[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <57626ADD.8080203@arm.com>
Date: Thu, 16 Jun 2016 10:01:17 +0100
From: Dietmar Eggemann <dietmar.eggemann@....com>
To: Mike Galbraith <umgwanakikbuti@...il.com>,
Peter Zijlstra <peterz@...radead.org>
Cc: Yuyang Du <yuyang.du@...el.com>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [rfc patch] sched/fair: Use instantaneous load for fork/exec
balancing
On 16/06/16 04:33, Mike Galbraith wrote:
> On Wed, 2016-06-15 at 20:03 +0100, Dietmar Eggemann wrote:
>
>> Isn't there a theoretical problem with the scale_load() on CONFIG_64BIT
>> machines on tip/sched/core? load.weight has a higher resolution than
>> runnable_load_avg (and so the values in the rq->cpu_load[] array).
>> Theoretically because [forkexec|wake]_idx is 0 so [target|source]_load()
>> is nothing else than weighted_cpuload().
>
> I see a not so theoretical problem with my rfc in that I forgot to
> scale_load_down() if that's what you mean.
Yup. Theoretical in the sense that this_load and min_load will be
affected both the same way as long as load_idx = 0.
>
> (changes nothing, reality was just extra special unadulterated;)
Agreed.
>
> -Mike
>
Powered by blists - more mailing lists