This is a note to let you know that I've just added the patch titled
blk-mq: fix potential io hang by wrong 'wake_batch'
to the 6.3-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
blk-mq-fix-potential-io-hang-by-wrong-wake_batch.patch
and it can be found in the queue-6.3 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
commit d74e29b02e06af5636acb504e0771b64f38b2a4d
Author: Yu Kuai <yukuai3@xxxxxxxxxx>
Date: Sat Jun 10 10:30:43 2023 +0800
blk-mq: fix potential io hang by wrong 'wake_batch'
[ Upstream commit 4f1731df60f9033669f024d06ae26a6301260b55 ]
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@xxxxxxxxxx>
Reviewed-by: Jan Kara <jack@xxxxxxx>
Link: https://lore.kernel.org/r/20230610023043.2559121-1-yukuai1@xxxxxxxxxxxxxxx
Signed-off-by: Jens Axboe <axboe@xxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
diff --git a/block/blk-mq-debugfs.c b/block/blk-mq-debugfs.c
index b01818f8e216e..c152276736832 100644
--- a/block/blk-mq-debugfs.c
+++ b/block/blk-mq-debugfs.c
@@ -427,7 +427,7 @@ static void blk_mq_debugfs_tags_show(struct seq_file *m,
seq_printf(m, "nr_tags=%u\n", tags->nr_tags);
seq_printf(m, "nr_reserved_tags=%u\n", tags->nr_reserved_tags);
seq_printf(m, "active_queues=%d\n",
- atomic_read(&tags->active_queues));
+ READ_ONCE(tags->active_queues));
seq_puts(m, "\nbitmap_tags:\n");
sbitmap_queue_show(&tags->bitmap_tags, m);
diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c
index a80d7c62bdfe6..100889c276c3f 100644
--- a/block/blk-mq-tag.c
+++ b/block/blk-mq-tag.c
@@ -40,6 +40,7 @@ static void blk_mq_update_wake_batch(struct blk_mq_tags *tags,
void __blk_mq_tag_busy(struct blk_mq_hw_ctx *hctx)
{
unsigned int users;
+ struct blk_mq_tags *tags = hctx->tags;
/*
* calling test_bit() prior to test_and_set_bit() is intentional,
@@ -57,9 +58,11 @@ void __blk_mq_tag_busy(struct blk_mq_hw_ctx *hctx)
return;
}
- users = atomic_inc_return(&hctx->tags->active_queues);
-
- blk_mq_update_wake_batch(hctx->tags, users);
+ spin_lock_irq(&tags->lock);
+ users = tags->active_queues + 1;
+ WRITE_ONCE(tags->active_queues, users);
+ blk_mq_update_wake_batch(tags, users);
+ spin_unlock_irq(&tags->lock);
}
/*
@@ -92,9 +95,11 @@ void __blk_mq_tag_idle(struct blk_mq_hw_ctx *hctx)
return;
}
- users = atomic_dec_return(&tags->active_queues);
-
+ spin_lock_irq(&tags->lock);
+ users = tags->active_queues - 1;
+ WRITE_ONCE(tags->active_queues, users);
blk_mq_update_wake_batch(tags, users);
+ spin_unlock_irq(&tags->lock);
blk_mq_tag_wakeup_all(tags, false);
}
diff --git a/block/blk-mq.h b/block/blk-mq.h
index a7482d2cc82e7..4542308c8e62f 100644
--- a/block/blk-mq.h
+++ b/block/blk-mq.h
@@ -362,8 +362,7 @@ static inline bool hctx_may_queue(struct blk_mq_hw_ctx *hctx,
return true;
}
- users = atomic_read(&hctx->tags->active_queues);
-
+ users = READ_ONCE(hctx->tags->active_queues);
if (!users)
return true;
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index de0b0c3e7395a..4110d6e99b2b9 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -748,8 +748,7 @@ 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;
