Prev: httpd PID and finding the linked web site
Next: 7.n/8.n on the Xeon --how kludgey is that, anyway?
From: Indi on 11 Jun 2010 13:41
After running freebsd-update and portmaster -av apparently
successfuly, I rebooted and got this:

Fatal trap 12: page fault while in kernel mode
cpuid = 0; apic id = 00
fault virtual address = 0x4
fault code = supervisor read, page not present

Subsequent boot attempts with acpi disabled, safe mode, and
even restoring the GENERIC kernel (using livefs cd) have been
no help at all.

This is FreeBSD 8.0-RELEASE-p3 on intel mac mini 2.0GHz.
The machine dual-boots OS X no problem, so it's unlikely to be a
hardware issue.

Any help would be greatly appreciated.

--
Caveat utilitor,
indi
From: Indi on 11 Jun 2010 14:04
On 2010-06-11, Indi <indi(a)satcidananda.16x108.merseine.nu> wrote:
> After running freebsd-update and portmaster -av apparently
> successfuly, I rebooted and got this:
>
> Fatal trap 12: page fault while in kernel mode
> cpuid = 0; apic id = 00
> fault virtual address = 0x4
> fault code = supervisor read, page not present
>
> Subsequent boot attempts with acpi disabled, safe mode, and
> even restoring the GENERIC kernel (using livefs cd) have been
> no help at all.
>
> This is FreeBSD 8.0-RELEASE-p3 on intel mac mini 2.0GHz.
> The machine dual-boots OS X no problem, so it's unlikely to be a
> hardware issue.
>
> Any help would be greatly appreciated.
>

Just a little more info:
This is the i386 version of FreeBSD (should have mentioned that above).

I booted from the livefs again and moved the kernel out of the way,
then manually installed the generic kernel from the installer.
Attempting a normal boot on that kernel, I got as far as:

firewire 0: 1 nodes, maxhop <= 0 cable IRM irm (0) (me)
firewire: bus manager 0

And then it hangs and will go no further.

Same kernel attempting to boot without acpi and again in safe mode
yield the same original "fatal trap 12" error.

Something in the latest updates apparently doesn't like my hardware...
Or at least I hope it's only on this hardware -- later when I have
time I will update the installation on my laptop and pray it doesn't
fail too, unless I read that others are having the same issue of course.

--
Caveat utilitor,
indi
From: Torfinn Ingolfsen on 11 Jun 2010 16:42
Indi wrote:
> firewire 0: 1 nodes, maxhop <= 0 cable IRM irm (0) (me)
> firewire: bus manager 0
>
> And then it hangs and will go no further.

How long did you wait? Give it at least 10 - 15 minutes, something could
be broken.

> Same kernel attempting to boot without acpi and again in safe mode
> yield the same original "fatal trap 12" error.
>
> Something in the latest updates apparently doesn't like my hardware...
> Or at least I hope it's only on this hardware -- later when I have
> time I will update the installation on my laptop and pray it doesn't
> fail too, unless I read that others are having the same issue of course.

Do you have anything in /boot/modules? Some ports create modules there,
and if one of those is out of sync, bad things can happen.

Have you checked /boot/loader.conf?
HTH
--
Torfinn Ingolfsen,
Norway
From: Indi on 11 Jun 2010 22:30
On 2010-06-11, Torfinn Ingolfsen <tingo(a)start.no> wrote:
>
> How long did you wait? Give it at least 10 - 15 minutes, something could
> be broken.
>
>> Same kernel attempting to boot without acpi and again in safe mode
>> yield the same original "fatal trap 12" error.
>>
>> Something in the latest updates apparently doesn't like my hardware...
>> Or at least I hope it's only on this hardware -- later when I have
>> time I will update the installation on my laptop and pray it doesn't
>> fail too, unless I read that others are having the same issue of course.
>
> Do you have anything in /boot/modules? Some ports create modules there,
> and if one of those is out of sync, bad things can happen.
>
> Have you checked /boot/loader.conf?
> HTH

Thanks for the very helpful reply!
You were spot on, it's the vboxnetflt module.
It seems that it will stop the boot process but can be loaded manually
after booting without a problem (so far, it's only been 20 minutes or
so). Also I tried windows in virtualbox and it works just as before.

--
Caveat utilitor,
indi

From: Warren Block on 12 Jun 2010 10:30
Indi <indi(a)satcidananda.16x108.merseine.nu> wrote:
> On 2010-06-11, Torfinn Ingolfsen <tingo(a)start.no> wrote:
>>
....
>> Do you have anything in /boot/modules? Some ports create modules there,
>> and if one of those is out of sync, bad things can happen.
>>
>> Have you checked /boot/loader.conf?
>> HTH
>
> Thanks for the very helpful reply!
> You were spot on, it's the vboxnetflt module.
> It seems that it will stop the boot process but can be loaded manually
> after booting without a problem (so far, it's only been 20 minutes or
> so). Also I tried windows in virtualbox and it works just as before.

Rebuilding the virtualbox-ose-kmod port may help. Actually, kernel
modules are supposed to be rebuilt after kernel upgrades, but there's no
mechanism for notifying about that.

If you can live with the VM being behind NAT, that mode doesn't need the
vboxnet module at all.

--
Warren Block * Rapid City, South Dakota * USA
 |  Next  |  Last
Pages: 1 2
Prev: httpd PID and finding the linked web site
Next: 7.n/8.n on the Xeon --how kludgey is that, anyway?