Re: fortran bug posted to gcc.gnu.org

[Jerry DeLisle <jvdelisle at frontier dot com>]

On 12/31/2010 09:40 AM, Steve Kargl wrote:
> On Fri, Dec 31, 2010 at 03:58:56PM +0000, Davis, Jimmie wrote:
> > has anyone answered this yet ?
> >
> > http://gcc.gnu.org/ml/gcc/2010-12/msg00472.html
>
> I haven't.  The poster should send email to fortran@
> or open a PR.  The poster did not send the code and
> data files that show the problem, so there is nothing to
> debug.  In fact, the poster seems to have that
> "1.2345<EOF>" in read statement is invalid for input,
> yet feels that gfortran should ignore the problem
> and follow his END=.  The real problem is his code
> code should use ERR=.
Jimmie, thanks for pointing this out.  I did not see it.

Well I think Steve is probably right. END= is distinctly different from ERR=. 
Also, if gfortran sees a CR by itself on a UNIX system, it is likely to not work 
because the usual end of record marker is an LF.

I have attempted to copy Lawrence here.

Lawrence, please send us a small code snippet that causes the issue you see and 
the resulting output.  We will see if we can help you with this. (Note 
fortraners monitor fortran@gcc.gnu.org)

Regards,

Jerry