From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754383Ab1I3Bnr (ORCPT ); Thu, 29 Sep 2011 21:43:47 -0400 Received: from mail-yw0-f46.google.com ([209.85.213.46]:44325 "EHLO mail-yw0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751346Ab1I3Bnp (ORCPT ); Thu, 29 Sep 2011 21:43:45 -0400 Date: Thu, 29 Sep 2011 20:43:34 -0500 From: Jonathan Nieder To: Linus Torvalds Cc: linux-kernel@vger.kernel.org, Andrew Morton , Nelson Elhage , Steven Rostedt , Catalin Marinas Subject: [PATCH 3.1-rc8 0/2] Two false positives from debugging tools (lockdep, kmemleak) Message-ID: <20110930014334.GA2368@elie.chipublib.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.5.21+46 (b01d63af6fea) (2011-07-01) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, Especially during the -rc period, it is nice to be able to take advantage of lockdep and kmemleak to find bugs. False positives make it harder to find real problems amid the noise. At boot time, I get: | allocated 14680064 bytes of page_cgroup | please try 'cgroup_disable=memory' option if you don't want memory cgroups | kmemleak: Trying to color unknown object at 0xffff88006f500000 as Grey | Pid: 0, comm: swapper Not tainted 3.1.0-rc8-00031-gd5767c53535a #12 | Call Trace: | [] paint_ptr+0x56/0x8e | [] kmemleak_not_leak+0x23/0x42 | [] kmemleak_init+0x1db/0x23a | [] start_kernel+0x31c/0x3b1 | [] x86_64_start_reservations+0xaf/0xb3 | [] ? early_idt_handlers+0x140/0x140 | [] x86_64_start_kernel+0x102/0x111 | kmemleak: Trying to color unknown object at 0xffff88006f600000 as Grey [and so on, for 13 addresses in all] Also: | [ 222.425445] systemd-logind/908 is trying to acquire lock: | [ 222.425447] (&ep->mtx){+.+.+.}, at: [] ep_scan_ready_list+0x3c/0x1a7 | [ 222.425456] | [ 222.425457] but task is already holding lock: | [ 222.425459] (&ep->mtx){+.+.+.}, at: [] sys_epoll_ctl+0x11d/0x520 Luckily both false positives seem to be known and well understood. The first of the patches below (the kmemleak one) seems to have fallen under the radar --- I can't even find it in linux-next. The second (the lockdep one) is taken from akpm's -mm patchset. Thoughts? Nelson Elhage (1): epoll: fix spurious lockdep warnings Steven Rostedt (1): cgroup/kmemleak: Annotate alloc_page() for cgroup allocations fs/eventpoll.c | 25 ++++++++++++++++++------- mm/page_cgroup.c | 7 +++++-- 2 files changed, 23 insertions(+), 9 deletions(-) -- 1.7.7.rc1