From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753067Ab0HXMrk (ORCPT ); Tue, 24 Aug 2010 08:47:40 -0400 Received: from e33.co.us.ibm.com ([32.97.110.151]:42808 "EHLO e33.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751375Ab0HXMri (ORCPT ); Tue, 24 Aug 2010 08:47:38 -0400 Date: Tue, 24 Aug 2010 18:17:26 +0530 From: Balbir Singh To: Peter Zijlstra Cc: Venkatesh Pallipadi , Martin Schwidefsky , Ingo Molnar , "H. Peter Anvin" , Thomas Gleixner , Paul Menage , linux-kernel@vger.kernel.org, Paul Turner , Heiko Carstens , Paul Mackerras , Tony Luck Subject: Re: [PATCH 0/4] Finer granularity and task/cgroup irq time accounting Message-ID: <20100824124726.GQ4684@balbir.in.ibm.com> Reply-To: balbir@linux.vnet.ibm.com References: <1279583835-22854-1-git-send-email-venki@google.com> <20100720095546.2f899e04@mschwide.boeblingen.de.ibm.com> <20100722131239.208d9501@mschwide.boeblingen.de.ibm.com> <1282636286.2605.2307.camel@laptop> <20100824080515.GK4684@balbir.in.ibm.com> <1282640953.2605.2428.camel@laptop> <20100824113801.GO4684@balbir.in.ibm.com> <1282650835.2605.2629.camel@laptop> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <1282650835.2605.2629.camel@laptop> User-Agent: Mutt/1.5.20 (2009-12-10) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org * Peter Zijlstra [2010-08-24 13:53:55]: > On Tue, 2010-08-24 at 17:08 +0530, Balbir Singh wrote: > > > > The point is for containers it is more likely to give the right answer > > and so on. Yes, the results are not 100% accurate. > > Consider one group heavily dirtying pages, it stuffs the IO queues full > and gets blocked on IO completion. Since the CPU is then free to > schedule something else we start running things from another group, > those IO completions will come in while we run other group and get > accounted to other group -- FAIL. > > s/group/task/ etc.. > > That just really doesn't work, accounting async work, esp stuff that is > not under software control it very tricky indeed. Yes, we don't have sufficient context to charge the correct context. I think openvz has some technology there, we will too when we have I/O cgroups at a cgroup level, but the instances of such operations are too many to accurately identify them all. > > So what are you wanting to do, and why. Do you really need accounting > madness? I think Venki gave the answer in the posting "There are usecases where reporting this time against task or task groups or cgroups will be useful for user/administrator in terms of resource planning and utilization charging" I don't have any specific use cases, I was just reviewing the patchset and trying to understand how to solve the problem. -- Three Cheers, Balbir