Re: [RFCv6 PATCH 09/10] sched: deadline: use deadline bandwidth in scale_rt_capacity

From: Luca Abeni <luca.abeni@unitn.it>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Vincent Guittot <vincent.guittot@linaro.org>,
	Steve Muckle <steve.muckle@linaro.org>,
	Ingo Molnar <mingo@redhat.com>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	"linux-pm@vger.kernel.org" <linux-pm@vger.kernel.org>,
	Morten Rasmussen <morten.rasmussen@arm.com>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	Juri Lelli <Juri.Lelli@arm.com>,
	Patrick Bellasi <patrick.bellasi@arm.com>,
	Michael Turquette <mturquette@baylibre.com>
Subject: Re: [RFCv6 PATCH 09/10] sched: deadline: use deadline bandwidth in scale_rt_capacity
Date: Tue, 15 Dec 2015 14:18:11 +0100	[thread overview]
Message-ID: <56701313.8000800@unitn.it> (raw)
In-Reply-To: <20151215122015.GA6357@twins.programming.kicks-ass.net>

On 12/15/2015 01:20 PM, Peter Zijlstra wrote:
> On Tue, Dec 15, 2015 at 09:50:14AM +0100, Luca Abeni wrote:
>> On 12/15/2015 05:59 AM, Vincent Guittot wrote:
>
>>> The 2nd definition is used to compute the remaining capacity for the
>>> CFS scheduler. This one doesn't need to be updated at each wake/sleep
>>> of a deadline task but should reflect the capacity used by deadline in
>>> a larger time scale. The latter will be used by the CFS scheduler  at
>>> the periodic load balance pace
>
>> Ok, so as I wrote above this really looks like an average utilisation.
>> My impression (but I do not know the CFS code too much) is that the mainline
>> kernel is currently doing the right thing to compute it, so maybe there is no
>> need to change the current code in this regard.
>> If the current code is not acceptable for some reason, an alternative would
>> be to measure the active utilisation for frequency scaling, and then apply a
>> low-pass filter to it for CFS.
>
> So CFS really only needs a 'vague' average idea on how much time it will
> not get. Its best effort etc., so being a little wrong isn't a problem.
>
> The current code suffices, but I think the reason its been changed in
> this series is that they want/need separate tracking for fifo/rr and
> deadline in the next patch, and taking out deadline like proposed was
> the easiest way of achieving that.
Ah, ok. Thanks for explaining.
So, I agree that this patch is not a good idea for estimating the average
utilisation needed by CFS.


				Luca