| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | /* | ||
| 2 | * Copyright (C) 2012 Michael Niedermayer (michaelni@gmx.at) | ||
| 3 | * | ||
| 4 | * This file is part of libswresample | ||
| 5 | * | ||
| 6 | * libswresample 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 | * libswresample 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 libswresample; 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 "libavutil/mem.h" | ||
| 23 | #include "libavutil/x86/cpu.h" | ||
| 24 | #include "libswresample/swresample_internal.h" | ||
| 25 | |||
| 26 | #define D(type, simd) \ | ||
| 27 | mix_1_1_func_type ff_mix_1_1_a_## type ## _ ## simd;\ | ||
| 28 | mix_2_1_func_type ff_mix_2_1_a_## type ## _ ## simd; | ||
| 29 | |||
| 30 | D(float, sse) | ||
| 31 | D(float, avx) | ||
| 32 | D(int16, sse2) | ||
| 33 | |||
| 34 | 25 | av_cold int swri_rematrix_init_x86(struct SwrContext *s){ | |
| 35 | #if HAVE_X86ASM | ||
| 36 | 25 | int mm_flags = av_get_cpu_flags(); | |
| 37 | 25 | int nb_in = s->used_ch_layout.nb_channels; | |
| 38 | 25 | int nb_out = s->out.ch_count; | |
| 39 | 25 | int num = nb_in * nb_out; | |
| 40 | int i,j; | ||
| 41 | |||
| 42 | 25 | s->mix_1_1_simd = NULL; | |
| 43 | 25 | s->mix_2_1_simd = NULL; | |
| 44 | |||
| 45 |
2/2✓ Branch 0 taken 18 times.
✓ Branch 1 taken 7 times.
|
25 | if (s->midbuf.fmt == AV_SAMPLE_FMT_S16P){ |
| 46 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 18 times.
|
18 | if(EXTERNAL_SSE2(mm_flags)) { |
| 47 | ✗ | s->mix_1_1_simd = ff_mix_1_1_a_int16_sse2; | |
| 48 | ✗ | s->mix_2_1_simd = ff_mix_2_1_a_int16_sse2; | |
| 49 | } | ||
| 50 | 18 | s->native_simd_matrix = av_calloc(num, 2 * sizeof(int16_t)); | |
| 51 | 18 | s->native_simd_one = av_mallocz(2 * sizeof(int16_t)); | |
| 52 |
2/4✓ Branch 0 taken 18 times.
✗ Branch 1 not taken.
✗ Branch 2 not taken.
✓ Branch 3 taken 18 times.
|
18 | if (!s->native_simd_matrix || !s->native_simd_one) |
| 53 | ✗ | return AVERROR(ENOMEM); | |
| 54 | |||
| 55 |
2/2✓ Branch 0 taken 40 times.
✓ Branch 1 taken 18 times.
|
58 | for(i=0; i<nb_out; i++){ |
| 56 | 40 | int sh = 0; | |
| 57 |
2/2✓ Branch 0 taken 150 times.
✓ Branch 1 taken 40 times.
|
190 | for(j=0; j<nb_in; j++) |
| 58 | 150 | sh = FFMAX(sh, FFABS(((int*)s->native_matrix)[i * nb_in + j])); | |
| 59 | 40 | sh = FFMAX(av_log2(sh) - 14, 0); | |
| 60 |
2/2✓ Branch 0 taken 150 times.
✓ Branch 1 taken 40 times.
|
190 | for(j=0; j<nb_in; j++) { |
| 61 | 150 | ((int16_t*)s->native_simd_matrix)[2*(i * nb_in + j)+1] = 15 - sh; | |
| 62 | 150 | ((int16_t*)s->native_simd_matrix)[2*(i * nb_in + j)] = | |
| 63 | 150 | ((((int*)s->native_matrix)[i * nb_in + j]) + (1<<sh>>1)) >> sh; | |
| 64 | } | ||
| 65 | } | ||
| 66 | 18 | ((int16_t*)s->native_simd_one)[1] = 14; | |
| 67 | 18 | ((int16_t*)s->native_simd_one)[0] = 16384; | |
| 68 |
1/2✓ Branch 0 taken 7 times.
✗ Branch 1 not taken.
|
7 | } else if(s->midbuf.fmt == AV_SAMPLE_FMT_FLTP){ |
| 69 |
2/2✓ Branch 0 taken 1 times.
✓ Branch 1 taken 6 times.
|
7 | if(EXTERNAL_SSE(mm_flags)) { |
| 70 | 1 | s->mix_1_1_simd = ff_mix_1_1_a_float_sse; | |
| 71 | 1 | s->mix_2_1_simd = ff_mix_2_1_a_float_sse; | |
| 72 | } | ||
| 73 |
3/4✓ Branch 0 taken 1 times.
✓ Branch 1 taken 6 times.
✓ Branch 2 taken 1 times.
✗ Branch 3 not taken.
|
7 | if(EXTERNAL_AVX_FAST(mm_flags)) { |
| 74 | 1 | s->mix_1_1_simd = ff_mix_1_1_a_float_avx; | |
| 75 | 1 | s->mix_2_1_simd = ff_mix_2_1_a_float_avx; | |
| 76 | } | ||
| 77 | 7 | s->native_simd_matrix = av_calloc(num, sizeof(float)); | |
| 78 | 7 | s->native_simd_one = av_mallocz(sizeof(float)); | |
| 79 |
2/4✓ Branch 0 taken 7 times.
✗ Branch 1 not taken.
✗ Branch 2 not taken.
✓ Branch 3 taken 7 times.
|
7 | if (!s->native_simd_matrix || !s->native_simd_one) |
| 80 | ✗ | return AVERROR(ENOMEM); | |
| 81 | 7 | memcpy(s->native_simd_matrix, s->native_matrix, num * sizeof(float)); | |
| 82 | 7 | memcpy(s->native_simd_one, s->native_one, sizeof(float)); | |
| 83 | } | ||
| 84 | #endif | ||
| 85 | |||
| 86 | 25 | return 0; | |
| 87 | } | ||
| 88 |