960823: problem building FORTRAN API under UNICOS: m4(1)

Carl,

>Date: Thu, 22 Aug 1996 16:07:57 -0700 
>From: address@hidden (Carl M. Lund)
>Organization: Los Alamos National Laboratory
>To: Steve Emmerson <address@hidden>
>Subject: Re: 960822: problem building FORTRAN API under UNICOS: m4(1) 
>Keywords: 199608221440.AA29398

In the above message you wrote:

> I got your file on known bugs, and still have problems.  I built
> jackets.c e= tc on a sun, as shown in the following dialog:
> 
> 
...
> >seagal.lanl.gov|cmlund|29> fortc -L . -O unicos jackets.src > jackets.c
> >seagal.lanl.gov|cmlund|30> fortc -L . -O unicos common.inc > netcdf.inc
> >seagal.lanl.gov|cmlund|31> fortc -L . -O unicos ftest.src > ftest.f
> >seagal.lanl.gov|cmlund|32> ls
...
> I then repeated the installation process on the Cray
...
> >making `all' in directory /usr/tmp/cmlund/dnetcdf/netcdf-2.4.3/src/fortran
> >
> >       c89 -c -O -I../libsrc  jackets.c
> >cc-506 c89: ERROR File =3D jackets.c, Line =3D 579
> >  A derived type of "function returning function" is not valid.
...
> As you can see, the Cray was not happy with jackets.c, netcdf.inc and
> ftest.f.  Any suggestions will will be eagerly acted upon!

The above doesn't happen on our CRAY Y-MP under UNICOS 8.0.4 so I'm not
sure what to look for.  What does the file `jackets.c' look like?  What
is the compiler unhappy about?

[Aside: my mailer was very unhappy with the format of your last message.
Would you please simplify the next one.  Thanks.]

--------
Steve Emmerson   <address@hidden>