Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Tue, 18 Jul 2023 09:25:25 -0400
From:   Phil Auld <pauld@redhat.com>
To:     Tejun Heo <tj@kernel.org>
Cc:     linux-kernel@vger.kernel.org, Juri Lelli <juri.lelli@redhat.com>,
        Ingo Molnar <mingo@redhat.com>,
        Daniel Bristot de Oliveira <bristot@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Vincent Guittot <vincent.guittot@linaro.org>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Valentin Schneider <vschneid@redhat.com>,
        Ben Segall <bsegall@google.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Mel Gorman <mgorman@suse.de>,
        Frederic Weisbecker <frederic@kernel.org>
Subject: Re: [PATCH v3 1/2] sched, cgroup: Restore meaning to
 hierarchical_quota
Message-ID: <20230718132525.GB126587@lorien.usersys.redhat.com>
References: <xm268rbkg4tg.fsf@google.com>
 <20230714125746.812891-1-pauld@redhat.com>
 <ZLWIDC2nlG8cb3VE@slm.duckdns.org>
 <20230718125759.GA126587@lorien.usersys.redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230718125759.GA126587@lorien.usersys.redhat.com>
Precedence: bulk

Hi Tejun,

On Tue, Jul 18, 2023 at 08:57:59AM -0400 Phil Auld wrote:
> On Mon, Jul 17, 2023 at 08:27:24AM -1000 Tejun Heo wrote:
> > On Fri, Jul 14, 2023 at 08:57:46AM -0400, Phil Auld wrote:
> > > In cgroupv2 cfs_b->hierarchical_quota is set to -1 for all task
> > > groups due to the previous fix simply taking the min.  It should
> > > reflect a limit imposed at that level or by an ancestor. Even
> > > though cgroupv2 does not require child quota to be less than or
> > > equal to that of its ancestors the task group will still be
> > > constrained by such a quota so this should be shown here. Cgroupv1
> > > continues to set this correctly.
> > > 
> > > In both cases, add initialization when a new task group is created
> > > based on the current parent's value (or RUNTIME_INF in the case of
> > > root_task_group). Otherwise, the field is wrong until a quota is
> > > changed after creation and __cfs_schedulable() is called.
> > > 
> > > Fixes: c53593e5cb69 ("sched, cgroup: Don't reject lower cpu.max on ancestors")
> > 
> > Does this really fix anything observable? I wonder whether this is more
> > misleading than helpful. In cgroup2, the value simply wasn't being used,
> > right?
> >

(Sorry, my editor bit me...I had added:)

I don't feel strongly about the fixes. What was there seemed broken to me
so ... "Fixes". But it doesn't matter. 


> 
> It wasn't being used but was actively being set wrong. I mean if we are
> going to bother doing the __cfs_schedulable() tg tree walk we might as
> well have not been setting a bogus value. But that said, no it was not
> observable until I tried to use it.
>

We have a field called hierarchical_quota, that was being unconditionally
set to -1 for cgroup2. I figured it would be more correct to reflect
the hieratchical quota. :)


> I'm fine if that's dropped. I just wanted it set right going forward :)
> 
> 
> > > Signed-off-by: Phil Auld <pauld@redhat.com>
> > > Reviewed-by: Ben Segall <bsegall@google.com>
> > > Cc: Ingo Molnar <mingo@redhat.com>
> > > Cc: Peter Zijlstra <peterz@infradead.org>
> > > Cc: Vincent Guittot <vincent.guittot@linaro.org>
> > > Cc: Juri Lelli <juri.lelli@redhat.com>
> > > Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
> > > Cc: Valentin Schneider <vschneid@redhat.com>
> > > Cc: Ben Segall <bsegall@google.com>
> > > Cc: Frederic Weisbecker <frederic@kernel.org>
> > > Cc: Tejun Heo <tj@kernel.org>
> > 
> > Acked-by: Tejun Heo <tj@kernel.org>
> >
> 
> Thanks!
> 
> 
> > > +		 * always take the non-RUNTIME_INF min.  On cgroup1, only
> > > +		 * inherit when no limit is set. In both cases this is used
> > > +		 * by the scheduler to determine if a given CFS task has a
> > > +		 * bandwidth constraint at some higher level.
> > 
> > The discussion on this comment is stretching too long and this is fine too
> > but what's worth commenting for cgroup2 is that the limit value itself
> > doesn't mean anything and we're just latching onto the value used by cgroup1
> > to track whether there's any limit active or not.
> 
> I thought that was implied by the wording. "If a given task has a bandwidth
> contraint" not "what a given task's bandwidth constraint is".  In both cases
> that's how the other parts of the scheduler are using it. The actual
> non-RUNTIME_INF value only matters in this function (and only for cgroup1
> indeed).
> 
> But... the value is just as accurate for cgroup2 and cgroup1.  The task is
> still going to be limited by that bandwidth constraint even if its own
> bandwidth limit is nominally higher, right? 
> 
> 
> Cheers,
> Phil
> 
> > 
> > Thanks.
> > 
> > -- 
> > tejun
> > 
> 
> -- 
> 

--