linux-kernel - Re: [PATCH] sched/fair: check for idle core

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]

Message-ID: <2376963.UiH3HBYXtl@kreacher>
Date:   Wed, 21 Oct 2020 22:25:59 +0200
From:   "Rafael J. Wysocki" <rjw@...ysocki.net>
To:     Julia Lawall <julia.lawall@...ia.fr>
Cc:     Peter Zijlstra <peterz@...radead.org>,
        Mel Gorman <mgorman@...e.de>, Ingo Molnar <mingo@...hat.com>,
        kernel-janitors@...r.kernel.org,
        Juri Lelli <juri.lelli@...hat.com>,
        Vincent Guittot <vincent.guittot@...aro.org>,
        Dietmar Eggemann <dietmar.eggemann@....com>,
        Steven Rostedt <rostedt@...dmis.org>,
        Ben Segall <bsegall@...gle.com>,
        Daniel Bristot de Oliveira <bristot@...hat.com>,
        linux-kernel@...r.kernel.org,
        Valentin Schneider <valentin.schneider@....com>,
        Gilles Muller <Gilles.Muller@...ia.fr>,
        viresh.kumar@...aro.org, srinivas.pandruvada@...ux.intel.com
Subject: Re: [PATCH] sched/fair: check for idle core

On Wednesday, October 21, 2020 9:47:51 PM CEST Julia Lawall wrote:
> 
> On Wed, 21 Oct 2020, Rafael J. Wysocki wrote:
> 
> > On Wednesday, October 21, 2020 2:42:20 PM CEST Julia Lawall wrote:
> > >
> > > On Wed, 21 Oct 2020, Peter Zijlstra wrote:
> > >
> > > > On Wed, Oct 21, 2020 at 01:56:55PM +0200, Julia Lawall wrote:
> > > > > Prior to 5.8, my machine was using intel_pstate and had few background
> > > > > tasks.  Thus the problem wasn't visible in practice.  Starting with 5.8
> > > > > the kernel decided that intel_cpufreq would be more appropriate, which
> > > > > introduced kworkers every 0.004 seconds on all cores.
> > > >
> > > > That still doesn't make any sense. Are you running the legacy on-demand
> > > > thing or something?
> > > >
> > > > Rafael, Srinivas, Viresh, how come it defaults to that?
> > >
> > > The relevant commits are 33aa46f252c7, and 39a188b88332 that fixes a small
> > > bug.  I have a Intel(R) Xeon(R) CPU E7-8870 v4 @ 2.10GHz that does not
> > > have the HWP feature, even though the cores seemed to be able to change
> > > their frequencies at the hardware level.
> >
> > That's in the range of "turbo" P-states (if a P-state above a certain threshold
> > is requested by the governor, the processor has a license to choose P-states
> > in the range above this threshold by itself).
> 
> Sorry, but I don't understand this answer at all.

Well, sorry about that and let me rephrase then.

Contemporary CPUs have two ranges of P-states, the so called "guaranteed
performance" range and the "turbo" range.

In the "guaranteed performance" range the CPU runs in the P-state requested
by the governor, unless a higher P-state has been requested for another CPU in
its frequency domain (usually covering the entire processor package) , in which
case that higher P-state will be used (the effective P-state for all CPUs in the
frequency domain is the maximum of all P-states requested for individual CPUs).

However, if the governor requests a P-state from the "turbo" range, the
processor is not required to take that request literally and the PM unit in it may
override the governor's choice and cause the CPU to run in a different P-state
(also from the "turbo" range), even if lower P-states have been requested for
the other CPUs in the processor package.

This is also explained in Documentation/admin-guide/pm/intel_pstate.rst (in the
Turbo P-states Support section), in more detail and hopefully more clearly.

Cheers!