Re: [PATCH v2 1/3] sched: Stop nohz stats when decayed

[Peter Zijlstra]

On Thu, Feb 08, 2018 at 05:52:00PM +0100, Vincent Guittot wrote:
> On 8 February 2018 at 16:44, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
> > On Thu, Feb 08, 2018 at 04:05:58PM +0100, Vincent Guittot wrote:
> >> On 8 February 2018 at 15:00, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
> >> > On Tue, Feb 06, 2018 at 08:23:05PM +0100, Vincent Guittot wrote:
> >> >
> >> >> @@ -9207,13 +9231,15 @@ void nohz_balance_enter_idle(int cpu)
> >> >>       if (!housekeeping_cpu(cpu, HK_FLAG_SCHED))
> >> >>               return;
> >> >>
> >> >> +     rq->has_blocked_load = 1;
> >
> > Should we not set that with rq->lock held? We already clear it while
> > holding rq->lock.
> 
> I think it's safe because it is used to re-enable the periodic decay
> unconditionally.
> It is cleared with  rq->lock held to prevents any update of the cfs_rq
> *_avg  while deciding if we can clear has_blocked_load

Yes, and that clearing must then have observed the new addition. But
please put so in a comment.