Re: [PATCH] sched: minor optimization

[Nikanth Karthikesan <knikanth@suse.de>]

> On Fri, Nov 23, 2007 at  5:48 PM, "Dmitry Adamushko"
<dmitry.adamushko@gmail.com> wrote:
> The only legitimate possibility of having the fair_sched_class
> returning no task in this case is when 'rq->nr_running ==
> rq->cfs.nr_running == 0'.

Yes, I think so.

> iow, a possible optimization would be just the following check :
>
> if (rq->nr_running == 0)
>         return idle_sched_class.pick_next_task(rq);
> at the beginning of pick_next_task().
>
> (or maybe put it at the beginning of the
> if (likely(rq->nr_running == rq->cfs.nr_running)) {} block as we
> already have 'likely()' there).
>

But that might add a test before the case we want to optimize the most.
I just thought of taking advantage of a case where we know
rq->nr_running==0, instead of throwing away that information.

Thanks
Nikanth