Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Newer
Older
100644 749 lines (619 sloc) 28.091 kb
ff4ec49 license/copyright change
Fabrice Bellard authored
1 /*
2 * DSP utils
3 * Copyright (c) 2000, 2001, 2002 Fabrice Bellard.
8f2ab83 @michaelni copyright year update of the files i touched and remembered, things look...
michaelni authored
4 * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
ff4ec49 license/copyright change
Fabrice Bellard authored
5 *
b78e719 @DonDiego Change license headers to say 'FFmpeg' instead of 'this program/this lib...
DonDiego authored
6 * This file is part of FFmpeg.
7 *
8 * FFmpeg is free software; you can redistribute it and/or
ff4ec49 license/copyright change
Fabrice Bellard authored
9 * modify it under the terms of the GNU Lesser General Public
10 * License as published by the Free Software Foundation; either
b78e719 @DonDiego Change license headers to say 'FFmpeg' instead of 'this program/this lib...
DonDiego authored
11 * version 2.1 of the License, or (at your option) any later version.
ff4ec49 license/copyright change
Fabrice Bellard authored
12 *
b78e719 @DonDiego Change license headers to say 'FFmpeg' instead of 'this program/this lib...
DonDiego authored
13 * FFmpeg is distributed in the hope that it will be useful,
ff4ec49 license/copyright change
Fabrice Bellard authored
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
16 * Lesser General Public License for more details.
17 *
18 * You should have received a copy of the GNU Lesser General Public
b78e719 @DonDiego Change license headers to say 'FFmpeg' instead of 'this program/this lib...
DonDiego authored
19 * License along with FFmpeg; if not, write to the Free Software
5509bff @DonDiego Update licensing information: The FSF changed postal address.
DonDiego authored
20 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
ff4ec49 license/copyright change
Fabrice Bellard authored
21 */
2464118 @michaelni more dox
michaelni authored
22
23 /**
24 * @file dsputil.h
983e324 @michaelni per file doxy
michaelni authored
25 * DSP utils.
5755c27 @michaelni doc fixes (forgot to commit)
michaelni authored
26 * note, many functions in here may use MMX which trashes the FPU state, it is
27 * absolutely necessary to call emms_c() between dsp & float/double code
2464118 @michaelni more dox
michaelni authored
28 */
29
de6d9b6 Initial revision
Fabrice Bellard authored
30 #ifndef DSPUTIL_H
31 #define DSPUTIL_H
32
43f1708 - Added PSNR feature to libavcodec and ffmpeg. By now just Y PSNR until ...
Juanjo authored
33 #include "avcodec.h"
de6d9b6 Initial revision
Fabrice Bellard authored
34
2464118 @michaelni more dox
michaelni authored
35
44eb495 @michaelni divx5-gmc support
michaelni authored
36 //#define DEBUG
de6d9b6 Initial revision
Fabrice Bellard authored
37 /* dct code */
38 typedef short DCTELEM;
059715a First part of a series of speed-enchancing patches.
Robert Edele authored
39 typedef int DWTELEM;
de6d9b6 Initial revision
Fabrice Bellard authored
40
03c94ed avoid name clash with libjpeg - added missing externs
Fabrice Bellard authored
41 void fdct_ifast (DCTELEM *data);
48b1f80 @rvs * adding integer/floating point AAN implementations for DCT 2-4-8
rvs authored
42 void fdct_ifast248 (DCTELEM *data);
28db7fc @michaelni slow but accurate integer dct from IJG (should be ok with the LGPL as th...
michaelni authored
43 void ff_jpeg_fdct_islow (DCTELEM *data);
10acc47 @rvs * introducing dct248 into the DSP context.
rvs authored
44 void ff_fdct248_islow (DCTELEM *data);
de6d9b6 Initial revision
Fabrice Bellard authored
45
46 void j_rev_dct (DCTELEM *data);
178fcca @michaelni 1/2 resolution decoding
michaelni authored
47 void j_rev_dct4 (DCTELEM *data);
9ca358b @michaelni 1/4 resolution decoding
michaelni authored
48 void j_rev_dct2 (DCTELEM *data);
1aa8c57 @michaelni 1/8 resolution decoding
michaelni authored
49 void j_rev_dct1 (DCTELEM *data);
de6d9b6 Initial revision
Fabrice Bellard authored
50
3f09f52 @arpitest fdct_mmx -> ff_fdct_mmx (renamed to avoid namespace conflict with xvid)
arpitest authored
51 void ff_fdct_mmx(DCTELEM *block);
cf3bf5b @michaelni minor mmx2 optimization if the dct
michaelni authored
52 void ff_fdct_mmx2(DCTELEM *block);
8fd19ab @michaelni SSE2 fdct by (Balatoni Denes <pnis at coder dot hu>)
michaelni authored
53 void ff_fdct_sse2(DCTELEM *block);
de6d9b6 Initial revision
Fabrice Bellard authored
54
43efd19 decode H.264 with 8x8 transform.
Loren Merritt authored
55 void ff_h264_idct8_add_c(uint8_t *dst, DCTELEM *block, int stride);
0fa8158 @michaelni move h264 idct to its own file and call via function pointer in DspConte...
michaelni authored
56 void ff_h264_idct_add_c(uint8_t *dst, DCTELEM *block, int stride);
ef9d1d1 h264: special case dc-only idct. ~1% faster overall
Loren Merritt authored
57 void ff_h264_idct8_dc_add_c(uint8_t *dst, DCTELEM *block, int stride);
58 void ff_h264_idct_dc_add_c(uint8_t *dst, DCTELEM *block, int stride);
0fa8158 @michaelni move h264 idct to its own file and call via function pointer in DspConte...
michaelni authored
59 void ff_h264_lowres_idct_add_c(uint8_t *dst, int stride, DCTELEM *block);
60 void ff_h264_lowres_idct_put_c(uint8_t *dst, int stride, DCTELEM *block);
61
eb4825b sse and 3dnow implementations of float->int conversion and mdct windowin...
Loren Merritt authored
62 void ff_vector_fmul_add_add_c(float *dst, const float *src0, const float *src1,
63 const float *src2, int src3, int blocksize, int step);
64 void ff_float_to_int16_c(int16_t *dst, const float *src, int len);
65
e0eac44 added block permutation functions
Fabrice Bellard authored
66 /* encoding scans */
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
67 extern const uint8_t ff_alternate_horizontal_scan[64];
68 extern const uint8_t ff_alternate_vertical_scan[64];
69 extern const uint8_t ff_zigzag_direct[64];
10acc47 @rvs * introducing dct248 into the DSP context.
rvs authored
70 extern const uint8_t ff_zigzag248_direct[64];
5a24083 @michaelni fix for non-mmx runtimedetect encoding bugs - patch by Michael Niedermay...
michaelni authored
71
de6d9b6 Initial revision
Fabrice Bellard authored
72 /* pixel operations */
f2e92ef @michaelni increase size of crop table
michaelni authored
73 #define MAX_NEG_CROP 1024
de6d9b6 Initial revision
Fabrice Bellard authored
74
75 /* temporary */
1d50395 @mansr rename squareTbl -> ff_squareTbl
mansr authored
76 extern uint32_t ff_squareTbl[512];
55fde95 @mansr rename cropTbl -> ff_cropTbl
mansr authored
77 extern uint8_t ff_cropTbl[256 + 2 * MAX_NEG_CROP];
de6d9b6 Initial revision
Fabrice Bellard authored
78
44cb64e @multimediamike seperated out the C-based VP3 DSP functions into a different file; also
multimediamike authored
79 /* VP3 DSP functions */
8b6103d @michaelni porting vp3 idct over to lavc idct api
michaelni authored
80 void ff_vp3_idct_c(DCTELEM *block/* align 16*/);
81 void ff_vp3_idct_put_c(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*align 16*/);
82 void ff_vp3_idct_add_c(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*align 16*/);
de6d9b6 Initial revision
Fabrice Bellard authored
83
54009d4 @michaelni remove mpegvideo.c img resample dependancy
michaelni authored
84 /* 1/2^n downscaling functions from imgconvert.c */
85 void ff_img_copy_plane(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
86 void ff_shrink22(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
87 void ff_shrink44(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
88 void ff_shrink88(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
703c819 mmx implementation of 3-point GMC. (5x faster than C)
Loren Merritt authored
89
90 void ff_gmc_c(uint8_t *dst, uint8_t *src, int stride, int h, int ox, int oy,
91 int dxx, int dxy, int dyx, int dyy, int shift, int r, int width, int height);
54009d4 @michaelni remove mpegvideo.c img resample dependancy
michaelni authored
92
b7c27ee @michaelni alignment info
michaelni authored
93 /* minimum alignment rules ;)
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
94 if u notice errors in the align stuff, need more alignment for some asm code for some cpu
b7c27ee @michaelni alignment info
michaelni authored
95 or need to use a function with less aligned data then send a mail to the ffmpeg-dev list, ...
96
97 !warning these alignments might not match reallity, (missing attribute((align)) stuff somewhere possible)
98 i (michael) didnt check them, these are just the alignents which i think could be reached easily ...
de6d9b6 Initial revision
Fabrice Bellard authored
99
b7c27ee @michaelni alignment info
michaelni authored
100 !future video codecs might need functions with less strict alignment
101 */
102
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
103 /*
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
104 void get_pixels_c(DCTELEM *block, const uint8_t *pixels, int line_size);
105 void diff_pixels_c(DCTELEM *block, const uint8_t *s1, const uint8_t *s2, int stride);
106 void put_pixels_clamped_c(const DCTELEM *block, uint8_t *pixels, int line_size);
107 void add_pixels_clamped_c(const DCTELEM *block, uint8_t *pixels, int line_size);
649c00c @michaelni sizeof(s->block) isnt 64*6*2 anymore bugfix
michaelni authored
108 void clear_blocks_c(DCTELEM *blocks);
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
109 */
de6d9b6 Initial revision
Fabrice Bellard authored
110
111 /* add and put pixel (decoding) */
b7c27ee @michaelni alignment info
michaelni authored
112 // blocksizes for op_pixels_func are 8x4,8x8 16x8 16x16
7d67aa9 @michaelni info about h and tpel
michaelni authored
113 //h for op_pixels_func is limited to {width/2, width} but never larger than 16 and never smaller then 4
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
114 typedef void (*op_pixels_func)(uint8_t *block/*align width (8 or 16)*/, const uint8_t *pixels/*align 1*/, int line_size, int h);
669ac79 @michaelni moving the svq3 motion compensation stuff to dsputil (this also means th...
michaelni authored
115 typedef void (*tpel_mc_func)(uint8_t *block/*align width (8 or 16)*/, const uint8_t *pixels/*align 1*/, int line_size, int w, int h);
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
116 typedef void (*qpel_mc_func)(uint8_t *dst/*align width (8 or 16)*/, uint8_t *src/*align 1*/, int stride);
0da7126 @michaelni H264 decoder & demuxer
michaelni authored
117 typedef void (*h264_chroma_mc_func)(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int srcStride, int h, int x, int y);
9f2d1b4 H.264 weighted prediction.
Loren Merritt authored
118 typedef void (*h264_weight_func)(uint8_t *block, int stride, int log2_denom, int weight, int offset);
e8b5620 tweak h264_biweight
Loren Merritt authored
119 typedef void (*h264_biweight_func)(uint8_t *dst, uint8_t *src, int stride, int log2_denom, int weightd, int weights, int offset);
b318477 @michaelni put/avg_pixels16
michaelni authored
120
db79495 @michaelni qpel fix
michaelni authored
121 #define DEF_OLD_QPEL(name)\
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
122 void ff_put_ ## name (uint8_t *dst/*align width (8 or 16)*/, uint8_t *src/*align 1*/, int stride);\
123 void ff_put_no_rnd_ ## name (uint8_t *dst/*align width (8 or 16)*/, uint8_t *src/*align 1*/, int stride);\
124 void ff_avg_ ## name (uint8_t *dst/*align width (8 or 16)*/, uint8_t *src/*align 1*/, int stride);
db79495 @michaelni qpel fix
michaelni authored
125
126 DEF_OLD_QPEL(qpel16_mc11_old_c)
127 DEF_OLD_QPEL(qpel16_mc31_old_c)
128 DEF_OLD_QPEL(qpel16_mc12_old_c)
129 DEF_OLD_QPEL(qpel16_mc32_old_c)
130 DEF_OLD_QPEL(qpel16_mc13_old_c)
131 DEF_OLD_QPEL(qpel16_mc33_old_c)
132 DEF_OLD_QPEL(qpel8_mc11_old_c)
133 DEF_OLD_QPEL(qpel8_mc31_old_c)
134 DEF_OLD_QPEL(qpel8_mc12_old_c)
135 DEF_OLD_QPEL(qpel8_mc32_old_c)
136 DEF_OLD_QPEL(qpel8_mc13_old_c)
137 DEF_OLD_QPEL(qpel8_mc33_old_c)
b318477 @michaelni put/avg_pixels16
michaelni authored
138
139 #define CALL_2X_PIXELS(a, b, n)\
140 static void a(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
141 b(block , pixels , line_size, h);\
142 b(block+n, pixels+n, line_size, h);\
143 }
44eb495 @michaelni divx5-gmc support
michaelni authored
144
de6d9b6 Initial revision
Fabrice Bellard authored
145 /* motion estimation */
7d67aa9 @michaelni info about h and tpel
michaelni authored
146 // h is limited to {width/2, width, 2*width} but never larger than 16 and never smaller then 2
755bfea @DonDiego misc spelling fixes
DonDiego authored
147 // although currently h<4 is not used as functions with width <8 are neither used nor implemented
bb198e1 @michaelni interlaced motion estimation
michaelni authored
148 typedef int (*me_cmp_func)(void /*MpegEncContext*/ *s, uint8_t *blk1/*align width (8 or 16)*/, uint8_t *blk2/*align 1*/, int line_size, int h)/* __attribute__ ((const))*/;
1457ab5 @michaelni qpel encoding
michaelni authored
149
0da7126 @michaelni H264 decoder & demuxer
michaelni authored
150
059715a First part of a series of speed-enchancing patches.
Robert Edele authored
151 // for snow slices
152 typedef struct slice_buffer_s slice_buffer;
153
2464118 @michaelni more dox
michaelni authored
154 /**
155 * DSPContext.
156 */
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
157 typedef struct DSPContext {
158 /* pixel ops : interface with DCT */
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
159 void (*get_pixels)(DCTELEM *block/*align 16*/, const uint8_t *pixels/*align 8*/, int line_size);
160 void (*diff_pixels)(DCTELEM *block/*align 16*/, const uint8_t *s1/*align 8*/, const uint8_t *s2/*align 8*/, int stride);
161 void (*put_pixels_clamped)(const DCTELEM *block/*align 16*/, uint8_t *pixels/*align 8*/, int line_size);
f9ed9d8 @multimediamike separate out put_signed_pixels_clamped() into its own function and
multimediamike authored
162 void (*put_signed_pixels_clamped)(const DCTELEM *block/*align 16*/, uint8_t *pixels/*align 8*/, int line_size);
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
163 void (*add_pixels_clamped)(const DCTELEM *block/*align 16*/, uint8_t *pixels/*align 8*/, int line_size);
36940ec H.264 lossless mode
Loren Merritt authored
164 void (*add_pixels8)(uint8_t *pixels, DCTELEM *block, int line_size);
165 void (*add_pixels4)(uint8_t *pixels, DCTELEM *block, int line_size);
1edbfe1 factor sum_abs_dctelem out of dct_sad, and simd it.
Loren Merritt authored
166 int (*sum_abs_dctelem)(DCTELEM *block/*align 16*/);
d518aeb @michaelni halfpel MC doxy
michaelni authored
167 /**
168 * translational global motion compensation.
169 */
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
170 void (*gmc1)(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int srcStride, int h, int x16, int y16, int rounder);
d518aeb @michaelni halfpel MC doxy
michaelni authored
171 /**
172 * global motion compensation.
173 */
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
174 void (*gmc )(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int ox, int oy,
bb270c0 @DonDiego COSMETICS: tabs --> spaces, some prettyprinting
DonDiego authored
175 int dxx, int dxy, int dyx, int dyy, int shift, int r, int width, int height);
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
176 void (*clear_blocks)(DCTELEM *blocks/*align 16*/);
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
177 int (*pix_sum)(uint8_t * pix, int line_size);
178 int (*pix_norm1)(uint8_t * pix, int line_size);
bb198e1 @michaelni interlaced motion estimation
michaelni authored
179 // 16x16 8x8 4x4 2x2 16x8 8x4 4x2 8x16 4x8 2x4
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
180
622348f @michaelni interlaced dct decision cleanup
michaelni authored
181 me_cmp_func sad[5]; /* identical to pix_absAxA except additional void * */
182 me_cmp_func sse[5];
183 me_cmp_func hadamard8_diff[5];
184 me_cmp_func dct_sad[5];
185 me_cmp_func quant_psnr[5];
186 me_cmp_func bit[5];
187 me_cmp_func rd[5];
188 me_cmp_func vsad[5];
189 me_cmp_func vsse[5];
e6a2ac3 @michaelni noise preserving sum of squares comparission function
michaelni authored
190 me_cmp_func nsse[5];
26efc54 @michaelni 4x4 SSE compare function
michaelni authored
191 me_cmp_func w53[5];
192 me_cmp_func w97[5];
0fd6aea @michaelni more flexible frame skip decission
michaelni authored
193 me_cmp_func dct_max[5];
27c61ac @michaelni 8x8 integer dct from x264 as cmp function (under CONFIG_GPL)
michaelni authored
194 me_cmp_func dct264_sad[5];
1457ab5 @michaelni qpel encoding
michaelni authored
195
bb198e1 @michaelni interlaced motion estimation
michaelni authored
196 me_cmp_func me_pre_cmp[5];
197 me_cmp_func me_cmp[5];
198 me_cmp_func me_sub_cmp[5];
199 me_cmp_func mb_cmp[5];
622348f @michaelni interlaced dct decision cleanup
michaelni authored
200 me_cmp_func ildct_cmp[5]; //only width 16 used
0fd6aea @michaelni more flexible frame skip decission
michaelni authored
201 me_cmp_func frame_skip_cmp[5]; //only width 8 used
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
202
5900637 mmx 16-bit ssd. 2.3x faster svq1 encoding.
Loren Merritt authored
203 int (*ssd_int8_vs_int16)(int8_t *pix1, int16_t *pix2, int size);
204
d518aeb @michaelni halfpel MC doxy
michaelni authored
205 /**
206 * Halfpel motion compensation with rounding (a+b+1)>>1.
30f1505 @lu-zero Typos in comments
lu-zero authored
207 * this is an array[4][4] of motion compensation functions for 4
e5771f4 @michaelni put_pixels width comment fix
michaelni authored
208 * horizontal blocksizes (8,16) and the 4 halfpel positions<br>
5755c27 @michaelni doc fixes (forgot to commit)
michaelni authored
209 * *pixels_tab[ 0->16xH 1->8xH ][ xhalfpel + 2*yhalfpel ]
d518aeb @michaelni halfpel MC doxy
michaelni authored
210 * @param block destination where the result is stored
211 * @param pixels source
212 * @param line_size number of bytes in a horizontal line of block
213 * @param h height
214 */
669ac79 @michaelni moving the svq3 motion compensation stuff to dsputil (this also means th...
michaelni authored
215 op_pixels_func put_pixels_tab[4][4];
d518aeb @michaelni halfpel MC doxy
michaelni authored
216
217 /**
218 * Halfpel motion compensation with rounding (a+b+1)>>1.
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
219 * This is an array[4][4] of motion compensation functions for 4
e5771f4 @michaelni put_pixels width comment fix
michaelni authored
220 * horizontal blocksizes (8,16) and the 4 halfpel positions<br>
5755c27 @michaelni doc fixes (forgot to commit)
michaelni authored
221 * *pixels_tab[ 0->16xH 1->8xH ][ xhalfpel + 2*yhalfpel ]
d518aeb @michaelni halfpel MC doxy
michaelni authored
222 * @param block destination into which the result is averaged (a+b+1)>>1
223 * @param pixels source
224 * @param line_size number of bytes in a horizontal line of block
225 * @param h height
226 */
da3b975 @multimediamike added support for B-frames and multiple slices
multimediamike authored
227 op_pixels_func avg_pixels_tab[4][4];
d518aeb @michaelni halfpel MC doxy
michaelni authored
228
229 /**
230 * Halfpel motion compensation with no rounding (a+b)>>1.
30f1505 @lu-zero Typos in comments
lu-zero authored
231 * this is an array[2][4] of motion compensation functions for 2
eb14c71 @michaelni doxy
michaelni authored
232 * horizontal blocksizes (8,16) and the 4 halfpel positions<br>
5755c27 @michaelni doc fixes (forgot to commit)
michaelni authored
233 * *pixels_tab[ 0->16xH 1->8xH ][ xhalfpel + 2*yhalfpel ]
d518aeb @michaelni halfpel MC doxy
michaelni authored
234 * @param block destination where the result is stored
235 * @param pixels source
236 * @param line_size number of bytes in a horizontal line of block
237 * @param h height
238 */
dbc56b3 @michaelni chroma me fix
michaelni authored
239 op_pixels_func put_no_rnd_pixels_tab[4][4];
d518aeb @michaelni halfpel MC doxy
michaelni authored
240
241 /**
242 * Halfpel motion compensation with no rounding (a+b)>>1.
30f1505 @lu-zero Typos in comments
lu-zero authored
243 * this is an array[2][4] of motion compensation functions for 2
eb14c71 @michaelni doxy
michaelni authored
244 * horizontal blocksizes (8,16) and the 4 halfpel positions<br>
5755c27 @michaelni doc fixes (forgot to commit)
michaelni authored
245 * *pixels_tab[ 0->16xH 1->8xH ][ xhalfpel + 2*yhalfpel ]
d518aeb @michaelni halfpel MC doxy
michaelni authored
246 * @param block destination into which the result is averaged (a+b)>>1
247 * @param pixels source
248 * @param line_size number of bytes in a horizontal line of block
249 * @param h height
250 */
dbc56b3 @michaelni chroma me fix
michaelni authored
251 op_pixels_func avg_no_rnd_pixels_tab[4][4];
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
252
c0a0170 @michaelni fix motion compensation with (x+1/2,y+1/2) MVs
michaelni authored
253 void (*put_no_rnd_pixels_l2[2])(uint8_t *block/*align width (8 or 16)*/, const uint8_t *a/*align 1*/, const uint8_t *b/*align 1*/, int line_size, int h);
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
254
669ac79 @michaelni moving the svq3 motion compensation stuff to dsputil (this also means th...
michaelni authored
255 /**
256 * Thirdpel motion compensation with rounding (a+b+1)>>1.
30f1505 @lu-zero Typos in comments
lu-zero authored
257 * this is an array[12] of motion compensation functions for the 9 thirdpe
258 * positions<br>
669ac79 @michaelni moving the svq3 motion compensation stuff to dsputil (this also means th...
michaelni authored
259 * *pixels_tab[ xthirdpel + 4*ythirdpel ]
260 * @param block destination where the result is stored
261 * @param pixels source
262 * @param line_size number of bytes in a horizontal line of block
263 * @param h height
264 */
265 tpel_mc_func put_tpel_pixels_tab[11]; //FIXME individual func ptr per width?
da3b975 @multimediamike added support for B-frames and multiple slices
multimediamike authored
266 tpel_mc_func avg_tpel_pixels_tab[11]; //FIXME individual func ptr per width?
267
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
268 qpel_mc_func put_qpel_pixels_tab[2][16];
269 qpel_mc_func avg_qpel_pixels_tab[2][16];
270 qpel_mc_func put_no_rnd_qpel_pixels_tab[2][16];
271 qpel_mc_func avg_no_rnd_qpel_pixels_tab[2][16];
1457ab5 @michaelni qpel encoding
michaelni authored
272 qpel_mc_func put_mspel_pixels_tab[8];
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
273
0da7126 @michaelni H264 decoder & demuxer
michaelni authored
274 /**
30f1505 @lu-zero Typos in comments
lu-zero authored
275 * h264 Chroma MC
0da7126 @michaelni H264 decoder & demuxer
michaelni authored
276 */
277 h264_chroma_mc_func put_h264_chroma_pixels_tab[3];
e34350a New function for chroma MC (will be used in VC-1)
Kostya Shishkov authored
278 /* This is really one func used in VC-1 decoding */
279 h264_chroma_mc_func put_no_rnd_h264_chroma_pixels_tab[3];
0da7126 @michaelni H264 decoder & demuxer
michaelni authored
280 h264_chroma_mc_func avg_h264_chroma_pixels_tab[3];
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
281
80e44bc @michaelni use h264 MC functions for 2xX Xx2 blocks in snow too
michaelni authored
282 qpel_mc_func put_h264_qpel_pixels_tab[4][16];
283 qpel_mc_func avg_h264_qpel_pixels_tab[4][16];
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
284
2833fc4 approximate qpel functions: sacrifice some quality for some decoding spe...
Loren Merritt authored
285 qpel_mc_func put_2tap_qpel_pixels_tab[4][16];
286 qpel_mc_func avg_2tap_qpel_pixels_tab[4][16];
287
9f2d1b4 H.264 weighted prediction.
Loren Merritt authored
288 h264_weight_func weight_h264_pixels_tab[10];
289 h264_biweight_func biweight_h264_pixels_tab[10];
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
290
b482e2d @michaelni CAVS decoder by (Stefan Gehrer stefan.gehrer gmx.de)
michaelni authored
291 /* AVS specific */
292 qpel_mc_func put_cavs_qpel_pixels_tab[2][16];
293 qpel_mc_func avg_cavs_qpel_pixels_tab[2][16];
294 void (*cavs_filter_lv)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
295 void (*cavs_filter_lh)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
296 void (*cavs_filter_cv)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
297 void (*cavs_filter_ch)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
298 void (*cavs_idct8_add)(uint8_t *dst, DCTELEM *block, int stride);
299
bb198e1 @michaelni interlaced motion estimation
michaelni authored
300 me_cmp_func pix_abs[2][4];
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
301
11f18fa @michaelni huffyuv
michaelni authored
302 /* huffyuv specific */
303 void (*add_bytes)(uint8_t *dst/*align 16*/, uint8_t *src/*align 16*/, int w);
1457ab5 @michaelni qpel encoding
michaelni authored
304 void (*diff_bytes)(uint8_t *dst/*align 16*/, uint8_t *src1/*align 16*/, uint8_t *src2/*align 1*/,int w);
8470540 @michaelni mmx2 optimization of huffyuv median encoding
michaelni authored
305 /**
306 * subtract huffyuv's variant of median prediction
307 * note, this might read from src1[-1], src2[-1]
308 */
309 void (*sub_hfyu_median_prediction)(uint8_t *dst, uint8_t *src1, uint8_t *src2, int w, int *left, int *left_top);
3d2e8cc @michaelni ASV1 codec
michaelni authored
310 void (*bswap_buf)(uint32_t *dst, uint32_t *src, int w);
42251a2 MMX for H.264 deblocking filter
Loren Merritt authored
311
5cf08f2 H.264 deblocking optimizations (mmx for chroma_bS4 case, convert existin...
Loren Merritt authored
312 void (*h264_v_loop_filter_luma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0);
313 void (*h264_h_loop_filter_luma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0);
314 void (*h264_v_loop_filter_chroma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0);
315 void (*h264_h_loop_filter_chroma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0);
316 void (*h264_v_loop_filter_chroma_intra)(uint8_t *pix, int stride, int alpha, int beta);
317 void (*h264_h_loop_filter_chroma_intra)(uint8_t *pix, int stride, int alpha, int beta);
3e20143 mmx implementation of deblocking strength decision.
Loren Merritt authored
318 // h264_loop_filter_strength: simd only. the C version is inlined in h264.c
319 void (*h264_loop_filter_strength)(int16_t bS[2][4][4], uint8_t nnz[40], int8_t ref[2][40], int16_t mv[2][40][2],
320 int bidir, int edges, int step, int mask_mv0, int mask_mv1);
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
321
332f9ac @michaelni h263 loop filter
michaelni authored
322 void (*h263_v_loop_filter)(uint8_t *src, int stride, int qscale);
323 void (*h263_h_loop_filter)(uint8_t *src, int stride, int qscale);
324
fdbbf2e @michaelni rewrite h261 loop filter
michaelni authored
325 void (*h261_loop_filter)(uint8_t *src, int stride);
c6148de @michaelni h261 decoder by (Maarten Daniels <maarten.daniels at student dot luc dot...
michaelni authored
326
eb4825b sse and 3dnow implementations of float->int conversion and mdct windowin...
Loren Merritt authored
327 /* assume len is a multiple of 4, and arrays are 16-byte aligned */
2dac4ac sse & sse2 implementations of vorbis channel coupling.
Loren Merritt authored
328 void (*vorbis_inverse_coupling)(float *mag, float *ang, int blocksize);
eb4825b sse and 3dnow implementations of float->int conversion and mdct windowin...
Loren Merritt authored
329 /* assume len is a multiple of 8, and arrays are 16-byte aligned */
0697205 vorbis simd tweaks
Loren Merritt authored
330 void (*vector_fmul)(float *dst, const float *src, int len);
eb4825b sse and 3dnow implementations of float->int conversion and mdct windowin...
Loren Merritt authored
331 void (*vector_fmul_reverse)(float *dst, const float *src0, const float *src1, int len);
332 /* assume len is a multiple of 8, and src arrays are 16-byte aligned */
333 void (*vector_fmul_add_add)(float *dst, const float *src0, const float *src1, const float *src2, int src3, int len, int step);
334
335 /* C version: convert floats from the range [384.0,386.0] to ints in [-32768,32767]
9814587 @lu-zero Align the input buffer in ffplay, introduce a public macro for aligned d...
lu-zero authored
336 * simd versions: convert floats from [-32768.0,32767.0] without rescaling and arrays are 16byte aligned */
eb4825b sse and 3dnow implementations of float->int conversion and mdct windowin...
Loren Merritt authored
337 void (*float_to_int16)(int16_t *dst, const float *src, int len);
2dac4ac sse & sse2 implementations of vorbis channel coupling.
Loren Merritt authored
338
b036883 @michaelni MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
michaelni authored
339 /* (I)DCT */
340 void (*fdct)(DCTELEM *block/* align 16*/);
10acc47 @rvs * introducing dct248 into the DSP context.
rvs authored
341 void (*fdct248)(DCTELEM *block/* align 16*/);
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
342
4fb518c @michaelni the return of the idct with 16bit output by ("Ivan Kalvachev" <ivan at c...
michaelni authored
343 /* IDCT really*/
344 void (*idct)(DCTELEM *block/* align 16*/);
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
345
2464118 @michaelni more dox
michaelni authored
346 /**
77c92c2 @michaelni cosmetics
michaelni authored
347 * block -> idct -> clip to unsigned 8 bit -> dest.
2464118 @michaelni more dox
michaelni authored
348 * (-1392, 0, 0, ...) -> idct -> (-174, -174, ...) -> put -> (0, 0, ...)
a1adf43 @michaelni typo
michaelni authored
349 * @param line_size size in bytes of a horizotal line of dest
2464118 @michaelni more dox
michaelni authored
350 */
b036883 @michaelni MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
michaelni authored
351 void (*idct_put)(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*align 16*/);
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
352
2464118 @michaelni more dox
michaelni authored
353 /**
354 * block -> idct -> add dest -> clip to unsigned 8 bit -> dest.
a1adf43 @michaelni typo
michaelni authored
355 * @param line_size size in bytes of a horizotal line of dest
2464118 @michaelni more dox
michaelni authored
356 */
b036883 @michaelni MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
michaelni authored
357 void (*idct_add)(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*align 16*/);
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
358
2464118 @michaelni more dox
michaelni authored
359 /**
77c92c2 @michaelni cosmetics
michaelni authored
360 * idct input permutation.
0549302 @michaelni more idct_permute docs
michaelni authored
361 * several optimized IDCTs need a permutated input (relative to the normal order of the reference
362 * IDCT)
363 * this permutation must be performed before the idct_put/add, note, normally this can be merged
364 * with the zigzag/alternate scan<br>
2464118 @michaelni more dox
michaelni authored
365 * an example to avoid confusion:
366 * - (->decode coeffs -> zigzag reorder -> dequant -> reference idct ->...)
367 * - (x -> referece dct -> reference idct -> x)
368 * - (x -> referece dct -> simple_mmx_perm = idct_permutation -> simple_idct_mmx -> x)
369 * - (->decode coeffs -> zigzag reorder -> simple_mmx_perm -> dequant -> simple_idct_mmx ->...)
370 */
b036883 @michaelni MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
michaelni authored
371 uint8_t idct_permutation[64];
372 int idct_permutation_type;
373 #define FF_NO_IDCT_PERM 1
374 #define FF_LIBMPEG2_IDCT_PERM 2
375 #define FF_SIMPLE_IDCT_PERM 3
376 #define FF_TRANSPOSE_IDCT_PERM 4
5773a74 @michaelni porting the mmx&sse2 (sse2 untested) vp3 idcts to the lavc idct API
michaelni authored
377 #define FF_PARTTRANS_IDCT_PERM 5
b036883 @michaelni MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
michaelni authored
378
364a179 @michaelni quantizer noise shaping optimization
michaelni authored
379 int (*try_8x8basis)(int16_t rem[64], int16_t weight[64], int16_t basis[64], int scale);
380 void (*add_8x8basis)(int16_t rem[64], int16_t basis[64], int scale);
381 #define BASIS_SHIFT 16
382 #define RECON_SHIFT 6
115329f @DonDiego COSMETICS: Remove all trailing whitespace.
DonDiego authored
383
ea9f5d6 Add comment denoting group of H.264 functions
Panagiotis Issaris authored
384 /* h264 functions */
0fa8158 @michaelni move h264 idct to its own file and call via function pointer in DspConte...
michaelni authored
385 void (*h264_idct_add)(uint8_t *dst, DCTELEM *block, int stride);
43efd19 decode H.264 with 8x8 transform.
Loren Merritt authored
386 void (*h264_idct8_add)(uint8_t *dst, DCTELEM *block, int stride);
ef9d1d1 h264: special case dc-only idct. ~1% faster overall
Loren Merritt authored
387 void (*h264_idct_dc_add)(uint8_t *dst, DCTELEM *block, int stride);
388 void (*h264_idct8_dc_add)(uint8_t *dst, DCTELEM *block, int stride);
f186212 Added entry for H.264 DCT to the DSP context.
Panagiotis Issaris authored
389 void (*h264_dct)(DCTELEM block[4][4]);
059715a First part of a series of speed-enchancing patches.
Robert Edele authored
390
391 /* snow wavelet */
392 void (*vertical_compose97i)(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2, DWTELEM *b3, DWTELEM *b4, DWTELEM *b5, int width);
393 void (*horizontal_compose97i)(DWTELEM *b, int width);
9dd6c80 Add the const specifier as needed to reduce the number of warnings.
Panagiotis Issaris authored
394 void (*inner_add_yblock)(const uint8_t *obmc, const int obmc_stride, uint8_t * * block, int b_w, int b_h, int src_x, int src_y, int src_stride, slice_buffer * sb, int add, uint8_t * dst8);
513fbd8 prefetch pixels for future motion compensation. 2-5% faster h264.
Loren Merritt authored
395
396 void (*prefetch)(void *mem, int stride, int h);
54009d4 @michaelni remove mpegvideo.c img resample dependancy
michaelni authored
397
398 void (*shrink[4])(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
64db55a New functions in DSPContext for VC-1 decoding
Kostya Shishkov authored
399
400 /* vc1 functions */
401 void (*vc1_inv_trans_8x8)(DCTELEM *b);
402 void (*vc1_inv_trans_8x4)(DCTELEM *b, int n);
403 void (*vc1_inv_trans_4x8)(DCTELEM *b, int n);
404 void (*vc1_inv_trans_4x4)(DCTELEM *b, int n);
61f5b14 Correct rounding values in overlap filtering
Kostya Shishkov authored
405 void (*vc1_v_overlap)(uint8_t* src, int stride);
406 void (*vc1_h_overlap)(uint8_t* src, int stride);
64db55a New functions in DSPContext for VC-1 decoding
Kostya Shishkov authored
407 /* put 8x8 block with bicubic interpolation and quarterpel precision
408 * last argument is actually round value instead of height
409 */
410 op_pixels_func put_vc1_mspel_pixels_tab[16];
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
411 } DSPContext;
412
486497e @mansr revert bad checkin
mansr authored
413 void dsputil_static_init(void);
b036883 @michaelni MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
michaelni authored
414 void dsputil_init(DSPContext* p, AVCodecContext *avctx);
de6d9b6 Initial revision
Fabrice Bellard authored
415
6dc7d5d @michaelni ff_check_alignment to warn the user about a missaligned stack
michaelni authored
416 int ff_check_alignment(void);
417
7801d21 @michaelni optimize block_permute()
michaelni authored
418 /**
419 * permute block according to permuatation.
420 * @param last last non zero element in scantable order
421 */
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
422 void ff_block_permute(DCTELEM *block, uint8_t *permutation, const uint8_t *scantable, int last);
e0eac44 added block permutation functions
Fabrice Bellard authored
423
622348f @michaelni interlaced dct decision cleanup
michaelni authored
424 void ff_set_cmp(DSPContext* c, me_cmp_func *cmp, int type);
425
bb270c0 @DonDiego COSMETICS: tabs --> spaces, some prettyprinting
DonDiego authored
426 #define BYTE_VEC32(c) ((c)*0x01010101UL)
d8085ea @michaelni cleanup
michaelni authored
427
428 static inline uint32_t rnd_avg32(uint32_t a, uint32_t b)
429 {
430 return (a | b) - (((a ^ b) & ~BYTE_VEC32(0x01)) >> 1);
431 }
432
433 static inline uint32_t no_rnd_avg32(uint32_t a, uint32_t b)
434 {
435 return (a & b) + (((a ^ b) & ~BYTE_VEC32(0x01)) >> 1);
436 }
437
26efc54 @michaelni 4x4 SSE compare function
michaelni authored
438 static inline int get_penalty_factor(int lambda, int lambda2, int type){
439 switch(type&0xFF){
440 default:
441 case FF_CMP_SAD:
442 return lambda>>FF_LAMBDA_SHIFT;
443 case FF_CMP_DCT:
444 return (3*lambda)>>(FF_LAMBDA_SHIFT+1);
445 case FF_CMP_W53:
446 return (4*lambda)>>(FF_LAMBDA_SHIFT);
447 case FF_CMP_W97:
448 return (2*lambda)>>(FF_LAMBDA_SHIFT);
449 case FF_CMP_SATD:
27c61ac @michaelni 8x8 integer dct from x264 as cmp function (under CONFIG_GPL)
michaelni authored
450 case FF_CMP_DCT264:
26efc54 @michaelni 4x4 SSE compare function
michaelni authored
451 return (2*lambda)>>FF_LAMBDA_SHIFT;
452 case FF_CMP_RD:
453 case FF_CMP_PSNR:
454 case FF_CMP_SSE:
455 case FF_CMP_NSSE:
456 return lambda2>>FF_LAMBDA_SHIFT;
457 case FF_CMP_BIT:
458 return 1;
459 }
460 }
461
2464118 @michaelni more dox
michaelni authored
462 /**
77c92c2 @michaelni cosmetics
michaelni authored
463 * Empty mmx state.
2464118 @michaelni more dox
michaelni authored
464 * this must be called between any dsp function and float/double code.
465 * for example sin(); dsp->idct_put(); emms_c(); cos()
466 */
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
467 #define emms_c()
468
e629ab6 altivec jumbo patch by (Romain Dolbeau <dolbeaur at club-internet dot fr...
Romain Dolbeau authored
469 /* should be defined by architectures supporting
470 one or more MultiMedia extension */
471 int mm_support(void);
472
a1bee08 @mansr move dsputil_init_foo() out from #ifdef
mansr authored
473 void dsputil_init_alpha(DSPContext* c, AVCodecContext *avctx);
474 void dsputil_init_armv4l(DSPContext* c, AVCodecContext *avctx);
475 void dsputil_init_bfin(DSPContext* c, AVCodecContext *avctx);
476 void dsputil_init_mlib(DSPContext* c, AVCodecContext *avctx);
477 void dsputil_init_mmi(DSPContext* c, AVCodecContext *avctx);
478 void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx);
479 void dsputil_init_ppc(DSPContext* c, AVCodecContext *avctx);
480 void dsputil_init_sh4(DSPContext* c, AVCodecContext *avctx);
481 void dsputil_init_vis(DSPContext* c, AVCodecContext *avctx);
482
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
483 #define DECLARE_ALIGNED_16(t, v) DECLARE_ALIGNED(16, t, v)
92a69cf @multimediamike move __align16 some place where non-MMX machines can see it
multimediamike authored
484
3d03c0a arm optimizations
Fabrice Bellard authored
485 #if defined(HAVE_MMX)
de6d9b6 Initial revision
Fabrice Bellard authored
486
18f7701 * fixing some minor const warnings
Zdenek Kabelac authored
487 #undef emms_c
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
488
de6d9b6 Initial revision
Fabrice Bellard authored
489 #define MM_MMX 0x0001 /* standard MMX */
490 #define MM_3DNOW 0x0004 /* AMD 3DNOW */
491 #define MM_MMXEXT 0x0002 /* SSE integer functions or AMD MMX ext */
492 #define MM_SSE 0x0008 /* SSE functions */
493 #define MM_SSE2 0x0010 /* PIV SSE2 functions */
e42a152 @michaelni x86 cpu capabilities detection rewrite / cleanup
michaelni authored
494 #define MM_3DNOWEXT 0x0020 /* AMD 3DNowExt */
392f6da @zuxy Remove unused and unsupported Cyrix's "Extended MMX",
zuxy authored
495 #define MM_SSE3 0x0040 /* Prescott SSE3 functions */
5a5c770 Add SSSE3 (Core2 aka Conroe/Merom/Woodcrester new instructions) detectio...
Guillaume Poirier authored
496 #define MM_SSSE3 0x0080 /* Conroe SSSE3 functions */
de6d9b6 Initial revision
Fabrice Bellard authored
497
486497e @mansr revert bad checkin
mansr authored
498 extern int mm_flags;
499
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
500 void add_pixels_clamped_mmx(const DCTELEM *block, uint8_t *pixels, int line_size);
501 void put_pixels_clamped_mmx(const DCTELEM *block, uint8_t *pixels, int line_size);
f9ed9d8 @multimediamike separate out put_signed_pixels_clamped() into its own function and
multimediamike authored
502 void put_signed_pixels_clamped_mmx(const DCTELEM *block, uint8_t *pixels, int line_size);
de6d9b6 Initial revision
Fabrice Bellard authored
503
504 static inline void emms(void)
505 {
fb16b7e added emms_c() macro which should can used in c code in both mmx/non mmx...
Fabrice Bellard authored
506 __asm __volatile ("emms;":::"memory");
507 }
508
1457ab5 @michaelni qpel encoding
michaelni authored
509
fb16b7e added emms_c() macro which should can used in c code in both mmx/non mmx...
Fabrice Bellard authored
510 #define emms_c() \
511 {\
486497e @mansr revert bad checkin
mansr authored
512 if (mm_flags & MM_MMX)\
fb16b7e added emms_c() macro which should can used in c code in both mmx/non mmx...
Fabrice Bellard authored
513 emms();\
de6d9b6 Initial revision
Fabrice Bellard authored
514 }
515
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
516 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(8, t, v)
68b51e5 MSVC-compatible __align8/__align16 declaration
Steve L'Homme authored
517
3237f73 @michaelni stride align cleanup
michaelni authored
518 #define STRIDE_ALIGN 8
de6d9b6 Initial revision
Fabrice Bellard authored
519
b036883 @michaelni MpegEncContext.(i)dct_* -> DspContext.(i)dct_*
michaelni authored
520 void dsputil_init_pix_mmx(DSPContext* c, AVCodecContext *avctx);
2720569 * prototypes at one place
Zdenek Kabelac authored
521
3d03c0a arm optimizations
Fabrice Bellard authored
522 #elif defined(ARCH_ARMV4L)
523
524 /* This is to use 4 bytes read to the IDCT pointers for some 'zero'
92a69cf @multimediamike move __align16 some place where non-MMX machines can see it
multimediamike authored
525 line optimizations */
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
526 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(4, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
527 #define STRIDE_ALIGN 4
3d03c0a arm optimizations
Fabrice Bellard authored
528
eba9ae3 IWMMXT configure support + runtime selection patch by (Gildas Bazin, gba...
Gildas Bazin authored
529 #define MM_IWMMXT 0x0100 /* XScale IWMMXT */
530
486497e @mansr revert bad checkin
mansr authored
531 extern int mm_flags;
532
c34270f mlib merge
Fabrice Bellard authored
533 #elif defined(HAVE_MLIB)
534
535 /* SPARC/VIS IDCT needs 8-byte aligned DCT blocks */
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
536 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(8, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
537 #define STRIDE_ALIGN 8
c34270f mlib merge
Fabrice Bellard authored
538
44f54ce @michaelni VIS optimized motion compensation code. by (David S. Miller <davem at re...
michaelni authored
539 #elif defined(ARCH_SPARC)
540
541 /* SPARC/VIS IDCT needs 8-byte aligned DCT blocks */
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
542 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(8, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
543 #define STRIDE_ALIGN 8
44f54ce @michaelni VIS optimized motion compensation code. by (David S. Miller <davem at re...
michaelni authored
544
1e98dff Alpha optimizations by Falk Hueffner <falk.hueffner@student.uni-tuebinge...
Nick Kurshev authored
545 #elif defined(ARCH_ALPHA)
546
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
547 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(8, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
548 #define STRIDE_ALIGN 8
1e98dff Alpha optimizations by Falk Hueffner <falk.hueffner@student.uni-tuebinge...
Nick Kurshev authored
549
59925ef first cut at altivec support on darwin patch by (Brian Foley <bfoley at ...
Brian Foley authored
550 #elif defined(ARCH_POWERPC)
551
404d224 altivec accelerated v-resample patch by (Brian Foley <bfoley at compsoc ...
Brian Foley authored
552 #define MM_ALTIVEC 0x0001 /* standard AltiVec */
553
486497e @mansr revert bad checkin
mansr authored
554 extern int mm_flags;
555
3b991c5 Altivec on non darwin systems patch by Romain Dolbeau
Romain Dolbeau authored
556 #if defined(HAVE_ALTIVEC) && !defined(CONFIG_DARWIN)
9a197a2 @michaelni altivec.h bug workaround (untested)
michaelni authored
557 #define pixel altivec_pixel
3b991c5 Altivec on non darwin systems patch by Romain Dolbeau
Romain Dolbeau authored
558 #include <altivec.h>
9a197a2 @michaelni altivec.h bug workaround (untested)
michaelni authored
559 #undef pixel
3b991c5 Altivec on non darwin systems patch by Romain Dolbeau
Romain Dolbeau authored
560 #endif
561
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
562 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(16, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
563 #define STRIDE_ALIGN 16
59925ef first cut at altivec support on darwin patch by (Brian Foley <bfoley at ...
Brian Foley authored
564
d46aba2 ps2 idct patch by (Leon van Stuivenberg <leonvs at iae dot nl>)
Leon van Stuivenberg authored
565 #elif defined(HAVE_MMI)
566
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
567 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(16, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
568 #define STRIDE_ALIGN 16
d46aba2 ps2 idct patch by (Leon van Stuivenberg <leonvs at iae dot nl>)
Leon van Stuivenberg authored
569
0c6bd2e sh4 optimized idct & bswap patch by (BERO <bero at geocities dot co dot ...
BERO authored
570 #elif defined(ARCH_SH4)
571
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
572 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(8, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
573 #define STRIDE_ALIGN 8
0c6bd2e sh4 optimized idct & bswap patch by (BERO <bero at geocities dot co dot ...
BERO authored
574
5ca030f ADI Blackfin optimizations
Michael Benjamin authored
575 #elif defined(ARCH_BFIN)
576
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
577 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(8, t, v)
5ca030f ADI Blackfin optimizations
Michael Benjamin authored
578 #define STRIDE_ALIGN 8
579
de6d9b6 Initial revision
Fabrice Bellard authored
580 #else
581
43c7c7c @axic use DECLARE_ALIGNED from mem.h (which is included through common.h)
axic authored
582 #define DECLARE_ALIGNED_8(t, v) DECLARE_ALIGNED(8, t, v)
3237f73 @michaelni stride align cleanup
michaelni authored
583 #define STRIDE_ALIGN 8
de6d9b6 Initial revision
Fabrice Bellard authored
584
585 #endif
586
43f1708 - Added PSNR feature to libavcodec and ffmpeg. By now just Y PSNR until ...
Juanjo authored
587 /* PSNR */
0c1a9ed * UINTX -> uintx_t INTX -> intx_t
Zdenek Kabelac authored
588 void get_psnr(uint8_t *orig_image[3], uint8_t *coded_image[3],
43f1708 - Added PSNR feature to libavcodec and ffmpeg. By now just Y PSNR until ...
Juanjo authored
589 int orig_linesize[3], int coded_linesize,
590 AVCodecContext *avctx);
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
591
592 /* FFT computation */
593
594 /* NOTE: soon integer code will be added, so you must use the
595 FFTSample type */
596 typedef float FFTSample;
597
bcfa3e5 3dnow2 implementation of imdct.
Loren Merritt authored
598 struct MDCTContext;
599
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
600 typedef struct FFTComplex {
601 FFTSample re, im;
602 } FFTComplex;
603
604 typedef struct FFTContext {
605 int nbits;
606 int inverse;
607 uint16_t *revtab;
608 FFTComplex *exptab;
609 FFTComplex *exptab1; /* only used by SSE code */
610 void (*fft_calc)(struct FFTContext *s, FFTComplex *z);
bcfa3e5 3dnow2 implementation of imdct.
Loren Merritt authored
611 void (*imdct_calc)(struct MDCTContext *s, FFTSample *output,
612 const FFTSample *input, FFTSample *tmp);
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
613 } FFTContext;
614
68951ec fft_*() renamed into ff_fft_*() patch by (Gildas Bazin <gbazin at altern...
Gildas Bazin authored
615 int ff_fft_init(FFTContext *s, int nbits, int inverse);
616 void ff_fft_permute(FFTContext *s, FFTComplex *z);
617 void ff_fft_calc_c(FFTContext *s, FFTComplex *z);
618 void ff_fft_calc_sse(FFTContext *s, FFTComplex *z);
82eb4b0 @zuxy 3DNow! & Extended 3DNow! versions of FFT
zuxy authored
619 void ff_fft_calc_3dn(FFTContext *s, FFTComplex *z);
620 void ff_fft_calc_3dn2(FFTContext *s, FFTComplex *z);
68951ec fft_*() renamed into ff_fft_*() patch by (Gildas Bazin <gbazin at altern...
Gildas Bazin authored
621 void ff_fft_calc_altivec(FFTContext *s, FFTComplex *z);
8d268a7 fft altivec by Romain Dolbeau - simplified patch, test it on PPC with ff...
Fabrice Bellard authored
622
68951ec fft_*() renamed into ff_fft_*() patch by (Gildas Bazin <gbazin at altern...
Gildas Bazin authored
623 static inline void ff_fft_calc(FFTContext *s, FFTComplex *z)
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
624 {
625 s->fft_calc(s, z);
626 }
68951ec fft_*() renamed into ff_fft_*() patch by (Gildas Bazin <gbazin at altern...
Gildas Bazin authored
627 void ff_fft_end(FFTContext *s);
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
628
629 /* MDCT computation */
630
631 typedef struct MDCTContext {
632 int n; /* size of MDCT (i.e. number of input data * 2) */
633 int nbits; /* n = 2^nbits */
634 /* pre/post rotation tables */
635 FFTSample *tcos;
636 FFTSample *tsin;
637 FFTContext fft;
638 } MDCTContext;
639
82696be avoid name clash
Fabrice Bellard authored
640 int ff_mdct_init(MDCTContext *s, int nbits, int inverse);
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
641 void ff_imdct_calc(MDCTContext *s, FFTSample *output,
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
642 const FFTSample *input, FFTSample *tmp);
bcfa3e5 3dnow2 implementation of imdct.
Loren Merritt authored
643 void ff_imdct_calc_3dn2(MDCTContext *s, FFTSample *output,
644 const FFTSample *input, FFTSample *tmp);
ebbafcb sse implementation of imdct.
Loren Merritt authored
645 void ff_imdct_calc_sse(MDCTContext *s, FFTSample *output,
646 const FFTSample *input, FFTSample *tmp);
eb4b3dd * using DSPContext - so each codec could use its local (sub)set of CPU e...
Zdenek Kabelac authored
647 void ff_mdct_calc(MDCTContext *s, FFTSample *out,
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
648 const FFTSample *input, FFTSample *tmp);
82696be avoid name clash
Fabrice Bellard authored
649 void ff_mdct_end(MDCTContext *s);
bb6f569 new generic FFT/MDCT code for audio codecs
Fabrice Bellard authored
650
bb198e1 @michaelni interlaced motion estimation
michaelni authored
651 #define WARPER8_16(name8, name16)\
652 static int name16(void /*MpegEncContext*/ *s, uint8_t *dst, uint8_t *src, int stride, int h){\
653 return name8(s, dst , src , stride, h)\
654 +name8(s, dst+8 , src+8 , stride, h);\
655 }
656
657 #define WARPER8_16_SQ(name8, name16)\
658 static int name16(void /*MpegEncContext*/ *s, uint8_t *dst, uint8_t *src, int stride, int h){\
659 int score=0;\
660 score +=name8(s, dst , src , stride, 8);\
661 score +=name8(s, dst+8 , src+8 , stride, 8);\
662 if(h==16){\
663 dst += 8*stride;\
664 src += 8*stride;\
665 score +=name8(s, dst , src , stride, 8);\
666 score +=name8(s, dst+8 , src+8 , stride, 8);\
667 }\
668 return score;\
1457ab5 @michaelni qpel encoding
michaelni authored
669 }
670
49cef74 Move copy_block* functions to dsp dsputil.h
Brian Foley authored
671
672 static inline void copy_block2(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h)
673 {
674 int i;
675 for(i=0; i<h; i++)
676 {
677 ST16(dst , LD16(src ));
678 dst+=dstStride;
679 src+=srcStride;
680 }
681 }
682
683 static inline void copy_block4(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h)
684 {
685 int i;
686 for(i=0; i<h; i++)
687 {
688 ST32(dst , LD32(src ));
689 dst+=dstStride;
690 src+=srcStride;
691 }
692 }
693
694 static inline void copy_block8(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h)
695 {
696 int i;
697 for(i=0; i<h; i++)
698 {
699 ST32(dst , LD32(src ));
700 ST32(dst+4 , LD32(src+4 ));
701 dst+=dstStride;
702 src+=srcStride;
703 }
704 }
705
706 static inline void copy_block9(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h)
707 {
708 int i;
709 for(i=0; i<h; i++)
710 {
711 ST32(dst , LD32(src ));
712 ST32(dst+4 , LD32(src+4 ));
713 dst[8]= src[8];
714 dst+=dstStride;
715 src+=srcStride;
716 }
717 }
718
719 static inline void copy_block16(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h)
720 {
721 int i;
722 for(i=0; i<h; i++)
723 {
724 ST32(dst , LD32(src ));
725 ST32(dst+4 , LD32(src+4 ));
726 ST32(dst+8 , LD32(src+8 ));
727 ST32(dst+12, LD32(src+12));
728 dst+=dstStride;
729 src+=srcStride;
730 }
731 }
732
733 static inline void copy_block17(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h)
734 {
735 int i;
736 for(i=0; i<h; i++)
737 {
738 ST32(dst , LD32(src ));
739 ST32(dst+4 , LD32(src+4 ));
740 ST32(dst+8 , LD32(src+8 ));
741 ST32(dst+12, LD32(src+12));
742 dst[16]= src[16];
743 dst+=dstStride;
744 src+=srcStride;
745 }
746 }
747
de6d9b6 Initial revision
Fabrice Bellard authored
748 #endif
Something went wrong with that request. Please try again.