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
|
/* WARNING: This file is generated by combine.pl from combine.inc.
Please edit one of those files rather than this one. */
#line 1 "combine.inc"
#define COMPONENT_SIZE 16
#define MASK 0xffffULL
#define ONE_HALF 0x8000ULL
#define G_SHIFT 16
#define B_SHIFT 16 * 2
#define A_SHIFT 16 * 3
#define G_MASK 0xffff0000ULL
#define B_MASK 0xffff00000000ULL
#define A_MASK 0xffff000000000000ULL
#define RB_MASK 0xffff0000ffffULL
#define AG_MASK 0xffff0000ffff0000ULL
#define RB_ONE_HALF 0x800000008000ULL
#define RB_MASK_PLUS_ONE 0x10000000010000ULL
#define Alpha(x) ((x) >> A_SHIFT)
/*
* Helper macros.
*/
#define IntMult(a,b,t) ( (t) = (a) * (b) + ONE_HALF, ( ( ( (t)>>G_SHIFT ) + (t) )>>G_SHIFT ) )
#define IntDiv(a,b) (((uint32_t) (a) * MASK) / (b))
#define GetComp(v,i) ((uint32_t) (uint16_t) ((v) >> i))
#define Add(x,y,i,t) ((t) = GetComp(x,i) + GetComp(y,i), \
(uint64_t) ((uint16_t) ((t) | (0 - ((t) >> G_SHIFT)))) << (i))
#define FbGen(x,y,i,ax,ay,t,u,v) ((t) = (IntMult(GetComp(y,i),ay,(u)) + \
IntMult(GetComp(x,i),ax,(v))), \
(uint64_t) ((uint16_t) ((t) | \
(0 - ((t) >> G_SHIFT)))) << (i))
/*
The methods below use some tricks to be able to do two color
components at the same time.
*/
/*
x_c = (x_c * a) / 255
*/
#define FbByteMul(x, a) do { \
uint64_t t = ((x & RB_MASK) * a) + RB_ONE_HALF; \
t = (t + ((t >> COMPONENT_SIZE) & RB_MASK)) >> COMPONENT_SIZE; \
t &= RB_MASK; \
\
x = (((x >> COMPONENT_SIZE) & RB_MASK) * a) + RB_ONE_HALF; \
x = (x + ((x >> COMPONENT_SIZE) & RB_MASK)); \
x &= RB_MASK << COMPONENT_SIZE; \
x += t; \
} while (0)
/*
x_c = (x_c * a) / 255 + y
*/
#define FbByteMulAdd(x, a, y) do { \
/* multiply and divide: trunc((i + 128)*257/65536) */ \
uint64_t t = ((x & RB_MASK) * a) + RB_ONE_HALF; \
t = (t + ((t >> COMPONENT_SIZE) & RB_MASK)) >> COMPONENT_SIZE; \
t &= RB_MASK; \
\
/* add */ \
t += y & RB_MASK; \
\
/* saturate */ \
t |= RB_MASK_PLUS_ONE - ((t >> COMPONENT_SIZE) & RB_MASK); \
t &= RB_MASK; \
\
/* multiply and divide */ \
x = (((x >> COMPONENT_SIZE) & RB_MASK) * a) + RB_ONE_HALF; \
x = (x + ((x >> COMPONENT_SIZE) & RB_MASK)) >> COMPONENT_SIZE; \
x &= RB_MASK; \
\
/* add */ \
x += (y >> COMPONENT_SIZE) & RB_MASK; \
\
/* saturate */ \
x |= RB_MASK_PLUS_ONE - ((x >> COMPONENT_SIZE) & RB_MASK); \
x &= RB_MASK; \
\
/* recombine */ \
x <<= COMPONENT_SIZE; \
x += t; \
} while (0)
/*
x_c = (x_c * a + y_c * b) / 255
*/
#define FbByteAddMul(x, a, y, b) do { \
uint64_t t; \
uint64_t r = (x >> A_SHIFT) * a + (y >> A_SHIFT) * b + ONE_HALF; \
r += (r >> G_SHIFT); \
r >>= G_SHIFT; \
\
t = (x & G_MASK) * a + (y & G_MASK) * b; \
t += (t >> G_SHIFT) + (ONE_HALF << G_SHIFT); \
t >>= B_SHIFT; \
\
t |= r << B_SHIFT; \
t |= RB_MASK_PLUS_ONE - ((t >> G_SHIFT) & RB_MASK); \
t &= RB_MASK; \
t <<= G_SHIFT; \
\
r = ((x >> B_SHIFT) & MASK) * a + \
((y >> B_SHIFT) & MASK) * b + ONE_HALF; \
r += (r >> G_SHIFT); \
r >>= G_SHIFT; \
\
x = (x & MASK) * a + (y & MASK) * b + ONE_HALF; \
x += (x >> G_SHIFT); \
x >>= G_SHIFT; \
x |= r << B_SHIFT; \
x |= RB_MASK_PLUS_ONE - ((x >> G_SHIFT) & RB_MASK); \
x &= RB_MASK; \
x |= t; \
} while (0)
/*
x_c = (x_c * a + y_c *b) / 256
*/
#define FbByteAddMul_256(x, a, y, b) do { \
uint64_t t = (x & RB_MASK) * a + (y & RB_MASK) * b; \
t >>= G_SHIFT; \
t &= RB_MASK; \
\
x = ((x >> G_SHIFT) & RB_MASK) * a + \
((y >> G_SHIFT) & RB_MASK) * b; \
x &= AG_MASK; \
x += t; \
} while (0)
/*
x_c = (x_c * a_c) / 255
*/
#define FbByteMulC(x, a) do { \
uint64_t t; \
uint64_t r = (x & MASK) * (a & MASK); \
r |= (x & B_MASK) * ((a >> B_SHIFT) & MASK); \
r += RB_ONE_HALF; \
r = (r + ((r >> G_SHIFT) & RB_MASK)) >> G_SHIFT; \
r &= RB_MASK; \
\
x >>= G_SHIFT; \
t = (x & MASK) * ((a >> G_SHIFT) & MASK); \
t |= (x & B_MASK) * (a >> A_SHIFT); \
t += RB_ONE_HALF; \
t = t + ((t >> G_SHIFT) & RB_MASK); \
x = r | (t & AG_MASK); \
} while (0)
/*
x_c = (x_c * a) / 255 + y
*/
#define FbByteMulAddC(x, a, y) do { \
uint64_t t; \
uint64_t r = (x & MASK) * (a & MASK); \
r |= (x & B_MASK) * ((a >> B_SHIFT) & MASK); \
r += RB_ONE_HALF; \
r = (r + ((r >> G_SHIFT) & RB_MASK)) >> G_SHIFT; \
r &= RB_MASK; \
r += y & RB_MASK; \
r |= RB_MASK_PLUS_ONE - ((r >> G_SHIFT) & RB_MASK); \
r &= RB_MASK; \
\
x >>= G_SHIFT; \
t = (x & MASK) * ((a >> G_SHIFT) & MASK); \
t |= (x & B_MASK) * (a >> A_SHIFT); \
t += RB_ONE_HALF; \
t = (t + ((t >> G_SHIFT) & RB_MASK)) >> G_SHIFT; \
t &= RB_MASK; \
t += (y >> G_SHIFT) & RB_MASK; \
t |= RB_MASK_PLUS_ONE - ((t >> G_SHIFT) & RB_MASK); \
t &= RB_MASK; \
x = r | (t << G_SHIFT); \
} while (0)
/*
x_c = (x_c * a_c + y_c * b) / 255
*/
#define FbByteAddMulC(x, a, y, b) do { \
uint64_t t; \
uint64_t r = (x >> A_SHIFT) * (a >> A_SHIFT) + \
(y >> A_SHIFT) * b; \
r += (r >> G_SHIFT) + ONE_HALF; \
r >>= G_SHIFT; \
\
t = (x & G_MASK) * ((a >> G_SHIFT) & MASK) + (y & G_MASK) * b; \
t += (t >> G_SHIFT) + (ONE_HALF << G_SHIFT); \
t >>= B_SHIFT; \
\
t |= r << B_SHIFT; \
t |= RB_MASK_PLUS_ONE - ((t >> G_SHIFT) & RB_MASK); \
t &= RB_MASK; \
t <<= G_SHIFT; \
\
r = ((x >> B_SHIFT) & MASK) * ((a >> B_SHIFT) & MASK) + \
((y >> B_SHIFT) & MASK) * b + ONE_HALF; \
r += (r >> G_SHIFT); \
r >>= G_SHIFT; \
\
x = (x & MASK) * (a & MASK) + (y & MASK) * b + ONE_HALF; \
x += (x >> G_SHIFT); \
x >>= G_SHIFT; \
x |= r << B_SHIFT; \
x |= RB_MASK_PLUS_ONE - ((x >> G_SHIFT) & RB_MASK); \
x &= RB_MASK; \
x |= t; \
} while (0)
/*
x_c = min(x_c + y_c, 255)
*/
#define FbByteAdd(x, y) do { \
uint64_t t; \
uint64_t r = (x & RB_MASK) + (y & RB_MASK); \
r |= RB_MASK_PLUS_ONE - ((r >> G_SHIFT) & RB_MASK); \
r &= RB_MASK; \
\
t = ((x >> G_SHIFT) & RB_MASK) + ((y >> G_SHIFT) & RB_MASK); \
t |= RB_MASK_PLUS_ONE - ((t >> G_SHIFT) & RB_MASK); \
r |= (t & RB_MASK) << G_SHIFT; \
x = r; \
} while (0)
|