Re: [RFC PATCH] sched/fair: Refactor can_migrate_task() to elimate looping

[I Hsin Cheng]

On Thu, Jan 09, 2025 at 02:28:27PM +0530, Nysal Jan K.A. wrote:
> On Thu, Jan 09, 2025 at 01:29:47AM +0800, I Hsin Cheng wrote:
> > @@ -9404,12 +9404,16 @@ int can_migrate_task(struct task_struct *p, struct lb_env *env)
> >  			return 0;
> >  
> >  		/* Prevent to re-select dst_cpu via env's CPUs: */
> > -		for_each_cpu_and(cpu, env->dst_grpmask, env->cpus) {
> > -			if (cpumask_test_cpu(cpu, p->cpus_ptr)) {
> > -				env->flags |= LBF_DST_PINNED;
> > -				env->new_dst_cpu = cpu;
> > -				break;
> > -			}
> > +		struct cpumask dst_mask;
> > +
> > +		cpumask_and(&dst_mask, env->dst_grpmask, env->cpus);
> > +		cpumask_and(&dst_mask, &dst_mask, p->cpus_ptr);
> > +
> > +		cpu = cpumask_first(&dst_mask);
> 
> Can cpumask_first_and_and() be used instead?
> 
> > +
> > +		if (cpu < nr_cpu_ids) {
> > +			env->flags |= LBF_DST_PINNED;
> > +			env->new_dst_cpu = cpu;
> >  		}
> >  
> >  		return 0;
> > -- 
> > 2.43.0
> > 
> 
> --Nysal

> Can cpumask_first_and_and() be used instead?

That's a great idea! thank you.
That way we don't have to put redundant cpumask on stack, I'll test it
real soon and send the test result up.

--Richard