-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathavx_sort_network.cpp
More file actions
387 lines (267 loc) · 11.1 KB
/
Copy pathavx_sort_network.cpp
File metadata and controls
387 lines (267 loc) · 11.1 KB
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
#include <stdio.h>
#include <stdlib.h>
#include <stdint.h>
#include <stdbool.h>
#include <sys/time.h>
#include <memory.h>
#include <math.h>
#include <immintrin.h>
#include <xmmintrin.h>
#include <algorithm>
#include <iostream>
using namespace std;
//http://parallelcomp.uw.hu/ch09lev1sec2.html
//float x[] = {10.f, 5.0f, 3.0f, 12.0f, 90.0f, 60.0f, 23.0f, 95.0f};
//float y[] = {20.0f, 9.0f, 8.0f, 14.0f, 0.0f, 40.0f, 35.0f, 18.0f};
float x[] = {67.0f, 173.0f, 196.0f, 103.0f, 117.0f, 105.0f, 80.0f, 114.0f, 142.0f, 31.0f};
float y[] = {69.0f, 156.0f, 144.0f, 118.0f, 51.0f, 56.0f, 192.0f, 190.0f, 194.0f, 139.0f};
//float x[] = {0.000000,0.000985,0.041631,0.050832,0.091331,0.092298,0.176643,0.233178};
//float y[] = {0.364602,0.454433,0.487217,0.526750,0.556094,0.568060,0.831292,0.931731};
void _swap_reg(float &a, float &b, int idx){
float c = a <= b ? a:b;
float d = a > b ? a:b;
*(x+idx) = c;
*(y+idx) = d;
}
static void
inspectLH_avx(__m256 l, __m256 h)
{
float f1[8];
float f2[8];
_mm256_storeu_ps(f1, l);
_mm256_storeu_ps(f2, h);
for (int i = 0; i < 8; ++i)
{
/* code */
printf("RegA_%d: %.1f RegB_%d: %.1f\n", i, f1[i], i, f2[i]);
}
puts("=================================");
/*printf("[L:%.1f H:%.1f] [L:%.1f H:%.1f] [L:%.1f H:%.1f] [L:%.1f H:%.1f] [L:%.1f H:%.1f] [L:%.1f H:%.1f] [L:%.1f H:%.1f] [L:%.1f H:%.1f] \n",
f1[0], f2[0], f1[1], f2[1], f1[2], f2[2], f1[3], f2[3], f1[4], f2[4], f1[5], f2[5], f1[6], f2[6], f1[7], f2[7]);*/
}
static void inline
inspect_avx(__m256 v)
{
float f[8];
_mm256_storeu_ps(f, v);
for (int i = 0; i < 8; ++i)
{
/* code */
printf("Reg%d: %.1f\n", i, f[i]);
}
puts("=================================");
/*printf("[%.1f,%.1f,%.1f,%.1f,%.1f,%.1f,%.1f,%.1f]\n",
f[0], f[1], f[2], f[3],f[4], f[5], f[6], f[7]);*/
}
static void inline
bitonic_merge_kernel_avx(__m256 *a, __m256 *b)
{
puts("[dbg] PHASE 1");
__m256 a1 = _mm256_min_ps(*a, *b);
__m256 b1 = _mm256_max_ps(*a, *b);
__m256 a1_out = _mm256_blend_ps (a1, b1, _MM_SHUFFLE(2,2,2,2));
__m256 b1_out = _mm256_blend_ps (b1, a1, _MM_SHUFFLE(2,2,2,2));
inspectLH_avx(a1_out, b1_out);
puts("[dbg] PHASE 2");
__m256 a1_mod = _mm256_shuffle_ps(a1_out, a1_out, _MM_SHUFFLE(2,3,0,1));
inspectLH_avx(a1_mod, b1_out);
__m256 b1_mod = _mm256_shuffle_ps(b1_out, b1_out, _MM_SHUFFLE(2,3,0,1));
inspectLH_avx(a1_out, b1_mod);
__m256 a21_blend = _mm256_blend_ps (a1_out, b1_mod, _MM_SHUFFLE(2,2,2,2));
__m256 b21_blend = _mm256_blend_ps (a1_mod, b1_out, _MM_SHUFFLE(2,2,2,2));
inspectLH_avx(a21_blend, b21_blend);
__m256 a21_out = _mm256_min_ps(a21_blend, b21_blend);
__m256 b21_out = _mm256_max_ps(a21_blend, b21_blend);
inspectLH_avx(a21_out, b21_out);
__m256 regx = _mm256_blend_ps (a21_out, b21_out, _MM_SHUFFLE(3,0,3,0));
__m256 regy = _mm256_blend_ps (a21_out, b21_out, _MM_SHUFFLE(0,3,0,3));
inspectLH_avx(regx, regy);
__m256 regx1 = _mm256_permute_ps(regx, _MM_SHUFFLE(2,3,0,1) );
__m256 regy1 = _mm256_permute_ps(regy, _MM_SHUFFLE(2,3,0,1) );
inspectLH_avx(regx1, regy1);
__m256 regx2 = _mm256_blend_ps (regx, regy1, _MM_SHUFFLE(2,2,2,2));
__m256 regy2 = _mm256_blend_ps (regx1, regy, _MM_SHUFFLE(2,2,2,2));
inspectLH_avx(regx2, regy2);
regx = _mm256_min_ps(regx2, regy2);
regy = _mm256_max_ps(regx2, regy2);
inspectLH_avx(regx, regy);
regx1 = _mm256_blend_ps (regx, regy, _MM_SHUFFLE(3,0,3,0));
regy1 = _mm256_blend_ps (regx, regy, _MM_SHUFFLE(0,3,0,3));
inspectLH_avx(regx1, regy1);
regx = regx1;
regy = regy1;
//output from the second big box;
regx1 = (_mm256_unpacklo_ps(regx, regy)) ;
regy1 = (_mm256_unpackhi_ps(regx, regy)) ;
regx = _mm256_min_ps(regx1, regy1);
regy = _mm256_max_ps(regx1, regy1);
inspectLH_avx(regx, regy);
regx1 = _mm256_blend_ps (regx, regy, _MM_SHUFFLE(3,3,0,0));
regy1 = _mm256_blend_ps (regx, regy, _MM_SHUFFLE(0,0,3,3));
inspectLH_avx(regx1, regy1);
regx = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(1,0,1,0));
regy = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx, regy);
regx1 = (_mm256_unpacklo_ps(regx, regy)) ;
regy1 = (_mm256_unpackhi_ps(regx, regy)) ;
inspectLH_avx(regx1, regy1);
regx = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(1,0,1,0));
regy = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx, regy);
regx1 = (_mm256_unpacklo_ps(regx, regy)) ;
regy1 = (_mm256_unpackhi_ps(regx, regy)) ;
// inspectLH_avx(regx1, regy1);
regx = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(1,0,1,0));
regy = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx, regy);
regx1 = _mm256_min_ps(regx, regy);
regy1 = _mm256_max_ps(regx, regy);
inspectLH_avx(regx1, regy1);
regx = _mm256_blend_ps (regx1, regy1, _MM_SHUFFLE(3,3,0,0));
regy = _mm256_blend_ps (regx1, regy1, _MM_SHUFFLE(0,0,3,3));
inspectLH_avx(regx, regy);
regx1 = (_mm256_unpacklo_ps(regx, regy)) ;
regy1 = (_mm256_unpackhi_ps(regx, regy)) ;
// inspectLH_avx(regx1, regy1);
regx = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(1,0,1,0));
regy = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx, regy);
regx1 = _mm256_min_ps(regx, regy);
regy1 = _mm256_max_ps(regx, regy);
inspectLH_avx(regx1, regy1);
regx = _mm256_blend_ps (regx1, regy1, _MM_SHUFFLE(3,3,0,0));
regy = _mm256_blend_ps (regx1, regy1, _MM_SHUFFLE(0,0,3,3));
inspectLH_avx(regx, regy);
//Box phase 2
regx1 = (_mm256_unpacklo_ps(regx, regy)) ;
regy1 = (_mm256_unpackhi_ps(regx, regy)) ;
regx = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(1,0,1,0));
regy = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(3,2,3,2));
// inspectLH_avx(regx, regy);
regx1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(1,0,1,0));
regy1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(3,2,3,2));
// inspectLH_avx(regx1, regy1);
//
regx = _mm256_permute2f128_ps(regx1, regy1, 0x20);
regy = _mm256_permute2f128_ps(regx1, regy1, 0x31);
inspectLH_avx(regx, regy );
regx1 = _mm256_min_ps(regx, regy);
regy1 = _mm256_max_ps(regx, regy);
inspectLH_avx(regx1, regy1); //output
//===========================================
regx = (_mm256_unpacklo_ps(regx1, regy1)) ;
regy = (_mm256_unpackhi_ps(regx1, regy1)) ;
regx1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(1,0,1,0));
regy1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx1, regy1);
//THE FIRST LONG PHASE FINISH HERE;
//( __m256i )0+ 2<<3 + 4 <<6 + 6<<9 + 1<<12 + 3<<15 + 5<<18 + 7<<21
int _perm[8];
_perm[0] = 0;
_perm[1] = 2;
_perm[2] = 4;
_perm[3] = 6;
_perm[4] = 1;
_perm[5] = 3;
_perm[6] = 5;
_perm[7] = 7;
__m256i idx = _mm256_load_si256((__m256i *) _perm);
regx = _mm256_permutevar8x32_ps(regx1, idx);
regy = _mm256_permutevar8x32_ps(regy1, idx);
inspectLH_avx(regx, regy);
regx1 = (_mm256_unpacklo_ps(regx, regy)) ;
regy1 = (_mm256_unpackhi_ps(regx, regy)) ;
regx = _mm256_min_ps(regx1, regy1);
regy = _mm256_max_ps(regx1, regy1);
inspectLH_avx(regx, regy);
regx1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(1,0,1,0));
regy1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx1, regy1);
//===========================================
regx = (_mm256_unpacklo_ps(regx1, regy1)) ;
regy = (_mm256_unpackhi_ps(regx1, regy1)) ;
regx1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(1,0,1,0));
regy1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx1, regy1);
regx = (_mm256_unpacklo_ps(regx1, regy1)) ;
regy = (_mm256_unpackhi_ps(regx1, regy1)) ;
regx1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(1,0,1,0));
regy1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(3,2,3,2));
regx = _mm256_min_ps(regx1, regy1);
regy = _mm256_max_ps(regx1, regy1);
inspectLH_avx(regx, regy);
regx1 = (_mm256_unpacklo_ps(regx, regy)) ;
regy1 = (_mm256_unpackhi_ps(regx, regy)) ;
regx = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(1,0,1,0));
regy = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx, regy);
regx1 = _mm256_min_ps(regx, regy);
regy1 = _mm256_max_ps(regx, regy);
inspectLH_avx(regx1, regy1); //<good result here;
regx = (_mm256_unpacklo_ps(regx1, regy1)) ;
regy = (_mm256_unpackhi_ps(regx1, regy1)) ;
regx1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(1,0,1,0));
regy1 = _mm256_shuffle_ps (regx, regy, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx1, regy1);
regx = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(1,0,1,0));
regy = _mm256_shuffle_ps (regx1, regy1, _MM_SHUFFLE(3,2,3,2));
inspectLH_avx(regx, regy);
_perm[0] = 4;
_perm[1] = 5;
_perm[2] = 6;
_perm[3] = 7;
_perm[4] = 0;
_perm[5] = 1;
_perm[6] = 2;
_perm[7] = 3;
idx = _mm256_load_si256((__m256i *) _perm);
regx1 = _mm256_permutevar8x32_ps(regx, idx);
regy1 = regy;
inspectLH_avx(regx1, regy1);
regx = _mm256_blend_ps (regx1, regy1, _MM_SHUFFLE(0,0,3,3));
regy = _mm256_blend_ps (regx1, regy1, _MM_SHUFFLE(3,3,0,0));
inspectLH_avx(regx, regy);
regx1 = _mm256_permutevar8x32_ps(regx, idx);
regy1 = regy;
inspectLH_avx(regx1, regy1);
/*float ret[16];
_mm256_storeu_ps(ret, regx1);
_mm256_storeu_ps(ret + 8, regy1);
puts("[dbg] output the final result");
for(int i = 0;i < 16;i++){
printf("%f\n", ret[i]);
}
puts("");*/
/*for(int i = 0;i < 16;i++){
printf("%.1f ", ret[i]);
}*/
//__m256 a3_2_hi = _mm256_unpackhi_ps (a1, b1);
// inspect_avx(a3_2_hi);
return;
}
static void
simple_bitonic_test_avx(void)
{
//float x[] = {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f};
//float y[] = {1.1f, 1.2f, 1.3f, 1.4f, 1.5f, 1.6f, 1.7f, 1.8f};
std::reverse(x, x + 8);
std::reverse(y ,y + 8);
__m256 a = _mm256_set_ps(x[0], x[1], x[2], x[3], x[4], x[5], x[6], x[7]);
__m256 b = _mm256_set_ps(y[0], y[1], y[2], y[3], y[4], y[5], y[6], y[7]);
//puts("[dbg] output the xy value:");inspect_xy();
//puts("[dbg] output the xy value");inspect_xy();
bitonic_merge_kernel_avx(&a, &b);
// inspect_avx(a);
// inspect_avx(b);
/* int i;
*/
}
int main(int argc, char const *argv[])
{
/*float test_reg_a = 1.6;
float test_reg_b = 1.6;
_swap_reg(test_reg_a, test_reg_b);
cout<< "test_reg_a: " << test_reg_a << " test_reg_b: " << test_reg_b << endl;*/
simple_bitonic_test_avx();
return 0;
}