[Patch, Fortran] PR 85841: [F2018] reject deleted features

[Janus Weil]

2018-05-21 0:19 GMT+02:00 Steve Kargl <sgk@troutmask.apl.washington.edu>:
> On Sun, May 20, 2018 at 09:44:47PM +0200, Janus Weil wrote:
>>
>> >> The patch still regtests cleanly. Ok for trunk?
>> >
>> > Patch looks good to me.  The only thing that worries me is
>> > whether the patch will cause the SPEC benchmark to throw
>> > an error or warning that it did not before.  As I don't have
>> > SPEC benchmark and it cost $$$, I'm not going to let it
>> > bother too much.
>>
>> Unfortunately I don't have access to SPEC either. Is anyone in a
>> position to check this?
>
> We'll find out evidently as one of the C/C++ developers run SPEC.

Ok, we have plenty of time to fix up things before the next release anyway.


>> If it is indeed a problem, I could remove the warning for the F2018
>> deleted features with default flags (-std=gnu). However, I'd only like
>> to do that if it's really necessary.
>
> I would rather see the warnings to encourage users to fix
> their codes.

Me too. Good that we can agree on something for once ;)


>> >> Btw, with the arrival of the F2018 standard, I wonder whether it
>> >> actually makes sense to keep the option -std=f2008ts, or to remove it
>> >> in favor of -std=f2018, since the two Technical Specifications covered
>> >> by this flag are now part of F2018 (and pretty much the main part!).
>> >> Any opinions on this?
>> >
>> > I think that f2008ts can go away.  We may need to do this
>> > in two step as some users may have f2008ts hardcoded in
>> > Makefiles.  Probably, issue warning for -std=2008ts and
>> > map it to -std=2018 for a brief period (3 to 6 months?)
>> > and then finally remove it.
>>
>> Yes, I agree that it would be a good idea to keep -std=f2008ts as an
>> alias for f2018 for a while (possibly until the next release, 9.0),
>> and document is as deprecated. I can take care of removing the
>> GFC_STD_F2008_TS macro in a follow-up patch.
>
> Sounds good to me.
>
>> Actually there is one other thing I do not like about my previous
>> patch: The duplication of GFC_STD_* flags in gfc_match_stopcode. To
>> get rid of this, I'm introducing additional macros in libgfortran.h,
>> see new patch in the attachment. I hope you agree that this is an
>> improvement. (Again: Regtests cleanly.)
>
> Again, looks good.

Thanks. I have committed this version of the patch as r260433.


> Welcome back to gfortran development.

Well, I do have a continued interest in improving gfortran. As usual,
I cannot make any promises (too many boundary conditions), but I will
certainly try to make the occasional contribution if I find the time.

It would be amazing to finally have complete Fortran 2003 support with
gfortran-9. I guess we're really only a handful of PRs away from that
goal now. Many of the small F2018 improvements should not be too hard
to implement either.

Cheers,
Janus