gcc 3.4.2 cpu32 patch

[Chris Johns]

Ralf Corsepius wrote:
> 
>>>So, we would have to reflect this change into rtems:
>>>#if __mc68020__
>>>#if __cpu32__
>>>..
>>>#else
>>>#endif
>>>#endif
>>
>>In the long term, I think adopting this new scheme would
>>be best for user-code such as RTEMS.
> 
> 
> OK, fine with me. I think we have a solution for this problem.
> 
> 

I can implement the above but it will not be compatible to versions 
before 3.4.x. The 68020 test could be:

#if defined(__mc68020__) && !defined(__cpu32__)

Given the change will be on the mainline and you need other important 
tool changes I prefer your suggested change.

> 
>>A much better scheme would be adding predefines for the
>>different m68k ISAs instead of trying to capture them with
>>chip names.  This scheme would still be full of pitfalls
>>because there's no clear subset/superset relation between
>>different members of the m68k family.
> 
> 
> This is the same problem we are facing with other CPUs. GCC is moving
> towards ISAs and ABIs, while RTEMS internals historically is "structured
> by CPU-variants", because this often allows a finer grained/more
> detailed optimization than GCC needs.
> 

I am pleased to see GCC move to be defined in terms of ISA etc.


Thanks Peter and Bernardo.


-- 
  Chris Johns