linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Warning: NO_HZ FULL will not work with unstable sched clock
@ 2013-11-09  8:32 Alexander Beregalov
  2013-11-11 17:20 ` Frederic Weisbecker
  0 siblings, 1 reply; 2+ messages in thread
From: Alexander Beregalov @ 2013-11-09  8:32 UTC (permalink / raw)
  To: Linux Kernel Mailing List
  Cc: Frederic Weisbecker, Peter Zijlstra, Ingo Molnar

Hi,
It first appeared some time after 3.9 or 3.10 and still persist.
I guess bisect is not required?

WARNING: CPU: 1 PID: 6 at kernel/time/tick-sched.c:187
can_stop_full_tick+0x8d/0x9a()
NO_HZ FULL will not work with unstable sched clock
Modules linked in:

CPU: 1 PID: 6 Comm: kworker/u8:0 Not tainted 3.12.0-01714-ga1212d2 #1
Hardware name:                  /D525MW, BIOS
MWPNT10N.86A.0084.2011.0726.1300 07/26/2011
 00000000000000bb ffff8800bea83e88 ffffffff813ba3b0 0000000000000050
 ffff8800bea83ed8 ffff8800bea83ec8 ffffffff81031441 ffff8800bb07a170
 ffffffff810687db 0000000000000000 0000000000000001 ffff8800bea91ac0
Call Trace:
 <IRQ>  [<ffffffff813ba3b0>] dump_stack+0x46/0x58
 [<ffffffff81031441>] warn_slowpath_common+0x78/0x92
 [<ffffffff810687db>] ? can_stop_full_tick+0x8d/0x9a
 [<ffffffff810314f1>] warn_slowpath_fmt+0x42/0x44
 [<ffffffff81052353>] ? account_system_time+0x114/0x15c
 [<ffffffff810687db>] can_stop_full_tick+0x8d/0x9a
 [<ffffffff810689ef>] tick_nohz_irq_exit+0x77/0x93
 [<ffffffff81034c28>] irq_exit+0xd3/0xdc
 [<ffffffff81021b7c>] smp_apic_timer_interrupt+0x40/0x4d
 [<ffffffff813c474a>] apic_timer_interrupt+0x6a/0x70
 <EOI>  [<ffffffff8104da85>] ? finish_task_switch+0x6f/0xd1
 [<ffffffff8104da5a>] ? finish_task_switch+0x44/0xd1
 [<ffffffff813bd0ae>] __schedule+0x75a/0x7f3
 [<ffffffff813be291>] ? _raw_spin_unlock_irqrestore+0x9/0xb
 [<ffffffff8104d704>] ? complete+0x43/0x4b
 [<ffffffff8104068c>] ? umh_complete+0x17/0x20
 [<ffffffff81040710>] ? __call_usermodehelper+0x7b/0x80
 [<ffffffff81042b96>] ? process_one_work+0x2f4/0x304
 [<ffffffff813bd3f6>] schedule+0x5f/0x61
 [<ffffffff810433ba>] worker_thread+0x2a8/0x2ad
 [<ffffffff81043112>] ? manage_workers.isra.31+0x1b9/0x1b9
 [<ffffffff81047b86>] kthread+0xb6/0xbe
 [<ffffffff81047ad0>] ? __kthread_parkme+0x68/0x68
 [<ffffffff813c39fc>] ret_from_fork+0x7c/0xb0
 [<ffffffff81047ad0>] ? __kthread_parkme+0x68/0x68

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: Warning: NO_HZ FULL will not work with unstable sched clock
  2013-11-09  8:32 Warning: NO_HZ FULL will not work with unstable sched clock Alexander Beregalov
@ 2013-11-11 17:20 ` Frederic Weisbecker
  0 siblings, 0 replies; 2+ messages in thread
From: Frederic Weisbecker @ 2013-11-11 17:20 UTC (permalink / raw)
  To: Alexander Beregalov
  Cc: Linux Kernel Mailing List, Peter Zijlstra, Ingo Molnar

On Sat, Nov 09, 2013 at 12:32:06PM +0400, Alexander Beregalov wrote:
> Hi,
> It first appeared some time after 3.9 or 3.10 and still persist.
> I guess bisect is not required?
> 
> WARNING: CPU: 1 PID: 6 at kernel/time/tick-sched.c:187
> can_stop_full_tick+0x8d/0x9a()
> NO_HZ FULL will not work with unstable sched clock

Right, it's not a bug, just a warning we've put to make sure that users of
full dynticks are aware of limitations when the sched clock is unstable.

That's something we want to fix in the long run as it seems that quite
a lot of people have unstable tsc.

Thanks.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2013-11-11 17:20 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-11-09  8:32 Warning: NO_HZ FULL will not work with unstable sched clock Alexander Beregalov
2013-11-11 17:20 ` 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).