[U-Boot-Users] Breakage of board ports on new features.

[Kumar Gala]

On Dec 5, 2006, at 9:35 AM, Wolfgang Denk wrote:

> In message <F245C92A-F4A7-40AF-A73B- 
> EB453262CB55 at kernel.crashing.org> you wrote:
>>
>> Ok, but I think the several of the boards are capable of running with
>> all flavors of MPC834x what do we do for them?  I know the MDS is and
>> I'm questing TQM834x is as well?
>
> I think so, but I cannot answer the before I've seen what TQ does  on
> the TQM8347 :-(

Are you ok with having CONFIG_MPC8349 set in tqm834x.h?

>> I understand the desire, I just feel that it doesn't add anything
>> because we've done a decent job of making the differences between
>> them covered by feature configs.
>
> As Tolunay explained: it's for consistency with existing boards.  And
> I  really  think it does not hurt ao much to have both CONFIG_MPC834X
> and CONFIG_MPC8349 in the board config file if this is  a  8349  only
> board,  even  if  CONFIG_MPC8349 should never be used anywhere in the
> code.

So how should we set this for TQM834x?  Clearly MPC8349EMDS shouldn't  
have it set, and MPC8349ITX should by your current description.

Also, let me ask, what exactly does CONFIG_MPC8349 mean?  Does it  
mean I'm on the subset of all MPC8349 revisions.  Rev3.0 of MPC834x  
adds DDR2 support so clearly CONFIG_MPC8349 can only be useful to  
distinguish the subset of MPC8349 revisions.

Personally, I feel that defining it before it's clear there is a user  
that needs it lends the define to be incorrectly used in the future.

- kumar