2 * ARM NEON optimised HEVC IDCT
3 * Copyright (c) 2017 Alexandra Hájková
5 * This file is part of Libav.
7 * Libav 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.
12 * Libav 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.
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with Libav; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
22 #include "libavutil/attributes.h"
23 #include "libavutil/cpu.h"
24 #include "libavutil/arm/cpu.h"
26 #include "libavcodec/hevcdsp.h"
29 void ff_hevc_add_residual_4x4_8_neon(uint8_t *_dst, int16_t *coeffs,
31 void ff_hevc_add_residual_4x4_10_neon(uint8_t *_dst, int16_t *coeffs,
33 void ff_hevc_add_residual_8x8_8_neon(uint8_t *_dst, int16_t *coeffs,
35 void ff_hevc_add_residual_8x8_10_neon(uint8_t *_dst, int16_t *coeffs,
37 void ff_hevc_add_residual_16x16_8_neon(uint8_t *_dst, int16_t *coeffs,
39 void ff_hevc_add_residual_16x16_10_neon(uint8_t *_dst, int16_t *coeffs,
41 void ff_hevc_add_residual_32x32_8_neon(uint8_t *_dst, int16_t *coeffs,
43 void ff_hevc_add_residual_32x32_10_neon(uint8_t *_dst, int16_t *coeffs,
46 void ff_hevc_idct_4x4_dc_8_neon(int16_t *coeffs);
47 void ff_hevc_idct_8x8_dc_8_neon(int16_t *coeffs);
48 void ff_hevc_idct_16x16_dc_8_neon(int16_t *coeffs);
49 void ff_hevc_idct_32x32_dc_8_neon(int16_t *coeffs);
50 void ff_hevc_idct_4x4_dc_10_neon(int16_t *coeffs);
51 void ff_hevc_idct_8x8_dc_10_neon(int16_t *coeffs);
52 void ff_hevc_idct_16x16_dc_10_neon(int16_t *coeffs);
53 void ff_hevc_idct_32x32_dc_10_neon(int16_t *coeffs);
55 void ff_hevc_idct_4x4_8_neon(int16_t *coeffs, int col_limit);
56 void ff_hevc_idct_8x8_8_neon(int16_t *coeffs, int col_limit);
57 void ff_hevc_idct_16x16_8_neon(int16_t *coeffs, int col_limit);
58 void ff_hevc_idct_32x32_8_neon(int16_t *coeffs, int col_limit);
59 void ff_hevc_idct_4x4_10_neon(int16_t *coeffs, int col_limit);
60 void ff_hevc_idct_8x8_10_neon(int16_t *coeffs, int col_limit);
61 void ff_hevc_idct_16x16_10_neon(int16_t *coeffs, int col_limit);
62 void ff_hevc_idct_32x32_10_neon(int16_t *coeffs, int col_limit);
64 av_cold void ff_hevc_dsp_init_arm(HEVCDSPContext *c, int bit_depth)
66 int cpu_flags = av_get_cpu_flags();
68 if (have_neon(cpu_flags)) {
70 c->add_residual[0] = ff_hevc_add_residual_4x4_8_neon;
71 c->add_residual[1] = ff_hevc_add_residual_8x8_8_neon;
72 c->add_residual[2] = ff_hevc_add_residual_16x16_8_neon;
73 c->add_residual[3] = ff_hevc_add_residual_32x32_8_neon;
75 c->idct_dc[0] = ff_hevc_idct_4x4_dc_8_neon;
76 c->idct_dc[1] = ff_hevc_idct_8x8_dc_8_neon;
77 c->idct_dc[2] = ff_hevc_idct_16x16_dc_8_neon;
78 c->idct_dc[3] = ff_hevc_idct_32x32_dc_8_neon;
80 c->idct[0] = ff_hevc_idct_4x4_8_neon;
81 c->idct[1] = ff_hevc_idct_8x8_8_neon;
82 c->idct[2] = ff_hevc_idct_16x16_8_neon;
83 c->idct[3] = ff_hevc_idct_32x32_8_neon;
85 if (bit_depth == 10) {
86 c->add_residual[0] = ff_hevc_add_residual_4x4_10_neon;
87 c->add_residual[1] = ff_hevc_add_residual_8x8_10_neon;
88 c->add_residual[2] = ff_hevc_add_residual_16x16_10_neon;
89 c->add_residual[3] = ff_hevc_add_residual_32x32_10_neon;
91 c->idct_dc[0] = ff_hevc_idct_4x4_dc_10_neon;
92 c->idct_dc[1] = ff_hevc_idct_8x8_dc_10_neon;
93 c->idct_dc[2] = ff_hevc_idct_16x16_dc_10_neon;
94 c->idct_dc[3] = ff_hevc_idct_32x32_dc_10_neon;
96 c->idct[0] = ff_hevc_idct_4x4_10_neon;
97 c->idct[1] = ff_hevc_idct_8x8_10_neon;
98 c->idct[2] = ff_hevc_idct_16x16_10_neon;
99 c->idct[3] = ff_hevc_idct_32x32_10_neon;