-
Notifications
You must be signed in to change notification settings - Fork 2
/
pm_nonperiodic.c
1675 lines (1372 loc) · 50.5 KB
/
pm_nonperiodic.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math.h>
#include <mpi.h>
/*! \file pm_nonperiodic.c
* \brief code for non-periodic FFT to compute long-range PM force
*/
#ifdef PMGRID
#if !defined (PERIODIC) || defined (PLACEHIGHRESREGION)
#ifdef NOTYPEPREFIX_FFTW
#include <rfftw_mpi.h>
#else
#ifdef DOUBLEPRECISION_FFTW
#include <drfftw_mpi.h> /* double precision FFTW */
#else
#include <srfftw_mpi.h>
#endif
#endif
#include "allvars.h"
#include "proto.h"
#define GRID (2*PMGRID)
#define GRID2 (2*(GRID/2 + 1))
static rfftwnd_mpi_plan fft_forward_plan, fft_inverse_plan;
static int slab_to_task[GRID];
static int *slabs_per_task;
static int *first_slab_of_task;
static int *meshmin_list, *meshmax_list;
static int slabstart_x, nslab_x, slabstart_y, nslab_y;
static int fftsize, maxfftsize;
static fftw_real *kernel[2][N_GRAVS][N_GRAVS], *rhogrid, *forcegrid, *workspace;
static fftw_complex *fft_of_kernel[2][N_GRAVS][N_GRAVS], *fft_of_rhogrid;
/*! This function determines the particle extension of all particles, and for
* those types selected with PLACEHIGHRESREGION if this is used, and then
* determines the boundaries of the non-periodic FFT-mesh that can be placed
* on this region. Note that a sufficient buffer region at the rim of the
* occupied part of the mesh needs to be reserved in order to allow a correct
* finite differencing using a 4-point formula. In addition, to allow
* non-periodic boundaries, the actual FFT mesh used is twice as large in
* each dimension compared with PMGRID.
*/
void pm_init_regionsize(void)
{
double meshinner[2], xmin[2][3], xmax[2][3];
int i, j, t;
/* find enclosing rectangle */
for(j = 0; j < 3; j++)
{
xmin[0][j] = xmin[1][j] = 1.0e36;
xmax[0][j] = xmax[1][j] = -1.0e36;
}
for(i = 0; i < NumPart; i++)
for(j = 0; j < 3; j++)
{
t = 0;
#ifdef PLACEHIGHRESREGION
if(((1 << P[i].Type) & (PLACEHIGHRESREGION)))
t = 1;
#endif
if(P[i].Pos[j] > xmax[t][j])
xmax[t][j] = P[i].Pos[j];
if(P[i].Pos[j] < xmin[t][j])
xmin[t][j] = P[i].Pos[j];
}
MPI_Allreduce(xmin, All.Xmintot, 6, MPI_DOUBLE, MPI_MIN, MPI_COMM_WORLD);
MPI_Allreduce(xmax, All.Xmaxtot, 6, MPI_DOUBLE, MPI_MAX, MPI_COMM_WORLD);
for(j = 0; j < 2; j++)
{
All.TotalMeshSize[j] = All.Xmaxtot[j][0] - All.Xmintot[j][0];
All.TotalMeshSize[j] = dmax(All.TotalMeshSize[j], All.Xmaxtot[j][1] - All.Xmintot[j][1]);
All.TotalMeshSize[j] = dmax(All.TotalMeshSize[j], All.Xmaxtot[j][2] - All.Xmintot[j][2]);
#ifdef ENLARGEREGION
All.TotalMeshSize[j] *= ENLARGEREGION;
#endif
/* symmetrize the box onto the center */
for(i = 0; i < 3; i++)
{
All.Xmintot[j][i] = (All.Xmintot[j][i] + All.Xmaxtot[j][i]) / 2 - All.TotalMeshSize[j] / 2;
All.Xmaxtot[j][i] = All.Xmintot[j][i] + All.TotalMeshSize[j];
}
}
/* this will produce enough room for zero-padding and buffer region to
allow finite differencing of the potential */
for(j = 0; j < 2; j++)
{
meshinner[j] = All.TotalMeshSize[j];
All.TotalMeshSize[j] *= 2.001 * (GRID) / ((double) (GRID - 2 - 8));
}
/* move lower left corner by two cells to allow finite differencing of the potential by a 4-point function */
for(j = 0; j < 2; j++)
for(i = 0; i < 3; i++)
{
All.Corner[j][i] = All.Xmintot[j][i] - 2.0005 * All.TotalMeshSize[j] / GRID;
All.UpperCorner[j][i] = All.Corner[j][i] + (GRID / 2 - 1) * (All.TotalMeshSize[j] / GRID);
}
#ifndef PERIODIC
All.Asmth[0] = ASMTH * All.TotalMeshSize[0] / GRID;
All.Rcut[0] = RCUT * All.Asmth[0];
#endif
#ifdef PLACEHIGHRESREGION
All.Asmth[1] = ASMTH * All.TotalMeshSize[1] / GRID;
All.Rcut[1] = RCUT * All.Asmth[1];
#endif
#ifdef PLACEHIGHRESREGION
if(2 * All.TotalMeshSize[1] / GRID < All.Rcut[0])
{
All.TotalMeshSize[1] = 2 * (meshinner[1] + 2 * All.Rcut[0]) * (GRID) / ((double) (GRID - 2));
for(i = 0; i < 3; i++)
{
All.Corner[1][i] = All.Xmintot[1][i] - 1.0001 * All.Rcut[0];
All.UpperCorner[1][i] = All.Corner[1][i] + (GRID / 2 - 1) * (All.TotalMeshSize[1] / GRID);
}
if(2 * All.TotalMeshSize[1] / GRID > All.Rcut[0])
{
All.TotalMeshSize[1] = 2 * (meshinner[1] + 2 * All.Rcut[0]) * (GRID) / ((double) (GRID - 10));
for(i = 0; i < 3; i++)
{
All.Corner[1][i] = All.Xmintot[1][i] - 1.0001 * (All.Rcut[0] + 2 * All.TotalMeshSize[1] / GRID);
All.UpperCorner[1][i] = All.Corner[1][i] + (GRID / 2 - 1) * (All.TotalMeshSize[1] / GRID);
}
}
All.Asmth[1] = ASMTH * All.TotalMeshSize[1] / GRID;
All.Rcut[1] = RCUT * All.Asmth[1];
}
#endif
if(ThisTask == 0)
{
#ifndef PERIODIC
printf("\nAllowed region for isolated PM mesh (coarse):\n");
printf("(%g|%g|%g) -> (%g|%g|%g) ext=%g totmeshsize=%g meshsize=%g\n\n",
All.Xmintot[0][0], All.Xmintot[0][1], All.Xmintot[0][2],
All.Xmaxtot[0][0], All.Xmaxtot[0][1], All.Xmaxtot[0][2], meshinner[0], All.TotalMeshSize[0],
All.TotalMeshSize[0] / GRID);
#endif
#ifdef PLACEHIGHRESREGION
printf("\nAllowed region for isolated PM mesh (high-res):\n");
printf("(%g|%g|%g) -> (%g|%g|%g) ext=%g totmeshsize=%g meshsize=%g\n\n",
All.Xmintot[1][0], All.Xmintot[1][1], All.Xmintot[1][2],
All.Xmaxtot[1][0], All.Xmaxtot[1][1], All.Xmaxtot[1][2],
meshinner[1], All.TotalMeshSize[1], All.TotalMeshSize[1] / GRID);
#endif
}
}
/*! Initialization of the non-periodic PM routines. The plan-files for FFTW
* are created. Finally, the routine to set-up the non-periodic Greens
* function is called.
*/
void pm_init_nonperiodic(void)
{
int i, slab_to_task_local[GRID];
double bytes_tot = 0;
size_t bytes;
int n, m;
/* Set up the FFTW plan files. */
fft_forward_plan = rfftw3d_mpi_create_plan(MPI_COMM_WORLD, GRID, GRID, GRID,
FFTW_REAL_TO_COMPLEX, FFTW_ESTIMATE | FFTW_IN_PLACE);
fft_inverse_plan = rfftw3d_mpi_create_plan(MPI_COMM_WORLD, GRID, GRID, GRID,
FFTW_COMPLEX_TO_REAL, FFTW_ESTIMATE | FFTW_IN_PLACE);
/* Workspace out the ranges on each processor. */
rfftwnd_mpi_local_sizes(fft_forward_plan, &nslab_x, &slabstart_x, &nslab_y, &slabstart_y, &fftsize);
for(i = 0; i < GRID; i++)
slab_to_task_local[i] = 0;
for(i = 0; i < nslab_x; i++)
slab_to_task_local[slabstart_x + i] = ThisTask;
MPI_Allreduce(slab_to_task_local, slab_to_task, GRID, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
slabs_per_task = malloc(NTask * sizeof(int));
MPI_Allgather(&nslab_x, 1, MPI_INT, slabs_per_task, 1, MPI_INT, MPI_COMM_WORLD);
#ifndef PERIODIC
if(ThisTask == 0)
{
for(i = 0; i < NTask; i++)
printf("Task=%d FFT-Slabs=%d\n", i, slabs_per_task[i]);
}
#endif
first_slab_of_task = malloc(NTask * sizeof(int));
MPI_Allgather(&slabstart_x, 1, MPI_INT, first_slab_of_task, 1, MPI_INT, MPI_COMM_WORLD);
meshmin_list = malloc(3 * NTask * sizeof(int));
meshmax_list = malloc(3 * NTask * sizeof(int));
MPI_Allreduce(&fftsize, &maxfftsize, 1, MPI_INT, MPI_MAX, MPI_COMM_WORLD);
/* now allocate memory to hold the FFT fields */
#if !defined(PERIODIC)
for(n = 0; n < N_GRAVS; ++n) {
for(m = 0; m < N_GRAVS; ++m) {
if(!(kernel[0][n][m] = (fftw_real *) malloc(bytes = fftsize * sizeof(fftw_real))))
{
printf("failed to allocate memory for `FFT-kernel[0]' (%g MB).\n", bytes / (1024.0 * 1024.0));
endrun(1);
}
bytes_tot += bytes;
fft_of_kernel[0][n][m] = (fftw_complex *) kernel[0][n][m];
}
}
#endif
#if defined(PLACEHIGHRESREGION)
for(n = 0; n < N_GRAVS; ++n) {
for(m = 0; m < N_GRAVS; ++m) {
if(!(kernel[1][n][m] = (fftw_real *) malloc(bytes = fftsize * sizeof(fftw_real))))
{
printf("failed to allocate memory for `FFT-kernel[1]' (%g MB).\n", bytes / (1024.0 * 1024.0));
endrun(1);
}
bytes_tot += bytes;
fft_of_kernel[1][n][m] = (fftw_complex *) kernel[1][n][m];
}
}
#endif
if(ThisTask == 0)
printf("\nAllocated %g MByte for FFT kernel(s).\n\n", bytes_tot / (1024.0 * 1024.0));
}
/*! This function allocates the workspace needed for the non-periodic FFT
* algorithm. Three fields are used, one for the density/potential fields,
* one to hold the force field obtained by finite differencing, and finally
* an additional workspace which is used both in the parallel FFT itself, and
* as a buffer for the communication algorithm.
*/
void pm_init_nonperiodic_allocate(int dimprod)
{
static int first_alloc = 1;
int dimprodmax;
double bytes_tot = 0;
size_t bytes;
MPI_Allreduce(&dimprod, &dimprodmax, 1, MPI_INT, MPI_MAX, MPI_COMM_WORLD);
if(!(rhogrid = (fftw_real *) malloc(bytes = fftsize * sizeof(fftw_real))))
{
printf("failed to allocate memory for `FFT-rhogrid' (%g MB).\n", bytes / (1024.0 * 1024.0));
endrun(1);
}
bytes_tot += bytes;
fft_of_rhogrid = (fftw_complex *) rhogrid;
if(!(forcegrid = (fftw_real *) malloc(bytes = imax(fftsize, dimprodmax) * sizeof(fftw_real))))
{
printf("failed to allocate memory for `FFT-forcegrid' (%g MB).\n", bytes / (1024.0 * 1024.0));
endrun(1);
}
bytes_tot += bytes;
if(!(workspace = (fftw_real *) malloc(bytes = imax(maxfftsize, dimprodmax) * sizeof(fftw_real))))
{
printf("failed to allocate memory for `FFT-workspace' (%g MB).\n", bytes / (1024.0 * 1024.0));
endrun(1);
}
bytes_tot += bytes;
if(first_alloc == 1)
{
first_alloc = 0;
if(ThisTask == 0)
printf("\nUsing %g MByte for non-periodic FFT computation.\n\n", bytes_tot / (1024.0 * 1024.0));
}
}
/*! This function frees the memory allocated for the non-periodic FFT
* computation. (With the exception of the Greens function(s), which are kept
* statically in memory for the next force computation.)
*/
void pm_init_nonperiodic_free(void)
{
/* deallocate memory */
free(workspace);
free(forcegrid);
free(rhogrid);
}
/*! This function sets-up the Greens function for the non-periodic potential
* in real space, and then converts it to Fourier space by means of a FFT.
*/
// KC 10/17/14
// This function will need initialize N_GRAVS^2 more kernels
// both for the coarse mesh and the fine mesh.
// Now we gotta do N_GRAVS^2 of them
// which is kind of wasteful because we should only need to do N_GRAVS(N_GRAVS + 1)/2 of
// them by symmetry. However, this step is only carried out once on any processor, and so
// for clarity of code, we do it the dumb way.
//
void pm_setup_nonperiodic_kernel(void)
{
int i, j, k;
double x, y, z, r, u, fac;
double kx, ky, kz, k2, fx, fy, fz, ff;
int ip;
// ngravs
int n, m;
double asmth2;
/* now set up kernel and its Fourier transform */
pm_init_nonperiodic_allocate(0);
#if !defined(PERIODIC)
for(n = 0; n < N_GRAVS; ++n)
for(m = 0; m < N_GRAVS; ++m)
for(i = 0; i < fftsize; i++) /* clear local density field */
kernel[0][n][m][i] = 0;
for(n = 0; n < N_GRAVS; ++n)
for(m = 0; m < N_GRAVS; ++m) {
// KC 11/12/15
//
// New strategy: we will sample the k-space function, but do so at 1/2.
// This will let us use the existing GRID^3 dimension workspace as a,
// well, direct workspace. (so we pass null to the fftw routines here)
//
// We will then trilinearly interpolate to the k-space kernel the result
// using exact value when we have them. If trilinear is good enough for
// Ewald, then its probably good enough here. We won't need ghost cells
// for this because we have the exact values on the edges already.
//
/* for(i = slabstart_x; i < (slabstart_x + nslab_x); i++) */
/* for(j = 0; j < GRID; j++) */
/* for(k = 0; k < GRID; k++) */
//
// Note, we gotta start in transposed order!
// And we gotta get the halfcomplex assignment correct. Ugh.
// KC 12/12/15
fac = All.G / (M_PI * All.BoxSize); /* to get potential */
asmth2 = (2 * M_PI) * ASMTH / (double) GRID; ///All.Asmth[0] / All.BoxSize;
asmth2 *= asmth2;
fprintf(stderr, "# fac = %.15e, asmth2 = %.15e\n", fac, asmth2);
for(y = slabstart_y; y < slabstart_y + nslab_y; y++)
for(x = 0; x < GRID; x++)
for(z = 0; z < GRID / 2 + 1; z++)
{
if(x > GRID / 2)
kx = x - GRID;
else
kx = x;
if(y > GRID / 2)
ky = y - GRID;
else
ky = y;
if(z > GRID / 2)
kz = z - GRID;
else
kz = z;
k2 = kx * kx + ky * ky + kz * kz;
if(k2 > 0)
{
fx = fy = fz = 1;
if(kx != 0)
{
fx = (M_PI * kx) / GRID;
fx = sin(fx) / fx;
}
if(ky != 0)
{
fy = (M_PI * ky) / GRID;
fy = sin(fy) / fy;
}
if(kz != 0)
{
fz = (M_PI * kz) / GRID;
fz = sin(fz) / fz;
}
ff = 1 / (fx * fy * fz);
ff = ff * ff * ff * ff;
ip = GRID * (GRID / 2 + 1) * (y - slabstart_y) + (GRID / 2 + 1) * x + z;
// XXX
// This is only for the non-highres mesh
// Assign the real and complex parts
//
// NOTE: added stretch of 2x to the exp, so that we can sample more low power
//
((fftw_complex *) workspace)[ip].re =
-exp(-k2 * asmth2) * (*GreensFxns[n][m])(All.MassTable[n], All.MassTable[m], k2, 0.0, 1) * fac / k2;
((fftw_complex *) workspace)[ip].im = 0.0;
// fprintf(stderr, "%.15e %.15e\n", sqrt(k2), ((fftw_complex *) workspace)[ip].re);
}
else {
// XXX? It should be zero. In pm_periodic.c, everthing is zeroed before the computation
((fftw_complex *) workspace)[ip].re = 0.0;
((fftw_complex *) workspace)[ip].im = 0.0;
}
}
// KC 11/12/15
// Here we do it with null, so we can use the workspace itself without having to allocate new
// stuff everywhere
rfftwnd_mpi(fft_inverse_plan, 1, workspace, NULL, FFTW_TRANSPOSED_ORDER);
// workspace now contains the non-periodic r-space kernel ** IN THE LOWER OCTANT **
// and at ** HALF RESOLUTION **, addressable as fftw_reals
// DEBUG
fprintf(stderr, "\n# Now the transforms\n");
for(i = slabstart_x; i < (slabstart_x + nslab_x); i++)
for(j = 0; j < GRID; j++)
for(k = 0; k < GRID; k++)
{
x = ((double) i) / GRID;
y = ((double) j) / GRID;
z = ((double) k) / GRID;
if(x >= 0.5)
x -= 1.0;
if(y >= 0.5)
y -= 1.0;
if(z >= 0.5)
z -= 1.0;
r = sqrt(x * x + y * y + z * z);
fprintf(stderr, "%.15e %.15e\n", r, workspace[GRID * GRID2 * (i -slabstart_x) + GRID2 * j + k]);
}
fprintf(stderr, "\n");
endrun(8686);
// ORIGINAL CODE
for(i = slabstart_x; i < (slabstart_x + nslab_x); i++)
for(j = 0; j < GRID; j++)
for(k = 0; k < GRID; k++)
{
x = ((double) i) / GRID;
y = ((double) j) / GRID;
z = ((double) k) / GRID;
if(x >= 0.5)
x -= 1.0;
if(y >= 0.5)
y -= 1.0;
if(z >= 0.5)
z -= 1.0;
r = sqrt(x * x + y * y + z * z);
u = 0.5 * r / (((double) ASMTH) / GRID);
fprintf(stderr, "%.15e\n", u);
if(r > 0) {
// XXX
// Note also that the PotentialFxns had better be hardcoding the masses
// as parameters, or else shortrange_fourier_pot will be mistabulated.
/* kernel[0][n][m][GRID * GRID2 * (i -slabstart_x) + GRID2 * j + k] = */
/* -(*PotentialFxns[n][m])(All.MassTable[n], All.MassTable[m], 0.0, r, 1) */
/* + utorwpi * csplineInterpolate(npFourierTable[n][m], npFourierIntTable[n][m], u); */
}
else if(r > 0) {
// XXX
// It won't quite be dead yet, so this will ring loud and bad
kernel[0][n][m][GRID * GRID2 * (i -slabstart_x) + GRID2 * j + k] = 0.0;
}
else {
// KC 1/25/15
kernel[0][n][m][GRID * GRID2 * (i - slabstart_x) + GRID2 * j + k] = PotentialZero[n][m];
}
}
endrun(8989);
/* do the forward transform of the kernel */
rfftwnd_mpi(fft_forward_plan, 1, kernel[0][n][m], workspace, FFTW_TRANSPOSED_ORDER);
}
#endif
#if defined(PLACEHIGHRESREGION)
for(n = 0; n < N_GRAVS; ++n)
for(m = 0; m < N_GRAVS; ++m) {
for(i = 0; i < fftsize; i++) /* clear local density field */
kernel[1][n][m][i] = 0;
for(i = slabstart_x; i < (slabstart_x + nslab_x); i++)
for(j = 0; j < GRID; j++)
for(k = 0; k < GRID; k++)
{
x = ((double) i) / GRID;
y = ((double) j) / GRID;
z = ((double) k) / GRID;
if(x >= 0.5)
x -= 1.0;
if(y >= 0.5)
y -= 1.0;
if(z >= 0.5)
z -= 1.0;
r = sqrt(x * x + y * y + z * z);
u = 0.5 * r / (((double) ASMTH) / GRID);
// KC 11/1/15
// XXX!!
// See above comments.
// This one is really bad though as the smoothing scales, which change,
// would need to enter the tabulation. No dice.
fac = erfc(u * All.Asmth[1] / All.Asmth[0]) - erfc(u);
if(r > 0) {
kernel[1][n][m][GRID * GRID2 * (i - slabstart_x) + GRID2 * j + k] =
-fac * (*PotentialFxns[n][m])(MassTable[nA], MassTable[nB], 0.0, r, 1);
}
else {
fac = 1 - All.Asmth[1] / All.Asmth[0];
kernel[1][n][m][GRID * GRID2 * (i - slabstart_x) + GRID2 * j + k] =
-fac * PotentialZero[n][m]; /// (sqrt(M_PI) * (((double) ASMTH) / GRID));
}
}
/* do the forward transform of the kernel */
rfftwnd_mpi(fft_forward_plan, 1, kernel[1][n][m], workspace, FFTW_TRANSPOSED_ORDER);
}
#endif
// KC 11/23/14
// This function must not even be called if we are not doing non-periodic styles
/* deconvolve the Greens function twice with the CIC kernel */
for(y = slabstart_y; y < slabstart_y + nslab_y; y++)
for(x = 0; x < GRID; x++)
for(z = 0; z < GRID / 2 + 1; z++)
{
if(x > GRID / 2)
kx = x - GRID;
else
kx = x;
if(y > GRID / 2)
ky = y - GRID;
else
ky = y;
if(z > GRID / 2)
kz = z - GRID;
else
kz = z;
k2 = kx * kx + ky * ky + kz * kz;
if(k2 > 0)
{
fx = fy = fz = 1;
if(kx != 0)
{
fx = (M_PI * kx) / GRID;
fx = sin(fx) / fx;
}
if(ky != 0)
{
fy = (M_PI * ky) / GRID;
fy = sin(fy) / fy;
}
if(kz != 0)
{
fz = (M_PI * kz) / GRID;
fz = sin(fz) / fz;
}
ff = 1 / (fx * fy * fz);
ff = ff * ff * ff * ff;
ip = GRID * (GRID / 2 + 1) * (y - slabstart_y) + (GRID / 2 + 1) * x + z;
for(n = 0; n < N_GRAVS; ++n)
for(m = 0; m < N_GRAVS; ++m) {
#if !defined(PERIODIC)
fft_of_kernel[0][n][m][ip].re *= ff;
fft_of_kernel[0][n][m][ip].im *= ff;
#endif
#if defined(PLACEHIGHRESREGION)
fft_of_kernel[1][n][m][ip].re *= ff;
fft_of_kernel[1][n][m][ip].im *= ff;
#endif
}
}
}
/* end deconvolution */
pm_init_nonperiodic_free();
}
/*! Calculates the long-range non-periodic forces using the PM method. The
* potential is Gaussian filtered with Asmth, given in mesh-cell units. The
* potential is finite differenced using a 4-point finite differencing
* formula to obtain the force fields, which are then interpolated to the
* particle positions. We carry out a CIC charge assignment, and compute the
* potenial by Fourier transform methods. The CIC kernel is deconvolved.
*/
int pmforce_nonperiodic(int grnr)
{
double dx, dy, dz;
double fac, to_slab_fac;
double re, im, acc_dim;
int i, j, slab, level, sendTask, recvTask, flag, flagsum;
int x, y, z, xl, yl, zl, xr, yr, zr, xll, yll, zll, xrr, yrr, zrr, ip, dim;
int slab_x, slab_y, slab_z;
int slab_xx, slab_yy, slab_zz;
int meshmin[3], meshmax[3], sendmin, sendmax, recvmin, recvmax;
int dimx, dimy, dimz, recv_dimx, recv_dimy, recv_dimz;
MPI_Status status;
// ngravs extension
int nA, nB;
int offsets[N_GRAVS + 1];
int k;
if(ThisTask == 0)
printf("Starting non-periodic PM calculation (grid=%d).\n", grnr);
fac = All.G / pow(All.TotalMeshSize[grnr], 4) * pow(All.TotalMeshSize[grnr] / GRID, 3); /* to get potential */
fac *= 1 / (2 * All.TotalMeshSize[grnr] / GRID); /* for finite differencing */
to_slab_fac = GRID / All.TotalMeshSize[grnr];
/* first, establish the extension of the local patch in GRID (for binning) */
for(j = 0; j < 3; j++)
{
meshmin[j] = GRID;
meshmax[j] = 0;
}
// KC 11/7/14
// Compute the gravitational offsets here
offsets[0] = 0;
for(k = 1; k < N_GRAVS; ++k)
offsets[k] = offsets[k-1] + NgravLocal[k-1];
offsets[N_GRAVS] = NumPart;
for(i = 0, flag = 0; i < NumPart; i++)
{
#ifdef PLACEHIGHRESREGION
if(grnr == 0 || (grnr == 1 && ((1 << P[i].Type) & (PLACEHIGHRESREGION))))
#endif
{
for(j = 0; j < 3; j++)
{
if(P[i].Pos[j] < All.Xmintot[grnr][j] || P[i].Pos[j] > All.Xmaxtot[grnr][j])
{
if(flag == 0)
{
printf
("Particle Id=%d on task=%d with coordinates (%g|%g|%g) lies outside PM mesh.\nStopping\n",
(int)P[i].ID, ThisTask, P[i].Pos[0], P[i].Pos[1], P[i].Pos[2]);
fflush(stdout);
}
flag++;
break;
}
}
}
if(flag > 0)
continue;
// KC 11/7/14
// grnr - GridNumberReference ?
if(P[i].Pos[0] >= All.Corner[grnr][0] && P[i].Pos[0] < All.UpperCorner[grnr][0])
if(P[i].Pos[1] >= All.Corner[grnr][1] && P[i].Pos[1] < All.UpperCorner[grnr][1])
if(P[i].Pos[2] >= All.Corner[grnr][2] && P[i].Pos[2] < All.UpperCorner[grnr][2])
{
for(j = 0; j < 3; j++)
{
slab = to_slab_fac * (P[i].Pos[j] - All.Corner[grnr][j]);
if(slab < meshmin[j])
meshmin[j] = slab;
if(slab > meshmax[j])
meshmax[j] = slab;
}
}
}
MPI_Allreduce(&flag, &flagsum, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
if(flagsum > 0)
{
if(ThisTask == 0)
{
printf("In total %d particles were outside allowed range.\n", flagsum);
fflush(stdout);
}
return 1; /* error - need to return because particle were outside allowed range */
}
MPI_Allgather(meshmin, 3, MPI_INT, meshmin_list, 3, MPI_INT, MPI_COMM_WORLD);
MPI_Allgather(meshmax, 3, MPI_INT, meshmax_list, 3, MPI_INT, MPI_COMM_WORLD);
dimx = meshmax[0] - meshmin[0] + 2;
dimy = meshmax[1] - meshmin[1] + 2;
dimz = meshmax[2] - meshmin[2] + 2;
force_treefree();
pm_init_nonperiodic_allocate((dimx + 4) * (dimy + 4) * (dimz + 4));
// Sources
for(nA = 0; nA < N_GRAVS; ++nA) {
// Receivers
for(nB = 0; nB < N_GRAVS; ++nB) {
for(i = 0; i < dimx * dimy * dimz; i++)
workspace[i] = 0;
for(i = offsets[nA]; i < offsets[nA+1]; ++i) {
if(P[i].Pos[0] < All.Corner[grnr][0] || P[i].Pos[0] >= All.UpperCorner[grnr][0])
continue;
if(P[i].Pos[1] < All.Corner[grnr][1] || P[i].Pos[1] >= All.UpperCorner[grnr][1])
continue;
if(P[i].Pos[2] < All.Corner[grnr][2] || P[i].Pos[2] >= All.UpperCorner[grnr][2])
continue;
slab_x = to_slab_fac * (P[i].Pos[0] - All.Corner[grnr][0]);
dx = to_slab_fac * (P[i].Pos[0] - All.Corner[grnr][0]) - slab_x;
slab_x -= meshmin[0];
slab_xx = slab_x + 1;
slab_y = to_slab_fac * (P[i].Pos[1] - All.Corner[grnr][1]);
dy = to_slab_fac * (P[i].Pos[1] - All.Corner[grnr][1]) - slab_y;
slab_y -= meshmin[1];
slab_yy = slab_y + 1;
slab_z = to_slab_fac * (P[i].Pos[2] - All.Corner[grnr][2]);
dz = to_slab_fac * (P[i].Pos[2] - All.Corner[grnr][2]) - slab_z;
slab_z -= meshmin[2];
slab_zz = slab_z + 1;
workspace[(slab_x * dimy + slab_y) * dimz + slab_z] += P[i].Mass * (1.0 - dx) * (1.0 - dy) * (1.0 - dz);
workspace[(slab_x * dimy + slab_yy) * dimz + slab_z] += P[i].Mass * (1.0 - dx) * dy * (1.0 - dz);
workspace[(slab_x * dimy + slab_y) * dimz + slab_zz] += P[i].Mass * (1.0 - dx) * (1.0 - dy) * dz;
workspace[(slab_x * dimy + slab_yy) * dimz + slab_zz] += P[i].Mass * (1.0 - dx) * dy * dz;
workspace[(slab_xx * dimy + slab_y) * dimz + slab_z] += P[i].Mass * (dx) * (1.0 - dy) * (1.0 - dz);
workspace[(slab_xx * dimy + slab_yy) * dimz + slab_z] += P[i].Mass * (dx) * dy * (1.0 - dz);
workspace[(slab_xx * dimy + slab_y) * dimz + slab_zz] += P[i].Mass * (dx) * (1.0 - dy) * dz;
workspace[(slab_xx * dimy + slab_yy) * dimz + slab_zz] += P[i].Mass * (dx) * dy * dz;
}
for(i = 0; i < fftsize; i++) /* clear local density field */
rhogrid[i] = 0;
for(level = 0; level < (1 << PTask); level++) /* note: for level=0, target is the same task */
{
sendTask = ThisTask;
recvTask = ThisTask ^ level;
if(recvTask < NTask)
{
/* check how much we have to send */
sendmin = 2 * GRID;
sendmax = -1;
for(slab_x = meshmin[0]; slab_x < meshmax[0] + 2; slab_x++)
if(slab_to_task[slab_x] == recvTask)
{
if(slab_x < sendmin)
sendmin = slab_x;
if(slab_x > sendmax)
sendmax = slab_x;
}
if(sendmax == -1)
sendmin = 0;
/* check how much we have to receive */
recvmin = 2 * GRID;
recvmax = -1;
for(slab_x = meshmin_list[3 * recvTask]; slab_x < meshmax_list[3 * recvTask] + 2; slab_x++)
if(slab_to_task[slab_x] == sendTask)
{
if(slab_x < recvmin)
recvmin = slab_x;
if(slab_x > recvmax)
recvmax = slab_x;
}
if(recvmax == -1)
recvmin = 0;
if((recvmax - recvmin) >= 0 || (sendmax - sendmin) >= 0) /* ok, we have a contribution to the slab */
{
recv_dimx = meshmax_list[3 * recvTask + 0] - meshmin_list[3 * recvTask + 0] + 2;
recv_dimy = meshmax_list[3 * recvTask + 1] - meshmin_list[3 * recvTask + 1] + 2;
recv_dimz = meshmax_list[3 * recvTask + 2] - meshmin_list[3 * recvTask + 2] + 2;
if(level > 0)
{
MPI_Sendrecv(workspace + (sendmin - meshmin[0]) * dimy * dimz,
(sendmax - sendmin + 1) * dimy * dimz * sizeof(fftw_real), MPI_BYTE, recvTask,
TAG_NONPERIOD_A | (nA << 6) | (nB << 9), forcegrid,
(recvmax - recvmin + 1) * recv_dimy * recv_dimz * sizeof(fftw_real), MPI_BYTE,
recvTask, TAG_NONPERIOD_A | (nA << 6) | (nB << 9), MPI_COMM_WORLD, &status);
}
else
{
memcpy(forcegrid, workspace + (sendmin - meshmin[0]) * dimy * dimz,
(sendmax - sendmin + 1) * dimy * dimz * sizeof(fftw_real));
}
for(slab_x = recvmin; slab_x <= recvmax; slab_x++)
{
slab_xx = slab_x - first_slab_of_task[ThisTask];
if(slab_xx >= 0 && slab_xx < slabs_per_task[ThisTask])
{
for(slab_y = meshmin_list[3 * recvTask + 1];
slab_y <= meshmax_list[3 * recvTask + 1] + 1; slab_y++)
{
slab_yy = slab_y;
for(slab_z = meshmin_list[3 * recvTask + 2];
slab_z <= meshmax_list[3 * recvTask + 2] + 1; slab_z++)
{
slab_zz = slab_z;
rhogrid[GRID * GRID2 * slab_xx + GRID2 * slab_yy + slab_zz] +=
forcegrid[((slab_x - recvmin) * recv_dimy +
(slab_y - meshmin_list[3 * recvTask + 1])) * recv_dimz +
(slab_z - meshmin_list[3 * recvTask + 2])];
}
}
}
}
}
}
}
/* Do the FFT of the density field */
rfftwnd_mpi(fft_forward_plan, 1, rhogrid, workspace, FFTW_TRANSPOSED_ORDER);
/* multiply with the Fourier transform of the Green's function (kernel) */
for(y = 0; y < nslab_y; y++)
for(x = 0; x < GRID; x++)
for(z = 0; z < GRID / 2 + 1; z++)
{
ip = GRID * (GRID / 2 + 1) * y + (GRID / 2 + 1) * x + z;
re =
fft_of_rhogrid[ip].re * fft_of_kernel[grnr][nA][nB][ip].re -
fft_of_rhogrid[ip].im * fft_of_kernel[grnr][nA][nB][ip].im;
im =
fft_of_rhogrid[ip].re * fft_of_kernel[grnr][nA][nB][ip].im +
fft_of_rhogrid[ip].im * fft_of_kernel[grnr][nA][nB][ip].re;
fft_of_rhogrid[ip].re = re;
fft_of_rhogrid[ip].im = im;
}
/* get the potential by inverse FFT */
rfftwnd_mpi(fft_inverse_plan, 1, rhogrid, workspace, FFTW_TRANSPOSED_ORDER);
/* Now rhogrid holds the potential */
/* construct the potential for the local patch */
/* if we have a high-res mesh, establish the extension of the local patch in GRID (for reading out the
* forces)
*/
// KC 11/7/14
// This code is baffling. Because exactly the same meshmin/max and meshlists will
// be computed (on all the processors?) as were computed the first time this
// same code was executed when we checked for grid OOB problems. Here we
// test for grnr == 1 and consider *only* contributions from the high-resolution type.
// These are the same conditions above (because grnr will never change within any one
// run of this method.) I am commenting this out and we will see.
/* #ifdef PLACEHIGHRESREGION */
/* if(grnr == 1) */
/* { */
/* for(j = 0; j < 3; j++) */
/* { */
/* meshmin[j] = GRID; */
/* meshmax[j] = 0; */
/* } */
/* for(i = 0; i < NumPart; i++) */
/* { */
/* if(!((1 << P[i].Type) & (PLACEHIGHRESREGION))) */
/* continue; */
/* if(P[i].Pos[0] >= All.Corner[grnr][0] && P[i].Pos[0] < All.UpperCorner[grnr][0]) */
/* if(P[i].Pos[1] >= All.Corner[grnr][1] && P[i].Pos[1] < All.UpperCorner[grnr][1]) */
/* if(P[i].Pos[2] >= All.Corner[grnr][2] && P[i].Pos[2] < All.UpperCorner[grnr][2]) */
/* { */
/* for(j = 0; j < 3; j++) */
/* { */
/* slab = to_slab_fac * (P[i].Pos[j] - All.Corner[grnr][j]); */
/* if(slab < meshmin[j]) */
/* meshmin[j] = slab; */
/* if(slab > meshmax[j]) */
/* meshmax[j] = slab; */
/* } */
/* } */
/* } */
/* MPI_Allgather(meshmin, 3, MPI_INT, meshmin_list, 3, MPI_INT, MPI_COMM_WORLD); */
/* MPI_Allgather(meshmax, 3, MPI_INT, meshmax_list, 3, MPI_INT, MPI_COMM_WORLD); */
/* } */
/* #endif */
// /XXX? Resume porting...
dimx = meshmax[0] - meshmin[0] + 6;
dimy = meshmax[1] - meshmin[1] + 6;
dimz = meshmax[2] - meshmin[2] + 6;
for(j = 0; j < 3; j++)
{
if(meshmin[j] < 2)
endrun(131231);
if(meshmax[j] > GRID / 2 - 3)
endrun(131288);
}
for(level = 0; level < (1 << PTask); level++) /* note: for level=0, target is the same task */
{
sendTask = ThisTask;
recvTask = ThisTask ^ level;
if(recvTask < NTask)
{
/* check how much we have to send */
sendmin = 2 * GRID;
sendmax = -GRID;
for(slab_x = meshmin_list[3 * recvTask] - 2; slab_x < meshmax_list[3 * recvTask] + 4; slab_x++)