Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!cbosgd!gatech!seismo!brl-tgr!gwyn
From: gwyn@brl-tgr.ARPA (Doug Gwyn <gwyn>)
Newsgroups: net.unix
Subject: Re: Calling f77 routines from C: HELP!
Message-ID: <1996@brl-tgr.ARPA>
Date: Sat, 25-Jan-86 23:41:55 EST
Article-I.D.: brl-tgr.1996
Posted: Sat Jan 25 23:41:55 1986
Date-Received: Wed, 29-Jan-86 04:15:11 EST
References: <1930@uwmacc.UUCP>
Distribution: net
Organization: Ballistic Research Lab
Lines: 24

> I am trying to call a f77 subroutine from a C program,
> and I find that the supposedly "preconnected" i/o units 5 and 6
> are not connected.  That is, when I do a simple print in the
> Fortran subroutine, like:
> 
> 	print *, 'hi!'
> 
> I get an entry in a file called fort.6.  What makes this especially
> unfortunate, is that this works fine in 2.8BSD, and we have found that
> the C/f77 programs that we wrote under 2.8BSD are failing to run
> as expected under Ultrix (for the reason just described).
> 
> Is there a way to preserve the "preconnections" advertised in the
> f77 documentation?

In general, it is unwise to mix C and Fortran unless the main
module is Fortran, and even then there are restrictions on
what is safe to do with respect to I/O.  This is due to the
need to set up Fortran I/O for the run-time system before
doing any Fortran I/O; this is normally done for you in the
Fortran start-off module.  You can try doing it "by hand" by
calling the function f_init(); some "f77" implementations
(e.g., SVR2) keep track of whether or not initialization has
been done and do it automatically if necessary.