Re: [PATCH v4 6/8] sched/fair: Add sched group latency support

[Tejun Heo <tj@kernel.org>]

Hello,

On Mon, Sep 19, 2022 at 01:55:15PM +0200, Dietmar Eggemann wrote:
> s/valentin.schneider@arm.com//
> 
> On 16/09/2022 10:03, Vincent Guittot wrote:
> > Task can set its latency priority, which is then used to decide to preempt
> > the current running entity of the cfs, but sched group entities still have
> > the default latency offset.
> > 
> > Add a latency field in task group to set the latency offset of the
> > sched_eneities of the group, which will be used against other entities in
> 
> s/sched_eneities/sched_entity
> 
> > the parent cfs when deciding which entity to schedule first.
> 
> So latency for cgroups does not follow any (existing) Resource
> Distribution Model/Scheme (Documentation/admin-guide/cgroup-v2.rst)?
> Latency values are only used to compare sched entities at the same level.

I think it'd still result in a hierarchical behavior as scheduling is done
recursively top-down. Right, vincent?

It doesn't fit any of the resource distribution model. But it's rather
difficult to map latencies to existing concepts of resources and we have a
precedence in the cpu controller (.idle) which behaves similarly, so as long
as the behavior is hierarchical, I think it's okay.

Thanks.

-- 
tejun