[Bug libfortran/26499] New: gfortran - End of File incorrectly positioned after binary I/O.

[Bug libfortran/26499] New: gfortran - End of File incorrectly positioned after binary I/O.

[dir at lanl dot gov]

The end of file is not correctly positioned by gfortran after this program
writes, closes and reopens a binary file (It works Ok with fort77) -

[dranta:~/tests/gfortran-D] dir% fort77 -o write27 write27.f
** test   === End of Compilation 1 ===
1501-510  Compilation successful for file write27.f.
[dranta:~/tests/gfortran-D] dir% write27
[dranta:~/tests/gfortran-D] dir% gfortran -o write27 write27.f
[dranta:~/tests/gfortran-D] dir% write27
  eof at wrong location            4           3
Abort
[dranta:~/tests/gfortran-D] dir% cat write27.f
      program test
      integer at,eof
      dimension idata(1011)
      open(unit=11,form='unformatted')
       write(11)idata
       write(11)idata
       write(11)idata
       backspace 11
       backspace 11
       write(11)idata
      close(11,status='keep')  
      eof=3
      open(unit=11,form='unformatted')
      rewind 11
      do at=1,eof+1
        iat=at
        read(11,end=250)idata
      enddo 
      write(6,*)' eof not read ',iat,eof 
      call abort()
      goto 300
 250  continue
      if(iat.ne.eof)then
        write(6,*)' eof at wrong location ',iat,eof 
        call abort     
      endif
 300  continue
      close(11,status='delete')  
      stop
      end

[dranta:~/tests/gfortran-D] dir% gfortran --v
Using built-in specs.
Target: powerpc-apple-darwin8.5.0
Configured with: ../gcc/configure --prefix=/Users/dir/gfortran
--enable-languages=c,f95
Thread model: posix
gcc version 4.2.0 20060228 (experimental)


-- 
           Summary: gfortran - End of File incorrectly positioned after
                    binary I/O.
           Product: gcc
           Version: 4.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libfortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: dir at lanl dot gov
  GCC host triplet: powerpc-apple-darwin8.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #1 from jvdelisle at gcc dot gnu dot org  2006-03-01 00:57 -------
Confirmed.


-- 

jvdelisle at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |jvdelisle at gcc dot gnu dot
                   |dot org                     |org
             Status|UNCONFIRMED                 |ASSIGNED
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2006-03-01 00:57:46
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #2 from jvdelisle at gcc dot gnu dot org  2006-03-01 07:19 -------
Take a closer look at this folks.  The program should jump to line 250 on the
fourth READ.  If you change the test line to: if(iat.ne.eof+1)then you will see
that gfortran is correctly jumping out on the fourth READ.  On the third READ
you have not hit the end-of-file yet, just almost.

Trying this slightly improved test case with ifort gives a segfault at run
time. 

Can I get a second opinion on this?  I think this is a NOT-A-BUG.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[dir at lanl dot gov]

------- Comment #3 from dir at lanl dot gov  2006-03-01 13:50 -------
Hi Jerry,

   The program does 
      3 writes      - eof after record 3 - write pointer after record 3
      2 backspaces  - eof after record 3 - write pointer after record 1
      1 write       - eof after record 2 - write pointer after record 2

      The two backspaces backup over the last two writes and the forth write
adds one additional record for a total of 2 records of the file - there should
only be 2 records with the eof jump happening on the third read.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at verizon dot net]

------- Comment #4 from jvdelisle at verizon dot net  2006-03-01 14:36 -------
Subject: Re:  gfortran - End of File incorrectly
 positioned after binary I/O.

dir at lanl dot gov wrote:
> ------- Comment #3 from dir at lanl dot gov  2006-03-01 13:50 -------
> Hi Jerry,
> 
>    The program does 
>       3 writes      - eof after record 3 - write pointer after record 3
>       2 backspaces  - eof after record 3 - write pointer after record 1
>       1 write       - eof after record 2 - write pointer after record 2
> 
>       The two backspaces backup over the last two writes and the forth write
> adds one additional record for a total of 2 records of the file - there should
> only be 2 records with the eof jump happening on the third read.
> 
> 
Hi Dale,

My "conceptual" problem is - does the last write truncate the file?  I remember 
reading about this somewhere, but it is vague in my mind.  Thats why when I 
started looking at this I just had to stop and ask.  I could see the difference 
in behavior between gfortran and g77. It puzzled me.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[dir at lanl dot gov]

------- Comment #5 from dir at lanl dot gov  2006-03-01 17:37 -------
With a sequential access file, the eof is always positioned after the location
of the last write. Thus, a file needs to be truncated to that location when it
is closed, otherwise you are saving data that is beyond the "end of file". On a
sequential tape, an eof mark seperates files and the eoi (end of information)
is usually two consecutive eof marks.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[patchapp at dberlin dot org]

------- Comment #6 from patchapp at dberlin dot org  2006-03-02 06:35 -------
Subject: Bug number PR26499

A patch for this bug has been added to the patch tracker.
The mailing list url for the patch is
http://gcc.gnu.org/ml/gcc-patches/2006-03/msg00115.html


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[dir at lanl dot gov]

------- Comment #7 from dir at lanl dot gov  2006-03-02 14:02 -------
    Hi Jerry,

    As you may have guessed, I added this problem to the things that my program
looks for. You got that one and all the ones like it, but here is another one
from a slightly different class (rewind after reading a eof instead of a
backspace after reading a eof) -

[dranta:~/tests/gfortran-D] dir% gfortran -o write28 write28.f
[dranta:~/tests/gfortran-D] dir% write28
Abort
[dranta:~/tests/gfortran-D] dir% cat write28.f
      program test
      dimension idata(1011)
      open(unit=11,form='unformatted')
       write(11)idata
       write(11)idata
       read(11,end=        1000 )idata
       call abort()
 1000  continue
       rewind 11
       write(11)idata
       close(11,status='keep')        
      open(unit=11,form='unformatted')
      rewind 11
      read(11)idata
      read(11, end=250)idata
      call abort()
 250  continue
      close(11,status='delete')  
      stop
      end


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #8 from jvdelisle at gcc dot gnu dot org  2006-03-03 01:51 -------
Add a flush just before the truncate in st_write_done and comment seven is
fixed.    I had it in there at first, then I took it out to see if things would
still work, and they did with the cases I had.  A revised patch has been
submitted for review.  Dale, I also copied you on this new patch.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[dir at lanl dot gov]

------- Comment #9 from dir at lanl dot gov  2006-03-03 14:51 -------
That got the one in comment #7. Here is the next one -
[dranta:~/tests/gfortran-D] dir% g77 -o write29 write29.f
[dranta:~/tests/gfortran-D] dir% write29
[dranta:~/tests/gfortran-D] dir% gfortran -o write29 write29.f
[dranta:~/tests/gfortran-D] dir% write29
Abort
[dranta:~/tests/gfortran-D] dir% cat write29.f
      program test
      dimension idata(1011)
      open(unit=11,form='unformatted')
        idata(1)   = -705
        idata(  1011)   = -706
       write(11)idata
        idata(1)   = -706
        idata(  1011)   = -707
       write(11)idata
        idata(1)   = -707
        idata(  1011)   = -708
       write(11)idata
       read(11,end=        1000 )idata
       call abort()
 1000  continue
       rewind 11
       read(11,end=        1001 )idata
        if(idata(1).ne. -705.or.idata(  1011).ne. -706)call abort()
 1001  continue
       close(11,status='keep')        
      open(unit=11,form='unformatted')
      rewind 11
      read(11)idata
      if(idata(1).ne.-705)then
      call abort()
      endif
      read(11)idata
      if(idata(1).ne.-706)then
      call abort()
      endif
      read(11)idata
      if(idata(1).ne.-707)then
      call abort()
      endif
      close(11,status='delete')  
      stop
      end


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #10 from jvdelisle at gcc dot gnu dot org  2006-03-03 23:05 -------
The case in Comment #9 is also very simple.  I got myself again.  I saw this
before, but because I did not have a test case that failed, I left it alone. 
With this fix in file_pos.c all of Dale's rewind tests pass up to number of
operations equals 9 (262144 tests).  I will submit the updated patch soon.

In st_rewind change this:


          if (u->mode == WRITING)
            {
              flush (u->s);
              struncate (u->s);
            }
          u->mode = READING;

to this:
          flush (u->s);
          if (u->mode == WRITING)
            struncate (u->s);
          u->mode = READING;


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #11 from jvdelisle at gcc dot gnu dot org  2006-03-05 06:20 -------
Final patch here:

http://gcc.gnu.org/ml/fortran/2006-03/msg00046.html

Waiting for OK to commit.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[dir at lanl dot gov]

------- Comment #12 from dir at lanl dot gov  2006-03-06 14:06 -------
It works great on the Macintosh. Now, if someone could just get the windows
version of gfortran under MinGW to pass these I/O tests, it might become
usuable. My programs compile under MinGW, but they all crash when I try to run
them - mostly with I/O problems.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #13 from jvdelisle at gcc dot gnu dot org  2006-03-10 03:15 -------
Subject: Bug 26499

Author: jvdelisle
Date: Fri Mar 10 03:15:36 2006
New Revision: 111924

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=111924
Log:
2006-03-09  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

        PR libgfortran/26499
        * io/file_pos (st_rewind): Flush always.
        * io/unix.c (fd_truncate): Return SUCCESS rather than FAILURE for
        special files like /dev/null.
        * io/transfer.c (st_write_done): Remove broken logic that prevented
        calling fd_truncate.

Modified:
    trunk/libgfortran/ChangeLog
    trunk/libgfortran/io/file_pos.c
    trunk/libgfortran/io/transfer.c
    trunk/libgfortran/io/unix.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #14 from jvdelisle at gcc dot gnu dot org  2006-03-10 03:23 -------
Subject: Bug 26499

Author: jvdelisle
Date: Fri Mar 10 03:23:28 2006
New Revision: 111925

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=111925
Log:
2006-03-09  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

        PR libgfortran/26499
        * gfortran.dg/write_back.f: New test.
        * gfortran.dg/write_rewind_1.f: New test.
        * gfortran.dg/write_rewind_2.f: New test.

Added:
    trunk/gcc/testsuite/gfortran.dg/write_back.f
    trunk/gcc/testsuite/gfortran.dg/write_rewind_1.f
    trunk/gcc/testsuite/gfortran.dg/write_rewind_2.f
Modified:
    trunk/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #15 from jvdelisle at gcc dot gnu dot org  2006-03-15 05:34 -------
Subject: Bug 26499

Author: jvdelisle
Date: Wed Mar 15 05:34:05 2006
New Revision: 112076

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=112076
Log:
2006-03-14  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

        PR libgfortran/26499
        * io/file_pos (st_rewind): Flush always.
        * io/unix.c (fd_truncate): Return SUCCESS rather than FAILURE for
        special files like /dev/null.
        * io/transfer.c (st_write_done): Remove broken logic that
        prevented calling fd_truncate.

Modified:
    branches/gcc-4_1-branch/libgfortran/ChangeLog
    branches/gcc-4_1-branch/libgfortran/io/file_pos.c
    branches/gcc-4_1-branch/libgfortran/io/transfer.c
    branches/gcc-4_1-branch/libgfortran/io/unix.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #16 from jvdelisle at gcc dot gnu dot org  2006-03-15 05:40 -------
Subject: Bug 26499

Author: jvdelisle
Date: Wed Mar 15 05:40:20 2006
New Revision: 112077

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=112077
Log:
2006-03-14  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

        PR libgfortran/26499
        * gfortran.dg/write_back.f: New test.
        * gfortran.dg/write_rewind_1.f: New test.
        * gfortran.dg/write_rewind_2.f: New test.

Added:
    branches/gcc-4_1-branch/gcc/testsuite/gfortran.dg/write_back.f
    branches/gcc-4_1-branch/gcc/testsuite/gfortran.dg/write_rewind_1.f
    branches/gcc-4_1-branch/gcc/testsuite/gfortran.dg/write_rewind_2.f
Modified:
    branches/gcc-4_1-branch/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499

[Bug libfortran/26499] gfortran - End of File incorrectly positioned after binary I/O.

[jvdelisle at gcc dot gnu dot org]

------- Comment #17 from jvdelisle at gcc dot gnu dot org  2006-03-15 07:10 -------
Fixed on 4.2 before, now fixed on 4.1.1 as well.


-- 

jvdelisle at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.1.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26499