Re: [RFC] Thread Migration Preemption - v4

[Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>]

* Peter Zijlstra (a.p.zijlstra@chello.nl) wrote:
> On Sat, 2007-07-14 at 14:42 -0400, Mathieu Desnoyers wrote:
> 
> > Note: (or we could say FIXME)
> > Is we ever want to check migration pending in assembly code, we will have to
> > make sure we test the right thread flag bits on each architectures. Care should
> > also be taken to check that the thread flags used won't trigger false positives
> > in non selective asm thread flag checks.
> > 
> > FIXME (HOTPLUG) :
> > 
> > > >     /* Affinity changed (again). */
> > > >     if (!cpu_isset(dest_cpu, p->cpus_allowed))
> > > >             goto out;
> > > >  
> > > >     on_rq = p->se.on_rq;
> > > > +#ifdef CONFIG_PREEMPT
> > > > +   if (!on_rq && task_thread_info(p)->migrate_count)
> > > > +           goto out;
> > > > +#endif
> > > 
> > > This means that move_task_off_dead_cpu() will spin until the task will be
> > > scheduled
> > > on the dead CPU. Given that we hold tasklist_lock and irqs are disabled, this
> > > may
> > > never happen.
> > >
> > 
> > Yes. My idea to fix this issue is the following:
> > 
> > If a thread has non zero migrate_count, we should still move it to a
> > different CPU upon hotplug cpu removal, even if this thread resists
> > migration. Care should be taken to send _all_ such threads to the _same_
> > CPU so they don't race for the per-cpu ressources. Does it make sense ?
> > 
> > We would have to keep the CPU affinity of the threads running on the
> > wrong CPU until they end their migrate disabled section, so that we can
> > put them back on their original CPU if it goes back online, otherwise we
> > could end up with concurrent per-cpu variables accesses.
> > 
> > (I'll wait for reply before coding a solution for this CPU HOTPLUG
> > related problem)
> 
> What would, aside from technical issues, be the problem with making
> migration_disable() delay CPU_DOWN until migration_enable()?
> 

Because if we thing a little further, migration disabling could be a
very interesting way to provide cheap per-cpu data structure access to
user-space. But we would not want user-space processes to hold CPU_DOWN
forever...

-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/