Re: [PATCH 2/4] sched: cpufreq: Keep track of cpufreq utilization update flags

From: Peter Zijlstra <peterz@infradead.org>
To: Viresh Kumar <viresh.kumar@linaro.org>
Cc: Rafael Wysocki <rjw@rjwysocki.net>,
	Ingo Molnar <mingo@redhat.com>,
	linux-pm@vger.kernel.org,
	Vincent Guittot <vincent.guittot@linaro.org>,
	dietmar.eggemann@arm.com, morten.rasmussen@arm.com,
	juri.lelli@redhat.com, tkjos@android.com, joelaf@google.com,
	linux-kernel@vger.kernel.org, patrick.bellasi@arm.com
Subject: Re: [PATCH 2/4] sched: cpufreq: Keep track of cpufreq utilization update flags
Date: Wed, 20 Dec 2017 10:17:12 +0100	[thread overview]
Message-ID: <20171220091712.kd2nx5rf3dgila45@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20171220084859.GY19815@vireshk-i7>

On Wed, Dec 20, 2017 at 02:18:59PM +0530, Viresh Kumar wrote:
> On 20-12-17, 09:31, Peter Zijlstra wrote:
> What about this case: A CFS task is running currently and an RT task
> is enqueued.
> 
> - Is it always the case that the CFS task is preempted immediately and
>   the CPU is given to RT task ? I was talking to Vincent earlier and
>   he told me that for certain configurations the CFS task may keep
>   running until the next tick.
> 
> - What if the CFS task has disabled preemption ?
> 
> - More corner cases like this ?

If there is a runnable RT task we'll schedule to it ASAP. This means
for:

 CONFIG_PREEMPT=y

   either immediately or when next preempt_enable() hits a 0
   preempt_count.

 CONFIG_PREEMPT=n

   on any next kernel preemption point or when returning to userspace.

But this is not something we should worry about.