Re: [PATCH v4 4/4] sched/fair: Kill wake_cap()

[Valentin Schneider]

On 07/02/2020 11:19, Quentin Perret wrote:
> On Thursday 06 Feb 2020 at 19:19:57 (+0000), Valentin Schneider wrote:
>> From: Morten Rasmussen <morten.rasmussen@xxxxxxx>
>>
>> Capacity-awareness in the wake-up path previously involved disabling
>> wake_affine in certain scenarios. We have just made select_idle_sibling()
>> capacity-aware, so this isn't needed anymore.
>>
>> Remove wake_cap() entirely.
>>
>> Signed-off-by: Morten Rasmussen <morten.rasmussen@xxxxxxx>
>> [Changelog tweaks]
>> Signed-off-by: Valentin Schneider <valentin.schneider@xxxxxxx>
> 
> Reviewed-by: Quentin Perret <qperret@xxxxxxxxxx>
> 
> I wanted to suggest removing the CA code from update_sg_wakeup_stats()
> which is now called only on fork/exec, but I suppose we still want it
> for util_clamp, so n/m.
> 

Good point, I hadn't thought about this. As you say we probably want to
keep it since we can fork/exec into a cgroup that has uclamp values already
set up, and that would drive task_fits_capacity().

> Thanks for the series,

Thanks for the review!

> Quentin
>