From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S967088AbXFHAaL (ORCPT ); Thu, 7 Jun 2007 20:30:11 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S966437AbXFHA3X (ORCPT ); Thu, 7 Jun 2007 20:29:23 -0400 Received: from www.osadl.org ([213.239.205.134]:53137 "EHLO mail.tglx.de" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S966405AbXFHA3W (ORCPT ); Thu, 7 Jun 2007 20:29:22 -0400 Message-Id: <20070607235801.494414000@linutronix.de> References: <20070607235106.387346000@linutronix.de> User-Agent: quilt/0.46-1 Date: Fri, 08 Jun 2007 00:29:20 -0000 From: Thomas Gleixner To: Andrew Morton Cc: LKML , Ingo Molnar , Steven Rostedt , Alexey Kuznetsov , Ulrich Drepper Subject: [patch 2/4] rt-mutex: Fix chain walk early wakeup bug Content-Disposition: inline; filename=rtmutex-fix-chain-walk.patch Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Alexey Kuznetsov found some problems in the pi-futex code. One of the root causes is: When a wakeup happens, we do not to stop the chain walk so we follow a not longer relevant locking chain. Drop out when this happens. Signed-off-by: Thomas Gleixner --- kernel/rtmutex.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) Index: linux-2.6.22-rc4/kernel/rtmutex.c =================================================================== --- linux-2.6.22-rc4.orig/kernel/rtmutex.c 2007-06-08 01:39:38.000000000 +0200 +++ linux-2.6.22-rc4/kernel/rtmutex.c 2007-06-08 01:39:38.000000000 +0200 @@ -189,6 +189,19 @@ int rt_mutex_adjust_prio_chain(struct ta if (!waiter || !waiter->task) goto out_unlock_pi; + /* + * Check the orig_waiter state. After we dropped the locks, + * the previous owner of the lock might have released the lock + * and made us the pending owner: + */ + if (orig_waiter && !orig_waiter->task) + goto out_unlock_pi; + + /* + * Drop out, when the task has no waiters. Note, + * top_waiter can be NULL, when we are in the deboosting + * mode! + */ if (top_waiter && (!task_has_pi_waiters(task) || top_waiter != task_top_pi_waiter(task))) goto out_unlock_pi; --