Re: [PATCH 1/4] sched/fair: Be less aggressive in calling cpufreq_update_util()

[Qais Yousef]

On 12/12/23 11:46, Dietmar Eggemann wrote:
> On 08/12/2023 02:52, Qais Yousef wrote:
> > Due to the way code is structured, it makes a lot of sense to trigger
> > cpufreq_update_util() from update_load_avg(). But this is too aggressive
> > as in most cases we are iterating through entities in a loop to
> > update_load_avg() in the hierarchy. So we end up sending too many
> > request in an loop as we're updating the hierarchy.
> 
> But update_load_avg() calls cfs_rq_util_change() which only issues a
> cpufreq_update_util() call for the root cfs_rq?

Yes I've noticed that and wondered. Maybe my analysis was flawed and I was just
hitting the issue of iowait boost request conflicting with update_load_avg()
request.

Let me have another look. I think we'll still end up needing to take the update
out of util_avg to be able to combine the two calls.


Cheers

--
Qais Yousef

> 
> So the 'iterating through entities' should be for a task in a non-root
> taskgroup which the condition (1) takes care of.
> 
> cfs_rq_util_change()
> 
>     ...
>     if (&rq->cfs == cfs_rq) (1)
> 
>         cpufreq_update_util()
> 
> [...]