Jeff Johnson wrote:

> Havard Eidnes wrote:
>
>>> I find it interesting that bge0 will cause a GPF/crash but bge1
>>> functions perfectly.  They look identical by comparison. The
>>> only difference would be their pci addresses and the ordering
>>> in which they appear to the kernel. It would be interesting if
>>> I could somehow reverse their order so the controller/phy that
>>> is now bge1 was seen as bge0 by the kernel and vice-versa.
>>>   
>>
>>
>> That should be relatively straight-forward to do.
>>
>> Looking at your earlier posted dmesg, I'd say insert
>>
>> bge0 at pci2 dev 3 function 0
>> bge1 at pci2 dev 2 function 0
>>
>> in your kernel config file, and you'll have the bge's swapped.
>>
>> Regards,
>>
>> - Håvard
>>  
>>
> I recompiled the 2.0.2 amd64 kernel source and rearranged the order in 
> the kernel config file so bge0 became bge1 and bge1 became bge0. I 
> booted this new kernel and I got the same failure of a kernel panic 
> when the port when ifconfig brought the port up during the boot 
> sequence. In other words the kernel panic condition follows the 
> physical port regardless of whether it is set up as bge0 or bge1. 
> Review of dmesg shows the controllers and phys between both ports to 
> be identical.
>
> Anyone have any ideas?
>
Update.. I removed an LSI dual channel PCI card (mpt driver) from the 
system and now it can boot up and activate the bge0 or bge1 with no 
kernel panic or failures. The LSI card was on the same pci bus (pci2) so 
I will move it around and try and come up with a more definitive view of 
the problem.

More to come.

Jeff