10 files changed, 102 insertions, 136 deletions

diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c
index 115ac6919533..b6c4b3795c4a 100644
--- a/fs/xfs/linux-2.6/xfs_acl.c
+++ b/fs/xfs/linux-2.6/xfs_acl.c
@@ -114,6 +114,8 @@ xfs_get_acl(struct inode *inode, int type)
 	if (acl != ACL_NOT_CACHED)
 		return acl;
 
+	trace_xfs_get_acl(ip);
+
 	switch (type) {
 	case ACL_TYPE_ACCESS:
 		ea_name = SGI_ACL_FILE;
@@ -218,42 +220,8 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 	return error;
 }
 
-int
-xfs_check_acl(struct inode *inode, int mask, unsigned int flags)
-{
-	struct xfs_inode *ip;
-	struct posix_acl *acl;
-	int error = -EAGAIN;
-
-	ip = XFS_I(inode);
-	trace_xfs_check_acl(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;
-
-	if (flags & IPERM_FLAG_RCU) {
-		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
-			return -ECHILD;
-		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)
+xfs_set_mode(struct inode *inode, umode_t mode)
 {
 	int error = 0;
 
@@ -297,29 +265,23 @@ posix_acl_default_exists(struct inode *inode)
  * 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)
+xfs_inherit_acl(struct inode *inode, struct posix_acl *acl)
 {
-	struct posix_acl *clone;
-	mode_t mode;
+	umode_t mode = inode->i_mode;
 	int error = 0, inherit = 0;
 
 	if (S_ISDIR(inode->i_mode)) {
-		error = xfs_set_acl(inode, ACL_TYPE_DEFAULT, default_acl);
+		error = xfs_set_acl(inode, ACL_TYPE_DEFAULT, acl);
 		if (error)
-			return error;
+			goto out;
 	}
 
-	clone = posix_acl_clone(default_acl, GFP_KERNEL);
-	if (!clone)
-		return -ENOMEM;
-
-	mode = inode->i_mode;
-	error = posix_acl_create_masq(clone, &mode);
+	error = posix_acl_create(&acl, GFP_KERNEL, &mode);
 	if (error < 0)
-		goto out_release_clone;
+		return error;
 
 	/*
-	 * If posix_acl_create_masq returns a positive value we need to
+	 * If posix_acl_create 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.
 	 */
@@ -328,20 +290,20 @@ xfs_inherit_acl(struct inode *inode, struct posix_acl *default_acl)
 
 	error = xfs_set_mode(inode, mode);
 	if (error)
-		goto out_release_clone;
+		goto out;
 
 	if (inherit)
-		error = xfs_set_acl(inode, ACL_TYPE_ACCESS, clone);
+		error = xfs_set_acl(inode, ACL_TYPE_ACCESS, acl);
 
- out_release_clone:
-	posix_acl_release(clone);
+out:
+	posix_acl_release(acl);
 	return error;
 }
 
 int
 xfs_acl_chmod(struct inode *inode)
 {
-	struct posix_acl *acl, *clone;
+	struct posix_acl *acl;
 	int error;
 
 	if (S_ISLNK(inode->i_mode))
@@ -351,16 +313,12 @@ xfs_acl_chmod(struct inode *inode)
 	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);
+	error = posix_acl_chmod(&acl, GFP_KERNEL, inode->i_mode);
+	if (error)
+		return error;
 
-	posix_acl_release(clone);
+	error = xfs_set_acl(inode, ACL_TYPE_ACCESS, acl);
+	posix_acl_release(acl);
 	return error;
 }
 
@@ -423,7 +381,7 @@ xfs_xattr_acl_set(struct dentry *dentry, const char *name,
 		goto out_release;
 
 	if (type == ACL_TYPE_ACCESS) {
-		mode_t mode = inode->i_mode;
+		umode_t mode = inode->i_mode;
 		error = posix_acl_equiv_mode(acl, &mode);
 
 		if (error <= 0) {
diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index 26384fe3f26d..63e971e2b837 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -1329,6 +1329,9 @@ xfs_end_io_direct_write(
 	} else {
 		xfs_finish_ioend_sync(ioend);
 	}
+
+	/* XXX: probably should move into the real I/O completion handler */
+	inode_dio_done(ioend->io_inode);
 }
 
 STATIC ssize_t
diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c
index 6bddce40de72..c57836dc778f 100644
--- a/fs/xfs/linux-2.6/xfs_buf.c
+++ b/fs/xfs/linux-2.6/xfs_buf.c
@@ -1223,6 +1223,9 @@ _xfs_buf_ioapply(
 		rw = READ;
 	}
 
+	/* we only use the buffer cache for meta-data */
+	rw |= REQ_META;
+
 next_chunk:
 	atomic_inc(&bp->b_io_remaining);
 	nr_pages = BIO_MAX_SECTORS >> (PAGE_SHIFT - BBSHIFT);
diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c
index 8073f61efb8e..7f7b42469ea7 100644
--- a/fs/xfs/linux-2.6/xfs_file.c
+++ b/fs/xfs/linux-2.6/xfs_file.c
@@ -127,6 +127,8 @@ xfs_iozero(
 STATIC int
 xfs_file_fsync(
 	struct file		*file,
+	loff_t			start,
+	loff_t			end,
 	int			datasync)
 {
 	struct inode		*inode = file->f_mapping->host;
@@ -138,12 +140,18 @@ xfs_file_fsync(
 
 	trace_xfs_file_fsync(ip);
 
+	error = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (error)
+		return error;
+
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -XFS_ERROR(EIO);
 
 	xfs_iflags_clear(ip, XFS_ITRUNCATED);
 
+	xfs_ilock(ip, XFS_IOLOCK_SHARED);
 	xfs_ioend_wait(ip);
+	xfs_iunlock(ip, XFS_IOLOCK_SHARED);
 
 	if (mp->m_flags & XFS_MOUNT_BARRIER) {
 		/*
@@ -875,18 +883,14 @@ xfs_file_aio_write(
 	/* Handle various SYNC-type writes */
 	if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) {
 		loff_t end = pos + ret - 1;
-		int error, error2;
+		int error;
 
 		xfs_rw_iunlock(ip, iolock);
-		error = filemap_write_and_wait_range(mapping, pos, end);
+		error = xfs_file_fsync(file, pos, end,
+				      (file->f_flags & __O_SYNC) ? 0 : 1);
 		xfs_rw_ilock(ip, iolock);
-
-		error2 = -xfs_file_fsync(file,
-					 (file->f_flags & __O_SYNC) ? 0 : 1);
 		if (error)
 			ret = error;
-		else if (error2)
-			ret = error2;
 	}
 
 out_unlock:
diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c
index acca2c5ca3fa..f7ce7debe14c 100644
--- a/fs/xfs/linux-2.6/xfs_ioctl.c
+++ b/fs/xfs/linux-2.6/xfs_ioctl.c
@@ -265,7 +265,7 @@ xfs_open_by_handle(
 		return PTR_ERR(filp);
 	}
 
-	if (inode->i_mode & S_IFREG) {
+	if (S_ISREG(inode->i_mode)) {
 		filp->f_flags |= O_NOATIME;
 		filp->f_mode |= FMODE_NOCMTIME;
 	}
@@ -850,14 +850,14 @@ xfs_set_diflags(
 		di_flags |= XFS_DIFLAG_NODEFRAG;
 	if (xflags & XFS_XFLAG_FILESTREAM)
 		di_flags |= XFS_DIFLAG_FILESTREAM;
-	if ((ip->i_d.di_mode & S_IFMT) == S_IFDIR) {
+	if (S_ISDIR(ip->i_d.di_mode)) {
 		if (xflags & XFS_XFLAG_RTINHERIT)
 			di_flags |= XFS_DIFLAG_RTINHERIT;
 		if (xflags & XFS_XFLAG_NOSYMLINKS)
 			di_flags |= XFS_DIFLAG_NOSYMLINKS;
 		if (xflags & XFS_XFLAG_EXTSZINHERIT)
 			di_flags |= XFS_DIFLAG_EXTSZINHERIT;
-	} else if ((ip->i_d.di_mode & S_IFMT) == S_IFREG) {
+	} else if (S_ISREG(ip->i_d.di_mode)) {
 		if (xflags & XFS_XFLAG_REALTIME)
 			di_flags |= XFS_DIFLAG_REALTIME;
 		if (xflags & XFS_XFLAG_EXTSIZE)
diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c
index 501e4f630548..b9c172b3fbbe 100644
--- a/fs/xfs/linux-2.6/xfs_iops.c
+++ b/fs/xfs/linux-2.6/xfs_iops.c
@@ -202,9 +202,9 @@ xfs_vn_mknod(
 
 	if (default_acl) {
 		error = -xfs_inherit_acl(inode, default_acl);
+		default_acl = NULL;
 		if (unlikely(error))
 			goto out_cleanup_inode;
-		posix_acl_release(default_acl);
 	}
 
 
@@ -1022,7 +1022,7 @@ xfs_vn_fiemap(
 }
 
 static const struct inode_operations xfs_inode_operations = {
-	.check_acl		= xfs_check_acl,
+	.get_acl		= xfs_get_acl,
 	.getattr		= xfs_vn_getattr,
 	.setattr		= xfs_vn_setattr,
 	.setxattr		= generic_setxattr,
@@ -1048,7 +1048,7 @@ static const struct inode_operations xfs_dir_inode_operations = {
 	.rmdir			= xfs_vn_unlink,
 	.mknod			= xfs_vn_mknod,
 	.rename			= xfs_vn_rename,
-	.check_acl		= xfs_check_acl,
+	.get_acl		= xfs_get_acl,
 	.getattr		= xfs_vn_getattr,
 	.setattr		= xfs_vn_setattr,
 	.setxattr		= generic_setxattr,
@@ -1073,7 +1073,7 @@ static const struct inode_operations xfs_dir_ci_inode_operations = {
 	.rmdir			= xfs_vn_unlink,
 	.mknod			= xfs_vn_mknod,
 	.rename			= xfs_vn_rename,
-	.check_acl		= xfs_check_acl,
+	.get_acl		= xfs_get_acl,
 	.getattr		= xfs_vn_getattr,
 	.setattr		= xfs_vn_setattr,
 	.setxattr		= generic_setxattr,
@@ -1086,7 +1086,7 @@ static const struct inode_operations xfs_symlink_inode_operations = {
 	.readlink		= generic_readlink,
 	.follow_link		= xfs_vn_follow_link,
 	.put_link		= xfs_vn_put_link,
-	.check_acl		= xfs_check_acl,
+	.get_acl		= xfs_get_acl,
 	.getattr		= xfs_vn_getattr,
 	.setattr		= xfs_vn_setattr,
 	.setxattr		= generic_setxattr,
@@ -1194,6 +1194,15 @@ xfs_setup_inode(
 		break;
 	}
 
+	/*
+	 * If there is no attribute fork no ACL can exist on this inode,
+	 * and it can't have any file capabilities attached to it either.
+	 */
+	if (!XFS_IFORK_Q(ip)) {
+		inode_has_no_xattr(inode);
+		cache_no_acl(inode);
+	}
+
 	xfs_iflags_clear(ip, XFS_INEW);
 	barrier();
 
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index 25fd2cd6c8b0..9a72dda58bd0 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -1024,11 +1024,6 @@ xfs_fs_put_super(
 {
 	struct xfs_mount	*mp = XFS_M(sb);
 
-	/*
-	 * Unregister the memory shrinker before we tear down the mount
-	 * structure so we don't have memory reclaim racing with us here.
-	 */
-	xfs_inode_shrinker_unregister(mp);
 	xfs_syncd_stop(mp);
 
 	/*
@@ -1411,8 +1406,6 @@ xfs_fs_fill_super(
 	sb->s_time_gran = 1;
 	set_posix_acl_flag(sb);
 
-	xfs_inode_shrinker_register(mp);
-
 	error = xfs_mountfs(mp);
 	if (error)
 		goto out_filestream_unmount;
@@ -1439,7 +1432,6 @@ xfs_fs_fill_super(
 	return 0;
 
  out_filestream_unmount:
-	xfs_inode_shrinker_unregister(mp);
 	xfs_filestream_unmount(mp);
  out_free_sb:
 	xfs_freesb(mp);
@@ -1458,8 +1450,6 @@ xfs_fs_fill_super(
  out_syncd_stop:
 	xfs_syncd_stop(mp);
  out_unmount:
-	xfs_inode_shrinker_unregister(mp);
-
 	/*
 	 * Blow away any referenced inode in the filestreams cache.
 	 * This can and will cause log traffic as inodes go inactive
@@ -1483,6 +1473,21 @@ xfs_fs_mount(
 	return mount_bdev(fs_type, flags, dev_name, data, xfs_fs_fill_super);
 }
 
+static int
+xfs_fs_nr_cached_objects(
+	struct super_block	*sb)
+{
+	return xfs_reclaim_inodes_count(XFS_M(sb));
+}
+
+static void
+xfs_fs_free_cached_objects(
+	struct super_block	*sb,
+	int			nr_to_scan)
+{
+	xfs_reclaim_inodes_nr(XFS_M(sb), nr_to_scan);
+}
+
 static const struct super_operations xfs_super_operations = {
 	.alloc_inode		= xfs_fs_alloc_inode,
 	.destroy_inode		= xfs_fs_destroy_inode,
@@ -1496,6 +1501,8 @@ static const struct super_operations xfs_super_operations = {
 	.statfs			= xfs_fs_statfs,
 	.remount_fs		= xfs_fs_remount,
 	.show_options		= xfs_fs_show_options,
+	.nr_cached_objects	= xfs_fs_nr_cached_objects,
+	.free_cached_objects	= xfs_fs_free_cached_objects,
 };
 
 static struct file_system_type xfs_fs_type = {
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index a8500e92ae72..4604f90f86a3 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -179,6 +179,8 @@ restart:
 		if (error == EFSCORRUPTED)
 			break;
 
+		cond_resched();
+
 	} while (nr_found && !done);
 
 	if (skipped) {
@@ -984,6 +986,8 @@ restart:
 
 			*nr_to_scan -= XFS_LOOKUP_BATCH;
 
+			cond_resched();
+
 		} while (nr_found && !done && *nr_to_scan > 0);
 
 		if (trylock && !done)
@@ -1001,7 +1005,7 @@ restart:
 	 * ensure that when we get more reclaimers than AGs we block rather
 	 * than spin trying to execute reclaim.
 	 */
-	if (trylock && skipped && *nr_to_scan > 0) {
+	if (skipped && (flags & SYNC_WAIT) && *nr_to_scan > 0) {
 		trylock = 0;
 		goto restart;
 	}
@@ -1019,44 +1023,38 @@ xfs_reclaim_inodes(
 }
 
 /*
- * Inode cache shrinker.
+ * Scan a certain number of inodes for reclaim.
  *
  * When called we make sure that there is a background (fast) inode reclaim in
- * progress, while we will throttle the speed of reclaim via doiing synchronous
+ * progress, while we will throttle the speed of reclaim via doing synchronous
  * reclaim of inodes. That means if we come across dirty inodes, we wait for
  * them to be cleaned, which we hope will not be very long due to the
  * background walker having already kicked the IO off on those dirty inodes.
  */
-static int
-xfs_reclaim_inode_shrink(
-	struct shrinker	*shrink,
-	struct shrink_control *sc)
+void
+xfs_reclaim_inodes_nr(
+	struct xfs_mount	*mp,
+	int			nr_to_scan)
 {
-	struct xfs_mount *mp;
-	struct xfs_perag *pag;
-	xfs_agnumber_t	ag;
-	int		reclaimable;
-	int nr_to_scan = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
-
-	mp = container_of(shrink, struct xfs_mount, m_inode_shrink);
-	if (nr_to_scan) {
-		/* kick background reclaimer and push the AIL */
-		xfs_syncd_queue_reclaim(mp);
-		xfs_ail_push_all(mp->m_ail);
+	/* kick background reclaimer and push the AIL */
+	xfs_syncd_queue_reclaim(mp);
+	xfs_ail_push_all(mp->m_ail);
 
-		if (!(gfp_mask & __GFP_FS))
-			return -1;
+	xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT, &nr_to_scan);
+}
 
-		xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT,
-					&nr_to_scan);
-		/* terminate if we don't exhaust the scan */
-		if (nr_to_scan > 0)
-			return -1;
-       }
+/*
+ * Return the number of reclaimable inodes in the filesystem for
+ * the shrinker to determine how much to reclaim.
+ */
+int
+xfs_reclaim_inodes_count(
+	struct xfs_mount	*mp)
+{
+	struct xfs_perag	*pag;
+	xfs_agnumber_t		ag = 0;
+	int			reclaimable = 0;
 
-	reclaimable = 0;
-	ag = 0;
 	while ((pag = xfs_perag_get_tag(mp, ag, XFS_ICI_RECLAIM_TAG))) {
 		ag = pag->pag_agno + 1;
 		reclaimable += pag->pag_ici_reclaimable;
@@ -1065,18 +1063,3 @@ xfs_reclaim_inode_shrink(
 	return reclaimable;
 }
 
-void
-xfs_inode_shrinker_register(
-	struct xfs_mount	*mp)
-{
-	mp->m_inode_shrink.shrink = xfs_reclaim_inode_shrink;
-	mp->m_inode_shrink.seeks = DEFAULT_SEEKS;
-	register_shrinker(&mp->m_inode_shrink);
-}
-
-void
-xfs_inode_shrinker_unregister(
-	struct xfs_mount	*mp)
-{
-	unregister_shrinker(&mp->m_inode_shrink);
-}
diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h
index e914fd621746..941202e7ac6e 100644
--- a/fs/xfs/linux-2.6/xfs_sync.h
+++ b/fs/xfs/linux-2.6/xfs_sync.h
@@ -35,6 +35,8 @@ void xfs_quiesce_attr(struct xfs_mount *mp);
 void xfs_flush_inodes(struct xfs_inode *ip);
 
 int xfs_reclaim_inodes(struct xfs_mount *mp, int mode);
+int xfs_reclaim_inodes_count(struct xfs_mount *mp);
+void xfs_reclaim_inodes_nr(struct xfs_mount *mp, int nr_to_scan);
 
 void xfs_inode_set_reclaim_tag(struct xfs_inode *ip);
 void __xfs_inode_set_reclaim_tag(struct xfs_perag *pag, struct xfs_inode *ip);
@@ -46,7 +48,4 @@ int xfs_inode_ag_iterator(struct xfs_mount *mp,
 	int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag, int flags),
 	int flags);
 
-void xfs_inode_shrinker_register(struct xfs_mount *mp);
-void xfs_inode_shrinker_unregister(struct xfs_mount *mp);
-
 #endif
diff --git a/fs/xfs/linux-2.6/xfs_trace.h b/fs/xfs/linux-2.6/xfs_trace.h
index fda0708ef2ea..690fc7a7bd72 100644
--- a/fs/xfs/linux-2.6/xfs_trace.h
+++ b/fs/xfs/linux-2.6/xfs_trace.h
@@ -571,7 +571,7 @@ DEFINE_INODE_EVENT(xfs_alloc_file_space);
 DEFINE_INODE_EVENT(xfs_free_file_space);
 DEFINE_INODE_EVENT(xfs_readdir);
 #ifdef CONFIG_XFS_POSIX_ACL
-DEFINE_INODE_EVENT(xfs_check_acl);
+DEFINE_INODE_EVENT(xfs_get_acl);
 #endif
 DEFINE_INODE_EVENT(xfs_vm_bmap);
 DEFINE_INODE_EVENT(xfs_file_ioctl);