Re: [avr-libc-dev] pgmspace.h compatibility with -mint8

[Theodore A. Roth]

On Wed, 17 Sep 2003, Tomas Vanek wrote:

> Theodore A. Roth wrote:
> >
> > I think using -int8 will break the pgm_*_far() macros. Might need to
> > remove those if -int8 is used.
> >
> > Also, should also change 'unsigned long' to uint32_t in the pgm_*_far
> > macros.
>
> Yes, -mint8 will break many things. I fully accept a global statement
> that »avr-libc is basically not compatible with -mint8. As far as I
> know, the hardest pain of -mint8 is that there is no way how to typedef
> (u)int32.
>
> I was wrong saying "patch to make compatible". I should better say
> patch to make pgmspace.h a little bit more usable with -mint8
> (and makes no harm to others).
>
> And this little bit means on all MCUs with less then 128kB flash.
>
> I have no other reason to enable -mint8 except to save couple of words
> when I run out of code space on ATmega8. And might be usable when
> optimising interrupt service routine as the last resort before rewriting
> it in asm.

Fair enough. I've committed your patch to HEAD and 1.0 branch.

I decided to split that off from my changes to the *_far() macros. We
still need to work on pgmspace.h though for the uint32_t cases, but
I'll start a new thread for that.

Ted Roth