| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | /* | ||
| 2 | * The simplest mpeg encoder (well, it was the simplest!) | ||
| 3 | * Copyright (c) 2000,2001 Fabrice Bellard | ||
| 4 | * | ||
| 5 | * This file is part of FFmpeg. | ||
| 6 | * | ||
| 7 | * FFmpeg 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. | ||
| 11 | * | ||
| 12 | * FFmpeg 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. | ||
| 16 | * | ||
| 17 | * You should have received a copy of the GNU Lesser General Public | ||
| 18 | * License along with FFmpeg; if not, write to the Free Software | ||
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA | ||
| 20 | */ | ||
| 21 | |||
| 22 | #include "libavutil/attributes.h" | ||
| 23 | #include "libavutil/cpu.h" | ||
| 24 | #include "libavutil/mem_internal.h" | ||
| 25 | #include "libavutil/x86/asm.h" | ||
| 26 | #include "libavutil/x86/cpu.h" | ||
| 27 | #include "libavcodec/avcodec.h" | ||
| 28 | #include "libavcodec/mpegvideoenc.h" | ||
| 29 | |||
| 30 | /* not permutated inverse zigzag_direct + 1 for MMX quantizer */ | ||
| 31 | DECLARE_ALIGNED(16, static const uint16_t, inv_zigzag_direct16)[64] = { | ||
| 32 | 1, 2, 6, 7, 15, 16, 28, 29, | ||
| 33 | 3, 5, 8, 14, 17, 27, 30, 43, | ||
| 34 | 4, 9, 13, 18, 26, 31, 42, 44, | ||
| 35 | 10, 12, 19, 25, 32, 41, 45, 54, | ||
| 36 | 11, 20, 24, 33, 40, 46, 53, 55, | ||
| 37 | 21, 23, 34, 39, 47, 52, 56, 61, | ||
| 38 | 22, 35, 38, 48, 51, 57, 60, 62, | ||
| 39 | 36, 37, 49, 50, 58, 59, 63, 64, | ||
| 40 | }; | ||
| 41 | |||
| 42 | #if HAVE_6REGS | ||
| 43 | |||
| 44 | #if HAVE_SSE2_INLINE | ||
| 45 | #define COMPILE_TEMPLATE_SSSE3 0 | ||
| 46 | #define RENAME(a) a ## _sse2 | ||
| 47 | #include "mpegvideoenc_template.c" | ||
| 48 | #endif /* HAVE_SSE2_INLINE */ | ||
| 49 | |||
| 50 | #if HAVE_SSSE3_INLINE | ||
| 51 | #undef COMPILE_TEMPLATE_SSSE3 | ||
| 52 | #define COMPILE_TEMPLATE_SSSE3 1 | ||
| 53 | #undef RENAME | ||
| 54 | #define RENAME(a) a ## _ssse3 | ||
| 55 | #include "mpegvideoenc_template.c" | ||
| 56 | #endif /* HAVE_SSSE3_INLINE */ | ||
| 57 | |||
| 58 | #endif /* HAVE_6REGS */ | ||
| 59 | |||
| 60 | #if HAVE_INLINE_ASM | ||
| 61 | #if HAVE_SSE2_INLINE | ||
| 62 | ✗ | static void denoise_dct_sse2(MPVEncContext *const s, int16_t block[]) | |
| 63 | { | ||
| 64 | ✗ | const int intra = s->c.mb_intra; | |
| 65 | ✗ | int *sum= s->dct_error_sum[intra]; | |
| 66 | ✗ | uint16_t *offset= s->dct_offset[intra]; | |
| 67 | |||
| 68 | ✗ | s->dct_count[intra]++; | |
| 69 | |||
| 70 | ✗ | __asm__ volatile( | |
| 71 | "pxor %%xmm7, %%xmm7 \n\t" | ||
| 72 | "1: \n\t" | ||
| 73 | "pxor %%xmm0, %%xmm0 \n\t" | ||
| 74 | "pxor %%xmm1, %%xmm1 \n\t" | ||
| 75 | "movdqa (%0), %%xmm2 \n\t" | ||
| 76 | "movdqa 16(%0), %%xmm3 \n\t" | ||
| 77 | "pcmpgtw %%xmm2, %%xmm0 \n\t" | ||
| 78 | "pcmpgtw %%xmm3, %%xmm1 \n\t" | ||
| 79 | "pxor %%xmm0, %%xmm2 \n\t" | ||
| 80 | "pxor %%xmm1, %%xmm3 \n\t" | ||
| 81 | "psubw %%xmm0, %%xmm2 \n\t" | ||
| 82 | "psubw %%xmm1, %%xmm3 \n\t" | ||
| 83 | "movdqa %%xmm2, %%xmm4 \n\t" | ||
| 84 | "movdqa %%xmm3, %%xmm5 \n\t" | ||
| 85 | "psubusw (%2), %%xmm2 \n\t" | ||
| 86 | "psubusw 16(%2), %%xmm3 \n\t" | ||
| 87 | "pxor %%xmm0, %%xmm2 \n\t" | ||
| 88 | "pxor %%xmm1, %%xmm3 \n\t" | ||
| 89 | "psubw %%xmm0, %%xmm2 \n\t" | ||
| 90 | "psubw %%xmm1, %%xmm3 \n\t" | ||
| 91 | "movdqa %%xmm2, (%0) \n\t" | ||
| 92 | "movdqa %%xmm3, 16(%0) \n\t" | ||
| 93 | "movdqa %%xmm4, %%xmm6 \n\t" | ||
| 94 | "movdqa %%xmm5, %%xmm0 \n\t" | ||
| 95 | "punpcklwd %%xmm7, %%xmm4 \n\t" | ||
| 96 | "punpckhwd %%xmm7, %%xmm6 \n\t" | ||
| 97 | "punpcklwd %%xmm7, %%xmm5 \n\t" | ||
| 98 | "punpckhwd %%xmm7, %%xmm0 \n\t" | ||
| 99 | "paddd (%1), %%xmm4 \n\t" | ||
| 100 | "paddd 16(%1), %%xmm6 \n\t" | ||
| 101 | "paddd 32(%1), %%xmm5 \n\t" | ||
| 102 | "paddd 48(%1), %%xmm0 \n\t" | ||
| 103 | "movdqa %%xmm4, (%1) \n\t" | ||
| 104 | "movdqa %%xmm6, 16(%1) \n\t" | ||
| 105 | "movdqa %%xmm5, 32(%1) \n\t" | ||
| 106 | "movdqa %%xmm0, 48(%1) \n\t" | ||
| 107 | "add $32, %0 \n\t" | ||
| 108 | "add $64, %1 \n\t" | ||
| 109 | "add $32, %2 \n\t" | ||
| 110 | "cmp %3, %0 \n\t" | ||
| 111 | " jb 1b \n\t" | ||
| 112 | : "+r" (block), "+r" (sum), "+r" (offset) | ||
| 113 | ✗ | : "r"(block+64) | |
| 114 | XMM_CLOBBERS_ONLY("%xmm0", "%xmm1", "%xmm2", "%xmm3", | ||
| 115 | "%xmm4", "%xmm5", "%xmm6", "%xmm7") | ||
| 116 | ); | ||
| 117 | ✗ | } | |
| 118 | #endif /* HAVE_SSE2_INLINE */ | ||
| 119 | #endif /* HAVE_INLINE_ASM */ | ||
| 120 | |||
| 121 | 277 | av_cold void ff_dct_encode_init_x86(MPVEncContext *const s) | |
| 122 | { | ||
| 123 | 277 | const int dct_algo = s->c.avctx->dct_algo; | |
| 124 | |||
| 125 |
3/4✓ Branch 0 taken 273 times.
✓ Branch 1 taken 4 times.
✗ Branch 2 not taken.
✓ Branch 3 taken 273 times.
|
277 | if (dct_algo == FF_DCT_AUTO || dct_algo == FF_DCT_MMX) { |
| 126 | #if HAVE_MMX_INLINE | ||
| 127 | 4 | int cpu_flags = av_get_cpu_flags(); | |
| 128 | #if HAVE_SSE2_INLINE | ||
| 129 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 4 times.
|
4 | if (INLINE_SSE2(cpu_flags)) { |
| 130 | #if HAVE_6REGS | ||
| 131 | ✗ | s->dct_quantize = dct_quantize_sse2; | |
| 132 | #endif | ||
| 133 | ✗ | s->denoise_dct = denoise_dct_sse2; | |
| 134 | } | ||
| 135 | #endif | ||
| 136 | #if HAVE_6REGS && HAVE_SSSE3_INLINE | ||
| 137 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 4 times.
|
4 | if (INLINE_SSSE3(cpu_flags)) |
| 138 | ✗ | s->dct_quantize = dct_quantize_ssse3; | |
| 139 | #endif | ||
| 140 | #endif | ||
| 141 | } | ||
| 142 | 277 | } | |
| 143 |