From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755933Ab1F1IJR (ORCPT ); Tue, 28 Jun 2011 04:09:17 -0400 Received: from cantor2.suse.de ([195.135.220.15]:59340 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756463Ab1F1IIw (ORCPT ); Tue, 28 Jun 2011 04:08:52 -0400 Date: Tue, 28 Jun 2011 10:08:47 +0200 From: Michal Hocko To: KAMEZAWA Hiroyuki Cc: Andrew Morton , torvalds@linux-foundation.org, kamezawa.hiroyu@jp.fujitsu.com, nishimura@mxp.nes.nec.co.jp, yinghan@google.com, linux-mm@kvack.org, LKML Subject: Re: [patch 14/22] memcg: fix direct softlimit reclaim to be called in limit path Message-ID: <20110628080847.GA16518@tiehlicka.suse.cz> References: <201106272318.p5RNICJW001465@imap1.linux-foundation.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <201106272318.p5RNICJW001465@imap1.linux-foundation.org> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org I am sorry, that I am answering that late but I didn't get to this sooner. On Mon 27-06-11 16:18:12, Andrew Morton wrote: > From: KAMEZAWA Hiroyuki > > commit d149e3b ("memcg: add the soft_limit reclaim in global direct > reclaim") adds a softlimit hook to shrink_zones(). By this, soft limit is > called as > > try_to_free_pages() > do_try_to_free_pages() > shrink_zones() > mem_cgroup_soft_limit_reclaim() > > Then, direct reclaim is memcg softlimit hint aware, now. > > But, the memory cgroup's "limit" path can call softlimit shrinker. > > try_to_free_mem_cgroup_pages() > do_try_to_free_pages() > shrink_zones() > mem_cgroup_soft_limit_reclaim() > > This will cause a global reclaim when a memcg hits limit. Sorry, I do not get it. How does it cause the global reclaim? Did you mean soft reclaim? > > This is bug. soft_limit_reclaim() should be called when > scanning_global_lru(sc) == true. Agreed > > And the commit adds a variable "total_scanned" for counting softlimit > scanned pages....it's not "total". This patch removes the variable and > update sc->nr_scanned instead of it. This will affect shrink_slab()'s > scan condition but, global LRU is scanned by softlimit and I think this > change makes sense. Yes, the previous semantic was really confusing (part of the scanned accounting is returned by shrink_zones and the other part in sc). This could be ideally a separate change because it fixes a different bug (softlimit is not considered for slab srhinking) > > TODO: avoid too much scanning of a zone when softlimit did enough work. > > Signed-off-by: KAMEZAWA Hiroyuki > Cc: Daisuke Nishimura > Cc: Ying Han > Cc: Michal Hocko > Signed-off-by: Andrew Morton Just in case it is not late yet. Reviewed-by: Michal Hocko -- Michal Hocko SUSE Labs SUSE LINUX s.r.o. Lihovarska 1060/12 190 00 Praha 9 Czech Republic