Re: [RFC v5 9/9] sched/deadline: also reclaim bandwidth not used by dl tasks

[Luca Abeni]

Hi Peter,

On Mon, 27 Mar 2017 16:03:41 +0200
Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:

> On Fri, Mar 24, 2017 at 04:53:02AM +0100, luca abeni wrote:
> 
> > +static inline
> > +void __dl_update(struct dl_bw *dl_b, s64 bw)
> > +{
> > +	struct root_domain *rd = container_of(dl_b, struct
> > root_domain, dl_bw);
> > +	int i;
> > +
> > +	RCU_LOCKDEP_WARN(!rcu_read_lock_sched_held(),
> > +			 "sched RCU must be held");
> > +	for_each_cpu_and(i, rd->span, cpu_active_mask) {
> > +		struct rq *rq = cpu_rq(i);
> > +
> > +		rq->dl.extra_bw += bw;
> > +	}  
> 
> So this is unfortunate (and we already have one such instance).
> 
> It effectively does an for_each_online_cpu() with IRQs disabled, and
> on SGI class hardware that takes _forever_.

I have to admit I copied this code from somewhere else... :)

I am happy to discuss a better solution.


			Thanks,
				Luca

> 
> This is also what I got stuck on trying to rewrite AC to use Tommaso's
> recoverable thing. In the end I had to do a 2 stage try/commit
> variant. Which ended up being a pain and I didn't finish.
> 
> I'm not saying this patch is bad, but this is something we need to
> thing about.