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
|
<?xml version="1.0" encoding="UTF-8" ?>
<!DOCTYPE html
PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html" />
<title>How to compile SFMT</title>
<style type="text/css">
BLOCKQUOTE {background-color:#a0ffa0;
padding-left: 1em;}
</style>
</head>
<body>
<h2> How to compile SFMT </h2>
<p>
This document explains how to compile SFMT for users who
are using UNIX like systems (for example Linux, Free BSD,
cygwin, osx, etc) on terminal. I can't help those who use IDE
(Integrated Development Environment,) please see your IDE's help
to use SIMD feature of your CPU.
</p>
<h3>1. First Step: Compile test programs using Makefile.</h3>
<h4>1-1. Compile standard C test program.</h4>
<p>
Check if SFMT.c and Makefile are in your current directory.
If not, <strong>cd</strong> to the directory where they exist.
Then, type
</p>
<blockquote>
<pre>make std</pre>
</blockquote>
<p>
If it causes an error, try to type
</p>
<blockquote>
<pre>cc -DSFMT_MEXP=19937 -o test-std-M19937 test.c SFMT.c</pre>
</blockquote>
<p>
or try to type
</p>
<blockquote>
<pre>gcc -DSFMT_MEXP=19937 -o test-std-M19937 test.c SFMT.c</pre>
</blockquote>
<p>
If success, then check the test program. Type
</p>
<blockquote>
<pre>./test-std-M19937 -b32</pre>
</blockquote>
<p>
You will see many random numbers displayed on your screen.
If you want to check these random numbers are correct output,
redirect output to a file and <strong>diff</strong> it with
<strong>SFMT.19937.out.txt</strong>, like this:</p>
<blockquote>
<pre>./test-std-M19937 -b32 > foo.txt
diff -w foo.txt SFMT.19937.out.txt</pre>
</blockquote>
<p>
Silence means they are the same because <strong>diff</strong>
reports the difference of two file.
</p>
<p>
If you want to know the generation speed of SFMT, type
</p>
<blockquote>
<pre>./test-std-M19937 -s</pre>
</blockquote>
<p>
It is very slow. To make it fast, compile it
with <strong>-O3</strong> option. If your compiler is gcc, you
should specify <strong>-fno-strict-aliasing</strong> option
with <strong>-O3</strong>. type
</p>
<blockquote>
<pre>gcc -O3 -fno-strict-aliasing -DSFMT_MEXP=19937 -o test-std-M19937 test.c SFMT.c
./test-std-M19937 -s</pre>
</blockquote>
<h4>1-2. Compile SSE2 test program.</h4>
<p>
If your CPU supports SSE2 and you can use gcc version 3.4 or later,
you can make test-sse2-Mxxx. To do this, type
</p>
<blockquote>
<pre>make sse2</pre>
</blockquote>
<p>or type</p>
<blockquote>
<pre>gcc -O3 -msse2 -fno-strict-aliasing -DHAVE_SSE2=1 -DSFMT_MEXP=19937 -o test-sse2-M19937 test.c SFMT.c</pre>
</blockquote>
<p>If everything works well,</p>
<blockquote>
<pre>./test-sse2-M19937 -s</pre>
</blockquote>
<p>will show much shorter time than <strong>test-std-M19937 -s</strong>.</p>
<!--h4>1-3. Compile AltiVec test program.</h4>
<p>
If you are using Macintosh computer with PowerPC G4 or G5, and
your gcc version is later 3.3, you can make test-alti-M19937. To
do this, type
</p>
<blockquote>
<pre>make osx-alti</pre>
</blockquote>
<p>or type</p>
<blockquote>
<pre>gcc -O3 -faltivec -fno-strict-aliasing -DHAVE_ALTIVEC=1 -DSFMT_MEXP=19937 -o test-alti-M19937 test.c</pre>
</blockquote>
<p>If everything works well,</p>
<blockquote>
<pre>./test-alti-M19937 -s</pre>
</blockquote>
<p>shows much shorter time than <strong>test-std-M19937 -s</strong>.</p>
<p>If you are using a CPU which supports AltiVec under Linux, use
<strong>alti</strong> instead of <strong>osx-alti</strong>.</p-->
<h4>1-4. Compile and check output automatically.</h4>
<p>
To make test program and check 32-bit output
automatically for all supported MEXPs of SFMT, type
</p>
<blockquote>
<pre>make std-check</pre>
</blockquote>
<!--p>
To check test program optimized for 64bit output of big endian CPU, type
</p>
<blockquote>
<pre>make big-check</pre>
</blockquote-->
<p>
To check test program optimized for SSE2, type
</p>
<blockquote>
<pre>make sse2-check</pre>
</blockquote>
<!--p>
To check test program optimized for OSX AltiVec, type
</p>
<blockquote>
<pre>make osx-alti-check</pre>
</blockquote>
<p>
To check test program optimized for OSX AltiVec and 64bit output, type
</p>
<blockquote>
<pre>make osx-altibig-check</pre>
</blockquote-->
<p>
These commands may take some time.
</p>
<h3>2. Second Step: Use SFMT pseudorandom number generator with
your C program.</h3>
<h4>2-1. Use sequential call and static link.</h4>
<p>
Here is a very simple program <strong>sample1.c</strong> which
calculates PI using Monte-Carlo method.
</p>
<blockquote>
<pre>
#include <stdio.h>
#include <stdlib.h>
#include "SFMT.h"
int main(int argc, char* argv[]) {
int i, cnt, seed;
double x, y, pi;
const int NUM = 10000;
sfmt_t sfmt;
if (argc >= 2) {
seed = strtol(argv[1], NULL, 10);
} else {
seed = 12345;
}
cnt = 0;
sfmt_init_gen_rand(&sfmt, seed);
for (i = 0; i < NUM; i++) {
x = sfmt_genrand_res53(&sfmt);
y = sfmt_genrand_res53(&sfmt);
if (x * x + y * y < 1.0) {
cnt++;
}
}
pi = (double)cnt / NUM * 4;
printf("%lf\n", pi);
return 0;
}
</pre>
</blockquote>
<p>To compile <strong>sample1.c</strong> with SFMT.c with the period of
2<sup>607</sup>, type</p>
<blockquote>
<pre>gcc -O3 -DSFMT_MEXP=607 -o sample1 SFMT.c sample1.c</pre>
</blockquote>
<!--p>If your CPU is BIG ENDIAN you need to type</p>
<blockquote>
<pre>gcc -DSFMT_MEXP=607 -DBIG_ENDIAN64 -o sample1 SFMT.c sample1.c</pre>
</blockquote>
<p>because genrand_res53() uses gen_rand64().</p-->
<p>If your CPU supports SSE2 and you want to use optimized SFMT for
SSE2, type</p>
<blockquote>
<pre>gcc -O3 -msse2 -DHAVE_SSE2 -DSFMT_MEXP=607 -o sample1 SFMT.c sample1.c</pre>
</blockquote>
<!--p>If your CPU supports AltiVec and you want to use optimized SFMT
for AltiVec, type</p>
<blockquote>
<pre>gcc -faltivec -DBIG_ENDIAN64 -DHAVE_ALTIVEC -DSFMT_MEXP=607 -o sample1 SFMT.c sample1.c</pre>
</blockquote-->
<h4>2-2. Use block call and static link.</h4>
<p>
Here is <strong>sample2.c</strong> which modifies sample1.c.
The block call <strong>fill_array64</strong> is much faster than
sequential call, but it needs an aligned memory. The standard function
to get an aligned memory is <strong>posix_memalign</strong>, but
it isn't usable in every OS.
</p>
<blockquote>
<pre>
#include <stdio.h>
#define _XOPEN_SOURCE 600
#include <stdlib.h>
#include "SFMT.h"
int main(int argc, char* argv[]) {
int i, j, cnt, seed;
double x, y, pi;
const int NUM = 10000;
const int R_SIZE = 2 * NUM;
int size;
uint64_t *array;
sfmt_t sfmt;
if (argc >= 2) {
seed = strtol(argv[1], NULL, 10);
} else {
seed = 12345;
}
size = sfmt_get_min_array_size64(&sfmt);
if (size < R_SIZE) {
size = R_SIZE;
}
#if defined(__APPLE__) || \
(defined(__FreeBSD__) && __FreeBSD__ >= 3 && __FreeBSD__ <= 6)
printf("malloc used\n");
array = malloc(sizeof(double) * size);
if (array == NULL) {
printf("can't allocate memory.\n");
return 1;
}
#elif defined(_POSIX_C_SOURCE)
printf("posix_memalign used\n");
if (posix_memalign((void **)&array, 16, sizeof(double) * size) != 0) {
printf("can't allocate memory.\n");
return 1;
}
#elif defined(__GNUC__) && (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 3))
printf("memalign used\n");
array = memalign(16, sizeof(double) * size);
if (array == NULL) {
printf("can't allocate memory.\n");
return 1;
}
#else /* in this case, gcc doesn't support SSE2 */
printf("malloc used\n");
array = malloc(sizeof(double) * size);
if (array == NULL) {
printf("can't allocate memory.\n");
return 1;
}
#endif
cnt = 0;
j = 0;
sfmt_init_gen_rand(&sfmt, seed);
sfmt_fill_array64(&sfmt, array, size);
for (i = 0; i < NUM; i++) {
x = sfmt_to_res53(array[j++]);
y = sfmt_to_res53(array[j++]);
if (x * x + y * y < 1.0) {
cnt++;
}
}
free(array);
pi = (double)cnt / NUM * 4;
printf("%lf\n", pi);
return 0;
}
</pre>
</blockquote>
<p>To compile <strong>sample2.c</strong> with SFMT.c with the period of
2<sup>2281</sup>, type</p>
<blockquote>
<pre>gcc -O3 -DSFMT_MEXP=2281 -o sample2 SFMT.c sample2.c</pre>
</blockquote>
<!--p>or </p>
<blockquote>
<pre>gcc -DSFMT_MEXP=2281 -DBIG_ENDIAN64 -o sample2 SFMT.c sample2.c</pre>
</blockquote -->
<p>If your CPU supports SSE2 and you want to use optimized SFMT for
SSE2, type</p>
<blockquote>
<pre>gcc -O3 -msse2 -DHAVE_SSE2 -DSFMT_MEXP=2281 -o sample2 SFMT.c sample2.c</pre>
</blockquote>
<!--p>If your CPU supports AltiVec and you want to use optimized SFMT
for AltiVec, type</p>
<blockquote>
<pre>gcc -faltivec -DHAVE_ALTIVEC -DSFMT_MEXP=2281 -DBIG_ENDIAN64 -o sample2 SFMT.c sample2.c</pre>
</blockquote>
<p>or type</p>
<blockquote>
<pre>gcc -faltivec -DHAVE_ALTIVEC -DBIG_ENDIAN64 -DONLY64 -DSFMT_MEXP=2281 -o sample2 SFMT.c sample2.c</pre>
</blockquote>
<p>The effect of the option -DONLY64 is:
When -DONLY64 option is used, the executive file can generate
64-bit integers faster but 32-bit output is not supported.
</p-->
<!--h4>2-3. Use sequential call and inline functions.</h4>
<p>
Here is <strong>sample3.c</strong> which modifies sample1.c.
This is very similar to sample1.c. The difference is only one line.
It include <strong>"SFMT.c"</strong> instead of <strong>"SFMT.h"
</strong>.
</p>
<blockquote>
<pre>
#include <stdio.h>
#include <stdlib.h>
#include "SFMT.c"
int main(int argc, char* argv[]) {
int i, cnt, seed;
double x, y, pi;
const int NUM = 10000;
if (argc >= 2) {
seed = strtol(argv[1], NULL, 10);
} else {
seed = 12345;
}
cnt = 0;
init_gen_rand(seed);
for (i = 0; i < NUM; i++) {
x = genrand_res53();
y = genrand_res53();
if (x * x + y * y < 1.0) {
cnt++;
}
}
pi = (double)cnt / NUM * 4;
printf("%lf\n", pi);
return 0;
}
</pre>
</blockquote>
<p>To compile <strong>sample3.c</strong>, type</p>
<blockquote>
<pre>gcc -DSFMT_MEXP=1279 -o sample3 sample3.c</pre>
</blockquote>
<p> or </p>
<blockquote>
<pre>gcc -DSFMT_MEXP=1279 -DBIG_ENDIAN64 -o sample3 sample3.c</pre>
</blockquote>
<p>If your CPU supports SSE2 and you want to use optimized SFMT for
SSE2, then type</p>
<blockquote>
<pre>gcc -msse2 -DHAVE_SSE2 -DSFMT_MEXP=1279 -o sample3 sample3.c</pre>
</blockquote>
<p>If your CPU supports AltiVec and you want to use optimized SFMT
for AltiVec, type</p>
<blockquote>
<pre>gcc -faltivec -DHAVE_ALTIVEC -DBIG_ENDIAN64 -DSFMT_MEXP=1279 -o sample3 sample3.c</pre>
</blockquote>
<p>or type</p>
<blockquote>
<pre>gcc -faltivec -DHAVE_ALTIVEC -DBIG_ENDIAN64 -DONLY64 -DSFMT_MEXP=1279 -o sample3 sample3.c</pre>
</blockquote-->
<h4>2-4. Initialize SFMT using sfmt_init_by_array function.</h4>
<p>
Here is <strong>sample4.c</strong> which modifies sample1.c.
The 32-bit integer seed can only make 2<sup>32</sup> kinds of
initial state, to avoid this problem, SFMT
provides <strong>sfmt_init_by_array</strong> function. This sample
uses sfmt_init_by_array function which initialize the internal state
array with an array of 32-bit. The size of an array can be
larger than the internal state array and all elements of the
array are used for initialization, but too large array is
wasteful.
</p>
<blockquote>
<pre>
#include <stdio.h>
#include <string.h>
#include "SFMT.h"
int main(int argc, char* argv[]) {
int i, cnt, seed_cnt;
double x, y, pi;
const int NUM = 10000;
uint32_t seeds[100];
sfmt_t sfmt;
if (argc >= 2) {
seed_cnt = 0;
for (i = 0; (i < 100) && (i < strlen(argv[1])); i++) {
seeds[i] = argv[1][i];
seed_cnt++;
}
} else {
seeds[0] = 12345;
seed_cnt = 1;
}
cnt = 0;
sfmt_init_by_array(&sfmt, seeds, seed_cnt);
for (i = 0; i < NUM; i++) {
x = sfmt_genrand_res53(&sfmt);
y = sfmt_genrand_res53(&sfmt);
if (x * x + y * y < 1.0) {
cnt++;
}
}
pi = (double)cnt / NUM * 4;
printf("%lf\n", pi);
return 0;
}
</pre>
</blockquote>
<p>To compile <strong>sample4.c</strong>, type</p>
<blockquote>
<pre>gcc -O3 -DSFMT_MEXP=19937 -o sample4 SFMT.c sample4.c</pre>
</blockquote>
<!--p>or</p>
<blockquote>
<pre>gcc -DSFMT_MEXP=19937 -DBIG_ENDIAN64 -o sample4 SFMT.c sample4.c</pre>
</blockquote-->
<p>Now, seed can be a string. Like this:</p>
<blockquote>
<pre>./sample4 your-full-name</pre>
</blockquote>
<h3>Appendix: C preprocessor definitions</h3>
<p>
Here is a list of C preprocessor definitions that users can
specify to control code generation. These macros must be set
just after -D compiler option.
</p>
<dl>
<dt>SFMT_MEXP</dt>
<dd>This macro is required. This macro means Mersenne exponent
and the period of generated code will be 2<sup>SFMT_MEXP</sup>-1.
SFMT_MEXP must be one of 607, 1279, 2281, 4253, 11213, 19937,
44497, 86243, 132049, 216091.
</dd>
<dt>HAVE_SSE2</dt>
<dd>This is optional. If this macro is specified, optimized code
for SSE2 will be generated.</dd>
<dt>HAVE_ALTIVEC</dt>
<dd>This is optional. If this macro is specified, optimized code
for AltiVec will be generated. This macro automatically turns on
BIG_ENDIAN64 macro. <b>This macro of SFMT ver. 1.4 is not tested
at all.</b></dd>
<dt>BIG_ENDIAN64</dt>
<dd>This macro is required when your CPU is BIG ENDIAN and you
use 64-bit output. If __BIG_ENDIAN__ macro is defined, this macro
is automatically turned on. GCC defines __BIG_ENDIAN__ macro on
BIG ENDIAN CPUs. <b>This macro of SFMT ver. 1.4 is not tested
at all.</b></dd>
<dt>ONLY64</dt>
<dd>This macro is optional. If this macro is specified,
optimized code for 64-bit output for BIG ENDIAN CPUs will be
generated and code for 32-bit output won't be
generated. BIG_ENDIAN64 macro must be specified with this macro
by user or automatically. <b>This macro of SFMT ver. 1.4 is not tested
at all.</b></dd>
</dl>
<table border="1" align="center">
<tr><td></td><td>32-bit output</td><td>LITTLE ENDIAN 64-bit output</td>
<td>BIG ENDIAN 64-bit output</td></tr>
<tr><td>required</td><td>SFMT_MEXP</td><td>SFMT_MEXP</td><td>SFMT_MEXP,
<strong>BIG_ENDIAN64</strong></td></tr>
<tr><td>optional</td><td>HAVE_SSE2,
HAVE_ALTIVEC</td><td>HAVE_SSE2</td><td>HAVE_ALTIVEC, ONLY64</td>
</tr>
</table>
</body>
</html>
|
