Re: [PATCH 2/2] sched_ext: Dump the stall CPU first in watchdog exit

[Changwoo Min]

Hi all,

Thanks for the feedback, Tejun and Andrea,

On 4/9/26 3:28 PM, Tejun Heo wrote:
> Hello,
>
> On Thu, Apr 09, 2026 at 07:52:23AM +0200, Andrea Righi wrote:
> > > Would it make sense to generalize this so that the exit record the CPU the
> > > exit is triggered on and always dump that CPU first? That should include
> > > stall case, is likely useful for different cases too and we don't have to
> > > add @stall_cpu to the exit functions.
> >
> > But if we record the current CPU the exit is triggered on, in
> > check_rq_for_timeouts() we would prioritize the watchdog worker's CPU instead of
> > the stalled one, right?
>
> Ah, right. Can we at least generalize it to culprit or offending CPU? For
> regular scx_error() triggers, we'd want to see what that particular CPU was
> doing too.

We can generalize to prioritize the CPU that was caught by the watchdog
or called scx_exit(). Regarding the field name, I think exit_cpu makes
sense, meaning a CPU that makes the scheduler exit.

I will send out v2 soon.

Regards,
Changwoo Min