/* * Copyright (c) 2018, Alliance for Open Media. All rights reserved. * * This source code is subject to the terms of the BSD 2 Clause License and * the Alliance for Open Media Patent License 1.0. If the BSD 2 Clause License * was not distributed with this source code in the LICENSE file, you can * obtain it at www.aomedia.org/license/software. If the Alliance for Open * Media Patent License 1.0 was not distributed with this source code in the * PATENTS file, you can obtain it at www.aomedia.org/license/patent. */ #include "config/av1_rtcd.h" #include "av1/common/enums.h" #include "av1/common/av1_txfm.h" #include "av1/encoder/x86/av1_fwd_txfm_avx2.h" #include "av1/common/x86/av1_txfm_sse2.h" #include "av1/encoder/av1_fwd_txfm1d_cfg.h" #include "av1/encoder/x86/av1_txfm1d_sse4.h" #include "av1/encoder/x86/av1_fwd_txfm_sse2.h" #include "aom_dsp/x86/txfm_common_avx2.h" static inline void fdct16x16_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fdct16x32_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fdct16x64_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fdct32_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fdct64_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fadst16x16_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fidentity16x16_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fidentity16x32_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void store_output_32bit_w16(int32_t *const out, const __m256i *const in1, const __m256i *const in2, const int stride, const int out_size) { … } // Store 8 16 bit values. Sign extend the values. static inline void store_buffer_16bit_to_32bit_w16_avx2(const __m256i *const in, int32_t *out, const int stride, const int out_size) { … } static inline void store_rect_16bit_to_32bit_avx2(const __m256i a, int32_t *const b) { … } static inline void store_rect_buffer_16bit_to_32bit_w16_avx2( const __m256i *const in, int32_t *const out, const int stride, const int out_size) { … } transform_1d_avx2; static const transform_1d_avx2 col_txfm16x32_arr[TX_TYPES] = …; static const transform_1d_avx2 row_txfm16x32_arr[TX_TYPES] = …; static const transform_1d_avx2 col_txfm16x16_arr[TX_TYPES] = …; static const transform_1d_avx2 row_txfm16x16_arr[TX_TYPES] = …; static const transform_1d_sse2 col_txfm8x8_arr[TX_TYPES] = …; static const transform_1d_sse2 row_txfm8x8_arr[TX_TYPES] = …; static inline void load_buffer_and_round_shift(const int16_t *in, int stride, __m128i *out, int bit) { … } static inline void load_buffer_and_flip_round_shift(const int16_t *in, int stride, __m128i *out, int bit) { … } #define TRANSPOSE_8X8_AVX2() … static inline void transpose_round_shift_flip_8x8(__m128i *const in, __m128i *const out, int bit) { … } static inline void transpose_round_shift_8x8(__m128i *const in, __m128i *const out, int bit) { … } static inline void store_buffer_16bit_to_32bit_w8_avx2(const __m128i *const in, int32_t *const out, const int stride, const int out_size) { … } static void av1_lowbd_fwd_txfm2d_8x8_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_16x16_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_32x32_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_64x64_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_16x32_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_32x16_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_64x32_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_32x64_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_16x64_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_64x16_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static inline void btf_16_avx2(__m256i *w0, __m256i *w1, __m256i *in0, __m256i *in1, __m128i *out0, __m128i *out1, __m128i *out2, __m128i *out3, const __m256i *__rounding, int8_t *cos_bit) { … } static inline void fdct8x8_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fadst8x8_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fidentity8x8_new_avx2(const __m256i *input, __m256i *output, int8_t cos_bit) { … } static inline void fdct8x16_new_avx2(const __m128i *input, __m128i *output, int8_t cos_bit) { … } static inline void fadst8x16_new_avx2(const __m128i *input, __m128i *output, int8_t cos_bit) { … } static inline void fidentity8x16_new_avx2(const __m128i *input, __m128i *output, int8_t cos_bit) { … } static const transform_1d_avx2 row_txfm8x16_arr[TX_TYPES] = …; static const transform_1d_sse2 col_txfm8x16_arr[TX_TYPES] = …; static const transform_1d_avx2 col_txfm16x8_arr[TX_TYPES] = …; static const transform_1d_sse2 row_txfm16x8_arr[TX_TYPES] = …; static void lowbd_fwd_txfm2d_8x16_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static void lowbd_fwd_txfm2d_16x8_avx2(const int16_t *input, int32_t *output, int stride, TX_TYPE tx_type, int bd) { … } static FwdTxfm2dFunc fwd_txfm2d_func_ls[TX_SIZES_ALL] = …; void av1_lowbd_fwd_txfm_avx2(const int16_t *src_diff, tran_low_t *coeff, int diff_stride, TxfmParam *txfm_param) { … }
Codebase Browser
chromium