Re: [PATCH V3 3/3] blk-mq: centralise related handling into blk_mq_get_driver_tag

[Ming Lei <ming.lei@redhat.com>]

On Thu, Jul 02, 2020 at 12:37:21AM -0400, Qian Cai wrote:
> On Tue, Jun 30, 2020 at 10:03:57PM +0800, Ming Lei wrote:
> > Move .nr_active update and request assignment into blk_mq_get_driver_tag(),
> > all are good to do during getting driver tag.
> > 
> > Meantime blk-flush related code is simplified and flush request needn't
> > to update the request table manually any more.
> > 
> > Cc: Christoph Hellwig <hch@infradead.org>
> > Signed-off-by: Ming Lei <ming.lei@redhat.com>
> 
> Reverting this commit on the top of next-20200701 fixed an issue where
> swapping is unable to move progress for hours while it will only take
> 5-min after the reverting.

Hi Qian,

Could you apply the following patch and see if it makes a difference?

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 65e0846fd065..e89ce9ae51fd 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1147,7 +1147,8 @@ static bool blk_mq_get_driver_tag(struct request *rq)
 	if (rq->tag == BLK_MQ_NO_TAG && !__blk_mq_get_driver_tag(rq))
 		return false;
 
-	if (hctx->flags & BLK_MQ_F_TAG_SHARED) {
+	if ((hctx->flags & BLK_MQ_F_TAG_SHARED) &&
+			!(rq->rq_flags & RQF_MQ_INFLIGHT)) {
 		rq->rq_flags |= RQF_MQ_INFLIGHT;
 		atomic_inc(&hctx->nr_active);
 	}


Thanks,
Ming