Re: [PATCH] sched/fair: fix mul overflow on 32-bit systems

From: Steve Muckle <steve.muckle@linaro.org>
To: Yuyang Du <yuyang.du@intel.com>, Peter Zijlstra <peterz@infradead.org>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>,
	Andrey Ryabinin <aryabinin@virtuozzo.com>,
	mingo@redhat.com, linux-kernel@vger.kernel.org,
	Paul Turner <pjt@google.com>, Ben Segall <bsegall@google.com>
Subject: Re: [PATCH] sched/fair: fix mul overflow on 32-bit systems
Date: Tue, 15 Dec 2015 13:56:01 -0800	[thread overview]
Message-ID: <56708C71.90701@linaro.org> (raw)
In-Reply-To: <20151215022229.GF28098@intel.com>

On 12/14/2015 06:22 PM, Yuyang Du wrote:
> what if we just init the util_avg to 0?

With scheduler-guided frequency [1] we will rely on the initial util_avg
value to determine the CPU frequency response when new tasks are created.

There is sure to be a lot of debate on what sort of default policy is
used (and how much that can be changed with tunables). IMO at least for
now, new tasks should cause the target CPU to go to fmax, especially
given how slow PELT is to respond to changes in task behavior. This
would imply leaving the initial task util_avg at 100% (or the equivalent
necessary to take care of the overflow issues).

[1] https://lkml.org/lkml/2015/12/9/35

thanks,
Steve