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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
|
/*
* Copyright 2011-2013 Blender Foundation
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License
*/
#ifndef __UTIL_MATH_H__
#define __UTIL_MATH_H__
/* Math
*
* Basic math functions on scalar and vector types. This header is used by
* both the kernel code when compiled as C++, and other C++ non-kernel code. */
#ifndef __KERNEL_OPENCL__
#ifdef _MSC_VER
# define _USE_MATH_DEFINES
#endif
#include <float.h>
#include <math.h>
#include <stdio.h>
#endif
#include "util_types.h"
CCL_NAMESPACE_BEGIN
/* Float Pi variations */
/* Division */
#ifndef M_PI_F
#define M_PI_F ((float)3.14159265358979323846264338327950288) /* pi */
#endif
#ifndef M_PI_2_F
#define M_PI_2_F ((float)1.57079632679489661923132169163975144) /* pi/2 */
#endif
#ifndef M_PI_4_F
#define M_PI_4_F ((float)0.785398163397448309615660845819875721) /* pi/4 */
#endif
#ifndef M_1_PI_F
#define M_1_PI_F ((float)0.318309886183790671537767526745028724) /* 1/pi */
#endif
#ifndef M_2_PI_F
#define M_2_PI_F ((float)0.636619772367581343075535053490057448) /* 2/pi */
#endif
/* Multiplication */
#ifndef M_2PI_F
#define M_2PI_F ((float)6.283185307179586476925286766559005768) /* 2*pi */
#endif
#ifndef M_4PI_F
#define M_4PI_F ((float)12.56637061435917295385057353311801153) /* 4*pi */
#endif
/* Float sqrt variations */
#ifndef M_SQRT2_F
#define M_SQRT2_F ((float)1.41421356237309504880) /* sqrt(2) */
#endif
/* Scalar */
#ifdef _WIN32
#ifndef __KERNEL_GPU__
#if(!defined(FREE_WINDOWS))
#define copysignf(x, y) ((float)_copysign(x, y))
#define hypotf(x, y) _hypotf(x, y)
#define isnan(x) _isnan(x)
#define isfinite(x) _finite(x)
#endif
#endif
#ifndef __KERNEL_OPENCL__
ccl_device_inline float fmaxf(float a, float b)
{
return (a > b)? a: b;
}
ccl_device_inline float fminf(float a, float b)
{
return (a < b)? a: b;
}
#endif
#endif
#ifndef __KERNEL_GPU__
ccl_device_inline int max(int a, int b)
{
return (a > b)? a: b;
}
ccl_device_inline int min(int a, int b)
{
return (a < b)? a: b;
}
ccl_device_inline float max(float a, float b)
{
return (a > b)? a: b;
}
ccl_device_inline float min(float a, float b)
{
return (a < b)? a: b;
}
ccl_device_inline double max(double a, double b)
{
return (a > b)? a: b;
}
ccl_device_inline double min(double a, double b)
{
return (a < b)? a: b;
}
#endif
ccl_device_inline float min4(float a, float b, float c, float d)
{
return min(min(a, b), min(c, d));
}
ccl_device_inline float max4(float a, float b, float c, float d)
{
return max(max(a, b), max(c, d));
}
#ifndef __KERNEL_OPENCL__
ccl_device_inline int clamp(int a, int mn, int mx)
{
return min(max(a, mn), mx);
}
ccl_device_inline float clamp(float a, float mn, float mx)
{
return min(max(a, mn), mx);
}
#endif
ccl_device_inline int float_to_int(float f)
{
#if defined(__KERNEL_SSE2__) && !defined(_MSC_VER)
return _mm_cvtt_ss2si(_mm_load_ss(&f));
#else
return (int)f;
#endif
}
ccl_device_inline int floor_to_int(float f)
{
return float_to_int(floorf(f));
}
ccl_device_inline int ceil_to_int(float f)
{
return float_to_int(ceilf(f));
}
ccl_device_inline float signf(float f)
{
return (f < 0.0f)? -1.0f: 1.0f;
}
ccl_device_inline float nonzerof(float f, float eps)
{
if(fabsf(f) < eps)
return signf(f)*eps;
else
return f;
}
ccl_device_inline float smoothstepf(float f)
{
float ff = f*f;
return (3.0f*ff - 2.0f*ff*f);
}
/* Float2 Vector */
#ifndef __KERNEL_OPENCL__
ccl_device_inline bool is_zero(const float2 a)
{
return (a.x == 0.0f && a.y == 0.0f);
}
#endif
#ifndef __KERNEL_OPENCL__
ccl_device_inline float average(const float2 a)
{
return (a.x + a.y)*(1.0f/2.0f);
}
#endif
#ifndef __KERNEL_OPENCL__
ccl_device_inline float2 operator-(const float2 a)
{
return make_float2(-a.x, -a.y);
}
ccl_device_inline float2 operator*(const float2 a, const float2 b)
{
return make_float2(a.x*b.x, a.y*b.y);
}
ccl_device_inline float2 operator*(const float2 a, float f)
{
return make_float2(a.x*f, a.y*f);
}
ccl_device_inline float2 operator*(float f, const float2 a)
{
return make_float2(a.x*f, a.y*f);
}
ccl_device_inline float2 operator/(float f, const float2 a)
{
return make_float2(f/a.x, f/a.y);
}
ccl_device_inline float2 operator/(const float2 a, float f)
{
float invf = 1.0f/f;
return make_float2(a.x*invf, a.y*invf);
}
ccl_device_inline float2 operator/(const float2 a, const float2 b)
{
return make_float2(a.x/b.x, a.y/b.y);
}
ccl_device_inline float2 operator+(const float2 a, const float2 b)
{
return make_float2(a.x+b.x, a.y+b.y);
}
ccl_device_inline float2 operator-(const float2 a, const float2 b)
{
return make_float2(a.x-b.x, a.y-b.y);
}
ccl_device_inline float2 operator+=(float2& a, const float2 b)
{
return a = a + b;
}
ccl_device_inline float2 operator*=(float2& a, const float2 b)
{
return a = a * b;
}
ccl_device_inline float2 operator*=(float2& a, float f)
{
return a = a * f;
}
ccl_device_inline float2 operator/=(float2& a, const float2 b)
{
return a = a / b;
}
ccl_device_inline float2 operator/=(float2& a, float f)
{
float invf = 1.0f/f;
return a = a * invf;
}
ccl_device_inline float dot(const float2 a, const float2 b)
{
return a.x*b.x + a.y*b.y;
}
ccl_device_inline float cross(const float2 a, const float2 b)
{
return (a.x*b.y - a.y*b.x);
}
#endif
#ifndef __KERNEL_OPENCL__
ccl_device_inline bool operator==(const int2 a, const int2 b)
{
return (a.x == b.x && a.y == b.y);
}
ccl_device_inline float len(const float2 a)
{
return sqrtf(dot(a, a));
}
ccl_device_inline float2 normalize(const float2 a)
{
return a/len(a);
}
ccl_device_inline float2 normalize_len(const float2 a, float *t)
{
*t = len(a);
return a/(*t);
}
ccl_device_inline bool operator==(const float2 a, const float2 b)
{
return (a.x == b.x && a.y == b.y);
}
ccl_device_inline bool operator!=(const float2 a, const float2 b)
{
return !(a == b);
}
ccl_device_inline float2 min(float2 a, float2 b)
{
return make_float2(min(a.x, b.x), min(a.y, b.y));
}
ccl_device_inline float2 max(float2 a, float2 b)
{
return make_float2(max(a.x, b.x), max(a.y, b.y));
}
ccl_device_inline float2 clamp(float2 a, float2 mn, float2 mx)
{
return min(max(a, mn), mx);
}
ccl_device_inline float2 fabs(float2 a)
{
return make_float2(fabsf(a.x), fabsf(a.y));
}
ccl_device_inline float2 as_float2(const float4 a)
{
return make_float2(a.x, a.y);
}
#endif
#ifndef __KERNEL_GPU__
ccl_device_inline void print_float2(const char *label, const float2& a)
{
printf("%s: %.8f %.8f\n", label, (double)a.x, (double)a.y);
}
#endif
#ifndef __KERNEL_OPENCL__
ccl_device_inline float2 interp(float2 a, float2 b, float t)
{
return a + t*(b - a);
}
#endif
/* Float3 Vector */
#ifndef __KERNEL_OPENCL__
ccl_device_inline float3 operator-(const float3 a)
{
return make_float3(-a.x, -a.y, -a.z);
}
ccl_device_inline float3 operator*(const float3 a, const float3 b)
{
return make_float3(a.x*b.x, a.y*b.y, a.z*b.z);
}
ccl_device_inline float3 operator*(const float3 a, float f)
{
return make_float3(a.x*f, a.y*f, a.z*f);
}
ccl_device_inline float3 operator*(float f, const float3 a)
{
return make_float3(a.x*f, a.y*f, a.z*f);
}
ccl_device_inline float3 operator/(float f, const float3 a)
{
return make_float3(f/a.x, f/a.y, f/a.z);
}
ccl_device_inline float3 operator/(const float3 a, float f)
{
float invf = 1.0f/f;
return make_float3(a.x*invf, a.y*invf, a.z*invf);
}
ccl_device_inline float3 operator/(const float3 a, const float3 b)
{
return make_float3(a.x/b.x, a.y/b.y, a.z/b.z);
}
ccl_device_inline float3 operator+(const float3 a, const float3 b)
{
return make_float3(a.x+b.x, a.y+b.y, a.z+b.z);
}
ccl_device_inline float3 operator-(const float3 a, const float3 b)
{
return make_float3(a.x-b.x, a.y-b.y, a.z-b.z);
}
ccl_device_inline float3 operator+=(float3& a, const float3 b)
{
return a = a + b;
}
ccl_device_inline float3 operator*=(float3& a, const float3 b)
{
return a = a * b;
}
ccl_device_inline float3 operator*=(float3& a, float f)
{
return a = a * f;
}
ccl_device_inline float3 operator/=(float3& a, const float3 b)
{
return a = a / b;
}
ccl_device_inline float3 operator/=(float3& a, float f)
{
float invf = 1.0f/f;
return a = a * invf;
}
ccl_device_inline float dot(const float3 a, const float3 b)
{
#if defined(__KERNEL_SSE41__) && defined(__KERNEL_SSE__)
return _mm_cvtss_f32(_mm_dp_ps(a, b, 0x7F));
#else
return a.x*b.x + a.y*b.y + a.z*b.z;
#endif
}
ccl_device_inline float3 cross(const float3 a, const float3 b)
{
float3 r = make_float3(a.y*b.z - a.z*b.y, a.z*b.x - a.x*b.z, a.x*b.y - a.y*b.x);
return r;
}
#endif
ccl_device_inline float len(const float3 a)
{
#if defined(__KERNEL_SSE41__) && defined(__KERNEL_SSE__)
return _mm_cvtss_f32(_mm_sqrt_ss(_mm_dp_ps(a.m128, a.m128, 0x7F)));
#else
return sqrtf(dot(a, a));
#endif
}
ccl_device_inline float len_squared(const float3 a)
{
return dot(a, a);
}
#ifndef __KERNEL_OPENCL__
ccl_device_inline float3 normalize(const float3 a)
{
#if defined(__KERNEL_SSE41__) && defined(__KERNEL_SSE__)
__m128 norm = _mm_sqrt_ps(_mm_dp_ps(a.m128, a.m128, 0x7F));
return _mm_div_ps(a.m128, norm);
#else
return a/len(a);
#endif
}
#endif
ccl_device_inline float3 normalize_len(const float3 a, float *t)
{
*t = len(a);
return a/(*t);
}
#ifndef __KERNEL_OPENCL__
ccl_device_inline bool operator==(const float3 a, const float3 b)
{
#ifdef __KERNEL_SSE__
return (_mm_movemask_ps(_mm_cmpeq_ps(a.m128, b.m128)) & 7) == 7;
#else
return (a.x == b.x && a.y == b.y && a.z == b.z);
#endif
}
ccl_device_inline bool operator!=(const float3 a, const float3 b)
{
return !(a == b);
}
ccl_device_inline float3 min(float3 a, float3 b)
{
#ifdef __KERNEL_SSE__
return _mm_min_ps(a.m128, b.m128);
#else
return make_float3(min(a.x, b.x), min(a.y, b.y), min(a.z, b.z));
#endif
}
ccl_device_inline float3 max(float3 a, float3 b)
{
#ifdef __KERNEL_SSE__
return _mm_max_ps(a.m128, b.m128);
#else
return make_float3(max(a.x, b.x), max(a.y, b.y), max(a.z, b.z));
#endif
}
ccl_device_inline float3 clamp(float3 a, float3 mn, float3 mx)
{
return min(max(a, mn), mx);
}
ccl_device_inline float3 fabs(float3 a)
{
#ifdef __KERNEL_SSE__
__m128 mask = _mm_castsi128_ps(_mm_set1_epi32(0x7fffffff));
return _mm_and_ps(a.m128, mask);
#else
return make_float3(fabsf(a.x), fabsf(a.y), fabsf(a.z));
#endif
}
#endif
ccl_device_inline float3 float2_to_float3(const float2 a)
{
return make_float3(a.x, a.y, 0.0f);
}
ccl_device_inline float3 float4_to_float3(const float4 a)
{
return make_float3(a.x, a.y, a.z);
}
ccl_device_inline float4 float3_to_float4(const float3 a)
{
return make_float4(a.x, a.y, a.z, 1.0f);
}
#ifndef __KERNEL_GPU__
ccl_device_inline void print_float3(const char *label, const float3& a)
{
printf("%s: %.8f %.8f %.8f\n", label, (double)a.x, (double)a.y, (double)a.z);
}
ccl_device_inline float3 rcp(const float3& a)
{
#ifdef __KERNEL_SSE__
float4 r = _mm_rcp_ps(a.m128);
return _mm_sub_ps(_mm_add_ps(r, r), _mm_mul_ps(_mm_mul_ps(r, r), a));
#else
return make_float3(1.0f/a.x, 1.0f/a.y, 1.0f/a.z);
#endif
}
#endif
ccl_device_inline float3 interp(float3 a, float3 b, float t)
{
return a + t*(b - a);
}
ccl_device_inline bool is_zero(const float3 a)
{
#ifdef __KERNEL_SSE__
return a == make_float3(0.0f);
#else
return (a.x == 0.0f && a.y == 0.0f && a.z == 0.0f);
#endif
}
ccl_device_inline float reduce_add(const float3 a)
{
#ifdef __KERNEL_SSE__
return (a.x + a.y + a.z);
#else
return (a.x + a.y + a.z);
#endif
}
ccl_device_inline float average(const float3 a)
{
return reduce_add(a)*(1.0f/3.0f);
}
/* Float4 Vector */
#ifdef __KERNEL_SSE__
template<size_t index_0, size_t index_1, size_t index_2, size_t index_3> __forceinline const float4 shuffle(const float4& b)
{
return _mm_castsi128_ps(_mm_shuffle_epi32(_mm_castps_si128(b), _MM_SHUFFLE(index_3, index_2, index_1, index_0)));
}
template<> __forceinline const float4 shuffle<0, 0, 2, 2>(const float4& b)
{
return _mm_moveldup_ps(b);
}
template<> __forceinline const float4 shuffle<1, 1, 3, 3>(const float4& b)
{
return _mm_movehdup_ps(b);
}
template<> __forceinline const float4 shuffle<0, 1, 0, 1>(const float4& b)
{
return _mm_castpd_ps(_mm_movedup_pd(_mm_castps_pd(b)));
}
#endif
#ifndef __KERNEL_OPENCL__
ccl_device_inline float4 operator-(const float4& a)
{
#ifdef __KERNEL_SSE__
__m128 mask = _mm_castsi128_ps(_mm_set1_epi32(0x80000000));
return _mm_xor_ps(a.m128, mask);
#else
return make_float4(-a.x, -a.y, -a.z, -a.w);
#endif
}
ccl_device_inline float4 operator*(const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
return _mm_mul_ps(a.m128, b.m128);
#else
return make_float4(a.x*b.x, a.y*b.y, a.z*b.z, a.w*b.w);
#endif
}
ccl_device_inline float4 operator*(const float4& a, float f)
{
#ifdef __KERNEL_SSE__
return a * make_float4(f);
#else
return make_float4(a.x*f, a.y*f, a.z*f, a.w*f);
#endif
}
ccl_device_inline float4 operator*(float f, const float4& a)
{
return a * f;
}
ccl_device_inline float4 rcp(const float4& a)
{
#ifdef __KERNEL_SSE__
float4 r = _mm_rcp_ps(a.m128);
return _mm_sub_ps(_mm_add_ps(r, r), _mm_mul_ps(_mm_mul_ps(r, r), a));
#else
return make_float4(1.0f/a.x, 1.0f/a.y, 1.0f/a.z, 1.0f/a.w);
#endif
}
ccl_device_inline float4 operator/(const float4& a, float f)
{
return a * (1.0f/f);
}
ccl_device_inline float4 operator/(const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
return a * rcp(b);
#else
return make_float4(a.x/b.x, a.y/b.y, a.z/b.z, a.w/b.w);
#endif
}
ccl_device_inline float4 operator+(const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
return _mm_add_ps(a.m128, b.m128);
#else
return make_float4(a.x+b.x, a.y+b.y, a.z+b.z, a.w+b.w);
#endif
}
ccl_device_inline float4 operator-(const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
return _mm_sub_ps(a.m128, b.m128);
#else
return make_float4(a.x-b.x, a.y-b.y, a.z-b.z, a.w-b.w);
#endif
}
ccl_device_inline float4 operator+=(float4& a, const float4& b)
{
return a = a + b;
}
ccl_device_inline float4 operator*=(float4& a, const float4& b)
{
return a = a * b;
}
ccl_device_inline float4 operator/=(float4& a, float f)
{
return a = a / f;
}
ccl_device_inline int4 operator<(const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
return _mm_cvtps_epi32(_mm_cmplt_ps(a.m128, b.m128)); /* todo: avoid cvt */
#else
return make_int4(a.x < b.x, a.y < b.y, a.z < b.z, a.w < b.w);
#endif
}
ccl_device_inline int4 operator>=(float4 a, float4 b)
{
#ifdef __KERNEL_SSE__
return _mm_cvtps_epi32(_mm_cmpge_ps(a.m128, b.m128)); /* todo: avoid cvt */
#else
return make_int4(a.x >= b.x, a.y >= b.y, a.z >= b.z, a.w >= b.w);
#endif
}
ccl_device_inline int4 operator<=(const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
return _mm_cvtps_epi32(_mm_cmple_ps(a.m128, b.m128)); /* todo: avoid cvt */
#else
return make_int4(a.x <= b.x, a.y <= b.y, a.z <= b.z, a.w <= b.w);
#endif
}
ccl_device_inline bool operator==(const float4 a, const float4 b)
{
#ifdef __KERNEL_SSE__
return (_mm_movemask_ps(_mm_cmpeq_ps(a.m128, b.m128)) & 15) == 15;
#else
return (a.x == b.x && a.y == b.y && a.z == b.z && a.w == b.w);
#endif
}
ccl_device_inline float4 cross(const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
return (shuffle<1,2,0,0>(a)*shuffle<2,0,1,0>(b)) - (shuffle<2,0,1,0>(a)*shuffle<1,2,0,0>(b));
#else
return make_float4(a.y*b.z - a.z*b.y, a.z*b.x - a.x*b.z, a.x*b.y - a.y*b.x, 0.0f);
#endif
}
ccl_device_inline bool is_zero(const float4& a)
{
#ifdef __KERNEL_SSE__
return a == make_float4(0.0f);
#else
return (a.x == 0.0f && a.y == 0.0f && a.z == 0.0f && a.w == 0.0f);
#endif
}
ccl_device_inline float reduce_add(const float4& a)
{
#ifdef __KERNEL_SSE__
float4 h = shuffle<1,0,3,2>(a) + a;
return _mm_cvtss_f32(shuffle<2,3,0,1>(h) + h); /* todo: efficiency? */
#else
return ((a.x + a.y) + (a.z + a.w));
#endif
}
ccl_device_inline float average(const float4& a)
{
return reduce_add(a) * 0.25f;
}
ccl_device_inline float dot(const float4& a, const float4& b)
{
return reduce_add(a * b);
}
ccl_device_inline float len(const float4 a)
{
return sqrtf(dot(a, a));
}
ccl_device_inline float4 normalize(const float4 a)
{
return a/len(a);
}
ccl_device_inline float4 min(float4 a, float4 b)
{
#ifdef __KERNEL_SSE__
return _mm_min_ps(a.m128, b.m128);
#else
return make_float4(min(a.x, b.x), min(a.y, b.y), min(a.z, b.z), min(a.w, b.w));
#endif
}
ccl_device_inline float4 max(float4 a, float4 b)
{
#ifdef __KERNEL_SSE__
return _mm_max_ps(a.m128, b.m128);
#else
return make_float4(max(a.x, b.x), max(a.y, b.y), max(a.z, b.z), max(a.w, b.w));
#endif
}
#endif
#ifndef __KERNEL_GPU__
ccl_device_inline float4 select(const int4& mask, const float4& a, const float4& b)
{
#ifdef __KERNEL_SSE__
/* blendv is sse4, and apparently broken on vs2008 */
return _mm_or_ps(_mm_and_ps(_mm_cvtepi32_ps(mask), a), _mm_andnot_ps(_mm_cvtepi32_ps(mask), b)); /* todo: avoid cvt */
#else
return make_float4((mask.x)? a.x: b.x, (mask.y)? a.y: b.y, (mask.z)? a.z: b.z, (mask.w)? a.w: b.w);
#endif
}
ccl_device_inline float4 reduce_min(const float4& a)
{
#ifdef __KERNEL_SSE__
float4 h = min(shuffle<1,0,3,2>(a), a);
return min(shuffle<2,3,0,1>(h), h);
#else
return make_float4(min(min(a.x, a.y), min(a.z, a.w)));
#endif
}
ccl_device_inline float4 reduce_max(const float4& a)
{
#ifdef __KERNEL_SSE__
float4 h = max(shuffle<1,0,3,2>(a), a);
return max(shuffle<2,3,0,1>(h), h);
#else
return make_float4(max(max(a.x, a.y), max(a.z, a.w)));
#endif
}
#if 0
ccl_device_inline float4 reduce_add(const float4& a)
{
#ifdef __KERNEL_SSE__
float4 h = shuffle<1,0,3,2>(a) + a;
return shuffle<2,3,0,1>(h) + h;
#else
return make_float4((a.x + a.y) + (a.z + a.w));
#endif
}
#endif
ccl_device_inline void print_float4(const char *label, const float4& a)
{
printf("%s: %.8f %.8f %.8f %.8f\n", label, (double)a.x, (double)a.y, (double)a.z, (double)a.w);
}
#endif
/* Int3 */
#ifndef __KERNEL_OPENCL__
ccl_device_inline int3 min(int3 a, int3 b)
{
#ifdef __KERNEL_SSE__
return _mm_min_epi32(a.m128, b.m128);
#else
return make_int3(min(a.x, b.x), min(a.y, b.y), min(a.z, b.z));
#endif
}
ccl_device_inline int3 max(int3 a, int3 b)
{
#ifdef __KERNEL_SSE__
return _mm_max_epi32(a.m128, b.m128);
#else
return make_int3(max(a.x, b.x), max(a.y, b.y), max(a.z, b.z));
#endif
}
ccl_device_inline int3 clamp(const int3& a, int mn, int mx)
{
#ifdef __KERNEL_SSE__
return min(max(a, make_int3(mn)), make_int3(mx));
#else
return make_int3(clamp(a.x, mn, mx), clamp(a.y, mn, mx), clamp(a.z, mn, mx));
#endif
}
ccl_device_inline int3 clamp(const int3& a, int3& mn, int mx)
{
#ifdef __KERNEL_SSE__
return min(max(a, mn), make_int3(mx));
#else
return make_int3(clamp(a.x, mn.x, mx), clamp(a.y, mn.y, mx), clamp(a.z, mn.z, mx));
#endif
}
#endif
#ifndef __KERNEL_GPU__
ccl_device_inline void print_int3(const char *label, const int3& a)
{
printf("%s: %d %d %d\n", label, a.x, a.y, a.z);
}
#endif
/* Int4 */
#ifndef __KERNEL_GPU__
ccl_device_inline int4 operator+(const int4& a, const int4& b)
{
#ifdef __KERNEL_SSE__
return _mm_add_epi32(a.m128, b.m128);
#else
return make_int4(a.x+b.x, a.y+b.y, a.z+b.z, a.w+b.w);
#endif
}
ccl_device_inline int4 operator+=(int4& a, const int4& b)
{
return a = a + b;
}
ccl_device_inline int4 operator>>(const int4& a, int i)
{
#ifdef __KERNEL_SSE__
return _mm_srai_epi32(a.m128, i);
#else
return make_int4(a.x >> i, a.y >> i, a.z >> i, a.w >> i);
#endif
}
ccl_device_inline int4 min(int4 a, int4 b)
{
#ifdef __KERNEL_SSE__
return _mm_min_epi32(a.m128, b.m128);
#else
return make_int4(min(a.x, b.x), min(a.y, b.y), min(a.z, b.z), min(a.w, b.w));
#endif
}
ccl_device_inline int4 max(int4 a, int4 b)
{
#ifdef __KERNEL_SSE__
return _mm_max_epi32(a.m128, b.m128);
#else
return make_int4(max(a.x, b.x), max(a.y, b.y), max(a.z, b.z), max(a.w, b.w));
#endif
}
ccl_device_inline int4 clamp(const int4& a, const int4& mn, const int4& mx)
{
return min(max(a, mn), mx);
}
ccl_device_inline int4 select(const int4& mask, const int4& a, const int4& b)
{
#ifdef __KERNEL_SSE__
__m128 m = _mm_cvtepi32_ps(mask);
return _mm_castps_si128(_mm_or_ps(_mm_and_ps(m, _mm_castsi128_ps(a)), _mm_andnot_ps(m, _mm_castsi128_ps(b)))); /* todo: avoid cvt */
#else
return make_int4((mask.x)? a.x: b.x, (mask.y)? a.y: b.y, (mask.z)? a.z: b.z, (mask.w)? a.w: b.w);
#endif
}
ccl_device_inline void print_int4(const char *label, const int4& a)
{
printf("%s: %d %d %d %d\n", label, a.x, a.y, a.z, a.w);
}
#endif
/* Int/Float conversion */
#ifndef __KERNEL_OPENCL__
ccl_device_inline int as_int(uint i)
{
union { uint ui; int i; } u;
u.ui = i;
return u.i;
}
ccl_device_inline uint as_uint(int i)
{
union { uint ui; int i; } u;
u.i = i;
return u.ui;
}
ccl_device_inline uint as_uint(float f)
{
union { uint i; float f; } u;
u.f = f;
return u.i;
}
ccl_device_inline int __float_as_int(float f)
{
union { int i; float f; } u;
u.f = f;
return u.i;
}
ccl_device_inline float __int_as_float(int i)
{
union { int i; float f; } u;
u.i = i;
return u.f;
}
ccl_device_inline uint __float_as_uint(float f)
{
union { uint i; float f; } u;
u.f = f;
return u.i;
}
ccl_device_inline float __uint_as_float(uint i)
{
union { uint i; float f; } u;
u.i = i;
return u.f;
}
/* Interpolation */
template<class A, class B> A lerp(const A& a, const A& b, const B& t)
{
return (A)(a * ((B)1 - t) + b * t);
}
/* Triangle */
ccl_device_inline float triangle_area(const float3 v1, const float3 v2, const float3 v3)
{
return len(cross(v3 - v2, v1 - v2))*0.5f;
}
#endif
/* Orthonormal vectors */
ccl_device_inline void make_orthonormals(const float3 N, float3 *a, float3 *b)
{
#if 0
if(fabsf(N.y) >= 0.999f) {
*a = make_float3(1, 0, 0);
*b = make_float3(0, 0, 1);
return;
}
if(fabsf(N.z) >= 0.999f) {
*a = make_float3(1, 0, 0);
*b = make_float3(0, 1, 0);
return;
}
#endif
if(N.x != N.y || N.x != N.z)
*a = make_float3(N.z-N.y, N.x-N.z, N.y-N.x); //(1,1,1)x N
else
*a = make_float3(N.z-N.y, N.x+N.z, -N.y-N.x); //(-1,1,1)x N
*a = normalize(*a);
*b = cross(N, *a);
}
/* Color division */
ccl_device_inline float3 safe_divide_color(float3 a, float3 b)
{
float x, y, z;
x = (b.x != 0.0f)? a.x/b.x: 0.0f;
y = (b.y != 0.0f)? a.y/b.y: 0.0f;
z = (b.z != 0.0f)? a.z/b.z: 0.0f;
return make_float3(x, y, z);
}
ccl_device_inline float3 safe_divide_even_color(float3 a, float3 b)
{
float x, y, z;
x = (b.x != 0.0f)? a.x/b.x: 0.0f;
y = (b.y != 0.0f)? a.y/b.y: 0.0f;
z = (b.z != 0.0f)? a.z/b.z: 0.0f;
/* try to get grey even if b is zero */
if(b.x == 0.0f) {
if(b.y == 0.0f) {
x = z;
y = z;
}
else if(b.z == 0.0f) {
x = y;
z = y;
}
else
x = 0.5f*(y + z);
}
else if(b.y == 0.0f) {
if(b.z == 0.0f) {
y = x;
z = x;
}
else
y = 0.5f*(x + z);
}
else if(b.z == 0.0f) {
z = 0.5f*(x + y);
}
return make_float3(x, y, z);
}
/* Rotation of point around axis and angle */
ccl_device_inline float3 rotate_around_axis(float3 p, float3 axis, float angle)
{
float costheta = cosf(angle);
float sintheta = sinf(angle);
float3 r;
r.x = ((costheta + (1 - costheta) * axis.x * axis.x) * p.x) +
(((1 - costheta) * axis.x * axis.y - axis.z * sintheta) * p.y) +
(((1 - costheta) * axis.x * axis.z + axis.y * sintheta) * p.z);
r.y = (((1 - costheta) * axis.x * axis.y + axis.z * sintheta) * p.x) +
((costheta + (1 - costheta) * axis.y * axis.y) * p.y) +
(((1 - costheta) * axis.y * axis.z - axis.x * sintheta) * p.z);
r.z = (((1 - costheta) * axis.x * axis.z - axis.y * sintheta) * p.x) +
(((1 - costheta) * axis.y * axis.z + axis.x * sintheta) * p.y) +
((costheta + (1 - costheta) * axis.z * axis.z) * p.z);
return r;
}
/* NaN-safe math ops */
ccl_device_inline float safe_sqrtf(float f)
{
return sqrtf(max(f, 0.0f));
}
ccl_device float safe_asinf(float a)
{
if(a <= -1.0f)
return -M_PI_2_F;
else if(a >= 1.0f)
return M_PI_2_F;
return asinf(a);
}
ccl_device float safe_acosf(float a)
{
if(a <= -1.0f)
return M_PI_F;
else if(a >= 1.0f)
return 0.0f;
return acosf(a);
}
ccl_device float compatible_powf(float x, float y)
{
/* GPU pow doesn't accept negative x, do manual checks here */
if(x < 0.0f) {
if(fmodf(-y, 2.0f) == 0.0f)
return powf(-x, y);
else
return -powf(-x, y);
}
else if(x == 0.0f)
return 0.0f;
return powf(x, y);
}
ccl_device float safe_powf(float a, float b)
{
if(b == 0.0f)
return 1.0f;
if(a == 0.0f)
return 0.0f;
if(a < 0.0f && b != float_to_int(b))
return 0.0f;
return compatible_powf(a, b);
}
ccl_device float safe_logf(float a, float b)
{
if(a < 0.0f || b < 0.0f)
return 0.0f;
return logf(a)/logf(b);
}
ccl_device float safe_divide(float a, float b)
{
return (b != 0.0f)? a/b: 0.0f;
}
ccl_device float safe_modulo(float a, float b)
{
return (b != 0.0f)? fmodf(a, b): 0.0f;
}
/* Ray Intersection */
ccl_device bool ray_sphere_intersect(
float3 ray_P, float3 ray_D, float ray_t,
float3 sphere_P, float sphere_radius,
float3 *isect_P, float *isect_t)
{
float3 d = sphere_P - ray_P;
float radiussq = sphere_radius*sphere_radius;
float tsq = dot(d, d);
if(tsq > radiussq) { /* ray origin outside sphere */
float tp = dot(d, ray_D);
if(tp < 0.0f) /* dir points away from sphere */
return false;
float dsq = tsq - tp*tp; /* pythagoras */
if(dsq > radiussq) /* closest point on ray outside sphere */
return false;
float t = tp - sqrtf(radiussq - dsq); /* pythagoras */
if(t < ray_t) {
*isect_t = t;
*isect_P = ray_P + ray_D*t;
return true;
}
}
return false;
}
ccl_device bool ray_aligned_disk_intersect(
float3 ray_P, float3 ray_D, float ray_t,
float3 disk_P, float disk_radius,
float3 *isect_P, float *isect_t)
{
/* aligned disk normal */
float disk_t;
float3 disk_N = normalize_len(ray_P - disk_P, &disk_t);
float div = dot(ray_D, disk_N);
if(div == 0.0f)
return false;
/* compute t to intersection point */
float t = -disk_t/div;
if(t < 0.0f || t > ray_t)
return false;
/* test if within radius */
float3 P = ray_P + ray_D*t;
if(len_squared(P - disk_P) > disk_radius*disk_radius)
return false;
*isect_P = P;
*isect_t = t;
return true;
}
ccl_device bool ray_triangle_intersect(
float3 ray_P, float3 ray_D, float ray_t,
float3 v0, float3 v1, float3 v2,
float3 *isect_P, float *isect_t)
{
/* Calculate intersection */
float3 e1 = v1 - v0;
float3 e2 = v2 - v0;
float3 s1 = cross(ray_D, e2);
const float divisor = dot(s1, e1);
if(divisor == 0.0f)
return false;
const float invdivisor = 1.0f/divisor;
/* compute first barycentric coordinate */
const float3 d = ray_P - v0;
const float u = dot(d, s1)*invdivisor;
if(u < 0.0f)
return false;
/* Compute second barycentric coordinate */
const float3 s2 = cross(d, e1);
const float v = dot(ray_D, s2)*invdivisor;
if(v < 0.0f)
return false;
const float b0 = 1.0f - u - v;
if(b0 < 0.0f)
return false;
/* compute t to intersection point */
const float t = dot(e2, s2)*invdivisor;
if(t < 0.0f || t > ray_t)
return false;
*isect_t = t;
*isect_P = ray_P + ray_D*t;
return true;
}
ccl_device bool ray_quad_intersect(
float3 ray_P, float3 ray_D, float ray_t,
float3 quad_P, float3 quad_u, float3 quad_v,
float3 *isect_P, float *isect_t)
{
float3 v0 = quad_P - quad_u*0.5f - quad_v*0.5f;
float3 v1 = quad_P + quad_u*0.5f - quad_v*0.5f;
float3 v2 = quad_P + quad_u*0.5f + quad_v*0.5f;
float3 v3 = quad_P - quad_u*0.5f + quad_v*0.5f;
if(ray_triangle_intersect(ray_P, ray_D, ray_t, v0, v1, v2, isect_P, isect_t))
return true;
else if(ray_triangle_intersect(ray_P, ray_D, ray_t, v0, v2, v3, isect_P, isect_t))
return true;
return false;
}
CCL_NAMESPACE_END
#endif /* __UTIL_MATH_H__ */
|