btrfs/085: check flakey device instead of backend device
authorNaohiro Aota <naohiro.aota@wdc.com>
Wed, 20 Nov 2019 06:44:06 +0000 (15:44 +0900)
committerEryu Guan <guaneryu@gmail.com>
Sun, 24 Nov 2019 16:07:06 +0000 (00:07 +0800)
btrfs/085 mount and works on $FLAKEY_DEV, but it's inspecting the
underlying $SCRATCH_DEV. Since writing to dm-flakey does not
invalidate page cache of $SCRATCH_DEV, the btrfs command can read a
stalled data from page cache. We should check the flakey device to
avoid such cache inconsistency.

Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
tests/btrfs/085

index 4773ed8..2a31eef 100755 (executable)
@@ -46,7 +46,7 @@ rm -f $seqres.full
 has_orphan_item()
 {
        INO=$1
 has_orphan_item()
 {
        INO=$1
-       if $BTRFS_UTIL_PROG inspect-internal dump-tree $SCRATCH_DEV | \
+       if $BTRFS_UTIL_PROG inspect-internal dump-tree $FLAKEY_DEV | \
                grep -q "key (ORPHAN ORPHAN_ITEM $INO)"; then
                return 0
        fi
                grep -q "key (ORPHAN ORPHAN_ITEM $INO)"; then
                return 0
        fi