linux-kernel - [PATCH for-5.13/block] blk-iocost: don't ignore vrate

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-ID: <YIIo1HuyNmhDeiNx@slm.duckdns.org>
Date:   Thu, 22 Apr 2021 21:54:28 -0400
From:   Tejun Heo <tj@...nel.org>
To:     Jens Axboe <axboe@...nel.dk>
Cc:     linux-block@...r.kernel.org, kernel-team@...com, dskarlat@...com,
        dschatzberg@...com, linux-kernel@...r.kernel.org
Subject: [PATCH for-5.13/block] blk-iocost: don't ignore vrate_min on QD
 contention

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@...nel.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