[Bug fortran/37523] New: data syncronisation error with openmp

[Bug fortran/37523] New: data syncronisation error with openmp

[j at jasper dot es]

When running the attached code (compiled with -fopenmp) I see intermittent
'WUH' in my output, which should not happen. Wasting some cycles between the
initialization of the array and the start of the DO loop (by printing some
data) works.

jbook:optidus-tng spaans$ LANG=C gfortran -v
Using built-in specs.
Target: i386-apple-darwin9.4.0
Configured with: ../gcc/configure --prefix=/usr/local/gfortran/
--with-gmp=/opt/local --with-mpfr=/opt/local --enable-languages=c,fortran
--no-create --no-recursion
Thread model: posix
gcc version 4.4.0 20080909 (experimental) (GCC) 
jbook:optidus-tng spaans$ gfortran -fopenmp gfortranbug-testcase.f90 -o
gfortranbug-testcase
jbook:optidus-tng spaans$ for a in `gseq 1 1000`; do ./gfortranbug-testcase  ;
done |grep WUH
STOP WUH!
...


-- 
           Summary: data syncronisation error with openmp
           Product: gcc
           Version: 4.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: j at jasper dot es
 GCC build triplet: i386-apple-darwin9.4.0
  GCC host triplet: i386-apple-darwin9.4.0
GCC target triplet: i386-apple-darwin9.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37523

[Bug fortran/37523] data syncronisation error with openmp

[j at jasper dot es]

------- Comment #1 from j at jasper dot es  2008-09-15 08:45 -------
Created an attachment (id=16321)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16321&action=view)
source code for testcase


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37523

[Bug fortran/37523] data syncronisation error with openmp

[j at jasper dot es]

------- Comment #2 from j at jasper dot es  2008-09-15 08:52 -------
Whoops, sorry, this was an error on my side, as the test case was dumbed down
too much. Will try to make a better and valid testcase, as I have something
like this occurring to me.


-- 

j at jasper dot es changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37523