From: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
To: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Peter Zijlstra <peterz@infradead.org>,
Mel Gorman <mgorman@techsingularity.net>,
Mike Galbraith <efault@gmx.de>,
Matt Fleming <matt@codeblueprint.co.uk>,
LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 4/4] sched/fair: Use a recently used CPU as an idle candidate and the basis for SIS
Date: Fri, 02 Feb 2018 06:54:24 -0800 [thread overview]
Message-ID: <1517583264.18051.60.camel@linux.intel.com> (raw)
In-Reply-To: <2447536.u3g27UoP4q@aspire.rjw.lan>
On Fri, 2018-02-02 at 12:00 +0100, Rafael J. Wysocki wrote:
> On Thursday, February 1, 2018 2:18:12 PM CET Srinivas Pandruvada
> wrote:
> >
> > On Thu, 2018-02-01 at 10:11 +0100, Peter Zijlstra wrote:
> > >
> > > On Thu, Feb 01, 2018 at 08:50:28AM +0100, Rafael J. Wysocki
> > > wrote:
> > > >
> > > >
> > > > On Wednesday, January 31, 2018 11:17:10 AM CET Peter Zijlstra
> > > > wrote:
> > > > >
> > > > >
> > > > > On Wed, Jan 31, 2018 at 10:22:49AM +0100, Rafael J. Wysocki
> > > > > wrote:
> > > > > >
> > > > > >
> > > > > > On Tuesday, January 30, 2018 2:15:31 PM CET Peter Zijlstra
> > > > > > wrote:
> > > > > >
> > > > > > >
> > > > > > >
> > > > > > > IA32_HWP_REQUEST has "Minimum_Performance",
> > > > > > > "Maximum_Performance" and
> > > > > > > "Desired_Performance" fields which can be used to give
> > > > > > > explicit
> > > > > > > frequency hints. And we really _should_ be doing that.
> > > > > > >
> > > > > > > Because, esp. in this scenario; a task migrating; the
> > > > > > > hardware really
> > > > > > > can't do anything sensible, whereas the OS _knows_.
> > > > > > But IA32_HWP_REQUEST is not a cheap MSR to write to.
> > > > > That just means we might need to throttle writing to it, like
> > > > > it
> > > > > already
> > > > > does for the regular pstate (PERF_CTRL) msr in any case
> > > > > (also, is
> > > > > that a
> > > > > cheap msr?)
> > > > >
> > > > > Not touching it at all seems silly.
> > > > OK
> > > >
> > > > So what field precisely would you touch? "desired"? If so,
> > > > does
> > > > that actually
> > > > guarantee anything to happen?
> > > No idea, desired would be the one I would start with, it matches
> > > with
> > > the intent here. But I've no idea what our current HWP
> > > implementation
> > > actually does with it.
> > Desired !=0 will disable HWP autonomous mode of frequency
> > selection.
> But I don't think it will just run at "desired" then, will it?
HWP all are these hints only not a guarantee.
There are totally different way HWP is handled in client an servers.
If you set desired all heuristics they collected will be dumped, so
they suggest don't set desired when you are in autonomous mode. If we
really want a boost set the EPP. We know that EPP makes lots of
measurable difference.
next prev parent reply other threads:[~2018-02-02 14:54 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-01-30 10:45 [PATCH 0/4] Reduce migrations and unnecessary spreading of load to multiple CPUs Mel Gorman
2018-01-30 10:45 ` [PATCH 1/4] sched/fair: Remove unnecessary parameters from wake_affine_idle Mel Gorman
2018-02-06 11:55 ` [tip:sched/urgent] sched/fair: Remove unnecessary parameters from wake_affine_idle() tip-bot for Mel Gorman
2018-01-30 10:45 ` [PATCH 2/4] sched/fair: Restructure wake_affine to return a CPU id Mel Gorman
2018-02-06 11:56 ` [tip:sched/urgent] sched/fair: Restructure wake_affine*() " tip-bot for Mel Gorman
2018-01-30 10:45 ` [PATCH 3/4] sched/fair: Do not migrate if the prev_cpu is idle Mel Gorman
2018-02-06 11:56 ` [tip:sched/urgent] " tip-bot for Mel Gorman
2018-01-30 10:45 ` [PATCH 4/4] sched/fair: Use a recently used CPU as an idle candidate and the basis for SIS Mel Gorman
2018-01-30 11:50 ` Peter Zijlstra
2018-01-30 12:57 ` Mel Gorman
2018-01-30 13:15 ` Peter Zijlstra
2018-01-30 13:25 ` Mel Gorman
2018-01-30 13:40 ` Peter Zijlstra
2018-01-30 14:06 ` Mel Gorman
2018-01-31 9:22 ` Rafael J. Wysocki
2018-01-31 10:17 ` Peter Zijlstra
2018-01-31 11:54 ` Mel Gorman
2018-01-31 17:44 ` Srinivas Pandruvada
2018-02-01 9:11 ` Peter Zijlstra
2018-02-01 7:50 ` Rafael J. Wysocki
2018-02-01 9:11 ` Peter Zijlstra
2018-02-01 13:18 ` Srinivas Pandruvada
2018-02-02 11:00 ` Rafael J. Wysocki
2018-02-02 14:54 ` Srinivas Pandruvada [this message]
2018-02-02 19:48 ` Mel Gorman
2018-02-02 20:01 ` Srinivas Pandruvada
2018-02-05 11:10 ` Mel Gorman
2018-02-05 17:04 ` Srinivas Pandruvada
2018-02-05 17:50 ` Mel Gorman
2018-02-04 8:42 ` Rafael J. Wysocki
2018-02-04 8:38 ` Rafael J. Wysocki
2018-02-02 11:42 ` Rafael J. Wysocki
2018-02-02 12:46 ` Peter Zijlstra
2018-02-02 12:55 ` Peter Zijlstra
2018-02-02 14:08 ` Peter Zijlstra
2018-02-03 16:30 ` Srinivas Pandruvada
2018-02-05 10:44 ` Peter Zijlstra
2018-02-05 10:58 ` Ingo Molnar
2018-02-02 12:58 ` Peter Zijlstra
2018-02-02 13:27 ` Mel Gorman
2018-01-30 13:15 ` Mike Galbraith
2018-01-30 13:25 ` Peter Zijlstra
2018-01-30 13:35 ` Mike Galbraith
2018-01-30 11:53 ` Peter Zijlstra
2018-01-30 12:59 ` Mel Gorman
2018-01-30 13:06 ` Peter Zijlstra
2018-01-30 13:18 ` Mel Gorman
2018-02-06 11:56 ` [tip:sched/urgent] " tip-bot for 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=1517583264.18051.60.camel@linux.intel.com \
--to=srinivas.pandruvada@linux.intel.com \
--cc=efault@gmx.de \
--cc=linux-kernel@vger.kernel.org \
--cc=matt@codeblueprint.co.uk \
--cc=mgorman@techsingularity.net \
--cc=peterz@infradead.org \
--cc=rjw@rjwysocki.net \
/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).