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
144
145
|
/*
* Copyright (C) 2013 Xiaolei Yu <dreifachstein@gmail.com>
*
* 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
*/
#include "config.h"
#include "libswscale/swscale.h"
#include "libswscale/swscale_internal.h"
#include "libavutil/arm/cpu.h"
#if 0
extern void rgbx_to_nv12_neon_32(const uint8_t *src, uint8_t *y, uint8_t *chroma,
int width, int height,
int y_stride, int c_stride, int src_stride,
int32_t coeff_tbl[9]);
extern void rgbx_to_nv12_neon_16(const uint8_t *src, uint8_t *y, uint8_t *chroma,
int width, int height,
int y_stride, int c_stride, int src_stride,
int32_t coeff_tbl[9]);
static int rgbx_to_nv12_neon_32_wrapper(SwsContext *context, const uint8_t *src[],
int srcStride[], int srcSliceY, int srcSliceH,
uint8_t *dst[], int dstStride[]) {
rgbx_to_nv12_neon_32(src[0] + srcSliceY * srcStride[0],
dst[0] + srcSliceY * dstStride[0],
dst[1] + (srcSliceY / 2) * dstStride[1],
context->srcW, srcSliceH,
dstStride[0], dstStride[1], srcStride[0],
context->input_rgb2yuv_table);
return 0;
}
static int rgbx_to_nv12_neon_16_wrapper(SwsContext *context, const uint8_t *src[],
int srcStride[], int srcSliceY, int srcSliceH,
uint8_t *dst[], int dstStride[]) {
rgbx_to_nv12_neon_16(src[0] + srcSliceY * srcStride[0],
dst[0] + srcSliceY * dstStride[0],
dst[1] + (srcSliceY / 2) * dstStride[1],
context->srcW, srcSliceH,
dstStride[0], dstStride[1], srcStride[0],
context->input_rgb2yuv_table);
return 0;
}
#endif
#define DECLARE_FF_NVX_TO_RGBX_FUNCS(ifmt, ofmt) \
int ff_##ifmt##_to_##ofmt##_neon(int w, int h, \
uint8_t *dst, int linesize, \
const uint8_t *srcY, int linesizeY, \
const uint8_t *srcC, int linesizeC, \
const int16_t *table, \
int y_offset, \
int y_coeff); \
\
static int ifmt##_to_##ofmt##_neon_wrapper(SwsContext *c, const uint8_t *src[], \
int srcStride[], int srcSliceY, int srcSliceH, \
uint8_t *dst[], int dstStride[]) { \
const int16_t yuv2rgb_table[] = { \
c->yuv2rgb_v2r_coeff, \
c->yuv2rgb_u2g_coeff, \
c->yuv2rgb_v2g_coeff, \
c->yuv2rgb_u2b_coeff, \
}; \
\
ff_##ifmt##_to_##ofmt##_neon(c->srcW, srcSliceH, \
dst[0] + srcSliceY * dstStride[0], dstStride[0], \
src[0] + srcSliceY * srcStride[0], srcStride[0], \
src[1] + (srcSliceY / 2) * srcStride[1], srcStride[1], \
yuv2rgb_table, \
c->yuv2rgb_y_offset >> 9, \
c->yuv2rgb_y_coeff); \
\
return 0; \
}
#define DECLARE_FF_NVX_TO_ALL_RGBX_FUNCS(nvx) \
DECLARE_FF_NVX_TO_RGBX_FUNCS(nvx, argb) \
DECLARE_FF_NVX_TO_RGBX_FUNCS(nvx, rgba) \
DECLARE_FF_NVX_TO_RGBX_FUNCS(nvx, abgr) \
DECLARE_FF_NVX_TO_RGBX_FUNCS(nvx, bgra) \
DECLARE_FF_NVX_TO_ALL_RGBX_FUNCS(nv12)
DECLARE_FF_NVX_TO_ALL_RGBX_FUNCS(nv21)
/* We need a 16 pixel width alignment. This constraint can easily be removed
* for input reading but for the output which is 4-bytes per pixel (RGBA) the
* assembly might be writing as much as 4*15=60 extra bytes at the end of the
* line, which won't fit the 32-bytes buffer alignment. */
#define SET_FF_NVX_TO_RGBX_FUNC(ifmt, IFMT, ofmt, OFMT) do { \
if (c->srcFormat == AV_PIX_FMT_##IFMT \
&& c->dstFormat == AV_PIX_FMT_##OFMT \
&& !(c->srcH & 1) \
&& !(c->srcW & 15)) { \
c->swscale = ifmt##_to_##ofmt##_neon_wrapper; \
} \
} while (0)
#define SET_FF_NVX_TO_ALL_RGBX_FUNC(nvx, NVX) do { \
SET_FF_NVX_TO_RGBX_FUNC(nvx, NVX, argb, ARGB); \
SET_FF_NVX_TO_RGBX_FUNC(nvx, NVX, rgba, RGBA); \
SET_FF_NVX_TO_RGBX_FUNC(nvx, NVX, abgr, ABGR); \
SET_FF_NVX_TO_RGBX_FUNC(nvx, NVX, bgra, BGRA); \
} while (0)
static void get_unscaled_swscale_neon(SwsContext *c) {
#if 0
int accurate_rnd = c->flags & SWS_ACCURATE_RND;
if (c->srcFormat == AV_PIX_FMT_RGBA
&& c->dstFormat == AV_PIX_FMT_NV12
&& (c->srcW >= 16)) {
c->swscale = accurate_rnd ? rgbx_to_nv12_neon_32_wrapper
: rgbx_to_nv12_neon_16_wrapper;
}
#endif
SET_FF_NVX_TO_ALL_RGBX_FUNC(nv12, NV12);
SET_FF_NVX_TO_ALL_RGBX_FUNC(nv21, NV21);
}
void ff_get_unscaled_swscale_arm(SwsContext *c)
{
int cpu_flags = av_get_cpu_flags();
if (have_neon(cpu_flags))
get_unscaled_swscale_neon(c);
}
|
