[PATCH] xfs: don't check agf_btreeblks on pre-lazysbcount filesystems

[PATCH] xfs: don't check agf_btreeblks on pre-lazysbcount filesystems

[Darrick J. Wong]

From: Darrick J. Wong <djwong@kernel.org>

The AGF free space btree block counter wasn't added until the
lazysbcount feature was added to XFS midway through the life of the V4
format, so ignore the field when checking.  Online AGF repair requires
rmapbt, so it doesn't need the feature check.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
 fs/xfs/scrub/agheader.c   |    7 ++++++-
 fs/xfs/scrub/fscounters.c |    3 ++-
 2 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/fs/xfs/scrub/agheader.c b/fs/xfs/scrub/agheader.c
index 749faa17f8e2..7a2f9b5f2db5 100644
--- a/fs/xfs/scrub/agheader.c
+++ b/fs/xfs/scrub/agheader.c
@@ -416,6 +416,10 @@ xchk_agf_xref_btreeblks(
 	xfs_agblock_t		btreeblks;
 	int			error;
 
+	/* agf_btreeblks didn't exist before lazysbcount */
+	if (!xfs_sb_version_haslazysbcount(&sc->mp->m_sb))
+		return;
+
 	/* Check agf_rmap_blocks; set up for agf_btreeblks check */
 	if (sc->sa.rmap_cur) {
 		error = xfs_btree_count_blocks(sc->sa.rmap_cur, &blocks);
@@ -581,7 +585,8 @@ xchk_agf(
 		xchk_block_set_corrupt(sc, sc->sa.agf_bp);
 	if (pag->pagf_flcount != be32_to_cpu(agf->agf_flcount))
 		xchk_block_set_corrupt(sc, sc->sa.agf_bp);
-	if (pag->pagf_btreeblks != be32_to_cpu(agf->agf_btreeblks))
+	if (xfs_sb_version_haslazysbcount(&sc->mp->m_sb) &&
+	    pag->pagf_btreeblks != be32_to_cpu(agf->agf_btreeblks))
 		xchk_block_set_corrupt(sc, sc->sa.agf_bp);
 	xfs_perag_put(pag);
 
diff --git a/fs/xfs/scrub/fscounters.c b/fs/xfs/scrub/fscounters.c
index 7b4386c78fbf..318b81c0f90d 100644
--- a/fs/xfs/scrub/fscounters.c
+++ b/fs/xfs/scrub/fscounters.c
@@ -182,7 +182,8 @@ xchk_fscount_aggregate_agcounts(
 		/* Add up the free/freelist/bnobt/cntbt blocks */
 		fsc->fdblocks += pag->pagf_freeblks;
 		fsc->fdblocks += pag->pagf_flcount;
-		fsc->fdblocks += pag->pagf_btreeblks;
+		if (xfs_sb_version_haslazysbcount(&sc->mp->m_sb))
+			fsc->fdblocks += pag->pagf_btreeblks;
 
 		/*
 		 * Per-AG reservations are taken out of the incore counters,

Re: [PATCH] xfs: don't check agf_btreeblks on pre-lazysbcount filesystems

[Dave Chinner]

On Sun, Apr 25, 2021 at 08:46:59AM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <djwong@kernel.org>
> 
> The AGF free space btree block counter wasn't added until the
> lazysbcount feature was added to XFS midway through the life of the V4
> format, so ignore the field when checking.  Online AGF repair requires
> rmapbt, so it doesn't need the feature check.
> 
> Signed-off-by: Darrick J. Wong <djwong@kernel.org>

Looks good.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
-- 
Dave Chinner
david@fromorbit.com