xfs: merge xfs_itobp into xfs_imap_to_bp

[deliverable/linux.git] / fs / xfs / xfs_sync.c

diff --git a/fs/xfs/xfs_sync.c b/fs/xfs/xfs_sync.c
index 1e9ee064dbb28c7cb491d4c41dce53007eebc7e9..e61fc15190732a6d9a512ed9ad049bef93dfe95d 100644 (file)
--- a/fs/xfs/xfs_sync.c
+++ b/fs/xfs/xfs_sync.c
@@ -712,8 +712,8 @@ restart:
         * Note that xfs_iflush will never block on the inode buffer lock, as
         * xfs_ifree_cluster() can lock the inode buffer before it locks the
         * ip->i_lock, and we are doing the exact opposite here.  As a result,
-        * doing a blocking xfs_itobp() to get the cluster buffer would result
-        * in an ABBA deadlock with xfs_ifree_cluster().
+        * doing a blocking xfs_imap_to_bp() to get the cluster buffer would
+        * result in an ABBA deadlock with xfs_ifree_cluster().
         *
         * As xfs_ifree_cluser() must gather all inodes that are active in the
         * cache to mark them stale, if we hit this case we don't actually want