source: CIVL/examples/omp/dataracebench-1.3.2/micro-benchmarks-fortran/DRB121-reduction-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)
!!!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~!!!

!Number of threads is empirical: We need enough threads so that
!the reduction is really performed hierarchically in the barrier!
!There is no data race.

program DRB121_reduction_orig_no
    use omp_lib
    implicit none

    integer :: var, i, sum1, sum2

    var = 0
    sum1 = 0
    sum2 = 0

    !$omp parallel reduction(+: var)
        !$omp do schedule(static) reduction(+: sum1)
        do i = 1, 5
            sum1 = sum1+i
        end do
        !$omp end do

        !$omp do schedule(static) reduction(+: sum2)
        do i = 1, 5
            sum2 = sum2+i
        end do
        !$omp end do

        var = sum1 + sum2
    !$omp end parallel

    print 100, var
    100 format ('var =',3i8)
end program