Call for Testers: VirtualBox 4.0.6 (PBIs now available)
rysto32 at gmail.com
Fri May 6 22:49:26 UTC 2011
On Fri, May 6, 2011 at 6:47 PM, Ryan Stone <rysto32 at gmail.com> wrote:
> On Wed, Apr 27, 2011 at 3:28 PM, Ryan Stone <rysto32 at gmail.com> wrote:
>> (Sorry for the noise earlier about the PBIs not working under PC-BSD;
>> I'm not sure how I missed that had been already reported).
>> I tried the new amd64 PBI and I am able to successfully start VMs now.
>> I had one VM(running some relatively recent version of amd64 HEAD)
>> boot up fine, but a second one (also running amd64 HEAD, but maybe a
>> different svn revision) gets a Guru Meditation during startup. I'm
>> not sure what is causing the crash. In their original configurations
>> the working VM booted off of an emulated IDE disk while the broken VM
>> booted of an emulated SATA disk, however I just tried changing the
>> SATA disk to instead be an IDE disk and it doesn't seems to have
>> resolved the problem.
>> I'm not getting a corefile for the crash(or I'm unable to find it). I
>> do have kern.sugid_coredump=1 and I'm running the PBI. Is this
>> expected for a Guru Meditation, or should it be putting a core
>> somewhere? I've put the VBox.log for the most recent crash here:
>> This seems to be easy to reproduce so let me know if there's any more
>> information that I can gather.
>> Also, the working VM was emulating uniprocessor machine. I tried
>> adding a second CPU and that VM started crashing, too. I tried
>> changing the broken VM to have only one core but it still crashes.
>> I'm not sure if it's related to the first crash or not.
> To follow up, I tried playing around with the "working" VM
> configuration. I can only make it fail if I emulate a machine with
> multiple processors and I leave SMP enabled in the kernel. If I
> either emulate a uniprocessor machine or disable SMP via tunable, the
> VM is able to boot without crashing.
> To summarize, I believe that I am running into two separate issues
> with VirtualBox 4.0.6:
> - Booting off of an emulated SATA drive leads to crashes during boot:
> http://people.freebsd.org/~rstone/vbox-4.0.6/VBox.log (Guru Meditation
> -4001 (VERR_VMX_INVALID_VMCS_PTR)
> - Booting a VM with multiple virtual processors on a kernel with SMP
> enabled crashes during boot:
> http://people.freebsd.org/~rstone/vbox-4.0.6/VBox-SMP.log (Guru
> Meditation -4002 (VERR_VMX_INVALID_VMXON_PTR))
It occurs to me that I've left out a very important piece of
information. I was previously able to boot the SATA, dual-core VM
under VirtualBox 4.0.4, so this is a very recent regression.
More information about the freebsd-ports