public inbox for linux-ia64@vger.kernel.org
 help / color / mirror / Atom feed
* [Linux-ia64] hanging process cannot be killed
@ 2001-05-16 16:55 Erich Focht
  2001-05-16 18:49 ` David Mosberger
  0 siblings, 1 reply; 2+ messages in thread
From: Erich Focht @ 2001-05-16 16:55 UTC (permalink / raw)
  To: linux-ia64

Hello,

how can it happen that a process isn't killable with kill -9 any more? I
have encountered this a few times and from within kdb two processes seem
to be in "run" state on the same CPU! The one marked as "current" shows
the following traceback:
   __ia64_save_fpu+0x20
   schedule+0xa40
   invoke_schedule+0x20
   ia64_leave_kernel+0x30

This process sits on the CPU, eats up cycles and the scheduler doesn't
seem to touch it any more. I only can get rid of it by rebooting.

Another traceback of a similar situation (with the same executable) was:
   __ia64_save_fpu+0x280
   schedule+0xa40
   ia64_do_signal+0x670
   handle_signal_delivery+0x30
   ia64_leave_kernel+0x30

I've seen this with 2.4.0 and 2.4.4 kernels, with B1 stepping CPUs. With
2.4.4 in single dispersal mode the problem doesn't seem to occur...

Thanks in advance for any help, ideas, comments...

Best regards,
Erich

---
Erich Focht                                    <efocht@ess.nec.de>
NEC European Supercomputer Systems, European HPC Technology Center
Hessbruehlstr. 21B, 70565 Stuttgart, Germany
phone: +49-711-78055-15                    fax  : +49-711-78055-25




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

* Re: [Linux-ia64] hanging process cannot be killed
  2001-05-16 16:55 [Linux-ia64] hanging process cannot be killed Erich Focht
@ 2001-05-16 18:49 ` David Mosberger
  0 siblings, 0 replies; 2+ messages in thread
From: David Mosberger @ 2001-05-16 18:49 UTC (permalink / raw)
  To: linux-ia64

>>>>> On Wed, 16 May 2001 18:55:20 +0200 (MEST), Erich Focht <focht@ess.nec.de> said:

  Erich> Hello, how can it happen that a process isn't killable with
  Erich> kill -9 any more?

Can happen easily if it's doing a non-interruptible sleep.

  Erich> I have encountered this a few times and from within kdb two
  Erich> processes seem to be in "run" state on the same CPU! The one
  Erich> marked as "current" shows the following traceback:
  Erich> __ia64_save_fpu+0x20 schedule+0xa40 invoke_schedule+0x20
  Erich> ia64_leave_kernel+0x30

  Erich> [snip...]

  Erich> I've seen this with 2.4.0 and 2.4.4 kernels, with B1 stepping
  Erich> CPUs. With 2.4.4 in single dispersal mode the problem doesn't
  Erich> seem to occur...

I'd recommend to upgrade to B3 or C0.  Earlier steppings had some
issues.  Most of them can be worked around, but it's not very pretty
and I'd not recommend spending a lot of time on it.

	--david


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

end of thread, other threads:[~2001-05-16 18:49 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2001-05-16 16:55 [Linux-ia64] hanging process cannot be killed Erich Focht
2001-05-16 18:49 ` David Mosberger

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox