APCS-32 & 'float' argument passing

Nick Burrett nick at sqrt.co.uk
Fri Aug 20 04:43:50 PDT 2004


John Tytgat said:
> In message <40C968F1.9040908 at dsvr.net>
>           Nick Burrett <nick at dsvr.net> wrote:
>
>> I looked into this problem in more detail last night.  The changes to
>> GCC are non-trivial since there is no actual mechanism in the middle-end
>> or back-end to do this auto-promotion.
>>
>> Making this change would be a violation of the ISO C99 standard.

Further on this, I suspect that the 2.7 compiler probably handled things
correctly because it would have had the code to support the old K&R style
of argument promotion.

> Picking up an old thread : instead of trying to persue that the ABI of
> GCC is 100% in sync with the one of Norcroft, would it be acceptable to go
> for a solution based on a compile option for UnixLib instead ? I.e. an
> option defining whether you're building a GCC or Norcroft version of
> UnixLib ?

It certainly is a possibility.  I don't have any concerns about such a
change being made.

Nick.





More information about the gcc mailing list