xfs: mark extended attr corrupt when lookup-by-hash fails
authorDarrick J. Wong <darrick.wong@oracle.com>
Wed, 11 Mar 2020 17:38:09 +0000 (10:38 -0700)
committerDarrick J. Wong <darrick.wong@oracle.com>
Thu, 12 Mar 2020 14:58:13 +0000 (07:58 -0700)
In xchk_xattr_listent, we attempt to validate the extended attribute
hash structures by performing a attr lookup by (hashed) name.  If the
lookup returns ENODATA, that means that the hash information is corrupt.
The _process_error functions don't catch this, so we have to add that
explicitly.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
fs/xfs/scrub/attr.c

index 0d3b5c0..9faddb3 100644 (file)
@@ -160,6 +160,9 @@ xchk_xattr_listent(
        args.valuelen = valuelen;
 
        error = xfs_attr_get_ilocked(&args);
+       /* ENODATA means the hash lookup failed and the attr is bad */
+       if (error == -ENODATA)
+               error = -EFSCORRUPTED;
        if (!xchk_fblock_process_error(sx->sc, XFS_ATTR_FORK, args.blkno,
                        &error))
                goto fail_xref;