From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758691AbYEHGIx (ORCPT ); Thu, 8 May 2008 02:08:53 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754201AbYEHGIp (ORCPT ); Thu, 8 May 2008 02:08:45 -0400 Received: from mga07.intel.com ([143.182.124.22]:4174 "EHLO azsmga101.ch.intel.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753501AbYEHGIo (ORCPT ); Thu, 8 May 2008 02:08:44 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.27,453,1204531200"; d="scan'208";a="243693784" Subject: Re: volanoMark regression with kernel 2.6.26-rc1 From: "Zhang, Yanmin" To: Peter Zijlstra Cc: Ingo Molnar , LKML , Dhaval Giani , Srivatsa Vaddagiri In-Reply-To: <1210186728.6525.12.camel@lappy.programming.kicks-ass.net> References: <1210039590.3453.18.camel@ymzhang> <20080507091748.GB30645@elte.hu> <1210181657.6525.7.camel@lappy.programming.kicks-ass.net> <1210186728.6525.12.camel@lappy.programming.kicks-ass.net> Content-Type: text/plain Date: Thu, 08 May 2008 14:07:30 +0800 Message-Id: <1210226850.3453.119.camel@ymzhang> Mime-Version: 1.0 X-Mailer: Evolution 2.21.5 (2.21.5-2.fc9) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2008-05-07 at 20:58 +0200, Peter Zijlstra wrote: > On Wed, 2008-05-07 at 19:34 +0200, Peter Zijlstra wrote: > > On Wed, 2008-05-07 at 11:17 +0200, Ingo Molnar wrote: > > > * Zhang, Yanmin wrote: > > > > > > > Comparing with 2.6.25, volanoMark has big regression with kernel > > > > 2.6.26-rc1. It's about 50% on my 8-core stoakley, 16-core tigerton, > > > > and Itanium Montecito. > > > > > > > > With bisect, I located below patch. > > > > > > thanks Yanmin, i've queued up your patch that reverts this change. > > > > Is this really needed now that GROUP_SCHED defaults to 'n' ? > > > > Yanmin, this is with GROUP_SCHED=y, right or is this without? > > Its a long shot, but does the below help? I tested it on 8-core stoakley. The result is equal to the one of pure 2.6.26-rc1. > > --- > Subject: sched: fixup SMP load-balance > > Keeping the aggregate on the first cpu of the sched domain has two problems: > - it could collide between different sched domains on different cpus > - it could slow things down because of the remote accesses > > Signed-off-by: Peter Zijlstra > --- > include/linux/sched.h | 1 > kernel/sched.c | 113 +++++++++++++++++++++++--------------------------- > kernel/sched_fair.c | 12 ++--- > 3 files changed, 60 insertions(+), 66 deletions(-)