-fdefault-integer-8 again!-(

[Dominique Dhumieres]

Paul,

I am sorry that my mail upset you, it was not meant to.

As a general comment I'ld ask the native English speakers to not overreact 
to the wording of nonnative English ones.  In the particular case I should 
have written "embarassing for me" instead of "annoying" (sloppy transation 
for "ennyeux") since I did the testing after the commit and not before.

To go back to gfortran, I have filled pr38347 for the ICE on
alloc_comp_constructor_1.f90 and added a comment to pr38324.
Note that I did not debugged directly alloc_comp_constructor_1.f90, but I 
saw bound errors in some of its variants I have in store.

> However, this platform is becoming so aberrant as to be unmaintainable
> for the likes of me.  I do not have this machine nor do I see the
> problems that it has.

I would add to the jack's answer:

> I am well aware of how annoying it will become to support PowerPC going
> forward considering that the main hardware source, Apple, is no longer
> creatin new hardware for that architecture.  However I think it is still
> worth the effort because an IA32/X86_64 monoculture for FSF gcc is less
> likely to trigger corner cases that would remain latent otherwise.  Just
> my 2 cents on the issue.

two things:

(1) ppc/intel Darwin are not more aberrant aberrant than other platforms. 
Most of the problems (beside some Aplle's bugs) are due to endianness 
problems on ppc and -fpic problems not tested on IA32/X86_64.

(2) In addition of the testing I am already doing, I can offer a few 
accesses to my G5 (as I did for FX or Andy Vaught in the past).

Cheers

Dominique