Here are the standard boot error when I'm using Bios defaults ( USB + 
Onboard Lan + Onboard Audio + ... )

uvm_fault(0x08f1be0, 0xbff18000, 2) -> 0xe
kernel: supervisor trap page fault, code=0
Stopped in pid 0.1 ( swapper ) at netbsd:_bus_dmamem_map+0x177: movl 
%eax,0xb fc00000(%edx)
db>

I forgot, here is my configuration :

Mother Board : GigaByte GA-K8N51GMF-9
( includes : nVidia geforce 6100 / nForce 430 / Socket 939 / USB 2 / 
SATA 3GBs / IEEE1394 )
CPU : AMD Athlon 64 X2 4400+
RAM : 2x 1GB DDR Dual Channel

Thanks

I'll continue my tests

On 3 avr. 06, at 19:39, KlinT wrote:

>
> Hi all,
>
> On 2 avr. 06, at 14:50, Manuel Bouyer wrote:
>
>> On Sun, Apr 02, 2006 at 01:44:47AM +0200, KlinT wrote:
>>> Hi all,
>>>
>>> Well I'm trying to setup Xen 2.0.7 on my new Athlon 64 X2 computer 
>>> ...
>>> I'm running NetBSD 3.99.17 and everything runs ok ( maybe except for
>>> the SATA.II feature ) ...
>>>
>>> So I install Xen as usual, according to the documentation ...
>>>
>>> The problem is when I'm trying to boot the XEN0 kernel. I got a 
>>> kernel
>>> error ( swapper error ) and then I enter the debugger ... :(
>>> I've found that problem comes with USB/FireWire components, so I
>>> disable them in the bios and try to boot again the XEN0 kernel ...
>>> Same problem but not at the same place :(
>>
>> Can you post the panic message, and eventually the stack trace ?
>
> I'd like too but there is no "panic message" just a system freeze
>
> I've just build a fresh XEN0 kernel ... Here are the very last lines :
>
> boot device:wd0
> root on wd0a dumps on wd0b
> mountroot: trying lsf ...
> mountroot: trying ffs ...
> root file system type: ffs
> init: copying out path '/bin/init' 11
>
> ---> The kernel stops there without any informations ...
>
>>
>>>
>>> I've tried to build my own kernel from the source. I've successfully
>>> built the toolchain, and a XEN0 kernel using build.sh.
>>> but trying to boot this kernel fails too ... Now, the system freeze 
>>> ...
>>> but never at the same place.
>>> Some times the boot process reach network configuration and then 
>>> freeze
>>> ... I was not able to boot entirely my system.
>>>
>>> Any idea ??
>>>
>>> I think that I have a similar problem than "Florent Parent" who post
>>> this thread "XEN0 kernel with options MULTIPROCESSOR".
>>> And my kernel is not able to deal with the dual-core architecture of 
>>> my
>>> cpu ...
>>
>> No, it's not the same problem. With Xen-2, all multiprocessor support 
>> is
>> in the Xen kernel, not in the dom0 kernel. The domain kernels don't 
>> even
>> know if they're on a multiprocessor or not.
>
> Understood ...
>
>>
>>>
>>> I've also try to optimize the build using some options in 
>>> /etc/mk.conf
>>> and the package pkgsrc/devel/cpuflags ... Don't know if it really 
>>> works
>>> ...
>>
>> Can you try without the flags ? It's possible that some special flags
>> triggers compilers bugs, optimisations issues or use some instructions
>> that are not handled by Xen.
>
> I'll try this now ...
> I'll let you know about the results ...
>
>>
>> -- 
>> Manuel Bouyer <bouyer@antioche.eu.org>
>>      NetBSD: 26 ans d'experience feront toujours la difference
>> --
>>
>>
> Thanks a lot Manuel.
>
>