3 * Copyright (C) 2000-2003 Michel Lespinasse <walken@zoy.org>
4 * Copyright (C) 1999-2000 Aaron Holtzman <aholtzma@ess.engr.uvic.ca>
6 * This file is part of mpeg2dec, a free MPEG-2 video stream decoder.
7 * See http://libmpeg2.sourceforge.net/ for updates.
9 * mpeg2dec is free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License as published by
11 * the Free Software Foundation; either version 2 of the License, or
12 * (at your option) any later version.
14 * mpeg2dec is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 * GNU General Public License for more details.
19 * You should have received a copy of the GNU General Public License
20 * along with this program; if not, write to the Free Software
21 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
24 * libmpeg2 sync history:
25 * 2008-07-01 - CVS revision 1.36
30 #include "mpeg2dec_config.h"
33 #include "attributes.h"
34 #include "mpeg2_internal.h"
36 #if defined(CPU_COLDFIRE) || defined (CPU_ARM)
42 #define W1 2841 /* 2048 * sqrt (2) * cos (1 * pi / 16) */
43 #define W2 2676 /* 2048 * sqrt (2) * cos (2 * pi / 16) */
44 #define W3 2408 /* 2048 * sqrt (2) * cos (3 * pi / 16) */
45 #define W5 1609 /* 2048 * sqrt (2) * cos (5 * pi / 16) */
46 #define W6 1108 /* 2048 * sqrt (2) * cos (6 * pi / 16) */
47 #define W7 565 /* 2048 * sqrt (2) * cos (7 * pi / 16) */
50 * In legal streams, the IDCT output should be between -384 and +384.
51 * In corrupted streams, it is possible to force the IDCT output to go
52 * to +-3826 - this is the worst case for a column IDCT where the
53 * column inputs are 16-bit values.
56 ({ typeof (i) _i = (i); \
57 if ((_i & 0xff) != _i) \
58 _i = ~(_i >> (8*sizeof(_i) - 1)); \
62 #define BUTTERFLY(t0,t1,W0,W1,d0,d1) \
64 t0 = W0 * d0 + W1 * d1; \
65 t1 = W0 * d1 - W1 * d0; \
68 #define BUTTERFLY(t0,t1,W0,W1,d0,d1) \
70 int tmp = W0 * (d0 + d1); \
71 t0 = tmp + (W1 - W0) * d1; \
72 t1 = tmp - (W1 + W0) * d0; \
76 static inline void idct_row (int16_t * const block
)
79 int a0
, a1
, a2
, a3
, b0
, b1
, b2
, b3
;
83 if (likely (!(block
[1] | ((int32_t *)block
)[1] | ((int32_t *)block
)[2] |
84 ((int32_t *)block
)[3])))
86 uint32_t tmp
= (uint16_t) (block
[0] >> 1);
88 ((int32_t *)block
)[0] = tmp
;
89 ((int32_t *)block
)[1] = tmp
;
90 ((int32_t *)block
)[2] = tmp
;
91 ((int32_t *)block
)[3] = tmp
;
95 d0
= (block
[0] << 11) + 2048;
101 BUTTERFLY (t2
, t3
, W6
, W2
, d3
, d1
);
111 BUTTERFLY (t0
, t1
, W7
, W1
, d3
, d0
);
112 BUTTERFLY (t2
, t3
, W3
, W5
, d1
, d2
);
117 b1
= ((t0
+ t1
) >> 8) * 181;
118 b2
= ((t0
- t1
) >> 8) * 181;
120 block
[0] = (a0
+ b0
) >> 12;
121 block
[1] = (a1
+ b1
) >> 12;
122 block
[2] = (a2
+ b2
) >> 12;
123 block
[3] = (a3
+ b3
) >> 12;
124 block
[4] = (a3
- b3
) >> 12;
125 block
[5] = (a2
- b2
) >> 12;
126 block
[6] = (a1
- b1
) >> 12;
127 block
[7] = (a0
- b0
) >> 12;
130 static inline void idct_col (int16_t * const block
)
133 int a0
, a1
, a2
, a3
, b0
, b1
, b2
, b3
;
136 d0
= (block
[8*0] << 11) + 65536;
138 d2
= block
[8*2] << 11;
142 BUTTERFLY (t2
, t3
, W6
, W2
, d3
, d1
);
152 BUTTERFLY (t0
, t1
, W7
, W1
, d3
, d0
);
153 BUTTERFLY (t2
, t3
, W3
, W5
, d1
, d2
);
158 b1
= ((t0
+ t1
) >> 8) * 181;
159 b2
= ((t0
- t1
) >> 8) * 181;
161 block
[8*0] = (a0
+ b0
) >> 17;
162 block
[8*1] = (a1
+ b1
) >> 17;
163 block
[8*2] = (a2
+ b2
) >> 17;
164 block
[8*3] = (a3
+ b3
) >> 17;
165 block
[8*4] = (a3
- b3
) >> 17;
166 block
[8*5] = (a2
- b2
) >> 17;
167 block
[8*6] = (a1
- b1
) >> 17;
168 block
[8*7] = (a0
- b0
) >> 17;
171 void mpeg2_idct_copy (int16_t * block
, uint8_t * dest
,
176 for (i
= 0; i
< 8; i
++)
177 idct_row (block
+ 8 * i
);
179 for (i
= 0; i
< 8; i
++)
180 idct_col (block
+ i
);
184 dest
[0] = CLIP (block
[0]);
185 dest
[1] = CLIP (block
[1]);
186 dest
[2] = CLIP (block
[2]);
187 dest
[3] = CLIP (block
[3]);
188 dest
[4] = CLIP (block
[4]);
189 dest
[5] = CLIP (block
[5]);
190 dest
[6] = CLIP (block
[6]);
191 dest
[7] = CLIP (block
[7]);
193 ((int32_t *)block
)[0] = 0;
194 ((int32_t *)block
)[1] = 0;
195 ((int32_t *)block
)[2] = 0;
196 ((int32_t *)block
)[3] = 0;
204 void mpeg2_idct_add (const int last
, int16_t * block
,
205 uint8_t * dest
, const int stride
)
209 if (last
!= 129 || (block
[0] & (7 << 4)) == (4 << 4))
211 for (i
= 0; i
< 8; i
++)
212 idct_row (block
+ 8 * i
);
214 for (i
= 0; i
< 8; i
++)
215 idct_col (block
+ i
);
219 dest
[0] = CLIP (block
[0] + dest
[0]);
220 dest
[1] = CLIP (block
[1] + dest
[1]);
221 dest
[2] = CLIP (block
[2] + dest
[2]);
222 dest
[3] = CLIP (block
[3] + dest
[3]);
223 dest
[4] = CLIP (block
[4] + dest
[4]);
224 dest
[5] = CLIP (block
[5] + dest
[5]);
225 dest
[6] = CLIP (block
[6] + dest
[6]);
226 dest
[7] = CLIP (block
[7] + dest
[7]);
228 ((int32_t *)block
)[0] = 0;
229 ((int32_t *)block
)[1] = 0;
230 ((int32_t *)block
)[2] = 0;
231 ((int32_t *)block
)[3] = 0;
240 int DC
= (block
[0] + 64) >> 7;
241 block
[0] = block
[63] = 0;
246 dest
[0] = CLIP (DC
+ dest
[0]);
247 dest
[1] = CLIP (DC
+ dest
[1]);
248 dest
[2] = CLIP (DC
+ dest
[2]);
249 dest
[3] = CLIP (DC
+ dest
[3]);
250 dest
[4] = CLIP (DC
+ dest
[4]);
251 dest
[5] = CLIP (DC
+ dest
[5]);
252 dest
[6] = CLIP (DC
+ dest
[6]);
253 dest
[7] = CLIP (DC
+ dest
[7]);
260 #endif /* IDCT_ASM */
262 void mpeg2_idct_init (void)
266 for (i
= 0; i
< 64; i
++)
268 j
= default_mpeg2_scan_norm
[i
];
269 mpeg2_scan_norm
[i
] = ((j
& 0x36) >> 1) | ((j
& 0x09) << 2);
271 j
= default_mpeg2_scan_alt
[i
];
272 mpeg2_scan_alt
[i
] = ((j
& 0x36) >> 1) | ((j
& 0x09) << 2);