From: Vincent Guittot <vincent.guittot@linaro.org>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>,
linux-kernel <linux-kernel@vger.kernel.org>,
Matt Fleming <matt@codeblueprint.co.uk>,
Morten Rasmussen <Morten.Rasmussen@arm.com>,
Dietmar Eggemann <dietmar.eggemann@arm.com>,
Wanpeng Li <kernellwp@gmail.com>,
yuyang.du@intel.comc, Mike Galbraith <umgwanakikbuti@gmail.com>
Subject: Re: [PATCH 2/2 v3] sched: use load_avg for selecting idlest group
Date: Fri, 9 Dec 2016 17:28:33 +0100 [thread overview]
Message-ID: <CAKfTPtAYJ_6C0=GZJoO-UQz=5GRQuf=6YXVBLvG6pojp9CDEdg@mail.gmail.com> (raw)
In-Reply-To: <20161209152242.GU3124@twins.programming.kicks-ass.net>
On 9 December 2016 at 16:22, Peter Zijlstra <peterz@infradead.org> wrote:
> On Thu, Dec 08, 2016 at 05:56:54PM +0100, Vincent Guittot wrote:
>> @@ -5449,14 +5456,32 @@ find_idlest_group(struct sched_domain *sd, struct task_struct *p,
>> }
>>
>> /* Adjust by relative CPU capacity of the group */
>> - avg_load = (avg_load * SCHED_CAPACITY_SCALE) / group->sgc->capacity;
>> + avg_load = (avg_load * SCHED_CAPACITY_SCALE) /
>> + group->sgc->capacity;
>> + runnable_load = (runnable_load * SCHED_CAPACITY_SCALE) /
>> + group->sgc->capacity;
>>
>> if (local_group) {
>> - this_load = avg_load;
>> + this_runnable_load = runnable_load;
>> + this_avg_load = avg_load;
>> this_spare = max_spare_cap;
>> } else {
>> - if (avg_load < min_load) {
>> - min_load = avg_load;
>> + if (min_runnable_load > (runnable_load + imbalance)) {
>> + /*
>> + * The runnable load is significantly smaller
>> + * so we can pick this new cpu
>> + */
>> + min_runnable_load = runnable_load;
>> + min_avg_load = avg_load;
>> + idlest = group;
>> + } else if ((runnable_load < (min_runnable_load + imbalance)) &&
>> + (100*min_avg_load > imbalance_scale*avg_load)) {
>> + /*
>> + * The runnable loads are close so we take
>> + * into account blocked load through avg_load
>> + * which is blocked + runnable load
>> + */
>> + min_avg_load = avg_load;
>> idlest = group;
>> }
>>
>> @@ -5480,13 +5505,16 @@ find_idlest_group(struct sched_domain *sd, struct task_struct *p,
>> goto skip_spare;
>>
>> if (this_spare > task_util(p) / 2 &&
>> - imbalance*this_spare > 100*most_spare)
>> + imbalance_scale*this_spare > 100*most_spare)
>> return NULL;
>> else if (most_spare > task_util(p) / 2)
>> return most_spare_sg;
>>
>> skip_spare:
>> - if (!idlest || 100*this_load < imbalance*min_load)
>> + if (!idlest ||
>> + (min_runnable_load > (this_runnable_load + imbalance)) ||
>> + ((this_runnable_load < (min_runnable_load + imbalance)) &&
>> + (100*this_avg_load < imbalance_scale*min_avg_load)))
>> return NULL;
>> return idlest;
>> }
>
> I did the below on top for readability.
Changes looks good to me
>
> ---
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -5469,17 +5469,16 @@ find_idlest_group(struct sched_domain *s
> if (min_runnable_load > (runnable_load + imbalance)) {
> /*
> * The runnable load is significantly smaller
> - * so we can pick this new cpu
> + * so we can pick this new cpu
> */
> min_runnable_load = runnable_load;
> min_avg_load = avg_load;
> idlest = group;
> } else if ((runnable_load < (min_runnable_load + imbalance)) &&
> - (100*min_avg_load > imbalance_scale*avg_load)) {
> + (100*min_avg_load > imbalance_scale*avg_load)) {
> /*
> - * The runnable loads are close so we take
> - * into account blocked load through avg_load
> - * which is blocked + runnable load
> + * The runnable loads are close so take the
> + * blocked load into account through avg_load.
> */
> min_avg_load = avg_load;
> idlest = group;
> @@ -5509,15 +5508,21 @@ find_idlest_group(struct sched_domain *s
> if (this_spare > task_util(p) / 2 &&
> imbalance_scale*this_spare > 100*most_spare)
> return NULL;
> - else if (most_spare > task_util(p) / 2)
> +
> + if (most_spare > task_util(p) / 2)
> return most_spare_sg;
>
> skip_spare:
> - if (!idlest ||
> - (min_runnable_load > (this_runnable_load + imbalance)) ||
> - ((this_runnable_load < (min_runnable_load + imbalance)) &&
> - (100*this_avg_load < imbalance_scale*min_avg_load)))
> + if (!idlest)
> + return NULL;
> +
> + if (min_runnable_load > (this_runnable_load + imbalance))
> return NULL;
> +
> + if ((this_runnable_load < (min_runnable_load + imbalance)) &&
> + (100*this_avg_load < imbalance_scale*min_avg_load))
> + return NULL;
> +
> return idlest;
> }
>
next prev parent reply other threads:[~2016-12-09 16:28 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-12-08 16:56 [PATCH 0/2 v3] sched: improve spread of tasks during fork Vincent Guittot
2016-12-08 16:56 ` [PATCH 1/2 v3] sched: fix find_idlest_group for fork Vincent Guittot
2016-12-09 13:18 ` Matt Fleming
2016-12-12 6:50 ` [tip:sched/core] sched/core: Fix find_idlest_group() " tip-bot for Vincent Guittot
2016-12-08 16:56 ` [PATCH 2/2 v3] sched: use load_avg for selecting idlest group Vincent Guittot
2016-12-09 13:22 ` Matt Fleming
2016-12-09 15:22 ` Peter Zijlstra
2016-12-09 16:28 ` Vincent Guittot [this message]
2016-12-12 6:51 ` [tip:sched/core] sched/core: Use " tip-bot for Vincent Guittot
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='CAKfTPtAYJ_6C0=GZJoO-UQz=5GRQuf=6YXVBLvG6pojp9CDEdg@mail.gmail.com' \
--to=vincent.guittot@linaro.org \
--cc=Morten.Rasmussen@arm.com \
--cc=dietmar.eggemann@arm.com \
--cc=kernellwp@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=matt@codeblueprint.co.uk \
--cc=mingo@kernel.org \
--cc=peterz@infradead.org \
--cc=umgwanakikbuti@gmail.com \
--cc=yuyang.du@intel.comc \
/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).