flacdsp_init.c (5704B)
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 26 void ff_flac_lpc_16_sse4(int32_t *samples, const int coeffs[32], int order, 27 int qlevel, int len); 28 void ff_flac_lpc_32_sse4(int32_t *samples, const int coeffs[32], int order, 29 int qlevel, int len); 30 void ff_flac_lpc_32_xop(int32_t *samples, const int coeffs[32], int order, 31 int qlevel, int len); 32 33 void ff_flac_wasted_32_sse2(int32_t *decoded, int wasted, int len); 34 void ff_flac_wasted_33_sse4(int64_t *decoded, const int32_t *residual, int wasted, int len); 35 36 #define DECORRELATE_FUNCS(fmt, opt) \ 37 void ff_flac_decorrelate_ls_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ 38 int len, int shift); \ 39 void ff_flac_decorrelate_rs_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ 40 int len, int shift); \ 41 void ff_flac_decorrelate_ms_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ 42 int len, int shift) 43 44 #define DECORRELATE_IFUNCS(fmt, opt) \ 45 void ff_flac_decorrelate_indep2_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ 46 int len, int shift); \ 47 void ff_flac_decorrelate_indep4_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ 48 int len, int shift); \ 49 void ff_flac_decorrelate_indep6_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ 50 int len, int shift); \ 51 void ff_flac_decorrelate_indep8_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ 52 int len, int shift) 53 54 DECORRELATE_FUNCS(16, sse2); 55 DECORRELATE_FUNCS(16, avx); 56 DECORRELATE_FUNCS(32, sse2); 57 DECORRELATE_FUNCS(32, avx); 58 DECORRELATE_IFUNCS(16, ssse3); 59 DECORRELATE_IFUNCS(16, avx); 60 DECORRELATE_IFUNCS(32, ssse3); 61 DECORRELATE_IFUNCS(32, avx); 62 63 av_cold void ff_flacdsp_init_x86(FLACDSPContext *c, enum AVSampleFormat fmt, int channels) 64 { 65 #if HAVE_X86ASM 66 int cpu_flags = av_get_cpu_flags(); 67 68 if (EXTERNAL_SSE2(cpu_flags)) { 69 c->wasted32 = ff_flac_wasted_32_sse2; 70 if (fmt == AV_SAMPLE_FMT_S16) { 71 c->decorrelate[1] = ff_flac_decorrelate_ls_16_sse2; 72 c->decorrelate[2] = ff_flac_decorrelate_rs_16_sse2; 73 c->decorrelate[3] = ff_flac_decorrelate_ms_16_sse2; 74 } else if (fmt == AV_SAMPLE_FMT_S32) { 75 c->decorrelate[1] = ff_flac_decorrelate_ls_32_sse2; 76 c->decorrelate[2] = ff_flac_decorrelate_rs_32_sse2; 77 c->decorrelate[3] = ff_flac_decorrelate_ms_32_sse2; 78 } 79 } 80 if (EXTERNAL_SSSE3(cpu_flags)) { 81 if (fmt == AV_SAMPLE_FMT_S16) { 82 if (channels == 2) 83 c->decorrelate[0] = ff_flac_decorrelate_indep2_16_ssse3; 84 else if (channels == 4) 85 c->decorrelate[0] = ff_flac_decorrelate_indep4_16_ssse3; 86 else if (channels == 6) 87 c->decorrelate[0] = ff_flac_decorrelate_indep6_16_ssse3; 88 else if (ARCH_X86_64 && channels == 8) 89 c->decorrelate[0] = ff_flac_decorrelate_indep8_16_ssse3; 90 } else if (fmt == AV_SAMPLE_FMT_S32) { 91 if (channels == 2) 92 c->decorrelate[0] = ff_flac_decorrelate_indep2_32_ssse3; 93 else if (channels == 4) 94 c->decorrelate[0] = ff_flac_decorrelate_indep4_32_ssse3; 95 else if (channels == 6) 96 c->decorrelate[0] = ff_flac_decorrelate_indep6_32_ssse3; 97 else if (ARCH_X86_64 && channels == 8) 98 c->decorrelate[0] = ff_flac_decorrelate_indep8_32_ssse3; 99 } 100 } 101 if (EXTERNAL_SSE4(cpu_flags)) { 102 c->lpc16 = ff_flac_lpc_16_sse4; 103 c->lpc32 = ff_flac_lpc_32_sse4; 104 c->wasted33 = ff_flac_wasted_33_sse4; 105 } 106 if (EXTERNAL_AVX(cpu_flags)) { 107 if (fmt == AV_SAMPLE_FMT_S16) { 108 if (ARCH_X86_64 && channels == 8) 109 c->decorrelate[0] = ff_flac_decorrelate_indep8_16_avx; 110 } else if (fmt == AV_SAMPLE_FMT_S32) { 111 if (channels == 4) 112 c->decorrelate[0] = ff_flac_decorrelate_indep4_32_avx; 113 else if (channels == 6) 114 c->decorrelate[0] = ff_flac_decorrelate_indep6_32_avx; 115 else if (ARCH_X86_64 && channels == 8) 116 c->decorrelate[0] = ff_flac_decorrelate_indep8_32_avx; 117 } 118 } 119 if (EXTERNAL_XOP(cpu_flags)) { 120 c->lpc32 = ff_flac_lpc_32_xop; 121 } 122 #endif /* HAVE_X86ASM */ 123 }