source: CIVL/examples/omp/dataracebench-1.3.2/micro-benchmarks-fortran/DRB098-simd2-orig-no.f95

!!!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~!!!
!!! Copyright (c) 2017-20, Lawrence Livermore National Security, LLC
!!! and DataRaceBench project contributors. See the DataRaceBench/COPYRIGHT file for details.
!!!
!!! SPDX-License-Identifier: (BSD-3-Clause)
!!!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~!!!

!Two-dimension array computation with a vetorization directive
!collapse(2) makes simd associate with 2 loops.
!Loop iteration variables should be predetermined as lastprivate. No data race pairs.

program DRB098_simd2_orig_no
    use omp_lib
    implicit none

    integer, parameter :: dp = kind(1.0d0)
    real(dp), dimension(:,:), allocatable :: a,b,c
    integer :: len, i, j

    len = 100
    allocate (a(len,len))
    allocate (b(len,len))
    allocate (c(len,len))

    do i = 1, len
        do j = 1, len
            a(i,j) = real(i,dp)/2.0
            b(i,j) = real(i,dp)/3.0
            c(i,j) = real(i,dp)/7.0
        end do
    end do

    !$omp simd collapse(2)
    do i = 1, len
        do j = 1, len
            c(i,j)=a(i,j)*b(i,j)
        end do
    end do
    !$omp end simd

    print*,'c(50,50) =',c(50,50)

    deallocate(a,b,c)
end program