[coreboot] unable to boot linux with coreboot, now with FreeBSD

[Edwin Beasant]

I think you may well have interrupt configuration problems:
There are messages in the boot log about interrupt storm, and
dissabling interrupts, which will then basically fubar any disk access/
Dma.
If you can boot Linux with options 'irqfixup irqpoll', it will start
to give you pointers.
A strong possibility is that the actual pins wired to the pci
interrupts are misconfigured early on, resulting in a permenantly 'on'
interrupt source.
I would look carefully at the interrupt routing table, and also the
PCI interrupt setup? Looking for things like interrupt overlap,
misconfigured GPIO etc...

Hope this helps!
Edwin

On 14 Jun 2010, at 21:40, "mbertens" <mbertens at xs4all.nl> wrote:

> Hi all,
>
> A new boot log but now with freeBSD.
>
> It seems to me that the memory gets trashed by something, but what?
>
> FreeBSD is able too boot, but encouters a interrupt storm. Normally
> this
> is due to faulty PIRQ table, and the only thing that i can think of is
> that the memory table gets trashed too.
>
> Anyone having any ideas.....
>
>
> On Sun, 2010-06-13 at 11:34 +0200, mbertens wrote:
>> The second bootlog
>>
>> On Sun, 2010-06-13 at 11:30 +0200, mbertens wrote:
>>> Hi all,
>>>
>>> an update:
>>> I installed a secondary build system (ubuntu), i build everything
>>> from
>>> scratch; seabios and coreboot r5575, r5600 and r5631, all these
>>> where
>>> working fine. But booting them all failed.
>>>
>>> I tried  everything i could think of;
>>> 1. Swapped hardware; memory, system board, harddisks, ide-controller
>>> (pri. versus sec.).
>>> 2. Swapped development systems (debian versus ubuntu)
>>> 3.
>>>
>>>
>>> I attached two bootlogs. for analise
>>> 1.  with 80Gb harddisk.
>>>    - up to 7 secs in the boot process all is fine
>>>    - then 120 secs of silence, and last the message from udevd
>>> 2.  with 1Gb compact flash disk (which has a DMA problem, but its
>>> able to
>>> boot futher in the Linux boot secuence.
>>>    - up to 6 secs in the boot process all is fine
>>>    - then 120 secs are lost to the DMA problem of the CF card, the
>>> system
>>> booted before with de DMA failures just fine it takes only 120 secs
>>> extra.
>>>    - then 120 secs of silence, and the message from udevd
>>>    - then some more output on starting the cardbus and others.
>>>    - final message adding the swap on /dev/hda5 and then silence
>>> again.
>>>
>>> I hope that anyone can help me with this silly problem, please don't
>>> just throw a term through the mail, like "fix the DMA problem",
>>> but give
>>> some pointers how to solve it in my configuration.
>>>
>>> Marc
>>>
>>>
>>>
>>> On Fri, 2010-06-11 at 20:43 +0200, mbertens wrote:
>>>> On Thu, 2010-06-10 at 12:31 -0600, Myles Watson wrote:
>>>>>>>>> Anyone any ideas
>>>>> Have you tried to boot the factory BIOS?  Are these all cold
>>>>> boots?  Did
>>>>> cold boot always work?
>>>>>
>>>>> Thanks,
>>>>> Myles
>>>>>
>>>> I have two systems here on the table;
>>>> 1.A.      original BIOS (insyde) boots just fine with FreeBSD
>>>>  B       with external ROM on the JDEBUG connector (see
>>>> http://www.coreboot.org/Tutorials:Nokia_IP530#J-DEBUG ) im not
>>>> able to
>>>> boot coreboot (R5628 or 5600 with my patches which worked before).
>>>> 2.        The on-board flash contains R5512 that does not boot as
>>>> well and this
>>>> one booted before too.
>>>>
>>>> So as you can understand i'm a little lost
>>>>
>>>> I tried adding bayou with payloads seabios and coreinfo, seabios is
>>>> starting as before, but coreinfo doesn't do anything.
>>>>
>>>>
>>>> Marc
>>>>
>>>>
>>>>
>>> --
>>> coreboot mailing list: coreboot at coreboot.org
>>> http://www.coreboot.org/mailman/listinfo/coreboot
>> --
>> coreboot mailing list: coreboot at coreboot.org
>> http://www.coreboot.org/mailman/listinfo/coreboot
> <dump.log>
> --
> coreboot mailing list: coreboot at coreboot.org
> http://www.coreboot.org/mailman/listinfo/coreboot