diff options
| author | Yu Kuai <yukuai3@huawei.com> | 2023-06-10 10:30:43 +0800 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2023-06-12 09:55:53 -0600 |
| commit | 4f1731df60f9033669f024d06ae26a6301260b55 (patch) | |
| tree | 8b3d6dbff9e497e9b2ecf9603284499be48a2bde /include/linux/blk-mq.h | |
| parent | 0733ad8002916b9dbbbcfe6e92ad44d2657de1c1 (diff) | |
| download | linux-stable-4f1731df60f9033669f024d06ae26a6301260b55.tar.gz linux-stable-4f1731df60f9033669f024d06ae26a6301260b55.tar.bz2 linux-stable-4f1731df60f9033669f024d06ae26a6301260b55.zip | |
blk-mq: fix potential io hang by wrong 'wake_batch'
In __blk_mq_tag_busy/idle(), updating 'active_queues' and calculating
'wake_batch' is not atomic:
t1: t2:
_blk_mq_tag_busy blk_mq_tag_busy
inc active_queues
// assume 1->2
inc active_queues
// 2 -> 3
blk_mq_update_wake_batch
// calculate based on 3
blk_mq_update_wake_batch
/* calculate based on 2, while active_queues is actually 3. */
Fix this problem by protecting them wih 'tags->lock', this is not a hot
path, so performance should not be concerned. And now that all writers
are inside the lock, switch 'actives_queues' from atomic to unsigned
int.
Fixes: 180dccb0dba4 ("blk-mq: fix tag_get wait task can't be awakened")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230610023043.2559121-1-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'include/linux/blk-mq.h')
| -rw-r--r-- | include/linux/blk-mq.h | 3 |
1 files changed, 1 insertions, 2 deletions
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h index 59b52ec155b1..f401067ac03a 100644 --- a/include/linux/blk-mq.h +++ b/include/linux/blk-mq.h @@ -739,8 +739,7 @@ struct request *blk_mq_alloc_request_hctx(struct request_queue *q, struct blk_mq_tags { unsigned int nr_tags; unsigned int nr_reserved_tags; - - atomic_t active_queues; + unsigned int active_queues; struct sbitmap_queue bitmap_tags; struct sbitmap_queue breserved_tags; |