Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S935862Ab3DJGzl (ORCPT ); Wed, 10 Apr 2013 02:55:41 -0400 Received: from mail-ob0-f175.google.com ([209.85.214.175]:45964 "EHLO mail-ob0-f175.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752725Ab3DJGzj (ORCPT ); Wed, 10 Apr 2013 02:55:39 -0400 MIME-Version: 1.0 In-Reply-To: <51650185.9060905@linux.vnet.ibm.com> References: <1364873008-3169-1-git-send-email-alex.shi@intel.com> <1364873008-3169-7-git-send-email-alex.shi@intel.com> <51650185.9060905@linux.vnet.ibm.com> Date: Wed, 10 Apr 2013 08:55:37 +0200 Message-ID: Subject: Re: [patch v3 6/8] sched: consider runnable load average in move_tasks From: Vincent Guittot To: Michael Wang , Alex Shi Cc: "mingo@redhat.com" , Peter Zijlstra , Thomas Gleixner , Andrew Morton , Arjan van de Ven , Borislav Petkov , Paul Turner , Namhyung Kim , Mike Galbraith , Morten Rasmussen , gregkh@linuxfoundation.org, Preeti U Murthy , Viresh Kumar , linux-kernel , Len Brown , rafael.j.wysocki@intel.com, jkosina@suse.cz, Clark Williams , "tony.luck@intel.com" , keescook@chromium.org, Mel Gorman , riel@redhat.com Content-Type: text/plain; charset=ISO-8859-1 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3004 Lines: 86 On 10 April 2013 08:07, Michael Wang wrote: > On 04/09/2013 03:08 PM, Vincent Guittot wrote: >> On 2 April 2013 05:23, Alex Shi wrote: >>> Except using runnable load average in background, move_tasks is also >>> the key functions in load balance. We need consider the runnable load >>> average in it in order to the apple to apple load comparison. >>> >>> Signed-off-by: Alex Shi >>> --- >>> kernel/sched/fair.c | 11 ++++++++++- >>> 1 file changed, 10 insertions(+), 1 deletion(-) >>> >>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c >>> index 1f9026e..bf4e0d4 100644 >>> --- a/kernel/sched/fair.c >>> +++ b/kernel/sched/fair.c >>> @@ -3966,6 +3966,15 @@ static unsigned long task_h_load(struct task_struct *p); >>> >>> static const unsigned int sched_nr_migrate_break = 32; >>> >>> +static unsigned long task_h_load_avg(struct task_struct *p) >>> +{ >>> + u32 period = p->se.avg.runnable_avg_period; >>> + if (!period) >>> + return 0; >>> + >>> + return task_h_load(p) * p->se.avg.runnable_avg_sum / period; >> >> How do you ensure that runnable_avg_period and runnable_avg_sum are >> coherent ? an update of the statistic can occur in the middle of your >> sequence. > > Hi, Vincent > > Don't we have the 'rq->lock' to protect it? > > move_tasks() was invoked with double locked, for all the se on src and > dst rq, no update should happen, isn't it? you're right, the double lock protect against concurrent access my remark doesn't apply here Regards, Vincent > > Regards, > Michael Wang > >> >> Vincent >> >>> +} >>> + >>> /* >>> * move_tasks tries to move up to imbalance weighted load from busiest to >>> * this_rq, as part of a balancing operation within domain "sd". >>> @@ -4001,7 +4010,7 @@ static int move_tasks(struct lb_env *env) >>> if (throttled_lb_pair(task_group(p), env->src_cpu, env->dst_cpu)) >>> goto next; >>> >>> - load = task_h_load(p); >>> + load = task_h_load_avg(p); >>> >>> if (sched_feat(LB_MIN) && load < 16 && !env->sd->nr_balance_failed) >>> goto next; >>> -- >>> 1.7.12 >>> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> Please read the FAQ at http://www.tux.org/lkml/ >> > > -- > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/ -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/