From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756223AbYDJJLX (ORCPT ); Thu, 10 Apr 2008 05:11:23 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754806AbYDJJLQ (ORCPT ); Thu, 10 Apr 2008 05:11:16 -0400 Received: from e28smtp06.in.ibm.com ([59.145.155.6]:41614 "EHLO e28smtp06.in.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754630AbYDJJLP (ORCPT ); Thu, 10 Apr 2008 05:11:15 -0400 Message-ID: <47FDD947.8020600@linux.vnet.ibm.com> Date: Thu, 10 Apr 2008 14:39:27 +0530 From: Balbir Singh Reply-To: balbir@linux.vnet.ibm.com Organization: IBM User-Agent: Thunderbird 2.0.0.12 (X11/20080226) MIME-Version: 1.0 To: Paul Menage CC: Pavel Emelianov , Hugh Dickins , Sudhir Kumar , YAMAMOTO Takashi , linux-kernel@vger.kernel.org, taka@valinux.co.jp, linux-mm@kvack.org, David Rientjes , Andrew Morton , KAMEZAWA Hiroyuki Subject: Re: [-mm] Add an owner to the mm_struct (v8) References: <20080404080544.26313.38199.sendpatchset@localhost.localdomain> <47F7BB69.3000502@linux.vnet.ibm.com> <6599ad830804051057n2f2802e4w6179f2e108467494@mail.gmail.com> <47F7CC08.4090209@linux.vnet.ibm.com> <6599ad830804051629k3649dbc4na92bb3d0cd7a0492@mail.gmail.com> <47F861C8.7080700@linux.vnet.ibm.com> <6599ad830804072337g2e7b4613hdcc05062dc2ca4e0@mail.gmail.com> <47FB162D.1020506@linux.vnet.ibm.com> <6599ad830804072357o2fd5e9bco3309d151e270e62e@mail.gmail.com> <47FB193A.8070801@linux.vnet.ibm.com> <6599ad830804080029v1d8f2ff7g5254f32362fd7cb9@mail.gmail.com> In-Reply-To: <6599ad830804080029v1d8f2ff7g5254f32362fd7cb9@mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Paul Menage wrote: > On Tue, Apr 8, 2008 at 12:05 AM, Balbir Singh wrote: >> Paul Menage wrote: >> > On Mon, Apr 7, 2008 at 11:52 PM, Balbir Singh wrote: >> >> I agree, but like I said earlier, this was the easily available ready made >> >> application I found. Do you know of any other highly threaded micro benchmark? >> >> >> > >> > How about a simple program that creates N threads that just sleep, >> > then has the main thread exit? >> > >> >> That is not really representative of anything. I have that program handy. How do >> we measure the impact on throughput? > > It's very representative of how much additional overhead in terms of > mm->owner churn there is in a large multi-threaded application > exiting, which is the thing that you're trying to optimize with the > delayed thread group leader checks. > I see almost no overhead after the notification change optimization (notify only if owner belongs to a different cgroup). My program creates n processes with k threads each and forces the thread group leader to exit. For my experiment I created 10 processes with 800 threads each (NOTE: you need to change ulimit -s for this to work). I am going to remove the delay_group_leader() optimization and submit v9. -- Warm Regards, Balbir Singh Linux Technology Center IBM, ISTL