Re: Dom0 Hang for large VCPU counts > PCPU

["Steven Hand" <steven.hand@cl.cam.ac.uk>]

With FC6 (or any 2.6.18 kernel) there's an issue using the tsc clocksource 
(we're working
on a fix); may not be related but could be worth booting with 
"clocksource=acpi_pm".

cheers,

S.

----- Original Message ----- 
From: "Zulauf, John" <john.zulauf@intel.com>
To: "Keir Fraser" <Keir.Fraser@cl.cam.ac.uk>; 
<xen-devel@lists.xensource.com>
Sent: Wednesday, January 24, 2007 7:28 PM
Subject: RE: [Xen-devel] Dom0 Hang for large VCPU counts > PCPU



________________________________________
On January 24, 2007 11:11 AM Keir Fraser [mailto:Keir.Fraser@cl.cam.ac.uk] 
wrote
> On 24/1/07 6:56 pm, "Zulauf, John" <john.zulauf@intel.com> wrote:
>
> > Using xen-3.0.4-testing with an HVM booting the FC6 DVD on a Core2 Duo 
> > (i.e. 2 PCPU) on a 965 chipset
> >
> > VCPU=17 - works
> > VCPU=20 - works (takes a very long time)
> > VCPU=24 - lockup (whole machine, yes I mean Dom0)
> > VCPU=31 - same
> > VCPU=32 - same
>
> Have you worked out where it gets stuck? Do many-VCPU domU guests fare 
> better?

It IS dying for a domU guests.  Do you perhaps mean PV guests?  No we 
haven't tried them.  I don't have any at the moment to test.

Clarifying the initial report:

Host is 2 core (Core 2 Duo/965 chipset) box with VPCU=PCPU for Dom0.
Issue occurs when attempting to start an HVM DomU with VCPU >= 24 (though we 
haven't checked 21-23 yet).
The DomU never(apparently) starts, and Dom0 hard hangs. (i.e. not even the 
numlock key works)

Given that Dom0 hangs hard when this occurs, we don't have any ideas as to 
the exact cause of death.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel