[Bug fortran/78797] It is time perhaps to implement -std=f2015

[pault at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78797

--- Comment #2 from Paul Thomas <pault at gcc dot gnu.org> ---
(In reply to janus from comment #1)
> +1
> 
> Can you remind us which F15 features we have already?
> 
> On https://gcc.gnu.org/wiki/GFortran/News I see:
> * Support for IMPLICIT NONE (external, type).
> * ERROR STOP is now permitted in pure procedures.
> 
> Anything else?

Assumed type and assumed rank(partial). I thought that we had implemented the
RANK intrinsic as well but the Chivers et al. table shows N.

There are rather a lot of features that are obsolete or deleted in F2015:
Labeled DO loops have been redundant since Fortran 90 and are now specified to
be obsolescent. The arithmetic IF statement has been deleted. The EQUIVALENCE
and COMMON statements and the block data program unit have been redundant since
Fortran 90 and are now specified to be obsolescent. The nonblock DO construct
has been deleted. The FORALL is now specified to be obsolescent.

What has been done with intrinsics, I am not sure; eg. has COSHAPE been
implemented?

I will post the existence of this PR to the list over the weekend togetehr
with: ftp://ftp.nag.co.uk/sc22wg5/n1551-n1600/n1579.pdf

Cheers

Paul