[LinuxBIOS] OK, I'm lost

[Corey Osgood]

Peter Stuge wrote:
>> In the src/northbridge and src/sourthbridge directories I assume
>> I'll need to create a cx700 directory.
> 
> northbridge yes, southbridge maybe. The southbridge seems to be a
> vt8237 core, perhaps you don't need to make a special southbridge
> directory for it.
> 

I'm rather curious, I'm wondering if CX700 = CN700 + VT8237. If that's
the case, CX700 should work with CN700 code, which should be ready by
the 27th. I'll take a look later, right now I'm off to work.

Also, do you have the CX700 datasheets? And is your board using ddr or ddr2?

> What about CPU? Does your board have a C3 or C7 CPU? Currently
> there's only code for C3 in LB.

C7 *should* work (if I'm reading the docs right, that is) with the C3
code, the only thing that needs to be changed is the device IDs added to
src/cpu/via/model_centaur/model_centaur_init.c, C7 is 0x06A[0-F]. Other
issues may crop up after ram is working, but again should be working by
the 27th, if all goes well.

>> But once again I'm at a loss as to which directory to use as a base
>> platform.  Everything I know about the CX700 seems so different
>> that I'm concerned nothing will work.  If I can just get it started
>> to a point where I start getting serial output
> 
> This is definately the starting point, but it already requires the
> superio to be working properly. There's no support for vt8237 but
> perhaps the vt8235 is at least a little similar. (Wasn't the 8237
> being worked on already by the way?)

Yes, it is ;)

>> I think I'll have a place where I can at least start working and
>> try things.  If I use the EPIA-M setup, do you think I'll at least
>> get that far?
> 
> Maybe if you're lucky. epia-m has vt8235. Try it. :)

VT8235 has its own superio, and epia-m uses it. VT8237 doesn't. Also,
the device IDs are entirely different, and so are some of the registers.
I used epia-m as a starting point also, but it requires a lot of gutting
out and fixing up.

-Corey