From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754225Ab3H2NlM (ORCPT ); Thu, 29 Aug 2013 09:41:12 -0400 Received: from mail-qe0-f46.google.com ([209.85.128.46]:39773 "EHLO mail-qe0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753106Ab3H2NlL (ORCPT ); Thu, 29 Aug 2013 09:41:11 -0400 Date: Thu, 29 Aug 2013 09:41:07 -0400 From: Tejun Heo To: Linus Torvalds Cc: linux-kernel@vger.kernel.org, Jamie Liu Subject: [GIT PULL] workqueue fixes for v3.11-rc7 Message-ID: <20130829134107.GA3985@htj.dyndns.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline 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 Hello, Linus. This pull request contains one fix which could lead to system-wide lockup on !PREEMPT kernels. It's very late in the cycle but this definitely is a -stable material. The problem is that workqueue worker tasks may process unlimited number of work items back-to-back without every yielding inbetween. This usually isn't noticeable but a work item which re-queues itself waiting for someone else to do something can deadlock with stop_machine. stop_machine will ensure nothing else happens on all other cpus and the requeueing work item will reqeueue itself indefinitely without ever yielding and thus preventing the CPU from entering stop_machine. Kudos to Jamie Liu for spotting and diagnosing the problem. This can be trivially fixed by adding cond_resched() after processing each work item. Thanks. The following changes since commit c95389b4cd6a4b52af78bea706a274453e886251: Merge branch 'akpm' (patches from Andrew Morton) (2013-08-28 19:31:33 -0700) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq.git for-3.11-fixes for you to fetch changes up to b22ce2785d97423846206cceec4efee0c4afd980: workqueue: cond_resched() after processing each work item (2013-08-29 09:19:28 -0400) ---------------------------------------------------------------- Tejun Heo (1): workqueue: cond_resched() after processing each work item kernel/workqueue.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 7f5d4be..e93f7b9 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -2201,6 +2201,15 @@ __acquires(&pool->lock) dump_stack(); } + /* + * The following prevents a kworker from hogging CPU on !PREEMPT + * kernels, where a requeueing work item waiting for something to + * happen could deadlock with stop_machine as such work item could + * indefinitely requeue itself while all other CPUs are trapped in + * stop_machine. + */ + cond_resched(); + spin_lock_irq(&pool->lock); /* clear cpu intensive status */