* watchdog kicked in while shrinking inactive list.
@ 2014-09-25 14:14 Dave Jones
2014-09-25 14:35 ` hrtimer deadlock caused by nohz_full Thomas Gleixner
0 siblings, 1 reply; 3+ messages in thread
From: Dave Jones @ 2014-09-25 14:14 UTC (permalink / raw)
To: linux-mm; +Cc: Thomas Gleixner, fweisbec
Got this on a box that had been fuzzing for 12 hours or so.
There's also some timer stuff going on htere, so cc'ing the usual suspects.
Kernel panic - not syncing: Watchdog detected hard LOCKUP on cpu 3
CPU: 3 PID: 16759 Comm: trinity-c25 Not tainted 3.17.0-rc6+ #59
0000000000000000 00000000ff743b71 ffff880244e06c00 ffffffffa1818ac4
ffffffffa1c633c0 ffff880244e06c80 ffffffffa18130a3 ffff880200000010
ffff880244e06c90 ffff880244e06c30 00000000ff743b71 ffffffffa117ead5
Call Trace:
<NMI> [<ffffffffa1818ac4>] dump_stack+0x4e/0x7a
[<ffffffffa18130a3>] panic+0xd7/0x20a
[<ffffffffa117ead5>] ? __perf_event_header__init_id+0xe5/0xf0
[<ffffffffa1146220>] ? restart_watchdog_hrtimer+0x50/0x50
[<ffffffffa1146338>] watchdog_overflow_callback+0x118/0x120
[<ffffffffa118792c>] __perf_event_overflow+0xac/0x350
[<ffffffffa1019ece>] ? x86_perf_event_set_period+0xde/0x150
[<ffffffffa1188844>] perf_event_overflow+0x14/0x20
[<ffffffffa101fc26>] intel_pmu_handle_irq+0x206/0x410
[<ffffffffa1018d2b>] perf_event_nmi_handler+0x2b/0x50
[<ffffffffa1007752>] nmi_handle+0xd2/0x3b0
[<ffffffffa1007685>] ? nmi_handle+0x5/0x3b0
[<ffffffffa10d103d>] ? lock_release+0x1d/0x300
[<ffffffffa1007c62>] default_do_nmi+0x72/0x1c0
[<ffffffffa1007e68>] do_nmi+0xb8/0xf0
[<ffffffffa18268aa>] end_repeat_nmi+0x1e/0x2e
[<ffffffffa10f95f8>] ? hrtimer_try_to_cancel+0x58/0x1f0
[<ffffffffa10d103d>] ? lock_release+0x1d/0x300
[<ffffffffa10d103d>] ? lock_release+0x1d/0x300
[<ffffffffa10f95f8>] ? hrtimer_try_to_cancel+0x58/0x1f0
[<ffffffffa10d103d>] ? lock_release+0x1d/0x300
<<EOE>> <IRQ> [<ffffffffa1823ac4>] _raw_spin_unlock_irqrestore+0x24/0x70
[<ffffffffa10f95f8>] hrtimer_try_to_cancel+0x58/0x1f0
[<ffffffffa10f97aa>] hrtimer_cancel+0x1a/0x30
[<ffffffffa110a0e7>] tick_nohz_restart+0x17/0x90
[<ffffffffa110af38>] __tick_nohz_full_check+0xc8/0xe0
[<ffffffffa110af5e>] nohz_full_kick_work_func+0xe/0x10
[<ffffffffa117c9bf>] irq_work_run_list+0x4f/0x70
[<ffffffffa117ca0a>] irq_work_run+0x2a/0x60
[<ffffffffa10f82eb>] update_process_times+0x5b/0x70
[<ffffffffa1109dc5>] tick_sched_handle.isra.21+0x25/0x60
[<ffffffffa110a0b1>] tick_sched_timer+0x41/0x60
[<ffffffffa10f8c71>] __run_hrtimer+0x81/0x480
[<ffffffffa110a070>] ? tick_sched_do_timer+0x90/0x90
[<ffffffffa10f9b27>] hrtimer_interrupt+0x107/0x260
[<ffffffffa10331a4>] local_apic_timer_interrupt+0x34/0x60
[<ffffffffa182734f>] smp_apic_timer_interrupt+0x3f/0x60
[<ffffffffa182576f>] apic_timer_interrupt+0x6f/0x80
<EOI> [<ffffffffa1823b42>] ? _raw_spin_unlock_irq+0x32/0x60
[<ffffffffa11a795b>] shrink_inactive_list+0x1cb/0x620
[<ffffffffa11a8663>] shrink_lruvec+0x563/0x6b0
[<ffffffffa119720f>] ? zone_watermark_ok+0x1f/0x30
[<ffffffffa11a87fe>] shrink_zone+0x4e/0x130
[<ffffffffa11a905f>] shrink_zones.constprop.63+0x20f/0x410
[<ffffffffa11a93bb>] try_to_free_pages+0x15b/0x3e0
[<ffffffffa119bacb>] __alloc_pages_nodemask+0x7eb/0xc20
[<ffffffffa11e3756>] alloc_pages_current+0x106/0x1e0
[<ffffffffa11ec905>] ? new_slab+0x2b5/0x390
[<ffffffffa11ec905>] new_slab+0x2b5/0x390
[<ffffffffa1817152>] __slab_alloc+0x348/0x56f
[<ffffffffa11f2116>] ? kmem_cache_alloc+0x256/0x330
[<ffffffffa16b31dd>] ? sock_alloc_inode+0x1d/0xc0
[<ffffffffa10cb687>] ? __lock_is_held+0x57/0x80
[<ffffffffa16b31dd>] ? sock_alloc_inode+0x1d/0xc0
[<ffffffffa11f2116>] kmem_cache_alloc+0x256/0x330
[<ffffffffa16b31dd>] sock_alloc_inode+0x1d/0xc0
[<ffffffffa1223c4d>] alloc_inode+0x1d/0xa0
[<ffffffffa1225df1>] new_inode_pseudo+0x11/0x60
[<ffffffffa16b2dac>] sock_alloc+0x1c/0x90
[<ffffffffa16b3c4d>] sock_create_lite+0x4d/0x90
[<ffffffffa1709d71>] __netlink_kernel_create+0x71/0x280
[<ffffffffa16ed337>] diag_net_init+0x47/0x80
[<ffffffffa16ed2b0>] ? diag_net_exit+0x30/0x30
[<ffffffffa16c85f1>] ops_init+0x41/0x190
[<ffffffffa16c87c3>] setup_net+0x83/0x140
[<ffffffffa16c8d7e>] copy_net_ns+0x7e/0x130
[<ffffffffa109f279>] create_new_namespaces+0xf9/0x190
[<ffffffffa109f5fa>] unshare_nsproxy_namespaces+0x5a/0xc0
[<ffffffffa1077883>] SyS_unshare+0x173/0x330
[<ffffffffa1824a24>] tracesys+0xdd/0xe2
--
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: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: hrtimer deadlock caused by nohz_full
2014-09-25 14:14 watchdog kicked in while shrinking inactive list Dave Jones
@ 2014-09-25 14:35 ` Thomas Gleixner
2014-09-25 23:46 ` Frederic Weisbecker
0 siblings, 1 reply; 3+ messages in thread
From: Thomas Gleixner @ 2014-09-25 14:35 UTC (permalink / raw)
To: Dave Jones; +Cc: linux-mm, Frederic Weisbecker, Peter Zijlstra, LKML
On Thu, 25 Sep 2014, Dave Jones wrote:
> Got this on a box that had been fuzzing for 12 hours or so.
> There's also some timer stuff going on htere, so cc'ing the usual suspects.
And it's a hrtimer lockup
> [<ffffffffa10f95f8>] ? hrtimer_try_to_cancel+0x58/0x1f0
> [<ffffffffa10d103d>] ? lock_release+0x1d/0x300
> [<ffffffffa10d103d>] ? lock_release+0x1d/0x300
> [<ffffffffa10f95f8>] ? hrtimer_try_to_cancel+0x58/0x1f0
> [<ffffffffa10d103d>] ? lock_release+0x1d/0x300
> <<EOE>> <IRQ> [<ffffffffa1823ac4>] _raw_spin_unlock_irqrestore+0x24/0x70
> [<ffffffffa10f95f8>] hrtimer_try_to_cancel+0x58/0x1f0
> [<ffffffffa10f97aa>] hrtimer_cancel+0x1a/0x30
> [<ffffffffa110a0e7>] tick_nohz_restart+0x17/0x90
> [<ffffffffa110af38>] __tick_nohz_full_check+0xc8/0xe0
> [<ffffffffa110af5e>] nohz_full_kick_work_func+0xe/0x10
> [<ffffffffa117c9bf>] irq_work_run_list+0x4f/0x70
> [<ffffffffa117ca0a>] irq_work_run+0x2a/0x60
> [<ffffffffa10f82eb>] update_process_times+0x5b/0x70
> [<ffffffffa1109dc5>] tick_sched_handle.isra.21+0x25/0x60
> [<ffffffffa110a0b1>] tick_sched_timer+0x41/0x60
> [<ffffffffa10f8c71>] __run_hrtimer+0x81/0x480
> [<ffffffffa110a070>] ? tick_sched_do_timer+0x90/0x90
> [<ffffffffa10f9b27>] hrtimer_interrupt+0x107/0x260
> [<ffffffffa10331a4>] local_apic_timer_interrupt+0x34/0x60
> [<ffffffffa182734f>] smp_apic_timer_interrupt+0x3f/0x60
> [<ffffffffa182576f>] apic_timer_interrupt+0x6f/0x80
hrtimer_interrupt
tick_sched_timer
tick_sched_handle
update_process_times
irq_work_run
irq_work_run_list
nohz_full_kick_work_func
__tick_nohz_full_check
tick_nohz_restart
hrtimer_cancel
And that hrtimer_cancel is:
static void tick_nohz_restart(struct tick_sched *ts, ktime_t now)
{
hrtimer_cancel(&ts->sched_timer);
Now, that's really bad because we are in the timer callback of
ts->sched_timer. So hrtimer_cancel will loop forever waiting for the
callback to complete.
Frederic !?!?
Thanks,
tglx
--
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: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: hrtimer deadlock caused by nohz_full
2014-09-25 14:35 ` hrtimer deadlock caused by nohz_full Thomas Gleixner
@ 2014-09-25 23:46 ` Frederic Weisbecker
0 siblings, 0 replies; 3+ messages in thread
From: Frederic Weisbecker @ 2014-09-25 23:46 UTC (permalink / raw)
To: Thomas Gleixner; +Cc: Dave Jones, linux-mm, Peter Zijlstra, LKML
On Thu, Sep 25, 2014 at 04:35:49PM +0200, Thomas Gleixner wrote:
> On Thu, 25 Sep 2014, Dave Jones wrote:
>
> > Got this on a box that had been fuzzing for 12 hours or so.
> > There's also some timer stuff going on htere, so cc'ing the usual suspects.
>
> And it's a hrtimer lockup
>
> > [<ffffffffa10f95f8>] ? hrtimer_try_to_cancel+0x58/0x1f0
> > [<ffffffffa10d103d>] ? lock_release+0x1d/0x300
> > [<ffffffffa10d103d>] ? lock_release+0x1d/0x300
> > [<ffffffffa10f95f8>] ? hrtimer_try_to_cancel+0x58/0x1f0
> > [<ffffffffa10d103d>] ? lock_release+0x1d/0x300
> > <<EOE>> <IRQ> [<ffffffffa1823ac4>] _raw_spin_unlock_irqrestore+0x24/0x70
> > [<ffffffffa10f95f8>] hrtimer_try_to_cancel+0x58/0x1f0
> > [<ffffffffa10f97aa>] hrtimer_cancel+0x1a/0x30
> > [<ffffffffa110a0e7>] tick_nohz_restart+0x17/0x90
> > [<ffffffffa110af38>] __tick_nohz_full_check+0xc8/0xe0
> > [<ffffffffa110af5e>] nohz_full_kick_work_func+0xe/0x10
> > [<ffffffffa117c9bf>] irq_work_run_list+0x4f/0x70
> > [<ffffffffa117ca0a>] irq_work_run+0x2a/0x60
> > [<ffffffffa10f82eb>] update_process_times+0x5b/0x70
> > [<ffffffffa1109dc5>] tick_sched_handle.isra.21+0x25/0x60
> > [<ffffffffa110a0b1>] tick_sched_timer+0x41/0x60
> > [<ffffffffa10f8c71>] __run_hrtimer+0x81/0x480
> > [<ffffffffa110a070>] ? tick_sched_do_timer+0x90/0x90
> > [<ffffffffa10f9b27>] hrtimer_interrupt+0x107/0x260
> > [<ffffffffa10331a4>] local_apic_timer_interrupt+0x34/0x60
> > [<ffffffffa182734f>] smp_apic_timer_interrupt+0x3f/0x60
> > [<ffffffffa182576f>] apic_timer_interrupt+0x6f/0x80
>
> hrtimer_interrupt
> tick_sched_timer
> tick_sched_handle
> update_process_times
> irq_work_run
> irq_work_run_list
> nohz_full_kick_work_func
> __tick_nohz_full_check
> tick_nohz_restart
> hrtimer_cancel
>
> And that hrtimer_cancel is:
>
> static void tick_nohz_restart(struct tick_sched *ts, ktime_t now)
> {
> hrtimer_cancel(&ts->sched_timer);
>
> Now, that's really bad because we are in the timer callback of
> ts->sched_timer. So hrtimer_cancel will loop forever waiting for the
> callback to complete.
>
> Frederic !?!?
Right, this patchset fixes it: "[PATCH 0/8] nohz: Fix nohz kick irq work on tick v3"
I was about to make the pull request, the branch is acked by peterz.
Would you like to pull it? It's all merge window material.
git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks.git
nohz/fixes-v3
Thanks.
--
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: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2014-09-25 23:47 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-09-25 14:14 watchdog kicked in while shrinking inactive list Dave Jones
2014-09-25 14:35 ` hrtimer deadlock caused by nohz_full Thomas Gleixner
2014-09-25 23:46 ` Frederic Weisbecker
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).