Re: [PATCH] sched_ext: Refresh idle state when kicking CPUs

[Andrea Righi <arighi@nvidia.com>]

On Thu, Jan 02, 2025 at 12:33:06PM -1000, Tejun Heo wrote:
> Hello,
> 
> On Wed, Jan 01, 2025 at 07:24:49PM +0100, Andrea Righi wrote:
> ...
> > @@ -6104,6 +6117,7 @@ static bool kick_one_cpu(s32 cpu, struct rq *this_rq, unsigned long *pseqs)
> >  			should_wait = true;
> >  		}
> >  
> > +		refresh_idle_state_on_kick(rq);
> >  		resched_curr(rq);
> 
> I'm not sure this is quite correct. e.g. This can cause multiple
> back-to-back busy->busy transitions and can incorrectly assert idle when the
> CPU ends up running non-idle tasks afterwards.

Right, we solve one issue, but it might result in some CPUs being marked as
idle even though they're actively running a task.

> 
> When the put_prev/set_next paths were reorganized, we lost the signal on the
> CPU re-entering idle from idle. However, that signal is still available if
> we hook into idle_class->pick_task(), right? So, if we move
> update_idle(true) call there and make sure that we don't generate an event
> on busy->busy transitions, we should be able to restore the previous
> behavior?

Which is basically what I did here:
https://lore.kernel.org/lkml/20241015111539.12136-1-andrea.righi@linux.dev/

We didn't fully like this, because it'd introduce unbalanced transitions,
as update_idle(cpu, true) can be generated multiple times. But it's
probably fine, at the end we would just restore the original behavior and
it'd allow to solve both the "pick_idle + kick CPU" and the "kick from
update_idle()" scenarios.

If we like this approach I can send a new patch updating the comment to
better clarify the scenarios that we are trying to solve. What do you
think?

Thanks,
-Andrea