From: Eric Sandeen <sandeen@redhat.com>
Date: Wed, 25 Mar 2009 15:57:14 -0500
Subject: [fs] xfs: misc upstream fixes
Message-id: 49CA9AAA.60404@redhat.com
O-Subject: [PATCH 10/10] misc upstream fixes
Bugzilla: 470845
RH-Acked-by: Josef Bacik <josef@redhat.com>

Misc upstream fixes not in 2.6.28.x.  Includes:

commit 49739140e57a65114d9e1976c4c158d2145595fb
Author: Christoph Hellwig <hch@lst.de>
    xfs: fix bad_features2 fixups for the root filesystem

commit 43f3f057c56d030546145696627f13f95735be95
Author: Felix Blyakher <felixb@sgi.com>
    [XFS] Warn on transaction in flight on read-only remount

commit 2e6560929d8ab4b650fecc3a87013852b34f0922
Author: Dave Chinner <david@fromorbit.com>
    [XFS] fix error inversion problems with data flushing

diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.c b/fs/xfs/linux-2.6/xfs_fs_subr.c
index 36caa6d..5aeb777 100644
--- a/fs/xfs/linux-2.6/xfs_fs_subr.c
+++ b/fs/xfs/linux-2.6/xfs_fs_subr.c
@@ -24,6 +24,10 @@ int  fs_noerr(void) { return 0; }
 int  fs_nosys(void) { return ENOSYS; }
 void fs_noval(void) { return; }
 
+/*
+ * note: all filemap functions return negative error codes. These
+ * need to be inverted before returning to the xfs core functions.
+ */
 void
 xfs_tosspages(
 	xfs_inode_t	*ip,
@@ -53,7 +57,7 @@ xfs_flushinval_pages(
 		if (!ret)
 			truncate_inode_pages(mapping, first);
 	}
-	return ret;
+	return -ret;
 }
 
 int
@@ -72,10 +76,23 @@ xfs_flush_pages(
 		xfs_iflags_clear(ip, XFS_ITRUNCATED);
 		ret = filemap_fdatawrite(mapping);
 		if (flags & XFS_B_ASYNC)
-			return ret;
+			return -ret;
 		ret2 = filemap_fdatawait(mapping);
 		if (!ret)
 			ret = ret2;
 	}
-	return ret;
+	return -ret;
+}
+
+int
+xfs_wait_on_pages(
+	xfs_inode_t	*ip,
+	xfs_off_t	first,
+	xfs_off_t	last)
+{
+	struct address_space *mapping = VFS_I(ip)->i_mapping;
+
+	if (mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK))
+		return -filemap_fdatawait(mapping);
+	return 0;
 }
diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c
index 617788f..2bcb148 100644
--- a/fs/xfs/linux-2.6/xfs_lrw.c
+++ b/fs/xfs/linux-2.6/xfs_lrw.c
@@ -243,7 +243,7 @@ xfs_read(
 
 	if (unlikely(ioflags & IO_ISDIRECT)) {
 		if (inode->i_mapping->nrpages)
-			ret = xfs_flushinval_pages(ip, (*offset & PAGE_CACHE_MASK),
+			ret = -xfs_flushinval_pages(ip, (*offset & PAGE_CACHE_MASK),
 						    -1, FI_REMAPF_LOCKED);
 		mutex_unlock(&inode->i_mutex);
 		if (ret) {
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index a7b60f2..ca3a593 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -901,21 +901,26 @@ xfs_fs_write_inode(
 	struct inode		*inode,
 	int			sync)
 {
+	struct xfs_inode	*ip = XFS_I(inode);
 	int			error = 0;
 	int			flags = 0;
 
-	xfs_itrace_entry(XFS_I(inode));
+	xfs_itrace_entry(ip);
 	if (sync) {
-		filemap_fdatawait(inode->i_mapping);
+		error = xfs_wait_on_pages(ip, 0, -1);
+		if (error)
+			goto out_error;
 		flags |= FLUSH_SYNC;
 	}
-	error = xfs_inode_flush(XFS_I(inode), flags);
+	error = xfs_inode_flush(ip, flags);
+
+out_error:
 	/*
 	 * if we failed to write out the inode then mark
 	 * it dirty again so we'll try again later.
 	 */
 	if (error)
-		mark_inode_dirty_sync(inode);
+		xfs_mark_inode_dirty_sync(ip);
 
 	return -error;
 }
@@ -1280,6 +1285,7 @@ xfs_fs_remount(
 	struct xfs_mount	*mp = XFS_M(sb);
 	substring_t		args[MAX_OPT_ARGS];
 	char			*p;
+	int			error;
 
 	while ((p = strsep(&options, ",")) != NULL) {
 		int token;
@@ -1330,11 +1336,25 @@ xfs_fs_remount(
 		}
 	}
 
-	/* rw/ro -> rw */
+	/* ro -> rw */
 	if ((mp->m_flags & XFS_MOUNT_RDONLY) && !(*flags & MS_RDONLY)) {
 		mp->m_flags &= ~XFS_MOUNT_RDONLY;
 		if (mp->m_flags & XFS_MOUNT_BARRIER)
 			xfs_mountfs_check_barriers(mp);
+
+		/*
+		 * If this is the first remount to writeable state we
+		 * might have some superblock changes to update.
+		 */
+		if (mp->m_update_flags) {
+			error = xfs_mount_log_sb(mp, mp->m_update_flags);
+			if (error) {
+				cmn_err(CE_WARN,
+					"XFS: failed to write sb changes");
+				return error;
+			}
+			mp->m_update_flags = 0;
+		}
 	}
 
 	/* rw -> ro */
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index 0c1a768..cda71c0 100644
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -45,7 +45,6 @@
 #include "xfs_fsops.h"
 #include "xfs_utils.h"
 
-STATIC int	xfs_mount_log_sb(xfs_mount_t *, __int64_t);
 STATIC int	xfs_uuid_mount(xfs_mount_t *);
 STATIC void	xfs_unmountfs_wait(xfs_mount_t *);
 
@@ -690,7 +689,7 @@ xfs_initialize_perag_data(xfs_mount_t *mp, xfs_agnumber_t agcount)
  * Update alignment values based on mount options and sb values
  */
 STATIC int
-xfs_update_alignment(xfs_mount_t *mp, __uint64_t *update_flags)
+xfs_update_alignment(xfs_mount_t *mp)
 {
 	xfs_sb_t	*sbp = &(mp->m_sb);
 
@@ -744,11 +743,11 @@ xfs_update_alignment(xfs_mount_t *mp, __uint64_t *update_flags)
 		if (xfs_sb_version_hasdalign(sbp)) {
 			if (sbp->sb_unit != mp->m_dalign) {
 				sbp->sb_unit = mp->m_dalign;
-				*update_flags |= XFS_SB_UNIT;
+				mp->m_update_flags |= XFS_SB_UNIT;
 			}
 			if (sbp->sb_width != mp->m_swidth) {
 				sbp->sb_width = mp->m_swidth;
-				*update_flags |= XFS_SB_WIDTH;
+				mp->m_update_flags |= XFS_SB_WIDTH;
 			}
 		}
 	} else if ((mp->m_flags & XFS_MOUNT_NOALIGN) != XFS_MOUNT_NOALIGN &&
@@ -913,7 +912,6 @@ xfs_mountfs(
 	xfs_sb_t	*sbp = &(mp->m_sb);
 	xfs_inode_t	*rip;
 	__uint64_t	resblks;
-	__int64_t	update_flags = 0LL;
 	uint		quotamount, quotaflags;
 	int		uuid_mounted = 0;
 	int		error = 0;
@@ -941,7 +939,7 @@ xfs_mountfs(
 			"XFS: correcting sb_features alignment problem");
 		sbp->sb_features2 |= sbp->sb_bad_features2;
 		sbp->sb_bad_features2 = sbp->sb_features2;
-		update_flags |= XFS_SB_FEATURES2 | XFS_SB_BAD_FEATURES2;
+		mp->m_update_flags |= XFS_SB_FEATURES2 | XFS_SB_BAD_FEATURES2;
 
 		/*
 		 * Re-check for ATTR2 in case it was found in bad_features2
@@ -955,11 +953,11 @@ xfs_mountfs(
 	if (xfs_sb_version_hasattr2(&mp->m_sb) &&
 	   (mp->m_flags & XFS_MOUNT_NOATTR2)) {
 		xfs_sb_version_removeattr2(&mp->m_sb);
-		update_flags |= XFS_SB_FEATURES2;
+		mp->m_update_flags |= XFS_SB_FEATURES2;
 
 		/* update sb_versionnum for the clearing of the morebits */
 		if (!sbp->sb_features2)
-			update_flags |= XFS_SB_VERSIONNUM;
+			mp->m_update_flags |= XFS_SB_VERSIONNUM;
 	}
 
 	/*
@@ -968,7 +966,7 @@ xfs_mountfs(
 	 * allocator alignment is within an ag, therefore ag has
 	 * to be aligned at stripe boundary.
 	 */
-	error = xfs_update_alignment(mp, &update_flags);
+	error = xfs_update_alignment(mp);
 	if (error)
 		goto error1;
 
@@ -1145,10 +1143,12 @@ xfs_mountfs(
 	}
 
 	/*
-	 * If fs is not mounted readonly, then update the superblock changes.
+	 * If this is a read-only mount defer the superblock updates until
+	 * the next remount into writeable mode.  Otherwise we would never
+	 * perform the update e.g. for the root filesystem.
 	 */
-	if (update_flags && !(mp->m_flags & XFS_MOUNT_RDONLY)) {
-		error = xfs_mount_log_sb(mp, update_flags);
+	if (mp->m_update_flags && !(mp->m_flags & XFS_MOUNT_RDONLY)) {
+		error = xfs_mount_log_sb(mp, mp->m_update_flags);
 		if (error) {
 			cmn_err(CE_WARN, "XFS: failed to write sb changes");
 			goto error4;
@@ -1849,7 +1849,7 @@ xfs_uuid_mount(
  * be altered by the mount options, as well as any potential sb_features2
  * fixup. Only the first superblock is updated.
  */
-STATIC int
+int
 xfs_mount_log_sb(
 	xfs_mount_t	*mp,
 	__int64_t	fields)
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index f3c1024..a7e0864 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -341,6 +341,8 @@ typedef struct xfs_mount {
 	spinlock_t		m_sync_lock;	/* work item list lock */
 	int			m_sync_seq;	/* sync thread generation no. */
 	wait_queue_head_t	m_wait_single_sync_task;
+	__int64_t		m_update_flags;	/* sb flags we need to update
+						   on the next remount,rw */
 } xfs_mount_t;
 
 /*
@@ -521,6 +523,7 @@ extern int	xfs_mod_incore_sb_unlocked(xfs_mount_t *, xfs_sb_field_t,
 			int64_t, int);
 extern int	xfs_mod_incore_sb_batch(xfs_mount_t *, xfs_mod_sb_t *,
 			uint, int);
+extern int	xfs_mount_log_sb(xfs_mount_t *, __int64_t);
 extern struct xfs_buf *xfs_getsb(xfs_mount_t *, int);
 extern int	xfs_readsb(xfs_mount_t *, int);
 extern void	xfs_freesb(xfs_mount_t *);
diff --git a/fs/xfs/xfs_vfsops.c b/fs/xfs/xfs_vfsops.c
index 439dd39..4da2583 100644
--- a/fs/xfs/xfs_vfsops.c
+++ b/fs/xfs/xfs_vfsops.c
@@ -102,7 +102,11 @@ xfs_attr_quiesce(
 	/* flush inodes and push all remaining buffers out to disk */
 	xfs_quiesce_fs(mp);
 
-	ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0);
+	/*
+	 * Just warn here till VFS can correctly support
+	 * read-only remount without racing.
+	 */
+	WARN_ON(atomic_read(&mp->m_active_trans) != 0);
 
 	/* Push the superblock and write an unmount record */
 	error = xfs_log_sbcount(mp, 1);
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index 3bc89c8..2ba3789 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -713,7 +713,7 @@ xfs_fsync(
 		return XFS_ERROR(EIO);
 
 	/* capture size updates in I/O completion before writing the inode. */
-	error = filemap_fdatawait(VFS_I(ip)->i_mapping);
+	error = xfs_wait_on_pages(ip, 0, -1);
 	if (error)
 		return XFS_ERROR(error);
 
diff --git a/fs/xfs/xfs_vnodeops.h b/fs/xfs/xfs_vnodeops.h
index e932a96..f9cd376 100644
--- a/fs/xfs/xfs_vnodeops.h
+++ b/fs/xfs/xfs_vnodeops.h
@@ -78,5 +78,6 @@ int xfs_flushinval_pages(struct xfs_inode *ip, xfs_off_t first,
 		xfs_off_t last, int fiopt);
 int xfs_flush_pages(struct xfs_inode *ip, xfs_off_t first,
 		xfs_off_t last, uint64_t flags, int fiopt);
+int xfs_wait_on_pages(struct xfs_inode *ip, xfs_off_t first, xfs_off_t last);
 
 #endif /* _XFS_VNODEOPS_H */