[3] | 1 | MODULE lib_mpp |
---|
[13] | 2 | !!====================================================================== |
---|
| 3 | !! *** MODULE lib_mpp *** |
---|
[1344] | 4 | !! Ocean numerics: massively parallel processing library |
---|
[13] | 5 | !!===================================================================== |
---|
[1344] | 6 | !! History : OPA ! 1994 (M. Guyon, J. Escobar, M. Imbard) Original code |
---|
| 7 | !! 7.0 ! 1997 (A.M. Treguier) SHMEM additions |
---|
| 8 | !! 8.0 ! 1998 (M. Imbard, J. Escobar, L. Colombet ) SHMEM and MPI |
---|
| 9 | !! ! 1998 (J.M. Molines) Open boundary conditions |
---|
| 10 | !! NEMO 1.0 ! 2003 (J.-M. Molines, G. Madec) F90, free form |
---|
| 11 | !! ! 2003 (J.M. Molines) add mpp_ini_north(_3d,_2d) |
---|
| 12 | !! - ! 2004 (R. Bourdalle Badie) isend option in mpi |
---|
| 13 | !! ! 2004 (J.M. Molines) minloc, maxloc |
---|
| 14 | !! - ! 2005 (G. Madec, S. Masson) npolj=5,6 F-point & ice cases |
---|
| 15 | !! - ! 2005 (R. Redler) Replacement of MPI_COMM_WORLD except for MPI_Abort |
---|
| 16 | !! - ! 2005 (R. Benshila, G. Madec) add extra halo case |
---|
| 17 | !! - ! 2008 (R. Benshila) add mpp_ini_ice |
---|
| 18 | !! 3.2 ! 2009 (R. Benshila) SHMEM suppression, north fold in lbc_nfd |
---|
[1345] | 19 | !! 3.2 ! 2009 (O. Marti) add mpp_ini_znl |
---|
[13] | 20 | !!---------------------------------------------------------------------- |
---|
[1344] | 21 | #if defined key_mpp_mpi |
---|
[13] | 22 | !!---------------------------------------------------------------------- |
---|
[1344] | 23 | !! 'key_mpp_mpi' MPI massively parallel processing library |
---|
| 24 | !!---------------------------------------------------------------------- |
---|
| 25 | !! mynode : indentify the processor unit |
---|
| 26 | !! mpp_lnk : interface (defined in lbclnk) for message passing of 2d or 3d arrays (mpp_lnk_2d, mpp_lnk_3d) |
---|
[473] | 27 | !! mpp_lnk_3d_gather : Message passing manadgement for two 3D arrays |
---|
[1344] | 28 | !! mpp_lnk_e : interface (defined in lbclnk) for message passing of 2d array with extra halo (mpp_lnk_2d_e) |
---|
| 29 | !! mpprecv : |
---|
[1345] | 30 | !! mppsend : SUBROUTINE mpp_ini_znl |
---|
[1344] | 31 | !! mppscatter : |
---|
| 32 | !! mppgather : |
---|
| 33 | !! mpp_min : generic interface for mppmin_int , mppmin_a_int , mppmin_real, mppmin_a_real |
---|
| 34 | !! mpp_max : generic interface for mppmax_int , mppmax_a_int , mppmax_real, mppmax_a_real |
---|
| 35 | !! mpp_sum : generic interface for mppsum_int , mppsum_a_int , mppsum_real, mppsum_a_real |
---|
| 36 | !! mpp_minloc : |
---|
| 37 | !! mpp_maxloc : |
---|
| 38 | !! mppsync : |
---|
| 39 | !! mppstop : |
---|
| 40 | !! mppobc : variant of mpp_lnk for open boundary condition |
---|
| 41 | !! mpp_ini_north : initialisation of north fold |
---|
| 42 | !! mpp_lbc_north : north fold processors gathering |
---|
| 43 | !! mpp_lbc_north_e : variant of mpp_lbc_north for extra outer halo |
---|
[13] | 44 | !!---------------------------------------------------------------------- |
---|
| 45 | !! History : |
---|
| 46 | !! ! 94 (M. Guyon, J. Escobar, M. Imbard) Original code |
---|
| 47 | !! ! 97 (A.M. Treguier) SHMEM additions |
---|
| 48 | !! ! 98 (M. Imbard, J. Escobar, L. Colombet ) SHMEM and MPI |
---|
| 49 | !! 9.0 ! 03 (J.-M. Molines, G. Madec) F90, free form |
---|
[233] | 50 | !! ! 04 (R. Bourdalle Badie) isend option in mpi |
---|
| 51 | !! ! 05 (G. Madec, S. Masson) npolj=5,6 F-point & ice cases |
---|
[532] | 52 | !! ! 05 (R. Redler) Replacement of MPI_COMM_WORLD except for MPI_Abort |
---|
[1344] | 53 | !! ! 09 (R. Benshila) SHMEM suppression, north fold in lbc_nfd |
---|
[13] | 54 | !!---------------------------------------------------------------------- |
---|
[473] | 55 | USE dom_oce ! ocean space and time domain |
---|
| 56 | USE in_out_manager ! I/O manager |
---|
[1344] | 57 | USE lbcnfd ! north fold treatment |
---|
[3] | 58 | |
---|
[13] | 59 | IMPLICIT NONE |
---|
[415] | 60 | PRIVATE |
---|
[1344] | 61 | |
---|
| 62 | PUBLIC mynode, mppstop, mppsync, mpp_comm_free |
---|
| 63 | PUBLIC mpp_ini_north, mpp_lbc_north, mpp_lbc_north_e |
---|
| 64 | PUBLIC mpp_min, mpp_max, mpp_sum, mpp_minloc, mpp_maxloc |
---|
| 65 | PUBLIC mpp_lnk_3d, mpp_lnk_3d_gather, mpp_lnk_2d, mpp_lnk_2d_e |
---|
[1528] | 66 | PUBLIC mppobc, mpp_ini_ice, mpp_ini_znl |
---|
[415] | 67 | |
---|
[13] | 68 | !! * Interfaces |
---|
| 69 | !! define generic interface for these routine as they are called sometimes |
---|
[1344] | 70 | !! with scalar arguments instead of array arguments, which causes problems |
---|
| 71 | !! for the compilation on AIX system as well as NEC and SGI. Ok on COMPACQ |
---|
[13] | 72 | INTERFACE mpp_min |
---|
| 73 | MODULE PROCEDURE mppmin_a_int, mppmin_int, mppmin_a_real, mppmin_real |
---|
| 74 | END INTERFACE |
---|
| 75 | INTERFACE mpp_max |
---|
[681] | 76 | MODULE PROCEDURE mppmax_a_int, mppmax_int, mppmax_a_real, mppmax_real |
---|
[13] | 77 | END INTERFACE |
---|
| 78 | INTERFACE mpp_sum |
---|
[2304] | 79 | # if defined key_mpp_rep |
---|
[1976] | 80 | MODULE PROCEDURE mppsum_a_int, mppsum_int, mppsum_a_real, mppsum_real, & |
---|
| 81 | mppsum_realdd, mppsum_a_realdd |
---|
| 82 | # else |
---|
[13] | 83 | MODULE PROCEDURE mppsum_a_int, mppsum_int, mppsum_a_real, mppsum_real |
---|
[1976] | 84 | # endif |
---|
[13] | 85 | END INTERFACE |
---|
| 86 | INTERFACE mpp_lbc_north |
---|
| 87 | MODULE PROCEDURE mpp_lbc_north_3d, mpp_lbc_north_2d |
---|
| 88 | END INTERFACE |
---|
[1344] | 89 | INTERFACE mpp_minloc |
---|
| 90 | MODULE PROCEDURE mpp_minloc2d ,mpp_minloc3d |
---|
| 91 | END INTERFACE |
---|
| 92 | INTERFACE mpp_maxloc |
---|
| 93 | MODULE PROCEDURE mpp_maxloc2d ,mpp_maxloc3d |
---|
| 94 | END INTERFACE |
---|
[1976] | 95 | |
---|
[51] | 96 | !! ========================= !! |
---|
| 97 | !! MPI variable definition !! |
---|
| 98 | !! ========================= !! |
---|
[1629] | 99 | !$AGRIF_DO_NOT_TREAT |
---|
[2004] | 100 | INCLUDE 'mpif.h' |
---|
[1629] | 101 | !$AGRIF_END_DO_NOT_TREAT |
---|
[1344] | 102 | |
---|
| 103 | LOGICAL, PUBLIC, PARAMETER :: lk_mpp = .TRUE. !: mpp flag |
---|
[3] | 104 | |
---|
[1344] | 105 | INTEGER, PARAMETER :: nprocmax = 2**10 ! maximun dimension (required to be a power of 2) |
---|
| 106 | |
---|
| 107 | INTEGER :: mppsize ! number of process |
---|
| 108 | INTEGER :: mpprank ! process number [ 0 - size-1 ] |
---|
[2363] | 109 | !$AGRIF_DO_NOT_TREAT |
---|
[2249] | 110 | INTEGER, PUBLIC :: mpi_comm_opa ! opa local communicator |
---|
[2363] | 111 | !$AGRIF_END_DO_NOT_TREAT |
---|
[3] | 112 | |
---|
[2480] | 113 | # if defined key_mpp_rep |
---|
| 114 | INTEGER :: MPI_SUMDD |
---|
| 115 | # endif |
---|
[1976] | 116 | |
---|
[869] | 117 | ! variables used in case of sea-ice |
---|
[1344] | 118 | INTEGER, PUBLIC :: ncomm_ice !: communicator made by the processors with sea-ice |
---|
[1345] | 119 | INTEGER :: ngrp_ice ! group ID for the ice processors (for rheology) |
---|
| 120 | INTEGER :: ndim_rank_ice ! number of 'ice' processors |
---|
| 121 | INTEGER :: n_ice_root ! number (in the comm_ice) of proc 0 in the ice comm |
---|
[1344] | 122 | INTEGER, DIMENSION(:), ALLOCATABLE :: nrank_ice ! dimension ndim_rank_ice |
---|
[1345] | 123 | |
---|
| 124 | ! variables used for zonal integration |
---|
| 125 | INTEGER, PUBLIC :: ncomm_znl !: communicator made by the processors on the same zonal average |
---|
| 126 | LOGICAL, PUBLIC :: l_znl_root ! True on the 'left'most processor on the same row |
---|
| 127 | INTEGER :: ngrp_znl ! group ID for the znl processors |
---|
| 128 | INTEGER :: ndim_rank_znl ! number of processors on the same zonal average |
---|
| 129 | INTEGER, DIMENSION(:), ALLOCATABLE :: nrank_znl ! dimension ndim_rank_znl, number of the procs into the same znl domain |
---|
[1344] | 130 | |
---|
| 131 | ! North fold condition in mpp_mpi with jpni > 1 |
---|
| 132 | INTEGER :: ngrp_world ! group ID for the world processors |
---|
[1345] | 133 | INTEGER :: ngrp_opa ! group ID for the opa processors |
---|
[1344] | 134 | INTEGER :: ngrp_north ! group ID for the northern processors (to be fold) |
---|
| 135 | INTEGER :: ncomm_north ! communicator made by the processors belonging to ngrp_north |
---|
| 136 | INTEGER :: ndim_rank_north ! number of 'sea' processor in the northern line (can be /= jpni !) |
---|
| 137 | INTEGER :: njmppmax ! value of njmpp for the processors of the northern line |
---|
| 138 | INTEGER :: north_root ! number (in the comm_opa) of proc 0 in the northern comm |
---|
| 139 | INTEGER, DIMENSION(:), ALLOCATABLE :: nrank_north ! dimension ndim_rank_north |
---|
[3] | 140 | |
---|
[1344] | 141 | ! Type of send : standard, buffered, immediate |
---|
[1601] | 142 | CHARACTER(len=1) :: cn_mpi_send = 'S' ! type od mpi send/recieve (S=standard, B=bsend, I=isend) |
---|
| 143 | LOGICAL :: l_isend = .FALSE. ! isend use indicator (T if cn_mpi_send='I') |
---|
[1344] | 144 | INTEGER :: nn_buffer = 0 ! size of the buffer in case of mpi_bsend |
---|
| 145 | |
---|
| 146 | REAL(wp), ALLOCATABLE, DIMENSION(:) :: tampon ! buffer in case of bsend |
---|
[3] | 147 | |
---|
[1344] | 148 | ! message passing arrays |
---|
| 149 | REAL(wp), DIMENSION(jpi,jprecj,jpk,2,2) :: t4ns, t4sn ! 2 x 3d for north-south & south-north |
---|
| 150 | REAL(wp), DIMENSION(jpj,jpreci,jpk,2,2) :: t4ew, t4we ! 2 x 3d for east-west & west-east |
---|
| 151 | REAL(wp), DIMENSION(jpi,jprecj,jpk,2,2) :: t4p1, t4p2 ! 2 x 3d for north fold |
---|
| 152 | REAL(wp), DIMENSION(jpi,jprecj,jpk,2) :: t3ns, t3sn ! 3d for north-south & south-north |
---|
| 153 | REAL(wp), DIMENSION(jpj,jpreci,jpk,2) :: t3ew, t3we ! 3d for east-west & west-east |
---|
| 154 | REAL(wp), DIMENSION(jpi,jprecj,jpk,2) :: t3p1, t3p2 ! 3d for north fold |
---|
| 155 | REAL(wp), DIMENSION(jpi,jprecj,2) :: t2ns, t2sn ! 2d for north-south & south-north |
---|
| 156 | REAL(wp), DIMENSION(jpj,jpreci,2) :: t2ew, t2we ! 2d for east-west & west-east |
---|
| 157 | REAL(wp), DIMENSION(jpi,jprecj,2) :: t2p1, t2p2 ! 2d for north fold |
---|
| 158 | REAL(wp), DIMENSION(1-jpr2di:jpi+jpr2di,jprecj+jpr2dj,2) :: tr2ns, tr2sn ! 2d for north-south & south-north + extra outer halo |
---|
| 159 | REAL(wp), DIMENSION(1-jpr2dj:jpj+jpr2dj,jpreci+jpr2di,2) :: tr2ew, tr2we ! 2d for east-west & west-east + extra outer halo |
---|
[51] | 160 | !!---------------------------------------------------------------------- |
---|
[2287] | 161 | !! NEMO/OPA 3.3 , NEMO Consortium (2010) |
---|
[888] | 162 | !! $Id$ |
---|
[2287] | 163 | !! Software governed by the CeCILL licence (NEMOGCM/NEMO_CeCILL.txt) |
---|
[1344] | 164 | !!---------------------------------------------------------------------- |
---|
[3] | 165 | |
---|
| 166 | CONTAINS |
---|
| 167 | |
---|
[1579] | 168 | FUNCTION mynode(ldtxt, localComm) |
---|
[51] | 169 | !!---------------------------------------------------------------------- |
---|
| 170 | !! *** routine mynode *** |
---|
| 171 | !! |
---|
| 172 | !! ** Purpose : Find processor unit |
---|
| 173 | !! |
---|
| 174 | !!---------------------------------------------------------------------- |
---|
[1579] | 175 | CHARACTER(len=*),DIMENSION(:), INTENT( out) :: ldtxt |
---|
| 176 | INTEGER, OPTIONAL , INTENT(in ) :: localComm |
---|
[2481] | 177 | INTEGER :: mynode, ierr, code, ji, ii |
---|
[532] | 178 | LOGICAL :: mpi_was_called |
---|
[1579] | 179 | |
---|
[1601] | 180 | NAMELIST/nammpp/ cn_mpi_send, nn_buffer |
---|
[51] | 181 | !!---------------------------------------------------------------------- |
---|
[1344] | 182 | ! |
---|
[2481] | 183 | ii = 1 |
---|
| 184 | WRITE(ldtxt(ii),*) ; ii = ii + 1 |
---|
| 185 | WRITE(ldtxt(ii),*) 'mynode : mpi initialisation' ; ii = ii + 1 |
---|
| 186 | WRITE(ldtxt(ii),*) '~~~~~~ ' ; ii = ii + 1 |
---|
[1344] | 187 | ! |
---|
| 188 | REWIND( numnam ) ! Namelist namrun : parameters of the run |
---|
[1601] | 189 | READ ( numnam, nammpp ) |
---|
[1344] | 190 | ! ! control print |
---|
[2481] | 191 | WRITE(ldtxt(ii),*) ' Namelist nammpp' ; ii = ii + 1 |
---|
| 192 | WRITE(ldtxt(ii),*) ' mpi send type cn_mpi_send = ', cn_mpi_send ; ii = ii + 1 |
---|
| 193 | WRITE(ldtxt(ii),*) ' size in bytes of exported buffer nn_buffer = ', nn_buffer ; ii = ii + 1 |
---|
[300] | 194 | |
---|
[2480] | 195 | CALL mpi_initialized ( mpi_was_called, code ) |
---|
| 196 | IF( code /= MPI_SUCCESS ) THEN |
---|
[2481] | 197 | DO ji = 1, SIZE(ldtxt) |
---|
| 198 | IF( TRIM(ldtxt(ji)) /= '' ) WRITE(*,*) ldtxt(ji) ! control print of mynode |
---|
| 199 | END DO |
---|
[2480] | 200 | WRITE(*, cform_err) |
---|
| 201 | WRITE(*, *) 'lib_mpp: Error in routine mpi_initialized' |
---|
| 202 | CALL mpi_abort( mpi_comm_world, code, ierr ) |
---|
| 203 | ENDIF |
---|
[415] | 204 | |
---|
[2480] | 205 | IF( mpi_was_called ) THEN |
---|
| 206 | ! |
---|
| 207 | SELECT CASE ( cn_mpi_send ) |
---|
| 208 | CASE ( 'S' ) ! Standard mpi send (blocking) |
---|
[2481] | 209 | WRITE(ldtxt(ii),*) ' Standard blocking mpi send (send)' ; ii = ii + 1 |
---|
[2480] | 210 | CASE ( 'B' ) ! Buffer mpi send (blocking) |
---|
[2481] | 211 | WRITE(ldtxt(ii),*) ' Buffer blocking mpi send (bsend)' ; ii = ii + 1 |
---|
| 212 | IF( Agrif_Root() ) CALL mpi_init_opa( ldtxt, ii, ierr ) |
---|
[2480] | 213 | CASE ( 'I' ) ! Immediate mpi send (non-blocking send) |
---|
[2481] | 214 | WRITE(ldtxt(ii),*) ' Immediate non-blocking send (isend)' ; ii = ii + 1 |
---|
[2480] | 215 | l_isend = .TRUE. |
---|
| 216 | CASE DEFAULT |
---|
[2481] | 217 | WRITE(ldtxt(ii),cform_err) ; ii = ii + 1 |
---|
| 218 | WRITE(ldtxt(ii),*) ' bad value for cn_mpi_send = ', cn_mpi_send ; ii = ii + 1 |
---|
[1579] | 219 | nstop = nstop + 1 |
---|
[2480] | 220 | END SELECT |
---|
| 221 | ELSE IF ( PRESENT(localComm) .and. .not. mpi_was_called ) THEN |
---|
[2481] | 222 | WRITE(ldtxt(ii),*) ' lib_mpp: You cannot provide a local communicator ' ; ii = ii + 1 |
---|
| 223 | WRITE(ldtxt(ii),*) ' without calling MPI_Init before ! ' ; ii = ii + 1 |
---|
[2480] | 224 | nstop = nstop + 1 |
---|
[532] | 225 | ELSE |
---|
[1601] | 226 | SELECT CASE ( cn_mpi_send ) |
---|
[524] | 227 | CASE ( 'S' ) ! Standard mpi send (blocking) |
---|
[2481] | 228 | WRITE(ldtxt(ii),*) ' Standard blocking mpi send (send)' ; ii = ii + 1 |
---|
[2480] | 229 | CALL mpi_init( ierr ) |
---|
[524] | 230 | CASE ( 'B' ) ! Buffer mpi send (blocking) |
---|
[2481] | 231 | WRITE(ldtxt(ii),*) ' Buffer blocking mpi send (bsend)' ; ii = ii + 1 |
---|
| 232 | IF( Agrif_Root() ) CALL mpi_init_opa( ldtxt, ii, ierr ) |
---|
[524] | 233 | CASE ( 'I' ) ! Immediate mpi send (non-blocking send) |
---|
[2481] | 234 | WRITE(ldtxt(ii),*) ' Immediate non-blocking send (isend)' ; ii = ii + 1 |
---|
[524] | 235 | l_isend = .TRUE. |
---|
[2480] | 236 | CALL mpi_init( ierr ) |
---|
[524] | 237 | CASE DEFAULT |
---|
[2481] | 238 | WRITE(ldtxt(ii),cform_err) ; ii = ii + 1 |
---|
| 239 | WRITE(ldtxt(ii),*) ' bad value for cn_mpi_send = ', cn_mpi_send ; ii = ii + 1 |
---|
[524] | 240 | nstop = nstop + 1 |
---|
| 241 | END SELECT |
---|
[2480] | 242 | ! |
---|
[415] | 243 | ENDIF |
---|
[570] | 244 | |
---|
[2480] | 245 | IF( PRESENT(localComm) ) THEN |
---|
| 246 | IF( Agrif_Root() ) THEN |
---|
| 247 | mpi_comm_opa = localComm |
---|
| 248 | ENDIF |
---|
| 249 | ELSE |
---|
| 250 | CALL mpi_comm_dup( mpi_comm_world, mpi_comm_opa, code) |
---|
| 251 | IF( code /= MPI_SUCCESS ) THEN |
---|
[2481] | 252 | DO ji = 1, SIZE(ldtxt) |
---|
| 253 | IF( TRIM(ldtxt(ji)) /= '' ) WRITE(*,*) ldtxt(ji) ! control print of mynode |
---|
| 254 | END DO |
---|
[2480] | 255 | WRITE(*, cform_err) |
---|
| 256 | WRITE(*, *) ' lib_mpp: Error in routine mpi_comm_dup' |
---|
| 257 | CALL mpi_abort( mpi_comm_world, code, ierr ) |
---|
| 258 | ENDIF |
---|
| 259 | ENDIF |
---|
| 260 | |
---|
[1344] | 261 | CALL mpi_comm_rank( mpi_comm_opa, mpprank, ierr ) |
---|
| 262 | CALL mpi_comm_size( mpi_comm_opa, mppsize, ierr ) |
---|
[629] | 263 | mynode = mpprank |
---|
[2480] | 264 | ! |
---|
[2304] | 265 | #if defined key_mpp_rep |
---|
[1976] | 266 | CALL MPI_OP_CREATE(DDPDD_MPI, .TRUE., MPI_SUMDD, ierr) |
---|
| 267 | #endif |
---|
| 268 | ! |
---|
[51] | 269 | END FUNCTION mynode |
---|
[3] | 270 | |
---|
| 271 | |
---|
[888] | 272 | SUBROUTINE mpp_lnk_3d( ptab, cd_type, psgn, cd_mpp, pval ) |
---|
[51] | 273 | !!---------------------------------------------------------------------- |
---|
| 274 | !! *** routine mpp_lnk_3d *** |
---|
| 275 | !! |
---|
| 276 | !! ** Purpose : Message passing manadgement |
---|
| 277 | !! |
---|
| 278 | !! ** Method : Use mppsend and mpprecv function for passing mask |
---|
| 279 | !! between processors following neighboring subdomains. |
---|
| 280 | !! domain parameters |
---|
| 281 | !! nlci : first dimension of the local subdomain |
---|
| 282 | !! nlcj : second dimension of the local subdomain |
---|
| 283 | !! nbondi : mark for "east-west local boundary" |
---|
| 284 | !! nbondj : mark for "north-south local boundary" |
---|
| 285 | !! noea : number for local neighboring processors |
---|
| 286 | !! nowe : number for local neighboring processors |
---|
| 287 | !! noso : number for local neighboring processors |
---|
| 288 | !! nono : number for local neighboring processors |
---|
| 289 | !! |
---|
| 290 | !! ** Action : ptab with update value at its periphery |
---|
| 291 | !! |
---|
| 292 | !!---------------------------------------------------------------------- |
---|
[1344] | 293 | REAL(wp), DIMENSION(jpi,jpj,jpk), INTENT(inout) :: ptab ! 3D array on which the boundary condition is applied |
---|
| 294 | CHARACTER(len=1) , INTENT(in ) :: cd_type ! define the nature of ptab array grid-points |
---|
| 295 | ! ! = T , U , V , F , W points |
---|
| 296 | REAL(wp) , INTENT(in ) :: psgn ! =-1 the sign change across the north fold boundary |
---|
| 297 | ! ! = 1. , the sign is kept |
---|
| 298 | CHARACTER(len=3), OPTIONAL , INTENT(in ) :: cd_mpp ! fill the overlap area only |
---|
| 299 | REAL(wp) , OPTIONAL , INTENT(in ) :: pval ! background value (used at closed boundaries) |
---|
| 300 | !! |
---|
[1718] | 301 | INTEGER :: ji, jj, jk, jl ! dummy loop indices |
---|
[1344] | 302 | INTEGER :: imigr, iihom, ijhom ! temporary integers |
---|
| 303 | INTEGER :: ml_req1, ml_req2, ml_err ! for key_mpi_isend |
---|
[888] | 304 | REAL(wp) :: zland |
---|
[1344] | 305 | INTEGER, DIMENSION(MPI_STATUS_SIZE) :: ml_stat ! for key_mpi_isend |
---|
[51] | 306 | !!---------------------------------------------------------------------- |
---|
[3] | 307 | |
---|
[1344] | 308 | IF( PRESENT( pval ) ) THEN ; zland = pval ! set land value |
---|
| 309 | ELSE ; zland = 0.e0 ! zero by default |
---|
| 310 | ENDIF |
---|
| 311 | |
---|
[51] | 312 | ! 1. standard boundary treatment |
---|
| 313 | ! ------------------------------ |
---|
[1718] | 314 | IF( PRESENT( cd_mpp ) ) THEN ! only fill added line/raw with existing values |
---|
[1344] | 315 | ! |
---|
[1718] | 316 | ! WARNING ptab is defined only between nld and nle |
---|
| 317 | DO jk = 1, jpk |
---|
| 318 | DO jj = nlcj+1, jpj ! added line(s) (inner only) |
---|
| 319 | ptab(nldi :nlei , jj ,jk) = ptab(nldi:nlei, nlej,jk) |
---|
| 320 | ptab(1 :nldi-1, jj ,jk) = ptab(nldi , nlej,jk) |
---|
| 321 | ptab(nlei+1:nlci , jj ,jk) = ptab( nlei, nlej,jk) |
---|
| 322 | END DO |
---|
| 323 | DO ji = nlci+1, jpi ! added column(s) (full) |
---|
| 324 | ptab(ji ,nldj :nlej ,jk) = ptab( nlei,nldj:nlej,jk) |
---|
| 325 | ptab(ji ,1 :nldj-1,jk) = ptab( nlei,nldj ,jk) |
---|
| 326 | ptab(ji ,nlej+1:jpj ,jk) = ptab( nlei, nlej,jk) |
---|
| 327 | END DO |
---|
[610] | 328 | END DO |
---|
[1344] | 329 | ! |
---|
| 330 | ELSE ! standard close or cyclic treatment |
---|
| 331 | ! |
---|
| 332 | ! ! East-West boundaries |
---|
| 333 | ! !* Cyclic east-west |
---|
| 334 | IF( nbondi == 2 .AND. (nperio == 1 .OR. nperio == 4 .OR. nperio == 6) ) THEN |
---|
[473] | 335 | ptab( 1 ,:,:) = ptab(jpim1,:,:) |
---|
| 336 | ptab(jpi,:,:) = ptab( 2 ,:,:) |
---|
[1344] | 337 | ELSE !* closed |
---|
| 338 | IF( .NOT. cd_type == 'F' ) ptab( 1 :jpreci,:,:) = zland ! south except F-point |
---|
| 339 | ptab(nlci-jpreci+1:jpi ,:,:) = zland ! north |
---|
[473] | 340 | ENDIF |
---|
[1344] | 341 | ! ! North-South boundaries (always closed) |
---|
| 342 | IF( .NOT. cd_type == 'F' ) ptab(:, 1 :jprecj,:) = zland ! south except F-point |
---|
| 343 | ptab(:,nlcj-jprecj+1:jpj ,:) = zland ! north |
---|
| 344 | ! |
---|
[51] | 345 | ENDIF |
---|
[3] | 346 | |
---|
[51] | 347 | ! 2. East and west directions exchange |
---|
| 348 | ! ------------------------------------ |
---|
[1344] | 349 | ! we play with the neigbours AND the row number because of the periodicity |
---|
| 350 | ! |
---|
| 351 | SELECT CASE ( nbondi ) ! Read Dirichlet lateral conditions |
---|
| 352 | CASE ( -1, 0, 1 ) ! all exept 2 (i.e. close case) |
---|
[51] | 353 | iihom = nlci-nreci |
---|
| 354 | DO jl = 1, jpreci |
---|
| 355 | t3ew(:,jl,:,1) = ptab(jpreci+jl,:,:) |
---|
| 356 | t3we(:,jl,:,1) = ptab(iihom +jl,:,:) |
---|
| 357 | END DO |
---|
[1345] | 358 | END SELECT |
---|
[1344] | 359 | ! |
---|
| 360 | ! ! Migrations |
---|
[51] | 361 | imigr = jpreci * jpj * jpk |
---|
[1344] | 362 | ! |
---|
[51] | 363 | SELECT CASE ( nbondi ) |
---|
| 364 | CASE ( -1 ) |
---|
[181] | 365 | CALL mppsend( 2, t3we(1,1,1,1), imigr, noea, ml_req1 ) |
---|
[51] | 366 | CALL mpprecv( 1, t3ew(1,1,1,2), imigr ) |
---|
[300] | 367 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
[51] | 368 | CASE ( 0 ) |
---|
[181] | 369 | CALL mppsend( 1, t3ew(1,1,1,1), imigr, nowe, ml_req1 ) |
---|
| 370 | CALL mppsend( 2, t3we(1,1,1,1), imigr, noea, ml_req2 ) |
---|
[51] | 371 | CALL mpprecv( 1, t3ew(1,1,1,2), imigr ) |
---|
| 372 | CALL mpprecv( 2, t3we(1,1,1,2), imigr ) |
---|
[300] | 373 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 374 | IF(l_isend) CALL mpi_wait(ml_req2, ml_stat, ml_err) |
---|
[51] | 375 | CASE ( 1 ) |
---|
[181] | 376 | CALL mppsend( 1, t3ew(1,1,1,1), imigr, nowe, ml_req1 ) |
---|
[51] | 377 | CALL mpprecv( 2, t3we(1,1,1,2), imigr ) |
---|
[300] | 378 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
[51] | 379 | END SELECT |
---|
[1344] | 380 | ! |
---|
| 381 | ! ! Write Dirichlet lateral conditions |
---|
[51] | 382 | iihom = nlci-jpreci |
---|
[1344] | 383 | ! |
---|
[51] | 384 | SELECT CASE ( nbondi ) |
---|
| 385 | CASE ( -1 ) |
---|
| 386 | DO jl = 1, jpreci |
---|
| 387 | ptab(iihom+jl,:,:) = t3ew(:,jl,:,2) |
---|
| 388 | END DO |
---|
| 389 | CASE ( 0 ) |
---|
| 390 | DO jl = 1, jpreci |
---|
| 391 | ptab(jl ,:,:) = t3we(:,jl,:,2) |
---|
| 392 | ptab(iihom+jl,:,:) = t3ew(:,jl,:,2) |
---|
| 393 | END DO |
---|
| 394 | CASE ( 1 ) |
---|
| 395 | DO jl = 1, jpreci |
---|
| 396 | ptab(jl ,:,:) = t3we(:,jl,:,2) |
---|
| 397 | END DO |
---|
| 398 | END SELECT |
---|
[3] | 399 | |
---|
| 400 | |
---|
[51] | 401 | ! 3. North and south directions |
---|
| 402 | ! ----------------------------- |
---|
[1344] | 403 | ! always closed : we play only with the neigbours |
---|
| 404 | ! |
---|
| 405 | IF( nbondj /= 2 ) THEN ! Read Dirichlet lateral conditions |
---|
[51] | 406 | ijhom = nlcj-nrecj |
---|
| 407 | DO jl = 1, jprecj |
---|
| 408 | t3sn(:,jl,:,1) = ptab(:,ijhom +jl,:) |
---|
| 409 | t3ns(:,jl,:,1) = ptab(:,jprecj+jl,:) |
---|
| 410 | END DO |
---|
| 411 | ENDIF |
---|
[1344] | 412 | ! |
---|
| 413 | ! ! Migrations |
---|
[51] | 414 | imigr = jprecj * jpi * jpk |
---|
[1344] | 415 | ! |
---|
[51] | 416 | SELECT CASE ( nbondj ) |
---|
| 417 | CASE ( -1 ) |
---|
[181] | 418 | CALL mppsend( 4, t3sn(1,1,1,1), imigr, nono, ml_req1 ) |
---|
[51] | 419 | CALL mpprecv( 3, t3ns(1,1,1,2), imigr ) |
---|
[300] | 420 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
[51] | 421 | CASE ( 0 ) |
---|
[181] | 422 | CALL mppsend( 3, t3ns(1,1,1,1), imigr, noso, ml_req1 ) |
---|
| 423 | CALL mppsend( 4, t3sn(1,1,1,1), imigr, nono, ml_req2 ) |
---|
[51] | 424 | CALL mpprecv( 3, t3ns(1,1,1,2), imigr ) |
---|
| 425 | CALL mpprecv( 4, t3sn(1,1,1,2), imigr ) |
---|
[300] | 426 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 427 | IF(l_isend) CALL mpi_wait(ml_req2, ml_stat, ml_err) |
---|
[51] | 428 | CASE ( 1 ) |
---|
[181] | 429 | CALL mppsend( 3, t3ns(1,1,1,1), imigr, noso, ml_req1 ) |
---|
[51] | 430 | CALL mpprecv( 4, t3sn(1,1,1,2), imigr ) |
---|
[300] | 431 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
[51] | 432 | END SELECT |
---|
[1344] | 433 | ! |
---|
| 434 | ! ! Write Dirichlet lateral conditions |
---|
[51] | 435 | ijhom = nlcj-jprecj |
---|
[1344] | 436 | ! |
---|
[51] | 437 | SELECT CASE ( nbondj ) |
---|
| 438 | CASE ( -1 ) |
---|
| 439 | DO jl = 1, jprecj |
---|
| 440 | ptab(:,ijhom+jl,:) = t3ns(:,jl,:,2) |
---|
| 441 | END DO |
---|
| 442 | CASE ( 0 ) |
---|
| 443 | DO jl = 1, jprecj |
---|
| 444 | ptab(:,jl ,:) = t3sn(:,jl,:,2) |
---|
| 445 | ptab(:,ijhom+jl,:) = t3ns(:,jl,:,2) |
---|
| 446 | END DO |
---|
| 447 | CASE ( 1 ) |
---|
| 448 | DO jl = 1, jprecj |
---|
| 449 | ptab(:,jl,:) = t3sn(:,jl,:,2) |
---|
| 450 | END DO |
---|
| 451 | END SELECT |
---|
[3] | 452 | |
---|
| 453 | |
---|
[51] | 454 | ! 4. north fold treatment |
---|
| 455 | ! ----------------------- |
---|
[1344] | 456 | ! |
---|
| 457 | IF( npolj /= 0 .AND. .NOT. PRESENT(cd_mpp) ) THEN |
---|
| 458 | ! |
---|
| 459 | SELECT CASE ( jpni ) |
---|
| 460 | CASE ( 1 ) ; CALL lbc_nfd ( ptab, cd_type, psgn ) ! only 1 northern proc, no mpp |
---|
| 461 | CASE DEFAULT ; CALL mpp_lbc_north( ptab, cd_type, psgn ) ! for all northern procs. |
---|
| 462 | END SELECT |
---|
| 463 | ! |
---|
[473] | 464 | ENDIF |
---|
[1344] | 465 | ! |
---|
[51] | 466 | END SUBROUTINE mpp_lnk_3d |
---|
[3] | 467 | |
---|
| 468 | |
---|
[888] | 469 | SUBROUTINE mpp_lnk_2d( pt2d, cd_type, psgn, cd_mpp, pval ) |
---|
[51] | 470 | !!---------------------------------------------------------------------- |
---|
| 471 | !! *** routine mpp_lnk_2d *** |
---|
| 472 | !! |
---|
| 473 | !! ** Purpose : Message passing manadgement for 2d array |
---|
| 474 | !! |
---|
| 475 | !! ** Method : Use mppsend and mpprecv function for passing mask |
---|
| 476 | !! between processors following neighboring subdomains. |
---|
| 477 | !! domain parameters |
---|
| 478 | !! nlci : first dimension of the local subdomain |
---|
| 479 | !! nlcj : second dimension of the local subdomain |
---|
| 480 | !! nbondi : mark for "east-west local boundary" |
---|
| 481 | !! nbondj : mark for "north-south local boundary" |
---|
| 482 | !! noea : number for local neighboring processors |
---|
| 483 | !! nowe : number for local neighboring processors |
---|
| 484 | !! noso : number for local neighboring processors |
---|
| 485 | !! nono : number for local neighboring processors |
---|
| 486 | !! |
---|
| 487 | !!---------------------------------------------------------------------- |
---|
[1344] | 488 | REAL(wp), DIMENSION(jpi,jpj), INTENT(inout) :: pt2d ! 2D array on which the boundary condition is applied |
---|
| 489 | CHARACTER(len=1) , INTENT(in ) :: cd_type ! define the nature of ptab array grid-points |
---|
| 490 | ! ! = T , U , V , F , W and I points |
---|
| 491 | REAL(wp) , INTENT(in ) :: psgn ! =-1 the sign change across the north fold boundary |
---|
| 492 | ! ! = 1. , the sign is kept |
---|
| 493 | CHARACTER(len=3), OPTIONAL , INTENT(in ) :: cd_mpp ! fill the overlap area only |
---|
| 494 | REAL(wp) , OPTIONAL , INTENT(in ) :: pval ! background value (used at closed boundaries) |
---|
| 495 | !! |
---|
| 496 | INTEGER :: ji, jj, jl ! dummy loop indices |
---|
| 497 | INTEGER :: imigr, iihom, ijhom ! temporary integers |
---|
| 498 | INTEGER :: ml_req1, ml_req2, ml_err ! for key_mpi_isend |
---|
[888] | 499 | REAL(wp) :: zland |
---|
[1344] | 500 | INTEGER, DIMENSION(MPI_STATUS_SIZE) :: ml_stat ! for key_mpi_isend |
---|
[51] | 501 | !!---------------------------------------------------------------------- |
---|
[3] | 502 | |
---|
[1344] | 503 | IF( PRESENT( pval ) ) THEN ; zland = pval ! set land value |
---|
| 504 | ELSE ; zland = 0.e0 ! zero by default |
---|
[888] | 505 | ENDIF |
---|
| 506 | |
---|
[51] | 507 | ! 1. standard boundary treatment |
---|
| 508 | ! ------------------------------ |
---|
[1344] | 509 | ! |
---|
[1718] | 510 | IF( PRESENT( cd_mpp ) ) THEN ! only fill added line/raw with existing values |
---|
[1344] | 511 | ! |
---|
[1718] | 512 | ! WARNING pt2d is defined only between nld and nle |
---|
| 513 | DO jj = nlcj+1, jpj ! added line(s) (inner only) |
---|
| 514 | pt2d(nldi :nlei , jj ) = pt2d(nldi:nlei, nlej) |
---|
| 515 | pt2d(1 :nldi-1, jj ) = pt2d(nldi , nlej) |
---|
| 516 | pt2d(nlei+1:nlci , jj ) = pt2d( nlei, nlej) |
---|
[610] | 517 | END DO |
---|
[1718] | 518 | DO ji = nlci+1, jpi ! added column(s) (full) |
---|
| 519 | pt2d(ji ,nldj :nlej ) = pt2d( nlei,nldj:nlej) |
---|
| 520 | pt2d(ji ,1 :nldj-1) = pt2d( nlei,nldj ) |
---|
| 521 | pt2d(ji ,nlej+1:jpj ) = pt2d( nlei, nlej) |
---|
[1344] | 522 | END DO |
---|
| 523 | ! |
---|
| 524 | ELSE ! standard close or cyclic treatment |
---|
| 525 | ! |
---|
| 526 | ! ! East-West boundaries |
---|
| 527 | IF( nbondi == 2 .AND. & ! Cyclic east-west |
---|
[473] | 528 | & (nperio == 1 .OR. nperio == 4 .OR. nperio == 6) ) THEN |
---|
[1344] | 529 | pt2d( 1 ,:) = pt2d(jpim1,:) ! west |
---|
| 530 | pt2d(jpi,:) = pt2d( 2 ,:) ! east |
---|
| 531 | ELSE ! closed |
---|
| 532 | IF( .NOT. cd_type == 'F' ) pt2d( 1 :jpreci,:) = zland ! south except F-point |
---|
| 533 | pt2d(nlci-jpreci+1:jpi ,:) = zland ! north |
---|
[473] | 534 | ENDIF |
---|
[1344] | 535 | ! ! North-South boundaries (always closed) |
---|
| 536 | IF( .NOT. cd_type == 'F' ) pt2d(:, 1 :jprecj) = zland !south except F-point |
---|
| 537 | pt2d(:,nlcj-jprecj+1:jpj ) = zland ! north |
---|
| 538 | ! |
---|
[51] | 539 | ENDIF |
---|
[3] | 540 | |
---|
[1344] | 541 | ! 2. East and west directions exchange |
---|
| 542 | ! ------------------------------------ |
---|
| 543 | ! we play with the neigbours AND the row number because of the periodicity |
---|
| 544 | ! |
---|
| 545 | SELECT CASE ( nbondi ) ! Read Dirichlet lateral conditions |
---|
| 546 | CASE ( -1, 0, 1 ) ! all exept 2 (i.e. close case) |
---|
[51] | 547 | iihom = nlci-nreci |
---|
| 548 | DO jl = 1, jpreci |
---|
| 549 | t2ew(:,jl,1) = pt2d(jpreci+jl,:) |
---|
| 550 | t2we(:,jl,1) = pt2d(iihom +jl,:) |
---|
| 551 | END DO |
---|
| 552 | END SELECT |
---|
[1344] | 553 | ! |
---|
| 554 | ! ! Migrations |
---|
[51] | 555 | imigr = jpreci * jpj |
---|
[1344] | 556 | ! |
---|
[51] | 557 | SELECT CASE ( nbondi ) |
---|
| 558 | CASE ( -1 ) |
---|
[181] | 559 | CALL mppsend( 2, t2we(1,1,1), imigr, noea, ml_req1 ) |
---|
[51] | 560 | CALL mpprecv( 1, t2ew(1,1,2), imigr ) |
---|
[300] | 561 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
[51] | 562 | CASE ( 0 ) |
---|
[181] | 563 | CALL mppsend( 1, t2ew(1,1,1), imigr, nowe, ml_req1 ) |
---|
| 564 | CALL mppsend( 2, t2we(1,1,1), imigr, noea, ml_req2 ) |
---|
[51] | 565 | CALL mpprecv( 1, t2ew(1,1,2), imigr ) |
---|
| 566 | CALL mpprecv( 2, t2we(1,1,2), imigr ) |
---|
[300] | 567 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 568 | IF(l_isend) CALL mpi_wait(ml_req2,ml_stat,ml_err) |
---|
[51] | 569 | CASE ( 1 ) |
---|
[181] | 570 | CALL mppsend( 1, t2ew(1,1,1), imigr, nowe, ml_req1 ) |
---|
[51] | 571 | CALL mpprecv( 2, t2we(1,1,2), imigr ) |
---|
[300] | 572 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
[51] | 573 | END SELECT |
---|
[1344] | 574 | ! |
---|
| 575 | ! ! Write Dirichlet lateral conditions |
---|
[51] | 576 | iihom = nlci - jpreci |
---|
[1344] | 577 | ! |
---|
[51] | 578 | SELECT CASE ( nbondi ) |
---|
| 579 | CASE ( -1 ) |
---|
| 580 | DO jl = 1, jpreci |
---|
| 581 | pt2d(iihom+jl,:) = t2ew(:,jl,2) |
---|
| 582 | END DO |
---|
| 583 | CASE ( 0 ) |
---|
| 584 | DO jl = 1, jpreci |
---|
| 585 | pt2d(jl ,:) = t2we(:,jl,2) |
---|
| 586 | pt2d(iihom+jl,:) = t2ew(:,jl,2) |
---|
| 587 | END DO |
---|
| 588 | CASE ( 1 ) |
---|
| 589 | DO jl = 1, jpreci |
---|
| 590 | pt2d(jl ,:) = t2we(:,jl,2) |
---|
| 591 | END DO |
---|
| 592 | END SELECT |
---|
[3] | 593 | |
---|
| 594 | |
---|
[51] | 595 | ! 3. North and south directions |
---|
| 596 | ! ----------------------------- |
---|
[1344] | 597 | ! always closed : we play only with the neigbours |
---|
| 598 | ! |
---|
| 599 | IF( nbondj /= 2 ) THEN ! Read Dirichlet lateral conditions |
---|
[51] | 600 | ijhom = nlcj-nrecj |
---|
| 601 | DO jl = 1, jprecj |
---|
| 602 | t2sn(:,jl,1) = pt2d(:,ijhom +jl) |
---|
| 603 | t2ns(:,jl,1) = pt2d(:,jprecj+jl) |
---|
| 604 | END DO |
---|
| 605 | ENDIF |
---|
[1344] | 606 | ! |
---|
| 607 | ! ! Migrations |
---|
[51] | 608 | imigr = jprecj * jpi |
---|
[1344] | 609 | ! |
---|
[51] | 610 | SELECT CASE ( nbondj ) |
---|
| 611 | CASE ( -1 ) |
---|
[181] | 612 | CALL mppsend( 4, t2sn(1,1,1), imigr, nono, ml_req1 ) |
---|
[51] | 613 | CALL mpprecv( 3, t2ns(1,1,2), imigr ) |
---|
[300] | 614 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
[51] | 615 | CASE ( 0 ) |
---|
[181] | 616 | CALL mppsend( 3, t2ns(1,1,1), imigr, noso, ml_req1 ) |
---|
| 617 | CALL mppsend( 4, t2sn(1,1,1), imigr, nono, ml_req2 ) |
---|
[51] | 618 | CALL mpprecv( 3, t2ns(1,1,2), imigr ) |
---|
| 619 | CALL mpprecv( 4, t2sn(1,1,2), imigr ) |
---|
[300] | 620 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 621 | IF(l_isend) CALL mpi_wait(ml_req2,ml_stat,ml_err) |
---|
[51] | 622 | CASE ( 1 ) |
---|
[181] | 623 | CALL mppsend( 3, t2ns(1,1,1), imigr, noso, ml_req1 ) |
---|
[51] | 624 | CALL mpprecv( 4, t2sn(1,1,2), imigr ) |
---|
[300] | 625 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
[51] | 626 | END SELECT |
---|
[1344] | 627 | ! |
---|
| 628 | ! ! Write Dirichlet lateral conditions |
---|
[51] | 629 | ijhom = nlcj - jprecj |
---|
[1344] | 630 | ! |
---|
[51] | 631 | SELECT CASE ( nbondj ) |
---|
| 632 | CASE ( -1 ) |
---|
| 633 | DO jl = 1, jprecj |
---|
| 634 | pt2d(:,ijhom+jl) = t2ns(:,jl,2) |
---|
| 635 | END DO |
---|
| 636 | CASE ( 0 ) |
---|
| 637 | DO jl = 1, jprecj |
---|
| 638 | pt2d(:,jl ) = t2sn(:,jl,2) |
---|
| 639 | pt2d(:,ijhom+jl) = t2ns(:,jl,2) |
---|
| 640 | END DO |
---|
| 641 | CASE ( 1 ) |
---|
| 642 | DO jl = 1, jprecj |
---|
| 643 | pt2d(:,jl ) = t2sn(:,jl,2) |
---|
| 644 | END DO |
---|
[1344] | 645 | END SELECT |
---|
[3] | 646 | |
---|
[1344] | 647 | |
---|
[51] | 648 | ! 4. north fold treatment |
---|
| 649 | ! ----------------------- |
---|
[1344] | 650 | ! |
---|
| 651 | IF( npolj /= 0 .AND. .NOT. PRESENT(cd_mpp) ) THEN |
---|
| 652 | ! |
---|
| 653 | SELECT CASE ( jpni ) |
---|
| 654 | CASE ( 1 ) ; CALL lbc_nfd ( pt2d, cd_type, psgn ) ! only 1 northern proc, no mpp |
---|
| 655 | CASE DEFAULT ; CALL mpp_lbc_north( pt2d, cd_type, psgn ) ! for all northern procs. |
---|
| 656 | END SELECT |
---|
| 657 | ! |
---|
[473] | 658 | ENDIF |
---|
[1344] | 659 | ! |
---|
[51] | 660 | END SUBROUTINE mpp_lnk_2d |
---|
[3] | 661 | |
---|
| 662 | |
---|
[473] | 663 | SUBROUTINE mpp_lnk_3d_gather( ptab1, cd_type1, ptab2, cd_type2, psgn ) |
---|
| 664 | !!---------------------------------------------------------------------- |
---|
| 665 | !! *** routine mpp_lnk_3d_gather *** |
---|
| 666 | !! |
---|
| 667 | !! ** Purpose : Message passing manadgement for two 3D arrays |
---|
| 668 | !! |
---|
| 669 | !! ** Method : Use mppsend and mpprecv function for passing mask |
---|
| 670 | !! between processors following neighboring subdomains. |
---|
| 671 | !! domain parameters |
---|
| 672 | !! nlci : first dimension of the local subdomain |
---|
| 673 | !! nlcj : second dimension of the local subdomain |
---|
| 674 | !! nbondi : mark for "east-west local boundary" |
---|
| 675 | !! nbondj : mark for "north-south local boundary" |
---|
| 676 | !! noea : number for local neighboring processors |
---|
| 677 | !! nowe : number for local neighboring processors |
---|
| 678 | !! noso : number for local neighboring processors |
---|
| 679 | !! nono : number for local neighboring processors |
---|
| 680 | !! |
---|
| 681 | !! ** Action : ptab1 and ptab2 with update value at its periphery |
---|
| 682 | !! |
---|
| 683 | !!---------------------------------------------------------------------- |
---|
[1344] | 684 | REAL(wp), DIMENSION(jpi,jpj,jpk), INTENT(inout) :: ptab1 ! first and second 3D array on which |
---|
| 685 | REAL(wp), DIMENSION(jpi,jpj,jpk), INTENT(inout) :: ptab2 ! the boundary condition is applied |
---|
| 686 | CHARACTER(len=1) , INTENT(in ) :: cd_type1 ! nature of ptab1 and ptab2 arrays |
---|
| 687 | CHARACTER(len=1) , INTENT(in ) :: cd_type2 ! i.e. grid-points = T , U , V , F or W points |
---|
| 688 | REAL(wp) , INTENT(in ) :: psgn ! =-1 the sign change across the north fold boundary |
---|
| 689 | !! ! = 1. , the sign is kept |
---|
| 690 | INTEGER :: jl ! dummy loop indices |
---|
| 691 | INTEGER :: imigr, iihom, ijhom ! temporary integers |
---|
| 692 | INTEGER :: ml_req1, ml_req2, ml_err ! for key_mpi_isend |
---|
| 693 | INTEGER, DIMENSION(MPI_STATUS_SIZE) :: ml_stat ! for key_mpi_isend |
---|
[473] | 694 | !!---------------------------------------------------------------------- |
---|
| 695 | |
---|
| 696 | ! 1. standard boundary treatment |
---|
| 697 | ! ------------------------------ |
---|
[1344] | 698 | ! ! East-West boundaries |
---|
| 699 | ! !* Cyclic east-west |
---|
| 700 | IF( nbondi == 2 .AND. (nperio == 1 .OR. nperio == 4 .OR. nperio == 6) ) THEN |
---|
[473] | 701 | ptab1( 1 ,:,:) = ptab1(jpim1,:,:) |
---|
| 702 | ptab1(jpi,:,:) = ptab1( 2 ,:,:) |
---|
| 703 | ptab2( 1 ,:,:) = ptab2(jpim1,:,:) |
---|
| 704 | ptab2(jpi,:,:) = ptab2( 2 ,:,:) |
---|
[1344] | 705 | ELSE !* closed |
---|
| 706 | IF( .NOT. cd_type1 == 'F' ) ptab1( 1 :jpreci,:,:) = 0.e0 ! south except at F-point |
---|
| 707 | IF( .NOT. cd_type2 == 'F' ) ptab2( 1 :jpreci,:,:) = 0.e0 |
---|
| 708 | ptab1(nlci-jpreci+1:jpi ,:,:) = 0.e0 ! north |
---|
| 709 | ptab2(nlci-jpreci+1:jpi ,:,:) = 0.e0 |
---|
[473] | 710 | ENDIF |
---|
| 711 | |
---|
[1344] | 712 | |
---|
| 713 | ! ! North-South boundaries |
---|
| 714 | IF( .NOT. cd_type1 == 'F' ) ptab1(:, 1 :jprecj,:) = 0.e0 ! south except at F-point |
---|
| 715 | IF( .NOT. cd_type2 == 'F' ) ptab2(:, 1 :jprecj,:) = 0.e0 |
---|
| 716 | ptab1(:,nlcj-jprecj+1:jpj ,:) = 0.e0 ! north |
---|
| 717 | ptab2(:,nlcj-jprecj+1:jpj ,:) = 0.e0 |
---|
[473] | 718 | |
---|
| 719 | |
---|
| 720 | ! 2. East and west directions exchange |
---|
| 721 | ! ------------------------------------ |
---|
[1344] | 722 | ! we play with the neigbours AND the row number because of the periodicity |
---|
| 723 | ! |
---|
| 724 | SELECT CASE ( nbondi ) ! Read Dirichlet lateral conditions |
---|
| 725 | CASE ( -1, 0, 1 ) ! all exept 2 (i.e. close case) |
---|
[473] | 726 | iihom = nlci-nreci |
---|
| 727 | DO jl = 1, jpreci |
---|
| 728 | t4ew(:,jl,:,1,1) = ptab1(jpreci+jl,:,:) |
---|
| 729 | t4we(:,jl,:,1,1) = ptab1(iihom +jl,:,:) |
---|
| 730 | t4ew(:,jl,:,2,1) = ptab2(jpreci+jl,:,:) |
---|
| 731 | t4we(:,jl,:,2,1) = ptab2(iihom +jl,:,:) |
---|
| 732 | END DO |
---|
| 733 | END SELECT |
---|
[1344] | 734 | ! |
---|
| 735 | ! ! Migrations |
---|
[473] | 736 | imigr = jpreci * jpj * jpk *2 |
---|
[1344] | 737 | ! |
---|
[473] | 738 | SELECT CASE ( nbondi ) |
---|
| 739 | CASE ( -1 ) |
---|
| 740 | CALL mppsend( 2, t4we(1,1,1,1,1), imigr, noea, ml_req1 ) |
---|
| 741 | CALL mpprecv( 1, t4ew(1,1,1,1,2), imigr ) |
---|
| 742 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 743 | CASE ( 0 ) |
---|
| 744 | CALL mppsend( 1, t4ew(1,1,1,1,1), imigr, nowe, ml_req1 ) |
---|
| 745 | CALL mppsend( 2, t4we(1,1,1,1,1), imigr, noea, ml_req2 ) |
---|
| 746 | CALL mpprecv( 1, t4ew(1,1,1,1,2), imigr ) |
---|
| 747 | CALL mpprecv( 2, t4we(1,1,1,1,2), imigr ) |
---|
| 748 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 749 | IF(l_isend) CALL mpi_wait(ml_req2, ml_stat, ml_err) |
---|
| 750 | CASE ( 1 ) |
---|
| 751 | CALL mppsend( 1, t4ew(1,1,1,1,1), imigr, nowe, ml_req1 ) |
---|
| 752 | CALL mpprecv( 2, t4we(1,1,1,1,2), imigr ) |
---|
| 753 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 754 | END SELECT |
---|
[1344] | 755 | ! |
---|
| 756 | ! ! Write Dirichlet lateral conditions |
---|
| 757 | iihom = nlci - jpreci |
---|
| 758 | ! |
---|
[473] | 759 | SELECT CASE ( nbondi ) |
---|
| 760 | CASE ( -1 ) |
---|
| 761 | DO jl = 1, jpreci |
---|
| 762 | ptab1(iihom+jl,:,:) = t4ew(:,jl,:,1,2) |
---|
| 763 | ptab2(iihom+jl,:,:) = t4ew(:,jl,:,2,2) |
---|
| 764 | END DO |
---|
| 765 | CASE ( 0 ) |
---|
| 766 | DO jl = 1, jpreci |
---|
| 767 | ptab1(jl ,:,:) = t4we(:,jl,:,1,2) |
---|
| 768 | ptab1(iihom+jl,:,:) = t4ew(:,jl,:,1,2) |
---|
| 769 | ptab2(jl ,:,:) = t4we(:,jl,:,2,2) |
---|
| 770 | ptab2(iihom+jl,:,:) = t4ew(:,jl,:,2,2) |
---|
| 771 | END DO |
---|
| 772 | CASE ( 1 ) |
---|
| 773 | DO jl = 1, jpreci |
---|
| 774 | ptab1(jl ,:,:) = t4we(:,jl,:,1,2) |
---|
| 775 | ptab2(jl ,:,:) = t4we(:,jl,:,2,2) |
---|
| 776 | END DO |
---|
| 777 | END SELECT |
---|
| 778 | |
---|
| 779 | |
---|
| 780 | ! 3. North and south directions |
---|
| 781 | ! ----------------------------- |
---|
[1344] | 782 | ! always closed : we play only with the neigbours |
---|
| 783 | ! |
---|
| 784 | IF( nbondj /= 2 ) THEN ! Read Dirichlet lateral conditions |
---|
| 785 | ijhom = nlcj - nrecj |
---|
[473] | 786 | DO jl = 1, jprecj |
---|
| 787 | t4sn(:,jl,:,1,1) = ptab1(:,ijhom +jl,:) |
---|
| 788 | t4ns(:,jl,:,1,1) = ptab1(:,jprecj+jl,:) |
---|
| 789 | t4sn(:,jl,:,2,1) = ptab2(:,ijhom +jl,:) |
---|
| 790 | t4ns(:,jl,:,2,1) = ptab2(:,jprecj+jl,:) |
---|
| 791 | END DO |
---|
| 792 | ENDIF |
---|
[1344] | 793 | ! |
---|
| 794 | ! ! Migrations |
---|
[473] | 795 | imigr = jprecj * jpi * jpk * 2 |
---|
[1344] | 796 | ! |
---|
[473] | 797 | SELECT CASE ( nbondj ) |
---|
| 798 | CASE ( -1 ) |
---|
| 799 | CALL mppsend( 4, t4sn(1,1,1,1,1), imigr, nono, ml_req1 ) |
---|
| 800 | CALL mpprecv( 3, t4ns(1,1,1,1,2), imigr ) |
---|
| 801 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 802 | CASE ( 0 ) |
---|
| 803 | CALL mppsend( 3, t4ns(1,1,1,1,1), imigr, noso, ml_req1 ) |
---|
| 804 | CALL mppsend( 4, t4sn(1,1,1,1,1), imigr, nono, ml_req2 ) |
---|
| 805 | CALL mpprecv( 3, t4ns(1,1,1,1,2), imigr ) |
---|
| 806 | CALL mpprecv( 4, t4sn(1,1,1,1,2), imigr ) |
---|
| 807 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 808 | IF(l_isend) CALL mpi_wait(ml_req2, ml_stat, ml_err) |
---|
| 809 | CASE ( 1 ) |
---|
| 810 | CALL mppsend( 3, t4ns(1,1,1,1,1), imigr, noso, ml_req1 ) |
---|
| 811 | CALL mpprecv( 4, t4sn(1,1,1,1,2), imigr ) |
---|
| 812 | IF(l_isend) CALL mpi_wait(ml_req1, ml_stat, ml_err) |
---|
| 813 | END SELECT |
---|
[1344] | 814 | ! |
---|
| 815 | ! ! Write Dirichlet lateral conditions |
---|
| 816 | ijhom = nlcj - jprecj |
---|
| 817 | ! |
---|
[473] | 818 | SELECT CASE ( nbondj ) |
---|
| 819 | CASE ( -1 ) |
---|
| 820 | DO jl = 1, jprecj |
---|
| 821 | ptab1(:,ijhom+jl,:) = t4ns(:,jl,:,1,2) |
---|
| 822 | ptab2(:,ijhom+jl,:) = t4ns(:,jl,:,2,2) |
---|
| 823 | END DO |
---|
| 824 | CASE ( 0 ) |
---|
| 825 | DO jl = 1, jprecj |
---|
| 826 | ptab1(:,jl ,:) = t4sn(:,jl,:,1,2) |
---|
| 827 | ptab1(:,ijhom+jl,:) = t4ns(:,jl,:,1,2) |
---|
| 828 | ptab2(:,jl ,:) = t4sn(:,jl,:,2,2) |
---|
| 829 | ptab2(:,ijhom+jl,:) = t4ns(:,jl,:,2,2) |
---|
| 830 | END DO |
---|
| 831 | CASE ( 1 ) |
---|
| 832 | DO jl = 1, jprecj |
---|
| 833 | ptab1(:,jl,:) = t4sn(:,jl,:,1,2) |
---|
| 834 | ptab2(:,jl,:) = t4sn(:,jl,:,2,2) |
---|
| 835 | END DO |
---|
| 836 | END SELECT |
---|
| 837 | |
---|
| 838 | |
---|
| 839 | ! 4. north fold treatment |
---|
| 840 | ! ----------------------- |
---|
[1344] | 841 | IF( npolj /= 0 ) THEN |
---|
| 842 | ! |
---|
| 843 | SELECT CASE ( jpni ) |
---|
| 844 | CASE ( 1 ) |
---|
| 845 | CALL lbc_nfd ( ptab1, cd_type1, psgn ) ! only for northern procs. |
---|
| 846 | CALL lbc_nfd ( ptab2, cd_type2, psgn ) |
---|
| 847 | CASE DEFAULT |
---|
| 848 | CALL mpp_lbc_north( ptab1, cd_type1, psgn ) ! for all northern procs. |
---|
| 849 | CALL mpp_lbc_north (ptab2, cd_type2, psgn) |
---|
| 850 | END SELECT |
---|
| 851 | ! |
---|
| 852 | ENDIF |
---|
| 853 | ! |
---|
[473] | 854 | END SUBROUTINE mpp_lnk_3d_gather |
---|
| 855 | |
---|
| 856 | |
---|
[311] | 857 | SUBROUTINE mpp_lnk_2d_e( pt2d, cd_type, psgn ) |
---|
| 858 | !!---------------------------------------------------------------------- |
---|
| 859 | !! *** routine mpp_lnk_2d_e *** |
---|
| 860 | !! |
---|
| 861 | !! ** Purpose : Message passing manadgement for 2d array (with halo) |
---|
| 862 | !! |
---|
| 863 | !! ** Method : Use mppsend and mpprecv function for passing mask |
---|
| 864 | !! between processors following neighboring subdomains. |
---|
| 865 | !! domain parameters |
---|
| 866 | !! nlci : first dimension of the local subdomain |
---|
| 867 | !! nlcj : second dimension of the local subdomain |
---|
| 868 | !! jpr2di : number of rows for extra outer halo |
---|
| 869 | !! jpr2dj : number of columns for extra outer halo |
---|
| 870 | !! nbondi : mark for "east-west local boundary" |
---|
| 871 | !! nbondj : mark for "north-south local boundary" |
---|
| 872 | !! noea : number for local neighboring processors |
---|
| 873 | !! nowe : number for local neighboring processors |
---|
| 874 | !! noso : number for local neighboring processors |
---|
| 875 | !! nono : number for local neighboring processors |
---|
| 876 | !! |
---|
| 877 | !!---------------------------------------------------------------------- |
---|
[1344] | 878 | REAL(wp), DIMENSION(1-jpr2di:jpi+jpr2di,1-jpr2dj:jpj+jpr2dj), INTENT(inout) :: pt2d ! 2D array with extra halo |
---|
| 879 | CHARACTER(len=1) , INTENT(in ) :: cd_type ! nature of ptab array grid-points |
---|
| 880 | ! ! = T , U , V , F , W and I points |
---|
| 881 | REAL(wp) , INTENT(in ) :: psgn ! =-1 the sign change across the |
---|
| 882 | !! ! north boundary, = 1. otherwise |
---|
| 883 | INTEGER :: jl ! dummy loop indices |
---|
| 884 | INTEGER :: imigr, iihom, ijhom ! temporary integers |
---|
| 885 | INTEGER :: ipreci, iprecj ! temporary integers |
---|
| 886 | INTEGER :: ml_req1, ml_req2, ml_err ! for key_mpi_isend |
---|
| 887 | INTEGER, DIMENSION(MPI_STATUS_SIZE) :: ml_stat ! for key_mpi_isend |
---|
| 888 | !!---------------------------------------------------------------------- |
---|
[311] | 889 | |
---|
[1344] | 890 | ipreci = jpreci + jpr2di ! take into account outer extra 2D overlap area |
---|
[311] | 891 | iprecj = jprecj + jpr2dj |
---|
| 892 | |
---|
| 893 | |
---|
| 894 | ! 1. standard boundary treatment |
---|
| 895 | ! ------------------------------ |
---|
[1344] | 896 | ! Order matters Here !!!! |
---|
| 897 | ! |
---|
| 898 | ! !* North-South boundaries (always colsed) |
---|
| 899 | IF( .NOT. cd_type == 'F' ) pt2d(:, 1-jpr2dj : jprecj ) = 0.e0 ! south except at F-point |
---|
| 900 | pt2d(:,nlcj-jprecj+1:jpj+jpr2dj) = 0.e0 ! north |
---|
| 901 | |
---|
| 902 | ! ! East-West boundaries |
---|
| 903 | ! !* Cyclic east-west |
---|
| 904 | IF( nbondi == 2 .AND. (nperio == 1 .OR. nperio == 4 .OR. nperio == 6) ) THEN |
---|
| 905 | pt2d(1-jpr2di: 1 ,:) = pt2d(jpim1-jpr2di: jpim1 ,:) ! east |
---|
| 906 | pt2d( jpi :jpi+jpr2di,:) = pt2d( 2 :2+jpr2di,:) ! west |
---|
| 907 | ! |
---|
| 908 | ELSE !* closed |
---|
| 909 | IF( .NOT. cd_type == 'F' ) pt2d( 1-jpr2di :jpreci ,:) = 0.e0 ! south except at F-point |
---|
| 910 | pt2d(nlci-jpreci+1:jpi+jpr2di,:) = 0.e0 ! north |
---|
| 911 | ENDIF |
---|
| 912 | ! |
---|
[311] | 913 | |
---|
[1344] | 914 | ! north fold treatment |
---|
| 915 | ! ----------------------- |
---|
| 916 | IF( npolj /= 0 ) THEN |
---|
| 917 | ! |
---|
| 918 | SELECT CASE ( jpni ) |
---|
| 919 | CASE ( 1 ) ; CALL lbc_nfd ( pt2d(1:jpi,1:jpj+jpr2dj), cd_type, psgn, pr2dj=jpr2dj ) |
---|
| 920 | CASE DEFAULT ; CALL mpp_lbc_north_e( pt2d , cd_type, psgn ) |
---|
| 921 | END SELECT |
---|
| 922 | ! |
---|
[311] | 923 | ENDIF |
---|
| 924 | |
---|
[1344] | 925 | ! 2. East and west directions exchange |
---|
| 926 | ! ------------------------------------ |
---|
| 927 | ! we play with the neigbours AND the row number because of the periodicity |
---|
| 928 | ! |
---|
| 929 | SELECT CASE ( nbondi ) ! Read Dirichlet lateral conditions |
---|
| 930 | CASE ( -1, 0, 1 ) ! all exept 2 (i.e. close case) |
---|
[311] | 931 | iihom = nlci-nreci-jpr2di |
---|
| 932 | DO jl = 1, ipreci |
---|
| 933 | tr2ew(:,jl,1) = pt2d(jpreci+jl,:) |
---|
| 934 | tr2we(:,jl,1) = pt2d(iihom +jl,:) |
---|
| 935 | END DO |
---|
| 936 | END SELECT |
---|
[1344] | 937 | ! |
---|
| 938 | ! ! Migrations |
---|
[311] | 939 | imigr = ipreci * ( jpj + 2*jpr2dj) |
---|
[1344] | 940 | ! |
---|
[311] | 941 | SELECT CASE ( nbondi ) |
---|
| 942 | CASE ( -1 ) |
---|
| 943 | CALL mppsend( 2, tr2we(1-jpr2dj,1,1), imigr, noea, ml_req1 ) |
---|
| 944 | CALL mpprecv( 1, tr2ew(1-jpr2dj,1,2), imigr ) |
---|
| 945 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 946 | CASE ( 0 ) |
---|
| 947 | CALL mppsend( 1, tr2ew(1-jpr2dj,1,1), imigr, nowe, ml_req1 ) |
---|
| 948 | CALL mppsend( 2, tr2we(1-jpr2dj,1,1), imigr, noea, ml_req2 ) |
---|
| 949 | CALL mpprecv( 1, tr2ew(1-jpr2dj,1,2), imigr ) |
---|
| 950 | CALL mpprecv( 2, tr2we(1-jpr2dj,1,2), imigr ) |
---|
| 951 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 952 | IF(l_isend) CALL mpi_wait(ml_req2,ml_stat,ml_err) |
---|
| 953 | CASE ( 1 ) |
---|
| 954 | CALL mppsend( 1, tr2ew(1-jpr2dj,1,1), imigr, nowe, ml_req1 ) |
---|
| 955 | CALL mpprecv( 2, tr2we(1-jpr2dj,1,2), imigr ) |
---|
| 956 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 957 | END SELECT |
---|
[1344] | 958 | ! |
---|
| 959 | ! ! Write Dirichlet lateral conditions |
---|
[311] | 960 | iihom = nlci - jpreci |
---|
[1344] | 961 | ! |
---|
[311] | 962 | SELECT CASE ( nbondi ) |
---|
| 963 | CASE ( -1 ) |
---|
| 964 | DO jl = 1, ipreci |
---|
| 965 | pt2d(iihom+jl,:) = tr2ew(:,jl,2) |
---|
| 966 | END DO |
---|
| 967 | CASE ( 0 ) |
---|
| 968 | DO jl = 1, ipreci |
---|
| 969 | pt2d(jl-jpr2di,:) = tr2we(:,jl,2) |
---|
| 970 | pt2d( iihom+jl,:) = tr2ew(:,jl,2) |
---|
| 971 | END DO |
---|
| 972 | CASE ( 1 ) |
---|
| 973 | DO jl = 1, ipreci |
---|
| 974 | pt2d(jl-jpr2di,:) = tr2we(:,jl,2) |
---|
| 975 | END DO |
---|
| 976 | END SELECT |
---|
| 977 | |
---|
| 978 | |
---|
| 979 | ! 3. North and south directions |
---|
| 980 | ! ----------------------------- |
---|
[1344] | 981 | ! always closed : we play only with the neigbours |
---|
| 982 | ! |
---|
| 983 | IF( nbondj /= 2 ) THEN ! Read Dirichlet lateral conditions |
---|
[311] | 984 | ijhom = nlcj-nrecj-jpr2dj |
---|
| 985 | DO jl = 1, iprecj |
---|
| 986 | tr2sn(:,jl,1) = pt2d(:,ijhom +jl) |
---|
| 987 | tr2ns(:,jl,1) = pt2d(:,jprecj+jl) |
---|
| 988 | END DO |
---|
| 989 | ENDIF |
---|
[1344] | 990 | ! |
---|
| 991 | ! ! Migrations |
---|
[311] | 992 | imigr = iprecj * ( jpi + 2*jpr2di ) |
---|
[1344] | 993 | ! |
---|
[311] | 994 | SELECT CASE ( nbondj ) |
---|
| 995 | CASE ( -1 ) |
---|
| 996 | CALL mppsend( 4, tr2sn(1-jpr2di,1,1), imigr, nono, ml_req1 ) |
---|
| 997 | CALL mpprecv( 3, tr2ns(1-jpr2di,1,2), imigr ) |
---|
| 998 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 999 | CASE ( 0 ) |
---|
| 1000 | CALL mppsend( 3, tr2ns(1-jpr2di,1,1), imigr, noso, ml_req1 ) |
---|
| 1001 | CALL mppsend( 4, tr2sn(1-jpr2di,1,1), imigr, nono, ml_req2 ) |
---|
| 1002 | CALL mpprecv( 3, tr2ns(1-jpr2di,1,2), imigr ) |
---|
| 1003 | CALL mpprecv( 4, tr2sn(1-jpr2di,1,2), imigr ) |
---|
| 1004 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 1005 | IF(l_isend) CALL mpi_wait(ml_req2,ml_stat,ml_err) |
---|
| 1006 | CASE ( 1 ) |
---|
| 1007 | CALL mppsend( 3, tr2ns(1-jpr2di,1,1), imigr, noso, ml_req1 ) |
---|
| 1008 | CALL mpprecv( 4, tr2sn(1-jpr2di,1,2), imigr ) |
---|
| 1009 | IF(l_isend) CALL mpi_wait(ml_req1,ml_stat,ml_err) |
---|
| 1010 | END SELECT |
---|
[1344] | 1011 | ! |
---|
| 1012 | ! ! Write Dirichlet lateral conditions |
---|
[311] | 1013 | ijhom = nlcj - jprecj |
---|
[1344] | 1014 | ! |
---|
[311] | 1015 | SELECT CASE ( nbondj ) |
---|
| 1016 | CASE ( -1 ) |
---|
| 1017 | DO jl = 1, iprecj |
---|
| 1018 | pt2d(:,ijhom+jl) = tr2ns(:,jl,2) |
---|
| 1019 | END DO |
---|
| 1020 | CASE ( 0 ) |
---|
| 1021 | DO jl = 1, iprecj |
---|
| 1022 | pt2d(:,jl-jpr2dj) = tr2sn(:,jl,2) |
---|
| 1023 | pt2d(:,ijhom+jl ) = tr2ns(:,jl,2) |
---|
| 1024 | END DO |
---|
| 1025 | CASE ( 1 ) |
---|
| 1026 | DO jl = 1, iprecj |
---|
| 1027 | pt2d(:,jl-jpr2dj) = tr2sn(:,jl,2) |
---|
| 1028 | END DO |
---|
[1344] | 1029 | END SELECT |
---|
[311] | 1030 | |
---|
| 1031 | END SUBROUTINE mpp_lnk_2d_e |
---|
| 1032 | |
---|
| 1033 | |
---|
[1344] | 1034 | SUBROUTINE mppsend( ktyp, pmess, kbytes, kdest, md_req ) |
---|
[51] | 1035 | !!---------------------------------------------------------------------- |
---|
| 1036 | !! *** routine mppsend *** |
---|
| 1037 | !! |
---|
| 1038 | !! ** Purpose : Send messag passing array |
---|
| 1039 | !! |
---|
| 1040 | !!---------------------------------------------------------------------- |
---|
[1344] | 1041 | REAL(wp), INTENT(inout) :: pmess(*) ! array of real |
---|
| 1042 | INTEGER , INTENT(in ) :: kbytes ! size of the array pmess |
---|
| 1043 | INTEGER , INTENT(in ) :: kdest ! receive process number |
---|
| 1044 | INTEGER , INTENT(in ) :: ktyp ! tag of the message |
---|
| 1045 | INTEGER , INTENT(in ) :: md_req ! argument for isend |
---|
| 1046 | !! |
---|
| 1047 | INTEGER :: iflag |
---|
[51] | 1048 | !!---------------------------------------------------------------------- |
---|
[1344] | 1049 | ! |
---|
[1601] | 1050 | SELECT CASE ( cn_mpi_send ) |
---|
[300] | 1051 | CASE ( 'S' ) ! Standard mpi send (blocking) |
---|
[1344] | 1052 | CALL mpi_send ( pmess, kbytes, mpi_double_precision, kdest , ktyp, mpi_comm_opa , iflag ) |
---|
[300] | 1053 | CASE ( 'B' ) ! Buffer mpi send (blocking) |
---|
[1344] | 1054 | CALL mpi_bsend( pmess, kbytes, mpi_double_precision, kdest , ktyp, mpi_comm_opa , iflag ) |
---|
[300] | 1055 | CASE ( 'I' ) ! Immediate mpi send (non-blocking send) |
---|
[1344] | 1056 | ! be carefull, one more argument here : the mpi request identifier.. |
---|
| 1057 | CALL mpi_isend( pmess, kbytes, mpi_double_precision, kdest , ktyp, mpi_comm_opa, md_req, iflag ) |
---|
[300] | 1058 | END SELECT |
---|
[1344] | 1059 | ! |
---|
[51] | 1060 | END SUBROUTINE mppsend |
---|
[3] | 1061 | |
---|
| 1062 | |
---|
[51] | 1063 | SUBROUTINE mpprecv( ktyp, pmess, kbytes ) |
---|
| 1064 | !!---------------------------------------------------------------------- |
---|
| 1065 | !! *** routine mpprecv *** |
---|
| 1066 | !! |
---|
| 1067 | !! ** Purpose : Receive messag passing array |
---|
| 1068 | !! |
---|
| 1069 | !!---------------------------------------------------------------------- |
---|
[1344] | 1070 | REAL(wp), INTENT(inout) :: pmess(*) ! array of real |
---|
| 1071 | INTEGER , INTENT(in ) :: kbytes ! suze of the array pmess |
---|
| 1072 | INTEGER , INTENT(in ) :: ktyp ! Tag of the recevied message |
---|
| 1073 | !! |
---|
[51] | 1074 | INTEGER :: istatus(mpi_status_size) |
---|
| 1075 | INTEGER :: iflag |
---|
[1344] | 1076 | !!---------------------------------------------------------------------- |
---|
| 1077 | ! |
---|
| 1078 | CALL mpi_recv( pmess, kbytes, mpi_double_precision, mpi_any_source, ktyp, mpi_comm_opa, istatus, iflag ) |
---|
| 1079 | ! |
---|
[51] | 1080 | END SUBROUTINE mpprecv |
---|
[3] | 1081 | |
---|
| 1082 | |
---|
[51] | 1083 | SUBROUTINE mppgather( ptab, kp, pio ) |
---|
| 1084 | !!---------------------------------------------------------------------- |
---|
| 1085 | !! *** routine mppgather *** |
---|
| 1086 | !! |
---|
| 1087 | !! ** Purpose : Transfert between a local subdomain array and a work |
---|
| 1088 | !! array which is distributed following the vertical level. |
---|
| 1089 | !! |
---|
[1344] | 1090 | !!---------------------------------------------------------------------- |
---|
| 1091 | REAL(wp), DIMENSION(jpi,jpj), INTENT(in ) :: ptab ! subdomain input array |
---|
| 1092 | INTEGER , INTENT(in ) :: kp ! record length |
---|
| 1093 | REAL(wp), DIMENSION(jpi,jpj,jpnij), INTENT( out) :: pio ! subdomain input array |
---|
[51] | 1094 | !! |
---|
[1344] | 1095 | INTEGER :: itaille, ierror ! temporary integer |
---|
[51] | 1096 | !!--------------------------------------------------------------------- |
---|
[1344] | 1097 | ! |
---|
| 1098 | itaille = jpi * jpj |
---|
| 1099 | CALL mpi_gather( ptab, itaille, mpi_double_precision, pio, itaille , & |
---|
[532] | 1100 | & mpi_double_precision, kp , mpi_comm_opa, ierror ) |
---|
[1344] | 1101 | ! |
---|
[51] | 1102 | END SUBROUTINE mppgather |
---|
[3] | 1103 | |
---|
| 1104 | |
---|
[51] | 1105 | SUBROUTINE mppscatter( pio, kp, ptab ) |
---|
| 1106 | !!---------------------------------------------------------------------- |
---|
| 1107 | !! *** routine mppscatter *** |
---|
| 1108 | !! |
---|
| 1109 | !! ** Purpose : Transfert between awork array which is distributed |
---|
| 1110 | !! following the vertical level and the local subdomain array. |
---|
| 1111 | !! |
---|
| 1112 | !!---------------------------------------------------------------------- |
---|
| 1113 | REAL(wp), DIMENSION(jpi,jpj,jpnij) :: pio ! output array |
---|
| 1114 | INTEGER :: kp ! Tag (not used with MPI |
---|
| 1115 | REAL(wp), DIMENSION(jpi,jpj) :: ptab ! subdomain array input |
---|
[1344] | 1116 | !! |
---|
| 1117 | INTEGER :: itaille, ierror ! temporary integer |
---|
[51] | 1118 | !!--------------------------------------------------------------------- |
---|
[1344] | 1119 | ! |
---|
[51] | 1120 | itaille=jpi*jpj |
---|
[1344] | 1121 | ! |
---|
| 1122 | CALL mpi_scatter( pio, itaille, mpi_double_precision, ptab, itaille , & |
---|
| 1123 | & mpi_double_precision, kp , mpi_comm_opa, ierror ) |
---|
| 1124 | ! |
---|
[51] | 1125 | END SUBROUTINE mppscatter |
---|
[3] | 1126 | |
---|
| 1127 | |
---|
[869] | 1128 | SUBROUTINE mppmax_a_int( ktab, kdim, kcom ) |
---|
[681] | 1129 | !!---------------------------------------------------------------------- |
---|
| 1130 | !! *** routine mppmax_a_int *** |
---|
| 1131 | !! |
---|
| 1132 | !! ** Purpose : Find maximum value in an integer layout array |
---|
| 1133 | !! |
---|
| 1134 | !!---------------------------------------------------------------------- |
---|
[1344] | 1135 | INTEGER , INTENT(in ) :: kdim ! size of array |
---|
| 1136 | INTEGER , INTENT(inout), DIMENSION(kdim) :: ktab ! input array |
---|
| 1137 | INTEGER , INTENT(in ), OPTIONAL :: kcom ! |
---|
| 1138 | !! |
---|
| 1139 | INTEGER :: ierror, localcomm ! temporary integer |
---|
[681] | 1140 | INTEGER, DIMENSION(kdim) :: iwork |
---|
[1344] | 1141 | !!---------------------------------------------------------------------- |
---|
| 1142 | ! |
---|
[869] | 1143 | localcomm = mpi_comm_opa |
---|
[1344] | 1144 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1145 | ! |
---|
| 1146 | CALL mpi_allreduce( ktab, iwork, kdim, mpi_integer, mpi_max, localcomm, ierror ) |
---|
| 1147 | ! |
---|
[681] | 1148 | ktab(:) = iwork(:) |
---|
[1344] | 1149 | ! |
---|
[681] | 1150 | END SUBROUTINE mppmax_a_int |
---|
| 1151 | |
---|
| 1152 | |
---|
[869] | 1153 | SUBROUTINE mppmax_int( ktab, kcom ) |
---|
[681] | 1154 | !!---------------------------------------------------------------------- |
---|
| 1155 | !! *** routine mppmax_int *** |
---|
| 1156 | !! |
---|
[1344] | 1157 | !! ** Purpose : Find maximum value in an integer layout array |
---|
[681] | 1158 | !! |
---|
| 1159 | !!---------------------------------------------------------------------- |
---|
[1344] | 1160 | INTEGER, INTENT(inout) :: ktab ! ??? |
---|
| 1161 | INTEGER, INTENT(in ), OPTIONAL :: kcom ! ??? |
---|
| 1162 | !! |
---|
| 1163 | INTEGER :: ierror, iwork, localcomm ! temporary integer |
---|
| 1164 | !!---------------------------------------------------------------------- |
---|
| 1165 | ! |
---|
[869] | 1166 | localcomm = mpi_comm_opa |
---|
[1344] | 1167 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1168 | ! |
---|
| 1169 | CALL mpi_allreduce( ktab, iwork, 1, mpi_integer, mpi_max, localcomm, ierror) |
---|
| 1170 | ! |
---|
[681] | 1171 | ktab = iwork |
---|
[1344] | 1172 | ! |
---|
[681] | 1173 | END SUBROUTINE mppmax_int |
---|
| 1174 | |
---|
| 1175 | |
---|
[869] | 1176 | SUBROUTINE mppmin_a_int( ktab, kdim, kcom ) |
---|
[51] | 1177 | !!---------------------------------------------------------------------- |
---|
| 1178 | !! *** routine mppmin_a_int *** |
---|
| 1179 | !! |
---|
| 1180 | !! ** Purpose : Find minimum value in an integer layout array |
---|
| 1181 | !! |
---|
| 1182 | !!---------------------------------------------------------------------- |
---|
| 1183 | INTEGER , INTENT( in ) :: kdim ! size of array |
---|
| 1184 | INTEGER , INTENT(inout), DIMENSION(kdim) :: ktab ! input array |
---|
[888] | 1185 | INTEGER , INTENT( in ), OPTIONAL :: kcom ! input array |
---|
[1344] | 1186 | !! |
---|
| 1187 | INTEGER :: ierror, localcomm ! temporary integer |
---|
[51] | 1188 | INTEGER, DIMENSION(kdim) :: iwork |
---|
[1344] | 1189 | !!---------------------------------------------------------------------- |
---|
| 1190 | ! |
---|
[869] | 1191 | localcomm = mpi_comm_opa |
---|
[1344] | 1192 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1193 | ! |
---|
| 1194 | CALL mpi_allreduce( ktab, iwork, kdim, mpi_integer, mpi_min, localcomm, ierror ) |
---|
| 1195 | ! |
---|
[51] | 1196 | ktab(:) = iwork(:) |
---|
[1344] | 1197 | ! |
---|
[51] | 1198 | END SUBROUTINE mppmin_a_int |
---|
[3] | 1199 | |
---|
[13] | 1200 | |
---|
[1345] | 1201 | SUBROUTINE mppmin_int( ktab, kcom ) |
---|
[51] | 1202 | !!---------------------------------------------------------------------- |
---|
| 1203 | !! *** routine mppmin_int *** |
---|
| 1204 | !! |
---|
[1344] | 1205 | !! ** Purpose : Find minimum value in an integer layout array |
---|
[51] | 1206 | !! |
---|
| 1207 | !!---------------------------------------------------------------------- |
---|
| 1208 | INTEGER, INTENT(inout) :: ktab ! ??? |
---|
[1345] | 1209 | INTEGER , INTENT( in ), OPTIONAL :: kcom ! input array |
---|
[1344] | 1210 | !! |
---|
[1345] | 1211 | INTEGER :: ierror, iwork, localcomm |
---|
[1344] | 1212 | !!---------------------------------------------------------------------- |
---|
| 1213 | ! |
---|
[1345] | 1214 | localcomm = mpi_comm_opa |
---|
| 1215 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
[1344] | 1216 | ! |
---|
[1345] | 1217 | CALL mpi_allreduce( ktab, iwork, 1, mpi_integer, mpi_min, localcomm, ierror ) |
---|
| 1218 | ! |
---|
[51] | 1219 | ktab = iwork |
---|
[1344] | 1220 | ! |
---|
[51] | 1221 | END SUBROUTINE mppmin_int |
---|
[3] | 1222 | |
---|
[13] | 1223 | |
---|
[51] | 1224 | SUBROUTINE mppsum_a_int( ktab, kdim ) |
---|
| 1225 | !!---------------------------------------------------------------------- |
---|
| 1226 | !! *** routine mppsum_a_int *** |
---|
| 1227 | !! |
---|
[1344] | 1228 | !! ** Purpose : Global integer sum, 1D array case |
---|
[51] | 1229 | !! |
---|
| 1230 | !!---------------------------------------------------------------------- |
---|
[1344] | 1231 | INTEGER, INTENT(in ) :: kdim ! ??? |
---|
[51] | 1232 | INTEGER, INTENT(inout), DIMENSION (kdim) :: ktab ! ??? |
---|
[1344] | 1233 | !! |
---|
[51] | 1234 | INTEGER :: ierror |
---|
| 1235 | INTEGER, DIMENSION (kdim) :: iwork |
---|
[1344] | 1236 | !!---------------------------------------------------------------------- |
---|
| 1237 | ! |
---|
| 1238 | CALL mpi_allreduce( ktab, iwork, kdim, mpi_integer, mpi_sum, mpi_comm_opa, ierror ) |
---|
| 1239 | ! |
---|
[51] | 1240 | ktab(:) = iwork(:) |
---|
[1344] | 1241 | ! |
---|
[51] | 1242 | END SUBROUTINE mppsum_a_int |
---|
[3] | 1243 | |
---|
[13] | 1244 | |
---|
[1344] | 1245 | SUBROUTINE mppsum_int( ktab ) |
---|
| 1246 | !!---------------------------------------------------------------------- |
---|
| 1247 | !! *** routine mppsum_int *** |
---|
| 1248 | !! |
---|
| 1249 | !! ** Purpose : Global integer sum |
---|
| 1250 | !! |
---|
| 1251 | !!---------------------------------------------------------------------- |
---|
| 1252 | INTEGER, INTENT(inout) :: ktab |
---|
| 1253 | !! |
---|
| 1254 | INTEGER :: ierror, iwork |
---|
| 1255 | !!---------------------------------------------------------------------- |
---|
| 1256 | ! |
---|
| 1257 | CALL mpi_allreduce( ktab, iwork, 1, mpi_integer, mpi_sum, mpi_comm_opa, ierror ) |
---|
| 1258 | ! |
---|
| 1259 | ktab = iwork |
---|
| 1260 | ! |
---|
| 1261 | END SUBROUTINE mppsum_int |
---|
[3] | 1262 | |
---|
[13] | 1263 | |
---|
[1344] | 1264 | SUBROUTINE mppmax_a_real( ptab, kdim, kcom ) |
---|
| 1265 | !!---------------------------------------------------------------------- |
---|
| 1266 | !! *** routine mppmax_a_real *** |
---|
| 1267 | !! |
---|
| 1268 | !! ** Purpose : Maximum |
---|
| 1269 | !! |
---|
| 1270 | !!---------------------------------------------------------------------- |
---|
| 1271 | INTEGER , INTENT(in ) :: kdim |
---|
| 1272 | REAL(wp), INTENT(inout), DIMENSION(kdim) :: ptab |
---|
| 1273 | INTEGER , INTENT(in ), OPTIONAL :: kcom |
---|
| 1274 | !! |
---|
| 1275 | INTEGER :: ierror, localcomm |
---|
| 1276 | REAL(wp), DIMENSION(kdim) :: zwork |
---|
| 1277 | !!---------------------------------------------------------------------- |
---|
| 1278 | ! |
---|
| 1279 | localcomm = mpi_comm_opa |
---|
| 1280 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1281 | ! |
---|
| 1282 | CALL mpi_allreduce( ptab, zwork, kdim, mpi_double_precision, mpi_max, localcomm, ierror ) |
---|
| 1283 | ptab(:) = zwork(:) |
---|
| 1284 | ! |
---|
| 1285 | END SUBROUTINE mppmax_a_real |
---|
[3] | 1286 | |
---|
| 1287 | |
---|
[1344] | 1288 | SUBROUTINE mppmax_real( ptab, kcom ) |
---|
| 1289 | !!---------------------------------------------------------------------- |
---|
| 1290 | !! *** routine mppmax_real *** |
---|
| 1291 | !! |
---|
| 1292 | !! ** Purpose : Maximum |
---|
| 1293 | !! |
---|
| 1294 | !!---------------------------------------------------------------------- |
---|
| 1295 | REAL(wp), INTENT(inout) :: ptab ! ??? |
---|
| 1296 | INTEGER , INTENT(in ), OPTIONAL :: kcom ! ??? |
---|
| 1297 | !! |
---|
| 1298 | INTEGER :: ierror, localcomm |
---|
| 1299 | REAL(wp) :: zwork |
---|
| 1300 | !!---------------------------------------------------------------------- |
---|
| 1301 | ! |
---|
| 1302 | localcomm = mpi_comm_opa |
---|
| 1303 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1304 | ! |
---|
| 1305 | CALL mpi_allreduce( ptab, zwork, 1, mpi_double_precision, mpi_max, localcomm, ierror ) |
---|
| 1306 | ptab = zwork |
---|
| 1307 | ! |
---|
| 1308 | END SUBROUTINE mppmax_real |
---|
[13] | 1309 | |
---|
[3] | 1310 | |
---|
[1344] | 1311 | SUBROUTINE mppmin_a_real( ptab, kdim, kcom ) |
---|
| 1312 | !!---------------------------------------------------------------------- |
---|
| 1313 | !! *** routine mppmin_a_real *** |
---|
| 1314 | !! |
---|
| 1315 | !! ** Purpose : Minimum of REAL, array case |
---|
| 1316 | !! |
---|
| 1317 | !!----------------------------------------------------------------------- |
---|
| 1318 | INTEGER , INTENT(in ) :: kdim |
---|
| 1319 | REAL(wp), INTENT(inout), DIMENSION(kdim) :: ptab |
---|
| 1320 | INTEGER , INTENT(in ), OPTIONAL :: kcom |
---|
| 1321 | !! |
---|
| 1322 | INTEGER :: ierror, localcomm |
---|
| 1323 | REAL(wp), DIMENSION(kdim) :: zwork |
---|
| 1324 | !!----------------------------------------------------------------------- |
---|
| 1325 | ! |
---|
| 1326 | localcomm = mpi_comm_opa |
---|
| 1327 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1328 | ! |
---|
| 1329 | CALL mpi_allreduce( ptab, zwork, kdim, mpi_double_precision, mpi_min, localcomm, ierror ) |
---|
| 1330 | ptab(:) = zwork(:) |
---|
| 1331 | ! |
---|
| 1332 | END SUBROUTINE mppmin_a_real |
---|
[3] | 1333 | |
---|
| 1334 | |
---|
[1344] | 1335 | SUBROUTINE mppmin_real( ptab, kcom ) |
---|
| 1336 | !!---------------------------------------------------------------------- |
---|
| 1337 | !! *** routine mppmin_real *** |
---|
| 1338 | !! |
---|
| 1339 | !! ** Purpose : minimum of REAL, scalar case |
---|
| 1340 | !! |
---|
| 1341 | !!----------------------------------------------------------------------- |
---|
| 1342 | REAL(wp), INTENT(inout) :: ptab ! |
---|
| 1343 | INTEGER , INTENT(in ), OPTIONAL :: kcom |
---|
| 1344 | !! |
---|
| 1345 | INTEGER :: ierror |
---|
| 1346 | REAL(wp) :: zwork |
---|
| 1347 | INTEGER :: localcomm |
---|
| 1348 | !!----------------------------------------------------------------------- |
---|
| 1349 | ! |
---|
| 1350 | localcomm = mpi_comm_opa |
---|
| 1351 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1352 | ! |
---|
| 1353 | CALL mpi_allreduce( ptab, zwork, 1, mpi_double_precision, mpi_min, localcomm, ierror ) |
---|
| 1354 | ptab = zwork |
---|
| 1355 | ! |
---|
| 1356 | END SUBROUTINE mppmin_real |
---|
[13] | 1357 | |
---|
[3] | 1358 | |
---|
[1344] | 1359 | SUBROUTINE mppsum_a_real( ptab, kdim, kcom ) |
---|
| 1360 | !!---------------------------------------------------------------------- |
---|
| 1361 | !! *** routine mppsum_a_real *** |
---|
| 1362 | !! |
---|
| 1363 | !! ** Purpose : global sum, REAL ARRAY argument case |
---|
| 1364 | !! |
---|
| 1365 | !!----------------------------------------------------------------------- |
---|
| 1366 | INTEGER , INTENT( in ) :: kdim ! size of ptab |
---|
| 1367 | REAL(wp), DIMENSION(kdim), INTENT( inout ) :: ptab ! input array |
---|
| 1368 | INTEGER , INTENT( in ), OPTIONAL :: kcom |
---|
| 1369 | !! |
---|
| 1370 | INTEGER :: ierror ! temporary integer |
---|
| 1371 | INTEGER :: localcomm |
---|
| 1372 | REAL(wp), DIMENSION(kdim) :: zwork ! temporary workspace |
---|
| 1373 | !!----------------------------------------------------------------------- |
---|
| 1374 | ! |
---|
| 1375 | localcomm = mpi_comm_opa |
---|
| 1376 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1377 | ! |
---|
| 1378 | CALL mpi_allreduce( ptab, zwork, kdim, mpi_double_precision, mpi_sum, localcomm, ierror ) |
---|
| 1379 | ptab(:) = zwork(:) |
---|
| 1380 | ! |
---|
| 1381 | END SUBROUTINE mppsum_a_real |
---|
[869] | 1382 | |
---|
[3] | 1383 | |
---|
[1344] | 1384 | SUBROUTINE mppsum_real( ptab, kcom ) |
---|
| 1385 | !!---------------------------------------------------------------------- |
---|
| 1386 | !! *** routine mppsum_real *** |
---|
| 1387 | !! |
---|
| 1388 | !! ** Purpose : global sum, SCALAR argument case |
---|
| 1389 | !! |
---|
| 1390 | !!----------------------------------------------------------------------- |
---|
| 1391 | REAL(wp), INTENT(inout) :: ptab ! input scalar |
---|
| 1392 | INTEGER , INTENT(in ), OPTIONAL :: kcom |
---|
| 1393 | !! |
---|
| 1394 | INTEGER :: ierror, localcomm |
---|
| 1395 | REAL(wp) :: zwork |
---|
| 1396 | !!----------------------------------------------------------------------- |
---|
| 1397 | ! |
---|
| 1398 | localcomm = mpi_comm_opa |
---|
| 1399 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1400 | ! |
---|
| 1401 | CALL mpi_allreduce( ptab, zwork, 1, mpi_double_precision, mpi_sum, localcomm, ierror ) |
---|
| 1402 | ptab = zwork |
---|
| 1403 | ! |
---|
| 1404 | END SUBROUTINE mppsum_real |
---|
[3] | 1405 | |
---|
[2304] | 1406 | # if defined key_mpp_rep |
---|
[1976] | 1407 | SUBROUTINE mppsum_realdd( ytab, kcom ) |
---|
| 1408 | !!---------------------------------------------------------------------- |
---|
| 1409 | !! *** routine mppsum_realdd *** |
---|
| 1410 | !! |
---|
| 1411 | !! ** Purpose : global sum in Massively Parallel Processing |
---|
| 1412 | !! SCALAR argument case for double-double precision |
---|
| 1413 | !! |
---|
| 1414 | !!----------------------------------------------------------------------- |
---|
| 1415 | COMPLEX(wp), INTENT(inout) :: ytab ! input scalar |
---|
| 1416 | INTEGER , INTENT( in ), OPTIONAL :: kcom |
---|
[3] | 1417 | |
---|
[1976] | 1418 | !! * Local variables (MPI version) |
---|
| 1419 | INTEGER :: ierror |
---|
| 1420 | INTEGER :: localcomm |
---|
| 1421 | COMPLEX(wp) :: zwork |
---|
| 1422 | |
---|
| 1423 | localcomm = mpi_comm_opa |
---|
| 1424 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1425 | |
---|
| 1426 | ! reduce local sums into global sum |
---|
| 1427 | CALL MPI_ALLREDUCE (ytab, zwork, 1, MPI_DOUBLE_COMPLEX, & |
---|
| 1428 | MPI_SUMDD,localcomm,ierror) |
---|
| 1429 | ytab = zwork |
---|
| 1430 | |
---|
| 1431 | END SUBROUTINE mppsum_realdd |
---|
| 1432 | |
---|
| 1433 | |
---|
| 1434 | SUBROUTINE mppsum_a_realdd( ytab, kdim, kcom ) |
---|
| 1435 | !!---------------------------------------------------------------------- |
---|
| 1436 | !! *** routine mppsum_a_realdd *** |
---|
| 1437 | !! |
---|
| 1438 | !! ** Purpose : global sum in Massively Parallel Processing |
---|
| 1439 | !! COMPLEX ARRAY case for double-double precision |
---|
| 1440 | !! |
---|
| 1441 | !!----------------------------------------------------------------------- |
---|
| 1442 | INTEGER , INTENT( in ) :: kdim ! size of ytab |
---|
| 1443 | COMPLEX(wp), DIMENSION(kdim), INTENT( inout ) :: ytab ! input array |
---|
| 1444 | INTEGER , INTENT( in ), OPTIONAL :: kcom |
---|
| 1445 | |
---|
| 1446 | !! * Local variables (MPI version) |
---|
| 1447 | INTEGER :: ierror ! temporary integer |
---|
| 1448 | INTEGER :: localcomm |
---|
| 1449 | COMPLEX(wp), DIMENSION(kdim) :: zwork ! temporary workspace |
---|
| 1450 | |
---|
| 1451 | localcomm = mpi_comm_opa |
---|
| 1452 | IF( PRESENT(kcom) ) localcomm = kcom |
---|
| 1453 | |
---|
| 1454 | CALL MPI_ALLREDUCE (ytab, zwork, kdim, MPI_DOUBLE_COMPLEX, & |
---|
| 1455 | MPI_SUMDD,localcomm,ierror) |
---|
| 1456 | ytab(:) = zwork(:) |
---|
| 1457 | |
---|
| 1458 | END SUBROUTINE mppsum_a_realdd |
---|
| 1459 | # endif |
---|
| 1460 | |
---|
[1344] | 1461 | SUBROUTINE mpp_minloc2d( ptab, pmask, pmin, ki,kj ) |
---|
| 1462 | !!------------------------------------------------------------------------ |
---|
| 1463 | !! *** routine mpp_minloc *** |
---|
| 1464 | !! |
---|
| 1465 | !! ** Purpose : Compute the global minimum of an array ptab |
---|
| 1466 | !! and also give its global position |
---|
| 1467 | !! |
---|
| 1468 | !! ** Method : Use MPI_ALLREDUCE with MPI_MINLOC |
---|
| 1469 | !! |
---|
| 1470 | !!-------------------------------------------------------------------------- |
---|
| 1471 | REAL(wp), DIMENSION (jpi,jpj), INTENT(in ) :: ptab ! Local 2D array |
---|
| 1472 | REAL(wp), DIMENSION (jpi,jpj), INTENT(in ) :: pmask ! Local mask |
---|
| 1473 | REAL(wp) , INTENT( out) :: pmin ! Global minimum of ptab |
---|
| 1474 | INTEGER , INTENT( out) :: ki, kj ! index of minimum in global frame |
---|
| 1475 | !! |
---|
| 1476 | INTEGER , DIMENSION(2) :: ilocs |
---|
| 1477 | INTEGER :: ierror |
---|
| 1478 | REAL(wp) :: zmin ! local minimum |
---|
| 1479 | REAL(wp), DIMENSION(2,1) :: zain, zaout |
---|
| 1480 | !!----------------------------------------------------------------------- |
---|
| 1481 | ! |
---|
| 1482 | zmin = MINVAL( ptab(:,:) , mask= pmask == 1.e0 ) |
---|
| 1483 | ilocs = MINLOC( ptab(:,:) , mask= pmask == 1.e0 ) |
---|
| 1484 | ! |
---|
| 1485 | ki = ilocs(1) + nimpp - 1 |
---|
| 1486 | kj = ilocs(2) + njmpp - 1 |
---|
| 1487 | ! |
---|
| 1488 | zain(1,:)=zmin |
---|
| 1489 | zain(2,:)=ki+10000.*kj |
---|
| 1490 | ! |
---|
| 1491 | CALL MPI_ALLREDUCE( zain,zaout, 1, MPI_2DOUBLE_PRECISION,MPI_MINLOC,MPI_COMM_OPA,ierror) |
---|
| 1492 | ! |
---|
| 1493 | pmin = zaout(1,1) |
---|
| 1494 | kj = INT(zaout(2,1)/10000.) |
---|
| 1495 | ki = INT(zaout(2,1) - 10000.*kj ) |
---|
| 1496 | ! |
---|
| 1497 | END SUBROUTINE mpp_minloc2d |
---|
[13] | 1498 | |
---|
[3] | 1499 | |
---|
[1344] | 1500 | SUBROUTINE mpp_minloc3d( ptab, pmask, pmin, ki, kj ,kk) |
---|
| 1501 | !!------------------------------------------------------------------------ |
---|
| 1502 | !! *** routine mpp_minloc *** |
---|
| 1503 | !! |
---|
| 1504 | !! ** Purpose : Compute the global minimum of an array ptab |
---|
| 1505 | !! and also give its global position |
---|
| 1506 | !! |
---|
| 1507 | !! ** Method : Use MPI_ALLREDUCE with MPI_MINLOC |
---|
| 1508 | !! |
---|
| 1509 | !!-------------------------------------------------------------------------- |
---|
| 1510 | REAL(wp), DIMENSION (jpi,jpj,jpk), INTENT(in ) :: ptab ! Local 2D array |
---|
| 1511 | REAL(wp), DIMENSION (jpi,jpj,jpk), INTENT(in ) :: pmask ! Local mask |
---|
| 1512 | REAL(wp) , INTENT( out) :: pmin ! Global minimum of ptab |
---|
| 1513 | INTEGER , INTENT( out) :: ki, kj, kk ! index of minimum in global frame |
---|
| 1514 | !! |
---|
| 1515 | INTEGER :: ierror |
---|
| 1516 | REAL(wp) :: zmin ! local minimum |
---|
| 1517 | INTEGER , DIMENSION(3) :: ilocs |
---|
| 1518 | REAL(wp), DIMENSION(2,1) :: zain, zaout |
---|
| 1519 | !!----------------------------------------------------------------------- |
---|
| 1520 | ! |
---|
| 1521 | zmin = MINVAL( ptab(:,:,:) , mask= pmask == 1.e0 ) |
---|
| 1522 | ilocs = MINLOC( ptab(:,:,:) , mask= pmask == 1.e0 ) |
---|
| 1523 | ! |
---|
| 1524 | ki = ilocs(1) + nimpp - 1 |
---|
| 1525 | kj = ilocs(2) + njmpp - 1 |
---|
| 1526 | kk = ilocs(3) |
---|
| 1527 | ! |
---|
| 1528 | zain(1,:)=zmin |
---|
| 1529 | zain(2,:)=ki+10000.*kj+100000000.*kk |
---|
| 1530 | ! |
---|
| 1531 | CALL MPI_ALLREDUCE( zain,zaout, 1, MPI_2DOUBLE_PRECISION,MPI_MINLOC,MPI_COMM_OPA,ierror) |
---|
| 1532 | ! |
---|
| 1533 | pmin = zaout(1,1) |
---|
| 1534 | kk = INT( zaout(2,1) / 100000000. ) |
---|
| 1535 | kj = INT( zaout(2,1) - kk * 100000000. ) / 10000 |
---|
| 1536 | ki = INT( zaout(2,1) - kk * 100000000. -kj * 10000. ) |
---|
| 1537 | ! |
---|
| 1538 | END SUBROUTINE mpp_minloc3d |
---|
[13] | 1539 | |
---|
[3] | 1540 | |
---|
[1344] | 1541 | SUBROUTINE mpp_maxloc2d( ptab, pmask, pmax, ki, kj ) |
---|
| 1542 | !!------------------------------------------------------------------------ |
---|
| 1543 | !! *** routine mpp_maxloc *** |
---|
| 1544 | !! |
---|
| 1545 | !! ** Purpose : Compute the global maximum of an array ptab |
---|
| 1546 | !! and also give its global position |
---|
| 1547 | !! |
---|
| 1548 | !! ** Method : Use MPI_ALLREDUCE with MPI_MINLOC |
---|
| 1549 | !! |
---|
| 1550 | !!-------------------------------------------------------------------------- |
---|
| 1551 | REAL(wp), DIMENSION (jpi,jpj), INTENT(in ) :: ptab ! Local 2D array |
---|
| 1552 | REAL(wp), DIMENSION (jpi,jpj), INTENT(in ) :: pmask ! Local mask |
---|
| 1553 | REAL(wp) , INTENT( out) :: pmax ! Global maximum of ptab |
---|
| 1554 | INTEGER , INTENT( out) :: ki, kj ! index of maximum in global frame |
---|
| 1555 | !! |
---|
| 1556 | INTEGER :: ierror |
---|
| 1557 | INTEGER, DIMENSION (2) :: ilocs |
---|
| 1558 | REAL(wp) :: zmax ! local maximum |
---|
| 1559 | REAL(wp), DIMENSION(2,1) :: zain, zaout |
---|
| 1560 | !!----------------------------------------------------------------------- |
---|
| 1561 | ! |
---|
| 1562 | zmax = MAXVAL( ptab(:,:) , mask= pmask == 1.e0 ) |
---|
| 1563 | ilocs = MAXLOC( ptab(:,:) , mask= pmask == 1.e0 ) |
---|
| 1564 | ! |
---|
| 1565 | ki = ilocs(1) + nimpp - 1 |
---|
| 1566 | kj = ilocs(2) + njmpp - 1 |
---|
| 1567 | ! |
---|
| 1568 | zain(1,:) = zmax |
---|
| 1569 | zain(2,:) = ki + 10000. * kj |
---|
| 1570 | ! |
---|
| 1571 | CALL MPI_ALLREDUCE( zain,zaout, 1, MPI_2DOUBLE_PRECISION,MPI_MAXLOC,MPI_COMM_OPA,ierror) |
---|
| 1572 | ! |
---|
| 1573 | pmax = zaout(1,1) |
---|
| 1574 | kj = INT( zaout(2,1) / 10000. ) |
---|
| 1575 | ki = INT( zaout(2,1) - 10000.* kj ) |
---|
| 1576 | ! |
---|
| 1577 | END SUBROUTINE mpp_maxloc2d |
---|
[3] | 1578 | |
---|
[13] | 1579 | |
---|
[1344] | 1580 | SUBROUTINE mpp_maxloc3d( ptab, pmask, pmax, ki, kj, kk ) |
---|
| 1581 | !!------------------------------------------------------------------------ |
---|
| 1582 | !! *** routine mpp_maxloc *** |
---|
| 1583 | !! |
---|
| 1584 | !! ** Purpose : Compute the global maximum of an array ptab |
---|
| 1585 | !! and also give its global position |
---|
| 1586 | !! |
---|
| 1587 | !! ** Method : Use MPI_ALLREDUCE with MPI_MINLOC |
---|
| 1588 | !! |
---|
| 1589 | !!-------------------------------------------------------------------------- |
---|
| 1590 | REAL(wp), DIMENSION (jpi,jpj,jpk), INTENT(in ) :: ptab ! Local 2D array |
---|
| 1591 | REAL(wp), DIMENSION (jpi,jpj,jpk), INTENT(in ) :: pmask ! Local mask |
---|
| 1592 | REAL(wp) , INTENT( out) :: pmax ! Global maximum of ptab |
---|
| 1593 | INTEGER , INTENT( out) :: ki, kj, kk ! index of maximum in global frame |
---|
| 1594 | !! |
---|
| 1595 | REAL(wp) :: zmax ! local maximum |
---|
| 1596 | REAL(wp), DIMENSION(2,1) :: zain, zaout |
---|
| 1597 | INTEGER , DIMENSION(3) :: ilocs |
---|
| 1598 | INTEGER :: ierror |
---|
| 1599 | !!----------------------------------------------------------------------- |
---|
| 1600 | ! |
---|
| 1601 | zmax = MAXVAL( ptab(:,:,:) , mask= pmask == 1.e0 ) |
---|
| 1602 | ilocs = MAXLOC( ptab(:,:,:) , mask= pmask == 1.e0 ) |
---|
| 1603 | ! |
---|
| 1604 | ki = ilocs(1) + nimpp - 1 |
---|
| 1605 | kj = ilocs(2) + njmpp - 1 |
---|
| 1606 | kk = ilocs(3) |
---|
| 1607 | ! |
---|
| 1608 | zain(1,:)=zmax |
---|
| 1609 | zain(2,:)=ki+10000.*kj+100000000.*kk |
---|
| 1610 | ! |
---|
| 1611 | CALL MPI_ALLREDUCE( zain,zaout, 1, MPI_2DOUBLE_PRECISION,MPI_MAXLOC,MPI_COMM_OPA,ierror) |
---|
| 1612 | ! |
---|
| 1613 | pmax = zaout(1,1) |
---|
| 1614 | kk = INT( zaout(2,1) / 100000000. ) |
---|
| 1615 | kj = INT( zaout(2,1) - kk * 100000000. ) / 10000 |
---|
| 1616 | ki = INT( zaout(2,1) - kk * 100000000. -kj * 10000. ) |
---|
| 1617 | ! |
---|
| 1618 | END SUBROUTINE mpp_maxloc3d |
---|
[3] | 1619 | |
---|
[869] | 1620 | |
---|
[1344] | 1621 | SUBROUTINE mppsync() |
---|
| 1622 | !!---------------------------------------------------------------------- |
---|
| 1623 | !! *** routine mppsync *** |
---|
| 1624 | !! |
---|
| 1625 | !! ** Purpose : Massively parallel processors, synchroneous |
---|
| 1626 | !! |
---|
| 1627 | !!----------------------------------------------------------------------- |
---|
| 1628 | INTEGER :: ierror |
---|
| 1629 | !!----------------------------------------------------------------------- |
---|
| 1630 | ! |
---|
| 1631 | CALL mpi_barrier( mpi_comm_opa, ierror ) |
---|
| 1632 | ! |
---|
| 1633 | END SUBROUTINE mppsync |
---|
[3] | 1634 | |
---|
| 1635 | |
---|
[1344] | 1636 | SUBROUTINE mppstop |
---|
| 1637 | !!---------------------------------------------------------------------- |
---|
| 1638 | !! *** routine mppstop *** |
---|
| 1639 | !! |
---|
| 1640 | !! ** purpose : Stop massilively parallel processors method |
---|
| 1641 | !! |
---|
| 1642 | !!---------------------------------------------------------------------- |
---|
| 1643 | INTEGER :: info |
---|
| 1644 | !!---------------------------------------------------------------------- |
---|
| 1645 | ! |
---|
| 1646 | CALL mppsync |
---|
| 1647 | CALL mpi_finalize( info ) |
---|
| 1648 | ! |
---|
| 1649 | END SUBROUTINE mppstop |
---|
[3] | 1650 | |
---|
| 1651 | |
---|
[1344] | 1652 | SUBROUTINE mppobc( ptab, kd1, kd2, kl, kk, ktype, kij ) |
---|
| 1653 | !!---------------------------------------------------------------------- |
---|
| 1654 | !! *** routine mppobc *** |
---|
| 1655 | !! |
---|
| 1656 | !! ** Purpose : Message passing manadgement for open boundary |
---|
| 1657 | !! conditions array |
---|
| 1658 | !! |
---|
| 1659 | !! ** Method : Use mppsend and mpprecv function for passing mask |
---|
| 1660 | !! between processors following neighboring subdomains. |
---|
| 1661 | !! domain parameters |
---|
| 1662 | !! nlci : first dimension of the local subdomain |
---|
| 1663 | !! nlcj : second dimension of the local subdomain |
---|
| 1664 | !! nbondi : mark for "east-west local boundary" |
---|
| 1665 | !! nbondj : mark for "north-south local boundary" |
---|
| 1666 | !! noea : number for local neighboring processors |
---|
| 1667 | !! nowe : number for local neighboring processors |
---|
| 1668 | !! noso : number for local neighboring processors |
---|
| 1669 | !! nono : number for local neighboring processors |
---|
| 1670 | !! |
---|
| 1671 | !!---------------------------------------------------------------------- |
---|
| 1672 | INTEGER , INTENT(in ) :: kd1, kd2 ! starting and ending indices |
---|
| 1673 | INTEGER , INTENT(in ) :: kl ! index of open boundary |
---|
| 1674 | INTEGER , INTENT(in ) :: kk ! vertical dimension |
---|
| 1675 | INTEGER , INTENT(in ) :: ktype ! define north/south or east/west cdt |
---|
| 1676 | ! ! = 1 north/south ; = 2 east/west |
---|
| 1677 | INTEGER , INTENT(in ) :: kij ! horizontal dimension |
---|
| 1678 | REAL(wp), INTENT(inout), DIMENSION(kij,kk) :: ptab ! variable array |
---|
| 1679 | !! |
---|
| 1680 | INTEGER :: ji, jj, jk, jl ! dummy loop indices |
---|
| 1681 | INTEGER :: iipt0, iipt1, ilpt1 ! temporary integers |
---|
| 1682 | INTEGER :: ijpt0, ijpt1 ! - - |
---|
| 1683 | INTEGER :: imigr, iihom, ijhom ! - - |
---|
| 1684 | INTEGER :: ml_req1, ml_req2, ml_err ! for key_mpi_isend |
---|
| 1685 | INTEGER :: ml_stat(MPI_STATUS_SIZE) ! for key_mpi_isend |
---|
| 1686 | REAL(wp), DIMENSION(jpi,jpj) :: ztab ! temporary workspace |
---|
| 1687 | !!---------------------------------------------------------------------- |
---|
[3] | 1688 | |
---|
[1344] | 1689 | ! boundary condition initialization |
---|
| 1690 | ! --------------------------------- |
---|
| 1691 | ztab(:,:) = 0.e0 |
---|
| 1692 | ! |
---|
| 1693 | IF( ktype==1 ) THEN ! north/south boundaries |
---|
| 1694 | iipt0 = MAX( 1, MIN(kd1 - nimpp+1, nlci ) ) |
---|
| 1695 | iipt1 = MAX( 0, MIN(kd2 - nimpp+1, nlci - 1 ) ) |
---|
| 1696 | ilpt1 = MAX( 1, MIN(kd2 - nimpp+1, nlci ) ) |
---|
| 1697 | ijpt0 = MAX( 1, MIN(kl - njmpp+1, nlcj ) ) |
---|
| 1698 | ijpt1 = MAX( 0, MIN(kl - njmpp+1, nlcj - 1 ) ) |
---|
| 1699 | ELSEIF( ktype==2 ) THEN ! east/west boundaries |
---|
| 1700 | iipt0 = MAX( 1, MIN(kl - nimpp+1, nlci ) ) |
---|
| 1701 | iipt1 = MAX( 0, MIN(kl - nimpp+1, nlci - 1 ) ) |
---|
| 1702 | ijpt0 = MAX( 1, MIN(kd1 - njmpp+1, nlcj ) ) |
---|
| 1703 | ijpt1 = MAX( 0, MIN(kd2 - njmpp+1, nlcj - 1 ) ) |
---|
| 1704 | ilpt1 = MAX( 1, MIN(kd2 - njmpp+1, nlcj ) ) |
---|
| 1705 | ELSE |
---|
| 1706 | CALL ctl_stop( 'mppobc: bad ktype' ) |
---|
| 1707 | ENDIF |
---|
| 1708 | |
---|
| 1709 | ! Communication level by level |
---|
| 1710 | ! ---------------------------- |
---|
| 1711 | !!gm Remark : this is very time consumming!!! |
---|
| 1712 | ! ! ------------------------ ! |
---|
| 1713 | DO jk = 1, kk ! Loop over the levels ! |
---|
| 1714 | ! ! ------------------------ ! |
---|
| 1715 | ! |
---|
| 1716 | IF( ktype == 1 ) THEN ! north/south boundaries |
---|
| 1717 | DO jj = ijpt0, ijpt1 |
---|
| 1718 | DO ji = iipt0, iipt1 |
---|
| 1719 | ztab(ji,jj) = ptab(ji,jk) |
---|
| 1720 | END DO |
---|
| 1721 | END DO |
---|
| 1722 | ELSEIF( ktype == 2 ) THEN ! east/west boundaries |
---|
| 1723 | DO jj = ijpt0, ijpt1 |
---|
| 1724 | DO ji = iipt0, iipt1 |
---|
| 1725 | ztab(ji,jj) = ptab(jj,jk) |
---|
| 1726 | END DO |
---|
| 1727 | END DO |
---|
| 1728 | ENDIF |
---|
[13] | 1729 | |
---|
[3] | 1730 | |
---|
[1344] | 1731 | ! 1. East and west directions |
---|
| 1732 | ! --------------------------- |
---|
| 1733 | ! |
---|
| 1734 | IF( nbondi /= 2 ) THEN ! Read Dirichlet lateral conditions |
---|
| 1735 | iihom = nlci-nreci |
---|
| 1736 | DO jl = 1, jpreci |
---|
| 1737 | t2ew(:,jl,1) = ztab(jpreci+jl,:) |
---|
| 1738 | t2we(:,jl,1) = ztab(iihom +jl,:) |
---|
| 1739 | END DO |
---|
| 1740 | ENDIF |
---|
| 1741 | ! |
---|
| 1742 | ! ! Migrations |
---|
| 1743 | imigr=jpreci*jpj |
---|
| 1744 | ! |
---|
| 1745 | IF( nbondi == -1 ) THEN |
---|
| 1746 | CALL mppsend( 2, t2we(1,1,1), imigr, noea, ml_req1 ) |
---|
| 1747 | CALL mpprecv( 1, t2ew(1,1,2), imigr ) |
---|
| 1748 | IF(l_isend) CALL mpi_wait( ml_req1, ml_stat, ml_err ) |
---|
| 1749 | ELSEIF( nbondi == 0 ) THEN |
---|
| 1750 | CALL mppsend( 1, t2ew(1,1,1), imigr, nowe, ml_req1 ) |
---|
| 1751 | CALL mppsend( 2, t2we(1,1,1), imigr, noea, ml_req2 ) |
---|
| 1752 | CALL mpprecv( 1, t2ew(1,1,2), imigr ) |
---|
| 1753 | CALL mpprecv( 2, t2we(1,1,2), imigr ) |
---|
| 1754 | IF(l_isend) CALL mpi_wait( ml_req1, ml_stat, ml_err ) |
---|
| 1755 | IF(l_isend) CALL mpi_wait( ml_req2, ml_stat, ml_err ) |
---|
| 1756 | ELSEIF( nbondi == 1 ) THEN |
---|
| 1757 | CALL mppsend( 1, t2ew(1,1,1), imigr, nowe, ml_req1 ) |
---|
| 1758 | CALL mpprecv( 2, t2we(1,1,2), imigr ) |
---|
| 1759 | IF(l_isend) CALL mpi_wait( ml_req1, ml_stat, ml_err ) |
---|
| 1760 | ENDIF |
---|
| 1761 | ! |
---|
| 1762 | ! ! Write Dirichlet lateral conditions |
---|
| 1763 | iihom = nlci-jpreci |
---|
| 1764 | ! |
---|
| 1765 | IF( nbondi == 0 .OR. nbondi == 1 ) THEN |
---|
| 1766 | DO jl = 1, jpreci |
---|
| 1767 | ztab(jl,:) = t2we(:,jl,2) |
---|
| 1768 | END DO |
---|
| 1769 | ENDIF |
---|
| 1770 | IF( nbondi == -1 .OR. nbondi == 0 ) THEN |
---|
| 1771 | DO jl = 1, jpreci |
---|
| 1772 | ztab(iihom+jl,:) = t2ew(:,jl,2) |
---|
| 1773 | END DO |
---|
| 1774 | ENDIF |
---|
[3] | 1775 | |
---|
| 1776 | |
---|
[1344] | 1777 | ! 2. North and south directions |
---|
| 1778 | ! ----------------------------- |
---|
| 1779 | ! |
---|
| 1780 | IF( nbondj /= 2 ) THEN ! Read Dirichlet lateral conditions |
---|
| 1781 | ijhom = nlcj-nrecj |
---|
| 1782 | DO jl = 1, jprecj |
---|
| 1783 | t2sn(:,jl,1) = ztab(:,ijhom +jl) |
---|
| 1784 | t2ns(:,jl,1) = ztab(:,jprecj+jl) |
---|
| 1785 | END DO |
---|
| 1786 | ENDIF |
---|
| 1787 | ! |
---|
| 1788 | ! ! Migrations |
---|
| 1789 | imigr = jprecj * jpi |
---|
| 1790 | ! |
---|
| 1791 | IF( nbondj == -1 ) THEN |
---|
| 1792 | CALL mppsend( 4, t2sn(1,1,1), imigr, nono, ml_req1 ) |
---|
| 1793 | CALL mpprecv( 3, t2ns(1,1,2), imigr ) |
---|
| 1794 | IF(l_isend) CALL mpi_wait( ml_req1, ml_stat, ml_err ) |
---|
| 1795 | ELSEIF( nbondj == 0 ) THEN |
---|
| 1796 | CALL mppsend( 3, t2ns(1,1,1), imigr, noso, ml_req1 ) |
---|
| 1797 | CALL mppsend( 4, t2sn(1,1,1), imigr, nono, ml_req2 ) |
---|
| 1798 | CALL mpprecv( 3, t2ns(1,1,2), imigr ) |
---|
| 1799 | CALL mpprecv( 4, t2sn(1,1,2), imigr ) |
---|
| 1800 | IF( l_isend ) CALL mpi_wait( ml_req1, ml_stat, ml_err ) |
---|
| 1801 | IF( l_isend ) CALL mpi_wait( ml_req2, ml_stat, ml_err ) |
---|
| 1802 | ELSEIF( nbondj == 1 ) THEN |
---|
| 1803 | CALL mppsend( 3, t2ns(1,1,1), imigr, noso, ml_req1 ) |
---|
| 1804 | CALL mpprecv( 4, t2sn(1,1,2), imigr) |
---|
| 1805 | IF( l_isend ) CALL mpi_wait( ml_req1, ml_stat, ml_err ) |
---|
| 1806 | ENDIF |
---|
| 1807 | ! |
---|
| 1808 | ! ! Write Dirichlet lateral conditions |
---|
| 1809 | ijhom = nlcj - jprecj |
---|
| 1810 | IF( nbondj == 0 .OR. nbondj == 1 ) THEN |
---|
| 1811 | DO jl = 1, jprecj |
---|
| 1812 | ztab(:,jl) = t2sn(:,jl,2) |
---|
| 1813 | END DO |
---|
| 1814 | ENDIF |
---|
| 1815 | IF( nbondj == 0 .OR. nbondj == -1 ) THEN |
---|
| 1816 | DO jl = 1, jprecj |
---|
| 1817 | ztab(:,ijhom+jl) = t2ns(:,jl,2) |
---|
| 1818 | END DO |
---|
| 1819 | ENDIF |
---|
| 1820 | IF( ktype==1 .AND. kd1 <= jpi+nimpp-1 .AND. nimpp <= kd2 ) THEN |
---|
| 1821 | DO jj = ijpt0, ijpt1 ! north/south boundaries |
---|
| 1822 | DO ji = iipt0,ilpt1 |
---|
| 1823 | ptab(ji,jk) = ztab(ji,jj) |
---|
| 1824 | END DO |
---|
| 1825 | END DO |
---|
| 1826 | ELSEIF( ktype==2 .AND. kd1 <= jpj+njmpp-1 .AND. njmpp <= kd2 ) THEN |
---|
| 1827 | DO jj = ijpt0, ilpt1 ! east/west boundaries |
---|
| 1828 | DO ji = iipt0,iipt1 |
---|
| 1829 | ptab(jj,jk) = ztab(ji,jj) |
---|
| 1830 | END DO |
---|
| 1831 | END DO |
---|
| 1832 | ENDIF |
---|
| 1833 | ! |
---|
| 1834 | END DO |
---|
| 1835 | ! |
---|
| 1836 | END SUBROUTINE mppobc |
---|
| 1837 | |
---|
[13] | 1838 | |
---|
[1344] | 1839 | SUBROUTINE mpp_comm_free( kcom ) |
---|
| 1840 | !!---------------------------------------------------------------------- |
---|
| 1841 | !!---------------------------------------------------------------------- |
---|
| 1842 | INTEGER, INTENT(in) :: kcom |
---|
| 1843 | !! |
---|
| 1844 | INTEGER :: ierr |
---|
| 1845 | !!---------------------------------------------------------------------- |
---|
| 1846 | ! |
---|
| 1847 | CALL MPI_COMM_FREE(kcom, ierr) |
---|
| 1848 | ! |
---|
| 1849 | END SUBROUTINE mpp_comm_free |
---|
[3] | 1850 | |
---|
[869] | 1851 | |
---|
[1344] | 1852 | SUBROUTINE mpp_ini_ice( pindic ) |
---|
| 1853 | !!---------------------------------------------------------------------- |
---|
| 1854 | !! *** routine mpp_ini_ice *** |
---|
| 1855 | !! |
---|
| 1856 | !! ** Purpose : Initialize special communicator for ice areas |
---|
| 1857 | !! condition together with global variables needed in the ddmpp folding |
---|
| 1858 | !! |
---|
| 1859 | !! ** Method : - Look for ice processors in ice routines |
---|
| 1860 | !! - Put their number in nrank_ice |
---|
| 1861 | !! - Create groups for the world processors and the ice processors |
---|
| 1862 | !! - Create a communicator for ice processors |
---|
| 1863 | !! |
---|
| 1864 | !! ** output |
---|
| 1865 | !! njmppmax = njmpp for northern procs |
---|
| 1866 | !! ndim_rank_ice = number of processors with ice |
---|
| 1867 | !! nrank_ice (ndim_rank_ice) = ice processors |
---|
| 1868 | !! ngrp_world = group ID for the world processors |
---|
| 1869 | !! ngrp_ice = group ID for the ice processors |
---|
| 1870 | !! ncomm_ice = communicator for the ice procs. |
---|
| 1871 | !! n_ice_root = number (in the world) of proc 0 in the ice comm. |
---|
| 1872 | !! |
---|
| 1873 | !!---------------------------------------------------------------------- |
---|
| 1874 | INTEGER, INTENT(in) :: pindic |
---|
| 1875 | !! |
---|
| 1876 | INTEGER :: ierr |
---|
| 1877 | INTEGER :: jjproc |
---|
| 1878 | INTEGER :: ii |
---|
| 1879 | INTEGER, DIMENSION(jpnij) :: kice |
---|
| 1880 | INTEGER, DIMENSION(jpnij) :: zwork |
---|
| 1881 | !!---------------------------------------------------------------------- |
---|
| 1882 | ! |
---|
| 1883 | ! Look for how many procs with sea-ice |
---|
| 1884 | ! |
---|
| 1885 | kice = 0 |
---|
| 1886 | DO jjproc = 1, jpnij |
---|
| 1887 | IF( jjproc == narea .AND. pindic .GT. 0 ) kice(jjproc) = 1 |
---|
| 1888 | END DO |
---|
| 1889 | ! |
---|
| 1890 | zwork = 0 |
---|
| 1891 | CALL MPI_ALLREDUCE( kice, zwork, jpnij, mpi_integer, mpi_sum, mpi_comm_opa, ierr ) |
---|
| 1892 | ndim_rank_ice = SUM( zwork ) |
---|
[3] | 1893 | |
---|
[1344] | 1894 | ! Allocate the right size to nrank_north |
---|
[1208] | 1895 | #if ! defined key_agrif |
---|
[1441] | 1896 | IF( ALLOCATED ( nrank_ice ) ) DEALLOCATE( nrank_ice ) |
---|
[1208] | 1897 | #else |
---|
[1441] | 1898 | IF( ASSOCIATED( nrank_ice ) ) DEALLOCATE( nrank_ice ) |
---|
[1208] | 1899 | #endif |
---|
[1344] | 1900 | ALLOCATE( nrank_ice(ndim_rank_ice) ) |
---|
| 1901 | ! |
---|
| 1902 | ii = 0 |
---|
| 1903 | nrank_ice = 0 |
---|
| 1904 | DO jjproc = 1, jpnij |
---|
| 1905 | IF( zwork(jjproc) == 1) THEN |
---|
| 1906 | ii = ii + 1 |
---|
| 1907 | nrank_ice(ii) = jjproc -1 |
---|
| 1908 | ENDIF |
---|
| 1909 | END DO |
---|
[1208] | 1910 | |
---|
[1344] | 1911 | ! Create the world group |
---|
| 1912 | CALL MPI_COMM_GROUP( mpi_comm_opa, ngrp_world, ierr ) |
---|
[869] | 1913 | |
---|
[1344] | 1914 | ! Create the ice group from the world group |
---|
| 1915 | CALL MPI_GROUP_INCL( ngrp_world, ndim_rank_ice, nrank_ice, ngrp_ice, ierr ) |
---|
[869] | 1916 | |
---|
[1344] | 1917 | ! Create the ice communicator , ie the pool of procs with sea-ice |
---|
| 1918 | CALL MPI_COMM_CREATE( mpi_comm_opa, ngrp_ice, ncomm_ice, ierr ) |
---|
[869] | 1919 | |
---|
[1344] | 1920 | ! Find proc number in the world of proc 0 in the north |
---|
| 1921 | ! The following line seems to be useless, we just comment & keep it as reminder |
---|
| 1922 | ! CALL MPI_GROUP_TRANSLATE_RANKS(ngrp_ice,1,0,ngrp_world,n_ice_root,ierr) |
---|
| 1923 | ! |
---|
| 1924 | END SUBROUTINE mpp_ini_ice |
---|
[869] | 1925 | |
---|
| 1926 | |
---|
[1345] | 1927 | SUBROUTINE mpp_ini_znl |
---|
| 1928 | !!---------------------------------------------------------------------- |
---|
| 1929 | !! *** routine mpp_ini_znl *** |
---|
| 1930 | !! |
---|
| 1931 | !! ** Purpose : Initialize special communicator for computing zonal sum |
---|
| 1932 | !! |
---|
| 1933 | !! ** Method : - Look for processors in the same row |
---|
| 1934 | !! - Put their number in nrank_znl |
---|
| 1935 | !! - Create group for the znl processors |
---|
| 1936 | !! - Create a communicator for znl processors |
---|
| 1937 | !! - Determine if processor should write znl files |
---|
| 1938 | !! |
---|
| 1939 | !! ** output |
---|
| 1940 | !! ndim_rank_znl = number of processors on the same row |
---|
| 1941 | !! ngrp_znl = group ID for the znl processors |
---|
| 1942 | !! ncomm_znl = communicator for the ice procs. |
---|
| 1943 | !! n_znl_root = number (in the world) of proc 0 in the ice comm. |
---|
| 1944 | !! |
---|
| 1945 | !!---------------------------------------------------------------------- |
---|
| 1946 | INTEGER :: ierr |
---|
| 1947 | INTEGER :: jproc |
---|
| 1948 | INTEGER :: ii |
---|
| 1949 | INTEGER, DIMENSION(jpnij) :: kwork |
---|
| 1950 | ! |
---|
| 1951 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - ngrp_world : ', ngrp_world |
---|
| 1952 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - mpi_comm_world : ', mpi_comm_world |
---|
| 1953 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - mpi_comm_opa : ', mpi_comm_opa |
---|
| 1954 | ! |
---|
| 1955 | IF ( jpnj == 1 ) THEN |
---|
| 1956 | ngrp_znl = ngrp_world |
---|
| 1957 | ncomm_znl = mpi_comm_opa |
---|
| 1958 | ELSE |
---|
| 1959 | ! |
---|
| 1960 | CALL MPI_ALLGATHER ( njmpp, 1, mpi_integer, kwork, 1, mpi_integer, mpi_comm_opa, ierr ) |
---|
| 1961 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - kwork pour njmpp : ', kwork |
---|
| 1962 | !-$$ CALL flush(numout) |
---|
| 1963 | ! |
---|
| 1964 | ! Count number of processors on the same row |
---|
| 1965 | ndim_rank_znl = 0 |
---|
| 1966 | DO jproc=1,jpnij |
---|
| 1967 | IF ( kwork(jproc) == njmpp ) THEN |
---|
| 1968 | ndim_rank_znl = ndim_rank_znl + 1 |
---|
| 1969 | ENDIF |
---|
| 1970 | END DO |
---|
| 1971 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - ndim_rank_znl : ', ndim_rank_znl |
---|
| 1972 | !-$$ CALL flush(numout) |
---|
| 1973 | ! Allocate the right size to nrank_znl |
---|
| 1974 | #if ! defined key_agrif |
---|
[1441] | 1975 | IF (ALLOCATED (nrank_znl)) DEALLOCATE(nrank_znl) |
---|
[1345] | 1976 | #else |
---|
[1441] | 1977 | IF (ASSOCIATED(nrank_znl)) DEALLOCATE(nrank_znl) |
---|
[1345] | 1978 | #endif |
---|
| 1979 | ALLOCATE(nrank_znl(ndim_rank_znl)) |
---|
| 1980 | ii = 0 |
---|
| 1981 | nrank_znl (:) = 0 |
---|
| 1982 | DO jproc=1,jpnij |
---|
| 1983 | IF ( kwork(jproc) == njmpp) THEN |
---|
| 1984 | ii = ii + 1 |
---|
| 1985 | nrank_znl(ii) = jproc -1 |
---|
| 1986 | ENDIF |
---|
| 1987 | END DO |
---|
| 1988 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - nrank_znl : ', nrank_znl |
---|
| 1989 | !-$$ CALL flush(numout) |
---|
| 1990 | |
---|
| 1991 | ! Create the opa group |
---|
| 1992 | CALL MPI_COMM_GROUP(mpi_comm_opa,ngrp_opa,ierr) |
---|
| 1993 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - ngrp_opa : ', ngrp_opa |
---|
| 1994 | !-$$ CALL flush(numout) |
---|
| 1995 | |
---|
| 1996 | ! Create the znl group from the opa group |
---|
| 1997 | CALL MPI_GROUP_INCL ( ngrp_opa, ndim_rank_znl, nrank_znl, ngrp_znl, ierr ) |
---|
| 1998 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - ngrp_znl ', ngrp_znl |
---|
| 1999 | !-$$ CALL flush(numout) |
---|
| 2000 | |
---|
| 2001 | ! Create the znl communicator from the opa communicator, ie the pool of procs in the same row |
---|
| 2002 | CALL MPI_COMM_CREATE ( mpi_comm_opa, ngrp_znl, ncomm_znl, ierr ) |
---|
| 2003 | !-$$ WRITE (numout,*) 'mpp_ini_znl ', nproc, ' - ncomm_znl ', ncomm_znl |
---|
| 2004 | !-$$ CALL flush(numout) |
---|
| 2005 | ! |
---|
| 2006 | END IF |
---|
| 2007 | |
---|
| 2008 | ! Determines if processor if the first (starting from i=1) on the row |
---|
| 2009 | IF ( jpni == 1 ) THEN |
---|
| 2010 | l_znl_root = .TRUE. |
---|
| 2011 | ELSE |
---|
| 2012 | l_znl_root = .FALSE. |
---|
| 2013 | kwork (1) = nimpp |
---|
| 2014 | CALL mpp_min ( kwork(1), kcom = ncomm_znl) |
---|
| 2015 | IF ( nimpp == kwork(1)) l_znl_root = .TRUE. |
---|
| 2016 | END IF |
---|
| 2017 | |
---|
| 2018 | END SUBROUTINE mpp_ini_znl |
---|
| 2019 | |
---|
| 2020 | |
---|
[1344] | 2021 | SUBROUTINE mpp_ini_north |
---|
| 2022 | !!---------------------------------------------------------------------- |
---|
| 2023 | !! *** routine mpp_ini_north *** |
---|
| 2024 | !! |
---|
| 2025 | !! ** Purpose : Initialize special communicator for north folding |
---|
| 2026 | !! condition together with global variables needed in the mpp folding |
---|
| 2027 | !! |
---|
| 2028 | !! ** Method : - Look for northern processors |
---|
| 2029 | !! - Put their number in nrank_north |
---|
| 2030 | !! - Create groups for the world processors and the north processors |
---|
| 2031 | !! - Create a communicator for northern processors |
---|
| 2032 | !! |
---|
| 2033 | !! ** output |
---|
| 2034 | !! njmppmax = njmpp for northern procs |
---|
| 2035 | !! ndim_rank_north = number of processors in the northern line |
---|
| 2036 | !! nrank_north (ndim_rank_north) = number of the northern procs. |
---|
| 2037 | !! ngrp_world = group ID for the world processors |
---|
| 2038 | !! ngrp_north = group ID for the northern processors |
---|
| 2039 | !! ncomm_north = communicator for the northern procs. |
---|
| 2040 | !! north_root = number (in the world) of proc 0 in the northern comm. |
---|
| 2041 | !! |
---|
| 2042 | !!---------------------------------------------------------------------- |
---|
| 2043 | INTEGER :: ierr |
---|
| 2044 | INTEGER :: jjproc |
---|
| 2045 | INTEGER :: ii, ji |
---|
| 2046 | !!---------------------------------------------------------------------- |
---|
| 2047 | ! |
---|
| 2048 | njmppmax = MAXVAL( njmppt ) |
---|
| 2049 | ! |
---|
| 2050 | ! Look for how many procs on the northern boundary |
---|
| 2051 | ndim_rank_north = 0 |
---|
| 2052 | DO jjproc = 1, jpnij |
---|
| 2053 | IF( njmppt(jjproc) == njmppmax ) ndim_rank_north = ndim_rank_north + 1 |
---|
| 2054 | END DO |
---|
| 2055 | ! |
---|
| 2056 | ! Allocate the right size to nrank_north |
---|
[1441] | 2057 | #if ! defined key_agrif |
---|
| 2058 | IF (ALLOCATED (nrank_north)) DEALLOCATE(nrank_north) |
---|
| 2059 | #else |
---|
| 2060 | IF (ASSOCIATED(nrank_north)) DEALLOCATE(nrank_north) |
---|
| 2061 | #endif |
---|
[1344] | 2062 | ALLOCATE( nrank_north(ndim_rank_north) ) |
---|
[869] | 2063 | |
---|
[1344] | 2064 | ! Fill the nrank_north array with proc. number of northern procs. |
---|
| 2065 | ! Note : the rank start at 0 in MPI |
---|
| 2066 | ii = 0 |
---|
| 2067 | DO ji = 1, jpnij |
---|
| 2068 | IF ( njmppt(ji) == njmppmax ) THEN |
---|
| 2069 | ii=ii+1 |
---|
| 2070 | nrank_north(ii)=ji-1 |
---|
| 2071 | END IF |
---|
| 2072 | END DO |
---|
| 2073 | ! |
---|
| 2074 | ! create the world group |
---|
| 2075 | CALL MPI_COMM_GROUP( mpi_comm_opa, ngrp_world, ierr ) |
---|
| 2076 | ! |
---|
| 2077 | ! Create the North group from the world group |
---|
| 2078 | CALL MPI_GROUP_INCL( ngrp_world, ndim_rank_north, nrank_north, ngrp_north, ierr ) |
---|
| 2079 | ! |
---|
| 2080 | ! Create the North communicator , ie the pool of procs in the north group |
---|
| 2081 | CALL MPI_COMM_CREATE( mpi_comm_opa, ngrp_north, ncomm_north, ierr ) |
---|
| 2082 | ! |
---|
| 2083 | END SUBROUTINE mpp_ini_north |
---|
[869] | 2084 | |
---|
| 2085 | |
---|
[1344] | 2086 | SUBROUTINE mpp_lbc_north_3d( pt3d, cd_type, psgn ) |
---|
[51] | 2087 | !!--------------------------------------------------------------------- |
---|
| 2088 | !! *** routine mpp_lbc_north_3d *** |
---|
| 2089 | !! |
---|
[1344] | 2090 | !! ** Purpose : Ensure proper north fold horizontal bondary condition |
---|
| 2091 | !! in mpp configuration in case of jpn1 > 1 |
---|
[51] | 2092 | !! |
---|
[1344] | 2093 | !! ** Method : North fold condition and mpp with more than one proc |
---|
| 2094 | !! in i-direction require a specific treatment. We gather |
---|
| 2095 | !! the 4 northern lines of the global domain on 1 processor |
---|
| 2096 | !! and apply lbc north-fold on this sub array. Then we |
---|
| 2097 | !! scatter the north fold array back to the processors. |
---|
[51] | 2098 | !! |
---|
| 2099 | !!---------------------------------------------------------------------- |
---|
[1344] | 2100 | REAL(wp), DIMENSION(jpi,jpj,jpk), INTENT(inout) :: pt3d ! 3D array on which the b.c. is applied |
---|
| 2101 | CHARACTER(len=1) , INTENT(in ) :: cd_type ! nature of pt3d grid-points |
---|
| 2102 | ! ! = T , U , V , F or W gridpoints |
---|
| 2103 | REAL(wp) , INTENT(in ) :: psgn ! = -1. the sign change across the north fold |
---|
| 2104 | !! ! = 1. , the sign is kept |
---|
| 2105 | INTEGER :: ji, jj, jr |
---|
| 2106 | INTEGER :: ierr, itaille, ildi, ilei, iilb |
---|
| 2107 | INTEGER :: ijpj, ijpjm1, ij, iproc |
---|
| 2108 | REAL(wp), DIMENSION(jpiglo,4,jpk) :: ztab |
---|
| 2109 | REAL(wp), DIMENSION(jpi ,4,jpk) :: znorthloc |
---|
| 2110 | REAL(wp), DIMENSION(jpi ,4,jpk,jpni) :: znorthgloio |
---|
[51] | 2111 | !!---------------------------------------------------------------------- |
---|
[1344] | 2112 | ! |
---|
| 2113 | ijpj = 4 |
---|
| 2114 | ijpjm1 = 3 |
---|
[2480] | 2115 | ztab(:,:,:) = 0.e0 |
---|
[1344] | 2116 | ! |
---|
| 2117 | DO jj = nlcj - ijpj +1, nlcj ! put in znorthloc the last 4 jlines of pt3d |
---|
| 2118 | ij = jj - nlcj + ijpj |
---|
| 2119 | znorthloc(:,ij,:) = pt3d(:,jj,:) |
---|
| 2120 | END DO |
---|
| 2121 | ! |
---|
| 2122 | ! ! Build in procs of ncomm_north the znorthgloio |
---|
| 2123 | itaille = jpi * jpk * ijpj |
---|
| 2124 | CALL MPI_ALLGATHER( znorthloc , itaille, MPI_DOUBLE_PRECISION, & |
---|
| 2125 | & znorthgloio, itaille, MPI_DOUBLE_PRECISION, ncomm_north, ierr ) |
---|
| 2126 | ! |
---|
| 2127 | ! ! recover the global north array |
---|
| 2128 | DO jr = 1, ndim_rank_north |
---|
| 2129 | iproc = nrank_north(jr) + 1 |
---|
| 2130 | ildi = nldit (iproc) |
---|
| 2131 | ilei = nleit (iproc) |
---|
| 2132 | iilb = nimppt(iproc) |
---|
| 2133 | DO jj = 1, 4 |
---|
| 2134 | DO ji = ildi, ilei |
---|
| 2135 | ztab(ji+iilb-1,jj,:) = znorthgloio(ji,jj,:,jr) |
---|
| 2136 | END DO |
---|
| 2137 | END DO |
---|
| 2138 | END DO |
---|
| 2139 | ! |
---|
| 2140 | CALL lbc_nfd( ztab, cd_type, psgn ) ! North fold boundary condition |
---|
| 2141 | ! |
---|
| 2142 | DO jj = nlcj-ijpj+1, nlcj ! Scatter back to pt3d |
---|
| 2143 | ij = jj - nlcj + ijpj |
---|
| 2144 | DO ji= 1, nlci |
---|
| 2145 | pt3d(ji,jj,:) = ztab(ji+nimpp-1,ij,:) |
---|
| 2146 | END DO |
---|
| 2147 | END DO |
---|
| 2148 | ! |
---|
| 2149 | END SUBROUTINE mpp_lbc_north_3d |
---|
[3] | 2150 | |
---|
| 2151 | |
---|
[1344] | 2152 | SUBROUTINE mpp_lbc_north_2d( pt2d, cd_type, psgn) |
---|
| 2153 | !!--------------------------------------------------------------------- |
---|
| 2154 | !! *** routine mpp_lbc_north_2d *** |
---|
| 2155 | !! |
---|
| 2156 | !! ** Purpose : Ensure proper north fold horizontal bondary condition |
---|
| 2157 | !! in mpp configuration in case of jpn1 > 1 (for 2d array ) |
---|
| 2158 | !! |
---|
| 2159 | !! ** Method : North fold condition and mpp with more than one proc |
---|
| 2160 | !! in i-direction require a specific treatment. We gather |
---|
| 2161 | !! the 4 northern lines of the global domain on 1 processor |
---|
| 2162 | !! and apply lbc north-fold on this sub array. Then we |
---|
| 2163 | !! scatter the north fold array back to the processors. |
---|
| 2164 | !! |
---|
| 2165 | !!---------------------------------------------------------------------- |
---|
| 2166 | REAL(wp), DIMENSION(jpi,jpj), INTENT(inout) :: pt2d ! 3D array on which the b.c. is applied |
---|
| 2167 | CHARACTER(len=1) , INTENT(in ) :: cd_type ! nature of pt3d grid-points |
---|
| 2168 | ! ! = T , U , V , F or W gridpoints |
---|
| 2169 | REAL(wp) , INTENT(in ) :: psgn ! = -1. the sign change across the north fold |
---|
| 2170 | !! ! = 1. , the sign is kept |
---|
| 2171 | INTEGER :: ji, jj, jr |
---|
| 2172 | INTEGER :: ierr, itaille, ildi, ilei, iilb |
---|
| 2173 | INTEGER :: ijpj, ijpjm1, ij, iproc |
---|
| 2174 | REAL(wp), DIMENSION(jpiglo,4) :: ztab |
---|
| 2175 | REAL(wp), DIMENSION(jpi ,4) :: znorthloc |
---|
| 2176 | REAL(wp), DIMENSION(jpi ,4,jpni) :: znorthgloio |
---|
| 2177 | !!---------------------------------------------------------------------- |
---|
| 2178 | ! |
---|
| 2179 | ijpj = 4 |
---|
| 2180 | ijpjm1 = 3 |
---|
[2480] | 2181 | ztab(:,:) = 0.e0 |
---|
[1344] | 2182 | ! |
---|
| 2183 | DO jj = nlcj-ijpj+1, nlcj ! put in znorthloc the last 4 jlines of pt2d |
---|
| 2184 | ij = jj - nlcj + ijpj |
---|
| 2185 | znorthloc(:,ij) = pt2d(:,jj) |
---|
| 2186 | END DO |
---|
[3] | 2187 | |
---|
[1344] | 2188 | ! ! Build in procs of ncomm_north the znorthgloio |
---|
| 2189 | itaille = jpi * ijpj |
---|
| 2190 | CALL MPI_ALLGATHER( znorthloc , itaille, MPI_DOUBLE_PRECISION, & |
---|
| 2191 | & znorthgloio, itaille, MPI_DOUBLE_PRECISION, ncomm_north, ierr ) |
---|
| 2192 | ! |
---|
| 2193 | DO jr = 1, ndim_rank_north ! recover the global north array |
---|
| 2194 | iproc = nrank_north(jr) + 1 |
---|
| 2195 | ildi=nldit (iproc) |
---|
| 2196 | ilei=nleit (iproc) |
---|
| 2197 | iilb=nimppt(iproc) |
---|
| 2198 | DO jj = 1, 4 |
---|
| 2199 | DO ji = ildi, ilei |
---|
| 2200 | ztab(ji+iilb-1,jj) = znorthgloio(ji,jj,jr) |
---|
[13] | 2201 | END DO |
---|
| 2202 | END DO |
---|
[1344] | 2203 | END DO |
---|
| 2204 | ! |
---|
| 2205 | CALL lbc_nfd( ztab, cd_type, psgn ) ! North fold boundary condition |
---|
| 2206 | ! |
---|
| 2207 | ! |
---|
| 2208 | DO jj = nlcj-ijpj+1, nlcj ! Scatter back to pt2d |
---|
[13] | 2209 | ij = jj - nlcj + ijpj |
---|
[1344] | 2210 | DO ji = 1, nlci |
---|
| 2211 | pt2d(ji,jj) = ztab(ji+nimpp-1,ij) |
---|
| 2212 | END DO |
---|
[13] | 2213 | END DO |
---|
[1344] | 2214 | ! |
---|
[13] | 2215 | END SUBROUTINE mpp_lbc_north_2d |
---|
[3] | 2216 | |
---|
| 2217 | |
---|
[1344] | 2218 | SUBROUTINE mpp_lbc_north_e( pt2d, cd_type, psgn) |
---|
| 2219 | !!--------------------------------------------------------------------- |
---|
| 2220 | !! *** routine mpp_lbc_north_2d *** |
---|
| 2221 | !! |
---|
| 2222 | !! ** Purpose : Ensure proper north fold horizontal bondary condition |
---|
| 2223 | !! in mpp configuration in case of jpn1 > 1 and for 2d |
---|
| 2224 | !! array with outer extra halo |
---|
| 2225 | !! |
---|
| 2226 | !! ** Method : North fold condition and mpp with more than one proc |
---|
| 2227 | !! in i-direction require a specific treatment. We gather |
---|
| 2228 | !! the 4+2*jpr2dj northern lines of the global domain on 1 |
---|
| 2229 | !! processor and apply lbc north-fold on this sub array. |
---|
| 2230 | !! Then we scatter the north fold array back to the processors. |
---|
| 2231 | !! |
---|
| 2232 | !!---------------------------------------------------------------------- |
---|
| 2233 | REAL(wp), DIMENSION(1-jpr2di:jpi+jpr2di,1-jpr2dj:jpj+jpr2dj), INTENT(inout) :: pt2d ! 2D array with extra halo |
---|
| 2234 | CHARACTER(len=1) , INTENT(in ) :: cd_type ! nature of pt3d grid-points |
---|
| 2235 | ! ! = T , U , V , F or W -points |
---|
| 2236 | REAL(wp) , INTENT(in ) :: psgn ! = -1. the sign change across the |
---|
| 2237 | !! ! north fold, = 1. otherwise |
---|
| 2238 | INTEGER :: ji, jj, jr |
---|
| 2239 | INTEGER :: ierr, itaille, ildi, ilei, iilb |
---|
| 2240 | INTEGER :: ijpj, ij, iproc |
---|
| 2241 | REAL(wp), DIMENSION(jpiglo,4+2*jpr2dj) :: ztab |
---|
| 2242 | REAL(wp), DIMENSION(jpi ,4+2*jpr2dj) :: znorthloc |
---|
| 2243 | REAL(wp), DIMENSION(jpi ,4+2*jpr2dj,jpni) :: znorthgloio |
---|
| 2244 | !!---------------------------------------------------------------------- |
---|
| 2245 | ! |
---|
| 2246 | ijpj=4 |
---|
[2480] | 2247 | ztab(:,:) = 0.e0 |
---|
[311] | 2248 | |
---|
[1344] | 2249 | ij=0 |
---|
| 2250 | ! put in znorthloc the last 4 jlines of pt2d |
---|
| 2251 | DO jj = nlcj - ijpj + 1 - jpr2dj, nlcj +jpr2dj |
---|
| 2252 | ij = ij + 1 |
---|
| 2253 | DO ji = 1, jpi |
---|
| 2254 | znorthloc(ji,ij)=pt2d(ji,jj) |
---|
| 2255 | END DO |
---|
| 2256 | END DO |
---|
| 2257 | ! |
---|
| 2258 | itaille = jpi * ( ijpj + 2 * jpr2dj ) |
---|
| 2259 | CALL MPI_ALLGATHER( znorthloc(1,1) , itaille, MPI_DOUBLE_PRECISION, & |
---|
| 2260 | & znorthgloio(1,1,1), itaille, MPI_DOUBLE_PRECISION, ncomm_north, ierr ) |
---|
| 2261 | ! |
---|
| 2262 | DO jr = 1, ndim_rank_north ! recover the global north array |
---|
| 2263 | iproc = nrank_north(jr) + 1 |
---|
| 2264 | ildi = nldit (iproc) |
---|
| 2265 | ilei = nleit (iproc) |
---|
| 2266 | iilb = nimppt(iproc) |
---|
| 2267 | DO jj = 1, ijpj+2*jpr2dj |
---|
| 2268 | DO ji = ildi, ilei |
---|
| 2269 | ztab(ji+iilb-1,jj) = znorthgloio(ji,jj,jr) |
---|
[311] | 2270 | END DO |
---|
[1344] | 2271 | END DO |
---|
| 2272 | END DO |
---|
[311] | 2273 | |
---|
| 2274 | |
---|
[1344] | 2275 | ! 2. North-Fold boundary conditions |
---|
| 2276 | ! ---------------------------------- |
---|
| 2277 | CALL lbc_nfd( ztab(:,:), cd_type, psgn, pr2dj = jpr2dj ) |
---|
[311] | 2278 | |
---|
[1344] | 2279 | ij = jpr2dj |
---|
| 2280 | !! Scatter back to pt2d |
---|
| 2281 | DO jj = nlcj - ijpj + 1 , nlcj +jpr2dj |
---|
| 2282 | ij = ij +1 |
---|
| 2283 | DO ji= 1, nlci |
---|
| 2284 | pt2d(ji,jj) = ztab(ji+nimpp-1,ij) |
---|
[311] | 2285 | END DO |
---|
| 2286 | END DO |
---|
[1344] | 2287 | ! |
---|
[311] | 2288 | END SUBROUTINE mpp_lbc_north_e |
---|
| 2289 | |
---|
[389] | 2290 | |
---|
[2481] | 2291 | SUBROUTINE mpi_init_opa( ldtxt, ksft, code ) |
---|
[1344] | 2292 | !!--------------------------------------------------------------------- |
---|
| 2293 | !! *** routine mpp_init.opa *** |
---|
| 2294 | !! |
---|
| 2295 | !! ** Purpose :: export and attach a MPI buffer for bsend |
---|
| 2296 | !! |
---|
| 2297 | !! ** Method :: define buffer size in namelist, if 0 no buffer attachment |
---|
| 2298 | !! but classical mpi_init |
---|
| 2299 | !! |
---|
| 2300 | !! History :: 01/11 :: IDRIS initial version for IBM only |
---|
| 2301 | !! 08/04 :: R. Benshila, generalisation |
---|
| 2302 | !!--------------------------------------------------------------------- |
---|
[2481] | 2303 | CHARACTER(len=*),DIMENSION(:), INTENT( out) :: ldtxt |
---|
| 2304 | INTEGER , INTENT(inout) :: ksft |
---|
| 2305 | INTEGER , INTENT( out) :: code |
---|
| 2306 | INTEGER :: ierr, ji |
---|
| 2307 | LOGICAL :: mpi_was_called |
---|
[1344] | 2308 | !!--------------------------------------------------------------------- |
---|
| 2309 | ! |
---|
| 2310 | CALL mpi_initialized( mpi_was_called, code ) ! MPI initialization |
---|
[532] | 2311 | IF ( code /= MPI_SUCCESS ) THEN |
---|
[2481] | 2312 | DO ji = 1, SIZE(ldtxt) |
---|
| 2313 | IF( TRIM(ldtxt(ji)) /= '' ) WRITE(*,*) ldtxt(ji) ! control print of mynode |
---|
| 2314 | END DO |
---|
| 2315 | WRITE(*, cform_err) |
---|
| 2316 | WRITE(*, *) ' lib_mpp: Error in routine mpi_initialized' |
---|
[1344] | 2317 | CALL mpi_abort( mpi_comm_world, code, ierr ) |
---|
[532] | 2318 | ENDIF |
---|
[1344] | 2319 | ! |
---|
| 2320 | IF( .NOT. mpi_was_called ) THEN |
---|
| 2321 | CALL mpi_init( code ) |
---|
| 2322 | CALL mpi_comm_dup( mpi_comm_world, mpi_comm_opa, code ) |
---|
[532] | 2323 | IF ( code /= MPI_SUCCESS ) THEN |
---|
[2481] | 2324 | DO ji = 1, SIZE(ldtxt) |
---|
| 2325 | IF( TRIM(ldtxt(ji)) /= '' ) WRITE(*,*) ldtxt(ji) ! control print of mynode |
---|
| 2326 | END DO |
---|
| 2327 | WRITE(*, cform_err) |
---|
| 2328 | WRITE(*, *) ' lib_mpp: Error in routine mpi_comm_dup' |
---|
[532] | 2329 | CALL mpi_abort( mpi_comm_world, code, ierr ) |
---|
| 2330 | ENDIF |
---|
| 2331 | ENDIF |
---|
[1344] | 2332 | ! |
---|
[897] | 2333 | IF( nn_buffer > 0 ) THEN |
---|
[2481] | 2334 | WRITE(ldtxt(ksft),*) 'mpi_bsend, buffer allocation of : ', nn_buffer ; ksft = ksft + 1 |
---|
[897] | 2335 | ! Buffer allocation and attachment |
---|
[2481] | 2336 | ALLOCATE( tampon(nn_buffer), stat = ierr ) |
---|
| 2337 | IF (ierr /= 0) THEN |
---|
| 2338 | DO ji = 1, SIZE(ldtxt) |
---|
| 2339 | IF( TRIM(ldtxt(ji)) /= '' ) WRITE(*,*) ldtxt(ji) ! control print of mynode |
---|
| 2340 | END DO |
---|
| 2341 | WRITE(*, cform_err) |
---|
| 2342 | WRITE(*, *) ' lib_mpp: Error in ALLOCATE', ierr |
---|
| 2343 | CALL mpi_abort( mpi_comm_world, code, ierr ) |
---|
| 2344 | END IF |
---|
| 2345 | CALL mpi_buffer_attach( tampon, nn_buffer, code ) |
---|
[897] | 2346 | ENDIF |
---|
[1344] | 2347 | ! |
---|
[13] | 2348 | END SUBROUTINE mpi_init_opa |
---|
[3] | 2349 | |
---|
[2304] | 2350 | #if defined key_mpp_rep |
---|
[1976] | 2351 | SUBROUTINE DDPDD_MPI (ydda, yddb, ilen, itype) |
---|
| 2352 | !!--------------------------------------------------------------------- |
---|
| 2353 | !! Routine DDPDD_MPI: used by reduction operator MPI_SUMDD |
---|
| 2354 | !! |
---|
| 2355 | !! Modification of original codes written by David H. Bailey |
---|
| 2356 | !! This subroutine computes yddb(i) = ydda(i)+yddb(i) |
---|
| 2357 | !!--------------------------------------------------------------------- |
---|
| 2358 | INTEGER, INTENT(in) :: ilen, itype |
---|
| 2359 | COMPLEX(wp), DIMENSION(ilen), INTENT(in) :: ydda |
---|
| 2360 | COMPLEX(wp), DIMENSION(ilen), INTENT(inout) :: yddb |
---|
| 2361 | ! |
---|
| 2362 | REAL(wp) :: zerr, zt1, zt2 ! local work variables |
---|
| 2363 | INTEGER :: ji, ztmp ! local scalar |
---|
| 2364 | |
---|
| 2365 | ztmp = itype ! avoid compilation warning |
---|
| 2366 | |
---|
| 2367 | DO ji=1,ilen |
---|
| 2368 | ! Compute ydda + yddb using Knuth's trick. |
---|
| 2369 | zt1 = real(ydda(ji)) + real(yddb(ji)) |
---|
| 2370 | zerr = zt1 - real(ydda(ji)) |
---|
| 2371 | zt2 = ((real(yddb(ji)) - zerr) + (real(ydda(ji)) - (zt1 - zerr))) & |
---|
| 2372 | + aimag(ydda(ji)) + aimag(yddb(ji)) |
---|
| 2373 | |
---|
| 2374 | ! The result is zt1 + zt2, after normalization. |
---|
| 2375 | yddb(ji) = cmplx ( zt1 + zt2, zt2 - ((zt1 + zt2) - zt1),wp ) |
---|
| 2376 | END DO |
---|
| 2377 | |
---|
| 2378 | END SUBROUTINE DDPDD_MPI |
---|
| 2379 | #endif |
---|
| 2380 | |
---|
[13] | 2381 | #else |
---|
| 2382 | !!---------------------------------------------------------------------- |
---|
| 2383 | !! Default case: Dummy module share memory computing |
---|
| 2384 | !!---------------------------------------------------------------------- |
---|
[1976] | 2385 | |
---|
[13] | 2386 | INTERFACE mpp_sum |
---|
[2480] | 2387 | MODULE PROCEDURE mpp_sum_a2s, mpp_sum_as, mpp_sum_ai, mpp_sum_s, mpp_sum_i |
---|
[13] | 2388 | END INTERFACE |
---|
| 2389 | INTERFACE mpp_max |
---|
[681] | 2390 | MODULE PROCEDURE mppmax_a_int, mppmax_int, mppmax_a_real, mppmax_real |
---|
[13] | 2391 | END INTERFACE |
---|
| 2392 | INTERFACE mpp_min |
---|
| 2393 | MODULE PROCEDURE mppmin_a_int, mppmin_int, mppmin_a_real, mppmin_real |
---|
| 2394 | END INTERFACE |
---|
| 2395 | INTERFACE mppobc |
---|
| 2396 | MODULE PROCEDURE mppobc_1d, mppobc_2d, mppobc_3d, mppobc_4d |
---|
| 2397 | END INTERFACE |
---|
[1344] | 2398 | INTERFACE mpp_minloc |
---|
| 2399 | MODULE PROCEDURE mpp_minloc2d ,mpp_minloc3d |
---|
| 2400 | END INTERFACE |
---|
| 2401 | INTERFACE mpp_maxloc |
---|
| 2402 | MODULE PROCEDURE mpp_maxloc2d ,mpp_maxloc3d |
---|
| 2403 | END INTERFACE |
---|
[3] | 2404 | |
---|
[2480] | 2405 | |
---|
[13] | 2406 | LOGICAL, PUBLIC, PARAMETER :: lk_mpp = .FALSE. !: mpp flag |
---|
[869] | 2407 | INTEGER :: ncomm_ice |
---|
[3] | 2408 | |
---|
[13] | 2409 | CONTAINS |
---|
[3] | 2410 | |
---|
[1579] | 2411 | FUNCTION mynode( ldtxt, localComm ) RESULT (function_value) |
---|
| 2412 | CHARACTER(len=*),DIMENSION(:), INTENT( out) :: ldtxt |
---|
| 2413 | INTEGER, OPTIONAL , INTENT(in ) :: localComm |
---|
[1559] | 2414 | IF( PRESENT( localComm ) .OR. .NOT.PRESENT( localComm ) ) function_value = 0 |
---|
[1579] | 2415 | IF( .FALSE. ) ldtxt(:) = 'never done' |
---|
[13] | 2416 | END FUNCTION mynode |
---|
[3] | 2417 | |
---|
[13] | 2418 | SUBROUTINE mppsync ! Dummy routine |
---|
| 2419 | END SUBROUTINE mppsync |
---|
[3] | 2420 | |
---|
[869] | 2421 | SUBROUTINE mpp_sum_as( parr, kdim, kcom ) ! Dummy routine |
---|
[13] | 2422 | REAL , DIMENSION(:) :: parr |
---|
| 2423 | INTEGER :: kdim |
---|
[869] | 2424 | INTEGER, OPTIONAL :: kcom |
---|
| 2425 | WRITE(*,*) 'mpp_sum_as: You should not have seen this print! error?', kdim, parr(1), kcom |
---|
[13] | 2426 | END SUBROUTINE mpp_sum_as |
---|
[3] | 2427 | |
---|
[869] | 2428 | SUBROUTINE mpp_sum_a2s( parr, kdim, kcom ) ! Dummy routine |
---|
[13] | 2429 | REAL , DIMENSION(:,:) :: parr |
---|
| 2430 | INTEGER :: kdim |
---|
[869] | 2431 | INTEGER, OPTIONAL :: kcom |
---|
| 2432 | WRITE(*,*) 'mpp_sum_a2s: You should not have seen this print! error?', kdim, parr(1,1), kcom |
---|
[13] | 2433 | END SUBROUTINE mpp_sum_a2s |
---|
[3] | 2434 | |
---|
[869] | 2435 | SUBROUTINE mpp_sum_ai( karr, kdim, kcom ) ! Dummy routine |
---|
[13] | 2436 | INTEGER, DIMENSION(:) :: karr |
---|
| 2437 | INTEGER :: kdim |
---|
[869] | 2438 | INTEGER, OPTIONAL :: kcom |
---|
| 2439 | WRITE(*,*) 'mpp_sum_ai: You should not have seen this print! error?', kdim, karr(1), kcom |
---|
[13] | 2440 | END SUBROUTINE mpp_sum_ai |
---|
[3] | 2441 | |
---|
[869] | 2442 | SUBROUTINE mpp_sum_s( psca, kcom ) ! Dummy routine |
---|
[13] | 2443 | REAL :: psca |
---|
[869] | 2444 | INTEGER, OPTIONAL :: kcom |
---|
| 2445 | WRITE(*,*) 'mpp_sum_s: You should not have seen this print! error?', psca, kcom |
---|
[13] | 2446 | END SUBROUTINE mpp_sum_s |
---|
[2480] | 2447 | |
---|
[869] | 2448 | SUBROUTINE mpp_sum_i( kint, kcom ) ! Dummy routine |
---|
[13] | 2449 | integer :: kint |
---|
[2480] | 2450 | INTEGER, OPTIONAL :: kcom |
---|
[869] | 2451 | WRITE(*,*) 'mpp_sum_i: You should not have seen this print! error?', kint, kcom |
---|
[13] | 2452 | END SUBROUTINE mpp_sum_i |
---|
| 2453 | |
---|
[869] | 2454 | SUBROUTINE mppmax_a_real( parr, kdim, kcom ) |
---|
[13] | 2455 | REAL , DIMENSION(:) :: parr |
---|
| 2456 | INTEGER :: kdim |
---|
[869] | 2457 | INTEGER, OPTIONAL :: kcom |
---|
| 2458 | WRITE(*,*) 'mppmax_a_real: You should not have seen this print! error?', kdim, parr(1), kcom |
---|
[13] | 2459 | END SUBROUTINE mppmax_a_real |
---|
| 2460 | |
---|
[869] | 2461 | SUBROUTINE mppmax_real( psca, kcom ) |
---|
[13] | 2462 | REAL :: psca |
---|
[869] | 2463 | INTEGER, OPTIONAL :: kcom |
---|
| 2464 | WRITE(*,*) 'mppmax_real: You should not have seen this print! error?', psca, kcom |
---|
[13] | 2465 | END SUBROUTINE mppmax_real |
---|
| 2466 | |
---|
[869] | 2467 | SUBROUTINE mppmin_a_real( parr, kdim, kcom ) |
---|
[13] | 2468 | REAL , DIMENSION(:) :: parr |
---|
| 2469 | INTEGER :: kdim |
---|
[869] | 2470 | INTEGER, OPTIONAL :: kcom |
---|
| 2471 | WRITE(*,*) 'mppmin_a_real: You should not have seen this print! error?', kdim, parr(1), kcom |
---|
[13] | 2472 | END SUBROUTINE mppmin_a_real |
---|
| 2473 | |
---|
[869] | 2474 | SUBROUTINE mppmin_real( psca, kcom ) |
---|
[13] | 2475 | REAL :: psca |
---|
[869] | 2476 | INTEGER, OPTIONAL :: kcom |
---|
| 2477 | WRITE(*,*) 'mppmin_real: You should not have seen this print! error?', psca, kcom |
---|
[13] | 2478 | END SUBROUTINE mppmin_real |
---|
| 2479 | |
---|
[869] | 2480 | SUBROUTINE mppmax_a_int( karr, kdim ,kcom) |
---|
[681] | 2481 | INTEGER, DIMENSION(:) :: karr |
---|
| 2482 | INTEGER :: kdim |
---|
[869] | 2483 | INTEGER, OPTIONAL :: kcom |
---|
[888] | 2484 | WRITE(*,*) 'mppmax_a_int: You should not have seen this print! error?', kdim, karr(1), kcom |
---|
[681] | 2485 | END SUBROUTINE mppmax_a_int |
---|
| 2486 | |
---|
[869] | 2487 | SUBROUTINE mppmax_int( kint, kcom) |
---|
[681] | 2488 | INTEGER :: kint |
---|
[869] | 2489 | INTEGER, OPTIONAL :: kcom |
---|
| 2490 | WRITE(*,*) 'mppmax_int: You should not have seen this print! error?', kint, kcom |
---|
[681] | 2491 | END SUBROUTINE mppmax_int |
---|
| 2492 | |
---|
[869] | 2493 | SUBROUTINE mppmin_a_int( karr, kdim, kcom ) |
---|
[13] | 2494 | INTEGER, DIMENSION(:) :: karr |
---|
| 2495 | INTEGER :: kdim |
---|
[869] | 2496 | INTEGER, OPTIONAL :: kcom |
---|
| 2497 | WRITE(*,*) 'mppmin_a_int: You should not have seen this print! error?', kdim, karr(1), kcom |
---|
[13] | 2498 | END SUBROUTINE mppmin_a_int |
---|
| 2499 | |
---|
[869] | 2500 | SUBROUTINE mppmin_int( kint, kcom ) |
---|
[13] | 2501 | INTEGER :: kint |
---|
[869] | 2502 | INTEGER, OPTIONAL :: kcom |
---|
| 2503 | WRITE(*,*) 'mppmin_int: You should not have seen this print! error?', kint, kcom |
---|
[13] | 2504 | END SUBROUTINE mppmin_int |
---|
| 2505 | |
---|
| 2506 | SUBROUTINE mppobc_1d( parr, kd1, kd2, kl, kk, ktype, kij ) |
---|
[1344] | 2507 | INTEGER :: kd1, kd2, kl , kk, ktype, kij |
---|
| 2508 | REAL, DIMENSION(:) :: parr ! variable array |
---|
| 2509 | WRITE(*,*) 'mppobc: You should not have seen this print! error?', parr(1), kd1, kd2, kl, kk, ktype, kij |
---|
[13] | 2510 | END SUBROUTINE mppobc_1d |
---|
| 2511 | |
---|
| 2512 | SUBROUTINE mppobc_2d( parr, kd1, kd2, kl, kk, ktype, kij ) |
---|
[1344] | 2513 | INTEGER :: kd1, kd2, kl , kk, ktype, kij |
---|
| 2514 | REAL, DIMENSION(:,:) :: parr ! variable array |
---|
| 2515 | WRITE(*,*) 'mppobc: You should not have seen this print! error?', parr(1,1), kd1, kd2, kl, kk, ktype, kij |
---|
[13] | 2516 | END SUBROUTINE mppobc_2d |
---|
| 2517 | |
---|
| 2518 | SUBROUTINE mppobc_3d( parr, kd1, kd2, kl, kk, ktype, kij ) |
---|
[1344] | 2519 | INTEGER :: kd1, kd2, kl , kk, ktype, kij |
---|
| 2520 | REAL, DIMENSION(:,:,:) :: parr ! variable array |
---|
| 2521 | WRITE(*,*) 'mppobc: You should not have seen this print! error?', parr(1,1,1), kd1, kd2, kl, kk, ktype, kij |
---|
[13] | 2522 | END SUBROUTINE mppobc_3d |
---|
| 2523 | |
---|
| 2524 | SUBROUTINE mppobc_4d( parr, kd1, kd2, kl, kk, ktype, kij ) |
---|
[1344] | 2525 | INTEGER :: kd1, kd2, kl , kk, ktype, kij |
---|
| 2526 | REAL, DIMENSION(:,:,:,:) :: parr ! variable array |
---|
| 2527 | WRITE(*,*) 'mppobc: You should not have seen this print! error?', parr(1,1,1,1), kd1, kd2, kl, kk, ktype, kij |
---|
[13] | 2528 | END SUBROUTINE mppobc_4d |
---|
| 2529 | |
---|
[1344] | 2530 | SUBROUTINE mpp_minloc2d( ptab, pmask, pmin, ki, kj ) |
---|
[181] | 2531 | REAL :: pmin |
---|
| 2532 | REAL , DIMENSION (:,:) :: ptab, pmask |
---|
| 2533 | INTEGER :: ki, kj |
---|
[1528] | 2534 | WRITE(*,*) 'mpp_minloc2d: You should not have seen this print! error?', pmin, ki, kj, ptab(1,1), pmask(1,1) |
---|
[181] | 2535 | END SUBROUTINE mpp_minloc2d |
---|
| 2536 | |
---|
[1344] | 2537 | SUBROUTINE mpp_minloc3d( ptab, pmask, pmin, ki, kj, kk ) |
---|
[181] | 2538 | REAL :: pmin |
---|
| 2539 | REAL , DIMENSION (:,:,:) :: ptab, pmask |
---|
| 2540 | INTEGER :: ki, kj, kk |
---|
[1528] | 2541 | WRITE(*,*) 'mpp_minloc3d: You should not have seen this print! error?', pmin, ki, kj, kk, ptab(1,1,1), pmask(1,1,1) |
---|
[181] | 2542 | END SUBROUTINE mpp_minloc3d |
---|
| 2543 | |
---|
[1344] | 2544 | SUBROUTINE mpp_maxloc2d( ptab, pmask, pmax, ki, kj ) |
---|
[181] | 2545 | REAL :: pmax |
---|
| 2546 | REAL , DIMENSION (:,:) :: ptab, pmask |
---|
| 2547 | INTEGER :: ki, kj |
---|
[1528] | 2548 | WRITE(*,*) 'mpp_maxloc2d: You should not have seen this print! error?', pmax, ki, kj, ptab(1,1), pmask(1,1) |
---|
[181] | 2549 | END SUBROUTINE mpp_maxloc2d |
---|
| 2550 | |
---|
[1344] | 2551 | SUBROUTINE mpp_maxloc3d( ptab, pmask, pmax, ki, kj, kk ) |
---|
[181] | 2552 | REAL :: pmax |
---|
| 2553 | REAL , DIMENSION (:,:,:) :: ptab, pmask |
---|
| 2554 | INTEGER :: ki, kj, kk |
---|
[1528] | 2555 | WRITE(*,*) 'mpp_maxloc3d: You should not have seen this print! error?', pmax, ki, kj, kk, ptab(1,1,1), pmask(1,1,1) |
---|
[181] | 2556 | END SUBROUTINE mpp_maxloc3d |
---|
| 2557 | |
---|
[51] | 2558 | SUBROUTINE mppstop |
---|
| 2559 | WRITE(*,*) 'mppstop: You should not have seen this print! error?' |
---|
| 2560 | END SUBROUTINE mppstop |
---|
| 2561 | |
---|
[1344] | 2562 | SUBROUTINE mpp_ini_ice( kcom ) |
---|
[888] | 2563 | INTEGER :: kcom |
---|
[1344] | 2564 | WRITE(*,*) 'mpp_ini_ice: You should not have seen this print! error?', kcom |
---|
[888] | 2565 | END SUBROUTINE mpp_ini_ice |
---|
[869] | 2566 | |
---|
[1345] | 2567 | SUBROUTINE mpp_ini_znl |
---|
| 2568 | WRITE(*,*) 'mpp_ini_znl: You should not have seen this print! error?' |
---|
| 2569 | END SUBROUTINE mpp_ini_znl |
---|
| 2570 | |
---|
[1344] | 2571 | SUBROUTINE mpp_comm_free( kcom ) |
---|
[869] | 2572 | INTEGER :: kcom |
---|
[1344] | 2573 | WRITE(*,*) 'mpp_comm_free: You should not have seen this print! error?', kcom |
---|
[869] | 2574 | END SUBROUTINE mpp_comm_free |
---|
[3] | 2575 | #endif |
---|
[13] | 2576 | !!---------------------------------------------------------------------- |
---|
[3] | 2577 | END MODULE lib_mpp |
---|