Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757758AbZFWHbi (ORCPT ); Tue, 23 Jun 2009 03:31:38 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752510AbZFWHbb (ORCPT ); Tue, 23 Jun 2009 03:31:31 -0400 Received: from fgwmail5.fujitsu.co.jp ([192.51.44.35]:45163 "EHLO fgwmail5.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752490AbZFWHba (ORCPT ); Tue, 23 Jun 2009 03:31:30 -0400 Date: Tue, 23 Jun 2009 16:29:51 +0900 From: KAMEZAWA Hiroyuki To: Daisuke Nishimura Cc: Ingo Molnar , "linux-mm@kvack.org" , "linux-kernel@vger.kernel.org" , "balbir@linux.vnet.ibm.com" , "menage@google.com" , "akpm@linux-foundation.org" Subject: Re: [BUGFIX][PATCH] fix bad page removal from LRU (Was Re: [RFC][PATCH] cgroup: fix permanent wait in rmdir Message-Id: <20090623162951.b4057104.kamezawa.hiroyu@jp.fujitsu.com> In-Reply-To: <20090623162759.94ba43d0.nishimura@mxp.nes.nec.co.jp> References: <20090622183707.dd9e665b.kamezawa.hiroyu@jp.fujitsu.com> <20090622105231.GA17242@elte.hu> <20090623085755.9cf75da2.kamezawa.hiroyu@jp.fujitsu.com> <20090623162759.94ba43d0.nishimura@mxp.nes.nec.co.jp> Organization: FUJITSU Co. LTD. X-Mailer: Sylpheed 2.5.0 (GTK+ 2.10.14; i686-pc-mingw32) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 6119 Lines: 109 On Tue, 23 Jun 2009 16:27:59 +0900 Daisuke Nishimura wrote: > On Tue, 23 Jun 2009 08:57:55 +0900, KAMEZAWA Hiroyuki wrote: > > On Mon, 22 Jun 2009 12:52:31 +0200 > > Ingo Molnar wrote: > > > > > > > > FYI, there's a new cgroup related list corruption warning/crash that > > > i've seen a lot of times in latest -tip tests: > > > > > > [ 478.555544] ------------[ cut here ]------------ > > > [ 478.556523] WARNING: at lib/list_debug.c:26 __list_add+0x70/0xa0() > > > [ 478.556523] Hardware name: > > > [ 478.556523] list_add corruption. next->prev should be prev (ffff88003e640448), but was ffff88003fa1a6e8. (next=ffff88003fa1a8a0). > > > [ 478.556523] Modules linked in: > > > [ 478.556523] Pid: 470, comm: kswapd0 Not tainted 2.6.30-tip #10989 > > > [ 478.556523] Call Trace: > > > [ 478.556523] [] ? __list_add+0x70/0xa0 > > > [ 478.556523] [] warn_slowpath_common+0x8c/0xc0 > > > [ 478.556523] [] warn_slowpath_fmt+0x69/0x70 > > > [ 478.556523] [] ? __lock_acquired+0x18b/0x2b0 > > > [ 478.556523] [] ? page_check_address+0x110/0x1a0 > > > [ 478.556523] [] ? cpumask_any_but+0x42/0xb0 > > > [ 478.556523] [] ? __lock_release+0x38/0x90 > > > [ 478.556523] [] ? page_referenced_one+0x91/0x120 > > > [ 478.556523] [] __list_add+0x70/0xa0 > > > [ 478.556523] [] mem_cgroup_add_lru_list+0x63/0x70 > > > [ 478.556523] [] move_active_pages_to_lru+0xf4/0x180 > > > [ 478.556523] [] ? shrink_active_list+0x1f8/0x2a0 > > > [ 478.556523] [] ? shrink_active_list+0x1f8/0x2a0 > > > [ 478.556523] [] shrink_active_list+0x234/0x2a0 > > > [ 478.556523] [] shrink_zone+0x173/0x1f0 > > > [ 478.556523] [] balance_pgdat+0x4da/0x4e0 > > > [ 478.556523] [] ? isolate_pages_global+0x0/0x60 > > > [ 478.556523] [] kswapd+0x106/0x150 > > > [ 478.556523] [] ? autoremove_wake_function+0x0/0x40 > > > [ 478.556523] [] ? kswapd+0x0/0x150 > > > [ 478.556523] [] kthread+0x9e/0xb0 > > > [ 478.556523] [] child_rip+0xa/0x20 > > > [ 478.556523] [] ? restore_args+0x0/0x30 > > > [ 478.556523] [] ? kthreadd+0xb5/0x100 > > > [ 478.556523] [] ? kthread+0x0/0xb0 > > > [ 478.556523] [] ? child_rip+0x0/0x20 > > > [ 478.556523] ---[ end trace 9f3122957c34141e ]--- > > > [ 484.923530] ------------[ cut here ]------------ > > > [ 484.924525] WARNING: at lib/list_debug.c:26 __list_add+0x70/0xa0() > > > [ 484.924525] Hardware name: > > > [ 484.924525] list_add corruption. next->prev should be prev (ffff88003e640448), but was ffff88003fa192e8. (next=ffff88003fa14d88). > > > [ 484.941152] Modules linked in: > > > [ 484.941152] Pid: 470, comm: kswapd0 Tainted: G W 2.6.30-tip #10989 > > > [ 484.941152] Call Trace: > > > [ 484.941152] [] ? __list_add+0x70/0xa0 > > > [ 484.941152] [] warn_slowpath_common+0x8c/0xc0 > > > [ 484.941152] [] warn_slowpath_fmt+0x69/0x70 > > > [ 484.941152] [] ? __lock_acquired+0x18b/0x2b0 > > > [ 484.941152] [] ? page_check_address+0x110/0x1a0 > > > [ 484.941152] [] ? cpumask_any_but+0x42/0xb0 > > > [ 484.941152] [] ? __lock_release+0x38/0x90 > > > [ 484.941152] [] ? page_referenced_one+0x91/0x120 > > > [ 484.941152] [] __list_add+0x70/0xa0 > > > [ 484.941152] [] mem_cgroup_add_lru_list+0x63/0x70 > > > [ 484.941152] [] move_active_pages_to_lru+0xf4/0x180 > > > [ 484.941152] [] ? shrink_active_list+0x1f8/0x2a0 > > > [ 484.941152] [] ? shrink_active_list+0x1f8/0x2a0 > > > [ 484.941152] [] shrink_active_list+0x234/0x2a0 > > > [ 484.941152] [] shrink_zone+0x173/0x1f0 > > > [ 484.941152] [] balance_pgdat+0x4da/0x4e0 > > > [ 484.941152] [] ? isolate_pages_global+0x0/0x60 > > > [ 484.941152] [] kswapd+0x106/0x150 > > > [ 484.941152] [] ? autoremove_wake_function+0x0/0x40 > > > [ 484.941152] [] ? kswapd+0x0/0x150 > > > [ 484.941152] [] kthread+0x9e/0xb0 > > > [ 484.941152] [] child_rip+0xa/0x20 > > > [ 484.941152] [] ? restore_args+0x0/0x30 > > > [ 484.941152] [] ? kthreadd+0xb5/0x100 > > > [ 484.941152] [] ? kthread+0x0/0xb0 > > > [ 484.941152] [] ? child_rip+0x0/0x20 > > > [ 484.941152] ---[ end trace 9f3122957c34141f ]--- > > > [ 485.365631] ------------[ cut here ]------------ > > > [ 485.368029] WARNING: at lib/list_debug.c:26 __list_add+0x70/0xa0() > > > > > > has this been reported before? Is there a fix for it i missed? > > > > > > > I think this is a fix for the problem. Sorry for regression. > > fix for "memcg: fix lru rotation in isolate_pages" patch in 2.6.30-git18. > > > > == > > From: KAMEZAWA Hiroyuki > > > > A page isolated is "cursor_page" not "page". > > This causes list corruption finally. > > > > Signed-off-by: KAMEZAWA Hiroyuki > > I reproduced this bug and have confirmed that this patch fixes it. > > Tested-by: Daisuke Nishimura > Thank you!! -Kame -- 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/