SUBROUTINE transpose_xy( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Current revisions: ! ----------------- ! ! ! Former revisions: ! ----------------- ! $Id: transpose.f90 684 2011-02-09 14:49:31Z banzhafs $ ! ! 683 2011-02-09 14:25:15Z raasch ! openMP parallelization of transpositions for 2d-domain-decomposition ! ! 622 2010-12-10 08:08:13Z raasch ! optional barriers included in order to speed up collective operations ! ! 164 2008-05-15 08:46:15Z raasch ! f_inv changed from subroutine argument to automatic array in order to do ! re-ordering from f_in to f_inv in one step, one array work is needed instead ! of work1 and work2 ! ! February 2007 ! RCS Log replace by Id keyword, revision history cleaned up ! ! Revision 1.2 2004/04/30 13:12:17 raasch ! Switched from mpi_alltoallv to the simpler mpi_alltoall, ! all former transpose-routine files collected in this file, enlarged ! transposition arrays introduced ! ! Revision 1.1 2004/04/30 13:08:16 raasch ! Initial revision (collection of former routines transpose_xy, transpose_xz, ! transpose_yx, transpose_yz, transpose_zx, transpose_zy) ! ! Revision 1.1 1997/07/24 11:25:18 raasch ! Initial revision ! ! ! Description: ! ------------ ! Transposition of input array (f_in) from x to y. For the input array, all ! elements along x reside on the same PE, while after transposition, all ! elements along y reside on the same PE. !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, ys REAL :: f_in(0:nxa,nys_x:nyn_xa,nzb_x:nzt_xa), & f_inv(nys_x:nyn_xa,nzb_x:nzt_xa,0:nxa), & f_out(0:nya,nxl_y:nxr_ya,nzb_y:nzt_ya), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- Rearrange indices of input array in order to make data to be send !-- by MPI contiguous !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO i = 0, nxa DO k = nzb_x, nzt_xa DO j = nys_x, nyn_xa f_inv(j,k,i) = f_in(i,j,k) ENDDO ENDDO ENDDO !$OMP END PARALLEL ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( f_inv(nys_x,nzb_x,0), sendrecvcount_xy, MPI_REAL, & work(1), sendrecvcount_xy, MPI_REAL, & comm1dy, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array !$OMP PARALLEL PRIVATE ( i, j, k, l, m, ys ) !$OMP DO DO l = 0, pdims(2) - 1 m = l * ( nxr_ya - nxl_y + 1 ) * ( nzt_ya - nzb_y + 1 ) * & ( nyn_xa - nys_x + 1 ) ys = 0 + l * ( nyn_xa - nys_x + 1 ) DO i = nxl_y, nxr_ya DO k = nzb_y, nzt_ya DO j = ys, ys + nyn_xa - nys_x m = m + 1 f_out(j,i,k) = work(m) ENDDO ENDDO ENDDO ENDDO !$OMP END PARALLEL #endif END SUBROUTINE transpose_xy SUBROUTINE transpose_xz( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Description: ! ------------ ! Transposition of input array (f_in) from x to z. For the input array, all ! elements along x reside on the same PE, while after transposition, all ! elements along z reside on the same PE. !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, xs REAL :: f_in(0:nxa,nys_x:nyn_xa,nzb_x:nzt_xa), & f_inv(nys:nyna,nxl:nxra,1:nza), & f_out(1:nza,nys:nyna,nxl:nxra), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- If the PE grid is one-dimensional along y, the array has only to be !-- reordered locally and therefore no transposition has to be done. IF ( pdims(1) /= 1 ) THEN ! !-- Reorder input array for transposition !$OMP PARALLEL PRIVATE ( i, j, k, l, m, xs ) !$OMP DO DO l = 0, pdims(1) - 1 m = l * ( nzt_xa - nzb_x + 1 ) * nnx * ( nyn_xa - nys_x + 1 ) xs = 0 + l * nnx DO k = nzb_x, nzt_xa DO i = xs, xs + nnx - 1 DO j = nys_x, nyn_xa m = m + 1 work(m) = f_in(i,j,k) ENDDO ENDDO ENDDO ENDDO !$OMP END PARALLEL ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( work(1), sendrecvcount_zx, MPI_REAL, & f_inv(nys,nxl,1), sendrecvcount_zx, MPI_REAL, & comm1dx, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array in a way that the z index is in first position !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO k = 1, nza DO i = nxl, nxra DO j = nys, nyna f_out(k,j,i) = f_inv(j,i,k) ENDDO ENDDO ENDDO !$OMP END PARALLEL ELSE ! !-- Reorder the array in a way that the z index is in first position !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO i = nxl, nxra DO j = nys, nyna DO k = 1, nza f_inv(j,i,k) = f_in(i,j,k) ENDDO ENDDO ENDDO !$OMP END PARALLEL !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO k = 1, nza DO i = nxl, nxra DO j = nys, nyna f_out(k,j,i) = f_inv(j,i,k) ENDDO ENDDO ENDDO !$OMP END PARALLEL ENDIF #endif END SUBROUTINE transpose_xz SUBROUTINE transpose_yx( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Description: ! ------------ ! Transposition of input array (f_in) from y to x. For the input array, all ! elements along y reside on the same PE, while after transposition, all ! elements along x reside on the same PE. !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, ys REAL :: f_in(0:nya,nxl_y:nxr_ya,nzb_y:nzt_ya), & f_inv(nys_x:nyn_xa,nzb_x:nzt_xa,0:nxa), & f_out(0:nxa,nys_x:nyn_xa,nzb_x:nzt_xa), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- Reorder input array for transposition !$OMP PARALLEL PRIVATE ( i, j, k, l, m, ys ) !$OMP DO DO l = 0, pdims(2) - 1 m = l * ( nxr_ya - nxl_y + 1 ) * ( nzt_ya - nzb_y + 1 ) * & ( nyn_xa - nys_x + 1 ) ys = 0 + l * ( nyn_xa - nys_x + 1 ) DO i = nxl_y, nxr_ya DO k = nzb_y, nzt_ya DO j = ys, ys + nyn_xa - nys_x m = m + 1 work(m) = f_in(j,i,k) ENDDO ENDDO ENDDO ENDDO !$OMP END PARALLEL ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( work(1), sendrecvcount_xy, MPI_REAL, & f_inv(nys_x,nzb_x,0), sendrecvcount_xy, MPI_REAL, & comm1dy, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array in a way that the x index is in first position !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO i = 0, nxa DO k = nzb_x, nzt_xa DO j = nys_x, nyn_xa f_out(i,j,k) = f_inv(j,k,i) ENDDO ENDDO ENDDO !$OMP END PARALLEL #endif END SUBROUTINE transpose_yx SUBROUTINE transpose_yxd( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Description: ! ------------ ! Transposition of input array (f_in) from y to x. For the input array, all ! elements along y reside on the same PE, while after transposition, all ! elements along x reside on the same PE. ! This is a direct transposition for arrays with indices in regular order ! (k,j,i) (cf. transpose_yx). !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, recvcount_yx, sendcount_yx, xs REAL :: f_in(1:nza,nys:nyna,nxl:nxra), f_inv(nxl:nxra,1:nza,nys:nyna), & f_out(0:nxa,nys_x:nyn_xa,nzb_x:nzt_xa), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- Rearrange indices of input array in order to make data to be send !-- by MPI contiguous DO k = 1, nza DO j = nys, nyna DO i = nxl, nxra f_inv(i,k,j) = f_in(k,j,i) ENDDO ENDDO ENDDO ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( f_inv(nxl,1,nys), sendrecvcount_xy, MPI_REAL, & work(1), sendrecvcount_xy, MPI_REAL, & comm1dx, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array m = 0 DO l = 0, pdims(1) - 1 xs = 0 + l * nnx DO j = nys_x, nyn_xa DO k = 1, nza DO i = xs, xs + nnx - 1 m = m + 1 f_out(i,j,k) = work(m) ENDDO ENDDO ENDDO ENDDO #endif END SUBROUTINE transpose_yxd SUBROUTINE transpose_yz( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Description: ! ------------ ! Transposition of input array (f_in) from y to z. For the input array, all ! elements along y reside on the same PE, while after transposition, all ! elements along z reside on the same PE. !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, zs REAL :: f_in(0:nya,nxl_y:nxr_ya,nzb_y:nzt_ya), & f_inv(nxl_y:nxr_ya,nzb_y:nzt_ya,0:nya), & f_out(nxl_z:nxr_za,nys_z:nyn_za,1:nza), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- Rearrange indices of input array in order to make data to be send !-- by MPI contiguous !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO j = 0, nya DO k = nzb_y, nzt_ya DO i = nxl_y, nxr_ya f_inv(i,k,j) = f_in(j,i,k) ENDDO ENDDO ENDDO !$OMP END PARALLEL ! !-- Move data to different array, because memory location of work1 is !-- needed further below (work1 = work2). !-- If the PE grid is one-dimensional along y, only local reordering !-- of the data is necessary and no transposition has to be done. IF ( pdims(1) == 1 ) THEN !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO j = 0, nya DO k = nzb_y, nzt_ya DO i = nxl_y, nxr_ya f_out(i,j,k) = f_inv(i,k,j) ENDDO ENDDO ENDDO !$OMP END PARALLEL RETURN ENDIF ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( f_inv(nxl_y,nzb_y,0), sendrecvcount_yz, MPI_REAL, & work(1), sendrecvcount_yz, MPI_REAL, & comm1dx, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array !$OMP PARALLEL PRIVATE ( i, j, k, l, m, zs ) !$OMP DO DO l = 0, pdims(1) - 1 m = l * ( nyn_za - nys_z + 1 ) * ( nzt_ya - nzb_y + 1 ) * & ( nxr_za - nxl_z + 1 ) zs = 1 + l * ( nzt_ya - nzb_y + 1 ) DO j = nys_z, nyn_za DO k = zs, zs + nzt_ya - nzb_y DO i = nxl_z, nxr_za m = m + 1 f_out(i,j,k) = work(m) ENDDO ENDDO ENDDO ENDDO !$OMP END PARALLEL #endif END SUBROUTINE transpose_yz SUBROUTINE transpose_zx( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Description: ! ------------ ! Transposition of input array (f_in) from z to x. For the input array, all ! elements along z reside on the same PE, while after transposition, all ! elements along x reside on the same PE. !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, xs REAL :: f_in(1:nza,nys:nyna,nxl:nxra), f_inv(nys:nyna,nxl:nxra,1:nza), & f_out(0:nxa,nys_x:nyn_xa,nzb_x:nzt_xa), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- Rearrange indices of input array in order to make data to be send !-- by MPI contiguous !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO k = 1,nza DO i = nxl, nxra DO j = nys, nyna f_inv(j,i,k) = f_in(k,j,i) ENDDO ENDDO ENDDO !$OMP END PARALLEL ! !-- Move data to different array, because memory location of work1 is !-- needed further below (work1 = work2). !-- If the PE grid is one-dimensional along y, only local reordering !-- of the data is necessary and no transposition has to be done. IF ( pdims(1) == 1 ) THEN !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO k = 1, nza DO i = nxl, nxra DO j = nys, nyna f_out(i,j,k) = f_inv(j,i,k) ENDDO ENDDO ENDDO !$OMP END PARALLEL RETURN ENDIF ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( f_inv(nys,nxl,1), sendrecvcount_zx, MPI_REAL, & work(1), sendrecvcount_zx, MPI_REAL, & comm1dx, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array !$OMP PARALLEL PRIVATE ( i, j, k, l, m, xs ) !$OMP DO DO l = 0, pdims(1) - 1 m = l * ( nzt_xa - nzb_x + 1 ) * nnx * ( nyn_xa - nys_x + 1 ) xs = 0 + l * nnx DO k = nzb_x, nzt_xa DO i = xs, xs + nnx - 1 DO j = nys_x, nyn_xa m = m + 1 f_out(i,j,k) = work(m) ENDDO ENDDO ENDDO ENDDO !$OMP END PARALLEL #endif END SUBROUTINE transpose_zx SUBROUTINE transpose_zy( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Description: ! ------------ ! Transposition of input array (f_in) from z to y. For the input array, all ! elements along z reside on the same PE, while after transposition, all ! elements along y reside on the same PE. !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, zs REAL :: f_in(nxl_z:nxr_za,nys_z:nyn_za,1:nza), & f_inv(nxl_y:nxr_ya,nzb_y:nzt_ya,0:nya), & f_out(0:nya,nxl_y:nxr_ya,nzb_y:nzt_ya), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- If the PE grid is one-dimensional along y, the array has only to be !-- reordered locally and therefore no transposition has to be done. IF ( pdims(1) /= 1 ) THEN ! !-- Reorder input array for transposition !$OMP PARALLEL PRIVATE ( i, j, k, l, m, zs ) !$OMP DO DO l = 0, pdims(1) - 1 m = l * ( nyn_za - nys_z + 1 ) * ( nzt_ya - nzb_y + 1 ) * & ( nxr_za - nxl_z + 1 ) zs = 1 + l * ( nzt_ya - nzb_y + 1 ) DO j = nys_z, nyn_za DO k = zs, zs + nzt_ya - nzb_y DO i = nxl_z, nxr_za m = m + 1 work(m) = f_in(i,j,k) ENDDO ENDDO ENDDO ENDDO !$OMP END PARALLEL ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( work(1), sendrecvcount_yz, MPI_REAL, & f_inv(nxl_y,nzb_y,0), sendrecvcount_yz, MPI_REAL, & comm1dx, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array in a way that the y index is in first position !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO j = 0, nya DO k = nzb_y, nzt_ya DO i = nxl_y, nxr_ya f_out(j,i,k) = f_inv(i,k,j) ENDDO ENDDO ENDDO !$OMP END PARALLEL ELSE ! !-- Reorder the array in a way that the y index is in first position !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO k = nzb_y, nzt_ya DO j = 0, nya DO i = nxl_y, nxr_ya f_inv(i,k,j) = f_in(i,j,k) ENDDO ENDDO ENDDO !$OMP END PARALLEL ! !-- Move data to output array !$OMP PARALLEL PRIVATE ( i, j, k ) !$OMP DO DO k = nzb_y, nzt_ya DO i = nxl_y, nxr_ya DO j = 0, nya f_out(j,i,k) = f_inv(i,k,j) ENDDO ENDDO ENDDO !$OMP END PARALLEL ENDIF #endif END SUBROUTINE transpose_zy SUBROUTINE transpose_zyd( f_in, work, f_out ) !------------------------------------------------------------------------------! ! Description: ! ------------ ! Transposition of input array (f_in) from z to y. For the input array, all ! elements along z reside on the same PE, while after transposition, all ! elements along y reside on the same PE. ! This is a direct transposition for arrays with indices in regular order ! (k,j,i) (cf. transpose_zy). !------------------------------------------------------------------------------! USE cpulog USE indices USE interfaces USE pegrid USE transpose_indices IMPLICIT NONE INTEGER :: i, j, k, l, m, ys REAL :: f_in(1:nza,nys:nyna,nxl:nxra), f_inv(nys:nyna,nxl:nxra,1:nza), & f_out(0:nya,nxl_yd:nxr_yda,nzb_yd:nzt_yda), & work(nnx*nny*nnz) #if defined( __parallel ) ! !-- Rearrange indices of input array in order to make data to be send !-- by MPI contiguous DO i = nxl, nxra DO j = nys, nyna DO k = 1, nza f_inv(j,i,k) = f_in(k,j,i) ENDDO ENDDO ENDDO ! !-- Move data to different array, because memory location of work1 is !-- needed further below (work1 = work2). !-- If the PE grid is one-dimensional along x, only local reordering !-- of the data is necessary and no transposition has to be done. IF ( pdims(2) == 1 ) THEN DO k = 1, nza DO i = nxl, nxra DO j = nys, nyna f_out(j,i,k) = f_inv(j,i,k) ENDDO ENDDO ENDDO RETURN ENDIF ! !-- Transpose array CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'start' ) IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) CALL MPI_ALLTOALL( f_inv(nys,nxl,1), sendrecvcount_zyd, MPI_REAL, & work(1), sendrecvcount_zyd, MPI_REAL, & comm1dy, ierr ) CALL cpu_log( log_point_s(32), 'mpi_alltoall', 'stop' ) ! !-- Reorder transposed array m = 0 DO l = 0, pdims(2) - 1 ys = 0 + l * nny DO k = nzb_yd, nzt_yda DO i = nxl_yd, nxr_yda DO j = ys, ys + nny - 1 m = m + 1 f_out(j,i,k) = work(m) ENDDO ENDDO ENDDO ENDDO #endif END SUBROUTINE transpose_zyd