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
|
/*
* Fast C2P (Chunky-to-Planar) Conversion
*
* Copyright (C) 2003-2008 Geert Uytterhoeven
*
* NOTES:
* - This code was inspired by Scout's C2P tutorial
* - It assumes to run on a big endian system
*
* This file is subject to the terms and conditions of the GNU General Public
* License. See the file COPYING in the main directory of this archive
* for more details.
*/
#include <linux/module.h>
#include <linux/string.h>
#include <asm/unaligned.h>
#include "c2p.h"
/*
* Basic transpose step
*/
static inline void _transp(u32 d[], unsigned int i1, unsigned int i2,
unsigned int shift, u32 mask)
{
u32 t = (d[i1] ^ (d[i2] >> shift)) & mask;
d[i1] ^= t;
d[i2] ^= t << shift;
}
extern void c2p_unsupported(void);
static inline u32 get_mask(unsigned int n)
{
switch (n) {
case 1:
return 0x55555555;
case 2:
return 0x33333333;
case 4:
return 0x0f0f0f0f;
case 8:
return 0x00ff00ff;
case 16:
return 0x0000ffff;
}
c2p_unsupported();
return 0;
}
static inline void transp8(u32 d[], unsigned int n, unsigned int m)
{
u32 mask = get_mask(n);
switch (m) {
case 1:
/* First n x 1 block */
_transp(d, 0, 1, n, mask);
/* Second n x 1 block */
_transp(d, 2, 3, n, mask);
/* Third n x 1 block */
_transp(d, 4, 5, n, mask);
/* Fourth n x 1 block */
_transp(d, 6, 7, n, mask);
return;
case 2:
/* First n x 2 block */
_transp(d, 0, 2, n, mask);
_transp(d, 1, 3, n, mask);
/* Second n x 2 block */
_transp(d, 4, 6, n, mask);
_transp(d, 5, 7, n, mask);
return;
case 4:
/* Single n x 4 block */
_transp(d, 0, 4, n, mask);
_transp(d, 1, 5, n, mask);
_transp(d, 2, 6, n, mask);
_transp(d, 3, 7, n, mask);
return;
}
c2p_unsupported();
}
/*
* Perform a full C2P step on 32 8-bit pixels, stored in 8 32-bit words
* containing
* - 32 8-bit chunky pixels on input
* - permutated planar data (1 plane per 32-bit word) on output
*/
static void c2p_32x8(u32 d[8])
{
transp8(d, 16, 4);
transp8(d, 8, 2);
transp8(d, 4, 1);
transp8(d, 2, 4);
transp8(d, 1, 2);
}
/*
* Array containing the permutation indices of the planar data after c2p
*/
static const int perm_c2p_32x8[8] = { 7, 5, 3, 1, 6, 4, 2, 0 };
/*
* Compose two values, using a bitmask as decision value
* This is equivalent to (a & mask) | (b & ~mask)
*/
static inline u32 comp(u32 a, u32 b, u32 mask)
{
return ((a ^ b) & mask) ^ b;
}
/*
* Store a full block of planar data after c2p conversion
*/
static inline void store_planar(void *dst, u32 dst_inc, u32 bpp, u32 d[8])
{
int i;
for (i = 0; i < bpp; i++, dst += dst_inc)
put_unaligned_be32(d[perm_c2p_32x8[i]], dst);
}
/*
* Store a partial block of planar data after c2p conversion
*/
static inline void store_planar_masked(void *dst, u32 dst_inc, u32 bpp,
u32 d[8], u32 mask)
{
int i;
for (i = 0; i < bpp; i++, dst += dst_inc)
put_unaligned_be32(comp(d[perm_c2p_32x8[i]],
get_unaligned_be32(dst), mask),
dst);
}
/*
* c2p - Copy 8-bit chunky image data to a planar frame buffer
* @dst: Starting address of the planar frame buffer
* @dx: Horizontal destination offset (in pixels)
* @dy: Vertical destination offset (in pixels)
* @width: Image width (in pixels)
* @height: Image height (in pixels)
* @dst_nextline: Frame buffer offset to the next line (in bytes)
* @dst_nextplane: Frame buffer offset to the next plane (in bytes)
* @src_nextline: Image offset to the next line (in bytes)
* @bpp: Bits per pixel of the planar frame buffer (1-8)
*/
void c2p(void *dst, const void *src, u32 dx, u32 dy, u32 width, u32 height,
u32 dst_nextline, u32 dst_nextplane, u32 src_nextline, u32 bpp)
{
union {
u8 pixels[32];
u32 words[8];
} d;
u32 dst_idx, first, last, w;
const u8 *c;
void *p;
dst += dy*dst_nextline+(dx & ~31);
dst_idx = dx % 32;
first = 0xffffffffU >> dst_idx;
last = ~(0xffffffffU >> ((dst_idx+width) % 32));
while (height--) {
c = src;
p = dst;
w = width;
if (dst_idx+width <= 32) {
/* Single destination word */
first &= last;
memset(d.pixels, 0, sizeof(d));
memcpy(d.pixels+dst_idx, c, width);
c += width;
c2p_32x8(d.words);
store_planar_masked(p, dst_nextplane, bpp, d.words,
first);
p += 4;
} else {
/* Multiple destination words */
w = width;
/* Leading bits */
if (dst_idx) {
w = 32 - dst_idx;
memset(d.pixels, 0, dst_idx);
memcpy(d.pixels+dst_idx, c, w);
c += w;
c2p_32x8(d.words);
store_planar_masked(p, dst_nextplane, bpp,
d.words, first);
p += 4;
w = width-w;
}
/* Main chunk */
while (w >= 32) {
memcpy(d.pixels, c, 32);
c += 32;
c2p_32x8(d.words);
store_planar(p, dst_nextplane, bpp, d.words);
p += 4;
w -= 32;
}
/* Trailing bits */
w %= 32;
if (w > 0) {
memcpy(d.pixels, c, w);
memset(d.pixels+w, 0, 32-w);
c2p_32x8(d.words);
store_planar_masked(p, dst_nextplane, bpp,
d.words, last);
}
}
src += src_nextline;
dst += dst_nextline;
}
}
EXPORT_SYMBOL_GPL(c2p);
MODULE_LICENSE("GPL");
|
