/* * 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/aom_config.h" #include "config/av1_rtcd.h" #include "av1/common/av1_inv_txfm1d_cfg.h" #include "av1/common/x86/av1_inv_txfm_ssse3.h" #include "av1/common/x86/av1_txfm_sse2.h" // TODO([email protected]): move this to header file // Sqrt2, Sqrt2^2, Sqrt2^3, Sqrt2^4, Sqrt2^5 static int32_t NewSqrt2list[TX_SIZES] = …; // TODO([email protected]): replace some for loop with do {} while static void idct4_sse2(const __m128i *input, __m128i *output) { … } static void idct4_w4_sse2(const __m128i *input, __m128i *output) { … } void av1_idct8_low1_ssse3(const __m128i *input, __m128i *output) { … } void av1_idct8_sse2(const __m128i *input, __m128i *output) { … } static void idct8_w4_sse2(const __m128i *input, __m128i *output) { … } static inline void idct16_stage5_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct16_stage6_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct16_stage7_sse2(__m128i *output, __m128i *x) { … } static void idct16_low1_ssse3(const __m128i *input, __m128i *output) { … } static void idct16_low8_ssse3(const __m128i *input, __m128i *output) { … } static void idct16_sse2(const __m128i *input, __m128i *output) { … } static void idct16_w4_sse2(const __m128i *input, __m128i *output) { … } static inline void idct32_high16_stage3_sse2(__m128i *x) { … } static inline void idct32_high16_stage4_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct32_high24_stage5_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct32_high28_stage6_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct32_stage7_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct32_stage8_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct32_stage9_sse2(__m128i *output, __m128i *x) { … } static void idct32_low1_ssse3(const __m128i *input, __m128i *output) { … } static void idct32_low8_ssse3(const __m128i *input, __m128i *output) { … } static void idct32_low16_ssse3(const __m128i *input, __m128i *output) { … } static void idct32_sse2(const __m128i *input, __m128i *output) { … } static inline void idct64_stage4_high32_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage5_high48_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage6_high32_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage6_high48_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage7_high48_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage8_high48_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage9_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage10_sse2(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void idct64_stage11_sse2(__m128i *output, __m128i *x) { … } static void idct64_low1_ssse3(const __m128i *input, __m128i *output) { … } static void idct64_low8_ssse3(const __m128i *input, __m128i *output) { … } static void idct64_low16_ssse3(const __m128i *input, __m128i *output) { … } static void idct64_low32_ssse3(const __m128i *input, __m128i *output) { … } static void iadst4_sse2(const __m128i *input, __m128i *output) { … } static void iadst4_w4_sse2(const __m128i *input, __m128i *output) { … } void av1_iadst8_low1_ssse3(const __m128i *input, __m128i *output) { … } void av1_iadst8_sse2(const __m128i *input, __m128i *output) { … } static void iadst8_w4_sse2(const __m128i *input, __m128i *output) { … } static inline void iadst16_stage3_ssse3(__m128i *x) { … } static inline void iadst16_stage4_ssse3(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void iadst16_stage5_ssse3(__m128i *x) { … } static inline void iadst16_stage6_ssse3(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void iadst16_stage7_ssse3(__m128i *x) { … } static inline void iadst16_stage8_ssse3(__m128i *x, const int32_t *cospi, const __m128i __rounding, int8_t cos_bit) { … } static inline void iadst16_stage9_ssse3(__m128i *output, __m128i *x) { … } static void iadst16_low1_ssse3(const __m128i *input, __m128i *output) { … } static void iadst16_low8_ssse3(const __m128i *input, __m128i *output) { … } static void iadst16_sse2(const __m128i *input, __m128i *output) { … } static void iadst16_w4_sse2(const __m128i *input, __m128i *output) { … } static void iidentity4_ssse3(const __m128i *input, __m128i *output) { … } static void iidentity8_sse2(const __m128i *input, __m128i *output) { … } static void iidentity16_ssse3(const __m128i *input, __m128i *output) { … } static inline __m128i lowbd_get_recon_8x8_sse2(const __m128i pred, __m128i res) { … } static inline void lowbd_write_buffer_4xn_sse2(__m128i *in, uint8_t *output, int stride, int flipud, const int height) { … } static inline void lowbd_write_buffer_8xn_sse2(__m128i *in, uint8_t *output, int stride, int flipud, const int height) { … } // 1D functions process process 8 pixels at one time. static const transform_1d_ssse3 lowbd_txfm_all_1d_w8_arr[TX_SIZES][ITX_TYPES_1D] = …; // functions for blocks with eob at DC and within // topleft 8x8, 16x16, 32x32 corner static const transform_1d_ssse3 lowbd_txfm_all_1d_zeros_w8_arr[TX_SIZES][ITX_TYPES_1D][4] = …; // 1D functions process process 4 pixels at one time. // used in 4x4, 4x8, 4x16, 8x4, 16x4 static const transform_1d_ssse3 lowbd_txfm_all_1d_w4_arr[TX_SIZES][ITX_TYPES_1D] = …; static inline void iidentity_row_8xn_ssse3(__m128i *out, const int32_t *input, int stride, int shift, int height, int txw_idx, int rect_type) { … } static inline void iidentity_col_8xn_ssse3(uint8_t *output, int stride, __m128i *buf, int shift, int height, int txh_idx) { … } void av1_lowbd_inv_txfm2d_add_idtx_ssse3(const int32_t *input, uint8_t *output, int stride, TX_SIZE tx_size) { … } static void lowbd_inv_txfm2d_add_4x4_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size_, int eob) { … } static inline __m128i lowbd_get_recon_16x16_sse2(const __m128i pred, __m128i res0, __m128i res1) { … } static inline void lowbd_write_buffer_16xn_sse2(__m128i *in, uint8_t *output, int stride, int flipud, int height) { … } static inline void round_shift_ssse3(const __m128i *input, __m128i *output, int size) { … } static inline void lowbd_inv_txfm2d_add_no_identity_ssse3( const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size, int eob) { … } void av1_lowbd_inv_txfm2d_add_h_identity_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size, int eob) { … } void av1_lowbd_inv_txfm2d_add_v_identity_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size, int eob) { … } // for 32x32,32x64,64x32,64x64,32x8,8x32,16x32,32x16,64x16,16x64 static inline void lowbd_inv_txfm2d_add_universe_ssse3( const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size, int eob) { … } static void lowbd_inv_txfm2d_add_4x8_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size_, int eob) { … } static void lowbd_inv_txfm2d_add_8x4_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size_, int eob) { … } static void lowbd_inv_txfm2d_add_4x16_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size_, int eob) { … } static void lowbd_inv_txfm2d_add_16x4_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size_, int eob) { … } void av1_lowbd_inv_txfm2d_add_ssse3(const int32_t *input, uint8_t *output, int stride, TX_TYPE tx_type, TX_SIZE tx_size, int eob) { … } void av1_inv_txfm_add_ssse3(const tran_low_t *dqcoeff, uint8_t *dst, int stride, const TxfmParam *txfm_param) { … }
Codebase Browser
chromium