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
|
// Copyright (C) 2003-2009 Dolphin Project.
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, version 2.0.
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License 2.0 for more details.
// A copy of the GPL 2.0 should have been included with the program.
// If not, see http://www.gnu.org/licenses/
// Official SVN repository and contact information can be found at
// http://code.google.com/p/dolphin-emu/
#include <xmmintrin.h>
#include "Common.h"
#include "MathUtil.h"
#include <cmath>
namespace {
static u32 saved_sse_state = _mm_getcsr();
static const u32 default_sse_state = _mm_getcsr();
}
namespace MathUtil
{
int ClassifyFP(double dvalue)
{
// TODO: Optimize the below to be as fast as possible.
IntDouble value;
value.d = dvalue;
// 5 bits (C, <, >, =, ?)
// easy cases first
if (value.i == 0) {
// positive zero
return 0x2;
} else if (value.i == 0x8000000000000000ULL) {
// negative zero
return 0x12;
} else if (value.i == 0x7FF0000000000000ULL) {
// positive inf
return 0x5;
} else if (value.i == 0xFFF0000000000000ULL) {
// negative inf
return 0x9;
} else {
// OK let's dissect this thing.
int sign = value.i >> 63;
int exp = (int)((value.i >> 52) & 0x7FF);
if (exp >= 1 && exp <= 2046) {
// Nice normalized number.
if (sign) {
return 0x8; // negative
} else {
return 0x4; // positive
}
}
u64 mantissa = value.i & 0x000FFFFFFFFFFFFFULL;
if (exp == 0 && mantissa) {
// Denormalized number.
if (sign) {
return 0x18;
} else {
return 0x14;
}
} else if (exp == 0x7FF && mantissa /* && mantissa_top*/) {
return 0x11; // Quiet NAN
}
}
return 0x4;
}
} // namespace
void LoadDefaultSSEState()
{
_mm_setcsr(default_sse_state);
}
void LoadSSEState()
{
_mm_setcsr(saved_sse_state);
}
void SaveSSEState()
{
saved_sse_state = _mm_getcsr();
}
void MatrixMul(int n, const float *a, const float *b, float *result)
{
for(int i = 0; i < n; ++i) {
for(int j= 0; j < n; ++j) {
float temp = 0;
for(int k = 0; k < n; ++k) {
temp += a[i * n + k] * b[k * n + j];
}
result[i * n + j] = temp;
}
}
}
// Calculate sum of a float list
float MathFloatVectorSum(std::vector<float> Vec)
{
float Sum = 0.0;
for(int i = 0; i < Vec.size(); i++)
{
Sum += Vec.at(i);
}
return Sum;
}
void Matrix33::LoadIdentity(Matrix33 &mtx)
{
memset(mtx.data, 0, sizeof(mtx.data));
mtx.data[0] = 1.0f;
mtx.data[4] = 1.0f;
mtx.data[8] = 1.0f;
}
void Matrix33::RotateX(Matrix33 &mtx, float rad)
{
float s = sin(rad);
float c = cos(rad);
memset(mtx.data, 0, sizeof(mtx.data));
mtx.data[0] = 1;
mtx.data[4] = c;
mtx.data[5] = -s;
mtx.data[7] = s;
mtx.data[8] = c;
}
void Matrix33::RotateY(Matrix33 &mtx, float rad)
{
float s = sin(rad);
float c = cos(rad);
memset(mtx.data, 0, sizeof(mtx.data));
mtx.data[0] = c;
mtx.data[2] = s;
mtx.data[4] = 1;
mtx.data[6] = -s;
mtx.data[8] = c;
}
void Matrix33::Multiply(const Matrix33 &a, const Matrix33 &b, Matrix33 &result)
{
MatrixMul(3, a.data, b.data, result.data);
}
void Matrix33::Multiply(const Matrix33 &a, const float vec[3], float result[3])
{
for(int i = 0; i < 3; ++i) {
result[i] = 0;
for(int k = 0; k < 3; ++k) {
result[i] += a.data[i * 3 + k] * vec[k];
}
}
}
void Matrix44::LoadIdentity(Matrix44 &mtx)
{
memset(mtx.data, 0, sizeof(mtx.data));
mtx.data[0] = 1.0f;
mtx.data[5] = 1.0f;
mtx.data[10] = 1.0f;
mtx.data[15] = 1.0f;
}
void Matrix44::LoadMatrix33(Matrix44 &mtx, const Matrix33 &m33)
{
for(int i = 0; i < 3; ++i) {
for(int j = 0; j < 3; ++j) {
mtx.data[i * 4 + j] = m33.data[i * 3 + j];
}
}
for(int i = 0; i < 3; ++i) {
mtx.data[i * 4 + 3] = 0;
mtx.data[i + 12] = 0;
}
mtx.data[15] = 1.0f;
}
void Matrix44::Set(Matrix44 &mtx, const float mtxArray[16])
{
for(int i = 0; i < 16; ++i) {
mtx.data[i] = mtxArray[i];
}
}
void Matrix44::Translate(Matrix44 &mtx, const float vec[3])
{
LoadIdentity(mtx);
mtx.data[3] = vec[0];
mtx.data[7] = vec[1];
mtx.data[11] = vec[2];
}
void Matrix44::Multiply(const Matrix44 &a, const Matrix44 &b, Matrix44 &result)
{
MatrixMul(4, a.data, b.data, result.data);
}
|
