[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZshyPVEc9w4sqXJy@fedora>
Date: Fri, 23 Aug 2024 19:27:57 +0800
From: Ming Lei <ming.lei@...hat.com>
To: Muchun Song <songmuchun@...edance.com>
Cc: axboe@...nel.dk, linux-block@...r.kernel.org,
linux-kernel@...r.kernel.org, ming.lei@...hat.com
Subject: Re: [PATCH 4/4] block: fix fix ordering between checking
QUEUE_FLAG_QUIESCED and adding requests to hctx->dispatch
On Sun, Aug 11, 2024 at 06:19:21PM +0800, Muchun Song wrote:
> Supposing the following scenario.
>
> CPU0 CPU1
>
> blk_mq_request_issue_directly() blk_mq_unquiesce_queue()
> if (blk_queue_quiesced()) blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store
> blk_mq_insert_request() blk_mq_run_hw_queues()
> /* blk_mq_run_hw_queue()
> * Add request to dispatch list or set bitmap of if (!blk_mq_hctx_has_pending()) 4) load
> * software queue. 1) store return
> */
> blk_mq_run_hw_queue()
> if (blk_queue_quiesced()) 2) load
> return
> blk_mq_sched_dispatch_requests()
>
> The full memory barrier should be inserted between 1) and 2), as well as
> between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is
> cleared or CPU1 sees dispatch list or setting of bitmap of software queue.
> Otherwise, either CPU will not re-run the hardware queue causing starvation.
Memory barrier shouldn't serve as bug fix for two slow code paths.
One simple fix is to add helper of blk_queue_quiesced_lock(), and
call the following check on CPU0:
if (blk_queue_quiesced_lock())
blk_mq_run_hw_queue();
thanks,
Ming
Powered by blists - more mailing lists