Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759606AbZLLNGs (ORCPT ); Sat, 12 Dec 2009 08:06:48 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1759506AbZLLNGr (ORCPT ); Sat, 12 Dec 2009 08:06:47 -0500 Received: from fg-out-1718.google.com ([72.14.220.156]:9715 "EHLO fg-out-1718.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758565AbZLLNGq convert rfc822-to-8bit (ORCPT ); Sat, 12 Dec 2009 08:06:46 -0500 MIME-Version: 1.0 In-Reply-To: <20091212125046.14df3134.d-nishimura@mtf.biglobe.ne.jp> References: <747ea0ec22b9348208c80f86f7a813728bf8e50a.1260571675.git.kirill@shutemov.name> <20091212125046.14df3134.d-nishimura@mtf.biglobe.ne.jp> Date: Sat, 12 Dec 2009 15:06:52 +0200 Message-ID: Subject: Re: [PATCH RFC v2 3/4] memcg: rework usage of stats by soft limit From: "Kirill A. Shutemov" To: nishimura@mxp.nes.nec.co.jp Cc: containers@lists.linux-foundation.org, linux-mm@kvack.org, Paul Menage , Li Zefan , Andrew Morton , KAMEZAWA Hiroyuki , Balbir Singh , Pavel Emelyanov , Dan Malek , Vladislav Buzov , linux-kernel@vger.kernel.org Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1519 Lines: 35 On Sat, Dec 12, 2009 at 5:50 AM, Daisuke Nishimura wrote: > Sorry, I disagree this change. > > mem_cgroup_soft_limit_check() is used for checking how much current usage exceeds > the soft_limit_in_bytes and updating softlimit tree asynchronously, instead of > checking every charge/uncharge. What if you change the soft_limit_in_bytes, > but the number of charges and uncharges are very balanced afterwards ? > The softlimit tree will not be updated for a long time. I don't see how my patch affects the logic you've described. Statistics updates and checks in the same place. It just uses decrement instead of increment. > > And IIUC, it's the same for your threshold feature, right ? > I think it would be better: > > - discard this change. > - in 4/4, rename mem_cgroup_soft_limit_check to mem_cgroup_event_check, >  and instead of adding a new STAT counter, do like: > >        if (mem_cgroup_event_check(mem)) { >                mem_cgroup_update_tree(mem, page); >                mem_cgroup_threshold(mem); >        } I think that mem_cgroup_update_tree() and mem_cgroup_threshold() should be run with different frequency. How to share MEM_CGROUP_STAT_EVENTS between soft limits and thresholds in this case? -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/