🚀 go-pugleaf

unicbe@fysik.uni-c.dk (Claus Bendtsen) writes:

>I first sent this message to bug-gcc but since I have received no
>response I now try help-gcc.
>
>I would like to use g++ on each node of a Thinking Machines, CM5 and in
>the process of doing that I've run into a few problems.
>
>I currently have a working model of the following kind,
>
>	a) main program in C, calls C++ main program adding _FiPPc.
>	b) C++ program works, calls fortran in order to use the CM5, adding
>	   _ to all calls, which further more are defined as extern "C".
>	c) Linking using fortran linker.
>
>Now my question is, since I would like not to have to include the C
>main program I need the gcc to be able to generate some routines (e.g.
>main) without expanding them, just treat them as ordinary C in order
>for the fortran linker to find them.

That's easy - just declare these routines as extern "C".
See Q106 in the comp.lang.c++ FAQ list.

>Further more I would like not to
>have to add the underscore for my fortran routines by using a different
>language style in the extern "??" definition. I can't find any language
>styles in the documentation but the C++ standard has "room" for it.

I'm afraid you're out of luck here. 
GNU C++ doesn't support extern "Fortran".

-- 
Fergus Henderson        |   "People who brook no compromise in programming
                        |   languages should program in lambda calculus or
fjh@munta.cs.mu.OZ.AU   |   machine language, depending." --Andrew Koenig.