From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751601Ab2KJFcd (ORCPT ); Sat, 10 Nov 2012 00:32:33 -0500 Received: from order.stressinduktion.org ([87.106.68.36]:44987 "EHLO order.stressinduktion.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750704Ab2KJFca (ORCPT ); Sat, 10 Nov 2012 00:32:30 -0500 Date: Sat, 10 Nov 2012 06:32:28 +0100 From: Hannes Frederic Sowa To: linux-kernel@vger.kernel.org Cc: a.p.zijlstra@chello.nl, paulus@samba.org, mingo@redhat.com, acme@ghostprotocols.net Subject: [PATCH] perf: call perf_event_comm under task_lock to fix suspicious rcu usage Message-ID: <20121110053228.GB11317@order.stressinduktion.org> Mail-Followup-To: linux-kernel@vger.kernel.org, a.p.zijlstra@chello.nl, paulus@samba.org, mingo@redhat.com, acme@ghostprotocols.net Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Following RCU warning showed up while executing a shebang-script under perf-record (could even be an empty script) on a 3.7-rc4 stable kernel: [ 32.185108] [ 32.185332] =============================== [ 32.185602] [ INFO: suspicious RCU usage. ] [ 32.185903] 3.7.0-rc4 #1 Not tainted [ 32.186021] ------------------------------- [ 32.186021] include/linux/cgroup.h:566 suspicious rcu_dereference_check() usage! [ 32.186021] [ 32.186021] other info that might help us debug this: [ 32.186021] [ 32.186021] [ 32.186021] rcu_scheduler_active = 1, debug_locks = 0 [ 32.186021] 1 lock held by empty.sh/556: [ 32.186021] #0: (&sig->cred_guard_mutex){+.+.+.}, at: [] prepare_bprm_creds+0x36/0x80 [ 32.186021] [ 32.186021] stack backtrace: [ 32.186021] Pid: 556, comm: empty.sh Not tainted 3.7.0-rc4 #1 [ 32.186021] Call Trace: [ 32.186021] [] lockdep_rcu_suspicious+0xfd/0x130 [ 32.186021] [] perf_event_comm+0x436/0x610 [ 32.186021] [] ? trace_hardirqs_off+0xd/0x10 [ 32.186021] [] ? local_clock+0x6f/0x80 [ 32.186021] [] ? lock_release_holdtime.part.26+0xf/0x180 [ 32.186021] [] set_task_comm+0x73/0x180 [ 32.186021] [] setup_new_exec+0x9a/0x210 [ 32.186021] [] load_elf_binary+0x3e3/0x1ab0 [ 32.186021] [] ? sched_clock_local+0x25/0xa0 [ 32.186021] [] ? sched_clock_cpu+0xa8/0x120 [ 32.186021] [] ? trace_hardirqs_off+0xd/0x10 [ 32.186021] [] ? local_clock+0x6f/0x80 [ 32.186021] [] ? load_elf_library+0x240/0x240 [ 32.186021] [] ? load_elf_library+0x240/0x240 [ 32.186021] [] search_binary_handler+0x194/0x4f0 [ 32.186021] [] ? search_binary_handler+0x5f/0x4f0 [ 32.186021] [] ? compat_sys_ioctl+0x1510/0x1510 [ 32.186021] [] load_script+0x294/0x2c0 [ 32.186021] [] ? lock_release_holdtime.part.26+0xf/0x180 [ 32.186021] [] ? compat_sys_ioctl+0x1510/0x1510 [ 32.186021] [] search_binary_handler+0x194/0x4f0 [ 32.186021] [] ? search_binary_handler+0x5f/0x4f0 [ 32.186021] [] do_execve_common.isra.25+0x50b/0x5b0 [ 32.186021] [] ? do_execve_common.isra.25+0x12a/0x5b0 [ 32.186021] [] do_execve+0x1b/0x20 [ 32.186021] [] sys_execve+0x54/0x80 [ 32.186021] [] stub_execve+0x69/0xc0 I think this dereference qualifies for the task_lock exception (as noted in kernel/cgroup.c), thus this patch ensures calling perf_event_comm before giving up the task_lock. Changelog -v2 (since <20121103235758.GD18106@order.stressinduktion.org>): 1) rebased to 3.7-rc4 2) slightly improved/updated commit msg and added more people to Cc Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Ingo Molnar Cc: Arnaldo Carvalho de Melo Signed-off-by: Hannes Frederic Sowa --- fs/exec.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/exec.c b/fs/exec.c index 0039055..a961b9d 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1038,8 +1038,8 @@ void set_task_comm(struct task_struct *tsk, char *buf) memset(tsk->comm, 0, TASK_COMM_LEN); wmb(); strlcpy(tsk->comm, buf, sizeof(tsk->comm)); - task_unlock(tsk); perf_event_comm(tsk); + task_unlock(tsk); } static void filename_to_taskname(char *tcomm, const char *fn, unsigned int len)