On Sun, Feb 22, 2009 at 6:36 PM, James Harper
<james.harper@xxxxxxxxxxxxxxxx> wrote:
> There is a bug in 3.3.1 that comes about when you have more than one CPU
> and you restrict Dom0 to one CPU by putting "(dom0-cpus 1)" in your
> config file. The symptoms are that the domain won't shut down until you
> do an 'xm start' again, and even then bits of it hang around in the
> hypervisor.
>
> I'm not sure if that bug existed in 3.2 though.
Okay. Now that's just plain weird.
I use RHEL 5.3 x86_64, Xen 3.3.1 from Gitco, (dom0-cpus 1),
(dom0-min-mem 256), and dom0_mem=512M dom0_vcpus_pin. I've been using
both PV Linux and HVM Windows domUs on it and they can shutdown and
reboot just fine.
Any ideas where this bug is located? Is it the hypervisor, kernel, or
userland tools? If what you're saying is true (in that the bug affects
3.3.1 as well) then my only guess is that somehow my setup is immune
to this bug. I'd really like to know how.
Regards,
Fajar
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
Home