From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756561Ab3A1LTr (ORCPT ); Mon, 28 Jan 2013 06:19:47 -0500 Received: from moutng.kundenserver.de ([212.227.126.171]:62874 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754205Ab3A1LTp (ORCPT ); Mon, 28 Jan 2013 06:19:45 -0500 Message-ID: <1359371965.5783.127.camel@marge.simpson.net> Subject: Re: [patch] sched: minimalist select_idle_sibling() bouncing cow syndrome fix From: Mike Galbraith To: Ingo Molnar Cc: LKML , Ingo Molnar , Peter Zijlstra Date: Mon, 28 Jan 2013 12:19:25 +0100 In-Reply-To: <20130128105304.GC20263@gmail.com> References: <1359273021.5803.80.camel@marge.simpson.net> <20130128105304.GC20263@gmail.com> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.2.3 Content-Transfer-Encoding: 7bit Mime-Version: 1.0 X-Provags-ID: V02:K0:tOU8rTuDblKA0YTb2fxRTBxA+5Ou6eJrH18w7OIh2tZ PPiD4aY21CXaeopy18Etwc7b2nZhpRIVIVJh1Y5mR25DKbPQH7 T/WUhjv0WbxI4gBVBFpS1III25sMK32+CFUJC/bGxPQZ5c4TAd L1DWTc3Rj2G1XZqsmRpFucElqxxnUFnicPUc8kI98c7DzxQ1Nk rleN6UBI9rS6rJ+roelHcoTjIB3+4j5BFcnp15C5tyNNbjYZdd zezECZ3Iv1puy7x2bjDsSF2EZmnwdBwMhWjmSJGItoVBWB7Zhg b2UGmbjs9VG3x7AYFksXewdWZ0FDiRIkTyrGWxpEjDzLqGw1HE 4g2DgKHffiFJY1CXHMeIRwtcFDlznUAIWE7qvloxkVbuO6/LP4 3RzY1Z2cPLLWg== Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 2013-01-28 at 11:53 +0100, Ingo Molnar wrote: > * Mike Galbraith wrote: > > > If the previous CPU is cache affine and idle, select it. > > No objections in principle - but would be nice to have a > changelog with numbers, % of improvement included and so? Well, that like was my changelog, guess it needs improvement. Take 2. sched: minimalist select_idle_sibling() bouncing cow syndrome fix If the previous CPU is cache affine and idle, select it. The current implementation simply traverses the sd_llc domain, taking the first idle CPU encountered, which walks buddy pairs hand in hand over the package, inflicting excruciating pain. 1 tbench pair (worst case) in a 10 core + SMT package: pre 15.22 MB/sec 1 procs post 252.01 MB/sec 1 procs Signed-off-by: Mike Galbraith --- kernel/sched/fair.c | 21 +++++++-------------- 1 file changed, 7 insertions(+), 14 deletions(-) --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -3270,25 +3270,18 @@ find_idlest_cpu(struct sched_group *grou */ static int select_idle_sibling(struct task_struct *p, int target) { - int cpu = smp_processor_id(); - int prev_cpu = task_cpu(p); struct sched_domain *sd; struct sched_group *sg; - int i; + int i = task_cpu(p); - /* - * If the task is going to be woken-up on this cpu and if it is - * already idle, then it is the right target. - */ - if (target == cpu && idle_cpu(cpu)) - return cpu; + if (idle_cpu(target)) + return target; /* - * If the task is going to be woken-up on the cpu where it previously - * ran and if it is currently idle, then it the right target. + * If the prevous cpu is cache affine and idle, don't be stupid. */ - if (target == prev_cpu && idle_cpu(prev_cpu)) - return prev_cpu; + if (i != target && cpus_share_cache(i, target) && idle_cpu(i)) + return i; /* * Otherwise, iterate the domains and find an elegible idle cpu. @@ -3302,7 +3295,7 @@ static int select_idle_sibling(struct ta goto next; for_each_cpu(i, sched_group_cpus(sg)) { - if (!idle_cpu(i)) + if (i == target || !idle_cpu(i)) goto next; }