FWIW, I can reliably get smbd (the samba daemon) to ignore SIGKILL. I'm
running 2.2.12-20 with samba 2.0.7. Here are the steps for reproducing
this:
1) define a share "tmp" that points to /tmp on the machinen
2) smbmount //host/tmp /tmp/mnt
3) ls /tmp/mnt/mnt
At this point, the ls hangs, and smbd can't be killed w/ kill -9.
Killing the ls cleans this up, and allows smbd to exit, so I presume the
issue here is that the smbd process is providing a filesystem interface
to the kernel, and the ongoing "ls" makes that filesystem busy. (I
believe the fact that the ls hangs is a samba bug, but that is another
story.)
Perhaps something similar is happening the cases others have observed?
Is this appropriate behavior, or should kill -9 still be able to wipe
out the sbmd process?
-Rob von Behren
2.2.12
Chris Wedgwood wrote:
>
> On Tue, May 23, 2000 at 12:29:03PM -0700, Linda Walsh wrote:
>
> Have seen the same -- httpd hung when coming up -- I guess
> something was misconfigured. It was on a lab machine, so didn't
> care so much, but kill -9 on the <pid> had no effect. I think it
> was waiting on WCHAN 'down'.
>
> Weird.
>
> On a server that has gone mad, kill -9 may not appear to work
> instantly if there are many pages swapped out.
>
> If there isn't swapping and a process is stuck in down, then
> something has deadlocked and needs to be fixed.
>
> --cw
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.rutgers.edu
> Please read the FAQ at http://www.tux.org/lkml/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue May 23 2000 - 21:00:25 EST