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
|
/**
* @file llviewerjointmesh.cpp
* @brief LLV4* class header file - vector processor enabled math
*
* Copyright (c) 2007-$CurrentYear$, Linden Research, Inc.
* $License$
*/
#ifndef LL_LLV4MATRIX3_H
#define LL_LLV4MATRIX3_H
#include "llv4math.h"
#include "llv4vector3.h"
#include "m3math.h" // for operator LLMatrix3()
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
// LLV4Matrix3
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
LL_LLV4MATH_ALIGN_PREFIX
class LLV4Matrix3
{
public:
union {
F32 mMatrix[LLV4_NUM_AXIS][LLV4_NUM_AXIS];
V4F32 mV[LLV4_NUM_AXIS];
};
void lerp(const LLV4Matrix3 &a, const LLV4Matrix3 &b, const F32 &w);
void multiply(const LLVector3 &a, LLVector3& out) const;
void multiply(const LLVector4 &a, LLV4Vector3& out) const;
void multiply(const LLVector3 &a, LLV4Vector3& out) const;
const LLV4Matrix3& transpose();
const LLV4Matrix3& operator=(const LLMatrix3& a);
operator LLMatrix3() const { return (reinterpret_cast<const LLMatrix4*>(const_cast<const F32*>(&mMatrix[0][0])))->getMat3(); }
friend LLVector3 operator*(const LLVector3& a, const LLV4Matrix3& b);
}
LL_LLV4MATH_ALIGN_POSTFIX;
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
// LLV4Matrix3 - SSE
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
#if LL_VECTORIZE
inline void LLV4Matrix3::lerp(const LLV4Matrix3 &a, const LLV4Matrix3 &b, const F32 &w)
{
__m128 vw = _mm_set1_ps(w);
mV[VX] = _mm_add_ps(_mm_mul_ps(_mm_sub_ps(b.mV[VX], a.mV[VX]), vw), a.mV[VX]); // ( b - a ) * w + a
mV[VY] = _mm_add_ps(_mm_mul_ps(_mm_sub_ps(b.mV[VY], a.mV[VY]), vw), a.mV[VY]);
mV[VZ] = _mm_add_ps(_mm_mul_ps(_mm_sub_ps(b.mV[VZ], a.mV[VZ]), vw), a.mV[VZ]);
}
inline void LLV4Matrix3::multiply(const LLVector3 &a, LLVector3& o) const
{
LLV4Vector3 j;
j.v = _mm_mul_ps(_mm_set1_ps(a.mV[VX]), mV[VX]); // ( ax * vx ) + ...
j.v = _mm_add_ps(j.v , _mm_mul_ps(_mm_set1_ps(a.mV[VY]), mV[VY]));
j.v = _mm_add_ps(j.v , _mm_mul_ps(_mm_set1_ps(a.mV[VZ]), mV[VZ]));
o.setVec(j.mV);
}
inline void LLV4Matrix3::multiply(const LLVector4 &a, LLV4Vector3& o) const
{
o.v = _mm_mul_ps(_mm_set1_ps(a.mV[VX]), mV[VX]); // ( ax * vx ) + ...
o.v = _mm_add_ps(o.v , _mm_mul_ps(_mm_set1_ps(a.mV[VY]), mV[VY]));
o.v = _mm_add_ps(o.v , _mm_mul_ps(_mm_set1_ps(a.mV[VZ]), mV[VZ]));
}
inline void LLV4Matrix3::multiply(const LLVector3 &a, LLV4Vector3& o) const
{
o.v = _mm_mul_ps(_mm_set1_ps(a.mV[VX]), mV[VX]); // ( ax * vx ) + ...
o.v = _mm_add_ps(o.v , _mm_mul_ps(_mm_set1_ps(a.mV[VY]), mV[VY]));
o.v = _mm_add_ps(o.v , _mm_mul_ps(_mm_set1_ps(a.mV[VZ]), mV[VZ]));
}
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
// LLV4Matrix3
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
#else
inline void LLV4Matrix3::lerp(const LLV4Matrix3 &a, const LLV4Matrix3 &b, const F32 &w)
{
mMatrix[VX][VX] = llv4lerp(a.mMatrix[VX][VX], b.mMatrix[VX][VX], w);
mMatrix[VX][VY] = llv4lerp(a.mMatrix[VX][VY], b.mMatrix[VX][VY], w);
mMatrix[VX][VZ] = llv4lerp(a.mMatrix[VX][VZ], b.mMatrix[VX][VZ], w);
mMatrix[VY][VX] = llv4lerp(a.mMatrix[VY][VX], b.mMatrix[VY][VX], w);
mMatrix[VY][VY] = llv4lerp(a.mMatrix[VY][VY], b.mMatrix[VY][VY], w);
mMatrix[VY][VZ] = llv4lerp(a.mMatrix[VY][VZ], b.mMatrix[VY][VZ], w);
mMatrix[VZ][VX] = llv4lerp(a.mMatrix[VZ][VX], b.mMatrix[VZ][VX], w);
mMatrix[VZ][VY] = llv4lerp(a.mMatrix[VZ][VY], b.mMatrix[VZ][VY], w);
mMatrix[VZ][VZ] = llv4lerp(a.mMatrix[VZ][VZ], b.mMatrix[VZ][VZ], w);
}
inline void LLV4Matrix3::multiply(const LLVector3 &a, LLVector3& o) const
{
o.setVec( a.mV[VX] * mMatrix[VX][VX] +
a.mV[VY] * mMatrix[VY][VX] +
a.mV[VZ] * mMatrix[VZ][VX],
a.mV[VX] * mMatrix[VX][VY] +
a.mV[VY] * mMatrix[VY][VY] +
a.mV[VZ] * mMatrix[VZ][VY],
a.mV[VX] * mMatrix[VX][VZ] +
a.mV[VY] * mMatrix[VY][VZ] +
a.mV[VZ] * mMatrix[VZ][VZ]);
}
inline void LLV4Matrix3::multiply(const LLVector4 &a, LLV4Vector3& o) const
{
o.setVec( a.mV[VX] * mMatrix[VX][VX] +
a.mV[VY] * mMatrix[VY][VX] +
a.mV[VZ] * mMatrix[VZ][VX],
a.mV[VX] * mMatrix[VX][VY] +
a.mV[VY] * mMatrix[VY][VY] +
a.mV[VZ] * mMatrix[VZ][VY],
a.mV[VX] * mMatrix[VX][VZ] +
a.mV[VY] * mMatrix[VY][VZ] +
a.mV[VZ] * mMatrix[VZ][VZ]);
}
inline void LLV4Matrix3::multiply(const LLVector3 &a, LLV4Vector3& o) const
{
o.setVec( a.mV[VX] * mMatrix[VX][VX] +
a.mV[VY] * mMatrix[VY][VX] +
a.mV[VZ] * mMatrix[VZ][VX],
a.mV[VX] * mMatrix[VX][VY] +
a.mV[VY] * mMatrix[VY][VY] +
a.mV[VZ] * mMatrix[VZ][VY],
a.mV[VX] * mMatrix[VX][VZ] +
a.mV[VY] * mMatrix[VY][VZ] +
a.mV[VZ] * mMatrix[VZ][VZ]);
}
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
// LLV4Matrix3
//-----------------------------------------------------------------------------
//-----------------------------------------------------------------------------
#endif
inline const LLV4Matrix3& LLV4Matrix3::transpose()
{
#if LL_VECTORIZE && defined(_MM_TRANSPOSE4_PS)
_MM_TRANSPOSE4_PS(mV[VX], mV[VY], mV[VZ], mV[VW]);
return *this;
#else
F32 temp;
temp = mMatrix[VX][VY]; mMatrix[VX][VY] = mMatrix[VY][VX]; mMatrix[VY][VX] = temp;
temp = mMatrix[VX][VZ]; mMatrix[VX][VZ] = mMatrix[VZ][VX]; mMatrix[VZ][VX] = temp;
temp = mMatrix[VY][VZ]; mMatrix[VY][VZ] = mMatrix[VZ][VY]; mMatrix[VZ][VY] = temp;
#endif
return *this;
}
inline const LLV4Matrix3& LLV4Matrix3::operator=(const LLMatrix3& a)
{
memcpy(mMatrix[VX], a.mMatrix[VX], sizeof(F32) * 3 );
memcpy(mMatrix[VY], a.mMatrix[VY], sizeof(F32) * 3 );
memcpy(mMatrix[VZ], a.mMatrix[VZ], sizeof(F32) * 3 );
return *this;
}
inline LLVector3 operator*(const LLVector3& a, const LLV4Matrix3& b)
{
return LLVector3(
a.mV[VX] * b.mMatrix[VX][VX] +
a.mV[VY] * b.mMatrix[VY][VX] +
a.mV[VZ] * b.mMatrix[VZ][VX],
a.mV[VX] * b.mMatrix[VX][VY] +
a.mV[VY] * b.mMatrix[VY][VY] +
a.mV[VZ] * b.mMatrix[VZ][VY],
a.mV[VX] * b.mMatrix[VX][VZ] +
a.mV[VY] * b.mMatrix[VY][VZ] +
a.mV[VZ] * b.mMatrix[VZ][VZ] );
}
#endif
|
