Re: [PATCH 2/2] sched/fair: take into account scheduling domain in select_idle_core()

[Peter Zijlstra <peterz@infradead.org>]

On Thu, Dec 14, 2023 at 06:55:51PM +0100, Keisuke Nishimura wrote:
> When picking out a CPU on a task wakeup, select_idle_smt() has to take
> into account the scheduling domain where the function looks for the CPU.
> This is because cpusets and isolcpus can remove CPUs from the domain
> to isolate them from other SMT siblings.

Same question as before, when cpusets, the cpu should also be unset from
p->cpus_ptr. So I'm thinking you're one of those isolcpus users I wish
that would go away ;-)

> This change replaces the set of CPUs allowed to run the task from
> p->cpus_ptr by the intersection of p->cpus_ptr and sched_domain_span(sd)
> which is stored in the cpus argument provided by select_idle_cpu.
> 
> Fixes: 9fe1f127b913 ("sched/fair: Merge select_idle_core/cpu()")
> Signed-off-by: Keisuke Nishimura <keisuke.nishimura@inria.fr>
> Signed-off-by: Julia Lawall <julia.lawall@inria.fr>
> ---
>  kernel/sched/fair.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 71306b48cf68..3b7d32632674 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -7262,7 +7262,7 @@ static int select_idle_core(struct task_struct *p, int core, struct cpumask *cpu
>  		if (!available_idle_cpu(cpu)) {
>  			idle = false;
>  			if (*idle_cpu == -1) {
> -				if (sched_idle_cpu(cpu) && cpumask_test_cpu(cpu, p->cpus_ptr)) {
> +				if (sched_idle_cpu(cpu) && cpumask_test_cpu(cpu, cpus)) {
>  					*idle_cpu = cpu;
>  					break;
>  				}
> @@ -7270,7 +7270,7 @@ static int select_idle_core(struct task_struct *p, int core, struct cpumask *cpu
>  			}
>  			break;
>  		}
> -		if (*idle_cpu == -1 && cpumask_test_cpu(cpu, p->cpus_ptr))
> +		if (*idle_cpu == -1 && cpumask_test_cpu(cpu, cpus))
>  			*idle_cpu = cpu;
>  	}

Aside of that, the actual patch seems to be fine, just the rationale
needs work.