From: Saravana Kannan <skannan@codeaurora.org>
To: Ingo Molnar <mingo@elte.hu>
Cc: linaro-kernel@lists.linaro.org,
Russell King <linux@arm.linux.org.uk>,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
Nicolas Pitre <nico@fluxnic.net>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Oleg Nesterov <oleg@redhat.com>,
cpufreq@vger.kernel.org, linux-kernel@vger.kernel.org,
Anton Vorontsov <anton.vorontsov@linaro.org>,
"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
Mike Chan <mike@android.com>, Dave Jones <davej@redhat.com>,
Todd Poynor <toddpoynor@google.com>,
kernel-team@android.com, linux-arm-kernel@lists.infradead.org,
Arjan Van De Ven <arjan@infradead.org>
Subject: Re: [PATCH RFC 0/4] Scheduler idle notifiers and users
Date: Tue, 14 Feb 2012 15:20:46 -0800 [thread overview]
Message-ID: <4F3AEC4E.9000303@codeaurora.org> (raw)
In-Reply-To: <20120211144530.GA497@elte.hu>
On 02/11/2012 06:45 AM, Ingo Molnar wrote:
>
> * Saravana Kannan<skannan@codeaurora.org> wrote:
>
>> When you say accommodate all hardware, does it mean we will
>> keep around CPUfreq and allow attempts at improving it? Or we
>> will completely move to scheduler based CPU freq scaling, but
>> won't try to force atomicity? Say, may be queue up a
>> notification to a CPU driver to scale up the frequency as soon
>> as it can?
>
> I don't think we should (or even could) force atomicity - we
> adapt to whatever the hardware can do.
May be I misread the emails from Peter and you, but it sounded like the
idea being proposed was to directly do a freq change from the scheduler.
That would force the freq change API to be atomic (if it can be
implemented is another issue). That's what I was referring to when I
loosely used the terms "force atomicity".
> But the design should be directed at systems where frequency
> changes can be done in a reasonably fast manner. That is what he
> future is - any change we initiate today takes years to reach
> actual products/systems.
As long as the new design doesn't treat archs needing schedulable
context to set freq as a second class citizen, I think we would all be
happy. Because it's not just a matter of it being old hardware.
Sometimes the decision to let the SW do the voltage scaling also comes
down to HW cost. Considering Linux runs on such a wide set of archs, I
think we shouldn't treat the need for schedulable context for freq
setting as "broken" or "not sane".
>> IMHO, I think the problem with CPUfreq and its dynamic
>> governors today is that they do a timer based sampling of the
>> CPU load instead of getting some hints from the scheduler when
>> the scheduler knows that the load average is quite high.
>
> Yes - that is one of the "frequency changes are slow"
> assumptions - which is wrong.
Thanks,
Saravana
--
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
WARNING: multiple messages have this Message-ID (diff)
From: skannan@codeaurora.org (Saravana Kannan)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH RFC 0/4] Scheduler idle notifiers and users
Date: Tue, 14 Feb 2012 15:20:46 -0800 [thread overview]
Message-ID: <4F3AEC4E.9000303@codeaurora.org> (raw)
In-Reply-To: <20120211144530.GA497@elte.hu>
On 02/11/2012 06:45 AM, Ingo Molnar wrote:
>
> * Saravana Kannan<skannan@codeaurora.org> wrote:
>
>> When you say accommodate all hardware, does it mean we will
>> keep around CPUfreq and allow attempts at improving it? Or we
>> will completely move to scheduler based CPU freq scaling, but
>> won't try to force atomicity? Say, may be queue up a
>> notification to a CPU driver to scale up the frequency as soon
>> as it can?
>
> I don't think we should (or even could) force atomicity - we
> adapt to whatever the hardware can do.
May be I misread the emails from Peter and you, but it sounded like the
idea being proposed was to directly do a freq change from the scheduler.
That would force the freq change API to be atomic (if it can be
implemented is another issue). That's what I was referring to when I
loosely used the terms "force atomicity".
> But the design should be directed at systems where frequency
> changes can be done in a reasonably fast manner. That is what he
> future is - any change we initiate today takes years to reach
> actual products/systems.
As long as the new design doesn't treat archs needing schedulable
context to set freq as a second class citizen, I think we would all be
happy. Because it's not just a matter of it being old hardware.
Sometimes the decision to let the SW do the voltage scaling also comes
down to HW cost. Considering Linux runs on such a wide set of archs, I
think we shouldn't treat the need for schedulable context for freq
setting as "broken" or "not sane".
>> IMHO, I think the problem with CPUfreq and its dynamic
>> governors today is that they do a timer based sampling of the
>> CPU load instead of getting some hints from the scheduler when
>> the scheduler knows that the load average is quite high.
>
> Yes - that is one of the "frequency changes are slow"
> assumptions - which is wrong.
Thanks,
Saravana
--
Sent by an employee of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.
next prev parent reply other threads:[~2012-02-14 23:20 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-08 1:39 [PATCH RFC 0/4] Scheduler idle notifiers and users Anton Vorontsov
2012-02-08 1:39 ` Anton Vorontsov
2012-02-08 1:41 ` [PATCH 1/4] sched: Introduce idle notifiers API Anton Vorontsov
2012-02-08 1:41 ` Anton Vorontsov
2012-02-08 1:43 ` [PATCH 2/4] sched: Wire up idle notifiers Anton Vorontsov
2012-02-08 1:43 ` Anton Vorontsov
2012-02-08 1:44 ` [PATCH 3/4] cpufreq: New 'interactive' governor Anton Vorontsov
2012-02-08 1:44 ` Anton Vorontsov
2012-02-08 23:00 ` Vincent Guittot
2012-02-08 23:00 ` Vincent Guittot
2012-02-09 0:32 ` Anton Vorontsov
2012-02-09 0:32 ` Anton Vorontsov
2012-02-08 1:44 ` [PATCH 4/4] ARM: Move leds idle start/stop calls to sched idle notifiers Anton Vorontsov
2012-02-08 1:44 ` Anton Vorontsov
2012-02-08 3:05 ` [PATCH RFC 0/4] Scheduler idle notifiers and users Peter Zijlstra
2012-02-08 3:05 ` Peter Zijlstra
2012-02-08 20:23 ` Dave Jones
2012-02-08 20:23 ` Dave Jones
2012-02-08 21:33 ` Benjamin Herrenschmidt
2012-02-08 21:33 ` Benjamin Herrenschmidt
2012-02-09 7:51 ` Ingo Molnar
2012-02-09 7:51 ` Ingo Molnar
2012-02-11 3:15 ` Saravana Kannan
2012-02-11 3:15 ` Saravana Kannan
2012-02-11 3:15 ` Saravana Kannan
2012-02-11 14:39 ` Mark Brown
2012-02-11 14:39 ` Mark Brown
2012-02-11 14:53 ` Peter Zijlstra
2012-02-11 14:53 ` Peter Zijlstra
2012-02-11 15:33 ` Mark Brown
2012-02-11 15:33 ` Mark Brown
2012-02-15 13:38 ` Peter Zijlstra
2012-02-15 13:38 ` Peter Zijlstra
2012-02-15 13:38 ` Peter Zijlstra
2012-02-15 16:04 ` Mark Brown
2012-02-15 16:04 ` Mark Brown
2012-02-12 21:33 ` Benjamin Herrenschmidt
2012-02-12 21:33 ` Benjamin Herrenschmidt
2012-02-11 14:45 ` Ingo Molnar
2012-02-11 14:45 ` Ingo Molnar
2012-02-14 23:20 ` Saravana Kannan [this message]
2012-02-14 23:20 ` Saravana Kannan
2012-02-15 13:38 ` Peter Zijlstra
2012-02-15 13:38 ` Peter Zijlstra
2012-02-15 13:38 ` Peter Zijlstra
2012-02-15 14:02 ` Russell King - ARM Linux
2012-02-15 14:02 ` Russell King - ARM Linux
2012-02-15 15:01 ` Peter Zijlstra
2012-02-15 15:01 ` Peter Zijlstra
2012-02-15 15:01 ` Peter Zijlstra
2012-02-15 16:00 ` Russell King - ARM Linux
2012-02-15 16:00 ` Russell King - ARM Linux
2012-02-15 16:09 ` Peter Zijlstra
2012-02-15 16:09 ` Peter Zijlstra
2012-02-16 3:31 ` Benjamin Herrenschmidt
2012-02-16 3:31 ` Benjamin Herrenschmidt
2012-02-16 10:14 ` Peter Zijlstra
2012-02-16 10:14 ` Peter Zijlstra
2012-02-17 9:00 ` Dominik Brodowski
2012-02-17 9:00 ` Dominik Brodowski
2012-02-20 11:03 ` Peter Zijlstra
2012-02-20 11:03 ` Peter Zijlstra
2012-02-21 12:38 ` Pantelis Antoniou
2012-02-21 12:38 ` Pantelis Antoniou
2012-02-21 12:38 ` Pantelis Antoniou
2012-02-21 12:56 ` Peter Zijlstra
2012-02-21 12:56 ` Peter Zijlstra
2012-02-21 13:31 ` Pantelis Antoniou
2012-02-21 13:31 ` Pantelis Antoniou
2012-02-21 14:52 ` Amit Kucheria
2012-02-21 14:52 ` Amit Kucheria
2012-02-21 17:06 ` Pantelis Antoniou
2012-02-21 17:06 ` Pantelis Antoniou
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=4F3AEC4E.9000303@codeaurora.org \
--to=skannan@codeaurora.org \
--cc=a.p.zijlstra@chello.nl \
--cc=anton.vorontsov@linaro.org \
--cc=arjan@infradead.org \
--cc=benh@kernel.crashing.org \
--cc=cpufreq@vger.kernel.org \
--cc=davej@redhat.com \
--cc=kernel-team@android.com \
--cc=linaro-kernel@lists.linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@arm.linux.org.uk \
--cc=mike@android.com \
--cc=mingo@elte.hu \
--cc=nico@fluxnic.net \
--cc=oleg@redhat.com \
--cc=paulmck@linux.vnet.ibm.com \
--cc=toddpoynor@google.com \
/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.