Home

Context Navigation

exchange_horiz_2d.f90 @ 707

Last change on this file since 707 was 707, checked in by raasch, 13 years ago

New:
---

In case of multigrid method, on coarse grid levels, gathered data are
identically processed on all PEs (before, on PE0 only), so that the subsequent
scattering of data is not neccessary any more. (modules, init_pegrid, poismg)

Changed:

Calculation of weighted average of p is now handled in the same way
regardless of the number of ghost layers (advection scheme). (pres)

multigrid and sor method are using p_loc for iterative
advancements of pressure. p_sub removed. (init_3d_model, modules, poismg, pres, sor)

bc_lr and bc_ns replaced by bc_lr_dirrad, bc_lr_raddir, bc_ns_dirrad, bc_ns_raddir
for speed optimization. (calc_spectra, check_parameters, exchange_horiz,
exchange_horiz_2d, header, init_3d_model, init_grid, init_pegrid, modules,
poismg, pres, sor, time_integration, timestep)

grid_level directly used as index for MPI data type arrays. (exchange_horiz,
poismg)

initial assignments of zero to array p for iterative solvers only (init_3d_model)

Errors:

localsum calculation modified for proper OpenMP reduction. (pres)

Bugfix: bottom (nzb) and top (nzt+1) boundary conditions set in routines
resid and restrict. They were missed before, which may have led to
unpredictable results. (poismg)

Property svn:keywords set to Id

File size: 6.7 KB

Line
1	SUBROUTINE exchange_horiz_2d( ar )
2
3	!------------------------------------------------------------------------------!
4	! Current revisions:
5	! -----------------
6	! bc_lr/ns replaced by bc_lr/ns_cyc
7	!
8	! Former revisions:
9	! -----------------
10	! $Id: exchange_horiz_2d.f90 707 2011-03-29 11:39:40Z raasch $
11	!
12	! 702 2011-03-24 19:33:15Z suehring
13	! Bugfix in declaration of ar in exchange_horiz_2d_int and number of MPI-blocks
14	! in MPI_SENDRECV().
15	!
16	! 667 2010-12-23 12:06:00Z suehring/gryschka
17	! Dynamic exchange of ghost points with nbgp, which depends on the advection
18	! scheme. Exchange between left and right PEs is now done with MPI-vectors.
19	!
20	! 73 2007-03-20 08:33:14Z raasch
21	! Neumann boundary conditions at inflow/outflow in case of non-cyclic boundary
22	! conditions
23	!
24	! RCS Log replace by Id keyword, revision history cleaned up
25	!
26	! Revision 1.9 2006/05/12 19:15:52 letzel
27	! MPI_REAL replaced by MPI_INTEGER in exchange_horiz_2d_int
28	!
29	! Revision 1.1 1998/01/23 09:58:21 raasch
30	! Initial revision
31	!
32	!
33	! Description:
34	! ------------
35	! Exchange of lateral (ghost) boundaries (parallel computers) and cyclic
36	! boundary conditions, respectively, for 2D-arrays.
37	!------------------------------------------------------------------------------!
38
39	USE control_parameters
40	USE cpulog
41	USE indices
42	USE interfaces
43	USE pegrid
44
45	IMPLICIT NONE
46
47	REAL :: ar(nysg:nyng,nxlg:nxrg)
48	INTEGER :: i
49
50
51	CALL cpu_log( log_point_s(13), 'exchange_horiz_2d', 'start' )
52
53	#if defined( __parallel )
54
55	!
56	!-- Exchange of lateral boundary values for parallel computers
57	IF ( pdims(1) == 1 ) THEN
58
59	!
60	!-- One-dimensional decomposition along y, boundary values can be exchanged
61	!-- within the PE memory
62	ar(:,nxlg:nxl-1) = ar(:,nxr-nbgp+1:nxr)
63	ar(:,nxr+1:nxrg) = ar(:,nxl:nxl+nbgp-1)
64
65	ELSE
66	!
67	!-- Send left boundary, receive right one
68
69	CALL MPI_SENDRECV( ar(nysg,nxl), 1, type_y, pleft, 0, &
70	ar(nysg,nxr+1), 1, type_y, pright, 0, &
71	comm2d, status, ierr )
72	!
73	!-- Send right boundary, receive left one
74	CALL MPI_SENDRECV( ar(nysg,nxr+1-nbgp), 1, type_y, pright, 1, &
75	ar(nysg,nxlg), 1, type_y, pleft, 1, &
76	comm2d, status, ierr )
77
78
79	ENDIF
80
81	IF ( pdims(2) == 1 ) THEN
82	!
83	!-- One-dimensional decomposition along x, boundary values can be exchanged
84	!-- within the PE memory
85	ar(nysg:nys-1,:) = ar(nyn-nbgp+1:nyn,:)
86	ar(nyn+1:nyng,:) = ar(nys:nys+nbgp-1,:)
87
88	ELSE
89	!
90	!-- Send front boundary, receive rear one
91
92	CALL MPI_SENDRECV( ar(nys,nxlg), 1, type_x, psouth, 0, &
93	ar(nyn+1,nxlg), 1, type_x, pnorth, 0, &
94	comm2d, status, ierr )
95	!
96	!-- Send rear boundary, receive front one
97	CALL MPI_SENDRECV( ar(nyn+1-nbgp,nxlg), 1, type_x, pnorth, 1, &
98	ar(nysg,nxlg), 1, type_x, psouth, 1, &
99	comm2d, status, ierr )
100
101	ENDIF
102
103	#else
104
105	!
106	!-- Lateral boundary conditions in the non-parallel case
107	IF ( bc_lr_cyc ) THEN
108	ar(:,nxlg:nxl-1) = ar(:,nxr-nbgp+1:nxr)
109	ar(:,nxr+1:nxrg) = ar(:,nxl:nxl+nbgp-1)
110	ENDIF
111
112	IF ( bc_ns_cyc ) THEN
113	ar(nysg:nys-1,:) = ar(nyn-nbgp+1:nyn,:)
114	ar(nyn+1:nyng,:) = ar(nys:nys+nbgp-1,:)
115	ENDIF
116
117
118	#endif
119
120	!
121	!-- Neumann-conditions at inflow/outflow in case of non-cyclic boundary
122	!-- conditions
123	IF ( inflow_l .OR. outflow_l ) THEN
124	DO i=nbgp, 1, -1
125	ar(:,nxl-i) = ar(:,nxl)
126	END DO
127	END IF
128	IF ( inflow_r .OR. outflow_r ) THEN
129	DO i=1, nbgp
130	ar(:,nxr+i) = ar(:,nxr)
131	END DO
132	END IF
133	IF ( inflow_s .OR. outflow_s ) THEN
134	DO i=nbgp, 1, -1
135	ar(nys-i,:) = ar(nys,:)
136	END DO
137	END IF
138	IF ( inflow_n .OR. outflow_n ) THEN
139	DO i=1, nbgp
140	ar(nyn+i,:) = ar(nyn,:)
141	END DO
142	END IF
143	CALL cpu_log( log_point_s(13), 'exchange_horiz_2d', 'stop' )
144
145	END SUBROUTINE exchange_horiz_2d
146
147
148
149	SUBROUTINE exchange_horiz_2d_int( ar )
150
151	!------------------------------------------------------------------------------!
152	! Description:
153	! ------------
154	! Exchange of lateral (ghost) boundaries (parallel computers) and cyclic
155	! boundary conditions, respectively, for 2D integer arrays.
156	!------------------------------------------------------------------------------!
157
158	USE control_parameters
159	USE cpulog
160	USE indices
161	USE interfaces
162	USE pegrid
163
164	IMPLICIT NONE
165
166	INTEGER :: ar(nysg:nyng,nxlg:nxrg)
167	INTEGER :: i
168
169	CALL cpu_log( log_point_s(13), 'exchange_horiz_2d', 'start' )
170
171	#if defined( __parallel )
172
173	!
174	!-- Exchange of lateral boundary values for parallel computers
175	IF ( pdims(1) == 1 ) THEN
176
177	!
178	!-- One-dimensional decomposition along y, boundary values can be exchanged
179	!-- within the PE memory
180	ar(:,nxlg:nxl-1) = ar(:,nxr-nbgp+1:nxr)
181	ar(:,nxr+1:nxrg) = ar(:,nxl:nxl+nbgp-1)
182
183
184	ELSE
185	!
186	!-- Send left boundary, receive right one
187	CALL MPI_SENDRECV( ar(nysg,nxl), 1, type_y_int, pleft, 0, &
188	ar(nysg,nxr+1), 1, type_y_int, pright, 0, &
189	comm2d, status, ierr )
190	!
191	!-- Send right boundary, receive left one
192	CALL MPI_SENDRECV( ar(nysg,nxr+1-nbgp), 1, type_y_int, pright, 1, &
193	ar(nysg,nxlg), 1, type_y_int, pleft, 1, &
194	comm2d, status, ierr )
195
196	ENDIF
197
198	IF ( pdims(2) == 1 ) THEN
199	!
200	!-- One-dimensional decomposition along x, boundary values can be exchanged
201	!-- within the PE memory
202	ar(nysg:nys-1,:) = ar(nyn+1-nbgp:nyn,:)
203	ar(nyn+1:nyng,:) = ar(nys:nys-1+nbgp,:)
204
205
206	ELSE
207	!
208	!-- Send front boundary, receive rear one
209	CALL MPI_SENDRECV( ar(nys,nxlg), 1, type_x_int, psouth, 0, &
210	ar(nyn+1,nxlg), 1, type_x_int, pnorth, 0, &
211	comm2d, status, ierr )
212
213	!
214	!-- Send rear boundary, receive front one
215	CALL MPI_SENDRECV( ar(nyn+1-nbgp,nxlg), 1, type_x_int, pnorth, 1, &
216	ar(nysg,nxlg), 1, type_x_int, psouth, 1, &
217	comm2d, status, ierr )
218
219	ENDIF
220
221	#else
222
223	!
224	!-- Lateral boundary conditions in the non-parallel case
225	IF ( bc_lr_cyc ) THEN
226	ar(:,nxlg:nxl-1) = ar(:,nxr-nbgp+1:nxr)
227	ar(:,nxr+1:nxrg) = ar(:,nxl:nxl+nbgp-1)
228	ENDIF
229
230	IF ( bc_ns_cyc ) THEN
231	ar(nysg:nys-1,:) = ar(nyn+1-nbgp:nyn,:)
232	ar(nyn+1:nyng,:) = ar(nys:nys-1+nbgp,:)
233	ENDIF
234
235	#endif
236	CALL cpu_log( log_point_s(13), 'exchange_horiz_2d', 'stop' )
237
238	END SUBROUTINE exchange_horiz_2d_int

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

Context Navigation

source: palm/trunk/SOURCE/exchange_horiz_2d.f90 @ 707

Download in other formats: