Skip to contentSkip to conversation

block: Consider only dispatched requests for inflight statistic

PATCH RESEND Superseded Gabriel Krisman Bertazi 2 participants 6 messages Oct 6, 2020 Block layer

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH RESEND v2 1/1 block: Consider only dispatched requests for inflight statistic

Hi Jens,

Reviving this, hopefully for the upcoming merge window, following the suggestion that we let it sit there and see if people complain about the metric fix. As I mentioned, I'm also happy to change the documentation plus create a new in_device metric to track only requests submitted, if you prefer.

Thanks,

Quoted 1 line

8

According to Documentation/block/stat.rst, inflight should not include I/O requests that are in the queue but not yet dispatched to the device, but blk-mq identifies as inflight any request that has a tag allocated, which, for queues without elevator, happens at request allocation time and before it is queued in the ctx (default case in blk_mq_submit_bio).

In addition, current behavior is different for queues with elevator from queues without it, since for the former the driver tag is allocated at dispatch time. A more precise approach would be to only consider requests with state MQ_RQ_IN_FLIGHT.

This effectively reverts commit 6131837b1de6 ("blk-mq: count allocated but not started requests in iostats inflight") to consolidate blk-mq behavior with itself (elevator case) and with original documentation, but it differs from the behavior used by the legacy path.

This version differs from v1 by using blk_mq_rq_state to access the state attribute. Avoid using blk_mq_request_started, which was suggested, since we don't want to include MQ_RQ_COMPLETE.

---
 block/blk-mq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
block/blk-mq.c +0 -0
index 2e4b3cad2a61..c5fefd39d0c0 100644
Signature · 1 line
2.28.0
#1 lore raw

PATCH RESEND v2 1/1 block: Consider only dispatched requests for inflight statistic

Oops, I have no idea what happened, but something ate the hunk at the last submission. My apologies. Please find it below.

Quoted 1 line

8

According to Documentation/block/stat.rst, inflight should not include I/O requests that are in the queue but not yet dispatched to the device, but blk-mq identifies as inflight any request that has a tag allocated, which, for queues without elevator, happens at request allocation time and before it is queued in the ctx (default case in blk_mq_submit_bio).

In addition, current behavior is different for queues with elevator from queues without it, since for the former the driver tag is allocated at dispatch time. A more precise approach would be to only consider requests with state MQ_RQ_IN_FLIGHT.

This effectively reverts commit 6131837b1de6 ("blk-mq: count allocated but not started requests in iostats inflight") to consolidate blk-mq behavior with itself (elevator case) and with original documentation, but it differs from the behavior used by the legacy path.

This version differs from v1 by using blk_mq_rq_state to access the state attribute. Avoid using blk_mq_request_started, which was suggested, since we don't want to include MQ_RQ_COMPLETE.

---
 block/blk-mq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
block/blk-mq.c +1 -1
index 2e4b3cad2a61..c5fefd39d0c0 100644
@@ -105,7 +105,7 @@ static bool blk_mq_check_inflight(struct blk_mq_hw_ctx *hctx,
 {
 	struct mq_inflight *mi = priv;
 
-	if (rq->part == mi->part)
+	if (rq->part == mi->part && blk_mq_rq_state(rq) == MQ_RQ_IN_FLIGHT)
 		mi->inflight[rq_data_dir(rq)]++;
 
 	return true;
Signature · 1 line
2.28.0
maintainer
#2 lore raw
Quoted 3 lines · Gabriel Krisman Bertazi

Oops, I have no idea what happened, but something ate the hunk at the last submission. My apologies. Please find it below.

Care to just resend a fixed up one? Saves me the time from fixing things up.

I guess we'll just try and see if this flies, not sure how else to make progress on it.

Signature · 1 line
Jens Axboe
#3 lore raw
Quoted 7 lines · Jens Axboe
Quoted 3 lines · Gabriel Krisman Bertazi

Oops, I have no idea what happened, but something ate the hunk at the last submission. My apologies. Please find it below.

Care to just resend a fixed up one? Saves me the time from fixing things up.

hm, the first submission had an empty patch and the email you quoted had the entire fixed patch ready to apply with scissors. It should be good to apply it, I think. Or, what do you mean?

Signature · 1 line
Gabriel Krisman Bertazi
maintainer
#4 lore raw
Quoted 7 lines · Gabriel Krisman Bertazi
Quoted 5 lines · Jens Axboe
Quoted 3 lines · Gabriel Krisman Bertazi

Oops, I have no idea what happened, but something ate the hunk at the last submission. My apologies. Please find it below.

Quoted 6 lines
Quoted 2 lines

Care to just resend a fixed up one? Saves me the time from fixing things up.

hm, the first submission had an empty patch and the email you quoted had the entire fixed patch ready to apply with scissors. It should be good to apply it, I think. Or, what do you mean?

The point is that I need to manually fiddle with it.

Signature · 1 line
Jens Axboe
#5 lore raw
Quoted 8 lines · Jens Axboe
Quoted 7 lines · Gabriel Krisman Bertazi
Quoted 5 lines · Jens Axboe
Quoted 3 lines · Gabriel Krisman Bertazi

Oops, I have no idea what happened, but something ate the hunk at the last submission. My apologies. Please find it below.

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

Care to just resend a fixed up one? Saves me the time from fixing things up.

hm, the first submission had an empty patch and the email you quoted had the entire fixed patch ready to apply with scissors. It should be good to apply it, I think. Or, what do you mean?

The point is that I need to manually fiddle with it.

Done, sorry for the noise. The v3 should apply on top of your for-next branch cleanly.

Signature · 1 line
Gabriel Krisman Bertazi