Merge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfasheh/ocfs2

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfasheh/ocfs2: (138 commits)
  ocfs2: Access the right buffer_head in ocfs2_merge_rec_left.
  ocfs2: use min_t in ocfs2_quota_read()
  ocfs2: remove unneeded lvb casts
  ocfs2: Add xattr support checking in init_security
  ocfs2: alloc xattr bucket in ocfs2_xattr_set_handle
  ocfs2: calculate and reserve credits for xattr value in mknod
  ocfs2/xattr: fix credits calculation during index create
  ocfs2/xattr: Always updating ctime during xattr set.
  ocfs2/xattr: Remove extend_trans call and add its credits from the beginning
  ocfs2/dlm: Fix race during lockres mastery
  ocfs2/dlm: Fix race in adding/removing lockres' to/from the tracking list
  ocfs2/dlm: Hold off sending lockres drop ref message while lockres is migrating
  ocfs2/dlm: Clean up errors in dlm_proxy_ast_handler()
  ocfs2/dlm: Fix a race between migrate request and exit domain
  ocfs2: One more hamming code optimization.
  ocfs2: Another hamming code optimization.
  ocfs2: Don't hand-code xor in ocfs2_hamming_encode().
  ocfs2: Enable metadata checksums.
  ocfs2: Validate superblock with checksum and ecc.
  ocfs2: Checksum and ECC for directory blocks.
  ...

Documentation/filesystems/ocfs2.txt

@@ -31,7 +31,6 @@ Features which OCFS2 does not support yet:
 	- quotas
 	- Directory change notification (F_NOTIFY)
 	- Distributed Caching (F_SETLEASE/F_GETLEASE/break_lease)
-	- POSIX ACLs
 
 Mount options
 =============
@@ -79,3 +78,5 @@ inode64			Indicates that Ocfs2 is allowed to create inodes at
 			bits of significance.
 user_xattr	(*)	Enables Extended User Attributes.
 nouser_xattr		Disables Extended User Attributes.
+acl			Enables POSIX Access Control Lists support.
+noacl		(*)	Disables POSIX Access Control Lists support.

fs/Kconfig

@@ -189,6 +189,8 @@ config OCFS2_FS
 	select CONFIGFS_FS
 	select JBD2
 	select CRC32
+	select QUOTA
+	select QUOTA_TREE
 	help
 	  OCFS2 is a general purpose extent based shared disk cluster file
 	  system with many similarities to ext3. It supports 64 bit inode
@@ -258,15 +260,14 @@ config OCFS2_DEBUG_FS
 	  this option for debugging only as it is likely to decrease
 	  performance of the filesystem.
 
-config OCFS2_COMPAT_JBD
-	bool "Use JBD for compatibility"
+config OCFS2_FS_POSIX_ACL
+	bool "OCFS2 POSIX Access Control Lists"
 	depends on OCFS2_FS
+	select FS_POSIX_ACL
 	default n
-	select JBD
 	help
-	  The ocfs2 filesystem now uses JBD2 for its journalling.  JBD2
-	  is backwards compatible with JBD.  It is safe to say N here.
-	  However, if you really want to use the original JBD, say Y here.
+	  Posix Access Control Lists (ACLs) support permissions for users and
+	  groups beyond the owner/group/world scheme.
 
 endif # BLOCK
 
@@ -303,6 +304,10 @@ config PRINT_QUOTA_WARNING
 	  Note that this behavior is currently deprecated and may go away in
 	  future. Please use notification via netlink socket instead.
 
+# Generic support for tree structured quota files. Seleted when needed.
+config QUOTA_TREE
+	 tristate
+
 config QFMT_V1
 	tristate "Old quota format support"
 	depends on QUOTA
@@ -314,6 +319,7 @@ config QFMT_V1
 config QFMT_V2
 	tristate "Quota format v2 support"
 	depends on QUOTA
+	select QUOTA_TREE
 	help
 	  This quota format allows using quotas with 32-bit UIDs/GIDs. If you
 	  need this functionality say Y here.

fs/Makefile

@@ -54,6 +54,7 @@ obj-$(CONFIG_GENERIC_ACL)	+= generic_acl.o
 obj-$(CONFIG_QUOTA)		+= dquot.o
 obj-$(CONFIG_QFMT_V1)		+= quota_v1.o
 obj-$(CONFIG_QFMT_V2)		+= quota_v2.o
+obj-$(CONFIG_QUOTA_TREE)	+= quota_tree.o
 obj-$(CONFIG_QUOTACTL)		+= quota.o
 
 obj-$(CONFIG_PROC_FS)		+= proc/

fs/dquot.c

@@ -211,8 +211,6 @@ static struct hlist_head *dquot_hash;
 
 struct dqstats dqstats;
 
-static void dqput(struct dquot *dquot);
-
 static inline unsigned int
 hashfn(const struct super_block *sb, unsigned int id, int type)
 {
@@ -415,6 +413,17 @@ out_dqlock:
 	return ret;
 }
 
+void dquot_destroy(struct dquot *dquot)
+{
+	kmem_cache_free(dquot_cachep, dquot);
+}
+EXPORT_SYMBOL(dquot_destroy);
+
+static inline void do_destroy_dquot(struct dquot *dquot)
+{
+	dquot->dq_sb->dq_op->destroy_dquot(dquot);
+}
+
 /* Invalidate all dquots on the list. Note that this function is called after
  * quota is disabled and pointers from inodes removed so there cannot be new
  * quota users. There can still be some users of quotas due to inodes being
@@ -463,9 +472,44 @@ restart:
 		remove_dquot_hash(dquot);
 		remove_free_dquot(dquot);
 		remove_inuse(dquot);
-		kmem_cache_free(dquot_cachep, dquot);
+		do_destroy_dquot(dquot);
+	}
+	spin_unlock(&dq_list_lock);
+}
+
+/* Call callback for every active dquot on given filesystem */
+int dquot_scan_active(struct super_block *sb,
+		      int (*fn)(struct dquot *dquot, unsigned long priv),
+		      unsigned long priv)
+{
+	struct dquot *dquot, *old_dquot = NULL;
+	int ret = 0;
+
+	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
+	spin_lock(&dq_list_lock);
+	list_for_each_entry(dquot, &inuse_list, dq_inuse) {
+		if (!test_bit(DQ_ACTIVE_B, &dquot->dq_flags))
+			continue;
+		if (dquot->dq_sb != sb)
+			continue;
+		/* Now we have active dquot so we can just increase use count */
+		atomic_inc(&dquot->dq_count);
+		dqstats.lookups++;
+		spin_unlock(&dq_list_lock);
+		dqput(old_dquot);
+		old_dquot = dquot;
+		ret = fn(dquot, priv);
+		if (ret < 0)
+			goto out;
+		spin_lock(&dq_list_lock);
+		/* We are safe to continue now because our dquot could not
+		 * be moved out of the inuse list while we hold the reference */
 	}
 	spin_unlock(&dq_list_lock);
+out:
+	dqput(old_dquot);
+	mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
+	return ret;
 }
 
 int vfs_quota_sync(struct super_block *sb, int type)
@@ -479,7 +523,7 @@ int vfs_quota_sync(struct super_block *sb, int type)
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (type != -1 && cnt != type)
 			continue;
-		if (!sb_has_quota_enabled(sb, cnt))
+		if (!sb_has_quota_active(sb, cnt))
 			continue;
 		spin_lock(&dq_list_lock);
 		dirty = &dqopt->info[cnt].dqi_dirty_list;
@@ -504,8 +548,8 @@ int vfs_quota_sync(struct super_block *sb, int type)
 	}
 
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
-		if ((cnt == type || type == -1) && sb_has_quota_enabled(sb, cnt)
-			&& info_dirty(&dqopt->info[cnt]))
+		if ((cnt == type || type == -1) && sb_has_quota_active(sb, cnt)
+		    && info_dirty(&dqopt->info[cnt]))
 			sb->dq_op->write_info(sb, cnt);
 	spin_lock(&dq_list_lock);
 	dqstats.syncs++;
@@ -527,7 +571,7 @@ static void prune_dqcache(int count)
 		remove_dquot_hash(dquot);
 		remove_free_dquot(dquot);
 		remove_inuse(dquot);
-		kmem_cache_free(dquot_cachep, dquot);
+		do_destroy_dquot(dquot);
 		count--;
 		head = free_dquots.prev;
 	}
@@ -558,7 +602,7 @@ static struct shrinker dqcache_shrinker = {
  * NOTE: If you change this function please check whether dqput_blocks() works right...
  * MUST be called with either dqptr_sem or dqonoff_mutex held
  */
-static void dqput(struct dquot *dquot)
+void dqput(struct dquot *dquot)
 {
 	int ret;
 
@@ -584,7 +628,7 @@ we_slept:
 		/* We have more than one user... nothing to do */
 		atomic_dec(&dquot->dq_count);
 		/* Releasing dquot during quotaoff phase? */
-		if (!sb_has_quota_enabled(dquot->dq_sb, dquot->dq_type) &&
+		if (!sb_has_quota_active(dquot->dq_sb, dquot->dq_type) &&
 		    atomic_read(&dquot->dq_count) == 1)
 			wake_up(&dquot->dq_wait_unused);
 		spin_unlock(&dq_list_lock);
@@ -625,11 +669,17 @@ we_slept:
 	spin_unlock(&dq_list_lock);
 }
 
+struct dquot *dquot_alloc(struct super_block *sb, int type)
+{
+	return kmem_cache_zalloc(dquot_cachep, GFP_NOFS);
+}
+EXPORT_SYMBOL(dquot_alloc);
+
 static struct dquot *get_empty_dquot(struct super_block *sb, int type)
 {
 	struct dquot *dquot;
 
-	dquot = kmem_cache_zalloc(dquot_cachep, GFP_NOFS);
+	dquot = sb->dq_op->alloc_dquot(sb, type);
 	if(!dquot)
 		return NODQUOT;
 
@@ -646,16 +696,34 @@ static struct dquot *get_empty_dquot(struct super_block *sb, int type)
 	return dquot;
 }
 
+/*
+ * Check whether dquot is in memory.
+ * MUST be called with either dqptr_sem or dqonoff_mutex held
+ */
+int dquot_is_cached(struct super_block *sb, unsigned int id, int type)
+{
+	unsigned int hashent = hashfn(sb, id, type);
+	int ret = 0;
+
+        if (!sb_has_quota_active(sb, type))
+		return 0;
+	spin_lock(&dq_list_lock);
+	if (find_dquot(hashent, sb, id, type) != NODQUOT)
+		ret = 1;
+	spin_unlock(&dq_list_lock);
+	return ret;
+}
+
 /*
  * Get reference to dquot
  * MUST be called with either dqptr_sem or dqonoff_mutex held
  */
-static struct dquot *dqget(struct super_block *sb, unsigned int id, int type)
+struct dquot *dqget(struct super_block *sb, unsigned int id, int type)
 {
 	unsigned int hashent = hashfn(sb, id, type);
 	struct dquot *dquot, *empty = NODQUOT;
 
-        if (!sb_has_quota_enabled(sb, type))
+        if (!sb_has_quota_active(sb, type))
 		return NODQUOT;
 we_slept:
 	spin_lock(&dq_list_lock);
@@ -682,7 +750,7 @@ we_slept:
 		dqstats.lookups++;
 		spin_unlock(&dq_list_lock);
 		if (empty)
-			kmem_cache_free(dquot_cachep, empty);
+			do_destroy_dquot(empty);
 	}
 	/* Wait for dq_lock - after this we know that either dquot_release() is already
 	 * finished or it will be canceled due to dq_count > 1 test */
@@ -820,7 +888,7 @@ static void drop_dquot_ref(struct super_block *sb, int type)
 	}
 }
 
-static inline void dquot_incr_inodes(struct dquot *dquot, unsigned long number)
+static inline void dquot_incr_inodes(struct dquot *dquot, qsize_t number)
 {
 	dquot->dq_dqb.dqb_curinodes += number;
 }
@@ -830,9 +898,10 @@ static inline void dquot_incr_space(struct dquot *dquot, qsize_t number)
 	dquot->dq_dqb.dqb_curspace += number;
 }
 
-static inline void dquot_decr_inodes(struct dquot *dquot, unsigned long number)
+static inline void dquot_decr_inodes(struct dquot *dquot, qsize_t number)
 {
-	if (dquot->dq_dqb.dqb_curinodes > number)
+	if (sb_dqopt(dquot->dq_sb)->flags & DQUOT_NEGATIVE_USAGE ||
+	    dquot->dq_dqb.dqb_curinodes >= number)
 		dquot->dq_dqb.dqb_curinodes -= number;
 	else
 		dquot->dq_dqb.dqb_curinodes = 0;
@@ -843,11 +912,12 @@ static inline void dquot_decr_inodes(struct dquot *dquot, unsigned long number)
 
 static inline void dquot_decr_space(struct dquot *dquot, qsize_t number)
 {
-	if (dquot->dq_dqb.dqb_curspace > number)
+	if (sb_dqopt(dquot->dq_sb)->flags & DQUOT_NEGATIVE_USAGE ||
+	    dquot->dq_dqb.dqb_curspace >= number)
 		dquot->dq_dqb.dqb_curspace -= number;
 	else
 		dquot->dq_dqb.dqb_curspace = 0;
-	if (toqb(dquot->dq_dqb.dqb_curspace) <= dquot->dq_dqb.dqb_bsoftlimit)
+	if (dquot->dq_dqb.dqb_curspace <= dquot->dq_dqb.dqb_bsoftlimit)
 		dquot->dq_dqb.dqb_btime = (time_t) 0;
 	clear_bit(DQ_BLKS_B, &dquot->dq_flags);
 }
@@ -1023,10 +1093,11 @@ static inline char ignore_hardlimit(struct dquot *dquot)
 }
 
 /* needs dq_data_lock */
-static int check_idq(struct dquot *dquot, ulong inodes, char *warntype)
+static int check_idq(struct dquot *dquot, qsize_t inodes, char *warntype)
 {
 	*warntype = QUOTA_NL_NOWARN;
-	if (inodes <= 0 || test_bit(DQ_FAKE_B, &dquot->dq_flags))
+	if (!sb_has_quota_limits_enabled(dquot->dq_sb, dquot->dq_type) ||
+	    test_bit(DQ_FAKE_B, &dquot->dq_flags))
 		return QUOTA_OK;
 
 	if (dquot->dq_dqb.dqb_ihardlimit &&
@@ -1058,11 +1129,12 @@ static int check_idq(struct dquot *dquot, ulong inodes, char *warntype)
 static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *warntype)
 {
 	*warntype = QUOTA_NL_NOWARN;
-	if (space <= 0 || test_bit(DQ_FAKE_B, &dquot->dq_flags))
+	if (!sb_has_quota_limits_enabled(dquot->dq_sb, dquot->dq_type) ||
+	    test_bit(DQ_FAKE_B, &dquot->dq_flags))
 		return QUOTA_OK;
 
 	if (dquot->dq_dqb.dqb_bhardlimit &&
-	   toqb(dquot->dq_dqb.dqb_curspace + space) > dquot->dq_dqb.dqb_bhardlimit &&
+	    dquot->dq_dqb.dqb_curspace + space > dquot->dq_dqb.dqb_bhardlimit &&
             !ignore_hardlimit(dquot)) {
 		if (!prealloc)
 			*warntype = QUOTA_NL_BHARDWARN;
@@ -1070,7 +1142,7 @@ static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *war
 	}
 
 	if (dquot->dq_dqb.dqb_bsoftlimit &&
-	   toqb(dquot->dq_dqb.dqb_curspace + space) > dquot->dq_dqb.dqb_bsoftlimit &&
+	    dquot->dq_dqb.dqb_curspace + space > dquot->dq_dqb.dqb_bsoftlimit &&
 	    dquot->dq_dqb.dqb_btime && get_seconds() >= dquot->dq_dqb.dqb_btime &&
             !ignore_hardlimit(dquot)) {
 		if (!prealloc)
@@ -1079,7 +1151,7 @@ static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *war
 	}
 
 	if (dquot->dq_dqb.dqb_bsoftlimit &&
-	   toqb(dquot->dq_dqb.dqb_curspace + space) > dquot->dq_dqb.dqb_bsoftlimit &&
+	    dquot->dq_dqb.dqb_curspace + space > dquot->dq_dqb.dqb_bsoftlimit &&
 	    dquot->dq_dqb.dqb_btime == 0) {
 		if (!prealloc) {
 			*warntype = QUOTA_NL_BSOFTWARN;
@@ -1096,10 +1168,11 @@ static int check_bdq(struct dquot *dquot, qsize_t space, int prealloc, char *war
 	return QUOTA_OK;
 }
 
-static int info_idq_free(struct dquot *dquot, ulong inodes)
+static int info_idq_free(struct dquot *dquot, qsize_t inodes)
 {
 	if (test_bit(DQ_FAKE_B, &dquot->dq_flags) ||
-	    dquot->dq_dqb.dqb_curinodes <= dquot->dq_dqb.dqb_isoftlimit)
+	    dquot->dq_dqb.dqb_curinodes <= dquot->dq_dqb.dqb_isoftlimit ||
+	    !sb_has_quota_limits_enabled(dquot->dq_sb, dquot->dq_type))
 		return QUOTA_NL_NOWARN;
 
 	if (dquot->dq_dqb.dqb_curinodes - inodes <= dquot->dq_dqb.dqb_isoftlimit)
@@ -1113,15 +1186,13 @@ static int info_idq_free(struct dquot *dquot, ulong inodes)
 static int info_bdq_free(struct dquot *dquot, qsize_t space)
 {
 	if (test_bit(DQ_FAKE_B, &dquot->dq_flags) ||
-	    toqb(dquot->dq_dqb.dqb_curspace) <= dquot->dq_dqb.dqb_bsoftlimit)
+	    dquot->dq_dqb.dqb_curspace <= dquot->dq_dqb.dqb_bsoftlimit)
 		return QUOTA_NL_NOWARN;
 
-	if (toqb(dquot->dq_dqb.dqb_curspace - space) <=
-	    dquot->dq_dqb.dqb_bsoftlimit)
+	if (dquot->dq_dqb.dqb_curspace - space <= dquot->dq_dqb.dqb_bsoftlimit)
 		return QUOTA_NL_BSOFTBELOW;
-	if (toqb(dquot->dq_dqb.dqb_curspace) >= dquot->dq_dqb.dqb_bhardlimit &&
-	    toqb(dquot->dq_dqb.dqb_curspace - space) <
-						dquot->dq_dqb.dqb_bhardlimit)
+	if (dquot->dq_dqb.dqb_curspace >= dquot->dq_dqb.dqb_bhardlimit &&
+	    dquot->dq_dqb.dqb_curspace - space < dquot->dq_dqb.dqb_bhardlimit)
 		return QUOTA_NL_BHARDBELOW;
 	return QUOTA_NL_NOWARN;
 }
@@ -1166,17 +1237,23 @@ out_err:
  * 	Release all quotas referenced by inode
  *	Transaction must be started at an entry
  */
-int dquot_drop(struct inode *inode)
+int dquot_drop_locked(struct inode *inode)
 {
 	int cnt;
 
-	down_write(&sb_dqopt(inode->i_sb)->dqptr_sem);
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (inode->i_dquot[cnt] != NODQUOT) {
 			dqput(inode->i_dquot[cnt]);
 			inode->i_dquot[cnt] = NODQUOT;
 		}
 	}
+	return 0;
+}
+
+int dquot_drop(struct inode *inode)
+{
+	down_write(&sb_dqopt(inode->i_sb)->dqptr_sem);
+	dquot_drop_locked(inode);
 	up_write(&sb_dqopt(inode->i_sb)->dqptr_sem);
 	return 0;
 }
@@ -1264,7 +1341,7 @@ warn_put_all:
 /*
  * This operation can block, but only after everything is updated
  */
-int dquot_alloc_inode(const struct inode *inode, unsigned long number)
+int dquot_alloc_inode(const struct inode *inode, qsize_t number)
 {
 	int cnt, ret = NO_QUOTA;
 	char warntype[MAXQUOTAS];
@@ -1349,7 +1426,7 @@ out_sub:
 /*
  * This operation can block, but only after everything is updated
  */
-int dquot_free_inode(const struct inode *inode, unsigned long number)
+int dquot_free_inode(const struct inode *inode, qsize_t number)
 {
 	unsigned int cnt;
 	char warntype[MAXQUOTAS];
@@ -1495,7 +1572,7 @@ warn_put_all:
 /* Wrapper for transferring ownership of an inode */
 int vfs_dq_transfer(struct inode *inode, struct iattr *iattr)
 {
-	if (sb_any_quota_enabled(inode->i_sb) && !IS_NOQUOTA(inode)) {
+	if (sb_any_quota_active(inode->i_sb) && !IS_NOQUOTA(inode)) {
 		vfs_dq_init(inode);
 		if (inode->i_sb->dq_op->transfer(inode, iattr) == NO_QUOTA)
 			return 1;
@@ -1533,54 +1610,27 @@ struct dquot_operations dquot_operations = {
 	.acquire_dquot	= dquot_acquire,
 	.release_dquot	= dquot_release,
 	.mark_dirty	= dquot_mark_dquot_dirty,
-	.write_info	= dquot_commit_info
+	.write_info	= dquot_commit_info,
+	.alloc_dquot	= dquot_alloc,
+	.destroy_dquot	= dquot_destroy,
 };
 
-static inline void set_enable_flags(struct quota_info *dqopt, int type)
-{
-	switch (type) {
-		case USRQUOTA:
-			dqopt->flags |= DQUOT_USR_ENABLED;
-			dqopt->flags &= ~DQUOT_USR_SUSPENDED;
-			break;
-		case GRPQUOTA:
-			dqopt->flags |= DQUOT_GRP_ENABLED;
-			dqopt->flags &= ~DQUOT_GRP_SUSPENDED;
-			break;
-	}
-}
-
-static inline void reset_enable_flags(struct quota_info *dqopt, int type,
-				      int remount)
-{
-	switch (type) {
-		case USRQUOTA:
-			dqopt->flags &= ~DQUOT_USR_ENABLED;
-			if (remount)
-				dqopt->flags |= DQUOT_USR_SUSPENDED;
-			else
-				dqopt->flags &= ~DQUOT_USR_SUSPENDED;
-			break;
-		case GRPQUOTA:
-			dqopt->flags &= ~DQUOT_GRP_ENABLED;
-			if (remount)
-				dqopt->flags |= DQUOT_GRP_SUSPENDED;
-			else
-				dqopt->flags &= ~DQUOT_GRP_SUSPENDED;
-			break;
-	}
-}
-
-
 /*
  * Turn quota off on a device. type == -1 ==> quotaoff for all types (umount)
  */
-int vfs_quota_off(struct super_block *sb, int type, int remount)
+int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags)
 {
 	int cnt, ret = 0;
 	struct quota_info *dqopt = sb_dqopt(sb);
 	struct inode *toputinode[MAXQUOTAS];
 
+	/* Cannot turn off usage accounting without turning off limits, or
+	 * suspend quotas and simultaneously turn quotas off. */
+	if ((flags & DQUOT_USAGE_ENABLED && !(flags & DQUOT_LIMITS_ENABLED))
+	    || (flags & DQUOT_SUSPENDED && flags & (DQUOT_LIMITS_ENABLED |
+	    DQUOT_USAGE_ENABLED)))
+		return -EINVAL;
+
 	/* We need to serialize quota_off() for device */
 	mutex_lock(&dqopt->dqonoff_mutex);
 
@@ -1589,7 +1639,7 @@ int vfs_quota_off(struct super_block *sb, int type, int remount)
 	 * sometimes we are called when fill_super() failed and calling
 	 * sync_fs() in such cases does no good.
 	 */
-	if (!sb_any_quota_enabled(sb) && !sb_any_quota_suspended(sb)) {
+	if (!sb_any_quota_loaded(sb)) {
 		mutex_unlock(&dqopt->dqonoff_mutex);
 		return 0;
 	}
@@ -1597,17 +1647,28 @@ int vfs_quota_off(struct super_block *sb, int type, int remount)
 		toputinode[cnt] = NULL;
 		if (type != -1 && cnt != type)
 			continue;
-		/* If we keep inodes of quota files after remount and quotaoff
-		 * is called, drop kept inodes. */
-		if (!remount && sb_has_quota_suspended(sb, cnt)) {
-			iput(dqopt->files[cnt]);
-			dqopt->files[cnt] = NULL;
-			reset_enable_flags(dqopt, cnt, 0);
+		if (!sb_has_quota_loaded(sb, cnt))
 			continue;
+
+		if (flags & DQUOT_SUSPENDED) {
+			dqopt->flags |=
+				dquot_state_flag(DQUOT_SUSPENDED, cnt);
+		} else {
+			dqopt->flags &= ~dquot_state_flag(flags, cnt);
+			/* Turning off suspended quotas? */
+			if (!sb_has_quota_loaded(sb, cnt) &&
+			    sb_has_quota_suspended(sb, cnt)) {
+				dqopt->flags &=	~dquot_state_flag(
+							DQUOT_SUSPENDED, cnt);
+				iput(dqopt->files[cnt]);
+				dqopt->files[cnt] = NULL;
+				continue;
+			}
 		}
-		if (!sb_has_quota_enabled(sb, cnt))
+
+		/* We still have to keep quota loaded? */
+		if (sb_has_quota_loaded(sb, cnt) && !(flags & DQUOT_SUSPENDED))
 			continue;
-		reset_enable_flags(dqopt, cnt, remount);
 
 		/* Note: these are blocking operations */
 		drop_dquot_ref(sb, cnt);
@@ -1623,7 +1684,7 @@ int vfs_quota_off(struct super_block *sb, int type, int remount)
 		put_quota_format(dqopt->info[cnt].dqi_format);
 
 		toputinode[cnt] = dqopt->files[cnt];
-		if (!remount)
+		if (!sb_has_quota_loaded(sb, cnt))
 			dqopt->files[cnt] = NULL;
 		dqopt->info[cnt].dqi_flags = 0;
 		dqopt->info[cnt].dqi_igrace = 0;
@@ -1631,6 +1692,11 @@ int vfs_quota_off(struct super_block *sb, int type, int remount)
 		dqopt->ops[cnt] = NULL;
 	}
 	mutex_unlock(&dqopt->dqonoff_mutex);
+
+	/* Skip syncing and setting flags if quota files are hidden */
+	if (dqopt->flags & DQUOT_QUOTA_SYS_FILE)
+		goto put_inodes;
+
 	/* Sync the superblock so that buffers with quota data are written to
 	 * disk (and so userspace sees correct data afterwards). */
 	if (sb->s_op->sync_fs)
@@ -1646,7 +1712,7 @@ int vfs_quota_off(struct super_block *sb, int type, int remount)
 			mutex_lock(&dqopt->dqonoff_mutex);
 			/* If quota was reenabled in the meantime, we have
 			 * nothing to do */
-			if (!sb_has_quota_enabled(sb, cnt)) {
+			if (!sb_has_quota_loaded(sb, cnt)) {
 				mutex_lock_nested(&toputinode[cnt]->i_mutex, I_MUTEX_QUOTA);
 				toputinode[cnt]->i_flags &= ~(S_IMMUTABLE |
 				  S_NOATIME | S_NOQUOTA);
@@ -1655,26 +1721,43 @@ int vfs_quota_off(struct super_block *sb, int type, int remount)
 				mark_inode_dirty(toputinode[cnt]);
 			}
 			mutex_unlock(&dqopt->dqonoff_mutex);
+		}
+	if (sb->s_bdev)
+		invalidate_bdev(sb->s_bdev);
+put_inodes:
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
+		if (toputinode[cnt]) {
 			/* On remount RO, we keep the inode pointer so that we
-			 * can reenable quota on the subsequent remount RW.
-			 * But we have better not keep inode pointer when there
-			 * is pending delete on the quota file... */
-			if (!remount)
+			 * can reenable quota on the subsequent remount RW. We
+			 * have to check 'flags' variable and not use sb_has_
+			 * function because another quotaon / quotaoff could
+			 * change global state before we got here. We refuse
+			 * to suspend quotas when there is pending delete on
+			 * the quota file... */
+			if (!(flags & DQUOT_SUSPENDED))
 				iput(toputinode[cnt]);
 			else if (!toputinode[cnt]->i_nlink)
 				ret = -EBUSY;
 		}
-	if (sb->s_bdev)
-		invalidate_bdev(sb->s_bdev);
 	return ret;
 }
 
+int vfs_quota_off(struct super_block *sb, int type, int remount)
+{
+	return vfs_quota_disable(sb, type, remount ? DQUOT_SUSPENDED :
+				 (DQUOT_USAGE_ENABLED | DQUOT_LIMITS_ENABLED));
+}
+
 /*
  *	Turn quotas on on a device
  */
 
-/* Helper function when we already have the inode */
-static int vfs_quota_on_inode(struct inode *inode, int type, int format_id)
+/*
+ * Helper function to turn quotas on when we already have the inode of
+ * quota file and no quota information is loaded.
+ */
+static int vfs_load_quota_inode(struct inode *inode, int type, int format_id,
+	unsigned int flags)
 {
 	struct quota_format_type *fmt = find_quota_format(format_id);
 	struct super_block *sb = inode->i_sb;
@@ -1696,27 +1779,37 @@ static int vfs_quota_on_inode(struct inode *inode, int type, int format_id)
 		error = -EINVAL;
 		goto out_fmt;
 	}
+	/* Usage always has to be set... */
+	if (!(flags & DQUOT_USAGE_ENABLED)) {
+		error = -EINVAL;
+		goto out_fmt;
+	}
 
-	/* As we bypass the pagecache we must now flush the inode so that
-	 * we see all the changes from userspace... */
-	write_inode_now(inode, 1);
-	/* And now flush the block cache so that kernel sees the changes */
-	invalidate_bdev(sb->s_bdev);
+	if (!(dqopt->flags & DQUOT_QUOTA_SYS_FILE)) {
+		/* As we bypass the pagecache we must now flush the inode so
+		 * that we see all the changes from userspace... */
+		write_inode_now(inode, 1);
+		/* And now flush the block cache so that kernel sees the
+		 * changes */
+		invalidate_bdev(sb->s_bdev);
+	}
 	mutex_lock(&inode->i_mutex);
 	mutex_lock(&dqopt->dqonoff_mutex);
-	if (sb_has_quota_enabled(sb, type) ||
-			sb_has_quota_suspended(sb, type)) {
+	if (sb_has_quota_loaded(sb, type)) {
 		error = -EBUSY;
 		goto out_lock;
 	}
-	/* We don't want quota and atime on quota files (deadlocks possible)
-	 * Also nobody should write to the file - we use special IO operations
-	 * which ignore the immutable bit. */
-	down_write(&dqopt->dqptr_sem);
-	oldflags = inode->i_flags & (S_NOATIME | S_IMMUTABLE | S_NOQUOTA);
-	inode->i_flags |= S_NOQUOTA | S_NOATIME | S_IMMUTABLE;
-	up_write(&dqopt->dqptr_sem);
-	sb->dq_op->drop(inode);
+
+	if (!(dqopt->flags & DQUOT_QUOTA_SYS_FILE)) {
+		/* We don't want quota and atime on quota files (deadlocks
+		 * possible) Also nobody should write to the file - we use
+		 * special IO operations which ignore the immutable bit. */
+		down_write(&dqopt->dqptr_sem);
+		oldflags = inode->i_flags & (S_NOATIME | S_IMMUTABLE | S_NOQUOTA);
+		inode->i_flags |= S_NOQUOTA | S_NOATIME | S_IMMUTABLE;
+		up_write(&dqopt->dqptr_sem);
+		sb->dq_op->drop(inode);
+	}
 
 	error = -EIO;
 	dqopt->files[type] = igrab(inode);
@@ -1737,7 +1830,7 @@ static int vfs_quota_on_inode(struct inode *inode, int type, int format_id)
 	}
 	mutex_unlock(&dqopt->dqio_mutex);
 	mutex_unlock(&inode->i_mutex);
-	set_enable_flags(dqopt, type);
+	dqopt->flags |= dquot_state_flag(flags, type);
 
 	add_dquot_ref(sb, type);
 	mutex_unlock(&dqopt->dqonoff_mutex);
@@ -1770,20 +1863,23 @@ static int vfs_quota_on_remount(struct super_block *sb, int type)
 	struct quota_info *dqopt = sb_dqopt(sb);
 	struct inode *inode;
 	int ret;
+	unsigned int flags;
 
 	mutex_lock(&dqopt->dqonoff_mutex);
 	if (!sb_has_quota_suspended(sb, type)) {
 		mutex_unlock(&dqopt->dqonoff_mutex);
 		return 0;
 	}
-	BUG_ON(sb_has_quota_enabled(sb, type));
-
 	inode = dqopt->files[type];
 	dqopt->files[type] = NULL;
-	reset_enable_flags(dqopt, type, 0);
+	flags = dqopt->flags & dquot_state_flag(DQUOT_USAGE_ENABLED |
+						DQUOT_LIMITS_ENABLED, type);
+	dqopt->flags &= ~dquot_state_flag(DQUOT_STATE_FLAGS, type);
 	mutex_unlock(&dqopt->dqonoff_mutex);
 
-	ret = vfs_quota_on_inode(inode, type, dqopt->info[type].dqi_fmt_id);
+	flags = dquot_generic_flag(flags, type);
+	ret = vfs_load_quota_inode(inode, type, dqopt->info[type].dqi_fmt_id,
+				   flags);
 	iput(inode);
 
 	return ret;
@@ -1799,12 +1895,12 @@ int vfs_quota_on_path(struct super_block *sb, int type, int format_id,
 	if (path->mnt->mnt_sb != sb)
 		error = -EXDEV;
 	else
-		error = vfs_quota_on_inode(path->dentry->d_inode, type,
-					   format_id);
+		error = vfs_load_quota_inode(path->dentry->d_inode, type,
+					     format_id, DQUOT_USAGE_ENABLED |
+					     DQUOT_LIMITS_ENABLED);
 	return error;
 }
 
-/* Actual function called from quotactl() */
 int vfs_quota_on(struct super_block *sb, int type, int format_id, char *name,
 		 int remount)
 {
@@ -1822,6 +1918,50 @@ int vfs_quota_on(struct super_block *sb, int type, int format_id, char *name,
 	return error;
 }
 
+/*
+ * More powerful function for turning on quotas allowing setting
+ * of individual quota flags
+ */
+int vfs_quota_enable(struct inode *inode, int type, int format_id,
+		unsigned int flags)
+{
+	int ret = 0;
+	struct super_block *sb = inode->i_sb;
+	struct quota_info *dqopt = sb_dqopt(sb);
+
+	/* Just unsuspend quotas? */
+	if (flags & DQUOT_SUSPENDED)
+		return vfs_quota_on_remount(sb, type);
+	if (!flags)
+		return 0;
+	/* Just updating flags needed? */
+	if (sb_has_quota_loaded(sb, type)) {
+		mutex_lock(&dqopt->dqonoff_mutex);
+		/* Now do a reliable test... */
+		if (!sb_has_quota_loaded(sb, type)) {
+			mutex_unlock(&dqopt->dqonoff_mutex);
+			goto load_quota;
+		}
+		if (flags & DQUOT_USAGE_ENABLED &&
+		    sb_has_quota_usage_enabled(sb, type)) {
+			ret = -EBUSY;
+			goto out_lock;
+		}
+		if (flags & DQUOT_LIMITS_ENABLED &&
+		    sb_has_quota_limits_enabled(sb, type)) {
+			ret = -EBUSY;
+			goto out_lock;
+		}
+		sb_dqopt(sb)->flags |= dquot_state_flag(flags, type);
+out_lock:
+		mutex_unlock(&dqopt->dqonoff_mutex);
+		return ret;
+	}
+
+load_quota:
+	return vfs_load_quota_inode(inode, type, format_id, flags);
+}
+
 /*
  * This function is used when filesystem needs to initialize quotas
  * during mount time.
@@ -1843,7 +1983,8 @@ int vfs_quota_on_mount(struct super_block *sb, char *qf_name,
 
 	error = security_quota_on(dentry);
 	if (!error)
-		error = vfs_quota_on_inode(dentry->d_inode, type, format_id);
+		error = vfs_load_quota_inode(dentry->d_inode, type, format_id,
+				DQUOT_USAGE_ENABLED | DQUOT_LIMITS_ENABLED);
 
 out:
 	dput(dentry);
@@ -1866,14 +2007,24 @@ int vfs_dq_quota_on_remount(struct super_block *sb)
 	return ret;
 }
 
+static inline qsize_t qbtos(qsize_t blocks)
+{
+	return blocks << QIF_DQBLKSIZE_BITS;
+}
+
+static inline qsize_t stoqb(qsize_t space)
+{
+	return (space + QIF_DQBLKSIZE - 1) >> QIF_DQBLKSIZE_BITS;
+}
+
 /* Generic routine for getting common part of quota structure */
 static void do_get_dqblk(struct dquot *dquot, struct if_dqblk *di)
 {
 	struct mem_dqblk *dm = &dquot->dq_dqb;
 
 	spin_lock(&dq_data_lock);
-	di->dqb_bhardlimit = dm->dqb_bhardlimit;
-	di->dqb_bsoftlimit = dm->dqb_bsoftlimit;
+	di->dqb_bhardlimit = stoqb(dm->dqb_bhardlimit);
+	di->dqb_bsoftlimit = stoqb(dm->dqb_bsoftlimit);
 	di->dqb_curspace = dm->dqb_curspace;
 	di->dqb_ihardlimit = dm->dqb_ihardlimit;
 	di->dqb_isoftlimit = dm->dqb_isoftlimit;
@@ -1918,28 +2069,36 @@ static int do_set_dqblk(struct dquot *dquot, struct if_dqblk *di)
 	if (di->dqb_valid & QIF_SPACE) {
 		dm->dqb_curspace = di->dqb_curspace;
 		check_blim = 1;
+		__set_bit(DQ_LASTSET_B + QIF_SPACE_B, &dquot->dq_flags);
 	}
 	if (di->dqb_valid & QIF_BLIMITS) {
-		dm->dqb_bsoftlimit = di->dqb_bsoftlimit;
-		dm->dqb_bhardlimit = di->dqb_bhardlimit;
+		dm->dqb_bsoftlimit = qbtos(di->dqb_bsoftlimit);
+		dm->dqb_bhardlimit = qbtos(di->dqb_bhardlimit);
 		check_blim = 1;
+		__set_bit(DQ_LASTSET_B + QIF_BLIMITS_B, &dquot->dq_flags);
 	}
 	if (di->dqb_valid & QIF_INODES) {
 		dm->dqb_curinodes = di->dqb_curinodes;
 		check_ilim = 1;
+		__set_bit(DQ_LASTSET_B + QIF_INODES_B, &dquot->dq_flags);
 	}
 	if (di->dqb_valid & QIF_ILIMITS) {
 		dm->dqb_isoftlimit = di->dqb_isoftlimit;
 		dm->dqb_ihardlimit = di->dqb_ihardlimit;
 		check_ilim = 1;
+		__set_bit(DQ_LASTSET_B + QIF_ILIMITS_B, &dquot->dq_flags);
 	}
-	if (di->dqb_valid & QIF_BTIME)
+	if (di->dqb_valid & QIF_BTIME) {
 		dm->dqb_btime = di->dqb_btime;
-	if (di->dqb_valid & QIF_ITIME)
+		__set_bit(DQ_LASTSET_B + QIF_BTIME_B, &dquot->dq_flags);
+	}
+	if (di->dqb_valid & QIF_ITIME) {
 		dm->dqb_itime = di->dqb_itime;
+		__set_bit(DQ_LASTSET_B + QIF_ITIME_B, &dquot->dq_flags);
+	}
 
 	if (check_blim) {
-		if (!dm->dqb_bsoftlimit || toqb(dm->dqb_curspace) < dm->dqb_bsoftlimit) {
+		if (!dm->dqb_bsoftlimit || dm->dqb_curspace < dm->dqb_bsoftlimit) {
 			dm->dqb_btime = 0;
 			clear_bit(DQ_BLKS_B, &dquot->dq_flags);
 		}
@@ -1970,12 +2129,14 @@ int vfs_set_dqblk(struct super_block *sb, int type, qid_t id, struct if_dqblk *d
 	int rc;
 
 	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
-	if (!(dquot = dqget(sb, id, type))) {
-		mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
-		return -ESRCH;
+	dquot = dqget(sb, id, type);
+	if (!dquot) {
+		rc = -ESRCH;
+		goto out;
 	}
 	rc = do_set_dqblk(dquot, di);
 	dqput(dquot);
+out:
 	mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
 	return rc;
 }
@@ -1986,7 +2147,7 @@ int vfs_get_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii)
 	struct mem_dqinfo *mi;
   
 	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
-	if (!sb_has_quota_enabled(sb, type)) {
+	if (!sb_has_quota_active(sb, type)) {
 		mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
 		return -ESRCH;
 	}
@@ -2005,11 +2166,12 @@ int vfs_get_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii)
 int vfs_set_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii)
 {
 	struct mem_dqinfo *mi;
+	int err = 0;
 
 	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
-	if (!sb_has_quota_enabled(sb, type)) {
-		mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
-		return -ESRCH;
+	if (!sb_has_quota_active(sb, type)) {
+		err = -ESRCH;
+		goto out;
 	}
 	mi = sb_dqopt(sb)->info + type;
 	spin_lock(&dq_data_lock);
@@ -2023,8 +2185,9 @@ int vfs_set_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii)
 	mark_info_dirty(sb, type);
 	/* Force write to disk */
 	sb->dq_op->write_info(sb, type);
+out:
 	mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
-	return 0;
+	return err;
 }
 
 struct quotactl_ops vfs_quotactl_ops = {
@@ -2186,10 +2349,13 @@ EXPORT_SYMBOL(register_quota_format);
 EXPORT_SYMBOL(unregister_quota_format);
 EXPORT_SYMBOL(dqstats);
 EXPORT_SYMBOL(dq_data_lock);
+EXPORT_SYMBOL(vfs_quota_enable);
 EXPORT_SYMBOL(vfs_quota_on);
 EXPORT_SYMBOL(vfs_quota_on_path);
 EXPORT_SYMBOL(vfs_quota_on_mount);
+EXPORT_SYMBOL(vfs_quota_disable);
 EXPORT_SYMBOL(vfs_quota_off);
+EXPORT_SYMBOL(dquot_scan_active);
 EXPORT_SYMBOL(vfs_quota_sync);
 EXPORT_SYMBOL(vfs_get_dqinfo);
 EXPORT_SYMBOL(vfs_set_dqinfo);
@@ -2202,7 +2368,11 @@ EXPORT_SYMBOL(dquot_release);
 EXPORT_SYMBOL(dquot_mark_dquot_dirty);
 EXPORT_SYMBOL(dquot_initialize);
 EXPORT_SYMBOL(dquot_drop);
+EXPORT_SYMBOL(dquot_drop_locked);
 EXPORT_SYMBOL(vfs_dq_drop);
+EXPORT_SYMBOL(dqget);
+EXPORT_SYMBOL(dqput);
+EXPORT_SYMBOL(dquot_is_cached);
 EXPORT_SYMBOL(dquot_alloc_space);
 EXPORT_SYMBOL(dquot_alloc_inode);
 EXPORT_SYMBOL(dquot_free_space);

fs/ext3/super.c

@@ -713,7 +713,9 @@ static struct dquot_operations ext3_quota_operations = {
 	.acquire_dquot	= ext3_acquire_dquot,
 	.release_dquot	= ext3_release_dquot,
 	.mark_dirty	= ext3_mark_dquot_dirty,
-	.write_info	= ext3_write_info
+	.write_info	= ext3_write_info,
+	.alloc_dquot	= dquot_alloc,
+	.destroy_dquot	= dquot_destroy,
 };
 
 static struct quotactl_ops ext3_qctl_operations = {
@@ -1035,8 +1037,7 @@ static int parse_options (char *options, struct super_block *sb,
 		case Opt_grpjquota:
 			qtype = GRPQUOTA;
 set_qf_name:
-			if ((sb_any_quota_enabled(sb) ||
-			     sb_any_quota_suspended(sb)) &&
+			if (sb_any_quota_loaded(sb) &&
 			    !sbi->s_qf_names[qtype]) {
 				printk(KERN_ERR
 					"EXT3-fs: Cannot change journaled "
@@ -1075,8 +1076,7 @@ set_qf_name:
 		case Opt_offgrpjquota:
 			qtype = GRPQUOTA;
 clear_qf_name:
-			if ((sb_any_quota_enabled(sb) ||
-			     sb_any_quota_suspended(sb)) &&
+			if (sb_any_quota_loaded(sb) &&
 			    sbi->s_qf_names[qtype]) {
 				printk(KERN_ERR "EXT3-fs: Cannot change "
 					"journaled quota options when "
@@ -1095,8 +1095,7 @@ clear_qf_name:
 		case Opt_jqfmt_vfsv0:
 			qfmt = QFMT_VFS_V0;
 set_qf_format:
-			if ((sb_any_quota_enabled(sb) ||
-			     sb_any_quota_suspended(sb)) &&
+			if (sb_any_quota_loaded(sb) &&
 			    sbi->s_jquota_fmt != qfmt) {
 				printk(KERN_ERR "EXT3-fs: Cannot change "
 					"journaled quota options when "
@@ -1115,8 +1114,7 @@ set_qf_format:
 			set_opt(sbi->s_mount_opt, GRPQUOTA);
 			break;
 		case Opt_noquota:
-			if (sb_any_quota_enabled(sb) ||
-			    sb_any_quota_suspended(sb)) {
+			if (sb_any_quota_loaded(sb)) {
 				printk(KERN_ERR "EXT3-fs: Cannot change quota "
 					"options when quota turned on.\n");
 				return 0;

fs/ext4/super.c

@@ -803,7 +803,9 @@ static struct dquot_operations ext4_quota_operations = {
 	.acquire_dquot	= ext4_acquire_dquot,
 	.release_dquot	= ext4_release_dquot,
 	.mark_dirty	= ext4_mark_dquot_dirty,
-	.write_info	= ext4_write_info
+	.write_info	= ext4_write_info,
+	.alloc_dquot	= dquot_alloc,
+	.destroy_dquot	= dquot_destroy,
 };
 
 static struct quotactl_ops ext4_qctl_operations = {
@@ -1142,8 +1144,7 @@ static int parse_options(char *options, struct super_block *sb,
 		case Opt_grpjquota:
 			qtype = GRPQUOTA;
 set_qf_name:
-			if ((sb_any_quota_enabled(sb) ||
-			     sb_any_quota_suspended(sb)) &&
+			if (sb_any_quota_loaded(sb) &&
 			    !sbi->s_qf_names[qtype]) {
 				printk(KERN_ERR
 				       "EXT4-fs: Cannot change journaled "
@@ -1182,8 +1183,7 @@ set_qf_name:
 		case Opt_offgrpjquota:
 			qtype = GRPQUOTA;
 clear_qf_name:
-			if ((sb_any_quota_enabled(sb) ||
-			     sb_any_quota_suspended(sb)) &&
+			if (sb_any_quota_loaded(sb) &&
 			    sbi->s_qf_names[qtype]) {
 				printk(KERN_ERR "EXT4-fs: Cannot change "
 					"journaled quota options when "
@@ -1202,8 +1202,7 @@ clear_qf_name:
 		case Opt_jqfmt_vfsv0:
 			qfmt = QFMT_VFS_V0;
 set_qf_format:
-			if ((sb_any_quota_enabled(sb) ||
-			     sb_any_quota_suspended(sb)) &&
+			if (sb_any_quota_loaded(sb) &&
 			    sbi->s_jquota_fmt != qfmt) {
 				printk(KERN_ERR "EXT4-fs: Cannot change "
 					"journaled quota options when "
@@ -1222,7 +1221,7 @@ set_qf_format:
 			set_opt(sbi->s_mount_opt, GRPQUOTA);
 			break;
 		case Opt_noquota:
-			if (sb_any_quota_enabled(sb)) {
+			if (sb_any_quota_loaded(sb)) {
 				printk(KERN_ERR "EXT4-fs: Cannot change quota "
 					"options when quota turned on.\n");
 				return 0;

fs/jbd2/commit.c

@@ -509,6 +509,10 @@ void jbd2_journal_commit_transaction(journal_t *journal)
 		if (is_journal_aborted(journal)) {
 			clear_buffer_jbddirty(jh2bh(jh));
 			JBUFFER_TRACE(jh, "journal is aborting: refile");
+			jbd2_buffer_abort_trigger(jh,
+						  jh->b_frozen_data ?
+						  jh->b_frozen_triggers :
+						  jh->b_triggers);
 			jbd2_journal_refile_buffer(journal, jh);
 			/* If that was the last one, we need to clean up
 			 * any descriptor buffers which may have been
@@ -844,6 +848,9 @@ restart_loop:
 		 * data.
 		 *
 		 * Otherwise, we can just throw away the frozen data now.
+		 *
+		 * We also know that the frozen data has already fired
+		 * its triggers if they exist, so we can clear that too.
 		 */
 		if (jh->b_committed_data) {
 			jbd2_free(jh->b_committed_data, bh->b_size);
@@ -851,10 +858,12 @@ restart_loop:
 			if (jh->b_frozen_data) {
 				jh->b_committed_data = jh->b_frozen_data;
 				jh->b_frozen_data = NULL;
+				jh->b_frozen_triggers = NULL;
 			}
 		} else if (jh->b_frozen_data) {
 			jbd2_free(jh->b_frozen_data, bh->b_size);
 			jh->b_frozen_data = NULL;
+			jh->b_frozen_triggers = NULL;
 		}
 
 		spin_lock(&journal->j_list_lock);

fs/jbd2/journal.c

@@ -50,6 +50,7 @@ EXPORT_SYMBOL(jbd2_journal_unlock_updates);
 EXPORT_SYMBOL(jbd2_journal_get_write_access);
 EXPORT_SYMBOL(jbd2_journal_get_create_access);
 EXPORT_SYMBOL(jbd2_journal_get_undo_access);
+EXPORT_SYMBOL(jbd2_journal_set_triggers);
 EXPORT_SYMBOL(jbd2_journal_dirty_metadata);
 EXPORT_SYMBOL(jbd2_journal_release_buffer);
 EXPORT_SYMBOL(jbd2_journal_forget);
@@ -290,6 +291,7 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction,
 	struct page *new_page;
 	unsigned int new_offset;
 	struct buffer_head *bh_in = jh2bh(jh_in);
+	struct jbd2_buffer_trigger_type *triggers;
 
 	/*
 	 * The buffer really shouldn't be locked: only the current committing
@@ -314,12 +316,22 @@ repeat:
 		done_copy_out = 1;
 		new_page = virt_to_page(jh_in->b_frozen_data);
 		new_offset = offset_in_page(jh_in->b_frozen_data);
+		triggers = jh_in->b_frozen_triggers;
 	} else {
 		new_page = jh2bh(jh_in)->b_page;
 		new_offset = offset_in_page(jh2bh(jh_in)->b_data);
+		triggers = jh_in->b_triggers;
 	}
 
 	mapped_data = kmap_atomic(new_page, KM_USER0);
+	/*
+	 * Fire any commit trigger.  Do this before checking for escaping,
+	 * as the trigger may modify the magic offset.  If a copy-out
+	 * happens afterwards, it will have the correct data in the buffer.
+	 */
+	jbd2_buffer_commit_trigger(jh_in, mapped_data + new_offset,
+				   triggers);
+
 	/*
 	 * Check for escaping
 	 */
@@ -352,6 +364,13 @@ repeat:
 		new_page = virt_to_page(tmp);
 		new_offset = offset_in_page(tmp);
 		done_copy_out = 1;
+
+		/*
+		 * This isn't strictly necessary, as we're using frozen
+		 * data for the escaping, but it keeps consistency with
+		 * b_frozen_data usage.
+		 */
+		jh_in->b_frozen_triggers = jh_in->b_triggers;
 	}
 
 	/*

fs/jbd2/transaction.c

@@ -741,6 +741,12 @@ done:
 		source = kmap_atomic(page, KM_USER0);
 		memcpy(jh->b_frozen_data, source+offset, jh2bh(jh)->b_size);
 		kunmap_atomic(source, KM_USER0);
+
+		/*
+		 * Now that the frozen data is saved off, we need to store
+		 * any matching triggers.
+		 */
+		jh->b_frozen_triggers = jh->b_triggers;
 	}
 	jbd_unlock_bh_state(bh);
 
@@ -943,6 +949,47 @@ out:
 	return err;
 }
 
+/**
+ * void jbd2_journal_set_triggers() - Add triggers for commit writeout
+ * @bh: buffer to trigger on
+ * @type: struct jbd2_buffer_trigger_type containing the trigger(s).
+ *
+ * Set any triggers on this journal_head.  This is always safe, because
+ * triggers for a committing buffer will be saved off, and triggers for
+ * a running transaction will match the buffer in that transaction.
+ *
+ * Call with NULL to clear the triggers.
+ */
+void jbd2_journal_set_triggers(struct buffer_head *bh,
+			       struct jbd2_buffer_trigger_type *type)
+{
+	struct journal_head *jh = bh2jh(bh);
+
+	jh->b_triggers = type;
+}
+
+void jbd2_buffer_commit_trigger(struct journal_head *jh, void *mapped_data,
+				struct jbd2_buffer_trigger_type *triggers)
+{
+	struct buffer_head *bh = jh2bh(jh);
+
+	if (!triggers || !triggers->t_commit)
+		return;
+
+	triggers->t_commit(triggers, bh, mapped_data, bh->b_size);
+}
+
+void jbd2_buffer_abort_trigger(struct journal_head *jh,
+			       struct jbd2_buffer_trigger_type *triggers)
+{
+	if (!triggers || !triggers->t_abort)
+		return;
+
+	triggers->t_abort(triggers, jh2bh(jh));
+}
+
+
+
 /**
  * int jbd2_journal_dirty_metadata() -  mark a buffer as containing dirty metadata
  * @handle: transaction to add buffer to.

fs/ocfs2/Makefile

@@ -12,6 +12,7 @@ obj-$(CONFIG_OCFS2_FS_USERSPACE_CLUSTER) += ocfs2_stack_user.o
 ocfs2-objs := \
 	alloc.o 		\
 	aops.o 			\
+	blockcheck.o		\
 	buffer_head_io.o	\
 	dcache.o 		\
 	dir.o 			\
@@ -35,8 +36,14 @@ ocfs2-objs := \
 	sysfile.o 		\
 	uptodate.o		\
 	ver.o			\
+	quota_local.o		\
+	quota_global.o		\
 	xattr.o
 
+ifeq ($(CONFIG_OCFS2_FS_POSIX_ACL),y)
+ocfs2-objs += acl.o
+endif
+
 ocfs2_stackglue-objs := stackglue.o
 ocfs2_stack_o2cb-objs := stack_o2cb.o
 ocfs2_stack_user-objs := stack_user.o

fs/ocfs2/acl.c

@@ -0,0 +1,479 @@
+/* -*- mode: c; c-basic-offset: 8; -*-
+ * vim: noexpandtab sw=8 ts=8 sts=0:
+ *
+ * acl.c
+ *
+ * Copyright (C) 2004, 2008 Oracle.  All rights reserved.
+ *
+ * CREDITS:
+ * Lots of code in this file is copy from linux/fs/ext3/acl.c.
+ * Copyright (C) 2001-2003 Andreas Gruenbacher, <agruen@suse.de>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/string.h>
+
+#define MLOG_MASK_PREFIX ML_INODE
+#include <cluster/masklog.h>
+
+#include "ocfs2.h"
+#include "alloc.h"
+#include "dlmglue.h"
+#include "file.h"
+#include "ocfs2_fs.h"
+
+#include "xattr.h"
+#include "acl.h"
+
+/*
+ * Convert from xattr value to acl struct.
+ */
+static struct posix_acl *ocfs2_acl_from_xattr(const void *value, size_t size)
+{
+	int n, count;
+	struct posix_acl *acl;
+
+	if (!value)
+		return NULL;
+	if (size < sizeof(struct posix_acl_entry))
+		return ERR_PTR(-EINVAL);
+
+	count = size / sizeof(struct posix_acl_entry);
+	if (count < 0)
+		return ERR_PTR(-EINVAL);
+	if (count == 0)
+		return NULL;
+
+	acl = posix_acl_alloc(count, GFP_NOFS);
+	if (!acl)
+		return ERR_PTR(-ENOMEM);
+	for (n = 0; n < count; n++) {
+		struct ocfs2_acl_entry *entry =
+			(struct ocfs2_acl_entry *)value;
+
+		acl->a_entries[n].e_tag  = le16_to_cpu(entry->e_tag);
+		acl->a_entries[n].e_perm = le16_to_cpu(entry->e_perm);
+		acl->a_entries[n].e_id   = le32_to_cpu(entry->e_id);
+		value += sizeof(struct posix_acl_entry);
+
+	}
+	return acl;
+}
+
+/*
+ * Convert acl struct to xattr value.
+ */
+static void *ocfs2_acl_to_xattr(const struct posix_acl *acl, size_t *size)
+{
+	struct ocfs2_acl_entry *entry = NULL;
+	char *ocfs2_acl;
+	size_t n;
+
+	*size = acl->a_count * sizeof(struct posix_acl_entry);
+
+	ocfs2_acl = kmalloc(*size, GFP_NOFS);
+	if (!ocfs2_acl)
+		return ERR_PTR(-ENOMEM);
+
+	entry = (struct ocfs2_acl_entry *)ocfs2_acl;
+	for (n = 0; n < acl->a_count; n++, entry++) {
+		entry->e_tag  = cpu_to_le16(acl->a_entries[n].e_tag);
+		entry->e_perm = cpu_to_le16(acl->a_entries[n].e_perm);
+		entry->e_id   = cpu_to_le32(acl->a_entries[n].e_id);
+	}
+	return ocfs2_acl;
+}
+
+static struct posix_acl *ocfs2_get_acl_nolock(struct inode *inode,
+					      int type,
+					      struct buffer_head *di_bh)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	int name_index;
+	char *value = NULL;
+	struct posix_acl *acl;
+	int retval;
+
+	if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+		return NULL;
+
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		name_index = OCFS2_XATTR_INDEX_POSIX_ACL_ACCESS;
+		break;
+	case ACL_TYPE_DEFAULT:
+		name_index = OCFS2_XATTR_INDEX_POSIX_ACL_DEFAULT;
+		break;
+	default:
+		return ERR_PTR(-EINVAL);
+	}
+
+	retval = ocfs2_xattr_get_nolock(inode, di_bh, name_index, "", NULL, 0);
+	if (retval > 0) {
+		value = kmalloc(retval, GFP_NOFS);
+		if (!value)
+			return ERR_PTR(-ENOMEM);
+		retval = ocfs2_xattr_get_nolock(inode, di_bh, name_index,
+						"", value, retval);
+	}
+
+	if (retval > 0)
+		acl = ocfs2_acl_from_xattr(value, retval);
+	else if (retval == -ENODATA || retval == 0)
+		acl = NULL;
+	else
+		acl = ERR_PTR(retval);
+
+	kfree(value);
+
+	return acl;
+}
+
+
+/*
+ * Get posix acl.
+ */
+static struct posix_acl *ocfs2_get_acl(struct inode *inode, int type)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct buffer_head *di_bh = NULL;
+	struct posix_acl *acl;
+	int ret;
+
+	if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+		return NULL;
+
+	ret = ocfs2_inode_lock(inode, &di_bh, 0);
+	if (ret < 0) {
+		mlog_errno(ret);
+		acl = ERR_PTR(ret);
+		return acl;
+	}
+
+	acl = ocfs2_get_acl_nolock(inode, type, di_bh);
+
+	ocfs2_inode_unlock(inode, 0);
+
+	brelse(di_bh);
+
+	return acl;
+}
+
+/*
+ * Set the access or default ACL of an inode.
+ */
+static int ocfs2_set_acl(handle_t *handle,
+			 struct inode *inode,
+			 struct buffer_head *di_bh,
+			 int type,
+			 struct posix_acl *acl,
+			 struct ocfs2_alloc_context *meta_ac,
+			 struct ocfs2_alloc_context *data_ac)
+{
+	int name_index;
+	void *value = NULL;
+	size_t size = 0;
+	int ret;
+
+	if (S_ISLNK(inode->i_mode))
+		return -EOPNOTSUPP;
+
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		name_index = OCFS2_XATTR_INDEX_POSIX_ACL_ACCESS;
+		if (acl) {
+			mode_t mode = inode->i_mode;
+			ret = posix_acl_equiv_mode(acl, &mode);
+			if (ret < 0)
+				return ret;
+			else {
+				inode->i_mode = mode;
+				if (ret == 0)
+					acl = NULL;
+			}
+		}
+		break;
+	case ACL_TYPE_DEFAULT:
+		name_index = OCFS2_XATTR_INDEX_POSIX_ACL_DEFAULT;
+		if (!S_ISDIR(inode->i_mode))
+			return acl ? -EACCES : 0;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (acl) {
+		value = ocfs2_acl_to_xattr(acl, &size);
+		if (IS_ERR(value))
+			return (int)PTR_ERR(value);
+	}
+
+	if (handle)
+		ret = ocfs2_xattr_set_handle(handle, inode, di_bh, name_index,
+					     "", value, size, 0,
+					     meta_ac, data_ac);
+	else
+		ret = ocfs2_xattr_set(inode, name_index, "", value, size, 0);
+
+	kfree(value);
+
+	return ret;
+}
+
+int ocfs2_check_acl(struct inode *inode, int mask)
+{
+	struct posix_acl *acl = ocfs2_get_acl(inode, ACL_TYPE_ACCESS);
+
+	if (IS_ERR(acl))
+		return PTR_ERR(acl);
+	if (acl) {
+		int ret = posix_acl_permission(inode, acl, mask);
+		posix_acl_release(acl);
+		return ret;
+	}
+
+	return -EAGAIN;
+}
+
+int ocfs2_acl_chmod(struct inode *inode)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct posix_acl *acl, *clone;
+	int ret;
+
+	if (S_ISLNK(inode->i_mode))
+		return -EOPNOTSUPP;
+
+	if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+		return 0;
+
+	acl = ocfs2_get_acl(inode, ACL_TYPE_ACCESS);
+	if (IS_ERR(acl) || !acl)
+		return PTR_ERR(acl);
+	clone = posix_acl_clone(acl, GFP_KERNEL);
+	posix_acl_release(acl);
+	if (!clone)
+		return -ENOMEM;
+	ret = posix_acl_chmod_masq(clone, inode->i_mode);
+	if (!ret)
+		ret = ocfs2_set_acl(NULL, inode, NULL, ACL_TYPE_ACCESS,
+				    clone, NULL, NULL);
+	posix_acl_release(clone);
+	return ret;
+}
+
+/*
+ * Initialize the ACLs of a new inode. If parent directory has default ACL,
+ * then clone to new inode. Called from ocfs2_mknod.
+ */
+int ocfs2_init_acl(handle_t *handle,
+		   struct inode *inode,
+		   struct inode *dir,
+		   struct buffer_head *di_bh,
+		   struct buffer_head *dir_bh,
+		   struct ocfs2_alloc_context *meta_ac,
+		   struct ocfs2_alloc_context *data_ac)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct posix_acl *acl = NULL;
+	int ret = 0;
+
+	if (!S_ISLNK(inode->i_mode)) {
+		if (osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) {
+			acl = ocfs2_get_acl_nolock(dir, ACL_TYPE_DEFAULT,
+						   dir_bh);
+			if (IS_ERR(acl))
+				return PTR_ERR(acl);
+		}
+		if (!acl)
+			inode->i_mode &= ~current->fs->umask;
+	}
+	if ((osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) && acl) {
+		struct posix_acl *clone;
+		mode_t mode;
+
+		if (S_ISDIR(inode->i_mode)) {
+			ret = ocfs2_set_acl(handle, inode, di_bh,
+					    ACL_TYPE_DEFAULT, acl,
+					    meta_ac, data_ac);
+			if (ret)
+				goto cleanup;
+		}
+		clone = posix_acl_clone(acl, GFP_NOFS);
+		ret = -ENOMEM;
+		if (!clone)
+			goto cleanup;
+
+		mode = inode->i_mode;
+		ret = posix_acl_create_masq(clone, &mode);
+		if (ret >= 0) {
+			inode->i_mode = mode;
+			if (ret > 0) {
+				ret = ocfs2_set_acl(handle, inode,
+						    di_bh, ACL_TYPE_ACCESS,
+						    clone, meta_ac, data_ac);
+			}
+		}
+		posix_acl_release(clone);
+	}
+cleanup:
+	posix_acl_release(acl);
+	return ret;
+}
+
+static size_t ocfs2_xattr_list_acl_access(struct inode *inode,
+					  char *list,
+					  size_t list_len,
+					  const char *name,
+					  size_t name_len)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	const size_t size = sizeof(POSIX_ACL_XATTR_ACCESS);
+
+	if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+		return 0;
+
+	if (list && size <= list_len)
+		memcpy(list, POSIX_ACL_XATTR_ACCESS, size);
+	return size;
+}
+
+static size_t ocfs2_xattr_list_acl_default(struct inode *inode,
+					   char *list,
+					   size_t list_len,
+					   const char *name,
+					   size_t name_len)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	const size_t size = sizeof(POSIX_ACL_XATTR_DEFAULT);
+
+	if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+		return 0;
+
+	if (list && size <= list_len)
+		memcpy(list, POSIX_ACL_XATTR_DEFAULT, size);
+	return size;
+}
+
+static int ocfs2_xattr_get_acl(struct inode *inode,
+			       int type,
+			       void *buffer,
+			       size_t size)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct posix_acl *acl;
+	int ret;
+
+	if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+		return -EOPNOTSUPP;
+
+	acl = ocfs2_get_acl(inode, type);
+	if (IS_ERR(acl))
+		return PTR_ERR(acl);
+	if (acl == NULL)
+		return -ENODATA;
+	ret = posix_acl_to_xattr(acl, buffer, size);
+	posix_acl_release(acl);
+
+	return ret;
+}
+
+static int ocfs2_xattr_get_acl_access(struct inode *inode,
+				      const char *name,
+				      void *buffer,
+				      size_t size)
+{
+	if (strcmp(name, "") != 0)
+		return -EINVAL;
+	return ocfs2_xattr_get_acl(inode, ACL_TYPE_ACCESS, buffer, size);
+}
+
+static int ocfs2_xattr_get_acl_default(struct inode *inode,
+				       const char *name,
+				       void *buffer,
+				       size_t size)
+{
+	if (strcmp(name, "") != 0)
+		return -EINVAL;
+	return ocfs2_xattr_get_acl(inode, ACL_TYPE_DEFAULT, buffer, size);
+}
+
+static int ocfs2_xattr_set_acl(struct inode *inode,
+			       int type,
+			       const void *value,
+			       size_t size)
+{
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct posix_acl *acl;
+	int ret = 0;
+
+	if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
+		return -EOPNOTSUPP;
+
+	if (!is_owner_or_cap(inode))
+		return -EPERM;
+
+	if (value) {
+		acl = posix_acl_from_xattr(value, size);
+		if (IS_ERR(acl))
+			return PTR_ERR(acl);
+		else if (acl) {
+			ret = posix_acl_valid(acl);
+			if (ret)
+				goto cleanup;
+		}
+	} else
+		acl = NULL;
+
+	ret = ocfs2_set_acl(NULL, inode, NULL, type, acl, NULL, NULL);
+
+cleanup:
+	posix_acl_release(acl);
+	return ret;
+}
+
+static int ocfs2_xattr_set_acl_access(struct inode *inode,
+				      const char *name,
+				      const void *value,
+				      size_t size,
+				      int flags)
+{
+	if (strcmp(name, "") != 0)
+		return -EINVAL;
+	return ocfs2_xattr_set_acl(inode, ACL_TYPE_ACCESS, value, size);
+}
+
+static int ocfs2_xattr_set_acl_default(struct inode *inode,
+				       const char *name,
+				       const void *value,
+				       size_t size,
+				       int flags)
+{
+	if (strcmp(name, "") != 0)
+		return -EINVAL;
+	return ocfs2_xattr_set_acl(inode, ACL_TYPE_DEFAULT, value, size);
+}
+
+struct xattr_handler ocfs2_xattr_acl_access_handler = {
+	.prefix	= POSIX_ACL_XATTR_ACCESS,
+	.list	= ocfs2_xattr_list_acl_access,
+	.get	= ocfs2_xattr_get_acl_access,
+	.set	= ocfs2_xattr_set_acl_access,
+};
+
+struct xattr_handler ocfs2_xattr_acl_default_handler = {
+	.prefix	= POSIX_ACL_XATTR_DEFAULT,
+	.list	= ocfs2_xattr_list_acl_default,
+	.get	= ocfs2_xattr_get_acl_default,
+	.set	= ocfs2_xattr_set_acl_default,
+};

fs/ocfs2/acl.h

@@ -0,0 +1,58 @@
+/* -*- mode: c; c-basic-offset: 8; -*-
+ * vim: noexpandtab sw=8 ts=8 sts=0:
+ *
+ * acl.h
+ *
+ * Copyright (C) 2004, 2008 Oracle.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef OCFS2_ACL_H
+#define OCFS2_ACL_H
+
+#include <linux/posix_acl_xattr.h>
+
+struct ocfs2_acl_entry {
+	__le16 e_tag;
+	__le16 e_perm;
+	__le32 e_id;
+};
+
+#ifdef CONFIG_OCFS2_FS_POSIX_ACL
+
+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 *,
+			  struct ocfs2_alloc_context *,
+			  struct ocfs2_alloc_context *);
+
+#else /* CONFIG_OCFS2_FS_POSIX_ACL*/
+
+#define ocfs2_check_acl NULL
+static inline int ocfs2_acl_chmod(struct inode *inode)
+{
+	return 0;
+}
+static inline int ocfs2_init_acl(handle_t *handle,
+				 struct inode *inode,
+				 struct inode *dir,
+				 struct buffer_head *di_bh,
+				 struct buffer_head *dir_bh,
+				 struct ocfs2_alloc_context *meta_ac,
+				 struct ocfs2_alloc_context *data_ac)
+{
+	return 0;
+}
+
+#endif /* CONFIG_OCFS2_FS_POSIX_ACL*/
+
+#endif /* OCFS2_ACL_H */

fs/ocfs2/alloc.c

@@ -28,6 +28,7 @@
 #include <linux/slab.h>
 #include <linux/highmem.h>
 #include <linux/swap.h>
+#include <linux/quotaops.h>
 
 #define MLOG_MASK_PREFIX ML_DISK_ALLOC
 #include <cluster/masklog.h>
@@ -36,6 +37,7 @@
 
 #include "alloc.h"
 #include "aops.h"
+#include "blockcheck.h"
 #include "dlmglue.h"
 #include "extent_map.h"
 #include "inode.h"
@@ -46,6 +48,7 @@
 #include "file.h"
 #include "super.h"
 #include "uptodate.h"
+#include "xattr.h"
 
 #include "buffer_head_io.h"
 
@@ -187,20 +190,12 @@ static int ocfs2_dinode_insert_check(struct inode *inode,
 static int ocfs2_dinode_sanity_check(struct inode *inode,
 				     struct ocfs2_extent_tree *et)
 {
-	int ret = 0;
-	struct ocfs2_dinode *di;
+	struct ocfs2_dinode *di = et->et_object;
 
 	BUG_ON(et->et_ops != &ocfs2_dinode_et_ops);
+	BUG_ON(!OCFS2_IS_VALID_DINODE(di));
 
-	di = et->et_object;
-	if (!OCFS2_IS_VALID_DINODE(di)) {
-		ret = -EIO;
-		ocfs2_error(inode->i_sb,
-			"Inode %llu has invalid path root",
-			(unsigned long long)OCFS2_I(inode)->ip_blkno);
-	}
-
-	return ret;
+	return 0;
 }
 
 static void ocfs2_dinode_fill_root_el(struct ocfs2_extent_tree *et)
@@ -213,36 +208,33 @@ static void ocfs2_dinode_fill_root_el(struct ocfs2_extent_tree *et)
 
 static void ocfs2_xattr_value_fill_root_el(struct ocfs2_extent_tree *et)
 {
-	struct ocfs2_xattr_value_root *xv = et->et_object;
+	struct ocfs2_xattr_value_buf *vb = et->et_object;
 
-	et->et_root_el = &xv->xr_list;
+	et->et_root_el = &vb->vb_xv->xr_list;
 }
 
 static void ocfs2_xattr_value_set_last_eb_blk(struct ocfs2_extent_tree *et,
 					      u64 blkno)
 {
-	struct ocfs2_xattr_value_root *xv =
-		(struct ocfs2_xattr_value_root *)et->et_object;
+	struct ocfs2_xattr_value_buf *vb = et->et_object;
 
-	xv->xr_last_eb_blk = cpu_to_le64(blkno);
+	vb->vb_xv->xr_last_eb_blk = cpu_to_le64(blkno);
 }
 
 static u64 ocfs2_xattr_value_get_last_eb_blk(struct ocfs2_extent_tree *et)
 {
-	struct ocfs2_xattr_value_root *xv =
-		(struct ocfs2_xattr_value_root *) et->et_object;
+	struct ocfs2_xattr_value_buf *vb = et->et_object;
 
-	return le64_to_cpu(xv->xr_last_eb_blk);
+	return le64_to_cpu(vb->vb_xv->xr_last_eb_blk);
 }
 
 static void ocfs2_xattr_value_update_clusters(struct inode *inode,
 					      struct ocfs2_extent_tree *et,
 					      u32 clusters)
 {
-	struct ocfs2_xattr_value_root *xv =
-		(struct ocfs2_xattr_value_root *)et->et_object;
+	struct ocfs2_xattr_value_buf *vb = et->et_object;
 
-	le32_add_cpu(&xv->xr_clusters, clusters);
+	le32_add_cpu(&vb->vb_xv->xr_clusters, clusters);
 }
 
 static struct ocfs2_extent_tree_operations ocfs2_xattr_value_et_ops = {
@@ -304,11 +296,13 @@ static struct ocfs2_extent_tree_operations ocfs2_xattr_tree_et_ops = {
 static void __ocfs2_init_extent_tree(struct ocfs2_extent_tree *et,
 				     struct inode *inode,
 				     struct buffer_head *bh,
+				     ocfs2_journal_access_func access,
 				     void *obj,
 				     struct ocfs2_extent_tree_operations *ops)
 {
 	et->et_ops = ops;
 	et->et_root_bh = bh;
+	et->et_root_journal_access = access;
 	if (!obj)
 		obj = (void *)bh->b_data;
 	et->et_object = obj;
@@ -324,23 +318,23 @@ void ocfs2_init_dinode_extent_tree(struct ocfs2_extent_tree *et,
 				   struct inode *inode,
 				   struct buffer_head *bh)
 {
-	__ocfs2_init_extent_tree(et, inode, bh, NULL, &ocfs2_dinode_et_ops);
+	__ocfs2_init_extent_tree(et, inode, bh, ocfs2_journal_access_di,
+				 NULL, &ocfs2_dinode_et_ops);
 }
 
 void ocfs2_init_xattr_tree_extent_tree(struct ocfs2_extent_tree *et,
 				       struct inode *inode,
 				       struct buffer_head *bh)
 {
-	__ocfs2_init_extent_tree(et, inode, bh, NULL,
-				 &ocfs2_xattr_tree_et_ops);
+	__ocfs2_init_extent_tree(et, inode, bh, ocfs2_journal_access_xb,
+				 NULL, &ocfs2_xattr_tree_et_ops);
 }
 
 void ocfs2_init_xattr_value_extent_tree(struct ocfs2_extent_tree *et,
 					struct inode *inode,
-					struct buffer_head *bh,
-					struct ocfs2_xattr_value_root *xv)
+					struct ocfs2_xattr_value_buf *vb)
 {
-	__ocfs2_init_extent_tree(et, inode, bh, xv,
+	__ocfs2_init_extent_tree(et, inode, vb->vb_bh, vb->vb_access, vb,
 				 &ocfs2_xattr_value_et_ops);
 }
 
@@ -362,6 +356,15 @@ static inline void ocfs2_et_update_clusters(struct inode *inode,
 	et->et_ops->eo_update_clusters(inode, et, clusters);
 }
 
+static inline int ocfs2_et_root_journal_access(handle_t *handle,
+					       struct inode *inode,
+					       struct ocfs2_extent_tree *et,
+					       int type)
+{
+	return et->et_root_journal_access(handle, inode, et->et_root_bh,
+					  type);
+}
+
 static inline int ocfs2_et_insert_check(struct inode *inode,
 					struct ocfs2_extent_tree *et,
 					struct ocfs2_extent_rec *rec)
@@ -402,12 +405,14 @@ struct ocfs2_path_item {
 #define OCFS2_MAX_PATH_DEPTH	5
 
 struct ocfs2_path {
-	int			p_tree_depth;
-	struct ocfs2_path_item	p_node[OCFS2_MAX_PATH_DEPTH];
+	int				p_tree_depth;
+	ocfs2_journal_access_func	p_root_access;
+	struct ocfs2_path_item		p_node[OCFS2_MAX_PATH_DEPTH];
 };
 
 #define path_root_bh(_path) ((_path)->p_node[0].bh)
 #define path_root_el(_path) ((_path)->p_node[0].el)
+#define path_root_access(_path)((_path)->p_root_access)
 #define path_leaf_bh(_path) ((_path)->p_node[(_path)->p_tree_depth].bh)
 #define path_leaf_el(_path) ((_path)->p_node[(_path)->p_tree_depth].el)
 #define path_num_items(_path) ((_path)->p_tree_depth + 1)
@@ -440,6 +445,8 @@ static void ocfs2_reinit_path(struct ocfs2_path *path, int keep_root)
 	 */
 	if (keep_root)
 		depth = le16_to_cpu(path_root_el(path)->l_tree_depth);
+	else
+		path_root_access(path) = NULL;
 
 	path->p_tree_depth = depth;
 }
@@ -465,6 +472,7 @@ static void ocfs2_cp_path(struct ocfs2_path *dest, struct ocfs2_path *src)
 
 	BUG_ON(path_root_bh(dest) != path_root_bh(src));
 	BUG_ON(path_root_el(dest) != path_root_el(src));
+	BUG_ON(path_root_access(dest) != path_root_access(src));
 
 	ocfs2_reinit_path(dest, 1);
 
@@ -486,6 +494,7 @@ static void ocfs2_mv_path(struct ocfs2_path *dest, struct ocfs2_path *src)
 	int i;
 
 	BUG_ON(path_root_bh(dest) != path_root_bh(src));
+	BUG_ON(path_root_access(dest) != path_root_access(src));
 
 	for(i = 1; i < OCFS2_MAX_PATH_DEPTH; i++) {
 		brelse(dest->p_node[i].bh);
@@ -521,7 +530,8 @@ static inline void ocfs2_path_insert_eb(struct ocfs2_path *path, int index,
 }
 
 static struct ocfs2_path *ocfs2_new_path(struct buffer_head *root_bh,
-					 struct ocfs2_extent_list *root_el)
+					 struct ocfs2_extent_list *root_el,
+					 ocfs2_journal_access_func access)
 {
 	struct ocfs2_path *path;
 
@@ -533,11 +543,48 @@ static struct ocfs2_path *ocfs2_new_path(struct buffer_head *root_bh,
 		get_bh(root_bh);
 		path_root_bh(path) = root_bh;
 		path_root_el(path) = root_el;
+		path_root_access(path) = access;
 	}
 
 	return path;
 }
 
+static struct ocfs2_path *ocfs2_new_path_from_path(struct ocfs2_path *path)
+{
+	return ocfs2_new_path(path_root_bh(path), path_root_el(path),
+			      path_root_access(path));
+}
+
+static struct ocfs2_path *ocfs2_new_path_from_et(struct ocfs2_extent_tree *et)
+{
+	return ocfs2_new_path(et->et_root_bh, et->et_root_el,
+			      et->et_root_journal_access);
+}
+
+/*
+ * Journal the buffer at depth idx.  All idx>0 are extent_blocks,
+ * otherwise it's the root_access function.
+ *
+ * I don't like the way this function's name looks next to
+ * ocfs2_journal_access_path(), but I don't have a better one.
+ */
+static int ocfs2_path_bh_journal_access(handle_t *handle,
+					struct inode *inode,
+					struct ocfs2_path *path,
+					int idx)
+{
+	ocfs2_journal_access_func access = path_root_access(path);
+
+	if (!access)
+		access = ocfs2_journal_access;
+
+	if (idx)
+		access = ocfs2_journal_access_eb;
+
+	return access(handle, inode, path->p_node[idx].bh,
+		      OCFS2_JOURNAL_ACCESS_WRITE);
+}
+
 /*
  * Convenience function to journal all components in a path.
  */
@@ -550,8 +597,7 @@ static int ocfs2_journal_access_path(struct inode *inode, handle_t *handle,
 		goto out;
 
 	for(i = 0; i < path_num_items(path); i++) {
-		ret = ocfs2_journal_access(handle, inode, path->p_node[i].bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode, path, i);
 		if (ret < 0) {
 			mlog_errno(ret);
 			goto out;
@@ -686,6 +732,80 @@ struct ocfs2_merge_ctxt {
 	int			c_split_covers_rec;
 };
 
+static int ocfs2_validate_extent_block(struct super_block *sb,
+				       struct buffer_head *bh)
+{
+	int rc;
+	struct ocfs2_extent_block *eb =
+		(struct ocfs2_extent_block *)bh->b_data;
+
+	mlog(0, "Validating extent block %llu\n",
+	     (unsigned long long)bh->b_blocknr);
+
+	BUG_ON(!buffer_uptodate(bh));
+
+	/*
+	 * If the ecc fails, we return the error but otherwise
+	 * leave the filesystem running.  We know any error is
+	 * local to this block.
+	 */
+	rc = ocfs2_validate_meta_ecc(sb, bh->b_data, &eb->h_check);
+	if (rc) {
+		mlog(ML_ERROR, "Checksum failed for extent block %llu\n",
+		     (unsigned long long)bh->b_blocknr);
+		return rc;
+	}
+
+	/*
+	 * Errors after here are fatal.
+	 */
+
+	if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
+		ocfs2_error(sb,
+			    "Extent block #%llu has bad signature %.*s",
+			    (unsigned long long)bh->b_blocknr, 7,
+			    eb->h_signature);
+		return -EINVAL;
+	}
+
+	if (le64_to_cpu(eb->h_blkno) != bh->b_blocknr) {
+		ocfs2_error(sb,
+			    "Extent block #%llu has an invalid h_blkno "
+			    "of %llu",
+			    (unsigned long long)bh->b_blocknr,
+			    (unsigned long long)le64_to_cpu(eb->h_blkno));
+		return -EINVAL;
+	}
+
+	if (le32_to_cpu(eb->h_fs_generation) != OCFS2_SB(sb)->fs_generation) {
+		ocfs2_error(sb,
+			    "Extent block #%llu has an invalid "
+			    "h_fs_generation of #%u",
+			    (unsigned long long)bh->b_blocknr,
+			    le32_to_cpu(eb->h_fs_generation));
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+int ocfs2_read_extent_block(struct inode *inode, u64 eb_blkno,
+			    struct buffer_head **bh)
+{
+	int rc;
+	struct buffer_head *tmp = *bh;
+
+	rc = ocfs2_read_block(inode, eb_blkno, &tmp,
+			      ocfs2_validate_extent_block);
+
+	/* If ocfs2_read_block() got us a new bh, pass it up. */
+	if (!rc && !*bh)
+		*bh = tmp;
+
+	return rc;
+}
+
+
 /*
  * How many free extents have we got before we need more meta data?
  */
@@ -705,8 +825,7 @@ int ocfs2_num_free_extents(struct ocfs2_super *osb,
 	last_eb_blk = ocfs2_et_get_last_eb_blk(et);
 
 	if (last_eb_blk) {
-		retval = ocfs2_read_block(inode, last_eb_blk,
-					  &eb_bh);
+		retval = ocfs2_read_extent_block(inode, last_eb_blk, &eb_bh);
 		if (retval < 0) {
 			mlog_errno(retval);
 			goto bail;
@@ -768,8 +887,8 @@ static int ocfs2_create_new_meta_bhs(struct ocfs2_super *osb,
 			}
 			ocfs2_set_new_buffer_uptodate(inode, bhs[i]);
 
-			status = ocfs2_journal_access(handle, inode, bhs[i],
-						      OCFS2_JOURNAL_ACCESS_CREATE);
+			status = ocfs2_journal_access_eb(handle, inode, bhs[i],
+							 OCFS2_JOURNAL_ACCESS_CREATE);
 			if (status < 0) {
 				mlog_errno(status);
 				goto bail;
@@ -908,15 +1027,12 @@ static int ocfs2_add_branch(struct ocfs2_super *osb,
 	for(i = 0; i < new_blocks; i++) {
 		bh = new_eb_bhs[i];
 		eb = (struct ocfs2_extent_block *) bh->b_data;
-		if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-			OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-			status = -EIO;
-			goto bail;
-		}
+		/* ocfs2_create_new_meta_bhs() should create it right! */
+		BUG_ON(!OCFS2_IS_VALID_EXTENT_BLOCK(eb));
 		eb_el = &eb->h_list;
 
-		status = ocfs2_journal_access(handle, inode, bh,
-					      OCFS2_JOURNAL_ACCESS_CREATE);
+		status = ocfs2_journal_access_eb(handle, inode, bh,
+						 OCFS2_JOURNAL_ACCESS_CREATE);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
@@ -955,21 +1071,21 @@ static int ocfs2_add_branch(struct ocfs2_super *osb,
 	 * journal_dirty erroring as it won't unless we've aborted the
 	 * handle (in which case we would never be here) so reserving
 	 * the write with journal_access is all we need to do. */
-	status = ocfs2_journal_access(handle, inode, *last_eb_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_eb(handle, inode, *last_eb_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
-	status = ocfs2_journal_access(handle, inode, et->et_root_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_et_root_journal_access(handle, inode, et,
+					      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
 	if (eb_bh) {
-		status = ocfs2_journal_access(handle, inode, eb_bh,
-					      OCFS2_JOURNAL_ACCESS_WRITE);
+		status = ocfs2_journal_access_eb(handle, inode, eb_bh,
+						 OCFS2_JOURNAL_ACCESS_WRITE);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
@@ -1052,17 +1168,14 @@ static int ocfs2_shift_tree_depth(struct ocfs2_super *osb,
 	}
 
 	eb = (struct ocfs2_extent_block *) new_eb_bh->b_data;
-	if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-		OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-		status = -EIO;
-		goto bail;
-	}
+	/* ocfs2_create_new_meta_bhs() should create it right! */
+	BUG_ON(!OCFS2_IS_VALID_EXTENT_BLOCK(eb));
 
 	eb_el = &eb->h_list;
 	root_el = et->et_root_el;
 
-	status = ocfs2_journal_access(handle, inode, new_eb_bh,
-				      OCFS2_JOURNAL_ACCESS_CREATE);
+	status = ocfs2_journal_access_eb(handle, inode, new_eb_bh,
+					 OCFS2_JOURNAL_ACCESS_CREATE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -1080,8 +1193,8 @@ static int ocfs2_shift_tree_depth(struct ocfs2_super *osb,
 		goto bail;
 	}
 
-	status = ocfs2_journal_access(handle, inode, et->et_root_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_et_root_journal_access(handle, inode, et,
+					      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -1176,18 +1289,13 @@ static int ocfs2_find_branch_target(struct ocfs2_super *osb,
 		brelse(bh);
 		bh = NULL;
 
-		status = ocfs2_read_block(inode, blkno, &bh);
+		status = ocfs2_read_extent_block(inode, blkno, &bh);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
 		}
 
 		eb = (struct ocfs2_extent_block *) bh->b_data;
-		if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-			OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-			status = -EIO;
-			goto bail;
-		}
 		el = &eb->h_list;
 
 		if (le16_to_cpu(el->l_next_free_rec) <
@@ -1540,7 +1648,7 @@ static int __ocfs2_find_path(struct inode *inode,
 
 		brelse(bh);
 		bh = NULL;
-		ret = ocfs2_read_block(inode, blkno, &bh);
+		ret = ocfs2_read_extent_block(inode, blkno, &bh);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -1548,11 +1656,6 @@ static int __ocfs2_find_path(struct inode *inode,
 
 		eb = (struct ocfs2_extent_block *) bh->b_data;
 		el = &eb->h_list;
-		if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-			OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-			ret = -EIO;
-			goto out;
-		}
 
 		if (le16_to_cpu(el->l_next_free_rec) >
 		    le16_to_cpu(el->l_count)) {
@@ -1860,25 +1963,23 @@ static int ocfs2_rotate_subtree_right(struct inode *inode,
 	root_bh = left_path->p_node[subtree_index].bh;
 	BUG_ON(root_bh != right_path->p_node[subtree_index].bh);
 
-	ret = ocfs2_journal_access(handle, inode, root_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_path_bh_journal_access(handle, inode, right_path,
+					   subtree_index);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
 	for(i = subtree_index + 1; i < path_num_items(right_path); i++) {
-		ret = ocfs2_journal_access(handle, inode,
-					   right_path->p_node[i].bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode,
+						   right_path, i);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
 
-		ret = ocfs2_journal_access(handle, inode,
-					   left_path->p_node[i].bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode,
+						   left_path, i);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -2102,8 +2203,7 @@ static int ocfs2_rotate_tree_right(struct inode *inode,
 
 	*ret_left_path = NULL;
 
-	left_path = ocfs2_new_path(path_root_bh(right_path),
-				   path_root_el(right_path));
+	left_path = ocfs2_new_path_from_path(right_path);
 	if (!left_path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -2398,9 +2498,9 @@ static int ocfs2_rotate_subtree_left(struct inode *inode, handle_t *handle,
 			return -EAGAIN;
 
 		if (le16_to_cpu(right_leaf_el->l_next_free_rec) > 1) {
-			ret = ocfs2_journal_access(handle, inode,
-						   path_leaf_bh(right_path),
-						   OCFS2_JOURNAL_ACCESS_WRITE);
+			ret = ocfs2_journal_access_eb(handle, inode,
+						      path_leaf_bh(right_path),
+						      OCFS2_JOURNAL_ACCESS_WRITE);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
@@ -2417,8 +2517,8 @@ static int ocfs2_rotate_subtree_left(struct inode *inode, handle_t *handle,
 		 * We have to update i_last_eb_blk during the meta
 		 * data delete.
 		 */
-		ret = ocfs2_journal_access(handle, inode, et_root_bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_et_root_journal_access(handle, inode, et,
+						   OCFS2_JOURNAL_ACCESS_WRITE);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -2433,25 +2533,23 @@ static int ocfs2_rotate_subtree_left(struct inode *inode, handle_t *handle,
 	 */
 	BUG_ON(right_has_empty && !del_right_subtree);
 
-	ret = ocfs2_journal_access(handle, inode, root_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_path_bh_journal_access(handle, inode, right_path,
+					   subtree_index);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
 	for(i = subtree_index + 1; i < path_num_items(right_path); i++) {
-		ret = ocfs2_journal_access(handle, inode,
-					   right_path->p_node[i].bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode,
+						   right_path, i);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
 
-		ret = ocfs2_journal_access(handle, inode,
-					   left_path->p_node[i].bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode,
+						   left_path, i);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -2596,16 +2694,17 @@ out:
 
 static int ocfs2_rotate_rightmost_leaf_left(struct inode *inode,
 					    handle_t *handle,
-					    struct buffer_head *bh,
-					    struct ocfs2_extent_list *el)
+					    struct ocfs2_path *path)
 {
 	int ret;
+	struct buffer_head *bh = path_leaf_bh(path);
+	struct ocfs2_extent_list *el = path_leaf_el(path);
 
 	if (!ocfs2_is_empty_extent(&el->l_recs[0]))
 		return 0;
 
-	ret = ocfs2_journal_access(handle, inode, bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_path_bh_journal_access(handle, inode, path,
+					   path_num_items(path) - 1);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -2644,8 +2743,7 @@ static int __ocfs2_rotate_tree_left(struct inode *inode,
 		goto out;
 	}
 
-	left_path = ocfs2_new_path(path_root_bh(path),
-				   path_root_el(path));
+	left_path = ocfs2_new_path_from_path(path);
 	if (!left_path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -2654,8 +2752,7 @@ static int __ocfs2_rotate_tree_left(struct inode *inode,
 
 	ocfs2_cp_path(left_path, path);
 
-	right_path = ocfs2_new_path(path_root_bh(path),
-				    path_root_el(path));
+	right_path = ocfs2_new_path_from_path(path);
 	if (!right_path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -2689,9 +2786,8 @@ static int __ocfs2_rotate_tree_left(struct inode *inode,
 		 * Caller might still want to make changes to the
 		 * tree root, so re-add it to the journal here.
 		 */
-		ret = ocfs2_journal_access(handle, inode,
-					   path_root_bh(left_path),
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode,
+						   left_path, 0);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -2785,8 +2881,7 @@ static int ocfs2_remove_rightmost_path(struct inode *inode, handle_t *handle,
 		 * We have a path to the left of this one - it needs
 		 * an update too.
 		 */
-		left_path = ocfs2_new_path(path_root_bh(path),
-					   path_root_el(path));
+		left_path = ocfs2_new_path_from_path(path);
 		if (!left_path) {
 			ret = -ENOMEM;
 			mlog_errno(ret);
@@ -2875,8 +2970,7 @@ rightmost_no_delete:
 		 * it up front.
 		 */
 		ret = ocfs2_rotate_rightmost_leaf_left(inode, handle,
-						       path_leaf_bh(path),
-						       path_leaf_el(path));
+						       path);
 		if (ret)
 			mlog_errno(ret);
 		goto out;
@@ -3027,8 +3121,7 @@ static int ocfs2_get_right_path(struct inode *inode,
 	/* This function shouldn't be called for the rightmost leaf. */
 	BUG_ON(right_cpos == 0);
 
-	right_path = ocfs2_new_path(path_root_bh(left_path),
-				    path_root_el(left_path));
+	right_path = ocfs2_new_path_from_path(left_path);
 	if (!right_path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -3111,8 +3204,8 @@ static int ocfs2_merge_rec_right(struct inode *inode,
 		root_bh = left_path->p_node[subtree_index].bh;
 		BUG_ON(root_bh != right_path->p_node[subtree_index].bh);
 
-		ret = ocfs2_journal_access(handle, inode, root_bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode, right_path,
+						   subtree_index);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -3120,17 +3213,15 @@ static int ocfs2_merge_rec_right(struct inode *inode,
 
 		for (i = subtree_index + 1;
 		     i < path_num_items(right_path); i++) {
-			ret = ocfs2_journal_access(handle, inode,
-						   right_path->p_node[i].bh,
-						   OCFS2_JOURNAL_ACCESS_WRITE);
+			ret = ocfs2_path_bh_journal_access(handle, inode,
+							   right_path, i);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
 			}
 
-			ret = ocfs2_journal_access(handle, inode,
-						   left_path->p_node[i].bh,
-						   OCFS2_JOURNAL_ACCESS_WRITE);
+			ret = ocfs2_path_bh_journal_access(handle, inode,
+							   left_path, i);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
@@ -3142,8 +3233,8 @@ static int ocfs2_merge_rec_right(struct inode *inode,
 		right_rec = &el->l_recs[index + 1];
 	}
 
-	ret = ocfs2_journal_access(handle, inode, bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_path_bh_journal_access(handle, inode, left_path,
+					   path_num_items(left_path) - 1);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -3199,8 +3290,7 @@ static int ocfs2_get_left_path(struct inode *inode,
 	/* This function shouldn't be called for the leftmost leaf. */
 	BUG_ON(left_cpos == 0);
 
-	left_path = ocfs2_new_path(path_root_bh(right_path),
-				   path_root_el(right_path));
+	left_path = ocfs2_new_path_from_path(right_path);
 	if (!left_path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -3283,8 +3373,8 @@ static int ocfs2_merge_rec_left(struct inode *inode,
 		root_bh = left_path->p_node[subtree_index].bh;
 		BUG_ON(root_bh != right_path->p_node[subtree_index].bh);
 
-		ret = ocfs2_journal_access(handle, inode, root_bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_path_bh_journal_access(handle, inode, right_path,
+						   subtree_index);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -3292,17 +3382,15 @@ static int ocfs2_merge_rec_left(struct inode *inode,
 
 		for (i = subtree_index + 1;
 		     i < path_num_items(right_path); i++) {
-			ret = ocfs2_journal_access(handle, inode,
-						   right_path->p_node[i].bh,
-						   OCFS2_JOURNAL_ACCESS_WRITE);
+			ret = ocfs2_path_bh_journal_access(handle, inode,
+							   right_path, i);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
 			}
 
-			ret = ocfs2_journal_access(handle, inode,
-						   left_path->p_node[i].bh,
-						   OCFS2_JOURNAL_ACCESS_WRITE);
+			ret = ocfs2_path_bh_journal_access(handle, inode,
+							   left_path, i);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
@@ -3314,8 +3402,8 @@ static int ocfs2_merge_rec_left(struct inode *inode,
 			has_empty_extent = 1;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_path_bh_journal_access(handle, inode, right_path,
+					   path_num_items(right_path) - 1);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -3732,8 +3820,7 @@ static int ocfs2_append_rec_to_path(struct inode *inode, handle_t *handle,
 		 * leftmost leaf.
 		 */
 		if (left_cpos) {
-			left_path = ocfs2_new_path(path_root_bh(right_path),
-						   path_root_el(right_path));
+			left_path = ocfs2_new_path_from_path(right_path);
 			if (!left_path) {
 				ret = -ENOMEM;
 				mlog_errno(ret);
@@ -3958,8 +4045,8 @@ static int ocfs2_do_insert_extent(struct inode *inode,
 
 	el = et->et_root_el;
 
-	ret = ocfs2_journal_access(handle, inode, et->et_root_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_et_root_journal_access(handle, inode, et,
+					   OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -3970,7 +4057,7 @@ static int ocfs2_do_insert_extent(struct inode *inode,
 		goto out_update_clusters;
 	}
 
-	right_path = ocfs2_new_path(et->et_root_bh, et->et_root_el);
+	right_path = ocfs2_new_path_from_et(et);
 	if (!right_path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -4020,8 +4107,8 @@ static int ocfs2_do_insert_extent(struct inode *inode,
 		 * ocfs2_rotate_tree_right() might have extended the
 		 * transaction without re-journaling our tree root.
 		 */
-		ret = ocfs2_journal_access(handle, inode, et->et_root_bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = ocfs2_et_root_journal_access(handle, inode, et,
+						   OCFS2_JOURNAL_ACCESS_WRITE);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -4082,8 +4169,7 @@ ocfs2_figure_merge_contig_type(struct inode *inode, struct ocfs2_path *path,
 			goto out;
 
 		if (left_cpos != 0) {
-			left_path = ocfs2_new_path(path_root_bh(path),
-						   path_root_el(path));
+			left_path = ocfs2_new_path_from_path(path);
 			if (!left_path)
 				goto out;
 
@@ -4097,8 +4183,15 @@ ocfs2_figure_merge_contig_type(struct inode *inode, struct ocfs2_path *path,
 			    le16_to_cpu(new_el->l_count)) {
 				bh = path_leaf_bh(left_path);
 				eb = (struct ocfs2_extent_block *)bh->b_data;
-				OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb,
-								 eb);
+				ocfs2_error(inode->i_sb,
+					    "Extent block #%llu has an "
+					    "invalid l_next_free_rec of "
+					    "%d.  It should have "
+					    "matched the l_count of %d",
+					    (unsigned long long)le64_to_cpu(eb->h_blkno),
+					    le16_to_cpu(new_el->l_next_free_rec),
+					    le16_to_cpu(new_el->l_count));
+				status = -EINVAL;
 				goto out;
 			}
 			rec = &new_el->l_recs[
@@ -4132,8 +4225,7 @@ ocfs2_figure_merge_contig_type(struct inode *inode, struct ocfs2_path *path,
 		if (right_cpos == 0)
 			goto out;
 
-		right_path = ocfs2_new_path(path_root_bh(path),
-					    path_root_el(path));
+		right_path = ocfs2_new_path_from_path(path);
 		if (!right_path)
 			goto out;
 
@@ -4147,8 +4239,12 @@ ocfs2_figure_merge_contig_type(struct inode *inode, struct ocfs2_path *path,
 			if (le16_to_cpu(new_el->l_next_free_rec) <= 1) {
 				bh = path_leaf_bh(right_path);
 				eb = (struct ocfs2_extent_block *)bh->b_data;
-				OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb,
-								 eb);
+				ocfs2_error(inode->i_sb,
+					    "Extent block #%llu has an "
+					    "invalid l_next_free_rec of %d",
+					    (unsigned long long)le64_to_cpu(eb->h_blkno),
+					    le16_to_cpu(new_el->l_next_free_rec));
+				status = -EINVAL;
 				goto out;
 			}
 			rec = &new_el->l_recs[1];
@@ -4294,7 +4390,9 @@ static int ocfs2_figure_insert_type(struct inode *inode,
 		 * ocfs2_figure_insert_type() and ocfs2_add_branch()
 		 * may want it later.
 		 */
-		ret = ocfs2_read_block(inode, ocfs2_et_get_last_eb_blk(et), &bh);
+		ret = ocfs2_read_extent_block(inode,
+					      ocfs2_et_get_last_eb_blk(et),
+					      &bh);
 		if (ret) {
 			mlog_exit(ret);
 			goto out;
@@ -4320,7 +4418,7 @@ static int ocfs2_figure_insert_type(struct inode *inode,
 		return 0;
 	}
 
-	path = ocfs2_new_path(et->et_root_bh, et->et_root_el);
+	path = ocfs2_new_path_from_et(et);
 	if (!path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -4531,9 +4629,9 @@ int ocfs2_add_clusters_in_btree(struct ocfs2_super *osb,
 
 	BUG_ON(num_bits > clusters_to_add);
 
-	/* reserve our write early -- insert_extent may update the inode */
-	status = ocfs2_journal_access(handle, inode, et->et_root_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	/* reserve our write early -- insert_extent may update the tree root */
+	status = ocfs2_et_root_journal_access(handle, inode, et,
+					      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -4760,20 +4858,15 @@ static int __ocfs2_mark_extent_written(struct inode *inode,
 	if (path->p_tree_depth) {
 		struct ocfs2_extent_block *eb;
 
-		ret = ocfs2_read_block(inode, ocfs2_et_get_last_eb_blk(et),
-				       &last_eb_bh);
+		ret = ocfs2_read_extent_block(inode,
+					      ocfs2_et_get_last_eb_blk(et),
+					      &last_eb_bh);
 		if (ret) {
 			mlog_exit(ret);
 			goto out;
 		}
 
 		eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
-		if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-			OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-			ret = -EROFS;
-			goto out;
-		}
-
 		rightmost_el = &eb->h_list;
 	} else
 		rightmost_el = path_root_el(path);
@@ -4854,7 +4947,7 @@ int ocfs2_mark_extent_written(struct inode *inode,
 	if (et->et_ops == &ocfs2_dinode_et_ops)
 		ocfs2_extent_map_trunc(inode, 0);
 
-	left_path = ocfs2_new_path(et->et_root_bh, et->et_root_el);
+	left_path = ocfs2_new_path_from_et(et);
 	if (!left_path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -4918,8 +5011,9 @@ static int ocfs2_split_tree(struct inode *inode, struct ocfs2_extent_tree *et,
 
 	depth = path->p_tree_depth;
 	if (depth > 0) {
-		ret = ocfs2_read_block(inode, ocfs2_et_get_last_eb_blk(et),
-				       &last_eb_bh);
+		ret = ocfs2_read_extent_block(inode,
+					      ocfs2_et_get_last_eb_blk(et),
+					      &last_eb_bh);
 		if (ret < 0) {
 			mlog_errno(ret);
 			goto out;
@@ -5025,8 +5119,7 @@ static int ocfs2_truncate_rec(struct inode *inode, handle_t *handle,
 		}
 
 		if (left_cpos && le16_to_cpu(el->l_next_free_rec) > 1) {
-			left_path = ocfs2_new_path(path_root_bh(path),
-						   path_root_el(path));
+			left_path = ocfs2_new_path_from_path(path);
 			if (!left_path) {
 				ret = -ENOMEM;
 				mlog_errno(ret);
@@ -5135,7 +5228,7 @@ int ocfs2_remove_extent(struct inode *inode,
 
 	ocfs2_extent_map_trunc(inode, 0);
 
-	path = ocfs2_new_path(et->et_root_bh, et->et_root_el);
+	path = ocfs2_new_path_from_et(et);
 	if (!path) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
@@ -5255,6 +5348,78 @@ out:
 	return ret;
 }
 
+int ocfs2_remove_btree_range(struct inode *inode,
+			     struct ocfs2_extent_tree *et,
+			     u32 cpos, u32 phys_cpos, u32 len,
+			     struct ocfs2_cached_dealloc_ctxt *dealloc)
+{
+	int ret;
+	u64 phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos);
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct inode *tl_inode = osb->osb_tl_inode;
+	handle_t *handle;
+	struct ocfs2_alloc_context *meta_ac = NULL;
+
+	ret = ocfs2_lock_allocators(inode, et, 0, 1, NULL, &meta_ac);
+	if (ret) {
+		mlog_errno(ret);
+		return ret;
+	}
+
+	mutex_lock(&tl_inode->i_mutex);
+
+	if (ocfs2_truncate_log_needs_flush(osb)) {
+		ret = __ocfs2_flush_truncate_log(osb);
+		if (ret < 0) {
+			mlog_errno(ret);
+			goto out;
+		}
+	}
+
+	handle = ocfs2_start_trans(osb, ocfs2_remove_extent_credits(osb->sb));
+	if (IS_ERR(handle)) {
+		ret = PTR_ERR(handle);
+		mlog_errno(ret);
+		goto out;
+	}
+
+	ret = ocfs2_et_root_journal_access(handle, inode, et,
+					   OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
+
+	ret = ocfs2_remove_extent(inode, et, cpos, len, handle, meta_ac,
+				  dealloc);
+	if (ret) {
+		mlog_errno(ret);
+		goto out_commit;
+	}
+
+	ocfs2_et_update_clusters(inode, et, -len);
+
+	ret = ocfs2_journal_dirty(handle, et->et_root_bh);
+	if (ret) {
+		mlog_errno(ret);
+		goto out_commit;
+	}
+
+	ret = ocfs2_truncate_log_append(osb, handle, phys_blkno, len);
+	if (ret)
+		mlog_errno(ret);
+
+out_commit:
+	ocfs2_commit_trans(osb, handle);
+out:
+	mutex_unlock(&tl_inode->i_mutex);
+
+	if (meta_ac)
+		ocfs2_free_alloc_context(meta_ac);
+
+	return ret;
+}
+
 int ocfs2_truncate_log_needs_flush(struct ocfs2_super *osb)
 {
 	struct buffer_head *tl_bh = osb->osb_tl_bh;
@@ -5308,13 +5473,13 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb,
 	start_cluster = ocfs2_blocks_to_clusters(osb->sb, start_blk);
 
 	di = (struct ocfs2_dinode *) tl_bh->b_data;
-	tl = &di->id2.i_dealloc;
-	if (!OCFS2_IS_VALID_DINODE(di)) {
-		OCFS2_RO_ON_INVALID_DINODE(osb->sb, di);
-		status = -EIO;
-		goto bail;
-	}
 
+	/* tl_bh is loaded from ocfs2_truncate_log_init().  It's validated
+	 * by the underlying call to ocfs2_read_inode_block(), so any
+	 * corruption is a code bug */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(di));
+
+	tl = &di->id2.i_dealloc;
 	tl_count = le16_to_cpu(tl->tl_count);
 	mlog_bug_on_msg(tl_count > ocfs2_truncate_recs_per_inode(osb->sb) ||
 			tl_count == 0,
@@ -5332,8 +5497,8 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb,
 		goto bail;
 	}
 
-	status = ocfs2_journal_access(handle, tl_inode, tl_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, tl_inode, tl_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -5394,8 +5559,8 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb,
 	while (i >= 0) {
 		/* Caller has given us at least enough credits to
 		 * update the truncate log dinode */
-		status = ocfs2_journal_access(handle, tl_inode, tl_bh,
-					      OCFS2_JOURNAL_ACCESS_WRITE);
+		status = ocfs2_journal_access_di(handle, tl_inode, tl_bh,
+						 OCFS2_JOURNAL_ACCESS_WRITE);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
@@ -5464,13 +5629,13 @@ int __ocfs2_flush_truncate_log(struct ocfs2_super *osb)
 	BUG_ON(mutex_trylock(&tl_inode->i_mutex));
 
 	di = (struct ocfs2_dinode *) tl_bh->b_data;
-	tl = &di->id2.i_dealloc;
-	if (!OCFS2_IS_VALID_DINODE(di)) {
-		OCFS2_RO_ON_INVALID_DINODE(osb->sb, di);
-		status = -EIO;
-		goto out;
-	}
 
+	/* tl_bh is loaded from ocfs2_truncate_log_init().  It's validated
+	 * by the underlying call to ocfs2_read_inode_block(), so any
+	 * corruption is a code bug */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(di));
+
+	tl = &di->id2.i_dealloc;
 	num_to_flush = le16_to_cpu(tl->tl_used);
 	mlog(0, "Flush %u records from truncate log #%llu\n",
 	     num_to_flush, (unsigned long long)OCFS2_I(tl_inode)->ip_blkno);
@@ -5586,7 +5751,7 @@ static int ocfs2_get_truncate_log_info(struct ocfs2_super *osb,
 		goto bail;
 	}
 
-	status = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, &bh);
+	status = ocfs2_read_inode_block(inode, &bh);
 	if (status < 0) {
 		iput(inode);
 		mlog_errno(status);
@@ -5625,13 +5790,13 @@ int ocfs2_begin_truncate_log_recovery(struct ocfs2_super *osb,
 	}
 
 	di = (struct ocfs2_dinode *) tl_bh->b_data;
-	tl = &di->id2.i_dealloc;
-	if (!OCFS2_IS_VALID_DINODE(di)) {
-		OCFS2_RO_ON_INVALID_DINODE(tl_inode->i_sb, di);
-		status = -EIO;
-		goto bail;
-	}
 
+	/* tl_bh is loaded from ocfs2_get_truncate_log_info().  It's
+	 * validated by the underlying call to ocfs2_read_inode_block(),
+	 * so any corruption is a code bug */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(di));
+
+	tl = &di->id2.i_dealloc;
 	if (le16_to_cpu(tl->tl_used)) {
 		mlog(0, "We'll have %u logs to recover\n",
 		     le16_to_cpu(tl->tl_used));
@@ -5651,6 +5816,7 @@ int ocfs2_begin_truncate_log_recovery(struct ocfs2_super *osb,
 		 * tl_used. */
 		tl->tl_used = 0;
 
+		ocfs2_compute_meta_ecc(osb->sb, tl_bh->b_data, &di->i_check);
 		status = ocfs2_write_block(osb, tl_bh, tl_inode);
 		if (status < 0) {
 			mlog_errno(status);
@@ -5800,7 +5966,10 @@ int ocfs2_truncate_log_init(struct ocfs2_super *osb)
  */
 
 /*
- * Describes a single block free from a suballocator
+ * Describe a single bit freed from a suballocator.  For the block
+ * suballocators, it represents one block.  For the global cluster
+ * allocator, it represents some clusters and free_bit indicates
+ * clusters number.
  */
 struct ocfs2_cached_block_free {
 	struct ocfs2_cached_block_free		*free_next;
@@ -5815,10 +5984,10 @@ struct ocfs2_per_slot_free_list {
 	struct ocfs2_cached_block_free		*f_first;
 };
 
-static int ocfs2_free_cached_items(struct ocfs2_super *osb,
-				   int sysfile_type,
-				   int slot,
-				   struct ocfs2_cached_block_free *head)
+static int ocfs2_free_cached_blocks(struct ocfs2_super *osb,
+				    int sysfile_type,
+				    int slot,
+				    struct ocfs2_cached_block_free *head)
 {
 	int ret;
 	u64 bg_blkno;
@@ -5893,6 +6062,82 @@ out:
 	return ret;
 }
 
+int ocfs2_cache_cluster_dealloc(struct ocfs2_cached_dealloc_ctxt *ctxt,
+				u64 blkno, unsigned int bit)
+{
+	int ret = 0;
+	struct ocfs2_cached_block_free *item;
+
+	item = kmalloc(sizeof(*item), GFP_NOFS);
+	if (item == NULL) {
+		ret = -ENOMEM;
+		mlog_errno(ret);
+		return ret;
+	}
+
+	mlog(0, "Insert clusters: (bit %u, blk %llu)\n",
+	     bit, (unsigned long long)blkno);
+
+	item->free_blk = blkno;
+	item->free_bit = bit;
+	item->free_next = ctxt->c_global_allocator;
+
+	ctxt->c_global_allocator = item;
+	return ret;
+}
+
+static int ocfs2_free_cached_clusters(struct ocfs2_super *osb,
+				      struct ocfs2_cached_block_free *head)
+{
+	struct ocfs2_cached_block_free *tmp;
+	struct inode *tl_inode = osb->osb_tl_inode;
+	handle_t *handle;
+	int ret = 0;
+
+	mutex_lock(&tl_inode->i_mutex);
+
+	while (head) {
+		if (ocfs2_truncate_log_needs_flush(osb)) {
+			ret = __ocfs2_flush_truncate_log(osb);
+			if (ret < 0) {
+				mlog_errno(ret);
+				break;
+			}
+		}
+
+		handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE);
+		if (IS_ERR(handle)) {
+			ret = PTR_ERR(handle);
+			mlog_errno(ret);
+			break;
+		}
+
+		ret = ocfs2_truncate_log_append(osb, handle, head->free_blk,
+						head->free_bit);
+
+		ocfs2_commit_trans(osb, handle);
+		tmp = head;
+		head = head->free_next;
+		kfree(tmp);
+
+		if (ret < 0) {
+			mlog_errno(ret);
+			break;
+		}
+	}
+
+	mutex_unlock(&tl_inode->i_mutex);
+
+	while (head) {
+		/* Premature exit may have left some dangling items. */
+		tmp = head;
+		head = head->free_next;
+		kfree(tmp);
+	}
+
+	return ret;
+}
+
 int ocfs2_run_deallocs(struct ocfs2_super *osb,
 		       struct ocfs2_cached_dealloc_ctxt *ctxt)
 {
@@ -5908,8 +6153,10 @@ int ocfs2_run_deallocs(struct ocfs2_super *osb,
 		if (fl->f_first) {
 			mlog(0, "Free items: (type %u, slot %d)\n",
 			     fl->f_inode_type, fl->f_slot);
-			ret2 = ocfs2_free_cached_items(osb, fl->f_inode_type,
-						       fl->f_slot, fl->f_first);
+			ret2 = ocfs2_free_cached_blocks(osb,
+							fl->f_inode_type,
+							fl->f_slot,
+							fl->f_first);
 			if (ret2)
 				mlog_errno(ret2);
 			if (!ret)
@@ -5920,6 +6167,17 @@ int ocfs2_run_deallocs(struct ocfs2_super *osb,
 		kfree(fl);
 	}
 
+	if (ctxt->c_global_allocator) {
+		ret2 = ocfs2_free_cached_clusters(osb,
+						  ctxt->c_global_allocator);
+		if (ret2)
+			mlog_errno(ret2);
+		if (!ret)
+			ret = ret2;
+
+		ctxt->c_global_allocator = NULL;
+	}
+
 	return ret;
 }
 
@@ -6075,11 +6333,10 @@ static int ocfs2_find_new_last_ext_blk(struct inode *inode,
 
 	eb = (struct ocfs2_extent_block *) bh->b_data;
 	el = &eb->h_list;
-	if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-		OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-		ret = -EROFS;
-		goto out;
-	}
+
+	/* ocfs2_find_leaf() gets the eb from ocfs2_read_extent_block().
+	 * Any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_EXTENT_BLOCK(eb));
 
 	*new_last_eb = bh;
 	get_bh(*new_last_eb);
@@ -6326,8 +6583,8 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb,
 	}
 
 	if (last_eb_bh) {
-		status = ocfs2_journal_access(handle, inode, last_eb_bh,
-					      OCFS2_JOURNAL_ACCESS_WRITE);
+		status = ocfs2_journal_access_eb(handle, inode, last_eb_bh,
+						 OCFS2_JOURNAL_ACCESS_WRITE);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
@@ -6350,6 +6607,8 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb,
 		goto bail;
 	}
 
+	vfs_dq_free_space_nodirty(inode,
+			ocfs2_clusters_to_bytes(osb->sb, clusters_to_del));
 	spin_lock(&OCFS2_I(inode)->ip_lock);
 	OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters) -
 				      clusters_to_del;
@@ -6436,11 +6695,6 @@ static void ocfs2_map_and_dirty_page(struct inode *inode, handle_t *handle,
 		mlog_errno(ret);
 	else if (ocfs2_should_order_data(inode)) {
 		ret = ocfs2_jbd2_file_inode(handle, inode);
-#ifdef CONFIG_OCFS2_COMPAT_JBD
-		ret = walk_page_buffers(handle, page_buffers(page),
-					from, to, &partial,
-					ocfs2_journal_dirty_data);
-#endif
 		if (ret < 0)
 			mlog_errno(ret);
 	}
@@ -6663,6 +6917,7 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 	struct page **pages = NULL;
 	loff_t end = osb->s_clustersize;
 	struct ocfs2_extent_tree et;
+	int did_quota = 0;
 
 	has_data = i_size_read(inode) ? 1 : 0;
 
@@ -6682,15 +6937,16 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_INLINE_TO_EXTENTS_CREDITS);
+	handle = ocfs2_start_trans(osb,
+				   ocfs2_inline_to_extents_credits(osb->sb));
 	if (IS_ERR(handle)) {
 		ret = PTR_ERR(handle);
 		mlog_errno(ret);
 		goto out_unlock;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -6701,6 +6957,13 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 		unsigned int page_end;
 		u64 phys;
 
+		if (vfs_dq_alloc_space_nodirty(inode,
+				       ocfs2_clusters_to_bytes(osb->sb, 1))) {
+			ret = -EDQUOT;
+			goto out_commit;
+		}
+		did_quota = 1;
+
 		ret = ocfs2_claim_clusters(osb, handle, data_ac, 1, &bit_off,
 					   &num);
 		if (ret) {
@@ -6774,6 +7037,10 @@ int ocfs2_convert_inline_data_to_extents(struct inode *inode,
 	}
 
 out_commit:
+	if (ret < 0 && did_quota)
+		vfs_dq_free_space_nodirty(inode,
+					  ocfs2_clusters_to_bytes(osb->sb, 1));
+
 	ocfs2_commit_trans(osb, handle);
 
 out_unlock:
@@ -6813,7 +7080,8 @@ int ocfs2_commit_truncate(struct ocfs2_super *osb,
 	new_highest_cpos = ocfs2_clusters_for_bytes(osb->sb,
 						     i_size_read(inode));
 
-	path = ocfs2_new_path(fe_bh, &di->id2.i_list);
+	path = ocfs2_new_path(fe_bh, &di->id2.i_list,
+			      ocfs2_journal_access_di);
 	if (!path) {
 		status = -ENOMEM;
 		mlog_errno(status);
@@ -6984,20 +7252,14 @@ int ocfs2_prepare_truncate(struct ocfs2_super *osb,
 	ocfs2_init_dealloc_ctxt(&(*tc)->tc_dealloc);
 
 	if (fe->id2.i_list.l_tree_depth) {
-		status = ocfs2_read_block(inode, le64_to_cpu(fe->i_last_eb_blk),
-					  &last_eb_bh);
+		status = ocfs2_read_extent_block(inode,
+						 le64_to_cpu(fe->i_last_eb_blk),
+						 &last_eb_bh);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
 		}
 		eb = (struct ocfs2_extent_block *) last_eb_bh->b_data;
-		if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-			OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-
-			brelse(last_eb_bh);
-			status = -EIO;
-			goto bail;
-		}
 	}
 
 	(*tc)->tc_last_eb_bh = last_eb_bh;
@@ -7052,8 +7314,8 @@ int ocfs2_truncate_inline(struct inode *inode, struct buffer_head *di_bh,
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;

fs/ocfs2/alloc.h

@@ -45,7 +45,9 @@
  *
  * ocfs2_extent_tree contains info for the root of the b-tree, it must have a
  * root ocfs2_extent_list and a root_bh so that they can be used in the b-tree
- * functions.
+ * functions.  With metadata ecc, we now call different journal_access
+ * functions for each type of metadata, so it must have the
+ * root_journal_access function.
  * ocfs2_extent_tree_operations abstract the normal operations we do for
  * the root of extent b-tree.
  */
@@ -54,6 +56,7 @@ struct ocfs2_extent_tree {
 	struct ocfs2_extent_tree_operations	*et_ops;
 	struct buffer_head			*et_root_bh;
 	struct ocfs2_extent_list		*et_root_el;
+	ocfs2_journal_access_func		et_root_journal_access;
 	void					*et_object;
 	unsigned int				et_max_leaf_clusters;
 };
@@ -68,10 +71,18 @@ void ocfs2_init_dinode_extent_tree(struct ocfs2_extent_tree *et,
 void ocfs2_init_xattr_tree_extent_tree(struct ocfs2_extent_tree *et,
 				       struct inode *inode,
 				       struct buffer_head *bh);
+struct ocfs2_xattr_value_buf;
 void ocfs2_init_xattr_value_extent_tree(struct ocfs2_extent_tree *et,
 					struct inode *inode,
-					struct buffer_head *bh,
-					struct ocfs2_xattr_value_root *xv);
+					struct ocfs2_xattr_value_buf *vb);
+
+/*
+ * Read an extent block into *bh.  If *bh is NULL, a bh will be
+ * allocated.  This is a cached read.  The extent block will be validated
+ * with ocfs2_validate_extent_block().
+ */
+int ocfs2_read_extent_block(struct inode *inode, u64 eb_blkno,
+			    struct buffer_head **bh);
 
 struct ocfs2_alloc_context;
 int ocfs2_insert_extent(struct ocfs2_super *osb,
@@ -110,6 +121,11 @@ int ocfs2_remove_extent(struct inode *inode,
 			u32 cpos, u32 len, handle_t *handle,
 			struct ocfs2_alloc_context *meta_ac,
 			struct ocfs2_cached_dealloc_ctxt *dealloc);
+int ocfs2_remove_btree_range(struct inode *inode,
+			     struct ocfs2_extent_tree *et,
+			     u32 cpos, u32 phys_cpos, u32 len,
+			     struct ocfs2_cached_dealloc_ctxt *dealloc);
+
 int ocfs2_num_free_extents(struct ocfs2_super *osb,
 			   struct inode *inode,
 			   struct ocfs2_extent_tree *et);
@@ -167,10 +183,18 @@ int __ocfs2_flush_truncate_log(struct ocfs2_super *osb);
  */
 struct ocfs2_cached_dealloc_ctxt {
 	struct ocfs2_per_slot_free_list		*c_first_suballocator;
+	struct ocfs2_cached_block_free 		*c_global_allocator;
 };
 static inline void ocfs2_init_dealloc_ctxt(struct ocfs2_cached_dealloc_ctxt *c)
 {
 	c->c_first_suballocator = NULL;
+	c->c_global_allocator = NULL;
+}
+int ocfs2_cache_cluster_dealloc(struct ocfs2_cached_dealloc_ctxt *ctxt,
+				u64 blkno, unsigned int bit);
+static inline int ocfs2_dealloc_has_cluster(struct ocfs2_cached_dealloc_ctxt *c)
+{
+	return c->c_global_allocator != NULL;
 }
 int ocfs2_run_deallocs(struct ocfs2_super *osb,
 		       struct ocfs2_cached_dealloc_ctxt *ctxt);

fs/ocfs2/aops.c

@@ -27,6 +27,7 @@
 #include <linux/swap.h>
 #include <linux/pipe_fs_i.h>
 #include <linux/mpage.h>
+#include <linux/quotaops.h>
 
 #define MLOG_MASK_PREFIX ML_FILE_IO
 #include <cluster/masklog.h>
@@ -68,20 +69,13 @@ static int ocfs2_symlink_get_block(struct inode *inode, sector_t iblock,
 		goto bail;
 	}
 
-	status = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, &bh);
+	status = ocfs2_read_inode_block(inode, &bh);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
 	fe = (struct ocfs2_dinode *) bh->b_data;
 
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		mlog(ML_ERROR, "Invalid dinode #%llu: signature = %.*s\n",
-		     (unsigned long long)le64_to_cpu(fe->i_blkno), 7,
-		     fe->i_signature);
-		goto bail;
-	}
-
 	if ((u64)iblock >= ocfs2_clusters_to_blocks(inode->i_sb,
 						    le32_to_cpu(fe->i_clusters))) {
 		mlog(ML_ERROR, "block offset is outside the allocated size: "
@@ -262,7 +256,7 @@ static int ocfs2_readpage_inline(struct inode *inode, struct page *page)
 	BUG_ON(!PageLocked(page));
 	BUG_ON(!(OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL));
 
-	ret = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, &di_bh);
+	ret = ocfs2_read_inode_block(inode, &di_bh);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -481,12 +475,6 @@ handle_t *ocfs2_start_walk_page_trans(struct inode *inode,
 
 	if (ocfs2_should_order_data(inode)) {
 		ret = ocfs2_jbd2_file_inode(handle, inode);
-#ifdef CONFIG_OCFS2_COMPAT_JBD
-		ret = walk_page_buffers(handle,
-					page_buffers(page),
-					from, to, NULL,
-					ocfs2_journal_dirty_data);
-#endif
 		if (ret < 0)
 			mlog_errno(ret);
 	}
@@ -1072,15 +1060,8 @@ static void ocfs2_write_failure(struct inode *inode,
 		tmppage = wc->w_pages[i];
 
 		if (page_has_buffers(tmppage)) {
-			if (ocfs2_should_order_data(inode)) {
+			if (ocfs2_should_order_data(inode))
 				ocfs2_jbd2_file_inode(wc->w_handle, inode);
-#ifdef CONFIG_OCFS2_COMPAT_JBD
-				walk_page_buffers(wc->w_handle,
-						  page_buffers(tmppage),
-						  from, to, NULL,
-						  ocfs2_journal_dirty_data);
-#endif
-			}
 
 			block_commit_write(tmppage, from, to);
 		}
@@ -1531,8 +1512,8 @@ static int ocfs2_write_begin_inline(struct address_space *mapping,
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, wc->w_di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, wc->w_di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		ocfs2_commit_trans(osb, handle);
 
@@ -1750,15 +1731,20 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 
 	wc->w_handle = handle;
 
+	if (clusters_to_alloc && vfs_dq_alloc_space_nodirty(inode,
+			ocfs2_clusters_to_bytes(osb->sb, clusters_to_alloc))) {
+		ret = -EDQUOT;
+		goto out_commit;
+	}
 	/*
 	 * We don't want this to fail in ocfs2_write_end(), so do it
 	 * here.
 	 */
-	ret = ocfs2_journal_access(handle, inode, wc->w_di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, wc->w_di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out_quota;
 	}
 
 	/*
@@ -1771,14 +1757,14 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 					 mmap_page);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out_quota;
 	}
 
 	ret = ocfs2_write_cluster_by_desc(mapping, data_ac, meta_ac, wc, pos,
 					  len);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out_quota;
 	}
 
 	if (data_ac)
@@ -1790,6 +1776,10 @@ success:
 	*pagep = wc->w_target_page;
 	*fsdata = wc;
 	return 0;
+out_quota:
+	if (clusters_to_alloc)
+		vfs_dq_free_space(inode,
+			  ocfs2_clusters_to_bytes(osb->sb, clusters_to_alloc));
 out_commit:
 	ocfs2_commit_trans(osb, handle);
 
@@ -1919,15 +1909,8 @@ int ocfs2_write_end_nolock(struct address_space *mapping,
 		}
 
 		if (page_has_buffers(tmppage)) {
-			if (ocfs2_should_order_data(inode)) {
+			if (ocfs2_should_order_data(inode))
 				ocfs2_jbd2_file_inode(wc->w_handle, inode);
-#ifdef CONFIG_OCFS2_COMPAT_JBD
-				walk_page_buffers(wc->w_handle,
-						  page_buffers(tmppage),
-						  from, to, NULL,
-						  ocfs2_journal_dirty_data);
-#endif
-			}
 			block_commit_write(tmppage, from, to);
 		}
 	}

fs/ocfs2/blockcheck.c

@@ -0,0 +1,477 @@
+/* -*- mode: c; c-basic-offset: 8; -*-
+ * vim: noexpandtab sw=8 ts=8 sts=0:
+ *
+ * blockcheck.c
+ *
+ * Checksum and ECC codes for the OCFS2 userspace library.
+ *
+ * Copyright (C) 2006, 2008 Oracle.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License, version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/crc32.h>
+#include <linux/buffer_head.h>
+#include <linux/bitops.h>
+#include <asm/byteorder.h>
+
+#include <cluster/masklog.h>
+
+#include "ocfs2.h"
+
+#include "blockcheck.h"
+
+
+/*
+ * We use the following conventions:
+ *
+ * d = # data bits
+ * p = # parity bits
+ * c = # total code bits (d + p)
+ */
+
+
+/*
+ * Calculate the bit offset in the hamming code buffer based on the bit's
+ * offset in the data buffer.  Since the hamming code reserves all
+ * power-of-two bits for parity, the data bit number and the code bit
+ * number are offest by all the parity bits beforehand.
+ *
+ * Recall that bit numbers in hamming code are 1-based.  This function
+ * takes the 0-based data bit from the caller.
+ *
+ * An example.  Take bit 1 of the data buffer.  1 is a power of two (2^0),
+ * so it's a parity bit.  2 is a power of two (2^1), so it's a parity bit.
+ * 3 is not a power of two.  So bit 1 of the data buffer ends up as bit 3
+ * in the code buffer.
+ *
+ * The caller can pass in *p if it wants to keep track of the most recent
+ * number of parity bits added.  This allows the function to start the
+ * calculation at the last place.
+ */
+static unsigned int calc_code_bit(unsigned int i, unsigned int *p_cache)
+{
+	unsigned int b, p = 0;
+
+	/*
+	 * Data bits are 0-based, but we're talking code bits, which
+	 * are 1-based.
+	 */
+	b = i + 1;
+
+	/* Use the cache if it is there */
+	if (p_cache)
+		p = *p_cache;
+        b += p;
+
+	/*
+	 * For every power of two below our bit number, bump our bit.
+	 *
+	 * We compare with (b + 1) because we have to compare with what b
+	 * would be _if_ it were bumped up by the parity bit.  Capice?
+	 *
+	 * p is set above.
+	 */
+	for (; (1 << p) < (b + 1); p++)
+		b++;
+
+	if (p_cache)
+		*p_cache = p;
+
+	return b;
+}
+
+/*
+ * This is the low level encoder function.  It can be called across
+ * multiple hunks just like the crc32 code.  'd' is the number of bits
+ * _in_this_hunk_.  nr is the bit offset of this hunk.  So, if you had
+ * two 512B buffers, you would do it like so:
+ *
+ * parity = ocfs2_hamming_encode(0, buf1, 512 * 8, 0);
+ * parity = ocfs2_hamming_encode(parity, buf2, 512 * 8, 512 * 8);
+ *
+ * If you just have one buffer, use ocfs2_hamming_encode_block().
+ */
+u32 ocfs2_hamming_encode(u32 parity, void *data, unsigned int d, unsigned int nr)
+{
+	unsigned int i, b, p = 0;
+
+	BUG_ON(!d);
+
+	/*
+	 * b is the hamming code bit number.  Hamming code specifies a
+	 * 1-based array, but C uses 0-based.  So 'i' is for C, and 'b' is
+	 * for the algorithm.
+	 *
+	 * The i++ in the for loop is so that the start offset passed
+	 * to ocfs2_find_next_bit_set() is one greater than the previously
+	 * found bit.
+	 */
+	for (i = 0; (i = ocfs2_find_next_bit(data, d, i)) < d; i++)
+	{
+		/*
+		 * i is the offset in this hunk, nr + i is the total bit
+		 * offset.
+		 */
+		b = calc_code_bit(nr + i, &p);
+
+		/*
+		 * Data bits in the resultant code are checked by
+		 * parity bits that are part of the bit number
+		 * representation.  Huh?
+		 *
+		 * <wikipedia href="http://en.wikipedia.org/wiki/Hamming_code">
+		 * In other words, the parity bit at position 2^k
+		 * checks bits in positions having bit k set in
+		 * their binary representation.  Conversely, for
+		 * instance, bit 13, i.e. 1101(2), is checked by
+		 * bits 1000(2) = 8, 0100(2)=4 and 0001(2) = 1.
+		 * </wikipedia>
+		 *
+		 * Note that 'k' is the _code_ bit number.  'b' in
+		 * our loop.
+		 */
+		parity ^= b;
+	}
+
+	/* While the data buffer was treated as little endian, the
+	 * return value is in host endian. */
+	return parity;
+}
+
+u32 ocfs2_hamming_encode_block(void *data, unsigned int blocksize)
+{
+	return ocfs2_hamming_encode(0, data, blocksize * 8, 0);
+}
+
+/*
+ * Like ocfs2_hamming_encode(), this can handle hunks.  nr is the bit
+ * offset of the current hunk.  If bit to be fixed is not part of the
+ * current hunk, this does nothing.
+ *
+ * If you only have one hunk, use ocfs2_hamming_fix_block().
+ */
+void ocfs2_hamming_fix(void *data, unsigned int d, unsigned int nr,
+		       unsigned int fix)
+{
+	unsigned int i, b;
+
+	BUG_ON(!d);
+
+	/*
+	 * If the bit to fix has an hweight of 1, it's a parity bit.  One
+	 * busted parity bit is its own error.  Nothing to do here.
+	 */
+	if (hweight32(fix) == 1)
+		return;
+
+	/*
+	 * nr + d is the bit right past the data hunk we're looking at.
+	 * If fix after that, nothing to do
+	 */
+	if (fix >= calc_code_bit(nr + d, NULL))
+		return;
+
+	/*
+	 * nr is the offset in the data hunk we're starting at.  Let's
+	 * start b at the offset in the code buffer.  See hamming_encode()
+	 * for a more detailed description of 'b'.
+	 */
+	b = calc_code_bit(nr, NULL);
+	/* If the fix is before this hunk, nothing to do */
+	if (fix < b)
+		return;
+
+	for (i = 0; i < d; i++, b++)
+	{
+		/* Skip past parity bits */
+		while (hweight32(b) == 1)
+			b++;
+
+		/*
+		 * i is the offset in this data hunk.
+		 * nr + i is the offset in the total data buffer.
+		 * b is the offset in the total code buffer.
+		 *
+		 * Thus, when b == fix, bit i in the current hunk needs
+		 * fixing.
+		 */
+		if (b == fix)
+		{
+			if (ocfs2_test_bit(i, data))
+				ocfs2_clear_bit(i, data);
+			else
+				ocfs2_set_bit(i, data);
+			break;
+		}
+	}
+}
+
+void ocfs2_hamming_fix_block(void *data, unsigned int blocksize,
+			     unsigned int fix)
+{
+	ocfs2_hamming_fix(data, blocksize * 8, 0, fix);
+}
+
+/*
+ * This function generates check information for a block.
+ * data is the block to be checked.  bc is a pointer to the
+ * ocfs2_block_check structure describing the crc32 and the ecc.
+ *
+ * bc should be a pointer inside data, as the function will
+ * take care of zeroing it before calculating the check information.  If
+ * bc does not point inside data, the caller must make sure any inline
+ * ocfs2_block_check structures are zeroed.
+ *
+ * The data buffer must be in on-disk endian (little endian for ocfs2).
+ * bc will be filled with little-endian values and will be ready to go to
+ * disk.
+ */
+void ocfs2_block_check_compute(void *data, size_t blocksize,
+			       struct ocfs2_block_check *bc)
+{
+	u32 crc;
+	u32 ecc;
+
+	memset(bc, 0, sizeof(struct ocfs2_block_check));
+
+	crc = crc32_le(~0, data, blocksize);
+	ecc = ocfs2_hamming_encode_block(data, blocksize);
+
+	/*
+	 * No ecc'd ocfs2 structure is larger than 4K, so ecc will be no
+	 * larger than 16 bits.
+	 */
+	BUG_ON(ecc > USHORT_MAX);
+
+	bc->bc_crc32e = cpu_to_le32(crc);
+	bc->bc_ecc = cpu_to_le16((u16)ecc);
+}
+
+/*
+ * This function validates existing check information.  Like _compute,
+ * the function will take care of zeroing bc before calculating check codes.
+ * If bc is not a pointer inside data, the caller must have zeroed any
+ * inline ocfs2_block_check structures.
+ *
+ * Again, the data passed in should be the on-disk endian.
+ */
+int ocfs2_block_check_validate(void *data, size_t blocksize,
+			       struct ocfs2_block_check *bc)
+{
+	int rc = 0;
+	struct ocfs2_block_check check;
+	u32 crc, ecc;
+
+	check.bc_crc32e = le32_to_cpu(bc->bc_crc32e);
+	check.bc_ecc = le16_to_cpu(bc->bc_ecc);
+
+	memset(bc, 0, sizeof(struct ocfs2_block_check));
+
+	/* Fast path - if the crc32 validates, we're good to go */
+	crc = crc32_le(~0, data, blocksize);
+	if (crc == check.bc_crc32e)
+		goto out;
+
+	mlog(ML_ERROR,
+	     "CRC32 failed: stored: %u, computed %u.  Applying ECC.\n",
+	     (unsigned int)check.bc_crc32e, (unsigned int)crc);
+
+	/* Ok, try ECC fixups */
+	ecc = ocfs2_hamming_encode_block(data, blocksize);
+	ocfs2_hamming_fix_block(data, blocksize, ecc ^ check.bc_ecc);
+
+	/* And check the crc32 again */
+	crc = crc32_le(~0, data, blocksize);
+	if (crc == check.bc_crc32e)
+		goto out;
+
+	mlog(ML_ERROR, "Fixed CRC32 failed: stored: %u, computed %u\n",
+	     (unsigned int)check.bc_crc32e, (unsigned int)crc);
+
+	rc = -EIO;
+
+out:
+	bc->bc_crc32e = cpu_to_le32(check.bc_crc32e);
+	bc->bc_ecc = cpu_to_le16(check.bc_ecc);
+
+	return rc;
+}
+
+/*
+ * This function generates check information for a list of buffer_heads.
+ * bhs is the blocks to be checked.  bc is a pointer to the
+ * ocfs2_block_check structure describing the crc32 and the ecc.
+ *
+ * bc should be a pointer inside data, as the function will
+ * take care of zeroing it before calculating the check information.  If
+ * bc does not point inside data, the caller must make sure any inline
+ * ocfs2_block_check structures are zeroed.
+ *
+ * The data buffer must be in on-disk endian (little endian for ocfs2).
+ * bc will be filled with little-endian values and will be ready to go to
+ * disk.
+ */
+void ocfs2_block_check_compute_bhs(struct buffer_head **bhs, int nr,
+				   struct ocfs2_block_check *bc)
+{
+	int i;
+	u32 crc, ecc;
+
+	BUG_ON(nr < 0);
+
+	if (!nr)
+		return;
+
+	memset(bc, 0, sizeof(struct ocfs2_block_check));
+
+	for (i = 0, crc = ~0, ecc = 0; i < nr; i++) {
+		crc = crc32_le(crc, bhs[i]->b_data, bhs[i]->b_size);
+		/*
+		 * The number of bits in a buffer is obviously b_size*8.
+		 * The offset of this buffer is b_size*i, so the bit offset
+		 * of this buffer is b_size*8*i.
+		 */
+		ecc = (u16)ocfs2_hamming_encode(ecc, bhs[i]->b_data,
+						bhs[i]->b_size * 8,
+						bhs[i]->b_size * 8 * i);
+	}
+
+	/*
+	 * No ecc'd ocfs2 structure is larger than 4K, so ecc will be no
+	 * larger than 16 bits.
+	 */
+	BUG_ON(ecc > USHORT_MAX);
+
+	bc->bc_crc32e = cpu_to_le32(crc);
+	bc->bc_ecc = cpu_to_le16((u16)ecc);
+}
+
+/*
+ * This function validates existing check information on a list of
+ * buffer_heads.  Like _compute_bhs, the function will take care of
+ * zeroing bc before calculating check codes.  If bc is not a pointer
+ * inside data, the caller must have zeroed any inline
+ * ocfs2_block_check structures.
+ *
+ * Again, the data passed in should be the on-disk endian.
+ */
+int ocfs2_block_check_validate_bhs(struct buffer_head **bhs, int nr,
+				   struct ocfs2_block_check *bc)
+{
+	int i, rc = 0;
+	struct ocfs2_block_check check;
+	u32 crc, ecc, fix;
+
+	BUG_ON(nr < 0);
+
+	if (!nr)
+		return 0;
+
+	check.bc_crc32e = le32_to_cpu(bc->bc_crc32e);
+	check.bc_ecc = le16_to_cpu(bc->bc_ecc);
+
+	memset(bc, 0, sizeof(struct ocfs2_block_check));
+
+	/* Fast path - if the crc32 validates, we're good to go */
+	for (i = 0, crc = ~0; i < nr; i++)
+		crc = crc32_le(crc, bhs[i]->b_data, bhs[i]->b_size);
+	if (crc == check.bc_crc32e)
+		goto out;
+
+	mlog(ML_ERROR,
+	     "CRC32 failed: stored: %u, computed %u.  Applying ECC.\n",
+	     (unsigned int)check.bc_crc32e, (unsigned int)crc);
+
+	/* Ok, try ECC fixups */
+	for (i = 0, ecc = 0; i < nr; i++) {
+		/*
+		 * The number of bits in a buffer is obviously b_size*8.
+		 * The offset of this buffer is b_size*i, so the bit offset
+		 * of this buffer is b_size*8*i.
+		 */
+		ecc = (u16)ocfs2_hamming_encode(ecc, bhs[i]->b_data,
+						bhs[i]->b_size * 8,
+						bhs[i]->b_size * 8 * i);
+	}
+	fix = ecc ^ check.bc_ecc;
+	for (i = 0; i < nr; i++) {
+		/*
+		 * Try the fix against each buffer.  It will only affect
+		 * one of them.
+		 */
+		ocfs2_hamming_fix(bhs[i]->b_data, bhs[i]->b_size * 8,
+				  bhs[i]->b_size * 8 * i, fix);
+	}
+
+	/* And check the crc32 again */
+	for (i = 0, crc = ~0; i < nr; i++)
+		crc = crc32_le(crc, bhs[i]->b_data, bhs[i]->b_size);
+	if (crc == check.bc_crc32e)
+		goto out;
+
+	mlog(ML_ERROR, "Fixed CRC32 failed: stored: %u, computed %u\n",
+	     (unsigned int)check.bc_crc32e, (unsigned int)crc);
+
+	rc = -EIO;
+
+out:
+	bc->bc_crc32e = cpu_to_le32(check.bc_crc32e);
+	bc->bc_ecc = cpu_to_le16(check.bc_ecc);
+
+	return rc;
+}
+
+/*
+ * These are the main API.  They check the superblock flag before
+ * calling the underlying operations.
+ *
+ * They expect the buffer(s) to be in disk format.
+ */
+void ocfs2_compute_meta_ecc(struct super_block *sb, void *data,
+			    struct ocfs2_block_check *bc)
+{
+	if (ocfs2_meta_ecc(OCFS2_SB(sb)))
+		ocfs2_block_check_compute(data, sb->s_blocksize, bc);
+}
+
+int ocfs2_validate_meta_ecc(struct super_block *sb, void *data,
+			    struct ocfs2_block_check *bc)
+{
+	int rc = 0;
+
+	if (ocfs2_meta_ecc(OCFS2_SB(sb)))
+		rc = ocfs2_block_check_validate(data, sb->s_blocksize, bc);
+
+	return rc;
+}
+
+void ocfs2_compute_meta_ecc_bhs(struct super_block *sb,
+				struct buffer_head **bhs, int nr,
+				struct ocfs2_block_check *bc)
+{
+	if (ocfs2_meta_ecc(OCFS2_SB(sb)))
+		ocfs2_block_check_compute_bhs(bhs, nr, bc);
+}
+
+int ocfs2_validate_meta_ecc_bhs(struct super_block *sb,
+				struct buffer_head **bhs, int nr,
+				struct ocfs2_block_check *bc)
+{
+	int rc = 0;
+
+	if (ocfs2_meta_ecc(OCFS2_SB(sb)))
+		rc = ocfs2_block_check_validate_bhs(bhs, nr, bc);
+
+	return rc;
+}
+

fs/ocfs2/blockcheck.h

@@ -0,0 +1,82 @@
+/* -*- mode: c; c-basic-offset: 8; -*-
+ * vim: noexpandtab sw=8 ts=8 sts=0:
+ *
+ * blockcheck.h
+ *
+ * Checksum and ECC codes for the OCFS2 userspace library.
+ *
+ * Copyright (C) 2004, 2008 Oracle.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License, version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef OCFS2_BLOCKCHECK_H
+#define OCFS2_BLOCKCHECK_H
+
+
+/* High level block API */
+void ocfs2_compute_meta_ecc(struct super_block *sb, void *data,
+			    struct ocfs2_block_check *bc);
+int ocfs2_validate_meta_ecc(struct super_block *sb, void *data,
+			    struct ocfs2_block_check *bc);
+void ocfs2_compute_meta_ecc_bhs(struct super_block *sb,
+				struct buffer_head **bhs, int nr,
+				struct ocfs2_block_check *bc);
+int ocfs2_validate_meta_ecc_bhs(struct super_block *sb,
+				struct buffer_head **bhs, int nr,
+				struct ocfs2_block_check *bc);
+
+/* Lower level API */
+void ocfs2_block_check_compute(void *data, size_t blocksize,
+			       struct ocfs2_block_check *bc);
+int ocfs2_block_check_validate(void *data, size_t blocksize,
+			       struct ocfs2_block_check *bc);
+void ocfs2_block_check_compute_bhs(struct buffer_head **bhs, int nr,
+				   struct ocfs2_block_check *bc);
+int ocfs2_block_check_validate_bhs(struct buffer_head **bhs, int nr,
+				   struct ocfs2_block_check *bc);
+
+/*
+ * Hamming code functions
+ */
+
+/*
+ * Encoding hamming code parity bits for a buffer.
+ *
+ * This is the low level encoder function.  It can be called across
+ * multiple hunks just like the crc32 code.  'd' is the number of bits
+ * _in_this_hunk_.  nr is the bit offset of this hunk.  So, if you had
+ * two 512B buffers, you would do it like so:
+ *
+ * parity = ocfs2_hamming_encode(0, buf1, 512 * 8, 0);
+ * parity = ocfs2_hamming_encode(parity, buf2, 512 * 8, 512 * 8);
+ *
+ * If you just have one buffer, use ocfs2_hamming_encode_block().
+ */
+u32 ocfs2_hamming_encode(u32 parity, void *data, unsigned int d,
+			 unsigned int nr);
+/*
+ * Fix a buffer with a bit error.  The 'fix' is the original parity
+ * xor'd with the parity calculated now.
+ *
+ * Like ocfs2_hamming_encode(), this can handle hunks.  nr is the bit
+ * offset of the current hunk.  If bit to be fixed is not part of the
+ * current hunk, this does nothing.
+ *
+ * If you only have one buffer, use ocfs2_hamming_fix_block().
+ */
+void ocfs2_hamming_fix(void *data, unsigned int d, unsigned int nr,
+		       unsigned int fix);
+
+/* Convenience wrappers for a single buffer of data */
+extern u32 ocfs2_hamming_encode_block(void *data, unsigned int blocksize);
+extern void ocfs2_hamming_fix_block(void *data, unsigned int blocksize,
+				    unsigned int fix);
+#endif

fs/ocfs2/buffer_head_io.c

@@ -39,6 +39,18 @@
 
 #include "buffer_head_io.h"
 
+/*
+ * Bits on bh->b_state used by ocfs2.
+ *
+ * These MUST be after the JBD2 bits.  Hence, we use BH_JBDPrivateStart.
+ */
+enum ocfs2_state_bits {
+	BH_NeedsValidate = BH_JBDPrivateStart,
+};
+
+/* Expand the magic b_state functions */
+BUFFER_FNS(NeedsValidate, needs_validate);
+
 int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
 		      struct inode *inode)
 {
@@ -166,7 +178,9 @@ bail:
 }
 
 int ocfs2_read_blocks(struct inode *inode, u64 block, int nr,
-		      struct buffer_head *bhs[], int flags)
+		      struct buffer_head *bhs[], int flags,
+		      int (*validate)(struct super_block *sb,
+				      struct buffer_head *bh))
 {
 	int status = 0;
 	int i, ignore_cache = 0;
@@ -298,6 +312,8 @@ int ocfs2_read_blocks(struct inode *inode, u64 block, int nr,
 
 			clear_buffer_uptodate(bh);
 			get_bh(bh); /* for end_buffer_read_sync() */
+			if (validate)
+				set_buffer_needs_validate(bh);
 			bh->b_end_io = end_buffer_read_sync;
 			submit_bh(READ, bh);
 			continue;
@@ -328,6 +344,20 @@ int ocfs2_read_blocks(struct inode *inode, u64 block, int nr,
 				bhs[i] = NULL;
 				continue;
 			}
+
+			if (buffer_needs_validate(bh)) {
+				/* We never set NeedsValidate if the
+				 * buffer was held by the journal, so
+				 * that better not have changed */
+				BUG_ON(buffer_jbd(bh));
+				clear_buffer_needs_validate(bh);
+				status = validate(inode->i_sb, bh);
+				if (status) {
+					put_bh(bh);
+					bhs[i] = NULL;
+					continue;
+				}
+			}
 		}
 
 		/* Always set the buffer in the cache, even if it was

fs/ocfs2/buffer_head_io.h

@@ -31,21 +31,24 @@
 void ocfs2_end_buffer_io_sync(struct buffer_head *bh,
 			     int uptodate);
 
-static inline int ocfs2_read_block(struct inode	       *inode,
-				   u64                  off,
-				   struct buffer_head **bh);
-
 int ocfs2_write_block(struct ocfs2_super          *osb,
 		      struct buffer_head  *bh,
 		      struct inode        *inode);
-int ocfs2_read_blocks(struct inode	  *inode,
-		      u64                  block,
-		      int                  nr,
-		      struct buffer_head  *bhs[],
-		      int                  flags);
 int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block,
 			   unsigned int nr, struct buffer_head *bhs[]);
 
+/*
+ * If not NULL, validate() will be called on a buffer that is freshly
+ * read from disk.  It will not be called if the buffer was in cache.
+ * Note that if validate() is being used for this buffer, it needs to
+ * be set even for a READAHEAD call, as it marks the buffer for later
+ * validation.
+ */
+int ocfs2_read_blocks(struct inode *inode, u64 block, int nr,
+		      struct buffer_head *bhs[], int flags,
+		      int (*validate)(struct super_block *sb,
+				      struct buffer_head *bh));
+
 int ocfs2_write_super_or_backup(struct ocfs2_super *osb,
 				struct buffer_head *bh);
 
@@ -53,7 +56,9 @@ int ocfs2_write_super_or_backup(struct ocfs2_super *osb,
 #define OCFS2_BH_READAHEAD         8
 
 static inline int ocfs2_read_block(struct inode *inode, u64 off,
-				   struct buffer_head **bh)
+				   struct buffer_head **bh,
+				   int (*validate)(struct super_block *sb,
+						   struct buffer_head *bh))
 {
 	int status = 0;
 
@@ -63,7 +68,7 @@ static inline int ocfs2_read_block(struct inode *inode, u64 off,
 		goto bail;
 	}
 
-	status = ocfs2_read_blocks(inode, off, 1, bh, 0);
+	status = ocfs2_read_blocks(inode, off, 1, bh, 0, validate);
 
 bail:
 	return status;

fs/ocfs2/cluster/masklog.c

@@ -110,6 +110,7 @@ static struct mlog_attribute mlog_attrs[MLOG_MAX_BITS] = {
 	define_mask(QUORUM),
 	define_mask(EXPORT),
 	define_mask(XATTR),
+	define_mask(QUOTA),
 	define_mask(ERROR),
 	define_mask(NOTICE),
 	define_mask(KTHREAD),

fs/ocfs2/cluster/masklog.h

@@ -113,6 +113,7 @@
 #define ML_QUORUM	0x0000000008000000ULL /* net connection quorum */
 #define ML_EXPORT	0x0000000010000000ULL /* ocfs2 export operations */
 #define ML_XATTR	0x0000000020000000ULL /* ocfs2 extended attributes */
+#define ML_QUOTA	0x0000000040000000ULL /* ocfs2 quota operations */
 /* bits that are infrequently given and frequently matched in the high word */
 #define ML_ERROR	0x0000000100000000ULL /* sent to KERN_ERR */
 #define ML_NOTICE	0x0000000200000000ULL /* setn to KERN_NOTICE */

fs/ocfs2/dir.c

@@ -40,6 +40,7 @@
 #include <linux/types.h>
 #include <linux/slab.h>
 #include <linux/highmem.h>
+#include <linux/quotaops.h>
 
 #define MLOG_MASK_PREFIX ML_NAMEI
 #include <cluster/masklog.h>
@@ -47,6 +48,7 @@
 #include "ocfs2.h"
 
 #include "alloc.h"
+#include "blockcheck.h"
 #include "dir.h"
 #include "dlmglue.h"
 #include "extent_map.h"
@@ -82,47 +84,72 @@ static int ocfs2_do_extend_dir(struct super_block *sb,
 			       struct ocfs2_alloc_context *meta_ac,
 			       struct buffer_head **new_bh);
 
-static struct buffer_head *ocfs2_bread(struct inode *inode,
-				       int block, int *err, int reada)
+/*
+ * These are distinct checks because future versions of the file system will
+ * want to have a trailing dirent structure independent of indexing.
+ */
+static int ocfs2_dir_has_trailer(struct inode *dir)
 {
-	struct buffer_head *bh = NULL;
-	int tmperr;
-	u64 p_blkno;
-	int readflags = 0;
+	if (OCFS2_I(dir)->ip_dyn_features & OCFS2_INLINE_DATA_FL)
+		return 0;
 
-	if (reada)
-		readflags |= OCFS2_BH_READAHEAD;
+	return ocfs2_meta_ecc(OCFS2_SB(dir->i_sb));
+}
 
-	if (((u64)block << inode->i_sb->s_blocksize_bits) >=
-	    i_size_read(inode)) {
-		BUG_ON(!reada);
-		return NULL;
-	}
+static int ocfs2_supports_dir_trailer(struct ocfs2_super *osb)
+{
+	return ocfs2_meta_ecc(osb);
+}
 
-	down_read(&OCFS2_I(inode)->ip_alloc_sem);
-	tmperr = ocfs2_extent_map_get_blocks(inode, block, &p_blkno, NULL,
-					     NULL);
-	up_read(&OCFS2_I(inode)->ip_alloc_sem);
-	if (tmperr < 0) {
-		mlog_errno(tmperr);
-		goto fail;
-	}
+static inline unsigned int ocfs2_dir_trailer_blk_off(struct super_block *sb)
+{
+	return sb->s_blocksize - sizeof(struct ocfs2_dir_block_trailer);
+}
 
-	tmperr = ocfs2_read_blocks(inode, p_blkno, 1, &bh, readflags);
-	if (tmperr < 0)
-		goto fail;
+#define ocfs2_trailer_from_bh(_bh, _sb) ((struct ocfs2_dir_block_trailer *) ((_bh)->b_data + ocfs2_dir_trailer_blk_off((_sb))))
 
-	tmperr = 0;
+/* XXX ocfs2_block_dqtrailer() is similar but not quite - can we make
+ * them more consistent? */
+struct ocfs2_dir_block_trailer *ocfs2_dir_trailer_from_size(int blocksize,
+							    void *data)
+{
+	char *p = data;
 
-	*err = 0;
-	return bh;
+	p += blocksize - sizeof(struct ocfs2_dir_block_trailer);
+	return (struct ocfs2_dir_block_trailer *)p;
+}
 
-fail:
-	brelse(bh);
-	bh = NULL;
+/*
+ * XXX: This is executed once on every dirent. We should consider optimizing
+ * it.
+ */
+static int ocfs2_skip_dir_trailer(struct inode *dir,
+				  struct ocfs2_dir_entry *de,
+				  unsigned long offset,
+				  unsigned long blklen)
+{
+	unsigned long toff = blklen - sizeof(struct ocfs2_dir_block_trailer);
 
-	*err = -EIO;
-	return NULL;
+	if (!ocfs2_dir_has_trailer(dir))
+		return 0;
+
+	if (offset != toff)
+		return 0;
+
+	return 1;
+}
+
+static void ocfs2_init_dir_trailer(struct inode *inode,
+				   struct buffer_head *bh)
+{
+	struct ocfs2_dir_block_trailer *trailer;
+
+	trailer = ocfs2_trailer_from_bh(bh, inode->i_sb);
+	strcpy(trailer->db_signature, OCFS2_DIR_TRAILER_SIGNATURE);
+	trailer->db_compat_rec_len =
+			cpu_to_le16(sizeof(struct ocfs2_dir_block_trailer));
+	trailer->db_parent_dinode = cpu_to_le64(OCFS2_I(inode)->ip_blkno);
+	trailer->db_blkno = cpu_to_le64(bh->b_blocknr);
 }
 
 /*
@@ -231,7 +258,7 @@ static struct buffer_head *ocfs2_find_entry_id(const char *name,
 	struct ocfs2_dinode *di;
 	struct ocfs2_inline_data *data;
 
-	ret = ocfs2_read_block(dir, OCFS2_I(dir)->ip_blkno, &di_bh);
+	ret = ocfs2_read_inode_block(dir, &di_bh);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -250,6 +277,108 @@ out:
 	return NULL;
 }
 
+static int ocfs2_validate_dir_block(struct super_block *sb,
+				    struct buffer_head *bh)
+{
+	int rc;
+	struct ocfs2_dir_block_trailer *trailer =
+		ocfs2_trailer_from_bh(bh, sb);
+
+
+	/*
+	 * We don't validate dirents here, that's handled
+	 * in-place when the code walks them.
+	 */
+	mlog(0, "Validating dirblock %llu\n",
+	     (unsigned long long)bh->b_blocknr);
+
+	BUG_ON(!buffer_uptodate(bh));
+
+	/*
+	 * If the ecc fails, we return the error but otherwise
+	 * leave the filesystem running.  We know any error is
+	 * local to this block.
+	 *
+	 * Note that we are safe to call this even if the directory
+	 * doesn't have a trailer.  Filesystems without metaecc will do
+	 * nothing, and filesystems with it will have one.
+	 */
+	rc = ocfs2_validate_meta_ecc(sb, bh->b_data, &trailer->db_check);
+	if (rc)
+		mlog(ML_ERROR, "Checksum failed for dinode %llu\n",
+		     (unsigned long long)bh->b_blocknr);
+
+	return rc;
+}
+
+/*
+ * This function forces all errors to -EIO for consistency with its
+ * predecessor, ocfs2_bread().  We haven't audited what returning the
+ * real error codes would do to callers.  We log the real codes with
+ * mlog_errno() before we squash them.
+ */
+static int ocfs2_read_dir_block(struct inode *inode, u64 v_block,
+				struct buffer_head **bh, int flags)
+{
+	int rc = 0;
+	struct buffer_head *tmp = *bh;
+	struct ocfs2_dir_block_trailer *trailer;
+
+	rc = ocfs2_read_virt_blocks(inode, v_block, 1, &tmp, flags,
+				    ocfs2_validate_dir_block);
+	if (rc) {
+		mlog_errno(rc);
+		goto out;
+	}
+
+	/*
+	 * We check the trailer here rather than in
+	 * ocfs2_validate_dir_block() because that function doesn't have
+	 * the inode to test.
+	 */
+	if (!(flags & OCFS2_BH_READAHEAD) &&
+	    ocfs2_dir_has_trailer(inode)) {
+		trailer = ocfs2_trailer_from_bh(tmp, inode->i_sb);
+		if (!OCFS2_IS_VALID_DIR_TRAILER(trailer)) {
+			rc = -EINVAL;
+			ocfs2_error(inode->i_sb,
+				    "Invalid dirblock #%llu: "
+				    "signature = %.*s\n",
+				    (unsigned long long)tmp->b_blocknr, 7,
+				    trailer->db_signature);
+			goto out;
+		}
+		if (le64_to_cpu(trailer->db_blkno) != tmp->b_blocknr) {
+			rc = -EINVAL;
+			ocfs2_error(inode->i_sb,
+				    "Directory block #%llu has an invalid "
+				    "db_blkno of %llu",
+				    (unsigned long long)tmp->b_blocknr,
+				    (unsigned long long)le64_to_cpu(trailer->db_blkno));
+			goto out;
+		}
+		if (le64_to_cpu(trailer->db_parent_dinode) !=
+		    OCFS2_I(inode)->ip_blkno) {
+			rc = -EINVAL;
+			ocfs2_error(inode->i_sb,
+				    "Directory block #%llu on dinode "
+				    "#%llu has an invalid parent_dinode "
+				    "of %llu",
+				    (unsigned long long)tmp->b_blocknr,
+				    (unsigned long long)OCFS2_I(inode)->ip_blkno,
+				    (unsigned long long)le64_to_cpu(trailer->db_blkno));
+			goto out;
+		}
+	}
+
+	/* If ocfs2_read_virt_blocks() got us a new bh, pass it up. */
+	if (!*bh)
+		*bh = tmp;
+
+out:
+	return rc ? -EIO : 0;
+}
+
 static struct buffer_head *ocfs2_find_entry_el(const char *name, int namelen,
 					       struct inode *dir,
 					       struct ocfs2_dir_entry **res_dir)
@@ -296,15 +425,17 @@ restart:
 				}
 				num++;
 
-				bh = ocfs2_bread(dir, b++, &err, 1);
+				bh = NULL;
+				err = ocfs2_read_dir_block(dir, b++, &bh,
+							   OCFS2_BH_READAHEAD);
 				bh_use[ra_max] = bh;
 			}
 		}
 		if ((bh = bh_use[ra_ptr++]) == NULL)
 			goto next;
-		if (ocfs2_read_block(dir, block, &bh)) {
+		if (ocfs2_read_dir_block(dir, block, &bh, 0)) {
 			/* read error, skip block & hope for the best.
-			 * ocfs2_read_block() has released the bh. */
+			 * ocfs2_read_dir_block() has released the bh. */
 			ocfs2_error(dir->i_sb, "reading directory %llu, "
 				    "offset %lu\n",
 				    (unsigned long long)OCFS2_I(dir)->ip_blkno,
@@ -381,14 +512,18 @@ int ocfs2_update_entry(struct inode *dir, handle_t *handle,
 		       struct inode *new_entry_inode)
 {
 	int ret;
+	ocfs2_journal_access_func access = ocfs2_journal_access_db;
 
 	/*
 	 * The same code works fine for both inline-data and extent
-	 * based directories, so no need to split this up.
+	 * based directories, so no need to split this up.  The only
+	 * difference is the journal_access function.
 	 */
 
-	ret = ocfs2_journal_access(handle, dir, de_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	if (OCFS2_I(dir)->ip_dyn_features & OCFS2_INLINE_DATA_FL)
+		access = ocfs2_journal_access_di;
+
+	ret = access(handle, dir, de_bh, OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -410,9 +545,13 @@ static int __ocfs2_delete_entry(handle_t *handle, struct inode *dir,
 {
 	struct ocfs2_dir_entry *de, *pde;
 	int i, status = -ENOENT;
+	ocfs2_journal_access_func access = ocfs2_journal_access_db;
 
 	mlog_entry("(0x%p, 0x%p, 0x%p, 0x%p)\n", handle, dir, de_del, bh);
 
+	if (OCFS2_I(dir)->ip_dyn_features & OCFS2_INLINE_DATA_FL)
+		access = ocfs2_journal_access_di;
+
 	i = 0;
 	pde = NULL;
 	de = (struct ocfs2_dir_entry *) first_de;
@@ -423,8 +562,8 @@ static int __ocfs2_delete_entry(handle_t *handle, struct inode *dir,
 			goto bail;
 		}
 		if (de == de_del)  {
-			status = ocfs2_journal_access(handle, dir, bh,
-						      OCFS2_JOURNAL_ACCESS_WRITE);
+			status = access(handle, dir, bh,
+					OCFS2_JOURNAL_ACCESS_WRITE);
 			if (status < 0) {
 				status = -EIO;
 				mlog_errno(status);
@@ -458,7 +597,7 @@ static inline int ocfs2_delete_entry_id(handle_t *handle,
 	struct ocfs2_dinode *di;
 	struct ocfs2_inline_data *data;
 
-	ret = ocfs2_read_block(dir, OCFS2_I(dir)->ip_blkno, &di_bh);
+	ret = ocfs2_read_inode_block(dir, &di_bh);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -576,6 +715,16 @@ int __ocfs2_add_entry(handle_t *handle,
 			goto bail;
 		}
 
+		/* We're guaranteed that we should have space, so we
+		 * can't possibly have hit the trailer...right? */
+		mlog_bug_on_msg(ocfs2_skip_dir_trailer(dir, de, offset, size),
+				"Hit dir trailer trying to insert %.*s "
+			        "(namelen %d) into directory %llu.  "
+				"offset is %lu, trailer offset is %d\n",
+				namelen, name, namelen,
+				(unsigned long long)parent_fe_bh->b_blocknr,
+				offset, ocfs2_dir_trailer_blk_off(dir->i_sb));
+
 		if (ocfs2_dirent_would_fit(de, rec_len)) {
 			dir->i_mtime = dir->i_ctime = CURRENT_TIME;
 			retval = ocfs2_mark_inode_dirty(handle, dir, parent_fe_bh);
@@ -584,8 +733,14 @@ int __ocfs2_add_entry(handle_t *handle,
 				goto bail;
 			}
 
-			status = ocfs2_journal_access(handle, dir, insert_bh,
-						      OCFS2_JOURNAL_ACCESS_WRITE);
+			if (insert_bh == parent_fe_bh)
+				status = ocfs2_journal_access_di(handle, dir,
+								 insert_bh,
+								 OCFS2_JOURNAL_ACCESS_WRITE);
+			else
+				status = ocfs2_journal_access_db(handle, dir,
+								 insert_bh,
+								 OCFS2_JOURNAL_ACCESS_WRITE);
 			/* By now the buffer is marked for journaling */
 			offset += le16_to_cpu(de->rec_len);
 			if (le64_to_cpu(de->inode)) {
@@ -611,6 +766,7 @@ int __ocfs2_add_entry(handle_t *handle,
 			retval = 0;
 			goto bail;
 		}
+
 		offset += le16_to_cpu(de->rec_len);
 		de = (struct ocfs2_dir_entry *) ((char *) de + le16_to_cpu(de->rec_len));
 	}
@@ -636,7 +792,7 @@ static int ocfs2_dir_foreach_blk_id(struct inode *inode,
 	struct ocfs2_inline_data *data;
 	struct ocfs2_dir_entry *de;
 
-	ret = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, &di_bh);
+	ret = ocfs2_read_inode_block(inode, &di_bh);
 	if (ret) {
 		mlog(ML_ERROR, "Unable to read inode block for dir %llu\n",
 		     (unsigned long long)OCFS2_I(inode)->ip_blkno);
@@ -724,7 +880,6 @@ static int ocfs2_dir_foreach_blk_el(struct inode *inode,
 	int i, stored;
 	struct buffer_head * bh, * tmp;
 	struct ocfs2_dir_entry * de;
-	int err;
 	struct super_block * sb = inode->i_sb;
 	unsigned int ra_sectors = 16;
 
@@ -735,12 +890,8 @@ static int ocfs2_dir_foreach_blk_el(struct inode *inode,
 
 	while (!error && !stored && *f_pos < i_size_read(inode)) {
 		blk = (*f_pos) >> sb->s_blocksize_bits;
-		bh = ocfs2_bread(inode, blk, &err, 0);
-		if (!bh) {
-			mlog(ML_ERROR,
-			     "directory #%llu contains a hole at offset %lld\n",
-			     (unsigned long long)OCFS2_I(inode)->ip_blkno,
-			     *f_pos);
+		if (ocfs2_read_dir_block(inode, blk, &bh, 0)) {
+			/* Skip the corrupt dirblock and keep trying */
 			*f_pos += sb->s_blocksize - offset;
 			continue;
 		}
@@ -754,8 +905,10 @@ static int ocfs2_dir_foreach_blk_el(struct inode *inode,
 		    || (((last_ra_blk - blk) << 9) <= (ra_sectors / 2))) {
 			for (i = ra_sectors >> (sb->s_blocksize_bits - 9);
 			     i > 0; i--) {
-				tmp = ocfs2_bread(inode, ++blk, &err, 1);
-				brelse(tmp);
+				tmp = NULL;
+				if (!ocfs2_read_dir_block(inode, ++blk, &tmp,
+							  OCFS2_BH_READAHEAD))
+					brelse(tmp);
 			}
 			last_ra_blk = blk;
 			ra_sectors = 8;
@@ -828,6 +981,7 @@ revalidate:
 		}
 		offset = 0;
 		brelse(bh);
+		bh = NULL;
 	}
 
 	stored = 0;
@@ -1050,9 +1204,15 @@ int ocfs2_empty_dir(struct inode *inode)
 	return !priv.seen_other;
 }
 
-static void ocfs2_fill_initial_dirents(struct inode *inode,
-				       struct inode *parent,
-				       char *start, unsigned int size)
+/*
+ * Fills "." and ".." dirents in a new directory block. Returns dirent for
+ * "..", which might be used during creation of a directory with a trailing
+ * header. It is otherwise safe to ignore the return code.
+ */
+static struct ocfs2_dir_entry *ocfs2_fill_initial_dirents(struct inode *inode,
+							  struct inode *parent,
+							  char *start,
+							  unsigned int size)
 {
 	struct ocfs2_dir_entry *de = (struct ocfs2_dir_entry *)start;
 
@@ -1069,6 +1229,8 @@ static void ocfs2_fill_initial_dirents(struct inode *inode,
 	de->name_len = 2;
 	strcpy(de->name, "..");
 	ocfs2_set_de_type(de, S_IFDIR);
+
+	return de;
 }
 
 /*
@@ -1086,8 +1248,8 @@ static int ocfs2_fill_new_dir_id(struct ocfs2_super *osb,
 	struct ocfs2_inline_data *data = &di->id2.i_data;
 	unsigned int size = le16_to_cpu(data->id_count);
 
-	ret = ocfs2_journal_access(handle, inode, di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -1121,10 +1283,15 @@ static int ocfs2_fill_new_dir_el(struct ocfs2_super *osb,
 				 struct ocfs2_alloc_context *data_ac)
 {
 	int status;
+	unsigned int size = osb->sb->s_blocksize;
 	struct buffer_head *new_bh = NULL;
+	struct ocfs2_dir_entry *de;
 
 	mlog_entry_void();
 
+	if (ocfs2_supports_dir_trailer(osb))
+		size = ocfs2_dir_trailer_blk_off(parent->i_sb);
+
 	status = ocfs2_do_extend_dir(osb->sb, handle, inode, fe_bh,
 				     data_ac, NULL, &new_bh);
 	if (status < 0) {
@@ -1134,16 +1301,17 @@ static int ocfs2_fill_new_dir_el(struct ocfs2_super *osb,
 
 	ocfs2_set_new_buffer_uptodate(inode, new_bh);
 
-	status = ocfs2_journal_access(handle, inode, new_bh,
-				      OCFS2_JOURNAL_ACCESS_CREATE);
+	status = ocfs2_journal_access_db(handle, inode, new_bh,
+					 OCFS2_JOURNAL_ACCESS_CREATE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
 	memset(new_bh->b_data, 0, osb->sb->s_blocksize);
 
-	ocfs2_fill_initial_dirents(inode, parent, new_bh->b_data,
-				   osb->sb->s_blocksize);
+	de = ocfs2_fill_initial_dirents(inode, parent, new_bh->b_data, size);
+	if (ocfs2_supports_dir_trailer(osb))
+		ocfs2_init_dir_trailer(inode, new_bh);
 
 	status = ocfs2_journal_dirty(handle, new_bh);
 	if (status < 0) {
@@ -1184,13 +1352,27 @@ int ocfs2_fill_new_dir(struct ocfs2_super *osb,
 				     data_ac);
 }
 
+/*
+ * Expand rec_len of the rightmost dirent in a directory block so that it
+ * contains the end of our valid space for dirents. We do this during
+ * expansion from an inline directory to one with extents. The first dir block
+ * in that case is taken from the inline data portion of the inode block.
+ *
+ * We add the dir trailer if this filesystem wants it.
+ */
 static void ocfs2_expand_last_dirent(char *start, unsigned int old_size,
-				     unsigned int new_size)
+				     struct super_block *sb)
 {
 	struct ocfs2_dir_entry *de;
 	struct ocfs2_dir_entry *prev_de;
 	char *de_buf, *limit;
-	unsigned int bytes = new_size - old_size;
+	unsigned int new_size = sb->s_blocksize;
+	unsigned int bytes;
+
+	if (ocfs2_supports_dir_trailer(OCFS2_SB(sb)))
+		new_size = ocfs2_dir_trailer_blk_off(sb);
+
+	bytes = new_size - old_size;
 
 	limit = start + old_size;
 	de_buf = start;
@@ -1216,9 +1398,9 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 				   unsigned int blocks_wanted,
 				   struct buffer_head **first_block_bh)
 {
-	int ret, credits = OCFS2_INLINE_TO_EXTENTS_CREDITS;
 	u32 alloc, bit_off, len;
 	struct super_block *sb = dir->i_sb;
+	int ret, credits = ocfs2_inline_to_extents_credits(sb);
 	u64 blkno, bytes = blocks_wanted << sb->s_blocksize_bits;
 	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
 	struct ocfs2_inode_info *oi = OCFS2_I(dir);
@@ -1227,6 +1409,7 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
 	handle_t *handle;
 	struct ocfs2_extent_tree et;
+	int did_quota = 0;
 
 	ocfs2_init_dinode_extent_tree(&et, dir, di_bh);
 
@@ -1264,6 +1447,12 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 		goto out_sem;
 	}
 
+	if (vfs_dq_alloc_space_nodirty(dir,
+				ocfs2_clusters_to_bytes(osb->sb, alloc))) {
+		ret = -EDQUOT;
+		goto out_commit;
+	}
+	did_quota = 1;
 	/*
 	 * Try to claim as many clusters as the bitmap can give though
 	 * if we only get one now, that's enough to continue. The rest
@@ -1290,8 +1479,8 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 
 	ocfs2_set_new_buffer_uptodate(dir, dirdata_bh);
 
-	ret = ocfs2_journal_access(handle, dir, dirdata_bh,
-				   OCFS2_JOURNAL_ACCESS_CREATE);
+	ret = ocfs2_journal_access_db(handle, dir, dirdata_bh,
+				      OCFS2_JOURNAL_ACCESS_CREATE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -1300,8 +1489,9 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 	memcpy(dirdata_bh->b_data, di->id2.i_data.id_data, i_size_read(dir));
 	memset(dirdata_bh->b_data + i_size_read(dir), 0,
 	       sb->s_blocksize - i_size_read(dir));
-	ocfs2_expand_last_dirent(dirdata_bh->b_data, i_size_read(dir),
-				 sb->s_blocksize);
+	ocfs2_expand_last_dirent(dirdata_bh->b_data, i_size_read(dir), sb);
+	if (ocfs2_supports_dir_trailer(osb))
+		ocfs2_init_dir_trailer(dir, dirdata_bh);
 
 	ret = ocfs2_journal_dirty(handle, dirdata_bh);
 	if (ret) {
@@ -1317,8 +1507,8 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 	 * We let the later dirent insert modify c/mtime - to the user
 	 * the data hasn't changed.
 	 */
-	ret = ocfs2_journal_access(handle, dir, di_bh,
-				   OCFS2_JOURNAL_ACCESS_CREATE);
+	ret = ocfs2_journal_access_di(handle, dir, di_bh,
+				      OCFS2_JOURNAL_ACCESS_CREATE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -1386,6 +1576,9 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh,
 	dirdata_bh = NULL;
 
 out_commit:
+	if (ret < 0 && did_quota)
+		vfs_dq_free_space_nodirty(dir,
+			ocfs2_clusters_to_bytes(osb->sb, 2));
 	ocfs2_commit_trans(osb, handle);
 
 out_sem:
@@ -1410,7 +1603,7 @@ static int ocfs2_do_extend_dir(struct super_block *sb,
 			       struct buffer_head **new_bh)
 {
 	int status;
-	int extend;
+	int extend, did_quota = 0;
 	u64 p_blkno, v_blkno;
 
 	spin_lock(&OCFS2_I(dir)->ip_lock);
@@ -1420,6 +1613,13 @@ static int ocfs2_do_extend_dir(struct super_block *sb,
 	if (extend) {
 		u32 offset = OCFS2_I(dir)->ip_clusters;
 
+		if (vfs_dq_alloc_space_nodirty(dir,
+					ocfs2_clusters_to_bytes(sb, 1))) {
+			status = -EDQUOT;
+			goto bail;
+		}
+		did_quota = 1;
+
 		status = ocfs2_add_inode_data(OCFS2_SB(sb), dir, &offset,
 					      1, 0, parent_fe_bh, handle,
 					      data_ac, meta_ac, NULL);
@@ -1445,6 +1645,8 @@ static int ocfs2_do_extend_dir(struct super_block *sb,
 	}
 	status = 0;
 bail:
+	if (did_quota && status < 0)
+		vfs_dq_free_space_nodirty(dir, ocfs2_clusters_to_bytes(sb, 1));
 	mlog_exit(status);
 	return status;
 }
@@ -1569,16 +1771,22 @@ do_extend:
 
 	ocfs2_set_new_buffer_uptodate(dir, new_bh);
 
-	status = ocfs2_journal_access(handle, dir, new_bh,
-				      OCFS2_JOURNAL_ACCESS_CREATE);
+	status = ocfs2_journal_access_db(handle, dir, new_bh,
+					 OCFS2_JOURNAL_ACCESS_CREATE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
 	memset(new_bh->b_data, 0, sb->s_blocksize);
+
 	de = (struct ocfs2_dir_entry *) new_bh->b_data;
 	de->inode = 0;
-	de->rec_len = cpu_to_le16(sb->s_blocksize);
+	if (ocfs2_dir_has_trailer(dir)) {
+		de->rec_len = cpu_to_le16(ocfs2_dir_trailer_blk_off(sb));
+		ocfs2_init_dir_trailer(dir, new_bh);
+	} else {
+		de->rec_len = cpu_to_le16(sb->s_blocksize);
+	}
 	status = ocfs2_journal_dirty(handle, new_bh);
 	if (status < 0) {
 		mlog_errno(status);
@@ -1620,11 +1828,21 @@ static int ocfs2_find_dir_space_id(struct inode *dir, struct buffer_head *di_bh,
 				   unsigned int *blocks_wanted)
 {
 	int ret;
+	struct super_block *sb = dir->i_sb;
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
 	struct ocfs2_dir_entry *de, *last_de = NULL;
 	char *de_buf, *limit;
 	unsigned long offset = 0;
-	unsigned int rec_len, new_rec_len;
+	unsigned int rec_len, new_rec_len, free_space = dir->i_sb->s_blocksize;
+
+	/*
+	 * This calculates how many free bytes we'd have in block zero, should
+	 * this function force expansion to an extent tree.
+	 */
+	if (ocfs2_supports_dir_trailer(OCFS2_SB(sb)))
+		free_space = ocfs2_dir_trailer_blk_off(sb) - i_size_read(dir);
+	else
+		free_space = dir->i_sb->s_blocksize - i_size_read(dir);
 
 	de_buf = di->id2.i_data.id_data;
 	limit = de_buf + i_size_read(dir);
@@ -1641,6 +1859,11 @@ static int ocfs2_find_dir_space_id(struct inode *dir, struct buffer_head *di_bh,
 			ret = -EEXIST;
 			goto out;
 		}
+		/*
+		 * No need to check for a trailing dirent record here as
+		 * they're not used for inline dirs.
+		 */
+
 		if (ocfs2_dirent_would_fit(de, rec_len)) {
 			/* Ok, we found a spot. Return this bh and let
 			 * the caller actually fill it in. */
@@ -1661,7 +1884,7 @@ static int ocfs2_find_dir_space_id(struct inode *dir, struct buffer_head *di_bh,
 	 * dirent can be found.
 	 */
 	*blocks_wanted = 1;
-	new_rec_len = le16_to_cpu(last_de->rec_len) + (dir->i_sb->s_blocksize - i_size_read(dir));
+	new_rec_len = le16_to_cpu(last_de->rec_len) + free_space;
 	if (new_rec_len < (rec_len + OCFS2_DIR_REC_LEN(last_de->name_len)))
 		*blocks_wanted = 2;
 
@@ -1679,9 +1902,10 @@ static int ocfs2_find_dir_space_el(struct inode *dir, const char *name,
 	struct ocfs2_dir_entry *de;
 	struct super_block *sb = dir->i_sb;
 	int status;
+	int blocksize = dir->i_sb->s_blocksize;
 
-	bh = ocfs2_bread(dir, 0, &status, 0);
-	if (!bh) {
+	status = ocfs2_read_dir_block(dir, 0, &bh, 0);
+	if (status) {
 		mlog_errno(status);
 		goto bail;
 	}
@@ -1702,11 +1926,10 @@ static int ocfs2_find_dir_space_el(struct inode *dir, const char *name,
 				status = -ENOSPC;
 				goto bail;
 			}
-			bh = ocfs2_bread(dir,
-					 offset >> sb->s_blocksize_bits,
-					 &status,
-					 0);
-			if (!bh) {
+			status = ocfs2_read_dir_block(dir,
+					     offset >> sb->s_blocksize_bits,
+					     &bh, 0);
+			if (status) {
 				mlog_errno(status);
 				goto bail;
 			}
@@ -1721,6 +1944,11 @@ static int ocfs2_find_dir_space_el(struct inode *dir, const char *name,
 			status = -EEXIST;
 			goto bail;
 		}
+
+		if (ocfs2_skip_dir_trailer(dir, de, offset % blocksize,
+					   blocksize))
+			goto next;
+
 		if (ocfs2_dirent_would_fit(de, rec_len)) {
 			/* Ok, we found a spot. Return this bh and let
 			 * the caller actually fill it in. */
@@ -1729,6 +1957,7 @@ static int ocfs2_find_dir_space_el(struct inode *dir, const char *name,
 			status = 0;
 			goto bail;
 		}
+next:
 		offset += le16_to_cpu(de->rec_len);
 		de = (struct ocfs2_dir_entry *)((char *) de + le16_to_cpu(de->rec_len));
 	}

fs/ocfs2/dir.h

@@ -83,4 +83,6 @@ int ocfs2_fill_new_dir(struct ocfs2_super *osb,
 		       struct buffer_head *fe_bh,
 		       struct ocfs2_alloc_context *data_ac);
 
+struct ocfs2_dir_block_trailer *ocfs2_dir_trailer_from_size(int blocksize,
+							    void *data);
 #endif /* OCFS2_DIR_H */

fs/ocfs2/dlm/dlmast.c

@@ -275,6 +275,7 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 	struct list_head *iter, *head=NULL;
 	u64 cookie;
 	u32 flags;
+	u8 node;
 
 	if (!dlm_grab(dlm)) {
 		dlm_error(DLM_REJECTED);
@@ -286,18 +287,21 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 
 	name = past->name;
 	locklen = past->namelen;
-	cookie = be64_to_cpu(past->cookie);
+	cookie = past->cookie;
 	flags = be32_to_cpu(past->flags);
+	node = past->node_idx;
 
 	if (locklen > DLM_LOCKID_NAME_MAX) {
 		ret = DLM_IVBUFLEN;
-		mlog(ML_ERROR, "Invalid name length in proxy ast handler!\n");
+		mlog(ML_ERROR, "Invalid name length (%d) in proxy ast "
+		     "handler!\n", locklen);
 		goto leave;
 	}
 
 	if ((flags & (LKM_PUT_LVB|LKM_GET_LVB)) ==
 	     (LKM_PUT_LVB|LKM_GET_LVB)) {
-		mlog(ML_ERROR, "both PUT and GET lvb specified\n");
+		mlog(ML_ERROR, "Both PUT and GET lvb specified, (0x%x)\n",
+		     flags);
 		ret = DLM_BADARGS;
 		goto leave;
 	}
@@ -310,22 +314,21 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 	if (past->type != DLM_AST &&
 	    past->type != DLM_BAST) {
 		mlog(ML_ERROR, "Unknown ast type! %d, cookie=%u:%llu"
-		     "name=%.*s\n", past->type, 
-		     dlm_get_lock_cookie_node(cookie),
-		     dlm_get_lock_cookie_seq(cookie),
-		     locklen, name);
+		     "name=%.*s, node=%u\n", past->type,
+		     dlm_get_lock_cookie_node(be64_to_cpu(cookie)),
+		     dlm_get_lock_cookie_seq(be64_to_cpu(cookie)),
+		     locklen, name, node);
 		ret = DLM_IVLOCKID;
 		goto leave;
 	}
 
 	res = dlm_lookup_lockres(dlm, name, locklen);
 	if (!res) {
-		mlog(0, "got %sast for unknown lockres! "
-		     "cookie=%u:%llu, name=%.*s, namelen=%u\n",
-		     past->type == DLM_AST ? "" : "b",
-		     dlm_get_lock_cookie_node(cookie),
-		     dlm_get_lock_cookie_seq(cookie),
-		     locklen, name, locklen);
+		mlog(0, "Got %sast for unknown lockres! cookie=%u:%llu, "
+		     "name=%.*s, node=%u\n", (past->type == DLM_AST ? "" : "b"),
+		     dlm_get_lock_cookie_node(be64_to_cpu(cookie)),
+		     dlm_get_lock_cookie_seq(be64_to_cpu(cookie)),
+		     locklen, name, node);
 		ret = DLM_IVLOCKID;
 		goto leave;
 	}
@@ -337,12 +340,12 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 
 	spin_lock(&res->spinlock);
 	if (res->state & DLM_LOCK_RES_RECOVERING) {
-		mlog(0, "responding with DLM_RECOVERING!\n");
+		mlog(0, "Responding with DLM_RECOVERING!\n");
 		ret = DLM_RECOVERING;
 		goto unlock_out;
 	}
 	if (res->state & DLM_LOCK_RES_MIGRATING) {
-		mlog(0, "responding with DLM_MIGRATING!\n");
+		mlog(0, "Responding with DLM_MIGRATING!\n");
 		ret = DLM_MIGRATING;
 		goto unlock_out;
 	}
@@ -351,7 +354,7 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 	lock = NULL;
 	list_for_each(iter, head) {
 		lock = list_entry (iter, struct dlm_lock, list);
-		if (be64_to_cpu(lock->ml.cookie) == cookie)
+		if (lock->ml.cookie == cookie)
 			goto do_ast;
 	}
 
@@ -363,15 +366,15 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data,
 
 	list_for_each(iter, head) {
 		lock = list_entry (iter, struct dlm_lock, list);
-		if (be64_to_cpu(lock->ml.cookie) == cookie)
+		if (lock->ml.cookie == cookie)
 			goto do_ast;
 	}
 
-	mlog(0, "got %sast for unknown lock!  cookie=%u:%llu, "
-	     "name=%.*s, namelen=%u\n", past->type == DLM_AST ? "" : "b", 
-	     dlm_get_lock_cookie_node(cookie),
-	     dlm_get_lock_cookie_seq(cookie),
-	     locklen, name, locklen);
+	mlog(0, "Got %sast for unknown lock! cookie=%u:%llu, name=%.*s, "
+	     "node=%u\n", past->type == DLM_AST ? "" : "b",
+	     dlm_get_lock_cookie_node(be64_to_cpu(cookie)),
+	     dlm_get_lock_cookie_seq(be64_to_cpu(cookie)),
+	     locklen, name, node);
 
 	ret = DLM_NORMAL;
 unlock_out:
@@ -383,8 +386,8 @@ do_ast:
 	if (past->type == DLM_AST) {
 		/* do not alter lock refcount.  switching lists. */
 		list_move_tail(&lock->list, &res->granted);
-		mlog(0, "ast: adding to granted list... type=%d, "
-			  "convert_type=%d\n", lock->ml.type, lock->ml.convert_type);
+		mlog(0, "ast: Adding to granted list... type=%d, "
+		     "convert_type=%d\n", lock->ml.type, lock->ml.convert_type);
 		if (lock->ml.convert_type != LKM_IVMODE) {
 			lock->ml.type = lock->ml.convert_type;
 			lock->ml.convert_type = LKM_IVMODE;
@@ -408,7 +411,6 @@ do_ast:
 		dlm_do_local_bast(dlm, res, lock, past->blocked_type);
 
 leave:
-
 	if (res)
 		dlm_lockres_put(res);
 

fs/ocfs2/dlm/dlmcommon.h

@@ -140,6 +140,7 @@ struct dlm_ctxt
 	unsigned int purge_count;
 	spinlock_t spinlock;
 	spinlock_t ast_lock;
+	spinlock_t track_lock;
 	char *name;
 	u8 node_num;
 	u32 key;
@@ -316,6 +317,8 @@ struct dlm_lock_resource
 	 * put on a list for the dlm thread to run. */
 	unsigned long    last_used;
 
+	struct dlm_ctxt *dlm;
+
 	unsigned migration_pending:1;
 	atomic_t asts_reserved;
 	spinlock_t spinlock;

fs/ocfs2/dlm/dlmdebug.c

@@ -630,43 +630,38 @@ static void *lockres_seq_start(struct seq_file *m, loff_t *pos)
 {
 	struct debug_lockres *dl = m->private;
 	struct dlm_ctxt *dlm = dl->dl_ctxt;
+	struct dlm_lock_resource *oldres = dl->dl_res;
 	struct dlm_lock_resource *res = NULL;
+	struct list_head *track_list;
 
-	spin_lock(&dlm->spinlock);
+	spin_lock(&dlm->track_lock);
+	if (oldres)
+		track_list = &oldres->tracking;
+	else
+		track_list = &dlm->tracking_list;
 
-	if (dl->dl_res) {
-		list_for_each_entry(res, &dl->dl_res->tracking, tracking) {
-			if (dl->dl_res) {
-				dlm_lockres_put(dl->dl_res);
-				dl->dl_res = NULL;
-			}
-			if (&res->tracking == &dlm->tracking_list) {
-				mlog(0, "End of list found, %p\n", res);
-				dl = NULL;
-				break;
-			}
+	list_for_each_entry(res, track_list, tracking) {
+		if (&res->tracking == &dlm->tracking_list)
+			res = NULL;
+		else
 			dlm_lockres_get(res);
-			dl->dl_res = res;
-			break;
-		}
-	} else {
-		if (!list_empty(&dlm->tracking_list)) {
-			list_for_each_entry(res, &dlm->tracking_list, tracking)
-				break;
-			dlm_lockres_get(res);
-			dl->dl_res = res;
-		} else
-			dl = NULL;
+		break;
 	}
+	spin_unlock(&dlm->track_lock);
 
-	if (dl) {
-		spin_lock(&dl->dl_res->spinlock);
-		dump_lockres(dl->dl_res, dl->dl_buf, dl->dl_len - 1);
-		spin_unlock(&dl->dl_res->spinlock);
-	}
+	if (oldres)
+		dlm_lockres_put(oldres);
 
-	spin_unlock(&dlm->spinlock);
+	dl->dl_res = res;
+
+	if (res) {
+		spin_lock(&res->spinlock);
+		dump_lockres(res, dl->dl_buf, dl->dl_len - 1);
+		spin_unlock(&res->spinlock);
+	} else
+		dl = NULL;
 
+	/* passed to seq_show */
 	return dl;
 }
 

fs/ocfs2/dlm/dlmdomain.c

@@ -1550,6 +1550,7 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain,
 	spin_lock_init(&dlm->spinlock);
 	spin_lock_init(&dlm->master_lock);
 	spin_lock_init(&dlm->ast_lock);
+	spin_lock_init(&dlm->track_lock);
 	INIT_LIST_HEAD(&dlm->list);
 	INIT_LIST_HEAD(&dlm->dirty_list);
 	INIT_LIST_HEAD(&dlm->reco.resources);

fs/ocfs2/dlm/dlmmaster.c

@@ -505,8 +505,10 @@ void dlm_change_lockres_owner(struct dlm_ctxt *dlm,
 static void dlm_lockres_release(struct kref *kref)
 {
 	struct dlm_lock_resource *res;
+	struct dlm_ctxt *dlm;
 
 	res = container_of(kref, struct dlm_lock_resource, refs);
+	dlm = res->dlm;
 
 	/* This should not happen -- all lockres' have a name
 	 * associated with them at init time. */
@@ -515,6 +517,7 @@ static void dlm_lockres_release(struct kref *kref)
 	mlog(0, "destroying lockres %.*s\n", res->lockname.len,
 	     res->lockname.name);
 
+	spin_lock(&dlm->track_lock);
 	if (!list_empty(&res->tracking))
 		list_del_init(&res->tracking);
 	else {
@@ -522,6 +525,9 @@ static void dlm_lockres_release(struct kref *kref)
 		     res->lockname.len, res->lockname.name);
 		dlm_print_one_lock_resource(res);
 	}
+	spin_unlock(&dlm->track_lock);
+
+	dlm_put(dlm);
 
 	if (!hlist_unhashed(&res->hash_node) ||
 	    !list_empty(&res->granted) ||
@@ -595,6 +601,10 @@ static void dlm_init_lockres(struct dlm_ctxt *dlm,
 	res->migration_pending = 0;
 	res->inflight_locks = 0;
 
+	/* put in dlm_lockres_release */
+	dlm_grab(dlm);
+	res->dlm = dlm;
+
 	kref_init(&res->refs);
 
 	/* just for consistency */
@@ -722,14 +732,21 @@ lookup:
 	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 */
+		if (tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN) {
+			__dlm_wait_on_lockres(tmpres);
+			BUG_ON(tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN);
+		}
+
 		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);
-		spin_unlock(&dlm->spinlock);
 
 		/* wait until done messaging the master, drop our ref to allow
 		 * the lockres to be purged, start over. */
@@ -2949,7 +2966,7 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
 				  struct dlm_node_iter *iter)
 {
 	struct dlm_migrate_request migrate;
-	int ret, status = 0;
+	int ret, skip, status = 0;
 	int nodenum;
 
 	memset(&migrate, 0, sizeof(migrate));
@@ -2966,12 +2983,27 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
 		    nodenum == new_master)
 			continue;
 
+		/* We could race exit domain. If exited, skip. */
+		spin_lock(&dlm->spinlock);
+		skip = (!test_bit(nodenum, dlm->domain_map));
+		spin_unlock(&dlm->spinlock);
+		if (skip) {
+			clear_bit(nodenum, iter->node_map);
+			continue;
+		}
+
 		ret = o2net_send_message(DLM_MIGRATE_REQUEST_MSG, dlm->key,
 					 &migrate, sizeof(migrate), nodenum,
 					 &status);
-		if (ret < 0)
-			mlog_errno(ret);
-		else if (status < 0) {
+		if (ret < 0) {
+			mlog(0, "migrate_request returned %d!\n", ret);
+			if (!dlm_is_host_down(ret)) {
+				mlog(ML_ERROR, "unhandled error=%d!\n", ret);
+				BUG();
+			}
+			clear_bit(nodenum, iter->node_map);
+			ret = 0;
+		} else if (status < 0) {
 			mlog(0, "migrate request (node %u) returned %d!\n",
 			     nodenum, status);
 			ret = status;

fs/ocfs2/dlm/dlmthread.c

@@ -181,7 +181,8 @@ static int dlm_purge_lockres(struct dlm_ctxt *dlm,
 
 		spin_lock(&res->spinlock);
 		/* This ensures that clear refmap is sent after the set */
-		__dlm_wait_on_lockres_flags(res, DLM_LOCK_RES_SETREF_INPROG);
+		__dlm_wait_on_lockres_flags(res, (DLM_LOCK_RES_SETREF_INPROG |
+						  DLM_LOCK_RES_MIGRATING));
 		spin_unlock(&res->spinlock);
 
 		/* clear our bit from the master's refmap, ignore errors */

fs/ocfs2/dlmglue.c

@@ -32,6 +32,7 @@
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
 #include <linux/time.h>
+#include <linux/quotaops.h>
 
 #define MLOG_MASK_PREFIX ML_DLM_GLUE
 #include <cluster/masklog.h>
@@ -51,6 +52,7 @@
 #include "slot_map.h"
 #include "super.h"
 #include "uptodate.h"
+#include "quota.h"
 
 #include "buffer_head_io.h"
 
@@ -68,6 +70,7 @@ struct ocfs2_mask_waiter {
 static struct ocfs2_super *ocfs2_get_dentry_osb(struct ocfs2_lock_res *lockres);
 static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres);
 static struct ocfs2_super *ocfs2_get_file_osb(struct ocfs2_lock_res *lockres);
+static struct ocfs2_super *ocfs2_get_qinfo_osb(struct ocfs2_lock_res *lockres);
 
 /*
  * Return value from ->downconvert_worker functions.
@@ -102,6 +105,7 @@ static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres,
 static void ocfs2_dentry_post_unlock(struct ocfs2_super *osb,
 				     struct ocfs2_lock_res *lockres);
 
+static void ocfs2_set_qinfo_lvb(struct ocfs2_lock_res *lockres);
 
 #define mlog_meta_lvb(__level, __lockres) ocfs2_dump_meta_lvb_info(__level, __PRETTY_FUNCTION__, __LINE__, __lockres)
 
@@ -111,8 +115,7 @@ static void ocfs2_dump_meta_lvb_info(u64 level,
 				     unsigned int line,
 				     struct ocfs2_lock_res *lockres)
 {
-	struct ocfs2_meta_lvb *lvb =
-		(struct ocfs2_meta_lvb *)ocfs2_dlm_lvb(&lockres->l_lksb);
+	struct ocfs2_meta_lvb *lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
 
 	mlog(level, "LVB information for %s (called from %s:%u):\n",
 	     lockres->l_name, function, line);
@@ -258,6 +261,12 @@ static struct ocfs2_lock_res_ops ocfs2_flock_lops = {
 	.flags		= 0,
 };
 
+static struct ocfs2_lock_res_ops ocfs2_qinfo_lops = {
+	.set_lvb	= ocfs2_set_qinfo_lvb,
+	.get_osb	= ocfs2_get_qinfo_osb,
+	.flags		= LOCK_TYPE_REQUIRES_REFRESH | LOCK_TYPE_USES_LVB,
+};
+
 static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres)
 {
 	return lockres->l_type == OCFS2_LOCK_TYPE_META ||
@@ -279,6 +288,13 @@ static inline struct ocfs2_dentry_lock *ocfs2_lock_res_dl(struct ocfs2_lock_res
 	return (struct ocfs2_dentry_lock *)lockres->l_priv;
 }
 
+static inline struct ocfs2_mem_dqinfo *ocfs2_lock_res_qinfo(struct ocfs2_lock_res *lockres)
+{
+	BUG_ON(lockres->l_type != OCFS2_LOCK_TYPE_QINFO);
+
+	return (struct ocfs2_mem_dqinfo *)lockres->l_priv;
+}
+
 static inline struct ocfs2_super *ocfs2_get_lockres_osb(struct ocfs2_lock_res *lockres)
 {
 	if (lockres->l_ops->get_osb)
@@ -507,6 +523,13 @@ static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres)
 	return OCFS2_SB(inode->i_sb);
 }
 
+static struct ocfs2_super *ocfs2_get_qinfo_osb(struct ocfs2_lock_res *lockres)
+{
+	struct ocfs2_mem_dqinfo *info = lockres->l_priv;
+
+	return OCFS2_SB(info->dqi_gi.dqi_sb);
+}
+
 static struct ocfs2_super *ocfs2_get_file_osb(struct ocfs2_lock_res *lockres)
 {
 	struct ocfs2_file_private *fp = lockres->l_priv;
@@ -609,6 +632,17 @@ void ocfs2_file_lock_res_init(struct ocfs2_lock_res *lockres,
 	lockres->l_flags |= OCFS2_LOCK_NOCACHE;
 }
 
+void ocfs2_qinfo_lock_res_init(struct ocfs2_lock_res *lockres,
+			       struct ocfs2_mem_dqinfo *info)
+{
+	ocfs2_lock_res_init_once(lockres);
+	ocfs2_build_lock_name(OCFS2_LOCK_TYPE_QINFO, info->dqi_gi.dqi_type,
+			      0, lockres->l_name);
+	ocfs2_lock_res_init_common(OCFS2_SB(info->dqi_gi.dqi_sb), lockres,
+				   OCFS2_LOCK_TYPE_QINFO, &ocfs2_qinfo_lops,
+				   info);
+}
+
 void ocfs2_lock_res_free(struct ocfs2_lock_res *res)
 {
 	mlog_entry_void();
@@ -1829,7 +1863,7 @@ static void __ocfs2_stuff_meta_lvb(struct inode *inode)
 
 	mlog_entry_void();
 
-	lvb = (struct ocfs2_meta_lvb *)ocfs2_dlm_lvb(&lockres->l_lksb);
+	lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
 
 	/*
 	 * Invalidate the LVB of a deleted inode - this way other
@@ -1881,7 +1915,7 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode)
 
 	mlog_meta_lvb(0, lockres);
 
-	lvb = (struct ocfs2_meta_lvb *)ocfs2_dlm_lvb(&lockres->l_lksb);
+	lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
 
 	/* We're safe here without the lockres lock... */
 	spin_lock(&oi->ip_lock);
@@ -1916,8 +1950,7 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode)
 static inline int ocfs2_meta_lvb_is_trustable(struct inode *inode,
 					      struct ocfs2_lock_res *lockres)
 {
-	struct ocfs2_meta_lvb *lvb =
-		(struct ocfs2_meta_lvb *)ocfs2_dlm_lvb(&lockres->l_lksb);
+	struct ocfs2_meta_lvb *lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
 
 	if (lvb->lvb_version == OCFS2_LVB_VERSION
 	    && be32_to_cpu(lvb->lvb_igeneration) == inode->i_generation)
@@ -2024,7 +2057,7 @@ static int ocfs2_inode_lock_update(struct inode *inode,
 	} else {
 		/* Boo, we have to go to disk. */
 		/* read bh, cast, ocfs2_refresh_inode */
-		status = ocfs2_read_block(inode, oi->ip_blkno, bh);
+		status = ocfs2_read_inode_block(inode, bh);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail_refresh;
@@ -2032,18 +2065,14 @@ static int ocfs2_inode_lock_update(struct inode *inode,
 		fe = (struct ocfs2_dinode *) (*bh)->b_data;
 
 		/* This is a good chance to make sure we're not
-		 * locking an invalid object.
+		 * locking an invalid object.  ocfs2_read_inode_block()
+		 * already checked that the inode block is sane.
 		 *
 		 * We bug on a stale inode here because we checked
 		 * above whether it was wiped from disk. The wiping
 		 * node provides a guarantee that we receive that
 		 * message and can mark the inode before dropping any
 		 * locks associated with it. */
-		if (!OCFS2_IS_VALID_DINODE(fe)) {
-			OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
-			status = -EIO;
-			goto bail_refresh;
-		}
 		mlog_bug_on_msg(inode->i_generation !=
 				le32_to_cpu(fe->i_generation),
 				"Invalid dinode %llu disk generation: %u "
@@ -2085,7 +2114,7 @@ static int ocfs2_assign_bh(struct inode *inode,
 		return 0;
 	}
 
-	status = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, ret_bh);
+	status = ocfs2_read_inode_block(inode, ret_bh);
 	if (status < 0)
 		mlog_errno(status);
 
@@ -3449,6 +3478,117 @@ static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres,
 	return UNBLOCK_CONTINUE_POST;
 }
 
+static void ocfs2_set_qinfo_lvb(struct ocfs2_lock_res *lockres)
+{
+	struct ocfs2_qinfo_lvb *lvb;
+	struct ocfs2_mem_dqinfo *oinfo = ocfs2_lock_res_qinfo(lockres);
+	struct mem_dqinfo *info = sb_dqinfo(oinfo->dqi_gi.dqi_sb,
+					    oinfo->dqi_gi.dqi_type);
+
+	mlog_entry_void();
+
+	lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
+	lvb->lvb_version = OCFS2_QINFO_LVB_VERSION;
+	lvb->lvb_bgrace = cpu_to_be32(info->dqi_bgrace);
+	lvb->lvb_igrace = cpu_to_be32(info->dqi_igrace);
+	lvb->lvb_syncms = cpu_to_be32(oinfo->dqi_syncms);
+	lvb->lvb_blocks = cpu_to_be32(oinfo->dqi_gi.dqi_blocks);
+	lvb->lvb_free_blk = cpu_to_be32(oinfo->dqi_gi.dqi_free_blk);
+	lvb->lvb_free_entry = cpu_to_be32(oinfo->dqi_gi.dqi_free_entry);
+
+	mlog_exit_void();
+}
+
+void ocfs2_qinfo_unlock(struct ocfs2_mem_dqinfo *oinfo, int ex)
+{
+	struct ocfs2_lock_res *lockres = &oinfo->dqi_gqlock;
+	struct ocfs2_super *osb = OCFS2_SB(oinfo->dqi_gi.dqi_sb);
+	int level = ex ? DLM_LOCK_EX : DLM_LOCK_PR;
+
+	mlog_entry_void();
+	if (!ocfs2_is_hard_readonly(osb) && !ocfs2_mount_local(osb))
+		ocfs2_cluster_unlock(osb, lockres, level);
+	mlog_exit_void();
+}
+
+static int ocfs2_refresh_qinfo(struct ocfs2_mem_dqinfo *oinfo)
+{
+	struct mem_dqinfo *info = sb_dqinfo(oinfo->dqi_gi.dqi_sb,
+					    oinfo->dqi_gi.dqi_type);
+	struct ocfs2_lock_res *lockres = &oinfo->dqi_gqlock;
+	struct ocfs2_qinfo_lvb *lvb = ocfs2_dlm_lvb(&lockres->l_lksb);
+	struct buffer_head *bh = NULL;
+	struct ocfs2_global_disk_dqinfo *gdinfo;
+	int status = 0;
+
+	if (lvb->lvb_version == OCFS2_QINFO_LVB_VERSION) {
+		info->dqi_bgrace = be32_to_cpu(lvb->lvb_bgrace);
+		info->dqi_igrace = be32_to_cpu(lvb->lvb_igrace);
+		oinfo->dqi_syncms = be32_to_cpu(lvb->lvb_syncms);
+		oinfo->dqi_gi.dqi_blocks = be32_to_cpu(lvb->lvb_blocks);
+		oinfo->dqi_gi.dqi_free_blk = be32_to_cpu(lvb->lvb_free_blk);
+		oinfo->dqi_gi.dqi_free_entry =
+					be32_to_cpu(lvb->lvb_free_entry);
+	} else {
+		status = ocfs2_read_quota_block(oinfo->dqi_gqinode, 0, &bh);
+		if (status) {
+			mlog_errno(status);
+			goto bail;
+		}
+		gdinfo = (struct ocfs2_global_disk_dqinfo *)
+					(bh->b_data + OCFS2_GLOBAL_INFO_OFF);
+		info->dqi_bgrace = le32_to_cpu(gdinfo->dqi_bgrace);
+		info->dqi_igrace = le32_to_cpu(gdinfo->dqi_igrace);
+		oinfo->dqi_syncms = le32_to_cpu(gdinfo->dqi_syncms);
+		oinfo->dqi_gi.dqi_blocks = le32_to_cpu(gdinfo->dqi_blocks);
+		oinfo->dqi_gi.dqi_free_blk = le32_to_cpu(gdinfo->dqi_free_blk);
+		oinfo->dqi_gi.dqi_free_entry =
+					le32_to_cpu(gdinfo->dqi_free_entry);
+		brelse(bh);
+		ocfs2_track_lock_refresh(lockres);
+	}
+
+bail:
+	return status;
+}
+
+/* Lock quota info, this function expects at least shared lock on the quota file
+ * so that we can safely refresh quota info from disk. */
+int ocfs2_qinfo_lock(struct ocfs2_mem_dqinfo *oinfo, int ex)
+{
+	struct ocfs2_lock_res *lockres = &oinfo->dqi_gqlock;
+	struct ocfs2_super *osb = OCFS2_SB(oinfo->dqi_gi.dqi_sb);
+	int level = ex ? DLM_LOCK_EX : DLM_LOCK_PR;
+	int status = 0;
+
+	mlog_entry_void();
+
+	/* On RO devices, locking really isn't needed... */
+	if (ocfs2_is_hard_readonly(osb)) {
+		if (ex)
+			status = -EROFS;
+		goto bail;
+	}
+	if (ocfs2_mount_local(osb))
+		goto bail;
+
+	status = ocfs2_cluster_lock(osb, lockres, level, 0, 0);
+	if (status < 0) {
+		mlog_errno(status);
+		goto bail;
+	}
+	if (!ocfs2_should_refresh_lock_res(lockres))
+		goto bail;
+	/* OK, we have the lock but we need to refresh the quota info */
+	status = ocfs2_refresh_qinfo(oinfo);
+	if (status)
+		ocfs2_qinfo_unlock(oinfo, ex);
+	ocfs2_complete_lock_res_refresh(lockres, status);
+bail:
+	mlog_exit(status);
+	return status;
+}
+
 /*
  * This is the filesystem locking protocol.  It provides the lock handling
  * hooks for the underlying DLM.  It has a maximum version number.

fs/ocfs2/dlmglue.h

@@ -49,6 +49,19 @@ struct ocfs2_meta_lvb {
 	__be32       lvb_reserved2;
 };
 
+#define OCFS2_QINFO_LVB_VERSION 1
+
+struct ocfs2_qinfo_lvb {
+	__u8	lvb_version;
+	__u8	lvb_reserved[3];
+	__be32	lvb_bgrace;
+	__be32	lvb_igrace;
+	__be32	lvb_syncms;
+	__be32	lvb_blocks;
+	__be32	lvb_free_blk;
+	__be32	lvb_free_entry;
+};
+
 /* ocfs2_inode_lock_full() 'arg_flags' flags */
 /* don't wait on recovery. */
 #define OCFS2_META_LOCK_RECOVERY	(0x01)
@@ -69,6 +82,9 @@ void ocfs2_dentry_lock_res_init(struct ocfs2_dentry_lock *dl,
 struct ocfs2_file_private;
 void ocfs2_file_lock_res_init(struct ocfs2_lock_res *lockres,
 			      struct ocfs2_file_private *fp);
+struct ocfs2_mem_dqinfo;
+void ocfs2_qinfo_lock_res_init(struct ocfs2_lock_res *lockres,
+                               struct ocfs2_mem_dqinfo *info);
 void ocfs2_lock_res_free(struct ocfs2_lock_res *res);
 int ocfs2_create_new_inode_locks(struct inode *inode);
 int ocfs2_drop_inode_locks(struct inode *inode);
@@ -103,6 +119,9 @@ int ocfs2_dentry_lock(struct dentry *dentry, int ex);
 void ocfs2_dentry_unlock(struct dentry *dentry, int ex);
 int ocfs2_file_lock(struct file *file, int ex, int trylock);
 void ocfs2_file_unlock(struct file *file);
+int ocfs2_qinfo_lock(struct ocfs2_mem_dqinfo *oinfo, int ex);
+void ocfs2_qinfo_unlock(struct ocfs2_mem_dqinfo *oinfo, int ex);
+
 
 void ocfs2_mark_lockres_freeing(struct ocfs2_lock_res *lockres);
 void ocfs2_simple_drop_lockres(struct ocfs2_super *osb,

fs/ocfs2/extent_map.c

@@ -293,7 +293,7 @@ static int ocfs2_last_eb_is_empty(struct inode *inode,
 	struct ocfs2_extent_block *eb;
 	struct ocfs2_extent_list *el;
 
-	ret = ocfs2_read_block(inode, last_eb_blk, &eb_bh);
+	ret = ocfs2_read_extent_block(inode, last_eb_blk, &eb_bh);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -302,12 +302,6 @@ static int ocfs2_last_eb_is_empty(struct inode *inode,
 	eb = (struct ocfs2_extent_block *) eb_bh->b_data;
 	el = &eb->h_list;
 
-	if (!OCFS2_IS_VALID_EXTENT_BLOCK(eb)) {
-		ret = -EROFS;
-		OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, eb);
-		goto out;
-	}
-
 	if (el->l_tree_depth) {
 		ocfs2_error(inode->i_sb,
 			    "Inode %lu has non zero tree depth in "
@@ -381,23 +375,16 @@ static int ocfs2_figure_hole_clusters(struct inode *inode,
 		if (le64_to_cpu(eb->h_next_leaf_blk) == 0ULL)
 			goto no_more_extents;
 
-		ret = ocfs2_read_block(inode,
-				       le64_to_cpu(eb->h_next_leaf_blk),
-				       &next_eb_bh);
+		ret = ocfs2_read_extent_block(inode,
+					      le64_to_cpu(eb->h_next_leaf_blk),
+					      &next_eb_bh);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
-		next_eb = (struct ocfs2_extent_block *)next_eb_bh->b_data;
-
-		if (!OCFS2_IS_VALID_EXTENT_BLOCK(next_eb)) {
-			ret = -EROFS;
-			OCFS2_RO_ON_INVALID_EXTENT_BLOCK(inode->i_sb, next_eb);
-			goto out;
-		}
 
+		next_eb = (struct ocfs2_extent_block *)next_eb_bh->b_data;
 		el = &next_eb->h_list;
-
 		i = ocfs2_search_for_hole_index(el, v_cluster);
 	}
 
@@ -630,7 +617,7 @@ int ocfs2_get_clusters(struct inode *inode, u32 v_cluster,
 	if (ret == 0)
 		goto out;
 
-	ret = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, &di_bh);
+	ret = ocfs2_read_inode_block(inode, &di_bh);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -819,3 +806,74 @@ out:
 
 	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,
+					   struct buffer_head *bh))
+{
+	int rc = 0;
+	u64 p_block, p_count;
+	int i, count, done = 0;
+
+	mlog_entry("(inode = %p, v_block = %llu, nr = %d, bhs = %p, "
+		   "flags = %x, validate = %p)\n",
+		   inode, (unsigned long long)v_block, nr, bhs, flags,
+		   validate);
+
+	if (((v_block + nr - 1) << inode->i_sb->s_blocksize_bits) >=
+	    i_size_read(inode)) {
+		BUG_ON(!(flags & OCFS2_BH_READAHEAD));
+		goto out;
+	}
+
+	while (done < nr) {
+		down_read(&OCFS2_I(inode)->ip_alloc_sem);
+		rc = ocfs2_extent_map_get_blocks(inode, v_block + done,
+						 &p_block, &p_count, NULL);
+		up_read(&OCFS2_I(inode)->ip_alloc_sem);
+		if (rc) {
+			mlog_errno(rc);
+			break;
+		}
+
+		if (!p_block) {
+			rc = -EIO;
+			mlog(ML_ERROR,
+			     "Inode #%llu contains a hole at offset %llu\n",
+			     (unsigned long long)OCFS2_I(inode)->ip_blkno,
+			     (unsigned long long)(v_block + done) <<
+			     inode->i_sb->s_blocksize_bits);
+			break;
+		}
+
+		count = nr - done;
+		if (p_count < count)
+			count = p_count;
+
+		/*
+		 * If the caller passed us bhs, they should have come
+		 * from a previous readahead call to this function.  Thus,
+		 * they should have the right b_blocknr.
+		 */
+		for (i = 0; i < count; i++) {
+			if (!bhs[done + i])
+				continue;
+			BUG_ON(bhs[done + i]->b_blocknr != (p_block + i));
+		}
+
+		rc = ocfs2_read_blocks(inode, p_block, count, bhs + done,
+				       flags, validate);
+		if (rc) {
+			mlog_errno(rc);
+			break;
+		}
+		done += count;
+	}
+
+out:
+	mlog_exit(rc);
+	return rc;
+}
+
+

fs/ocfs2/extent_map.h

@@ -57,4 +57,28 @@ int ocfs2_xattr_get_clusters(struct inode *inode, u32 v_cluster,
 			     u32 *p_cluster, u32 *num_clusters,
 			     struct ocfs2_extent_list *el);
 
+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,
+					   struct buffer_head *bh));
+static inline int ocfs2_read_virt_block(struct inode *inode, u64 v_block,
+					struct buffer_head **bh,
+					int (*validate)(struct super_block *sb,
+							struct buffer_head *bh))
+{
+	int status = 0;
+
+	if (bh == NULL) {
+		printk("ocfs2: bh == NULL\n");
+		status = -EINVAL;
+		goto bail;
+	}
+
+	status = ocfs2_read_virt_blocks(inode, v_block, 1, bh, 0, validate);
+
+bail:
+	return status;
+}
+
+
 #endif  /* _EXTENT_MAP_H */

fs/ocfs2/file.c

@@ -35,6 +35,7 @@
 #include <linux/mount.h>
 #include <linux/writeback.h>
 #include <linux/falloc.h>
+#include <linux/quotaops.h>
 
 #define MLOG_MASK_PREFIX ML_INODE
 #include <cluster/masklog.h>
@@ -56,6 +57,8 @@
 #include "suballoc.h"
 #include "super.h"
 #include "xattr.h"
+#include "acl.h"
+#include "quota.h"
 
 #include "buffer_head_io.h"
 
@@ -253,8 +256,8 @@ int ocfs2_update_inode_atime(struct inode *inode,
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -303,9 +306,9 @@ bail:
 	return status;
 }
 
-static int ocfs2_simple_size_update(struct inode *inode,
-				    struct buffer_head *di_bh,
-				    u64 new_i_size)
+int ocfs2_simple_size_update(struct inode *inode,
+			     struct buffer_head *di_bh,
+			     u64 new_i_size)
 {
 	int ret;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
@@ -350,8 +353,8 @@ static int ocfs2_orphan_for_truncate(struct ocfs2_super *osb,
 		goto out;
 	}
 
-	status = ocfs2_journal_access(handle, inode, fe_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, inode, fe_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto out_commit;
@@ -401,12 +404,9 @@ static int ocfs2_truncate_file(struct inode *inode,
 		   (unsigned long long)OCFS2_I(inode)->ip_blkno,
 		   (unsigned long long)new_i_size);
 
+	/* We trust di_bh because it comes from ocfs2_inode_lock(), which
+	 * already validated it */
 	fe = (struct ocfs2_dinode *) di_bh->b_data;
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
-		status = -EIO;
-		goto bail;
-	}
 
 	mlog_bug_on_msg(le64_to_cpu(fe->i_size) != i_size_read(inode),
 			"Inode %llu, inode i_size = %lld != di "
@@ -536,6 +536,7 @@ static int __ocfs2_extend_allocation(struct inode *inode, u32 logical_start,
 	enum ocfs2_alloc_restarted why;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct ocfs2_extent_tree et;
+	int did_quota = 0;
 
 	mlog_entry("(clusters_to_add = %u)\n", clusters_to_add);
 
@@ -545,18 +546,12 @@ static int __ocfs2_extend_allocation(struct inode *inode, u32 logical_start,
 	 */
 	BUG_ON(mark_unwritten && !ocfs2_sparse_alloc(osb));
 
-	status = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, &bh);
+	status = ocfs2_read_inode_block(inode, &bh);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
 	}
-
 	fe = (struct ocfs2_dinode *) bh->b_data;
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
-		status = -EIO;
-		goto leave;
-	}
 
 restart_all:
 	BUG_ON(le32_to_cpu(fe->i_clusters) != OCFS2_I(inode)->ip_clusters);
@@ -585,11 +580,18 @@ restart_all:
 	}
 
 restarted_transaction:
+	if (vfs_dq_alloc_space_nodirty(inode, ocfs2_clusters_to_bytes(osb->sb,
+	    clusters_to_add))) {
+		status = -EDQUOT;
+		goto leave;
+	}
+	did_quota = 1;
+
 	/* reserve a write to the file entry early on - that we if we
 	 * run out of credits in the allocation path, we can still
 	 * update i_size. */
-	status = ocfs2_journal_access(handle, inode, bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, inode, bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -622,6 +624,10 @@ restarted_transaction:
 	spin_lock(&OCFS2_I(inode)->ip_lock);
 	clusters_to_add -= (OCFS2_I(inode)->ip_clusters - prev_clusters);
 	spin_unlock(&OCFS2_I(inode)->ip_lock);
+	/* Release unused quota reservation */
+	vfs_dq_free_space(inode,
+			ocfs2_clusters_to_bytes(osb->sb, clusters_to_add));
+	did_quota = 0;
 
 	if (why != RESTART_NONE && clusters_to_add) {
 		if (why == RESTART_META) {
@@ -654,6 +660,9 @@ restarted_transaction:
 	     OCFS2_I(inode)->ip_clusters, (long long)i_size_read(inode));
 
 leave:
+	if (status < 0 && did_quota)
+		vfs_dq_free_space(inode,
+			ocfs2_clusters_to_bytes(osb->sb, clusters_to_add));
 	if (handle) {
 		ocfs2_commit_trans(osb, handle);
 		handle = NULL;
@@ -885,6 +894,9 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 	struct ocfs2_super *osb = OCFS2_SB(sb);
 	struct buffer_head *bh = NULL;
 	handle_t *handle = NULL;
+	int locked[MAXQUOTAS] = {0, 0};
+	int credits, qtype;
+	struct ocfs2_mem_dqinfo *oinfo;
 
 	mlog_entry("(0x%p, '%.*s')\n", dentry,
 	           dentry->d_name.len, dentry->d_name.name);
@@ -955,11 +967,47 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
-	if (IS_ERR(handle)) {
-		status = PTR_ERR(handle);
-		mlog_errno(status);
-		goto bail_unlock;
+	if ((attr->ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
+	    (attr->ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) {
+		credits = OCFS2_INODE_UPDATE_CREDITS;
+		if (attr->ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid
+		    && OCFS2_HAS_RO_COMPAT_FEATURE(sb,
+		    OCFS2_FEATURE_RO_COMPAT_USRQUOTA)) {
+			oinfo = sb_dqinfo(sb, USRQUOTA)->dqi_priv;
+			status = ocfs2_lock_global_qf(oinfo, 1);
+			if (status < 0)
+				goto bail_unlock;
+			credits += ocfs2_calc_qinit_credits(sb, USRQUOTA) +
+				ocfs2_calc_qdel_credits(sb, USRQUOTA);
+			locked[USRQUOTA] = 1;
+		}
+		if (attr->ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid
+		    && OCFS2_HAS_RO_COMPAT_FEATURE(sb,
+		    OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)) {
+			oinfo = sb_dqinfo(sb, GRPQUOTA)->dqi_priv;
+			status = ocfs2_lock_global_qf(oinfo, 1);
+			if (status < 0)
+				goto bail_unlock;
+			credits += ocfs2_calc_qinit_credits(sb, GRPQUOTA) +
+				   ocfs2_calc_qdel_credits(sb, GRPQUOTA);
+			locked[GRPQUOTA] = 1;
+		}
+		handle = ocfs2_start_trans(osb, credits);
+		if (IS_ERR(handle)) {
+			status = PTR_ERR(handle);
+			mlog_errno(status);
+			goto bail_unlock;
+		}
+		status = vfs_dq_transfer(inode, attr) ? -EDQUOT : 0;
+		if (status < 0)
+			goto bail_commit;
+	} else {
+		handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
+		if (IS_ERR(handle)) {
+			status = PTR_ERR(handle);
+			mlog_errno(status);
+			goto bail_unlock;
+		}
 	}
 
 	/*
@@ -982,6 +1030,12 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 bail_commit:
 	ocfs2_commit_trans(osb, handle);
 bail_unlock:
+	for (qtype = 0; qtype < MAXQUOTAS; qtype++) {
+		if (!locked[qtype])
+			continue;
+		oinfo = sb_dqinfo(sb, qtype)->dqi_priv;
+		ocfs2_unlock_global_qf(oinfo, 1);
+	}
 	ocfs2_inode_unlock(inode, 1);
 bail_unlock_rw:
 	if (size_change)
@@ -989,6 +1043,12 @@ bail_unlock_rw:
 bail:
 	brelse(bh);
 
+	if (!status && attr->ia_valid & ATTR_MODE) {
+		status = ocfs2_acl_chmod(inode);
+		if (status < 0)
+			mlog_errno(status);
+	}
+
 	mlog_exit(status);
 	return status;
 }
@@ -1035,7 +1095,7 @@ int ocfs2_permission(struct inode *inode, int mask)
 		goto out;
 	}
 
-	ret = generic_permission(inode, mask, NULL);
+	ret = generic_permission(inode, mask, ocfs2_check_acl);
 
 	ocfs2_inode_unlock(inode, 0);
 out:
@@ -1061,8 +1121,8 @@ static int __ocfs2_write_remove_suid(struct inode *inode,
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out_trans;
@@ -1128,9 +1188,8 @@ static int ocfs2_write_remove_suid(struct inode *inode)
 {
 	int ret;
 	struct buffer_head *bh = NULL;
-	struct ocfs2_inode_info *oi = OCFS2_I(inode);
 
-	ret = ocfs2_read_block(inode, oi->ip_blkno, &bh);
+	ret = ocfs2_read_inode_block(inode, &bh);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
@@ -1156,8 +1215,7 @@ static int ocfs2_allocate_unwritten_extents(struct inode *inode,
 	struct buffer_head *di_bh = NULL;
 
 	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
-		ret = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno,
-				       &di_bh);
+		ret = ocfs2_read_inode_block(inode, &di_bh);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -1226,83 +1284,6 @@ out:
 	return ret;
 }
 
-static int __ocfs2_remove_inode_range(struct inode *inode,
-				      struct buffer_head *di_bh,
-				      u32 cpos, u32 phys_cpos, u32 len,
-				      struct ocfs2_cached_dealloc_ctxt *dealloc)
-{
-	int ret;
-	u64 phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos);
-	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	struct inode *tl_inode = osb->osb_tl_inode;
-	handle_t *handle;
-	struct ocfs2_alloc_context *meta_ac = NULL;
-	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
-	struct ocfs2_extent_tree et;
-
-	ocfs2_init_dinode_extent_tree(&et, inode, di_bh);
-
-	ret = ocfs2_lock_allocators(inode, &et, 0, 1, NULL, &meta_ac);
-	if (ret) {
-		mlog_errno(ret);
-		return ret;
-	}
-
-	mutex_lock(&tl_inode->i_mutex);
-
-	if (ocfs2_truncate_log_needs_flush(osb)) {
-		ret = __ocfs2_flush_truncate_log(osb);
-		if (ret < 0) {
-			mlog_errno(ret);
-			goto out;
-		}
-	}
-
-	handle = ocfs2_start_trans(osb, OCFS2_REMOVE_EXTENT_CREDITS);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		mlog_errno(ret);
-		goto out;
-	}
-
-	ret = ocfs2_journal_access(handle, inode, di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
-
-	ret = ocfs2_remove_extent(inode, &et, cpos, len, handle, meta_ac,
-				  dealloc);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
-
-	OCFS2_I(inode)->ip_clusters -= len;
-	di->i_clusters = cpu_to_le32(OCFS2_I(inode)->ip_clusters);
-
-	ret = ocfs2_journal_dirty(handle, di_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
-
-	ret = ocfs2_truncate_log_append(osb, handle, phys_blkno, len);
-	if (ret)
-		mlog_errno(ret);
-
-out_commit:
-	ocfs2_commit_trans(osb, handle);
-out:
-	mutex_unlock(&tl_inode->i_mutex);
-
-	if (meta_ac)
-		ocfs2_free_alloc_context(meta_ac);
-
-	return ret;
-}
-
 /*
  * Truncate a byte range, avoiding pages within partial clusters. This
  * preserves those pages for the zeroing code to write to.
@@ -1402,7 +1383,9 @@ static int ocfs2_remove_inode_range(struct inode *inode,
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct ocfs2_cached_dealloc_ctxt dealloc;
 	struct address_space *mapping = inode->i_mapping;
+	struct ocfs2_extent_tree et;
 
+	ocfs2_init_dinode_extent_tree(&et, inode, di_bh);
 	ocfs2_init_dealloc_ctxt(&dealloc);
 
 	if (byte_len == 0)
@@ -1458,9 +1441,9 @@ static int ocfs2_remove_inode_range(struct inode *inode,
 
 		/* Only do work for non-holes */
 		if (phys_cpos != 0) {
-			ret = __ocfs2_remove_inode_range(inode, di_bh, cpos,
-							 phys_cpos, alloc_size,
-							 &dealloc);
+			ret = ocfs2_remove_btree_range(inode, &et, cpos,
+						       phys_cpos, alloc_size,
+						       &dealloc);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;

fs/ocfs2/file.h

@@ -51,6 +51,9 @@ int ocfs2_add_inode_data(struct ocfs2_super *osb,
 			 struct ocfs2_alloc_context *data_ac,
 			 struct ocfs2_alloc_context *meta_ac,
 			 enum ocfs2_alloc_restarted *reason_ret);
+int ocfs2_simple_size_update(struct inode *inode,
+			     struct buffer_head *di_bh,
+			     u64 new_i_size);
 int ocfs2_extend_no_holes(struct inode *inode, u64 new_i_size,
 			  u64 zero_to);
 int ocfs2_setattr(struct dentry *dentry, struct iattr *attr);

fs/ocfs2/inode.c

@@ -28,6 +28,7 @@
 #include <linux/slab.h>
 #include <linux/highmem.h>
 #include <linux/pagemap.h>
+#include <linux/quotaops.h>
 
 #include <asm/byteorder.h>
 
@@ -37,6 +38,7 @@
 #include "ocfs2.h"
 
 #include "alloc.h"
+#include "blockcheck.h"
 #include "dlmglue.h"
 #include "extent_map.h"
 #include "file.h"
@@ -214,12 +216,11 @@ static int ocfs2_init_locked_inode(struct inode *inode, void *opaque)
 	return 0;
 }
 
-int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
-		     	 int create_ino)
+void ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
+			  int create_ino)
 {
 	struct super_block *sb;
 	struct ocfs2_super *osb;
-	int status = -EINVAL;
 	int use_plocks = 1;
 
 	mlog_entry("(0x%p, size:%llu)\n", inode,
@@ -232,25 +233,17 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
 	    ocfs2_mount_local(osb) || !ocfs2_stack_supports_plocks())
 		use_plocks = 0;
 
-	/* this means that read_inode cannot create a superblock inode
-	 * today.  change if needed. */
-	if (!OCFS2_IS_VALID_DINODE(fe) ||
-	    !(fe->i_flags & cpu_to_le32(OCFS2_VALID_FL))) {
-		mlog(0, "Invalid dinode: i_ino=%lu, i_blkno=%llu, "
-		     "signature = %.*s, flags = 0x%x\n",
-		     inode->i_ino,
-		     (unsigned long long)le64_to_cpu(fe->i_blkno), 7,
-		     fe->i_signature, le32_to_cpu(fe->i_flags));
-		goto bail;
-	}
+	/*
+	 * These have all been checked by ocfs2_read_inode_block() or set
+	 * by ocfs2_mknod_locked(), so a failure is a code bug.
+	 */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));  /* This means that read_inode
+						cannot create a superblock
+						inode today.  change if
+						that is needed. */
+	BUG_ON(!(fe->i_flags & cpu_to_le32(OCFS2_VALID_FL)));
+	BUG_ON(le32_to_cpu(fe->i_fs_generation) != osb->fs_generation);
 
-	if (le32_to_cpu(fe->i_fs_generation) != osb->fs_generation) {
-		mlog(ML_ERROR, "file entry generation does not match "
-		     "superblock! osb->fs_generation=%x, "
-		     "fe->i_fs_generation=%x\n",
-		     osb->fs_generation, le32_to_cpu(fe->i_fs_generation));
-		goto bail;
-	}
 
 	OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
 	OCFS2_I(inode)->ip_attr = le32_to_cpu(fe->i_attr);
@@ -284,14 +277,18 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
 
 	inode->i_nlink = le16_to_cpu(fe->i_links_count);
 
-	if (fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL))
+	if (fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)) {
 		OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SYSTEM_FILE;
+		inode->i_flags |= S_NOQUOTA;
+	}
 
 	if (fe->i_flags & cpu_to_le32(OCFS2_LOCAL_ALLOC_FL)) {
 		OCFS2_I(inode)->ip_flags |= OCFS2_INODE_BITMAP;
 		mlog(0, "local alloc inode: i_ino=%lu\n", inode->i_ino);
 	} else if (fe->i_flags & cpu_to_le32(OCFS2_BITMAP_FL)) {
 		OCFS2_I(inode)->ip_flags |= OCFS2_INODE_BITMAP;
+	} else if (fe->i_flags & cpu_to_le32(OCFS2_QUOTA_FL)) {
+		inode->i_flags |= S_NOQUOTA;
 	} else if (fe->i_flags & cpu_to_le32(OCFS2_SUPER_BLOCK_FL)) {
 		mlog(0, "superblock inode: i_ino=%lu\n", inode->i_ino);
 		/* we can't actually hit this as read_inode can't
@@ -354,10 +351,7 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
 
 	ocfs2_set_inode_flags(inode);
 
-	status = 0;
-bail:
-	mlog_exit(status);
-	return status;
+	mlog_exit_void();
 }
 
 static int ocfs2_read_locked_inode(struct inode *inode,
@@ -460,11 +454,14 @@ static int ocfs2_read_locked_inode(struct inode *inode,
 		}
 	}
 
-	if (can_lock)
-		status = ocfs2_read_blocks(inode, args->fi_blkno, 1, &bh,
-					   OCFS2_BH_IGNORE_CACHE);
-	else
+	if (can_lock) {
+		status = ocfs2_read_inode_block_full(inode, &bh,
+						     OCFS2_BH_IGNORE_CACHE);
+	} else {
 		status = ocfs2_read_blocks_sync(osb, args->fi_blkno, 1, &bh);
+		if (!status)
+			status = ocfs2_validate_inode_block(osb->sb, bh);
+	}
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -472,12 +469,6 @@ static int ocfs2_read_locked_inode(struct inode *inode,
 
 	status = -EINVAL;
 	fe = (struct ocfs2_dinode *) bh->b_data;
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		mlog(0, "Invalid dinode #%llu: signature = %.*s\n",
-		     (unsigned long long)args->fi_blkno, 7,
-		     fe->i_signature);
-		goto bail;
-	}
 
 	/*
 	 * This is a code bug. Right now the caller needs to
@@ -491,10 +482,9 @@ static int ocfs2_read_locked_inode(struct inode *inode,
 
 	if (S_ISCHR(le16_to_cpu(fe->i_mode)) ||
 	    S_ISBLK(le16_to_cpu(fe->i_mode)))
-    		inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev));
+		inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev));
 
-	if (ocfs2_populate_inode(inode, fe, 0) < 0)
-		goto bail;
+	ocfs2_populate_inode(inode, fe, 0);
 
 	BUG_ON(args->fi_blkno != le64_to_cpu(fe->i_blkno));
 
@@ -547,8 +537,8 @@ static int ocfs2_truncate_for_delete(struct ocfs2_super *osb,
 			goto out;
 		}
 
-		status = ocfs2_journal_access(handle, inode, fe_bh,
-					      OCFS2_JOURNAL_ACCESS_WRITE);
+		status = ocfs2_journal_access_di(handle, inode, fe_bh,
+						 OCFS2_JOURNAL_ACCESS_WRITE);
 		if (status < 0) {
 			mlog_errno(status);
 			goto out;
@@ -615,7 +605,8 @@ static int ocfs2_remove_inode(struct inode *inode,
 		goto bail;
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_DELETE_INODE_CREDITS);
+	handle = ocfs2_start_trans(osb, OCFS2_DELETE_INODE_CREDITS +
+					ocfs2_quota_trans_credits(inode->i_sb));
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		mlog_errno(status);
@@ -630,8 +621,8 @@ static int ocfs2_remove_inode(struct inode *inode,
 	}
 
 	/* set the inodes dtime */
-	status = ocfs2_journal_access(handle, inode, di_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, inode, di_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail_commit;
@@ -647,6 +638,7 @@ static int ocfs2_remove_inode(struct inode *inode,
 	}
 
 	ocfs2_remove_from_cache(inode, di_bh);
+	vfs_dq_free_inode(inode);
 
 	status = ocfs2_free_dinode(handle, inode_alloc_inode,
 				   inode_alloc_bh, di);
@@ -929,7 +921,10 @@ void ocfs2_delete_inode(struct inode *inode)
 
 	mlog_entry("(inode->i_ino = %lu)\n", inode->i_ino);
 
-	if (is_bad_inode(inode)) {
+	/* When we fail in read_inode() we mark inode as bad. The second test
+	 * catches the case when inode allocation fails before allocating
+	 * a block for inode. */
+	if (is_bad_inode(inode) || !OCFS2_I(inode)->ip_blkno) {
 		mlog(0, "Skipping delete of bad inode\n");
 		goto bail;
 	}
@@ -1195,8 +1190,8 @@ int ocfs2_mark_inode_dirty(handle_t *handle,
 	mlog_entry("(inode %llu)\n",
 		   (unsigned long long)OCFS2_I(inode)->ip_blkno);
 
-	status = ocfs2_journal_access(handle, inode, bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, inode, bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -1264,3 +1259,89 @@ void ocfs2_refresh_inode(struct inode *inode,
 
 	spin_unlock(&OCFS2_I(inode)->ip_lock);
 }
+
+int ocfs2_validate_inode_block(struct super_block *sb,
+			       struct buffer_head *bh)
+{
+	int rc;
+	struct ocfs2_dinode *di = (struct ocfs2_dinode *)bh->b_data;
+
+	mlog(0, "Validating dinode %llu\n",
+	     (unsigned long long)bh->b_blocknr);
+
+	BUG_ON(!buffer_uptodate(bh));
+
+	/*
+	 * If the ecc fails, we return the error but otherwise
+	 * leave the filesystem running.  We know any error is
+	 * local to this block.
+	 */
+	rc = ocfs2_validate_meta_ecc(sb, bh->b_data, &di->i_check);
+	if (rc) {
+		mlog(ML_ERROR, "Checksum failed for dinode %llu\n",
+		     (unsigned long long)bh->b_blocknr);
+		goto bail;
+	}
+
+	/*
+	 * Errors after here are fatal.
+	 */
+
+	rc = -EINVAL;
+
+	if (!OCFS2_IS_VALID_DINODE(di)) {
+		ocfs2_error(sb, "Invalid dinode #%llu: signature = %.*s\n",
+			    (unsigned long long)bh->b_blocknr, 7,
+			    di->i_signature);
+		goto bail;
+	}
+
+	if (le64_to_cpu(di->i_blkno) != bh->b_blocknr) {
+		ocfs2_error(sb, "Invalid dinode #%llu: i_blkno is %llu\n",
+			    (unsigned long long)bh->b_blocknr,
+			    (unsigned long long)le64_to_cpu(di->i_blkno));
+		goto bail;
+	}
+
+	if (!(di->i_flags & cpu_to_le32(OCFS2_VALID_FL))) {
+		ocfs2_error(sb,
+			    "Invalid dinode #%llu: OCFS2_VALID_FL not set\n",
+			    (unsigned long long)bh->b_blocknr);
+		goto bail;
+	}
+
+	if (le32_to_cpu(di->i_fs_generation) !=
+	    OCFS2_SB(sb)->fs_generation) {
+		ocfs2_error(sb,
+			    "Invalid dinode #%llu: fs_generation is %u\n",
+			    (unsigned long long)bh->b_blocknr,
+			    le32_to_cpu(di->i_fs_generation));
+		goto bail;
+	}
+
+	rc = 0;
+
+bail:
+	return rc;
+}
+
+int ocfs2_read_inode_block_full(struct inode *inode, struct buffer_head **bh,
+				int flags)
+{
+	int rc;
+	struct buffer_head *tmp = *bh;
+
+	rc = ocfs2_read_blocks(inode, OCFS2_I(inode)->ip_blkno, 1, &tmp,
+			       flags, ocfs2_validate_inode_block);
+
+	/* If ocfs2_read_blocks() got us a new bh, pass it up. */
+	if (!rc && !*bh)
+		*bh = tmp;
+
+	return rc;
+}
+
+int ocfs2_read_inode_block(struct inode *inode, struct buffer_head **bh)
+{
+	return ocfs2_read_inode_block_full(inode, bh, 0);
+}

fs/ocfs2/inode.h

@@ -128,8 +128,8 @@ struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 feoff, unsigned flags,
 			 int sysfile_type);
 int ocfs2_inode_init_private(struct inode *inode);
 int ocfs2_inode_revalidate(struct dentry *dentry);
-int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
-			 int create_ino);
+void ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
+			  int create_ino);
 void ocfs2_read_inode(struct inode *inode);
 void ocfs2_read_inode2(struct inode *inode, void *opaque);
 ssize_t ocfs2_rw_direct(int rw, struct file *filp, char *buf,
@@ -142,6 +142,8 @@ int ocfs2_mark_inode_dirty(handle_t *handle,
 			   struct buffer_head *bh);
 int ocfs2_aio_read(struct file *file, struct kiocb *req, struct iocb *iocb);
 int ocfs2_aio_write(struct file *file, struct kiocb *req, struct iocb *iocb);
+struct buffer_head *ocfs2_bread(struct inode *inode,
+				int block, int *err, int reada);
 
 void ocfs2_set_inode_flags(struct inode *inode);
 void ocfs2_get_inode_flags(struct ocfs2_inode_info *oi);
@@ -153,4 +155,16 @@ static inline blkcnt_t ocfs2_inode_sector_count(struct inode *inode)
 	return (blkcnt_t)(OCFS2_I(inode)->ip_clusters << c_to_s_bits);
 }
 
+/* Validate that a bh contains a valid inode */
+int ocfs2_validate_inode_block(struct super_block *sb,
+			       struct buffer_head *bh);
+/*
+ * Read an inode block into *bh.  If *bh is NULL, a bh will be allocated.
+ * This is a cached read.  The inode will be validated with
+ * ocfs2_validate_inode_block().
+ */
+int ocfs2_read_inode_block(struct inode *inode, struct buffer_head **bh);
+/* The same, but can be passed OCFS2_BH_* flags */
+int ocfs2_read_inode_block_full(struct inode *inode, struct buffer_head **bh,
+				int flags);
 #endif /* OCFS2_INODE_H */

fs/ocfs2/journal.c

@@ -35,6 +35,7 @@
 #include "ocfs2.h"
 
 #include "alloc.h"
+#include "blockcheck.h"
 #include "dir.h"
 #include "dlmglue.h"
 #include "extent_map.h"
@@ -45,6 +46,7 @@
 #include "slot_map.h"
 #include "super.h"
 #include "sysfile.h"
+#include "quota.h"
 
 #include "buffer_head_io.h"
 
@@ -52,10 +54,10 @@ DEFINE_SPINLOCK(trans_inc_lock);
 
 static int ocfs2_force_read_journal(struct inode *inode);
 static int ocfs2_recover_node(struct ocfs2_super *osb,
-			      int node_num);
+			      int node_num, int slot_num);
 static int __ocfs2_recovery_thread(void *arg);
 static int ocfs2_commit_cache(struct ocfs2_super *osb);
-static int ocfs2_wait_on_mount(struct ocfs2_super *osb);
+static int __ocfs2_wait_on_mount(struct ocfs2_super *osb, int quota);
 static int ocfs2_journal_toggle_dirty(struct ocfs2_super *osb,
 				      int dirty, int replayed);
 static int ocfs2_trylock_journal(struct ocfs2_super *osb,
@@ -64,6 +66,17 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
 				 int slot);
 static int ocfs2_commit_thread(void *arg);
 
+static inline int ocfs2_wait_on_mount(struct ocfs2_super *osb)
+{
+	return __ocfs2_wait_on_mount(osb, 0);
+}
+
+static inline int ocfs2_wait_on_quotas(struct ocfs2_super *osb)
+{
+	return __ocfs2_wait_on_mount(osb, 1);
+}
+
+
 
 /*
  * The recovery_list is a simple linked list of node numbers to recover.
@@ -256,11 +269,9 @@ handle_t *ocfs2_start_trans(struct ocfs2_super *osb, int max_buffs)
 	BUG_ON(osb->journal->j_state == OCFS2_JOURNAL_FREE);
 	BUG_ON(max_buffs <= 0);
 
-	/* JBD might support this, but our journalling code doesn't yet. */
-	if (journal_current_handle()) {
-		mlog(ML_ERROR, "Recursive transaction attempted!\n");
-		BUG();
-	}
+	/* Nested transaction? Just return the handle... */
+	if (journal_current_handle())
+		return jbd2_journal_start(journal, max_buffs);
 
 	down_read(&osb->journal->j_trans_barrier);
 
@@ -285,16 +296,18 @@ handle_t *ocfs2_start_trans(struct ocfs2_super *osb, int max_buffs)
 int ocfs2_commit_trans(struct ocfs2_super *osb,
 		       handle_t *handle)
 {
-	int ret;
+	int ret, nested;
 	struct ocfs2_journal *journal = osb->journal;
 
 	BUG_ON(!handle);
 
+	nested = handle->h_ref > 1;
 	ret = jbd2_journal_stop(handle);
 	if (ret < 0)
 		mlog_errno(ret);
 
-	up_read(&journal->j_trans_barrier);
+	if (!nested)
+		up_read(&journal->j_trans_barrier);
 
 	return ret;
 }
@@ -357,10 +370,137 @@ bail:
 	return status;
 }
 
-int ocfs2_journal_access(handle_t *handle,
-			 struct inode *inode,
-			 struct buffer_head *bh,
-			 int type)
+struct ocfs2_triggers {
+	struct jbd2_buffer_trigger_type	ot_triggers;
+	int				ot_offset;
+};
+
+static inline struct ocfs2_triggers *to_ocfs2_trigger(struct jbd2_buffer_trigger_type *triggers)
+{
+	return container_of(triggers, struct ocfs2_triggers, ot_triggers);
+}
+
+static void ocfs2_commit_trigger(struct jbd2_buffer_trigger_type *triggers,
+				 struct buffer_head *bh,
+				 void *data, size_t size)
+{
+	struct ocfs2_triggers *ot = to_ocfs2_trigger(triggers);
+
+	/*
+	 * We aren't guaranteed to have the superblock here, so we
+	 * must unconditionally compute the ecc data.
+	 * __ocfs2_journal_access() will only set the triggers if
+	 * metaecc is enabled.
+	 */
+	ocfs2_block_check_compute(data, size, data + ot->ot_offset);
+}
+
+/*
+ * Quota blocks have their own trigger because the struct ocfs2_block_check
+ * offset depends on the blocksize.
+ */
+static void ocfs2_dq_commit_trigger(struct jbd2_buffer_trigger_type *triggers,
+				 struct buffer_head *bh,
+				 void *data, size_t size)
+{
+	struct ocfs2_disk_dqtrailer *dqt =
+		ocfs2_block_dqtrailer(size, data);
+
+	/*
+	 * We aren't guaranteed to have the superblock here, so we
+	 * must unconditionally compute the ecc data.
+	 * __ocfs2_journal_access() will only set the triggers if
+	 * metaecc is enabled.
+	 */
+	ocfs2_block_check_compute(data, size, &dqt->dq_check);
+}
+
+/*
+ * Directory blocks also have their own trigger because the
+ * struct ocfs2_block_check offset depends on the blocksize.
+ */
+static void ocfs2_db_commit_trigger(struct jbd2_buffer_trigger_type *triggers,
+				 struct buffer_head *bh,
+				 void *data, size_t size)
+{
+	struct ocfs2_dir_block_trailer *trailer =
+		ocfs2_dir_trailer_from_size(size, data);
+
+	/*
+	 * We aren't guaranteed to have the superblock here, so we
+	 * must unconditionally compute the ecc data.
+	 * __ocfs2_journal_access() will only set the triggers if
+	 * metaecc is enabled.
+	 */
+	ocfs2_block_check_compute(data, size, &trailer->db_check);
+}
+
+static void ocfs2_abort_trigger(struct jbd2_buffer_trigger_type *triggers,
+				struct buffer_head *bh)
+{
+	mlog(ML_ERROR,
+	     "ocfs2_abort_trigger called by JBD2.  bh = 0x%lx, "
+	     "bh->b_blocknr = %llu\n",
+	     (unsigned long)bh,
+	     (unsigned long long)bh->b_blocknr);
+
+	/* We aren't guaranteed to have the superblock here - but if we
+	 * don't, it'll just crash. */
+	ocfs2_error(bh->b_assoc_map->host->i_sb,
+		    "JBD2 has aborted our journal, ocfs2 cannot continue\n");
+}
+
+static struct ocfs2_triggers di_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+	.ot_offset	= offsetof(struct ocfs2_dinode, i_check),
+};
+
+static struct ocfs2_triggers eb_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+	.ot_offset	= offsetof(struct ocfs2_extent_block, h_check),
+};
+
+static struct ocfs2_triggers gd_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+	.ot_offset	= offsetof(struct ocfs2_group_desc, bg_check),
+};
+
+static struct ocfs2_triggers db_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_db_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+};
+
+static struct ocfs2_triggers xb_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+	.ot_offset	= offsetof(struct ocfs2_xattr_block, xb_check),
+};
+
+static struct ocfs2_triggers dq_triggers = {
+	.ot_triggers = {
+		.t_commit = ocfs2_dq_commit_trigger,
+		.t_abort = ocfs2_abort_trigger,
+	},
+};
+
+static int __ocfs2_journal_access(handle_t *handle,
+				  struct inode *inode,
+				  struct buffer_head *bh,
+				  struct ocfs2_triggers *triggers,
+				  int type)
 {
 	int status;
 
@@ -406,6 +546,8 @@ int ocfs2_journal_access(handle_t *handle,
 		status = -EINVAL;
 		mlog(ML_ERROR, "Uknown access type!\n");
 	}
+	if (!status && ocfs2_meta_ecc(OCFS2_SB(inode->i_sb)) && triggers)
+		jbd2_journal_set_triggers(bh, &triggers->ot_triggers);
 	mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
 
 	if (status < 0)
@@ -416,6 +558,54 @@ int ocfs2_journal_access(handle_t *handle,
 	return status;
 }
 
+int ocfs2_journal_access_di(handle_t *handle, struct inode *inode,
+			       struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &di_triggers,
+				      type);
+}
+
+int ocfs2_journal_access_eb(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &eb_triggers,
+				      type);
+}
+
+int ocfs2_journal_access_gd(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &gd_triggers,
+				      type);
+}
+
+int ocfs2_journal_access_db(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &db_triggers,
+				      type);
+}
+
+int ocfs2_journal_access_xb(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &xb_triggers,
+				      type);
+}
+
+int ocfs2_journal_access_dq(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, &dq_triggers,
+				      type);
+}
+
+int ocfs2_journal_access(handle_t *handle, struct inode *inode,
+			 struct buffer_head *bh, int type)
+{
+	return __ocfs2_journal_access(handle, inode, bh, NULL, type);
+}
+
 int ocfs2_journal_dirty(handle_t *handle,
 			struct buffer_head *bh)
 {
@@ -434,20 +624,6 @@ int ocfs2_journal_dirty(handle_t *handle,
 	return status;
 }
 
-#ifdef CONFIG_OCFS2_COMPAT_JBD
-int ocfs2_journal_dirty_data(handle_t *handle,
-			     struct buffer_head *bh)
-{
-	int err = journal_dirty_data(handle, bh);
-	if (err)
-		mlog_errno(err);
-	/* TODO: When we can handle it, abort the handle and go RO on
-	 * error here. */
-
-	return err;
-}
-#endif
-
 #define OCFS2_DEFAULT_COMMIT_INTERVAL	(HZ * JBD2_DEFAULT_MAX_COMMIT_AGE)
 
 void ocfs2_set_journal_params(struct ocfs2_super *osb)
@@ -587,17 +763,11 @@ static int ocfs2_journal_toggle_dirty(struct ocfs2_super *osb,
 	mlog_entry_void();
 
 	fe = (struct ocfs2_dinode *)bh->b_data;
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		/* This is called from startup/shutdown which will
-		 * handle the errors in a specific manner, so no need
-		 * to call ocfs2_error() here. */
-		mlog(ML_ERROR, "Journal dinode %llu  has invalid "
-		     "signature: %.*s",
-		     (unsigned long long)le64_to_cpu(fe->i_blkno), 7,
-		     fe->i_signature);
-		status = -EIO;
-		goto out;
-	}
+
+	/* The journal bh on the osb always comes from ocfs2_journal_init()
+	 * and was validated there inside ocfs2_inode_lock_full().  It's a
+	 * code bug if we mess it up. */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
 
 	flags = le32_to_cpu(fe->id1.journal1.ij_flags);
 	if (dirty)
@@ -609,11 +779,11 @@ static int ocfs2_journal_toggle_dirty(struct ocfs2_super *osb,
 	if (replayed)
 		ocfs2_bump_recovery_generation(fe);
 
+	ocfs2_compute_meta_ecc(osb->sb, bh->b_data, &fe->i_check);
 	status = ocfs2_write_block(osb, bh, journal->j_inode);
 	if (status < 0)
 		mlog_errno(status);
 
-out:
 	mlog_exit(status);
 	return status;
 }
@@ -878,6 +1048,7 @@ struct ocfs2_la_recovery_item {
 	int			lri_slot;
 	struct ocfs2_dinode	*lri_la_dinode;
 	struct ocfs2_dinode	*lri_tl_dinode;
+	struct ocfs2_quota_recovery *lri_qrec;
 };
 
 /* Does the second half of the recovery process. By this point, the
@@ -898,6 +1069,7 @@ void ocfs2_complete_recovery(struct work_struct *work)
 	struct ocfs2_super *osb = journal->j_osb;
 	struct ocfs2_dinode *la_dinode, *tl_dinode;
 	struct ocfs2_la_recovery_item *item, *n;
+	struct ocfs2_quota_recovery *qrec;
 	LIST_HEAD(tmp_la_list);
 
 	mlog_entry_void();
@@ -913,6 +1085,8 @@ void ocfs2_complete_recovery(struct work_struct *work)
 
 		mlog(0, "Complete recovery for slot %d\n", item->lri_slot);
 
+		ocfs2_wait_on_quotas(osb);
+
 		la_dinode = item->lri_la_dinode;
 		if (la_dinode) {
 			mlog(0, "Clean up local alloc %llu\n",
@@ -943,6 +1117,16 @@ void ocfs2_complete_recovery(struct work_struct *work)
 		if (ret < 0)
 			mlog_errno(ret);
 
+		qrec = item->lri_qrec;
+		if (qrec) {
+			mlog(0, "Recovering quota files");
+			ret = ocfs2_finish_quota_recovery(osb, qrec,
+							  item->lri_slot);
+			if (ret < 0)
+				mlog_errno(ret);
+			/* Recovery info is already freed now */
+		}
+
 		kfree(item);
 	}
 
@@ -956,7 +1140,8 @@ void ocfs2_complete_recovery(struct work_struct *work)
 static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal,
 					    int slot_num,
 					    struct ocfs2_dinode *la_dinode,
-					    struct ocfs2_dinode *tl_dinode)
+					    struct ocfs2_dinode *tl_dinode,
+					    struct ocfs2_quota_recovery *qrec)
 {
 	struct ocfs2_la_recovery_item *item;
 
@@ -971,6 +1156,9 @@ static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal,
 		if (tl_dinode)
 			kfree(tl_dinode);
 
+		if (qrec)
+			ocfs2_free_quota_recovery(qrec);
+
 		mlog_errno(-ENOMEM);
 		return;
 	}
@@ -979,6 +1167,7 @@ static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal,
 	item->lri_la_dinode = la_dinode;
 	item->lri_slot = slot_num;
 	item->lri_tl_dinode = tl_dinode;
+	item->lri_qrec = qrec;
 
 	spin_lock(&journal->j_lock);
 	list_add_tail(&item->lri_list, &journal->j_la_cleanups);
@@ -998,6 +1187,7 @@ void ocfs2_complete_mount_recovery(struct ocfs2_super *osb)
 		ocfs2_queue_recovery_completion(journal,
 						osb->slot_num,
 						osb->local_alloc_copy,
+						NULL,
 						NULL);
 		ocfs2_schedule_truncate_log_flush(osb, 0);
 
@@ -1006,11 +1196,26 @@ void ocfs2_complete_mount_recovery(struct ocfs2_super *osb)
 	}
 }
 
+void ocfs2_complete_quota_recovery(struct ocfs2_super *osb)
+{
+	if (osb->quota_rec) {
+		ocfs2_queue_recovery_completion(osb->journal,
+						osb->slot_num,
+						NULL,
+						NULL,
+						osb->quota_rec);
+		osb->quota_rec = NULL;
+	}
+}
+
 static int __ocfs2_recovery_thread(void *arg)
 {
-	int status, node_num;
+	int status, node_num, slot_num;
 	struct ocfs2_super *osb = arg;
 	struct ocfs2_recovery_map *rm = osb->recovery_map;
+	int *rm_quota = NULL;
+	int rm_quota_used = 0, i;
+	struct ocfs2_quota_recovery *qrec;
 
 	mlog_entry_void();
 
@@ -1019,6 +1224,11 @@ static int __ocfs2_recovery_thread(void *arg)
 		goto bail;
 	}
 
+	rm_quota = kzalloc(osb->max_slots * sizeof(int), GFP_NOFS);
+	if (!rm_quota) {
+		status = -ENOMEM;
+		goto bail;
+	}
 restart:
 	status = ocfs2_super_lock(osb, 1);
 	if (status < 0) {
@@ -1032,8 +1242,28 @@ restart:
 		 * clear it until ocfs2_recover_node() has succeeded. */
 		node_num = rm->rm_entries[0];
 		spin_unlock(&osb->osb_lock);
-
-		status = ocfs2_recover_node(osb, node_num);
+		mlog(0, "checking node %d\n", node_num);
+		slot_num = ocfs2_node_num_to_slot(osb, node_num);
+		if (slot_num == -ENOENT) {
+			status = 0;
+			mlog(0, "no slot for this node, so no recovery"
+			     "required.\n");
+			goto skip_recovery;
+		}
+		mlog(0, "node %d was using slot %d\n", node_num, slot_num);
+
+		/* It is a bit subtle with quota recovery. We cannot do it
+		 * immediately because we have to obtain cluster locks from
+		 * quota files and we also don't want to just skip it because
+		 * then quota usage would be out of sync until some node takes
+		 * the slot. So we remember which nodes need quota recovery
+		 * and when everything else is done, we recover quotas. */
+		for (i = 0; i < rm_quota_used && rm_quota[i] != slot_num; i++);
+		if (i == rm_quota_used)
+			rm_quota[rm_quota_used++] = slot_num;
+
+		status = ocfs2_recover_node(osb, node_num, slot_num);
+skip_recovery:
 		if (!status) {
 			ocfs2_recovery_map_clear(osb, node_num);
 		} else {
@@ -1055,13 +1285,27 @@ restart:
 	if (status < 0)
 		mlog_errno(status);
 
+	/* Now it is right time to recover quotas... We have to do this under
+	 * superblock lock so that noone can start using the slot (and crash)
+	 * before we recover it */
+	for (i = 0; i < rm_quota_used; i++) {
+		qrec = ocfs2_begin_quota_recovery(osb, rm_quota[i]);
+		if (IS_ERR(qrec)) {
+			status = PTR_ERR(qrec);
+			mlog_errno(status);
+			continue;
+		}
+		ocfs2_queue_recovery_completion(osb->journal, rm_quota[i],
+						NULL, NULL, qrec);
+	}
+
 	ocfs2_super_unlock(osb, 1);
 
 	/* We always run recovery on our own orphan dir - the dead
 	 * node(s) may have disallowd a previos inode delete. Re-processing
 	 * is therefore required. */
 	ocfs2_queue_recovery_completion(osb->journal, osb->slot_num, NULL,
-					NULL);
+					NULL, NULL);
 
 bail:
 	mutex_lock(&osb->recovery_lock);
@@ -1076,6 +1320,9 @@ bail:
 
 	mutex_unlock(&osb->recovery_lock);
 
+	if (rm_quota)
+		kfree(rm_quota);
+
 	mlog_exit(status);
 	/* no one is callint kthread_stop() for us so the kthread() api
 	 * requires that we call do_exit().  And it isn't exported, but
@@ -1135,8 +1382,7 @@ static int ocfs2_read_journal_inode(struct ocfs2_super *osb,
 	}
 	SET_INODE_JOURNAL(inode);
 
-	status = ocfs2_read_blocks(inode, OCFS2_I(inode)->ip_blkno, 1, bh,
-				   OCFS2_BH_IGNORE_CACHE);
+	status = ocfs2_read_inode_block_full(inode, bh, OCFS2_BH_IGNORE_CACHE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -1268,6 +1514,7 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
 	osb->slot_recovery_generations[slot_num] =
 					ocfs2_get_recovery_generation(fe);
 
+	ocfs2_compute_meta_ecc(osb->sb, bh->b_data, &fe->i_check);
 	status = ocfs2_write_block(osb, bh, inode);
 	if (status < 0)
 		mlog_errno(status);
@@ -1304,31 +1551,19 @@ done:
  * far less concerning.
  */
 static int ocfs2_recover_node(struct ocfs2_super *osb,
-			      int node_num)
+			      int node_num, int slot_num)
 {
 	int status = 0;
-	int slot_num;
 	struct ocfs2_dinode *la_copy = NULL;
 	struct ocfs2_dinode *tl_copy = NULL;
 
-	mlog_entry("(node_num=%d, osb->node_num = %d)\n",
-		   node_num, osb->node_num);
-
-	mlog(0, "checking node %d\n", node_num);
+	mlog_entry("(node_num=%d, slot_num=%d, osb->node_num = %d)\n",
+		   node_num, slot_num, osb->node_num);
 
 	/* Should not ever be called to recover ourselves -- in that
 	 * case we should've called ocfs2_journal_load instead. */
 	BUG_ON(osb->node_num == node_num);
 
-	slot_num = ocfs2_node_num_to_slot(osb, node_num);
-	if (slot_num == -ENOENT) {
-		status = 0;
-		mlog(0, "no slot for this node, so no recovery required.\n");
-		goto done;
-	}
-
-	mlog(0, "node %d was using slot %d\n", node_num, slot_num);
-
 	status = ocfs2_replay_journal(osb, node_num, slot_num);
 	if (status < 0) {
 		if (status == -EBUSY) {
@@ -1364,7 +1599,7 @@ static int ocfs2_recover_node(struct ocfs2_super *osb,
 
 	/* This will kfree the memory pointed to by la_copy and tl_copy */
 	ocfs2_queue_recovery_completion(osb->journal, slot_num, la_copy,
-					tl_copy);
+					tl_copy, NULL);
 
 	status = 0;
 done:
@@ -1659,13 +1894,14 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
 	return ret;
 }
 
-static int ocfs2_wait_on_mount(struct ocfs2_super *osb)
+static int __ocfs2_wait_on_mount(struct ocfs2_super *osb, int quota)
 {
 	/* This check is good because ocfs2 will wait on our recovery
 	 * thread before changing it to something other than MOUNTED
 	 * or DISABLED. */
 	wait_event(osb->osb_mount_event,
-		   atomic_read(&osb->vol_state) == VOLUME_MOUNTED ||
+		  (!quota && atomic_read(&osb->vol_state) == VOLUME_MOUNTED) ||
+		   atomic_read(&osb->vol_state) == VOLUME_MOUNTED_QUOTAS ||
 		   atomic_read(&osb->vol_state) == VOLUME_DISABLED);
 
 	/* If there's an error on mount, then we may never get to the

fs/ocfs2/journal.h

@@ -27,12 +27,7 @@
 #define OCFS2_JOURNAL_H
 
 #include <linux/fs.h>
-#ifndef CONFIG_OCFS2_COMPAT_JBD
-# include <linux/jbd2.h>
-#else
-# include <linux/jbd.h>
-# include "ocfs2_jbd_compat.h"
-#endif
+#include <linux/jbd2.h>
 
 enum ocfs2_journal_state {
 	OCFS2_JOURNAL_FREE = 0,
@@ -173,6 +168,7 @@ void   ocfs2_recovery_thread(struct ocfs2_super *osb,
 			     int node_num);
 int    ocfs2_mark_dead_nodes(struct ocfs2_super *osb);
 void   ocfs2_complete_mount_recovery(struct ocfs2_super *osb);
+void ocfs2_complete_quota_recovery(struct ocfs2_super *osb);
 
 static inline void ocfs2_start_checkpoint(struct ocfs2_super *osb)
 {
@@ -216,9 +212,12 @@ static inline void ocfs2_checkpoint_inode(struct inode *inode)
  *  ocfs2_extend_trans     - Extend a handle by nblocks credits. This may
  *                          commit the handle to disk in the process, but will
  *                          not release any locks taken during the transaction.
- *  ocfs2_journal_access   - Notify the handle that we want to journal this
+ *  ocfs2_journal_access* - Notify the handle that we want to journal this
  *                          buffer. Will have to call ocfs2_journal_dirty once
  *                          we've actually dirtied it. Type is one of . or .
+ *                          Always call the specific flavor of
+ *                          ocfs2_journal_access_*() unless you intend to
+ *                          manage the checksum by hand.
  *  ocfs2_journal_dirty    - Mark a journalled buffer as having dirty data.
  *  ocfs2_jbd2_file_inode  - Mark an inode so that its data goes out before
  *                           the current handle commits.
@@ -248,10 +247,29 @@ int			     ocfs2_extend_trans(handle_t *handle, int nblocks);
 #define OCFS2_JOURNAL_ACCESS_WRITE  1
 #define OCFS2_JOURNAL_ACCESS_UNDO   2
 
-int                  ocfs2_journal_access(handle_t *handle,
-					  struct inode *inode,
-					  struct buffer_head *bh,
-					  int type);
+
+/* ocfs2_inode */
+int ocfs2_journal_access_di(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
+/* ocfs2_extent_block */
+int ocfs2_journal_access_eb(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
+/* ocfs2_group_desc */
+int ocfs2_journal_access_gd(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
+/* ocfs2_xattr_block */
+int ocfs2_journal_access_xb(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
+/* quota blocks */
+int ocfs2_journal_access_dq(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
+/* dirblock */
+int ocfs2_journal_access_db(handle_t *handle, struct inode *inode,
+			    struct buffer_head *bh, int type);
+/* Anything that has no ecc */
+int ocfs2_journal_access(handle_t *handle, struct inode *inode,
+			 struct buffer_head *bh, int type);
+
 /*
  * A word about the journal_access/journal_dirty "dance". It is
  * entirely legal to journal_access a buffer more than once (as long
@@ -273,10 +291,6 @@ int                  ocfs2_journal_access(handle_t *handle,
  */
 int                  ocfs2_journal_dirty(handle_t *handle,
 					 struct buffer_head *bh);
-#ifdef CONFIG_OCFS2_COMPAT_JBD
-int                  ocfs2_journal_dirty_data(handle_t *handle,
-					      struct buffer_head *bh);
-#endif
 
 /*
  *  Credit Macros:
@@ -293,6 +307,37 @@ int                  ocfs2_journal_dirty_data(handle_t *handle,
 /* extended attribute block update */
 #define OCFS2_XATTR_BLOCK_UPDATE_CREDITS 1
 
+/* global quotafile inode update, data block */
+#define OCFS2_QINFO_WRITE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1)
+
+/*
+ * The two writes below can accidentally see global info dirty due
+ * to set_info() quotactl so make them prepared for the writes.
+ */
+/* quota data block, global info */
+/* Write to local quota file */
+#define OCFS2_QWRITE_CREDITS (OCFS2_QINFO_WRITE_CREDITS + 1)
+
+/* global quota data block, local quota data block, global quota inode,
+ * global quota info */
+#define OCFS2_QSYNC_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 3)
+
+static inline int ocfs2_quota_trans_credits(struct super_block *sb)
+{
+	int credits = 0;
+
+	if (OCFS2_HAS_RO_COMPAT_FEATURE(sb, OCFS2_FEATURE_RO_COMPAT_USRQUOTA))
+		credits += OCFS2_QWRITE_CREDITS;
+	if (OCFS2_HAS_RO_COMPAT_FEATURE(sb, OCFS2_FEATURE_RO_COMPAT_GRPQUOTA))
+		credits += OCFS2_QWRITE_CREDITS;
+	return credits;
+}
+
+/* Number of credits needed for removing quota structure from file */
+int ocfs2_calc_qdel_credits(struct super_block *sb, int type);
+/* Number of credits needed for initialization of new quota structure */
+int ocfs2_calc_qinit_credits(struct super_block *sb, int type);
+
 /* group extend. inode update and last group update. */
 #define OCFS2_GROUP_EXTEND_CREDITS	(OCFS2_INODE_UPDATE_CREDITS + 1)
 
@@ -303,8 +348,11 @@ int                  ocfs2_journal_dirty_data(handle_t *handle,
  * prev. group desc. if we relink. */
 #define OCFS2_SUBALLOC_ALLOC (3)
 
-#define OCFS2_INLINE_TO_EXTENTS_CREDITS (OCFS2_SUBALLOC_ALLOC		\
-					 + OCFS2_INODE_UPDATE_CREDITS)
+static inline int ocfs2_inline_to_extents_credits(struct super_block *sb)
+{
+	return OCFS2_SUBALLOC_ALLOC + OCFS2_INODE_UPDATE_CREDITS +
+	       ocfs2_quota_trans_credits(sb);
+}
 
 /* dinode + group descriptor update. We don't relink on free yet. */
 #define OCFS2_SUBALLOC_FREE  (2)
@@ -313,16 +361,23 @@ int                  ocfs2_journal_dirty_data(handle_t *handle,
 #define OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC (OCFS2_SUBALLOC_FREE 		      \
 					 + OCFS2_TRUNCATE_LOG_UPDATE)
 
-#define OCFS2_REMOVE_EXTENT_CREDITS (OCFS2_TRUNCATE_LOG_UPDATE + OCFS2_INODE_UPDATE_CREDITS)
+static inline int ocfs2_remove_extent_credits(struct super_block *sb)
+{
+	return OCFS2_TRUNCATE_LOG_UPDATE + OCFS2_INODE_UPDATE_CREDITS +
+	       ocfs2_quota_trans_credits(sb);
+}
 
 /* data block for new dir/symlink, 2 for bitmap updates (bitmap fe +
  * bitmap block for the new bit) */
 #define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + 2)
 
 /* parent fe, parent block, new file entry, inode alloc fe, inode alloc
- * group descriptor + mkdir/symlink blocks */
-#define OCFS2_MKNOD_CREDITS (3 + OCFS2_SUBALLOC_ALLOC                         \
-			    + OCFS2_DIR_LINK_ADDITIONAL_CREDITS)
+ * group descriptor + mkdir/symlink blocks + quota update */
+static inline int ocfs2_mknod_credits(struct super_block *sb)
+{
+	return 3 + OCFS2_SUBALLOC_ALLOC + OCFS2_DIR_LINK_ADDITIONAL_CREDITS +
+	       ocfs2_quota_trans_credits(sb);
+}
 
 /* local alloc metadata change + main bitmap updates */
 #define OCFS2_WINDOW_MOVE_CREDITS (OCFS2_INODE_UPDATE_CREDITS                 \
@@ -332,13 +387,21 @@ int                  ocfs2_journal_dirty_data(handle_t *handle,
  * for the dinode, one for the new block. */
 #define OCFS2_SIMPLE_DIR_EXTEND_CREDITS (2)
 
-/* file update (nlink, etc) + directory mtime/ctime + dir entry block */
-#define OCFS2_LINK_CREDITS  (2*OCFS2_INODE_UPDATE_CREDITS + 1)
+/* file update (nlink, etc) + directory mtime/ctime + dir entry block + quota
+ * update on dir */
+static inline int ocfs2_link_credits(struct super_block *sb)
+{
+	return 2*OCFS2_INODE_UPDATE_CREDITS + 1 +
+	       ocfs2_quota_trans_credits(sb);
+}
 
 /* inode + dir inode (if we unlink a dir), + dir entry block + orphan
  * dir inode link */
-#define OCFS2_UNLINK_CREDITS  (2 * OCFS2_INODE_UPDATE_CREDITS + 1             \
-			      + OCFS2_LINK_CREDITS)
+static inline int ocfs2_unlink_credits(struct super_block *sb)
+{
+	/* The quota update from ocfs2_link_credits is unused here... */
+	return 2 * OCFS2_INODE_UPDATE_CREDITS + 1 + ocfs2_link_credits(sb);
+}
 
 /* dinode + orphan dir dinode + inode alloc dinode + orphan dir entry +
  * inode alloc group descriptor */
@@ -347,8 +410,10 @@ int                  ocfs2_journal_dirty_data(handle_t *handle,
 /* dinode update, old dir dinode update, new dir dinode update, old
  * dir dir entry, new dir dir entry, dir entry update for renaming
  * directory + target unlink */
-#define OCFS2_RENAME_CREDITS (3 * OCFS2_INODE_UPDATE_CREDITS + 3              \
-			     + OCFS2_UNLINK_CREDITS)
+static inline int ocfs2_rename_credits(struct super_block *sb)
+{
+	return 3 * OCFS2_INODE_UPDATE_CREDITS + 3 + ocfs2_unlink_credits(sb);
+}
 
 /* global bitmap dinode, group desc., relinked group,
  * suballocator dinode, group desc., relinked group,
@@ -386,18 +451,19 @@ static inline int ocfs2_calc_extend_credits(struct super_block *sb,
 	 * credit for the dinode there. */
 	extent_blocks = 1 + 1 + le16_to_cpu(root_el->l_tree_depth);
 
-	return bitmap_blocks + sysfile_bitmap_blocks + extent_blocks;
+	return bitmap_blocks + sysfile_bitmap_blocks + extent_blocks +
+	       ocfs2_quota_trans_credits(sb);
 }
 
 static inline int ocfs2_calc_symlink_credits(struct super_block *sb)
 {
-	int blocks = OCFS2_MKNOD_CREDITS;
+	int blocks = ocfs2_mknod_credits(sb);
 
 	/* links can be longer than one block so we may update many
 	 * within our single allocated extent. */
 	blocks += ocfs2_clusters_to_blocks(sb, 1);
 
-	return blocks;
+	return blocks + ocfs2_quota_trans_credits(sb);
 }
 
 static inline int ocfs2_calc_group_alloc_credits(struct super_block *sb,
@@ -434,6 +500,8 @@ static inline int ocfs2_calc_tree_trunc_credits(struct super_block *sb,
 	/* update to the truncate log. */
 	credits += OCFS2_TRUNCATE_LOG_UPDATE;
 
+	credits += ocfs2_quota_trans_credits(sb);
+
 	return credits;
 }
 

fs/ocfs2/localalloc.c

@@ -36,6 +36,7 @@
 #include "ocfs2.h"
 
 #include "alloc.h"
+#include "blockcheck.h"
 #include "dlmglue.h"
 #include "inode.h"
 #include "journal.h"
@@ -248,8 +249,8 @@ int ocfs2_load_local_alloc(struct ocfs2_super *osb)
 		goto bail;
 	}
 
-	status = ocfs2_read_blocks(inode, OCFS2_I(inode)->ip_blkno, 1,
-				   &alloc_bh, OCFS2_BH_IGNORE_CACHE);
+	status = ocfs2_read_inode_block_full(inode, &alloc_bh,
+					     OCFS2_BH_IGNORE_CACHE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -382,8 +383,8 @@ void ocfs2_shutdown_local_alloc(struct ocfs2_super *osb)
 	}
 	memcpy(alloc_copy, alloc, bh->b_size);
 
-	status = ocfs2_journal_access(handle, local_alloc_inode, bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, local_alloc_inode, bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto out_commit;
@@ -459,8 +460,8 @@ int ocfs2_begin_local_alloc_recovery(struct ocfs2_super *osb,
 
 	mutex_lock(&inode->i_mutex);
 
-	status = ocfs2_read_blocks(inode, OCFS2_I(inode)->ip_blkno, 1,
-				   &alloc_bh, OCFS2_BH_IGNORE_CACHE);
+	status = ocfs2_read_inode_block_full(inode, &alloc_bh,
+					     OCFS2_BH_IGNORE_CACHE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -476,6 +477,7 @@ int ocfs2_begin_local_alloc_recovery(struct ocfs2_super *osb,
 	alloc = (struct ocfs2_dinode *) alloc_bh->b_data;
 	ocfs2_clear_local_alloc(alloc);
 
+	ocfs2_compute_meta_ecc(osb->sb, alloc_bh->b_data, &alloc->i_check);
 	status = ocfs2_write_block(osb, alloc_bh, inode);
 	if (status < 0)
 		mlog_errno(status);
@@ -762,9 +764,9 @@ int ocfs2_claim_local_alloc_bits(struct ocfs2_super *osb,
 	 * delete bits from it! */
 	*num_bits = bits_wanted;
 
-	status = ocfs2_journal_access(handle, local_alloc_inode,
-				      osb->local_alloc_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, local_alloc_inode,
+					 osb->local_alloc_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -1240,9 +1242,9 @@ static int ocfs2_local_alloc_slide_window(struct ocfs2_super *osb,
 	}
 	memcpy(alloc_copy, alloc, osb->local_alloc_bh->b_size);
 
-	status = ocfs2_journal_access(handle, local_alloc_inode,
-				      osb->local_alloc_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, local_alloc_inode,
+					 osb->local_alloc_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;

fs/ocfs2/namei.c

@@ -40,6 +40,7 @@
 #include <linux/types.h>
 #include <linux/slab.h>
 #include <linux/highmem.h>
+#include <linux/quotaops.h>
 
 #define MLOG_MASK_PREFIX ML_NAMEI
 #include <cluster/masklog.h>
@@ -61,17 +62,18 @@
 #include "sysfile.h"
 #include "uptodate.h"
 #include "xattr.h"
+#include "acl.h"
 
 #include "buffer_head_io.h"
 
 static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 			      struct inode *dir,
-			      struct dentry *dentry, int mode,
+			      struct inode *inode,
+			      struct dentry *dentry,
 			      dev_t dev,
 			      struct buffer_head **new_fe_bh,
 			      struct buffer_head *parent_fe_bh,
 			      handle_t *handle,
-			      struct inode **ret_inode,
 			      struct ocfs2_alloc_context *inode_ac);
 
 static int ocfs2_prepare_orphan_dir(struct ocfs2_super *osb,
@@ -186,6 +188,35 @@ bail:
 	return ret;
 }
 
+static struct inode *ocfs2_get_init_inode(struct inode *dir, int mode)
+{
+	struct inode *inode;
+
+	inode = new_inode(dir->i_sb);
+	if (!inode) {
+		mlog(ML_ERROR, "new_inode failed!\n");
+		return NULL;
+	}
+
+	/* populate as many fields early on as possible - many of
+	 * these are used by the support functions here and in
+	 * callers. */
+	if (S_ISDIR(mode))
+		inode->i_nlink = 2;
+	else
+		inode->i_nlink = 1;
+	inode->i_uid = current_fsuid();
+	if (dir->i_mode & S_ISGID) {
+		inode->i_gid = dir->i_gid;
+		if (S_ISDIR(mode))
+			mode |= S_ISGID;
+	} else
+		inode->i_gid = current_fsgid();
+	inode->i_mode = mode;
+	vfs_dq_init(inode);
+	return inode;
+}
+
 static int ocfs2_mknod(struct inode *dir,
 		       struct dentry *dentry,
 		       int mode,
@@ -201,6 +232,13 @@ static int ocfs2_mknod(struct inode *dir,
 	struct inode *inode = NULL;
 	struct ocfs2_alloc_context *inode_ac = NULL;
 	struct ocfs2_alloc_context *data_ac = NULL;
+	struct ocfs2_alloc_context *xattr_ac = NULL;
+	int want_clusters = 0;
+	int xattr_credits = 0;
+	struct ocfs2_security_xattr_info si = {
+		.enable = 1,
+	};
+	int did_quota_inode = 0;
 
 	mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry, mode,
 		   (unsigned long)dev, dentry->d_name.len,
@@ -250,17 +288,46 @@ static int ocfs2_mknod(struct inode *dir,
 		goto leave;
 	}
 
-	/* Reserve a cluster if creating an extent based directory. */
-	if (S_ISDIR(mode) && !ocfs2_supports_inline_data(osb)) {
-		status = ocfs2_reserve_clusters(osb, 1, &data_ac);
-		if (status < 0) {
-			if (status != -ENOSPC)
-				mlog_errno(status);
+	inode = ocfs2_get_init_inode(dir, mode);
+	if (!inode) {
+		status = -ENOMEM;
+		mlog_errno(status);
+		goto leave;
+	}
+
+	/* get security xattr */
+	status = ocfs2_init_security_get(inode, dir, &si);
+	if (status) {
+		if (status == -EOPNOTSUPP)
+			si.enable = 0;
+		else {
+			mlog_errno(status);
 			goto leave;
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_MKNOD_CREDITS);
+	/* calculate meta data/clusters for setting security and acl xattr */
+	status = ocfs2_calc_xattr_init(dir, parent_fe_bh, mode,
+					&si, &want_clusters,
+					&xattr_credits, &xattr_ac);
+	if (status < 0) {
+		mlog_errno(status);
+		goto leave;
+	}
+
+	/* Reserve a cluster if creating an extent based directory. */
+	if (S_ISDIR(mode) && !ocfs2_supports_inline_data(osb))
+		want_clusters += 1;
+
+	status = ocfs2_reserve_clusters(osb, want_clusters, &data_ac);
+	if (status < 0) {
+		if (status != -ENOSPC)
+			mlog_errno(status);
+		goto leave;
+	}
+
+	handle = ocfs2_start_trans(osb, ocfs2_mknod_credits(osb->sb) +
+				   xattr_credits);
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		handle = NULL;
@@ -268,10 +335,19 @@ static int ocfs2_mknod(struct inode *dir,
 		goto leave;
 	}
 
+	/* We don't use standard VFS wrapper because we don't want vfs_dq_init
+	 * to be called. */
+	if (sb_any_quota_active(osb->sb) &&
+	    osb->sb->dq_op->alloc_inode(inode, 1) == NO_QUOTA) {
+		status = -EDQUOT;
+		goto leave;
+	}
+	did_quota_inode = 1;
+
 	/* do the real work now. */
-	status = ocfs2_mknod_locked(osb, dir, dentry, mode, dev,
+	status = ocfs2_mknod_locked(osb, dir, inode, dentry, dev,
 				    &new_fe_bh, parent_fe_bh, handle,
-				    &inode, inode_ac);
+				    inode_ac);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -285,8 +361,8 @@ static int ocfs2_mknod(struct inode *dir,
 			goto leave;
 		}
 
-		status = ocfs2_journal_access(handle, dir, parent_fe_bh,
-					      OCFS2_JOURNAL_ACCESS_WRITE);
+		status = ocfs2_journal_access_di(handle, dir, parent_fe_bh,
+						 OCFS2_JOURNAL_ACCESS_WRITE);
 		if (status < 0) {
 			mlog_errno(status);
 			goto leave;
@@ -300,6 +376,22 @@ static int ocfs2_mknod(struct inode *dir,
 		inc_nlink(dir);
 	}
 
+	status = ocfs2_init_acl(handle, inode, dir, new_fe_bh, parent_fe_bh,
+				xattr_ac, data_ac);
+	if (status < 0) {
+		mlog_errno(status);
+		goto leave;
+	}
+
+	if (si.enable) {
+		status = ocfs2_init_security_set(handle, inode, new_fe_bh, &si,
+						 xattr_ac, data_ac);
+		if (status < 0) {
+			mlog_errno(status);
+			goto leave;
+		}
+	}
+
 	status = ocfs2_add_entry(handle, dentry, inode,
 				 OCFS2_I(inode)->ip_blkno, parent_fe_bh,
 				 de_bh);
@@ -320,6 +412,8 @@ static int ocfs2_mknod(struct inode *dir,
 	d_instantiate(dentry, inode);
 	status = 0;
 leave:
+	if (status < 0 && did_quota_inode)
+		vfs_dq_free_inode(inode);
 	if (handle)
 		ocfs2_commit_trans(osb, handle);
 
@@ -331,9 +425,13 @@ leave:
 	brelse(new_fe_bh);
 	brelse(de_bh);
 	brelse(parent_fe_bh);
+	kfree(si.name);
+	kfree(si.value);
 
-	if ((status < 0) && inode)
+	if ((status < 0) && inode) {
+		clear_nlink(inode);
 		iput(inode);
+	}
 
 	if (inode_ac)
 		ocfs2_free_alloc_context(inode_ac);
@@ -341,6 +439,9 @@ leave:
 	if (data_ac)
 		ocfs2_free_alloc_context(data_ac);
 
+	if (xattr_ac)
+		ocfs2_free_alloc_context(xattr_ac);
+
 	mlog_exit(status);
 
 	return status;
@@ -348,12 +449,12 @@ leave:
 
 static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 			      struct inode *dir,
-			      struct dentry *dentry, int mode,
+			      struct inode *inode,
+			      struct dentry *dentry,
 			      dev_t dev,
 			      struct buffer_head **new_fe_bh,
 			      struct buffer_head *parent_fe_bh,
 			      handle_t *handle,
-			      struct inode **ret_inode,
 			      struct ocfs2_alloc_context *inode_ac)
 {
 	int status = 0;
@@ -361,14 +462,12 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 	struct ocfs2_extent_list *fel;
 	u64 fe_blkno = 0;
 	u16 suballoc_bit;
-	struct inode *inode = NULL;
 
-	mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry, mode,
-		   (unsigned long)dev, dentry->d_name.len,
+	mlog_entry("(0x%p, 0x%p, %d, %lu, '%.*s')\n", dir, dentry,
+		   inode->i_mode, (unsigned long)dev, dentry->d_name.len,
 		   dentry->d_name.name);
 
 	*new_fe_bh = NULL;
-	*ret_inode = NULL;
 
 	status = ocfs2_claim_new_inode(osb, handle, inode_ac, &suballoc_bit,
 				       &fe_blkno);
@@ -377,23 +476,11 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 		goto leave;
 	}
 
-	inode = new_inode(dir->i_sb);
-	if (!inode) {
-		status = -ENOMEM;
-		mlog(ML_ERROR, "new_inode failed!\n");
-		goto leave;
-	}
-
 	/* populate as many fields early on as possible - many of
 	 * these are used by the support functions here and in
 	 * callers. */
 	inode->i_ino = ino_from_blkno(osb->sb, fe_blkno);
 	OCFS2_I(inode)->ip_blkno = fe_blkno;
-	if (S_ISDIR(mode))
-		inode->i_nlink = 2;
-	else
-		inode->i_nlink = 1;
-	inode->i_mode = mode;
 	spin_lock(&osb->osb_lock);
 	inode->i_generation = osb->s_next_generation++;
 	spin_unlock(&osb->osb_lock);
@@ -406,8 +493,8 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 	}
 	ocfs2_set_new_buffer_uptodate(inode, *new_fe_bh);
 
-	status = ocfs2_journal_access(handle, inode, *new_fe_bh,
-				      OCFS2_JOURNAL_ACCESS_CREATE);
+	status = ocfs2_journal_access_di(handle, inode, *new_fe_bh,
+					 OCFS2_JOURNAL_ACCESS_CREATE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -421,17 +508,11 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 	fe->i_blkno = cpu_to_le64(fe_blkno);
 	fe->i_suballoc_bit = cpu_to_le16(suballoc_bit);
 	fe->i_suballoc_slot = cpu_to_le16(inode_ac->ac_alloc_slot);
-	fe->i_uid = cpu_to_le32(current_fsuid());
-	if (dir->i_mode & S_ISGID) {
-		fe->i_gid = cpu_to_le32(dir->i_gid);
-		if (S_ISDIR(mode))
-			mode |= S_ISGID;
-	} else
-		fe->i_gid = cpu_to_le32(current_fsgid());
-	fe->i_mode = cpu_to_le16(mode);
-	if (S_ISCHR(mode) || S_ISBLK(mode))
+	fe->i_uid = cpu_to_le32(inode->i_uid);
+	fe->i_gid = cpu_to_le32(inode->i_gid);
+	fe->i_mode = cpu_to_le16(inode->i_mode);
+	if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode))
 		fe->id1.dev1.i_rdev = cpu_to_le64(huge_encode_dev(dev));
-
 	fe->i_links_count = cpu_to_le16(inode->i_nlink);
 
 	fe->i_last_eb_blk = 0;
@@ -446,7 +527,7 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 	/*
 	 * If supported, directories start with inline data.
 	 */
-	if (S_ISDIR(mode) && ocfs2_supports_inline_data(osb)) {
+	if (S_ISDIR(inode->i_mode) && ocfs2_supports_inline_data(osb)) {
 		u16 feat = le16_to_cpu(fe->i_dyn_features);
 
 		fe->i_dyn_features = cpu_to_le16(feat | OCFS2_INLINE_DATA_FL);
@@ -465,15 +546,7 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 		goto leave;
 	}
 
-	if (ocfs2_populate_inode(inode, fe, 1) < 0) {
-		mlog(ML_ERROR, "populate inode failed! bh->b_blocknr=%llu, "
-		     "i_blkno=%llu, i_ino=%lu\n",
-		     (unsigned long long)(*new_fe_bh)->b_blocknr,
-		     (unsigned long long)le64_to_cpu(fe->i_blkno),
-		     inode->i_ino);
-		BUG();
-	}
-
+	ocfs2_populate_inode(inode, fe, 1);
 	ocfs2_inode_set_new(osb, inode);
 	if (!ocfs2_mount_local(osb)) {
 		status = ocfs2_create_new_inode_locks(inode);
@@ -484,17 +557,12 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb,
 	status = 0; /* error in ocfs2_create_new_inode_locks is not
 		     * critical */
 
-	*ret_inode = inode;
 leave:
 	if (status < 0) {
 		if (*new_fe_bh) {
 			brelse(*new_fe_bh);
 			*new_fe_bh = NULL;
 		}
-		if (inode) {
-			clear_nlink(inode);
-			iput(inode);
-		}
 	}
 
 	mlog_exit(status);
@@ -588,7 +656,7 @@ static int ocfs2_link(struct dentry *old_dentry,
 		goto out_unlock_inode;
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_LINK_CREDITS);
+	handle = ocfs2_start_trans(osb, ocfs2_link_credits(osb->sb));
 	if (IS_ERR(handle)) {
 		err = PTR_ERR(handle);
 		handle = NULL;
@@ -596,8 +664,8 @@ static int ocfs2_link(struct dentry *old_dentry,
 		goto out_unlock_inode;
 	}
 
-	err = ocfs2_journal_access(handle, inode, fe_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	err = ocfs2_journal_access_di(handle, inode, fe_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (err < 0) {
 		mlog_errno(err);
 		goto out_commit;
@@ -775,7 +843,7 @@ static int ocfs2_unlink(struct inode *dir,
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_UNLINK_CREDITS);
+	handle = ocfs2_start_trans(osb, ocfs2_unlink_credits(osb->sb));
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		handle = NULL;
@@ -783,8 +851,8 @@ static int ocfs2_unlink(struct inode *dir,
 		goto leave;
 	}
 
-	status = ocfs2_journal_access(handle, inode, fe_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, inode, fe_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -1181,7 +1249,7 @@ static int ocfs2_rename(struct inode *old_dir,
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_RENAME_CREDITS);
+	handle = ocfs2_start_trans(osb, ocfs2_rename_credits(osb->sb));
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		handle = NULL;
@@ -1197,8 +1265,8 @@ static int ocfs2_rename(struct inode *old_dir,
 				goto bail;
 			}
 		}
-		status = ocfs2_journal_access(handle, new_inode, newfe_bh,
-					      OCFS2_JOURNAL_ACCESS_WRITE);
+		status = ocfs2_journal_access_di(handle, new_inode, newfe_bh,
+						 OCFS2_JOURNAL_ACCESS_WRITE);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
@@ -1244,8 +1312,8 @@ static int ocfs2_rename(struct inode *old_dir,
 	old_inode->i_ctime = CURRENT_TIME;
 	mark_inode_dirty(old_inode);
 
-	status = ocfs2_journal_access(handle, old_inode, old_inode_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, old_inode, old_inode_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status >= 0) {
 		old_di = (struct ocfs2_dinode *) old_inode_bh->b_data;
 
@@ -1321,9 +1389,9 @@ static int ocfs2_rename(struct inode *old_dir,
 			     (int)old_dir_nlink, old_dir->i_nlink);
 		} else {
 			struct ocfs2_dinode *fe;
-			status = ocfs2_journal_access(handle, old_dir,
-						      old_dir_bh,
-						      OCFS2_JOURNAL_ACCESS_WRITE);
+			status = ocfs2_journal_access_di(handle, old_dir,
+							 old_dir_bh,
+							 OCFS2_JOURNAL_ACCESS_WRITE);
 			fe = (struct ocfs2_dinode *) old_dir_bh->b_data;
 			fe->i_links_count = cpu_to_le16(old_dir->i_nlink);
 			status = ocfs2_journal_dirty(handle, old_dir_bh);
@@ -1496,6 +1564,13 @@ static int ocfs2_symlink(struct inode *dir,
 	handle_t *handle = NULL;
 	struct ocfs2_alloc_context *inode_ac = NULL;
 	struct ocfs2_alloc_context *data_ac = NULL;
+	struct ocfs2_alloc_context *xattr_ac = NULL;
+	int want_clusters = 0;
+	int xattr_credits = 0;
+	struct ocfs2_security_xattr_info si = {
+		.enable = 1,
+	};
+	int did_quota = 0, did_quota_inode = 0;
 
 	mlog_entry("(0x%p, 0x%p, symname='%s' actual='%.*s')\n", dir,
 		   dentry, symname, dentry->d_name.len, dentry->d_name.name);
@@ -1542,17 +1617,46 @@ static int ocfs2_symlink(struct inode *dir,
 		goto bail;
 	}
 
-	/* don't reserve bitmap space for fast symlinks. */
-	if (l > ocfs2_fast_symlink_chars(sb)) {
-		status = ocfs2_reserve_clusters(osb, 1, &data_ac);
+	inode = ocfs2_get_init_inode(dir, S_IFLNK | S_IRWXUGO);
+	if (!inode) {
+		status = -ENOMEM;
+		mlog_errno(status);
+		goto bail;
+	}
+
+	/* get security xattr */
+	status = ocfs2_init_security_get(inode, dir, &si);
+	if (status) {
+		if (status == -EOPNOTSUPP)
+			si.enable = 0;
+		else {
+			mlog_errno(status);
+			goto bail;
+		}
+	}
+
+	/* calculate meta data/clusters for setting security xattr */
+	if (si.enable) {
+		status = ocfs2_calc_security_init(dir, &si, &want_clusters,
+						  &xattr_credits, &xattr_ac);
 		if (status < 0) {
-			if (status != -ENOSPC)
-				mlog_errno(status);
+			mlog_errno(status);
 			goto bail;
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, credits);
+	/* don't reserve bitmap space for fast symlinks. */
+	if (l > ocfs2_fast_symlink_chars(sb))
+		want_clusters += 1;
+
+	status = ocfs2_reserve_clusters(osb, want_clusters, &data_ac);
+	if (status < 0) {
+		if (status != -ENOSPC)
+			mlog_errno(status);
+		goto bail;
+	}
+
+	handle = ocfs2_start_trans(osb, credits + xattr_credits);
 	if (IS_ERR(handle)) {
 		status = PTR_ERR(handle);
 		handle = NULL;
@@ -1560,10 +1664,18 @@ static int ocfs2_symlink(struct inode *dir,
 		goto bail;
 	}
 
-	status = ocfs2_mknod_locked(osb, dir, dentry,
-				    S_IFLNK | S_IRWXUGO, 0,
-				    &new_fe_bh, parent_fe_bh, handle,
-				    &inode, inode_ac);
+	/* We don't use standard VFS wrapper because we don't want vfs_dq_init
+	 * to be called. */
+	if (sb_any_quota_active(osb->sb) &&
+	    osb->sb->dq_op->alloc_inode(inode, 1) == NO_QUOTA) {
+		status = -EDQUOT;
+		goto bail;
+	}
+	did_quota_inode = 1;
+
+	status = ocfs2_mknod_locked(osb, dir, inode, dentry,
+				    0, &new_fe_bh, parent_fe_bh, handle,
+				    inode_ac);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -1576,6 +1688,12 @@ static int ocfs2_symlink(struct inode *dir,
 		u32 offset = 0;
 
 		inode->i_op = &ocfs2_symlink_inode_operations;
+		if (vfs_dq_alloc_space_nodirty(inode,
+		    ocfs2_clusters_to_bytes(osb->sb, 1))) {
+			status = -EDQUOT;
+			goto bail;
+		}
+		did_quota = 1;
 		status = ocfs2_add_inode_data(osb, inode, &offset, 1, 0,
 					      new_fe_bh,
 					      handle, data_ac, NULL,
@@ -1614,6 +1732,15 @@ static int ocfs2_symlink(struct inode *dir,
 		}
 	}
 
+	if (si.enable) {
+		status = ocfs2_init_security_set(handle, inode, new_fe_bh, &si,
+						 xattr_ac, data_ac);
+		if (status < 0) {
+			mlog_errno(status);
+			goto bail;
+		}
+	}
+
 	status = ocfs2_add_entry(handle, dentry, inode,
 				 le64_to_cpu(fe->i_blkno), parent_fe_bh,
 				 de_bh);
@@ -1632,6 +1759,11 @@ static int ocfs2_symlink(struct inode *dir,
 	dentry->d_op = &ocfs2_dentry_ops;
 	d_instantiate(dentry, inode);
 bail:
+	if (status < 0 && did_quota)
+		vfs_dq_free_space_nodirty(inode,
+					ocfs2_clusters_to_bytes(osb->sb, 1));
+	if (status < 0 && did_quota_inode)
+		vfs_dq_free_inode(inode);
 	if (handle)
 		ocfs2_commit_trans(osb, handle);
 
@@ -1640,12 +1772,18 @@ bail:
 	brelse(new_fe_bh);
 	brelse(parent_fe_bh);
 	brelse(de_bh);
+	kfree(si.name);
+	kfree(si.value);
 	if (inode_ac)
 		ocfs2_free_alloc_context(inode_ac);
 	if (data_ac)
 		ocfs2_free_alloc_context(data_ac);
-	if ((status < 0) && inode)
+	if (xattr_ac)
+		ocfs2_free_alloc_context(xattr_ac);
+	if ((status < 0) && inode) {
+		clear_nlink(inode);
 		iput(inode);
+	}
 
 	mlog_exit(status);
 
@@ -1754,16 +1892,14 @@ static int ocfs2_orphan_add(struct ocfs2_super *osb,
 
 	mlog_entry("(inode->i_ino = %lu)\n", inode->i_ino);
 
-	status = ocfs2_read_block(orphan_dir_inode,
-				  OCFS2_I(orphan_dir_inode)->ip_blkno,
-				  &orphan_dir_bh);
+	status = ocfs2_read_inode_block(orphan_dir_inode, &orphan_dir_bh);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
 	}
 
-	status = ocfs2_journal_access(handle, orphan_dir_inode, orphan_dir_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, orphan_dir_inode, orphan_dir_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
@@ -1850,8 +1986,8 @@ int ocfs2_orphan_del(struct ocfs2_super *osb,
 		goto leave;
 	}
 
-	status = ocfs2_journal_access(handle,orphan_dir_inode,  orphan_dir_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle,orphan_dir_inode,  orphan_dir_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;

fs/ocfs2/ocfs2.h

@@ -161,6 +161,7 @@ enum ocfs2_vol_state
 {
 	VOLUME_INIT = 0,
 	VOLUME_MOUNTED,
+	VOLUME_MOUNTED_QUOTAS,
 	VOLUME_DISMOUNTED,
 	VOLUME_DISABLED
 };
@@ -195,6 +196,9 @@ enum ocfs2_mount_options
 	OCFS2_MOUNT_LOCALFLOCKS = 1 << 5, /* No cluster aware user file locks */
 	OCFS2_MOUNT_NOUSERXATTR = 1 << 6, /* No user xattr */
 	OCFS2_MOUNT_INODE64 = 1 << 7,	/* Allow inode numbers > 2^32 */
+	OCFS2_MOUNT_POSIX_ACL = 1 << 8,	/* POSIX access control lists */
+	OCFS2_MOUNT_USRQUOTA = 1 << 9, /* We support user quotas */
+	OCFS2_MOUNT_GRPQUOTA = 1 << 10, /* We support group quotas */
 };
 
 #define OCFS2_OSB_SOFT_RO	0x0001
@@ -205,6 +209,7 @@ enum ocfs2_mount_options
 struct ocfs2_journal;
 struct ocfs2_slot_info;
 struct ocfs2_recovery_map;
+struct ocfs2_quota_recovery;
 struct ocfs2_super
 {
 	struct task_struct *commit_task;
@@ -286,10 +291,11 @@ struct ocfs2_super
 	char *local_alloc_debug_buf;
 #endif
 
-	/* Next two fields are for local node slot recovery during
+	/* Next three fields are for local node slot recovery during
 	 * mount. */
 	int dirty;
 	struct ocfs2_dinode *local_alloc_copy;
+	struct ocfs2_quota_recovery *quota_rec;
 
 	struct ocfs2_alloc_stats alloc_stats;
 	char dev_str[20];		/* "major,minor" of the device */
@@ -333,6 +339,10 @@ struct ocfs2_super
 
 #define OCFS2_SB(sb)	    ((struct ocfs2_super *)(sb)->s_fs_info)
 
+/* Useful typedef for passing around journal access functions */
+typedef int (*ocfs2_journal_access_func)(handle_t *handle, struct inode *inode,
+					 struct buffer_head *bh, int type);
+
 static inline int ocfs2_should_order_data(struct inode *inode)
 {
 	if (!S_ISREG(inode->i_mode))
@@ -376,6 +386,13 @@ static inline int ocfs2_supports_xattr(struct ocfs2_super *osb)
 	return 0;
 }
 
+static inline int ocfs2_meta_ecc(struct ocfs2_super *osb)
+{
+	if (osb->s_feature_incompat & OCFS2_FEATURE_INCOMPAT_META_ECC)
+		return 1;
+	return 0;
+}
+
 /* set / clear functions because cluster events can make these happen
  * in parallel so we want the transitions to be atomic. this also
  * means that any future flags osb_flags must be protected by spinlock
@@ -443,39 +460,19 @@ static inline int ocfs2_uses_extended_slot_map(struct ocfs2_super *osb)
 #define OCFS2_IS_VALID_DINODE(ptr)					\
 	(!strcmp((ptr)->i_signature, OCFS2_INODE_SIGNATURE))
 
-#define OCFS2_RO_ON_INVALID_DINODE(__sb, __di)	do {			\
-	typeof(__di) ____di = (__di);					\
-	ocfs2_error((__sb), 						\
-		"Dinode # %llu has bad signature %.*s",			\
-		(unsigned long long)le64_to_cpu((____di)->i_blkno), 7, 	\
-		(____di)->i_signature);					\
-} while (0)
-
 #define OCFS2_IS_VALID_EXTENT_BLOCK(ptr)				\
 	(!strcmp((ptr)->h_signature, OCFS2_EXTENT_BLOCK_SIGNATURE))
 
-#define OCFS2_RO_ON_INVALID_EXTENT_BLOCK(__sb, __eb)	do {		\
-	typeof(__eb) ____eb = (__eb);					\
-	ocfs2_error((__sb), 						\
-		"Extent Block # %llu has bad signature %.*s",		\
-		(unsigned long long)le64_to_cpu((____eb)->h_blkno), 7,	\
-		(____eb)->h_signature);					\
-} while (0)
-
 #define OCFS2_IS_VALID_GROUP_DESC(ptr)					\
 	(!strcmp((ptr)->bg_signature, OCFS2_GROUP_DESC_SIGNATURE))
 
-#define OCFS2_RO_ON_INVALID_GROUP_DESC(__sb, __gd)	do {		\
-	typeof(__gd) ____gd = (__gd);					\
-		ocfs2_error((__sb),					\
-		"Group Descriptor # %llu has bad signature %.*s",	\
-		(unsigned long long)le64_to_cpu((____gd)->bg_blkno), 7, \
-		(____gd)->bg_signature);				\
-} while (0)
 
 #define OCFS2_IS_VALID_XATTR_BLOCK(ptr)					\
 	(!strcmp((ptr)->xb_signature, OCFS2_XATTR_BLOCK_SIGNATURE))
 
+#define OCFS2_IS_VALID_DIR_TRAILER(ptr)					\
+	(!strcmp((ptr)->db_signature, OCFS2_DIR_TRAILER_SIGNATURE))
+
 static inline unsigned long ino_from_blkno(struct super_block *sb,
 					   u64 blkno)
 {
@@ -632,5 +629,6 @@ static inline s16 ocfs2_get_inode_steal_slot(struct ocfs2_super *osb)
 #define ocfs2_clear_bit ext2_clear_bit
 #define ocfs2_test_bit ext2_test_bit
 #define ocfs2_find_next_zero_bit ext2_find_next_zero_bit
+#define ocfs2_find_next_bit ext2_find_next_bit
 #endif  /* OCFS2_H */
 

fs/ocfs2/ocfs2_fs.h

@@ -65,6 +65,7 @@
 #define OCFS2_EXTENT_BLOCK_SIGNATURE	"EXBLK01"
 #define OCFS2_GROUP_DESC_SIGNATURE      "GROUP01"
 #define OCFS2_XATTR_BLOCK_SIGNATURE	"XATTR01"
+#define OCFS2_DIR_TRAILER_SIGNATURE	"DIRTRL1"
 
 /* Compatibility flags */
 #define OCFS2_HAS_COMPAT_FEATURE(sb,mask)			\
@@ -93,8 +94,11 @@
 					 | OCFS2_FEATURE_INCOMPAT_INLINE_DATA \
 					 | OCFS2_FEATURE_INCOMPAT_EXTENDED_SLOT_MAP \
 					 | OCFS2_FEATURE_INCOMPAT_USERSPACE_STACK \
-					 | OCFS2_FEATURE_INCOMPAT_XATTR)
-#define OCFS2_FEATURE_RO_COMPAT_SUPP	OCFS2_FEATURE_RO_COMPAT_UNWRITTEN
+					 | OCFS2_FEATURE_INCOMPAT_XATTR \
+					 | OCFS2_FEATURE_INCOMPAT_META_ECC)
+#define OCFS2_FEATURE_RO_COMPAT_SUPP	(OCFS2_FEATURE_RO_COMPAT_UNWRITTEN \
+					 | OCFS2_FEATURE_RO_COMPAT_USRQUOTA \
+					 | OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)
 
 /*
  * Heartbeat-only devices are missing journals and other files.  The
@@ -147,6 +151,9 @@
 /* Support for extended attributes */
 #define OCFS2_FEATURE_INCOMPAT_XATTR		0x0200
 
+/* Metadata checksum and error correction */
+#define OCFS2_FEATURE_INCOMPAT_META_ECC		0x0800
+
 /*
  * backup superblock flag is used to indicate that this volume
  * has backup superblocks.
@@ -163,6 +170,12 @@
  */
 #define OCFS2_FEATURE_RO_COMPAT_UNWRITTEN	0x0001
 
+/*
+ * Maintain quota information for this filesystem
+ */
+#define OCFS2_FEATURE_RO_COMPAT_USRQUOTA	0x0002
+#define OCFS2_FEATURE_RO_COMPAT_GRPQUOTA	0x0004
+
 /* The byte offset of the first backup block will be 1G.
  * The following will be 4G, 16G, 64G, 256G and 1T.
  */
@@ -192,6 +205,7 @@
 #define OCFS2_HEARTBEAT_FL	(0x00000200)	/* Heartbeat area */
 #define OCFS2_CHAIN_FL		(0x00000400)	/* Chain allocator */
 #define OCFS2_DEALLOC_FL	(0x00000800)	/* Truncate log */
+#define OCFS2_QUOTA_FL		(0x00001000)	/* Quota file */
 
 /*
  * Flags on ocfs2_dinode.i_dyn_features
@@ -329,13 +343,17 @@ enum {
 #define OCFS2_FIRST_ONLINE_SYSTEM_INODE SLOT_MAP_SYSTEM_INODE
 	HEARTBEAT_SYSTEM_INODE,
 	GLOBAL_BITMAP_SYSTEM_INODE,
-#define OCFS2_LAST_GLOBAL_SYSTEM_INODE GLOBAL_BITMAP_SYSTEM_INODE
+	USER_QUOTA_SYSTEM_INODE,
+	GROUP_QUOTA_SYSTEM_INODE,
+#define OCFS2_LAST_GLOBAL_SYSTEM_INODE GROUP_QUOTA_SYSTEM_INODE
 	ORPHAN_DIR_SYSTEM_INODE,
 	EXTENT_ALLOC_SYSTEM_INODE,
 	INODE_ALLOC_SYSTEM_INODE,
 	JOURNAL_SYSTEM_INODE,
 	LOCAL_ALLOC_SYSTEM_INODE,
 	TRUNCATE_LOG_SYSTEM_INODE,
+	LOCAL_USER_QUOTA_SYSTEM_INODE,
+	LOCAL_GROUP_QUOTA_SYSTEM_INODE,
 	NUM_SYSTEM_INODES
 };
 
@@ -349,6 +367,8 @@ static struct ocfs2_system_inode_info ocfs2_system_inodes[NUM_SYSTEM_INODES] = {
 	[SLOT_MAP_SYSTEM_INODE]			= { "slot_map", 0, S_IFREG | 0644 },
 	[HEARTBEAT_SYSTEM_INODE]		= { "heartbeat", OCFS2_HEARTBEAT_FL, S_IFREG | 0644 },
 	[GLOBAL_BITMAP_SYSTEM_INODE]		= { "global_bitmap", 0, S_IFREG | 0644 },
+	[USER_QUOTA_SYSTEM_INODE]		= { "aquota.user", OCFS2_QUOTA_FL, S_IFREG | 0644 },
+	[GROUP_QUOTA_SYSTEM_INODE]		= { "aquota.group", OCFS2_QUOTA_FL, S_IFREG | 0644 },
 
 	/* Slot-specific system inodes (one copy per slot) */
 	[ORPHAN_DIR_SYSTEM_INODE]		= { "orphan_dir:%04d", 0, S_IFDIR | 0755 },
@@ -356,7 +376,9 @@ static struct ocfs2_system_inode_info ocfs2_system_inodes[NUM_SYSTEM_INODES] = {
 	[INODE_ALLOC_SYSTEM_INODE]		= { "inode_alloc:%04d", OCFS2_BITMAP_FL | OCFS2_CHAIN_FL, S_IFREG | 0644 },
 	[JOURNAL_SYSTEM_INODE]			= { "journal:%04d", OCFS2_JOURNAL_FL, S_IFREG | 0644 },
 	[LOCAL_ALLOC_SYSTEM_INODE]		= { "local_alloc:%04d", OCFS2_BITMAP_FL | OCFS2_LOCAL_ALLOC_FL, S_IFREG | 0644 },
-	[TRUNCATE_LOG_SYSTEM_INODE]		= { "truncate_log:%04d", OCFS2_DEALLOC_FL, S_IFREG | 0644 }
+	[TRUNCATE_LOG_SYSTEM_INODE]		= { "truncate_log:%04d", OCFS2_DEALLOC_FL, S_IFREG | 0644 },
+	[LOCAL_USER_QUOTA_SYSTEM_INODE]		= { "aquota.user:%04d", OCFS2_QUOTA_FL, S_IFREG | 0644 },
+	[LOCAL_GROUP_QUOTA_SYSTEM_INODE]	= { "aquota.group:%04d", OCFS2_QUOTA_FL, S_IFREG | 0644 },
 };
 
 /* Parameter passed from mount.ocfs2 to module */
@@ -409,6 +431,22 @@ static unsigned char ocfs2_type_by_mode[S_IFMT >> S_SHIFT] = {
  */
 #define OCFS2_RAW_SB(dinode)		(&((dinode)->id2.i_super))
 
+/*
+ * Block checking structure.  This is used in metadata to validate the
+ * contents.  If OCFS2_FEATURE_INCOMPAT_META_ECC is not set, it is all
+ * zeros.
+ */
+struct ocfs2_block_check {
+/*00*/	__le32 bc_crc32e;	/* 802.3 Ethernet II CRC32 */
+	__le16 bc_ecc;		/* Single-error-correction parity vector.
+				   This is a simple Hamming code dependant
+				   on the blocksize.  OCFS2's maximum
+				   blocksize, 4K, requires 16 parity bits,
+				   so we fit in __le16. */
+	__le16 bc_reserved1;
+/*08*/
+};
+
 /*
  * On disk extent record for OCFS2
  * It describes a range of clusters on disk.
@@ -496,7 +534,7 @@ struct ocfs2_truncate_log {
 struct ocfs2_extent_block
 {
 /*00*/	__u8 h_signature[8];		/* Signature for verification */
-	__le64 h_reserved1;
+	struct ocfs2_block_check h_check;	/* Error checking */
 /*10*/	__le16 h_suballoc_slot;		/* Slot suballocator this
 					   extent_header belongs to */
 	__le16 h_suballoc_bit;		/* Bit offset in suballocator
@@ -666,7 +704,8 @@ struct ocfs2_dinode {
 					   was set in i_flags */
 	__le16 i_dyn_features;
 	__le64 i_xattr_loc;
-/*80*/	__le64 i_reserved2[7];
+/*80*/	struct ocfs2_block_check i_check;	/* Error checking */
+/*88*/	__le64 i_reserved2[6];
 /*B8*/	union {
 		__le64 i_pad1;		/* Generic way to refer to this
 					   64bit union */
@@ -714,6 +753,34 @@ struct ocfs2_dir_entry {
 /* Actual on-disk length specified by rec_len */
 } __attribute__ ((packed));
 
+/*
+ * Per-block record for the unindexed directory btree. This is carefully
+ * crafted so that the rec_len and name_len records of an ocfs2_dir_entry are
+ * mirrored. That way, the directory manipulation code needs a minimal amount
+ * of update.
+ *
+ * NOTE: Keep this structure aligned to a multiple of 4 bytes.
+ */
+struct ocfs2_dir_block_trailer {
+/*00*/	__le64		db_compat_inode;	/* Always zero. Was inode */
+
+	__le16		db_compat_rec_len;	/* Backwards compatible with
+						 * ocfs2_dir_entry. */
+	__u8		db_compat_name_len;	/* Always zero. Was name_len */
+	__u8		db_reserved0;
+	__le16		db_reserved1;
+	__le16		db_free_rec_len;	/* Size of largest empty hole
+						 * in this block. (unused) */
+/*10*/	__u8		db_signature[8];	/* Signature for verification */
+	__le64		db_reserved2;
+	__le64		db_free_next;		/* Next block in list (unused) */
+/*20*/	__le64		db_blkno;		/* Offset on disk, in blocks */
+	__le64		db_parent_dinode;	/* dinode which owns me, in
+						   blocks */
+/*30*/	struct ocfs2_block_check db_check;	/* Error checking */
+/*40*/
+};
+
 /*
  * On disk allocator group structure for OCFS2
  */
@@ -733,7 +800,8 @@ struct ocfs2_group_desc
 /*20*/	__le64   bg_parent_dinode;       /* dinode which owns me, in
 					   blocks */
 	__le64   bg_blkno;               /* Offset on disk, in blocks */
-/*30*/	__le64   bg_reserved2[2];
+/*30*/	struct ocfs2_block_check bg_check;	/* Error checking */
+	__le64   bg_reserved2;
 /*40*/	__u8    bg_bitmap[0];
 };
 
@@ -776,7 +844,12 @@ struct ocfs2_xattr_header {
 						   in this extent record,
 						   only valid in the first
 						   bucket. */
-	__le64  xh_csum;
+	struct ocfs2_block_check xh_check;	/* Error checking
+						   (Note, this is only
+						    used for xattr
+						    buckets.  A block uses
+						    xb_check and sets
+						    this field to zero.) */
 	struct ocfs2_xattr_entry xh_entries[0]; /* xattr entry list. */
 };
 
@@ -827,7 +900,7 @@ struct ocfs2_xattr_block {
 					block group */
 	__le32	xb_fs_generation;    /* Must match super block */
 /*10*/	__le64	xb_blkno;            /* Offset on disk, in blocks */
-	__le64	xb_csum;
+	struct ocfs2_block_check xb_check;	/* Error checking */
 /*20*/	__le16	xb_flags;            /* Indicates whether this block contains
 					real xattr or a xattr tree. */
 	__le16	xb_reserved0;
@@ -868,6 +941,128 @@ static inline int ocfs2_xattr_get_type(struct ocfs2_xattr_entry *xe)
 	return xe->xe_type & OCFS2_XATTR_TYPE_MASK;
 }
 
+/*
+ *  On disk structures for global quota file
+ */
+
+/* Magic numbers and known versions for global quota files */
+#define OCFS2_GLOBAL_QMAGICS {\
+	0x0cf52470, /* USRQUOTA */ \
+	0x0cf52471  /* GRPQUOTA */ \
+}
+
+#define OCFS2_GLOBAL_QVERSIONS {\
+	0, \
+	0, \
+}
+
+
+/* Each block of each quota file has a certain fixed number of bytes reserved
+ * for OCFS2 internal use at its end. OCFS2 can use it for things like
+ * checksums, etc. */
+#define OCFS2_QBLK_RESERVED_SPACE 8
+
+/* Generic header of all quota files */
+struct ocfs2_disk_dqheader {
+	__le32 dqh_magic;	/* Magic number identifying file */
+	__le32 dqh_version;	/* Quota format version */
+};
+
+#define OCFS2_GLOBAL_INFO_OFF (sizeof(struct ocfs2_disk_dqheader))
+
+/* Information header of global quota file (immediately follows the generic
+ * header) */
+struct ocfs2_global_disk_dqinfo {
+/*00*/	__le32 dqi_bgrace;	/* Grace time for space softlimit excess */
+	__le32 dqi_igrace;	/* Grace time for inode softlimit excess */
+	__le32 dqi_syncms;	/* Time after which we sync local changes to
+				 * global quota file */
+	__le32 dqi_blocks;	/* Number of blocks in quota file */
+/*10*/	__le32 dqi_free_blk;	/* First free block in quota file */
+	__le32 dqi_free_entry;	/* First block with free dquot entry in quota
+				 * file */
+};
+
+/* Structure with global user / group information. We reserve some space
+ * for future use. */
+struct ocfs2_global_disk_dqblk {
+/*00*/	__le32 dqb_id;          /* ID the structure belongs to */
+	__le32 dqb_use_count;   /* Number of nodes having reference to this structure */
+	__le64 dqb_ihardlimit;  /* absolute limit on allocated inodes */
+/*10*/	__le64 dqb_isoftlimit;  /* preferred inode limit */
+	__le64 dqb_curinodes;   /* current # allocated inodes */
+/*20*/	__le64 dqb_bhardlimit;  /* absolute limit on disk space */
+	__le64 dqb_bsoftlimit;  /* preferred limit on disk space */
+/*30*/	__le64 dqb_curspace;    /* current space occupied */
+	__le64 dqb_btime;       /* time limit for excessive disk use */
+/*40*/	__le64 dqb_itime;       /* time limit for excessive inode use */
+	__le64 dqb_pad1;
+/*50*/	__le64 dqb_pad2;
+};
+
+/*
+ *  On-disk structures for local quota file
+ */
+
+/* Magic numbers and known versions for local quota files */
+#define OCFS2_LOCAL_QMAGICS {\
+	0x0cf524c0, /* USRQUOTA */ \
+	0x0cf524c1  /* GRPQUOTA */ \
+}
+
+#define OCFS2_LOCAL_QVERSIONS {\
+	0, \
+	0, \
+}
+
+/* Quota flags in dqinfo header */
+#define OLQF_CLEAN	0x0001	/* Quota file is empty (this should be after\
+				 * quota has been cleanly turned off) */
+
+#define OCFS2_LOCAL_INFO_OFF (sizeof(struct ocfs2_disk_dqheader))
+
+/* Information header of local quota file (immediately follows the generic
+ * header) */
+struct ocfs2_local_disk_dqinfo {
+	__le32 dqi_flags;	/* Flags for quota file */
+	__le32 dqi_chunks;	/* Number of chunks of quota structures
+				 * with a bitmap */
+	__le32 dqi_blocks;	/* Number of blocks allocated for quota file */
+};
+
+/* Header of one chunk of a quota file */
+struct ocfs2_local_disk_chunk {
+	__le32 dqc_free;	/* Number of free entries in the bitmap */
+	u8 dqc_bitmap[0];	/* Bitmap of entries in the corresponding
+				 * chunk of quota file */
+};
+
+/* One entry in local quota file */
+struct ocfs2_local_disk_dqblk {
+/*00*/	__le64 dqb_id;		/* id this quota applies to */
+	__le64 dqb_spacemod;	/* Change in the amount of used space */
+/*10*/	__le64 dqb_inodemod;	/* Change in the amount of used inodes */
+};
+
+
+/*
+ * The quota trailer lives at the end of each quota block.
+ */
+
+struct ocfs2_disk_dqtrailer {
+/*00*/	struct ocfs2_block_check dq_check;	/* Error checking */
+/*08*/	/* Cannot be larger than OCFS2_QBLK_RESERVED_SPACE */
+};
+
+static inline struct ocfs2_disk_dqtrailer *ocfs2_block_dqtrailer(int blocksize,
+								 void *buf)
+{
+	char *ptr = buf;
+	ptr += blocksize - OCFS2_QBLK_RESERVED_SPACE;
+
+	return (struct ocfs2_disk_dqtrailer *)ptr;
+}
+
 #ifdef __KERNEL__
 static inline int ocfs2_fast_symlink_chars(struct super_block *sb)
 {

fs/ocfs2/ocfs2_jbd_compat.h

@@ -1,82 +0,0 @@
-/* -*- mode: c; c-basic-offset: 8; -*-
- * vim: noexpandtab sw=8 ts=8 sts=0:
- *
- * ocfs2_jbd_compat.h
- *
- * Compatibility defines for JBD.
- *
- * Copyright (C) 2008 Oracle.  All rights reserved.
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public
- * License version 2 as published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
- * General Public License for more details.
- */
-
-#ifndef OCFS2_JBD_COMPAT_H
-#define OCFS2_JBD_COMPAT_H
-
-#ifndef CONFIG_OCFS2_COMPAT_JBD
-# error Should not have been included
-#endif
-
-struct jbd2_inode {
-	unsigned int dummy;
-};
-
-#define JBD2_BARRIER			JFS_BARRIER
-#define JBD2_DEFAULT_MAX_COMMIT_AGE	JBD_DEFAULT_MAX_COMMIT_AGE
-
-#define jbd2_journal_ack_err			journal_ack_err
-#define jbd2_journal_clear_err			journal_clear_err
-#define jbd2_journal_destroy			journal_destroy
-#define jbd2_journal_dirty_metadata		journal_dirty_metadata
-#define jbd2_journal_errno			journal_errno
-#define jbd2_journal_extend			journal_extend
-#define jbd2_journal_flush			journal_flush
-#define jbd2_journal_force_commit		journal_force_commit
-#define jbd2_journal_get_write_access		journal_get_write_access
-#define jbd2_journal_get_undo_access		journal_get_undo_access
-#define jbd2_journal_init_inode			journal_init_inode
-#define jbd2_journal_invalidatepage		journal_invalidatepage
-#define jbd2_journal_load			journal_load
-#define jbd2_journal_lock_updates		journal_lock_updates
-#define jbd2_journal_restart			journal_restart
-#define jbd2_journal_start			journal_start
-#define jbd2_journal_start_commit		journal_start_commit
-#define jbd2_journal_stop			journal_stop
-#define jbd2_journal_try_to_free_buffers	journal_try_to_free_buffers
-#define jbd2_journal_unlock_updates		journal_unlock_updates
-#define jbd2_journal_wipe			journal_wipe
-#define jbd2_log_wait_commit			log_wait_commit
-
-static inline int jbd2_journal_file_inode(handle_t *handle,
-					  struct jbd2_inode *inode)
-{
-	return 0;
-}
-
-static inline int jbd2_journal_begin_ordered_truncate(struct jbd2_inode *inode,
-						      loff_t new_size)
-{
-	return 0;
-}
-
-static inline void jbd2_journal_init_jbd_inode(struct jbd2_inode *jinode,
-					       struct inode *inode)
-{
-	return;
-}
-
-static inline void jbd2_journal_release_jbd_inode(journal_t *journal,
-						  struct jbd2_inode *jinode)
-{
-	return;
-}
-
-
-#endif  /* OCFS2_JBD_COMPAT_H */

fs/ocfs2/ocfs2_lockid.h

@@ -46,6 +46,7 @@ enum ocfs2_lock_type {
 	OCFS2_LOCK_TYPE_DENTRY,
 	OCFS2_LOCK_TYPE_OPEN,
 	OCFS2_LOCK_TYPE_FLOCK,
+	OCFS2_LOCK_TYPE_QINFO,
 	OCFS2_NUM_LOCK_TYPES
 };
 
@@ -77,6 +78,9 @@ static inline char ocfs2_lock_type_char(enum ocfs2_lock_type type)
 		case OCFS2_LOCK_TYPE_FLOCK:
 			c = 'F';
 			break;
+		case OCFS2_LOCK_TYPE_QINFO:
+			c = 'Q';
+			break;
 		default:
 			c = '\0';
 	}
@@ -95,6 +99,7 @@ static char *ocfs2_lock_type_strings[] = {
 	[OCFS2_LOCK_TYPE_DENTRY] = "Dentry",
 	[OCFS2_LOCK_TYPE_OPEN] = "Open",
 	[OCFS2_LOCK_TYPE_FLOCK] = "Flock",
+	[OCFS2_LOCK_TYPE_QINFO] = "Quota",
 };
 
 static inline const char *ocfs2_lock_type_string(enum ocfs2_lock_type type)

fs/ocfs2/quota.h

@@ -0,0 +1,119 @@
+/*
+ * quota.h for OCFS2
+ *
+ * On disk quota structures for local and global quota file, in-memory
+ * structures.
+ *
+ */
+
+#ifndef _OCFS2_QUOTA_H
+#define _OCFS2_QUOTA_H
+
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/quota.h>
+#include <linux/list.h>
+#include <linux/dqblk_qtree.h>
+
+#include "ocfs2.h"
+
+/* Common stuff */
+/* id number of quota format */
+#define QFMT_OCFS2 3
+
+/*
+ * In-memory structures
+ */
+struct ocfs2_dquot {
+	struct dquot dq_dquot;	/* Generic VFS dquot */
+	loff_t dq_local_off;	/* Offset in the local quota file */
+	struct ocfs2_quota_chunk *dq_chunk;	/* Chunk dquot is in */
+	unsigned int dq_use_count;	/* Number of nodes having reference to this entry in global quota file */
+	s64 dq_origspace;	/* Last globally synced space usage */
+	s64 dq_originodes;	/* Last globally synced inode usage */
+};
+
+/* Description of one chunk to recover in memory */
+struct ocfs2_recovery_chunk {
+	struct list_head rc_list;	/* List of chunks */
+	int rc_chunk;			/* Chunk number */
+	unsigned long *rc_bitmap;	/* Bitmap of entries to recover */
+};
+
+struct ocfs2_quota_recovery {
+	struct list_head r_list[MAXQUOTAS];	/* List of chunks to recover */
+};
+
+/* In-memory structure with quota header information */
+struct ocfs2_mem_dqinfo {
+	unsigned int dqi_type;		/* Quota type this structure describes */
+	unsigned int dqi_chunks;	/* Number of chunks in local quota file */
+	unsigned int dqi_blocks;	/* Number of blocks allocated for local quota file */
+	unsigned int dqi_syncms;	/* How often should we sync with other nodes */
+	unsigned int dqi_syncjiff;	/* Precomputed dqi_syncms in jiffies */
+	struct list_head dqi_chunk;	/* List of chunks */
+	struct inode *dqi_gqinode;	/* Global quota file inode */
+	struct ocfs2_lock_res dqi_gqlock;	/* Lock protecting quota information structure */
+	struct buffer_head *dqi_gqi_bh;	/* Buffer head with global quota file inode - set only if inode lock is obtained */
+	int dqi_gqi_count;		/* Number of holders of dqi_gqi_bh */
+	struct buffer_head *dqi_lqi_bh;	/* Buffer head with local quota file inode */
+	struct buffer_head *dqi_ibh;	/* Buffer with information header */
+	struct qtree_mem_dqinfo dqi_gi;	/* Info about global file */
+	struct delayed_work dqi_sync_work;	/* Work for syncing dquots */
+	struct ocfs2_quota_recovery *dqi_rec;	/* Pointer to recovery
+						 * information, in case we
+						 * enable quotas on file
+						 * needing it */
+};
+
+static inline struct ocfs2_dquot *OCFS2_DQUOT(struct dquot *dquot)
+{
+	return container_of(dquot, struct ocfs2_dquot, dq_dquot);
+}
+
+struct ocfs2_quota_chunk {
+	struct list_head qc_chunk;	/* List of quotafile chunks */
+	int qc_num;			/* Number of quota chunk */
+	struct buffer_head *qc_headerbh;	/* Buffer head with chunk header */
+};
+
+extern struct kmem_cache *ocfs2_dquot_cachep;
+extern struct kmem_cache *ocfs2_qf_chunk_cachep;
+
+extern struct qtree_fmt_operations ocfs2_global_ops;
+
+struct ocfs2_quota_recovery *ocfs2_begin_quota_recovery(
+				struct ocfs2_super *osb, int slot_num);
+int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
+				struct ocfs2_quota_recovery *rec,
+				int slot_num);
+void ocfs2_free_quota_recovery(struct ocfs2_quota_recovery *rec);
+ssize_t ocfs2_quota_read(struct super_block *sb, int type, char *data,
+			 size_t len, loff_t off);
+ssize_t ocfs2_quota_write(struct super_block *sb, int type,
+			  const char *data, size_t len, loff_t off);
+int ocfs2_global_read_info(struct super_block *sb, int type);
+int ocfs2_global_write_info(struct super_block *sb, int type);
+int ocfs2_global_read_dquot(struct dquot *dquot);
+int __ocfs2_sync_dquot(struct dquot *dquot, int freeing);
+static inline int ocfs2_sync_dquot(struct dquot *dquot)
+{
+	return __ocfs2_sync_dquot(dquot, 0);
+}
+static inline int ocfs2_global_release_dquot(struct dquot *dquot)
+{
+	return __ocfs2_sync_dquot(dquot, 1);
+}
+
+int ocfs2_lock_global_qf(struct ocfs2_mem_dqinfo *oinfo, int ex);
+void ocfs2_unlock_global_qf(struct ocfs2_mem_dqinfo *oinfo, int ex);
+int ocfs2_read_quota_block(struct inode *inode, u64 v_block,
+			   struct buffer_head **bh);
+
+extern struct dquot_operations ocfs2_quota_operations;
+extern struct quota_format_type ocfs2_quota_format;
+
+int ocfs2_quota_setup(void);
+void ocfs2_quota_shutdown(void);
+
+#endif /* _OCFS2_QUOTA_H */

fs/ocfs2/quota_global.c

@@ -0,0 +1,1025 @@
+/*
+ *  Implementation of operations over global quota file
+ */
+#include <linux/spinlock.h>
+#include <linux/fs.h>
+#include <linux/quota.h>
+#include <linux/quotaops.h>
+#include <linux/dqblk_qtree.h>
+#include <linux/jiffies.h>
+#include <linux/writeback.h>
+#include <linux/workqueue.h>
+
+#define MLOG_MASK_PREFIX ML_QUOTA
+#include <cluster/masklog.h>
+
+#include "ocfs2_fs.h"
+#include "ocfs2.h"
+#include "alloc.h"
+#include "blockcheck.h"
+#include "inode.h"
+#include "journal.h"
+#include "file.h"
+#include "sysfile.h"
+#include "dlmglue.h"
+#include "uptodate.h"
+#include "quota.h"
+
+static struct workqueue_struct *ocfs2_quota_wq = NULL;
+
+static void qsync_work_fn(struct work_struct *work);
+
+static void ocfs2_global_disk2memdqb(struct dquot *dquot, void *dp)
+{
+	struct ocfs2_global_disk_dqblk *d = dp;
+	struct mem_dqblk *m = &dquot->dq_dqb;
+
+	/* Update from disk only entries not set by the admin */
+	if (!test_bit(DQ_LASTSET_B + QIF_ILIMITS_B, &dquot->dq_flags)) {
+		m->dqb_ihardlimit = le64_to_cpu(d->dqb_ihardlimit);
+		m->dqb_isoftlimit = le64_to_cpu(d->dqb_isoftlimit);
+	}
+	if (!test_bit(DQ_LASTSET_B + QIF_INODES_B, &dquot->dq_flags))
+		m->dqb_curinodes = le64_to_cpu(d->dqb_curinodes);
+	if (!test_bit(DQ_LASTSET_B + QIF_BLIMITS_B, &dquot->dq_flags)) {
+		m->dqb_bhardlimit = le64_to_cpu(d->dqb_bhardlimit);
+		m->dqb_bsoftlimit = le64_to_cpu(d->dqb_bsoftlimit);
+	}
+	if (!test_bit(DQ_LASTSET_B + QIF_SPACE_B, &dquot->dq_flags))
+		m->dqb_curspace = le64_to_cpu(d->dqb_curspace);
+	if (!test_bit(DQ_LASTSET_B + QIF_BTIME_B, &dquot->dq_flags))
+		m->dqb_btime = le64_to_cpu(d->dqb_btime);
+	if (!test_bit(DQ_LASTSET_B + QIF_ITIME_B, &dquot->dq_flags))
+		m->dqb_itime = le64_to_cpu(d->dqb_itime);
+	OCFS2_DQUOT(dquot)->dq_use_count = le32_to_cpu(d->dqb_use_count);
+}
+
+static void ocfs2_global_mem2diskdqb(void *dp, struct dquot *dquot)
+{
+	struct ocfs2_global_disk_dqblk *d = dp;
+	struct mem_dqblk *m = &dquot->dq_dqb;
+
+	d->dqb_id = cpu_to_le32(dquot->dq_id);
+	d->dqb_use_count = cpu_to_le32(OCFS2_DQUOT(dquot)->dq_use_count);
+	d->dqb_ihardlimit = cpu_to_le64(m->dqb_ihardlimit);
+	d->dqb_isoftlimit = cpu_to_le64(m->dqb_isoftlimit);
+	d->dqb_curinodes = cpu_to_le64(m->dqb_curinodes);
+	d->dqb_bhardlimit = cpu_to_le64(m->dqb_bhardlimit);
+	d->dqb_bsoftlimit = cpu_to_le64(m->dqb_bsoftlimit);
+	d->dqb_curspace = cpu_to_le64(m->dqb_curspace);
+	d->dqb_btime = cpu_to_le64(m->dqb_btime);
+	d->dqb_itime = cpu_to_le64(m->dqb_itime);
+}
+
+static int ocfs2_global_is_id(void *dp, struct dquot *dquot)
+{
+	struct ocfs2_global_disk_dqblk *d = dp;
+	struct ocfs2_mem_dqinfo *oinfo =
+			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
+
+	if (qtree_entry_unused(&oinfo->dqi_gi, dp))
+		return 0;
+	return le32_to_cpu(d->dqb_id) == dquot->dq_id;
+}
+
+struct qtree_fmt_operations ocfs2_global_ops = {
+	.mem2disk_dqblk = ocfs2_global_mem2diskdqb,
+	.disk2mem_dqblk = ocfs2_global_disk2memdqb,
+	.is_id = ocfs2_global_is_id,
+};
+
+static int ocfs2_validate_quota_block(struct super_block *sb,
+				      struct buffer_head *bh)
+{
+	struct ocfs2_disk_dqtrailer *dqt =
+		ocfs2_block_dqtrailer(sb->s_blocksize, bh->b_data);
+
+	mlog(0, "Validating quota block %llu\n",
+	     (unsigned long long)bh->b_blocknr);
+
+	BUG_ON(!buffer_uptodate(bh));
+
+	/*
+	 * If the ecc fails, we return the error but otherwise
+	 * leave the filesystem running.  We know any error is
+	 * local to this block.
+	 */
+	return ocfs2_validate_meta_ecc(sb, bh->b_data, &dqt->dq_check);
+}
+
+int ocfs2_read_quota_block(struct inode *inode, u64 v_block,
+			   struct buffer_head **bh)
+{
+	int rc = 0;
+	struct buffer_head *tmp = *bh;
+
+	rc = ocfs2_read_virt_blocks(inode, v_block, 1, &tmp, 0,
+				    ocfs2_validate_quota_block);
+	if (rc)
+		mlog_errno(rc);
+
+	/* If ocfs2_read_virt_blocks() got us a new bh, pass it up. */
+	if (!rc && !*bh)
+		*bh = tmp;
+
+	return rc;
+}
+
+static int ocfs2_get_quota_block(struct inode *inode, int block,
+				 struct buffer_head **bh)
+{
+	u64 pblock, pcount;
+	int err;
+
+	down_read(&OCFS2_I(inode)->ip_alloc_sem);
+	err = ocfs2_extent_map_get_blocks(inode, block, &pblock, &pcount, NULL);
+	up_read(&OCFS2_I(inode)->ip_alloc_sem);
+	if (err) {
+		mlog_errno(err);
+		return err;
+	}
+	*bh = sb_getblk(inode->i_sb, pblock);
+	if (!*bh) {
+		err = -EIO;
+		mlog_errno(err);
+	}
+	return err;;
+}
+
+/* Read data from global quotafile - avoid pagecache and such because we cannot
+ * afford acquiring the locks... We use quota cluster lock to serialize
+ * operations. Caller is responsible for acquiring it. */
+ssize_t ocfs2_quota_read(struct super_block *sb, int type, char *data,
+			 size_t len, loff_t off)
+{
+	struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv;
+	struct inode *gqinode = oinfo->dqi_gqinode;
+	loff_t i_size = i_size_read(gqinode);
+	int offset = off & (sb->s_blocksize - 1);
+	sector_t blk = off >> sb->s_blocksize_bits;
+	int err = 0;
+	struct buffer_head *bh;
+	size_t toread, tocopy;
+
+	if (off > i_size)
+		return 0;
+	if (off + len > i_size)
+		len = i_size - off;
+	toread = len;
+	while (toread > 0) {
+		tocopy = min_t(size_t, (sb->s_blocksize - offset), toread);
+		bh = NULL;
+		err = ocfs2_read_quota_block(gqinode, blk, &bh);
+		if (err) {
+			mlog_errno(err);
+			return err;
+		}
+		memcpy(data, bh->b_data + offset, tocopy);
+		brelse(bh);
+		offset = 0;
+		toread -= tocopy;
+		data += tocopy;
+		blk++;
+	}
+	return len;
+}
+
+/* Write to quotafile (we know the transaction is already started and has
+ * enough credits) */
+ssize_t ocfs2_quota_write(struct super_block *sb, int type,
+			  const char *data, size_t len, loff_t off)
+{
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	struct inode *gqinode = oinfo->dqi_gqinode;
+	int offset = off & (sb->s_blocksize - 1);
+	sector_t blk = off >> sb->s_blocksize_bits;
+	int err = 0, new = 0, ja_type;
+	struct buffer_head *bh = NULL;
+	handle_t *handle = journal_current_handle();
+
+	if (!handle) {
+		mlog(ML_ERROR, "Quota write (off=%llu, len=%llu) cancelled "
+		     "because transaction was not started.\n",
+		     (unsigned long long)off, (unsigned long long)len);
+		return -EIO;
+	}
+	if (len > sb->s_blocksize - OCFS2_QBLK_RESERVED_SPACE - offset) {
+		WARN_ON(1);
+		len = sb->s_blocksize - OCFS2_QBLK_RESERVED_SPACE - offset;
+	}
+
+	mutex_lock_nested(&gqinode->i_mutex, I_MUTEX_QUOTA);
+	if (gqinode->i_size < off + len) {
+		down_write(&OCFS2_I(gqinode)->ip_alloc_sem);
+		err = ocfs2_extend_no_holes(gqinode, off + len, off);
+		up_write(&OCFS2_I(gqinode)->ip_alloc_sem);
+		if (err < 0)
+			goto out;
+		err = ocfs2_simple_size_update(gqinode,
+					       oinfo->dqi_gqi_bh,
+					       off + len);
+		if (err < 0)
+			goto out;
+		new = 1;
+	}
+	/* Not rewriting whole block? */
+	if ((offset || len < sb->s_blocksize - OCFS2_QBLK_RESERVED_SPACE) &&
+	    !new) {
+		err = ocfs2_read_quota_block(gqinode, blk, &bh);
+		ja_type = OCFS2_JOURNAL_ACCESS_WRITE;
+	} else {
+		err = ocfs2_get_quota_block(gqinode, blk, &bh);
+		ja_type = OCFS2_JOURNAL_ACCESS_CREATE;
+	}
+	if (err) {
+		mlog_errno(err);
+		return err;
+	}
+	lock_buffer(bh);
+	if (new)
+		memset(bh->b_data, 0, sb->s_blocksize);
+	memcpy(bh->b_data + offset, data, len);
+	flush_dcache_page(bh->b_page);
+	set_buffer_uptodate(bh);
+	unlock_buffer(bh);
+	ocfs2_set_buffer_uptodate(gqinode, bh);
+	err = ocfs2_journal_access_dq(handle, gqinode, bh, ja_type);
+	if (err < 0) {
+		brelse(bh);
+		goto out;
+	}
+	err = ocfs2_journal_dirty(handle, bh);
+	brelse(bh);
+	if (err < 0)
+		goto out;
+out:
+	if (err) {
+		mutex_unlock(&gqinode->i_mutex);
+		mlog_errno(err);
+		return err;
+	}
+	gqinode->i_version++;
+	ocfs2_mark_inode_dirty(handle, gqinode, oinfo->dqi_gqi_bh);
+	mutex_unlock(&gqinode->i_mutex);
+	return len;
+}
+
+int ocfs2_lock_global_qf(struct ocfs2_mem_dqinfo *oinfo, int ex)
+{
+	int status;
+	struct buffer_head *bh = NULL;
+
+	status = ocfs2_inode_lock(oinfo->dqi_gqinode, &bh, ex);
+	if (status < 0)
+		return status;
+	spin_lock(&dq_data_lock);
+	if (!oinfo->dqi_gqi_count++)
+		oinfo->dqi_gqi_bh = bh;
+	else
+		WARN_ON(bh != oinfo->dqi_gqi_bh);
+	spin_unlock(&dq_data_lock);
+	return 0;
+}
+
+void ocfs2_unlock_global_qf(struct ocfs2_mem_dqinfo *oinfo, int ex)
+{
+	ocfs2_inode_unlock(oinfo->dqi_gqinode, ex);
+	brelse(oinfo->dqi_gqi_bh);
+	spin_lock(&dq_data_lock);
+	if (!--oinfo->dqi_gqi_count)
+		oinfo->dqi_gqi_bh = NULL;
+	spin_unlock(&dq_data_lock);
+}
+
+/* Read information header from global quota file */
+int ocfs2_global_read_info(struct super_block *sb, int type)
+{
+	struct inode *gqinode = NULL;
+	unsigned int ino[MAXQUOTAS] = { USER_QUOTA_SYSTEM_INODE,
+					GROUP_QUOTA_SYSTEM_INODE };
+	struct ocfs2_global_disk_dqinfo dinfo;
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	int status;
+
+	mlog_entry_void();
+
+	/* Read global header */
+	gqinode = ocfs2_get_system_file_inode(OCFS2_SB(sb), ino[type],
+			OCFS2_INVALID_SLOT);
+	if (!gqinode) {
+		mlog(ML_ERROR, "failed to get global quota inode (type=%d)\n",
+			type);
+		status = -EINVAL;
+		goto out_err;
+	}
+	oinfo->dqi_gi.dqi_sb = sb;
+	oinfo->dqi_gi.dqi_type = type;
+	ocfs2_qinfo_lock_res_init(&oinfo->dqi_gqlock, oinfo);
+	oinfo->dqi_gi.dqi_entry_size = sizeof(struct ocfs2_global_disk_dqblk);
+	oinfo->dqi_gi.dqi_ops = &ocfs2_global_ops;
+	oinfo->dqi_gqi_bh = NULL;
+	oinfo->dqi_gqi_count = 0;
+	oinfo->dqi_gqinode = gqinode;
+	status = ocfs2_lock_global_qf(oinfo, 0);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_err;
+	}
+	status = sb->s_op->quota_read(sb, type, (char *)&dinfo,
+				      sizeof(struct ocfs2_global_disk_dqinfo),
+				      OCFS2_GLOBAL_INFO_OFF);
+	ocfs2_unlock_global_qf(oinfo, 0);
+	if (status != sizeof(struct ocfs2_global_disk_dqinfo)) {
+		mlog(ML_ERROR, "Cannot read global quota info (%d).\n",
+		     status);
+		if (status >= 0)
+			status = -EIO;
+		mlog_errno(status);
+		goto out_err;
+	}
+	info->dqi_bgrace = le32_to_cpu(dinfo.dqi_bgrace);
+	info->dqi_igrace = le32_to_cpu(dinfo.dqi_igrace);
+	oinfo->dqi_syncms = le32_to_cpu(dinfo.dqi_syncms);
+	oinfo->dqi_syncjiff = msecs_to_jiffies(oinfo->dqi_syncms);
+	oinfo->dqi_gi.dqi_blocks = le32_to_cpu(dinfo.dqi_blocks);
+	oinfo->dqi_gi.dqi_free_blk = le32_to_cpu(dinfo.dqi_free_blk);
+	oinfo->dqi_gi.dqi_free_entry = le32_to_cpu(dinfo.dqi_free_entry);
+	oinfo->dqi_gi.dqi_blocksize_bits = sb->s_blocksize_bits;
+	oinfo->dqi_gi.dqi_usable_bs = sb->s_blocksize -
+						OCFS2_QBLK_RESERVED_SPACE;
+	oinfo->dqi_gi.dqi_qtree_depth = qtree_depth(&oinfo->dqi_gi);
+	INIT_DELAYED_WORK(&oinfo->dqi_sync_work, qsync_work_fn);
+	queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work,
+			   oinfo->dqi_syncjiff);
+
+out_err:
+	mlog_exit(status);
+	return status;
+}
+
+/* Write information to global quota file. Expects exlusive lock on quota
+ * file inode and quota info */
+static int __ocfs2_global_write_info(struct super_block *sb, int type)
+{
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	struct ocfs2_global_disk_dqinfo dinfo;
+	ssize_t size;
+
+	spin_lock(&dq_data_lock);
+	info->dqi_flags &= ~DQF_INFO_DIRTY;
+	dinfo.dqi_bgrace = cpu_to_le32(info->dqi_bgrace);
+	dinfo.dqi_igrace = cpu_to_le32(info->dqi_igrace);
+	spin_unlock(&dq_data_lock);
+	dinfo.dqi_syncms = cpu_to_le32(oinfo->dqi_syncms);
+	dinfo.dqi_blocks = cpu_to_le32(oinfo->dqi_gi.dqi_blocks);
+	dinfo.dqi_free_blk = cpu_to_le32(oinfo->dqi_gi.dqi_free_blk);
+	dinfo.dqi_free_entry = cpu_to_le32(oinfo->dqi_gi.dqi_free_entry);
+	size = sb->s_op->quota_write(sb, type, (char *)&dinfo,
+				     sizeof(struct ocfs2_global_disk_dqinfo),
+				     OCFS2_GLOBAL_INFO_OFF);
+	if (size != sizeof(struct ocfs2_global_disk_dqinfo)) {
+		mlog(ML_ERROR, "Cannot write global quota info structure\n");
+		if (size >= 0)
+			size = -EIO;
+		return size;
+	}
+	return 0;
+}
+
+int ocfs2_global_write_info(struct super_block *sb, int type)
+{
+	int err;
+	struct ocfs2_mem_dqinfo *info = sb_dqinfo(sb, type)->dqi_priv;
+
+	err = ocfs2_qinfo_lock(info, 1);
+	if (err < 0)
+		return err;
+	err = __ocfs2_global_write_info(sb, type);
+	ocfs2_qinfo_unlock(info, 1);
+	return err;
+}
+
+/* Read in information from global quota file and acquire a reference to it.
+ * dquot_acquire() has already started the transaction and locked quota file */
+int ocfs2_global_read_dquot(struct dquot *dquot)
+{
+	int err, err2, ex = 0;
+	struct ocfs2_mem_dqinfo *info =
+			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
+
+	err = ocfs2_qinfo_lock(info, 0);
+	if (err < 0)
+		goto out;
+	err = qtree_read_dquot(&info->dqi_gi, dquot);
+	if (err < 0)
+		goto out_qlock;
+	OCFS2_DQUOT(dquot)->dq_use_count++;
+	OCFS2_DQUOT(dquot)->dq_origspace = dquot->dq_dqb.dqb_curspace;
+	OCFS2_DQUOT(dquot)->dq_originodes = dquot->dq_dqb.dqb_curinodes;
+	if (!dquot->dq_off) {	/* No real quota entry? */
+		/* Upgrade to exclusive lock for allocation */
+		err = ocfs2_qinfo_lock(info, 1);
+		if (err < 0)
+			goto out_qlock;
+		ex = 1;
+	}
+	err = qtree_write_dquot(&info->dqi_gi, dquot);
+	if (ex && info_dirty(sb_dqinfo(dquot->dq_sb, dquot->dq_type))) {
+		err2 = __ocfs2_global_write_info(dquot->dq_sb, dquot->dq_type);
+		if (!err)
+			err = err2;
+	}
+out_qlock:
+	if (ex)
+		ocfs2_qinfo_unlock(info, 1);
+	ocfs2_qinfo_unlock(info, 0);
+out:
+	if (err < 0)
+		mlog_errno(err);
+	return err;
+}
+
+/* Sync local information about quota modifications with global quota file.
+ * Caller must have started the transaction and obtained exclusive lock for
+ * global quota file inode */
+int __ocfs2_sync_dquot(struct dquot *dquot, int freeing)
+{
+	int err, err2;
+	struct super_block *sb = dquot->dq_sb;
+	int type = dquot->dq_type;
+	struct ocfs2_mem_dqinfo *info = sb_dqinfo(sb, type)->dqi_priv;
+	struct ocfs2_global_disk_dqblk dqblk;
+	s64 spacechange, inodechange;
+	time_t olditime, oldbtime;
+
+	err = sb->s_op->quota_read(sb, type, (char *)&dqblk,
+				   sizeof(struct ocfs2_global_disk_dqblk),
+				   dquot->dq_off);
+	if (err != sizeof(struct ocfs2_global_disk_dqblk)) {
+		if (err >= 0) {
+			mlog(ML_ERROR, "Short read from global quota file "
+				       "(%u read)\n", err);
+			err = -EIO;
+		}
+		goto out;
+	}
+
+	/* Update space and inode usage. Get also other information from
+	 * global quota file so that we don't overwrite any changes there.
+	 * We are */
+	spin_lock(&dq_data_lock);
+	spacechange = dquot->dq_dqb.dqb_curspace -
+					OCFS2_DQUOT(dquot)->dq_origspace;
+	inodechange = dquot->dq_dqb.dqb_curinodes -
+					OCFS2_DQUOT(dquot)->dq_originodes;
+	olditime = dquot->dq_dqb.dqb_itime;
+	oldbtime = dquot->dq_dqb.dqb_btime;
+	ocfs2_global_disk2memdqb(dquot, &dqblk);
+	mlog(0, "Syncing global dquot %u space %lld+%lld, inodes %lld+%lld\n",
+	     dquot->dq_id, dquot->dq_dqb.dqb_curspace, (long long)spacechange,
+	     dquot->dq_dqb.dqb_curinodes, (long long)inodechange);
+	if (!test_bit(DQ_LASTSET_B + QIF_SPACE_B, &dquot->dq_flags))
+		dquot->dq_dqb.dqb_curspace += spacechange;
+	if (!test_bit(DQ_LASTSET_B + QIF_INODES_B, &dquot->dq_flags))
+		dquot->dq_dqb.dqb_curinodes += inodechange;
+	/* Set properly space grace time... */
+	if (dquot->dq_dqb.dqb_bsoftlimit &&
+	    dquot->dq_dqb.dqb_curspace > dquot->dq_dqb.dqb_bsoftlimit) {
+		if (!test_bit(DQ_LASTSET_B + QIF_BTIME_B, &dquot->dq_flags) &&
+		    oldbtime > 0) {
+			if (dquot->dq_dqb.dqb_btime > 0)
+				dquot->dq_dqb.dqb_btime =
+					min(dquot->dq_dqb.dqb_btime, oldbtime);
+			else
+				dquot->dq_dqb.dqb_btime = oldbtime;
+		}
+	} else {
+		dquot->dq_dqb.dqb_btime = 0;
+		clear_bit(DQ_BLKS_B, &dquot->dq_flags);
+	}
+	/* Set properly inode grace time... */
+	if (dquot->dq_dqb.dqb_isoftlimit &&
+	    dquot->dq_dqb.dqb_curinodes > dquot->dq_dqb.dqb_isoftlimit) {
+		if (!test_bit(DQ_LASTSET_B + QIF_ITIME_B, &dquot->dq_flags) &&
+		    olditime > 0) {
+			if (dquot->dq_dqb.dqb_itime > 0)
+				dquot->dq_dqb.dqb_itime =
+					min(dquot->dq_dqb.dqb_itime, olditime);
+			else
+				dquot->dq_dqb.dqb_itime = olditime;
+		}
+	} else {
+		dquot->dq_dqb.dqb_itime = 0;
+		clear_bit(DQ_INODES_B, &dquot->dq_flags);
+	}
+	/* All information is properly updated, clear the flags */
+	__clear_bit(DQ_LASTSET_B + QIF_SPACE_B, &dquot->dq_flags);
+	__clear_bit(DQ_LASTSET_B + QIF_INODES_B, &dquot->dq_flags);
+	__clear_bit(DQ_LASTSET_B + QIF_BLIMITS_B, &dquot->dq_flags);
+	__clear_bit(DQ_LASTSET_B + QIF_ILIMITS_B, &dquot->dq_flags);
+	__clear_bit(DQ_LASTSET_B + QIF_BTIME_B, &dquot->dq_flags);
+	__clear_bit(DQ_LASTSET_B + QIF_ITIME_B, &dquot->dq_flags);
+	OCFS2_DQUOT(dquot)->dq_origspace = dquot->dq_dqb.dqb_curspace;
+	OCFS2_DQUOT(dquot)->dq_originodes = dquot->dq_dqb.dqb_curinodes;
+	spin_unlock(&dq_data_lock);
+	err = ocfs2_qinfo_lock(info, freeing);
+	if (err < 0) {
+		mlog(ML_ERROR, "Failed to lock quota info, loosing quota write"
+			       " (type=%d, id=%u)\n", dquot->dq_type,
+			       (unsigned)dquot->dq_id);
+		goto out;
+	}
+	if (freeing)
+		OCFS2_DQUOT(dquot)->dq_use_count--;
+	err = qtree_write_dquot(&info->dqi_gi, dquot);
+	if (err < 0)
+		goto out_qlock;
+	if (freeing && !OCFS2_DQUOT(dquot)->dq_use_count) {
+		err = qtree_release_dquot(&info->dqi_gi, dquot);
+		if (info_dirty(sb_dqinfo(sb, type))) {
+			err2 = __ocfs2_global_write_info(sb, type);
+			if (!err)
+				err = err2;
+		}
+	}
+out_qlock:
+	ocfs2_qinfo_unlock(info, freeing);
+out:
+	if (err < 0)
+		mlog_errno(err);
+	return err;
+}
+
+/*
+ *  Functions for periodic syncing of dquots with global file
+ */
+static int ocfs2_sync_dquot_helper(struct dquot *dquot, unsigned long type)
+{
+	handle_t *handle;
+	struct super_block *sb = dquot->dq_sb;
+	struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv;
+	struct ocfs2_super *osb = OCFS2_SB(sb);
+	int status = 0;
+
+	mlog_entry("id=%u qtype=%u type=%lu device=%s\n", dquot->dq_id,
+		   dquot->dq_type, type, sb->s_id);
+	if (type != dquot->dq_type)
+		goto out;
+	status = ocfs2_lock_global_qf(oinfo, 1);
+	if (status < 0)
+		goto out;
+
+	handle = ocfs2_start_trans(osb, OCFS2_QSYNC_CREDITS);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out_ilock;
+	}
+	mutex_lock(&sb_dqopt(sb)->dqio_mutex);
+	status = ocfs2_sync_dquot(dquot);
+	mutex_unlock(&sb_dqopt(sb)->dqio_mutex);
+	if (status < 0)
+		mlog_errno(status);
+	/* We have to write local structure as well... */
+	dquot_mark_dquot_dirty(dquot);
+	status = dquot_commit(dquot);
+	if (status < 0)
+		mlog_errno(status);
+	ocfs2_commit_trans(osb, handle);
+out_ilock:
+	ocfs2_unlock_global_qf(oinfo, 1);
+out:
+	mlog_exit(status);
+	return status;
+}
+
+static void qsync_work_fn(struct work_struct *work)
+{
+	struct ocfs2_mem_dqinfo *oinfo = container_of(work,
+						      struct ocfs2_mem_dqinfo,
+						      dqi_sync_work.work);
+	struct super_block *sb = oinfo->dqi_gqinode->i_sb;
+
+	dquot_scan_active(sb, ocfs2_sync_dquot_helper, oinfo->dqi_type);
+	queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work,
+			   oinfo->dqi_syncjiff);
+}
+
+/*
+ *  Wrappers for generic quota functions
+ */
+
+static int ocfs2_write_dquot(struct dquot *dquot)
+{
+	handle_t *handle;
+	struct ocfs2_super *osb = OCFS2_SB(dquot->dq_sb);
+	int status = 0;
+
+	mlog_entry("id=%u, type=%d", dquot->dq_id, dquot->dq_type);
+
+	handle = ocfs2_start_trans(osb, OCFS2_QWRITE_CREDITS);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out;
+	}
+	status = dquot_commit(dquot);
+	ocfs2_commit_trans(osb, handle);
+out:
+	mlog_exit(status);
+	return status;
+}
+
+int ocfs2_calc_qdel_credits(struct super_block *sb, int type)
+{
+	struct ocfs2_mem_dqinfo *oinfo;
+	int features[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
+				    OCFS2_FEATURE_RO_COMPAT_GRPQUOTA };
+
+	if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, features[type]))
+		return 0;
+
+	oinfo = sb_dqinfo(sb, type)->dqi_priv;
+	/* We modify tree, leaf block, global info, local chunk header,
+	 * global and local inode */
+	return oinfo->dqi_gi.dqi_qtree_depth + 2 + 1 +
+	       2 * OCFS2_INODE_UPDATE_CREDITS;
+}
+
+static int ocfs2_release_dquot(struct dquot *dquot)
+{
+	handle_t *handle;
+	struct ocfs2_mem_dqinfo *oinfo =
+			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
+	struct ocfs2_super *osb = OCFS2_SB(dquot->dq_sb);
+	int status = 0;
+
+	mlog_entry("id=%u, type=%d", dquot->dq_id, dquot->dq_type);
+
+	status = ocfs2_lock_global_qf(oinfo, 1);
+	if (status < 0)
+		goto out;
+	handle = ocfs2_start_trans(osb,
+		ocfs2_calc_qdel_credits(dquot->dq_sb, dquot->dq_type));
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out_ilock;
+	}
+	status = dquot_release(dquot);
+	ocfs2_commit_trans(osb, handle);
+out_ilock:
+	ocfs2_unlock_global_qf(oinfo, 1);
+out:
+	mlog_exit(status);
+	return status;
+}
+
+int ocfs2_calc_qinit_credits(struct super_block *sb, int type)
+{
+	struct ocfs2_mem_dqinfo *oinfo;
+	int features[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
+				    OCFS2_FEATURE_RO_COMPAT_GRPQUOTA };
+	struct ocfs2_dinode *lfe, *gfe;
+
+	if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, features[type]))
+		return 0;
+
+	oinfo = sb_dqinfo(sb, type)->dqi_priv;
+	gfe = (struct ocfs2_dinode *)oinfo->dqi_gqi_bh->b_data;
+	lfe = (struct ocfs2_dinode *)oinfo->dqi_lqi_bh->b_data;
+	/* We can extend local file + global file. In local file we
+	 * can modify info, chunk header block and dquot block. In
+	 * global file we can modify info, tree and leaf block */
+	return ocfs2_calc_extend_credits(sb, &lfe->id2.i_list, 0) +
+	       ocfs2_calc_extend_credits(sb, &gfe->id2.i_list, 0) +
+	       3 + oinfo->dqi_gi.dqi_qtree_depth + 2;
+}
+
+static int ocfs2_acquire_dquot(struct dquot *dquot)
+{
+	handle_t *handle;
+	struct ocfs2_mem_dqinfo *oinfo =
+			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
+	struct ocfs2_super *osb = OCFS2_SB(dquot->dq_sb);
+	int status = 0;
+
+	mlog_entry("id=%u, type=%d", dquot->dq_id, dquot->dq_type);
+	/* We need an exclusive lock, because we're going to update use count
+	 * and instantiate possibly new dquot structure */
+	status = ocfs2_lock_global_qf(oinfo, 1);
+	if (status < 0)
+		goto out;
+	handle = ocfs2_start_trans(osb,
+		ocfs2_calc_qinit_credits(dquot->dq_sb, dquot->dq_type));
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out_ilock;
+	}
+	status = dquot_acquire(dquot);
+	ocfs2_commit_trans(osb, handle);
+out_ilock:
+	ocfs2_unlock_global_qf(oinfo, 1);
+out:
+	mlog_exit(status);
+	return status;
+}
+
+static int ocfs2_mark_dquot_dirty(struct dquot *dquot)
+{
+	unsigned long mask = (1 << (DQ_LASTSET_B + QIF_ILIMITS_B)) |
+			     (1 << (DQ_LASTSET_B + QIF_BLIMITS_B)) |
+			     (1 << (DQ_LASTSET_B + QIF_INODES_B)) |
+			     (1 << (DQ_LASTSET_B + QIF_SPACE_B)) |
+			     (1 << (DQ_LASTSET_B + QIF_BTIME_B)) |
+			     (1 << (DQ_LASTSET_B + QIF_ITIME_B));
+	int sync = 0;
+	int status;
+	struct super_block *sb = dquot->dq_sb;
+	int type = dquot->dq_type;
+	struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv;
+	handle_t *handle;
+	struct ocfs2_super *osb = OCFS2_SB(sb);
+
+	mlog_entry("id=%u, type=%d", dquot->dq_id, type);
+	dquot_mark_dquot_dirty(dquot);
+
+	/* In case user set some limits, sync dquot immediately to global
+	 * quota file so that information propagates quicker */
+	spin_lock(&dq_data_lock);
+	if (dquot->dq_flags & mask)
+		sync = 1;
+	spin_unlock(&dq_data_lock);
+	if (!sync) {
+		status = ocfs2_write_dquot(dquot);
+		goto out;
+	}
+	status = ocfs2_lock_global_qf(oinfo, 1);
+	if (status < 0)
+		goto out;
+	handle = ocfs2_start_trans(osb, OCFS2_QSYNC_CREDITS);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out_ilock;
+	}
+	status = ocfs2_sync_dquot(dquot);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	/* Now write updated local dquot structure */
+	status = dquot_commit(dquot);
+out_trans:
+	ocfs2_commit_trans(osb, handle);
+out_ilock:
+	ocfs2_unlock_global_qf(oinfo, 1);
+out:
+	mlog_exit(status);
+	return status;
+}
+
+/* This should happen only after set_dqinfo(). */
+static int ocfs2_write_info(struct super_block *sb, int type)
+{
+	handle_t *handle;
+	int status = 0;
+	struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv;
+
+	mlog_entry_void();
+
+	status = ocfs2_lock_global_qf(oinfo, 1);
+	if (status < 0)
+		goto out;
+	handle = ocfs2_start_trans(OCFS2_SB(sb), OCFS2_QINFO_WRITE_CREDITS);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out_ilock;
+	}
+	status = dquot_commit_info(sb, type);
+	ocfs2_commit_trans(OCFS2_SB(sb), handle);
+out_ilock:
+	ocfs2_unlock_global_qf(oinfo, 1);
+out:
+	mlog_exit(status);
+	return status;
+}
+
+/* This is difficult. We have to lock quota inode and start transaction
+ * in this function but we don't want to take the penalty of exlusive
+ * quota file lock when we are just going to use cached structures. So
+ * we just take read lock check whether we have dquot cached and if so,
+ * we don't have to take the write lock... */
+static int ocfs2_dquot_initialize(struct inode *inode, int type)
+{
+	handle_t *handle = NULL;
+	int status = 0;
+	struct super_block *sb = inode->i_sb;
+	struct ocfs2_mem_dqinfo *oinfo;
+	int exclusive = 0;
+	int cnt;
+	qid_t id;
+
+	mlog_entry_void();
+
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		if (type != -1 && cnt != type)
+			continue;
+		if (!sb_has_quota_active(sb, cnt))
+			continue;
+		oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
+		status = ocfs2_lock_global_qf(oinfo, 0);
+		if (status < 0)
+			goto out;
+		/* This is just a performance optimization not a reliable test.
+		 * Since we hold an inode lock, noone can actually release
+		 * the structure until we are finished with initialization. */
+		if (inode->i_dquot[cnt] != NODQUOT) {
+			ocfs2_unlock_global_qf(oinfo, 0);
+			continue;
+		}
+		/* When we have inode lock, we know that no dquot_release() can
+		 * run and thus we can safely check whether we need to
+		 * read+modify global file to get quota information or whether
+		 * our node already has it. */
+		if (cnt == USRQUOTA)
+			id = inode->i_uid;
+		else if (cnt == GRPQUOTA)
+			id = inode->i_gid;
+		else
+			BUG();
+		/* Obtain exclusion from quota off... */
+		down_write(&sb_dqopt(sb)->dqptr_sem);
+		exclusive = !dquot_is_cached(sb, id, cnt);
+		up_write(&sb_dqopt(sb)->dqptr_sem);
+		if (exclusive) {
+			status = ocfs2_lock_global_qf(oinfo, 1);
+			if (status < 0) {
+				exclusive = 0;
+				mlog_errno(status);
+				goto out_ilock;
+			}
+			handle = ocfs2_start_trans(OCFS2_SB(sb),
+					ocfs2_calc_qinit_credits(sb, cnt));
+			if (IS_ERR(handle)) {
+				status = PTR_ERR(handle);
+				mlog_errno(status);
+				goto out_ilock;
+			}
+		}
+		dquot_initialize(inode, cnt);
+		if (exclusive) {
+			ocfs2_commit_trans(OCFS2_SB(sb), handle);
+			ocfs2_unlock_global_qf(oinfo, 1);
+		}
+		ocfs2_unlock_global_qf(oinfo, 0);
+	}
+	mlog_exit(0);
+	return 0;
+out_ilock:
+	if (exclusive)
+		ocfs2_unlock_global_qf(oinfo, 1);
+	ocfs2_unlock_global_qf(oinfo, 0);
+out:
+	mlog_exit(status);
+	return status;
+}
+
+static int ocfs2_dquot_drop_slow(struct inode *inode)
+{
+	int status = 0;
+	int cnt;
+	int got_lock[MAXQUOTAS] = {0, 0};
+	handle_t *handle;
+	struct super_block *sb = inode->i_sb;
+	struct ocfs2_mem_dqinfo *oinfo;
+
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		if (!sb_has_quota_active(sb, cnt))
+			continue;
+		oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
+		status = ocfs2_lock_global_qf(oinfo, 1);
+		if (status < 0)
+			goto out;
+		got_lock[cnt] = 1;
+	}
+	handle = ocfs2_start_trans(OCFS2_SB(sb),
+			ocfs2_calc_qinit_credits(sb, USRQUOTA) +
+			ocfs2_calc_qinit_credits(sb, GRPQUOTA));
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out;
+	}
+	dquot_drop(inode);
+	ocfs2_commit_trans(OCFS2_SB(sb), handle);
+out:
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
+		if (got_lock[cnt]) {
+			oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
+			ocfs2_unlock_global_qf(oinfo, 1);
+		}
+	return status;
+}
+
+/* See the comment before ocfs2_dquot_initialize. */
+static int ocfs2_dquot_drop(struct inode *inode)
+{
+	int status = 0;
+	struct super_block *sb = inode->i_sb;
+	struct ocfs2_mem_dqinfo *oinfo;
+	int exclusive = 0;
+	int cnt;
+	int got_lock[MAXQUOTAS] = {0, 0};
+
+	mlog_entry_void();
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		if (!sb_has_quota_active(sb, cnt))
+			continue;
+		oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
+		status = ocfs2_lock_global_qf(oinfo, 0);
+		if (status < 0)
+			goto out;
+		got_lock[cnt] = 1;
+	}
+	/* Lock against anyone releasing references so that when when we check
+	 * we know we are not going to be last ones to release dquot */
+	down_write(&sb_dqopt(sb)->dqptr_sem);
+	/* Urgh, this is a terrible hack :( */
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
+		if (inode->i_dquot[cnt] != NODQUOT &&
+		    atomic_read(&inode->i_dquot[cnt]->dq_count) > 1) {
+			exclusive = 1;
+			break;
+		}
+	}
+	if (!exclusive)
+		dquot_drop_locked(inode);
+	up_write(&sb_dqopt(sb)->dqptr_sem);
+out:
+	for (cnt = 0; cnt < MAXQUOTAS; cnt++)
+		if (got_lock[cnt]) {
+			oinfo = sb_dqinfo(sb, cnt)->dqi_priv;
+			ocfs2_unlock_global_qf(oinfo, 0);
+		}
+	/* In case we bailed out because we had to do expensive locking
+	 * do it now... */
+	if (exclusive)
+		status = ocfs2_dquot_drop_slow(inode);
+	mlog_exit(status);
+	return status;
+}
+
+static struct dquot *ocfs2_alloc_dquot(struct super_block *sb, int type)
+{
+	struct ocfs2_dquot *dquot =
+				kmem_cache_zalloc(ocfs2_dquot_cachep, GFP_NOFS);
+
+	if (!dquot)
+		return NULL;
+	return &dquot->dq_dquot;
+}
+
+static void ocfs2_destroy_dquot(struct dquot *dquot)
+{
+	kmem_cache_free(ocfs2_dquot_cachep, dquot);
+}
+
+struct dquot_operations ocfs2_quota_operations = {
+	.initialize	= ocfs2_dquot_initialize,
+	.drop		= ocfs2_dquot_drop,
+	.alloc_space	= dquot_alloc_space,
+	.alloc_inode	= dquot_alloc_inode,
+	.free_space	= dquot_free_space,
+	.free_inode	= dquot_free_inode,
+	.transfer	= dquot_transfer,
+	.write_dquot	= ocfs2_write_dquot,
+	.acquire_dquot	= ocfs2_acquire_dquot,
+	.release_dquot	= ocfs2_release_dquot,
+	.mark_dirty	= ocfs2_mark_dquot_dirty,
+	.write_info	= ocfs2_write_info,
+	.alloc_dquot	= ocfs2_alloc_dquot,
+	.destroy_dquot	= ocfs2_destroy_dquot,
+};
+
+int ocfs2_quota_setup(void)
+{
+	ocfs2_quota_wq = create_workqueue("o2quot");
+	if (!ocfs2_quota_wq)
+		return -ENOMEM;
+	return 0;
+}
+
+void ocfs2_quota_shutdown(void)
+{
+	if (ocfs2_quota_wq) {
+		flush_workqueue(ocfs2_quota_wq);
+		destroy_workqueue(ocfs2_quota_wq);
+		ocfs2_quota_wq = NULL;
+	}
+}

fs/ocfs2/quota_local.c

@@ -0,0 +1,1253 @@
+/*
+ *  Implementation of operations over local quota file
+ */
+
+#include <linux/fs.h>
+#include <linux/quota.h>
+#include <linux/quotaops.h>
+#include <linux/module.h>
+
+#define MLOG_MASK_PREFIX ML_QUOTA
+#include <cluster/masklog.h>
+
+#include "ocfs2_fs.h"
+#include "ocfs2.h"
+#include "inode.h"
+#include "alloc.h"
+#include "file.h"
+#include "buffer_head_io.h"
+#include "journal.h"
+#include "sysfile.h"
+#include "dlmglue.h"
+#include "quota.h"
+
+/* Number of local quota structures per block */
+static inline unsigned int ol_quota_entries_per_block(struct super_block *sb)
+{
+	return ((sb->s_blocksize - OCFS2_QBLK_RESERVED_SPACE) /
+		sizeof(struct ocfs2_local_disk_dqblk));
+}
+
+/* Number of blocks with entries in one chunk */
+static inline unsigned int ol_chunk_blocks(struct super_block *sb)
+{
+	return ((sb->s_blocksize - sizeof(struct ocfs2_local_disk_chunk) -
+		 OCFS2_QBLK_RESERVED_SPACE) << 3) /
+	       ol_quota_entries_per_block(sb);
+}
+
+/* Number of entries in a chunk bitmap */
+static unsigned int ol_chunk_entries(struct super_block *sb)
+{
+	return ol_chunk_blocks(sb) * ol_quota_entries_per_block(sb);
+}
+
+/* Offset of the chunk in quota file */
+static unsigned int ol_quota_chunk_block(struct super_block *sb, int c)
+{
+	/* 1 block for local quota file info, 1 block per chunk for chunk info */
+	return 1 + (ol_chunk_blocks(sb) + 1) * c;
+}
+
+static unsigned int ol_dqblk_block(struct super_block *sb, int c, int off)
+{
+	int epb = ol_quota_entries_per_block(sb);
+
+	return ol_quota_chunk_block(sb, c) + 1 + off / epb;
+}
+
+static unsigned int ol_dqblk_block_off(struct super_block *sb, int c, int off)
+{
+	int epb = ol_quota_entries_per_block(sb);
+
+	return (off % epb) * sizeof(struct ocfs2_local_disk_dqblk);
+}
+
+/* Offset of the dquot structure in the quota file */
+static loff_t ol_dqblk_off(struct super_block *sb, int c, int off)
+{
+	return (ol_dqblk_block(sb, c, off) << sb->s_blocksize_bits) +
+	       ol_dqblk_block_off(sb, c, off);
+}
+
+/* Compute block number from given offset */
+static inline unsigned int ol_dqblk_file_block(struct super_block *sb, loff_t off)
+{
+	return off >> sb->s_blocksize_bits;
+}
+
+static inline unsigned int ol_dqblk_block_offset(struct super_block *sb, loff_t off)
+{
+	return off & ((1 << sb->s_blocksize_bits) - 1);
+}
+
+/* Compute offset in the chunk of a structure with the given offset */
+static int ol_dqblk_chunk_off(struct super_block *sb, int c, loff_t off)
+{
+	int epb = ol_quota_entries_per_block(sb);
+
+	return ((off >> sb->s_blocksize_bits) -
+			ol_quota_chunk_block(sb, c) - 1) * epb
+	       + ((unsigned int)(off & ((1 << sb->s_blocksize_bits) - 1))) /
+		 sizeof(struct ocfs2_local_disk_dqblk);
+}
+
+/* Write bufferhead into the fs */
+static int ocfs2_modify_bh(struct inode *inode, struct buffer_head *bh,
+		void (*modify)(struct buffer_head *, void *), void *private)
+{
+	struct super_block *sb = inode->i_sb;
+	handle_t *handle;
+	int status;
+
+	handle = ocfs2_start_trans(OCFS2_SB(sb), 1);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		return status;
+	}
+	status = ocfs2_journal_access_dq(handle, inode, bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
+	if (status < 0) {
+		mlog_errno(status);
+		ocfs2_commit_trans(OCFS2_SB(sb), handle);
+		return status;
+	}
+	lock_buffer(bh);
+	modify(bh, private);
+	unlock_buffer(bh);
+	status = ocfs2_journal_dirty(handle, bh);
+	if (status < 0) {
+		mlog_errno(status);
+		ocfs2_commit_trans(OCFS2_SB(sb), handle);
+		return status;
+	}
+	status = ocfs2_commit_trans(OCFS2_SB(sb), handle);
+	if (status < 0) {
+		mlog_errno(status);
+		return status;
+	}
+	return 0;
+}
+
+/* Check whether we understand format of quota files */
+static int ocfs2_local_check_quota_file(struct super_block *sb, int type)
+{
+	unsigned int lmagics[MAXQUOTAS] = OCFS2_LOCAL_QMAGICS;
+	unsigned int lversions[MAXQUOTAS] = OCFS2_LOCAL_QVERSIONS;
+	unsigned int gmagics[MAXQUOTAS] = OCFS2_GLOBAL_QMAGICS;
+	unsigned int gversions[MAXQUOTAS] = OCFS2_GLOBAL_QVERSIONS;
+	unsigned int ino[MAXQUOTAS] = { USER_QUOTA_SYSTEM_INODE,
+					GROUP_QUOTA_SYSTEM_INODE };
+	struct buffer_head *bh = NULL;
+	struct inode *linode = sb_dqopt(sb)->files[type];
+	struct inode *ginode = NULL;
+	struct ocfs2_disk_dqheader *dqhead;
+	int status, ret = 0;
+
+	/* First check whether we understand local quota file */
+	status = ocfs2_read_quota_block(linode, 0, &bh);
+	if (status) {
+		mlog_errno(status);
+		mlog(ML_ERROR, "failed to read quota file header (type=%d)\n",
+			type);
+		goto out_err;
+	}
+	dqhead = (struct ocfs2_disk_dqheader *)(bh->b_data);
+	if (le32_to_cpu(dqhead->dqh_magic) != lmagics[type]) {
+		mlog(ML_ERROR, "quota file magic does not match (%u != %u),"
+			" type=%d\n", le32_to_cpu(dqhead->dqh_magic),
+			lmagics[type], type);
+		goto out_err;
+	}
+	if (le32_to_cpu(dqhead->dqh_version) != lversions[type]) {
+		mlog(ML_ERROR, "quota file version does not match (%u != %u),"
+			" type=%d\n", le32_to_cpu(dqhead->dqh_version),
+			lversions[type], type);
+		goto out_err;
+	}
+	brelse(bh);
+	bh = NULL;
+
+	/* Next check whether we understand global quota file */
+	ginode = ocfs2_get_system_file_inode(OCFS2_SB(sb), ino[type],
+						OCFS2_INVALID_SLOT);
+	if (!ginode) {
+		mlog(ML_ERROR, "cannot get global quota file inode "
+				"(type=%d)\n", type);
+		goto out_err;
+	}
+	/* Since the header is read only, we don't care about locking */
+	status = ocfs2_read_quota_block(ginode, 0, &bh);
+	if (status) {
+		mlog_errno(status);
+		mlog(ML_ERROR, "failed to read global quota file header "
+				"(type=%d)\n", type);
+		goto out_err;
+	}
+	dqhead = (struct ocfs2_disk_dqheader *)(bh->b_data);
+	if (le32_to_cpu(dqhead->dqh_magic) != gmagics[type]) {
+		mlog(ML_ERROR, "global quota file magic does not match "
+			"(%u != %u), type=%d\n",
+			le32_to_cpu(dqhead->dqh_magic), gmagics[type], type);
+		goto out_err;
+	}
+	if (le32_to_cpu(dqhead->dqh_version) != gversions[type]) {
+		mlog(ML_ERROR, "global quota file version does not match "
+			"(%u != %u), type=%d\n",
+			le32_to_cpu(dqhead->dqh_version), gversions[type],
+			type);
+		goto out_err;
+	}
+
+	ret = 1;
+out_err:
+	brelse(bh);
+	iput(ginode);
+	return ret;
+}
+
+/* Release given list of quota file chunks */
+static void ocfs2_release_local_quota_bitmaps(struct list_head *head)
+{
+	struct ocfs2_quota_chunk *pos, *next;
+
+	list_for_each_entry_safe(pos, next, head, qc_chunk) {
+		list_del(&pos->qc_chunk);
+		brelse(pos->qc_headerbh);
+		kmem_cache_free(ocfs2_qf_chunk_cachep, pos);
+	}
+}
+
+/* Load quota bitmaps into memory */
+static int ocfs2_load_local_quota_bitmaps(struct inode *inode,
+			struct ocfs2_local_disk_dqinfo *ldinfo,
+			struct list_head *head)
+{
+	struct ocfs2_quota_chunk *newchunk;
+	int i, status;
+
+	INIT_LIST_HEAD(head);
+	for (i = 0; i < le32_to_cpu(ldinfo->dqi_chunks); i++) {
+		newchunk = kmem_cache_alloc(ocfs2_qf_chunk_cachep, GFP_NOFS);
+		if (!newchunk) {
+			ocfs2_release_local_quota_bitmaps(head);
+			return -ENOMEM;
+		}
+		newchunk->qc_num = i;
+		newchunk->qc_headerbh = NULL;
+		status = ocfs2_read_quota_block(inode,
+				ol_quota_chunk_block(inode->i_sb, i),
+				&newchunk->qc_headerbh);
+		if (status) {
+			mlog_errno(status);
+			kmem_cache_free(ocfs2_qf_chunk_cachep, newchunk);
+			ocfs2_release_local_quota_bitmaps(head);
+			return status;
+		}
+		list_add_tail(&newchunk->qc_chunk, head);
+	}
+	return 0;
+}
+
+static void olq_update_info(struct buffer_head *bh, void *private)
+{
+	struct mem_dqinfo *info = private;
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	struct ocfs2_local_disk_dqinfo *ldinfo;
+
+	ldinfo = (struct ocfs2_local_disk_dqinfo *)(bh->b_data +
+						OCFS2_LOCAL_INFO_OFF);
+	spin_lock(&dq_data_lock);
+	ldinfo->dqi_flags = cpu_to_le32(info->dqi_flags & DQF_MASK);
+	ldinfo->dqi_chunks = cpu_to_le32(oinfo->dqi_chunks);
+	ldinfo->dqi_blocks = cpu_to_le32(oinfo->dqi_blocks);
+	spin_unlock(&dq_data_lock);
+}
+
+static int ocfs2_add_recovery_chunk(struct super_block *sb,
+				    struct ocfs2_local_disk_chunk *dchunk,
+				    int chunk,
+				    struct list_head *head)
+{
+	struct ocfs2_recovery_chunk *rc;
+
+	rc = kmalloc(sizeof(struct ocfs2_recovery_chunk), GFP_NOFS);
+	if (!rc)
+		return -ENOMEM;
+	rc->rc_chunk = chunk;
+	rc->rc_bitmap = kmalloc(sb->s_blocksize, GFP_NOFS);
+	if (!rc->rc_bitmap) {
+		kfree(rc);
+		return -ENOMEM;
+	}
+	memcpy(rc->rc_bitmap, dchunk->dqc_bitmap,
+	       (ol_chunk_entries(sb) + 7) >> 3);
+	list_add_tail(&rc->rc_list, head);
+	return 0;
+}
+
+static void free_recovery_list(struct list_head *head)
+{
+	struct ocfs2_recovery_chunk *next;
+	struct ocfs2_recovery_chunk *rchunk;
+
+	list_for_each_entry_safe(rchunk, next, head, rc_list) {
+		list_del(&rchunk->rc_list);
+		kfree(rchunk->rc_bitmap);
+		kfree(rchunk);
+	}
+}
+
+void ocfs2_free_quota_recovery(struct ocfs2_quota_recovery *rec)
+{
+	int type;
+
+	for (type = 0; type < MAXQUOTAS; type++)
+		free_recovery_list(&(rec->r_list[type]));
+	kfree(rec);
+}
+
+/* Load entries in our quota file we have to recover*/
+static int ocfs2_recovery_load_quota(struct inode *lqinode,
+				     struct ocfs2_local_disk_dqinfo *ldinfo,
+				     int type,
+				     struct list_head *head)
+{
+	struct super_block *sb = lqinode->i_sb;
+	struct buffer_head *hbh;
+	struct ocfs2_local_disk_chunk *dchunk;
+	int i, chunks = le32_to_cpu(ldinfo->dqi_chunks);
+	int status = 0;
+
+	for (i = 0; i < chunks; i++) {
+		hbh = NULL;
+		status = ocfs2_read_quota_block(lqinode,
+						ol_quota_chunk_block(sb, i),
+						&hbh);
+		if (status) {
+			mlog_errno(status);
+			break;
+		}
+		dchunk = (struct ocfs2_local_disk_chunk *)hbh->b_data;
+		if (le32_to_cpu(dchunk->dqc_free) < ol_chunk_entries(sb))
+			status = ocfs2_add_recovery_chunk(sb, dchunk, i, head);
+		brelse(hbh);
+		if (status < 0)
+			break;
+	}
+	if (status < 0)
+		free_recovery_list(head);
+	return status;
+}
+
+static struct ocfs2_quota_recovery *ocfs2_alloc_quota_recovery(void)
+{
+	int type;
+	struct ocfs2_quota_recovery *rec;
+
+	rec = kmalloc(sizeof(struct ocfs2_quota_recovery), GFP_NOFS);
+	if (!rec)
+		return NULL;
+	for (type = 0; type < MAXQUOTAS; type++)
+		INIT_LIST_HEAD(&(rec->r_list[type]));
+	return rec;
+}
+
+/* Load information we need for quota recovery into memory */
+struct ocfs2_quota_recovery *ocfs2_begin_quota_recovery(
+						struct ocfs2_super *osb,
+						int slot_num)
+{
+	unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
+					    OCFS2_FEATURE_RO_COMPAT_GRPQUOTA};
+	unsigned int ino[MAXQUOTAS] = { LOCAL_USER_QUOTA_SYSTEM_INODE,
+					LOCAL_GROUP_QUOTA_SYSTEM_INODE };
+	struct super_block *sb = osb->sb;
+	struct ocfs2_local_disk_dqinfo *ldinfo;
+	struct inode *lqinode;
+	struct buffer_head *bh;
+	int type;
+	int status = 0;
+	struct ocfs2_quota_recovery *rec;
+
+	mlog(ML_NOTICE, "Beginning quota recovery in slot %u\n", slot_num);
+	rec = ocfs2_alloc_quota_recovery();
+	if (!rec)
+		return ERR_PTR(-ENOMEM);
+	/* First init... */
+
+	for (type = 0; type < MAXQUOTAS; type++) {
+		if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, feature[type]))
+			continue;
+		/* At this point, journal of the slot is already replayed so
+		 * we can trust metadata and data of the quota file */
+		lqinode = ocfs2_get_system_file_inode(osb, ino[type], slot_num);
+		if (!lqinode) {
+			status = -ENOENT;
+			goto out;
+		}
+		status = ocfs2_inode_lock_full(lqinode, NULL, 1,
+					       OCFS2_META_LOCK_RECOVERY);
+		if (status < 0) {
+			mlog_errno(status);
+			goto out_put;
+		}
+		/* Now read local header */
+		bh = NULL;
+		status = ocfs2_read_quota_block(lqinode, 0, &bh);
+		if (status) {
+			mlog_errno(status);
+			mlog(ML_ERROR, "failed to read quota file info header "
+				"(slot=%d type=%d)\n", slot_num, type);
+			goto out_lock;
+		}
+		ldinfo = (struct ocfs2_local_disk_dqinfo *)(bh->b_data +
+							OCFS2_LOCAL_INFO_OFF);
+		status = ocfs2_recovery_load_quota(lqinode, ldinfo, type,
+						   &rec->r_list[type]);
+		brelse(bh);
+out_lock:
+		ocfs2_inode_unlock(lqinode, 1);
+out_put:
+		iput(lqinode);
+		if (status < 0)
+			break;
+	}
+out:
+	if (status < 0) {
+		ocfs2_free_quota_recovery(rec);
+		rec = ERR_PTR(status);
+	}
+	return rec;
+}
+
+/* Sync changes in local quota file into global quota file and
+ * reinitialize local quota file.
+ * The function expects local quota file to be already locked and
+ * dqonoff_mutex locked. */
+static int ocfs2_recover_local_quota_file(struct inode *lqinode,
+					  int type,
+					  struct ocfs2_quota_recovery *rec)
+{
+	struct super_block *sb = lqinode->i_sb;
+	struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv;
+	struct ocfs2_local_disk_chunk *dchunk;
+	struct ocfs2_local_disk_dqblk *dqblk;
+	struct dquot *dquot;
+	handle_t *handle;
+	struct buffer_head *hbh = NULL, *qbh = NULL;
+	int status = 0;
+	int bit, chunk;
+	struct ocfs2_recovery_chunk *rchunk, *next;
+	qsize_t spacechange, inodechange;
+
+	mlog_entry("ino=%lu type=%u", (unsigned long)lqinode->i_ino, type);
+
+	status = ocfs2_lock_global_qf(oinfo, 1);
+	if (status < 0)
+		goto out;
+
+	list_for_each_entry_safe(rchunk, next, &(rec->r_list[type]), rc_list) {
+		chunk = rchunk->rc_chunk;
+		hbh = NULL;
+		status = ocfs2_read_quota_block(lqinode,
+						ol_quota_chunk_block(sb, chunk),
+						&hbh);
+		if (status) {
+			mlog_errno(status);
+			break;
+		}
+		dchunk = (struct ocfs2_local_disk_chunk *)hbh->b_data;
+		for_each_bit(bit, rchunk->rc_bitmap, ol_chunk_entries(sb)) {
+			qbh = NULL;
+			status = ocfs2_read_quota_block(lqinode,
+						ol_dqblk_block(sb, chunk, bit),
+						&qbh);
+			if (status) {
+				mlog_errno(status);
+				break;
+			}
+			dqblk = (struct ocfs2_local_disk_dqblk *)(qbh->b_data +
+				ol_dqblk_block_off(sb, chunk, bit));
+			dquot = dqget(sb, le64_to_cpu(dqblk->dqb_id), type);
+			if (!dquot) {
+				status = -EIO;
+				mlog(ML_ERROR, "Failed to get quota structure "
+				     "for id %u, type %d. Cannot finish quota "
+				     "file recovery.\n",
+				     (unsigned)le64_to_cpu(dqblk->dqb_id),
+				     type);
+				goto out_put_bh;
+			}
+			handle = ocfs2_start_trans(OCFS2_SB(sb),
+						   OCFS2_QSYNC_CREDITS);
+			if (IS_ERR(handle)) {
+				status = PTR_ERR(handle);
+				mlog_errno(status);
+				goto out_put_dquot;
+			}
+			mutex_lock(&sb_dqopt(sb)->dqio_mutex);
+			spin_lock(&dq_data_lock);
+			/* Add usage from quota entry into quota changes
+			 * of our node. Auxiliary variables are important
+			 * due to signedness */
+			spacechange = le64_to_cpu(dqblk->dqb_spacemod);
+			inodechange = le64_to_cpu(dqblk->dqb_inodemod);
+			dquot->dq_dqb.dqb_curspace += spacechange;
+			dquot->dq_dqb.dqb_curinodes += inodechange;
+			spin_unlock(&dq_data_lock);
+			/* We want to drop reference held by the crashed
+			 * node. Since we have our own reference we know
+			 * global structure actually won't be freed. */
+			status = ocfs2_global_release_dquot(dquot);
+			if (status < 0) {
+				mlog_errno(status);
+				goto out_commit;
+			}
+			/* Release local quota file entry */
+			status = ocfs2_journal_access_dq(handle, lqinode,
+					qbh, OCFS2_JOURNAL_ACCESS_WRITE);
+			if (status < 0) {
+				mlog_errno(status);
+				goto out_commit;
+			}
+			lock_buffer(qbh);
+			WARN_ON(!ocfs2_test_bit(bit, dchunk->dqc_bitmap));
+			ocfs2_clear_bit(bit, dchunk->dqc_bitmap);
+			le32_add_cpu(&dchunk->dqc_free, 1);
+			unlock_buffer(qbh);
+			status = ocfs2_journal_dirty(handle, qbh);
+			if (status < 0)
+				mlog_errno(status);
+out_commit:
+			mutex_unlock(&sb_dqopt(sb)->dqio_mutex);
+			ocfs2_commit_trans(OCFS2_SB(sb), handle);
+out_put_dquot:
+			dqput(dquot);
+out_put_bh:
+			brelse(qbh);
+			if (status < 0)
+				break;
+		}
+		brelse(hbh);
+		list_del(&rchunk->rc_list);
+		kfree(rchunk->rc_bitmap);
+		kfree(rchunk);
+		if (status < 0)
+			break;
+	}
+	ocfs2_unlock_global_qf(oinfo, 1);
+out:
+	if (status < 0)
+		free_recovery_list(&(rec->r_list[type]));
+	mlog_exit(status);
+	return status;
+}
+
+/* Recover local quota files for given node different from us */
+int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
+				struct ocfs2_quota_recovery *rec,
+				int slot_num)
+{
+	unsigned int ino[MAXQUOTAS] = { LOCAL_USER_QUOTA_SYSTEM_INODE,
+					LOCAL_GROUP_QUOTA_SYSTEM_INODE };
+	struct super_block *sb = osb->sb;
+	struct ocfs2_local_disk_dqinfo *ldinfo;
+	struct buffer_head *bh;
+	handle_t *handle;
+	int type;
+	int status = 0;
+	struct inode *lqinode;
+	unsigned int flags;
+
+	mlog(ML_NOTICE, "Finishing quota recovery in slot %u\n", slot_num);
+	mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
+	for (type = 0; type < MAXQUOTAS; type++) {
+		if (list_empty(&(rec->r_list[type])))
+			continue;
+		mlog(0, "Recovering quota in slot %d\n", slot_num);
+		lqinode = ocfs2_get_system_file_inode(osb, ino[type], slot_num);
+		if (!lqinode) {
+			status = -ENOENT;
+			goto out;
+		}
+		status = ocfs2_inode_lock_full(lqinode, NULL, 1,
+						       OCFS2_META_LOCK_NOQUEUE);
+		/* 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);
+			status = 0;
+			goto out_put;
+		} else if (status < 0) {
+			mlog_errno(status);
+			goto out_put;
+		}
+		/* Now read local header */
+		bh = NULL;
+		status = ocfs2_read_quota_block(lqinode, 0, &bh);
+		if (status) {
+			mlog_errno(status);
+			mlog(ML_ERROR, "failed to read quota file info header "
+				"(slot=%d type=%d)\n", slot_num, type);
+			goto out_lock;
+		}
+		ldinfo = (struct ocfs2_local_disk_dqinfo *)(bh->b_data +
+							OCFS2_LOCAL_INFO_OFF);
+		/* Is recovery still needed? */
+		flags = le32_to_cpu(ldinfo->dqi_flags);
+		if (!(flags & OLQF_CLEAN))
+			status = ocfs2_recover_local_quota_file(lqinode,
+								type,
+								rec);
+		/* We don't want to mark file as clean when it is actually
+		 * active */
+		if (slot_num == osb->slot_num)
+			goto out_bh;
+		/* Mark quota file as clean if we are recovering quota file of
+		 * some other node. */
+		handle = ocfs2_start_trans(osb, 1);
+		if (IS_ERR(handle)) {
+			status = PTR_ERR(handle);
+			mlog_errno(status);
+			goto out_bh;
+		}
+		status = ocfs2_journal_access_dq(handle, lqinode, bh,
+						 OCFS2_JOURNAL_ACCESS_WRITE);
+		if (status < 0) {
+			mlog_errno(status);
+			goto out_trans;
+		}
+		lock_buffer(bh);
+		ldinfo->dqi_flags = cpu_to_le32(flags | OLQF_CLEAN);
+		unlock_buffer(bh);
+		status = ocfs2_journal_dirty(handle, bh);
+		if (status < 0)
+			mlog_errno(status);
+out_trans:
+		ocfs2_commit_trans(osb, handle);
+out_bh:
+		brelse(bh);
+out_lock:
+		ocfs2_inode_unlock(lqinode, 1);
+out_put:
+		iput(lqinode);
+		if (status < 0)
+			break;
+	}
+out:
+	mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex);
+	kfree(rec);
+	return status;
+}
+
+/* Read information header from quota file */
+static int ocfs2_local_read_info(struct super_block *sb, int type)
+{
+	struct ocfs2_local_disk_dqinfo *ldinfo;
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo;
+	struct inode *lqinode = sb_dqopt(sb)->files[type];
+	int status;
+	struct buffer_head *bh = NULL;
+	struct ocfs2_quota_recovery *rec;
+	int locked = 0;
+
+	info->dqi_maxblimit = 0x7fffffffffffffffLL;
+	info->dqi_maxilimit = 0x7fffffffffffffffLL;
+	oinfo = kmalloc(sizeof(struct ocfs2_mem_dqinfo), GFP_NOFS);
+	if (!oinfo) {
+		mlog(ML_ERROR, "failed to allocate memory for ocfs2 quota"
+			       " info.");
+		goto out_err;
+	}
+	info->dqi_priv = oinfo;
+	oinfo->dqi_type = type;
+	INIT_LIST_HEAD(&oinfo->dqi_chunk);
+	oinfo->dqi_rec = NULL;
+	oinfo->dqi_lqi_bh = NULL;
+	oinfo->dqi_ibh = NULL;
+
+	status = ocfs2_global_read_info(sb, type);
+	if (status < 0)
+		goto out_err;
+
+	status = ocfs2_inode_lock(lqinode, &oinfo->dqi_lqi_bh, 1);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_err;
+	}
+	locked = 1;
+
+	/* Now read local header */
+	status = ocfs2_read_quota_block(lqinode, 0, &bh);
+	if (status) {
+		mlog_errno(status);
+		mlog(ML_ERROR, "failed to read quota file info header "
+			"(type=%d)\n", type);
+		goto out_err;
+	}
+	ldinfo = (struct ocfs2_local_disk_dqinfo *)(bh->b_data +
+						OCFS2_LOCAL_INFO_OFF);
+	info->dqi_flags = le32_to_cpu(ldinfo->dqi_flags);
+	oinfo->dqi_chunks = le32_to_cpu(ldinfo->dqi_chunks);
+	oinfo->dqi_blocks = le32_to_cpu(ldinfo->dqi_blocks);
+	oinfo->dqi_ibh = bh;
+
+	/* We crashed when using local quota file? */
+	if (!(info->dqi_flags & OLQF_CLEAN)) {
+		rec = OCFS2_SB(sb)->quota_rec;
+		if (!rec) {
+			rec = ocfs2_alloc_quota_recovery();
+			if (!rec) {
+				status = -ENOMEM;
+				mlog_errno(status);
+				goto out_err;
+			}
+			OCFS2_SB(sb)->quota_rec = rec;
+		}
+
+		status = ocfs2_recovery_load_quota(lqinode, ldinfo, type,
+                                                   &rec->r_list[type]);
+		if (status < 0) {
+			mlog_errno(status);
+			goto out_err;
+		}
+	}
+
+	status = ocfs2_load_local_quota_bitmaps(lqinode,
+						ldinfo,
+						&oinfo->dqi_chunk);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_err;
+	}
+
+	/* Now mark quota file as used */
+	info->dqi_flags &= ~OLQF_CLEAN;
+	status = ocfs2_modify_bh(lqinode, bh, olq_update_info, info);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_err;
+	}
+
+	return 0;
+out_err:
+	if (oinfo) {
+		iput(oinfo->dqi_gqinode);
+		ocfs2_simple_drop_lockres(OCFS2_SB(sb), &oinfo->dqi_gqlock);
+		ocfs2_lock_res_free(&oinfo->dqi_gqlock);
+		brelse(oinfo->dqi_lqi_bh);
+		if (locked)
+			ocfs2_inode_unlock(lqinode, 1);
+		ocfs2_release_local_quota_bitmaps(&oinfo->dqi_chunk);
+		kfree(oinfo);
+	}
+	brelse(bh);
+	return -1;
+}
+
+/* Write local info to quota file */
+static int ocfs2_local_write_info(struct super_block *sb, int type)
+{
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct buffer_head *bh = ((struct ocfs2_mem_dqinfo *)info->dqi_priv)
+						->dqi_ibh;
+	int status;
+
+	status = ocfs2_modify_bh(sb_dqopt(sb)->files[type], bh, olq_update_info,
+				 info);
+	if (status < 0) {
+		mlog_errno(status);
+		return -1;
+	}
+
+	return 0;
+}
+
+/* Release info from memory */
+static int ocfs2_local_free_info(struct super_block *sb, int type)
+{
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	struct ocfs2_quota_chunk *chunk;
+	struct ocfs2_local_disk_chunk *dchunk;
+	int mark_clean = 1, len;
+	int status;
+
+	/* At this point we know there are no more dquots and thus
+	 * even if there's some sync in the pdflush queue, it won't
+	 * find any dquots and return without doing anything */
+	cancel_delayed_work_sync(&oinfo->dqi_sync_work);
+	iput(oinfo->dqi_gqinode);
+	ocfs2_simple_drop_lockres(OCFS2_SB(sb), &oinfo->dqi_gqlock);
+	ocfs2_lock_res_free(&oinfo->dqi_gqlock);
+	list_for_each_entry(chunk, &oinfo->dqi_chunk, qc_chunk) {
+		dchunk = (struct ocfs2_local_disk_chunk *)
+					(chunk->qc_headerbh->b_data);
+		if (chunk->qc_num < oinfo->dqi_chunks - 1) {
+			len = ol_chunk_entries(sb);
+		} else {
+			len = (oinfo->dqi_blocks -
+			       ol_quota_chunk_block(sb, chunk->qc_num) - 1)
+			      * ol_quota_entries_per_block(sb);
+		}
+		/* Not all entries free? Bug! */
+		if (le32_to_cpu(dchunk->dqc_free) != len) {
+			mlog(ML_ERROR, "releasing quota file with used "
+					"entries (type=%d)\n", type);
+			mark_clean = 0;
+		}
+	}
+	ocfs2_release_local_quota_bitmaps(&oinfo->dqi_chunk);
+
+	/* dqonoff_mutex protects us against racing with recovery thread... */
+	if (oinfo->dqi_rec) {
+		ocfs2_free_quota_recovery(oinfo->dqi_rec);
+		mark_clean = 0;
+	}
+
+	if (!mark_clean)
+		goto out;
+
+	/* Mark local file as clean */
+	info->dqi_flags |= OLQF_CLEAN;
+	status = ocfs2_modify_bh(sb_dqopt(sb)->files[type],
+				 oinfo->dqi_ibh,
+				 olq_update_info,
+				 info);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+
+out:
+	ocfs2_inode_unlock(sb_dqopt(sb)->files[type], 1);
+	brelse(oinfo->dqi_ibh);
+	brelse(oinfo->dqi_lqi_bh);
+	kfree(oinfo);
+	return 0;
+}
+
+static void olq_set_dquot(struct buffer_head *bh, void *private)
+{
+	struct ocfs2_dquot *od = private;
+	struct ocfs2_local_disk_dqblk *dqblk;
+	struct super_block *sb = od->dq_dquot.dq_sb;
+
+	dqblk = (struct ocfs2_local_disk_dqblk *)(bh->b_data
+		+ ol_dqblk_block_offset(sb, od->dq_local_off));
+
+	dqblk->dqb_id = cpu_to_le64(od->dq_dquot.dq_id);
+	spin_lock(&dq_data_lock);
+	dqblk->dqb_spacemod = cpu_to_le64(od->dq_dquot.dq_dqb.dqb_curspace -
+					  od->dq_origspace);
+	dqblk->dqb_inodemod = cpu_to_le64(od->dq_dquot.dq_dqb.dqb_curinodes -
+					  od->dq_originodes);
+	spin_unlock(&dq_data_lock);
+	mlog(0, "Writing local dquot %u space %lld inodes %lld\n",
+	     od->dq_dquot.dq_id, (long long)le64_to_cpu(dqblk->dqb_spacemod),
+	     (long long)le64_to_cpu(dqblk->dqb_inodemod));
+}
+
+/* Write dquot to local quota file */
+static int ocfs2_local_write_dquot(struct dquot *dquot)
+{
+	struct super_block *sb = dquot->dq_sb;
+	struct ocfs2_dquot *od = OCFS2_DQUOT(dquot);
+	struct buffer_head *bh = NULL;
+	int status;
+
+	status = ocfs2_read_quota_block(sb_dqopt(sb)->files[dquot->dq_type],
+				    ol_dqblk_file_block(sb, od->dq_local_off),
+				    &bh);
+	if (status) {
+		mlog_errno(status);
+		goto out;
+	}
+	status = ocfs2_modify_bh(sb_dqopt(sb)->files[dquot->dq_type], bh,
+				 olq_set_dquot, od);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+out:
+	brelse(bh);
+	return status;
+}
+
+/* Find free entry in local quota file */
+static struct ocfs2_quota_chunk *ocfs2_find_free_entry(struct super_block *sb,
+						       int type,
+						       int *offset)
+{
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	struct ocfs2_quota_chunk *chunk;
+	struct ocfs2_local_disk_chunk *dchunk;
+	int found = 0, len;
+
+	list_for_each_entry(chunk, &oinfo->dqi_chunk, qc_chunk) {
+		dchunk = (struct ocfs2_local_disk_chunk *)
+						chunk->qc_headerbh->b_data;
+		if (le32_to_cpu(dchunk->dqc_free) > 0) {
+			found = 1;
+			break;
+		}
+	}
+	if (!found)
+		return NULL;
+
+	if (chunk->qc_num < oinfo->dqi_chunks - 1) {
+		len = ol_chunk_entries(sb);
+	} else {
+		len = (oinfo->dqi_blocks -
+		       ol_quota_chunk_block(sb, chunk->qc_num) - 1)
+		      * ol_quota_entries_per_block(sb);
+	}
+
+	found = ocfs2_find_next_zero_bit(dchunk->dqc_bitmap, len, 0);
+	/* We failed? */
+	if (found == len) {
+		mlog(ML_ERROR, "Did not find empty entry in chunk %d with %u"
+		     " entries free (type=%d)\n", chunk->qc_num,
+		     le32_to_cpu(dchunk->dqc_free), type);
+		return ERR_PTR(-EIO);
+	}
+	*offset = found;
+	return chunk;
+}
+
+/* Add new chunk to the local quota file */
+static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk(
+							struct super_block *sb,
+							int type,
+							int *offset)
+{
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	struct inode *lqinode = sb_dqopt(sb)->files[type];
+	struct ocfs2_quota_chunk *chunk = NULL;
+	struct ocfs2_local_disk_chunk *dchunk;
+	int status;
+	handle_t *handle;
+	struct buffer_head *bh = NULL;
+	u64 p_blkno;
+
+	/* We are protected by dqio_sem so no locking needed */
+	status = ocfs2_extend_no_holes(lqinode,
+				       lqinode->i_size + 2 * sb->s_blocksize,
+				       lqinode->i_size);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	status = ocfs2_simple_size_update(lqinode, oinfo->dqi_lqi_bh,
+					  lqinode->i_size + 2 * sb->s_blocksize);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+
+	chunk = kmem_cache_alloc(ocfs2_qf_chunk_cachep, GFP_NOFS);
+	if (!chunk) {
+		status = -ENOMEM;
+		mlog_errno(status);
+		goto out;
+	}
+
+	down_read(&OCFS2_I(lqinode)->ip_alloc_sem);
+	status = ocfs2_extent_map_get_blocks(lqinode, oinfo->dqi_blocks,
+					     &p_blkno, NULL, NULL);
+	up_read(&OCFS2_I(lqinode)->ip_alloc_sem);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	bh = sb_getblk(sb, p_blkno);
+	if (!bh) {
+		status = -ENOMEM;
+		mlog_errno(status);
+		goto out;
+	}
+	dchunk = (struct ocfs2_local_disk_chunk *)bh->b_data;
+
+	handle = ocfs2_start_trans(OCFS2_SB(sb), 2);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out;
+	}
+
+	status = ocfs2_journal_access_dq(handle, lqinode, bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	lock_buffer(bh);
+	dchunk->dqc_free = cpu_to_le32(ol_quota_entries_per_block(sb));
+	memset(dchunk->dqc_bitmap, 0,
+	       sb->s_blocksize - sizeof(struct ocfs2_local_disk_chunk) -
+	       OCFS2_QBLK_RESERVED_SPACE);
+	set_buffer_uptodate(bh);
+	unlock_buffer(bh);
+	status = ocfs2_journal_dirty(handle, bh);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+
+	oinfo->dqi_blocks += 2;
+	oinfo->dqi_chunks++;
+	status = ocfs2_local_write_info(sb, type);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	status = ocfs2_commit_trans(OCFS2_SB(sb), handle);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+
+	list_add_tail(&chunk->qc_chunk, &oinfo->dqi_chunk);
+	chunk->qc_num = list_entry(chunk->qc_chunk.prev,
+				   struct ocfs2_quota_chunk,
+				   qc_chunk)->qc_num + 1;
+	chunk->qc_headerbh = bh;
+	*offset = 0;
+	return chunk;
+out_trans:
+	ocfs2_commit_trans(OCFS2_SB(sb), handle);
+out:
+	brelse(bh);
+	kmem_cache_free(ocfs2_qf_chunk_cachep, chunk);
+	return ERR_PTR(status);
+}
+
+/* Find free entry in local quota file */
+static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file(
+						       struct super_block *sb,
+						       int type,
+						       int *offset)
+{
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct ocfs2_mem_dqinfo *oinfo = info->dqi_priv;
+	struct ocfs2_quota_chunk *chunk;
+	struct inode *lqinode = sb_dqopt(sb)->files[type];
+	struct ocfs2_local_disk_chunk *dchunk;
+	int epb = ol_quota_entries_per_block(sb);
+	unsigned int chunk_blocks;
+	int status;
+	handle_t *handle;
+
+	if (list_empty(&oinfo->dqi_chunk))
+		return ocfs2_local_quota_add_chunk(sb, type, offset);
+	/* Is the last chunk full? */
+	chunk = list_entry(oinfo->dqi_chunk.prev,
+			struct ocfs2_quota_chunk, qc_chunk);
+	chunk_blocks = oinfo->dqi_blocks -
+			ol_quota_chunk_block(sb, chunk->qc_num) - 1;
+	if (ol_chunk_blocks(sb) == chunk_blocks)
+		return ocfs2_local_quota_add_chunk(sb, type, offset);
+
+	/* We are protected by dqio_sem so no locking needed */
+	status = ocfs2_extend_no_holes(lqinode,
+				       lqinode->i_size + sb->s_blocksize,
+				       lqinode->i_size);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	status = ocfs2_simple_size_update(lqinode, oinfo->dqi_lqi_bh,
+					  lqinode->i_size + sb->s_blocksize);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	handle = ocfs2_start_trans(OCFS2_SB(sb), 2);
+	if (IS_ERR(handle)) {
+		status = PTR_ERR(handle);
+		mlog_errno(status);
+		goto out;
+	}
+	status = ocfs2_journal_access_dq(handle, lqinode, chunk->qc_headerbh,
+				 OCFS2_JOURNAL_ACCESS_WRITE);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+
+	dchunk = (struct ocfs2_local_disk_chunk *)chunk->qc_headerbh->b_data;
+	lock_buffer(chunk->qc_headerbh);
+	le32_add_cpu(&dchunk->dqc_free, ol_quota_entries_per_block(sb));
+	unlock_buffer(chunk->qc_headerbh);
+	status = ocfs2_journal_dirty(handle, chunk->qc_headerbh);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+	oinfo->dqi_blocks++;
+	status = ocfs2_local_write_info(sb, type);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_trans;
+	}
+
+	status = ocfs2_commit_trans(OCFS2_SB(sb), handle);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	*offset = chunk_blocks * epb;
+	return chunk;
+out_trans:
+	ocfs2_commit_trans(OCFS2_SB(sb), handle);
+out:
+	return ERR_PTR(status);
+}
+
+static void olq_alloc_dquot(struct buffer_head *bh, void *private)
+{
+	int *offset = private;
+	struct ocfs2_local_disk_chunk *dchunk;
+
+	dchunk = (struct ocfs2_local_disk_chunk *)bh->b_data;
+	ocfs2_set_bit(*offset, dchunk->dqc_bitmap);
+	le32_add_cpu(&dchunk->dqc_free, -1);
+}
+
+/* Create dquot in the local file for given id */
+static int ocfs2_create_local_dquot(struct dquot *dquot)
+{
+	struct super_block *sb = dquot->dq_sb;
+	int type = dquot->dq_type;
+	struct inode *lqinode = sb_dqopt(sb)->files[type];
+	struct ocfs2_quota_chunk *chunk;
+	struct ocfs2_dquot *od = OCFS2_DQUOT(dquot);
+	int offset;
+	int status;
+
+	chunk = ocfs2_find_free_entry(sb, type, &offset);
+	if (!chunk) {
+		chunk = ocfs2_extend_local_quota_file(sb, type, &offset);
+		if (IS_ERR(chunk))
+			return PTR_ERR(chunk);
+	} else if (IS_ERR(chunk)) {
+		return PTR_ERR(chunk);
+	}
+	od->dq_local_off = ol_dqblk_off(sb, chunk->qc_num, offset);
+	od->dq_chunk = chunk;
+
+	/* Initialize dquot structure on disk */
+	status = ocfs2_local_write_dquot(dquot);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+
+	/* Mark structure as allocated */
+	status = ocfs2_modify_bh(lqinode, chunk->qc_headerbh, olq_alloc_dquot,
+				 &offset);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+out:
+	return status;
+}
+
+/* Create entry in local file for dquot, load data from the global file */
+static int ocfs2_local_read_dquot(struct dquot *dquot)
+{
+	int status;
+
+	mlog_entry("id=%u, type=%d\n", dquot->dq_id, dquot->dq_type);
+
+	status = ocfs2_global_read_dquot(dquot);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_err;
+	}
+
+	/* Now create entry in the local quota file */
+	status = ocfs2_create_local_dquot(dquot);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out_err;
+	}
+	mlog_exit(0);
+	return 0;
+out_err:
+	mlog_exit(status);
+	return status;
+}
+
+/* Release dquot structure from local quota file. ocfs2_release_dquot() has
+ * already started a transaction and obtained exclusive lock for global
+ * quota file. */
+static int ocfs2_local_release_dquot(struct dquot *dquot)
+{
+	int status;
+	int type = dquot->dq_type;
+	struct ocfs2_dquot *od = OCFS2_DQUOT(dquot);
+	struct super_block *sb = dquot->dq_sb;
+	struct ocfs2_local_disk_chunk *dchunk;
+	int offset;
+	handle_t *handle = journal_current_handle();
+
+	BUG_ON(!handle);
+	/* First write all local changes to global file */
+	status = ocfs2_global_release_dquot(dquot);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+
+	status = ocfs2_journal_access_dq(handle, sb_dqopt(sb)->files[type],
+			od->dq_chunk->qc_headerbh, OCFS2_JOURNAL_ACCESS_WRITE);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	offset = ol_dqblk_chunk_off(sb, od->dq_chunk->qc_num,
+					     od->dq_local_off);
+	dchunk = (struct ocfs2_local_disk_chunk *)
+			(od->dq_chunk->qc_headerbh->b_data);
+	/* Mark structure as freed */
+	lock_buffer(od->dq_chunk->qc_headerbh);
+	ocfs2_clear_bit(offset, dchunk->dqc_bitmap);
+	le32_add_cpu(&dchunk->dqc_free, 1);
+	unlock_buffer(od->dq_chunk->qc_headerbh);
+	status = ocfs2_journal_dirty(handle, od->dq_chunk->qc_headerbh);
+	if (status < 0) {
+		mlog_errno(status);
+		goto out;
+	}
+	status = 0;
+out:
+	/* Clear the read bit so that next time someone uses this
+	 * dquot he reads fresh info from disk and allocates local
+	 * dquot structure */
+	clear_bit(DQ_READ_B, &dquot->dq_flags);
+	return status;
+}
+
+static struct quota_format_ops ocfs2_format_ops = {
+	.check_quota_file	= ocfs2_local_check_quota_file,
+	.read_file_info		= ocfs2_local_read_info,
+	.write_file_info	= ocfs2_global_write_info,
+	.free_file_info		= ocfs2_local_free_info,
+	.read_dqblk		= ocfs2_local_read_dquot,
+	.commit_dqblk		= ocfs2_local_write_dquot,
+	.release_dqblk		= ocfs2_local_release_dquot,
+};
+
+struct quota_format_type ocfs2_quota_format = {
+	.qf_fmt_id = QFMT_OCFS2,
+	.qf_ops = &ocfs2_format_ops,
+	.qf_owner = THIS_MODULE
+};

fs/ocfs2/resize.c

@@ -106,8 +106,8 @@ static int ocfs2_update_last_group_and_inode(handle_t *handle,
 	mlog_entry("(new_clusters=%d, first_new_cluster = %u)\n",
 		   new_clusters, first_new_cluster);
 
-	ret = ocfs2_journal_access(handle, bm_inode, group_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_gd(handle, bm_inode, group_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
@@ -141,8 +141,8 @@ static int ocfs2_update_last_group_and_inode(handle_t *handle,
 	}
 
 	/* update the inode accordingly. */
-	ret = ocfs2_journal_access(handle, bm_inode, bm_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, bm_inode, bm_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out_rollback;
@@ -314,6 +314,10 @@ int ocfs2_group_extend(struct inode * inode, int new_clusters)
 
 	fe = (struct ocfs2_dinode *)main_bm_bh->b_data;
 
+	/* main_bm_bh is validated by inode read inside ocfs2_inode_lock(),
+	 * so any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
+
 	if (le16_to_cpu(fe->id2.i_chain.cl_cpg) !=
 				 ocfs2_group_bitmap_size(osb->sb) * 8) {
 		mlog(ML_ERROR, "The disk is too old and small. "
@@ -322,30 +326,18 @@ int ocfs2_group_extend(struct inode * inode, int new_clusters)
 		goto out_unlock;
 	}
 
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		OCFS2_RO_ON_INVALID_DINODE(main_bm_inode->i_sb, fe);
-		ret = -EIO;
-		goto out_unlock;
-	}
-
 	first_new_cluster = le32_to_cpu(fe->i_clusters);
 	lgd_blkno = ocfs2_which_cluster_group(main_bm_inode,
 					      first_new_cluster - 1);
 
-	ret = ocfs2_read_block(main_bm_inode, lgd_blkno, &group_bh);
+	ret = ocfs2_read_group_descriptor(main_bm_inode, fe, lgd_blkno,
+					  &group_bh);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out_unlock;
 	}
-
 	group = (struct ocfs2_group_desc *)group_bh->b_data;
 
-	ret = ocfs2_check_group_descriptor(inode->i_sb, fe, group);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_unlock;
-	}
-
 	cl_bpc = le16_to_cpu(fe->id2.i_chain.cl_bpc);
 	if (le16_to_cpu(group->bg_bits) / cl_bpc + new_clusters >
 		le16_to_cpu(fe->id2.i_chain.cl_cpg)) {
@@ -398,41 +390,16 @@ static int ocfs2_check_new_group(struct inode *inode,
 				 struct buffer_head *group_bh)
 {
 	int ret;
-	struct ocfs2_group_desc *gd;
+	struct ocfs2_group_desc *gd =
+		(struct ocfs2_group_desc *)group_bh->b_data;
 	u16 cl_bpc = le16_to_cpu(di->id2.i_chain.cl_bpc);
-	unsigned int max_bits = le16_to_cpu(di->id2.i_chain.cl_cpg) *
-				le16_to_cpu(di->id2.i_chain.cl_bpc);
-
 
-	gd = (struct ocfs2_group_desc *)group_bh->b_data;
+	ret = ocfs2_check_group_descriptor(inode->i_sb, di, group_bh);
+	if (ret)
+		goto out;
 
-	ret = -EIO;
-	if (!OCFS2_IS_VALID_GROUP_DESC(gd))
-		mlog(ML_ERROR, "Group descriptor # %llu isn't valid.\n",
-		     (unsigned long long)le64_to_cpu(gd->bg_blkno));
-	else if (di->i_blkno != gd->bg_parent_dinode)
-		mlog(ML_ERROR, "Group descriptor # %llu has bad parent "
-		     "pointer (%llu, expected %llu)\n",
-		     (unsigned long long)le64_to_cpu(gd->bg_blkno),
-		     (unsigned long long)le64_to_cpu(gd->bg_parent_dinode),
-		     (unsigned long long)le64_to_cpu(di->i_blkno));
-	else if (le16_to_cpu(gd->bg_bits) > max_bits)
-		mlog(ML_ERROR, "Group descriptor # %llu has bit count of %u\n",
-		     (unsigned long long)le64_to_cpu(gd->bg_blkno),
-		     le16_to_cpu(gd->bg_bits));
-	else if (le16_to_cpu(gd->bg_free_bits_count) > le16_to_cpu(gd->bg_bits))
-		mlog(ML_ERROR, "Group descriptor # %llu has bit count %u but "
-		     "claims that %u are free\n",
-		     (unsigned long long)le64_to_cpu(gd->bg_blkno),
-		     le16_to_cpu(gd->bg_bits),
-		     le16_to_cpu(gd->bg_free_bits_count));
-	else if (le16_to_cpu(gd->bg_bits) > (8 * le16_to_cpu(gd->bg_size)))
-		mlog(ML_ERROR, "Group descriptor # %llu has bit count %u but "
-		     "max bitmap bits of %u\n",
-		     (unsigned long long)le64_to_cpu(gd->bg_blkno),
-		     le16_to_cpu(gd->bg_bits),
-		     8 * le16_to_cpu(gd->bg_size));
-	else if (le16_to_cpu(gd->bg_chain) != input->chain)
+	ret = -EINVAL;
+	if (le16_to_cpu(gd->bg_chain) != input->chain)
 		mlog(ML_ERROR, "Group descriptor # %llu has bad chain %u "
 		     "while input has %u set.\n",
 		     (unsigned long long)le64_to_cpu(gd->bg_blkno),
@@ -451,6 +418,7 @@ static int ocfs2_check_new_group(struct inode *inode,
 	else
 		ret = 0;
 
+out:
 	return ret;
 }
 
@@ -568,8 +536,8 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input)
 	cl = &fe->id2.i_chain;
 	cr = &cl->cl_recs[input->chain];
 
-	ret = ocfs2_journal_access(handle, main_bm_inode, group_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_gd(handle, main_bm_inode, group_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -584,8 +552,8 @@ int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input)
 		goto out_commit;
 	}
 
-	ret = ocfs2_journal_access(handle, main_bm_inode, main_bm_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, main_bm_inode, main_bm_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out_commit;

fs/ocfs2/slot_map.c

@@ -151,7 +151,7 @@ int ocfs2_refresh_slot_info(struct ocfs2_super *osb)
 	 * this is not true, the read of -1 (UINT64_MAX) will fail.
 	 */
 	ret = ocfs2_read_blocks(si->si_inode, -1, si->si_blocks, si->si_bh,
-				OCFS2_BH_IGNORE_CACHE);
+				OCFS2_BH_IGNORE_CACHE, NULL);
 	if (ret == 0) {
 		spin_lock(&osb->osb_lock);
 		ocfs2_update_slot_info(si);
@@ -405,7 +405,7 @@ static int ocfs2_map_slot_buffers(struct ocfs2_super *osb,
 
 		bh = NULL;  /* Acquire a fresh bh */
 		status = ocfs2_read_blocks(si->si_inode, blkno, 1, &bh,
-					   OCFS2_BH_IGNORE_CACHE);
+					   OCFS2_BH_IGNORE_CACHE, NULL);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;

fs/ocfs2/suballoc.c

@@ -35,6 +35,7 @@
 #include "ocfs2.h"
 
 #include "alloc.h"
+#include "blockcheck.h"
 #include "dlmglue.h"
 #include "inode.h"
 #include "journal.h"
@@ -145,62 +146,183 @@ static u32 ocfs2_bits_per_group(struct ocfs2_chain_list *cl)
 	return (u32)le16_to_cpu(cl->cl_cpg) * (u32)le16_to_cpu(cl->cl_bpc);
 }
 
-/* somewhat more expensive than our other checks, so use sparingly. */
-int ocfs2_check_group_descriptor(struct super_block *sb,
-				 struct ocfs2_dinode *di,
-				 struct ocfs2_group_desc *gd)
+#define do_error(fmt, ...)						\
+	do{								\
+		if (clean_error)					\
+			mlog(ML_ERROR, fmt "\n", ##__VA_ARGS__);	\
+		else							\
+			ocfs2_error(sb, fmt, ##__VA_ARGS__);		\
+	} while (0)
+
+static int ocfs2_validate_gd_self(struct super_block *sb,
+				  struct buffer_head *bh,
+				  int clean_error)
 {
-	unsigned int max_bits;
+	struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *)bh->b_data;
 
 	if (!OCFS2_IS_VALID_GROUP_DESC(gd)) {
-		OCFS2_RO_ON_INVALID_GROUP_DESC(sb, gd);
-		return -EIO;
+		do_error("Group descriptor #%llu has bad signature %.*s",
+			 (unsigned long long)bh->b_blocknr, 7,
+			 gd->bg_signature);
+		return -EINVAL;
 	}
 
+	if (le64_to_cpu(gd->bg_blkno) != bh->b_blocknr) {
+		do_error("Group descriptor #%llu has an invalid bg_blkno "
+			 "of %llu",
+			 (unsigned long long)bh->b_blocknr,
+			 (unsigned long long)le64_to_cpu(gd->bg_blkno));
+		return -EINVAL;
+	}
+
+	if (le32_to_cpu(gd->bg_generation) != OCFS2_SB(sb)->fs_generation) {
+		do_error("Group descriptor #%llu has an invalid "
+			 "fs_generation of #%u",
+			 (unsigned long long)bh->b_blocknr,
+			 le32_to_cpu(gd->bg_generation));
+		return -EINVAL;
+	}
+
+	if (le16_to_cpu(gd->bg_free_bits_count) > le16_to_cpu(gd->bg_bits)) {
+		do_error("Group descriptor #%llu has bit count %u but "
+			 "claims that %u are free",
+			 (unsigned long long)bh->b_blocknr,
+			 le16_to_cpu(gd->bg_bits),
+			 le16_to_cpu(gd->bg_free_bits_count));
+		return -EINVAL;
+	}
+
+	if (le16_to_cpu(gd->bg_bits) > (8 * le16_to_cpu(gd->bg_size))) {
+		do_error("Group descriptor #%llu has bit count %u but "
+			 "max bitmap bits of %u",
+			 (unsigned long long)bh->b_blocknr,
+			 le16_to_cpu(gd->bg_bits),
+			 8 * le16_to_cpu(gd->bg_size));
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int ocfs2_validate_gd_parent(struct super_block *sb,
+				    struct ocfs2_dinode *di,
+				    struct buffer_head *bh,
+				    int clean_error)
+{
+	unsigned int max_bits;
+	struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *)bh->b_data;
+
 	if (di->i_blkno != gd->bg_parent_dinode) {
-		ocfs2_error(sb, "Group descriptor # %llu has bad parent "
-			    "pointer (%llu, expected %llu)",
-			    (unsigned long long)le64_to_cpu(gd->bg_blkno),
-			    (unsigned long long)le64_to_cpu(gd->bg_parent_dinode),
-			    (unsigned long long)le64_to_cpu(di->i_blkno));
-		return -EIO;
+		do_error("Group descriptor #%llu has bad parent "
+			 "pointer (%llu, expected %llu)",
+			 (unsigned long long)bh->b_blocknr,
+			 (unsigned long long)le64_to_cpu(gd->bg_parent_dinode),
+			 (unsigned long long)le64_to_cpu(di->i_blkno));
+		return -EINVAL;
 	}
 
 	max_bits = le16_to_cpu(di->id2.i_chain.cl_cpg) * le16_to_cpu(di->id2.i_chain.cl_bpc);
 	if (le16_to_cpu(gd->bg_bits) > max_bits) {
-		ocfs2_error(sb, "Group descriptor # %llu has bit count of %u",
-			    (unsigned long long)le64_to_cpu(gd->bg_blkno),
-			    le16_to_cpu(gd->bg_bits));
-		return -EIO;
+		do_error("Group descriptor #%llu has bit count of %u",
+			 (unsigned long long)bh->b_blocknr,
+			 le16_to_cpu(gd->bg_bits));
+		return -EINVAL;
 	}
 
 	if (le16_to_cpu(gd->bg_chain) >=
 	    le16_to_cpu(di->id2.i_chain.cl_next_free_rec)) {
-		ocfs2_error(sb, "Group descriptor # %llu has bad chain %u",
-			    (unsigned long long)le64_to_cpu(gd->bg_blkno),
-			    le16_to_cpu(gd->bg_chain));
-		return -EIO;
+		do_error("Group descriptor #%llu has bad chain %u",
+			 (unsigned long long)bh->b_blocknr,
+			 le16_to_cpu(gd->bg_chain));
+		return -EINVAL;
 	}
 
-	if (le16_to_cpu(gd->bg_free_bits_count) > le16_to_cpu(gd->bg_bits)) {
-		ocfs2_error(sb, "Group descriptor # %llu has bit count %u but "
-			    "claims that %u are free",
-			    (unsigned long long)le64_to_cpu(gd->bg_blkno),
-			    le16_to_cpu(gd->bg_bits),
-			    le16_to_cpu(gd->bg_free_bits_count));
-		return -EIO;
-	}
+	return 0;
+}
 
-	if (le16_to_cpu(gd->bg_bits) > (8 * le16_to_cpu(gd->bg_size))) {
-		ocfs2_error(sb, "Group descriptor # %llu has bit count %u but "
-			    "max bitmap bits of %u",
-			    (unsigned long long)le64_to_cpu(gd->bg_blkno),
-			    le16_to_cpu(gd->bg_bits),
-			    8 * le16_to_cpu(gd->bg_size));
-		return -EIO;
+#undef do_error
+
+/*
+ * This version only prints errors.  It does not fail the filesystem, and
+ * exists only for resize.
+ */
+int ocfs2_check_group_descriptor(struct super_block *sb,
+				 struct ocfs2_dinode *di,
+				 struct buffer_head *bh)
+{
+	int rc;
+	struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *)bh->b_data;
+
+	BUG_ON(!buffer_uptodate(bh));
+
+	/*
+	 * If the ecc fails, we return the error but otherwise
+	 * leave the filesystem running.  We know any error is
+	 * local to this block.
+	 */
+	rc = ocfs2_validate_meta_ecc(sb, bh->b_data, &gd->bg_check);
+	if (rc) {
+		mlog(ML_ERROR,
+		     "Checksum failed for group descriptor %llu\n",
+		     (unsigned long long)bh->b_blocknr);
+	} else
+		rc = ocfs2_validate_gd_self(sb, bh, 1);
+	if (!rc)
+		rc = ocfs2_validate_gd_parent(sb, di, bh, 1);
+
+	return rc;
+}
+
+static int ocfs2_validate_group_descriptor(struct super_block *sb,
+					   struct buffer_head *bh)
+{
+	int rc;
+	struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *)bh->b_data;
+
+	mlog(0, "Validating group descriptor %llu\n",
+	     (unsigned long long)bh->b_blocknr);
+
+	BUG_ON(!buffer_uptodate(bh));
+
+	/*
+	 * If the ecc fails, we return the error but otherwise
+	 * leave the filesystem running.  We know any error is
+	 * local to this block.
+	 */
+	rc = ocfs2_validate_meta_ecc(sb, bh->b_data, &gd->bg_check);
+	if (rc)
+		return rc;
+
+	/*
+	 * Errors after here are fatal.
+	 */
+
+	return ocfs2_validate_gd_self(sb, bh, 0);
+}
+
+int ocfs2_read_group_descriptor(struct inode *inode, struct ocfs2_dinode *di,
+				u64 gd_blkno, struct buffer_head **bh)
+{
+	int rc;
+	struct buffer_head *tmp = *bh;
+
+	rc = ocfs2_read_block(inode, gd_blkno, &tmp,
+			      ocfs2_validate_group_descriptor);
+	if (rc)
+		goto out;
+
+	rc = ocfs2_validate_gd_parent(inode->i_sb, di, tmp, 0);
+	if (rc) {
+		brelse(tmp);
+		goto out;
 	}
 
-	return 0;
+	/* If ocfs2_read_block() got us a new bh, pass it up. */
+	if (!*bh)
+		*bh = tmp;
+
+out:
+	return rc;
 }
 
 static int ocfs2_block_group_fill(handle_t *handle,
@@ -225,10 +347,10 @@ static int ocfs2_block_group_fill(handle_t *handle,
 		goto bail;
 	}
 
-	status = ocfs2_journal_access(handle,
-				      alloc_inode,
-				      bg_bh,
-				      OCFS2_JOURNAL_ACCESS_CREATE);
+	status = ocfs2_journal_access_gd(handle,
+					 alloc_inode,
+					 bg_bh,
+					 OCFS2_JOURNAL_ACCESS_CREATE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -358,8 +480,8 @@ static int ocfs2_block_group_alloc(struct ocfs2_super *osb,
 
 	bg = (struct ocfs2_group_desc *) bg_bh->b_data;
 
-	status = ocfs2_journal_access(handle, alloc_inode,
-				      bh, OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, alloc_inode,
+					 bh, OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -441,11 +563,11 @@ static int ocfs2_reserve_suballoc_bits(struct ocfs2_super *osb,
 	ac->ac_alloc_slot = slot;
 
 	fe = (struct ocfs2_dinode *) bh->b_data;
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		OCFS2_RO_ON_INVALID_DINODE(alloc_inode->i_sb, fe);
-		status = -EIO;
-		goto bail;
-	}
+
+	/* The bh was validated by the inode read inside
+	 * ocfs2_inode_lock().  Any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
+
 	if (!(fe->i_flags & cpu_to_le32(OCFS2_CHAIN_FL))) {
 		ocfs2_error(alloc_inode->i_sb, "Invalid chain allocator %llu",
 			    (unsigned long long)le64_to_cpu(fe->i_blkno));
@@ -790,10 +912,9 @@ static int ocfs2_block_group_find_clear_bits(struct ocfs2_super *osb,
 	int offset, start, found, status = 0;
 	struct ocfs2_group_desc *bg = (struct ocfs2_group_desc *) bg_bh->b_data;
 
-	if (!OCFS2_IS_VALID_GROUP_DESC(bg)) {
-		OCFS2_RO_ON_INVALID_GROUP_DESC(osb->sb, bg);
-		return -EIO;
-	}
+	/* Callers got this descriptor from
+	 * ocfs2_read_group_descriptor().  Any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_GROUP_DESC(bg));
 
 	found = start = best_offset = best_size = 0;
 	bitmap = bg->bg_bitmap;
@@ -858,11 +979,9 @@ static inline int ocfs2_block_group_set_bits(handle_t *handle,
 
 	mlog_entry_void();
 
-	if (!OCFS2_IS_VALID_GROUP_DESC(bg)) {
-		OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, bg);
-		status = -EIO;
-		goto bail;
-	}
+	/* All callers get the descriptor via
+	 * ocfs2_read_group_descriptor().  Any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_GROUP_DESC(bg));
 	BUG_ON(le16_to_cpu(bg->bg_free_bits_count) < num_bits);
 
 	mlog(0, "block_group_set_bits: off = %u, num = %u\n", bit_off,
@@ -871,10 +990,10 @@ static inline int ocfs2_block_group_set_bits(handle_t *handle,
 	if (ocfs2_is_cluster_bitmap(alloc_inode))
 		journal_type = OCFS2_JOURNAL_ACCESS_UNDO;
 
-	status = ocfs2_journal_access(handle,
-				      alloc_inode,
-				      group_bh,
-				      journal_type);
+	status = ocfs2_journal_access_gd(handle,
+					 alloc_inode,
+					 group_bh,
+					 journal_type);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -931,21 +1050,10 @@ static int ocfs2_relink_block_group(handle_t *handle,
 	struct ocfs2_group_desc *bg = (struct ocfs2_group_desc *) bg_bh->b_data;
 	struct ocfs2_group_desc *prev_bg = (struct ocfs2_group_desc *) prev_bg_bh->b_data;
 
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		OCFS2_RO_ON_INVALID_DINODE(alloc_inode->i_sb, fe);
-		status = -EIO;
-		goto out;
-	}
-	if (!OCFS2_IS_VALID_GROUP_DESC(bg)) {
-		OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, bg);
-		status = -EIO;
-		goto out;
-	}
-	if (!OCFS2_IS_VALID_GROUP_DESC(prev_bg)) {
-		OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, prev_bg);
-		status = -EIO;
-		goto out;
-	}
+	/* The caller got these descriptors from
+	 * ocfs2_read_group_descriptor().  Any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_GROUP_DESC(bg));
+	BUG_ON(!OCFS2_IS_VALID_GROUP_DESC(prev_bg));
 
 	mlog(0, "Suballoc %llu, chain %u, move group %llu to top, prev = %llu\n",
 	     (unsigned long long)le64_to_cpu(fe->i_blkno), chain,
@@ -956,8 +1064,8 @@ static int ocfs2_relink_block_group(handle_t *handle,
 	bg_ptr = le64_to_cpu(bg->bg_next_group);
 	prev_bg_ptr = le64_to_cpu(prev_bg->bg_next_group);
 
-	status = ocfs2_journal_access(handle, alloc_inode, prev_bg_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_gd(handle, alloc_inode, prev_bg_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto out_rollback;
@@ -971,8 +1079,8 @@ static int ocfs2_relink_block_group(handle_t *handle,
 		goto out_rollback;
 	}
 
-	status = ocfs2_journal_access(handle, alloc_inode, bg_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_gd(handle, alloc_inode, bg_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto out_rollback;
@@ -986,8 +1094,8 @@ static int ocfs2_relink_block_group(handle_t *handle,
 		goto out_rollback;
 	}
 
-	status = ocfs2_journal_access(handle, alloc_inode, fe_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, alloc_inode, fe_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto out_rollback;
@@ -1008,7 +1116,7 @@ out_rollback:
 		bg->bg_next_group = cpu_to_le64(bg_ptr);
 		prev_bg->bg_next_group = cpu_to_le64(prev_bg_ptr);
 	}
-out:
+
 	mlog_exit(status);
 	return status;
 }
@@ -1138,8 +1246,8 @@ static int ocfs2_alloc_dinode_update_counts(struct inode *inode,
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *) di_bh->b_data;
 	struct ocfs2_chain_list *cl = (struct ocfs2_chain_list *) &di->id2.i_chain;
 
-	ret = ocfs2_journal_access(handle, inode, di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
@@ -1170,21 +1278,17 @@ static int ocfs2_search_one_group(struct ocfs2_alloc_context *ac,
 	u16 found;
 	struct buffer_head *group_bh = NULL;
 	struct ocfs2_group_desc *gd;
+	struct ocfs2_dinode *di = (struct ocfs2_dinode *)ac->ac_bh->b_data;
 	struct inode *alloc_inode = ac->ac_inode;
 
-	ret = ocfs2_read_block(alloc_inode, gd_blkno, &group_bh);
+	ret = ocfs2_read_group_descriptor(alloc_inode, di, gd_blkno,
+					  &group_bh);
 	if (ret < 0) {
 		mlog_errno(ret);
 		return ret;
 	}
 
 	gd = (struct ocfs2_group_desc *) group_bh->b_data;
-	if (!OCFS2_IS_VALID_GROUP_DESC(gd)) {
-		OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, gd);
-		ret = -EIO;
-		goto out;
-	}
-
 	ret = ac->ac_group_search(alloc_inode, group_bh, bits_wanted, min_bits,
 				  ac->ac_max_block, bit_off, &found);
 	if (ret < 0) {
@@ -1241,19 +1345,14 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac,
 	     bits_wanted, chain,
 	     (unsigned long long)OCFS2_I(alloc_inode)->ip_blkno);
 
-	status = ocfs2_read_block(alloc_inode,
-				  le64_to_cpu(cl->cl_recs[chain].c_blkno),
-				  &group_bh);
+	status = ocfs2_read_group_descriptor(alloc_inode, fe,
+					     le64_to_cpu(cl->cl_recs[chain].c_blkno),
+					     &group_bh);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
 	bg = (struct ocfs2_group_desc *) group_bh->b_data;
-	status = ocfs2_check_group_descriptor(alloc_inode->i_sb, fe, bg);
-	if (status) {
-		mlog_errno(status);
-		goto bail;
-	}
 
 	status = -ENOSPC;
 	/* for now, the chain search is a bit simplistic. We just use
@@ -1271,18 +1370,13 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac,
 		next_group = le64_to_cpu(bg->bg_next_group);
 		prev_group_bh = group_bh;
 		group_bh = NULL;
-		status = ocfs2_read_block(alloc_inode,
-					  next_group, &group_bh);
+		status = ocfs2_read_group_descriptor(alloc_inode, fe,
+						     next_group, &group_bh);
 		if (status < 0) {
 			mlog_errno(status);
 			goto bail;
 		}
 		bg = (struct ocfs2_group_desc *) group_bh->b_data;
-		status = ocfs2_check_group_descriptor(alloc_inode->i_sb, fe, bg);
-		if (status) {
-			mlog_errno(status);
-			goto bail;
-		}
 	}
 	if (status < 0) {
 		if (status != -ENOSPC)
@@ -1324,10 +1418,10 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac,
 
 	/* Ok, claim our bits now: set the info on dinode, chainlist
 	 * and then the group */
-	status = ocfs2_journal_access(handle,
-				      alloc_inode,
-				      ac->ac_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle,
+					 alloc_inode,
+					 ac->ac_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -1392,11 +1486,11 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb,
 	BUG_ON(!ac->ac_bh);
 
 	fe = (struct ocfs2_dinode *) ac->ac_bh->b_data;
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		OCFS2_RO_ON_INVALID_DINODE(osb->sb, fe);
-		status = -EIO;
-		goto bail;
-	}
+
+	/* The bh was validated by the inode read during
+	 * ocfs2_reserve_suballoc_bits().  Any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
+
 	if (le32_to_cpu(fe->id1.bitmap1.i_used) >=
 	    le32_to_cpu(fe->id1.bitmap1.i_total)) {
 		ocfs2_error(osb->sb, "Chain allocator dinode %llu has %u used "
@@ -1725,19 +1819,17 @@ static inline int ocfs2_block_group_clear_bits(handle_t *handle,
 
 	mlog_entry_void();
 
-	if (!OCFS2_IS_VALID_GROUP_DESC(bg)) {
-		OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, bg);
-		status = -EIO;
-		goto bail;
-	}
+	/* The caller got this descriptor from
+	 * ocfs2_read_group_descriptor().  Any corruption is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_GROUP_DESC(bg));
 
 	mlog(0, "off = %u, num = %u\n", bit_off, num_bits);
 
 	if (ocfs2_is_cluster_bitmap(alloc_inode))
 		journal_type = OCFS2_JOURNAL_ACCESS_UNDO;
 
-	status = ocfs2_journal_access(handle, alloc_inode, group_bh,
-				      journal_type);
+	status = ocfs2_journal_access_gd(handle, alloc_inode, group_bh,
+					 journal_type);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
@@ -1782,29 +1874,26 @@ int ocfs2_free_suballoc_bits(handle_t *handle,
 
 	mlog_entry_void();
 
-	if (!OCFS2_IS_VALID_DINODE(fe)) {
-		OCFS2_RO_ON_INVALID_DINODE(alloc_inode->i_sb, fe);
-		status = -EIO;
-		goto bail;
-	}
+	/* The alloc_bh comes from ocfs2_free_dinode() or
+	 * ocfs2_free_clusters().  The callers have all locked the
+	 * allocator and gotten alloc_bh from the lock call.  This
+	 * validates the dinode buffer.  Any corruption that has happended
+	 * is a code bug. */
+	BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
 	BUG_ON((count + start_bit) > ocfs2_bits_per_group(cl));
 
 	mlog(0, "%llu: freeing %u bits from group %llu, starting at %u\n",
 	     (unsigned long long)OCFS2_I(alloc_inode)->ip_blkno, count,
 	     (unsigned long long)bg_blkno, start_bit);
 
-	status = ocfs2_read_block(alloc_inode, bg_blkno, &group_bh);
+	status = ocfs2_read_group_descriptor(alloc_inode, fe, bg_blkno,
+					     &group_bh);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;
 	}
-
 	group = (struct ocfs2_group_desc *) group_bh->b_data;
-	status = ocfs2_check_group_descriptor(alloc_inode->i_sb, fe, group);
-	if (status) {
-		mlog_errno(status);
-		goto bail;
-	}
+
 	BUG_ON((count + start_bit) > le16_to_cpu(group->bg_bits));
 
 	status = ocfs2_block_group_clear_bits(handle, alloc_inode,
@@ -1815,8 +1904,8 @@ int ocfs2_free_suballoc_bits(handle_t *handle,
 		goto bail;
 	}
 
-	status = ocfs2_journal_access(handle, alloc_inode, alloc_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = ocfs2_journal_access_di(handle, alloc_inode, alloc_bh,
+					 OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto bail;

fs/ocfs2/suballoc.h

@@ -164,10 +164,24 @@ void ocfs2_free_ac_resource(struct ocfs2_alloc_context *ac);
  * and return that block offset. */
 u64 ocfs2_which_cluster_group(struct inode *inode, u32 cluster);
 
-/* somewhat more expensive than our other checks, so use sparingly. */
+/*
+ * By default, ocfs2_read_group_descriptor() calls ocfs2_error() when it
+ * finds a problem.  A caller that wants to check a group descriptor
+ * without going readonly should read the block with ocfs2_read_block[s]()
+ * and then checking it with this function.  This is only resize, really.
+ * Everyone else should be using ocfs2_read_group_descriptor().
+ */
 int ocfs2_check_group_descriptor(struct super_block *sb,
 				 struct ocfs2_dinode *di,
-				 struct ocfs2_group_desc *gd);
+				 struct buffer_head *bh);
+/*
+ * Read a group descriptor block into *bh.  If *bh is NULL, a bh will be
+ * allocated.  This is a cached read.  The descriptor will be validated with
+ * ocfs2_validate_group_descriptor().
+ */
+int ocfs2_read_group_descriptor(struct inode *inode, struct ocfs2_dinode *di,
+				u64 gd_blkno, struct buffer_head **bh);
+
 int ocfs2_lock_allocators(struct inode *inode, struct ocfs2_extent_tree *et,
 			  u32 clusters_to_add, u32 extents_to_split,
 			  struct ocfs2_alloc_context **data_ac,

fs/ocfs2/super.c

@@ -41,6 +41,7 @@
 #include <linux/debugfs.h>
 #include <linux/mount.h>
 #include <linux/seq_file.h>
+#include <linux/quotaops.h>
 
 #define MLOG_MASK_PREFIX ML_SUPER
 #include <cluster/masklog.h>
@@ -51,6 +52,7 @@
 #include "ocfs1_fs_compat.h"
 
 #include "alloc.h"
+#include "blockcheck.h"
 #include "dlmglue.h"
 #include "export.h"
 #include "extent_map.h"
@@ -65,10 +67,13 @@
 #include "uptodate.h"
 #include "ver.h"
 #include "xattr.h"
+#include "quota.h"
 
 #include "buffer_head_io.h"
 
 static struct kmem_cache *ocfs2_inode_cachep = NULL;
+struct kmem_cache *ocfs2_dquot_cachep;
+struct kmem_cache *ocfs2_qf_chunk_cachep;
 
 /* OCFS2 needs to schedule several differnt types of work which
  * require cluster locking, disk I/O, recovery waits, etc. Since these
@@ -124,6 +129,9 @@ static int ocfs2_get_sector(struct super_block *sb,
 static void ocfs2_write_super(struct super_block *sb);
 static struct inode *ocfs2_alloc_inode(struct super_block *sb);
 static void ocfs2_destroy_inode(struct inode *inode);
+static int ocfs2_susp_quotas(struct ocfs2_super *osb, int unsuspend);
+static int ocfs2_enable_quotas(struct ocfs2_super *osb);
+static void ocfs2_disable_quotas(struct ocfs2_super *osb);
 
 static const struct super_operations ocfs2_sops = {
 	.statfs		= ocfs2_statfs,
@@ -137,6 +145,8 @@ static const struct super_operations ocfs2_sops = {
 	.put_super	= ocfs2_put_super,
 	.remount_fs	= ocfs2_remount,
 	.show_options   = ocfs2_show_options,
+	.quota_read	= ocfs2_quota_read,
+	.quota_write	= ocfs2_quota_write,
 };
 
 enum {
@@ -158,6 +168,10 @@ enum {
 	Opt_user_xattr,
 	Opt_nouser_xattr,
 	Opt_inode64,
+	Opt_acl,
+	Opt_noacl,
+	Opt_usrquota,
+	Opt_grpquota,
 	Opt_err,
 };
 
@@ -180,6 +194,10 @@ static const match_table_t tokens = {
 	{Opt_user_xattr, "user_xattr"},
 	{Opt_nouser_xattr, "nouser_xattr"},
 	{Opt_inode64, "inode64"},
+	{Opt_acl, "acl"},
+	{Opt_noacl, "noacl"},
+	{Opt_usrquota, "usrquota"},
+	{Opt_grpquota, "grpquota"},
 	{Opt_err, NULL}
 };
 
@@ -221,6 +239,19 @@ static int ocfs2_sync_fs(struct super_block *sb, int wait)
 	return 0;
 }
 
+static int ocfs2_need_system_inode(struct ocfs2_super *osb, int ino)
+{
+	if (!OCFS2_HAS_RO_COMPAT_FEATURE(osb->sb, OCFS2_FEATURE_RO_COMPAT_USRQUOTA)
+	    && (ino == USER_QUOTA_SYSTEM_INODE
+		|| ino == LOCAL_USER_QUOTA_SYSTEM_INODE))
+		return 0;
+	if (!OCFS2_HAS_RO_COMPAT_FEATURE(osb->sb, OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)
+	    && (ino == GROUP_QUOTA_SYSTEM_INODE
+		|| ino == LOCAL_GROUP_QUOTA_SYSTEM_INODE))
+		return 0;
+	return 1;
+}
+
 static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb)
 {
 	struct inode *new = NULL;
@@ -247,6 +278,8 @@ static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb)
 
 	for (i = OCFS2_FIRST_ONLINE_SYSTEM_INODE;
 	     i <= OCFS2_LAST_GLOBAL_SYSTEM_INODE; i++) {
+		if (!ocfs2_need_system_inode(osb, i))
+			continue;
 		new = ocfs2_get_system_file_inode(osb, i, osb->slot_num);
 		if (!new) {
 			ocfs2_release_system_inodes(osb);
@@ -277,6 +310,8 @@ static int ocfs2_init_local_system_inodes(struct ocfs2_super *osb)
 	for (i = OCFS2_LAST_GLOBAL_SYSTEM_INODE + 1;
 	     i < NUM_SYSTEM_INODES;
 	     i++) {
+		if (!ocfs2_need_system_inode(osb, i))
+			continue;
 		new = ocfs2_get_system_file_inode(osb, i, osb->slot_num);
 		if (!new) {
 			ocfs2_release_system_inodes(osb);
@@ -426,6 +461,12 @@ static int ocfs2_remount(struct super_block *sb, int *flags, char *data)
 
 	/* We're going to/from readonly mode. */
 	if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY)) {
+		/* Disable quota accounting before remounting RO */
+		if (*flags & MS_RDONLY) {
+			ret = ocfs2_susp_quotas(osb, 0);
+			if (ret < 0)
+				goto out;
+		}
 		/* Lock here so the check of HARD_RO and the potential
 		 * setting of SOFT_RO is atomic. */
 		spin_lock(&osb->osb_lock);
@@ -461,11 +502,28 @@ static int ocfs2_remount(struct super_block *sb, int *flags, char *data)
 		}
 unlock_osb:
 		spin_unlock(&osb->osb_lock);
+		/* Enable quota accounting after remounting RW */
+		if (!ret && !(*flags & MS_RDONLY)) {
+			if (sb_any_quota_suspended(sb))
+				ret = ocfs2_susp_quotas(osb, 1);
+			else
+				ret = ocfs2_enable_quotas(osb);
+			if (ret < 0) {
+				/* Return back changes... */
+				spin_lock(&osb->osb_lock);
+				sb->s_flags |= MS_RDONLY;
+				osb->osb_flags |= OCFS2_OSB_SOFT_RO;
+				spin_unlock(&osb->osb_lock);
+				goto out;
+			}
+		}
 	}
 
 	if (!ret) {
 		/* Only save off the new mount options in case of a successful
 		 * remount. */
+		if (!(osb->s_feature_incompat & OCFS2_FEATURE_INCOMPAT_XATTR))
+			parsed_options.mount_opt &= ~OCFS2_MOUNT_POSIX_ACL;
 		osb->s_mount_opt = parsed_options.mount_opt;
 		osb->s_atime_quantum = parsed_options.atime_quantum;
 		osb->preferred_slot = parsed_options.slot;
@@ -619,6 +677,131 @@ static int ocfs2_verify_userspace_stack(struct ocfs2_super *osb,
 	return 0;
 }
 
+static int ocfs2_susp_quotas(struct ocfs2_super *osb, int unsuspend)
+{
+	int type;
+	struct super_block *sb = osb->sb;
+	unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
+					     OCFS2_FEATURE_RO_COMPAT_GRPQUOTA};
+	int status = 0;
+
+	for (type = 0; type < MAXQUOTAS; type++) {
+		if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, feature[type]))
+			continue;
+		if (unsuspend)
+			status = vfs_quota_enable(
+					sb_dqopt(sb)->files[type],
+					type, QFMT_OCFS2,
+					DQUOT_SUSPENDED);
+		else
+			status = vfs_quota_disable(sb, type,
+						   DQUOT_SUSPENDED);
+		if (status < 0)
+			break;
+	}
+	if (status < 0)
+		mlog(ML_ERROR, "Failed to suspend/unsuspend quotas on "
+		     "remount (error = %d).\n", status);
+	return status;
+}
+
+static int ocfs2_enable_quotas(struct ocfs2_super *osb)
+{
+	struct inode *inode[MAXQUOTAS] = { NULL, NULL };
+	struct super_block *sb = osb->sb;
+	unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
+					     OCFS2_FEATURE_RO_COMPAT_GRPQUOTA};
+	unsigned int ino[MAXQUOTAS] = { LOCAL_USER_QUOTA_SYSTEM_INODE,
+					LOCAL_GROUP_QUOTA_SYSTEM_INODE };
+	int status;
+	int type;
+
+	sb_dqopt(sb)->flags |= DQUOT_QUOTA_SYS_FILE | DQUOT_NEGATIVE_USAGE;
+	for (type = 0; type < MAXQUOTAS; type++) {
+		if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, feature[type]))
+			continue;
+		inode[type] = ocfs2_get_system_file_inode(osb, ino[type],
+							osb->slot_num);
+		if (!inode[type]) {
+			status = -ENOENT;
+			goto out_quota_off;
+		}
+		status = vfs_quota_enable(inode[type], type, QFMT_OCFS2,
+						DQUOT_USAGE_ENABLED);
+		if (status < 0)
+			goto out_quota_off;
+	}
+
+	for (type = 0; type < MAXQUOTAS; type++)
+		iput(inode[type]);
+	return 0;
+out_quota_off:
+	ocfs2_disable_quotas(osb);
+	for (type = 0; type < MAXQUOTAS; type++)
+		iput(inode[type]);
+	mlog_errno(status);
+	return status;
+}
+
+static void ocfs2_disable_quotas(struct ocfs2_super *osb)
+{
+	int type;
+	struct inode *inode;
+	struct super_block *sb = osb->sb;
+
+	/* We mostly ignore errors in this function because there's not much
+	 * we can do when we see them */
+	for (type = 0; type < MAXQUOTAS; type++) {
+		if (!sb_has_quota_loaded(sb, type))
+			continue;
+		inode = igrab(sb->s_dquot.files[type]);
+		/* Turn off quotas. This will remove all dquot structures from
+		 * memory and so they will be automatically synced to global
+		 * quota files */
+		vfs_quota_disable(sb, type, DQUOT_USAGE_ENABLED |
+					    DQUOT_LIMITS_ENABLED);
+		if (!inode)
+			continue;
+		iput(inode);
+	}
+}
+
+/* Handle quota on quotactl */
+static int ocfs2_quota_on(struct super_block *sb, int type, int format_id,
+			  char *path, int remount)
+{
+	unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
+					     OCFS2_FEATURE_RO_COMPAT_GRPQUOTA};
+
+	if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, feature[type]))
+		return -EINVAL;
+
+	if (remount)
+		return 0;	/* Just ignore it has been handled in
+				 * ocfs2_remount() */
+	return vfs_quota_enable(sb_dqopt(sb)->files[type], type,
+				    format_id, DQUOT_LIMITS_ENABLED);
+}
+
+/* Handle quota off quotactl */
+static int ocfs2_quota_off(struct super_block *sb, int type, int remount)
+{
+	if (remount)
+		return 0;	/* Ignore now and handle later in
+				 * ocfs2_remount() */
+	return vfs_quota_disable(sb, type, DQUOT_LIMITS_ENABLED);
+}
+
+static struct quotactl_ops ocfs2_quotactl_ops = {
+	.quota_on	= ocfs2_quota_on,
+	.quota_off	= ocfs2_quota_off,
+	.quota_sync	= vfs_quota_sync,
+	.get_info	= vfs_get_dqinfo,
+	.set_info	= vfs_set_dqinfo,
+	.get_dqblk	= vfs_get_dqblk,
+	.set_dqblk	= vfs_set_dqblk,
+};
+
 static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 {
 	struct dentry *root;
@@ -651,12 +834,32 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 	}
 	brelse(bh);
 	bh = NULL;
+
+	if (!(osb->s_feature_incompat & OCFS2_FEATURE_INCOMPAT_XATTR))
+		parsed_options.mount_opt &= ~OCFS2_MOUNT_POSIX_ACL;
+
 	osb->s_mount_opt = parsed_options.mount_opt;
 	osb->s_atime_quantum = parsed_options.atime_quantum;
 	osb->preferred_slot = parsed_options.slot;
 	osb->osb_commit_interval = parsed_options.commit_interval;
 	osb->local_alloc_default_bits = ocfs2_megabytes_to_clusters(sb, parsed_options.localalloc_opt);
 	osb->local_alloc_bits = osb->local_alloc_default_bits;
+	if (osb->s_mount_opt & OCFS2_MOUNT_USRQUOTA &&
+	    !OCFS2_HAS_RO_COMPAT_FEATURE(sb,
+					 OCFS2_FEATURE_RO_COMPAT_USRQUOTA)) {
+		status = -EINVAL;
+		mlog(ML_ERROR, "User quotas were requested, but this "
+		     "filesystem does not have the feature enabled.\n");
+		goto read_super_error;
+	}
+	if (osb->s_mount_opt & OCFS2_MOUNT_GRPQUOTA &&
+	    !OCFS2_HAS_RO_COMPAT_FEATURE(sb,
+					 OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)) {
+		status = -EINVAL;
+		mlog(ML_ERROR, "Group quotas were requested, but this "
+		     "filesystem does not have the feature enabled.\n");
+		goto read_super_error;
+	}
 
 	status = ocfs2_verify_userspace_stack(osb, &parsed_options);
 	if (status)
@@ -664,6 +867,9 @@ 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) |
+		((osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
+
 	/* Hard readonly mode only if: bdev_read_only, MS_RDONLY,
 	 * heartbeat=none */
 	if (bdev_read_only(sb->s_bdev)) {
@@ -758,6 +964,28 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
 	atomic_set(&osb->vol_state, VOLUME_MOUNTED);
 	wake_up(&osb->osb_mount_event);
 
+	/* Now we can initialize quotas because we can afford to wait
+	 * for cluster locks recovery now. That also means that truncation
+	 * log recovery can happen but that waits for proper quota setup */
+	if (!(sb->s_flags & MS_RDONLY)) {
+		status = ocfs2_enable_quotas(osb);
+		if (status < 0) {
+			/* We have to err-out specially here because
+			 * s_root is already set */
+			mlog_errno(status);
+			atomic_set(&osb->vol_state, VOLUME_DISABLED);
+			wake_up(&osb->osb_mount_event);
+			mlog_exit(status);
+			return status;
+		}
+	}
+
+	ocfs2_complete_quota_recovery(osb);
+
+	/* Now we wake up again for processes waiting for quotas */
+	atomic_set(&osb->vol_state, VOLUME_MOUNTED_QUOTAS);
+	wake_up(&osb->osb_mount_event);
+
 	mlog_exit(status);
 	return status;
 
@@ -945,6 +1173,41 @@ static int ocfs2_parse_options(struct super_block *sb,
 		case Opt_inode64:
 			mopt->mount_opt |= OCFS2_MOUNT_INODE64;
 			break;
+		case Opt_usrquota:
+			/* We check only on remount, otherwise features
+			 * aren't yet initialized. */
+			if (is_remount && !OCFS2_HAS_RO_COMPAT_FEATURE(sb,
+			    OCFS2_FEATURE_RO_COMPAT_USRQUOTA)) {
+				mlog(ML_ERROR, "User quota requested but "
+				     "filesystem feature is not set\n");
+				status = 0;
+				goto bail;
+			}
+			mopt->mount_opt |= OCFS2_MOUNT_USRQUOTA;
+			break;
+		case Opt_grpquota:
+			if (is_remount && !OCFS2_HAS_RO_COMPAT_FEATURE(sb,
+			    OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)) {
+				mlog(ML_ERROR, "Group quota requested but "
+				     "filesystem feature is not set\n");
+				status = 0;
+				goto bail;
+			}
+			mopt->mount_opt |= OCFS2_MOUNT_GRPQUOTA;
+			break;
+#ifdef CONFIG_OCFS2_FS_POSIX_ACL
+		case Opt_acl:
+			mopt->mount_opt |= OCFS2_MOUNT_POSIX_ACL;
+			break;
+		case Opt_noacl:
+			mopt->mount_opt &= ~OCFS2_MOUNT_POSIX_ACL;
+			break;
+#else
+		case Opt_acl:
+		case Opt_noacl:
+			printk(KERN_INFO "ocfs2 (no)acl options not supported\n");
+			break;
+#endif
 		default:
 			mlog(ML_ERROR,
 			     "Unrecognized mount option \"%s\" "
@@ -1008,6 +1271,10 @@ static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt)
 	if (osb->osb_cluster_stack[0])
 		seq_printf(s, ",cluster_stack=%.*s", OCFS2_STACK_LABEL_LEN,
 			   osb->osb_cluster_stack);
+	if (opts & OCFS2_MOUNT_USRQUOTA)
+		seq_printf(s, ",usrquota");
+	if (opts & OCFS2_MOUNT_GRPQUOTA)
+		seq_printf(s, ",grpquota");
 
 	if (opts & OCFS2_MOUNT_NOUSERXATTR)
 		seq_printf(s, ",nouser_xattr");
@@ -1017,6 +1284,13 @@ static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt)
 	if (opts & OCFS2_MOUNT_INODE64)
 		seq_printf(s, ",inode64");
 
+#ifdef CONFIG_OCFS2_FS_POSIX_ACL
+	if (opts & OCFS2_MOUNT_POSIX_ACL)
+		seq_printf(s, ",acl");
+	else
+		seq_printf(s, ",noacl");
+#endif
+
 	return 0;
 }
 
@@ -1052,10 +1326,16 @@ static int __init ocfs2_init(void)
 		mlog(ML_ERROR, "Unable to create ocfs2 debugfs root.\n");
 	}
 
+	status = ocfs2_quota_setup();
+	if (status)
+		goto leave;
+
 	ocfs2_set_locking_protocol();
 
+	status = register_quota_format(&ocfs2_quota_format);
 leave:
 	if (status < 0) {
+		ocfs2_quota_shutdown();
 		ocfs2_free_mem_caches();
 		exit_ocfs2_uptodate_cache();
 	}
@@ -1072,11 +1352,15 @@ static void __exit ocfs2_exit(void)
 {
 	mlog_entry_void();
 
+	ocfs2_quota_shutdown();
+
 	if (ocfs2_wq) {
 		flush_workqueue(ocfs2_wq);
 		destroy_workqueue(ocfs2_wq);
 	}
 
+	unregister_quota_format(&ocfs2_quota_format);
+
 	debugfs_remove(ocfs2_debugfs_root);
 
 	ocfs2_free_mem_caches();
@@ -1192,8 +1476,27 @@ static int ocfs2_initialize_mem_caches(void)
 				       (SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT|
 						SLAB_MEM_SPREAD),
 				       ocfs2_inode_init_once);
-	if (!ocfs2_inode_cachep)
+	ocfs2_dquot_cachep = kmem_cache_create("ocfs2_dquot_cache",
+					sizeof(struct ocfs2_dquot),
+					0,
+					(SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT|
+						SLAB_MEM_SPREAD),
+					NULL);
+	ocfs2_qf_chunk_cachep = kmem_cache_create("ocfs2_qf_chunk_cache",
+					sizeof(struct ocfs2_quota_chunk),
+					0,
+					(SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD),
+					NULL);
+	if (!ocfs2_inode_cachep || !ocfs2_dquot_cachep ||
+	    !ocfs2_qf_chunk_cachep) {
+		if (ocfs2_inode_cachep)
+			kmem_cache_destroy(ocfs2_inode_cachep);
+		if (ocfs2_dquot_cachep)
+			kmem_cache_destroy(ocfs2_dquot_cachep);
+		if (ocfs2_qf_chunk_cachep)
+			kmem_cache_destroy(ocfs2_qf_chunk_cachep);
 		return -ENOMEM;
+	}
 
 	return 0;
 }
@@ -1202,8 +1505,15 @@ static void ocfs2_free_mem_caches(void)
 {
 	if (ocfs2_inode_cachep)
 		kmem_cache_destroy(ocfs2_inode_cachep);
-
 	ocfs2_inode_cachep = NULL;
+
+	if (ocfs2_dquot_cachep)
+		kmem_cache_destroy(ocfs2_dquot_cachep);
+	ocfs2_dquot_cachep = NULL;
+
+	if (ocfs2_qf_chunk_cachep)
+		kmem_cache_destroy(ocfs2_qf_chunk_cachep);
+	ocfs2_qf_chunk_cachep = NULL;
 }
 
 static int ocfs2_get_sector(struct super_block *sb,
@@ -1303,6 +1613,8 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
 	osb = OCFS2_SB(sb);
 	BUG_ON(!osb);
 
+	ocfs2_disable_quotas(osb);
+
 	ocfs2_shutdown_local_alloc(osb);
 
 	ocfs2_truncate_log_shutdown(osb);
@@ -1413,6 +1725,8 @@ static int ocfs2_initialize_super(struct super_block *sb,
 	sb->s_fs_info = osb;
 	sb->s_op = &ocfs2_sops;
 	sb->s_export_op = &ocfs2_export_ops;
+	sb->s_qcop = &ocfs2_quotactl_ops;
+	sb->dq_op = &ocfs2_quota_operations;
 	sb->s_xattr = ocfs2_xattr_handlers;
 	sb->s_time_gran = 1;
 	sb->s_flags |= MS_NOATIME;
@@ -1676,6 +1990,15 @@ static int ocfs2_verify_volume(struct ocfs2_dinode *di,
 
 	if (memcmp(di->i_signature, OCFS2_SUPER_BLOCK_SIGNATURE,
 		   strlen(OCFS2_SUPER_BLOCK_SIGNATURE)) == 0) {
+		/* We have to do a raw check of the feature here */
+		if (le32_to_cpu(di->id2.i_super.s_feature_incompat) &
+		    OCFS2_FEATURE_INCOMPAT_META_ECC) {
+			status = ocfs2_block_check_validate(bh->b_data,
+							    bh->b_size,
+							    &di->i_check);
+			if (status)
+				goto out;
+		}
 		status = -EINVAL;
 		if ((1 << le32_to_cpu(di->id2.i_super.s_blocksize_bits)) != blksz) {
 			mlog(ML_ERROR, "found superblock with incorrect block "
@@ -1717,6 +2040,7 @@ static int ocfs2_verify_volume(struct ocfs2_dinode *di,
 		}
 	}
 
+out:
 	mlog_exit(status);
 	return status;
 }

fs/ocfs2/symlink.c

@@ -84,7 +84,7 @@ static char *ocfs2_fast_symlink_getlink(struct inode *inode,
 
 	mlog_entry_void();
 
-	status = ocfs2_read_block(inode, OCFS2_I(inode)->ip_blkno, bh);
+	status = ocfs2_read_inode_block(inode, bh);
 	if (status < 0) {
 		mlog_errno(status);
 		link = ERR_PTR(status);

fs/ocfs2/xattr.c

@@ -35,12 +35,14 @@
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/string.h>
+#include <linux/security.h>
 
 #define MLOG_MASK_PREFIX ML_XATTR
 #include <cluster/masklog.h>
 
 #include "ocfs2.h"
 #include "alloc.h"
+#include "blockcheck.h"
 #include "dlmglue.h"
 #include "file.h"
 #include "symlink.h"
@@ -61,12 +63,32 @@ struct ocfs2_xattr_def_value_root {
 };
 
 struct ocfs2_xattr_bucket {
-	struct buffer_head *bhs[OCFS2_XATTR_MAX_BLOCKS_PER_BUCKET];
-	struct ocfs2_xattr_header *xh;
+	/* The inode these xattrs are associated with */
+	struct inode *bu_inode;
+
+	/* The actual buffers that make up the bucket */
+	struct buffer_head *bu_bhs[OCFS2_XATTR_MAX_BLOCKS_PER_BUCKET];
+
+	/* How many blocks make up one bucket for this filesystem */
+	int bu_blocks;
+};
+
+struct ocfs2_xattr_set_ctxt {
+	handle_t *handle;
+	struct ocfs2_alloc_context *meta_ac;
+	struct ocfs2_alloc_context *data_ac;
+	struct ocfs2_cached_dealloc_ctxt dealloc;
 };
 
 #define OCFS2_XATTR_ROOT_SIZE	(sizeof(struct ocfs2_xattr_def_value_root))
 #define OCFS2_XATTR_INLINE_SIZE	80
+#define OCFS2_XATTR_FREE_IN_IBODY	(OCFS2_MIN_XATTR_INLINE_SIZE \
+					 - sizeof(struct ocfs2_xattr_header) \
+					 - sizeof(__u32))
+#define OCFS2_XATTR_FREE_IN_BLOCK(ptr)	((ptr)->i_sb->s_blocksize \
+					 - sizeof(struct ocfs2_xattr_block) \
+					 - sizeof(struct ocfs2_xattr_header) \
+					 - sizeof(__u32))
 
 static struct ocfs2_xattr_def_value_root def_xv = {
 	.xv.xr_list.l_count = cpu_to_le16(1),
@@ -74,13 +96,25 @@ static struct ocfs2_xattr_def_value_root def_xv = {
 
 struct xattr_handler *ocfs2_xattr_handlers[] = {
 	&ocfs2_xattr_user_handler,
+#ifdef CONFIG_OCFS2_FS_POSIX_ACL
+	&ocfs2_xattr_acl_access_handler,
+	&ocfs2_xattr_acl_default_handler,
+#endif
 	&ocfs2_xattr_trusted_handler,
+	&ocfs2_xattr_security_handler,
 	NULL
 };
 
 static struct xattr_handler *ocfs2_xattr_handler_map[OCFS2_XATTR_MAX] = {
 	[OCFS2_XATTR_INDEX_USER]	= &ocfs2_xattr_user_handler,
+#ifdef CONFIG_OCFS2_FS_POSIX_ACL
+	[OCFS2_XATTR_INDEX_POSIX_ACL_ACCESS]
+					= &ocfs2_xattr_acl_access_handler,
+	[OCFS2_XATTR_INDEX_POSIX_ACL_DEFAULT]
+					= &ocfs2_xattr_acl_default_handler,
+#endif
 	[OCFS2_XATTR_INDEX_TRUSTED]	= &ocfs2_xattr_trusted_handler,
+	[OCFS2_XATTR_INDEX_SECURITY]	= &ocfs2_xattr_security_handler,
 };
 
 struct ocfs2_xattr_info {
@@ -98,7 +132,7 @@ struct ocfs2_xattr_search {
 	 */
 	struct buffer_head *xattr_bh;
 	struct ocfs2_xattr_header *header;
-	struct ocfs2_xattr_bucket bucket;
+	struct ocfs2_xattr_bucket *bucket;
 	void *base;
 	void *end;
 	struct ocfs2_xattr_entry *here;
@@ -127,14 +161,20 @@ static int ocfs2_xattr_tree_list_index_block(struct inode *inode,
 					size_t buffer_size);
 
 static int ocfs2_xattr_create_index_block(struct inode *inode,
-					  struct ocfs2_xattr_search *xs);
+					  struct ocfs2_xattr_search *xs,
+					  struct ocfs2_xattr_set_ctxt *ctxt);
 
 static int ocfs2_xattr_set_entry_index_block(struct inode *inode,
 					     struct ocfs2_xattr_info *xi,
-					     struct ocfs2_xattr_search *xs);
+					     struct ocfs2_xattr_search *xs,
+					     struct ocfs2_xattr_set_ctxt *ctxt);
 
 static int ocfs2_delete_xattr_index_block(struct inode *inode,
 					  struct buffer_head *xb_bh);
+static int ocfs2_mv_xattr_buckets(struct inode *inode, handle_t *handle,
+				  u64 src_blk, u64 last_blk, u64 to_blk,
+				  unsigned int start_bucket,
+				  u32 *first_hash);
 
 static inline u16 ocfs2_xattr_buckets_per_cluster(struct ocfs2_super *osb)
 {
@@ -154,6 +194,216 @@ static inline u16 ocfs2_xattr_max_xe_in_bucket(struct super_block *sb)
 	return len / sizeof(struct ocfs2_xattr_entry);
 }
 
+#define bucket_blkno(_b) ((_b)->bu_bhs[0]->b_blocknr)
+#define bucket_block(_b, _n) ((_b)->bu_bhs[(_n)]->b_data)
+#define bucket_xh(_b) ((struct ocfs2_xattr_header *)bucket_block((_b), 0))
+
+static struct ocfs2_xattr_bucket *ocfs2_xattr_bucket_new(struct inode *inode)
+{
+	struct ocfs2_xattr_bucket *bucket;
+	int blks = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+
+	BUG_ON(blks > OCFS2_XATTR_MAX_BLOCKS_PER_BUCKET);
+
+	bucket = kzalloc(sizeof(struct ocfs2_xattr_bucket), GFP_NOFS);
+	if (bucket) {
+		bucket->bu_inode = inode;
+		bucket->bu_blocks = blks;
+	}
+
+	return bucket;
+}
+
+static void ocfs2_xattr_bucket_relse(struct ocfs2_xattr_bucket *bucket)
+{
+	int i;
+
+	for (i = 0; i < bucket->bu_blocks; i++) {
+		brelse(bucket->bu_bhs[i]);
+		bucket->bu_bhs[i] = NULL;
+	}
+}
+
+static void ocfs2_xattr_bucket_free(struct ocfs2_xattr_bucket *bucket)
+{
+	if (bucket) {
+		ocfs2_xattr_bucket_relse(bucket);
+		bucket->bu_inode = NULL;
+		kfree(bucket);
+	}
+}
+
+/*
+ * A bucket that has never been written to disk doesn't need to be
+ * read.  We just need the buffer_heads.  Don't call this for
+ * buckets that are already on disk.  ocfs2_read_xattr_bucket() initializes
+ * them fully.
+ */
+static int ocfs2_init_xattr_bucket(struct ocfs2_xattr_bucket *bucket,
+				   u64 xb_blkno)
+{
+	int i, rc = 0;
+
+	for (i = 0; i < bucket->bu_blocks; i++) {
+		bucket->bu_bhs[i] = sb_getblk(bucket->bu_inode->i_sb,
+					      xb_blkno + i);
+		if (!bucket->bu_bhs[i]) {
+			rc = -EIO;
+			mlog_errno(rc);
+			break;
+		}
+
+		if (!ocfs2_buffer_uptodate(bucket->bu_inode,
+					   bucket->bu_bhs[i]))
+			ocfs2_set_new_buffer_uptodate(bucket->bu_inode,
+						      bucket->bu_bhs[i]);
+	}
+
+	if (rc)
+		ocfs2_xattr_bucket_relse(bucket);
+	return rc;
+}
+
+/* Read the xattr bucket at xb_blkno */
+static int ocfs2_read_xattr_bucket(struct ocfs2_xattr_bucket *bucket,
+				   u64 xb_blkno)
+{
+	int rc;
+
+	rc = ocfs2_read_blocks(bucket->bu_inode, xb_blkno,
+			       bucket->bu_blocks, bucket->bu_bhs, 0,
+			       NULL);
+	if (!rc) {
+		rc = ocfs2_validate_meta_ecc_bhs(bucket->bu_inode->i_sb,
+						 bucket->bu_bhs,
+						 bucket->bu_blocks,
+						 &bucket_xh(bucket)->xh_check);
+		if (rc)
+			mlog_errno(rc);
+	}
+
+	if (rc)
+		ocfs2_xattr_bucket_relse(bucket);
+	return rc;
+}
+
+static int ocfs2_xattr_bucket_journal_access(handle_t *handle,
+					     struct ocfs2_xattr_bucket *bucket,
+					     int type)
+{
+	int i, rc = 0;
+
+	for (i = 0; i < bucket->bu_blocks; i++) {
+		rc = ocfs2_journal_access(handle, bucket->bu_inode,
+					  bucket->bu_bhs[i], type);
+		if (rc) {
+			mlog_errno(rc);
+			break;
+		}
+	}
+
+	return rc;
+}
+
+static void ocfs2_xattr_bucket_journal_dirty(handle_t *handle,
+					     struct ocfs2_xattr_bucket *bucket)
+{
+	int i;
+
+	ocfs2_compute_meta_ecc_bhs(bucket->bu_inode->i_sb,
+				   bucket->bu_bhs, bucket->bu_blocks,
+				   &bucket_xh(bucket)->xh_check);
+
+	for (i = 0; i < bucket->bu_blocks; i++)
+		ocfs2_journal_dirty(handle, bucket->bu_bhs[i]);
+}
+
+static void ocfs2_xattr_bucket_copy_data(struct ocfs2_xattr_bucket *dest,
+					 struct ocfs2_xattr_bucket *src)
+{
+	int i;
+	int blocksize = src->bu_inode->i_sb->s_blocksize;
+
+	BUG_ON(dest->bu_blocks != src->bu_blocks);
+	BUG_ON(dest->bu_inode != src->bu_inode);
+
+	for (i = 0; i < src->bu_blocks; i++) {
+		memcpy(bucket_block(dest, i), bucket_block(src, i),
+		       blocksize);
+	}
+}
+
+static int ocfs2_validate_xattr_block(struct super_block *sb,
+				      struct buffer_head *bh)
+{
+	int rc;
+	struct ocfs2_xattr_block *xb =
+		(struct ocfs2_xattr_block *)bh->b_data;
+
+	mlog(0, "Validating xattr block %llu\n",
+	     (unsigned long long)bh->b_blocknr);
+
+	BUG_ON(!buffer_uptodate(bh));
+
+	/*
+	 * If the ecc fails, we return the error but otherwise
+	 * leave the filesystem running.  We know any error is
+	 * local to this block.
+	 */
+	rc = ocfs2_validate_meta_ecc(sb, bh->b_data, &xb->xb_check);
+	if (rc)
+		return rc;
+
+	/*
+	 * Errors after here are fatal
+	 */
+
+	if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) {
+		ocfs2_error(sb,
+			    "Extended attribute block #%llu has bad "
+			    "signature %.*s",
+			    (unsigned long long)bh->b_blocknr, 7,
+			    xb->xb_signature);
+		return -EINVAL;
+	}
+
+	if (le64_to_cpu(xb->xb_blkno) != bh->b_blocknr) {
+		ocfs2_error(sb,
+			    "Extended attribute block #%llu has an "
+			    "invalid xb_blkno of %llu",
+			    (unsigned long long)bh->b_blocknr,
+			    (unsigned long long)le64_to_cpu(xb->xb_blkno));
+		return -EINVAL;
+	}
+
+	if (le32_to_cpu(xb->xb_fs_generation) != OCFS2_SB(sb)->fs_generation) {
+		ocfs2_error(sb,
+			    "Extended attribute block #%llu has an invalid "
+			    "xb_fs_generation of #%u",
+			    (unsigned long long)bh->b_blocknr,
+			    le32_to_cpu(xb->xb_fs_generation));
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int ocfs2_read_xattr_block(struct inode *inode, u64 xb_blkno,
+				  struct buffer_head **bh)
+{
+	int rc;
+	struct buffer_head *tmp = *bh;
+
+	rc = ocfs2_read_block(inode, xb_blkno, &tmp,
+			      ocfs2_validate_xattr_block);
+
+	/* If ocfs2_read_block() got us a new bh, pass it up. */
+	if (!rc && !*bh)
+		*bh = tmp;
+
+	return rc;
+}
+
 static inline const char *ocfs2_xattr_prefix(int name_index)
 {
 	struct xattr_handler *handler = NULL;
@@ -200,54 +450,163 @@ static void ocfs2_xattr_hash_entry(struct inode *inode,
 	return;
 }
 
+static int ocfs2_xattr_entry_real_size(int name_len, size_t value_len)
+{
+	int size = 0;
+
+	if (value_len <= OCFS2_XATTR_INLINE_SIZE)
+		size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_SIZE(value_len);
+	else
+		size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
+	size += sizeof(struct ocfs2_xattr_entry);
+
+	return size;
+}
+
+int ocfs2_calc_security_init(struct inode *dir,
+			     struct ocfs2_security_xattr_info *si,
+			     int *want_clusters,
+			     int *xattr_credits,
+			     struct ocfs2_alloc_context **xattr_ac)
+{
+	int ret = 0;
+	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
+	int s_size = ocfs2_xattr_entry_real_size(strlen(si->name),
+						 si->value_len);
+
+	/*
+	 * The max space of security xattr taken inline is
+	 * 256(name) + 80(value) + 16(entry) = 352 bytes,
+	 * So reserve one metadata block for it is ok.
+	 */
+	if (dir->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE ||
+	    s_size > OCFS2_XATTR_FREE_IN_IBODY) {
+		ret = ocfs2_reserve_new_metadata_blocks(osb, 1, xattr_ac);
+		if (ret) {
+			mlog_errno(ret);
+			return ret;
+		}
+		*xattr_credits += OCFS2_XATTR_BLOCK_CREATE_CREDITS;
+	}
+
+	/* reserve clusters for xattr value which will be set in B tree*/
+	if (si->value_len > OCFS2_XATTR_INLINE_SIZE) {
+		int new_clusters = ocfs2_clusters_for_bytes(dir->i_sb,
+							    si->value_len);
+
+		*xattr_credits += ocfs2_clusters_to_blocks(dir->i_sb,
+							   new_clusters);
+		*want_clusters += new_clusters;
+	}
+	return ret;
+}
+
+int ocfs2_calc_xattr_init(struct inode *dir,
+			  struct buffer_head *dir_bh,
+			  int mode,
+			  struct ocfs2_security_xattr_info *si,
+			  int *want_clusters,
+			  int *xattr_credits,
+			  struct ocfs2_alloc_context **xattr_ac)
+{
+	int ret = 0;
+	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
+	int s_size = 0, a_size = 0, acl_len = 0, new_clusters;
+
+	if (si->enable)
+		s_size = ocfs2_xattr_entry_real_size(strlen(si->name),
+						     si->value_len);
+
+	if (osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) {
+		acl_len = ocfs2_xattr_get_nolock(dir, dir_bh,
+					OCFS2_XATTR_INDEX_POSIX_ACL_DEFAULT,
+					"", NULL, 0);
+		if (acl_len > 0) {
+			a_size = ocfs2_xattr_entry_real_size(0, acl_len);
+			if (S_ISDIR(mode))
+				a_size <<= 1;
+		} else if (acl_len != 0 && acl_len != -ENODATA) {
+			mlog_errno(ret);
+			return ret;
+		}
+	}
+
+	if (!(s_size + a_size))
+		return ret;
+
+	/*
+	 * The max space of security xattr taken inline is
+	 * 256(name) + 80(value) + 16(entry) = 352 bytes,
+	 * The max space of acl xattr taken inline is
+	 * 80(value) + 16(entry) * 2(if directory) = 192 bytes,
+	 * when blocksize = 512, may reserve one more cluser for
+	 * xattr bucket, otherwise reserve one metadata block
+	 * for them is ok.
+	 */
+	if (dir->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE ||
+	    (s_size + a_size) > OCFS2_XATTR_FREE_IN_IBODY) {
+		ret = ocfs2_reserve_new_metadata_blocks(osb, 1, xattr_ac);
+		if (ret) {
+			mlog_errno(ret);
+			return ret;
+		}
+		*xattr_credits += OCFS2_XATTR_BLOCK_CREATE_CREDITS;
+	}
+
+	if (dir->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE &&
+	    (s_size + a_size) > OCFS2_XATTR_FREE_IN_BLOCK(dir)) {
+		*want_clusters += 1;
+		*xattr_credits += ocfs2_blocks_per_xattr_bucket(dir->i_sb);
+	}
+
+	/*
+	 * reserve credits and clusters for xattrs which has large value
+	 * and have to be set outside
+	 */
+	if (si->enable && si->value_len > OCFS2_XATTR_INLINE_SIZE) {
+		new_clusters = ocfs2_clusters_for_bytes(dir->i_sb,
+							si->value_len);
+		*xattr_credits += ocfs2_clusters_to_blocks(dir->i_sb,
+							   new_clusters);
+		*want_clusters += new_clusters;
+	}
+	if (osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL &&
+	    acl_len > OCFS2_XATTR_INLINE_SIZE) {
+		/* for directory, it has DEFAULT and ACCESS two types of acls */
+		new_clusters = (S_ISDIR(mode) ? 2 : 1) *
+				ocfs2_clusters_for_bytes(dir->i_sb, acl_len);
+		*xattr_credits += ocfs2_clusters_to_blocks(dir->i_sb,
+							   new_clusters);
+		*want_clusters += new_clusters;
+	}
+
+	return ret;
+}
+
 static int ocfs2_xattr_extend_allocation(struct inode *inode,
 					 u32 clusters_to_add,
-					 struct buffer_head *xattr_bh,
-					 struct ocfs2_xattr_value_root *xv)
+					 struct ocfs2_xattr_value_buf *vb,
+					 struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	int status = 0;
-	int restart_func = 0;
-	int credits = 0;
-	handle_t *handle = NULL;
-	struct ocfs2_alloc_context *data_ac = NULL;
-	struct ocfs2_alloc_context *meta_ac = NULL;
+	handle_t *handle = ctxt->handle;
 	enum ocfs2_alloc_restarted why;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	u32 prev_clusters, logical_start = le32_to_cpu(xv->xr_clusters);
+	u32 prev_clusters, logical_start = le32_to_cpu(vb->vb_xv->xr_clusters);
 	struct ocfs2_extent_tree et;
 
 	mlog(0, "(clusters_to_add for xattr= %u)\n", clusters_to_add);
 
-	ocfs2_init_xattr_value_extent_tree(&et, inode, xattr_bh, xv);
-
-restart_all:
-
-	status = ocfs2_lock_allocators(inode, &et, clusters_to_add, 0,
-				       &data_ac, &meta_ac);
-	if (status) {
-		mlog_errno(status);
-		goto leave;
-	}
-
-	credits = ocfs2_calc_extend_credits(osb->sb, et.et_root_el,
-					    clusters_to_add);
-	handle = ocfs2_start_trans(osb, credits);
-	if (IS_ERR(handle)) {
-		status = PTR_ERR(handle);
-		handle = NULL;
-		mlog_errno(status);
-		goto leave;
-	}
+	ocfs2_init_xattr_value_extent_tree(&et, inode, vb);
 
-restarted_transaction:
-	status = ocfs2_journal_access(handle, inode, xattr_bh,
-				      OCFS2_JOURNAL_ACCESS_WRITE);
+	status = vb->vb_access(handle, inode, vb->vb_bh,
+			      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
 	}
 
-	prev_clusters = le32_to_cpu(xv->xr_clusters);
+	prev_clusters = le32_to_cpu(vb->vb_xv->xr_clusters);
 	status = ocfs2_add_clusters_in_btree(osb,
 					     inode,
 					     &logical_start,
@@ -255,157 +614,84 @@ restarted_transaction:
 					     0,
 					     &et,
 					     handle,
-					     data_ac,
-					     meta_ac,
+					     ctxt->data_ac,
+					     ctxt->meta_ac,
 					     &why);
-	if ((status < 0) && (status != -EAGAIN)) {
-		if (status != -ENOSPC)
-			mlog_errno(status);
+	if (status < 0) {
+		mlog_errno(status);
 		goto leave;
 	}
 
-	status = ocfs2_journal_dirty(handle, xattr_bh);
+	status = ocfs2_journal_dirty(handle, vb->vb_bh);
 	if (status < 0) {
 		mlog_errno(status);
 		goto leave;
 	}
 
-	clusters_to_add -= le32_to_cpu(xv->xr_clusters) - prev_clusters;
+	clusters_to_add -= le32_to_cpu(vb->vb_xv->xr_clusters) - prev_clusters;
 
-	if (why != RESTART_NONE && clusters_to_add) {
-		if (why == RESTART_META) {
-			mlog(0, "restarting function.\n");
-			restart_func = 1;
-		} else {
-			BUG_ON(why != RESTART_TRANS);
-
-			mlog(0, "restarting transaction.\n");
-			/* TODO: This can be more intelligent. */
-			credits = ocfs2_calc_extend_credits(osb->sb,
-							    et.et_root_el,
-							    clusters_to_add);
-			status = ocfs2_extend_trans(handle, credits);
-			if (status < 0) {
-				/* handle still has to be committed at
-				 * this point. */
-				status = -ENOMEM;
-				mlog_errno(status);
-				goto leave;
-			}
-			goto restarted_transaction;
-		}
-	}
+	/*
+	 * We should have already allocated enough space before the transaction,
+	 * so no need to restart.
+	 */
+	BUG_ON(why != RESTART_NONE || clusters_to_add);
 
 leave:
-	if (handle) {
-		ocfs2_commit_trans(osb, handle);
-		handle = NULL;
-	}
-	if (data_ac) {
-		ocfs2_free_alloc_context(data_ac);
-		data_ac = NULL;
-	}
-	if (meta_ac) {
-		ocfs2_free_alloc_context(meta_ac);
-		meta_ac = NULL;
-	}
-	if ((!status) && restart_func) {
-		restart_func = 0;
-		goto restart_all;
-	}
 
 	return status;
 }
 
 static int __ocfs2_remove_xattr_range(struct inode *inode,
-				      struct buffer_head *root_bh,
-				      struct ocfs2_xattr_value_root *xv,
+				      struct ocfs2_xattr_value_buf *vb,
 				      u32 cpos, u32 phys_cpos, u32 len,
-				      struct ocfs2_cached_dealloc_ctxt *dealloc)
+				      struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	int ret;
 	u64 phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos);
-	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	struct inode *tl_inode = osb->osb_tl_inode;
-	handle_t *handle;
-	struct ocfs2_alloc_context *meta_ac = NULL;
+	handle_t *handle = ctxt->handle;
 	struct ocfs2_extent_tree et;
 
-	ocfs2_init_xattr_value_extent_tree(&et, inode, root_bh, xv);
+	ocfs2_init_xattr_value_extent_tree(&et, inode, vb);
 
-	ret = ocfs2_lock_allocators(inode, &et, 0, 1, NULL, &meta_ac);
+	ret = vb->vb_access(handle, inode, vb->vb_bh,
+			    OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
-		mlog_errno(ret);
-		return ret;
-	}
-
-	mutex_lock(&tl_inode->i_mutex);
-
-	if (ocfs2_truncate_log_needs_flush(osb)) {
-		ret = __ocfs2_flush_truncate_log(osb);
-		if (ret < 0) {
-			mlog_errno(ret);
-			goto out;
-		}
-	}
-
-	handle = ocfs2_start_trans(osb, OCFS2_REMOVE_EXTENT_CREDITS);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, root_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
-
-	ret = ocfs2_remove_extent(inode, &et, cpos, len, handle, meta_ac,
-				  dealloc);
+	ret = ocfs2_remove_extent(inode, &et, cpos, len, handle, ctxt->meta_ac,
+				  &ctxt->dealloc);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 
-	le32_add_cpu(&xv->xr_clusters, -len);
+	le32_add_cpu(&vb->vb_xv->xr_clusters, -len);
 
-	ret = ocfs2_journal_dirty(handle, root_bh);
+	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 
-	ret = ocfs2_truncate_log_append(osb, handle, phys_blkno, len);
+	ret = ocfs2_cache_cluster_dealloc(&ctxt->dealloc, phys_blkno, len);
 	if (ret)
 		mlog_errno(ret);
 
-out_commit:
-	ocfs2_commit_trans(osb, handle);
 out:
-	mutex_unlock(&tl_inode->i_mutex);
-
-	if (meta_ac)
-		ocfs2_free_alloc_context(meta_ac);
-
 	return ret;
 }
 
 static int ocfs2_xattr_shrink_size(struct inode *inode,
 				   u32 old_clusters,
 				   u32 new_clusters,
-				   struct buffer_head *root_bh,
-				   struct ocfs2_xattr_value_root *xv)
+				   struct ocfs2_xattr_value_buf *vb,
+				   struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	int ret = 0;
 	u32 trunc_len, cpos, phys_cpos, alloc_size;
 	u64 block;
-	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	struct ocfs2_cached_dealloc_ctxt dealloc;
-
-	ocfs2_init_dealloc_ctxt(&dealloc);
 
 	if (old_clusters <= new_clusters)
 		return 0;
@@ -414,7 +700,8 @@ static int ocfs2_xattr_shrink_size(struct inode *inode,
 	trunc_len = old_clusters - new_clusters;
 	while (trunc_len) {
 		ret = ocfs2_xattr_get_clusters(inode, cpos, &phys_cpos,
-					       &alloc_size, &xv->xr_list);
+					       &alloc_size,
+					       &vb->vb_xv->xr_list);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -423,9 +710,9 @@ static int ocfs2_xattr_shrink_size(struct inode *inode,
 		if (alloc_size > trunc_len)
 			alloc_size = trunc_len;
 
-		ret = __ocfs2_remove_xattr_range(inode, root_bh, xv, cpos,
+		ret = __ocfs2_remove_xattr_range(inode, vb, cpos,
 						 phys_cpos, alloc_size,
-						 &dealloc);
+						 ctxt);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
@@ -439,20 +726,17 @@ static int ocfs2_xattr_shrink_size(struct inode *inode,
 	}
 
 out:
-	ocfs2_schedule_truncate_log_flush(osb, 1);
-	ocfs2_run_deallocs(osb, &dealloc);
-
 	return ret;
 }
 
 static int ocfs2_xattr_value_truncate(struct inode *inode,
-				      struct buffer_head *root_bh,
-				      struct ocfs2_xattr_value_root *xv,
-				      int len)
+				      struct ocfs2_xattr_value_buf *vb,
+				      int len,
+				      struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	int ret;
 	u32 new_clusters = ocfs2_clusters_for_bytes(inode->i_sb, len);
-	u32 old_clusters = le32_to_cpu(xv->xr_clusters);
+	u32 old_clusters = le32_to_cpu(vb->vb_xv->xr_clusters);
 
 	if (new_clusters == old_clusters)
 		return 0;
@@ -460,11 +744,11 @@ static int ocfs2_xattr_value_truncate(struct inode *inode,
 	if (new_clusters > old_clusters)
 		ret = ocfs2_xattr_extend_allocation(inode,
 						    new_clusters - old_clusters,
-						    root_bh, xv);
+						    vb, ctxt);
 	else
 		ret = ocfs2_xattr_shrink_size(inode,
 					      old_clusters, new_clusters,
-					      root_bh, xv);
+					      vb, ctxt);
 
 	return ret;
 }
@@ -554,18 +838,14 @@ static int ocfs2_xattr_block_list(struct inode *inode,
 	if (!di->i_xattr_loc)
 		return ret;
 
-	ret = ocfs2_read_block(inode, le64_to_cpu(di->i_xattr_loc), &blk_bh);
+	ret = ocfs2_read_xattr_block(inode, le64_to_cpu(di->i_xattr_loc),
+				     &blk_bh);
 	if (ret < 0) {
 		mlog_errno(ret);
 		return ret;
 	}
 
 	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
-	if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) {
-		ret = -EIO;
-		goto cleanup;
-	}
-
 	if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) {
 		struct ocfs2_xattr_header *header = &xb->xb_attrs.xb_header;
 		ret = ocfs2_xattr_list_entries(inode, header,
@@ -575,7 +855,7 @@ static int ocfs2_xattr_block_list(struct inode *inode,
 		ret = ocfs2_xattr_tree_list_index_block(inode, xt,
 						   buffer, buffer_size);
 	}
-cleanup:
+
 	brelse(blk_bh);
 
 	return ret;
@@ -685,7 +965,7 @@ static int ocfs2_xattr_get_value_outside(struct inode *inode,
 		blkno = ocfs2_clusters_to_blocks(inode->i_sb, p_cluster);
 		/* Copy ocfs2_xattr_value */
 		for (i = 0; i < num_clusters * bpc; i++, blkno++) {
-			ret = ocfs2_read_block(inode, blkno, &bh);
+			ret = ocfs2_read_block(inode, blkno, &bh, NULL);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
@@ -769,7 +1049,12 @@ static int ocfs2_xattr_block_get(struct inode *inode,
 	size_t size;
 	int ret = -ENODATA, name_offset, name_len, block_off, i;
 
-	memset(&xs->bucket, 0, sizeof(xs->bucket));
+	xs->bucket = ocfs2_xattr_bucket_new(inode);
+	if (!xs->bucket) {
+		ret = -ENOMEM;
+		mlog_errno(ret);
+		goto cleanup;
+	}
 
 	ret = ocfs2_xattr_block_find(inode, name_index, name, xs);
 	if (ret) {
@@ -795,11 +1080,11 @@ static int ocfs2_xattr_block_get(struct inode *inode,
 
 		if (le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED) {
 			ret = ocfs2_xattr_bucket_get_name_value(inode,
-								xs->bucket.xh,
+								bucket_xh(xs->bucket),
 								i,
 								&block_off,
 								&name_offset);
-			xs->base = xs->bucket.bhs[block_off]->b_data;
+			xs->base = bucket_block(xs->bucket, block_off);
 		}
 		if (ocfs2_xattr_is_local(xs->here)) {
 			memcpy(buffer, (void *)xs->base +
@@ -817,21 +1102,15 @@ static int ocfs2_xattr_block_get(struct inode *inode,
 	}
 	ret = size;
 cleanup:
-	for (i = 0; i < OCFS2_XATTR_MAX_BLOCKS_PER_BUCKET; i++)
-		brelse(xs->bucket.bhs[i]);
-	memset(&xs->bucket, 0, sizeof(xs->bucket));
+	ocfs2_xattr_bucket_free(xs->bucket);
 
 	brelse(xs->xattr_bh);
 	xs->xattr_bh = NULL;
 	return ret;
 }
 
-/* ocfs2_xattr_get()
- *
- * Copy an extended attribute into the buffer provided.
- * Buffer is NULL to compute the size of buffer required.
- */
-static int ocfs2_xattr_get(struct inode *inode,
+int ocfs2_xattr_get_nolock(struct inode *inode,
+			   struct buffer_head *di_bh,
 			   int name_index,
 			   const char *name,
 			   void *buffer,
@@ -839,7 +1118,6 @@ static int ocfs2_xattr_get(struct inode *inode,
 {
 	int ret;
 	struct ocfs2_dinode *di = NULL;
-	struct buffer_head *di_bh = NULL;
 	struct ocfs2_inode_info *oi = OCFS2_I(inode);
 	struct ocfs2_xattr_search xis = {
 		.not_found = -ENODATA,
@@ -854,11 +1132,6 @@ static int ocfs2_xattr_get(struct inode *inode,
 	if (!(oi->ip_dyn_features & OCFS2_HAS_XATTR_FL))
 		ret = -ENODATA;
 
-	ret = ocfs2_inode_lock(inode, &di_bh, 0);
-	if (ret < 0) {
-		mlog_errno(ret);
-		return ret;
-	}
 	xis.inode_bh = xbs.inode_bh = di_bh;
 	di = (struct ocfs2_dinode *)di_bh->b_data;
 
@@ -869,52 +1142,70 @@ static int ocfs2_xattr_get(struct inode *inode,
 		ret = ocfs2_xattr_block_get(inode, name_index, name, buffer,
 					    buffer_size, &xbs);
 	up_read(&oi->ip_xattr_sem);
-	ocfs2_inode_unlock(inode, 0);
-
-	brelse(di_bh);
 
 	return ret;
 }
 
-static int __ocfs2_xattr_set_value_outside(struct inode *inode,
-					   struct ocfs2_xattr_value_root *xv,
-					   const void *value,
-					   int value_len)
+/* ocfs2_xattr_get()
+ *
+ * Copy an extended attribute into the buffer provided.
+ * Buffer is NULL to compute the size of buffer required.
+ */
+static int ocfs2_xattr_get(struct inode *inode,
+			   int name_index,
+			   const char *name,
+			   void *buffer,
+			   size_t buffer_size)
 {
-	int ret = 0, i, cp_len, credits;
-	u16 blocksize = inode->i_sb->s_blocksize;
-	u32 p_cluster, num_clusters;
-	u32 cpos = 0, bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
+	int ret;
+	struct buffer_head *di_bh = NULL;
+
+	ret = ocfs2_inode_lock(inode, &di_bh, 0);
+	if (ret < 0) {
+		mlog_errno(ret);
+		return ret;
+	}
+	ret = ocfs2_xattr_get_nolock(inode, di_bh, name_index,
+				     name, buffer, buffer_size);
+
+	ocfs2_inode_unlock(inode, 0);
+
+	brelse(di_bh);
+
+	return ret;
+}
+
+static int __ocfs2_xattr_set_value_outside(struct inode *inode,
+					   handle_t *handle,
+					   struct ocfs2_xattr_value_root *xv,
+					   const void *value,
+					   int value_len)
+{
+	int ret = 0, i, cp_len;
+	u16 blocksize = inode->i_sb->s_blocksize;
+	u32 p_cluster, num_clusters;
+	u32 cpos = 0, bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
 	u32 clusters = ocfs2_clusters_for_bytes(inode->i_sb, value_len);
 	u64 blkno;
 	struct buffer_head *bh = NULL;
-	handle_t *handle;
 
 	BUG_ON(clusters > le32_to_cpu(xv->xr_clusters));
 
-	credits = clusters * bpc;
-	handle = ocfs2_start_trans(OCFS2_SB(inode->i_sb), credits);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		mlog_errno(ret);
-		goto out;
-	}
-
 	while (cpos < clusters) {
 		ret = ocfs2_xattr_get_clusters(inode, cpos, &p_cluster,
 					       &num_clusters, &xv->xr_list);
 		if (ret) {
 			mlog_errno(ret);
-			goto out_commit;
+			goto out;
 		}
 
 		blkno = ocfs2_clusters_to_blocks(inode->i_sb, p_cluster);
 
 		for (i = 0; i < num_clusters * bpc; i++, blkno++) {
-			ret = ocfs2_read_block(inode, blkno, &bh);
+			ret = ocfs2_read_block(inode, blkno, &bh, NULL);
 			if (ret) {
 				mlog_errno(ret);
-				goto out_commit;
+				goto out;
 			}
 
 			ret = ocfs2_journal_access(handle,
@@ -923,7 +1214,7 @@ static int __ocfs2_xattr_set_value_outside(struct inode *inode,
 						   OCFS2_JOURNAL_ACCESS_WRITE);
 			if (ret < 0) {
 				mlog_errno(ret);
-				goto out_commit;
+				goto out;
 			}
 
 			cp_len = value_len > blocksize ? blocksize : value_len;
@@ -937,7 +1228,7 @@ static int __ocfs2_xattr_set_value_outside(struct inode *inode,
 			ret = ocfs2_journal_dirty(handle, bh);
 			if (ret < 0) {
 				mlog_errno(ret);
-				goto out_commit;
+				goto out;
 			}
 			brelse(bh);
 			bh = NULL;
@@ -951,8 +1242,6 @@ static int __ocfs2_xattr_set_value_outside(struct inode *inode,
 		}
 		cpos += num_clusters;
 	}
-out_commit:
-	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
 out:
 	brelse(bh);
 
@@ -960,28 +1249,22 @@ out:
 }
 
 static int ocfs2_xattr_cleanup(struct inode *inode,
+			       handle_t *handle,
 			       struct ocfs2_xattr_info *xi,
 			       struct ocfs2_xattr_search *xs,
+			       struct ocfs2_xattr_value_buf *vb,
 			       size_t offs)
 {
-	handle_t *handle = NULL;
 	int ret = 0;
 	size_t name_len = strlen(xi->name);
 	void *val = xs->base + offs;
 	size_t size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
 
-	handle = ocfs2_start_trans((OCFS2_SB(inode->i_sb)),
-				   OCFS2_XATTR_BLOCK_UPDATE_CREDITS);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		mlog_errno(ret);
-		goto out;
-	}
-	ret = ocfs2_journal_access(handle, inode, xs->xattr_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = vb->vb_access(handle, inode, vb->vb_bh,
+			    OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 	/* Decrease xattr count */
 	le16_add_cpu(&xs->header->xh_count, -1);
@@ -989,35 +1272,27 @@ static int ocfs2_xattr_cleanup(struct inode *inode,
 	memset((void *)xs->here, 0, sizeof(struct ocfs2_xattr_entry));
 	memset(val, 0, size);
 
-	ret = ocfs2_journal_dirty(handle, xs->xattr_bh);
+	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
 	if (ret < 0)
 		mlog_errno(ret);
-out_commit:
-	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
 out:
 	return ret;
 }
 
 static int ocfs2_xattr_update_entry(struct inode *inode,
+				    handle_t *handle,
 				    struct ocfs2_xattr_info *xi,
 				    struct ocfs2_xattr_search *xs,
+				    struct ocfs2_xattr_value_buf *vb,
 				    size_t offs)
 {
-	handle_t *handle = NULL;
-	int ret = 0;
+	int ret;
 
-	handle = ocfs2_start_trans((OCFS2_SB(inode->i_sb)),
-				   OCFS2_XATTR_BLOCK_UPDATE_CREDITS);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		mlog_errno(ret);
-		goto out;
-	}
-	ret = ocfs2_journal_access(handle, inode, xs->xattr_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = vb->vb_access(handle, inode, vb->vb_bh,
+			    OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 
 	xs->here->xe_name_offset = cpu_to_le16(offs);
@@ -1028,11 +1303,9 @@ static int ocfs2_xattr_update_entry(struct inode *inode,
 		ocfs2_xattr_set_local(xs->here, 0);
 	ocfs2_xattr_hash_entry(inode, xs->header, xs->here);
 
-	ret = ocfs2_journal_dirty(handle, xs->xattr_bh);
+	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
 	if (ret < 0)
 		mlog_errno(ret);
-out_commit:
-	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
 out:
 	return ret;
 }
@@ -1045,6 +1318,8 @@ out:
 static int ocfs2_xattr_set_value_outside(struct inode *inode,
 					 struct ocfs2_xattr_info *xi,
 					 struct ocfs2_xattr_search *xs,
+					 struct ocfs2_xattr_set_ctxt *ctxt,
+					 struct ocfs2_xattr_value_buf *vb,
 					 size_t offs)
 {
 	size_t name_len = strlen(xi->name);
@@ -1062,20 +1337,20 @@ static int ocfs2_xattr_set_value_outside(struct inode *inode,
 	xv->xr_list.l_tree_depth = 0;
 	xv->xr_list.l_count = cpu_to_le16(1);
 	xv->xr_list.l_next_free_rec = 0;
+	vb->vb_xv = xv;
 
-	ret = ocfs2_xattr_value_truncate(inode, xs->xattr_bh, xv,
-					 xi->value_len);
+	ret = ocfs2_xattr_value_truncate(inode, vb, xi->value_len, ctxt);
 	if (ret < 0) {
 		mlog_errno(ret);
 		return ret;
 	}
-	ret = __ocfs2_xattr_set_value_outside(inode, xv, xi->value,
-					      xi->value_len);
+	ret = ocfs2_xattr_update_entry(inode, ctxt->handle, xi, xs, vb, offs);
 	if (ret < 0) {
 		mlog_errno(ret);
 		return ret;
 	}
-	ret = ocfs2_xattr_update_entry(inode, xi, xs, offs);
+	ret = __ocfs2_xattr_set_value_outside(inode, ctxt->handle, vb->vb_xv,
+					      xi->value, xi->value_len);
 	if (ret < 0)
 		mlog_errno(ret);
 
@@ -1195,6 +1470,7 @@ static void ocfs2_xattr_set_entry_local(struct inode *inode,
 static int ocfs2_xattr_set_entry(struct inode *inode,
 				 struct ocfs2_xattr_info *xi,
 				 struct ocfs2_xattr_search *xs,
+				 struct ocfs2_xattr_set_ctxt *ctxt,
 				 int flag)
 {
 	struct ocfs2_xattr_entry *last;
@@ -1202,7 +1478,7 @@ static int ocfs2_xattr_set_entry(struct inode *inode,
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
 	size_t min_offs = xs->end - xs->base, name_len = strlen(xi->name);
 	size_t size_l = 0;
-	handle_t *handle = NULL;
+	handle_t *handle = ctxt->handle;
 	int free, i, ret;
 	struct ocfs2_xattr_info xi_l = {
 		.name_index = xi->name_index,
@@ -1210,6 +1486,16 @@ static int ocfs2_xattr_set_entry(struct inode *inode,
 		.value = xi->value,
 		.value_len = xi->value_len,
 	};
+	struct ocfs2_xattr_value_buf vb = {
+		.vb_bh = xs->xattr_bh,
+		.vb_access = ocfs2_journal_access_di,
+	};
+
+	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
+		BUG_ON(xs->xattr_bh == xs->inode_bh);
+		vb.vb_access = ocfs2_journal_access_xb;
+	} else
+		BUG_ON(xs->xattr_bh != xs->inode_bh);
 
 	/* Compute min_offs, last and free space. */
 	last = xs->header->xh_entries;
@@ -1265,15 +1551,14 @@ static int ocfs2_xattr_set_entry(struct inode *inode,
 		if (ocfs2_xattr_is_local(xs->here) && size == size_l) {
 			/* Replace existing local xattr with tree root */
 			ret = ocfs2_xattr_set_value_outside(inode, xi, xs,
-							    offs);
+							    ctxt, &vb, offs);
 			if (ret < 0)
 				mlog_errno(ret);
 			goto out;
 		} else if (!ocfs2_xattr_is_local(xs->here)) {
 			/* For existing xattr which has value outside */
-			struct ocfs2_xattr_value_root *xv = NULL;
-			xv = (struct ocfs2_xattr_value_root *)(val +
-				OCFS2_XATTR_SIZE(name_len));
+			vb.vb_xv = (struct ocfs2_xattr_value_root *)
+				(val + OCFS2_XATTR_SIZE(name_len));
 
 			if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
 				/*
@@ -1282,27 +1567,30 @@ static int ocfs2_xattr_set_entry(struct inode *inode,
 				 * then set new value with set_value_outside().
 				 */
 				ret = ocfs2_xattr_value_truncate(inode,
-								 xs->xattr_bh,
-								 xv,
-								 xi->value_len);
+								 &vb,
+								 xi->value_len,
+								 ctxt);
 				if (ret < 0) {
 					mlog_errno(ret);
 					goto out;
 				}
 
-				ret = __ocfs2_xattr_set_value_outside(inode,
-								xv,
-								xi->value,
-								xi->value_len);
+				ret = ocfs2_xattr_update_entry(inode,
+							       handle,
+							       xi,
+							       xs,
+							       &vb,
+							       offs);
 				if (ret < 0) {
 					mlog_errno(ret);
 					goto out;
 				}
 
-				ret = ocfs2_xattr_update_entry(inode,
-							       xi,
-							       xs,
-							       offs);
+				ret = __ocfs2_xattr_set_value_outside(inode,
+								handle,
+								vb.vb_xv,
+								xi->value,
+								xi->value_len);
 				if (ret < 0)
 					mlog_errno(ret);
 				goto out;
@@ -1312,44 +1600,28 @@ static int ocfs2_xattr_set_entry(struct inode *inode,
 				 * just trucate old value to zero.
 				 */
 				 ret = ocfs2_xattr_value_truncate(inode,
-								 xs->xattr_bh,
-								 xv,
-								 0);
+								  &vb,
+								  0,
+								  ctxt);
 				if (ret < 0)
 					mlog_errno(ret);
 			}
 		}
 	}
 
-	handle = ocfs2_start_trans((OCFS2_SB(inode->i_sb)),
-				   OCFS2_INODE_UPDATE_CREDITS);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		mlog_errno(ret);
-		goto out;
-	}
-
-	ret = ocfs2_journal_access(handle, inode, xs->inode_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, xs->inode_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 
 	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
-		/* set extended attribute in external block. */
-		ret = ocfs2_extend_trans(handle,
-					 OCFS2_INODE_UPDATE_CREDITS +
-					 OCFS2_XATTR_BLOCK_UPDATE_CREDITS);
-		if (ret) {
-			mlog_errno(ret);
-			goto out_commit;
-		}
-		ret = ocfs2_journal_access(handle, inode, xs->xattr_bh,
-					   OCFS2_JOURNAL_ACCESS_WRITE);
+		ret = vb.vb_access(handle, inode, vb.vb_bh,
+				   OCFS2_JOURNAL_ACCESS_WRITE);
 		if (ret) {
 			mlog_errno(ret);
-			goto out_commit;
+			goto out;
 		}
 	}
 
@@ -1363,7 +1635,7 @@ static int ocfs2_xattr_set_entry(struct inode *inode,
 		ret = ocfs2_journal_dirty(handle, xs->xattr_bh);
 		if (ret < 0) {
 			mlog_errno(ret);
-			goto out_commit;
+			goto out;
 		}
 	}
 
@@ -1391,25 +1663,19 @@ static int ocfs2_xattr_set_entry(struct inode *inode,
 	oi->ip_dyn_features |= flag;
 	di->i_dyn_features = cpu_to_le16(oi->ip_dyn_features);
 	spin_unlock(&oi->ip_lock);
-	/* Update inode ctime */
-	inode->i_ctime = CURRENT_TIME;
-	di->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec);
-	di->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec);
 
 	ret = ocfs2_journal_dirty(handle, xs->inode_bh);
 	if (ret < 0)
 		mlog_errno(ret);
 
-out_commit:
-	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
-
 	if (!ret && xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
 		/*
 		 * Set value outside in B tree.
 		 * This is the second step for value size > INLINE_SIZE.
 		 */
 		size_t offs = le16_to_cpu(xs->here->xe_name_offset);
-		ret = ocfs2_xattr_set_value_outside(inode, xi, xs, offs);
+		ret = ocfs2_xattr_set_value_outside(inode, xi, xs, ctxt,
+						    &vb, offs);
 		if (ret < 0) {
 			int ret2;
 
@@ -1418,41 +1684,56 @@ out_commit:
 			 * If set value outside failed, we have to clean
 			 * the junk tree root we have already set in local.
 			 */
-			ret2 = ocfs2_xattr_cleanup(inode, xi, xs, offs);
+			ret2 = ocfs2_xattr_cleanup(inode, ctxt->handle,
+						   xi, xs, &vb, offs);
 			if (ret2 < 0)
 				mlog_errno(ret2);
 		}
 	}
 out:
 	return ret;
-
 }
 
 static int ocfs2_remove_value_outside(struct inode*inode,
-				      struct buffer_head *bh,
+				      struct ocfs2_xattr_value_buf *vb,
 				      struct ocfs2_xattr_header *header)
 {
 	int ret = 0, i;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct ocfs2_xattr_set_ctxt ctxt = { NULL, NULL, };
+
+	ocfs2_init_dealloc_ctxt(&ctxt.dealloc);
+
+	ctxt.handle = ocfs2_start_trans(osb,
+					ocfs2_remove_extent_credits(osb->sb));
+	if (IS_ERR(ctxt.handle)) {
+		ret = PTR_ERR(ctxt.handle);
+		mlog_errno(ret);
+		goto out;
+	}
 
 	for (i = 0; i < le16_to_cpu(header->xh_count); i++) {
 		struct ocfs2_xattr_entry *entry = &header->xh_entries[i];
 
 		if (!ocfs2_xattr_is_local(entry)) {
-			struct ocfs2_xattr_value_root *xv;
 			void *val;
 
 			val = (void *)header +
 				le16_to_cpu(entry->xe_name_offset);
-			xv = (struct ocfs2_xattr_value_root *)
+			vb->vb_xv = (struct ocfs2_xattr_value_root *)
 				(val + OCFS2_XATTR_SIZE(entry->xe_name_len));
-			ret = ocfs2_xattr_value_truncate(inode, bh, xv, 0);
+			ret = ocfs2_xattr_value_truncate(inode, vb, 0, &ctxt);
 			if (ret < 0) {
 				mlog_errno(ret);
-				return ret;
+				break;
 			}
 		}
 	}
 
+	ocfs2_commit_trans(osb, ctxt.handle);
+	ocfs2_schedule_truncate_log_flush(osb, 1);
+	ocfs2_run_deallocs(osb, &ctxt.dealloc);
+out:
 	return ret;
 }
 
@@ -1463,12 +1744,16 @@ static int ocfs2_xattr_ibody_remove(struct inode *inode,
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
 	struct ocfs2_xattr_header *header;
 	int ret;
+	struct ocfs2_xattr_value_buf vb = {
+		.vb_bh = di_bh,
+		.vb_access = ocfs2_journal_access_di,
+	};
 
 	header = (struct ocfs2_xattr_header *)
 		 ((void *)di + inode->i_sb->s_blocksize -
 		 le16_to_cpu(di->i_xattr_inline_size));
 
-	ret = ocfs2_remove_value_outside(inode, di_bh, header);
+	ret = ocfs2_remove_value_outside(inode, &vb, header);
 
 	return ret;
 }
@@ -1478,11 +1763,15 @@ static int ocfs2_xattr_block_remove(struct inode *inode,
 {
 	struct ocfs2_xattr_block *xb;
 	int ret = 0;
+	struct ocfs2_xattr_value_buf vb = {
+		.vb_bh = blk_bh,
+		.vb_access = ocfs2_journal_access_xb,
+	};
 
 	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
 	if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) {
 		struct ocfs2_xattr_header *header = &(xb->xb_attrs.xb_header);
-		ret = ocfs2_remove_value_outside(inode, blk_bh, header);
+		ret = ocfs2_remove_value_outside(inode, &vb, header);
 	} else
 		ret = ocfs2_delete_xattr_index_block(inode, blk_bh);
 
@@ -1502,24 +1791,19 @@ static int ocfs2_xattr_free_block(struct inode *inode,
 	u64 blk, bg_blkno;
 	u16 bit;
 
-	ret = ocfs2_read_block(inode, block, &blk_bh);
+	ret = ocfs2_read_xattr_block(inode, block, &blk_bh);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
-	if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) {
-		ret = -EIO;
-		goto out;
-	}
-
 	ret = ocfs2_xattr_block_remove(inode, blk_bh);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
 	}
 
+	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
 	blk = le64_to_cpu(xb->xb_blkno);
 	bit = le16_to_cpu(xb->xb_suballoc_bit);
 	bg_blkno = ocfs2_which_suballoc_group(blk, bit);
@@ -1606,8 +1890,8 @@ int ocfs2_xattr_remove(struct inode *inode, struct buffer_head *di_bh)
 		mlog_errno(ret);
 		goto out;
 	}
-	ret = ocfs2_journal_access(handle, inode, di_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_di(handle, inode, di_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -1714,7 +1998,8 @@ static int ocfs2_xattr_ibody_find(struct inode *inode,
  */
 static int ocfs2_xattr_ibody_set(struct inode *inode,
 				 struct ocfs2_xattr_info *xi,
-				 struct ocfs2_xattr_search *xs)
+				 struct ocfs2_xattr_search *xs,
+				 struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	struct ocfs2_inode_info *oi = OCFS2_I(inode);
 	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
@@ -1731,7 +2016,7 @@ static int ocfs2_xattr_ibody_set(struct inode *inode,
 		}
 	}
 
-	ret = ocfs2_xattr_set_entry(inode, xi, xs,
+	ret = ocfs2_xattr_set_entry(inode, xi, xs, ctxt,
 				(OCFS2_INLINE_XATTR_FL | OCFS2_HAS_XATTR_FL));
 out:
 	up_write(&oi->ip_alloc_sem);
@@ -1758,19 +2043,15 @@ static int ocfs2_xattr_block_find(struct inode *inode,
 	if (!di->i_xattr_loc)
 		return ret;
 
-	ret = ocfs2_read_block(inode, le64_to_cpu(di->i_xattr_loc), &blk_bh);
+	ret = ocfs2_read_xattr_block(inode, le64_to_cpu(di->i_xattr_loc),
+				     &blk_bh);
 	if (ret < 0) {
 		mlog_errno(ret);
 		return ret;
 	}
 
-	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
-	if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) {
-		ret = -EIO;
-		goto cleanup;
-	}
-
 	xs->xattr_bh = blk_bh;
+	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
 
 	if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) {
 		xs->header = &xb->xb_attrs.xb_header;
@@ -1804,13 +2085,13 @@ cleanup:
  */
 static int ocfs2_xattr_block_set(struct inode *inode,
 				 struct ocfs2_xattr_info *xi,
-				 struct ocfs2_xattr_search *xs)
+				 struct ocfs2_xattr_search *xs,
+				 struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	struct buffer_head *new_bh = NULL;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct ocfs2_dinode *di =  (struct ocfs2_dinode *)xs->inode_bh->b_data;
-	struct ocfs2_alloc_context *meta_ac = NULL;
-	handle_t *handle = NULL;
+	handle_t *handle = ctxt->handle;
 	struct ocfs2_xattr_block *xblk = NULL;
 	u16 suballoc_bit_start;
 	u32 num_got;
@@ -1818,45 +2099,29 @@ static int ocfs2_xattr_block_set(struct inode *inode,
 	int ret;
 
 	if (!xs->xattr_bh) {
-		/*
-		 * Alloc one external block for extended attribute
-		 * outside of inode.
-		 */
-		ret = ocfs2_reserve_new_metadata_blocks(osb, 1, &meta_ac);
+		ret = ocfs2_journal_access_di(handle, inode, xs->inode_bh,
+					      OCFS2_JOURNAL_ACCESS_CREATE);
 		if (ret < 0) {
 			mlog_errno(ret);
-			goto out;
-		}
-		handle = ocfs2_start_trans(osb,
-					   OCFS2_XATTR_BLOCK_CREATE_CREDITS);
-		if (IS_ERR(handle)) {
-			ret = PTR_ERR(handle);
-			mlog_errno(ret);
-			goto out;
-		}
-		ret = ocfs2_journal_access(handle, inode, xs->inode_bh,
-					   OCFS2_JOURNAL_ACCESS_CREATE);
-		if (ret < 0) {
-			mlog_errno(ret);
-			goto out_commit;
+			goto end;
 		}
 
-		ret = ocfs2_claim_metadata(osb, handle, meta_ac, 1,
+		ret = ocfs2_claim_metadata(osb, handle, ctxt->meta_ac, 1,
 					   &suballoc_bit_start, &num_got,
 					   &first_blkno);
 		if (ret < 0) {
 			mlog_errno(ret);
-			goto out_commit;
+			goto end;
 		}
 
 		new_bh = sb_getblk(inode->i_sb, first_blkno);
 		ocfs2_set_new_buffer_uptodate(inode, new_bh);
 
-		ret = ocfs2_journal_access(handle, inode, new_bh,
-					   OCFS2_JOURNAL_ACCESS_CREATE);
+		ret = ocfs2_journal_access_xb(handle, inode, new_bh,
+					      OCFS2_JOURNAL_ACCESS_CREATE);
 		if (ret < 0) {
 			mlog_errno(ret);
-			goto out_commit;
+			goto end;
 		}
 
 		/* Initialize ocfs2_xattr_block */
@@ -1874,44 +2139,555 @@ static int ocfs2_xattr_block_set(struct inode *inode,
 		xs->end = (void *)xblk + inode->i_sb->s_blocksize;
 		xs->here = xs->header->xh_entries;
 
-
 		ret = ocfs2_journal_dirty(handle, new_bh);
 		if (ret < 0) {
 			mlog_errno(ret);
-			goto out_commit;
+			goto end;
 		}
 		di->i_xattr_loc = cpu_to_le64(first_blkno);
-		ret = ocfs2_journal_dirty(handle, xs->inode_bh);
-		if (ret < 0)
-			mlog_errno(ret);
-out_commit:
-		ocfs2_commit_trans(osb, handle);
-out:
-		if (meta_ac)
-			ocfs2_free_alloc_context(meta_ac);
-		if (ret < 0)
-			return ret;
+		ocfs2_journal_dirty(handle, xs->inode_bh);
 	} else
 		xblk = (struct ocfs2_xattr_block *)xs->xattr_bh->b_data;
 
 	if (!(le16_to_cpu(xblk->xb_flags) & OCFS2_XATTR_INDEXED)) {
 		/* Set extended attribute into external block */
-		ret = ocfs2_xattr_set_entry(inode, xi, xs, OCFS2_HAS_XATTR_FL);
+		ret = ocfs2_xattr_set_entry(inode, xi, xs, ctxt,
+					    OCFS2_HAS_XATTR_FL);
 		if (!ret || ret != -ENOSPC)
 			goto end;
 
-		ret = ocfs2_xattr_create_index_block(inode, xs);
+		ret = ocfs2_xattr_create_index_block(inode, xs, ctxt);
 		if (ret)
 			goto end;
 	}
 
-	ret = ocfs2_xattr_set_entry_index_block(inode, xi, xs);
+	ret = ocfs2_xattr_set_entry_index_block(inode, xi, xs, ctxt);
 
 end:
 
 	return ret;
 }
 
+/* Check whether the new xattr can be inserted into the inode. */
+static int ocfs2_xattr_can_be_in_inode(struct inode *inode,
+				       struct ocfs2_xattr_info *xi,
+				       struct ocfs2_xattr_search *xs)
+{
+	u64 value_size;
+	struct ocfs2_xattr_entry *last;
+	int free, i;
+	size_t min_offs = xs->end - xs->base;
+
+	if (!xs->header)
+		return 0;
+
+	last = xs->header->xh_entries;
+
+	for (i = 0; i < le16_to_cpu(xs->header->xh_count); i++) {
+		size_t offs = le16_to_cpu(last->xe_name_offset);
+		if (offs < min_offs)
+			min_offs = offs;
+		last += 1;
+	}
+
+	free = min_offs - ((void *)last - xs->base) - sizeof(__u32);
+	if (free < 0)
+		return 0;
+
+	BUG_ON(!xs->not_found);
+
+	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE)
+		value_size = OCFS2_XATTR_ROOT_SIZE;
+	else
+		value_size = OCFS2_XATTR_SIZE(xi->value_len);
+
+	if (free >= sizeof(struct ocfs2_xattr_entry) +
+		   OCFS2_XATTR_SIZE(strlen(xi->name)) + value_size)
+		return 1;
+
+	return 0;
+}
+
+static int ocfs2_calc_xattr_set_need(struct inode *inode,
+				     struct ocfs2_dinode *di,
+				     struct ocfs2_xattr_info *xi,
+				     struct ocfs2_xattr_search *xis,
+				     struct ocfs2_xattr_search *xbs,
+				     int *clusters_need,
+				     int *meta_need,
+				     int *credits_need)
+{
+	int ret = 0, old_in_xb = 0;
+	int clusters_add = 0, meta_add = 0, credits = 0;
+	struct buffer_head *bh = NULL;
+	struct ocfs2_xattr_block *xb = NULL;
+	struct ocfs2_xattr_entry *xe = NULL;
+	struct ocfs2_xattr_value_root *xv = NULL;
+	char *base = NULL;
+	int name_offset, name_len = 0;
+	u32 new_clusters = ocfs2_clusters_for_bytes(inode->i_sb,
+						    xi->value_len);
+	u64 value_size;
+
+	/*
+	 * Calculate the clusters we need to write.
+	 * No matter whether we replace an old one or add a new one,
+	 * we need this for writing.
+	 */
+	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE)
+		credits += new_clusters *
+			   ocfs2_clusters_to_blocks(inode->i_sb, 1);
+
+	if (xis->not_found && xbs->not_found) {
+		credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+
+		if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
+			clusters_add += new_clusters;
+			credits += ocfs2_calc_extend_credits(inode->i_sb,
+							&def_xv.xv.xr_list,
+							new_clusters);
+		}
+
+		goto meta_guess;
+	}
+
+	if (!xis->not_found) {
+		xe = xis->here;
+		name_offset = le16_to_cpu(xe->xe_name_offset);
+		name_len = OCFS2_XATTR_SIZE(xe->xe_name_len);
+		base = xis->base;
+		credits += OCFS2_INODE_UPDATE_CREDITS;
+	} else {
+		int i, block_off = 0;
+		xb = (struct ocfs2_xattr_block *)xbs->xattr_bh->b_data;
+		xe = xbs->here;
+		name_offset = le16_to_cpu(xe->xe_name_offset);
+		name_len = OCFS2_XATTR_SIZE(xe->xe_name_len);
+		i = xbs->here - xbs->header->xh_entries;
+		old_in_xb = 1;
+
+		if (le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED) {
+			ret = ocfs2_xattr_bucket_get_name_value(inode,
+							bucket_xh(xbs->bucket),
+							i, &block_off,
+							&name_offset);
+			base = bucket_block(xbs->bucket, block_off);
+			credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+		} else {
+			base = xbs->base;
+			credits += OCFS2_XATTR_BLOCK_UPDATE_CREDITS;
+		}
+	}
+
+	/*
+	 * delete a xattr doesn't need metadata and cluster allocation.
+	 * so just calculate the credits and return.
+	 *
+	 * The credits for removing the value tree will be extended
+	 * by ocfs2_remove_extent itself.
+	 */
+	if (!xi->value) {
+		if (!ocfs2_xattr_is_local(xe))
+			credits += ocfs2_remove_extent_credits(inode->i_sb);
+
+		goto out;
+	}
+
+	/* do cluster allocation guess first. */
+	value_size = le64_to_cpu(xe->xe_value_size);
+
+	if (old_in_xb) {
+		/*
+		 * In xattr set, we always try to set the xe in inode first,
+		 * so if it can be inserted into inode successfully, the old
+		 * one will be removed from the xattr block, and this xattr
+		 * will be inserted into inode as a new xattr in inode.
+		 */
+		if (ocfs2_xattr_can_be_in_inode(inode, xi, xis)) {
+			clusters_add += new_clusters;
+			credits += ocfs2_remove_extent_credits(inode->i_sb) +
+				    OCFS2_INODE_UPDATE_CREDITS;
+			if (!ocfs2_xattr_is_local(xe))
+				credits += ocfs2_calc_extend_credits(
+							inode->i_sb,
+							&def_xv.xv.xr_list,
+							new_clusters);
+			goto out;
+		}
+	}
+
+	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
+		/* the new values will be stored outside. */
+		u32 old_clusters = 0;
+
+		if (!ocfs2_xattr_is_local(xe)) {
+			old_clusters =	ocfs2_clusters_for_bytes(inode->i_sb,
+								 value_size);
+			xv = (struct ocfs2_xattr_value_root *)
+			     (base + name_offset + name_len);
+			value_size = OCFS2_XATTR_ROOT_SIZE;
+		} else
+			xv = &def_xv.xv;
+
+		if (old_clusters >= new_clusters) {
+			credits += ocfs2_remove_extent_credits(inode->i_sb);
+			goto out;
+		} else {
+			meta_add += ocfs2_extend_meta_needed(&xv->xr_list);
+			clusters_add += new_clusters - old_clusters;
+			credits += ocfs2_calc_extend_credits(inode->i_sb,
+							     &xv->xr_list,
+							     new_clusters -
+							     old_clusters);
+			if (value_size >= OCFS2_XATTR_ROOT_SIZE)
+				goto out;
+		}
+	} else {
+		/*
+		 * Now the new value will be stored inside. So if the new
+		 * value is smaller than the size of value root or the old
+		 * value, we don't need any allocation, otherwise we have
+		 * to guess metadata allocation.
+		 */
+		if ((ocfs2_xattr_is_local(xe) && value_size >= xi->value_len) ||
+		    (!ocfs2_xattr_is_local(xe) &&
+		     OCFS2_XATTR_ROOT_SIZE >= xi->value_len))
+			goto out;
+	}
+
+meta_guess:
+	/* calculate metadata allocation. */
+	if (di->i_xattr_loc) {
+		if (!xbs->xattr_bh) {
+			ret = ocfs2_read_xattr_block(inode,
+						     le64_to_cpu(di->i_xattr_loc),
+						     &bh);
+			if (ret) {
+				mlog_errno(ret);
+				goto out;
+			}
+
+			xb = (struct ocfs2_xattr_block *)bh->b_data;
+		} else
+			xb = (struct ocfs2_xattr_block *)xbs->xattr_bh->b_data;
+
+		/*
+		 * If there is already an xattr tree, good, we can calculate
+		 * like other b-trees. Otherwise we may have the chance of
+		 * create a tree, the credit calculation is borrowed from
+		 * ocfs2_calc_extend_credits with root_el = NULL. And the
+		 * new tree will be cluster based, so no meta is needed.
+		 */
+		if (le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED) {
+			struct ocfs2_extent_list *el =
+				 &xb->xb_attrs.xb_root.xt_list;
+			meta_add += ocfs2_extend_meta_needed(el);
+			credits += ocfs2_calc_extend_credits(inode->i_sb,
+							     el, 1);
+		} else
+			credits += OCFS2_SUBALLOC_ALLOC + 1;
+
+		/*
+		 * This cluster will be used either for new bucket or for
+		 * new xattr block.
+		 * If the cluster size is the same as the bucket size, one
+		 * more is needed since we may need to extend the bucket
+		 * also.
+		 */
+		clusters_add += 1;
+		credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+		if (OCFS2_XATTR_BUCKET_SIZE ==
+			OCFS2_SB(inode->i_sb)->s_clustersize) {
+			credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+			clusters_add += 1;
+		}
+	} else {
+		meta_add += 1;
+		credits += OCFS2_XATTR_BLOCK_CREATE_CREDITS;
+	}
+out:
+	if (clusters_need)
+		*clusters_need = clusters_add;
+	if (meta_need)
+		*meta_need = meta_add;
+	if (credits_need)
+		*credits_need = credits;
+	brelse(bh);
+	return ret;
+}
+
+static int ocfs2_init_xattr_set_ctxt(struct inode *inode,
+				     struct ocfs2_dinode *di,
+				     struct ocfs2_xattr_info *xi,
+				     struct ocfs2_xattr_search *xis,
+				     struct ocfs2_xattr_search *xbs,
+				     struct ocfs2_xattr_set_ctxt *ctxt,
+				     int *credits)
+{
+	int clusters_add, meta_add, ret;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+
+	memset(ctxt, 0, sizeof(struct ocfs2_xattr_set_ctxt));
+
+	ocfs2_init_dealloc_ctxt(&ctxt->dealloc);
+
+	ret = ocfs2_calc_xattr_set_need(inode, di, xi, xis, xbs,
+					&clusters_add, &meta_add, credits);
+	if (ret) {
+		mlog_errno(ret);
+		return ret;
+	}
+
+	mlog(0, "Set xattr %s, reserve meta blocks = %d, clusters = %d, "
+	     "credits = %d\n", xi->name, meta_add, clusters_add, *credits);
+
+	if (meta_add) {
+		ret = ocfs2_reserve_new_metadata_blocks(osb, meta_add,
+							&ctxt->meta_ac);
+		if (ret) {
+			mlog_errno(ret);
+			goto out;
+		}
+	}
+
+	if (clusters_add) {
+		ret = ocfs2_reserve_clusters(osb, clusters_add, &ctxt->data_ac);
+		if (ret)
+			mlog_errno(ret);
+	}
+out:
+	if (ret) {
+		if (ctxt->meta_ac) {
+			ocfs2_free_alloc_context(ctxt->meta_ac);
+			ctxt->meta_ac = NULL;
+		}
+
+		/*
+		 * We cannot have an error and a non null ctxt->data_ac.
+		 */
+	}
+
+	return ret;
+}
+
+static int __ocfs2_xattr_set_handle(struct inode *inode,
+				    struct ocfs2_dinode *di,
+				    struct ocfs2_xattr_info *xi,
+				    struct ocfs2_xattr_search *xis,
+				    struct ocfs2_xattr_search *xbs,
+				    struct ocfs2_xattr_set_ctxt *ctxt)
+{
+	int ret = 0, credits, old_found;
+
+	if (!xi->value) {
+		/* Remove existing extended attribute */
+		if (!xis->not_found)
+			ret = ocfs2_xattr_ibody_set(inode, xi, xis, ctxt);
+		else if (!xbs->not_found)
+			ret = ocfs2_xattr_block_set(inode, xi, xbs, ctxt);
+	} else {
+		/* We always try to set extended attribute into inode first*/
+		ret = ocfs2_xattr_ibody_set(inode, xi, xis, ctxt);
+		if (!ret && !xbs->not_found) {
+			/*
+			 * If succeed and that extended attribute existing in
+			 * external block, then we will remove it.
+			 */
+			xi->value = NULL;
+			xi->value_len = 0;
+
+			old_found = xis->not_found;
+			xis->not_found = -ENODATA;
+			ret = ocfs2_calc_xattr_set_need(inode,
+							di,
+							xi,
+							xis,
+							xbs,
+							NULL,
+							NULL,
+							&credits);
+			xis->not_found = old_found;
+			if (ret) {
+				mlog_errno(ret);
+				goto out;
+			}
+
+			ret = ocfs2_extend_trans(ctxt->handle, credits +
+					ctxt->handle->h_buffer_credits);
+			if (ret) {
+				mlog_errno(ret);
+				goto out;
+			}
+			ret = ocfs2_xattr_block_set(inode, xi, xbs, ctxt);
+		} else if (ret == -ENOSPC) {
+			if (di->i_xattr_loc && !xbs->xattr_bh) {
+				ret = ocfs2_xattr_block_find(inode,
+							     xi->name_index,
+							     xi->name, xbs);
+				if (ret)
+					goto out;
+
+				old_found = xis->not_found;
+				xis->not_found = -ENODATA;
+				ret = ocfs2_calc_xattr_set_need(inode,
+								di,
+								xi,
+								xis,
+								xbs,
+								NULL,
+								NULL,
+								&credits);
+				xis->not_found = old_found;
+				if (ret) {
+					mlog_errno(ret);
+					goto out;
+				}
+
+				ret = ocfs2_extend_trans(ctxt->handle, credits +
+					ctxt->handle->h_buffer_credits);
+				if (ret) {
+					mlog_errno(ret);
+					goto out;
+				}
+			}
+			/*
+			 * If no space in inode, we will set extended attribute
+			 * into external block.
+			 */
+			ret = ocfs2_xattr_block_set(inode, xi, xbs, ctxt);
+			if (ret)
+				goto out;
+			if (!xis->not_found) {
+				/*
+				 * If succeed and that extended attribute
+				 * existing in inode, we will remove it.
+				 */
+				xi->value = NULL;
+				xi->value_len = 0;
+				xbs->not_found = -ENODATA;
+				ret = ocfs2_calc_xattr_set_need(inode,
+								di,
+								xi,
+								xis,
+								xbs,
+								NULL,
+								NULL,
+								&credits);
+				if (ret) {
+					mlog_errno(ret);
+					goto out;
+				}
+
+				ret = ocfs2_extend_trans(ctxt->handle, credits +
+						ctxt->handle->h_buffer_credits);
+				if (ret) {
+					mlog_errno(ret);
+					goto out;
+				}
+				ret = ocfs2_xattr_ibody_set(inode, xi,
+							    xis, ctxt);
+			}
+		}
+	}
+
+	if (!ret) {
+		/* Update inode ctime. */
+		ret = ocfs2_journal_access(ctxt->handle, inode, xis->inode_bh,
+					   OCFS2_JOURNAL_ACCESS_WRITE);
+		if (ret) {
+			mlog_errno(ret);
+			goto out;
+		}
+
+		inode->i_ctime = CURRENT_TIME;
+		di->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec);
+		di->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec);
+		ocfs2_journal_dirty(ctxt->handle, xis->inode_bh);
+	}
+out:
+	return ret;
+}
+
+/*
+ * This function only called duing creating inode
+ * for init security/acl xattrs of the new inode.
+ * All transanction credits have been reserved in mknod.
+ */
+int ocfs2_xattr_set_handle(handle_t *handle,
+			   struct inode *inode,
+			   struct buffer_head *di_bh,
+			   int name_index,
+			   const char *name,
+			   const void *value,
+			   size_t value_len,
+			   int flags,
+			   struct ocfs2_alloc_context *meta_ac,
+			   struct ocfs2_alloc_context *data_ac)
+{
+	struct ocfs2_dinode *di;
+	int ret;
+
+	struct ocfs2_xattr_info xi = {
+		.name_index = name_index,
+		.name = name,
+		.value = value,
+		.value_len = value_len,
+	};
+
+	struct ocfs2_xattr_search xis = {
+		.not_found = -ENODATA,
+	};
+
+	struct ocfs2_xattr_search xbs = {
+		.not_found = -ENODATA,
+	};
+
+	struct ocfs2_xattr_set_ctxt ctxt = {
+		.handle = handle,
+		.meta_ac = meta_ac,
+		.data_ac = data_ac,
+	};
+
+	if (!ocfs2_supports_xattr(OCFS2_SB(inode->i_sb)))
+		return -EOPNOTSUPP;
+
+	/*
+	 * In extreme situation, may need xattr bucket when
+	 * block size is too small. And we have already reserved
+	 * the credits for bucket in mknod.
+	 */
+	if (inode->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE) {
+		xbs.bucket = ocfs2_xattr_bucket_new(inode);
+		if (!xbs.bucket) {
+			mlog_errno(-ENOMEM);
+			return -ENOMEM;
+		}
+	}
+
+	xis.inode_bh = xbs.inode_bh = di_bh;
+	di = (struct ocfs2_dinode *)di_bh->b_data;
+
+	down_write(&OCFS2_I(inode)->ip_xattr_sem);
+
+	ret = ocfs2_xattr_ibody_find(inode, name_index, name, &xis);
+	if (ret)
+		goto cleanup;
+	if (xis.not_found) {
+		ret = ocfs2_xattr_block_find(inode, name_index, name, &xbs);
+		if (ret)
+			goto cleanup;
+	}
+
+	ret = __ocfs2_xattr_set_handle(inode, di, &xi, &xis, &xbs, &ctxt);
+
+cleanup:
+	up_write(&OCFS2_I(inode)->ip_xattr_sem);
+	brelse(xbs.xattr_bh);
+	ocfs2_xattr_bucket_free(xbs.bucket);
+
+	return ret;
+}
+
 /*
  * ocfs2_xattr_set()
  *
@@ -1928,8 +2704,10 @@ int ocfs2_xattr_set(struct inode *inode,
 {
 	struct buffer_head *di_bh = NULL;
 	struct ocfs2_dinode *di;
-	int ret;
-	u16 i, blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+	int ret, credits;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct inode *tl_inode = osb->osb_tl_inode;
+	struct ocfs2_xattr_set_ctxt ctxt = { NULL, NULL, };
 
 	struct ocfs2_xattr_info xi = {
 		.name_index = name_index,
@@ -1949,10 +2727,20 @@ int ocfs2_xattr_set(struct inode *inode,
 	if (!ocfs2_supports_xattr(OCFS2_SB(inode->i_sb)))
 		return -EOPNOTSUPP;
 
+	/*
+	 * Only xbs will be used on indexed trees.  xis doesn't need a
+	 * bucket.
+	 */
+	xbs.bucket = ocfs2_xattr_bucket_new(inode);
+	if (!xbs.bucket) {
+		mlog_errno(-ENOMEM);
+		return -ENOMEM;
+	}
+
 	ret = ocfs2_inode_lock(inode, &di_bh, 1);
 	if (ret < 0) {
 		mlog_errno(ret);
-		return ret;
+		goto cleanup_nolock;
 	}
 	xis.inode_bh = xbs.inode_bh = di_bh;
 	di = (struct ocfs2_dinode *)di_bh->b_data;
@@ -1984,55 +2772,53 @@ int ocfs2_xattr_set(struct inode *inode,
 			goto cleanup;
 	}
 
-	if (!value) {
-		/* Remove existing extended attribute */
-		if (!xis.not_found)
-			ret = ocfs2_xattr_ibody_set(inode, &xi, &xis);
-		else if (!xbs.not_found)
-			ret = ocfs2_xattr_block_set(inode, &xi, &xbs);
-	} else {
-		/* We always try to set extended attribute into inode first*/
-		ret = ocfs2_xattr_ibody_set(inode, &xi, &xis);
-		if (!ret && !xbs.not_found) {
-			/*
-			 * If succeed and that extended attribute existing in
-			 * external block, then we will remove it.
-			 */
-			xi.value = NULL;
-			xi.value_len = 0;
-			ret = ocfs2_xattr_block_set(inode, &xi, &xbs);
-		} else if (ret == -ENOSPC) {
-			if (di->i_xattr_loc && !xbs.xattr_bh) {
-				ret = ocfs2_xattr_block_find(inode, name_index,
-							     name, &xbs);
-				if (ret)
-					goto cleanup;
-			}
-			/*
-			 * If no space in inode, we will set extended attribute
-			 * into external block.
-			 */
-			ret = ocfs2_xattr_block_set(inode, &xi, &xbs);
-			if (ret)
-				goto cleanup;
-			if (!xis.not_found) {
-				/*
-				 * If succeed and that extended attribute
-				 * existing in inode, we will remove it.
-				 */
-				xi.value = NULL;
-				xi.value_len = 0;
-				ret = ocfs2_xattr_ibody_set(inode, &xi, &xis);
-			}
-		}
+
+	mutex_lock(&tl_inode->i_mutex);
+
+	if (ocfs2_truncate_log_needs_flush(osb)) {
+		ret = __ocfs2_flush_truncate_log(osb);
+		if (ret < 0) {
+			mutex_unlock(&tl_inode->i_mutex);
+			mlog_errno(ret);
+			goto cleanup;
+		}
+	}
+	mutex_unlock(&tl_inode->i_mutex);
+
+	ret = ocfs2_init_xattr_set_ctxt(inode, di, &xi, &xis,
+					&xbs, &ctxt, &credits);
+	if (ret) {
+		mlog_errno(ret);
+		goto cleanup;
+	}
+
+	/* we need to update inode's ctime field, so add credit for it. */
+	credits += OCFS2_INODE_UPDATE_CREDITS;
+	ctxt.handle = ocfs2_start_trans(osb, credits);
+	if (IS_ERR(ctxt.handle)) {
+		ret = PTR_ERR(ctxt.handle);
+		mlog_errno(ret);
+		goto cleanup;
 	}
+
+	ret = __ocfs2_xattr_set_handle(inode, di, &xi, &xis, &xbs, &ctxt);
+
+	ocfs2_commit_trans(osb, ctxt.handle);
+
+	if (ctxt.data_ac)
+		ocfs2_free_alloc_context(ctxt.data_ac);
+	if (ctxt.meta_ac)
+		ocfs2_free_alloc_context(ctxt.meta_ac);
+	if (ocfs2_dealloc_has_cluster(&ctxt.dealloc))
+		ocfs2_schedule_truncate_log_flush(osb, 1);
+	ocfs2_run_deallocs(osb, &ctxt.dealloc);
 cleanup:
 	up_write(&OCFS2_I(inode)->ip_xattr_sem);
 	ocfs2_inode_unlock(inode, 1);
+cleanup_nolock:
 	brelse(di_bh);
 	brelse(xbs.xattr_bh);
-	for (i = 0; i < blk_per_bucket; i++)
-		brelse(xbs.bucket.bhs[i]);
+	ocfs2_xattr_bucket_free(xbs.bucket);
 
 	return ret;
 }
@@ -2107,7 +2893,7 @@ typedef int (xattr_bucket_func)(struct inode *inode,
 				void *para);
 
 static int ocfs2_find_xe_in_bucket(struct inode *inode,
-				   struct buffer_head *header_bh,
+				   struct ocfs2_xattr_bucket *bucket,
 				   int name_index,
 				   const char *name,
 				   u32 name_hash,
@@ -2115,11 +2901,9 @@ static int ocfs2_find_xe_in_bucket(struct inode *inode,
 				   int *found)
 {
 	int i, ret = 0, cmp = 1, block_off, new_offset;
-	struct ocfs2_xattr_header *xh =
-			(struct ocfs2_xattr_header *)header_bh->b_data;
+	struct ocfs2_xattr_header *xh = bucket_xh(bucket);
 	size_t name_len = strlen(name);
 	struct ocfs2_xattr_entry *xe = NULL;
-	struct buffer_head *name_bh = NULL;
 	char *xe_name;
 
 	/*
@@ -2150,19 +2934,9 @@ static int ocfs2_find_xe_in_bucket(struct inode *inode,
 			break;
 		}
 
-		ret = ocfs2_read_block(inode, header_bh->b_blocknr + block_off,
-				       &name_bh);
-		if (ret) {
-			mlog_errno(ret);
-			break;
-		}
-		xe_name = name_bh->b_data + new_offset;
-
-		cmp = memcmp(name, xe_name, name_len);
-		brelse(name_bh);
-		name_bh = NULL;
 
-		if (cmp == 0) {
+		xe_name = bucket_block(bucket, block_off) + new_offset;
+		if (!memcmp(name, xe_name, name_len)) {
 			*xe_index = i;
 			*found = 1;
 			ret = 0;
@@ -2192,39 +2966,42 @@ static int ocfs2_xattr_bucket_find(struct inode *inode,
 				   struct ocfs2_xattr_search *xs)
 {
 	int ret, found = 0;
-	struct buffer_head *bh = NULL;
-	struct buffer_head *lower_bh = NULL;
 	struct ocfs2_xattr_header *xh = NULL;
 	struct ocfs2_xattr_entry *xe = NULL;
 	u16 index = 0;
 	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
 	int low_bucket = 0, bucket, high_bucket;
+	struct ocfs2_xattr_bucket *search;
 	u32 last_hash;
-	u64 blkno;
+	u64 blkno, lower_blkno = 0;
+
+	search = ocfs2_xattr_bucket_new(inode);
+	if (!search) {
+		ret = -ENOMEM;
+		mlog_errno(ret);
+		goto out;
+	}
 
-	ret = ocfs2_read_block(inode, p_blkno, &bh);
+	ret = ocfs2_read_xattr_bucket(search, p_blkno);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	xh = (struct ocfs2_xattr_header *)bh->b_data;
+	xh = bucket_xh(search);
 	high_bucket = le16_to_cpu(xh->xh_num_buckets) - 1;
-
 	while (low_bucket <= high_bucket) {
-		brelse(bh);
-		bh = NULL;
-		bucket = (low_bucket + high_bucket) / 2;
+		ocfs2_xattr_bucket_relse(search);
 
+		bucket = (low_bucket + high_bucket) / 2;
 		blkno = p_blkno + bucket * blk_per_bucket;
-
-		ret = ocfs2_read_block(inode, blkno, &bh);
+		ret = ocfs2_read_xattr_bucket(search, blkno);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
 
-		xh = (struct ocfs2_xattr_header *)bh->b_data;
+		xh = bucket_xh(search);
 		xe = &xh->xh_entries[0];
 		if (name_hash < le32_to_cpu(xe->xe_name_hash)) {
 			high_bucket = bucket - 1;
@@ -2241,10 +3018,8 @@ static int ocfs2_xattr_bucket_find(struct inode *inode,
 
 		last_hash = le32_to_cpu(xe->xe_name_hash);
 
-		/* record lower_bh which may be the insert place. */
-		brelse(lower_bh);
-		lower_bh = bh;
-		bh = NULL;
+		/* record lower_blkno which may be the insert place. */
+		lower_blkno = blkno;
 
 		if (name_hash > le32_to_cpu(xe->xe_name_hash)) {
 			low_bucket = bucket + 1;
@@ -2252,7 +3027,7 @@ static int ocfs2_xattr_bucket_find(struct inode *inode,
 		}
 
 		/* the searched xattr should reside in this bucket if exists. */
-		ret = ocfs2_find_xe_in_bucket(inode, lower_bh,
+		ret = ocfs2_find_xe_in_bucket(inode, search,
 					      name_index, name, name_hash,
 					      &index, &found);
 		if (ret) {
@@ -2267,46 +3042,29 @@ static int ocfs2_xattr_bucket_find(struct inode *inode,
 	 * When the xattr's hash value is in the gap of 2 buckets, we will
 	 * always set it to the previous bucket.
 	 */
-	if (!lower_bh) {
-		/*
-		 * We can't find any bucket whose first name_hash is less
-		 * than the find name_hash.
-		 */
-		BUG_ON(bh->b_blocknr != p_blkno);
-		lower_bh = bh;
-		bh = NULL;
+	if (!lower_blkno)
+		lower_blkno = p_blkno;
+
+	/* This should be in cache - we just read it during the search */
+	ret = ocfs2_read_xattr_bucket(xs->bucket, lower_blkno);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
 	}
-	xs->bucket.bhs[0] = lower_bh;
-	xs->bucket.xh = (struct ocfs2_xattr_header *)
-					xs->bucket.bhs[0]->b_data;
-	lower_bh = NULL;
 
-	xs->header = xs->bucket.xh;
-	xs->base = xs->bucket.bhs[0]->b_data;
+	xs->header = bucket_xh(xs->bucket);
+	xs->base = bucket_block(xs->bucket, 0);
 	xs->end = xs->base + inode->i_sb->s_blocksize;
 
 	if (found) {
-		/*
-		 * If we have found the xattr enty, read all the blocks in
-		 * this bucket.
-		 */
-		ret = ocfs2_read_blocks(inode, xs->bucket.bhs[0]->b_blocknr + 1,
-					blk_per_bucket - 1, &xs->bucket.bhs[1],
-					0);
-		if (ret) {
-			mlog_errno(ret);
-			goto out;
-		}
-
 		xs->here = &xs->header->xh_entries[index];
 		mlog(0, "find xattr %s in bucket %llu, entry = %u\n", name,
-		     (unsigned long long)xs->bucket.bhs[0]->b_blocknr, index);
+		     (unsigned long long)bucket_blkno(xs->bucket), index);
 	} else
 		ret = -ENODATA;
 
 out:
-	brelse(bh);
-	brelse(lower_bh);
+	ocfs2_xattr_bucket_free(search);
 	return ret;
 }
 
@@ -2357,53 +3115,50 @@ static int ocfs2_iterate_xattr_buckets(struct inode *inode,
 				       xattr_bucket_func *func,
 				       void *para)
 {
-	int i, j, ret = 0;
-	int blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+	int i, ret = 0;
 	u32 bpc = ocfs2_xattr_buckets_per_cluster(OCFS2_SB(inode->i_sb));
 	u32 num_buckets = clusters * bpc;
-	struct ocfs2_xattr_bucket bucket;
+	struct ocfs2_xattr_bucket *bucket;
 
-	memset(&bucket, 0, sizeof(bucket));
+	bucket = ocfs2_xattr_bucket_new(inode);
+	if (!bucket) {
+		mlog_errno(-ENOMEM);
+		return -ENOMEM;
+	}
 
 	mlog(0, "iterating xattr buckets in %u clusters starting from %llu\n",
 	     clusters, (unsigned long long)blkno);
 
-	for (i = 0; i < num_buckets; i++, blkno += blk_per_bucket) {
-		ret = ocfs2_read_blocks(inode, blkno, blk_per_bucket,
-					bucket.bhs, 0);
+	for (i = 0; i < num_buckets; i++, blkno += bucket->bu_blocks) {
+		ret = ocfs2_read_xattr_bucket(bucket, blkno);
 		if (ret) {
 			mlog_errno(ret);
-			goto out;
+			break;
 		}
 
-		bucket.xh = (struct ocfs2_xattr_header *)bucket.bhs[0]->b_data;
 		/*
 		 * The real bucket num in this series of blocks is stored
 		 * in the 1st bucket.
 		 */
 		if (i == 0)
-			num_buckets = le16_to_cpu(bucket.xh->xh_num_buckets);
+			num_buckets = le16_to_cpu(bucket_xh(bucket)->xh_num_buckets);
 
 		mlog(0, "iterating xattr bucket %llu, first hash %u\n",
 		     (unsigned long long)blkno,
-		     le32_to_cpu(bucket.xh->xh_entries[0].xe_name_hash));
+		     le32_to_cpu(bucket_xh(bucket)->xh_entries[0].xe_name_hash));
 		if (func) {
-			ret = func(inode, &bucket, para);
-			if (ret) {
+			ret = func(inode, bucket, para);
+			if (ret)
 				mlog_errno(ret);
-				break;
-			}
+			/* Fall through to bucket_relse() */
 		}
 
-		for (j = 0; j < blk_per_bucket; j++)
-			brelse(bucket.bhs[j]);
-		memset(&bucket, 0, sizeof(bucket));
+		ocfs2_xattr_bucket_relse(bucket);
+		if (ret)
+			break;
 	}
 
-out:
-	for (j = 0; j < blk_per_bucket; j++)
-		brelse(bucket.bhs[j]);
-
+	ocfs2_xattr_bucket_free(bucket);
 	return ret;
 }
 
@@ -2441,21 +3196,21 @@ static int ocfs2_list_xattr_bucket(struct inode *inode,
 	int i, block_off, new_offset;
 	const char *prefix, *name;
 
-	for (i = 0 ; i < le16_to_cpu(bucket->xh->xh_count); i++) {
-		struct ocfs2_xattr_entry *entry = &bucket->xh->xh_entries[i];
+	for (i = 0 ; i < le16_to_cpu(bucket_xh(bucket)->xh_count); i++) {
+		struct ocfs2_xattr_entry *entry = &bucket_xh(bucket)->xh_entries[i];
 		type = ocfs2_xattr_get_type(entry);
 		prefix = ocfs2_xattr_prefix(type);
 
 		if (prefix) {
 			ret = ocfs2_xattr_bucket_get_name_value(inode,
-								bucket->xh,
+								bucket_xh(bucket),
 								i,
 								&block_off,
 								&new_offset);
 			if (ret)
 				break;
 
-			name = (const char *)bucket->bhs[block_off]->b_data +
+			name = (const char *)bucket_block(bucket, block_off) +
 				new_offset;
 			ret = ocfs2_xattr_list_entry(xl->buffer,
 						     xl->buffer_size,
@@ -2540,32 +3295,34 @@ static void swap_xe(void *a, void *b, int size)
 /*
  * When the ocfs2_xattr_block is filled up, new bucket will be created
  * and all the xattr entries will be moved to the new bucket.
+ * The header goes at the start of the bucket, and the names+values are
+ * filled from the end.  This is why *target starts as the last buffer.
  * Note: we need to sort the entries since they are not saved in order
  * in the ocfs2_xattr_block.
  */
 static void ocfs2_cp_xattr_block_to_bucket(struct inode *inode,
 					   struct buffer_head *xb_bh,
-					   struct buffer_head *xh_bh,
-					   struct buffer_head *data_bh)
+					   struct ocfs2_xattr_bucket *bucket)
 {
 	int i, blocksize = inode->i_sb->s_blocksize;
+	int blks = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
 	u16 offset, size, off_change;
 	struct ocfs2_xattr_entry *xe;
 	struct ocfs2_xattr_block *xb =
 				(struct ocfs2_xattr_block *)xb_bh->b_data;
 	struct ocfs2_xattr_header *xb_xh = &xb->xb_attrs.xb_header;
-	struct ocfs2_xattr_header *xh =
-				(struct ocfs2_xattr_header *)xh_bh->b_data;
+	struct ocfs2_xattr_header *xh = bucket_xh(bucket);
 	u16 count = le16_to_cpu(xb_xh->xh_count);
-	char *target = xh_bh->b_data, *src = xb_bh->b_data;
+	char *src = xb_bh->b_data;
+	char *target = bucket_block(bucket, blks - 1);
 
 	mlog(0, "cp xattr from block %llu to bucket %llu\n",
 	     (unsigned long long)xb_bh->b_blocknr,
-	     (unsigned long long)xh_bh->b_blocknr);
+	     (unsigned long long)bucket_blkno(bucket));
+
+	for (i = 0; i < blks; i++)
+		memset(bucket_block(bucket, i), 0, blocksize);
 
-	memset(xh_bh->b_data, 0, blocksize);
-	if (data_bh)
-		memset(data_bh->b_data, 0, blocksize);
 	/*
 	 * Since the xe_name_offset is based on ocfs2_xattr_header,
 	 * there is a offset change corresponding to the change of
@@ -2577,8 +3334,6 @@ static void ocfs2_cp_xattr_block_to_bucket(struct inode *inode,
 	size = blocksize - offset;
 
 	/* copy all the names and values. */
-	if (data_bh)
-		target = data_bh->b_data;
 	memcpy(target + offset, src + offset, size);
 
 	/* Init new header now. */
@@ -2588,7 +3343,7 @@ static void ocfs2_cp_xattr_block_to_bucket(struct inode *inode,
 	xh->xh_free_start = cpu_to_le16(OCFS2_XATTR_BUCKET_SIZE - size);
 
 	/* copy all the entries. */
-	target = xh_bh->b_data;
+	target = bucket_block(bucket, 0);
 	offset = offsetof(struct ocfs2_xattr_header, xh_entries);
 	size = count * sizeof(struct ocfs2_xattr_entry);
 	memcpy(target + offset, (char *)xb_xh + offset, size);
@@ -2614,73 +3369,47 @@ static void ocfs2_cp_xattr_block_to_bucket(struct inode *inode,
  * While if the entry is in index b-tree, "bucket" indicates the
  * real place of the xattr.
  */
-static int ocfs2_xattr_update_xattr_search(struct inode *inode,
-					   struct ocfs2_xattr_search *xs,
-					   struct buffer_head *old_bh,
-					   struct buffer_head *new_bh)
+static void ocfs2_xattr_update_xattr_search(struct inode *inode,
+					    struct ocfs2_xattr_search *xs,
+					    struct buffer_head *old_bh)
 {
-	int ret = 0;
 	char *buf = old_bh->b_data;
 	struct ocfs2_xattr_block *old_xb = (struct ocfs2_xattr_block *)buf;
 	struct ocfs2_xattr_header *old_xh = &old_xb->xb_attrs.xb_header;
-	int i, blocksize = inode->i_sb->s_blocksize;
-	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-
-	xs->bucket.bhs[0] = new_bh;
-	get_bh(new_bh);
-	xs->bucket.xh = (struct ocfs2_xattr_header *)xs->bucket.bhs[0]->b_data;
-	xs->header = xs->bucket.xh;
+	int i;
 
-	xs->base = new_bh->b_data;
+	xs->header = bucket_xh(xs->bucket);
+	xs->base = bucket_block(xs->bucket, 0);
 	xs->end = xs->base + inode->i_sb->s_blocksize;
 
-	if (!xs->not_found) {
-		if (OCFS2_XATTR_BUCKET_SIZE != blocksize) {
-			ret = ocfs2_read_blocks(inode,
-					xs->bucket.bhs[0]->b_blocknr + 1,
-					blk_per_bucket - 1, &xs->bucket.bhs[1],
-					0);
-			if (ret) {
-				mlog_errno(ret);
-				return ret;
-			}
-
-		}
-		i = xs->here - old_xh->xh_entries;
-		xs->here = &xs->header->xh_entries[i];
-	}
+	if (xs->not_found)
+		return;
 
-	return ret;
+	i = xs->here - old_xh->xh_entries;
+	xs->here = &xs->header->xh_entries[i];
 }
 
 static int ocfs2_xattr_create_index_block(struct inode *inode,
-					  struct ocfs2_xattr_search *xs)
+					  struct ocfs2_xattr_search *xs,
+					  struct ocfs2_xattr_set_ctxt *ctxt)
 {
-	int ret, credits = OCFS2_SUBALLOC_ALLOC;
+	int ret;
 	u32 bit_off, len;
 	u64 blkno;
-	handle_t *handle;
+	handle_t *handle = ctxt->handle;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct ocfs2_inode_info *oi = OCFS2_I(inode);
-	struct ocfs2_alloc_context *data_ac;
-	struct buffer_head *xh_bh = NULL, *data_bh = NULL;
 	struct buffer_head *xb_bh = xs->xattr_bh;
 	struct ocfs2_xattr_block *xb =
 			(struct ocfs2_xattr_block *)xb_bh->b_data;
 	struct ocfs2_xattr_tree_root *xr;
 	u16 xb_flags = le16_to_cpu(xb->xb_flags);
-	u16 bpb = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
 
 	mlog(0, "create xattr index block for %llu\n",
 	     (unsigned long long)xb_bh->b_blocknr);
 
 	BUG_ON(xb_flags & OCFS2_XATTR_INDEXED);
-
-	ret = ocfs2_reserve_clusters(osb, 1, &data_ac);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
+	BUG_ON(!xs->bucket);
 
 	/*
 	 * XXX:
@@ -2689,29 +3418,18 @@ static int ocfs2_xattr_create_index_block(struct inode *inode,
 	 */
 	down_write(&oi->ip_alloc_sem);
 
-	/*
-	 * 3 more credits, one for xattr block update, one for the 1st block
-	 * of the new xattr bucket and one for the value/data.
-	 */
-	credits += 3;
-	handle = ocfs2_start_trans(osb, credits);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		mlog_errno(ret);
-		goto out_sem;
-	}
-
-	ret = ocfs2_journal_access(handle, inode, xb_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_xb(handle, inode, xb_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 
-	ret = ocfs2_claim_clusters(osb, handle, data_ac, 1, &bit_off, &len);
+	ret = __ocfs2_claim_clusters(osb, handle, ctxt->data_ac,
+				     1, 1, &bit_off, &len);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 
 	/*
@@ -2724,51 +3442,23 @@ static int ocfs2_xattr_create_index_block(struct inode *inode,
 	mlog(0, "allocate 1 cluster from %llu to xattr block\n",
 	     (unsigned long long)blkno);
 
-	xh_bh = sb_getblk(inode->i_sb, blkno);
-	if (!xh_bh) {
-		ret = -EIO;
+	ret = ocfs2_init_xattr_bucket(xs->bucket, blkno);
+	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		goto out;
 	}
 
-	ocfs2_set_new_buffer_uptodate(inode, xh_bh);
-
-	ret = ocfs2_journal_access(handle, inode, xh_bh,
-				   OCFS2_JOURNAL_ACCESS_CREATE);
+	ret = ocfs2_xattr_bucket_journal_access(handle, xs->bucket,
+						OCFS2_JOURNAL_ACCESS_CREATE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
-	}
-
-	if (bpb > 1) {
-		data_bh = sb_getblk(inode->i_sb, blkno + bpb - 1);
-		if (!data_bh) {
-			ret = -EIO;
-			mlog_errno(ret);
-			goto out_commit;
-		}
-
-		ocfs2_set_new_buffer_uptodate(inode, data_bh);
-
-		ret = ocfs2_journal_access(handle, inode, data_bh,
-					   OCFS2_JOURNAL_ACCESS_CREATE);
-		if (ret) {
-			mlog_errno(ret);
-			goto out_commit;
-		}
+		goto out;
 	}
 
-	ocfs2_cp_xattr_block_to_bucket(inode, xb_bh, xh_bh, data_bh);
+	ocfs2_cp_xattr_block_to_bucket(inode, xb_bh, xs->bucket);
+	ocfs2_xattr_bucket_journal_dirty(handle, xs->bucket);
 
-	ocfs2_journal_dirty(handle, xh_bh);
-	if (data_bh)
-		ocfs2_journal_dirty(handle, data_bh);
-
-	ret = ocfs2_xattr_update_xattr_search(inode, xs, xb_bh, xh_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
+	ocfs2_xattr_update_xattr_search(inode, xs, xb_bh);
 
 	/* Change from ocfs2_xattr_header to ocfs2_xattr_tree_root */
 	memset(&xb->xb_attrs, 0, inode->i_sb->s_blocksize -
@@ -2787,24 +3477,10 @@ static int ocfs2_xattr_create_index_block(struct inode *inode,
 
 	xb->xb_flags = cpu_to_le16(xb_flags | OCFS2_XATTR_INDEXED);
 
-	ret = ocfs2_journal_dirty(handle, xb_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
-
-out_commit:
-	ocfs2_commit_trans(osb, handle);
-
-out_sem:
-	up_write(&oi->ip_alloc_sem);
+	ocfs2_journal_dirty(handle, xb_bh);
 
 out:
-	if (data_ac)
-		ocfs2_free_alloc_context(data_ac);
-
-	brelse(xh_bh);
-	brelse(data_bh);
+	up_write(&oi->ip_alloc_sem);
 
 	return ret;
 }
@@ -2829,29 +3505,18 @@ static int cmp_xe_offset(const void *a, const void *b)
  * so that we can spare some space for insertion.
  */
 static int ocfs2_defrag_xattr_bucket(struct inode *inode,
+				     handle_t *handle,
 				     struct ocfs2_xattr_bucket *bucket)
 {
 	int ret, i;
 	size_t end, offset, len, value_len;
 	struct ocfs2_xattr_header *xh;
 	char *entries, *buf, *bucket_buf = NULL;
-	u64 blkno = bucket->bhs[0]->b_blocknr;
-	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+	u64 blkno = bucket_blkno(bucket);
 	u16 xh_free_start;
 	size_t blocksize = inode->i_sb->s_blocksize;
-	handle_t *handle;
-	struct buffer_head **bhs;
 	struct ocfs2_xattr_entry *xe;
 
-	bhs = kzalloc(sizeof(struct buffer_head *) * blk_per_bucket,
-			GFP_NOFS);
-	if (!bhs)
-		return -ENOMEM;
-
-	ret = ocfs2_read_blocks(inode, blkno, blk_per_bucket, bhs, 0);
-	if (ret)
-		goto out;
-
 	/*
 	 * In order to make the operation more efficient and generic,
 	 * we copy all the blocks into a contiguous memory and do the
@@ -2865,26 +3530,16 @@ static int ocfs2_defrag_xattr_bucket(struct inode *inode,
 	}
 
 	buf = bucket_buf;
-	for (i = 0; i < blk_per_bucket; i++, buf += blocksize)
-		memcpy(buf, bhs[i]->b_data, blocksize);
+	for (i = 0; i < bucket->bu_blocks; i++, buf += blocksize)
+		memcpy(buf, bucket_block(bucket, i), blocksize);
 
-	handle = ocfs2_start_trans((OCFS2_SB(inode->i_sb)), blk_per_bucket);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		handle = NULL;
+	ret = ocfs2_xattr_bucket_journal_access(handle, bucket,
+						OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	for (i = 0; i < blk_per_bucket; i++) {
-		ret = ocfs2_journal_access(handle, inode, bhs[i],
-					   OCFS2_JOURNAL_ACCESS_WRITE);
-		if (ret < 0) {
-			mlog_errno(ret);
-			goto commit;
-		}
-	}
-
 	xh = (struct ocfs2_xattr_header *)bucket_buf;
 	entries = (char *)xh->xh_entries;
 	xh_free_start = le16_to_cpu(xh->xh_free_start);
@@ -2940,7 +3595,7 @@ static int ocfs2_defrag_xattr_bucket(struct inode *inode,
 			"bucket %llu\n", (unsigned long long)blkno);
 
 	if (xh_free_start == end)
-		goto commit;
+		goto out;
 
 	memset(bucket_buf + xh_free_start, 0, end - xh_free_start);
 	xh->xh_free_start = cpu_to_le16(end);
@@ -2951,169 +3606,94 @@ static int ocfs2_defrag_xattr_bucket(struct inode *inode,
 	     cmp_xe, swap_xe);
 
 	buf = bucket_buf;
-	for (i = 0; i < blk_per_bucket; i++, buf += blocksize) {
-		memcpy(bhs[i]->b_data, buf, blocksize);
-		ocfs2_journal_dirty(handle, bhs[i]);
-	}
+	for (i = 0; i < bucket->bu_blocks; i++, buf += blocksize)
+		memcpy(bucket_block(bucket, i), buf, blocksize);
+	ocfs2_xattr_bucket_journal_dirty(handle, bucket);
 
-commit:
-	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
 out:
-
-	if (bhs) {
-		for (i = 0; i < blk_per_bucket; i++)
-			brelse(bhs[i]);
-	}
-	kfree(bhs);
-
 	kfree(bucket_buf);
 	return ret;
 }
 
 /*
- * Move half nums of the xattr bucket in the previous cluster to this new
- * cluster. We only touch the last cluster of the previous extend record.
+ * prev_blkno points to the start of an existing extent.  new_blkno
+ * points to a newly allocated extent.  Because we know each of our
+ * clusters contains more than bucket, we can easily split one cluster
+ * at a bucket boundary.  So we take the last cluster of the existing
+ * extent and split it down the middle.  We move the last half of the
+ * buckets in the last cluster of the existing extent over to the new
+ * extent.
  *
- * first_bh is the first buffer_head of a series of bucket in the same
- * extent rec and header_bh is the header of one bucket in this cluster.
- * They will be updated if we move the data header_bh contains to the new
- * cluster. first_hash will be set as the 1st xe's name_hash of the new cluster.
+ * first_bh is the buffer at prev_blkno so we can update the existing
+ * extent's bucket count.  header_bh is the bucket were we were hoping
+ * to insert our xattr.  If the bucket move places the target in the new
+ * extent, we'll update first_bh and header_bh after modifying the old
+ * extent.
+ *
+ * first_hash will be set as the 1st xe's name_hash in the new extent.
  */
 static int ocfs2_mv_xattr_bucket_cross_cluster(struct inode *inode,
 					       handle_t *handle,
-					       struct buffer_head **first_bh,
-					       struct buffer_head **header_bh,
+					       struct ocfs2_xattr_bucket *first,
+					       struct ocfs2_xattr_bucket *target,
 					       u64 new_blkno,
-					       u64 prev_blkno,
 					       u32 num_clusters,
 					       u32 *first_hash)
 {
-	int i, ret, credits;
-	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	int bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
-	int num_buckets = ocfs2_xattr_buckets_per_cluster(osb);
-	int blocksize = inode->i_sb->s_blocksize;
-	struct buffer_head *old_bh, *new_bh, *prev_bh, *new_first_bh = NULL;
-	struct ocfs2_xattr_header *new_xh;
-	struct ocfs2_xattr_header *xh =
-			(struct ocfs2_xattr_header *)((*first_bh)->b_data);
-
-	BUG_ON(le16_to_cpu(xh->xh_num_buckets) < num_buckets);
-	BUG_ON(OCFS2_XATTR_BUCKET_SIZE == osb->s_clustersize);
-
-	prev_bh = *first_bh;
-	get_bh(prev_bh);
-	xh = (struct ocfs2_xattr_header *)prev_bh->b_data;
+	int ret;
+	struct super_block *sb = inode->i_sb;
+	int blks_per_bucket = ocfs2_blocks_per_xattr_bucket(sb);
+	int num_buckets = ocfs2_xattr_buckets_per_cluster(OCFS2_SB(sb));
+	int to_move = num_buckets / 2;
+	u64 src_blkno;
+	u64 last_cluster_blkno = bucket_blkno(first) +
+		((num_clusters - 1) * ocfs2_clusters_to_blocks(sb, 1));
 
-	prev_blkno += (num_clusters - 1) * bpc + bpc / 2;
+	BUG_ON(le16_to_cpu(bucket_xh(first)->xh_num_buckets) < num_buckets);
+	BUG_ON(OCFS2_XATTR_BUCKET_SIZE == OCFS2_SB(sb)->s_clustersize);
 
 	mlog(0, "move half of xattrs in cluster %llu to %llu\n",
-	     (unsigned long long)prev_blkno, (unsigned long long)new_blkno);
+	     (unsigned long long)last_cluster_blkno, (unsigned long long)new_blkno);
 
-	/*
-	 * We need to update the 1st half of the new cluster and
-	 * 1 more for the update of the 1st bucket of the previous
-	 * extent record.
-	 */
-	credits = bpc / 2 + 1;
-	ret = ocfs2_extend_trans(handle, credits);
+	ret = ocfs2_mv_xattr_buckets(inode, handle, bucket_blkno(first),
+				     last_cluster_blkno, new_blkno,
+				     to_move, first_hash);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, prev_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
+	/* This is the first bucket that got moved */
+	src_blkno = last_cluster_blkno + (to_move * blks_per_bucket);
 
-	for (i = 0; i < bpc / 2; i++, prev_blkno++, new_blkno++) {
-		old_bh = new_bh = NULL;
-		new_bh = sb_getblk(inode->i_sb, new_blkno);
-		if (!new_bh) {
-			ret = -EIO;
-			mlog_errno(ret);
-			goto out;
-		}
+	/*
+	 * If the target bucket was part of the moved buckets, we need to
+	 * update first and target.
+	 */
+	if (bucket_blkno(target) >= src_blkno) {
+		/* Find the block for the new target bucket */
+		src_blkno = new_blkno +
+			(bucket_blkno(target) - src_blkno);
 
-		ocfs2_set_new_buffer_uptodate(inode, new_bh);
+		ocfs2_xattr_bucket_relse(first);
+		ocfs2_xattr_bucket_relse(target);
 
-		ret = ocfs2_journal_access(handle, inode, new_bh,
-					   OCFS2_JOURNAL_ACCESS_CREATE);
-		if (ret < 0) {
+		/*
+		 * These shouldn't fail - the buffers are in the
+		 * journal from ocfs2_cp_xattr_bucket().
+		 */
+		ret = ocfs2_read_xattr_bucket(first, new_blkno);
+		if (ret) {
 			mlog_errno(ret);
-			brelse(new_bh);
 			goto out;
 		}
-
-		ret = ocfs2_read_block(inode, prev_blkno, &old_bh);
-		if (ret < 0) {
+		ret = ocfs2_read_xattr_bucket(target, src_blkno);
+		if (ret)
 			mlog_errno(ret);
-			brelse(new_bh);
-			goto out;
-		}
-
-		memcpy(new_bh->b_data, old_bh->b_data, blocksize);
-
-		if (i == 0) {
-			new_xh = (struct ocfs2_xattr_header *)new_bh->b_data;
-			new_xh->xh_num_buckets = cpu_to_le16(num_buckets / 2);
 
-			if (first_hash)
-				*first_hash = le32_to_cpu(
-					new_xh->xh_entries[0].xe_name_hash);
-			new_first_bh = new_bh;
-			get_bh(new_first_bh);
-		}
-
-		ocfs2_journal_dirty(handle, new_bh);
-
-		if (*header_bh == old_bh) {
-			brelse(*header_bh);
-			*header_bh = new_bh;
-			get_bh(*header_bh);
-
-			brelse(*first_bh);
-			*first_bh = new_first_bh;
-			get_bh(*first_bh);
-		}
-		brelse(new_bh);
-		brelse(old_bh);
 	}
 
-	le16_add_cpu(&xh->xh_num_buckets, -(num_buckets / 2));
-
-	ocfs2_journal_dirty(handle, prev_bh);
 out:
-	brelse(prev_bh);
-	brelse(new_first_bh);
-	return ret;
-}
-
-static int ocfs2_read_xattr_bucket(struct inode *inode,
-				   u64 blkno,
-				   struct buffer_head **bhs,
-				   int new)
-{
-	int ret = 0;
-	u16 i, blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-
-	if (!new)
-		return ocfs2_read_blocks(inode, blkno,
-					 blk_per_bucket, bhs, 0);
-
-	for (i = 0; i < blk_per_bucket; i++) {
-		bhs[i] = sb_getblk(inode->i_sb, blkno + i);
-		if (bhs[i] == NULL) {
-			ret = -EIO;
-			mlog_errno(ret);
-			break;
-		}
-		ocfs2_set_new_buffer_uptodate(inode, bhs[i]);
-	}
-
 	return ret;
 }
 
@@ -3178,8 +3758,7 @@ static int ocfs2_divide_xattr_bucket(struct inode *inode,
 {
 	int ret, i;
 	int count, start, len, name_value_len = 0, xe_len, name_offset = 0;
-	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-	struct buffer_head **s_bhs, **t_bhs = NULL;
+	struct ocfs2_xattr_bucket *s_bucket = NULL, *t_bucket = NULL;
 	struct ocfs2_xattr_header *xh;
 	struct ocfs2_xattr_entry *xe;
 	int blocksize = inode->i_sb->s_blocksize;
@@ -3187,47 +3766,52 @@ static int ocfs2_divide_xattr_bucket(struct inode *inode,
 	mlog(0, "move some of xattrs from bucket %llu to %llu\n",
 	     (unsigned long long)blk, (unsigned long long)new_blk);
 
-	s_bhs = kcalloc(blk_per_bucket, sizeof(struct buffer_head *), GFP_NOFS);
-	if (!s_bhs)
-		return -ENOMEM;
-
-	ret = ocfs2_read_xattr_bucket(inode, blk, s_bhs, 0);
-	if (ret) {
+	s_bucket = ocfs2_xattr_bucket_new(inode);
+	t_bucket = ocfs2_xattr_bucket_new(inode);
+	if (!s_bucket || !t_bucket) {
+		ret = -ENOMEM;
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, s_bhs[0],
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_read_xattr_bucket(s_bucket, blk);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	t_bhs = kcalloc(blk_per_bucket, sizeof(struct buffer_head *), GFP_NOFS);
-	if (!t_bhs) {
-		ret = -ENOMEM;
+	ret = ocfs2_xattr_bucket_journal_access(handle, s_bucket,
+						OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret) {
+		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_read_xattr_bucket(inode, new_blk, t_bhs, new_bucket_head);
+	/*
+	 * Even if !new_bucket_head, we're overwriting t_bucket.  Thus,
+	 * there's no need to read it.
+	 */
+	ret = ocfs2_init_xattr_bucket(t_bucket, new_blk);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	for (i = 0; i < blk_per_bucket; i++) {
-		ret = ocfs2_journal_access(handle, inode, t_bhs[i],
-					   new_bucket_head ?
-					   OCFS2_JOURNAL_ACCESS_CREATE :
-					   OCFS2_JOURNAL_ACCESS_WRITE);
-		if (ret) {
-			mlog_errno(ret);
-			goto out;
-		}
+	/*
+	 * Hey, if we're overwriting t_bucket, what difference does
+	 * ACCESS_CREATE vs ACCESS_WRITE make?  See the comment in the
+	 * same part of ocfs2_cp_xattr_bucket().
+	 */
+	ret = ocfs2_xattr_bucket_journal_access(handle, t_bucket,
+						new_bucket_head ?
+						OCFS2_JOURNAL_ACCESS_CREATE :
+						OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
 	}
 
-	xh = (struct ocfs2_xattr_header *)s_bhs[0]->b_data;
+	xh = bucket_xh(s_bucket);
 	count = le16_to_cpu(xh->xh_count);
 	start = ocfs2_xattr_find_divide_pos(xh);
 
@@ -3239,10 +3823,10 @@ static int ocfs2_divide_xattr_bucket(struct inode *inode,
 		 * The hash value is set as one larger than
 		 * that of the last entry in the previous bucket.
 		 */
-		for (i = 0; i < blk_per_bucket; i++)
-			memset(t_bhs[i]->b_data, 0, blocksize);
+		for (i = 0; i < t_bucket->bu_blocks; i++)
+			memset(bucket_block(t_bucket, i), 0, blocksize);
 
-		xh = (struct ocfs2_xattr_header *)t_bhs[0]->b_data;
+		xh = bucket_xh(t_bucket);
 		xh->xh_free_start = cpu_to_le16(blocksize);
 		xh->xh_entries[0].xe_name_hash = xe->xe_name_hash;
 		le32_add_cpu(&xh->xh_entries[0].xe_name_hash, 1);
@@ -3251,11 +3835,10 @@ static int ocfs2_divide_xattr_bucket(struct inode *inode,
 	}
 
 	/* copy the whole bucket to the new first. */
-	for (i = 0; i < blk_per_bucket; i++)
-		memcpy(t_bhs[i]->b_data, s_bhs[i]->b_data, blocksize);
+	ocfs2_xattr_bucket_copy_data(t_bucket, s_bucket);
 
 	/* update the new bucket. */
-	xh = (struct ocfs2_xattr_header *)t_bhs[0]->b_data;
+	xh = bucket_xh(t_bucket);
 
 	/*
 	 * Calculate the total name/value len and xh_free_start for
@@ -3319,11 +3902,7 @@ set_num_buckets:
 	else
 		xh->xh_num_buckets = 0;
 
-	for (i = 0; i < blk_per_bucket; i++) {
-		ocfs2_journal_dirty(handle, t_bhs[i]);
-		if (ret)
-			mlog_errno(ret);
-	}
+	ocfs2_xattr_bucket_journal_dirty(handle, t_bucket);
 
 	/* store the first_hash of the new bucket. */
 	if (first_hash)
@@ -3337,29 +3916,18 @@ set_num_buckets:
 	if (start == count)
 		goto out;
 
-	xh = (struct ocfs2_xattr_header *)s_bhs[0]->b_data;
+	xh = bucket_xh(s_bucket);
 	memset(&xh->xh_entries[start], 0,
 	       sizeof(struct ocfs2_xattr_entry) * (count - start));
 	xh->xh_count = cpu_to_le16(start);
 	xh->xh_free_start = cpu_to_le16(name_offset);
 	xh->xh_name_value_len = cpu_to_le16(name_value_len);
 
-	ocfs2_journal_dirty(handle, s_bhs[0]);
-	if (ret)
-		mlog_errno(ret);
+	ocfs2_xattr_bucket_journal_dirty(handle, s_bucket);
 
 out:
-	if (s_bhs) {
-		for (i = 0; i < blk_per_bucket; i++)
-			brelse(s_bhs[i]);
-	}
-	kfree(s_bhs);
-
-	if (t_bhs) {
-		for (i = 0; i < blk_per_bucket; i++)
-			brelse(t_bhs[i]);
-	}
-	kfree(t_bhs);
+	ocfs2_xattr_bucket_free(s_bucket);
+	ocfs2_xattr_bucket_free(t_bucket);
 
 	return ret;
 }
@@ -3376,103 +3944,124 @@ static int ocfs2_cp_xattr_bucket(struct inode *inode,
 				 u64 t_blkno,
 				 int t_is_new)
 {
-	int ret, i;
-	int blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-	int blocksize = inode->i_sb->s_blocksize;
-	struct buffer_head **s_bhs, **t_bhs = NULL;
+	int ret;
+	struct ocfs2_xattr_bucket *s_bucket = NULL, *t_bucket = NULL;
 
 	BUG_ON(s_blkno == t_blkno);
 
 	mlog(0, "cp bucket %llu to %llu, target is %d\n",
 	     (unsigned long long)s_blkno, (unsigned long long)t_blkno,
-	     t_is_new);
-
-	s_bhs = kzalloc(sizeof(struct buffer_head *) * blk_per_bucket,
-			GFP_NOFS);
-	if (!s_bhs)
-		return -ENOMEM;
-
-	ret = ocfs2_read_xattr_bucket(inode, s_blkno, s_bhs, 0);
-	if (ret)
-		goto out;
+	     t_is_new);
 
-	t_bhs = kzalloc(sizeof(struct buffer_head *) * blk_per_bucket,
-			GFP_NOFS);
-	if (!t_bhs) {
+	s_bucket = ocfs2_xattr_bucket_new(inode);
+	t_bucket = ocfs2_xattr_bucket_new(inode);
+	if (!s_bucket || !t_bucket) {
 		ret = -ENOMEM;
+		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_read_xattr_bucket(inode, t_blkno, t_bhs, t_is_new);
+	ret = ocfs2_read_xattr_bucket(s_bucket, s_blkno);
 	if (ret)
 		goto out;
 
-	for (i = 0; i < blk_per_bucket; i++) {
-		ret = ocfs2_journal_access(handle, inode, t_bhs[i],
-					   t_is_new ?
-					   OCFS2_JOURNAL_ACCESS_CREATE :
-					   OCFS2_JOURNAL_ACCESS_WRITE);
-		if (ret)
-			goto out;
-	}
+	/*
+	 * Even if !t_is_new, we're overwriting t_bucket.  Thus,
+	 * there's no need to read it.
+	 */
+	ret = ocfs2_init_xattr_bucket(t_bucket, t_blkno);
+	if (ret)
+		goto out;
 
-	for (i = 0; i < blk_per_bucket; i++) {
-		memcpy(t_bhs[i]->b_data, s_bhs[i]->b_data, blocksize);
-		ocfs2_journal_dirty(handle, t_bhs[i]);
-	}
+	/*
+	 * Hey, if we're overwriting t_bucket, what difference does
+	 * ACCESS_CREATE vs ACCESS_WRITE make?  Well, if we allocated a new
+	 * cluster to fill, we came here from
+	 * ocfs2_mv_xattr_buckets(), and it is really new -
+	 * ACCESS_CREATE is required.  But we also might have moved data
+	 * out of t_bucket before extending back into it.
+	 * ocfs2_add_new_xattr_bucket() can do this - its call to
+	 * ocfs2_add_new_xattr_cluster() may have created a new extent
+	 * and copied out the end of the old extent.  Then it re-extends
+	 * the old extent back to create space for new xattrs.  That's
+	 * how we get here, and the bucket isn't really new.
+	 */
+	ret = ocfs2_xattr_bucket_journal_access(handle, t_bucket,
+						t_is_new ?
+						OCFS2_JOURNAL_ACCESS_CREATE :
+						OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret)
+		goto out;
 
-out:
-	if (s_bhs) {
-		for (i = 0; i < blk_per_bucket; i++)
-			brelse(s_bhs[i]);
-	}
-	kfree(s_bhs);
+	ocfs2_xattr_bucket_copy_data(t_bucket, s_bucket);
+	ocfs2_xattr_bucket_journal_dirty(handle, t_bucket);
 
-	if (t_bhs) {
-		for (i = 0; i < blk_per_bucket; i++)
-			brelse(t_bhs[i]);
-	}
-	kfree(t_bhs);
+out:
+	ocfs2_xattr_bucket_free(t_bucket);
+	ocfs2_xattr_bucket_free(s_bucket);
 
 	return ret;
 }
 
 /*
- * Copy one xattr cluster from src_blk to to_blk.
- * The to_blk will become the first bucket header of the cluster, so its
- * xh_num_buckets will be initialized as the bucket num in the cluster.
+ * src_blk points to the start of an existing extent.  last_blk points to
+ * last cluster in that extent.  to_blk points to a newly allocated
+ * extent.  We copy the buckets from the cluster at last_blk to the new
+ * extent.  If start_bucket is non-zero, we skip that many buckets before
+ * we start copying.  The new extent's xh_num_buckets gets set to the
+ * number of buckets we copied.  The old extent's xh_num_buckets shrinks
+ * by the same amount.
  */
-static int ocfs2_cp_xattr_cluster(struct inode *inode,
-				  handle_t *handle,
-				  struct buffer_head *first_bh,
-				  u64 src_blk,
-				  u64 to_blk,
+static int ocfs2_mv_xattr_buckets(struct inode *inode, handle_t *handle,
+				  u64 src_blk, u64 last_blk, u64 to_blk,
+				  unsigned int start_bucket,
 				  u32 *first_hash)
 {
 	int i, ret, credits;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	int bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
+	int blks_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
 	int num_buckets = ocfs2_xattr_buckets_per_cluster(osb);
-	struct buffer_head *bh = NULL;
-	struct ocfs2_xattr_header *xh;
-	u64 to_blk_start = to_blk;
+	struct ocfs2_xattr_bucket *old_first, *new_first;
+
+	mlog(0, "mv xattrs from cluster %llu to %llu\n",
+	     (unsigned long long)last_blk, (unsigned long long)to_blk);
+
+	BUG_ON(start_bucket >= num_buckets);
+	if (start_bucket) {
+		num_buckets -= start_bucket;
+		last_blk += (start_bucket * blks_per_bucket);
+	}
 
-	mlog(0, "cp xattrs from cluster %llu to %llu\n",
-	     (unsigned long long)src_blk, (unsigned long long)to_blk);
+	/* The first bucket of the original extent */
+	old_first = ocfs2_xattr_bucket_new(inode);
+	/* The first bucket of the new extent */
+	new_first = ocfs2_xattr_bucket_new(inode);
+	if (!old_first || !new_first) {
+		ret = -ENOMEM;
+		mlog_errno(ret);
+		goto out;
+	}
+
+	ret = ocfs2_read_xattr_bucket(old_first, src_blk);
+	if (ret) {
+		mlog_errno(ret);
+		goto out;
+	}
 
 	/*
-	 * We need to update the new cluster and 1 more for the update of
-	 * the 1st bucket of the previous extent rec.
+	 * We need to update the first bucket of the old extent and all
+	 * the buckets going to the new extent.
 	 */
-	credits = bpc + 1;
+	credits = ((num_buckets + 1) * blks_per_bucket) +
+		handle->h_buffer_credits;
 	ret = ocfs2_extend_trans(handle, credits);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, first_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_xattr_bucket_journal_access(handle, old_first,
+						OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -3480,45 +4069,45 @@ static int ocfs2_cp_xattr_cluster(struct inode *inode,
 
 	for (i = 0; i < num_buckets; i++) {
 		ret = ocfs2_cp_xattr_bucket(inode, handle,
-					    src_blk, to_blk, 1);
+					    last_blk + (i * blks_per_bucket),
+					    to_blk + (i * blks_per_bucket),
+					    1);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
-
-		src_blk += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-		to_blk += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
 	}
 
-	/* update the old bucket header. */
-	xh = (struct ocfs2_xattr_header *)first_bh->b_data;
-	le16_add_cpu(&xh->xh_num_buckets, -num_buckets);
-
-	ocfs2_journal_dirty(handle, first_bh);
-
-	/* update the new bucket header. */
-	ret = ocfs2_read_block(inode, to_blk_start, &bh);
-	if (ret < 0) {
+	/*
+	 * Get the new bucket ready before we dirty anything
+	 * (This actually shouldn't fail, because we already dirtied
+	 * it once in ocfs2_cp_xattr_bucket()).
+	 */
+	ret = ocfs2_read_xattr_bucket(new_first, to_blk);
+	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
-
-	ret = ocfs2_journal_access(handle, inode, bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_xattr_bucket_journal_access(handle, new_first,
+						OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	xh = (struct ocfs2_xattr_header *)bh->b_data;
-	xh->xh_num_buckets = cpu_to_le16(num_buckets);
+	/* Now update the headers */
+	le16_add_cpu(&bucket_xh(old_first)->xh_num_buckets, -num_buckets);
+	ocfs2_xattr_bucket_journal_dirty(handle, old_first);
 
-	ocfs2_journal_dirty(handle, bh);
+	bucket_xh(new_first)->xh_num_buckets = cpu_to_le16(num_buckets);
+	ocfs2_xattr_bucket_journal_dirty(handle, new_first);
 
 	if (first_hash)
-		*first_hash = le32_to_cpu(xh->xh_entries[0].xe_name_hash);
+		*first_hash = le32_to_cpu(bucket_xh(new_first)->xh_entries[0].xe_name_hash);
+
 out:
-	brelse(bh);
+	ocfs2_xattr_bucket_free(new_first);
+	ocfs2_xattr_bucket_free(old_first);
 	return ret;
 }
 
@@ -3534,7 +4123,7 @@ static int ocfs2_divide_xattr_cluster(struct inode *inode,
 				      u32 *first_hash)
 {
 	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-	int ret, credits = 2 * blk_per_bucket;
+	int ret, credits = 2 * blk_per_bucket + handle->h_buffer_credits;
 
 	BUG_ON(OCFS2_XATTR_BUCKET_SIZE < OCFS2_SB(inode->i_sb)->s_clustersize);
 
@@ -3577,43 +4166,49 @@ static int ocfs2_divide_xattr_cluster(struct inode *inode,
  */
 static int ocfs2_adjust_xattr_cross_cluster(struct inode *inode,
 					    handle_t *handle,
-					    struct buffer_head **first_bh,
-					    struct buffer_head **header_bh,
+					    struct ocfs2_xattr_bucket *first,
+					    struct ocfs2_xattr_bucket *target,
 					    u64 new_blk,
-					    u64 prev_blk,
 					    u32 prev_clusters,
 					    u32 *v_start,
 					    int *extend)
 {
-	int ret = 0;
-	int bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
+	int ret;
 
 	mlog(0, "adjust xattrs from cluster %llu len %u to %llu\n",
-	     (unsigned long long)prev_blk, prev_clusters,
+	     (unsigned long long)bucket_blkno(first), prev_clusters,
 	     (unsigned long long)new_blk);
 
-	if (ocfs2_xattr_buckets_per_cluster(OCFS2_SB(inode->i_sb)) > 1)
+	if (ocfs2_xattr_buckets_per_cluster(OCFS2_SB(inode->i_sb)) > 1) {
 		ret = ocfs2_mv_xattr_bucket_cross_cluster(inode,
 							  handle,
-							  first_bh,
-							  header_bh,
+							  first, target,
 							  new_blk,
-							  prev_blk,
 							  prev_clusters,
 							  v_start);
-	else {
-		u64 last_blk = prev_blk + bpc * (prev_clusters - 1);
-
-		if (prev_clusters > 1 && (*header_bh)->b_blocknr != last_blk)
-			ret = ocfs2_cp_xattr_cluster(inode, handle, *first_bh,
-						     last_blk, new_blk,
+		if (ret)
+			mlog_errno(ret);
+	} else {
+		/* The start of the last cluster in the first extent */
+		u64 last_blk = bucket_blkno(first) +
+			((prev_clusters - 1) *
+			 ocfs2_clusters_to_blocks(inode->i_sb, 1));
+
+		if (prev_clusters > 1 && bucket_blkno(target) != last_blk) {
+			ret = ocfs2_mv_xattr_buckets(inode, handle,
+						     bucket_blkno(first),
+						     last_blk, new_blk, 0,
 						     v_start);
-		else {
+			if (ret)
+				mlog_errno(ret);
+		} else {
 			ret = ocfs2_divide_xattr_cluster(inode, handle,
 							 last_blk, new_blk,
 							 v_start);
+			if (ret)
+				mlog_errno(ret);
 
-			if ((*header_bh)->b_blocknr == last_blk && extend)
+			if ((bucket_blkno(target) == last_blk) && extend)
 				*extend = 0;
 		}
 	}
@@ -3639,56 +4234,37 @@ static int ocfs2_adjust_xattr_cross_cluster(struct inode *inode,
  */
 static int ocfs2_add_new_xattr_cluster(struct inode *inode,
 				       struct buffer_head *root_bh,
-				       struct buffer_head **first_bh,
-				       struct buffer_head **header_bh,
+				       struct ocfs2_xattr_bucket *first,
+				       struct ocfs2_xattr_bucket *target,
 				       u32 *num_clusters,
 				       u32 prev_cpos,
-				       u64 prev_blkno,
-				       int *extend)
+				       int *extend,
+				       struct ocfs2_xattr_set_ctxt *ctxt)
 {
-	int ret, credits;
+	int ret;
 	u16 bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
 	u32 prev_clusters = *num_clusters;
 	u32 clusters_to_add = 1, bit_off, num_bits, v_start = 0;
 	u64 block;
-	handle_t *handle = NULL;
-	struct ocfs2_alloc_context *data_ac = NULL;
-	struct ocfs2_alloc_context *meta_ac = NULL;
+	handle_t *handle = ctxt->handle;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct ocfs2_extent_tree et;
 
 	mlog(0, "Add new xattr cluster for %llu, previous xattr hash = %u, "
 	     "previous xattr blkno = %llu\n",
 	     (unsigned long long)OCFS2_I(inode)->ip_blkno,
-	     prev_cpos, (unsigned long long)prev_blkno);
+	     prev_cpos, (unsigned long long)bucket_blkno(first));
 
 	ocfs2_init_xattr_tree_extent_tree(&et, inode, root_bh);
 
-	ret = ocfs2_lock_allocators(inode, &et, clusters_to_add, 0,
-				    &data_ac, &meta_ac);
-	if (ret) {
-		mlog_errno(ret);
-		goto leave;
-	}
-
-	credits = ocfs2_calc_extend_credits(osb->sb, et.et_root_el,
-					    clusters_to_add);
-	handle = ocfs2_start_trans(osb, credits);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		handle = NULL;
-		mlog_errno(ret);
-		goto leave;
-	}
-
-	ret = ocfs2_journal_access(handle, inode, root_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_xb(handle, inode, root_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto leave;
 	}
 
-	ret = __ocfs2_claim_clusters(osb, handle, data_ac, 1,
+	ret = __ocfs2_claim_clusters(osb, handle, ctxt->data_ac, 1,
 				     clusters_to_add, &bit_off, &num_bits);
 	if (ret < 0) {
 		if (ret != -ENOSPC)
@@ -3702,7 +4278,7 @@ static int ocfs2_add_new_xattr_cluster(struct inode *inode,
 	mlog(0, "Allocating %u clusters at block %u for xattr in inode %llu\n",
 	     num_bits, bit_off, (unsigned long long)OCFS2_I(inode)->ip_blkno);
 
-	if (prev_blkno + prev_clusters * bpc == block &&
+	if (bucket_blkno(first) + (prev_clusters * bpc) == block &&
 	    (prev_clusters + num_bits) << osb->s_clustersize_bits <=
 	     OCFS2_MAX_XATTR_TREE_LEAF_SIZE) {
 		/*
@@ -3721,10 +4297,9 @@ static int ocfs2_add_new_xattr_cluster(struct inode *inode,
 	} else {
 		ret = ocfs2_adjust_xattr_cross_cluster(inode,
 						       handle,
-						       first_bh,
-						       header_bh,
+						       first,
+						       target,
 						       block,
-						       prev_blkno,
 						       prev_clusters,
 						       &v_start,
 						       extend);
@@ -3734,149 +4309,137 @@ static int ocfs2_add_new_xattr_cluster(struct inode *inode,
 		}
 	}
 
-	if (handle->h_buffer_credits < credits) {
-		/*
-		 * The journal has been restarted before, and don't
-		 * have enough space for the insertion, so extend it
-		 * here.
-		 */
-		ret = ocfs2_extend_trans(handle, credits);
-		if (ret) {
-			mlog_errno(ret);
-			goto leave;
-		}
-	}
 	mlog(0, "Insert %u clusters at block %llu for xattr at %u\n",
 	     num_bits, (unsigned long long)block, v_start);
 	ret = ocfs2_insert_extent(osb, handle, inode, &et, v_start, block,
-				  num_bits, 0, meta_ac);
+				  num_bits, 0, ctxt->meta_ac);
 	if (ret < 0) {
 		mlog_errno(ret);
 		goto leave;
 	}
 
 	ret = ocfs2_journal_dirty(handle, root_bh);
-	if (ret < 0) {
+	if (ret < 0)
 		mlog_errno(ret);
-		goto leave;
-	}
 
 leave:
-	if (handle)
-		ocfs2_commit_trans(osb, handle);
-	if (data_ac)
-		ocfs2_free_alloc_context(data_ac);
-	if (meta_ac)
-		ocfs2_free_alloc_context(meta_ac);
-
 	return ret;
 }
 
 /*
- * Extend a new xattr bucket and move xattrs to the end one by one until
- * We meet with start_bh. Only move half of the xattrs to the bucket after it.
+ * We are given an extent.  'first' is the bucket at the very front of
+ * the extent.  The extent has space for an additional bucket past
+ * bucket_xh(first)->xh_num_buckets.  'target_blkno' is the block number
+ * of the target bucket.  We wish to shift every bucket past the target
+ * down one, filling in that additional space.  When we get back to the
+ * target, we split the target between itself and the now-empty bucket
+ * at target+1 (aka, target_blkno + blks_per_bucket).
  */
 static int ocfs2_extend_xattr_bucket(struct inode *inode,
-				     struct buffer_head *first_bh,
-				     struct buffer_head *start_bh,
+				     handle_t *handle,
+				     struct ocfs2_xattr_bucket *first,
+				     u64 target_blk,
 				     u32 num_clusters)
 {
 	int ret, credits;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-	u64 start_blk = start_bh->b_blocknr, end_blk;
-	u32 num_buckets = num_clusters * ocfs2_xattr_buckets_per_cluster(osb);
-	handle_t *handle;
-	struct ocfs2_xattr_header *first_xh =
-				(struct ocfs2_xattr_header *)first_bh->b_data;
-	u16 bucket = le16_to_cpu(first_xh->xh_num_buckets);
+	u64 end_blk;
+	u16 new_bucket = le16_to_cpu(bucket_xh(first)->xh_num_buckets);
 
 	mlog(0, "extend xattr bucket in %llu, xattr extend rec starting "
-	     "from %llu, len = %u\n", (unsigned long long)start_blk,
-	     (unsigned long long)first_bh->b_blocknr, num_clusters);
+	     "from %llu, len = %u\n", (unsigned long long)target_blk,
+	     (unsigned long long)bucket_blkno(first), num_clusters);
 
-	BUG_ON(bucket >= num_buckets);
+	/* The extent must have room for an additional bucket */
+	BUG_ON(new_bucket >=
+	       (num_clusters * ocfs2_xattr_buckets_per_cluster(osb)));
 
-	end_blk = first_bh->b_blocknr + (bucket - 1) * blk_per_bucket;
+	/* end_blk points to the last existing bucket */
+	end_blk = bucket_blkno(first) + ((new_bucket - 1) * blk_per_bucket);
 
 	/*
-	 * We will touch all the buckets after the start_bh(include it).
-	 * Add one more bucket and modify the first_bh.
+	 * end_blk is the start of the last existing bucket.
+	 * Thus, (end_blk - target_blk) covers the target bucket and
+	 * every bucket after it up to, but not including, the last
+	 * existing bucket.  Then we add the last existing bucket, the
+	 * new bucket, and the first bucket (3 * blk_per_bucket).
 	 */
-	credits = end_blk - start_blk + 2 * blk_per_bucket + 1;
-	handle = ocfs2_start_trans(osb, credits);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		handle = NULL;
+	credits = (end_blk - target_blk) + (3 * blk_per_bucket) +
+		  handle->h_buffer_credits;
+	ret = ocfs2_extend_trans(handle, credits);
+	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, first_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_xattr_bucket_journal_access(handle, first,
+						OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto commit;
+		goto out;
 	}
 
-	while (end_blk != start_blk) {
+	while (end_blk != target_blk) {
 		ret = ocfs2_cp_xattr_bucket(inode, handle, end_blk,
 					    end_blk + blk_per_bucket, 0);
 		if (ret)
-			goto commit;
+			goto out;
 		end_blk -= blk_per_bucket;
 	}
 
-	/* Move half of the xattr in start_blk to the next bucket. */
-	ret = ocfs2_divide_xattr_bucket(inode, handle, start_blk,
-					start_blk + blk_per_bucket, NULL, 0);
+	/* Move half of the xattr in target_blkno to the next bucket. */
+	ret = ocfs2_divide_xattr_bucket(inode, handle, target_blk,
+					target_blk + blk_per_bucket, NULL, 0);
 
-	le16_add_cpu(&first_xh->xh_num_buckets, 1);
-	ocfs2_journal_dirty(handle, first_bh);
+	le16_add_cpu(&bucket_xh(first)->xh_num_buckets, 1);
+	ocfs2_xattr_bucket_journal_dirty(handle, first);
 
-commit:
-	ocfs2_commit_trans(osb, handle);
 out:
 	return ret;
 }
 
 /*
- * Add new xattr bucket in an extent record and adjust the buckets accordingly.
- * xb_bh is the ocfs2_xattr_block.
- * We will move all the buckets starting from header_bh to the next place. As
- * for this one, half num of its xattrs will be moved to the next one.
+ * Add new xattr bucket in an extent record and adjust the buckets
+ * accordingly.  xb_bh is the ocfs2_xattr_block, and target is the
+ * bucket we want to insert into.
  *
- * We will allocate a new cluster if current cluster is full and adjust
- * header_bh and first_bh if the insert place is moved to the new cluster.
+ * In the easy case, we will move all the buckets after target down by
+ * one. Half of target's xattrs will be moved to the next bucket.
+ *
+ * If current cluster is full, we'll allocate a new one.  This may not
+ * be contiguous.  The underlying calls will make sure that there is
+ * space for the insert, shifting buckets around if necessary.
+ * 'target' may be moved by those calls.
  */
 static int ocfs2_add_new_xattr_bucket(struct inode *inode,
 				      struct buffer_head *xb_bh,
-				      struct buffer_head *header_bh)
+				      struct ocfs2_xattr_bucket *target,
+				      struct ocfs2_xattr_set_ctxt *ctxt)
 {
-	struct ocfs2_xattr_header *first_xh = NULL;
-	struct buffer_head *first_bh = NULL;
 	struct ocfs2_xattr_block *xb =
 			(struct ocfs2_xattr_block *)xb_bh->b_data;
 	struct ocfs2_xattr_tree_root *xb_root = &xb->xb_attrs.xb_root;
 	struct ocfs2_extent_list *el = &xb_root->xt_list;
-	struct ocfs2_xattr_header *xh =
-			(struct ocfs2_xattr_header *)header_bh->b_data;
-	u32 name_hash = le32_to_cpu(xh->xh_entries[0].xe_name_hash);
-	struct super_block *sb = inode->i_sb;
-	struct ocfs2_super *osb = OCFS2_SB(sb);
+	u32 name_hash =
+		le32_to_cpu(bucket_xh(target)->xh_entries[0].xe_name_hash);
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	int ret, num_buckets, extend = 1;
 	u64 p_blkno;
 	u32 e_cpos, num_clusters;
+	/* The bucket at the front of the extent */
+	struct ocfs2_xattr_bucket *first;
 
-	mlog(0, "Add new xattr bucket starting form %llu\n",
-	     (unsigned long long)header_bh->b_blocknr);
+	mlog(0, "Add new xattr bucket starting from %llu\n",
+	     (unsigned long long)bucket_blkno(target));
 
-	/*
-	 * Add refrence for header_bh here because it may be
-	 * changed in ocfs2_add_new_xattr_cluster and we need
-	 * to free it in the end.
-	 */
-	get_bh(header_bh);
+	/* The first bucket of the original extent */
+	first = ocfs2_xattr_bucket_new(inode);
+	if (!first) {
+		ret = -ENOMEM;
+		mlog_errno(ret);
+		goto out;
+	}
 
 	ret = ocfs2_xattr_get_rec(inode, name_hash, &p_blkno, &e_cpos,
 				  &num_clusters, el);
@@ -3885,40 +4448,45 @@ static int ocfs2_add_new_xattr_bucket(struct inode *inode,
 		goto out;
 	}
 
-	ret = ocfs2_read_block(inode, p_blkno, &first_bh);
+	ret = ocfs2_read_xattr_bucket(first, p_blkno);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
 	num_buckets = ocfs2_xattr_buckets_per_cluster(osb) * num_clusters;
-	first_xh = (struct ocfs2_xattr_header *)first_bh->b_data;
-
-	if (num_buckets == le16_to_cpu(first_xh->xh_num_buckets)) {
+	if (num_buckets == le16_to_cpu(bucket_xh(first)->xh_num_buckets)) {
+		/*
+		 * This can move first+target if the target bucket moves
+		 * to the new extent.
+		 */
 		ret = ocfs2_add_new_xattr_cluster(inode,
 						  xb_bh,
-						  &first_bh,
-						  &header_bh,
+						  first,
+						  target,
 						  &num_clusters,
 						  e_cpos,
-						  p_blkno,
-						  &extend);
+						  &extend,
+						  ctxt);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
 	}
 
-	if (extend)
+	if (extend) {
 		ret = ocfs2_extend_xattr_bucket(inode,
-						first_bh,
-						header_bh,
+						ctxt->handle,
+						first,
+						bucket_blkno(target),
 						num_clusters);
-	if (ret)
-		mlog_errno(ret);
+		if (ret)
+			mlog_errno(ret);
+	}
+
 out:
-	brelse(first_bh);
-	brelse(header_bh);
+	ocfs2_xattr_bucket_free(first);
+
 	return ret;
 }
 
@@ -3929,7 +4497,7 @@ static inline char *ocfs2_xattr_bucket_get_val(struct inode *inode,
 	int block_off = offs >> inode->i_sb->s_blocksize_bits;
 
 	offs = offs % inode->i_sb->s_blocksize;
-	return bucket->bhs[block_off]->b_data + offs;
+	return bucket_block(bucket, block_off) + offs;
 }
 
 /*
@@ -3984,7 +4552,7 @@ static void ocfs2_xattr_set_entry_normal(struct inode *inode,
 				xe->xe_value_size = 0;
 
 			val = ocfs2_xattr_bucket_get_val(inode,
-							 &xs->bucket, offs);
+							 xs->bucket, offs);
 			memset(val + OCFS2_XATTR_SIZE(name_len), 0,
 			       size - OCFS2_XATTR_SIZE(name_len));
 			if (OCFS2_XATTR_SIZE(xi->value_len) > 0)
@@ -4062,8 +4630,7 @@ set_new_name_value:
 		xh->xh_free_start = cpu_to_le16(offs);
 	}
 
-	val = ocfs2_xattr_bucket_get_val(inode,
-					 &xs->bucket, offs - size);
+	val = ocfs2_xattr_bucket_get_val(inode, xs->bucket, offs - size);
 	xe->xe_name_offset = cpu_to_le16(offs - size);
 
 	memset(val, 0, size);
@@ -4079,125 +4646,45 @@ set_new_name_value:
 	return;
 }
 
-static int ocfs2_xattr_bucket_handle_journal(struct inode *inode,
-					     handle_t *handle,
-					     struct ocfs2_xattr_search *xs,
-					     struct buffer_head **bhs,
-					     u16 bh_num)
-{
-	int ret = 0, off, block_off;
-	struct ocfs2_xattr_entry *xe = xs->here;
-
-	/*
-	 * First calculate all the blocks we should journal_access
-	 * and journal_dirty. The first block should always be touched.
-	 */
-	ret = ocfs2_journal_dirty(handle, bhs[0]);
-	if (ret)
-		mlog_errno(ret);
-
-	/* calc the data. */
-	off = le16_to_cpu(xe->xe_name_offset);
-	block_off = off >> inode->i_sb->s_blocksize_bits;
-	ret = ocfs2_journal_dirty(handle, bhs[block_off]);
-	if (ret)
-		mlog_errno(ret);
-
-	return ret;
-}
-
 /*
  * Set the xattr entry in the specified bucket.
  * The bucket is indicated by xs->bucket and it should have the enough
  * space for the xattr insertion.
  */
 static int ocfs2_xattr_set_entry_in_bucket(struct inode *inode,
+					   handle_t *handle,
 					   struct ocfs2_xattr_info *xi,
 					   struct ocfs2_xattr_search *xs,
 					   u32 name_hash,
 					   int local)
 {
-	int i, ret;
-	handle_t *handle = NULL;
-	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
-	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	int ret;
+	u64 blkno;
 
 	mlog(0, "Set xattr entry len = %lu index = %d in bucket %llu\n",
 	     (unsigned long)xi->value_len, xi->name_index,
-	     (unsigned long long)xs->bucket.bhs[0]->b_blocknr);
+	     (unsigned long long)bucket_blkno(xs->bucket));
 
-	if (!xs->bucket.bhs[1]) {
-		ret = ocfs2_read_blocks(inode,
-					xs->bucket.bhs[0]->b_blocknr + 1,
-					blk_per_bucket - 1, &xs->bucket.bhs[1],
-					0);
+	if (!xs->bucket->bu_bhs[1]) {
+		blkno = bucket_blkno(xs->bucket);
+		ocfs2_xattr_bucket_relse(xs->bucket);
+		ret = ocfs2_read_xattr_bucket(xs->bucket, blkno);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, blk_per_bucket);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		handle = NULL;
+	ret = ocfs2_xattr_bucket_journal_access(handle, xs->bucket,
+						OCFS2_JOURNAL_ACCESS_WRITE);
+	if (ret < 0) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	for (i = 0; i < blk_per_bucket; i++) {
-		ret = ocfs2_journal_access(handle, inode, xs->bucket.bhs[i],
-					   OCFS2_JOURNAL_ACCESS_WRITE);
-		if (ret < 0) {
-			mlog_errno(ret);
-			goto out;
-		}
-	}
-
 	ocfs2_xattr_set_entry_normal(inode, xi, xs, name_hash, local);
+	ocfs2_xattr_bucket_journal_dirty(handle, xs->bucket);
 
-	/*Only dirty the blocks we have touched in set xattr. */
-	ret = ocfs2_xattr_bucket_handle_journal(inode, handle, xs,
-						xs->bucket.bhs, blk_per_bucket);
-	if (ret)
-		mlog_errno(ret);
-out:
-	ocfs2_commit_trans(osb, handle);
-
-	return ret;
-}
-
-static int ocfs2_xattr_value_update_size(struct inode *inode,
-					 struct buffer_head *xe_bh,
-					 struct ocfs2_xattr_entry *xe,
-					 u64 new_size)
-{
-	int ret;
-	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-	handle_t *handle = NULL;
-
-	handle = ocfs2_start_trans(osb, 1);
-	if (IS_ERR(handle)) {
-		ret = -ENOMEM;
-		mlog_errno(ret);
-		goto out;
-	}
-
-	ret = ocfs2_journal_access(handle, inode, xe_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
-	if (ret < 0) {
-		mlog_errno(ret);
-		goto out_commit;
-	}
-
-	xe->xe_value_size = cpu_to_le64(new_size);
-
-	ret = ocfs2_journal_dirty(handle, xe_bh);
-	if (ret < 0)
-		mlog_errno(ret);
-
-out_commit:
-	ocfs2_commit_trans(osb, handle);
 out:
 	return ret;
 }
@@ -4210,18 +4697,19 @@ out:
  * Copy the new updated xe and xe_value_root to new_xe and new_xv if needed.
  */
 static int ocfs2_xattr_bucket_value_truncate(struct inode *inode,
-					     struct buffer_head *header_bh,
+					     struct ocfs2_xattr_bucket *bucket,
 					     int xe_off,
-					     int len)
+					     int len,
+					     struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	int ret, offset;
 	u64 value_blk;
-	struct buffer_head *value_bh = NULL;
-	struct ocfs2_xattr_value_root *xv;
 	struct ocfs2_xattr_entry *xe;
-	struct ocfs2_xattr_header *xh =
-			(struct ocfs2_xattr_header *)header_bh->b_data;
+	struct ocfs2_xattr_header *xh = bucket_xh(bucket);
 	size_t blocksize = inode->i_sb->s_blocksize;
+	struct ocfs2_xattr_value_buf vb = {
+		.vb_access = ocfs2_journal_access,
+	};
 
 	xe = &xh->xh_entries[xe_off];
 
@@ -4234,49 +4722,58 @@ static int ocfs2_xattr_bucket_value_truncate(struct inode *inode,
 
 	/* We don't allow ocfs2_xattr_value to be stored in different block. */
 	BUG_ON(value_blk != (offset + OCFS2_XATTR_ROOT_SIZE - 1) / blocksize);
-	value_blk += header_bh->b_blocknr;
 
-	ret = ocfs2_read_block(inode, value_blk, &value_bh);
-	if (ret) {
-		mlog_errno(ret);
-		goto out;
-	}
+	vb.vb_bh = bucket->bu_bhs[value_blk];
+	BUG_ON(!vb.vb_bh);
 
-	xv = (struct ocfs2_xattr_value_root *)
-		(value_bh->b_data + offset % blocksize);
+	vb.vb_xv = (struct ocfs2_xattr_value_root *)
+		(vb.vb_bh->b_data + offset % blocksize);
 
-	mlog(0, "truncate %u in xattr bucket %llu to %d bytes.\n",
-	     xe_off, (unsigned long long)header_bh->b_blocknr, len);
-	ret = ocfs2_xattr_value_truncate(inode, value_bh, xv, len);
+	ret = ocfs2_xattr_bucket_journal_access(ctxt->handle, bucket,
+						OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_xattr_value_update_size(inode, header_bh, xe, len);
+	/*
+	 * From here on out we have to dirty the bucket.  The generic
+	 * value calls only modify one of the bucket's bhs, but we need
+	 * to send the bucket at once.  So if they error, they *could* have
+	 * modified something.  We have to assume they did, and dirty
+	 * the whole bucket.  This leaves us in a consistent state.
+	 */
+	mlog(0, "truncate %u in xattr bucket %llu to %d bytes.\n",
+	     xe_off, (unsigned long long)bucket_blkno(bucket), len);
+	ret = ocfs2_xattr_value_truncate(inode, &vb, len, ctxt);
 	if (ret) {
 		mlog_errno(ret);
-		goto out;
+		goto out_dirty;
 	}
 
+	xe->xe_value_size = cpu_to_le64(len);
+
+out_dirty:
+	ocfs2_xattr_bucket_journal_dirty(ctxt->handle, bucket);
+
 out:
-	brelse(value_bh);
 	return ret;
 }
 
 static int ocfs2_xattr_bucket_value_truncate_xs(struct inode *inode,
-						struct ocfs2_xattr_search *xs,
-						int len)
+					struct ocfs2_xattr_search *xs,
+					int len,
+					struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	int ret, offset;
 	struct ocfs2_xattr_entry *xe = xs->here;
 	struct ocfs2_xattr_header *xh = (struct ocfs2_xattr_header *)xs->base;
 
-	BUG_ON(!xs->bucket.bhs[0] || !xe || ocfs2_xattr_is_local(xe));
+	BUG_ON(!xs->bucket->bu_bhs[0] || !xe || ocfs2_xattr_is_local(xe));
 
 	offset = xe - xh->xh_entries;
-	ret = ocfs2_xattr_bucket_value_truncate(inode, xs->bucket.bhs[0],
-						offset, len);
+	ret = ocfs2_xattr_bucket_value_truncate(inode, xs->bucket,
+						offset, len, ctxt);
 	if (ret)
 		mlog_errno(ret);
 
@@ -4284,6 +4781,7 @@ static int ocfs2_xattr_bucket_value_truncate_xs(struct inode *inode,
 }
 
 static int ocfs2_xattr_bucket_set_value_outside(struct inode *inode,
+						handle_t *handle,
 						struct ocfs2_xattr_search *xs,
 						char *val,
 						int value_len)
@@ -4299,7 +4797,8 @@ static int ocfs2_xattr_bucket_set_value_outside(struct inode *inode,
 
 	xv = (struct ocfs2_xattr_value_root *)(xs->base + offset);
 
-	return __ocfs2_xattr_set_value_outside(inode, xv, val, value_len);
+	return __ocfs2_xattr_set_value_outside(inode, handle,
+					       xv, val, value_len);
 }
 
 static int ocfs2_rm_xattr_cluster(struct inode *inode,
@@ -4343,15 +4842,15 @@ static int ocfs2_rm_xattr_cluster(struct inode *inode,
 		}
 	}
 
-	handle = ocfs2_start_trans(osb, OCFS2_REMOVE_EXTENT_CREDITS);
+	handle = ocfs2_start_trans(osb, ocfs2_remove_extent_credits(osb->sb));
 	if (IS_ERR(handle)) {
 		ret = -ENOMEM;
 		mlog_errno(ret);
 		goto out;
 	}
 
-	ret = ocfs2_journal_access(handle, inode, root_bh,
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_journal_access_xb(handle, inode, root_bh,
+				      OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
 		goto out_commit;
@@ -4392,26 +4891,19 @@ out:
 }
 
 static void ocfs2_xattr_bucket_remove_xs(struct inode *inode,
+					 handle_t *handle,
 					 struct ocfs2_xattr_search *xs)
 {
-	handle_t *handle = NULL;
-	struct ocfs2_xattr_header *xh = xs->bucket.xh;
+	struct ocfs2_xattr_header *xh = bucket_xh(xs->bucket);
 	struct ocfs2_xattr_entry *last = &xh->xh_entries[
 						le16_to_cpu(xh->xh_count) - 1];
 	int ret = 0;
 
-	handle = ocfs2_start_trans((OCFS2_SB(inode->i_sb)), 1);
-	if (IS_ERR(handle)) {
-		ret = PTR_ERR(handle);
-		mlog_errno(ret);
-		return;
-	}
-
-	ret = ocfs2_journal_access(handle, inode, xs->bucket.bhs[0],
-				   OCFS2_JOURNAL_ACCESS_WRITE);
+	ret = ocfs2_xattr_bucket_journal_access(handle, xs->bucket,
+						OCFS2_JOURNAL_ACCESS_WRITE);
 	if (ret) {
 		mlog_errno(ret);
-		goto out_commit;
+		return;
 	}
 
 	/* Remove the old entry. */
@@ -4420,11 +4912,7 @@ static void ocfs2_xattr_bucket_remove_xs(struct inode *inode,
 	memset(last, 0, sizeof(struct ocfs2_xattr_entry));
 	le16_add_cpu(&xh->xh_count, -1);
 
-	ret = ocfs2_journal_dirty(handle, xs->bucket.bhs[0]);
-	if (ret < 0)
-		mlog_errno(ret);
-out_commit:
-	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
+	ocfs2_xattr_bucket_journal_dirty(handle, xs->bucket);
 }
 
 /*
@@ -4440,7 +4928,8 @@ out_commit:
  */
 static int ocfs2_xattr_set_in_bucket(struct inode *inode,
 				     struct ocfs2_xattr_info *xi,
-				     struct ocfs2_xattr_search *xs)
+				     struct ocfs2_xattr_search *xs,
+				     struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	int ret, local = 1;
 	size_t value_len;
@@ -4468,7 +4957,8 @@ static int ocfs2_xattr_set_in_bucket(struct inode *inode,
 			value_len = 0;
 
 		ret = ocfs2_xattr_bucket_value_truncate_xs(inode, xs,
-							   value_len);
+							   value_len,
+							   ctxt);
 		if (ret)
 			goto out;
 
@@ -4488,7 +4978,8 @@ static int ocfs2_xattr_set_in_bucket(struct inode *inode,
 		xi->value_len = OCFS2_XATTR_ROOT_SIZE;
 	}
 
-	ret = ocfs2_xattr_set_entry_in_bucket(inode, xi, xs, name_hash, local);
+	ret = ocfs2_xattr_set_entry_in_bucket(inode, ctxt->handle, xi, xs,
+					      name_hash, local);
 	if (ret) {
 		mlog_errno(ret);
 		goto out;
@@ -4499,7 +4990,7 @@ static int ocfs2_xattr_set_in_bucket(struct inode *inode,
 
 	/* allocate the space now for the outside block storage. */
 	ret = ocfs2_xattr_bucket_value_truncate_xs(inode, xs,
-						   value_len);
+						   value_len, ctxt);
 	if (ret) {
 		mlog_errno(ret);
 
@@ -4509,13 +5000,14 @@ static int ocfs2_xattr_set_in_bucket(struct inode *inode,
 			 * storage and we have allocated xattr already,
 			 * so need to remove it.
 			 */
-			ocfs2_xattr_bucket_remove_xs(inode, xs);
+			ocfs2_xattr_bucket_remove_xs(inode, ctxt->handle, xs);
 		}
 		goto out;
 	}
 
 set_value_outside:
-	ret = ocfs2_xattr_bucket_set_value_outside(inode, xs, val, value_len);
+	ret = ocfs2_xattr_bucket_set_value_outside(inode, ctxt->handle,
+						   xs, val, value_len);
 out:
 	return ret;
 }
@@ -4530,7 +5022,7 @@ static int ocfs2_check_xattr_bucket_collision(struct inode *inode,
 					      struct ocfs2_xattr_bucket *bucket,
 					      const char *name)
 {
-	struct ocfs2_xattr_header *xh = bucket->xh;
+	struct ocfs2_xattr_header *xh = bucket_xh(bucket);
 	u32 name_hash = ocfs2_xattr_name_hash(inode, name, strlen(name));
 
 	if (name_hash != le32_to_cpu(xh->xh_entries[0].xe_name_hash))
@@ -4540,7 +5032,7 @@ static int ocfs2_check_xattr_bucket_collision(struct inode *inode,
 	    xh->xh_entries[0].xe_name_hash) {
 		mlog(ML_ERROR, "Too much hash collision in xattr bucket %llu, "
 		     "hash = %u\n",
-		     (unsigned long long)bucket->bhs[0]->b_blocknr,
+		     (unsigned long long)bucket_blkno(bucket),
 		     le32_to_cpu(xh->xh_entries[0].xe_name_hash));
 		return -ENOSPC;
 	}
@@ -4550,16 +5042,16 @@ static int ocfs2_check_xattr_bucket_collision(struct inode *inode,
 
 static int ocfs2_xattr_set_entry_index_block(struct inode *inode,
 					     struct ocfs2_xattr_info *xi,
-					     struct ocfs2_xattr_search *xs)
+					     struct ocfs2_xattr_search *xs,
+					     struct ocfs2_xattr_set_ctxt *ctxt)
 {
 	struct ocfs2_xattr_header *xh;
 	struct ocfs2_xattr_entry *xe;
 	u16 count, header_size, xh_free_start;
-	int i, free, max_free, need, old;
+	int free, max_free, need, old;
 	size_t value_size = 0, name_len = strlen(xi->name);
 	size_t blocksize = inode->i_sb->s_blocksize;
 	int ret, allocation = 0;
-	u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb);
 
 	mlog_entry("Set xattr %s in xattr index block\n", xi->name);
 
@@ -4574,7 +5066,7 @@ try_again:
 
 	mlog_bug_on_msg(header_size > blocksize, "bucket %llu has header size "
 			"of %u which exceed block size\n",
-			(unsigned long long)xs->bucket.bhs[0]->b_blocknr,
+			(unsigned long long)bucket_blkno(xs->bucket),
 			header_size);
 
 	if (xi->value && xi->value_len > OCFS2_XATTR_INLINE_SIZE)
@@ -4614,11 +5106,13 @@ try_again:
 	mlog(0, "xs->not_found = %d, in xattr bucket %llu: free = %d, "
 	     "need = %d, max_free = %d, xh_free_start = %u, xh_name_value_len ="
 	     " %u\n", xs->not_found,
-	     (unsigned long long)xs->bucket.bhs[0]->b_blocknr,
+	     (unsigned long long)bucket_blkno(xs->bucket),
 	     free, need, max_free, le16_to_cpu(xh->xh_free_start),
 	     le16_to_cpu(xh->xh_name_value_len));
 
-	if (free < need || count == ocfs2_xattr_max_xe_in_bucket(inode->i_sb)) {
+	if (free < need ||
+	    (xs->not_found &&
+	     count == ocfs2_xattr_max_xe_in_bucket(inode->i_sb))) {
 		if (need <= max_free &&
 		    count < ocfs2_xattr_max_xe_in_bucket(inode->i_sb)) {
 			/*
@@ -4626,7 +5120,8 @@ try_again:
 			 * name/value will be moved, the xe shouldn't be changed
 			 * in xs.
 			 */
-			ret = ocfs2_defrag_xattr_bucket(inode, &xs->bucket);
+			ret = ocfs2_defrag_xattr_bucket(inode, ctxt->handle,
+							xs->bucket);
 			if (ret) {
 				mlog_errno(ret);
 				goto out;
@@ -4658,7 +5153,7 @@ try_again:
 		 * add a new bucket for the insert.
 		 */
 		ret = ocfs2_check_xattr_bucket_collision(inode,
-							 &xs->bucket,
+							 xs->bucket,
 							 xi->name);
 		if (ret) {
 			mlog_errno(ret);
@@ -4667,17 +5162,21 @@ try_again:
 
 		ret = ocfs2_add_new_xattr_bucket(inode,
 						 xs->xattr_bh,
-						 xs->bucket.bhs[0]);
+						 xs->bucket,
+						 ctxt);
 		if (ret) {
 			mlog_errno(ret);
 			goto out;
 		}
 
-		for (i = 0; i < blk_per_bucket; i++)
-			brelse(xs->bucket.bhs[i]);
-
-		memset(&xs->bucket, 0, sizeof(xs->bucket));
-
+		/*
+		 * ocfs2_add_new_xattr_bucket() will have updated
+		 * xs->bucket if it moved, but it will not have updated
+		 * any of the other search fields.  Thus, we drop it and
+		 * re-search.  Everything should be cached, so it'll be
+		 * quick.
+		 */
+		ocfs2_xattr_bucket_relse(xs->bucket);
 		ret = ocfs2_xattr_index_block_find(inode, xs->xattr_bh,
 						   xi->name_index,
 						   xi->name, xs);
@@ -4689,7 +5188,7 @@ try_again:
 	}
 
 xattr_set:
-	ret = ocfs2_xattr_set_in_bucket(inode, xi, xs);
+	ret = ocfs2_xattr_set_in_bucket(inode, xi, xs, ctxt);
 out:
 	mlog_exit(ret);
 	return ret;
@@ -4700,24 +5199,41 @@ static int ocfs2_delete_xattr_in_bucket(struct inode *inode,
 					void *para)
 {
 	int ret = 0;
-	struct ocfs2_xattr_header *xh = bucket->xh;
+	struct ocfs2_xattr_header *xh = bucket_xh(bucket);
 	u16 i;
 	struct ocfs2_xattr_entry *xe;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	struct ocfs2_xattr_set_ctxt ctxt = {NULL, NULL,};
+	int credits = ocfs2_remove_extent_credits(osb->sb) +
+		ocfs2_blocks_per_xattr_bucket(inode->i_sb);
+
+
+	ocfs2_init_dealloc_ctxt(&ctxt.dealloc);
 
 	for (i = 0; i < le16_to_cpu(xh->xh_count); i++) {
 		xe = &xh->xh_entries[i];
 		if (ocfs2_xattr_is_local(xe))
 			continue;
 
-		ret = ocfs2_xattr_bucket_value_truncate(inode,
-							bucket->bhs[0],
-							i, 0);
+		ctxt.handle = ocfs2_start_trans(osb, credits);
+		if (IS_ERR(ctxt.handle)) {
+			ret = PTR_ERR(ctxt.handle);
+			mlog_errno(ret);
+			break;
+		}
+
+		ret = ocfs2_xattr_bucket_value_truncate(inode, bucket,
+							i, 0, &ctxt);
+
+		ocfs2_commit_trans(osb, ctxt.handle);
 		if (ret) {
 			mlog_errno(ret);
 			break;
 		}
 	}
 
+	ocfs2_schedule_truncate_log_flush(osb, 1);
+	ocfs2_run_deallocs(osb, &ctxt.dealloc);
 	return ret;
 }
 
@@ -4767,6 +5283,74 @@ out:
 	return ret;
 }
 
+/*
+ * 'security' attributes support
+ */
+static size_t ocfs2_xattr_security_list(struct inode *inode, char *list,
+					size_t list_size, const char *name,
+					size_t name_len)
+{
+	const size_t prefix_len = XATTR_SECURITY_PREFIX_LEN;
+	const size_t total_len = prefix_len + name_len + 1;
+
+	if (list && total_len <= list_size) {
+		memcpy(list, XATTR_SECURITY_PREFIX, prefix_len);
+		memcpy(list + prefix_len, name, name_len);
+		list[prefix_len + name_len] = '\0';
+	}
+	return total_len;
+}
+
+static int ocfs2_xattr_security_get(struct inode *inode, const char *name,
+				    void *buffer, size_t size)
+{
+	if (strcmp(name, "") == 0)
+		return -EINVAL;
+	return ocfs2_xattr_get(inode, OCFS2_XATTR_INDEX_SECURITY, name,
+			       buffer, size);
+}
+
+static int ocfs2_xattr_security_set(struct inode *inode, const char *name,
+				    const void *value, size_t size, int flags)
+{
+	if (strcmp(name, "") == 0)
+		return -EINVAL;
+
+	return ocfs2_xattr_set(inode, OCFS2_XATTR_INDEX_SECURITY, name, value,
+			       size, flags);
+}
+
+int ocfs2_init_security_get(struct inode *inode,
+			    struct inode *dir,
+			    struct ocfs2_security_xattr_info *si)
+{
+	/* check whether ocfs2 support feature xattr */
+	if (!ocfs2_supports_xattr(OCFS2_SB(dir->i_sb)))
+		return -EOPNOTSUPP;
+	return security_inode_init_security(inode, dir, &si->name, &si->value,
+					    &si->value_len);
+}
+
+int ocfs2_init_security_set(handle_t *handle,
+			    struct inode *inode,
+			    struct buffer_head *di_bh,
+			    struct ocfs2_security_xattr_info *si,
+			    struct ocfs2_alloc_context *xattr_ac,
+			    struct ocfs2_alloc_context *data_ac)
+{
+	return ocfs2_xattr_set_handle(handle, inode, di_bh,
+				     OCFS2_XATTR_INDEX_SECURITY,
+				     si->name, si->value, si->value_len, 0,
+				     xattr_ac, data_ac);
+}
+
+struct xattr_handler ocfs2_xattr_security_handler = {
+	.prefix	= XATTR_SECURITY_PREFIX,
+	.list	= ocfs2_xattr_security_list,
+	.get	= ocfs2_xattr_security_get,
+	.set	= ocfs2_xattr_security_set,
+};
+
 /*
  * 'trusted' attributes support
  */

fs/ocfs2/xattr.h

@@ -30,13 +30,58 @@ enum ocfs2_xattr_type {
 	OCFS2_XATTR_MAX
 };
 
+struct ocfs2_security_xattr_info {
+	int enable;
+	char *name;
+	void *value;
+	size_t value_len;
+};
+
 extern struct xattr_handler ocfs2_xattr_user_handler;
 extern struct xattr_handler ocfs2_xattr_trusted_handler;
+extern struct xattr_handler ocfs2_xattr_security_handler;
+#ifdef CONFIG_OCFS2_FS_POSIX_ACL
+extern struct xattr_handler ocfs2_xattr_acl_access_handler;
+extern struct xattr_handler ocfs2_xattr_acl_default_handler;
+#endif
 extern struct xattr_handler *ocfs2_xattr_handlers[];
 
 ssize_t ocfs2_listxattr(struct dentry *, char *, size_t);
+int ocfs2_xattr_get_nolock(struct inode *, struct buffer_head *, int,
+			   const char *, void *, size_t);
 int ocfs2_xattr_set(struct inode *, int, const char *, const void *,
 		    size_t, int);
+int ocfs2_xattr_set_handle(handle_t *, struct inode *, struct buffer_head *,
+			   int, const char *, const void *, size_t, int,
+			   struct ocfs2_alloc_context *,
+			   struct ocfs2_alloc_context *);
 int ocfs2_xattr_remove(struct inode *, struct buffer_head *);
+int ocfs2_init_security_get(struct inode *, struct inode *,
+			    struct ocfs2_security_xattr_info *);
+int ocfs2_init_security_set(handle_t *, struct inode *,
+			    struct buffer_head *,
+			    struct ocfs2_security_xattr_info *,
+			    struct ocfs2_alloc_context *,
+			    struct ocfs2_alloc_context *);
+int ocfs2_calc_security_init(struct inode *,
+			     struct ocfs2_security_xattr_info *,
+			     int *, int *, struct ocfs2_alloc_context **);
+int ocfs2_calc_xattr_init(struct inode *, struct buffer_head *,
+			  int, struct ocfs2_security_xattr_info *,
+			  int *, int *, struct ocfs2_alloc_context **);
+
+/*
+ * xattrs can live inside an inode, as part of an external xattr block,
+ * or inside an xattr bucket, which is the leaf of a tree rooted in an
+ * xattr block.  Some of the xattr calls, especially the value setting
+ * functions, want to treat each of these locations as equal.  Let's wrap
+ * them in a structure that we can pass around instead of raw buffer_heads.
+ */
+struct ocfs2_xattr_value_buf {
+	struct buffer_head		*vb_bh;
+	ocfs2_journal_access_func	vb_access;
+	struct ocfs2_xattr_value_root	*vb_xv;
+};
+
 
 #endif /* OCFS2_XATTR_H */

fs/quota.c

@@ -73,7 +73,7 @@ static int generic_quotactl_valid(struct super_block *sb, int type, int cmd, qid
 		case Q_SETQUOTA:
 		case Q_GETQUOTA:
 			/* This is just informative test so we are satisfied without a lock */
-			if (!sb_has_quota_enabled(sb, type))
+			if (!sb_has_quota_active(sb, type))
 				return -ESRCH;
 	}
 
@@ -160,6 +160,9 @@ static void quota_sync_sb(struct super_block *sb, int type)
 	int cnt;
 
 	sb->s_qcop->quota_sync(sb, type);
+
+	if (sb_dqopt(sb)->flags & DQUOT_QUOTA_SYS_FILE)
+		return;
 	/* This is not very clever (and fast) but currently I don't know about
 	 * any other simple way of getting quota data to disk and we must get
 	 * them there for userspace to be visible... */
@@ -175,7 +178,7 @@ static void quota_sync_sb(struct super_block *sb, int type)
 	for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 		if (type != -1 && cnt != type)
 			continue;
-		if (!sb_has_quota_enabled(sb, cnt))
+		if (!sb_has_quota_active(sb, cnt))
 			continue;
 		mutex_lock_nested(&sb_dqopt(sb)->files[cnt]->i_mutex, I_MUTEX_QUOTA);
 		truncate_inode_pages(&sb_dqopt(sb)->files[cnt]->i_data, 0);
@@ -201,7 +204,7 @@ restart:
 		for (cnt = 0; cnt < MAXQUOTAS; cnt++) {
 			if (type != -1 && type != cnt)
 				continue;
-			if (!sb_has_quota_enabled(sb, cnt))
+			if (!sb_has_quota_active(sb, cnt))
 				continue;
 			if (!info_dirty(&sb_dqopt(sb)->info[cnt]) &&
 			    list_empty(&sb_dqopt(sb)->info[cnt].dqi_dirty_list))
@@ -245,7 +248,7 @@ static int do_quotactl(struct super_block *sb, int type, int cmd, qid_t id, void
 			__u32 fmt;
 
 			down_read(&sb_dqopt(sb)->dqptr_sem);
-			if (!sb_has_quota_enabled(sb, type)) {
+			if (!sb_has_quota_active(sb, type)) {
 				up_read(&sb_dqopt(sb)->dqptr_sem);
 				return -ESRCH;
 			}

fs/quota_tree.c

@@ -0,0 +1,645 @@
+/*
+ *	vfsv0 quota IO operations on file
+ */
+
+#include <linux/errno.h>
+#include <linux/fs.h>
+#include <linux/mount.h>
+#include <linux/dqblk_v2.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/quotaops.h>
+
+#include <asm/byteorder.h>
+
+#include "quota_tree.h"
+
+MODULE_AUTHOR("Jan Kara");
+MODULE_DESCRIPTION("Quota trie support");
+MODULE_LICENSE("GPL");
+
+#define __QUOTA_QT_PARANOIA
+
+typedef char *dqbuf_t;
+
+static int get_index(struct qtree_mem_dqinfo *info, qid_t id, int depth)
+{
+	unsigned int epb = info->dqi_usable_bs >> 2;
+
+	depth = info->dqi_qtree_depth - depth - 1;
+	while (depth--)
+		id /= epb;
+	return id % epb;
+}
+
+/* Number of entries in one blocks */
+static inline int qtree_dqstr_in_blk(struct qtree_mem_dqinfo *info)
+{
+	return (info->dqi_usable_bs - sizeof(struct qt_disk_dqdbheader))
+	       / info->dqi_entry_size;
+}
+
+static dqbuf_t getdqbuf(size_t size)
+{
+	dqbuf_t buf = kmalloc(size, GFP_NOFS);
+	if (!buf)
+		printk(KERN_WARNING "VFS: Not enough memory for quota buffers.\n");
+	return buf;
+}
+
+static inline void freedqbuf(dqbuf_t buf)
+{
+	kfree(buf);
+}
+
+static inline ssize_t read_blk(struct qtree_mem_dqinfo *info, uint blk, dqbuf_t buf)
+{
+	struct super_block *sb = info->dqi_sb;
+
+	memset(buf, 0, info->dqi_usable_bs);
+	return sb->s_op->quota_read(sb, info->dqi_type, (char *)buf,
+	       info->dqi_usable_bs, blk << info->dqi_blocksize_bits);
+}
+
+static inline ssize_t write_blk(struct qtree_mem_dqinfo *info, uint blk, dqbuf_t buf)
+{
+	struct super_block *sb = info->dqi_sb;
+
+	return sb->s_op->quota_write(sb, info->dqi_type, (char *)buf,
+	       info->dqi_usable_bs, blk << info->dqi_blocksize_bits);
+}
+
+/* Remove empty block from list and return it */
+static int get_free_dqblk(struct qtree_mem_dqinfo *info)
+{
+	dqbuf_t buf = getdqbuf(info->dqi_usable_bs);
+	struct qt_disk_dqdbheader *dh = (struct qt_disk_dqdbheader *)buf;
+	int ret, blk;
+
+	if (!buf)
+		return -ENOMEM;
+	if (info->dqi_free_blk) {
+		blk = info->dqi_free_blk;
+		ret = read_blk(info, blk, buf);
+		if (ret < 0)
+			goto out_buf;
+		info->dqi_free_blk = le32_to_cpu(dh->dqdh_next_free);
+	}
+	else {
+		memset(buf, 0, info->dqi_usable_bs);
+		/* Assure block allocation... */
+		ret = write_blk(info, info->dqi_blocks, buf);
+		if (ret < 0)
+			goto out_buf;
+		blk = info->dqi_blocks++;
+	}
+	mark_info_dirty(info->dqi_sb, info->dqi_type);
+	ret = blk;
+out_buf:
+	freedqbuf(buf);
+	return ret;
+}
+
+/* Insert empty block to the list */
+static int put_free_dqblk(struct qtree_mem_dqinfo *info, dqbuf_t buf, uint blk)
+{
+	struct qt_disk_dqdbheader *dh = (struct qt_disk_dqdbheader *)buf;
+	int err;
+
+	dh->dqdh_next_free = cpu_to_le32(info->dqi_free_blk);
+	dh->dqdh_prev_free = cpu_to_le32(0);
+	dh->dqdh_entries = cpu_to_le16(0);
+	err = write_blk(info, blk, buf);
+	if (err < 0)
+		return err;
+	info->dqi_free_blk = blk;
+	mark_info_dirty(info->dqi_sb, info->dqi_type);
+	return 0;
+}
+
+/* Remove given block from the list of blocks with free entries */
+static int remove_free_dqentry(struct qtree_mem_dqinfo *info, dqbuf_t buf, uint blk)
+{
+	dqbuf_t tmpbuf = getdqbuf(info->dqi_usable_bs);
+	struct qt_disk_dqdbheader *dh = (struct qt_disk_dqdbheader *)buf;
+	uint nextblk = le32_to_cpu(dh->dqdh_next_free);
+	uint prevblk = le32_to_cpu(dh->dqdh_prev_free);
+	int err;
+
+	if (!tmpbuf)
+		return -ENOMEM;
+	if (nextblk) {
+		err = read_blk(info, nextblk, tmpbuf);
+		if (err < 0)
+			goto out_buf;
+		((struct qt_disk_dqdbheader *)tmpbuf)->dqdh_prev_free =
+							dh->dqdh_prev_free;
+		err = write_blk(info, nextblk, tmpbuf);
+		if (err < 0)
+			goto out_buf;
+	}
+	if (prevblk) {
+		err = read_blk(info, prevblk, tmpbuf);
+		if (err < 0)
+			goto out_buf;
+		((struct qt_disk_dqdbheader *)tmpbuf)->dqdh_next_free =
+							dh->dqdh_next_free;
+		err = write_blk(info, prevblk, tmpbuf);
+		if (err < 0)
+			goto out_buf;
+	} else {
+		info->dqi_free_entry = nextblk;
+		mark_info_dirty(info->dqi_sb, info->dqi_type);
+	}
+	freedqbuf(tmpbuf);
+	dh->dqdh_next_free = dh->dqdh_prev_free = cpu_to_le32(0);
+	/* No matter whether write succeeds block is out of list */
+	if (write_blk(info, blk, buf) < 0)
+		printk(KERN_ERR "VFS: Can't write block (%u) with free entries.\n", blk);
+	return 0;
+out_buf:
+	freedqbuf(tmpbuf);
+	return err;
+}
+
+/* Insert given block to the beginning of list with free entries */
+static int insert_free_dqentry(struct qtree_mem_dqinfo *info, dqbuf_t buf, uint blk)
+{
+	dqbuf_t tmpbuf = getdqbuf(info->dqi_usable_bs);
+	struct qt_disk_dqdbheader *dh = (struct qt_disk_dqdbheader *)buf;
+	int err;
+
+	if (!tmpbuf)
+		return -ENOMEM;
+	dh->dqdh_next_free = cpu_to_le32(info->dqi_free_entry);
+	dh->dqdh_prev_free = cpu_to_le32(0);
+	err = write_blk(info, blk, buf);
+	if (err < 0)
+		goto out_buf;
+	if (info->dqi_free_entry) {
+		err = read_blk(info, info->dqi_free_entry, tmpbuf);
+		if (err < 0)
+			goto out_buf;
+		((struct qt_disk_dqdbheader *)tmpbuf)->dqdh_prev_free =
+							cpu_to_le32(blk);
+		err = write_blk(info, info->dqi_free_entry, tmpbuf);
+		if (err < 0)
+			goto out_buf;
+	}
+	freedqbuf(tmpbuf);
+	info->dqi_free_entry = blk;
+	mark_info_dirty(info->dqi_sb, info->dqi_type);
+	return 0;
+out_buf:
+	freedqbuf(tmpbuf);
+	return err;
+}
+
+/* Is the entry in the block free? */
+int qtree_entry_unused(struct qtree_mem_dqinfo *info, char *disk)
+{
+	int i;
+
+	for (i = 0; i < info->dqi_entry_size; i++)
+		if (disk[i])
+			return 0;
+	return 1;
+}
+EXPORT_SYMBOL(qtree_entry_unused);
+
+/* Find space for dquot */
+static uint find_free_dqentry(struct qtree_mem_dqinfo *info,
+			      struct dquot *dquot, int *err)
+{
+	uint blk, i;
+	struct qt_disk_dqdbheader *dh;
+	dqbuf_t buf = getdqbuf(info->dqi_usable_bs);
+	char *ddquot;
+
+	*err = 0;
+	if (!buf) {
+		*err = -ENOMEM;
+		return 0;
+	}
+	dh = (struct qt_disk_dqdbheader *)buf;
+	if (info->dqi_free_entry) {
+		blk = info->dqi_free_entry;
+		*err = read_blk(info, blk, buf);
+		if (*err < 0)
+			goto out_buf;
+	} else {
+		blk = get_free_dqblk(info);
+		if ((int)blk < 0) {
+			*err = blk;
+			freedqbuf(buf);
+			return 0;
+		}
+		memset(buf, 0, info->dqi_usable_bs);
+		/* This is enough as block is already zeroed and entry list is empty... */
+		info->dqi_free_entry = blk;
+		mark_info_dirty(dquot->dq_sb, dquot->dq_type);
+	}
+	/* Block will be full? */
+	if (le16_to_cpu(dh->dqdh_entries) + 1 >= qtree_dqstr_in_blk(info)) {
+		*err = remove_free_dqentry(info, buf, blk);
+		if (*err < 0) {
+			printk(KERN_ERR "VFS: find_free_dqentry(): Can't "
+			       "remove block (%u) from entry free list.\n",
+			       blk);
+			goto out_buf;
+		}
+	}
+	le16_add_cpu(&dh->dqdh_entries, 1);
+	/* Find free structure in block */
+	for (i = 0, ddquot = ((char *)buf) + sizeof(struct qt_disk_dqdbheader);
+	     i < qtree_dqstr_in_blk(info) && !qtree_entry_unused(info, ddquot);
+	     i++, ddquot += info->dqi_entry_size);
+#ifdef __QUOTA_QT_PARANOIA
+	if (i == qtree_dqstr_in_blk(info)) {
+		printk(KERN_ERR "VFS: find_free_dqentry(): Data block full "
+				"but it shouldn't.\n");
+		*err = -EIO;
+		goto out_buf;
+	}
+#endif
+	*err = write_blk(info, blk, buf);
+	if (*err < 0) {
+		printk(KERN_ERR "VFS: find_free_dqentry(): Can't write quota "
+				"data block %u.\n", blk);
+		goto out_buf;
+	}
+	dquot->dq_off = (blk << info->dqi_blocksize_bits) +
+			sizeof(struct qt_disk_dqdbheader) +
+			i * info->dqi_entry_size;
+	freedqbuf(buf);
+	return blk;
+out_buf:
+	freedqbuf(buf);
+	return 0;
+}
+
+/* Insert reference to structure into the trie */
+static int do_insert_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot,
+			  uint *treeblk, int depth)
+{
+	dqbuf_t buf = getdqbuf(info->dqi_usable_bs);
+	int ret = 0, newson = 0, newact = 0;
+	__le32 *ref;
+	uint newblk;
+
+	if (!buf)
+		return -ENOMEM;
+	if (!*treeblk) {
+		ret = get_free_dqblk(info);
+		if (ret < 0)
+			goto out_buf;
+		*treeblk = ret;
+		memset(buf, 0, info->dqi_usable_bs);
+		newact = 1;
+	} else {
+		ret = read_blk(info, *treeblk, buf);
+		if (ret < 0) {
+			printk(KERN_ERR "VFS: Can't read tree quota block "
+					"%u.\n", *treeblk);
+			goto out_buf;
+		}
+	}
+	ref = (__le32 *)buf;
+	newblk = le32_to_cpu(ref[get_index(info, dquot->dq_id, depth)]);
+	if (!newblk)
+		newson = 1;
+	if (depth == info->dqi_qtree_depth - 1) {
+#ifdef __QUOTA_QT_PARANOIA
+		if (newblk) {
+			printk(KERN_ERR "VFS: Inserting already present quota "
+					"entry (block %u).\n",
+			       le32_to_cpu(ref[get_index(info,
+						dquot->dq_id, depth)]));
+			ret = -EIO;
+			goto out_buf;
+		}
+#endif
+		newblk = find_free_dqentry(info, dquot, &ret);
+	} else {
+		ret = do_insert_tree(info, dquot, &newblk, depth+1);
+	}
+	if (newson && ret >= 0) {
+		ref[get_index(info, dquot->dq_id, depth)] =
+							cpu_to_le32(newblk);
+		ret = write_blk(info, *treeblk, buf);
+	} else if (newact && ret < 0) {
+		put_free_dqblk(info, buf, *treeblk);
+	}
+out_buf:
+	freedqbuf(buf);
+	return ret;
+}
+
+/* Wrapper for inserting quota structure into tree */
+static inline int dq_insert_tree(struct qtree_mem_dqinfo *info,
+				 struct dquot *dquot)
+{
+	int tmp = QT_TREEOFF;
+	return do_insert_tree(info, dquot, &tmp, 0);
+}
+
+/*
+ *	We don't have to be afraid of deadlocks as we never have quotas on quota files...
+ */
+int qtree_write_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
+{
+	int type = dquot->dq_type;
+	struct super_block *sb = dquot->dq_sb;
+	ssize_t ret;
+	dqbuf_t ddquot = getdqbuf(info->dqi_entry_size);
+
+	if (!ddquot)
+		return -ENOMEM;
+
+	/* dq_off is guarded by dqio_mutex */
+	if (!dquot->dq_off) {
+		ret = dq_insert_tree(info, dquot);
+		if (ret < 0) {
+			printk(KERN_ERR "VFS: Error %zd occurred while "
+					"creating quota.\n", ret);
+			freedqbuf(ddquot);
+			return ret;
+		}
+	}
+	spin_lock(&dq_data_lock);
+	info->dqi_ops->mem2disk_dqblk(ddquot, dquot);
+	spin_unlock(&dq_data_lock);
+	ret = sb->s_op->quota_write(sb, type, (char *)ddquot,
+					info->dqi_entry_size, dquot->dq_off);
+	if (ret != info->dqi_entry_size) {
+		printk(KERN_WARNING "VFS: dquota write failed on dev %s\n",
+		       sb->s_id);
+		if (ret >= 0)
+			ret = -ENOSPC;
+	} else {
+		ret = 0;
+	}
+	dqstats.writes++;
+	freedqbuf(ddquot);
+
+	return ret;
+}
+EXPORT_SYMBOL(qtree_write_dquot);
+
+/* Free dquot entry in data block */
+static int free_dqentry(struct qtree_mem_dqinfo *info, struct dquot *dquot,
+			uint blk)
+{
+	struct qt_disk_dqdbheader *dh;
+	dqbuf_t buf = getdqbuf(info->dqi_usable_bs);
+	int ret = 0;
+
+	if (!buf)
+		return -ENOMEM;
+	if (dquot->dq_off >> info->dqi_blocksize_bits != blk) {
+		printk(KERN_ERR "VFS: Quota structure has offset to other "
+		  "block (%u) than it should (%u).\n", blk,
+		  (uint)(dquot->dq_off >> info->dqi_blocksize_bits));
+		goto out_buf;
+	}
+	ret = read_blk(info, blk, buf);
+	if (ret < 0) {
+		printk(KERN_ERR "VFS: Can't read quota data block %u\n", blk);
+		goto out_buf;
+	}
+	dh = (struct qt_disk_dqdbheader *)buf;
+	le16_add_cpu(&dh->dqdh_entries, -1);
+	if (!le16_to_cpu(dh->dqdh_entries)) {	/* Block got free? */
+		ret = remove_free_dqentry(info, buf, blk);
+		if (ret >= 0)
+			ret = put_free_dqblk(info, buf, blk);
+		if (ret < 0) {
+			printk(KERN_ERR "VFS: Can't move quota data block (%u) "
+			  "to free list.\n", blk);
+			goto out_buf;
+		}
+	} else {
+		memset(buf +
+		       (dquot->dq_off & ((1 << info->dqi_blocksize_bits) - 1)),
+		       0, info->dqi_entry_size);
+		if (le16_to_cpu(dh->dqdh_entries) ==
+		    qtree_dqstr_in_blk(info) - 1) {
+			/* Insert will write block itself */
+			ret = insert_free_dqentry(info, buf, blk);
+			if (ret < 0) {
+				printk(KERN_ERR "VFS: Can't insert quota data "
+				       "block (%u) to free entry list.\n", blk);
+				goto out_buf;
+			}
+		} else {
+			ret = write_blk(info, blk, buf);
+			if (ret < 0) {
+				printk(KERN_ERR "VFS: Can't write quota data "
+				  "block %u\n", blk);
+				goto out_buf;
+			}
+		}
+	}
+	dquot->dq_off = 0;	/* Quota is now unattached */
+out_buf:
+	freedqbuf(buf);
+	return ret;
+}
+
+/* Remove reference to dquot from tree */
+static int remove_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot,
+		       uint *blk, int depth)
+{
+	dqbuf_t buf = getdqbuf(info->dqi_usable_bs);
+	int ret = 0;
+	uint newblk;
+	__le32 *ref = (__le32 *)buf;
+
+	if (!buf)
+		return -ENOMEM;
+	ret = read_blk(info, *blk, buf);
+	if (ret < 0) {
+		printk(KERN_ERR "VFS: Can't read quota data block %u\n", *blk);
+		goto out_buf;
+	}
+	newblk = le32_to_cpu(ref[get_index(info, dquot->dq_id, depth)]);
+	if (depth == info->dqi_qtree_depth - 1) {
+		ret = free_dqentry(info, dquot, newblk);
+		newblk = 0;
+	} else {
+		ret = remove_tree(info, dquot, &newblk, depth+1);
+	}
+	if (ret >= 0 && !newblk) {
+		int i;
+		ref[get_index(info, dquot->dq_id, depth)] = cpu_to_le32(0);
+		/* Block got empty? */
+		for (i = 0;
+		     i < (info->dqi_usable_bs >> 2) && !ref[i];
+		     i++);
+		/* Don't put the root block into the free block list */
+		if (i == (info->dqi_usable_bs >> 2)
+		    && *blk != QT_TREEOFF) {
+			put_free_dqblk(info, buf, *blk);
+			*blk = 0;
+		} else {
+			ret = write_blk(info, *blk, buf);
+			if (ret < 0)
+				printk(KERN_ERR "VFS: Can't write quota tree "
+				  "block %u.\n", *blk);
+		}
+	}
+out_buf:
+	freedqbuf(buf);
+	return ret;
+}
+
+/* Delete dquot from tree */
+int qtree_delete_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
+{
+	uint tmp = QT_TREEOFF;
+
+	if (!dquot->dq_off)	/* Even not allocated? */
+		return 0;
+	return remove_tree(info, dquot, &tmp, 0);
+}
+EXPORT_SYMBOL(qtree_delete_dquot);
+
+/* Find entry in block */
+static loff_t find_block_dqentry(struct qtree_mem_dqinfo *info,
+				 struct dquot *dquot, uint blk)
+{
+	dqbuf_t buf = getdqbuf(info->dqi_usable_bs);
+	loff_t ret = 0;
+	int i;
+	char *ddquot;
+
+	if (!buf)
+		return -ENOMEM;
+	ret = read_blk(info, blk, buf);
+	if (ret < 0) {
+		printk(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk);
+		goto out_buf;
+	}
+	for (i = 0, ddquot = ((char *)buf) + sizeof(struct qt_disk_dqdbheader);
+	     i < qtree_dqstr_in_blk(info) && !info->dqi_ops->is_id(ddquot, dquot);
+	     i++, ddquot += info->dqi_entry_size);
+	if (i == qtree_dqstr_in_blk(info)) {
+		printk(KERN_ERR "VFS: Quota for id %u referenced "
+		  "but not present.\n", dquot->dq_id);
+		ret = -EIO;
+		goto out_buf;
+	} else {
+		ret = (blk << info->dqi_blocksize_bits) + sizeof(struct
+		  qt_disk_dqdbheader) + i * info->dqi_entry_size;
+	}
+out_buf:
+	freedqbuf(buf);
+	return ret;
+}
+
+/* Find entry for given id in the tree */
+static loff_t find_tree_dqentry(struct qtree_mem_dqinfo *info,
+				struct dquot *dquot, uint blk, int depth)
+{
+	dqbuf_t buf = getdqbuf(info->dqi_usable_bs);
+	loff_t ret = 0;
+	__le32 *ref = (__le32 *)buf;
+
+	if (!buf)
+		return -ENOMEM;
+	ret = read_blk(info, blk, buf);
+	if (ret < 0) {
+		printk(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk);
+		goto out_buf;
+	}
+	ret = 0;
+	blk = le32_to_cpu(ref[get_index(info, dquot->dq_id, depth)]);
+	if (!blk)	/* No reference? */
+		goto out_buf;
+	if (depth < info->dqi_qtree_depth - 1)
+		ret = find_tree_dqentry(info, dquot, blk, depth+1);
+	else
+		ret = find_block_dqentry(info, dquot, blk);
+out_buf:
+	freedqbuf(buf);
+	return ret;
+}
+
+/* Find entry for given id in the tree - wrapper function */
+static inline loff_t find_dqentry(struct qtree_mem_dqinfo *info,
+				  struct dquot *dquot)
+{
+	return find_tree_dqentry(info, dquot, QT_TREEOFF, 0);
+}
+
+int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
+{
+	int type = dquot->dq_type;
+	struct super_block *sb = dquot->dq_sb;
+	loff_t offset;
+	dqbuf_t ddquot;
+	int ret = 0;
+
+#ifdef __QUOTA_QT_PARANOIA
+	/* Invalidated quota? */
+	if (!sb_dqopt(dquot->dq_sb)->files[type]) {
+		printk(KERN_ERR "VFS: Quota invalidated while reading!\n");
+		return -EIO;
+	}
+#endif
+	/* Do we know offset of the dquot entry in the quota file? */
+	if (!dquot->dq_off) {
+		offset = find_dqentry(info, dquot);
+		if (offset <= 0) {	/* Entry not present? */
+			if (offset < 0)
+				printk(KERN_ERR "VFS: Can't read quota "
+				  "structure for id %u.\n", dquot->dq_id);
+			dquot->dq_off = 0;
+			set_bit(DQ_FAKE_B, &dquot->dq_flags);
+			memset(&dquot->dq_dqb, 0, sizeof(struct mem_dqblk));
+			ret = offset;
+			goto out;
+		}
+		dquot->dq_off = offset;
+	}
+	ddquot = getdqbuf(info->dqi_entry_size);
+	if (!ddquot)
+		return -ENOMEM;
+	ret = sb->s_op->quota_read(sb, type, (char *)ddquot,
+				   info->dqi_entry_size, dquot->dq_off);
+	if (ret != info->dqi_entry_size) {
+		if (ret >= 0)
+			ret = -EIO;
+		printk(KERN_ERR "VFS: Error while reading quota "
+				"structure for id %u.\n", dquot->dq_id);
+		set_bit(DQ_FAKE_B, &dquot->dq_flags);
+		memset(&dquot->dq_dqb, 0, sizeof(struct mem_dqblk));
+		freedqbuf(ddquot);
+		goto out;
+	}
+	spin_lock(&dq_data_lock);
+	info->dqi_ops->disk2mem_dqblk(dquot, ddquot);
+	if (!dquot->dq_dqb.dqb_bhardlimit &&
+	    !dquot->dq_dqb.dqb_bsoftlimit &&
+	    !dquot->dq_dqb.dqb_ihardlimit &&
+	    !dquot->dq_dqb.dqb_isoftlimit)
+		set_bit(DQ_FAKE_B, &dquot->dq_flags);
+	spin_unlock(&dq_data_lock);
+	freedqbuf(ddquot);
+out:
+	dqstats.reads++;
+	return ret;
+}
+EXPORT_SYMBOL(qtree_read_dquot);
+
+/* Check whether dquot should not be deleted. We know we are
+ * the only one operating on dquot (thanks to dq_lock) */
+int qtree_release_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot)
+{
+	if (test_bit(DQ_FAKE_B, &dquot->dq_flags) && !(dquot->dq_dqb.dqb_curinodes | dquot->dq_dqb.dqb_curspace))
+		return qtree_delete_dquot(info, dquot);
+	return 0;
+}
+EXPORT_SYMBOL(qtree_release_dquot);

fs/quota_tree.h

@@ -0,0 +1,25 @@
+/*
+ *	Definitions of structures for vfsv0 quota format
+ */
+
+#ifndef _LINUX_QUOTA_TREE_H
+#define _LINUX_QUOTA_TREE_H
+
+#include <linux/types.h>
+#include <linux/quota.h>
+
+/*
+ *  Structure of header of block with quota structures. It is padded to 16 bytes so
+ *  there will be space for exactly 21 quota-entries in a block
+ */
+struct qt_disk_dqdbheader {
+	__le32 dqdh_next_free;	/* Number of next block with free entry */
+	__le32 dqdh_prev_free;	/* Number of previous block with free entry */
+	__le16 dqdh_entries;	/* Number of valid entries in block */
+	__le16 dqdh_pad1;
+	__le32 dqdh_pad2;
+};
+
+#define QT_TREEOFF	1		/* Offset of tree in file in blocks */
+
+#endif /* _LINUX_QUOTAIO_TREE_H */

fs/quota_v1.c

@@ -3,25 +3,39 @@
 #include <linux/quota.h>
 #include <linux/quotaops.h>
 #include <linux/dqblk_v1.h>
-#include <linux/quotaio_v1.h>
 #include <linux/kernel.h>
 #include <linux/init.h>
 #include <linux/module.h>
 
 #include <asm/byteorder.h>
 
+#include "quotaio_v1.h"
+
 MODULE_AUTHOR("Jan Kara");
 MODULE_DESCRIPTION("Old quota format support");
 MODULE_LICENSE("GPL");
 
+#define QUOTABLOCK_BITS 10
+#define QUOTABLOCK_SIZE (1 << QUOTABLOCK_BITS)
+
+static inline qsize_t v1_stoqb(qsize_t space)
+{
+	return (space + QUOTABLOCK_SIZE - 1) >> QUOTABLOCK_BITS;
+}
+
+static inline qsize_t v1_qbtos(qsize_t blocks)
+{
+	return blocks << QUOTABLOCK_BITS;
+}
+
 static void v1_disk2mem_dqblk(struct mem_dqblk *m, struct v1_disk_dqblk *d)
 {
 	m->dqb_ihardlimit = d->dqb_ihardlimit;
 	m->dqb_isoftlimit = d->dqb_isoftlimit;
 	m->dqb_curinodes = d->dqb_curinodes;
-	m->dqb_bhardlimit = d->dqb_bhardlimit;
-	m->dqb_bsoftlimit = d->dqb_bsoftlimit;
-	m->dqb_curspace = ((qsize_t)d->dqb_curblocks) << QUOTABLOCK_BITS;
+	m->dqb_bhardlimit = v1_qbtos(d->dqb_bhardlimit);
+	m->dqb_bsoftlimit = v1_qbtos(d->dqb_bsoftlimit);
+	m->dqb_curspace = v1_qbtos(d->dqb_curblocks);
 	m->dqb_itime = d->dqb_itime;
 	m->dqb_btime = d->dqb_btime;
 }
@@ -31,9 +45,9 @@ static void v1_mem2disk_dqblk(struct v1_disk_dqblk *d, struct mem_dqblk *m)
 	d->dqb_ihardlimit = m->dqb_ihardlimit;
 	d->dqb_isoftlimit = m->dqb_isoftlimit;
 	d->dqb_curinodes = m->dqb_curinodes;
-	d->dqb_bhardlimit = m->dqb_bhardlimit;
-	d->dqb_bsoftlimit = m->dqb_bsoftlimit;
-	d->dqb_curblocks = toqb(m->dqb_curspace);
+	d->dqb_bhardlimit = v1_stoqb(m->dqb_bhardlimit);
+	d->dqb_bsoftlimit = v1_stoqb(m->dqb_bsoftlimit);
+	d->dqb_curblocks = v1_stoqb(m->dqb_curspace);
 	d->dqb_itime = m->dqb_itime;
 	d->dqb_btime = m->dqb_btime;
 }

fs/quota_v2.c

@@ -6,7 +6,6 @@
 #include <linux/fs.h>
 #include <linux/mount.h>
 #include <linux/dqblk_v2.h>
-#include <linux/quotaio_v2.h>
 #include <linux/kernel.h>
 #include <linux/init.h>
 #include <linux/module.h>
@@ -15,16 +14,37 @@
 
 #include <asm/byteorder.h>
 
+#include "quota_tree.h"
+#include "quotaio_v2.h"
+
 MODULE_AUTHOR("Jan Kara");
 MODULE_DESCRIPTION("Quota format v2 support");
 MODULE_LICENSE("GPL");
 
 #define __QUOTA_V2_PARANOIA
 
-typedef char *dqbuf_t;
+static void v2_mem2diskdqb(void *dp, struct dquot *dquot);
+static void v2_disk2memdqb(struct dquot *dquot, void *dp);
+static int v2_is_id(void *dp, struct dquot *dquot);
+
+static struct qtree_fmt_operations v2_qtree_ops = {
+	.mem2disk_dqblk = v2_mem2diskdqb,
+	.disk2mem_dqblk = v2_disk2memdqb,
+	.is_id = v2_is_id,
+};
+
+#define QUOTABLOCK_BITS 10
+#define QUOTABLOCK_SIZE (1 << QUOTABLOCK_BITS)
 
-#define GETIDINDEX(id, depth) (((id) >> ((V2_DQTREEDEPTH-(depth)-1)*8)) & 0xff)
-#define GETENTRIES(buf) ((struct v2_disk_dqblk *)(((char *)buf)+sizeof(struct v2_disk_dqdbheader)))
+static inline qsize_t v2_stoqb(qsize_t space)
+{
+	return (space + QUOTABLOCK_SIZE - 1) >> QUOTABLOCK_BITS;
+}
+
+static inline qsize_t v2_qbtos(qsize_t blocks)
+{
+	return blocks << QUOTABLOCK_BITS;
+}
 
 /* Check whether given file is really vfsv0 quotafile */
 static int v2_check_quota_file(struct super_block *sb, int type)
@@ -50,7 +70,8 @@ static int v2_check_quota_file(struct super_block *sb, int type)
 static int v2_read_file_info(struct super_block *sb, int type)
 {
 	struct v2_disk_dqinfo dinfo;
-	struct mem_dqinfo *info = sb_dqopt(sb)->info+type;
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct qtree_mem_dqinfo *qinfo;
 	ssize_t size;
 
 	size = sb->s_op->quota_read(sb, type, (char *)&dinfo,
@@ -60,15 +81,29 @@ static int v2_read_file_info(struct super_block *sb, int type)
 			sb->s_id);
 		return -1;
 	}
+	info->dqi_priv = kmalloc(sizeof(struct qtree_mem_dqinfo), GFP_NOFS);
+	if (!info->dqi_priv) {
+		printk(KERN_WARNING
+		       "Not enough memory for quota information structure.\n");
+		return -1;
+	}
+	qinfo = info->dqi_priv;
 	/* limits are stored as unsigned 32-bit data */
 	info->dqi_maxblimit = 0xffffffff;
 	info->dqi_maxilimit = 0xffffffff;
 	info->dqi_bgrace = le32_to_cpu(dinfo.dqi_bgrace);
 	info->dqi_igrace = le32_to_cpu(dinfo.dqi_igrace);
 	info->dqi_flags = le32_to_cpu(dinfo.dqi_flags);
-	info->u.v2_i.dqi_blocks = le32_to_cpu(dinfo.dqi_blocks);
-	info->u.v2_i.dqi_free_blk = le32_to_cpu(dinfo.dqi_free_blk);
-	info->u.v2_i.dqi_free_entry = le32_to_cpu(dinfo.dqi_free_entry);
+	qinfo->dqi_sb = sb;
+	qinfo->dqi_type = type;
+	qinfo->dqi_blocks = le32_to_cpu(dinfo.dqi_blocks);
+	qinfo->dqi_free_blk = le32_to_cpu(dinfo.dqi_free_blk);
+	qinfo->dqi_free_entry = le32_to_cpu(dinfo.dqi_free_entry);
+	qinfo->dqi_blocksize_bits = V2_DQBLKSIZE_BITS;
+	qinfo->dqi_usable_bs = 1 << V2_DQBLKSIZE_BITS;
+	qinfo->dqi_qtree_depth = qtree_depth(qinfo);
+	qinfo->dqi_entry_size = sizeof(struct v2_disk_dqblk);
+	qinfo->dqi_ops = &v2_qtree_ops;
 	return 0;
 }
 
@@ -76,7 +111,8 @@ static int v2_read_file_info(struct super_block *sb, int type)
 static int v2_write_file_info(struct super_block *sb, int type)
 {
 	struct v2_disk_dqinfo dinfo;
-	struct mem_dqinfo *info = sb_dqopt(sb)->info+type;
+	struct mem_dqinfo *info = sb_dqinfo(sb, type);
+	struct qtree_mem_dqinfo *qinfo = info->dqi_priv;
 	ssize_t size;
 
 	spin_lock(&dq_data_lock);
@@ -85,9 +121,9 @@ static int v2_write_file_info(struct super_block *sb, int type)
 	dinfo.dqi_igrace = cpu_to_le32(info->dqi_igrace);
 	dinfo.dqi_flags = cpu_to_le32(info->dqi_flags & DQF_MASK);
 	spin_unlock(&dq_data_lock);
-	dinfo.dqi_blocks = cpu_to_le32(info->u.v2_i.dqi_blocks);
-	dinfo.dqi_free_blk = cpu_to_le32(info->u.v2_i.dqi_free_blk);
-	dinfo.dqi_free_entry = cpu_to_le32(info->u.v2_i.dqi_free_entry);
+	dinfo.dqi_blocks = cpu_to_le32(qinfo->dqi_blocks);
+	dinfo.dqi_free_blk = cpu_to_le32(qinfo->dqi_free_blk);
+	dinfo.dqi_free_entry = cpu_to_le32(qinfo->dqi_free_entry);
 	size = sb->s_op->quota_write(sb, type, (char *)&dinfo,
 	       sizeof(struct v2_disk_dqinfo), V2_DQINFOOFF);
 	if (size != sizeof(struct v2_disk_dqinfo)) {
@@ -98,574 +134,75 @@ static int v2_write_file_info(struct super_block *sb, int type)
 	return 0;
 }
 
-static void disk2memdqb(struct mem_dqblk *m, struct v2_disk_dqblk *d)
+static void v2_disk2memdqb(struct dquot *dquot, void *dp)
 {
+	struct v2_disk_dqblk *d = dp, empty;
+	struct mem_dqblk *m = &dquot->dq_dqb;
+
 	m->dqb_ihardlimit = le32_to_cpu(d->dqb_ihardlimit);
 	m->dqb_isoftlimit = le32_to_cpu(d->dqb_isoftlimit);
 	m->dqb_curinodes = le32_to_cpu(d->dqb_curinodes);
 	m->dqb_itime = le64_to_cpu(d->dqb_itime);
-	m->dqb_bhardlimit = le32_to_cpu(d->dqb_bhardlimit);
-	m->dqb_bsoftlimit = le32_to_cpu(d->dqb_bsoftlimit);
+	m->dqb_bhardlimit = v2_qbtos(le32_to_cpu(d->dqb_bhardlimit));
+	m->dqb_bsoftlimit = v2_qbtos(le32_to_cpu(d->dqb_bsoftlimit));
 	m->dqb_curspace = le64_to_cpu(d->dqb_curspace);
 	m->dqb_btime = le64_to_cpu(d->dqb_btime);
+	/* We need to escape back all-zero structure */
+	memset(&empty, 0, sizeof(struct v2_disk_dqblk));
+	empty.dqb_itime = cpu_to_le64(1);
+	if (!memcmp(&empty, dp, sizeof(struct v2_disk_dqblk)))
+		m->dqb_itime = 0;
 }
 
-static void mem2diskdqb(struct v2_disk_dqblk *d, struct mem_dqblk *m, qid_t id)
+static void v2_mem2diskdqb(void *dp, struct dquot *dquot)
 {
+	struct v2_disk_dqblk *d = dp;
+	struct mem_dqblk *m = &dquot->dq_dqb;
+	struct qtree_mem_dqinfo *info =
+			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
+
 	d->dqb_ihardlimit = cpu_to_le32(m->dqb_ihardlimit);
 	d->dqb_isoftlimit = cpu_to_le32(m->dqb_isoftlimit);
 	d->dqb_curinodes = cpu_to_le32(m->dqb_curinodes);
 	d->dqb_itime = cpu_to_le64(m->dqb_itime);
-	d->dqb_bhardlimit = cpu_to_le32(m->dqb_bhardlimit);
-	d->dqb_bsoftlimit = cpu_to_le32(m->dqb_bsoftlimit);
+	d->dqb_bhardlimit = cpu_to_le32(v2_stoqb(m->dqb_bhardlimit));
+	d->dqb_bsoftlimit = cpu_to_le32(v2_stoqb(m->dqb_bsoftlimit));
 	d->dqb_curspace = cpu_to_le64(m->dqb_curspace);
 	d->dqb_btime = cpu_to_le64(m->dqb_btime);
-	d->dqb_id = cpu_to_le32(id);
-}
-
-static dqbuf_t getdqbuf(void)
-{
-	dqbuf_t buf = kmalloc(V2_DQBLKSIZE, GFP_NOFS);
-	if (!buf)
-		printk(KERN_WARNING "VFS: Not enough memory for quota buffers.\n");
-	return buf;
-}
-
-static inline void freedqbuf(dqbuf_t buf)
-{
-	kfree(buf);
-}
-
-static inline ssize_t read_blk(struct super_block *sb, int type, uint blk, dqbuf_t buf)
-{
-	memset(buf, 0, V2_DQBLKSIZE);
-	return sb->s_op->quota_read(sb, type, (char *)buf,
-	       V2_DQBLKSIZE, blk << V2_DQBLKSIZE_BITS);
-}
-
-static inline ssize_t write_blk(struct super_block *sb, int type, uint blk, dqbuf_t buf)
-{
-	return sb->s_op->quota_write(sb, type, (char *)buf,
-	       V2_DQBLKSIZE, blk << V2_DQBLKSIZE_BITS);
-}
-
-/* Remove empty block from list and return it */
-static int get_free_dqblk(struct super_block *sb, int type)
-{
-	dqbuf_t buf = getdqbuf();
-	struct mem_dqinfo *info = sb_dqinfo(sb, type);
-	struct v2_disk_dqdbheader *dh = (struct v2_disk_dqdbheader *)buf;
-	int ret, blk;
-
-	if (!buf)
-		return -ENOMEM;
-	if (info->u.v2_i.dqi_free_blk) {
-		blk = info->u.v2_i.dqi_free_blk;
-		if ((ret = read_blk(sb, type, blk, buf)) < 0)
-			goto out_buf;
-		info->u.v2_i.dqi_free_blk = le32_to_cpu(dh->dqdh_next_free);
-	}
-	else {
-		memset(buf, 0, V2_DQBLKSIZE);
-		/* Assure block allocation... */
-		if ((ret = write_blk(sb, type, info->u.v2_i.dqi_blocks, buf)) < 0)
-			goto out_buf;
-		blk = info->u.v2_i.dqi_blocks++;
-	}
-	mark_info_dirty(sb, type);
-	ret = blk;
-out_buf:
-	freedqbuf(buf);
-	return ret;
-}
-
-/* Insert empty block to the list */
-static int put_free_dqblk(struct super_block *sb, int type, dqbuf_t buf, uint blk)
-{
-	struct mem_dqinfo *info = sb_dqinfo(sb, type);
-	struct v2_disk_dqdbheader *dh = (struct v2_disk_dqdbheader *)buf;
-	int err;
-
-	dh->dqdh_next_free = cpu_to_le32(info->u.v2_i.dqi_free_blk);
-	dh->dqdh_prev_free = cpu_to_le32(0);
-	dh->dqdh_entries = cpu_to_le16(0);
-	info->u.v2_i.dqi_free_blk = blk;
-	mark_info_dirty(sb, type);
-	/* Some strange block. We had better leave it... */
-	if ((err = write_blk(sb, type, blk, buf)) < 0)
-		return err;
-	return 0;
+	d->dqb_id = cpu_to_le32(dquot->dq_id);
+	if (qtree_entry_unused(info, dp))
+		d->dqb_itime = cpu_to_le64(1);
 }
 
-/* Remove given block from the list of blocks with free entries */
-static int remove_free_dqentry(struct super_block *sb, int type, dqbuf_t buf, uint blk)
+static int v2_is_id(void *dp, struct dquot *dquot)
 {
-	dqbuf_t tmpbuf = getdqbuf();
-	struct mem_dqinfo *info = sb_dqinfo(sb, type);
-	struct v2_disk_dqdbheader *dh = (struct v2_disk_dqdbheader *)buf;
-	uint nextblk = le32_to_cpu(dh->dqdh_next_free), prevblk = le32_to_cpu(dh->dqdh_prev_free);
-	int err;
+	struct v2_disk_dqblk *d = dp;
+	struct qtree_mem_dqinfo *info =
+			sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv;
 
-	if (!tmpbuf)
-		return -ENOMEM;
-	if (nextblk) {
-		if ((err = read_blk(sb, type, nextblk, tmpbuf)) < 0)
-			goto out_buf;
-		((struct v2_disk_dqdbheader *)tmpbuf)->dqdh_prev_free = dh->dqdh_prev_free;
-		if ((err = write_blk(sb, type, nextblk, tmpbuf)) < 0)
-			goto out_buf;
-	}
-	if (prevblk) {
-		if ((err = read_blk(sb, type, prevblk, tmpbuf)) < 0)
-			goto out_buf;
-		((struct v2_disk_dqdbheader *)tmpbuf)->dqdh_next_free = dh->dqdh_next_free;
-		if ((err = write_blk(sb, type, prevblk, tmpbuf)) < 0)
-			goto out_buf;
-	}
-	else {
-		info->u.v2_i.dqi_free_entry = nextblk;
-		mark_info_dirty(sb, type);
-	}
-	freedqbuf(tmpbuf);
-	dh->dqdh_next_free = dh->dqdh_prev_free = cpu_to_le32(0);
-	/* No matter whether write succeeds block is out of list */
-	if (write_blk(sb, type, blk, buf) < 0)
-		printk(KERN_ERR "VFS: Can't write block (%u) with free entries.\n", blk);
-	return 0;
-out_buf:
-	freedqbuf(tmpbuf);
-	return err;
-}
-
-/* Insert given block to the beginning of list with free entries */
-static int insert_free_dqentry(struct super_block *sb, int type, dqbuf_t buf, uint blk)
-{
-	dqbuf_t tmpbuf = getdqbuf();
-	struct mem_dqinfo *info = sb_dqinfo(sb, type);
-	struct v2_disk_dqdbheader *dh = (struct v2_disk_dqdbheader *)buf;
-	int err;
-
-	if (!tmpbuf)
-		return -ENOMEM;
-	dh->dqdh_next_free = cpu_to_le32(info->u.v2_i.dqi_free_entry);
-	dh->dqdh_prev_free = cpu_to_le32(0);
-	if ((err = write_blk(sb, type, blk, buf)) < 0)
-		goto out_buf;
-	if (info->u.v2_i.dqi_free_entry) {
-		if ((err = read_blk(sb, type, info->u.v2_i.dqi_free_entry, tmpbuf)) < 0)
-			goto out_buf;
-		((struct v2_disk_dqdbheader *)tmpbuf)->dqdh_prev_free = cpu_to_le32(blk);
-		if ((err = write_blk(sb, type, info->u.v2_i.dqi_free_entry, tmpbuf)) < 0)
-			goto out_buf;
-	}
-	freedqbuf(tmpbuf);
-	info->u.v2_i.dqi_free_entry = blk;
-	mark_info_dirty(sb, type);
-	return 0;
-out_buf:
-	freedqbuf(tmpbuf);
-	return err;
-}
-
-/* Find space for dquot */
-static uint find_free_dqentry(struct dquot *dquot, int *err)
-{
-	struct super_block *sb = dquot->dq_sb;
-	struct mem_dqinfo *info = sb_dqopt(sb)->info+dquot->dq_type;
-	uint blk, i;
-	struct v2_disk_dqdbheader *dh;
-	struct v2_disk_dqblk *ddquot;
-	struct v2_disk_dqblk fakedquot;
-	dqbuf_t buf;
-
-	*err = 0;
-	if (!(buf = getdqbuf())) {
-		*err = -ENOMEM;
+	if (qtree_entry_unused(info, dp))
 		return 0;
-	}
-	dh = (struct v2_disk_dqdbheader *)buf;
-	ddquot = GETENTRIES(buf);
-	if (info->u.v2_i.dqi_free_entry) {
-		blk = info->u.v2_i.dqi_free_entry;
-		if ((*err = read_blk(sb, dquot->dq_type, blk, buf)) < 0)
-			goto out_buf;
-	}
-	else {
-		blk = get_free_dqblk(sb, dquot->dq_type);
-		if ((int)blk < 0) {
-			*err = blk;
-			freedqbuf(buf);
-			return 0;
-		}
-		memset(buf, 0, V2_DQBLKSIZE);
-		/* This is enough as block is already zeroed and entry list is empty... */
-		info->u.v2_i.dqi_free_entry = blk;
-		mark_info_dirty(sb, dquot->dq_type);
-	}
-	if (le16_to_cpu(dh->dqdh_entries)+1 >= V2_DQSTRINBLK)	/* Block will be full? */
-		if ((*err = remove_free_dqentry(sb, dquot->dq_type, buf, blk)) < 0) {
-			printk(KERN_ERR "VFS: find_free_dqentry(): Can't remove block (%u) from entry free list.\n", blk);
-			goto out_buf;
-		}
-	le16_add_cpu(&dh->dqdh_entries, 1);
-	memset(&fakedquot, 0, sizeof(struct v2_disk_dqblk));
-	/* Find free structure in block */
-	for (i = 0; i < V2_DQSTRINBLK && memcmp(&fakedquot, ddquot+i, sizeof(struct v2_disk_dqblk)); i++);
-#ifdef __QUOTA_V2_PARANOIA
-	if (i == V2_DQSTRINBLK) {
-		printk(KERN_ERR "VFS: find_free_dqentry(): Data block full but it shouldn't.\n");
-		*err = -EIO;
-		goto out_buf;
-	}
-#endif
-	if ((*err = write_blk(sb, dquot->dq_type, blk, buf)) < 0) {
-		printk(KERN_ERR "VFS: find_free_dqentry(): Can't write quota data block %u.\n", blk);
-		goto out_buf;
-	}
-	dquot->dq_off = (blk<<V2_DQBLKSIZE_BITS)+sizeof(struct v2_disk_dqdbheader)+i*sizeof(struct v2_disk_dqblk);
-	freedqbuf(buf);
-	return blk;
-out_buf:
-	freedqbuf(buf);
-	return 0;
-}
-
-/* Insert reference to structure into the trie */
-static int do_insert_tree(struct dquot *dquot, uint *treeblk, int depth)
-{
-	struct super_block *sb = dquot->dq_sb;
-	dqbuf_t buf;
-	int ret = 0, newson = 0, newact = 0;
-	__le32 *ref;
-	uint newblk;
-
-	if (!(buf = getdqbuf()))
-		return -ENOMEM;
-	if (!*treeblk) {
-		ret = get_free_dqblk(sb, dquot->dq_type);
-		if (ret < 0)
-			goto out_buf;
-		*treeblk = ret;
-		memset(buf, 0, V2_DQBLKSIZE);
-		newact = 1;
-	}
-	else {
-		if ((ret = read_blk(sb, dquot->dq_type, *treeblk, buf)) < 0) {
-			printk(KERN_ERR "VFS: Can't read tree quota block %u.\n", *treeblk);
-			goto out_buf;
-		}
-	}
-	ref = (__le32 *)buf;
-	newblk = le32_to_cpu(ref[GETIDINDEX(dquot->dq_id, depth)]);
-	if (!newblk)
-		newson = 1;
-	if (depth == V2_DQTREEDEPTH-1) {
-#ifdef __QUOTA_V2_PARANOIA
-		if (newblk) {
-			printk(KERN_ERR "VFS: Inserting already present quota entry (block %u).\n", le32_to_cpu(ref[GETIDINDEX(dquot->dq_id, depth)]));
-			ret = -EIO;
-			goto out_buf;
-		}
-#endif
-		newblk = find_free_dqentry(dquot, &ret);
-	}
-	else
-		ret = do_insert_tree(dquot, &newblk, depth+1);
-	if (newson && ret >= 0) {
-		ref[GETIDINDEX(dquot->dq_id, depth)] = cpu_to_le32(newblk);
-		ret = write_blk(sb, dquot->dq_type, *treeblk, buf);
-	}
-	else if (newact && ret < 0)
-		put_free_dqblk(sb, dquot->dq_type, buf, *treeblk);
-out_buf:
-	freedqbuf(buf);
-	return ret;
+	return le32_to_cpu(d->dqb_id) == dquot->dq_id;
 }
 
-/* Wrapper for inserting quota structure into tree */
-static inline int dq_insert_tree(struct dquot *dquot)
+static int v2_read_dquot(struct dquot *dquot)
 {
-	int tmp = V2_DQTREEOFF;
-	return do_insert_tree(dquot, &tmp, 0);
+	return qtree_read_dquot(sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv, dquot);
 }
 
-/*
- *	We don't have to be afraid of deadlocks as we never have quotas on quota files...
- */
 static int v2_write_dquot(struct dquot *dquot)
 {
-	int type = dquot->dq_type;
-	ssize_t ret;
-	struct v2_disk_dqblk ddquot, empty;
-
-	/* dq_off is guarded by dqio_mutex */
-	if (!dquot->dq_off)
-		if ((ret = dq_insert_tree(dquot)) < 0) {
-			printk(KERN_ERR "VFS: Error %zd occurred while creating quota.\n", ret);
-			return ret;
-		}
-	spin_lock(&dq_data_lock);
-	mem2diskdqb(&ddquot, &dquot->dq_dqb, dquot->dq_id);
-	/* Argh... We may need to write structure full of zeroes but that would be
-	 * treated as an empty place by the rest of the code. Format change would
-	 * be definitely cleaner but the problems probably are not worth it */
-	memset(&empty, 0, sizeof(struct v2_disk_dqblk));
-	if (!memcmp(&empty, &ddquot, sizeof(struct v2_disk_dqblk)))
-		ddquot.dqb_itime = cpu_to_le64(1);
-	spin_unlock(&dq_data_lock);
-	ret = dquot->dq_sb->s_op->quota_write(dquot->dq_sb, type,
-	      (char *)&ddquot, sizeof(struct v2_disk_dqblk), dquot->dq_off);
-	if (ret != sizeof(struct v2_disk_dqblk)) {
-		printk(KERN_WARNING "VFS: dquota write failed on dev %s\n", dquot->dq_sb->s_id);
-		if (ret >= 0)
-			ret = -ENOSPC;
-	}
-	else
-		ret = 0;
-	dqstats.writes++;
-
-	return ret;
+	return qtree_write_dquot(sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv, dquot);
 }
 
-/* Free dquot entry in data block */
-static int free_dqentry(struct dquot *dquot, uint blk)
-{
-	struct super_block *sb = dquot->dq_sb;
-	int type = dquot->dq_type;
-	struct v2_disk_dqdbheader *dh;
-	dqbuf_t buf = getdqbuf();
-	int ret = 0;
-
-	if (!buf)
-		return -ENOMEM;
-	if (dquot->dq_off >> V2_DQBLKSIZE_BITS != blk) {
-		printk(KERN_ERR "VFS: Quota structure has offset to other "
-		  "block (%u) than it should (%u).\n", blk,
-		  (uint)(dquot->dq_off >> V2_DQBLKSIZE_BITS));
-		goto out_buf;
-	}
-	if ((ret = read_blk(sb, type, blk, buf)) < 0) {
-		printk(KERN_ERR "VFS: Can't read quota data block %u\n", blk);
-		goto out_buf;
-	}
-	dh = (struct v2_disk_dqdbheader *)buf;
-	le16_add_cpu(&dh->dqdh_entries, -1);
-	if (!le16_to_cpu(dh->dqdh_entries)) {	/* Block got free? */
-		if ((ret = remove_free_dqentry(sb, type, buf, blk)) < 0 ||
-		    (ret = put_free_dqblk(sb, type, buf, blk)) < 0) {
-			printk(KERN_ERR "VFS: Can't move quota data block (%u) "
-			  "to free list.\n", blk);
-			goto out_buf;
-		}
-	}
-	else {
-		memset(buf+(dquot->dq_off & ((1 << V2_DQBLKSIZE_BITS)-1)), 0,
-		  sizeof(struct v2_disk_dqblk));
-		if (le16_to_cpu(dh->dqdh_entries) == V2_DQSTRINBLK-1) {
-			/* Insert will write block itself */
-			if ((ret = insert_free_dqentry(sb, type, buf, blk)) < 0) {
-				printk(KERN_ERR "VFS: Can't insert quota data block (%u) to free entry list.\n", blk);
-				goto out_buf;
-			}
-		}
-		else
-			if ((ret = write_blk(sb, type, blk, buf)) < 0) {
-				printk(KERN_ERR "VFS: Can't write quota data "
-				  "block %u\n", blk);
-				goto out_buf;
-			}
-	}
-	dquot->dq_off = 0;	/* Quota is now unattached */
-out_buf:
-	freedqbuf(buf);
-	return ret;
-}
-
-/* Remove reference to dquot from tree */
-static int remove_tree(struct dquot *dquot, uint *blk, int depth)
-{
-	struct super_block *sb = dquot->dq_sb;
-	int type = dquot->dq_type;
-	dqbuf_t buf = getdqbuf();
-	int ret = 0;
-	uint newblk;
-	__le32 *ref = (__le32 *)buf;
-	
-	if (!buf)
-		return -ENOMEM;
-	if ((ret = read_blk(sb, type, *blk, buf)) < 0) {
-		printk(KERN_ERR "VFS: Can't read quota data block %u\n", *blk);
-		goto out_buf;
-	}
-	newblk = le32_to_cpu(ref[GETIDINDEX(dquot->dq_id, depth)]);
-	if (depth == V2_DQTREEDEPTH-1) {
-		ret = free_dqentry(dquot, newblk);
-		newblk = 0;
-	}
-	else
-		ret = remove_tree(dquot, &newblk, depth+1);
-	if (ret >= 0 && !newblk) {
-		int i;
-		ref[GETIDINDEX(dquot->dq_id, depth)] = cpu_to_le32(0);
-		for (i = 0; i < V2_DQBLKSIZE && !buf[i]; i++);	/* Block got empty? */
-		/* Don't put the root block into the free block list */
-		if (i == V2_DQBLKSIZE && *blk != V2_DQTREEOFF) {
-			put_free_dqblk(sb, type, buf, *blk);
-			*blk = 0;
-		}
-		else
-			if ((ret = write_blk(sb, type, *blk, buf)) < 0)
-				printk(KERN_ERR "VFS: Can't write quota tree "
-				  "block %u.\n", *blk);
-	}
-out_buf:
-	freedqbuf(buf);
-	return ret;	
-}
-
-/* Delete dquot from tree */
-static int v2_delete_dquot(struct dquot *dquot)
-{
-	uint tmp = V2_DQTREEOFF;
-
-	if (!dquot->dq_off)	/* Even not allocated? */
-		return 0;
-	return remove_tree(dquot, &tmp, 0);
-}
-
-/* Find entry in block */
-static loff_t find_block_dqentry(struct dquot *dquot, uint blk)
-{
-	dqbuf_t buf = getdqbuf();
-	loff_t ret = 0;
-	int i;
-	struct v2_disk_dqblk *ddquot = GETENTRIES(buf);
-
-	if (!buf)
-		return -ENOMEM;
-	if ((ret = read_blk(dquot->dq_sb, dquot->dq_type, blk, buf)) < 0) {
-		printk(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk);
-		goto out_buf;
-	}
-	if (dquot->dq_id)
-		for (i = 0; i < V2_DQSTRINBLK &&
-		     le32_to_cpu(ddquot[i].dqb_id) != dquot->dq_id; i++);
-	else {	/* ID 0 as a bit more complicated searching... */
-		struct v2_disk_dqblk fakedquot;
-
-		memset(&fakedquot, 0, sizeof(struct v2_disk_dqblk));
-		for (i = 0; i < V2_DQSTRINBLK; i++)
-			if (!le32_to_cpu(ddquot[i].dqb_id) &&
-			    memcmp(&fakedquot, ddquot+i, sizeof(struct v2_disk_dqblk)))
-				break;
-	}
-	if (i == V2_DQSTRINBLK) {
-		printk(KERN_ERR "VFS: Quota for id %u referenced "
-		  "but not present.\n", dquot->dq_id);
-		ret = -EIO;
-		goto out_buf;
-	}
-	else
-		ret = (blk << V2_DQBLKSIZE_BITS) + sizeof(struct
-		  v2_disk_dqdbheader) + i * sizeof(struct v2_disk_dqblk);
-out_buf:
-	freedqbuf(buf);
-	return ret;
-}
-
-/* Find entry for given id in the tree */
-static loff_t find_tree_dqentry(struct dquot *dquot, uint blk, int depth)
-{
-	dqbuf_t buf = getdqbuf();
-	loff_t ret = 0;
-	__le32 *ref = (__le32 *)buf;
-
-	if (!buf)
-		return -ENOMEM;
-	if ((ret = read_blk(dquot->dq_sb, dquot->dq_type, blk, buf)) < 0) {
-		printk(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk);
-		goto out_buf;
-	}
-	ret = 0;
-	blk = le32_to_cpu(ref[GETIDINDEX(dquot->dq_id, depth)]);
-	if (!blk)	/* No reference? */
-		goto out_buf;
-	if (depth < V2_DQTREEDEPTH-1)
-		ret = find_tree_dqentry(dquot, blk, depth+1);
-	else
-		ret = find_block_dqentry(dquot, blk);
-out_buf:
-	freedqbuf(buf);
-	return ret;
-}
-
-/* Find entry for given id in the tree - wrapper function */
-static inline loff_t find_dqentry(struct dquot *dquot)
-{
-	return find_tree_dqentry(dquot, V2_DQTREEOFF, 0);
-}
-
-static int v2_read_dquot(struct dquot *dquot)
+static int v2_release_dquot(struct dquot *dquot)
 {
-	int type = dquot->dq_type;
-	loff_t offset;
-	struct v2_disk_dqblk ddquot, empty;
-	int ret = 0;
-
-#ifdef __QUOTA_V2_PARANOIA
-	/* Invalidated quota? */
-	if (!dquot->dq_sb || !sb_dqopt(dquot->dq_sb)->files[type]) {
-		printk(KERN_ERR "VFS: Quota invalidated while reading!\n");
-		return -EIO;
-	}
-#endif
-	offset = find_dqentry(dquot);
-	if (offset <= 0) {	/* Entry not present? */
-		if (offset < 0)
-			printk(KERN_ERR "VFS: Can't read quota "
-			  "structure for id %u.\n", dquot->dq_id);
-		dquot->dq_off = 0;
-		set_bit(DQ_FAKE_B, &dquot->dq_flags);
-		memset(&dquot->dq_dqb, 0, sizeof(struct mem_dqblk));
-		ret = offset;
-	}
-	else {
-		dquot->dq_off = offset;
-		if ((ret = dquot->dq_sb->s_op->quota_read(dquot->dq_sb, type,
-		    (char *)&ddquot, sizeof(struct v2_disk_dqblk), offset))
-		    != sizeof(struct v2_disk_dqblk)) {
-			if (ret >= 0)
-				ret = -EIO;
-			printk(KERN_ERR "VFS: Error while reading quota "
-			  "structure for id %u.\n", dquot->dq_id);
-			memset(&ddquot, 0, sizeof(struct v2_disk_dqblk));
-		}
-		else {
-			ret = 0;
-			/* We need to escape back all-zero structure */
-			memset(&empty, 0, sizeof(struct v2_disk_dqblk));
-			empty.dqb_itime = cpu_to_le64(1);
-			if (!memcmp(&empty, &ddquot, sizeof(struct v2_disk_dqblk)))
-				ddquot.dqb_itime = 0;
-		}
-		disk2memdqb(&dquot->dq_dqb, &ddquot);
-		if (!dquot->dq_dqb.dqb_bhardlimit &&
-			!dquot->dq_dqb.dqb_bsoftlimit &&
-			!dquot->dq_dqb.dqb_ihardlimit &&
-			!dquot->dq_dqb.dqb_isoftlimit)
-			set_bit(DQ_FAKE_B, &dquot->dq_flags);
-	}
-	dqstats.reads++;
-
-	return ret;
+	return qtree_release_dquot(sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv, dquot);
 }
 
-/* Check whether dquot should not be deleted. We know we are
- * the only one operating on dquot (thanks to dq_lock) */
-static int v2_release_dquot(struct dquot *dquot)
+static int v2_free_file_info(struct super_block *sb, int type)
 {
-	if (test_bit(DQ_FAKE_B, &dquot->dq_flags) && !(dquot->dq_dqb.dqb_curinodes | dquot->dq_dqb.dqb_curspace))
-		return v2_delete_dquot(dquot);
+	kfree(sb_dqinfo(sb, type)->dqi_priv);
 	return 0;
 }
 
@@ -673,7 +210,7 @@ static struct quota_format_ops v2_format_ops = {
 	.check_quota_file	= v2_check_quota_file,
 	.read_file_info		= v2_read_file_info,
 	.write_file_info	= v2_write_file_info,
-	.free_file_info		= NULL,
+	.free_file_info		= v2_free_file_info,
 	.read_dqblk		= v2_read_dquot,
 	.commit_dqblk		= v2_write_dquot,
 	.release_dqblk		= v2_release_dquot,

include/linux/quotaio_v2.h → fs/quotaio_v2.h

@@ -21,6 +21,12 @@
 	0		/* GRPQUOTA */\
 }
 
+/* First generic header */
+struct v2_disk_dqheader {
+	__le32 dqh_magic;	/* Magic number identifying file */
+	__le32 dqh_version;	/* File version */
+};
+
 /*
  * The following structure defines the format of the disk quota file
  * (as it appears on disk) - the file is a radix tree whose leaves point
@@ -38,15 +44,6 @@ struct v2_disk_dqblk {
 	__le64 dqb_itime;	/* time limit for excessive inode use */
 };
 
-/*
- * Here are header structures as written on disk and their in-memory copies
- */
-/* First generic header */
-struct v2_disk_dqheader {
-	__le32 dqh_magic;	/* Magic number identifying file */
-	__le32 dqh_version;	/* File version */
-};
-
 /* Header with type and version specific information */
 struct v2_disk_dqinfo {
 	__le32 dqi_bgrace;	/* Time before block soft limit becomes hard limit */
@@ -57,23 +54,7 @@ struct v2_disk_dqinfo {
 	__le32 dqi_free_entry;	/* Number of block with at least one free entry */
 };
 
-/*
- *  Structure of header of block with quota structures. It is padded to 16 bytes so
- *  there will be space for exactly 21 quota-entries in a block
- */
-struct v2_disk_dqdbheader {
-	__le32 dqdh_next_free;	/* Number of next block with free entry */
-	__le32 dqdh_prev_free;	/* Number of previous block with free entry */
-	__le16 dqdh_entries;	/* Number of valid entries in block */
-	__le16 dqdh_pad1;
-	__le32 dqdh_pad2;
-};
-
 #define V2_DQINFOOFF	sizeof(struct v2_disk_dqheader)	/* Offset of info header in file */
-#define V2_DQBLKSIZE_BITS	10
-#define V2_DQBLKSIZE	(1 << V2_DQBLKSIZE_BITS)	/* Size of block with quota structures */
-#define V2_DQTREEOFF	1		/* Offset of tree in file in blocks */
-#define V2_DQTREEDEPTH	4		/* Depth of quota tree */
-#define V2_DQSTRINBLK	((V2_DQBLKSIZE - sizeof(struct v2_disk_dqdbheader)) / sizeof(struct v2_disk_dqblk))	/* Number of entries in one blocks */
+#define V2_DQBLKSIZE_BITS 10				/* Size of leaf block in tree */
 
 #endif /* _LINUX_QUOTAIO_V2_H */

fs/reiserfs/super.c

@@ -649,6 +649,8 @@ static struct dquot_operations reiserfs_quota_operations = {
 	.release_dquot = reiserfs_release_dquot,
 	.mark_dirty = reiserfs_mark_dquot_dirty,
 	.write_info = reiserfs_write_info,
+	.alloc_dquot	= dquot_alloc,
+	.destroy_dquot	= dquot_destroy,
 };
 
 static struct quotactl_ops reiserfs_qctl_operations = {
@@ -994,8 +996,7 @@ static int reiserfs_parse_options(struct super_block *s, char *options,	/* strin
 		if (c == 'u' || c == 'g') {
 			int qtype = c == 'u' ? USRQUOTA : GRPQUOTA;
 
-			if ((sb_any_quota_enabled(s) ||
-			     sb_any_quota_suspended(s)) &&
+			if (sb_any_quota_loaded(s) &&
 			    (!*arg != !REISERFS_SB(s)->s_qf_names[qtype])) {
 				reiserfs_warning(s,
 						 "reiserfs_parse_options: cannot change journaled quota options when quota turned on.");
@@ -1041,8 +1042,7 @@ static int reiserfs_parse_options(struct super_block *s, char *options,	/* strin
 						 "reiserfs_parse_options: unknown quota format specified.");
 				return 0;
 			}
-			if ((sb_any_quota_enabled(s) ||
-			     sb_any_quota_suspended(s)) &&
+			if (sb_any_quota_loaded(s) &&
 			    *qfmt != REISERFS_SB(s)->s_jquota_fmt) {
 				reiserfs_warning(s,
 						 "reiserfs_parse_options: cannot change journaled quota options when quota turned on.");
@@ -1067,7 +1067,7 @@ static int reiserfs_parse_options(struct super_block *s, char *options,	/* strin
 	}
 	/* This checking is not precise wrt the quota type but for our purposes it is sufficient */
 	if (!(*mount_options & (1 << REISERFS_QUOTA))
-	    && sb_any_quota_enabled(s)) {
+	    && sb_any_quota_loaded(s)) {
 		reiserfs_warning(s,
 				 "reiserfs_parse_options: quota options must be present when quota is turned on.");
 		return 0;

include/linux/Kbuild

@@ -56,8 +56,6 @@ header-y += dlm_device.h
 header-y += dlm_netlink.h
 header-y += dm-ioctl.h
 header-y += dn.h
-header-y += dqblk_v1.h
-header-y += dqblk_v2.h
 header-y += dqblk_xfs.h
 header-y += efs_fs_sb.h
 header-y += elf-fdpic.h
@@ -134,8 +132,6 @@ header-y += posix_types.h
 header-y += ppdev.h
 header-y += prctl.h
 header-y += qnxtypes.h
-header-y += quotaio_v1.h
-header-y += quotaio_v2.h
 header-y += radeonfb.h
 header-y += raw.h
 header-y += resource.h

include/linux/dqblk_qtree.h

@@ -0,0 +1,56 @@
+/*
+ *	Definitions of structures and functions for quota formats using trie
+ */
+
+#ifndef _LINUX_DQBLK_QTREE_H
+#define _LINUX_DQBLK_QTREE_H
+
+#include <linux/types.h>
+
+/* Numbers of blocks needed for updates - we count with the smallest
+ * possible block size (1024) */
+#define QTREE_INIT_ALLOC 4
+#define QTREE_INIT_REWRITE 2
+#define QTREE_DEL_ALLOC 0
+#define QTREE_DEL_REWRITE 6
+
+struct dquot;
+
+/* Operations */
+struct qtree_fmt_operations {
+	void (*mem2disk_dqblk)(void *disk, struct dquot *dquot);	/* Convert given entry from in memory format to disk one */
+	void (*disk2mem_dqblk)(struct dquot *dquot, void *disk);	/* Convert given entry from disk format to in memory one */
+	int (*is_id)(void *disk, struct dquot *dquot);	/* Is this structure for given id? */
+};
+
+/* Inmemory copy of version specific information */
+struct qtree_mem_dqinfo {
+	struct super_block *dqi_sb;	/* Sb quota is on */
+	int dqi_type;			/* Quota type */
+	unsigned int dqi_blocks;	/* # of blocks in quota file */
+	unsigned int dqi_free_blk;	/* First block in list of free blocks */
+	unsigned int dqi_free_entry;	/* First block with free entry */
+	unsigned int dqi_blocksize_bits;	/* Block size of quota file */
+	unsigned int dqi_entry_size;	/* Size of quota entry in quota file */
+	unsigned int dqi_usable_bs;	/* Space usable in block for quota data */
+	unsigned int dqi_qtree_depth;	/* Precomputed depth of quota tree */
+	struct qtree_fmt_operations *dqi_ops;	/* Operations for entry manipulation */
+};
+
+int qtree_write_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot);
+int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot);
+int qtree_delete_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot);
+int qtree_release_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot);
+int qtree_entry_unused(struct qtree_mem_dqinfo *info, char *disk);
+static inline int qtree_depth(struct qtree_mem_dqinfo *info)
+{
+	unsigned int epb = info->dqi_usable_bs >> 2;
+	unsigned long long entries = epb;
+	int i;
+
+	for (i = 1; entries < (1ULL << 32); i++)
+		entries *= epb;
+	return i;
+}
+
+#endif /* _LINUX_DQBLK_QTREE_H */

include/linux/dqblk_v1.h

@@ -5,9 +5,6 @@
 #ifndef _LINUX_DQBLK_V1_H
 #define _LINUX_DQBLK_V1_H
 
-/* Id of quota format */
-#define QFMT_VFS_OLD 1
-
 /* Root squash turned on */
 #define V1_DQF_RSQUASH 1
 
@@ -17,8 +14,4 @@
 #define V1_DEL_ALLOC 0
 #define V1_DEL_REWRITE 2
 
-/* Special information about quotafile */
-struct v1_mem_dqinfo {
-};
-
 #endif	/* _LINUX_DQBLK_V1_H */

include/linux/dqblk_v2.h

@@ -1,26 +1,16 @@
 /*
- *	Definitions of structures for vfsv0 quota format
+ *  Definitions for vfsv0 quota format
  */
 
 #ifndef _LINUX_DQBLK_V2_H
 #define _LINUX_DQBLK_V2_H
 
-#include <linux/types.h>
-
-/* id numbers of quota format */
-#define QFMT_VFS_V0 2
+#include <linux/dqblk_qtree.h>
 
 /* Numbers of blocks needed for updates */
-#define V2_INIT_ALLOC 4
-#define V2_INIT_REWRITE 2
-#define V2_DEL_ALLOC 0
-#define V2_DEL_REWRITE 6
-
-/* Inmemory copy of version specific information */
-struct v2_mem_dqinfo {
-	unsigned int dqi_blocks;
-	unsigned int dqi_free_blk;
-	unsigned int dqi_free_entry;
-};
+#define V2_INIT_ALLOC QTREE_INIT_ALLOC
+#define V2_INIT_REWRITE QTREE_INIT_REWRITE
+#define V2_DEL_ALLOC QTREE_DEL_ALLOC
+#define V2_DEL_REWRITE QTREE_DEL_REWRITE
 
 #endif /* _LINUX_DQBLK_V2_H */

include/linux/jbd2.h

@@ -329,6 +329,7 @@ enum jbd_state_bits {
 	BH_State,		/* Pins most journal_head state */
 	BH_JournalHead,		/* Pins bh->b_private and jh->b_bh */
 	BH_Unshadow,		/* Dummy bit, for BJ_Shadow wakeup filtering */
+	BH_JBDPrivateStart,	/* First bit available for private use by FS */
 };
 
 BUFFER_FNS(JBD, jbd)
@@ -1007,6 +1008,35 @@ int __jbd2_journal_clean_checkpoint_list(journal_t *journal);
 int __jbd2_journal_remove_checkpoint(struct journal_head *);
 void __jbd2_journal_insert_checkpoint(struct journal_head *, transaction_t *);
 
+
+/*
+ * Triggers
+ */
+
+struct jbd2_buffer_trigger_type {
+	/*
+	 * Fired just before a buffer is written to the journal.
+	 * mapped_data is a mapped buffer that is the frozen data for
+	 * commit.
+	 */
+	void (*t_commit)(struct jbd2_buffer_trigger_type *type,
+			 struct buffer_head *bh, void *mapped_data,
+			 size_t size);
+
+	/*
+	 * Fired during journal abort for dirty buffers that will not be
+	 * committed.
+	 */
+	void (*t_abort)(struct jbd2_buffer_trigger_type *type,
+			struct buffer_head *bh);
+};
+
+extern void jbd2_buffer_commit_trigger(struct journal_head *jh,
+				       void *mapped_data,
+				       struct jbd2_buffer_trigger_type *triggers);
+extern void jbd2_buffer_abort_trigger(struct journal_head *jh,
+				      struct jbd2_buffer_trigger_type *triggers);
+
 /* Buffer IO */
 extern int
 jbd2_journal_write_metadata_buffer(transaction_t	  *transaction,
@@ -1045,6 +1075,8 @@ extern int	 jbd2_journal_extend (handle_t *, int nblocks);
 extern int	 jbd2_journal_get_write_access(handle_t *, struct buffer_head *);
 extern int	 jbd2_journal_get_create_access (handle_t *, struct buffer_head *);
 extern int	 jbd2_journal_get_undo_access(handle_t *, struct buffer_head *);
+void		 jbd2_journal_set_triggers(struct buffer_head *,
+					   struct jbd2_buffer_trigger_type *type);
 extern int	 jbd2_journal_dirty_metadata (handle_t *, struct buffer_head *);
 extern void	 jbd2_journal_release_buffer (handle_t *, struct buffer_head *);
 extern int	 jbd2_journal_forget (handle_t *, struct buffer_head *);

include/linux/journal-head.h

@@ -12,6 +12,8 @@
 
 typedef unsigned int		tid_t;		/* Unique transaction ID */
 typedef struct transaction_s	transaction_t;	/* Compound transaction type */
+
+
 struct buffer_head;
 
 struct journal_head {
@@ -87,6 +89,12 @@ struct journal_head {
 	 * [j_list_lock]
 	 */
 	struct journal_head *b_cpnext, *b_cpprev;
+
+	/* Trigger type */
+	struct jbd2_buffer_trigger_type *b_triggers;
+
+	/* Trigger type for the committing transaction's frozen data */
+	struct jbd2_buffer_trigger_type *b_frozen_triggers;
 };
 
 #endif		/* JOURNAL_HEAD_H_INCLUDED */

include/linux/quota.h

@@ -36,17 +36,7 @@
 #include <linux/errno.h>
 #include <linux/types.h>
 
-#define __DQUOT_VERSION__	"dquot_6.5.1"
-#define __DQUOT_NUM_VERSION__	6*10000+5*100+1
-
-/* Size of blocks in which are counted size limits */
-#define QUOTABLOCK_BITS 10
-#define QUOTABLOCK_SIZE (1 << QUOTABLOCK_BITS)
-
-/* Conversion routines from and to quota blocks */
-#define qb2kb(x) ((x) << (QUOTABLOCK_BITS-10))
-#define kb2qb(x) ((x) >> (QUOTABLOCK_BITS-10))
-#define toqb(x) (((x) + QUOTABLOCK_SIZE - 1) >> QUOTABLOCK_BITS)
+#define __DQUOT_VERSION__	"dquot_6.5.2"
 
 #define MAXQUOTAS 2
 #define USRQUOTA  0		/* element used for user quotas */
@@ -80,16 +70,34 @@
 #define Q_GETQUOTA 0x800007	/* get user quota structure */
 #define Q_SETQUOTA 0x800008	/* set user quota structure */
 
+/* Quota format type IDs */
+#define	QFMT_VFS_OLD 1
+#define	QFMT_VFS_V0 2
+
+/* Size of block in which space limits are passed through the quota
+ * interface */
+#define QIF_DQBLKSIZE_BITS 10
+#define QIF_DQBLKSIZE (1 << QIF_DQBLKSIZE_BITS)
+
 /*
  * Quota structure used for communication with userspace via quotactl
  * Following flags are used to specify which fields are valid
  */
-#define QIF_BLIMITS	1
-#define QIF_SPACE	2
-#define QIF_ILIMITS	4
-#define QIF_INODES	8
-#define QIF_BTIME	16
-#define QIF_ITIME	32
+enum {
+	QIF_BLIMITS_B = 0,
+	QIF_SPACE_B,
+	QIF_ILIMITS_B,
+	QIF_INODES_B,
+	QIF_BTIME_B,
+	QIF_ITIME_B,
+};
+
+#define QIF_BLIMITS	(1 << QIF_BLIMITS_B)
+#define QIF_SPACE	(1 << QIF_SPACE_B)
+#define QIF_ILIMITS	(1 << QIF_ILIMITS_B)
+#define QIF_INODES	(1 << QIF_INODES_B)
+#define QIF_BTIME	(1 << QIF_BTIME_B)
+#define QIF_ITIME	(1 << QIF_ITIME_B)
 #define QIF_LIMITS	(QIF_BLIMITS | QIF_ILIMITS)
 #define QIF_USAGE	(QIF_SPACE | QIF_INODES)
 #define QIF_TIMES	(QIF_BTIME | QIF_ITIME)
@@ -172,7 +180,7 @@ enum {
 #include <asm/atomic.h>
 
 typedef __kernel_uid32_t qid_t; /* Type in which we store ids in memory */
-typedef __u64 qsize_t;          /* Type in which we store sizes */
+typedef long long qsize_t;	/* Type in which we store sizes */
 
 extern spinlock_t dq_data_lock;
 
@@ -187,12 +195,12 @@ extern spinlock_t dq_data_lock;
  * Data for one user/group kept in memory
  */
 struct mem_dqblk {
-	__u32 dqb_bhardlimit;	/* absolute limit on disk blks alloc */
-	__u32 dqb_bsoftlimit;	/* preferred limit on disk blks */
+	qsize_t dqb_bhardlimit;	/* absolute limit on disk blks alloc */
+	qsize_t dqb_bsoftlimit;	/* preferred limit on disk blks */
 	qsize_t dqb_curspace;	/* current used space */
-	__u32 dqb_ihardlimit;	/* absolute limit on allocated inodes */
-	__u32 dqb_isoftlimit;	/* preferred inode limit */
-	__u32 dqb_curinodes;	/* current # allocated inodes */
+	qsize_t dqb_ihardlimit;	/* absolute limit on allocated inodes */
+	qsize_t dqb_isoftlimit;	/* preferred inode limit */
+	qsize_t dqb_curinodes;	/* current # allocated inodes */
 	time_t dqb_btime;	/* time limit for excessive disk use */
 	time_t dqb_itime;	/* time limit for excessive inode use */
 };
@@ -212,10 +220,7 @@ struct mem_dqinfo {
 	unsigned int dqi_igrace;
 	qsize_t dqi_maxblimit;
 	qsize_t dqi_maxilimit;
-	union {
-		struct v1_mem_dqinfo v1_i;
-		struct v2_mem_dqinfo v2_i;
-	} u;
+	void *dqi_priv;
 };
 
 struct super_block;
@@ -249,6 +254,11 @@ extern struct dqstats dqstats;
 #define DQ_FAKE_B	3	/* no limits only usage */
 #define DQ_READ_B	4	/* dquot was read into memory */
 #define DQ_ACTIVE_B	5	/* dquot is active (dquot_release not called) */
+#define DQ_LASTSET_B	6	/* Following 6 bits (see QIF_) are reserved\
+				 * for the mask of entries set via SETQUOTA\
+				 * quotactl. They are set under dq_data_lock\
+				 * and the quota format handling dquot can\
+				 * clear them when it sees fit. */
 
 struct dquot {
 	struct hlist_node dq_hash;	/* Hash list in memory */
@@ -287,11 +297,13 @@ struct dquot_operations {
 	int (*initialize) (struct inode *, int);
 	int (*drop) (struct inode *);
 	int (*alloc_space) (struct inode *, qsize_t, int);
-	int (*alloc_inode) (const struct inode *, unsigned long);
+	int (*alloc_inode) (const struct inode *, qsize_t);
 	int (*free_space) (struct inode *, qsize_t);
-	int (*free_inode) (const struct inode *, unsigned long);
+	int (*free_inode) (const struct inode *, qsize_t);
 	int (*transfer) (struct inode *, struct iattr *);
 	int (*write_dquot) (struct dquot *);		/* Ordinary dquot write */
+	struct dquot *(*alloc_dquot)(struct super_block *, int);	/* Allocate memory for new dquot */
+	void (*destroy_dquot)(struct dquot *);		/* Free memory for dquot */
 	int (*acquire_dquot) (struct dquot *);		/* Quota is going to be created on disk */
 	int (*release_dquot) (struct dquot *);		/* Quota is going to be deleted from disk */
 	int (*mark_dirty) (struct dquot *);		/* Dquot is marked dirty */
@@ -320,12 +332,42 @@ struct quota_format_type {
 	struct quota_format_type *qf_next;
 };
 
-#define DQUOT_USR_ENABLED	0x01		/* User diskquotas enabled */
-#define DQUOT_GRP_ENABLED	0x02		/* Group diskquotas enabled */
-#define DQUOT_USR_SUSPENDED	0x04		/* User diskquotas are off, but
+/* Quota state flags - they actually come in two flavors - for users and groups */
+enum {
+	_DQUOT_USAGE_ENABLED = 0,		/* Track disk usage for users */
+	_DQUOT_LIMITS_ENABLED,			/* Enforce quota limits for users */
+	_DQUOT_SUSPENDED,			/* User diskquotas are off, but
 						 * we have necessary info in
 						 * memory to turn them on */
-#define DQUOT_GRP_SUSPENDED	0x08		/* The same for group quotas */
+	_DQUOT_STATE_FLAGS
+};
+#define DQUOT_USAGE_ENABLED	(1 << _DQUOT_USAGE_ENABLED)
+#define DQUOT_LIMITS_ENABLED	(1 << _DQUOT_LIMITS_ENABLED)
+#define DQUOT_SUSPENDED		(1 << _DQUOT_SUSPENDED)
+#define DQUOT_STATE_FLAGS	(DQUOT_USAGE_ENABLED | DQUOT_LIMITS_ENABLED | \
+				 DQUOT_SUSPENDED)
+/* Other quota flags */
+#define DQUOT_QUOTA_SYS_FILE	(1 << 6)	/* Quota file is a special
+						 * system file and user cannot
+						 * touch it. Filesystem is
+						 * responsible for setting
+						 * S_NOQUOTA, S_NOATIME flags
+						 */
+#define DQUOT_NEGATIVE_USAGE	(1 << 7)	/* Allow negative quota usage */
+
+static inline unsigned int dquot_state_flag(unsigned int flags, int type)
+{
+	if (type == USRQUOTA)
+		return flags;
+	return flags << _DQUOT_STATE_FLAGS;
+}
+
+static inline unsigned int dquot_generic_flag(unsigned int flags, int type)
+{
+	if (type == USRQUOTA)
+		return flags;
+	return flags >> _DQUOT_STATE_FLAGS;
+}
 
 struct quota_info {
 	unsigned int flags;			/* Flags for diskquotas on this device */

include/linux/quotaops.h

@@ -24,12 +24,21 @@ void sync_dquots(struct super_block *sb, int type);
 
 int dquot_initialize(struct inode *inode, int type);
 int dquot_drop(struct inode *inode);
+int dquot_drop_locked(struct inode *inode);
+struct dquot *dqget(struct super_block *sb, unsigned int id, int type);
+void dqput(struct dquot *dquot);
+int dquot_is_cached(struct super_block *sb, unsigned int id, int type);
+int dquot_scan_active(struct super_block *sb,
+		      int (*fn)(struct dquot *dquot, unsigned long priv),
+		      unsigned long priv);
+struct dquot *dquot_alloc(struct super_block *sb, int type);
+void dquot_destroy(struct dquot *dquot);
 
 int dquot_alloc_space(struct inode *inode, qsize_t number, int prealloc);
-int dquot_alloc_inode(const struct inode *inode, unsigned long number);
+int dquot_alloc_inode(const struct inode *inode, qsize_t number);
 
 int dquot_free_space(struct inode *inode, qsize_t number);
-int dquot_free_inode(const struct inode *inode, unsigned long number);
+int dquot_free_inode(const struct inode *inode, qsize_t number);
 
 int dquot_transfer(struct inode *inode, struct iattr *iattr);
 int dquot_commit(struct dquot *dquot);
@@ -40,11 +49,14 @@ int dquot_mark_dquot_dirty(struct dquot *dquot);
 
 int vfs_quota_on(struct super_block *sb, int type, int format_id,
  	char *path, int remount);
+int vfs_quota_enable(struct inode *inode, int type, int format_id,
+	unsigned int flags);
 int vfs_quota_on_path(struct super_block *sb, int type, int format_id,
  	struct path *path);
 int vfs_quota_on_mount(struct super_block *sb, char *qf_name,
  	int format_id, int type);
 int vfs_quota_off(struct super_block *sb, int type, int remount);
+int vfs_quota_disable(struct super_block *sb, int type, unsigned int flags);
 int vfs_quota_sync(struct super_block *sb, int type);
 int vfs_get_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii);
 int vfs_set_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii);
@@ -64,24 +76,22 @@ static inline struct mem_dqinfo *sb_dqinfo(struct super_block *sb, int type)
  * Functions for checking status of quota
  */
 
-static inline int sb_has_quota_enabled(struct super_block *sb, int type)
+static inline int sb_has_quota_usage_enabled(struct super_block *sb, int type)
 {
-	if (type == USRQUOTA)
-		return sb_dqopt(sb)->flags & DQUOT_USR_ENABLED;
-	return sb_dqopt(sb)->flags & DQUOT_GRP_ENABLED;
+	return sb_dqopt(sb)->flags &
+				dquot_state_flag(DQUOT_USAGE_ENABLED, type);
 }
 
-static inline int sb_any_quota_enabled(struct super_block *sb)
+static inline int sb_has_quota_limits_enabled(struct super_block *sb, int type)
 {
-	return sb_has_quota_enabled(sb, USRQUOTA) ||
-		sb_has_quota_enabled(sb, GRPQUOTA);
+	return sb_dqopt(sb)->flags &
+				dquot_state_flag(DQUOT_LIMITS_ENABLED, type);
 }
 
 static inline int sb_has_quota_suspended(struct super_block *sb, int type)
 {
-	if (type == USRQUOTA)
-		return sb_dqopt(sb)->flags & DQUOT_USR_SUSPENDED;
-	return sb_dqopt(sb)->flags & DQUOT_GRP_SUSPENDED;
+	return sb_dqopt(sb)->flags &
+				dquot_state_flag(DQUOT_SUSPENDED, type);
 }
 
 static inline int sb_any_quota_suspended(struct super_block *sb)
@@ -90,6 +100,31 @@ static inline int sb_any_quota_suspended(struct super_block *sb)
 		sb_has_quota_suspended(sb, GRPQUOTA);
 }
 
+/* Does kernel know about any quota information for given sb + type? */
+static inline int sb_has_quota_loaded(struct super_block *sb, int type)
+{
+	/* Currently if anything is on, then quota usage is on as well */
+	return sb_has_quota_usage_enabled(sb, type);
+}
+
+static inline int sb_any_quota_loaded(struct super_block *sb)
+{
+	return sb_has_quota_loaded(sb, USRQUOTA) ||
+		sb_has_quota_loaded(sb, GRPQUOTA);
+}
+
+static inline int sb_has_quota_active(struct super_block *sb, int type)
+{
+	return sb_has_quota_loaded(sb, type) &&
+	       !sb_has_quota_suspended(sb, type);
+}
+
+static inline int sb_any_quota_active(struct super_block *sb)
+{
+	return sb_has_quota_active(sb, USRQUOTA) ||
+	       sb_has_quota_active(sb, GRPQUOTA);
+}
+
 /*
  * Operations supported for diskquotas.
  */
@@ -104,7 +139,7 @@ extern struct quotactl_ops vfs_quotactl_ops;
 static inline void vfs_dq_init(struct inode *inode)
 {
 	BUG_ON(!inode->i_sb);
-	if (sb_any_quota_enabled(inode->i_sb) && !IS_NOQUOTA(inode))
+	if (sb_any_quota_active(inode->i_sb) && !IS_NOQUOTA(inode))
 		inode->i_sb->dq_op->initialize(inode, -1);
 }
 
@@ -112,7 +147,7 @@ static inline void vfs_dq_init(struct inode *inode)
  * a transaction (deadlocks possible otherwise) */
 static inline int vfs_dq_prealloc_space_nodirty(struct inode *inode, qsize_t nr)
 {
-	if (sb_any_quota_enabled(inode->i_sb)) {
+	if (sb_any_quota_active(inode->i_sb)) {
 		/* Used space is updated in alloc_space() */
 		if (inode->i_sb->dq_op->alloc_space(inode, nr, 1) == NO_QUOTA)
 			return 1;
@@ -132,7 +167,7 @@ static inline int vfs_dq_prealloc_space(struct inode *inode, qsize_t nr)
 
 static inline int vfs_dq_alloc_space_nodirty(struct inode *inode, qsize_t nr)
 {
-	if (sb_any_quota_enabled(inode->i_sb)) {
+	if (sb_any_quota_active(inode->i_sb)) {
 		/* Used space is updated in alloc_space() */
 		if (inode->i_sb->dq_op->alloc_space(inode, nr, 0) == NO_QUOTA)
 			return 1;
@@ -152,7 +187,7 @@ static inline int vfs_dq_alloc_space(struct inode *inode, qsize_t nr)
 
 static inline int vfs_dq_alloc_inode(struct inode *inode)
 {
-	if (sb_any_quota_enabled(inode->i_sb)) {
+	if (sb_any_quota_active(inode->i_sb)) {
 		vfs_dq_init(inode);
 		if (inode->i_sb->dq_op->alloc_inode(inode, 1) == NO_QUOTA)
 			return 1;
@@ -162,7 +197,7 @@ static inline int vfs_dq_alloc_inode(struct inode *inode)
 
 static inline void vfs_dq_free_space_nodirty(struct inode *inode, qsize_t nr)
 {
-	if (sb_any_quota_enabled(inode->i_sb))
+	if (sb_any_quota_active(inode->i_sb))
 		inode->i_sb->dq_op->free_space(inode, nr);
 	else
 		inode_sub_bytes(inode, nr);
@@ -176,7 +211,7 @@ static inline void vfs_dq_free_space(struct inode *inode, qsize_t nr)
 
 static inline void vfs_dq_free_inode(struct inode *inode)
 {
-	if (sb_any_quota_enabled(inode->i_sb))
+	if (sb_any_quota_active(inode->i_sb))
 		inode->i_sb->dq_op->free_inode(inode, 1);
 }
 
@@ -197,12 +232,12 @@ static inline int vfs_dq_off(struct super_block *sb, int remount)
 
 #else
 
-static inline int sb_has_quota_enabled(struct super_block *sb, int type)
+static inline int sb_has_quota_usage_enabled(struct super_block *sb, int type)
 {
 	return 0;
 }
 
-static inline int sb_any_quota_enabled(struct super_block *sb)
+static inline int sb_has_quota_limits_enabled(struct super_block *sb, int type)
 {
 	return 0;
 }
@@ -217,6 +252,27 @@ static inline int sb_any_quota_suspended(struct super_block *sb)
 	return 0;
 }
 
+/* Does kernel know about any quota information for given sb + type? */
+static inline int sb_has_quota_loaded(struct super_block *sb, int type)
+{
+	return 0;
+}
+
+static inline int sb_any_quota_loaded(struct super_block *sb)
+{
+	return 0;
+}
+
+static inline int sb_has_quota_active(struct super_block *sb, int type)
+{
+	return 0;
+}
+
+static inline int sb_any_quota_active(struct super_block *sb)
+{
+	return 0;
+}
+
 /*
  * NO-OP when quota not configured.
  */