Home

Context Navigation

source: palm/trunk/SOURCE/init_pegrid.f90 @ 1723

Last change on this file since 1723 was 1683, checked in by knoop, 9 years ago
last commit documented
Property svn:keywords set to `Id`
File size: 42.3 KB

Line
1	!> @file init_pegrid.f90
2	!--------------------------------------------------------------------------------!
3	! This file is part of PALM.
4	!
5	! PALM is free software: you can redistribute it and/or modify it under the terms
6	! of the GNU General Public License as published by the Free Software Foundation,
7	! either version 3 of the License, or (at your option) any later version.
8	!
9	! PALM is distributed in the hope that it will be useful, but WITHOUT ANY
10	! WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
11	! A PARTICULAR PURPOSE. See the GNU General Public License for more details.
12	!
13	! You should have received a copy of the GNU General Public License along with
14	! PALM. If not, see <http://www.gnu.org/licenses/>.
15	!
16	! Copyright 1997-2014 Leibniz Universitaet Hannover
17	!--------------------------------------------------------------------------------!
18	!
19	! Current revisions:
20	! ------------------
21	!
22	!
23	! Former revisions:
24	! -----------------
25	! $Id: init_pegrid.f90 1683 2015-10-07 23:57:51Z boeske $
26	!
27	! 1682 2015-10-07 23:56:08Z knoop
28	! Code annotations made doxygen readable
29	!
30	! 1677 2015-10-02 13:25:23Z boeske
31	! New MPI-data types for exchange of 3D integer arrays.
32	!
33	! 1575 2015-03-27 09:56:27Z raasch
34	! adjustments for psolver-queries, calculation of ngp_xz added
35	!
36	! 1565 2015-03-09 20:59:31Z suehring
37	! Refine if-clause for setting nbgp.
38	!
39	! 1557 2015-03-05 16:43:04Z suehring
40	! Adjustment for monotonic limiter
41	!
42	! 1468 2014-09-24 14:06:57Z maronga
43	! Adapted for use on up to 6-digit processor cores
44	!
45	! 1435 2014-07-21 10:37:02Z keck
46	! bugfix: added missing parameter coupling_mode_remote to ONLY-attribute
47	!
48	! 1402 2014-05-09 14:25:13Z raasch
49	! location messages modified
50	!
51	! 1384 2014-05-02 14:31:06Z raasch
52	! location messages added
53	!
54	! 1353 2014-04-08 15:21:23Z heinze
55	! REAL constants provided with KIND-attribute
56	!
57	! 1322 2014-03-20 16:38:49Z raasch
58	! REAL functions provided with KIND-attribute
59	!
60	! 1320 2014-03-20 08:40:49Z raasch
61	! ONLY-attribute added to USE-statements,
62	! kind-parameters added to all INTEGER and REAL declaration statements,
63	! kinds are defined in new module kinds,
64	! revision history before 2012 removed,
65	! comment fields (!:) to be used for variable explanations added to
66	! all variable declaration statements
67	!
68	! 1304 2014-03-12 10:29:42Z raasch
69	! bugfix: single core MPI runs missed some settings of transpose indices
70	!
71	! 1212 2013-08-15 08:46:27Z raasch
72	! error message for poisfft_hybrid removed
73	!
74	! 1159 2013-05-21 11:58:22Z fricke
75	! dirichlet/neumann and neumann/dirichlet removed
76	!
77	! 1139 2013-04-18 07:25:03Z raasch
78	! bugfix for calculating the id of the PE carrying the recycling plane
79	!
80	! 1111 2013-03-08 23:54:10Z raasch
81	! initialization of poisfft moved to module poisfft
82	!
83	! 1092 2013-02-02 11:24:22Z raasch
84	! unused variables removed
85	!
86	! 1056 2012-11-16 15:28:04Z raasch
87	! Indices for arrays n.._mg start from zero due to definition of arrays f2 and
88	! p2 as automatic arrays in recursive subroutine next_mg_level
89	!
90	! 1041 2012-11-06 02:36:29Z raasch
91	! a 2d virtual processor topology is used by default for all machines
92	!
93	! 1036 2012-10-22 13:43:42Z raasch
94	! code put under GPL (PALM 3.9)
95	!
96	! 1003 2012-09-14 14:35:53Z raasch
97	! subdomains must have identical size (grid matching = "match" removed)
98	!
99	! 1001 2012-09-13 14:08:46Z raasch
100	! all actions concerning upstream-spline-method removed
101	!
102	! 978 2012-08-09 08:28:32Z fricke
103	! dirichlet/neumann and neumann/dirichlet added
104	! nxlu and nysv are also calculated for inflow boundary
105	!
106	! 809 2012-01-30 13:32:58Z maronga
107	! Bugfix: replaced .AND. and .NOT. with && and ! in the preprocessor directives
108	!
109	! 807 2012-01-25 11:53:51Z maronga
110	! New cpp directive "__check" implemented which is used by check_namelist_files
111	!
112	! Revision 1.1 1997/07/24 11:15:09 raasch
113	! Initial revision
114	!
115	!
116	! Description:
117	! ------------
118	!> Determination of the virtual processor topology (if not prescribed by the
119	!> user)and computation of the grid point number and array bounds of the local
120	!> domains.
121	!------------------------------------------------------------------------------!
122	SUBROUTINE init_pegrid
123
124
125	USE control_parameters, &
126	ONLY: bc_lr, bc_ns, coupling_mode, coupling_mode_remote, &
127	coupling_topology, dt_dosp, gathered_size, grid_level, &
128	grid_level_count, host, inflow_l, inflow_n, inflow_r, inflow_s, &
129	io_blocks, io_group, maximum_grid_level, &
130	maximum_parallel_io_streams, message_string, &
131	mg_switch_to_pe0_level, momentum_advec, neutral, psolver, &
132	outflow_l, outflow_n, outflow_r, outflow_s, recycling_width, &
133	scalar_advec, subdomain_size
134
135	USE grid_variables, &
136	ONLY: dx
137
138	USE indices, &
139	ONLY: mg_loc_ind, nbgp, nnx, nny, nnz, nx, nx_a, nx_o, nxl, nxl_mg, &
140	nxlu, nxr, nxr_mg, ny, ny_a, ny_o, nyn, nyn_mg, nys, nys_mg, &
141	nysv, nz, nzb, nzt, nzt_mg, wall_flags_1, wall_flags_2, &
142	wall_flags_3, wall_flags_4, wall_flags_5, wall_flags_6, &
143	wall_flags_7, wall_flags_8, wall_flags_9, wall_flags_10
144
145	USE kinds
146
147	USE pegrid
148
149	USE transpose_indices, &
150	ONLY: nxl_y, nxl_yd, nxl_z, nxr_y, nxr_yd, nxr_z, nyn_x, nyn_z, nys_x,&
151	nys_z, nzb_x, nzb_y, nzb_yd, nzt_x, nzt_yd, nzt_y
152
153	IMPLICIT NONE
154
155	INTEGER(iwp) :: i !<
156	INTEGER(iwp) :: id_inflow_l !<
157	INTEGER(iwp) :: id_recycling_l !<
158	INTEGER(iwp) :: ind(5) !<
159	INTEGER(iwp) :: j !<
160	INTEGER(iwp) :: k !<
161	INTEGER(iwp) :: maximum_grid_level_l !<
162	INTEGER(iwp) :: mg_levels_x !<
163	INTEGER(iwp) :: mg_levels_y !<
164	INTEGER(iwp) :: mg_levels_z !<
165	INTEGER(iwp) :: mg_switch_to_pe0_level_l !<
166	INTEGER(iwp) :: nnx_y !<
167	INTEGER(iwp) :: nnx_z !<
168	INTEGER(iwp) :: nny_x !<
169	INTEGER(iwp) :: nny_z !<
170	INTEGER(iwp) :: nnz_x !<
171	INTEGER(iwp) :: nnz_y !<
172	INTEGER(iwp) :: numproc_sqr !<
173	INTEGER(iwp) :: nxl_l !<
174	INTEGER(iwp) :: nxr_l !<
175	INTEGER(iwp) :: nyn_l !<
176	INTEGER(iwp) :: nys_l !<
177	INTEGER(iwp) :: nzb_l !<
178	INTEGER(iwp) :: nzt_l !<
179	INTEGER(iwp) :: omp_get_num_threads !<
180
181	INTEGER(iwp), DIMENSION(:), ALLOCATABLE :: ind_all !<
182	INTEGER(iwp), DIMENSION(:), ALLOCATABLE :: nxlf !<
183	INTEGER(iwp), DIMENSION(:), ALLOCATABLE :: nxrf !<
184	INTEGER(iwp), DIMENSION(:), ALLOCATABLE :: nynf !<
185	INTEGER(iwp), DIMENSION(:), ALLOCATABLE :: nysf !<
186
187	INTEGER(iwp), DIMENSION(2) :: pdims_remote !<
188
189	#if defined( __mpi2 )
190	LOGICAL :: found !<
191	#endif
192
193	!
194	!-- Get the number of OpenMP threads
195	!$OMP PARALLEL
196	#if defined( __intel_openmp_bug )
197	threads_per_task = omp_get_num_threads()
198	#else
199	!$ threads_per_task = omp_get_num_threads()
200	#endif
201	!$OMP END PARALLEL
202
203
204	#if defined( __parallel )
205
206	CALL location_message( 'creating virtual PE grids + MPI derived data types', &
207	.FALSE. )
208	!
209	!-- Determine the processor topology or check it, if prescribed by the user
210	IF ( npex == -1 .AND. npey == -1 ) THEN
211
212	!
213	!-- Automatic determination of the topology
214	numproc_sqr = SQRT( REAL( numprocs, KIND=wp ) )
215	pdims(1) = MAX( numproc_sqr , 1 )
216	DO WHILE ( MOD( numprocs , pdims(1) ) /= 0 )
217	pdims(1) = pdims(1) - 1
218	ENDDO
219	pdims(2) = numprocs / pdims(1)
220
221	ELSEIF ( npex /= -1 .AND. npey /= -1 ) THEN
222
223	!
224	!-- Prescribed by user. Number of processors on the prescribed topology
225	!-- must be equal to the number of PEs available to the job
226	IF ( ( npex * npey ) /= numprocs ) THEN
227	WRITE( message_string, * ) 'number of PEs of the prescribed ', &
228	'topology (', npex*npey,') does not match & the number of ', &
229	'PEs available to the job (', numprocs, ')'
230	CALL message( 'init_pegrid', 'PA0221', 1, 2, 0, 6, 0 )
231	ENDIF
232	pdims(1) = npex
233	pdims(2) = npey
234
235	ELSE
236	!
237	!-- If the processor topology is prescribed by the user, the number of
238	!-- PEs must be given in both directions
239	message_string = 'if the processor topology is prescribed by the, ' // &
240	' user& both values of "npex" and "npey" must be given ' // &
241	'in the &NAMELIST-parameter file'
242	CALL message( 'init_pegrid', 'PA0222', 1, 2, 0, 6, 0 )
243
244	ENDIF
245
246	!
247	!-- For communication speedup, set barriers in front of collective
248	!-- communications by default on SGI-type systems
249	IF ( host(3:5) == 'sgi' ) collective_wait = .TRUE.
250
251	!
252	!-- If necessary, set horizontal boundary conditions to non-cyclic
253	IF ( bc_lr /= 'cyclic' ) cyclic(1) = .FALSE.
254	IF ( bc_ns /= 'cyclic' ) cyclic(2) = .FALSE.
255
256
257	#if ! defined( __check)
258	!
259	!-- Create the virtual processor grid
260	CALL MPI_CART_CREATE( comm_palm, ndim, pdims, cyclic, reorder, &
261	comm2d, ierr )
262	CALL MPI_COMM_RANK( comm2d, myid, ierr )
263	WRITE (myid_char,'(''_'',I6.6)') myid
264
265	CALL MPI_CART_COORDS( comm2d, myid, ndim, pcoord, ierr )
266	CALL MPI_CART_SHIFT( comm2d, 0, 1, pleft, pright, ierr )
267	CALL MPI_CART_SHIFT( comm2d, 1, 1, psouth, pnorth, ierr )
268
269	!
270	!-- Determine sub-topologies for transpositions
271	!-- Transposition from z to x:
272	remain_dims(1) = .TRUE.
273	remain_dims(2) = .FALSE.
274	CALL MPI_CART_SUB( comm2d, remain_dims, comm1dx, ierr )
275	CALL MPI_COMM_RANK( comm1dx, myidx, ierr )
276	!
277	!-- Transposition from x to y
278	remain_dims(1) = .FALSE.
279	remain_dims(2) = .TRUE.
280	CALL MPI_CART_SUB( comm2d, remain_dims, comm1dy, ierr )
281	CALL MPI_COMM_RANK( comm1dy, myidy, ierr )
282
283	#endif
284
285	!
286	!-- Calculate array bounds along x-direction for every PE.
287	ALLOCATE( nxlf(0:pdims(1)-1), nxrf(0:pdims(1)-1), nynf(0:pdims(2)-1), &
288	nysf(0:pdims(2)-1) )
289
290	IF ( MOD( nx+1 , pdims(1) ) /= 0 ) THEN
291	WRITE( message_string, * ) 'x-direction: gridpoint number (',nx+1,') ',&
292	'is not an& integral divisor of the number ', &
293	'processors (', pdims(1),')'
294	CALL message( 'init_pegrid', 'PA0225', 1, 2, 0, 6, 0 )
295	ELSE
296	nnx = ( nx + 1 ) / pdims(1)
297	IF ( nnx*pdims(1) - ( nx + 1) > nnx ) THEN
298	WRITE( message_string, * ) 'x-direction: nx does not match the', &
299	'requirements given by the number of PEs &used', &
300	'& please use nx = ', nx - ( pdims(1) - ( nnx*pdims(1) &
301	- ( nx + 1 ) ) ), ' instead of nx =', nx
302	CALL message( 'init_pegrid', 'PA0226', 1, 2, 0, 6, 0 )
303	ENDIF
304	ENDIF
305
306	!
307	!-- Left and right array bounds, number of gridpoints
308	DO i = 0, pdims(1)-1
309	nxlf(i) = i * nnx
310	nxrf(i) = ( i + 1 ) * nnx - 1
311	ENDDO
312
313	!
314	!-- Calculate array bounds in y-direction for every PE.
315	IF ( MOD( ny+1 , pdims(2) ) /= 0 ) THEN
316	WRITE( message_string, * ) 'y-direction: gridpoint number (',ny+1,') ', &
317	'is not an& integral divisor of the number of', &
318	'processors (', pdims(2),')'
319	CALL message( 'init_pegrid', 'PA0227', 1, 2, 0, 6, 0 )
320	ELSE
321	nny = ( ny + 1 ) / pdims(2)
322	IF ( nny*pdims(2) - ( ny + 1) > nny ) THEN
323	WRITE( message_string, * ) 'y-direction: ny does not match the', &
324	'requirements given by the number of PEs &used ', &
325	'& please use ny = ', ny - ( pdims(2) - ( nnx*pdims(2) &
326	- ( ny + 1 ) ) ), ' instead of ny =', ny
327	CALL message( 'init_pegrid', 'PA0228', 1, 2, 0, 6, 0 )
328	ENDIF
329	ENDIF
330
331	!
332	!-- South and north array bounds
333	DO j = 0, pdims(2)-1
334	nysf(j) = j * nny
335	nynf(j) = ( j + 1 ) * nny - 1
336	ENDDO
337
338	!
339	!-- Local array bounds of the respective PEs
340	nxl = nxlf(pcoord(1))
341	nxr = nxrf(pcoord(1))
342	nys = nysf(pcoord(2))
343	nyn = nynf(pcoord(2))
344	nzb = 0
345	nzt = nz
346	nnz = nz
347
348	!
349	!-- Set switches to define if the PE is situated at the border of the virtual
350	!-- processor grid
351	IF ( nxl == 0 ) left_border_pe = .TRUE.
352	IF ( nxr == nx ) right_border_pe = .TRUE.
353	IF ( nys == 0 ) south_border_pe = .TRUE.
354	IF ( nyn == ny ) north_border_pe = .TRUE.
355
356	!
357	!-- Calculate array bounds and gridpoint numbers for the transposed arrays
358	!-- (needed in the pressure solver)
359	!-- For the transposed arrays, cyclic boundaries as well as top and bottom
360	!-- boundaries are omitted, because they are obstructive to the transposition
361
362	!
363	!-- 1. transposition z --> x
364	!-- This transposition is not neccessary in case of a 1d-decomposition along x
365	nys_x = nys
366	nyn_x = nyn
367	nny_x = nny
368	nnz_x = nz / pdims(1)
369	nzb_x = 1 + myidx * nnz_x
370	nzt_x = ( myidx + 1 ) * nnz_x
371	sendrecvcount_zx = nnx * nny * nnz_x
372
373	IF ( pdims(2) /= 1 ) THEN
374	IF ( MOD( nz , pdims(1) ) /= 0 ) THEN
375	WRITE( message_string, * ) 'transposition z --> x:', &
376	'&nz=',nz,' is not an integral divisior of pdims(1)=', &
377	pdims(1)
378	CALL message( 'init_pegrid', 'PA0230', 1, 2, 0, 6, 0 )
379	ENDIF
380	ENDIF
381
382	!
383	!-- 2. transposition x --> y
384	nnz_y = nnz_x
385	nzb_y = nzb_x
386	nzt_y = nzt_x
387	IF ( MOD( nx+1 , pdims(2) ) /= 0 ) THEN
388	WRITE( message_string, * ) 'transposition x --> y:', &
389	'&nx+1=',nx+1,' is not an integral divisor of ',&
390	'pdims(2)=',pdims(2)
391	CALL message( 'init_pegrid', 'PA0231', 1, 2, 0, 6, 0 )
392	ENDIF
393	nnx_y = (nx+1) / pdims(2)
394	nxl_y = myidy * nnx_y
395	nxr_y = ( myidy + 1 ) * nnx_y - 1
396	sendrecvcount_xy = nnx_y * nny_x * nnz_y
397
398	!
399	!-- 3. transposition y --> z (ELSE: x --> y in case of 1D-decomposition
400	!-- along x)
401	nnx_z = nnx_y
402	nxl_z = nxl_y
403	nxr_z = nxr_y
404	nny_z = (ny+1) / pdims(1)
405	nys_z = myidx * nny_z
406	nyn_z = ( myidx + 1 ) * nny_z - 1
407	sendrecvcount_yz = nnx_y * nny_z * nnz_y
408
409	IF ( pdims(2) /= 1 ) THEN
410	!
411	!-- y --> z
412	!-- This transposition is not neccessary in case of a 1d-decomposition
413	!-- along x, except that the uptream-spline method is switched on
414	IF ( MOD( ny+1 , pdims(1) ) /= 0 ) THEN
415	WRITE( message_string, * ) 'transposition y --> z:', &
416	'& ny+1=',ny+1,' is not an integral divisor of ',&
417	'pdims(1)=',pdims(1)
418	CALL message( 'init_pegrid', 'PA0232', 1, 2, 0, 6, 0 )
419	ENDIF
420
421	ELSE
422	!
423	!-- x --> y. This condition must be fulfilled for a 1D-decomposition along x
424	IF ( MOD( ny+1 , pdims(1) ) /= 0 ) THEN
425	WRITE( message_string, * ) 'transposition x --> y:', &
426	'& ny+1=',ny+1,' is not an integral divisor of ',&
427	'pdims(1)=',pdims(1)
428	CALL message( 'init_pegrid', 'PA0233', 1, 2, 0, 6, 0 )
429	ENDIF
430
431	ENDIF
432
433	!
434	!-- Indices for direct transpositions z --> y (used for calculating spectra)
435	IF ( dt_dosp /= 9999999.9_wp ) THEN
436	IF ( MOD( nz, pdims(2) ) /= 0 ) THEN
437	WRITE( message_string, * ) 'direct transposition z --> y (needed ', &
438	'for spectra):& nz=',nz,' is not an integral divisor of ',&
439	'pdims(2)=',pdims(2)
440	CALL message( 'init_pegrid', 'PA0234', 1, 2, 0, 6, 0 )
441	ELSE
442	nxl_yd = nxl
443	nxr_yd = nxr
444	nzb_yd = 1 + myidy * ( nz / pdims(2) )
445	nzt_yd = ( myidy + 1 ) * ( nz / pdims(2) )
446	sendrecvcount_zyd = nnx * nny * ( nz / pdims(2) )
447	ENDIF
448	ENDIF
449
450	!
451	!-- Indices for direct transpositions y --> x (they are only possible in case
452	!-- of a 1d-decomposition along x)
453	IF ( pdims(2) == 1 ) THEN
454	nny_x = nny / pdims(1)
455	nys_x = myid * nny_x
456	nyn_x = ( myid + 1 ) * nny_x - 1
457	nzb_x = 1
458	nzt_x = nz
459	sendrecvcount_xy = nnx * nny_x * nz
460	ENDIF
461
462	!
463	!-- Indices for direct transpositions x --> y (they are only possible in case
464	!-- of a 1d-decomposition along y)
465	IF ( pdims(1) == 1 ) THEN
466	nnx_y = nnx / pdims(2)
467	nxl_y = myid * nnx_y
468	nxr_y = ( myid + 1 ) * nnx_y - 1
469	nzb_y = 1
470	nzt_y = nz
471	sendrecvcount_xy = nnx_y * nny * nz
472	ENDIF
473
474	!
475	!-- Arrays for storing the array bounds are needed any more
476	DEALLOCATE( nxlf , nxrf , nynf , nysf )
477
478
479	#if ! defined( __check)
480	!
481	!-- Collect index bounds from other PEs (to be written to restart file later)
482	ALLOCATE( hor_index_bounds(4,0:numprocs-1) )
483
484	IF ( myid == 0 ) THEN
485
486	hor_index_bounds(1,0) = nxl
487	hor_index_bounds(2,0) = nxr
488	hor_index_bounds(3,0) = nys
489	hor_index_bounds(4,0) = nyn
490
491	!
492	!-- Receive data from all other PEs
493	DO i = 1, numprocs-1
494	CALL MPI_RECV( ibuf, 4, MPI_INTEGER, i, MPI_ANY_TAG, comm2d, status, &
495	ierr )
496	hor_index_bounds(:,i) = ibuf(1:4)
497	ENDDO
498
499	ELSE
500	!
501	!-- Send index bounds to PE0
502	ibuf(1) = nxl
503	ibuf(2) = nxr
504	ibuf(3) = nys
505	ibuf(4) = nyn
506	CALL MPI_SEND( ibuf, 4, MPI_INTEGER, 0, myid, comm2d, ierr )
507
508	ENDIF
509
510	#endif
511
512	#if defined( __print )
513	!
514	!-- Control output
515	IF ( myid == 0 ) THEN
516	PRINT, ' processor topology *'
517	PRINT*, ' '
518	PRINT*, 'myid pcoord left right south north idx idy nxl: nxr',&
519	&' nys: nyn'
520	PRINT*, '------------------------------------------------------------',&
521	&'-----------'
522	WRITE (*,1000) 0, pcoord(1), pcoord(2), pleft, pright, psouth, pnorth, &
523	myidx, myidy, nxl, nxr, nys, nyn
524	1000 FORMAT (I4,2X,'(',I3,',',I3,')',3X,I4,2X,I4,3X,I4,2X,I4,2X,I3,1X,I3, &
525	2(2X,I4,':',I4))
526
527	!
528	!-- Receive data from the other PEs
529	DO i = 1,numprocs-1
530	CALL MPI_RECV( ibuf, 12, MPI_INTEGER, i, MPI_ANY_TAG, comm2d, status, &
531	ierr )
532	WRITE (*,1000) i, ( ibuf(j) , j = 1,12 )
533	ENDDO
534	ELSE
535
536	!
537	!-- Send data to PE0
538	ibuf(1) = pcoord(1); ibuf(2) = pcoord(2); ibuf(3) = pleft
539	ibuf(4) = pright; ibuf(5) = psouth; ibuf(6) = pnorth; ibuf(7) = myidx
540	ibuf(8) = myidy; ibuf(9) = nxl; ibuf(10) = nxr; ibuf(11) = nys
541	ibuf(12) = nyn
542	CALL MPI_SEND( ibuf, 12, MPI_INTEGER, 0, myid, comm2d, ierr )
543	ENDIF
544	#endif
545
546	#if defined( __parallel ) && ! defined( __check)
547	#if defined( __mpi2 )
548	!
549	!-- In case of coupled runs, get the port name on PE0 of the atmosphere model
550	!-- and pass it to PE0 of the ocean model
551	IF ( myid == 0 ) THEN
552
553	IF ( coupling_mode == 'atmosphere_to_ocean' ) THEN
554
555	CALL MPI_OPEN_PORT( MPI_INFO_NULL, port_name, ierr )
556
557	CALL MPI_PUBLISH_NAME( 'palm_coupler', MPI_INFO_NULL, port_name, &
558	ierr )
559
560	!
561	!-- Write a flag file for the ocean model and the other atmosphere
562	!-- processes.
563	!-- There seems to be a bug in MPICH2 which causes hanging processes
564	!-- in case that execution of LOOKUP_NAME is continued too early
565	!-- (i.e. before the port has been created)
566	OPEN( 90, FILE='COUPLING_PORT_OPENED', FORM='FORMATTED' )
567	WRITE ( 90, '(''TRUE'')' )
568	CLOSE ( 90 )
569
570	ELSEIF ( coupling_mode == 'ocean_to_atmosphere' ) THEN
571
572	!
573	!-- Continue only if the atmosphere model has created the port.
574	!-- There seems to be a bug in MPICH2 which causes hanging processes
575	!-- in case that execution of LOOKUP_NAME is continued too early
576	!-- (i.e. before the port has been created)
577	INQUIRE( FILE='COUPLING_PORT_OPENED', EXIST=found )
578	DO WHILE ( .NOT. found )
579	INQUIRE( FILE='COUPLING_PORT_OPENED', EXIST=found )
580	ENDDO
581
582	CALL MPI_LOOKUP_NAME( 'palm_coupler', MPI_INFO_NULL, port_name, ierr )
583
584	ENDIF
585
586	ENDIF
587
588	!
589	!-- In case of coupled runs, establish the connection between the atmosphere
590	!-- and the ocean model and define the intercommunicator (comm_inter)
591	CALL MPI_BARRIER( comm2d, ierr )
592	IF ( coupling_mode == 'atmosphere_to_ocean' ) THEN
593
594	CALL MPI_COMM_ACCEPT( port_name, MPI_INFO_NULL, 0, MPI_COMM_WORLD, &
595	comm_inter, ierr )
596	coupling_mode_remote = 'ocean_to_atmosphere'
597
598	ELSEIF ( coupling_mode == 'ocean_to_atmosphere' ) THEN
599
600	CALL MPI_COMM_CONNECT( port_name, MPI_INFO_NULL, 0, MPI_COMM_WORLD, &
601	comm_inter, ierr )
602	coupling_mode_remote = 'atmosphere_to_ocean'
603
604	ENDIF
605	#endif
606
607	!
608	!-- Determine the number of ghost point layers
609	IF ( ( scalar_advec == 'ws-scheme' .AND. .NOT. neutral ) .OR. &
610	scalar_advec == 'ws-scheme-mono' .OR. &
611	momentum_advec == 'ws-scheme' ) THEN
612	nbgp = 3
613	ELSE
614	nbgp = 1
615	ENDIF
616
617	!
618	!-- Create a new MPI derived datatype for the exchange of surface (xy) data,
619	!-- which is needed for coupled atmosphere-ocean runs.
620	!-- First, calculate number of grid points of an xy-plane.
621	ngp_xy = ( nxr - nxl + 1 + 2 * nbgp ) * ( nyn - nys + 1 + 2 * nbgp )
622	CALL MPI_TYPE_VECTOR( ngp_xy, 1, nzt-nzb+2, MPI_REAL, type_xy, ierr )
623	CALL MPI_TYPE_COMMIT( type_xy, ierr )
624
625	IF ( TRIM( coupling_mode ) /= 'uncoupled' ) THEN
626
627	!
628	!-- Pass the number of grid points of the atmosphere model to
629	!-- the ocean model and vice versa
630	IF ( coupling_mode == 'atmosphere_to_ocean' ) THEN
631
632	nx_a = nx
633	ny_a = ny
634
635	IF ( myid == 0 ) THEN
636
637	CALL MPI_SEND( nx_a, 1, MPI_INTEGER, numprocs, 1, comm_inter, &
638	ierr )
639	CALL MPI_SEND( ny_a, 1, MPI_INTEGER, numprocs, 2, comm_inter, &
640	ierr )
641	CALL MPI_SEND( pdims, 2, MPI_INTEGER, numprocs, 3, comm_inter, &
642	ierr )
643	CALL MPI_RECV( nx_o, 1, MPI_INTEGER, numprocs, 4, comm_inter, &
644	status, ierr )
645	CALL MPI_RECV( ny_o, 1, MPI_INTEGER, numprocs, 5, comm_inter, &
646	status, ierr )
647	CALL MPI_RECV( pdims_remote, 2, MPI_INTEGER, numprocs, 6, &
648	comm_inter, status, ierr )
649	ENDIF
650
651	CALL MPI_BCAST( nx_o, 1, MPI_INTEGER, 0, comm2d, ierr )
652	CALL MPI_BCAST( ny_o, 1, MPI_INTEGER, 0, comm2d, ierr )
653	CALL MPI_BCAST( pdims_remote, 2, MPI_INTEGER, 0, comm2d, ierr )
654
655	ELSEIF ( coupling_mode == 'ocean_to_atmosphere' ) THEN
656
657	nx_o = nx
658	ny_o = ny
659
660	IF ( myid == 0 ) THEN
661
662	CALL MPI_RECV( nx_a, 1, MPI_INTEGER, 0, 1, comm_inter, status, &
663	ierr )
664	CALL MPI_RECV( ny_a, 1, MPI_INTEGER, 0, 2, comm_inter, status, &
665	ierr )
666	CALL MPI_RECV( pdims_remote, 2, MPI_INTEGER, 0, 3, comm_inter, &
667	status, ierr )
668	CALL MPI_SEND( nx_o, 1, MPI_INTEGER, 0, 4, comm_inter, ierr )
669	CALL MPI_SEND( ny_o, 1, MPI_INTEGER, 0, 5, comm_inter, ierr )
670	CALL MPI_SEND( pdims, 2, MPI_INTEGER, 0, 6, comm_inter, ierr )
671	ENDIF
672
673	CALL MPI_BCAST( nx_a, 1, MPI_INTEGER, 0, comm2d, ierr)
674	CALL MPI_BCAST( ny_a, 1, MPI_INTEGER, 0, comm2d, ierr)
675	CALL MPI_BCAST( pdims_remote, 2, MPI_INTEGER, 0, comm2d, ierr)
676
677	ENDIF
678
679	ngp_a = ( nx_a+1 + 2 * nbgp ) * ( ny_a+1 + 2 * nbgp )
680	ngp_o = ( nx_o+1 + 2 * nbgp ) * ( ny_o+1 + 2 * nbgp )
681
682	!
683	!-- Determine if the horizontal grid and the number of PEs in ocean and
684	!-- atmosphere is same or not
685	IF ( nx_o == nx_a .AND. ny_o == ny_a .AND. &
686	pdims(1) == pdims_remote(1) .AND. pdims(2) == pdims_remote(2) ) &
687	THEN
688	coupling_topology = 0
689	ELSE
690	coupling_topology = 1
691	ENDIF
692
693	!
694	!-- Determine the target PEs for the exchange between ocean and
695	!-- atmosphere (comm2d)
696	IF ( coupling_topology == 0 ) THEN
697	!
698	!-- In case of identical topologies, every atmosphere PE has exactly one
699	!-- ocean PE counterpart and vice versa
700	IF ( TRIM( coupling_mode ) == 'atmosphere_to_ocean' ) THEN
701	target_id = myid + numprocs
702	ELSE
703	target_id = myid
704	ENDIF
705
706	ELSE
707	!
708	!-- In case of nonequivalent topology in ocean and atmosphere only for
709	!-- PE0 in ocean and PE0 in atmosphere a target_id is needed, since
710	!-- data echxchange between ocean and atmosphere will be done only
711	!-- between these PEs.
712	IF ( myid == 0 ) THEN
713
714	IF ( TRIM( coupling_mode ) == 'atmosphere_to_ocean' ) THEN
715	target_id = numprocs
716	ELSE
717	target_id = 0
718	ENDIF
719
720	ENDIF
721
722	ENDIF
723
724	ENDIF
725
726
727	#endif
728
729	#else
730
731	!
732	!-- Array bounds when running on a single PE (respectively a non-parallel
733	!-- machine)
734	nxl = 0
735	nxr = nx
736	nnx = nxr - nxl + 1
737	nys = 0
738	nyn = ny
739	nny = nyn - nys + 1
740	nzb = 0
741	nzt = nz
742	nnz = nz
743
744	ALLOCATE( hor_index_bounds(4,0:0) )
745	hor_index_bounds(1,0) = nxl
746	hor_index_bounds(2,0) = nxr
747	hor_index_bounds(3,0) = nys
748	hor_index_bounds(4,0) = nyn
749
750	!
751	!-- Array bounds for the pressure solver (in the parallel code, these bounds
752	!-- are the ones for the transposed arrays)
753	nys_x = nys
754	nyn_x = nyn
755	nzb_x = nzb + 1
756	nzt_x = nzt
757
758	nxl_y = nxl
759	nxr_y = nxr
760	nzb_y = nzb + 1
761	nzt_y = nzt
762
763	nxl_z = nxl
764	nxr_z = nxr
765	nys_z = nys
766	nyn_z = nyn
767
768	#endif
769
770	!
771	!-- Calculate number of grid levels necessary for the multigrid poisson solver
772	!-- as well as the gridpoint indices on each level
773	IF ( psolver(1:9) == 'multigrid' ) THEN
774
775	!
776	!-- First calculate number of possible grid levels for the subdomains
777	mg_levels_x = 1
778	mg_levels_y = 1
779	mg_levels_z = 1
780
781	i = nnx
782	DO WHILE ( MOD( i, 2 ) == 0 .AND. i /= 2 )
783	i = i / 2
784	mg_levels_x = mg_levels_x + 1
785	ENDDO
786
787	j = nny
788	DO WHILE ( MOD( j, 2 ) == 0 .AND. j /= 2 )
789	j = j / 2
790	mg_levels_y = mg_levels_y + 1
791	ENDDO
792
793	k = nz ! do not use nnz because it might be > nz due to transposition
794	! requirements
795	DO WHILE ( MOD( k, 2 ) == 0 .AND. k /= 2 )
796	k = k / 2
797	mg_levels_z = mg_levels_z + 1
798	ENDDO
799
800	maximum_grid_level = MIN( mg_levels_x, mg_levels_y, mg_levels_z )
801
802	!
803	!-- Find out, if the total domain allows more levels. These additional
804	!-- levels are identically processed on all PEs.
805	IF ( numprocs > 1 .AND. mg_switch_to_pe0_level /= -1 ) THEN
806
807	IF ( mg_levels_z > MIN( mg_levels_x, mg_levels_y ) ) THEN
808
809	mg_switch_to_pe0_level_l = maximum_grid_level
810
811	mg_levels_x = 1
812	mg_levels_y = 1
813
814	i = nx+1
815	DO WHILE ( MOD( i, 2 ) == 0 .AND. i /= 2 )
816	i = i / 2
817	mg_levels_x = mg_levels_x + 1
818	ENDDO
819
820	j = ny+1
821	DO WHILE ( MOD( j, 2 ) == 0 .AND. j /= 2 )
822	j = j / 2
823	mg_levels_y = mg_levels_y + 1
824	ENDDO
825
826	maximum_grid_level_l = MIN( mg_levels_x, mg_levels_y, mg_levels_z )
827
828	IF ( maximum_grid_level_l > mg_switch_to_pe0_level_l ) THEN
829	mg_switch_to_pe0_level_l = maximum_grid_level_l - &
830	mg_switch_to_pe0_level_l + 1
831	ELSE
832	mg_switch_to_pe0_level_l = 0
833	ENDIF
834
835	ELSE
836	mg_switch_to_pe0_level_l = 0
837	maximum_grid_level_l = maximum_grid_level
838
839	ENDIF
840
841	!
842	!-- Use switch level calculated above only if it is not pre-defined
843	!-- by user
844	IF ( mg_switch_to_pe0_level == 0 ) THEN
845	IF ( mg_switch_to_pe0_level_l /= 0 ) THEN
846	mg_switch_to_pe0_level = mg_switch_to_pe0_level_l
847	maximum_grid_level = maximum_grid_level_l
848	ENDIF
849
850	ELSE
851	!
852	!-- Check pre-defined value and reset to default, if neccessary
853	IF ( mg_switch_to_pe0_level < mg_switch_to_pe0_level_l .OR. &
854	mg_switch_to_pe0_level >= maximum_grid_level_l ) THEN
855	message_string = 'mg_switch_to_pe0_level ' // &
856	'out of range and reset to default (=0)'
857	CALL message( 'init_pegrid', 'PA0235', 0, 1, 0, 6, 0 )
858	mg_switch_to_pe0_level = 0
859	ELSE
860	!
861	!-- Use the largest number of possible levels anyway and recalculate
862	!-- the switch level to this largest number of possible values
863	maximum_grid_level = maximum_grid_level_l
864
865	ENDIF
866
867	ENDIF
868
869	ENDIF
870
871	ALLOCATE( grid_level_count(maximum_grid_level), &
872	nxl_mg(0:maximum_grid_level), nxr_mg(0:maximum_grid_level), &
873	nyn_mg(0:maximum_grid_level), nys_mg(0:maximum_grid_level), &
874	nzt_mg(0:maximum_grid_level) )
875
876	grid_level_count = 0
877	!
878	!-- Index zero required as dummy due to definition of arrays f2 and p2 in
879	!-- recursive subroutine next_mg_level
880	nxl_mg(0) = 0; nxr_mg(0) = 0; nyn_mg(0) = 0; nys_mg(0) = 0; nzt_mg(0) = 0
881
882	nxl_l = nxl; nxr_l = nxr; nys_l = nys; nyn_l = nyn; nzt_l = nzt
883
884	DO i = maximum_grid_level, 1 , -1
885
886	IF ( i == mg_switch_to_pe0_level ) THEN
887	#if defined( __parallel ) && ! defined( __check )
888	!
889	!-- Save the grid size of the subdomain at the switch level, because
890	!-- it is needed in poismg.
891	ind(1) = nxl_l; ind(2) = nxr_l
892	ind(3) = nys_l; ind(4) = nyn_l
893	ind(5) = nzt_l
894	ALLOCATE( ind_all(5*numprocs), mg_loc_ind(5,0:numprocs-1) )
895	CALL MPI_ALLGATHER( ind, 5, MPI_INTEGER, ind_all, 5, &
896	MPI_INTEGER, comm2d, ierr )
897	DO j = 0, numprocs-1
898	DO k = 1, 5
899	mg_loc_ind(k,j) = ind_all(k+j*5)
900	ENDDO
901	ENDDO
902	DEALLOCATE( ind_all )
903	!
904	!-- Calculate the grid size of the total domain
905	nxr_l = ( nxr_l-nxl_l+1 ) * pdims(1) - 1
906	nxl_l = 0
907	nyn_l = ( nyn_l-nys_l+1 ) * pdims(2) - 1
908	nys_l = 0
909	!
910	!-- The size of this gathered array must not be larger than the
911	!-- array tend, which is used in the multigrid scheme as a temporary
912	!-- array. Therefore the subdomain size of an PE is calculated and
913	!-- the size of the gathered grid. These values are used in
914	!-- routines pres and poismg
915	subdomain_size = ( nxr - nxl + 2 * nbgp + 1 ) * &
916	( nyn - nys + 2 * nbgp + 1 ) * ( nzt - nzb + 2 )
917	gathered_size = ( nxr_l - nxl_l + 3 ) * ( nyn_l - nys_l + 3 ) * &
918	( nzt_l - nzb + 2 )
919
920	#elif ! defined ( __parallel )
921	message_string = 'multigrid gather/scatter impossible ' // &
922	'in non parallel mode'
923	CALL message( 'init_pegrid', 'PA0237', 1, 2, 0, 6, 0 )
924	#endif
925	ENDIF
926
927	nxl_mg(i) = nxl_l
928	nxr_mg(i) = nxr_l
929	nys_mg(i) = nys_l
930	nyn_mg(i) = nyn_l
931	nzt_mg(i) = nzt_l
932
933	nxl_l = nxl_l / 2
934	nxr_l = nxr_l / 2
935	nys_l = nys_l / 2
936	nyn_l = nyn_l / 2
937	nzt_l = nzt_l / 2
938
939	ENDDO
940
941	!
942	!-- Temporary problem: Currently calculation of maxerror iin routine poismg crashes
943	!-- if grid data are collected on PE0 already on the finest grid level.
944	!-- To be solved later.
945	IF ( maximum_grid_level == mg_switch_to_pe0_level ) THEN
946	message_string = 'grid coarsening on subdomain level cannot be performed'
947	CALL message( 'poismg', 'PA0236', 1, 2, 0, 6, 0 )
948	ENDIF
949
950	ELSE
951
952	maximum_grid_level = 0
953
954	ENDIF
955
956	!
957	!-- Default level 0 tells exchange_horiz that all ghost planes have to be
958	!-- exchanged. grid_level is adjusted in poismg, where only one ghost plane
959	!-- is required.
960	grid_level = 0
961
962	#if defined( __parallel ) && ! defined ( __check )
963	!
964	!-- Gridpoint number for the exchange of ghost points (y-line for 2D-arrays)
965	ngp_y = nyn - nys + 1 + 2 * nbgp
966
967	!
968	!-- Define new MPI derived datatypes for the exchange of ghost points in
969	!-- x- and y-direction for 2D-arrays (line)
970	CALL MPI_TYPE_VECTOR( nxr-nxl+1+2*nbgp, nbgp, ngp_y, MPI_REAL, type_x, &
971	ierr )
972	CALL MPI_TYPE_COMMIT( type_x, ierr )
973	CALL MPI_TYPE_VECTOR( nxr-nxl+1+2*nbgp, nbgp, ngp_y, MPI_INTEGER, &
974	type_x_int, ierr )
975	CALL MPI_TYPE_COMMIT( type_x_int, ierr )
976
977	CALL MPI_TYPE_VECTOR( nbgp, ngp_y, ngp_y, MPI_REAL, type_y, ierr )
978	CALL MPI_TYPE_COMMIT( type_y, ierr )
979	CALL MPI_TYPE_VECTOR( nbgp, ngp_y, ngp_y, MPI_INTEGER, type_y_int, ierr )
980	CALL MPI_TYPE_COMMIT( type_y_int, ierr )
981
982
983	!
984	!-- Calculate gridpoint numbers for the exchange of ghost points along x
985	!-- (yz-plane for 3D-arrays) and define MPI derived data type(s) for the
986	!-- exchange of ghost points in y-direction (xz-plane).
987	!-- Do these calculations for the model grid and (if necessary) also
988	!-- for the coarser grid levels used in the multigrid method
989	ALLOCATE ( ngp_xz(0:maximum_grid_level), ngp_yz(0:maximum_grid_level), &
990	type_xz(0:maximum_grid_level), type_yz(0:maximum_grid_level) )
991
992	nxl_l = nxl; nxr_l = nxr; nys_l = nys; nyn_l = nyn; nzb_l = nzb; nzt_l = nzt
993
994	!
995	!-- Discern between the model grid, which needs nbgp ghost points and
996	!-- grid levels for the multigrid scheme. In the latter case only one
997	!-- ghost point is necessary.
998	!-- First definition of MPI-datatypes for exchange of ghost layers on normal
999	!-- grid. The following loop is needed for data exchange in poismg.f90.
1000	!
1001	!-- Determine number of grid points of yz-layer for exchange
1002	ngp_yz(0) = (nzt - nzb + 2) * (nyn - nys + 1 + 2 * nbgp)
1003
1004	!
1005	!-- Define an MPI-datatype for the exchange of left/right boundaries.
1006	!-- Although data are contiguous in physical memory (which does not
1007	!-- necessarily require an MPI-derived datatype), the data exchange between
1008	!-- left and right PE's using the MPI-derived type is 10% faster than without.
1009	CALL MPI_TYPE_VECTOR( nxr-nxl+1+2nbgp, nbgp(nzt-nzb+2), ngp_yz(0), &
1010	MPI_REAL, type_xz(0), ierr )
1011	CALL MPI_TYPE_COMMIT( type_xz(0), ierr )
1012
1013	CALL MPI_TYPE_VECTOR( nbgp, ngp_yz(0), ngp_yz(0), MPI_REAL, type_yz(0), &
1014	ierr )
1015	CALL MPI_TYPE_COMMIT( type_yz(0), ierr )
1016
1017	!
1018	!-- Definition of MPI-datatypes for multigrid method (coarser level grids)
1019	IF ( psolver(1:9) == 'multigrid' ) THEN
1020	!
1021	!-- Definition of MPI-datatyoe as above, but only 1 ghost level is used
1022	DO i = maximum_grid_level, 1 , -1
1023
1024	ngp_xz(i) = (nzt_l - nzb_l + 2) * (nxr_l - nxl_l + 3)
1025	ngp_yz(i) = (nzt_l - nzb_l + 2) * (nyn_l - nys_l + 3)
1026
1027	CALL MPI_TYPE_VECTOR( nxr_l-nxl_l+3, nzt_l-nzb_l+2, ngp_yz(i), &
1028	MPI_REAL, type_xz(i), ierr )
1029	CALL MPI_TYPE_COMMIT( type_xz(i), ierr )
1030
1031	CALL MPI_TYPE_VECTOR( 1, ngp_yz(i), ngp_yz(i), MPI_REAL, type_yz(i), &
1032	ierr )
1033	CALL MPI_TYPE_COMMIT( type_yz(i), ierr )
1034
1035	nxl_l = nxl_l / 2
1036	nxr_l = nxr_l / 2
1037	nys_l = nys_l / 2
1038	nyn_l = nyn_l / 2
1039	nzt_l = nzt_l / 2
1040
1041	ENDDO
1042
1043	ENDIF
1044	!
1045	!-- Define data types for exchange of 3D Integer arrays.
1046	ngp_yz_int = (nzt - nzb + 2) * (nyn - nys + 1 + 2 * nbgp)
1047
1048	CALL MPI_TYPE_VECTOR( nxr-nxl+1+2nbgp, nbgp(nzt-nzb+2), ngp_yz_int, &
1049	MPI_INTEGER, type_xz_int, ierr )
1050	CALL MPI_TYPE_COMMIT( type_xz_int, ierr )
1051
1052	CALL MPI_TYPE_VECTOR( nbgp, ngp_yz_int, ngp_yz_int, MPI_INTEGER, type_yz_int, &
1053	ierr )
1054	CALL MPI_TYPE_COMMIT( type_yz_int, ierr )
1055
1056	#endif
1057
1058	#if defined( __parallel ) && ! defined ( __check )
1059	!
1060	!-- Setting of flags for inflow/outflow conditions in case of non-cyclic
1061	!-- horizontal boundary conditions.
1062	IF ( pleft == MPI_PROC_NULL ) THEN
1063	IF ( bc_lr == 'dirichlet/radiation' ) THEN
1064	inflow_l = .TRUE.
1065	ELSEIF ( bc_lr == 'radiation/dirichlet' ) THEN
1066	outflow_l = .TRUE.
1067	ENDIF
1068	ENDIF
1069
1070	IF ( pright == MPI_PROC_NULL ) THEN
1071	IF ( bc_lr == 'dirichlet/radiation' ) THEN
1072	outflow_r = .TRUE.
1073	ELSEIF ( bc_lr == 'radiation/dirichlet' ) THEN
1074	inflow_r = .TRUE.
1075	ENDIF
1076	ENDIF
1077
1078	IF ( psouth == MPI_PROC_NULL ) THEN
1079	IF ( bc_ns == 'dirichlet/radiation' ) THEN
1080	outflow_s = .TRUE.
1081	ELSEIF ( bc_ns == 'radiation/dirichlet' ) THEN
1082	inflow_s = .TRUE.
1083	ENDIF
1084	ENDIF
1085
1086	IF ( pnorth == MPI_PROC_NULL ) THEN
1087	IF ( bc_ns == 'dirichlet/radiation' ) THEN
1088	inflow_n = .TRUE.
1089	ELSEIF ( bc_ns == 'radiation/dirichlet' ) THEN
1090	outflow_n = .TRUE.
1091	ENDIF
1092	ENDIF
1093
1094	!
1095	!-- Broadcast the id of the inflow PE
1096	IF ( inflow_l ) THEN
1097	id_inflow_l = myidx
1098	ELSE
1099	id_inflow_l = 0
1100	ENDIF
1101	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1102	CALL MPI_ALLREDUCE( id_inflow_l, id_inflow, 1, MPI_INTEGER, MPI_SUM, &
1103	comm1dx, ierr )
1104
1105	!
1106	!-- Broadcast the id of the recycling plane
1107	!-- WARNING: needs to be adjusted in case of inflows other than from left side!
1108	IF ( NINT( recycling_width / dx ) >= nxl .AND. &
1109	NINT( recycling_width / dx ) <= nxr ) THEN
1110	id_recycling_l = myidx
1111	ELSE
1112	id_recycling_l = 0
1113	ENDIF
1114	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1115	CALL MPI_ALLREDUCE( id_recycling_l, id_recycling, 1, MPI_INTEGER, MPI_SUM, &
1116	comm1dx, ierr )
1117
1118	CALL location_message( 'finished', .TRUE. )
1119
1120	#elif ! defined ( __parallel )
1121	IF ( bc_lr == 'dirichlet/radiation' ) THEN
1122	inflow_l = .TRUE.
1123	outflow_r = .TRUE.
1124	ELSEIF ( bc_lr == 'radiation/dirichlet' ) THEN
1125	outflow_l = .TRUE.
1126	inflow_r = .TRUE.
1127	ENDIF
1128
1129	IF ( bc_ns == 'dirichlet/radiation' ) THEN
1130	inflow_n = .TRUE.
1131	outflow_s = .TRUE.
1132	ELSEIF ( bc_ns == 'radiation/dirichlet' ) THEN
1133	outflow_n = .TRUE.
1134	inflow_s = .TRUE.
1135	ENDIF
1136	#endif
1137
1138	!
1139	!-- At the inflow or outflow, u or v, respectively, have to be calculated for
1140	!-- one more grid point.
1141	IF ( inflow_l .OR. outflow_l ) THEN
1142	nxlu = nxl + 1
1143	ELSE
1144	nxlu = nxl
1145	ENDIF
1146	IF ( inflow_s .OR. outflow_s ) THEN
1147	nysv = nys + 1
1148	ELSE
1149	nysv = nys
1150	ENDIF
1151
1152	!
1153	!-- Allocate wall flag arrays used in the multigrid solver
1154	IF ( psolver(1:9) == 'multigrid' ) THEN
1155
1156	DO i = maximum_grid_level, 1, -1
1157
1158	SELECT CASE ( i )
1159
1160	CASE ( 1 )
1161	ALLOCATE( wall_flags_1(nzb:nzt_mg(i)+1, &
1162	nys_mg(i)-1:nyn_mg(i)+1, &
1163	nxl_mg(i)-1:nxr_mg(i)+1) )
1164
1165	CASE ( 2 )
1166	ALLOCATE( wall_flags_2(nzb:nzt_mg(i)+1, &
1167	nys_mg(i)-1:nyn_mg(i)+1, &
1168	nxl_mg(i)-1:nxr_mg(i)+1) )
1169
1170	CASE ( 3 )
1171	ALLOCATE( wall_flags_3(nzb:nzt_mg(i)+1, &
1172	nys_mg(i)-1:nyn_mg(i)+1, &
1173	nxl_mg(i)-1:nxr_mg(i)+1) )
1174
1175	CASE ( 4 )
1176	ALLOCATE( wall_flags_4(nzb:nzt_mg(i)+1, &
1177	nys_mg(i)-1:nyn_mg(i)+1, &
1178	nxl_mg(i)-1:nxr_mg(i)+1) )
1179
1180	CASE ( 5 )
1181	ALLOCATE( wall_flags_5(nzb:nzt_mg(i)+1, &
1182	nys_mg(i)-1:nyn_mg(i)+1, &
1183	nxl_mg(i)-1:nxr_mg(i)+1) )
1184
1185	CASE ( 6 )
1186	ALLOCATE( wall_flags_6(nzb:nzt_mg(i)+1, &
1187	nys_mg(i)-1:nyn_mg(i)+1, &
1188	nxl_mg(i)-1:nxr_mg(i)+1) )
1189
1190	CASE ( 7 )
1191	ALLOCATE( wall_flags_7(nzb:nzt_mg(i)+1, &
1192	nys_mg(i)-1:nyn_mg(i)+1, &
1193	nxl_mg(i)-1:nxr_mg(i)+1) )
1194
1195	CASE ( 8 )
1196	ALLOCATE( wall_flags_8(nzb:nzt_mg(i)+1, &
1197	nys_mg(i)-1:nyn_mg(i)+1, &
1198	nxl_mg(i)-1:nxr_mg(i)+1) )
1199
1200	CASE ( 9 )
1201	ALLOCATE( wall_flags_9(nzb:nzt_mg(i)+1, &
1202	nys_mg(i)-1:nyn_mg(i)+1, &
1203	nxl_mg(i)-1:nxr_mg(i)+1) )
1204
1205	CASE ( 10 )
1206	ALLOCATE( wall_flags_10(nzb:nzt_mg(i)+1, &
1207	nys_mg(i)-1:nyn_mg(i)+1, &
1208	nxl_mg(i)-1:nxr_mg(i)+1) )
1209
1210	CASE DEFAULT
1211	message_string = 'more than 10 multigrid levels'
1212	CALL message( 'init_pegrid', 'PA0238', 1, 2, 0, 6, 0 )
1213
1214	END SELECT
1215
1216	ENDDO
1217
1218	ENDIF
1219
1220	!
1221	!-- Calculate the number of groups into which parallel I/O is split.
1222	!-- The default for files which are opened by all PEs (or where each
1223	!-- PE opens his own independent file) is, that all PEs are doing input/output
1224	!-- in parallel at the same time. This might cause performance or even more
1225	!-- severe problems depending on the configuration of the underlying file
1226	!-- system.
1227	!-- First, set the default:
1228	IF ( maximum_parallel_io_streams == -1 .OR. &
1229	maximum_parallel_io_streams > numprocs ) THEN
1230	maximum_parallel_io_streams = numprocs
1231	ENDIF
1232
1233	!
1234	!-- Now calculate the number of io_blocks and the io_group to which the
1235	!-- respective PE belongs. I/O of the groups is done in serial, but in parallel
1236	!-- for all PEs belonging to the same group. A preliminary setting with myid
1237	!-- based on MPI_COMM_WORLD has been done in parin.
1238	io_blocks = numprocs / maximum_parallel_io_streams
1239	io_group = MOD( myid+1, io_blocks )
1240
1241
1242	END SUBROUTINE init_pegrid

Note: See TracBrowser for help on using the repository browser.

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |