FFmpeg
Loading...
Searching...
No Matches
vp8dsp_init.c
Go to the documentation of this file.
1/*
2 * Copyright (c) 2024 Institute of Software Chinese Academy of Sciences (ISCAS).
3 *
4 * This file is part of FFmpeg.
5 *
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
10 *
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 */
20
21#include "config.h"
22
23#include "libavutil/attributes.h"
24#include "libavutil/cpu.h"
25#include "libavutil/riscv/cpu.h"
26#include "libavcodec/vp8dsp.h"
27#include "vp8dsp.h"
28
29void ff_vp8_luma_dc_wht_rvv(int16_t block[4][4][16], int16_t dc[16]);
30void ff_vp8_idct_add_rvv(uint8_t *dst, int16_t block[16], ptrdiff_t stride);
31void ff_vp8_idct_dc_add_rvv(uint8_t *dst, int16_t block[16], ptrdiff_t stride);
32void ff_vp8_idct_dc_add4y_rvv(uint8_t *dst, int16_t block[4][16], ptrdiff_t stride);
33void ff_vp8_idct_dc_add4uv_rvv(uint8_t *dst, int16_t block[4][16], ptrdiff_t stride);
34
35VP8_EPEL(16, rvi);
36VP8_EPEL(8, rvi);
37VP8_EPEL(4, rvi);
38VP8_EPEL(16, rvv);
39VP8_EPEL(8, rvv);
40VP8_EPEL(4, rvv);
41
42VP8_BILIN(16, rvv);
43VP8_BILIN(8, rvv);
44VP8_BILIN(4, rvv);
45
46av_cold void ff_vp78dsp_init_riscv(VP8DSPContext *c)
47{
48#if HAVE_RV
49 int flags = av_get_cpu_flags();
50 if (flags & AV_CPU_FLAG_RV_MISALIGNED) {
51#if __riscv_xlen >= 64
52 c->put_vp8_epel_pixels_tab[0][0][0] = ff_put_vp8_pixels16_rvi;
53 c->put_vp8_epel_pixels_tab[1][0][0] = ff_put_vp8_pixels8_rvi;
54 c->put_vp8_bilinear_pixels_tab[0][0][0] = ff_put_vp8_pixels16_rvi;
55 c->put_vp8_bilinear_pixels_tab[1][0][0] = ff_put_vp8_pixels8_rvi;
56#endif
57 c->put_vp8_epel_pixels_tab[2][0][0] = ff_put_vp8_pixels4_rvi;
58 c->put_vp8_bilinear_pixels_tab[2][0][0] = ff_put_vp8_pixels4_rvi;
59 }
60#if HAVE_RVV
61 if (flags & AV_CPU_FLAG_RVV_I32 && ff_rv_vlen_least(128)) {
62 c->put_vp8_bilinear_pixels_tab[0][0][1] = ff_put_vp8_bilin16_h_rvv;
63 c->put_vp8_bilinear_pixels_tab[0][0][2] = ff_put_vp8_bilin16_h_rvv;
64 c->put_vp8_bilinear_pixels_tab[1][0][1] = ff_put_vp8_bilin8_h_rvv;
65 c->put_vp8_bilinear_pixels_tab[1][0][2] = ff_put_vp8_bilin8_h_rvv;
66 c->put_vp8_bilinear_pixels_tab[2][0][1] = ff_put_vp8_bilin4_h_rvv;
67 c->put_vp8_bilinear_pixels_tab[2][0][2] = ff_put_vp8_bilin4_h_rvv;
68
69 c->put_vp8_bilinear_pixels_tab[0][1][0] = ff_put_vp8_bilin16_v_rvv;
70 c->put_vp8_bilinear_pixels_tab[0][2][0] = ff_put_vp8_bilin16_v_rvv;
71 c->put_vp8_bilinear_pixels_tab[1][1][0] = ff_put_vp8_bilin8_v_rvv;
72 c->put_vp8_bilinear_pixels_tab[1][2][0] = ff_put_vp8_bilin8_v_rvv;
73 c->put_vp8_bilinear_pixels_tab[2][1][0] = ff_put_vp8_bilin4_v_rvv;
74 c->put_vp8_bilinear_pixels_tab[2][2][0] = ff_put_vp8_bilin4_v_rvv;
75
76 c->put_vp8_bilinear_pixels_tab[0][1][1] = ff_put_vp8_bilin16_hv_rvv;
77 c->put_vp8_bilinear_pixels_tab[0][1][2] = ff_put_vp8_bilin16_hv_rvv;
78 c->put_vp8_bilinear_pixels_tab[0][2][1] = ff_put_vp8_bilin16_hv_rvv;
79 c->put_vp8_bilinear_pixels_tab[0][2][2] = ff_put_vp8_bilin16_hv_rvv;
80 c->put_vp8_bilinear_pixels_tab[1][1][1] = ff_put_vp8_bilin8_hv_rvv;
81 c->put_vp8_bilinear_pixels_tab[1][1][2] = ff_put_vp8_bilin8_hv_rvv;
82 c->put_vp8_bilinear_pixels_tab[1][2][1] = ff_put_vp8_bilin8_hv_rvv;
83 c->put_vp8_bilinear_pixels_tab[1][2][2] = ff_put_vp8_bilin8_hv_rvv;
84 c->put_vp8_bilinear_pixels_tab[2][1][1] = ff_put_vp8_bilin4_hv_rvv;
85 c->put_vp8_bilinear_pixels_tab[2][1][2] = ff_put_vp8_bilin4_hv_rvv;
86 c->put_vp8_bilinear_pixels_tab[2][2][1] = ff_put_vp8_bilin4_hv_rvv;
87 c->put_vp8_bilinear_pixels_tab[2][2][2] = ff_put_vp8_bilin4_hv_rvv;
88
89 if (flags & AV_CPU_FLAG_RVB) {
90 c->put_vp8_epel_pixels_tab[0][0][2] = ff_put_vp8_epel16_h6_rvv;
91 c->put_vp8_epel_pixels_tab[1][0][2] = ff_put_vp8_epel8_h6_rvv;
92 c->put_vp8_epel_pixels_tab[2][0][2] = ff_put_vp8_epel4_h6_rvv;
93 c->put_vp8_epel_pixels_tab[1][0][1] = ff_put_vp8_epel8_h4_rvv;
94 c->put_vp8_epel_pixels_tab[2][0][1] = ff_put_vp8_epel4_h4_rvv;
95
96 c->put_vp8_epel_pixels_tab[0][2][0] = ff_put_vp8_epel16_v6_rvv;
97 c->put_vp8_epel_pixels_tab[1][2][0] = ff_put_vp8_epel8_v6_rvv;
98 c->put_vp8_epel_pixels_tab[2][2][0] = ff_put_vp8_epel4_v6_rvv;
99 c->put_vp8_epel_pixels_tab[1][1][0] = ff_put_vp8_epel8_v4_rvv;
100 c->put_vp8_epel_pixels_tab[2][1][0] = ff_put_vp8_epel4_v4_rvv;
101#if __riscv_xlen <= 64
102 c->put_vp8_epel_pixels_tab[0][2][2] = ff_put_vp8_epel16_h6v6_rvv;
103 c->put_vp8_epel_pixels_tab[1][2][2] = ff_put_vp8_epel8_h6v6_rvv;
104 c->put_vp8_epel_pixels_tab[2][2][2] = ff_put_vp8_epel4_h6v6_rvv;
105 c->put_vp8_epel_pixels_tab[1][2][1] = ff_put_vp8_epel8_h4v6_rvv;
106 c->put_vp8_epel_pixels_tab[2][2][1] = ff_put_vp8_epel4_h4v6_rvv;
107 c->put_vp8_epel_pixels_tab[1][1][1] = ff_put_vp8_epel8_h4v4_rvv;
108 c->put_vp8_epel_pixels_tab[2][1][1] = ff_put_vp8_epel4_h4v4_rvv;
109 c->put_vp8_epel_pixels_tab[1][1][2] = ff_put_vp8_epel8_h6v4_rvv;
110 c->put_vp8_epel_pixels_tab[2][1][2] = ff_put_vp8_epel4_h6v4_rvv;
111#endif
112 }
113 }
114#endif
115#endif
116}
117
118av_cold void ff_vp8dsp_init_riscv(VP8DSPContext *c)
119{
120#if HAVE_RVV
121 int flags = av_get_cpu_flags();
122
123 if (flags & AV_CPU_FLAG_RVV_I32 && ff_rv_vlen_least(128)) {
124#if __riscv_xlen >= 64
125 if (flags & AV_CPU_FLAG_RVV_I64)
126 c->vp8_luma_dc_wht = ff_vp8_luma_dc_wht_rvv;
127 c->vp8_idct_add = ff_vp8_idct_add_rvv;
128#endif
129 c->vp8_idct_dc_add = ff_vp8_idct_dc_add_rvv;
130 c->vp8_idct_dc_add4y = ff_vp8_idct_dc_add4y_rvv;
131 if (flags & AV_CPU_FLAG_RVV_I64)
132 c->vp8_idct_dc_add4uv = ff_vp8_idct_dc_add4uv_rvv;
133 }
134#endif
135}
dst
uint8_t ptrdiff_t const uint8_t ptrdiff_t int intptr_t intptr_t int int16_t * dst
Definition dsp.h:87
VP8_BILIN
#define VP8_BILIN(w, opt)
Definition vp8dsp.h:70
VP8_EPEL
#define VP8_EPEL(w, opt)
Definition vp8dsp.h:59
flags
#define flags(name, subs,...)
Definition cbs_h264.c:74
block
static int16_t block[64]
Definition dct.c:125
attributes.h
Macro definitions for various function/variable attributes.
av_cold
#define av_cold
Definition attributes.h:117
av_get_cpu_flags
int av_get_cpu_flags(void)
Return the flags which specify extensions supported by the CPU.
Definition cpu.c:109
AV_CPU_FLAG_RVV_I64
#define AV_CPU_FLAG_RVV_I64
Vectors of 64-bit int's *‍/.
Definition cpu.h:99
AV_CPU_FLAG_RVB
#define AV_CPU_FLAG_RVB
B (bit manipulations)
Definition cpu.h:104
AV_CPU_FLAG_RVV_I32
#define AV_CPU_FLAG_RVV_I32
Vectors of 8/16/32-bit int's *‍/.
Definition cpu.h:97
AV_CPU_FLAG_RV_MISALIGNED
#define AV_CPU_FLAG_RV_MISALIGNED
Fast misaligned accesses.
Definition cpu.h:103
ff_vp8_idct_dc_add4uv_rvv
void ff_vp8_idct_dc_add4uv_rvv(uint8_t *dst, int16_t block[4][16], ptrdiff_t stride)
ff_vp8dsp_init_riscv
av_cold void ff_vp8dsp_init_riscv(VP8DSPContext *c)
Definition vp8dsp_init.c:118
ff_vp78dsp_init_riscv
av_cold void ff_vp78dsp_init_riscv(VP8DSPContext *c)
Definition vp8dsp_init.c:46
ff_vp8_idct_dc_add4y_rvv
void ff_vp8_idct_dc_add4y_rvv(uint8_t *dst, int16_t block[4][16], ptrdiff_t stride)
ff_vp8_idct_add_rvv
void ff_vp8_idct_add_rvv(uint8_t *dst, int16_t block[16], ptrdiff_t stride)
ff_vp8_luma_dc_wht_rvv
void ff_vp8_luma_dc_wht_rvv(int16_t block[4][4][16], int16_t dc[16])
ff_vp8_idct_dc_add_rvv
void ff_vp8_idct_dc_add_rvv(uint8_t *dst, int16_t block[16], ptrdiff_t stride)
stride
#define stride
Definition h264pred_template.c:536
vp8dsp.h
VP8 compatible video decoder.
c
static double c[64]
Definition vsrc_mptestsrc.c:89
Generated on Wed Aug 19 2026 20:20:52 for FFmpeg by doxygen 1.13.2