aboutsummaryrefslogtreecommitdiffstats
path: root/libavcodec/x86/fmtconvert_mmx.c
blob: ca0b29344ac87b84353dc48e48ad1f37022da1eb (plain) (blame)
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
/*
 * Format Conversion Utils
 * Copyright (c) 2000, 2001 Fabrice Bellard
 * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 *
 * MMX optimization by Nick Kurshev <nickols_k@mail.ru>
 */

#include "libavutil/cpu.h"
#include "libavutil/x86_cpu.h"
#include "libavcodec/fmtconvert.h"

#if HAVE_YASM

void ff_int32_to_float_fmul_scalar_sse (float *dst, const int *src, float mul, int len);
void ff_int32_to_float_fmul_scalar_sse2(float *dst, const int *src, float mul, int len);

void ff_float_to_int16_3dnow(int16_t *dst, const float *src, long len);
void ff_float_to_int16_sse  (int16_t *dst, const float *src, long len);
void ff_float_to_int16_sse2 (int16_t *dst, const float *src, long len);

void ff_float_to_int16_interleave2_3dnow(int16_t *dst, const float **src, long len);
void ff_float_to_int16_interleave2_sse  (int16_t *dst, const float **src, long len);
void ff_float_to_int16_interleave2_sse2 (int16_t *dst, const float **src, long len);

void ff_float_to_int16_interleave6_sse(int16_t *dst, const float **src, int len);
void ff_float_to_int16_interleave6_3dnow(int16_t *dst, const float **src, int len);
void ff_float_to_int16_interleave6_3dn2(int16_t *dst, const float **src, int len);

#define ff_float_to_int16_interleave6_sse2 ff_float_to_int16_interleave6_sse

#define FLOAT_TO_INT16_INTERLEAVE(cpu) \
/* gcc pessimizes register allocation if this is in the same function as float_to_int16_interleave_sse2*/\
static av_noinline void float_to_int16_interleave_misc_##cpu(int16_t *dst, const float **src, long len, int channels){\
    DECLARE_ALIGNED(16, int16_t, tmp)[len];\
    int i,j,c;\
    for(c=0; c<channels; c++){\
        ff_float_to_int16_##cpu(tmp, src[c], len);\
        for(i=0, j=c; i<len; i++, j+=channels)\
            dst[j] = tmp[i];\
    }\
}\
\
static void float_to_int16_interleave_##cpu(int16_t *dst, const float **src, long len, int channels){\
    if(channels==1)\
        ff_float_to_int16_##cpu(dst, src[0], len);\
    else if(channels==2){\
        ff_float_to_int16_interleave2_##cpu(dst, src, len);\
    }else if(channels==6){\
        ff_float_to_int16_interleave6_##cpu(dst, src, len);\
    }else\
        float_to_int16_interleave_misc_##cpu(dst, src, len, channels);\
}

FLOAT_TO_INT16_INTERLEAVE(3dnow)
FLOAT_TO_INT16_INTERLEAVE(sse)
FLOAT_TO_INT16_INTERLEAVE(sse2)

static void float_to_int16_interleave_3dn2(int16_t *dst, const float **src, long len, int channels){
    if(channels==6)
        ff_float_to_int16_interleave6_3dn2(dst, src, len);
    else
        float_to_int16_interleave_3dnow(dst, src, len, channels);
}

void ff_float_interleave2_mmx(float *dst, const float **src, unsigned int len);
void ff_float_interleave2_sse(float *dst, const float **src, unsigned int len);

void ff_float_interleave6_mmx(float *dst, const float **src, unsigned int len);
void ff_float_interleave6_sse(float *dst, const float **src, unsigned int len);

static void float_interleave_mmx(float *dst, const float **src,
                                 unsigned int len, int channels)
{
    if (channels == 2) {
        ff_float_interleave2_mmx(dst, src, len);
    } else if (channels == 6)
        ff_float_interleave6_mmx(dst, src, len);
    else
        ff_float_interleave_c(dst, src, len, channels);
}

static void float_interleave_sse(float *dst, const float **src,
                                 unsigned int len, int channels)
{
    if (channels == 2) {
        ff_float_interleave2_sse(dst, src, len);
    } else if (channels == 6)
        ff_float_interleave6_sse(dst, src, len);
    else
        ff_float_interleave_c(dst, src, len, channels);
}
#endif

void ff_fmt_convert_init_x86(FmtConvertContext *c, AVCodecContext *avctx)
{
#if HAVE_YASM
    int mm_flags = av_get_cpu_flags();

    if (mm_flags & AV_CPU_FLAG_MMX) {
        c->float_interleave = float_interleave_mmx;

        if (HAVE_AMD3DNOW && mm_flags & AV_CPU_FLAG_3DNOW) {
            if(!(avctx->flags & CODEC_FLAG_BITEXACT)){
                c->float_to_int16 = ff_float_to_int16_3dnow;
                c->float_to_int16_interleave = float_to_int16_interleave_3dnow;
            }
        }
        if (HAVE_AMD3DNOWEXT && mm_flags & AV_CPU_FLAG_3DNOWEXT) {
            if(!(avctx->flags & CODEC_FLAG_BITEXACT)){
                c->float_to_int16_interleave = float_to_int16_interleave_3dn2;
            }
        }
        if (HAVE_SSE && mm_flags & AV_CPU_FLAG_SSE) {
            c->int32_to_float_fmul_scalar = ff_int32_to_float_fmul_scalar_sse;
            c->float_to_int16 = ff_float_to_int16_sse;
            c->float_to_int16_interleave = float_to_int16_interleave_sse;
            c->float_interleave = float_interleave_sse;
        }
        if (HAVE_SSE && mm_flags & AV_CPU_FLAG_SSE2) {
            c->int32_to_float_fmul_scalar = ff_int32_to_float_fmul_scalar_sse2;
            c->float_to_int16 = ff_float_to_int16_sse2;
            c->float_to_int16_interleave = float_to_int16_interleave_sse2;
        }
    }
#endif
}