https://gcc.gnu.org/bugzilla/show_bug.cgi?id=128682
--- Comment #4 from Steve Kargl <kargl at gcc dot gnu.org> ---
(In reply to Harald Anlauf from comment #3)
> (In reply to Steve Kargl from comment #2)
> > Another workaround is to place parentheses around the actual
> > argument
> >
> > call show((transpose(reshape([(blk(i:i), i=1, len(blk))], [w, h]))))
>
> Indeed.
>
> Funny. I thought that Mikael recently addressed the issue with TRANSPOSE
> as actual argument.
I don't remember any commit, but I know he has don very cleanups
with descriptors.
The ICE is coming from an assert in gfc_get_expr_charlen() at
trans-expr.cc:2295
gcc_assert (e->expr_type == EXPR_VARIABLE
&& e->ts.type == BT_CHARACTER);
The comment above the function is
/* Get the character length of an expression, looking through gfc_refs
if necessary. */
The untested fix is to do
gcc_assert (e->ts.type == BT_CHARACTER);
or
gcc_assert ((e->expr_type == EXPR_VARIABLE || e->expr_type == EXPR_FUNCTION)
&& e->ts.type == BT_CHARACTER);