Statistics
| Branch: | Revision:

ffmpeg / libavcodec / i386 / dsputil_mmx.h @ 97d1d009

History | View | Annotate | Download (3.77 KB)

1
/*
2
 * MMX optimized DSP utils
3
 * Copyright (c) 2007  Aurelien Jacobs <aurel@gnuage.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
#ifndef FFMPEG_DSPUTIL_MMX_H
23
#define FFMPEG_DSPUTIL_MMX_H
24

    
25
#include <stdint.h>
26
#include "dsputil.h"
27

    
28
typedef struct { uint64_t a, b; } xmm_t;
29

    
30
extern const uint64_t ff_bone;
31
extern const uint64_t ff_wtwo;
32

    
33
extern const uint64_t ff_pdw_80000000[2];
34

    
35
extern const uint64_t ff_pw_3;
36
extern const uint64_t ff_pw_4;
37
extern const xmm_t    ff_pw_5;
38
extern const uint64_t ff_pw_8;
39
extern const uint64_t ff_pw_15;
40
extern const xmm_t    ff_pw_16;
41
extern const uint64_t ff_pw_20;
42
extern const xmm_t    ff_pw_32;
43
extern const uint64_t ff_pw_42;
44
extern const uint64_t ff_pw_64;
45
extern const uint64_t ff_pw_96;
46
extern const uint64_t ff_pw_128;
47
extern const uint64_t ff_pw_255;
48

    
49
extern const uint64_t ff_pb_1;
50
extern const uint64_t ff_pb_3;
51
extern const uint64_t ff_pb_7;
52
extern const uint64_t ff_pb_3F;
53
extern const uint64_t ff_pb_A1;
54
extern const uint64_t ff_pb_FC;
55

    
56
extern const double ff_pd_1[2];
57
extern const double ff_pd_2[2];
58

    
59
/* in/out: mma=mma+mmb, mmb=mmb-mma */
60
#define SUMSUB_BA( a, b ) \
61
    "paddw "#b", "#a" \n\t"\
62
    "paddw "#b", "#b" \n\t"\
63
    "psubw "#a", "#b" \n\t"
64

    
65
#define SBUTTERFLY(a,b,t,n,m)\
66
    "mov" #m " " #a ", " #t "         \n\t" /* abcd */\
67
    "punpckl" #n " " #b ", " #a "     \n\t" /* aebf */\
68
    "punpckh" #n " " #b ", " #t "     \n\t" /* cgdh */\
69

    
70
#define TRANSPOSE4(a,b,c,d,t)\
71
    SBUTTERFLY(a,b,t,wd,q) /* a=aebf t=cgdh */\
72
    SBUTTERFLY(c,d,b,wd,q) /* c=imjn b=kolp */\
73
    SBUTTERFLY(a,c,d,dq,q) /* a=aeim d=bfjn */\
74
    SBUTTERFLY(t,b,c,dq,q) /* t=cgko c=dhlp */
75

    
76
#ifdef ARCH_X86_64
77
// permutes 01234567 -> 05736421
78
#define TRANSPOSE8(a,b,c,d,e,f,g,h,t)\
79
    SBUTTERFLY(a,b,%%xmm8,wd,dqa)\
80
    SBUTTERFLY(c,d,b,wd,dqa)\
81
    SBUTTERFLY(e,f,d,wd,dqa)\
82
    SBUTTERFLY(g,h,f,wd,dqa)\
83
    SBUTTERFLY(a,c,h,dq,dqa)\
84
    SBUTTERFLY(%%xmm8,b,c,dq,dqa)\
85
    SBUTTERFLY(e,g,b,dq,dqa)\
86
    SBUTTERFLY(d,f,g,dq,dqa)\
87
    SBUTTERFLY(a,e,f,qdq,dqa)\
88
    SBUTTERFLY(%%xmm8,d,e,qdq,dqa)\
89
    SBUTTERFLY(h,b,d,qdq,dqa)\
90
    SBUTTERFLY(c,g,b,qdq,dqa)\
91
    "movdqa %%xmm8, "#g"              \n\t"
92
#else
93
#define TRANSPOSE8(a,b,c,d,e,f,g,h,t)\
94
    "movdqa "#h", "#t"                \n\t"\
95
    SBUTTERFLY(a,b,h,wd,dqa)\
96
    "movdqa "#h", 16"#t"              \n\t"\
97
    "movdqa "#t", "#h"                \n\t"\
98
    SBUTTERFLY(c,d,b,wd,dqa)\
99
    SBUTTERFLY(e,f,d,wd,dqa)\
100
    SBUTTERFLY(g,h,f,wd,dqa)\
101
    SBUTTERFLY(a,c,h,dq,dqa)\
102
    "movdqa "#h", "#t"                \n\t"\
103
    "movdqa 16"#t", "#h"              \n\t"\
104
    SBUTTERFLY(h,b,c,dq,dqa)\
105
    SBUTTERFLY(e,g,b,dq,dqa)\
106
    SBUTTERFLY(d,f,g,dq,dqa)\
107
    SBUTTERFLY(a,e,f,qdq,dqa)\
108
    SBUTTERFLY(h,d,e,qdq,dqa)\
109
    "movdqa "#h", 16"#t"              \n\t"\
110
    "movdqa "#t", "#h"                \n\t"\
111
    SBUTTERFLY(h,b,d,qdq,dqa)\
112
    SBUTTERFLY(c,g,b,qdq,dqa)\
113
    "movdqa 16"#t", "#g"              \n\t"
114
#endif
115

    
116
#define MOVQ_WONE(regd) \
117
    asm volatile ( \
118
    "pcmpeqd %%" #regd ", %%" #regd " \n\t" \
119
    "psrlw $15, %%" #regd ::)
120

    
121
void dsputilenc_init_mmx(DSPContext* c, AVCodecContext *avctx);
122

    
123
#endif /* FFMPEG_DSPUTIL_MMX_H */