From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx176.postini.com [74.125.245.176]) by kanga.kvack.org (Postfix) with SMTP id 2C9816B005A for ; Tue, 17 Jul 2012 07:58:33 -0400 (EDT) Received: from /spool/local by e6.ny.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Tue, 17 Jul 2012 07:58:28 -0400 Received: from d01relay04.pok.ibm.com (d01relay04.pok.ibm.com [9.56.227.236]) by d01dlp02.pok.ibm.com (Postfix) with ESMTP id 738E86E804C for ; Tue, 17 Jul 2012 07:57:26 -0400 (EDT) Received: from d03av01.boulder.ibm.com (d03av01.boulder.ibm.com [9.17.195.167]) by d01relay04.pok.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id q6HBvPU7427342 for ; Tue, 17 Jul 2012 07:57:25 -0400 Received: from d03av01.boulder.ibm.com (loopback [127.0.0.1]) by d03av01.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id q6HBvN7Z002730 for ; Tue, 17 Jul 2012 05:57:25 -0600 Date: Tue, 17 Jul 2012 04:57:21 -0700 From: "Paul E. McKenney" Subject: Re: mm,numad,rcu: hang on OOM Message-ID: <20120717115721.GC2438@linux.vnet.ibm.com> Reply-To: paulmck@linux.vnet.ibm.com References: <1340988281.2936.58.camel@lappy> <20120701011538.GD2907@linux.vnet.ibm.com> <50051F32.5040907@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <50051F32.5040907@gmail.com> Sender: owner-linux-mm@kvack.org List-ID: To: Sasha Levin Cc: Andrew Morton , Ingo Molnar , Peter Zijlstra , linux-mm , "linux-kernel@vger.kernel.org" On Tue, Jul 17, 2012 at 10:15:46AM +0200, Sasha Levin wrote: > Hi Paul, > > I've been running with your patch below for a while now, and haven't encountered the issue again. Thank you for the testing, Sasha! I have the full OOM patch queued for 3.7. Thanx, Paul > On 07/01/2012 03:15 AM, Paul E. McKenney wrote: > > On Fri, Jun 29, 2012 at 06:44:41PM +0200, Sasha Levin wrote: > >> Hi all, > >> > >> While fuzzing using trinity on a KVM tools guest with todays linux-next, I've hit the following lockup: > >> > >> [ 362.261729] INFO: task numad/2:27 blocked for more than 120 seconds. > >> [ 362.263974] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > >> [ 362.271684] numad/2 D 0000000000000001 5672 27 2 0x00000000 > >> [ 362.280052] ffff8800294c7c58 0000000000000046 ffff8800294c7c08 ffffffff81163dba > >> [ 362.294477] ffff8800294c6000 ffff8800294c6010 ffff8800294c7fd8 ffff8800294c6000 > >> [ 362.306631] ffff8800294c6010 ffff8800294c7fd8 ffff88000d5c3000 ffff8800294c8000 > >> [ 362.315395] Call Trace: > >> [ 362.318556] [] ? __lock_release+0x1ba/0x1d0 > >> [ 362.325411] [] schedule+0x55/0x60 > >> [ 362.328844] [] rwsem_down_failed_common+0xf5/0x130 > >> [ 362.332501] [] ? put_lock_stats+0xe/0x40 > >> [ 362.334496] [] ? __lock_contended+0x1f5/0x230 > >> [ 362.336723] [] rwsem_down_read_failed+0x15/0x17 > >> [ 362.339297] [] call_rwsem_down_read_failed+0x14/0x30 > >> [ 362.341768] [] ? down_read+0x79/0xa0 > >> [ 362.343669] [] ? lazy_migrate_process+0x22/0x60 > >> [ 362.345616] [] lazy_migrate_process+0x22/0x60 > >> [ 362.347464] [] process_mem_migrate+0x10/0x20 > >> [ 362.349340] [] move_processes+0x190/0x230 > >> [ 362.351398] [] numad_thread+0x7a/0x120 > >> [ 362.353245] [] ? find_busiest_node+0x310/0x310 > >> [ 362.355396] [] kthread+0xb2/0xc0 > >> [ 362.356996] [] kernel_thread_helper+0x4/0x10 > >> [ 362.359253] [] ? retint_restore_args+0x13/0x13 > >> [ 362.361168] [] ? __init_kthread_worker+0x70/0x70 > >> [ 362.363277] [] ? gs_change+0x13/0x13 > >> > >> I've hit sysrq-t to see what might be the cause, and it appears that an OOM was in progress, and was stuck on RCU: > >> > >> [ 578.086230] trinity-child69 D ffff8800277a54c8 3968 6658 6580 0x00000000 > >> [ 578.086230] ffff880022c5f518 0000000000000046 ffff880022c5f4c8 ffff88001b9d6e00 > >> [ 578.086230] ffff880022c5e000 ffff880022c5e010 ffff880022c5ffd8 ffff880022c5e000 > >> [ 578.086230] ffff880022c5e010 ffff880022c5ffd8 ffff880023c08000 ffff880022c33000 > >> [ 578.086230] Call Trace: > >> [ 578.086230] [] schedule+0x55/0x60 > >> [ 578.086230] [] schedule_timeout+0x38/0x2c0 > >> [ 578.086230] [] ? mark_held_locks+0xf6/0x120 > >> [ 578.086230] [] ? __lock_release+0x1ba/0x1d0 > >> [ 578.086230] [] ? _raw_spin_unlock_irq+0x2b/0x80 > >> [ 578.086230] [] wait_for_common+0xff/0x170 > >> [ 578.086230] [] ? try_to_wake_up+0x290/0x290 > >> [ 578.086230] [] wait_for_completion+0x18/0x20 > >> [ 578.086230] [] _rcu_barrier+0x4a7/0x4e0 > > > > Hmmm... Perhaps a blocking operation is not appropriate here. I have > > substituted a nonblocking approach, which is at -rcu (thus soon -next) > > at 1ee4c09d (Provide OOM handler to motivate lazy RCU callbacks). > > Patch below. > > > > Thanx, Paul > > > > ------------------------------------------------------------------------ > > > > rcu: Provide OOM handler to motivate lazy RCU callbacks > > > > In kernels built with CONFIG_RCU_FAST_NO_HZ=y, CPUs can accumulate a > > large number of lazy callbacks, which as the name implies will be slow > > to be invoked. This can be a problem on small-memory systems, where the > > default 6-second sleep for CPUs having only lazy RCU callbacks could well > > be fatal. This commit therefore installs an OOM hander that ensures that > > every CPU with non-lazy callbacks has at least one non-lazy callback, > > in turn ensuring timely advancement for these callbacks. > > > > Signed-off-by: Paul E. McKenney > > Signed-off-by: Paul E. McKenney > > > > diff --git a/kernel/rcutree.h b/kernel/rcutree.h > > index 4b47fbe..dab279f 100644 > > --- a/kernel/rcutree.h > > +++ b/kernel/rcutree.h > > @@ -314,8 +314,11 @@ struct rcu_data { > > unsigned long n_rp_need_fqs; > > unsigned long n_rp_need_nothing; > > > > - /* 6) _rcu_barrier() callback. */ > > + /* 6) _rcu_barrier() and OOM callbacks. */ > > struct rcu_head barrier_head; > > +#ifdef CONFIG_RCU_FAST_NO_HZ > > + struct rcu_head oom_head; > > +#endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */ > > > > int cpu; > > struct rcu_state *rsp; > > diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h > > index 81e53eb..1908847 100644 > > --- a/kernel/rcutree_plugin.h > > +++ b/kernel/rcutree_plugin.h > > @@ -25,6 +25,7 @@ > > */ > > > > #include > > +#include > > > > #define RCU_KTHREAD_PRIO 1 > > > > @@ -2128,6 +2129,90 @@ static void rcu_idle_count_callbacks_posted(void) > > __this_cpu_add(rcu_dynticks.nonlazy_posted, 1); > > } > > > > +/* > > + * Data for flushing lazy RCU callbacks at OOM time. > > + */ > > +static atomic_t oom_callback_count; > > +static DECLARE_WAIT_QUEUE_HEAD(oom_callback_wq); > > + > > +/* > > + * RCU OOM callback -- decrement the outstanding count and deliver the > > + * wake-up if we are the last one. > > + */ > > +static void rcu_oom_callback(struct rcu_head *rhp) > > +{ > > + if (atomic_dec_and_test(&oom_callback_count)) > > + wake_up(&oom_callback_wq); > > +} > > + > > +/* > > + * Post an rcu_oom_notify callback on the current CPU if it has at > > + * least one lazy callback. This will unnecessarily post callbacks > > + * to CPUs that already have a non-lazy callback at the end of their > > + * callback list, but this is an infrequent operation, so accept some > > + * extra overhead to keep things simple. > > + */ > > +static void rcu_oom_notify_cpu(void *flavor) > > +{ > > + struct rcu_state *rsp = flavor; > > + struct rcu_data *rdp = __this_cpu_ptr(rsp->rda); > > + > > + if (rdp->qlen_lazy != 0) { > > + atomic_inc(&oom_callback_count); > > + rsp->call(&rdp->oom_head, rcu_oom_callback); > > + } > > +} > > + > > +/* > > + * If low on memory, ensure that each CPU has a non-lazy callback. > > + * This will wake up CPUs that have only lazy callbacks, in turn > > + * ensuring that they free up the corresponding memory in a timely manner. > > + */ > > +static int rcu_oom_notify(struct notifier_block *self, > > + unsigned long notused, void *nfreed) > > +{ > > + int cpu; > > + > > + /* Wait for callbacks from earlier instance to complete. */ > > + wait_event(oom_callback_wq, atomic_read(&oom_callback_count) == 0); > > + > > + /* > > + * Prevent premature wakeup: ensure that all increments happen > > + * before there is a chance of the counter reaching zero. > > + */ > > + atomic_set(&oom_callback_count, 1); > > + > > + get_online_cpus(); > > + for_each_online_cpu(cpu) { > > +#ifdef CONFIG_PREEMPT_RCU > > + smp_call_function_single(cpu, rcu_oom_notify_cpu, > > + &rcu_preempt_state, 1); > > +#endif /* #ifdef CONFIG_PREEMPT_RCU */ > > + smp_call_function_single(cpu, rcu_oom_notify_cpu, > > + &rcu_bh_state, 1); > > + smp_call_function_single(cpu, rcu_oom_notify_cpu, > > + &rcu_sched_state, 1); > > + } > > + put_online_cpus(); > > + > > + /* Unconditionally decrement: no need to wake ourselves up. */ > > + atomic_dec(&oom_callback_count); > > + > > + *(unsigned long *)nfreed = 1; > > + return NOTIFY_OK; > > +} > > + > > +static struct notifier_block rcu_oom_nb = { > > + .notifier_call = rcu_oom_notify > > +}; > > + > > +static int __init rcu_register_oom_notifier(void) > > +{ > > + register_oom_notifier(&rcu_oom_nb); > > + return 0; > > +} > > +early_initcall(rcu_register_oom_notifier); > > + > > #endif /* #else #if !defined(CONFIG_RCU_FAST_NO_HZ) */ > > > > #ifdef CONFIG_RCU_CPU_STALL_INFO > > > > -- > > 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/ > > > > -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org