Re: [PATCH 0/2 v4] sched: Rewrite per entity runnable load average tracking

[Peter Zijlstra]

On Tue, Jul 29, 2014 at 09:17:13AM +0800, Yuyang Du wrote:
> On Mon, Jul 28, 2014 at 12:38:12PM +0200, Peter Zijlstra wrote:
> > On Mon, Jul 28, 2014 at 03:02:37AM +0800, Yuyang Du wrote:
> > > > Another thing that might be an issue is that the blocked of a terminated
> > > > task lives on for quite a while until has decayed away.
> > > 
> > > Good point. To do so, if I read correctly, we need to hook do_exit(), but probably
> > > we are gonna encounter rq->lock issue.
> > > 
> > > What is the opinion/guidance from the maintainers/others?
> > 
> > So the entire point of this per entity tracking was to make sure load
> > numbers reflect reality. We account migrations etc., it would be weird
> > to then throw all that out the window and let task exit accumulate crap.
> > 
> 
> Yes. So I will hook up do_exit. 

There is sched_class::task_dead(), pjt wanted to use that for this.

> Hope I did it right. Likewise, also do group entity
> in group destroy and group offline? 

group destroy, yes. Offline should mostly fix itself already due to it
actively migrating the actual load around I think.
--
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/