[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [Xen-users] RE: Windows assigned 4GB RAM BSOD with GPLPV (0.9.10)


  • To: <terii@xxxxxxxxxxxxxxx>, <xen-users@xxxxxxxxxxxxxxxxxxx>
  • From: "James Harper" <james.harper@xxxxxxxxxxxxxxxx>
  • Date: Fri, 10 Oct 2008 10:26:47 +1100
  • Cc:
  • Delivery-date: Thu, 09 Oct 2008 16:27:27 -0700
  • List-id: Xen user discussion <xen-users.lists.xensource.com>
  • Thread-index: AckqRVDuQfg/QV2sS9Kh4NCwQE4+pQAIFoWQ
  • Thread-topic: [Xen-users] RE: Windows assigned 4GB RAM BSOD with GPLPV (0.9.10)

> On Thu, 9 Oct 2008 09:11:58 +1100
> "James Harper" <james.harper@xxxxxxxxxxxxxxxx> wrote:
> 
> 
> > Can you please test a later version of the GPLPV drivers?
0.9.11-pre18
> > is the latest. I know I fixed a 4G problem at some point, but I
> > thought that was earlier than 0.9.10. Also, can you give me the BSoD
> > numbers?
> >
> 
> 0.9.11-pre18 seems to work great!
> 
> STOP: 0x0000007B (0xF789EA94,Oxc0000034,0x00000000,0x00000000)
> 
> Screenshot here:
> http://terii.kbcowboy.net/images/misc/xen_4gb-gplpv-bsod.jpg
> 

Yes, this appears to be the problem I was seeing. At some point (>3.5G?)
windows switches on PAE to get the available memory over 4G, and so even
though it is a 32bit system it behaves like a 64bit system in terms of
some of the addressing requirements, and the vbd driver wasn't reporting
to windows correctly that it supported this, so the vbd driver didn't
load and windows was left without a boot device (0x0000007B is
inaccessible boot device).

> > I have tested with 4CPU's and have not noticed a problem before.
> >
> 
> To be more specific... 4CPUs+4GB RAM causing reboot/shutdown to
freeze.
> This happens even with non-GPLPV enabled guests. So this is not GPLPV
> related.
> 
> Unfortunately I don't have enough time to coordinate with Windows team
> to test out stability of "pre" GPLPV drivers. I have to push out those
> guests as it is for now. I will continue to test out GPL PV drivers as
> new ones is released on my dev servers. Thanks again.

Don't read too much into the fact that the version of the drivers is
tagged -preXX. They are much more stable than the 0.9.10 version, the
only reason I was holding off on a 0.9.11 proper is that I had reports
of a few problems that I wanted to have fixed before 0.9.11.
Unfortunately I have never been able to reproduce these problems and I
think the reporting parties and I have run out of steam in terms of
trying to come up with a solution and/or a way to reproduce the problem.

I will roll the driver version over to 0.9.11 today if I get time, but
it will be the same as 0.9.11-pre18.

James


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


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.