With a simple data file:
$ cat file.dat
1
2
3
$ wc file.dat
3 3 6 file.dat
the following code:
$ cat aa.f90
program fred
implicit none
integer::i,badness
character::c
open(unit=10,file='file.dat')
do i=1,huge(1)
read(10,*,iostat=badness)
write(*,*)i,badness
if (badness/=0) exit
enddo
write(*,*)i
end
delays the nonzero iostat return with 4.3.0:
$ gfortran --version
GNU Fortran (GCC) 4.3.0 20071221 (experimental)
Copyright (C) 2007 Free Software Foundation, Inc.
GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
You may redistribute copies of GNU Fortran
under the terms of the GNU General Public License.
For more information about these matters, see the file named COPYING
$ gfortran -o aa aa.f90
$ ./aa
1 0
2 0
3 0
4 0
5 -1
5
That is, the forth read completes apparently successfully, even though we've
read all three lines previously.
4.2.2 does what's expected:
$ gfortran --version
GNU Fortran (GCC) 4.2.2
Copyright (C) 2007 Free Software Foundation, Inc.
GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
You may redistribute copies of GNU Fortran
under the terms of the GNU General Public License.
For more information about these matters, see the file named COPYING
$ gfortran -o aa aa.f90
$ ./aa
1 0
2 0
3 0
4 -1
4
If the read statement reads into something [ie line 7 of aa.f90 is changed to
read(10,*,iostat=badness)c] then iostat returns -1 on the forth read with
4.3.0, as expected.
(For those that insist on looking over their shoulder, ifort 10.1 and pgf90 7.0
return badness as -1 on the forth call, like gfortran used to.)
--
Summary: IO error delayed
Product: gcc
Version: 4.3.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: fortran
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: terry at chem dot gu dot se
GCC host triplet: x86_64-unknown-linux-gnu
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34676
