From: Peter Zijlstra <peterz@infradead.org>
To: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Steve Muckle <steve.muckle@linaro.org>,
Ingo Molnar <mingo@redhat.com>,
"Rafael J. Wysocki" <rafael@kernel.org>,
linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org,
Vincent Guittot <vincent.guittot@linaro.org>,
Morten Rasmussen <morten.rasmussen@arm.com>,
Dietmar Eggemann <dietmar.eggemann@arm.com>,
Juri Lelli <Juri.Lelli@arm.com>,
Patrick Bellasi <patrick.bellasi@arm.com>,
Michael Turquette <mturquette@baylibre.com>,
Ricky Liang <jcliang@chromium.org>
Subject: Re: [RFCv7 PATCH 03/10] sched: scheduler-driven cpu frequency selection
Date: Thu, 25 Feb 2016 10:28:37 +0100 [thread overview]
Message-ID: <20160225092837.GD6357@twins.programming.kicks-ass.net> (raw)
In-Reply-To: <8427745.Y8N2bqC3SO@vostro.rjw.lan>
On Thu, Feb 25, 2016 at 04:55:57AM +0100, Rafael J. Wysocki wrote:
> > +static void dummy(void *info) {}
> > +
> > +static int cpufreq_sched_stop(struct cpufreq_policy *policy)
> > +{
> > + struct gov_data *gd = policy->governor_data;
> > + int cpu;
> > +
> > + /*
> > + * The schedfreq static key is managed here so the global schedfreq
> > + * lock must be taken - a per-policy lock such as policy->rwsem is
> > + * not sufficient.
> > + */
> > + mutex_lock(&gov_enable_lock);
> > +
> > + /*
> > + * The governor stop path may or may not hold policy->rwsem. There
> > + * must be synchronization with the slow path however.
> > + */
> > + mutex_lock(&gd->slowpath_lock);
> > +
> > + /*
> > + * Stop new entries into the hot path for all CPUs. This will
> > + * potentially affect other policies which are still running but
> > + * this is an infrequent operation.
> > + */
> > + static_key_slow_dec(&__sched_freq);
> > + enabled_policies--;
> > +
> > + /*
> > + * Ensure that all CPUs currently part of this policy are out
> > + * of the hot path so that if this policy exits we can free gd.
> > + */
> > + preempt_disable();
> > + smp_call_function_many(policy->cpus, dummy, NULL, true);
> > + preempt_enable();
>
> I'm not sure how this works, can you please tell me?
I think it relies on the fact that rq->lock disables IRQs, so if we've
managed to IPI all relevant CPUs, it means they cannot be inside a
rq->lock section.
Its vile though; one should not spray IPIs if one can avoid it. Such
things are much better done with RCU. Sure sync_sched() takes a little
longer, but this isn't a fast path by any measure.
> > +
> > + /*
> > + * Other CPUs in other policies may still have the schedfreq
> > + * static key enabled. The percpu gd is used to signal which
> > + * CPUs are enabled in the sched gov during the hot path.
> > + */
> > + for_each_cpu(cpu, policy->cpus)
> > + per_cpu(cpu_gov_data, cpu) = NULL;
> > +
> > + /* Pause the slow path for this policy. */
> > + gd->enabled = 0;
> > +
> > + if (enabled_policies)
> > + static_key_slow_inc(&__sched_freq);
> > + mutex_unlock(&gd->slowpath_lock);
> > + mutex_unlock(&gov_enable_lock);
> > +
> > + return 0;
> > +}
next prev parent reply other threads:[~2016-02-25 9:28 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-23 1:22 [RFCv7 PATCH 00/10] sched: scheduler-driven CPU frequency selection Steve Muckle
2016-02-23 1:22 ` [RFCv7 PATCH 01/10] sched: Compute cpu capacity available at current frequency Steve Muckle
2016-02-23 1:41 ` Rafael J. Wysocki
2016-02-23 9:19 ` Peter Zijlstra
2016-02-26 1:37 ` Rafael J. Wysocki
2016-02-26 9:14 ` Peter Zijlstra
2016-02-23 1:22 ` [RFCv7 PATCH 02/10] cpufreq: introduce cpufreq_driver_is_slow Steve Muckle
2016-02-23 1:31 ` Rafael J. Wysocki
2016-02-26 0:50 ` Michael Turquette
2016-02-26 1:07 ` Steve Muckle
2016-02-26 1:16 ` Rafael J. Wysocki
[not found] ` <20160226185503.2278.20479@quark.deferred.io>
2016-02-26 21:00 ` Rafael J. Wysocki
2016-02-23 1:22 ` [RFCv7 PATCH 03/10] sched: scheduler-driven cpu frequency selection Steve Muckle
2016-02-25 3:55 ` Rafael J. Wysocki
2016-02-25 9:21 ` Peter Zijlstra
2016-02-25 21:04 ` Rafael J. Wysocki
2016-02-25 9:28 ` Peter Zijlstra [this message]
2016-02-25 21:08 ` Rafael J. Wysocki
2016-02-26 9:18 ` Peter Zijlstra
2016-02-27 0:08 ` Rafael J. Wysocki
2016-03-01 12:57 ` Peter Zijlstra
2016-03-01 19:44 ` Rafael J. Wysocki
2016-02-25 11:04 ` Rafael J. Wysocki
2016-02-26 0:34 ` Steve Muckle
2016-02-27 2:39 ` Rafael J. Wysocki
2016-02-27 4:17 ` Steve Muckle
2016-02-28 2:26 ` Rafael J. Wysocki
2016-03-01 14:31 ` Peter Zijlstra
2016-03-01 20:32 ` Rafael J. Wysocki
2016-03-01 13:26 ` Peter Zijlstra
2016-03-01 13:17 ` Peter Zijlstra
2016-03-02 7:49 ` Michael Turquette
2016-03-03 2:49 ` Rafael J. Wysocki
2016-03-03 3:50 ` Steve Muckle
2016-03-03 9:34 ` Juri Lelli
2016-03-03 13:03 ` Peter Zijlstra
2016-03-03 14:21 ` Ingo Molnar
2016-02-23 1:22 ` [RFCv7 PATCH 04/10] sched/fair: add triggers for OPP change requests Steve Muckle
2016-03-01 6:51 ` Ricky Liang
2016-03-03 3:55 ` Steve Muckle
2016-02-23 1:22 ` [RFCv7 PATCH 05/10] sched/{core,fair}: trigger OPP change request on fork() Steve Muckle
2016-02-23 1:22 ` [RFCv7 PATCH 06/10] sched/fair: cpufreq_sched triggers for load balancing Steve Muckle
2016-02-23 1:22 ` [RFCv7 PATCH 07/10] sched/fair: jump to max OPP when crossing UP threshold Steve Muckle
2016-02-23 1:22 ` [RFCv7 PATCH 08/10] sched: remove call of sched_avg_update from sched_rt_avg_update Steve Muckle
2016-02-23 1:22 ` [RFCv7 PATCH 09/10] sched/deadline: split rt_avg in 2 distincts metrics Steve Muckle
2016-02-23 1:22 ` [RFCv7 PATCH 10/10] sched: rt scheduler sets capacity requirement Steve Muckle
2016-02-23 1:33 ` [RFCv7 PATCH 00/10] sched: scheduler-driven CPU frequency selection Steve Muckle
2016-03-30 0:45 ` Yuyang Du
2016-03-31 1:35 ` Steve Muckle
2016-03-30 20:22 ` Yuyang Du
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=20160225092837.GD6357@twins.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=Juri.Lelli@arm.com \
--cc=dietmar.eggemann@arm.com \
--cc=jcliang@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=morten.rasmussen@arm.com \
--cc=mturquette@baylibre.com \
--cc=patrick.bellasi@arm.com \
--cc=rafael@kernel.org \
--cc=rjw@rjwysocki.net \
--cc=steve.muckle@linaro.org \
--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 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).