[PATCH 3/3] xfs: terminate perag iteration reliably on end agno

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



The for_each_perag*() set of macros are hacky in that some (i.e. those
based on sb_agcount) rely on the assumption that perag iteration
terminates naturally with a NULL perag at the specified end agno. Others
allow for the final AG to have a valid perag and require the calling
function to clean up any potential leftover xfs_perag reference on
termination of the loop.

Aside from providing a subtly inconsistent interface, the former variant
is racy with a potential growfs in progress because growfs can create
discoverable post-eofs perags before the final superblock update that
completes the grow operation and increases sb_agcount. This leads to
unexpected assert failures (reproduced by xfs/104) such as the following
in the superblock buffer write verifier path:

 XFS: Assertion failed: agno < mp->m_sb.sb_agcount, file: fs/xfs/libxfs/xfs_types.c, line: 22

This occurs because the perag loop in xfs_icount_range() finds and
attempts to process a perag struct where pag_agno == sb_agcount.

The following assert failure occasionally triggers during the xfs_perag
free path on unmount, presumably because one of the many
for_each_perag() loops in the code that is expected to terminate with a
NULL pag raced with a growfs and actually terminated with a non-NULL
reference to post-eofs (at the time) perag.

 XFS: Assertion failed: atomic_read(&pag->pag_ref) == 0, file: fs/xfs/libxfs/xfs_ag.c, line: 195

Rework the lower level perag iteration logic to explicitly terminate
on the specified end agno, not implicitly rely on pag == NULL as a
termination clause and thus avoid these problems. As of this change,
the remaining post-loop xfs_perag_put() checks that exist purely to
cover the natural termination case (i.e., not mid-loop breaks) are
spurious (yet harmless) and can be removed.

Signed-off-by: Brian Foster <bfoster@xxxxxxxxxx>
---
 fs/xfs/libxfs/xfs_ag.h | 20 ++++++++------------
 1 file changed, 8 insertions(+), 12 deletions(-)

diff --git a/fs/xfs/libxfs/xfs_ag.h b/fs/xfs/libxfs/xfs_ag.h
index d05c9217c3af..edcdd4fbc225 100644
--- a/fs/xfs/libxfs/xfs_ag.h
+++ b/fs/xfs/libxfs/xfs_ag.h
@@ -116,34 +116,30 @@ void xfs_perag_put(struct xfs_perag *pag);
 
 /*
  * Perag iteration APIs
- *
- * XXX: for_each_perag_range() usage really needs an iterator to clean up when
- * we terminate at end_agno because we may have taken a reference to the perag
- * beyond end_agno. Right now callers have to be careful to catch and clean that
- * up themselves. This is not necessary for the callers of for_each_perag() and
- * for_each_perag_from() because they terminate at sb_agcount where there are
- * no perag structures in tree beyond end_agno.
  */
 static inline
 struct xfs_perag *xfs_perag_next(
 	struct xfs_perag	*pag,
-	xfs_agnumber_t		*agno)
+	xfs_agnumber_t		*agno,
+	xfs_agnumber_t		end_agno)
 {
 	struct xfs_mount	*mp = pag->pag_mount;
 
 	*agno = pag->pag_agno + 1;
 	xfs_perag_put(pag);
-	pag = xfs_perag_get(mp, *agno);
+	pag = NULL;
+	if (*agno <= end_agno)
+		pag = xfs_perag_get(mp, *agno);
 	return pag;
 }
 
 #define for_each_perag_range(mp, agno, end_agno, pag) \
 	for ((pag) = xfs_perag_get((mp), (agno)); \
-		(pag) != NULL && (agno) <= (end_agno); \
-		(pag) = xfs_perag_next((pag), &(agno)))
+		(pag) != NULL; \
+		(pag) = xfs_perag_next((pag), &(agno), (end_agno)))
 
 #define for_each_perag_from(mp, agno, pag) \
-	for_each_perag_range((mp), (agno), (mp)->m_sb.sb_agcount, (pag))
+	for_each_perag_range((mp), (agno), (mp)->m_sb.sb_agcount - 1, (pag))
 
 
 #define for_each_perag(mp, agno, pag) \
-- 
2.31.1




[Index of Archives]     [XFS Filesystem Development (older mail)]     [Linux Filesystem Development]     [Linux Audio Users]     [Yosemite Trails]     [Linux Kernel]     [Linux RAID]     [Linux SCSI]


  Powered by Linux