> -----Original Message-----
> From: Iain Sandoe <idsan...@googlemail.com>
> Sent: Monday, May 5, 2025 14:46
> To: Robert Dubner <rdub...@symas.com>
> Cc: GCC Patches <gcc-patches@gcc.gnu.org>; Jakub Jelinek
> <ja...@redhat.com>; James K. Lowden <jklow...@cobolworx.com>
> Subject: Re: [PATCH] cobol: Rewrite exception handling. Partially refactor
> subscript/refmod calculations.
>
> Hi Bob,
>
> > On 5 May 2025, at 02:49, Robert Dubner <rdub...@symas.com> wrote:
> >
> > I know it's the weekend. But this has been niggling at me, so I decided
> > to push it.
>
> This causes bootstrap to fail on x86_64 darwin with:
>
> Undefined symbols for architecture x86_64:
> "_program_invocation_short_name", referenced from:
> ___gg__check_fatal_exception in libgcobol.o
> ld: symbol(s) not found for architecture x86_64
>
> Looking at the patch, it seems that the symbol is not, indeed, defined
> anywhere.
That symbol is very much supposed to be there. Many things won't work
without it.
The function is defined in libgcobol.cc, at line 11321
extern "C"
void
__gg__check_fatal_exception()
{
There are two leading underscores, not three.
>
> Is it intended to be undefined so that it can be bound at runtime?
> (although I don’t see anything in the FE that would be emitting it).
>
> If that’s so, then I can fix the Darwin build to allow it to be undefined.
>
> thanks
> Iain
