https://gcc.gnu.org/bugzilla/show_bug.cgi?id=128682
Mikael Morin <mikael at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |mikael at gcc dot gnu.org
--- Comment #9 from Mikael Morin <mikael at gcc dot gnu.org> ---
(In reply to Steve Kargl from comment #7)
> (In reply to Steve Kargl from comment #6)
> > (In reply to Harald Anlauf from comment #5)
> > >
> > > This function is very old code, so we need to exercise it a bit.
> > > Can you do that?
> >
> > I'll see what I can do. It's been a long time since I hacked on
> > gfortran source code.
>
> This simple patch fixes OP's issue. I haven't run check-fortran,
> as I need to update my gcc tree and rebuild everything first.
>
>
> diff --git a/gcc/fortran/trans-expr.cc b/gcc/fortran/trans-expr.cc
> index 289ae0311fa..28db117f1ae 100644
> --- a/gcc/fortran/trans-expr.cc
> +++ b/gcc/fortran/trans-expr.cc
> @@ -2288,16 +2288,14 @@ tree
> gfc_get_expr_charlen (gfc_expr *e)
> {
> gfc_ref *r;
> - tree length;
> + tree length = NULL_TREE;
> tree previous = NULL_TREE;
> gfc_se se;
>
> - gcc_assert (e->expr_type == EXPR_VARIABLE
> - && e->ts.type == BT_CHARACTER);
> + gcc_assert (e->ts.type == BT_CHARACTER);
>
> - length = NULL; /* To silence compiler warning. */
> -
> - if (is_subref_array (e) && e->ts.u.cl->length)
> + if ((is_subref_array (e) || e->expr_type == EXPR_FUNCTION)
> + && e->ts.u.cl->length)
> {
> gfc_se tmpse;
> gfc_init_se (&tmpse, NULL);
That's another way. It could work I guess.