xfs: don't warn on EAGAIN in inode reclaim
authorDave Chinner <dchinner@redhat.com>
Tue, 13 Apr 2010 05:06:45 +0000 (15:06 +1000)
committerAlex Elder <aelder@sgi.com>
Fri, 16 Apr 2010 18:51:44 +0000 (13:51 -0500)
Any inode reclaim flush that returns EAGAIN will result in the inode
reclaim being attempted again later. There is no need to issue a
warning into the logs about this situation.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
Signed-off-by: Alex Elder <aelder@sgi.com>
fs/xfs/linux-2.6/xfs_sync.c

index 05cd853..fd96982 100644 (file)
@@ -820,10 +820,10 @@ xfs_reclaim_inode(
         * call into reclaim to find it in a clean state instead of waiting for
         * it now. We also don't return errors here - if the error is transient
         * then the next reclaim pass will flush the inode, and if the error
-        * is permanent then the next sync reclaim will relcaim the inode and
+        * is permanent then the next sync reclaim will reclaim the inode and
         * pass on the error.
         */
-       if (error && !XFS_FORCED_SHUTDOWN(ip->i_mount)) {
+       if (error && error != EAGAIN && !XFS_FORCED_SHUTDOWN(ip->i_mount)) {
                xfs_fs_cmn_err(CE_WARN, ip->i_mount,
                        "inode 0x%llx background reclaim flush failed with %d",
                        (long long)ip->i_ino, error);