linux-kernel - [PATCH 2/5] iocost: replace nr_shortages cond in ioc_forgive_debts() with busy

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

Date:   Thu, 17 Sep 2020 20:44:53 -0400
From:   Tejun Heo <tj@...nel.org>
To:     axboe@...nel.dk
Cc:     linux-block@...r.kernel.org, cgroups@...r.kernel.org,
        kernel-team@...com, linux-kernel@...r.kernel.org,
        Tejun Heo <tj@...nel.org>
Subject: [PATCH 2/5] iocost: replace nr_shortages cond in ioc_forgive_debts() with busy_level one

Debt reduction was blocked if any iocg was short on budget in the past
period to avoid reducing debts while some iocgs are saturated. However, this
ends up unnecessarily blocking debt reduction due to temporary local
imbalances when the device is generally being underutilized, while also
failing to block when the underlying device is overwhelmed and the usage
becomes low from high latency.

Given that debt accumulation mostly happens with swapout bursts which can
significantly deteriorate the underlying device's latency response, the
current logic is not great.

Let's replace it with ioc->busy_level based condition so that we block debt
reduction when the underlying device is being saturated. ioc_forgive_debts()
call is moved after busy_level determination.

Signed-off-by: Tejun Heo <tj@...nel.org>
---
 block/blk-iocost.c | 13 +++++--------
 1 file changed, 5 insertions(+), 8 deletions(-)

diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index bbf30bb06c07..c0499c294da9 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -287,10 +287,7 @@ enum {
 	MIN_DELAY		= 250,
 	MAX_DELAY		= 250 * USEC_PER_MSEC,
 
-	/*
-	 * Halve debts if total usage keeps staying under 25% w/o any shortages
-	 * for over 100ms.
-	 */
+	/* halve debts if total usage keeps staying under 25% for over 100ms */
 	DEBT_BUSY_USAGE_PCT	= 25,
 	DEBT_REDUCTION_IDLE_DUR	= 100 * USEC_PER_MSEC,
 
@@ -1990,9 +1987,9 @@ static void transfer_surpluses(struct list_head *surpluses, struct ioc_now *now)
  * sufficiently idle for a while, the debts are halved.
  */
 static void ioc_forgive_debts(struct ioc *ioc, u64 usage_us_sum, int nr_debtors,
-			      int nr_shortages, struct ioc_now *now)
+			      struct ioc_now *now)
 {
-	if (nr_shortages ||
+	if (ioc->busy_level < 0 ||
 	    div64_u64(100 * usage_us_sum, now->now - ioc->period_at) >=
 	    DEBT_BUSY_USAGE_PCT)
 		ioc->debt_busy_at = now->now;
@@ -2205,8 +2202,6 @@ static void ioc_timer_fn(struct timer_list *timer)
 	list_for_each_entry_safe(iocg, tiocg, &surpluses, surplus_list)
 		list_del_init(&iocg->surplus_list);
 
-	ioc_forgive_debts(ioc, usage_us_sum, nr_debtors, nr_shortages, &now);
-
 	/*
 	 * If q is getting clogged or we're missing too much, we're issuing
 	 * too much IO and should lower vtime rate.  If we're not missing
@@ -2301,6 +2296,8 @@ static void ioc_timer_fn(struct timer_list *timer)
 
 	ioc_refresh_params(ioc, false);
 
+	ioc_forgive_debts(ioc, usage_us_sum, nr_debtors, &now);
+
 	/*
 	 * This period is done.  Move onto the next one.  If nothing's
 	 * going on with the device, stop the timer.
-- 
2.26.2