get_nohz_timer_target?

get_nohz_timer_target?

[Richard Cochran]

Looking at kernel/sched/core.c:get_nohz_timer_target(), I don't
understand the change made in:

    commit 9642d18eee2cd169b60c6ac0f20bda745b5a3d1e
    Author: Vatika Harlalka <vatikaharlalka@gmail.com>
    Date:   Tue Sep 1 16:50:59 2015 +0200
    nohz: Affine unpinned timers to housekeepers

After that change, the code now reads like this:

	int i, cpu = smp_processor_id();
	struct sched_domain *sd;

	if (!idle_cpu(cpu) && is_housekeeping_cpu(cpu))
		return cpu;

	rcu_read_lock();
	for_each_domain(cpu, sd) {
		for_each_cpu(i, sched_domain_span(sd)) {
			if (!idle_cpu(i) && is_housekeeping_cpu(cpu)) {
--------------------------------------------------------------- ^^^
Was this supposed to be 'i' instead?
If not, how does this test make any sense?
In any case, testing over and over again is surely wasteful.
---------------------------------------------------------------
				cpu = i;
				goto unlock;
			}
		}
	}

	if (!is_housekeeping_cpu(cpu))
		cpu = housekeeping_any_cpu();
unlock:
	rcu_read_unlock();
	return cpu;

Thanks,
Richard

Re: get_nohz_timer_target?

[Thomas Gleixner]

On Mon, 18 Apr 2016, Richard Cochran wrote:
> Looking at kernel/sched/core.c:get_nohz_timer_target(), I don't
> understand the change made in:
> 
>     commit 9642d18eee2cd169b60c6ac0f20bda745b5a3d1e
>     Author: Vatika Harlalka <vatikaharlalka@gmail.com>
>     Date:   Tue Sep 1 16:50:59 2015 +0200
>     nohz: Affine unpinned timers to housekeepers
> 
> After that change, the code now reads like this:
> 
> 	int i, cpu = smp_processor_id();
> 	struct sched_domain *sd;
> 
> 	if (!idle_cpu(cpu) && is_housekeeping_cpu(cpu))
> 		return cpu;
> 
> 	rcu_read_lock();
> 	for_each_domain(cpu, sd) {
> 		for_each_cpu(i, sched_domain_span(sd)) {
> 			if (!idle_cpu(i) && is_housekeeping_cpu(cpu)) {
> --------------------------------------------------------------- ^^^
> Was this supposed to be 'i' instead?

Yes. Care to send a patch?

Thanks,

	tglx

Re: get_nohz_timer_target?

[Frederic Weisbecker]

On Fri, Jul 08, 2016 at 11:05:58AM +0200, Thomas Gleixner wrote:
> On Mon, 18 Apr 2016, Richard Cochran wrote:
> > Looking at kernel/sched/core.c:get_nohz_timer_target(), I don't
> > understand the change made in:
> > 
> >     commit 9642d18eee2cd169b60c6ac0f20bda745b5a3d1e
> >     Author: Vatika Harlalka <vatikaharlalka@gmail.com>
> >     Date:   Tue Sep 1 16:50:59 2015 +0200
> >     nohz: Affine unpinned timers to housekeepers
> > 
> > After that change, the code now reads like this:
> > 
> > 	int i, cpu = smp_processor_id();
> > 	struct sched_domain *sd;
> > 
> > 	if (!idle_cpu(cpu) && is_housekeeping_cpu(cpu))
> > 		return cpu;
> > 
> > 	rcu_read_lock();
> > 	for_each_domain(cpu, sd) {
> > 		for_each_cpu(i, sched_domain_span(sd)) {
> > 			if (!idle_cpu(i) && is_housekeeping_cpu(cpu)) {
> > --------------------------------------------------------------- ^^^
> > Was this supposed to be 'i' instead?
> 
> Yes. Care to send a patch?

Ah this got fixed already: 444969223c81c7d0a95136b7b4cfdcfbc96ac5bd
("sched/nohz: Fix affine unpinned timers mess")

Thanks.

> 
> Thanks,
> 
> 	tglx

Re: get_nohz_timer_target?

[Thomas Gleixner]

On Fri, 8 Jul 2016, Frederic Weisbecker wrote:
> On Fri, Jul 08, 2016 at 11:05:58AM +0200, Thomas Gleixner wrote:
> > On Mon, 18 Apr 2016, Richard Cochran wrote:
> > > Looking at kernel/sched/core.c:get_nohz_timer_target(), I don't
> > > understand the change made in:
> > > 
> > >     commit 9642d18eee2cd169b60c6ac0f20bda745b5a3d1e
> > >     Author: Vatika Harlalka <vatikaharlalka@gmail.com>
> > >     Date:   Tue Sep 1 16:50:59 2015 +0200
> > >     nohz: Affine unpinned timers to housekeepers
> > > 
> > > After that change, the code now reads like this:
> > > 
> > > 	int i, cpu = smp_processor_id();
> > > 	struct sched_domain *sd;
> > > 
> > > 	if (!idle_cpu(cpu) && is_housekeeping_cpu(cpu))
> > > 		return cpu;
> > > 
> > > 	rcu_read_lock();
> > > 	for_each_domain(cpu, sd) {
> > > 		for_each_cpu(i, sched_domain_span(sd)) {
> > > 			if (!idle_cpu(i) && is_housekeeping_cpu(cpu)) {
> > > --------------------------------------------------------------- ^^^
> > > Was this supposed to be 'i' instead?
> > 
> > Yes. Care to send a patch?
> 
> Ah this got fixed already: 444969223c81c7d0a95136b7b4cfdcfbc96ac5bd
> ("sched/nohz: Fix affine unpinned timers mess")

Good. Though that patch misses a stable tag.

Thanks,

	tglx