From: Peter Zijlstra <peterz@infradead.org>
To: Alex Shi <alex.shi@intel.com>
Cc: torvalds@linux-foundation.org, mingo@redhat.com,
tglx@linutronix.de, akpm@linux-foundation.org,
arjan@linux.intel.com, bp@alien8.de, pjt@google.com,
namhyung@kernel.org, efault@gmx.de, vincent.guittot@linaro.org,
gregkh@linuxfoundation.org, preeti@linux.vnet.ibm.com,
viresh.kumar@linaro.org, linux-kernel@vger.kernel.org,
morten.rasmussen@arm.com
Subject: Re: [patch v5 09/15] sched: add power aware scheduling in fork/exec/wake
Date: Fri, 22 Feb 2013 09:54:39 +0100 [thread overview]
Message-ID: <1361523279.26780.45.camel@laptop> (raw)
In-Reply-To: <512631C7.8060103@intel.com>
On Thu, 2013-02-21 at 22:40 +0800, Alex Shi wrote:
> > The name is a secondary issue, first you need to explain why you
> think
> > nr_running is a useful metric at all.
> >
> > You can have a high nr_running and a low utilization (a burst of
> > wakeups, each waking a process that'll instantly go to sleep again),
> or
> > low nr_running and high utilization (a single process cpu bound
> > process).
>
> It is true in periodic balance. But in fork/exec/waking timing, the
> incoming processes usually need to do something before sleep again.
You'd be surprised, there's a fair number of workloads that have
negligible runtime on wakeup.
> I use nr_running to measure how the group busy, due to 3 reasons:
> 1, the current performance policy doesn't use utilization too.
We were planning to fix that now that its available.
> 2, the power policy don't care load weight.
Then its broken, it should very much still care about weight.
> 3, I tested some benchmarks, kbuild/tbench/hackbench/aim7 etc, some
> benchmark results looks clear bad when use utilization. if my memory
> right, the hackbench/aim7 both looks bad. I had tried many ways to
> engage utilization into this balance, like use utilization only, or
> use
> utilization * nr_running etc. but still can not find a way to recover
> the lose. But with nr_running, the performance seems doesn't lose much
> with power policy.
You're failing to explain why utilization performs bad and you don't
explain why nr_running is better. That things work simply isn't good
enough, you have to have at least a general idea (but much preferable a
very good idea) _why_ things work.
next prev parent reply other threads:[~2013-02-22 8:54 UTC|newest]
Thread overview: 90+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-02-18 5:07 [patch v5 0/15] power aware scheduling Alex Shi
2013-02-18 5:07 ` [patch v5 01/15] sched: set initial value for runnable avg of sched entities Alex Shi
2013-02-18 8:28 ` Joonsoo Kim
2013-02-18 9:16 ` Alex Shi
2013-02-18 5:07 ` [patch v5 02/15] sched: set initial load avg of new forked task Alex Shi
2013-02-20 6:20 ` Alex Shi
2013-02-24 10:57 ` Preeti U Murthy
2013-02-25 6:00 ` Alex Shi
2013-02-28 7:03 ` Preeti U Murthy
2013-02-25 7:12 ` Alex Shi
2013-02-18 5:07 ` [patch v5 03/15] Revert "sched: Introduce temporary FAIR_GROUP_SCHED dependency for load-tracking" Alex Shi
2013-02-18 5:07 ` [patch v5 04/15] sched: add sched balance policies in kernel Alex Shi
2013-02-20 9:37 ` Ingo Molnar
2013-02-20 13:40 ` Alex Shi
2013-02-20 15:41 ` Ingo Molnar
2013-02-21 1:43 ` Alex Shi
2013-02-18 5:07 ` [patch v5 05/15] sched: add sysfs interface for sched_balance_policy selection Alex Shi
2013-02-18 5:07 ` [patch v5 06/15] sched: log the cpu utilization at rq Alex Shi
2013-02-20 9:30 ` Peter Zijlstra
2013-02-20 12:09 ` Preeti U Murthy
2013-02-20 13:34 ` Peter Zijlstra
2013-02-20 14:36 ` Alex Shi
2013-02-20 14:33 ` Alex Shi
2013-02-20 15:20 ` Peter Zijlstra
2013-02-21 1:35 ` Alex Shi
2013-02-20 15:22 ` Peter Zijlstra
2013-02-25 2:26 ` Alex Shi
2013-03-22 8:49 ` Alex Shi
2013-02-20 12:19 ` Preeti U Murthy
2013-02-20 12:39 ` Alex Shi
2013-02-18 5:07 ` [patch v5 07/15] sched: add new sg/sd_lb_stats fields for incoming fork/exec/wake balancing Alex Shi
2013-02-20 9:38 ` Peter Zijlstra
2013-02-20 12:27 ` Alex Shi
2013-02-18 5:07 ` [patch v5 08/15] sched: move sg/sd_lb_stats struct ahead Alex Shi
2013-02-18 5:07 ` [patch v5 09/15] sched: add power aware scheduling in fork/exec/wake Alex Shi
2013-02-20 9:42 ` Peter Zijlstra
2013-02-20 12:09 ` Alex Shi
2013-02-20 13:36 ` Peter Zijlstra
2013-02-20 14:23 ` Alex Shi
2013-02-21 13:33 ` Peter Zijlstra
2013-02-21 14:40 ` Alex Shi
2013-02-22 8:54 ` Peter Zijlstra [this message]
2013-02-24 9:27 ` Alex Shi
2013-02-24 9:49 ` Preeti U Murthy
2013-02-24 11:55 ` Alex Shi
2013-02-24 17:51 ` Preeti U Murthy
2013-02-25 2:23 ` Alex Shi
2013-02-25 3:23 ` Mike Galbraith
2013-02-25 9:53 ` Alex Shi
2013-02-25 10:30 ` Mike Galbraith
2013-02-18 5:07 ` [patch v5 10/15] sched: packing transitory tasks in wake/exec power balancing Alex Shi
2013-02-18 8:44 ` Joonsoo Kim
2013-02-18 8:56 ` Alex Shi
2013-02-20 5:55 ` Alex Shi
2013-02-20 7:40 ` Mike Galbraith
2013-02-20 8:11 ` Alex Shi
2013-02-20 8:43 ` Mike Galbraith
2013-02-20 8:54 ` Alex Shi
2013-02-18 5:07 ` [patch v5 11/15] sched: add power/performance balance allow flag Alex Shi
2013-02-20 9:48 ` Peter Zijlstra
2013-02-20 12:04 ` Alex Shi
2013-02-20 13:37 ` Peter Zijlstra
2013-02-20 13:48 ` Peter Zijlstra
2013-02-20 14:08 ` Alex Shi
2013-02-20 13:52 ` Alex Shi
2013-02-20 12:12 ` Borislav Petkov
2013-02-20 14:20 ` Alex Shi
2013-02-20 15:22 ` Borislav Petkov
2013-02-21 1:32 ` Alex Shi
2013-02-21 9:42 ` Borislav Petkov
2013-02-21 14:52 ` Alex Shi
2013-02-18 5:07 ` [patch v5 12/15] sched: pull all tasks from source group Alex Shi
2013-02-18 5:07 ` [patch v5 13/15] sched: no balance for prefer_sibling in power scheduling Alex Shi
2013-02-18 5:07 ` [patch v5 14/15] sched: power aware load balance Alex Shi
2013-03-20 4:57 ` Preeti U Murthy
2013-03-21 7:43 ` Alex Shi
2013-03-21 8:41 ` Preeti U Murthy
2013-03-21 9:27 ` Alex Shi
2013-03-21 10:27 ` Preeti U Murthy
2013-03-22 1:30 ` Alex Shi
2013-03-22 5:14 ` Preeti U Murthy
2013-03-25 4:52 ` Alex Shi
2013-03-29 12:42 ` Preeti U Murthy
2013-03-29 13:39 ` Alex Shi
2013-03-30 11:25 ` Preeti U Murthy
2013-03-30 14:04 ` Alex Shi
2013-03-30 15:31 ` Preeti U Murthy
2013-02-18 5:07 ` [patch v5 15/15] sched: lazy power balance Alex Shi
2013-02-18 7:44 ` [patch v5 0/15] power aware scheduling Alex Shi
2013-02-19 12:08 ` Paul Turner
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=1361523279.26780.45.camel@laptop \
--to=peterz@infradead.org \
--cc=akpm@linux-foundation.org \
--cc=alex.shi@intel.com \
--cc=arjan@linux.intel.com \
--cc=bp@alien8.de \
--cc=efault@gmx.de \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=morten.rasmussen@arm.com \
--cc=namhyung@kernel.org \
--cc=pjt@google.com \
--cc=preeti@linux.vnet.ibm.com \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=vincent.guittot@linaro.org \
--cc=viresh.kumar@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).