From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755283Ab1DOCzC (ORCPT ); Thu, 14 Apr 2011 22:55:02 -0400 Received: from mga11.intel.com ([192.55.52.93]:30429 "EHLO mga11.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755033Ab1DOCzB (ORCPT ); Thu, 14 Apr 2011 22:55:01 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.64,215,1301900400"; d="scan'208";a="679320128" From: Alex Shi To: a.p.zijlstra@chello.nl, linux-kernel@vger.kernel.org Cc: tim.c.chen@intel.com, suresh.b.siddha@intel.com Subject: [PATCH] sched:make ads.avg_load update in time Date: Fri, 15 Apr 2011 10:11:37 +0800 Message-Id: <1302833497-19506-1-git-send-email-alex.shi@intel.com> X-Mailer: git-send-email 1.7.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org commit 866ab43efd325fae cause hackbench benchmark process mode performance dropping about 15% on our x86_64 machines. The patch works as its origin purpose, but it cause nearly double context switch on hackbench running. The sds.avg_load was not updated in time cause this. So when move the sds.avg_load update before group_imb checking, performance recovered totally. Signed-off-by: Alex Shi --- kernel/sched_fair.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 7f00772..036b660 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -3127,6 +3127,8 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, if (!sds.busiest || sds.busiest_nr_running == 0) goto out_balanced; + sds->avg_load = (SCHED_LOAD_SCALE * sds->total_load) / sds->total_pwr; + /* * If the busiest group is imbalanced the below checks don't * work because they assumes all things are equal, which typically @@ -3151,7 +3153,6 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, * Don't pull any tasks if this group is already above the domain * average load. */ - sds.avg_load = (SCHED_LOAD_SCALE * sds.total_load) / sds.total_pwr; if (sds.this_load >= sds.avg_load) goto out_balanced; -- 1.7.0