xfs: fix bad dquot buffer size in log recovery readahead
authorDave Chinner <dchinner@redhat.com>
Tue, 27 Aug 2013 03:25:43 +0000 (13:25 +1000)
committerBen Myers <bpm@sgi.com>
Thu, 29 Aug 2013 15:51:35 +0000 (10:51 -0500)
xfstests xfs/087 fails 100% reliably with this assert:

XFS (vdb): Mounting Filesystem
XFS (vdb): Starting recovery (logdev: internal)
XFS: Assertion failed: bp->b_flags & XBF_STALE, file: fs/xfs/xfs_buf.c, line: 548

while trying to read a dquot buffer in xlog_recover_dquot_ra_pass2().

The issue is that the buffer length to read that is passed to
xfs_buf_readahead is in units of filesystem blocks, not disk blocks.
(i.e. FSB, not daddr). Fix it but putting the correct conversion in
place.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
fs/xfs/xfs_log_recover.c

index 90b756f7117c94443fab0fe3f3939bbaa07dff3a..006ee288246d06821e5a9ed387be1ab88ff8e1b8 100644 (file)
@@ -3202,7 +3202,7 @@ xlog_recover_dquot_ra_pass2(
        ASSERT(dq_f->qlf_len == 1);
 
        xfs_buf_readahead(mp->m_ddev_targp, dq_f->qlf_blkno,
-                               dq_f->qlf_len, NULL);
+                         XFS_FSB_TO_BB(mp, dq_f->qlf_len), NULL);
 }
 
 STATIC void