From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932574Ab2ENWM4 (ORCPT ); Mon, 14 May 2012 18:12:56 -0400 Received: from mail-pz0-f46.google.com ([209.85.210.46]:64506 "EHLO mail-pz0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757900Ab2ENWMz (ORCPT ); Mon, 14 May 2012 18:12:55 -0400 Date: Mon, 14 May 2012 15:12:50 -0700 From: Tejun Heo To: "Paul E. McKenney" Cc: linux-kernel@vger.kernel.org Subject: [PATCH] workqueue: skip nr_running sanity check in worker_enter_idle() if trustee is active Message-ID: <20120514221250.GA8414@google.com> References: <20120506153814.GA25681@linux.vnet.ibm.com> <20120507194042.GG19417@google.com> <20120507205516.GD21152@linux.vnet.ibm.com> <20120507213449.GM19417@google.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20120507213449.GM19417@google.com> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org >>From 544ecf310f0e7f51fa057ac2a295fc1b3b35a9d3 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 14 May 2012 15:04:50 -0700 worker_enter_idle() has WARN_ON_ONCE() which triggers if nr_running isn't zero when every worker is idle. This can trigger spuriously while a cpu is going down due to the way trustee sets %WORKER_ROGUE and zaps nr_running. It first sets %WORKER_ROGUE on all workers without updating nr_running, releases gcwq->lock, schedules, regrabs gcwq->lock and then zaps nr_running. If the last running worker enters idle inbetween, it would see stale nr_running which hasn't been zapped yet and trigger the WARN_ON_ONCE(). Fix it by performing the sanity check iff the trustee is idle. Signed-off-by: Tejun Heo Reported-by: "Paul E. McKenney" Cc: stable@vger.kernel.org --- Sorry about the delay. After scratching my head quite a bit, I found where during cpu-offlining such discrepancy may happen. I'm fairly sure this is it but I might be wrong, so please include this patch in your test setup and let me know how it goes. Thank you. kernel/workqueue.c | 9 +++++++-- 1 files changed, 7 insertions(+), 2 deletions(-) diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 211eadb..c36c86c 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -1213,8 +1213,13 @@ static void worker_enter_idle(struct worker *worker) } else wake_up_all(&gcwq->trustee_wait); - /* sanity check nr_running */ - WARN_ON_ONCE(gcwq->nr_workers == gcwq->nr_idle && + /* + * Sanity check nr_running. Because trustee releases gcwq->lock + * between setting %WORKER_ROGUE and zapping nr_running, the + * warning may trigger spuriously. Check iff trustee is idle. + */ + WARN_ON_ONCE(gcwq->trustee_state == TRUSTEE_DONE && + gcwq->nr_workers == gcwq->nr_idle && atomic_read(get_gcwq_nr_running(gcwq->cpu))); } -- 1.7.7.3