From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755956Ab3GYNNt (ORCPT ); Thu, 25 Jul 2013 09:13:49 -0400 Received: from e23smtp06.au.ibm.com ([202.81.31.148]:52172 "EHLO e23smtp06.au.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755678Ab3GYNNr (ORCPT ); Thu, 25 Jul 2013 09:13:47 -0400 Message-ID: <51F1263A.6070704@linux.vnet.ibm.com> Date: Thu, 25 Jul 2013 18:50:58 +0530 From: Raghavendra K T Organization: IBM User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130625 Thunderbird/17.0.7 MIME-Version: 1.0 To: Eric Dumazet CC: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, "David S. Miller" , "Eric W. Biederman" , Eric Dumazet , Al Viro , Gao feng Subject: Re: Hack bench regression with CONFIG_SLUB_CPU_PARTIAL disabled (info only) References: <51F0DB11.2080508@linux.vnet.ibm.com> <1374750359.23313.17.camel@edumazet-glaptop> In-Reply-To: <1374750359.23313.17.camel@edumazet-glaptop> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-Content-Scanned: Fidelis XPS MAILER x-cbid: 13072513-7014-0000-0000-00000360668D Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 07/25/2013 04:35 PM, Eric Dumazet wrote: > On Thu, 2013-07-25 at 13:30 +0530, Raghavendra K T wrote: >> Hi, >> >> While testing hackbench with 3.11-rc1 I observed hackbench testcase >> took too long to complete (32 vcpu kvm guest on 32 core HT off machine). >> >> hackbench 1x degraded by 3 time (40sec vs 112sec) >> hackbench 2x degraded by around 20time (90sec vs 1800 sec) >> >> When I bisected, I found that I had CONFIG_SLUB_CPU_PARTIAL >> accidentally disabled. >> I understand that CONFIG_SLUB_CPU_PARTIAL=y by default and the config >> option is to provide determinism for RT kernels and thus it is not a >> problem. >> But thought of sharing this here so that if somebody runs into same >> issue accidently, it would save their time. > > Incidentally, what difference do you currently get with SLUB and SLAB ? > tested now, hackbench 1x is almost 2 times slower (40 vs 93) hackbench 2x is almost 6times slower (90sec vs 528)