Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756547AbcKWNIw (ORCPT ); Wed, 23 Nov 2016 08:08:52 -0500 Received: from mail-wm0-f67.google.com ([74.125.82.67]:36759 "EHLO mail-wm0-f67.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751008AbcKWNIv (ORCPT ); Wed, 23 Nov 2016 08:08:51 -0500 Date: Wed, 23 Nov 2016 14:08:48 +0100 From: Daniel Vetter To: Peter Zijlstra Cc: Nicolai =?iso-8859-1?Q?H=E4hnle?= , Nicolai =?iso-8859-1?Q?H=E4hnle?= , linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, Ingo Molnar , stable@vger.kernel.org, Maarten Lankhorst Subject: Re: [PATCH 1/4] locking/ww_mutex: Fix a deadlock affecting ww_mutexes Message-ID: <20161123130848.q6yw73fjdhttmbqh@phenom.ffwll.local> Mail-Followup-To: Peter Zijlstra , Nicolai =?iso-8859-1?Q?H=E4hnle?= , Nicolai =?iso-8859-1?Q?H=E4hnle?= , linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, Ingo Molnar , stable@vger.kernel.org, Maarten Lankhorst References: <1479900325-28358-1-git-send-email-nhaehnle@gmail.com> <20161123130046.GS3092@twins.programming.kicks-ass.net> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20161123130046.GS3092@twins.programming.kicks-ass.net> X-Operating-System: Linux phenom 4.8.0-1-amd64 User-Agent: NeoMutt/20161104 (1.7.1) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3386 Lines: 78 On Wed, Nov 23, 2016 at 02:00:46PM +0100, Peter Zijlstra wrote: > On Wed, Nov 23, 2016 at 12:25:22PM +0100, Nicolai H?hnle wrote: > > From: Nicolai H?hnle > > > > Fix a race condition involving 4 threads and 2 ww_mutexes as indicated in > > the following example. Acquire context stamps are ordered like the thread > > numbers, i.e. thread #1 should back off when it encounters a mutex locked > > by thread #0 etc. > > > > Thread #0 Thread #1 Thread #2 Thread #3 > > --------- --------- --------- --------- > > lock(ww) > > success > > lock(ww') > > success > > lock(ww) > > lock(ww) . > > . . unlock(ww) part 1 > > lock(ww) . . . > > success . . . > > . . unlock(ww) part 2 > > . back off > > lock(ww') . > > . . > > (stuck) (stuck) > > > > Here, unlock(ww) part 1 is the part that sets lock->base.count to 1 > > (without being protected by lock->base.wait_lock), meaning that thread #0 > > can acquire ww in the fast path or, much more likely, the medium path > > in mutex_optimistic_spin. Since lock->base.count == 0, thread #0 then > > won't wake up any of the waiters in ww_mutex_set_context_fastpath. > > > > Then, unlock(ww) part 2 wakes up _only_the_first_ waiter of ww. This is > > thread #2, since waiters are added at the tail. Thread #2 wakes up and > > backs off since it sees ww owned by a context with a lower stamp. > > > > Meanwhile, thread #1 is never woken up, and so it won't back off its lock > > on ww'. So thread #0 gets stuck waiting for ww' to be released. > > > > This patch fixes the deadlock by waking up all waiters in the slow path > > of ww_mutex_unlock. > > > > We have an internal test case for amdgpu which continuously submits > > command streams from tens of threads, where all command streams reference > > hundreds of GPU buffer objects with a lot of overlap in the buffer lists > > between command streams. This test reliably caused a deadlock, and while I > > haven't completely confirmed that it is exactly the scenario outlined > > above, this patch does fix the test case. > > > > v2: > > - use wake_q_add > > - add additional explanations > > > > Cc: Peter Zijlstra > > Cc: Ingo Molnar > > Cc: Chris Wilson > > Cc: Maarten Lankhorst > > Cc: dri-devel@lists.freedesktop.org > > Cc: stable@vger.kernel.org > > Reviewed-by: Christian K?nig (v1) > > Signed-off-by: Nicolai H?hnle > > Completely and utterly fails to apply; I think this patch is based on > code prior to the mutex rewrite. > > Please rebase on tip/locking/core. > > Also, is this a regression, or has this been a 'feature' of the ww_mutex > code from early on? Sorry forgot to mention that, but I checked. Seems to have been broken since day 1, at least looking at the original code the wake-single-waiter stuff is as old as the mutex code added in 2006. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch