Merge branch 'master' of git://oss.sgi.com/xfs/xfs into for-linus

9 files changed, 846 insertions, 374 deletions

diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c
new file mode 100644
index 000000000000..1e9d1246eebc
--- /dev/null
+++ b/fs/xfs/linux-2.6/xfs_acl.c
@@ -0,0 +1,523 @@
+/*
+ * Copyright (c) 2008, Christoph Hellwig
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+#include "xfs.h"
+#include "xfs_acl.h"
+#include "xfs_attr.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_inode.h"
+#include "xfs_vnodeops.h"
+#include <linux/xattr.h>
+#include <linux/posix_acl_xattr.h>
+
+
+#define XFS_ACL_NOT_CACHED	((void *)-1)
+
+/*
+ * Locking scheme:
+ *  - all ACL updates are protected by inode->i_mutex, which is taken before
+ *    calling into this file.
+ *  - access and updates to the ip->i_acl and ip->i_default_acl pointers are
+ *    protected by inode->i_lock.
+ */
+
+STATIC struct posix_acl *
+xfs_acl_from_disk(struct xfs_acl *aclp)
+{
+	struct posix_acl_entry *acl_e;
+	struct posix_acl *acl;
+	struct xfs_acl_entry *ace;
+	int count, i;
+
+	count = be32_to_cpu(aclp->acl_cnt);
+
+	acl = posix_acl_alloc(count, GFP_KERNEL);
+	if (!acl)
+		return ERR_PTR(-ENOMEM);
+
+	for (i = 0; i < count; i++) {
+		acl_e = &acl->a_entries[i];
+		ace = &aclp->acl_entry[i];
+
+		/*
+		 * The tag is 32 bits on disk and 16 bits in core.
+		 *
+		 * Because every access to it goes through the core
+		 * format first this is not a problem.
+		 */
+		acl_e->e_tag = be32_to_cpu(ace->ae_tag);
+		acl_e->e_perm = be16_to_cpu(ace->ae_perm);
+
+		switch (acl_e->e_tag) {
+		case ACL_USER:
+		case ACL_GROUP:
+			acl_e->e_id = be32_to_cpu(ace->ae_id);
+			break;
+		case ACL_USER_OBJ:
+		case ACL_GROUP_OBJ:
+		case ACL_MASK:
+		case ACL_OTHER:
+			acl_e->e_id = ACL_UNDEFINED_ID;
+			break;
+		default:
+			goto fail;
+		}
+	}
+	return acl;
+
+fail:
+	posix_acl_release(acl);
+	return ERR_PTR(-EINVAL);
+}
+
+STATIC void
+xfs_acl_to_disk(struct xfs_acl *aclp, const struct posix_acl *acl)
+{
+	const struct posix_acl_entry *acl_e;
+	struct xfs_acl_entry *ace;
+	int i;
+
+	aclp->acl_cnt = cpu_to_be32(acl->a_count);
+	for (i = 0; i < acl->a_count; i++) {
+		ace = &aclp->acl_entry[i];
+		acl_e = &acl->a_entries[i];
+
+		ace->ae_tag = cpu_to_be32(acl_e->e_tag);
+		ace->ae_id = cpu_to_be32(acl_e->e_id);
+		ace->ae_perm = cpu_to_be16(acl_e->e_perm);
+	}
+}
+
+/*
+ * Update the cached ACL pointer in the inode.
+ *
+ * Because we don't hold any locks while reading/writing the attribute
+ * from/to disk another thread could have raced and updated the cached
+ * ACL value before us. In that case we release the previous cached value
+ * and update it with our new value.
+ */
+STATIC void
+xfs_update_cached_acl(struct inode *inode, struct posix_acl **p_acl,
+		struct posix_acl *acl)
+{
+	spin_lock(&inode->i_lock);
+	if (*p_acl && *p_acl != XFS_ACL_NOT_CACHED)
+		posix_acl_release(*p_acl);
+	*p_acl = posix_acl_dup(acl);
+	spin_unlock(&inode->i_lock);
+}
+
+struct posix_acl *
+xfs_get_acl(struct inode *inode, int type)
+{
+	struct xfs_inode *ip = XFS_I(inode);
+	struct posix_acl *acl = NULL, **p_acl;
+	struct xfs_acl *xfs_acl;
+	int len = sizeof(struct xfs_acl);
+	char *ea_name;
+	int error;
+
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		ea_name = SGI_ACL_FILE;
+		p_acl = &ip->i_acl;
+		break;
+	case ACL_TYPE_DEFAULT:
+		ea_name = SGI_ACL_DEFAULT;
+		p_acl = &ip->i_default_acl;
+		break;
+	default:
+		return ERR_PTR(-EINVAL);
+	}
+
+	spin_lock(&inode->i_lock);
+	if (*p_acl != XFS_ACL_NOT_CACHED)
+		acl = posix_acl_dup(*p_acl);
+	spin_unlock(&inode->i_lock);
+
+	/*
+	 * If we have a cached ACLs value just return it, not need to
+	 * go out to the disk.
+	 */
+	if (acl)
+		return acl;
+
+	xfs_acl = kzalloc(sizeof(struct xfs_acl), GFP_KERNEL);
+	if (!xfs_acl)
+		return ERR_PTR(-ENOMEM);
+
+	error = -xfs_attr_get(ip, ea_name, (char *)xfs_acl, &len, ATTR_ROOT);
+	if (error) {
+		/*
+		 * If the attribute doesn't exist make sure we have a negative
+		 * cache entry, for any other error assume it is transient and
+		 * leave the cache entry as XFS_ACL_NOT_CACHED.
+		 */
+		if (error == -ENOATTR) {
+			acl = NULL;
+			goto out_update_cache;
+		}
+		goto out;
+	}
+
+	acl = xfs_acl_from_disk(xfs_acl);
+	if (IS_ERR(acl))
+		goto out;
+
+ out_update_cache:
+	xfs_update_cached_acl(inode, p_acl, acl);
+ out:
+	kfree(xfs_acl);
+	return acl;
+}
+
+STATIC int
+xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
+{
+	struct xfs_inode *ip = XFS_I(inode);
+	struct posix_acl **p_acl;
+	char *ea_name;
+	int error;
+
+	if (S_ISLNK(inode->i_mode))
+		return -EOPNOTSUPP;
+
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		ea_name = SGI_ACL_FILE;
+		p_acl = &ip->i_acl;
+		break;
+	case ACL_TYPE_DEFAULT:
+		if (!S_ISDIR(inode->i_mode))
+			return acl ? -EACCES : 0;
+		ea_name = SGI_ACL_DEFAULT;
+		p_acl = &ip->i_default_acl;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (acl) {
+		struct xfs_acl *xfs_acl;
+		int len;
+
+		xfs_acl = kzalloc(sizeof(struct xfs_acl), GFP_KERNEL);
+		if (!xfs_acl)
+			return -ENOMEM;
+
+		xfs_acl_to_disk(xfs_acl, acl);
+		len = sizeof(struct xfs_acl) -
+			(sizeof(struct xfs_acl_entry) *
+			 (XFS_ACL_MAX_ENTRIES - acl->a_count));
+
+		error = -xfs_attr_set(ip, ea_name, (char *)xfs_acl,
+				len, ATTR_ROOT);
+
+		kfree(xfs_acl);
+	} else {
+		/*
+		 * A NULL ACL argument means we want to remove the ACL.
+		 */
+		error = -xfs_attr_remove(ip, ea_name, ATTR_ROOT);
+
+		/*
+		 * If the attribute didn't exist to start with that's fine.
+		 */
+		if (error == -ENOATTR)
+			error = 0;
+	}
+
+	if (!error)
+		xfs_update_cached_acl(inode, p_acl, acl);
+	return error;
+}
+
+int
+xfs_check_acl(struct inode *inode, int mask)
+{
+	struct xfs_inode *ip = XFS_I(inode);
+	struct posix_acl *acl;
+	int error = -EAGAIN;
+
+	xfs_itrace_entry(ip);
+
+	/*
+	 * If there is no attribute fork no ACL exists on this inode and
+	 * we can skip the whole exercise.
+	 */
+	if (!XFS_IFORK_Q(ip))
+		return -EAGAIN;
+
+	acl = xfs_get_acl(inode, ACL_TYPE_ACCESS);
+	if (IS_ERR(acl))
+		return PTR_ERR(acl);
+	if (acl) {
+		error = posix_acl_permission(inode, acl, mask);
+		posix_acl_release(acl);
+	}
+
+	return error;
+}
+
+static int
+xfs_set_mode(struct inode *inode, mode_t mode)
+{
+	int error = 0;
+
+	if (mode != inode->i_mode) {
+		struct iattr iattr;
+
+		iattr.ia_valid = ATTR_MODE;
+		iattr.ia_mode = mode;
+
+		error = -xfs_setattr(XFS_I(inode), &iattr, XFS_ATTR_NOACL);
+	}
+
+	return error;
+}
+
+static int
+xfs_acl_exists(struct inode *inode, char *name)
+{
+	int len = sizeof(struct xfs_acl);
+
+	return (xfs_attr_get(XFS_I(inode), name, NULL, &len,
+			    ATTR_ROOT|ATTR_KERNOVAL) == 0);
+}
+
+int
+posix_acl_access_exists(struct inode *inode)
+{
+	return xfs_acl_exists(inode, SGI_ACL_FILE);
+}
+
+int
+posix_acl_default_exists(struct inode *inode)
+{
+	if (!S_ISDIR(inode->i_mode))
+		return 0;
+	return xfs_acl_exists(inode, SGI_ACL_DEFAULT);
+}
+
+/*
+ * No need for i_mutex because the inode is not yet exposed to the VFS.
+ */
+int
+xfs_inherit_acl(struct inode *inode, struct posix_acl *default_acl)
+{
+	struct posix_acl *clone;
+	mode_t mode;
+	int error = 0, inherit = 0;
+
+	if (S_ISDIR(inode->i_mode)) {
+		error = xfs_set_acl(inode, ACL_TYPE_DEFAULT, default_acl);
+		if (error)
+			return error;
+	}
+
+	clone = posix_acl_clone(default_acl, GFP_KERNEL);
+	if (!clone)
+		return -ENOMEM;
+
+	mode = inode->i_mode;
+	error = posix_acl_create_masq(clone, &mode);
+	if (error < 0)
+		goto out_release_clone;
+
+	/*
+	 * If posix_acl_create_masq returns a positive value we need to
+	 * inherit a permission that can't be represented using the Unix
+	 * mode bits and we actually need to set an ACL.
+	 */
+	if (error > 0)
+		inherit = 1;
+
+	error = xfs_set_mode(inode, mode);
+	if (error)
+		goto out_release_clone;
+
+	if (inherit)
+		error = xfs_set_acl(inode, ACL_TYPE_ACCESS, clone);
+
+ out_release_clone:
+	posix_acl_release(clone);
+	return error;
+}
+
+int
+xfs_acl_chmod(struct inode *inode)
+{
+	struct posix_acl *acl, *clone;
+	int error;
+
+	if (S_ISLNK(inode->i_mode))
+		return -EOPNOTSUPP;
+
+	acl = xfs_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;
+
+	error = posix_acl_chmod_masq(clone, inode->i_mode);
+	if (!error)
+		error = xfs_set_acl(inode, ACL_TYPE_ACCESS, clone);
+
+	posix_acl_release(clone);
+	return error;
+}
+
+void
+xfs_inode_init_acls(struct xfs_inode *ip)
+{
+	/*
+	 * No need for locking, inode is not live yet.
+	 */
+	ip->i_acl = XFS_ACL_NOT_CACHED;
+	ip->i_default_acl = XFS_ACL_NOT_CACHED;
+}
+
+void
+xfs_inode_clear_acls(struct xfs_inode *ip)
+{
+	/*
+	 * No need for locking here, the inode is not live anymore
+	 * and just about to be freed.
+	 */
+	if (ip->i_acl != XFS_ACL_NOT_CACHED)
+		posix_acl_release(ip->i_acl);
+	if (ip->i_default_acl != XFS_ACL_NOT_CACHED)
+		posix_acl_release(ip->i_default_acl);
+}
+
+
+/*
+ * System xattr handlers.
+ *
+ * Currently Posix ACLs are the only system namespace extended attribute
+ * handlers supported by XFS, so we just implement the handlers here.
+ * If we ever support other system extended attributes this will need
+ * some refactoring.
+ */
+
+static int
+xfs_decode_acl(const char *name)
+{
+	if (strcmp(name, "posix_acl_access") == 0)
+		return ACL_TYPE_ACCESS;
+	else if (strcmp(name, "posix_acl_default") == 0)
+		return ACL_TYPE_DEFAULT;
+	return -EINVAL;
+}
+
+static int
+xfs_xattr_system_get(struct inode *inode, const char *name,
+		void *value, size_t size)
+{
+	struct posix_acl *acl;
+	int type, error;
+
+	type = xfs_decode_acl(name);
+	if (type < 0)
+		return type;
+
+	acl = xfs_get_acl(inode, type);
+	if (IS_ERR(acl))
+		return PTR_ERR(acl);
+	if (acl == NULL)
+		return -ENODATA;
+
+	error = posix_acl_to_xattr(acl, value, size);
+	posix_acl_release(acl);
+
+	return error;
+}
+
+static int
+xfs_xattr_system_set(struct inode *inode, const char *name,
+		const void *value, size_t size, int flags)
+{
+	struct posix_acl *acl = NULL;
+	int error = 0, type;
+
+	type = xfs_decode_acl(name);
+	if (type < 0)
+		return type;
+	if (flags & XATTR_CREATE)
+		return -EINVAL;
+	if (type == ACL_TYPE_DEFAULT && !S_ISDIR(inode->i_mode))
+		return value ? -EACCES : 0;
+	if ((current_fsuid() != inode->i_uid) && !capable(CAP_FOWNER))
+		return -EPERM;
+
+	if (!value)
+		goto set_acl;
+
+	acl = posix_acl_from_xattr(value, size);
+	if (!acl) {
+		/*
+		 * acl_set_file(3) may request that we set default ACLs with
+		 * zero length -- defend (gracefully) against that here.
+		 */
+		goto out;
+	}
+	if (IS_ERR(acl)) {
+		error = PTR_ERR(acl);
+		goto out;
+	}
+
+	error = posix_acl_valid(acl);
+	if (error)
+		goto out_release;
+
+	error = -EINVAL;
+	if (acl->a_count > XFS_ACL_MAX_ENTRIES)
+		goto out_release;
+
+	if (type == ACL_TYPE_ACCESS) {
+		mode_t mode = inode->i_mode;
+		error = posix_acl_equiv_mode(acl, &mode);
+
+		if (error <= 0) {
+			posix_acl_release(acl);
+			acl = NULL;
+
+			if (error < 0)
+				return error;
+		}
+
+		error = xfs_set_mode(inode, mode);
+		if (error)
+			goto out_release;
+	}
+
+ set_acl:
+	error = xfs_set_acl(inode, type, acl);
+ out_release:
+	posix_acl_release(acl);
+ out:
+	return error;
+}
+
+struct xattr_handler xfs_xattr_system_handler = {
+	.prefix	= XATTR_SYSTEM_PREFIX,
+	.get	= xfs_xattr_system_get,
+	.set	= xfs_xattr_system_set,
+};
diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c
index 34eaab608e6e..5bb523d7f37e 100644
--- a/fs/xfs/linux-2.6/xfs_ioctl.c
+++ b/fs/xfs/linux-2.6/xfs_ioctl.c
@@ -41,7 +41,6 @@
 #include "xfs_itable.h"
 #include "xfs_error.h"
 #include "xfs_rw.h"
-#include "xfs_acl.h"
 #include "xfs_attr.h"
 #include "xfs_bmap.h"
 #include "xfs_buf_item.h"
@@ -899,7 +898,8 @@ xfs_ioctl_setattr(
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_trans	*tp;
 	unsigned int		lock_flags = 0;
-	struct xfs_dquot	*udqp = NULL, *gdqp = NULL;
+	struct xfs_dquot	*udqp = NULL;
+	struct xfs_dquot	*gdqp = NULL;
 	struct xfs_dquot	*olddquot = NULL;
 	int			code;
 
@@ -919,7 +919,7 @@ xfs_ioctl_setattr(
 	 * because the i_*dquot fields will get updated anyway.
 	 */
 	if (XFS_IS_QUOTA_ON(mp) && (mask & FSX_PROJID)) {
-		code = XFS_QM_DQVOPALLOC(mp, ip, ip->i_d.di_uid,
+		code = xfs_qm_vop_dqalloc(ip, ip->i_d.di_uid,
 					 ip->i_d.di_gid, fa->fsx_projid,
 					 XFS_QMOPT_PQUOTA, &udqp, &gdqp);
 		if (code)
@@ -954,10 +954,11 @@ xfs_ioctl_setattr(
 	 * Do a quota reservation only if projid is actually going to change.
 	 */
 	if (mask & FSX_PROJID) {
-		if (XFS_IS_PQUOTA_ON(mp) &&
+		if (XFS_IS_QUOTA_RUNNING(mp) &&
+		    XFS_IS_PQUOTA_ON(mp) &&
 		    ip->i_d.di_projid != fa->fsx_projid) {
 			ASSERT(tp);
-			code = XFS_QM_DQVOPCHOWNRESV(mp, tp, ip, udqp, gdqp,
+			code = xfs_qm_vop_chown_reserve(tp, ip, udqp, gdqp,
 						capable(CAP_FOWNER) ?
 						XFS_QMOPT_FORCE_RES : 0);
 			if (code)	/* out of quota */
@@ -1059,8 +1060,8 @@ xfs_ioctl_setattr(
 		 * in the transaction.
 		 */
 		if (ip->i_d.di_projid != fa->fsx_projid) {
-			if (XFS_IS_PQUOTA_ON(mp)) {
-				olddquot = XFS_QM_DQVOPCHOWN(mp, tp, ip,
+			if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_PQUOTA_ON(mp)) {
+				olddquot = xfs_qm_vop_chown(tp, ip,
 							&ip->i_gdquot, gdqp);
 			}
 			ip->i_d.di_projid = fa->fsx_projid;
@@ -1106,9 +1107,9 @@ xfs_ioctl_setattr(
 	/*
 	 * Release any dquot(s) the inode had kept before chown.
 	 */
-	XFS_QM_DQRELE(mp, olddquot);
-	XFS_QM_DQRELE(mp, udqp);
-	XFS_QM_DQRELE(mp, gdqp);
+	xfs_qm_dqrele(olddquot);
+	xfs_qm_dqrele(udqp);
+	xfs_qm_dqrele(gdqp);
 
 	if (code)
 		return code;
@@ -1122,8 +1123,8 @@ xfs_ioctl_setattr(
 	return 0;
 
  error_return:
-	XFS_QM_DQRELE(mp, udqp);
-	XFS_QM_DQRELE(mp, gdqp);
+	xfs_qm_dqrele(udqp);
+	xfs_qm_dqrele(gdqp);
 	xfs_trans_cancel(tp, 0);
 	if (lock_flags)
 		xfs_iunlock(ip, lock_flags);
diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c
index 6075382336d7..58973bb46038 100644
--- a/fs/xfs/linux-2.6/xfs_iops.c
+++ b/fs/xfs/linux-2.6/xfs_iops.c
@@ -17,6 +17,7 @@
  */
 #include "xfs.h"
 #include "xfs_fs.h"
+#include "xfs_acl.h"
 #include "xfs_bit.h"
 #include "xfs_log.h"
 #include "xfs_inum.h"
@@ -51,6 +52,7 @@
 #include <linux/capability.h>
 #include <linux/xattr.h>
 #include <linux/namei.h>
+#include <linux/posix_acl.h>
 #include <linux/security.h>
 #include <linux/falloc.h>
 #include <linux/fiemap.h>
@@ -202,9 +204,8 @@ xfs_vn_mknod(
 {
 	struct inode	*inode;
 	struct xfs_inode *ip = NULL;
-	xfs_acl_t	*default_acl = NULL;
+	struct posix_acl *default_acl = NULL;
 	struct xfs_name	name;
-	int (*test_default_acl)(struct inode *) = _ACL_DEFAULT_EXISTS;
 	int		error;
 
 	/*
@@ -219,18 +220,14 @@ xfs_vn_mknod(
 		rdev = 0;
 	}
 
-	if (test_default_acl && test_default_acl(dir)) {
-		if (!_ACL_ALLOC(default_acl)) {
-			return -ENOMEM;
-		}
-		if (!_ACL_GET_DEFAULT(dir, default_acl)) {
-			_ACL_FREE(default_acl);
-			default_acl = NULL;
-		}
-	}
+	if (IS_POSIXACL(dir)) {
+		default_acl = xfs_get_acl(dir, ACL_TYPE_DEFAULT);
+		if (IS_ERR(default_acl))
+			return -PTR_ERR(default_acl);
 
-	if (IS_POSIXACL(dir) && !default_acl)
-		mode &= ~current_umask();
+		if (!default_acl)
+			mode &= ~current_umask();
+	}
 
 	xfs_dentry_to_name(&name, dentry);
 	error = xfs_create(XFS_I(dir), &name, mode, rdev, &ip, NULL);
@@ -244,10 +241,10 @@ xfs_vn_mknod(
 		goto out_cleanup_inode;
 
 	if (default_acl) {
-		error = _ACL_INHERIT(inode, mode, default_acl);
+		error = -xfs_inherit_acl(inode, default_acl);
 		if (unlikely(error))
 			goto out_cleanup_inode;
-		_ACL_FREE(default_acl);
+		posix_acl_release(default_acl);
 	}
 
 
@@ -257,8 +254,7 @@ xfs_vn_mknod(
  out_cleanup_inode:
 	xfs_cleanup_inode(dir, inode, dentry);
  out_free_acl:
-	if (default_acl)
-		_ACL_FREE(default_acl);
+	posix_acl_release(default_acl);
 	return -error;
 }
 
@@ -488,26 +484,6 @@ xfs_vn_put_link(
 		kfree(s);
 }
 
-#ifdef CONFIG_XFS_POSIX_ACL
-STATIC int
-xfs_check_acl(
-	struct inode		*inode,
-	int			mask)
-{
-	struct xfs_inode	*ip = XFS_I(inode);
-	int			error;
-
-	xfs_itrace_entry(ip);
-
-	if (XFS_IFORK_Q(ip)) {
-		error = xfs_acl_iaccess(ip, mask, NULL);
-		if (error != -1)
-			return -error;
-	}
-
-	return -EAGAIN;
-}
-
 STATIC int
 xfs_vn_permission(
 	struct inode		*inode,
@@ -515,9 +491,6 @@ xfs_vn_permission(
 {
 	return generic_permission(inode, mask, xfs_check_acl);
 }
-#else
-#define xfs_vn_permission NULL
-#endif
 
 STATIC int
 xfs_vn_getattr(
diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c
index 9142192ccbe6..7078974a6eee 100644
--- a/fs/xfs/linux-2.6/xfs_lrw.c
+++ b/fs/xfs/linux-2.6/xfs_lrw.c
@@ -42,7 +42,6 @@
 #include "xfs_error.h"
 #include "xfs_itable.h"
 #include "xfs_rw.h"
-#include "xfs_acl.h"
 #include "xfs_attr.h"
 #include "xfs_inode_item.h"
 #include "xfs_buf_item.h"
diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c
index 94d9a633d3d9..cb6e2cca214f 100644
--- a/fs/xfs/linux-2.6/xfs_quotaops.c
+++ b/fs/xfs/linux-2.6/xfs_quotaops.c
@@ -50,9 +50,11 @@ xfs_fs_quota_sync(
 {
 	struct xfs_mount	*mp = XFS_M(sb);
 
+	if (sb->s_flags & MS_RDONLY)
+		return -EROFS;
 	if (!XFS_IS_QUOTA_RUNNING(mp))
 		return -ENOSYS;
-	return -xfs_sync_inodes(mp, SYNC_DELWRI);
+	return -xfs_sync_data(mp, 0);
 }
 
 STATIC int
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index 08d6bd9a3947..2e09efbca8db 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -43,7 +43,6 @@
 #include "xfs_itable.h"
 #include "xfs_fsops.h"
 #include "xfs_rw.h"
-#include "xfs_acl.h"
 #include "xfs_attr.h"
 #include "xfs_buf_item.h"
 #include "xfs_utils.h"
@@ -405,6 +404,14 @@ xfs_parseargs(
 		return EINVAL;
 	}
 
+#ifndef CONFIG_XFS_QUOTA
+	if (XFS_IS_QUOTA_RUNNING(mp)) {
+		cmn_err(CE_WARN,
+			"XFS: quota support not available in this kernel.");
+		return EINVAL;
+	}
+#endif
+
 	if ((mp->m_qflags & (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE)) &&
 	    (mp->m_qflags & (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE))) {
 		cmn_err(CE_WARN,
@@ -1063,7 +1070,18 @@ xfs_fs_put_super(
 	int			unmount_event_flags = 0;
 
 	xfs_syncd_stop(mp);
-	xfs_sync_inodes(mp, SYNC_ATTR|SYNC_DELWRI);
+
+	if (!(sb->s_flags & MS_RDONLY)) {
+		/*
+		 * XXX(hch): this should be SYNC_WAIT.
+		 *
+		 * Or more likely not needed at all because the VFS is already
+		 * calling ->sync_fs after shutting down all filestem
+		 * operations and just before calling ->put_super.
+		 */
+		xfs_sync_data(mp, 0);
+		xfs_sync_attr(mp, 0);
+	}
 
 #ifdef HAVE_DMAPI
 	if (mp->m_flags & XFS_MOUNT_DMAPI) {
@@ -1098,7 +1116,6 @@ xfs_fs_put_super(
 	xfs_freesb(mp);
 	xfs_icsb_destroy_counters(mp);
 	xfs_close_devices(mp);
-	xfs_qmops_put(mp);
 	xfs_dmops_put(mp);
 	xfs_free_fsname(mp);
 	kfree(mp);
@@ -1158,6 +1175,7 @@ xfs_fs_statfs(
 {
 	struct xfs_mount	*mp = XFS_M(dentry->d_sb);
 	xfs_sb_t		*sbp = &mp->m_sb;
+	struct xfs_inode	*ip = XFS_I(dentry->d_inode);
 	__uint64_t		fakeinos, id;
 	xfs_extlen_t		lsize;
 
@@ -1186,7 +1204,10 @@ xfs_fs_statfs(
 	statp->f_ffree = statp->f_files - (sbp->sb_icount - sbp->sb_ifree);
 	spin_unlock(&mp->m_sb_lock);
 
-	XFS_QM_DQSTATVFS(XFS_I(dentry->d_inode), statp);
+	if ((ip->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) ||
+	    ((mp->m_qflags & (XFS_PQUOTA_ACCT|XFS_OQUOTA_ENFD))) ==
+			      (XFS_PQUOTA_ACCT|XFS_OQUOTA_ENFD))
+		xfs_qm_statvfs(ip, statp);
 	return 0;
 }
 
@@ -1394,16 +1415,13 @@ xfs_fs_fill_super(
 	error = xfs_dmops_get(mp);
 	if (error)
 		goto out_free_fsname;
-	error = xfs_qmops_get(mp);
-	if (error)
-		goto out_put_dmops;
 
 	if (silent)
 		flags |= XFS_MFSI_QUIET;
 
 	error = xfs_open_devices(mp);
 	if (error)
-		goto out_put_qmops;
+		goto out_put_dmops;
 
 	if (xfs_icsb_init_counters(mp))
 		mp->m_flags |= XFS_MOUNT_NO_PERCPU_SB;
@@ -1471,8 +1489,6 @@ xfs_fs_fill_super(
  out_destroy_counters:
 	xfs_icsb_destroy_counters(mp);
 	xfs_close_devices(mp);
- out_put_qmops:
-	xfs_qmops_put(mp);
  out_put_dmops:
 	xfs_dmops_put(mp);
  out_free_fsname:
@@ -1706,18 +1722,8 @@ xfs_init_zones(void)
 	if (!xfs_ili_zone)
 		goto out_destroy_inode_zone;
 
-#ifdef CONFIG_XFS_POSIX_ACL
-	xfs_acl_zone = kmem_zone_init(sizeof(xfs_acl_t), "xfs_acl");
-	if (!xfs_acl_zone)
-		goto out_destroy_ili_zone;
-#endif
-
 	return 0;
 
-#ifdef CONFIG_XFS_POSIX_ACL
- out_destroy_ili_zone:
-#endif
-	kmem_zone_destroy(xfs_ili_zone);
  out_destroy_inode_zone:
 	kmem_zone_destroy(xfs_inode_zone);
  out_destroy_efi_zone:
@@ -1751,9 +1757,6 @@ xfs_init_zones(void)
 STATIC void
 xfs_destroy_zones(void)
 {
-#ifdef CONFIG_XFS_POSIX_ACL
-	kmem_zone_destroy(xfs_acl_zone);
-#endif
 	kmem_zone_destroy(xfs_ili_zone);
 	kmem_zone_destroy(xfs_inode_zone);
 	kmem_zone_destroy(xfs_efi_zone);
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index f7ba76633c29..b619d6b8ca43 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -43,166 +43,267 @@
 #include "xfs_buf_item.h"
 #include "xfs_inode_item.h"
 #include "xfs_rw.h"
+#include "xfs_quota.h"
 
 #include <linux/kthread.h>
 #include <linux/freezer.h>
 
-/*
- * Sync all the inodes in the given AG according to the
- * direction given by the flags.
- */
-STATIC int
-xfs_sync_inodes_ag(
-	xfs_mount_t	*mp,
-	int		ag,
-	int		flags)
-{
-	xfs_perag_t	*pag = &mp->m_perag[ag];
-	int		nr_found;
-	uint32_t	first_index = 0;
-	int		error = 0;
-	int		last_error = 0;
 
-	do {
-		struct inode	*inode;
-		xfs_inode_t	*ip = NULL;
-		int		lock_flags = XFS_ILOCK_SHARED;
+STATIC xfs_inode_t *
+xfs_inode_ag_lookup(
+	struct xfs_mount	*mp,
+	struct xfs_perag	*pag,
+	uint32_t		*first_index,
+	int			tag)
+{
+	int			nr_found;
+	struct xfs_inode	*ip;
 
-		/*
-		 * use a gang lookup to find the next inode in the tree
-		 * as the tree is sparse and a gang lookup walks to find
-		 * the number of objects requested.
-		 */
-		read_lock(&pag->pag_ici_lock);
+	/*
+	 * use a gang lookup to find the next inode in the tree
+	 * as the tree is sparse and a gang lookup walks to find
+	 * the number of objects requested.
+	 */
+	read_lock(&pag->pag_ici_lock);
+	if (tag == XFS_ICI_NO_TAG) {
 		nr_found = radix_tree_gang_lookup(&pag->pag_ici_root,
-				(void**)&ip, first_index, 1);
+				(void **)&ip, *first_index, 1);
+	} else {
+		nr_found = radix_tree_gang_lookup_tag(&pag->pag_ici_root,
+				(void **)&ip, *first_index, 1, tag);
+	}
+	if (!nr_found)
+		goto unlock;
 
-		if (!nr_found) {
-			read_unlock(&pag->pag_ici_lock);
-			break;
-		}
+	/*
+	 * Update the index for the next lookup. Catch overflows
+	 * into the next AG range which can occur if we have inodes
+	 * in the last block of the AG and we are currently
+	 * pointing to the last inode.
+	 */
+	*first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
+	if (*first_index < XFS_INO_TO_AGINO(mp, ip->i_ino))
+		goto unlock;
 
-		/*
-		 * Update the index for the next lookup. Catch overflows
-		 * into the next AG range which can occur if we have inodes
-		 * in the last block of the AG and we are currently
-		 * pointing to the last inode.
-		 */
-		first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
-		if (first_index < XFS_INO_TO_AGINO(mp, ip->i_ino)) {
-			read_unlock(&pag->pag_ici_lock);
-			break;
-		}
+	return ip;
 
-		/* nothing to sync during shutdown */
-		if (XFS_FORCED_SHUTDOWN(mp)) {
-			read_unlock(&pag->pag_ici_lock);
-			return 0;
-		}
+unlock:
+	read_unlock(&pag->pag_ici_lock);
+	return NULL;
+}
 
-		/*
-		 * If we can't get a reference on the inode, it must be
-		 * in reclaim. Leave it for the reclaim code to flush.
-		 */
-		inode = VFS_I(ip);
-		if (!igrab(inode)) {
-			read_unlock(&pag->pag_ici_lock);
-			continue;
-		}
-		read_unlock(&pag->pag_ici_lock);
+STATIC int
+xfs_inode_ag_walk(
+	struct xfs_mount	*mp,
+	xfs_agnumber_t		ag,
+	int			(*execute)(struct xfs_inode *ip,
+					   struct xfs_perag *pag, int flags),
+	int			flags,
+	int			tag)
+{
+	struct xfs_perag	*pag = &mp->m_perag[ag];
+	uint32_t		first_index;
+	int			last_error = 0;
+	int			skipped;
 
-		/* avoid new or bad inodes */
-		if (is_bad_inode(inode) ||
-		    xfs_iflags_test(ip, XFS_INEW)) {
-			IRELE(ip);
-			continue;
-		}
+restart:
+	skipped = 0;
+	first_index = 0;
+	do {
+		int		error = 0;
+		xfs_inode_t	*ip;
 
-		/*
-		 * If we have to flush data or wait for I/O completion
-		 * we need to hold the iolock.
-		 */
-		if (flags & SYNC_DELWRI) {
-			if (VN_DIRTY(inode)) {
-				if (flags & SYNC_TRYLOCK) {
-					if (xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED))
-						lock_flags |= XFS_IOLOCK_SHARED;
-				} else {
-					xfs_ilock(ip, XFS_IOLOCK_SHARED);
-					lock_flags |= XFS_IOLOCK_SHARED;
-				}
-				if (lock_flags & XFS_IOLOCK_SHARED) {
-					error = xfs_flush_pages(ip, 0, -1,
-							(flags & SYNC_WAIT) ? 0
-								: XFS_B_ASYNC,
-							FI_NONE);
-				}
-			}
-			if (VN_CACHED(inode) && (flags & SYNC_IOWAIT))
-				xfs_ioend_wait(ip);
-		}
-		xfs_ilock(ip, XFS_ILOCK_SHARED);
-
-		if ((flags & SYNC_ATTR) && !xfs_inode_clean(ip)) {
-			if (flags & SYNC_WAIT) {
-				xfs_iflock(ip);
-				if (!xfs_inode_clean(ip))
-					error = xfs_iflush(ip, XFS_IFLUSH_SYNC);
-				else
-					xfs_ifunlock(ip);
-			} else if (xfs_iflock_nowait(ip)) {
-				if (!xfs_inode_clean(ip))
-					error = xfs_iflush(ip, XFS_IFLUSH_DELWRI);
-				else
-					xfs_ifunlock(ip);
-			}
-		}
-		xfs_iput(ip, lock_flags);
+		ip = xfs_inode_ag_lookup(mp, pag, &first_index, tag);
+		if (!ip)
+			break;
 
+		error = execute(ip, pag, flags);
+		if (error == EAGAIN) {
+			skipped++;
+			continue;
+		}
 		if (error)
 			last_error = error;
 		/*
 		 * bail out if the filesystem is corrupted.
 		 */
 		if (error == EFSCORRUPTED)
-			return XFS_ERROR(error);
+			break;
 
-	} while (nr_found);
+	} while (1);
+
+	if (skipped) {
+		delay(1);
+		goto restart;
+	}
 
+	xfs_put_perag(mp, pag);
 	return last_error;
 }
 
 int
-xfs_sync_inodes(
-	xfs_mount_t	*mp,
-	int		flags)
+xfs_inode_ag_iterator(
+	struct xfs_mount	*mp,
+	int			(*execute)(struct xfs_inode *ip,
+					   struct xfs_perag *pag, int flags),
+	int			flags,
+	int			tag)
 {
-	int		error;
-	int		last_error;
-	int		i;
-	int		lflags = XFS_LOG_FORCE;
+	int			error = 0;
+	int			last_error = 0;
+	xfs_agnumber_t		ag;
 
-	if (mp->m_flags & XFS_MOUNT_RDONLY)
-		return 0;
-	error = 0;
-	last_error = 0;
+	for (ag = 0; ag < mp->m_sb.sb_agcount; ag++) {
+		if (!mp->m_perag[ag].pag_ici_init)
+			continue;
+		error = xfs_inode_ag_walk(mp, ag, execute, flags, tag);
+		if (error) {
+			last_error = error;
+			if (error == EFSCORRUPTED)
+				break;
+		}
+	}
+	return XFS_ERROR(last_error);
+}
+
+/* must be called with pag_ici_lock held and releases it */
+int
+xfs_sync_inode_valid(
+	struct xfs_inode	*ip,
+	struct xfs_perag	*pag)
+{
+	struct inode		*inode = VFS_I(ip);
+
+	/* nothing to sync during shutdown */
+	if (XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+		read_unlock(&pag->pag_ici_lock);
+		return EFSCORRUPTED;
+	}
+
+	/*
+	 * If we can't get a reference on the inode, it must be in reclaim.
+	 * Leave it for the reclaim code to flush. Also avoid inodes that
+	 * haven't been fully initialised.
+	 */
+	if (!igrab(inode)) {
+		read_unlock(&pag->pag_ici_lock);
+		return ENOENT;
+	}
+	read_unlock(&pag->pag_ici_lock);
+
+	if (is_bad_inode(inode) || xfs_iflags_test(ip, XFS_INEW)) {
+		IRELE(ip);
+		return ENOENT;
+	}
+
+	return 0;
+}
+
+STATIC int
+xfs_sync_inode_data(
+	struct xfs_inode	*ip,
+	struct xfs_perag	*pag,
+	int			flags)
+{
+	struct inode		*inode = VFS_I(ip);
+	struct address_space *mapping = inode->i_mapping;
+	int			error = 0;
+
+	error = xfs_sync_inode_valid(ip, pag);
+	if (error)
+		return error;
+
+	if (!mapping_tagged(mapping, PAGECACHE_TAG_DIRTY))
+		goto out_wait;
+
+	if (!xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED)) {
+		if (flags & SYNC_TRYLOCK)
+			goto out_wait;
+		xfs_ilock(ip, XFS_IOLOCK_SHARED);
+	}
+
+	error = xfs_flush_pages(ip, 0, -1, (flags & SYNC_WAIT) ?
+				0 : XFS_B_ASYNC, FI_NONE);
+	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
 
+ out_wait:
 	if (flags & SYNC_WAIT)
-		lflags |= XFS_LOG_SYNC;
+		xfs_ioend_wait(ip);
+	IRELE(ip);
+	return error;
+}
 
-	for (i = 0; i < mp->m_sb.sb_agcount; i++) {
-		if (!mp->m_perag[i].pag_ici_init)
-			continue;
-		error = xfs_sync_inodes_ag(mp, i, flags);
-		if (error)
-			last_error = error;
-		if (error == EFSCORRUPTED)
-			break;
+STATIC int
+xfs_sync_inode_attr(
+	struct xfs_inode	*ip,
+	struct xfs_perag	*pag,
+	int			flags)
+{
+	int			error = 0;
+
+	error = xfs_sync_inode_valid(ip, pag);
+	if (error)
+		return error;
+
+	xfs_ilock(ip, XFS_ILOCK_SHARED);
+	if (xfs_inode_clean(ip))
+		goto out_unlock;
+	if (!xfs_iflock_nowait(ip)) {
+		if (!(flags & SYNC_WAIT))
+			goto out_unlock;
+		xfs_iflock(ip);
 	}
-	if (flags & SYNC_DELWRI)
-		xfs_log_force(mp, 0, lflags);
 
-	return XFS_ERROR(last_error);
+	if (xfs_inode_clean(ip)) {
+		xfs_ifunlock(ip);
+		goto out_unlock;
+	}
+
+	error = xfs_iflush(ip, (flags & SYNC_WAIT) ?
+			   XFS_IFLUSH_SYNC : XFS_IFLUSH_DELWRI);
+
+ out_unlock:
+	xfs_iunlock(ip, XFS_ILOCK_SHARED);
+	IRELE(ip);
+	return error;
+}
+
+/*
+ * Write out pagecache data for the whole filesystem.
+ */
+int
+xfs_sync_data(
+	struct xfs_mount	*mp,
+	int			flags)
+{
+	int			error;
+
+	ASSERT((flags & ~(SYNC_TRYLOCK|SYNC_WAIT)) == 0);
+
+	error = xfs_inode_ag_iterator(mp, xfs_sync_inode_data, flags,
+				      XFS_ICI_NO_TAG);
+	if (error)
+		return XFS_ERROR(error);
+
+	xfs_log_force(mp, 0,
+		      (flags & SYNC_WAIT) ?
+		       XFS_LOG_FORCE | XFS_LOG_SYNC :
+		       XFS_LOG_FORCE);
+	return 0;
+}
+
+/*
+ * Write out inode metadata (attributes) for the whole filesystem.
+ */
+int
+xfs_sync_attr(
+	struct xfs_mount	*mp,
+	int			flags)
+{
+	ASSERT((flags & ~SYNC_WAIT) == 0);
+
+	return xfs_inode_ag_iterator(mp, xfs_sync_inode_attr, flags,
+				     XFS_ICI_NO_TAG);
 }
 
 STATIC int
@@ -252,7 +353,7 @@ xfs_sync_fsdata(
 	 * If this is xfssyncd() then only sync the superblock if we can
 	 * lock it without sleeping and it is not pinned.
 	 */
-	if (flags & SYNC_BDFLUSH) {
+	if (flags & SYNC_TRYLOCK) {
 		ASSERT(!(flags & SYNC_WAIT));
 
 		bp = xfs_getsb(mp, XFS_BUF_TRYLOCK);
@@ -316,13 +417,13 @@ xfs_quiesce_data(
 	int error;
 
 	/* push non-blocking */
-	xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_BDFLUSH);
-	XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
+	xfs_sync_data(mp, 0);
+	xfs_qm_sync(mp, SYNC_TRYLOCK);
 	xfs_filestream_flush(mp);
 
 	/* push and block */
-	xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_WAIT|SYNC_IOWAIT);
-	XFS_QM_DQSYNC(mp, SYNC_WAIT);
+	xfs_sync_data(mp, SYNC_WAIT);
+	xfs_qm_sync(mp, SYNC_WAIT);
 
 	/* write superblock and hoover up shutdown errors */
 	error = xfs_sync_fsdata(mp, 0);
@@ -341,7 +442,7 @@ xfs_quiesce_fs(
 	int	count = 0, pincount;
 
 	xfs_flush_buftarg(mp->m_ddev_targp, 0);
-	xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
+	xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
 
 	/*
 	 * This loop must run at least twice.  The first instance of the loop
@@ -350,7 +451,7 @@ xfs_quiesce_fs(
 	 * logged before we can write the unmount record.
 	 */
 	do {
-		xfs_sync_inodes(mp, SYNC_ATTR|SYNC_WAIT);
+		xfs_sync_attr(mp, SYNC_WAIT);
 		pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1);
 		if (!pincount) {
 			delay(50);
@@ -433,8 +534,8 @@ xfs_flush_inodes_work(
 	void		*arg)
 {
 	struct inode	*inode = arg;
-	xfs_sync_inodes(mp, SYNC_DELWRI | SYNC_TRYLOCK);
-	xfs_sync_inodes(mp, SYNC_DELWRI | SYNC_TRYLOCK | SYNC_IOWAIT);
+	xfs_sync_data(mp, SYNC_TRYLOCK);
+	xfs_sync_data(mp, SYNC_TRYLOCK | SYNC_WAIT);
 	iput(inode);
 }
 
@@ -465,10 +566,10 @@ xfs_sync_worker(
 
 	if (!(mp->m_flags & XFS_MOUNT_RDONLY)) {
 		xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
-		xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
+		xfs_reclaim_inodes(mp, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
 		/* dgc: errors ignored here */
-		error = XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
-		error = xfs_sync_fsdata(mp, SYNC_BDFLUSH);
+		error = xfs_qm_sync(mp, SYNC_TRYLOCK);
+		error = xfs_sync_fsdata(mp, SYNC_TRYLOCK);
 		if (xfs_log_need_covered(mp))
 			error = xfs_commit_dummy_trans(mp, XFS_LOG_FORCE);
 	}
@@ -569,7 +670,7 @@ xfs_reclaim_inode(
 			xfs_ifunlock(ip);
 			xfs_iunlock(ip, XFS_ILOCK_EXCL);
 		}
-		return 1;
+		return -EAGAIN;
 	}
 	__xfs_iflags_set(ip, XFS_IRECLAIM);
 	spin_unlock(&ip->i_flags_lock);
@@ -654,101 +755,27 @@ xfs_inode_clear_reclaim_tag(
 	xfs_put_perag(mp, pag);
 }
 
-
-STATIC void
-xfs_reclaim_inodes_ag(
-	xfs_mount_t	*mp,
-	int		ag,
-	int		noblock,
-	int		mode)
+STATIC int
+xfs_reclaim_inode_now(
+	struct xfs_inode	*ip,
+	struct xfs_perag	*pag,
+	int			flags)
 {
-	xfs_inode_t	*ip = NULL;
-	xfs_perag_t	*pag = &mp->m_perag[ag];
-	int		nr_found;
-	uint32_t	first_index;
-	int		skipped;
-
-restart:
-	first_index = 0;
-	skipped = 0;
-	do {
-		/*
-		 * use a gang lookup to find the next inode in the tree
-		 * as the tree is sparse and a gang lookup walks to find
-		 * the number of objects requested.
-		 */
-		read_lock(&pag->pag_ici_lock);
-		nr_found = radix_tree_gang_lookup_tag(&pag->pag_ici_root,
-					(void**)&ip, first_index, 1,
-					XFS_ICI_RECLAIM_TAG);
-
-		if (!nr_found) {
-			read_unlock(&pag->pag_ici_lock);
-			break;
-		}
-
-		/*
-		 * Update the index for the next lookup. Catch overflows
-		 * into the next AG range which can occur if we have inodes
-		 * in the last block of the AG and we are currently
-		 * pointing to the last inode.
-		 */
-		first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
-		if (first_index < XFS_INO_TO_AGINO(mp, ip->i_ino)) {
-			read_unlock(&pag->pag_ici_lock);
-			break;
-		}
-
-		/* ignore if already under reclaim */
-		if (xfs_iflags_test(ip, XFS_IRECLAIM)) {
-			read_unlock(&pag->pag_ici_lock);
-			continue;
-		}
-
-		if (noblock) {
-			if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
-				read_unlock(&pag->pag_ici_lock);
-				continue;
-			}
-			if (xfs_ipincount(ip) ||
-			    !xfs_iflock_nowait(ip)) {
-				xfs_iunlock(ip, XFS_ILOCK_EXCL);
-				read_unlock(&pag->pag_ici_lock);
-				continue;
-			}
-		}
+	/* ignore if already under reclaim */
+	if (xfs_iflags_test(ip, XFS_IRECLAIM)) {
 		read_unlock(&pag->pag_ici_lock);
-
-		/*
-		 * hmmm - this is an inode already in reclaim. Do
-		 * we even bother catching it here?
-		 */
-		if (xfs_reclaim_inode(ip, noblock, mode))
-			skipped++;
-	} while (nr_found);
-
-	if (skipped) {
-		delay(1);
-		goto restart;
+		return 0;
 	}
-	return;
+	read_unlock(&pag->pag_ici_lock);
 
+	return xfs_reclaim_inode(ip, 0, flags);
 }
 
 int
 xfs_reclaim_inodes(
 	xfs_mount_t	*mp,
-	int		 noblock,
 	int		mode)
 {
-	int		i;
-
-	for (i = 0; i < mp->m_sb.sb_agcount; i++) {
-		if (!mp->m_perag[i].pag_ici_init)
-			continue;
-		xfs_reclaim_inodes_ag(mp, i, noblock, mode);
-	}
-	return 0;
+	return xfs_inode_ag_iterator(mp, xfs_reclaim_inode_now, mode,
+					XFS_ICI_RECLAIM_TAG);
 }
-
-
diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h
index 308d5bf6dfbd..2a10301c99c7 100644
--- a/fs/xfs/linux-2.6/xfs_sync.h
+++ b/fs/xfs/linux-2.6/xfs_sync.h
@@ -29,17 +29,14 @@ typedef struct xfs_sync_work {
 	struct completion	*w_completion;
 } xfs_sync_work_t;
 
-#define SYNC_ATTR		0x0001	/* sync attributes */
-#define SYNC_DELWRI		0x0002	/* look at delayed writes */
-#define SYNC_WAIT		0x0004	/* wait for i/o to complete */
-#define SYNC_BDFLUSH		0x0008	/* BDFLUSH is calling -- don't block */
-#define SYNC_IOWAIT		0x0010  /* wait for all I/O to complete */
-#define SYNC_TRYLOCK		0x0020  /* only try to lock inodes */
+#define SYNC_WAIT		0x0001	/* wait for i/o to complete */
+#define SYNC_TRYLOCK		0x0002  /* only try to lock inodes */
 
 int xfs_syncd_init(struct xfs_mount *mp);
 void xfs_syncd_stop(struct xfs_mount *mp);
 
-int xfs_sync_inodes(struct xfs_mount *mp, int flags);
+int xfs_sync_attr(struct xfs_mount *mp, int flags);
+int xfs_sync_data(struct xfs_mount *mp, int flags);
 int xfs_sync_fsdata(struct xfs_mount *mp, int flags);
 
 int xfs_quiesce_data(struct xfs_mount *mp);
@@ -48,10 +45,16 @@ void xfs_quiesce_attr(struct xfs_mount *mp);
 void xfs_flush_inodes(struct xfs_inode *ip);
 
 int xfs_reclaim_inode(struct xfs_inode *ip, int locked, int sync_mode);
-int xfs_reclaim_inodes(struct xfs_mount *mp, int noblock, int mode);
+int xfs_reclaim_inodes(struct xfs_mount *mp, int mode);
 
 void xfs_inode_set_reclaim_tag(struct xfs_inode *ip);
 void xfs_inode_clear_reclaim_tag(struct xfs_inode *ip);
 void __xfs_inode_clear_reclaim_tag(struct xfs_mount *mp, struct xfs_perag *pag,
 				struct xfs_inode *ip);
+
+int xfs_sync_inode_valid(struct xfs_inode *ip, struct xfs_perag *pag);
+int xfs_inode_ag_iterator(struct xfs_mount *mp,
+	int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag, int flags),
+	int flags, int tag);
+
 #endif
diff --git a/fs/xfs/linux-2.6/xfs_xattr.c b/fs/xfs/linux-2.6/xfs_xattr.c
index 964621fde6ed..497c7fb75cc1 100644
--- a/fs/xfs/linux-2.6/xfs_xattr.c
+++ b/fs/xfs/linux-2.6/xfs_xattr.c
@@ -29,67 +29,6 @@
 #include <linux/xattr.h>
 
 
-/*
- * ACL handling.  Should eventually be moved into xfs_acl.c
- */
-
-static int
-xfs_decode_acl(const char *name)
-{
-	if (strcmp(name, "posix_acl_access") == 0)
-		return _ACL_TYPE_ACCESS;
-	else if (strcmp(name, "posix_acl_default") == 0)
-		return _ACL_TYPE_DEFAULT;
-	return -EINVAL;
-}
-
-/*
- * Get system extended attributes which at the moment only
- * includes Posix ACLs.
- */
-static int
-xfs_xattr_system_get(struct inode *inode, const char *name,
-		void *buffer, size_t size)
-{
-	int acl;
-
-	acl = xfs_decode_acl(name);
-	if (acl < 0)
-		return acl;
-
-	return xfs_acl_vget(inode, buffer, size, acl);
-}
-
-static int
-xfs_xattr_system_set(struct inode *inode, const char *name,
-		const void *value, size_t size, int flags)
-{
-	int acl;
-
-	acl = xfs_decode_acl(name);
-	if (acl < 0)
-		return acl;
-	if (flags & XATTR_CREATE)
-		return -EINVAL;
-
-	if (!value)
-		return xfs_acl_vremove(inode, acl);
-
-	return xfs_acl_vset(inode, (void *)value, size, acl);
-}
-
-static struct xattr_handler xfs_xattr_system_handler = {
-	.prefix	= XATTR_SYSTEM_PREFIX,
-	.get	= xfs_xattr_system_get,
-	.set	= xfs_xattr_system_set,
-};
-
-
-/*
- * Real xattr handling.  The only difference between the namespaces is
- * a flag passed to the low-level attr code.
- */
-
 static int
 __xfs_xattr_get(struct inode *inode, const char *name,
 		void *value, size_t size, int xflags)
@@ -199,7 +138,9 @@ struct xattr_handler *xfs_xattr_handlers[] = {
 	&xfs_xattr_user_handler,
 	&xfs_xattr_trusted_handler,
 	&xfs_xattr_security_handler,
+#ifdef CONFIG_XFS_POSIX_ACL
 	&xfs_xattr_system_handler,
+#endif
 	NULL
 };
 
@@ -310,7 +251,7 @@ xfs_vn_listxattr(struct dentry *dentry, char *data, size_t size)
 	/*
 	 * Then add the two synthetic ACL attributes.
 	 */
-	if (xfs_acl_vhasacl_access(inode)) {
+	if (posix_acl_access_exists(inode)) {
 		error = list_one_attr(POSIX_ACL_XATTR_ACCESS,
 				strlen(POSIX_ACL_XATTR_ACCESS) + 1,
 				data, size, &context.count);
@@ -318,7 +259,7 @@ xfs_vn_listxattr(struct dentry *dentry, char *data, size_t size)
 			return error;
 	}
 
-	if (xfs_acl_vhasacl_default(inode)) {
+	if (posix_acl_default_exists(inode)) {
 		error = list_one_attr(POSIX_ACL_XATTR_DEFAULT,
 				strlen(POSIX_ACL_XATTR_DEFAULT) + 1,
 				data, size, &context.count);