[PATCH RESEND] blk-mq: order adding requests to hctx->dispatch and checking SCHED_RESTART

* [PATCH RESEND] blk-mq: order adding requests to hctx->dispatch and checking SCHED_RESTART
@ 2020-08-17 10:01 Ming Lei
  2020-08-17 10:15 ` Christoph Hellwig
                   ` (2 more replies)
  0 siblings, 3 replies; 5+ messages in thread
From: Ming Lei @ 2020-08-17 10:01 UTC (permalink / raw)
  To: Jens Axboe
  Cc: linux-block, Ming Lei, Bart Van Assche, Christoph Hellwig,
	David Jeffery, kernel test robot, stable

SCHED_RESTART code path is relied to re-run queue for dispatch requests
in hctx->dispatch. Meantime the SCHED_RSTART flag is checked when adding
requests to hctx->dispatch.

memory barriers have to be used for ordering the following two pair of OPs:

1) adding requests to hctx->dispatch and checking SCHED_RESTART in
blk_mq_dispatch_rq_list()

2) clearing SCHED_RESTART and checking if there is request in hctx->dispatch
in blk_mq_sched_restart().

Without the added memory barrier, either:

1) blk_mq_sched_restart() may miss requests added to hctx->dispatch meantime
blk_mq_dispatch_rq_list() observes SCHED_RESTART, and not run queue in
dispatch side

or

2) blk_mq_dispatch_rq_list still sees SCHED_RESTART, and not run queue
in dispatch side, meantime checking if there is request in
hctx->dispatch from blk_mq_sched_restart() is missed.

IO hang in ltp/fs_fill test is reported by kernel test robot:

	https://lkml.org/lkml/2020/7/26/77

Turns out it is caused by the above out-of-order OPs. And the IO hang
can't be observed any more after applying this patch.

Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Jeffery <djeffery@redhat.com>
Reported-by: kernel test robot <rong.a.chen@intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
---
 block/blk-mq-sched.c | 9 +++++++++
 block/blk-mq.c       | 9 +++++++++
 2 files changed, 18 insertions(+)

diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c
index a19cdf159b75..d2790e5b06d1 100644
--- a/block/blk-mq-sched.c
+++ b/block/blk-mq-sched.c
@@ -78,6 +78,15 @@ void blk_mq_sched_restart(struct blk_mq_hw_ctx *hctx)
 		return;
 	clear_bit(BLK_MQ_S_SCHED_RESTART, &hctx->state);
 
+	/*
+	 * Order clearing SCHED_RESTART and list_empty_careful(&hctx->dispatch)
+	 * in blk_mq_run_hw_queue(). Its pair is the barrier in
+	 * blk_mq_dispatch_rq_list(). So dispatch code won't see SCHED_RESTART,
+	 * meantime new request added to hctx->dispatch is missed to check in
+	 * blk_mq_run_hw_queue().
+	 */
+	smp_mb();
+
 	blk_mq_run_hw_queue(hctx, true);
 }
 
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 0015a1892153..6c1c3ad175a9 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1437,6 +1437,15 @@ bool blk_mq_dispatch_rq_list(struct blk_mq_hw_ctx *hctx, struct list_head *list,
 		list_splice_tail_init(list, &hctx->dispatch);
 		spin_unlock(&hctx->lock);
 
+		/*
+		 * Order adding requests to hctx->dispatch and checking
+		 * SCHED_RESTART flag. The pair of this smp_mb() is the one
+		 * in blk_mq_sched_restart(). Avoid restart code path to
+		 * miss the new added requests to hctx->dispatch, meantime
+		 * SCHED_RESTART is observed here.
+		 */
+		smp_mb();
+
 		/*
 		 * If SCHED_RESTART was set by the caller of this function and
 		 * it is no longer set that means that it was cleared by another
-- 
2.25.2


^ permalink raw reply related	[flat|nested] 5+ messages in thread