* [PATCH for-5.13/block] blk-iocost: don't ignore vrate_min on QD contention
@ 2021-04-23 1:54 Tejun Heo
2021-04-23 14:55 ` Jens Axboe
0 siblings, 1 reply; 2+ messages in thread
From: Tejun Heo @ 2021-04-23 1:54 UTC (permalink / raw)
To: Jens Axboe; +Cc: linux-block, kernel-team, dskarlat, dschatzberg, linux-kernel
ioc_adjust_base_vrate() ignored vrate_min when rq_wait_pct indicates that
there is QD contention. The reasoning was that QD depletion always reliably
indicates device saturation and thus it's safe to override user specified
vrate_min. However, this sometimes leads to unnecessary throttling,
especially on really fast devices, because vrate adjustments have delays and
inertia. It also confuses users because the behavior violates the explicitly
specified configuration.
This patch drops the special case handling so that vrate_min is always
applied.
Signed-off-by: Tejun Heo <tj@kernel.org>
---
block/blk-iocost.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index 98d656bdb42b7..e0c4baa018578 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -987,10 +987,6 @@ static void ioc_adjust_base_vrate(struct ioc *ioc, u32 rq_wait_pct,
return;
}
- /* rq_wait signal is always reliable, ignore user vrate_min */
- if (rq_wait_pct > RQ_WAIT_BUSY_PCT)
- vrate_min = VRATE_MIN;
-
/*
* If vrate is out of bounds, apply clamp gradually as the
* bounds can change abruptly. Otherwise, apply busy_level
--
2.31.1
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH for-5.13/block] blk-iocost: don't ignore vrate_min on QD contention
2021-04-23 1:54 [PATCH for-5.13/block] blk-iocost: don't ignore vrate_min on QD contention Tejun Heo
@ 2021-04-23 14:55 ` Jens Axboe
0 siblings, 0 replies; 2+ messages in thread
From: Jens Axboe @ 2021-04-23 14:55 UTC (permalink / raw)
To: Tejun Heo; +Cc: linux-block, kernel-team, dskarlat, dschatzberg, linux-kernel
On 4/22/21 7:54 PM, Tejun Heo wrote:
> ioc_adjust_base_vrate() ignored vrate_min when rq_wait_pct indicates that
> there is QD contention. The reasoning was that QD depletion always reliably
> indicates device saturation and thus it's safe to override user specified
> vrate_min. However, this sometimes leads to unnecessary throttling,
> especially on really fast devices, because vrate adjustments have delays and
> inertia. It also confuses users because the behavior violates the explicitly
> specified configuration.
>
> This patch drops the special case handling so that vrate_min is always
> applied.
Applied, thanks.
--
Jens Axboe
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2021-04-23 14:55 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-04-23 1:54 [PATCH for-5.13/block] blk-iocost: don't ignore vrate_min on QD contention Tejun Heo
2021-04-23 14:55 ` Jens Axboe
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).