All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mel Gorman <mgorman@techsingularity.net>
To: "Li, Aubrey" <aubrey.li@linux.intel.com>
Cc: Barry Song <song.bao.hua@hisilicon.com>,
	Juri Lelli <juri.lelli@redhat.com>,
	Vincent Guittot <vincent.guittot@linaro.org>,
	Peter Ziljstra <peterz@infradead.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Ingo Molnar <mingo@redhat.com>,
	Valentin Schneider <valentin.schneider@arm.com>,
	Linux-ARM <linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH 2/4] sched/fair: Move avg_scan_cost calculations under SIS_PROP
Date: Thu, 10 Dec 2020 09:32:40 +0000	[thread overview]
Message-ID: <20201210093240.GQ3371@techsingularity.net> (raw)
In-Reply-To: <1963d0ca-054c-19f9-94e0-d019a2e8e259@linux.intel.com>

On Thu, Dec 10, 2020 at 01:18:05PM +0800, Li, Aubrey wrote:
> > diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> > index ac7b34e7372b..5c41875aec23 100644
> > --- a/kernel/sched/fair.c
> > +++ b/kernel/sched/fair.c
> > @@ -6153,6 +6153,8 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
> >  	if (!this_sd)
> >  		return -1;
> >  
> > +	cpumask_and(cpus, sched_domain_span(sd), p->cpus_ptr);
> > +
> >  	if (sched_feat(SIS_PROP)) {
> >  		u64 avg_cost, avg_idle, span_avg;
> >  
> > @@ -6168,11 +6170,9 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
> >  			nr = div_u64(span_avg, avg_cost);
> >  		else
> >  			nr = 4;
> > -	}
> > -
> > -	time = cpu_clock(this);
> >  
> > -	cpumask_and(cpus, sched_domain_span(sd), p->cpus_ptr);
> > +		time = cpu_clock(this);
> > +	}
> >  
> >  	for_each_cpu_wrap(cpu, cpus, target) {
> >  		if (!--nr)
> >			return -1;
> 
> I thought about this again and here seems not to be consistent:
> - even if nr reduces to 0, shouldn't avg_scan_cost be updated as well before return -1?

You're right, but it's outside the scope
of this patch. I noted that this was a problem in
lore.kernel.org/r/lore.kernel.org/r/20201203141124.7391-8-mgorman@techsingularity.net
It's neither a consistent win or loss to always account for it and so
was dropped for this series to keep the number of controversial patches
to a minimum.

> - if avg_scan_cost is not updated because nr is throttled, the first 
> 	time = cpu_clock(this);
>   can be optimized. As nr is calculated and we already know which of the weight of cpumask and nr is greater.
> 

That is also outside the scope of this patch. To do that, cpumask_weight()
would have to be calculated but it's likely to be a net loss. Even under
light load, nr will be smaller than the domain weight incurring both the
cost of cpumask_weight and the clock read in the common case.

-- 
Mel Gorman
SUSE Labs

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

WARNING: multiple messages have this Message-ID (diff)
From: Mel Gorman <mgorman@techsingularity.net>
To: "Li, Aubrey" <aubrey.li@linux.intel.com>
Cc: Peter Ziljstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>,
	LKML <linux-kernel@vger.kernel.org>,
	Barry Song <song.bao.hua@hisilicon.com>,
	Juri Lelli <juri.lelli@redhat.com>,
	Vincent Guittot <vincent.guittot@linaro.org>,
	Valentin Schneider <valentin.schneider@arm.com>,
	Linux-ARM <linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH 2/4] sched/fair: Move avg_scan_cost calculations under SIS_PROP
Date: Thu, 10 Dec 2020 09:32:40 +0000	[thread overview]
Message-ID: <20201210093240.GQ3371@techsingularity.net> (raw)
In-Reply-To: <1963d0ca-054c-19f9-94e0-d019a2e8e259@linux.intel.com>

On Thu, Dec 10, 2020 at 01:18:05PM +0800, Li, Aubrey wrote:
> > diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> > index ac7b34e7372b..5c41875aec23 100644
> > --- a/kernel/sched/fair.c
> > +++ b/kernel/sched/fair.c
> > @@ -6153,6 +6153,8 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
> >  	if (!this_sd)
> >  		return -1;
> >  
> > +	cpumask_and(cpus, sched_domain_span(sd), p->cpus_ptr);
> > +
> >  	if (sched_feat(SIS_PROP)) {
> >  		u64 avg_cost, avg_idle, span_avg;
> >  
> > @@ -6168,11 +6170,9 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
> >  			nr = div_u64(span_avg, avg_cost);
> >  		else
> >  			nr = 4;
> > -	}
> > -
> > -	time = cpu_clock(this);
> >  
> > -	cpumask_and(cpus, sched_domain_span(sd), p->cpus_ptr);
> > +		time = cpu_clock(this);
> > +	}
> >  
> >  	for_each_cpu_wrap(cpu, cpus, target) {
> >  		if (!--nr)
> >			return -1;
> 
> I thought about this again and here seems not to be consistent:
> - even if nr reduces to 0, shouldn't avg_scan_cost be updated as well before return -1?

You're right, but it's outside the scope
of this patch. I noted that this was a problem in
lore.kernel.org/r/lore.kernel.org/r/20201203141124.7391-8-mgorman@techsingularity.net
It's neither a consistent win or loss to always account for it and so
was dropped for this series to keep the number of controversial patches
to a minimum.

> - if avg_scan_cost is not updated because nr is throttled, the first 
> 	time = cpu_clock(this);
>   can be optimized. As nr is calculated and we already know which of the weight of cpumask and nr is greater.
> 

That is also outside the scope of this patch. To do that, cpumask_weight()
would have to be calculated but it's likely to be a net loss. Even under
light load, nr will be smaller than the domain weight incurring both the
cost of cpumask_weight and the clock read in the common case.

-- 
Mel Gorman
SUSE Labs

  reply	other threads:[~2020-12-10  9:33 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-08 15:34 [PATCH 0/4] Reduce scanning of runqueues in select_idle_sibling Mel Gorman
2020-12-08 15:34 ` Mel Gorman
2020-12-08 15:34 ` [PATCH 1/4] sched/fair: Remove SIS_AVG_CPU Mel Gorman
2020-12-08 15:34   ` Mel Gorman
2020-12-08 16:13   ` Vincent Guittot
2020-12-08 16:13     ` Vincent Guittot
2020-12-08 15:34 ` [PATCH 2/4] sched/fair: Move avg_scan_cost calculations under SIS_PROP Mel Gorman
2020-12-08 15:34   ` Mel Gorman
2020-12-08 16:03   ` Vincent Guittot
2020-12-08 16:03     ` Vincent Guittot
2020-12-08 16:30     ` Mel Gorman
2020-12-08 16:30       ` Mel Gorman
2020-12-09  5:28     ` Li, Aubrey
2020-12-09  5:28       ` Li, Aubrey
2020-12-09  9:05       ` Mel Gorman
2020-12-09  9:05         ` Mel Gorman
2020-12-09 11:07         ` Li, Aubrey
2020-12-09 11:07           ` Li, Aubrey
2020-12-09 11:33           ` Mel Gorman
2020-12-09 11:33             ` Mel Gorman
2020-12-10  5:18   ` Li, Aubrey
2020-12-10  5:18     ` Li, Aubrey
2020-12-10  9:32     ` Mel Gorman [this message]
2020-12-10  9:32       ` Mel Gorman
2020-12-08 15:35 ` [PATCH 3/4] sched/fair: Do not replace recent_used_cpu with the new target Mel Gorman
2020-12-08 15:35   ` Mel Gorman
2020-12-08 16:14   ` Vincent Guittot
2020-12-08 16:14     ` Vincent Guittot
2020-12-10  9:40     ` Vincent Guittot
2020-12-10  9:40       ` Vincent Guittot
2020-12-11  6:25   ` Hillf Danton
2020-12-11  9:02     ` Mel Gorman
2020-12-11  9:02       ` Mel Gorman
2020-12-11  9:34       ` Hillf Danton
2020-12-11  9:45         ` Mel Gorman
2020-12-11  9:45           ` Mel Gorman
2020-12-08 15:35 ` [PATCH 4/4] sched/fair: Return an idle cpu if one is found after a failed search for an idle core Mel Gorman
2020-12-08 15:35   ` Mel Gorman
2020-12-08 16:15   ` Vincent Guittot
2020-12-08 16:15     ` Vincent Guittot
2020-12-09 14:37 ` [PATCH 0/4] Reduce scanning of runqueues in select_idle_sibling Mel Gorman
2020-12-09 14:37   ` Mel Gorman
2020-12-10  8:00   ` Vincent Guittot
2020-12-10  8:00     ` Vincent Guittot
2020-12-10  9:38     ` Vincent Guittot
2020-12-10  9:38       ` Vincent Guittot
2020-12-10 11:04       ` Mel Gorman
2020-12-10 11:04         ` Mel Gorman
2020-12-11  9:51         ` Vincent Guittot
2020-12-11  9:51           ` Vincent Guittot
2020-12-11 10:23           ` Mel Gorman
2020-12-11 10:23             ` Mel Gorman
2020-12-12 10:02             ` Vincent Guittot
2020-12-12 10:02               ` Vincent Guittot
  -- strict thread matches above, loose matches on Subject: below --
2021-01-25  8:59 [PATCH v4 0/4] Scan for an idle sibling in a single pass Mel Gorman
2021-01-25  8:59 ` [PATCH 2/4] sched/fair: Move avg_scan_cost calculations under SIS_PROP Mel Gorman
2021-01-27 10:39   ` Vincent Guittot
2021-01-27 13:51 [PATCH v5 0/4] Scan for an idle sibling in a single pass Mel Gorman
2021-01-27 13:52 ` [PATCH 2/4] sched/fair: Move avg_scan_cost calculations under SIS_PROP Mel Gorman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20201210093240.GQ3371@techsingularity.net \
    --to=mgorman@techsingularity.net \
    --cc=aubrey.li@linux.intel.com \
    --cc=juri.lelli@redhat.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    --cc=song.bao.hua@hisilicon.com \
    --cc=valentin.schneider@arm.com \
    --cc=vincent.guittot@linaro.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.