From mboxrd@z Thu Jan  1 00:00:00 1970
From: Glauber Costa <glommer-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org>
Subject: Re: [PATCH v3 1/6] measure exec_clock for rt sched entities
Date: Wed, 30 May 2012 14:32:07 +0400
Message-ID: <4FC5F727.2040804@parallels.com>
References: <1338371317-5980-1-git-send-email-glommer@parallels.com>  <1338371317-5980-2-git-send-email-glommer@parallels.com> <1338373759.26856.238.camel@twins>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Return-path: <cgroups-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
In-Reply-To: <1338373759.26856.238.camel@twins>
Sender: cgroups-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
List-ID: <cgroups.vger.kernel.org>
Content-Type: text/plain; charset="us-ascii"; format="flowed"
To: Peter Zijlstra <a.p.zijlstra-/NLkJaSkS4VmR6Xm/wNWPw@public.gmane.org>
Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, devel-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org, Paul Turner <pjt-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>, Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>, "Eric W. Biederman" <ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>, handai.szj-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, Andrew.Phillips-xheW4WVAX9Y@public.gmane.org, Serge Hallyn <serge.hallyn-Z7WLFzj8eWMS+FvcfC7Uqw@public.gmane.org>

On 05/30/2012 02:29 PM, Peter Zijlstra wrote:
> On Wed, 2012-05-30 at 13:48 +0400, Glauber Costa wrote:
>> For simetry with the cfq tasks, measure exec_clock for the rt
>> sched entities (rt_se).
>
> Symmetry methinks..

=p bad me
> anyway, where is the symmetry?, fair.c:update_curr()
> doesn't do the for_each_sched_entity() thing.

It does implicitly, because fair.c:update_curr() is called from
within enqueue_task(), that is called for_each_sched_entity in
enqueue_task_fair().

>
>> This can be used in a number of fashions. For instance, to
>> compute total cpu usage in a cgroup that is generated by
>> rt tasks.
>>
>> Signed-off-by: Glauber Costa<glommer-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org>
>> CC: Peter Zijlstra<a.p.zijlstra-/NLkJaSkS4VmR6Xm/wNWPw@public.gmane.org>
>> CC: Paul Turner<pjt-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
>> ---
>>   kernel/sched/rt.c    |    5 +++++
>>   kernel/sched/sched.h |    1 +
>>   2 files changed, 6 insertions(+)
>>
>> diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
>> index c5565c3..30ee4e2 100644
>> --- a/kernel/sched/rt.c
>> +++ b/kernel/sched/rt.c
>> @@ -919,6 +919,11 @@ static void update_curr_rt(struct rq *rq)
>>
>>   	sched_rt_avg_update(rq, delta_exec);
>>
>> +	for_each_sched_rt_entity(rt_se) {
>> +		rt_rq = rt_rq_of_se(rt_se);
>> +		schedstat_add(rt_rq, exec_clock, delta_exec);
>> +	}
>> +
>>   	if (!rt_bandwidth_enabled())
>>   		return;
>
> See, this just makes me sad.. you now have a double
> for_each_sched_rt_entity() loop.

The way I read the rt.c code, it it is called from enqueue_task_rt only 
once.