On 21 March 2017 at 15:58, Peter Zijlstra <pet...@infradead.org> wrote: > > On Tue, Mar 21, 2017 at 03:16:19PM +0100, Vincent Guittot wrote: > > On 21 March 2017 at 15:03, Peter Zijlstra <pet...@infradead.org> wrote: > > > > > On Tue, Mar 21, 2017 at 02:37:08PM +0100, Vincent Guittot wrote: > > > > On 21 March 2017 at 14:22, Peter Zijlstra <pet...@infradead.org> wrote: > > > > > > > For the not overloaded case, it makes sense to immediately update to > > > > OPP to be aligned with the new utilization of the CPU even if it was > > > > not idle in the past couple of ticks > > > > > > Yeah, but we cannot know. Also, who cares? > > > > > > > embedded system that doesn't want to stay at higest OPP if significant part > > of the utilzation has moved away as an example > > AFAICT, schedutil tries to select the best OPP according to the current > > utilization of the CPU so if the utilization decreases, the OPP should also > > decrease > > Sure I get that; but given the lack of crystal ball instructions we > cannot know if this is the case or not.
cfs_rq->avg.load_avg account the waiting time of CPU (in addition to the weight of task) so i was wondering if we can't use it to detect if we are in the overloaded case or not even if utilization is not mac capacity because we have just migrated a task (and its utilization) out > > And if we really dropped below 100% utilization, we should hit idle > fairly soon.
