[dm-devel] [PATCH] dm: make sure dm_table is binded before queue request

Zhang Yi yi.zhang at huawei.com
Wed Feb 9 09:37:51 UTC 2022


We found a NULL pointer dereference problem when using dm-mpath target.
The problem is if we submit IO between loading and binding the table,
we could neither get a valid dm_target nor a valid dm table when
submitting request in dm_mq_queue_rq(). BIO based dm target could
handle this case in dm_submit_bio(). This patch fix this by checking
the mapping table before submitting request.

Signed-off-by: Zhang Yi <yi.zhang at huawei.com>
---
 drivers/md/dm-rq.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/drivers/md/dm-rq.c b/drivers/md/dm-rq.c
index 579ab6183d4d..af2cf71519e9 100644
--- a/drivers/md/dm-rq.c
+++ b/drivers/md/dm-rq.c
@@ -499,8 +499,15 @@ static blk_status_t dm_mq_queue_rq(struct blk_mq_hw_ctx *hctx,
 
 	if (unlikely(!ti)) {
 		int srcu_idx;
-		struct dm_table *map = dm_get_live_table(md, &srcu_idx);
-
+		struct dm_table *map;
+
+		map = dm_get_live_table(md, &srcu_idx);
+		if (!map) {
+			DMERR_LIMIT("%s: mapping table unavailable, erroring io",
+				    dm_device_name(md));
+			dm_put_live_table(md, srcu_idx);
+			return BLK_STS_IOERR;
+		}
 		ti = dm_table_find_target(map, 0);
 		dm_put_live_table(md, srcu_idx);
 	}
-- 
2.31.1





More information about the dm-devel mailing list