virtualbox.org

Hi All,

I've been using VB for a few weeks now and lately all my CentOS 6.2 x86 guests (Windows7 hosts) stopped working. The kernel panics with divide_zero error in init_tsc_clocksource on startup. Sometimes if I reboot over and over again it makes it through, but not always.

I'm stumped what could be causing this. I tried to change the kernel clocksource to 'jiffies' and 'acpi_pm' in grub.conf, but that doesn't appear to make any difference. Is this something new to the latest kernel: Linux centos32-vm1 2.6.32-220.7.1.el6.i686

Thanks for ANY help!! I'm dead in the water until I can figure this out.

Thanks!
Corey.

I would try the divider=10 in the boot line and see if that settles it down.

Didn't make any difference...

Does this only happen during a reboot (warm boot)? Does it happen/not happen on a shut down (cold boot)?

It appears to only happen on a cold boot. I attached a screenshot of the kernel panic, maybe there's something else that jumps out at you.
Corey.

I do see this but just the opposite. If I do a reboot it will hang, but if I do a shut down and start it works. I have reported a ticket about this Bugtracker ticket #10427 if you would like to add your information.

I added my info to the ticket. Good hunting.

Corey.

This just hit me for several RH-based VMs. Some googling found the answer on the centos forums. (I'd post a link, but I just created my account, so the rules block it.If you google for "Kernel panic with cent os kernel 2.6.32-220.7.1.el6.i686 virtualbox", it's the 2nd hit. First is this thread. )

To fix it, I had to change the chipset from PIIX3 to ICH9 and turn on IO APIC. This worked for several VMs that were experiencing the issue;
Oracle Linux: 2.6.32-300.21.1.el6uek.x86_64
Oracle Linux: 2.6.32-300.21.1.el6uek.x86_64
CentOS:2.6.32-220.13.1.el6.x86_64

All VMs are 64-bit on a WinXP host.

ICH9 is still experimental and has issues. Actually it has more issues than the annoyance of not being able to cold boot the machine.

I wasn't aware of that. Thanks for the heads-up!

I went with it since my affected VMs wouldn't boot at all. Not cold boot or reboot.

I was having the same issue when I would install 6.2 and then yum update to the .300 kernel.

Changing to the ICH9 Chipset seemed to do it.

Thanks

J

jfmatth wrote:I was having the same issue when I would install 6.2 and then yum update to the .300 kernel.

Changing to the ICH9 Chipset seemed to do it.
J

I'll add the infamous "me too". This fix was necessary to boot centos 6.2 kernel 2.6.32-220.23.1.el6.i686. I didn't experience this with any earlier kernel.

Could you also try if using the normal PIIX3 chipset with activating HPET (VBoxManage modifyvm VM_NAME --hpet on) works as well?