Thread (21 messages) 21 messages, 6 authors, 2022-03-29
STALE1558d LANDED: 2 (0M)
Revisions (2)
  1. v1 [diff vs current]
  2. v2 current

[PATCH v2 1/4] block/mq-deadline: Improve request accounting further

From: Bart Van Assche <bvanassche@acm.org>
Date: 2021-09-27 22:03:54
Subsystem: block layer, the rest · Maintainers: Jens Axboe, Linus Torvalds

The scheduler .insert_requests() callback is called when a request is
queued for the first time and also when it is requeued. Only count a
request the first time it is queued. Additionally, since the mq-deadline
scheduler only performs zone locking for requests that have been
inserted, skip the zone unlock code for requests that have not been
inserted into the mq-deadline scheduler.

Fixes: 38ba64d12d4c ("block/mq-deadline: Track I/O statistics")
Reviewed-by: Damien Le Moal <redacted>
Reviewed-by: Niklas Cassel <redacted>
Cc: Hannes Reinecke <hare@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
---
 block/mq-deadline.c | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)
diff --git a/block/mq-deadline.c b/block/mq-deadline.c
index 47f042fa6a68..c27b4347ca91 100644
--- a/block/mq-deadline.c
+++ b/block/mq-deadline.c
@@ -677,8 +677,10 @@ static void dd_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 	blk_req_zone_write_unlock(rq);
 
 	prio = ioprio_class_to_prio[ioprio_class];
-	dd_count(dd, inserted, prio);
-	rq->elv.priv[0] = (void *)(uintptr_t)1;
+	if (!rq->elv.priv[0]) {
+		dd_count(dd, inserted, prio);
+		rq->elv.priv[0] = (void *)(uintptr_t)1;
+	}
 
 	if (blk_mq_sched_try_insert_merge(q, rq, &free)) {
 		blk_mq_free_requests(&free);
@@ -759,12 +761,13 @@ static void dd_finish_request(struct request *rq)
 
 	/*
 	 * The block layer core may call dd_finish_request() without having
-	 * called dd_insert_requests(). Hence only update statistics for
-	 * requests for which dd_insert_requests() has been called. See also
-	 * blk_mq_request_bypass_insert().
+	 * called dd_insert_requests(). Skip requests that bypassed I/O
+	 * scheduling. See also blk_mq_request_bypass_insert().
 	 */
-	if (rq->elv.priv[0])
-		dd_count(dd, completed, prio);
+	if (!rq->elv.priv[0])
+		return;
+
+	dd_count(dd, completed, prio);
 
 	if (blk_queue_is_zoned(q)) {
 		unsigned long flags;
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help