From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail138.messagelabs.com (mail138.messagelabs.com [216.82.249.35]) by kanga.kvack.org (Postfix) with SMTP id 8A86E6B005C for ; Thu, 15 Jan 2009 02:22:34 -0500 (EST) Received: from m1.gw.fujitsu.co.jp ([10.0.50.71]) by fgwmail5.fujitsu.co.jp (Fujitsu Gateway) with ESMTP id n0F7MVtC032088 for (envelope-from kamezawa.hiroyu@jp.fujitsu.com); Thu, 15 Jan 2009 16:22:32 +0900 Received: from smail (m1 [127.0.0.1]) by outgoing.m1.gw.fujitsu.co.jp (Postfix) with ESMTP id 2998F45DD76 for ; Thu, 15 Jan 2009 16:22:33 +0900 (JST) Received: from s1.gw.fujitsu.co.jp (s1.gw.fujitsu.co.jp [10.0.50.91]) by m1.gw.fujitsu.co.jp (Postfix) with ESMTP id 05CE445DD75 for ; Thu, 15 Jan 2009 16:22:33 +0900 (JST) Received: from s1.gw.fujitsu.co.jp (localhost.localdomain [127.0.0.1]) by s1.gw.fujitsu.co.jp (Postfix) with ESMTP id 6E4371DB8046 for ; Thu, 15 Jan 2009 16:22:31 +0900 (JST) Received: from m105.s.css.fujitsu.com (m105.s.css.fujitsu.com [10.249.87.105]) by s1.gw.fujitsu.co.jp (Postfix) with ESMTP id 0F35A1DB803A for ; Thu, 15 Jan 2009 16:22:31 +0900 (JST) Date: Thu, 15 Jan 2009 16:21:26 +0900 From: KAMEZAWA Hiroyuki Subject: Re: [RFC] [PATCH] memcg: fix infinite loop Message-Id: <20090115162126.cf040c63.kamezawa.hiroyu@jp.fujitsu.com> In-Reply-To: <496EE25E.3030703@cn.fujitsu.com> References: <496ED2B7.5050902@cn.fujitsu.com> <20090115061557.GD30358@balbir.in.ibm.com> <20090115153134.632ebc85.kamezawa.hiroyu@jp.fujitsu.com> <496EE25E.3030703@cn.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org To: Li Zefan Cc: balbir@linux.vnet.ibm.com, Daisuke Nishimura , "linux-mm@kvack.org" List-ID: On Thu, 15 Jan 2009 15:14:38 +0800 Li Zefan wrote: > KAMEZAWA Hiroyuki wrote: > > On Thu, 15 Jan 2009 11:45:57 +0530 > > Balbir Singh wrote: > > > >> * Li Zefan [2009-01-15 14:07:51]: > >> > >>> 1. task p1 is in /memcg/0 > >>> 2. p1 does mmap(4096*2, MAP_LOCKED) > >>> 3. echo 4096 > /memcg/0/memory.limit_in_bytes > >>> > >>> The above 'echo' will never return, unless p1 exited or freed the memory. > >>> The cause is we can't reclaim memory from p1, so the while loop in > >>> mem_cgroup_resize_limit() won't break. > >>> > >>> This patch fixes it by decrementing retry_count regardless the return value > >>> of mem_cgroup_hierarchical_reclaim(). > >>> > >> The problem definitely seems to exist, shouldn't we fix reclaim to > >> return 0, so that we know progress is not made and retry count > >> decrements? > >> > > > > The behavior is correct. And we already check signal_pending() in the loop. > > Ctrl-C or SIGALARM will works better than checking retry count. > > But this behavior seems like a regression. Please try it in 2.6.28, you'll see > it returns EBUSY immediately. > > Looks like the return value of mem_cgroup_hierarchical_reclaim() is buggy ? > This is intentional behavior change by == memcg-make-oom-less-frequently.patch == try_to_free_page() returns positive value if try_to_free_page() reclaims at least 1 pages. It itself doesn't seem to be buggy. What buggy is resize_limit's retry-out check code, I think. How about following ? == while (1) { if (signal_pending()) break; try to set limit .... ... ret = mem_cgroup_hierarchical_reclaim(memcg, GFP_KERNEL, false); total_progress += ret; if (total_progress > (memcg->res.usage - val) * 2) { /* * It seems we reclaimed twice of necessary * pages...this memcg is busy */ ret = -EBUSY; break; } } == Thanks, -Kame -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org