Home

Context Navigation

source: palm/trunk/SOURCE/init_pegrid.f90 @ 164

Last change on this file since 164 was 163, checked in by raasch, 17 years ago
bugfixes for turbulent inflow in init_pegrid, inflow_turbulence, and init_3d_model
Property svn:keywords set to `Id`
File size: 32.8 KB

Line
1	SUBROUTINE init_pegrid
2
3	!------------------------------------------------------------------------------!
4	! Actual revisions:
5	! -----------------
6	! Collect on PE0 horizontal index bounds from all other PEs,
7	! broadcast the id of the inflow PE (using the respective communicator)
8	! TEST OUTPUT (TO BE REMOVED) logging mpi2 ierr values
9	!
10	! Former revisions:
11	! -----------------
12	! $Id: init_pegrid.f90 163 2008-05-05 14:09:05Z raasch $
13	!
14	! 114 2007-10-10 00:03:15Z raasch
15	! Allocation of wall flag arrays for multigrid solver
16	!
17	! 108 2007-08-24 15:10:38Z letzel
18	! Intercommunicator (comm_inter) and derived data type (type_xy) for
19	! coupled model runs created, assign coupling_mode_remote,
20	! indices nxlu and nysv are calculated (needed for non-cyclic boundary
21	! conditions)
22	!
23	! 82 2007-04-16 15:40:52Z raasch
24	! Cpp-directive lcmuk changed to intel_openmp_bug, setting of host on lcmuk by
25	! cpp-directive removed
26	!
27	! 75 2007-03-22 09:54:05Z raasch
28	! uxrp, vynp eliminated,
29	! dirichlet/neumann changed to dirichlet/radiation, etc.,
30	! poisfft_init is only called if fft-solver is switched on
31	!
32	! RCS Log replace by Id keyword, revision history cleaned up
33	!
34	! Revision 1.28 2006/04/26 13:23:32 raasch
35	! lcmuk does not understand the !$ comment so a cpp-directive is required
36	!
37	! Revision 1.1 1997/07/24 11:15:09 raasch
38	! Initial revision
39	!
40	!
41	! Description:
42	! ------------
43	! Determination of the virtual processor topology (if not prescribed by the
44	! user)and computation of the grid point number and array bounds of the local
45	! domains.
46	!------------------------------------------------------------------------------!
47
48	USE control_parameters
49	USE fft_xy
50	USE grid_variables
51	USE indices
52	USE pegrid
53	USE poisfft_mod
54	USE poisfft_hybrid_mod
55	USE statistics
56	USE transpose_indices
57
58
59	IMPLICIT NONE
60
61	INTEGER :: gathered_size, i, id_inflow_l, id_recycling_l, ind(5), j, k, &
62	maximum_grid_level_l, mg_switch_to_pe0_level_l, mg_levels_x, &
63	mg_levels_y, mg_levels_z, nnx_y, nnx_z, nny_x, nny_z, nnz_x, &
64	nnz_y, numproc_sqr, nx_total, nxl_l, nxr_l, nyn_l, nys_l, &
65	nzb_l, nzt_l, omp_get_num_threads, subdomain_size
66
67	INTEGER, DIMENSION(:), ALLOCATABLE :: ind_all, nxlf, nxrf, nynf, nysf
68
69	LOGICAL :: found
70
71	!
72	!-- Get the number of OpenMP threads
73	!$OMP PARALLEL
74	#if defined( __intel_openmp_bug )
75	threads_per_task = omp_get_num_threads()
76	#else
77	!$ threads_per_task = omp_get_num_threads()
78	#endif
79	!$OMP END PARALLEL
80
81
82	#if defined( __parallel )
83	!
84	!-- Determine the processor topology or check it, if prescribed by the user
85	IF ( npex == -1 .AND. npey == -1 ) THEN
86
87	!
88	!-- Automatic determination of the topology
89	!-- The default on SMP- and cluster-hosts is a 1d-decomposition along x
90	IF ( host(1:3) == 'ibm' .OR. host(1:3) == 'nec' .OR. &
91	host(1:2) == 'lc' .OR. host(1:3) == 'dec' ) THEN
92
93	pdims(1) = numprocs
94	pdims(2) = 1
95
96	ELSE
97
98	numproc_sqr = SQRT( REAL( numprocs ) )
99	pdims(1) = MAX( numproc_sqr , 1 )
100	DO WHILE ( MOD( numprocs , pdims(1) ) /= 0 )
101	pdims(1) = pdims(1) - 1
102	ENDDO
103	pdims(2) = numprocs / pdims(1)
104
105	ENDIF
106
107	ELSEIF ( npex /= -1 .AND. npey /= -1 ) THEN
108
109	!
110	!-- Prescribed by user. Number of processors on the prescribed topology
111	!-- must be equal to the number of PEs available to the job
112	IF ( ( npex * npey ) /= numprocs ) THEN
113	PRINT*, '+++ init_pegrid:'
114	PRINT, ' number of PEs of the prescribed topology (', npexnpey, &
115	') does not match the number of PEs available to the ', &
116	'job (', numprocs, ')'
117	CALL local_stop
118	ENDIF
119	pdims(1) = npex
120	pdims(2) = npey
121
122	ELSE
123	!
124	!-- If the processor topology is prescribed by the user, the number of
125	!-- PEs must be given in both directions
126	PRINT*, '+++ init_pegrid:'
127	PRINT*, ' if the processor topology is prescribed by the user, ', &
128	'both values of "npex" and "npey" must be given in the ', &
129	'NAMELIST-parameter file'
130	CALL local_stop
131
132	ENDIF
133
134	!
135	!-- The hybrid solver can only be used in case of a 1d-decomposition along x
136	IF ( pdims(2) /= 1 .AND. psolver == 'poisfft_hybrid' ) THEN
137	IF ( myid == 0 ) THEN
138	PRINT, '** init_pegrid: psolver = "poisfft_hybrid" can only be'
139	PRINT*, ' used in case of a 1d-decomposition along x'
140	ENDIF
141	ENDIF
142
143	!
144	!-- If necessary, set horizontal boundary conditions to non-cyclic
145	IF ( bc_lr /= 'cyclic' ) cyclic(1) = .FALSE.
146	IF ( bc_ns /= 'cyclic' ) cyclic(2) = .FALSE.
147
148	!
149	!-- Create the virtual processor grid
150	CALL MPI_CART_CREATE( comm_palm, ndim, pdims, cyclic, reorder, &
151	comm2d, ierr )
152	CALL MPI_COMM_RANK( comm2d, myid, ierr )
153	WRITE (myid_char,'(''_'',I4.4)') myid
154
155	CALL MPI_CART_COORDS( comm2d, myid, ndim, pcoord, ierr )
156	CALL MPI_CART_SHIFT( comm2d, 0, 1, pleft, pright, ierr )
157	CALL MPI_CART_SHIFT( comm2d, 1, 1, psouth, pnorth, ierr )
158
159	!
160	!-- Determine sub-topologies for transpositions
161	!-- Transposition from z to x:
162	remain_dims(1) = .TRUE.
163	remain_dims(2) = .FALSE.
164	CALL MPI_CART_SUB( comm2d, remain_dims, comm1dx, ierr )
165	CALL MPI_COMM_RANK( comm1dx, myidx, ierr )
166	!
167	!-- Transposition from x to y
168	remain_dims(1) = .FALSE.
169	remain_dims(2) = .TRUE.
170	CALL MPI_CART_SUB( comm2d, remain_dims, comm1dy, ierr )
171	CALL MPI_COMM_RANK( comm1dy, myidy, ierr )
172
173
174	!
175	!-- Find a grid (used for array d) which will match the transposition demands
176	IF ( grid_matching == 'strict' ) THEN
177
178	nxa = nx; nya = ny; nza = nz
179
180	ELSE
181
182	found = .FALSE.
183	xn: DO nxa = nx, 2*nx
184	!
185	!-- Meet conditions for nx
186	IF ( MOD( nxa+1, pdims(1) ) /= 0 .OR. &
187	MOD( nxa+1, pdims(2) ) /= 0 ) CYCLE xn
188
189	yn: DO nya = ny, 2*ny
190	!
191	!-- Meet conditions for ny
192	IF ( MOD( nya+1, pdims(2) ) /= 0 .OR. &
193	MOD( nya+1, pdims(1) ) /= 0 ) CYCLE yn
194
195
196	zn: DO nza = nz, 2*nz
197	!
198	!-- Meet conditions for nz
199	IF ( ( MOD( nza, pdims(1) ) /= 0 .AND. pdims(1) /= 1 .AND. &
200	pdims(2) /= 1 ) .OR. &
201	( MOD( nza, pdims(2) ) /= 0 .AND. dt_dosp /= 9999999.9 &
202	) ) THEN
203	CYCLE zn
204	ELSE
205	found = .TRUE.
206	EXIT xn
207	ENDIF
208
209	ENDDO zn
210
211	ENDDO yn
212
213	ENDDO xn
214
215	IF ( .NOT. found ) THEN
216	IF ( myid == 0 ) THEN
217	PRINT*,'+++ init_pegrid: no matching grid for transpositions found'
218	ENDIF
219	CALL local_stop
220	ENDIF
221
222	ENDIF
223
224	!
225	!-- Calculate array bounds in x-direction for every PE.
226	!-- The last PE along x may get less grid points than the others
227	ALLOCATE( nxlf(0:pdims(1)-1), nxrf(0:pdims(1)-1), nynf(0:pdims(2)-1), &
228	nysf(0:pdims(2)-1), nnx_pe(0:pdims(1)-1), nny_pe(0:pdims(2)-1) )
229
230	IF ( MOD( nxa+1 , pdims(1) ) /= 0 ) THEN
231	IF ( myid == 0 ) THEN
232	PRINT*,'+++ x-direction: gridpoint number (',nx+1,') is not an'
233	PRINT*,' integral divisor of the number of proces', &
234	&'sors (', pdims(1),')'
235	ENDIF
236	CALL local_stop
237	ELSE
238	nnx = ( nxa + 1 ) / pdims(1)
239	IF ( nnx*pdims(1) - ( nx + 1) > nnx ) THEN
240	IF ( myid == 0 ) THEN
241	PRINT*,'+++ x-direction: nx does not match the requirements ', &
242	'given by the number of PEs'
243	PRINT*,' used'
244	PRINT,' please use nx = ', nx - ( pdims(1) - ( nnxpdims(1) &
245	- ( nx + 1 ) ) ), ' instead of nx =', nx
246	ENDIF
247	CALL local_stop
248	ENDIF
249	ENDIF
250
251	!
252	!-- Left and right array bounds, number of gridpoints
253	DO i = 0, pdims(1)-1
254	nxlf(i) = i * nnx
255	nxrf(i) = ( i + 1 ) * nnx - 1
256	nnx_pe(i) = MIN( nx, nxrf(i) ) - nxlf(i) + 1
257	ENDDO
258
259	!
260	!-- Calculate array bounds in y-direction for every PE.
261	IF ( MOD( nya+1 , pdims(2) ) /= 0 ) THEN
262	IF ( myid == 0 ) THEN
263	PRINT*,'+++ y-direction: gridpoint number (',ny+1,') is not an'
264	PRINT*,' integral divisor of the number of proces', &
265	&'sors (', pdims(2),')'
266	ENDIF
267	CALL local_stop
268	ELSE
269	nny = ( nya + 1 ) / pdims(2)
270	IF ( nny*pdims(2) - ( ny + 1) > nny ) THEN
271	IF ( myid == 0 ) THEN
272	PRINT*,'+++ x-direction: nx does not match the requirements ', &
273	'given by the number of PEs'
274	PRINT*,' used'
275	PRINT,' please use nx = ', nx - ( pdims(1) - ( nnxpdims(1) &
276	- ( nx + 1 ) ) ), ' instead of nx =', nx
277	ENDIF
278	CALL local_stop
279	ENDIF
280	ENDIF
281
282	!
283	!-- South and north array bounds
284	DO j = 0, pdims(2)-1
285	nysf(j) = j * nny
286	nynf(j) = ( j + 1 ) * nny - 1
287	nny_pe(j) = MIN( ny, nynf(j) ) - nysf(j) + 1
288	ENDDO
289
290	!
291	!-- Local array bounds of the respective PEs
292	nxl = nxlf(pcoord(1))
293	nxra = nxrf(pcoord(1))
294	nxr = MIN( nx, nxra )
295	nys = nysf(pcoord(2))
296	nyna = nynf(pcoord(2))
297	nyn = MIN( ny, nyna )
298	nzb = 0
299	nzta = nza
300	nzt = MIN( nz, nzta )
301	nnz = nza
302
303	!
304	!-- Calculate array bounds and gridpoint numbers for the transposed arrays
305	!-- (needed in the pressure solver)
306	!-- For the transposed arrays, cyclic boundaries as well as top and bottom
307	!-- boundaries are omitted, because they are obstructive to the transposition
308
309	!
310	!-- 1. transposition z --> x
311	!-- This transposition is not neccessary in case of a 1d-decomposition along x,
312	!-- except that the uptream-spline method is switched on
313	IF ( pdims(2) /= 1 .OR. momentum_advec == 'ups-scheme' .OR. &
314	scalar_advec == 'ups-scheme' ) THEN
315
316	IF ( pdims(2) == 1 .AND. ( momentum_advec == 'ups-scheme' .OR. &
317	scalar_advec == 'ups-scheme' ) ) THEN
318	IF ( myid == 0 ) THEN
319	PRINT*,'+++ WARNING: init_pegrid: 1d-decomposition along x ', &
320	&'chosen but nz restrictions may occur'
321	PRINT*,' since ups-scheme is activated'
322	ENDIF
323	ENDIF
324	nys_x = nys
325	nyn_xa = nyna
326	nyn_x = nyn
327	nny_x = nny
328	IF ( MOD( nza , pdims(1) ) /= 0 ) THEN
329	IF ( myid == 0 ) THEN
330	PRINT*,'+++ transposition z --> x:'
331	PRINT*,' nz=',nz,' is not an integral divisior of pdims(1)=', &
332	&pdims(1)
333	ENDIF
334	CALL local_stop
335	ENDIF
336	nnz_x = nza / pdims(1)
337	nzb_x = 1 + myidx * nnz_x
338	nzt_xa = ( myidx + 1 ) * nnz_x
339	nzt_x = MIN( nzt, nzt_xa )
340
341	sendrecvcount_zx = nnx * nny * nnz_x
342
343	ENDIF
344
345	!
346	!-- 2. transposition x --> y
347	nnz_y = nnz_x
348	nzb_y = nzb_x
349	nzt_ya = nzt_xa
350	nzt_y = nzt_x
351	IF ( MOD( nxa+1 , pdims(2) ) /= 0 ) THEN
352	IF ( myid == 0 ) THEN
353	PRINT*,'+++ transposition x --> y:'
354	PRINT*,' nx+1=',nx+1,' is not an integral divisor of ',&
355	&'pdims(2)=',pdims(2)
356	ENDIF
357	CALL local_stop
358	ENDIF
359	nnx_y = (nxa+1) / pdims(2)
360	nxl_y = myidy * nnx_y
361	nxr_ya = ( myidy + 1 ) * nnx_y - 1
362	nxr_y = MIN( nx, nxr_ya )
363
364	sendrecvcount_xy = nnx_y * nny_x * nnz_y
365
366	!
367	!-- 3. transposition y --> z (ELSE: x --> y in case of 1D-decomposition
368	!-- along x)
369	IF ( pdims(2) /= 1 .OR. momentum_advec == 'ups-scheme' .OR. &
370	scalar_advec == 'ups-scheme' ) THEN
371	!
372	!-- y --> z
373	!-- This transposition is not neccessary in case of a 1d-decomposition
374	!-- along x, except that the uptream-spline method is switched on
375	nnx_z = nnx_y
376	nxl_z = nxl_y
377	nxr_za = nxr_ya
378	nxr_z = nxr_y
379	IF ( MOD( nya+1 , pdims(1) ) /= 0 ) THEN
380	IF ( myid == 0 ) THEN
381	PRINT*,'+++ Transposition y --> z:'
382	PRINT*,' ny+1=',ny+1,' is not an integral divisor of ',&
383	&'pdims(1)=',pdims(1)
384	ENDIF
385	CALL local_stop
386	ENDIF
387	nny_z = (nya+1) / pdims(1)
388	nys_z = myidx * nny_z
389	nyn_za = ( myidx + 1 ) * nny_z - 1
390	nyn_z = MIN( ny, nyn_za )
391
392	sendrecvcount_yz = nnx_y * nny_z * nnz_y
393
394	ELSE
395	!
396	!-- x --> y. This condition must be fulfilled for a 1D-decomposition along x
397	IF ( MOD( nya+1 , pdims(1) ) /= 0 ) THEN
398	IF ( myid == 0 ) THEN
399	PRINT*,'+++ Transposition x --> y:'
400	PRINT*,' ny+1=',ny+1,' is not an integral divisor of ',&
401	&'pdims(1)=',pdims(1)
402	ENDIF
403	CALL local_stop
404	ENDIF
405
406	ENDIF
407
408	!
409	!-- Indices for direct transpositions z --> y (used for calculating spectra)
410	IF ( dt_dosp /= 9999999.9 ) THEN
411	IF ( MOD( nza, pdims(2) ) /= 0 ) THEN
412	IF ( myid == 0 ) THEN
413	PRINT*,'+++ Direct transposition z --> y (needed for spectra):'
414	PRINT*,' nz=',nz,' is not an integral divisor of ',&
415	&'pdims(2)=',pdims(2)
416	ENDIF
417	CALL local_stop
418	ELSE
419	nxl_yd = nxl
420	nxr_yda = nxra
421	nxr_yd = nxr
422	nzb_yd = 1 + myidy * ( nza / pdims(2) )
423	nzt_yda = ( myidy + 1 ) * ( nza / pdims(2) )
424	nzt_yd = MIN( nzt, nzt_yda )
425
426	sendrecvcount_zyd = nnx * nny * ( nza / pdims(2) )
427	ENDIF
428	ENDIF
429
430	!
431	!-- Indices for direct transpositions y --> x (they are only possible in case
432	!-- of a 1d-decomposition along x)
433	IF ( pdims(2) == 1 ) THEN
434	nny_x = nny / pdims(1)
435	nys_x = myid * nny_x
436	nyn_xa = ( myid + 1 ) * nny_x - 1
437	nyn_x = MIN( ny, nyn_xa )
438	nzb_x = 1
439	nzt_xa = nza
440	nzt_x = nz
441	sendrecvcount_xy = nnx * nny_x * nza
442	ENDIF
443
444	!
445	!-- Indices for direct transpositions x --> y (they are only possible in case
446	!-- of a 1d-decomposition along y)
447	IF ( pdims(1) == 1 ) THEN
448	nnx_y = nnx / pdims(2)
449	nxl_y = myid * nnx_y
450	nxr_ya = ( myid + 1 ) * nnx_y - 1
451	nxr_y = MIN( nx, nxr_ya )
452	nzb_y = 1
453	nzt_ya = nza
454	nzt_y = nz
455	sendrecvcount_xy = nnx_y * nny * nza
456	ENDIF
457
458	!
459	!-- Arrays for storing the array bounds are needed any more
460	DEALLOCATE( nxlf , nxrf , nynf , nysf )
461
462	!
463	!-- Collect index bounds from other PEs (to be written to restart file later)
464	ALLOCATE( hor_index_bounds(4,0:numprocs-1) )
465
466	IF ( myid == 0 ) THEN
467
468	hor_index_bounds(1,0) = nxl
469	hor_index_bounds(2,0) = nxr
470	hor_index_bounds(3,0) = nys
471	hor_index_bounds(4,0) = nyn
472
473	!
474	!-- Receive data from all other PEs
475	DO i = 1, numprocs-1
476	CALL MPI_RECV( ibuf, 4, MPI_INTEGER, i, MPI_ANY_TAG, comm2d, status, &
477	ierr )
478	hor_index_bounds(:,i) = ibuf(1:4)
479	ENDDO
480
481	ELSE
482	!
483	!-- Send index bounds to PE0
484	ibuf(1) = nxl
485	ibuf(2) = nxr
486	ibuf(3) = nys
487	ibuf(4) = nyn
488	CALL MPI_SEND( ibuf, 4, MPI_INTEGER, 0, myid, comm2d, ierr )
489
490	ENDIF
491
492	#if defined( __print )
493	!
494	!-- Control output
495	IF ( myid == 0 ) THEN
496	PRINT, ' processor topology *'
497	PRINT*, ' '
498	PRINT*, 'myid pcoord left right south north idx idy nxl: nxr',&
499	&' nys: nyn'
500	PRINT*, '------------------------------------------------------------',&
501	&'-----------'
502	WRITE (*,1000) 0, pcoord(1), pcoord(2), pleft, pright, psouth, pnorth, &
503	myidx, myidy, nxl, nxr, nys, nyn
504	1000 FORMAT (I4,2X,'(',I3,',',I3,')',3X,I4,2X,I4,3X,I4,2X,I4,2X,I3,1X,I3, &
505	2(2X,I4,':',I4))
506
507	!
508	!-- Receive data from the other PEs
509	DO i = 1,numprocs-1
510	CALL MPI_RECV( ibuf, 12, MPI_INTEGER, i, MPI_ANY_TAG, comm2d, status, &
511	ierr )
512	WRITE (*,1000) i, ( ibuf(j) , j = 1,12 )
513	ENDDO
514	ELSE
515
516	!
517	!-- Send data to PE0
518	ibuf(1) = pcoord(1); ibuf(2) = pcoord(2); ibuf(3) = pleft
519	ibuf(4) = pright; ibuf(5) = psouth; ibuf(6) = pnorth; ibuf(7) = myidx
520	ibuf(8) = myidy; ibuf(9) = nxl; ibuf(10) = nxr; ibuf(11) = nys
521	ibuf(12) = nyn
522	CALL MPI_SEND( ibuf, 12, MPI_INTEGER, 0, myid, comm2d, ierr )
523	ENDIF
524	#endif
525
526	#if defined( __mpi2 )
527	!
528	!-- In case of coupled runs, get the port name on PE0 of the atmosphere model
529	!-- and pass it to PE0 of the ocean model
530	IF ( myid == 0 ) THEN
531
532	IF ( coupling_mode == 'atmosphere_to_ocean' ) THEN
533
534	CALL MPI_OPEN_PORT( MPI_INFO_NULL, port_name, ierr )
535	!
536	!-- TEST OUTPUT (TO BE REMOVED)
537	WRITE(9,*) TRIM( coupling_mode ), &
538	', ierr after MPI_OPEN_PORT: ', ierr
539	CALL LOCAL_FLUSH( 9 )
540
541	CALL MPI_PUBLISH_NAME( 'palm_coupler', MPI_INFO_NULL, port_name, &
542	ierr )
543	!
544	!-- TEST OUTPUT (TO BE REMOVED)
545	WRITE(9,*) TRIM( coupling_mode ), &
546	', ierr after MPI_PUBLISH_NAME: ', ierr
547	CALL LOCAL_FLUSH( 9 )
548
549	!
550	!-- Write a flag file for the ocean model and the other atmosphere
551	!-- processes.
552	!-- There seems to be a bug in MPICH2 which causes hanging processes
553	!-- in case that execution of LOOKUP_NAME is continued too early
554	!-- (i.e. before the port has been created)
555	OPEN( 90, FILE='COUPLING_PORT_OPENED', FORM='FORMATTED' )
556	WRITE ( 90, '(''TRUE'')' )
557	CLOSE ( 90 )
558
559	ELSEIF ( coupling_mode == 'ocean_to_atmosphere' ) THEN
560
561	!
562	!-- Continue only if the atmosphere model has created the port.
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	INQUIRE( FILE='COUPLING_PORT_OPENED', EXIST=found )
567	DO WHILE ( .NOT. found )
568	INQUIRE( FILE='COUPLING_PORT_OPENED', EXIST=found )
569	ENDDO
570
571	CALL MPI_LOOKUP_NAME( 'palm_coupler', MPI_INFO_NULL, port_name, ierr )
572	!
573	!-- TEST OUTPUT (TO BE REMOVED)
574	WRITE(9,*) TRIM( coupling_mode ), &
575	', ierr after MPI_LOOKUP_NAME: ', ierr
576	CALL LOCAL_FLUSH( 9 )
577
578
579	ENDIF
580
581	ENDIF
582
583	!
584	!-- In case of coupled runs, establish the connection between the atmosphere
585	!-- and the ocean model and define the intercommunicator (comm_inter)
586	CALL MPI_BARRIER( comm2d, ierr )
587	IF ( coupling_mode == 'atmosphere_to_ocean' ) THEN
588
589	print*, '... before COMM_ACCEPT'
590	CALL MPI_COMM_ACCEPT( port_name, MPI_INFO_NULL, 0, MPI_COMM_WORLD, &
591	comm_inter, ierr )
592	print*, '--- ierr = ', ierr
593	print*, '--- comm_inter atmosphere = ', comm_inter
594
595	coupling_mode_remote = 'ocean_to_atmosphere'
596
597	ELSEIF ( coupling_mode == 'ocean_to_atmosphere' ) THEN
598
599	IF ( myid == 0 ) PRINT, '** read: ', port_name, ' ierr = ', ierr
600	print*, '... before COMM_CONNECT'
601	CALL MPI_COMM_CONNECT( port_name, MPI_INFO_NULL, 0, MPI_COMM_WORLD, &
602	comm_inter, ierr )
603	print*, '--- ierr = ', ierr
604	print*, '--- comm_inter ocean = ', comm_inter
605
606	coupling_mode_remote = 'atmosphere_to_ocean'
607
608	ENDIF
609
610	!
611	!-- In case of coupled runs, create a new MPI derived datatype for the
612	!-- exchange of surface (xy) data .
613	!-- Gridpoint number for the exchange of ghost points (xy-plane)
614	ngp_xy = ( nxr - nxl + 3 ) * ( nyn - nys + 3 )
615
616	!
617	!-- Define a new MPI derived datatype for the exchange of ghost points in
618	!-- y-direction for 2D-arrays (line)
619	CALL MPI_TYPE_VECTOR( ngp_xy, 1, nzt-nzb+2, MPI_REAL, type_xy, ierr )
620	CALL MPI_TYPE_COMMIT( type_xy, ierr )
621	#endif
622
623	#else
624
625	!
626	!-- Array bounds when running on a single PE (respectively a non-parallel
627	!-- machine)
628	nxl = 0
629	nxr = nx
630	nxra = nx
631	nnx = nxr - nxl + 1
632	nys = 0
633	nyn = ny
634	nyna = ny
635	nny = nyn - nys + 1
636	nzb = 0
637	nzt = nz
638	nzta = nz
639	nnz = nz
640
641	ALLOCATE( hor_index_bounds(4,0:0) )
642	hor_index_bounds(1,0) = nxl
643	hor_index_bounds(2,0) = nxr
644	hor_index_bounds(3,0) = nys
645	hor_index_bounds(4,0) = nyn
646
647	!
648	!-- Array bounds for the pressure solver (in the parallel code, these bounds
649	!-- are the ones for the transposed arrays)
650	nys_x = nys
651	nyn_x = nyn
652	nyn_xa = nyn
653	nzb_x = nzb + 1
654	nzt_x = nzt
655	nzt_xa = nzt
656
657	nxl_y = nxl
658	nxr_y = nxr
659	nxr_ya = nxr
660	nzb_y = nzb + 1
661	nzt_y = nzt
662	nzt_ya = nzt
663
664	nxl_z = nxl
665	nxr_z = nxr
666	nxr_za = nxr
667	nys_z = nys
668	nyn_z = nyn
669	nyn_za = nyn
670
671	#endif
672
673	!
674	!-- Calculate number of grid levels necessary for the multigrid poisson solver
675	!-- as well as the gridpoint indices on each level
676	IF ( psolver == 'multigrid' ) THEN
677
678	!
679	!-- First calculate number of possible grid levels for the subdomains
680	mg_levels_x = 1
681	mg_levels_y = 1
682	mg_levels_z = 1
683
684	i = nnx
685	DO WHILE ( MOD( i, 2 ) == 0 .AND. i /= 2 )
686	i = i / 2
687	mg_levels_x = mg_levels_x + 1
688	ENDDO
689
690	j = nny
691	DO WHILE ( MOD( j, 2 ) == 0 .AND. j /= 2 )
692	j = j / 2
693	mg_levels_y = mg_levels_y + 1
694	ENDDO
695
696	k = nnz
697	DO WHILE ( MOD( k, 2 ) == 0 .AND. k /= 2 )
698	k = k / 2
699	mg_levels_z = mg_levels_z + 1
700	ENDDO
701
702	maximum_grid_level = MIN( mg_levels_x, mg_levels_y, mg_levels_z )
703
704	!
705	!-- Find out, if the total domain allows more levels. These additional
706	!-- levels are processed on PE0 only.
707	IF ( numprocs > 1 ) THEN
708	IF ( mg_levels_z > MIN( mg_levels_x, mg_levels_y ) ) THEN
709	mg_switch_to_pe0_level_l = maximum_grid_level
710
711	mg_levels_x = 1
712	mg_levels_y = 1
713
714	i = nx+1
715	DO WHILE ( MOD( i, 2 ) == 0 .AND. i /= 2 )
716	i = i / 2
717	mg_levels_x = mg_levels_x + 1
718	ENDDO
719
720	j = ny+1
721	DO WHILE ( MOD( j, 2 ) == 0 .AND. j /= 2 )
722	j = j / 2
723	mg_levels_y = mg_levels_y + 1
724	ENDDO
725
726	maximum_grid_level_l = MIN( mg_levels_x, mg_levels_y, mg_levels_z )
727
728	IF ( maximum_grid_level_l > mg_switch_to_pe0_level_l ) THEN
729	mg_switch_to_pe0_level_l = maximum_grid_level_l - &
730	mg_switch_to_pe0_level_l + 1
731	ELSE
732	mg_switch_to_pe0_level_l = 0
733	ENDIF
734	ELSE
735	mg_switch_to_pe0_level_l = 0
736	maximum_grid_level_l = maximum_grid_level
737	ENDIF
738
739	!
740	!-- Use switch level calculated above only if it is not pre-defined
741	!-- by user
742	IF ( mg_switch_to_pe0_level == 0 ) THEN
743
744	IF ( mg_switch_to_pe0_level_l /= 0 ) THEN
745	mg_switch_to_pe0_level = mg_switch_to_pe0_level_l
746	maximum_grid_level = maximum_grid_level_l
747	ENDIF
748
749	ELSE
750	!
751	!-- Check pre-defined value and reset to default, if neccessary
752	IF ( mg_switch_to_pe0_level < mg_switch_to_pe0_level_l .OR. &
753	mg_switch_to_pe0_level >= maximum_grid_level_l ) THEN
754	IF ( myid == 0 ) THEN
755	PRINT*, '+++ WARNING init_pegrid: mg_switch_to_pe0_level ', &
756	'out of range and reset to default (=0)'
757	ENDIF
758	mg_switch_to_pe0_level = 0
759	ELSE
760	!
761	!-- Use the largest number of possible levels anyway and recalculate
762	!-- the switch level to this largest number of possible values
763	maximum_grid_level = maximum_grid_level_l
764
765	ENDIF
766	ENDIF
767
768	ENDIF
769
770	ALLOCATE( grid_level_count(maximum_grid_level), &
771	nxl_mg(maximum_grid_level), nxr_mg(maximum_grid_level), &
772	nyn_mg(maximum_grid_level), nys_mg(maximum_grid_level), &
773	nzt_mg(maximum_grid_level) )
774
775	grid_level_count = 0
776	nxl_l = nxl; nxr_l = nxr; nys_l = nys; nyn_l = nyn; nzt_l = nzt
777
778	DO i = maximum_grid_level, 1 , -1
779
780	IF ( i == mg_switch_to_pe0_level ) THEN
781	#if defined( __parallel )
782	!
783	!-- Save the grid size of the subdomain at the switch level, because
784	!-- it is needed in poismg.
785	!-- Array bounds of the local subdomain grids are gathered on PE0
786	ind(1) = nxl_l; ind(2) = nxr_l
787	ind(3) = nys_l; ind(4) = nyn_l
788	ind(5) = nzt_l
789	ALLOCATE( ind_all(5*numprocs), mg_loc_ind(5,0:numprocs-1) )
790	CALL MPI_ALLGATHER( ind, 5, MPI_INTEGER, ind_all, 5, &
791	MPI_INTEGER, comm2d, ierr )
792	DO j = 0, numprocs-1
793	DO k = 1, 5
794	mg_loc_ind(k,j) = ind_all(k+j*5)
795	ENDDO
796	ENDDO
797	DEALLOCATE( ind_all )
798	!
799	!-- Calculate the grid size of the total domain gathered on PE0
800	nxr_l = ( nxr_l-nxl_l+1 ) * pdims(1) - 1
801	nxl_l = 0
802	nyn_l = ( nyn_l-nys_l+1 ) * pdims(2) - 1
803	nys_l = 0
804	!
805	!-- The size of this gathered array must not be larger than the
806	!-- array tend, which is used in the multigrid scheme as a temporary
807	!-- array
808	subdomain_size = ( nxr - nxl + 3 ) * ( nyn - nys + 3 ) * &
809	( nzt - nzb + 2 )
810	gathered_size = ( nxr_l - nxl_l + 3 ) * ( nyn_l - nys_l + 3 ) * &
811	( nzt_l - nzb + 2 )
812
813	IF ( gathered_size > subdomain_size ) THEN
814	IF ( myid == 0 ) THEN
815	PRINT*, '+++ init_pegrid: not enough memory for storing ', &
816	'gathered multigrid data on PE0'
817	ENDIF
818	CALL local_stop
819	ENDIF
820	#else
821	PRINT*, '+++ init_pegrid: multigrid gather/scatter impossible ', &
822	'in non parallel mode'
823	CALL local_stop
824	#endif
825	ENDIF
826
827	nxl_mg(i) = nxl_l
828	nxr_mg(i) = nxr_l
829	nys_mg(i) = nys_l
830	nyn_mg(i) = nyn_l
831	nzt_mg(i) = nzt_l
832
833	nxl_l = nxl_l / 2
834	nxr_l = nxr_l / 2
835	nys_l = nys_l / 2
836	nyn_l = nyn_l / 2
837	nzt_l = nzt_l / 2
838	ENDDO
839
840	ELSE
841
842	maximum_grid_level = 1
843
844	ENDIF
845
846	grid_level = maximum_grid_level
847
848	#if defined( __parallel )
849	!
850	!-- Gridpoint number for the exchange of ghost points (y-line for 2D-arrays)
851	ngp_y = nyn - nys + 1
852
853	!
854	!-- Define a new MPI derived datatype for the exchange of ghost points in
855	!-- y-direction for 2D-arrays (line)
856	CALL MPI_TYPE_VECTOR( nxr-nxl+3, 1, ngp_y+2, MPI_REAL, type_x, ierr )
857	CALL MPI_TYPE_COMMIT( type_x, ierr )
858	CALL MPI_TYPE_VECTOR( nxr-nxl+3, 1, ngp_y+2, MPI_INTEGER, type_x_int, ierr )
859	CALL MPI_TYPE_COMMIT( type_x_int, ierr )
860
861	!
862	!-- Calculate gridpoint numbers for the exchange of ghost points along x
863	!-- (yz-plane for 3D-arrays) and define MPI derived data type(s) for the
864	!-- exchange of ghost points in y-direction (xz-plane).
865	!-- Do these calculations for the model grid and (if necessary) also
866	!-- for the coarser grid levels used in the multigrid method
867	ALLOCATE ( ngp_yz(maximum_grid_level), type_xz(maximum_grid_level) )
868
869	nxl_l = nxl; nxr_l = nxr; nys_l = nys; nyn_l = nyn; nzb_l = nzb; nzt_l = nzt
870
871	DO i = maximum_grid_level, 1 , -1
872	ngp_yz(i) = (nzt_l - nzb_l + 2) * (nyn_l - nys_l + 3)
873
874	CALL MPI_TYPE_VECTOR( nxr_l-nxl_l+3, nzt_l-nzb_l+2, ngp_yz(i), &
875	MPI_REAL, type_xz(i), ierr )
876	CALL MPI_TYPE_COMMIT( type_xz(i), ierr )
877
878	nxl_l = nxl_l / 2
879	nxr_l = nxr_l / 2
880	nys_l = nys_l / 2
881	nyn_l = nyn_l / 2
882	nzt_l = nzt_l / 2
883	ENDDO
884	#endif
885
886	#if defined( __parallel )
887	!
888	!-- Setting of flags for inflow/outflow conditions in case of non-cyclic
889	!-- horizontal boundary conditions.
890	IF ( pleft == MPI_PROC_NULL ) THEN
891	IF ( bc_lr == 'dirichlet/radiation' ) THEN
892	inflow_l = .TRUE.
893	ELSEIF ( bc_lr == 'radiation/dirichlet' ) THEN
894	outflow_l = .TRUE.
895	ENDIF
896	ENDIF
897
898	IF ( pright == MPI_PROC_NULL ) THEN
899	IF ( bc_lr == 'dirichlet/radiation' ) THEN
900	outflow_r = .TRUE.
901	ELSEIF ( bc_lr == 'radiation/dirichlet' ) THEN
902	inflow_r = .TRUE.
903	ENDIF
904	ENDIF
905
906	IF ( psouth == MPI_PROC_NULL ) THEN
907	IF ( bc_ns == 'dirichlet/radiation' ) THEN
908	outflow_s = .TRUE.
909	ELSEIF ( bc_ns == 'radiation/dirichlet' ) THEN
910	inflow_s = .TRUE.
911	ENDIF
912	ENDIF
913
914	IF ( pnorth == MPI_PROC_NULL ) THEN
915	IF ( bc_ns == 'dirichlet/radiation' ) THEN
916	inflow_n = .TRUE.
917	ELSEIF ( bc_ns == 'radiation/dirichlet' ) THEN
918	outflow_n = .TRUE.
919	ENDIF
920	ENDIF
921
922	!
923	!-- Broadcast the id of the inflow PE
924	IF ( inflow_l ) THEN
925	id_inflow_l = myidx
926	ELSE
927	id_inflow_l = 0
928	ENDIF
929	CALL MPI_ALLREDUCE( id_inflow_l, id_inflow, 1, MPI_INTEGER, MPI_SUM, &
930	comm1dx, ierr )
931
932	!
933	!-- Broadcast the id of the recycling plane
934	!-- WARNING: needs to be adjusted in case of inflows other than from left side!
935	IF ( ( recycling_width / dx ) >= nxl .AND. ( recycling_width / dx ) <= nxr ) &
936	THEN
937	id_recycling_l = myidx
938	ELSE
939	id_recycling_l = 0
940	ENDIF
941	CALL MPI_ALLREDUCE( id_recycling_l, id_recycling, 1, MPI_INTEGER, MPI_SUM, &
942	comm1dx, ierr )
943
944	#else
945	IF ( bc_lr == 'dirichlet/radiation' ) THEN
946	inflow_l = .TRUE.
947	outflow_r = .TRUE.
948	ELSEIF ( bc_lr == 'radiation/dirichlet' ) THEN
949	outflow_l = .TRUE.
950	inflow_r = .TRUE.
951	ENDIF
952
953	IF ( bc_ns == 'dirichlet/radiation' ) THEN
954	inflow_n = .TRUE.
955	outflow_s = .TRUE.
956	ELSEIF ( bc_ns == 'radiation/dirichlet' ) THEN
957	outflow_n = .TRUE.
958	inflow_s = .TRUE.
959	ENDIF
960	#endif
961	!
962	!-- At the outflow, u or v, respectively, have to be calculated for one more
963	!-- grid point.
964	IF ( outflow_l ) THEN
965	nxlu = nxl + 1
966	ELSE
967	nxlu = nxl
968	ENDIF
969	IF ( outflow_s ) THEN
970	nysv = nys + 1
971	ELSE
972	nysv = nys
973	ENDIF
974
975	IF ( psolver == 'poisfft_hybrid' ) THEN
976	CALL poisfft_hybrid_ini
977	ELSEIF ( psolver == 'poisfft' ) THEN
978	CALL poisfft_init
979	ENDIF
980
981	!
982	!-- Allocate wall flag arrays used in the multigrid solver
983	IF ( psolver == 'multigrid' ) THEN
984
985	DO i = maximum_grid_level, 1, -1
986
987	SELECT CASE ( i )
988
989	CASE ( 1 )
990	ALLOCATE( wall_flags_1(nzb:nzt_mg(i)+1, &
991	nys_mg(i)-1:nyn_mg(i)+1, &
992	nxl_mg(i)-1:nxr_mg(i)+1) )
993
994	CASE ( 2 )
995	ALLOCATE( wall_flags_2(nzb:nzt_mg(i)+1, &
996	nys_mg(i)-1:nyn_mg(i)+1, &
997	nxl_mg(i)-1:nxr_mg(i)+1) )
998
999	CASE ( 3 )
1000	ALLOCATE( wall_flags_3(nzb:nzt_mg(i)+1, &
1001	nys_mg(i)-1:nyn_mg(i)+1, &
1002	nxl_mg(i)-1:nxr_mg(i)+1) )
1003
1004	CASE ( 4 )
1005	ALLOCATE( wall_flags_4(nzb:nzt_mg(i)+1, &
1006	nys_mg(i)-1:nyn_mg(i)+1, &
1007	nxl_mg(i)-1:nxr_mg(i)+1) )
1008
1009	CASE ( 5 )
1010	ALLOCATE( wall_flags_5(nzb:nzt_mg(i)+1, &
1011	nys_mg(i)-1:nyn_mg(i)+1, &
1012	nxl_mg(i)-1:nxr_mg(i)+1) )
1013
1014	CASE ( 6 )
1015	ALLOCATE( wall_flags_6(nzb:nzt_mg(i)+1, &
1016	nys_mg(i)-1:nyn_mg(i)+1, &
1017	nxl_mg(i)-1:nxr_mg(i)+1) )
1018
1019	CASE ( 7 )
1020	ALLOCATE( wall_flags_7(nzb:nzt_mg(i)+1, &
1021	nys_mg(i)-1:nyn_mg(i)+1, &
1022	nxl_mg(i)-1:nxr_mg(i)+1) )
1023
1024	CASE ( 8 )
1025	ALLOCATE( wall_flags_8(nzb:nzt_mg(i)+1, &
1026	nys_mg(i)-1:nyn_mg(i)+1, &
1027	nxl_mg(i)-1:nxr_mg(i)+1) )
1028
1029	CASE ( 9 )
1030	ALLOCATE( wall_flags_9(nzb:nzt_mg(i)+1, &
1031	nys_mg(i)-1:nyn_mg(i)+1, &
1032	nxl_mg(i)-1:nxr_mg(i)+1) )
1033
1034	CASE ( 10 )
1035	ALLOCATE( wall_flags_10(nzb:nzt_mg(i)+1, &
1036	nys_mg(i)-1:nyn_mg(i)+1, &
1037	nxl_mg(i)-1:nxr_mg(i)+1) )
1038
1039	CASE DEFAULT
1040	IF ( myid == 0 ) PRINT*, '+++ init_pegrid: more than 10 ', &
1041	' multigrid levels'
1042	CALL local_stop
1043
1044	END SELECT
1045
1046	ENDDO
1047
1048	ENDIF
1049
1050	END SUBROUTINE init_pegrid

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

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |