Re: Problem with gfortran and named pipes under unix

[Jerry DeLisle <jvdelisle at verizon dot net>]

Jerry DeLisle wrote:
> Bud Davis wrote:
> > --- Johannes Lotz <johannes.lotz@rwth-aachen.de>
> > wrote:
> >
> > > Hello.
> > >
> > > I have a problem using gfortran for a code dealing
> > > with named pipes.
>
>
> > Which looks to me like an error both on the writing
> > and the reading.
> >
> > Could you please try the above example, and if
> > possible simplify it a bit more...if this demonstrates
> > your problem, then file a bugzilla report.
> I have tried this here and can confirm the hangups.  We are going to 
> have to study this a bit.  I have not used named pipes myself much.
>
> Jerry
This snippet id from libgfortran/io/unix.c

  /* non-seekable files, like terminals and fifo's fail the lseek.
     Using ftruncate on a seekable special file (like /dev/null)
     is undefined, so we treat it as if the ftruncate succeeded.
  */
#ifdef HAVE_FTRUNCATE
  if (s->special_file || ftruncate (s->fd, s->logical_offset))
#else
#ifdef HAVE_CHSIZE
  if (s->special_file || chsize (s->fd, s->logical_offset))
#endif
#endif
    {
      s->physical_offset = s->file_length = 0;
      return SUCCESS;
    }

Further examination of unix.c reveals lseek being used in fd_flush, 
fd_alloc_r_at, fd_alloc_w_at.  We do have a helper function to test that a 
stream is seekable.  I think we need to incorporate this into those functions so 
that we handle this correctly.  If a stream is not seekable, then just allocate 
the next portion of the stream the number of bytes requested.

I opened PR 30162 for this.

Jerry