[coreboot] [RFC] Netbook Support

[Carl-Daniel Hailfinger]

On 13.12.2008 17:14, Jordan Crouse wrote:
> ron minnich wrote:
>> This is a place where I'd like us to pick a common most-appropriate
>> notebook and go for it.
>>
>> So which of these is the most likely to be doable? We'll have to solve
>> the EC problem.
>
> But don't be confused as to what the "EC problem" really is.  We've
> discussed this before on this very list:  Open firmware for the EC
> would be great, but it is a problem on the scope of coreboot itself,
> since every platform would need a slightly "custom" variant of the
> firmware, and over time you would end up needing to support different
> "chips" for different "mainboards".  Certainly that is an interesting
> problem to solve, but I don't think that it is _our_ problem to solve.
>
> Coreboot would be much better served by operating with the "stock"
> firmware.  Don't get me wrong - this is still a challenge because,
> like I said in IRC yesterday, we don't know what we don't know.  
> Behavior at runtime is fairly standardized, but we don't know what we
> need to do for initialization - do we need to set up registers, put in
> tables, kick things, or will it all Just Work (TM)?

There's also the interesting question about who owns the EC firmware. If
the manufacturer owns it and has full rights for it, the situation is
different compared to the BIOS. For example, Quanta (OLPC manufacturer)
holds some (if not all) rights to the OLPC XO EC code and not BIOS
vendor was involved.

> I guess the only way to find out is to start trying - enough
> documentation exists (especially for th ENE) that we can probably
> determine how large the firmware is and extract it from a stock ROM, and
> try it out.  From what I have heard, I think the VIA processors are
> the best bet right now for experimentation.  Intel of course, is out,
> and AMD processors aren't in any netbooks to speak of - and I'm sure
> if we go to this effort, we would rather have it pay off for a netbook
> rather then a dinosaur laptop.

If I'm right about EC code copyrights and usage rights, a cooperating
manufacturer could easily tell us about the interface for their EC code.

On a side note (or rather two):
- I seem to remember that VIA manufactures ECs as well. If that is true,
we could ask them nicely whether any netbooks have a VIA EC and start
with that.
- Some people here were successful in talking with Winbond about data
sheets. While these data sheets were not for ECs, it can't hurt to ask.

As long as we can avoid time consuming reverse engineering, we should
IMHO focus on getting stuff done.


Regards,
Carl-Daniel

-- 
http://www.hailfinger.org/