forked from GaijinEntertainment/vecmath
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdag_vecMath_const.h
174 lines (156 loc) · 7.97 KB
/
dag_vecMath_const.h
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
//
// Dagor Engine 6.5
// Copyright (C) 2023 Gaijin Games KFT. All rights reserved
// (for conditions of use see prog/license.txt)
//
#pragma once
#include "dag_vecMathDecl.h"
#ifdef __cplusplus
#include <cmath> // for INFINITY
#else
#include <math.h>
#endif
#if defined(_MSC_VER) && !defined(__clang__)
#define DECL_VEC_CONST extern const __declspec(selectany) __declspec(align(16))
#else
#define DECL_VEC_CONST alignas(16) static const
#endif
#define REPLICATE(v) v, v, v, v
#if _TARGET_SIMD_SSE
DECL_VEC_CONST vec4f_const V_C_HALF = { REPLICATE(0.5f) };
DECL_VEC_CONST vec4f_const V_C_HALF_MINUS_EPS = { REPLICATE(0.5f - 1.192092896e-07f * 32) };
DECL_VEC_CONST vec4f_const V_C_ONE = { REPLICATE(1.0f) };
DECL_VEC_CONST vec4f_const V_C_TWO = { REPLICATE(2.0f) };
DECL_VEC_CONST vec4f_const V_C_PI = { REPLICATE(3.14159265f) }; // pi
DECL_VEC_CONST vec4f_const V_C_HALFPI = { REPLICATE(3.14159265f / 2.f) }; // pi/2
DECL_VEC_CONST vec4f_const V_C_TWOPI = { REPLICATE(2.f * 3.14159265f) }; // pi*2
DECL_VEC_CONST vec4f_const V_C_PI_DIV_4 = { REPLICATE(0.78539816f) }; // pi/4
DECL_VEC_CONST vec4f_const V_C_2_DIV_PI = { REPLICATE(0.63661977f) }; // 2/pi
DECL_VEC_CONST vec4f_const V_C_4_DIV_PI = { REPLICATE(1.27323954f) }; // 4/pi
DECL_VEC_CONST vec4f_const V_C_MAX_VAL = { REPLICATE(1e32f) };
DECL_VEC_CONST vec4f_const V_C_MIN_VAL = { REPLICATE(-1e32f) };
DECL_VEC_CONST vec4f_const V_C_EPS_VAL = { REPLICATE(1.192092896e-07f) };
DECL_VEC_CONST vec4f_const V_C_VERY_SMALL_VAL = { REPLICATE(4e-19f) }; // ~sqrt(FLT_MIN), safe threshold for passing argument to rcp(x)
#if defined(__clang__) || defined(__GNUC__)
DECL_VEC_CONST vec4f_const V_C_INF = { REPLICATE(__builtin_inff()) };
#else
DECL_VEC_CONST vec4f_const V_C_INF = { REPLICATE(INFINITY) };
#endif
#define DECL_VECFLOAT4(X,Y,Z,W) {X,Y,Z,W}
#define DECL_VECUINT4(X,Y,Z,W) {X,Y,Z,W}
DECL_VEC_CONST vec4i_const V_CI_SIGN_MASK = { REPLICATE(0x80000000) };
DECL_VEC_CONST vec4i_const V_CI_INV_SIGN_MASK = { REPLICATE(0x7FFFFFFF) };
DECL_VEC_CONST vec4i_const V_CI_3FF = { REPLICATE(0x3FF) };
DECL_VEC_CONST vec4i_const V_CI_070 = { REPLICATE(0x70) };
DECL_VEC_CONST vec4i_const V_CI_0FF = { REPLICATE(0xFF) };
DECL_VEC_CONST vec4i_const V_CI_01F = { REPLICATE(0x1F) };
DECL_VEC_CONST vec4i_const V_CI_07FFFFF = { REPLICATE(0x7fffff) };
DECL_VEC_CONST vec4i_const V_CI_0 = { REPLICATE(0) };
DECL_VEC_CONST vec4i_const V_CI_1 = { REPLICATE(1) };
DECL_VEC_CONST vec4i_const V_CI_2 = { REPLICATE(2) };
DECL_VEC_CONST vec4i_const V_CI_3 = { REPLICATE(3) };
DECL_VEC_CONST vec4i_const V_CI_4 = { REPLICATE(4) };
DECL_VEC_CONST vec4f_const V_C_UNIT_1000 = {1.0f, 0.0f, 0.0f, 0.0f};
DECL_VEC_CONST vec4f_const V_C_UNIT_0100 = {0.0f, 1.0f, 0.0f, 0.0f};
DECL_VEC_CONST vec4f_const V_C_UNIT_0010 = {0.0f, 0.0f, 1.0f, 0.0f};
DECL_VEC_CONST vec4f_const V_C_UNIT_0001 = {0.0f, 0.0f, 0.0f, 1.0f};
DECL_VEC_CONST vec4f_const V_C_UNIT_1110 = {1.0f, 1.0f, 1.0f, 0.0f};
DECL_VEC_CONST vec4f_const V_C_UNIT_0011 = {0.0f, 0.0f, 1.0f, 1.0f};
DECL_VEC_CONST vec4i_const V_CI_MASK0000 = { 0, 0, 0, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK1000 = { 0xFFFFFFFF, 0, 0, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK0100 = { 0, 0xFFFFFFFF, 0, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK0010 = { 0, 0, 0xFFFFFFFF, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK0001 = { 0, 0, 0, 0xFFFFFFFF };
DECL_VEC_CONST vec4i_const V_CI_MASK1100 = { 0xFFFFFFFF, 0xFFFFFFFF, 0, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK0110 = { 0, 0xFFFFFFFF, 0xFFFFFFFF, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK0011 = { 0, 0, 0xFFFFFFFF, 0xFFFFFFFF };
DECL_VEC_CONST vec4i_const V_CI_MASK1010 = { 0xFFFFFFFF, 0, 0xFFFFFFFF, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK0101 = { 0, 0xFFFFFFFF, 0xFFFFFFFF, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK1110 = { 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0 };
DECL_VEC_CONST vec4i_const V_CI_MASK1101 = { 0xFFFFFFFF, 0xFFFFFFFF, 0, 0xFFFFFFFF };
DECL_VEC_CONST vec4i_const V_CI_MASK1011 = { 0xFFFFFFFF, 0, 0xFFFFFFFF, 0xFFFFFFFF };
DECL_VEC_CONST vec4i_const V_CI_MASK0111 = { 0, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF };
DECL_VEC_CONST vec4i_const V_CI_MASK1111 = { 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF };
#elif _TARGET_SIMD_NEON
#define V_C_HALF vdupq_n_f32(0.5f)
#define V_C_HALF_MINUS_EPS vdupq_n_f32(0.5f - 1.192092896e-07f * 32)
#define V_C_ONE vdupq_n_f32(1.0f)
#define V_C_TWO vdupq_n_f32(2.0f)
#define V_C_PI vdupq_n_f32(3.14159265f) // pi
#define V_C_HALFPI vdupq_n_f32(3.14159265f / 2.f) // pi/2
#define V_C_TWOPI vdupq_n_f32(2.f * 3.14159265f) // pi*2
#define V_C_PI_DIV_4 vdupq_n_f32(0.78539816f) // pi/4
#define V_C_2_DIV_PI vdupq_n_f32(0.63661977f) // 2/pi
#define V_C_4_DIV_PI vdupq_n_f32(1.27323954f) // 4/pi
#define V_C_MAX_VAL vdupq_n_f32(1e32f)
#define V_C_MIN_VAL vdupq_n_f32(-1e32f)
#define V_C_EPS_VAL vdupq_n_f32(1.192092896e-07f)
#define V_C_VERY_SMALL_VAL vdupq_n_f32(4e-19f) // ~sqrt(FLT_MIN), safe threshold for passing argument to rcp(x)
#define V_CI_SIGN_MASK vdupq_n_s32(0x80000000)
#define V_CI_INV_SIGN_MASK vdupq_n_s32(0x7FFFFFFF)
#define V_CI_3FF vdupq_n_s32(0x3FF)
#define V_CI_070 vdupq_n_s32(0x70)
#define V_CI_0FF vdupq_n_s32(0xFF)
#define V_CI_0FF vdupq_n_s32(0xFF)
#define V_CI_01F vdupq_n_s32(0x1F)
#define V_CI_07FFFFF vdupq_n_s32(0x7fffff)
#define V_CI_0 vdupq_n_s32(0)
#define V_CI_1 vdupq_n_s32(1)
#define V_CI_2 vdupq_n_s32(2)
#define V_CI_3 vdupq_n_s32(3)
#define V_CI_4 vdupq_n_s32(4)
#if defined(__clang__) || defined(__GNUC__)
#define V_C_INF vdupq_n_f32(__builtin_inff())
#define DECL_VECFLOAT4(X,Y,Z,W) (float32x4_t){X,Y,Z,W}
#define DECL_VECUINT4(X,Y,Z,W) (int32x4_t){X,Y,Z,W}
#elif defined(_MSC_VER)
#define V_C_INF vdupq_n_f32(INFINITY)
constexpr struct vec4f_const_hlp
{
constexpr vec4f_const_hlp(float x, float y, float z, float w)
{
v.n128_f32[0] = x;
v.n128_f32[1] = y;
v.n128_f32[2] = z;
v.n128_f32[3] = w;
}
__n128 v;
};
constexpr struct vec4i_const_hlp
{
constexpr vec4i_const_hlp(int x, int y, int z, int w)
{
v.n128_i32[0] = x;
v.n128_i32[1] = y;
v.n128_i32[2] = z;
v.n128_i32[3] = w;
}
__n128 v;
};
#define DECL_VECFLOAT4(X,Y,Z,W) vec4f_const_hlp(X,Y,Z,W).v
#define DECL_VECUINT4(X,Y,Z,W) vec4i_const_hlp(X,Y,Z,W).v
#endif
#define V_C_UNIT_1000 DECL_VECFLOAT4( 1.0f, 0.0f, 0.0f, 0.0f )
#define V_C_UNIT_0100 DECL_VECFLOAT4( 0.0f, 1.0f, 0.0f, 0.0f )
#define V_C_UNIT_0010 DECL_VECFLOAT4( 0.0f, 0.0f, 1.0f, 0.0f )
#define V_C_UNIT_0001 DECL_VECFLOAT4( 0.0f, 0.0f, 0.0f, 1.0f )
#define V_C_UNIT_1110 DECL_VECFLOAT4( 1.0f, 1.0f, 1.0f, 0.0f )
#define V_C_UNIT_0011 DECL_VECFLOAT4( 0.0f, 0.0f, 1.0f, 1.0f )
#define V_CI_MASK0000 vdupq_n_s32(0)
#define V_CI_MASK1000 DECL_VECUINT4( -1, 0, 0, 0 )
#define V_CI_MASK0100 DECL_VECUINT4( 0, -1, 0, 0 )
#define V_CI_MASK0010 DECL_VECUINT4( 0, 0, -1, 0 )
#define V_CI_MASK0001 DECL_VECUINT4( 0, 0, 0, -1 )
#define V_CI_MASK1100 DECL_VECUINT4( -1, -1, 0, 0 )
#define V_CI_MASK0110 DECL_VECUINT4( 0, -1, -1, 0 )
#define V_CI_MASK0011 DECL_VECUINT4( 0, 0, -1, -1 )
#define V_CI_MASK1010 DECL_VECUINT4( -1, 0, -1, 0 )
#define V_CI_MASK0101 DECL_VECUINT4( 0, -1, 0, -1 )
#define V_CI_MASK1110 DECL_VECUINT4( -1, -1, -1, 0 )
#define V_CI_MASK1101 DECL_VECUINT4( -1, -1, 0, -1 )
#define V_CI_MASK1011 DECL_VECUINT4( -1, 0, -1, -1 )
#define V_CI_MASK0111 DECL_VECUINT4( 0, -1, -1, -1 )
#define V_CI_MASK1111 DECL_VECUINT4( -1, -1, -1, -1 )
#endif
#undef REPLICATE
#undef DECL_VEC_CONST