RFC POSIX Fortran Interface

[Steve Kargl]

On Sun, Oct 14, 2007 at 01:54:01PM -0700, Walter Spector wrote:
> Steve Kargl wrote:
> > ...
> > I haven't read through the code in detail; however, I'm wondering
> > what your plans are.  In particular, do you plan to add the PXF
> > functions as gfortran intrinsic procedures?
> 
> There is nothing in PXF that requires the routines to be intrinsic.
> So I would suggest not.

That's good.  I haven't spent too much time reading the PXF docs
(other than to note that it shouldn't be hard to implement).

> > ...  This would allow someone (perhaps, Walt) to use their
> > own implementation of libpxf without worrying about whether gfortran
> > screws with the namespace.
> 
> Don't worry about me.  It is trivial to link my library ahead of any
> system supplied libraries.

Name mangling occurs before linking.  I suspect that you may actually
use an EXTERNAL statement in your code, so name mangling won't occur.
For example, the current behavior is

  program foo
    print *, iargc()  ! Name mangled to _gfortran_iargc_
  end program foo

  program bar
    external iargc
    print *, iargc()  ! Name not mangled; becomes iargc_
  end program bar

In bar, gfortran is expecting the programmer to supply an iargc
function.  In foo, gfortran assumes that iargc is an intrinsic
procedure.

-- 
Steve