Re: [PATCH v3 15/20] sched/core: Introduce a simple steal monitor

[Shrikanth Hegde]

On 6/8/26 9:55 PM, Ilya Leoshkevich wrote:
> On 5/14/26 17:21, Shrikanth Hegde wrote:
> > Start with a simple steal monitor.
> >
> > It is meant to look at steal time and make the decision to
> > reduce/increase the preferred CPUs.
> >
> > It has
> > - work function to execute the steal time calculations and decision
> >    making periodically.
> > - temporary cpumask, which will be used in the work function. This helps
> >    to avoid cpumask allocation in periodic work function.
> > - low and high thresholds for steal time.
> > - sampling period to control the frequency of steal time calculations.
> > - cache the previous decision to avoid oscillations
> >
> > Signed-off-by: Shrikanth Hegde <sshegde@xxxxxxxxxxxxx>
> > ---
> >   include/linux/sched.h | 13 +++++++++++++
> >   kernel/sched/core.c   | 24 ++++++++++++++++++++++++
> >   kernel/sched/sched.h  |  3 +++
> >   3 files changed, 40 insertions(+)
> >
> > diff --git a/include/linux/sched.h b/include/linux/sched.h
> > index dcfb57c90850..ee5f19a96118 100644
> > --- a/include/linux/sched.h
> > +++ b/include/linux/sched.h
>
> [...]
>
> > @@ -11351,4 +11353,26 @@ void 
> > sched_push_current_non_preferred_cpu(struct rq *rq)
> >                   push_task, this_cpu_ptr(&npc_push_task_work));
> >       local_irq_restore(flags);
> >   }
> > +
> > +struct steal_monitor_t steal_mon;
> > +
> > +void sched_init_steal_monitor(void)
> > +{
> > +    INIT_WORK(&steal_mon.work, sched_steal_detection_work);
> > +    zalloc_cpumask_var(&steal_mon.tmp_mask, GFP_KERNEL);
>
> Error handling is missing.
>
> [...]

I had seen llm complain the same. I think there very little chance for it since this is
during the boot. But yes.
I can add the relevant checks for tmp_mask and WARN_ON here. Its not like anything should
crash if one can;t enable this steal monitor thing.


PS:
we don't have similar error handling in other paths for zalloc_cpumask_var case too.
Many are in init time, so again very unlikely. One can add BUG_ON() for those,
but that's different story.