]> git.apps.os.sepia.ceph.com Git - xfstests-dev.git/commitdiff
fstests: btrfs/301: handle auto-removed qgroups
authorQu Wenruo <wqu@suse.com>
Tue, 7 May 2024 07:06:06 +0000 (16:36 +0930)
committerAnand Jain <anand.jain@oracle.com>
Fri, 24 May 2024 04:06:43 +0000 (12:06 +0800)
There are always attempts to auto-remove empty qgroups after dropping a
subvolume.

For squota mode, not all qgroups can or should be dropped, as there are
common cases where the dropped subvolume are still referred by other
snapshots.
In that case, the numbers can only be freed when the last referencer
got dropped.

The latest kernel attempt would only try to drop empty qgroups for
squota mode.
But even with such safe change, the test case still needs to handle
auto-removed qgroups, by explicitly echoing "0", or later calculation
would break bash grammar.

This patch would add extra handling for such removed qgroups, to be
future proof for qgroup auto-removal behavior change.

Reviewed-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Boris Burkov <boris@bur.io>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
tests/btrfs/301

index 05e0f4c88183542a2ac0eeed5be5010d19aa0a1a..62854a500414fb2c602a7d6fd71cca7d4e23a83e 100755 (executable)
@@ -47,9 +47,17 @@ _require_fio $fio_config
 get_qgroup_usage()
 {
        local qgroupid=$1
-
-       $BTRFS_UTIL_PROG qgroup show --sync --raw $SCRATCH_MNT | \
-                               grep "$qgroupid" | $AWK_PROG '{print $3}'
+       local output
+
+       output=$($BTRFS_UTIL_PROG qgroup show --sync --raw $SCRATCH_MNT | \
+                grep "$qgroupid" | $AWK_PROG '{print $3}')
+       # The qgroup is auto-removed, this can only happen if its numbers are
+       # already all zeros, so here we only need to explicitly echo "0".
+       if [ -z "$output" ]; then
+               echo "0"
+       else
+               echo "$output"
+       fi
 }
 
 get_subvol_usage()