Re: [RFC][PATCH 2/3] sched: Add support to pick functions to take rf

[Vincent Guittot]

On Mon, 13 Oct 2025 at 13:09, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>
> On Mon, Oct 13, 2025 at 01:04:49PM +0200, Peter Zijlstra wrote:
>
> > Bah; so yeah, this new behaviour is better for indeed always calling
> > newidle when it is needed, but you're also right that in case of ext
> > this might not be ideal.
> >
> > So I have a pile of newidle hacks here:
> >
> >   https://lkml.kernel.org/r/20251010170937.GG4067720@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> >
> > and while I don't particularly like NI_SPARE (the has_spare_tasks thing
> > is fickle); the idea seems to have some merit for this situation --
> > where we know we'll not be having fair tasks at all.
> >
> > I mean, we can always do something like this to sched_balance_newidle():
> >
> >       if (scx_switched_all())
> >               return 0;
> >
> > Not pretty, but should do the job.
>
> Oh, never mind, none of this is needed.
>
> __pick_next_task()
>
>         if (scx_enabled())
>           goto restart;
>
>         ...
> restart:
>         for_each_active_class(class) {
>                 ...
>         }
>
>
> And then we have next_active_class() skip fair_sched_class entirely when
> scx_switch_all().

Ah yes you're right. fair is not called in case of scx_switched_all()

>
> So in the common ext case, we'll not hit pick_next_task_fair() at all.