From: Peter Zijlstra <peterz@infradead.org>
To: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Andy Lutomirski <luto@kernel.org>, Ingo Molnar <mingo@kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: Abysmal scheduler performance in Linus' tree?
Date: Wed, 6 Sep 2017 12:44:20 +0200 [thread overview]
Message-ID: <20170906104420.ic5lbpacpyyz53w5@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <150469312649.28581.17626550155735691534@mail.alporthouse.com>
On Wed, Sep 06, 2017 at 11:18:46AM +0100, Chris Wilson wrote:
> > +static void get_llc_stats(struct llc_stats *stats, int cpu)
> > +{
> > + struct sched_domain_shared *sds = rcu_dereference(per_cpu(sd_llc_shared, cpu));
> > +
> > + if (!sds) {
> > + memset(&stats, 0, sizeof(*stats));
>
> Yes, I even sent you a mail about it ;)
Bah, too much email, sorry :-(
> > + /*
> > + * The has_capacity stuff is not SMT aware, but by trying to balance
> > + * the nr_running on both ends we try and fill the domain at equal
> > + * rates, thereby first consuming cores before siblings.
> > + */
> > +
> > + /* if the old cache has capacity, stay there */
> > + if (prev_stats.has_capacity && prev_stats.nr_running < this_stats.nr_running+1)
> > + return false;
> > +
> > + /* if this cache has capacity, come here */
> > + if (this_stats.has_capacity && this_stats.nr_running < prev_stats.nr_running+1)
> > + return true;
>
> This is still not working as intended, it should be
>
> if (this_stats.has_capacity && this_stats.nr_running+1 < prev_stats.nr_running)
> return true;
>
> to fix the regression.
Argh, you're quite right. Let me do a patch for that.
next prev parent reply other threads:[~2017-09-06 10:44 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-09-06 5:13 Abysmal scheduler performance in Linus' tree? Andy Lutomirski
2017-09-06 8:25 ` Peter Zijlstra
2017-09-06 8:59 ` Andy Lutomirski
2017-09-06 9:03 ` Peter Zijlstra
2017-09-06 16:14 ` Peter Zijlstra
2017-09-07 6:15 ` Mike Galbraith
2017-09-07 7:31 ` Ingo Molnar
2017-09-07 9:13 ` [PATCH] sched/cpuset/pm: Fix cpuset vs suspend-resume Peter Zijlstra
2017-09-07 9:26 ` Peter Zijlstra
2017-09-07 10:54 ` Rafael J. Wysocki
2017-09-07 20:38 ` Tejun Heo
2017-09-07 10:33 ` [tip:sched/urgent] sched/cpuset/pm: Fix cpuset vs. suspend-resume bugs tip-bot for Peter Zijlstra
2017-09-06 9:15 ` Abysmal scheduler performance in Linus' tree? Chris Wilson
2017-09-06 9:24 ` Peter Zijlstra
[not found] ` <150469312649.28581.17626550155735691534@mail.alporthouse.com>
2017-09-06 10:44 ` Peter Zijlstra [this message]
2017-09-06 10:51 ` Peter Zijlstra
2017-09-07 8:16 ` [tip:sched/urgent] sched/fair: Fix wake_affine_llc() balancing rules tip-bot for Peter Zijlstra
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=20170906104420.ic5lbpacpyyz53w5@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=chris@chris-wilson.co.uk \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@kernel.org \
--cc=mingo@kernel.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 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).