Re: [PATCH 1/2] cpuset: fix cpuset_cpus_allowed_fallback() don'tupdate tsk->rt.nr_cpus_allowed

[KOSAKI Motohiro]

(2011/05/13 16:43), Yong Zhang wrote:
> On Fri, May 13, 2011 at 3:33 PM, KOSAKI Motohiro
> <kosaki.motohiro@xxxxxxxxxxxxxx>  wrote:
> > Hello,
> >
> > > > > > (3) schedule() makes fallback and cpuset_cpus_allowed_fallback
> > > > > >      change task->cpus_allowed
> > > > >
> > > > > I'm failing to see how this is happening, surely that kthread isn't
> > > > > actually running that early?
> > > >
> > > > If my understand correctly, current call graph is below.
> > > >
> > > > kernel_init()
> > > >         smp_init();
> > > >                 cpu_up()
> > > >                         ... cpu hotplug notification
> > > >                                 kthread_create()
> > > >         sched_init_smp();
> > > >
> > > >
> > > > So, cpu hotplug event is happen before sched_init_smp(). The old rule is,
> > > > all kthreads of using cpu-up notification have to use kthread_bind(). It
> > > > protected from sched load balance.
> > > >
> > > > but, now cpuset_cpus_allowed_fallback() forcedly change kthread's
> > > > cpumask.
> > > > Why is this works? the point are two.
> > > >
> > > > - rcu_cpu_kthread_should_stop() call set_cpus_allowed_ptr() again
> > > > periodically.
> > > >   then, it can reset cpumask if cpuset_cpus_allowed_fallback() change it.
> > > >   my debug print obseve following cpumask change occur at boot time.
> > > >      1) kthread_bind: bind cpu1
> > > >      2) cpuset_cpus_allowed_fallback: bind possible cpu
> > > >      3) rcu_cpu_kthread_should_stop: rebind cpu1
> > > > - while tsk->rt.nr_cpus_allowed == 1, sched load balancer never be crash.
> > >
> > > Seems rcu_spawn_one_cpu_kthread() call wake_up_process() directly,
> > > which is under hotplug event CPU_UP_PREPARE. Maybe it should be
> > > under CPU_ONLINE.
> >
> > Hmm..
> >
> > I haven't catch your point. cpu_up() call both CPU_UP_PREPARE and CPU_ONLINE
> > notification. Thus, CPU_ONLINE still be called before sched_init_smp().
>
> The difference is that CPU_ONLINE will set cpu_online_mask and cpu_active_mask,
> so wake_up_process()->ttwu()->select_task_rq() will not go back to
> cpuset_cpus_allowed_fallback().

Right. Thank you. Now I got your point!

> But the comments for rcu_cpu_kthread_should_stop() shot me, since
> rcu_spawn_one_cpu_kthread() must be called under CPU_UP_PREPARE
> for some reason.
>
> Maybe we should find a more suitable way for your issue :)

ok, I'll try to read and understand rcu-kthread design and find it out.



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/