[uml-devel] process creep on host FC2, UML 2.4.27

[uml-devel] process creep on host FC2, UML 2.4.27

[Daniel Nurmi]

Greetings,

We're currently running a course using UML to do some kernel
developement but have run into a confusing snag.  Our host machines
are pretty much stock Fedora (FC2) boxes running kernel
2.6.9-1.6_FC2smp and our UML kernels are 2.4.27 (clean kernel source,
2.4.27uml patch).  

When we start a process under UML, a corresponding kernel thread is
started on the host (which is normal and to be expected) but when the
process under UML exits the corresponding host thread does not exit
(which is not expected).  After a short time, of course, the host is
swamped with thousands of processes, all of which exit cleanly when
the UML is halted or killed.  

I cannot replicate this behaviour on a clean 2.6.7, 2.6.7 SMP, or
2.6.8 SMP kernel (on other systems) and our admin tells us that this
problem did not exist on these same machines running FC1.  

If anyone has any input about a cause or fix to this issue, I would be
greatly appreciative!

Thanks,
-Dan Nurmi


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] process creep on host FC2, UML 2.4.27

[Blaisorblade]

On Wednesday 12 January 2005 20:08, Daniel Nurmi wrote:
> Greetings,
>
> We're currently running a course using UML to do some kernel
> developement but have run into a confusing snag.  Our host machines
> are pretty much stock Fedora (FC2) boxes running kernel
> 2.6.9-1.6_FC2smp and our UML kernels are 2.4.27 (clean kernel source,
> 2.4.27uml patch).
>
> When we start a process under UML, a corresponding kernel thread is
> started on the host (which is normal and to be expected) but when the
> process under UML exits the corresponding host thread does not exit
> (which is not expected).  After a short time, of course, the host is
> swamped with thousands of processes, all of which exit cleanly when
> the UML is halted or killed.
>
> I cannot replicate this behaviour on a clean 2.6.7, 2.6.7 SMP, or
> 2.6.8 SMP kernel (on other systems) and our admin tells us that this
> problem did not exist on these same machines running FC1.
>
> If anyone has any input about a cause or fix to this issue, I would be
> greatly appreciative!

This is a FAQ about 2.6.9. There are some patches out there (search for 
fix-kill by C. Aker for the 2.4 backport of the fix).

Bye
-- 
Paolo Giarrusso, aka Blaisorblade
Linux registered user n. 292729
http://www.user-mode-linux.org/~blaisorblade


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel