This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug fortran/73142] Reading a namelist fails - my code worked until 4.8.2
- From: "kargl at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Wed, 10 Aug 2016 21:45:40 +0000
- Subject: [Bug fortran/73142] Reading a namelist fails - my code worked until 4.8.2
- Auto-submitted: auto-generated
- References: <bug-73142-4@http.gcc.gnu.org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=73142
--- Comment #3 from kargl at gcc dot gnu.org ---
(In reply to Günter Weber from comment #0)
>
> I am using the GCC package for Windows provided by www.equation.com. It
> looks like after version 4.8.2 a bug was introduced in the reading of
> namelists. The reading is rubbish when the data file has a certain format.
> But at least in Windows when you open the data file in a common editor you
> see identical content to a version of the data file, which can be read
> correctly.
The files are not identical. fileA.dat terminates lines
with a line with a LF. fileB.dat terminates lines with
a CRLF.
With gfortran 5.4.1 I can read fileA.dat. Reading fileB.dat
generates
At line 14 of file minimum_example.f (unit = 31, file = 'fileB.dat')
Fortran runtime error: Cannot match namelist object name 30.00
With gfortran 6.1.1, I get the same results as with 5.4.1, except
the built executable aborts with a backtrace.
With gfortran 7.0.0, both files can be read. So, the bug is
fixed on trunk. You can update to trunk or ensure your files
have the format of fileA.dat.