Mercurial > libavcodec.hg
annotate eaidct.c @ 12197:fbf4d5b1b664 libavcodec
Remove FF_MM_SSE2/3 flags for CPUs where this is generally not faster than
regular MMX code. Examples of this are the Core1 CPU. Instead, set a new flag,
FF_MM_SSE2/3SLOW, which can be checked for particular SSE2/3 functions that
have been checked specifically on such CPUs and are actually faster than
their MMX counterparts.
In addition, use this flag to enable particular VP8 and LPC SSE2 functions
that are faster than their MMX counterparts.
Based on a patch by Loren Merritt <lorenm AT u washington edu>.
author | rbultje |
---|---|
date | Mon, 19 Jul 2010 22:38:23 +0000 |
parents | 7dd2a45249a9 |
children |
rev | line source |
---|---|
8120 | 1 /* |
2 * Electronic Arts TGQ/TQI/MAD IDCT algorithm | |
3 * Copyright (c) 2007-2008 Peter Ross <pross@xvid.org> | |
4 * | |
5 * This file is part of FFmpeg. | |
6 * | |
7 * FFmpeg is free software; you can redistribute it and/or | |
8 * modify it under the terms of the GNU Lesser General Public | |
9 * License as published by the Free Software Foundation; either | |
10 * version 2.1 of the License, or (at your option) any later version. | |
11 * | |
12 * FFmpeg is distributed in the hope that it will be useful, | |
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
15 * Lesser General Public License for more details. | |
16 * | |
17 * You should have received a copy of the GNU Lesser General Public | |
18 * License along with FFmpeg; if not, write to the Free Software | |
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA | |
20 */ | |
21 | |
22 /** | |
11644
7dd2a45249a9
Remove explicit filename from Doxygen @file commands.
diego
parents:
8718
diff
changeset
|
23 * @file |
8120 | 24 * Electronic Arts TGQ/TQI/MAD IDCT algorithm |
25 * @author Peter Ross <pross@xvid.org> | |
26 */ | |
27 | |
28 #include "dsputil.h" | |
29 | |
30 #define ASQRT 181 /* (1/sqrt(2))<<8 */ | |
31 #define A4 669 /* cos(pi/8)*sqrt(2)<<9 */ | |
32 #define A2 277 /* sin(pi/8)*sqrt(2)<<9 */ | |
33 #define A5 196 /* sin(pi/8)<<9 */ | |
34 | |
35 #define IDCT_TRANSFORM(dest,s0,s1,s2,s3,s4,s5,s6,s7,d0,d1,d2,d3,d4,d5,d6,d7,munge,src) {\ | |
36 const int a1 = (src)[s1] + (src)[s7]; \ | |
37 const int a7 = (src)[s1] - (src)[s7]; \ | |
38 const int a5 = (src)[s5] + (src)[s3]; \ | |
39 const int a3 = (src)[s5] - (src)[s3]; \ | |
40 const int a2 = (src)[s2] + (src)[s6]; \ | |
41 const int a6 = (ASQRT*((src)[s2] - (src)[s6]))>>8; \ | |
42 const int a0 = (src)[s0] + (src)[s4]; \ | |
43 const int a4 = (src)[s0] - (src)[s4]; \ | |
44 const int b0 = (((A4-A5)*a7 - A5*a3)>>9) + a1+a5; \ | |
45 const int b1 = (((A4-A5)*a7 - A5*a3)>>9) + ((ASQRT*(a1-a5))>>8); \ | |
46 const int b2 = (((A2+A5)*a3 + A5*a7)>>9) + ((ASQRT*(a1-a5))>>8); \ | |
47 const int b3 = ((A2+A5)*a3 + A5*a7)>>9; \ | |
48 (dest)[d0] = munge(a0+a2+a6+b0); \ | |
49 (dest)[d1] = munge(a4+a6 +b1); \ | |
50 (dest)[d2] = munge(a4-a6 +b2); \ | |
51 (dest)[d3] = munge(a0-a2-a6+b3); \ | |
52 (dest)[d4] = munge(a0-a2-a6-b3); \ | |
53 (dest)[d5] = munge(a4-a6 -b2); \ | |
54 (dest)[d6] = munge(a4+a6 -b1); \ | |
55 (dest)[d7] = munge(a0+a2+a6-b0); \ | |
56 } | |
57 /* end IDCT_TRANSFORM macro */ | |
58 | |
59 #define MUNGE_NONE(x) (x) | |
60 #define IDCT_COL(dest,src) IDCT_TRANSFORM(dest,0,8,16,24,32,40,48,56,0,8,16,24,32,40,48,56,MUNGE_NONE,src) | |
61 | |
62 #define MUNGE_8BIT(x) av_clip_uint8((x)>>4) | |
63 #define IDCT_ROW(dest,src) IDCT_TRANSFORM(dest,0,1,2,3,4,5,6,7,0,1,2,3,4,5,6,7,MUNGE_8BIT,src) | |
64 | |
65 static inline void ea_idct_col(DCTELEM *dest, const DCTELEM *src) { | |
66 if ((src[8]|src[16]|src[24]|src[32]|src[40]|src[48]|src[56])==0) { | |
67 dest[0] = | |
68 dest[8] = | |
69 dest[16] = | |
70 dest[24] = | |
71 dest[32] = | |
72 dest[40] = | |
73 dest[48] = | |
74 dest[56] = src[0]; | |
75 }else | |
76 IDCT_COL(dest, src); | |
77 } | |
78 | |
79 void ff_ea_idct_put_c(uint8_t *dest, int linesize, DCTELEM *block) { | |
80 int i; | |
81 DCTELEM temp[64]; | |
82 block[0] += 4; | |
83 for (i=0; i<8; i++) | |
84 ea_idct_col(&temp[i], &block[i]); | |
85 for (i=0; i<8; i++) | |
86 IDCT_ROW( (&dest[i*linesize]), (&temp[8*i]) ); | |
87 } |