Re: [PATCH RFC tip/core/rcu 8/9] rcu: Make RCU-tasks track exiting tasks

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Wed, Jul 30, 2014 at 07:04:42PM +0200, Oleg Nesterov wrote:
> On 07/28, Paul E. McKenney wrote:
> >
> > This commit adds synchronization with exiting tasks, so that RCU-tasks
> > avoids waiting on tasks that no longer exist.
> 
> I don't understand this patch yet, but it seems that it adds more than
> just synchronization with exiting tasks?

There was also a bit of code reorganization to keep indentation level
down to a dull roar.

> > +		ACCESS_ONCE(t->rcu_tasks_holdout) = 1;
> > +		spin_unlock(&t->rcu_tasks_lock);
> > +		smp_mb();  /* Order ->rcu_tasks_holdout store before "if". */
> > +		if (t == current || !ACCESS_ONCE(t->on_rq) || is_idle_task(t)) {
> > +			smp_store_release(&t->rcu_tasks_holdout, 0);
> > +			goto next_thread;
> > +		}
> 
> This should avoid the race with schedule()->rcu_note_voluntary_context_switch(),
> right?
> 
> > -		rcu_read_lock();
> > -		do_each_thread(g, t) {
> > -			if (t != current && ACCESS_ONCE(t->on_rq) &&
> > -			    !is_idle_task(t)) {
> > -				t->rcu_tasks_holdout = 1;
> 
> Because before this patch the code looks obviously racy, a task can do
> sleep(FOREVER) and block rcu_tasks_kthread() if it reads ->on_rq == 1
> after rcu_note_voluntary_context_switch() was already called.
> 
> However, I am not sure this race is actually closed even after this
> change... why rcu_note_voluntary_context_switch() can not miss
> ->rcu_tasks_holdout != 0 ?

Good point, I need to add a !ACCESS_ONCE(t->on_rq) when scanning the list
of tasks blocking the grace period.  I also need to handle NO_HZ_FULL,
but that comes later.

							Thanx, Paul

> OK, it seems that you are going to send the next version anyway, so
> please ignore.
> 
> Oleg.
>