xfs: reflink can skip remap existing mappings
authorDarrick J. Wong <darrick.wong@oracle.com>
Mon, 29 Jun 2020 21:47:19 +0000 (14:47 -0700)
committerDarrick J. Wong <darrick.wong@oracle.com>
Mon, 6 Jul 2020 17:46:57 +0000 (10:46 -0700)
If the source and destination map are identical, we can skip the remap
step to save some time.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
fs/xfs/xfs_reflink.c

index 35a17ca..0a36816 100644 (file)
@@ -1035,6 +1035,22 @@ xfs_reflink_remap_extent(
 
        trace_xfs_reflink_remap_extent_dest(ip, &smap);
 
+       /*
+        * Two extents mapped to the same physical block must not have
+        * different states; that's filesystem corruption.  Move on to the next
+        * extent if they're both holes or both the same physical extent.
+        */
+       if (dmap->br_startblock == smap.br_startblock) {
+               if (dmap->br_state != smap.br_state)
+                       error = -EFSCORRUPTED;
+               goto out_cancel;
+       }
+
+       /* If both extents are unwritten, leave them alone. */
+       if (dmap->br_state == XFS_EXT_UNWRITTEN &&
+           smap.br_state == XFS_EXT_UNWRITTEN)
+               goto out_cancel;
+
        /* No reflinking if the AG of the dest mapping is low on space. */
        if (dmap_written) {
                error = xfs_reflink_ag_has_free_space(mp,