Re: [Patch, Fortran] PR 38415: procedure pointer assignment to abstract interface

["Janus Weil" <jaydub66 at googlemail dot com>]

Btw I'm afraid I'll have to remove the gfc_compare_interfaces check in
gfc_check_pointer_assign again, since I just noticed that it has lots
of problems with intrinsics (both in lvalue and rvalue). Since these
cannot be fixed in 4.4 any more, I guess it's better to omit the
checking for some invalid cases, than to reject many valid ones. For
this patch, I just commented out the check, so that it can be
reactivated later (i.e. in 4.5), once all the troubles with intrinsics
are dealt with (one of those problems is that we don't copy the formal
args of intrinsic interfaces in PROCEDURE statements yet, and: the
typespec of the return value *is* copied, but too late).

Cheers,
Janus



2008/12/5 Janus Weil <jaydub66@googlemail.com>:
> Hi all,
>
> this is a very simple patch for a small accepts-invalid problem with
> procedure pointers that was recently reported by Tobias (thanks for
> finding this!). The patch is really trivial, and I'd say it's
> reasonable to still include it in 4.4. But of course I'll leave that
> decision to the reviewers. It's regtested on i686-pc-linux-gnu, and I
> think one can be sure that it doesn't introduce any new regressions.
> Ok for trunk/4.5?
>
> Cheers,
> Janus
>
>
> 2008-12-05  Janus Weil  <janus@gcc.gnu.org>
>
>        PR fortran/38415
>        * expr.c (gfc_check_pointer_assign): Added a check for abstract
>        interfaces in procedure pointer assignments.
>
>
> 2008-12-05  Janus Weil  <janus@gcc.gnu.org>
>
>        PR fortran/38415
>        * gfortran.dg/proc_ptr_2.f90: Extended.
>