[clug] What causes an unkillable process?

[Duncan Roe]

I have been finding recently that sometimes a process will be stopped (e.g.
because of SIGTT(IN|OU) and will not be killed by kill -9 until you first kill
it with SIGCONT. I don't remember if ps showed the process in state D (as here)
or T (stopped) - hope this is some help anyway.

Cheers ... Duncan.

On Sun, Nov 23, 2008 at 01:20:43PM +0100, Martijn van Oosterhout wrote:
> On Sun, Nov 23, 2008 at 04:11:27AM -0800, Jason Stokes wrote:
> > The following process:
> >
> > > 10685 jason     20   0  143m 8500 8492 D 64.9  1.6 937:50.21 totem
> >
> > Won't die with kill -9, not even as root.  Sure, I can just reboot, but I'm wondering what on earth causes such a thing.
>
> D state is unkillable. It means it's stuck in the kernel somewhere. I
> thought there was a way to find out exactly where, but I can't remember
> how. One of the sysrq commands?
>
> The most common causes are disk access, reading/wrting from a stuck NFS
> mount, broken hard disk, etc.
>
> Once it gets out of the kernel the process will be killed before it
> reaches userspace. So it's just a matter of finding out where it's
> stuck.
>
> Oh yes, kernel bugs can cause this too :)
>
> Have a niec day,
> --
> Martijn van Oosterhout   <kleptog at svana.org>   http://svana.org/kleptog/
> > Please line up in a tree and maintain the heap invariant while
> > boarding. Thank you for flying nlogn airlines.



> --
> linux mailing list
> linux at lists.samba.org
> https://lists.samba.org/mailman/listinfo/linux


--
 Please avoid sending me Word or PowerPoint attachments.
 See http://www.gnu.org/philosophy/no-word-attachments.html