This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
RE: Spec 2005 and gfortran
- From: THOMAS Paul Richard 169137 <prthomas at drfccad dot cea dot fr>
- To: "'Toby White'" <tow21 at cam dot ac dot uk>
- Cc: "'fortran at gcc dot gnu dot org'" <fortran at gcc dot gnu dot org>, "'stevenb at novell dot com'" <stevenb at novell dot com>, "'michael dot meissner at amd dot com'" <michael dot meissner at amd dot com>
- Date: Wed, 3 Aug 2005 08:37:09 +0200
- Subject: RE: Spec 2005 and gfortran
Toby,
Thank you sir. I will clear all the remaining PRs tonight.
>
> I've gone ahead & closed it. There's another, dependent, bug,
> PR20663, which is also now fixed. I can't close that since
> I wasn't the submitter.
>
> The identifying feature of 20482 is that it
> involved overloading an existing function, while 18108
> only involved overloading user-written functions.
>
> 20663 involves a generic name for a user-written function
> which happens to have the same name as a fortran statement.
>
> From the little I know of gfortran internals, I don't think
> these actually warrant a separate testcase, but it's easy
> enough to create one if anyone thinks it's worth it.
I am not sure that you are right, although I will have a think about it. The
mechanism that was failing was that of the host association of generic
interfaces; ie those interfaces that one way or another provide a single
call to a number of functions, selected by their interface. Over-riding
intrinsic or module functions is different in its detail, so I should
probably provide testcases; if for no other reason than to make sure that
the testsuite covers potential, future regressions.
Thanks
Paul Thomas