5 |
#ifdef ALLOW_USE_MPI |
#ifdef ALLOW_USE_MPI |
6 |
C HACK to avoid global_max |
C HACK to avoid global_max |
7 |
# define ALLOW_CONST_RHSMAX |
# define ALLOW_CONST_RHSMAX |
8 |
#endif |
#endif |
9 |
|
|
10 |
CML THIS DOES NOT WORK +++++ |
CML THIS DOES NOT WORK +++++ |
11 |
#undef ALLOW_LOOP_DIRECTIVE |
#undef ALLOW_LOOP_DIRECTIVE |
12 |
CBOP |
CBOP |
13 |
C !ROUTINE: CG2D_NSA |
C !ROUTINE: CG2D_NSA |
14 |
C !INTERFACE: |
C !INTERFACE: |
15 |
SUBROUTINE CG2D_NSA( |
SUBROUTINE CG2D_NSA( |
16 |
I cg2d_b, |
I cg2d_b, |
17 |
U cg2d_x, |
U cg2d_x, |
18 |
O firstResidual, |
O firstResidual, |
21 |
I myThid ) |
I myThid ) |
22 |
C !DESCRIPTION: \bv |
C !DESCRIPTION: \bv |
23 |
C *==========================================================* |
C *==========================================================* |
24 |
C | SUBROUTINE CG2D_NSA |
C | SUBROUTINE CG2D_NSA |
25 |
C | o Two-dimensional grid problem conjugate-gradient |
C | o Two-dimensional grid problem conjugate-gradient |
26 |
C | inverter (with preconditioner). |
C | inverter (with preconditioner). |
27 |
C | o This version is used only in the case when the matrix |
C | o This version is used only in the case when the matrix |
28 |
C | operator is not "self-adjoint" (NSA). Any remaining |
C | operator is not "self-adjoint" (NSA). Any remaining |
29 |
C | residuals will immediately reported to the department |
C | residuals will immediately reported to the department |
30 |
C | of homeland security. |
C | of homeland security. |
31 |
C *==========================================================* |
C *==========================================================* |
32 |
C | Con. grad is an iterative procedure for solving Ax = b. |
C | Con. grad is an iterative procedure for solving Ax = b. |
33 |
C | It requires the A be symmetric. |
C | It requires the A be symmetric. |
34 |
C | This implementation assumes A is a five-diagonal |
C | This implementation assumes A is a five-diagonal |
35 |
C | matrix of the form that arises in the discrete |
C | matrix of the form that arises in the discrete |
36 |
C | representation of the del^2 operator in a |
C | representation of the del^2 operator in a |
37 |
C | two-dimensional space. |
C | two-dimensional space. |
38 |
C | Notes: |
C | Notes: |
39 |
C | ====== |
C | ====== |
40 |
C | This implementation can support shared-memory |
C | This implementation can support shared-memory |
41 |
C | multi-threaded execution. In order to do this COMMON |
C | multi-threaded execution. In order to do this COMMON |
42 |
C | blocks are used for many of the arrays - even ones that |
C | blocks are used for many of the arrays - even ones that |
43 |
C | are only used for intermedaite results. This design is |
C | are only used for intermedaite results. This design is |
44 |
C | OK if you want to all the threads to collaborate on |
C | OK if you want to all the threads to collaborate on |
45 |
C | solving the same problem. On the other hand if you want |
C | solving the same problem. On the other hand if you want |
46 |
C | the threads to solve several different problems |
C | the threads to solve several different problems |
47 |
C | concurrently this implementation will not work. |
C | concurrently this implementation will not work. |
48 |
C *==========================================================* |
C *==========================================================* |
49 |
C \ev |
C \ev |
50 |
|
|
54 |
#include "SIZE.h" |
#include "SIZE.h" |
55 |
#include "EEPARAMS.h" |
#include "EEPARAMS.h" |
56 |
#include "PARAMS.h" |
#include "PARAMS.h" |
|
#include "GRID.h" |
|
57 |
#include "CG2D.h" |
#include "CG2D.h" |
58 |
#include "SURFACE.h" |
c#include "GRID.h" |
59 |
|
c#include "SURFACE.h" |
60 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
61 |
# include "tamc.h" |
# include "tamc.h" |
62 |
# include "tamc_keys.h" |
# include "tamc_keys.h" |
64 |
|
|
65 |
C !INPUT/OUTPUT PARAMETERS: |
C !INPUT/OUTPUT PARAMETERS: |
66 |
C === Routine arguments === |
C === Routine arguments === |
67 |
C myThid - Thread on which I am working. |
C cg2d_b :: The source term or "right hand side" |
68 |
C cg2d_b - The source term or "right hand side" |
C cg2d_x :: The solution |
69 |
C cg2d_x - The solution |
C firstResidual :: the initial residual before any iterations |
70 |
C firstResidual - the initial residual before any iterations |
C lastResidual :: the actual residual reached |
71 |
C lastResidual - the actual residual reached |
C numIters :: Entry: the maximum number of iterations allowed |
72 |
C numIters - Entry: the maximum number of iterations allowed |
C Exit: the actual number of iterations used |
73 |
C Exit: the actual number of iterations used |
C myThid :: Thread on which I am working. |
74 |
_RL cg2d_b(1-OLx:sNx+OLx,1-OLy:sNy+OLy,nSx,nSy) |
_RL cg2d_b(1-OLx:sNx+OLx,1-OLy:sNy+OLy,nSx,nSy) |
75 |
_RL cg2d_x(1-OLx:sNx+OLx,1-OLy:sNy+OLy,nSx,nSy) |
_RL cg2d_x(1-OLx:sNx+OLx,1-OLy:sNy+OLy,nSx,nSy) |
76 |
_RL firstResidual |
_RL firstResidual |
81 |
#ifdef ALLOW_CG2D_NSA |
#ifdef ALLOW_CG2D_NSA |
82 |
C !LOCAL VARIABLES: |
C !LOCAL VARIABLES: |
83 |
C === Local variables ==== |
C === Local variables ==== |
84 |
C actualIts - Number of iterations taken |
C actualIts :: Number of iterations taken |
85 |
C actualResidual - residual |
C actualResidual :: residual |
86 |
C bi - Block index in X and Y. |
C bi, bj :: Block index in X and Y. |
87 |
C bj |
C eta_qrN :: Used in computing search directions |
|
C eta_qrN - Used in computing search directions |
|
88 |
C eta_qrNM1 suffix N and NM1 denote current and |
C eta_qrNM1 suffix N and NM1 denote current and |
89 |
C cgBeta previous iterations respectively. |
C cgBeta previous iterations respectively. |
90 |
C recip_eta_qrNM1 reciprocal of eta_qrNM1 |
C recip_eta_qrNM1 :: reciprocal of eta_qrNM1 |
91 |
C alpha |
C alpha |
92 |
C alpha_aux - to avoid the statement: alpha = 1./alpha (for TAMC/TAF) |
C alpha_aux :: to avoid the statement: alpha = 1./alpha (for TAMC/TAF) |
93 |
C sumRHS - Sum of right-hand-side. Sometimes this is a |
C sumRHS :: Sum of right-hand-side. Sometimes this is a |
94 |
C useful debuggin/trouble shooting diagnostic. |
C useful debuggin/trouble shooting diagnostic. |
95 |
C For neumann problems sumRHS needs to be ~0. |
C For neumann problems sumRHS needs to be ~0. |
96 |
C or they converge at a non-zero residual. |
C or they converge at a non-zero residual. |
97 |
C err - Measure of residual of Ax - b, usually the norm. |
C err :: Measure of residual of Ax - b, usually the norm. |
98 |
C err_sq - square of err (for TAMC/TAF) |
C err_sq :: square of err (for TAMC/TAF) |
99 |
C I, J, N - Loop counters ( N counts CG iterations ) |
C I, J, it2d :: Loop counters ( it2d counts CG iterations ) |
100 |
INTEGER actualIts |
INTEGER actualIts |
101 |
_RL actualResidual |
_RL actualResidual |
102 |
INTEGER bi, bj |
INTEGER bi, bj |
103 |
INTEGER I, J, it2d |
INTEGER I, J, it2d |
104 |
|
|
105 |
_RL err |
_RL err |
118 |
|
|
119 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
120 |
IF ( numIters .GT. numItersMax ) THEN |
IF ( numIters .GT. numItersMax ) THEN |
121 |
WRITE(standardMessageUnit,'(A,I10)') |
WRITE(standardMessageUnit,'(A,I10)') |
122 |
& 'CG2D_NSA: numIters > numItersMax = ', numItersMax |
& 'CG2D_NSA: numIters > numItersMax = ', numItersMax |
123 |
STOP 'NON-NORMAL in CG2D_NSA' |
STOP 'NON-NORMAL in CG2D_NSA' |
124 |
ENDIF |
ENDIF |
149 |
|
|
150 |
C-- Normalise RHS |
C-- Normalise RHS |
151 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
152 |
CADJ STORE cg2d_b = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE cg2d_b = comlev1_cg2d, key = ikey, byte = isbyte |
153 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
154 |
rhsMax = 0. _d 0 |
rhsMax = 0. _d 0 |
155 |
DO bj=myByLo(myThid),myByHi(myThid) |
DO bj=myByLo(myThid),myByHi(myThid) |
177 |
#endif /* ALLOW_CONST_RHSMAX */ |
#endif /* ALLOW_CONST_RHSMAX */ |
178 |
#endif |
#endif |
179 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
180 |
CADJ STORE rhsNorm = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE rhsNorm = comlev1_cg2d, key = ikey, byte = isbyte |
181 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
182 |
IF ( rhsMaxGlobal .NE. 0. ) THEN |
IF ( rhsMaxGlobal .NE. 0. ) THEN |
183 |
rhsNorm = 1. _d 0 / rhsMaxGlobal |
rhsNorm = 1. _d 0 / rhsMaxGlobal |
186 |
ENDIF |
ENDIF |
187 |
|
|
188 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
189 |
CADJ STORE cg2d_b = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE cg2d_b = comlev1_cg2d, key = ikey, byte = isbyte |
190 |
CADJ STORE cg2d_x = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE cg2d_x = comlev1_cg2d, key = ikey, byte = isbyte |
191 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
192 |
DO bj=myByLo(myThid),myByHi(myThid) |
DO bj=myByLo(myThid),myByHi(myThid) |
193 |
DO bi=myBxLo(myThid),myBxHi(myThid) |
DO bi=myBxLo(myThid),myBxHi(myThid) |
203 |
ENDIF |
ENDIF |
204 |
|
|
205 |
C-- Update overlaps |
C-- Update overlaps |
206 |
_EXCH_XY_RL( cg2d_b, myThid ) |
c CALL EXCH_XY_RL( cg2d_b, myThid ) |
207 |
_EXCH_XY_RL( cg2d_x, myThid ) |
CALL EXCH_XY_RL( cg2d_x, myThid ) |
208 |
CcnhDebugStarts |
CcnhDebugStarts |
209 |
C CALL PLOT_FIELD_XYRL( cg2d_b, 'CG2D.1 CG2D_B' , 1, myThid ) |
C CALL PLOT_FIELD_XYRL( cg2d_b, 'CG2D.1 CG2D_B' , 1, myThid ) |
210 |
C suff = 'normalised' |
C suff = 'normalised' |
216 |
err_sq = 0. _d 0 |
err_sq = 0. _d 0 |
217 |
sumRHS = 0. _d 0 |
sumRHS = 0. _d 0 |
218 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
219 |
CADJ STORE cg2d_b = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE cg2d_b = comlev1_cg2d, key = ikey, byte = isbyte |
220 |
CADJ STORE cg2d_x = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE cg2d_x = comlev1_cg2d, key = ikey, byte = isbyte |
221 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
222 |
DO bj=myByLo(myThid),myByHi(myThid) |
DO bj=myByLo(myThid),myByHi(myThid) |
223 |
DO bi=myBxLo(myThid),myBxHi(myThid) |
DO bi=myBxLo(myThid),myBxHi(myThid) |
224 |
|
DO J=1-1,sNy+1 |
225 |
|
DO I=1-1,sNx+1 |
226 |
|
cg2d_s(I,J,bi,bj) = 0. |
227 |
|
ENDDO |
228 |
|
ENDDO |
229 |
DO J=1,sNy |
DO J=1,sNy |
230 |
DO I=1,sNx |
DO I=1,sNx |
|
cg2d_s(I,J,bi,bj) = 0. |
|
231 |
cg2d_r(I,J,bi,bj) = cg2d_b(I,J,bi,bj) - |
cg2d_r(I,J,bi,bj) = cg2d_b(I,J,bi,bj) - |
232 |
& (aW2d(I ,J ,bi,bj)*cg2d_x(I-1,J ,bi,bj) |
& (aW2d(I ,J ,bi,bj)*cg2d_x(I-1,J ,bi,bj) |
233 |
& +aW2d(I+1,J ,bi,bj)*cg2d_x(I+1,J ,bi,bj) |
& +aW2d(I+1,J ,bi,bj)*cg2d_x(I+1,J ,bi,bj) |
234 |
& +aS2d(I ,J ,bi,bj)*cg2d_x(I ,J-1,bi,bj) |
& +aS2d(I ,J ,bi,bj)*cg2d_x(I ,J-1,bi,bj) |
235 |
& +aS2d(I ,J+1,bi,bj)*cg2d_x(I ,J+1,bi,bj) |
& +aS2d(I ,J+1,bi,bj)*cg2d_x(I ,J+1,bi,bj) |
236 |
& -aW2d(I ,J ,bi,bj)*cg2d_x(I ,J ,bi,bj) |
& +aC2d(I ,J ,bi,bj)*cg2d_x(I ,J ,bi,bj) |
|
& -aW2d(I+1,J ,bi,bj)*cg2d_x(I ,J ,bi,bj) |
|
|
& -aS2d(I ,J ,bi,bj)*cg2d_x(I ,J ,bi,bj) |
|
|
& -aS2d(I ,J+1,bi,bj)*cg2d_x(I ,J ,bi,bj) |
|
|
& -freeSurfFac*_rA(i,j,bi,bj)*recip_Bo(i,j,bi,bj)* |
|
|
& cg2d_x(I ,J ,bi,bj)/deltaTMom/deltaTfreesurf*cg2dNorm |
|
|
CML & cg2d_x(I ,J ,bi,bj)/deltaTMom/deltaTMom*cg2dNorm |
|
237 |
& ) |
& ) |
238 |
err_sq = err_sq + |
c & -aW2d(I ,J ,bi,bj)*cg2d_x(I ,J ,bi,bj) |
239 |
|
c & -aW2d(I+1,J ,bi,bj)*cg2d_x(I ,J ,bi,bj) |
240 |
|
c & -aS2d(I ,J ,bi,bj)*cg2d_x(I ,J ,bi,bj) |
241 |
|
c & -aS2d(I ,J+1,bi,bj)*cg2d_x(I ,J ,bi,bj) |
242 |
|
c & -freeSurfFac*_rA(i,j,bi,bj)*recip_Bo(i,j,bi,bj)* |
243 |
|
c & cg2d_x(I ,J ,bi,bj)/deltaTMom/deltaTfreesurf*cg2dNorm |
244 |
|
cML & cg2d_x(I ,J ,bi,bj)/deltaTMom/deltaTMom*cg2dNorm |
245 |
|
c & ) |
246 |
|
err_sq = err_sq + |
247 |
& cg2d_r(I,J,bi,bj)*cg2d_r(I,J,bi,bj) |
& cg2d_r(I,J,bi,bj)*cg2d_r(I,J,bi,bj) |
248 |
sumRHS = sumRHS + |
sumRHS = sumRHS + |
249 |
& cg2d_b(I,J,bi,bj) |
& cg2d_b(I,J,bi,bj) |
252 |
ENDDO |
ENDDO |
253 |
ENDDO |
ENDDO |
254 |
|
|
255 |
#ifdef LETS_MAKE_JAM |
c CALL EXCH_S3D_RL( cg2d_r, 1, myThid ) |
256 |
CALL EXCH_XY_O1_R8_JAM( cg2d_r ) |
CALL EXCH_XY_RL ( cg2d_r, myThid ) |
257 |
CALL EXCH_XY_O1_R8_JAM( cg2d_s ) |
_GLOBAL_SUM_RL( sumRHS, myThid ) |
258 |
#else |
_GLOBAL_SUM_RL( err_sq, myThid ) |
259 |
_EXCH_XY_RL( cg2d_r, myThid ) |
IF ( err_sq .NE. 0. ) THEN |
|
_EXCH_XY_RL( cg2d_s, myThid ) |
|
|
#endif |
|
|
_GLOBAL_SUM_RL( sumRHS, myThid ) |
|
|
_GLOBAL_SUM_RL( err_sq, myThid ) |
|
|
if ( err_sq .ne. 0. ) then |
|
260 |
err = SQRT(err_sq) |
err = SQRT(err_sq) |
261 |
else |
ELSE |
262 |
err = 0. |
err = 0. |
263 |
end if |
ENDIF |
264 |
actualIts = 0 |
actualIts = 0 |
265 |
actualResidual = err |
actualResidual = err |
266 |
|
|
267 |
_BEGIN_MASTER( myThid ) |
IF ( debugLevel .GE. debLevZero ) THEN |
268 |
write(standardMessageUnit,'(A,1P2E22.14)') |
_BEGIN_MASTER( myThid ) |
269 |
& ' cg2d: Sum(rhs),rhsMax = ', sumRHS,rhsMaxGlobal |
WRITE(standardmessageunit,'(A,1P2E22.14)') |
270 |
_END_MASTER( ) |
& ' cg2d: Sum(rhs),rhsMax = ', sumRHS,rhsMaxGlobal |
271 |
|
_END_MASTER( myThid ) |
272 |
|
ENDIF |
273 |
C _BARRIER |
C _BARRIER |
274 |
c _BEGIN_MASTER( myThid ) |
c _BEGIN_MASTER( myThid ) |
275 |
c WRITE(*,'(A,I6,1PE30.14)') ' CG2D_NSA iters, err = ', |
c WRITE(*,'(A,I6,1PE30.14)') ' CG2D_NSA iters, err = ', |
276 |
c & actualIts, actualResidual |
c & actualIts, actualResidual |
277 |
c _END_MASTER( ) |
c _END_MASTER( myThid ) |
278 |
firstResidual=actualResidual |
firstResidual=actualResidual |
279 |
cg2dTolerance_sq = cg2dTolerance**2 |
cg2dTolerance_sq = cg2dTolerance**2 |
280 |
|
|
286 |
DO it2d=1, numIters |
DO it2d=1, numIters |
287 |
#ifdef ALLOW_LOOP_DIRECTIVE |
#ifdef ALLOW_LOOP_DIRECTIVE |
288 |
CML it2d = 0 |
CML it2d = 0 |
289 |
CML DO WHILE ( err_sq .GT. cg2dTolerance_sq .and. it2d .LT. numIters ) |
CML DO WHILE ( err_sq .GT. cg2dTolerance_sq .and. it2d .LT. numIters ) |
290 |
CML it2d = it2d+1 |
CML it2d = it2d+1 |
291 |
#endif /* ALLOW_LOOP_DIRECTIVE */ |
#endif /* ALLOW_LOOP_DIRECTIVE */ |
292 |
|
|
293 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
294 |
icg2dkey = (ikey-1)*numItersMax + it2d |
icg2dkey = (ikey-1)*numItersMax + it2d |
295 |
CMLCADJ STORE err = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CMLCADJ STORE err = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
296 |
CADJ STORE err_sq = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CADJ STORE err_sq = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
297 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
298 |
CML IF ( err .LT. cg2dTolerance ) THEN |
CML IF ( err .LT. cg2dTolerance ) THEN |
299 |
IF ( err_sq .LT. cg2dTolerance_sq ) THEN |
IF ( err_sq .LT. cg2dTolerance_sq ) THEN |
308 |
C-- conjugate direction vector "s". |
C-- conjugate direction vector "s". |
309 |
eta_qrN = 0. _d 0 |
eta_qrN = 0. _d 0 |
310 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
311 |
CADJ STORE cg2d_r = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CADJ STORE cg2d_r = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
312 |
CADJ STORE cg2d_s = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CADJ STORE cg2d_s = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
313 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
314 |
DO bj=myByLo(myThid),myByHi(myThid) |
DO bj=myByLo(myThid),myByHi(myThid) |
315 |
DO bi=myBxLo(myThid),myBxHi(myThid) |
DO bi=myBxLo(myThid),myBxHi(myThid) |
316 |
DO J=1,sNy |
DO J=1,sNy |
317 |
DO I=1,sNx |
DO I=1,sNx |
318 |
cg2d_z(I,J,bi,bj) = |
cg2d_z(I,J,bi,bj) = |
319 |
& pC(I ,J ,bi,bj)*cg2d_r(I ,J ,bi,bj) |
& pC(I ,J ,bi,bj)*cg2d_r(I ,J ,bi,bj) |
320 |
& +pW(I ,J ,bi,bj)*cg2d_r(I-1,J ,bi,bj) |
& +pW(I ,J ,bi,bj)*cg2d_r(I-1,J ,bi,bj) |
321 |
& +pW(I+1,J ,bi,bj)*cg2d_r(I+1,J ,bi,bj) |
& +pW(I+1,J ,bi,bj)*cg2d_r(I+1,J ,bi,bj) |
336 |
C WRITE(*,*) ' CG2D_NSA: Iteration ',it2d-1,' eta_qrN = ',eta_qrN |
C WRITE(*,*) ' CG2D_NSA: Iteration ',it2d-1,' eta_qrN = ',eta_qrN |
337 |
CcnhDebugEnds |
CcnhDebugEnds |
338 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
339 |
CMLCADJ STORE eta_qrNM1 = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CMLCADJ STORE eta_qrNM1 = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
340 |
CADJ STORE recip_eta_qrNM1 = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CADJ STORE recip_eta_qrNM1 = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
341 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
342 |
CML cgBeta = eta_qrN/eta_qrNM1 |
CML cgBeta = eta_qrN/eta_qrNM1 |
343 |
cgBeta = eta_qrN*recip_eta_qrNM1 |
cgBeta = eta_qrN*recip_eta_qrNM1 |
344 |
CcnhDebugStarts |
CcnhDebugStarts |
345 |
C WRITE(*,*) ' CG2D_NSA: Iteration ',it2d-1,' beta = ',cgBeta |
C WRITE(*,*) ' CG2D_NSA: Iteration ',it2d-1,' beta = ',cgBeta |
346 |
CcnhDebugEnds |
CcnhDebugEnds |
347 |
Cml store normalisation factor for the next interation |
Cml store normalisation factor for the next interation |
348 |
Cml (in case there is one). |
Cml (in case there is one). |
349 |
CML store the inverse of the normalization factor for higher precision |
CML store the inverse of the normalization factor for higher precision |
350 |
CML eta_qrNM1 = eta_qrN |
CML eta_qrNM1 = eta_qrN |
363 |
|
|
364 |
C-- Do exchanges that require messages i.e. between |
C-- Do exchanges that require messages i.e. between |
365 |
C-- processes. |
C-- processes. |
366 |
#ifdef LETS_MAKE_JAM |
c CALL EXCH_S3D_RL( cg2d_s, 1, myThid ) |
367 |
CALL EXCH_XY_O1_R8_JAM( cg2d_s ) |
CALL EXCH_XY_RL ( cg2d_s, myThid ) |
|
#else |
|
|
_EXCH_XY_RL( cg2d_s, myThid ) |
|
|
#endif |
|
368 |
|
|
369 |
C== Evaluate laplace operator on conjugate gradient vector |
C== Evaluate laplace operator on conjugate gradient vector |
370 |
C== q = A.s |
C== q = A.s |
372 |
alpha_aux = 0. _d 0 |
alpha_aux = 0. _d 0 |
373 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
374 |
#ifndef ALLOW_LOOP_DIRECTIVE |
#ifndef ALLOW_LOOP_DIRECTIVE |
375 |
CADJ STORE cg2d_s = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CADJ STORE cg2d_s = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
376 |
#endif /* not ALLOW_LOOP_DIRECTIVE */ |
#endif /* not ALLOW_LOOP_DIRECTIVE */ |
377 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
378 |
DO bj=myByLo(myThid),myByHi(myThid) |
DO bj=myByLo(myThid),myByHi(myThid) |
379 |
DO bi=myBxLo(myThid),myBxHi(myThid) |
DO bi=myBxLo(myThid),myBxHi(myThid) |
380 |
DO J=1,sNy |
DO J=1,sNy |
381 |
DO I=1,sNx |
DO I=1,sNx |
382 |
cg2d_q(I,J,bi,bj) = |
cg2d_q(I,J,bi,bj) = |
383 |
& aW2d(I ,J ,bi,bj)*cg2d_s(I-1,J ,bi,bj) |
& aW2d(I ,J ,bi,bj)*cg2d_s(I-1,J ,bi,bj) |
384 |
& +aW2d(I+1,J ,bi,bj)*cg2d_s(I+1,J ,bi,bj) |
& +aW2d(I+1,J ,bi,bj)*cg2d_s(I+1,J ,bi,bj) |
385 |
& +aS2d(I ,J ,bi,bj)*cg2d_s(I ,J-1,bi,bj) |
& +aS2d(I ,J ,bi,bj)*cg2d_s(I ,J-1,bi,bj) |
386 |
& +aS2d(I ,J+1,bi,bj)*cg2d_s(I ,J+1,bi,bj) |
& +aS2d(I ,J+1,bi,bj)*cg2d_s(I ,J+1,bi,bj) |
387 |
& -aW2d(I ,J ,bi,bj)*cg2d_s(I ,J ,bi,bj) |
& +aC2d(I ,J ,bi,bj)*cg2d_s(I ,J ,bi,bj) |
388 |
& -aW2d(I+1,J ,bi,bj)*cg2d_s(I ,J ,bi,bj) |
c & -aW2d(I ,J ,bi,bj)*cg2d_s(I ,J ,bi,bj) |
389 |
& -aS2d(I ,J ,bi,bj)*cg2d_s(I ,J ,bi,bj) |
c & -aW2d(I+1,J ,bi,bj)*cg2d_s(I ,J ,bi,bj) |
390 |
& -aS2d(I ,J+1,bi,bj)*cg2d_s(I ,J ,bi,bj) |
c & -aS2d(I ,J ,bi,bj)*cg2d_s(I ,J ,bi,bj) |
391 |
& -freeSurfFac*_rA(i,j,bi,bj)*recip_Bo(i,j,bi,bj)* |
c & -aS2d(I ,J+1,bi,bj)*cg2d_s(I ,J ,bi,bj) |
392 |
& cg2d_s(I ,J ,bi,bj)/deltaTMom/deltaTfreesurf*cg2dNorm |
c & -freeSurfFac*_rA(i,j,bi,bj)*recip_Bo(i,j,bi,bj)* |
393 |
CML & cg2d_s(I ,J ,bi,bj)/deltaTMom/deltaTMom*cg2dNorm |
c & cg2d_s(I ,J ,bi,bj)/deltaTMom/deltaTfreesurf*cg2dNorm |
394 |
alpha_aux = alpha_aux+cg2d_s(I,J,bi,bj)*cg2d_q(I,J,bi,bj) |
cML & cg2d_s(I ,J ,bi,bj)/deltaTMom/deltaTMom*cg2dNorm |
395 |
|
alpha_aux = alpha_aux+cg2d_s(I,J,bi,bj)*cg2d_q(I,J,bi,bj) |
396 |
ENDDO |
ENDDO |
397 |
ENDDO |
ENDDO |
398 |
ENDDO |
ENDDO |
405 |
CcnhDebugStarts |
CcnhDebugStarts |
406 |
C WRITE(*,*) ' CG2D_NSA: Iteration ',it2d-1,' alpha= ',alpha |
C WRITE(*,*) ' CG2D_NSA: Iteration ',it2d-1,' alpha= ',alpha |
407 |
CcnhDebugEnds |
CcnhDebugEnds |
408 |
|
|
409 |
C== Update solution and residual vectors |
C== Update solution and residual vectors |
410 |
C Now compute "interior" points. |
C Now compute "interior" points. |
411 |
err = 0. _d 0 |
err = 0. _d 0 |
412 |
err_sq = 0. _d 0 |
err_sq = 0. _d 0 |
413 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
414 |
#ifndef ALLOW_LOOP_DIRECTIVE |
#ifndef ALLOW_LOOP_DIRECTIVE |
415 |
CADJ STORE cg2d_r = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
CADJ STORE cg2d_r = comlev1_cg2d_iter, key = icg2dkey, byte = isbyte |
416 |
#endif /* ALLOW_LOOP_DIRECTIVE */ |
#endif /* ALLOW_LOOP_DIRECTIVE */ |
417 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
418 |
DO bj=myByLo(myThid),myByHi(myThid) |
DO bj=myByLo(myThid),myByHi(myThid) |
436 |
actualIts = it2d |
actualIts = it2d |
437 |
actualResidual = err |
actualResidual = err |
438 |
|
|
439 |
#ifdef LETS_MAKE_JAM |
c CALL EXCH_S3D_RL( cg2d_r, 1, myThid ) |
440 |
CALL EXCH_XY_O1_R8_JAM( cg2d_r ) |
CALL EXCH_XY_RL ( cg2d_r, myThid ) |
|
#else |
|
|
_EXCH_XY_RL( cg2d_r, myThid ) |
|
|
_EXCH_XY_RL( cg2d_x, myThid ) |
|
|
#endif |
|
441 |
|
|
442 |
Cml end of IF ( err .LT. cg2dTolerance ) THEN; ELSE |
Cml end of IF ( err .LT. cg2dTolerance ) THEN; ELSE |
443 |
ENDIF |
ENDIF |
446 |
|
|
447 |
IF (cg2dNormaliseRHS) THEN |
IF (cg2dNormaliseRHS) THEN |
448 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
449 |
CADJ STORE rhsNorm = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE rhsNorm = comlev1_cg2d, key = ikey, byte = isbyte |
450 |
CADJ STORE cg2d_x = comlev1_cg2d, key = ikey, byte = isbyte |
CADJ STORE cg2d_x = comlev1_cg2d, key = ikey, byte = isbyte |
451 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
452 |
C-- Un-normalise the answer |
C-- Un-normalise the answer |
453 |
DO bj=myByLo(myThid),myByHi(myThid) |
DO bj=myByLo(myThid),myByHi(myThid) |
465 |
C for compatibility with TAMC. |
C for compatibility with TAMC. |
466 |
C _EXCH_XY_RL(cg2d_x, myThid ) |
C _EXCH_XY_RL(cg2d_x, myThid ) |
467 |
c _BEGIN_MASTER( myThid ) |
c _BEGIN_MASTER( myThid ) |
468 |
c WRITE(*,'(A,I6,1PE30.14)') ' CG2D_NSA iters, err = ', |
c WRITE(*,'(A,I6,1PE30.14)') ' CG2D_NSA iters, err = ', |
469 |
c & actualIts, actualResidual |
c & actualIts, actualResidual |
470 |
c _END_MASTER( ) |
c _END_MASTER( ) |
471 |
|
|
492 |
character*(*) chardum |
character*(*) chardum |
493 |
integer int1, int2, int3, idow, icount |
integer int1, int2, int3, idow, icount |
494 |
|
|
495 |
C the length of this vector must be greater or equal |
C the length of this vector must be greater or equal |
496 |
C twice the number of timesteps |
C twice the number of timesteps |
497 |
integer nidow |
integer nidow |
498 |
#ifdef ALLOW_TAMC_CHECKPOINTING |
#ifdef ALLOW_TAMC_CHECKPOINTING |
503 |
integer istoreidow(nidow) |
integer istoreidow(nidow) |
504 |
common /istorecommon/ istoreidow |
common /istorecommon/ istoreidow |
505 |
|
|
506 |
print *, 'adstore: ', chardum, int1, idow, int2, int3, icount |
print *, 'adstore: ', chardum, int1, idow, int2, int3, icount |
507 |
|
|
508 |
if ( icount .gt. nidow ) then |
if ( icount .gt. nidow ) then |
509 |
print *, 'adstore: error: icount > nidow = ', nidow |
print *, 'adstore: error: icount > nidow = ', nidow |
526 |
integer int1, int2, int3, idow, icount |
integer int1, int2, int3, idow, icount |
527 |
|
|
528 |
|
|
529 |
C the length of this vector must be greater or equal |
C the length of this vector must be greater or equal |
530 |
C twice the number of timesteps |
C twice the number of timesteps |
531 |
integer nidow |
integer nidow |
532 |
#ifdef ALLOW_TAMC_CHECKPOINTING |
#ifdef ALLOW_TAMC_CHECKPOINTING |
537 |
integer istoreidow(nidow) |
integer istoreidow(nidow) |
538 |
common /istorecommon/ istoreidow |
common /istorecommon/ istoreidow |
539 |
|
|
540 |
print *, 'adresto: ', chardum, int1, idow, int2, int3, icount |
print *, 'adresto: ', chardum, int1, idow, int2, int3, icount |
541 |
|
|
542 |
if ( icount .gt. nidow ) then |
if ( icount .gt. nidow ) then |
543 |
print *, 'adstore: error: icount > nidow = ', nidow |
print *, 'adstore: error: icount > nidow = ', nidow |
550 |
end |
end |
551 |
# endif /* ALLOW_AUTODIFF_TAMC and ALLOW_LOOP_DIRECTIVE */ |
# endif /* ALLOW_AUTODIFF_TAMC and ALLOW_LOOP_DIRECTIVE */ |
552 |
|
|
|
|
|