/* * Copyright ( c ) 2023 , 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 "warp_plane_neon.h" 16 , static const uint8_t, usdot_permute_idx[48 ]) = {0 , 1 , 2 , 3 , 1 , 2 , 3 , 4 , 2 , 3 , 4 , 5 , 3 , 4 , 5 , 6 ,4 , 5 , 6 , 7 , 5 , 6 , 7 , 8 , 6 , 7 , 8 , 9 , 7 , 8 , 9 , 10 ,8 , 9 , 10 , 11 , 9 , 10 , 11 , 12 , 10 , 11 , 12 , 13 , 11 , 12 , 13 , 14 static AOM_FORCE_INLINE int16x8_t horizontal_filter_4x1_f4(const uint8x16_t in,int sx, int alpha) {// Only put the constant in every other lane to avoid double-counting when // performing the pairwise add later. const int32x4_t add_const =1 << (8 + FILTER_BITS - 1 )));// Loading the 8 filter taps 4 ];0 ]), vmovn_s16(f[1 ]));2 ]), vmovn_s16(f[3 ]));1 ));2 ));3 ));0 ));return vreinterpretq_s16_u16(res);static AOM_FORCE_INLINE int16x8_t horizontal_filter_8x1_f8(const uint8x16_t in,int sx, int alpha) {// Only put the constant in every other lane to avoid double-counting when // performing the pairwise add later. const int32x4_t add_const =1 << (8 + FILTER_BITS - 1 )));// Loading the 8 filter taps 8 ];0 ]), vmovn_s16(f[1 ]));2 ]), vmovn_s16(f[3 ]));4 ]), vmovn_s16(f[5 ]));6 ]), vmovn_s16(f[7 ]));1 ));2 ));3 ));4 ));5 ));6 ));7 ));return vreinterpretq_s16_u16(res);static AOM_FORCE_INLINE int16x8_tconst uint8x16_t in, int16x8_t f_s16) {const int32x4_t add_const = vdupq_n_s32(1 << (8 + FILTER_BITS - 1 ));0 ]);16 ]);// Permute samples ready for dot product. // { 0, 1, 2, 3, 1, 2, 3, 4, 2, 3, 4, 5, 3, 4, 5, 6 } // { 4, 5, 6, 7, 5, 6, 7, 8, 6, 7, 8, 9, 7, 8, 9, 10 } 0 );1 );0 ));return vreinterpretq_s16_u16(res);static AOM_FORCE_INLINE int16x8_t horizontal_filter_4x1_f1(const uint8x16_t in,int sx) {return horizontal_filter_4x1_f1_beta0(in, f_s16);static AOM_FORCE_INLINE int16x8_tconst uint8x16_t in, int16x8_t f_s16) {const int32x4_t add_const = vdupq_n_s32(1 << (8 + FILTER_BITS - 1 ));0 ]);16 ]);32 ]);// Permute samples ready for dot product. // { 0, 1, 2, 3, 1, 2, 3, 4, 2, 3, 4, 5, 3, 4, 5, 6 } // { 4, 5, 6, 7, 5, 6, 7, 8, 6, 7, 8, 9, 7, 8, 9, 10 } // { 8, 9, 10, 11, 9, 10, 11, 12, 10, 11, 12, 13, 11, 12, 13, 14 } 0 );1 );0 );1 );return vreinterpretq_s16_u16(res);static AOM_FORCE_INLINE int16x8_t horizontal_filter_8x1_f1(const uint8x16_t in,int sx) {return horizontal_filter_8x1_f1_beta0(in, f_s16);static AOM_FORCE_INLINE void vertical_filter_4x1_f1(const int16x8_t *src,int sy) {0 ]);1 ]);2 ]);3 ]);4 ]);5 ]);6 ]);7 ]);0 );1 );2 );3 );0 );1 );2 );3 );static AOM_FORCE_INLINE void vertical_filter_4x1_f4(const int16x8_t *src,int sy,int gamma) {0 ]), vget_low_s16(src[1 ]), vget_low_s16(src[2 ]),3 ]), vget_low_s16(src[4 ]), vget_low_s16(src[5 ]),6 ]), vget_low_s16(src[7 ]), &s0, &s1, &s2, &s3);4 ];0 ]));0 ]));1 ]));1 ]));2 ]));2 ]));3 ]));3 ]));static AOM_FORCE_INLINE void vertical_filter_8x1_f1(const int16x8_t *src,int sy) {0 ];1 ];2 ];3 ];4 ];5 ];6 ];7 ];0 );1 );2 );3 );0 );1 );2 );3 );0 );1 );2 );3 );0 );1 );2 );3 );static AOM_FORCE_INLINE void vertical_filter_8x1_f8(const int16x8_t *src,int sy,int gamma) {0 ];1 ];2 ];3 ];4 ];5 ];6 ];7 ];8 ];0 ]));0 ]));1 ]));1 ]));2 ]));2 ]));3 ]));3 ]));4 ]));4 ]));5 ]));5 ]));6 ]));6 ]));7 ]));7 ]));void av1_warp_affine_neon_i8mm(const int32_t *mat, const uint8_t *ref,int width, int height, int stride, uint8_t *pred,int p_col, int p_row, int p_width, int p_height,int p_stride, int subsampling_x,int subsampling_y, ConvolveParams *conv_params,
Messung V0.5 in Prozent C=81 H=100 G=90
¤ Dauer der Verarbeitung: 0.5 Sekunden
¤ *© Formatika GbR, Deutschland
