blk-mq: In blk_mq_dispatch_rq_list() "no budget" is a reason to kick

[ Upstream commit ab3cee3 ]

In blk_mq_dispatch_rq_list(), if blk_mq_sched_needs_restart() returns
true and the driver returns BLK_STS_RESOURCE then we'll kick the
queue.  However, there's another case where we might need to kick it.
If we were unable to get budget we can be in much the same state as
when the driver returns BLK_STS_RESOURCE, so we should treat it the
same.

It should be noted that even if we add a whole bunch of extra kicking
to the queue in other patches this patch is still important.
Specifically any kicking that happened before we re-spliced leftover
requests into 'hctx->dispatch' wouldn't have found any work, so we
really need to make sure we kick ourselves after we've done the
splicing.

Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>

block/blk-mq.c

@@ -1233,6 +1233,7 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
 	bool no_tag = false;
 	int errors, queued;
 	blk_status_t ret = BLK_STS_OK;
+	bool no_budget_avail = false;
 
 	if (list_empty(list))
 		return false;
@@ -1251,6 +1252,7 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
 		hctx = rq->mq_hctx;
 		if (!got_budget && !blk_mq_get_dispatch_budget(hctx)) {
 			blk_mq_put_driver_tag(rq);
+			no_budget_avail = true;
 			break;
 		}
 
@@ -1356,13 +1358,15 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
 		 *
 		 * If driver returns BLK_STS_RESOURCE and SCHED_RESTART
 		 * bit is set, run queue after a delay to avoid IO stalls
-		 * that could otherwise occur if the queue is idle.
+		 * that could otherwise occur if the queue is idle.  We'll do
+		 * similar if we couldn't get budget and SCHED_RESTART is set.
 		 */
 		needs_restart = blk_mq_sched_needs_restart(hctx);
 		if (!needs_restart ||
 		    (no_tag && list_empty_careful(&hctx->dispatch_wait.entry)))
 			blk_mq_run_hw_queue(hctx, true);
-		else if (needs_restart && (ret == BLK_STS_RESOURCE))
+		else if (needs_restart && (ret == BLK_STS_RESOURCE ||
+					   no_budget_avail))
 			blk_mq_delay_run_hw_queue(hctx, BLK_MQ_RESOURCE_DELAY);
 
 		blk_mq_update_dispatch_busy(hctx, true);