[Ksummit-2013-discuss] [ATTEND] CPU power management (cpufreq, cpuidle) integration with the scheduler

[Vincent Guittot]

On 11 July 2013 12:55, James Bottomley
<James.Bottomley at hansenpartnership.com> wrote:
> On Thu, 2013-07-11 at 12:08 +0200, Vincent Guittot wrote:
>> Hi,
>>
>> I'm working on the scheduler and the power management for Linaro and
>> especially on the activity of packing tasks (
>> http://lwn.net/Articles/548522/ for the latest version and my ongoing
>> work is available here
>> http://git.linaro.org/gitweb?p=people/vingu/kernel.git;a=shortlog;h=refs/heads/linux-linaro-sched
>> ).
>> With Daniel who is co-maintainer of cpuidle, we are interested to
>> discuss the integration of CPU power framework into the scheduler.
>>
>> Please consider our participation to the discussion
>
> Just for the benefit of others who might think about doing this, you
> just posted the textbook how not to have the programme committee
> consider your request for a debate.
>
> The patch set you point to is old (24 April), but the git tree has
> clearly been rebased (commit dates are 21 June).  So obviously you're
> not telling us something.  The something you're not telling us is
> probably this:
>
> http://thread.gmane.org/gmane.linux.kernel/1499202/
>
> The final word of which seems to be a request for a rewrite and a NAK
> from the scheduler maintainers to the patch set (even the obviously
> rebased one).  There's nothing for KS to discuss here (yet) ... if you
> got the impression it's the forum to override a maintainer NAK, it's
> not.
>

Sorry for the misunderstood but my goal was not to override a NAK but
to express my interest on the subject. We are currently working on
taking into account the comments of the thread that you have mentioned
above but such modification are not available in the git tree yet.

Vincent

> What needs to happen now is that you need to redo your thinking taking
> into account the feedback on the thread you didn't mention.  If after a
> rework, a repost to the lists and a healthy discussion on the list,
> there are still disagreements that might benefit from personal instead
> of mailing list interaction, *then* there might be something KS could
> help with (although probably not in the plenary setting).
>
> James
>
>