flags(int flag)
 {
-	if ((flag+1) & O_ACCMODE)
-		flag++;
+	if ((flag & O_ACCMODE) == 3)
+		flag--;
 	return flag;
 }
 
@@ -2314,35 +2263,29 @@ struct file *do_file_open_root(struct dentry *dentry, struct vfsmount *mnt,
 	return file;
 }
 
-/**
- * lookup_create - lookup a dentry, creating it if it doesn't exist
- * @nd: nameidata info
- * @is_dir: directory flag
- *
- * Simple function to lookup and return a dentry and create it
- * if it doesn't exist.  Is SMP-safe.
- *
- * Returns with nd->path.dentry->d_inode->i_mutex locked.
- */
-struct dentry *lookup_create(struct nameidata *nd, int is_dir)
+struct dentry *kern_path_create(int dfd, const char *pathname, struct path *path, int is_dir)
 {
 	struct dentry *dentry = ERR_PTR(-EEXIST);
+	struct nameidata nd;
+	int error = do_path_lookup(dfd, pathname, LOOKUP_PARENT, &nd);
+	if (error)
+		return ERR_PTR(error);
 
-	mutex_lock_nested(&nd->path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
 	/*
 	 * Yucky last component or no last component at all?
 	 * (foo/., foo/.., /////)
 	 */
-	if (nd->last_type != LAST_NORM)
-		goto fail;
-	nd->flags &= ~LOOKUP_PARENT;
-	nd->flags |= LOOKUP_CREATE | LOOKUP_EXCL;
-	nd->intent.open.flags = O_EXCL;
+	if (nd.last_type != LAST_NORM)
+		goto out;
+	nd.flags &= ~LOOKUP_PARENT;
+	nd.flags |= LOOKUP_CREATE | LOOKUP_EXCL;
+	nd.intent.open.flags = O_EXCL;
 
 	/*
 	 * Do the final lookup.
 	 */
-	dentry = lookup_hash(nd);
+	mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
+	dentry = lookup_hash(&nd);
 	if (IS_ERR(dentry))
 		goto fail;
 
@@ -2354,18 +2297,35 @@ struct dentry *lookup_create(struct nameidata *nd, int is_dir)
 	 * all is fine. Let's be bastards - you had / on the end, you've
 	 * been asking for (non-existent) directory. -ENOENT for you.
 	 */
-	if (unlikely(!is_dir && nd->last.name[nd->last.len])) {
+	if (unlikely(!is_dir && nd.last.name[nd.last.len])) {
 		dput(dentry);
 		dentry = ERR_PTR(-ENOENT);
+		goto fail;
 	}
+	*path = nd.path;
 	return dentry;
 eexist:
 	dput(dentry);
 	dentry = ERR_PTR(-EEXIST);
 fail:
+	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
+out:
+	path_put(&nd.path);
 	return dentry;
 }
-EXPORT_SYMBOL_GPL(lookup_create);
+EXPORT_SYMBOL(kern_path_create);
+
+struct dentry *user_path_create(int dfd, const char __user *pathname, struct path *path, int is_dir)
+{
+	char *tmp = getname(pathname);
+	struct dentry *res;
+	if (IS_ERR(tmp))
+		return ERR_CAST(tmp);
+	res = kern_path_create(dfd, tmp, path, is_dir);
+	putname(tmp);
+	return res;
+}
+EXPORT_SYMBOL(user_path_create);
 
 int vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev)
 {
@@ -2415,54 +2375,46 @@ static int may_mknod(mode_t mode)
 SYSCALL_DEFINE4(mknodat, int, dfd, const char __user *, filename, int, mode,
 		unsigned, dev)
 {
-	int error;
-	char *tmp;
 	struct dentry *dentry;
-	struct nameidata nd;
+	struct path path;
+	int error;
 
 	if (S_ISDIR(mode))
 		return -EPERM;
 
-	error = user_path_parent(dfd, filename, &nd, &tmp);
-	if (error)
-		return error;
+	dentry = user_path_create(dfd, filename, &path, 0);
+	if (IS_ERR(dentry))
+		return PTR_ERR(dentry);
 
-	dentry = lookup_create(&nd, 0);
-	if (IS_ERR(dentry)) {
-		error = PTR_ERR(dentry);
-		goto out_unlock;
-	}
-	if (!IS_POSIXACL(nd.path.dentry->d_inode))
+	if (!IS_POSIXACL(path.dentry->d_inode))
 		mode &= ~current_umask();
 	error = may_mknod(mode);
 	if (error)
 		goto out_dput;
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_mknod(&nd.path, dentry, mode, dev);
+	error = security_path_mknod(&path, dentry, mode, dev);
 	if (error)
 		goto out_drop_write;
 	switch (mode & S_IFMT) {
 		case 0: case S_IFREG:
-			error = vfs_create(nd.path.dentry->d_inode,dentry,mode,&nd);
+			error = vfs_create(path.dentry->d_inode,dentry,mode,NULL);
 			break;
 		case S_IFCHR: case S_IFBLK:
-			error = vfs_mknod(nd.path.dentry->d_inode,dentry,mode,
+			error = vfs_mknod(path.dentry->d_inode,dentry,mode,
 					new_decode_dev(dev));
 			break;
 		case S_IFIFO: case S_IFSOCK:
-			error = vfs_mknod(nd.path.dentry->d_inode,dentry,mode,0);
+			error = vfs_mknod(path.dentry->d_inode,dentry,mode,0);
 			break;
 	}
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(path.mnt);
 out_dput:
 	dput(dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-	path_put(&nd.path);
-	putname(tmp);
+	mutex_unlock(&path.dentry->d_inode->i_mutex);
+	path_put(&path);
 
 	return error;
 }
@@ -2495,38 +2447,29 @@ int vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 
 SYSCALL_DEFINE3(mkdirat, int, dfd, const char __user *, pathname, int, mode)
 {
-	int error = 0;
-	char * tmp;
 	struct dentry *dentry;
-	struct nameidata nd;
-
-	error = user_path_parent(dfd, pathname, &nd, &tmp);
-	if (error)
-		goto out_err;
+	struct path path;
+	int error;
 
-	dentry = lookup_create(&nd, 1);
-	error = PTR_ERR(dentry);
+	dentry = user_path_create(dfd, pathname, &path, 1);
 	if (IS_ERR(dentry))
-		goto out_unlock;
+		return PTR_ERR(dentry);
 
-	if (!IS_POSIXACL(nd.path.dentry->d_inode))
+	if (!IS_POSIXACL(path.dentry->d_inode))
 		mode &= ~current_umask();
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_mkdir(&nd.path, dentry, mode);
+	error = security_path_mkdir(&path, dentry, mode);
 	if (error)
 		goto out_drop_write;
-	error = vfs_mkdir(nd.path.dentry->d_inode, dentry, mode);
+	error = vfs_mkdir(path.dentry->d_inode, dentry, mode);
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(path.mnt);
 out_dput:
 	dput(dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-	path_put(&nd.path);
-	putname(tmp);
-out_err:
+	mutex_unlock(&path.dentry->d_inode->i_mutex);
+	path_put(&path);
 	return error;
 }
 
@@ -2624,6 +2567,10 @@ static long do_rmdir(int dfd, const char __user *pathname)
 	error = PTR_ERR(dentry);
 	if (IS_ERR(dentry))
 		goto exit2;
+	if (!dentry->d_inode) {
+		error = -ENOENT;
+		goto exit3;
+	}
 	error = mnt_want_write(nd.path.mnt);
 	if (error)
 		goto exit3;
@@ -2712,8 +2659,9 @@ static long do_unlinkat(int dfd, const char __user *pathname)
 		if (nd.last.name[nd.last.len])
 			goto slashes;
 		inode = dentry->d_inode;
-		if (inode)
-			ihold(inode);
+		if (!inode)
+			goto slashes;
+		ihold(inode);
 		error = mnt_want_write(nd.path.mnt);
 		if (error)
 			goto exit2;
@@ -2781,38 +2729,31 @@ SYSCALL_DEFINE3(symlinkat, const char __user *, oldname,
 {
 	int error;
 	char *from;
-	char *to;
 	struct dentry *dentry;
-	struct nameidata nd;
+	struct path path;
 
 	from = getname(oldname);
 	if (IS_ERR(from))
 		return PTR_ERR(from);
 
-	error = user_path_parent(newdfd, newname, &nd, &to);
-	if (error)
-		goto out_putname;
-
-	dentry = lookup_create(&nd, 0);
+	dentry = user_path_create(newdfd, newname, &path, 0);
 	error = PTR_ERR(dentry);
 	if (IS_ERR(dentry))
-		goto out_unlock;
+		goto out_putname;
 
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_symlink(&nd.path, dentry, from);
+	error = security_path_symlink(&path, dentry, from);
 	if (error)
 		goto out_drop_write;
-	error = vfs_symlink(nd.path.dentry->d_inode, dentry, from);
+	error = vfs_symlink(path.dentry->d_inode, dentry, from);
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(path.mnt);
 out_dput:
 	dput(dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-	path_put(&nd.path);
-	putname(to);
+	mutex_unlock(&path.dentry->d_inode->i_mutex);
+	path_put(&path);
 out_putname:
 	putname(from);
 	return error;
@@ -2877,11 +2818,9 @@ SYSCALL_DEFINE5(linkat, int, olddfd, const char __user *, oldname,
 		int, newdfd, const char __user *, newname, int, flags)
 {
 	struct dentry *new_dentry;
-	struct nameidata nd;
-	struct path old_path;
+	struct path old_path, new_path;
 	int how = 0;
 	int error;
-	char *to;
 
 	if ((flags & ~(AT_SYMLINK_FOLLOW | AT_EMPTY_PATH)) != 0)
 		return -EINVAL;
@@ -2903,32 +2842,27 @@ SYSCALL_DEFINE5(linkat, int, olddfd, const char __user *, oldname,
 	if (error)
 		return error;
 
-	error = user_path_parent(newdfd, newname, &nd, &to);
-	if (error)
-		goto out;
-	error = -EXDEV;
-	if (old_path.mnt != nd.path.mnt)
-		goto out_release;
-	new_dentry = lookup_create(&nd, 0);
+	new_dentry = user_path_create(newdfd, newname, &new_path, 0);
 	error = PTR_ERR(new_dentry);
 	if (IS_ERR(new_dentry))
-		goto out_unlock;
-	error = mnt_want_write(nd.path.mnt);
+		goto out;
+
+	error = -EXDEV;
+	if (old_path.mnt != new_path.mnt)
+		goto out_dput;
+	error = mnt_want_write(new_path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_link(old_path.dentry, &nd.path, new_dentry);
+	error = security_path_link(old_path.dentry, &new_path, new_dentry);
 	if (error)
 		goto out_drop_write;
-	error = vfs_link(old_path.dentry, nd.path.dentry->d_inode, new_dentry);
+	error = vfs_link(old_path.dentry, new_path.dentry->d_inode, new_dentry);
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(new_path.mnt);
 out_dput:
 	dput(new_dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-out_release:
-	path_put(&nd.path);
-	putname(to);
+	mutex_unlock(&new_path.dentry->d_inode->i_mutex);
+	path_put(&new_path);
 out:
 	path_put(&old_path);
 
@@ -3334,11 +3268,9 @@ EXPORT_SYMBOL(page_readlink);
 EXPORT_SYMBOL(__page_symlink);
 EXPORT_SYMBOL(page_symlink);
 EXPORT_SYMBOL(page_symlink_inode_operations);
-EXPORT_SYMBOL(kern_path_parent);
 EXPORT_SYMBOL(kern_path);
 EXPORT_SYMBOL(vfs_path_lookup);
 EXPORT_SYMBOL(inode_permission);
-EXPORT_SYMBOL(file_permission);
 EXPORT_SYMBOL(unlock_rename);
 EXPORT_SYMBOL(vfs_create);
 EXPORT_SYMBOL(vfs_follow_link);
diff --git a/fs/namespace.c b/fs/namespace.c
index fe59bd145d2..cda50fe9250 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -934,8 +934,8 @@ int mnt_had_events(struct proc_mounts *p)
 	int res = 0;
 
 	br_read_lock(vfsmount_lock);
-	if (p->event != ns->event) {
-		p->event = ns->event;
+	if (p->m.poll_event != ns->event) {
+		p->m.poll_event = ns->event;
 		res = 1;
 	}
 	br_read_unlock(vfsmount_lock);
diff --git a/fs/ncpfs/file.c b/fs/ncpfs/file.c
index 0ed65e0c3df..64a326418aa 100644
--- a/fs/ncpfs/file.c
+++ b/fs/ncpfs/file.c
@@ -20,9 +20,9 @@
 
 #include "ncp_fs.h"
 
-static int ncp_fsync(struct file *file, int datasync)
+static int ncp_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
-	return 0;
+	return filemap_write_and_wait_range(file->f_mapping, start, end);
 }
 
 /*
diff --git a/fs/nfs/cache_lib.c b/fs/nfs/cache_lib.c
index 84690319e62..c98b439332f 100644
--- a/fs/nfs/cache_lib.c
+++ b/fs/nfs/cache_lib.c
@@ -113,19 +113,18 @@ int nfs_cache_wait_for_upcall(struct nfs_cache_defer_req *dreq)
 
 int nfs_cache_register(struct cache_detail *cd)
 {
-	struct nameidata nd;
 	struct vfsmount *mnt;
+	struct path path;
 	int ret;
 
 	mnt = rpc_get_mount();
 	if (IS_ERR(mnt))
 		return PTR_ERR(mnt);
-	ret = vfs_path_lookup(mnt->mnt_root, mnt, "/cache", 0, &nd);
+	ret = vfs_path_lookup(mnt->mnt_root, mnt, "/cache", 0, &path);
 	if (ret)
 		goto err;
-	ret = sunrpc_cache_register_pipefs(nd.path.dentry,
-			cd->name, 0600, cd);
-	path_put(&nd.path);
+	ret = sunrpc_cache_register_pipefs(path.dentry, cd->name, 0600, cd);
+	path_put(&path);
 	if (!ret)
 		return ret;
 err:
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index ededdbd0db3..57f578e2560 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -56,7 +56,7 @@ static int nfs_link(struct dentry *, struct inode *, struct dentry *);
 static int nfs_mknod(struct inode *, struct dentry *, int, dev_t);
 static int nfs_rename(struct inode *, struct dentry *,
 		      struct inode *, struct dentry *);
-static int nfs_fsync_dir(struct file *, int);
+static int nfs_fsync_dir(struct file *, loff_t, loff_t, int);
 static loff_t nfs_llseek_dir(struct file *, loff_t, int);
 static void nfs_readdir_clear_array(struct page*);
 
@@ -945,15 +945,19 @@ out:
  * All directory operations under NFS are synchronous, so fsync()
  * is a dummy operation.
  */
-static int nfs_fsync_dir(struct file *filp, int datasync)
+static int nfs_fsync_dir(struct file *filp, loff_t start, loff_t end,
+			 int datasync)
 {
 	struct dentry *dentry = filp->f_path.dentry;
+	struct inode *inode = dentry->d_inode;
 
 	dfprintk(FILE, "NFS: fsync dir(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
+	mutex_lock(&inode->i_mutex);
 	nfs_inc_stats(dentry->d_inode, NFSIOS_VFSFSYNC);
+	mutex_unlock(&inode->i_mutex);
 	return 0;
 }
 
@@ -997,14 +1001,12 @@ static int nfs_check_verifier(struct inode *dir, struct dentry *dentry)
  * Return the intent data that applies to this particular path component
  *
  * Note that the current set of intents only apply to the very last
- * component of the path.
- * We check for this using LOOKUP_CONTINUE and LOOKUP_PARENT.
+ * component of the path and none of them is set before that last
+ * component.
  */
 static inline unsigned int nfs_lookup_check_intent(struct nameidata *nd,
 						unsigned int mask)
 {
-	if (nd->flags & (LOOKUP_CONTINUE|LOOKUP_PARENT))
-		return 0;
 	return nd->flags & mask;
 }
 
@@ -1338,25 +1340,31 @@ static int is_atomic_open(struct nameidata *nd)
 		return 0;
 	/* Are we trying to write to a read only partition? */
 	if (__mnt_is_readonly(nd->path.mnt) &&
-	    (nd->intent.open.flags & (O_CREAT|O_TRUNC|FMODE_WRITE)))
+	    (nd->intent.open.flags & (O_CREAT|O_TRUNC|O_ACCMODE)))
 		return 0;
 	return 1;
 }
 
-static struct nfs_open_context *nameidata_to_nfs_open_context(struct dentry *dentry, struct nameidata *nd)
+static fmode_t flags_to_mode(int flags)
+{
+	fmode_t res = (__force fmode_t)flags & FMODE_EXEC;
+	if ((flags & O_ACCMODE) != O_WRONLY)
+		res |= FMODE_READ;
+	if ((flags & O_ACCMODE) != O_RDONLY)
+		res |= FMODE_WRITE;
+	return res;
+}
+
+static struct nfs_open_context *create_nfs_open_context(struct dentry *dentry, int open_flags)
 {
-	struct path path = {
-		.mnt = nd->path.mnt,
-		.dentry = dentry,
-	};
 	struct nfs_open_context *ctx;
 	struct rpc_cred *cred;
-	fmode_t fmode = nd->intent.open.flags & (FMODE_READ | FMODE_WRITE | FMODE_EXEC);
+	fmode_t fmode = flags_to_mode(open_flags);
 
 	cred = rpc_lookup_cred();
 	if (IS_ERR(cred))
 		return ERR_CAST(cred);
-	ctx = alloc_nfs_open_context(&path, cred, fmode);
+	ctx = alloc_nfs_open_context(dentry, cred, fmode);
 	put_rpccred(cred);
 	if (ctx == NULL)
 		return ERR_PTR(-ENOMEM);
@@ -1376,13 +1384,13 @@ static int nfs_intent_set_file(struct nameidata *nd, struct nfs_open_context *ct
 
 	/* If the open_intent is for execute, we have an extra check to make */
 	if (ctx->mode & FMODE_EXEC) {
-		ret = nfs_may_open(ctx->path.dentry->d_inode,
+		ret = nfs_may_open(ctx->dentry->d_inode,
 				ctx->cred,
 				nd->intent.open.flags);
 		if (ret < 0)
 			goto out;
 	}
-	filp = lookup_instantiate_filp(nd, ctx->path.dentry, do_open);
+	filp = lookup_instantiate_filp(nd, ctx->dentry, do_open);
 	if (IS_ERR(filp))
 		ret = PTR_ERR(filp);
 	else
@@ -1420,12 +1428,13 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
 		goto out;
 	}
 
-	ctx = nameidata_to_nfs_open_context(dentry, nd);
+	open_flags = nd->intent.open.flags;
+
+	ctx = create_nfs_open_context(dentry, open_flags);
 	res = ERR_CAST(ctx);
 	if (IS_ERR(ctx))
 		goto out;
 
-	open_flags = nd->intent.open.flags;
 	if (nd->flags & LOOKUP_CREATE) {
 		attr.ia_mode = nd->intent.open.create_mode;
 		attr.ia_valid = ATTR_MODE;
@@ -1463,8 +1472,8 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
 	res = d_add_unique(dentry, inode);
 	nfs_unblock_sillyrename(dentry->d_parent);
 	if (res != NULL) {
-		dput(ctx->path.dentry);
-		ctx->path.dentry = dget(res);
+		dput(ctx->dentry);
+		ctx->dentry = dget(res);
 		dentry = res;
 	}
 	err = nfs_intent_set_file(nd, ctx);
@@ -1517,7 +1526,7 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd)
 	/* We can't create new files, or truncate existing ones here */
 	openflags &= ~(O_CREAT|O_EXCL|O_TRUNC);
 
-	ctx = nameidata_to_nfs_open_context(dentry, nd);
+	ctx = create_nfs_open_context(dentry, openflags);
 	ret = PTR_ERR(ctx);
 	if (IS_ERR(ctx))
 		goto out;
@@ -1570,7 +1579,7 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	struct nfs_open_context *ctx = NULL;
 	struct iattr attr;
 	int error;
-	int open_flags = 0;
+	int open_flags = O_CREAT|O_EXCL;
 
 	dfprintk(VFS, "NFS: create(%s/%ld), %s\n",
 			dir->i_sb->s_id, dir->i_ino, dentry->d_name.name);
@@ -1578,27 +1587,27 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	attr.ia_mode = mode;
 	attr.ia_valid = ATTR_MODE;
 
-	if ((nd->flags & LOOKUP_CREATE) != 0) {
+	if (nd)
 		open_flags = nd->intent.open.flags;
 
-		ctx = nameidata_to_nfs_open_context(dentry, nd);
-		error = PTR_ERR(ctx);
-		if (IS_ERR(ctx))
-			goto out_err_drop;
-	}
+	ctx = create_nfs_open_context(dentry, open_flags);
+	error = PTR_ERR(ctx);
+	if (IS_ERR(ctx))
+		goto out_err_drop;
 
 	error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, ctx);
 	if (error != 0)
 		goto out_put_ctx;
-	if (ctx != NULL) {
+	if (nd) {
 		error = nfs_intent_set_file(nd, ctx);
 		if (error < 0)
 			goto out_err;
+	} else {
+		put_nfs_open_context(ctx);
 	}
 	return 0;
 out_put_ctx:
-	if (ctx != NULL)
-		put_nfs_open_context(ctx);
+	put_nfs_open_context(ctx);
 out_err_drop:
 	d_drop(dentry);
 out_err:
@@ -1660,7 +1669,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode,
 {
 	struct iattr attr;
 	int error;
-	int open_flags = 0;
+	int open_flags = O_CREAT|O_EXCL;
 
 	dfprintk(VFS, "NFS: create(%s/%ld), %s\n",
 			dir->i_sb->s_id, dir->i_ino, dentry->d_name.name);
@@ -1668,7 +1677,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	attr.ia_mode = mode;
 	attr.ia_valid = ATTR_MODE;
 
-	if ((nd->flags & LOOKUP_CREATE) != 0)
+	if (nd)
 		open_flags = nd->intent.open.flags;
 
 	error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, NULL);
@@ -2259,11 +2268,11 @@ static int nfs_open_permission_mask(int openflags)
 {
 	int mask = 0;
 
-	if (openflags & FMODE_READ)
+	if ((openflags & O_ACCMODE) != O_WRONLY)
 		mask |= MAY_READ;
-	if (openflags & FMODE_WRITE)
+	if ((openflags & O_ACCMODE) != O_RDONLY)
 		mask |= MAY_WRITE;
-	if (openflags & FMODE_EXEC)
+	if (openflags & __FMODE_EXEC)
 		mask |= MAY_EXEC;
 	return mask;
 }
@@ -2273,12 +2282,12 @@ int nfs_may_open(struct inode *inode, struct rpc_cred *cred, int openflags)
 	return nfs_do_access(inode, cred, nfs_open_permission_mask(openflags));
 }
 
-int nfs_permission(struct inode *inode, int mask, unsigned int flags)
+int nfs_permission(struct inode *inode, int mask)
 {
 	struct rpc_cred *cred;
 	int res = 0;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	nfs_inc_stats(inode, NFSIOS_VFSACCESS);
@@ -2328,7 +2337,7 @@ out:
 out_notsup:
 	res = nfs_revalidate_inode(NFS_SERVER(inode), inode);
 	if (res == 0)
-		res = generic_permission(inode, mask, flags, NULL);
+		res = generic_permission(inode, mask);
 	goto out;
 }
 
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 8eea2536671..b35d25b98da 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -284,7 +284,7 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq,
 						loff_t pos)
 {
 	struct nfs_open_context *ctx = dreq->ctx;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 	unsigned long user_addr = (unsigned long)iov->iov_base;
 	size_t count = iov->iov_len;
 	size_t rsize = NFS_SERVER(inode)->rsize;
@@ -715,7 +715,7 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq,
 						 loff_t pos, int sync)
 {
 	struct nfs_open_context *ctx = dreq->ctx;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 	unsigned long user_addr = (unsigned long)iov->iov_base;
 	size_t count = iov->iov_len;
 	struct rpc_task *task;
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 2f093ed1698..28b8c3f3cda 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -55,7 +55,7 @@ static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe,
 static ssize_t nfs_file_write(struct kiocb *, const struct iovec *iov,
 				unsigned long nr_segs, loff_t pos);
 static int  nfs_file_flush(struct file *, fl_owner_t id);
-static int  nfs_file_fsync(struct file *, int datasync);
+static int  nfs_file_fsync(struct file *, loff_t, loff_t, int datasync);
 static int nfs_check_flags(int flags);
 static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl);
 static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl);
@@ -187,8 +187,11 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 			filp->f_path.dentry->d_name.name,
 			offset, origin);
 
-	/* origin == SEEK_END => we must revalidate the cached file length */
-	if (origin == SEEK_END) {
+	/*
+	 * origin == SEEK_END || SEEK_DATA || SEEK_HOLE => we must revalidate
+	 * the cached file length
+	 */
+	if (origin != SEEK_SET || origin != SEEK_CUR) {
 		struct inode *inode = filp->f_mapping->host;
 
 		int retval = nfs_revalidate_file_size(inode, filp);
@@ -305,7 +308,7 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma)
  * fall back to doing a synchronous write.
  */
 static int
-nfs_file_fsync(struct file *file, int datasync)
+nfs_file_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct dentry *dentry = file->f_path.dentry;
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
@@ -313,11 +316,15 @@ nfs_file_fsync(struct file *file, int datasync)
 	int have_error, status;
 	int ret = 0;
 
-
 	dprintk("NFS: fsync file(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	nfs_inc_stats(inode, NFSIOS_VFSFSYNC);
 	have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
 	status = nfs_commit_inode(inode, FLUSH_SYNC);
@@ -329,6 +336,7 @@ nfs_file_fsync(struct file *file, int datasync)
 	if (!ret && !datasync)
 		/* application has asked for meta-data sync */
 		ret = pnfs_layoutcommit_inode(inode, true);
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/nfs/fscache.c b/fs/nfs/fscache.c
index ce153a6b3ae..419119c371b 100644
--- a/fs/nfs/fscache.c
+++ b/fs/nfs/fscache.c
@@ -259,12 +259,10 @@ static void nfs_fscache_disable_inode_cookie(struct inode *inode)
 		dfprintk(FSCACHE,
 			 "NFS: nfsi 0x%p turning cache off\n", NFS_I(inode));
 
-		/* Need to invalidate any mapped pages that were read in before
-		 * turning off the cache.
+		/* Need to uncache any pages attached to this inode that
+		 * fscache knows about before turning off the cache.
 		 */
-		if (inode->i_mapping && inode->i_mapping->nrpages)
-			invalidate_inode_pages2(inode->i_mapping);
-
+		fscache_uncache_all_inode_pages(NFS_I(inode)->fscache, inode);
 		nfs_fscache_zap_inode_cookie(inode);
 	}
 }
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 144f2a3c718..fe1203797b2 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -256,7 +256,8 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 
 	nfs_attr_check_mountpoint(sb, fattr);
 
-	if ((fattr->valid & NFS_ATTR_FATTR_FILEID) == 0 && (fattr->valid & NFS_ATTR_FATTR_MOUNTPOINT) == 0)
+	if (((fattr->valid & NFS_ATTR_FATTR_FILEID) == 0) &&
+	    !nfs_attr_use_mounted_on_fileid(fattr))
 		goto out_no_inode;
 	if ((fattr->valid & NFS_ATTR_FATTR_TYPE) == 0)
 		goto out_no_inode;
@@ -566,7 +567,7 @@ static struct nfs_lock_context *__nfs_find_lock_context(struct nfs_open_context
 struct nfs_lock_context *nfs_get_lock_context(struct nfs_open_context *ctx)
 {
 	struct nfs_lock_context *res, *new = NULL;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 
 	spin_lock(&inode->i_lock);
 	res = __nfs_find_lock_context(ctx);
@@ -593,7 +594,7 @@ struct nfs_lock_context *nfs_get_lock_context(struct nfs_open_context *ctx)
 void nfs_put_lock_context(struct nfs_lock_context *l_ctx)
 {
 	struct nfs_open_context *ctx = l_ctx->open_context;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 
 	if (!atomic_dec_and_lock(&l_ctx->count, &inode->i_lock))
 		return;
@@ -619,7 +620,7 @@ void nfs_close_context(struct nfs_open_context *ctx, int is_sync)
 		return;
 	if (!is_sync)
 		return;
-	inode = ctx->path.dentry->d_inode;
+	inode = ctx->dentry->d_inode;
 	if (!list_empty(&NFS_I(inode)->open_files))
 		return;
 	server = NFS_SERVER(inode);
@@ -628,14 +629,14 @@ void nfs_close_context(struct nfs_open_context *ctx, int is_sync)
 	nfs_revalidate_inode(server, inode);
 }
 
-struct nfs_open_context *alloc_nfs_open_context(struct path *path, struct rpc_cred *cred, fmode_t f_mode)
+struct nfs_open_context *alloc_nfs_open_context(struct dentry *dentry, struct rpc_cred *cred, fmode_t f_mode)
 {
 	struct nfs_open_context *ctx;
 
 	ctx = kmalloc(sizeof(*ctx), GFP_KERNEL);
 	if (ctx != NULL) {
-		ctx->path = *path;
-		path_get(&ctx->path);
+		nfs_sb_active(dentry->d_sb);
+		ctx->dentry = dget(dentry);
 		ctx->cred = get_rpccred(cred);
 		ctx->state = NULL;
 		ctx->mode = f_mode;
@@ -657,7 +658,8 @@ struct nfs_open_context *get_nfs_open_context(struct nfs_open_context *ctx)
 
 static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync)
 {
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
+	struct super_block *sb = ctx->dentry->d_sb;
 
 	if (!list_empty(&ctx->list)) {
 		if (!atomic_dec_and_lock(&ctx->lock_context.count, &inode->i_lock))
@@ -670,7 +672,8 @@ static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync)
 		NFS_PROTO(inode)->close_context(ctx, is_sync);
 	if (ctx->cred != NULL)
 		put_rpccred(ctx->cred);
-	path_put(&ctx->path);
+	dput(ctx->dentry);
+	nfs_sb_deactive(sb);
 	kfree(ctx);
 }
 
@@ -740,7 +743,7 @@ 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, cred, filp->f_mode);
+	ctx = alloc_nfs_open_context(filp->f_path.dentry, cred, filp->f_mode);
 	put_rpccred(cred);
 	if (ctx == NULL)
 		return -ENOMEM;
@@ -1294,7 +1297,8 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 		if (new_isize != cur_isize) {
 			/* Do we perhaps have any outstanding writes, or has
 			 * the file grown beyond our last write? */
-			if (nfsi->npages == 0 || new_isize > cur_isize) {
+			if ((nfsi->npages == 0 && !test_bit(NFS_INO_LAYOUTCOMMIT, &nfsi->flags)) ||
+			     new_isize > cur_isize) {
 				i_size_write(inode, new_isize);
 				invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
 			}
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index b9056cbe68d..2a55347a2da 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -45,6 +45,17 @@ static inline void nfs_attr_check_mountpoint(struct super_block *parent, struct
 		fattr->valid |= NFS_ATTR_FATTR_MOUNTPOINT;
 }
 
+static inline int nfs_attr_use_mounted_on_fileid(struct nfs_fattr *fattr)
+{
+	if (((fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID) == 0) ||
+	    (((fattr->valid & NFS_ATTR_FATTR_MOUNTPOINT) == 0) &&
+	     ((fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL) == 0)))
+		return 0;
+
+	fattr->fileid = fattr->mounted_on_fileid;
+	return 1;
+}
+
 struct nfs_clone_mount {
 	const struct super_block *sb;
 	const struct dentry *dentry;
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index c4a69833dd0..b788f2eb1ba 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -238,7 +238,7 @@ extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *);
 extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *);
 extern int nfs4_init_clientid(struct nfs_client *, struct rpc_cred *);
 extern int nfs41_init_clientid(struct nfs_client *, struct rpc_cred *);
-extern int nfs4_do_close(struct path *path, struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc);
+extern int nfs4_do_close(struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc);
 extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle);
 extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name,
 		struct nfs4_fs_locations *fs_locations, struct page *page);
@@ -341,8 +341,8 @@ extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struc
 extern void nfs4_put_state_owner(struct nfs4_state_owner *);
 extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state_owner *);
 extern void nfs4_put_open_state(struct nfs4_state *);
-extern void nfs4_close_state(struct path *, struct nfs4_state *, fmode_t);
-extern void nfs4_close_sync(struct path *, struct nfs4_state *, fmode_t);
+extern void nfs4_close_state(struct nfs4_state *, fmode_t);
+extern void nfs4_close_sync(struct nfs4_state *, fmode_t);
 extern void nfs4_state_set_mode_locked(struct nfs4_state *, fmode_t);
 extern void nfs4_schedule_lease_recovery(struct nfs_client *);
 extern void nfs4_schedule_state_manager(struct nfs_client *);
@@ -373,8 +373,8 @@ extern struct svc_version nfs4_callback_version4;
 
 #else
 
-#define nfs4_close_state(a, b, c) do { } while (0)
-#define nfs4_close_sync(a, b, c) do { } while (0)
+#define nfs4_close_state(a, b) do { } while (0)
+#define nfs4_close_sync(a, b) do { } while (0)
 
 #endif /* CONFIG_NFS_V4 */
 #endif /* __LINUX_FS_NFS_NFS4_FS.H */
diff --git a/fs/nfs/nfs4filelayout.c b/fs/nfs/nfs4filelayout.c
index 426908809c9..f9d03abcd04 100644
--- a/fs/nfs/nfs4filelayout.c
+++ b/fs/nfs/nfs4filelayout.c
@@ -30,6 +30,7 @@
  */
 
 #include <linux/nfs_fs.h>
+#include <linux/nfs_page.h>
 
 #include "internal.h"
 #include "nfs4filelayout.h"
@@ -397,7 +398,6 @@ filelayout_write_pagelist(struct nfs_write_data *data, int sync)
 	 * this offset and save the original offset.
 	 */
 	data->args.offset = filelayout_get_dserver_offset(lseg, offset);
-	data->mds_offset = offset;
 
 	/* Perform an asynchronous write */
 	status = nfs_initiate_write(data, ds->ds_clp->cl_rpcclient,
@@ -552,13 +552,18 @@ filelayout_decode_layout(struct pnfs_layout_hdr *flo,
 		__func__, nfl_util, fl->num_fh, fl->first_stripe_index,
 		fl->pattern_offset);
 
-	if (!fl->num_fh)
+	/* Note that a zero value for num_fh is legal for STRIPE_SPARSE.
+	 * Futher checking is done in filelayout_check_layout */
+	if (fl->num_fh < 0 || fl->num_fh >
+	    max(NFS4_PNFS_MAX_STRIPE_CNT, NFS4_PNFS_MAX_MULTI_CNT))
 		goto out_err;
 
-	fl->fh_array = kzalloc(fl->num_fh * sizeof(struct nfs_fh *),
-			       gfp_flags);
-	if (!fl->fh_array)
-		goto out_err;
+	if (fl->num_fh > 0) {
+		fl->fh_array = kzalloc(fl->num_fh * sizeof(struct nfs_fh *),
+				       gfp_flags);
+		if (!fl->fh_array)
+			goto out_err;
+	}
 
 	for (i = 0; i < fl->num_fh; i++) {
 		/* Do we want to use a mempool here? */
@@ -661,8 +666,9 @@ filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
 	u64 p_stripe, r_stripe;
 	u32 stripe_unit;
 
-	if (!pnfs_generic_pg_test(pgio, prev, req))
-		return 0;
+	if (!pnfs_generic_pg_test(pgio, prev, req) ||
+	    !nfs_generic_pg_test(pgio, prev, req))
+		return false;
 
 	if (!pgio->pg_lseg)
 		return 1;
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index d2c4b59c896..26bece8f308 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -763,8 +763,8 @@ struct nfs4_opendata {
 	struct nfs_open_confirmres c_res;
 	struct nfs_fattr f_attr;
 	struct nfs_fattr dir_attr;
-	struct path path;
 	struct dentry *dir;
+	struct dentry *dentry;
 	struct nfs4_state_owner *owner;
 	struct nfs4_state *state;
 	struct iattr attrs;
@@ -786,12 +786,12 @@ static void nfs4_init_opendata_res(struct nfs4_opendata *p)
 	nfs_fattr_init(&p->dir_attr);
 }
 
-static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
+static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry,
 		struct nfs4_state_owner *sp, fmode_t fmode, int flags,
 		const struct iattr *attrs,
 		gfp_t gfp_mask)
 {
-	struct dentry *parent = dget_parent(path->dentry);
+	struct dentry *parent = dget_parent(dentry);
 	struct inode *dir = parent->d_inode;
 	struct nfs_server *server = NFS_SERVER(dir);
 	struct nfs4_opendata *p;
@@ -802,8 +802,8 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
 	p->o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid, gfp_mask);
 	if (p->o_arg.seqid == NULL)
 		goto err_free;
-	path_get(path);
-	p->path = *path;
+	nfs_sb_active(dentry->d_sb);
+	p->dentry = dget(dentry);
 	p->dir = parent;
 	p->owner = sp;
 	atomic_inc(&sp->so_count);
@@ -812,7 +812,7 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
 	p->o_arg.fmode = fmode & (FMODE_READ|FMODE_WRITE);
 	p->o_arg.clientid = server->nfs_client->cl_clientid;
 	p->o_arg.id = sp->so_owner_id.id;
-	p->o_arg.name = &p->path.dentry->d_name;
+	p->o_arg.name = &dentry->d_name;
 	p->o_arg.server = server;
 	p->o_arg.bitmask = server->attr_bitmask;
 	p->o_arg.claim = NFS4_OPEN_CLAIM_NULL;
@@ -842,13 +842,15 @@ static void nfs4_opendata_free(struct kref *kref)
 {
 	struct nfs4_opendata *p = container_of(kref,
 			struct nfs4_opendata, kref);
+	struct super_block *sb = p->dentry->d_sb;
 
 	nfs_free_seqid(p->o_arg.seqid);
 	if (p->state != NULL)
 		nfs4_put_open_state(p->state);
 	nfs4_put_state_owner(p->owner);
 	dput(p->dir);
-	path_put(&p->path);
+	dput(p->dentry);
+	nfs_sb_deactive(sb);
 	kfree(p);
 }
 
@@ -1130,7 +1132,7 @@ static struct nfs4_opendata *nfs4_open_recoverdata_alloc(struct nfs_open_context
 {
 	struct nfs4_opendata *opendata;
 
-	opendata = nfs4_opendata_alloc(&ctx->path, state->owner, 0, 0, NULL, GFP_NOFS);
+	opendata = nfs4_opendata_alloc(ctx->dentry, state->owner, 0, 0, NULL, GFP_NOFS);
 	if (opendata == NULL)
 		return ERR_PTR(-ENOMEM);
 	opendata->state = state;
@@ -1154,7 +1156,7 @@ static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, fmode_t fmod
 	newstate = nfs4_opendata_to_nfs4_state(opendata);
 	if (IS_ERR(newstate))
 		return PTR_ERR(newstate);
-	nfs4_close_state(&opendata->path, newstate, fmode);
+	nfs4_close_state(newstate, fmode);
 	*res = newstate;
 	return 0;
 }
@@ -1352,7 +1354,7 @@ static void nfs4_open_confirm_release(void *calldata)
 		goto out_free;
 	state = nfs4_opendata_to_nfs4_state(data);
 	if (!IS_ERR(state))
-		nfs4_close_state(&data->path, state, data->o_arg.fmode);
+		nfs4_close_state(state, data->o_arg.fmode);
 out_free:
 	nfs4_opendata_put(data);
 }
@@ -1497,7 +1499,7 @@ static void nfs4_open_release(void *calldata)
 		goto out_free;
 	state = nfs4_opendata_to_nfs4_state(data);
 	if (!IS_ERR(state))
-		nfs4_close_state(&data->path, state, data->o_arg.fmode);
+		nfs4_close_state(state, data->o_arg.fmode);
 out_free:
 	nfs4_opendata_put(data);
 }
@@ -1648,7 +1650,7 @@ static int _nfs4_open_expired(struct nfs_open_context *ctx, struct nfs4_state *s
 		return PTR_ERR(opendata);
 	ret = nfs4_open_recover(opendata, state);
 	if (ret == -ESTALE)
-		d_drop(ctx->path.dentry);
+		d_drop(ctx->dentry);
 	nfs4_opendata_put(opendata);
 	return ret;
 }
@@ -1706,7 +1708,7 @@ static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata, struct
 /*
  * Returns a referenced nfs4_state
  */
-static int _nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res)
+static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res)
 {
 	struct nfs4_state_owner  *sp;
 	struct nfs4_state     *state = NULL;
@@ -1723,15 +1725,15 @@ static int _nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, in
 	status = nfs4_recover_expired_lease(server);
 	if (status != 0)
 		goto err_put_state_owner;
-	if (path->dentry->d_inode != NULL)
-		nfs4_return_incompatible_delegation(path->dentry->d_inode, fmode);
+	if (dentry->d_inode != NULL)
+		nfs4_return_incompatible_delegation(dentry->d_inode, fmode);
 	status = -ENOMEM;
-	opendata = nfs4_opendata_alloc(path, sp, fmode, flags, sattr, GFP_KERNEL);
+	opendata = nfs4_opendata_alloc(dentry, sp, fmode, flags, sattr, GFP_KERNEL);
 	if (opendata == NULL)
 		goto err_put_state_owner;
 
-	if (path->dentry->d_inode != NULL)
-		opendata->state = nfs4_get_open_state(path->dentry->d_inode, sp);
+	if (dentry->d_inode != NULL)
+		opendata->state = nfs4_get_open_state(dentry->d_inode, sp);
 
 	status = _nfs4_proc_open(opendata);
 	if (status != 0)
@@ -1769,14 +1771,14 @@ out_err:
 }
 
 
-static struct nfs4_state *nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred)
+static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred)
 {
 	struct nfs4_exception exception = { };
 	struct nfs4_state *res;
 	int status;
 
 	do {
-		status = _nfs4_do_open(dir, path, fmode, flags, sattr, cred, &res);
+		status = _nfs4_do_open(dir, dentry, fmode, flags, sattr, cred, &res);
 		if (status == 0)
 			break;
 		/* NOTE: BAD_SEQID means the server and client disagree about the
@@ -1873,7 +1875,6 @@ static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 }
 
 struct nfs4_closedata {
-	struct path path;
 	struct inode *inode;
 	struct nfs4_state *state;
 	struct nfs_closeargs arg;
@@ -1888,13 +1889,14 @@ static void nfs4_free_closedata(void *data)
 {
 	struct nfs4_closedata *calldata = data;
 	struct nfs4_state_owner *sp = calldata->state->owner;
+	struct super_block *sb = calldata->state->inode->i_sb;
 
 	if (calldata->roc)
 		pnfs_roc_release(calldata->state->inode);
 	nfs4_put_open_state(calldata->state);
 	nfs_free_seqid(calldata->arg.seqid);
 	nfs4_put_state_owner(sp);
-	path_put(&calldata->path);
+	nfs_sb_deactive(sb);
 	kfree(calldata);
 }
 
@@ -2014,7 +2016,7 @@ static const struct rpc_call_ops nfs4_close_ops = {
  *
  * NOTE: Caller must be holding the sp->so_owner semaphore!
  */
-int nfs4_do_close(struct path *path, struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc)
+int nfs4_do_close(struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc)
 {
 	struct nfs_server *server = NFS_SERVER(state->inode);
 	struct nfs4_closedata *calldata;
@@ -2050,8 +2052,7 @@ int nfs4_do_close(struct path *path, struct nfs4_state *state, gfp_t gfp_mask, i
 	calldata->res.seqid = calldata->arg.seqid;
 	calldata->res.server = server;
 	calldata->roc = roc;
-	path_get(path);
-	calldata->path = *path;
+	nfs_sb_active(calldata->inode->i_sb);
 
 	msg.rpc_argp = &calldata->arg;
 	msg.rpc_resp = &calldata->res;
@@ -2080,7 +2081,7 @@ nfs4_atomic_open(struct inode *dir, struct nfs_open_context *ctx, int open_flags
 	struct nfs4_state *state;
 
 	/* Protect against concurrent sillydeletes */
-	state = nfs4_do_open(dir, &ctx->path, ctx->mode, open_flags, attr, ctx->cred);
+	state = nfs4_do_open(dir, ctx->dentry, ctx->mode, open_flags, attr, ctx->cred);
 	if (IS_ERR(state))
 		return ERR_CAST(state);
 	ctx->state = state;
@@ -2092,9 +2093,9 @@ static void nfs4_close_context(struct nfs_open_context *ctx, int is_sync)
 	if (ctx->state == NULL)
 		return;
 	if (is_sync)
-		nfs4_close_sync(&ctx->path, ctx->state, ctx->mode);
+		nfs4_close_sync(ctx->state, ctx->mode);
 	else
-		nfs4_close_state(&ctx->path, ctx->state, ctx->mode);
+		nfs4_close_state(ctx->state, ctx->mode);
 }
 
 static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle)
@@ -2265,12 +2266,14 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 	return nfs4_map_errors(status);
 }
 
+static void nfs_fixup_referral_attributes(struct nfs_fattr *fattr);
 /*
  * Get locations and (maybe) other attributes of a referral.
  * Note that we'll actually follow the referral later when
  * we detect fsid mismatch in inode revalidation
  */
-static int nfs4_get_referral(struct inode *dir, const struct qstr *name, struct nfs_fattr *fattr, struct nfs_fh *fhandle)
+static int nfs4_get_referral(struct inode *dir, const struct qstr *name,
+			     struct nfs_fattr *fattr, struct nfs_fh *fhandle)
 {
 	int status = -ENOMEM;
 	struct page *page = NULL;
@@ -2288,15 +2291,16 @@ static int nfs4_get_referral(struct inode *dir, const struct qstr *name, struct
 		goto out;
 	/* Make sure server returned a different fsid for the referral */
 	if (nfs_fsid_equal(&NFS_SERVER(dir)->fsid, &locations->fattr.fsid)) {
-		dprintk("%s: server did not return a different fsid for a referral at %s\n", __func__, name->name);
+		dprintk("%s: server did not return a different fsid for"
+			" a referral at %s\n", __func__, name->name);
 		status = -EIO;
 		goto out;
 	}
+	/* Fixup attributes for the nfs_lookup() call to nfs_fhget() */
+	nfs_fixup_referral_attributes(&locations->fattr);
 
+	/* replace the lookup nfs_fattr with the locations nfs_fattr */
 	memcpy(fattr, &locations->fattr, sizeof(struct nfs_fattr));
-	fattr->valid |= NFS_ATTR_FATTR_V4_REFERRAL;
-	if (!fattr->mode)
-		fattr->mode = S_IFDIR;
 	memset(fhandle, 0, sizeof(struct nfs_fh));
 out:
 	if (page)
@@ -2613,10 +2617,7 @@ static int
 nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
                  int flags, struct nfs_open_context *ctx)
 {
-	struct path my_path = {
-		.dentry = dentry,
-	};
-	struct path *path = &my_path;
+	struct dentry *de = dentry;
 	struct nfs4_state *state;
 	struct rpc_cred *cred = NULL;
 	fmode_t fmode = 0;
@@ -2624,11 +2625,11 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
 
 	if (ctx != NULL) {
 		cred = ctx->cred;
-		path = &ctx->path;
+		de = ctx->dentry;
 		fmode = ctx->mode;
 	}
 	sattr->ia_mode &= ~current_umask();
-	state = nfs4_do_open(dir, path, fmode, flags, sattr, cred);
+	state = nfs4_do_open(dir, de, fmode, flags, sattr, cred);
 	d_drop(dentry);
 	if (IS_ERR(state)) {
 		status = PTR_ERR(state);
@@ -2639,7 +2640,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
 	if (ctx != NULL)
 		ctx->state = state;
 	else
-		nfs4_close_sync(path, state, fmode);
+		nfs4_close_sync(state, fmode);
 out:
 	return status;
 }
@@ -4291,7 +4292,7 @@ static void nfs4_lock_done(struct rpc_task *task, void *calldata)
 		memcpy(data->lsp->ls_stateid.data, data->res.stateid.data,
 					sizeof(data->lsp->ls_stateid.data));
 		data->lsp->ls_flags |= NFS_LOCK_INITIALIZED;
-		renew_lease(NFS_SERVER(data->ctx->path.dentry->d_inode), data->timestamp);
+		renew_lease(NFS_SERVER(data->ctx->dentry->d_inode), data->timestamp);
 	}
 out:
 	dprintk("%s: done, ret = %d!\n", __func__, data->rpc_status);
@@ -4667,11 +4668,15 @@ static size_t nfs4_xattr_list_nfs4_acl(struct dentry *dentry, char *list,
 	return len;
 }
 
+/*
+ * nfs_fhget will use either the mounted_on_fileid or the fileid
+ */
 static void nfs_fixup_referral_attributes(struct nfs_fattr *fattr)
 {
-	if (!((fattr->valid & NFS_ATTR_FATTR_FILEID) &&
-		(fattr->valid & NFS_ATTR_FATTR_FSID) &&
-		(fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL)))
+	if (!(((fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID) ||
+	       (fattr->valid & NFS_ATTR_FATTR_FILEID)) &&
+	      (fattr->valid & NFS_ATTR_FATTR_FSID) &&
+	      (fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL)))
 		return;
 
 	fattr->valid |= NFS_ATTR_FATTR_TYPE | NFS_ATTR_FATTR_MODE |
@@ -4686,7 +4691,6 @@ int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name,
 	struct nfs_server *server = NFS_SERVER(dir);
 	u32 bitmask[2] = {
 		[0] = FATTR4_WORD0_FSID | FATTR4_WORD0_FS_LOCATIONS,
-		[1] = FATTR4_WORD1_MOUNTED_ON_FILEID,
 	};
 	struct nfs4_fs_locations_arg args = {
 		.dir_fh = NFS_FH(dir),
@@ -4705,11 +4709,18 @@ int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name,
 	int status;
 
 	dprintk("%s: start\n", __func__);
+
+	/* Ask for the fileid of the absent filesystem if mounted_on_fileid
+	 * is not supported */
+	if (NFS_SERVER(dir)->attr_bitmask[1] & FATTR4_WORD1_MOUNTED_ON_FILEID)
+		bitmask[1] |= FATTR4_WORD1_MOUNTED_ON_FILEID;
+	else
+		bitmask[0] |= FATTR4_WORD0_FILEID;
+
 	nfs_fattr_init(&fs_locations->fattr);
 	fs_locations->server = server;
 	fs_locations->nlocations = 0;
 	status = nfs4_call_sync(server->client, server, &msg, &args.seq_args, &res.seq_res, 0);
-	nfs_fixup_referral_attributes(&fs_locations->fattr);
 	dprintk("%s: returned status = %d\n", __func__, status);
 	return status;
 }
@@ -5098,7 +5109,6 @@ static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args)
 	if (mxresp_sz == 0)
 		mxresp_sz = NFS_MAX_FILE_IO_SIZE;
 	/* Fore channel attributes */
-	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_ops = NFS4_MAX_OPS;
@@ -5111,7 +5121,6 @@ static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args)
 		args->fc_attrs.max_ops, args->fc_attrs.max_reqs);
 
 	/* Back channel attributes */
-	args->bc_attrs.headerpadsz = 0;
 	args->bc_attrs.max_rqst_sz = PAGE_SIZE;
 	args->bc_attrs.max_resp_sz = PAGE_SIZE;
 	args->bc_attrs.max_resp_sz_cached = 0;
@@ -5131,8 +5140,6 @@ static int nfs4_verify_fore_channel_attrs(struct nfs41_create_session_args *args
 	struct nfs4_channel_attrs *sent = &args->fc_attrs;
 	struct nfs4_channel_attrs *rcvd = &session->fc_attrs;
 
-	if (rcvd->headerpadsz > sent->headerpadsz)
-		return -EINVAL;
 	if (rcvd->max_resp_sz > sent->max_resp_sz)
 		return -EINVAL;
 	/*
@@ -5697,6 +5704,7 @@ static void nfs4_layoutreturn_done(struct rpc_task *task, void *calldata)
 {
 	struct nfs4_layoutreturn *lrp = calldata;
 	struct nfs_server *server;
+	struct pnfs_layout_hdr *lo = NFS_I(lrp->args.inode)->layout;
 
 	dprintk("--> %s\n", __func__);
 
@@ -5708,16 +5716,15 @@ static void nfs4_layoutreturn_done(struct rpc_task *task, void *calldata)
 		nfs_restart_rpc(task, lrp->clp);
 		return;
 	}
+	spin_lock(&lo->plh_inode->i_lock);
 	if (task->tk_status == 0) {
-		struct pnfs_layout_hdr *lo = NFS_I(lrp->args.inode)->layout;
-
 		if (lrp->res.lrs_present) {
-			spin_lock(&lo->plh_inode->i_lock);
 			pnfs_set_layout_stateid(lo, &lrp->res.stateid, true);
-			spin_unlock(&lo->plh_inode->i_lock);
 		} else
 			BUG_ON(!list_empty(&lo->plh_segs));
 	}
+	lo->plh_block_lgets--;
+	spin_unlock(&lo->plh_inode->i_lock);
 	dprintk("<-- %s\n", __func__);
 }
 
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index e97dd219f84..7acfe884362 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -641,7 +641,7 @@ void nfs4_put_open_state(struct nfs4_state *state)
 /*
  * Close the current file.
  */
-static void __nfs4_close(struct path *path, struct nfs4_state *state,
+static void __nfs4_close(struct nfs4_state *state,
 		fmode_t fmode, gfp_t gfp_mask, int wait)
 {
 	struct nfs4_state_owner *owner = state->owner;
@@ -685,18 +685,18 @@ static void __nfs4_close(struct path *path, struct nfs4_state *state,
 	} else {
 		bool roc = pnfs_roc(state->inode);
 
-		nfs4_do_close(path, state, gfp_mask, wait, roc);
+		nfs4_do_close(state, gfp_mask, wait, roc);
 	}
 }
 
-void nfs4_close_state(struct path *path, struct nfs4_state *state, fmode_t fmode)
+void nfs4_close_state(struct nfs4_state *state, fmode_t fmode)
 {
-	__nfs4_close(path, state, fmode, GFP_NOFS, 0);
+	__nfs4_close(state, fmode, GFP_NOFS, 0);
 }
 
-void nfs4_close_sync(struct path *path, struct nfs4_state *state, fmode_t fmode)
+void nfs4_close_sync(struct nfs4_state *state, fmode_t fmode)
 {
-	__nfs4_close(path, state, fmode, GFP_KERNEL, 1);
+	__nfs4_close(state, fmode, GFP_KERNEL, 1);
 }
 
 /*
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index d869a5e5464..e6e8f3b9a1d 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -91,7 +91,7 @@ static int nfs4_stat_to_errno(int);
 #define encode_getfh_maxsz      (op_encode_hdr_maxsz)
 #define decode_getfh_maxsz      (op_decode_hdr_maxsz + 1 + \
 				((3+NFS4_FHSIZE) >> 2))
-#define nfs4_fattr_bitmap_maxsz 3
+#define nfs4_fattr_bitmap_maxsz 4
 #define encode_getattr_maxsz    (op_encode_hdr_maxsz + nfs4_fattr_bitmap_maxsz)
 #define nfs4_name_maxsz		(1 + ((3 + NFS4_MAXNAMLEN) >> 2))
 #define nfs4_path_maxsz		(1 + ((3 + NFS4_MAXPATHLEN) >> 2))
@@ -255,7 +255,7 @@ static int nfs4_stat_to_errno(int);
 #define decode_fs_locations_maxsz \
 				(0)
 #define encode_secinfo_maxsz	(op_encode_hdr_maxsz + nfs4_name_maxsz)
-#define decode_secinfo_maxsz	(op_decode_hdr_maxsz + 4 + (NFS_MAX_SECFLAVORS * (16 + GSS_OID_MAX_LEN)))
+#define decode_secinfo_maxsz	(op_decode_hdr_maxsz + 1 + ((NFS_MAX_SECFLAVORS * (16 + GSS_OID_MAX_LEN)) / 4))
 
 #if defined(CONFIG_NFS_V4_1)
 #define NFS4_MAX_MACHINE_NAME_LEN (64)
@@ -1725,7 +1725,7 @@ static void encode_create_session(struct xdr_stream *xdr,
 	*p++ = cpu_to_be32(args->flags);			/*flags */
 
 	/* Fore Channel */
-	*p++ = cpu_to_be32(args->fc_attrs.headerpadsz);	/* header padding size */
+	*p++ = cpu_to_be32(0);				/* 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(max_resp_sz_cached);		/* Max resp sz cached */
@@ -1734,7 +1734,7 @@ static void encode_create_session(struct xdr_stream *xdr,
 	*p++ = cpu_to_be32(0);				/* rdmachannel_attrs */
 
 	/* Back Channel */
-	*p++ = cpu_to_be32(args->fc_attrs.headerpadsz);	/* header padding size */
+	*p++ = cpu_to_be32(0);				/* header padding size */
 	*p++ = cpu_to_be32(args->bc_attrs.max_rqst_sz);	/* max req size */
 	*p++ = cpu_to_be32(args->bc_attrs.max_resp_sz);	/* max resp size */
 	*p++ = cpu_to_be32(args->bc_attrs.max_resp_sz_cached);	/* Max resp sz cached */
@@ -3098,7 +3098,7 @@ out_overflow:
 	return -EIO;
 }
 
-static int decode_attr_error(struct xdr_stream *xdr, uint32_t *bitmap)
+static int decode_attr_error(struct xdr_stream *xdr, uint32_t *bitmap, int32_t *res)
 {
 	__be32 *p;
 
@@ -3109,7 +3109,7 @@ static int decode_attr_error(struct xdr_stream *xdr, uint32_t *bitmap)
 		if (unlikely(!p))
 			goto out_overflow;
 		bitmap[0] &= ~FATTR4_WORD0_RDATTR_ERROR;
-		return -be32_to_cpup(p);
+		*res = -be32_to_cpup(p);
 	}
 	return 0;
 out_overflow:
@@ -4070,6 +4070,7 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap,
 	int status;
 	umode_t fmode = 0;
 	uint32_t type;
+	int32_t err;
 
 	status = decode_attr_type(xdr, bitmap, &type);
 	if (status < 0)
@@ -4095,13 +4096,12 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap,
 		goto xdr_error;
 	fattr->valid |= status;
 
-	status = decode_attr_error(xdr, bitmap);
-	if (status == -NFS4ERR_WRONGSEC) {
-		nfs_fixup_secinfo_attributes(fattr, fh);
-		status = 0;
-	}
+	err = 0;
+	status = decode_attr_error(xdr, bitmap, &err);
 	if (status < 0)
 		goto xdr_error;
+	if (err == -NFS4ERR_WRONGSEC)
+		nfs_fixup_secinfo_attributes(fattr, fh);
 
 	status = decode_attr_filehandle(xdr, bitmap, fh);
 	if (status < 0)
@@ -4997,12 +4997,14 @@ static int decode_chan_attrs(struct xdr_stream *xdr,
 			     struct nfs4_channel_attrs *attrs)
 {
 	__be32 *p;
-	u32 nr_attrs;
+	u32 nr_attrs, val;
 
 	p = xdr_inline_decode(xdr, 28);
 	if (unlikely(!p))
 		goto out_overflow;
-	attrs->headerpadsz = be32_to_cpup(p++);
+	val = be32_to_cpup(p++);	/* headerpadsz */
+	if (val)
+		return -EINVAL;		/* no support for header padding yet */
 	attrs->max_rqst_sz = be32_to_cpup(p++);
 	attrs->max_resp_sz = be32_to_cpup(p++);
 	attrs->max_resp_sz_cached = be32_to_cpup(p++);
diff --git a/fs/nfs/objlayout/objio_osd.c b/fs/nfs/objlayout/objio_osd.c
index 9cf208df1f2..8ff2ea3f10e 100644
--- a/fs/nfs/objlayout/objio_osd.c
+++ b/fs/nfs/objlayout/objio_osd.c
@@ -108,7 +108,6 @@ _dev_list_add(const struct nfs_server *nfss,
 		de = n;
 	}
 
-	atomic_inc(&de->id_node.ref);
 	return de;
 }
 
@@ -1001,6 +1000,9 @@ static bool objio_pg_test(struct nfs_pageio_descriptor *pgio,
 	if (!pnfs_generic_pg_test(pgio, prev, req))
 		return false;
 
+	if (pgio->pg_lseg == NULL)
+		return true;
+
 	return pgio->pg_count + req->wb_bytes <=
 			OBJIO_LSEG(pgio->pg_lseg)->max_io_size;
 }
diff --git a/fs/nfs/objlayout/objlayout.c b/fs/nfs/objlayout/objlayout.c
index dc3956c0de8..1d06f8e2ade 100644
--- a/fs/nfs/objlayout/objlayout.c
+++ b/fs/nfs/objlayout/objlayout.c
@@ -291,7 +291,7 @@ objlayout_read_done(struct objlayout_io_state *state, ssize_t status, bool sync)
 	struct nfs_read_data *rdata;
 
 	state->status = status;
-	dprintk("%s: Begin status=%ld eof=%d\n", __func__, status, eof);
+	dprintk("%s: Begin status=%zd eof=%d\n", __func__, status, eof);
 	rdata = state->rpcdata;
 	rdata->task.tk_status = status;
 	if (status >= 0) {
diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index 7913961aff2..18449f43c56 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -114,7 +114,7 @@ int nfs_set_page_tag_locked(struct nfs_page *req)
 	if (!nfs_lock_request_dontget(req))
 		return 0;
 	if (test_bit(PG_MAPPED, &req->wb_flags))
-		radix_tree_tag_set(&NFS_I(req->wb_context->path.dentry->d_inode)->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED);
+		radix_tree_tag_set(&NFS_I(req->wb_context->dentry->d_inode)->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED);
 	return 1;
 }
 
@@ -124,7 +124,7 @@ int nfs_set_page_tag_locked(struct nfs_page *req)
 void nfs_clear_page_tag_locked(struct nfs_page *req)
 {
 	if (test_bit(PG_MAPPED, &req->wb_flags)) {
-		struct inode *inode = req->wb_context->path.dentry->d_inode;
+		struct inode *inode = req->wb_context->dentry->d_inode;
 		struct nfs_inode *nfsi = NFS_I(inode);
 
 		spin_lock(&inode->i_lock);
@@ -204,7 +204,7 @@ nfs_wait_on_request(struct nfs_page *req)
 			TASK_UNINTERRUPTIBLE);
 }
 
-static bool nfs_generic_pg_test(struct nfs_pageio_descriptor *desc, struct nfs_page *prev, struct nfs_page *req)
+bool nfs_generic_pg_test(struct nfs_pageio_descriptor *desc, struct nfs_page *prev, struct nfs_page *req)
 {
 	/*
 	 * FIXME: ideally we should be able to coalesce all requests
@@ -218,6 +218,7 @@ static bool nfs_generic_pg_test(struct nfs_pageio_descriptor *desc, struct nfs_p
 
 	return desc->pg_count + req->wb_bytes <= desc->pg_bsize;
 }
+EXPORT_SYMBOL_GPL(nfs_generic_pg_test);
 
 /**
  * nfs_pageio_init - initialise a page io descriptor
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 8c1309d852a..29c0ca7fc34 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -634,14 +634,16 @@ _pnfs_return_layout(struct inode *ino)
 
 	spin_lock(&ino->i_lock);
 	lo = nfsi->layout;
-	if (!lo || !mark_matching_lsegs_invalid(lo, &tmp_list, NULL)) {
+	if (!lo) {
 		spin_unlock(&ino->i_lock);
-		dprintk("%s: no layout segments to return\n", __func__);
-		goto out;
+		dprintk("%s: no layout to return\n", __func__);
+		return status;
 	}
 	stateid = nfsi->layout->plh_stateid;
 	/* Reference matched in nfs4_layoutreturn_release */
 	get_layout_hdr(lo);
+	mark_matching_lsegs_invalid(lo, &tmp_list, NULL);
+	lo->plh_block_lgets++;
 	spin_unlock(&ino->i_lock);
 	pnfs_free_lseg_list(&tmp_list);
 
@@ -650,6 +652,9 @@ _pnfs_return_layout(struct inode *ino)
 	lrp = kzalloc(sizeof(*lrp), GFP_KERNEL);
 	if (unlikely(lrp == NULL)) {
 		status = -ENOMEM;
+		set_bit(NFS_LAYOUT_RW_FAILED, &lo->plh_flags);
+		set_bit(NFS_LAYOUT_RO_FAILED, &lo->plh_flags);
+		put_layout_hdr(lo);
 		goto out;
 	}
 
@@ -887,7 +892,7 @@ pnfs_find_lseg(struct pnfs_layout_hdr *lo,
 			ret = get_lseg(lseg);
 			break;
 		}
-		if (cmp_layout(range, &lseg->pls_range) > 0)
+		if (lseg->pls_range.offset > range->offset)
 			break;
 	}
 
@@ -1059,23 +1064,36 @@ pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
 		gfp_flags = GFP_NOFS;
 	}
 
-	if (pgio->pg_count == prev->wb_bytes) {
+	if (pgio->pg_lseg == NULL) {
+		if (pgio->pg_count != prev->wb_bytes)
+			return true;
 		/* This is first coelesce call for a series of nfs_pages */
 		pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
 						   prev->wb_context,
-						   req_offset(req),
+						   req_offset(prev),
 						   pgio->pg_count,
 						   access_type,
 						   gfp_flags);
-		return true;
+		if (pgio->pg_lseg == NULL)
+			return true;
 	}
 
-	if (pgio->pg_lseg &&
-	    req_offset(req) > end_offset(pgio->pg_lseg->pls_range.offset,
-					 pgio->pg_lseg->pls_range.length))
-		return false;
-
-	return true;
+	/*
+	 * Test if a nfs_page is fully contained in the pnfs_layout_range.
+	 * Note that this test makes several assumptions:
+	 * - that the previous nfs_page in the struct nfs_pageio_descriptor
+	 *   is known to lie within the range.
+	 *   - that the nfs_page being tested is known to be contiguous with the
+	 *   previous nfs_page.
+	 *   - Layout ranges are page aligned, so we only have to test the
+	 *   start offset of the request.
+	 *
+	 * Please also note that 'end_offset' is actually the offset of the
+	 * first byte that lies outside the pnfs_layout_range. FIXME?
+	 *
+	 */
+	return req_offset(req) < end_offset(pgio->pg_lseg->pls_range.offset,
+					 pgio->pg_lseg->pls_range.length);
 }
 EXPORT_SYMBOL_GPL(pnfs_generic_pg_test);
 
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index 48d0a8e4d06..96bf4e6f45b 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -186,6 +186,7 @@ int pnfs_ld_read_done(struct nfs_read_data *);
 /* pnfs_dev.c */
 struct nfs4_deviceid_node {
 	struct hlist_node		node;
+	struct hlist_node		tmpnode;
 	const struct pnfs_layoutdriver_type *ld;
 	const struct nfs_client		*nfs_client;
 	struct nfs4_deviceid		deviceid;
diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c
index c65e133ce9c..f0f8e1e22f6 100644
--- a/fs/nfs/pnfs_dev.c
+++ b/fs/nfs/pnfs_dev.c
@@ -174,6 +174,7 @@ nfs4_init_deviceid_node(struct nfs4_deviceid_node *d,
 			const struct nfs4_deviceid *id)
 {
 	INIT_HLIST_NODE(&d->node);
+	INIT_HLIST_NODE(&d->tmpnode);
 	d->ld = ld;
 	d->nfs_client = nfs_client;
 	d->deviceid = *id;
@@ -208,6 +209,7 @@ nfs4_insert_deviceid_node(struct nfs4_deviceid_node *new)
 
 	hlist_add_head_rcu(&new->node, &nfs4_deviceid_cache[hash]);
 	spin_unlock(&nfs4_deviceid_lock);
+	atomic_inc(&new->ref);
 
 	return new;
 }
@@ -238,24 +240,29 @@ static void
 _deviceid_purge_client(const struct nfs_client *clp, long hash)
 {
 	struct nfs4_deviceid_node *d;
-	struct hlist_node *n, *next;
+	struct hlist_node *n;
 	HLIST_HEAD(tmp);
 
+	spin_lock(&nfs4_deviceid_lock);
 	rcu_read_lock();
 	hlist_for_each_entry_rcu(d, n, &nfs4_deviceid_cache[hash], node)
 		if (d->nfs_client == clp && atomic_read(&d->ref)) {
 			hlist_del_init_rcu(&d->node);
-			hlist_add_head(&d->node, &tmp);
+			hlist_add_head(&d->tmpnode, &tmp);
 		}
 	rcu_read_unlock();
+	spin_unlock(&nfs4_deviceid_lock);
 
 	if (hlist_empty(&tmp))
 		return;
 
 	synchronize_rcu();
-	hlist_for_each_entry_safe(d, n, next, &tmp, node)
+	while (!hlist_empty(&tmp)) {
+		d = hlist_entry(tmp.first, struct nfs4_deviceid_node, tmpnode);
+		hlist_del(&d->tmpnode);
 		if (atomic_dec_and_test(&d->ref))
 			d->ld->free_deviceid_node(d);
+	}
 }
 
 void
@@ -263,8 +270,8 @@ nfs4_deviceid_purge_client(const struct nfs_client *clp)
 {
 	long h;
 
-	spin_lock(&nfs4_deviceid_lock);
+	if (!(clp->cl_exchange_flags & EXCHGID4_FLAG_USE_PNFS_MDS))
+		return;
 	for (h = 0; h < NFS4_DEVICE_ID_HASH_SIZE; h++)
 		_deviceid_purge_client(clp, h);
-	spin_unlock(&nfs4_deviceid_lock);
 }
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index 20a7f952e24..a68679f538f 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -144,7 +144,7 @@ int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
 
 static void nfs_readpage_release(struct nfs_page *req)
 {
-	struct inode *d_inode = req->wb_context->path.dentry->d_inode;
+	struct inode *d_inode = req->wb_context->dentry->d_inode;
 
 	if (PageUptodate(req->wb_page))
 		nfs_readpage_to_fscache(d_inode, req->wb_page, 0);
@@ -152,8 +152,8 @@ static void nfs_readpage_release(struct nfs_page *req)
 	unlock_page(req->wb_page);
 
 	dprintk("NFS: read done (%s/%Ld %d@%Ld)\n",
-			req->wb_context->path.dentry->d_inode->i_sb->s_id,
-			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
+			req->wb_context->dentry->d_inode->i_sb->s_id,
+			(long long)NFS_FILEID(req->wb_context->dentry->d_inode),
 			req->wb_bytes,
 			(long long)req_offset(req));
 	nfs_release_request(req);
@@ -207,7 +207,7 @@ static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data,
 		unsigned int count, unsigned int offset,
 		struct pnfs_layout_segment *lseg)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 
 	data->req	  = req;
 	data->inode	  = inode;
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index ce40e5c568b..b961ceac66b 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -2773,16 +2773,12 @@ static void nfs_referral_loop_unprotect(void)
 static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt,
 		const char *export_path)
 {
-	struct nameidata *nd = NULL;
 	struct mnt_namespace *ns_private;
 	struct super_block *s;
 	struct dentry *dentry;
+	struct path path;
 	int ret;
 
-	nd = kmalloc(sizeof(*nd), GFP_KERNEL);
-	if (nd == NULL)
-		return ERR_PTR(-ENOMEM);
-
 	ns_private = create_mnt_ns(root_mnt);
 	ret = PTR_ERR(ns_private);
 	if (IS_ERR(ns_private))
@@ -2793,7 +2789,7 @@ static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt,
 		goto out_put_mnt_ns;
 
 	ret = vfs_path_lookup(root_mnt->mnt_root, root_mnt,
-			export_path, LOOKUP_FOLLOW, nd);
+			export_path, LOOKUP_FOLLOW, &path);
 
 	nfs_referral_loop_unprotect();
 	put_mnt_ns(ns_private);
@@ -2801,12 +2797,11 @@ static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt,
 	if (ret != 0)
 		goto out_err;
 
-	s = nd->path.mnt->mnt_sb;
+	s = path.mnt->mnt_sb;
 	atomic_inc(&s->s_active);
-	dentry = dget(nd->path.dentry);
+	dentry = dget(path.dentry);
 
-	path_put(&nd->path);
-	kfree(nd);
+	path_put(&path);
 	down_write(&s->s_umount);
 	return dentry;
 out_put_mnt_ns:
@@ -2814,7 +2809,6 @@ out_put_mnt_ns:
 out_mntput:
 	mntput(root_mnt);
 out_err:
-	kfree(nd);
 	return ERR_PTR(ret);
 }
 
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index e268e3b2349..08579312c57 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -409,7 +409,7 @@ out:
  */
 static void nfs_inode_remove_request(struct nfs_page *req)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	BUG_ON (!NFS_WBACK_BUSY(req));
@@ -438,7 +438,7 @@ nfs_mark_request_dirty(struct nfs_page *req)
 static void
 nfs_mark_request_commit(struct nfs_page *req, struct pnfs_layout_segment *lseg)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	spin_lock(&inode->i_lock);
@@ -852,18 +852,20 @@ static int nfs_write_rpcsetup(struct nfs_page *req,
 		struct pnfs_layout_segment *lseg,
 		int how)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 
 	/* Set up the RPC argument and reply structs
 	 * NB: take care not to mess about with data->commit et al. */
 
 	data->req = req;
-	data->inode = inode = req->wb_context->path.dentry->d_inode;
+	data->inode = inode = req->wb_context->dentry->d_inode;
 	data->cred = req->wb_context->cred;
 	data->lseg = get_lseg(lseg);
 
 	data->args.fh     = NFS_FH(inode);
 	data->args.offset = req_offset(req) + offset;
+	/* pnfs_set_layoutcommit needs this */
+	data->mds_offset = data->args.offset;
 	data->args.pgbase = req->wb_pgbase + offset;
 	data->args.pages  = data->pagevec;
 	data->args.count  = count;
@@ -1051,9 +1053,9 @@ static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata)
 
 	dprintk("NFS: %5u write(%s/%lld %d@%lld)",
 		task->tk_pid,
-		data->req->wb_context->path.dentry->d_inode->i_sb->s_id,
+		data->req->wb_context->dentry->d_inode->i_sb->s_id,
 		(long long)
-		  NFS_FILEID(data->req->wb_context->path.dentry->d_inode),
+		  NFS_FILEID(data->req->wb_context->dentry->d_inode),
 		data->req->wb_bytes, (long long)req_offset(data->req));
 
 	nfs_writeback_done(task, data);
@@ -1146,8 +1148,8 @@ static void nfs_writeback_release_full(void *calldata)
 
 		dprintk("NFS: %5u write (%s/%lld %d@%lld)",
 			data->task.tk_pid,
-			req->wb_context->path.dentry->d_inode->i_sb->s_id,
-			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
+			req->wb_context->dentry->d_inode->i_sb->s_id,
+			(long long)NFS_FILEID(req->wb_context->dentry->d_inode),
 			req->wb_bytes,
 			(long long)req_offset(req));
 
@@ -1345,7 +1347,7 @@ void nfs_init_commit(struct nfs_write_data *data,
 			    struct pnfs_layout_segment *lseg)
 {
 	struct nfs_page *first = nfs_list_entry(head->next);
-	struct inode *inode = first->wb_context->path.dentry->d_inode;
+	struct inode *inode = first->wb_context->dentry->d_inode;
 
 	/* Set up the RPC argument and reply structs
 	 * NB: take care not to mess about with data->commit et al. */
@@ -1433,8 +1435,8 @@ void nfs_commit_release_pages(struct nfs_write_data *data)
 		nfs_clear_request_commit(req);
 
 		dprintk("NFS:       commit (%s/%lld %d@%lld)",
-			req->wb_context->path.dentry->d_inode->i_sb->s_id,
-			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
+			req->wb_context->dentry->d_sb->s_id,
+			(long long)NFS_FILEID(req->wb_context->dentry->d_inode),
 			req->wb_bytes,
 			(long long)req_offset(req));
 		if (status < 0) {
diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig
index 18b3e8975fe..fbb2a5ef581 100644
--- a/fs/nfsd/Kconfig
+++ b/fs/nfsd/Kconfig
@@ -82,6 +82,7 @@ config NFSD_V4
 	select NFSD_V3
 	select FS_POSIX_ACL
 	select SUNRPC_GSS
+	select CRYPTO
 	help
 	  This option enables support in your system's NFS server for
 	  version 4 of the NFS protocol (RFC 3530).
diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index ffb59ef6f82..29d77f60585 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -191,52 +191,42 @@ nfsd4_build_namelist(void *arg, const char *name, int namlen,
 }
 
 static int
-nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f)
+nfsd4_list_rec_dir(recdir_func *f)
 {
 	const struct cred *original_cred;
-	struct file *filp;
+	struct dentry *dir = rec_file->f_path.dentry;
 	LIST_HEAD(names);
-	struct name_list *entry;
-	struct dentry *dentry;
 	int status;
 
-	if (!rec_file)
-		return 0;
-
 	status = nfs4_save_creds(&original_cred);
 	if (status < 0)
 		return status;
 
-	filp = dentry_open(dget(dir), mntget(rec_file->f_path.mnt), O_RDONLY,
-			   current_cred());
-	status = PTR_ERR(filp);
-	if (IS_ERR(filp))
-		goto out;
-	status = vfs_readdir(filp, nfsd4_build_namelist, &names);
-	fput(filp);
+	status = vfs_llseek(rec_file, 0, SEEK_SET);
+	if (status < 0) {
+		nfs4_reset_creds(original_cred);
+		return status;
+	}
+
+	status = vfs_readdir(rec_file, nfsd4_build_namelist, &names);
 	mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_PARENT);
 	while (!list_empty(&names)) {
+		struct name_list *entry;
 		entry = list_entry(names.next, struct name_list, list);
-
-		dentry = lookup_one_len(entry->name, dir, HEXDIR_LEN-1);
-		if (IS_ERR(dentry)) {
-			status = PTR_ERR(dentry);
-			break;
+		if (!status) {
+			struct dentry *dentry;
+			dentry = lookup_one_len(entry->name, dir, HEXDIR_LEN-1);
+			if (IS_ERR(dentry)) {
+				status = PTR_ERR(dentry);
+				break;
+			}
+			status = f(dir, dentry);
+			dput(dentry);
 		}
-		status = f(dir, dentry);
-		dput(dentry);
-		if (status)
-			break;
 		list_del(&entry->list);
 		kfree(entry);
 	}
 	mutex_unlock(&dir->d_inode->i_mutex);
-out:
-	while (!list_empty(&names)) {
-		entry = list_entry(names.next, struct name_list, list);
-		list_del(&entry->list);
-		kfree(entry);
-	}
 	nfs4_reset_creds(original_cred);
 	return status;
 }
@@ -322,7 +312,7 @@ nfsd4_recdir_purge_old(void) {
 	status = mnt_want_write(rec_file->f_path.mnt);
 	if (status)
 		goto out;
-	status = nfsd4_list_rec_dir(rec_file->f_path.dentry, purge_old);
+	status = nfsd4_list_rec_dir(purge_old);
 	if (status == 0)
 		vfs_fsync(rec_file, 0);
 	mnt_drop_write(rec_file->f_path.mnt);
@@ -352,7 +342,7 @@ nfsd4_recdir_load(void) {
 	if (!rec_file)
 		return 0;
 
-	status = nfsd4_list_rec_dir(rec_file->f_path.dentry, load_recdir);
+	status = nfsd4_list_rec_dir(load_recdir);
 	if (status)
 		printk("nfsd4: failed loading clients from recovery"
 			" directory %s\n", rec_file->f_path.dentry->d_name.name);
diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index 1f5eae40f34..2b1449dd2f4 100644
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -13,6 +13,7 @@
 #include <linux/lockd/lockd.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/gss_api.h>
+#include <linux/sunrpc/gss_krb5_enctypes.h>
 
 #include "idmap.h"
 #include "nfsd.h"
@@ -189,18 +190,10 @@ static struct file_operations export_features_operations = {
 	.release	= single_release,
 };
 
-#ifdef CONFIG_SUNRPC_GSS
+#if defined(CONFIG_SUNRPC_GSS) || defined(CONFIG_SUNRPC_GSS_MODULE)
 static int supported_enctypes_show(struct seq_file *m, void *v)
 {
-	struct gss_api_mech *k5mech;
-
-	k5mech = gss_mech_get_by_name("krb5");
-	if (k5mech == NULL)
-		goto out;
-	if (k5mech->gm_upcall_enctypes != NULL)
-		seq_printf(m, k5mech->gm_upcall_enctypes);
-	gss_mech_put(k5mech);
-out:
+	seq_printf(m, KRB5_SUPPORTED_ENCTYPES);
 	return 0;
 }
 
@@ -215,7 +208,7 @@ static struct file_operations supported_enctypes_ops = {
 	.llseek		= seq_lseek,
 	.release	= single_release,
 };
-#endif /* CONFIG_SUNRPC_GSS */
+#endif /* CONFIG_SUNRPC_GSS or CONFIG_SUNRPC_GSS_MODULE */
 
 extern int nfsd_pool_stats_open(struct inode *inode, struct file *file);
 extern int nfsd_pool_stats_release(struct inode *inode, struct file *file);
@@ -1427,9 +1420,9 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent)
 		[NFSD_Versions] = {"versions", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Ports] = {"portlist", &transaction_ops, S_IWUSR|S_IRUGO},
 		[NFSD_MaxBlkSize] = {"max_block_size", &transaction_ops, S_IWUSR|S_IRUGO},
-#ifdef CONFIG_SUNRPC_GSS
+#if defined(CONFIG_SUNRPC_GSS) || defined(CONFIG_SUNRPC_GSS_MODULE)
 		[NFSD_SupportedEnctypes] = {"supported_krb5_enctypes", &supported_enctypes_ops, S_IRUGO},
-#endif /* CONFIG_SUNRPC_GSS */
+#endif /* CONFIG_SUNRPC_GSS or CONFIG_SUNRPC_GSS_MODULE */
 #ifdef CONFIG_NFSD_V4
 		[NFSD_Leasetime] = {"nfsv4leasetime", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Gracetime] = {"nfsv4gracetime", &transaction_ops, S_IWUSR|S_IRUSR},
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index d5718273bb3..fd0acca5370 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -696,7 +696,15 @@ nfsd_access(struct svc_rqst *rqstp, struct svc_fh *fhp, u32 *access, u32 *suppor
 }
 #endif /* CONFIG_NFSD_V3 */
 
+static int nfsd_open_break_lease(struct inode *inode, int access)
+{
+	unsigned int mode;
 
+	if (access & NFSD_MAY_NOT_BREAK_LEASE)
+		return 0;
+	mode = (access & NFSD_MAY_WRITE) ? O_WRONLY : O_RDONLY;
+	return break_lease(inode, mode | O_NONBLOCK);
+}
 
 /*
  * Open an existing file or directory.
@@ -744,12 +752,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	if (!inode->i_fop)
 		goto out;
 
-	/*
-	 * Check to see if there are any leases on this file.
-	 * This may block while leases are broken.
-	 */
-	if (!(access & NFSD_MAY_NOT_BREAK_LEASE))
-		host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? O_WRONLY : 0));
+	host_err = nfsd_open_break_lease(inode, access);
 	if (host_err) /* NOMEM or WOULDBLOCK */
 		goto out_nfserr;
 
@@ -1660,8 +1663,10 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 	if (!dold->d_inode)
 		goto out_drop_write;
 	host_err = nfsd_break_lease(dold->d_inode);
-	if (host_err)
+	if (host_err) {
+		err = nfserrno(host_err);
 		goto out_drop_write;
+	}
 	host_err = vfs_link(dold, dirp, dnew);
 	if (!host_err) {
 		err = nfserrno(commit_metadata(ffhp));
diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c
index 7eafe468a29..b2e3ff34762 100644
--- a/fs/nilfs2/btree.c
+++ b/fs/nilfs2/btree.c
@@ -1346,6 +1346,11 @@ static void nilfs_btree_shrink(struct nilfs_bmap *btree,
 	path[level].bp_bh = NULL;
 }
 
+static void nilfs_btree_nop(struct nilfs_bmap *btree,
+			    struct nilfs_btree_path *path,
+			    int level, __u64 *keyp, __u64 *ptrp)
+{
+}
 
 static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree,
 				      struct nilfs_btree_path *path,
@@ -1356,20 +1361,19 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree,
 	struct buffer_head *bh;
 	struct nilfs_btree_node *node, *parent, *sib;
 	__u64 sibptr;
-	int pindex, level, ncmin, ncmax, ncblk, ret;
+	int pindex, dindex, level, ncmin, ncmax, ncblk, ret;
 
 	ret = 0;
 	stats->bs_nblocks = 0;
 	ncmin = NILFS_BTREE_NODE_NCHILDREN_MIN(nilfs_btree_node_size(btree));
 	ncblk = nilfs_btree_nchildren_per_block(btree);
 
-	for (level = NILFS_BTREE_LEVEL_NODE_MIN;
+	for (level = NILFS_BTREE_LEVEL_NODE_MIN, dindex = path[level].bp_index;
 	     level < nilfs_btree_height(btree) - 1;
 	     level++) {
 		node = nilfs_btree_get_nonroot_node(path, level);
 		path[level].bp_oldreq.bpr_ptr =
-			nilfs_btree_node_get_ptr(node, path[level].bp_index,
-						 ncblk);
+			nilfs_btree_node_get_ptr(node, dindex, ncblk);
 		ret = nilfs_bmap_prepare_end_ptr(btree,
 						 &path[level].bp_oldreq, dat);
 		if (ret < 0)
@@ -1383,6 +1387,7 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree,
 
 		parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax);
 		pindex = path[level + 1].bp_index;
+		dindex = pindex;
 
 		if (pindex > 0) {
 			/* left sibling */
@@ -1421,6 +1426,14 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree,
 				path[level].bp_sib_bh = bh;
 				path[level].bp_op = nilfs_btree_concat_right;
 				stats->bs_nblocks++;
+				/*
+				 * When merging right sibling node
+				 * into the current node, pointer to
+				 * the right sibling node must be
+				 * terminated instead.  The adjustment
+				 * below is required for that.
+				 */
+				dindex = pindex + 1;
 				/* continue; */
 			}
 		} else {
@@ -1431,29 +1444,31 @@ static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree,
 			    NILFS_BTREE_ROOT_NCHILDREN_MAX) {
 				path[level].bp_op = nilfs_btree_shrink;
 				stats->bs_nblocks += 2;
+				level++;
+				path[level].bp_op = nilfs_btree_nop;
+				goto shrink_root_child;
 			} else {
 				path[level].bp_op = nilfs_btree_do_delete;
 				stats->bs_nblocks++;
+				goto out;
 			}
-
-			goto out;
-
 		}
 	}
 
+	/* child of the root node is deleted */
+	path[level].bp_op = nilfs_btree_do_delete;
+	stats->bs_nblocks++;
+
+shrink_root_child:
 	node = nilfs_btree_get_root(btree);
 	path[level].bp_oldreq.bpr_ptr =
-		nilfs_btree_node_get_ptr(node, path[level].bp_index,
+		nilfs_btree_node_get_ptr(node, dindex,
 					 NILFS_BTREE_ROOT_NCHILDREN_MAX);
 
 	ret = nilfs_bmap_prepare_end_ptr(btree, &path[level].bp_oldreq, dat);
 	if (ret < 0)
 		goto err_out_child_node;
 
-	/* child of the root node is deleted */
-	path[level].bp_op = nilfs_btree_do_delete;
-	stats->bs_nblocks++;
-
 	/* success */
  out:
 	*levelp = level;
diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c
index d7eeca62feb..26601529dc1 100644
--- a/fs/nilfs2/file.c
+++ b/fs/nilfs2/file.c
@@ -27,7 +27,7 @@
 #include "nilfs.h"
 #include "segment.h"
 
-int nilfs_sync_file(struct file *file, int datasync)
+int nilfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	/*
 	 * Called from fsync() system call
@@ -40,8 +40,15 @@ int nilfs_sync_file(struct file *file, int datasync)
 	struct inode *inode = file->f_mapping->host;
 	int err;
 
-	if (!nilfs_inode_dirty(inode))
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&inode->i_mutex);
+
+	if (!nilfs_inode_dirty(inode)) {
+		mutex_unlock(&inode->i_mutex);
 		return 0;
+	}
 
 	if (datasync)
 		err = nilfs_construct_dsync_segment(inode->i_sb, inode, 0,
@@ -49,6 +56,7 @@ int nilfs_sync_file(struct file *file, int datasync)
 	else
 		err = nilfs_construct_segment(inode->i_sb);
 
+	mutex_unlock(&inode->i_mutex);
 	return err;
 }
 
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index b954878ad6c..666628b395f 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -259,8 +259,8 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
 		return 0;
 
 	/* Needs synchronization with the cleaner */
-	size = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs, nilfs_get_block, NULL);
+	size = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				  nilfs_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
@@ -778,6 +778,8 @@ int nilfs_setattr(struct dentry *dentry, struct iattr *iattr)
 
 	if ((iattr->ia_valid & ATTR_SIZE) &&
 	    iattr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		err = vmtruncate(inode, iattr->ia_size);
 		if (unlikely(err))
 			goto out_err;
@@ -799,19 +801,14 @@ out_err:
 	return err;
 }
 
-int nilfs_permission(struct inode *inode, int mask, unsigned int flags)
+int nilfs_permission(struct inode *inode, int mask)
 {
-	struct nilfs_root *root;
-
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
-	root = NILFS_I(inode)->i_root;
+	struct nilfs_root *root = NILFS_I(inode)->i_root;
 	if ((mask & MAY_WRITE) && root &&
 	    root->cno != NILFS_CPTREE_CURRENT_CNO)
 		return -EROFS; /* snapshot is not writable */
 
-	return generic_permission(inode, mask, flags, NULL);
+	return generic_permission(inode, mask);
 }
 
 int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh)
diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c
index 546849b3e88..a3141990061 100644
--- a/fs/nilfs2/namei.c
+++ b/fs/nilfs2/namei.c
@@ -72,12 +72,7 @@ nilfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
 		return ERR_PTR(-ENAMETOOLONG);
 
 	ino = nilfs_inode_by_name(dir, &dentry->d_name);
-	inode = NULL;
-	if (ino) {
-		inode = nilfs_iget(dir->i_sb, NILFS_I(dir)->i_root, ino);
-		if (IS_ERR(inode))
-			return ERR_CAST(inode);
-	}
+	inode = ino ? nilfs_iget(dir->i_sb, NILFS_I(dir)->i_root, ino) : NULL;
 	return d_splice_alias(inode, dentry);
 }
 
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index f02b9ad43a2..255d5e1c03b 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -235,7 +235,7 @@ extern void nilfs_set_link(struct inode *, struct nilfs_dir_entry *,
 			   struct page *, struct inode *);
 
 /* file.c */
-extern int nilfs_sync_file(struct file *, int);
+extern int nilfs_sync_file(struct file *, loff_t, loff_t, int);
 
 /* ioctl.c */
 long nilfs_ioctl(struct file *, unsigned int, unsigned long);
@@ -264,7 +264,7 @@ extern void nilfs_update_inode(struct inode *, struct buffer_head *);
 extern void nilfs_truncate(struct inode *);
 extern void nilfs_evict_inode(struct inode *);
 extern int nilfs_setattr(struct dentry *, struct iattr *);
-int nilfs_permission(struct inode *inode, int mask, unsigned int flags);
+int nilfs_permission(struct inode *inode, int mask);
 int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh);
 extern int nilfs_inode_dirty(struct inode *);
 int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty);
diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index 141646e88fb..bb24ab6c282 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -2573,7 +2573,7 @@ static struct nilfs_sc_info *nilfs_segctor_new(struct super_block *sb,
 	sci->sc_watermark = NILFS_SC_DEFAULT_WATERMARK;
 
 	if (nilfs->ns_interval)
-		sci->sc_interval = nilfs->ns_interval;
+		sci->sc_interval = HZ * nilfs->ns_interval;
 	if (nilfs->ns_watermark)
 		sci->sc_watermark = nilfs->ns_watermark;
 	return sci;
diff --git a/fs/ntfs/dir.c b/fs/ntfs/dir.c
index 0f48e7c5d9e..99e36107ff6 100644
--- a/fs/ntfs/dir.c
+++ b/fs/ntfs/dir.c
@@ -1527,13 +1527,20 @@ static int ntfs_dir_open(struct inode *vi, struct file *filp)
  * this problem for now.  We do write the $BITMAP attribute if it is present
  * which is the important one for a directory so things are not too bad.
  */
-static int ntfs_dir_fsync(struct file *filp, int datasync)
+static int ntfs_dir_fsync(struct file *filp, loff_t start, loff_t end,
+			  int datasync)
 {
 	struct inode *bmp_vi, *vi = filp->f_mapping->host;
 	int err, ret;
 	ntfs_attr na;
 
 	ntfs_debug("Entering for inode 0x%lx.", vi->i_ino);
+
+	err = filemap_write_and_wait_range(vi->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&vi->i_mutex);
+
 	BUG_ON(!S_ISDIR(vi->i_mode));
 	/* If the bitmap attribute inode is in memory sync it, too. */
 	na.mft_no = vi->i_ino;
@@ -1555,6 +1562,7 @@ static int ntfs_dir_fsync(struct file *filp, int datasync)
 	else
 		ntfs_warning(vi->i_sb, "Failed to f%ssync inode 0x%lx.  Error "
 				"%u.", datasync ? "data" : "", vi->i_ino, -ret);
+	mutex_unlock(&vi->i_mutex);
 	return ret;
 }
 
diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c
index f4b1057abdd..c587e2d2718 100644
--- a/fs/ntfs/file.c
+++ b/fs/ntfs/file.c
@@ -1832,9 +1832,8 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb,
 	 * fails again.
 	 */
 	if (unlikely(NInoTruncateFailed(ni))) {
-		down_write(&vi->i_alloc_sem);
+		inode_dio_wait(vi);
 		err = ntfs_truncate(vi);
-		up_write(&vi->i_alloc_sem);
 		if (err || NInoTruncateFailed(ni)) {
 			if (!err)
 				err = -EIO;
@@ -2153,12 +2152,19 @@ static ssize_t ntfs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
  * with this inode but since we have no simple way of getting to them we ignore
  * this problem for now.
  */
-static int ntfs_file_fsync(struct file *filp, int datasync)
+static int ntfs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			   int datasync)
 {
 	struct inode *vi = filp->f_mapping->host;
 	int err, ret = 0;
 
 	ntfs_debug("Entering for inode 0x%lx.", vi->i_ino);
+
+	err = filemap_write_and_wait_range(vi->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&vi->i_mutex);
+
 	BUG_ON(S_ISDIR(vi->i_mode));
 	if (!datasync || !NInoNonResident(NTFS_I(vi)))
 		ret = __ntfs_write_inode(vi, 1);
@@ -2176,6 +2182,7 @@ static int ntfs_file_fsync(struct file *filp, int datasync)
 	else
 		ntfs_warning(vi->i_sb, "Failed to f%ssync inode 0x%lx.  Error "
 				"%u.", datasync ? "data" : "", vi->i_ino, -ret);
+	mutex_unlock(&vi->i_mutex);
 	return ret;
 }
 
diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c
index c05d6dcf77a..1371487da95 100644
--- a/fs/ntfs/inode.c
+++ b/fs/ntfs/inode.c
@@ -2357,12 +2357,7 @@ static const char *es = "  Leaving inconsistent metadata.  Unmount and run "
  *
  * Returns 0 on success or -errno on error.
  *
- * Called with ->i_mutex held.  In all but one case ->i_alloc_sem is held for
- * writing.  The only case in the kernel where ->i_alloc_sem is not held is
- * mm/filemap.c::generic_file_buffered_write() where vmtruncate() is called
- * with the current i_size as the offset.  The analogous place in NTFS is in
- * fs/ntfs/file.c::ntfs_file_buffered_write() where we call vmtruncate() again
- * without holding ->i_alloc_sem.
+ * Called with ->i_mutex held.
  */
 int ntfs_truncate(struct inode *vi)
 {
@@ -2887,8 +2882,7 @@ void ntfs_truncate_vfs(struct inode *vi) {
  * We also abort all changes of user, group, and mode as we do not implement
  * the NTFS ACLs yet.
  *
- * Called with ->i_mutex held.  For the ATTR_SIZE (i.e. ->truncate) case, also
- * called with ->i_alloc_sem held for writing.
+ * Called with ->i_mutex held.
  */
 int ntfs_setattr(struct dentry *dentry, struct iattr *attr)
 {
diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c
index e913ad130fd..1cee970eb55 100644
--- a/fs/ocfs2/acl.c
+++ b/fs/ocfs2/acl.c
@@ -290,14 +290,14 @@ static int ocfs2_set_acl(handle_t *handle,
 	return ret;
 }
 
-int ocfs2_check_acl(struct inode *inode, int mask, unsigned int flags)
+int ocfs2_check_acl(struct inode *inode, int mask)
 {
 	struct ocfs2_super *osb;
 	struct buffer_head *di_bh = NULL;
 	struct posix_acl *acl;
 	int ret = -EAGAIN;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	osb = OCFS2_SB(inode->i_sb);
diff --git a/fs/ocfs2/acl.h b/fs/ocfs2/acl.h
index 4fe7c9cf4bf..5c5d31f0585 100644
--- a/fs/ocfs2/acl.h
+++ b/fs/ocfs2/acl.h
@@ -26,7 +26,7 @@ struct ocfs2_acl_entry {
 	__le32 e_id;
 };
 
-extern int ocfs2_check_acl(struct inode *, int, unsigned int);
+extern int ocfs2_check_acl(struct inode *, int);
 extern int ocfs2_acl_chmod(struct inode *);
 extern int ocfs2_init_acl(handle_t *, struct inode *, struct inode *,
 			  struct buffer_head *, struct buffer_head *,
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index ba3ca1e63b5..78b68af3b0e 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -559,9 +559,8 @@ bail:
 
 /*
  * ocfs2_dio_end_io is called by the dio core when a dio is finished.  We're
- * particularly interested in the aio/dio case.  Like the core uses
- * i_alloc_sem, we use the rw_lock DLM lock to protect io on one node from
- * truncation on another.
+ * particularly interested in the aio/dio case.  We use the rw_lock DLM lock
+ * to protect io on one node from truncation on another.
  */
 static void ocfs2_dio_end_io(struct kiocb *iocb,
 			     loff_t offset,
@@ -577,10 +576,8 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 	/* this io's submitter should not have unlocked this before we could */
 	BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
 
-	if (ocfs2_iocb_is_sem_locked(iocb)) {
-		up_read(&inode->i_alloc_sem);
+	if (ocfs2_iocb_is_sem_locked(iocb))
 		ocfs2_iocb_clear_sem_locked(iocb);
-	}
 
 	if (ocfs2_iocb_is_unaligned_aio(iocb)) {
 		ocfs2_iocb_clear_unaligned_aio(iocb);
@@ -598,6 +595,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 
 	if (is_async)
 		aio_complete(iocb, ret, 0);
+	inode_dio_done(inode);
 }
 
 /*
@@ -883,6 +881,12 @@ struct ocfs2_write_ctxt {
 	struct page			*w_target_page;
 
 	/*
+	 * w_target_locked is used for page_mkwrite path indicating no unlocking
+	 * against w_target_page in ocfs2_write_end_nolock.
+	 */
+	unsigned int			w_target_locked:1;
+
+	/*
 	 * ocfs2_write_end() uses this to know what the real range to
 	 * write in the target should be.
 	 */
@@ -915,6 +919,24 @@ void ocfs2_unlock_and_free_pages(struct page **pages, int num_pages)
 
 static void ocfs2_free_write_ctxt(struct ocfs2_write_ctxt *wc)
 {
+	int i;
+
+	/*
+	 * w_target_locked is only set to true in the page_mkwrite() case.
+	 * The intent is to allow us to lock the target page from write_begin()
+	 * to write_end(). The caller must hold a ref on w_target_page.
+	 */
+	if (wc->w_target_locked) {
+		BUG_ON(!wc->w_target_page);
+		for (i = 0; i < wc->w_num_pages; i++) {
+			if (wc->w_target_page == wc->w_pages[i]) {
+				wc->w_pages[i] = NULL;
+				break;
+			}
+		}
+		mark_page_accessed(wc->w_target_page);
+		page_cache_release(wc->w_target_page);
+	}
 	ocfs2_unlock_and_free_pages(wc->w_pages, wc->w_num_pages);
 
 	brelse(wc->w_di_bh);
@@ -1152,20 +1174,17 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping,
 			 */
 			lock_page(mmap_page);
 
+			/* Exit and let the caller retry */
 			if (mmap_page->mapping != mapping) {
+				WARN_ON(mmap_page->mapping);
 				unlock_page(mmap_page);
-				/*
-				 * Sanity check - the locking in
-				 * ocfs2_pagemkwrite() should ensure
-				 * that this code doesn't trigger.
-				 */
-				ret = -EINVAL;
-				mlog_errno(ret);
+				ret = -EAGAIN;
 				goto out;
 			}
 
 			page_cache_get(mmap_page);
 			wc->w_pages[i] = mmap_page;
+			wc->w_target_locked = true;
 		} else {
 			wc->w_pages[i] = find_or_create_page(mapping, index,
 							     GFP_NOFS);
@@ -1180,6 +1199,8 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping,
 			wc->w_target_page = wc->w_pages[i];
 	}
 out:
+	if (ret)
+		wc->w_target_locked = false;
 	return ret;
 }
 
@@ -1837,11 +1858,23 @@ try_again:
 	 */
 	ret = ocfs2_grab_pages_for_write(mapping, wc, wc->w_cpos, pos, len,
 					 cluster_of_pages, mmap_page);
-	if (ret) {
+	if (ret && ret != -EAGAIN) {
 		mlog_errno(ret);
 		goto out_quota;
 	}
 
+	/*
+	 * ocfs2_grab_pages_for_write() returns -EAGAIN if it could not lock
+	 * the target page. In this case, we exit with no error and no target
+	 * page. This will trigger the caller, page_mkwrite(), to re-try
+	 * the operation.
+	 */
+	if (ret == -EAGAIN) {
+		BUG_ON(wc->w_target_page);
+		ret = 0;
+		goto out_quota;
+	}
+
 	ret = ocfs2_write_cluster_by_desc(mapping, data_ac, meta_ac, wc, pos,
 					  len);
 	if (ret) {
diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c
index 9a3e6bbff27..a4e855e3690 100644
--- a/fs/ocfs2/cluster/heartbeat.c
+++ b/fs/ocfs2/cluster/heartbeat.c
@@ -216,6 +216,7 @@ struct o2hb_region {
 
 	struct list_head	hr_all_item;
 	unsigned		hr_unclean_stop:1,
+				hr_aborted_start:1,
 				hr_item_pinned:1,
 				hr_item_dropped:1;
 
@@ -254,6 +255,10 @@ struct o2hb_region {
 	 * a more complete api that doesn't lead to this sort of fragility. */
 	atomic_t		hr_steady_iterations;
 
+	/* terminate o2hb thread if it does not reach steady state
+	 * (hr_steady_iterations == 0) within hr_unsteady_iterations */
+	atomic_t		hr_unsteady_iterations;
+
 	char			hr_dev_name[BDEVNAME_SIZE];
 
 	unsigned int		hr_timeout_ms;
@@ -324,6 +329,10 @@ static void o2hb_write_timeout(struct work_struct *work)
 
 static void o2hb_arm_write_timeout(struct o2hb_region *reg)
 {
+	/* Arm writeout only after thread reaches steady state */
+	if (atomic_read(&reg->hr_steady_iterations) != 0)
+		return;
+
 	mlog(ML_HEARTBEAT, "Queue write timeout for %u ms\n",
 	     O2HB_MAX_WRITE_TIMEOUT_MS);
 
@@ -537,9 +546,14 @@ static int o2hb_verify_crc(struct o2hb_region *reg,
 	return read == computed;
 }
 
-/* We want to make sure that nobody is heartbeating on top of us --
- * this will help detect an invalid configuration. */
-static void o2hb_check_last_timestamp(struct o2hb_region *reg)
+/*
+ * Compare the slot data with what we wrote in the last iteration.
+ * If the match fails, print an appropriate error message. This is to
+ * detect errors like... another node hearting on the same slot,
+ * flaky device that is losing writes, etc.
+ * Returns 1 if check succeeds, 0 otherwise.
+ */
+static int o2hb_check_own_slot(struct o2hb_region *reg)
 {
 	struct o2hb_disk_slot *slot;
 	struct o2hb_disk_heartbeat_block *hb_block;
@@ -548,13 +562,13 @@ static void o2hb_check_last_timestamp(struct o2hb_region *reg)
 	slot = &reg->hr_slots[o2nm_this_node()];
 	/* Don't check on our 1st timestamp */
 	if (!slot->ds_last_time)
-		return;
+		return 0;
 
 	hb_block = slot->ds_raw_block;
 	if (le64_to_cpu(hb_block->hb_seq) == slot->ds_last_time &&
 	    le64_to_cpu(hb_block->hb_generation) == slot->ds_last_generation &&
 	    hb_block->hb_node == slot->ds_node_num)
-		return;
+		return 1;
 
 #define ERRSTR1		"Another node is heartbeating on device"
 #define ERRSTR2		"Heartbeat generation mismatch on device"
@@ -574,6 +588,8 @@ static void o2hb_check_last_timestamp(struct o2hb_region *reg)
 	     (unsigned long long)slot->ds_last_time, hb_block->hb_node,
 	     (unsigned long long)le64_to_cpu(hb_block->hb_generation),
 	     (unsigned long long)le64_to_cpu(hb_block->hb_seq));
+
+	return 0;
 }
 
 static inline void o2hb_prepare_block(struct o2hb_region *reg,
@@ -719,17 +735,24 @@ static void o2hb_shutdown_slot(struct o2hb_disk_slot *slot)
 	o2nm_node_put(node);
 }
 
-static void o2hb_set_quorum_device(struct o2hb_region *reg,
-				   struct o2hb_disk_slot *slot)
+static void o2hb_set_quorum_device(struct o2hb_region *reg)
 {
-	assert_spin_locked(&o2hb_live_lock);
-
 	if (!o2hb_global_heartbeat_active())
 		return;
 
-	if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+	/* Prevent race with o2hb_heartbeat_group_drop_item() */
+	if (kthread_should_stop())
+		return;
+
+	/* Tag region as quorum only after thread reaches steady state */
+	if (atomic_read(&reg->hr_steady_iterations) != 0)
 		return;
 
+	spin_lock(&o2hb_live_lock);
+
+	if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+		goto unlock;
+
 	/*
 	 * A region can be added to the quorum only when it sees all
 	 * live nodes heartbeat on it. In other words, the region has been
@@ -737,13 +760,10 @@ static void o2hb_set_quorum_device(struct o2hb_region *reg,
 	 */
 	if (memcmp(reg->hr_live_node_bitmap, o2hb_live_node_bitmap,
 		   sizeof(o2hb_live_node_bitmap)))
-		return;
-
-	if (slot->ds_changed_samples < O2HB_LIVE_THRESHOLD)
-		return;
+		goto unlock;
 
-	printk(KERN_NOTICE "o2hb: Region %s is now a quorum device\n",
-	       config_item_name(&reg->hr_item));
+	printk(KERN_NOTICE "o2hb: Region %s (%s) is now a quorum device\n",
+	       config_item_name(&reg->hr_item), reg->hr_dev_name);
 
 	set_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
 
@@ -754,6 +774,8 @@ static void o2hb_set_quorum_device(struct o2hb_region *reg,
 	if (o2hb_pop_count(&o2hb_quorum_region_bitmap,
 			   O2NM_MAX_REGIONS) > O2HB_PIN_CUT_OFF)
 		o2hb_region_unpin(NULL);
+unlock:
+	spin_unlock(&o2hb_live_lock);
 }
 
 static int o2hb_check_slot(struct o2hb_region *reg,
@@ -925,8 +947,6 @@ fire_callbacks:
 		slot->ds_equal_samples = 0;
 	}
 out:
-	o2hb_set_quorum_device(reg, slot);
-
 	spin_unlock(&o2hb_live_lock);
 
 	o2hb_run_event_list(&event);
@@ -957,7 +977,8 @@ static int o2hb_highest_node(unsigned long *nodes,
 
 static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 {
-	int i, ret, highest_node, change = 0;
+	int i, ret, highest_node;
+	int membership_change = 0, own_slot_ok = 0;
 	unsigned long configured_nodes[BITS_TO_LONGS(O2NM_MAX_NODES)];
 	unsigned long live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
 	struct o2hb_bio_wait_ctxt write_wc;
@@ -966,7 +987,7 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 				       sizeof(configured_nodes));
 	if (ret) {
 		mlog_errno(ret);
-		return ret;
+		goto bail;
 	}
 
 	/*
@@ -982,8 +1003,9 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 
 	highest_node = o2hb_highest_node(configured_nodes, O2NM_MAX_NODES);
 	if (highest_node >= O2NM_MAX_NODES) {
-		mlog(ML_NOTICE, "ocfs2_heartbeat: no configured nodes found!\n");
-		return -EINVAL;
+		mlog(ML_NOTICE, "o2hb: No configured nodes found!\n");
+		ret = -EINVAL;
+		goto bail;
 	}
 
 	/* No sense in reading the slots of nodes that don't exist
@@ -993,29 +1015,27 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 	ret = o2hb_read_slots(reg, highest_node + 1);
 	if (ret < 0) {
 		mlog_errno(ret);
-		return ret;
+		goto bail;
 	}
 
 	/* With an up to date view of the slots, we can check that no
 	 * other node has been improperly configured to heartbeat in
 	 * our slot. */
-	o2hb_check_last_timestamp(reg);
+	own_slot_ok = o2hb_check_own_slot(reg);
 
 	/* fill in the proper info for our next heartbeat */
 	o2hb_prepare_block(reg, reg->hr_generation);
 
-	/* And fire off the write. Note that we don't wait on this I/O
-	 * until later. */
 	ret = o2hb_issue_node_write(reg, &write_wc);
 	if (ret < 0) {
 		mlog_errno(ret);
-		return ret;
+		goto bail;
 	}
 
 	i = -1;
 	while((i = find_next_bit(configured_nodes,
 				 O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES) {
-		change |= o2hb_check_slot(reg, &reg->hr_slots[i]);
+		membership_change |= o2hb_check_slot(reg, &reg->hr_slots[i]);
 	}
 
 	/*
@@ -1030,18 +1050,39 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
 		 * disk */
 		mlog(ML_ERROR, "Write error %d on device \"%s\"\n",
 		     write_wc.wc_error, reg->hr_dev_name);
-		return write_wc.wc_error;
+		ret = write_wc.wc_error;
+		goto bail;
 	}
 
-	o2hb_arm_write_timeout(reg);
+	/* Skip disarming the timeout if own slot has stale/bad data */
+	if (own_slot_ok) {
+		o2hb_set_quorum_device(reg);
+		o2hb_arm_write_timeout(reg);
+	}
 
+bail:
 	/* let the person who launched us know when things are steady */
-	if (!change && (atomic_read(&reg->hr_steady_iterations) != 0)) {
-		if (atomic_dec_and_test(&reg->hr_steady_iterations))
+	if (atomic_read(&reg->hr_steady_iterations) != 0) {
+		if (!ret && own_slot_ok && !membership_change) {
+			if (atomic_dec_and_test(&reg->hr_steady_iterations))
+				wake_up(&o2hb_steady_queue);
+		}
+	}
+
+	if (atomic_read(&reg->hr_steady_iterations) != 0) {
+		if (atomic_dec_and_test(&reg->hr_unsteady_iterations)) {
+			printk(KERN_NOTICE "o2hb: Unable to stabilize "
+			       "heartbeart on region %s (%s)\n",
+			       config_item_name(&reg->hr_item),
+			       reg->hr_dev_name);
+			atomic_set(&reg->hr_steady_iterations, 0);
+			reg->hr_aborted_start = 1;
 			wake_up(&o2hb_steady_queue);
+			ret = -EIO;
+		}
 	}
 
-	return 0;
+	return ret;
 }
 
 /* Subtract b from a, storing the result in a. a *must* have a larger
@@ -1095,7 +1136,8 @@ static int o2hb_thread(void *data)
 	/* Pin node */
 	o2nm_depend_this_node();
 
-	while (!kthread_should_stop() && !reg->hr_unclean_stop) {
+	while (!kthread_should_stop() &&
+	       !reg->hr_unclean_stop && !reg->hr_aborted_start) {
 		/* We track the time spent inside
 		 * o2hb_do_disk_heartbeat so that we avoid more than
 		 * hr_timeout_ms between disk writes. On busy systems
@@ -1103,10 +1145,7 @@ static int o2hb_thread(void *data)
 		 * likely to time itself out. */
 		do_gettimeofday(&before_hb);
 
-		i = 0;
-		do {
-			ret = o2hb_do_disk_heartbeat(reg);
-		} while (ret && ++i < 2);
+		ret = o2hb_do_disk_heartbeat(reg);
 
 		do_gettimeofday(&after_hb);
 		elapsed_msec = o2hb_elapsed_msecs(&before_hb, &after_hb);
@@ -1117,7 +1156,8 @@ static int o2hb_thread(void *data)
 		     after_hb.tv_sec, (unsigned long) after_hb.tv_usec,
 		     elapsed_msec);
 
-		if (elapsed_msec < reg->hr_timeout_ms) {
+		if (!kthread_should_stop() &&
+		    elapsed_msec < reg->hr_timeout_ms) {
 			/* the kthread api has blocked signals for us so no
 			 * need to record the return value. */
 			msleep_interruptible(reg->hr_timeout_ms - elapsed_msec);
@@ -1134,20 +1174,20 @@ static int o2hb_thread(void *data)
 	 * to timeout on this region when we could just as easily
 	 * write a clear generation - thus indicating to them that
 	 * this node has left this region.
-	 *
-	 * XXX: Should we skip this on unclean_stop? */
-	o2hb_prepare_block(reg, 0);
-	ret = o2hb_issue_node_write(reg, &write_wc);
-	if (ret == 0) {
-		o2hb_wait_on_io(reg, &write_wc);
-	} else {
-		mlog_errno(ret);
+	 */
+	if (!reg->hr_unclean_stop && !reg->hr_aborted_start) {
+		o2hb_prepare_block(reg, 0);
+		ret = o2hb_issue_node_write(reg, &write_wc);
+		if (ret == 0)
+			o2hb_wait_on_io(reg, &write_wc);
+		else
+			mlog_errno(ret);
 	}
 
 	/* Unpin node */
 	o2nm_undepend_this_node();
 
-	mlog(ML_HEARTBEAT|ML_KTHREAD, "hb thread exiting\n");
+	mlog(ML_HEARTBEAT|ML_KTHREAD, "o2hb thread exiting\n");
 
 	return 0;
 }
@@ -1158,6 +1198,7 @@ static int o2hb_debug_open(struct inode *inode, struct file *file)
 	struct o2hb_debug_buf *db = inode->i_private;
 	struct o2hb_region *reg;
 	unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)];
+	unsigned long lts;
 	char *buf = NULL;
 	int i = -1;
 	int out = 0;
@@ -1194,9 +1235,11 @@ static int o2hb_debug_open(struct inode *inode, struct file *file)
 
 	case O2HB_DB_TYPE_REGION_ELAPSED_TIME:
 		reg = (struct o2hb_region *)db->db_data;
-		out += snprintf(buf + out, PAGE_SIZE - out, "%u\n",
-				jiffies_to_msecs(jiffies -
-						 reg->hr_last_timeout_start));
+		lts = reg->hr_last_timeout_start;
+		/* If 0, it has never been set before */
+		if (lts)
+			lts = jiffies_to_msecs(jiffies - lts);
+		out += snprintf(buf + out, PAGE_SIZE - out, "%lu\n", lts);
 		goto done;
 
 	case O2HB_DB_TYPE_REGION_PINNED:
@@ -1426,6 +1469,8 @@ static void o2hb_region_release(struct config_item *item)
 	struct page *page;
 	struct o2hb_region *reg = to_o2hb_region(item);
 
+	mlog(ML_HEARTBEAT, "hb region release (%s)\n", reg->hr_dev_name);
+
 	if (reg->hr_tmp_block)
 		kfree(reg->hr_tmp_block);
 
@@ -1792,7 +1837,10 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 			live_threshold <<= 1;
 		spin_unlock(&o2hb_live_lock);
 	}
-	atomic_set(&reg->hr_steady_iterations, live_threshold + 1);
+	++live_threshold;
+	atomic_set(&reg->hr_steady_iterations, live_threshold);
+	/* unsteady_iterations is double the steady_iterations */
+	atomic_set(&reg->hr_unsteady_iterations, (live_threshold << 1));
 
 	hb_task = kthread_run(o2hb_thread, reg, "o2hb-%s",
 			      reg->hr_item.ci_name);
@@ -1809,14 +1857,12 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 	ret = wait_event_interruptible(o2hb_steady_queue,
 				atomic_read(&reg->hr_steady_iterations) == 0);
 	if (ret) {
-		/* We got interrupted (hello ptrace!).  Clean up */
-		spin_lock(&o2hb_live_lock);
-		hb_task = reg->hr_task;
-		reg->hr_task = NULL;
-		spin_unlock(&o2hb_live_lock);
+		atomic_set(&reg->hr_steady_iterations, 0);
+		reg->hr_aborted_start = 1;
+	}
 
-		if (hb_task)
-			kthread_stop(hb_task);
+	if (reg->hr_aborted_start) {
+		ret = -EIO;
 		goto out;
 	}
 
@@ -1833,8 +1879,8 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
 		ret = -EIO;
 
 	if (hb_task && o2hb_global_heartbeat_active())
-		printk(KERN_NOTICE "o2hb: Heartbeat started on region %s\n",
-		       config_item_name(&reg->hr_item));
+		printk(KERN_NOTICE "o2hb: Heartbeat started on region %s (%s)\n",
+		       config_item_name(&reg->hr_item), reg->hr_dev_name);
 
 out:
 	if (filp)
@@ -2092,13 +2138,6 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group,
 
 	/* stop the thread when the user removes the region dir */
 	spin_lock(&o2hb_live_lock);
-	if (o2hb_global_heartbeat_active()) {
-		clear_bit(reg->hr_region_num, o2hb_region_bitmap);
-		clear_bit(reg->hr_region_num, o2hb_live_region_bitmap);
-		if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
-			quorum_region = 1;
-		clear_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
-	}
 	hb_task = reg->hr_task;
 	reg->hr_task = NULL;
 	reg->hr_item_dropped = 1;
@@ -2107,19 +2146,30 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group,
 	if (hb_task)
 		kthread_stop(hb_task);
 
+	if (o2hb_global_heartbeat_active()) {
+		spin_lock(&o2hb_live_lock);
+		clear_bit(reg->hr_region_num, o2hb_region_bitmap);
+		clear_bit(reg->hr_region_num, o2hb_live_region_bitmap);
+		if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+			quorum_region = 1;
+		clear_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
+		spin_unlock(&o2hb_live_lock);
+		printk(KERN_NOTICE "o2hb: Heartbeat %s on region %s (%s)\n",
+		       ((atomic_read(&reg->hr_steady_iterations) == 0) ?
+			"stopped" : "start aborted"), config_item_name(item),
+		       reg->hr_dev_name);
+	}
+
 	/*
 	 * If we're racing a dev_write(), we need to wake them.  They will
 	 * check reg->hr_task
 	 */
 	if (atomic_read(&reg->hr_steady_iterations) != 0) {
+		reg->hr_aborted_start = 1;
 		atomic_set(&reg->hr_steady_iterations, 0);
 		wake_up(&o2hb_steady_queue);
 	}
 
-	if (o2hb_global_heartbeat_active())
-		printk(KERN_NOTICE "o2hb: Heartbeat stopped on region %s\n",
-		       config_item_name(&reg->hr_item));
-
 	config_item_put(item);
 
 	if (!o2hb_global_heartbeat_active() || !quorum_region)
diff --git a/fs/ocfs2/cluster/netdebug.c b/fs/ocfs2/cluster/netdebug.c
index 3a5835904b3..dc45deb19e6 100644
--- a/fs/ocfs2/cluster/netdebug.c
+++ b/fs/ocfs2/cluster/netdebug.c
@@ -47,6 +47,7 @@
 #define SC_DEBUG_NAME		"sock_containers"
 #define NST_DEBUG_NAME		"send_tracking"
 #define STATS_DEBUG_NAME	"stats"
+#define NODES_DEBUG_NAME	"connected_nodes"
 
 #define SHOW_SOCK_CONTAINERS	0
 #define SHOW_SOCK_STATS		1
@@ -55,6 +56,7 @@ static struct dentry *o2net_dentry;
 static struct dentry *sc_dentry;
 static struct dentry *nst_dentry;
 static struct dentry *stats_dentry;
+static struct dentry *nodes_dentry;
 
 static DEFINE_SPINLOCK(o2net_debug_lock);
 
@@ -491,53 +493,87 @@ static const struct file_operations sc_seq_fops = {
 	.release = sc_fop_release,
 };
 
-int o2net_debugfs_init(void)
+static int o2net_fill_bitmap(char *buf, int len)
 {
-	o2net_dentry = debugfs_create_dir(O2NET_DEBUG_DIR, NULL);
-	if (!o2net_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)];
+	int i = -1, out = 0;
 
-	nst_dentry = debugfs_create_file(NST_DEBUG_NAME, S_IFREG|S_IRUSR,
-					 o2net_dentry, NULL,
-					 &nst_seq_fops);
-	if (!nst_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	o2net_fill_node_map(map, sizeof(map));
 
-	sc_dentry = debugfs_create_file(SC_DEBUG_NAME, S_IFREG|S_IRUSR,
-					o2net_dentry, NULL,
-					&sc_seq_fops);
-	if (!sc_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	while ((i = find_next_bit(map, O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES)
+		out += snprintf(buf + out, PAGE_SIZE - out, "%d ", i);
+	out += snprintf(buf + out, PAGE_SIZE - out, "\n");
 
-	stats_dentry = debugfs_create_file(STATS_DEBUG_NAME, S_IFREG|S_IRUSR,
-					   o2net_dentry, NULL,
-					   &stats_seq_fops);
-	if (!stats_dentry) {
-		mlog_errno(-ENOMEM);
-		goto bail;
-	}
+	return out;
+}
+
+static int nodes_fop_open(struct inode *inode, struct file *file)
+{
+	char *buf;
+
+	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	i_size_write(inode, o2net_fill_bitmap(buf, PAGE_SIZE));
+
+	file->private_data = buf;
 
 	return 0;
-bail:
-	debugfs_remove(stats_dentry);
-	debugfs_remove(sc_dentry);
-	debugfs_remove(nst_dentry);
-	debugfs_remove(o2net_dentry);
-	return -ENOMEM;
 }
 
+static int o2net_debug_release(struct inode *inode, struct file *file)
+{
+	kfree(file->private_data);
+	return 0;
+}
+
+static ssize_t o2net_debug_read(struct file *file, char __user *buf,
+				size_t nbytes, loff_t *ppos)
+{
+	return simple_read_from_buffer(buf, nbytes, ppos, file->private_data,
+				       i_size_read(file->f_mapping->host));
+}
+
+static const struct file_operations nodes_fops = {
+	.open		= nodes_fop_open,
+	.release	= o2net_debug_release,
+	.read		= o2net_debug_read,
+	.llseek		= generic_file_llseek,
+};
+
 void o2net_debugfs_exit(void)
 {
+	debugfs_remove(nodes_dentry);
 	debugfs_remove(stats_dentry);
 	debugfs_remove(sc_dentry);
 	debugfs_remove(nst_dentry);
 	debugfs_remove(o2net_dentry);
 }
 
+int o2net_debugfs_init(void)
+{
+	mode_t mode = S_IFREG|S_IRUSR;
+
+	o2net_dentry = debugfs_create_dir(O2NET_DEBUG_DIR, NULL);
+	if (o2net_dentry)
+		nst_dentry = debugfs_create_file(NST_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &nst_seq_fops);
+	if (nst_dentry)
+		sc_dentry = debugfs_create_file(SC_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &sc_seq_fops);
+	if (sc_dentry)
+		stats_dentry = debugfs_create_file(STATS_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &stats_seq_fops);
+	if (stats_dentry)
+		nodes_dentry = debugfs_create_file(NODES_DEBUG_NAME, mode,
+					o2net_dentry, NULL, &nodes_fops);
+	if (nodes_dentry)
+		return 0;
+
+	o2net_debugfs_exit();
+	mlog_errno(-ENOMEM);
+	return -ENOMEM;
+}
+
 #endif	/* CONFIG_DEBUG_FS */
diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c
index db5ee4b4f47..ae13d5ca790 100644
--- a/fs/ocfs2/cluster/tcp.c
+++ b/fs/ocfs2/cluster/tcp.c
@@ -545,7 +545,7 @@ static void o2net_set_nn_state(struct o2net_node *nn,
 	}
 
 	if (was_valid && !valid) {
-		printk(KERN_NOTICE "o2net: no longer connected to "
+		printk(KERN_NOTICE "o2net: No longer connected to "
 		       SC_NODEF_FMT "\n", SC_NODEF_ARGS(old_sc));
 		o2net_complete_nodes_nsw(nn);
 	}
@@ -555,7 +555,7 @@ static void o2net_set_nn_state(struct o2net_node *nn,
 		cancel_delayed_work(&nn->nn_connect_expired);
 		printk(KERN_NOTICE "o2net: %s " SC_NODEF_FMT "\n",
 		       o2nm_this_node() > sc->sc_node->nd_num ?
-		       		"connected to" : "accepted connection from",
+		       "Connected to" : "Accepted connection from",
 		       SC_NODEF_ARGS(sc));
 	}
 
@@ -643,7 +643,7 @@ static void o2net_state_change(struct sock *sk)
 			o2net_sc_queue_work(sc, &sc->sc_connect_work);
 			break;
 		default:
-			printk(KERN_INFO "o2net: connection to " SC_NODEF_FMT
+			printk(KERN_INFO "o2net: Connection to " SC_NODEF_FMT
 			      " shutdown, state %d\n",
 			      SC_NODEF_ARGS(sc), sk->sk_state);
 			o2net_sc_queue_work(sc, &sc->sc_shutdown_work);
@@ -1034,6 +1034,25 @@ static int o2net_tx_can_proceed(struct o2net_node *nn,
 	return ret;
 }
 
+/* Get a map of all nodes to which this node is currently connected to */
+void o2net_fill_node_map(unsigned long *map, unsigned bytes)
+{
+	struct o2net_sock_container *sc;
+	int node, ret;
+
+	BUG_ON(bytes < (BITS_TO_LONGS(O2NM_MAX_NODES) * sizeof(unsigned long)));
+
+	memset(map, 0, bytes);
+	for (node = 0; node < O2NM_MAX_NODES; ++node) {
+		o2net_tx_can_proceed(o2net_nn_from_num(node), &sc, &ret);
+		if (!ret) {
+			set_bit(node, map);
+			sc_put(sc);
+		}
+	}
+}
+EXPORT_SYMBOL_GPL(o2net_fill_node_map);
+
 int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec,
 			   size_t caller_veclen, u8 target_node, int *status)
 {
@@ -1284,11 +1303,11 @@ static int o2net_check_handshake(struct o2net_sock_container *sc)
 	struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
 
 	if (hand->protocol_version != cpu_to_be64(O2NET_PROTOCOL_VERSION)) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " advertised net protocol "
-		     "version %llu but %llu is required, disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     (unsigned long long)be64_to_cpu(hand->protocol_version),
-		     O2NET_PROTOCOL_VERSION);
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " Advertised net "
+		       "protocol version %llu but %llu is required. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       (unsigned long long)be64_to_cpu(hand->protocol_version),
+		       O2NET_PROTOCOL_VERSION);
 
 		/* don't bother reconnecting if its the wrong version. */
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
@@ -1302,33 +1321,33 @@ static int o2net_check_handshake(struct o2net_sock_container *sc)
 	 */
 	if (be32_to_cpu(hand->o2net_idle_timeout_ms) !=
 				o2net_idle_timeout()) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " uses a network idle timeout of "
-		     "%u ms, but we use %u ms locally.  disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     be32_to_cpu(hand->o2net_idle_timeout_ms),
-		     o2net_idle_timeout());
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a network "
+		       "idle timeout of %u ms, but we use %u ms locally. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       be32_to_cpu(hand->o2net_idle_timeout_ms),
+		       o2net_idle_timeout());
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
 		return -1;
 	}
 
 	if (be32_to_cpu(hand->o2net_keepalive_delay_ms) !=
 			o2net_keepalive_delay()) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " uses a keepalive delay of "
-		     "%u ms, but we use %u ms locally.  disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     be32_to_cpu(hand->o2net_keepalive_delay_ms),
-		     o2net_keepalive_delay());
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a keepalive "
+		       "delay of %u ms, but we use %u ms locally. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       be32_to_cpu(hand->o2net_keepalive_delay_ms),
+		       o2net_keepalive_delay());
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
 		return -1;
 	}
 
 	if (be32_to_cpu(hand->o2hb_heartbeat_timeout_ms) !=
 			O2HB_MAX_WRITE_TIMEOUT_MS) {
-		mlog(ML_NOTICE, SC_NODEF_FMT " uses a heartbeat timeout of "
-		     "%u ms, but we use %u ms locally.  disconnecting\n",
-		     SC_NODEF_ARGS(sc),
-		     be32_to_cpu(hand->o2hb_heartbeat_timeout_ms),
-		     O2HB_MAX_WRITE_TIMEOUT_MS);
+		printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a heartbeat "
+		       "timeout of %u ms, but we use %u ms locally. "
+		       "Disconnecting.\n", SC_NODEF_ARGS(sc),
+		       be32_to_cpu(hand->o2hb_heartbeat_timeout_ms),
+		       O2HB_MAX_WRITE_TIMEOUT_MS);
 		o2net_ensure_shutdown(nn, sc, -ENOTCONN);
 		return -1;
 	}
@@ -1539,28 +1558,16 @@ static void o2net_idle_timer(unsigned long data)
 {
 	struct o2net_sock_container *sc = (struct o2net_sock_container *)data;
 	struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
-
 #ifdef CONFIG_DEBUG_FS
-	ktime_t now = ktime_get();
+	unsigned long msecs = ktime_to_ms(ktime_get()) -
+		ktime_to_ms(sc->sc_tv_timer);
+#else
+	unsigned long msecs = o2net_idle_timeout();
 #endif
 
-	printk(KERN_NOTICE "o2net: connection to " SC_NODEF_FMT " has been idle for %u.%u "
-	     "seconds, shutting it down.\n", SC_NODEF_ARGS(sc),
-		     o2net_idle_timeout() / 1000,
-		     o2net_idle_timeout() % 1000);
-
-#ifdef CONFIG_DEBUG_FS
-	mlog(ML_NOTICE, "Here are some times that might help debug the "
-	     "situation: (Timer: %lld, Now %lld, DataReady %lld, Advance %lld-%lld, "
-	     "Key 0x%08x, Func %u, FuncTime %lld-%lld)\n",
-	     (long long)ktime_to_us(sc->sc_tv_timer), (long long)ktime_to_us(now),
-	     (long long)ktime_to_us(sc->sc_tv_data_ready),
-	     (long long)ktime_to_us(sc->sc_tv_advance_start),
-	     (long long)ktime_to_us(sc->sc_tv_advance_stop),
-	     sc->sc_msg_key, sc->sc_msg_type,
-	     (long long)ktime_to_us(sc->sc_tv_func_start),
-	     (long long)ktime_to_us(sc->sc_tv_func_stop));
-#endif
+	printk(KERN_NOTICE "o2net: Connection to " SC_NODEF_FMT " has been "
+	       "idle for %lu.%lu secs, shutting it down.\n", SC_NODEF_ARGS(sc),
+	       msecs / 1000, msecs % 1000);
 
 	/*
 	 * Initialize the nn_timeout so that the next connection attempt
@@ -1693,8 +1700,8 @@ static void o2net_start_connect(struct work_struct *work)
 
 out:
 	if (ret) {
-		mlog(ML_NOTICE, "connect attempt to " SC_NODEF_FMT " failed "
-		     "with errno %d\n", SC_NODEF_ARGS(sc), ret);
+		printk(KERN_NOTICE "o2net: Connect attempt to " SC_NODEF_FMT
+		       " failed with errno %d\n", SC_NODEF_ARGS(sc), ret);
 		/* 0 err so that another will be queued and attempted
 		 * from set_nn_state */
 		if (sc)
@@ -1717,8 +1724,8 @@ static void o2net_connect_expired(struct work_struct *work)
 
 	spin_lock(&nn->nn_lock);
 	if (!nn->nn_sc_valid) {
-		mlog(ML_ERROR, "no connection established with node %u after "
-		     "%u.%u seconds, giving up and returning errors.\n",
+		printk(KERN_NOTICE "o2net: No connection established with "
+		       "node %u after %u.%u seconds, giving up.\n",
 		     o2net_num_from_nn(nn),
 		     o2net_idle_timeout() / 1000,
 		     o2net_idle_timeout() % 1000);
@@ -1861,21 +1868,21 @@ static int o2net_accept_one(struct socket *sock)
 
 	node = o2nm_get_node_by_ip(sin.sin_addr.s_addr);
 	if (node == NULL) {
-		mlog(ML_NOTICE, "attempt to connect from unknown node at %pI4:%d\n",
-		     &sin.sin_addr.s_addr, ntohs(sin.sin_port));
+		printk(KERN_NOTICE "o2net: Attempt to connect from unknown "
+		       "node at %pI4:%d\n", &sin.sin_addr.s_addr,
+		       ntohs(sin.sin_port));
 		ret = -EINVAL;
 		goto out;
 	}
 
 	if (o2nm_this_node() >= node->nd_num) {
 		local_node = o2nm_get_node_by_num(o2nm_this_node());
-		mlog(ML_NOTICE, "unexpected connect attempt seen at node '%s' ("
-		     "%u, %pI4:%d) from node '%s' (%u, %pI4:%d)\n",
-		     local_node->nd_name, local_node->nd_num,
-		     &(local_node->nd_ipv4_address),
-		     ntohs(local_node->nd_ipv4_port),
-		     node->nd_name, node->nd_num, &sin.sin_addr.s_addr,
-		     ntohs(sin.sin_port));
+		printk(KERN_NOTICE "o2net: Unexpected connect attempt seen "
+		       "at node '%s' (%u, %pI4:%d) from node '%s' (%u, "
+		       "%pI4:%d)\n", local_node->nd_name, local_node->nd_num,
+		       &(local_node->nd_ipv4_address),
+		       ntohs(local_node->nd_ipv4_port), node->nd_name,
+		       node->nd_num, &sin.sin_addr.s_addr, ntohs(sin.sin_port));
 		ret = -EINVAL;
 		goto out;
 	}
@@ -1900,10 +1907,10 @@ static int o2net_accept_one(struct socket *sock)
 		ret = 0;
 	spin_unlock(&nn->nn_lock);
 	if (ret) {
-		mlog(ML_NOTICE, "attempt to connect from node '%s' at "
-		     "%pI4:%d but it already has an open connection\n",
-		     node->nd_name, &sin.sin_addr.s_addr,
-		     ntohs(sin.sin_port));
+		printk(KERN_NOTICE "o2net: Attempt to connect from node '%s' "
+		       "at %pI4:%d but it already has an open connection\n",
+		       node->nd_name, &sin.sin_addr.s_addr,
+		       ntohs(sin.sin_port));
 		goto out;
 	}
 
@@ -1983,7 +1990,7 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port)
 
 	ret = sock_create(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
 	if (ret < 0) {
-		mlog(ML_ERROR, "unable to create socket, ret=%d\n", ret);
+		printk(KERN_ERR "o2net: Error %d while creating socket\n", ret);
 		goto out;
 	}
 
@@ -2000,16 +2007,15 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port)
 	sock->sk->sk_reuse = 1;
 	ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin));
 	if (ret < 0) {
-		mlog(ML_ERROR, "unable to bind socket at %pI4:%u, "
-		     "ret=%d\n", &addr, ntohs(port), ret);
+		printk(KERN_ERR "o2net: Error %d while binding socket at "
+		       "%pI4:%u\n", ret, &addr, ntohs(port)); 
 		goto out;
 	}
 
 	ret = sock->ops->listen(sock, 64);
-	if (ret < 0) {
-		mlog(ML_ERROR, "unable to listen on %pI4:%u, ret=%d\n",
-		     &addr, ntohs(port), ret);
-	}
+	if (ret < 0)
+		printk(KERN_ERR "o2net: Error %d while listening on %pI4:%u\n",
+		       ret, &addr, ntohs(port));
 
 out:
 	if (ret) {
diff --git a/fs/ocfs2/cluster/tcp.h b/fs/ocfs2/cluster/tcp.h
index fd6179eb26d..5bada2a69b5 100644
--- a/fs/ocfs2/cluster/tcp.h
+++ b/fs/ocfs2/cluster/tcp.h
@@ -106,6 +106,8 @@ int o2net_register_handler(u32 msg_type, u32 key, u32 max_len,
 			   struct list_head *unreg_list);
 void o2net_unregister_handler_list(struct list_head *list);
 
+void o2net_fill_node_map(unsigned long *map, unsigned bytes);
+
 struct o2nm_node;
 int o2net_register_hb_callbacks(void);
 void o2net_unregister_hb_callbacks(void);
diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h
index d602abb51b6..a5952ceecba 100644
--- a/fs/ocfs2/dlm/dlmcommon.h
+++ b/fs/ocfs2/dlm/dlmcommon.h
@@ -859,8 +859,8 @@ void dlm_complete_recovery_thread(struct dlm_ctxt *dlm);
 void dlm_wait_for_recovery(struct dlm_ctxt *dlm);
 void dlm_kick_recovery_thread(struct dlm_ctxt *dlm);
 int dlm_is_node_dead(struct dlm_ctxt *dlm, u8 node);
-int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout);
-int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout);
+void dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout);
+void dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout);
 
 void dlm_put(struct dlm_ctxt *dlm);
 struct dlm_ctxt *dlm_grab(struct dlm_ctxt *dlm);
@@ -877,9 +877,8 @@ static inline void dlm_lockres_get(struct dlm_lock_resource *res)
 	kref_get(&res->refs);
 }
 void dlm_lockres_put(struct dlm_lock_resource *res);
-void __dlm_unhash_lockres(struct dlm_lock_resource *res);
-void __dlm_insert_lockres(struct dlm_ctxt *dlm,
-			  struct dlm_lock_resource *res);
+void __dlm_unhash_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res);
+void __dlm_insert_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res);
 struct dlm_lock_resource * __dlm_lookup_lockres_full(struct dlm_ctxt *dlm,
 						     const char *name,
 						     unsigned int len,
@@ -902,46 +901,15 @@ struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm,
 					  const char *name,
 					  unsigned int namelen);
 
-#define dlm_lockres_set_refmap_bit(bit,res)  \
-	__dlm_lockres_set_refmap_bit(bit,res,__FILE__,__LINE__)
-#define dlm_lockres_clear_refmap_bit(bit,res)  \
-	__dlm_lockres_clear_refmap_bit(bit,res,__FILE__,__LINE__)
+void dlm_lockres_set_refmap_bit(struct dlm_ctxt *dlm,
+				struct dlm_lock_resource *res, int bit);
+void dlm_lockres_clear_refmap_bit(struct dlm_ctxt *dlm,
+				  struct dlm_lock_resource *res, int bit);
 
-static inline void __dlm_lockres_set_refmap_bit(int bit,
-						struct dlm_lock_resource *res,
-						const char *file,
-						int line)
-{
-	//printk("%s:%d:%.*s: setting bit %d\n", file, line,
-	//     res->lockname.len, res->lockname.name, bit);
-	set_bit(bit, res->refmap);
-}
-
-static inline void __dlm_lockres_clear_refmap_bit(int bit,
-						  struct dlm_lock_resource *res,
-						  const char *file,
-						  int line)
-{
-	//printk("%s:%d:%.*s: clearing bit %d\n", file, line,
-	//     res->lockname.len, res->lockname.name, bit);
-	clear_bit(bit, res->refmap);
-}
-
-void __dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   const char *file,
-				   int line);
-void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   int new_lockres,
-				   const char *file,
-				   int line);
-#define dlm_lockres_drop_inflight_ref(d,r)  \
-	__dlm_lockres_drop_inflight_ref(d,r,__FILE__,__LINE__)
-#define dlm_lockres_grab_inflight_ref(d,r)  \
-	__dlm_lockres_grab_inflight_ref(d,r,0,__FILE__,__LINE__)
-#define dlm_lockres_grab_inflight_ref_new(d,r)  \
-	__dlm_lockres_grab_inflight_ref(d,r,1,__FILE__,__LINE__)
+void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res);
+void dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res);
 
 void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
 void dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c
index 6ed6b95dcf9..92f2ead0fab 100644
--- a/fs/ocfs2/dlm/dlmdomain.c
+++ b/fs/ocfs2/dlm/dlmdomain.c
@@ -157,16 +157,18 @@ static int dlm_protocol_compare(struct dlm_protocol_version *existing,
 
 static void dlm_unregister_domain_handlers(struct dlm_ctxt *dlm);
 
-void __dlm_unhash_lockres(struct dlm_lock_resource *lockres)
+void __dlm_unhash_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 {
-	if (!hlist_unhashed(&lockres->hash_node)) {
-		hlist_del_init(&lockres->hash_node);
-		dlm_lockres_put(lockres);
-	}
+	if (hlist_unhashed(&res->hash_node))
+		return;
+
+	mlog(0, "%s: Unhash res %.*s\n", dlm->name, res->lockname.len,
+	     res->lockname.name);
+	hlist_del_init(&res->hash_node);
+	dlm_lockres_put(res);
 }
 
-void __dlm_insert_lockres(struct dlm_ctxt *dlm,
-		       struct dlm_lock_resource *res)
+void __dlm_insert_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 {
 	struct hlist_head *bucket;
 	struct qstr *q;
@@ -180,6 +182,9 @@ void __dlm_insert_lockres(struct dlm_ctxt *dlm,
 	dlm_lockres_get(res);
 
 	hlist_add_head(&res->hash_node, bucket);
+
+	mlog(0, "%s: Hash res %.*s\n", dlm->name, res->lockname.len,
+	     res->lockname.name);
 }
 
 struct dlm_lock_resource * __dlm_lookup_lockres_full(struct dlm_ctxt *dlm,
@@ -539,17 +544,17 @@ again:
 
 static void __dlm_print_nodes(struct dlm_ctxt *dlm)
 {
-	int node = -1;
+	int node = -1, num = 0;
 
 	assert_spin_locked(&dlm->spinlock);
 
-	printk(KERN_NOTICE "o2dlm: Nodes in domain %s: ", dlm->name);
-
+	printk("( ");
 	while ((node = find_next_bit(dlm->domain_map, O2NM_MAX_NODES,
 				     node + 1)) < O2NM_MAX_NODES) {
 		printk("%d ", node);
+		++num;
 	}
-	printk("\n");
+	printk(") %u nodes\n", num);
 }
 
 static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data,
@@ -566,11 +571,10 @@ static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data,
 
 	node = exit_msg->node_idx;
 
-	printk(KERN_NOTICE "o2dlm: Node %u leaves domain %s\n", node, dlm->name);
-
 	spin_lock(&dlm->spinlock);
 	clear_bit(node, dlm->domain_map);
 	clear_bit(node, dlm->exit_domain_map);
+	printk(KERN_NOTICE "o2dlm: Node %u leaves domain %s ", node, dlm->name);
 	__dlm_print_nodes(dlm);
 
 	/* notify anything attached to the heartbeat events */
@@ -755,6 +759,7 @@ void dlm_unregister_domain(struct dlm_ctxt *dlm)
 
 		dlm_mark_domain_leaving(dlm);
 		dlm_leave_domain(dlm);
+		printk(KERN_NOTICE "o2dlm: Leaving domain %s\n", dlm->name);
 		dlm_force_free_mles(dlm);
 		dlm_complete_dlm_shutdown(dlm);
 	}
@@ -970,7 +975,7 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data,
 		clear_bit(assert->node_idx, dlm->exit_domain_map);
 		__dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN);
 
-		printk(KERN_NOTICE "o2dlm: Node %u joins domain %s\n",
+		printk(KERN_NOTICE "o2dlm: Node %u joins domain %s ",
 		       assert->node_idx, dlm->name);
 		__dlm_print_nodes(dlm);
 
@@ -1701,8 +1706,10 @@ static int dlm_try_to_join_domain(struct dlm_ctxt *dlm)
 bail:
 	spin_lock(&dlm->spinlock);
 	__dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN);
-	if (!status)
+	if (!status) {
+		printk(KERN_NOTICE "o2dlm: Joining domain %s ", dlm->name);
 		__dlm_print_nodes(dlm);
+	}
 	spin_unlock(&dlm->spinlock);
 
 	if (ctxt) {
@@ -2131,13 +2138,6 @@ struct dlm_ctxt * dlm_register_domain(const char *domain,
 		goto leave;
 	}
 
-	if (!o2hb_check_local_node_heartbeating()) {
-		mlog(ML_ERROR, "the local node has not been configured, or is "
-		     "not heartbeating\n");
-		ret = -EPROTO;
-		goto leave;
-	}
-
 	mlog(0, "register called for domain \"%s\"\n", domain);
 
 retry:
diff --git a/fs/ocfs2/dlm/dlmlock.c b/fs/ocfs2/dlm/dlmlock.c
index 8d39e0fd66f..f32fcba0492 100644
--- a/fs/ocfs2/dlm/dlmlock.c
+++ b/fs/ocfs2/dlm/dlmlock.c
@@ -183,10 +183,6 @@ static enum dlm_status dlmlock_master(struct dlm_ctxt *dlm,
 			kick_thread = 1;
 		}
 	}
-	/* reduce the inflight count, this may result in the lockres
-	 * being purged below during calc_usage */
-	if (lock->ml.node == dlm->node_num)
-		dlm_lockres_drop_inflight_ref(dlm, res);
 
 	spin_unlock(&res->spinlock);
 	wake_up(&res->wq);
@@ -231,10 +227,16 @@ static enum dlm_status dlmlock_remote(struct dlm_ctxt *dlm,
 	     lock->ml.type, res->lockname.len,
 	     res->lockname.name, flags);
 
+	/*
+	 * Wait if resource is getting recovered, remastered, etc.
+	 * If the resource was remastered and new owner is self, then exit.
+	 */
 	spin_lock(&res->spinlock);
-
-	/* will exit this call with spinlock held */
 	__dlm_wait_on_lockres(res);
+	if (res->owner == dlm->node_num) {
+		spin_unlock(&res->spinlock);
+		return DLM_RECOVERING;
+	}
 	res->state |= DLM_LOCK_RES_IN_PROGRESS;
 
 	/* add lock to local (secondary) queue */
@@ -319,27 +321,23 @@ static enum dlm_status dlm_send_remote_lock_request(struct dlm_ctxt *dlm,
 	tmpret = o2net_send_message(DLM_CREATE_LOCK_MSG, dlm->key, &create,
 				    sizeof(create), res->owner, &status);
 	if (tmpret >= 0) {
-		// successfully sent and received
-		ret = status;  // this is already a dlm_status
+		ret = status;
 		if (ret == DLM_REJECTED) {
-			mlog(ML_ERROR, "%s:%.*s: BUG.  this is a stale lockres "
-			     "no longer owned by %u.  that node is coming back "
-			     "up currently.\n", dlm->name, create.namelen,
+			mlog(ML_ERROR, "%s: res %.*s, Stale lockres no longer "
+			     "owned by node %u. That node is coming back up "
+			     "currently.\n", dlm->name, create.namelen,
 			     create.name, res->owner);
 			dlm_print_one_lock_resource(res);
 			BUG();
 		}
 	} else {
-		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
-		     "node %u\n", tmpret, DLM_CREATE_LOCK_MSG, dlm->key,
-		     res->owner);
-		if (dlm_is_host_down(tmpret)) {
+		mlog(ML_ERROR, "%s: res %.*s, Error %d send CREATE LOCK to "
+		     "node %u\n", dlm->name, create.namelen, create.name,
+		     tmpret, res->owner);
+		if (dlm_is_host_down(tmpret))
 			ret = DLM_RECOVERING;
-			mlog(0, "node %u died so returning DLM_RECOVERING "
-			     "from lock message!\n", res->owner);
-		} else {
+		else
 			ret = dlm_err_to_dlm_status(tmpret);
-		}
 	}
 
 	return ret;
@@ -718,18 +716,10 @@ retry_lock:
 
 		if (status == DLM_RECOVERING || status == DLM_MIGRATING ||
 		    status == DLM_FORWARD) {
-			mlog(0, "retrying lock with migration/"
-			     "recovery/in progress\n");
 			msleep(100);
-			/* no waiting for dlm_reco_thread */
 			if (recovery) {
 				if (status != DLM_RECOVERING)
 					goto retry_lock;
-
-				mlog(0, "%s: got RECOVERING "
-				     "for $RECOVERY lock, master "
-				     "was %u\n", dlm->name,
-				     res->owner);
 				/* wait to see the node go down, then
 				 * drop down and allow the lockres to
 				 * get cleaned up.  need to remaster. */
@@ -741,6 +731,14 @@ retry_lock:
 			}
 		}
 
+		/* Inflight taken in dlm_get_lock_resource() is dropped here */
+		spin_lock(&res->spinlock);
+		dlm_lockres_drop_inflight_ref(dlm, res);
+		spin_unlock(&res->spinlock);
+
+		dlm_lockres_calc_usage(dlm, res);
+		dlm_kick_thread(dlm, res);
+
 		if (status != DLM_NORMAL) {
 			lock->lksb->flags &= ~DLM_LKSB_GET_LVB;
 			if (status != DLM_NOTQUEUED)
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index 11eefb8c12e..005261c333b 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -631,39 +631,54 @@ error:
 	return NULL;
 }
 
-void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   int new_lockres,
-				   const char *file,
-				   int line)
+void dlm_lockres_set_refmap_bit(struct dlm_ctxt *dlm,
+				struct dlm_lock_resource *res, int bit)
 {
-	if (!new_lockres)
-		assert_spin_locked(&res->spinlock);
+	assert_spin_locked(&res->spinlock);
+
+	mlog(0, "res %.*s, set node %u, %ps()\n", res->lockname.len,
+	     res->lockname.name, bit, __builtin_return_address(0));
+
+	set_bit(bit, res->refmap);
+}
+
+void dlm_lockres_clear_refmap_bit(struct dlm_ctxt *dlm,
+				  struct dlm_lock_resource *res, int bit)
+{
+	assert_spin_locked(&res->spinlock);
+
+	mlog(0, "res %.*s, clr node %u, %ps()\n", res->lockname.len,
+	     res->lockname.name, bit, __builtin_return_address(0));
+
+	clear_bit(bit, res->refmap);
+}
+
+
+void dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res)
+{
+	assert_spin_locked(&res->spinlock);
 
-	if (!test_bit(dlm->node_num, res->refmap)) {
-		BUG_ON(res->inflight_locks != 0);
-		dlm_lockres_set_refmap_bit(dlm->node_num, res);
-	}
 	res->inflight_locks++;
-	mlog(0, "%s:%.*s: inflight++: now %u\n",
-	     dlm->name, res->lockname.len, res->lockname.name,
-	     res->inflight_locks);
+
+	mlog(0, "%s: res %.*s, inflight++: now %u, %ps()\n", dlm->name,
+	     res->lockname.len, res->lockname.name, res->inflight_locks,
+	     __builtin_return_address(0));
 }
 
-void __dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
-				   struct dlm_lock_resource *res,
-				   const char *file,
-				   int line)
+void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
+				   struct dlm_lock_resource *res)
 {
 	assert_spin_locked(&res->spinlock);
 
 	BUG_ON(res->inflight_locks == 0);
+
 	res->inflight_locks--;
-	mlog(0, "%s:%.*s: inflight--: now %u\n",
-	     dlm->name, res->lockname.len, res->lockname.name,
-	     res->inflight_locks);
-	if (res->inflight_locks == 0)
-		dlm_lockres_clear_refmap_bit(dlm->node_num, res);
+
+	mlog(0, "%s: res %.*s, inflight--: now %u, %ps()\n", dlm->name,
+	     res->lockname.len, res->lockname.name, res->inflight_locks,
+	     __builtin_return_address(0));
+
 	wake_up(&res->wq);
 }
 
@@ -697,7 +712,6 @@ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
 	unsigned int hash;
 	int tries = 0;
 	int bit, wait_on_recovery = 0;
-	int drop_inflight_if_nonlocal = 0;
 
 	BUG_ON(!lockid);
 
@@ -709,36 +723,33 @@ lookup:
 	spin_lock(&dlm->spinlock);
 	tmpres = __dlm_lookup_lockres_full(dlm, lockid, namelen, hash);
 	if (tmpres) {
-		int dropping_ref = 0;
-
 		spin_unlock(&dlm->spinlock);
-
 		spin_lock(&tmpres->spinlock);
-		/* We wait for the other thread that is mastering the resource */
+		/* Wait on the thread that is mastering the resource */
 		if (tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN) {
 			__dlm_wait_on_lockres(tmpres);
 			BUG_ON(tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN);
+			spin_unlock(&tmpres->spinlock);
+			dlm_lockres_put(tmpres);
+			tmpres = NULL;
+			goto lookup;
 		}
 
-		if (tmpres->owner == dlm->node_num) {
-			BUG_ON(tmpres->state & DLM_LOCK_RES_DROPPING_REF);
-			dlm_lockres_grab_inflight_ref(dlm, tmpres);
-		} else if (tmpres->state & DLM_LOCK_RES_DROPPING_REF)
-			dropping_ref = 1;
-		spin_unlock(&tmpres->spinlock);
-
-		/* wait until done messaging the master, drop our ref to allow
-		 * the lockres to be purged, start over. */
-		if (dropping_ref) {
-			spin_lock(&tmpres->spinlock);
-			__dlm_wait_on_lockres_flags(tmpres, DLM_LOCK_RES_DROPPING_REF);
+		/* Wait on the resource purge to complete before continuing */
+		if (tmpres->state & DLM_LOCK_RES_DROPPING_REF) {
+			BUG_ON(tmpres->owner == dlm->node_num);
+			__dlm_wait_on_lockres_flags(tmpres,
+						    DLM_LOCK_RES_DROPPING_REF);
 			spin_unlock(&tmpres->spinlock);
 			dlm_lockres_put(tmpres);
 			tmpres = NULL;
 			goto lookup;
 		}
 
-		mlog(0, "found in hash!\n");
+		/* Grab inflight ref to pin the resource */
+		dlm_lockres_grab_inflight_ref(dlm, tmpres);
+
+		spin_unlock(&tmpres->spinlock);
 		if (res)
 			dlm_lockres_put(res);
 		res = tmpres;
@@ -829,8 +840,8 @@ lookup:
 		 * but they might own this lockres.  wait on them. */
 		bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
 		if (bit < O2NM_MAX_NODES) {
-			mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
-			     "recover before lock mastery can begin\n",
+			mlog(0, "%s: res %.*s, At least one node (%d) "
+			     "to recover before lock mastery can begin\n",
 			     dlm->name, namelen, (char *)lockid, bit);
 			wait_on_recovery = 1;
 		}
@@ -843,12 +854,11 @@ lookup:
 
 	/* finally add the lockres to its hash bucket */
 	__dlm_insert_lockres(dlm, res);
-	/* since this lockres is new it doesn't not require the spinlock */
-	dlm_lockres_grab_inflight_ref_new(dlm, res);
 
-	/* if this node does not become the master make sure to drop
-	 * this inflight reference below */
-	drop_inflight_if_nonlocal = 1;
+	/* Grab inflight ref to pin the resource */
+	spin_lock(&res->spinlock);
+	dlm_lockres_grab_inflight_ref(dlm, res);
+	spin_unlock(&res->spinlock);
 
 	/* get an extra ref on the mle in case this is a BLOCK
 	 * if so, the creator of the BLOCK may try to put the last
@@ -864,8 +874,8 @@ redo_request:
 		 * dlm spinlock would be detectable be a change on the mle,
 		 * so we only need to clear out the recovery map once. */
 		if (dlm_is_recovery_lock(lockid, namelen)) {
-			mlog(ML_NOTICE, "%s: recovery map is not empty, but "
-			     "must master $RECOVERY lock now\n", dlm->name);
+			mlog(0, "%s: Recovery map is not empty, but must "
+			     "master $RECOVERY lock now\n", dlm->name);
 			if (!dlm_pre_master_reco_lockres(dlm, res))
 				wait_on_recovery = 0;
 			else {
@@ -883,8 +893,8 @@ redo_request:
 		spin_lock(&dlm->spinlock);
 		bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
 		if (bit < O2NM_MAX_NODES) {
-			mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
-			     "recover before lock mastery can begin\n",
+			mlog(0, "%s: res %.*s, At least one node (%d) "
+			     "to recover before lock mastery can begin\n",
 			     dlm->name, namelen, (char *)lockid, bit);
 			wait_on_recovery = 1;
 		} else
@@ -913,8 +923,8 @@ redo_request:
 			 * yet, keep going until it does.  this is how the
 			 * master will know that asserts are needed back to
 			 * the lower nodes. */
-			mlog(0, "%s:%.*s: requests only up to %u but master "
-			     "is %u, keep going\n", dlm->name, namelen,
+			mlog(0, "%s: res %.*s, Requests only up to %u but "
+			     "master is %u, keep going\n", dlm->name, namelen,
 			     lockid, nodenum, mle->master);
 		}
 	}
@@ -924,13 +934,12 @@ wait:
 	ret = dlm_wait_for_lock_mastery(dlm, res, mle, &blocked);
 	if (ret < 0) {
 		wait_on_recovery = 1;
-		mlog(0, "%s:%.*s: node map changed, redo the "
-		     "master request now, blocked=%d\n",
-		     dlm->name, res->lockname.len,
+		mlog(0, "%s: res %.*s, Node map changed, redo the master "
+		     "request now, blocked=%d\n", dlm->name, res->lockname.len,
 		     res->lockname.name, blocked);
 		if (++tries > 20) {
-			mlog(ML_ERROR, "%s:%.*s: spinning on "
-			     "dlm_wait_for_lock_mastery, blocked=%d\n",
+			mlog(ML_ERROR, "%s: res %.*s, Spinning on "
+			     "dlm_wait_for_lock_mastery, blocked = %d\n",
 			     dlm->name, res->lockname.len,
 			     res->lockname.name, blocked);
 			dlm_print_one_lock_resource(res);
@@ -940,7 +949,8 @@ wait:
 		goto redo_request;
 	}
 
-	mlog(0, "lockres mastered by %u\n", res->owner);
+	mlog(0, "%s: res %.*s, Mastered by %u\n", dlm->name, res->lockname.len,
+	     res->lockname.name, res->owner);
 	/* make sure we never continue without this */
 	BUG_ON(res->owner == O2NM_MAX_NODES);
 
@@ -952,8 +962,6 @@ wait:
 
 wake_waiters:
 	spin_lock(&res->spinlock);
-	if (res->owner != dlm->node_num && drop_inflight_if_nonlocal)
-		dlm_lockres_drop_inflight_ref(dlm, res);
 	res->state &= ~DLM_LOCK_RES_IN_PROGRESS;
 	spin_unlock(&res->spinlock);
 	wake_up(&res->wq);
@@ -1426,9 +1434,7 @@ way_up_top:
 		}
 
 		if (res->owner == dlm->node_num) {
-			mlog(0, "%s:%.*s: setting bit %u in refmap\n",
-			     dlm->name, namelen, name, request->node_idx);
-			dlm_lockres_set_refmap_bit(request->node_idx, res);
+			dlm_lockres_set_refmap_bit(dlm, res, request->node_idx);
 			spin_unlock(&res->spinlock);
 			response = DLM_MASTER_RESP_YES;
 			if (mle)
@@ -1493,10 +1499,8 @@ way_up_top:
 				 * go back and clean the mles on any
 				 * other nodes */
 				dispatch_assert = 1;
-				dlm_lockres_set_refmap_bit(request->node_idx, res);
-				mlog(0, "%s:%.*s: setting bit %u in refmap\n",
-				     dlm->name, namelen, name,
-				     request->node_idx);
+				dlm_lockres_set_refmap_bit(dlm, res,
+							   request->node_idx);
 			} else
 				response = DLM_MASTER_RESP_NO;
 		} else {
@@ -1702,7 +1706,7 @@ again:
 			     "lockres, set the bit in the refmap\n",
 			     namelen, lockname, to);
 			spin_lock(&res->spinlock);
-			dlm_lockres_set_refmap_bit(to, res);
+			dlm_lockres_set_refmap_bit(dlm, res, to);
 			spin_unlock(&res->spinlock);
 		}
 	}
@@ -2187,8 +2191,6 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 	namelen = res->lockname.len;
 	BUG_ON(namelen > O2NM_MAX_NAME_LEN);
 
-	mlog(0, "%s:%.*s: sending deref to %d\n",
-	     dlm->name, namelen, lockname, res->owner);
 	memset(&deref, 0, sizeof(deref));
 	deref.node_idx = dlm->node_num;
 	deref.namelen = namelen;
@@ -2197,14 +2199,12 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
 	ret = o2net_send_message(DLM_DEREF_LOCKRES_MSG, dlm->key,
 				 &deref, sizeof(deref), res->owner, &r);
 	if (ret < 0)
-		mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
-		     "node %u\n", ret, DLM_DEREF_LOCKRES_MSG, dlm->key,
-		     res->owner);
+		mlog(ML_ERROR, "%s: res %.*s, error %d send DEREF to node %u\n",
+		     dlm->name, namelen, lockname, ret, res->owner);
 	else if (r < 0) {
 		/* BAD.  other node says I did not have a ref. */
-		mlog(ML_ERROR,"while dropping ref on %s:%.*s "
-		    "(master=%u) got %d.\n", dlm->name, namelen,
-		    lockname, res->owner, r);
+		mlog(ML_ERROR, "%s: res %.*s, DEREF to node %u got %d\n",
+		     dlm->name, namelen, lockname, res->owner, r);
 		dlm_print_one_lock_resource(res);
 		BUG();
 	}
@@ -2260,7 +2260,7 @@ int dlm_deref_lockres_handler(struct o2net_msg *msg, u32 len, void *data,
 	else {
 		BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
 		if (test_bit(node, res->refmap)) {
-			dlm_lockres_clear_refmap_bit(node, res);
+			dlm_lockres_clear_refmap_bit(dlm, res, node);
 			cleared = 1;
 		}
 	}
@@ -2320,7 +2320,7 @@ static void dlm_deref_lockres_worker(struct dlm_work_item *item, void *data)
 	BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
 	if (test_bit(node, res->refmap)) {
 		__dlm_wait_on_lockres_flags(res, DLM_LOCK_RES_SETREF_INPROG);
-		dlm_lockres_clear_refmap_bit(node, res);
+		dlm_lockres_clear_refmap_bit(dlm, res, node);
 		cleared = 1;
 	}
 	spin_unlock(&res->spinlock);
@@ -2802,7 +2802,8 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
 				BUG_ON(!list_empty(&lock->bast_list));
 				BUG_ON(lock->ast_pending);
 				BUG_ON(lock->bast_pending);
-				dlm_lockres_clear_refmap_bit(lock->ml.node, res);
+				dlm_lockres_clear_refmap_bit(dlm, res,
+							     lock->ml.node);
 				list_del_init(&lock->list);
 				dlm_lock_put(lock);
 				/* In a normal unlock, we would have added a
@@ -2823,7 +2824,7 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
 			mlog(0, "%s:%.*s: node %u had a ref to this "
 			     "migrating lockres, clearing\n", dlm->name,
 			     res->lockname.len, res->lockname.name, bit);
-			dlm_lockres_clear_refmap_bit(bit, res);
+			dlm_lockres_clear_refmap_bit(dlm, res, bit);
 		}
 		bit++;
 	}
@@ -2916,9 +2917,9 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
 					 &migrate, sizeof(migrate), nodenum,
 					 &status);
 		if (ret < 0) {
-			mlog(ML_ERROR, "Error %d when sending message %u (key "
-			     "0x%x) to node %u\n", ret, DLM_MIGRATE_REQUEST_MSG,
-			     dlm->key, nodenum);
+			mlog(ML_ERROR, "%s: res %.*s, Error %d send "
+			     "MIGRATE_REQUEST to node %u\n", dlm->name,
+			     migrate.namelen, migrate.name, ret, nodenum);
 			if (!dlm_is_host_down(ret)) {
 				mlog(ML_ERROR, "unhandled error=%d!\n", ret);
 				BUG();
@@ -2937,7 +2938,7 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
 			     dlm->name, res->lockname.len, res->lockname.name,
 			     nodenum);
 			spin_lock(&res->spinlock);
-			dlm_lockres_set_refmap_bit(nodenum, res);
+			dlm_lockres_set_refmap_bit(dlm, res, nodenum);
 			spin_unlock(&res->spinlock);
 		}
 	}
@@ -3271,7 +3272,7 @@ int dlm_finish_migration(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
 	 * mastery reference here since old_master will briefly have
 	 * a reference after the migration completes */
 	spin_lock(&res->spinlock);
-	dlm_lockres_set_refmap_bit(old_master, res);
+	dlm_lockres_set_refmap_bit(dlm, res, old_master);
 	spin_unlock(&res->spinlock);
 
 	mlog(0, "now time to do a migrate request to other nodes\n");
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 7efab6d28a2..01ebfd0bdad 100644
--- a/fs/ocfs2/dlm/dlmrecovery.c
+++ b/fs/ocfs2/dlm/dlmrecovery.c
@@ -362,40 +362,38 @@ static int dlm_is_node_recovered(struct dlm_ctxt *dlm, u8 node)
 }
 
 
-int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout)
+void dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout)
 {
-	if (timeout) {
-		mlog(ML_NOTICE, "%s: waiting %dms for notification of "
-		     "death of node %u\n", dlm->name, timeout, node);
+	if (dlm_is_node_dead(dlm, node))
+		return;
+
+	printk(KERN_NOTICE "o2dlm: Waiting on the death of node %u in "
+	       "domain %s\n", node, dlm->name);
+
+	if (timeout)
 		wait_event_timeout(dlm->dlm_reco_thread_wq,
-			   dlm_is_node_dead(dlm, node),
-			   msecs_to_jiffies(timeout));
-	} else {
-		mlog(ML_NOTICE, "%s: waiting indefinitely for notification "
-		     "of death of node %u\n", dlm->name, node);
+				   dlm_is_node_dead(dlm, node),
+				   msecs_to_jiffies(timeout));
+	else
 		wait_event(dlm->dlm_reco_thread_wq,
 			   dlm_is_node_dead(dlm, node));
-	}
-	/* for now, return 0 */
-	return 0;
 }
 
-int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout)
+void dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout)
 {
-	if (timeout) {
-		mlog(0, "%s: waiting %dms for notification of "
-		     "recovery of node %u\n", dlm->name, timeout, node);
+	if (dlm_is_node_recovered(dlm, node))
+		return;
+
+	printk(KERN_NOTICE "o2dlm: Waiting on the recovery of node %u in "
+	       "domain %s\n", node, dlm->name);
+
+	if (timeout)
 		wait_event_timeout(dlm->dlm_reco_thread_wq,
-			   dlm_is_node_recovered(dlm, node),
-			   msecs_to_jiffies(timeout));
-	} else {
-		mlog(0, "%s: waiting indefinitely for notification "
-		     "of recovery of node %u\n", dlm->name, node);
+				   dlm_is_node_recovered(dlm, node),
+				   msecs_to_jiffies(timeout));
+	else
 		wait_event(dlm->dlm_reco_thread_wq,
 			   dlm_is_node_recovered(dlm, node));
-	}
-	/* for now, return 0 */
-	return 0;
 }
 
 /* callers of the top-level api calls (dlmlock/dlmunlock) should
@@ -430,6 +428,8 @@ static void dlm_begin_recovery(struct dlm_ctxt *dlm)
 {
 	spin_lock(&dlm->spinlock);
 	BUG_ON(dlm->reco.state & DLM_RECO_STATE_ACTIVE);
+	printk(KERN_NOTICE "o2dlm: Begin recovery on domain %s for node %u\n",
+	       dlm->name, dlm->reco.dead_node);
 	dlm->reco.state |= DLM_RECO_STATE_ACTIVE;
 	spin_unlock(&dlm->spinlock);
 }
@@ -440,9 +440,18 @@ static void dlm_end_recovery(struct dlm_ctxt *dlm)
 	BUG_ON(!(dlm->reco.state & DLM_RECO_STATE_ACTIVE));
 	dlm->reco.state &= ~DLM_RECO_STATE_ACTIVE;
 	spin_unlock(&dlm->spinlock);
+	printk(KERN_NOTICE "o2dlm: End recovery on domain %s\n", dlm->name);
 	wake_up(&dlm->reco.event);
 }
 
+static void dlm_print_recovery_master(struct dlm_ctxt *dlm)
+{
+	printk(KERN_NOTICE "o2dlm: Node %u (%s) is the Recovery Master for the "
+	       "dead node %u in domain %s\n", dlm->reco.new_master,
+	       (dlm->node_num == dlm->reco.new_master ? "me" : "he"),
+	       dlm->reco.dead_node, dlm->name);
+}
+
 static int dlm_do_recovery(struct dlm_ctxt *dlm)
 {
 	int status = 0;
@@ -505,9 +514,8 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm)
 		}
 		mlog(0, "another node will master this recovery session.\n");
 	}
-	mlog(0, "dlm=%s (%d), new_master=%u, this node=%u, dead_node=%u\n",
-	     dlm->name, task_pid_nr(dlm->dlm_reco_thread_task), dlm->reco.new_master,
-	     dlm->node_num, dlm->reco.dead_node);
+
+	dlm_print_recovery_master(dlm);
 
 	/* it is safe to start everything back up here
 	 * because all of the dead node's lock resources
@@ -518,15 +526,13 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm)
 	return 0;
 
 master_here:
-	mlog(ML_NOTICE, "(%d) Node %u is the Recovery Master for the Dead Node "
-	     "%u for Domain %s\n", task_pid_nr(dlm->dlm_reco_thread_task),
-	     dlm->node_num, dlm->reco.dead_node, dlm->name);
+	dlm_print_recovery_master(dlm);
 
 	status = dlm_remaster_locks(dlm, dlm->reco.dead_node);
 	if (status < 0) {
 		/* we should never hit this anymore */
-		mlog(ML_ERROR, "error %d remastering locks for node %u, "
-		     "retrying.\n", status, dlm->reco.dead_node);
+		mlog(ML_ERROR, "%s: Error %d remastering locks for node %u, "
+		     "retrying.\n", dlm->name, status, dlm->reco.dead_node);
 		/* yield a bit to allow any final network messages
 		 * to get handled on remaining nodes */
 		msleep(100);
@@ -567,7 +573,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
 		BUG_ON(ndata->state != DLM_RECO_NODE_DATA_INIT);
 		ndata->state = DLM_RECO_NODE_DATA_REQUESTING;
 
-		mlog(0, "requesting lock info from node %u\n",
+		mlog(0, "%s: Requesting lock info from node %u\n", dlm->name,
 		     ndata->node_num);
 
 		if (ndata->node_num == dlm->node_num) {
@@ -640,7 +646,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
 		spin_unlock(&dlm_reco_state_lock);
 	}
 
-	mlog(0, "done requesting all lock info\n");
+	mlog(0, "%s: Done requesting all lock info\n", dlm->name);
 
 	/* nodes should be sending reco data now
 	 * just need to wait */
@@ -802,10 +808,9 @@ static int dlm_request_all_locks(struct dlm_ctxt *dlm, u8 request_from,
 
 	/* negative status is handled by caller */
 	if (ret < 0)
-		mlog(ML_ERROR, "Error %d when sending message %u (key "
-		     "0x%x) to node %u\n", ret, DLM_LOCK_REQUEST_MSG,
-		     dlm->key, request_from);
-
+		mlog(ML_ERROR, "%s: Error %d send LOCK_REQUEST to node %u "
+		     "to recover dead node %u\n", dlm->name, ret,
+		     request_from, dead_node);
 	// return from here, then
 	// sleep until all received or error
 	return ret;
@@ -956,9 +961,9 @@ static int dlm_send_all_done_msg(struct dlm_ctxt *dlm, u8 dead_node, u8 send_to)
 	ret = o2net_send_message(DLM_RECO_DATA_DONE_MSG, dlm->key, &done_msg,
 				 sizeof(done_msg), send_to, &tmpret);
 	if (ret < 0) {
-		mlog(ML_ERROR, "Error %d when sending message %u (key "
-		     "0x%x) to node %u\n", ret, DLM_RECO_DATA_DONE_MSG,
-		     dlm->key, send_to);
+		mlog(ML_ERROR, "%s: Error %d send RECO_DATA_DONE to node %u "
+		     "to recover dead node %u\n", dlm->name, ret, send_to,
+		     dead_node);
 		if (!dlm_is_host_down(ret)) {
 			BUG();
 		}
@@ -1127,9 +1132,11 @@ static int dlm_send_mig_lockres_msg(struct dlm_ctxt *dlm,
 	if (ret < 0) {
 		/* XXX: negative status is not handled.
 		 * this will end up killing this node. */
-		mlog(ML_ERROR, "Error %d when sending message %u (key "
-		     "0x%x) to node %u\n", ret, DLM_MIG_LOCKRES_MSG,
-		     dlm->key, send_to);
+		mlog(ML_ERROR, "%s: res %.*s, Error %d send MIG_LOCKRES to "
+		     "node %u (%s)\n", dlm->name, mres->lockname_len,
+		     mres->lockname, ret, send_to,
+		     (orig_flags & DLM_MRES_MIGRATION ?
+		      "migration" : "recovery"));
 	} else {
 		/* might get an -ENOMEM back here */
 		ret = status;
@@ -1767,7 +1774,7 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm,
 			     dlm->name, mres->lockname_len, mres->lockname,
 			     from);
 			spin_lock(&res->spinlock);
-			dlm_lockres_set_refmap_bit(from, res);
+			dlm_lockres_set_refmap_bit(dlm, res, from);
 			spin_unlock(&res->spinlock);
 			added++;
 			break;
@@ -1965,7 +1972,7 @@ skip_lvb:
 			mlog(0, "%s:%.*s: added lock for node %u, "
 			     "setting refmap bit\n", dlm->name,
 			     res->lockname.len, res->lockname.name, ml->node);
-			dlm_lockres_set_refmap_bit(ml->node, res);
+			dlm_lockres_set_refmap_bit(dlm, res, ml->node);
 			added++;
 		}
 		spin_unlock(&res->spinlock);
@@ -2084,6 +2091,9 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
 
 	list_for_each_entry_safe(res, next, &dlm->reco.resources, recovering) {
 		if (res->owner == dead_node) {
+			mlog(0, "%s: res %.*s, Changing owner from %u to %u\n",
+			     dlm->name, res->lockname.len, res->lockname.name,
+			     res->owner, new_master);
 			list_del_init(&res->recovering);
 			spin_lock(&res->spinlock);
 			/* new_master has our reference from
@@ -2105,40 +2115,30 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
 	for (i = 0; i < DLM_HASH_BUCKETS; i++) {
 		bucket = dlm_lockres_hash(dlm, i);
 		hlist_for_each_entry(res, hash_iter, bucket, hash_node) {
-			if (res->state & DLM_LOCK_RES_RECOVERING) {
-				if (res->owner == dead_node) {
-					mlog(0, "(this=%u) res %.*s owner=%u "
-					     "was not on recovering list, but "
-					     "clearing state anyway\n",
-					     dlm->node_num, res->lockname.len,
-					     res->lockname.name, new_master);
-				} else if (res->owner == dlm->node_num) {
-					mlog(0, "(this=%u) res %.*s owner=%u "
-					     "was not on recovering list, "
-					     "owner is THIS node, clearing\n",
-					     dlm->node_num, res->lockname.len,
-					     res->lockname.name, new_master);
-				} else
-					continue;
+			if (!(res->state & DLM_LOCK_RES_RECOVERING))
+				continue;
 
-				if (!list_empty(&res->recovering)) {
-					mlog(0, "%s:%.*s: lockres was "
-					     "marked RECOVERING, owner=%u\n",
-					     dlm->name, res->lockname.len,
-					     res->lockname.name, res->owner);
-					list_del_init(&res->recovering);
-					dlm_lockres_put(res);
-				}
-				spin_lock(&res->spinlock);
-				/* new_master has our reference from
-				 * the lock state sent during recovery */
-				dlm_change_lockres_owner(dlm, res, new_master);
-				res->state &= ~DLM_LOCK_RES_RECOVERING;
-				if (__dlm_lockres_has_locks(res))
-					__dlm_dirty_lockres(dlm, res);
-				spin_unlock(&res->spinlock);
-				wake_up(&res->wq);
+			if (res->owner != dead_node &&
+			    res->owner != dlm->node_num)
+				continue;
+
+			if (!list_empty(&res->recovering)) {
+				list_del_init(&res->recovering);
+				dlm_lockres_put(res);
 			}
+
+			/* new_master has our reference from
+			 * the lock state sent during recovery */
+			mlog(0, "%s: res %.*s, Changing owner from %u to %u\n",
+			     dlm->name, res->lockname.len, res->lockname.name,
+			     res->owner, new_master);
+			spin_lock(&res->spinlock);
+			dlm_change_lockres_owner(dlm, res, new_master);
+			res->state &= ~DLM_LOCK_RES_RECOVERING;
+			if (__dlm_lockres_has_locks(res))
+				__dlm_dirty_lockres(dlm, res);
+			spin_unlock(&res->spinlock);
+			wake_up(&res->wq);
 		}
 	}
 }
@@ -2252,12 +2252,12 @@ static void dlm_free_dead_locks(struct dlm_ctxt *dlm,
 			     res->lockname.len, res->lockname.name, freed, dead_node);
 			__dlm_print_one_lock_resource(res);
 		}
-		dlm_lockres_clear_refmap_bit(dead_node, res);
+		dlm_lockres_clear_refmap_bit(dlm, res, dead_node);
 	} else if (test_bit(dead_node, res->refmap)) {
 		mlog(0, "%s:%.*s: dead node %u had a ref, but had "
 		     "no locks and had not purged before dying\n", dlm->name,
 		     res->lockname.len, res->lockname.name, dead_node);
-		dlm_lockres_clear_refmap_bit(dead_node, res);
+		dlm_lockres_clear_refmap_bit(dlm, res, dead_node);
 	}
 
 	/* do not kick thread yet */
@@ -2324,9 +2324,9 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node)
 			dlm_revalidate_lvb(dlm, res, dead_node);
 			if (res->owner == dead_node) {
 				if (res->state & DLM_LOCK_RES_DROPPING_REF) {
-					mlog(ML_NOTICE, "Ignore %.*s for "
+					mlog(ML_NOTICE, "%s: res %.*s, Skip "
 					     "recovery as it is being freed\n",
-					     res->lockname.len,
+					     dlm->name, res->lockname.len,
 					     res->lockname.name);
 				} else
 					dlm_move_lockres_to_recovery_list(dlm,
diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c
index 1d6d1d22c47..e73c833fc2a 100644
--- a/fs/ocfs2/dlm/dlmthread.c
+++ b/fs/ocfs2/dlm/dlmthread.c
@@ -94,24 +94,26 @@ int __dlm_lockres_unused(struct dlm_lock_resource *res)
 {
 	int bit;
 
+	assert_spin_locked(&res->spinlock);
+
 	if (__dlm_lockres_has_locks(res))
 		return 0;
 
+	/* Locks are in the process of being created */
+	if (res->inflight_locks)
+		return 0;
+
 	if (!list_empty(&res->dirty) || res->state & DLM_LOCK_RES_DIRTY)
 		return 0;
 
 	if (res->state & DLM_LOCK_RES_RECOVERING)
 		return 0;
 
+	/* Another node has this resource with this node as the master */
 	bit = find_next_bit(res->refmap, O2NM_MAX_NODES, 0);
 	if (bit < O2NM_MAX_NODES)
 		return 0;
 
-	/*
-	 * since the bit for dlm->node_num is not set, inflight_locks better
-	 * be zero
-	 */
-	BUG_ON(res->inflight_locks != 0);
 	return 1;
 }
 
@@ -185,8 +187,6 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm,
 		/* clear our bit from the master's refmap, ignore errors */
 		ret = dlm_drop_lockres_ref(dlm, res);
 		if (ret < 0) {
-			mlog(ML_ERROR, "%s: deref %.*s failed %d\n", dlm->name,
-			     res->lockname.len, res->lockname.name, ret);
 			if (!dlm_is_host_down(ret))
 				BUG();
 		}
@@ -209,7 +209,7 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm,
 		BUG();
 	}
 
-	__dlm_unhash_lockres(res);
+	__dlm_unhash_lockres(dlm, res);
 
 	/* lockres is not in the hash now.  drop the flag and wake up
 	 * any processes waiting in dlm_get_lock_resource. */
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 23457b491e8..2f5b92ef0e5 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -832,6 +832,102 @@ out:
 	return ret;
 }
 
+int ocfs2_seek_data_hole_offset(struct file *file, loff_t *offset, int origin)
+{
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+	unsigned int is_last = 0, is_data = 0;
+	u16 cs_bits = OCFS2_SB(inode->i_sb)->s_clustersize_bits;
+	u32 cpos, cend, clen, hole_size;
+	u64 extoff, extlen;
+	struct buffer_head *di_bh = NULL;
+	struct ocfs2_extent_rec rec;
+
+	BUG_ON(origin != SEEK_DATA && origin != SEEK_HOLE);
+
+	ret = ocfs2_inode_lock(inode, &di_bh, 0);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
+
+	down_read(&OCFS2_I(inode)->ip_alloc_sem);
+
+	if (*offset >= inode->i_size) {
+		ret = -ENXIO;
+		goto out_unlock;
+	}
+
+	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
+		if (origin == SEEK_HOLE)
+			*offset = inode->i_size;
+		goto out_unlock;
+	}
+
+	clen = 0;
+	cpos = *offset >> cs_bits;
+	cend = ocfs2_clusters_for_bytes(inode->i_sb, inode->i_size);
+
+	while (cpos < cend && !is_last) {
+		ret = ocfs2_get_clusters_nocache(inode, di_bh, cpos, &hole_size,
+						 &rec, &is_last);
+		if (ret) {
+			mlog_errno(ret);
+			goto out_unlock;
+		}
+
+		extoff = cpos;
+		extoff <<= cs_bits;
+
+		if (rec.e_blkno == 0ULL) {
+			clen = hole_size;
+			is_data = 0;
+		} else {
+			clen = le16_to_cpu(rec.e_leaf_clusters) -
+				(cpos - le32_to_cpu(rec.e_cpos));
+			is_data = (rec.e_flags & OCFS2_EXT_UNWRITTEN) ?  0 : 1;
+		}
+
+		if ((!is_data && origin == SEEK_HOLE) ||
+		    (is_data && origin == SEEK_DATA)) {
+			if (extoff > *offset)
+				*offset = extoff;
+			goto out_unlock;
+		}
+
+		if (!is_last)
+			cpos += clen;
+	}
+
+	if (origin == SEEK_HOLE) {
+		extoff = cpos;
+		extoff <<= cs_bits;
+		extlen = clen;
+		extlen <<=  cs_bits;
+
+		if ((extoff + extlen) > inode->i_size)
+			extlen = inode->i_size - extoff;
+		extoff += extlen;
+		if (extoff > *offset)
+			*offset = extoff;
+		goto out_unlock;
+	}
+
+	ret = -ENXIO;
+
+out_unlock:
+
+	brelse(di_bh);
+
+	up_read(&OCFS2_I(inode)->ip_alloc_sem);
+
+	ocfs2_inode_unlock(inode, 0);
+out:
+	if (ret && ret != -ENXIO)
+		ret = -ENXIO;
+	return ret;
+}
+
 int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
 			   struct buffer_head *bhs[], int flags,
 			   int (*validate)(struct super_block *sb,
diff --git a/fs/ocfs2/extent_map.h b/fs/ocfs2/extent_map.h
index e79d41c2c90..67ea57d2fd5 100644
--- a/fs/ocfs2/extent_map.h
+++ b/fs/ocfs2/extent_map.h
@@ -53,6 +53,8 @@ int ocfs2_extent_map_get_blocks(struct inode *inode, u64 v_blkno, u64 *p_blkno,
 int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		 u64 map_start, u64 map_len);
 
+int ocfs2_seek_data_hole_offset(struct file *file, loff_t *offset, int origin);
+
 int ocfs2_xattr_get_clusters(struct inode *inode, u32 v_cluster,
 			     u32 *p_cluster, u32 *num_clusters,
 			     struct ocfs2_extent_list *el,
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 145f4533a93..5c4a74e04ab 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -171,7 +171,8 @@ static int ocfs2_dir_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static int ocfs2_sync_file(struct file *file, int datasync)
+static int ocfs2_sync_file(struct file *file, loff_t start, loff_t end,
+			   int datasync)
 {
 	int err = 0;
 	journal_t *journal;
@@ -184,6 +185,16 @@ static int ocfs2_sync_file(struct file *file, int datasync)
 			      file->f_path.dentry->d_name.name,
 			      (unsigned long long)datasync);
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	/*
+	 * Probably don't need the i_mutex at all in here, just putting it here
+	 * to be consistent with how fsync used to be called, someone more
+	 * familiar with the fs could possibly remove it.
+	 */
+	mutex_lock(&inode->i_mutex);
 	if (datasync && !(inode->i_state & I_DIRTY_DATASYNC)) {
 		/*
 		 * We still have to flush drive's caches to get data to the
@@ -200,6 +211,7 @@ static int ocfs2_sync_file(struct file *file, int datasync)
 bail:
 	if (err)
 		mlog_errno(err);
+	mutex_unlock(&inode->i_mutex);
 
 	return (err < 0) ? -EIO : 0;
 }
@@ -1142,6 +1154,8 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 		if (status)
 			goto bail_unlock;
 
+		inode_dio_wait(inode);
+
 		if (i_size_read(inode) > attr->ia_size) {
 			if (ocfs2_should_order_data(inode)) {
 				status = ocfs2_begin_ordered_truncate(inode,
@@ -1279,11 +1293,11 @@ bail:
 	return err;
 }
 
-int ocfs2_permission(struct inode *inode, int mask, unsigned int flags)
+int ocfs2_permission(struct inode *inode, int mask)
 {
 	int ret;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	ret = ocfs2_inode_lock(inode, NULL, 0);
@@ -1293,7 +1307,7 @@ int ocfs2_permission(struct inode *inode, int mask, unsigned int flags)
 		goto out;
 	}
 
-	ret = generic_permission(inode, mask, flags, ocfs2_check_acl);
+	ret = generic_permission(inode, mask);
 
 	ocfs2_inode_unlock(inode, 0);
 out:
@@ -2254,9 +2268,8 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
 	ocfs2_iocb_clear_sem_locked(iocb);
 
 relock:
-	/* to match setattr's i_mutex -> i_alloc_sem -> rw_lock ordering */
+	/* to match setattr's i_mutex -> rw_lock ordering */
 	if (direct_io) {
-		down_read(&inode->i_alloc_sem);
 		have_alloc_sem = 1;
 		/* communicate with ocfs2_dio_end_io */
 		ocfs2_iocb_set_sem_locked(iocb);
@@ -2312,7 +2325,6 @@ relock:
 	 */
 	if (direct_io && !can_do_direct) {
 		ocfs2_rw_unlock(inode, rw_level);
-		up_read(&inode->i_alloc_sem);
 
 		have_alloc_sem = 0;
 		rw_level = -1;
@@ -2395,8 +2407,7 @@ out_dio:
 	/*
 	 * deep in g_f_a_w_n()->ocfs2_direct_IO we pass in a ocfs2_dio_end_io
 	 * function pointer which is called when o_direct io completes so that
-	 * it can unlock our rw lock.  (it's the clustered equivalent of
-	 * i_alloc_sem; protects truncate from racing with pending ios).
+	 * it can unlock our rw lock.
 	 * Unfortunately there are error cases which call end_io and others
 	 * that don't.  so we don't have to unlock the rw_lock if either an
 	 * async dio is going to do it in the future or an end_io after an
@@ -2416,10 +2427,8 @@ out:
 		ocfs2_rw_unlock(inode, rw_level);
 
 out_sems:
-	if (have_alloc_sem) {
-		up_read(&inode->i_alloc_sem);
+	if (have_alloc_sem)
 		ocfs2_iocb_clear_sem_locked(iocb);
-	}
 
 	mutex_unlock(&inode->i_mutex);
 
@@ -2569,7 +2578,6 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
 	 * need locks to protect pending reads from racing with truncate.
 	 */
 	if (filp->f_flags & O_DIRECT) {
-		down_read(&inode->i_alloc_sem);
 		have_alloc_sem = 1;
 		ocfs2_iocb_set_sem_locked(iocb);
 
@@ -2612,16 +2620,66 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
 	}
 
 bail:
-	if (have_alloc_sem) {
-		up_read(&inode->i_alloc_sem);
+	if (have_alloc_sem)
 		ocfs2_iocb_clear_sem_locked(iocb);
-	}
+
 	if (rw_level != -1)
 		ocfs2_rw_unlock(inode, rw_level);
 
 	return ret;
 }
 
+/* Refer generic_file_llseek_unlocked() */
+static loff_t ocfs2_file_llseek(struct file *file, loff_t offset, int origin)
+{
+	struct inode *inode = file->f_mapping->host;
+	int ret = 0;
+
+	mutex_lock(&inode->i_mutex);
+
+	switch (origin) {
+	case SEEK_SET:
+		break;
+	case SEEK_END:
+		offset += inode->i_size;
+		break;
+	case SEEK_CUR:
+		if (offset == 0) {
+			offset = file->f_pos;
+			goto out;
+		}
+		offset += file->f_pos;
+		break;
+	case SEEK_DATA:
+	case SEEK_HOLE:
+		ret = ocfs2_seek_data_hole_offset(file, &offset, origin);
+		if (ret)
+			goto out;
+		break;
+	default:
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET))
+		ret = -EINVAL;
+	if (!ret && offset > inode->i_sb->s_maxbytes)
+		ret = -EINVAL;
+	if (ret)
+		goto out;
+
+	if (offset != file->f_pos) {
+		file->f_pos = offset;
+		file->f_version = 0;
+	}
+
+out:
+	mutex_unlock(&inode->i_mutex);
+	if (ret)
+		return ret;
+	return offset;
+}
+
 const struct inode_operations ocfs2_file_iops = {
 	.setattr	= ocfs2_setattr,
 	.getattr	= ocfs2_getattr,
@@ -2631,12 +2689,14 @@ const struct inode_operations ocfs2_file_iops = {
 	.listxattr	= ocfs2_listxattr,
 	.removexattr	= generic_removexattr,
 	.fiemap		= ocfs2_fiemap,
+	.check_acl	= ocfs2_check_acl,
 };
 
 const struct inode_operations ocfs2_special_file_iops = {
 	.setattr	= ocfs2_setattr,
 	.getattr	= ocfs2_getattr,
 	.permission	= ocfs2_permission,
+	.check_acl	= ocfs2_check_acl,
 };
 
 /*
@@ -2644,7 +2704,7 @@ const struct inode_operations ocfs2_special_file_iops = {
  * ocfs2_fops_no_plocks and ocfs2_dops_no_plocks!
  */
 const struct file_operations ocfs2_fops = {
-	.llseek		= generic_file_llseek,
+	.llseek		= ocfs2_file_llseek,
 	.read		= do_sync_read,
 	.write		= do_sync_write,
 	.mmap		= ocfs2_mmap,
@@ -2692,7 +2752,7 @@ const struct file_operations ocfs2_dops = {
  * the cluster.
  */
 const struct file_operations ocfs2_fops_no_plocks = {
-	.llseek		= generic_file_llseek,
+	.llseek		= ocfs2_file_llseek,
 	.read		= do_sync_read,
 	.write		= do_sync_write,
 	.mmap		= ocfs2_mmap,
diff --git a/fs/ocfs2/file.h b/fs/ocfs2/file.h
index f5afbbef670..97bf761c9e7 100644
--- a/fs/ocfs2/file.h
+++ b/fs/ocfs2/file.h
@@ -61,7 +61,7 @@ int ocfs2_zero_extend(struct inode *inode, struct buffer_head *di_bh,
 int ocfs2_setattr(struct dentry *dentry, struct iattr *attr);
 int ocfs2_getattr(struct vfsmount *mnt, struct dentry *dentry,
 		  struct kstat *stat);
-int ocfs2_permission(struct inode *inode, int mask, unsigned int flags);
+int ocfs2_permission(struct inode *inode, int mask);
 
 int ocfs2_should_update_atime(struct inode *inode,
 			      struct vfsmount *vfsmnt);
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index 295d56454e8..0a42ae96dca 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -1544,9 +1544,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
 	/* we need to run complete recovery for offline orphan slots */
 	ocfs2_replay_map_set_state(osb, REPLAY_NEEDED);
 
-	mlog(ML_NOTICE, "Recovering node %d from slot %d on device (%u,%u)\n",
-	     node_num, slot_num,
-	     MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));
+	printk(KERN_NOTICE "ocfs2: Begin replay journal (node %d, slot %d) on "\
+	       "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+	       MINOR(osb->sb->s_dev));
 
 	OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
 
@@ -1601,6 +1601,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
 
 	jbd2_journal_destroy(journal);
 
+	printk(KERN_NOTICE "ocfs2: End replay journal (node %d, slot %d) on "\
+	       "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+	       MINOR(osb->sb->s_dev));
 done:
 	/* drop the lock on this nodes journal */
 	if (got_lock)
@@ -1808,6 +1811,20 @@ static inline unsigned long ocfs2_orphan_scan_timeout(void)
  * every slot, queuing a recovery of the slot on the ocfs2_wq thread. This
  * is done to catch any orphans that are left over in orphan directories.
  *
+ * It scans all slots, even ones that are in use. It does so to handle the
+ * case described below:
+ *
+ *   Node 1 has an inode it was using. The dentry went away due to memory
+ *   pressure.  Node 1 closes the inode, but it's on the free list. The node
+ *   has the open lock.
+ *   Node 2 unlinks the inode. It grabs the dentry lock to notify others,
+ *   but node 1 has no dentry and doesn't get the message. It trylocks the
+ *   open lock, sees that another node has a PR, and does nothing.
+ *   Later node 2 runs its orphan dir. It igets the inode, trylocks the
+ *   open lock, sees the PR still, and does nothing.
+ *   Basically, we have to trigger an orphan iput on node 1. The only way
+ *   for this to happen is if node 1 runs node 2's orphan dir.
+ *
  * ocfs2_queue_orphan_scan gets called every ORPHAN_SCAN_SCHEDULE_TIMEOUT
  * seconds.  It gets an EX lock on os_lockres and checks sequence number
  * stored in LVB. If the sequence number has changed, it means some other
diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c
index 3e9393ca39e..9cd41083e99 100644
--- a/fs/ocfs2/mmap.c
+++ b/fs/ocfs2/mmap.c
@@ -61,7 +61,7 @@ static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
 static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
 				struct page *page)
 {
-	int ret;
+	int ret = VM_FAULT_NOPAGE;
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct address_space *mapping = inode->i_mapping;
 	loff_t pos = page_offset(page);
@@ -71,32 +71,25 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
 	void *fsdata;
 	loff_t size = i_size_read(inode);
 
-	/*
-	 * Another node might have truncated while we were waiting on
-	 * cluster locks.
-	 * We don't check size == 0 before the shift. This is borrowed
-	 * from do_generic_file_read.
-	 */
 	last_index = (size - 1) >> PAGE_CACHE_SHIFT;
-	if (unlikely(!size || page->index > last_index)) {
-		ret = -EINVAL;
-		goto out;
-	}
 
 	/*
-	 * The i_size check above doesn't catch the case where nodes
-	 * truncated and then re-extended the file. We'll re-check the
-	 * page mapping after taking the page lock inside of
-	 * ocfs2_write_begin_nolock().
+	 * There are cases that lead to the page no longer bebongs to the
+	 * mapping.
+	 * 1) pagecache truncates locally due to memory pressure.
+	 * 2) pagecache truncates when another is taking EX lock against 
+	 * inode lock. see ocfs2_data_convert_worker.
+	 * 
+	 * The i_size check doesn't catch the case where nodes truncated and
+	 * then re-extended the file. We'll re-check the page mapping after
+	 * taking the page lock inside of ocfs2_write_begin_nolock().
+	 *
+	 * Let VM retry with these cases.
 	 */
-	if (!PageUptodate(page) || page->mapping != inode->i_mapping) {
-		/*
-		 * the page has been umapped in ocfs2_data_downconvert_worker.
-		 * So return 0 here and let VFS retry.
-		 */
-		ret = 0;
+	if ((page->mapping != inode->i_mapping) ||
+	    (!PageUptodate(page)) ||
+	    (page_offset(page) >= size))
 		goto out;
-	}
 
 	/*
 	 * Call ocfs2_write_begin() and ocfs2_write_end() to take
@@ -116,17 +109,21 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
 	if (ret) {
 		if (ret != -ENOSPC)
 			mlog_errno(ret);
+		if (ret == -ENOMEM)
+			ret = VM_FAULT_OOM;
+		else
+			ret = VM_FAULT_SIGBUS;
 		goto out;
 	}
 
-	ret = ocfs2_write_end_nolock(mapping, pos, len, len, locked_page,
-				     fsdata);
-	if (ret < 0) {
-		mlog_errno(ret);
+	if (!locked_page) {
+		ret = VM_FAULT_NOPAGE;
 		goto out;
 	}
+	ret = ocfs2_write_end_nolock(mapping, pos, len, len, locked_page,
+				     fsdata);
 	BUG_ON(ret != len);
-	ret = 0;
+	ret = VM_FAULT_LOCKED;
 out:
 	return ret;
 }
@@ -168,8 +165,6 @@ static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 
 out:
 	ocfs2_unblock_signals(&oldset);
-	if (ret)
-		ret = VM_FAULT_SIGBUS;
 	return ret;
 }
 
diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index e5d738cd9cc..33889dc52dd 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -2498,4 +2498,5 @@ const struct inode_operations ocfs2_dir_iops = {
 	.listxattr	= ocfs2_listxattr,
 	.removexattr	= generic_removexattr,
 	.fiemap         = ocfs2_fiemap,
+	.check_acl	= ocfs2_check_acl,
 };
diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c
index dc8007fc924..942fd65bdad 100644
--- a/fs/ocfs2/quota_local.c
+++ b/fs/ocfs2/quota_local.c
@@ -404,7 +404,9 @@ struct ocfs2_quota_recovery *ocfs2_begin_quota_recovery(
 	int status = 0;
 	struct ocfs2_quota_recovery *rec;
 
-	mlog(ML_NOTICE, "Beginning quota recovery in slot %u\n", slot_num);
+	printk(KERN_NOTICE "ocfs2: Beginning quota recovery on device (%s) for "
+	       "slot %u\n", osb->dev_str, slot_num);
+
 	rec = ocfs2_alloc_quota_recovery();
 	if (!rec)
 		return ERR_PTR(-ENOMEM);
@@ -596,7 +598,9 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
 	struct inode *lqinode;
 	unsigned int flags;
 
-	mlog(ML_NOTICE, "Finishing quota recovery in slot %u\n", slot_num);
+	printk(KERN_NOTICE "ocfs2: Finishing quota recovery on device (%s) for "
+	       "slot %u\n", osb->dev_str, slot_num);
+
 	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
 	for (type = 0; type < MAXQUOTAS; type++) {
 		if (list_empty(&(rec->r_list[type])))
@@ -612,8 +616,9 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
 		/* Someone else is holding the lock? Then he must be
 		 * doing the recovery. Just skip the file... */
 		if (status == -EAGAIN) {
-			mlog(ML_NOTICE, "skipping quota recovery for slot %d "
-			     "because quota file is locked.\n", slot_num);
+			printk(KERN_NOTICE "ocfs2: Skipping quota recovery on "
+			       "device (%s) for slot %d because quota file is "
+			       "locked.\n", osb->dev_str, slot_num);
 			status = 0;
 			goto out_put;
 		} else if (status < 0) {
diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index ebfd3825f12..cf782338266 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -4368,25 +4368,6 @@ static inline int ocfs2_may_create(struct inode *dir, struct dentry *child)
 	return inode_permission(dir, MAY_WRITE | MAY_EXEC);
 }
 
-/* copied from user_path_parent. */
-static int ocfs2_user_path_parent(const char __user *path,
-				  struct nameidata *nd, char **name)
-{
-	char *s = getname(path);
-	int error;
-
-	if (IS_ERR(s))
-		return PTR_ERR(s);
-
-	error = kern_path_parent(s, nd);
-	if (error)
-		putname(s);
-	else
-		*name = s;
-
-	return error;
-}
-
 /**
  * ocfs2_vfs_reflink - Create a reference-counted link
  *
@@ -4460,10 +4441,8 @@ int ocfs2_reflink_ioctl(struct inode *inode,
 			bool preserve)
 {
 	struct dentry *new_dentry;
-	struct nameidata nd;
-	struct path old_path;
+	struct path old_path, new_path;
 	int error;
-	char *to = NULL;
 
 	if (!ocfs2_refcount_tree(OCFS2_SB(inode->i_sb)))
 		return -EOPNOTSUPP;
@@ -4474,39 +4453,33 @@ int ocfs2_reflink_ioctl(struct inode *inode,
 		return error;
 	}
 
-	error = ocfs2_user_path_parent(newname, &nd, &to);
-	if (error) {
+	new_dentry = user_path_create(AT_FDCWD, newname, &new_path, 0);
+	error = PTR_ERR(new_dentry);
+	if (IS_ERR(new_dentry)) {
 		mlog_errno(error);
 		goto out;
 	}
 
 	error = -EXDEV;
-	if (old_path.mnt != nd.path.mnt)
-		goto out_release;
-	new_dentry = lookup_create(&nd, 0);
-	error = PTR_ERR(new_dentry);
-	if (IS_ERR(new_dentry)) {
+	if (old_path.mnt != new_path.mnt) {
 		mlog_errno(error);
-		goto out_unlock;
+		goto out_dput;
 	}
 
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(new_path.mnt);
 	if (error) {
 		mlog_errno(error);
 		goto out_dput;
 	}
 
 	error = ocfs2_vfs_reflink(old_path.dentry,
-				  nd.path.dentry->d_inode,
+				  new_path.dentry->d_inode,
 				  new_dentry, preserve);
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(new_path.mnt);
 out_dput:
 	dput(new_dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-out_release:
-	path_put(&nd.path);
-	putname(to);
+	mutex_unlock(&new_path.dentry->d_inode->i_mutex);
+	path_put(&new_path);
 out:
 	path_put(&old_path);
 
diff --git a/fs/ocfs2/slot_map.c b/fs/ocfs2/slot_map.c
index 26fc0014d50..1424c151ccc 100644
--- a/fs/ocfs2/slot_map.c
+++ b/fs/ocfs2/slot_map.c
@@ -493,8 +493,8 @@ int ocfs2_find_slot(struct ocfs2_super *osb)
 			goto bail;
 		}
 	} else
-		mlog(ML_NOTICE, "slot %d is already allocated to this node!\n",
-		     slot);
+		printk(KERN_INFO "ocfs2: Slot %d on device (%s) was already "
+		       "allocated to this node!\n", slot, osb->dev_str);
 
 	ocfs2_set_slot(si, slot, osb->node_num);
 	osb->slot_num = slot;
diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c
index 19965b00c43..94368017edb 100644
--- a/fs/ocfs2/stack_o2cb.c
+++ b/fs/ocfs2/stack_o2cb.c
@@ -28,6 +28,7 @@
 #include "cluster/masklog.h"
 #include "cluster/nodemanager.h"
 #include "cluster/heartbeat.h"
+#include "cluster/tcp.h"
 
 #include "stackglue.h"
 
@@ -256,6 +257,61 @@ static void o2cb_dump_lksb(struct ocfs2_dlm_lksb *lksb)
 }
 
 /*
+ * Check if this node is heartbeating and is connected to all other
+ * heartbeating nodes.
+ */
+static int o2cb_cluster_check(void)
+{
+	u8 node_num;
+	int i;
+	unsigned long hbmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
+	unsigned long netmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
+
+	node_num = o2nm_this_node();
+	if (node_num == O2NM_MAX_NODES) {
+		printk(KERN_ERR "o2cb: This node has not been configured.\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * o2dlm expects o2net sockets to be created. If not, then
+	 * dlm_join_domain() fails with a stack of errors which are both cryptic
+	 * and incomplete. The idea here is to detect upfront whether we have
+	 * managed to connect to all nodes or not. If not, then list the nodes
+	 * to allow the user to check the configuration (incorrect IP, firewall,
+	 * etc.) Yes, this is racy. But its not the end of the world.
+	 */
+#define	O2CB_MAP_STABILIZE_COUNT	60
+	for (i = 0; i < O2CB_MAP_STABILIZE_COUNT; ++i) {
+		o2hb_fill_node_map(hbmap, sizeof(hbmap));
+		if (!test_bit(node_num, hbmap)) {
+			printk(KERN_ERR "o2cb: %s heartbeat has not been "
+			       "started.\n", (o2hb_global_heartbeat_active() ?
+					      "Global" : "Local"));
+			return -EINVAL;
+		}
+		o2net_fill_node_map(netmap, sizeof(netmap));
+		/* Force set the current node to allow easy compare */
+		set_bit(node_num, netmap);
+		if (!memcmp(hbmap, netmap, sizeof(hbmap)))
+			return 0;
+		if (i < O2CB_MAP_STABILIZE_COUNT)
+			msleep(1000);
+	}
+
+	printk(KERN_ERR "o2cb: This node could not connect to nodes:");
+	i = -1;
+	while ((i = find_next_bit(hbmap, O2NM_MAX_NODES,
+				  i + 1)) < O2NM_MAX_NODES) {
+		if (!test_bit(i, netmap))
+			printk(" %u", i);
+	}
+	printk(".\n");
+
+	return -ENOTCONN;
+}
+
+/*
  * Called from the dlm when it's about to evict a node. This is how the
  * classic stack signals node death.
  */
@@ -263,8 +319,8 @@ static void o2dlm_eviction_cb(int node_num, void *data)
 {
 	struct ocfs2_cluster_connection *conn = data;
 
-	mlog(ML_NOTICE, "o2dlm has evicted node %d from group %.*s\n",
-	     node_num, conn->cc_namelen, conn->cc_name);
+	printk(KERN_NOTICE "o2cb: o2dlm has evicted node %d from domain %.*s\n",
+	       node_num, conn->cc_namelen, conn->cc_name);
 
 	conn->cc_recovery_handler(node_num, conn->cc_recovery_data);
 }
@@ -280,12 +336,11 @@ static int o2cb_cluster_connect(struct ocfs2_cluster_connection *conn)
 	BUG_ON(conn == NULL);
 	BUG_ON(conn->cc_proto == NULL);
 
-	/* for now we only have one cluster/node, make sure we see it
-	 * in the heartbeat universe */
-	if (!o2hb_check_local_node_heartbeating()) {
-		if (o2hb_global_heartbeat_active())
-			mlog(ML_ERROR, "Global heartbeat not started\n");
-		rc = -EINVAL;
+	/* Ensure cluster stack is up and all nodes are connected */
+	rc = o2cb_cluster_check();
+	if (rc) {
+		printk(KERN_ERR "o2cb: Cluster check failed. Fix errors "
+		       "before retrying.\n");
 		goto out;
 	}
 
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 603f5fe9f81..938e2b2b0c9 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -1073,7 +1073,7 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 
 	sb->s_magic = OCFS2_SUPER_MAGIC;
 
-	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
+	sb->s_flags = (sb->s_flags & ~(MS_POSIXACL | MS_NOSEC)) |
 		((osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
 
 	/* Hard readonly mode only if: bdev_read_only, MS_RDONLY,
@@ -1108,9 +1108,9 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 
 		ocfs2_set_ro_flag(osb, 1);
 
-		printk(KERN_NOTICE "Readonly device detected. No cluster "
-		       "services will be utilized for this mount. Recovery "
-		       "will be skipped.\n");
+		printk(KERN_NOTICE "ocfs2: Readonly device (%s) detected. "
+		       "Cluster services will not be used for this mount. "
+		       "Recovery will be skipped.\n", osb->dev_str);
 	}
 
 	if (!ocfs2_is_hard_readonly(osb)) {
@@ -2469,8 +2469,8 @@ static int ocfs2_check_volume(struct ocfs2_super *osb)
 			goto finally;
 		}
 	} else {
-		mlog(ML_NOTICE, "File system was not unmounted cleanly, "
-		     "recovering volume.\n");
+		printk(KERN_NOTICE "ocfs2: File system on device (%s) was not "
+		       "unmounted cleanly, recovering it.\n", osb->dev_str);
 	}
 
 	local = ocfs2_mount_local(osb);
diff --git a/fs/omfs/file.c b/fs/omfs/file.c
index d738a7e493d..2c6d95257a4 100644
--- a/fs/omfs/file.c
+++ b/fs/omfs/file.c
@@ -4,7 +4,6 @@
  * Released under GPL v2.
  */
 
-#include <linux/version.h>
 #include <linux/module.h>
 #include <linux/fs.h>
 #include <linux/buffer_head.h>
diff --git a/fs/open.c b/fs/open.c
index b52cf013ffa..739b751aa73 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -793,7 +793,7 @@ out:
 	return nd->intent.open.file;
 out_err:
 	release_open_intent(nd);
-	nd->intent.open.file = (struct file *)dentry;
+	nd->intent.open.file = ERR_CAST(dentry);
 	goto out;
 }
 EXPORT_SYMBOL_GPL(lookup_instantiate_filp);
diff --git a/fs/partitions/check.c b/fs/partitions/check.c
index f82e762eeca..d545e97d99c 100644
--- a/fs/partitions/check.c
+++ b/fs/partitions/check.c
@@ -255,13 +255,7 @@ ssize_t part_discard_alignment_show(struct device *dev,
 				   struct device_attribute *attr, char *buf)
 {
 	struct hd_struct *p = dev_to_part(dev);
-	struct gendisk *disk = dev_to_disk(dev);
-	unsigned int alignment = 0;
-
-	if (disk->queue)
-		alignment = queue_limit_discard_alignment(&disk->queue->limits,
-								p->start_sect);
-	return sprintf(buf, "%u\n", alignment);
+	return sprintf(buf, "%u\n", p->discard_alignment);
 }
 
 ssize_t part_stat_show(struct device *dev,
@@ -455,6 +449,8 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno,
 	p->start_sect = 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/proc/array.c b/fs/proc/array.c
index 9b45ee84fbc..3a1dafd228d 100644
--- a/fs/proc/array.c
+++ b/fs/proc/array.c
@@ -172,7 +172,7 @@ static inline void task_state(struct seq_file *m, struct pid_namespace *ns,
 		task_tgid_nr_ns(rcu_dereference(p->real_parent), ns) : 0;
 	tpid = 0;
 	if (pid_alive(p)) {
-		struct task_struct *tracer = tracehook_tracer_task(p);
+		struct task_struct *tracer = ptrace_parent(p);
 		if (tracer)
 			tpid = task_pid_nr_ns(tracer, ns);
 	}
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 14def991d9d..91fb655a5cb 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -216,7 +216,7 @@ static struct mm_struct *__check_mem_permission(struct task_struct *task)
 	if (task_is_stopped_or_traced(task)) {
 		int match;
 		rcu_read_lock();
-		match = (tracehook_tracer_task(task) == current);
+		match = (ptrace_parent(task) == current);
 		rcu_read_unlock();
 		if (match && ptrace_may_access(task, PTRACE_MODE_ATTACH))
 			return mm;
@@ -673,7 +673,7 @@ static int mounts_open_common(struct inode *inode, struct file *file,
 	p->m.private = p;
 	p->ns = ns;
 	p->root = root;
-	p->event = ns->event;
+	p->m.poll_event = ns->event;
 
 	return 0;
 
@@ -2167,13 +2167,9 @@ static const struct file_operations proc_fd_operations = {
  * /proc/pid/fd needs a special permission handler so that a process can still
  * access /proc/self/fd after it has executed a setuid().
  */
-static int proc_fd_permission(struct inode *inode, int mask, unsigned int flags)
+static int proc_fd_permission(struct inode *inode, int mask)
 {
-	int rv;
-
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-	rv = generic_permission(inode, mask, flags, NULL);
+	int rv = generic_permission(inode, mask);
 	if (rv == 0)
 		return 0;
 	if (task_pid(current) == proc_pid(inode))
@@ -2712,6 +2708,9 @@ static int do_io_accounting(struct task_struct *task, char *buffer, int whole)
 	struct task_io_accounting acct = task->ioac;
 	unsigned long flags;
 
+	if (!ptrace_may_access(task, PTRACE_MODE_READ))
+		return -EACCES;
+
 	if (whole && lock_task_sighand(task, &flags)) {
 		struct task_struct *t = task;
 
@@ -2843,7 +2842,7 @@ static const struct pid_entry tgid_base_stuff[] = {
 	REG("coredump_filter", S_IRUGO|S_IWUSR, proc_coredump_filter_operations),
 #endif
 #ifdef CONFIG_TASK_IO_ACCOUNTING
-	INF("io",	S_IRUGO, proc_tgid_io_accounting),
+	INF("io",	S_IRUSR, proc_tgid_io_accounting),
 #endif
 #ifdef CONFIG_HARDWALL
 	INF("hardwall",   S_IRUGO, proc_pid_hardwall),
@@ -3185,7 +3184,7 @@ static const struct pid_entry tid_base_stuff[] = {
 	REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations),
 #endif
 #ifdef CONFIG_TASK_IO_ACCOUNTING
-	INF("io",	S_IRUGO, proc_tid_io_accounting),
+	INF("io",	S_IRUSR, proc_tid_io_accounting),
 #endif
 #ifdef CONFIG_HARDWALL
 	INF("hardwall",   S_IRUGO, proc_pid_hardwall),
diff --git a/fs/proc/namespaces.c b/fs/proc/namespaces.c
index 781dec5bd68..be177f702ac 100644
--- a/fs/proc/namespaces.c
+++ b/fs/proc/namespaces.c
@@ -38,18 +38,21 @@ static struct dentry *proc_ns_instantiate(struct inode *dir,
 	struct inode *inode;
 	struct proc_inode *ei;
 	struct dentry *error = ERR_PTR(-ENOENT);
+	void *ns;
 
 	inode = proc_pid_make_inode(dir->i_sb, task);
 	if (!inode)
 		goto out;
 
+	ns = ns_ops->get(task);
+	if (!ns)
+		goto out_iput;
+
 	ei = PROC_I(inode);
 	inode->i_mode = S_IFREG|S_IRUSR;
 	inode->i_fop  = &ns_file_operations;
 	ei->ns_ops    = ns_ops;
-	ei->ns	      = ns_ops->get(task);
-	if (!ei->ns)
-		goto out_iput;
+	ei->ns	      = ns;
 
 	dentry->d_op = &pid_dentry_operations;
 	d_add(dentry, inode);
diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index f50133c11c2..1a77dbef226 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -294,7 +294,7 @@ out:
 	return ret;
 }
 
-static int proc_sys_permission(struct inode *inode, int mask,unsigned int flags)
+static int proc_sys_permission(struct inode *inode, int mask)
 {
 	/*
 	 * sysctl entries that are not writeable,
@@ -304,9 +304,6 @@ static int proc_sys_permission(struct inode *inode, int mask,unsigned int flags)
 	struct ctl_table *table;
 	int error;
 
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
-
 	/* Executable files are not allowed under /proc/sys/ */
 	if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode))
 		return -EACCES;
@@ -319,7 +316,7 @@ static int proc_sys_permission(struct inode *inode, int mask,unsigned int flags)
 	if (!table) /* global root - r-xr-xr-x */
 		error = mask & MAY_WRITE ? -EACCES : 0;
 	else /* Use the permissions on the sysctl table entry */
-		error = sysctl_perm(head->root, table, mask);
+		error = sysctl_perm(head->root, table, mask & ~MAY_NOT_BLOCK);
 
 	sysctl_head_finish(head);
 	return error;
diff --git a/fs/proc/root.c b/fs/proc/root.c
index a9000e9cfee..d6c3b416529 100644
--- a/fs/proc/root.c
+++ b/fs/proc/root.c
@@ -28,11 +28,12 @@ static int proc_test_super(struct super_block *sb, void *data)
 
 static int proc_set_super(struct super_block *sb, void *data)
 {
-	struct pid_namespace *ns;
-
-	ns = (struct pid_namespace *)data;
-	sb->s_fs_info = get_pid_ns(ns);
-	return set_anon_super(sb, NULL);
+	int err = set_anon_super(sb, NULL);
+	if (!err) {
+		struct pid_namespace *ns = (struct pid_namespace *)data;
+		sb->s_fs_info = get_pid_ns(ns);
+	}
+	return err;
 }
 
 static struct dentry *proc_mount(struct file_system_type *fs_type,
diff --git a/fs/read_write.c b/fs/read_write.c
index 5520f8ad550..5907b49e4d7 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -64,6 +64,23 @@ generic_file_llseek_unlocked(struct file *file, loff_t offset, int origin)
 			return file->f_pos;
 		offset += file->f_pos;
 		break;
+	case SEEK_DATA:
+		/*
+		 * In the generic case the entire file is data, so as long as
+		 * offset isn't at the end of the file then the offset is data.
+		 */
+		if (offset >= inode->i_size)
+			return -ENXIO;
+		break;
+	case SEEK_HOLE:
+		/*
+		 * There is a virtual hole at the end of the file, so as long as
+		 * offset isn't i_size or larger, return i_size.
+		 */
+		if (offset >= inode->i_size)
+			return -ENXIO;
+		offset = inode->i_size;
+		break;
 	}
 
 	if (offset < 0 && !unsigned_offsets(file))
@@ -128,12 +145,13 @@ EXPORT_SYMBOL(no_llseek);
 
 loff_t default_llseek(struct file *file, loff_t offset, int origin)
 {
+	struct inode *inode = file->f_path.dentry->d_inode;
 	loff_t retval;
 
-	mutex_lock(&file->f_dentry->d_inode->i_mutex);
+	mutex_lock(&inode->i_mutex);
 	switch (origin) {
 		case SEEK_END:
-			offset += i_size_read(file->f_path.dentry->d_inode);
+			offset += i_size_read(inode);
 			break;
 		case SEEK_CUR:
 			if (offset == 0) {
@@ -141,6 +159,26 @@ loff_t default_llseek(struct file *file, loff_t offset, int origin)
 				goto out;
 			}
 			offset += file->f_pos;
+			break;
+		case SEEK_DATA:
+			/*
+			 * In the generic case the entire file is data, so as
+			 * long as offset isn't at the end of the file then the
+			 * offset is data.
+			 */
+			if (offset >= inode->i_size)
+				return -ENXIO;
+			break;
+		case SEEK_HOLE:
+			/*
+			 * There is a virtual hole at the end of the file, so
+			 * as long as offset isn't i_size or larger, return
+			 * i_size.
+			 */
+			if (offset >= inode->i_size)
+				return -ENXIO;
+			offset = inode->i_size;
+			break;
 	}
 	retval = -EINVAL;
 	if (offset >= 0 || unsigned_offsets(file)) {
@@ -151,7 +189,7 @@ loff_t default_llseek(struct file *file, loff_t offset, int origin)
 		retval = offset;
 	}
 out:
-	mutex_unlock(&file->f_dentry->d_inode->i_mutex);
+	mutex_unlock(&inode->i_mutex);
 	return retval;
 }
 EXPORT_SYMBOL(default_llseek);
diff --git a/fs/reiserfs/dir.c b/fs/reiserfs/dir.c
index 198dabf1b2b..133e9355dc6 100644
--- a/fs/reiserfs/dir.c
+++ b/fs/reiserfs/dir.c
@@ -14,7 +14,8 @@
 extern const struct reiserfs_key MIN_KEY;
 
 static int reiserfs_readdir(struct file *, void *, filldir_t);
-static int reiserfs_dir_fsync(struct file *filp, int datasync);
+static int reiserfs_dir_fsync(struct file *filp, loff_t start, loff_t end,
+			      int datasync);
 
 const struct file_operations reiserfs_dir_operations = {
 	.llseek = generic_file_llseek,
@@ -27,13 +28,21 @@ const struct file_operations reiserfs_dir_operations = {
 #endif
 };
 
-static int reiserfs_dir_fsync(struct file *filp, int datasync)
+static int reiserfs_dir_fsync(struct file *filp, loff_t start, loff_t end,
+			      int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int err;
+
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	reiserfs_write_lock(inode->i_sb);
 	err = reiserfs_commit_for_inode(inode);
 	reiserfs_write_unlock(inode->i_sb);
+	mutex_unlock(&inode->i_mutex);
 	if (err < 0)
 		return err;
 	return 0;
diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c
index 91f080cc76c..c7156dc39ce 100644
--- a/fs/reiserfs/file.c
+++ b/fs/reiserfs/file.c
@@ -140,12 +140,18 @@ static void reiserfs_vfs_truncate_file(struct inode *inode)
  * be removed...
  */
 
-static int reiserfs_sync_file(struct file *filp, int datasync)
+static int reiserfs_sync_file(struct file *filp, loff_t start, loff_t end,
+			      int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int err;
 	int barrier_done;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	BUG_ON(!S_ISREG(inode->i_mode));
 	err = sync_mapping_buffers(inode->i_mapping);
 	reiserfs_write_lock(inode->i_sb);
@@ -153,6 +159,7 @@ static int reiserfs_sync_file(struct file *filp, int datasync)
 	reiserfs_write_unlock(inode->i_sb);
 	if (barrier_done != 1 && reiserfs_barrier_flush(inode->i_sb))
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
+	mutex_unlock(&inode->i_mutex);
 	if (barrier_done < 0)
 		return barrier_done;
 	return (err < 0) ? -EIO : 0;
@@ -312,4 +319,5 @@ const struct inode_operations reiserfs_file_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
+	.check_acl = reiserfs_check_acl,
 };
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 4fd5bb33dbb..2922b90ceac 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -3068,9 +3068,8 @@ static ssize_t reiserfs_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs,
-				  reiserfs_get_blocks_direct_io, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				  reiserfs_get_blocks_direct_io);
 
 	/*
 	 * In case of error extending write may have instantiated a few
@@ -3114,6 +3113,9 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr)
 			error = -EFBIG;
 			goto out;
 		}
+
+		inode_dio_wait(inode);
+
 		/* fill in hole pointers in the expanding truncate case. */
 		if (attr->ia_size > inode->i_size) {
 			error = generic_cont_expand_simple(inode, attr->ia_size);
diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c
index 118662690cd..551f1b79dbc 100644
--- a/fs/reiserfs/namei.c
+++ b/fs/reiserfs/namei.c
@@ -1529,6 +1529,7 @@ const struct inode_operations reiserfs_dir_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
+	.check_acl = reiserfs_check_acl,
 };
 
 /*
@@ -1545,6 +1546,7 @@ const struct inode_operations reiserfs_symlink_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
+	.check_acl = reiserfs_check_acl,
 
 };
 
@@ -1558,5 +1560,5 @@ const struct inode_operations reiserfs_special_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
-
+	.check_acl = reiserfs_check_acl,
 };
diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c
index aa91089162c..14363b96b6a 100644
--- a/fs/reiserfs/super.c
+++ b/fs/reiserfs/super.c
@@ -1643,6 +1643,7 @@ static int reiserfs_fill_super(struct super_block *s, void *data, int silent)
 	/* Set default values for options: non-aggressive tails, RO on errors */
 	REISERFS_SB(s)->s_mount_opt |= (1 << REISERFS_SMALLTAIL);
 	REISERFS_SB(s)->s_mount_opt |= (1 << REISERFS_ERROR_RO);
+	REISERFS_SB(s)->s_mount_opt |= (1 << REISERFS_BARRIER_FLUSH);
 	/* no preallocation minimum, be smart in
 	   reiserfs_file_write instead */
 	REISERFS_SB(s)->s_alloc_options.preallocmin = 0;
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index e8a62f41b45..6938d8c68d6 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -555,11 +555,10 @@ reiserfs_xattr_set_handle(struct reiserfs_transaction_handle *th,
 
 		reiserfs_write_unlock(inode->i_sb);
 		mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_XATTR);
-		down_write(&dentry->d_inode->i_alloc_sem);
+		inode_dio_wait(dentry->d_inode);
 		reiserfs_write_lock(inode->i_sb);
 
 		err = reiserfs_setattr(dentry, &newattrs);
-		up_write(&dentry->d_inode->i_alloc_sem);
 		mutex_unlock(&dentry->d_inode->i_mutex);
 	} else
 		update_ctime(inode);
@@ -868,12 +867,18 @@ out:
 	return err;
 }
 
-static int reiserfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int reiserfs_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	int error = -EAGAIN; /* do regular unix permission checks by default */
 
-	if (flags & IPERM_FLAG_RCU)
+	/*
+	 * Stat data v1 doesn't support ACLs.
+	 */
+	if (get_inode_sd_version(inode) == STAT_DATA_V1)
+		return -EAGAIN;
+
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	acl = reiserfs_get_acl(inode, ACL_TYPE_ACCESS);
@@ -952,10 +957,8 @@ static int xattr_mount_check(struct super_block *s)
 	return 0;
 }
 
-int reiserfs_permission(struct inode *inode, int mask, unsigned int flags)
+int reiserfs_permission(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU)
-		return -ECHILD;
 	/*
 	 * We don't do permission checks on the internal objects.
 	 * Permissions are determined by the "owning" object.
@@ -963,15 +966,7 @@ int reiserfs_permission(struct inode *inode, int mask, unsigned int flags)
 	if (IS_PRIVATE(inode))
 		return 0;
 
-#ifdef CONFIG_REISERFS_FS_XATTR
-	/*
-	 * Stat data v1 doesn't support ACLs.
-	 */
-	if (get_inode_sd_version(inode) != STAT_DATA_V1)
-		return generic_permission(inode, mask, flags,
-					reiserfs_check_acl);
-#endif
-	return generic_permission(inode, mask, flags, NULL);
+	return generic_permission(inode, mask);
 }
 
 static int xattr_hide_revalidate(struct dentry *dentry, struct nameidata *nd)
diff --git a/fs/romfs/mmap-nommu.c b/fs/romfs/mmap-nommu.c
index f0511e81696..eed99428f10 100644
--- a/fs/romfs/mmap-nommu.c
+++ b/fs/romfs/mmap-nommu.c
@@ -27,14 +27,18 @@ static unsigned long romfs_get_unmapped_area(struct file *file,
 {
 	struct inode *inode = file->f_mapping->host;
 	struct mtd_info *mtd = inode->i_sb->s_mtd;
-	unsigned long isize, offset;
+	unsigned long isize, offset, maxpages, lpages;
 
 	if (!mtd)
 		goto cant_map_directly;
 
+	/* the mapping mustn't extend beyond the EOF */
+	lpages = (len + PAGE_SIZE - 1) >> PAGE_SHIFT;
 	isize = i_size_read(inode);
 	offset = pgoff << PAGE_SHIFT;
-	if (offset > isize || len > isize || offset > isize - len)
+
+	maxpages = (isize + PAGE_SIZE - 1) >> PAGE_SHIFT;
+	if ((pgoff >= maxpages) || (maxpages - pgoff < lpages))
 		return (unsigned long) -EINVAL;
 
 	/* we need to call down to the MTD layer to do the actual mapping */
diff --git a/fs/squashfs/namei.c b/fs/squashfs/namei.c
index 4bc63ac64bc..0682b38d7e3 100644
--- a/fs/squashfs/namei.c
+++ b/fs/squashfs/namei.c
@@ -220,11 +220,6 @@ static struct dentry *squashfs_lookup(struct inode *dir, struct dentry *dentry,
 					blk, off, ino_num);
 
 				inode = squashfs_iget(dir->i_sb, ino, ino_num);
-				if (IS_ERR(inode)) {
-					err = PTR_ERR(inode);
-					goto failed;
-				}
-
 				goto exit_lookup;
 			}
 		}
@@ -232,10 +227,7 @@ static struct dentry *squashfs_lookup(struct inode *dir, struct dentry *dentry,
 
 exit_lookup:
 	kfree(dire);
-	if (inode)
-		return d_splice_alias(inode, dentry);
-	d_add(dentry, inode);
-	return ERR_PTR(0);
+	return d_splice_alias(inode, dentry);
 
 data_error:
 	err = -EIO;
diff --git a/fs/super.c b/fs/super.c
index c75593953c5..7943f04cb3a 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -38,6 +38,69 @@
 LIST_HEAD(super_blocks);
 DEFINE_SPINLOCK(sb_lock);
 
+/*
+ * One thing we have to be careful of with a per-sb shrinker is that we don't
+ * drop the last active reference to the superblock from within the shrinker.
+ * If that happens we could trigger unregistering the shrinker from within the
+ * shrinker path and that leads to deadlock on the shrinker_rwsem. Hence we
+ * take a passive reference to the superblock to avoid this from occurring.
+ */
+static int prune_super(struct shrinker *shrink, struct shrink_control *sc)
+{
+	struct super_block *sb;
+	int	fs_objects = 0;
+	int	total_objects;
+
+	sb = container_of(shrink, struct super_block, s_shrink);
+
+	/*
+	 * Deadlock avoidance.  We may hold various FS locks, and we don't want
+	 * to recurse into the FS that called us in clear_inode() and friends..
+	 */
+	if (sc->nr_to_scan && !(sc->gfp_mask & __GFP_FS))
+		return -1;
+
+	if (!grab_super_passive(sb))
+		return -1;
+
+	if (sb->s_op && sb->s_op->nr_cached_objects)
+		fs_objects = sb->s_op->nr_cached_objects(sb);
+
+	total_objects = sb->s_nr_dentry_unused +
+			sb->s_nr_inodes_unused + fs_objects + 1;
+
+	if (sc->nr_to_scan) {
+		int	dentries;
+		int	inodes;
+
+		/* proportion the scan between the caches */
+		dentries = (sc->nr_to_scan * sb->s_nr_dentry_unused) /
+							total_objects;
+		inodes = (sc->nr_to_scan * sb->s_nr_inodes_unused) /
+							total_objects;
+		if (fs_objects)
+			fs_objects = (sc->nr_to_scan * fs_objects) /
+							total_objects;
+		/*
+		 * prune the dcache first as the icache is pinned by it, then
+		 * prune the icache, followed by the filesystem specific caches
+		 */
+		prune_dcache_sb(sb, dentries);
+		prune_icache_sb(sb, inodes);
+
+		if (fs_objects && sb->s_op->free_cached_objects) {
+			sb->s_op->free_cached_objects(sb, fs_objects);
+			fs_objects = sb->s_op->nr_cached_objects(sb);
+		}
+		total_objects = sb->s_nr_dentry_unused +
+				sb->s_nr_inodes_unused + fs_objects;
+	}
+
+	total_objects = (total_objects / 100) * sysctl_vfs_cache_pressure;
+	drop_super(sb);
+	return total_objects;
+}
+
 /**
  *	alloc_super	-	create new superblock
  *	@type:	filesystem type superblock should belong to
@@ -77,6 +140,8 @@ static struct super_block *alloc_super(struct file_system_type *type)
 		INIT_HLIST_BL_HEAD(&s->s_anon);
 		INIT_LIST_HEAD(&s->s_inodes);
 		INIT_LIST_HEAD(&s->s_dentry_lru);
+		INIT_LIST_HEAD(&s->s_inode_lru);
+		spin_lock_init(&s->s_inode_lru_lock);
 		init_rwsem(&s->s_umount);
 		mutex_init(&s->s_lock);
 		lockdep_set_class(&s->s_umount, &type->s_umount_key);
@@ -114,6 +179,10 @@ static struct super_block *alloc_super(struct file_system_type *type)
 		s->s_op = &default_op;
 		s->s_time_gran = 1000000000;
 		s->cleancache_poolid = -1;
+
+		s->s_shrink.seeks = DEFAULT_SEEKS;
+		s->s_shrink.shrink = prune_super;
+		s->s_shrink.batch = 1024;
 	}
 out:
 	return s;
@@ -181,6 +250,10 @@ void deactivate_locked_super(struct super_block *s)
 	if (atomic_dec_and_test(&s->s_active)) {
 		cleancache_flush_fs(s);
 		fs->kill_sb(s);
+
+		/* caches are now gone, we can safely kill the shrinker now */
+		unregister_shrinker(&s->s_shrink);
+
 		/*
 		 * We need to call rcu_barrier so all the delayed rcu free
 		 * inodes are flushed before we release the fs module.
@@ -241,6 +314,39 @@ static int grab_super(struct super_block *s) __releases(sb_lock)
 }
 
 /*
+ *	grab_super_passive - acquire a passive reference
+ *	@s: reference we are trying to grab
+ *
+ *	Tries to acquire a passive reference. This is used in places where we
+ *	cannot take an active reference but we need to ensure that the
+ *	superblock does not go away while we are working on it. It returns
+ *	false if a reference was not gained, and returns true with the s_umount
+ *	lock held in read mode if a reference is gained. On successful return,
+ *	the caller must drop the s_umount lock and the passive reference when
+ *	done.
+ */
+bool grab_super_passive(struct super_block *sb)
+{
+	spin_lock(&sb_lock);
+	if (list_empty(&sb->s_instances)) {
+		spin_unlock(&sb_lock);
+		return false;
+	}
+
+	sb->s_count++;
+	spin_unlock(&sb_lock);
+
+	if (down_read_trylock(&sb->s_umount)) {
+		if (sb->s_root)
+			return true;
+		up_read(&sb->s_umount);
+	}
+
+	put_super(sb);
+	return false;
+}
+
+/*
  * Superblock locking.  We really ought to get rid of these two.
  */
 void lock_super(struct super_block * sb)
@@ -276,7 +382,6 @@ void generic_shutdown_super(struct super_block *sb)
 {
 	const struct super_operations *sop = sb->s_op;
 
-
 	if (sb->s_root) {
 		shrink_dcache_for_umount(sb);
 		sync_filesystem(sb);
@@ -364,6 +469,7 @@ retry:
 	list_add(&s->s_instances, &type->fs_supers);
 	spin_unlock(&sb_lock);
 	get_filesystem(type);
+	register_shrinker(&s->s_shrink);
 	return s;
 }
 
@@ -452,6 +558,42 @@ void iterate_supers(void (*f)(struct super_block *, void *), void *arg)
 }
 
 /**
+ *	iterate_supers_type - call function for superblocks of given type
+ *	@type: fs type
+ *	@f: function to call
+ *	@arg: argument to pass to it
+ *
+ *	Scans the superblock list and calls given function, passing it
+ *	locked superblock and given argument.
+ */
+void iterate_supers_type(struct file_system_type *type,
+	void (*f)(struct super_block *, void *), void *arg)
+{
+	struct super_block *sb, *p = NULL;
+
+	spin_lock(&sb_lock);
+	list_for_each_entry(sb, &type->fs_supers, s_instances) {
+		sb->s_count++;
+		spin_unlock(&sb_lock);
+
+		down_read(&sb->s_umount);
+		if (sb->s_root)
+			f(sb, arg);
+		up_read(&sb->s_umount);
+
+		spin_lock(&sb_lock);
+		if (p)
+			__put_super(p);
+		p = sb;
+	}
+	if (p)
+		__put_super(p);
+	spin_unlock(&sb_lock);
+}
+
+EXPORT_SYMBOL(iterate_supers_type);
+
+/**
  *	get_super - get the superblock of a device
  *	@bdev: device to get the superblock for
  *	
@@ -657,7 +799,7 @@ static DEFINE_IDA(unnamed_dev_ida);
 static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */
 static int unnamed_dev_start = 0; /* don't bother trying below it */
 
-int set_anon_super(struct super_block *s, void *data)
+int get_anon_bdev(dev_t *p)
 {
 	int dev;
 	int error;
@@ -684,24 +826,38 @@ int set_anon_super(struct super_block *s, void *data)
 		spin_unlock(&unnamed_dev_lock);
 		return -EMFILE;
 	}
-	s->s_dev = MKDEV(0, dev & MINORMASK);
-	s->s_bdi = &noop_backing_dev_info;
+	*p = MKDEV(0, dev & MINORMASK);
 	return 0;
 }
+EXPORT_SYMBOL(get_anon_bdev);
 
-EXPORT_SYMBOL(set_anon_super);
-
-void kill_anon_super(struct super_block *sb)
+void free_anon_bdev(dev_t dev)
 {
-	int slot = MINOR(sb->s_dev);
-
-	generic_shutdown_super(sb);
+	int slot = MINOR(dev);
 	spin_lock(&unnamed_dev_lock);
 	ida_remove(&unnamed_dev_ida, slot);
 	if (slot < unnamed_dev_start)
 		unnamed_dev_start = slot;
 	spin_unlock(&unnamed_dev_lock);
 }
+EXPORT_SYMBOL(free_anon_bdev);
+
+int set_anon_super(struct super_block *s, void *data)
+{
+	int error = get_anon_bdev(&s->s_dev);
+	if (!error)
+		s->s_bdi = &noop_backing_dev_info;
+	return error;
+}
+
+EXPORT_SYMBOL(set_anon_super);
+
+void kill_anon_super(struct super_block *sb)
+{
+	dev_t dev = sb->s_dev;
+	generic_shutdown_super(sb);
+	free_anon_bdev(dev);
+}
 
 EXPORT_SYMBOL(kill_anon_super);
 
@@ -822,7 +978,7 @@ struct dentry *mount_bdev(struct file_system_type *fs_type,
 	} else {
 		char b[BDEVNAME_SIZE];
 
-		s->s_flags = flags;
+		s->s_flags = flags | MS_NOSEC;
 		s->s_mode = mode;
 		strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
 		sb_set_blocksize(s, block_size(bdev));
diff --git a/fs/sync.c b/fs/sync.c
index c38ec163da6..c98a7477edf 100644
--- a/fs/sync.c
+++ b/fs/sync.c
@@ -165,28 +165,9 @@ SYSCALL_DEFINE1(syncfs, int, fd)
  */
 int vfs_fsync_range(struct file *file, loff_t start, loff_t end, int datasync)
 {
-	struct address_space *mapping = file->f_mapping;
-	int err, ret;
-
-	if (!file->f_op || !file->f_op->fsync) {
-		ret = -EINVAL;
-		goto out;
-	}
-
-	ret = filemap_write_and_wait_range(mapping, start, end);
-
-	/*
-	 * We need to protect against concurrent writers, which could cause
-	 * livelocks in fsync_buffers_list().
-	 */
-	mutex_lock(&mapping->host->i_mutex);
-	err = file->f_op->fsync(file, datasync);
-	if (!ret)
-		ret = err;
-	mutex_unlock(&mapping->host->i_mutex);
-
-out:
-	return ret;
+	if (!file->f_op || !file->f_op->fsync)
+		return -EINVAL;
+	return file->f_op->fsync(file, start, end, datasync);
 }
 EXPORT_SYMBOL(vfs_fsync_range);
 
diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c
index 0a12eb89cd3..e3f091a81c7 100644
--- a/fs/sysfs/inode.c
+++ b/fs/sysfs/inode.c
@@ -349,11 +349,11 @@ int sysfs_hash_and_remove(struct sysfs_dirent *dir_sd, const void *ns, const cha
 		return -ENOENT;
 }
 
-int sysfs_permission(struct inode *inode, int mask, unsigned int flags)
+int sysfs_permission(struct inode *inode, int mask)
 {
 	struct sysfs_dirent *sd;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	sd = inode->i_private;
@@ -362,5 +362,5 @@ int sysfs_permission(struct inode *inode, int mask, unsigned int flags)
 	sysfs_refresh_inode(sd, inode);
 	mutex_unlock(&sysfs_mutex);
 
-	return generic_permission(inode, mask, flags, NULL);
+	return generic_permission(inode, mask);
 }
diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c
index 266895783b4..e34f0d99ea4 100644
--- a/fs/sysfs/mount.c
+++ b/fs/sysfs/mount.c
@@ -95,6 +95,14 @@ static int sysfs_set_super(struct super_block *sb, void *data)
 	return error;
 }
 
+static void free_sysfs_super_info(struct sysfs_super_info *info)
+{
+	int type;
+	for (type = KOBJ_NS_TYPE_NONE; type < KOBJ_NS_TYPES; type++)
+		kobj_ns_drop(type, info->ns[type]);
+	kfree(info);
+}
+
 static struct dentry *sysfs_mount(struct file_system_type *fs_type,
 	int flags, const char *dev_name, void *data)
 {
@@ -108,11 +116,11 @@ static struct dentry *sysfs_mount(struct file_system_type *fs_type,
 		return ERR_PTR(-ENOMEM);
 
 	for (type = KOBJ_NS_TYPE_NONE; type < KOBJ_NS_TYPES; type++)
-		info->ns[type] = kobj_ns_current(type);
+		info->ns[type] = kobj_ns_grab_current(type);
 
 	sb = sget(fs_type, sysfs_test_super, sysfs_set_super, info);
 	if (IS_ERR(sb) || sb->s_fs_info != info)
-		kfree(info);
+		free_sysfs_super_info(info);
 	if (IS_ERR(sb))
 		return ERR_CAST(sb);
 	if (!sb->s_root) {
@@ -131,12 +139,11 @@ static struct dentry *sysfs_mount(struct file_system_type *fs_type,
 static void sysfs_kill_sb(struct super_block *sb)
 {
 	struct sysfs_super_info *info = sysfs_info(sb);
-
 	/* Remove the superblock from fs_supers/s_instances
 	 * so we can't find it, before freeing sysfs_super_info.
 	 */
 	kill_anon_super(sb);
-	kfree(info);
+	free_sysfs_super_info(info);
 }
 
 static struct file_system_type sysfs_fs_type = {
@@ -145,28 +152,6 @@ static struct file_system_type sysfs_fs_type = {
 	.kill_sb	= sysfs_kill_sb,
 };
 
-void sysfs_exit_ns(enum kobj_ns_type type, const void *ns)
-{
-	struct super_block *sb;
-
-	mutex_lock(&sysfs_mutex);
-	spin_lock(&sb_lock);
-	list_for_each_entry(sb, &sysfs_fs_type.fs_supers, s_instances) {
-		struct sysfs_super_info *info = sysfs_info(sb);
-		/*
-		 * If we see a superblock on the fs_supers/s_instances
-		 * list the unmount has not completed and sb->s_fs_info
-		 * points to a valid struct sysfs_super_info.
-		 */
-		/* Ignore superblocks with the wrong ns */
-		if (info->ns[type] != ns)
-			continue;
-		info->ns[type] = NULL;
-	}
-	spin_unlock(&sb_lock);
-	mutex_unlock(&sysfs_mutex);
-}
-
 int __init sysfs_init(void)
 {
 	int err = -ENOMEM;
diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h
index 3d28af31d86..845ab3ad229 100644
--- a/fs/sysfs/sysfs.h
+++ b/fs/sysfs/sysfs.h
@@ -136,7 +136,7 @@ struct sysfs_addrm_cxt {
  * instance).
  */
 struct sysfs_super_info {
-	const void *ns[KOBJ_NS_TYPES];
+	void *ns[KOBJ_NS_TYPES];
 };
 #define sysfs_info(SB) ((struct sysfs_super_info *)(SB->s_fs_info))
 extern struct sysfs_dirent sysfs_root;
@@ -201,7 +201,7 @@ static inline void __sysfs_put(struct sysfs_dirent *sd)
 struct inode *sysfs_get_inode(struct super_block *sb, struct sysfs_dirent *sd);
 void sysfs_evict_inode(struct inode *inode);
 int sysfs_sd_setattr(struct sysfs_dirent *sd, struct iattr *iattr);
-int sysfs_permission(struct inode *inode, int mask, unsigned int flags);
+int sysfs_permission(struct inode *inode, int mask);
 int sysfs_setattr(struct dentry *dentry, struct iattr *iattr);
 int sysfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat);
 int sysfs_setxattr(struct dentry *dentry, const char *name, const void *value,
diff --git a/fs/timerfd.c b/fs/timerfd.c
index f67acbdda5e..dffeb3795af 100644
--- a/fs/timerfd.c
+++ b/fs/timerfd.c
@@ -61,7 +61,9 @@ static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr)
 
 /*
  * Called when the clock was set to cancel the timers in the cancel
- * list.
+ * list. This will wake up processes waiting on these timers. The
+ * wake-up requires ctx->ticks to be non zero, therefore we increment
+ * it before calling wake_up_locked().
  */
 void timerfd_clock_was_set(void)
 {
@@ -76,6 +78,7 @@ void timerfd_clock_was_set(void)
 		spin_lock_irqsave(&ctx->wqh.lock, flags);
 		if (ctx->moffs.tv64 != moffs.tv64) {
 			ctx->moffs.tv64 = KTIME_MAX;
+			ctx->ticks++;
 			wake_up_locked(&ctx->wqh);
 		}
 		spin_unlock_irqrestore(&ctx->wqh.lock, flags);
diff --git a/fs/ubifs/commit.c b/fs/ubifs/commit.c
index 87cd0ead863..fb3b5c813a3 100644
--- a/fs/ubifs/commit.c
+++ b/fs/ubifs/commit.c
@@ -78,7 +78,7 @@ static int nothing_to_commit(struct ubifs_info *c)
 	 * If the root TNC node is dirty, we definitely have something to
 	 * commit.
 	 */
-	if (c->zroot.znode && test_bit(DIRTY_ZNODE, &c->zroot.znode->flags))
+	if (c->zroot.znode && ubifs_zn_dirty(c->zroot.znode))
 		return 0;
 
 	/*
@@ -418,7 +418,7 @@ int ubifs_run_commit(struct ubifs_info *c)
 
 	spin_lock(&c->cs_lock);
 	if (c->cmt_state == COMMIT_BROKEN) {
-		err = -EINVAL;
+		err = -EROFS;
 		goto out;
 	}
 
@@ -444,7 +444,7 @@ int ubifs_run_commit(struct ubifs_info *c)
 	 * re-check it.
 	 */
 	if (c->cmt_state == COMMIT_BROKEN) {
-		err = -EINVAL;
+		err = -EROFS;
 		goto out_cmt_unlock;
 	}
 
@@ -576,7 +576,7 @@ int dbg_check_old_index(struct ubifs_info *c, struct ubifs_zbranch *zroot)
 	struct idx_node *i;
 	size_t sz;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_OLD_IDX))
+	if (!dbg_is_chk_index(c))
 		return 0;
 
 	INIT_LIST_HEAD(&list);
diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c
index 0bb2bcef0de..eef109a1a92 100644
--- a/fs/ubifs/debug.c
+++ b/fs/ubifs/debug.c
@@ -27,13 +27,12 @@
  * various local functions of those subsystems.
  */
 
-#define UBIFS_DBG_PRESERVE_UBI
-
-#include "ubifs.h"
 #include <linux/module.h>
-#include <linux/moduleparam.h>
 #include <linux/debugfs.h>
 #include <linux/math64.h>
+#include <linux/uaccess.h>
+#include <linux/random.h>
+#include "ubifs.h"
 
 #ifdef CONFIG_UBIFS_FS_DEBUG
 
@@ -42,15 +41,6 @@ DEFINE_SPINLOCK(dbg_lock);
 static char dbg_key_buf0[128];
 static char dbg_key_buf1[128];
 
-unsigned int ubifs_chk_flags;
-unsigned int ubifs_tst_flags;
-
-module_param_named(debug_chks, ubifs_chk_flags, uint, S_IRUGO | S_IWUSR);
-module_param_named(debug_tsts, ubifs_tst_flags, uint, S_IRUGO | S_IWUSR);
-
-MODULE_PARM_DESC(debug_chks, "Debug check flags");
-MODULE_PARM_DESC(debug_tsts, "Debug special test flags");
-
 static const char *get_key_fmt(int fmt)
 {
 	switch (fmt) {
@@ -91,6 +81,28 @@ static const char *get_key_type(int type)
 	}
 }
 
+static const char *get_dent_type(int type)
+{
+	switch (type) {
+	case UBIFS_ITYPE_REG:
+		return "file";
+	case UBIFS_ITYPE_DIR:
+		return "dir";
+	case UBIFS_ITYPE_LNK:
+		return "symlink";
+	case UBIFS_ITYPE_BLK:
+		return "blkdev";
+	case UBIFS_ITYPE_CHR:
+		return "char dev";
+	case UBIFS_ITYPE_FIFO:
+		return "fifo";
+	case UBIFS_ITYPE_SOCK:
+		return "socket";
+	default:
+		return "unknown/invalid type";
+	}
+}
+
 static void sprintf_key(const struct ubifs_info *c, const union ubifs_key *key,
 			char *buffer)
 {
@@ -234,9 +246,13 @@ static void dump_ch(const struct ubifs_ch *ch)
 	printk(KERN_DEBUG "\tlen            %u\n", le32_to_cpu(ch->len));
 }
 
-void dbg_dump_inode(const struct ubifs_info *c, const struct inode *inode)
+void dbg_dump_inode(struct ubifs_info *c, const struct inode *inode)
 {
 	const struct ubifs_inode *ui = ubifs_inode(inode);
+	struct qstr nm = { .name = NULL };
+	union ubifs_key key;
+	struct ubifs_dent_node *dent, *pdent = NULL;
+	int count = 2;
 
 	printk(KERN_DEBUG "Dump in-memory inode:");
 	printk(KERN_DEBUG "\tinode          %lu\n", inode->i_ino);
@@ -270,6 +286,32 @@ void dbg_dump_inode(const struct ubifs_info *c, const struct inode *inode)
 	printk(KERN_DEBUG "\tlast_page_read %lu\n", ui->last_page_read);
 	printk(KERN_DEBUG "\tread_in_a_row  %lu\n", ui->read_in_a_row);
 	printk(KERN_DEBUG "\tdata_len       %d\n", ui->data_len);
+
+	if (!S_ISDIR(inode->i_mode))
+		return;
+
+	printk(KERN_DEBUG "List of directory entries:\n");
+	ubifs_assert(!mutex_is_locked(&c->tnc_mutex));
+
+	lowest_dent_key(c, &key, inode->i_ino);
+	while (1) {
+		dent = ubifs_tnc_next_ent(c, &key, &nm);
+		if (IS_ERR(dent)) {
+			if (PTR_ERR(dent) != -ENOENT)
+				printk(KERN_DEBUG "error %ld\n", PTR_ERR(dent));
+			break;
+		}
+
+		printk(KERN_DEBUG "\t%d: %s (%s)\n",
+		       count++, dent->name, get_dent_type(dent->type));
+
+		nm.name = dent->name;
+		nm.len = le16_to_cpu(dent->nlen);
+		kfree(pdent);
+		pdent = dent;
+		key_read(c, &dent->key, &key);
+	}
+	kfree(pdent);
 }
 
 void dbg_dump_node(const struct ubifs_info *c, const void *node)
@@ -278,7 +320,7 @@ void dbg_dump_node(const struct ubifs_info *c, const void *node)
 	union ubifs_key key;
 	const struct ubifs_ch *ch = node;
 
-	if (dbg_failure_mode)
+	if (dbg_is_tst_rcvry(c))
 		return;
 
 	/* If the magic is incorrect, just hexdump the first bytes */
@@ -834,7 +876,7 @@ void dbg_dump_leb(const struct ubifs_info *c, int lnum)
 	struct ubifs_scan_node *snod;
 	void *buf;
 
-	if (dbg_failure_mode)
+	if (dbg_is_tst_rcvry(c))
 		return;
 
 	printk(KERN_DEBUG "(pid %d) start dumping LEB %d\n",
@@ -1080,6 +1122,7 @@ out:
 
 /**
  * dbg_check_synced_i_size - check synchronized inode size.
+ * @c: UBIFS file-system description object
  * @inode: inode to check
  *
  * If inode is clean, synchronized inode size has to be equivalent to current
@@ -1087,12 +1130,12 @@ out:
  * has to be locked). Returns %0 if synchronized inode size if correct, and
  * %-EINVAL if not.
  */
-int dbg_check_synced_i_size(struct inode *inode)
+int dbg_check_synced_i_size(const struct ubifs_info *c, struct inode *inode)
 {
 	int err = 0;
 	struct ubifs_inode *ui = ubifs_inode(inode);
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 	if (!S_ISREG(inode->i_mode))
 		return 0;
@@ -1125,7 +1168,7 @@ int dbg_check_synced_i_size(struct inode *inode)
  * Note, it is good idea to make sure the @dir->i_mutex is locked before
  * calling this function.
  */
-int dbg_check_dir_size(struct ubifs_info *c, const struct inode *dir)
+int dbg_check_dir(struct ubifs_info *c, const struct inode *dir)
 {
 	unsigned int nlink = 2;
 	union ubifs_key key;
@@ -1133,7 +1176,7 @@ int dbg_check_dir_size(struct ubifs_info *c, const struct inode *dir)
 	struct qstr nm = { .name = NULL };
 	loff_t size = UBIFS_INO_NODE_SZ;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 
 	if (!S_ISDIR(dir->i_mode))
@@ -1167,12 +1210,14 @@ int dbg_check_dir_size(struct ubifs_info *c, const struct inode *dir)
 			  "but calculated size is %llu", dir->i_ino,
 			  (unsigned long long)i_size_read(dir),
 			  (unsigned long long)size);
+		dbg_dump_inode(c, dir);
 		dump_stack();
 		return -EINVAL;
 	}
 	if (dir->i_nlink != nlink) {
 		ubifs_err("directory inode %lu has nlink %u, but calculated "
 			  "nlink is %u", dir->i_ino, dir->i_nlink, nlink);
+		dbg_dump_inode(c, dir);
 		dump_stack();
 		return -EINVAL;
 	}
@@ -1489,7 +1534,7 @@ int dbg_check_tnc(struct ubifs_info *c, int extra)
 	long clean_cnt = 0, dirty_cnt = 0;
 	int err, last;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_TNC))
+	if (!dbg_is_chk_index(c))
 		return 0;
 
 	ubifs_assert(mutex_is_locked(&c->tnc_mutex));
@@ -1736,7 +1781,7 @@ int dbg_check_idx_size(struct ubifs_info *c, long long idx_size)
 	int err;
 	long long calc = 0;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_IDX_SZ))
+	if (!dbg_is_chk_index(c))
 		return 0;
 
 	err = dbg_walk_index(c, NULL, add_size, &calc);
@@ -2312,7 +2357,7 @@ int dbg_check_filesystem(struct ubifs_info *c)
 	int err;
 	struct fsck_data fsckd;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_FS))
+	if (!dbg_is_chk_fs(c))
 		return 0;
 
 	fsckd.inodes = RB_ROOT;
@@ -2347,7 +2392,7 @@ int dbg_check_data_nodes_order(struct ubifs_info *c, struct list_head *head)
 	struct list_head *cur;
 	struct ubifs_scan_node *sa, *sb;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 
 	for (cur = head->next; cur->next != head; cur = cur->next) {
@@ -2414,7 +2459,7 @@ int dbg_check_nondata_nodes_order(struct ubifs_info *c, struct list_head *head)
 	struct list_head *cur;
 	struct ubifs_scan_node *sa, *sb;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 
 	for (cur = head->next; cur->next != head; cur = cur->next) {
@@ -2491,214 +2536,141 @@ error_dump:
 	return 0;
 }
 
-int dbg_force_in_the_gaps(void)
+static inline int chance(unsigned int n, unsigned int out_of)
 {
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
-		return 0;
+	return !!((random32() % out_of) + 1 <= n);
 
-	return !(random32() & 7);
 }
 
-/* Failure mode for recovery testing */
-
-#define chance(n, d) (simple_rand() <= (n) * 32768LL / (d))
-
-struct failure_mode_info {
-	struct list_head list;
-	struct ubifs_info *c;
-};
-
-static LIST_HEAD(fmi_list);
-static DEFINE_SPINLOCK(fmi_lock);
-
-static unsigned int next;
-
-static int simple_rand(void)
-{
-	if (next == 0)
-		next = current->pid;
-	next = next * 1103515245 + 12345;
-	return (next >> 16) & 32767;
-}
-
-static void failure_mode_init(struct ubifs_info *c)
-{
-	struct failure_mode_info *fmi;
-
-	fmi = kmalloc(sizeof(struct failure_mode_info), GFP_NOFS);
-	if (!fmi) {
-		ubifs_err("Failed to register failure mode - no memory");
-		return;
-	}
-	fmi->c = c;
-	spin_lock(&fmi_lock);
-	list_add_tail(&fmi->list, &fmi_list);
-	spin_unlock(&fmi_lock);
-}
-
-static void failure_mode_exit(struct ubifs_info *c)
+static int power_cut_emulated(struct ubifs_info *c, int lnum, int write)
 {
-	struct failure_mode_info *fmi, *tmp;
-
-	spin_lock(&fmi_lock);
-	list_for_each_entry_safe(fmi, tmp, &fmi_list, list)
-		if (fmi->c == c) {
-			list_del(&fmi->list);
-			kfree(fmi);
-		}
-	spin_unlock(&fmi_lock);
-}
-
-static struct ubifs_info *dbg_find_info(struct ubi_volume_desc *desc)
-{
-	struct failure_mode_info *fmi;
-
-	spin_lock(&fmi_lock);
-	list_for_each_entry(fmi, &fmi_list, list)
-		if (fmi->c->ubi == desc) {
-			struct ubifs_info *c = fmi->c;
-
-			spin_unlock(&fmi_lock);
-			return c;
-		}
-	spin_unlock(&fmi_lock);
-	return NULL;
-}
-
-static int in_failure_mode(struct ubi_volume_desc *desc)
-{
-	struct ubifs_info *c = dbg_find_info(desc);
-
-	if (c && dbg_failure_mode)
-		return c->dbg->failure_mode;
-	return 0;
-}
+	struct ubifs_debug_info *d = c->dbg;
 
-static int do_fail(struct ubi_volume_desc *desc, int lnum, int write)
-{
-	struct ubifs_info *c = dbg_find_info(desc);
-	struct ubifs_debug_info *d;
+	ubifs_assert(dbg_is_tst_rcvry(c));
 
-	if (!c || !dbg_failure_mode)
-		return 0;
-	d = c->dbg;
-	if (d->failure_mode)
-		return 1;
-	if (!d->fail_cnt) {
-		/* First call - decide delay to failure */
+	if (!d->pc_cnt) {
+		/* First call - decide delay to the power cut */
 		if (chance(1, 2)) {
-			unsigned int delay = 1 << (simple_rand() >> 11);
+			unsigned long delay;
 
 			if (chance(1, 2)) {
-				d->fail_delay = 1;
-				d->fail_timeout = jiffies +
-						  msecs_to_jiffies(delay);
-				dbg_rcvry("failing after %ums", delay);
+				d->pc_delay = 1;
+				/* Fail withing 1 minute */
+				delay = random32() % 60000;
+				d->pc_timeout = jiffies;
+				d->pc_timeout += msecs_to_jiffies(delay);
+				ubifs_warn("failing after %lums", delay);
 			} else {
-				d->fail_delay = 2;
-				d->fail_cnt_max = delay;
-				dbg_rcvry("failing after %u calls", delay);
+				d->pc_delay = 2;
+				delay = random32() % 10000;
+				/* Fail within 10000 operations */
+				d->pc_cnt_max = delay;
+				ubifs_warn("failing after %lu calls", delay);
 			}
 		}
-		d->fail_cnt += 1;
+
+		d->pc_cnt += 1;
 	}
+
 	/* Determine if failure delay has expired */
-	if (d->fail_delay == 1) {
-		if (time_before(jiffies, d->fail_timeout))
+	if (d->pc_delay == 1 && time_before(jiffies, d->pc_timeout))
 			return 0;
-	} else if (d->fail_delay == 2)
-		if (d->fail_cnt++ < d->fail_cnt_max)
+	if (d->pc_delay == 2 && d->pc_cnt++ < d->pc_cnt_max)
 			return 0;
+
 	if (lnum == UBIFS_SB_LNUM) {
-		if (write) {
-			if (chance(1, 2))
-				return 0;
-		} else if (chance(19, 20))
+		if (write && chance(1, 2))
+			return 0;
+		if (chance(19, 20))
 			return 0;
-		dbg_rcvry("failing in super block LEB %d", lnum);
+		ubifs_warn("failing in super block LEB %d", lnum);
 	} else if (lnum == UBIFS_MST_LNUM || lnum == UBIFS_MST_LNUM + 1) {
 		if (chance(19, 20))
 			return 0;
-		dbg_rcvry("failing in master LEB %d", lnum);
+		ubifs_warn("failing in master LEB %d", lnum);
 	} else if (lnum >= UBIFS_LOG_LNUM && lnum <= c->log_last) {
-		if (write) {
-			if (chance(99, 100))
-				return 0;
-		} else if (chance(399, 400))
+		if (write && chance(99, 100))
 			return 0;
-		dbg_rcvry("failing in log LEB %d", lnum);
+		if (chance(399, 400))
+			return 0;
+		ubifs_warn("failing in log LEB %d", lnum);
 	} else if (lnum >= c->lpt_first && lnum <= c->lpt_last) {
-		if (write) {
-			if (chance(7, 8))
-				return 0;
-		} else if (chance(19, 20))
+		if (write && chance(7, 8))
 			return 0;
-		dbg_rcvry("failing in LPT LEB %d", lnum);
+		if (chance(19, 20))
+			return 0;
+		ubifs_warn("failing in LPT LEB %d", lnum);
 	} else if (lnum >= c->orph_first && lnum <= c->orph_last) {
-		if (write) {
-			if (chance(1, 2))
-				return 0;
-		} else if (chance(9, 10))
+		if (write && chance(1, 2))
+			return 0;
+		if (chance(9, 10))
 			return 0;
-		dbg_rcvry("failing in orphan LEB %d", lnum);
+		ubifs_warn("failing in orphan LEB %d", lnum);
 	} else if (lnum == c->ihead_lnum) {
 		if (chance(99, 100))
 			return 0;
-		dbg_rcvry("failing in index head LEB %d", lnum);
+		ubifs_warn("failing in index head LEB %d", lnum);
 	} else if (c->jheads && lnum == c->jheads[GCHD].wbuf.lnum) {
 		if (chance(9, 10))
 			return 0;
-		dbg_rcvry("failing in GC head LEB %d", lnum);
+		ubifs_warn("failing in GC head LEB %d", lnum);
 	} else if (write && !RB_EMPTY_ROOT(&c->buds) &&
 		   !ubifs_search_bud(c, lnum)) {
 		if (chance(19, 20))
 			return 0;
-		dbg_rcvry("failing in non-bud LEB %d", lnum);
+		ubifs_warn("failing in non-bud LEB %d", lnum);
 	} else if (c->cmt_state == COMMIT_RUNNING_BACKGROUND ||
 		   c->cmt_state == COMMIT_RUNNING_REQUIRED) {
 		if (chance(999, 1000))
 			return 0;
-		dbg_rcvry("failing in bud LEB %d commit running", lnum);
+		ubifs_warn("failing in bud LEB %d commit running", lnum);
 	} else {
 		if (chance(9999, 10000))
 			return 0;
-		dbg_rcvry("failing in bud LEB %d commit not running", lnum);
+		ubifs_warn("failing in bud LEB %d commit not running", lnum);
 	}
-	ubifs_err("*** SETTING FAILURE MODE ON (LEB %d) ***", lnum);
-	d->failure_mode = 1;
+
+	d->pc_happened = 1;
+	ubifs_warn("========== Power cut emulated ==========");
 	dump_stack();
 	return 1;
 }
 
-static void cut_data(const void *buf, int len)
+static void cut_data(const void *buf, unsigned int len)
 {
-	int flen, i;
+	unsigned int from, to, i, ffs = chance(1, 2);
 	unsigned char *p = (void *)buf;
 
-	flen = (len * (long long)simple_rand()) >> 15;
-	for (i = flen; i < len; i++)
-		p[i] = 0xff;
-}
+	from = random32() % (len + 1);
+	if (chance(1, 2))
+		to = random32() % (len - from + 1);
+	else
+		to = len;
 
-int dbg_leb_read(struct ubi_volume_desc *desc, int lnum, char *buf, int offset,
-		 int len, int check)
-{
-	if (in_failure_mode(desc))
-		return -EROFS;
-	return ubi_leb_read(desc, lnum, buf, offset, len, check);
+	if (from < to)
+		ubifs_warn("filled bytes %u-%u with %s", from, to - 1,
+			   ffs ? "0xFFs" : "random data");
+
+	if (ffs)
+		for (i = from; i < to; i++)
+			p[i] = 0xFF;
+	else
+		for (i = from; i < to; i++)
+			p[i] = random32() % 0x100;
 }
 
-int dbg_leb_write(struct ubi_volume_desc *desc, int lnum, const void *buf,
-		  int offset, int len, int dtype)
+int dbg_leb_write(struct ubifs_info *c, int lnum, const void *buf,
+		  int offs, int len, int dtype)
 {
 	int err, failing;
 
-	if (in_failure_mode(desc))
+	if (c->dbg->pc_happened)
 		return -EROFS;
-	failing = do_fail(desc, lnum, 1);
+
+	failing = power_cut_emulated(c, lnum, 1);
 	if (failing)
 		cut_data(buf, len);
-	err = ubi_leb_write(desc, lnum, buf, offset, len, dtype);
+	err = ubi_leb_write(c->ubi, lnum, buf, offs, len, dtype);
 	if (err)
 		return err;
 	if (failing)
@@ -2706,162 +2678,207 @@ int dbg_leb_write(struct ubi_volume_desc *desc, int lnum, const void *buf,
 	return 0;
 }
 
-int dbg_leb_change(struct ubi_volume_desc *desc, int lnum, const void *buf,
+int dbg_leb_change(struct ubifs_info *c, int lnum, const void *buf,
 		   int len, int dtype)
 {
 	int err;
 
-	if (do_fail(desc, lnum, 1))
+	if (c->dbg->pc_happened)
 		return -EROFS;
-	err = ubi_leb_change(desc, lnum, buf, len, dtype);
+	if (power_cut_emulated(c, lnum, 1))
+		return -EROFS;
+	err = ubi_leb_change(c->ubi, lnum, buf, len, dtype);
 	if (err)
 		return err;
-	if (do_fail(desc, lnum, 1))
+	if (power_cut_emulated(c, lnum, 1))
 		return -EROFS;
 	return 0;
 }
 
-int dbg_leb_erase(struct ubi_volume_desc *desc, int lnum)
+int dbg_leb_unmap(struct ubifs_info *c, int lnum)
 {
 	int err;
 
-	if (do_fail(desc, lnum, 0))
+	if (c->dbg->pc_happened)
+		return -EROFS;
+	if (power_cut_emulated(c, lnum, 0))
 		return -EROFS;
-	err = ubi_leb_erase(desc, lnum);
+	err = ubi_leb_unmap(c->ubi, lnum);
 	if (err)
 		return err;
-	if (do_fail(desc, lnum, 0))
+	if (power_cut_emulated(c, lnum, 0))
 		return -EROFS;
 	return 0;
 }
 
-int dbg_leb_unmap(struct ubi_volume_desc *desc, int lnum)
+int dbg_leb_map(struct ubifs_info *c, int lnum, int dtype)
 {
 	int err;
 
-	if (do_fail(desc, lnum, 0))
+	if (c->dbg->pc_happened)
 		return -EROFS;
-	err = ubi_leb_unmap(desc, lnum);
+	if (power_cut_emulated(c, lnum, 0))
+		return -EROFS;
+	err = ubi_leb_map(c->ubi, lnum, dtype);
 	if (err)
 		return err;
-	if (do_fail(desc, lnum, 0))
+	if (power_cut_emulated(c, lnum, 0))
 		return -EROFS;
 	return 0;
 }
 
-int dbg_is_mapped(struct ubi_volume_desc *desc, int lnum)
-{
-	if (in_failure_mode(desc))
-		return -EROFS;
-	return ubi_is_mapped(desc, lnum);
-}
+/*
+ * Root directory for UBIFS stuff in debugfs. Contains sub-directories which
+ * contain the stuff specific to particular file-system mounts.
+ */
+static struct dentry *dfs_rootdir;
 
-int dbg_leb_map(struct ubi_volume_desc *desc, int lnum, int dtype)
+static int dfs_file_open(struct inode *inode, struct file *file)
 {
-	int err;
-
-	if (do_fail(desc, lnum, 0))
-		return -EROFS;
-	err = ubi_leb_map(desc, lnum, dtype);
-	if (err)
-		return err;
-	if (do_fail(desc, lnum, 0))
-		return -EROFS;
-	return 0;
+	file->private_data = inode->i_private;
+	return nonseekable_open(inode, file);
 }
 
 /**
- * ubifs_debugging_init - initialize UBIFS debugging.
- * @c: UBIFS file-system description object
+ * provide_user_output - provide output to the user reading a debugfs file.
+ * @val: boolean value for the answer
+ * @u: the buffer to store the answer at
+ * @count: size of the buffer
+ * @ppos: position in the @u output buffer
  *
- * This function initializes debugging-related data for the file system.
- * Returns zero in case of success and a negative error code in case of
+ * This is a simple helper function which stores @val boolean value in the user
+ * buffer when the user reads one of UBIFS debugfs files. Returns amount of
+ * bytes written to @u in case of success and a negative error code in case of
  * failure.
  */
-int ubifs_debugging_init(struct ubifs_info *c)
+static int provide_user_output(int val, char __user *u, size_t count,
+			       loff_t *ppos)
 {
-	c->dbg = kzalloc(sizeof(struct ubifs_debug_info), GFP_KERNEL);
-	if (!c->dbg)
-		return -ENOMEM;
+	char buf[3];
 
-	failure_mode_init(c);
-	return 0;
+	if (val)
+		buf[0] = '1';
+	else
+		buf[0] = '0';
+	buf[1] = '\n';
+	buf[2] = 0x00;
+
+	return simple_read_from_buffer(u, count, ppos, buf, 2);
 }
 
-/**
- * ubifs_debugging_exit - free debugging data.
- * @c: UBIFS file-system description object
- */
-void ubifs_debugging_exit(struct ubifs_info *c)
+static ssize_t dfs_file_read(struct file *file, char __user *u, size_t count,
+			     loff_t *ppos)
 {
-	failure_mode_exit(c);
-	kfree(c->dbg);
-}
+	struct dentry *dent = file->f_path.dentry;
+	struct ubifs_info *c = file->private_data;
+	struct ubifs_debug_info *d = c->dbg;
+	int val;
+
+	if (dent == d->dfs_chk_gen)
+		val = d->chk_gen;
+	else if (dent == d->dfs_chk_index)
+		val = d->chk_index;
+	else if (dent == d->dfs_chk_orph)
+		val = d->chk_orph;
+	else if (dent == d->dfs_chk_lprops)
+		val = d->chk_lprops;
+	else if (dent == d->dfs_chk_fs)
+		val = d->chk_fs;
+	else if (dent == d->dfs_tst_rcvry)
+		val = d->tst_rcvry;
+	else
+		return -EINVAL;
 
-/*
- * Root directory for UBIFS stuff in debugfs. Contains sub-directories which
- * contain the stuff specific to particular file-system mounts.
- */
-static struct dentry *dfs_rootdir;
+	return provide_user_output(val, u, count, ppos);
+}
 
 /**
- * dbg_debugfs_init - initialize debugfs file-system.
+ * interpret_user_input - interpret user debugfs file input.
+ * @u: user-provided buffer with the input
+ * @count: buffer size
  *
- * UBIFS uses debugfs file-system to expose various debugging knobs to
- * user-space. This function creates "ubifs" directory in the debugfs
- * file-system. Returns zero in case of success and a negative error code in
- * case of failure.
+ * This is a helper function which interpret user input to a boolean UBIFS
+ * debugfs file. Returns %0 or %1 in case of success and a negative error code
+ * in case of failure.
  */
-int dbg_debugfs_init(void)
+static int interpret_user_input(const char __user *u, size_t count)
 {
-	dfs_rootdir = debugfs_create_dir("ubifs", NULL);
-	if (IS_ERR(dfs_rootdir)) {
-		int err = PTR_ERR(dfs_rootdir);
-		ubifs_err("cannot create \"ubifs\" debugfs directory, "
-			  "error %d\n", err);
-		return err;
-	}
+	size_t buf_size;
+	char buf[8];
 
-	return 0;
-}
+	buf_size = min_t(size_t, count, (sizeof(buf) - 1));
+	if (copy_from_user(buf, u, buf_size))
+		return -EFAULT;
 
-/**
- * dbg_debugfs_exit - remove the "ubifs" directory from debugfs file-system.
- */
-void dbg_debugfs_exit(void)
-{
-	debugfs_remove(dfs_rootdir);
-}
+	if (buf[0] == '1')
+		return 1;
+	else if (buf[0] == '0')
+		return 0;
 
-static int open_debugfs_file(struct inode *inode, struct file *file)
-{
-	file->private_data = inode->i_private;
-	return nonseekable_open(inode, file);
+	return -EINVAL;
 }
 
-static ssize_t write_debugfs_file(struct file *file, const char __user *buf,
-				  size_t count, loff_t *ppos)
+static ssize_t dfs_file_write(struct file *file, const char __user *u,
+			      size_t count, loff_t *ppos)
 {
 	struct ubifs_info *c = file->private_data;
 	struct ubifs_debug_info *d = c->dbg;
+	struct dentry *dent = file->f_path.dentry;
+	int val;
 
-	if (file->f_path.dentry == d->dfs_dump_lprops)
+	/*
+	 * TODO: this is racy - the file-system might have already been
+	 * unmounted and we'd oops in this case. The plan is to fix it with
+	 * help of 'iterate_supers_type()' which we should have in v3.0: when
+	 * a debugfs opened, we rember FS's UUID in file->private_data. Then
+	 * whenever we access the FS via a debugfs file, we iterate all UBIFS
+	 * superblocks and fine the one with the same UUID, and take the
+	 * locking right.
+	 *
+	 * The other way to go suggested by Al Viro is to create a separate
+	 * 'ubifs-debug' file-system instead.
+	 */
+	if (file->f_path.dentry == d->dfs_dump_lprops) {
 		dbg_dump_lprops(c);
-	else if (file->f_path.dentry == d->dfs_dump_budg)
+		return count;
+	}
+	if (file->f_path.dentry == d->dfs_dump_budg) {
 		dbg_dump_budg(c, &c->bi);
-	else if (file->f_path.dentry == d->dfs_dump_tnc) {
+		return count;
+	}
+	if (file->f_path.dentry == d->dfs_dump_tnc) {
 		mutex_lock(&c->tnc_mutex);
 		dbg_dump_tnc(c);
 		mutex_unlock(&c->tnc_mutex);
-	} else
+		return count;
+	}
+
+	val = interpret_user_input(u, count);
+	if (val < 0)
+		return val;
+
+	if (dent == d->dfs_chk_gen)
+		d->chk_gen = val;
+	else if (dent == d->dfs_chk_index)
+		d->chk_index = val;
+	else if (dent == d->dfs_chk_orph)
+		d->chk_orph = val;
+	else if (dent == d->dfs_chk_lprops)
+		d->chk_lprops = val;
+	else if (dent == d->dfs_chk_fs)
+		d->chk_fs = val;
+	else if (dent == d->dfs_tst_rcvry)
+		d->tst_rcvry = val;
+	else
 		return -EINVAL;
 
 	return count;
 }
 
 static const struct file_operations dfs_fops = {
-	.open = open_debugfs_file,
-	.write = write_debugfs_file,
+	.open = dfs_file_open,
+	.read = dfs_file_read,
+	.write = dfs_file_write,
 	.owner = THIS_MODULE,
 	.llseek = no_llseek,
 };
@@ -2880,12 +2897,20 @@ static const struct file_operations dfs_fops = {
  */
 int dbg_debugfs_init_fs(struct ubifs_info *c)
 {
-	int err;
+	int err, n;
 	const char *fname;
 	struct dentry *dent;
 	struct ubifs_debug_info *d = c->dbg;
 
-	sprintf(d->dfs_dir_name, "ubi%d_%d", c->vi.ubi_num, c->vi.vol_id);
+	n = snprintf(d->dfs_dir_name, UBIFS_DFS_DIR_LEN + 1, UBIFS_DFS_DIR_NAME,
+		     c->vi.ubi_num, c->vi.vol_id);
+	if (n == UBIFS_DFS_DIR_LEN) {
+		/* The array size is too small */
+		fname = UBIFS_DFS_DIR_NAME;
+		dent = ERR_PTR(-EINVAL);
+		goto out;
+	}
+
 	fname = d->dfs_dir_name;
 	dent = debugfs_create_dir(fname, dfs_rootdir);
 	if (IS_ERR_OR_NULL(dent))
@@ -2910,13 +2935,55 @@ int dbg_debugfs_init_fs(struct ubifs_info *c)
 		goto out_remove;
 	d->dfs_dump_tnc = dent;
 
+	fname = "chk_general";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, d->dfs_dir, c,
+				   &dfs_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	d->dfs_chk_gen = dent;
+
+	fname = "chk_index";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, d->dfs_dir, c,
+				   &dfs_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	d->dfs_chk_index = dent;
+
+	fname = "chk_orphans";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, d->dfs_dir, c,
+				   &dfs_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	d->dfs_chk_orph = dent;
+
+	fname = "chk_lprops";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, d->dfs_dir, c,
+				   &dfs_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	d->dfs_chk_lprops = dent;
+
+	fname = "chk_fs";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, d->dfs_dir, c,
+				   &dfs_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	d->dfs_chk_fs = dent;
+
+	fname = "tst_recovery";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, d->dfs_dir, c,
+				   &dfs_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	d->dfs_tst_rcvry = dent;
+
 	return 0;
 
 out_remove:
 	debugfs_remove_recursive(d->dfs_dir);
 out:
 	err = dent ? PTR_ERR(dent) : -ENODEV;
-	ubifs_err("cannot create \"%s\" debugfs directory, error %d\n",
+	ubifs_err("cannot create \"%s\" debugfs file or directory, error %d\n",
 		  fname, err);
 	return err;
 }
@@ -2930,4 +2997,179 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c)
 	debugfs_remove_recursive(c->dbg->dfs_dir);
 }
 
+struct ubifs_global_debug_info ubifs_dbg;
+
+static struct dentry *dfs_chk_gen;
+static struct dentry *dfs_chk_index;
+static struct dentry *dfs_chk_orph;
+static struct dentry *dfs_chk_lprops;
+static struct dentry *dfs_chk_fs;
+static struct dentry *dfs_tst_rcvry;
+
+static ssize_t dfs_global_file_read(struct file *file, char __user *u,
+				    size_t count, loff_t *ppos)
+{
+	struct dentry *dent = file->f_path.dentry;
+	int val;
+
+	if (dent == dfs_chk_gen)
+		val = ubifs_dbg.chk_gen;
+	else if (dent == dfs_chk_index)
+		val = ubifs_dbg.chk_index;
+	else if (dent == dfs_chk_orph)
+		val = ubifs_dbg.chk_orph;
+	else if (dent == dfs_chk_lprops)
+		val = ubifs_dbg.chk_lprops;
+	else if (dent == dfs_chk_fs)
+		val = ubifs_dbg.chk_fs;
+	else if (dent == dfs_tst_rcvry)
+		val = ubifs_dbg.tst_rcvry;
+	else
+		return -EINVAL;
+
+	return provide_user_output(val, u, count, ppos);
+}
+
+static ssize_t dfs_global_file_write(struct file *file, const char __user *u,
+				     size_t count, loff_t *ppos)
+{
+	struct dentry *dent = file->f_path.dentry;
+	int val;
+
+	val = interpret_user_input(u, count);
+	if (val < 0)
+		return val;
+
+	if (dent == dfs_chk_gen)
+		ubifs_dbg.chk_gen = val;
+	else if (dent == dfs_chk_index)
+		ubifs_dbg.chk_index = val;
+	else if (dent == dfs_chk_orph)
+		ubifs_dbg.chk_orph = val;
+	else if (dent == dfs_chk_lprops)
+		ubifs_dbg.chk_lprops = val;
+	else if (dent == dfs_chk_fs)
+		ubifs_dbg.chk_fs = val;
+	else if (dent == dfs_tst_rcvry)
+		ubifs_dbg.tst_rcvry = val;
+	else
+		return -EINVAL;
+
+	return count;
+}
+
+static const struct file_operations dfs_global_fops = {
+	.read = dfs_global_file_read,
+	.write = dfs_global_file_write,
+	.owner = THIS_MODULE,
+	.llseek = no_llseek,
+};
+
+/**
+ * dbg_debugfs_init - initialize debugfs file-system.
+ *
+ * UBIFS uses debugfs file-system to expose various debugging knobs to
+ * user-space. This function creates "ubifs" directory in the debugfs
+ * file-system. Returns zero in case of success and a negative error code in
+ * case of failure.
+ */
+int dbg_debugfs_init(void)
+{
+	int err;
+	const char *fname;
+	struct dentry *dent;
+
+	fname = "ubifs";
+	dent = debugfs_create_dir(fname, NULL);
+	if (IS_ERR_OR_NULL(dent))
+		goto out;
+	dfs_rootdir = dent;
+
+	fname = "chk_general";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, dfs_rootdir, NULL,
+				   &dfs_global_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	dfs_chk_gen = dent;
+
+	fname = "chk_index";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, dfs_rootdir, NULL,
+				   &dfs_global_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	dfs_chk_index = dent;
+
+	fname = "chk_orphans";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, dfs_rootdir, NULL,
+				   &dfs_global_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	dfs_chk_orph = dent;
+
+	fname = "chk_lprops";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, dfs_rootdir, NULL,
+				   &dfs_global_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	dfs_chk_lprops = dent;
+
+	fname = "chk_fs";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, dfs_rootdir, NULL,
+				   &dfs_global_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	dfs_chk_fs = dent;
+
+	fname = "tst_recovery";
+	dent = debugfs_create_file(fname, S_IRUSR | S_IWUSR, dfs_rootdir, NULL,
+				   &dfs_global_fops);
+	if (IS_ERR_OR_NULL(dent))
+		goto out_remove;
+	dfs_tst_rcvry = dent;
+
+	return 0;
+
+out_remove:
+	debugfs_remove_recursive(dfs_rootdir);
+out:
+	err = dent ? PTR_ERR(dent) : -ENODEV;
+	ubifs_err("cannot create \"%s\" debugfs file or directory, error %d\n",
+		  fname, err);
+	return err;
+}
+
+/**
+ * dbg_debugfs_exit - remove the "ubifs" directory from debugfs file-system.
+ */
+void dbg_debugfs_exit(void)
+{
+	debugfs_remove_recursive(dfs_rootdir);
+}
+
+/**
+ * ubifs_debugging_init - initialize UBIFS debugging.
+ * @c: UBIFS file-system description object
+ *
+ * This function initializes debugging-related data for the file system.
+ * Returns zero in case of success and a negative error code in case of
+ * failure.
+ */
+int ubifs_debugging_init(struct ubifs_info *c)
+{
+	c->dbg = kzalloc(sizeof(struct ubifs_debug_info), GFP_KERNEL);
+	if (!c->dbg)
+		return -ENOMEM;
+
+	return 0;
+}
+
+/**
+ * ubifs_debugging_exit - free debugging data.
+ * @c: UBIFS file-system description object
+ */
+void ubifs_debugging_exit(struct ubifs_info *c)
+{
+	kfree(c->dbg);
+}
+
 #endif /* CONFIG_UBIFS_FS_DEBUG */
diff --git a/fs/ubifs/debug.h b/fs/ubifs/debug.h
index a811ac4a26b..45174b53437 100644
--- a/fs/ubifs/debug.h
+++ b/fs/ubifs/debug.h
@@ -31,18 +31,25 @@ typedef int (*dbg_znode_callback)(struct ubifs_info *c,
 
 #ifdef CONFIG_UBIFS_FS_DEBUG
 
-#include <linux/random.h>
+/*
+ * The UBIFS debugfs directory name pattern and maximum name length (3 for "ubi"
+ * + 1 for "_" and plus 2x2 for 2 UBI numbers and 1 for the trailing zero byte.
+ */
+#define UBIFS_DFS_DIR_NAME "ubi%d_%d"
+#define UBIFS_DFS_DIR_LEN  (3 + 1 + 2*2 + 1)
 
 /**
  * ubifs_debug_info - per-FS debugging information.
  * @old_zroot: old index root - used by 'dbg_check_old_index()'
  * @old_zroot_level: old index root level - used by 'dbg_check_old_index()'
  * @old_zroot_sqnum: old index root sqnum - used by 'dbg_check_old_index()'
- * @failure_mode: failure mode for recovery testing
- * @fail_delay: 0=>don't delay, 1=>delay a time, 2=>delay a number of calls
- * @fail_timeout: time in jiffies when delay of failure mode expires
- * @fail_cnt: current number of calls to failure mode I/O functions
- * @fail_cnt_max: number of calls by which to delay failure mode
+ *
+ * @pc_happened: non-zero if an emulated power cut happened
+ * @pc_delay: 0=>don't delay, 1=>delay a time, 2=>delay a number of calls
+ * @pc_timeout: time in jiffies when delay of failure mode expires
+ * @pc_cnt: current number of calls to failure mode I/O functions
+ * @pc_cnt_max: number of calls by which to delay failure mode
+ *
  * @chk_lpt_sz: used by LPT tree size checker
  * @chk_lpt_sz2: used by LPT tree size checker
  * @chk_lpt_wastage: used by LPT tree size checker
@@ -56,21 +63,36 @@ typedef int (*dbg_znode_callback)(struct ubifs_info *c,
  * @saved_free: saved amount of free space
  * @saved_idx_gc_cnt: saved value of @c->idx_gc_cnt
  *
+ * @chk_gen: if general extra checks are enabled
+ * @chk_index: if index xtra checks are enabled
+ * @chk_orph: if orphans extra checks are enabled
+ * @chk_lprops: if lprops extra checks are enabled
+ * @chk_fs: if UBIFS contents extra checks are enabled
+ * @tst_rcvry: if UBIFS recovery testing mode enabled
+ *
  * @dfs_dir_name: name of debugfs directory containing this file-system's files
  * @dfs_dir: direntry object of the file-system debugfs directory
  * @dfs_dump_lprops: "dump lprops" debugfs knob
  * @dfs_dump_budg: "dump budgeting information" debugfs knob
  * @dfs_dump_tnc: "dump TNC" debugfs knob
+ * @dfs_chk_gen: debugfs knob to enable UBIFS general extra checks
+ * @dfs_chk_index: debugfs knob to enable UBIFS index extra checks
+ * @dfs_chk_orph: debugfs knob to enable UBIFS orphans extra checks
+ * @dfs_chk_lprops: debugfs knob to enable UBIFS LEP properties extra checks
+ * @dfs_chk_fs: debugfs knob to enable UBIFS contents extra checks
+ * @dfs_tst_rcvry: debugfs knob to enable UBIFS recovery testing
  */
 struct ubifs_debug_info {
 	struct ubifs_zbranch old_zroot;
 	int old_zroot_level;
 	unsigned long long old_zroot_sqnum;
-	int failure_mode;
-	int fail_delay;
-	unsigned long fail_timeout;
-	unsigned int fail_cnt;
-	unsigned int fail_cnt_max;
+
+	int pc_happened;
+	int pc_delay;
+	unsigned long pc_timeout;
+	unsigned int pc_cnt;
+	unsigned int pc_cnt_max;
+
 	long long chk_lpt_sz;
 	long long chk_lpt_sz2;
 	long long chk_lpt_wastage;
@@ -84,11 +106,43 @@ struct ubifs_debug_info {
 	long long saved_free;
 	int saved_idx_gc_cnt;
 
-	char dfs_dir_name[100];
+	unsigned int chk_gen:1;
+	unsigned int chk_index:1;
+	unsigned int chk_orph:1;
+	unsigned int chk_lprops:1;
+	unsigned int chk_fs:1;
+	unsigned int tst_rcvry:1;
+
+	char dfs_dir_name[UBIFS_DFS_DIR_LEN + 1];
 	struct dentry *dfs_dir;
 	struct dentry *dfs_dump_lprops;
 	struct dentry *dfs_dump_budg;
 	struct dentry *dfs_dump_tnc;
+	struct dentry *dfs_chk_gen;
+	struct dentry *dfs_chk_index;
+	struct dentry *dfs_chk_orph;
+	struct dentry *dfs_chk_lprops;
+	struct dentry *dfs_chk_fs;
+	struct dentry *dfs_tst_rcvry;
+};
+
+/**
+ * ubifs_global_debug_info - global (not per-FS) UBIFS debugging information.
+ *
+ * @chk_gen: if general extra checks are enabled
+ * @chk_index: if index xtra checks are enabled
+ * @chk_orph: if orphans extra checks are enabled
+ * @chk_lprops: if lprops extra checks are enabled
+ * @chk_fs: if UBIFS contents extra checks are enabled
+ * @tst_rcvry: if UBIFS recovery testing mode enabled
+ */
+struct ubifs_global_debug_info {
+	unsigned int chk_gen:1;
+	unsigned int chk_index:1;
+	unsigned int chk_orph:1;
+	unsigned int chk_lprops:1;
+	unsigned int chk_fs:1;
+	unsigned int tst_rcvry:1;
 };
 
 #define ubifs_assert(expr) do {                                                \
@@ -127,6 +181,8 @@ const char *dbg_key_str1(const struct ubifs_info *c,
 #define DBGKEY(key) dbg_key_str0(c, (key))
 #define DBGKEY1(key) dbg_key_str1(c, (key))
 
+extern spinlock_t dbg_lock;
+
 #define ubifs_dbg_msg(type, fmt, ...) do {                        \
 	spin_lock(&dbg_lock);                                     \
 	pr_debug("UBIFS DBG " type ": " fmt "\n", ##__VA_ARGS__); \
@@ -162,41 +218,36 @@ const char *dbg_key_str1(const struct ubifs_info *c,
 /* Additional recovery messages */
 #define dbg_rcvry(fmt, ...) ubifs_dbg_msg("rcvry", fmt, ##__VA_ARGS__)
 
-/*
- * Debugging check flags.
- *
- * UBIFS_CHK_GEN: general checks
- * UBIFS_CHK_TNC: check TNC
- * UBIFS_CHK_IDX_SZ: check index size
- * UBIFS_CHK_ORPH: check orphans
- * UBIFS_CHK_OLD_IDX: check the old index
- * UBIFS_CHK_LPROPS: check lprops
- * UBIFS_CHK_FS: check the file-system
- */
-enum {
-	UBIFS_CHK_GEN     = 0x1,
-	UBIFS_CHK_TNC     = 0x2,
-	UBIFS_CHK_IDX_SZ  = 0x4,
-	UBIFS_CHK_ORPH    = 0x8,
-	UBIFS_CHK_OLD_IDX = 0x10,
-	UBIFS_CHK_LPROPS  = 0x20,
-	UBIFS_CHK_FS      = 0x40,
-};
-
-/*
- * Special testing flags.
- *
- * UBIFS_TST_RCVRY: failure mode for recovery testing
- */
-enum {
-	UBIFS_TST_RCVRY             = 0x4,
-};
-
-extern spinlock_t dbg_lock;
+extern struct ubifs_global_debug_info ubifs_dbg;
 
-extern unsigned int ubifs_msg_flags;
-extern unsigned int ubifs_chk_flags;
-extern unsigned int ubifs_tst_flags;
+static inline int dbg_is_chk_gen(const struct ubifs_info *c)
+{
+	return !!(ubifs_dbg.chk_gen || c->dbg->chk_gen);
+}
+static inline int dbg_is_chk_index(const struct ubifs_info *c)
+{
+	return !!(ubifs_dbg.chk_index || c->dbg->chk_index);
+}
+static inline int dbg_is_chk_orph(const struct ubifs_info *c)
+{
+	return !!(ubifs_dbg.chk_orph || c->dbg->chk_orph);
+}
+static inline int dbg_is_chk_lprops(const struct ubifs_info *c)
+{
+	return !!(ubifs_dbg.chk_lprops || c->dbg->chk_lprops);
+}
+static inline int dbg_is_chk_fs(const struct ubifs_info *c)
+{
+	return !!(ubifs_dbg.chk_fs || c->dbg->chk_fs);
+}
+static inline int dbg_is_tst_rcvry(const struct ubifs_info *c)
+{
+	return !!(ubifs_dbg.tst_rcvry || c->dbg->tst_rcvry);
+}
+static inline int dbg_is_power_cut(const struct ubifs_info *c)
+{
+	return !!c->dbg->pc_happened;
+}
 
 int ubifs_debugging_init(struct ubifs_info *c);
 void ubifs_debugging_exit(struct ubifs_info *c);
@@ -207,7 +258,7 @@ const char *dbg_cstate(int cmt_state);
 const char *dbg_jhead(int jhead);
 const char *dbg_get_key_dump(const struct ubifs_info *c,
 			     const union ubifs_key *key);
-void dbg_dump_inode(const struct ubifs_info *c, const struct inode *inode);
+void dbg_dump_inode(struct ubifs_info *c, const struct inode *inode);
 void dbg_dump_node(const struct ubifs_info *c, const void *node);
 void dbg_dump_lpt_node(const struct ubifs_info *c, void *node, int lnum,
 		       int offs);
@@ -240,8 +291,8 @@ int dbg_check_cats(struct ubifs_info *c);
 int dbg_check_ltab(struct ubifs_info *c);
 int dbg_chk_lpt_free_spc(struct ubifs_info *c);
 int dbg_chk_lpt_sz(struct ubifs_info *c, int action, int len);
-int dbg_check_synced_i_size(struct inode *inode);
-int dbg_check_dir_size(struct ubifs_info *c, const struct inode *dir);
+int dbg_check_synced_i_size(const struct ubifs_info *c, struct inode *inode);
+int dbg_check_dir(struct ubifs_info *c, const struct inode *dir);
 int dbg_check_tnc(struct ubifs_info *c, int extra);
 int dbg_check_idx_size(struct ubifs_info *c, long long idx_size);
 int dbg_check_filesystem(struct ubifs_info *c);
@@ -254,54 +305,12 @@ int dbg_check_inode_size(struct ubifs_info *c, const struct inode *inode,
 int dbg_check_data_nodes_order(struct ubifs_info *c, struct list_head *head);
 int dbg_check_nondata_nodes_order(struct ubifs_info *c, struct list_head *head);
 
-/* Force the use of in-the-gaps method for testing */
-static inline int dbg_force_in_the_gaps_enabled(void)
-{
-	return ubifs_chk_flags & UBIFS_CHK_GEN;
-}
-int dbg_force_in_the_gaps(void);
-
-/* Failure mode for recovery testing */
-#define dbg_failure_mode (ubifs_tst_flags & UBIFS_TST_RCVRY)
-
-#ifndef UBIFS_DBG_PRESERVE_UBI
-#define ubi_leb_read   dbg_leb_read
-#define ubi_leb_write  dbg_leb_write
-#define ubi_leb_change dbg_leb_change
-#define ubi_leb_erase  dbg_leb_erase
-#define ubi_leb_unmap  dbg_leb_unmap
-#define ubi_is_mapped  dbg_is_mapped
-#define ubi_leb_map    dbg_leb_map
-#endif
-
-int dbg_leb_read(struct ubi_volume_desc *desc, int lnum, char *buf, int offset,
-		 int len, int check);
-int dbg_leb_write(struct ubi_volume_desc *desc, int lnum, const void *buf,
-		  int offset, int len, int dtype);
-int dbg_leb_change(struct ubi_volume_desc *desc, int lnum, const void *buf,
-		   int len, int dtype);
-int dbg_leb_erase(struct ubi_volume_desc *desc, int lnum);
-int dbg_leb_unmap(struct ubi_volume_desc *desc, int lnum);
-int dbg_is_mapped(struct ubi_volume_desc *desc, int lnum);
-int dbg_leb_map(struct ubi_volume_desc *desc, int lnum, int dtype);
-
-static inline int dbg_read(struct ubi_volume_desc *desc, int lnum, char *buf,
-			   int offset, int len)
-{
-	return dbg_leb_read(desc, lnum, buf, offset, len, 0);
-}
-
-static inline int dbg_write(struct ubi_volume_desc *desc, int lnum,
-			    const void *buf, int offset, int len)
-{
-	return dbg_leb_write(desc, lnum, buf, offset, len, UBI_UNKNOWN);
-}
-
-static inline int dbg_change(struct ubi_volume_desc *desc, int lnum,
-				    const void *buf, int len)
-{
-	return dbg_leb_change(desc, lnum, buf, len, UBI_UNKNOWN);
-}
+int dbg_leb_write(struct ubifs_info *c, int lnum, const void *buf, int offs,
+		  int len, int dtype);
+int dbg_leb_change(struct ubifs_info *c, int lnum, const void *buf, int len,
+		   int dtype);
+int dbg_leb_unmap(struct ubifs_info *c, int lnum);
+int dbg_leb_map(struct ubifs_info *c, int lnum, int dtype);
 
 /* Debugfs-related stuff */
 int dbg_debugfs_init(void);
@@ -313,7 +322,7 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c);
 
 /* Use "if (0)" to make compiler check arguments even if debugging is off */
 #define ubifs_assert(expr)  do {                                               \
-	if (0 && (expr))                                                       \
+	if (0)                                                                 \
 		printk(KERN_CRIT "UBIFS assert failed in %s at %u (pid %d)\n", \
 		       __func__, __LINE__, current->pid);                      \
 } while (0)
@@ -323,6 +332,9 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c);
 		ubifs_err(fmt, ##__VA_ARGS__);     \
 } while (0)
 
+#define DBGKEY(key)  ((char *)(key))
+#define DBGKEY1(key) ((char *)(key))
+
 #define ubifs_dbg_msg(fmt, ...) do {               \
 	if (0)                                     \
 		pr_debug(fmt "\n", ##__VA_ARGS__); \
@@ -346,9 +358,6 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c);
 #define dbg_scan(fmt, ...)  ubifs_dbg_msg(fmt, ##__VA_ARGS__)
 #define dbg_rcvry(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__)
 
-#define DBGKEY(key)  ((char *)(key))
-#define DBGKEY1(key) ((char *)(key))
-
 static inline int ubifs_debugging_init(struct ubifs_info *c)      { return 0; }
 static inline void ubifs_debugging_exit(struct ubifs_info *c)     { return; }
 static inline const char *dbg_ntype(int type)                     { return ""; }
@@ -357,7 +366,7 @@ static inline const char *dbg_jhead(int jhead)                    { return ""; }
 static inline const char *
 dbg_get_key_dump(const struct ubifs_info *c,
 		 const union ubifs_key *key)                      { return ""; }
-static inline void dbg_dump_inode(const struct ubifs_info *c,
+static inline void dbg_dump_inode(struct ubifs_info *c,
 				  const struct inode *inode)      { return; }
 static inline void dbg_dump_node(const struct ubifs_info *c,
 				 const void *node)                { return; }
@@ -409,9 +418,11 @@ static inline int dbg_check_ltab(struct ubifs_info *c)            { return 0; }
 static inline int dbg_chk_lpt_free_spc(struct ubifs_info *c)      { return 0; }
 static inline int dbg_chk_lpt_sz(struct ubifs_info *c,
 				 int action, int len)             { return 0; }
-static inline int dbg_check_synced_i_size(struct inode *inode)    { return 0; }
-static inline int dbg_check_dir_size(struct ubifs_info *c,
-				     const struct inode *dir)     { return 0; }
+static inline int
+dbg_check_synced_i_size(const struct ubifs_info *c,
+			struct inode *inode)                      { return 0; }
+static inline int dbg_check_dir(struct ubifs_info *c,
+				const struct inode *dir)          { return 0; }
 static inline int dbg_check_tnc(struct ubifs_info *c, int extra)  { return 0; }
 static inline int dbg_check_idx_size(struct ubifs_info *c,
 				     long long idx_size)          { return 0; }
@@ -431,9 +442,23 @@ static inline int
 dbg_check_nondata_nodes_order(struct ubifs_info *c,
 			      struct list_head *head)             { return 0; }
 
-static inline int dbg_force_in_the_gaps(void)                     { return 0; }
-#define dbg_force_in_the_gaps_enabled() 0
-#define dbg_failure_mode                0
+static inline int dbg_leb_write(struct ubifs_info *c, int lnum,
+				const void *buf, int offset,
+				int len, int dtype)               { return 0; }
+static inline int dbg_leb_change(struct ubifs_info *c, int lnum,
+				 const void *buf, int len,
+				 int dtype)                       { return 0; }
+static inline int dbg_leb_unmap(struct ubifs_info *c, int lnum)   { return 0; }
+static inline int dbg_leb_map(struct ubifs_info *c, int lnum,
+			      int dtype)                          { return 0; }
+
+static inline int dbg_is_chk_gen(const struct ubifs_info *c)      { return 0; }
+static inline int dbg_is_chk_index(const struct ubifs_info *c)    { return 0; }
+static inline int dbg_is_chk_orph(const struct ubifs_info *c)     { return 0; }
+static inline int dbg_is_chk_lprops(const struct ubifs_info *c)   { return 0; }
+static inline int dbg_is_chk_fs(const struct ubifs_info *c)       { return 0; }
+static inline int dbg_is_tst_rcvry(const struct ubifs_info *c)    { return 0; }
+static inline int dbg_is_power_cut(const struct ubifs_info *c)    { return 0; }
 
 static inline int dbg_debugfs_init(void)                          { return 0; }
 static inline void dbg_debugfs_exit(void)                         { return; }
diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c
index ef5abd38f0b..68349204331 100644
--- a/fs/ubifs/dir.c
+++ b/fs/ubifs/dir.c
@@ -102,7 +102,7 @@ struct inode *ubifs_new_inode(struct ubifs_info *c, const struct inode *dir,
 	 * UBIFS has to fully control "clean <-> dirty" transitions of inodes
 	 * to make budgeting work.
 	 */
-	inode->i_flags |= (S_NOCMTIME);
+	inode->i_flags |= S_NOCMTIME;
 
 	inode_init_owner(inode, dir, mode);
 	inode->i_mtime = inode->i_atime = inode->i_ctime =
@@ -172,9 +172,11 @@ struct inode *ubifs_new_inode(struct ubifs_info *c, const struct inode *dir,
 
 #ifdef CONFIG_UBIFS_FS_DEBUG
 
-static int dbg_check_name(struct ubifs_dent_node *dent, struct qstr *nm)
+static int dbg_check_name(const struct ubifs_info *c,
+			  const struct ubifs_dent_node *dent,
+			  const struct qstr *nm)
 {
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 	if (le16_to_cpu(dent->nlen) != nm->len)
 		return -EINVAL;
@@ -185,7 +187,7 @@ static int dbg_check_name(struct ubifs_dent_node *dent, struct qstr *nm)
 
 #else
 
-#define dbg_check_name(dent, nm) 0
+#define dbg_check_name(c, dent, nm) 0
 
 #endif
 
@@ -219,7 +221,7 @@ static struct dentry *ubifs_lookup(struct inode *dir, struct dentry *dentry,
 		goto out;
 	}
 
-	if (dbg_check_name(dent, &dentry->d_name)) {
+	if (dbg_check_name(c, dent, &dentry->d_name)) {
 		err = -EINVAL;
 		goto out;
 	}
@@ -522,7 +524,7 @@ static int ubifs_link(struct dentry *old_dentry, struct inode *dir,
 	ubifs_assert(mutex_is_locked(&dir->i_mutex));
 	ubifs_assert(mutex_is_locked(&inode->i_mutex));
 
-	err = dbg_check_synced_i_size(inode);
+	err = dbg_check_synced_i_size(c, inode);
 	if (err)
 		return err;
 
@@ -577,7 +579,7 @@ static int ubifs_unlink(struct inode *dir, struct dentry *dentry)
 		inode->i_nlink, dir->i_ino);
 	ubifs_assert(mutex_is_locked(&dir->i_mutex));
 	ubifs_assert(mutex_is_locked(&inode->i_mutex));
-	err = dbg_check_synced_i_size(inode);
+	err = dbg_check_synced_i_size(c, inode);
 	if (err)
 		return err;
 
diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 5e7fccfc4b2..f9c234bf33d 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -1263,7 +1263,7 @@ int ubifs_setattr(struct dentry *dentry, struct iattr *attr)
 	if (err)
 		return err;
 
-	err = dbg_check_synced_i_size(inode);
+	err = dbg_check_synced_i_size(c, inode);
 	if (err)
 		return err;
 
@@ -1304,7 +1304,7 @@ static void *ubifs_follow_link(struct dentry *dentry, struct nameidata *nd)
 	return NULL;
 }
 
-int ubifs_fsync(struct file *file, int datasync)
+int ubifs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ubifs_info *c = inode->i_sb->s_fs_info;
@@ -1319,14 +1319,16 @@ int ubifs_fsync(struct file *file, int datasync)
 		 */
 		return 0;
 
-	/*
-	 * VFS has already synchronized dirty pages for this inode. Synchronize
-	 * the inode unless this is a 'datasync()' call.
-	 */
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&inode->i_mutex);
+
+	/* Synchronize 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, NULL);
 		if (err)
-			return err;
+			goto out;
 	}
 
 	/*
@@ -1334,10 +1336,9 @@ int ubifs_fsync(struct file *file, int datasync)
 	 * them.
 	 */
 	err = ubifs_sync_wbufs_by_inode(c, inode);
-	if (err)
-		return err;
-
-	return 0;
+out:
+	mutex_unlock(&inode->i_mutex);
+	return err;
 }
 
 /**
diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c
index 166951e0dcd..9228950a658 100644
--- a/fs/ubifs/io.c
+++ b/fs/ubifs/io.c
@@ -86,8 +86,125 @@ void ubifs_ro_mode(struct ubifs_info *c, int err)
 		c->no_chk_data_crc = 0;
 		c->vfs_sb->s_flags |= MS_RDONLY;
 		ubifs_warn("switched to read-only mode, error %d", err);
+		dump_stack();
+	}
+}
+
+/*
+ * Below are simple wrappers over UBI I/O functions which include some
+ * additional checks and UBIFS debugging stuff. See corresponding UBI function
+ * for more information.
+ */
+
+int ubifs_leb_read(const struct ubifs_info *c, int lnum, void *buf, int offs,
+		   int len, int even_ebadmsg)
+{
+	int err;
+
+	err = ubi_read(c->ubi, lnum, buf, offs, len);
+	/*
+	 * In case of %-EBADMSG print the error message only if the
+	 * @even_ebadmsg is true.
+	 */
+	if (err && (err != -EBADMSG || even_ebadmsg)) {
+		ubifs_err("reading %d bytes from LEB %d:%d failed, error %d",
+			  len, lnum, offs, err);
+		dbg_dump_stack();
+	}
+	return err;
+}
+
+int ubifs_leb_write(struct ubifs_info *c, int lnum, const void *buf, int offs,
+		    int len, int dtype)
+{
+	int err;
+
+	ubifs_assert(!c->ro_media && !c->ro_mount);
+	if (c->ro_error)
+		return -EROFS;
+	if (!dbg_is_tst_rcvry(c))
+		err = ubi_leb_write(c->ubi, lnum, buf, offs, len, dtype);
+	else
+		err = dbg_leb_write(c, lnum, buf, offs, len, dtype);
+	if (err) {
+		ubifs_err("writing %d bytes to LEB %d:%d failed, error %d",
+			  len, lnum, offs, err);
+		ubifs_ro_mode(c, err);
+		dbg_dump_stack();
+	}
+	return err;
+}
+
+int ubifs_leb_change(struct ubifs_info *c, int lnum, const void *buf, int len,
+		     int dtype)
+{
+	int err;
+
+	ubifs_assert(!c->ro_media && !c->ro_mount);
+	if (c->ro_error)
+		return -EROFS;
+	if (!dbg_is_tst_rcvry(c))
+		err = ubi_leb_change(c->ubi, lnum, buf, len, dtype);
+	else
+		err = dbg_leb_change(c, lnum, buf, len, dtype);
+	if (err) {
+		ubifs_err("changing %d bytes in LEB %d failed, error %d",
+			  len, lnum, err);
+		ubifs_ro_mode(c, err);
+		dbg_dump_stack();
+	}
+	return err;
+}
+
+int ubifs_leb_unmap(struct ubifs_info *c, int lnum)
+{
+	int err;
+
+	ubifs_assert(!c->ro_media && !c->ro_mount);
+	if (c->ro_error)
+		return -EROFS;
+	if (!dbg_is_tst_rcvry(c))
+		err = ubi_leb_unmap(c->ubi, lnum);
+	else
+		err = dbg_leb_unmap(c, lnum);
+	if (err) {
+		ubifs_err("unmap LEB %d failed, error %d", lnum, err);
+		ubifs_ro_mode(c, err);
+		dbg_dump_stack();
+	}
+	return err;
+}
+
+int ubifs_leb_map(struct ubifs_info *c, int lnum, int dtype)
+{
+	int err;
+
+	ubifs_assert(!c->ro_media && !c->ro_mount);
+	if (c->ro_error)
+		return -EROFS;
+	if (!dbg_is_tst_rcvry(c))
+		err = ubi_leb_map(c->ubi, lnum, dtype);
+	else
+		err = dbg_leb_map(c, lnum, dtype);
+	if (err) {
+		ubifs_err("mapping LEB %d failed, error %d", lnum, err);
+		ubifs_ro_mode(c, err);
+		dbg_dump_stack();
+	}
+	return err;
+}
+
+int ubifs_is_mapped(const struct ubifs_info *c, int lnum)
+{
+	int err;
+
+	err = ubi_is_mapped(c->ubi, lnum);
+	if (err < 0) {
+		ubifs_err("ubi_is_mapped failed for LEB %d, error %d",
+			  lnum, err);
 		dbg_dump_stack();
 	}
+	return err;
 }
 
 /**
@@ -406,14 +523,10 @@ int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf)
 	dirt = sync_len - wbuf->used;
 	if (dirt)
 		ubifs_pad(c, wbuf->buf + wbuf->used, dirt);
-	err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs,
-			    sync_len, wbuf->dtype);
-	if (err) {
-		ubifs_err("cannot write %d bytes to LEB %d:%d",
-			  sync_len, wbuf->lnum, wbuf->offs);
-		dbg_dump_stack();
+	err = ubifs_leb_write(c, wbuf->lnum, wbuf->buf, wbuf->offs, sync_len,
+			      wbuf->dtype);
+	if (err)
 		return err;
-	}
 
 	spin_lock(&wbuf->lock);
 	wbuf->offs += sync_len;
@@ -581,6 +694,7 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len)
 	ubifs_assert(wbuf->size % c->min_io_size == 0);
 	ubifs_assert(mutex_is_locked(&wbuf->io_mutex));
 	ubifs_assert(!c->ro_media && !c->ro_mount);
+	ubifs_assert(!c->space_fixup);
 	if (c->leb_size - wbuf->offs >= c->max_write_size)
 		ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size));
 
@@ -604,9 +718,9 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len)
 		if (aligned_len == wbuf->avail) {
 			dbg_io("flush jhead %s wbuf to LEB %d:%d",
 			       dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs);
-			err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf,
-					    wbuf->offs, wbuf->size,
-					    wbuf->dtype);
+			err = ubifs_leb_write(c, wbuf->lnum, wbuf->buf,
+					      wbuf->offs, wbuf->size,
+					      wbuf->dtype);
 			if (err)
 				goto out;
 
@@ -641,8 +755,8 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len)
 		dbg_io("flush jhead %s wbuf to LEB %d:%d",
 		       dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs);
 		memcpy(wbuf->buf + wbuf->used, buf, wbuf->avail);
-		err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs,
-				    wbuf->size, wbuf->dtype);
+		err = ubifs_leb_write(c, wbuf->lnum, wbuf->buf, wbuf->offs,
+				      wbuf->size, wbuf->dtype);
 		if (err)
 			goto out;
 
@@ -660,8 +774,8 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len)
 		 */
 		dbg_io("write %d bytes to LEB %d:%d",
 		       wbuf->size, wbuf->lnum, wbuf->offs);
-		err = ubi_leb_write(c->ubi, wbuf->lnum, buf, wbuf->offs,
-				    wbuf->size, wbuf->dtype);
+		err = ubifs_leb_write(c, wbuf->lnum, buf, wbuf->offs,
+				      wbuf->size, wbuf->dtype);
 		if (err)
 			goto out;
 
@@ -682,8 +796,8 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len)
 		n <<= c->max_write_shift;
 		dbg_io("write %d bytes to LEB %d:%d", n, wbuf->lnum,
 		       wbuf->offs);
-		err = ubi_leb_write(c->ubi, wbuf->lnum, buf + written,
-				    wbuf->offs, n, wbuf->dtype);
+		err = ubifs_leb_write(c, wbuf->lnum, buf + written,
+				      wbuf->offs, n, wbuf->dtype);
 		if (err)
 			goto out;
 		wbuf->offs += n;
@@ -759,18 +873,15 @@ int ubifs_write_node(struct ubifs_info *c, void *buf, int len, int lnum,
 	ubifs_assert(lnum >= 0 && lnum < c->leb_cnt && offs >= 0);
 	ubifs_assert(offs % c->min_io_size == 0 && offs < c->leb_size);
 	ubifs_assert(!c->ro_media && !c->ro_mount);
+	ubifs_assert(!c->space_fixup);
 
 	if (c->ro_error)
 		return -EROFS;
 
 	ubifs_prepare_node(c, buf, len, 1);
-	err = ubi_leb_write(c->ubi, lnum, buf, offs, buf_len, dtype);
-	if (err) {
-		ubifs_err("cannot write %d bytes to LEB %d:%d, error %d",
-			  buf_len, lnum, offs, err);
+	err = ubifs_leb_write(c, lnum, buf, offs, buf_len, dtype);
+	if (err)
 		dbg_dump_node(c, buf);
-		dbg_dump_stack();
-	}
 
 	return err;
 }
@@ -822,13 +933,9 @@ int ubifs_read_node_wbuf(struct ubifs_wbuf *wbuf, void *buf, int type, int len,
 
 	if (rlen > 0) {
 		/* Read everything that goes before write-buffer */
-		err = ubi_read(c->ubi, lnum, buf, offs, rlen);
-		if (err && err != -EBADMSG) {
-			ubifs_err("failed to read node %d from LEB %d:%d, "
-				  "error %d", type, lnum, offs, err);
-			dbg_dump_stack();
+		err = ubifs_leb_read(c, lnum, buf, offs, rlen, 0);
+		if (err && err != -EBADMSG)
 			return err;
-		}
 	}
 
 	if (type != ch->node_type) {
@@ -883,12 +990,9 @@ int ubifs_read_node(const struct ubifs_info *c, void *buf, int type, int len,
 	ubifs_assert(!(offs & 7) && offs < c->leb_size);
 	ubifs_assert(type >= 0 && type < UBIFS_NODE_TYPES_CNT);
 
-	err = ubi_read(c->ubi, lnum, buf, offs, len);
-	if (err && err != -EBADMSG) {
-		ubifs_err("cannot read node %d from LEB %d:%d, error %d",
-			  type, lnum, offs, err);
+	err = ubifs_leb_read(c, lnum, buf, offs, len, 0);
+	if (err && err != -EBADMSG)
 		return err;
-	}
 
 	if (type != ch->node_type) {
 		ubifs_err("bad node type (%d but expected %d)",
diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c
index 34b1679e6e3..cef0460f4c5 100644
--- a/fs/ubifs/journal.c
+++ b/fs/ubifs/journal.c
@@ -669,6 +669,7 @@ out_free:
 
 out_release:
 	release_head(c, BASEHD);
+	kfree(dent);
 out_ro:
 	ubifs_ro_mode(c, err);
 	if (last_reference)
diff --git a/fs/ubifs/log.c b/fs/ubifs/log.c
index affea9494ae..f9fd068d1ae 100644
--- a/fs/ubifs/log.c
+++ b/fs/ubifs/log.c
@@ -262,7 +262,7 @@ int ubifs_add_bud_to_log(struct ubifs_info *c, int jhead, int lnum, int offs)
 		 * an unclean reboot, because the target LEB might have been
 		 * unmapped, but not yet physically erased.
 		 */
-		err = ubi_leb_map(c->ubi, bud->lnum, UBI_SHORTTERM);
+		err = ubifs_leb_map(c, bud->lnum, UBI_SHORTTERM);
 		if (err)
 			goto out_unlock;
 	}
@@ -283,8 +283,6 @@ int ubifs_add_bud_to_log(struct ubifs_info *c, int jhead, int lnum, int offs)
 	return 0;
 
 out_unlock:
-	if (err != -EAGAIN)
-		ubifs_ro_mode(c, err);
 	mutex_unlock(&c->log_mutex);
 	kfree(ref);
 	kfree(bud);
@@ -752,7 +750,7 @@ static int dbg_check_bud_bytes(struct ubifs_info *c)
 	struct ubifs_bud *bud;
 	long long bud_bytes = 0;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 
 	spin_lock(&c->buds_lock);
diff --git a/fs/ubifs/lprops.c b/fs/ubifs/lprops.c
index 667884f4a61..f8a181e647c 100644
--- a/fs/ubifs/lprops.c
+++ b/fs/ubifs/lprops.c
@@ -504,7 +504,7 @@ static int is_lprops_dirty(struct ubifs_info *c, struct ubifs_lprops *lprops)
 	pnode = (struct ubifs_pnode *)container_of(lprops - pos,
 						   struct ubifs_pnode,
 						   lprops[0]);
-	return !test_bit(COW_ZNODE, &pnode->flags) &&
+	return !test_bit(COW_CNODE, &pnode->flags) &&
 	       test_bit(DIRTY_CNODE, &pnode->flags);
 }
 
@@ -860,7 +860,7 @@ int dbg_check_cats(struct ubifs_info *c)
 	struct list_head *pos;
 	int i, cat;
 
-	if (!(ubifs_chk_flags & (UBIFS_CHK_GEN | UBIFS_CHK_LPROPS)))
+	if (!dbg_is_chk_gen(c) && !dbg_is_chk_lprops(c))
 		return 0;
 
 	list_for_each_entry(lprops, &c->empty_list, list) {
@@ -958,7 +958,7 @@ void dbg_check_heap(struct ubifs_info *c, struct ubifs_lpt_heap *heap, int cat,
 {
 	int i = 0, j, err = 0;
 
-	if (!(ubifs_chk_flags & (UBIFS_CHK_GEN | UBIFS_CHK_LPROPS)))
+	if (!dbg_is_chk_gen(c) && !dbg_is_chk_lprops(c))
 		return;
 
 	for (i = 0; i < heap->cnt; i++) {
@@ -1262,7 +1262,7 @@ int dbg_check_lprops(struct ubifs_info *c)
 	int i, err;
 	struct ubifs_lp_stats lst;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS))
+	if (!dbg_is_chk_lprops(c))
 		return 0;
 
 	/*
diff --git a/fs/ubifs/lpt.c b/fs/ubifs/lpt.c
index ef5155e109a..6189c74d97f 100644
--- a/fs/ubifs/lpt.c
+++ b/fs/ubifs/lpt.c
@@ -701,8 +701,8 @@ int ubifs_create_dflt_lpt(struct ubifs_info *c, int *main_lebs, int lpt_first,
 			alen = ALIGN(len, c->min_io_size);
 			set_ltab(c, lnum, c->leb_size - alen, alen - len);
 			memset(p, 0xff, alen - len);
-			err = ubi_leb_change(c->ubi, lnum++, buf, alen,
-					     UBI_SHORTTERM);
+			err = ubifs_leb_change(c, lnum++, buf, alen,
+					       UBI_SHORTTERM);
 			if (err)
 				goto out;
 			p = buf;
@@ -732,8 +732,8 @@ int ubifs_create_dflt_lpt(struct ubifs_info *c, int *main_lebs, int lpt_first,
 				set_ltab(c, lnum, c->leb_size - alen,
 					    alen - len);
 				memset(p, 0xff, alen - len);
-				err = ubi_leb_change(c->ubi, lnum++, buf, alen,
-						     UBI_SHORTTERM);
+				err = ubifs_leb_change(c, lnum++, buf, alen,
+						       UBI_SHORTTERM);
 				if (err)
 					goto out;
 				p = buf;
@@ -780,8 +780,8 @@ int ubifs_create_dflt_lpt(struct ubifs_info *c, int *main_lebs, int lpt_first,
 			alen = ALIGN(len, c->min_io_size);
 			set_ltab(c, lnum, c->leb_size - alen, alen - len);
 			memset(p, 0xff, alen - len);
-			err = ubi_leb_change(c->ubi, lnum++, buf, alen,
-					     UBI_SHORTTERM);
+			err = ubifs_leb_change(c, lnum++, buf, alen,
+					       UBI_SHORTTERM);
 			if (err)
 				goto out;
 			p = buf;
@@ -806,7 +806,7 @@ int ubifs_create_dflt_lpt(struct ubifs_info *c, int *main_lebs, int lpt_first,
 		alen = ALIGN(len, c->min_io_size);
 		set_ltab(c, lnum, c->leb_size - alen, alen - len);
 		memset(p, 0xff, alen - len);
-		err = ubi_leb_change(c->ubi, lnum++, buf, alen, UBI_SHORTTERM);
+		err = ubifs_leb_change(c, lnum++, buf, alen, UBI_SHORTTERM);
 		if (err)
 			goto out;
 		p = buf;
@@ -826,7 +826,7 @@ int ubifs_create_dflt_lpt(struct ubifs_info *c, int *main_lebs, int lpt_first,
 
 	/* Write remaining buffer */
 	memset(p, 0xff, alen - len);
-	err = ubi_leb_change(c->ubi, lnum, buf, alen, UBI_SHORTTERM);
+	err = ubifs_leb_change(c, lnum, buf, alen, UBI_SHORTTERM);
 	if (err)
 		goto out;
 
@@ -1222,7 +1222,7 @@ int ubifs_read_nnode(struct ubifs_info *c, struct ubifs_nnode *parent, int iip)
 		if (c->big_lpt)
 			nnode->num = calc_nnode_num_from_parent(c, parent, iip);
 	} else {
-		err = ubi_read(c->ubi, lnum, buf, offs, c->nnode_sz);
+		err = ubifs_leb_read(c, lnum, buf, offs, c->nnode_sz, 1);
 		if (err)
 			goto out;
 		err = ubifs_unpack_nnode(c, buf, nnode);
@@ -1247,6 +1247,7 @@ int ubifs_read_nnode(struct ubifs_info *c, struct ubifs_nnode *parent, int iip)
 
 out:
 	ubifs_err("error %d reading nnode at %d:%d", err, lnum, offs);
+	dbg_dump_stack();
 	kfree(nnode);
 	return err;
 }
@@ -1290,7 +1291,7 @@ static int read_pnode(struct ubifs_info *c, struct ubifs_nnode *parent, int iip)
 			lprops->flags = ubifs_categorize_lprops(c, lprops);
 		}
 	} else {
-		err = ubi_read(c->ubi, lnum, buf, offs, c->pnode_sz);
+		err = ubifs_leb_read(c, lnum, buf, offs, c->pnode_sz, 1);
 		if (err)
 			goto out;
 		err = unpack_pnode(c, buf, pnode);
@@ -1312,6 +1313,7 @@ static int read_pnode(struct ubifs_info *c, struct ubifs_nnode *parent, int iip)
 out:
 	ubifs_err("error %d reading pnode at %d:%d", err, lnum, offs);
 	dbg_dump_pnode(c, pnode, parent, iip);
+	dbg_dump_stack();
 	dbg_msg("calc num: %d", calc_pnode_num_from_parent(c, parent, iip));
 	kfree(pnode);
 	return err;
@@ -1331,7 +1333,7 @@ static int read_ltab(struct ubifs_info *c)
 	buf = vmalloc(c->ltab_sz);
 	if (!buf)
 		return -ENOMEM;
-	err = ubi_read(c->ubi, c->ltab_lnum, buf, c->ltab_offs, c->ltab_sz);
+	err = ubifs_leb_read(c, c->ltab_lnum, buf, c->ltab_offs, c->ltab_sz, 1);
 	if (err)
 		goto out;
 	err = unpack_ltab(c, buf);
@@ -1354,7 +1356,8 @@ static int read_lsave(struct ubifs_info *c)
 	buf = vmalloc(c->lsave_sz);
 	if (!buf)
 		return -ENOMEM;
-	err = ubi_read(c->ubi, c->lsave_lnum, buf, c->lsave_offs, c->lsave_sz);
+	err = ubifs_leb_read(c, c->lsave_lnum, buf, c->lsave_offs,
+			     c->lsave_sz, 1);
 	if (err)
 		goto out;
 	err = unpack_lsave(c, buf);
@@ -1814,8 +1817,8 @@ static struct ubifs_nnode *scan_get_nnode(struct ubifs_info *c,
 		if (c->big_lpt)
 			nnode->num = calc_nnode_num_from_parent(c, parent, iip);
 	} else {
-		err = ubi_read(c->ubi, branch->lnum, buf, branch->offs,
-			       c->nnode_sz);
+		err = ubifs_leb_read(c, branch->lnum, buf, branch->offs,
+				     c->nnode_sz, 1);
 		if (err)
 			return ERR_PTR(err);
 		err = ubifs_unpack_nnode(c, buf, nnode);
@@ -1883,8 +1886,8 @@ static struct ubifs_pnode *scan_get_pnode(struct ubifs_info *c,
 		ubifs_assert(branch->lnum >= c->lpt_first &&
 			     branch->lnum <= c->lpt_last);
 		ubifs_assert(branch->offs >= 0 && branch->offs < c->leb_size);
-		err = ubi_read(c->ubi, branch->lnum, buf, branch->offs,
-			       c->pnode_sz);
+		err = ubifs_leb_read(c, branch->lnum, buf, branch->offs,
+				     c->pnode_sz, 1);
 		if (err)
 			return ERR_PTR(err);
 		err = unpack_pnode(c, buf, pnode);
@@ -2224,7 +2227,7 @@ int dbg_check_lpt_nodes(struct ubifs_info *c, struct ubifs_cnode *cnode,
 	struct ubifs_cnode *cn;
 	int num, iip = 0, err;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS))
+	if (!dbg_is_chk_lprops(c))
 		return 0;
 
 	while (cnode) {
diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c
index dfcb5748a7d..cddd6bd214f 100644
--- a/fs/ubifs/lpt_commit.c
+++ b/fs/ubifs/lpt_commit.c
@@ -27,6 +27,7 @@
 
 #include <linux/crc16.h>
 #include <linux/slab.h>
+#include <linux/random.h>
 #include "ubifs.h"
 
 #ifdef CONFIG_UBIFS_FS_DEBUG
@@ -116,8 +117,8 @@ static int get_cnodes_to_commit(struct ubifs_info *c)
 		return 0;
 	cnt += 1;
 	while (1) {
-		ubifs_assert(!test_bit(COW_ZNODE, &cnode->flags));
-		__set_bit(COW_ZNODE, &cnode->flags);
+		ubifs_assert(!test_bit(COW_CNODE, &cnode->flags));
+		__set_bit(COW_CNODE, &cnode->flags);
 		cnext = next_dirty_cnode(cnode);
 		if (!cnext) {
 			cnode->cnext = c->lpt_cnext;
@@ -465,7 +466,7 @@ static int write_cnodes(struct ubifs_info *c)
 		 */
 		clear_bit(DIRTY_CNODE, &cnode->flags);
 		smp_mb__before_clear_bit();
-		clear_bit(COW_ZNODE, &cnode->flags);
+		clear_bit(COW_CNODE, &cnode->flags);
 		smp_mb__after_clear_bit();
 		offs += len;
 		dbg_chk_lpt_sz(c, 1, len);
@@ -1160,11 +1161,11 @@ static int lpt_gc_lnum(struct ubifs_info *c, int lnum)
 	void *buf = c->lpt_buf;
 
 	dbg_lp("LEB %d", lnum);
-	err = ubi_read(c->ubi, lnum, buf, 0, c->leb_size);
-	if (err) {
-		ubifs_err("cannot read LEB %d, error %d", lnum, err);
+
+	err = ubifs_leb_read(c, lnum, buf, 0, c->leb_size, 1);
+	if (err)
 		return err;
-	}
+
 	while (1) {
 		if (!is_a_node(c, buf, len)) {
 			int pad_len;
@@ -1640,7 +1641,7 @@ static int dbg_check_ltab_lnum(struct ubifs_info *c, int lnum)
 	int ret;
 	void *buf, *p;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS))
+	if (!dbg_is_chk_lprops(c))
 		return 0;
 
 	buf = p = __vmalloc(c->leb_size, GFP_NOFS, PAGE_KERNEL);
@@ -1650,11 +1651,11 @@ static int dbg_check_ltab_lnum(struct ubifs_info *c, int lnum)
 	}
 
 	dbg_lp("LEB %d", lnum);
-	err = ubi_read(c->ubi, lnum, buf, 0, c->leb_size);
-	if (err) {
-		dbg_msg("ubi_read failed, LEB %d, error %d", lnum, err);
+
+	err = ubifs_leb_read(c, lnum, buf, 0, c->leb_size, 1);
+	if (err)
 		goto out;
-	}
+
 	while (1) {
 		if (!is_a_node(c, p, len)) {
 			int i, pad_len;
@@ -1711,7 +1712,7 @@ int dbg_check_ltab(struct ubifs_info *c)
 {
 	int lnum, err, i, cnt;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS))
+	if (!dbg_is_chk_lprops(c))
 		return 0;
 
 	/* Bring the entire tree into memory */
@@ -1754,7 +1755,7 @@ int dbg_chk_lpt_free_spc(struct ubifs_info *c)
 	long long free = 0;
 	int i;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS))
+	if (!dbg_is_chk_lprops(c))
 		return 0;
 
 	for (i = 0; i < c->lpt_lebs; i++) {
@@ -1796,7 +1797,7 @@ int dbg_chk_lpt_sz(struct ubifs_info *c, int action, int len)
 	long long chk_lpt_sz, lpt_sz;
 	int err = 0;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS))
+	if (!dbg_is_chk_lprops(c))
 		return 0;
 
 	switch (action) {
@@ -1901,11 +1902,10 @@ static void dump_lpt_leb(const struct ubifs_info *c, int lnum)
 		return;
 	}
 
-	err = ubi_read(c->ubi, lnum, buf, 0, c->leb_size);
-	if (err) {
-		ubifs_err("cannot read LEB %d, error %d", lnum, err);
+	err = ubifs_leb_read(c, lnum, buf, 0, c->leb_size, 1);
+	if (err)
 		goto out;
-	}
+
 	while (1) {
 		offs = c->leb_size - len;
 		if (!is_a_node(c, p, len)) {
@@ -2019,7 +2019,7 @@ static int dbg_populate_lsave(struct ubifs_info *c)
 	struct ubifs_lpt_heap *heap;
 	int i;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 	if (random32() & 3)
 		return 0;
diff --git a/fs/ubifs/misc.h b/fs/ubifs/misc.h
index 0b5296a9a4c..ee7cb5ebb6e 100644
--- a/fs/ubifs/misc.h
+++ b/fs/ubifs/misc.h
@@ -39,6 +39,29 @@ static inline int ubifs_zn_dirty(const struct ubifs_znode *znode)
 }
 
 /**
+ * ubifs_zn_obsolete - check if znode is obsolete.
+ * @znode: znode to check
+ *
+ * This helper function returns %1 if @znode is obsolete and %0 otherwise.
+ */
+static inline int ubifs_zn_obsolete(const struct ubifs_znode *znode)
+{
+	return !!test_bit(OBSOLETE_ZNODE, &znode->flags);
+}
+
+/**
+ * ubifs_zn_cow - check if znode has to be copied on write.
+ * @znode: znode to check
+ *
+ * This helper function returns %1 if @znode is has COW flag set and %0
+ * otherwise.
+ */
+static inline int ubifs_zn_cow(const struct ubifs_znode *znode)
+{
+	return !!test_bit(COW_ZNODE, &znode->flags);
+}
+
+/**
  * ubifs_wake_up_bgt - wake up background thread.
  * @c: UBIFS file-system description object
  */
@@ -122,86 +145,6 @@ static inline int ubifs_wbuf_sync(struct ubifs_wbuf *wbuf)
 }
 
 /**
- * ubifs_leb_unmap - unmap an LEB.
- * @c: UBIFS file-system description object
- * @lnum: LEB number to unmap
- *
- * This function returns %0 on success and a negative error code on failure.
- */
-static inline int ubifs_leb_unmap(const struct ubifs_info *c, int lnum)
-{
-	int err;
-
-	ubifs_assert(!c->ro_media && !c->ro_mount);
-	if (c->ro_error)
-		return -EROFS;
-	err = ubi_leb_unmap(c->ubi, lnum);
-	if (err) {
-		ubifs_err("unmap LEB %d failed, error %d", lnum, err);
-		return err;
-	}
-
-	return 0;
-}
-
-/**
- * ubifs_leb_write - write to a LEB.
- * @c: UBIFS file-system description object
- * @lnum: LEB number to write
- * @buf: buffer to write from
- * @offs: offset within LEB to write to
- * @len: length to write
- * @dtype: data type
- *
- * This function returns %0 on success and a negative error code on failure.
- */
-static inline int ubifs_leb_write(const struct ubifs_info *c, int lnum,
-				  const void *buf, int offs, int len, int dtype)
-{
-	int err;
-
-	ubifs_assert(!c->ro_media && !c->ro_mount);
-	if (c->ro_error)
-		return -EROFS;
-	err = ubi_leb_write(c->ubi, lnum, buf, offs, len, dtype);
-	if (err) {
-		ubifs_err("writing %d bytes at %d:%d, error %d",
-			  len, lnum, offs, err);
-		return err;
-	}
-
-	return 0;
-}
-
-/**
- * ubifs_leb_change - atomic LEB change.
- * @c: UBIFS file-system description object
- * @lnum: LEB number to write
- * @buf: buffer to write from
- * @len: length to write
- * @dtype: data type
- *
- * This function returns %0 on success and a negative error code on failure.
- */
-static inline int ubifs_leb_change(const struct ubifs_info *c, int lnum,
-				   const void *buf, int len, int dtype)
-{
-	int err;
-
-	ubifs_assert(!c->ro_media && !c->ro_mount);
-	if (c->ro_error)
-		return -EROFS;
-	err = ubi_leb_change(c->ubi, lnum, buf, len, dtype);
-	if (err) {
-		ubifs_err("changing %d bytes in LEB %d, error %d",
-			  len, lnum, err);
-		return err;
-	}
-
-	return 0;
-}
-
-/**
  * ubifs_encode_dev - encode device node IDs.
  * @dev: UBIFS device node information
  * @rdev: device IDs to encode
diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c
index bd644bf587a..c542c73cfa3 100644
--- a/fs/ubifs/orphan.c
+++ b/fs/ubifs/orphan.c
@@ -674,7 +674,7 @@ static int kill_orphans(struct ubifs_info *c)
 		if (IS_ERR(sleb)) {
 			if (PTR_ERR(sleb) == -EUCLEAN)
 				sleb = ubifs_recover_leb(c, lnum, 0,
-							 c->sbuf, 0);
+							 c->sbuf, -1);
 			if (IS_ERR(sleb)) {
 				err = PTR_ERR(sleb);
 				break;
@@ -929,7 +929,7 @@ static int dbg_check_orphans(struct ubifs_info *c)
 	struct check_info ci;
 	int err;
 
-	if (!(ubifs_chk_flags & UBIFS_CHK_ORPH))
+	if (!dbg_is_chk_orph(c))
 		return 0;
 
 	ci.last_ino = 0;
diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c
index 731d9e2e7b5..af02790d932 100644
--- a/fs/ubifs/recovery.c
+++ b/fs/ubifs/recovery.c
@@ -117,7 +117,7 @@ static int get_master_node(const struct ubifs_info *c, int lnum, void **pbuf,
 	if (!sbuf)
 		return -ENOMEM;
 
-	err = ubi_read(c->ubi, lnum, sbuf, 0, c->leb_size);
+	err = ubifs_leb_read(c, lnum, sbuf, 0, c->leb_size, 0);
 	if (err && err != -EBADMSG)
 		goto out_free;
 
@@ -213,10 +213,10 @@ static int write_rcvrd_mst_node(struct ubifs_info *c,
 	mst->flags |= cpu_to_le32(UBIFS_MST_RCVRY);
 
 	ubifs_prepare_node(c, mst, UBIFS_MST_NODE_SZ, 1);
-	err = ubi_leb_change(c->ubi, lnum, mst, sz, UBI_SHORTTERM);
+	err = ubifs_leb_change(c, lnum, mst, sz, UBI_SHORTTERM);
 	if (err)
 		goto out;
-	err = ubi_leb_change(c->ubi, lnum + 1, mst, sz, UBI_SHORTTERM);
+	err = ubifs_leb_change(c, lnum + 1, mst, sz, UBI_SHORTTERM);
 	if (err)
 		goto out;
 out:
@@ -274,7 +274,8 @@ int ubifs_recover_master_node(struct ubifs_info *c)
 				if (cor1)
 					goto out_err;
 				mst = mst1;
-			} else if (offs1 == 0 && offs2 + sz >= c->leb_size) {
+			} else if (offs1 == 0 &&
+				   c->leb_size - offs2 - sz < sz) {
 				/* 1st LEB was unmapped and written, 2nd not */
 				if (cor1)
 					goto out_err;
@@ -539,8 +540,8 @@ static int fix_unclean_leb(struct ubifs_info *c, struct ubifs_scan_leb *sleb,
 			int len = ALIGN(endpt, c->min_io_size);
 
 			if (start) {
-				err = ubi_read(c->ubi, lnum, sleb->buf, 0,
-					       start);
+				err = ubifs_leb_read(c, lnum, sleb->buf, 0,
+						     start, 1);
 				if (err)
 					return err;
 			}
@@ -554,8 +555,8 @@ static int fix_unclean_leb(struct ubifs_info *c, struct ubifs_scan_leb *sleb,
 					ubifs_pad(c, buf, pad_len);
 				}
 			}
-			err = ubi_leb_change(c->ubi, lnum, sleb->buf, len,
-					     UBI_UNKNOWN);
+			err = ubifs_leb_change(c, lnum, sleb->buf, len,
+					       UBI_UNKNOWN);
 			if (err)
 				return err;
 		}
@@ -564,19 +565,15 @@ static int fix_unclean_leb(struct ubifs_info *c, struct ubifs_scan_leb *sleb,
 }
 
 /**
- * drop_last_node - drop the last node or group of nodes.
+ * drop_last_group - drop the last group of nodes.
  * @sleb: scanned LEB information
  * @offs: offset of dropped nodes is returned here
- * @grouped: non-zero if whole group of nodes have to be dropped
  *
  * This is a helper function for 'ubifs_recover_leb()' which drops the last
- * node of the scanned LEB or the last group of nodes if @grouped is not zero.
- * This function returns %1 if a node was dropped and %0 otherwise.
+ * group of nodes of the scanned LEB.
  */
-static int drop_last_node(struct ubifs_scan_leb *sleb, int *offs, int grouped)
+static void drop_last_group(struct ubifs_scan_leb *sleb, int *offs)
 {
-	int dropped = 0;
-
 	while (!list_empty(&sleb->nodes)) {
 		struct ubifs_scan_node *snod;
 		struct ubifs_ch *ch;
@@ -585,17 +582,40 @@ static int drop_last_node(struct ubifs_scan_leb *sleb, int *offs, int grouped)
 				  list);
 		ch = snod->node;
 		if (ch->group_type != UBIFS_IN_NODE_GROUP)
-			return dropped;
-		dbg_rcvry("dropping node at %d:%d", sleb->lnum, snod->offs);
+			break;
+
+		dbg_rcvry("dropping grouped node at %d:%d",
+			  sleb->lnum, snod->offs);
+		*offs = snod->offs;
+		list_del(&snod->list);
+		kfree(snod);
+		sleb->nodes_cnt -= 1;
+	}
+}
+
+/**
+ * drop_last_node - drop the last node.
+ * @sleb: scanned LEB information
+ * @offs: offset of dropped nodes is returned here
+ * @grouped: non-zero if whole group of nodes have to be dropped
+ *
+ * This is a helper function for 'ubifs_recover_leb()' which drops the last
+ * node of the scanned LEB.
+ */
+static void drop_last_node(struct ubifs_scan_leb *sleb, int *offs)
+{
+	struct ubifs_scan_node *snod;
+
+	if (!list_empty(&sleb->nodes)) {
+		snod = list_entry(sleb->nodes.prev, struct ubifs_scan_node,
+				  list);
+
+		dbg_rcvry("dropping last node at %d:%d", sleb->lnum, snod->offs);
 		*offs = snod->offs;
 		list_del(&snod->list);
 		kfree(snod);
 		sleb->nodes_cnt -= 1;
-		dropped = 1;
-		if (!grouped)
-			break;
 	}
-	return dropped;
 }
 
 /**
@@ -604,7 +624,8 @@ static int drop_last_node(struct ubifs_scan_leb *sleb, int *offs, int grouped)
  * @lnum: LEB number
  * @offs: offset
  * @sbuf: LEB-sized buffer to use
- * @grouped: nodes may be grouped for recovery
+ * @jhead: journal head number this LEB belongs to (%-1 if the LEB does not
+ *         belong to any journal head)
  *
  * This function does a scan of a LEB, but caters for errors that might have
  * been caused by the unclean unmount from which we are attempting to recover.
@@ -612,13 +633,14 @@ static int drop_last_node(struct ubifs_scan_leb *sleb, int *offs, int grouped)
  * found, and a negative error code in case of failure.
  */
 struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum,
-					 int offs, void *sbuf, int grouped)
+					 int offs, void *sbuf, int jhead)
 {
 	int ret = 0, err, len = c->leb_size - offs, start = offs, min_io_unit;
+	int grouped = jhead == -1 ? 0 : c->jheads[jhead].grouped;
 	struct ubifs_scan_leb *sleb;
 	void *buf = sbuf + offs;
 
-	dbg_rcvry("%d:%d", lnum, offs);
+	dbg_rcvry("%d:%d, jhead %d, grouped %d", lnum, offs, jhead, grouped);
 
 	sleb = ubifs_start_scan(c, lnum, offs, sbuf);
 	if (IS_ERR(sleb))
@@ -635,7 +657,7 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum,
 		 * Scan quietly until there is an error from which we cannot
 		 * recover
 		 */
-		ret = ubifs_scan_a_node(c, buf, len, lnum, offs, 0);
+		ret = ubifs_scan_a_node(c, buf, len, lnum, offs, 1);
 		if (ret == SCANNED_A_NODE) {
 			/* A valid node, and not a padding node */
 			struct ubifs_ch *ch = buf;
@@ -695,59 +717,62 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum,
 		 * If nodes are grouped, always drop the incomplete group at
 		 * the end.
 		 */
-		drop_last_node(sleb, &offs, 1);
+		drop_last_group(sleb, &offs);
 
-	/*
-	 * While we are in the middle of the same min. I/O unit keep dropping
-	 * nodes. So basically, what we want is to make sure that the last min.
-	 * I/O unit where we saw the corruption is dropped completely with all
-	 * the uncorrupted node which may possibly sit there.
-	 *
-	 * In other words, let's name the min. I/O unit where the corruption
-	 * starts B, and the previous min. I/O unit A. The below code tries to
-	 * deal with a situation when half of B contains valid nodes or the end
-	 * of a valid node, and the second half of B contains corrupted data or
-	 * garbage. This means that UBIFS had been writing to B just before the
-	 * power cut happened. I do not know how realistic is this scenario
-	 * that half of the min. I/O unit had been written successfully and the
-	 * other half not, but this is possible in our 'failure mode emulation'
-	 * infrastructure at least.
-	 *
-	 * So what is the problem, why we need to drop those nodes? Whey can't
-	 * we just clean-up the second half of B by putting a padding node
-	 * there? We can, and this works fine with one exception which was
-	 * reproduced with power cut emulation testing and happens extremely
-	 * rarely. The description follows, but it is worth noting that that is
-	 * only about the GC head, so we could do this trick only if the bud
-	 * belongs to the GC head, but it does not seem to be worth an
-	 * additional "if" statement.
-	 *
-	 * So, imagine the file-system is full, we run GC which is moving valid
-	 * nodes from LEB X to LEB Y (obviously, LEB Y is the current GC head
-	 * LEB). The @c->gc_lnum is -1, which means that GC will retain LEB X
-	 * and will try to continue. Imagine that LEB X is currently the
-	 * dirtiest LEB, and the amount of used space in LEB Y is exactly the
-	 * same as amount of free space in LEB X.
-	 *
-	 * And a power cut happens when nodes are moved from LEB X to LEB Y. We
-	 * are here trying to recover LEB Y which is the GC head LEB. We find
-	 * the min. I/O unit B as described above. Then we clean-up LEB Y by
-	 * padding min. I/O unit. And later 'ubifs_rcvry_gc_commit()' function
-	 * fails, because it cannot find a dirty LEB which could be GC'd into
-	 * LEB Y! Even LEB X does not match because the amount of valid nodes
-	 * there does not fit the free space in LEB Y any more! And this is
-	 * because of the padding node which we added to LEB Y. The
-	 * user-visible effect of this which I once observed and analysed is
-	 * that we cannot mount the file-system with -ENOSPC error.
-	 *
-	 * So obviously, to make sure that situation does not happen we should
-	 * free min. I/O unit B in LEB Y completely and the last used min. I/O
-	 * unit in LEB Y should be A. This is basically what the below code
-	 * tries to do.
-	 */
-	while (min_io_unit == round_down(offs, c->min_io_size) &&
-	       min_io_unit != offs &&
-	       drop_last_node(sleb, &offs, grouped));
+	if (jhead == GCHD) {
+		/*
+		 * If this LEB belongs to the GC head then while we are in the
+		 * middle of the same min. I/O unit keep dropping nodes. So
+		 * basically, what we want is to make sure that the last min.
+		 * I/O unit where we saw the corruption is dropped completely
+		 * with all the uncorrupted nodes which may possibly sit there.
+		 *
+		 * In other words, let's name the min. I/O unit where the
+		 * corruption starts B, and the previous min. I/O unit A. The
+		 * below code tries to deal with a situation when half of B
+		 * contains valid nodes or the end of a valid node, and the
+		 * second half of B contains corrupted data or garbage. This
+		 * means that UBIFS had been writing to B just before the power
+		 * cut happened. I do not know how realistic is this scenario
+		 * that half of the min. I/O unit had been written successfully
+		 * and the other half not, but this is possible in our 'failure
+		 * mode emulation' infrastructure at least.
+		 *
+		 * So what is the problem, why we need to drop those nodes? Why
+		 * can't we just clean-up the second half of B by putting a
+		 * padding node there? We can, and this works fine with one
+		 * exception which was reproduced with power cut emulation
+		 * testing and happens extremely rarely.
+		 *
+		 * Imagine the file-system is full, we run GC which starts
+		 * moving valid nodes from LEB X to LEB Y (obviously, LEB Y is
+		 * the current GC head LEB). The @c->gc_lnum is -1, which means
+		 * that GC will retain LEB X and will try to continue. Imagine
+		 * that LEB X is currently the dirtiest LEB, and the amount of
+		 * used space in LEB Y is exactly the same as amount of free
+		 * space in LEB X.
+		 *
+		 * And a power cut happens when nodes are moved from LEB X to
+		 * LEB Y. We are here trying to recover LEB Y which is the GC
+		 * head LEB. We find the min. I/O unit B as described above.
+		 * Then we clean-up LEB Y by padding min. I/O unit. And later
+		 * 'ubifs_rcvry_gc_commit()' function fails, because it cannot
+		 * find a dirty LEB which could be GC'd into LEB Y! Even LEB X
+		 * does not match because the amount of valid nodes there does
+		 * not fit the free space in LEB Y any more! And this is
+		 * because of the padding node which we added to LEB Y. The
+		 * user-visible effect of this which I once observed and
+		 * analysed is that we cannot mount the file-system with
+		 * -ENOSPC error.
+		 *
+		 * So obviously, to make sure that situation does not happen we
+		 * should free min. I/O unit B in LEB Y completely and the last
+		 * used min. I/O unit in LEB Y should be A. This is basically
+		 * what the below code tries to do.
+		 */
+		while (offs > min_io_unit)
+			drop_last_node(sleb, &offs);
+	}
 
 	buf = sbuf + offs;
 	len = c->leb_size - offs;
@@ -795,7 +820,8 @@ static int get_cs_sqnum(struct ubifs_info *c, int lnum, int offs,
 		return -ENOMEM;
 	if (c->leb_size - offs < UBIFS_CS_NODE_SZ)
 		goto out_err;
-	err = ubi_read(c->ubi, lnum, (void *)cs_node, offs, UBIFS_CS_NODE_SZ);
+	err = ubifs_leb_read(c, lnum, (void *)cs_node, offs,
+			     UBIFS_CS_NODE_SZ, 0);
 	if (err && err != -EBADMSG)
 		goto out_free;
 	ret = ubifs_scan_a_node(c, cs_node, UBIFS_CS_NODE_SZ, lnum, offs, 0);
@@ -881,7 +907,7 @@ struct ubifs_scan_leb *ubifs_recover_log_leb(struct ubifs_info *c, int lnum,
 		}
 		ubifs_scan_destroy(sleb);
 	}
-	return ubifs_recover_leb(c, lnum, offs, sbuf, 0);
+	return ubifs_recover_leb(c, lnum, offs, sbuf, -1);
 }
 
 /**
@@ -895,8 +921,7 @@ struct ubifs_scan_leb *ubifs_recover_log_leb(struct ubifs_info *c, int lnum,
  *
  * This function returns %0 on success and a negative error code on failure.
  */
-static int recover_head(const struct ubifs_info *c, int lnum, int offs,
-			void *sbuf)
+static int recover_head(struct ubifs_info *c, int lnum, int offs, void *sbuf)
 {
 	int len = c->max_write_size, err;
 
@@ -907,15 +932,15 @@ static int recover_head(const struct ubifs_info *c, int lnum, int offs,
 		return 0;
 
 	/* Read at the head location and check it is empty flash */
-	err = ubi_read(c->ubi, lnum, sbuf, offs, len);
+	err = ubifs_leb_read(c, lnum, sbuf, offs, len, 1);
 	if (err || !is_empty(sbuf, len)) {
 		dbg_rcvry("cleaning head at %d:%d", lnum, offs);
 		if (offs == 0)
 			return ubifs_leb_unmap(c, lnum);
-		err = ubi_read(c->ubi, lnum, sbuf, 0, offs);
+		err = ubifs_leb_read(c, lnum, sbuf, 0, offs, 1);
 		if (err)
 			return err;
-		return ubi_leb_change(c->ubi, lnum, sbuf, offs, UBI_UNKNOWN);
+		return ubifs_leb_change(c, lnum, sbuf, offs, UBI_UNKNOWN);
 	}
 
 	return 0;
@@ -938,7 +963,7 @@ static int recover_head(const struct ubifs_info *c, int lnum, int offs,
  *
  * This function returns %0 on success and a negative error code on failure.
  */
-int ubifs_recover_inl_heads(const struct ubifs_info *c, void *sbuf)
+int ubifs_recover_inl_heads(struct ubifs_info *c, void *sbuf)
 {
 	int err;
 
@@ -969,7 +994,7 @@ int ubifs_recover_inl_heads(const struct ubifs_info *c, void *sbuf)
  *
  * This function returns %0 on success and a negative error code on failure.
  */
-static int clean_an_unclean_leb(const struct ubifs_info *c,
+static int clean_an_unclean_leb(struct ubifs_info *c,
 				struct ubifs_unclean_leb *ucleb, void *sbuf)
 {
 	int err, lnum = ucleb->lnum, offs = 0, len = ucleb->endpt, quiet = 1;
@@ -985,7 +1010,7 @@ static int clean_an_unclean_leb(const struct ubifs_info *c,
 		return 0;
 	}
 
-	err = ubi_read(c->ubi, lnum, buf, offs, len);
+	err = ubifs_leb_read(c, lnum, buf, offs, len, 0);
 	if (err && err != -EBADMSG)
 		return err;
 
@@ -1045,7 +1070,7 @@ static int clean_an_unclean_leb(const struct ubifs_info *c,
 	}
 
 	/* Write back the LEB atomically */
-	err = ubi_leb_change(c->ubi, lnum, sbuf, len, UBI_UNKNOWN);
+	err = ubifs_leb_change(c, lnum, sbuf, len, UBI_UNKNOWN);
 	if (err)
 		return err;
 
@@ -1065,7 +1090,7 @@ static int clean_an_unclean_leb(const struct ubifs_info *c,
  *
  * This function returns %0 on success and a negative error code on failure.
  */
-int ubifs_clean_lebs(const struct ubifs_info *c, void *sbuf)
+int ubifs_clean_lebs(struct ubifs_info *c, void *sbuf)
 {
 	dbg_rcvry("recovery");
 	while (!list_empty(&c->unclean_leb_list)) {
@@ -1430,7 +1455,7 @@ static int fix_size_in_place(struct ubifs_info *c, struct size_entry *e)
 	if (i_size >= e->d_size)
 		return 0;
 	/* Read the LEB */
-	err = ubi_read(c->ubi, lnum, c->sbuf, 0, c->leb_size);
+	err = ubifs_leb_read(c, lnum, c->sbuf, 0, c->leb_size, 1);
 	if (err)
 		goto out;
 	/* Change the size field and recalculate the CRC */
@@ -1446,7 +1471,7 @@ static int fix_size_in_place(struct ubifs_info *c, struct size_entry *e)
 		len -= 1;
 	len = ALIGN(len + 1, c->min_io_size);
 	/* Atomically write the fixed LEB back again */
-	err = ubi_leb_change(c->ubi, lnum, c->sbuf, len, UBI_UNKNOWN);
+	err = ubifs_leb_change(c, lnum, c->sbuf, len, UBI_UNKNOWN);
 	if (err)
 		goto out;
 	dbg_rcvry("inode %lu at %d:%d size %lld -> %lld",
diff --git a/fs/ubifs/replay.c b/fs/ubifs/replay.c
index 6617280d167..ccabaf1164b 100644
--- a/fs/ubifs/replay.c
+++ b/fs/ubifs/replay.c
@@ -523,8 +523,7 @@ static int is_last_bud(struct ubifs_info *c, struct ubifs_bud *bud)
 	if (!list_is_last(&next->list, &jh->buds_list))
 		return 0;
 
-	err = ubi_read(c->ubi, next->lnum, (char *)&data,
-		       next->start, 4);
+	err = ubifs_leb_read(c, next->lnum, (char *)&data, next->start, 4, 1);
 	if (err)
 		return 0;
 
@@ -557,8 +556,7 @@ static int replay_bud(struct ubifs_info *c, struct bud_entry *b)
 		 * these LEBs could possibly be written to at the power cut
 		 * time.
 		 */
-		sleb = ubifs_recover_leb(c, lnum, offs, c->sbuf,
-					 b->bud->jhead != GCHD);
+		sleb = ubifs_recover_leb(c, lnum, offs, c->sbuf, b->bud->jhead);
 	else
 		sleb = ubifs_scan(c, lnum, offs, c->sbuf, 0);
 	if (IS_ERR(sleb))
diff --git a/fs/ubifs/sb.c b/fs/ubifs/sb.c
index c606f010e8d..93d938ad3d2 100644
--- a/fs/ubifs/sb.c
+++ b/fs/ubifs/sb.c
@@ -674,15 +674,15 @@ static int fixup_leb(struct ubifs_info *c, int lnum, int len)
 
 	if (len == 0) {
 		dbg_mnt("unmap empty LEB %d", lnum);
-		return ubi_leb_unmap(c->ubi, lnum);
+		return ubifs_leb_unmap(c, lnum);
 	}
 
 	dbg_mnt("fixup LEB %d, data len %d", lnum, len);
-	err = ubi_read(c->ubi, lnum, c->sbuf, 0, len);
+	err = ubifs_leb_read(c, lnum, c->sbuf, 0, len, 1);
 	if (err)
 		return err;
 
-	return ubi_leb_change(c->ubi, lnum, c->sbuf, len, UBI_UNKNOWN);
+	return ubifs_leb_change(c, lnum, c->sbuf, len, UBI_UNKNOWN);
 }
 
 /**
diff --git a/fs/ubifs/scan.c b/fs/ubifs/scan.c
index 36216b46f77..37383e8011b 100644
--- a/fs/ubifs/scan.c
+++ b/fs/ubifs/scan.c
@@ -148,7 +148,7 @@ struct ubifs_scan_leb *ubifs_start_scan(const struct ubifs_info *c, int lnum,
 	INIT_LIST_HEAD(&sleb->nodes);
 	sleb->buf = sbuf;
 
-	err = ubi_read(c->ubi, lnum, sbuf + offs, offs, c->leb_size - offs);
+	err = ubifs_leb_read(c, lnum, sbuf + offs, offs, c->leb_size - offs, 0);
 	if (err && err != -EBADMSG) {
 		ubifs_err("cannot read %d bytes from LEB %d:%d,"
 			  " error %d", c->leb_size - offs, lnum, offs, err);
@@ -240,7 +240,7 @@ void ubifs_scanned_corruption(const struct ubifs_info *c, int lnum, int offs,
 	int len;
 
 	ubifs_err("corruption at LEB %d:%d", lnum, offs);
-	if (dbg_failure_mode)
+	if (dbg_is_tst_rcvry(c))
 		return;
 	len = c->leb_size - offs;
 	if (len > 8192)
diff --git a/fs/ubifs/shrinker.c b/fs/ubifs/shrinker.c
index ca953a94502..9e1d05666fe 100644
--- a/fs/ubifs/shrinker.c
+++ b/fs/ubifs/shrinker.c
@@ -284,7 +284,11 @@ int ubifs_shrinker(struct shrinker *shrink, struct shrink_control *sc)
 	long clean_zn_cnt = atomic_long_read(&ubifs_clean_zn_cnt);
 
 	if (nr == 0)
-		return clean_zn_cnt;
+		/*
+		 * Due to the way UBIFS updates the clean znode counter it may
+		 * temporarily be negative.
+		 */
+		return clean_zn_cnt >= 0 ? clean_zn_cnt : 1;
 
 	if (!clean_zn_cnt) {
 		/*
diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
index 1ab0d22e4c9..b28121278d4 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -85,7 +85,7 @@ static int validate_inode(struct ubifs_info *c, const struct inode *inode)
 	if (ui->data_len < 0 || ui->data_len > UBIFS_MAX_INO_DATA)
 		return 4;
 
-	if (ui->xattr && (inode->i_mode & S_IFMT) != S_IFREG)
+	if (ui->xattr && !S_ISREG(inode->i_mode))
 		return 5;
 
 	if (!ubifs_compr_present(ui->compr_type)) {
@@ -94,7 +94,7 @@ static int validate_inode(struct ubifs_info *c, const struct inode *inode)
 			   ubifs_compr_name(ui->compr_type));
 	}
 
-	err = dbg_check_dir_size(c, inode);
+	err = dbg_check_dir(c, inode);
 	return err;
 }
 
@@ -811,15 +811,18 @@ static int alloc_wbufs(struct ubifs_info *c)
 
 		c->jheads[i].wbuf.sync_callback = &bud_wbuf_callback;
 		c->jheads[i].wbuf.jhead = i;
+		c->jheads[i].grouped = 1;
 	}
 
 	c->jheads[BASEHD].wbuf.dtype = UBI_SHORTTERM;
 	/*
 	 * Garbage Collector head likely contains long-term data and
-	 * does not need to be synchronized by timer.
+	 * does not need to be synchronized by timer. Also GC head nodes are
+	 * not grouped.
 	 */
 	c->jheads[GCHD].wbuf.dtype = UBI_LONGTERM;
 	c->jheads[GCHD].wbuf.no_timer = 1;
+	c->jheads[GCHD].grouped = 0;
 
 	return 0;
 }
@@ -911,7 +914,7 @@ static int check_volume_empty(struct ubifs_info *c)
 
 	c->empty = 1;
 	for (lnum = 0; lnum < c->leb_cnt; lnum++) {
-		err = ubi_is_mapped(c->ubi, lnum);
+		err = ubifs_is_mapped(c, lnum);
 		if (unlikely(err < 0))
 			return err;
 		if (err == 1) {
@@ -1284,12 +1287,25 @@ static int mount_ubifs(struct ubifs_info *c)
 	if ((c->mst_node->flags & cpu_to_le32(UBIFS_MST_DIRTY)) != 0) {
 		ubifs_msg("recovery needed");
 		c->need_recovery = 1;
-		if (!c->ro_mount) {
-			err = ubifs_recover_inl_heads(c, c->sbuf);
-			if (err)
-				goto out_master;
-		}
-	} else if (!c->ro_mount) {
+	}
+
+	if (c->need_recovery && !c->ro_mount) {
+		err = ubifs_recover_inl_heads(c, c->sbuf);
+		if (err)
+			goto out_master;
+	}
+
+	err = ubifs_lpt_init(c, 1, !c->ro_mount);
+	if (err)
+		goto out_master;
+
+	if (!c->ro_mount && c->space_fixup) {
+		err = ubifs_fixup_free_space(c);
+		if (err)
+			goto out_master;
+	}
+
+	if (!c->ro_mount) {
 		/*
 		 * Set the "dirty" flag so that if we reboot uncleanly we
 		 * will notice this immediately on the next mount.
@@ -1297,13 +1313,9 @@ static int mount_ubifs(struct ubifs_info *c)
 		c->mst_node->flags |= cpu_to_le32(UBIFS_MST_DIRTY);
 		err = ubifs_write_master(c);
 		if (err)
-			goto out_master;
+			goto out_lpt;
 	}
 
-	err = ubifs_lpt_init(c, 1, !c->ro_mount);
-	if (err)
-		goto out_lpt;
-
 	err = dbg_check_idx_size(c, c->bi.old_idx_sz);
 	if (err)
 		goto out_lpt;
@@ -1396,12 +1408,6 @@ static int mount_ubifs(struct ubifs_info *c)
 	} else
 		ubifs_assert(c->lst.taken_empty_lebs > 0);
 
-	if (!c->ro_mount && c->space_fixup) {
-		err = ubifs_fixup_free_space(c);
-		if (err)
-			goto out_infos;
-	}
-
 	err = dbg_check_filesystem(c);
 	if (err)
 		goto out_infos;
@@ -1842,7 +1848,6 @@ static void ubifs_put_super(struct super_block *sb)
 	bdi_destroy(&c->bdi);
 	ubi_close_volume(c->ubi);
 	mutex_unlock(&c->umount_mutex);
-	kfree(c);
 }
 
 static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data)
@@ -1965,61 +1970,65 @@ static struct ubi_volume_desc *open_ubi(const char *name, int mode)
 	return ERR_PTR(-EINVAL);
 }
 
-static int ubifs_fill_super(struct super_block *sb, void *data, int silent)
+static struct ubifs_info *alloc_ubifs_info(struct ubi_volume_desc *ubi)
 {
-	struct ubi_volume_desc *ubi = sb->s_fs_info;
 	struct ubifs_info *c;
-	struct inode *root;
-	int err;
 
 	c = kzalloc(sizeof(struct ubifs_info), GFP_KERNEL);
-	if (!c)
-		return -ENOMEM;
+	if (c) {
+		spin_lock_init(&c->cnt_lock);
+		spin_lock_init(&c->cs_lock);
+		spin_lock_init(&c->buds_lock);
+		spin_lock_init(&c->space_lock);
+		spin_lock_init(&c->orphan_lock);
+		init_rwsem(&c->commit_sem);
+		mutex_init(&c->lp_mutex);
+		mutex_init(&c->tnc_mutex);
+		mutex_init(&c->log_mutex);
+		mutex_init(&c->mst_mutex);
+		mutex_init(&c->umount_mutex);
+		mutex_init(&c->bu_mutex);
+		mutex_init(&c->write_reserve_mutex);
+		init_waitqueue_head(&c->cmt_wq);
+		c->buds = RB_ROOT;
+		c->old_idx = RB_ROOT;
+		c->size_tree = RB_ROOT;
+		c->orph_tree = RB_ROOT;
+		INIT_LIST_HEAD(&c->infos_list);
+		INIT_LIST_HEAD(&c->idx_gc);
+		INIT_LIST_HEAD(&c->replay_list);
+		INIT_LIST_HEAD(&c->replay_buds);
+		INIT_LIST_HEAD(&c->uncat_list);
+		INIT_LIST_HEAD(&c->empty_list);
+		INIT_LIST_HEAD(&c->freeable_list);
+		INIT_LIST_HEAD(&c->frdi_idx_list);
+		INIT_LIST_HEAD(&c->unclean_leb_list);
+		INIT_LIST_HEAD(&c->old_buds);
+		INIT_LIST_HEAD(&c->orph_list);
+		INIT_LIST_HEAD(&c->orph_new);
+		c->no_chk_data_crc = 1;
+
+		c->highest_inum = UBIFS_FIRST_INO;
+		c->lhead_lnum = c->ltail_lnum = UBIFS_LOG_LNUM;
+
+		ubi_get_volume_info(ubi, &c->vi);
+		ubi_get_device_info(c->vi.ubi_num, &c->di);
+	}
+	return c;
+}
 
-	spin_lock_init(&c->cnt_lock);
-	spin_lock_init(&c->cs_lock);
-	spin_lock_init(&c->buds_lock);
-	spin_lock_init(&c->space_lock);
-	spin_lock_init(&c->orphan_lock);
-	init_rwsem(&c->commit_sem);
-	mutex_init(&c->lp_mutex);
-	mutex_init(&c->tnc_mutex);
-	mutex_init(&c->log_mutex);
-	mutex_init(&c->mst_mutex);
-	mutex_init(&c->umount_mutex);
-	mutex_init(&c->bu_mutex);
-	mutex_init(&c->write_reserve_mutex);
-	init_waitqueue_head(&c->cmt_wq);
-	c->buds = RB_ROOT;
-	c->old_idx = RB_ROOT;
-	c->size_tree = RB_ROOT;
-	c->orph_tree = RB_ROOT;
-	INIT_LIST_HEAD(&c->infos_list);
-	INIT_LIST_HEAD(&c->idx_gc);
-	INIT_LIST_HEAD(&c->replay_list);
-	INIT_LIST_HEAD(&c->replay_buds);
-	INIT_LIST_HEAD(&c->uncat_list);
-	INIT_LIST_HEAD(&c->empty_list);
-	INIT_LIST_HEAD(&c->freeable_list);
-	INIT_LIST_HEAD(&c->frdi_idx_list);
-	INIT_LIST_HEAD(&c->unclean_leb_list);
-	INIT_LIST_HEAD(&c->old_buds);
-	INIT_LIST_HEAD(&c->orph_list);
-	INIT_LIST_HEAD(&c->orph_new);
-	c->no_chk_data_crc = 1;
+static int ubifs_fill_super(struct super_block *sb, void *data, int silent)
+{
+	struct ubifs_info *c = sb->s_fs_info;
+	struct inode *root;
+	int err;
 
 	c->vfs_sb = sb;
-	c->highest_inum = UBIFS_FIRST_INO;
-	c->lhead_lnum = c->ltail_lnum = UBIFS_LOG_LNUM;
-
-	ubi_get_volume_info(ubi, &c->vi);
-	ubi_get_device_info(c->vi.ubi_num, &c->di);
-
 	/* Re-open the UBI device in read-write mode */
 	c->ubi = ubi_open_volume(c->vi.ubi_num, c->vi.vol_id, UBI_READWRITE);
 	if (IS_ERR(c->ubi)) {
 		err = PTR_ERR(c->ubi);
-		goto out_free;
+		goto out;
 	}
 
 	/*
@@ -2085,24 +2094,29 @@ out_bdi:
 	bdi_destroy(&c->bdi);
 out_close:
 	ubi_close_volume(c->ubi);
-out_free:
-	kfree(c);
+out:
 	return err;
 }
 
 static int sb_test(struct super_block *sb, void *data)
 {
-	dev_t *dev = data;
+	struct ubifs_info *c1 = data;
 	struct ubifs_info *c = sb->s_fs_info;
 
-	return c->vi.cdev == *dev;
+	return c->vi.cdev == c1->vi.cdev;
+}
+
+static int sb_set(struct super_block *sb, void *data)
+{
+	sb->s_fs_info = data;
+	return set_anon_super(sb, NULL);
 }
 
 static struct dentry *ubifs_mount(struct file_system_type *fs_type, int flags,
 			const char *name, void *data)
 {
 	struct ubi_volume_desc *ubi;
-	struct ubi_volume_info vi;
+	struct ubifs_info *c;
 	struct super_block *sb;
 	int err;
 
@@ -2119,19 +2133,25 @@ static struct dentry *ubifs_mount(struct file_system_type *fs_type, int flags,
 			name, (int)PTR_ERR(ubi));
 		return ERR_CAST(ubi);
 	}
-	ubi_get_volume_info(ubi, &vi);
 
-	dbg_gen("opened ubi%d_%d", vi.ubi_num, vi.vol_id);
+	c = alloc_ubifs_info(ubi);
+	if (!c) {
+		err = -ENOMEM;
+		goto out_close;
+	}
+
+	dbg_gen("opened ubi%d_%d", c->vi.ubi_num, c->vi.vol_id);
 
-	sb = sget(fs_type, &sb_test, &set_anon_super, &vi.cdev);
+	sb = sget(fs_type, sb_test, sb_set, c);
 	if (IS_ERR(sb)) {
 		err = PTR_ERR(sb);
+		kfree(c);
 		goto out_close;
 	}
 
 	if (sb->s_root) {
 		struct ubifs_info *c1 = sb->s_fs_info;
-
+		kfree(c);
 		/* A new mount point for already mounted UBIFS */
 		dbg_gen("this ubi volume is already mounted");
 		if (!!(flags & MS_RDONLY) != c1->ro_mount) {
@@ -2140,11 +2160,6 @@ static struct dentry *ubifs_mount(struct file_system_type *fs_type, int flags,
 		}
 	} else {
 		sb->s_flags = flags;
-		/*
-		 * Pass 'ubi' to 'fill_super()' in sb->s_fs_info where it is
-		 * replaced by 'c'.
-		 */
-		sb->s_fs_info = ubi;
 		err = ubifs_fill_super(sb, data, flags & MS_SILENT ? 1 : 0);
 		if (err)
 			goto out_deact;
@@ -2164,11 +2179,18 @@ out_close:
 	return ERR_PTR(err);
 }
 
+static void kill_ubifs_super(struct super_block *s)
+{
+	struct ubifs_info *c = s->s_fs_info;
+	kill_anon_super(s);
+	kfree(c);
+}
+
 static struct file_system_type ubifs_fs_type = {
 	.name    = "ubifs",
 	.owner   = THIS_MODULE,
 	.mount   = ubifs_mount,
-	.kill_sb = kill_anon_super,
+	.kill_sb = kill_ubifs_super,
 };
 
 /*
diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c
index 8119b1fd8d9..06673864768 100644
--- a/fs/ubifs/tnc.c
+++ b/fs/ubifs/tnc.c
@@ -223,7 +223,7 @@ static struct ubifs_znode *copy_znode(struct ubifs_info *c,
 	__set_bit(DIRTY_ZNODE, &zn->flags);
 	__clear_bit(COW_ZNODE, &zn->flags);
 
-	ubifs_assert(!test_bit(OBSOLETE_ZNODE, &znode->flags));
+	ubifs_assert(!ubifs_zn_obsolete(znode));
 	__set_bit(OBSOLETE_ZNODE, &znode->flags);
 
 	if (znode->level != 0) {
@@ -271,7 +271,7 @@ static struct ubifs_znode *dirty_cow_znode(struct ubifs_info *c,
 	struct ubifs_znode *zn;
 	int err;
 
-	if (!test_bit(COW_ZNODE, &znode->flags)) {
+	if (!ubifs_zn_cow(znode)) {
 		/* znode is not being committed */
 		if (!test_and_set_bit(DIRTY_ZNODE, &znode->flags)) {
 			atomic_long_inc(&c->dirty_zn_cnt);
@@ -462,7 +462,7 @@ static int try_read_node(const struct ubifs_info *c, void *buf, int type,
 
 	dbg_io("LEB %d:%d, %s, length %d", lnum, offs, dbg_ntype(type), len);
 
-	err = ubi_read(c->ubi, lnum, buf, offs, len);
+	err = ubifs_leb_read(c, lnum, buf, offs, len, 1);
 	if (err) {
 		ubifs_err("cannot read node type %d from LEB %d:%d, error %d",
 			  type, lnum, offs, err);
@@ -1666,7 +1666,7 @@ static int read_wbuf(struct ubifs_wbuf *wbuf, void *buf, int len, int lnum,
 	if (!overlap) {
 		/* We may safely unlock the write-buffer and read the data */
 		spin_unlock(&wbuf->lock);
-		return ubi_read(c->ubi, lnum, buf, offs, len);
+		return ubifs_leb_read(c, lnum, buf, offs, len, 0);
 	}
 
 	/* Don't read under wbuf */
@@ -1680,7 +1680,7 @@ static int read_wbuf(struct ubifs_wbuf *wbuf, void *buf, int len, int lnum,
 
 	if (rlen > 0)
 		/* Read everything that goes before write-buffer */
-		return ubi_read(c->ubi, lnum, buf, offs, rlen);
+		return ubifs_leb_read(c, lnum, buf, offs, rlen, 0);
 
 	return 0;
 }
@@ -1767,7 +1767,7 @@ int ubifs_tnc_bulk_read(struct ubifs_info *c, struct bu_info *bu)
 	if (wbuf)
 		err = read_wbuf(wbuf, bu->buf, len, lnum, offs);
 	else
-		err = ubi_read(c->ubi, lnum, bu->buf, offs, len);
+		err = ubifs_leb_read(c, lnum, bu->buf, offs, len, 0);
 
 	/* Check for a race with GC */
 	if (maybe_leb_gced(c, lnum, bu->gc_seq))
@@ -2423,7 +2423,7 @@ static int tnc_delete(struct ubifs_info *c, struct ubifs_znode *znode, int n)
 	 */
 
 	do {
-		ubifs_assert(!test_bit(OBSOLETE_ZNODE, &znode->flags));
+		ubifs_assert(!ubifs_zn_obsolete(znode));
 		ubifs_assert(ubifs_zn_dirty(znode));
 
 		zp = znode->parent;
@@ -2479,9 +2479,8 @@ static int tnc_delete(struct ubifs_info *c, struct ubifs_znode *znode, int n)
 			c->zroot.offs = zbr->offs;
 			c->zroot.len = zbr->len;
 			c->zroot.znode = znode;
-			ubifs_assert(!test_bit(OBSOLETE_ZNODE,
-				     &zp->flags));
-			ubifs_assert(test_bit(DIRTY_ZNODE, &zp->flags));
+			ubifs_assert(!ubifs_zn_obsolete(zp));
+			ubifs_assert(ubifs_zn_dirty(zp));
 			atomic_long_dec(&c->dirty_zn_cnt);
 
 			if (zp->cnext) {
@@ -2865,7 +2864,7 @@ static void tnc_destroy_cnext(struct ubifs_info *c)
 		struct ubifs_znode *znode = cnext;
 
 		cnext = cnext->cnext;
-		if (test_bit(OBSOLETE_ZNODE, &znode->flags))
+		if (ubifs_zn_obsolete(znode))
 			kfree(znode);
 	} while (cnext && cnext != c->cnext);
 }
@@ -2876,12 +2875,13 @@ static void tnc_destroy_cnext(struct ubifs_info *c)
  */
 void ubifs_tnc_close(struct ubifs_info *c)
 {
-	long clean_freed;
-
 	tnc_destroy_cnext(c);
 	if (c->zroot.znode) {
-		clean_freed = ubifs_destroy_tnc_subtree(c->zroot.znode);
-		atomic_long_sub(clean_freed, &ubifs_clean_zn_cnt);
+		long n;
+
+		ubifs_destroy_tnc_subtree(c->zroot.znode);
+		n = atomic_long_read(&c->clean_zn_cnt);
+		atomic_long_sub(n, &ubifs_clean_zn_cnt);
 	}
 	kfree(c->gap_lebs);
 	kfree(c->ilebs);
@@ -3300,7 +3300,7 @@ int dbg_check_inode_size(struct ubifs_info *c, const struct inode *inode,
 
 	if (!S_ISREG(inode->i_mode))
 		return 0;
-	if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+	if (!dbg_is_chk_gen(c))
 		return 0;
 
 	block = (size + UBIFS_BLOCK_SIZE - 1) >> UBIFS_BLOCK_SHIFT;
@@ -3336,9 +3336,10 @@ out_dump:
 	ubifs_err("inode %lu has size %lld, but there are data at offset %lld "
 		  "(data key %s)", (unsigned long)inode->i_ino, size,
 		  ((loff_t)block) << UBIFS_BLOCK_SHIFT, DBGKEY(key));
+	mutex_unlock(&c->tnc_mutex);
 	dbg_dump_inode(c, inode);
 	dbg_dump_stack();
-	err = -EINVAL;
+	return -EINVAL;
 
 out_unlock:
 	mutex_unlock(&c->tnc_mutex);
diff --git a/fs/ubifs/tnc_commit.c b/fs/ubifs/tnc_commit.c
index 41920f357bb..4c15f07a8bb 100644
--- a/fs/ubifs/tnc_commit.c
+++ b/fs/ubifs/tnc_commit.c
@@ -22,6 +22,7 @@
 
 /* This file implements TNC functions for committing */
 
+#include <linux/random.h>
 #include "ubifs.h"
 
 /**
@@ -87,8 +88,12 @@ static int make_idx_node(struct ubifs_info *c, struct ubifs_idx_node *idx,
 	atomic_long_dec(&c->dirty_zn_cnt);
 
 	ubifs_assert(ubifs_zn_dirty(znode));
-	ubifs_assert(test_bit(COW_ZNODE, &znode->flags));
+	ubifs_assert(ubifs_zn_cow(znode));
 
+	/*
+	 * Note, unlike 'write_index()' we do not add memory barriers here
+	 * because this function is called with @c->tnc_mutex locked.
+	 */
 	__clear_bit(DIRTY_ZNODE, &znode->flags);
 	__clear_bit(COW_ZNODE, &znode->flags);
 
@@ -377,7 +382,7 @@ static int layout_in_gaps(struct ubifs_info *c, int cnt)
 				c->gap_lebs = NULL;
 				return err;
 			}
-			if (dbg_force_in_the_gaps_enabled()) {
+			if (!dbg_is_chk_index(c)) {
 				/*
 				 * Do not print scary warnings if the debugging
 				 * option which forces in-the-gaps is enabled.
@@ -491,25 +496,6 @@ static int layout_in_empty_space(struct ubifs_info *c)
 		else
 			next_len = ubifs_idx_node_sz(c, cnext->child_cnt);
 
-		if (c->min_io_size == 1) {
-			buf_offs += ALIGN(len, 8);
-			if (next_len) {
-				if (buf_offs + next_len <= c->leb_size)
-					continue;
-				err = ubifs_update_one_lp(c, lnum, 0,
-						c->leb_size - buf_offs, 0, 0);
-				if (err)
-					return err;
-				lnum = -1;
-				continue;
-			}
-			err = ubifs_update_one_lp(c, lnum,
-					c->leb_size - buf_offs, 0, 0, 0);
-			if (err)
-				return err;
-			break;
-		}
-
 		/* Update buffer positions */
 		wlen = used + len;
 		used += ALIGN(len, 8);
@@ -658,7 +644,7 @@ static int get_znodes_to_commit(struct ubifs_info *c)
 	}
 	cnt += 1;
 	while (1) {
-		ubifs_assert(!test_bit(COW_ZNODE, &znode->flags));
+		ubifs_assert(!ubifs_zn_cow(znode));
 		__set_bit(COW_ZNODE, &znode->flags);
 		znode->alt = 0;
 		cnext = find_next_dirty(znode);
@@ -704,7 +690,7 @@ static int alloc_idx_lebs(struct ubifs_info *c, int cnt)
 		c->ilebs[c->ileb_cnt++] = lnum;
 		dbg_cmt("LEB %d", lnum);
 	}
-	if (dbg_force_in_the_gaps())
+	if (dbg_is_chk_index(c) && !(random32() & 7))
 		return -ENOSPC;
 	return 0;
 }
@@ -830,7 +816,7 @@ static int write_index(struct ubifs_info *c)
 	struct ubifs_idx_node *idx;
 	struct ubifs_znode *znode, *cnext;
 	int i, lnum, offs, len, next_len, buf_len, buf_offs, used;
-	int avail, wlen, err, lnum_pos = 0;
+	int avail, wlen, err, lnum_pos = 0, blen, nxt_offs;
 
 	cnext = c->enext;
 	if (!cnext)
@@ -907,7 +893,7 @@ static int write_index(struct ubifs_info *c)
 		cnext = znode->cnext;
 
 		ubifs_assert(ubifs_zn_dirty(znode));
-		ubifs_assert(test_bit(COW_ZNODE, &znode->flags));
+		ubifs_assert(ubifs_zn_cow(znode));
 
 		/*
 		 * It is important that other threads should see %DIRTY_ZNODE
@@ -922,6 +908,28 @@ static int write_index(struct ubifs_info *c)
 		clear_bit(COW_ZNODE, &znode->flags);
 		smp_mb__after_clear_bit();
 
+		/*
+		 * We have marked the znode as clean but have not updated the
+		 * @c->clean_zn_cnt counter. If this znode becomes dirty again
+		 * before 'free_obsolete_znodes()' is called, then
+		 * @c->clean_zn_cnt will be decremented before it gets
+		 * incremented (resulting in 2 decrements for the same znode).
+		 * This means that @c->clean_zn_cnt may become negative for a
+		 * while.
+		 *
+		 * Q: why we cannot increment @c->clean_zn_cnt?
+		 * A: because we do not have the @c->tnc_mutex locked, and the
+		 *    following code would be racy and buggy:
+		 *
+		 *    if (!ubifs_zn_obsolete(znode)) {
+		 *            atomic_long_inc(&c->clean_zn_cnt);
+		 *            atomic_long_inc(&ubifs_clean_zn_cnt);
+		 *    }
+		 *
+		 *    Thus, we just delay the @c->clean_zn_cnt update until we
+		 *    have the mutex locked.
+		 */
+
 		/* Do not access znode from this point on */
 
 		/* Update buffer positions */
@@ -938,65 +946,38 @@ static int write_index(struct ubifs_info *c)
 		else
 			next_len = ubifs_idx_node_sz(c, cnext->child_cnt);
 
-		if (c->min_io_size == 1) {
-			/*
-			 * Write the prepared index node immediately if there is
-			 * no minimum IO size
-			 */
-			err = ubifs_leb_write(c, lnum, c->cbuf, buf_offs,
-					      wlen, UBI_SHORTTERM);
-			if (err)
-				return err;
-			buf_offs += ALIGN(wlen, 8);
-			if (next_len) {
-				used = 0;
-				avail = buf_len;
-				if (buf_offs + next_len > c->leb_size) {
-					err = ubifs_update_one_lp(c, lnum,
-						LPROPS_NC, 0, 0, LPROPS_TAKEN);
-					if (err)
-						return err;
-					lnum = -1;
-				}
+		nxt_offs = buf_offs + used + next_len;
+		if (next_len && nxt_offs <= c->leb_size) {
+			if (avail > 0)
 				continue;
-			}
+			else
+				blen = buf_len;
 		} else {
-			int blen, nxt_offs = buf_offs + used + next_len;
-
-			if (next_len && nxt_offs <= c->leb_size) {
-				if (avail > 0)
-					continue;
-				else
-					blen = buf_len;
-			} else {
-				wlen = ALIGN(wlen, 8);
-				blen = ALIGN(wlen, c->min_io_size);
-				ubifs_pad(c, c->cbuf + wlen, blen - wlen);
-			}
-			/*
-			 * The buffer is full or there are no more znodes
-			 * to do
-			 */
-			err = ubifs_leb_write(c, lnum, c->cbuf, buf_offs,
-					      blen, UBI_SHORTTERM);
-			if (err)
-				return err;
-			buf_offs += blen;
-			if (next_len) {
-				if (nxt_offs > c->leb_size) {
-					err = ubifs_update_one_lp(c, lnum,
-						LPROPS_NC, 0, 0, LPROPS_TAKEN);
-					if (err)
-						return err;
-					lnum = -1;
-				}
-				used -= blen;
-				if (used < 0)
-					used = 0;
-				avail = buf_len - used;
-				memmove(c->cbuf, c->cbuf + blen, used);
-				continue;
+			wlen = ALIGN(wlen, 8);
+			blen = ALIGN(wlen, c->min_io_size);
+			ubifs_pad(c, c->cbuf + wlen, blen - wlen);
+		}
+
+		/* The buffer is full or there are no more znodes to do */
+		err = ubifs_leb_write(c, lnum, c->cbuf, buf_offs, blen,
+				      UBI_SHORTTERM);
+		if (err)
+			return err;
+		buf_offs += blen;
+		if (next_len) {
+			if (nxt_offs > c->leb_size) {
+				err = ubifs_update_one_lp(c, lnum, LPROPS_NC, 0,
+							  0, LPROPS_TAKEN);
+				if (err)
+					return err;
+				lnum = -1;
 			}
+			used -= blen;
+			if (used < 0)
+				used = 0;
+			avail = buf_len - used;
+			memmove(c->cbuf, c->cbuf + blen, used);
+			continue;
 		}
 		break;
 	}
@@ -1029,7 +1010,7 @@ static void free_obsolete_znodes(struct ubifs_info *c)
 	do {
 		znode = cnext;
 		cnext = znode->cnext;
-		if (test_bit(OBSOLETE_ZNODE, &znode->flags))
+		if (ubifs_zn_obsolete(znode))
 			kfree(znode);
 		else {
 			znode->cnext = NULL;
diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h
index a70d7b4ffb2..27f22551f80 100644
--- a/fs/ubifs/ubifs.h
+++ b/fs/ubifs/ubifs.h
@@ -230,14 +230,14 @@ enum {
  * LPT cnode flag bits.
  *
  * DIRTY_CNODE: cnode is dirty
- * COW_CNODE: cnode is being committed and must be copied before writing
  * OBSOLETE_CNODE: cnode is being committed and has been copied (or deleted),
- * so it can (and must) be freed when the commit is finished
+ *                 so it can (and must) be freed when the commit is finished
+ * COW_CNODE: cnode is being committed and must be copied before writing
  */
 enum {
 	DIRTY_CNODE    = 0,
-	COW_CNODE      = 1,
-	OBSOLETE_CNODE = 2,
+	OBSOLETE_CNODE = 1,
+	COW_CNODE      = 2,
 };
 
 /*
@@ -722,12 +722,14 @@ struct ubifs_bud {
  * struct ubifs_jhead - journal head.
  * @wbuf: head's write-buffer
  * @buds_list: list of bud LEBs belonging to this journal head
+ * @grouped: non-zero if UBIFS groups nodes when writing to this journal head
  *
  * Note, the @buds list is protected by the @c->buds_lock.
  */
 struct ubifs_jhead {
 	struct ubifs_wbuf wbuf;
 	struct list_head buds_list;
+	unsigned int grouped:1;
 };
 
 /**
@@ -1466,6 +1468,15 @@ extern struct ubifs_compressor *ubifs_compressors[UBIFS_COMPR_TYPES_CNT];
 
 /* io.c */
 void ubifs_ro_mode(struct ubifs_info *c, int err);
+int ubifs_leb_read(const struct ubifs_info *c, int lnum, void *buf, int offs,
+		   int len, int even_ebadmsg);
+int ubifs_leb_write(struct ubifs_info *c, int lnum, const void *buf, int offs,
+		    int len, int dtype);
+int ubifs_leb_change(struct ubifs_info *c, int lnum, const void *buf, int len,
+		     int dtype);
+int ubifs_leb_unmap(struct ubifs_info *c, int lnum);
+int ubifs_leb_map(struct ubifs_info *c, int lnum, int dtype);
+int ubifs_is_mapped(const struct ubifs_info *c, int lnum);
 int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len);
 int ubifs_wbuf_seek_nolock(struct ubifs_wbuf *wbuf, int lnum, int offs,
 			   int dtype);
@@ -1718,7 +1729,7 @@ const struct ubifs_lprops *ubifs_fast_find_frdi_idx(struct ubifs_info *c);
 int ubifs_calc_dark(const struct ubifs_info *c, int spc);
 
 /* file.c */
-int ubifs_fsync(struct file *file, int datasync);
+int ubifs_fsync(struct file *file, loff_t start, loff_t end, int datasync);
 int ubifs_setattr(struct dentry *dentry, struct iattr *attr);
 
 /* dir.c */
@@ -1742,11 +1753,11 @@ struct inode *ubifs_iget(struct super_block *sb, unsigned long inum);
 int ubifs_recover_master_node(struct ubifs_info *c);
 int ubifs_write_rcvrd_mst_node(struct ubifs_info *c);
 struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum,
-					 int offs, void *sbuf, int grouped);
+					 int offs, void *sbuf, int jhead);
 struct ubifs_scan_leb *ubifs_recover_log_leb(struct ubifs_info *c, int lnum,
 					     int offs, void *sbuf);
-int ubifs_recover_inl_heads(const struct ubifs_info *c, void *sbuf);
-int ubifs_clean_lebs(const struct ubifs_info *c, void *sbuf);
+int ubifs_recover_inl_heads(struct ubifs_info *c, void *sbuf);
+int ubifs_clean_lebs(struct ubifs_info *c, void *sbuf);
 int ubifs_rcvry_gc_commit(struct ubifs_info *c);
 int ubifs_recover_size_accum(struct ubifs_info *c, union ubifs_key *key,
 			     int deletion, loff_t new_size);
diff --git a/fs/udf/file.c b/fs/udf/file.c
index 2a346bb1d9f..d8ffa7cc661 100644
--- a/fs/udf/file.c
+++ b/fs/udf/file.c
@@ -150,7 +150,7 @@ long udf_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	long old_block, new_block;
 	int result = -EINVAL;
 
-	if (file_permission(filp, MAY_READ) != 0) {
+	if (inode_permission(inode, MAY_READ) != 0) {
 		udf_debug("no permission to access inode %lu\n", inode->i_ino);
 		result = -EPERM;
 		goto out;
diff --git a/fs/ufs/namei.c b/fs/ufs/namei.c
index 29309e25417..639d4916224 100644
--- a/fs/ufs/namei.c
+++ b/fs/ufs/namei.c
@@ -56,16 +56,10 @@ static struct dentry *ufs_lookup(struct inode * dir, struct dentry *dentry, stru
 
 	lock_ufs(dir->i_sb);
 	ino = ufs_inode_by_name(dir, &dentry->d_name);
-	if (ino) {
+	if (ino)
 		inode = ufs_iget(dir->i_sb, ino);
-		if (IS_ERR(inode)) {
-			unlock_ufs(dir->i_sb);
-			return ERR_CAST(inode);
-		}
-	}
 	unlock_ufs(dir->i_sb);
-	d_add(dentry, inode);
-	return NULL;
+	return d_splice_alias(inode, dentry);
 }
 
 /*
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 284a7c89697..75bb316529d 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -88,8 +88,6 @@ xfs-y				+= xfs_alloc.o \
 				   xfs_vnodeops.o \
 				   xfs_rw.o
 
-xfs-$(CONFIG_XFS_TRACE)		+= xfs_btree_trace.o
-
 # Objects in linux/
 xfs-y				+= $(addprefix $(XFS_LINUX)/, \
 				   kmem.o \
diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c
index 39f4f809bb6..cac48fe22ad 100644
--- a/fs/xfs/linux-2.6/xfs_acl.c
+++ b/fs/xfs/linux-2.6/xfs_acl.c
@@ -219,7 +219,7 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 }
 
 int
-xfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+xfs_check_acl(struct inode *inode, int mask)
 {
 	struct xfs_inode *ip;
 	struct posix_acl *acl;
@@ -235,7 +235,7 @@ xfs_check_acl(struct inode *inode, int mask, unsigned int flags)
 	if (!XFS_IFORK_Q(ip))
 		return -EAGAIN;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
@@ -264,7 +264,7 @@ xfs_set_mode(struct inode *inode, mode_t mode)
 		iattr.ia_mode = mode;
 		iattr.ia_ctime = current_fs_time(inode->i_sb);
 
-		error = -xfs_setattr(XFS_I(inode), &iattr, XFS_ATTR_NOACL);
+		error = -xfs_setattr_nonsize(XFS_I(inode), &iattr, XFS_ATTR_NOACL);
 	}
 
 	return error;
diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index 79ce38be15a..63e971e2b83 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -181,6 +181,7 @@ xfs_setfilesize(
 
 	isize = xfs_ioend_new_eof(ioend);
 	if (isize) {
+		trace_xfs_setfilesize(ip, ioend->io_offset, ioend->io_size);
 		ip->i_d.di_size = isize;
 		xfs_mark_inode_dirty(ip);
 	}
@@ -894,11 +895,6 @@ out_invalidate:
  * For unwritten space on the page we need to start the conversion to
  * regular allocated space.
  * For any other dirty buffer heads on the page we should flush them.
- *
- * If we detect that a transaction would be required to flush the page, we
- * have to check the process flags first, if we are already in a transaction
- * or disk I/O during allocations is off, we need to fail the writepage and
- * redirty the page.
  */
 STATIC int
 xfs_vm_writepage(
@@ -906,7 +902,6 @@ xfs_vm_writepage(
 	struct writeback_control *wbc)
 {
 	struct inode		*inode = page->mapping->host;
-	int			delalloc, unwritten;
 	struct buffer_head	*bh, *head;
 	struct xfs_bmbt_irec	imap;
 	xfs_ioend_t		*ioend = NULL, *iohead = NULL;
@@ -938,15 +933,10 @@ xfs_vm_writepage(
 		goto redirty;
 
 	/*
-	 * We need a transaction if there are delalloc or unwritten buffers
-	 * on the page.
-	 *
-	 * If we need a transaction and the process flags say we are already
-	 * in a transaction, or no IO is allowed then mark the page dirty
-	 * again and leave the page as is.
+	 * Given that we do not allow direct reclaim to call us, we should
+	 * never be called while in a filesystem transaction.
 	 */
-	xfs_count_page_state(page, &delalloc, &unwritten);
-	if ((current->flags & PF_FSTRANS) && (delalloc || unwritten))
+	if (WARN_ON(current->flags & PF_FSTRANS))
 		goto redirty;
 
 	/* Is this page beyond the end of the file? */
@@ -970,7 +960,7 @@ xfs_vm_writepage(
 	offset = page_offset(page);
 	type = IO_OVERWRITE;
 
-	if (wbc->sync_mode == WB_SYNC_NONE && wbc->nonblocking)
+	if (wbc->sync_mode == WB_SYNC_NONE)
 		nonblocking = 1;
 
 	do {
@@ -1339,6 +1329,9 @@ xfs_end_io_direct_write(
 	} else {
 		xfs_finish_ioend_sync(ioend);
 	}
+
+	/* XXX: probably should move into the real I/O completion handler */
+	inode_dio_done(ioend->io_inode);
 }
 
 STATIC ssize_t
diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c
index 5e68099db2a..b2b41198559 100644
--- a/fs/xfs/linux-2.6/xfs_buf.c
+++ b/fs/xfs/linux-2.6/xfs_buf.c
@@ -499,16 +499,14 @@ found:
 	spin_unlock(&pag->pag_buf_lock);
 	xfs_perag_put(pag);
 
-	if (xfs_buf_cond_lock(bp)) {
-		/* failed, so wait for the lock if requested. */
-		if (!(flags & XBF_TRYLOCK)) {
-			xfs_buf_lock(bp);
-			XFS_STATS_INC(xb_get_locked_waited);
-		} else {
+	if (!xfs_buf_trylock(bp)) {
+		if (flags & XBF_TRYLOCK) {
 			xfs_buf_rele(bp);
 			XFS_STATS_INC(xb_busy_locked);
 			return NULL;
 		}
+		xfs_buf_lock(bp);
+		XFS_STATS_INC(xb_get_locked_waited);
 	}
 
 	/*
@@ -594,10 +592,8 @@ _xfs_buf_read(
 	ASSERT(!(flags & (XBF_DELWRI|XBF_WRITE)));
 	ASSERT(bp->b_bn != XFS_BUF_DADDR_NULL);
 
-	bp->b_flags &= ~(XBF_WRITE | XBF_ASYNC | XBF_DELWRI | \
-			XBF_READ_AHEAD | _XBF_RUN_QUEUES);
-	bp->b_flags |= flags & (XBF_READ | XBF_ASYNC | \
-			XBF_READ_AHEAD | _XBF_RUN_QUEUES);
+	bp->b_flags &= ~(XBF_WRITE | XBF_ASYNC | XBF_DELWRI | XBF_READ_AHEAD);
+	bp->b_flags |= flags & (XBF_READ | XBF_ASYNC | XBF_READ_AHEAD);
 
 	status = xfs_buf_iorequest(bp);
 	if (status || XFS_BUF_ISERROR(bp) || (flags & XBF_ASYNC))
@@ -681,7 +677,6 @@ xfs_buf_read_uncached(
 		return NULL;
 
 	/* set up the buffer for a read IO */
-	xfs_buf_lock(bp);
 	XFS_BUF_SET_ADDR(bp, daddr);
 	XFS_BUF_READ(bp);
 	XFS_BUF_BUSY(bp);
@@ -816,8 +811,6 @@ xfs_buf_get_uncached(
 		goto fail_free_mem;
 	}
 
-	xfs_buf_unlock(bp);
-
 	trace_xfs_buf_get_uncached(bp, _RET_IP_);
 	return bp;
 
@@ -896,8 +889,8 @@ xfs_buf_rele(
  *	to push on stale inode buffers.
  */
 int
-xfs_buf_cond_lock(
-	xfs_buf_t		*bp)
+xfs_buf_trylock(
+	struct xfs_buf		*bp)
 {
 	int			locked;
 
@@ -907,15 +900,8 @@ xfs_buf_cond_lock(
 	else if (atomic_read(&bp->b_pin_count) && (bp->b_flags & XBF_STALE))
 		xfs_log_force(bp->b_target->bt_mount, 0);
 
-	trace_xfs_buf_cond_lock(bp, _RET_IP_);
-	return locked ? 0 : -EBUSY;
-}
-
-int
-xfs_buf_lock_value(
-	xfs_buf_t		*bp)
-{
-	return bp->b_sema.count;
+	trace_xfs_buf_trylock(bp, _RET_IP_);
+	return locked;
 }
 
 /*
@@ -929,7 +915,7 @@ xfs_buf_lock_value(
  */
 void
 xfs_buf_lock(
-	xfs_buf_t		*bp)
+	struct xfs_buf		*bp)
 {
 	trace_xfs_buf_lock(bp, _RET_IP_);
 
@@ -950,7 +936,7 @@ xfs_buf_lock(
  */
 void
 xfs_buf_unlock(
-	xfs_buf_t		*bp)
+	struct xfs_buf		*bp)
 {
 	if ((bp->b_flags & (XBF_DELWRI|_XBF_DELWRI_Q)) == XBF_DELWRI) {
 		atomic_inc(&bp->b_hold);
@@ -1121,7 +1107,7 @@ xfs_bioerror_relse(
 	XFS_BUF_UNDELAYWRITE(bp);
 	XFS_BUF_DONE(bp);
 	XFS_BUF_STALE(bp);
-	XFS_BUF_CLR_IODONE_FUNC(bp);
+	bp->b_iodone = NULL;
 	if (!(fl & XBF_ASYNC)) {
 		/*
 		 * Mark b_error and B_ERROR _both_.
@@ -1223,23 +1209,21 @@ _xfs_buf_ioapply(
 	total_nr_pages = bp->b_page_count;
 	map_i = 0;
 
-	if (bp->b_flags & XBF_ORDERED) {
-		ASSERT(!(bp->b_flags & XBF_READ));
-		rw = WRITE_FLUSH_FUA;
-	} else if (bp->b_flags & XBF_LOG_BUFFER) {
-		ASSERT(!(bp->b_flags & XBF_READ_AHEAD));
-		bp->b_flags &= ~_XBF_RUN_QUEUES;
-		rw = (bp->b_flags & XBF_WRITE) ? WRITE_SYNC : READ_SYNC;
-	} else if (bp->b_flags & _XBF_RUN_QUEUES) {
-		ASSERT(!(bp->b_flags & XBF_READ_AHEAD));
-		bp->b_flags &= ~_XBF_RUN_QUEUES;
-		rw = (bp->b_flags & XBF_WRITE) ? WRITE_META : READ_META;
+	if (bp->b_flags & XBF_WRITE) {
+		if (bp->b_flags & XBF_SYNCIO)
+			rw = WRITE_SYNC;
+		else
+			rw = WRITE;
+		if (bp->b_flags & XBF_FUA)
+			rw |= REQ_FUA;
+		if (bp->b_flags & XBF_FLUSH)
+			rw |= REQ_FLUSH;
+	} else if (bp->b_flags & XBF_READ_AHEAD) {
+		rw = READA;
 	} else {
-		rw = (bp->b_flags & XBF_WRITE) ? WRITE :
-		     (bp->b_flags & XBF_READ_AHEAD) ? READA : READ;
+		rw = READ;
 	}
 
-
 next_chunk:
 	atomic_inc(&bp->b_io_remaining);
 	nr_pages = BIO_MAX_SECTORS >> (PAGE_SHIFT - BBSHIFT);
@@ -1694,15 +1678,14 @@ xfs_buf_delwri_split(
 	list_for_each_entry_safe(bp, n, dwq, b_list) {
 		ASSERT(bp->b_flags & XBF_DELWRI);
 
-		if (!XFS_BUF_ISPINNED(bp) && !xfs_buf_cond_lock(bp)) {
+		if (!XFS_BUF_ISPINNED(bp) && xfs_buf_trylock(bp)) {
 			if (!force &&
 			    time_before(jiffies, bp->b_queuetime + age)) {
 				xfs_buf_unlock(bp);
 				break;
 			}
 
-			bp->b_flags &= ~(XBF_DELWRI|_XBF_DELWRI_Q|
-					 _XBF_RUN_QUEUES);
+			bp->b_flags &= ~(XBF_DELWRI | _XBF_DELWRI_Q);
 			bp->b_flags |= XBF_WRITE;
 			list_move_tail(&bp->b_list, list);
 			trace_xfs_buf_delwri_split(bp, _RET_IP_);
@@ -1738,14 +1721,6 @@ xfs_buf_cmp(
 	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)
diff --git a/fs/xfs/linux-2.6/xfs_buf.h b/fs/xfs/linux-2.6/xfs_buf.h
index 50a7d5fb3b7..6a83b46b4bc 100644
--- a/fs/xfs/linux-2.6/xfs_buf.h
+++ b/fs/xfs/linux-2.6/xfs_buf.h
@@ -46,43 +46,46 @@ typedef enum {
 
 #define XBF_READ	(1 << 0) /* buffer intended for reading from device */
 #define XBF_WRITE	(1 << 1) /* buffer intended for writing to device */
-#define XBF_MAPPED	(1 << 2) /* buffer mapped (b_addr valid) */
+#define XBF_READ_AHEAD	(1 << 2) /* asynchronous read-ahead */
+#define XBF_MAPPED	(1 << 3) /* buffer mapped (b_addr valid) */
 #define XBF_ASYNC	(1 << 4) /* initiator will not wait for completion */
 #define XBF_DONE	(1 << 5) /* all pages in the buffer uptodate */
 #define XBF_DELWRI	(1 << 6) /* buffer has dirty pages */
 #define XBF_STALE	(1 << 7) /* buffer has been staled, do not find it */
-#define XBF_ORDERED	(1 << 11)/* use ordered writes */
-#define XBF_READ_AHEAD	(1 << 12)/* asynchronous read-ahead */
-#define XBF_LOG_BUFFER	(1 << 13)/* this is a buffer used for the log */
+
+/* I/O hints for the BIO layer */
+#define XBF_SYNCIO	(1 << 10)/* treat this buffer as synchronous I/O */
+#define XBF_FUA		(1 << 11)/* force cache write through mode */
+#define XBF_FLUSH	(1 << 12)/* flush the disk cache before a write */
 
 /* flags used only as arguments to access routines */
-#define XBF_LOCK	(1 << 14)/* lock requested */
-#define XBF_TRYLOCK	(1 << 15)/* lock requested, but do not wait */
-#define XBF_DONT_BLOCK	(1 << 16)/* do not block in current thread */
+#define XBF_LOCK	(1 << 15)/* lock requested */
+#define XBF_TRYLOCK	(1 << 16)/* lock requested, but do not wait */
+#define XBF_DONT_BLOCK	(1 << 17)/* do not block in current thread */
 
 /* flags used only internally */
-#define _XBF_PAGES	(1 << 18)/* backed by refcounted pages */
-#define	_XBF_RUN_QUEUES	(1 << 19)/* run block device task queue	*/
-#define	_XBF_KMEM	(1 << 20)/* backed by heap memory */
-#define _XBF_DELWRI_Q	(1 << 21)/* buffer on delwri queue */
+#define _XBF_PAGES	(1 << 20)/* backed by refcounted pages */
+#define _XBF_KMEM	(1 << 21)/* backed by heap memory */
+#define _XBF_DELWRI_Q	(1 << 22)/* buffer on delwri queue */
 
 typedef unsigned int xfs_buf_flags_t;
 
 #define XFS_BUF_FLAGS \
 	{ XBF_READ,		"READ" }, \
 	{ XBF_WRITE,		"WRITE" }, \
+	{ XBF_READ_AHEAD,	"READ_AHEAD" }, \
 	{ XBF_MAPPED,		"MAPPED" }, \
 	{ XBF_ASYNC,		"ASYNC" }, \
 	{ XBF_DONE,		"DONE" }, \
 	{ XBF_DELWRI,		"DELWRI" }, \
 	{ XBF_STALE,		"STALE" }, \
-	{ XBF_ORDERED,		"ORDERED" }, \
-	{ XBF_READ_AHEAD,	"READ_AHEAD" }, \
+	{ XBF_SYNCIO,		"SYNCIO" }, \
+	{ XBF_FUA,		"FUA" }, \
+	{ XBF_FLUSH,		"FLUSH" }, \
 	{ XBF_LOCK,		"LOCK" },  	/* should never be set */\
 	{ XBF_TRYLOCK,		"TRYLOCK" }, 	/* ditto */\
 	{ XBF_DONT_BLOCK,	"DONT_BLOCK" },	/* ditto */\
 	{ _XBF_PAGES,		"PAGES" }, \
-	{ _XBF_RUN_QUEUES,	"RUN_QUEUES" }, \
 	{ _XBF_KMEM,		"KMEM" }, \
 	{ _XBF_DELWRI_Q,	"DELWRI_Q" }
 
@@ -91,11 +94,6 @@ typedef enum {
 	XBT_FORCE_FLUSH = 1,
 } xfs_buftarg_flags_t;
 
-typedef struct xfs_bufhash {
-	struct list_head	bh_list;
-	spinlock_t		bh_lock;
-} xfs_bufhash_t;
-
 typedef struct xfs_buftarg {
 	dev_t			bt_dev;
 	struct block_device	*bt_bdev;
@@ -151,7 +149,7 @@ typedef struct xfs_buf {
 	xfs_buf_iodone_t	b_iodone;	/* I/O completion function */
 	struct completion	b_iowait;	/* queue for I/O waiters */
 	void			*b_fspriv;
-	void			*b_fspriv2;
+	struct xfs_trans	*b_transp;
 	struct page		**b_pages;	/* array of page pointers */
 	struct page		*b_page_array[XB_PAGES]; /* inline pages */
 	unsigned long		b_queuetime;	/* time buffer was queued */
@@ -192,10 +190,11 @@ extern void xfs_buf_free(xfs_buf_t *);
 extern void xfs_buf_rele(xfs_buf_t *);
 
 /* Locking and Unlocking Buffers */
-extern int xfs_buf_cond_lock(xfs_buf_t *);
-extern int xfs_buf_lock_value(xfs_buf_t *);
+extern int xfs_buf_trylock(xfs_buf_t *);
 extern void xfs_buf_lock(xfs_buf_t *);
 extern void xfs_buf_unlock(xfs_buf_t *);
+#define xfs_buf_islocked(bp) \
+	((bp)->b_sema.count <= 0)
 
 /* Buffer Read and Write Routines */
 extern int xfs_bwrite(struct xfs_mount *mp, struct xfs_buf *bp);
@@ -234,8 +233,9 @@ extern void xfs_buf_terminate(void);
 
 
 #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_ZEROFLAGS(bp) \
+	((bp)->b_flags &= ~(XBF_READ|XBF_WRITE|XBF_ASYNC|XBF_DELWRI| \
+			    XBF_SYNCIO|XBF_FUA|XBF_FLUSH))
 
 void xfs_buf_stale(struct xfs_buf *bp);
 #define XFS_BUF_STALE(bp)	xfs_buf_stale(bp);
@@ -267,10 +267,6 @@ void xfs_buf_stale(struct xfs_buf *bp);
 #define XFS_BUF_UNASYNC(bp)	((bp)->b_flags &= ~XBF_ASYNC)
 #define XFS_BUF_ISASYNC(bp)	((bp)->b_flags & XBF_ASYNC)
 
-#define XFS_BUF_ORDERED(bp)	((bp)->b_flags |= XBF_ORDERED)
-#define XFS_BUF_UNORDERED(bp)	((bp)->b_flags &= ~XBF_ORDERED)
-#define XFS_BUF_ISORDERED(bp)	((bp)->b_flags & XBF_ORDERED)
-
 #define XFS_BUF_HOLD(bp)	xfs_buf_hold(bp)
 #define XFS_BUF_READ(bp)	((bp)->b_flags |= XBF_READ)
 #define XFS_BUF_UNREAD(bp)	((bp)->b_flags &= ~XBF_READ)
@@ -280,14 +276,6 @@ void xfs_buf_stale(struct xfs_buf *bp);
 #define XFS_BUF_UNWRITE(bp)	((bp)->b_flags &= ~XBF_WRITE)
 #define XFS_BUF_ISWRITE(bp)	((bp)->b_flags & XBF_WRITE)
 
-#define XFS_BUF_IODONE_FUNC(bp)			((bp)->b_iodone)
-#define XFS_BUF_SET_IODONE_FUNC(bp, func)	((bp)->b_iodone = (func))
-#define XFS_BUF_CLR_IODONE_FUNC(bp)		((bp)->b_iodone = NULL)
-
-#define XFS_BUF_FSPRIVATE(bp, type)		((type)(bp)->b_fspriv)
-#define XFS_BUF_SET_FSPRIVATE(bp, val)		((bp)->b_fspriv = (void*)(val))
-#define XFS_BUF_FSPRIVATE2(bp, type)		((type)(bp)->b_fspriv2)
-#define XFS_BUF_SET_FSPRIVATE2(bp, val)		((bp)->b_fspriv2 = (void*)(val))
 #define XFS_BUF_SET_START(bp)			do { } while (0)
 
 #define XFS_BUF_PTR(bp)			(xfs_caddr_t)((bp)->b_addr)
@@ -313,10 +301,6 @@ xfs_buf_set_ref(
 
 #define XFS_BUF_ISPINNED(bp)	atomic_read(&((bp)->b_pin_count))
 
-#define XFS_BUF_VALUSEMA(bp)	xfs_buf_lock_value(bp)
-#define XFS_BUF_CPSEMA(bp)	(xfs_buf_cond_lock(bp) == 0)
-#define XFS_BUF_VSEMA(bp)	xfs_buf_unlock(bp)
-#define XFS_BUF_PSEMA(bp,x)	xfs_buf_lock(bp)
 #define XFS_BUF_FINISH_IOWAIT(bp)	complete(&bp->b_iowait);
 
 #define XFS_BUF_SET_TARGET(bp, target)	((bp)->b_target = (target))
diff --git a/fs/xfs/linux-2.6/xfs_export.c b/fs/xfs/linux-2.6/xfs_export.c
index f4f878fc008..75e5d322e48 100644
--- a/fs/xfs/linux-2.6/xfs_export.c
+++ b/fs/xfs/linux-2.6/xfs_export.c
@@ -151,14 +151,14 @@ xfs_nfs_get_inode(
 		 * We don't use ESTALE directly down the chain to not
 		 * confuse applications using bulkstat that expect EINVAL.
 		 */
-		if (error == EINVAL)
+		if (error == EINVAL || error == ENOENT)
 			error = ESTALE;
 		return ERR_PTR(-error);
 	}
 
 	if (ip->i_d.di_gen != generation) {
 		IRELE(ip);
-		return ERR_PTR(-ENOENT);
+		return ERR_PTR(-ESTALE);
 	}
 
 	return VFS_I(ip);
diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c
index f4213ba1ff8..cca00f49e09 100644
--- a/fs/xfs/linux-2.6/xfs_file.c
+++ b/fs/xfs/linux-2.6/xfs_file.c
@@ -127,23 +127,44 @@ xfs_iozero(
 STATIC int
 xfs_file_fsync(
 	struct file		*file,
+	loff_t			start,
+	loff_t			end,
 	int			datasync)
 {
 	struct inode		*inode = file->f_mapping->host;
 	struct xfs_inode	*ip = XFS_I(inode);
+	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_trans	*tp;
 	int			error = 0;
 	int			log_flushed = 0;
 
 	trace_xfs_file_fsync(ip);
 
-	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
+	error = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (error)
+		return error;
+
+	if (XFS_FORCED_SHUTDOWN(mp))
 		return -XFS_ERROR(EIO);
 
 	xfs_iflags_clear(ip, XFS_ITRUNCATED);
 
 	xfs_ioend_wait(ip);
 
+	if (mp->m_flags & XFS_MOUNT_BARRIER) {
+		/*
+		 * If we have an RT and/or log subvolume we need to make sure
+		 * to flush the write cache the device used for file data
+		 * first.  This is to ensure newly written file data make
+		 * it to disk before logging the new inode size in case of
+		 * an extending write.
+		 */
+		if (XFS_IS_REALTIME_INODE(ip))
+			xfs_blkdev_issue_flush(mp->m_rtdev_targp);
+		else if (mp->m_logdev_targp != mp->m_ddev_targp)
+			xfs_blkdev_issue_flush(mp->m_ddev_targp);
+	}
+
 	/*
 	 * 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
@@ -175,9 +196,9 @@ xfs_file_fsync(
 		 * 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);
+		tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
 		error = xfs_trans_reserve(tp, 0,
-				XFS_FSYNC_TS_LOG_RES(ip->i_mount), 0, 0, 0);
+				XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
 		if (error) {
 			xfs_trans_cancel(tp, 0);
 			return -error;
@@ -209,28 +230,25 @@ xfs_file_fsync(
 		 * force the log.
 		 */
 		if (xfs_ipincount(ip)) {
-			error = _xfs_log_force_lsn(ip->i_mount,
+			error = _xfs_log_force_lsn(mp,
 					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);
-	}
+	/*
+	 * If we only have a single device, and the log force about was
+	 * a no-op we might have to flush the data device cache here.
+	 * This can only happen for fdatasync/O_DSYNC if we were overwriting
+	 * an already allocated file and thus do not have any metadata to
+	 * commit.
+	 */
+	if ((mp->m_flags & XFS_MOUNT_BARRIER) &&
+	    mp->m_logdev_targp == mp->m_ddev_targp &&
+	    !XFS_IS_REALTIME_INODE(ip) &&
+	    !log_flushed)
+		xfs_blkdev_issue_flush(mp->m_ddev_targp);
 
 	return -error;
 }
@@ -863,18 +881,11 @@ xfs_file_aio_write(
 	/* Handle various SYNC-type writes */
 	if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) {
 		loff_t end = pos + ret - 1;
-		int error, error2;
 
 		xfs_rw_iunlock(ip, iolock);
-		error = filemap_write_and_wait_range(mapping, pos, end);
+		ret = -xfs_file_fsync(file, pos, end,
+				      (file->f_flags & __O_SYNC) ? 0 : 1);
 		xfs_rw_ilock(ip, iolock);
-
-		error2 = -xfs_file_fsync(file,
-					 (file->f_flags & __O_SYNC) ? 0 : 1);
-		if (error)
-			ret = error;
-		else if (error2)
-			ret = error2;
 	}
 
 out_unlock:
@@ -932,7 +943,7 @@ xfs_file_fallocate(
 
 		iattr.ia_valid = ATTR_SIZE;
 		iattr.ia_size = new_size;
-		error = -xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK);
+		error = -xfs_setattr_size(ip, &iattr, XFS_ATTR_NOLOCK);
 	}
 
 out_unlock:
diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c
index dd21784525a..501e4f63054 100644
--- a/fs/xfs/linux-2.6/xfs_iops.c
+++ b/fs/xfs/linux-2.6/xfs_iops.c
@@ -39,6 +39,7 @@
 #include "xfs_buf_item.h"
 #include "xfs_utils.h"
 #include "xfs_vnodeops.h"
+#include "xfs_inode_item.h"
 #include "xfs_trace.h"
 
 #include <linux/capability.h>
@@ -182,7 +183,7 @@ xfs_vn_mknod(
 	if (IS_POSIXACL(dir)) {
 		default_acl = xfs_get_acl(dir, ACL_TYPE_DEFAULT);
 		if (IS_ERR(default_acl))
-			return -PTR_ERR(default_acl);
+			return PTR_ERR(default_acl);
 
 		if (!default_acl)
 			mode &= ~current_umask();
@@ -497,12 +498,442 @@ xfs_vn_getattr(
 	return 0;
 }
 
+int
+xfs_setattr_nonsize(
+	struct xfs_inode	*ip,
+	struct iattr		*iattr,
+	int			flags)
+{
+	xfs_mount_t		*mp = ip->i_mount;
+	struct inode		*inode = VFS_I(ip);
+	int			mask = iattr->ia_valid;
+	xfs_trans_t		*tp;
+	int			error;
+	uid_t			uid = 0, iuid = 0;
+	gid_t			gid = 0, igid = 0;
+	struct xfs_dquot	*udqp = NULL, *gdqp = NULL;
+	struct xfs_dquot	*olddquot1 = NULL, *olddquot2 = NULL;
+
+	trace_xfs_setattr(ip);
+
+	if (mp->m_flags & XFS_MOUNT_RDONLY)
+		return XFS_ERROR(EROFS);
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return XFS_ERROR(EIO);
+
+	error = -inode_change_ok(inode, iattr);
+	if (error)
+		return XFS_ERROR(error);
+
+	ASSERT((mask & ATTR_SIZE) == 0);
+
+	/*
+	 * If disk quotas is on, we make sure that the dquots do exist on disk,
+	 * before we start any other transactions. Trying to do this later
+	 * is messy. We don't care to take a readlock to look at the ids
+	 * in inode here, because we can't hold it across the trans_reserve.
+	 * If the IDs do change before we take the ilock, we're covered
+	 * because the i_*dquot fields will get updated anyway.
+	 */
+	if (XFS_IS_QUOTA_ON(mp) && (mask & (ATTR_UID|ATTR_GID))) {
+		uint	qflags = 0;
+
+		if ((mask & ATTR_UID) && XFS_IS_UQUOTA_ON(mp)) {
+			uid = iattr->ia_uid;
+			qflags |= XFS_QMOPT_UQUOTA;
+		} else {
+			uid = ip->i_d.di_uid;
+		}
+		if ((mask & ATTR_GID) && XFS_IS_GQUOTA_ON(mp)) {
+			gid = iattr->ia_gid;
+			qflags |= XFS_QMOPT_GQUOTA;
+		}  else {
+			gid = ip->i_d.di_gid;
+		}
+
+		/*
+		 * We take a reference when we initialize udqp and gdqp,
+		 * so it is important that we never blindly double trip on
+		 * the same variable. See xfs_create() for an example.
+		 */
+		ASSERT(udqp == NULL);
+		ASSERT(gdqp == NULL);
+		error = xfs_qm_vop_dqalloc(ip, uid, gid, xfs_get_projid(ip),
+					 qflags, &udqp, &gdqp);
+		if (error)
+			return error;
+	}
+
+	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
+	error = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0);
+	if (error)
+		goto out_dqrele;
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+
+	/*
+	 * Change file ownership.  Must be the owner or privileged.
+	 */
+	if (mask & (ATTR_UID|ATTR_GID)) {
+		/*
+		 * These IDs could have changed since we last looked at them.
+		 * But, we're assured that if the ownership did change
+		 * while we didn't have the inode locked, inode's dquot(s)
+		 * would have changed also.
+		 */
+		iuid = ip->i_d.di_uid;
+		igid = ip->i_d.di_gid;
+		gid = (mask & ATTR_GID) ? iattr->ia_gid : igid;
+		uid = (mask & ATTR_UID) ? iattr->ia_uid : iuid;
+
+		/*
+		 * Do a quota reservation only if uid/gid is actually
+		 * going to change.
+		 */
+		if (XFS_IS_QUOTA_RUNNING(mp) &&
+		    ((XFS_IS_UQUOTA_ON(mp) && iuid != uid) ||
+		     (XFS_IS_GQUOTA_ON(mp) && igid != gid))) {
+			ASSERT(tp);
+			error = xfs_qm_vop_chown_reserve(tp, ip, udqp, gdqp,
+						capable(CAP_FOWNER) ?
+						XFS_QMOPT_FORCE_RES : 0);
+			if (error)	/* out of quota */
+				goto out_trans_cancel;
+		}
+	}
+
+	xfs_trans_ijoin(tp, ip);
+
+	/*
+	 * Change file ownership.  Must be the owner or privileged.
+	 */
+	if (mask & (ATTR_UID|ATTR_GID)) {
+		/*
+		 * CAP_FSETID overrides the following restrictions:
+		 *
+		 * The set-user-ID and set-group-ID bits of a file will be
+		 * cleared upon successful return from chown()
+		 */
+		if ((ip->i_d.di_mode & (S_ISUID|S_ISGID)) &&
+		    !capable(CAP_FSETID))
+			ip->i_d.di_mode &= ~(S_ISUID|S_ISGID);
+
+		/*
+		 * Change the ownerships and register quota modifications
+		 * in the transaction.
+		 */
+		if (iuid != uid) {
+			if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_UQUOTA_ON(mp)) {
+				ASSERT(mask & ATTR_UID);
+				ASSERT(udqp);
+				olddquot1 = xfs_qm_vop_chown(tp, ip,
+							&ip->i_udquot, udqp);
+			}
+			ip->i_d.di_uid = uid;
+			inode->i_uid = uid;
+		}
+		if (igid != gid) {
+			if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_GQUOTA_ON(mp)) {
+				ASSERT(!XFS_IS_PQUOTA_ON(mp));
+				ASSERT(mask & ATTR_GID);
+				ASSERT(gdqp);
+				olddquot2 = xfs_qm_vop_chown(tp, ip,
+							&ip->i_gdquot, gdqp);
+			}
+			ip->i_d.di_gid = gid;
+			inode->i_gid = gid;
+		}
+	}
+
+	/*
+	 * Change file access modes.
+	 */
+	if (mask & ATTR_MODE) {
+		umode_t mode = iattr->ia_mode;
+
+		if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
+			mode &= ~S_ISGID;
+
+		ip->i_d.di_mode &= S_IFMT;
+		ip->i_d.di_mode |= mode & ~S_IFMT;
+
+		inode->i_mode &= S_IFMT;
+		inode->i_mode |= mode & ~S_IFMT;
+	}
+
+	/*
+	 * Change file access or modified times.
+	 */
+	if (mask & ATTR_ATIME) {
+		inode->i_atime = iattr->ia_atime;
+		ip->i_d.di_atime.t_sec = iattr->ia_atime.tv_sec;
+		ip->i_d.di_atime.t_nsec = iattr->ia_atime.tv_nsec;
+		ip->i_update_core = 1;
+	}
+	if (mask & ATTR_CTIME) {
+		inode->i_ctime = iattr->ia_ctime;
+		ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
+		ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
+		ip->i_update_core = 1;
+	}
+	if (mask & ATTR_MTIME) {
+		inode->i_mtime = iattr->ia_mtime;
+		ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
+		ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
+		ip->i_update_core = 1;
+	}
+
+	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+
+	XFS_STATS_INC(xs_ig_attrchg);
+
+	if (mp->m_flags & XFS_MOUNT_WSYNC)
+		xfs_trans_set_sync(tp);
+	error = xfs_trans_commit(tp, 0);
+
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+
+	/*
+	 * Release any dquot(s) the inode had kept before chown.
+	 */
+	xfs_qm_dqrele(olddquot1);
+	xfs_qm_dqrele(olddquot2);
+	xfs_qm_dqrele(udqp);
+	xfs_qm_dqrele(gdqp);
+
+	if (error)
+		return XFS_ERROR(error);
+
+	/*
+	 * XXX(hch): Updating the ACL entries is not atomic vs the i_mode
+	 * 	     update.  We could avoid this with linked transactions
+	 * 	     and passing down the transaction pointer all the way
+	 *	     to attr_set.  No previous user of the generic
+	 * 	     Posix ACL code seems to care about this issue either.
+	 */
+	if ((mask & ATTR_MODE) && !(flags & XFS_ATTR_NOACL)) {
+		error = -xfs_acl_chmod(inode);
+		if (error)
+			return XFS_ERROR(error);
+	}
+
+	return 0;
+
+out_trans_cancel:
+	xfs_trans_cancel(tp, 0);
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+out_dqrele:
+	xfs_qm_dqrele(udqp);
+	xfs_qm_dqrele(gdqp);
+	return error;
+}
+
+/*
+ * Truncate file.  Must have write permission and not be a directory.
+ */
+int
+xfs_setattr_size(
+	struct xfs_inode	*ip,
+	struct iattr		*iattr,
+	int			flags)
+{
+	struct xfs_mount	*mp = ip->i_mount;
+	struct inode		*inode = VFS_I(ip);
+	int			mask = iattr->ia_valid;
+	struct xfs_trans	*tp;
+	int			error;
+	uint			lock_flags;
+	uint			commit_flags = 0;
+
+	trace_xfs_setattr(ip);
+
+	if (mp->m_flags & XFS_MOUNT_RDONLY)
+		return XFS_ERROR(EROFS);
+
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return XFS_ERROR(EIO);
+
+	error = -inode_change_ok(inode, iattr);
+	if (error)
+		return XFS_ERROR(error);
+
+	ASSERT(S_ISREG(ip->i_d.di_mode));
+	ASSERT((mask & (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET|
+			ATTR_MTIME_SET|ATTR_KILL_SUID|ATTR_KILL_SGID|
+			ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0);
+
+	lock_flags = XFS_ILOCK_EXCL;
+	if (!(flags & XFS_ATTR_NOLOCK))
+		lock_flags |= XFS_IOLOCK_EXCL;
+	xfs_ilock(ip, lock_flags);
+
+	/*
+	 * Short circuit the truncate case for zero length files.
+	 */
+	if (iattr->ia_size == 0 &&
+	    ip->i_size == 0 && ip->i_d.di_nextents == 0) {
+		if (!(mask & (ATTR_CTIME|ATTR_MTIME)))
+			goto out_unlock;
+
+		/*
+		 * Use the regular setattr path to update the timestamps.
+		 */
+		xfs_iunlock(ip, lock_flags);
+		iattr->ia_valid &= ~ATTR_SIZE;
+		return xfs_setattr_nonsize(ip, iattr, 0);
+	}
+
+	/*
+	 * Make sure that the dquots are attached to the inode.
+	 */
+	error = xfs_qm_dqattach_locked(ip, 0);
+	if (error)
+		goto out_unlock;
+
+	/*
+	 * Now we can make the changes.  Before we join the inode to the
+	 * transaction, take care of the part of the truncation that must be
+	 * done without the inode lock.  This needs to be done before joining
+	 * the inode to the transaction, because the inode cannot be unlocked
+	 * once it is a part of the transaction.
+	 */
+	if (iattr->ia_size > ip->i_size) {
+		/*
+		 * Do the first part of growing a file: zero any data in the
+		 * last block that is beyond the old EOF.  We need to do this
+		 * before the inode is joined to the transaction to modify
+		 * i_size.
+		 */
+		error = xfs_zero_eof(ip, iattr->ia_size, ip->i_size);
+		if (error)
+			goto out_unlock;
+	}
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+	lock_flags &= ~XFS_ILOCK_EXCL;
+
+	/*
+	 * We are going to log the inode size change in this transaction so
+	 * any previous writes that are beyond the on disk EOF and the new
+	 * EOF that have not been written out need to be written here.  If we
+	 * do not write the data out, we expose ourselves to the null files
+	 * problem.
+	 *
+	 * Only flush from the on disk size to the smaller of the in memory
+	 * file size or the new size as that's the range we really care about
+	 * here and prevents waiting for other data not within the range we
+	 * care about here.
+	 */
+	if (ip->i_size != ip->i_d.di_size && iattr->ia_size > ip->i_d.di_size) {
+		error = xfs_flush_pages(ip, ip->i_d.di_size, iattr->ia_size,
+					XBF_ASYNC, FI_NONE);
+		if (error)
+			goto out_unlock;
+	}
+
+	/*
+	 * Wait for all I/O to complete.
+	 */
+	xfs_ioend_wait(ip);
+
+	error = -block_truncate_page(inode->i_mapping, iattr->ia_size,
+				     xfs_get_blocks);
+	if (error)
+		goto out_unlock;
+
+	tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
+	error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
+				 XFS_TRANS_PERM_LOG_RES,
+				 XFS_ITRUNCATE_LOG_COUNT);
+	if (error)
+		goto out_trans_cancel;
+
+	truncate_setsize(inode, iattr->ia_size);
+
+	commit_flags = XFS_TRANS_RELEASE_LOG_RES;
+	lock_flags |= XFS_ILOCK_EXCL;
+
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+
+	xfs_trans_ijoin(tp, ip);
+
+	/*
+	 * Only change the c/mtime if we are changing the size or we are
+	 * explicitly asked to change it.  This handles the semantic difference
+	 * between truncate() and ftruncate() as implemented in the VFS.
+	 *
+	 * The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a
+	 * special case where we need to update the times despite not having
+	 * these flags set.  For all other operations the VFS set these flags
+	 * explicitly if it wants a timestamp update.
+	 */
+	if (iattr->ia_size != ip->i_size &&
+	    (!(mask & (ATTR_CTIME | ATTR_MTIME)))) {
+		iattr->ia_ctime = iattr->ia_mtime =
+			current_fs_time(inode->i_sb);
+		mask |= ATTR_CTIME | ATTR_MTIME;
+	}
+
+	if (iattr->ia_size > ip->i_size) {
+		ip->i_d.di_size = iattr->ia_size;
+		ip->i_size = iattr->ia_size;
+	} else if (iattr->ia_size <= ip->i_size ||
+		   (iattr->ia_size == 0 && ip->i_d.di_nextents)) {
+		error = xfs_itruncate_data(&tp, ip, iattr->ia_size);
+		if (error)
+			goto out_trans_abort;
+
+		/*
+		 * Truncated "down", so we're removing references to old data
+		 * here - if we delay flushing for a long time, we expose
+		 * ourselves unduly to the notorious NULL files problem.  So,
+		 * we mark this inode and flush it when the file is closed,
+		 * and do not wait the usual (long) time for writeout.
+		 */
+		xfs_iflags_set(ip, XFS_ITRUNCATED);
+	}
+
+	if (mask & ATTR_CTIME) {
+		inode->i_ctime = iattr->ia_ctime;
+		ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
+		ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
+		ip->i_update_core = 1;
+	}
+	if (mask & ATTR_MTIME) {
+		inode->i_mtime = iattr->ia_mtime;
+		ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
+		ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
+		ip->i_update_core = 1;
+	}
+
+	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+
+	XFS_STATS_INC(xs_ig_attrchg);
+
+	if (mp->m_flags & XFS_MOUNT_WSYNC)
+		xfs_trans_set_sync(tp);
+
+	error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
+out_unlock:
+	if (lock_flags)
+		xfs_iunlock(ip, lock_flags);
+	return error;
+
+out_trans_abort:
+	commit_flags |= XFS_TRANS_ABORT;
+out_trans_cancel:
+	xfs_trans_cancel(tp, commit_flags);
+	goto out_unlock;
+}
+
 STATIC int
 xfs_vn_setattr(
 	struct dentry	*dentry,
 	struct iattr	*iattr)
 {
-	return -xfs_setattr(XFS_I(dentry->d_inode), iattr, 0);
+	if (iattr->ia_valid & ATTR_SIZE)
+		return -xfs_setattr_size(XFS_I(dentry->d_inode), iattr, 0);
+	return -xfs_setattr_nonsize(XFS_I(dentry->d_inode), iattr, 0);
 }
 
 #define XFS_FIEMAP_FLAGS	(FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)
diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h
index 8633521b3b2..d42f814e4d3 100644
--- a/fs/xfs/linux-2.6/xfs_linux.h
+++ b/fs/xfs/linux-2.6/xfs_linux.h
@@ -33,7 +33,6 @@
 #endif
 
 #include <xfs_types.h>
-#include <xfs_arch.h>
 
 #include <kmem.h>
 #include <mrlock.h>
@@ -88,6 +87,12 @@
 #include <xfs_buf.h>
 #include <xfs_message.h>
 
+#ifdef __BIG_ENDIAN
+#define XFS_NATIVE_HOST 1
+#else
+#undef XFS_NATIVE_HOST
+#endif
+
 /*
  * Feature macros (disable/enable)
  */
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index 1e3a7ce804d..9a72dda58bd 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -33,7 +33,6 @@
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_btree.h"
-#include "xfs_btree_trace.h"
 #include "xfs_ialloc.h"
 #include "xfs_bmap.h"
 #include "xfs_rtalloc.h"
@@ -627,68 +626,6 @@ xfs_blkdev_put(
 		blkdev_put(bdev, FMODE_READ|FMODE_WRITE|FMODE_EXCL);
 }
 
-/*
- * Try to write out the superblock using barriers.
- */
-STATIC int
-xfs_barrier_test(
-	xfs_mount_t	*mp)
-{
-	xfs_buf_t	*sbp = xfs_getsb(mp, 0);
-	int		error;
-
-	XFS_BUF_UNDONE(sbp);
-	XFS_BUF_UNREAD(sbp);
-	XFS_BUF_UNDELAYWRITE(sbp);
-	XFS_BUF_WRITE(sbp);
-	XFS_BUF_UNASYNC(sbp);
-	XFS_BUF_ORDERED(sbp);
-
-	xfsbdstrat(mp, sbp);
-	error = xfs_buf_iowait(sbp);
-
-	/*
-	 * Clear all the flags we set and possible error state in the
-	 * buffer.  We only did the write to try out whether barriers
-	 * worked and shouldn't leave any traces in the superblock
-	 * buffer.
-	 */
-	XFS_BUF_DONE(sbp);
-	XFS_BUF_ERROR(sbp, 0);
-	XFS_BUF_UNORDERED(sbp);
-
-	xfs_buf_relse(sbp);
-	return error;
-}
-
-STATIC void
-xfs_mountfs_check_barriers(xfs_mount_t *mp)
-{
-	int error;
-
-	if (mp->m_logdev_targp != mp->m_ddev_targp) {
-		xfs_notice(mp,
-		  "Disabling barriers, not supported with external log device");
-		mp->m_flags &= ~XFS_MOUNT_BARRIER;
-		return;
-	}
-
-	if (xfs_readonly_buftarg(mp->m_ddev_targp)) {
-		xfs_notice(mp,
-			"Disabling barriers, underlying device is readonly");
-		mp->m_flags &= ~XFS_MOUNT_BARRIER;
-		return;
-	}
-
-	error = xfs_barrier_test(mp);
-	if (error) {
-		xfs_notice(mp,
-			"Disabling barriers, trial barrier write failed");
-		mp->m_flags &= ~XFS_MOUNT_BARRIER;
-		return;
-	}
-}
-
 void
 xfs_blkdev_issue_flush(
 	xfs_buftarg_t		*buftarg)
@@ -1087,11 +1024,6 @@ xfs_fs_put_super(
 {
 	struct xfs_mount	*mp = XFS_M(sb);
 
-	/*
-	 * Unregister the memory shrinker before we tear down the mount
-	 * structure so we don't have memory reclaim racing with us here.
-	 */
-	xfs_inode_shrinker_unregister(mp);
 	xfs_syncd_stop(mp);
 
 	/*
@@ -1240,14 +1172,6 @@ xfs_fs_remount(
 		switch (token) {
 		case Opt_barrier:
 			mp->m_flags |= XFS_MOUNT_BARRIER;
-
-			/*
-			 * Test if barriers are actually working if we can,
-			 * else delay this check until the filesystem is
-			 * marked writeable.
-			 */
-			if (!(mp->m_flags & XFS_MOUNT_RDONLY))
-				xfs_mountfs_check_barriers(mp);
 			break;
 		case Opt_nobarrier:
 			mp->m_flags &= ~XFS_MOUNT_BARRIER;
@@ -1282,8 +1206,6 @@ xfs_fs_remount(
 	/* ro -> rw */
 	if ((mp->m_flags & XFS_MOUNT_RDONLY) && !(*flags & MS_RDONLY)) {
 		mp->m_flags &= ~XFS_MOUNT_RDONLY;
-		if (mp->m_flags & XFS_MOUNT_BARRIER)
-			xfs_mountfs_check_barriers(mp);
 
 		/*
 		 * If this is the first remount to writeable state we
@@ -1465,9 +1387,6 @@ xfs_fs_fill_super(
 	if (error)
 		goto out_free_sb;
 
-	if (mp->m_flags & XFS_MOUNT_BARRIER)
-		xfs_mountfs_check_barriers(mp);
-
 	error = xfs_filestream_mount(mp);
 	if (error)
 		goto out_free_sb;
@@ -1487,36 +1406,31 @@ xfs_fs_fill_super(
 	sb->s_time_gran = 1;
 	set_posix_acl_flag(sb);
 
-	error = xfs_syncd_init(mp);
+	error = xfs_mountfs(mp);
 	if (error)
 		goto out_filestream_unmount;
 
-	xfs_inode_shrinker_register(mp);
-
-	error = xfs_mountfs(mp);
+	error = xfs_syncd_init(mp);
 	if (error)
-		goto out_syncd_stop;
+		goto out_unmount;
 
 	root = igrab(VFS_I(mp->m_rootip));
 	if (!root) {
 		error = ENOENT;
-		goto fail_unmount;
+		goto out_syncd_stop;
 	}
 	if (is_bad_inode(root)) {
 		error = EINVAL;
-		goto fail_vnrele;
+		goto out_syncd_stop;
 	}
 	sb->s_root = d_alloc_root(root);
 	if (!sb->s_root) {
 		error = ENOMEM;
-		goto fail_vnrele;
+		goto out_iput;
 	}
 
 	return 0;
 
- out_syncd_stop:
-	xfs_inode_shrinker_unregister(mp);
-	xfs_syncd_stop(mp);
  out_filestream_unmount:
 	xfs_filestream_unmount(mp);
  out_free_sb:
@@ -1531,18 +1445,11 @@ xfs_fs_fill_super(
  out:
 	return -error;
 
- fail_vnrele:
-	if (sb->s_root) {
-		dput(sb->s_root);
-		sb->s_root = NULL;
-	} else {
-		iput(root);
-	}
-
- fail_unmount:
-	xfs_inode_shrinker_unregister(mp);
+ out_iput:
+	iput(root);
+ out_syncd_stop:
 	xfs_syncd_stop(mp);
-
+ out_unmount:
 	/*
 	 * Blow away any referenced inode in the filestreams cache.
 	 * This can and will cause log traffic as inodes go inactive
@@ -1566,6 +1473,21 @@ xfs_fs_mount(
 	return mount_bdev(fs_type, flags, dev_name, data, xfs_fs_fill_super);
 }
 
+static int
+xfs_fs_nr_cached_objects(
+	struct super_block	*sb)
+{
+	return xfs_reclaim_inodes_count(XFS_M(sb));
+}
+
+static void
+xfs_fs_free_cached_objects(
+	struct super_block	*sb,
+	int			nr_to_scan)
+{
+	xfs_reclaim_inodes_nr(XFS_M(sb), nr_to_scan);
+}
+
 static const struct super_operations xfs_super_operations = {
 	.alloc_inode		= xfs_fs_alloc_inode,
 	.destroy_inode		= xfs_fs_destroy_inode,
@@ -1579,6 +1501,8 @@ static const struct super_operations xfs_super_operations = {
 	.statfs			= xfs_fs_statfs,
 	.remount_fs		= xfs_fs_remount,
 	.show_options		= xfs_fs_show_options,
+	.nr_cached_objects	= xfs_fs_nr_cached_objects,
+	.free_cached_objects	= xfs_fs_free_cached_objects,
 };
 
 static struct file_system_type xfs_fs_type = {
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index 8ecad5ff9f9..e4c938afb91 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -179,6 +179,8 @@ restart:
 		if (error == EFSCORRUPTED)
 			break;
 
+		cond_resched();
+
 	} while (nr_found && !done);
 
 	if (skipped) {
@@ -359,14 +361,12 @@ xfs_quiesce_data(
 {
 	int			error, error2 = 0;
 
-	/* push non-blocking */
-	xfs_sync_data(mp, 0);
 	xfs_qm_sync(mp, SYNC_TRYLOCK);
-
-	/* push and block till complete */
-	xfs_sync_data(mp, SYNC_WAIT);
 	xfs_qm_sync(mp, SYNC_WAIT);
 
+	/* force out the newly dirtied log buffers */
+	xfs_log_force(mp, XFS_LOG_SYNC);
+
 	/* write superblock and hoover up shutdown errors */
 	error = xfs_sync_fsdata(mp);
 
@@ -436,7 +436,7 @@ xfs_quiesce_attr(
 	WARN_ON(atomic_read(&mp->m_active_trans) != 0);
 
 	/* Push the superblock and write an unmount record */
-	error = xfs_log_sbcount(mp, 1);
+	error = xfs_log_sbcount(mp);
 	if (error)
 		xfs_warn(mp, "xfs_attr_quiesce: failed to log sb changes. "
 				"Frozen image may not be consistent.");
@@ -986,6 +986,8 @@ restart:
 
 			*nr_to_scan -= XFS_LOOKUP_BATCH;
 
+			cond_resched();
+
 		} while (nr_found && !done && *nr_to_scan > 0);
 
 		if (trylock && !done)
@@ -1003,7 +1005,7 @@ restart:
 	 * ensure that when we get more reclaimers than AGs we block rather
 	 * than spin trying to execute reclaim.
 	 */
-	if (trylock && skipped && *nr_to_scan > 0) {
+	if (skipped && (flags & SYNC_WAIT) && *nr_to_scan > 0) {
 		trylock = 0;
 		goto restart;
 	}
@@ -1021,44 +1023,38 @@ xfs_reclaim_inodes(
 }
 
 /*
- * Inode cache shrinker.
+ * Scan a certain number of inodes for reclaim.
  *
  * When called we make sure that there is a background (fast) inode reclaim in
- * progress, while we will throttle the speed of reclaim via doiing synchronous
+ * progress, while we will throttle the speed of reclaim via doing synchronous
  * reclaim of inodes. That means if we come across dirty inodes, we wait for
  * them to be cleaned, which we hope will not be very long due to the
  * background walker having already kicked the IO off on those dirty inodes.
  */
-static int
-xfs_reclaim_inode_shrink(
-	struct shrinker	*shrink,
-	struct shrink_control *sc)
+void
+xfs_reclaim_inodes_nr(
+	struct xfs_mount	*mp,
+	int			nr_to_scan)
 {
-	struct xfs_mount *mp;
-	struct xfs_perag *pag;
-	xfs_agnumber_t	ag;
-	int		reclaimable;
-	int nr_to_scan = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
-
-	mp = container_of(shrink, struct xfs_mount, m_inode_shrink);
-	if (nr_to_scan) {
-		/* kick background reclaimer and push the AIL */
-		xfs_syncd_queue_reclaim(mp);
-		xfs_ail_push_all(mp->m_ail);
+	/* kick background reclaimer and push the AIL */
+	xfs_syncd_queue_reclaim(mp);
+	xfs_ail_push_all(mp->m_ail);
 
-		if (!(gfp_mask & __GFP_FS))
-			return -1;
+	xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT, &nr_to_scan);
+}
 
-		xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT,
-					&nr_to_scan);
-		/* terminate if we don't exhaust the scan */
-		if (nr_to_scan > 0)
-			return -1;
-       }
+/*
+ * Return the number of reclaimable inodes in the filesystem for
+ * the shrinker to determine how much to reclaim.
+ */
+int
+xfs_reclaim_inodes_count(
+	struct xfs_mount	*mp)
+{
+	struct xfs_perag	*pag;
+	xfs_agnumber_t		ag = 0;
+	int			reclaimable = 0;
 
-	reclaimable = 0;
-	ag = 0;
 	while ((pag = xfs_perag_get_tag(mp, ag, XFS_ICI_RECLAIM_TAG))) {
 		ag = pag->pag_agno + 1;
 		reclaimable += pag->pag_ici_reclaimable;
@@ -1067,18 +1063,3 @@ xfs_reclaim_inode_shrink(
 	return reclaimable;
 }
 
-void
-xfs_inode_shrinker_register(
-	struct xfs_mount	*mp)
-{
-	mp->m_inode_shrink.shrink = xfs_reclaim_inode_shrink;
-	mp->m_inode_shrink.seeks = DEFAULT_SEEKS;
-	register_shrinker(&mp->m_inode_shrink);
-}
-
-void
-xfs_inode_shrinker_unregister(
-	struct xfs_mount	*mp)
-{
-	unregister_shrinker(&mp->m_inode_shrink);
-}
diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h
index e3a6ad27415..941202e7ac6 100644
--- a/fs/xfs/linux-2.6/xfs_sync.h
+++ b/fs/xfs/linux-2.6/xfs_sync.h
@@ -21,14 +21,6 @@
 struct xfs_mount;
 struct xfs_perag;
 
-typedef struct xfs_sync_work {
-	struct list_head	w_list;
-	struct xfs_mount	*w_mount;
-	void			*w_data;	/* syncer routine argument */
-	void			(*w_syncer)(struct xfs_mount *, void *);
-	struct completion	*w_completion;
-} xfs_sync_work_t;
-
 #define SYNC_WAIT		0x0001	/* wait for i/o to complete */
 #define SYNC_TRYLOCK		0x0002  /* only try to lock inodes */
 
@@ -43,6 +35,8 @@ void xfs_quiesce_attr(struct xfs_mount *mp);
 void xfs_flush_inodes(struct xfs_inode *ip);
 
 int xfs_reclaim_inodes(struct xfs_mount *mp, int mode);
+int xfs_reclaim_inodes_count(struct xfs_mount *mp);
+void xfs_reclaim_inodes_nr(struct xfs_mount *mp, int nr_to_scan);
 
 void xfs_inode_set_reclaim_tag(struct xfs_inode *ip);
 void __xfs_inode_set_reclaim_tag(struct xfs_perag *pag, struct xfs_inode *ip);
@@ -54,7 +48,4 @@ int xfs_inode_ag_iterator(struct xfs_mount *mp,
 	int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag, int flags),
 	int flags);
 
-void xfs_inode_shrinker_register(struct xfs_mount *mp);
-void xfs_inode_shrinker_unregister(struct xfs_mount *mp);
-
 #endif
diff --git a/fs/xfs/linux-2.6/xfs_trace.h b/fs/xfs/linux-2.6/xfs_trace.h
index d48b7a579ae..fda0708ef2e 100644
--- a/fs/xfs/linux-2.6/xfs_trace.h
+++ b/fs/xfs/linux-2.6/xfs_trace.h
@@ -293,7 +293,7 @@ DECLARE_EVENT_CLASS(xfs_buf_class,
 		__entry->buffer_length = bp->b_buffer_length;
 		__entry->hold = atomic_read(&bp->b_hold);
 		__entry->pincount = atomic_read(&bp->b_pin_count);
-		__entry->lockval = xfs_buf_lock_value(bp);
+		__entry->lockval = bp->b_sema.count;
 		__entry->flags = bp->b_flags;
 		__entry->caller_ip = caller_ip;
 	),
@@ -323,7 +323,7 @@ DEFINE_BUF_EVENT(xfs_buf_bawrite);
 DEFINE_BUF_EVENT(xfs_buf_bdwrite);
 DEFINE_BUF_EVENT(xfs_buf_lock);
 DEFINE_BUF_EVENT(xfs_buf_lock_done);
-DEFINE_BUF_EVENT(xfs_buf_cond_lock);
+DEFINE_BUF_EVENT(xfs_buf_trylock);
 DEFINE_BUF_EVENT(xfs_buf_unlock);
 DEFINE_BUF_EVENT(xfs_buf_iowait);
 DEFINE_BUF_EVENT(xfs_buf_iowait_done);
@@ -366,7 +366,7 @@ DECLARE_EVENT_CLASS(xfs_buf_flags_class,
 		__entry->flags = flags;
 		__entry->hold = atomic_read(&bp->b_hold);
 		__entry->pincount = atomic_read(&bp->b_pin_count);
-		__entry->lockval = xfs_buf_lock_value(bp);
+		__entry->lockval = bp->b_sema.count;
 		__entry->caller_ip = caller_ip;
 	),
 	TP_printk("dev %d:%d bno 0x%llx len 0x%zx hold %d pincount %d "
@@ -409,7 +409,7 @@ TRACE_EVENT(xfs_buf_ioerror,
 		__entry->buffer_length = bp->b_buffer_length;
 		__entry->hold = atomic_read(&bp->b_hold);
 		__entry->pincount = atomic_read(&bp->b_pin_count);
-		__entry->lockval = xfs_buf_lock_value(bp);
+		__entry->lockval = bp->b_sema.count;
 		__entry->error = error;
 		__entry->flags = bp->b_flags;
 		__entry->caller_ip = caller_ip;
@@ -454,7 +454,7 @@ DECLARE_EVENT_CLASS(xfs_buf_item_class,
 		__entry->buf_flags = bip->bli_buf->b_flags;
 		__entry->buf_hold = atomic_read(&bip->bli_buf->b_hold);
 		__entry->buf_pincount = atomic_read(&bip->bli_buf->b_pin_count);
-		__entry->buf_lockval = xfs_buf_lock_value(bip->bli_buf);
+		__entry->buf_lockval = bip->bli_buf->b_sema.count;
 		__entry->li_desc = bip->bli_item.li_desc;
 		__entry->li_flags = bip->bli_item.li_flags;
 	),
@@ -998,7 +998,8 @@ DECLARE_EVENT_CLASS(xfs_simple_io_class,
 	TP_STRUCT__entry(
 		__field(dev_t, dev)
 		__field(xfs_ino_t, ino)
-		__field(loff_t, size)
+		__field(loff_t, isize)
+		__field(loff_t, disize)
 		__field(loff_t, new_size)
 		__field(loff_t, offset)
 		__field(size_t, count)
@@ -1006,16 +1007,18 @@ DECLARE_EVENT_CLASS(xfs_simple_io_class,
 	TP_fast_assign(
 		__entry->dev = VFS_I(ip)->i_sb->s_dev;
 		__entry->ino = ip->i_ino;
-		__entry->size = ip->i_d.di_size;
+		__entry->isize = ip->i_size;
+		__entry->disize = ip->i_d.di_size;
 		__entry->new_size = ip->i_new_size;
 		__entry->offset = offset;
 		__entry->count = count;
 	),
-	TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx "
+	TP_printk("dev %d:%d ino 0x%llx isize 0x%llx disize 0x%llx new_size 0x%llx "
 		  "offset 0x%llx count %zd",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
 		  __entry->ino,
-		  __entry->size,
+		  __entry->isize,
+		  __entry->disize,
 		  __entry->new_size,
 		  __entry->offset,
 		  __entry->count)
@@ -1028,40 +1031,7 @@ DEFINE_EVENT(xfs_simple_io_class, name,	\
 DEFINE_SIMPLE_IO_EVENT(xfs_delalloc_enospc);
 DEFINE_SIMPLE_IO_EVENT(xfs_unwritten_convert);
 DEFINE_SIMPLE_IO_EVENT(xfs_get_blocks_notfound);
-
-
-TRACE_EVENT(xfs_itruncate_start,
-	TP_PROTO(struct xfs_inode *ip, xfs_fsize_t new_size, int flag,
-		 xfs_off_t toss_start, xfs_off_t toss_finish),
-	TP_ARGS(ip, new_size, flag, toss_start, toss_finish),
-	TP_STRUCT__entry(
-		__field(dev_t, dev)
-		__field(xfs_ino_t, ino)
-		__field(xfs_fsize_t, size)
-		__field(xfs_fsize_t, new_size)
-		__field(xfs_off_t, toss_start)
-		__field(xfs_off_t, toss_finish)
-		__field(int, flag)
-	),
-	TP_fast_assign(
-		__entry->dev = VFS_I(ip)->i_sb->s_dev;
-		__entry->ino = ip->i_ino;
-		__entry->size = ip->i_d.di_size;
-		__entry->new_size = new_size;
-		__entry->toss_start = toss_start;
-		__entry->toss_finish = toss_finish;
-		__entry->flag = flag;
-	),
-	TP_printk("dev %d:%d ino 0x%llx %s size 0x%llx new_size 0x%llx "
-		  "toss start 0x%llx toss finish 0x%llx",
-		  MAJOR(__entry->dev), MINOR(__entry->dev),
-		  __entry->ino,
-		  __print_flags(__entry->flag, "|", XFS_ITRUNC_FLAGS),
-		  __entry->size,
-		  __entry->new_size,
-		  __entry->toss_start,
-		  __entry->toss_finish)
-);
+DEFINE_SIMPLE_IO_EVENT(xfs_setfilesize);
 
 DECLARE_EVENT_CLASS(xfs_itrunc_class,
 	TP_PROTO(struct xfs_inode *ip, xfs_fsize_t new_size),
@@ -1089,8 +1059,8 @@ DECLARE_EVENT_CLASS(xfs_itrunc_class,
 DEFINE_EVENT(xfs_itrunc_class, name, \
 	TP_PROTO(struct xfs_inode *ip, xfs_fsize_t new_size), \
 	TP_ARGS(ip, new_size))
-DEFINE_ITRUNC_EVENT(xfs_itruncate_finish_start);
-DEFINE_ITRUNC_EVENT(xfs_itruncate_finish_end);
+DEFINE_ITRUNC_EVENT(xfs_itruncate_data_start);
+DEFINE_ITRUNC_EVENT(xfs_itruncate_data_end);
 
 TRACE_EVENT(xfs_pagecache_inval,
 	TP_PROTO(struct xfs_inode *ip, xfs_off_t start, xfs_off_t finish),
diff --git a/fs/xfs/quota/xfs_dquot.c b/fs/xfs/quota/xfs_dquot.c
index 6fa21460381..837f31158d4 100644
--- a/fs/xfs/quota/xfs_dquot.c
+++ b/fs/xfs/quota/xfs_dquot.c
@@ -220,7 +220,7 @@ xfs_qm_adjust_dqtimers(
 {
 	ASSERT(d->d_id);
 
-#ifdef QUOTADEBUG
+#ifdef DEBUG
 	if (d->d_blk_hardlimit)
 		ASSERT(be64_to_cpu(d->d_blk_softlimit) <=
 		       be64_to_cpu(d->d_blk_hardlimit));
@@ -231,6 +231,7 @@ xfs_qm_adjust_dqtimers(
 		ASSERT(be64_to_cpu(d->d_rtb_softlimit) <=
 		       be64_to_cpu(d->d_rtb_hardlimit));
 #endif
+
 	if (!d->d_btimer) {
 		if ((d->d_blk_softlimit &&
 		     (be64_to_cpu(d->d_bcount) >=
@@ -318,7 +319,7 @@ xfs_qm_init_dquot_blk(
 
 	ASSERT(tp);
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
+	ASSERT(xfs_buf_islocked(bp));
 
 	d = (xfs_dqblk_t *)XFS_BUF_PTR(bp);
 
@@ -534,7 +535,7 @@ xfs_qm_dqtobp(
 	}
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
+	ASSERT(xfs_buf_islocked(bp));
 
 	/*
 	 * calculate the location of the dquot inside the buffer.
@@ -622,7 +623,7 @@ xfs_qm_dqread(
 	 * brelse it because we have the changes incore.
 	 */
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
+	ASSERT(xfs_buf_islocked(bp));
 	xfs_trans_brelse(tp, bp);
 
 	return (error);
@@ -1423,45 +1424,6 @@ xfs_qm_dqpurge(
 }
 
 
-#ifdef QUOTADEBUG
-void
-xfs_qm_dqprint(xfs_dquot_t *dqp)
-{
-	struct xfs_mount	*mp = dqp->q_mount;
-
-	xfs_debug(mp, "-----------KERNEL DQUOT----------------");
-	xfs_debug(mp, "---- dquotID =  %d",
-		(int)be32_to_cpu(dqp->q_core.d_id));
-	xfs_debug(mp, "---- type    =  %s", DQFLAGTO_TYPESTR(dqp));
-	xfs_debug(mp, "---- fs      =  0x%p", dqp->q_mount);
-	xfs_debug(mp, "---- blkno   =  0x%x", (int) dqp->q_blkno);
-	xfs_debug(mp, "---- boffset =  0x%x", (int) dqp->q_bufoffset);
-	xfs_debug(mp, "---- blkhlimit =  %Lu (0x%x)",
-		be64_to_cpu(dqp->q_core.d_blk_hardlimit),
-		(int)be64_to_cpu(dqp->q_core.d_blk_hardlimit));
-	xfs_debug(mp, "---- blkslimit =  %Lu (0x%x)",
-		be64_to_cpu(dqp->q_core.d_blk_softlimit),
-		(int)be64_to_cpu(dqp->q_core.d_blk_softlimit));
-	xfs_debug(mp, "---- inohlimit =  %Lu (0x%x)",
-		be64_to_cpu(dqp->q_core.d_ino_hardlimit),
-		(int)be64_to_cpu(dqp->q_core.d_ino_hardlimit));
-	xfs_debug(mp, "---- inoslimit =  %Lu (0x%x)",
-		be64_to_cpu(dqp->q_core.d_ino_softlimit),
-		(int)be64_to_cpu(dqp->q_core.d_ino_softlimit));
-	xfs_debug(mp, "---- bcount  =  %Lu (0x%x)",
-		be64_to_cpu(dqp->q_core.d_bcount),
-		(int)be64_to_cpu(dqp->q_core.d_bcount));
-	xfs_debug(mp, "---- icount  =  %Lu (0x%x)",
-		be64_to_cpu(dqp->q_core.d_icount),
-		(int)be64_to_cpu(dqp->q_core.d_icount));
-	xfs_debug(mp, "---- btimer  =  %d",
-		(int)be32_to_cpu(dqp->q_core.d_btimer));
-	xfs_debug(mp, "---- itimer  =  %d",
-		(int)be32_to_cpu(dqp->q_core.d_itimer));
-	xfs_debug(mp, "---------------------------");
-}
-#endif
-
 /*
  * Give the buffer a little push if it is incore and
  * wait on the flush lock.
diff --git a/fs/xfs/quota/xfs_dquot.h b/fs/xfs/quota/xfs_dquot.h
index 5da3a23b820..34b7e945dbf 100644
--- a/fs/xfs/quota/xfs_dquot.h
+++ b/fs/xfs/quota/xfs_dquot.h
@@ -116,12 +116,6 @@ static inline void xfs_dqfunlock(xfs_dquot_t *dqp)
 				     (XFS_IS_UQUOTA_ON((d)->q_mount)) : \
 				     (XFS_IS_OQUOTA_ON((d)->q_mount))))
 
-#ifdef QUOTADEBUG
-extern void		xfs_qm_dqprint(xfs_dquot_t *);
-#else
-#define xfs_qm_dqprint(a)
-#endif
-
 extern void		xfs_qm_dqdestroy(xfs_dquot_t *);
 extern int		xfs_qm_dqflush(xfs_dquot_t *, uint);
 extern int		xfs_qm_dqpurge(xfs_dquot_t *);
diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c
index b94dace4e78..46e54ad9a2d 100644
--- a/fs/xfs/quota/xfs_qm.c
+++ b/fs/xfs/quota/xfs_qm.c
@@ -67,32 +67,6 @@ static struct shrinker xfs_qm_shaker = {
 	.seeks = DEFAULT_SEEKS,
 };
 
-#ifdef DEBUG
-extern struct mutex	qcheck_lock;
-#endif
-
-#ifdef QUOTADEBUG
-static void
-xfs_qm_dquot_list_print(
-	struct xfs_mount *mp)
-{
-	xfs_dquot_t	*dqp;
-	int		i = 0;
-
-	list_for_each_entry(dqp, &mp->m_quotainfo->qi_dqlist_lock, qi_mplist) {
-		xfs_debug(mp, "   %d. \"%d (%s)\"   "
-				  "bcnt = %lld, icnt = %lld, refs = %d",
-			i++, be32_to_cpu(dqp->q_core.d_id),
-			DQFLAGTO_TYPESTR(dqp),
-			(long long)be64_to_cpu(dqp->q_core.d_bcount),
-			(long long)be64_to_cpu(dqp->q_core.d_icount),
-			dqp->q_nrefs);
-	}
-}
-#else
-static void xfs_qm_dquot_list_print(struct xfs_mount *mp) { }
-#endif
-
 /*
  * Initialize the XQM structure.
  * Note that there is not one quota manager per file system.
@@ -165,9 +139,6 @@ xfs_Gqm_init(void)
 	atomic_set(&xqm->qm_totaldquots, 0);
 	xqm->qm_dqfree_ratio = XFS_QM_DQFREE_RATIO;
 	xqm->qm_nrefs = 0;
-#ifdef DEBUG
-	mutex_init(&qcheck_lock);
-#endif
 	return xqm;
 
  out_free_udqhash:
@@ -204,9 +175,6 @@ xfs_qm_destroy(
 	mutex_lock(&xqm->qm_dqfrlist_lock);
 	list_for_each_entry_safe(dqp, n, &xqm->qm_dqfrlist, q_freelist) {
 		xfs_dqlock(dqp);
-#ifdef QUOTADEBUG
-		xfs_debug(dqp->q_mount, "FREELIST destroy 0x%p", dqp);
-#endif
 		list_del_init(&dqp->q_freelist);
 		xfs_Gqm->qm_dqfrlist_cnt--;
 		xfs_dqunlock(dqp);
@@ -214,9 +182,6 @@ xfs_qm_destroy(
 	}
 	mutex_unlock(&xqm->qm_dqfrlist_lock);
 	mutex_destroy(&xqm->qm_dqfrlist_lock);
-#ifdef DEBUG
-	mutex_destroy(&qcheck_lock);
-#endif
 	kmem_free(xqm);
 }
 
@@ -409,11 +374,6 @@ xfs_qm_mount_quotas(
 		xfs_warn(mp, "Failed to initialize disk quotas.");
 		return;
 	}
-
-#ifdef QUOTADEBUG
-	if (XFS_IS_QUOTA_ON(mp))
-		xfs_qm_internalqcheck(mp);
-#endif
 }
 
 /*
@@ -866,8 +826,8 @@ xfs_qm_dqattach_locked(
 	}
 
  done:
-#ifdef QUOTADEBUG
-	if (! error) {
+#ifdef DEBUG
+	if (!error) {
 		if (XFS_IS_UQUOTA_ON(mp))
 			ASSERT(ip->i_udquot);
 		if (XFS_IS_OQUOTA_ON(mp))
@@ -1733,8 +1693,6 @@ xfs_qm_quotacheck(
 	mp->m_qflags &= ~(XFS_OQUOTA_CHKD | XFS_UQUOTA_CHKD);
 	mp->m_qflags |= flags;
 
-	xfs_qm_dquot_list_print(mp);
-
  error_return:
 	if (error) {
 		xfs_warn(mp,
@@ -2096,9 +2054,6 @@ xfs_qm_write_sb_changes(
 	xfs_trans_t	*tp;
 	int		error;
 
-#ifdef QUOTADEBUG
-	xfs_notice(mp, "Writing superblock quota changes");
-#endif
 	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE);
 	if ((error = xfs_trans_reserve(tp, 0,
 				      mp->m_sb.sb_sectsize + 128, 0,
diff --git a/fs/xfs/quota/xfs_qm.h b/fs/xfs/quota/xfs_qm.h
index 567b29b9f1b..43b9abe1052 100644
--- a/fs/xfs/quota/xfs_qm.h
+++ b/fs/xfs/quota/xfs_qm.h
@@ -163,10 +163,4 @@ extern int		xfs_qm_scall_getqstat(xfs_mount_t *, fs_quota_stat_t *);
 extern int		xfs_qm_scall_quotaon(xfs_mount_t *, uint);
 extern int		xfs_qm_scall_quotaoff(xfs_mount_t *, uint);
 
-#ifdef DEBUG
-extern int		xfs_qm_internalqcheck(xfs_mount_t *);
-#else
-#define xfs_qm_internalqcheck(mp)	(0)
-#endif
-
 #endif /* __XFS_QM_H__ */
diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c
index 2dadb15d5ca..609246f42e6 100644
--- a/fs/xfs/quota/xfs_qm_syscalls.c
+++ b/fs/xfs/quota/xfs_qm_syscalls.c
@@ -263,7 +263,7 @@ xfs_qm_scall_trunc_qfile(
 	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_trans_ijoin(tp, ip);
 
-	error = xfs_itruncate_finish(&tp, ip, 0, XFS_DATA_FORK, 1);
+	error = xfs_itruncate_data(&tp, ip, 0);
 	if (error) {
 		xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES |
 				     XFS_TRANS_ABORT);
@@ -622,7 +622,6 @@ xfs_qm_scall_setqlim(
 	xfs_trans_log_dquot(tp, dqp);
 
 	error = xfs_trans_commit(tp, 0);
-	xfs_qm_dqprint(dqp);
 	xfs_qm_dqrele(dqp);
 
  out_unlock:
@@ -657,7 +656,6 @@ xfs_qm_scall_getquota(
 		xfs_qm_dqput(dqp);
 		return XFS_ERROR(ENOENT);
 	}
-	/* xfs_qm_dqprint(dqp); */
 	/*
 	 * Convert the disk dquot to the exportable format
 	 */
@@ -906,354 +904,3 @@ xfs_qm_dqrele_all_inodes(
 	ASSERT(mp->m_quotainfo);
 	xfs_inode_ag_iterator(mp, xfs_dqrele_inode, flags);
 }
-
-/*------------------------------------------------------------------------*/
-#ifdef DEBUG
-/*
- * This contains all the test functions for XFS disk quotas.
- * Currently it does a quota accounting check. ie. it walks through
- * all inodes in the file system, calculating the dquot accounting fields,
- * and prints out any inconsistencies.
- */
-xfs_dqhash_t *qmtest_udqtab;
-xfs_dqhash_t *qmtest_gdqtab;
-int	      qmtest_hashmask;
-int	      qmtest_nfails;
-struct mutex  qcheck_lock;
-
-#define DQTEST_HASHVAL(mp, id) (((__psunsigned_t)(mp) + \
-				 (__psunsigned_t)(id)) & \
-				(qmtest_hashmask - 1))
-
-#define DQTEST_HASH(mp, id, type)   ((type & XFS_DQ_USER) ? \
-				     (qmtest_udqtab + \
-				      DQTEST_HASHVAL(mp, id)) : \
-				     (qmtest_gdqtab + \
-				      DQTEST_HASHVAL(mp, id)))
-
-#define DQTEST_LIST_PRINT(l, NXT, title) \
-{ \
-	  xfs_dqtest_t	*dqp; int i = 0;\
-	  xfs_debug(NULL, "%s (#%d)", title, (int) (l)->qh_nelems); \
-	  for (dqp = (xfs_dqtest_t *)(l)->qh_next; dqp != NULL; \
-	       dqp = (xfs_dqtest_t *)dqp->NXT) { \
-		xfs_debug(dqp->q_mount,		\
-			"  %d. \"%d (%s)\"  bcnt = %d, icnt = %d", \
-			 ++i, dqp->d_id, DQFLAGTO_TYPESTR(dqp),	     \
-			 dqp->d_bcount, dqp->d_icount); } \
-}
-
-typedef struct dqtest {
-	uint		 dq_flags;	/* various flags (XFS_DQ_*) */
-	struct list_head q_hashlist;
-	xfs_dqhash_t	*q_hash;	/* the hashchain header */
-	xfs_mount_t	*q_mount;	/* filesystem this relates to */
-	xfs_dqid_t	d_id;		/* user id or group id */
-	xfs_qcnt_t	d_bcount;	/* # disk blocks owned by the user */
-	xfs_qcnt_t	d_icount;	/* # inodes owned by the user */
-} xfs_dqtest_t;
-
-STATIC void
-xfs_qm_hashinsert(xfs_dqhash_t *h, xfs_dqtest_t *dqp)
-{
-	list_add(&dqp->q_hashlist, &h->qh_list);
-	h->qh_version++;
-	h->qh_nelems++;
-}
-STATIC void
-xfs_qm_dqtest_print(
-	struct xfs_mount	*mp,
-	struct dqtest		*d)
-{
-	xfs_debug(mp, "-----------DQTEST DQUOT----------------");
-	xfs_debug(mp, "---- dquot ID = %d", d->d_id);
-	xfs_debug(mp, "---- fs       = 0x%p", d->q_mount);
-	xfs_debug(mp, "---- bcount   = %Lu (0x%x)",
-		d->d_bcount, (int)d->d_bcount);
-	xfs_debug(mp, "---- icount   = %Lu (0x%x)",
-		d->d_icount, (int)d->d_icount);
-	xfs_debug(mp, "---------------------------");
-}
-
-STATIC void
-xfs_qm_dqtest_failed(
-	xfs_dqtest_t	*d,
-	xfs_dquot_t	*dqp,
-	char		*reason,
-	xfs_qcnt_t	a,
-	xfs_qcnt_t	b,
-	int		error)
-{
-	qmtest_nfails++;
-	if (error)
-		xfs_debug(dqp->q_mount,
-			"quotacheck failed id=%d, err=%d\nreason: %s",
-			d->d_id, error, reason);
-	else
-		xfs_debug(dqp->q_mount,
-			"quotacheck failed id=%d (%s) [%d != %d]",
-			d->d_id, reason, (int)a, (int)b);
-	xfs_qm_dqtest_print(dqp->q_mount, d);
-	if (dqp)
-		xfs_qm_dqprint(dqp);
-}
-
-STATIC int
-xfs_dqtest_cmp2(
-	xfs_dqtest_t	*d,
-	xfs_dquot_t	*dqp)
-{
-	int err = 0;
-	if (be64_to_cpu(dqp->q_core.d_icount) != d->d_icount) {
-		xfs_qm_dqtest_failed(d, dqp, "icount mismatch",
-			be64_to_cpu(dqp->q_core.d_icount),
-			d->d_icount, 0);
-		err++;
-	}
-	if (be64_to_cpu(dqp->q_core.d_bcount) != d->d_bcount) {
-		xfs_qm_dqtest_failed(d, dqp, "bcount mismatch",
-			be64_to_cpu(dqp->q_core.d_bcount),
-			d->d_bcount, 0);
-		err++;
-	}
-	if (dqp->q_core.d_blk_softlimit &&
-	    be64_to_cpu(dqp->q_core.d_bcount) >=
-	    be64_to_cpu(dqp->q_core.d_blk_softlimit)) {
-		if (!dqp->q_core.d_btimer && dqp->q_core.d_id) {
-			xfs_debug(dqp->q_mount,
-				"%d [%s] BLK TIMER NOT STARTED",
-				d->d_id, DQFLAGTO_TYPESTR(d));
-			err++;
-		}
-	}
-	if (dqp->q_core.d_ino_softlimit &&
-	    be64_to_cpu(dqp->q_core.d_icount) >=
-	    be64_to_cpu(dqp->q_core.d_ino_softlimit)) {
-		if (!dqp->q_core.d_itimer && dqp->q_core.d_id) {
-			xfs_debug(dqp->q_mount,
-				"%d [%s] INO TIMER NOT STARTED",
-				d->d_id, DQFLAGTO_TYPESTR(d));
-			err++;
-		}
-	}
-#ifdef QUOTADEBUG
-	if (!err) {
-		xfs_debug(dqp->q_mount, "%d [%s] qchecked",
-			d->d_id, DQFLAGTO_TYPESTR(d));
-	}
-#endif
-	return (err);
-}
-
-STATIC void
-xfs_dqtest_cmp(
-	xfs_dqtest_t	*d)
-{
-	xfs_dquot_t	*dqp;
-	int		error;
-
-	/* xfs_qm_dqtest_print(d); */
-	if ((error = xfs_qm_dqget(d->q_mount, NULL, d->d_id, d->dq_flags, 0,
-				 &dqp))) {
-		xfs_qm_dqtest_failed(d, NULL, "dqget failed", 0, 0, error);
-		return;
-	}
-	xfs_dqtest_cmp2(d, dqp);
-	xfs_qm_dqput(dqp);
-}
-
-STATIC int
-xfs_qm_internalqcheck_dqget(
-	xfs_mount_t	*mp,
-	xfs_dqid_t	id,
-	uint		type,
-	xfs_dqtest_t	**O_dq)
-{
-	xfs_dqtest_t	*d;
-	xfs_dqhash_t	*h;
-
-	h = DQTEST_HASH(mp, id, type);
-	list_for_each_entry(d, &h->qh_list, q_hashlist) {
-		if (d->d_id == id && mp == d->q_mount) {
-			*O_dq = d;
-			return (0);
-		}
-	}
-	d = kmem_zalloc(sizeof(xfs_dqtest_t), KM_SLEEP);
-	d->dq_flags = type;
-	d->d_id = id;
-	d->q_mount = mp;
-	d->q_hash = h;
-	INIT_LIST_HEAD(&d->q_hashlist);
-	xfs_qm_hashinsert(h, d);
-	*O_dq = d;
-	return (0);
-}
-
-STATIC void
-xfs_qm_internalqcheck_get_dquots(
-	xfs_mount_t	*mp,
-	xfs_dqid_t	uid,
-	xfs_dqid_t	projid,
-	xfs_dqid_t	gid,
-	xfs_dqtest_t	**ud,
-	xfs_dqtest_t	**gd)
-{
-	if (XFS_IS_UQUOTA_ON(mp))
-		xfs_qm_internalqcheck_dqget(mp, uid, XFS_DQ_USER, ud);
-	if (XFS_IS_GQUOTA_ON(mp))
-		xfs_qm_internalqcheck_dqget(mp, gid, XFS_DQ_GROUP, gd);
-	else if (XFS_IS_PQUOTA_ON(mp))
-		xfs_qm_internalqcheck_dqget(mp, projid, XFS_DQ_PROJ, gd);
-}
-
-
-STATIC void
-xfs_qm_internalqcheck_dqadjust(
-	xfs_inode_t		*ip,
-	xfs_dqtest_t		*d)
-{
-	d->d_icount++;
-	d->d_bcount += (xfs_qcnt_t)ip->i_d.di_nblocks;
-}
-
-STATIC int
-xfs_qm_internalqcheck_adjust(
-	xfs_mount_t	*mp,		/* mount point for filesystem */
-	xfs_ino_t	ino,		/* inode number to get data for */
-	void		__user *buffer,	/* not used */
-	int		ubsize,		/* not used */
-	int		*ubused,	/* not used */
-	int		*res)		/* bulkstat result code */
-{
-	xfs_inode_t		*ip;
-	xfs_dqtest_t		*ud, *gd;
-	uint			lock_flags;
-	boolean_t		ipreleased;
-	int			error;
-
-	ASSERT(XFS_IS_QUOTA_RUNNING(mp));
-
-	if (ino == mp->m_sb.sb_uquotino || ino == mp->m_sb.sb_gquotino) {
-		*res = BULKSTAT_RV_NOTHING;
-		xfs_debug(mp, "%s: ino=%llu, uqino=%llu, gqino=%llu\n",
-			__func__, (unsigned long long) ino,
-			(unsigned long long) mp->m_sb.sb_uquotino,
-			(unsigned long long) mp->m_sb.sb_gquotino);
-		return XFS_ERROR(EINVAL);
-	}
-	ipreleased = B_FALSE;
- again:
-	lock_flags = XFS_ILOCK_SHARED;
-	if ((error = xfs_iget(mp, NULL, ino, 0, lock_flags, &ip))) {
-		*res = BULKSTAT_RV_NOTHING;
-		return (error);
-	}
-
-	/*
-	 * This inode can have blocks after eof which can get released
-	 * when we send it to inactive. Since we don't check the dquot
-	 * until the after all our calculations are done, we must get rid
-	 * of those now.
-	 */
-	if (! ipreleased) {
-		xfs_iunlock(ip, lock_flags);
-		IRELE(ip);
-		ipreleased = B_TRUE;
-		goto again;
-	}
-	xfs_qm_internalqcheck_get_dquots(mp,
-					(xfs_dqid_t) ip->i_d.di_uid,
-					(xfs_dqid_t) xfs_get_projid(ip),
-					(xfs_dqid_t) ip->i_d.di_gid,
-					&ud, &gd);
-	if (XFS_IS_UQUOTA_ON(mp)) {
-		ASSERT(ud);
-		xfs_qm_internalqcheck_dqadjust(ip, ud);
-	}
-	if (XFS_IS_OQUOTA_ON(mp)) {
-		ASSERT(gd);
-		xfs_qm_internalqcheck_dqadjust(ip, gd);
-	}
-	xfs_iunlock(ip, lock_flags);
-	IRELE(ip);
-	*res = BULKSTAT_RV_DIDONE;
-	return (0);
-}
-
-
-/* PRIVATE, debugging */
-int
-xfs_qm_internalqcheck(
-	xfs_mount_t	*mp)
-{
-	xfs_ino_t	lastino;
-	int		done, count;
-	int		i;
-	int		error;
-
-	lastino = 0;
-	qmtest_hashmask = 32;
-	count = 5;
-	done = 0;
-	qmtest_nfails = 0;
-
-	if (! XFS_IS_QUOTA_ON(mp))
-		return XFS_ERROR(ESRCH);
-
-	xfs_log_force(mp, XFS_LOG_SYNC);
-	XFS_bflush(mp->m_ddev_targp);
-	xfs_log_force(mp, XFS_LOG_SYNC);
-	XFS_bflush(mp->m_ddev_targp);
-
-	mutex_lock(&qcheck_lock);
-	/* There should be absolutely no quota activity while this
-	   is going on. */
-	qmtest_udqtab = kmem_zalloc(qmtest_hashmask *
-				    sizeof(xfs_dqhash_t), KM_SLEEP);
-	qmtest_gdqtab = kmem_zalloc(qmtest_hashmask *
-				    sizeof(xfs_dqhash_t), KM_SLEEP);
-	do {
-		/*
-		 * Iterate thru all the inodes in the file system,
-		 * adjusting the corresponding dquot counters
-		 */
-		error = xfs_bulkstat(mp, &lastino, &count,
-				 xfs_qm_internalqcheck_adjust,
-				 0, NULL, &done);
-		if (error) {
-			xfs_debug(mp, "Bulkstat returned error 0x%x", error);
-			break;
-		}
-	} while (!done);
-
-	xfs_debug(mp, "Checking results against system dquots");
-	for (i = 0; i < qmtest_hashmask; i++) {
-		xfs_dqtest_t	*d, *n;
-		xfs_dqhash_t	*h;
-
-		h = &qmtest_udqtab[i];
-		list_for_each_entry_safe(d, n, &h->qh_list, q_hashlist) {
-			xfs_dqtest_cmp(d);
-			kmem_free(d);
-		}
-		h = &qmtest_gdqtab[i];
-		list_for_each_entry_safe(d, n, &h->qh_list, q_hashlist) {
-			xfs_dqtest_cmp(d);
-			kmem_free(d);
-		}
-	}
-
-	if (qmtest_nfails) {
-		xfs_debug(mp, "******** quotacheck failed  ********");
-		xfs_debug(mp, "failures = %d", qmtest_nfails);
-	} else {
-		xfs_debug(mp, "******** quotacheck successful! ********");
-	}
-	kmem_free(qmtest_udqtab);
-	kmem_free(qmtest_gdqtab);
-	mutex_unlock(&qcheck_lock);
-	return (qmtest_nfails);
-}
-
-#endif /* DEBUG */
diff --git a/fs/xfs/quota/xfs_trans_dquot.c b/fs/xfs/quota/xfs_trans_dquot.c
index 2a364873133..4d00ee67792 100644
--- a/fs/xfs/quota/xfs_trans_dquot.c
+++ b/fs/xfs/quota/xfs_trans_dquot.c
@@ -59,7 +59,7 @@ xfs_trans_dqjoin(
 	xfs_trans_add_item(tp, &dqp->q_logitem.qli_item);
 
 	/*
-	 * Initialize i_transp so we can later determine if this dquot is
+	 * Initialize d_transp so we can later determine if this dquot is
 	 * associated with this transaction.
 	 */
 	dqp->q_transp = tp;
@@ -387,18 +387,18 @@ xfs_trans_apply_dquot_deltas(
 				qtrx->qt_delbcnt_delta;
 			totalrtbdelta = qtrx->qt_rtbcount_delta +
 				qtrx->qt_delrtb_delta;
-#ifdef QUOTADEBUG
+#ifdef DEBUG
 			if (totalbdelta < 0)
 				ASSERT(be64_to_cpu(d->d_bcount) >=
-				       (xfs_qcnt_t) -totalbdelta);
+				       -totalbdelta);
 
 			if (totalrtbdelta < 0)
 				ASSERT(be64_to_cpu(d->d_rtbcount) >=
-				       (xfs_qcnt_t) -totalrtbdelta);
+				       -totalrtbdelta);
 
 			if (qtrx->qt_icount_delta < 0)
 				ASSERT(be64_to_cpu(d->d_icount) >=
-				       (xfs_qcnt_t) -qtrx->qt_icount_delta);
+				       -qtrx->qt_icount_delta);
 #endif
 			if (totalbdelta)
 				be64_add_cpu(&d->d_bcount, (xfs_qcnt_t)totalbdelta);
@@ -642,11 +642,6 @@ xfs_trans_dqresv(
 	    ((XFS_IS_UQUOTA_ENFORCED(dqp->q_mount) && XFS_QM_ISUDQ(dqp)) ||
 	     (XFS_IS_OQUOTA_ENFORCED(dqp->q_mount) &&
 	      (XFS_QM_ISPDQ(dqp) || XFS_QM_ISGDQ(dqp))))) {
-#ifdef QUOTADEBUG
-		xfs_debug(mp,
-			"BLK Res: nblks=%ld + resbcount=%Ld > hardlimit=%Ld?",
-			nblks, *resbcountp, hardlimit);
-#endif
 		if (nblks > 0) {
 			/*
 			 * dquot is locked already. See if we'd go over the
diff --git a/fs/xfs/xfs.h b/fs/xfs/xfs.h
index 5ad8ad3a1dc..53ec3ea9a62 100644
--- a/fs/xfs/xfs.h
+++ b/fs/xfs/xfs.h
@@ -22,7 +22,6 @@
 #define STATIC
 #define DEBUG 1
 #define XFS_BUF_LOCK_TRACKING 1
-/* #define QUOTADEBUG 1 */
 #endif
 
 #include <linux-2.6/xfs_linux.h>
diff --git a/fs/xfs/xfs_acl.h b/fs/xfs/xfs_acl.h
index 11dd72070cb..0135e2a669d 100644
--- a/fs/xfs/xfs_acl.h
+++ b/fs/xfs/xfs_acl.h
@@ -42,7 +42,7 @@ struct xfs_acl {
 #define SGI_ACL_DEFAULT_SIZE	(sizeof(SGI_ACL_DEFAULT)-1)
 
 #ifdef CONFIG_XFS_POSIX_ACL
-extern int xfs_check_acl(struct inode *inode, int mask, unsigned int flags);
+extern int xfs_check_acl(struct inode *inode, int mask);
 extern struct posix_acl *xfs_get_acl(struct inode *inode, int type);
 extern int xfs_inherit_acl(struct inode *inode, struct posix_acl *default_acl);
 extern int xfs_acl_chmod(struct inode *inode);
diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 95862bbff56..1e00b3ef627 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -570,9 +570,7 @@ xfs_alloc_ag_vextent_exact(
 	xfs_agblock_t	tbno;	/* start block of trimmed extent */
 	xfs_extlen_t	tlen;	/* length of trimmed extent */
 	xfs_agblock_t	tend;	/* end block of trimmed extent */
-	xfs_agblock_t	end;	/* end of allocated extent */
 	int		i;	/* success/failure of operation */
-	xfs_extlen_t	rlen;	/* length of returned extent */
 
 	ASSERT(args->alignment == 1);
 
@@ -625,18 +623,16 @@ xfs_alloc_ag_vextent_exact(
 	 *
 	 * Fix the length according to mod and prod if given.
 	 */
-	end = XFS_AGBLOCK_MIN(tend, args->agbno + args->maxlen);
-	args->len = end - args->agbno;
+	args->len = XFS_AGBLOCK_MIN(tend, args->agbno + args->maxlen)
+						- args->agbno;
 	xfs_alloc_fix_len(args);
 	if (!xfs_alloc_fix_minleft(args))
 		goto not_found;
 
-	rlen = args->len;
-	ASSERT(args->agbno + rlen <= tend);
-	end = args->agbno + rlen;
+	ASSERT(args->agbno + args->len <= tend);
 
 	/*
-	 * We are allocating agbno for rlen [agbno .. end]
+	 * We are allocating agbno for args->len
 	 * Allocate/initialize a cursor for the by-size btree.
 	 */
 	cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp,
@@ -2127,7 +2123,7 @@ xfs_read_agf(
 	 * Validate the magic number of the agf block.
 	 */
 	agf_ok =
-		be32_to_cpu(agf->agf_magicnum) == XFS_AGF_MAGIC &&
+		agf->agf_magicnum == cpu_to_be32(XFS_AGF_MAGIC) &&
 		XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) &&
 		be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) &&
 		be32_to_cpu(agf->agf_flfirst) < XFS_AGFL_SIZE(mp) &&
diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c
index 2b3518826a6..ffb3386e45c 100644
--- a/fs/xfs/xfs_alloc_btree.c
+++ b/fs/xfs/xfs_alloc_btree.c
@@ -31,7 +31,6 @@
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_btree.h"
-#include "xfs_btree_trace.h"
 #include "xfs_alloc.h"
 #include "xfs_error.h"
 #include "xfs_trace.h"
@@ -311,72 +310,6 @@ xfs_allocbt_recs_inorder(
 }
 #endif	/* DEBUG */
 
-#ifdef XFS_BTREE_TRACE
-ktrace_t	*xfs_allocbt_trace_buf;
-
-STATIC void
-xfs_allocbt_trace_enter(
-	struct xfs_btree_cur	*cur,
-	const char		*func,
-	char			*s,
-	int			type,
-	int			line,
-	__psunsigned_t		a0,
-	__psunsigned_t		a1,
-	__psunsigned_t		a2,
-	__psunsigned_t		a3,
-	__psunsigned_t		a4,
-	__psunsigned_t		a5,
-	__psunsigned_t		a6,
-	__psunsigned_t		a7,
-	__psunsigned_t		a8,
-	__psunsigned_t		a9,
-	__psunsigned_t		a10)
-{
-	ktrace_enter(xfs_allocbt_trace_buf, (void *)(__psint_t)type,
-		(void *)func, (void *)s, NULL, (void *)cur,
-		(void *)a0, (void *)a1, (void *)a2, (void *)a3,
-		(void *)a4, (void *)a5, (void *)a6, (void *)a7,
-		(void *)a8, (void *)a9, (void *)a10);
-}
-
-STATIC void
-xfs_allocbt_trace_cursor(
-	struct xfs_btree_cur	*cur,
-	__uint32_t		*s0,
-	__uint64_t		*l0,
-	__uint64_t		*l1)
-{
-	*s0 = cur->bc_private.a.agno;
-	*l0 = cur->bc_rec.a.ar_startblock;
-	*l1 = cur->bc_rec.a.ar_blockcount;
-}
-
-STATIC void
-xfs_allocbt_trace_key(
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_key	*key,
-	__uint64_t		*l0,
-	__uint64_t		*l1)
-{
-	*l0 = be32_to_cpu(key->alloc.ar_startblock);
-	*l1 = be32_to_cpu(key->alloc.ar_blockcount);
-}
-
-STATIC void
-xfs_allocbt_trace_record(
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_rec	*rec,
-	__uint64_t		*l0,
-	__uint64_t		*l1,
-	__uint64_t		*l2)
-{
-	*l0 = be32_to_cpu(rec->alloc.ar_startblock);
-	*l1 = be32_to_cpu(rec->alloc.ar_blockcount);
-	*l2 = 0;
-}
-#endif /* XFS_BTREE_TRACE */
-
 static const struct xfs_btree_ops xfs_allocbt_ops = {
 	.rec_len		= sizeof(xfs_alloc_rec_t),
 	.key_len		= sizeof(xfs_alloc_key_t),
@@ -393,18 +326,10 @@ static const struct xfs_btree_ops xfs_allocbt_ops = {
 	.init_rec_from_cur	= xfs_allocbt_init_rec_from_cur,
 	.init_ptr_from_cur	= xfs_allocbt_init_ptr_from_cur,
 	.key_diff		= xfs_allocbt_key_diff,
-
 #ifdef DEBUG
 	.keys_inorder		= xfs_allocbt_keys_inorder,
 	.recs_inorder		= xfs_allocbt_recs_inorder,
 #endif
-
-#ifdef XFS_BTREE_TRACE
-	.trace_enter		= xfs_allocbt_trace_enter,
-	.trace_cursor		= xfs_allocbt_trace_cursor,
-	.trace_key		= xfs_allocbt_trace_key,
-	.trace_record		= xfs_allocbt_trace_record,
-#endif
 };
 
 /*
@@ -427,13 +352,16 @@ xfs_allocbt_init_cursor(
 
 	cur->bc_tp = tp;
 	cur->bc_mp = mp;
-	cur->bc_nlevels = be32_to_cpu(agf->agf_levels[btnum]);
 	cur->bc_btnum = btnum;
 	cur->bc_blocklog = mp->m_sb.sb_blocklog;
-
 	cur->bc_ops = &xfs_allocbt_ops;
-	if (btnum == XFS_BTNUM_CNT)
+
+	if (btnum == XFS_BTNUM_CNT) {
+		cur->bc_nlevels = be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNT]);
 		cur->bc_flags = XFS_BTREE_LASTREC_UPDATE;
+	} else {
+		cur->bc_nlevels = be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNO]);
+	}
 
 	cur->bc_private.a.agbp = agbp;
 	cur->bc_private.a.agno = agno;
diff --git a/fs/xfs/xfs_arch.h b/fs/xfs/xfs_arch.h
deleted file mode 100644
index 0902249354a..00000000000
--- a/fs/xfs/xfs_arch.h
+++ /dev/null
@@ -1,136 +0,0 @@
-/*
- * Copyright (c) 2000-2002,2005 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_ARCH_H__
-#define __XFS_ARCH_H__
-
-#ifndef XFS_BIG_INUMS
-# error XFS_BIG_INUMS must be defined true or false
-#endif
-
-#ifdef __KERNEL__
-
-#include <asm/byteorder.h>
-
-#ifdef __BIG_ENDIAN
-#define	XFS_NATIVE_HOST	1
-#else
-#undef XFS_NATIVE_HOST
-#endif
-
-#else /* __KERNEL__ */
-
-#if __BYTE_ORDER == __BIG_ENDIAN
-#define	XFS_NATIVE_HOST	1
-#else
-#undef XFS_NATIVE_HOST
-#endif
-
-#ifdef XFS_NATIVE_HOST
-#define cpu_to_be16(val)	((__force __be16)(__u16)(val))
-#define cpu_to_be32(val)	((__force __be32)(__u32)(val))
-#define cpu_to_be64(val)	((__force __be64)(__u64)(val))
-#define be16_to_cpu(val)	((__force __u16)(__be16)(val))
-#define be32_to_cpu(val)	((__force __u32)(__be32)(val))
-#define be64_to_cpu(val)	((__force __u64)(__be64)(val))
-#else
-#define cpu_to_be16(val)	((__force __be16)__swab16((__u16)(val)))
-#define cpu_to_be32(val)	((__force __be32)__swab32((__u32)(val)))
-#define cpu_to_be64(val)	((__force __be64)__swab64((__u64)(val)))
-#define be16_to_cpu(val)	(__swab16((__force __u16)(__be16)(val)))
-#define be32_to_cpu(val)	(__swab32((__force __u32)(__be32)(val)))
-#define be64_to_cpu(val)	(__swab64((__force __u64)(__be64)(val)))
-#endif
-
-static inline void be16_add_cpu(__be16 *a, __s16 b)
-{
-	*a = cpu_to_be16(be16_to_cpu(*a) + b);
-}
-
-static inline void be32_add_cpu(__be32 *a, __s32 b)
-{
-	*a = cpu_to_be32(be32_to_cpu(*a) + b);
-}
-
-static inline void be64_add_cpu(__be64 *a, __s64 b)
-{
-	*a = cpu_to_be64(be64_to_cpu(*a) + b);
-}
-
-#endif	/* __KERNEL__ */
-
-/*
- * get and set integers from potentially unaligned locations
- */
-
-#define INT_GET_UNALIGNED_16_BE(pointer) \
-   ((__u16)((((__u8*)(pointer))[0] << 8) | (((__u8*)(pointer))[1])))
-#define INT_SET_UNALIGNED_16_BE(pointer,value) \
-    { \
-	((__u8*)(pointer))[0] = (((value) >> 8) & 0xff); \
-	((__u8*)(pointer))[1] = (((value)     ) & 0xff); \
-    }
-
-/*
- * In directories inode numbers are stored as unaligned arrays of unsigned
- * 8bit integers on disk.
- *
- * For v1 directories or v2 directories that contain inode numbers that
- * do not fit into 32bit the array has eight members, but the first member
- * is always zero:
- *
- *  |unused|48-55|40-47|32-39|24-31|16-23| 8-15| 0- 7|
- *
- * For v2 directories that only contain entries with inode numbers that fit
- * into 32bits a four-member array is used:
- *
- *  |24-31|16-23| 8-15| 0- 7|
- */ 
-
-#define XFS_GET_DIR_INO4(di) \
-	(((__u32)(di).i[0] << 24) | ((di).i[1] << 16) | ((di).i[2] << 8) | ((di).i[3]))
-
-#define XFS_PUT_DIR_INO4(from, di) \
-do { \
-	(di).i[0] = (((from) & 0xff000000ULL) >> 24); \
-	(di).i[1] = (((from) & 0x00ff0000ULL) >> 16); \
-	(di).i[2] = (((from) & 0x0000ff00ULL) >> 8); \
-	(di).i[3] = ((from) & 0x000000ffULL); \
-} while (0)
-
-#define XFS_DI_HI(di) \
-	(((__u32)(di).i[1] << 16) | ((di).i[2] << 8) | ((di).i[3]))
-#define XFS_DI_LO(di) \
-	(((__u32)(di).i[4] << 24) | ((di).i[5] << 16) | ((di).i[6] << 8) | ((di).i[7]))
-
-#define XFS_GET_DIR_INO8(di)        \
-	(((xfs_ino_t)XFS_DI_LO(di) & 0xffffffffULL) | \
-	 ((xfs_ino_t)XFS_DI_HI(di) << 32))
-
-#define XFS_PUT_DIR_INO8(from, di) \
-do { \
-	(di).i[0] = 0; \
-	(di).i[1] = (((from) & 0x00ff000000000000ULL) >> 48); \
-	(di).i[2] = (((from) & 0x0000ff0000000000ULL) >> 40); \
-	(di).i[3] = (((from) & 0x000000ff00000000ULL) >> 32); \
-	(di).i[4] = (((from) & 0x00000000ff000000ULL) >> 24); \
-	(di).i[5] = (((from) & 0x0000000000ff0000ULL) >> 16); \
-	(di).i[6] = (((from) & 0x000000000000ff00ULL) >> 8); \
-	(di).i[7] = ((from) & 0x00000000000000ffULL); \
-} while (0)
-	
-#endif	/* __XFS_ARCH_H__ */
diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c
index c8637537881..cbae424fe1b 100644
--- a/fs/xfs/xfs_attr.c
+++ b/fs/xfs/xfs_attr.c
@@ -490,6 +490,13 @@ xfs_attr_remove_int(xfs_inode_t *dp, struct xfs_name *name, int flags)
 	args.whichfork = XFS_ATTR_FORK;
 
 	/*
+	 * we have no control over the attribute names that userspace passes us
+	 * to remove, so we have to allow the name lookup prior to attribute
+	 * removal to fail.
+	 */
+	args.op_flags = XFS_DA_OP_OKNOENT;
+
+	/*
 	 * Attach the dquots to the inode.
 	 */
 	error = xfs_qm_dqattach(dp, 0);
@@ -815,17 +822,21 @@ xfs_attr_inactive(xfs_inode_t *dp)
 	error = xfs_attr_root_inactive(&trans, dp);
 	if (error)
 		goto out;
+
 	/*
-	 * signal synchronous inactive transactions unless this
-	 * is a synchronous mount filesystem in which case we
-	 * know that we're here because we've been called out of
-	 * xfs_inactive which means that the last reference is gone
-	 * and the unlink transaction has already hit the disk so
-	 * async inactive transactions are safe.
+	 * Signal synchronous inactive transactions unless this is a
+	 * synchronous mount filesystem in which case we know that we're here
+	 * because we've been called out of xfs_inactive which means that the
+	 * last reference is gone and the unlink transaction has already hit
+	 * the disk so async inactive transactions are safe.
 	 */
-	if ((error = xfs_itruncate_finish(&trans, dp, 0LL, XFS_ATTR_FORK,
-				(!(mp->m_flags & XFS_MOUNT_WSYNC)
-				 ? 1 : 0))))
+	if (!(mp->m_flags & XFS_MOUNT_WSYNC)) {
+		if (dp->i_d.di_anextents > 0)
+			xfs_trans_set_sync(trans);
+	}
+
+	error = xfs_itruncate_extents(&trans, dp, XFS_ATTR_FORK, 0);
+	if (error)
 		goto out;
 
 	/*
@@ -1192,7 +1203,7 @@ xfs_attr_leaf_list(xfs_attr_list_context_t *context)
 		return XFS_ERROR(error);
 	ASSERT(bp != NULL);
 	leaf = bp->data;
-	if (unlikely(be16_to_cpu(leaf->hdr.info.magic) != XFS_ATTR_LEAF_MAGIC)) {
+	if (unlikely(leaf->hdr.info.magic != cpu_to_be16(XFS_ATTR_LEAF_MAGIC))) {
 		XFS_CORRUPTION_ERROR("xfs_attr_leaf_list", XFS_ERRLEVEL_LOW,
 				     context->dp->i_mount, leaf);
 		xfs_da_brelse(NULL, bp);
@@ -1599,9 +1610,8 @@ xfs_attr_node_removename(xfs_da_args_t *args)
 						     XFS_ATTR_FORK);
 		if (error)
 			goto out;
-		ASSERT(be16_to_cpu(((xfs_attr_leafblock_t *)
-				      bp->data)->hdr.info.magic)
-						       == XFS_ATTR_LEAF_MAGIC);
+		ASSERT((((xfs_attr_leafblock_t *)bp->data)->hdr.info.magic) ==
+		       cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 
 		if ((forkoff = xfs_attr_shortform_allfit(bp, dp))) {
 			xfs_bmap_init(args->flist, args->firstblock);
@@ -1866,11 +1876,11 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
 				return(XFS_ERROR(EFSCORRUPTED));
 			}
 			node = bp->data;
-			if (be16_to_cpu(node->hdr.info.magic)
-							== XFS_ATTR_LEAF_MAGIC)
+			if (node->hdr.info.magic ==
+			    cpu_to_be16(XFS_ATTR_LEAF_MAGIC))
 				break;
-			if (unlikely(be16_to_cpu(node->hdr.info.magic)
-							!= XFS_DA_NODE_MAGIC)) {
+			if (unlikely(node->hdr.info.magic !=
+				     cpu_to_be16(XFS_DA_NODE_MAGIC))) {
 				XFS_CORRUPTION_ERROR("xfs_attr_node_list(3)",
 						     XFS_ERRLEVEL_LOW,
 						     context->dp->i_mount,
@@ -1905,8 +1915,8 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
 	 */
 	for (;;) {
 		leaf = bp->data;
-		if (unlikely(be16_to_cpu(leaf->hdr.info.magic)
-						!= XFS_ATTR_LEAF_MAGIC)) {
+		if (unlikely(leaf->hdr.info.magic !=
+			     cpu_to_be16(XFS_ATTR_LEAF_MAGIC))) {
 			XFS_CORRUPTION_ERROR("xfs_attr_node_list(4)",
 					     XFS_ERRLEVEL_LOW,
 					     context->dp->i_mount, leaf);
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index 71e90dc2aeb..8fad9602542 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -731,7 +731,7 @@ xfs_attr_shortform_allfit(xfs_dabuf_t *bp, xfs_inode_t *dp)
 	int bytes, i;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 
 	entry = &leaf->entries[0];
 	bytes = sizeof(struct xfs_attr_sf_hdr);
@@ -777,7 +777,7 @@ xfs_attr_leaf_to_shortform(xfs_dabuf_t *bp, xfs_da_args_t *args, int forkoff)
 	ASSERT(bp != NULL);
 	memcpy(tmpbuffer, bp->data, XFS_LBSIZE(dp->i_mount));
 	leaf = (xfs_attr_leafblock_t *)tmpbuffer;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	memset(bp->data, 0, XFS_LBSIZE(dp->i_mount));
 
 	/*
@@ -872,7 +872,7 @@ xfs_attr_leaf_to_node(xfs_da_args_t *args)
 		goto out;
 	node = bp1->data;
 	leaf = bp2->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	/* both on-disk, don't endian-flip twice */
 	node->btree[0].hashval =
 		leaf->entries[be16_to_cpu(leaf->hdr.count)-1 ].hashval;
@@ -997,7 +997,7 @@ xfs_attr_leaf_add(xfs_dabuf_t *bp, xfs_da_args_t *args)
 	int tablesize, entsize, sum, tmp, i;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	ASSERT((args->index >= 0)
 		&& (args->index <= be16_to_cpu(leaf->hdr.count)));
 	hdr = &leaf->hdr;
@@ -1070,7 +1070,7 @@ xfs_attr_leaf_add_work(xfs_dabuf_t *bp, xfs_da_args_t *args, int mapindex)
 	int tmp, i;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	hdr = &leaf->hdr;
 	ASSERT((mapindex >= 0) && (mapindex < XFS_ATTR_LEAF_MAPSIZE));
 	ASSERT((args->index >= 0) && (args->index <= be16_to_cpu(hdr->count)));
@@ -1256,8 +1256,8 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
 	ASSERT(blk2->magic == XFS_ATTR_LEAF_MAGIC);
 	leaf1 = blk1->bp->data;
 	leaf2 = blk2->bp->data;
-	ASSERT(be16_to_cpu(leaf1->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
-	ASSERT(be16_to_cpu(leaf2->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf1->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
+	ASSERT(leaf2->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	args = state->args;
 
 	/*
@@ -1533,7 +1533,7 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
 	 */
 	blk = &state->path.blk[ state->path.active-1 ];
 	info = blk->bp->data;
-	ASSERT(be16_to_cpu(info->magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	leaf = (xfs_attr_leafblock_t *)info;
 	count = be16_to_cpu(leaf->hdr.count);
 	bytes = sizeof(xfs_attr_leaf_hdr_t) +
@@ -1596,7 +1596,7 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
 		bytes  = state->blocksize - (state->blocksize>>2);
 		bytes -= be16_to_cpu(leaf->hdr.usedbytes);
 		leaf = bp->data;
-		ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+		ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 		count += be16_to_cpu(leaf->hdr.count);
 		bytes -= be16_to_cpu(leaf->hdr.usedbytes);
 		bytes -= count * sizeof(xfs_attr_leaf_entry_t);
@@ -1650,7 +1650,7 @@ xfs_attr_leaf_remove(xfs_dabuf_t *bp, xfs_da_args_t *args)
 	xfs_mount_t *mp;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	hdr = &leaf->hdr;
 	mp = args->trans->t_mountp;
 	ASSERT((be16_to_cpu(hdr->count) > 0)
@@ -1813,8 +1813,8 @@ xfs_attr_leaf_unbalance(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
 	ASSERT(save_blk->magic == XFS_ATTR_LEAF_MAGIC);
 	drop_leaf = drop_blk->bp->data;
 	save_leaf = save_blk->bp->data;
-	ASSERT(be16_to_cpu(drop_leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
-	ASSERT(be16_to_cpu(save_leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(drop_leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
+	ASSERT(save_leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	drop_hdr = &drop_leaf->hdr;
 	save_hdr = &save_leaf->hdr;
 
@@ -1915,7 +1915,7 @@ xfs_attr_leaf_lookup_int(xfs_dabuf_t *bp, xfs_da_args_t *args)
 	xfs_dahash_t hashval;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	ASSERT(be16_to_cpu(leaf->hdr.count)
 					< (XFS_LBSIZE(args->dp->i_mount)/8));
 
@@ -2019,7 +2019,7 @@ xfs_attr_leaf_getvalue(xfs_dabuf_t *bp, xfs_da_args_t *args)
 	xfs_attr_leaf_name_remote_t *name_rmt;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	ASSERT(be16_to_cpu(leaf->hdr.count)
 					< (XFS_LBSIZE(args->dp->i_mount)/8));
 	ASSERT(args->index < be16_to_cpu(leaf->hdr.count));
@@ -2087,8 +2087,8 @@ xfs_attr_leaf_moveents(xfs_attr_leafblock_t *leaf_s, int start_s,
 	/*
 	 * Set up environment.
 	 */
-	ASSERT(be16_to_cpu(leaf_s->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
-	ASSERT(be16_to_cpu(leaf_d->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf_s->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
+	ASSERT(leaf_d->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	hdr_s = &leaf_s->hdr;
 	hdr_d = &leaf_d->hdr;
 	ASSERT((be16_to_cpu(hdr_s->count) > 0) &&
@@ -2222,8 +2222,8 @@ xfs_attr_leaf_order(xfs_dabuf_t *leaf1_bp, xfs_dabuf_t *leaf2_bp)
 
 	leaf1 = leaf1_bp->data;
 	leaf2 = leaf2_bp->data;
-	ASSERT((be16_to_cpu(leaf1->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC) &&
-	       (be16_to_cpu(leaf2->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC));
+	ASSERT((leaf1->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)) &&
+	       (leaf2->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)));
 	if ((be16_to_cpu(leaf1->hdr.count) > 0) &&
 	    (be16_to_cpu(leaf2->hdr.count) > 0) &&
 	    ((be32_to_cpu(leaf2->entries[0].hashval) <
@@ -2246,7 +2246,7 @@ xfs_attr_leaf_lasthash(xfs_dabuf_t *bp, int *count)
 	xfs_attr_leafblock_t *leaf;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	if (count)
 		*count = be16_to_cpu(leaf->hdr.count);
 	if (!leaf->hdr.count)
@@ -2265,7 +2265,7 @@ xfs_attr_leaf_entsize(xfs_attr_leafblock_t *leaf, int index)
 	xfs_attr_leaf_name_remote_t *name_rmt;
 	int size;
 
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	if (leaf->entries[index].flags & XFS_ATTR_LOCAL) {
 		name_loc = xfs_attr_leaf_name_local(leaf, index);
 		size = xfs_attr_leaf_entsize_local(name_loc->namelen,
@@ -2451,7 +2451,7 @@ xfs_attr_leaf_clearflag(xfs_da_args_t *args)
 	ASSERT(bp != NULL);
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	ASSERT(args->index < be16_to_cpu(leaf->hdr.count));
 	ASSERT(args->index >= 0);
 	entry = &leaf->entries[ args->index ];
@@ -2515,7 +2515,7 @@ xfs_attr_leaf_setflag(xfs_da_args_t *args)
 	ASSERT(bp != NULL);
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	ASSERT(args->index < be16_to_cpu(leaf->hdr.count));
 	ASSERT(args->index >= 0);
 	entry = &leaf->entries[ args->index ];
@@ -2585,13 +2585,13 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
 	}
 
 	leaf1 = bp1->data;
-	ASSERT(be16_to_cpu(leaf1->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf1->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	ASSERT(args->index < be16_to_cpu(leaf1->hdr.count));
 	ASSERT(args->index >= 0);
 	entry1 = &leaf1->entries[ args->index ];
 
 	leaf2 = bp2->data;
-	ASSERT(be16_to_cpu(leaf2->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf2->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	ASSERT(args->index2 < be16_to_cpu(leaf2->hdr.count));
 	ASSERT(args->index2 >= 0);
 	entry2 = &leaf2->entries[ args->index2 ];
@@ -2689,9 +2689,9 @@ xfs_attr_root_inactive(xfs_trans_t **trans, xfs_inode_t *dp)
 	 * This is a depth-first traversal!
 	 */
 	info = bp->data;
-	if (be16_to_cpu(info->magic) == XFS_DA_NODE_MAGIC) {
+	if (info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC)) {
 		error = xfs_attr_node_inactive(trans, dp, bp, 1);
-	} else if (be16_to_cpu(info->magic) == XFS_ATTR_LEAF_MAGIC) {
+	} else if (info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)) {
 		error = xfs_attr_leaf_inactive(trans, dp, bp);
 	} else {
 		error = XFS_ERROR(EIO);
@@ -2739,7 +2739,7 @@ xfs_attr_node_inactive(xfs_trans_t **trans, xfs_inode_t *dp, xfs_dabuf_t *bp,
 	}
 
 	node = bp->data;
-	ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	parent_blkno = xfs_da_blkno(bp);	/* save for re-read later */
 	count = be16_to_cpu(node->hdr.count);
 	if (!count) {
@@ -2773,10 +2773,10 @@ xfs_attr_node_inactive(xfs_trans_t **trans, xfs_inode_t *dp, xfs_dabuf_t *bp,
 			 * Invalidate the subtree, however we have to.
 			 */
 			info = child_bp->data;
-			if (be16_to_cpu(info->magic) == XFS_DA_NODE_MAGIC) {
+			if (info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC)) {
 				error = xfs_attr_node_inactive(trans, dp,
 						child_bp, level+1);
-			} else if (be16_to_cpu(info->magic) == XFS_ATTR_LEAF_MAGIC) {
+			} else if (info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)) {
 				error = xfs_attr_leaf_inactive(trans, dp,
 						child_bp);
 			} else {
@@ -2836,7 +2836,7 @@ xfs_attr_leaf_inactive(xfs_trans_t **trans, xfs_inode_t *dp, xfs_dabuf_t *bp)
 	int error, count, size, tmp, i;
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 
 	/*
 	 * Count the number of "remote" value extents.
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index e546a33214c..c51a3f90363 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -29,15 +29,11 @@
 #include "xfs_bmap_btree.h"
 #include "xfs_alloc_btree.h"
 #include "xfs_ialloc_btree.h"
-#include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_btree.h"
 #include "xfs_mount.h"
 #include "xfs_itable.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
 #include "xfs_inode_item.h"
 #include "xfs_extfree_item.h"
 #include "xfs_alloc.h"
@@ -94,6 +90,7 @@ xfs_bmap_add_attrfork_local(
  */
 STATIC int				/* error */
 xfs_bmap_add_extent_delay_real(
+	struct xfs_trans	*tp,	/* transaction pointer */
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	xfs_extnum_t		*idx,	/* extent number to update/insert */
 	xfs_btree_cur_t		**curp,	/* if *curp is null, not a btree */
@@ -439,6 +436,7 @@ xfs_bmap_add_attrfork_local(
  */
 STATIC int				/* error */
 xfs_bmap_add_extent(
+	struct xfs_trans	*tp,	/* transaction pointer */
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	xfs_extnum_t		*idx,	/* extent number to update/insert */
 	xfs_btree_cur_t		**curp,	/* if *curp is null, not a btree */
@@ -524,7 +522,7 @@ xfs_bmap_add_extent(
 				if (cur)
 					ASSERT(cur->bc_private.b.flags &
 						XFS_BTCUR_BPRV_WASDEL);
-				error = xfs_bmap_add_extent_delay_real(ip,
+				error = xfs_bmap_add_extent_delay_real(tp, ip,
 						idx, &cur, new, &da_new,
 						first, flist, &logflags);
 			} else {
@@ -561,7 +559,7 @@ xfs_bmap_add_extent(
 		int	tmp_logflags;	/* partial log flag return val */
 
 		ASSERT(cur == NULL);
-		error = xfs_bmap_extents_to_btree(ip->i_transp, ip, first,
+		error = xfs_bmap_extents_to_btree(tp, ip, first,
 			flist, &cur, da_old > 0, &tmp_logflags, whichfork);
 		logflags |= tmp_logflags;
 		if (error)
@@ -604,6 +602,7 @@ done:
  */
 STATIC int				/* error */
 xfs_bmap_add_extent_delay_real(
+	struct xfs_trans	*tp,	/* transaction pointer */
 	xfs_inode_t		*ip,	/* incore inode pointer */
 	xfs_extnum_t		*idx,	/* extent number to update/insert */
 	xfs_btree_cur_t		**curp,	/* if *curp is null, not a btree */
@@ -901,7 +900,7 @@ xfs_bmap_add_extent_delay_real(
 		}
 		if (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
 		    ip->i_d.di_nextents > ip->i_df.if_ext_max) {
-			error = xfs_bmap_extents_to_btree(ip->i_transp, ip,
+			error = xfs_bmap_extents_to_btree(tp, ip,
 					first, flist, &cur, 1, &tmp_rval,
 					XFS_DATA_FORK);
 			rval |= tmp_rval;
@@ -984,7 +983,7 @@ xfs_bmap_add_extent_delay_real(
 		}
 		if (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
 		    ip->i_d.di_nextents > ip->i_df.if_ext_max) {
-			error = xfs_bmap_extents_to_btree(ip->i_transp, ip,
+			error = xfs_bmap_extents_to_btree(tp, ip,
 				first, flist, &cur, 1, &tmp_rval,
 				XFS_DATA_FORK);
 			rval |= tmp_rval;
@@ -1052,7 +1051,7 @@ xfs_bmap_add_extent_delay_real(
 		}
 		if (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS &&
 		    ip->i_d.di_nextents > ip->i_df.if_ext_max) {
-			error = xfs_bmap_extents_to_btree(ip->i_transp, ip,
+			error = xfs_bmap_extents_to_btree(tp, ip,
 					first, flist, &cur, 1, &tmp_rval,
 					XFS_DATA_FORK);
 			rval |= tmp_rval;
@@ -2871,8 +2870,8 @@ xfs_bmap_del_extent(
 			len = del->br_blockcount;
 			do_div(bno, mp->m_sb.sb_rextsize);
 			do_div(len, mp->m_sb.sb_rextsize);
-			if ((error = xfs_rtfree_extent(ip->i_transp, bno,
-					(xfs_extlen_t)len)))
+			error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
+			if (error)
 				goto done;
 			do_fx = 0;
 			nblks = len * mp->m_sb.sb_rextsize;
@@ -4080,7 +4079,7 @@ xfs_bmap_sanity_check(
 {
 	struct xfs_btree_block  *block = XFS_BUF_TO_BLOCK(bp);
 
-	if (be32_to_cpu(block->bb_magic) != XFS_BMAP_MAGIC ||
+	if (block->bb_magic != cpu_to_be32(XFS_BMAP_MAGIC) ||
 	    be16_to_cpu(block->bb_level) != level ||
 	    be16_to_cpu(block->bb_numrecs) == 0 ||
 	    be16_to_cpu(block->bb_numrecs) > mp->m_bmap_dmxr[level != 0])
@@ -4662,7 +4661,7 @@ xfs_bmapi(
 				if (!wasdelay && (flags & XFS_BMAPI_PREALLOC))
 					got.br_state = XFS_EXT_UNWRITTEN;
 			}
-			error = xfs_bmap_add_extent(ip, &lastx, &cur, &got,
+			error = xfs_bmap_add_extent(tp, ip, &lastx, &cur, &got,
 				firstblock, flist, &tmp_logflags,
 				whichfork);
 			logflags |= tmp_logflags;
@@ -4763,7 +4762,7 @@ xfs_bmapi(
 			mval->br_state = (mval->br_state == XFS_EXT_UNWRITTEN)
 						? XFS_EXT_NORM
 						: XFS_EXT_UNWRITTEN;
-			error = xfs_bmap_add_extent(ip, &lastx, &cur, mval,
+			error = xfs_bmap_add_extent(tp, ip, &lastx, &cur, mval,
 				firstblock, flist, &tmp_logflags,
 				whichfork);
 			logflags |= tmp_logflags;
@@ -5117,7 +5116,7 @@ xfs_bunmapi(
 				del.br_blockcount = mod;
 			}
 			del.br_state = XFS_EXT_UNWRITTEN;
-			error = xfs_bmap_add_extent(ip, &lastx, &cur, &del,
+			error = xfs_bmap_add_extent(tp, ip, &lastx, &cur, &del,
 				firstblock, flist, &logflags,
 				XFS_DATA_FORK);
 			if (error)
@@ -5175,18 +5174,18 @@ xfs_bunmapi(
 				}
 				prev.br_state = XFS_EXT_UNWRITTEN;
 				lastx--;
-				error = xfs_bmap_add_extent(ip, &lastx, &cur,
-					&prev, firstblock, flist, &logflags,
-					XFS_DATA_FORK);
+				error = xfs_bmap_add_extent(tp, ip, &lastx,
+						&cur, &prev, firstblock, flist,
+						&logflags, XFS_DATA_FORK);
 				if (error)
 					goto error0;
 				goto nodelete;
 			} else {
 				ASSERT(del.br_state == XFS_EXT_NORM);
 				del.br_state = XFS_EXT_UNWRITTEN;
-				error = xfs_bmap_add_extent(ip, &lastx, &cur,
-					&del, firstblock, flist, &logflags,
-					XFS_DATA_FORK);
+				error = xfs_bmap_add_extent(tp, ip, &lastx,
+						&cur, &del, firstblock, flist,
+						&logflags, XFS_DATA_FORK);
 				if (error)
 					goto error0;
 				goto nodelete;
diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c
index 87d3c10b695..e2f5d59cbea 100644
--- a/fs/xfs/xfs_bmap_btree.c
+++ b/fs/xfs/xfs_bmap_btree.c
@@ -33,7 +33,6 @@
 #include "xfs_inode_item.h"
 #include "xfs_alloc.h"
 #include "xfs_btree.h"
-#include "xfs_btree_trace.h"
 #include "xfs_itable.h"
 #include "xfs_bmap.h"
 #include "xfs_error.h"
@@ -425,10 +424,10 @@ xfs_bmbt_to_bmdr(
 	xfs_bmbt_key_t		*tkp;
 	__be64			*tpp;
 
-	ASSERT(be32_to_cpu(rblock->bb_magic) == XFS_BMAP_MAGIC);
-	ASSERT(be64_to_cpu(rblock->bb_u.l.bb_leftsib) == NULLDFSBNO);
-	ASSERT(be64_to_cpu(rblock->bb_u.l.bb_rightsib) == NULLDFSBNO);
-	ASSERT(be16_to_cpu(rblock->bb_level) > 0);
+	ASSERT(rblock->bb_magic == cpu_to_be32(XFS_BMAP_MAGIC));
+	ASSERT(rblock->bb_u.l.bb_leftsib == cpu_to_be64(NULLDFSBNO));
+	ASSERT(rblock->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO));
+	ASSERT(rblock->bb_level != 0);
 	dblock->bb_level = rblock->bb_level;
 	dblock->bb_numrecs = rblock->bb_numrecs;
 	dmxr = xfs_bmdr_maxrecs(mp, dblocklen, 0);
@@ -732,95 +731,6 @@ xfs_bmbt_recs_inorder(
 }
 #endif	/* DEBUG */
 
-#ifdef XFS_BTREE_TRACE
-ktrace_t	*xfs_bmbt_trace_buf;
-
-STATIC void
-xfs_bmbt_trace_enter(
-	struct xfs_btree_cur	*cur,
-	const char		*func,
-	char			*s,
-	int			type,
-	int			line,
-	__psunsigned_t		a0,
-	__psunsigned_t		a1,
-	__psunsigned_t		a2,
-	__psunsigned_t		a3,
-	__psunsigned_t		a4,
-	__psunsigned_t		a5,
-	__psunsigned_t		a6,
-	__psunsigned_t		a7,
-	__psunsigned_t		a8,
-	__psunsigned_t		a9,
-	__psunsigned_t		a10)
-{
-	struct xfs_inode	*ip = cur->bc_private.b.ip;
-	int			whichfork = cur->bc_private.b.whichfork;
-
-	ktrace_enter(xfs_bmbt_trace_buf,
-		(void *)((__psint_t)type | (whichfork << 8) | (line << 16)),
-		(void *)func, (void *)s, (void *)ip, (void *)cur,
-		(void *)a0, (void *)a1, (void *)a2, (void *)a3,
-		(void *)a4, (void *)a5, (void *)a6, (void *)a7,
-		(void *)a8, (void *)a9, (void *)a10);
-}
-
-STATIC void
-xfs_bmbt_trace_cursor(
-	struct xfs_btree_cur	*cur,
-	__uint32_t		*s0,
-	__uint64_t		*l0,
-	__uint64_t		*l1)
-{
-	struct xfs_bmbt_rec_host r;
-
-	xfs_bmbt_set_all(&r, &cur->bc_rec.b);
-
-	*s0 = (cur->bc_nlevels << 24) |
-	      (cur->bc_private.b.flags << 16) |
-	       cur->bc_private.b.allocated;
-	*l0 = r.l0;
-	*l1 = r.l1;
-}
-
-STATIC void
-xfs_bmbt_trace_key(
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_key	*key,
-	__uint64_t		*l0,
-	__uint64_t		*l1)
-{
-	*l0 = be64_to_cpu(key->bmbt.br_startoff);
-	*l1 = 0;
-}
-
-/* Endian flipping versions of the bmbt extraction functions */
-STATIC void
-xfs_bmbt_disk_get_all(
-	xfs_bmbt_rec_t	*r,
-	xfs_bmbt_irec_t *s)
-{
-	__xfs_bmbt_get_all(get_unaligned_be64(&r->l0),
-				get_unaligned_be64(&r->l1), s);
-}
-
-STATIC void
-xfs_bmbt_trace_record(
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_rec	*rec,
-	__uint64_t		*l0,
-	__uint64_t		*l1,
-	__uint64_t		*l2)
-{
-	struct xfs_bmbt_irec	irec;
-
-	xfs_bmbt_disk_get_all(&rec->bmbt, &irec);
-	*l0 = irec.br_startoff;
-	*l1 = irec.br_startblock;
-	*l2 = irec.br_blockcount;
-}
-#endif /* XFS_BTREE_TRACE */
-
 static const struct xfs_btree_ops xfs_bmbt_ops = {
 	.rec_len		= sizeof(xfs_bmbt_rec_t),
 	.key_len		= sizeof(xfs_bmbt_key_t),
@@ -837,18 +747,10 @@ static const struct xfs_btree_ops xfs_bmbt_ops = {
 	.init_rec_from_cur	= xfs_bmbt_init_rec_from_cur,
 	.init_ptr_from_cur	= xfs_bmbt_init_ptr_from_cur,
 	.key_diff		= xfs_bmbt_key_diff,
-
 #ifdef DEBUG
 	.keys_inorder		= xfs_bmbt_keys_inorder,
 	.recs_inorder		= xfs_bmbt_recs_inorder,
 #endif
-
-#ifdef XFS_BTREE_TRACE
-	.trace_enter		= xfs_bmbt_trace_enter,
-	.trace_cursor		= xfs_bmbt_trace_cursor,
-	.trace_key		= xfs_bmbt_trace_key,
-	.trace_record		= xfs_bmbt_trace_record,
-#endif
 };
 
 /*
diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c
index 2f9e97c128a..cabf4b5604a 100644
--- a/fs/xfs/xfs_btree.c
+++ b/fs/xfs/xfs_btree.c
@@ -32,7 +32,6 @@
 #include "xfs_inode.h"
 #include "xfs_inode_item.h"
 #include "xfs_btree.h"
-#include "xfs_btree_trace.h"
 #include "xfs_error.h"
 #include "xfs_trace.h"
 
@@ -66,11 +65,11 @@ xfs_btree_check_lblock(
 		be16_to_cpu(block->bb_numrecs) <=
 			cur->bc_ops->get_maxrecs(cur, level) &&
 		block->bb_u.l.bb_leftsib &&
-		(be64_to_cpu(block->bb_u.l.bb_leftsib) == NULLDFSBNO ||
+		(block->bb_u.l.bb_leftsib == cpu_to_be64(NULLDFSBNO) ||
 		 XFS_FSB_SANITY_CHECK(mp,
 		 	be64_to_cpu(block->bb_u.l.bb_leftsib))) &&
 		block->bb_u.l.bb_rightsib &&
-		(be64_to_cpu(block->bb_u.l.bb_rightsib) == NULLDFSBNO ||
+		(block->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO) ||
 		 XFS_FSB_SANITY_CHECK(mp,
 		 	be64_to_cpu(block->bb_u.l.bb_rightsib)));
 	if (unlikely(XFS_TEST_ERROR(!lblock_ok, mp,
@@ -105,10 +104,10 @@ xfs_btree_check_sblock(
 		be16_to_cpu(block->bb_level) == level &&
 		be16_to_cpu(block->bb_numrecs) <=
 			cur->bc_ops->get_maxrecs(cur, level) &&
-		(be32_to_cpu(block->bb_u.s.bb_leftsib) == NULLAGBLOCK ||
+		(block->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK) ||
 		 be32_to_cpu(block->bb_u.s.bb_leftsib) < agflen) &&
 		block->bb_u.s.bb_leftsib &&
-		(be32_to_cpu(block->bb_u.s.bb_rightsib) == NULLAGBLOCK ||
+		(block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) ||
 		 be32_to_cpu(block->bb_u.s.bb_rightsib) < agflen) &&
 		block->bb_u.s.bb_rightsib;
 	if (unlikely(XFS_TEST_ERROR(!sblock_ok, cur->bc_mp,
@@ -511,9 +510,9 @@ xfs_btree_islastblock(
 	block = xfs_btree_get_block(cur, level, &bp);
 	xfs_btree_check_block(cur, block, level, bp);
 	if (cur->bc_flags & XFS_BTREE_LONG_PTRS)
-		return be64_to_cpu(block->bb_u.l.bb_rightsib) == NULLDFSBNO;
+		return block->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO);
 	else
-		return be32_to_cpu(block->bb_u.s.bb_rightsib) == NULLAGBLOCK;
+		return block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK);
 }
 
 /*
@@ -777,14 +776,14 @@ xfs_btree_setbuf(
 
 	b = XFS_BUF_TO_BLOCK(bp);
 	if (cur->bc_flags & XFS_BTREE_LONG_PTRS) {
-		if (be64_to_cpu(b->bb_u.l.bb_leftsib) == NULLDFSBNO)
+		if (b->bb_u.l.bb_leftsib == cpu_to_be64(NULLDFSBNO))
 			cur->bc_ra[lev] |= XFS_BTCUR_LEFTRA;
-		if (be64_to_cpu(b->bb_u.l.bb_rightsib) == NULLDFSBNO)
+		if (b->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO))
 			cur->bc_ra[lev] |= XFS_BTCUR_RIGHTRA;
 	} else {
-		if (be32_to_cpu(b->bb_u.s.bb_leftsib) == NULLAGBLOCK)
+		if (b->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK))
 			cur->bc_ra[lev] |= XFS_BTCUR_LEFTRA;
-		if (be32_to_cpu(b->bb_u.s.bb_rightsib) == NULLAGBLOCK)
+		if (b->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK))
 			cur->bc_ra[lev] |= XFS_BTCUR_RIGHTRA;
 	}
 }
@@ -795,9 +794,9 @@ xfs_btree_ptr_is_null(
 	union xfs_btree_ptr	*ptr)
 {
 	if (cur->bc_flags & XFS_BTREE_LONG_PTRS)
-		return be64_to_cpu(ptr->l) == NULLDFSBNO;
+		return ptr->l == cpu_to_be64(NULLDFSBNO);
 	else
-		return be32_to_cpu(ptr->s) == NULLAGBLOCK;
+		return ptr->s == cpu_to_be32(NULLAGBLOCK);
 }
 
 STATIC void
@@ -923,12 +922,12 @@ xfs_btree_ptr_to_daddr(
 	union xfs_btree_ptr	*ptr)
 {
 	if (cur->bc_flags & XFS_BTREE_LONG_PTRS) {
-		ASSERT(be64_to_cpu(ptr->l) != NULLDFSBNO);
+		ASSERT(ptr->l != cpu_to_be64(NULLDFSBNO));
 
 		return XFS_FSB_TO_DADDR(cur->bc_mp, be64_to_cpu(ptr->l));
 	} else {
 		ASSERT(cur->bc_private.a.agno != NULLAGNUMBER);
-		ASSERT(be32_to_cpu(ptr->s) != NULLAGBLOCK);
+		ASSERT(ptr->s != cpu_to_be32(NULLAGBLOCK));
 
 		return XFS_AGB_TO_DADDR(cur->bc_mp, cur->bc_private.a.agno,
 					be32_to_cpu(ptr->s));
diff --git a/fs/xfs/xfs_btree.h b/fs/xfs/xfs_btree.h
index 82fafc66bd1..8d05a6a46ce 100644
--- a/fs/xfs/xfs_btree.h
+++ b/fs/xfs/xfs_btree.h
@@ -199,25 +199,6 @@ struct xfs_btree_ops {
 				union xfs_btree_rec *r1,
 				union xfs_btree_rec *r2);
 #endif
-
-	/* btree tracing */
-#ifdef XFS_BTREE_TRACE
-	void		(*trace_enter)(struct xfs_btree_cur *, const char *,
-				       char *, int, int, __psunsigned_t,
-				       __psunsigned_t, __psunsigned_t,
-				       __psunsigned_t, __psunsigned_t,
-				       __psunsigned_t, __psunsigned_t,
-				       __psunsigned_t, __psunsigned_t,
-				       __psunsigned_t, __psunsigned_t);
-	void		(*trace_cursor)(struct xfs_btree_cur *, __uint32_t *,
-					__uint64_t *, __uint64_t *);
-	void		(*trace_key)(struct xfs_btree_cur *,
-				     union xfs_btree_key *, __uint64_t *,
-				     __uint64_t *);
-	void		(*trace_record)(struct xfs_btree_cur *,
-					union xfs_btree_rec *, __uint64_t *,
-					__uint64_t *, __uint64_t *);
-#endif
 };
 
 /*
@@ -452,4 +433,23 @@ static inline int xfs_btree_get_level(struct xfs_btree_block *block)
 	(XFS_FSB_TO_AGNO(mp, fsb) < mp->m_sb.sb_agcount && \
 		XFS_FSB_TO_AGBNO(mp, fsb) < mp->m_sb.sb_agblocks)
 
+/*
+ * Trace hooks.  Currently not implemented as they need to be ported
+ * over to the generic tracing functionality, which is some effort.
+ *
+ * i,j = integer (32 bit)
+ * b = btree block buffer (xfs_buf_t)
+ * p = btree ptr
+ * r = btree record
+ * k = btree key
+ */
+#define	XFS_BTREE_TRACE_ARGBI(c, b, i)
+#define	XFS_BTREE_TRACE_ARGBII(c, b, i, j)
+#define	XFS_BTREE_TRACE_ARGI(c, i)
+#define	XFS_BTREE_TRACE_ARGIPK(c, i, p, s)
+#define	XFS_BTREE_TRACE_ARGIPR(c, i, p, r)
+#define	XFS_BTREE_TRACE_ARGIK(c, i, k)
+#define XFS_BTREE_TRACE_ARGR(c, r)
+#define	XFS_BTREE_TRACE_CURSOR(c, t)
+
 #endif	/* __XFS_BTREE_H__ */
diff --git a/fs/xfs/xfs_btree_trace.c b/fs/xfs/xfs_btree_trace.c
deleted file mode 100644
index 44ff942a0fd..00000000000
--- a/fs/xfs/xfs_btree_trace.c
+++ /dev/null
@@ -1,249 +0,0 @@
-/*
- * Copyright (c) 2008 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_types.h"
-#include "xfs_inum.h"
-#include "xfs_bmap_btree.h"
-#include "xfs_alloc_btree.h"
-#include "xfs_ialloc_btree.h"
-#include "xfs_inode.h"
-#include "xfs_btree.h"
-#include "xfs_btree_trace.h"
-
-STATIC void
-xfs_btree_trace_ptr(
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_ptr	ptr,
-	__psunsigned_t		*high,
-	__psunsigned_t		*low)
-{
-	if (cur->bc_flags & XFS_BTREE_LONG_PTRS) {
-		__u64 val = be64_to_cpu(ptr.l);
-		*high = val >> 32;
-		*low = (int)val;
-	} else {
-		*high = 0;
-		*low = be32_to_cpu(ptr.s);
-	}
-}
-
-/*
- * Add a trace buffer entry for arguments, for a buffer & 1 integer arg.
- */
-void
-xfs_btree_trace_argbi(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	struct xfs_buf		*b,
-	int			i,
-	int			line)
-{
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGBI,
-				 line, (__psunsigned_t)b, i, 0, 0, 0, 0, 0,
-				 0, 0, 0, 0);
-}
-
-/*
- * Add a trace buffer entry for arguments, for a buffer & 2 integer args.
- */
-void
-xfs_btree_trace_argbii(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	struct xfs_buf		*b,
-	int			i0,
-	int			i1,
-	int			line)
-{
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGBII,
-				 line, (__psunsigned_t)b, i0, i1, 0, 0, 0, 0,
-				 0, 0, 0, 0);
-}
-
-/*
- * Add a trace buffer entry for arguments, for 3 block-length args
- * and an integer arg.
- */
-void
-xfs_btree_trace_argfffi(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	xfs_dfiloff_t		o,
-	xfs_dfsbno_t		b,
-	xfs_dfilblks_t		i,
-	int			j,
-	int			line)
-{
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGFFFI,
-				 line,
-				 o >> 32, (int)o,
-				 b >> 32, (int)b,
-				 i >> 32, (int)i,
-				 (int)j, 0, 0, 0, 0);
-}
-
-/*
- * Add a trace buffer entry for arguments, for one integer arg.
- */
-void
-xfs_btree_trace_argi(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	int			i,
-	int			line)
-{
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGI,
-				 line, i, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0);
-}
-
-/*
- * Add a trace buffer entry for arguments, for int, fsblock, key.
- */
-void
-xfs_btree_trace_argipk(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	int			i,
-	union xfs_btree_ptr	ptr,
-	union xfs_btree_key	*key,
-	int			line)
-{
-	__psunsigned_t		high, low;
-	__uint64_t		l0, l1;
-
-	xfs_btree_trace_ptr(cur, ptr, &high, &low);
-	cur->bc_ops->trace_key(cur, key, &l0, &l1);
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGIPK,
-				 line, i, high, low,
-				 l0 >> 32, (int)l0,
-				 l1 >> 32, (int)l1,
-				 0, 0, 0, 0);
-}
-
-/*
- * Add a trace buffer entry for arguments, for int, fsblock, rec.
- */
-void
-xfs_btree_trace_argipr(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	int			i,
-	union xfs_btree_ptr	ptr,
-	union xfs_btree_rec	*rec,
-	int			line)
-{
-	__psunsigned_t		high, low;
-	__uint64_t		l0, l1, l2;
-
-	xfs_btree_trace_ptr(cur, ptr, &high, &low);
-	cur->bc_ops->trace_record(cur, rec, &l0, &l1, &l2);
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGIPR,
-			      line, i,
-			      high, low,
-			      l0 >> 32, (int)l0,
-			      l1 >> 32, (int)l1,
-			      l2 >> 32, (int)l2,
-			      0, 0);
-}
-
-/*
- * Add a trace buffer entry for arguments, for int, key.
- */
-void
-xfs_btree_trace_argik(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	int			i,
-	union xfs_btree_key	*key,
-	int			line)
-{
-	__uint64_t		l0, l1;
-
-	cur->bc_ops->trace_key(cur, key, &l0, &l1);
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGIK,
-				 line, i,
-				 l0 >> 32, (int)l0,
-				 l1 >> 32, (int)l1,
-				 0, 0, 0, 0, 0, 0);
-}
-
-/*
- * Add a trace buffer entry for arguments, for record.
- */
-void
-xfs_btree_trace_argr(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_rec	*rec,
-	int			line)
-{
-	__uint64_t		l0, l1, l2;
-
-	cur->bc_ops->trace_record(cur, rec, &l0, &l1, &l2);
-	cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGR,
-			      line,
-			      l0 >> 32, (int)l0,
-			      l1 >> 32, (int)l1,
-			      l2 >> 32, (int)l2,
-			      0, 0, 0, 0, 0);
-}
-
-/*
- * Add a trace buffer entry for the cursor/operation.
- */
-void
-xfs_btree_trace_cursor(
-	const char		*func,
-	struct xfs_btree_cur	*cur,
-	int			type,
-	int			line)
-{
-	__uint32_t		s0;
-	__uint64_t		l0, l1;
-	char			*s;
-
-	switch (type) {
-	case XBT_ARGS:
-		s = "args";
-		break;
-	case XBT_ENTRY:
-		s = "entry";
-		break;
-	case XBT_ERROR:
-		s = "error";
-		break;
-	case XBT_EXIT:
-		s = "exit";
-		break;
-	default:
-		s = "unknown";
-		break;
-	}
-
-	cur->bc_ops->trace_cursor(cur, &s0, &l0, &l1);
-	cur->bc_ops->trace_enter(cur, func, s, XFS_BTREE_KTRACE_CUR, line,
-				 s0,
-				 l0 >> 32, (int)l0,
-				 l1 >> 32, (int)l1,
-				 (__psunsigned_t)cur->bc_bufs[0],
-				 (__psunsigned_t)cur->bc_bufs[1],
-				 (__psunsigned_t)cur->bc_bufs[2],
-				 (__psunsigned_t)cur->bc_bufs[3],
-				 (cur->bc_ptrs[0] << 16) | cur->bc_ptrs[1],
-				 (cur->bc_ptrs[2] << 16) | cur->bc_ptrs[3]);
-}
diff --git a/fs/xfs/xfs_btree_trace.h b/fs/xfs/xfs_btree_trace.h
deleted file mode 100644
index 2d8a309873e..00000000000
--- a/fs/xfs/xfs_btree_trace.h
+++ /dev/null
@@ -1,99 +0,0 @@
-/*
- * Copyright (c) 2008 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_BTREE_TRACE_H__
-#define	__XFS_BTREE_TRACE_H__
-
-struct xfs_btree_cur;
-struct xfs_buf;
-
-
-/*
- * Trace hooks.
- * i,j = integer (32 bit)
- * b = btree block buffer (xfs_buf_t)
- * p = btree ptr
- * r = btree record
- * k = btree key
- */
-
-#ifdef XFS_BTREE_TRACE
-
-/*
- * Trace buffer entry types.
- */
-#define XFS_BTREE_KTRACE_ARGBI   1
-#define XFS_BTREE_KTRACE_ARGBII  2
-#define XFS_BTREE_KTRACE_ARGFFFI 3
-#define XFS_BTREE_KTRACE_ARGI    4
-#define XFS_BTREE_KTRACE_ARGIPK  5
-#define XFS_BTREE_KTRACE_ARGIPR  6
-#define XFS_BTREE_KTRACE_ARGIK   7
-#define XFS_BTREE_KTRACE_ARGR	 8
-#define XFS_BTREE_KTRACE_CUR     9
-
-/*
- * Sub-types for cursor traces.
- */
-#define XBT_ARGS	0
-#define XBT_ENTRY	1
-#define XBT_ERROR	2
-#define XBT_EXIT	3
-
-void xfs_btree_trace_argbi(const char *, struct xfs_btree_cur *,
-		struct xfs_buf *, int, int);
-void xfs_btree_trace_argbii(const char *, struct xfs_btree_cur *,
-		struct xfs_buf *, int, int, int);
-void xfs_btree_trace_argi(const char *, struct xfs_btree_cur *, int, int);
-void xfs_btree_trace_argipk(const char *, struct xfs_btree_cur *, int,
-		union xfs_btree_ptr, union xfs_btree_key *, int);
-void xfs_btree_trace_argipr(const char *, struct xfs_btree_cur *, int,
-		union xfs_btree_ptr, union xfs_btree_rec *, int);
-void xfs_btree_trace_argik(const char *, struct xfs_btree_cur *, int,
-		union xfs_btree_key *, int);
-void xfs_btree_trace_argr(const char *, struct xfs_btree_cur *,
-		union xfs_btree_rec *, int);
-void xfs_btree_trace_cursor(const char *, struct xfs_btree_cur *, int, int);
-
-#define	XFS_BTREE_TRACE_ARGBI(c, b, i)	\
-	xfs_btree_trace_argbi(__func__, c, b, i, __LINE__)
-#define	XFS_BTREE_TRACE_ARGBII(c, b, i, j)	\
-	xfs_btree_trace_argbii(__func__, c, b, i, j, __LINE__)
-#define	XFS_BTREE_TRACE_ARGI(c, i)	\
-	xfs_btree_trace_argi(__func__, c, i, __LINE__)
-#define	XFS_BTREE_TRACE_ARGIPK(c, i, p, k)	\
-	xfs_btree_trace_argipk(__func__, c, i, p, k, __LINE__)
-#define	XFS_BTREE_TRACE_ARGIPR(c, i, p, r)	\
-	xfs_btree_trace_argipr(__func__, c, i, p, r, __LINE__)
-#define	XFS_BTREE_TRACE_ARGIK(c, i, k)	\
-	xfs_btree_trace_argik(__func__, c, i, k, __LINE__)
-#define XFS_BTREE_TRACE_ARGR(c, r)	\
-	xfs_btree_trace_argr(__func__, c, r, __LINE__)
-#define	XFS_BTREE_TRACE_CURSOR(c, t)	\
-	xfs_btree_trace_cursor(__func__, c, t, __LINE__)
-#else
-#define	XFS_BTREE_TRACE_ARGBI(c, b, i)
-#define	XFS_BTREE_TRACE_ARGBII(c, b, i, j)
-#define	XFS_BTREE_TRACE_ARGI(c, i)
-#define	XFS_BTREE_TRACE_ARGIPK(c, i, p, s)
-#define	XFS_BTREE_TRACE_ARGIPR(c, i, p, r)
-#define	XFS_BTREE_TRACE_ARGIK(c, i, k)
-#define XFS_BTREE_TRACE_ARGR(c, r)
-#define	XFS_BTREE_TRACE_CURSOR(c, t)
-#endif	/* XFS_BTREE_TRACE */
-
-#endif /* __XFS_BTREE_TRACE_H__ */
diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index 7b7e005e3dc..88492916c3d 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -90,13 +90,11 @@ xfs_buf_item_flush_log_debug(
 	uint		first,
 	uint		last)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 	uint			nbytes;
 
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
-	if ((bip == NULL) || (bip->bli_item.li_type != XFS_LI_BUF)) {
+	if (bip == NULL || (bip->bli_item.li_type != XFS_LI_BUF))
 		return;
-	}
 
 	ASSERT(bip->bli_logged != NULL);
 	nbytes = last - first + 1;
@@ -408,7 +406,7 @@ xfs_buf_item_unpin(
 	int		stale = bip->bli_flags & XFS_BLI_STALE;
 	int		freed;
 
-	ASSERT(XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *) == bip);
+	ASSERT(bp->b_fspriv == bip);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 
 	trace_xfs_buf_item_unpin(bip);
@@ -420,7 +418,7 @@ xfs_buf_item_unpin(
 
 	if (freed && stale) {
 		ASSERT(bip->bli_flags & XFS_BLI_STALE);
-		ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
+		ASSERT(xfs_buf_islocked(bp));
 		ASSERT(!(XFS_BUF_ISDELAYWRITE(bp)));
 		ASSERT(XFS_BUF_ISSTALE(bp));
 		ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL);
@@ -443,7 +441,7 @@ xfs_buf_item_unpin(
 			 * Since the transaction no longer refers to the buffer,
 			 * the buffer should no longer refer to the transaction.
 			 */
-			XFS_BUF_SET_FSPRIVATE2(bp, NULL);
+			bp->b_transp = NULL;
 		}
 
 		/*
@@ -454,13 +452,13 @@ xfs_buf_item_unpin(
 		 */
 		if (bip->bli_flags & XFS_BLI_STALE_INODE) {
 			xfs_buf_do_callbacks(bp);
-			XFS_BUF_SET_FSPRIVATE(bp, NULL);
-			XFS_BUF_CLR_IODONE_FUNC(bp);
+			bp->b_fspriv = NULL;
+			bp->b_iodone = NULL;
 		} else {
 			spin_lock(&ailp->xa_lock);
 			xfs_trans_ail_delete(ailp, (xfs_log_item_t *)bip);
 			xfs_buf_item_relse(bp);
-			ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL);
+			ASSERT(bp->b_fspriv == NULL);
 		}
 		xfs_buf_relse(bp);
 	}
@@ -483,7 +481,7 @@ xfs_buf_item_trylock(
 
 	if (XFS_BUF_ISPINNED(bp))
 		return XFS_ITEM_PINNED;
-	if (!XFS_BUF_CPSEMA(bp))
+	if (!xfs_buf_trylock(bp))
 		return XFS_ITEM_LOCKED;
 
 	/* take a reference to the buffer.  */
@@ -525,7 +523,7 @@ xfs_buf_item_unlock(
 	uint			hold;
 
 	/* Clear the buffer's association with this transaction. */
-	XFS_BUF_SET_FSPRIVATE2(bp, NULL);
+	bp->b_transp = NULL;
 
 	/*
 	 * If this is a transaction abort, don't return early.  Instead, allow
@@ -684,7 +682,7 @@ xfs_buf_item_init(
 	xfs_buf_t	*bp,
 	xfs_mount_t	*mp)
 {
-	xfs_log_item_t		*lip;
+	xfs_log_item_t		*lip = bp->b_fspriv;
 	xfs_buf_log_item_t	*bip;
 	int			chunks;
 	int			map_size;
@@ -696,12 +694,8 @@ xfs_buf_item_init(
 	 * nothing to do here so return.
 	 */
 	ASSERT(bp->b_target->bt_mount == mp);
-	if (XFS_BUF_FSPRIVATE(bp, void *) != NULL) {
-		lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
-		if (lip->li_type == XFS_LI_BUF) {
-			return;
-		}
-	}
+	if (lip != NULL && lip->li_type == XFS_LI_BUF)
+		return;
 
 	/*
 	 * chunks is the number of XFS_BLF_CHUNK size pieces
@@ -740,11 +734,9 @@ xfs_buf_item_init(
 	 * Put the buf item into the list of items attached to the
 	 * buffer at the front.
 	 */
-	if (XFS_BUF_FSPRIVATE(bp, void *) != NULL) {
-		bip->bli_item.li_bio_list =
-				XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
-	}
-	XFS_BUF_SET_FSPRIVATE(bp, bip);
+	if (bp->b_fspriv)
+		bip->bli_item.li_bio_list = bp->b_fspriv;
+	bp->b_fspriv = bip;
 }
 
 
@@ -876,12 +868,11 @@ xfs_buf_item_relse(
 
 	trace_xfs_buf_item_relse(bp, _RET_IP_);
 
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
-	XFS_BUF_SET_FSPRIVATE(bp, bip->bli_item.li_bio_list);
-	if ((XFS_BUF_FSPRIVATE(bp, void *) == NULL) &&
-	    (XFS_BUF_IODONE_FUNC(bp) != NULL)) {
-		XFS_BUF_CLR_IODONE_FUNC(bp);
-	}
+	bip = bp->b_fspriv;
+	bp->b_fspriv = bip->bli_item.li_bio_list;
+	if (bp->b_fspriv == NULL)
+		bp->b_iodone = NULL;
+
 	xfs_buf_rele(bp);
 	xfs_buf_item_free(bip);
 }
@@ -905,20 +896,20 @@ xfs_buf_attach_iodone(
 	xfs_log_item_t	*head_lip;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
+	ASSERT(xfs_buf_islocked(bp));
 
 	lip->li_cb = cb;
-	if (XFS_BUF_FSPRIVATE(bp, void *) != NULL) {
-		head_lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
+	head_lip = bp->b_fspriv;
+	if (head_lip) {
 		lip->li_bio_list = head_lip->li_bio_list;
 		head_lip->li_bio_list = lip;
 	} else {
-		XFS_BUF_SET_FSPRIVATE(bp, lip);
+		bp->b_fspriv = lip;
 	}
 
-	ASSERT((XFS_BUF_IODONE_FUNC(bp) == xfs_buf_iodone_callbacks) ||
-	       (XFS_BUF_IODONE_FUNC(bp) == NULL));
-	XFS_BUF_SET_IODONE_FUNC(bp, xfs_buf_iodone_callbacks);
+	ASSERT(bp->b_iodone == NULL ||
+	       bp->b_iodone == xfs_buf_iodone_callbacks);
+	bp->b_iodone = xfs_buf_iodone_callbacks;
 }
 
 /*
@@ -939,8 +930,8 @@ xfs_buf_do_callbacks(
 {
 	struct xfs_log_item	*lip;
 
-	while ((lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *)) != NULL) {
-		XFS_BUF_SET_FSPRIVATE(bp, lip->li_bio_list);
+	while ((lip = bp->b_fspriv) != NULL) {
+		bp->b_fspriv = lip->li_bio_list;
 		ASSERT(lip->li_cb != NULL);
 		/*
 		 * Clear the next pointer so we don't have any
@@ -1007,7 +998,7 @@ xfs_buf_iodone_callbacks(
 			XFS_BUF_DONE(bp);
 			XFS_BUF_SET_START(bp);
 		}
-		ASSERT(XFS_BUF_IODONE_FUNC(bp));
+		ASSERT(bp->b_iodone != NULL);
 		trace_xfs_buf_item_iodone_async(bp, _RET_IP_);
 		xfs_buf_relse(bp);
 		return;
@@ -1026,8 +1017,8 @@ xfs_buf_iodone_callbacks(
 
 do_callbacks:
 	xfs_buf_do_callbacks(bp);
-	XFS_BUF_SET_FSPRIVATE(bp, NULL);
-	XFS_BUF_CLR_IODONE_FUNC(bp);
+	bp->b_fspriv = NULL;
+	bp->b_iodone = NULL;
 	xfs_buf_ioend(bp, 0);
 }
 
diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c
index 6102ac6d1df..2925726529f 100644
--- a/fs/xfs/xfs_da_btree.c
+++ b/fs/xfs/xfs_da_btree.c
@@ -24,11 +24,12 @@
 #include "xfs_trans.h"
 #include "xfs_sb.h"
 #include "xfs_ag.h"
-#include "xfs_dir2.h"
 #include "xfs_mount.h"
 #include "xfs_da_btree.h"
 #include "xfs_bmap_btree.h"
-#include "xfs_dir2_sf.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_format.h"
+#include "xfs_dir2_priv.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_inode_item.h"
@@ -36,10 +37,6 @@
 #include "xfs_bmap.h"
 #include "xfs_attr.h"
 #include "xfs_attr_leaf.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
-#include "xfs_dir2_node.h"
 #include "xfs_error.h"
 #include "xfs_trace.h"
 
@@ -89,7 +86,7 @@ STATIC void xfs_da_node_unbalance(xfs_da_state_t *state,
  */
 STATIC uint	xfs_da_node_lasthash(xfs_dabuf_t *bp, int *count);
 STATIC int	xfs_da_node_order(xfs_dabuf_t *node1_bp, xfs_dabuf_t *node2_bp);
-STATIC xfs_dabuf_t *xfs_da_buf_make(int nbuf, xfs_buf_t **bps, inst_t *ra);
+STATIC xfs_dabuf_t *xfs_da_buf_make(int nbuf, xfs_buf_t **bps);
 STATIC int	xfs_da_blk_unlink(xfs_da_state_t *state,
 				  xfs_da_state_blk_t *drop_blk,
 				  xfs_da_state_blk_t *save_blk);
@@ -321,11 +318,11 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
 	ASSERT(bp != NULL);
 	node = bp->data;
 	oldroot = blk1->bp->data;
-	if (be16_to_cpu(oldroot->hdr.info.magic) == XFS_DA_NODE_MAGIC) {
+	if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC)) {
 		size = (int)((char *)&oldroot->btree[be16_to_cpu(oldroot->hdr.count)] -
 			     (char *)oldroot);
 	} else {
-		ASSERT(be16_to_cpu(oldroot->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+		ASSERT(oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 		leaf = (xfs_dir2_leaf_t *)oldroot;
 		size = (int)((char *)&leaf->ents[be16_to_cpu(leaf->hdr.count)] -
 			     (char *)leaf);
@@ -352,7 +349,7 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
 	node->hdr.count = cpu_to_be16(2);
 
 #ifdef DEBUG
-	if (be16_to_cpu(oldroot->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC) {
+	if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)) {
 		ASSERT(blk1->blkno >= mp->m_dirleafblk &&
 		       blk1->blkno < mp->m_dirfreeblk);
 		ASSERT(blk2->blkno >= mp->m_dirleafblk &&
@@ -384,7 +381,7 @@ xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
 	int useextra;
 
 	node = oldblk->bp->data;
-	ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 
 	/*
 	 * With V2 dirs the extra block is data or freespace.
@@ -483,8 +480,8 @@ xfs_da_node_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
 		node1 = node2;
 		node2 = tmpnode;
 	}
-	ASSERT(be16_to_cpu(node1->hdr.info.magic) == XFS_DA_NODE_MAGIC);
-	ASSERT(be16_to_cpu(node2->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(node1->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
+	ASSERT(node2->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	count = (be16_to_cpu(node1->hdr.count) - be16_to_cpu(node2->hdr.count)) / 2;
 	if (count == 0)
 		return;
@@ -578,7 +575,7 @@ xfs_da_node_add(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
 	int tmp;
 
 	node = oldblk->bp->data;
-	ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	ASSERT((oldblk->index >= 0) && (oldblk->index <= be16_to_cpu(node->hdr.count)));
 	ASSERT(newblk->blkno != 0);
 	if (state->args->whichfork == XFS_DATA_FORK)
@@ -714,7 +711,7 @@ xfs_da_root_join(xfs_da_state_t *state, xfs_da_state_blk_t *root_blk)
 	ASSERT(args != NULL);
 	ASSERT(root_blk->magic == XFS_DA_NODE_MAGIC);
 	oldroot = root_blk->bp->data;
-	ASSERT(be16_to_cpu(oldroot->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(oldroot->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	ASSERT(!oldroot->hdr.info.forw);
 	ASSERT(!oldroot->hdr.info.back);
 
@@ -737,10 +734,10 @@ xfs_da_root_join(xfs_da_state_t *state, xfs_da_state_blk_t *root_blk)
 	ASSERT(bp != NULL);
 	blkinfo = bp->data;
 	if (be16_to_cpu(oldroot->hdr.level) == 1) {
-		ASSERT(be16_to_cpu(blkinfo->magic) == XFS_DIR2_LEAFN_MAGIC ||
-		       be16_to_cpu(blkinfo->magic) == XFS_ATTR_LEAF_MAGIC);
+		ASSERT(blkinfo->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+		       blkinfo->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 	} else {
-		ASSERT(be16_to_cpu(blkinfo->magic) == XFS_DA_NODE_MAGIC);
+		ASSERT(blkinfo->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	}
 	ASSERT(!blkinfo->forw);
 	ASSERT(!blkinfo->back);
@@ -776,7 +773,7 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
 	 */
 	blk = &state->path.blk[ state->path.active-1 ];
 	info = blk->bp->data;
-	ASSERT(be16_to_cpu(info->magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	node = (xfs_da_intnode_t *)info;
 	count = be16_to_cpu(node->hdr.count);
 	if (count > (state->node_ents >> 1)) {
@@ -836,7 +833,7 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
 		count -= state->node_ents >> 2;
 		count -= be16_to_cpu(node->hdr.count);
 		node = bp->data;
-		ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+		ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 		count -= be16_to_cpu(node->hdr.count);
 		xfs_da_brelse(state->args->trans, bp);
 		if (count >= 0)
@@ -911,7 +908,7 @@ xfs_da_fixhashpath(xfs_da_state_t *state, xfs_da_state_path_t *path)
 	}
 	for (blk--, level--; level >= 0; blk--, level--) {
 		node = blk->bp->data;
-		ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+		ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 		btree = &node->btree[ blk->index ];
 		if (be32_to_cpu(btree->hashval) == lasthash)
 			break;
@@ -979,8 +976,8 @@ xfs_da_node_unbalance(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
 
 	drop_node = drop_blk->bp->data;
 	save_node = save_blk->bp->data;
-	ASSERT(be16_to_cpu(drop_node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
-	ASSERT(be16_to_cpu(save_node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(drop_node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
+	ASSERT(save_node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	tp = state->args->trans;
 
 	/*
@@ -1278,8 +1275,8 @@ xfs_da_node_order(xfs_dabuf_t *node1_bp, xfs_dabuf_t *node2_bp)
 
 	node1 = node1_bp->data;
 	node2 = node2_bp->data;
-	ASSERT((be16_to_cpu(node1->hdr.info.magic) == XFS_DA_NODE_MAGIC) &&
-	       (be16_to_cpu(node2->hdr.info.magic) == XFS_DA_NODE_MAGIC));
+	ASSERT(node1->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC) &&
+	       node2->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	if ((be16_to_cpu(node1->hdr.count) > 0) && (be16_to_cpu(node2->hdr.count) > 0) &&
 	    ((be32_to_cpu(node2->btree[0].hashval) <
 	      be32_to_cpu(node1->btree[0].hashval)) ||
@@ -1299,7 +1296,7 @@ xfs_da_node_lasthash(xfs_dabuf_t *bp, int *count)
 	xfs_da_intnode_t *node;
 
 	node = bp->data;
-	ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+	ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 	if (count)
 		*count = be16_to_cpu(node->hdr.count);
 	if (!node->hdr.count)
@@ -1412,7 +1409,7 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
 	for (blk = &path->blk[level]; level >= 0; blk--, level--) {
 		ASSERT(blk->bp != NULL);
 		node = blk->bp->data;
-		ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC);
+		ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 		if (forward && (blk->index < be16_to_cpu(node->hdr.count)-1)) {
 			blk->index++;
 			blkno = be32_to_cpu(node->btree[blk->index].before);
@@ -1451,9 +1448,9 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
 			return(error);
 		ASSERT(blk->bp != NULL);
 		info = blk->bp->data;
-		ASSERT(be16_to_cpu(info->magic) == XFS_DA_NODE_MAGIC ||
-		       be16_to_cpu(info->magic) == XFS_DIR2_LEAFN_MAGIC ||
-		       be16_to_cpu(info->magic) == XFS_ATTR_LEAF_MAGIC);
+		ASSERT(info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
+		       info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+		       info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
 		blk->magic = be16_to_cpu(info->magic);
 		if (blk->magic == XFS_DA_NODE_MAGIC) {
 			node = (xfs_da_intnode_t *)info;
@@ -1546,79 +1543,62 @@ const struct xfs_nameops xfs_default_nameops = {
 	.compname	= xfs_da_compname
 };
 
-/*
- * Add a block to the btree ahead of the file.
- * Return the new block number to the caller.
- */
 int
-xfs_da_grow_inode(xfs_da_args_t *args, xfs_dablk_t *new_blkno)
+xfs_da_grow_inode_int(
+	struct xfs_da_args	*args,
+	xfs_fileoff_t		*bno,
+	int			count)
 {
-	xfs_fileoff_t bno, b;
-	xfs_bmbt_irec_t map;
-	xfs_bmbt_irec_t	*mapp;
-	xfs_inode_t *dp;
-	int nmap, error, w, count, c, got, i, mapi;
-	xfs_trans_t *tp;
-	xfs_mount_t *mp;
-	xfs_drfsbno_t	nblks;
+	struct xfs_trans	*tp = args->trans;
+	struct xfs_inode	*dp = args->dp;
+	int			w = args->whichfork;
+	xfs_drfsbno_t		nblks = dp->i_d.di_nblocks;
+	struct xfs_bmbt_irec	map, *mapp;
+	int			nmap, error, got, i, mapi;
 
-	dp = args->dp;
-	mp = dp->i_mount;
-	w = args->whichfork;
-	tp = args->trans;
-	nblks = dp->i_d.di_nblocks;
-
-	/*
-	 * For new directories adjust the file offset and block count.
-	 */
-	if (w == XFS_DATA_FORK) {
-		bno = mp->m_dirleafblk;
-		count = mp->m_dirblkfsbs;
-	} else {
-		bno = 0;
-		count = 1;
-	}
 	/*
 	 * Find a spot in the file space to put the new block.
 	 */
-	if ((error = xfs_bmap_first_unused(tp, dp, count, &bno, w)))
+	error = xfs_bmap_first_unused(tp, dp, count, bno, w);
+	if (error)
 		return error;
-	if (w == XFS_DATA_FORK)
-		ASSERT(bno >= mp->m_dirleafblk && bno < mp->m_dirfreeblk);
+
 	/*
 	 * Try mapping it in one filesystem block.
 	 */
 	nmap = 1;
 	ASSERT(args->firstblock != NULL);
-	if ((error = xfs_bmapi(tp, dp, bno, count,
+	error = xfs_bmapi(tp, dp, *bno, count,
 			xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE|XFS_BMAPI_METADATA|
 			XFS_BMAPI_CONTIG,
 			args->firstblock, args->total, &map, &nmap,
-			args->flist))) {
+			args->flist);
+	if (error)
 		return error;
-	}
+
 	ASSERT(nmap <= 1);
 	if (nmap == 1) {
 		mapp = &map;
 		mapi = 1;
-	}
-	/*
-	 * If we didn't get it and the block might work if fragmented,
-	 * try without the CONTIG flag.  Loop until we get it all.
-	 */
-	else if (nmap == 0 && count > 1) {
+	} else if (nmap == 0 && count > 1) {
+		xfs_fileoff_t		b;
+		int			c;
+
+		/*
+		 * If we didn't get it and the block might work if fragmented,
+		 * try without the CONTIG flag.  Loop until we get it all.
+		 */
 		mapp = kmem_alloc(sizeof(*mapp) * count, KM_SLEEP);
-		for (b = bno, mapi = 0; b < bno + count; ) {
+		for (b = *bno, mapi = 0; b < *bno + count; ) {
 			nmap = MIN(XFS_BMAP_MAX_NMAP, count);
-			c = (int)(bno + count - b);
-			if ((error = xfs_bmapi(tp, dp, b, c,
+			c = (int)(*bno + count - b);
+			error = xfs_bmapi(tp, dp, b, c,
 					xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE|
 					XFS_BMAPI_METADATA,
 					args->firstblock, args->total,
-					&mapp[mapi], &nmap, args->flist))) {
-				kmem_free(mapp);
-				return error;
-			}
+					&mapp[mapi], &nmap, args->flist);
+			if (error)
+				goto out_free_map;
 			if (nmap < 1)
 				break;
 			mapi += nmap;
@@ -1629,24 +1609,53 @@ xfs_da_grow_inode(xfs_da_args_t *args, xfs_dablk_t *new_blkno)
 		mapi = 0;
 		mapp = NULL;
 	}
+
 	/*
 	 * Count the blocks we got, make sure it matches the total.
 	 */
 	for (i = 0, got = 0; i < mapi; i++)
 		got += mapp[i].br_blockcount;
-	if (got != count || mapp[0].br_startoff != bno ||
+	if (got != count || mapp[0].br_startoff != *bno ||
 	    mapp[mapi - 1].br_startoff + mapp[mapi - 1].br_blockcount !=
-	    bno + count) {
-		if (mapp != &map)
-			kmem_free(mapp);
-		return XFS_ERROR(ENOSPC);
+	    *bno + count) {
+		error = XFS_ERROR(ENOSPC);
+		goto out_free_map;
 	}
-	if (mapp != &map)
-		kmem_free(mapp);
+
 	/* account for newly allocated blocks in reserved blocks total */
 	args->total -= dp->i_d.di_nblocks - nblks;
-	*new_blkno = (xfs_dablk_t)bno;
-	return 0;
+
+out_free_map:
+	if (mapp != &map)
+		kmem_free(mapp);
+	return error;
+}
+
+/*
+ * Add a block to the btree ahead of the file.
+ * Return the new block number to the caller.
+ */
+int
+xfs_da_grow_inode(
+	struct xfs_da_args	*args,
+	xfs_dablk_t		*new_blkno)
+{
+	xfs_fileoff_t		bno;
+	int			count;
+	int			error;
+
+	if (args->whichfork == XFS_DATA_FORK) {
+		bno = args->dp->i_mount->m_dirleafblk;
+		count = args->dp->i_mount->m_dirblkfsbs;
+	} else {
+		bno = 0;
+		count = 1;
+	}
+
+	error = xfs_da_grow_inode_int(args, &bno, count);
+	if (!error)
+		*new_blkno = (xfs_dablk_t)bno;
+	return error;
 }
 
 /*
@@ -1704,12 +1713,12 @@ xfs_da_swap_lastblock(xfs_da_args_t *args, xfs_dablk_t *dead_blknop,
 	/*
 	 * Get values from the moved block.
 	 */
-	if (be16_to_cpu(dead_info->magic) == XFS_DIR2_LEAFN_MAGIC) {
+	if (dead_info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)) {
 		dead_leaf2 = (xfs_dir2_leaf_t *)dead_info;
 		dead_level = 0;
 		dead_hash = be32_to_cpu(dead_leaf2->ents[be16_to_cpu(dead_leaf2->hdr.count) - 1].hashval);
 	} else {
-		ASSERT(be16_to_cpu(dead_info->magic) == XFS_DA_NODE_MAGIC);
+		ASSERT(dead_info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
 		dead_node = (xfs_da_intnode_t *)dead_info;
 		dead_level = be16_to_cpu(dead_node->hdr.level);
 		dead_hash = be32_to_cpu(dead_node->btree[be16_to_cpu(dead_node->hdr.count) - 1].hashval);
@@ -1768,8 +1777,8 @@ xfs_da_swap_lastblock(xfs_da_args_t *args, xfs_dablk_t *dead_blknop,
 		if ((error = xfs_da_read_buf(tp, ip, par_blkno, -1, &par_buf, w)))
 			goto done;
 		par_node = par_buf->data;
-		if (unlikely(
-		    be16_to_cpu(par_node->hdr.info.magic) != XFS_DA_NODE_MAGIC ||
+		if (unlikely(par_node->hdr.info.magic !=
+		    cpu_to_be16(XFS_DA_NODE_MAGIC) ||
 		    (level >= 0 && level != be16_to_cpu(par_node->hdr.level) + 1))) {
 			XFS_ERROR_REPORT("xfs_da_swap_lastblock(4)",
 					 XFS_ERRLEVEL_LOW, mp);
@@ -1820,7 +1829,7 @@ xfs_da_swap_lastblock(xfs_da_args_t *args, xfs_dablk_t *dead_blknop,
 		par_node = par_buf->data;
 		if (unlikely(
 		    be16_to_cpu(par_node->hdr.level) != level ||
-		    be16_to_cpu(par_node->hdr.info.magic) != XFS_DA_NODE_MAGIC)) {
+		    par_node->hdr.info.magic != cpu_to_be16(XFS_DA_NODE_MAGIC))) {
 			XFS_ERROR_REPORT("xfs_da_swap_lastblock(7)",
 					 XFS_ERRLEVEL_LOW, mp);
 			error = XFS_ERROR(EFSCORRUPTED);
@@ -1930,8 +1939,7 @@ xfs_da_do_buf(
 	xfs_daddr_t	*mappedbnop,
 	xfs_dabuf_t	**bpp,
 	int		whichfork,
-	int		caller,
-	inst_t		*ra)
+	int		caller)
 {
 	xfs_buf_t	*bp = NULL;
 	xfs_buf_t	**bplist;
@@ -2070,25 +2078,22 @@ xfs_da_do_buf(
 	 * Build a dabuf structure.
 	 */
 	if (bplist) {
-		rbp = xfs_da_buf_make(nbplist, bplist, ra);
+		rbp = xfs_da_buf_make(nbplist, bplist);
 	} else if (bp)
-		rbp = xfs_da_buf_make(1, &bp, ra);
+		rbp = xfs_da_buf_make(1, &bp);
 	else
 		rbp = NULL;
 	/*
 	 * For read_buf, check the magic number.
 	 */
 	if (caller == 1) {
-		xfs_dir2_data_t		*data;
-		xfs_dir2_free_t		*free;
-		xfs_da_blkinfo_t	*info;
+		xfs_dir2_data_hdr_t	*hdr = rbp->data;
+		xfs_dir2_free_t		*free = rbp->data;
+		xfs_da_blkinfo_t	*info = rbp->data;
 		uint			magic, magic1;
 
-		info = rbp->data;
-		data = rbp->data;
-		free = rbp->data;
 		magic = be16_to_cpu(info->magic);
-		magic1 = be32_to_cpu(data->hdr.magic);
+		magic1 = be32_to_cpu(hdr->magic);
 		if (unlikely(
 		    XFS_TEST_ERROR((magic != XFS_DA_NODE_MAGIC) &&
 				   (magic != XFS_ATTR_LEAF_MAGIC) &&
@@ -2096,7 +2101,7 @@ xfs_da_do_buf(
 				   (magic != XFS_DIR2_LEAFN_MAGIC) &&
 				   (magic1 != XFS_DIR2_BLOCK_MAGIC) &&
 				   (magic1 != XFS_DIR2_DATA_MAGIC) &&
-				   (be32_to_cpu(free->hdr.magic) != XFS_DIR2_FREE_MAGIC),
+				   (free->hdr.magic != cpu_to_be32(XFS_DIR2_FREE_MAGIC)),
 				mp, XFS_ERRTAG_DA_READ_BUF,
 				XFS_RANDOM_DA_READ_BUF))) {
 			trace_xfs_da_btree_corrupt(rbp->bps[0], _RET_IP_);
@@ -2143,8 +2148,7 @@ xfs_da_get_buf(
 	xfs_dabuf_t	**bpp,
 	int		whichfork)
 {
-	return xfs_da_do_buf(trans, dp, bno, &mappedbno, bpp, whichfork, 0,
-						 (inst_t *)__return_address);
+	return xfs_da_do_buf(trans, dp, bno, &mappedbno, bpp, whichfork, 0);
 }
 
 /*
@@ -2159,8 +2163,7 @@ xfs_da_read_buf(
 	xfs_dabuf_t	**bpp,
 	int		whichfork)
 {
-	return xfs_da_do_buf(trans, dp, bno, &mappedbno, bpp, whichfork, 1,
-		(inst_t *)__return_address);
+	return xfs_da_do_buf(trans, dp, bno, &mappedbno, bpp, whichfork, 1);
 }
 
 /*
@@ -2176,8 +2179,7 @@ xfs_da_reada_buf(
 	xfs_daddr_t		rval;
 
 	rval = -1;
-	if (xfs_da_do_buf(trans, dp, bno, &rval, NULL, whichfork, 3,
-			(inst_t *)__return_address))
+	if (xfs_da_do_buf(trans, dp, bno, &rval, NULL, whichfork, 3))
 		return -1;
 	else
 		return rval;
@@ -2235,17 +2237,12 @@ xfs_da_state_free(xfs_da_state_t *state)
 	kmem_zone_free(xfs_da_state_zone, state);
 }
 
-#ifdef XFS_DABUF_DEBUG
-xfs_dabuf_t	*xfs_dabuf_global_list;
-static DEFINE_SPINLOCK(xfs_dabuf_global_lock);
-#endif
-
 /*
  * Create a dabuf.
  */
 /* ARGSUSED */
 STATIC xfs_dabuf_t *
-xfs_da_buf_make(int nbuf, xfs_buf_t **bps, inst_t *ra)
+xfs_da_buf_make(int nbuf, xfs_buf_t **bps)
 {
 	xfs_buf_t	*bp;
 	xfs_dabuf_t	*dabuf;
@@ -2257,11 +2254,6 @@ xfs_da_buf_make(int nbuf, xfs_buf_t **bps, inst_t *ra)
 	else
 		dabuf = kmem_alloc(XFS_DA_BUF_SIZE(nbuf), KM_NOFS);
 	dabuf->dirty = 0;
-#ifdef XFS_DABUF_DEBUG
-	dabuf->ra = ra;
-	dabuf->target = XFS_BUF_TARGET(bps[0]);
-	dabuf->blkno = XFS_BUF_ADDR(bps[0]);
-#endif
 	if (nbuf == 1) {
 		dabuf->nbuf = 1;
 		bp = bps[0];
@@ -2281,23 +2273,6 @@ xfs_da_buf_make(int nbuf, xfs_buf_t **bps, inst_t *ra)
 				XFS_BUF_COUNT(bp));
 		}
 	}
-#ifdef XFS_DABUF_DEBUG
-	{
-		xfs_dabuf_t	*p;
-
-		spin_lock(&xfs_dabuf_global_lock);
-		for (p = xfs_dabuf_global_list; p; p = p->next) {
-			ASSERT(p->blkno != dabuf->blkno ||
-			       p->target != dabuf->target);
-		}
-		dabuf->prev = NULL;
-		if (xfs_dabuf_global_list)
-			xfs_dabuf_global_list->prev = dabuf;
-		dabuf->next = xfs_dabuf_global_list;
-		xfs_dabuf_global_list = dabuf;
-		spin_unlock(&xfs_dabuf_global_lock);
-	}
-#endif
 	return dabuf;
 }
 
@@ -2333,25 +2308,12 @@ xfs_da_buf_done(xfs_dabuf_t *dabuf)
 	ASSERT(dabuf->nbuf && dabuf->data && dabuf->bbcount && dabuf->bps[0]);
 	if (dabuf->dirty)
 		xfs_da_buf_clean(dabuf);
-	if (dabuf->nbuf > 1)
+	if (dabuf->nbuf > 1) {
 		kmem_free(dabuf->data);
-#ifdef XFS_DABUF_DEBUG
-	{
-		spin_lock(&xfs_dabuf_global_lock);
-		if (dabuf->prev)
-			dabuf->prev->next = dabuf->next;
-		else
-			xfs_dabuf_global_list = dabuf->next;
-		if (dabuf->next)
-			dabuf->next->prev = dabuf->prev;
-		spin_unlock(&xfs_dabuf_global_lock);
-	}
-	memset(dabuf, 0, XFS_DA_BUF_SIZE(dabuf->nbuf));
-#endif
-	if (dabuf->nbuf == 1)
-		kmem_zone_free(xfs_dabuf_zone, dabuf);
-	else
 		kmem_free(dabuf);
+	} else {
+		kmem_zone_free(xfs_dabuf_zone, dabuf);
+	}
 }
 
 /*
diff --git a/fs/xfs/xfs_da_btree.h b/fs/xfs/xfs_da_btree.h
index fe9f5a8c1d2..dbf7c074ae7 100644
--- a/fs/xfs/xfs_da_btree.h
+++ b/fs/xfs/xfs_da_btree.h
@@ -145,22 +145,11 @@ typedef struct xfs_dabuf {
 	short		dirty;		/* data needs to be copied back */
 	short		bbcount;	/* how large is data in bbs */
 	void		*data;		/* pointer for buffers' data */
-#ifdef XFS_DABUF_DEBUG
-	inst_t		*ra;		/* return address of caller to make */
-	struct xfs_dabuf *next;		/* next in global chain */
-	struct xfs_dabuf *prev;		/* previous in global chain */
-	struct xfs_buftarg *target;	/* device for buffer */
-	xfs_daddr_t	blkno;		/* daddr first in bps[0] */
-#endif
 	struct xfs_buf	*bps[1];	/* actually nbuf of these */
 } xfs_dabuf_t;
 #define	XFS_DA_BUF_SIZE(n)	\
 	(sizeof(xfs_dabuf_t) + sizeof(struct xfs_buf *) * ((n) - 1))
 
-#ifdef XFS_DABUF_DEBUG
-extern xfs_dabuf_t	*xfs_dabuf_global_list;
-#endif
-
 /*
  * Storage for holding state during Btree searches and split/join ops.
  *
@@ -248,6 +237,8 @@ int	xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
  * Utility routines.
  */
 int	xfs_da_grow_inode(xfs_da_args_t *args, xfs_dablk_t *new_blkno);
+int	xfs_da_grow_inode_int(struct xfs_da_args *args, xfs_fileoff_t *bno,
+			      int count);
 int	xfs_da_get_buf(struct xfs_trans *trans, struct xfs_inode *dp,
 			      xfs_dablk_t bno, xfs_daddr_t mappedbno,
 			      xfs_dabuf_t **bp, int whichfork);
diff --git a/fs/xfs/xfs_dir2.c b/fs/xfs/xfs_dir2.c
index dba7a71cedf..4580ce00aeb 100644
--- a/fs/xfs/xfs_dir2.c
+++ b/fs/xfs/xfs_dir2.c
@@ -24,20 +24,17 @@
 #include "xfs_trans.h"
 #include "xfs_sb.h"
 #include "xfs_ag.h"
-#include "xfs_dir2.h"
 #include "xfs_mount.h"
 #include "xfs_da_btree.h"
 #include "xfs_bmap_btree.h"
 #include "xfs_alloc_btree.h"
-#include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_inode_item.h"
 #include "xfs_bmap.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
-#include "xfs_dir2_node.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_format.h"
+#include "xfs_dir2_priv.h"
 #include "xfs_error.h"
 #include "xfs_vnodeops.h"
 #include "xfs_trace.h"
@@ -122,15 +119,15 @@ int
 xfs_dir_isempty(
 	xfs_inode_t	*dp)
 {
-	xfs_dir2_sf_t	*sfp;
+	xfs_dir2_sf_hdr_t	*sfp;
 
 	ASSERT((dp->i_d.di_mode & S_IFMT) == S_IFDIR);
 	if (dp->i_d.di_size == 0)	/* might happen during shutdown. */
 		return 1;
 	if (dp->i_d.di_size > XFS_IFORK_DSIZE(dp))
 		return 0;
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	return !sfp->hdr.count;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	return !sfp->count;
 }
 
 /*
@@ -500,129 +497,34 @@ xfs_dir_canenter(
 
 /*
  * Add a block to the directory.
- * This routine is for data and free blocks, not leaf/node blocks
- * which are handled by xfs_da_grow_inode.
+ *
+ * This routine is for data and free blocks, not leaf/node blocks which are
+ * handled by xfs_da_grow_inode.
  */
 int
 xfs_dir2_grow_inode(
-	xfs_da_args_t	*args,
-	int		space,		/* v2 dir's space XFS_DIR2_xxx_SPACE */
-	xfs_dir2_db_t	*dbp)		/* out: block number added */
+	struct xfs_da_args	*args,
+	int			space,	/* v2 dir's space XFS_DIR2_xxx_SPACE */
+	xfs_dir2_db_t		*dbp)	/* out: block number added */
 {
-	xfs_fileoff_t	bno;		/* directory offset of new block */
-	int		count;		/* count of filesystem blocks */
-	xfs_inode_t	*dp;		/* incore directory inode */
-	int		error;
-	int		got;		/* blocks actually mapped */
-	int		i;
-	xfs_bmbt_irec_t	map;		/* single structure for bmap */
-	int		mapi;		/* mapping index */
-	xfs_bmbt_irec_t	*mapp;		/* bmap mapping structure(s) */
-	xfs_mount_t	*mp;
-	int		nmap;		/* number of bmap entries */
-	xfs_trans_t	*tp;
-	xfs_drfsbno_t	nblks;
+	struct xfs_inode	*dp = args->dp;
+	struct xfs_mount	*mp = dp->i_mount;
+	xfs_fileoff_t		bno;	/* directory offset of new block */
+	int			count;	/* count of filesystem blocks */
+	int			error;
 
 	trace_xfs_dir2_grow_inode(args, space);
 
-	dp = args->dp;
-	tp = args->trans;
-	mp = dp->i_mount;
-	nblks = dp->i_d.di_nblocks;
 	/*
 	 * Set lowest possible block in the space requested.
 	 */
 	bno = XFS_B_TO_FSBT(mp, space * XFS_DIR2_SPACE_SIZE);
 	count = mp->m_dirblkfsbs;
-	/*
-	 * Find the first hole for our block.
-	 */
-	if ((error = xfs_bmap_first_unused(tp, dp, count, &bno, XFS_DATA_FORK)))
-		return error;
-	nmap = 1;
-	ASSERT(args->firstblock != NULL);
-	/*
-	 * Try mapping the new block contiguously (one extent).
-	 */
-	if ((error = xfs_bmapi(tp, dp, bno, count,
-			XFS_BMAPI_WRITE|XFS_BMAPI_METADATA|XFS_BMAPI_CONTIG,
-			args->firstblock, args->total, &map, &nmap,
-			args->flist)))
-		return error;
-	ASSERT(nmap <= 1);
-	if (nmap == 1) {
-		mapp = &map;
-		mapi = 1;
-	}
-	/*
-	 * Didn't work and this is a multiple-fsb directory block.
-	 * Try again with contiguous flag turned on.
-	 */
-	else if (nmap == 0 && count > 1) {
-		xfs_fileoff_t	b;	/* current file offset */
 
-		/*
-		 * Space for maximum number of mappings.
-		 */
-		mapp = kmem_alloc(sizeof(*mapp) * count, KM_SLEEP);
-		/*
-		 * Iterate until we get to the end of our block.
-		 */
-		for (b = bno, mapi = 0; b < bno + count; ) {
-			int	c;	/* current fsb count */
-
-			/*
-			 * Can't map more than MAX_NMAP at once.
-			 */
-			nmap = MIN(XFS_BMAP_MAX_NMAP, count);
-			c = (int)(bno + count - b);
-			if ((error = xfs_bmapi(tp, dp, b, c,
-					XFS_BMAPI_WRITE|XFS_BMAPI_METADATA,
-					args->firstblock, args->total,
-					&mapp[mapi], &nmap, args->flist))) {
-				kmem_free(mapp);
-				return error;
-			}
-			if (nmap < 1)
-				break;
-			/*
-			 * Add this bunch into our table, go to the next offset.
-			 */
-			mapi += nmap;
-			b = mapp[mapi - 1].br_startoff +
-			    mapp[mapi - 1].br_blockcount;
-		}
-	}
-	/*
-	 * Didn't work.
-	 */
-	else {
-		mapi = 0;
-		mapp = NULL;
-	}
-	/*
-	 * See how many fsb's we got.
-	 */
-	for (i = 0, got = 0; i < mapi; i++)
-		got += mapp[i].br_blockcount;
-	/*
-	 * Didn't get enough fsb's, or the first/last block's are wrong.
-	 */
-	if (got != count || mapp[0].br_startoff != bno ||
-	    mapp[mapi - 1].br_startoff + mapp[mapi - 1].br_blockcount !=
-	    bno + count) {
-		if (mapp != &map)
-			kmem_free(mapp);
-		return XFS_ERROR(ENOSPC);
-	}
-	/*
-	 * Done with the temporary mapping table.
-	 */
-	if (mapp != &map)
-		kmem_free(mapp);
+	error = xfs_da_grow_inode_int(args, &bno, count);
+	if (error)
+		return error;
 
-	/* account for newly allocated blocks in reserved blocks total */
-	args->total -= dp->i_d.di_nblocks - nblks;
 	*dbp = xfs_dir2_da_to_db(mp, (xfs_dablk_t)bno);
 
 	/*
@@ -634,7 +536,7 @@ xfs_dir2_grow_inode(
 		size = XFS_FSB_TO_B(mp, bno + count);
 		if (size > dp->i_d.di_size) {
 			dp->i_d.di_size = size;
-			xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
+			xfs_trans_log_inode(args->trans, dp, XFS_ILOG_CORE);
 		}
 	}
 	return 0;
diff --git a/fs/xfs/xfs_dir2.h b/fs/xfs/xfs_dir2.h
index 74a3b105768..e937d9991c1 100644
--- a/fs/xfs/xfs_dir2.h
+++ b/fs/xfs/xfs_dir2.h
@@ -16,49 +16,14 @@
  * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
  */
 #ifndef __XFS_DIR2_H__
-#define	__XFS_DIR2_H__
+#define __XFS_DIR2_H__
 
-struct uio;
-struct xfs_dabuf;
-struct xfs_da_args;
-struct xfs_dir2_put_args;
 struct xfs_bmap_free;
+struct xfs_da_args;
 struct xfs_inode;
 struct xfs_mount;
 struct xfs_trans;
 
-/*
- * Directory version 2.
- * There are 4 possible formats:
- *	shortform
- *	single block - data with embedded leaf at the end
- *	multiple data blocks, single leaf+freeindex block
- *	data blocks, node&leaf blocks (btree), freeindex blocks
- *
- *	The shortform format is in xfs_dir2_sf.h.
- *	The single block format is in xfs_dir2_block.h.
- *	The data block format is in xfs_dir2_data.h.
- *	The leaf and freeindex block formats are in xfs_dir2_leaf.h.
- *	Node blocks are the same as the other version, in xfs_da_btree.h.
- */
-
-/*
- * Byte offset in data block and shortform entry.
- */
-typedef	__uint16_t	xfs_dir2_data_off_t;
-#define	NULLDATAOFF	0xffffU
-typedef uint		xfs_dir2_data_aoff_t;	/* argument form */
-
-/*
- * Directory block number (logical dirblk in file)
- */
-typedef	__uint32_t	xfs_dir2_db_t;
-
-/*
- * Byte offset in a directory.
- */
-typedef	xfs_off_t	xfs_dir2_off_t;
-
 extern struct xfs_name	xfs_name_dotdot;
 
 /*
@@ -86,21 +51,10 @@ extern int xfs_dir_replace(struct xfs_trans *tp, struct xfs_inode *dp,
 				struct xfs_bmap_free *flist, xfs_extlen_t tot);
 extern int xfs_dir_canenter(struct xfs_trans *tp, struct xfs_inode *dp,
 				struct xfs_name *name, uint resblks);
-extern int xfs_dir_ino_validate(struct xfs_mount *mp, xfs_ino_t ino);
 
 /*
- * Utility routines for v2 directories.
+ * Direct call from the bmap code, bypassing the generic directory layer.
  */
-extern int xfs_dir2_grow_inode(struct xfs_da_args *args, int space,
-				xfs_dir2_db_t *dbp);
-extern int xfs_dir2_isblock(struct xfs_trans *tp, struct xfs_inode *dp,
-				int *vp);
-extern int xfs_dir2_isleaf(struct xfs_trans *tp, struct xfs_inode *dp,
-				int *vp);
-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 unsigned char *name, int len);
+extern int xfs_dir2_sf_to_block(struct xfs_da_args *args);
 
 #endif	/* __XFS_DIR2_H__ */
diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c
index 580d99cef9e..9245e029b8e 100644
--- a/fs/xfs/xfs_dir2_block.c
+++ b/fs/xfs/xfs_dir2_block.c
@@ -23,17 +23,14 @@
 #include "xfs_trans.h"
 #include "xfs_sb.h"
 #include "xfs_ag.h"
-#include "xfs_dir2.h"
 #include "xfs_mount.h"
 #include "xfs_da_btree.h"
 #include "xfs_bmap_btree.h"
-#include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_inode_item.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
+#include "xfs_dir2_format.h"
+#include "xfs_dir2_priv.h"
 #include "xfs_error.h"
 #include "xfs_trace.h"
 
@@ -67,7 +64,7 @@ xfs_dir2_block_addname(
 	xfs_da_args_t		*args)		/* directory op arguments */
 {
 	xfs_dir2_data_free_t	*bf;		/* bestfree table in block */
-	xfs_dir2_block_t	*block;		/* directory block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_leaf_entry_t	*blp;		/* block leaf entries */
 	xfs_dabuf_t		*bp;		/* buffer for block */
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
@@ -105,13 +102,13 @@ xfs_dir2_block_addname(
 		return error;
 	}
 	ASSERT(bp != NULL);
-	block = bp->data;
+	hdr = bp->data;
 	/*
 	 * Check the magic number, corrupted if wrong.
 	 */
-	if (unlikely(be32_to_cpu(block->hdr.magic) != XFS_DIR2_BLOCK_MAGIC)) {
+	if (unlikely(hdr->magic != cpu_to_be32(XFS_DIR2_BLOCK_MAGIC))) {
 		XFS_CORRUPTION_ERROR("xfs_dir2_block_addname",
-				     XFS_ERRLEVEL_LOW, mp, block);
+				     XFS_ERRLEVEL_LOW, mp, hdr);
 		xfs_da_brelse(tp, bp);
 		return XFS_ERROR(EFSCORRUPTED);
 	}
@@ -119,8 +116,8 @@ xfs_dir2_block_addname(
 	/*
 	 * Set up pointers to parts of the block.
 	 */
-	bf = block->hdr.bestfree;
-	btp = xfs_dir2_block_tail_p(mp, block);
+	bf = hdr->bestfree;
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * No stale entries?  Need space for entry and new leaf.
@@ -133,7 +130,7 @@ xfs_dir2_block_addname(
 		/*
 		 * Data object just before the first leaf entry.
 		 */
-		enddup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
+		enddup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp));
 		/*
 		 * If it's not free then can't do this add without cleaning up:
 		 * the space before the first leaf entry needs to be free so it
@@ -146,7 +143,7 @@ xfs_dir2_block_addname(
 		 */
 		else {
 			dup = (xfs_dir2_data_unused_t *)
-			      ((char *)block + be16_to_cpu(bf[0].offset));
+			      ((char *)hdr + be16_to_cpu(bf[0].offset));
 			if (dup == enddup) {
 				/*
 				 * It is the biggest freespace, is it too small
@@ -159,7 +156,7 @@ xfs_dir2_block_addname(
 					 */
 					if (be16_to_cpu(bf[1].length) >= len)
 						dup = (xfs_dir2_data_unused_t *)
-						      ((char *)block +
+						      ((char *)hdr +
 						       be16_to_cpu(bf[1].offset));
 					else
 						dup = NULL;
@@ -182,7 +179,7 @@ xfs_dir2_block_addname(
 	 */
 	else if (be16_to_cpu(bf[0].length) >= len) {
 		dup = (xfs_dir2_data_unused_t *)
-		      ((char *)block + be16_to_cpu(bf[0].offset));
+		      ((char *)hdr + be16_to_cpu(bf[0].offset));
 		compact = 0;
 	}
 	/*
@@ -196,7 +193,7 @@ xfs_dir2_block_addname(
 		/*
 		 * Data object just before the first leaf entry.
 		 */
-		dup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
+		dup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp));
 		/*
 		 * If it's not free then the data will go where the
 		 * leaf data starts now, if it works at all.
@@ -255,7 +252,8 @@ xfs_dir2_block_addname(
 			highstale = lfloghigh = -1;
 		     fromidx >= 0;
 		     fromidx--) {
-			if (be32_to_cpu(blp[fromidx].address) == XFS_DIR2_NULL_DATAPTR) {
+			if (blp[fromidx].address ==
+			    cpu_to_be32(XFS_DIR2_NULL_DATAPTR)) {
 				if (highstale == -1)
 					highstale = toidx;
 				else {
@@ -272,7 +270,7 @@ xfs_dir2_block_addname(
 		lfloghigh -= be32_to_cpu(btp->stale) - 1;
 		be32_add_cpu(&btp->count, -(be32_to_cpu(btp->stale) - 1));
 		xfs_dir2_data_make_free(tp, bp,
-			(xfs_dir2_data_aoff_t)((char *)blp - (char *)block),
+			(xfs_dir2_data_aoff_t)((char *)blp - (char *)hdr),
 			(xfs_dir2_data_aoff_t)((be32_to_cpu(btp->stale) - 1) * sizeof(*blp)),
 			&needlog, &needscan);
 		blp += be32_to_cpu(btp->stale) - 1;
@@ -282,7 +280,7 @@ xfs_dir2_block_addname(
 		 * This needs to happen before the next call to use_free.
 		 */
 		if (needscan) {
-			xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
+			xfs_dir2_data_freescan(mp, hdr, &needlog);
 			needscan = 0;
 		}
 	}
@@ -318,7 +316,7 @@ xfs_dir2_block_addname(
 		 */
 		xfs_dir2_data_use_free(tp, bp, enddup,
 			(xfs_dir2_data_aoff_t)
-			((char *)enddup - (char *)block + be16_to_cpu(enddup->length) -
+			((char *)enddup - (char *)hdr + be16_to_cpu(enddup->length) -
 			 sizeof(*blp)),
 			(xfs_dir2_data_aoff_t)sizeof(*blp),
 			&needlog, &needscan);
@@ -331,8 +329,7 @@ xfs_dir2_block_addname(
 		 * This needs to happen before the next call to use_free.
 		 */
 		if (needscan) {
-			xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block,
-				&needlog);
+			xfs_dir2_data_freescan(mp, hdr, &needlog);
 			needscan = 0;
 		}
 		/*
@@ -353,12 +350,14 @@ xfs_dir2_block_addname(
 	else {
 		for (lowstale = mid;
 		     lowstale >= 0 &&
-			be32_to_cpu(blp[lowstale].address) != XFS_DIR2_NULL_DATAPTR;
+			blp[lowstale].address !=
+			cpu_to_be32(XFS_DIR2_NULL_DATAPTR);
 		     lowstale--)
 			continue;
 		for (highstale = mid + 1;
 		     highstale < be32_to_cpu(btp->count) &&
-			be32_to_cpu(blp[highstale].address) != XFS_DIR2_NULL_DATAPTR &&
+			blp[highstale].address !=
+			cpu_to_be32(XFS_DIR2_NULL_DATAPTR) &&
 			(lowstale < 0 || mid - lowstale > highstale - mid);
 		     highstale++)
 			continue;
@@ -397,13 +396,13 @@ xfs_dir2_block_addname(
 	 */
 	blp[mid].hashval = cpu_to_be32(args->hashval);
 	blp[mid].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
-				(char *)dep - (char *)block));
+				(char *)dep - (char *)hdr));
 	xfs_dir2_block_log_leaf(tp, bp, lfloglow, lfloghigh);
 	/*
 	 * Mark space for the data entry used.
 	 */
 	xfs_dir2_data_use_free(tp, bp, dup,
-		(xfs_dir2_data_aoff_t)((char *)dup - (char *)block),
+		(xfs_dir2_data_aoff_t)((char *)dup - (char *)hdr),
 		(xfs_dir2_data_aoff_t)len, &needlog, &needscan);
 	/*
 	 * Create the new data entry.
@@ -412,12 +411,12 @@ xfs_dir2_block_addname(
 	dep->namelen = args->namelen;
 	memcpy(dep->name, args->name, args->namelen);
 	tagp = xfs_dir2_data_entry_tag_p(dep);
-	*tagp = cpu_to_be16((char *)dep - (char *)block);
+	*tagp = cpu_to_be16((char *)dep - (char *)hdr);
 	/*
 	 * Clean up the bestfree array and log the header, tail, and entry.
 	 */
 	if (needscan)
-		xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	if (needlog)
 		xfs_dir2_data_log_header(tp, bp);
 	xfs_dir2_block_log_tail(tp, bp);
@@ -437,7 +436,7 @@ xfs_dir2_block_getdents(
 	xfs_off_t		*offset,
 	filldir_t		filldir)
 {
-	xfs_dir2_block_t	*block;		/* directory block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dabuf_t		*bp;		/* buffer for block */
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
 	xfs_dir2_data_entry_t	*dep;		/* block data entry */
@@ -470,13 +469,13 @@ xfs_dir2_block_getdents(
 	 * We'll skip entries before this.
 	 */
 	wantoff = xfs_dir2_dataptr_to_off(mp, *offset);
-	block = bp->data;
+	hdr = bp->data;
 	xfs_dir2_data_check(dp, bp);
 	/*
 	 * Set up values for the loop.
 	 */
-	btp = xfs_dir2_block_tail_p(mp, block);
-	ptr = (char *)block->u;
+	btp = xfs_dir2_block_tail_p(mp, hdr);
+	ptr = (char *)(hdr + 1);
 	endptr = (char *)xfs_dir2_block_leaf_p(btp);
 
 	/*
@@ -502,11 +501,11 @@ xfs_dir2_block_getdents(
 		/*
 		 * The entry is before the desired starting point, skip it.
 		 */
-		if ((char *)dep - (char *)block < wantoff)
+		if ((char *)dep - (char *)hdr < wantoff)
 			continue;
 
 		cook = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
-					    (char *)dep - (char *)block);
+					    (char *)dep - (char *)hdr);
 
 		/*
 		 * If it didn't fit, set the final offset to here & return.
@@ -540,17 +539,14 @@ xfs_dir2_block_log_leaf(
 	int			first,		/* index of first logged leaf */
 	int			last)		/* index of last logged leaf */
 {
-	xfs_dir2_block_t	*block;		/* directory block structure */
-	xfs_dir2_leaf_entry_t	*blp;		/* block leaf entries */
-	xfs_dir2_block_tail_t	*btp;		/* block tail */
-	xfs_mount_t		*mp;		/* filesystem mount point */
+	xfs_dir2_data_hdr_t	*hdr = bp->data;
+	xfs_dir2_leaf_entry_t	*blp;
+	xfs_dir2_block_tail_t	*btp;
 
-	mp = tp->t_mountp;
-	block = bp->data;
-	btp = xfs_dir2_block_tail_p(mp, block);
+	btp = xfs_dir2_block_tail_p(tp->t_mountp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
-	xfs_da_log_buf(tp, bp, (uint)((char *)&blp[first] - (char *)block),
-		(uint)((char *)&blp[last + 1] - (char *)block - 1));
+	xfs_da_log_buf(tp, bp, (uint)((char *)&blp[first] - (char *)hdr),
+		(uint)((char *)&blp[last + 1] - (char *)hdr - 1));
 }
 
 /*
@@ -561,15 +557,12 @@ xfs_dir2_block_log_tail(
 	xfs_trans_t		*tp,		/* transaction structure */
 	xfs_dabuf_t		*bp)		/* block buffer */
 {
-	xfs_dir2_block_t	*block;		/* directory block structure */
-	xfs_dir2_block_tail_t	*btp;		/* block tail */
-	xfs_mount_t		*mp;		/* filesystem mount point */
+	xfs_dir2_data_hdr_t	*hdr = bp->data;
+	xfs_dir2_block_tail_t	*btp;
 
-	mp = tp->t_mountp;
-	block = bp->data;
-	btp = xfs_dir2_block_tail_p(mp, block);
-	xfs_da_log_buf(tp, bp, (uint)((char *)btp - (char *)block),
-		(uint)((char *)(btp + 1) - (char *)block - 1));
+	btp = xfs_dir2_block_tail_p(tp->t_mountp, hdr);
+	xfs_da_log_buf(tp, bp, (uint)((char *)btp - (char *)hdr),
+		(uint)((char *)(btp + 1) - (char *)hdr - 1));
 }
 
 /*
@@ -580,7 +573,7 @@ int						/* error */
 xfs_dir2_block_lookup(
 	xfs_da_args_t		*args)		/* dir lookup arguments */
 {
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_leaf_entry_t	*blp;		/* block leaf entries */
 	xfs_dabuf_t		*bp;		/* block buffer */
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
@@ -600,14 +593,14 @@ xfs_dir2_block_lookup(
 		return error;
 	dp = args->dp;
 	mp = dp->i_mount;
-	block = bp->data;
+	hdr = bp->data;
 	xfs_dir2_data_check(dp, bp);
-	btp = xfs_dir2_block_tail_p(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Get the offset from the leaf entry, to point to the data.
 	 */
-	dep = (xfs_dir2_data_entry_t *)((char *)block +
+	dep = (xfs_dir2_data_entry_t *)((char *)hdr +
 		xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
 	/*
 	 * Fill in inode number, CI name if appropriate, release the block.
@@ -628,7 +621,7 @@ xfs_dir2_block_lookup_int(
 	int			*entno)		/* returned entry number */
 {
 	xfs_dir2_dataptr_t	addr;		/* data entry address */
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_leaf_entry_t	*blp;		/* block leaf entries */
 	xfs_dabuf_t		*bp;		/* block buffer */
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
@@ -654,9 +647,9 @@ xfs_dir2_block_lookup_int(
 		return error;
 	}
 	ASSERT(bp != NULL);
-	block = bp->data;
+	hdr = bp->data;
 	xfs_dir2_data_check(dp, bp);
-	btp = xfs_dir2_block_tail_p(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Loop doing a binary search for our hash value.
@@ -694,7 +687,7 @@ xfs_dir2_block_lookup_int(
 		 * Get pointer to the entry from the leaf.
 		 */
 		dep = (xfs_dir2_data_entry_t *)
-			((char *)block + xfs_dir2_dataptr_to_off(mp, addr));
+			((char *)hdr + xfs_dir2_dataptr_to_off(mp, addr));
 		/*
 		 * Compare name and if it's an exact match, return the index
 		 * and buffer. If it's the first case-insensitive match, store
@@ -733,7 +726,7 @@ int						/* error */
 xfs_dir2_block_removename(
 	xfs_da_args_t		*args)		/* directory operation args */
 {
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_leaf_entry_t	*blp;		/* block leaf pointer */
 	xfs_dabuf_t		*bp;		/* block buffer */
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
@@ -760,20 +753,20 @@ xfs_dir2_block_removename(
 	dp = args->dp;
 	tp = args->trans;
 	mp = dp->i_mount;
-	block = bp->data;
-	btp = xfs_dir2_block_tail_p(mp, block);
+	hdr = bp->data;
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Point to the data entry using the leaf entry.
 	 */
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)block + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
+	      ((char *)hdr + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
 	/*
 	 * Mark the data entry's space free.
 	 */
 	needlog = needscan = 0;
 	xfs_dir2_data_make_free(tp, bp,
-		(xfs_dir2_data_aoff_t)((char *)dep - (char *)block),
+		(xfs_dir2_data_aoff_t)((char *)dep - (char *)hdr),
 		xfs_dir2_data_entsize(dep->namelen), &needlog, &needscan);
 	/*
 	 * Fix up the block tail.
@@ -789,15 +782,15 @@ xfs_dir2_block_removename(
 	 * Fix up bestfree, log the header if necessary.
 	 */
 	if (needscan)
-		xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	if (needlog)
 		xfs_dir2_data_log_header(tp, bp);
 	xfs_dir2_data_check(dp, bp);
 	/*
 	 * See if the size as a shortform is good enough.
 	 */
-	if ((size = xfs_dir2_block_sfsize(dp, block, &sfh)) >
-	    XFS_IFORK_DSIZE(dp)) {
+	size = xfs_dir2_block_sfsize(dp, hdr, &sfh);
+	if (size > XFS_IFORK_DSIZE(dp)) {
 		xfs_da_buf_done(bp);
 		return 0;
 	}
@@ -815,7 +808,7 @@ int						/* error */
 xfs_dir2_block_replace(
 	xfs_da_args_t		*args)		/* directory operation args */
 {
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_leaf_entry_t	*blp;		/* block leaf entries */
 	xfs_dabuf_t		*bp;		/* block buffer */
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
@@ -836,14 +829,14 @@ xfs_dir2_block_replace(
 	}
 	dp = args->dp;
 	mp = dp->i_mount;
-	block = bp->data;
-	btp = xfs_dir2_block_tail_p(mp, block);
+	hdr = bp->data;
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Point to the data entry we need to change.
 	 */
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)block + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
+	      ((char *)hdr + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
 	ASSERT(be64_to_cpu(dep->inumber) != args->inumber);
 	/*
 	 * Change the inode number to the new value.
@@ -882,7 +875,7 @@ xfs_dir2_leaf_to_block(
 	xfs_dabuf_t		*dbp)		/* data buffer */
 {
 	__be16			*bestsp;	/* leaf bests table */
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
 	xfs_inode_t		*dp;		/* incore directory inode */
 	xfs_dir2_data_unused_t	*dup;		/* unused data entry */
@@ -906,7 +899,7 @@ xfs_dir2_leaf_to_block(
 	tp = args->trans;
 	mp = dp->i_mount;
 	leaf = lbp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
 	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	/*
 	 * If there are data blocks other than the first one, take this
@@ -917,7 +910,7 @@ xfs_dir2_leaf_to_block(
 	while (dp->i_d.di_size > mp->m_dirblksize) {
 		bestsp = xfs_dir2_leaf_bests_p(ltp);
 		if (be16_to_cpu(bestsp[be32_to_cpu(ltp->bestcount) - 1]) ==
-		    mp->m_dirblksize - (uint)sizeof(block->hdr)) {
+		    mp->m_dirblksize - (uint)sizeof(*hdr)) {
 			if ((error =
 			    xfs_dir2_leaf_trim_data(args, lbp,
 				    (xfs_dir2_db_t)(be32_to_cpu(ltp->bestcount) - 1))))
@@ -935,18 +928,18 @@ xfs_dir2_leaf_to_block(
 		    XFS_DATA_FORK))) {
 		goto out;
 	}
-	block = dbp->data;
-	ASSERT(be32_to_cpu(block->hdr.magic) == XFS_DIR2_DATA_MAGIC);
+	hdr = dbp->data;
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC));
 	/*
 	 * Size of the "leaf" area in the block.
 	 */
-	size = (uint)sizeof(block->tail) +
+	size = (uint)sizeof(xfs_dir2_block_tail_t) +
 	       (uint)sizeof(*lep) * (be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale));
 	/*
 	 * Look at the last data entry.
 	 */
-	tagp = (__be16 *)((char *)block + mp->m_dirblksize) - 1;
-	dup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
+	tagp = (__be16 *)((char *)hdr + mp->m_dirblksize) - 1;
+	dup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp));
 	/*
 	 * If it's not free or is too short we can't do it.
 	 */
@@ -958,7 +951,7 @@ xfs_dir2_leaf_to_block(
 	/*
 	 * Start converting it to block form.
 	 */
-	block->hdr.magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
+	hdr->magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
 	needlog = 1;
 	needscan = 0;
 	/*
@@ -969,7 +962,7 @@ xfs_dir2_leaf_to_block(
 	/*
 	 * Initialize the block tail.
 	 */
-	btp = xfs_dir2_block_tail_p(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	btp->count = cpu_to_be32(be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale));
 	btp->stale = 0;
 	xfs_dir2_block_log_tail(tp, dbp);
@@ -978,7 +971,8 @@ xfs_dir2_leaf_to_block(
 	 */
 	lep = xfs_dir2_block_leaf_p(btp);
 	for (from = to = 0; from < be16_to_cpu(leaf->hdr.count); from++) {
-		if (be32_to_cpu(leaf->ents[from].address) == XFS_DIR2_NULL_DATAPTR)
+		if (leaf->ents[from].address ==
+		    cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
 			continue;
 		lep[to++] = leaf->ents[from];
 	}
@@ -988,7 +982,7 @@ xfs_dir2_leaf_to_block(
 	 * Scan the bestfree if we need it and log the data block header.
 	 */
 	if (needscan)
-		xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	if (needlog)
 		xfs_dir2_data_log_header(tp, dbp);
 	/*
@@ -1002,8 +996,8 @@ xfs_dir2_leaf_to_block(
 	/*
 	 * Now see if the resulting block can be shrunken to shortform.
 	 */
-	if ((size = xfs_dir2_block_sfsize(dp, block, &sfh)) >
-	    XFS_IFORK_DSIZE(dp)) {
+	size = xfs_dir2_block_sfsize(dp, hdr, &sfh);
+	if (size > XFS_IFORK_DSIZE(dp)) {
 		error = 0;
 		goto out;
 	}
@@ -1024,12 +1018,10 @@ xfs_dir2_sf_to_block(
 	xfs_da_args_t		*args)		/* operation arguments */
 {
 	xfs_dir2_db_t		blkno;		/* dir-relative block # (0) */
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_leaf_entry_t	*blp;		/* block leaf entries */
 	xfs_dabuf_t		*bp;		/* block buffer */
 	xfs_dir2_block_tail_t	*btp;		/* block tail pointer */
-	char			*buf;		/* sf buffer */
-	int			buf_len;
 	xfs_dir2_data_entry_t	*dep;		/* data entry pointer */
 	xfs_inode_t		*dp;		/* incore directory inode */
 	int			dummy;		/* trash */
@@ -1043,7 +1035,8 @@ xfs_dir2_sf_to_block(
 	int			newoffset;	/* offset from current entry */
 	int			offset;		/* target block offset */
 	xfs_dir2_sf_entry_t	*sfep;		/* sf entry pointer */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*oldsfp;	/* old shortform header  */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform header  */
 	__be16			*tagp;		/* end of data entry */
 	xfs_trans_t		*tp;		/* transaction pointer */
 	struct xfs_name		name;
@@ -1061,32 +1054,30 @@ xfs_dir2_sf_to_block(
 		ASSERT(XFS_FORCED_SHUTDOWN(mp));
 		return XFS_ERROR(EIO);
 	}
+
+	oldsfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(oldsfp->i8count));
+
 	/*
-	 * Copy the directory into the stack buffer.
+	 * Copy the directory into a temporary buffer.
 	 * Then pitch the incore inode data so we can make extents.
 	 */
+	sfp = kmem_alloc(dp->i_df.if_bytes, KM_SLEEP);
+	memcpy(sfp, oldsfp, dp->i_df.if_bytes);
 
-	buf_len = dp->i_df.if_bytes;
-	buf = kmem_alloc(buf_len, KM_SLEEP);
-
-	memcpy(buf, sfp, buf_len);
-	xfs_idata_realloc(dp, -buf_len, XFS_DATA_FORK);
+	xfs_idata_realloc(dp, -dp->i_df.if_bytes, XFS_DATA_FORK);
 	dp->i_d.di_size = 0;
 	xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
-	/*
-	 * Reset pointer - old sfp is gone.
-	 */
-	sfp = (xfs_dir2_sf_t *)buf;
+
 	/*
 	 * Add block 0 to the inode.
 	 */
 	error = xfs_dir2_grow_inode(args, XFS_DIR2_DATA_SPACE, &blkno);
 	if (error) {
-		kmem_free(buf);
+		kmem_free(sfp);
 		return error;
 	}
 	/*
@@ -1094,21 +1085,21 @@ xfs_dir2_sf_to_block(
 	 */
 	error = xfs_dir2_data_init(args, blkno, &bp);
 	if (error) {
-		kmem_free(buf);
+		kmem_free(sfp);
 		return error;
 	}
-	block = bp->data;
-	block->hdr.magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
+	hdr = bp->data;
+	hdr->magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
 	/*
 	 * Compute size of block "tail" area.
 	 */
 	i = (uint)sizeof(*btp) +
-	    (sfp->hdr.count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t);
+	    (sfp->count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t);
 	/*
 	 * The whole thing is initialized to free by the init routine.
 	 * Say we're using the leaf and tail area.
 	 */
-	dup = (xfs_dir2_data_unused_t *)block->u;
+	dup = (xfs_dir2_data_unused_t *)(hdr + 1);
 	needlog = needscan = 0;
 	xfs_dir2_data_use_free(tp, bp, dup, mp->m_dirblksize - i, i, &needlog,
 		&needscan);
@@ -1116,50 +1107,51 @@ xfs_dir2_sf_to_block(
 	/*
 	 * Fill in the tail.
 	 */
-	btp = xfs_dir2_block_tail_p(mp, block);
-	btp->count = cpu_to_be32(sfp->hdr.count + 2);	/* ., .. */
+	btp = xfs_dir2_block_tail_p(mp, hdr);
+	btp->count = cpu_to_be32(sfp->count + 2);	/* ., .. */
 	btp->stale = 0;
 	blp = xfs_dir2_block_leaf_p(btp);
-	endoffset = (uint)((char *)blp - (char *)block);
+	endoffset = (uint)((char *)blp - (char *)hdr);
 	/*
 	 * Remove the freespace, we'll manage it.
 	 */
 	xfs_dir2_data_use_free(tp, bp, dup,
-		(xfs_dir2_data_aoff_t)((char *)dup - (char *)block),
+		(xfs_dir2_data_aoff_t)((char *)dup - (char *)hdr),
 		be16_to_cpu(dup->length), &needlog, &needscan);
 	/*
 	 * Create entry for .
 	 */
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)block + XFS_DIR2_DATA_DOT_OFFSET);
+	      ((char *)hdr + XFS_DIR2_DATA_DOT_OFFSET);
 	dep->inumber = cpu_to_be64(dp->i_ino);
 	dep->namelen = 1;
 	dep->name[0] = '.';
 	tagp = xfs_dir2_data_entry_tag_p(dep);
-	*tagp = cpu_to_be16((char *)dep - (char *)block);
+	*tagp = cpu_to_be16((char *)dep - (char *)hdr);
 	xfs_dir2_data_log_entry(tp, bp, dep);
 	blp[0].hashval = cpu_to_be32(xfs_dir_hash_dot);
 	blp[0].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
-				(char *)dep - (char *)block));
+				(char *)dep - (char *)hdr));
 	/*
 	 * Create entry for ..
 	 */
 	dep = (xfs_dir2_data_entry_t *)
-		((char *)block + XFS_DIR2_DATA_DOTDOT_OFFSET);
-	dep->inumber = cpu_to_be64(xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent));
+		((char *)hdr + XFS_DIR2_DATA_DOTDOT_OFFSET);
+	dep->inumber = cpu_to_be64(xfs_dir2_sf_get_parent_ino(sfp));
 	dep->namelen = 2;
 	dep->name[0] = dep->name[1] = '.';
 	tagp = xfs_dir2_data_entry_tag_p(dep);
-	*tagp = cpu_to_be16((char *)dep - (char *)block);
+	*tagp = cpu_to_be16((char *)dep - (char *)hdr);
 	xfs_dir2_data_log_entry(tp, bp, dep);
 	blp[1].hashval = cpu_to_be32(xfs_dir_hash_dotdot);
 	blp[1].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
-				(char *)dep - (char *)block));
+				(char *)dep - (char *)hdr));
 	offset = XFS_DIR2_DATA_FIRST_OFFSET;
 	/*
 	 * Loop over existing entries, stuff them in.
 	 */
-	if ((i = 0) == sfp->hdr.count)
+	i = 0;
+	if (!sfp->count)
 		sfep = NULL;
 	else
 		sfep = xfs_dir2_sf_firstentry(sfp);
@@ -1179,43 +1171,40 @@ xfs_dir2_sf_to_block(
 		 * There should be a hole here, make one.
 		 */
 		if (offset < newoffset) {
-			dup = (xfs_dir2_data_unused_t *)
-			      ((char *)block + offset);
+			dup = (xfs_dir2_data_unused_t *)((char *)hdr + offset);
 			dup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 			dup->length = cpu_to_be16(newoffset - offset);
 			*xfs_dir2_data_unused_tag_p(dup) = cpu_to_be16(
-				((char *)dup - (char *)block));
+				((char *)dup - (char *)hdr));
 			xfs_dir2_data_log_unused(tp, bp, dup);
-			(void)xfs_dir2_data_freeinsert((xfs_dir2_data_t *)block,
-				dup, &dummy);
+			xfs_dir2_data_freeinsert(hdr, dup, &dummy);
 			offset += be16_to_cpu(dup->length);
 			continue;
 		}
 		/*
 		 * Copy a real entry.
 		 */
-		dep = (xfs_dir2_data_entry_t *)((char *)block + newoffset);
-		dep->inumber = cpu_to_be64(xfs_dir2_sf_get_inumber(sfp,
-				xfs_dir2_sf_inumberp(sfep)));
+		dep = (xfs_dir2_data_entry_t *)((char *)hdr + newoffset);
+		dep->inumber = cpu_to_be64(xfs_dir2_sfe_get_ino(sfp, sfep));
 		dep->namelen = sfep->namelen;
 		memcpy(dep->name, sfep->name, dep->namelen);
 		tagp = xfs_dir2_data_entry_tag_p(dep);
-		*tagp = cpu_to_be16((char *)dep - (char *)block);
+		*tagp = cpu_to_be16((char *)dep - (char *)hdr);
 		xfs_dir2_data_log_entry(tp, bp, dep);
 		name.name = sfep->name;
 		name.len = sfep->namelen;
 		blp[2 + i].hashval = cpu_to_be32(mp->m_dirnameops->
 							hashname(&name));
 		blp[2 + i].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
-						 (char *)dep - (char *)block));
-		offset = (int)((char *)(tagp + 1) - (char *)block);
-		if (++i == sfp->hdr.count)
+						 (char *)dep - (char *)hdr));
+		offset = (int)((char *)(tagp + 1) - (char *)hdr);
+		if (++i == sfp->count)
 			sfep = NULL;
 		else
 			sfep = xfs_dir2_sf_nextentry(sfp, sfep);
 	}
 	/* Done with the temporary buffer */
-	kmem_free(buf);
+	kmem_free(sfp);
 	/*
 	 * Sort the leaf entries by hash value.
 	 */
diff --git a/fs/xfs/xfs_dir2_block.h b/fs/xfs/xfs_dir2_block.h
deleted file mode 100644
index 10e68967638..00000000000
--- a/fs/xfs/xfs_dir2_block.h
+++ /dev/null
@@ -1,92 +0,0 @@
-/*
- * Copyright (c) 2000-2001,2005 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_DIR2_BLOCK_H__
-#define	__XFS_DIR2_BLOCK_H__
-
-/*
- * xfs_dir2_block.h
- * Directory version 2, single block format structures
- */
-
-struct uio;
-struct xfs_dabuf;
-struct xfs_da_args;
-struct xfs_dir2_data_hdr;
-struct xfs_dir2_leaf_entry;
-struct xfs_inode;
-struct xfs_mount;
-struct xfs_trans;
-
-/*
- * The single block format is as follows:
- * xfs_dir2_data_hdr_t structure
- * xfs_dir2_data_entry_t and xfs_dir2_data_unused_t structures
- * xfs_dir2_leaf_entry_t structures
- * xfs_dir2_block_tail_t structure
- */
-
-#define	XFS_DIR2_BLOCK_MAGIC	0x58443242	/* XD2B: for one block dirs */
-
-typedef struct xfs_dir2_block_tail {
-	__be32		count;			/* count of leaf entries */
-	__be32		stale;			/* count of stale lf entries */
-} xfs_dir2_block_tail_t;
-
-/*
- * Generic single-block structure, for xfs_db.
- */
-typedef struct xfs_dir2_block {
-	xfs_dir2_data_hdr_t	hdr;		/* magic XFS_DIR2_BLOCK_MAGIC */
-	xfs_dir2_data_union_t	u[1];
-	xfs_dir2_leaf_entry_t	leaf[1];
-	xfs_dir2_block_tail_t	tail;
-} xfs_dir2_block_t;
-
-/*
- * Pointer to the leaf header embedded in a data block (1-block format)
- */
-static inline xfs_dir2_block_tail_t *
-xfs_dir2_block_tail_p(struct xfs_mount *mp, xfs_dir2_block_t *block)
-{
-	return (((xfs_dir2_block_tail_t *)
-		((char *)(block) + (mp)->m_dirblksize)) - 1);
-}
-
-/*
- * Pointer to the leaf entries embedded in a data block (1-block format)
- */
-static inline struct xfs_dir2_leaf_entry *
-xfs_dir2_block_leaf_p(xfs_dir2_block_tail_t *btp)
-{
-	return ((struct xfs_dir2_leaf_entry *)btp) - be32_to_cpu(btp->count);
-}
-
-/*
- * Function declarations.
- */
-extern int xfs_dir2_block_addname(struct xfs_da_args *args);
-extern int xfs_dir2_block_getdents(struct xfs_inode *dp, void *dirent,
-				   xfs_off_t *offset, filldir_t filldir);
-extern int xfs_dir2_block_lookup(struct xfs_da_args *args);
-extern int xfs_dir2_block_removename(struct xfs_da_args *args);
-extern int xfs_dir2_block_replace(struct xfs_da_args *args);
-extern int xfs_dir2_leaf_to_block(struct xfs_da_args *args,
-				  struct xfs_dabuf *lbp, struct xfs_dabuf *dbp);
-extern int xfs_dir2_sf_to_block(struct xfs_da_args *args);
-
-#endif	/* __XFS_DIR2_BLOCK_H__ */
diff --git a/fs/xfs/xfs_dir2_data.c b/fs/xfs/xfs_dir2_data.c
index 921595b84f5..5bbe2a8a023 100644
--- a/fs/xfs/xfs_dir2_data.c
+++ b/fs/xfs/xfs_dir2_data.c
@@ -23,18 +23,18 @@
 #include "xfs_trans.h"
 #include "xfs_sb.h"
 #include "xfs_ag.h"
-#include "xfs_dir2.h"
 #include "xfs_mount.h"
 #include "xfs_da_btree.h"
 #include "xfs_bmap_btree.h"
-#include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
+#include "xfs_dir2_format.h"
+#include "xfs_dir2_priv.h"
 #include "xfs_error.h"
 
+STATIC xfs_dir2_data_free_t *
+xfs_dir2_data_freefind(xfs_dir2_data_hdr_t *hdr, xfs_dir2_data_unused_t *dup);
+
 #ifdef DEBUG
 /*
  * Check the consistency of the data block.
@@ -50,7 +50,7 @@ xfs_dir2_data_check(
 	xfs_dir2_data_free_t	*bf;		/* bestfree table */
 	xfs_dir2_block_tail_t	*btp=NULL;	/* block tail */
 	int			count;		/* count of entries found */
-	xfs_dir2_data_t		*d;		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dir2_data_entry_t	*dep;		/* data entry */
 	xfs_dir2_data_free_t	*dfp;		/* bestfree entry */
 	xfs_dir2_data_unused_t	*dup;		/* unused entry */
@@ -66,17 +66,19 @@ xfs_dir2_data_check(
 	struct xfs_name		name;
 
 	mp = dp->i_mount;
-	d = bp->data;
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
-	bf = d->hdr.bestfree;
-	p = (char *)d->u;
-	if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC) {
-		btp = xfs_dir2_block_tail_p(mp, (xfs_dir2_block_t *)d);
+	hdr = bp->data;
+	bf = hdr->bestfree;
+	p = (char *)(hdr + 1);
+
+	if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) {
+		btp = xfs_dir2_block_tail_p(mp, hdr);
 		lep = xfs_dir2_block_leaf_p(btp);
 		endp = (char *)lep;
-	} else
-		endp = (char *)d + mp->m_dirblksize;
+	} else {
+		ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC));
+		endp = (char *)hdr + mp->m_dirblksize;
+	}
+
 	count = lastfree = freeseen = 0;
 	/*
 	 * Account for zero bestfree entries.
@@ -108,8 +110,8 @@ xfs_dir2_data_check(
 		if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) {
 			ASSERT(lastfree == 0);
 			ASSERT(be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)) ==
-			       (char *)dup - (char *)d);
-			dfp = xfs_dir2_data_freefind(d, dup);
+			       (char *)dup - (char *)hdr);
+			dfp = xfs_dir2_data_freefind(hdr, dup);
 			if (dfp) {
 				i = (int)(dfp - bf);
 				ASSERT((freeseen & (1 << i)) == 0);
@@ -132,13 +134,13 @@ xfs_dir2_data_check(
 		ASSERT(dep->namelen != 0);
 		ASSERT(xfs_dir_ino_validate(mp, be64_to_cpu(dep->inumber)) == 0);
 		ASSERT(be16_to_cpu(*xfs_dir2_data_entry_tag_p(dep)) ==
-		       (char *)dep - (char *)d);
+		       (char *)dep - (char *)hdr);
 		count++;
 		lastfree = 0;
-		if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC) {
+		if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) {
 			addr = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 				(xfs_dir2_data_aoff_t)
-				((char *)dep - (char *)d));
+				((char *)dep - (char *)hdr));
 			name.name = dep->name;
 			name.len = dep->namelen;
 			hash = mp->m_dirnameops->hashname(&name);
@@ -155,9 +157,10 @@ xfs_dir2_data_check(
 	 * Need to have seen all the entries and all the bestfree slots.
 	 */
 	ASSERT(freeseen == 7);
-	if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC) {
+	if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) {
 		for (i = stale = 0; i < be32_to_cpu(btp->count); i++) {
-			if (be32_to_cpu(lep[i].address) == XFS_DIR2_NULL_DATAPTR)
+			if (lep[i].address ==
+			    cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
 				stale++;
 			if (i > 0)
 				ASSERT(be32_to_cpu(lep[i].hashval) >= be32_to_cpu(lep[i - 1].hashval));
@@ -172,9 +175,9 @@ xfs_dir2_data_check(
  * Given a data block and an unused entry from that block,
  * return the bestfree entry if any that corresponds to it.
  */
-xfs_dir2_data_free_t *
+STATIC xfs_dir2_data_free_t *
 xfs_dir2_data_freefind(
-	xfs_dir2_data_t		*d,		/* data block */
+	xfs_dir2_data_hdr_t	*hdr,		/* data block */
 	xfs_dir2_data_unused_t	*dup)		/* data unused entry */
 {
 	xfs_dir2_data_free_t	*dfp;		/* bestfree entry */
@@ -184,17 +187,17 @@ xfs_dir2_data_freefind(
 	int			seenzero;	/* saw a 0 bestfree entry */
 #endif
 
-	off = (xfs_dir2_data_aoff_t)((char *)dup - (char *)d);
+	off = (xfs_dir2_data_aoff_t)((char *)dup - (char *)hdr);
 #if defined(DEBUG) && defined(__KERNEL__)
 	/*
 	 * Validate some consistency in the bestfree table.
 	 * Check order, non-overlapping entries, and if we find the
 	 * one we're looking for it has to be exact.
 	 */
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
-	for (dfp = &d->hdr.bestfree[0], seenzero = matched = 0;
-	     dfp < &d->hdr.bestfree[XFS_DIR2_DATA_FD_COUNT];
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+	for (dfp = &hdr->bestfree[0], seenzero = matched = 0;
+	     dfp < &hdr->bestfree[XFS_DIR2_DATA_FD_COUNT];
 	     dfp++) {
 		if (!dfp->offset) {
 			ASSERT(!dfp->length);
@@ -210,7 +213,7 @@ xfs_dir2_data_freefind(
 		else
 			ASSERT(be16_to_cpu(dfp->offset) + be16_to_cpu(dfp->length) <= off);
 		ASSERT(matched || be16_to_cpu(dfp->length) >= be16_to_cpu(dup->length));
-		if (dfp > &d->hdr.bestfree[0])
+		if (dfp > &hdr->bestfree[0])
 			ASSERT(be16_to_cpu(dfp[-1].length) >= be16_to_cpu(dfp[0].length));
 	}
 #endif
@@ -219,13 +222,13 @@ xfs_dir2_data_freefind(
 	 * it can't be there since they're sorted.
 	 */
 	if (be16_to_cpu(dup->length) <
-	    be16_to_cpu(d->hdr.bestfree[XFS_DIR2_DATA_FD_COUNT - 1].length))
+	    be16_to_cpu(hdr->bestfree[XFS_DIR2_DATA_FD_COUNT - 1].length))
 		return NULL;
 	/*
 	 * Look at the three bestfree entries for our guy.
 	 */
-	for (dfp = &d->hdr.bestfree[0];
-	     dfp < &d->hdr.bestfree[XFS_DIR2_DATA_FD_COUNT];
+	for (dfp = &hdr->bestfree[0];
+	     dfp < &hdr->bestfree[XFS_DIR2_DATA_FD_COUNT];
 	     dfp++) {
 		if (!dfp->offset)
 			return NULL;
@@ -243,7 +246,7 @@ xfs_dir2_data_freefind(
  */
 xfs_dir2_data_free_t *				/* entry inserted */
 xfs_dir2_data_freeinsert(
-	xfs_dir2_data_t		*d,		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr,		/* data block pointer */
 	xfs_dir2_data_unused_t	*dup,		/* unused space */
 	int			*loghead)	/* log the data header (out) */
 {
@@ -251,12 +254,13 @@ xfs_dir2_data_freeinsert(
 	xfs_dir2_data_free_t	new;		/* new bestfree entry */
 
 #ifdef __KERNEL__
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
 #endif
-	dfp = d->hdr.bestfree;
+	dfp = hdr->bestfree;
 	new.length = dup->length;
-	new.offset = cpu_to_be16((char *)dup - (char *)d);
+	new.offset = cpu_to_be16((char *)dup - (char *)hdr);
+
 	/*
 	 * Insert at position 0, 1, or 2; or not at all.
 	 */
@@ -286,36 +290,36 @@ xfs_dir2_data_freeinsert(
  */
 STATIC void
 xfs_dir2_data_freeremove(
-	xfs_dir2_data_t		*d,		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr,		/* data block header */
 	xfs_dir2_data_free_t	*dfp,		/* bestfree entry pointer */
 	int			*loghead)	/* out: log data header */
 {
 #ifdef __KERNEL__
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
 #endif
 	/*
 	 * It's the first entry, slide the next 2 up.
 	 */
-	if (dfp == &d->hdr.bestfree[0]) {
-		d->hdr.bestfree[0] = d->hdr.bestfree[1];
-		d->hdr.bestfree[1] = d->hdr.bestfree[2];
+	if (dfp == &hdr->bestfree[0]) {
+		hdr->bestfree[0] = hdr->bestfree[1];
+		hdr->bestfree[1] = hdr->bestfree[2];
 	}
 	/*
 	 * It's the second entry, slide the 3rd entry up.
 	 */
-	else if (dfp == &d->hdr.bestfree[1])
-		d->hdr.bestfree[1] = d->hdr.bestfree[2];
+	else if (dfp == &hdr->bestfree[1])
+		hdr->bestfree[1] = hdr->bestfree[2];
 	/*
 	 * Must be the last entry.
 	 */
 	else
-		ASSERT(dfp == &d->hdr.bestfree[2]);
+		ASSERT(dfp == &hdr->bestfree[2]);
 	/*
 	 * Clear the 3rd entry, must be zero now.
 	 */
-	d->hdr.bestfree[2].length = 0;
-	d->hdr.bestfree[2].offset = 0;
+	hdr->bestfree[2].length = 0;
+	hdr->bestfree[2].offset = 0;
 	*loghead = 1;
 }
 
@@ -325,7 +329,7 @@ xfs_dir2_data_freeremove(
 void
 xfs_dir2_data_freescan(
 	xfs_mount_t		*mp,		/* filesystem mount point */
-	xfs_dir2_data_t		*d,		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr,		/* data block header */
 	int			*loghead)	/* out: log data header */
 {
 	xfs_dir2_block_tail_t	*btp;		/* block tail */
@@ -335,23 +339,23 @@ xfs_dir2_data_freescan(
 	char			*p;		/* current entry pointer */
 
 #ifdef __KERNEL__
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
 #endif
 	/*
 	 * Start by clearing the table.
 	 */
-	memset(d->hdr.bestfree, 0, sizeof(d->hdr.bestfree));
+	memset(hdr->bestfree, 0, sizeof(hdr->bestfree));
 	*loghead = 1;
 	/*
 	 * Set up pointers.
 	 */
-	p = (char *)d->u;
-	if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC) {
-		btp = xfs_dir2_block_tail_p(mp, (xfs_dir2_block_t *)d);
+	p = (char *)(hdr + 1);
+	if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) {
+		btp = xfs_dir2_block_tail_p(mp, hdr);
 		endp = (char *)xfs_dir2_block_leaf_p(btp);
 	} else
-		endp = (char *)d + mp->m_dirblksize;
+		endp = (char *)hdr + mp->m_dirblksize;
 	/*
 	 * Loop over the block's entries.
 	 */
@@ -361,9 +365,9 @@ xfs_dir2_data_freescan(
 		 * If it's a free entry, insert it.
 		 */
 		if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) {
-			ASSERT((char *)dup - (char *)d ==
+			ASSERT((char *)dup - (char *)hdr ==
 			       be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)));
-			xfs_dir2_data_freeinsert(d, dup, loghead);
+			xfs_dir2_data_freeinsert(hdr, dup, loghead);
 			p += be16_to_cpu(dup->length);
 		}
 		/*
@@ -371,7 +375,7 @@ xfs_dir2_data_freescan(
 		 */
 		else {
 			dep = (xfs_dir2_data_entry_t *)p;
-			ASSERT((char *)dep - (char *)d ==
+			ASSERT((char *)dep - (char *)hdr ==
 			       be16_to_cpu(*xfs_dir2_data_entry_tag_p(dep)));
 			p += xfs_dir2_data_entsize(dep->namelen);
 		}
@@ -389,7 +393,7 @@ xfs_dir2_data_init(
 	xfs_dabuf_t		**bpp)		/* output block buffer */
 {
 	xfs_dabuf_t		*bp;		/* block buffer */
-	xfs_dir2_data_t		*d;		/* pointer to block */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_inode_t		*dp;		/* incore directory inode */
 	xfs_dir2_data_unused_t	*dup;		/* unused entry pointer */
 	int			error;		/* error return value */
@@ -410,26 +414,28 @@ xfs_dir2_data_init(
 		return error;
 	}
 	ASSERT(bp != NULL);
+
 	/*
 	 * Initialize the header.
 	 */
-	d = bp->data;
-	d->hdr.magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
-	d->hdr.bestfree[0].offset = cpu_to_be16(sizeof(d->hdr));
+	hdr = bp->data;
+	hdr->magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
+	hdr->bestfree[0].offset = cpu_to_be16(sizeof(*hdr));
 	for (i = 1; i < XFS_DIR2_DATA_FD_COUNT; i++) {
-		d->hdr.bestfree[i].length = 0;
-		d->hdr.bestfree[i].offset = 0;
+		hdr->bestfree[i].length = 0;
+		hdr->bestfree[i].offset = 0;
 	}
+
 	/*
 	 * Set up an unused entry for the block's body.
 	 */
-	dup = &d->u[0].unused;
+	dup = (xfs_dir2_data_unused_t *)(hdr + 1);
 	dup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 
-	t=mp->m_dirblksize - (uint)sizeof(d->hdr);
-	d->hdr.bestfree[0].length = cpu_to_be16(t);
+	t = mp->m_dirblksize - (uint)sizeof(*hdr);
+	hdr->bestfree[0].length = cpu_to_be16(t);
 	dup->length = cpu_to_be16(t);
-	*xfs_dir2_data_unused_tag_p(dup) = cpu_to_be16((char *)dup - (char *)d);
+	*xfs_dir2_data_unused_tag_p(dup) = cpu_to_be16((char *)dup - (char *)hdr);
 	/*
 	 * Log it and return it.
 	 */
@@ -448,14 +454,14 @@ xfs_dir2_data_log_entry(
 	xfs_dabuf_t		*bp,		/* block buffer */
 	xfs_dir2_data_entry_t	*dep)		/* data entry pointer */
 {
-	xfs_dir2_data_t		*d;		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr = bp->data;
+
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
 
-	d = bp->data;
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
-	xfs_da_log_buf(tp, bp, (uint)((char *)dep - (char *)d),
+	xfs_da_log_buf(tp, bp, (uint)((char *)dep - (char *)hdr),
 		(uint)((char *)(xfs_dir2_data_entry_tag_p(dep) + 1) -
-		       (char *)d - 1));
+		       (char *)hdr - 1));
 }
 
 /*
@@ -466,13 +472,12 @@ xfs_dir2_data_log_header(
 	xfs_trans_t		*tp,		/* transaction pointer */
 	xfs_dabuf_t		*bp)		/* block buffer */
 {
-	xfs_dir2_data_t		*d;		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr = bp->data;
 
-	d = bp->data;
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
-	xfs_da_log_buf(tp, bp, (uint)((char *)&d->hdr - (char *)d),
-		(uint)(sizeof(d->hdr) - 1));
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+
+	xfs_da_log_buf(tp, bp, 0, sizeof(*hdr) - 1);
 }
 
 /*
@@ -484,23 +489,23 @@ xfs_dir2_data_log_unused(
 	xfs_dabuf_t		*bp,		/* block buffer */
 	xfs_dir2_data_unused_t	*dup)		/* data unused pointer */
 {
-	xfs_dir2_data_t		*d;		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr = bp->data;
+
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
 
-	d = bp->data;
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
 	/*
 	 * Log the first part of the unused entry.
 	 */
-	xfs_da_log_buf(tp, bp, (uint)((char *)dup - (char *)d),
+	xfs_da_log_buf(tp, bp, (uint)((char *)dup - (char *)hdr),
 		(uint)((char *)&dup->length + sizeof(dup->length) -
-		       1 - (char *)d));
+		       1 - (char *)hdr));
 	/*
 	 * Log the end (tag) of the unused entry.
 	 */
 	xfs_da_log_buf(tp, bp,
-		(uint)((char *)xfs_dir2_data_unused_tag_p(dup) - (char *)d),
-		(uint)((char *)xfs_dir2_data_unused_tag_p(dup) - (char *)d +
+		(uint)((char *)xfs_dir2_data_unused_tag_p(dup) - (char *)hdr),
+		(uint)((char *)xfs_dir2_data_unused_tag_p(dup) - (char *)hdr +
 		       sizeof(xfs_dir2_data_off_t) - 1));
 }
 
@@ -517,7 +522,7 @@ xfs_dir2_data_make_free(
 	int			*needlogp,	/* out: log header */
 	int			*needscanp)	/* out: regen bestfree */
 {
-	xfs_dir2_data_t		*d;		/* data block pointer */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block pointer */
 	xfs_dir2_data_free_t	*dfp;		/* bestfree pointer */
 	char			*endptr;	/* end of data area */
 	xfs_mount_t		*mp;		/* filesystem mount point */
@@ -527,28 +532,29 @@ xfs_dir2_data_make_free(
 	xfs_dir2_data_unused_t	*prevdup;	/* unused entry before us */
 
 	mp = tp->t_mountp;
-	d = bp->data;
+	hdr = bp->data;
+
 	/*
 	 * Figure out where the end of the data area is.
 	 */
-	if (be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC)
-		endptr = (char *)d + mp->m_dirblksize;
+	if (hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC))
+		endptr = (char *)hdr + mp->m_dirblksize;
 	else {
 		xfs_dir2_block_tail_t	*btp;	/* block tail */
 
-		ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
-		btp = xfs_dir2_block_tail_p(mp, (xfs_dir2_block_t *)d);
+		ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+		btp = xfs_dir2_block_tail_p(mp, hdr);
 		endptr = (char *)xfs_dir2_block_leaf_p(btp);
 	}
 	/*
 	 * If this isn't the start of the block, then back up to
 	 * the previous entry and see if it's free.
 	 */
-	if (offset > sizeof(d->hdr)) {
+	if (offset > sizeof(*hdr)) {
 		__be16			*tagp;	/* tag just before us */
 
-		tagp = (__be16 *)((char *)d + offset) - 1;
-		prevdup = (xfs_dir2_data_unused_t *)((char *)d + be16_to_cpu(*tagp));
+		tagp = (__be16 *)((char *)hdr + offset) - 1;
+		prevdup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp));
 		if (be16_to_cpu(prevdup->freetag) != XFS_DIR2_DATA_FREE_TAG)
 			prevdup = NULL;
 	} else
@@ -557,9 +563,9 @@ xfs_dir2_data_make_free(
 	 * If this isn't the end of the block, see if the entry after
 	 * us is free.
 	 */
-	if ((char *)d + offset + len < endptr) {
+	if ((char *)hdr + offset + len < endptr) {
 		postdup =
-			(xfs_dir2_data_unused_t *)((char *)d + offset + len);
+			(xfs_dir2_data_unused_t *)((char *)hdr + offset + len);
 		if (be16_to_cpu(postdup->freetag) != XFS_DIR2_DATA_FREE_TAG)
 			postdup = NULL;
 	} else
@@ -576,21 +582,21 @@ xfs_dir2_data_make_free(
 		/*
 		 * See if prevdup and/or postdup are in bestfree table.
 		 */
-		dfp = xfs_dir2_data_freefind(d, prevdup);
-		dfp2 = xfs_dir2_data_freefind(d, postdup);
+		dfp = xfs_dir2_data_freefind(hdr, prevdup);
+		dfp2 = xfs_dir2_data_freefind(hdr, postdup);
 		/*
 		 * We need a rescan unless there are exactly 2 free entries
 		 * namely our two.  Then we know what's happening, otherwise
 		 * since the third bestfree is there, there might be more
 		 * entries.
 		 */
-		needscan = (d->hdr.bestfree[2].length != 0);
+		needscan = (hdr->bestfree[2].length != 0);
 		/*
 		 * Fix up the new big freespace.
 		 */
 		be16_add_cpu(&prevdup->length, len + be16_to_cpu(postdup->length));
 		*xfs_dir2_data_unused_tag_p(prevdup) =
-			cpu_to_be16((char *)prevdup - (char *)d);
+			cpu_to_be16((char *)prevdup - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, prevdup);
 		if (!needscan) {
 			/*
@@ -600,18 +606,18 @@ xfs_dir2_data_make_free(
 			 * Remove entry 1 first then entry 0.
 			 */
 			ASSERT(dfp && dfp2);
-			if (dfp == &d->hdr.bestfree[1]) {
-				dfp = &d->hdr.bestfree[0];
+			if (dfp == &hdr->bestfree[1]) {
+				dfp = &hdr->bestfree[0];
 				ASSERT(dfp2 == dfp);
-				dfp2 = &d->hdr.bestfree[1];
+				dfp2 = &hdr->bestfree[1];
 			}
-			xfs_dir2_data_freeremove(d, dfp2, needlogp);
-			xfs_dir2_data_freeremove(d, dfp, needlogp);
+			xfs_dir2_data_freeremove(hdr, dfp2, needlogp);
+			xfs_dir2_data_freeremove(hdr, dfp, needlogp);
 			/*
 			 * Now insert the new entry.
 			 */
-			dfp = xfs_dir2_data_freeinsert(d, prevdup, needlogp);
-			ASSERT(dfp == &d->hdr.bestfree[0]);
+			dfp = xfs_dir2_data_freeinsert(hdr, prevdup, needlogp);
+			ASSERT(dfp == &hdr->bestfree[0]);
 			ASSERT(dfp->length == prevdup->length);
 			ASSERT(!dfp[1].length);
 			ASSERT(!dfp[2].length);
@@ -621,10 +627,10 @@ xfs_dir2_data_make_free(
 	 * The entry before us is free, merge with it.
 	 */
 	else if (prevdup) {
-		dfp = xfs_dir2_data_freefind(d, prevdup);
+		dfp = xfs_dir2_data_freefind(hdr, prevdup);
 		be16_add_cpu(&prevdup->length, len);
 		*xfs_dir2_data_unused_tag_p(prevdup) =
-			cpu_to_be16((char *)prevdup - (char *)d);
+			cpu_to_be16((char *)prevdup - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, prevdup);
 		/*
 		 * If the previous entry was in the table, the new entry
@@ -632,27 +638,27 @@ xfs_dir2_data_make_free(
 		 * the old one and add the new one.
 		 */
 		if (dfp) {
-			xfs_dir2_data_freeremove(d, dfp, needlogp);
-			(void)xfs_dir2_data_freeinsert(d, prevdup, needlogp);
+			xfs_dir2_data_freeremove(hdr, dfp, needlogp);
+			xfs_dir2_data_freeinsert(hdr, prevdup, needlogp);
 		}
 		/*
 		 * Otherwise we need a scan if the new entry is big enough.
 		 */
 		else {
 			needscan = be16_to_cpu(prevdup->length) >
-				   be16_to_cpu(d->hdr.bestfree[2].length);
+				   be16_to_cpu(hdr->bestfree[2].length);
 		}
 	}
 	/*
 	 * The following entry is free, merge with it.
 	 */
 	else if (postdup) {
-		dfp = xfs_dir2_data_freefind(d, postdup);
-		newdup = (xfs_dir2_data_unused_t *)((char *)d + offset);
+		dfp = xfs_dir2_data_freefind(hdr, postdup);
+		newdup = (xfs_dir2_data_unused_t *)((char *)hdr + offset);
 		newdup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup->length = cpu_to_be16(len + be16_to_cpu(postdup->length));
 		*xfs_dir2_data_unused_tag_p(newdup) =
-			cpu_to_be16((char *)newdup - (char *)d);
+			cpu_to_be16((char *)newdup - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		/*
 		 * If the following entry was in the table, the new entry
@@ -660,28 +666,28 @@ xfs_dir2_data_make_free(
 		 * the old one and add the new one.
 		 */
 		if (dfp) {
-			xfs_dir2_data_freeremove(d, dfp, needlogp);
-			(void)xfs_dir2_data_freeinsert(d, newdup, needlogp);
+			xfs_dir2_data_freeremove(hdr, dfp, needlogp);
+			xfs_dir2_data_freeinsert(hdr, newdup, needlogp);
 		}
 		/*
 		 * Otherwise we need a scan if the new entry is big enough.
 		 */
 		else {
 			needscan = be16_to_cpu(newdup->length) >
-				   be16_to_cpu(d->hdr.bestfree[2].length);
+				   be16_to_cpu(hdr->bestfree[2].length);
 		}
 	}
 	/*
 	 * Neither neighbor is free.  Make a new entry.
 	 */
 	else {
-		newdup = (xfs_dir2_data_unused_t *)((char *)d + offset);
+		newdup = (xfs_dir2_data_unused_t *)((char *)hdr + offset);
 		newdup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup->length = cpu_to_be16(len);
 		*xfs_dir2_data_unused_tag_p(newdup) =
-			cpu_to_be16((char *)newdup - (char *)d);
+			cpu_to_be16((char *)newdup - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
-		(void)xfs_dir2_data_freeinsert(d, newdup, needlogp);
+		xfs_dir2_data_freeinsert(hdr, newdup, needlogp);
 	}
 	*needscanp = needscan;
 }
@@ -699,7 +705,7 @@ xfs_dir2_data_use_free(
 	int			*needlogp,	/* out: need to log header */
 	int			*needscanp)	/* out: need regen bestfree */
 {
-	xfs_dir2_data_t		*d;		/* data block */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dir2_data_free_t	*dfp;		/* bestfree pointer */
 	int			matchback;	/* matches end of freespace */
 	int			matchfront;	/* matches start of freespace */
@@ -708,24 +714,24 @@ xfs_dir2_data_use_free(
 	xfs_dir2_data_unused_t	*newdup2;	/* another new unused entry */
 	int			oldlen;		/* old unused entry's length */
 
-	d = bp->data;
-	ASSERT(be32_to_cpu(d->hdr.magic) == XFS_DIR2_DATA_MAGIC ||
-	       be32_to_cpu(d->hdr.magic) == XFS_DIR2_BLOCK_MAGIC);
+	hdr = bp->data;
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+	       hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
 	ASSERT(be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG);
-	ASSERT(offset >= (char *)dup - (char *)d);
-	ASSERT(offset + len <= (char *)dup + be16_to_cpu(dup->length) - (char *)d);
-	ASSERT((char *)dup - (char *)d == be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)));
+	ASSERT(offset >= (char *)dup - (char *)hdr);
+	ASSERT(offset + len <= (char *)dup + be16_to_cpu(dup->length) - (char *)hdr);
+	ASSERT((char *)dup - (char *)hdr == be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)));
 	/*
 	 * Look up the entry in the bestfree table.
 	 */
-	dfp = xfs_dir2_data_freefind(d, dup);
+	dfp = xfs_dir2_data_freefind(hdr, dup);
 	oldlen = be16_to_cpu(dup->length);
-	ASSERT(dfp || oldlen <= be16_to_cpu(d->hdr.bestfree[2].length));
+	ASSERT(dfp || oldlen <= be16_to_cpu(hdr->bestfree[2].length));
 	/*
 	 * Check for alignment with front and back of the entry.
 	 */
-	matchfront = (char *)dup - (char *)d == offset;
-	matchback = (char *)dup + oldlen - (char *)d == offset + len;
+	matchfront = (char *)dup - (char *)hdr == offset;
+	matchback = (char *)dup + oldlen - (char *)hdr == offset + len;
 	ASSERT(*needscanp == 0);
 	needscan = 0;
 	/*
@@ -734,9 +740,9 @@ xfs_dir2_data_use_free(
 	 */
 	if (matchfront && matchback) {
 		if (dfp) {
-			needscan = (d->hdr.bestfree[2].offset != 0);
+			needscan = (hdr->bestfree[2].offset != 0);
 			if (!needscan)
-				xfs_dir2_data_freeremove(d, dfp, needlogp);
+				xfs_dir2_data_freeremove(hdr, dfp, needlogp);
 		}
 	}
 	/*
@@ -744,27 +750,27 @@ xfs_dir2_data_use_free(
 	 * Make a new entry with the remaining freespace.
 	 */
 	else if (matchfront) {
-		newdup = (xfs_dir2_data_unused_t *)((char *)d + offset + len);
+		newdup = (xfs_dir2_data_unused_t *)((char *)hdr + offset + len);
 		newdup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup->length = cpu_to_be16(oldlen - len);
 		*xfs_dir2_data_unused_tag_p(newdup) =
-			cpu_to_be16((char *)newdup - (char *)d);
+			cpu_to_be16((char *)newdup - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		/*
 		 * If it was in the table, remove it and add the new one.
 		 */
 		if (dfp) {
-			xfs_dir2_data_freeremove(d, dfp, needlogp);
-			dfp = xfs_dir2_data_freeinsert(d, newdup, needlogp);
+			xfs_dir2_data_freeremove(hdr, dfp, needlogp);
+			dfp = xfs_dir2_data_freeinsert(hdr, newdup, needlogp);
 			ASSERT(dfp != NULL);
 			ASSERT(dfp->length == newdup->length);
-			ASSERT(be16_to_cpu(dfp->offset) == (char *)newdup - (char *)d);
+			ASSERT(be16_to_cpu(dfp->offset) == (char *)newdup - (char *)hdr);
 			/*
 			 * If we got inserted at the last slot,
 			 * that means we don't know if there was a better
 			 * choice for the last slot, or not.  Rescan.
 			 */
-			needscan = dfp == &d->hdr.bestfree[2];
+			needscan = dfp == &hdr->bestfree[2];
 		}
 	}
 	/*
@@ -773,25 +779,25 @@ xfs_dir2_data_use_free(
 	 */
 	else if (matchback) {
 		newdup = dup;
-		newdup->length = cpu_to_be16(((char *)d + offset) - (char *)newdup);
+		newdup->length = cpu_to_be16(((char *)hdr + offset) - (char *)newdup);
 		*xfs_dir2_data_unused_tag_p(newdup) =
-			cpu_to_be16((char *)newdup - (char *)d);
+			cpu_to_be16((char *)newdup - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
 		/*
 		 * If it was in the table, remove it and add the new one.
 		 */
 		if (dfp) {
-			xfs_dir2_data_freeremove(d, dfp, needlogp);
-			dfp = xfs_dir2_data_freeinsert(d, newdup, needlogp);
+			xfs_dir2_data_freeremove(hdr, dfp, needlogp);
+			dfp = xfs_dir2_data_freeinsert(hdr, newdup, needlogp);
 			ASSERT(dfp != NULL);
 			ASSERT(dfp->length == newdup->length);
-			ASSERT(be16_to_cpu(dfp->offset) == (char *)newdup - (char *)d);
+			ASSERT(be16_to_cpu(dfp->offset) == (char *)newdup - (char *)hdr);
 			/*
 			 * If we got inserted at the last slot,
 			 * that means we don't know if there was a better
 			 * choice for the last slot, or not.  Rescan.
 			 */
-			needscan = dfp == &d->hdr.bestfree[2];
+			needscan = dfp == &hdr->bestfree[2];
 		}
 	}
 	/*
@@ -800,15 +806,15 @@ xfs_dir2_data_use_free(
 	 */
 	else {
 		newdup = dup;
-		newdup->length = cpu_to_be16(((char *)d + offset) - (char *)newdup);
+		newdup->length = cpu_to_be16(((char *)hdr + offset) - (char *)newdup);
 		*xfs_dir2_data_unused_tag_p(newdup) =
-			cpu_to_be16((char *)newdup - (char *)d);
+			cpu_to_be16((char *)newdup - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, newdup);
-		newdup2 = (xfs_dir2_data_unused_t *)((char *)d + offset + len);
+		newdup2 = (xfs_dir2_data_unused_t *)((char *)hdr + offset + len);
 		newdup2->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
 		newdup2->length = cpu_to_be16(oldlen - len - be16_to_cpu(newdup->length));
 		*xfs_dir2_data_unused_tag_p(newdup2) =
-			cpu_to_be16((char *)newdup2 - (char *)d);
+			cpu_to_be16((char *)newdup2 - (char *)hdr);
 		xfs_dir2_data_log_unused(tp, bp, newdup2);
 		/*
 		 * If the old entry was in the table, we need to scan
@@ -819,13 +825,12 @@ xfs_dir2_data_use_free(
 		 * the 2 new will work.
 		 */
 		if (dfp) {
-			needscan = (d->hdr.bestfree[2].length != 0);
+			needscan = (hdr->bestfree[2].length != 0);
 			if (!needscan) {
-				xfs_dir2_data_freeremove(d, dfp, needlogp);
-				(void)xfs_dir2_data_freeinsert(d, newdup,
-					needlogp);
-				(void)xfs_dir2_data_freeinsert(d, newdup2,
-					needlogp);
+				xfs_dir2_data_freeremove(hdr, dfp, needlogp);
+				xfs_dir2_data_freeinsert(hdr, newdup, needlogp);
+				xfs_dir2_data_freeinsert(hdr, newdup2,
+							 needlogp);
 			}
 		}
 	}
diff --git a/fs/xfs/xfs_dir2_data.h b/fs/xfs/xfs_dir2_data.h
deleted file mode 100644
index efbc290c7fe..00000000000
--- a/fs/xfs/xfs_dir2_data.h
+++ /dev/null
@@ -1,184 +0,0 @@
-/*
- * Copyright (c) 2000,2005 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_DIR2_DATA_H__
-#define	__XFS_DIR2_DATA_H__
-
-/*
- * Directory format 2, data block structures.
- */
-
-struct xfs_dabuf;
-struct xfs_da_args;
-struct xfs_inode;
-struct xfs_trans;
-
-/*
- * Constants.
- */
-#define	XFS_DIR2_DATA_MAGIC	0x58443244	/* XD2D: for multiblock dirs */
-#define	XFS_DIR2_DATA_ALIGN_LOG	3		/* i.e., 8 bytes */
-#define	XFS_DIR2_DATA_ALIGN	(1 << XFS_DIR2_DATA_ALIGN_LOG)
-#define	XFS_DIR2_DATA_FREE_TAG	0xffff
-#define	XFS_DIR2_DATA_FD_COUNT	3
-
-/*
- * Directory address space divided into sections,
- * spaces separated by 32GB.
- */
-#define	XFS_DIR2_SPACE_SIZE	(1ULL << (32 + XFS_DIR2_DATA_ALIGN_LOG))
-#define	XFS_DIR2_DATA_SPACE	0
-#define	XFS_DIR2_DATA_OFFSET	(XFS_DIR2_DATA_SPACE * XFS_DIR2_SPACE_SIZE)
-#define	XFS_DIR2_DATA_FIRSTDB(mp)	\
-	xfs_dir2_byte_to_db(mp, XFS_DIR2_DATA_OFFSET)
-
-/*
- * Offsets of . and .. in data space (always block 0)
- */
-#define	XFS_DIR2_DATA_DOT_OFFSET	\
-	((xfs_dir2_data_aoff_t)sizeof(xfs_dir2_data_hdr_t))
-#define	XFS_DIR2_DATA_DOTDOT_OFFSET	\
-	(XFS_DIR2_DATA_DOT_OFFSET + xfs_dir2_data_entsize(1))
-#define	XFS_DIR2_DATA_FIRST_OFFSET		\
-	(XFS_DIR2_DATA_DOTDOT_OFFSET + xfs_dir2_data_entsize(2))
-
-/*
- * Structures.
- */
-
-/*
- * Describe a free area in the data block.
- * The freespace will be formatted as a xfs_dir2_data_unused_t.
- */
-typedef struct xfs_dir2_data_free {
-	__be16			offset;		/* start of freespace */
-	__be16			length;		/* length of freespace */
-} xfs_dir2_data_free_t;
-
-/*
- * Header for the data blocks.
- * Always at the beginning of a directory-sized block.
- * The code knows that XFS_DIR2_DATA_FD_COUNT is 3.
- */
-typedef struct xfs_dir2_data_hdr {
-	__be32			magic;		/* XFS_DIR2_DATA_MAGIC */
-						/* or XFS_DIR2_BLOCK_MAGIC */
-	xfs_dir2_data_free_t	bestfree[XFS_DIR2_DATA_FD_COUNT];
-} xfs_dir2_data_hdr_t;
-
-/*
- * Active entry in a data block.  Aligned to 8 bytes.
- * Tag appears as the last 2 bytes.
- */
-typedef struct xfs_dir2_data_entry {
-	__be64			inumber;	/* inode number */
-	__u8			namelen;	/* name length */
-	__u8			name[1];	/* name bytes, no null */
-						/* variable offset */
-	__be16			tag;		/* starting offset of us */
-} xfs_dir2_data_entry_t;
-
-/*
- * Unused entry in a data block.  Aligned to 8 bytes.
- * Tag appears as the last 2 bytes.
- */
-typedef struct xfs_dir2_data_unused {
-	__be16			freetag;	/* XFS_DIR2_DATA_FREE_TAG */
-	__be16			length;		/* total free length */
-						/* variable offset */
-	__be16			tag;		/* starting offset of us */
-} xfs_dir2_data_unused_t;
-
-typedef union {
-	xfs_dir2_data_entry_t	entry;
-	xfs_dir2_data_unused_t	unused;
-} xfs_dir2_data_union_t;
-
-/*
- * Generic data block structure, for xfs_db.
- */
-typedef struct xfs_dir2_data {
-	xfs_dir2_data_hdr_t	hdr;		/* magic XFS_DIR2_DATA_MAGIC */
-	xfs_dir2_data_union_t	u[1];
-} xfs_dir2_data_t;
-
-/*
- * Macros.
- */
-
-/*
- * Size of a data entry.
- */
-static inline int xfs_dir2_data_entsize(int n)
-{
-	return (int)roundup(offsetof(xfs_dir2_data_entry_t, name[0]) + (n) + \
-		 (uint)sizeof(xfs_dir2_data_off_t), XFS_DIR2_DATA_ALIGN);
-}
-
-/*
- * Pointer to an entry's tag word.
- */
-static inline __be16 *
-xfs_dir2_data_entry_tag_p(xfs_dir2_data_entry_t *dep)
-{
-	return (__be16 *)((char *)dep +
-		xfs_dir2_data_entsize(dep->namelen) - sizeof(__be16));
-}
-
-/*
- * Pointer to a freespace's tag word.
- */
-static inline __be16 *
-xfs_dir2_data_unused_tag_p(xfs_dir2_data_unused_t *dup)
-{
-	return (__be16 *)((char *)dup +
-			be16_to_cpu(dup->length) - sizeof(__be16));
-}
-
-/*
- * Function declarations.
- */
-#ifdef DEBUG
-extern void xfs_dir2_data_check(struct xfs_inode *dp, struct xfs_dabuf *bp);
-#else
-#define	xfs_dir2_data_check(dp,bp)
-#endif
-extern xfs_dir2_data_free_t *xfs_dir2_data_freefind(xfs_dir2_data_t *d,
-				xfs_dir2_data_unused_t *dup);
-extern xfs_dir2_data_free_t *xfs_dir2_data_freeinsert(xfs_dir2_data_t *d,
-				xfs_dir2_data_unused_t *dup, int *loghead);
-extern void xfs_dir2_data_freescan(struct xfs_mount *mp, xfs_dir2_data_t *d,
-				int *loghead);
-extern int xfs_dir2_data_init(struct xfs_da_args *args, xfs_dir2_db_t blkno,
-				struct xfs_dabuf **bpp);
-extern void xfs_dir2_data_log_entry(struct xfs_trans *tp, struct xfs_dabuf *bp,
-				xfs_dir2_data_entry_t *dep);
-extern void xfs_dir2_data_log_header(struct xfs_trans *tp,
-				struct xfs_dabuf *bp);
-extern void xfs_dir2_data_log_unused(struct xfs_trans *tp, struct xfs_dabuf *bp,
-				xfs_dir2_data_unused_t *dup);
-extern void xfs_dir2_data_make_free(struct xfs_trans *tp, struct xfs_dabuf *bp,
-				xfs_dir2_data_aoff_t offset,
-				xfs_dir2_data_aoff_t len, int *needlogp,
-				int *needscanp);
-extern void xfs_dir2_data_use_free(struct xfs_trans *tp, struct xfs_dabuf *bp,
-			       xfs_dir2_data_unused_t *dup,
-			       xfs_dir2_data_aoff_t offset,
-			       xfs_dir2_data_aoff_t len, int *needlogp,
-			       int *needscanp);
-
-#endif	/* __XFS_DIR2_DATA_H__ */
diff --git a/fs/xfs/xfs_dir2_format.h b/fs/xfs/xfs_dir2_format.h
new file mode 100644
index 00000000000..07270981f48
--- /dev/null
+++ b/fs/xfs/xfs_dir2_format.h
@@ -0,0 +1,597 @@
+/*
+ * Copyright (c) 2000-2001,2005 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_DIR2_FORMAT_H__
+#define __XFS_DIR2_FORMAT_H__
+
+/*
+ * Directory version 2.
+ *
+ * There are 4 possible formats:
+ *  - shortform - embedded into the inode
+ *  - single block - data with embedded leaf at the end
+ *  - multiple data blocks, single leaf+freeindex block
+ *  - data blocks, node and leaf blocks (btree), freeindex blocks
+ *
+ * Note: many node blocks structures and constants are shared with the attr
+ * code and defined in xfs_da_btree.h.
+ */
+
+#define	XFS_DIR2_BLOCK_MAGIC	0x58443242	/* XD2B: single block dirs */
+#define	XFS_DIR2_DATA_MAGIC	0x58443244	/* XD2D: multiblock dirs */
+#define	XFS_DIR2_FREE_MAGIC	0x58443246	/* XD2F: free index blocks */
+
+/*
+ * Byte offset in data block and shortform entry.
+ */
+typedef	__uint16_t	xfs_dir2_data_off_t;
+#define	NULLDATAOFF	0xffffU
+typedef uint		xfs_dir2_data_aoff_t;	/* argument form */
+
+/*
+ * Normalized offset (in a data block) of the entry, really xfs_dir2_data_off_t.
+ * Only need 16 bits, this is the byte offset into the single block form.
+ */
+typedef struct { __uint8_t i[2]; } __arch_pack xfs_dir2_sf_off_t;
+
+/*
+ * Offset in data space of a data entry.
+ */
+typedef	__uint32_t	xfs_dir2_dataptr_t;
+#define	XFS_DIR2_MAX_DATAPTR	((xfs_dir2_dataptr_t)0xffffffff)
+#define	XFS_DIR2_NULL_DATAPTR	((xfs_dir2_dataptr_t)0)
+
+/*
+ * Byte offset in a directory.
+ */
+typedef	xfs_off_t	xfs_dir2_off_t;
+
+/*
+ * Directory block number (logical dirblk in file)
+ */
+typedef	__uint32_t	xfs_dir2_db_t;
+
+/*
+ * Inode number stored as 8 8-bit values.
+ */
+typedef	struct { __uint8_t i[8]; } xfs_dir2_ino8_t;
+
+/*
+ * Inode number stored as 4 8-bit values.
+ * Works a lot of the time, when all the inode numbers in a directory
+ * fit in 32 bits.
+ */
+typedef struct { __uint8_t i[4]; } xfs_dir2_ino4_t;
+
+typedef union {
+	xfs_dir2_ino8_t	i8;
+	xfs_dir2_ino4_t	i4;
+} xfs_dir2_inou_t;
+#define	XFS_DIR2_MAX_SHORT_INUM	((xfs_ino_t)0xffffffffULL)
+
+/*
+ * Directory layout when stored internal to an inode.
+ *
+ * Small directories are packed as tightly as possible so as to fit into the
+ * literal area of the inode.  These "shortform" directories consist of a
+ * single xfs_dir2_sf_hdr header followed by zero or more xfs_dir2_sf_entry
+ * structures.  Due the different inode number storage size and the variable
+ * length name field in the xfs_dir2_sf_entry all these structure are
+ * variable length, and the accessors in this file should be used to iterate
+ * over them.
+ */
+typedef struct xfs_dir2_sf_hdr {
+	__uint8_t		count;		/* count of entries */
+	__uint8_t		i8count;	/* count of 8-byte inode #s */
+	xfs_dir2_inou_t		parent;		/* parent dir inode number */
+} __arch_pack xfs_dir2_sf_hdr_t;
+
+typedef struct xfs_dir2_sf_entry {
+	__u8			namelen;	/* actual name length */
+	xfs_dir2_sf_off_t	offset;		/* saved offset */
+	__u8			name[];		/* name, variable size */
+	/*
+	 * A xfs_dir2_ino8_t or xfs_dir2_ino4_t follows here, at a
+	 * variable offset after the name.
+	 */
+} __arch_pack xfs_dir2_sf_entry_t;
+
+static inline int xfs_dir2_sf_hdr_size(int i8count)
+{
+	return sizeof(struct xfs_dir2_sf_hdr) -
+		(i8count == 0) *
+		(sizeof(xfs_dir2_ino8_t) - sizeof(xfs_dir2_ino4_t));
+}
+
+static inline xfs_dir2_data_aoff_t
+xfs_dir2_sf_get_offset(xfs_dir2_sf_entry_t *sfep)
+{
+	return get_unaligned_be16(&sfep->offset.i);
+}
+
+static inline void
+xfs_dir2_sf_put_offset(xfs_dir2_sf_entry_t *sfep, xfs_dir2_data_aoff_t off)
+{
+	put_unaligned_be16(off, &sfep->offset.i);
+}
+
+static inline int
+xfs_dir2_sf_entsize(struct xfs_dir2_sf_hdr *hdr, int len)
+{
+	return sizeof(struct xfs_dir2_sf_entry) +	/* namelen + offset */
+		len +					/* name */
+		(hdr->i8count ?				/* ino */
+		 sizeof(xfs_dir2_ino8_t) :
+		 sizeof(xfs_dir2_ino4_t));
+}
+
+static inline struct xfs_dir2_sf_entry *
+xfs_dir2_sf_firstentry(struct xfs_dir2_sf_hdr *hdr)
+{
+	return (struct xfs_dir2_sf_entry *)
+		((char *)hdr + xfs_dir2_sf_hdr_size(hdr->i8count));
+}
+
+static inline struct xfs_dir2_sf_entry *
+xfs_dir2_sf_nextentry(struct xfs_dir2_sf_hdr *hdr,
+		struct xfs_dir2_sf_entry *sfep)
+{
+	return (struct xfs_dir2_sf_entry *)
+		((char *)sfep + xfs_dir2_sf_entsize(hdr, sfep->namelen));
+}
+
+
+/*
+ * Data block structures.
+ *
+ * A pure data block looks like the following drawing on disk:
+ *
+ *    +-------------------------------------------------+
+ *    | xfs_dir2_data_hdr_t                             |
+ *    +-------------------------------------------------+
+ *    | xfs_dir2_data_entry_t OR xfs_dir2_data_unused_t |
+ *    | xfs_dir2_data_entry_t OR xfs_dir2_data_unused_t |
+ *    | xfs_dir2_data_entry_t OR xfs_dir2_data_unused_t |
+ *    | ...                                             |
+ *    +-------------------------------------------------+
+ *    | unused space                                    |
+ *    +-------------------------------------------------+
+ *
+ * As all the entries are variable size structures the accessors below should
+ * be used to iterate over them.
+ *
+ * In addition to the pure data blocks for the data and node formats,
+ * most structures are also used for the combined data/freespace "block"
+ * format below.
+ */
+
+#define	XFS_DIR2_DATA_ALIGN_LOG	3		/* i.e., 8 bytes */
+#define	XFS_DIR2_DATA_ALIGN	(1 << XFS_DIR2_DATA_ALIGN_LOG)
+#define	XFS_DIR2_DATA_FREE_TAG	0xffff
+#define	XFS_DIR2_DATA_FD_COUNT	3
+
+/*
+ * Directory address space divided into sections,
+ * spaces separated by 32GB.
+ */
+#define	XFS_DIR2_SPACE_SIZE	(1ULL << (32 + XFS_DIR2_DATA_ALIGN_LOG))
+#define	XFS_DIR2_DATA_SPACE	0
+#define	XFS_DIR2_DATA_OFFSET	(XFS_DIR2_DATA_SPACE * XFS_DIR2_SPACE_SIZE)
+#define	XFS_DIR2_DATA_FIRSTDB(mp)	\
+	xfs_dir2_byte_to_db(mp, XFS_DIR2_DATA_OFFSET)
+
+/*
+ * Offsets of . and .. in data space (always block 0)
+ */
+#define	XFS_DIR2_DATA_DOT_OFFSET	\
+	((xfs_dir2_data_aoff_t)sizeof(struct xfs_dir2_data_hdr))
+#define	XFS_DIR2_DATA_DOTDOT_OFFSET	\
+	(XFS_DIR2_DATA_DOT_OFFSET + xfs_dir2_data_entsize(1))
+#define	XFS_DIR2_DATA_FIRST_OFFSET		\
+	(XFS_DIR2_DATA_DOTDOT_OFFSET + xfs_dir2_data_entsize(2))
+
+/*
+ * Describe a free area in the data block.
+ *
+ * The freespace will be formatted as a xfs_dir2_data_unused_t.
+ */
+typedef struct xfs_dir2_data_free {
+	__be16			offset;		/* start of freespace */
+	__be16			length;		/* length of freespace */
+} xfs_dir2_data_free_t;
+
+/*
+ * Header for the data blocks.
+ *
+ * The code knows that XFS_DIR2_DATA_FD_COUNT is 3.
+ */
+typedef struct xfs_dir2_data_hdr {
+	__be32			magic;		/* XFS_DIR2_DATA_MAGIC or */
+						/* XFS_DIR2_BLOCK_MAGIC */
+	xfs_dir2_data_free_t	bestfree[XFS_DIR2_DATA_FD_COUNT];
+} xfs_dir2_data_hdr_t;
+
+/*
+ * Active entry in a data block.
+ *
+ * Aligned to 8 bytes.  After the variable length name field there is a
+ * 2 byte tag field, which can be accessed using xfs_dir2_data_entry_tag_p.
+ */
+typedef struct xfs_dir2_data_entry {
+	__be64			inumber;	/* inode number */
+	__u8			namelen;	/* name length */
+	__u8			name[];		/* name bytes, no null */
+     /*	__be16                  tag; */		/* starting offset of us */
+} xfs_dir2_data_entry_t;
+
+/*
+ * Unused entry in a data block.
+ *
+ * Aligned to 8 bytes.  Tag appears as the last 2 bytes and must be accessed
+ * using xfs_dir2_data_unused_tag_p.
+ */
+typedef struct xfs_dir2_data_unused {
+	__be16			freetag;	/* XFS_DIR2_DATA_FREE_TAG */
+	__be16			length;		/* total free length */
+						/* variable offset */
+	__be16			tag;		/* starting offset of us */
+} xfs_dir2_data_unused_t;
+
+/*
+ * Size of a data entry.
+ */
+static inline int xfs_dir2_data_entsize(int n)
+{
+	return (int)roundup(offsetof(struct xfs_dir2_data_entry, name[0]) + n +
+		 (uint)sizeof(xfs_dir2_data_off_t), XFS_DIR2_DATA_ALIGN);
+}
+
+/*
+ * Pointer to an entry's tag word.
+ */
+static inline __be16 *
+xfs_dir2_data_entry_tag_p(struct xfs_dir2_data_entry *dep)
+{
+	return (__be16 *)((char *)dep +
+		xfs_dir2_data_entsize(dep->namelen) - sizeof(__be16));
+}
+
+/*
+ * Pointer to a freespace's tag word.
+ */
+static inline __be16 *
+xfs_dir2_data_unused_tag_p(struct xfs_dir2_data_unused *dup)
+{
+	return (__be16 *)((char *)dup +
+			be16_to_cpu(dup->length) - sizeof(__be16));
+}
+
+/*
+ * Leaf block structures.
+ *
+ * A pure leaf block looks like the following drawing on disk:
+ *
+ *    +---------------------------+
+ *    | xfs_dir2_leaf_hdr_t       |
+ *    +---------------------------+
+ *    | xfs_dir2_leaf_entry_t     |
+ *    | xfs_dir2_leaf_entry_t     |
+ *    | xfs_dir2_leaf_entry_t     |
+ *    | xfs_dir2_leaf_entry_t     |
+ *    | ...                       |
+ *    +---------------------------+
+ *    | xfs_dir2_data_off_t       |
+ *    | xfs_dir2_data_off_t       |
+ *    | xfs_dir2_data_off_t       |
+ *    | ...                       |
+ *    +---------------------------+
+ *    | xfs_dir2_leaf_tail_t      |
+ *    +---------------------------+
+ *
+ * The xfs_dir2_data_off_t members (bests) and tail are at the end of the block
+ * for single-leaf (magic = XFS_DIR2_LEAF1_MAGIC) blocks only, but not present
+ * for directories with separate leaf nodes and free space blocks
+ * (magic = XFS_DIR2_LEAFN_MAGIC).
+ *
+ * As all the entries are variable size structures the accessors below should
+ * be used to iterate over them.
+ */
+
+/*
+ * Offset of the leaf/node space.  First block in this space
+ * is the btree root.
+ */
+#define	XFS_DIR2_LEAF_SPACE	1
+#define	XFS_DIR2_LEAF_OFFSET	(XFS_DIR2_LEAF_SPACE * XFS_DIR2_SPACE_SIZE)
+#define	XFS_DIR2_LEAF_FIRSTDB(mp)	\
+	xfs_dir2_byte_to_db(mp, XFS_DIR2_LEAF_OFFSET)
+
+/*
+ * Leaf block header.
+ */
+typedef struct xfs_dir2_leaf_hdr {
+	xfs_da_blkinfo_t	info;		/* header for da routines */
+	__be16			count;		/* count of entries */
+	__be16			stale;		/* count of stale entries */
+} xfs_dir2_leaf_hdr_t;
+
+/*
+ * Leaf block entry.
+ */
+typedef struct xfs_dir2_leaf_entry {
+	__be32			hashval;	/* hash value of name */
+	__be32			address;	/* address of data entry */
+} xfs_dir2_leaf_entry_t;
+
+/*
+ * Leaf block tail.
+ */
+typedef struct xfs_dir2_leaf_tail {
+	__be32			bestcount;
+} xfs_dir2_leaf_tail_t;
+
+/*
+ * Leaf block.
+ */
+typedef struct xfs_dir2_leaf {
+	xfs_dir2_leaf_hdr_t	hdr;		/* leaf header */
+	xfs_dir2_leaf_entry_t	ents[];		/* entries */
+} xfs_dir2_leaf_t;
+
+/*
+ * DB blocks here are logical directory block numbers, not filesystem blocks.
+ */
+
+static inline int xfs_dir2_max_leaf_ents(struct xfs_mount *mp)
+{
+	return (mp->m_dirblksize - (uint)sizeof(struct xfs_dir2_leaf_hdr)) /
+		(uint)sizeof(struct xfs_dir2_leaf_entry);
+}
+
+/*
+ * Get address of the bestcount field in the single-leaf block.
+ */
+static inline struct xfs_dir2_leaf_tail *
+xfs_dir2_leaf_tail_p(struct xfs_mount *mp, struct xfs_dir2_leaf *lp)
+{
+	return (struct xfs_dir2_leaf_tail *)
+		((char *)lp + mp->m_dirblksize -
+		  sizeof(struct xfs_dir2_leaf_tail));
+}
+
+/*
+ * Get address of the bests array in the single-leaf block.
+ */
+static inline __be16 *
+xfs_dir2_leaf_bests_p(struct xfs_dir2_leaf_tail *ltp)
+{
+	return (__be16 *)ltp - be32_to_cpu(ltp->bestcount);
+}
+
+/*
+ * Convert dataptr to byte in file space
+ */
+static inline xfs_dir2_off_t
+xfs_dir2_dataptr_to_byte(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
+{
+	return (xfs_dir2_off_t)dp << XFS_DIR2_DATA_ALIGN_LOG;
+}
+
+/*
+ * Convert byte in file space to dataptr.  It had better be aligned.
+ */
+static inline xfs_dir2_dataptr_t
+xfs_dir2_byte_to_dataptr(struct xfs_mount *mp, xfs_dir2_off_t by)
+{
+	return (xfs_dir2_dataptr_t)(by >> XFS_DIR2_DATA_ALIGN_LOG);
+}
+
+/*
+ * Convert byte in space to (DB) block
+ */
+static inline xfs_dir2_db_t
+xfs_dir2_byte_to_db(struct xfs_mount *mp, xfs_dir2_off_t by)
+{
+	return (xfs_dir2_db_t)
+		(by >> (mp->m_sb.sb_blocklog + mp->m_sb.sb_dirblklog));
+}
+
+/*
+ * Convert dataptr to a block number
+ */
+static inline xfs_dir2_db_t
+xfs_dir2_dataptr_to_db(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
+{
+	return xfs_dir2_byte_to_db(mp, xfs_dir2_dataptr_to_byte(mp, dp));
+}
+
+/*
+ * Convert byte in space to offset in a block
+ */
+static inline xfs_dir2_data_aoff_t
+xfs_dir2_byte_to_off(struct xfs_mount *mp, xfs_dir2_off_t by)
+{
+	return (xfs_dir2_data_aoff_t)(by &
+		((1 << (mp->m_sb.sb_blocklog + mp->m_sb.sb_dirblklog)) - 1));
+}
+
+/*
+ * Convert dataptr to a byte offset in a block
+ */
+static inline xfs_dir2_data_aoff_t
+xfs_dir2_dataptr_to_off(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
+{
+	return xfs_dir2_byte_to_off(mp, xfs_dir2_dataptr_to_byte(mp, dp));
+}
+
+/*
+ * Convert block and offset to byte in space
+ */
+static inline xfs_dir2_off_t
+xfs_dir2_db_off_to_byte(struct xfs_mount *mp, xfs_dir2_db_t db,
+			xfs_dir2_data_aoff_t o)
+{
+	return ((xfs_dir2_off_t)db <<
+		(mp->m_sb.sb_blocklog + mp->m_sb.sb_dirblklog)) + o;
+}
+
+/*
+ * Convert block (DB) to block (dablk)
+ */
+static inline xfs_dablk_t
+xfs_dir2_db_to_da(struct xfs_mount *mp, xfs_dir2_db_t db)
+{
+	return (xfs_dablk_t)(db << mp->m_sb.sb_dirblklog);
+}
+
+/*
+ * Convert byte in space to (DA) block
+ */
+static inline xfs_dablk_t
+xfs_dir2_byte_to_da(struct xfs_mount *mp, xfs_dir2_off_t by)
+{
+	return xfs_dir2_db_to_da(mp, xfs_dir2_byte_to_db(mp, by));
+}
+
+/*
+ * Convert block and offset to dataptr
+ */
+static inline xfs_dir2_dataptr_t
+xfs_dir2_db_off_to_dataptr(struct xfs_mount *mp, xfs_dir2_db_t db,
+			   xfs_dir2_data_aoff_t o)
+{
+	return xfs_dir2_byte_to_dataptr(mp, xfs_dir2_db_off_to_byte(mp, db, o));
+}
+
+/*
+ * Convert block (dablk) to block (DB)
+ */
+static inline xfs_dir2_db_t
+xfs_dir2_da_to_db(struct xfs_mount *mp, xfs_dablk_t da)
+{
+	return (xfs_dir2_db_t)(da >> mp->m_sb.sb_dirblklog);
+}
+
+/*
+ * Convert block (dablk) to byte offset in space
+ */
+static inline xfs_dir2_off_t
+xfs_dir2_da_to_byte(struct xfs_mount *mp, xfs_dablk_t da)
+{
+	return xfs_dir2_db_off_to_byte(mp, xfs_dir2_da_to_db(mp, da), 0);
+}
+
+/*
+ * Free space block defintions for the node format.
+ */
+
+/*
+ * Offset of the freespace index.
+ */
+#define	XFS_DIR2_FREE_SPACE	2
+#define	XFS_DIR2_FREE_OFFSET	(XFS_DIR2_FREE_SPACE * XFS_DIR2_SPACE_SIZE)
+#define	XFS_DIR2_FREE_FIRSTDB(mp)	\
+	xfs_dir2_byte_to_db(mp, XFS_DIR2_FREE_OFFSET)
+
+typedef	struct xfs_dir2_free_hdr {
+	__be32			magic;		/* XFS_DIR2_FREE_MAGIC */
+	__be32			firstdb;	/* db of first entry */
+	__be32			nvalid;		/* count of valid entries */
+	__be32			nused;		/* count of used entries */
+} xfs_dir2_free_hdr_t;
+
+typedef struct xfs_dir2_free {
+	xfs_dir2_free_hdr_t	hdr;		/* block header */
+	__be16			bests[];	/* best free counts */
+						/* unused entries are -1 */
+} xfs_dir2_free_t;
+
+static inline int xfs_dir2_free_max_bests(struct xfs_mount *mp)
+{
+	return (mp->m_dirblksize - sizeof(struct xfs_dir2_free_hdr)) /
+		sizeof(xfs_dir2_data_off_t);
+}
+
+/*
+ * Convert data space db to the corresponding free db.
+ */
+static inline xfs_dir2_db_t
+xfs_dir2_db_to_fdb(struct xfs_mount *mp, xfs_dir2_db_t db)
+{
+	return XFS_DIR2_FREE_FIRSTDB(mp) + db / xfs_dir2_free_max_bests(mp);
+}
+
+/*
+ * Convert data space db to the corresponding index in a free db.
+ */
+static inline int
+xfs_dir2_db_to_fdindex(struct xfs_mount *mp, xfs_dir2_db_t db)
+{
+	return db % xfs_dir2_free_max_bests(mp);
+}
+
+/*
+ * Single block format.
+ *
+ * The single block format looks like the following drawing on disk:
+ *
+ *    +-------------------------------------------------+
+ *    | xfs_dir2_data_hdr_t                             |
+ *    +-------------------------------------------------+
+ *    | xfs_dir2_data_entry_t OR xfs_dir2_data_unused_t |
+ *    | xfs_dir2_data_entry_t OR xfs_dir2_data_unused_t |
+ *    | xfs_dir2_data_entry_t OR xfs_dir2_data_unused_t :
+ *    | ...                                             |
+ *    +-------------------------------------------------+
+ *    | unused space                                    |
+ *    +-------------------------------------------------+
+ *    | ...                                             |
+ *    | xfs_dir2_leaf_entry_t                           |
+ *    | xfs_dir2_leaf_entry_t                           |
+ *    +-------------------------------------------------+
+ *    | xfs_dir2_block_tail_t                           |
+ *    +-------------------------------------------------+
+ *
+ * As all the entries are variable size structures the accessors below should
+ * be used to iterate over them.
+ */
+
+typedef struct xfs_dir2_block_tail {
+	__be32		count;			/* count of leaf entries */
+	__be32		stale;			/* count of stale lf entries */
+} xfs_dir2_block_tail_t;
+
+/*
+ * Pointer to the leaf header embedded in a data block (1-block format)
+ */
+static inline struct xfs_dir2_block_tail *
+xfs_dir2_block_tail_p(struct xfs_mount *mp, struct xfs_dir2_data_hdr *hdr)
+{
+	return ((struct xfs_dir2_block_tail *)
+		((char *)hdr + mp->m_dirblksize)) - 1;
+}
+
+/*
+ * Pointer to the leaf entries embedded in a data block (1-block format)
+ */
+static inline struct xfs_dir2_leaf_entry *
+xfs_dir2_block_leaf_p(struct xfs_dir2_block_tail *btp)
+{
+	return ((struct xfs_dir2_leaf_entry *)btp) - be32_to_cpu(btp->count);
+}
+
+#endif /* __XFS_DIR2_FORMAT_H__ */
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index ae891223be9..ca2386d82cd 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -24,18 +24,14 @@
 #include "xfs_trans.h"
 #include "xfs_sb.h"
 #include "xfs_ag.h"
-#include "xfs_dir2.h"
 #include "xfs_mount.h"
 #include "xfs_da_btree.h"
 #include "xfs_bmap_btree.h"
-#include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_bmap.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
-#include "xfs_dir2_node.h"
+#include "xfs_dir2_format.h"
+#include "xfs_dir2_priv.h"
 #include "xfs_error.h"
 #include "xfs_trace.h"
 
@@ -64,7 +60,7 @@ xfs_dir2_block_to_leaf(
 {
 	__be16			*bestsp;	/* leaf's bestsp entries */
 	xfs_dablk_t		blkno;		/* leaf block's bno */
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_leaf_entry_t	*blp;		/* block's leaf entries */
 	xfs_dir2_block_tail_t	*btp;		/* block's tail */
 	xfs_inode_t		*dp;		/* incore directory inode */
@@ -101,9 +97,9 @@ xfs_dir2_block_to_leaf(
 	}
 	ASSERT(lbp != NULL);
 	leaf = lbp->data;
-	block = dbp->data;
+	hdr = dbp->data;
 	xfs_dir2_data_check(dp, dbp);
-	btp = xfs_dir2_block_tail_p(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
 	/*
 	 * Set the counts in the leaf header.
@@ -123,23 +119,23 @@ xfs_dir2_block_to_leaf(
 	 * tail be free.
 	 */
 	xfs_dir2_data_make_free(tp, dbp,
-		(xfs_dir2_data_aoff_t)((char *)blp - (char *)block),
-		(xfs_dir2_data_aoff_t)((char *)block + mp->m_dirblksize -
+		(xfs_dir2_data_aoff_t)((char *)blp - (char *)hdr),
+		(xfs_dir2_data_aoff_t)((char *)hdr + mp->m_dirblksize -
 				       (char *)blp),
 		&needlog, &needscan);
 	/*
 	 * Fix up the block header, make it a data block.
 	 */
-	block->hdr.magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
+	hdr->magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
 	if (needscan)
-		xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	/*
 	 * Set up leaf tail and bests table.
 	 */
 	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	ltp->bestcount = cpu_to_be32(1);
 	bestsp = xfs_dir2_leaf_bests_p(ltp);
-	bestsp[0] =  block->hdr.bestfree[0].length;
+	bestsp[0] =  hdr->bestfree[0].length;
 	/*
 	 * Log the data header and leaf bests table.
 	 */
@@ -152,6 +148,131 @@ xfs_dir2_block_to_leaf(
 	return 0;
 }
 
+STATIC void
+xfs_dir2_leaf_find_stale(
+	struct xfs_dir2_leaf	*leaf,
+	int			index,
+	int			*lowstale,
+	int			*highstale)
+{
+	/*
+	 * Find the first stale entry before our index, if any.
+	 */
+	for (*lowstale = index - 1; *lowstale >= 0; --*lowstale) {
+		if (leaf->ents[*lowstale].address ==
+		    cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
+			break;
+	}
+
+	/*
+	 * Find the first stale entry at or after our index, if any.
+	 * Stop if the result would require moving more entries than using
+	 * lowstale.
+	 */
+	for (*highstale = index;
+	     *highstale < be16_to_cpu(leaf->hdr.count);
+	     ++*highstale) {
+		if (leaf->ents[*highstale].address ==
+		    cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
+			break;
+		if (*lowstale >= 0 && index - *lowstale <= *highstale - index)
+			break;
+	}
+}
+
+struct xfs_dir2_leaf_entry *
+xfs_dir2_leaf_find_entry(
+	xfs_dir2_leaf_t		*leaf,		/* leaf structure */
+	int			index,		/* leaf table position */
+	int			compact,	/* need to compact leaves */
+	int			lowstale,	/* index of prev stale leaf */
+	int			highstale,	/* index of next stale leaf */
+	int			*lfloglow,	/* low leaf logging index */
+	int			*lfloghigh)	/* high leaf logging index */
+{
+	if (!leaf->hdr.stale) {
+		xfs_dir2_leaf_entry_t	*lep;	/* leaf entry table pointer */
+
+		/*
+		 * Now we need to make room to insert the leaf entry.
+		 *
+		 * If there are no stale entries, just insert a hole at index.
+		 */
+		lep = &leaf->ents[index];
+		if (index < be16_to_cpu(leaf->hdr.count))
+			memmove(lep + 1, lep,
+				(be16_to_cpu(leaf->hdr.count) - index) *
+				 sizeof(*lep));
+
+		/*
+		 * Record low and high logging indices for the leaf.
+		 */
+		*lfloglow = index;
+		*lfloghigh = be16_to_cpu(leaf->hdr.count);
+		be16_add_cpu(&leaf->hdr.count, 1);
+		return lep;
+	}
+
+	/*
+	 * There are stale entries.
+	 *
+	 * We will use one of them for the new entry.  It's probably not at
+	 * the right location, so we'll have to shift some up or down first.
+	 *
+	 * If we didn't compact before, we need to find the nearest stale
+	 * entries before and after our insertion point.
+	 */
+	if (compact == 0)
+		xfs_dir2_leaf_find_stale(leaf, index, &lowstale, &highstale);
+
+	/*
+	 * If the low one is better, use it.
+	 */
+	if (lowstale >= 0 &&
+	    (highstale == be16_to_cpu(leaf->hdr.count) ||
+	     index - lowstale - 1 < highstale - index)) {
+		ASSERT(index - lowstale - 1 >= 0);
+		ASSERT(leaf->ents[lowstale].address ==
+		       cpu_to_be32(XFS_DIR2_NULL_DATAPTR));
+
+		/*
+		 * Copy entries up to cover the stale entry and make room
+		 * for the new entry.
+		 */
+		if (index - lowstale - 1 > 0) {
+			memmove(&leaf->ents[lowstale],
+				&leaf->ents[lowstale + 1],
+				(index - lowstale - 1) *
+				sizeof(xfs_dir2_leaf_entry_t));
+		}
+		*lfloglow = MIN(lowstale, *lfloglow);
+		*lfloghigh = MAX(index - 1, *lfloghigh);
+		be16_add_cpu(&leaf->hdr.stale, -1);
+		return &leaf->ents[index - 1];
+	}
+
+	/*
+	 * The high one is better, so use that one.
+	 */
+	ASSERT(highstale - index >= 0);
+	ASSERT(leaf->ents[highstale].address ==
+	       cpu_to_be32(XFS_DIR2_NULL_DATAPTR));
+
+	/*
+	 * Copy entries down to cover the stale entry and make room for the
+	 * new entry.
+	 */
+	if (highstale - index > 0) {
+		memmove(&leaf->ents[index + 1],
+			&leaf->ents[index],
+			(highstale - index) * sizeof(xfs_dir2_leaf_entry_t));
+	}
+	*lfloglow = MIN(index, *lfloglow);
+	*lfloghigh = MAX(highstale, *lfloghigh);
+	be16_add_cpu(&leaf->hdr.stale, -1);
+	return &leaf->ents[index];
+}
+
 /*
  * Add an entry to a leaf form directory.
  */
@@ -161,7 +282,7 @@ xfs_dir2_leaf_addname(
 {
 	__be16			*bestsp;	/* freespace table in leaf */
 	int			compact;	/* need to compact leaves */
-	xfs_dir2_data_t		*data;		/* data block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dabuf_t		*dbp;		/* data block buffer */
 	xfs_dir2_data_entry_t	*dep;		/* data block entry */
 	xfs_inode_t		*dp;		/* incore directory inode */
@@ -225,7 +346,7 @@ xfs_dir2_leaf_addname(
 			continue;
 		i = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
 		ASSERT(i < be32_to_cpu(ltp->bestcount));
-		ASSERT(be16_to_cpu(bestsp[i]) != NULLDATAOFF);
+		ASSERT(bestsp[i] != cpu_to_be16(NULLDATAOFF));
 		if (be16_to_cpu(bestsp[i]) >= length) {
 			use_block = i;
 			break;
@@ -239,7 +360,8 @@ xfs_dir2_leaf_addname(
 			/*
 			 * Remember a block we see that's missing.
 			 */
-			if (be16_to_cpu(bestsp[i]) == NULLDATAOFF && use_block == -1)
+			if (bestsp[i] == cpu_to_be16(NULLDATAOFF) &&
+			    use_block == -1)
 				use_block = i;
 			else if (be16_to_cpu(bestsp[i]) >= length) {
 				use_block = i;
@@ -250,14 +372,17 @@ xfs_dir2_leaf_addname(
 	/*
 	 * How many bytes do we need in the leaf block?
 	 */
-	needbytes =
-		(leaf->hdr.stale ? 0 : (uint)sizeof(leaf->ents[0])) +
-		(use_block != -1 ? 0 : (uint)sizeof(leaf->bests[0]));
+	needbytes = 0;
+	if (!leaf->hdr.stale)
+		needbytes += sizeof(xfs_dir2_leaf_entry_t);
+	if (use_block == -1)
+		needbytes += sizeof(xfs_dir2_data_off_t);
+
 	/*
 	 * Now kill use_block if it refers to a missing block, so we
 	 * can use it as an indication of allocation needed.
 	 */
-	if (use_block != -1 && be16_to_cpu(bestsp[use_block]) == NULLDATAOFF)
+	if (use_block != -1 && bestsp[use_block] == cpu_to_be16(NULLDATAOFF))
 		use_block = -1;
 	/*
 	 * If we don't have enough free bytes but we can make enough
@@ -369,8 +494,8 @@ xfs_dir2_leaf_addname(
 		 */
 		else
 			xfs_dir2_leaf_log_bests(tp, lbp, use_block, use_block);
-		data = dbp->data;
-		bestsp[use_block] = data->hdr.bestfree[0].length;
+		hdr = dbp->data;
+		bestsp[use_block] = hdr->bestfree[0].length;
 		grown = 1;
 	}
 	/*
@@ -384,7 +509,7 @@ xfs_dir2_leaf_addname(
 			xfs_da_brelse(tp, lbp);
 			return error;
 		}
-		data = dbp->data;
+		hdr = dbp->data;
 		grown = 0;
 	}
 	xfs_dir2_data_check(dp, dbp);
@@ -392,14 +517,14 @@ xfs_dir2_leaf_addname(
 	 * Point to the biggest freespace in our data block.
 	 */
 	dup = (xfs_dir2_data_unused_t *)
-	      ((char *)data + be16_to_cpu(data->hdr.bestfree[0].offset));
+	      ((char *)hdr + be16_to_cpu(hdr->bestfree[0].offset));
 	ASSERT(be16_to_cpu(dup->length) >= length);
 	needscan = needlog = 0;
 	/*
 	 * Mark the initial part of our freespace in use for the new entry.
 	 */
 	xfs_dir2_data_use_free(tp, dbp, dup,
-		(xfs_dir2_data_aoff_t)((char *)dup - (char *)data), length,
+		(xfs_dir2_data_aoff_t)((char *)dup - (char *)hdr), length,
 		&needlog, &needscan);
 	/*
 	 * Initialize our new entry (at last).
@@ -409,12 +534,12 @@ xfs_dir2_leaf_addname(
 	dep->namelen = args->namelen;
 	memcpy(dep->name, args->name, dep->namelen);
 	tagp = xfs_dir2_data_entry_tag_p(dep);
-	*tagp = cpu_to_be16((char *)dep - (char *)data);
+	*tagp = cpu_to_be16((char *)dep - (char *)hdr);
 	/*
 	 * Need to scan fix up the bestfree table.
 	 */
 	if (needscan)
-		xfs_dir2_data_freescan(mp, data, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	/*
 	 * Need to log the data block's header.
 	 */
@@ -425,107 +550,15 @@ xfs_dir2_leaf_addname(
 	 * If the bests table needs to be changed, do it.
 	 * Log the change unless we've already done that.
 	 */
-	if (be16_to_cpu(bestsp[use_block]) != be16_to_cpu(data->hdr.bestfree[0].length)) {
-		bestsp[use_block] = data->hdr.bestfree[0].length;
+	if (be16_to_cpu(bestsp[use_block]) != be16_to_cpu(hdr->bestfree[0].length)) {
+		bestsp[use_block] = hdr->bestfree[0].length;
 		if (!grown)
 			xfs_dir2_leaf_log_bests(tp, lbp, use_block, use_block);
 	}
-	/*
-	 * Now we need to make room to insert the leaf entry.
-	 * If there are no stale entries, we just insert a hole at index.
-	 */
-	if (!leaf->hdr.stale) {
-		/*
-		 * lep is still good as the index leaf entry.
-		 */
-		if (index < be16_to_cpu(leaf->hdr.count))
-			memmove(lep + 1, lep,
-				(be16_to_cpu(leaf->hdr.count) - index) * sizeof(*lep));
-		/*
-		 * Record low and high logging indices for the leaf.
-		 */
-		lfloglow = index;
-		lfloghigh = be16_to_cpu(leaf->hdr.count);
-		be16_add_cpu(&leaf->hdr.count, 1);
-	}
-	/*
-	 * There are stale entries.
-	 * We will use one of them for the new entry.
-	 * It's probably not at the right location, so we'll have to
-	 * shift some up or down first.
-	 */
-	else {
-		/*
-		 * If we didn't compact before, we need to find the nearest
-		 * stale entries before and after our insertion point.
-		 */
-		if (compact == 0) {
-			/*
-			 * Find the first stale entry before the insertion
-			 * point, if any.
-			 */
-			for (lowstale = index - 1;
-			     lowstale >= 0 &&
-				be32_to_cpu(leaf->ents[lowstale].address) !=
-				XFS_DIR2_NULL_DATAPTR;
-			     lowstale--)
-				continue;
-			/*
-			 * Find the next stale entry at or after the insertion
-			 * point, if any.   Stop if we go so far that the
-			 * lowstale entry would be better.
-			 */
-			for (highstale = index;
-			     highstale < be16_to_cpu(leaf->hdr.count) &&
-				be32_to_cpu(leaf->ents[highstale].address) !=
-				XFS_DIR2_NULL_DATAPTR &&
-				(lowstale < 0 ||
-				 index - lowstale - 1 >= highstale - index);
-			     highstale++)
-				continue;
-		}
-		/*
-		 * If the low one is better, use it.
-		 */
-		if (lowstale >= 0 &&
-		    (highstale == be16_to_cpu(leaf->hdr.count) ||
-		     index - lowstale - 1 < highstale - index)) {
-			ASSERT(index - lowstale - 1 >= 0);
-			ASSERT(be32_to_cpu(leaf->ents[lowstale].address) ==
-			       XFS_DIR2_NULL_DATAPTR);
-			/*
-			 * Copy entries up to cover the stale entry
-			 * and make room for the new entry.
-			 */
-			if (index - lowstale - 1 > 0)
-				memmove(&leaf->ents[lowstale],
-					&leaf->ents[lowstale + 1],
-					(index - lowstale - 1) * sizeof(*lep));
-			lep = &leaf->ents[index - 1];
-			lfloglow = MIN(lowstale, lfloglow);
-			lfloghigh = MAX(index - 1, lfloghigh);
-		}
-		/*
-		 * The high one is better, so use that one.
-		 */
-		else {
-			ASSERT(highstale - index >= 0);
-			ASSERT(be32_to_cpu(leaf->ents[highstale].address) ==
-			       XFS_DIR2_NULL_DATAPTR);
-			/*
-			 * Copy entries down to cover the stale entry
-			 * and make room for the new entry.
-			 */
-			if (highstale - index > 0)
-				memmove(&leaf->ents[index + 1],
-					&leaf->ents[index],
-					(highstale - index) * sizeof(*lep));
-			lep = &leaf->ents[index];
-			lfloglow = MIN(index, lfloglow);
-			lfloghigh = MAX(highstale, lfloghigh);
-		}
-		be16_add_cpu(&leaf->hdr.stale, -1);
-	}
+
+	lep = xfs_dir2_leaf_find_entry(leaf, index, compact, lowstale,
+				       highstale, &lfloglow, &lfloghigh);
+
 	/*
 	 * Fill in the new leaf entry.
 	 */
@@ -562,7 +595,7 @@ xfs_dir2_leaf_check(
 
 	leaf = bp->data;
 	mp = dp->i_mount;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
 	/*
 	 * This value is not restrictive enough.
 	 * Should factor in the size of the bests table as well.
@@ -582,7 +615,7 @@ xfs_dir2_leaf_check(
 		if (i + 1 < be16_to_cpu(leaf->hdr.count))
 			ASSERT(be32_to_cpu(leaf->ents[i].hashval) <=
 			       be32_to_cpu(leaf->ents[i + 1].hashval));
-		if (be32_to_cpu(leaf->ents[i].address) == XFS_DIR2_NULL_DATAPTR)
+		if (leaf->ents[i].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
 			stale++;
 	}
 	ASSERT(be16_to_cpu(leaf->hdr.stale) == stale);
@@ -611,7 +644,8 @@ xfs_dir2_leaf_compact(
 	 * Compress out the stale entries in place.
 	 */
 	for (from = to = 0, loglow = -1; from < be16_to_cpu(leaf->hdr.count); from++) {
-		if (be32_to_cpu(leaf->ents[from].address) == XFS_DIR2_NULL_DATAPTR)
+		if (leaf->ents[from].address ==
+		    cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
 			continue;
 		/*
 		 * Only actually copy the entries that are different.
@@ -663,24 +697,9 @@ xfs_dir2_leaf_compact_x1(
 	leaf = bp->data;
 	ASSERT(be16_to_cpu(leaf->hdr.stale) > 1);
 	index = *indexp;
-	/*
-	 * Find the first stale entry before our index, if any.
-	 */
-	for (lowstale = index - 1;
-	     lowstale >= 0 &&
-		be32_to_cpu(leaf->ents[lowstale].address) != XFS_DIR2_NULL_DATAPTR;
-	     lowstale--)
-		continue;
-	/*
-	 * Find the first stale entry at or after our index, if any.
-	 * Stop if the answer would be worse than lowstale.
-	 */
-	for (highstale = index;
-	     highstale < be16_to_cpu(leaf->hdr.count) &&
-		be32_to_cpu(leaf->ents[highstale].address) != XFS_DIR2_NULL_DATAPTR &&
-		(lowstale < 0 || index - lowstale > highstale - index);
-	     highstale++)
-		continue;
+
+	xfs_dir2_leaf_find_stale(leaf, index, &lowstale, &highstale);
+
 	/*
 	 * Pick the better of lowstale and highstale.
 	 */
@@ -701,7 +720,8 @@ xfs_dir2_leaf_compact_x1(
 		if (index == from)
 			newindex = to;
 		if (from != keepstale &&
-		    be32_to_cpu(leaf->ents[from].address) == XFS_DIR2_NULL_DATAPTR) {
+		    leaf->ents[from].address ==
+		    cpu_to_be32(XFS_DIR2_NULL_DATAPTR)) {
 			if (from == to)
 				*lowlogp = to;
 			continue;
@@ -760,7 +780,7 @@ xfs_dir2_leaf_getdents(
 	int			byteoff;	/* offset in current block */
 	xfs_dir2_db_t		curdb;		/* db for current block */
 	xfs_dir2_off_t		curoff;		/* current overall offset */
-	xfs_dir2_data_t		*data;		/* data block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dir2_data_entry_t	*dep;		/* data entry */
 	xfs_dir2_data_unused_t	*dup;		/* unused entry */
 	int			error = 0;	/* error return value */
@@ -1018,23 +1038,23 @@ xfs_dir2_leaf_getdents(
 			else if (curoff > newoff)
 				ASSERT(xfs_dir2_byte_to_db(mp, curoff) ==
 				       curdb);
-			data = bp->data;
+			hdr = bp->data;
 			xfs_dir2_data_check(dp, bp);
 			/*
 			 * Find our position in the block.
 			 */
-			ptr = (char *)&data->u;
+			ptr = (char *)(hdr + 1);
 			byteoff = xfs_dir2_byte_to_off(mp, curoff);
 			/*
 			 * Skip past the header.
 			 */
 			if (byteoff == 0)
-				curoff += (uint)sizeof(data->hdr);
+				curoff += (uint)sizeof(*hdr);
 			/*
 			 * Skip past entries until we reach our offset.
 			 */
 			else {
-				while ((char *)ptr - (char *)data < byteoff) {
+				while ((char *)ptr - (char *)hdr < byteoff) {
 					dup = (xfs_dir2_data_unused_t *)ptr;
 
 					if (be16_to_cpu(dup->freetag)
@@ -1055,8 +1075,8 @@ xfs_dir2_leaf_getdents(
 				curoff =
 					xfs_dir2_db_off_to_byte(mp,
 					    xfs_dir2_byte_to_db(mp, curoff),
-					    (char *)ptr - (char *)data);
-				if (ptr >= (char *)data + mp->m_dirblksize) {
+					    (char *)ptr - (char *)hdr);
+				if (ptr >= (char *)hdr + mp->m_dirblksize) {
 					continue;
 				}
 			}
@@ -1179,7 +1199,7 @@ xfs_dir2_leaf_log_bests(
 	xfs_dir2_leaf_tail_t	*ltp;		/* leaf tail structure */
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
 	ltp = xfs_dir2_leaf_tail_p(tp->t_mountp, leaf);
 	firstb = xfs_dir2_leaf_bests_p(ltp) + first;
 	lastb = xfs_dir2_leaf_bests_p(ltp) + last;
@@ -1202,8 +1222,8 @@ xfs_dir2_leaf_log_ents(
 	xfs_dir2_leaf_t		*leaf;		/* leaf structure */
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC ||
-	       be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC) ||
+	       leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	firstlep = &leaf->ents[first];
 	lastlep = &leaf->ents[last];
 	xfs_da_log_buf(tp, bp, (uint)((char *)firstlep - (char *)leaf),
@@ -1221,8 +1241,8 @@ xfs_dir2_leaf_log_header(
 	xfs_dir2_leaf_t		*leaf;		/* leaf structure */
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC ||
-	       be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC) ||
+	       leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	xfs_da_log_buf(tp, bp, (uint)((char *)&leaf->hdr - (char *)leaf),
 		(uint)(sizeof(leaf->hdr) - 1));
 }
@@ -1241,7 +1261,7 @@ xfs_dir2_leaf_log_tail(
 
 	mp = tp->t_mountp;
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
 	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	xfs_da_log_buf(tp, bp, (uint)((char *)ltp - (char *)leaf),
 		(uint)(mp->m_dirblksize - 1));
@@ -1437,7 +1457,7 @@ xfs_dir2_leaf_removename(
 	xfs_da_args_t		*args)		/* operation arguments */
 {
 	__be16			*bestsp;	/* leaf block best freespace */
-	xfs_dir2_data_t		*data;		/* data block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dir2_db_t		db;		/* data block number */
 	xfs_dabuf_t		*dbp;		/* data block buffer */
 	xfs_dir2_data_entry_t	*dep;		/* data entry structure */
@@ -1467,7 +1487,7 @@ xfs_dir2_leaf_removename(
 	tp = args->trans;
 	mp = dp->i_mount;
 	leaf = lbp->data;
-	data = dbp->data;
+	hdr = dbp->data;
 	xfs_dir2_data_check(dp, dbp);
 	/*
 	 * Point to the leaf entry, use that to point to the data entry.
@@ -1475,9 +1495,9 @@ xfs_dir2_leaf_removename(
 	lep = &leaf->ents[index];
 	db = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
 	dep = (xfs_dir2_data_entry_t *)
-	      ((char *)data + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address)));
+	      ((char *)hdr + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address)));
 	needscan = needlog = 0;
-	oldbest = be16_to_cpu(data->hdr.bestfree[0].length);
+	oldbest = be16_to_cpu(hdr->bestfree[0].length);
 	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
 	bestsp = xfs_dir2_leaf_bests_p(ltp);
 	ASSERT(be16_to_cpu(bestsp[db]) == oldbest);
@@ -1485,7 +1505,7 @@ xfs_dir2_leaf_removename(
 	 * Mark the former data entry unused.
 	 */
 	xfs_dir2_data_make_free(tp, dbp,
-		(xfs_dir2_data_aoff_t)((char *)dep - (char *)data),
+		(xfs_dir2_data_aoff_t)((char *)dep - (char *)hdr),
 		xfs_dir2_data_entsize(dep->namelen), &needlog, &needscan);
 	/*
 	 * We just mark the leaf entry stale by putting a null in it.
@@ -1499,23 +1519,23 @@ xfs_dir2_leaf_removename(
 	 * log the data block header if necessary.
 	 */
 	if (needscan)
-		xfs_dir2_data_freescan(mp, data, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	if (needlog)
 		xfs_dir2_data_log_header(tp, dbp);
 	/*
 	 * If the longest freespace in the data block has changed,
 	 * put the new value in the bests table and log that.
 	 */
-	if (be16_to_cpu(data->hdr.bestfree[0].length) != oldbest) {
-		bestsp[db] = data->hdr.bestfree[0].length;
+	if (be16_to_cpu(hdr->bestfree[0].length) != oldbest) {
+		bestsp[db] = hdr->bestfree[0].length;
 		xfs_dir2_leaf_log_bests(tp, lbp, db, db);
 	}
 	xfs_dir2_data_check(dp, dbp);
 	/*
 	 * If the data block is now empty then get rid of the data block.
 	 */
-	if (be16_to_cpu(data->hdr.bestfree[0].length) ==
-	    mp->m_dirblksize - (uint)sizeof(data->hdr)) {
+	if (be16_to_cpu(hdr->bestfree[0].length) ==
+	    mp->m_dirblksize - (uint)sizeof(*hdr)) {
 		ASSERT(db != mp->m_dirdatablk);
 		if ((error = xfs_dir2_shrink_inode(args, db, dbp))) {
 			/*
@@ -1542,7 +1562,7 @@ xfs_dir2_leaf_removename(
 			 * Look for the last active entry (i).
 			 */
 			for (i = db - 1; i > 0; i--) {
-				if (be16_to_cpu(bestsp[i]) != NULLDATAOFF)
+				if (bestsp[i] != cpu_to_be16(NULLDATAOFF))
 					break;
 			}
 			/*
@@ -1686,9 +1706,6 @@ xfs_dir2_leaf_trim_data(
 	xfs_dir2_db_t		db)		/* data block number */
 {
 	__be16			*bestsp;	/* leaf bests table */
-#ifdef DEBUG
-	xfs_dir2_data_t		*data;		/* data block structure */
-#endif
 	xfs_dabuf_t		*dbp;		/* data block buffer */
 	xfs_inode_t		*dp;		/* incore directory inode */
 	int			error;		/* error return value */
@@ -1707,20 +1724,21 @@ xfs_dir2_leaf_trim_data(
 			XFS_DATA_FORK))) {
 		return error;
 	}
-#ifdef DEBUG
-	data = dbp->data;
-	ASSERT(be32_to_cpu(data->hdr.magic) == XFS_DIR2_DATA_MAGIC);
-#endif
-	/* this seems to be an error
-	 * data is only valid if DEBUG is defined?
-	 * RMC 09/08/1999
-	 */
 
 	leaf = lbp->data;
 	ltp = xfs_dir2_leaf_tail_p(mp, leaf);
-	ASSERT(be16_to_cpu(data->hdr.bestfree[0].length) ==
-	       mp->m_dirblksize - (uint)sizeof(data->hdr));
+
+#ifdef DEBUG
+{
+	struct xfs_dir2_data_hdr *hdr = dbp->data;
+
+	ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC));
+	ASSERT(be16_to_cpu(hdr->bestfree[0].length) ==
+	       mp->m_dirblksize - (uint)sizeof(*hdr));
 	ASSERT(db == be32_to_cpu(ltp->bestcount) - 1);
+}
+#endif
+
 	/*
 	 * Get rid of the data block.
 	 */
@@ -1740,6 +1758,20 @@ xfs_dir2_leaf_trim_data(
 	return 0;
 }
 
+static inline size_t
+xfs_dir2_leaf_size(
+	struct xfs_dir2_leaf_hdr	*hdr,
+	int				counts)
+{
+	int			entries;
+
+	entries = be16_to_cpu(hdr->count) - be16_to_cpu(hdr->stale);
+	return sizeof(xfs_dir2_leaf_hdr_t) +
+	    entries * sizeof(xfs_dir2_leaf_entry_t) +
+	    counts * sizeof(xfs_dir2_data_off_t) +
+	    sizeof(xfs_dir2_leaf_tail_t);
+}
+
 /*
  * Convert node form directory to leaf form directory.
  * The root of the node form dir needs to already be a LEAFN block.
@@ -1810,7 +1842,7 @@ xfs_dir2_node_to_leaf(
 		return 0;
 	lbp = state->path.blk[0].bp;
 	leaf = lbp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	/*
 	 * Read the freespace block.
 	 */
@@ -1819,20 +1851,19 @@ xfs_dir2_node_to_leaf(
 		return error;
 	}
 	free = fbp->data;
-	ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+	ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 	ASSERT(!free->hdr.firstdb);
+
 	/*
 	 * Now see if the leafn and free data will fit in a leaf1.
 	 * If not, release the buffer and give up.
 	 */
-	if ((uint)sizeof(leaf->hdr) +
-	    (be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale)) * (uint)sizeof(leaf->ents[0]) +
-	    be32_to_cpu(free->hdr.nvalid) * (uint)sizeof(leaf->bests[0]) +
-	    (uint)sizeof(leaf->tail) >
-	    mp->m_dirblksize) {
+	if (xfs_dir2_leaf_size(&leaf->hdr, be32_to_cpu(free->hdr.nvalid)) >
+			mp->m_dirblksize) {
 		xfs_da_brelse(tp, fbp);
 		return 0;
 	}
+
 	/*
 	 * If the leaf has any stale entries in it, compress them out.
 	 * The compact routine will log the header.
@@ -1851,7 +1882,7 @@ xfs_dir2_node_to_leaf(
 	 * Set up the leaf bests table.
 	 */
 	memcpy(xfs_dir2_leaf_bests_p(ltp), free->bests,
-		be32_to_cpu(ltp->bestcount) * sizeof(leaf->bests[0]));
+		be32_to_cpu(ltp->bestcount) * sizeof(xfs_dir2_data_off_t));
 	xfs_dir2_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
 	xfs_dir2_leaf_log_tail(tp, lbp);
 	xfs_dir2_leaf_check(dp, lbp);
diff --git a/fs/xfs/xfs_dir2_leaf.h b/fs/xfs/xfs_dir2_leaf.h
deleted file mode 100644
index 6c9539f0698..00000000000
--- a/fs/xfs/xfs_dir2_leaf.h
+++ /dev/null
@@ -1,253 +0,0 @@
-/*
- * Copyright (c) 2000-2001,2005 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_DIR2_LEAF_H__
-#define	__XFS_DIR2_LEAF_H__
-
-struct uio;
-struct xfs_dabuf;
-struct xfs_da_args;
-struct xfs_inode;
-struct xfs_mount;
-struct xfs_trans;
-
-/*
- * Offset of the leaf/node space.  First block in this space
- * is the btree root.
- */
-#define	XFS_DIR2_LEAF_SPACE	1
-#define	XFS_DIR2_LEAF_OFFSET	(XFS_DIR2_LEAF_SPACE * XFS_DIR2_SPACE_SIZE)
-#define	XFS_DIR2_LEAF_FIRSTDB(mp)	\
-	xfs_dir2_byte_to_db(mp, XFS_DIR2_LEAF_OFFSET)
-
-/*
- * Offset in data space of a data entry.
- */
-typedef	__uint32_t	xfs_dir2_dataptr_t;
-#define	XFS_DIR2_MAX_DATAPTR	((xfs_dir2_dataptr_t)0xffffffff)
-#define	XFS_DIR2_NULL_DATAPTR	((xfs_dir2_dataptr_t)0)
-
-/*
- * Leaf block header.
- */
-typedef struct xfs_dir2_leaf_hdr {
-	xfs_da_blkinfo_t	info;		/* header for da routines */
-	__be16			count;		/* count of entries */
-	__be16			stale;		/* count of stale entries */
-} xfs_dir2_leaf_hdr_t;
-
-/*
- * Leaf block entry.
- */
-typedef struct xfs_dir2_leaf_entry {
-	__be32			hashval;	/* hash value of name */
-	__be32			address;	/* address of data entry */
-} xfs_dir2_leaf_entry_t;
-
-/*
- * Leaf block tail.
- */
-typedef struct xfs_dir2_leaf_tail {
-	__be32			bestcount;
-} xfs_dir2_leaf_tail_t;
-
-/*
- * Leaf block.
- * bests and tail are at the end of the block for single-leaf only
- * (magic = XFS_DIR2_LEAF1_MAGIC not XFS_DIR2_LEAFN_MAGIC).
- */
-typedef struct xfs_dir2_leaf {
-	xfs_dir2_leaf_hdr_t	hdr;		/* leaf header */
-	xfs_dir2_leaf_entry_t	ents[1];	/* entries */
-						/* ... */
-	xfs_dir2_data_off_t	bests[1];	/* best free counts */
-	xfs_dir2_leaf_tail_t	tail;		/* leaf tail */
-} xfs_dir2_leaf_t;
-
-/*
- * DB blocks here are logical directory block numbers, not filesystem blocks.
- */
-
-static inline int xfs_dir2_max_leaf_ents(struct xfs_mount *mp)
-{
-	return (int)(((mp)->m_dirblksize - (uint)sizeof(xfs_dir2_leaf_hdr_t)) /
-	       (uint)sizeof(xfs_dir2_leaf_entry_t));
-}
-
-/*
- * Get address of the bestcount field in the single-leaf block.
- */
-static inline xfs_dir2_leaf_tail_t *
-xfs_dir2_leaf_tail_p(struct xfs_mount *mp, xfs_dir2_leaf_t *lp)
-{
-	return (xfs_dir2_leaf_tail_t *)
-		((char *)(lp) + (mp)->m_dirblksize - 
-		  (uint)sizeof(xfs_dir2_leaf_tail_t));
-}
-
-/*
- * Get address of the bests array in the single-leaf block.
- */
-static inline __be16 *
-xfs_dir2_leaf_bests_p(xfs_dir2_leaf_tail_t *ltp)
-{
-	return (__be16 *)ltp - be32_to_cpu(ltp->bestcount);
-}
-
-/*
- * Convert dataptr to byte in file space
- */
-static inline xfs_dir2_off_t
-xfs_dir2_dataptr_to_byte(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
-{
-	return (xfs_dir2_off_t)(dp) << XFS_DIR2_DATA_ALIGN_LOG;
-}
-
-/*
- * Convert byte in file space to dataptr.  It had better be aligned.
- */
-static inline xfs_dir2_dataptr_t
-xfs_dir2_byte_to_dataptr(struct xfs_mount *mp, xfs_dir2_off_t by)
-{
-	return (xfs_dir2_dataptr_t)((by) >> XFS_DIR2_DATA_ALIGN_LOG);
-}
-
-/*
- * Convert byte in space to (DB) block
- */
-static inline xfs_dir2_db_t
-xfs_dir2_byte_to_db(struct xfs_mount *mp, xfs_dir2_off_t by)
-{
-	return (xfs_dir2_db_t)((by) >> \
-		 ((mp)->m_sb.sb_blocklog + (mp)->m_sb.sb_dirblklog));
-}
-
-/*
- * Convert dataptr to a block number
- */
-static inline xfs_dir2_db_t
-xfs_dir2_dataptr_to_db(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
-{
-	return xfs_dir2_byte_to_db(mp, xfs_dir2_dataptr_to_byte(mp, dp));
-}
-
-/*
- * Convert byte in space to offset in a block
- */
-static inline xfs_dir2_data_aoff_t
-xfs_dir2_byte_to_off(struct xfs_mount *mp, xfs_dir2_off_t by)
-{
-	return (xfs_dir2_data_aoff_t)((by) & \
-		((1 << ((mp)->m_sb.sb_blocklog + (mp)->m_sb.sb_dirblklog)) - 1));
-}
-
-/*
- * Convert dataptr to a byte offset in a block
- */
-static inline xfs_dir2_data_aoff_t
-xfs_dir2_dataptr_to_off(struct xfs_mount *mp, xfs_dir2_dataptr_t dp)
-{
-	return xfs_dir2_byte_to_off(mp, xfs_dir2_dataptr_to_byte(mp, dp));
-}
-
-/*
- * Convert block and offset to byte in space
- */
-static inline xfs_dir2_off_t
-xfs_dir2_db_off_to_byte(struct xfs_mount *mp, xfs_dir2_db_t db,
-			xfs_dir2_data_aoff_t o)
-{
-	return ((xfs_dir2_off_t)(db) << \
-		((mp)->m_sb.sb_blocklog + (mp)->m_sb.sb_dirblklog)) + (o);
-}
-
-/*
- * Convert block (DB) to block (dablk)
- */
-static inline xfs_dablk_t
-xfs_dir2_db_to_da(struct xfs_mount *mp, xfs_dir2_db_t db)
-{
-	return (xfs_dablk_t)((db) << (mp)->m_sb.sb_dirblklog);
-}
-
-/*
- * Convert byte in space to (DA) block
- */
-static inline xfs_dablk_t
-xfs_dir2_byte_to_da(struct xfs_mount *mp, xfs_dir2_off_t by)
-{
-	return xfs_dir2_db_to_da(mp, xfs_dir2_byte_to_db(mp, by));
-}
-
-/*
- * Convert block and offset to dataptr
- */
-static inline xfs_dir2_dataptr_t
-xfs_dir2_db_off_to_dataptr(struct xfs_mount *mp, xfs_dir2_db_t db,
-			   xfs_dir2_data_aoff_t o)
-{
-	return xfs_dir2_byte_to_dataptr(mp, xfs_dir2_db_off_to_byte(mp, db, o));
-}
-
-/*
- * Convert block (dablk) to block (DB)
- */
-static inline xfs_dir2_db_t
-xfs_dir2_da_to_db(struct xfs_mount *mp, xfs_dablk_t da)
-{
-	return (xfs_dir2_db_t)((da) >> (mp)->m_sb.sb_dirblklog);
-}
-
-/*
- * Convert block (dablk) to byte offset in space
- */
-static inline xfs_dir2_off_t
-xfs_dir2_da_to_byte(struct xfs_mount *mp, xfs_dablk_t da)
-{
-	return xfs_dir2_db_off_to_byte(mp, xfs_dir2_da_to_db(mp, da), 0);
-}
-
-/*
- * Function declarations.
- */
-extern int xfs_dir2_block_to_leaf(struct xfs_da_args *args,
-				  struct xfs_dabuf *dbp);
-extern int xfs_dir2_leaf_addname(struct xfs_da_args *args);
-extern void xfs_dir2_leaf_compact(struct xfs_da_args *args,
-				  struct xfs_dabuf *bp);
-extern void xfs_dir2_leaf_compact_x1(struct xfs_dabuf *bp, int *indexp,
-				     int *lowstalep, int *highstalep,
-				     int *lowlogp, int *highlogp);
-extern int xfs_dir2_leaf_getdents(struct xfs_inode *dp, void *dirent,
-				  size_t bufsize, xfs_off_t *offset,
-				  filldir_t filldir);
-extern int xfs_dir2_leaf_init(struct xfs_da_args *args, xfs_dir2_db_t bno,
-			      struct xfs_dabuf **bpp, int magic);
-extern void xfs_dir2_leaf_log_ents(struct xfs_trans *tp, struct xfs_dabuf *bp,
-				   int first, int last);
-extern void xfs_dir2_leaf_log_header(struct xfs_trans *tp,
-				     struct xfs_dabuf *bp);
-extern int xfs_dir2_leaf_lookup(struct xfs_da_args *args);
-extern int xfs_dir2_leaf_removename(struct xfs_da_args *args);
-extern int xfs_dir2_leaf_replace(struct xfs_da_args *args);
-extern int xfs_dir2_leaf_search_hash(struct xfs_da_args *args,
-				     struct xfs_dabuf *lbp);
-extern int xfs_dir2_leaf_trim_data(struct xfs_da_args *args,
-				   struct xfs_dabuf *lbp, xfs_dir2_db_t db);
-extern int xfs_dir2_node_to_leaf(struct xfs_da_state *state);
-
-#endif	/* __XFS_DIR2_LEAF_H__ */
diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c
index a0aab7d3294..084b3247d63 100644
--- a/fs/xfs/xfs_dir2_node.c
+++ b/fs/xfs/xfs_dir2_node.c
@@ -23,18 +23,14 @@
 #include "xfs_trans.h"
 #include "xfs_sb.h"
 #include "xfs_ag.h"
-#include "xfs_dir2.h"
 #include "xfs_mount.h"
 #include "xfs_da_btree.h"
 #include "xfs_bmap_btree.h"
-#include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_bmap.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
-#include "xfs_dir2_node.h"
+#include "xfs_dir2_format.h"
+#include "xfs_dir2_priv.h"
 #include "xfs_error.h"
 #include "xfs_trace.h"
 
@@ -73,7 +69,7 @@ xfs_dir2_free_log_bests(
 	xfs_dir2_free_t		*free;		/* freespace structure */
 
 	free = bp->data;
-	ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+	ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 	xfs_da_log_buf(tp, bp,
 		(uint)((char *)&free->bests[first] - (char *)free),
 		(uint)((char *)&free->bests[last] - (char *)free +
@@ -91,7 +87,7 @@ xfs_dir2_free_log_header(
 	xfs_dir2_free_t		*free;		/* freespace structure */
 
 	free = bp->data;
-	ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+	ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 	xfs_da_log_buf(tp, bp, (uint)((char *)&free->hdr - (char *)free),
 		(uint)(sizeof(xfs_dir2_free_hdr_t) - 1));
 }
@@ -244,89 +240,13 @@ xfs_dir2_leafn_add(
 		lfloglow = be16_to_cpu(leaf->hdr.count);
 		lfloghigh = -1;
 	}
-	/*
-	 * No stale entries, just insert a space for the new entry.
-	 */
-	if (!leaf->hdr.stale) {
-		lep = &leaf->ents[index];
-		if (index < be16_to_cpu(leaf->hdr.count))
-			memmove(lep + 1, lep,
-				(be16_to_cpu(leaf->hdr.count) - index) * sizeof(*lep));
-		lfloglow = index;
-		lfloghigh = be16_to_cpu(leaf->hdr.count);
-		be16_add_cpu(&leaf->hdr.count, 1);
-	}
-	/*
-	 * There are stale entries.  We'll use one for the new entry.
-	 */
-	else {
-		/*
-		 * If we didn't do a compact then we need to figure out
-		 * which stale entry will be used.
-		 */
-		if (compact == 0) {
-			/*
-			 * Find first stale entry before our insertion point.
-			 */
-			for (lowstale = index - 1;
-			     lowstale >= 0 &&
-				be32_to_cpu(leaf->ents[lowstale].address) !=
-				XFS_DIR2_NULL_DATAPTR;
-			     lowstale--)
-				continue;
-			/*
-			 * Find next stale entry after insertion point.
-			 * Stop looking if the answer would be worse than
-			 * lowstale already found.
-			 */
-			for (highstale = index;
-			     highstale < be16_to_cpu(leaf->hdr.count) &&
-				be32_to_cpu(leaf->ents[highstale].address) !=
-				XFS_DIR2_NULL_DATAPTR &&
-				(lowstale < 0 ||
-				 index - lowstale - 1 >= highstale - index);
-			     highstale++)
-				continue;
-		}
-		/*
-		 * Using the low stale entry.
-		 * Shift entries up toward the stale slot.
-		 */
-		if (lowstale >= 0 &&
-		    (highstale == be16_to_cpu(leaf->hdr.count) ||
-		     index - lowstale - 1 < highstale - index)) {
-			ASSERT(be32_to_cpu(leaf->ents[lowstale].address) ==
-			       XFS_DIR2_NULL_DATAPTR);
-			ASSERT(index - lowstale - 1 >= 0);
-			if (index - lowstale - 1 > 0)
-				memmove(&leaf->ents[lowstale],
-					&leaf->ents[lowstale + 1],
-					(index - lowstale - 1) * sizeof(*lep));
-			lep = &leaf->ents[index - 1];
-			lfloglow = MIN(lowstale, lfloglow);
-			lfloghigh = MAX(index - 1, lfloghigh);
-		}
-		/*
-		 * Using the high stale entry.
-		 * Shift entries down toward the stale slot.
-		 */
-		else {
-			ASSERT(be32_to_cpu(leaf->ents[highstale].address) ==
-			       XFS_DIR2_NULL_DATAPTR);
-			ASSERT(highstale - index >= 0);
-			if (highstale - index > 0)
-				memmove(&leaf->ents[index + 1],
-					&leaf->ents[index],
-					(highstale - index) * sizeof(*lep));
-			lep = &leaf->ents[index];
-			lfloglow = MIN(index, lfloglow);
-			lfloghigh = MAX(highstale, lfloghigh);
-		}
-		be16_add_cpu(&leaf->hdr.stale, -1);
-	}
+
 	/*
 	 * Insert the new entry, log everything.
 	 */
+	lep = xfs_dir2_leaf_find_entry(leaf, index, compact, lowstale,
+				       highstale, &lfloglow, &lfloghigh);
+
 	lep->hashval = cpu_to_be32(args->hashval);
 	lep->address = cpu_to_be32(xfs_dir2_db_off_to_dataptr(mp,
 				args->blkno, args->index));
@@ -352,14 +272,14 @@ xfs_dir2_leafn_check(
 
 	leaf = bp->data;
 	mp = dp->i_mount;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	ASSERT(be16_to_cpu(leaf->hdr.count) <= xfs_dir2_max_leaf_ents(mp));
 	for (i = stale = 0; i < be16_to_cpu(leaf->hdr.count); i++) {
 		if (i + 1 < be16_to_cpu(leaf->hdr.count)) {
 			ASSERT(be32_to_cpu(leaf->ents[i].hashval) <=
 			       be32_to_cpu(leaf->ents[i + 1].hashval));
 		}
-		if (be32_to_cpu(leaf->ents[i].address) == XFS_DIR2_NULL_DATAPTR)
+		if (leaf->ents[i].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
 			stale++;
 	}
 	ASSERT(be16_to_cpu(leaf->hdr.stale) == stale);
@@ -378,7 +298,7 @@ xfs_dir2_leafn_lasthash(
 	xfs_dir2_leaf_t	*leaf;			/* leaf structure */
 
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	if (count)
 		*count = be16_to_cpu(leaf->hdr.count);
 	if (!leaf->hdr.count)
@@ -417,7 +337,7 @@ xfs_dir2_leafn_lookup_for_addname(
 	tp = args->trans;
 	mp = dp->i_mount;
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 #ifdef __KERNEL__
 	ASSERT(be16_to_cpu(leaf->hdr.count) > 0);
 #endif
@@ -434,7 +354,7 @@ xfs_dir2_leafn_lookup_for_addname(
 		curbp = state->extrablk.bp;
 		curfdb = state->extrablk.blkno;
 		free = curbp->data;
-		ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+		ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 	}
 	length = xfs_dir2_data_entsize(args->namelen);
 	/*
@@ -488,7 +408,7 @@ xfs_dir2_leafn_lookup_for_addname(
 				ASSERT(be32_to_cpu(free->hdr.magic) ==
 					XFS_DIR2_FREE_MAGIC);
 				ASSERT((be32_to_cpu(free->hdr.firstdb) %
-					XFS_DIR2_MAX_FREE_BESTS(mp)) == 0);
+					xfs_dir2_free_max_bests(mp)) == 0);
 				ASSERT(be32_to_cpu(free->hdr.firstdb) <= curdb);
 				ASSERT(curdb < be32_to_cpu(free->hdr.firstdb) +
 					be32_to_cpu(free->hdr.nvalid));
@@ -500,7 +420,8 @@ xfs_dir2_leafn_lookup_for_addname(
 			/*
 			 * If it has room, return it.
 			 */
-			if (unlikely(be16_to_cpu(free->bests[fi]) == NULLDATAOFF)) {
+			if (unlikely(free->bests[fi] ==
+			    cpu_to_be16(NULLDATAOFF))) {
 				XFS_ERROR_REPORT("xfs_dir2_leafn_lookup_int",
 							XFS_ERRLEVEL_LOW, mp);
 				if (curfdb != newfdb)
@@ -561,7 +482,7 @@ xfs_dir2_leafn_lookup_for_entry(
 	tp = args->trans;
 	mp = dp->i_mount;
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 #ifdef __KERNEL__
 	ASSERT(be16_to_cpu(leaf->hdr.count) > 0);
 #endif
@@ -742,7 +663,8 @@ xfs_dir2_leafn_moveents(
 		int	i;			/* temp leaf index */
 
 		for (i = start_s, stale = 0; i < start_s + count; i++) {
-			if (be32_to_cpu(leaf_s->ents[i].address) == XFS_DIR2_NULL_DATAPTR)
+			if (leaf_s->ents[i].address ==
+			    cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
 				stale++;
 		}
 	} else
@@ -789,8 +711,8 @@ xfs_dir2_leafn_order(
 
 	leaf1 = leaf1_bp->data;
 	leaf2 = leaf2_bp->data;
-	ASSERT(be16_to_cpu(leaf1->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
-	ASSERT(be16_to_cpu(leaf2->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf1->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+	ASSERT(leaf2->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	if (be16_to_cpu(leaf1->hdr.count) > 0 &&
 	    be16_to_cpu(leaf2->hdr.count) > 0 &&
 	    (be32_to_cpu(leaf2->ents[0].hashval) < be32_to_cpu(leaf1->ents[0].hashval) ||
@@ -918,7 +840,7 @@ xfs_dir2_leafn_remove(
 	xfs_da_state_blk_t	*dblk,		/* data block */
 	int			*rval)		/* resulting block needs join */
 {
-	xfs_dir2_data_t		*data;		/* data block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dir2_db_t		db;		/* data block number */
 	xfs_dabuf_t		*dbp;		/* data block buffer */
 	xfs_dir2_data_entry_t	*dep;		/* data block entry */
@@ -938,7 +860,7 @@ xfs_dir2_leafn_remove(
 	tp = args->trans;
 	mp = dp->i_mount;
 	leaf = bp->data;
-	ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	/*
 	 * Point to the entry we're removing.
 	 */
@@ -963,9 +885,9 @@ xfs_dir2_leafn_remove(
 	 * in the data block in case it changes.
 	 */
 	dbp = dblk->bp;
-	data = dbp->data;
-	dep = (xfs_dir2_data_entry_t *)((char *)data + off);
-	longest = be16_to_cpu(data->hdr.bestfree[0].length);
+	hdr = dbp->data;
+	dep = (xfs_dir2_data_entry_t *)((char *)hdr + off);
+	longest = be16_to_cpu(hdr->bestfree[0].length);
 	needlog = needscan = 0;
 	xfs_dir2_data_make_free(tp, dbp, off,
 		xfs_dir2_data_entsize(dep->namelen), &needlog, &needscan);
@@ -974,7 +896,7 @@ xfs_dir2_leafn_remove(
 	 * Log the data block header if needed.
 	 */
 	if (needscan)
-		xfs_dir2_data_freescan(mp, data, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	if (needlog)
 		xfs_dir2_data_log_header(tp, dbp);
 	xfs_dir2_data_check(dp, dbp);
@@ -982,7 +904,7 @@ xfs_dir2_leafn_remove(
 	 * If the longest data block freespace changes, need to update
 	 * the corresponding freeblock entry.
 	 */
-	if (longest < be16_to_cpu(data->hdr.bestfree[0].length)) {
+	if (longest < be16_to_cpu(hdr->bestfree[0].length)) {
 		int		error;		/* error return value */
 		xfs_dabuf_t	*fbp;		/* freeblock buffer */
 		xfs_dir2_db_t	fdb;		/* freeblock block number */
@@ -1000,27 +922,27 @@ xfs_dir2_leafn_remove(
 			return error;
 		}
 		free = fbp->data;
-		ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+		ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 		ASSERT(be32_to_cpu(free->hdr.firstdb) ==
-		       XFS_DIR2_MAX_FREE_BESTS(mp) *
+		       xfs_dir2_free_max_bests(mp) *
 		       (fdb - XFS_DIR2_FREE_FIRSTDB(mp)));
 		/*
 		 * Calculate which entry we need to fix.
 		 */
 		findex = xfs_dir2_db_to_fdindex(mp, db);
-		longest = be16_to_cpu(data->hdr.bestfree[0].length);
+		longest = be16_to_cpu(hdr->bestfree[0].length);
 		/*
 		 * If the data block is now empty we can get rid of it
 		 * (usually).
 		 */
-		if (longest == mp->m_dirblksize - (uint)sizeof(data->hdr)) {
+		if (longest == mp->m_dirblksize - (uint)sizeof(*hdr)) {
 			/*
 			 * Try to punch out the data block.
 			 */
 			error = xfs_dir2_shrink_inode(args, db, dbp);
 			if (error == 0) {
 				dblk->bp = NULL;
-				data = NULL;
+				hdr = NULL;
 			}
 			/*
 			 * We can get ENOSPC if there's no space reservation.
@@ -1036,7 +958,7 @@ xfs_dir2_leafn_remove(
 		 * If we got rid of the data block, we can eliminate that entry
 		 * in the free block.
 		 */
-		if (data == NULL) {
+		if (hdr == NULL) {
 			/*
 			 * One less used entry in the free table.
 			 */
@@ -1052,7 +974,8 @@ xfs_dir2_leafn_remove(
 				int	i;		/* free entry index */
 
 				for (i = findex - 1;
-				     i >= 0 && be16_to_cpu(free->bests[i]) == NULLDATAOFF;
+				     i >= 0 &&
+				     free->bests[i] == cpu_to_be16(NULLDATAOFF);
 				     i--)
 					continue;
 				free->hdr.nvalid = cpu_to_be32(i + 1);
@@ -1209,7 +1132,7 @@ xfs_dir2_leafn_toosmall(
 	 */
 	blk = &state->path.blk[state->path.active - 1];
 	info = blk->bp->data;
-	ASSERT(be16_to_cpu(info->magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	leaf = (xfs_dir2_leaf_t *)info;
 	count = be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale);
 	bytes = (uint)sizeof(leaf->hdr) + count * (uint)sizeof(leaf->ents[0]);
@@ -1268,7 +1191,7 @@ xfs_dir2_leafn_toosmall(
 		count = be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale);
 		bytes = state->blocksize - (state->blocksize >> 2);
 		leaf = bp->data;
-		ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+		ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 		count += be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale);
 		bytes -= count * (uint)sizeof(leaf->ents[0]);
 		/*
@@ -1327,8 +1250,8 @@ xfs_dir2_leafn_unbalance(
 	ASSERT(save_blk->magic == XFS_DIR2_LEAFN_MAGIC);
 	drop_leaf = drop_blk->bp->data;
 	save_leaf = save_blk->bp->data;
-	ASSERT(be16_to_cpu(drop_leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
-	ASSERT(be16_to_cpu(save_leaf->hdr.info.magic) == XFS_DIR2_LEAFN_MAGIC);
+	ASSERT(drop_leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+	ASSERT(save_leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
 	/*
 	 * If there are any stale leaf entries, take this opportunity
 	 * to purge them.
@@ -1432,7 +1355,7 @@ xfs_dir2_node_addname_int(
 	xfs_da_args_t		*args,		/* operation arguments */
 	xfs_da_state_blk_t	*fblk)		/* optional freespace block */
 {
-	xfs_dir2_data_t		*data;		/* data block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dir2_db_t		dbno;		/* data block number */
 	xfs_dabuf_t		*dbp;		/* data block buffer */
 	xfs_dir2_data_entry_t	*dep;		/* data entry pointer */
@@ -1469,7 +1392,7 @@ xfs_dir2_node_addname_int(
 		 */
 		ifbno = fblk->blkno;
 		free = fbp->data;
-		ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+		ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 		findex = fblk->index;
 		/*
 		 * This means the free entry showed that the data block had
@@ -1553,7 +1476,7 @@ xfs_dir2_node_addname_int(
 				continue;
 			}
 			free = fbp->data;
-			ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+			ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 			findex = 0;
 		}
 		/*
@@ -1680,12 +1603,12 @@ xfs_dir2_node_addname_int(
 			free->hdr.magic = cpu_to_be32(XFS_DIR2_FREE_MAGIC);
 			free->hdr.firstdb = cpu_to_be32(
 				(fbno - XFS_DIR2_FREE_FIRSTDB(mp)) *
-				XFS_DIR2_MAX_FREE_BESTS(mp));
+				xfs_dir2_free_max_bests(mp));
 			free->hdr.nvalid = 0;
 			free->hdr.nused = 0;
 		} else {
 			free = fbp->data;
-			ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+			ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 		}
 
 		/*
@@ -1697,7 +1620,7 @@ xfs_dir2_node_addname_int(
 		 * freespace block, extend that table.
 		 */
 		if (findex >= be32_to_cpu(free->hdr.nvalid)) {
-			ASSERT(findex < XFS_DIR2_MAX_FREE_BESTS(mp));
+			ASSERT(findex < xfs_dir2_free_max_bests(mp));
 			free->hdr.nvalid = cpu_to_be32(findex + 1);
 			/*
 			 * Tag new entry so nused will go up.
@@ -1708,7 +1631,7 @@ xfs_dir2_node_addname_int(
 		 * If this entry was for an empty data block
 		 * (this should always be true) then update the header.
 		 */
-		if (be16_to_cpu(free->bests[findex]) == NULLDATAOFF) {
+		if (free->bests[findex] == cpu_to_be16(NULLDATAOFF)) {
 			be32_add_cpu(&free->hdr.nused, 1);
 			xfs_dir2_free_log_header(tp, fbp);
 		}
@@ -1717,8 +1640,8 @@ xfs_dir2_node_addname_int(
 		 * We haven't allocated the data entry yet so this will
 		 * change again.
 		 */
-		data = dbp->data;
-		free->bests[findex] = data->hdr.bestfree[0].length;
+		hdr = dbp->data;
+		free->bests[findex] = hdr->bestfree[0].length;
 		logfree = 1;
 	}
 	/*
@@ -1743,21 +1666,21 @@ xfs_dir2_node_addname_int(
 				xfs_da_buf_done(fbp);
 			return error;
 		}
-		data = dbp->data;
+		hdr = dbp->data;
 		logfree = 0;
 	}
-	ASSERT(be16_to_cpu(data->hdr.bestfree[0].length) >= length);
+	ASSERT(be16_to_cpu(hdr->bestfree[0].length) >= length);
 	/*
 	 * Point to the existing unused space.
 	 */
 	dup = (xfs_dir2_data_unused_t *)
-	      ((char *)data + be16_to_cpu(data->hdr.bestfree[0].offset));
+	      ((char *)hdr + be16_to_cpu(hdr->bestfree[0].offset));
 	needscan = needlog = 0;
 	/*
 	 * Mark the first part of the unused space, inuse for us.
 	 */
 	xfs_dir2_data_use_free(tp, dbp, dup,
-		(xfs_dir2_data_aoff_t)((char *)dup - (char *)data), length,
+		(xfs_dir2_data_aoff_t)((char *)dup - (char *)hdr), length,
 		&needlog, &needscan);
 	/*
 	 * Fill in the new entry and log it.
@@ -1767,13 +1690,13 @@ xfs_dir2_node_addname_int(
 	dep->namelen = args->namelen;
 	memcpy(dep->name, args->name, dep->namelen);
 	tagp = xfs_dir2_data_entry_tag_p(dep);
-	*tagp = cpu_to_be16((char *)dep - (char *)data);
+	*tagp = cpu_to_be16((char *)dep - (char *)hdr);
 	xfs_dir2_data_log_entry(tp, dbp, dep);
 	/*
 	 * Rescan the block for bestfree if needed.
 	 */
 	if (needscan)
-		xfs_dir2_data_freescan(mp, data, &needlog);
+		xfs_dir2_data_freescan(mp, hdr, &needlog);
 	/*
 	 * Log the data block header if needed.
 	 */
@@ -1782,8 +1705,8 @@ xfs_dir2_node_addname_int(
 	/*
 	 * If the freespace entry is now wrong, update it.
 	 */
-	if (be16_to_cpu(free->bests[findex]) != be16_to_cpu(data->hdr.bestfree[0].length)) {
-		free->bests[findex] = data->hdr.bestfree[0].length;
+	if (be16_to_cpu(free->bests[findex]) != be16_to_cpu(hdr->bestfree[0].length)) {
+		free->bests[findex] = hdr->bestfree[0].length;
 		logfree = 1;
 	}
 	/*
@@ -1933,7 +1856,7 @@ xfs_dir2_node_replace(
 	xfs_da_args_t		*args)		/* operation arguments */
 {
 	xfs_da_state_blk_t	*blk;		/* leaf block */
-	xfs_dir2_data_t		*data;		/* data block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* data block header */
 	xfs_dir2_data_entry_t	*dep;		/* data entry changed */
 	int			error;		/* error return value */
 	int			i;		/* btree level */
@@ -1977,10 +1900,10 @@ xfs_dir2_node_replace(
 		/*
 		 * Point to the data entry.
 		 */
-		data = state->extrablk.bp->data;
-		ASSERT(be32_to_cpu(data->hdr.magic) == XFS_DIR2_DATA_MAGIC);
+		hdr = state->extrablk.bp->data;
+		ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC));
 		dep = (xfs_dir2_data_entry_t *)
-		      ((char *)data +
+		      ((char *)hdr +
 		       xfs_dir2_dataptr_to_off(state->mp, be32_to_cpu(lep->address)));
 		ASSERT(inum != be64_to_cpu(dep->inumber));
 		/*
@@ -2044,7 +1967,7 @@ xfs_dir2_node_trim_free(
 		return 0;
 	}
 	free = bp->data;
-	ASSERT(be32_to_cpu(free->hdr.magic) == XFS_DIR2_FREE_MAGIC);
+	ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
 	/*
 	 * If there are used entries, there's nothing to do.
 	 */
diff --git a/fs/xfs/xfs_dir2_node.h b/fs/xfs/xfs_dir2_node.h
deleted file mode 100644
index 82dfe714719..00000000000
--- a/fs/xfs/xfs_dir2_node.h
+++ /dev/null
@@ -1,100 +0,0 @@
-/*
- * Copyright (c) 2000,2005 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_DIR2_NODE_H__
-#define	__XFS_DIR2_NODE_H__
-
-/*
- * Directory version 2, btree node format structures
- */
-
-struct uio;
-struct xfs_dabuf;
-struct xfs_da_args;
-struct xfs_da_state;
-struct xfs_da_state_blk;
-struct xfs_inode;
-struct xfs_trans;
-
-/*
- * Offset of the freespace index.
- */
-#define	XFS_DIR2_FREE_SPACE	2
-#define	XFS_DIR2_FREE_OFFSET	(XFS_DIR2_FREE_SPACE * XFS_DIR2_SPACE_SIZE)
-#define	XFS_DIR2_FREE_FIRSTDB(mp)	\
-	xfs_dir2_byte_to_db(mp, XFS_DIR2_FREE_OFFSET)
-
-#define	XFS_DIR2_FREE_MAGIC	0x58443246	/* XD2F */
-
-typedef	struct xfs_dir2_free_hdr {
-	__be32			magic;		/* XFS_DIR2_FREE_MAGIC */
-	__be32			firstdb;	/* db of first entry */
-	__be32			nvalid;		/* count of valid entries */
-	__be32			nused;		/* count of used entries */
-} xfs_dir2_free_hdr_t;
-
-typedef struct xfs_dir2_free {
-	xfs_dir2_free_hdr_t	hdr;		/* block header */
-	__be16			bests[1];	/* best free counts */
-						/* unused entries are -1 */
-} xfs_dir2_free_t;
-
-#define	XFS_DIR2_MAX_FREE_BESTS(mp)	\
-	(((mp)->m_dirblksize - (uint)sizeof(xfs_dir2_free_hdr_t)) / \
-	 (uint)sizeof(xfs_dir2_data_off_t))
-
-/*
- * Convert data space db to the corresponding free db.
- */
-static inline xfs_dir2_db_t
-xfs_dir2_db_to_fdb(struct xfs_mount *mp, xfs_dir2_db_t db)
-{
-	return (XFS_DIR2_FREE_FIRSTDB(mp) + (db) / XFS_DIR2_MAX_FREE_BESTS(mp));
-}
-
-/*
- * Convert data space db to the corresponding index in a free db.
- */
-static inline int
-xfs_dir2_db_to_fdindex(struct xfs_mount *mp, xfs_dir2_db_t db)
-{
-	return ((db) % XFS_DIR2_MAX_FREE_BESTS(mp));
-}
-
-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);
-extern int xfs_dir2_leafn_lookup_int(struct xfs_dabuf *bp,
-				     struct xfs_da_args *args, int *indexp,
-				     struct xfs_da_state *state);
-extern int xfs_dir2_leafn_order(struct xfs_dabuf *leaf1_bp,
-				struct xfs_dabuf *leaf2_bp);
-extern int xfs_dir2_leafn_split(struct xfs_da_state *state,
-				struct xfs_da_state_blk *oldblk,
-				struct xfs_da_state_blk *newblk);
-extern int xfs_dir2_leafn_toosmall(struct xfs_da_state *state, int *action);
-extern void xfs_dir2_leafn_unbalance(struct xfs_da_state *state,
-				     struct xfs_da_state_blk *drop_blk,
-				     struct xfs_da_state_blk *save_blk);
-extern int xfs_dir2_node_addname(struct xfs_da_args *args);
-extern int xfs_dir2_node_lookup(struct xfs_da_args *args);
-extern int xfs_dir2_node_removename(struct xfs_da_args *args);
-extern int xfs_dir2_node_replace(struct xfs_da_args *args);
-extern int xfs_dir2_node_trim_free(struct xfs_da_args *args, xfs_fileoff_t fo,
-				   int *rvalp);
-
-#endif	/* __XFS_DIR2_NODE_H__ */
diff --git a/fs/xfs/xfs_dir2_priv.h b/fs/xfs/xfs_dir2_priv.h
new file mode 100644
index 00000000000..067f403ecf8
--- /dev/null
+++ b/fs/xfs/xfs_dir2_priv.h
@@ -0,0 +1,135 @@
+/*
+ * Copyright (c) 2000-2001,2005 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_DIR2_PRIV_H__
+#define __XFS_DIR2_PRIV_H__
+
+/* xfs_dir2.c */
+extern int xfs_dir_ino_validate(struct xfs_mount *mp, xfs_ino_t ino);
+extern int xfs_dir2_isblock(struct xfs_trans *tp, struct xfs_inode *dp, int *r);
+extern int xfs_dir2_isleaf(struct xfs_trans *tp, struct xfs_inode *dp, int *r);
+extern int xfs_dir2_grow_inode(struct xfs_da_args *args, int space,
+				xfs_dir2_db_t *dbp);
+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 unsigned char *name, int len);
+
+/* xfs_dir2_block.c */
+extern int xfs_dir2_block_addname(struct xfs_da_args *args);
+extern int xfs_dir2_block_getdents(struct xfs_inode *dp, void *dirent,
+		xfs_off_t *offset, filldir_t filldir);
+extern int xfs_dir2_block_lookup(struct xfs_da_args *args);
+extern int xfs_dir2_block_removename(struct xfs_da_args *args);
+extern int xfs_dir2_block_replace(struct xfs_da_args *args);
+extern int xfs_dir2_leaf_to_block(struct xfs_da_args *args,
+		struct xfs_dabuf *lbp, struct xfs_dabuf *dbp);
+
+/* xfs_dir2_data.c */
+#ifdef DEBUG
+extern void xfs_dir2_data_check(struct xfs_inode *dp, struct xfs_dabuf *bp);
+#else
+#define	xfs_dir2_data_check(dp,bp)
+#endif
+extern struct xfs_dir2_data_free *
+xfs_dir2_data_freeinsert(struct xfs_dir2_data_hdr *hdr,
+		struct xfs_dir2_data_unused *dup, int *loghead);
+extern void xfs_dir2_data_freescan(struct xfs_mount *mp,
+		struct xfs_dir2_data_hdr *hdr, int *loghead);
+extern int xfs_dir2_data_init(struct xfs_da_args *args, xfs_dir2_db_t blkno,
+		struct xfs_dabuf **bpp);
+extern void xfs_dir2_data_log_entry(struct xfs_trans *tp, struct xfs_dabuf *bp,
+		struct xfs_dir2_data_entry *dep);
+extern void xfs_dir2_data_log_header(struct xfs_trans *tp,
+		struct xfs_dabuf *bp);
+extern void xfs_dir2_data_log_unused(struct xfs_trans *tp, struct xfs_dabuf *bp,
+		struct xfs_dir2_data_unused *dup);
+extern void xfs_dir2_data_make_free(struct xfs_trans *tp, struct xfs_dabuf *bp,
+		xfs_dir2_data_aoff_t offset, xfs_dir2_data_aoff_t len,
+		int *needlogp, int *needscanp);
+extern void xfs_dir2_data_use_free(struct xfs_trans *tp, struct xfs_dabuf *bp,
+		struct xfs_dir2_data_unused *dup, xfs_dir2_data_aoff_t offset,
+		xfs_dir2_data_aoff_t len, int *needlogp, int *needscanp);
+
+/* xfs_dir2_leaf.c */
+extern int xfs_dir2_block_to_leaf(struct xfs_da_args *args,
+		struct xfs_dabuf *dbp);
+extern int xfs_dir2_leaf_addname(struct xfs_da_args *args);
+extern void xfs_dir2_leaf_compact(struct xfs_da_args *args,
+		struct xfs_dabuf *bp);
+extern void xfs_dir2_leaf_compact_x1(struct xfs_dabuf *bp, int *indexp,
+		int *lowstalep, int *highstalep, int *lowlogp, int *highlogp);
+extern int xfs_dir2_leaf_getdents(struct xfs_inode *dp, void *dirent,
+		size_t bufsize, xfs_off_t *offset, filldir_t filldir);
+extern int xfs_dir2_leaf_init(struct xfs_da_args *args, xfs_dir2_db_t bno,
+		struct xfs_dabuf **bpp, int magic);
+extern void xfs_dir2_leaf_log_ents(struct xfs_trans *tp, struct xfs_dabuf *bp,
+		int first, int last);
+extern void xfs_dir2_leaf_log_header(struct xfs_trans *tp,
+		struct xfs_dabuf *bp);
+extern int xfs_dir2_leaf_lookup(struct xfs_da_args *args);
+extern int xfs_dir2_leaf_removename(struct xfs_da_args *args);
+extern int xfs_dir2_leaf_replace(struct xfs_da_args *args);
+extern int xfs_dir2_leaf_search_hash(struct xfs_da_args *args,
+		struct xfs_dabuf *lbp);
+extern int xfs_dir2_leaf_trim_data(struct xfs_da_args *args,
+		struct xfs_dabuf *lbp, xfs_dir2_db_t db);
+extern struct xfs_dir2_leaf_entry *
+xfs_dir2_leaf_find_entry(struct xfs_dir2_leaf *leaf, int index, int compact,
+		int lowstale, int highstale,
+		int *lfloglow, int *lfloghigh);
+extern int xfs_dir2_node_to_leaf(struct xfs_da_state *state);
+
+/* xfs_dir2_node.c */
+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);
+extern int xfs_dir2_leafn_lookup_int(struct xfs_dabuf *bp,
+		struct xfs_da_args *args, int *indexp,
+		struct xfs_da_state *state);
+extern int xfs_dir2_leafn_order(struct xfs_dabuf *leaf1_bp,
+		struct xfs_dabuf *leaf2_bp);
+extern int xfs_dir2_leafn_split(struct xfs_da_state *state,
+	struct xfs_da_state_blk *oldblk, struct xfs_da_state_blk *newblk);
+extern int xfs_dir2_leafn_toosmall(struct xfs_da_state *state, int *action);
+extern void xfs_dir2_leafn_unbalance(struct xfs_da_state *state,
+		struct xfs_da_state_blk *drop_blk,
+		struct xfs_da_state_blk *save_blk);
+extern int xfs_dir2_node_addname(struct xfs_da_args *args);
+extern int xfs_dir2_node_lookup(struct xfs_da_args *args);
+extern int xfs_dir2_node_removename(struct xfs_da_args *args);
+extern int xfs_dir2_node_replace(struct xfs_da_args *args);
+extern int xfs_dir2_node_trim_free(struct xfs_da_args *args, xfs_fileoff_t fo,
+		int *rvalp);
+
+/* xfs_dir2_sf.c */
+extern xfs_ino_t xfs_dir2_sf_get_parent_ino(struct xfs_dir2_sf_hdr *sfp);
+extern xfs_ino_t xfs_dir2_sfe_get_ino(struct xfs_dir2_sf_hdr *sfp,
+		struct xfs_dir2_sf_entry *sfep);
+extern int xfs_dir2_block_sfsize(struct xfs_inode *dp,
+		struct xfs_dir2_data_hdr *block, struct xfs_dir2_sf_hdr *sfhp);
+extern int xfs_dir2_block_to_sf(struct xfs_da_args *args, struct xfs_dabuf *bp,
+		int size, xfs_dir2_sf_hdr_t *sfhp);
+extern int xfs_dir2_sf_addname(struct xfs_da_args *args);
+extern int xfs_dir2_sf_create(struct xfs_da_args *args, xfs_ino_t pino);
+extern int xfs_dir2_sf_getdents(struct xfs_inode *dp, void *dirent,
+		xfs_off_t *offset, filldir_t filldir);
+extern int xfs_dir2_sf_lookup(struct xfs_da_args *args);
+extern int xfs_dir2_sf_removename(struct xfs_da_args *args);
+extern int xfs_dir2_sf_replace(struct xfs_da_args *args);
+
+#endif /* __XFS_DIR2_PRIV_H__ */
diff --git a/fs/xfs/xfs_dir2_sf.c b/fs/xfs/xfs_dir2_sf.c
index b1bae6b1eed..79d05e84e29 100644
--- a/fs/xfs/xfs_dir2_sf.c
+++ b/fs/xfs/xfs_dir2_sf.c
@@ -23,18 +23,16 @@
 #include "xfs_trans.h"
 #include "xfs_sb.h"
 #include "xfs_ag.h"
-#include "xfs_dir2.h"
 #include "xfs_mount.h"
 #include "xfs_da_btree.h"
 #include "xfs_bmap_btree.h"
-#include "xfs_dir2_sf.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_inode_item.h"
 #include "xfs_error.h"
-#include "xfs_dir2_data.h"
-#include "xfs_dir2_leaf.h"
-#include "xfs_dir2_block.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_format.h"
+#include "xfs_dir2_priv.h"
 #include "xfs_trace.h"
 
 /*
@@ -60,6 +58,82 @@ static void xfs_dir2_sf_toino8(xfs_da_args_t *args);
 #endif /* XFS_BIG_INUMS */
 
 /*
+ * Inode numbers in short-form directories can come in two versions,
+ * either 4 bytes or 8 bytes wide.  These helpers deal with the
+ * two forms transparently by looking at the headers i8count field.
+ *
+ * For 64-bit inode number the most significant byte must be zero.
+ */
+static xfs_ino_t
+xfs_dir2_sf_get_ino(
+	struct xfs_dir2_sf_hdr	*hdr,
+	xfs_dir2_inou_t		*from)
+{
+	if (hdr->i8count)
+		return get_unaligned_be64(&from->i8.i) & 0x00ffffffffffffffULL;
+	else
+		return get_unaligned_be32(&from->i4.i);
+}
+
+static void
+xfs_dir2_sf_put_ino(
+	struct xfs_dir2_sf_hdr	*hdr,
+	xfs_dir2_inou_t		*to,
+	xfs_ino_t		ino)
+{
+	ASSERT((ino & 0xff00000000000000ULL) == 0);
+
+	if (hdr->i8count)
+		put_unaligned_be64(ino, &to->i8.i);
+	else
+		put_unaligned_be32(ino, &to->i4.i);
+}
+
+xfs_ino_t
+xfs_dir2_sf_get_parent_ino(
+	struct xfs_dir2_sf_hdr	*hdr)
+{
+	return xfs_dir2_sf_get_ino(hdr, &hdr->parent);
+}
+
+static void
+xfs_dir2_sf_put_parent_ino(
+	struct xfs_dir2_sf_hdr	*hdr,
+	xfs_ino_t		ino)
+{
+	xfs_dir2_sf_put_ino(hdr, &hdr->parent, ino);
+}
+
+/*
+ * In short-form directory entries the inode numbers are stored at variable
+ * offset behind the entry name.  The inode numbers may only be accessed
+ * through the helpers below.
+ */
+static xfs_dir2_inou_t *
+xfs_dir2_sfe_inop(
+	struct xfs_dir2_sf_entry *sfep)
+{
+	return (xfs_dir2_inou_t *)&sfep->name[sfep->namelen];
+}
+
+xfs_ino_t
+xfs_dir2_sfe_get_ino(
+	struct xfs_dir2_sf_hdr	*hdr,
+	struct xfs_dir2_sf_entry *sfep)
+{
+	return xfs_dir2_sf_get_ino(hdr, xfs_dir2_sfe_inop(sfep));
+}
+
+static void
+xfs_dir2_sfe_put_ino(
+	struct xfs_dir2_sf_hdr	*hdr,
+	struct xfs_dir2_sf_entry *sfep,
+	xfs_ino_t		ino)
+{
+	xfs_dir2_sf_put_ino(hdr, xfs_dir2_sfe_inop(sfep), ino);
+}
+
+/*
  * Given a block directory (dp/block), calculate its size as a shortform (sf)
  * directory and a header for the sf directory, if it will fit it the
  * space currently present in the inode.  If it won't fit, the output
@@ -68,7 +142,7 @@ static void xfs_dir2_sf_toino8(xfs_da_args_t *args);
 int						/* size for sf form */
 xfs_dir2_block_sfsize(
 	xfs_inode_t		*dp,		/* incore inode pointer */
-	xfs_dir2_block_t	*block,		/* block directory data */
+	xfs_dir2_data_hdr_t	*hdr,		/* block directory data */
 	xfs_dir2_sf_hdr_t	*sfhp)		/* output: header for sf form */
 {
 	xfs_dir2_dataptr_t	addr;		/* data entry address */
@@ -88,7 +162,7 @@ xfs_dir2_block_sfsize(
 	mp = dp->i_mount;
 
 	count = i8count = namelen = 0;
-	btp = xfs_dir2_block_tail_p(mp, block);
+	btp = xfs_dir2_block_tail_p(mp, hdr);
 	blp = xfs_dir2_block_leaf_p(btp);
 
 	/*
@@ -101,7 +175,7 @@ xfs_dir2_block_sfsize(
 		 * Calculate the pointer to the entry at hand.
 		 */
 		dep = (xfs_dir2_data_entry_t *)
-		      ((char *)block + xfs_dir2_dataptr_to_off(mp, addr));
+		      ((char *)hdr + xfs_dir2_dataptr_to_off(mp, addr));
 		/*
 		 * Detect . and .., so we can special-case them.
 		 * . is not included in sf directories.
@@ -138,7 +212,7 @@ xfs_dir2_block_sfsize(
 	 */
 	sfhp->count = count;
 	sfhp->i8count = i8count;
-	xfs_dir2_sf_put_inumber((xfs_dir2_sf_t *)sfhp, &parent, &sfhp->parent);
+	xfs_dir2_sf_put_parent_ino(sfhp, parent);
 	return size;
 }
 
@@ -153,7 +227,7 @@ xfs_dir2_block_to_sf(
 	int			size,		/* shortform directory size */
 	xfs_dir2_sf_hdr_t	*sfhp)		/* shortform directory hdr */
 {
-	xfs_dir2_block_t	*block;		/* block structure */
+	xfs_dir2_data_hdr_t	*hdr;		/* block header */
 	xfs_dir2_block_tail_t	*btp;		/* block tail pointer */
 	xfs_dir2_data_entry_t	*dep;		/* data entry pointer */
 	xfs_inode_t		*dp;		/* incore directory inode */
@@ -164,8 +238,7 @@ xfs_dir2_block_to_sf(
 	xfs_mount_t		*mp;		/* filesystem mount point */
 	char			*ptr;		/* current data pointer */
 	xfs_dir2_sf_entry_t	*sfep;		/* shortform entry */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
-	xfs_ino_t               temp;
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform directory header */
 
 	trace_xfs_dir2_block_to_sf(args);
 
@@ -176,13 +249,14 @@ xfs_dir2_block_to_sf(
 	 * Make a copy of the block data, so we can shrink the inode
 	 * and add local data.
 	 */
-	block = kmem_alloc(mp->m_dirblksize, KM_SLEEP);
-	memcpy(block, bp->data, mp->m_dirblksize);
+	hdr = kmem_alloc(mp->m_dirblksize, KM_SLEEP);
+	memcpy(hdr, bp->data, mp->m_dirblksize);
 	logflags = XFS_ILOG_CORE;
 	if ((error = xfs_dir2_shrink_inode(args, mp->m_dirdatablk, bp))) {
 		ASSERT(error != ENOSPC);
 		goto out;
 	}
+
 	/*
 	 * The buffer is now unconditionally gone, whether
 	 * xfs_dir2_shrink_inode worked or not.
@@ -198,14 +272,14 @@ xfs_dir2_block_to_sf(
 	/*
 	 * Copy the header into the newly allocate local space.
 	 */
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	memcpy(sfp, sfhp, xfs_dir2_sf_hdr_size(sfhp->i8count));
 	dp->i_d.di_size = size;
 	/*
 	 * Set up to loop over the block's entries.
 	 */
-	btp = xfs_dir2_block_tail_p(mp, block);
-	ptr = (char *)block->u;
+	btp = xfs_dir2_block_tail_p(mp, hdr);
+	ptr = (char *)(hdr + 1);
 	endptr = (char *)xfs_dir2_block_leaf_p(btp);
 	sfep = xfs_dir2_sf_firstentry(sfp);
 	/*
@@ -233,7 +307,7 @@ xfs_dir2_block_to_sf(
 		else if (dep->namelen == 2 &&
 			 dep->name[0] == '.' && dep->name[1] == '.')
 			ASSERT(be64_to_cpu(dep->inumber) ==
-			       xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent));
+			       xfs_dir2_sf_get_parent_ino(sfp));
 		/*
 		 * Normal entry, copy it into shortform.
 		 */
@@ -241,11 +315,11 @@ xfs_dir2_block_to_sf(
 			sfep->namelen = dep->namelen;
 			xfs_dir2_sf_put_offset(sfep,
 				(xfs_dir2_data_aoff_t)
-				((char *)dep - (char *)block));
+				((char *)dep - (char *)hdr));
 			memcpy(sfep->name, dep->name, dep->namelen);
-			temp = be64_to_cpu(dep->inumber);
-			xfs_dir2_sf_put_inumber(sfp, &temp,
-				xfs_dir2_sf_inumberp(sfep));
+			xfs_dir2_sfe_put_ino(sfp, sfep,
+					     be64_to_cpu(dep->inumber));
+
 			sfep = xfs_dir2_sf_nextentry(sfp, sfep);
 		}
 		ptr += xfs_dir2_data_entsize(dep->namelen);
@@ -254,7 +328,7 @@ xfs_dir2_block_to_sf(
 	xfs_dir2_sf_check(args);
 out:
 	xfs_trans_log_inode(args->trans, dp, logflags);
-	kmem_free(block);
+	kmem_free(hdr);
 	return error;
 }
 
@@ -277,7 +351,7 @@ xfs_dir2_sf_addname(
 	xfs_dir2_data_aoff_t	offset = 0;	/* offset for new entry */
 	int			old_isize;	/* di_size before adding name */
 	int			pick;		/* which algorithm to use */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 	xfs_dir2_sf_entry_t	*sfep = NULL;	/* shortform entry */
 
 	trace_xfs_dir2_sf_addname(args);
@@ -294,19 +368,19 @@ xfs_dir2_sf_addname(
 	}
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->i8count));
 	/*
 	 * Compute entry (and change in) size.
 	 */
-	add_entsize = xfs_dir2_sf_entsize_byname(sfp, args->namelen);
+	add_entsize = xfs_dir2_sf_entsize(sfp, args->namelen);
 	incr_isize = add_entsize;
 	objchange = 0;
 #if XFS_BIG_INUMS
 	/*
 	 * Do we have to change to 8 byte inodes?
 	 */
-	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->hdr.i8count == 0) {
+	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->i8count == 0) {
 		/*
 		 * Yes, adjust the entry size and the total size.
 		 */
@@ -314,7 +388,7 @@ xfs_dir2_sf_addname(
 			(uint)sizeof(xfs_dir2_ino8_t) -
 			(uint)sizeof(xfs_dir2_ino4_t);
 		incr_isize +=
-			(sfp->hdr.count + 2) *
+			(sfp->count + 2) *
 			((uint)sizeof(xfs_dir2_ino8_t) -
 			 (uint)sizeof(xfs_dir2_ino4_t));
 		objchange = 1;
@@ -384,21 +458,21 @@ xfs_dir2_sf_addname_easy(
 {
 	int			byteoff;	/* byte offset in sf dir */
 	xfs_inode_t		*dp;		/* incore directory inode */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 
 	dp = args->dp;
 
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	byteoff = (int)((char *)sfep - (char *)sfp);
 	/*
 	 * Grow the in-inode space.
 	 */
-	xfs_idata_realloc(dp, xfs_dir2_sf_entsize_byname(sfp, args->namelen),
+	xfs_idata_realloc(dp, xfs_dir2_sf_entsize(sfp, args->namelen),
 		XFS_DATA_FORK);
 	/*
 	 * Need to set up again due to realloc of the inode data.
 	 */
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	sfep = (xfs_dir2_sf_entry_t *)((char *)sfp + byteoff);
 	/*
 	 * Fill in the new entry.
@@ -406,15 +480,14 @@ xfs_dir2_sf_addname_easy(
 	sfep->namelen = args->namelen;
 	xfs_dir2_sf_put_offset(sfep, offset);
 	memcpy(sfep->name, args->name, sfep->namelen);
-	xfs_dir2_sf_put_inumber(sfp, &args->inumber,
-		xfs_dir2_sf_inumberp(sfep));
+	xfs_dir2_sfe_put_ino(sfp, sfep, args->inumber);
 	/*
 	 * Update the header and inode.
 	 */
-	sfp->hdr.count++;
+	sfp->count++;
 #if XFS_BIG_INUMS
 	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM)
-		sfp->hdr.i8count++;
+		sfp->i8count++;
 #endif
 	dp->i_d.di_size = new_isize;
 	xfs_dir2_sf_check(args);
@@ -444,19 +517,19 @@ xfs_dir2_sf_addname_hard(
 	xfs_dir2_data_aoff_t	offset;		/* current offset value */
 	int			old_isize;	/* previous di_size */
 	xfs_dir2_sf_entry_t	*oldsfep;	/* entry in original dir */
-	xfs_dir2_sf_t		*oldsfp;	/* original shortform dir */
+	xfs_dir2_sf_hdr_t	*oldsfp;	/* original shortform dir */
 	xfs_dir2_sf_entry_t	*sfep;		/* entry in new dir */
-	xfs_dir2_sf_t		*sfp;		/* new shortform dir */
+	xfs_dir2_sf_hdr_t	*sfp;		/* new shortform dir */
 
 	/*
 	 * Copy the old directory to the stack buffer.
 	 */
 	dp = args->dp;
 
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	old_isize = (int)dp->i_d.di_size;
 	buf = kmem_alloc(old_isize, KM_SLEEP);
-	oldsfp = (xfs_dir2_sf_t *)buf;
+	oldsfp = (xfs_dir2_sf_hdr_t *)buf;
 	memcpy(oldsfp, sfp, old_isize);
 	/*
 	 * Loop over the old directory finding the place we're going
@@ -485,7 +558,7 @@ xfs_dir2_sf_addname_hard(
 	/*
 	 * Reset the pointer since the buffer was reallocated.
 	 */
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	/*
 	 * Copy the first part of the directory, including the header.
 	 */
@@ -498,12 +571,11 @@ xfs_dir2_sf_addname_hard(
 	sfep->namelen = args->namelen;
 	xfs_dir2_sf_put_offset(sfep, offset);
 	memcpy(sfep->name, args->name, sfep->namelen);
-	xfs_dir2_sf_put_inumber(sfp, &args->inumber,
-		xfs_dir2_sf_inumberp(sfep));
-	sfp->hdr.count++;
+	xfs_dir2_sfe_put_ino(sfp, sfep, args->inumber);
+	sfp->count++;
 #if XFS_BIG_INUMS
 	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && !objchange)
-		sfp->hdr.i8count++;
+		sfp->i8count++;
 #endif
 	/*
 	 * If there's more left to copy, do that.
@@ -537,14 +609,14 @@ xfs_dir2_sf_addname_pick(
 	xfs_mount_t		*mp;		/* filesystem mount point */
 	xfs_dir2_data_aoff_t	offset;		/* data block offset */
 	xfs_dir2_sf_entry_t	*sfep;		/* shortform entry */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 	int			size;		/* entry's data size */
 	int			used;		/* data bytes used */
 
 	dp = args->dp;
 	mp = dp->i_mount;
 
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	size = xfs_dir2_data_entsize(args->namelen);
 	offset = XFS_DIR2_DATA_FIRST_OFFSET;
 	sfep = xfs_dir2_sf_firstentry(sfp);
@@ -554,7 +626,7 @@ xfs_dir2_sf_addname_pick(
 	 * Keep track of data offset and whether we've seen a place
 	 * to insert the new entry.
 	 */
-	for (i = 0; i < sfp->hdr.count; i++) {
+	for (i = 0; i < sfp->count; i++) {
 		if (!holefit)
 			holefit = offset + size <= xfs_dir2_sf_get_offset(sfep);
 		offset = xfs_dir2_sf_get_offset(sfep) +
@@ -566,7 +638,7 @@ xfs_dir2_sf_addname_pick(
 	 * was a data block (block form directory).
 	 */
 	used = offset +
-	       (sfp->hdr.count + 3) * (uint)sizeof(xfs_dir2_leaf_entry_t) +
+	       (sfp->count + 3) * (uint)sizeof(xfs_dir2_leaf_entry_t) +
 	       (uint)sizeof(xfs_dir2_block_tail_t);
 	/*
 	 * If it won't fit in a block form then we can't insert it,
@@ -612,30 +684,30 @@ xfs_dir2_sf_check(
 	xfs_ino_t		ino;		/* entry inode number */
 	int			offset;		/* data offset */
 	xfs_dir2_sf_entry_t	*sfep;		/* shortform dir entry */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 
 	dp = args->dp;
 
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	offset = XFS_DIR2_DATA_FIRST_OFFSET;
-	ino = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
+	ino = xfs_dir2_sf_get_parent_ino(sfp);
 	i8count = ino > XFS_DIR2_MAX_SHORT_INUM;
 
 	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
-	     i < sfp->hdr.count;
+	     i < sfp->count;
 	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 		ASSERT(xfs_dir2_sf_get_offset(sfep) >= offset);
-		ino = xfs_dir2_sf_get_inumber(sfp, xfs_dir2_sf_inumberp(sfep));
+		ino = xfs_dir2_sfe_get_ino(sfp, sfep);
 		i8count += ino > XFS_DIR2_MAX_SHORT_INUM;
 		offset =
 			xfs_dir2_sf_get_offset(sfep) +
 			xfs_dir2_data_entsize(sfep->namelen);
 	}
-	ASSERT(i8count == sfp->hdr.i8count);
+	ASSERT(i8count == sfp->i8count);
 	ASSERT(XFS_BIG_INUMS || i8count == 0);
 	ASSERT((char *)sfep - (char *)sfp == dp->i_d.di_size);
 	ASSERT(offset +
-	       (sfp->hdr.count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t) +
+	       (sfp->count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t) +
 	       (uint)sizeof(xfs_dir2_block_tail_t) <=
 	       dp->i_mount->m_dirblksize);
 }
@@ -651,7 +723,7 @@ xfs_dir2_sf_create(
 {
 	xfs_inode_t	*dp;		/* incore directory inode */
 	int		i8count;	/* parent inode is an 8-byte number */
-	xfs_dir2_sf_t	*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t *sfp;		/* shortform structure */
 	int		size;		/* directory size */
 
 	trace_xfs_dir2_sf_create(args);
@@ -681,13 +753,13 @@ xfs_dir2_sf_create(
 	/*
 	 * Fill in the header,
 	 */
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	sfp->hdr.i8count = i8count;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	sfp->i8count = i8count;
 	/*
 	 * Now can put in the inode number, since i8count is set.
 	 */
-	xfs_dir2_sf_put_inumber(sfp, &pino, &sfp->hdr.parent);
-	sfp->hdr.count = 0;
+	xfs_dir2_sf_put_parent_ino(sfp, pino);
+	sfp->count = 0;
 	dp->i_d.di_size = size;
 	xfs_dir2_sf_check(args);
 	xfs_trans_log_inode(args->trans, dp, XFS_ILOG_CORE | XFS_ILOG_DDATA);
@@ -705,7 +777,7 @@ xfs_dir2_sf_getdents(
 	xfs_mount_t		*mp;		/* filesystem mount point */
 	xfs_dir2_dataptr_t	off;		/* current entry's offset */
 	xfs_dir2_sf_entry_t	*sfep;		/* shortform directory entry */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 	xfs_dir2_dataptr_t	dot_offset;
 	xfs_dir2_dataptr_t	dotdot_offset;
 	xfs_ino_t		ino;
@@ -724,9 +796,9 @@ xfs_dir2_sf_getdents(
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
 
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 
-	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->i8count));
 
 	/*
 	 * If the block number in the offset is out of range, we're done.
@@ -759,7 +831,7 @@ xfs_dir2_sf_getdents(
 	 * Put .. entry unless we're starting past it.
 	 */
 	if (*offset <= dotdot_offset) {
-		ino = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
+		ino = xfs_dir2_sf_get_parent_ino(sfp);
 		if (filldir(dirent, "..", 2, dotdot_offset & 0x7fffffff, ino, DT_DIR)) {
 			*offset = dotdot_offset & 0x7fffffff;
 			return 0;
@@ -770,7 +842,7 @@ xfs_dir2_sf_getdents(
 	 * Loop while there are more entries and put'ing works.
 	 */
 	sfep = xfs_dir2_sf_firstentry(sfp);
-	for (i = 0; i < sfp->hdr.count; i++) {
+	for (i = 0; i < sfp->count; i++) {
 		off = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
 				xfs_dir2_sf_get_offset(sfep));
 
@@ -779,7 +851,7 @@ xfs_dir2_sf_getdents(
 			continue;
 		}
 
-		ino = xfs_dir2_sf_get_inumber(sfp, xfs_dir2_sf_inumberp(sfep));
+		ino = xfs_dir2_sfe_get_ino(sfp, sfep);
 		if (filldir(dirent, (char *)sfep->name, sfep->namelen,
 			    off & 0x7fffffff, ino, DT_UNKNOWN)) {
 			*offset = off & 0x7fffffff;
@@ -805,7 +877,7 @@ xfs_dir2_sf_lookup(
 	int			i;		/* entry index */
 	int			error;
 	xfs_dir2_sf_entry_t	*sfep;		/* shortform directory entry */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 	enum xfs_dacmp		cmp;		/* comparison result */
 	xfs_dir2_sf_entry_t	*ci_sfep;	/* case-insens. entry */
 
@@ -824,8 +896,8 @@ xfs_dir2_sf_lookup(
 	}
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->i8count));
 	/*
 	 * Special case for .
 	 */
@@ -839,7 +911,7 @@ xfs_dir2_sf_lookup(
 	 */
 	if (args->namelen == 2 &&
 	    args->name[0] == '.' && args->name[1] == '.') {
-		args->inumber = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
+		args->inumber = xfs_dir2_sf_get_parent_ino(sfp);
 		args->cmpresult = XFS_CMP_EXACT;
 		return XFS_ERROR(EEXIST);
 	}
@@ -847,7 +919,7 @@ xfs_dir2_sf_lookup(
 	 * Loop over all the entries trying to match ours.
 	 */
 	ci_sfep = NULL;
-	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp); i < sfp->hdr.count;
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp); i < sfp->count;
 				i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 		/*
 		 * Compare name and if it's an exact match, return the inode
@@ -858,8 +930,7 @@ xfs_dir2_sf_lookup(
 								sfep->namelen);
 		if (cmp != XFS_CMP_DIFFERENT && cmp != args->cmpresult) {
 			args->cmpresult = cmp;
-			args->inumber = xfs_dir2_sf_get_inumber(sfp,
-						xfs_dir2_sf_inumberp(sfep));
+			args->inumber = xfs_dir2_sfe_get_ino(sfp, sfep);
 			if (cmp == XFS_CMP_EXACT)
 				return XFS_ERROR(EEXIST);
 			ci_sfep = sfep;
@@ -891,7 +962,7 @@ xfs_dir2_sf_removename(
 	int			newsize;	/* new inode size */
 	int			oldsize;	/* old inode size */
 	xfs_dir2_sf_entry_t	*sfep;		/* shortform directory entry */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 
 	trace_xfs_dir2_sf_removename(args);
 
@@ -908,32 +979,31 @@ xfs_dir2_sf_removename(
 	}
 	ASSERT(dp->i_df.if_bytes == oldsize);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(oldsize >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	ASSERT(oldsize >= xfs_dir2_sf_hdr_size(sfp->i8count));
 	/*
 	 * Loop over the old directory entries.
 	 * Find the one we're deleting.
 	 */
-	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp); i < sfp->hdr.count;
+	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp); i < sfp->count;
 				i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 		if (xfs_da_compname(args, sfep->name, sfep->namelen) ==
 								XFS_CMP_EXACT) {
-			ASSERT(xfs_dir2_sf_get_inumber(sfp,
-						xfs_dir2_sf_inumberp(sfep)) ==
-								args->inumber);
+			ASSERT(xfs_dir2_sfe_get_ino(sfp, sfep) ==
+			       args->inumber);
 			break;
 		}
 	}
 	/*
 	 * Didn't find it.
 	 */
-	if (i == sfp->hdr.count)
+	if (i == sfp->count)
 		return XFS_ERROR(ENOENT);
 	/*
 	 * Calculate sizes.
 	 */
 	byteoff = (int)((char *)sfep - (char *)sfp);
-	entsize = xfs_dir2_sf_entsize_byname(sfp, args->namelen);
+	entsize = xfs_dir2_sf_entsize(sfp, args->namelen);
 	newsize = oldsize - entsize;
 	/*
 	 * Copy the part if any after the removed entry, sliding it down.
@@ -944,22 +1014,22 @@ xfs_dir2_sf_removename(
 	/*
 	 * Fix up the header and file size.
 	 */
-	sfp->hdr.count--;
+	sfp->count--;
 	dp->i_d.di_size = newsize;
 	/*
 	 * Reallocate, making it smaller.
 	 */
 	xfs_idata_realloc(dp, newsize - oldsize, XFS_DATA_FORK);
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 #if XFS_BIG_INUMS
 	/*
 	 * Are we changing inode number size?
 	 */
 	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM) {
-		if (sfp->hdr.i8count == 1)
+		if (sfp->i8count == 1)
 			xfs_dir2_sf_toino4(args);
 		else
-			sfp->hdr.i8count--;
+			sfp->i8count--;
 	}
 #endif
 	xfs_dir2_sf_check(args);
@@ -983,7 +1053,7 @@ xfs_dir2_sf_replace(
 	int			i8elevated;	/* sf_toino8 set i8count=1 */
 #endif
 	xfs_dir2_sf_entry_t	*sfep;		/* shortform directory entry */
-	xfs_dir2_sf_t		*sfp;		/* shortform structure */
+	xfs_dir2_sf_hdr_t	*sfp;		/* shortform structure */
 
 	trace_xfs_dir2_sf_replace(args);
 
@@ -999,19 +1069,19 @@ xfs_dir2_sf_replace(
 	}
 	ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
 	ASSERT(dp->i_df.if_u1.if_data != NULL);
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->i8count));
 #if XFS_BIG_INUMS
 	/*
 	 * New inode number is large, and need to convert to 8-byte inodes.
 	 */
-	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->hdr.i8count == 0) {
+	if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->i8count == 0) {
 		int	error;			/* error return value */
 		int	newsize;		/* new inode size */
 
 		newsize =
 			dp->i_df.if_bytes +
-			(sfp->hdr.count + 1) *
+			(sfp->count + 1) *
 			((uint)sizeof(xfs_dir2_ino8_t) -
 			 (uint)sizeof(xfs_dir2_ino4_t));
 		/*
@@ -1029,7 +1099,7 @@ xfs_dir2_sf_replace(
 		 */
 		xfs_dir2_sf_toino8(args);
 		i8elevated = 1;
-		sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+		sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	} else
 		i8elevated = 0;
 #endif
@@ -1040,34 +1110,32 @@ xfs_dir2_sf_replace(
 	if (args->namelen == 2 &&
 	    args->name[0] == '.' && args->name[1] == '.') {
 #if XFS_BIG_INUMS || defined(DEBUG)
-		ino = xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent);
+		ino = xfs_dir2_sf_get_parent_ino(sfp);
 		ASSERT(args->inumber != ino);
 #endif
-		xfs_dir2_sf_put_inumber(sfp, &args->inumber, &sfp->hdr.parent);
+		xfs_dir2_sf_put_parent_ino(sfp, args->inumber);
 	}
 	/*
 	 * Normal entry, look for the name.
 	 */
 	else {
 		for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
-				i < sfp->hdr.count;
+				i < sfp->count;
 				i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
 			if (xfs_da_compname(args, sfep->name, sfep->namelen) ==
 								XFS_CMP_EXACT) {
 #if XFS_BIG_INUMS || defined(DEBUG)
-				ino = xfs_dir2_sf_get_inumber(sfp,
-					xfs_dir2_sf_inumberp(sfep));
+				ino = xfs_dir2_sfe_get_ino(sfp, sfep);
 				ASSERT(args->inumber != ino);
 #endif
-				xfs_dir2_sf_put_inumber(sfp, &args->inumber,
-					xfs_dir2_sf_inumberp(sfep));
+				xfs_dir2_sfe_put_ino(sfp, sfep, args->inumber);
 				break;
 			}
 		}
 		/*
 		 * Didn't find it.
 		 */
-		if (i == sfp->hdr.count) {
+		if (i == sfp->count) {
 			ASSERT(args->op_flags & XFS_DA_OP_OKNOENT);
 #if XFS_BIG_INUMS
 			if (i8elevated)
@@ -1085,10 +1153,10 @@ xfs_dir2_sf_replace(
 		/*
 		 * And the old count was one, so need to convert to small.
 		 */
-		if (sfp->hdr.i8count == 1)
+		if (sfp->i8count == 1)
 			xfs_dir2_sf_toino4(args);
 		else
-			sfp->hdr.i8count--;
+			sfp->i8count--;
 	}
 	/*
 	 * See if the old number was small, the new number is large.
@@ -1099,9 +1167,9 @@ xfs_dir2_sf_replace(
 		 * add to the i8count unless we just converted to 8-byte
 		 * inodes (which does an implied i8count = 1)
 		 */
-		ASSERT(sfp->hdr.i8count != 0);
+		ASSERT(sfp->i8count != 0);
 		if (!i8elevated)
-			sfp->hdr.i8count++;
+			sfp->i8count++;
 	}
 #endif
 	xfs_dir2_sf_check(args);
@@ -1121,13 +1189,12 @@ xfs_dir2_sf_toino4(
 	char			*buf;		/* old dir's buffer */
 	xfs_inode_t		*dp;		/* incore directory inode */
 	int			i;		/* entry index */
-	xfs_ino_t		ino;		/* entry inode number */
 	int			newsize;	/* new inode size */
 	xfs_dir2_sf_entry_t	*oldsfep;	/* old sf entry */
-	xfs_dir2_sf_t		*oldsfp;	/* old sf directory */
+	xfs_dir2_sf_hdr_t	*oldsfp;	/* old sf directory */
 	int			oldsize;	/* old inode size */
 	xfs_dir2_sf_entry_t	*sfep;		/* new sf entry */
-	xfs_dir2_sf_t		*sfp;		/* new sf directory */
+	xfs_dir2_sf_hdr_t	*sfp;		/* new sf directory */
 
 	trace_xfs_dir2_sf_toino4(args);
 
@@ -1140,44 +1207,42 @@ xfs_dir2_sf_toino4(
 	 */
 	oldsize = dp->i_df.if_bytes;
 	buf = kmem_alloc(oldsize, KM_SLEEP);
-	oldsfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(oldsfp->hdr.i8count == 1);
+	oldsfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	ASSERT(oldsfp->i8count == 1);
 	memcpy(buf, oldsfp, oldsize);
 	/*
 	 * Compute the new inode size.
 	 */
 	newsize =
 		oldsize -
-		(oldsfp->hdr.count + 1) *
+		(oldsfp->count + 1) *
 		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t));
 	xfs_idata_realloc(dp, -oldsize, XFS_DATA_FORK);
 	xfs_idata_realloc(dp, newsize, XFS_DATA_FORK);
 	/*
 	 * Reset our pointers, the data has moved.
 	 */
-	oldsfp = (xfs_dir2_sf_t *)buf;
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	oldsfp = (xfs_dir2_sf_hdr_t *)buf;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	/*
 	 * Fill in the new header.
 	 */
-	sfp->hdr.count = oldsfp->hdr.count;
-	sfp->hdr.i8count = 0;
-	ino = xfs_dir2_sf_get_inumber(oldsfp, &oldsfp->hdr.parent);
-	xfs_dir2_sf_put_inumber(sfp, &ino, &sfp->hdr.parent);
+	sfp->count = oldsfp->count;
+	sfp->i8count = 0;
+	xfs_dir2_sf_put_parent_ino(sfp, xfs_dir2_sf_get_parent_ino(oldsfp));
 	/*
 	 * Copy the entries field by field.
 	 */
 	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp),
 		    oldsfep = xfs_dir2_sf_firstentry(oldsfp);
-	     i < sfp->hdr.count;
+	     i < sfp->count;
 	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep),
 		  oldsfep = xfs_dir2_sf_nextentry(oldsfp, oldsfep)) {
 		sfep->namelen = oldsfep->namelen;
 		sfep->offset = oldsfep->offset;
 		memcpy(sfep->name, oldsfep->name, sfep->namelen);
-		ino = xfs_dir2_sf_get_inumber(oldsfp,
-			xfs_dir2_sf_inumberp(oldsfep));
-		xfs_dir2_sf_put_inumber(sfp, &ino, xfs_dir2_sf_inumberp(sfep));
+		xfs_dir2_sfe_put_ino(sfp, sfep,
+			xfs_dir2_sfe_get_ino(oldsfp, oldsfep));
 	}
 	/*
 	 * Clean up the inode.
@@ -1199,13 +1264,12 @@ xfs_dir2_sf_toino8(
 	char			*buf;		/* old dir's buffer */
 	xfs_inode_t		*dp;		/* incore directory inode */
 	int			i;		/* entry index */
-	xfs_ino_t		ino;		/* entry inode number */
 	int			newsize;	/* new inode size */
 	xfs_dir2_sf_entry_t	*oldsfep;	/* old sf entry */
-	xfs_dir2_sf_t		*oldsfp;	/* old sf directory */
+	xfs_dir2_sf_hdr_t	*oldsfp;	/* old sf directory */
 	int			oldsize;	/* old inode size */
 	xfs_dir2_sf_entry_t	*sfep;		/* new sf entry */
-	xfs_dir2_sf_t		*sfp;		/* new sf directory */
+	xfs_dir2_sf_hdr_t	*sfp;		/* new sf directory */
 
 	trace_xfs_dir2_sf_toino8(args);
 
@@ -1218,44 +1282,42 @@ xfs_dir2_sf_toino8(
 	 */
 	oldsize = dp->i_df.if_bytes;
 	buf = kmem_alloc(oldsize, KM_SLEEP);
-	oldsfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
-	ASSERT(oldsfp->hdr.i8count == 0);
+	oldsfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
+	ASSERT(oldsfp->i8count == 0);
 	memcpy(buf, oldsfp, oldsize);
 	/*
 	 * Compute the new inode size.
 	 */
 	newsize =
 		oldsize +
-		(oldsfp->hdr.count + 1) *
+		(oldsfp->count + 1) *
 		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t));
 	xfs_idata_realloc(dp, -oldsize, XFS_DATA_FORK);
 	xfs_idata_realloc(dp, newsize, XFS_DATA_FORK);
 	/*
 	 * Reset our pointers, the data has moved.
 	 */
-	oldsfp = (xfs_dir2_sf_t *)buf;
-	sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
+	oldsfp = (xfs_dir2_sf_hdr_t *)buf;
+	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	/*
 	 * Fill in the new header.
 	 */
-	sfp->hdr.count = oldsfp->hdr.count;
-	sfp->hdr.i8count = 1;
-	ino = xfs_dir2_sf_get_inumber(oldsfp, &oldsfp->hdr.parent);
-	xfs_dir2_sf_put_inumber(sfp, &ino, &sfp->hdr.parent);
+	sfp->count = oldsfp->count;
+	sfp->i8count = 1;
+	xfs_dir2_sf_put_parent_ino(sfp, xfs_dir2_sf_get_parent_ino(oldsfp));
 	/*
 	 * Copy the entries field by field.
 	 */
 	for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp),
 		    oldsfep = xfs_dir2_sf_firstentry(oldsfp);
-	     i < sfp->hdr.count;
+	     i < sfp->count;
 	     i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep),
 		  oldsfep = xfs_dir2_sf_nextentry(oldsfp, oldsfep)) {
 		sfep->namelen = oldsfep->namelen;
 		sfep->offset = oldsfep->offset;
 		memcpy(sfep->name, oldsfep->name, sfep->namelen);
-		ino = xfs_dir2_sf_get_inumber(oldsfp,
-			xfs_dir2_sf_inumberp(oldsfep));
-		xfs_dir2_sf_put_inumber(sfp, &ino, xfs_dir2_sf_inumberp(sfep));
+		xfs_dir2_sfe_put_ino(sfp, sfep,
+			xfs_dir2_sfe_get_ino(oldsfp, oldsfep));
 	}
 	/*
 	 * Clean up the inode.
diff --git a/fs/xfs/xfs_dir2_sf.h b/fs/xfs/xfs_dir2_sf.h
deleted file mode 100644
index 6ac44b550d3..00000000000
--- a/fs/xfs/xfs_dir2_sf.h
+++ /dev/null
@@ -1,171 +0,0 @@
-/*
- * Copyright (c) 2000-2001,2005 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_DIR2_SF_H__
-#define	__XFS_DIR2_SF_H__
-
-/*
- * Directory layout when stored internal to an inode.
- *
- * Small directories are packed as tightly as possible so as to
- * fit into the literal area of the inode.
- */
-
-struct uio;
-struct xfs_dabuf;
-struct xfs_da_args;
-struct xfs_dir2_block;
-struct xfs_inode;
-struct xfs_mount;
-struct xfs_trans;
-
-/*
- * Inode number stored as 8 8-bit values.
- */
-typedef	struct { __uint8_t i[8]; } xfs_dir2_ino8_t;
-
-/*
- * Inode number stored as 4 8-bit values.
- * Works a lot of the time, when all the inode numbers in a directory
- * fit in 32 bits.
- */
-typedef struct { __uint8_t i[4]; } xfs_dir2_ino4_t;
-
-typedef union {
-	xfs_dir2_ino8_t	i8;
-	xfs_dir2_ino4_t	i4;
-} xfs_dir2_inou_t;
-#define	XFS_DIR2_MAX_SHORT_INUM	((xfs_ino_t)0xffffffffULL)
-
-/*
- * Normalized offset (in a data block) of the entry, really xfs_dir2_data_off_t.
- * Only need 16 bits, this is the byte offset into the single block form.
- */
-typedef struct { __uint8_t i[2]; } __arch_pack xfs_dir2_sf_off_t;
-
-/*
- * The parent directory has a dedicated field, and the self-pointer must
- * be calculated on the fly.
- *
- * Entries are packed toward the top as tightly as possible.  The header
- * and the elements must be memcpy'd out into a work area to get correct
- * alignment for the inode number fields.
- */
-typedef struct xfs_dir2_sf_hdr {
-	__uint8_t		count;		/* count of entries */
-	__uint8_t		i8count;	/* count of 8-byte inode #s */
-	xfs_dir2_inou_t		parent;		/* parent dir inode number */
-} __arch_pack xfs_dir2_sf_hdr_t;
-
-typedef struct xfs_dir2_sf_entry {
-	__uint8_t		namelen;	/* actual name length */
-	xfs_dir2_sf_off_t	offset;		/* saved offset */
-	__uint8_t		name[1];	/* name, variable size */
-	xfs_dir2_inou_t		inumber;	/* inode number, var. offset */
-} __arch_pack xfs_dir2_sf_entry_t; 
-
-typedef struct xfs_dir2_sf {
-	xfs_dir2_sf_hdr_t	hdr;		/* shortform header */
-	xfs_dir2_sf_entry_t	list[1];	/* shortform entries */
-} xfs_dir2_sf_t;
-
-static inline int xfs_dir2_sf_hdr_size(int i8count)
-{
-	return ((uint)sizeof(xfs_dir2_sf_hdr_t) - \
-		((i8count) == 0) * \
-		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t)));
-}
-
-static inline xfs_dir2_inou_t *xfs_dir2_sf_inumberp(xfs_dir2_sf_entry_t *sfep)
-{
-	return (xfs_dir2_inou_t *)&(sfep)->name[(sfep)->namelen];
-}
-
-static inline xfs_intino_t
-xfs_dir2_sf_get_inumber(xfs_dir2_sf_t *sfp, xfs_dir2_inou_t *from)
-{
-	return ((sfp)->hdr.i8count == 0 ? \
-		(xfs_intino_t)XFS_GET_DIR_INO4((from)->i4) : \
-		(xfs_intino_t)XFS_GET_DIR_INO8((from)->i8));
-}
-
-static inline void xfs_dir2_sf_put_inumber(xfs_dir2_sf_t *sfp, xfs_ino_t *from,
-						xfs_dir2_inou_t *to)
-{
-	if ((sfp)->hdr.i8count == 0)
-		XFS_PUT_DIR_INO4(*(from), (to)->i4);
-	else
-		XFS_PUT_DIR_INO8(*(from), (to)->i8);
-}
-
-static inline xfs_dir2_data_aoff_t
-xfs_dir2_sf_get_offset(xfs_dir2_sf_entry_t *sfep)
-{
-	return INT_GET_UNALIGNED_16_BE(&(sfep)->offset.i);
-}
-
-static inline void
-xfs_dir2_sf_put_offset(xfs_dir2_sf_entry_t *sfep, xfs_dir2_data_aoff_t off)
-{
-	INT_SET_UNALIGNED_16_BE(&(sfep)->offset.i, off);
-}
-
-static inline int xfs_dir2_sf_entsize_byname(xfs_dir2_sf_t *sfp, int len)
-{
-	return ((uint)sizeof(xfs_dir2_sf_entry_t) - 1 + (len) - \
-		((sfp)->hdr.i8count == 0) * \
-		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t)));
-}
-
-static inline int
-xfs_dir2_sf_entsize_byentry(xfs_dir2_sf_t *sfp, xfs_dir2_sf_entry_t *sfep)
-{
-	return ((uint)sizeof(xfs_dir2_sf_entry_t) - 1 + (sfep)->namelen - \
-		((sfp)->hdr.i8count == 0) * \
-		((uint)sizeof(xfs_dir2_ino8_t) - (uint)sizeof(xfs_dir2_ino4_t)));
-}
-
-static inline xfs_dir2_sf_entry_t *xfs_dir2_sf_firstentry(xfs_dir2_sf_t *sfp)
-{
-	return ((xfs_dir2_sf_entry_t *) \
-		((char *)(sfp) + xfs_dir2_sf_hdr_size(sfp->hdr.i8count)));
-}
-
-static inline xfs_dir2_sf_entry_t *
-xfs_dir2_sf_nextentry(xfs_dir2_sf_t *sfp, xfs_dir2_sf_entry_t *sfep)
-{
-	return ((xfs_dir2_sf_entry_t *) \
-		((char *)(sfep) + xfs_dir2_sf_entsize_byentry(sfp,sfep)));
-}
-
-/*
- * Functions.
- */
-extern int xfs_dir2_block_sfsize(struct xfs_inode *dp,
-				 struct xfs_dir2_block *block,
-				 xfs_dir2_sf_hdr_t *sfhp);
-extern int xfs_dir2_block_to_sf(struct xfs_da_args *args, struct xfs_dabuf *bp,
-				int size, xfs_dir2_sf_hdr_t *sfhp);
-extern int xfs_dir2_sf_addname(struct xfs_da_args *args);
-extern int xfs_dir2_sf_create(struct xfs_da_args *args, xfs_ino_t pino);
-extern int xfs_dir2_sf_getdents(struct xfs_inode *dp, void *dirent,
-				xfs_off_t *offset, filldir_t filldir);
-extern int xfs_dir2_sf_lookup(struct xfs_da_args *args);
-extern int xfs_dir2_sf_removename(struct xfs_da_args *args);
-extern int xfs_dir2_sf_replace(struct xfs_da_args *args);
-
-#endif	/* __XFS_DIR2_SF_H__ */
diff --git a/fs/xfs/xfs_fs.h b/fs/xfs/xfs_fs.h
index 8f6fc1a9638..c13fed8c394 100644
--- a/fs/xfs/xfs_fs.h
+++ b/fs/xfs/xfs_fs.h
@@ -249,6 +249,11 @@ typedef struct xfs_fsop_resblks {
 #define XFS_MAX_LOG_BYTES \
 	((2 * 1024 * 1024 * 1024ULL) - XFS_MIN_LOG_BYTES)
 
+/* Used for sanity checks on superblock */
+#define XFS_MAX_DBLOCKS(s) ((xfs_drfsbno_t)(s)->sb_agcount * (s)->sb_agblocks)
+#define XFS_MIN_DBLOCKS(s) ((xfs_drfsbno_t)((s)->sb_agcount - 1) *	\
+			 (s)->sb_agblocks + XFS_MIN_AG_BLOCKS)
+
 /*
  * Structures for XFS_IOC_FSGROWFSDATA, XFS_IOC_FSGROWFSLOG & XFS_IOC_FSGROWFSRT
  */
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c
index 84ebeec1664..dd5628bd8d0 100644
--- a/fs/xfs/xfs_ialloc.c
+++ b/fs/xfs/xfs_ialloc.c
@@ -683,7 +683,7 @@ xfs_dialloc(
 			return 0;
 		}
 		agi = XFS_BUF_TO_AGI(agbp);
-		ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
+		ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
 	} else {
 		/*
 		 * Continue where we left off before.  In this case, we
@@ -691,7 +691,7 @@ xfs_dialloc(
 		 */
 		agbp = *IO_agbp;
 		agi = XFS_BUF_TO_AGI(agbp);
-		ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
+		ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
 		ASSERT(be32_to_cpu(agi->agi_freecount) > 0);
 	}
 	mp = tp->t_mountp;
@@ -775,7 +775,7 @@ nextag:
 		if (error)
 			goto nextag;
 		agi = XFS_BUF_TO_AGI(agbp);
-		ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
+		ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
 	}
 	/*
 	 * Here with an allocation group that has a free inode.
@@ -944,7 +944,7 @@ nextag:
 	 * See if the most recently allocated block has any free.
 	 */
 newino:
-	if (be32_to_cpu(agi->agi_newino) != NULLAGINO) {
+	if (agi->agi_newino != cpu_to_be32(NULLAGINO)) {
 		error = xfs_inobt_lookup(cur, be32_to_cpu(agi->agi_newino),
 					 XFS_LOOKUP_EQ, &i);
 		if (error)
@@ -1085,7 +1085,7 @@ xfs_difree(
 		return error;
 	}
 	agi = XFS_BUF_TO_AGI(agbp);
-	ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
+	ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
 	ASSERT(agbno < be32_to_cpu(agi->agi_length));
 	/*
 	 * Initialize the cursor.
@@ -1438,7 +1438,7 @@ xfs_ialloc_log_agi(
 	xfs_agi_t		*agi;	/* allocation group header */
 
 	agi = XFS_BUF_TO_AGI(bp);
-	ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
+	ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
 #endif
 	/*
 	 * Compute byte offsets for the first and last fields.
@@ -1492,7 +1492,7 @@ xfs_read_agi(
 	/*
 	 * Validate the magic number of the agi block.
 	 */
-	agi_ok = be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC &&
+	agi_ok = agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC) &&
 		XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)) &&
 		be32_to_cpu(agi->agi_seqno) == agno;
 	if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI,
diff --git a/fs/xfs/xfs_ialloc_btree.c b/fs/xfs/xfs_ialloc_btree.c
index 16921f55c54..c6a75815aea 100644
--- a/fs/xfs/xfs_ialloc_btree.c
+++ b/fs/xfs/xfs_ialloc_btree.c
@@ -31,7 +31,6 @@
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_btree.h"
-#include "xfs_btree_trace.h"
 #include "xfs_ialloc.h"
 #include "xfs_alloc.h"
 #include "xfs_error.h"
@@ -205,72 +204,6 @@ xfs_inobt_recs_inorder(
 }
 #endif	/* DEBUG */
 
-#ifdef XFS_BTREE_TRACE
-ktrace_t	*xfs_inobt_trace_buf;
-
-STATIC void
-xfs_inobt_trace_enter(
-	struct xfs_btree_cur	*cur,
-	const char		*func,
-	char			*s,
-	int			type,
-	int			line,
-	__psunsigned_t		a0,
-	__psunsigned_t		a1,
-	__psunsigned_t		a2,
-	__psunsigned_t		a3,
-	__psunsigned_t		a4,
-	__psunsigned_t		a5,
-	__psunsigned_t		a6,
-	__psunsigned_t		a7,
-	__psunsigned_t		a8,
-	__psunsigned_t		a9,
-	__psunsigned_t		a10)
-{
-	ktrace_enter(xfs_inobt_trace_buf, (void *)(__psint_t)type,
-		(void *)func, (void *)s, NULL, (void *)cur,
-		(void *)a0, (void *)a1, (void *)a2, (void *)a3,
-		(void *)a4, (void *)a5, (void *)a6, (void *)a7,
-		(void *)a8, (void *)a9, (void *)a10);
-}
-
-STATIC void
-xfs_inobt_trace_cursor(
-	struct xfs_btree_cur	*cur,
-	__uint32_t		*s0,
-	__uint64_t		*l0,
-	__uint64_t		*l1)
-{
-	*s0 = cur->bc_private.a.agno;
-	*l0 = cur->bc_rec.i.ir_startino;
-	*l1 = cur->bc_rec.i.ir_free;
-}
-
-STATIC void
-xfs_inobt_trace_key(
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_key	*key,
-	__uint64_t		*l0,
-	__uint64_t		*l1)
-{
-	*l0 = be32_to_cpu(key->inobt.ir_startino);
-	*l1 = 0;
-}
-
-STATIC void
-xfs_inobt_trace_record(
-	struct xfs_btree_cur	*cur,
-	union xfs_btree_rec	*rec,
-	__uint64_t		*l0,
-	__uint64_t		*l1,
-	__uint64_t		*l2)
-{
-	*l0 = be32_to_cpu(rec->inobt.ir_startino);
-	*l1 = be32_to_cpu(rec->inobt.ir_freecount);
-	*l2 = be64_to_cpu(rec->inobt.ir_free);
-}
-#endif /* XFS_BTREE_TRACE */
-
 static const struct xfs_btree_ops xfs_inobt_ops = {
 	.rec_len		= sizeof(xfs_inobt_rec_t),
 	.key_len		= sizeof(xfs_inobt_key_t),
@@ -286,18 +219,10 @@ static const struct xfs_btree_ops xfs_inobt_ops = {
 	.init_rec_from_cur	= xfs_inobt_init_rec_from_cur,
 	.init_ptr_from_cur	= xfs_inobt_init_ptr_from_cur,
 	.key_diff		= xfs_inobt_key_diff,
-
 #ifdef DEBUG
 	.keys_inorder		= xfs_inobt_keys_inorder,
 	.recs_inorder		= xfs_inobt_recs_inorder,
 #endif
-
-#ifdef XFS_BTREE_TRACE
-	.trace_enter		= xfs_inobt_trace_enter,
-	.trace_cursor		= xfs_inobt_trace_cursor,
-	.trace_key		= xfs_inobt_trace_key,
-	.trace_record		= xfs_inobt_trace_record,
-#endif
 };
 
 /*
diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c
index cb9b6d1469f..7759812c1bb 100644
--- a/fs/xfs/xfs_iget.c
+++ b/fs/xfs/xfs_iget.c
@@ -38,7 +38,6 @@
 #include "xfs_trans_priv.h"
 #include "xfs_inode_item.h"
 #include "xfs_bmap.h"
-#include "xfs_btree_trace.h"
 #include "xfs_trace.h"
 
 
@@ -253,16 +252,21 @@ xfs_iget_cache_hit(
 			rcu_read_lock();
 			spin_lock(&ip->i_flags_lock);
 
-			ip->i_flags &= ~XFS_INEW;
-			ip->i_flags |= XFS_IRECLAIMABLE;
-			__xfs_inode_set_reclaim_tag(pag, ip);
+			ip->i_flags &= ~(XFS_INEW | XFS_IRECLAIM);
+			ASSERT(ip->i_flags & XFS_IRECLAIMABLE);
 			trace_xfs_iget_reclaim_fail(ip);
 			goto out_error;
 		}
 
 		spin_lock(&pag->pag_ici_lock);
 		spin_lock(&ip->i_flags_lock);
-		ip->i_flags &= ~(XFS_IRECLAIMABLE | XFS_IRECLAIM);
+
+		/*
+		 * Clear the per-lifetime state in the inode as we are now
+		 * effectively a new inode and need to return to the initial
+		 * state before reuse occurs.
+		 */
+		ip->i_flags &= ~XFS_IRECLAIM_RESET_FLAGS;
 		ip->i_flags |= XFS_INEW;
 		__xfs_inode_clear_reclaim_tag(mp, pag, ip);
 		inode->i_state = I_NEW;
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index a098a20ca63..3cc21ddf9f7 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -37,7 +37,6 @@
 #include "xfs_buf_item.h"
 #include "xfs_inode_item.h"
 #include "xfs_btree.h"
-#include "xfs_btree_trace.h"
 #include "xfs_alloc.h"
 #include "xfs_ialloc.h"
 #include "xfs_bmap.h"
@@ -52,7 +51,7 @@ kmem_zone_t *xfs_ifork_zone;
 kmem_zone_t *xfs_inode_zone;
 
 /*
- * Used in xfs_itruncate().  This is the maximum number of extents
+ * Used in xfs_itruncate_extents().  This is the maximum number of extents
  * freed from a file in a single transaction.
  */
 #define	XFS_ITRUNC_MAX_EXTENTS	2
@@ -167,7 +166,7 @@ xfs_imap_to_bp(
 
 		dip = (xfs_dinode_t *)xfs_buf_offset(bp,
 					(i << mp->m_sb.sb_inodelog));
-		di_ok = be16_to_cpu(dip->di_magic) == XFS_DINODE_MAGIC &&
+		di_ok = dip->di_magic == cpu_to_be16(XFS_DINODE_MAGIC) &&
 			    XFS_DINODE_GOOD_VERSION(dip->di_version);
 		if (unlikely(XFS_TEST_ERROR(!di_ok, mp,
 						XFS_ERRTAG_ITOBP_INOTOBP,
@@ -802,7 +801,7 @@ xfs_iread(
 	 * If we got something that isn't an inode it means someone
 	 * (nfs or dmi) has a stale handle.
 	 */
-	if (be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC) {
+	if (dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC)) {
 #ifdef DEBUG
 		xfs_alert(mp,
 			"%s: dip->di_magic (0x%x) != XFS_DINODE_MAGIC (0x%x)",
@@ -1179,15 +1178,15 @@ xfs_ialloc(
  * at least do it for regular files.
  */
 #ifdef DEBUG
-void
+STATIC void
 xfs_isize_check(
-	xfs_mount_t	*mp,
-	xfs_inode_t	*ip,
-	xfs_fsize_t	isize)
+	struct xfs_inode	*ip,
+	xfs_fsize_t		isize)
 {
-	xfs_fileoff_t	map_first;
-	int		nimaps;
-	xfs_bmbt_irec_t	imaps[2];
+	struct xfs_mount	*mp = ip->i_mount;
+	xfs_fileoff_t		map_first;
+	int			nimaps;
+	xfs_bmbt_irec_t		imaps[2];
 
 	if ((ip->i_d.di_mode & S_IFMT) != S_IFREG)
 		return;
@@ -1214,168 +1213,14 @@ xfs_isize_check(
 	ASSERT(nimaps == 1);
 	ASSERT(imaps[0].br_startblock == HOLESTARTBLOCK);
 }
+#else	/* DEBUG */
+#define xfs_isize_check(ip, isize)
 #endif	/* DEBUG */
 
 /*
- * Calculate the last possible buffered byte in a file.  This must
- * include data that was buffered beyond the EOF by the write code.
- * This also needs to deal with overflowing the xfs_fsize_t type
- * which can happen for sizes near the limit.
- *
- * We also need to take into account any blocks beyond the EOF.  It
- * may be the case that they were buffered by a write which failed.
- * In that case the pages will still be in memory, but the inode size
- * will never have been updated.
- */
-STATIC xfs_fsize_t
-xfs_file_last_byte(
-	xfs_inode_t	*ip)
-{
-	xfs_mount_t	*mp;
-	xfs_fsize_t	last_byte;
-	xfs_fileoff_t	last_block;
-	xfs_fileoff_t	size_last_block;
-	int		error;
-
-	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED));
-
-	mp = ip->i_mount;
-	/*
-	 * Only check for blocks beyond the EOF if the extents have
-	 * been read in.  This eliminates the need for the inode lock,
-	 * and it also saves us from looking when it really isn't
-	 * necessary.
-	 */
-	if (ip->i_df.if_flags & XFS_IFEXTENTS) {
-		xfs_ilock(ip, XFS_ILOCK_SHARED);
-		error = xfs_bmap_last_offset(NULL, ip, &last_block,
-			XFS_DATA_FORK);
-		xfs_iunlock(ip, XFS_ILOCK_SHARED);
-		if (error) {
-			last_block = 0;
-		}
-	} else {
-		last_block = 0;
-	}
-	size_last_block = XFS_B_TO_FSB(mp, (xfs_ufsize_t)ip->i_size);
-	last_block = XFS_FILEOFF_MAX(last_block, size_last_block);
-
-	last_byte = XFS_FSB_TO_B(mp, last_block);
-	if (last_byte < 0) {
-		return XFS_MAXIOFFSET(mp);
-	}
-	last_byte += (1 << mp->m_writeio_log);
-	if (last_byte < 0) {
-		return XFS_MAXIOFFSET(mp);
-	}
-	return last_byte;
-}
-
-/*
- * Start the truncation of the file to new_size.  The new size
- * must be smaller than the current size.  This routine will
- * clear the buffer and page caches of file data in the removed
- * range, and xfs_itruncate_finish() will remove the underlying
- * disk blocks.
- *
- * The inode must have its I/O lock locked EXCLUSIVELY, and it
- * must NOT have the inode lock held at all.  This is because we're
- * calling into the buffer/page cache code and we can't hold the
- * inode lock when we do so.
- *
- * We need to wait for any direct I/Os in flight to complete before we
- * proceed with the truncate. This is needed to prevent the extents
- * being read or written by the direct I/Os from being removed while the
- * I/O is in flight as there is no other method of synchronising
- * direct I/O with the truncate operation.  Also, because we hold
- * the IOLOCK in exclusive mode, we prevent new direct I/Os from being
- * started until the truncate completes and drops the lock. Essentially,
- * the xfs_ioend_wait() call forms an I/O barrier that provides strict
- * ordering between direct I/Os and the truncate operation.
- *
- * The flags parameter can have either the value XFS_ITRUNC_DEFINITE
- * or XFS_ITRUNC_MAYBE.  The XFS_ITRUNC_MAYBE value should be used
- * in the case that the caller is locking things out of order and
- * may not be able to call xfs_itruncate_finish() with the inode lock
- * held without dropping the I/O lock.  If the caller must drop the
- * I/O lock before calling xfs_itruncate_finish(), then xfs_itruncate_start()
- * must be called again with all the same restrictions as the initial
- * call.
- */
-int
-xfs_itruncate_start(
-	xfs_inode_t	*ip,
-	uint		flags,
-	xfs_fsize_t	new_size)
-{
-	xfs_fsize_t	last_byte;
-	xfs_off_t	toss_start;
-	xfs_mount_t	*mp;
-	int		error = 0;
-
-	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
-	ASSERT((new_size == 0) || (new_size <= ip->i_size));
-	ASSERT((flags == XFS_ITRUNC_DEFINITE) ||
-	       (flags == XFS_ITRUNC_MAYBE));
-
-	mp = ip->i_mount;
-
-	/* wait for the completion of any pending DIOs */
-	if (new_size == 0 || new_size < ip->i_size)
-		xfs_ioend_wait(ip);
-
-	/*
-	 * Call toss_pages or flushinval_pages to get rid of pages
-	 * overlapping the region being removed.  We have to use
-	 * the less efficient flushinval_pages in the case that the
-	 * caller may not be able to finish the truncate without
-	 * dropping the inode's I/O lock.  Make sure
-	 * to catch any pages brought in by buffers overlapping
-	 * the EOF by searching out beyond the isize by our
-	 * block size. We round new_size up to a block boundary
-	 * so that we don't toss things on the same block as
-	 * new_size but before it.
-	 *
-	 * Before calling toss_page or flushinval_pages, make sure to
-	 * call remapf() over the same region if the file is mapped.
-	 * This frees up mapped file references to the pages in the
-	 * given range and for the flushinval_pages case it ensures
-	 * that we get the latest mapped changes flushed out.
-	 */
-	toss_start = XFS_B_TO_FSB(mp, (xfs_ufsize_t)new_size);
-	toss_start = XFS_FSB_TO_B(mp, toss_start);
-	if (toss_start < 0) {
-		/*
-		 * The place to start tossing is beyond our maximum
-		 * file size, so there is no way that the data extended
-		 * out there.
-		 */
-		return 0;
-	}
-	last_byte = xfs_file_last_byte(ip);
-	trace_xfs_itruncate_start(ip, new_size, flags, toss_start, last_byte);
-	if (last_byte > toss_start) {
-		if (flags & XFS_ITRUNC_DEFINITE) {
-			xfs_tosspages(ip, toss_start,
-					-1, FI_REMAPF_LOCKED);
-		} else {
-			error = xfs_flushinval_pages(ip, toss_start,
-					-1, FI_REMAPF_LOCKED);
-		}
-	}
-
-#ifdef DEBUG
-	if (new_size == 0) {
-		ASSERT(VN_CACHED(VFS_I(ip)) == 0);
-	}
-#endif
-	return error;
-}
-
-/*
- * Shrink the file to the given new_size.  The new size must be smaller than
- * the current size.  This will free up the underlying blocks in the removed
- * range after a call to xfs_itruncate_start() or xfs_atruncate_start().
+ * Free up the underlying blocks past new_size.  The new size must be smaller
+ * than the current size.  This routine can be used both for the attribute and
+ * data fork, and does not modify the inode size, which is left to the caller.
  *
  * The transaction passed to this routine must have made a permanent log
  * reservation of at least XFS_ITRUNCATE_LOG_RES.  This routine may commit the
@@ -1387,31 +1232,6 @@ xfs_itruncate_start(
  * will be "held" within the returned transaction.  This routine does NOT
  * require any disk space to be reserved for it within the transaction.
  *
- * The fork parameter must be either xfs_attr_fork or xfs_data_fork, and it
- * indicates the fork which is to be truncated.  For the attribute fork we only
- * support truncation to size 0.
- *
- * We use the sync parameter to indicate whether or not the first transaction
- * we perform might have to be synchronous.  For the attr fork, it needs to be
- * so if the unlink of the inode is not yet known to be permanent in the log.
- * This keeps us from freeing and reusing the blocks of the attribute fork
- * before the unlink of the inode becomes permanent.
- *
- * For the data fork, we normally have to run synchronously if we're being
- * called out of the inactive path or we're being called out of the create path
- * where we're truncating an existing file.  Either way, the truncate needs to
- * be sync so blocks don't reappear in the file with altered data in case of a
- * crash.  wsync filesystems can run the first case async because anything that
- * shrinks the inode has to run sync so by the time we're called here from
- * inactive, the inode size is permanently set to 0.
- *
- * Calls from the truncate path always need to be sync unless we're in a wsync
- * filesystem and the file has already been unlinked.
- *
- * The caller is responsible for correctly setting the sync parameter.  It gets
- * too hard for us to guess here which path we're being called out of just
- * based on inode state.
- *
  * If we get an error, we must return with the inode locked and linked into the
  * current transaction. This keeps things simple for the higher level code,
  * because it always knows that the inode is locked and held in the transaction
@@ -1419,124 +1239,30 @@ xfs_itruncate_start(
  * dirty on error so that transactions can be easily aborted if possible.
  */
 int
-xfs_itruncate_finish(
-	xfs_trans_t	**tp,
-	xfs_inode_t	*ip,
-	xfs_fsize_t	new_size,
-	int		fork,
-	int		sync)
+xfs_itruncate_extents(
+	struct xfs_trans	**tpp,
+	struct xfs_inode	*ip,
+	int			whichfork,
+	xfs_fsize_t		new_size)
 {
-	xfs_fsblock_t	first_block;
-	xfs_fileoff_t	first_unmap_block;
-	xfs_fileoff_t	last_block;
-	xfs_filblks_t	unmap_len=0;
-	xfs_mount_t	*mp;
-	xfs_trans_t	*ntp;
-	int		done;
-	int		committed;
-	xfs_bmap_free_t	free_list;
-	int		error;
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_trans	*tp = *tpp;
+	struct xfs_trans	*ntp;
+	xfs_bmap_free_t		free_list;
+	xfs_fsblock_t		first_block;
+	xfs_fileoff_t		first_unmap_block;
+	xfs_fileoff_t		last_block;
+	xfs_filblks_t		unmap_len;
+	int			committed;
+	int			error = 0;
+	int			done = 0;
 
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL));
-	ASSERT((new_size == 0) || (new_size <= ip->i_size));
-	ASSERT(*tp != NULL);
-	ASSERT((*tp)->t_flags & XFS_TRANS_PERM_LOG_RES);
-	ASSERT(ip->i_transp == *tp);
+	ASSERT(new_size <= ip->i_size);
+	ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
 	ASSERT(ip->i_itemp != NULL);
 	ASSERT(ip->i_itemp->ili_lock_flags == 0);
-
-
-	ntp = *tp;
-	mp = (ntp)->t_mountp;
-	ASSERT(! XFS_NOT_DQATTACHED(mp, ip));
-
-	/*
-	 * We only support truncating the entire attribute fork.
-	 */
-	if (fork == XFS_ATTR_FORK) {
-		new_size = 0LL;
-	}
-	first_unmap_block = XFS_B_TO_FSB(mp, (xfs_ufsize_t)new_size);
-	trace_xfs_itruncate_finish_start(ip, new_size);
-
-	/*
-	 * The first thing we do is set the size to new_size permanently
-	 * on disk.  This way we don't have to worry about anyone ever
-	 * being able to look at the data being freed even in the face
-	 * of a crash.  What we're getting around here is the case where
-	 * we free a block, it is allocated to another file, it is written
-	 * to, and then we crash.  If the new data gets written to the
-	 * file but the log buffers containing the free and reallocation
-	 * don't, then we'd end up with garbage in the blocks being freed.
-	 * As long as we make the new_size permanent before actually
-	 * freeing any blocks it doesn't matter if they get written to.
-	 *
-	 * The callers must signal into us whether or not the size
-	 * setting here must be synchronous.  There are a few cases
-	 * where it doesn't have to be synchronous.  Those cases
-	 * occur if the file is unlinked and we know the unlink is
-	 * permanent or if the blocks being truncated are guaranteed
-	 * to be beyond the inode eof (regardless of the link count)
-	 * and the eof value is permanent.  Both of these cases occur
-	 * only on wsync-mounted filesystems.  In those cases, we're
-	 * guaranteed that no user will ever see the data in the blocks
-	 * that are being truncated so the truncate can run async.
-	 * In the free beyond eof case, the file may wind up with
-	 * more blocks allocated to it than it needs if we crash
-	 * and that won't get fixed until the next time the file
-	 * is re-opened and closed but that's ok as that shouldn't
-	 * be too many blocks.
-	 *
-	 * However, we can't just make all wsync xactions run async
-	 * because there's one call out of the create path that needs
-	 * to run sync where it's truncating an existing file to size
-	 * 0 whose size is > 0.
-	 *
-	 * It's probably possible to come up with a test in this
-	 * routine that would correctly distinguish all the above
-	 * cases from the values of the function parameters and the
-	 * inode state but for sanity's sake, I've decided to let the
-	 * layers above just tell us.  It's simpler to correctly figure
-	 * out in the layer above exactly under what conditions we
-	 * can run async and I think it's easier for others read and
-	 * follow the logic in case something has to be changed.
-	 * cscope is your friend -- rcc.
-	 *
-	 * The attribute fork is much simpler.
-	 *
-	 * For the attribute fork we allow the caller to tell us whether
-	 * the unlink of the inode that led to this call is yet permanent
-	 * in the on disk log.  If it is not and we will be freeing extents
-	 * in this inode then we make the first transaction synchronous
-	 * to make sure that the unlink is permanent by the time we free
-	 * the blocks.
-	 */
-	if (fork == XFS_DATA_FORK) {
-		if (ip->i_d.di_nextents > 0) {
-			/*
-			 * If we are not changing the file size then do
-			 * not update the on-disk file size - we may be
-			 * called from xfs_inactive_free_eofblocks().  If we
-			 * update the on-disk file size and then the system
-			 * crashes before the contents of the file are
-			 * flushed to disk then the files may be full of
-			 * holes (ie NULL files bug).
-			 */
-			if (ip->i_size != new_size) {
-				ip->i_d.di_size = new_size;
-				ip->i_size = new_size;
-				xfs_trans_log_inode(ntp, ip, XFS_ILOG_CORE);
-			}
-		}
-	} else if (sync) {
-		ASSERT(!(mp->m_flags & XFS_MOUNT_WSYNC));
-		if (ip->i_d.di_anextents > 0)
-			xfs_trans_set_sync(ntp);
-	}
-	ASSERT(fork == XFS_DATA_FORK ||
-		(fork == XFS_ATTR_FORK &&
-			((sync && !(mp->m_flags & XFS_MOUNT_WSYNC)) ||
-			 (sync == 0 && (mp->m_flags & XFS_MOUNT_WSYNC)))));
+	ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
 
 	/*
 	 * Since it is possible for space to become allocated beyond
@@ -1547,128 +1273,142 @@ xfs_itruncate_finish(
 	 * beyond the maximum file size (ie it is the same as last_block),
 	 * then there is nothing to do.
 	 */
+	first_unmap_block = XFS_B_TO_FSB(mp, (xfs_ufsize_t)new_size);
 	last_block = XFS_B_TO_FSB(mp, (xfs_ufsize_t)XFS_MAXIOFFSET(mp));
-	ASSERT(first_unmap_block <= last_block);
-	done = 0;
-	if (last_block == first_unmap_block) {
-		done = 1;
-	} else {
-		unmap_len = last_block - first_unmap_block + 1;
-	}
+	if (first_unmap_block == last_block)
+		return 0;
+
+	ASSERT(first_unmap_block < last_block);
+	unmap_len = last_block - first_unmap_block + 1;
 	while (!done) {
-		/*
-		 * Free up up to XFS_ITRUNC_MAX_EXTENTS.  xfs_bunmapi()
-		 * will tell us whether it freed the entire range or
-		 * not.  If this is a synchronous mount (wsync),
-		 * then we can tell bunmapi to keep all the
-		 * transactions asynchronous since the unlink
-		 * transaction that made this inode inactive has
-		 * already hit the disk.  There's no danger of
-		 * the freed blocks being reused, there being a
-		 * crash, and the reused blocks suddenly reappearing
-		 * in this file with garbage in them once recovery
-		 * runs.
-		 */
 		xfs_bmap_init(&free_list, &first_block);
-		error = xfs_bunmapi(ntp, ip,
+		error = xfs_bunmapi(tp, ip,
 				    first_unmap_block, unmap_len,
-				    xfs_bmapi_aflag(fork),
+				    xfs_bmapi_aflag(whichfork),
 				    XFS_ITRUNC_MAX_EXTENTS,
 				    &first_block, &free_list,
 				    &done);
-		if (error) {
-			/*
-			 * If the bunmapi call encounters an error,
-			 * return to the caller where the transaction
-			 * can be properly aborted.  We just need to
-			 * make sure we're not holding any resources
-			 * that we were not when we came in.
-			 */
-			xfs_bmap_cancel(&free_list);
-			return error;
-		}
+		if (error)
+			goto out_bmap_cancel;
 
 		/*
 		 * Duplicate the transaction that has the permanent
 		 * reservation and commit the old transaction.
 		 */
-		error = xfs_bmap_finish(tp, &free_list, &committed);
-		ntp = *tp;
+		error = xfs_bmap_finish(&tp, &free_list, &committed);
 		if (committed)
-			xfs_trans_ijoin(ntp, ip);
-
-		if (error) {
-			/*
-			 * If the bmap finish call encounters an error, return
-			 * to the caller where the transaction can be properly
-			 * aborted.  We just need to make sure we're not
-			 * holding any resources that we were not when we came
-			 * in.
-			 *
-			 * Aborting from this point might lose some blocks in
-			 * the file system, but oh well.
-			 */
-			xfs_bmap_cancel(&free_list);
-			return error;
-		}
+			xfs_trans_ijoin(tp, ip);
+		if (error)
+			goto out_bmap_cancel;
 
 		if (committed) {
 			/*
 			 * Mark the inode dirty so it will be logged and
 			 * moved forward in the log as part of every commit.
 			 */
-			xfs_trans_log_inode(ntp, ip, XFS_ILOG_CORE);
+			xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 		}
 
-		ntp = xfs_trans_dup(ntp);
-		error = xfs_trans_commit(*tp, 0);
-		*tp = ntp;
+		ntp = xfs_trans_dup(tp);
+		error = xfs_trans_commit(tp, 0);
+		tp = ntp;
 
-		xfs_trans_ijoin(ntp, ip);
+		xfs_trans_ijoin(tp, ip);
 
 		if (error)
-			return error;
+			goto out;
+
 		/*
-		 * transaction commit worked ok so we can drop the extra ticket
+		 * Transaction commit worked ok so we can drop the extra ticket
 		 * reference that we gained in xfs_trans_dup()
 		 */
-		xfs_log_ticket_put(ntp->t_ticket);
-		error = xfs_trans_reserve(ntp, 0,
+		xfs_log_ticket_put(tp->t_ticket);
+		error = xfs_trans_reserve(tp, 0,
 					XFS_ITRUNCATE_LOG_RES(mp), 0,
 					XFS_TRANS_PERM_LOG_RES,
 					XFS_ITRUNCATE_LOG_COUNT);
 		if (error)
-			return error;
+			goto out;
 	}
+
+out:
+	*tpp = tp;
+	return error;
+out_bmap_cancel:
 	/*
-	 * Only update the size in the case of the data fork, but
-	 * always re-log the inode so that our permanent transaction
-	 * can keep on rolling it forward in the log.
+	 * If the bunmapi call encounters an error, return to the caller where
+	 * the transaction can be properly aborted.  We just need to make sure
+	 * we're not holding any resources that we were not when we came in.
 	 */
-	if (fork == XFS_DATA_FORK) {
-		xfs_isize_check(mp, ip, new_size);
+	xfs_bmap_cancel(&free_list);
+	goto out;
+}
+
+int
+xfs_itruncate_data(
+	struct xfs_trans	**tpp,
+	struct xfs_inode	*ip,
+	xfs_fsize_t		new_size)
+{
+	int			error;
+
+	trace_xfs_itruncate_data_start(ip, new_size);
+
+	/*
+	 * The first thing we do is set the size to new_size permanently on
+	 * disk.  This way we don't have to worry about anyone ever being able
+	 * to look at the data being freed even in the face of a crash.
+	 * What we're getting around here is the case where we free a block, it
+	 * is allocated to another file, it is written to, and then we crash.
+	 * If the new data gets written to the file but the log buffers
+	 * containing the free and reallocation don't, then we'd end up with
+	 * garbage in the blocks being freed.  As long as we make the new_size
+	 * permanent before actually freeing any blocks it doesn't matter if
+	 * they get written to.
+	 */
+	if (ip->i_d.di_nextents > 0) {
 		/*
-		 * If we are not changing the file size then do
-		 * not update the on-disk file size - we may be
-		 * called from xfs_inactive_free_eofblocks().  If we
-		 * update the on-disk file size and then the system
-		 * crashes before the contents of the file are
-		 * flushed to disk then the files may be full of
-		 * holes (ie NULL files bug).
+		 * If we are not changing the file size then do not update
+		 * the on-disk file size - we may be called from
+		 * xfs_inactive_free_eofblocks().  If we update the on-disk
+		 * file size and then the system crashes before the contents
+		 * of the file are flushed to disk then the files may be
+		 * full of holes (ie NULL files bug).
 		 */
 		if (ip->i_size != new_size) {
 			ip->i_d.di_size = new_size;
 			ip->i_size = new_size;
+			xfs_trans_log_inode(*tpp, ip, XFS_ILOG_CORE);
 		}
 	}
-	xfs_trans_log_inode(ntp, ip, XFS_ILOG_CORE);
-	ASSERT((new_size != 0) ||
-	       (fork == XFS_ATTR_FORK) ||
-	       (ip->i_delayed_blks == 0));
-	ASSERT((new_size != 0) ||
-	       (fork == XFS_ATTR_FORK) ||
-	       (ip->i_d.di_nextents == 0));
-	trace_xfs_itruncate_finish_end(ip, new_size);
+
+	error = xfs_itruncate_extents(tpp, ip, XFS_DATA_FORK, new_size);
+	if (error)
+		return error;
+
+	/*
+	 * If we are not changing the file size then do not update the on-disk
+	 * file size - we may be called from xfs_inactive_free_eofblocks().
+	 * If we update the on-disk file size and then the system crashes
+	 * before the contents of the file are flushed to disk then the files
+	 * may be full of holes (ie NULL files bug).
+	 */
+	xfs_isize_check(ip, new_size);
+	if (ip->i_size != new_size) {
+		ip->i_d.di_size = new_size;
+		ip->i_size = new_size;
+	}
+
+	ASSERT(new_size != 0 || ip->i_delayed_blks == 0);
+	ASSERT(new_size != 0 || ip->i_d.di_nextents == 0);
+
+	/*
+	 * Always re-log the inode so that our permanent transaction can keep
+	 * on rolling it forward in the log.
+	 */
+	xfs_trans_log_inode(*tpp, ip, XFS_ILOG_CORE);
+
+	trace_xfs_itruncate_data_end(ip, new_size);
 	return 0;
 }
 
@@ -1694,7 +1434,6 @@ xfs_iunlink(
 
 	ASSERT(ip->i_d.di_nlink == 0);
 	ASSERT(ip->i_d.di_mode != 0);
-	ASSERT(ip->i_transp == tp);
 
 	mp = tp->t_mountp;
 
@@ -1717,7 +1456,7 @@ xfs_iunlink(
 	ASSERT(agi->agi_unlinked[bucket_index]);
 	ASSERT(be32_to_cpu(agi->agi_unlinked[bucket_index]) != agino);
 
-	if (be32_to_cpu(agi->agi_unlinked[bucket_index]) != NULLAGINO) {
+	if (agi->agi_unlinked[bucket_index] != cpu_to_be32(NULLAGINO)) {
 		/*
 		 * There is already another inode in the bucket we need
 		 * to add ourselves to.  Add us at the front of the list.
@@ -1728,8 +1467,7 @@ xfs_iunlink(
 		if (error)
 			return error;
 
-		ASSERT(be32_to_cpu(dip->di_next_unlinked) == NULLAGINO);
-		/* both on-disk, don't endian flip twice */
+		ASSERT(dip->di_next_unlinked == cpu_to_be32(NULLAGINO));
 		dip->di_next_unlinked = agi->agi_unlinked[bucket_index];
 		offset = ip->i_imap.im_boffset +
 			offsetof(xfs_dinode_t, di_next_unlinked);
@@ -1794,7 +1532,7 @@ xfs_iunlink_remove(
 	agino = XFS_INO_TO_AGINO(mp, ip->i_ino);
 	ASSERT(agino != 0);
 	bucket_index = agino % XFS_AGI_UNLINKED_BUCKETS;
-	ASSERT(be32_to_cpu(agi->agi_unlinked[bucket_index]) != NULLAGINO);
+	ASSERT(agi->agi_unlinked[bucket_index] != cpu_to_be32(NULLAGINO));
 	ASSERT(agi->agi_unlinked[bucket_index]);
 
 	if (be32_to_cpu(agi->agi_unlinked[bucket_index]) == agino) {
@@ -1959,7 +1697,7 @@ xfs_ifree_cluster(
 		 * stale first, we will not attempt to lock them in the loop
 		 * below as the XFS_ISTALE flag will be set.
 		 */
-		lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
+		lip = bp->b_fspriv;
 		while (lip) {
 			if (lip->li_type == XFS_LI_INODE) {
 				iip = (xfs_inode_log_item_t *)lip;
@@ -2086,7 +1824,6 @@ xfs_ifree(
 	xfs_buf_t       	*ibp;
 
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
-	ASSERT(ip->i_transp == tp);
 	ASSERT(ip->i_d.di_nlink == 0);
 	ASSERT(ip->i_d.di_nextents == 0);
 	ASSERT(ip->i_d.di_anextents == 0);
@@ -2733,7 +2470,7 @@ cluster_corrupt_out:
 		 * mark the buffer as an error and call them.  Otherwise
 		 * mark it as stale and brelse.
 		 */
-		if (XFS_BUF_IODONE_FUNC(bp)) {
+		if (bp->b_iodone) {
 			XFS_BUF_UNDONE(bp);
 			XFS_BUF_STALE(bp);
 			XFS_BUF_ERROR(bp,EIO);
@@ -2920,7 +2657,7 @@ xfs_iflush_int(
 	 */
 	xfs_synchronize_times(ip);
 
-	if (XFS_TEST_ERROR(be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC,
+	if (XFS_TEST_ERROR(dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC),
 			       mp, XFS_ERRTAG_IFLUSH_1, XFS_RANDOM_IFLUSH_1)) {
 		xfs_alert_tag(mp, XFS_PTAG_IFLUSH,
 			"%s: Bad inode %Lu magic number 0x%x, ptr 0x%p",
@@ -3073,8 +2810,8 @@ xfs_iflush_int(
 		 */
 		xfs_buf_attach_iodone(bp, xfs_iflush_done, &iip->ili_item);
 
-		ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
-		ASSERT(XFS_BUF_IODONE_FUNC(bp) != NULL);
+		ASSERT(bp->b_fspriv != NULL);
+		ASSERT(bp->b_iodone != NULL);
 	} else {
 		/*
 		 * We're flushing an inode which is not in the AIL and has
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 3ae6d58e547..a97644ab945 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -241,7 +241,6 @@ typedef struct xfs_inode {
 	xfs_ifork_t		i_df;		/* data fork */
 
 	/* Transaction and locking information. */
-	struct xfs_trans	*i_transp;	/* ptr to owning transaction*/
 	struct xfs_inode_log_item *i_itemp;	/* logging information */
 	mrlock_t		i_lock;		/* inode lock */
 	mrlock_t		i_iolock;	/* inode IO lock */
@@ -384,6 +383,16 @@ static inline void xfs_ifunlock(xfs_inode_t *ip)
 #define XFS_IDIRTY_RELEASE	0x0040	/* dirty release already seen */
 
 /*
+ * Per-lifetime flags need to be reset when re-using a reclaimable inode during
+ * inode lookup. Thi prevents unintended behaviour on the new inode from
+ * ocurring.
+ */
+#define XFS_IRECLAIM_RESET_FLAGS	\
+	(XFS_IRECLAIMABLE | XFS_IRECLAIM | \
+	 XFS_IDIRTY_RELEASE | XFS_ITRUNCATED | \
+	 XFS_IFILESTREAM);
+
+/*
  * Flags for inode locking.
  * Bit ranges:	1<<1  - 1<<16-1 -- iolock/ilock modes (bitfield)
  *		1<<16 - 1<<32-1 -- lockdep annotation (integers)
@@ -448,16 +457,6 @@ static inline void xfs_ifunlock(xfs_inode_t *ip)
 extern struct lock_class_key xfs_iolock_reclaimable;
 
 /*
- * Flags for xfs_itruncate_start().
- */
-#define	XFS_ITRUNC_DEFINITE	0x1
-#define	XFS_ITRUNC_MAYBE	0x2
-
-#define XFS_ITRUNC_FLAGS \
-	{ XFS_ITRUNC_DEFINITE,	"DEFINITE" }, \
-	{ XFS_ITRUNC_MAYBE,	"MAYBE" }
-
-/*
  * For multiple groups support: if S_ISGID bit is set in the parent
  * directory, group of new file is set to that of the parent, and
  * new subdirectory gets S_ISGID bit from parent.
@@ -491,9 +490,10 @@ uint		xfs_ip2xflags(struct xfs_inode *);
 uint		xfs_dic2xflags(struct xfs_dinode *);
 int		xfs_ifree(struct xfs_trans *, xfs_inode_t *,
 			   struct xfs_bmap_free *);
-int		xfs_itruncate_start(xfs_inode_t *, uint, xfs_fsize_t);
-int		xfs_itruncate_finish(struct xfs_trans **, xfs_inode_t *,
-				     xfs_fsize_t, int, int);
+int		xfs_itruncate_extents(struct xfs_trans **, struct xfs_inode *,
+				      int, xfs_fsize_t);
+int		xfs_itruncate_data(struct xfs_trans **, struct xfs_inode *,
+				   xfs_fsize_t);
 int		xfs_iunlink(struct xfs_trans *, xfs_inode_t *);
 
 void		xfs_iext_realloc(xfs_inode_t *, int, int);
@@ -569,13 +569,6 @@ void		xfs_iext_irec_update_extoffs(xfs_ifork_t *, int, int);
 
 #define xfs_ipincount(ip)	((unsigned int) atomic_read(&ip->i_pincount))
 
-#ifdef DEBUG
-void		xfs_isize_check(struct xfs_mount *, struct xfs_inode *,
-				xfs_fsize_t);
-#else	/* DEBUG */
-#define xfs_isize_check(mp, ip, isize)
-#endif	/* DEBUG */
-
 #if defined(DEBUG)
 void		xfs_inobp_check(struct xfs_mount *, struct xfs_buf *);
 #else
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index 09983a3344a..588406dc6a3 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -632,13 +632,8 @@ xfs_inode_item_unlock(
 	struct xfs_inode	*ip = iip->ili_inode;
 	unsigned short		lock_flags;
 
-	ASSERT(iip->ili_inode->i_itemp != NULL);
-	ASSERT(xfs_isilocked(iip->ili_inode, XFS_ILOCK_EXCL));
-
-	/*
-	 * Clear the transaction pointer in the inode.
-	 */
-	ip->i_transp = NULL;
+	ASSERT(ip->i_itemp != NULL);
+	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
 
 	/*
 	 * If the inode needed a separate buffer with which to log
@@ -664,8 +659,8 @@ xfs_inode_item_unlock(
 	lock_flags = iip->ili_lock_flags;
 	iip->ili_lock_flags = 0;
 	if (lock_flags) {
-		xfs_iunlock(iip->ili_inode, lock_flags);
-		IRELE(iip->ili_inode);
+		xfs_iunlock(ip, lock_flags);
+		IRELE(ip);
 	}
 }
 
@@ -681,15 +676,15 @@ xfs_inode_item_unlock(
  * where the cluster buffer may be unpinned before the inode is inserted into
  * the AIL during transaction committed processing. If the buffer is unpinned
  * before the inode item has been committed and inserted, then it is possible
- * for the buffer to be written and IO completions before the inode is inserted
+ * for the buffer to be written and IO completes before the inode is inserted
  * into the AIL. In that case, we'd be inserting a clean, stale inode into the
  * AIL which will never get removed. It will, however, get reclaimed which
  * triggers an assert in xfs_inode_free() complaining about freein an inode
  * still in the AIL.
  *
- * To avoid this, return a lower LSN than the one passed in so that the
- * transaction committed code will not move the inode forward in the AIL but
- * will still unpin it properly.
+ * To avoid this, just unpin the inode directly and return a LSN of -1 so the
+ * transaction committed code knows that it does not need to do any further
+ * processing on the item.
  */
 STATIC xfs_lsn_t
 xfs_inode_item_committed(
@@ -699,8 +694,10 @@ xfs_inode_item_committed(
 	struct xfs_inode_log_item *iip = INODE_ITEM(lip);
 	struct xfs_inode	*ip = iip->ili_inode;
 
-	if (xfs_iflags_test(ip, XFS_ISTALE))
-		return lsn - 1;
+	if (xfs_iflags_test(ip, XFS_ISTALE)) {
+		xfs_inode_item_unpin(lip, 0);
+		return -1;
+	}
 	return lsn;
 }
 
@@ -877,7 +874,7 @@ xfs_iflush_done(
 	 * Scan the buffer IO completions for other inodes being completed and
 	 * attach them to the current inode log item.
 	 */
-	blip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
+	blip = bp->b_fspriv;
 	prev = NULL;
 	while (blip != NULL) {
 		if (lip->li_cb != xfs_iflush_done) {
@@ -889,7 +886,7 @@ xfs_iflush_done(
 		/* remove from list */
 		next = blip->li_bio_list;
 		if (!prev) {
-			XFS_BUF_SET_FSPRIVATE(bp, next);
+			bp->b_fspriv = next;
 		} else {
 			prev->li_bio_list = next;
 		}
diff --git a/fs/xfs/xfs_inum.h b/fs/xfs/xfs_inum.h
index b8e4ee4e89a..b253c0ea5be 100644
--- a/fs/xfs/xfs_inum.h
+++ b/fs/xfs/xfs_inum.h
@@ -28,17 +28,6 @@
 
 typedef	__uint32_t	xfs_agino_t;	/* within allocation grp inode number */
 
-/*
- * Useful inode bits for this kernel.
- * Used in some places where having 64-bits in the 32-bit kernels
- * costs too much.
- */
-#if XFS_BIG_INUMS
-typedef	xfs_ino_t	xfs_intino_t;
-#else
-typedef	__uint32_t	xfs_intino_t;
-#endif
-
 #define	NULLFSINO	((xfs_ino_t)-1)
 #define	NULLAGINO	((xfs_agino_t)-1)
 
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 211930246f2..06ff8437ed8 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -871,15 +871,9 @@ xlog_space_left(
 void
 xlog_iodone(xfs_buf_t *bp)
 {
-	xlog_in_core_t	*iclog;
-	xlog_t		*l;
-	int		aborted;
-
-	iclog = XFS_BUF_FSPRIVATE(bp, xlog_in_core_t *);
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, unsigned long) == (unsigned long) 2);
-	XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1);
-	aborted = 0;
-	l = iclog->ic_log;
+	xlog_in_core_t	*iclog = bp->b_fspriv;
+	xlog_t		*l = iclog->ic_log;
+	int		aborted = 0;
 
 	/*
 	 * Race to shutdown the filesystem if we see an error.
@@ -1056,10 +1050,9 @@ xlog_alloc_log(xfs_mount_t	*mp,
 	bp = xfs_buf_get_empty(log->l_iclog_size, mp->m_logdev_targp);
 	if (!bp)
 		goto out_free_log;
-	XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone);
-	XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1);
+	bp->b_iodone = xlog_iodone;
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
+	ASSERT(xfs_buf_islocked(bp));
 	log->l_xbuf = bp;
 
 	spin_lock_init(&log->l_icloglock);
@@ -1090,10 +1083,8 @@ xlog_alloc_log(xfs_mount_t	*mp,
 						log->l_iclog_size, 0);
 		if (!bp)
 			goto out_free_iclog;
-		if (!XFS_BUF_CPSEMA(bp))
-			ASSERT(0);
-		XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone);
-		XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1);
+
+		bp->b_iodone = xlog_iodone;
 		iclog->ic_bp = bp;
 		iclog->ic_data = bp->b_addr;
 #ifdef DEBUG
@@ -1118,7 +1109,7 @@ xlog_alloc_log(xfs_mount_t	*mp,
 		iclog->ic_datap = (char *)iclog->ic_data + log->l_iclog_hsize;
 
 		ASSERT(XFS_BUF_ISBUSY(iclog->ic_bp));
-		ASSERT(XFS_BUF_VALUSEMA(iclog->ic_bp) <= 0);
+		ASSERT(xfs_buf_islocked(iclog->ic_bp));
 		init_waitqueue_head(&iclog->ic_force_wait);
 		init_waitqueue_head(&iclog->ic_write_wait);
 
@@ -1254,9 +1245,8 @@ STATIC int
 xlog_bdstrat(
 	struct xfs_buf		*bp)
 {
-	struct xlog_in_core	*iclog;
+	struct xlog_in_core	*iclog = bp->b_fspriv;
 
-	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);
@@ -1269,7 +1259,6 @@ xlog_bdstrat(
 		return 0;
 	}
 
-	bp->b_flags |= _XBF_RUN_QUEUES;
 	xfs_buf_iorequest(bp);
 	return 0;
 }
@@ -1351,8 +1340,6 @@ xlog_sync(xlog_t		*log,
 	}
 
 	bp = iclog->ic_bp;
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, unsigned long) == (unsigned long)1);
-	XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)2);
 	XFS_BUF_SET_ADDR(bp, BLOCK_LSN(be64_to_cpu(iclog->ic_header.h_lsn)));
 
 	XFS_STATS_ADD(xs_log_blocks, BTOBB(count));
@@ -1366,14 +1353,29 @@ xlog_sync(xlog_t		*log,
 		iclog->ic_bwritecnt = 1;
 	}
 	XFS_BUF_SET_COUNT(bp, count);
-	XFS_BUF_SET_FSPRIVATE(bp, iclog);	/* save for later */
+	bp->b_fspriv = iclog;
 	XFS_BUF_ZEROFLAGS(bp);
 	XFS_BUF_BUSY(bp);
 	XFS_BUF_ASYNC(bp);
-	bp->b_flags |= XBF_LOG_BUFFER;
+	bp->b_flags |= XBF_SYNCIO;
 
-	if (log->l_mp->m_flags & XFS_MOUNT_BARRIER)
-		XFS_BUF_ORDERED(bp);
+	if (log->l_mp->m_flags & XFS_MOUNT_BARRIER) {
+		bp->b_flags |= XBF_FUA;
+
+		/*
+		 * Flush the data device before flushing the log to make
+		 * sure all meta data written back from the AIL actually made
+		 * it to disk before stamping the new log tail LSN into the
+		 * log buffer.  For an external log we need to issue the
+		 * flush explicitly, and unfortunately synchronously here;
+		 * for an internal log we can simply use the block layer
+		 * state machine for preflushes.
+		 */
+		if (log->l_mp->m_logdev_targp != log->l_mp->m_ddev_targp)
+			xfs_blkdev_issue_flush(log->l_mp->m_ddev_targp);
+		else
+			bp->b_flags |= XBF_FLUSH;
+	}
 
 	ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1);
 	ASSERT(XFS_BUF_ADDR(bp) + BTOBB(count) <= log->l_logBBsize);
@@ -1395,19 +1397,16 @@ xlog_sync(xlog_t		*log,
 	}
 	if (split) {
 		bp = iclog->ic_log->l_xbuf;
-		ASSERT(XFS_BUF_FSPRIVATE2(bp, unsigned long) ==
-							(unsigned long)1);
-		XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)2);
 		XFS_BUF_SET_ADDR(bp, 0);	     /* logical 0 */
 		XFS_BUF_SET_PTR(bp, (xfs_caddr_t)((__psint_t)&(iclog->ic_header)+
 					    (__psint_t)count), split);
-		XFS_BUF_SET_FSPRIVATE(bp, iclog);
+		bp->b_fspriv = iclog;
 		XFS_BUF_ZEROFLAGS(bp);
 		XFS_BUF_BUSY(bp);
 		XFS_BUF_ASYNC(bp);
-		bp->b_flags |= XBF_LOG_BUFFER;
+		bp->b_flags |= XBF_SYNCIO;
 		if (log->l_mp->m_flags & XFS_MOUNT_BARRIER)
-			XFS_BUF_ORDERED(bp);
+			bp->b_flags |= XBF_FUA;
 		dptr = XFS_BUF_PTR(bp);
 		/*
 		 * Bump the cycle numbers at the start of each block
@@ -3512,13 +3511,13 @@ xlog_verify_iclog(xlog_t	 *log,
 	spin_unlock(&log->l_icloglock);
 
 	/* check log magic numbers */
-	if (be32_to_cpu(iclog->ic_header.h_magicno) != XLOG_HEADER_MAGIC_NUM)
+	if (iclog->ic_header.h_magicno != cpu_to_be32(XLOG_HEADER_MAGIC_NUM))
 		xfs_emerg(log->l_mp, "%s: invalid magic num", __func__);
 
 	ptr = (xfs_caddr_t) &iclog->ic_header;
 	for (ptr += BBSIZE; ptr < ((xfs_caddr_t)&iclog->ic_header) + count;
 	     ptr += BBSIZE) {
-		if (be32_to_cpu(*(__be32 *)ptr) == XLOG_HEADER_MAGIC_NUM)
+		if (*(__be32 *)ptr == cpu_to_be32(XLOG_HEADER_MAGIC_NUM))
 			xfs_emerg(log->l_mp, "%s: unexpected magic num",
 				__func__);
 	}
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 04142caedb2..8fe4206de05 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -91,6 +91,8 @@ xlog_get_bp(
 	xlog_t		*log,
 	int		nbblks)
 {
+	struct xfs_buf	*bp;
+
 	if (!xlog_buf_bbcount_valid(log, nbblks)) {
 		xfs_warn(log->l_mp, "Invalid block length (0x%x) for buffer",
 			nbblks);
@@ -118,8 +120,10 @@ xlog_get_bp(
 		nbblks += log->l_sectBBsize;
 	nbblks = round_up(nbblks, log->l_sectBBsize);
 
-	return xfs_buf_get_uncached(log->l_mp->m_logdev_targp,
-					BBTOB(nbblks), 0);
+	bp = xfs_buf_get_uncached(log->l_mp->m_logdev_targp, BBTOB(nbblks), 0);
+	if (bp)
+		xfs_buf_unlock(bp);
+	return bp;
 }
 
 STATIC void
@@ -264,7 +268,7 @@ xlog_bwrite(
 	XFS_BUF_ZEROFLAGS(bp);
 	XFS_BUF_BUSY(bp);
 	XFS_BUF_HOLD(bp);
-	XFS_BUF_PSEMA(bp, PRIBIO);
+	xfs_buf_lock(bp);
 	XFS_BUF_SET_COUNT(bp, BBTOB(nbblks));
 	XFS_BUF_SET_TARGET(bp, log->l_mp->m_logdev_targp);
 
@@ -300,14 +304,14 @@ xlog_header_check_recover(
 	xfs_mount_t		*mp,
 	xlog_rec_header_t	*head)
 {
-	ASSERT(be32_to_cpu(head->h_magicno) == XLOG_HEADER_MAGIC_NUM);
+	ASSERT(head->h_magicno == cpu_to_be32(XLOG_HEADER_MAGIC_NUM));
 
 	/*
 	 * IRIX doesn't write the h_fmt field and leaves it zeroed
 	 * (XLOG_FMT_UNKNOWN). This stops us from trying to recover
 	 * a dirty log created in IRIX.
 	 */
-	if (unlikely(be32_to_cpu(head->h_fmt) != XLOG_FMT)) {
+	if (unlikely(head->h_fmt != cpu_to_be32(XLOG_FMT))) {
 		xfs_warn(mp,
 	"dirty log written in incompatible format - can't recover");
 		xlog_header_check_dump(mp, head);
@@ -333,7 +337,7 @@ xlog_header_check_mount(
 	xfs_mount_t		*mp,
 	xlog_rec_header_t	*head)
 {
-	ASSERT(be32_to_cpu(head->h_magicno) == XLOG_HEADER_MAGIC_NUM);
+	ASSERT(head->h_magicno == cpu_to_be32(XLOG_HEADER_MAGIC_NUM));
 
 	if (uuid_is_nil(&head->h_fs_uuid)) {
 		/*
@@ -367,7 +371,7 @@ xlog_recover_iodone(
 		xfs_force_shutdown(bp->b_target->bt_mount,
 					SHUTDOWN_META_IO_ERROR);
 	}
-	XFS_BUF_CLR_IODONE_FUNC(bp);
+	bp->b_iodone = NULL;
 	xfs_buf_ioend(bp, 0);
 }
 
@@ -534,7 +538,7 @@ xlog_find_verify_log_record(
 
 		head = (xlog_rec_header_t *)offset;
 
-		if (XLOG_HEADER_MAGIC_NUM == be32_to_cpu(head->h_magicno))
+		if (head->h_magicno == cpu_to_be32(XLOG_HEADER_MAGIC_NUM))
 			break;
 
 		if (!smallmem)
@@ -916,7 +920,7 @@ xlog_find_tail(
 		if (error)
 			goto done;
 
-		if (XLOG_HEADER_MAGIC_NUM == be32_to_cpu(*(__be32 *)offset)) {
+		if (*(__be32 *)offset == cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) {
 			found = 1;
 			break;
 		}
@@ -933,8 +937,8 @@ xlog_find_tail(
 			if (error)
 				goto done;
 
-			if (XLOG_HEADER_MAGIC_NUM ==
-			    be32_to_cpu(*(__be32 *)offset)) {
+			if (*(__be32 *)offset ==
+			    cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) {
 				found = 2;
 				break;
 			}
@@ -1947,7 +1951,7 @@ xfs_qm_dqcheck(
 	 * This is all fine; things are still consistent, and we haven't lost
 	 * any quota information. Just don't complain about bad dquot blks.
 	 */
-	if (be16_to_cpu(ddq->d_magic) != XFS_DQUOT_MAGIC) {
+	if (ddq->d_magic != cpu_to_be16(XFS_DQUOT_MAGIC)) {
 		if (flags & XFS_QMOPT_DOWARN)
 			xfs_alert(mp,
 			"%s : XFS dquot ID 0x%x, magic 0x%x != 0x%x",
@@ -2174,7 +2178,7 @@ xlog_recover_buffer_pass2(
 		error = xfs_bwrite(mp, bp);
 	} else {
 		ASSERT(bp->b_target->bt_mount == mp);
-		XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
+		bp->b_iodone = xlog_recover_iodone;
 		xfs_bdwrite(mp, bp);
 	}
 
@@ -2238,7 +2242,7 @@ xlog_recover_inode_pass2(
 	 * Make sure the place we're flushing out to really looks
 	 * like an inode!
 	 */
-	if (unlikely(be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC)) {
+	if (unlikely(dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC))) {
 		xfs_buf_relse(bp);
 		xfs_alert(mp,
 	"%s: Bad inode magic number, dip = 0x%p, dino bp = 0x%p, ino = %Ld",
@@ -2434,7 +2438,7 @@ xlog_recover_inode_pass2(
 
 write_inode_buffer:
 	ASSERT(bp->b_target->bt_mount == mp);
-	XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
+	bp->b_iodone = xlog_recover_iodone;
 	xfs_bdwrite(mp, bp);
 error:
 	if (need_free)
@@ -2556,7 +2560,7 @@ xlog_recover_dquot_pass2(
 
 	ASSERT(dq_f->qlf_size == 2);
 	ASSERT(bp->b_target->bt_mount == mp);
-	XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone);
+	bp->b_iodone = xlog_recover_iodone;
 	xfs_bdwrite(mp, bp);
 
 	return (0);
@@ -3295,7 +3299,7 @@ xlog_valid_rec_header(
 {
 	int			hlen;
 
-	if (unlikely(be32_to_cpu(rhead->h_magicno) != XLOG_HEADER_MAGIC_NUM)) {
+	if (unlikely(rhead->h_magicno != cpu_to_be32(XLOG_HEADER_MAGIC_NUM))) {
 		XFS_ERROR_REPORT("xlog_valid_rec_header(1)",
 				XFS_ERRLEVEL_LOW, log->l_mp);
 		return XFS_ERROR(EFSCORRUPTED);
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index b49b82363d2..7f25245da28 100644
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -348,7 +348,7 @@ xfs_mount_validate_sb(
 	}
 
 	/*
-	 * More sanity checking. These were stolen directly from
+	 * More sanity checking.  Most of these were stolen directly from
 	 * xfs_repair.
 	 */
 	if (unlikely(
@@ -371,23 +371,13 @@ xfs_mount_validate_sb(
 	    (sbp->sb_blocklog - sbp->sb_inodelog != sbp->sb_inopblog)	||
 	    (sbp->sb_rextsize * sbp->sb_blocksize > XFS_MAX_RTEXTSIZE)	||
 	    (sbp->sb_rextsize * sbp->sb_blocksize < XFS_MIN_RTEXTSIZE)	||
-	    (sbp->sb_imax_pct > 100 /* zero sb_imax_pct is valid */))) {
+	    (sbp->sb_imax_pct > 100 /* zero sb_imax_pct is valid */)	||
+	    sbp->sb_dblocks == 0					||
+	    sbp->sb_dblocks > XFS_MAX_DBLOCKS(sbp)			||
+	    sbp->sb_dblocks < XFS_MIN_DBLOCKS(sbp))) {
 		if (loud)
-			xfs_warn(mp, "SB sanity check 1 failed");
-		return XFS_ERROR(EFSCORRUPTED);
-	}
-
-	/*
-	 * Sanity check AG count, size fields against data size field
-	 */
-	if (unlikely(
-	    sbp->sb_dblocks == 0 ||
-	    sbp->sb_dblocks >
-	     (xfs_drfsbno_t)sbp->sb_agcount * sbp->sb_agblocks ||
-	    sbp->sb_dblocks < (xfs_drfsbno_t)(sbp->sb_agcount - 1) *
-			      sbp->sb_agblocks + XFS_MIN_AG_BLOCKS)) {
-		if (loud)
-			xfs_warn(mp, "SB sanity check 2 failed");
+			XFS_CORRUPTION_ERROR("SB sanity check failed",
+				XFS_ERRLEVEL_LOW, mp, sbp);
 		return XFS_ERROR(EFSCORRUPTED);
 	}
 
@@ -864,7 +854,8 @@ xfs_update_alignment(xfs_mount_t *mp)
 		if ((BBTOB(mp->m_dalign) & mp->m_blockmask) ||
 		    (BBTOB(mp->m_swidth) & mp->m_blockmask)) {
 			if (mp->m_flags & XFS_MOUNT_RETERR) {
-				xfs_warn(mp, "alignment check 1 failed");
+				xfs_warn(mp, "alignment check failed: "
+					 "(sunit/swidth vs. blocksize)");
 				return XFS_ERROR(EINVAL);
 			}
 			mp->m_dalign = mp->m_swidth = 0;
@@ -875,6 +866,8 @@ xfs_update_alignment(xfs_mount_t *mp)
 			mp->m_dalign = XFS_BB_TO_FSBT(mp, mp->m_dalign);
 			if (mp->m_dalign && (sbp->sb_agblocks % mp->m_dalign)) {
 				if (mp->m_flags & XFS_MOUNT_RETERR) {
+					xfs_warn(mp, "alignment check failed: "
+						 "(sunit/swidth vs. ag size)");
 					return XFS_ERROR(EINVAL);
 				}
 				xfs_warn(mp,
@@ -889,8 +882,8 @@ xfs_update_alignment(xfs_mount_t *mp)
 				mp->m_swidth = XFS_BB_TO_FSBT(mp, mp->m_swidth);
 			} else {
 				if (mp->m_flags & XFS_MOUNT_RETERR) {
-					xfs_warn(mp,
-		"stripe alignment turned off: sunit(%d) less than bsize(%d)",
+					xfs_warn(mp, "alignment check failed: "
+						"sunit(%d) less than bsize(%d)",
 						mp->m_dalign,
 						mp->m_blockmask +1);
 					return XFS_ERROR(EINVAL);
@@ -1096,10 +1089,6 @@ xfs_mount_reset_sbqflags(
 	if (mp->m_flags & XFS_MOUNT_RDONLY)
 		return 0;
 
-#ifdef QUOTADEBUG
-	xfs_notice(mp, "Writing superblock quota changes");
-#endif
-
 	tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE);
 	error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0,
 				      XFS_DEFAULT_LOG_COUNT);
@@ -1532,7 +1521,7 @@ xfs_unmountfs(
 		xfs_warn(mp, "Unable to free reserved block pool. "
 				"Freespace may not be correct on next mount.");
 
-	error = xfs_log_sbcount(mp, 1);
+	error = xfs_log_sbcount(mp);
 	if (error)
 		xfs_warn(mp, "Unable to update superblock counters. "
 				"Freespace may not be correct on next mount.");
@@ -1568,18 +1557,14 @@ xfs_fs_writable(xfs_mount_t *mp)
 /*
  * xfs_log_sbcount
  *
- * Called either periodically to keep the on disk superblock values
- * roughly up to date or from unmount to make sure the values are
- * correct on a clean unmount.
+ * Sync the superblock counters to disk.
  *
  * Note this code can be called during the process of freezing, so
- * we may need to use the transaction allocator which does not not
+ * we may need to use the transaction allocator which does not
  * block when the transaction subsystem is in its frozen state.
  */
 int
-xfs_log_sbcount(
-	xfs_mount_t	*mp,
-	uint		sync)
+xfs_log_sbcount(xfs_mount_t *mp)
 {
 	xfs_trans_t	*tp;
 	int		error;
@@ -1605,8 +1590,7 @@ xfs_log_sbcount(
 	}
 
 	xfs_mod_sb(tp, XFS_SB_IFREE | XFS_SB_ICOUNT | XFS_SB_FDBLOCKS);
-	if (sync)
-		xfs_trans_set_sync(tp);
+	xfs_trans_set_sync(tp);
 	error = xfs_trans_commit(tp, 0);
 	return error;
 }
@@ -1941,22 +1925,19 @@ unwind:
  * the superblock buffer if it can be locked without sleeping.
  * If it can't then we'll return NULL.
  */
-xfs_buf_t *
+struct xfs_buf *
 xfs_getsb(
-	xfs_mount_t	*mp,
-	int		flags)
+	struct xfs_mount	*mp,
+	int			flags)
 {
-	xfs_buf_t	*bp;
+	struct xfs_buf		*bp = mp->m_sb_bp;
 
-	ASSERT(mp->m_sb_bp != NULL);
-	bp = mp->m_sb_bp;
-	if (flags & XBF_TRYLOCK) {
-		if (!XFS_BUF_CPSEMA(bp)) {
+	if (!xfs_buf_trylock(bp)) {
+		if (flags & XBF_TRYLOCK)
 			return NULL;
-		}
-	} else {
-		XFS_BUF_PSEMA(bp, PRIBIO);
+		xfs_buf_lock(bp);
 	}
+
 	XFS_BUF_HOLD(bp);
 	ASSERT(XFS_BUF_ISDONE(bp));
 	return bp;
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 3d68bb267c5..bb24dac42a2 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -371,7 +371,7 @@ typedef struct xfs_mod_sb {
 	int64_t		msb_delta;	/* Change to make to specified field */
 } xfs_mod_sb_t;
 
-extern int	xfs_log_sbcount(xfs_mount_t *, uint);
+extern int	xfs_log_sbcount(xfs_mount_t *);
 extern __uint64_t xfs_default_resblks(xfs_mount_t *mp);
 extern int	xfs_mountfs(xfs_mount_t *mp);
 
diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index 7c7bc2b786b..efc147f0e9b 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -1361,7 +1361,7 @@ xfs_trans_item_committed(
 		lip->li_flags |= XFS_LI_ABORTED;
 	item_lsn = IOP_COMMITTED(lip, commit_lsn);
 
-	/* If the committed routine returns -1, item has been freed. */
+	/* item_lsn of -1 means the item needs no further processing */
 	if (XFS_LSN_CMP(item_lsn, (xfs_lsn_t)-1) == 0)
 		return;
 
@@ -1426,6 +1426,7 @@ xfs_trans_committed(
 static inline void
 xfs_log_item_batch_insert(
 	struct xfs_ail		*ailp,
+	struct xfs_ail_cursor	*cur,
 	struct xfs_log_item	**log_items,
 	int			nr_items,
 	xfs_lsn_t		commit_lsn)
@@ -1434,7 +1435,7 @@ xfs_log_item_batch_insert(
 
 	spin_lock(&ailp->xa_lock);
 	/* xfs_trans_ail_update_bulk drops ailp->xa_lock */
-	xfs_trans_ail_update_bulk(ailp, log_items, nr_items, commit_lsn);
+	xfs_trans_ail_update_bulk(ailp, cur, log_items, nr_items, commit_lsn);
 
 	for (i = 0; i < nr_items; i++)
 		IOP_UNPIN(log_items[i], 0);
@@ -1452,6 +1453,13 @@ xfs_log_item_batch_insert(
  * as an iclog write error even though we haven't started any IO yet. Hence in
  * this case all we need to do is IOP_COMMITTED processing, followed by an
  * IOP_UNPIN(aborted) call.
+ *
+ * The AIL cursor is used to optimise the insert process. If commit_lsn is not
+ * at the end of the AIL, the insert cursor avoids the need to walk
+ * the AIL to find the insertion point on every xfs_log_item_batch_insert()
+ * call. This saves a lot of needless list walking and is a net win, even
+ * though it slightly increases that amount of AIL lock traffic to set it up
+ * and tear it down.
  */
 void
 xfs_trans_committed_bulk(
@@ -1463,8 +1471,13 @@ xfs_trans_committed_bulk(
 #define LOG_ITEM_BATCH_SIZE	32
 	struct xfs_log_item	*log_items[LOG_ITEM_BATCH_SIZE];
 	struct xfs_log_vec	*lv;
+	struct xfs_ail_cursor	cur;
 	int			i = 0;
 
+	spin_lock(&ailp->xa_lock);
+	xfs_trans_ail_cursor_last(ailp, &cur, commit_lsn);
+	spin_unlock(&ailp->xa_lock);
+
 	/* unpin all the log items */
 	for (lv = log_vector; lv; lv = lv->lv_next ) {
 		struct xfs_log_item	*lip = lv->lv_item;
@@ -1474,7 +1487,7 @@ xfs_trans_committed_bulk(
 			lip->li_flags |= XFS_LI_ABORTED;
 		item_lsn = IOP_COMMITTED(lip, commit_lsn);
 
-		/* item_lsn of -1 means the item was freed */
+		/* item_lsn of -1 means the item needs no further processing */
 		if (XFS_LSN_CMP(item_lsn, (xfs_lsn_t)-1) == 0)
 			continue;
 
@@ -1493,7 +1506,9 @@ xfs_trans_committed_bulk(
 			/*
 			 * Not a bulk update option due to unusual item_lsn.
 			 * Push into AIL immediately, rechecking the lsn once
-			 * we have the ail lock. Then unpin the item.
+			 * we have the ail lock. Then unpin the item. This does
+			 * not affect the AIL cursor the bulk insert path is
+			 * using.
 			 */
 			spin_lock(&ailp->xa_lock);
 			if (XFS_LSN_CMP(item_lsn, lip->li_lsn) > 0)
@@ -1507,7 +1522,7 @@ xfs_trans_committed_bulk(
 		/* Item is a candidate for bulk AIL insert.  */
 		log_items[i++] = lv->lv_item;
 		if (i >= LOG_ITEM_BATCH_SIZE) {
-			xfs_log_item_batch_insert(ailp, log_items,
+			xfs_log_item_batch_insert(ailp, &cur, log_items,
 					LOG_ITEM_BATCH_SIZE, commit_lsn);
 			i = 0;
 		}
@@ -1515,7 +1530,11 @@ xfs_trans_committed_bulk(
 
 	/* make sure we insert the remainder! */
 	if (i)
-		xfs_log_item_batch_insert(ailp, log_items, i, commit_lsn);
+		xfs_log_item_batch_insert(ailp, &cur, log_items, i, commit_lsn);
+
+	spin_lock(&ailp->xa_lock);
+	xfs_trans_ail_cursor_done(ailp, &cur);
+	spin_unlock(&ailp->xa_lock);
 }
 
 /*
diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c
index 5fc2380092c..43233e92f0f 100644
--- a/fs/xfs/xfs_trans_ail.c
+++ b/fs/xfs/xfs_trans_ail.c
@@ -163,17 +163,11 @@ xfs_ail_max_lsn(
 }
 
 /*
- * AIL traversal cursor initialisation.
- *
- * The cursor keeps track of where our current traversal is up
- * to by tracking the next ƣtem in the list for us. However, for
- * this to be safe, removing an object from the AIL needs to invalidate
- * any cursor that points to it. hence the traversal cursor needs to
- * be linked to the struct xfs_ail so that deletion can search all the
- * active cursors for invalidation.
- *
- * We don't link the push cursor because it is embedded in the struct
- * xfs_ail and hence easily findable.
+ * The cursor keeps track of where our current traversal is up to by tracking
+ * the next item in the list for us. However, for this to be safe, removing an
+ * object from the AIL needs to invalidate any cursor that points to it. hence
+ * the traversal cursor needs to be linked to the struct xfs_ail so that
+ * deletion can search all the active cursors for invalidation.
  */
 STATIC void
 xfs_trans_ail_cursor_init(
@@ -181,31 +175,12 @@ xfs_trans_ail_cursor_init(
 	struct xfs_ail_cursor	*cur)
 {
 	cur->item = NULL;
-	if (cur == &ailp->xa_cursors)
-		return;
-
-	cur->next = ailp->xa_cursors.next;
-	ailp->xa_cursors.next = cur;
-}
-
-/*
- * Set the cursor to the next item, because when we look
- * up the cursor the current item may have been freed.
- */
-STATIC void
-xfs_trans_ail_cursor_set(
-	struct xfs_ail		*ailp,
-	struct xfs_ail_cursor	*cur,
-	struct xfs_log_item	*lip)
-{
-	if (lip)
-		cur->item = xfs_ail_next(ailp, lip);
+	list_add_tail(&cur->list, &ailp->xa_cursors);
 }
 
 /*
- * Get the next item in the traversal and advance the cursor.
- * If the cursor was invalidated (inidicated by a lip of 1),
- * restart the traversal.
+ * Get the next item in the traversal and advance the cursor.  If the cursor
+ * was invalidated (indicated by a lip of 1), restart the traversal.
  */
 struct xfs_log_item *
 xfs_trans_ail_cursor_next(
@@ -216,45 +191,31 @@ xfs_trans_ail_cursor_next(
 
 	if ((__psint_t)lip & 1)
 		lip = xfs_ail_min(ailp);
-	xfs_trans_ail_cursor_set(ailp, cur, lip);
+	if (lip)
+		cur->item = xfs_ail_next(ailp, lip);
 	return lip;
 }
 
 /*
- * Now that the traversal is complete, we need to remove the cursor
- * from the list of traversing cursors. Avoid removing the embedded
- * push cursor, but use the fact it is always present to make the
- * list deletion simple.
+ * When the traversal is complete, we need to remove the cursor from the list
+ * of traversing cursors.
  */
 void
 xfs_trans_ail_cursor_done(
 	struct xfs_ail		*ailp,
-	struct xfs_ail_cursor	*done)
+	struct xfs_ail_cursor	*cur)
 {
-	struct xfs_ail_cursor	*prev = NULL;
-	struct xfs_ail_cursor	*cur;
-
-	done->item = NULL;
-	if (done == &ailp->xa_cursors)
-		return;
-	prev = &ailp->xa_cursors;
-	for (cur = prev->next; cur; prev = cur, cur = prev->next) {
-		if (cur == done) {
-			prev->next = cur->next;
-			break;
-		}
-	}
-	ASSERT(cur);
+	cur->item = NULL;
+	list_del_init(&cur->list);
 }
 
 /*
- * Invalidate any cursor that is pointing to this item. This is
- * called when an item is removed from the AIL. Any cursor pointing
- * to this object is now invalid and the traversal needs to be
- * terminated so it doesn't reference a freed object. We set the
- * cursor item to a value of 1 so we can distinguish between an
- * invalidation and the end of the list when getting the next item
- * from the cursor.
+ * Invalidate any cursor that is pointing to this item. This is called when an
+ * item is removed from the AIL. Any cursor pointing to this object is now
+ * invalid and the traversal needs to be terminated so it doesn't reference a
+ * freed object. We set the low bit of the cursor item pointer so we can
+ * distinguish between an invalidation and the end of the list when getting the
+ * next item from the cursor.
  */
 STATIC void
 xfs_trans_ail_cursor_clear(
@@ -263,8 +224,7 @@ xfs_trans_ail_cursor_clear(
 {
 	struct xfs_ail_cursor	*cur;
 
-	/* need to search all cursors */
-	for (cur = &ailp->xa_cursors; cur; cur = cur->next) {
+	list_for_each_entry(cur, &ailp->xa_cursors, list) {
 		if (cur->item == lip)
 			cur->item = (struct xfs_log_item *)
 					((__psint_t)cur->item | 1);
@@ -272,9 +232,10 @@ xfs_trans_ail_cursor_clear(
 }
 
 /*
- * Return the item in the AIL with the current lsn.
- * Return the current tree generation number for use
- * in calls to xfs_trans_next_ail().
+ * Find the first item in the AIL with the given @lsn by searching in ascending
+ * LSN order and initialise the cursor to point to the next item for a
+ * ascending traversal.  Pass a @lsn of zero to initialise the cursor to the
+ * first item in the AIL. Returns NULL if the list is empty.
  */
 xfs_log_item_t *
 xfs_trans_ail_cursor_first(
@@ -285,46 +246,112 @@ xfs_trans_ail_cursor_first(
 	xfs_log_item_t		*lip;
 
 	xfs_trans_ail_cursor_init(ailp, cur);
-	lip = xfs_ail_min(ailp);
-	if (lsn == 0)
+
+	if (lsn == 0) {
+		lip = xfs_ail_min(ailp);
 		goto out;
+	}
 
 	list_for_each_entry(lip, &ailp->xa_ail, li_ail) {
 		if (XFS_LSN_CMP(lip->li_lsn, lsn) >= 0)
 			goto out;
 	}
-	lip = NULL;
+	return NULL;
+
 out:
-	xfs_trans_ail_cursor_set(ailp, cur, lip);
+	if (lip)
+		cur->item = xfs_ail_next(ailp, lip);
 	return lip;
 }
 
+static struct xfs_log_item *
+__xfs_trans_ail_cursor_last(
+	struct xfs_ail		*ailp,
+	xfs_lsn_t		lsn)
+{
+	xfs_log_item_t		*lip;
+
+	list_for_each_entry_reverse(lip, &ailp->xa_ail, li_ail) {
+		if (XFS_LSN_CMP(lip->li_lsn, lsn) <= 0)
+			return lip;
+	}
+	return NULL;
+}
+
+/*
+ * Find the last item in the AIL with the given @lsn by searching in descending
+ * LSN order and initialise the cursor to point to that item.  If there is no
+ * item with the value of @lsn, then it sets the cursor to the last item with an
+ * LSN lower than @lsn.  Returns NULL if the list is empty.
+ */
+struct xfs_log_item *
+xfs_trans_ail_cursor_last(
+	struct xfs_ail		*ailp,
+	struct xfs_ail_cursor	*cur,
+	xfs_lsn_t		lsn)
+{
+	xfs_trans_ail_cursor_init(ailp, cur);
+	cur->item = __xfs_trans_ail_cursor_last(ailp, lsn);
+	return cur->item;
+}
+
 /*
- * splice the log item list into the AIL at the given LSN.
+ * Splice the log item list into the AIL at the given LSN. We splice to the
+ * tail of the given LSN to maintain insert order for push traversals. The
+ * cursor is optional, allowing repeated updates to the same LSN to avoid
+ * repeated traversals.
  */
 static void
 xfs_ail_splice(
-	struct xfs_ail  *ailp,
-	struct list_head *list,
-	xfs_lsn_t       lsn)
+	struct xfs_ail		*ailp,
+	struct xfs_ail_cursor	*cur,
+	struct list_head	*list,
+	xfs_lsn_t		lsn)
 {
-	xfs_log_item_t  *next_lip;
+	struct xfs_log_item	*lip = cur ? cur->item : NULL;
+	struct xfs_log_item	*next_lip;
 
-	/* If the list is empty, just insert the item.  */
-	if (list_empty(&ailp->xa_ail)) {
-		list_splice(list, &ailp->xa_ail);
-		return;
+	/*
+	 * Get a new cursor if we don't have a placeholder or the existing one
+	 * has been invalidated.
+	 */
+	if (!lip || (__psint_t)lip & 1) {
+		lip = __xfs_trans_ail_cursor_last(ailp, lsn);
+
+		if (!lip) {
+			/* The list is empty, so just splice and return.  */
+			if (cur)
+				cur->item = NULL;
+			list_splice(list, &ailp->xa_ail);
+			return;
+		}
 	}
 
-	list_for_each_entry_reverse(next_lip, &ailp->xa_ail, li_ail) {
-		if (XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0)
-			break;
+	/*
+	 * Our cursor points to the item we want to insert _after_, so we have
+	 * to update the cursor to point to the end of the list we are splicing
+	 * in so that it points to the correct location for the next splice.
+	 * i.e. before the splice
+	 *
+	 *  lsn -> lsn -> lsn + x -> lsn + x ...
+	 *          ^
+	 *          | cursor points here
+	 *
+	 * After the splice we have:
+	 *
+	 *  lsn -> lsn -> lsn -> lsn -> .... -> lsn -> lsn + x -> lsn + x ...
+	 *          ^                            ^
+	 *          | cursor points here         | needs to move here
+	 *
+	 * So we set the cursor to the last item in the list to be spliced
+	 * before we execute the splice, resulting in the cursor pointing to
+	 * the correct item after the splice occurs.
+	 */
+	if (cur) {
+		next_lip = list_entry(list->prev, struct xfs_log_item, li_ail);
+		cur->item = next_lip;
 	}
-
-	ASSERT(&next_lip->li_ail == &ailp->xa_ail ||
-	       XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0);
-
-	list_splice_init(list, &next_lip->li_ail);
+	list_splice(list, &lip->li_ail);
 }
 
 /*
@@ -351,7 +378,7 @@ xfs_ail_worker(
 	struct xfs_ail		*ailp = container_of(to_delayed_work(work),
 					struct xfs_ail, xa_work);
 	xfs_mount_t		*mp = ailp->xa_mount;
-	struct xfs_ail_cursor	*cur = &ailp->xa_cursors;
+	struct xfs_ail_cursor	cur;
 	xfs_log_item_t		*lip;
 	xfs_lsn_t		lsn;
 	xfs_lsn_t		target;
@@ -363,13 +390,12 @@ xfs_ail_worker(
 
 	spin_lock(&ailp->xa_lock);
 	target = ailp->xa_target;
-	xfs_trans_ail_cursor_init(ailp, cur);
-	lip = xfs_trans_ail_cursor_first(ailp, cur, ailp->xa_last_pushed_lsn);
+	lip = xfs_trans_ail_cursor_first(ailp, &cur, ailp->xa_last_pushed_lsn);
 	if (!lip || XFS_FORCED_SHUTDOWN(mp)) {
 		/*
 		 * AIL is empty or our push has reached the end.
 		 */
-		xfs_trans_ail_cursor_done(ailp, cur);
+		xfs_trans_ail_cursor_done(ailp, &cur);
 		spin_unlock(&ailp->xa_lock);
 		goto out_done;
 	}
@@ -457,12 +483,12 @@ xfs_ail_worker(
 		if (stuck > 100)
 			break;
 
-		lip = xfs_trans_ail_cursor_next(ailp, cur);
+		lip = xfs_trans_ail_cursor_next(ailp, &cur);
 		if (lip == NULL)
 			break;
 		lsn = lip->li_lsn;
 	}
-	xfs_trans_ail_cursor_done(ailp, cur);
+	xfs_trans_ail_cursor_done(ailp, &cur);
 	spin_unlock(&ailp->xa_lock);
 
 	if (flush_log) {
@@ -645,6 +671,7 @@ xfs_trans_unlocked_item(
 void
 xfs_trans_ail_update_bulk(
 	struct xfs_ail		*ailp,
+	struct xfs_ail_cursor	*cur,
 	struct xfs_log_item	**log_items,
 	int			nr_items,
 	xfs_lsn_t		lsn) __releases(ailp->xa_lock)
@@ -674,7 +701,7 @@ xfs_trans_ail_update_bulk(
 		list_add(&lip->li_ail, &tmp);
 	}
 
-	xfs_ail_splice(ailp, &tmp, lsn);
+	xfs_ail_splice(ailp, cur, &tmp, lsn);
 
 	if (!mlip_changed) {
 		spin_unlock(&ailp->xa_lock);
@@ -793,6 +820,7 @@ xfs_trans_ail_init(
 
 	ailp->xa_mount = mp;
 	INIT_LIST_HEAD(&ailp->xa_ail);
+	INIT_LIST_HEAD(&ailp->xa_cursors);
 	spin_lock_init(&ailp->xa_lock);
 	INIT_DELAYED_WORK(&ailp->xa_work, xfs_ail_worker);
 	mp->m_ail = ailp;
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index 03b3b7f85a3..15584fc3ed7 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -81,7 +81,7 @@ _xfs_trans_bjoin(
 	struct xfs_buf_log_item	*bip;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, void *) == NULL);
+	ASSERT(bp->b_transp == NULL);
 
 	/*
 	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
@@ -89,7 +89,7 @@ _xfs_trans_bjoin(
 	 * 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 *);
+	bip = bp->b_fspriv;
 	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
 	ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
 	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
@@ -110,7 +110,7 @@ _xfs_trans_bjoin(
 	 * 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);
+	bp->b_transp = tp;
 
 }
 
@@ -160,7 +160,7 @@ xfs_trans_get_buf(xfs_trans_t	*tp,
 	 */
 	bp = xfs_trans_buf_item_match(tp, target_dev, blkno, len);
 	if (bp != NULL) {
-		ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
+		ASSERT(xfs_buf_islocked(bp));
 		if (XFS_FORCED_SHUTDOWN(tp->t_mountp))
 			XFS_BUF_SUPER_STALE(bp);
 
@@ -172,8 +172,8 @@ xfs_trans_get_buf(xfs_trans_t	*tp,
 		else if (XFS_BUF_ISSTALE(bp))
 			ASSERT(!XFS_BUF_ISDELAYWRITE(bp));
 
-		ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-		bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+		ASSERT(bp->b_transp == tp);
+		bip = bp->b_fspriv;
 		ASSERT(bip != NULL);
 		ASSERT(atomic_read(&bip->bli_refcount) > 0);
 		bip->bli_recur++;
@@ -232,8 +232,8 @@ xfs_trans_getsb(xfs_trans_t	*tp,
 	 * recursion count and return the buffer to the caller.
 	 */
 	bp = mp->m_sb_bp;
-	if (XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp) {
-		bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
+	if (bp->b_transp == tp) {
+		bip = bp->b_fspriv;
 		ASSERT(bip != NULL);
 		ASSERT(atomic_read(&bip->bli_refcount) > 0);
 		bip->bli_recur++;
@@ -327,9 +327,9 @@ xfs_trans_read_buf(
 	 */
 	bp = xfs_trans_buf_item_match(tp, target, blkno, len);
 	if (bp != NULL) {
-		ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
-		ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-		ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
+		ASSERT(xfs_buf_islocked(bp));
+		ASSERT(bp->b_transp == tp);
+		ASSERT(bp->b_fspriv != NULL);
 		ASSERT((XFS_BUF_ISERROR(bp)) == 0);
 		if (!(XFS_BUF_ISDONE(bp))) {
 			trace_xfs_trans_read_buf_io(bp, _RET_IP_);
@@ -363,7 +363,7 @@ xfs_trans_read_buf(
 		}
 
 
-		bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
+		bip = bp->b_fspriv;
 		bip->bli_recur++;
 
 		ASSERT(atomic_read(&bip->bli_refcount) > 0);
@@ -460,32 +460,30 @@ xfs_trans_brelse(xfs_trans_t	*tp,
 		 xfs_buf_t	*bp)
 {
 	xfs_buf_log_item_t	*bip;
-	xfs_log_item_t		*lip;
 
 	/*
 	 * Default to a normal brelse() call if the tp is NULL.
 	 */
 	if (tp == NULL) {
-		ASSERT(XFS_BUF_FSPRIVATE2(bp, void *) == NULL);
+		struct xfs_log_item	*lip = bp->b_fspriv;
+
+		ASSERT(bp->b_transp == NULL);
+
 		/*
 		 * If there's a buf log item attached to the buffer,
 		 * then let the AIL know that the buffer is being
 		 * unlocked.
 		 */
-		if (XFS_BUF_FSPRIVATE(bp, void *) != NULL) {
-			lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
-			if (lip->li_type == XFS_LI_BUF) {
-				bip = XFS_BUF_FSPRIVATE(bp,xfs_buf_log_item_t*);
-				xfs_trans_unlocked_item(bip->bli_item.li_ailp,
-							lip);
-			}
+		if (lip != NULL && lip->li_type == XFS_LI_BUF) {
+			bip = bp->b_fspriv;
+			xfs_trans_unlocked_item(bip->bli_item.li_ailp, lip);
 		}
 		xfs_buf_relse(bp);
 		return;
 	}
 
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(bp->b_transp == tp);
+	bip = bp->b_fspriv;
 	ASSERT(bip->bli_item.li_type == XFS_LI_BUF);
 	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
 	ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
@@ -556,7 +554,7 @@ xfs_trans_brelse(xfs_trans_t	*tp,
 		xfs_buf_item_relse(bp);
 		bip = NULL;
 	}
-	XFS_BUF_SET_FSPRIVATE2(bp, NULL);
+	bp->b_transp = NULL;
 
 	/*
 	 * If we've still got a buf log item on the buffer, then
@@ -581,16 +579,15 @@ void
 xfs_trans_bhold(xfs_trans_t	*tp,
 		xfs_buf_t	*bp)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
-
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
 	ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
+
 	bip->bli_flags |= XFS_BLI_HOLD;
 	trace_xfs_trans_bhold(bip);
 }
@@ -603,19 +600,17 @@ void
 xfs_trans_bhold_release(xfs_trans_t	*tp,
 			xfs_buf_t	*bp)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
-
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
 	ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL));
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 	ASSERT(bip->bli_flags & XFS_BLI_HOLD);
-	bip->bli_flags &= ~XFS_BLI_HOLD;
 
+	bip->bli_flags &= ~XFS_BLI_HOLD;
 	trace_xfs_trans_bhold_release(bip);
 }
 
@@ -634,14 +629,14 @@ xfs_trans_log_buf(xfs_trans_t	*tp,
 		  uint		first,
 		  uint		last)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT((first <= last) && (last < XFS_BUF_COUNT(bp)));
-	ASSERT((XFS_BUF_IODONE_FUNC(bp) == NULL) ||
-	       (XFS_BUF_IODONE_FUNC(bp) == xfs_buf_iodone_callbacks));
+	ASSERT(bp->b_iodone == NULL ||
+	       bp->b_iodone == xfs_buf_iodone_callbacks);
 
 	/*
 	 * Mark the buffer as needing to be written out eventually,
@@ -656,9 +651,8 @@ xfs_trans_log_buf(xfs_trans_t	*tp,
 	XFS_BUF_DELAYWRITE(bp);
 	XFS_BUF_DONE(bp);
 
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
-	XFS_BUF_SET_IODONE_FUNC(bp, xfs_buf_iodone_callbacks);
+	bp->b_iodone = xfs_buf_iodone_callbacks;
 	bip->bli_item.li_cb = xfs_buf_iodone;
 
 	trace_xfs_trans_log_buf(bip);
@@ -706,13 +700,11 @@ xfs_trans_binval(
 	xfs_trans_t	*tp,
 	xfs_buf_t	*bp)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
-
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 
 	trace_xfs_trans_binval(bip);
@@ -780,13 +772,11 @@ xfs_trans_inode_buf(
 	xfs_trans_t	*tp,
 	xfs_buf_t	*bp)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
-
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 
 	bip->bli_flags |= XFS_BLI_INODE_BUF;
@@ -806,13 +796,11 @@ xfs_trans_stale_inode_buf(
 	xfs_trans_t	*tp,
 	xfs_buf_t	*bp)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
-
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 
 	bip->bli_flags |= XFS_BLI_STALE_INODE;
@@ -833,13 +821,11 @@ xfs_trans_inode_alloc_buf(
 	xfs_trans_t	*tp,
 	xfs_buf_t	*bp)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
-
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 
 	bip->bli_flags |= XFS_BLI_INODE_ALLOC_BUF;
@@ -863,16 +849,14 @@ xfs_trans_dquot_buf(
 	xfs_buf_t	*bp,
 	uint		type)
 {
-	xfs_buf_log_item_t	*bip;
+	xfs_buf_log_item_t	*bip = bp->b_fspriv;
 
 	ASSERT(XFS_BUF_ISBUSY(bp));
-	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
-	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
+	ASSERT(bp->b_transp == tp);
+	ASSERT(bip != NULL);
 	ASSERT(type == XFS_BLF_UDQUOT_BUF ||
 	       type == XFS_BLF_PDQUOT_BUF ||
 	       type == XFS_BLF_GDQUOT_BUF);
-
-	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
 	ASSERT(atomic_read(&bip->bli_refcount) > 0);
 
 	bip->bli_format.blf_flags |= type;
diff --git a/fs/xfs/xfs_trans_inode.c b/fs/xfs/xfs_trans_inode.c
index 048b0c689d3..c8dea2fd7e6 100644
--- a/fs/xfs/xfs_trans_inode.c
+++ b/fs/xfs/xfs_trans_inode.c
@@ -55,7 +55,6 @@ xfs_trans_ijoin(
 {
 	xfs_inode_log_item_t	*iip;
 
-	ASSERT(ip->i_transp == NULL);
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
 	if (ip->i_itemp == NULL)
 		xfs_inode_item_init(ip, ip->i_mount);
@@ -68,12 +67,6 @@ xfs_trans_ijoin(
 	xfs_trans_add_item(tp, &iip->ili_item);
 
 	xfs_trans_inode_broot_debug(ip);
-
-	/*
-	 * Initialize i_transp so we can find it with xfs_inode_incore()
-	 * in xfs_trans_iget() above.
-	 */
-	ip->i_transp = tp;
 }
 
 /*
@@ -111,7 +104,6 @@ xfs_trans_ichgtime(
 
 	ASSERT(tp);
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
-	ASSERT(ip->i_transp == tp);
 
 	tv = current_fs_time(inode->i_sb);
 
@@ -140,7 +132,6 @@ xfs_trans_log_inode(
 	xfs_inode_t	*ip,
 	uint		flags)
 {
-	ASSERT(ip->i_transp == tp);
 	ASSERT(ip->i_itemp != NULL);
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
 
diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h
index 6b164e9e9a1..212946b9723 100644
--- a/fs/xfs/xfs_trans_priv.h
+++ b/fs/xfs/xfs_trans_priv.h
@@ -53,7 +53,7 @@ void	xfs_trans_committed_bulk(struct xfs_ail *ailp, struct xfs_log_vec *lv,
  * of the list to trigger traversal restarts.
  */
 struct xfs_ail_cursor {
-	struct xfs_ail_cursor	*next;
+	struct list_head	list;
 	struct xfs_log_item	*item;
 };
 
@@ -66,7 +66,7 @@ struct xfs_ail {
 	struct xfs_mount	*xa_mount;
 	struct list_head	xa_ail;
 	xfs_lsn_t		xa_target;
-	struct xfs_ail_cursor	xa_cursors;
+	struct list_head	xa_cursors;
 	spinlock_t		xa_lock;
 	struct delayed_work	xa_work;
 	xfs_lsn_t		xa_last_pushed_lsn;
@@ -82,6 +82,7 @@ struct xfs_ail {
 extern struct workqueue_struct	*xfs_ail_wq;	/* AIL workqueue */
 
 void	xfs_trans_ail_update_bulk(struct xfs_ail *ailp,
+				struct xfs_ail_cursor *cur,
 				struct xfs_log_item **log_items, int nr_items,
 				xfs_lsn_t lsn) __releases(ailp->xa_lock);
 static inline void
@@ -90,7 +91,7 @@ xfs_trans_ail_update(
 	struct xfs_log_item	*lip,
 	xfs_lsn_t		lsn) __releases(ailp->xa_lock)
 {
-	xfs_trans_ail_update_bulk(ailp, &lip, 1, lsn);
+	xfs_trans_ail_update_bulk(ailp, NULL, &lip, 1, lsn);
 }
 
 void	xfs_trans_ail_delete_bulk(struct xfs_ail *ailp,
@@ -111,10 +112,13 @@ xfs_lsn_t		xfs_ail_min_lsn(struct xfs_ail *ailp);
 void			xfs_trans_unlocked_item(struct xfs_ail *,
 					xfs_log_item_t *);
 
-struct xfs_log_item	*xfs_trans_ail_cursor_first(struct xfs_ail *ailp,
+struct xfs_log_item *	xfs_trans_ail_cursor_first(struct xfs_ail *ailp,
 					struct xfs_ail_cursor *cur,
 					xfs_lsn_t lsn);
-struct xfs_log_item	*xfs_trans_ail_cursor_next(struct xfs_ail *ailp,
+struct xfs_log_item *	xfs_trans_ail_cursor_last(struct xfs_ail *ailp,
+					struct xfs_ail_cursor *cur,
+					xfs_lsn_t lsn);
+struct xfs_log_item *	xfs_trans_ail_cursor_next(struct xfs_ail *ailp,
 					struct xfs_ail_cursor *cur);
 void			xfs_trans_ail_cursor_done(struct xfs_ail *ailp,
 					struct xfs_ail_cursor *cur);
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index b7a5fe7c52c..88d121486c5 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -50,430 +50,6 @@
 #include "xfs_vnodeops.h"
 #include "xfs_trace.h"
 
-int
-xfs_setattr(
-	struct xfs_inode	*ip,
-	struct iattr		*iattr,
-	int			flags)
-{
-	xfs_mount_t		*mp = ip->i_mount;
-	struct inode		*inode = VFS_I(ip);
-	int			mask = iattr->ia_valid;
-	xfs_trans_t		*tp;
-	int			code;
-	uint			lock_flags;
-	uint			commit_flags=0;
-	uid_t			uid=0, iuid=0;
-	gid_t			gid=0, igid=0;
-	struct xfs_dquot	*udqp, *gdqp, *olddquot1, *olddquot2;
-	int			need_iolock = 1;
-
-	trace_xfs_setattr(ip);
-
-	if (mp->m_flags & XFS_MOUNT_RDONLY)
-		return XFS_ERROR(EROFS);
-
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return XFS_ERROR(EIO);
-
-	code = -inode_change_ok(inode, iattr);
-	if (code)
-		return code;
-
-	olddquot1 = olddquot2 = NULL;
-	udqp = gdqp = NULL;
-
-	/*
-	 * If disk quotas is on, we make sure that the dquots do exist on disk,
-	 * before we start any other transactions. Trying to do this later
-	 * is messy. We don't care to take a readlock to look at the ids
-	 * in inode here, because we can't hold it across the trans_reserve.
-	 * If the IDs do change before we take the ilock, we're covered
-	 * because the i_*dquot fields will get updated anyway.
-	 */
-	if (XFS_IS_QUOTA_ON(mp) && (mask & (ATTR_UID|ATTR_GID))) {
-		uint	qflags = 0;
-
-		if ((mask & ATTR_UID) && XFS_IS_UQUOTA_ON(mp)) {
-			uid = iattr->ia_uid;
-			qflags |= XFS_QMOPT_UQUOTA;
-		} else {
-			uid = ip->i_d.di_uid;
-		}
-		if ((mask & ATTR_GID) && XFS_IS_GQUOTA_ON(mp)) {
-			gid = iattr->ia_gid;
-			qflags |= XFS_QMOPT_GQUOTA;
-		}  else {
-			gid = ip->i_d.di_gid;
-		}
-
-		/*
-		 * We take a reference when we initialize udqp and gdqp,
-		 * so it is important that we never blindly double trip on
-		 * the same variable. See xfs_create() for an example.
-		 */
-		ASSERT(udqp == NULL);
-		ASSERT(gdqp == NULL);
-		code = xfs_qm_vop_dqalloc(ip, uid, gid, xfs_get_projid(ip),
-					 qflags, &udqp, &gdqp);
-		if (code)
-			return code;
-	}
-
-	/*
-	 * For the other attributes, we acquire the inode lock and
-	 * first do an error checking pass.
-	 */
-	tp = NULL;
-	lock_flags = XFS_ILOCK_EXCL;
-	if (flags & XFS_ATTR_NOLOCK)
-		need_iolock = 0;
-	if (!(mask & ATTR_SIZE)) {
-		tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
-		commit_flags = 0;
-		code = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp),
-					 0, 0, 0);
-		if (code) {
-			lock_flags = 0;
-			goto error_return;
-		}
-	} else {
-		if (need_iolock)
-			lock_flags |= XFS_IOLOCK_EXCL;
-	}
-
-	xfs_ilock(ip, lock_flags);
-
-	/*
-	 * Change file ownership.  Must be the owner or privileged.
-	 */
-	if (mask & (ATTR_UID|ATTR_GID)) {
-		/*
-		 * These IDs could have changed since we last looked at them.
-		 * But, we're assured that if the ownership did change
-		 * while we didn't have the inode locked, inode's dquot(s)
-		 * would have changed also.
-		 */
-		iuid = ip->i_d.di_uid;
-		igid = ip->i_d.di_gid;
-		gid = (mask & ATTR_GID) ? iattr->ia_gid : igid;
-		uid = (mask & ATTR_UID) ? iattr->ia_uid : iuid;
-
-		/*
-		 * Do a quota reservation only if uid/gid is actually
-		 * going to change.
-		 */
-		if (XFS_IS_QUOTA_RUNNING(mp) &&
-		    ((XFS_IS_UQUOTA_ON(mp) && iuid != uid) ||
-		     (XFS_IS_GQUOTA_ON(mp) && igid != gid))) {
-			ASSERT(tp);
-			code = xfs_qm_vop_chown_reserve(tp, ip, udqp, gdqp,
-						capable(CAP_FOWNER) ?
-						XFS_QMOPT_FORCE_RES : 0);
-			if (code)	/* out of quota */
-				goto error_return;
-		}
-	}
-
-	/*
-	 * Truncate file.  Must have write permission and not be a directory.
-	 */
-	if (mask & ATTR_SIZE) {
-		/* Short circuit the truncate case for zero length files */
-		if (iattr->ia_size == 0 &&
-		    ip->i_size == 0 && ip->i_d.di_nextents == 0) {
-			xfs_iunlock(ip, XFS_ILOCK_EXCL);
-			lock_flags &= ~XFS_ILOCK_EXCL;
-			if (mask & ATTR_CTIME) {
-				inode->i_mtime = inode->i_ctime =
-						current_fs_time(inode->i_sb);
-				xfs_mark_inode_dirty_sync(ip);
-			}
-			code = 0;
-			goto error_return;
-		}
-
-		if (S_ISDIR(ip->i_d.di_mode)) {
-			code = XFS_ERROR(EISDIR);
-			goto error_return;
-		} else if (!S_ISREG(ip->i_d.di_mode)) {
-			code = XFS_ERROR(EINVAL);
-			goto error_return;
-		}
-
-		/*
-		 * Make sure that the dquots are attached to the inode.
-		 */
-		code = xfs_qm_dqattach_locked(ip, 0);
-		if (code)
-			goto error_return;
-
-		/*
-		 * Now we can make the changes.  Before we join the inode
-		 * to the transaction, if ATTR_SIZE is set then take care of
-		 * the part of the truncation that must be done without the
-		 * inode lock.  This needs to be done before joining the inode
-		 * to the transaction, because the inode cannot be unlocked
-		 * once it is a part of the transaction.
-		 */
-		if (iattr->ia_size > ip->i_size) {
-			/*
-			 * Do the first part of growing a file: zero any data
-			 * in the last block that is beyond the old EOF.  We
-			 * need to do this before the inode is joined to the
-			 * transaction to modify the i_size.
-			 */
-			code = xfs_zero_eof(ip, iattr->ia_size, ip->i_size);
-			if (code)
-				goto error_return;
-		}
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-		lock_flags &= ~XFS_ILOCK_EXCL;
-
-		/*
-		 * We are going to log the inode size change in this
-		 * transaction so any previous writes that are beyond the on
-		 * disk EOF and the new EOF that have not been written out need
-		 * to be written here. If we do not write the data out, we
-		 * expose ourselves to the null files problem.
-		 *
-		 * Only flush from the on disk size to the smaller of the in
-		 * memory file size or the new size as that's the range we
-		 * really care about here and prevents waiting for other data
-		 * not within the range we care about here.
-		 */
-		if (ip->i_size != ip->i_d.di_size &&
-		    iattr->ia_size > ip->i_d.di_size) {
-			code = xfs_flush_pages(ip,
-					ip->i_d.di_size, iattr->ia_size,
-					XBF_ASYNC, FI_NONE);
-			if (code)
-				goto error_return;
-		}
-
-		/* wait for all I/O to complete */
-		xfs_ioend_wait(ip);
-
-		code = -block_truncate_page(inode->i_mapping, iattr->ia_size,
-					    xfs_get_blocks);
-		if (code)
-			goto error_return;
-
-		tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
-		code = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
-					 XFS_TRANS_PERM_LOG_RES,
-					 XFS_ITRUNCATE_LOG_COUNT);
-		if (code)
-			goto error_return;
-
-		truncate_setsize(inode, iattr->ia_size);
-
-		commit_flags = XFS_TRANS_RELEASE_LOG_RES;
-		lock_flags |= XFS_ILOCK_EXCL;
-
-		xfs_ilock(ip, XFS_ILOCK_EXCL);
-
-		xfs_trans_ijoin(tp, ip);
-
-		/*
-		 * Only change the c/mtime if we are changing the size
-		 * or we are explicitly asked to change it. This handles
-		 * the semantic difference between truncate() and ftruncate()
-		 * as implemented in the VFS.
-		 *
-		 * The regular truncate() case without ATTR_CTIME and ATTR_MTIME
-		 * is a special case where we need to update the times despite
-		 * not having these flags set.  For all other operations the
-		 * VFS set these flags explicitly if it wants a timestamp
-		 * update.
-		 */
-		if (iattr->ia_size != ip->i_size &&
-		    (!(mask & (ATTR_CTIME | ATTR_MTIME)))) {
-			iattr->ia_ctime = iattr->ia_mtime =
-				current_fs_time(inode->i_sb);
-			mask |= ATTR_CTIME | ATTR_MTIME;
-		}
-
-		if (iattr->ia_size > ip->i_size) {
-			ip->i_d.di_size = iattr->ia_size;
-			ip->i_size = iattr->ia_size;
-			xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-		} else if (iattr->ia_size <= ip->i_size ||
-			   (iattr->ia_size == 0 && ip->i_d.di_nextents)) {
-			/*
-			 * signal a sync transaction unless
-			 * we're truncating an already unlinked
-			 * file on a wsync filesystem
-			 */
-			code = xfs_itruncate_finish(&tp, ip, iattr->ia_size,
-					    XFS_DATA_FORK,
-					    ((ip->i_d.di_nlink != 0 ||
-					      !(mp->m_flags & XFS_MOUNT_WSYNC))
-					     ? 1 : 0));
-			if (code)
-				goto abort_return;
-			/*
-			 * Truncated "down", so we're removing references
-			 * to old data here - if we now delay flushing for
-			 * a long time, we expose ourselves unduly to the
-			 * notorious NULL files problem.  So, we mark this
-			 * vnode and flush it when the file is closed, and
-			 * do not wait the usual (long) time for writeout.
-			 */
-			xfs_iflags_set(ip, XFS_ITRUNCATED);
-		}
-	} else if (tp) {
-		xfs_trans_ijoin(tp, ip);
-	}
-
-	/*
-	 * Change file ownership.  Must be the owner or privileged.
-	 */
-	if (mask & (ATTR_UID|ATTR_GID)) {
-		/*
-		 * CAP_FSETID overrides the following restrictions:
-		 *
-		 * The set-user-ID and set-group-ID bits of a file will be
-		 * cleared upon successful return from chown()
-		 */
-		if ((ip->i_d.di_mode & (S_ISUID|S_ISGID)) &&
-		    !capable(CAP_FSETID)) {
-			ip->i_d.di_mode &= ~(S_ISUID|S_ISGID);
-		}
-
-		/*
-		 * Change the ownerships and register quota modifications
-		 * in the transaction.
-		 */
-		if (iuid != uid) {
-			if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_UQUOTA_ON(mp)) {
-				ASSERT(mask & ATTR_UID);
-				ASSERT(udqp);
-				olddquot1 = xfs_qm_vop_chown(tp, ip,
-							&ip->i_udquot, udqp);
-			}
-			ip->i_d.di_uid = uid;
-			inode->i_uid = uid;
-		}
-		if (igid != gid) {
-			if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_GQUOTA_ON(mp)) {
-				ASSERT(!XFS_IS_PQUOTA_ON(mp));
-				ASSERT(mask & ATTR_GID);
-				ASSERT(gdqp);
-				olddquot2 = xfs_qm_vop_chown(tp, ip,
-							&ip->i_gdquot, gdqp);
-			}
-			ip->i_d.di_gid = gid;
-			inode->i_gid = gid;
-		}
-	}
-
-	/*
-	 * Change file access modes.
-	 */
-	if (mask & ATTR_MODE) {
-		umode_t mode = iattr->ia_mode;
-
-		if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
-			mode &= ~S_ISGID;
-
-		ip->i_d.di_mode &= S_IFMT;
-		ip->i_d.di_mode |= mode & ~S_IFMT;
-
-		inode->i_mode &= S_IFMT;
-		inode->i_mode |= mode & ~S_IFMT;
-	}
-
-	/*
-	 * Change file access or modified times.
-	 */
-	if (mask & ATTR_ATIME) {
-		inode->i_atime = iattr->ia_atime;
-		ip->i_d.di_atime.t_sec = iattr->ia_atime.tv_sec;
-		ip->i_d.di_atime.t_nsec = iattr->ia_atime.tv_nsec;
-		ip->i_update_core = 1;
-	}
-	if (mask & ATTR_CTIME) {
-		inode->i_ctime = iattr->ia_ctime;
-		ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
-		ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
-		ip->i_update_core = 1;
-	}
-	if (mask & ATTR_MTIME) {
-		inode->i_mtime = iattr->ia_mtime;
-		ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
-		ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
-		ip->i_update_core = 1;
-	}
-
-	/*
-	 * And finally, log the inode core if any attribute in it
-	 * has been changed.
-	 */
-	if (mask & (ATTR_UID|ATTR_GID|ATTR_MODE|
-		    ATTR_ATIME|ATTR_CTIME|ATTR_MTIME))
-		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-
-	XFS_STATS_INC(xs_ig_attrchg);
-
-	/*
-	 * If this is a synchronous mount, make sure that the
-	 * transaction goes to disk before returning to the user.
-	 * This is slightly sub-optimal in that truncates require
-	 * two sync transactions instead of one for wsync filesystems.
-	 * One for the truncate and one for the timestamps since we
-	 * don't want to change the timestamps unless we're sure the
-	 * truncate worked.  Truncates are less than 1% of the laddis
-	 * mix so this probably isn't worth the trouble to optimize.
-	 */
-	code = 0;
-	if (mp->m_flags & XFS_MOUNT_WSYNC)
-		xfs_trans_set_sync(tp);
-
-	code = xfs_trans_commit(tp, commit_flags);
-
-	xfs_iunlock(ip, lock_flags);
-
-	/*
-	 * Release any dquot(s) the inode had kept before chown.
-	 */
-	xfs_qm_dqrele(olddquot1);
-	xfs_qm_dqrele(olddquot2);
-	xfs_qm_dqrele(udqp);
-	xfs_qm_dqrele(gdqp);
-
-	if (code)
-		return code;
-
-	/*
-	 * XXX(hch): Updating the ACL entries is not atomic vs the i_mode
-	 * 	     update.  We could avoid this with linked transactions
-	 * 	     and passing down the transaction pointer all the way
-	 *	     to attr_set.  No previous user of the generic
-	 * 	     Posix ACL code seems to care about this issue either.
-	 */
-	if ((mask & ATTR_MODE) && !(flags & XFS_ATTR_NOACL)) {
-		code = -xfs_acl_chmod(inode);
-		if (code)
-			return XFS_ERROR(code);
-	}
-
-	return 0;
-
- abort_return:
-	commit_flags |= XFS_TRANS_ABORT;
- error_return:
-	xfs_qm_dqrele(udqp);
-	xfs_qm_dqrele(gdqp);
-	if (tp) {
-		xfs_trans_cancel(tp, commit_flags);
-	}
-	if (lock_flags != 0) {
-		xfs_iunlock(ip, lock_flags);
-	}
-	return code;
-}
-
 /*
  * The maximum pathlen is 1024 bytes. Since the minimum file system
  * blocksize is 512 bytes, we can get a max of 2 extents back from
@@ -621,13 +197,6 @@ xfs_free_eofblocks(
 		 */
 		tp = xfs_trans_alloc(mp, XFS_TRANS_INACTIVE);
 
-		/*
-		 * Do the xfs_itruncate_start() call before
-		 * reserving any log space because
-		 * itruncate_start will call into the buffer
-		 * cache and we can't
-		 * do that within a transaction.
-		 */
 		if (flags & XFS_FREE_EOF_TRYLOCK) {
 			if (!xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL)) {
 				xfs_trans_cancel(tp, 0);
@@ -636,13 +205,6 @@ xfs_free_eofblocks(
 		} else {
 			xfs_ilock(ip, XFS_IOLOCK_EXCL);
 		}
-		error = xfs_itruncate_start(ip, XFS_ITRUNC_DEFINITE,
-				    ip->i_size);
-		if (error) {
-			xfs_trans_cancel(tp, 0);
-			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
-			return error;
-		}
 
 		error = xfs_trans_reserve(tp, 0,
 					  XFS_ITRUNCATE_LOG_RES(mp),
@@ -658,15 +220,12 @@ xfs_free_eofblocks(
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
 		xfs_trans_ijoin(tp, ip);
 
-		error = xfs_itruncate_finish(&tp, ip,
-					     ip->i_size,
-					     XFS_DATA_FORK,
-					     0);
-		/*
-		 * If we get an error at this point we
-		 * simply don't bother truncating the file.
-		 */
+		error = xfs_itruncate_data(&tp, ip, ip->i_size);
 		if (error) {
+			/*
+			 * If we get an error at this point we simply don't
+			 * bother truncating the file.
+			 */
 			xfs_trans_cancel(tp,
 					 (XFS_TRANS_RELEASE_LOG_RES |
 					  XFS_TRANS_ABORT));
@@ -960,8 +519,11 @@ xfs_release(
 		 * be exposed to that problem.
 		 */
 		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, XBF_ASYNC, FI_NONE);
+		if (truncated) {
+			xfs_iflags_clear(ip, XFS_IDIRTY_RELEASE);
+			if (VN_DIRTY(VFS_I(ip)) && ip->i_delayed_blks > 0)
+				xfs_flush_pages(ip, 0, -1, XBF_ASYNC, FI_NONE);
+		}
 	}
 
 	if (ip->i_d.di_nlink == 0)
@@ -1081,20 +643,9 @@ xfs_inactive(
 
 	tp = xfs_trans_alloc(mp, XFS_TRANS_INACTIVE);
 	if (truncate) {
-		/*
-		 * Do the xfs_itruncate_start() call before
-		 * reserving any log space because itruncate_start
-		 * will call into the buffer cache and we can't
-		 * do that within a transaction.
-		 */
 		xfs_ilock(ip, XFS_IOLOCK_EXCL);
 
-		error = xfs_itruncate_start(ip, XFS_ITRUNC_DEFINITE, 0);
-		if (error) {
-			xfs_trans_cancel(tp, 0);
-			xfs_iunlock(ip, XFS_IOLOCK_EXCL);
-			return VN_INACTIVE_CACHE;
-		}
+		xfs_ioend_wait(ip);
 
 		error = xfs_trans_reserve(tp, 0,
 					  XFS_ITRUNCATE_LOG_RES(mp),
@@ -1111,16 +662,7 @@ xfs_inactive(
 		xfs_ilock(ip, XFS_ILOCK_EXCL);
 		xfs_trans_ijoin(tp, ip);
 
-		/*
-		 * normally, we have to run xfs_itruncate_finish sync.
-		 * But if filesystem is wsync and we're in the inactive
-		 * path, then we know that nlink == 0, and that the
-		 * xaction that made nlink == 0 is permanently committed
-		 * since xfs_remove runs as a synchronous transaction.
-		 */
-		error = xfs_itruncate_finish(&tp, ip, 0, XFS_DATA_FORK,
-				(!(mp->m_flags & XFS_MOUNT_WSYNC) ? 1 : 0));
-
+		error = xfs_itruncate_data(&tp, ip, 0);
 		if (error) {
 			xfs_trans_cancel(tp,
 				XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT);
@@ -2427,6 +1969,8 @@ xfs_zero_remaining_bytes(
 	if (!bp)
 		return XFS_ERROR(ENOMEM);
 
+	xfs_buf_unlock(bp);
+
 	for (offset = startoff; offset <= endoff; offset = lastoffset + 1) {
 		offset_fsb = XFS_B_TO_FSBT(mp, offset);
 		nimap = 1;
@@ -2781,7 +2325,7 @@ xfs_change_file_space(
 		iattr.ia_valid = ATTR_SIZE;
 		iattr.ia_size = startoffset;
 
-		error = xfs_setattr(ip, &iattr, attr_flags);
+		error = xfs_setattr_size(ip, &iattr, attr_flags);
 
 		if (error)
 			return error;
diff --git a/fs/xfs/xfs_vnodeops.h b/fs/xfs/xfs_vnodeops.h
index 3bcd23353d6..35d3d513e1e 100644
--- a/fs/xfs/xfs_vnodeops.h
+++ b/fs/xfs/xfs_vnodeops.h
@@ -13,7 +13,8 @@ struct xfs_inode;
 struct xfs_iomap;
 
 
-int xfs_setattr(struct xfs_inode *ip, struct iattr *vap, int flags);
+int xfs_setattr_nonsize(struct xfs_inode *ip, struct iattr *vap, int flags);
+int xfs_setattr_size(struct xfs_inode *ip, struct iattr *vap, int flags);
 #define	XFS_ATTR_DMI		0x01	/* invocation from a DMI function */
 #define	XFS_ATTR_NONBLOCK	0x02	/* return EAGAIN if operation would block */
 #define XFS_ATTR_NOLOCK		0x04	/* Don't grab any conflicting locks */