Re: [PATCH 3/5] sched/fair: Make select_idle_cpu() proportional to cores

[Mel Gorman]

On Wed, Jan 13, 2021 at 05:49:58PM +0100, Vincent Guittot wrote:
> > @@ -7444,11 +7444,20 @@ int sched_cpu_activate(unsigned int cpu)
> >         balance_push_set(cpu, false);
> >
> >  #ifdef CONFIG_SCHED_SMT
> > -       /*
> > -        * When going up, increment the number of cores with SMT present.
> > -        */
> > -       if (cpumask_weight(cpu_smt_mask(cpu)) == 2)
> > -               static_branch_inc_cpuslocked(&sched_smt_present);
> > +       do {
> > +               int weight = cpumask_weight(cpu_smt_mask(cpu));
> > +               extern int sched_smt_weight;
> 
> coding style problem
> 

Presumably you are referring to an extern defined in a C file. That can
move to kernel/sched/sched.h in this patch.

> > <SNIP>
> >  /*
> >   * Scan the LLC domain for idle CPUs; this is dynamically regulated by
> >   * comparing the average scan cost (tracked in sd->avg_scan_cost) against the
> > @@ -6166,10 +6172,12 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
> >                 avg_cost = this_sd->avg_scan_cost + 1;
> >
> >                 span_avg = sd->span_weight * avg_idle;
> > -               if (span_avg > 4*avg_cost)
> > +               if (span_avg > sis_min_cores*avg_cost)
> >                         nr = div_u64(span_avg, avg_cost);
> >                 else
> > -                       nr = 4;
> > +                       nr = sis_min_cores;
> > +
> > +               nr *= sched_smt_weight;
> 
> Also,  patch 5 will look at all CPUs of a core in select_idle_core so
> nr will decrement by 1 per core so i don't see the need to multiply by
> sched_smt_weight one patch 5 is applied
> 

It makes sense in the context of this patch but can be removed again in
the last patch and then I think sched_smt_weight only exists in core.c

-- 
Mel Gorman
SUSE Labs