/
lbm.c
336 lines (288 loc) · 10.7 KB
/
lbm.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
// mpicc lbm.c -std=gnu11 -o lbm_out -O3
#include <stdio.h>
#include <stdlib.h> // malloc
#include <string.h> //memcpy
#include <math.h>
/* Use MPI */
#include "mpi.h"
#define RED 0
#define BLACK 1
#define K 9 // number of velocity directions
#define IDX_YX(y, x) ((int) (y*M + x))
#define IDX_YXI(y, x, i) ((int) (i*M*M + y*M + x))
#define GET_MACRO(_1,_2,_3,NAME,...) NAME
#define idx(...) GET_MACRO(__VA_ARGS__, IDX_YXI, IDX_YX)(__VA_ARGS__)
#define OUTPUT_FILE "csim_data.bin"
#define dprintf if(DEBUG)printf
void write_data_to_row(double* A, double* data, int y, int i, int M) {
for (int x = 0; x < M; x++) {
A[idx(y, x, i)] = data[x];
}
}
void write_data_to_col(double* A, double* data, int x, int i, int M) {
for (int y = 0; y < M; y++) {
A[idx(y, x, i)] = data[y];
}
}
void write_row_to_data(double* A, double* data, int y, int i, int M) {
for (int x = 0; x < M; x++) {
data[x] = A[idx(y, x, i)];
}
}
void write_col_to_data(double* A, double* data, int x, int i, int M) {
for (int y = 0; y < M; y++) {
data[y] = A[idx(y, x, i)];
}
}
int cyclic_shift(int px, int shift, int sqrtP) {
px = (px + shift) % sqrtP;
if (px < 0) px += sqrtP;
return px;
}
int get_p(int py, int px, int dy, int dx, int sqrtP) {
return cyclic_shift(py, dy, sqrtP) * sqrtP + cyclic_shift(px, dx, sqrtP);
}
int main(int argc, char *argv[]) {
MPI_Status status;
int ghost_tag = 100;
int print_tag = 200;
int P, p;
/* Initialize MPI */
MPI_Init(&argc, &argv);
MPI_Comm_size(MPI_COMM_WORLD, &P);
MPI_Comm_rank(MPI_COMM_WORLD, &p);
if (p == 0) {
if (argc < 2) printf("Missing length L (position 2)!\n" );
if (argc < 3) printf("Missing time T (position 3)!\n" );
if (argc < 4) printf("Missing strength G (position 4)!\n" );
if (argc < 5) printf("Missing DEBUG flag (position 5)!\n" );
if (argc < 6) printf("Missing SAVE flag (position 6)!\n" );
if (argc < 6) exit(1);
}
int L = atoi(argv[1]);
int T = atoi(argv[2]);
double G = atof(argv[3]);
int DEBUG = atoi(argv[4]);
int SAVE = atoi(argv[5]);
if (p == 0) {
printf("L = %d\n", L);
printf("T = %d\n", T);
printf("G = %f\n", G);
printf("DEBUG = %d\n", DEBUG);
printf("SAVE = %d\n", SAVE);
}
if (floor(sqrt(P)) != sqrt(P)) {
fprintf(stdout, "Number of processors must be a square number.\n");
exit(1);
}
int sqrtP = (int) sqrt(P);
if (L % sqrtP != 0) {
fprintf(stdout, "sqrt(P) must evenly divide L.\n");
exit(1);
}
int px = p % sqrtP;
int py = (p - px) / sqrtP;
//dprintf("Indx: p:%d px:%d py:%d sqrtP:%d\n", p, px, py, sqrtP);
int colorx = px % 2 == 0 ? RED : BLACK;
int colory = py % 2 == 0 ? RED : BLACK;
int M = L/sqrtP + 2;
double ex[] = {0, 1, 0, -1, 0, 1, -1, -1, 1};
double ey[] = {0, 0, 1, 0, -1, 1, 1, -1, -1};
double w[] = {4.0/9, 1.0/9, 1.0/9, 1.0/9, 1.0/9, 1.0/36, 1.0/36, 1.0/36, 1.0/36};
int updown[] = {2, 5, 6, 4, 7, 8};
int rightleft[] = {1, 5, 8, 3, 6, 7};
int corners[] = {5, 6, 7, 8};
double* f = (double *) malloc(M*M*K*sizeof(double));
double* rho = (double *) malloc(M*M*sizeof(double));
double* ux = (double *) malloc(M*M*sizeof(double));
double* uy = (double *) malloc(M*M*sizeof(double));
double* psi = (double *) malloc(M*M*sizeof(double));
double* psisend = (double *) malloc(M*sizeof(double));
double* psirecv = (double *) malloc(M*sizeof(double));
double* fsend = (double *) malloc(M*3*sizeof(double));
double* frecv = (double *) malloc(M*3*sizeof(double));
srandom(p);
for (int x = 1; x < M-1; x++) {
for (int y = 1; y < M-1; y++) {
double r = ((double) random())/RAND_MAX;
double initial_rho = 2.0 + r/100.0;
for (int i = 0; i < K; i++) {
f[idx(y, x, i)] = w[i] * initial_rho;
}
}
}
if (SAVE && p == 0) {
char write_mode = 'w';
FILE *file = fopen(OUTPUT_FILE, &write_mode);
fwrite(&L, sizeof(int), 1, file);
fwrite(&T, sizeof(int), 1, file);
fwrite(&P, sizeof(int), 1, file);
fclose (file);
}
for (int t = 0; t < T; t++) {
for (int x = 1; x < M-1; x++) {
for (int y = 1; y < M-1; y++) {
rho[idx(y, x)] = 0;
ux[idx(y, x)] = 0;
uy[idx(y, x)] = 0;
for (int i = 0; i < K; i++) {
rho[idx(y, x)] += f[idx(y, x, i)];
ux[idx(y, x)] += f[idx(y, x, i)] * ex[i];
uy[idx(y, x)] += f[idx(y, x, i)] * ey[i];
}
ux[idx(y, x)] /= rho[idx(y, x)];
uy[idx(y, x)] /= rho[idx(y, x)];
psi[idx(y, x)] = 2*exp(-2/rho[idx(y, x)]);
}
}
if (SAVE && t % 10 == 0) {
char dummy;
if (p > 0) {
MPI_Recv(&dummy, 1, MPI_BYTE, p-1, print_tag, MPI_COMM_WORLD, &status);
}
char append_mode = 'a';
FILE *file = fopen(OUTPUT_FILE, &append_mode);
fwrite(rho, sizeof(double), M*M, file);
fclose (file);
if (p < P-1) {
MPI_Send(&dummy, 1, MPI_BYTE, p+1, print_tag, MPI_COMM_WORLD);
}
}
// up-down
for (int d = 0; d < 2; d++) {
int dir = ey[updown[d*3]];
if (colory == RED) {
int send_row_idx = dir < 0 ? 1 : M-2;
int recv_row_idx = dir < 0 ? 0 : M-1;
int partner = get_p(py, px, -dir, 0, sqrtP);
write_row_to_data(psi, psisend, send_row_idx, 0, M);
MPI_Send(psisend, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
MPI_Recv(psirecv, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
write_data_to_row(psi, psirecv, recv_row_idx, 0, M);
} else {
int send_row_idx = dir < 0 ? M-2 : 1;
int recv_row_idx = dir < 0 ? M-1 : 0;
int partner = get_p(py, px, dir, 0, sqrtP);
write_row_to_data(psi, psisend, send_row_idx, 0, M);
MPI_Recv(psirecv, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
MPI_Send(psisend, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
write_data_to_row(psi, psirecv, recv_row_idx, 0, M);
}
}
// right-left
for (int d = 0; d < 2; d++) {
int dir = ex[rightleft[d*3]];
if (colorx == RED) {
int send_col_idx = dir < 0 ? 1 : M-2;
int recv_col_idx = dir < 0 ? 0 : M-1;
int partner = get_p(py, px, 0, dir, sqrtP);
write_col_to_data(psi, psisend, send_col_idx, 0, M);
MPI_Send(psisend, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
MPI_Recv(psirecv, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
write_data_to_col(psi, psirecv, recv_col_idx, 0, M);
} else {
int send_col_idx = dir < 0 ? M-2 : 1;
int recv_col_idx = dir < 0 ? M-1 : 0;
int partner = get_p(py, px, 0, -dir, sqrtP);
write_col_to_data(psi, psisend, send_col_idx, 0, M);
MPI_Recv(psirecv, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
MPI_Send(psisend, M, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
write_data_to_col(psi, psirecv, recv_col_idx, 0, M);
}
}
// corners
for (int d = 0; d < 4; d++) {
int dirx = ex[corners[d]];
int diry = ey[corners[d]];
if (colorx == RED) {
int send_x = dirx < 0 ? 1 : M-2;
int recv_x = dirx < 0 ? 0 : M-1;
int send_y = diry < 0 ? 1 : M-2;
int recv_y = diry < 0 ? 0 : M-1;
int partner = get_p(py, px, -diry, dirx, sqrtP);
MPI_Send(psi + idx(send_y, send_x), 1, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
MPI_Recv(psi + idx(recv_y, recv_x), 1, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
} else {
int send_x = dirx < 0 ? M-2 : 1;
int recv_x = dirx < 0 ? M-1 : 0;
int send_y = diry < 0 ? M-2 : 1;
int recv_y = diry < 0 ? M-1 : 0;
int partner = get_p(py, px, diry, -dirx, sqrtP);
MPI_Recv(psi + idx(recv_y, recv_x), 1, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
MPI_Send(psi + idx(send_y, send_x), 1, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
}
}
for (int x = 1; x < M-1; x++) {
for (int y = 1; y < M-1; y++) {
for (int i = 0; i < K; i++) {
int neighbor_x = x - (int)ex[i];
int neighbor_y = y - (int)ey[i];
double Fxi = G * w[i] * psi[idx(y, x)] * psi[idx(neighbor_y, neighbor_x)] * ex[i];
double Fyi = G * w[i] * psi[idx(y, x)] * psi[idx(neighbor_y, neighbor_x)] * ey[i];
ux[idx(y, x)] = ux[idx(y, x)] + Fxi / rho[idx(y, x)];
uy[idx(y, x)] = uy[idx(y, x)] + Fyi / rho[idx(y, x)];
}
for (int i = 0; i < K; i++) {
double edotu = ux[idx(y, x)] * ex[i] + uy[idx(y, x)] * ey[i];
double udotu = ux[idx(y, x)] * ux[idx(y, x)] + uy[idx(y, x)] * uy[idx(y, x)];
int stream_x = x + (int)ex[i];
int stream_y = y + (int)ey[i];
f[idx(stream_y, stream_x, i)] = w[i] * rho[idx(y, x)] * (1 + 3*edotu + 4.5*edotu*edotu - 1.5*udotu);
}
}
}
// up-down
for (int d = 0; d < 2; d++) {
int dir = ey[updown[d*3]];
if (colory == RED) {
int send_row_idx = dir < 0 ? 0 : M-1;
int recv_row_idx = dir < 0 ? 1 : M-2;
int partner = get_p(py, px, -dir, 0, sqrtP);
for (int j = 0; j < 3; j++)
write_row_to_data(f, fsend+M*j, send_row_idx, updown[j+d*3], M);
MPI_Send(fsend, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
MPI_Recv(frecv, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
for (int j = 0; j < 3; j++)
write_data_to_row(f, frecv+M*j, recv_row_idx, updown[j+(1-d)*3], M);
} else {
int send_row_idx = dir < 0 ? M-1 : 0;
int recv_row_idx = dir < 0 ? M-2 : 1;
int partner = get_p(py, px, dir, 0, sqrtP);
for (int j = 0; j < 3; j++)
write_row_to_data(f, fsend+M*j, send_row_idx, updown[j+(1-d)*3], M);
MPI_Recv(frecv, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
MPI_Send(fsend, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
for (int j = 0; j < 3; j++)
write_data_to_row(f, frecv+M*j, recv_row_idx, updown[j+d*3], M);
}
}
// right-left
for (int d = 0; d < 2; d++) {
int dir = ex[rightleft[d*3]];
if (colorx == RED) {
int send_col_idx = dir < 0 ? 0 : M-1;
int recv_col_idx = dir < 0 ? 1 : M-2;
int partner = get_p(py, px, 0, dir, sqrtP);
for (int j = 0; j < 3; j++)
write_col_to_data(f, fsend+M*j, send_col_idx, rightleft[j+d*3], M);
MPI_Send(fsend, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
MPI_Recv(frecv, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
for (int j = 0; j < 3; j++)
write_data_to_col(f, frecv+M*j, recv_col_idx, rightleft[j+(1-d)*3], M);
} else {
int send_col_idx = dir < 0 ? M-1 : 0;
int recv_col_idx = dir < 0 ? M-2 : 1;
int partner = get_p(py, px, 0, -dir, sqrtP);
for (int j = 0; j < 3; j++)
write_col_to_data(f, fsend+M*j, send_col_idx, rightleft[j+(1-d)*3], M);
MPI_Recv(frecv, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD, &status);
MPI_Send(fsend, M*3, MPI_DOUBLE, partner, ghost_tag, MPI_COMM_WORLD);
for (int j = 0; j < 3; j++)
write_data_to_col(f, frecv+M*j, recv_col_idx, rightleft[j+d*3], M);
}
}
dprintf("%d\n", t);
}
MPI_Finalize();
return(0);
}