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, 1 Aug 2023 17:37:31 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Phil Auld <pauld@redhat.com>
Cc:     linux-kernel@vger.kernel.org, Juri Lelli <juri.lelli@redhat.com>,
        Ingo Molnar <mingo@redhat.com>,
        Daniel Bristot de Oliveira <bristot@redhat.com>,
        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>,
        Tejun Heo <tj@kernel.org>
Subject: Re: [PATCH v6 2/2] Sched/fair: Block nohz tick_stop when cfs
 bandwidth in use
Message-ID: <20230801153731.GD11704@hirez.programming.kicks-ass.net>
References: <20230712133357.381137-1-pauld@redhat.com>
 <20230712133357.381137-3-pauld@redhat.com>
 <20230731224934.GD51835@hirez.programming.kicks-ass.net>
 <20230801111342.GA268019@lorien.usersys.redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230801111342.GA268019@lorien.usersys.redhat.com>
Precedence: bulk

On Tue, Aug 01, 2023 at 07:13:42AM -0400, Phil Auld wrote:
> On Tue, Aug 01, 2023 at 12:49:34AM +0200 Peter Zijlstra wrote:
> > On Wed, Jul 12, 2023 at 09:33:57AM -0400, Phil Auld wrote:
> > > CFS bandwidth limits and NOHZ full don't play well together.  Tasks
> > > can easily run well past their quotas before a remote tick does
> > > accounting.  This leads to long, multi-period stalls before such
> > > tasks can run again. Currently, when presented with these conflicting
> > > requirements the scheduler is favoring nohz_full and letting the tick
> > > be stopped. However, nohz tick stopping is already best-effort, there
> > > are a number of conditions that can prevent it, whereas cfs runtime
> > > bandwidth is expected to be enforced.
> > > 
> > > Make the scheduler favor bandwidth over stopping the tick by setting
> > > TICK_DEP_BIT_SCHED when the only running task is a cfs task with
> > > runtime limit enabled. We use cfs_b->hierarchical_quota to
> > > determine if the task requires the tick.
> > > 
> > > Add check in pick_next_task_fair() as well since that is where
> > > we have a handle on the task that is actually going to be running.
> > > 
> > > Add check in sched_can_stop_tick() to cover some edge cases such
> > > as nr_running going from 2->1 and the 1 remains the running task.
> > 
> > These appear fine to me, except:
> >
> > > Add sched_feat HZ_BW (off by default) to control the tick_stop
> > > behavior.
> > 
> > What was the thinking here? This means nobody will be using this -- why
> > would you want this default disabled?
> > 
> 
> That was just a hedge in case it caused issues. I'd probably have had to
> enable it in RHEL anyway. Using a feature was to make it inocuous when
> disabled.  Would you prefer me to enable it or remove the sched_feat
> entirely? (or do you want to just switch that to true when you apply it?)

I've edited it to default enabled -- we can pull the feature flag
eventually I suppose.

Things didn't readily apply, so I've kicked at it a little. Should be in
queue/sched/core for the robots to chew on.