[Bug fortran/27446] New: ICE on OpenMP array REDUCTION clause and -fbounds-check

[benjamin dot reveille at gmail dot com]

This bug is linked to PR27395 because the same 2 fortran test cases are the
ones triggering the problem. (The test cases can be found further bellow or as
their equivalents PR27395_1.f90 and PR27395_2.f90 in the libgomp dejagnu
testsuite)

Once PR27395 was corrected (Thanks Jakub) I played around with compile options.

adding the -fbounds-check options triggers the following ICE for both testcases

>> gfortran -v
Using built-in specs.
Target: i386-linux
Configured with: ../gcc/configure
--prefix=/cosmic/coudert/tmp/gfortran-20060504/irun
--enable-languages=c,fortran --host=i386-linux
--with-gmp=/cosmic/coudert/tmp/gfortran-20060504/gfortran_libs
Thread model: posix
gcc version 4.2.0 20060504 (experimental)

 >> gfortran -fopenmp -fbounds-check array_reduction1.f90
array_reduction1.f90: In function 'foo1':
array_reduction1.f90:24: internal compiler error: in declare_tmp_vars, at
gimplify.c:661
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

 >> gfortran -fopenmp -fbounds-check array_reduction2.f90
array_reduction2.f90: In function 'foo2':
array_reduction2.f90:20: internal compiler error: in declare_tmp_vars, at
gimplify.c:661
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.


 >> more array_reduction1
program array_reduction1
   implicit none
   integer, parameter :: n=10,m=1000
   integer :: i
   integer, dimension(n) :: sumarray
   !
   call foo1(n,m,sumarray)
   do i=1,n
      print*,sumarray(i)
   end do
end program array_reduction1

subroutine foo1(n,m,sumarray)
   use omp_lib, only : omp_get_thread_num
   implicit none
   integer, intent(in) :: n,m
   integer, dimension(n), intent(out) :: sumarray
   !
   integer :: i,j
   !
   sumarray(:)=0
!$OMP PARALLEL DEFAULT(shared)
   print*,'hello from thread ',omp_get_thread_num()
!$OMP DO PRIVATE(j,i), REDUCTION(+:sumarray)
   do j=1,m
      do i=1,n
         sumarray(i)=sumarray(i)+i
      end do
   end do
!$OMP END DO
!$OMP END PARALLEL
end subroutine foo
-----------------------------------------------------------

 >> more array_reduction2.f90
program array_reduction2
   implicit none
   integer, parameter :: n=10,m=1000
   integer :: i
   !
   call foo2(n,m)
end program array_reduction2

subroutine foo2(n,m)
   use omp_lib, only : omp_get_thread_num
   implicit none
   integer, intent(in) :: n,m
   !
   integer :: i,j
   integer, dimension(n) :: sumarray
   !
   sumarray(:)=0
!$OMP PARALLEL DEFAULT(shared)
   print*,'hello from thread ',omp_get_thread_num()
!$OMP DO PRIVATE(j,i), REDUCTION(+:sumarray)
   do j=1,m
      do i=1,n
         sumarray(i)=sumarray(i)+i
      end do
   end do
!$OMP END DO
!$OMP END PARALLEL
   do i=1,n
      print*,sumarray(i)
   end do
end subroutine foo2




Routines are added again here for completeness


-- 
           Summary: ICE on OpenMP array REDUCTION clause and -fbounds-check
           Product: gcc
           Version: 4.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: benjamin dot reveille at gmail dot com
 GCC build triplet: i386-gnu-linux
  GCC host triplet: i386-gnu-linux
GCC target triplet: i386-gnu-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27446