Re: [PATCH] sched/fair: schedutil: update only with all info available

[Joel Fernandes]

On Thu, Apr 12, 2018 at 12:01 AM, Vincent Guittot
<vincent.guittot@xxxxxxxxxx> wrote:

>>
>> Also that aside, the "running util" is what was used to drive the CFS
>> util before Peter's patch (8f111bc357a). That was accounting the
>> blocked and decaying utilization but that patch changed the behavior.
>> It seems logical we should just use that not check for h_nr_running
>> for CFS so we don't miss on the decayed  utilization. What is the use
>> of checking h_nr_running or state of runqueue for CFS? I am sure to be
>> missing something here. :-(
>
> As Peter mentioned, the change in commit (8f111bc357a) was to remove
> the test that was arbitrary removing the util_avg of a cpu that has
> not been updated since a tick
>
> But with the update of blocked idle load, we don't need to handle the
> case of stalled load/utilization

Thanks a lot for the clarification. It makes sense now.

- Joel