1/*
2 * Copyright (c) 2014 James Almer
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 "libavutil/attributes.h"
22#include "libavcodec/flacdsp.h"
23#include "libavutil/x86/cpu.h"
24#include "config.h"
25#include "config_components.h"
26
27void ff_flac_lpc_32_sse4(int32_t *samples, const int coeffs[32], int order,
28                         int qlevel, int len);
29void ff_flac_lpc_32_xop(int32_t *samples, const int coeffs[32], int order,
30                        int qlevel, int len);
31
32void ff_flac_enc_lpc_16_sse4(int32_t *, const int32_t *, int, int, const int32_t *,int);
33
34#define DECORRELATE_FUNCS(fmt, opt)                                                      \
35void ff_flac_decorrelate_ls_##fmt##_##opt(uint8_t **out, int32_t **in, int channels,     \
36                                          int len, int shift);                           \
37void ff_flac_decorrelate_rs_##fmt##_##opt(uint8_t **out, int32_t **in, int channels,     \
38                                          int len, int shift);                           \
39void ff_flac_decorrelate_ms_##fmt##_##opt(uint8_t **out, int32_t **in, int channels,     \
40                                          int len, int shift);                           \
41void ff_flac_decorrelate_indep2_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \
42                                             int len, int shift);                        \
43void ff_flac_decorrelate_indep4_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \
44                                              int len, int shift);                       \
45void ff_flac_decorrelate_indep6_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \
46                                              int len, int shift);                       \
47void ff_flac_decorrelate_indep8_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \
48                                              int len, int shift)
49
50DECORRELATE_FUNCS(16, sse2);
51DECORRELATE_FUNCS(16,  avx);
52DECORRELATE_FUNCS(32, sse2);
53DECORRELATE_FUNCS(32,  avx);
54
55av_cold void ff_flacdsp_init_x86(FLACDSPContext *c, enum AVSampleFormat fmt, int channels,
56                                 int bps)
57{
58#if HAVE_X86ASM
59    int cpu_flags = av_get_cpu_flags();
60
61#if CONFIG_FLAC_DECODER
62    if (EXTERNAL_SSE2(cpu_flags)) {
63        if (fmt == AV_SAMPLE_FMT_S16) {
64            if (channels == 2)
65                c->decorrelate[0] = ff_flac_decorrelate_indep2_16_sse2;
66            else if (channels == 4)
67                c->decorrelate[0] = ff_flac_decorrelate_indep4_16_sse2;
68            else if (channels == 6)
69                c->decorrelate[0] = ff_flac_decorrelate_indep6_16_sse2;
70            else if (ARCH_X86_64 && channels == 8)
71                c->decorrelate[0] = ff_flac_decorrelate_indep8_16_sse2;
72            c->decorrelate[1] = ff_flac_decorrelate_ls_16_sse2;
73            c->decorrelate[2] = ff_flac_decorrelate_rs_16_sse2;
74            c->decorrelate[3] = ff_flac_decorrelate_ms_16_sse2;
75        } else if (fmt == AV_SAMPLE_FMT_S32) {
76            if (channels == 2)
77                c->decorrelate[0] = ff_flac_decorrelate_indep2_32_sse2;
78            else if (channels == 4)
79                c->decorrelate[0] = ff_flac_decorrelate_indep4_32_sse2;
80            else if (channels == 6)
81                c->decorrelate[0] = ff_flac_decorrelate_indep6_32_sse2;
82            else if (ARCH_X86_64 && channels == 8)
83                c->decorrelate[0] = ff_flac_decorrelate_indep8_32_sse2;
84            c->decorrelate[1] = ff_flac_decorrelate_ls_32_sse2;
85            c->decorrelate[2] = ff_flac_decorrelate_rs_32_sse2;
86            c->decorrelate[3] = ff_flac_decorrelate_ms_32_sse2;
87        }
88    }
89    if (EXTERNAL_SSE4(cpu_flags)) {
90        c->lpc32 = ff_flac_lpc_32_sse4;
91    }
92    if (EXTERNAL_AVX(cpu_flags)) {
93        if (fmt == AV_SAMPLE_FMT_S16) {
94            if (ARCH_X86_64 && channels == 8)
95                c->decorrelate[0] = ff_flac_decorrelate_indep8_16_avx;
96        } else if (fmt == AV_SAMPLE_FMT_S32) {
97            if (channels == 4)
98                c->decorrelate[0] = ff_flac_decorrelate_indep4_32_avx;
99            else if (channels == 6)
100                c->decorrelate[0] = ff_flac_decorrelate_indep6_32_avx;
101            else if (ARCH_X86_64 && channels == 8)
102                c->decorrelate[0] = ff_flac_decorrelate_indep8_32_avx;
103        }
104    }
105    if (EXTERNAL_XOP(cpu_flags)) {
106        c->lpc32 = ff_flac_lpc_32_xop;
107    }
108#endif
109
110#if CONFIG_FLAC_ENCODER
111    if (EXTERNAL_SSE4(cpu_flags)) {
112        if (CONFIG_GPL)
113            c->lpc16_encode = ff_flac_enc_lpc_16_sse4;
114    }
115#endif
116#endif /* HAVE_X86ASM */
117}
118