[Bug libfortran/21621] New: Inconsistency with binary sequential output

[Bug libfortran/21621] New: Inconsistency with binary sequential output

[P dot Schaffnit at access dot rwth-aachen dot de]

Hi!

I think I've found something strange with binary data written SEQUENTIALly...

I was having trouble reading binary data written by gfortran, so I tried to read
that DIRECTly, and... it doesn't to work! (why?)

Here is a small example in f90, and (for the record the same thing in f77, which
worked with g77).

Can someone help me with that?

Thanks!

Philippe

PS: the f90 sources:
!
      PROGRAM Check_Record_Length
!
      IMPLICIT NONE
!
      INTEGER                           :: IO_Stat
      INTEGER, PARAMETER                :: Unit_Number = 13
      INTEGER, PARAMETER                :: Record_Length = 16
!       INTEGER, PARAMETER                :: Record_Length = 20
      INTEGER                           :: Record_Length_1
      INTEGER                           :: Record_Length_2
      INTEGER                           :: Integer_Variable
!
      REAL                              :: Real_Variable
      DOUBLE PRECISION                  :: Double_Variable
!
      CHARACTER ( LEN = * ), PARAMETER  :: Name = "Binary_Data"
!
!
! Write 'sequential'
      OPEN ( UNIT = Unit_Number, ACTION = "WRITE", FILE = Name,        &
     &                     FORM = "UNFORMATTED", ACCESS = "SEQUENTIAL" )
!
      Real_Variable    = 666
      Integer_Variable = 13
      WRITE ( UNIT = Unit_Number, IOSTAT = IO_Stat )                   &
     &                                   Real_Variable, Integer_Variable
!
      CLOSE ( UNIT = Unit_Number )
!
!
! Read 'sequential'
      OPEN ( UNIT = Unit_Number, ACTION = "READ", FILE = Name,         &
     &                     FORM = "UNFORMATTED", ACCESS = "SEQUENTIAL" )
!
      READ ( UNIT = Unit_Number, IOSTAT = IO_Stat )                    &
     &                                   Real_Variable, Integer_Variable
!
      WRITE ( 6, "(/G13.5,I13,2I13/)" ) Real_Variable, Integer_Variable
!
      CLOSE ( UNIT = Unit_Number )
!
!
! Read 'direct'
      OPEN ( UNIT = Unit_Number, ACTION = "READ", FILE = Name,         &
     &   FORM = "UNFORMATTED", ACCESS = "DIRECT", RECL = Record_Length )
!
      READ ( UNIT = Unit_Number, REC = 1 )                             &
     & Record_Length_1, Real_Variable, Integer_Variable, Record_Length_2
!      & Record_Length_1, Double_Variable, Integer_Variable, Record_Length_2
!       Real_Variable = Double_Variable
!
      WRITE ( 6, "(/G13.5,I13,2I13/)" ) Real_Variable,                 &
     &                Integer_Variable, Record_Length_1, Record_Length_2
!
      CLOSE ( UNIT = Unit_Number )
!
!
      END PROGRAM Check_Record_Length
!


PPS: the f77 ones:
!
      PROGRAM Check_Record_Length
!
      IMPLICIT NONE
!
      INTEGER IO_Stat
      INTEGER Unit_Number
      PARAMETER ( Unit_Number = 13 )
      INTEGER Rec_Length
      PARAMETER ( Rec_Length = 16 )
      INTEGER Record_Length, Record_Length_1, Record_Length_2
      INTEGER Integer_Variable
!
      REAL Real_Variable
      DOUBLE PRECISION Double_Variable
!
      CHARACTER ( 11 ) Name
      PARAMETER ( Name = "Binary_Data" )
!
!
! Write 'sequential'
      OPEN ( UNIT = Unit_Number, FILE = Name,
     &                     FORM = "UNFORMATTED", ACCESS = "SEQUENTIAL" )
!
      Real_Variable    = 666
      Integer_Variable = 13
      WRITE ( UNIT = Unit_Number, IOSTAT = IO_Stat )
     &                                   Real_Variable, Integer_Variable
!
      CLOSE ( UNIT = Unit_Number )
!
!
! Read 'sequential'
      OPEN ( UNIT = Unit_Number, FILE = Name,
     &                     FORM = "UNFORMATTED", ACCESS = "SEQUENTIAL" )
!
      READ ( UNIT = Unit_Number, IOSTAT = IO_Stat )
     &                                   Real_Variable, Integer_Variable
!
      WRITE ( 6, "(/G13.5,I13,2I13/)" ) Real_Variable, Integer_Variable
!
      CLOSE ( UNIT = Unit_Number )
!
!
! Read 'direct'
      OPEN ( UNIT = Unit_Number, FILE = Name,
     &     FORM = "UNFORMATTED", ACCESS = "DIRECT", RECL = Rec_Length )
!
      READ ( UNIT = Unit_Number, REC = 1 )
     & Record_Length_1, Real_Variable, Integer_Variable, Record_Length_2
!
      WRITE ( 6, "(/G13.5,I13,2I13/)" ) Real_Variable,
     &                Integer_Variable, Record_Length_1, Record_Length_2
!
      CLOSE ( UNIT = Unit_Number )
!
!
      END PROGRAM Check_Record_Length
!

-- 
           Summary: Inconsistency with binary sequential output
           Product: gcc
           Version: 4.1.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: libfortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: P dot Schaffnit at access dot rwth-aachen dot de
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: GCC 4.1.0 20050517
  GCC host triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621

[Bug libfortran/21621] Inconsistency with binary sequential output

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
  GCC build triplet|GCC 4.1.0 20050517          |
   GCC host triplet|i686-pc-linux-gnu           |
 GCC target triplet|                            |i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621

[Bug libfortran/21621] Inconsistency with binary sequential output

[tobi at gcc dot gnu dot org]

------- Additional Comments From tobi at gcc dot gnu dot org  2005-05-19 16:43 -------
Can you be a little more specific on what you're trying to do and why you think
that what gfortran does is wrong?

I don't think the standard requires sequential access files to be in any
reliable format.  I find the results surprising, but I'm not yet sure it's a bug.

[tobi@marktplatz tests]$ gfortran pr21621.f
[tobi@marktplatz tests]$ ./a.out

   666.00               13

   0.0000       1143373824            8           13

[tobi@marktplatz tests]$ g77 pr21621.f
[tobi@marktplatz tests]$ ./a.out

   666.00               13

   666.00               13            8            8

[tobi@marktplatz tests]$


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tobi at gcc dot gnu dot org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2005-05-19 16:43:58
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621

[Bug libfortran/21621] Inconsistency with binary sequential output

[tobi at gcc dot gnu dot org]

------- Additional Comments From tobi at gcc dot gnu dot org  2005-05-19 16:44 -------
Um, I meant to set this to waiting ...

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621

[Bug libfortran/21621] Inconsistency with binary sequential output

[P dot Schaffnit at access dot rwth-aachen dot de]

------- Additional Comments From P dot Schaffnit at access dot rwth-aachen dot de  2005-05-23 09:05 -------

Hi!

I was trying to read some binary data written by a GFortran binary with binaries
written with other compilers (LF, SGI), and I couldn't get anywhere, so I just
started to poke around to see where it came from...

I tried to have a look at the Fortran standard, but I didn't get anywhere, but
my experience with several compilers is that they do use this REC-LENGTH thing,
so (and g77 did also!), so I would say it would be a pity to have trouble
reading old binary data from g77 with gfortran...

It's actually a rather self-centered comment, but maybe I'm not the only one...

Cheers!

Philippe

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621

[Bug libfortran/21621] Inconsistency with binary sequential output

[jblomqvi at cc dot hut dot fi]

------- Additional Comments From jblomqvi at cc dot hut dot fi  2005-05-24 20:02 -------
Well no, the standard does not specify how unformatted sequential record markers
are implemented. In short, gfortran uses markers of type off_t, which is 64 bits
on operating systems with large file (LFS) support. Today, that is most OS:es in
common use. g77 AFAIK uses markers of type long, that is 32-bits on most 32-bit
platforms and 64 bits on 64-bit platforms. See also PR 19303. Also note that Bud
Davis is apparently currently working on pluggable record markers, which should
allow gfortran to add some kind of g77 compatibility flag.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621

[Bug libfortran/21621] Inconsistency with binary sequential output

[bdavis9659 at comcast dot net]

------- Additional Comments From bdavis9659 at comcast dot net  2005-05-31 02:36 -------
proposed patch here:

http://gcc.gnu.org/ml/fortran/2005-05/msg00431.html

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621

[Bug libfortran/21621] Inconsistency with binary sequential output

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-05-31 02:45 -------


*** This bug has been marked as a duplicate of 19303 ***

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |RESOLVED
         Resolution|                            |DUPLICATE


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21621