Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965132Ab2JCQwf (ORCPT ); Wed, 3 Oct 2012 12:52:35 -0400 Received: from e32.co.us.ibm.com ([32.97.110.150]:50924 "EHLO e32.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S965019Ab2JCQwd (ORCPT ); Wed, 3 Oct 2012 12:52:33 -0400 Date: Wed, 3 Oct 2012 09:00:45 -0700 From: "Paul E. McKenney" To: "Srivatsa S. Bhat" Cc: Jiri Kosina , Christoph Lameter , Pekka Enberg , "Paul E. McKenney" , Josh Triplett , linux-kernel@vger.kernel.org, linux-mm@kvack.org Subject: Re: [PATCH v2] [RFC] mm, slab: release slab_mutex earlier in kmem_cache_destroy() Message-ID: <20121003160045.GD2527@linux.vnet.ibm.com> Reply-To: paulmck@linux.vnet.ibm.com References: <20121002233138.GD2465@linux.vnet.ibm.com> <20121003001530.GF2465@linux.vnet.ibm.com> <506C2E02.9080804@linux.vnet.ibm.com> <20121003145047.GA2527@linux.vnet.ibm.com> <506C51E0.1000602@linux.vnet.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <506C51E0.1000602@linux.vnet.ibm.com> User-Agent: Mutt/1.5.21 (2010-09-15) X-Content-Scanned: Fidelis XPS MAILER x-cbid: 12100316-5406-0000-0000-000000C8DA14 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 9853 Lines: 220 On Wed, Oct 03, 2012 at 08:25:28PM +0530, Srivatsa S. Bhat wrote: > On 10/03/2012 08:20 PM, Paul E. McKenney wrote: > > On Wed, Oct 03, 2012 at 05:52:26PM +0530, Srivatsa S. Bhat wrote: > >> On 10/03/2012 03:16 PM, Jiri Kosina wrote: > >>> On Wed, 3 Oct 2012, Jiri Kosina wrote: > >>> > >>>> Good question. I believe it should be safe to drop slab_mutex earlier, as > >>>> cachep has already been unlinked. I am adding slab people and linux-mm to > >>>> CC (the whole thread on LKML can be found at > >>>> https://lkml.org/lkml/2012/10/2/296 for reference). > >>>> > >>>> How about the patch below? Pekka, Christoph, please? > >>> > >>> It turns out that lockdep is actually getting this wrong, so the changelog > >>> in the previous version wasn't accurate. > >>> > >>> Please find patch with updated changelog below. Pekka, Christoph, could > >>> you please check whether it makes sense to you as well? Thanks. > >>> > >>> > >>> > >>> > >>> From: Jiri Kosina > >>> Subject: [PATCH] mm, slab: release slab_mutex earlier in kmem_cache_destroy() > >>> > >>> Commit 1331e7a1bbe1 ("rcu: Remove _rcu_barrier() dependency on > >>> __stop_machine()") introduced slab_mutex -> cpu_hotplug.lock > >>> dependency through kmem_cache_destroy() -> rcu_barrier() -> > >>> _rcu_barrier() -> get_online_cpus(). > >>> > >>> Lockdep thinks that this might actually result in ABBA deadlock, > >>> and reports it as below: > >>> > >>> === [ cut here ] === > >>> ====================================================== > >>> [ INFO: possible circular locking dependency detected ] > >>> 3.6.0-rc5-00004-g0d8ee37 #143 Not tainted > >>> ------------------------------------------------------- > >>> kworker/u:2/40 is trying to acquire lock: > >>> (rcu_sched_state.barrier_mutex){+.+...}, at: [] _rcu_barrier+0x26/0x1e0 > >>> > >>> but task is already holding lock: > >>> (slab_mutex){+.+.+.}, at: [] kmem_cache_destroy+0x45/0xe0 > >>> > >>> which lock already depends on the new lock. > >>> > >>> the existing dependency chain (in reverse order) is: > >>> > >>> -> #2 (slab_mutex){+.+.+.}: > >>> [] validate_chain+0x632/0x720 > >>> [] __lock_acquire+0x309/0x530 > >>> [] lock_acquire+0x121/0x190 > >>> [] __mutex_lock_common+0x5c/0x450 > >>> [] mutex_lock_nested+0x3e/0x50 > >>> [] cpuup_callback+0x2f/0xbe > >>> [] notifier_call_chain+0x93/0x140 > >>> [] __raw_notifier_call_chain+0x9/0x10 > >>> [] _cpu_up+0xba/0x14e > >>> [] cpu_up+0xbc/0x117 > >>> [] smp_init+0x6b/0x9f > >>> [] kernel_init+0x147/0x1dc > >>> [] kernel_thread_helper+0x4/0x10 > >>> > >>> -> #1 (cpu_hotplug.lock){+.+.+.}: > >>> [] validate_chain+0x632/0x720 > >>> [] __lock_acquire+0x309/0x530 > >>> [] lock_acquire+0x121/0x190 > >>> [] __mutex_lock_common+0x5c/0x450 > >>> [] mutex_lock_nested+0x3e/0x50 > >>> [] get_online_cpus+0x37/0x50 > >>> [] _rcu_barrier+0xbb/0x1e0 > >>> [] rcu_barrier_sched+0x10/0x20 > >>> [] rcu_barrier+0x9/0x10 > >>> [] deactivate_locked_super+0x49/0x90 > >>> [] deactivate_super+0x61/0x70 > >>> [] mntput_no_expire+0x127/0x180 > >>> [] sys_umount+0x6e/0xd0 > >>> [] system_call_fastpath+0x16/0x1b > >>> > >>> -> #0 (rcu_sched_state.barrier_mutex){+.+...}: > >>> [] check_prev_add+0x3de/0x440 > >>> [] validate_chain+0x632/0x720 > >>> [] __lock_acquire+0x309/0x530 > >>> [] lock_acquire+0x121/0x190 > >>> [] __mutex_lock_common+0x5c/0x450 > >>> [] mutex_lock_nested+0x3e/0x50 > >>> [] _rcu_barrier+0x26/0x1e0 > >>> [] rcu_barrier_sched+0x10/0x20 > >>> [] rcu_barrier+0x9/0x10 > >>> [] kmem_cache_destroy+0xd1/0xe0 > >>> [] nf_conntrack_cleanup_net+0xe4/0x110 [nf_conntrack] > >>> [] nf_conntrack_cleanup+0x2a/0x70 [nf_conntrack] > >>> [] nf_conntrack_net_exit+0x5e/0x80 [nf_conntrack] > >>> [] ops_exit_list+0x39/0x60 > >>> [] cleanup_net+0xfb/0x1b0 > >>> [] process_one_work+0x26b/0x4c0 > >>> [] worker_thread+0x12e/0x320 > >>> [] kthread+0x9e/0xb0 > >>> [] kernel_thread_helper+0x4/0x10 > >>> > >>> other info that might help us debug this: > >>> > >>> Chain exists of: > >>> rcu_sched_state.barrier_mutex --> cpu_hotplug.lock --> slab_mutex > >>> > >>> Possible unsafe locking scenario: > >>> > >>> CPU0 CPU1 > >>> ---- ---- > >>> lock(slab_mutex); > >>> lock(cpu_hotplug.lock); > >>> lock(slab_mutex); > >>> lock(rcu_sched_state.barrier_mutex); > >>> > >>> *** DEADLOCK *** > >>> === [ cut here ] === > >>> > >>> This is actually a false positive. Lockdep has no way of knowing the fact > >>> that the ABBA can actually never happen, because of special semantics of > >>> cpu_hotplug.refcount and itss handling in cpu_hotplug_begin(); the mutual > >>> exclusion there is not achieved through mutex, but through > >>> cpu_hotplug.refcount. > >>> > >>> The "neither cpu_up() nor cpu_down() will proceed past cpu_hotplug_begin() > >>> until everyone who called get_online_cpus() will call put_online_cpus()" > >>> semantics is totally invisible to lockdep. > >>> > >>> This patch therefore moves the unlock of slab_mutex so that rcu_barrier() > >>> is being called with it unlocked. It has two advantages: > >>> > >>> - it slightly reduces hold time of slab_mutex; as it's used to protect > >>> the cachep list, it's not necessary to hold it over __kmem_cache_destroy() > >>> call any more > >>> - it silences the lockdep false positive warning, as it avoids lockdep ever > >>> learning about slab_mutex -> cpu_hotplug.lock dependency > >>> > >>> Reviewed-by: Paul E. McKenney > >>> Signed-off-by: Jiri Kosina > >> > >> Reviewed-by: Srivatsa S. Bhat > >> > >> Earlier I was under the wrong impression that all the calltraces that lockdep > >> spewed were reflecting the same point in time. So, sorry about that noise! :-) > >> It is indeed a false-positive and there is no real bug here, and the fix looks > >> good, provided __kmem_cache_destroy() doesn't expect slab_mutex to be held. > > > > I am not so sure about it being a false positive. Consider the following > > sequence of events: > > > > o Thread A starts a CPU-hotplug operation, acquiring the > > hotplug mutex. > > > > o Thread B does a kmem_cache_destroy(), acquiring the slab mutex. > > This can't happen. Because kmem_cache_destroy() will call get_online_cpus() before > trying to acquire slab mutex. And it sleeps waiting at get_online_cpus() because > the hotplug lock has already been acquired by Thread A. Good point!!! False positive it is! Thanx, Paul > > o Thread A reaches the slab CPU-hotplug notifier, but cannot acquire > > the slab mutex because Thread B hold it. > > > > o Thread B enters rcu_barrier(), but cannot acquire the hotplug > > mutex because Thread A holds it. > > > > So I would argue that lockdep's output was a bit confusing, but that > > the deadlock it flagged is real. Or am I still missing something? > > > > So the key point is, Thread A is a hotplug writer. Thread B becomes a hotplug reader > the moment it calls get_online_cpus(). So they can't co-exist/run together. They will > get serialized. That is, Thread A runs to completion, releases hotplug lock. Only then > thread B gets past get_online_cpus(). > > Regards, > Srivatsa S. Bhat > > >> But, I'm also quite surprised that the put_online_cpus() code as it stands today > >> doesn't have any checks for the refcount going negative. I believe that such a > >> check would be valuable to help catch cases where we might end up inadvertently > >> causing an imbalance between get_online_cpus() and put_online_cpus(). I'll post > >> that as a separate patch. > >> > >> Regards, > >> Srivatsa S. Bhat > >> > >>> --- > >>> mm/slab.c | 2 +- > >>> 1 files changed, 1 insertions(+), 1 deletions(-) > >>> > >>> diff --git a/mm/slab.c b/mm/slab.c > >>> index 1133911..693c7cb 100644 > >>> --- a/mm/slab.c > >>> +++ b/mm/slab.c > >>> @@ -2801,12 +2801,12 @@ void kmem_cache_destroy(struct kmem_cache *cachep) > >>> put_online_cpus(); > >>> return; > >>> } > >>> + mutex_unlock(&slab_mutex); > >>> > >>> if (unlikely(cachep->flags & SLAB_DESTROY_BY_RCU)) > >>> rcu_barrier(); > >>> > >>> __kmem_cache_destroy(cachep); > >>> - mutex_unlock(&slab_mutex); > >>> put_online_cpus(); > >>> } > >>> EXPORT_SYMBOL(kmem_cache_destroy); > >>> > >> > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/