This turned out to be a xen install problem. The default install
location for the build was /usr/local but my system was using /usr. The
root cause was hvmloader created by the new build was not actually the
one being used.
On 2017-02-27 11:59, Konrad Rzeszutek Wilk wrote:
On Wed, Feb 22, 2017 at 03:50:28PM -0500, Jason Dickens wrote:
I'm trying to find a solution to an immediate VM crash which occurs
by
simply adding "bios='ovmf' to my configuration?
I started with a standard Ubuntu install which contained Xen 4.6.0
and had
the crash. The VM works fine booting w/ SeaBIOS once the
configuration line
is removed. It also works fine with OVMF using just QEMU.
Then I was led to believe that i needed to rebuild Xen with
--enable-ovmf ,
using the source from apt-get source, this proved to be difficult
and
ultimately I had to disable stub domains to get it to work. (in
case, that
is relevant)
Still I get the crash almost immediately after the VM is launched.
With
sdl=1 I can see the window flash up and disappear.
I should mention that ultimately I plan to use a non-standard build
of the
OVMF kernel, but for now I'm happy to get it working with any one.
Can you try
pci_platform=0 in your guest config pleae?
At this point, I have been studying the source to try and determine
where
things diverge. I find very little use of LIBXL_BIOS_TYPE_OVMF and
can't
find exactly where setting hvm.bios that even matters?
In any case, I would appreciate help on how to avoid the crash
and/or
understanding the Xen modifications for OVMF.
Jason
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
